diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,42423 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9978900335112324,
+  "eval_steps": 504,
+  "global_step": 6042,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0004964627032394191,
+      "grad_norm": 0.48562569977411674,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 0.8809,
+      "step": 1
+    },
+    {
+      "epoch": 0.0004964627032394191,
+      "eval_loss": 0.9288526773452759,
+      "eval_runtime": 135.0527,
+      "eval_samples_per_second": 224.749,
+      "eval_steps_per_second": 28.1,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009929254064788382,
+      "grad_norm": 0.49840719985370363,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.8877,
+      "step": 2
+    },
+    {
+      "epoch": 0.0014893881097182574,
+      "grad_norm": 0.5073931939592379,
+      "learning_rate": 7.5e-07,
+      "loss": 0.912,
+      "step": 3
+    },
+    {
+      "epoch": 0.0019858508129576764,
+      "grad_norm": 0.4677170410046681,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.953,
+      "step": 4
+    },
+    {
+      "epoch": 0.002482313516197096,
+      "grad_norm": 0.48958186047439367,
+      "learning_rate": 1.25e-06,
+      "loss": 0.9888,
+      "step": 5
+    },
+    {
+      "epoch": 0.002978776219436515,
+      "grad_norm": 0.47591967932406565,
+      "learning_rate": 1.5e-06,
+      "loss": 0.9378,
+      "step": 6
+    },
+    {
+      "epoch": 0.0034752389226759338,
+      "grad_norm": 0.4775015628908847,
+      "learning_rate": 1.75e-06,
+      "loss": 0.9636,
+      "step": 7
+    },
+    {
+      "epoch": 0.003971701625915353,
+      "grad_norm": 0.4636218067902225,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.9179,
+      "step": 8
+    },
+    {
+      "epoch": 0.004468164329154772,
+      "grad_norm": 0.4627241033337522,
+      "learning_rate": 2.25e-06,
+      "loss": 0.9676,
+      "step": 9
+    },
+    {
+      "epoch": 0.004964627032394192,
+      "grad_norm": 0.4866597760611659,
+      "learning_rate": 2.5e-06,
+      "loss": 0.9696,
+      "step": 10
+    },
+    {
+      "epoch": 0.00546108973563361,
+      "grad_norm": 0.4791346551664399,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.907,
+      "step": 11
+    },
+    {
+      "epoch": 0.00595755243887303,
+      "grad_norm": 0.4769913096433881,
+      "learning_rate": 3e-06,
+      "loss": 0.8921,
+      "step": 12
+    },
+    {
+      "epoch": 0.006454015142112449,
+      "grad_norm": 0.34294431274413834,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.926,
+      "step": 13
+    },
+    {
+      "epoch": 0.0069504778453518675,
+      "grad_norm": 0.34641042017011436,
+      "learning_rate": 3.5e-06,
+      "loss": 0.9275,
+      "step": 14
+    },
+    {
+      "epoch": 0.007446940548591287,
+      "grad_norm": 0.34905473483457555,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.8703,
+      "step": 15
+    },
+    {
+      "epoch": 0.007943403251830706,
+      "grad_norm": 0.3172635186340447,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.9291,
+      "step": 16
+    },
+    {
+      "epoch": 0.008439865955070125,
+      "grad_norm": 0.2879283799693857,
+      "learning_rate": 4.25e-06,
+      "loss": 0.8477,
+      "step": 17
+    },
+    {
+      "epoch": 0.008936328658309544,
+      "grad_norm": 0.27723518517408907,
+      "learning_rate": 4.5e-06,
+      "loss": 0.8719,
+      "step": 18
+    },
+    {
+      "epoch": 0.009432791361548964,
+      "grad_norm": 0.3064644489149753,
+      "learning_rate": 4.75e-06,
+      "loss": 0.8433,
+      "step": 19
+    },
+    {
+      "epoch": 0.009929254064788383,
+      "grad_norm": 0.29412214176584833,
+      "learning_rate": 5e-06,
+      "loss": 0.859,
+      "step": 20
+    },
+    {
+      "epoch": 0.010425716768027803,
+      "grad_norm": 0.2822069566103716,
+      "learning_rate": 4.999999808957543e-06,
+      "loss": 0.8315,
+      "step": 21
+    },
+    {
+      "epoch": 0.01092217947126722,
+      "grad_norm": 0.2959170002125076,
+      "learning_rate": 4.9999992358301984e-06,
+      "loss": 0.8518,
+      "step": 22
+    },
+    {
+      "epoch": 0.01141864217450664,
+      "grad_norm": 0.2847305128418995,
+      "learning_rate": 4.9999982806180555e-06,
+      "loss": 0.9263,
+      "step": 23
+    },
+    {
+      "epoch": 0.01191510487774606,
+      "grad_norm": 0.2552021781953879,
+      "learning_rate": 4.99999694332126e-06,
+      "loss": 0.8462,
+      "step": 24
+    },
+    {
+      "epoch": 0.012411567580985479,
+      "grad_norm": 0.24750889112085914,
+      "learning_rate": 4.9999952239400165e-06,
+      "loss": 0.8324,
+      "step": 25
+    },
+    {
+      "epoch": 0.012908030284224898,
+      "grad_norm": 0.23795567702694292,
+      "learning_rate": 4.9999931224745864e-06,
+      "loss": 0.8167,
+      "step": 26
+    },
+    {
+      "epoch": 0.013404492987464317,
+      "grad_norm": 0.25300589875266577,
+      "learning_rate": 4.9999906389252926e-06,
+      "loss": 0.8979,
+      "step": 27
+    },
+    {
+      "epoch": 0.013900955690703735,
+      "grad_norm": 0.25172921500056106,
+      "learning_rate": 4.9999877732925135e-06,
+      "loss": 0.8962,
+      "step": 28
+    },
+    {
+      "epoch": 0.014397418393943155,
+      "grad_norm": 0.24565475616345825,
+      "learning_rate": 4.999984525576688e-06,
+      "loss": 0.8919,
+      "step": 29
+    },
+    {
+      "epoch": 0.014893881097182574,
+      "grad_norm": 0.23769439461231937,
+      "learning_rate": 4.999980895778312e-06,
+      "loss": 0.8506,
+      "step": 30
+    },
+    {
+      "epoch": 0.015390343800421993,
+      "grad_norm": 0.23162836109371954,
+      "learning_rate": 4.999976883897939e-06,
+      "loss": 0.9284,
+      "step": 31
+    },
+    {
+      "epoch": 0.01588680650366141,
+      "grad_norm": 0.2291685811405672,
+      "learning_rate": 4.999972489936185e-06,
+      "loss": 0.8594,
+      "step": 32
+    },
+    {
+      "epoch": 0.016383269206900832,
+      "grad_norm": 0.22080073693898697,
+      "learning_rate": 4.9999677138937185e-06,
+      "loss": 0.8606,
+      "step": 33
+    },
+    {
+      "epoch": 0.01687973191014025,
+      "grad_norm": 0.2168698203296081,
+      "learning_rate": 4.999962555771272e-06,
+      "loss": 0.8781,
+      "step": 34
+    },
+    {
+      "epoch": 0.01737619461337967,
+      "grad_norm": 0.22362106702009418,
+      "learning_rate": 4.999957015569632e-06,
+      "loss": 0.8172,
+      "step": 35
+    },
+    {
+      "epoch": 0.01787265731661909,
+      "grad_norm": 0.21848314476434746,
+      "learning_rate": 4.999951093289645e-06,
+      "loss": 0.823,
+      "step": 36
+    },
+    {
+      "epoch": 0.018369120019858506,
+      "grad_norm": 0.21837246279483064,
+      "learning_rate": 4.9999447889322195e-06,
+      "loss": 0.8576,
+      "step": 37
+    },
+    {
+      "epoch": 0.018865582723097928,
+      "grad_norm": 0.21667015902394673,
+      "learning_rate": 4.999938102498315e-06,
+      "loss": 0.8648,
+      "step": 38
+    },
+    {
+      "epoch": 0.019362045426337345,
+      "grad_norm": 0.2202153889422801,
+      "learning_rate": 4.9999310339889554e-06,
+      "loss": 0.9336,
+      "step": 39
+    },
+    {
+      "epoch": 0.019858508129576766,
+      "grad_norm": 0.22077174681701484,
+      "learning_rate": 4.9999235834052204e-06,
+      "loss": 0.8857,
+      "step": 40
+    },
+    {
+      "epoch": 0.020354970832816184,
+      "grad_norm": 0.20143789350780875,
+      "learning_rate": 4.9999157507482485e-06,
+      "loss": 0.8106,
+      "step": 41
+    },
+    {
+      "epoch": 0.020851433536055605,
+      "grad_norm": 0.21195838433230027,
+      "learning_rate": 4.999907536019238e-06,
+      "loss": 0.8221,
+      "step": 42
+    },
+    {
+      "epoch": 0.021347896239295023,
+      "grad_norm": 0.20602305261000692,
+      "learning_rate": 4.999898939219443e-06,
+      "loss": 0.8578,
+      "step": 43
+    },
+    {
+      "epoch": 0.02184435894253444,
+      "grad_norm": 0.21153710576925158,
+      "learning_rate": 4.999889960350179e-06,
+      "loss": 0.8217,
+      "step": 44
+    },
+    {
+      "epoch": 0.022340821645773862,
+      "grad_norm": 0.21344887165712584,
+      "learning_rate": 4.9998805994128166e-06,
+      "loss": 0.8838,
+      "step": 45
+    },
+    {
+      "epoch": 0.02283728434901328,
+      "grad_norm": 0.21002096278796756,
+      "learning_rate": 4.999870856408787e-06,
+      "loss": 0.8782,
+      "step": 46
+    },
+    {
+      "epoch": 0.0233337470522527,
+      "grad_norm": 0.20886971438867197,
+      "learning_rate": 4.999860731339579e-06,
+      "loss": 0.8295,
+      "step": 47
+    },
+    {
+      "epoch": 0.02383020975549212,
+      "grad_norm": 0.22095165664633118,
+      "learning_rate": 4.999850224206741e-06,
+      "loss": 0.8687,
+      "step": 48
+    },
+    {
+      "epoch": 0.02432667245873154,
+      "grad_norm": 0.21253844707230096,
+      "learning_rate": 4.999839335011878e-06,
+      "loss": 0.8782,
+      "step": 49
+    },
+    {
+      "epoch": 0.024823135161970957,
+      "grad_norm": 0.21604040244789194,
+      "learning_rate": 4.999828063756655e-06,
+      "loss": 0.8385,
+      "step": 50
+    },
+    {
+      "epoch": 0.025319597865210375,
+      "grad_norm": 0.21102395354985817,
+      "learning_rate": 4.999816410442794e-06,
+      "loss": 0.8903,
+      "step": 51
+    },
+    {
+      "epoch": 0.025816060568449796,
+      "grad_norm": 0.2062210328133959,
+      "learning_rate": 4.999804375072076e-06,
+      "loss": 0.8717,
+      "step": 52
+    },
+    {
+      "epoch": 0.026312523271689214,
+      "grad_norm": 0.22296430893025995,
+      "learning_rate": 4.999791957646341e-06,
+      "loss": 0.8369,
+      "step": 53
+    },
+    {
+      "epoch": 0.026808985974928635,
+      "grad_norm": 0.2172722843835763,
+      "learning_rate": 4.9997791581674855e-06,
+      "loss": 0.9088,
+      "step": 54
+    },
+    {
+      "epoch": 0.027305448678168052,
+      "grad_norm": 0.2097980373922686,
+      "learning_rate": 4.999765976637467e-06,
+      "loss": 0.8559,
+      "step": 55
+    },
+    {
+      "epoch": 0.02780191138140747,
+      "grad_norm": 0.20153765827962083,
+      "learning_rate": 4.9997524130583e-06,
+      "loss": 0.8233,
+      "step": 56
+    },
+    {
+      "epoch": 0.02829837408464689,
+      "grad_norm": 0.20769665963331085,
+      "learning_rate": 4.999738467432057e-06,
+      "loss": 0.8321,
+      "step": 57
+    },
+    {
+      "epoch": 0.02879483678788631,
+      "grad_norm": 0.21216448449114564,
+      "learning_rate": 4.999724139760869e-06,
+      "loss": 0.847,
+      "step": 58
+    },
+    {
+      "epoch": 0.02929129949112573,
+      "grad_norm": 0.2063639919391254,
+      "learning_rate": 4.999709430046926e-06,
+      "loss": 0.8064,
+      "step": 59
+    },
+    {
+      "epoch": 0.029787762194365148,
+      "grad_norm": 0.21452334497936051,
+      "learning_rate": 4.999694338292478e-06,
+      "loss": 0.8838,
+      "step": 60
+    },
+    {
+      "epoch": 0.03028422489760457,
+      "grad_norm": 0.20061672335186068,
+      "learning_rate": 4.999678864499828e-06,
+      "loss": 0.7818,
+      "step": 61
+    },
+    {
+      "epoch": 0.030780687600843987,
+      "grad_norm": 0.2046523532499392,
+      "learning_rate": 4.999663008671344e-06,
+      "loss": 0.8581,
+      "step": 62
+    },
+    {
+      "epoch": 0.03127715030408341,
+      "grad_norm": 0.20186628257277983,
+      "learning_rate": 4.999646770809449e-06,
+      "loss": 0.8445,
+      "step": 63
+    },
+    {
+      "epoch": 0.03177361300732282,
+      "grad_norm": 0.2012482614306437,
+      "learning_rate": 4.9996301509166225e-06,
+      "loss": 0.8138,
+      "step": 64
+    },
+    {
+      "epoch": 0.03227007571056224,
+      "grad_norm": 0.20616591148685806,
+      "learning_rate": 4.999613148995406e-06,
+      "loss": 0.854,
+      "step": 65
+    },
+    {
+      "epoch": 0.032766538413801664,
+      "grad_norm": 0.20274972415160764,
+      "learning_rate": 4.999595765048399e-06,
+      "loss": 0.8223,
+      "step": 66
+    },
+    {
+      "epoch": 0.033263001117041086,
+      "grad_norm": 0.20826308052556233,
+      "learning_rate": 4.9995779990782556e-06,
+      "loss": 0.8669,
+      "step": 67
+    },
+    {
+      "epoch": 0.0337594638202805,
+      "grad_norm": 0.20529737714316182,
+      "learning_rate": 4.999559851087694e-06,
+      "loss": 0.856,
+      "step": 68
+    },
+    {
+      "epoch": 0.03425592652351992,
+      "grad_norm": 0.2098299357116322,
+      "learning_rate": 4.999541321079486e-06,
+      "loss": 0.895,
+      "step": 69
+    },
+    {
+      "epoch": 0.03475238922675934,
+      "grad_norm": 0.21085272972883423,
+      "learning_rate": 4.9995224090564645e-06,
+      "loss": 0.8135,
+      "step": 70
+    },
+    {
+      "epoch": 0.035248851929998756,
+      "grad_norm": 0.20815593231474888,
+      "learning_rate": 4.9995031150215194e-06,
+      "loss": 0.8325,
+      "step": 71
+    },
+    {
+      "epoch": 0.03574531463323818,
+      "grad_norm": 0.2030397777365442,
+      "learning_rate": 4.9994834389776e-06,
+      "loss": 0.8028,
+      "step": 72
+    },
+    {
+      "epoch": 0.0362417773364776,
+      "grad_norm": 0.2093993861940969,
+      "learning_rate": 4.999463380927713e-06,
+      "loss": 0.8222,
+      "step": 73
+    },
+    {
+      "epoch": 0.03673824003971701,
+      "grad_norm": 0.20310206636138728,
+      "learning_rate": 4.9994429408749235e-06,
+      "loss": 0.8987,
+      "step": 74
+    },
+    {
+      "epoch": 0.037234702742956434,
+      "grad_norm": 0.19635518255581516,
+      "learning_rate": 4.999422118822357e-06,
+      "loss": 0.8257,
+      "step": 75
+    },
+    {
+      "epoch": 0.037731165446195855,
+      "grad_norm": 0.20613686945426812,
+      "learning_rate": 4.999400914773193e-06,
+      "loss": 0.8519,
+      "step": 76
+    },
+    {
+      "epoch": 0.038227628149435276,
+      "grad_norm": 0.2148909458983486,
+      "learning_rate": 4.999379328730676e-06,
+      "loss": 0.8431,
+      "step": 77
+    },
+    {
+      "epoch": 0.03872409085267469,
+      "grad_norm": 0.20131996731845622,
+      "learning_rate": 4.999357360698103e-06,
+      "loss": 0.8298,
+      "step": 78
+    },
+    {
+      "epoch": 0.03922055355591411,
+      "grad_norm": 0.19799106731473673,
+      "learning_rate": 4.999335010678831e-06,
+      "loss": 0.8438,
+      "step": 79
+    },
+    {
+      "epoch": 0.03971701625915353,
+      "grad_norm": 0.20461179643215718,
+      "learning_rate": 4.999312278676276e-06,
+      "loss": 0.8501,
+      "step": 80
+    },
+    {
+      "epoch": 0.04021347896239295,
+      "grad_norm": 0.19950036085187078,
+      "learning_rate": 4.999289164693913e-06,
+      "loss": 0.802,
+      "step": 81
+    },
+    {
+      "epoch": 0.04070994166563237,
+      "grad_norm": 0.20647834483552271,
+      "learning_rate": 4.999265668735274e-06,
+      "loss": 0.8004,
+      "step": 82
+    },
+    {
+      "epoch": 0.04120640436887179,
+      "grad_norm": 0.20535506020956135,
+      "learning_rate": 4.99924179080395e-06,
+      "loss": 0.8398,
+      "step": 83
+    },
+    {
+      "epoch": 0.04170286707211121,
+      "grad_norm": 0.20379519814931177,
+      "learning_rate": 4.999217530903592e-06,
+      "loss": 0.8493,
+      "step": 84
+    },
+    {
+      "epoch": 0.042199329775350625,
+      "grad_norm": 0.19569632518773314,
+      "learning_rate": 4.999192889037905e-06,
+      "loss": 0.7781,
+      "step": 85
+    },
+    {
+      "epoch": 0.042695792478590046,
+      "grad_norm": 0.2022603888395999,
+      "learning_rate": 4.999167865210656e-06,
+      "loss": 0.8202,
+      "step": 86
+    },
+    {
+      "epoch": 0.04319225518182947,
+      "grad_norm": 0.2061084529311429,
+      "learning_rate": 4.999142459425671e-06,
+      "loss": 0.8112,
+      "step": 87
+    },
+    {
+      "epoch": 0.04368871788506888,
+      "grad_norm": 0.20386460325471262,
+      "learning_rate": 4.999116671686832e-06,
+      "loss": 0.8314,
+      "step": 88
+    },
+    {
+      "epoch": 0.0441851805883083,
+      "grad_norm": 0.2001689787137877,
+      "learning_rate": 4.9990905019980795e-06,
+      "loss": 0.8493,
+      "step": 89
+    },
+    {
+      "epoch": 0.044681643291547724,
+      "grad_norm": 0.19704962097184103,
+      "learning_rate": 4.999063950363413e-06,
+      "loss": 0.8125,
+      "step": 90
+    },
+    {
+      "epoch": 0.045178105994787145,
+      "grad_norm": 0.20288945888095392,
+      "learning_rate": 4.999037016786891e-06,
+      "loss": 0.8529,
+      "step": 91
+    },
+    {
+      "epoch": 0.04567456869802656,
+      "grad_norm": 0.20036286496607184,
+      "learning_rate": 4.999009701272632e-06,
+      "loss": 0.818,
+      "step": 92
+    },
+    {
+      "epoch": 0.04617103140126598,
+      "grad_norm": 0.19540166087377867,
+      "learning_rate": 4.998982003824807e-06,
+      "loss": 0.8044,
+      "step": 93
+    },
+    {
+      "epoch": 0.0466674941045054,
+      "grad_norm": 0.19654628372196933,
+      "learning_rate": 4.998953924447652e-06,
+      "loss": 0.8276,
+      "step": 94
+    },
+    {
+      "epoch": 0.047163956807744815,
+      "grad_norm": 0.20737336312858767,
+      "learning_rate": 4.998925463145456e-06,
+      "loss": 0.792,
+      "step": 95
+    },
+    {
+      "epoch": 0.04766041951098424,
+      "grad_norm": 0.2059319669770031,
+      "learning_rate": 4.998896619922571e-06,
+      "loss": 0.8635,
+      "step": 96
+    },
+    {
+      "epoch": 0.04815688221422366,
+      "grad_norm": 0.20528654692184178,
+      "learning_rate": 4.9988673947834045e-06,
+      "loss": 0.8605,
+      "step": 97
+    },
+    {
+      "epoch": 0.04865334491746308,
+      "grad_norm": 0.20472541684877216,
+      "learning_rate": 4.998837787732422e-06,
+      "loss": 0.87,
+      "step": 98
+    },
+    {
+      "epoch": 0.04914980762070249,
+      "grad_norm": 0.20212979201560113,
+      "learning_rate": 4.998807798774151e-06,
+      "loss": 0.8236,
+      "step": 99
+    },
+    {
+      "epoch": 0.049646270323941914,
+      "grad_norm": 0.19931273766196522,
+      "learning_rate": 4.998777427913172e-06,
+      "loss": 0.8541,
+      "step": 100
+    },
+    {
+      "epoch": 0.050142733027181335,
+      "grad_norm": 0.2060934643730322,
+      "learning_rate": 4.998746675154129e-06,
+      "loss": 0.8765,
+      "step": 101
+    },
+    {
+      "epoch": 0.05063919573042075,
+      "grad_norm": 0.19905720321276965,
+      "learning_rate": 4.99871554050172e-06,
+      "loss": 0.8193,
+      "step": 102
+    },
+    {
+      "epoch": 0.05113565843366017,
+      "grad_norm": 0.19654055207721635,
+      "learning_rate": 4.998684023960705e-06,
+      "loss": 0.804,
+      "step": 103
+    },
+    {
+      "epoch": 0.05163212113689959,
+      "grad_norm": 0.1982198351814607,
+      "learning_rate": 4.998652125535901e-06,
+      "loss": 0.7891,
+      "step": 104
+    },
+    {
+      "epoch": 0.052128583840139006,
+      "grad_norm": 0.39942501787277024,
+      "learning_rate": 4.998619845232181e-06,
+      "loss": 0.8563,
+      "step": 105
+    },
+    {
+      "epoch": 0.05262504654337843,
+      "grad_norm": 0.1955611794394933,
+      "learning_rate": 4.998587183054481e-06,
+      "loss": 0.8287,
+      "step": 106
+    },
+    {
+      "epoch": 0.05312150924661785,
+      "grad_norm": 0.20389389297242477,
+      "learning_rate": 4.9985541390077915e-06,
+      "loss": 0.8284,
+      "step": 107
+    },
+    {
+      "epoch": 0.05361797194985727,
+      "grad_norm": 0.21045434439040575,
+      "learning_rate": 4.998520713097164e-06,
+      "loss": 0.8608,
+      "step": 108
+    },
+    {
+      "epoch": 0.054114434653096684,
+      "grad_norm": 0.20116892551504365,
+      "learning_rate": 4.998486905327704e-06,
+      "loss": 0.8359,
+      "step": 109
+    },
+    {
+      "epoch": 0.054610897356336105,
+      "grad_norm": 0.21230325768339284,
+      "learning_rate": 4.9984527157045825e-06,
+      "loss": 0.8218,
+      "step": 110
+    },
+    {
+      "epoch": 0.055107360059575526,
+      "grad_norm": 0.19875350435158548,
+      "learning_rate": 4.998418144233023e-06,
+      "loss": 0.8115,
+      "step": 111
+    },
+    {
+      "epoch": 0.05560382276281494,
+      "grad_norm": 0.19722218299068556,
+      "learning_rate": 4.998383190918309e-06,
+      "loss": 0.8409,
+      "step": 112
+    },
+    {
+      "epoch": 0.05610028546605436,
+      "grad_norm": 0.19442312486929395,
+      "learning_rate": 4.998347855765783e-06,
+      "loss": 0.7861,
+      "step": 113
+    },
+    {
+      "epoch": 0.05659674816929378,
+      "grad_norm": 0.19986199846633768,
+      "learning_rate": 4.998312138780845e-06,
+      "loss": 0.8412,
+      "step": 114
+    },
+    {
+      "epoch": 0.057093210872533204,
+      "grad_norm": 0.19565952095909392,
+      "learning_rate": 4.998276039968953e-06,
+      "loss": 0.8415,
+      "step": 115
+    },
+    {
+      "epoch": 0.05758967357577262,
+      "grad_norm": 0.19154584275348677,
+      "learning_rate": 4.998239559335627e-06,
+      "loss": 0.8631,
+      "step": 116
+    },
+    {
+      "epoch": 0.05808613627901204,
+      "grad_norm": 0.19665878881747706,
+      "learning_rate": 4.99820269688644e-06,
+      "loss": 0.8598,
+      "step": 117
+    },
+    {
+      "epoch": 0.05858259898225146,
+      "grad_norm": 0.2028047172295385,
+      "learning_rate": 4.998165452627025e-06,
+      "loss": 0.8683,
+      "step": 118
+    },
+    {
+      "epoch": 0.059079061685490875,
+      "grad_norm": 0.2037639166726157,
+      "learning_rate": 4.998127826563077e-06,
+      "loss": 0.8112,
+      "step": 119
+    },
+    {
+      "epoch": 0.059575524388730296,
+      "grad_norm": 0.1959085397655898,
+      "learning_rate": 4.998089818700344e-06,
+      "loss": 0.8463,
+      "step": 120
+    },
+    {
+      "epoch": 0.06007198709196972,
+      "grad_norm": 0.19554313432021866,
+      "learning_rate": 4.998051429044638e-06,
+      "loss": 0.814,
+      "step": 121
+    },
+    {
+      "epoch": 0.06056844979520914,
+      "grad_norm": 0.21327067999617655,
+      "learning_rate": 4.998012657601823e-06,
+      "loss": 0.8369,
+      "step": 122
+    },
+    {
+      "epoch": 0.06106491249844855,
+      "grad_norm": 0.19482066937524098,
+      "learning_rate": 4.997973504377826e-06,
+      "loss": 0.8503,
+      "step": 123
+    },
+    {
+      "epoch": 0.06156137520168797,
+      "grad_norm": 0.1921263735287261,
+      "learning_rate": 4.99793396937863e-06,
+      "loss": 0.8235,
+      "step": 124
+    },
+    {
+      "epoch": 0.062057837904927395,
+      "grad_norm": 0.1912317820741509,
+      "learning_rate": 4.99789405261028e-06,
+      "loss": 0.7959,
+      "step": 125
+    },
+    {
+      "epoch": 0.06255430060816682,
+      "grad_norm": 0.1930012329719747,
+      "learning_rate": 4.997853754078873e-06,
+      "loss": 0.8312,
+      "step": 126
+    },
+    {
+      "epoch": 0.06305076331140623,
+      "grad_norm": 0.1854436986074747,
+      "learning_rate": 4.997813073790571e-06,
+      "loss": 0.7915,
+      "step": 127
+    },
+    {
+      "epoch": 0.06354722601464564,
+      "grad_norm": 0.19266710662549197,
+      "learning_rate": 4.997772011751589e-06,
+      "loss": 0.8953,
+      "step": 128
+    },
+    {
+      "epoch": 0.06404368871788507,
+      "grad_norm": 0.18944442325778477,
+      "learning_rate": 4.9977305679682044e-06,
+      "loss": 0.7999,
+      "step": 129
+    },
+    {
+      "epoch": 0.06454015142112449,
+      "grad_norm": 0.19586350088433074,
+      "learning_rate": 4.99768874244675e-06,
+      "loss": 0.8091,
+      "step": 130
+    },
+    {
+      "epoch": 0.0650366141243639,
+      "grad_norm": 0.19084476465979933,
+      "learning_rate": 4.997646535193618e-06,
+      "loss": 0.8733,
+      "step": 131
+    },
+    {
+      "epoch": 0.06553307682760333,
+      "grad_norm": 0.1884614630341233,
+      "learning_rate": 4.997603946215262e-06,
+      "loss": 0.8106,
+      "step": 132
+    },
+    {
+      "epoch": 0.06602953953084274,
+      "grad_norm": 0.1932654239334205,
+      "learning_rate": 4.9975609755181875e-06,
+      "loss": 0.9149,
+      "step": 133
+    },
+    {
+      "epoch": 0.06652600223408217,
+      "grad_norm": 0.1872082345438042,
+      "learning_rate": 4.997517623108964e-06,
+      "loss": 0.7794,
+      "step": 134
+    },
+    {
+      "epoch": 0.06702246493732159,
+      "grad_norm": 0.1873912754282765,
+      "learning_rate": 4.997473888994215e-06,
+      "loss": 0.8577,
+      "step": 135
+    },
+    {
+      "epoch": 0.067518927640561,
+      "grad_norm": 0.18175316620522922,
+      "learning_rate": 4.997429773180627e-06,
+      "loss": 0.8226,
+      "step": 136
+    },
+    {
+      "epoch": 0.06801539034380043,
+      "grad_norm": 0.17684485991325427,
+      "learning_rate": 4.997385275674942e-06,
+      "loss": 0.7802,
+      "step": 137
+    },
+    {
+      "epoch": 0.06851185304703984,
+      "grad_norm": 0.1824706517270434,
+      "learning_rate": 4.99734039648396e-06,
+      "loss": 0.7985,
+      "step": 138
+    },
+    {
+      "epoch": 0.06900831575027926,
+      "grad_norm": 0.1799259476989057,
+      "learning_rate": 4.997295135614539e-06,
+      "loss": 0.8036,
+      "step": 139
+    },
+    {
+      "epoch": 0.06950477845351868,
+      "grad_norm": 0.17823278052011338,
+      "learning_rate": 4.997249493073598e-06,
+      "loss": 0.8167,
+      "step": 140
+    },
+    {
+      "epoch": 0.0700012411567581,
+      "grad_norm": 0.18252365188173084,
+      "learning_rate": 4.997203468868113e-06,
+      "loss": 0.8295,
+      "step": 141
+    },
+    {
+      "epoch": 0.07049770385999751,
+      "grad_norm": 0.19736868569754917,
+      "learning_rate": 4.997157063005117e-06,
+      "loss": 0.8908,
+      "step": 142
+    },
+    {
+      "epoch": 0.07099416656323694,
+      "grad_norm": 0.17947955446070774,
+      "learning_rate": 4.997110275491702e-06,
+      "loss": 0.8115,
+      "step": 143
+    },
+    {
+      "epoch": 0.07149062926647635,
+      "grad_norm": 0.19402407753710812,
+      "learning_rate": 4.997063106335021e-06,
+      "loss": 0.8454,
+      "step": 144
+    },
+    {
+      "epoch": 0.07198709196971577,
+      "grad_norm": 0.1887305860550766,
+      "learning_rate": 4.99701555554228e-06,
+      "loss": 0.8945,
+      "step": 145
+    },
+    {
+      "epoch": 0.0724835546729552,
+      "grad_norm": 0.17558377313650073,
+      "learning_rate": 4.99696762312075e-06,
+      "loss": 0.8009,
+      "step": 146
+    },
+    {
+      "epoch": 0.07298001737619461,
+      "grad_norm": 0.1718469916803291,
+      "learning_rate": 4.9969193090777526e-06,
+      "loss": 0.8143,
+      "step": 147
+    },
+    {
+      "epoch": 0.07347648007943403,
+      "grad_norm": 0.17617757928734015,
+      "learning_rate": 4.996870613420675e-06,
+      "loss": 0.8217,
+      "step": 148
+    },
+    {
+      "epoch": 0.07397294278267345,
+      "grad_norm": 0.1706971696924589,
+      "learning_rate": 4.996821536156958e-06,
+      "loss": 0.7952,
+      "step": 149
+    },
+    {
+      "epoch": 0.07446940548591287,
+      "grad_norm": 0.18103564748682863,
+      "learning_rate": 4.996772077294103e-06,
+      "loss": 0.8214,
+      "step": 150
+    },
+    {
+      "epoch": 0.0749658681891523,
+      "grad_norm": 0.18622213669793447,
+      "learning_rate": 4.9967222368396686e-06,
+      "loss": 0.8843,
+      "step": 151
+    },
+    {
+      "epoch": 0.07546233089239171,
+      "grad_norm": 0.1707461817930418,
+      "learning_rate": 4.9966720148012714e-06,
+      "loss": 0.7733,
+      "step": 152
+    },
+    {
+      "epoch": 0.07595879359563112,
+      "grad_norm": 0.15640519366869854,
+      "learning_rate": 4.996621411186589e-06,
+      "loss": 0.7887,
+      "step": 153
+    },
+    {
+      "epoch": 0.07645525629887055,
+      "grad_norm": 0.1624602295916605,
+      "learning_rate": 4.996570426003354e-06,
+      "loss": 0.8471,
+      "step": 154
+    },
+    {
+      "epoch": 0.07695171900210997,
+      "grad_norm": 0.18290118651942194,
+      "learning_rate": 4.996519059259358e-06,
+      "loss": 0.8467,
+      "step": 155
+    },
+    {
+      "epoch": 0.07744818170534938,
+      "grad_norm": 0.15912757799549243,
+      "learning_rate": 4.996467310962453e-06,
+      "loss": 0.8005,
+      "step": 156
+    },
+    {
+      "epoch": 0.07794464440858881,
+      "grad_norm": 0.1613389987950997,
+      "learning_rate": 4.996415181120547e-06,
+      "loss": 0.7892,
+      "step": 157
+    },
+    {
+      "epoch": 0.07844110711182822,
+      "grad_norm": 0.1570677524283997,
+      "learning_rate": 4.996362669741609e-06,
+      "loss": 0.8286,
+      "step": 158
+    },
+    {
+      "epoch": 0.07893756981506764,
+      "grad_norm": 0.16537699681870446,
+      "learning_rate": 4.996309776833661e-06,
+      "loss": 0.8273,
+      "step": 159
+    },
+    {
+      "epoch": 0.07943403251830707,
+      "grad_norm": 0.17043045232874193,
+      "learning_rate": 4.99625650240479e-06,
+      "loss": 0.7966,
+      "step": 160
+    },
+    {
+      "epoch": 0.07993049522154648,
+      "grad_norm": 0.15656286097485822,
+      "learning_rate": 4.9962028464631365e-06,
+      "loss": 0.8332,
+      "step": 161
+    },
+    {
+      "epoch": 0.0804269579247859,
+      "grad_norm": 0.1776462779512006,
+      "learning_rate": 4.9961488090169015e-06,
+      "loss": 0.819,
+      "step": 162
+    },
+    {
+      "epoch": 0.08092342062802532,
+      "grad_norm": 0.15898779205316405,
+      "learning_rate": 4.996094390074345e-06,
+      "loss": 0.8326,
+      "step": 163
+    },
+    {
+      "epoch": 0.08141988333126474,
+      "grad_norm": 0.16851878425262465,
+      "learning_rate": 4.996039589643782e-06,
+      "loss": 0.8636,
+      "step": 164
+    },
+    {
+      "epoch": 0.08191634603450416,
+      "grad_norm": 0.14941991731340487,
+      "learning_rate": 4.995984407733588e-06,
+      "loss": 0.7804,
+      "step": 165
+    },
+    {
+      "epoch": 0.08241280873774358,
+      "grad_norm": 0.16021285793408913,
+      "learning_rate": 4.995928844352198e-06,
+      "loss": 0.8614,
+      "step": 166
+    },
+    {
+      "epoch": 0.08290927144098299,
+      "grad_norm": 0.1592609781313818,
+      "learning_rate": 4.995872899508103e-06,
+      "loss": 0.8355,
+      "step": 167
+    },
+    {
+      "epoch": 0.08340573414422242,
+      "grad_norm": 0.14893046373707292,
+      "learning_rate": 4.995816573209854e-06,
+      "loss": 0.8146,
+      "step": 168
+    },
+    {
+      "epoch": 0.08390219684746184,
+      "grad_norm": 0.14994377964296016,
+      "learning_rate": 4.995759865466059e-06,
+      "loss": 0.8073,
+      "step": 169
+    },
+    {
+      "epoch": 0.08439865955070125,
+      "grad_norm": 0.15057311866648648,
+      "learning_rate": 4.995702776285385e-06,
+      "loss": 0.8195,
+      "step": 170
+    },
+    {
+      "epoch": 0.08489512225394068,
+      "grad_norm": 0.1462021842716386,
+      "learning_rate": 4.995645305676558e-06,
+      "loss": 0.8054,
+      "step": 171
+    },
+    {
+      "epoch": 0.08539158495718009,
+      "grad_norm": 0.15612484218678557,
+      "learning_rate": 4.995587453648359e-06,
+      "loss": 0.8538,
+      "step": 172
+    },
+    {
+      "epoch": 0.0858880476604195,
+      "grad_norm": 0.15826299784330564,
+      "learning_rate": 4.995529220209633e-06,
+      "loss": 0.7967,
+      "step": 173
+    },
+    {
+      "epoch": 0.08638451036365893,
+      "grad_norm": 0.14934453023970579,
+      "learning_rate": 4.9954706053692766e-06,
+      "loss": 0.8191,
+      "step": 174
+    },
+    {
+      "epoch": 0.08688097306689835,
+      "grad_norm": 0.14487398595953221,
+      "learning_rate": 4.995411609136252e-06,
+      "loss": 0.7913,
+      "step": 175
+    },
+    {
+      "epoch": 0.08737743577013776,
+      "grad_norm": 0.2361253677063094,
+      "learning_rate": 4.995352231519572e-06,
+      "loss": 0.8506,
+      "step": 176
+    },
+    {
+      "epoch": 0.08787389847337719,
+      "grad_norm": 0.1542240974761206,
+      "learning_rate": 4.995292472528315e-06,
+      "loss": 0.8198,
+      "step": 177
+    },
+    {
+      "epoch": 0.0883703611766166,
+      "grad_norm": 0.1533907925673139,
+      "learning_rate": 4.9952323321716114e-06,
+      "loss": 0.8095,
+      "step": 178
+    },
+    {
+      "epoch": 0.08886682387985602,
+      "grad_norm": 0.14883299168355707,
+      "learning_rate": 4.995171810458654e-06,
+      "loss": 0.8492,
+      "step": 179
+    },
+    {
+      "epoch": 0.08936328658309545,
+      "grad_norm": 0.14717378708658913,
+      "learning_rate": 4.995110907398693e-06,
+      "loss": 0.7989,
+      "step": 180
+    },
+    {
+      "epoch": 0.08985974928633486,
+      "grad_norm": 0.1432697352230233,
+      "learning_rate": 4.995049623001036e-06,
+      "loss": 0.8112,
+      "step": 181
+    },
+    {
+      "epoch": 0.09035621198957429,
+      "grad_norm": 0.14420342048963478,
+      "learning_rate": 4.994987957275048e-06,
+      "loss": 0.8113,
+      "step": 182
+    },
+    {
+      "epoch": 0.0908526746928137,
+      "grad_norm": 0.1484614971553416,
+      "learning_rate": 4.994925910230156e-06,
+      "loss": 0.7928,
+      "step": 183
+    },
+    {
+      "epoch": 0.09134913739605312,
+      "grad_norm": 0.14303299403206174,
+      "learning_rate": 4.994863481875842e-06,
+      "loss": 0.821,
+      "step": 184
+    },
+    {
+      "epoch": 0.09184560009929255,
+      "grad_norm": 0.14990402029979874,
+      "learning_rate": 4.9948006722216456e-06,
+      "loss": 0.7919,
+      "step": 185
+    },
+    {
+      "epoch": 0.09234206280253196,
+      "grad_norm": 0.15676689121485238,
+      "learning_rate": 4.9947374812771675e-06,
+      "loss": 0.8475,
+      "step": 186
+    },
+    {
+      "epoch": 0.09283852550577137,
+      "grad_norm": 0.14836071218925942,
+      "learning_rate": 4.994673909052067e-06,
+      "loss": 0.81,
+      "step": 187
+    },
+    {
+      "epoch": 0.0933349882090108,
+      "grad_norm": 0.14051092402883178,
+      "learning_rate": 4.994609955556057e-06,
+      "loss": 0.7896,
+      "step": 188
+    },
+    {
+      "epoch": 0.09383145091225022,
+      "grad_norm": 0.13964829094096,
+      "learning_rate": 4.994545620798914e-06,
+      "loss": 0.8095,
+      "step": 189
+    },
+    {
+      "epoch": 0.09432791361548963,
+      "grad_norm": 0.13916395003469767,
+      "learning_rate": 4.994480904790469e-06,
+      "loss": 0.7818,
+      "step": 190
+    },
+    {
+      "epoch": 0.09482437631872906,
+      "grad_norm": 0.13992841292678654,
+      "learning_rate": 4.994415807540616e-06,
+      "loss": 0.8511,
+      "step": 191
+    },
+    {
+      "epoch": 0.09532083902196847,
+      "grad_norm": 0.1382517091115567,
+      "learning_rate": 4.9943503290593e-06,
+      "loss": 0.804,
+      "step": 192
+    },
+    {
+      "epoch": 0.09581730172520789,
+      "grad_norm": 0.13828160588381802,
+      "learning_rate": 4.99428446935653e-06,
+      "loss": 0.7945,
+      "step": 193
+    },
+    {
+      "epoch": 0.09631376442844732,
+      "grad_norm": 0.14059132138612335,
+      "learning_rate": 4.9942182284423715e-06,
+      "loss": 0.8127,
+      "step": 194
+    },
+    {
+      "epoch": 0.09681022713168673,
+      "grad_norm": 0.14342500046524775,
+      "learning_rate": 4.994151606326949e-06,
+      "loss": 0.7869,
+      "step": 195
+    },
+    {
+      "epoch": 0.09730668983492616,
+      "grad_norm": 0.15328257822811497,
+      "learning_rate": 4.994084603020444e-06,
+      "loss": 0.7969,
+      "step": 196
+    },
+    {
+      "epoch": 0.09780315253816557,
+      "grad_norm": 0.15036264302541547,
+      "learning_rate": 4.9940172185330975e-06,
+      "loss": 0.8021,
+      "step": 197
+    },
+    {
+      "epoch": 0.09829961524140499,
+      "grad_norm": 0.13631880610103,
+      "learning_rate": 4.993949452875208e-06,
+      "loss": 0.8062,
+      "step": 198
+    },
+    {
+      "epoch": 0.09879607794464441,
+      "grad_norm": 0.1370149045617051,
+      "learning_rate": 4.993881306057131e-06,
+      "loss": 0.7964,
+      "step": 199
+    },
+    {
+      "epoch": 0.09929254064788383,
+      "grad_norm": 0.14003683139998516,
+      "learning_rate": 4.993812778089283e-06,
+      "loss": 0.7829,
+      "step": 200
+    },
+    {
+      "epoch": 0.09978900335112324,
+      "grad_norm": 0.13364321729142267,
+      "learning_rate": 4.993743868982137e-06,
+      "loss": 0.7435,
+      "step": 201
+    },
+    {
+      "epoch": 0.10028546605436267,
+      "grad_norm": 0.1432644408384139,
+      "learning_rate": 4.993674578746225e-06,
+      "loss": 0.8404,
+      "step": 202
+    },
+    {
+      "epoch": 0.10078192875760209,
+      "grad_norm": 0.14022940831071956,
+      "learning_rate": 4.9936049073921365e-06,
+      "loss": 0.7916,
+      "step": 203
+    },
+    {
+      "epoch": 0.1012783914608415,
+      "grad_norm": 0.1431480895121353,
+      "learning_rate": 4.99353485493052e-06,
+      "loss": 0.7825,
+      "step": 204
+    },
+    {
+      "epoch": 0.10177485416408093,
+      "grad_norm": 0.1537315485269007,
+      "learning_rate": 4.993464421372081e-06,
+      "loss": 0.8018,
+      "step": 205
+    },
+    {
+      "epoch": 0.10227131686732034,
+      "grad_norm": 0.13149782403488902,
+      "learning_rate": 4.993393606727587e-06,
+      "loss": 0.748,
+      "step": 206
+    },
+    {
+      "epoch": 0.10276777957055976,
+      "grad_norm": 0.13795599982306228,
+      "learning_rate": 4.993322411007857e-06,
+      "loss": 0.7745,
+      "step": 207
+    },
+    {
+      "epoch": 0.10326424227379918,
+      "grad_norm": 0.14263948789480718,
+      "learning_rate": 4.993250834223774e-06,
+      "loss": 0.8286,
+      "step": 208
+    },
+    {
+      "epoch": 0.1037607049770386,
+      "grad_norm": 0.1322915151191568,
+      "learning_rate": 4.9931788763862774e-06,
+      "loss": 0.7765,
+      "step": 209
+    },
+    {
+      "epoch": 0.10425716768027801,
+      "grad_norm": 0.1411982251684298,
+      "learning_rate": 4.993106537506365e-06,
+      "loss": 0.7978,
+      "step": 210
+    },
+    {
+      "epoch": 0.10475363038351744,
+      "grad_norm": 0.14287120534895575,
+      "learning_rate": 4.993033817595092e-06,
+      "loss": 0.8119,
+      "step": 211
+    },
+    {
+      "epoch": 0.10525009308675685,
+      "grad_norm": 0.1457313365849308,
+      "learning_rate": 4.992960716663572e-06,
+      "loss": 0.8391,
+      "step": 212
+    },
+    {
+      "epoch": 0.10574655578999628,
+      "grad_norm": 0.1374862675337484,
+      "learning_rate": 4.992887234722978e-06,
+      "loss": 0.8284,
+      "step": 213
+    },
+    {
+      "epoch": 0.1062430184932357,
+      "grad_norm": 0.13532768487001728,
+      "learning_rate": 4.992813371784542e-06,
+      "loss": 0.7933,
+      "step": 214
+    },
+    {
+      "epoch": 0.10673948119647511,
+      "grad_norm": 0.14619055294203817,
+      "learning_rate": 4.99273912785955e-06,
+      "loss": 0.8232,
+      "step": 215
+    },
+    {
+      "epoch": 0.10723594389971454,
+      "grad_norm": 0.13334770368704213,
+      "learning_rate": 4.992664502959351e-06,
+      "loss": 0.7987,
+      "step": 216
+    },
+    {
+      "epoch": 0.10773240660295395,
+      "grad_norm": 0.1386527129272213,
+      "learning_rate": 4.99258949709535e-06,
+      "loss": 0.8385,
+      "step": 217
+    },
+    {
+      "epoch": 0.10822886930619337,
+      "grad_norm": 0.13307297628479753,
+      "learning_rate": 4.99251411027901e-06,
+      "loss": 0.7945,
+      "step": 218
+    },
+    {
+      "epoch": 0.1087253320094328,
+      "grad_norm": 0.1405179423764166,
+      "learning_rate": 4.992438342521851e-06,
+      "loss": 0.8591,
+      "step": 219
+    },
+    {
+      "epoch": 0.10922179471267221,
+      "grad_norm": 0.14156609276835416,
+      "learning_rate": 4.992362193835456e-06,
+      "loss": 0.8333,
+      "step": 220
+    },
+    {
+      "epoch": 0.10971825741591162,
+      "grad_norm": 0.135514778316713,
+      "learning_rate": 4.992285664231462e-06,
+      "loss": 0.7784,
+      "step": 221
+    },
+    {
+      "epoch": 0.11021472011915105,
+      "grad_norm": 0.13722582342328407,
+      "learning_rate": 4.992208753721564e-06,
+      "loss": 0.8339,
+      "step": 222
+    },
+    {
+      "epoch": 0.11071118282239047,
+      "grad_norm": 0.1354247771990746,
+      "learning_rate": 4.992131462317518e-06,
+      "loss": 0.828,
+      "step": 223
+    },
+    {
+      "epoch": 0.11120764552562988,
+      "grad_norm": 0.13994897440630472,
+      "learning_rate": 4.992053790031136e-06,
+      "loss": 0.8453,
+      "step": 224
+    },
+    {
+      "epoch": 0.11170410822886931,
+      "grad_norm": 0.13384734719071256,
+      "learning_rate": 4.9919757368742895e-06,
+      "loss": 0.8066,
+      "step": 225
+    },
+    {
+      "epoch": 0.11220057093210872,
+      "grad_norm": 0.13127250934166673,
+      "learning_rate": 4.991897302858908e-06,
+      "loss": 0.7916,
+      "step": 226
+    },
+    {
+      "epoch": 0.11269703363534815,
+      "grad_norm": 0.14039582163447972,
+      "learning_rate": 4.9918184879969765e-06,
+      "loss": 0.8656,
+      "step": 227
+    },
+    {
+      "epoch": 0.11319349633858757,
+      "grad_norm": 0.1470635342507985,
+      "learning_rate": 4.991739292300544e-06,
+      "loss": 0.8278,
+      "step": 228
+    },
+    {
+      "epoch": 0.11368995904182698,
+      "grad_norm": 0.14174626233120474,
+      "learning_rate": 4.991659715781712e-06,
+      "loss": 0.823,
+      "step": 229
+    },
+    {
+      "epoch": 0.11418642174506641,
+      "grad_norm": 0.13862626411027507,
+      "learning_rate": 4.991579758452644e-06,
+      "loss": 0.8278,
+      "step": 230
+    },
+    {
+      "epoch": 0.11468288444830582,
+      "grad_norm": 0.13959224442195398,
+      "learning_rate": 4.991499420325558e-06,
+      "loss": 0.8252,
+      "step": 231
+    },
+    {
+      "epoch": 0.11517934715154524,
+      "grad_norm": 0.13884826767573266,
+      "learning_rate": 4.991418701412735e-06,
+      "loss": 0.8404,
+      "step": 232
+    },
+    {
+      "epoch": 0.11567580985478466,
+      "grad_norm": 0.1389057416612581,
+      "learning_rate": 4.991337601726509e-06,
+      "loss": 0.8106,
+      "step": 233
+    },
+    {
+      "epoch": 0.11617227255802408,
+      "grad_norm": 0.13638785848050972,
+      "learning_rate": 4.991256121279277e-06,
+      "loss": 0.8069,
+      "step": 234
+    },
+    {
+      "epoch": 0.11666873526126349,
+      "grad_norm": 0.1338672028203034,
+      "learning_rate": 4.991174260083491e-06,
+      "loss": 0.8019,
+      "step": 235
+    },
+    {
+      "epoch": 0.11716519796450292,
+      "grad_norm": 0.14097764226298978,
+      "learning_rate": 4.991092018151663e-06,
+      "loss": 0.846,
+      "step": 236
+    },
+    {
+      "epoch": 0.11766166066774233,
+      "grad_norm": 0.13455102270597805,
+      "learning_rate": 4.991009395496361e-06,
+      "loss": 0.8056,
+      "step": 237
+    },
+    {
+      "epoch": 0.11815812337098175,
+      "grad_norm": 0.14555871706504694,
+      "learning_rate": 4.9909263921302135e-06,
+      "loss": 0.8075,
+      "step": 238
+    },
+    {
+      "epoch": 0.11865458607422118,
+      "grad_norm": 0.14461201544386468,
+      "learning_rate": 4.990843008065905e-06,
+      "loss": 0.8704,
+      "step": 239
+    },
+    {
+      "epoch": 0.11915104877746059,
+      "grad_norm": 0.1419964925817203,
+      "learning_rate": 4.9907592433161815e-06,
+      "loss": 0.792,
+      "step": 240
+    },
+    {
+      "epoch": 0.1196475114807,
+      "grad_norm": 0.13759060310549415,
+      "learning_rate": 4.990675097893843e-06,
+      "loss": 0.7836,
+      "step": 241
+    },
+    {
+      "epoch": 0.12014397418393943,
+      "grad_norm": 0.14605868485481172,
+      "learning_rate": 4.9905905718117505e-06,
+      "loss": 0.836,
+      "step": 242
+    },
+    {
+      "epoch": 0.12064043688717885,
+      "grad_norm": 0.1373934253609785,
+      "learning_rate": 4.990505665082824e-06,
+      "loss": 0.7723,
+      "step": 243
+    },
+    {
+      "epoch": 0.12113689959041828,
+      "grad_norm": 0.13865838334876657,
+      "learning_rate": 4.9904203777200375e-06,
+      "loss": 0.7898,
+      "step": 244
+    },
+    {
+      "epoch": 0.12163336229365769,
+      "grad_norm": 0.1390399518577959,
+      "learning_rate": 4.990334709736428e-06,
+      "loss": 0.7869,
+      "step": 245
+    },
+    {
+      "epoch": 0.1221298249968971,
+      "grad_norm": 0.14077068851100413,
+      "learning_rate": 4.990248661145087e-06,
+      "loss": 0.8108,
+      "step": 246
+    },
+    {
+      "epoch": 0.12262628770013653,
+      "grad_norm": 0.1337469353425711,
+      "learning_rate": 4.9901622319591665e-06,
+      "loss": 0.8248,
+      "step": 247
+    },
+    {
+      "epoch": 0.12312275040337595,
+      "grad_norm": 0.13685347748788432,
+      "learning_rate": 4.9900754221918766e-06,
+      "loss": 0.8252,
+      "step": 248
+    },
+    {
+      "epoch": 0.12361921310661536,
+      "grad_norm": 0.13611377917936257,
+      "learning_rate": 4.989988231856483e-06,
+      "loss": 0.7827,
+      "step": 249
+    },
+    {
+      "epoch": 0.12411567580985479,
+      "grad_norm": 0.13974150757425385,
+      "learning_rate": 4.989900660966312e-06,
+      "loss": 0.8795,
+      "step": 250
+    },
+    {
+      "epoch": 0.1246121385130942,
+      "grad_norm": 0.13278222798977063,
+      "learning_rate": 4.9898127095347475e-06,
+      "loss": 0.8178,
+      "step": 251
+    },
+    {
+      "epoch": 0.12510860121633363,
+      "grad_norm": 0.13505628901465874,
+      "learning_rate": 4.989724377575231e-06,
+      "loss": 0.7909,
+      "step": 252
+    },
+    {
+      "epoch": 0.12560506391957305,
+      "grad_norm": 0.1366672871248607,
+      "learning_rate": 4.989635665101263e-06,
+      "loss": 0.8262,
+      "step": 253
+    },
+    {
+      "epoch": 0.12610152662281246,
+      "grad_norm": 0.14405211149440367,
+      "learning_rate": 4.989546572126402e-06,
+      "loss": 0.8606,
+      "step": 254
+    },
+    {
+      "epoch": 0.12659798932605187,
+      "grad_norm": 0.14227617759563377,
+      "learning_rate": 4.9894570986642655e-06,
+      "loss": 0.8379,
+      "step": 255
+    },
+    {
+      "epoch": 0.1270944520292913,
+      "grad_norm": 0.1469637596335691,
+      "learning_rate": 4.989367244728526e-06,
+      "loss": 0.8578,
+      "step": 256
+    },
+    {
+      "epoch": 0.12759091473253073,
+      "grad_norm": 0.13275558591812953,
+      "learning_rate": 4.989277010332917e-06,
+      "loss": 0.7559,
+      "step": 257
+    },
+    {
+      "epoch": 0.12808737743577014,
+      "grad_norm": 0.14791426685196862,
+      "learning_rate": 4.989186395491229e-06,
+      "loss": 0.844,
+      "step": 258
+    },
+    {
+      "epoch": 0.12858384013900956,
+      "grad_norm": 0.13244945426039428,
+      "learning_rate": 4.989095400217312e-06,
+      "loss": 0.8062,
+      "step": 259
+    },
+    {
+      "epoch": 0.12908030284224897,
+      "grad_norm": 0.13778000814954997,
+      "learning_rate": 4.9890040245250725e-06,
+      "loss": 0.8399,
+      "step": 260
+    },
+    {
+      "epoch": 0.1295767655454884,
+      "grad_norm": 0.143797201748363,
+      "learning_rate": 4.9889122684284765e-06,
+      "loss": 0.7978,
+      "step": 261
+    },
+    {
+      "epoch": 0.1300732282487278,
+      "grad_norm": 0.13685020353927613,
+      "learning_rate": 4.988820131941547e-06,
+      "loss": 0.7712,
+      "step": 262
+    },
+    {
+      "epoch": 0.13056969095196724,
+      "grad_norm": 0.13664331543020586,
+      "learning_rate": 4.988727615078365e-06,
+      "loss": 0.7458,
+      "step": 263
+    },
+    {
+      "epoch": 0.13106615365520666,
+      "grad_norm": 0.1499586905570859,
+      "learning_rate": 4.988634717853071e-06,
+      "loss": 0.8261,
+      "step": 264
+    },
+    {
+      "epoch": 0.13156261635844607,
+      "grad_norm": 0.13982088441840343,
+      "learning_rate": 4.988541440279862e-06,
+      "loss": 0.8105,
+      "step": 265
+    },
+    {
+      "epoch": 0.13205907906168549,
+      "grad_norm": 0.13605433414869017,
+      "learning_rate": 4.988447782372996e-06,
+      "loss": 0.7822,
+      "step": 266
+    },
+    {
+      "epoch": 0.1325555417649249,
+      "grad_norm": 0.13252713875584649,
+      "learning_rate": 4.988353744146784e-06,
+      "loss": 0.7532,
+      "step": 267
+    },
+    {
+      "epoch": 0.13305200446816434,
+      "grad_norm": 0.140502196927792,
+      "learning_rate": 4.988259325615601e-06,
+      "loss": 0.8733,
+      "step": 268
+    },
+    {
+      "epoch": 0.13354846717140376,
+      "grad_norm": 0.13822527446456542,
+      "learning_rate": 4.988164526793877e-06,
+      "loss": 0.7752,
+      "step": 269
+    },
+    {
+      "epoch": 0.13404492987464317,
+      "grad_norm": 0.13966558823867067,
+      "learning_rate": 4.988069347696098e-06,
+      "loss": 0.8037,
+      "step": 270
+    },
+    {
+      "epoch": 0.13454139257788258,
+      "grad_norm": 0.14070793181761732,
+      "learning_rate": 4.987973788336814e-06,
+      "loss": 0.801,
+      "step": 271
+    },
+    {
+      "epoch": 0.135037855281122,
+      "grad_norm": 0.13548015995353577,
+      "learning_rate": 4.987877848730627e-06,
+      "loss": 0.7917,
+      "step": 272
+    },
+    {
+      "epoch": 0.1355343179843614,
+      "grad_norm": 0.13914099615548198,
+      "learning_rate": 4.987781528892201e-06,
+      "loss": 0.7604,
+      "step": 273
+    },
+    {
+      "epoch": 0.13603078068760086,
+      "grad_norm": 0.1439655316233177,
+      "learning_rate": 4.987684828836257e-06,
+      "loss": 0.8128,
+      "step": 274
+    },
+    {
+      "epoch": 0.13652724339084027,
+      "grad_norm": 0.13334244239454465,
+      "learning_rate": 4.987587748577574e-06,
+      "loss": 0.7762,
+      "step": 275
+    },
+    {
+      "epoch": 0.13702370609407968,
+      "grad_norm": 0.13033286848134193,
+      "learning_rate": 4.98749028813099e-06,
+      "loss": 0.794,
+      "step": 276
+    },
+    {
+      "epoch": 0.1375201687973191,
+      "grad_norm": 0.1263283049211498,
+      "learning_rate": 4.987392447511398e-06,
+      "loss": 0.7245,
+      "step": 277
+    },
+    {
+      "epoch": 0.1380166315005585,
+      "grad_norm": 0.1346177373563915,
+      "learning_rate": 4.987294226733753e-06,
+      "loss": 0.825,
+      "step": 278
+    },
+    {
+      "epoch": 0.13851309420379793,
+      "grad_norm": 0.13797046885885128,
+      "learning_rate": 4.987195625813066e-06,
+      "loss": 0.7799,
+      "step": 279
+    },
+    {
+      "epoch": 0.13900955690703737,
+      "grad_norm": 0.13671450378929442,
+      "learning_rate": 4.987096644764407e-06,
+      "loss": 0.8186,
+      "step": 280
+    },
+    {
+      "epoch": 0.13950601961027678,
+      "grad_norm": 0.13733243307369106,
+      "learning_rate": 4.986997283602903e-06,
+      "loss": 0.8646,
+      "step": 281
+    },
+    {
+      "epoch": 0.1400024823135162,
+      "grad_norm": 0.13818658844648463,
+      "learning_rate": 4.986897542343741e-06,
+      "loss": 0.8053,
+      "step": 282
+    },
+    {
+      "epoch": 0.1404989450167556,
+      "grad_norm": 0.1376323705656438,
+      "learning_rate": 4.9867974210021634e-06,
+      "loss": 0.8298,
+      "step": 283
+    },
+    {
+      "epoch": 0.14099540771999503,
+      "grad_norm": 0.13132512407016342,
+      "learning_rate": 4.986696919593473e-06,
+      "loss": 0.7656,
+      "step": 284
+    },
+    {
+      "epoch": 0.14149187042323447,
+      "grad_norm": 0.135693807444876,
+      "learning_rate": 4.986596038133029e-06,
+      "loss": 0.8259,
+      "step": 285
+    },
+    {
+      "epoch": 0.14198833312647388,
+      "grad_norm": 0.14566445753533755,
+      "learning_rate": 4.986494776636251e-06,
+      "loss": 0.8459,
+      "step": 286
+    },
+    {
+      "epoch": 0.1424847958297133,
+      "grad_norm": 0.14082471675828692,
+      "learning_rate": 4.986393135118614e-06,
+      "loss": 0.8634,
+      "step": 287
+    },
+    {
+      "epoch": 0.1429812585329527,
+      "grad_norm": 0.13675780982750446,
+      "learning_rate": 4.9862911135956525e-06,
+      "loss": 0.7724,
+      "step": 288
+    },
+    {
+      "epoch": 0.14347772123619212,
+      "grad_norm": 0.1387994309453966,
+      "learning_rate": 4.986188712082959e-06,
+      "loss": 0.7802,
+      "step": 289
+    },
+    {
+      "epoch": 0.14397418393943154,
+      "grad_norm": 0.13354684023511523,
+      "learning_rate": 4.986085930596184e-06,
+      "loss": 0.8134,
+      "step": 290
+    },
+    {
+      "epoch": 0.14447064664267098,
+      "grad_norm": 0.14594678291314142,
+      "learning_rate": 4.985982769151035e-06,
+      "loss": 0.799,
+      "step": 291
+    },
+    {
+      "epoch": 0.1449671093459104,
+      "grad_norm": 0.14334448087535562,
+      "learning_rate": 4.985879227763281e-06,
+      "loss": 0.8697,
+      "step": 292
+    },
+    {
+      "epoch": 0.1454635720491498,
+      "grad_norm": 0.13442178316157122,
+      "learning_rate": 4.985775306448743e-06,
+      "loss": 0.794,
+      "step": 293
+    },
+    {
+      "epoch": 0.14596003475238922,
+      "grad_norm": 0.14958434833387912,
+      "learning_rate": 4.985671005223308e-06,
+      "loss": 0.8104,
+      "step": 294
+    },
+    {
+      "epoch": 0.14645649745562864,
+      "grad_norm": 0.14089328532302936,
+      "learning_rate": 4.985566324102913e-06,
+      "loss": 0.8334,
+      "step": 295
+    },
+    {
+      "epoch": 0.14695296015886805,
+      "grad_norm": 0.14930089897209584,
+      "learning_rate": 4.98546126310356e-06,
+      "loss": 0.8472,
+      "step": 296
+    },
+    {
+      "epoch": 0.1474494228621075,
+      "grad_norm": 0.13838634635947197,
+      "learning_rate": 4.9853558222413025e-06,
+      "loss": 0.8081,
+      "step": 297
+    },
+    {
+      "epoch": 0.1479458855653469,
+      "grad_norm": 0.13657833547456122,
+      "learning_rate": 4.985250001532258e-06,
+      "loss": 0.8046,
+      "step": 298
+    },
+    {
+      "epoch": 0.14844234826858632,
+      "grad_norm": 0.13416933719640858,
+      "learning_rate": 4.9851438009925985e-06,
+      "loss": 0.7718,
+      "step": 299
+    },
+    {
+      "epoch": 0.14893881097182574,
+      "grad_norm": 0.13305701512405635,
+      "learning_rate": 4.985037220638556e-06,
+      "loss": 0.7332,
+      "step": 300
+    },
+    {
+      "epoch": 0.14943527367506515,
+      "grad_norm": 0.12255084911798615,
+      "learning_rate": 4.9849302604864176e-06,
+      "loss": 0.7301,
+      "step": 301
+    },
+    {
+      "epoch": 0.1499317363783046,
+      "grad_norm": 0.1349582502588026,
+      "learning_rate": 4.9848229205525325e-06,
+      "loss": 0.8268,
+      "step": 302
+    },
+    {
+      "epoch": 0.150428199081544,
+      "grad_norm": 0.13941501409088294,
+      "learning_rate": 4.984715200853305e-06,
+      "loss": 0.8187,
+      "step": 303
+    },
+    {
+      "epoch": 0.15092466178478342,
+      "grad_norm": 0.13492743679489108,
+      "learning_rate": 4.9846071014051985e-06,
+      "loss": 0.7948,
+      "step": 304
+    },
+    {
+      "epoch": 0.15142112448802283,
+      "grad_norm": 0.1652920462764793,
+      "learning_rate": 4.984498622224734e-06,
+      "loss": 0.834,
+      "step": 305
+    },
+    {
+      "epoch": 0.15191758719126225,
+      "grad_norm": 0.13639513644830636,
+      "learning_rate": 4.984389763328491e-06,
+      "loss": 0.8263,
+      "step": 306
+    },
+    {
+      "epoch": 0.15241404989450166,
+      "grad_norm": 0.13894535392298077,
+      "learning_rate": 4.984280524733107e-06,
+      "loss": 0.8399,
+      "step": 307
+    },
+    {
+      "epoch": 0.1529105125977411,
+      "grad_norm": 0.1427172968985807,
+      "learning_rate": 4.984170906455277e-06,
+      "loss": 0.8176,
+      "step": 308
+    },
+    {
+      "epoch": 0.15340697530098052,
+      "grad_norm": 0.13916573482217912,
+      "learning_rate": 4.984060908511755e-06,
+      "loss": 0.84,
+      "step": 309
+    },
+    {
+      "epoch": 0.15390343800421993,
+      "grad_norm": 0.13640696194757118,
+      "learning_rate": 4.983950530919352e-06,
+      "loss": 0.8378,
+      "step": 310
+    },
+    {
+      "epoch": 0.15439990070745935,
+      "grad_norm": 0.1269776959706502,
+      "learning_rate": 4.983839773694937e-06,
+      "loss": 0.7752,
+      "step": 311
+    },
+    {
+      "epoch": 0.15489636341069876,
+      "grad_norm": 0.13735768715592372,
+      "learning_rate": 4.983728636855438e-06,
+      "loss": 0.8528,
+      "step": 312
+    },
+    {
+      "epoch": 0.1553928261139382,
+      "grad_norm": 0.13601796887750978,
+      "learning_rate": 4.983617120417841e-06,
+      "loss": 0.7813,
+      "step": 313
+    },
+    {
+      "epoch": 0.15588928881717762,
+      "grad_norm": 0.1394868538656493,
+      "learning_rate": 4.983505224399188e-06,
+      "loss": 0.8121,
+      "step": 314
+    },
+    {
+      "epoch": 0.15638575152041703,
+      "grad_norm": 0.1393241198294074,
+      "learning_rate": 4.983392948816582e-06,
+      "loss": 0.8167,
+      "step": 315
+    },
+    {
+      "epoch": 0.15688221422365645,
+      "grad_norm": 0.13601273115897386,
+      "learning_rate": 4.9832802936871815e-06,
+      "loss": 0.815,
+      "step": 316
+    },
+    {
+      "epoch": 0.15737867692689586,
+      "grad_norm": 0.13309082027141247,
+      "learning_rate": 4.983167259028205e-06,
+      "loss": 0.7845,
+      "step": 317
+    },
+    {
+      "epoch": 0.15787513963013527,
+      "grad_norm": 0.1328710618339079,
+      "learning_rate": 4.983053844856928e-06,
+      "loss": 0.837,
+      "step": 318
+    },
+    {
+      "epoch": 0.15837160233337472,
+      "grad_norm": 0.1382512821366624,
+      "learning_rate": 4.982940051190682e-06,
+      "loss": 0.7642,
+      "step": 319
+    },
+    {
+      "epoch": 0.15886806503661413,
+      "grad_norm": 0.13347990081206224,
+      "learning_rate": 4.982825878046862e-06,
+      "loss": 0.7786,
+      "step": 320
+    },
+    {
+      "epoch": 0.15936452773985355,
+      "grad_norm": 0.1350212390870091,
+      "learning_rate": 4.9827113254429144e-06,
+      "loss": 0.7822,
+      "step": 321
+    },
+    {
+      "epoch": 0.15986099044309296,
+      "grad_norm": 0.13659963300872746,
+      "learning_rate": 4.982596393396348e-06,
+      "loss": 0.7519,
+      "step": 322
+    },
+    {
+      "epoch": 0.16035745314633237,
+      "grad_norm": 0.14307889455512066,
+      "learning_rate": 4.982481081924728e-06,
+      "loss": 0.8533,
+      "step": 323
+    },
+    {
+      "epoch": 0.1608539158495718,
+      "grad_norm": 0.1330601424580335,
+      "learning_rate": 4.982365391045679e-06,
+      "loss": 0.8239,
+      "step": 324
+    },
+    {
+      "epoch": 0.16135037855281123,
+      "grad_norm": 0.15026890457234857,
+      "learning_rate": 4.982249320776882e-06,
+      "loss": 0.8806,
+      "step": 325
+    },
+    {
+      "epoch": 0.16184684125605064,
+      "grad_norm": 0.12974893713522884,
+      "learning_rate": 4.982132871136075e-06,
+      "loss": 0.7378,
+      "step": 326
+    },
+    {
+      "epoch": 0.16234330395929006,
+      "grad_norm": 0.1444247544686782,
+      "learning_rate": 4.9820160421410575e-06,
+      "loss": 0.8243,
+      "step": 327
+    },
+    {
+      "epoch": 0.16283976666252947,
+      "grad_norm": 0.1323450801185765,
+      "learning_rate": 4.981898833809684e-06,
+      "loss": 0.7696,
+      "step": 328
+    },
+    {
+      "epoch": 0.1633362293657689,
+      "grad_norm": 0.1353207749212687,
+      "learning_rate": 4.981781246159867e-06,
+      "loss": 0.7677,
+      "step": 329
+    },
+    {
+      "epoch": 0.16383269206900833,
+      "grad_norm": 0.12787497196232483,
+      "learning_rate": 4.98166327920958e-06,
+      "loss": 0.7655,
+      "step": 330
+    },
+    {
+      "epoch": 0.16432915477224774,
+      "grad_norm": 0.13359250170701767,
+      "learning_rate": 4.9815449329768505e-06,
+      "loss": 0.8392,
+      "step": 331
+    },
+    {
+      "epoch": 0.16482561747548716,
+      "grad_norm": 0.13162070870865217,
+      "learning_rate": 4.981426207479767e-06,
+      "loss": 0.8023,
+      "step": 332
+    },
+    {
+      "epoch": 0.16532208017872657,
+      "grad_norm": 0.1340847314964679,
+      "learning_rate": 4.981307102736474e-06,
+      "loss": 0.8049,
+      "step": 333
+    },
+    {
+      "epoch": 0.16581854288196599,
+      "grad_norm": 0.13799078318837465,
+      "learning_rate": 4.981187618765175e-06,
+      "loss": 0.801,
+      "step": 334
+    },
+    {
+      "epoch": 0.1663150055852054,
+      "grad_norm": 0.1352495083171962,
+      "learning_rate": 4.981067755584131e-06,
+      "loss": 0.8028,
+      "step": 335
+    },
+    {
+      "epoch": 0.16681146828844484,
+      "grad_norm": 0.13424131175066562,
+      "learning_rate": 4.9809475132116624e-06,
+      "loss": 0.8148,
+      "step": 336
+    },
+    {
+      "epoch": 0.16730793099168426,
+      "grad_norm": 0.13380623256179228,
+      "learning_rate": 4.980826891666145e-06,
+      "loss": 0.7906,
+      "step": 337
+    },
+    {
+      "epoch": 0.16780439369492367,
+      "grad_norm": 0.13711848434016552,
+      "learning_rate": 4.980705890966014e-06,
+      "loss": 0.7818,
+      "step": 338
+    },
+    {
+      "epoch": 0.16830085639816308,
+      "grad_norm": 0.1408936385231929,
+      "learning_rate": 4.980584511129763e-06,
+      "loss": 0.8319,
+      "step": 339
+    },
+    {
+      "epoch": 0.1687973191014025,
+      "grad_norm": 0.1361241205412377,
+      "learning_rate": 4.980462752175943e-06,
+      "loss": 0.7816,
+      "step": 340
+    },
+    {
+      "epoch": 0.1692937818046419,
+      "grad_norm": 0.13336550378881848,
+      "learning_rate": 4.980340614123162e-06,
+      "loss": 0.7656,
+      "step": 341
+    },
+    {
+      "epoch": 0.16979024450788135,
+      "grad_norm": 0.1291163574188604,
+      "learning_rate": 4.980218096990087e-06,
+      "loss": 0.7573,
+      "step": 342
+    },
+    {
+      "epoch": 0.17028670721112077,
+      "grad_norm": 0.1331261115959964,
+      "learning_rate": 4.980095200795443e-06,
+      "loss": 0.7838,
+      "step": 343
+    },
+    {
+      "epoch": 0.17078316991436018,
+      "grad_norm": 0.13606378865652022,
+      "learning_rate": 4.979971925558014e-06,
+      "loss": 0.817,
+      "step": 344
+    },
+    {
+      "epoch": 0.1712796326175996,
+      "grad_norm": 0.12897271331540627,
+      "learning_rate": 4.979848271296639e-06,
+      "loss": 0.7615,
+      "step": 345
+    },
+    {
+      "epoch": 0.171776095320839,
+      "grad_norm": 0.14023828743325562,
+      "learning_rate": 4.979724238030217e-06,
+      "loss": 0.7834,
+      "step": 346
+    },
+    {
+      "epoch": 0.17227255802407845,
+      "grad_norm": 0.13830905999777027,
+      "learning_rate": 4.979599825777704e-06,
+      "loss": 0.8302,
+      "step": 347
+    },
+    {
+      "epoch": 0.17276902072731787,
+      "grad_norm": 0.1428666780776237,
+      "learning_rate": 4.979475034558115e-06,
+      "loss": 0.8588,
+      "step": 348
+    },
+    {
+      "epoch": 0.17326548343055728,
+      "grad_norm": 0.128916884054928,
+      "learning_rate": 4.979349864390523e-06,
+      "loss": 0.7776,
+      "step": 349
+    },
+    {
+      "epoch": 0.1737619461337967,
+      "grad_norm": 0.13925327403441065,
+      "learning_rate": 4.9792243152940576e-06,
+      "loss": 0.8802,
+      "step": 350
+    },
+    {
+      "epoch": 0.1742584088370361,
+      "grad_norm": 0.1363578403401138,
+      "learning_rate": 4.979098387287907e-06,
+      "loss": 0.8118,
+      "step": 351
+    },
+    {
+      "epoch": 0.17475487154027552,
+      "grad_norm": 0.14256811683930512,
+      "learning_rate": 4.978972080391317e-06,
+      "loss": 0.7879,
+      "step": 352
+    },
+    {
+      "epoch": 0.17525133424351497,
+      "grad_norm": 0.13404514324949623,
+      "learning_rate": 4.978845394623591e-06,
+      "loss": 0.7738,
+      "step": 353
+    },
+    {
+      "epoch": 0.17574779694675438,
+      "grad_norm": 0.14672131580692716,
+      "learning_rate": 4.978718330004093e-06,
+      "loss": 0.8091,
+      "step": 354
+    },
+    {
+      "epoch": 0.1762442596499938,
+      "grad_norm": 0.13482860481948047,
+      "learning_rate": 4.978590886552241e-06,
+      "loss": 0.7855,
+      "step": 355
+    },
+    {
+      "epoch": 0.1767407223532332,
+      "grad_norm": 0.1390863998270458,
+      "learning_rate": 4.978463064287513e-06,
+      "loss": 0.8136,
+      "step": 356
+    },
+    {
+      "epoch": 0.17723718505647262,
+      "grad_norm": 0.12766825836109758,
+      "learning_rate": 4.978334863229445e-06,
+      "loss": 0.7519,
+      "step": 357
+    },
+    {
+      "epoch": 0.17773364775971204,
+      "grad_norm": 0.13957492610638372,
+      "learning_rate": 4.97820628339763e-06,
+      "loss": 0.8374,
+      "step": 358
+    },
+    {
+      "epoch": 0.17823011046295148,
+      "grad_norm": 0.13624109938717846,
+      "learning_rate": 4.97807732481172e-06,
+      "loss": 0.8468,
+      "step": 359
+    },
+    {
+      "epoch": 0.1787265731661909,
+      "grad_norm": 0.1371796025283825,
+      "learning_rate": 4.977947987491424e-06,
+      "loss": 0.8134,
+      "step": 360
+    },
+    {
+      "epoch": 0.1792230358694303,
+      "grad_norm": 0.13791288379959807,
+      "learning_rate": 4.977818271456508e-06,
+      "loss": 0.8523,
+      "step": 361
+    },
+    {
+      "epoch": 0.17971949857266972,
+      "grad_norm": 0.13682556966382917,
+      "learning_rate": 4.977688176726799e-06,
+      "loss": 0.7998,
+      "step": 362
+    },
+    {
+      "epoch": 0.18021596127590914,
+      "grad_norm": 0.12914631502678342,
+      "learning_rate": 4.977557703322178e-06,
+      "loss": 0.7715,
+      "step": 363
+    },
+    {
+      "epoch": 0.18071242397914858,
+      "grad_norm": 0.13450502572602166,
+      "learning_rate": 4.977426851262588e-06,
+      "loss": 0.7999,
+      "step": 364
+    },
+    {
+      "epoch": 0.181208886682388,
+      "grad_norm": 0.1324950964120947,
+      "learning_rate": 4.977295620568025e-06,
+      "loss": 0.773,
+      "step": 365
+    },
+    {
+      "epoch": 0.1817053493856274,
+      "grad_norm": 0.13291470130623328,
+      "learning_rate": 4.977164011258547e-06,
+      "loss": 0.8047,
+      "step": 366
+    },
+    {
+      "epoch": 0.18220181208886682,
+      "grad_norm": 0.14303063683318232,
+      "learning_rate": 4.977032023354269e-06,
+      "loss": 0.8361,
+      "step": 367
+    },
+    {
+      "epoch": 0.18269827479210624,
+      "grad_norm": 0.14299058880067156,
+      "learning_rate": 4.976899656875361e-06,
+      "loss": 0.8172,
+      "step": 368
+    },
+    {
+      "epoch": 0.18319473749534565,
+      "grad_norm": 0.13663382187886108,
+      "learning_rate": 4.976766911842056e-06,
+      "loss": 0.7529,
+      "step": 369
+    },
+    {
+      "epoch": 0.1836912001985851,
+      "grad_norm": 0.13831178738187494,
+      "learning_rate": 4.9766337882746395e-06,
+      "loss": 0.7863,
+      "step": 370
+    },
+    {
+      "epoch": 0.1841876629018245,
+      "grad_norm": 0.13456481368960774,
+      "learning_rate": 4.976500286193458e-06,
+      "loss": 0.8505,
+      "step": 371
+    },
+    {
+      "epoch": 0.18468412560506392,
+      "grad_norm": 0.13582472070675122,
+      "learning_rate": 4.976366405618916e-06,
+      "loss": 0.8036,
+      "step": 372
+    },
+    {
+      "epoch": 0.18518058830830333,
+      "grad_norm": 0.13588434233209187,
+      "learning_rate": 4.976232146571476e-06,
+      "loss": 0.7814,
+      "step": 373
+    },
+    {
+      "epoch": 0.18567705101154275,
+      "grad_norm": 0.13408060178137982,
+      "learning_rate": 4.976097509071654e-06,
+      "loss": 0.824,
+      "step": 374
+    },
+    {
+      "epoch": 0.1861735137147822,
+      "grad_norm": 0.13739925696360844,
+      "learning_rate": 4.975962493140029e-06,
+      "loss": 0.8233,
+      "step": 375
+    },
+    {
+      "epoch": 0.1866699764180216,
+      "grad_norm": 0.12982209493306304,
+      "learning_rate": 4.9758270987972356e-06,
+      "loss": 0.7537,
+      "step": 376
+    },
+    {
+      "epoch": 0.18716643912126102,
+      "grad_norm": 0.13310617104329844,
+      "learning_rate": 4.975691326063968e-06,
+      "loss": 0.7771,
+      "step": 377
+    },
+    {
+      "epoch": 0.18766290182450043,
+      "grad_norm": 0.13823931193705333,
+      "learning_rate": 4.9755551749609755e-06,
+      "loss": 0.7723,
+      "step": 378
+    },
+    {
+      "epoch": 0.18815936452773985,
+      "grad_norm": 0.1382970645097716,
+      "learning_rate": 4.975418645509066e-06,
+      "loss": 0.8002,
+      "step": 379
+    },
+    {
+      "epoch": 0.18865582723097926,
+      "grad_norm": 0.13785374560341188,
+      "learning_rate": 4.975281737729109e-06,
+      "loss": 0.7799,
+      "step": 380
+    },
+    {
+      "epoch": 0.1891522899342187,
+      "grad_norm": 0.13630037342754214,
+      "learning_rate": 4.975144451642024e-06,
+      "loss": 0.7825,
+      "step": 381
+    },
+    {
+      "epoch": 0.18964875263745812,
+      "grad_norm": 0.13742010823689682,
+      "learning_rate": 4.975006787268797e-06,
+      "loss": 0.7832,
+      "step": 382
+    },
+    {
+      "epoch": 0.19014521534069753,
+      "grad_norm": 0.1372266789447449,
+      "learning_rate": 4.974868744630467e-06,
+      "loss": 0.7805,
+      "step": 383
+    },
+    {
+      "epoch": 0.19064167804393695,
+      "grad_norm": 0.13335756294028897,
+      "learning_rate": 4.974730323748129e-06,
+      "loss": 0.7747,
+      "step": 384
+    },
+    {
+      "epoch": 0.19113814074717636,
+      "grad_norm": 0.1331325911806612,
+      "learning_rate": 4.974591524642942e-06,
+      "loss": 0.7626,
+      "step": 385
+    },
+    {
+      "epoch": 0.19163460345041577,
+      "grad_norm": 0.13528286793265076,
+      "learning_rate": 4.974452347336116e-06,
+      "loss": 0.7906,
+      "step": 386
+    },
+    {
+      "epoch": 0.19213106615365522,
+      "grad_norm": 0.13710197959682258,
+      "learning_rate": 4.974312791848925e-06,
+      "loss": 0.7775,
+      "step": 387
+    },
+    {
+      "epoch": 0.19262752885689463,
+      "grad_norm": 0.13646907256302476,
+      "learning_rate": 4.974172858202695e-06,
+      "loss": 0.8266,
+      "step": 388
+    },
+    {
+      "epoch": 0.19312399156013405,
+      "grad_norm": 0.14002294190339723,
+      "learning_rate": 4.974032546418816e-06,
+      "loss": 0.7807,
+      "step": 389
+    },
+    {
+      "epoch": 0.19362045426337346,
+      "grad_norm": 0.14404269798290067,
+      "learning_rate": 4.973891856518728e-06,
+      "loss": 0.8195,
+      "step": 390
+    },
+    {
+      "epoch": 0.19411691696661287,
+      "grad_norm": 0.13811976877791707,
+      "learning_rate": 4.973750788523937e-06,
+      "loss": 0.8008,
+      "step": 391
+    },
+    {
+      "epoch": 0.19461337966985232,
+      "grad_norm": 0.14072300476936328,
+      "learning_rate": 4.9736093424560005e-06,
+      "loss": 0.7705,
+      "step": 392
+    },
+    {
+      "epoch": 0.19510984237309173,
+      "grad_norm": 0.14423164849233763,
+      "learning_rate": 4.973467518336538e-06,
+      "loss": 0.7988,
+      "step": 393
+    },
+    {
+      "epoch": 0.19560630507633114,
+      "grad_norm": 0.13418208376903598,
+      "learning_rate": 4.973325316187225e-06,
+      "loss": 0.7761,
+      "step": 394
+    },
+    {
+      "epoch": 0.19610276777957056,
+      "grad_norm": 0.1316889322126393,
+      "learning_rate": 4.973182736029793e-06,
+      "loss": 0.847,
+      "step": 395
+    },
+    {
+      "epoch": 0.19659923048280997,
+      "grad_norm": 0.14201288602040998,
+      "learning_rate": 4.973039777886035e-06,
+      "loss": 0.7743,
+      "step": 396
+    },
+    {
+      "epoch": 0.1970956931860494,
+      "grad_norm": 0.14278977007281732,
+      "learning_rate": 4.9728964417777986e-06,
+      "loss": 0.7947,
+      "step": 397
+    },
+    {
+      "epoch": 0.19759215588928883,
+      "grad_norm": 0.13398922706134572,
+      "learning_rate": 4.972752727726992e-06,
+      "loss": 0.75,
+      "step": 398
+    },
+    {
+      "epoch": 0.19808861859252824,
+      "grad_norm": 0.1375355686898203,
+      "learning_rate": 4.972608635755577e-06,
+      "loss": 0.781,
+      "step": 399
+    },
+    {
+      "epoch": 0.19858508129576766,
+      "grad_norm": 0.1371602165214636,
+      "learning_rate": 4.972464165885579e-06,
+      "loss": 0.823,
+      "step": 400
+    },
+    {
+      "epoch": 0.19908154399900707,
+      "grad_norm": 0.14577906731021556,
+      "learning_rate": 4.972319318139074e-06,
+      "loss": 0.8288,
+      "step": 401
+    },
+    {
+      "epoch": 0.19957800670224649,
+      "grad_norm": 0.15599884531010932,
+      "learning_rate": 4.972174092538203e-06,
+      "loss": 0.8007,
+      "step": 402
+    },
+    {
+      "epoch": 0.2000744694054859,
+      "grad_norm": 0.13716250162699573,
+      "learning_rate": 4.97202848910516e-06,
+      "loss": 0.763,
+      "step": 403
+    },
+    {
+      "epoch": 0.20057093210872534,
+      "grad_norm": 0.13013318815804806,
+      "learning_rate": 4.9718825078622e-06,
+      "loss": 0.7654,
+      "step": 404
+    },
+    {
+      "epoch": 0.20106739481196476,
+      "grad_norm": 0.14489783733703743,
+      "learning_rate": 4.971736148831631e-06,
+      "loss": 0.8095,
+      "step": 405
+    },
+    {
+      "epoch": 0.20156385751520417,
+      "grad_norm": 0.14178082134535308,
+      "learning_rate": 4.971589412035823e-06,
+      "loss": 0.7955,
+      "step": 406
+    },
+    {
+      "epoch": 0.20206032021844358,
+      "grad_norm": 0.14527202694186261,
+      "learning_rate": 4.971442297497202e-06,
+      "loss": 0.8,
+      "step": 407
+    },
+    {
+      "epoch": 0.202556782921683,
+      "grad_norm": 0.13886094340580113,
+      "learning_rate": 4.971294805238252e-06,
+      "loss": 0.8482,
+      "step": 408
+    },
+    {
+      "epoch": 0.20305324562492244,
+      "grad_norm": 0.15143893194787292,
+      "learning_rate": 4.971146935281517e-06,
+      "loss": 0.7927,
+      "step": 409
+    },
+    {
+      "epoch": 0.20354970832816185,
+      "grad_norm": 0.1636495293029831,
+      "learning_rate": 4.970998687649593e-06,
+      "loss": 0.8153,
+      "step": 410
+    },
+    {
+      "epoch": 0.20404617103140127,
+      "grad_norm": 0.13453836074071723,
+      "learning_rate": 4.97085006236514e-06,
+      "loss": 0.7805,
+      "step": 411
+    },
+    {
+      "epoch": 0.20454263373464068,
+      "grad_norm": 0.16194606706618664,
+      "learning_rate": 4.970701059450872e-06,
+      "loss": 0.821,
+      "step": 412
+    },
+    {
+      "epoch": 0.2050390964378801,
+      "grad_norm": 0.13277679173675005,
+      "learning_rate": 4.970551678929562e-06,
+      "loss": 0.7262,
+      "step": 413
+    },
+    {
+      "epoch": 0.2055355591411195,
+      "grad_norm": 0.13842208760542138,
+      "learning_rate": 4.970401920824039e-06,
+      "loss": 0.7754,
+      "step": 414
+    },
+    {
+      "epoch": 0.20603202184435895,
+      "grad_norm": 0.13680895396236756,
+      "learning_rate": 4.970251785157193e-06,
+      "loss": 0.7984,
+      "step": 415
+    },
+    {
+      "epoch": 0.20652848454759837,
+      "grad_norm": 0.1332819441058511,
+      "learning_rate": 4.9701012719519694e-06,
+      "loss": 0.7827,
+      "step": 416
+    },
+    {
+      "epoch": 0.20702494725083778,
+      "grad_norm": 0.13155005386495136,
+      "learning_rate": 4.969950381231371e-06,
+      "loss": 0.7598,
+      "step": 417
+    },
+    {
+      "epoch": 0.2075214099540772,
+      "grad_norm": 0.13812357614760978,
+      "learning_rate": 4.969799113018459e-06,
+      "loss": 0.7805,
+      "step": 418
+    },
+    {
+      "epoch": 0.2080178726573166,
+      "grad_norm": 0.15001010565178097,
+      "learning_rate": 4.9696474673363536e-06,
+      "loss": 0.7916,
+      "step": 419
+    },
+    {
+      "epoch": 0.20851433536055602,
+      "grad_norm": 0.14065673015444682,
+      "learning_rate": 4.96949544420823e-06,
+      "loss": 0.7902,
+      "step": 420
+    },
+    {
+      "epoch": 0.20901079806379547,
+      "grad_norm": 0.13896425621312414,
+      "learning_rate": 4.969343043657323e-06,
+      "loss": 0.78,
+      "step": 421
+    },
+    {
+      "epoch": 0.20950726076703488,
+      "grad_norm": 0.14937608116406495,
+      "learning_rate": 4.969190265706926e-06,
+      "loss": 0.8289,
+      "step": 422
+    },
+    {
+      "epoch": 0.2100037234702743,
+      "grad_norm": 0.13459240584003007,
+      "learning_rate": 4.969037110380387e-06,
+      "loss": 0.7906,
+      "step": 423
+    },
+    {
+      "epoch": 0.2105001861735137,
+      "grad_norm": 0.1484626362599989,
+      "learning_rate": 4.968883577701112e-06,
+      "loss": 0.8244,
+      "step": 424
+    },
+    {
+      "epoch": 0.21099664887675312,
+      "grad_norm": 0.13015898606801415,
+      "learning_rate": 4.9687296676925686e-06,
+      "loss": 0.7431,
+      "step": 425
+    },
+    {
+      "epoch": 0.21149311157999257,
+      "grad_norm": 0.1377331802046174,
+      "learning_rate": 4.96857538037828e-06,
+      "loss": 0.8043,
+      "step": 426
+    },
+    {
+      "epoch": 0.21198957428323198,
+      "grad_norm": 0.1363174078519366,
+      "learning_rate": 4.968420715781823e-06,
+      "loss": 0.7584,
+      "step": 427
+    },
+    {
+      "epoch": 0.2124860369864714,
+      "grad_norm": 0.14109517205489622,
+      "learning_rate": 4.9682656739268385e-06,
+      "loss": 0.7868,
+      "step": 428
+    },
+    {
+      "epoch": 0.2129824996897108,
+      "grad_norm": 0.14476636902871076,
+      "learning_rate": 4.968110254837022e-06,
+      "loss": 0.8051,
+      "step": 429
+    },
+    {
+      "epoch": 0.21347896239295022,
+      "grad_norm": 0.13488134316489944,
+      "learning_rate": 4.967954458536126e-06,
+      "loss": 0.7816,
+      "step": 430
+    },
+    {
+      "epoch": 0.21397542509618964,
+      "grad_norm": 0.1311388707406287,
+      "learning_rate": 4.967798285047961e-06,
+      "loss": 0.7856,
+      "step": 431
+    },
+    {
+      "epoch": 0.21447188779942908,
+      "grad_norm": 0.13733380279723623,
+      "learning_rate": 4.967641734396397e-06,
+      "loss": 0.8495,
+      "step": 432
+    },
+    {
+      "epoch": 0.2149683505026685,
+      "grad_norm": 0.13619991947733362,
+      "learning_rate": 4.967484806605359e-06,
+      "loss": 0.7803,
+      "step": 433
+    },
+    {
+      "epoch": 0.2154648132059079,
+      "grad_norm": 0.13862255789954744,
+      "learning_rate": 4.967327501698831e-06,
+      "loss": 0.7749,
+      "step": 434
+    },
+    {
+      "epoch": 0.21596127590914732,
+      "grad_norm": 0.1451665257035334,
+      "learning_rate": 4.967169819700856e-06,
+      "loss": 0.7966,
+      "step": 435
+    },
+    {
+      "epoch": 0.21645773861238674,
+      "grad_norm": 0.13309881671792553,
+      "learning_rate": 4.967011760635532e-06,
+      "loss": 0.749,
+      "step": 436
+    },
+    {
+      "epoch": 0.21695420131562618,
+      "grad_norm": 0.14123567913488688,
+      "learning_rate": 4.966853324527015e-06,
+      "loss": 0.7611,
+      "step": 437
+    },
+    {
+      "epoch": 0.2174506640188656,
+      "grad_norm": 0.1398898539392954,
+      "learning_rate": 4.966694511399521e-06,
+      "loss": 0.8463,
+      "step": 438
+    },
+    {
+      "epoch": 0.217947126722105,
+      "grad_norm": 0.1372879573733997,
+      "learning_rate": 4.9665353212773215e-06,
+      "loss": 0.7858,
+      "step": 439
+    },
+    {
+      "epoch": 0.21844358942534442,
+      "grad_norm": 0.13654858802550288,
+      "learning_rate": 4.966375754184746e-06,
+      "loss": 0.8215,
+      "step": 440
+    },
+    {
+      "epoch": 0.21894005212858383,
+      "grad_norm": 0.1319581844818004,
+      "learning_rate": 4.966215810146181e-06,
+      "loss": 0.8257,
+      "step": 441
+    },
+    {
+      "epoch": 0.21943651483182325,
+      "grad_norm": 0.13477793443314756,
+      "learning_rate": 4.966055489186072e-06,
+      "loss": 0.8037,
+      "step": 442
+    },
+    {
+      "epoch": 0.2199329775350627,
+      "grad_norm": 0.13368048885447226,
+      "learning_rate": 4.965894791328924e-06,
+      "loss": 0.7658,
+      "step": 443
+    },
+    {
+      "epoch": 0.2204294402383021,
+      "grad_norm": 0.13455924110626613,
+      "learning_rate": 4.965733716599292e-06,
+      "loss": 0.8163,
+      "step": 444
+    },
+    {
+      "epoch": 0.22092590294154152,
+      "grad_norm": 0.1389198409505174,
+      "learning_rate": 4.965572265021798e-06,
+      "loss": 0.8417,
+      "step": 445
+    },
+    {
+      "epoch": 0.22142236564478093,
+      "grad_norm": 0.13388284038750456,
+      "learning_rate": 4.965410436621115e-06,
+      "loss": 0.7774,
+      "step": 446
+    },
+    {
+      "epoch": 0.22191882834802035,
+      "grad_norm": 0.132820811008653,
+      "learning_rate": 4.965248231421977e-06,
+      "loss": 0.8015,
+      "step": 447
+    },
+    {
+      "epoch": 0.22241529105125976,
+      "grad_norm": 0.13543397173414973,
+      "learning_rate": 4.965085649449175e-06,
+      "loss": 0.8054,
+      "step": 448
+    },
+    {
+      "epoch": 0.2229117537544992,
+      "grad_norm": 0.13391298366876886,
+      "learning_rate": 4.964922690727555e-06,
+      "loss": 0.8157,
+      "step": 449
+    },
+    {
+      "epoch": 0.22340821645773862,
+      "grad_norm": 0.13119644805742203,
+      "learning_rate": 4.964759355282024e-06,
+      "loss": 0.7737,
+      "step": 450
+    },
+    {
+      "epoch": 0.22390467916097803,
+      "grad_norm": 0.13789404014972487,
+      "learning_rate": 4.964595643137544e-06,
+      "loss": 0.8227,
+      "step": 451
+    },
+    {
+      "epoch": 0.22440114186421745,
+      "grad_norm": 0.13702173759120256,
+      "learning_rate": 4.964431554319138e-06,
+      "loss": 0.7777,
+      "step": 452
+    },
+    {
+      "epoch": 0.22489760456745686,
+      "grad_norm": 0.167958896728115,
+      "learning_rate": 4.964267088851883e-06,
+      "loss": 0.8295,
+      "step": 453
+    },
+    {
+      "epoch": 0.2253940672706963,
+      "grad_norm": 0.1362052877173801,
+      "learning_rate": 4.964102246760915e-06,
+      "loss": 0.8139,
+      "step": 454
+    },
+    {
+      "epoch": 0.22589052997393572,
+      "grad_norm": 0.13619122221235858,
+      "learning_rate": 4.963937028071427e-06,
+      "loss": 0.7996,
+      "step": 455
+    },
+    {
+      "epoch": 0.22638699267717513,
+      "grad_norm": 0.13307543777024503,
+      "learning_rate": 4.96377143280867e-06,
+      "loss": 0.7634,
+      "step": 456
+    },
+    {
+      "epoch": 0.22688345538041454,
+      "grad_norm": 0.13177673869505957,
+      "learning_rate": 4.963605460997954e-06,
+      "loss": 0.7707,
+      "step": 457
+    },
+    {
+      "epoch": 0.22737991808365396,
+      "grad_norm": 0.13107526362132932,
+      "learning_rate": 4.963439112664644e-06,
+      "loss": 0.7883,
+      "step": 458
+    },
+    {
+      "epoch": 0.22787638078689337,
+      "grad_norm": 0.13905062257304818,
+      "learning_rate": 4.963272387834163e-06,
+      "loss": 0.7869,
+      "step": 459
+    },
+    {
+      "epoch": 0.22837284349013282,
+      "grad_norm": 0.14228112591091902,
+      "learning_rate": 4.963105286531994e-06,
+      "loss": 0.7684,
+      "step": 460
+    },
+    {
+      "epoch": 0.22886930619337223,
+      "grad_norm": 0.1278341075030752,
+      "learning_rate": 4.962937808783675e-06,
+      "loss": 0.7381,
+      "step": 461
+    },
+    {
+      "epoch": 0.22936576889661164,
+      "grad_norm": 0.13836582166759392,
+      "learning_rate": 4.962769954614802e-06,
+      "loss": 0.8072,
+      "step": 462
+    },
+    {
+      "epoch": 0.22986223159985106,
+      "grad_norm": 0.13221136159222038,
+      "learning_rate": 4.962601724051029e-06,
+      "loss": 0.7667,
+      "step": 463
+    },
+    {
+      "epoch": 0.23035869430309047,
+      "grad_norm": 0.1359691390529077,
+      "learning_rate": 4.962433117118067e-06,
+      "loss": 0.8016,
+      "step": 464
+    },
+    {
+      "epoch": 0.2308551570063299,
+      "grad_norm": 0.13838185689016483,
+      "learning_rate": 4.962264133841686e-06,
+      "loss": 0.7823,
+      "step": 465
+    },
+    {
+      "epoch": 0.23135161970956933,
+      "grad_norm": 0.13217829549783422,
+      "learning_rate": 4.96209477424771e-06,
+      "loss": 0.7233,
+      "step": 466
+    },
+    {
+      "epoch": 0.23184808241280874,
+      "grad_norm": 0.12783947356664202,
+      "learning_rate": 4.9619250383620256e-06,
+      "loss": 0.7493,
+      "step": 467
+    },
+    {
+      "epoch": 0.23234454511604816,
+      "grad_norm": 0.13083482079204406,
+      "learning_rate": 4.961754926210572e-06,
+      "loss": 0.7389,
+      "step": 468
+    },
+    {
+      "epoch": 0.23284100781928757,
+      "grad_norm": 0.13464481187124352,
+      "learning_rate": 4.9615844378193505e-06,
+      "loss": 0.78,
+      "step": 469
+    },
+    {
+      "epoch": 0.23333747052252699,
+      "grad_norm": 0.13684478446270137,
+      "learning_rate": 4.961413573214415e-06,
+      "loss": 0.8143,
+      "step": 470
+    },
+    {
+      "epoch": 0.23383393322576643,
+      "grad_norm": 0.14377264892348118,
+      "learning_rate": 4.9612423324218816e-06,
+      "loss": 0.8265,
+      "step": 471
+    },
+    {
+      "epoch": 0.23433039592900584,
+      "grad_norm": 0.13245307939757578,
+      "learning_rate": 4.961070715467921e-06,
+      "loss": 0.8093,
+      "step": 472
+    },
+    {
+      "epoch": 0.23482685863224526,
+      "grad_norm": 0.1373381453121188,
+      "learning_rate": 4.9608987223787606e-06,
+      "loss": 0.7547,
+      "step": 473
+    },
+    {
+      "epoch": 0.23532332133548467,
+      "grad_norm": 0.1370145480054436,
+      "learning_rate": 4.960726353180688e-06,
+      "loss": 0.8459,
+      "step": 474
+    },
+    {
+      "epoch": 0.23581978403872408,
+      "grad_norm": 0.1375438658427002,
+      "learning_rate": 4.960553607900047e-06,
+      "loss": 0.7813,
+      "step": 475
+    },
+    {
+      "epoch": 0.2363162467419635,
+      "grad_norm": 0.13778611231342322,
+      "learning_rate": 4.96038048656324e-06,
+      "loss": 0.7739,
+      "step": 476
+    },
+    {
+      "epoch": 0.23681270944520294,
+      "grad_norm": 0.13166121819124513,
+      "learning_rate": 4.9602069891967245e-06,
+      "loss": 0.7708,
+      "step": 477
+    },
+    {
+      "epoch": 0.23730917214844235,
+      "grad_norm": 0.1396399426393052,
+      "learning_rate": 4.9600331158270175e-06,
+      "loss": 0.7725,
+      "step": 478
+    },
+    {
+      "epoch": 0.23780563485168177,
+      "grad_norm": 0.1392271735139755,
+      "learning_rate": 4.959858866480691e-06,
+      "loss": 0.7779,
+      "step": 479
+    },
+    {
+      "epoch": 0.23830209755492118,
+      "grad_norm": 0.13001276373757875,
+      "learning_rate": 4.959684241184379e-06,
+      "loss": 0.7653,
+      "step": 480
+    },
+    {
+      "epoch": 0.2387985602581606,
+      "grad_norm": 0.13464254447792998,
+      "learning_rate": 4.959509239964768e-06,
+      "loss": 0.7932,
+      "step": 481
+    },
+    {
+      "epoch": 0.2392950229614,
+      "grad_norm": 0.13597319212243217,
+      "learning_rate": 4.959333862848605e-06,
+      "loss": 0.7934,
+      "step": 482
+    },
+    {
+      "epoch": 0.23979148566463945,
+      "grad_norm": 0.14655344478202936,
+      "learning_rate": 4.959158109862694e-06,
+      "loss": 0.798,
+      "step": 483
+    },
+    {
+      "epoch": 0.24028794836787887,
+      "grad_norm": 0.1305172241900655,
+      "learning_rate": 4.958981981033895e-06,
+      "loss": 0.7791,
+      "step": 484
+    },
+    {
+      "epoch": 0.24078441107111828,
+      "grad_norm": 0.13605164916015544,
+      "learning_rate": 4.958805476389127e-06,
+      "loss": 0.8128,
+      "step": 485
+    },
+    {
+      "epoch": 0.2412808737743577,
+      "grad_norm": 0.13833174537228857,
+      "learning_rate": 4.958628595955366e-06,
+      "loss": 0.7694,
+      "step": 486
+    },
+    {
+      "epoch": 0.2417773364775971,
+      "grad_norm": 0.1358842375740616,
+      "learning_rate": 4.958451339759645e-06,
+      "loss": 0.7786,
+      "step": 487
+    },
+    {
+      "epoch": 0.24227379918083655,
+      "grad_norm": 0.13943485857687118,
+      "learning_rate": 4.9582737078290556e-06,
+      "loss": 0.8026,
+      "step": 488
+    },
+    {
+      "epoch": 0.24277026188407597,
+      "grad_norm": 0.1341222818235527,
+      "learning_rate": 4.958095700190745e-06,
+      "loss": 0.821,
+      "step": 489
+    },
+    {
+      "epoch": 0.24326672458731538,
+      "grad_norm": 0.13304092780884888,
+      "learning_rate": 4.957917316871919e-06,
+      "loss": 0.8058,
+      "step": 490
+    },
+    {
+      "epoch": 0.2437631872905548,
+      "grad_norm": 0.13446637676436832,
+      "learning_rate": 4.957738557899841e-06,
+      "loss": 0.8165,
+      "step": 491
+    },
+    {
+      "epoch": 0.2442596499937942,
+      "grad_norm": 0.1355754492629554,
+      "learning_rate": 4.9575594233018305e-06,
+      "loss": 0.84,
+      "step": 492
+    },
+    {
+      "epoch": 0.24475611269703362,
+      "grad_norm": 0.1326880486992906,
+      "learning_rate": 4.957379913105267e-06,
+      "loss": 0.7595,
+      "step": 493
+    },
+    {
+      "epoch": 0.24525257540027307,
+      "grad_norm": 0.13614000382079813,
+      "learning_rate": 4.957200027337585e-06,
+      "loss": 0.7988,
+      "step": 494
+    },
+    {
+      "epoch": 0.24574903810351248,
+      "grad_norm": 0.12773189423591907,
+      "learning_rate": 4.957019766026277e-06,
+      "loss": 0.7577,
+      "step": 495
+    },
+    {
+      "epoch": 0.2462455008067519,
+      "grad_norm": 0.14086258045001926,
+      "learning_rate": 4.956839129198892e-06,
+      "loss": 0.8078,
+      "step": 496
+    },
+    {
+      "epoch": 0.2467419635099913,
+      "grad_norm": 0.13555212844444037,
+      "learning_rate": 4.95665811688304e-06,
+      "loss": 0.8094,
+      "step": 497
+    },
+    {
+      "epoch": 0.24723842621323072,
+      "grad_norm": 0.13674333466288918,
+      "learning_rate": 4.9564767291063844e-06,
+      "loss": 0.7807,
+      "step": 498
+    },
+    {
+      "epoch": 0.24773488891647016,
+      "grad_norm": 0.1304786699944242,
+      "learning_rate": 4.956294965896647e-06,
+      "loss": 0.7614,
+      "step": 499
+    },
+    {
+      "epoch": 0.24823135161970958,
+      "grad_norm": 0.13848355027573125,
+      "learning_rate": 4.956112827281607e-06,
+      "loss": 0.8078,
+      "step": 500
+    },
+    {
+      "epoch": 0.248727814322949,
+      "grad_norm": 0.13212981378633934,
+      "learning_rate": 4.955930313289102e-06,
+      "loss": 0.7895,
+      "step": 501
+    },
+    {
+      "epoch": 0.2492242770261884,
+      "grad_norm": 0.134424879179094,
+      "learning_rate": 4.955747423947027e-06,
+      "loss": 0.8235,
+      "step": 502
+    },
+    {
+      "epoch": 0.24972073972942782,
+      "grad_norm": 0.12855995535715545,
+      "learning_rate": 4.955564159283334e-06,
+      "loss": 0.7454,
+      "step": 503
+    },
+    {
+      "epoch": 0.25021720243266726,
+      "grad_norm": 0.13938742116853411,
+      "learning_rate": 4.95538051932603e-06,
+      "loss": 0.7975,
+      "step": 504
+    },
+    {
+      "epoch": 0.25021720243266726,
+      "eval_loss": 0.7929754853248596,
+      "eval_runtime": 135.44,
+      "eval_samples_per_second": 224.107,
+      "eval_steps_per_second": 28.02,
+      "step": 504
+    },
+    {
+      "epoch": 0.25071366513590665,
+      "grad_norm": 0.13609034177916032,
+      "learning_rate": 4.9551965041031835e-06,
+      "loss": 0.7783,
+      "step": 505
+    },
+    {
+      "epoch": 0.2512101278391461,
+      "grad_norm": 0.13139635870746627,
+      "learning_rate": 4.955012113642916e-06,
+      "loss": 0.7706,
+      "step": 506
+    },
+    {
+      "epoch": 0.2517065905423855,
+      "grad_norm": 0.13959142837209643,
+      "learning_rate": 4.954827347973412e-06,
+      "loss": 0.8491,
+      "step": 507
+    },
+    {
+      "epoch": 0.2522030532456249,
+      "grad_norm": 0.15436767440806537,
+      "learning_rate": 4.954642207122907e-06,
+      "loss": 0.8424,
+      "step": 508
+    },
+    {
+      "epoch": 0.25269951594886436,
+      "grad_norm": 0.14047650581011453,
+      "learning_rate": 4.954456691119698e-06,
+      "loss": 0.7782,
+      "step": 509
+    },
+    {
+      "epoch": 0.25319597865210375,
+      "grad_norm": 0.13512624112535368,
+      "learning_rate": 4.954270799992138e-06,
+      "loss": 0.7552,
+      "step": 510
+    },
+    {
+      "epoch": 0.2536924413553432,
+      "grad_norm": 0.13165016338907254,
+      "learning_rate": 4.954084533768637e-06,
+      "loss": 0.7652,
+      "step": 511
+    },
+    {
+      "epoch": 0.2541889040585826,
+      "grad_norm": 0.13432352709536072,
+      "learning_rate": 4.953897892477664e-06,
+      "loss": 0.7549,
+      "step": 512
+    },
+    {
+      "epoch": 0.254685366761822,
+      "grad_norm": 0.13825583852964196,
+      "learning_rate": 4.953710876147743e-06,
+      "loss": 0.7461,
+      "step": 513
+    },
+    {
+      "epoch": 0.25518182946506146,
+      "grad_norm": 0.13800340620252008,
+      "learning_rate": 4.953523484807456e-06,
+      "loss": 0.8118,
+      "step": 514
+    },
+    {
+      "epoch": 0.25567829216830085,
+      "grad_norm": 0.1357923592978122,
+      "learning_rate": 4.9533357184854454e-06,
+      "loss": 0.8104,
+      "step": 515
+    },
+    {
+      "epoch": 0.2561747548715403,
+      "grad_norm": 0.13697526329337478,
+      "learning_rate": 4.953147577210406e-06,
+      "loss": 0.853,
+      "step": 516
+    },
+    {
+      "epoch": 0.2566712175747797,
+      "grad_norm": 0.13197113130506147,
+      "learning_rate": 4.952959061011091e-06,
+      "loss": 0.7424,
+      "step": 517
+    },
+    {
+      "epoch": 0.2571676802780191,
+      "grad_norm": 0.13803258269495886,
+      "learning_rate": 4.952770169916316e-06,
+      "loss": 0.8018,
+      "step": 518
+    },
+    {
+      "epoch": 0.25766414298125856,
+      "grad_norm": 0.13462239799861447,
+      "learning_rate": 4.952580903954946e-06,
+      "loss": 0.7604,
+      "step": 519
+    },
+    {
+      "epoch": 0.25816060568449795,
+      "grad_norm": 0.13185857943402207,
+      "learning_rate": 4.95239126315591e-06,
+      "loss": 0.7809,
+      "step": 520
+    },
+    {
+      "epoch": 0.2586570683877374,
+      "grad_norm": 0.13281620469964295,
+      "learning_rate": 4.95220124754819e-06,
+      "loss": 0.7726,
+      "step": 521
+    },
+    {
+      "epoch": 0.2591535310909768,
+      "grad_norm": 0.14105017206726814,
+      "learning_rate": 4.952010857160828e-06,
+      "loss": 0.785,
+      "step": 522
+    },
+    {
+      "epoch": 0.2596499937942162,
+      "grad_norm": 0.1439691078823198,
+      "learning_rate": 4.951820092022921e-06,
+      "loss": 0.7934,
+      "step": 523
+    },
+    {
+      "epoch": 0.2601464564974556,
+      "grad_norm": 0.1382892013792061,
+      "learning_rate": 4.951628952163625e-06,
+      "loss": 0.7892,
+      "step": 524
+    },
+    {
+      "epoch": 0.26064291920069504,
+      "grad_norm": 0.13861843984858735,
+      "learning_rate": 4.951437437612152e-06,
+      "loss": 0.8235,
+      "step": 525
+    },
+    {
+      "epoch": 0.2611393819039345,
+      "grad_norm": 0.14191335563414534,
+      "learning_rate": 4.951245548397773e-06,
+      "loss": 0.8068,
+      "step": 526
+    },
+    {
+      "epoch": 0.2616358446071739,
+      "grad_norm": 0.13893025334242587,
+      "learning_rate": 4.951053284549815e-06,
+      "loss": 0.7718,
+      "step": 527
+    },
+    {
+      "epoch": 0.2621323073104133,
+      "grad_norm": 0.128500724810349,
+      "learning_rate": 4.950860646097661e-06,
+      "loss": 0.7775,
+      "step": 528
+    },
+    {
+      "epoch": 0.2626287700136527,
+      "grad_norm": 0.14778687944277374,
+      "learning_rate": 4.950667633070755e-06,
+      "loss": 0.8259,
+      "step": 529
+    },
+    {
+      "epoch": 0.26312523271689214,
+      "grad_norm": 0.1482567507544276,
+      "learning_rate": 4.950474245498594e-06,
+      "loss": 0.751,
+      "step": 530
+    },
+    {
+      "epoch": 0.2636216954201316,
+      "grad_norm": 0.14168154296888727,
+      "learning_rate": 4.950280483410735e-06,
+      "loss": 0.7766,
+      "step": 531
+    },
+    {
+      "epoch": 0.26411815812337097,
+      "grad_norm": 0.13937007291923947,
+      "learning_rate": 4.950086346836792e-06,
+      "loss": 0.8215,
+      "step": 532
+    },
+    {
+      "epoch": 0.2646146208266104,
+      "grad_norm": 0.13948603039348756,
+      "learning_rate": 4.949891835806434e-06,
+      "loss": 0.764,
+      "step": 533
+    },
+    {
+      "epoch": 0.2651110835298498,
+      "grad_norm": 0.1398432077501712,
+      "learning_rate": 4.9496969503493905e-06,
+      "loss": 0.7917,
+      "step": 534
+    },
+    {
+      "epoch": 0.26560754623308924,
+      "grad_norm": 0.13864095868027326,
+      "learning_rate": 4.949501690495446e-06,
+      "loss": 0.7852,
+      "step": 535
+    },
+    {
+      "epoch": 0.2661040089363287,
+      "grad_norm": 0.137626101201187,
+      "learning_rate": 4.949306056274443e-06,
+      "loss": 0.7971,
+      "step": 536
+    },
+    {
+      "epoch": 0.26660047163956807,
+      "grad_norm": 0.13991644106960857,
+      "learning_rate": 4.949110047716281e-06,
+      "loss": 0.8049,
+      "step": 537
+    },
+    {
+      "epoch": 0.2670969343428075,
+      "grad_norm": 0.13755972862852978,
+      "learning_rate": 4.948913664850917e-06,
+      "loss": 0.7761,
+      "step": 538
+    },
+    {
+      "epoch": 0.2675933970460469,
+      "grad_norm": 0.14095099190685437,
+      "learning_rate": 4.9487169077083645e-06,
+      "loss": 0.7995,
+      "step": 539
+    },
+    {
+      "epoch": 0.26808985974928634,
+      "grad_norm": 0.13772198796480103,
+      "learning_rate": 4.948519776318694e-06,
+      "loss": 0.775,
+      "step": 540
+    },
+    {
+      "epoch": 0.2685863224525257,
+      "grad_norm": 0.1474365567813288,
+      "learning_rate": 4.948322270712036e-06,
+      "loss": 0.775,
+      "step": 541
+    },
+    {
+      "epoch": 0.26908278515576517,
+      "grad_norm": 0.14740248795105346,
+      "learning_rate": 4.948124390918574e-06,
+      "loss": 0.8823,
+      "step": 542
+    },
+    {
+      "epoch": 0.2695792478590046,
+      "grad_norm": 0.13090799079120452,
+      "learning_rate": 4.947926136968551e-06,
+      "loss": 0.7722,
+      "step": 543
+    },
+    {
+      "epoch": 0.270075710562244,
+      "grad_norm": 0.1398513276367715,
+      "learning_rate": 4.947727508892268e-06,
+      "loss": 0.7622,
+      "step": 544
+    },
+    {
+      "epoch": 0.27057217326548344,
+      "grad_norm": 0.1402504056708596,
+      "learning_rate": 4.947528506720082e-06,
+      "loss": 0.7855,
+      "step": 545
+    },
+    {
+      "epoch": 0.2710686359687228,
+      "grad_norm": 0.13124650948584116,
+      "learning_rate": 4.947329130482407e-06,
+      "loss": 0.738,
+      "step": 546
+    },
+    {
+      "epoch": 0.27156509867196227,
+      "grad_norm": 0.13017696798038783,
+      "learning_rate": 4.947129380209713e-06,
+      "loss": 0.7761,
+      "step": 547
+    },
+    {
+      "epoch": 0.2720615613752017,
+      "grad_norm": 0.14381395142272502,
+      "learning_rate": 4.9469292559325316e-06,
+      "loss": 0.7887,
+      "step": 548
+    },
+    {
+      "epoch": 0.2725580240784411,
+      "grad_norm": 0.13338140216734587,
+      "learning_rate": 4.946728757681446e-06,
+      "loss": 0.8003,
+      "step": 549
+    },
+    {
+      "epoch": 0.27305448678168054,
+      "grad_norm": 0.13560694970854092,
+      "learning_rate": 4.946527885487101e-06,
+      "loss": 0.7377,
+      "step": 550
+    },
+    {
+      "epoch": 0.2735509494849199,
+      "grad_norm": 0.13758839128075348,
+      "learning_rate": 4.946326639380194e-06,
+      "loss": 0.8017,
+      "step": 551
+    },
+    {
+      "epoch": 0.27404741218815937,
+      "grad_norm": 0.13753618486867644,
+      "learning_rate": 4.946125019391486e-06,
+      "loss": 0.7786,
+      "step": 552
+    },
+    {
+      "epoch": 0.2745438748913988,
+      "grad_norm": 0.13936503272255338,
+      "learning_rate": 4.945923025551789e-06,
+      "loss": 0.7771,
+      "step": 553
+    },
+    {
+      "epoch": 0.2750403375946382,
+      "grad_norm": 0.13630518143436557,
+      "learning_rate": 4.945720657891975e-06,
+      "loss": 0.7896,
+      "step": 554
+    },
+    {
+      "epoch": 0.27553680029787764,
+      "grad_norm": 0.13452520454118158,
+      "learning_rate": 4.945517916442971e-06,
+      "loss": 0.82,
+      "step": 555
+    },
+    {
+      "epoch": 0.276033263001117,
+      "grad_norm": 0.15078185753975815,
+      "learning_rate": 4.945314801235766e-06,
+      "loss": 0.7977,
+      "step": 556
+    },
+    {
+      "epoch": 0.27652972570435647,
+      "grad_norm": 0.13571286379008968,
+      "learning_rate": 4.9451113123014e-06,
+      "loss": 0.8183,
+      "step": 557
+    },
+    {
+      "epoch": 0.27702618840759585,
+      "grad_norm": 0.14134176469144435,
+      "learning_rate": 4.9449074496709756e-06,
+      "loss": 0.8391,
+      "step": 558
+    },
+    {
+      "epoch": 0.2775226511108353,
+      "grad_norm": 0.13575461049209747,
+      "learning_rate": 4.944703213375648e-06,
+      "loss": 0.7913,
+      "step": 559
+    },
+    {
+      "epoch": 0.27801911381407474,
+      "grad_norm": 0.13627699697738124,
+      "learning_rate": 4.944498603446633e-06,
+      "loss": 0.7737,
+      "step": 560
+    },
+    {
+      "epoch": 0.2785155765173141,
+      "grad_norm": 0.1436068512980607,
+      "learning_rate": 4.9442936199152e-06,
+      "loss": 0.7665,
+      "step": 561
+    },
+    {
+      "epoch": 0.27901203922055356,
+      "grad_norm": 0.1419116978490336,
+      "learning_rate": 4.944088262812679e-06,
+      "loss": 0.8012,
+      "step": 562
+    },
+    {
+      "epoch": 0.27950850192379295,
+      "grad_norm": 0.14345533439447763,
+      "learning_rate": 4.943882532170454e-06,
+      "loss": 0.7487,
+      "step": 563
+    },
+    {
+      "epoch": 0.2800049646270324,
+      "grad_norm": 0.1482619360446857,
+      "learning_rate": 4.94367642801997e-06,
+      "loss": 0.7929,
+      "step": 564
+    },
+    {
+      "epoch": 0.28050142733027184,
+      "grad_norm": 0.1379616939658588,
+      "learning_rate": 4.943469950392724e-06,
+      "loss": 0.7792,
+      "step": 565
+    },
+    {
+      "epoch": 0.2809978900335112,
+      "grad_norm": 0.14304974445218507,
+      "learning_rate": 4.943263099320275e-06,
+      "loss": 0.7879,
+      "step": 566
+    },
+    {
+      "epoch": 0.28149435273675066,
+      "grad_norm": 0.1428969404097643,
+      "learning_rate": 4.943055874834236e-06,
+      "loss": 0.7217,
+      "step": 567
+    },
+    {
+      "epoch": 0.28199081543999005,
+      "grad_norm": 0.13838491143778672,
+      "learning_rate": 4.942848276966278e-06,
+      "loss": 0.7751,
+      "step": 568
+    },
+    {
+      "epoch": 0.2824872781432295,
+      "grad_norm": 0.13637716635230512,
+      "learning_rate": 4.942640305748128e-06,
+      "loss": 0.7978,
+      "step": 569
+    },
+    {
+      "epoch": 0.28298374084646893,
+      "grad_norm": 0.14002084365223352,
+      "learning_rate": 4.942431961211573e-06,
+      "loss": 0.8041,
+      "step": 570
+    },
+    {
+      "epoch": 0.2834802035497083,
+      "grad_norm": 0.13976480972292757,
+      "learning_rate": 4.942223243388454e-06,
+      "loss": 0.801,
+      "step": 571
+    },
+    {
+      "epoch": 0.28397666625294776,
+      "grad_norm": 0.14031361844161488,
+      "learning_rate": 4.9420141523106705e-06,
+      "loss": 0.7837,
+      "step": 572
+    },
+    {
+      "epoch": 0.28447312895618715,
+      "grad_norm": 0.14188180474957363,
+      "learning_rate": 4.941804688010178e-06,
+      "loss": 0.8129,
+      "step": 573
+    },
+    {
+      "epoch": 0.2849695916594266,
+      "grad_norm": 0.13739560633335732,
+      "learning_rate": 4.941594850518991e-06,
+      "loss": 0.8599,
+      "step": 574
+    },
+    {
+      "epoch": 0.285466054362666,
+      "grad_norm": 0.13161274663378267,
+      "learning_rate": 4.9413846398691775e-06,
+      "loss": 0.7992,
+      "step": 575
+    },
+    {
+      "epoch": 0.2859625170659054,
+      "grad_norm": 0.13571724186099784,
+      "learning_rate": 4.941174056092868e-06,
+      "loss": 0.7426,
+      "step": 576
+    },
+    {
+      "epoch": 0.28645897976914486,
+      "grad_norm": 0.14657992209581697,
+      "learning_rate": 4.940963099222244e-06,
+      "loss": 0.7474,
+      "step": 577
+    },
+    {
+      "epoch": 0.28695544247238425,
+      "grad_norm": 0.14054315894814753,
+      "learning_rate": 4.94075176928955e-06,
+      "loss": 0.8116,
+      "step": 578
+    },
+    {
+      "epoch": 0.2874519051756237,
+      "grad_norm": 0.14667038794719667,
+      "learning_rate": 4.940540066327082e-06,
+      "loss": 0.839,
+      "step": 579
+    },
+    {
+      "epoch": 0.2879483678788631,
+      "grad_norm": 0.13180298446137564,
+      "learning_rate": 4.940327990367196e-06,
+      "loss": 0.7865,
+      "step": 580
+    },
+    {
+      "epoch": 0.2884448305821025,
+      "grad_norm": 0.14864936755923036,
+      "learning_rate": 4.940115541442303e-06,
+      "loss": 0.7855,
+      "step": 581
+    },
+    {
+      "epoch": 0.28894129328534196,
+      "grad_norm": 0.13838905353793282,
+      "learning_rate": 4.939902719584875e-06,
+      "loss": 0.7736,
+      "step": 582
+    },
+    {
+      "epoch": 0.28943775598858135,
+      "grad_norm": 0.13826980328271227,
+      "learning_rate": 4.939689524827436e-06,
+      "loss": 0.7738,
+      "step": 583
+    },
+    {
+      "epoch": 0.2899342186918208,
+      "grad_norm": 0.13562760080033687,
+      "learning_rate": 4.939475957202572e-06,
+      "loss": 0.7803,
+      "step": 584
+    },
+    {
+      "epoch": 0.2904306813950602,
+      "grad_norm": 0.1477129197436419,
+      "learning_rate": 4.939262016742921e-06,
+      "loss": 0.8892,
+      "step": 585
+    },
+    {
+      "epoch": 0.2909271440982996,
+      "grad_norm": 0.13608868262673587,
+      "learning_rate": 4.939047703481182e-06,
+      "loss": 0.7787,
+      "step": 586
+    },
+    {
+      "epoch": 0.29142360680153906,
+      "grad_norm": 0.13532164943155106,
+      "learning_rate": 4.938833017450108e-06,
+      "loss": 0.7919,
+      "step": 587
+    },
+    {
+      "epoch": 0.29192006950477845,
+      "grad_norm": 0.1354528176328873,
+      "learning_rate": 4.938617958682511e-06,
+      "loss": 0.7187,
+      "step": 588
+    },
+    {
+      "epoch": 0.2924165322080179,
+      "grad_norm": 0.14287654528543972,
+      "learning_rate": 4.93840252721126e-06,
+      "loss": 0.7962,
+      "step": 589
+    },
+    {
+      "epoch": 0.2929129949112573,
+      "grad_norm": 0.13500520944212716,
+      "learning_rate": 4.9381867230692795e-06,
+      "loss": 0.7953,
+      "step": 590
+    },
+    {
+      "epoch": 0.2934094576144967,
+      "grad_norm": 0.13701507538430724,
+      "learning_rate": 4.937970546289551e-06,
+      "loss": 0.8274,
+      "step": 591
+    },
+    {
+      "epoch": 0.2939059203177361,
+      "grad_norm": 0.13546590818001467,
+      "learning_rate": 4.937753996905115e-06,
+      "loss": 0.7731,
+      "step": 592
+    },
+    {
+      "epoch": 0.29440238302097554,
+      "grad_norm": 0.13811700535991556,
+      "learning_rate": 4.937537074949067e-06,
+      "loss": 0.7647,
+      "step": 593
+    },
+    {
+      "epoch": 0.294898845724215,
+      "grad_norm": 0.12945879834966295,
+      "learning_rate": 4.937319780454559e-06,
+      "loss": 0.744,
+      "step": 594
+    },
+    {
+      "epoch": 0.2953953084274544,
+      "grad_norm": 0.13498731326365385,
+      "learning_rate": 4.937102113454803e-06,
+      "loss": 0.7759,
+      "step": 595
+    },
+    {
+      "epoch": 0.2958917711306938,
+      "grad_norm": 0.1292963266491359,
+      "learning_rate": 4.936884073983065e-06,
+      "loss": 0.7647,
+      "step": 596
+    },
+    {
+      "epoch": 0.2963882338339332,
+      "grad_norm": 0.1364683596899809,
+      "learning_rate": 4.9366656620726685e-06,
+      "loss": 0.782,
+      "step": 597
+    },
+    {
+      "epoch": 0.29688469653717264,
+      "grad_norm": 0.14354575937871145,
+      "learning_rate": 4.936446877756994e-06,
+      "loss": 0.7965,
+      "step": 598
+    },
+    {
+      "epoch": 0.2973811592404121,
+      "grad_norm": 0.14297046663924878,
+      "learning_rate": 4.936227721069481e-06,
+      "loss": 0.8565,
+      "step": 599
+    },
+    {
+      "epoch": 0.29787762194365147,
+      "grad_norm": 0.13515162331475536,
+      "learning_rate": 4.936008192043621e-06,
+      "loss": 0.7897,
+      "step": 600
+    },
+    {
+      "epoch": 0.2983740846468909,
+      "grad_norm": 0.14366874023945117,
+      "learning_rate": 4.935788290712969e-06,
+      "loss": 0.8526,
+      "step": 601
+    },
+    {
+      "epoch": 0.2988705473501303,
+      "grad_norm": 0.13501986218993212,
+      "learning_rate": 4.935568017111131e-06,
+      "loss": 0.7843,
+      "step": 602
+    },
+    {
+      "epoch": 0.29936701005336974,
+      "grad_norm": 0.13501818290266443,
+      "learning_rate": 4.935347371271772e-06,
+      "loss": 0.785,
+      "step": 603
+    },
+    {
+      "epoch": 0.2998634727566092,
+      "grad_norm": 0.1357480400677297,
+      "learning_rate": 4.9351263532286165e-06,
+      "loss": 0.809,
+      "step": 604
+    },
+    {
+      "epoch": 0.30035993545984857,
+      "grad_norm": 0.13963197307606626,
+      "learning_rate": 4.934904963015442e-06,
+      "loss": 0.7573,
+      "step": 605
+    },
+    {
+      "epoch": 0.300856398163088,
+      "grad_norm": 0.13904388544231225,
+      "learning_rate": 4.934683200666084e-06,
+      "loss": 0.76,
+      "step": 606
+    },
+    {
+      "epoch": 0.3013528608663274,
+      "grad_norm": 0.13273586015038175,
+      "learning_rate": 4.934461066214436e-06,
+      "loss": 0.777,
+      "step": 607
+    },
+    {
+      "epoch": 0.30184932356956684,
+      "grad_norm": 0.1410913256356068,
+      "learning_rate": 4.934238559694448e-06,
+      "loss": 0.8199,
+      "step": 608
+    },
+    {
+      "epoch": 0.3023457862728063,
+      "grad_norm": 0.13415630992181152,
+      "learning_rate": 4.9340156811401265e-06,
+      "loss": 0.8143,
+      "step": 609
+    },
+    {
+      "epoch": 0.30284224897604567,
+      "grad_norm": 0.1350060544192735,
+      "learning_rate": 4.9337924305855335e-06,
+      "loss": 0.7607,
+      "step": 610
+    },
+    {
+      "epoch": 0.3033387116792851,
+      "grad_norm": 0.13671649355566554,
+      "learning_rate": 4.933568808064791e-06,
+      "loss": 0.8062,
+      "step": 611
+    },
+    {
+      "epoch": 0.3038351743825245,
+      "grad_norm": 0.13241711189350516,
+      "learning_rate": 4.933344813612076e-06,
+      "loss": 0.7463,
+      "step": 612
+    },
+    {
+      "epoch": 0.30433163708576394,
+      "grad_norm": 0.145086762190584,
+      "learning_rate": 4.933120447261621e-06,
+      "loss": 0.8496,
+      "step": 613
+    },
+    {
+      "epoch": 0.3048280997890033,
+      "grad_norm": 0.13576833005384956,
+      "learning_rate": 4.932895709047719e-06,
+      "loss": 0.7619,
+      "step": 614
+    },
+    {
+      "epoch": 0.30532456249224277,
+      "grad_norm": 0.13080739610003392,
+      "learning_rate": 4.932670599004715e-06,
+      "loss": 0.7436,
+      "step": 615
+    },
+    {
+      "epoch": 0.3058210251954822,
+      "grad_norm": 0.13931345500666475,
+      "learning_rate": 4.932445117167016e-06,
+      "loss": 0.7806,
+      "step": 616
+    },
+    {
+      "epoch": 0.3063174878987216,
+      "grad_norm": 0.1403456162971794,
+      "learning_rate": 4.932219263569082e-06,
+      "loss": 0.8203,
+      "step": 617
+    },
+    {
+      "epoch": 0.30681395060196104,
+      "grad_norm": 0.14545314884277688,
+      "learning_rate": 4.93199303824543e-06,
+      "loss": 0.7991,
+      "step": 618
+    },
+    {
+      "epoch": 0.3073104133052004,
+      "grad_norm": 0.13361188513192082,
+      "learning_rate": 4.931766441230637e-06,
+      "loss": 0.7669,
+      "step": 619
+    },
+    {
+      "epoch": 0.30780687600843987,
+      "grad_norm": 0.14211532439526364,
+      "learning_rate": 4.931539472559335e-06,
+      "loss": 0.8197,
+      "step": 620
+    },
+    {
+      "epoch": 0.3083033387116793,
+      "grad_norm": 0.1384426265495238,
+      "learning_rate": 4.93131213226621e-06,
+      "loss": 0.7982,
+      "step": 621
+    },
+    {
+      "epoch": 0.3087998014149187,
+      "grad_norm": 0.13479886625937776,
+      "learning_rate": 4.931084420386009e-06,
+      "loss": 0.7696,
+      "step": 622
+    },
+    {
+      "epoch": 0.30929626411815814,
+      "grad_norm": 0.13591593582520675,
+      "learning_rate": 4.9308563369535335e-06,
+      "loss": 0.8021,
+      "step": 623
+    },
+    {
+      "epoch": 0.3097927268213975,
+      "grad_norm": 0.13924380730728358,
+      "learning_rate": 4.930627882003644e-06,
+      "loss": 0.8394,
+      "step": 624
+    },
+    {
+      "epoch": 0.31028918952463697,
+      "grad_norm": 0.13692960315786307,
+      "learning_rate": 4.930399055571253e-06,
+      "loss": 0.7905,
+      "step": 625
+    },
+    {
+      "epoch": 0.3107856522278764,
+      "grad_norm": 0.1339979894901406,
+      "learning_rate": 4.930169857691336e-06,
+      "loss": 0.7881,
+      "step": 626
+    },
+    {
+      "epoch": 0.3112821149311158,
+      "grad_norm": 0.1367431893133873,
+      "learning_rate": 4.929940288398921e-06,
+      "loss": 0.7887,
+      "step": 627
+    },
+    {
+      "epoch": 0.31177857763435524,
+      "grad_norm": 0.13841200775795232,
+      "learning_rate": 4.929710347729094e-06,
+      "loss": 0.7653,
+      "step": 628
+    },
+    {
+      "epoch": 0.3122750403375946,
+      "grad_norm": 0.13600960600331674,
+      "learning_rate": 4.929480035716997e-06,
+      "loss": 0.7811,
+      "step": 629
+    },
+    {
+      "epoch": 0.31277150304083406,
+      "grad_norm": 0.13653488024445395,
+      "learning_rate": 4.9292493523978315e-06,
+      "loss": 0.7919,
+      "step": 630
+    },
+    {
+      "epoch": 0.31326796574407345,
+      "grad_norm": 0.13176835024293457,
+      "learning_rate": 4.929018297806852e-06,
+      "loss": 0.7634,
+      "step": 631
+    },
+    {
+      "epoch": 0.3137644284473129,
+      "grad_norm": 0.1469495293000222,
+      "learning_rate": 4.928786871979372e-06,
+      "loss": 0.8497,
+      "step": 632
+    },
+    {
+      "epoch": 0.31426089115055234,
+      "grad_norm": 0.1431964358331162,
+      "learning_rate": 4.928555074950761e-06,
+      "loss": 0.8058,
+      "step": 633
+    },
+    {
+      "epoch": 0.3147573538537917,
+      "grad_norm": 0.13850195286656924,
+      "learning_rate": 4.928322906756446e-06,
+      "loss": 0.7735,
+      "step": 634
+    },
+    {
+      "epoch": 0.31525381655703116,
+      "grad_norm": 0.13463773472502238,
+      "learning_rate": 4.92809036743191e-06,
+      "loss": 0.7184,
+      "step": 635
+    },
+    {
+      "epoch": 0.31575027926027055,
+      "grad_norm": 0.1364605714313478,
+      "learning_rate": 4.927857457012693e-06,
+      "loss": 0.7585,
+      "step": 636
+    },
+    {
+      "epoch": 0.31624674196351,
+      "grad_norm": 0.1366334427972066,
+      "learning_rate": 4.927624175534391e-06,
+      "loss": 0.7739,
+      "step": 637
+    },
+    {
+      "epoch": 0.31674320466674943,
+      "grad_norm": 0.13226666489304473,
+      "learning_rate": 4.927390523032658e-06,
+      "loss": 0.7801,
+      "step": 638
+    },
+    {
+      "epoch": 0.3172396673699888,
+      "grad_norm": 0.1321397495808711,
+      "learning_rate": 4.927156499543203e-06,
+      "loss": 0.7691,
+      "step": 639
+    },
+    {
+      "epoch": 0.31773613007322826,
+      "grad_norm": 0.14417418984304473,
+      "learning_rate": 4.926922105101795e-06,
+      "loss": 0.8335,
+      "step": 640
+    },
+    {
+      "epoch": 0.31823259277646765,
+      "grad_norm": 0.1373833193656603,
+      "learning_rate": 4.926687339744255e-06,
+      "loss": 0.8057,
+      "step": 641
+    },
+    {
+      "epoch": 0.3187290554797071,
+      "grad_norm": 0.13780220647272026,
+      "learning_rate": 4.926452203506464e-06,
+      "loss": 0.8188,
+      "step": 642
+    },
+    {
+      "epoch": 0.31922551818294653,
+      "grad_norm": 0.1304815912554088,
+      "learning_rate": 4.926216696424359e-06,
+      "loss": 0.7549,
+      "step": 643
+    },
+    {
+      "epoch": 0.3197219808861859,
+      "grad_norm": 0.14230394590112047,
+      "learning_rate": 4.9259808185339344e-06,
+      "loss": 0.8688,
+      "step": 644
+    },
+    {
+      "epoch": 0.32021844358942536,
+      "grad_norm": 0.13197144711432499,
+      "learning_rate": 4.925744569871238e-06,
+      "loss": 0.7638,
+      "step": 645
+    },
+    {
+      "epoch": 0.32071490629266475,
+      "grad_norm": 0.1377381360289958,
+      "learning_rate": 4.925507950472378e-06,
+      "loss": 0.7865,
+      "step": 646
+    },
+    {
+      "epoch": 0.3212113689959042,
+      "grad_norm": 0.14064672696348562,
+      "learning_rate": 4.9252709603735184e-06,
+      "loss": 0.8655,
+      "step": 647
+    },
+    {
+      "epoch": 0.3217078316991436,
+      "grad_norm": 0.17099346992203562,
+      "learning_rate": 4.925033599610879e-06,
+      "loss": 0.7557,
+      "step": 648
+    },
+    {
+      "epoch": 0.322204294402383,
+      "grad_norm": 0.13649366857460205,
+      "learning_rate": 4.9247958682207365e-06,
+      "loss": 0.8075,
+      "step": 649
+    },
+    {
+      "epoch": 0.32270075710562246,
+      "grad_norm": 0.13376375859315848,
+      "learning_rate": 4.924557766239424e-06,
+      "loss": 0.8125,
+      "step": 650
+    },
+    {
+      "epoch": 0.32319721980886185,
+      "grad_norm": 0.1395328981536072,
+      "learning_rate": 4.9243192937033304e-06,
+      "loss": 0.7883,
+      "step": 651
+    },
+    {
+      "epoch": 0.3236936825121013,
+      "grad_norm": 0.13819391735551326,
+      "learning_rate": 4.924080450648905e-06,
+      "loss": 0.8368,
+      "step": 652
+    },
+    {
+      "epoch": 0.3241901452153407,
+      "grad_norm": 0.13868224090210812,
+      "learning_rate": 4.92384123711265e-06,
+      "loss": 0.7756,
+      "step": 653
+    },
+    {
+      "epoch": 0.3246866079185801,
+      "grad_norm": 0.13995057823515514,
+      "learning_rate": 4.923601653131125e-06,
+      "loss": 0.7953,
+      "step": 654
+    },
+    {
+      "epoch": 0.32518307062181956,
+      "grad_norm": 0.14384219577990262,
+      "learning_rate": 4.923361698740946e-06,
+      "loss": 0.7864,
+      "step": 655
+    },
+    {
+      "epoch": 0.32567953332505895,
+      "grad_norm": 0.1359832308842288,
+      "learning_rate": 4.923121373978789e-06,
+      "loss": 0.7315,
+      "step": 656
+    },
+    {
+      "epoch": 0.3261759960282984,
+      "grad_norm": 0.1332816844154388,
+      "learning_rate": 4.92288067888138e-06,
+      "loss": 0.771,
+      "step": 657
+    },
+    {
+      "epoch": 0.3266724587315378,
+      "grad_norm": 0.13971086373297162,
+      "learning_rate": 4.922639613485508e-06,
+      "loss": 0.7904,
+      "step": 658
+    },
+    {
+      "epoch": 0.3271689214347772,
+      "grad_norm": 0.13132115501434735,
+      "learning_rate": 4.922398177828015e-06,
+      "loss": 0.7416,
+      "step": 659
+    },
+    {
+      "epoch": 0.32766538413801666,
+      "grad_norm": 0.13992347458321633,
+      "learning_rate": 4.9221563719458e-06,
+      "loss": 0.814,
+      "step": 660
+    },
+    {
+      "epoch": 0.32816184684125604,
+      "grad_norm": 0.13633398687640888,
+      "learning_rate": 4.921914195875821e-06,
+      "loss": 0.7611,
+      "step": 661
+    },
+    {
+      "epoch": 0.3286583095444955,
+      "grad_norm": 0.14882463890187234,
+      "learning_rate": 4.921671649655088e-06,
+      "loss": 0.7434,
+      "step": 662
+    },
+    {
+      "epoch": 0.3291547722477349,
+      "grad_norm": 0.13772949607044616,
+      "learning_rate": 4.921428733320674e-06,
+      "loss": 0.7717,
+      "step": 663
+    },
+    {
+      "epoch": 0.3296512349509743,
+      "grad_norm": 0.1414530607047641,
+      "learning_rate": 4.921185446909702e-06,
+      "loss": 0.782,
+      "step": 664
+    },
+    {
+      "epoch": 0.3301476976542137,
+      "grad_norm": 0.14471648591939731,
+      "learning_rate": 4.920941790459355e-06,
+      "loss": 0.7662,
+      "step": 665
+    },
+    {
+      "epoch": 0.33064416035745314,
+      "grad_norm": 0.14485133167971112,
+      "learning_rate": 4.920697764006872e-06,
+      "loss": 0.8213,
+      "step": 666
+    },
+    {
+      "epoch": 0.3311406230606926,
+      "grad_norm": 0.14192892706466312,
+      "learning_rate": 4.920453367589548e-06,
+      "loss": 0.8613,
+      "step": 667
+    },
+    {
+      "epoch": 0.33163708576393197,
+      "grad_norm": 0.14207950507096373,
+      "learning_rate": 4.920208601244737e-06,
+      "loss": 0.826,
+      "step": 668
+    },
+    {
+      "epoch": 0.3321335484671714,
+      "grad_norm": 0.14408452937358693,
+      "learning_rate": 4.919963465009846e-06,
+      "loss": 0.7818,
+      "step": 669
+    },
+    {
+      "epoch": 0.3326300111704108,
+      "grad_norm": 0.13736006931932224,
+      "learning_rate": 4.919717958922341e-06,
+      "loss": 0.7589,
+      "step": 670
+    },
+    {
+      "epoch": 0.33312647387365024,
+      "grad_norm": 0.14298168063992314,
+      "learning_rate": 4.919472083019743e-06,
+      "loss": 0.8034,
+      "step": 671
+    },
+    {
+      "epoch": 0.3336229365768897,
+      "grad_norm": 0.1376542857034229,
+      "learning_rate": 4.91922583733963e-06,
+      "loss": 0.7863,
+      "step": 672
+    },
+    {
+      "epoch": 0.33411939928012907,
+      "grad_norm": 0.1390010637369658,
+      "learning_rate": 4.918979221919637e-06,
+      "loss": 0.842,
+      "step": 673
+    },
+    {
+      "epoch": 0.3346158619833685,
+      "grad_norm": 0.14484040584209332,
+      "learning_rate": 4.918732236797456e-06,
+      "loss": 0.7629,
+      "step": 674
+    },
+    {
+      "epoch": 0.3351123246866079,
+      "grad_norm": 0.13890387799381465,
+      "learning_rate": 4.918484882010833e-06,
+      "loss": 0.778,
+      "step": 675
+    },
+    {
+      "epoch": 0.33560878738984734,
+      "grad_norm": 0.13863969701935344,
+      "learning_rate": 4.918237157597574e-06,
+      "loss": 0.8279,
+      "step": 676
+    },
+    {
+      "epoch": 0.3361052500930868,
+      "grad_norm": 0.1426640315283534,
+      "learning_rate": 4.917989063595539e-06,
+      "loss": 0.8762,
+      "step": 677
+    },
+    {
+      "epoch": 0.33660171279632617,
+      "grad_norm": 0.13500863262637763,
+      "learning_rate": 4.917740600042645e-06,
+      "loss": 0.7536,
+      "step": 678
+    },
+    {
+      "epoch": 0.3370981754995656,
+      "grad_norm": 0.12795948317244496,
+      "learning_rate": 4.917491766976865e-06,
+      "loss": 0.7169,
+      "step": 679
+    },
+    {
+      "epoch": 0.337594638202805,
+      "grad_norm": 0.13710750398142116,
+      "learning_rate": 4.917242564436231e-06,
+      "loss": 0.738,
+      "step": 680
+    },
+    {
+      "epoch": 0.33809110090604444,
+      "grad_norm": 0.13190657220996582,
+      "learning_rate": 4.916992992458828e-06,
+      "loss": 0.7693,
+      "step": 681
+    },
+    {
+      "epoch": 0.3385875636092838,
+      "grad_norm": 0.13830630412972159,
+      "learning_rate": 4.9167430510828e-06,
+      "loss": 0.8153,
+      "step": 682
+    },
+    {
+      "epoch": 0.33908402631252327,
+      "grad_norm": 0.13304884196362807,
+      "learning_rate": 4.916492740346346e-06,
+      "loss": 0.7945,
+      "step": 683
+    },
+    {
+      "epoch": 0.3395804890157627,
+      "grad_norm": 0.12774428215124484,
+      "learning_rate": 4.916242060287723e-06,
+      "loss": 0.7052,
+      "step": 684
+    },
+    {
+      "epoch": 0.3400769517190021,
+      "grad_norm": 0.1265371082556476,
+      "learning_rate": 4.9159910109452416e-06,
+      "loss": 0.7483,
+      "step": 685
+    },
+    {
+      "epoch": 0.34057341442224154,
+      "grad_norm": 0.13343444419954362,
+      "learning_rate": 4.9157395923572716e-06,
+      "loss": 0.7653,
+      "step": 686
+    },
+    {
+      "epoch": 0.3410698771254809,
+      "grad_norm": 0.13166700692340655,
+      "learning_rate": 4.9154878045622385e-06,
+      "loss": 0.7587,
+      "step": 687
+    },
+    {
+      "epoch": 0.34156633982872037,
+      "grad_norm": 0.14076974443956647,
+      "learning_rate": 4.915235647598624e-06,
+      "loss": 0.7912,
+      "step": 688
+    },
+    {
+      "epoch": 0.3420628025319598,
+      "grad_norm": 0.12977776060064555,
+      "learning_rate": 4.914983121504966e-06,
+      "loss": 0.7723,
+      "step": 689
+    },
+    {
+      "epoch": 0.3425592652351992,
+      "grad_norm": 0.13812993400540244,
+      "learning_rate": 4.914730226319859e-06,
+      "loss": 0.7968,
+      "step": 690
+    },
+    {
+      "epoch": 0.34305572793843864,
+      "grad_norm": 0.14598669976661624,
+      "learning_rate": 4.914476962081954e-06,
+      "loss": 0.8009,
+      "step": 691
+    },
+    {
+      "epoch": 0.343552190641678,
+      "grad_norm": 0.13314085383077542,
+      "learning_rate": 4.9142233288299595e-06,
+      "loss": 0.8293,
+      "step": 692
+    },
+    {
+      "epoch": 0.34404865334491747,
+      "grad_norm": 0.14232577807503202,
+      "learning_rate": 4.9139693266026375e-06,
+      "loss": 0.7755,
+      "step": 693
+    },
+    {
+      "epoch": 0.3445451160481569,
+      "grad_norm": 0.13552034485159273,
+      "learning_rate": 4.91371495543881e-06,
+      "loss": 0.8046,
+      "step": 694
+    },
+    {
+      "epoch": 0.3450415787513963,
+      "grad_norm": 0.13920010636352773,
+      "learning_rate": 4.913460215377351e-06,
+      "loss": 0.7898,
+      "step": 695
+    },
+    {
+      "epoch": 0.34553804145463574,
+      "grad_norm": 0.1365082779225381,
+      "learning_rate": 4.9132051064571965e-06,
+      "loss": 0.7949,
+      "step": 696
+    },
+    {
+      "epoch": 0.3460345041578751,
+      "grad_norm": 0.13329039576470306,
+      "learning_rate": 4.912949628717334e-06,
+      "loss": 0.7691,
+      "step": 697
+    },
+    {
+      "epoch": 0.34653096686111456,
+      "grad_norm": 0.13821152204427162,
+      "learning_rate": 4.912693782196808e-06,
+      "loss": 0.782,
+      "step": 698
+    },
+    {
+      "epoch": 0.34702742956435395,
+      "grad_norm": 0.1351784114090911,
+      "learning_rate": 4.912437566934724e-06,
+      "loss": 0.7467,
+      "step": 699
+    },
+    {
+      "epoch": 0.3475238922675934,
+      "grad_norm": 0.13348674748078376,
+      "learning_rate": 4.912180982970237e-06,
+      "loss": 0.7762,
+      "step": 700
+    },
+    {
+      "epoch": 0.34802035497083283,
+      "grad_norm": 0.12985538825056495,
+      "learning_rate": 4.911924030342563e-06,
+      "loss": 0.7902,
+      "step": 701
+    },
+    {
+      "epoch": 0.3485168176740722,
+      "grad_norm": 0.1368260578507066,
+      "learning_rate": 4.911666709090974e-06,
+      "loss": 0.7878,
+      "step": 702
+    },
+    {
+      "epoch": 0.34901328037731166,
+      "grad_norm": 0.1360969813881036,
+      "learning_rate": 4.911409019254797e-06,
+      "loss": 0.7653,
+      "step": 703
+    },
+    {
+      "epoch": 0.34950974308055105,
+      "grad_norm": 0.13601859820292503,
+      "learning_rate": 4.911150960873414e-06,
+      "loss": 0.7416,
+      "step": 704
+    },
+    {
+      "epoch": 0.3500062057837905,
+      "grad_norm": 0.13523409685451968,
+      "learning_rate": 4.910892533986268e-06,
+      "loss": 0.7683,
+      "step": 705
+    },
+    {
+      "epoch": 0.35050266848702993,
+      "grad_norm": 0.1307299504782071,
+      "learning_rate": 4.9106337386328524e-06,
+      "loss": 0.7706,
+      "step": 706
+    },
+    {
+      "epoch": 0.3509991311902693,
+      "grad_norm": 0.1324922818188745,
+      "learning_rate": 4.910374574852722e-06,
+      "loss": 0.7753,
+      "step": 707
+    },
+    {
+      "epoch": 0.35149559389350876,
+      "grad_norm": 0.13283375926300728,
+      "learning_rate": 4.910115042685486e-06,
+      "loss": 0.7988,
+      "step": 708
+    },
+    {
+      "epoch": 0.35199205659674815,
+      "grad_norm": 0.1364833655942737,
+      "learning_rate": 4.909855142170809e-06,
+      "loss": 0.7662,
+      "step": 709
+    },
+    {
+      "epoch": 0.3524885192999876,
+      "grad_norm": 0.13240082659711724,
+      "learning_rate": 4.909594873348412e-06,
+      "loss": 0.7692,
+      "step": 710
+    },
+    {
+      "epoch": 0.35298498200322703,
+      "grad_norm": 0.13859318485445407,
+      "learning_rate": 4.909334236258073e-06,
+      "loss": 0.8092,
+      "step": 711
+    },
+    {
+      "epoch": 0.3534814447064664,
+      "grad_norm": 0.14669171230394829,
+      "learning_rate": 4.909073230939628e-06,
+      "loss": 0.849,
+      "step": 712
+    },
+    {
+      "epoch": 0.35397790740970586,
+      "grad_norm": 0.1329208698752514,
+      "learning_rate": 4.908811857432966e-06,
+      "loss": 0.8112,
+      "step": 713
+    },
+    {
+      "epoch": 0.35447437011294525,
+      "grad_norm": 0.13420771737338105,
+      "learning_rate": 4.908550115778032e-06,
+      "loss": 0.8081,
+      "step": 714
+    },
+    {
+      "epoch": 0.3549708328161847,
+      "grad_norm": 0.13869844814379267,
+      "learning_rate": 4.908288006014833e-06,
+      "loss": 0.7924,
+      "step": 715
+    },
+    {
+      "epoch": 0.3554672955194241,
+      "grad_norm": 0.13159030247433556,
+      "learning_rate": 4.9080255281834255e-06,
+      "loss": 0.7896,
+      "step": 716
+    },
+    {
+      "epoch": 0.3559637582226635,
+      "grad_norm": 0.12782379383553066,
+      "learning_rate": 4.907762682323926e-06,
+      "loss": 0.7492,
+      "step": 717
+    },
+    {
+      "epoch": 0.35646022092590296,
+      "grad_norm": 0.12984107500991426,
+      "learning_rate": 4.907499468476506e-06,
+      "loss": 0.7713,
+      "step": 718
+    },
+    {
+      "epoch": 0.35695668362914235,
+      "grad_norm": 0.13563855815235779,
+      "learning_rate": 4.907235886681394e-06,
+      "loss": 0.7198,
+      "step": 719
+    },
+    {
+      "epoch": 0.3574531463323818,
+      "grad_norm": 0.1410105313395293,
+      "learning_rate": 4.906971936978874e-06,
+      "loss": 0.7584,
+      "step": 720
+    },
+    {
+      "epoch": 0.3579496090356212,
+      "grad_norm": 0.13491737996113276,
+      "learning_rate": 4.906707619409285e-06,
+      "loss": 0.8048,
+      "step": 721
+    },
+    {
+      "epoch": 0.3584460717388606,
+      "grad_norm": 0.15901372063337899,
+      "learning_rate": 4.906442934013026e-06,
+      "loss": 0.8157,
+      "step": 722
+    },
+    {
+      "epoch": 0.35894253444210006,
+      "grad_norm": 0.1318933579561297,
+      "learning_rate": 4.906177880830548e-06,
+      "loss": 0.7624,
+      "step": 723
+    },
+    {
+      "epoch": 0.35943899714533945,
+      "grad_norm": 0.13581410637368269,
+      "learning_rate": 4.905912459902362e-06,
+      "loss": 0.8205,
+      "step": 724
+    },
+    {
+      "epoch": 0.3599354598485789,
+      "grad_norm": 0.1370308459267444,
+      "learning_rate": 4.905646671269032e-06,
+      "loss": 0.8298,
+      "step": 725
+    },
+    {
+      "epoch": 0.3604319225518183,
+      "grad_norm": 0.1358006289221788,
+      "learning_rate": 4.90538051497118e-06,
+      "loss": 0.8233,
+      "step": 726
+    },
+    {
+      "epoch": 0.3609283852550577,
+      "grad_norm": 0.1316316195921008,
+      "learning_rate": 4.905113991049484e-06,
+      "loss": 0.8134,
+      "step": 727
+    },
+    {
+      "epoch": 0.36142484795829716,
+      "grad_norm": 0.1335957967976374,
+      "learning_rate": 4.904847099544676e-06,
+      "loss": 0.7753,
+      "step": 728
+    },
+    {
+      "epoch": 0.36192131066153654,
+      "grad_norm": 0.13467968539337932,
+      "learning_rate": 4.904579840497549e-06,
+      "loss": 0.7678,
+      "step": 729
+    },
+    {
+      "epoch": 0.362417773364776,
+      "grad_norm": 0.13328087348761786,
+      "learning_rate": 4.904312213948948e-06,
+      "loss": 0.7422,
+      "step": 730
+    },
+    {
+      "epoch": 0.3629142360680154,
+      "grad_norm": 0.13791325862150497,
+      "learning_rate": 4.904044219939775e-06,
+      "loss": 0.81,
+      "step": 731
+    },
+    {
+      "epoch": 0.3634106987712548,
+      "grad_norm": 0.14664569154055215,
+      "learning_rate": 4.9037758585109886e-06,
+      "loss": 0.7829,
+      "step": 732
+    },
+    {
+      "epoch": 0.36390716147449426,
+      "grad_norm": 0.13819662122726,
+      "learning_rate": 4.9035071297036045e-06,
+      "loss": 0.7533,
+      "step": 733
+    },
+    {
+      "epoch": 0.36440362417773364,
+      "grad_norm": 0.14379447561879494,
+      "learning_rate": 4.903238033558692e-06,
+      "loss": 0.7812,
+      "step": 734
+    },
+    {
+      "epoch": 0.3649000868809731,
+      "grad_norm": 0.1348693038250325,
+      "learning_rate": 4.90296857011738e-06,
+      "loss": 0.7406,
+      "step": 735
+    },
+    {
+      "epoch": 0.36539654958421247,
+      "grad_norm": 0.13919538306325066,
+      "learning_rate": 4.90269873942085e-06,
+      "loss": 0.7716,
+      "step": 736
+    },
+    {
+      "epoch": 0.3658930122874519,
+      "grad_norm": 0.13768912705501787,
+      "learning_rate": 4.902428541510342e-06,
+      "loss": 0.7797,
+      "step": 737
+    },
+    {
+      "epoch": 0.3663894749906913,
+      "grad_norm": 0.13652014063796708,
+      "learning_rate": 4.902157976427152e-06,
+      "loss": 0.7549,
+      "step": 738
+    },
+    {
+      "epoch": 0.36688593769393074,
+      "grad_norm": 0.13524592628689325,
+      "learning_rate": 4.901887044212631e-06,
+      "loss": 0.7481,
+      "step": 739
+    },
+    {
+      "epoch": 0.3673824003971702,
+      "grad_norm": 0.1311044839852732,
+      "learning_rate": 4.9016157449081855e-06,
+      "loss": 0.8021,
+      "step": 740
+    },
+    {
+      "epoch": 0.36787886310040957,
+      "grad_norm": 0.13954525535721593,
+      "learning_rate": 4.901344078555282e-06,
+      "loss": 0.7548,
+      "step": 741
+    },
+    {
+      "epoch": 0.368375325803649,
+      "grad_norm": 0.136268642834534,
+      "learning_rate": 4.901072045195437e-06,
+      "loss": 0.7602,
+      "step": 742
+    },
+    {
+      "epoch": 0.3688717885068884,
+      "grad_norm": 0.23363764465279754,
+      "learning_rate": 4.90079964487023e-06,
+      "loss": 0.776,
+      "step": 743
+    },
+    {
+      "epoch": 0.36936825121012784,
+      "grad_norm": 0.13274057394053615,
+      "learning_rate": 4.90052687762129e-06,
+      "loss": 0.7351,
+      "step": 744
+    },
+    {
+      "epoch": 0.3698647139133673,
+      "grad_norm": 0.14127484700099766,
+      "learning_rate": 4.900253743490307e-06,
+      "loss": 0.7798,
+      "step": 745
+    },
+    {
+      "epoch": 0.37036117661660667,
+      "grad_norm": 0.12815346443371342,
+      "learning_rate": 4.8999802425190235e-06,
+      "loss": 0.7075,
+      "step": 746
+    },
+    {
+      "epoch": 0.3708576393198461,
+      "grad_norm": 0.1368536898030887,
+      "learning_rate": 4.899706374749242e-06,
+      "loss": 0.757,
+      "step": 747
+    },
+    {
+      "epoch": 0.3713541020230855,
+      "grad_norm": 0.133763511827003,
+      "learning_rate": 4.899432140222816e-06,
+      "loss": 0.807,
+      "step": 748
+    },
+    {
+      "epoch": 0.37185056472632494,
+      "grad_norm": 0.14086274715481756,
+      "learning_rate": 4.899157538981661e-06,
+      "loss": 0.7545,
+      "step": 749
+    },
+    {
+      "epoch": 0.3723470274295644,
+      "grad_norm": 0.13940663877796536,
+      "learning_rate": 4.898882571067742e-06,
+      "loss": 0.7895,
+      "step": 750
+    },
+    {
+      "epoch": 0.37284349013280377,
+      "grad_norm": 0.14364436254867233,
+      "learning_rate": 4.898607236523086e-06,
+      "loss": 0.824,
+      "step": 751
+    },
+    {
+      "epoch": 0.3733399528360432,
+      "grad_norm": 0.1363424647021028,
+      "learning_rate": 4.898331535389772e-06,
+      "loss": 0.7689,
+      "step": 752
+    },
+    {
+      "epoch": 0.3738364155392826,
+      "grad_norm": 0.134333700914737,
+      "learning_rate": 4.898055467709938e-06,
+      "loss": 0.781,
+      "step": 753
+    },
+    {
+      "epoch": 0.37433287824252204,
+      "grad_norm": 0.1354244860930839,
+      "learning_rate": 4.897779033525775e-06,
+      "loss": 0.7425,
+      "step": 754
+    },
+    {
+      "epoch": 0.3748293409457614,
+      "grad_norm": 0.13641351699049295,
+      "learning_rate": 4.897502232879533e-06,
+      "loss": 0.8106,
+      "step": 755
+    },
+    {
+      "epoch": 0.37532580364900087,
+      "grad_norm": 0.1397668428766841,
+      "learning_rate": 4.897225065813515e-06,
+      "loss": 0.7725,
+      "step": 756
+    },
+    {
+      "epoch": 0.3758222663522403,
+      "grad_norm": 0.1388989852014346,
+      "learning_rate": 4.896947532370083e-06,
+      "loss": 0.8095,
+      "step": 757
+    },
+    {
+      "epoch": 0.3763187290554797,
+      "grad_norm": 0.13452943247011598,
+      "learning_rate": 4.896669632591652e-06,
+      "loss": 0.7768,
+      "step": 758
+    },
+    {
+      "epoch": 0.37681519175871914,
+      "grad_norm": 0.13898005901785496,
+      "learning_rate": 4.896391366520695e-06,
+      "loss": 0.8469,
+      "step": 759
+    },
+    {
+      "epoch": 0.3773116544619585,
+      "grad_norm": 0.1344241818366532,
+      "learning_rate": 4.8961127341997425e-06,
+      "loss": 0.7871,
+      "step": 760
+    },
+    {
+      "epoch": 0.37780811716519797,
+      "grad_norm": 0.13570943105691186,
+      "learning_rate": 4.895833735671376e-06,
+      "loss": 0.7702,
+      "step": 761
+    },
+    {
+      "epoch": 0.3783045798684374,
+      "grad_norm": 0.1341854190460997,
+      "learning_rate": 4.895554370978238e-06,
+      "loss": 0.7409,
+      "step": 762
+    },
+    {
+      "epoch": 0.3788010425716768,
+      "grad_norm": 0.13050283705876012,
+      "learning_rate": 4.895274640163023e-06,
+      "loss": 0.7793,
+      "step": 763
+    },
+    {
+      "epoch": 0.37929750527491624,
+      "grad_norm": 0.1329571983648645,
+      "learning_rate": 4.894994543268486e-06,
+      "loss": 0.7539,
+      "step": 764
+    },
+    {
+      "epoch": 0.3797939679781556,
+      "grad_norm": 0.14453559821774697,
+      "learning_rate": 4.894714080337433e-06,
+      "loss": 0.8057,
+      "step": 765
+    },
+    {
+      "epoch": 0.38029043068139506,
+      "grad_norm": 0.14541141571134905,
+      "learning_rate": 4.894433251412729e-06,
+      "loss": 0.8173,
+      "step": 766
+    },
+    {
+      "epoch": 0.3807868933846345,
+      "grad_norm": 0.1365214842205778,
+      "learning_rate": 4.894152056537295e-06,
+      "loss": 0.7515,
+      "step": 767
+    },
+    {
+      "epoch": 0.3812833560878739,
+      "grad_norm": 0.13104231373031291,
+      "learning_rate": 4.893870495754106e-06,
+      "loss": 0.7388,
+      "step": 768
+    },
+    {
+      "epoch": 0.38177981879111333,
+      "grad_norm": 0.1293760148802387,
+      "learning_rate": 4.8935885691061955e-06,
+      "loss": 0.7202,
+      "step": 769
+    },
+    {
+      "epoch": 0.3822762814943527,
+      "grad_norm": 0.1413055453494607,
+      "learning_rate": 4.893306276636649e-06,
+      "loss": 0.7529,
+      "step": 770
+    },
+    {
+      "epoch": 0.38277274419759216,
+      "grad_norm": 0.13778756966339015,
+      "learning_rate": 4.893023618388612e-06,
+      "loss": 0.7854,
+      "step": 771
+    },
+    {
+      "epoch": 0.38326920690083155,
+      "grad_norm": 0.12932306420294923,
+      "learning_rate": 4.892740594405285e-06,
+      "loss": 0.7832,
+      "step": 772
+    },
+    {
+      "epoch": 0.383765669604071,
+      "grad_norm": 0.13397702650716775,
+      "learning_rate": 4.892457204729923e-06,
+      "loss": 0.8013,
+      "step": 773
+    },
+    {
+      "epoch": 0.38426213230731043,
+      "grad_norm": 0.14230173874593646,
+      "learning_rate": 4.892173449405837e-06,
+      "loss": 0.749,
+      "step": 774
+    },
+    {
+      "epoch": 0.3847585950105498,
+      "grad_norm": 0.13494471486274012,
+      "learning_rate": 4.891889328476395e-06,
+      "loss": 0.7644,
+      "step": 775
+    },
+    {
+      "epoch": 0.38525505771378926,
+      "grad_norm": 0.13174879824024627,
+      "learning_rate": 4.89160484198502e-06,
+      "loss": 0.7659,
+      "step": 776
+    },
+    {
+      "epoch": 0.38575152041702865,
+      "grad_norm": 0.1340875139550934,
+      "learning_rate": 4.891319989975191e-06,
+      "loss": 0.8022,
+      "step": 777
+    },
+    {
+      "epoch": 0.3862479831202681,
+      "grad_norm": 0.1317344878554217,
+      "learning_rate": 4.891034772490444e-06,
+      "loss": 0.7821,
+      "step": 778
+    },
+    {
+      "epoch": 0.38674444582350753,
+      "grad_norm": 0.12863325352249577,
+      "learning_rate": 4.890749189574369e-06,
+      "loss": 0.729,
+      "step": 779
+    },
+    {
+      "epoch": 0.3872409085267469,
+      "grad_norm": 0.13603104401195087,
+      "learning_rate": 4.8904632412706135e-06,
+      "loss": 0.7692,
+      "step": 780
+    },
+    {
+      "epoch": 0.38773737122998636,
+      "grad_norm": 0.1371314105411586,
+      "learning_rate": 4.890176927622879e-06,
+      "loss": 0.816,
+      "step": 781
+    },
+    {
+      "epoch": 0.38823383393322575,
+      "grad_norm": 0.13666348082090357,
+      "learning_rate": 4.889890248674926e-06,
+      "loss": 0.7781,
+      "step": 782
+    },
+    {
+      "epoch": 0.3887302966364652,
+      "grad_norm": 0.12972574048556007,
+      "learning_rate": 4.889603204470566e-06,
+      "loss": 0.7418,
+      "step": 783
+    },
+    {
+      "epoch": 0.38922675933970463,
+      "grad_norm": 0.1380447797265697,
+      "learning_rate": 4.889315795053671e-06,
+      "loss": 0.7666,
+      "step": 784
+    },
+    {
+      "epoch": 0.389723222042944,
+      "grad_norm": 0.1302714912154436,
+      "learning_rate": 4.889028020468167e-06,
+      "loss": 0.783,
+      "step": 785
+    },
+    {
+      "epoch": 0.39021968474618346,
+      "grad_norm": 0.1467555997648217,
+      "learning_rate": 4.8887398807580345e-06,
+      "loss": 0.7373,
+      "step": 786
+    },
+    {
+      "epoch": 0.39071614744942285,
+      "grad_norm": 0.13456233284077948,
+      "learning_rate": 4.888451375967313e-06,
+      "loss": 0.7843,
+      "step": 787
+    },
+    {
+      "epoch": 0.3912126101526623,
+      "grad_norm": 0.13579468578021064,
+      "learning_rate": 4.888162506140093e-06,
+      "loss": 0.7926,
+      "step": 788
+    },
+    {
+      "epoch": 0.3917090728559017,
+      "grad_norm": 0.13486742182263928,
+      "learning_rate": 4.887873271320526e-06,
+      "loss": 0.7848,
+      "step": 789
+    },
+    {
+      "epoch": 0.3922055355591411,
+      "grad_norm": 0.13060505166884362,
+      "learning_rate": 4.887583671552815e-06,
+      "loss": 0.7316,
+      "step": 790
+    },
+    {
+      "epoch": 0.39270199826238056,
+      "grad_norm": 0.13869981464559716,
+      "learning_rate": 4.887293706881224e-06,
+      "loss": 0.8121,
+      "step": 791
+    },
+    {
+      "epoch": 0.39319846096561994,
+      "grad_norm": 0.13166122616659146,
+      "learning_rate": 4.887003377350066e-06,
+      "loss": 0.7251,
+      "step": 792
+    },
+    {
+      "epoch": 0.3936949236688594,
+      "grad_norm": 0.14046352512611365,
+      "learning_rate": 4.886712683003715e-06,
+      "loss": 0.758,
+      "step": 793
+    },
+    {
+      "epoch": 0.3941913863720988,
+      "grad_norm": 0.14464252163844996,
+      "learning_rate": 4.886421623886598e-06,
+      "loss": 0.815,
+      "step": 794
+    },
+    {
+      "epoch": 0.3946878490753382,
+      "grad_norm": 0.13409840602053655,
+      "learning_rate": 4.886130200043199e-06,
+      "loss": 0.7752,
+      "step": 795
+    },
+    {
+      "epoch": 0.39518431177857766,
+      "grad_norm": 0.13116167627946584,
+      "learning_rate": 4.885838411518058e-06,
+      "loss": 0.7308,
+      "step": 796
+    },
+    {
+      "epoch": 0.39568077448181704,
+      "grad_norm": 0.1366342943883782,
+      "learning_rate": 4.885546258355769e-06,
+      "loss": 0.8026,
+      "step": 797
+    },
+    {
+      "epoch": 0.3961772371850565,
+      "grad_norm": 0.13731477100332815,
+      "learning_rate": 4.885253740600985e-06,
+      "loss": 0.7204,
+      "step": 798
+    },
+    {
+      "epoch": 0.39667369988829587,
+      "grad_norm": 0.14004464166577948,
+      "learning_rate": 4.88496085829841e-06,
+      "loss": 0.7729,
+      "step": 799
+    },
+    {
+      "epoch": 0.3971701625915353,
+      "grad_norm": 0.13691778866799595,
+      "learning_rate": 4.884667611492808e-06,
+      "loss": 0.7497,
+      "step": 800
+    },
+    {
+      "epoch": 0.39766662529477476,
+      "grad_norm": 0.13321987767176327,
+      "learning_rate": 4.884374000228998e-06,
+      "loss": 0.7415,
+      "step": 801
+    },
+    {
+      "epoch": 0.39816308799801414,
+      "grad_norm": 0.1398138150913019,
+      "learning_rate": 4.884080024551851e-06,
+      "loss": 0.7733,
+      "step": 802
+    },
+    {
+      "epoch": 0.3986595507012536,
+      "grad_norm": 0.1470854981776522,
+      "learning_rate": 4.8837856845062994e-06,
+      "loss": 0.7704,
+      "step": 803
+    },
+    {
+      "epoch": 0.39915601340449297,
+      "grad_norm": 0.13800002748758905,
+      "learning_rate": 4.883490980137327e-06,
+      "loss": 0.781,
+      "step": 804
+    },
+    {
+      "epoch": 0.3996524761077324,
+      "grad_norm": 0.13129465917522473,
+      "learning_rate": 4.883195911489974e-06,
+      "loss": 0.7406,
+      "step": 805
+    },
+    {
+      "epoch": 0.4001489388109718,
+      "grad_norm": 0.13483055098605387,
+      "learning_rate": 4.882900478609338e-06,
+      "loss": 0.7982,
+      "step": 806
+    },
+    {
+      "epoch": 0.40064540151421124,
+      "grad_norm": 0.13355656690282036,
+      "learning_rate": 4.8826046815405705e-06,
+      "loss": 0.8013,
+      "step": 807
+    },
+    {
+      "epoch": 0.4011418642174507,
+      "grad_norm": 0.13207774630124497,
+      "learning_rate": 4.882308520328879e-06,
+      "loss": 0.7445,
+      "step": 808
+    },
+    {
+      "epoch": 0.40163832692069007,
+      "grad_norm": 0.13638111186259064,
+      "learning_rate": 4.882011995019529e-06,
+      "loss": 0.7928,
+      "step": 809
+    },
+    {
+      "epoch": 0.4021347896239295,
+      "grad_norm": 0.13865080146765882,
+      "learning_rate": 4.881715105657837e-06,
+      "loss": 0.7972,
+      "step": 810
+    },
+    {
+      "epoch": 0.4026312523271689,
+      "grad_norm": 0.14025214556857968,
+      "learning_rate": 4.88141785228918e-06,
+      "loss": 0.8052,
+      "step": 811
+    },
+    {
+      "epoch": 0.40312771503040834,
+      "grad_norm": 0.13674600726600722,
+      "learning_rate": 4.881120234958986e-06,
+      "loss": 0.8034,
+      "step": 812
+    },
+    {
+      "epoch": 0.4036241777336478,
+      "grad_norm": 0.1360386191646674,
+      "learning_rate": 4.8808222537127436e-06,
+      "loss": 0.7882,
+      "step": 813
+    },
+    {
+      "epoch": 0.40412064043688717,
+      "grad_norm": 0.14178158974636348,
+      "learning_rate": 4.8805239085959936e-06,
+      "loss": 0.781,
+      "step": 814
+    },
+    {
+      "epoch": 0.4046171031401266,
+      "grad_norm": 0.1352509016655136,
+      "learning_rate": 4.880225199654331e-06,
+      "loss": 0.7834,
+      "step": 815
+    },
+    {
+      "epoch": 0.405113565843366,
+      "grad_norm": 0.13284596151325054,
+      "learning_rate": 4.879926126933412e-06,
+      "loss": 0.7846,
+      "step": 816
+    },
+    {
+      "epoch": 0.40561002854660544,
+      "grad_norm": 0.13744074794376318,
+      "learning_rate": 4.8796266904789445e-06,
+      "loss": 0.7567,
+      "step": 817
+    },
+    {
+      "epoch": 0.4061064912498449,
+      "grad_norm": 0.13859156903074973,
+      "learning_rate": 4.8793268903366905e-06,
+      "loss": 0.793,
+      "step": 818
+    },
+    {
+      "epoch": 0.40660295395308427,
+      "grad_norm": 0.14032883681460276,
+      "learning_rate": 4.879026726552471e-06,
+      "loss": 0.8204,
+      "step": 819
+    },
+    {
+      "epoch": 0.4070994166563237,
+      "grad_norm": 0.13375088164936702,
+      "learning_rate": 4.878726199172162e-06,
+      "loss": 0.7397,
+      "step": 820
+    },
+    {
+      "epoch": 0.4075958793595631,
+      "grad_norm": 0.13828272771245217,
+      "learning_rate": 4.878425308241693e-06,
+      "loss": 0.7657,
+      "step": 821
+    },
+    {
+      "epoch": 0.40809234206280254,
+      "grad_norm": 0.14138136285113104,
+      "learning_rate": 4.87812405380705e-06,
+      "loss": 0.7708,
+      "step": 822
+    },
+    {
+      "epoch": 0.4085888047660419,
+      "grad_norm": 0.13690922786942494,
+      "learning_rate": 4.8778224359142775e-06,
+      "loss": 0.7559,
+      "step": 823
+    },
+    {
+      "epoch": 0.40908526746928137,
+      "grad_norm": 0.1428920733606056,
+      "learning_rate": 4.87752045460947e-06,
+      "loss": 0.7543,
+      "step": 824
+    },
+    {
+      "epoch": 0.4095817301725208,
+      "grad_norm": 0.13259517168171056,
+      "learning_rate": 4.877218109938782e-06,
+      "loss": 0.7468,
+      "step": 825
+    },
+    {
+      "epoch": 0.4100781928757602,
+      "grad_norm": 0.13286837812829952,
+      "learning_rate": 4.876915401948421e-06,
+      "loss": 0.7498,
+      "step": 826
+    },
+    {
+      "epoch": 0.41057465557899964,
+      "grad_norm": 0.1505604133945408,
+      "learning_rate": 4.876612330684652e-06,
+      "loss": 0.8083,
+      "step": 827
+    },
+    {
+      "epoch": 0.411071118282239,
+      "grad_norm": 0.14013330639827198,
+      "learning_rate": 4.876308896193795e-06,
+      "loss": 0.7805,
+      "step": 828
+    },
+    {
+      "epoch": 0.41156758098547847,
+      "grad_norm": 0.141751989941827,
+      "learning_rate": 4.876005098522224e-06,
+      "loss": 0.789,
+      "step": 829
+    },
+    {
+      "epoch": 0.4120640436887179,
+      "grad_norm": 0.1439633625106176,
+      "learning_rate": 4.87570093771637e-06,
+      "loss": 0.7825,
+      "step": 830
+    },
+    {
+      "epoch": 0.4125605063919573,
+      "grad_norm": 0.13578927525549853,
+      "learning_rate": 4.875396413822719e-06,
+      "loss": 0.7434,
+      "step": 831
+    },
+    {
+      "epoch": 0.41305696909519674,
+      "grad_norm": 0.14301935596353957,
+      "learning_rate": 4.875091526887813e-06,
+      "loss": 0.7866,
+      "step": 832
+    },
+    {
+      "epoch": 0.4135534317984361,
+      "grad_norm": 0.13962888610646845,
+      "learning_rate": 4.8747862769582485e-06,
+      "loss": 0.797,
+      "step": 833
+    },
+    {
+      "epoch": 0.41404989450167556,
+      "grad_norm": 0.13762036122185398,
+      "learning_rate": 4.874480664080679e-06,
+      "loss": 0.7647,
+      "step": 834
+    },
+    {
+      "epoch": 0.414546357204915,
+      "grad_norm": 0.14874331327029203,
+      "learning_rate": 4.874174688301811e-06,
+      "loss": 0.7742,
+      "step": 835
+    },
+    {
+      "epoch": 0.4150428199081544,
+      "grad_norm": 0.14145690460721833,
+      "learning_rate": 4.873868349668409e-06,
+      "loss": 0.7794,
+      "step": 836
+    },
+    {
+      "epoch": 0.41553928261139383,
+      "grad_norm": 0.13516936437887206,
+      "learning_rate": 4.873561648227292e-06,
+      "loss": 0.772,
+      "step": 837
+    },
+    {
+      "epoch": 0.4160357453146332,
+      "grad_norm": 0.13997107273973253,
+      "learning_rate": 4.873254584025335e-06,
+      "loss": 0.7802,
+      "step": 838
+    },
+    {
+      "epoch": 0.41653220801787266,
+      "grad_norm": 0.15576247682108174,
+      "learning_rate": 4.8729471571094665e-06,
+      "loss": 0.8451,
+      "step": 839
+    },
+    {
+      "epoch": 0.41702867072111205,
+      "grad_norm": 0.1391550838396009,
+      "learning_rate": 4.872639367526672e-06,
+      "loss": 0.74,
+      "step": 840
+    },
+    {
+      "epoch": 0.4175251334243515,
+      "grad_norm": 0.13906947337724887,
+      "learning_rate": 4.872331215323993e-06,
+      "loss": 0.7883,
+      "step": 841
+    },
+    {
+      "epoch": 0.41802159612759093,
+      "grad_norm": 0.1369675811107513,
+      "learning_rate": 4.872022700548525e-06,
+      "loss": 0.7384,
+      "step": 842
+    },
+    {
+      "epoch": 0.4185180588308303,
+      "grad_norm": 0.1443749383496745,
+      "learning_rate": 4.87171382324742e-06,
+      "loss": 0.773,
+      "step": 843
+    },
+    {
+      "epoch": 0.41901452153406976,
+      "grad_norm": 0.13524935011822023,
+      "learning_rate": 4.871404583467884e-06,
+      "loss": 0.8338,
+      "step": 844
+    },
+    {
+      "epoch": 0.41951098423730915,
+      "grad_norm": 0.13600833744156696,
+      "learning_rate": 4.8710949812571805e-06,
+      "loss": 0.7715,
+      "step": 845
+    },
+    {
+      "epoch": 0.4200074469405486,
+      "grad_norm": 0.13910800980465646,
+      "learning_rate": 4.870785016662627e-06,
+      "loss": 0.7653,
+      "step": 846
+    },
+    {
+      "epoch": 0.42050390964378803,
+      "grad_norm": 0.14716154207375018,
+      "learning_rate": 4.870474689731596e-06,
+      "loss": 0.7753,
+      "step": 847
+    },
+    {
+      "epoch": 0.4210003723470274,
+      "grad_norm": 0.13651844849366374,
+      "learning_rate": 4.870164000511516e-06,
+      "loss": 0.7252,
+      "step": 848
+    },
+    {
+      "epoch": 0.42149683505026686,
+      "grad_norm": 0.13107938750960044,
+      "learning_rate": 4.869852949049872e-06,
+      "loss": 0.7403,
+      "step": 849
+    },
+    {
+      "epoch": 0.42199329775350625,
+      "grad_norm": 0.1405132580685121,
+      "learning_rate": 4.8695415353942025e-06,
+      "loss": 0.7816,
+      "step": 850
+    },
+    {
+      "epoch": 0.4224897604567457,
+      "grad_norm": 0.16019940582846015,
+      "learning_rate": 4.869229759592101e-06,
+      "loss": 0.7977,
+      "step": 851
+    },
+    {
+      "epoch": 0.42298622315998513,
+      "grad_norm": 0.1386861474814064,
+      "learning_rate": 4.868917621691219e-06,
+      "loss": 0.8859,
+      "step": 852
+    },
+    {
+      "epoch": 0.4234826858632245,
+      "grad_norm": 0.1438580035371236,
+      "learning_rate": 4.868605121739261e-06,
+      "loss": 0.7708,
+      "step": 853
+    },
+    {
+      "epoch": 0.42397914856646396,
+      "grad_norm": 0.13851415036446404,
+      "learning_rate": 4.868292259783988e-06,
+      "loss": 0.7379,
+      "step": 854
+    },
+    {
+      "epoch": 0.42447561126970335,
+      "grad_norm": 0.13928629569080855,
+      "learning_rate": 4.867979035873216e-06,
+      "loss": 0.7663,
+      "step": 855
+    },
+    {
+      "epoch": 0.4249720739729428,
+      "grad_norm": 0.1326009455578934,
+      "learning_rate": 4.8676654500548156e-06,
+      "loss": 0.7432,
+      "step": 856
+    },
+    {
+      "epoch": 0.4254685366761822,
+      "grad_norm": 0.13141434025922472,
+      "learning_rate": 4.867351502376714e-06,
+      "loss": 0.7558,
+      "step": 857
+    },
+    {
+      "epoch": 0.4259649993794216,
+      "grad_norm": 0.1301537847703651,
+      "learning_rate": 4.867037192886893e-06,
+      "loss": 0.7003,
+      "step": 858
+    },
+    {
+      "epoch": 0.42646146208266106,
+      "grad_norm": 0.13958306777392412,
+      "learning_rate": 4.866722521633389e-06,
+      "loss": 0.8252,
+      "step": 859
+    },
+    {
+      "epoch": 0.42695792478590044,
+      "grad_norm": 0.1371175213676672,
+      "learning_rate": 4.866407488664296e-06,
+      "loss": 0.7807,
+      "step": 860
+    },
+    {
+      "epoch": 0.4274543874891399,
+      "grad_norm": 0.1326784777780711,
+      "learning_rate": 4.866092094027761e-06,
+      "loss": 0.7701,
+      "step": 861
+    },
+    {
+      "epoch": 0.4279508501923793,
+      "grad_norm": 0.13589123146778564,
+      "learning_rate": 4.865776337771986e-06,
+      "loss": 0.7374,
+      "step": 862
+    },
+    {
+      "epoch": 0.4284473128956187,
+      "grad_norm": 0.13307442584248397,
+      "learning_rate": 4.86546021994523e-06,
+      "loss": 0.7581,
+      "step": 863
+    },
+    {
+      "epoch": 0.42894377559885816,
+      "grad_norm": 0.12935156456607097,
+      "learning_rate": 4.865143740595807e-06,
+      "loss": 0.7259,
+      "step": 864
+    },
+    {
+      "epoch": 0.42944023830209754,
+      "grad_norm": 0.13856977776450893,
+      "learning_rate": 4.864826899772086e-06,
+      "loss": 0.7921,
+      "step": 865
+    },
+    {
+      "epoch": 0.429936701005337,
+      "grad_norm": 0.13098700749992728,
+      "learning_rate": 4.864509697522489e-06,
+      "loss": 0.7581,
+      "step": 866
+    },
+    {
+      "epoch": 0.43043316370857637,
+      "grad_norm": 0.1277394417952265,
+      "learning_rate": 4.864192133895499e-06,
+      "loss": 0.7424,
+      "step": 867
+    },
+    {
+      "epoch": 0.4309296264118158,
+      "grad_norm": 0.1297413964613426,
+      "learning_rate": 4.8638742089396464e-06,
+      "loss": 0.7869,
+      "step": 868
+    },
+    {
+      "epoch": 0.43142608911505526,
+      "grad_norm": 0.1395569604985153,
+      "learning_rate": 4.863555922703523e-06,
+      "loss": 0.7837,
+      "step": 869
+    },
+    {
+      "epoch": 0.43192255181829464,
+      "grad_norm": 0.14789922672713987,
+      "learning_rate": 4.863237275235774e-06,
+      "loss": 0.8032,
+      "step": 870
+    },
+    {
+      "epoch": 0.4324190145215341,
+      "grad_norm": 0.148230222225613,
+      "learning_rate": 4.8629182665850995e-06,
+      "loss": 0.7581,
+      "step": 871
+    },
+    {
+      "epoch": 0.43291547722477347,
+      "grad_norm": 0.13520814483911175,
+      "learning_rate": 4.862598896800254e-06,
+      "loss": 0.7961,
+      "step": 872
+    },
+    {
+      "epoch": 0.4334119399280129,
+      "grad_norm": 0.1400523186641088,
+      "learning_rate": 4.862279165930049e-06,
+      "loss": 0.7443,
+      "step": 873
+    },
+    {
+      "epoch": 0.43390840263125235,
+      "grad_norm": 0.14000148595084386,
+      "learning_rate": 4.861959074023348e-06,
+      "loss": 0.7092,
+      "step": 874
+    },
+    {
+      "epoch": 0.43440486533449174,
+      "grad_norm": 0.13363969656101354,
+      "learning_rate": 4.8616386211290755e-06,
+      "loss": 0.7532,
+      "step": 875
+    },
+    {
+      "epoch": 0.4349013280377312,
+      "grad_norm": 0.1338415278751024,
+      "learning_rate": 4.861317807296205e-06,
+      "loss": 0.7598,
+      "step": 876
+    },
+    {
+      "epoch": 0.43539779074097057,
+      "grad_norm": 0.13445264018650427,
+      "learning_rate": 4.860996632573769e-06,
+      "loss": 0.7517,
+      "step": 877
+    },
+    {
+      "epoch": 0.43589425344421,
+      "grad_norm": 0.13560845548955244,
+      "learning_rate": 4.860675097010853e-06,
+      "loss": 0.7565,
+      "step": 878
+    },
+    {
+      "epoch": 0.4363907161474494,
+      "grad_norm": 0.14803665445404562,
+      "learning_rate": 4.860353200656599e-06,
+      "loss": 0.8122,
+      "step": 879
+    },
+    {
+      "epoch": 0.43688717885068884,
+      "grad_norm": 0.1373213861484435,
+      "learning_rate": 4.860030943560204e-06,
+      "loss": 0.7541,
+      "step": 880
+    },
+    {
+      "epoch": 0.4373836415539283,
+      "grad_norm": 0.12955793875119787,
+      "learning_rate": 4.859708325770919e-06,
+      "loss": 0.7502,
+      "step": 881
+    },
+    {
+      "epoch": 0.43788010425716767,
+      "grad_norm": 0.1441160209192175,
+      "learning_rate": 4.859385347338052e-06,
+      "loss": 0.7734,
+      "step": 882
+    },
+    {
+      "epoch": 0.4383765669604071,
+      "grad_norm": 0.13942948203768102,
+      "learning_rate": 4.8590620083109645e-06,
+      "loss": 0.7528,
+      "step": 883
+    },
+    {
+      "epoch": 0.4388730296636465,
+      "grad_norm": 0.14399699173850394,
+      "learning_rate": 4.858738308739073e-06,
+      "loss": 0.7628,
+      "step": 884
+    },
+    {
+      "epoch": 0.43936949236688594,
+      "grad_norm": 0.137394867443982,
+      "learning_rate": 4.858414248671851e-06,
+      "loss": 0.8637,
+      "step": 885
+    },
+    {
+      "epoch": 0.4398659550701254,
+      "grad_norm": 0.14448436375711424,
+      "learning_rate": 4.8580898281588255e-06,
+      "loss": 0.8147,
+      "step": 886
+    },
+    {
+      "epoch": 0.44036241777336477,
+      "grad_norm": 0.1341227002790187,
+      "learning_rate": 4.8577650472495785e-06,
+      "loss": 0.7602,
+      "step": 887
+    },
+    {
+      "epoch": 0.4408588804766042,
+      "grad_norm": 0.14377468127267348,
+      "learning_rate": 4.857439905993748e-06,
+      "loss": 0.7913,
+      "step": 888
+    },
+    {
+      "epoch": 0.4413553431798436,
+      "grad_norm": 0.13340052111413453,
+      "learning_rate": 4.857114404441027e-06,
+      "loss": 0.7482,
+      "step": 889
+    },
+    {
+      "epoch": 0.44185180588308304,
+      "grad_norm": 0.13574880212233634,
+      "learning_rate": 4.856788542641162e-06,
+      "loss": 0.7771,
+      "step": 890
+    },
+    {
+      "epoch": 0.4423482685863225,
+      "grad_norm": 0.1394549058341552,
+      "learning_rate": 4.856462320643957e-06,
+      "loss": 0.727,
+      "step": 891
+    },
+    {
+      "epoch": 0.44284473128956187,
+      "grad_norm": 0.16206005047704644,
+      "learning_rate": 4.856135738499269e-06,
+      "loss": 0.7901,
+      "step": 892
+    },
+    {
+      "epoch": 0.4433411939928013,
+      "grad_norm": 0.14512700342806611,
+      "learning_rate": 4.855808796257012e-06,
+      "loss": 0.8094,
+      "step": 893
+    },
+    {
+      "epoch": 0.4438376566960407,
+      "grad_norm": 0.1531915103698981,
+      "learning_rate": 4.855481493967152e-06,
+      "loss": 0.8711,
+      "step": 894
+    },
+    {
+      "epoch": 0.44433411939928014,
+      "grad_norm": 0.14141629017985097,
+      "learning_rate": 4.855153831679713e-06,
+      "loss": 0.7529,
+      "step": 895
+    },
+    {
+      "epoch": 0.4448305821025195,
+      "grad_norm": 0.14793605224512887,
+      "learning_rate": 4.854825809444773e-06,
+      "loss": 0.7606,
+      "step": 896
+    },
+    {
+      "epoch": 0.44532704480575896,
+      "grad_norm": 0.14449987137890363,
+      "learning_rate": 4.854497427312465e-06,
+      "loss": 0.8164,
+      "step": 897
+    },
+    {
+      "epoch": 0.4458235075089984,
+      "grad_norm": 0.14487270020396967,
+      "learning_rate": 4.854168685332977e-06,
+      "loss": 0.7833,
+      "step": 898
+    },
+    {
+      "epoch": 0.4463199702122378,
+      "grad_norm": 0.14100131641479818,
+      "learning_rate": 4.853839583556551e-06,
+      "loss": 0.7785,
+      "step": 899
+    },
+    {
+      "epoch": 0.44681643291547724,
+      "grad_norm": 0.1402338208402533,
+      "learning_rate": 4.853510122033486e-06,
+      "loss": 0.7922,
+      "step": 900
+    },
+    {
+      "epoch": 0.4473128956187166,
+      "grad_norm": 0.15070020553438185,
+      "learning_rate": 4.853180300814135e-06,
+      "loss": 0.7818,
+      "step": 901
+    },
+    {
+      "epoch": 0.44780935832195606,
+      "grad_norm": 0.14340004090339828,
+      "learning_rate": 4.8528501199489045e-06,
+      "loss": 0.7831,
+      "step": 902
+    },
+    {
+      "epoch": 0.4483058210251955,
+      "grad_norm": 0.12886555141368564,
+      "learning_rate": 4.852519579488258e-06,
+      "loss": 0.7358,
+      "step": 903
+    },
+    {
+      "epoch": 0.4488022837284349,
+      "grad_norm": 0.13659666264950565,
+      "learning_rate": 4.852188679482715e-06,
+      "loss": 0.797,
+      "step": 904
+    },
+    {
+      "epoch": 0.44929874643167433,
+      "grad_norm": 0.1344704194549763,
+      "learning_rate": 4.851857419982845e-06,
+      "loss": 0.7746,
+      "step": 905
+    },
+    {
+      "epoch": 0.4497952091349137,
+      "grad_norm": 0.14038386138367254,
+      "learning_rate": 4.8515258010392786e-06,
+      "loss": 0.7897,
+      "step": 906
+    },
+    {
+      "epoch": 0.45029167183815316,
+      "grad_norm": 0.1310362323726338,
+      "learning_rate": 4.851193822702698e-06,
+      "loss": 0.7891,
+      "step": 907
+    },
+    {
+      "epoch": 0.4507881345413926,
+      "grad_norm": 0.13456562483222215,
+      "learning_rate": 4.850861485023839e-06,
+      "loss": 0.8137,
+      "step": 908
+    },
+    {
+      "epoch": 0.451284597244632,
+      "grad_norm": 0.13867498023241415,
+      "learning_rate": 4.8505287880534954e-06,
+      "loss": 0.755,
+      "step": 909
+    },
+    {
+      "epoch": 0.45178105994787143,
+      "grad_norm": 0.13709552350038648,
+      "learning_rate": 4.8501957318425145e-06,
+      "loss": 0.7789,
+      "step": 910
+    },
+    {
+      "epoch": 0.4522775226511108,
+      "grad_norm": 0.20256799536792502,
+      "learning_rate": 4.849862316441799e-06,
+      "loss": 0.7916,
+      "step": 911
+    },
+    {
+      "epoch": 0.45277398535435026,
+      "grad_norm": 0.13805282049183487,
+      "learning_rate": 4.849528541902304e-06,
+      "loss": 0.7578,
+      "step": 912
+    },
+    {
+      "epoch": 0.45327044805758965,
+      "grad_norm": 0.13846494457743488,
+      "learning_rate": 4.849194408275045e-06,
+      "loss": 0.7993,
+      "step": 913
+    },
+    {
+      "epoch": 0.4537669107608291,
+      "grad_norm": 0.13624434622740592,
+      "learning_rate": 4.848859915611087e-06,
+      "loss": 0.7218,
+      "step": 914
+    },
+    {
+      "epoch": 0.45426337346406853,
+      "grad_norm": 0.13735589522446118,
+      "learning_rate": 4.848525063961551e-06,
+      "loss": 0.7927,
+      "step": 915
+    },
+    {
+      "epoch": 0.4547598361673079,
+      "grad_norm": 0.1435760229715099,
+      "learning_rate": 4.848189853377615e-06,
+      "loss": 0.7959,
+      "step": 916
+    },
+    {
+      "epoch": 0.45525629887054736,
+      "grad_norm": 0.1374310227347919,
+      "learning_rate": 4.8478542839105105e-06,
+      "loss": 0.8102,
+      "step": 917
+    },
+    {
+      "epoch": 0.45575276157378675,
+      "grad_norm": 0.13784054445459668,
+      "learning_rate": 4.847518355611524e-06,
+      "loss": 0.8705,
+      "step": 918
+    },
+    {
+      "epoch": 0.4562492242770262,
+      "grad_norm": 0.1429652700234789,
+      "learning_rate": 4.8471820685319965e-06,
+      "loss": 0.8273,
+      "step": 919
+    },
+    {
+      "epoch": 0.45674568698026563,
+      "grad_norm": 0.13776620485826613,
+      "learning_rate": 4.8468454227233235e-06,
+      "loss": 0.7659,
+      "step": 920
+    },
+    {
+      "epoch": 0.457242149683505,
+      "grad_norm": 0.13044770694847052,
+      "learning_rate": 4.8465084182369566e-06,
+      "loss": 0.7336,
+      "step": 921
+    },
+    {
+      "epoch": 0.45773861238674446,
+      "grad_norm": 0.13757280161779586,
+      "learning_rate": 4.846171055124401e-06,
+      "loss": 0.7948,
+      "step": 922
+    },
+    {
+      "epoch": 0.45823507508998385,
+      "grad_norm": 0.1391821959776347,
+      "learning_rate": 4.845833333437219e-06,
+      "loss": 0.7553,
+      "step": 923
+    },
+    {
+      "epoch": 0.4587315377932233,
+      "grad_norm": 0.1412266279982853,
+      "learning_rate": 4.845495253227023e-06,
+      "loss": 0.8104,
+      "step": 924
+    },
+    {
+      "epoch": 0.45922800049646273,
+      "grad_norm": 0.13037017997721642,
+      "learning_rate": 4.845156814545485e-06,
+      "loss": 0.755,
+      "step": 925
+    },
+    {
+      "epoch": 0.4597244631997021,
+      "grad_norm": 0.13938907235913728,
+      "learning_rate": 4.844818017444331e-06,
+      "loss": 0.7582,
+      "step": 926
+    },
+    {
+      "epoch": 0.46022092590294156,
+      "grad_norm": 0.13642206078365174,
+      "learning_rate": 4.8444788619753375e-06,
+      "loss": 0.7709,
+      "step": 927
+    },
+    {
+      "epoch": 0.46071738860618094,
+      "grad_norm": 0.13281685905144847,
+      "learning_rate": 4.844139348190342e-06,
+      "loss": 0.7613,
+      "step": 928
+    },
+    {
+      "epoch": 0.4612138513094204,
+      "grad_norm": 0.1376296253823425,
+      "learning_rate": 4.843799476141231e-06,
+      "loss": 0.8199,
+      "step": 929
+    },
+    {
+      "epoch": 0.4617103140126598,
+      "grad_norm": 0.1430226666279928,
+      "learning_rate": 4.843459245879952e-06,
+      "loss": 0.7879,
+      "step": 930
+    },
+    {
+      "epoch": 0.4622067767158992,
+      "grad_norm": 0.13853491478085572,
+      "learning_rate": 4.8431186574585e-06,
+      "loss": 0.799,
+      "step": 931
+    },
+    {
+      "epoch": 0.46270323941913866,
+      "grad_norm": 0.14106561846325905,
+      "learning_rate": 4.842777710928932e-06,
+      "loss": 0.799,
+      "step": 932
+    },
+    {
+      "epoch": 0.46319970212237804,
+      "grad_norm": 0.13514878380767298,
+      "learning_rate": 4.842436406343353e-06,
+      "loss": 0.7928,
+      "step": 933
+    },
+    {
+      "epoch": 0.4636961648256175,
+      "grad_norm": 0.12994346068694584,
+      "learning_rate": 4.842094743753929e-06,
+      "loss": 0.7338,
+      "step": 934
+    },
+    {
+      "epoch": 0.46419262752885687,
+      "grad_norm": 0.13540353265751,
+      "learning_rate": 4.841752723212874e-06,
+      "loss": 0.7365,
+      "step": 935
+    },
+    {
+      "epoch": 0.4646890902320963,
+      "grad_norm": 0.1383233625428097,
+      "learning_rate": 4.841410344772464e-06,
+      "loss": 0.7931,
+      "step": 936
+    },
+    {
+      "epoch": 0.46518555293533576,
+      "grad_norm": 0.13984478325484884,
+      "learning_rate": 4.841067608485024e-06,
+      "loss": 0.8022,
+      "step": 937
+    },
+    {
+      "epoch": 0.46568201563857514,
+      "grad_norm": 0.13182848523521054,
+      "learning_rate": 4.840724514402936e-06,
+      "loss": 0.7207,
+      "step": 938
+    },
+    {
+      "epoch": 0.4661784783418146,
+      "grad_norm": 0.13720103035170697,
+      "learning_rate": 4.8403810625786366e-06,
+      "loss": 0.7699,
+      "step": 939
+    },
+    {
+      "epoch": 0.46667494104505397,
+      "grad_norm": 0.13805399872377858,
+      "learning_rate": 4.840037253064617e-06,
+      "loss": 0.7676,
+      "step": 940
+    },
+    {
+      "epoch": 0.4671714037482934,
+      "grad_norm": 0.133084203123112,
+      "learning_rate": 4.839693085913423e-06,
+      "loss": 0.7366,
+      "step": 941
+    },
+    {
+      "epoch": 0.46766786645153285,
+      "grad_norm": 0.1338565121410207,
+      "learning_rate": 4.839348561177656e-06,
+      "loss": 0.7738,
+      "step": 942
+    },
+    {
+      "epoch": 0.46816432915477224,
+      "grad_norm": 0.14005655784202467,
+      "learning_rate": 4.839003678909968e-06,
+      "loss": 0.8011,
+      "step": 943
+    },
+    {
+      "epoch": 0.4686607918580117,
+      "grad_norm": 0.13866968197048227,
+      "learning_rate": 4.838658439163072e-06,
+      "loss": 0.7869,
+      "step": 944
+    },
+    {
+      "epoch": 0.46915725456125107,
+      "grad_norm": 0.14490477371225619,
+      "learning_rate": 4.838312841989731e-06,
+      "loss": 0.8122,
+      "step": 945
+    },
+    {
+      "epoch": 0.4696537172644905,
+      "grad_norm": 0.13501188505213316,
+      "learning_rate": 4.837966887442764e-06,
+      "loss": 0.8048,
+      "step": 946
+    },
+    {
+      "epoch": 0.4701501799677299,
+      "grad_norm": 0.13061796475992138,
+      "learning_rate": 4.837620575575045e-06,
+      "loss": 0.7505,
+      "step": 947
+    },
+    {
+      "epoch": 0.47064664267096934,
+      "grad_norm": 0.13123329580779844,
+      "learning_rate": 4.837273906439501e-06,
+      "loss": 0.7514,
+      "step": 948
+    },
+    {
+      "epoch": 0.4711431053742088,
+      "grad_norm": 0.13931347467033603,
+      "learning_rate": 4.836926880089117e-06,
+      "loss": 0.7545,
+      "step": 949
+    },
+    {
+      "epoch": 0.47163956807744817,
+      "grad_norm": 0.13794961687851362,
+      "learning_rate": 4.836579496576928e-06,
+      "loss": 0.8052,
+      "step": 950
+    },
+    {
+      "epoch": 0.4721360307806876,
+      "grad_norm": 0.13580616802094855,
+      "learning_rate": 4.836231755956028e-06,
+      "loss": 0.755,
+      "step": 951
+    },
+    {
+      "epoch": 0.472632493483927,
+      "grad_norm": 0.1375813327214774,
+      "learning_rate": 4.835883658279562e-06,
+      "loss": 0.7844,
+      "step": 952
+    },
+    {
+      "epoch": 0.47312895618716644,
+      "grad_norm": 0.1334419288379686,
+      "learning_rate": 4.835535203600732e-06,
+      "loss": 0.7886,
+      "step": 953
+    },
+    {
+      "epoch": 0.4736254188904059,
+      "grad_norm": 0.13849333665474736,
+      "learning_rate": 4.835186391972795e-06,
+      "loss": 0.7719,
+      "step": 954
+    },
+    {
+      "epoch": 0.47412188159364527,
+      "grad_norm": 0.13814260935997866,
+      "learning_rate": 4.834837223449058e-06,
+      "loss": 0.7593,
+      "step": 955
+    },
+    {
+      "epoch": 0.4746183442968847,
+      "grad_norm": 0.12665160870314987,
+      "learning_rate": 4.834487698082888e-06,
+      "loss": 0.6989,
+      "step": 956
+    },
+    {
+      "epoch": 0.4751148070001241,
+      "grad_norm": 0.13161442511625795,
+      "learning_rate": 4.834137815927705e-06,
+      "loss": 0.7502,
+      "step": 957
+    },
+    {
+      "epoch": 0.47561126970336354,
+      "grad_norm": 0.13088405996355415,
+      "learning_rate": 4.833787577036981e-06,
+      "loss": 0.7672,
+      "step": 958
+    },
+    {
+      "epoch": 0.476107732406603,
+      "grad_norm": 0.14037857434636752,
+      "learning_rate": 4.833436981464246e-06,
+      "loss": 0.7766,
+      "step": 959
+    },
+    {
+      "epoch": 0.47660419510984237,
+      "grad_norm": 0.14031622164392288,
+      "learning_rate": 4.833086029263081e-06,
+      "loss": 0.7815,
+      "step": 960
+    },
+    {
+      "epoch": 0.4771006578130818,
+      "grad_norm": 0.13531078299513247,
+      "learning_rate": 4.832734720487125e-06,
+      "loss": 0.7547,
+      "step": 961
+    },
+    {
+      "epoch": 0.4775971205163212,
+      "grad_norm": 0.13337784164706815,
+      "learning_rate": 4.8323830551900705e-06,
+      "loss": 0.7955,
+      "step": 962
+    },
+    {
+      "epoch": 0.47809358321956064,
+      "grad_norm": 0.1347558468252713,
+      "learning_rate": 4.832031033425663e-06,
+      "loss": 0.7486,
+      "step": 963
+    },
+    {
+      "epoch": 0.4785900459228,
+      "grad_norm": 0.12872155796573553,
+      "learning_rate": 4.831678655247702e-06,
+      "loss": 0.7768,
+      "step": 964
+    },
+    {
+      "epoch": 0.47908650862603946,
+      "grad_norm": 0.1348281268570624,
+      "learning_rate": 4.831325920710045e-06,
+      "loss": 0.7895,
+      "step": 965
+    },
+    {
+      "epoch": 0.4795829713292789,
+      "grad_norm": 0.13513666322029744,
+      "learning_rate": 4.830972829866601e-06,
+      "loss": 0.7915,
+      "step": 966
+    },
+    {
+      "epoch": 0.4800794340325183,
+      "grad_norm": 0.13491478767782572,
+      "learning_rate": 4.830619382771334e-06,
+      "loss": 0.7665,
+      "step": 967
+    },
+    {
+      "epoch": 0.48057589673575773,
+      "grad_norm": 0.13770336867680258,
+      "learning_rate": 4.830265579478263e-06,
+      "loss": 0.8023,
+      "step": 968
+    },
+    {
+      "epoch": 0.4810723594389971,
+      "grad_norm": 0.13544943102350623,
+      "learning_rate": 4.829911420041461e-06,
+      "loss": 0.7393,
+      "step": 969
+    },
+    {
+      "epoch": 0.48156882214223656,
+      "grad_norm": 0.13435327527712077,
+      "learning_rate": 4.829556904515056e-06,
+      "loss": 0.7506,
+      "step": 970
+    },
+    {
+      "epoch": 0.482065284845476,
+      "grad_norm": 0.13966597982630186,
+      "learning_rate": 4.82920203295323e-06,
+      "loss": 0.786,
+      "step": 971
+    },
+    {
+      "epoch": 0.4825617475487154,
+      "grad_norm": 0.1424347215239213,
+      "learning_rate": 4.828846805410219e-06,
+      "loss": 0.7934,
+      "step": 972
+    },
+    {
+      "epoch": 0.48305821025195483,
+      "grad_norm": 0.13234179930319845,
+      "learning_rate": 4.828491221940313e-06,
+      "loss": 0.7655,
+      "step": 973
+    },
+    {
+      "epoch": 0.4835546729551942,
+      "grad_norm": 0.14012843827614116,
+      "learning_rate": 4.82813528259786e-06,
+      "loss": 0.7496,
+      "step": 974
+    },
+    {
+      "epoch": 0.48405113565843366,
+      "grad_norm": 0.13715136273058942,
+      "learning_rate": 4.827778987437256e-06,
+      "loss": 0.8375,
+      "step": 975
+    },
+    {
+      "epoch": 0.4845475983616731,
+      "grad_norm": 0.1947144750999217,
+      "learning_rate": 4.827422336512958e-06,
+      "loss": 0.8667,
+      "step": 976
+    },
+    {
+      "epoch": 0.4850440610649125,
+      "grad_norm": 0.13741283362319834,
+      "learning_rate": 4.827065329879473e-06,
+      "loss": 0.8308,
+      "step": 977
+    },
+    {
+      "epoch": 0.48554052376815193,
+      "grad_norm": 0.13386724756044246,
+      "learning_rate": 4.826707967591364e-06,
+      "loss": 0.7704,
+      "step": 978
+    },
+    {
+      "epoch": 0.4860369864713913,
+      "grad_norm": 0.14441549526350253,
+      "learning_rate": 4.826350249703249e-06,
+      "loss": 0.7697,
+      "step": 979
+    },
+    {
+      "epoch": 0.48653344917463076,
+      "grad_norm": 0.14033203475651918,
+      "learning_rate": 4.825992176269797e-06,
+      "loss": 0.8019,
+      "step": 980
+    },
+    {
+      "epoch": 0.48702991187787015,
+      "grad_norm": 0.13714343033359203,
+      "learning_rate": 4.825633747345736e-06,
+      "loss": 0.7529,
+      "step": 981
+    },
+    {
+      "epoch": 0.4875263745811096,
+      "grad_norm": 0.12941529603226035,
+      "learning_rate": 4.825274962985845e-06,
+      "loss": 0.7562,
+      "step": 982
+    },
+    {
+      "epoch": 0.48802283728434903,
+      "grad_norm": 0.13974390480489013,
+      "learning_rate": 4.82491582324496e-06,
+      "loss": 0.8017,
+      "step": 983
+    },
+    {
+      "epoch": 0.4885192999875884,
+      "grad_norm": 0.1366449796259129,
+      "learning_rate": 4.824556328177968e-06,
+      "loss": 0.7515,
+      "step": 984
+    },
+    {
+      "epoch": 0.48901576269082786,
+      "grad_norm": 0.14711958138650388,
+      "learning_rate": 4.824196477839812e-06,
+      "loss": 0.795,
+      "step": 985
+    },
+    {
+      "epoch": 0.48951222539406725,
+      "grad_norm": 0.14257192599317975,
+      "learning_rate": 4.8238362722854905e-06,
+      "loss": 0.7937,
+      "step": 986
+    },
+    {
+      "epoch": 0.4900086880973067,
+      "grad_norm": 0.1347130511121721,
+      "learning_rate": 4.823475711570055e-06,
+      "loss": 0.7963,
+      "step": 987
+    },
+    {
+      "epoch": 0.49050515080054613,
+      "grad_norm": 0.13038662073702864,
+      "learning_rate": 4.823114795748611e-06,
+      "loss": 0.7671,
+      "step": 988
+    },
+    {
+      "epoch": 0.4910016135037855,
+      "grad_norm": 0.14016752297712926,
+      "learning_rate": 4.8227535248763185e-06,
+      "loss": 0.7237,
+      "step": 989
+    },
+    {
+      "epoch": 0.49149807620702496,
+      "grad_norm": 0.14127823957650343,
+      "learning_rate": 4.8223918990083925e-06,
+      "loss": 0.8157,
+      "step": 990
+    },
+    {
+      "epoch": 0.49199453891026435,
+      "grad_norm": 0.15260187773192319,
+      "learning_rate": 4.8220299182001014e-06,
+      "loss": 0.7994,
+      "step": 991
+    },
+    {
+      "epoch": 0.4924910016135038,
+      "grad_norm": 0.13943785694471836,
+      "learning_rate": 4.821667582506768e-06,
+      "loss": 0.7663,
+      "step": 992
+    },
+    {
+      "epoch": 0.49298746431674323,
+      "grad_norm": 0.13275215903917895,
+      "learning_rate": 4.8213048919837694e-06,
+      "loss": 0.7614,
+      "step": 993
+    },
+    {
+      "epoch": 0.4934839270199826,
+      "grad_norm": 0.14101255565793344,
+      "learning_rate": 4.820941846686538e-06,
+      "loss": 0.7854,
+      "step": 994
+    },
+    {
+      "epoch": 0.49398038972322206,
+      "grad_norm": 0.13656223313177246,
+      "learning_rate": 4.820578446670559e-06,
+      "loss": 0.7944,
+      "step": 995
+    },
+    {
+      "epoch": 0.49447685242646144,
+      "grad_norm": 0.13690268510235684,
+      "learning_rate": 4.820214691991372e-06,
+      "loss": 0.7417,
+      "step": 996
+    },
+    {
+      "epoch": 0.4949733151297009,
+      "grad_norm": 0.14131395637674096,
+      "learning_rate": 4.81985058270457e-06,
+      "loss": 0.798,
+      "step": 997
+    },
+    {
+      "epoch": 0.49546977783294033,
+      "grad_norm": 0.13522955049735227,
+      "learning_rate": 4.819486118865804e-06,
+      "loss": 0.7744,
+      "step": 998
+    },
+    {
+      "epoch": 0.4959662405361797,
+      "grad_norm": 0.13340560525311815,
+      "learning_rate": 4.819121300530774e-06,
+      "loss": 0.7248,
+      "step": 999
+    },
+    {
+      "epoch": 0.49646270323941916,
+      "grad_norm": 0.14852492729160868,
+      "learning_rate": 4.8187561277552376e-06,
+      "loss": 0.8108,
+      "step": 1000
+    },
+    {
+      "epoch": 0.49695916594265854,
+      "grad_norm": 0.14319486748419052,
+      "learning_rate": 4.818390600595005e-06,
+      "loss": 0.7618,
+      "step": 1001
+    },
+    {
+      "epoch": 0.497455628645898,
+      "grad_norm": 0.13608599816764147,
+      "learning_rate": 4.818024719105942e-06,
+      "loss": 0.7732,
+      "step": 1002
+    },
+    {
+      "epoch": 0.49795209134913737,
+      "grad_norm": 0.13154303448443375,
+      "learning_rate": 4.817658483343967e-06,
+      "loss": 0.7478,
+      "step": 1003
+    },
+    {
+      "epoch": 0.4984485540523768,
+      "grad_norm": 0.13240308227842207,
+      "learning_rate": 4.817291893365055e-06,
+      "loss": 0.7901,
+      "step": 1004
+    },
+    {
+      "epoch": 0.49894501675561626,
+      "grad_norm": 0.14667939413445374,
+      "learning_rate": 4.816924949225231e-06,
+      "loss": 0.7964,
+      "step": 1005
+    },
+    {
+      "epoch": 0.49944147945885564,
+      "grad_norm": 0.13188227123187843,
+      "learning_rate": 4.816557650980578e-06,
+      "loss": 0.6981,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4999379421620951,
+      "grad_norm": 0.12344883892983644,
+      "learning_rate": 4.816189998687231e-06,
+      "loss": 0.717,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5004344048653345,
+      "grad_norm": 0.14130934547471188,
+      "learning_rate": 4.81582199240138e-06,
+      "loss": 0.7951,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5004344048653345,
+      "eval_loss": 0.7746918201446533,
+      "eval_runtime": 135.3234,
+      "eval_samples_per_second": 224.3,
+      "eval_steps_per_second": 28.044,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5009308675685739,
+      "grad_norm": 0.134048946932314,
+      "learning_rate": 4.815453632179269e-06,
+      "loss": 0.7824,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5014273302718133,
+      "grad_norm": 0.13698117562535206,
+      "learning_rate": 4.815084918077196e-06,
+      "loss": 0.8098,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5019237929750527,
+      "grad_norm": 0.14425364984140046,
+      "learning_rate": 4.8147158501515125e-06,
+      "loss": 0.7688,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5024202556782922,
+      "grad_norm": 0.1332199024954632,
+      "learning_rate": 4.814346428458624e-06,
+      "loss": 0.7414,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5029167183815316,
+      "grad_norm": 0.1325832205983266,
+      "learning_rate": 4.813976653054993e-06,
+      "loss": 0.8369,
+      "step": 1013
+    },
+    {
+      "epoch": 0.503413181084771,
+      "grad_norm": 0.13394042417772226,
+      "learning_rate": 4.813606523997132e-06,
+      "loss": 0.7683,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5039096437880104,
+      "grad_norm": 0.15382272131461908,
+      "learning_rate": 4.81323604134161e-06,
+      "loss": 0.7601,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5044061064912498,
+      "grad_norm": 0.14301524433648288,
+      "learning_rate": 4.8128652051450485e-06,
+      "loss": 0.7507,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5049025691944893,
+      "grad_norm": 0.13250662485949544,
+      "learning_rate": 4.812494015464124e-06,
+      "loss": 0.7627,
+      "step": 1017
+    },
+    {
+      "epoch": 0.5053990318977287,
+      "grad_norm": 0.136342520846831,
+      "learning_rate": 4.812122472355569e-06,
+      "loss": 0.7467,
+      "step": 1018
+    },
+    {
+      "epoch": 0.505895494600968,
+      "grad_norm": 0.1487816899919886,
+      "learning_rate": 4.811750575876164e-06,
+      "loss": 0.8047,
+      "step": 1019
+    },
+    {
+      "epoch": 0.5063919573042075,
+      "grad_norm": 0.13428718634545492,
+      "learning_rate": 4.811378326082751e-06,
+      "loss": 0.7221,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5068884200074469,
+      "grad_norm": 0.1439447750897544,
+      "learning_rate": 4.811005723032219e-06,
+      "loss": 0.7775,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5073848827106864,
+      "grad_norm": 0.13831804614048945,
+      "learning_rate": 4.810632766781519e-06,
+      "loss": 0.7779,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5078813454139258,
+      "grad_norm": 0.1322159420042874,
+      "learning_rate": 4.810259457387647e-06,
+      "loss": 0.7703,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5083778081171652,
+      "grad_norm": 0.13665579834505667,
+      "learning_rate": 4.80988579490766e-06,
+      "loss": 0.773,
+      "step": 1024
+    },
+    {
+      "epoch": 0.5088742708204046,
+      "grad_norm": 0.13881136535201452,
+      "learning_rate": 4.809511779398665e-06,
+      "loss": 0.7572,
+      "step": 1025
+    },
+    {
+      "epoch": 0.509370733523644,
+      "grad_norm": 0.12477421280336831,
+      "learning_rate": 4.809137410917825e-06,
+      "loss": 0.6892,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5098671962268835,
+      "grad_norm": 0.1333532065223667,
+      "learning_rate": 4.808762689522356e-06,
+      "loss": 0.7785,
+      "step": 1027
+    },
+    {
+      "epoch": 0.5103636589301229,
+      "grad_norm": 0.139072803813187,
+      "learning_rate": 4.8083876152695285e-06,
+      "loss": 0.7882,
+      "step": 1028
+    },
+    {
+      "epoch": 0.5108601216333623,
+      "grad_norm": 0.13361520481016406,
+      "learning_rate": 4.808012188216665e-06,
+      "loss": 0.7488,
+      "step": 1029
+    },
+    {
+      "epoch": 0.5113565843366017,
+      "grad_norm": 0.13878508065904147,
+      "learning_rate": 4.807636408421146e-06,
+      "loss": 0.8114,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5118530470398411,
+      "grad_norm": 0.13851247233308997,
+      "learning_rate": 4.807260275940401e-06,
+      "loss": 0.8061,
+      "step": 1031
+    },
+    {
+      "epoch": 0.5123495097430806,
+      "grad_norm": 0.1373268203823735,
+      "learning_rate": 4.806883790831918e-06,
+      "loss": 0.7767,
+      "step": 1032
+    },
+    {
+      "epoch": 0.51284597244632,
+      "grad_norm": 0.13370723442504853,
+      "learning_rate": 4.806506953153235e-06,
+      "loss": 0.7912,
+      "step": 1033
+    },
+    {
+      "epoch": 0.5133424351495594,
+      "grad_norm": 0.1339473130089365,
+      "learning_rate": 4.806129762961946e-06,
+      "loss": 0.7391,
+      "step": 1034
+    },
+    {
+      "epoch": 0.5138388978527988,
+      "grad_norm": 0.15400773010538693,
+      "learning_rate": 4.805752220315699e-06,
+      "loss": 0.7607,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5143353605560382,
+      "grad_norm": 0.13291573497843698,
+      "learning_rate": 4.8053743252721954e-06,
+      "loss": 0.7504,
+      "step": 1036
+    },
+    {
+      "epoch": 0.5148318232592777,
+      "grad_norm": 0.14283555501521533,
+      "learning_rate": 4.804996077889189e-06,
+      "loss": 0.8104,
+      "step": 1037
+    },
+    {
+      "epoch": 0.5153282859625171,
+      "grad_norm": 0.13534285097280963,
+      "learning_rate": 4.8046174782244915e-06,
+      "loss": 0.7844,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5158247486657564,
+      "grad_norm": 0.14292129810329937,
+      "learning_rate": 4.804238526335963e-06,
+      "loss": 0.7838,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5163212113689959,
+      "grad_norm": 0.13392734203551412,
+      "learning_rate": 4.803859222281522e-06,
+      "loss": 0.7902,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5168176740722353,
+      "grad_norm": 0.19974575949479675,
+      "learning_rate": 4.803479566119138e-06,
+      "loss": 0.7953,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5173141367754748,
+      "grad_norm": 0.13159644138104465,
+      "learning_rate": 4.803099557906836e-06,
+      "loss": 0.769,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5178105994787141,
+      "grad_norm": 0.1388552332573298,
+      "learning_rate": 4.802719197702694e-06,
+      "loss": 0.8074,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5183070621819535,
+      "grad_norm": 0.13679177955633,
+      "learning_rate": 4.802338485564843e-06,
+      "loss": 0.7777,
+      "step": 1044
+    },
+    {
+      "epoch": 0.518803524885193,
+      "grad_norm": 0.13979496307917497,
+      "learning_rate": 4.8019574215514705e-06,
+      "loss": 0.8268,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5192999875884324,
+      "grad_norm": 0.12800445973569957,
+      "learning_rate": 4.801576005720816e-06,
+      "loss": 0.7296,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5197964502916719,
+      "grad_norm": 0.13332186476475139,
+      "learning_rate": 4.801194238131171e-06,
+      "loss": 0.7315,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5202929129949112,
+      "grad_norm": 0.13956881166358395,
+      "learning_rate": 4.800812118840884e-06,
+      "loss": 0.7848,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5207893756981506,
+      "grad_norm": 0.14486219818800664,
+      "learning_rate": 4.800429647908354e-06,
+      "loss": 0.7346,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5212858384013901,
+      "grad_norm": 0.13700325136931357,
+      "learning_rate": 4.800046825392039e-06,
+      "loss": 0.7827,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5217823011046295,
+      "grad_norm": 0.13238624342923797,
+      "learning_rate": 4.799663651350444e-06,
+      "loss": 0.737,
+      "step": 1051
+    },
+    {
+      "epoch": 0.522278763807869,
+      "grad_norm": 0.1409117185449216,
+      "learning_rate": 4.799280125842133e-06,
+      "loss": 0.7865,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5227752265111083,
+      "grad_norm": 0.16712731559406424,
+      "learning_rate": 4.79889624892572e-06,
+      "loss": 0.7599,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5232716892143477,
+      "grad_norm": 0.14064208506068193,
+      "learning_rate": 4.798512020659876e-06,
+      "loss": 0.7839,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5237681519175872,
+      "grad_norm": 0.13461779032052776,
+      "learning_rate": 4.7981274411033225e-06,
+      "loss": 0.7552,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5242646146208266,
+      "grad_norm": 0.13913729985952666,
+      "learning_rate": 4.797742510314838e-06,
+      "loss": 0.7936,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5247610773240661,
+      "grad_norm": 0.136526668795062,
+      "learning_rate": 4.797357228353252e-06,
+      "loss": 0.7271,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5252575400273054,
+      "grad_norm": 0.13602099516938532,
+      "learning_rate": 4.796971595277449e-06,
+      "loss": 0.7623,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5257540027305448,
+      "grad_norm": 0.13692035663882426,
+      "learning_rate": 4.796585611146367e-06,
+      "loss": 0.8187,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5262504654337843,
+      "grad_norm": 0.13511454024474553,
+      "learning_rate": 4.7961992760189975e-06,
+      "loss": 0.7717,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5267469281370237,
+      "grad_norm": 0.13653868560447394,
+      "learning_rate": 4.795812589954385e-06,
+      "loss": 0.7493,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5272433908402632,
+      "grad_norm": 0.14284294774022338,
+      "learning_rate": 4.795425553011629e-06,
+      "loss": 0.7828,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5277398535435025,
+      "grad_norm": 0.13641287616715173,
+      "learning_rate": 4.795038165249882e-06,
+      "loss": 0.7575,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5282363162467419,
+      "grad_norm": 0.13399693271167426,
+      "learning_rate": 4.794650426728349e-06,
+      "loss": 0.7907,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5287327789499814,
+      "grad_norm": 0.13237777235979964,
+      "learning_rate": 4.79426233750629e-06,
+      "loss": 0.7839,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5292292416532208,
+      "grad_norm": 0.1390129774707514,
+      "learning_rate": 4.793873897643019e-06,
+      "loss": 0.7499,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5297257043564603,
+      "grad_norm": 0.15109478098320664,
+      "learning_rate": 4.793485107197902e-06,
+      "loss": 0.7653,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5302221670596996,
+      "grad_norm": 0.14002670175893508,
+      "learning_rate": 4.793095966230359e-06,
+      "loss": 0.7565,
+      "step": 1068
+    },
+    {
+      "epoch": 0.530718629762939,
+      "grad_norm": 0.13739224347859277,
+      "learning_rate": 4.792706474799865e-06,
+      "loss": 0.698,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5312150924661785,
+      "grad_norm": 0.14191602937165326,
+      "learning_rate": 4.792316632965947e-06,
+      "loss": 0.7827,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5317115551694179,
+      "grad_norm": 0.13560549121655127,
+      "learning_rate": 4.791926440788186e-06,
+      "loss": 0.7651,
+      "step": 1071
+    },
+    {
+      "epoch": 0.5322080178726574,
+      "grad_norm": 0.1476565893427292,
+      "learning_rate": 4.791535898326217e-06,
+      "loss": 0.7883,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5327044805758967,
+      "grad_norm": 0.13008136272495874,
+      "learning_rate": 4.791145005639729e-06,
+      "loss": 0.78,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5332009432791361,
+      "grad_norm": 0.1327752708528825,
+      "learning_rate": 4.790753762788461e-06,
+      "loss": 0.7236,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5336974059823756,
+      "grad_norm": 0.14247698450976287,
+      "learning_rate": 4.79036216983221e-06,
+      "loss": 0.7687,
+      "step": 1075
+    },
+    {
+      "epoch": 0.534193868685615,
+      "grad_norm": 0.13898239220917705,
+      "learning_rate": 4.789970226830825e-06,
+      "loss": 0.7758,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5346903313888545,
+      "grad_norm": 0.13280694141514973,
+      "learning_rate": 4.789577933844207e-06,
+      "loss": 0.7538,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5351867940920938,
+      "grad_norm": 0.14265951347289108,
+      "learning_rate": 4.7891852909323135e-06,
+      "loss": 0.7814,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5356832567953332,
+      "grad_norm": 0.13154992991885675,
+      "learning_rate": 4.788792298155152e-06,
+      "loss": 0.7445,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5361797194985727,
+      "grad_norm": 0.13519381517124315,
+      "learning_rate": 4.788398955572786e-06,
+      "loss": 0.7487,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5366761822018121,
+      "grad_norm": 0.13044574075599727,
+      "learning_rate": 4.788005263245331e-06,
+      "loss": 0.7455,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5371726449050515,
+      "grad_norm": 0.13266919082399994,
+      "learning_rate": 4.787611221232957e-06,
+      "loss": 0.7757,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5376691076082909,
+      "grad_norm": 0.13376136036691982,
+      "learning_rate": 4.787216829595887e-06,
+      "loss": 0.7519,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5381655703115303,
+      "grad_norm": 0.13418604670303574,
+      "learning_rate": 4.786822088394397e-06,
+      "loss": 0.7926,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5386620330147698,
+      "grad_norm": 0.13482820857673963,
+      "learning_rate": 4.786426997688817e-06,
+      "loss": 0.7779,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5391584957180092,
+      "grad_norm": 0.130173504550406,
+      "learning_rate": 4.786031557539532e-06,
+      "loss": 0.717,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5396549584212486,
+      "grad_norm": 0.14790797249373314,
+      "learning_rate": 4.785635768006975e-06,
+      "loss": 0.8366,
+      "step": 1087
+    },
+    {
+      "epoch": 0.540151421124488,
+      "grad_norm": 0.13839226189444587,
+      "learning_rate": 4.78523962915164e-06,
+      "loss": 0.7947,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5406478838277274,
+      "grad_norm": 0.13079738699092058,
+      "learning_rate": 4.784843141034068e-06,
+      "loss": 0.7289,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5411443465309669,
+      "grad_norm": 0.13393354009715142,
+      "learning_rate": 4.784446303714856e-06,
+      "loss": 0.7756,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5416408092342063,
+      "grad_norm": 0.14144453676659285,
+      "learning_rate": 4.784049117254656e-06,
+      "loss": 0.793,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5421372719374457,
+      "grad_norm": 0.1300964867769227,
+      "learning_rate": 4.783651581714169e-06,
+      "loss": 0.7597,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5426337346406851,
+      "grad_norm": 0.1357913073042723,
+      "learning_rate": 4.7832536971541546e-06,
+      "loss": 0.788,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5431301973439245,
+      "grad_norm": 0.13222437636600876,
+      "learning_rate": 4.7828554636354216e-06,
+      "loss": 0.7809,
+      "step": 1094
+    },
+    {
+      "epoch": 0.543626660047164,
+      "grad_norm": 0.13258977130957175,
+      "learning_rate": 4.782456881218834e-06,
+      "loss": 0.7448,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5441231227504034,
+      "grad_norm": 0.13730375474418763,
+      "learning_rate": 4.782057949965307e-06,
+      "loss": 0.7784,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5446195854536428,
+      "grad_norm": 0.13642288559333154,
+      "learning_rate": 4.781658669935813e-06,
+      "loss": 0.781,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5451160481568822,
+      "grad_norm": 0.1401061103174583,
+      "learning_rate": 4.7812590411913755e-06,
+      "loss": 0.7655,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5456125108601216,
+      "grad_norm": 0.13382843638430478,
+      "learning_rate": 4.780859063793071e-06,
+      "loss": 0.761,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5461089735633611,
+      "grad_norm": 0.13467015132419832,
+      "learning_rate": 4.780458737802028e-06,
+      "loss": 0.8425,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5466054362666005,
+      "grad_norm": 0.13538805042285004,
+      "learning_rate": 4.780058063279432e-06,
+      "loss": 0.8023,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5471018989698399,
+      "grad_norm": 0.13332827992090326,
+      "learning_rate": 4.779657040286519e-06,
+      "loss": 0.7454,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5475983616730793,
+      "grad_norm": 0.13714991044262492,
+      "learning_rate": 4.779255668884579e-06,
+      "loss": 0.7781,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5480948243763187,
+      "grad_norm": 0.13230879275054402,
+      "learning_rate": 4.778853949134956e-06,
+      "loss": 0.7512,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5485912870795582,
+      "grad_norm": 0.13456603648064147,
+      "learning_rate": 4.778451881099045e-06,
+      "loss": 0.8209,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5490877497827976,
+      "grad_norm": 0.1308167779233721,
+      "learning_rate": 4.7780494648382955e-06,
+      "loss": 0.764,
+      "step": 1106
+    },
+    {
+      "epoch": 0.549584212486037,
+      "grad_norm": 0.16101308796343475,
+      "learning_rate": 4.777646700414211e-06,
+      "loss": 0.7786,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5500806751892764,
+      "grad_norm": 0.1459790001808385,
+      "learning_rate": 4.777243587888348e-06,
+      "loss": 0.7868,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5505771378925158,
+      "grad_norm": 0.13657235299783674,
+      "learning_rate": 4.776840127322316e-06,
+      "loss": 0.8113,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5510736005957553,
+      "grad_norm": 0.14294798523667074,
+      "learning_rate": 4.7764363187777765e-06,
+      "loss": 0.7649,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5515700632989947,
+      "grad_norm": 0.13846612184838433,
+      "learning_rate": 4.776032162316445e-06,
+      "loss": 0.7307,
+      "step": 1111
+    },
+    {
+      "epoch": 0.552066526002234,
+      "grad_norm": 0.14690601721943153,
+      "learning_rate": 4.775627658000091e-06,
+      "loss": 0.7695,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5525629887054735,
+      "grad_norm": 0.14581794719191593,
+      "learning_rate": 4.775222805890537e-06,
+      "loss": 0.8154,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5530594514087129,
+      "grad_norm": 0.14444888719097665,
+      "learning_rate": 4.7748176060496574e-06,
+      "loss": 0.8383,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5535559141119524,
+      "grad_norm": 0.13254092300237755,
+      "learning_rate": 4.77441205853938e-06,
+      "loss": 0.8386,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5540523768151917,
+      "grad_norm": 0.1347220636449837,
+      "learning_rate": 4.774006163421687e-06,
+      "loss": 0.752,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5545488395184311,
+      "grad_norm": 0.13548044148354343,
+      "learning_rate": 4.773599920758614e-06,
+      "loss": 0.7409,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5550453022216706,
+      "grad_norm": 0.14102047024443465,
+      "learning_rate": 4.773193330612246e-06,
+      "loss": 0.7561,
+      "step": 1118
+    },
+    {
+      "epoch": 0.55554176492491,
+      "grad_norm": 0.13374789735214362,
+      "learning_rate": 4.772786393044726e-06,
+      "loss": 0.7837,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5560382276281495,
+      "grad_norm": 0.13483434548142853,
+      "learning_rate": 4.772379108118247e-06,
+      "loss": 0.7457,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5565346903313888,
+      "grad_norm": 0.13493609143971075,
+      "learning_rate": 4.7719714758950565e-06,
+      "loss": 0.7414,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5570311530346282,
+      "grad_norm": 0.14080955544969456,
+      "learning_rate": 4.771563496437454e-06,
+      "loss": 0.7954,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5575276157378677,
+      "grad_norm": 0.15315063885453276,
+      "learning_rate": 4.771155169807793e-06,
+      "loss": 0.787,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5580240784411071,
+      "grad_norm": 0.1343038161486333,
+      "learning_rate": 4.770746496068479e-06,
+      "loss": 0.7636,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5585205411443466,
+      "grad_norm": 0.1323279813259488,
+      "learning_rate": 4.770337475281972e-06,
+      "loss": 0.7246,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5590170038475859,
+      "grad_norm": 0.12987821538488584,
+      "learning_rate": 4.769928107510784e-06,
+      "loss": 0.753,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5595134665508253,
+      "grad_norm": 0.14121741134888693,
+      "learning_rate": 4.7695183928174804e-06,
+      "loss": 0.7721,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5600099292540648,
+      "grad_norm": 0.13095267808109395,
+      "learning_rate": 4.76910833126468e-06,
+      "loss": 0.7727,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5605063919573042,
+      "grad_norm": 0.13160852813062984,
+      "learning_rate": 4.768697922915053e-06,
+      "loss": 0.7499,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5610028546605437,
+      "grad_norm": 0.13131472336248304,
+      "learning_rate": 4.768287167831323e-06,
+      "loss": 0.7271,
+      "step": 1130
+    },
+    {
+      "epoch": 0.561499317363783,
+      "grad_norm": 0.14344145034960104,
+      "learning_rate": 4.767876066076271e-06,
+      "loss": 0.8335,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5619957800670224,
+      "grad_norm": 0.13562056297798028,
+      "learning_rate": 4.7674646177127236e-06,
+      "loss": 0.8303,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5624922427702619,
+      "grad_norm": 0.1397414253999263,
+      "learning_rate": 4.767052822803565e-06,
+      "loss": 0.7954,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5629887054735013,
+      "grad_norm": 0.13474425151028716,
+      "learning_rate": 4.7666406814117324e-06,
+      "loss": 0.7554,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5634851681767408,
+      "grad_norm": 0.13073363088146278,
+      "learning_rate": 4.7662281936002155e-06,
+      "loss": 0.7659,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5639816308799801,
+      "grad_norm": 0.13266775105459727,
+      "learning_rate": 4.765815359432054e-06,
+      "loss": 0.8103,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5644780935832195,
+      "grad_norm": 0.14135466000280938,
+      "learning_rate": 4.765402178970345e-06,
+      "loss": 0.7477,
+      "step": 1137
+    },
+    {
+      "epoch": 0.564974556286459,
+      "grad_norm": 0.1364663264168024,
+      "learning_rate": 4.764988652278235e-06,
+      "loss": 0.7801,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5654710189896984,
+      "grad_norm": 0.133069946323726,
+      "learning_rate": 4.764574779418927e-06,
+      "loss": 0.7814,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5659674816929379,
+      "grad_norm": 0.13667566517665522,
+      "learning_rate": 4.7641605604556725e-06,
+      "loss": 0.7193,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5664639443961772,
+      "grad_norm": 0.14023909922018574,
+      "learning_rate": 4.763745995451781e-06,
+      "loss": 0.7794,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5669604070994166,
+      "grad_norm": 0.1343364576457988,
+      "learning_rate": 4.763331084470609e-06,
+      "loss": 0.7651,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5674568698026561,
+      "grad_norm": 0.13304454158801782,
+      "learning_rate": 4.762915827575571e-06,
+      "loss": 0.741,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5679533325058955,
+      "grad_norm": 0.13140270388519174,
+      "learning_rate": 4.762500224830132e-06,
+      "loss": 0.7353,
+      "step": 1144
+    },
+    {
+      "epoch": 0.568449795209135,
+      "grad_norm": 0.13074327953386966,
+      "learning_rate": 4.762084276297811e-06,
+      "loss": 0.765,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5689462579123743,
+      "grad_norm": 0.13281986443064125,
+      "learning_rate": 4.761667982042176e-06,
+      "loss": 0.7372,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5694427206156137,
+      "grad_norm": 0.13419581835055136,
+      "learning_rate": 4.7612513421268546e-06,
+      "loss": 0.7767,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5699391833188532,
+      "grad_norm": 0.13612919993671027,
+      "learning_rate": 4.760834356615521e-06,
+      "loss": 0.7834,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5704356460220926,
+      "grad_norm": 0.13693184772625464,
+      "learning_rate": 4.760417025571907e-06,
+      "loss": 0.7544,
+      "step": 1149
+    },
+    {
+      "epoch": 0.570932108725332,
+      "grad_norm": 0.13395312495200518,
+      "learning_rate": 4.759999349059793e-06,
+      "loss": 0.7411,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.1481697018642554,
+      "learning_rate": 4.759581327143015e-06,
+      "loss": 0.7968,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5719250341318108,
+      "grad_norm": 0.1371234108294726,
+      "learning_rate": 4.7591629598854595e-06,
+      "loss": 0.8194,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5724214968350503,
+      "grad_norm": 0.13604199584938104,
+      "learning_rate": 4.7587442473510705e-06,
+      "loss": 0.7528,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5729179595382897,
+      "grad_norm": 0.1346379831014206,
+      "learning_rate": 4.758325189603838e-06,
+      "loss": 0.7171,
+      "step": 1154
+    },
+    {
+      "epoch": 0.573414422241529,
+      "grad_norm": 0.14237506813976503,
+      "learning_rate": 4.757905786707811e-06,
+      "loss": 0.7842,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5739108849447685,
+      "grad_norm": 0.13403550334730185,
+      "learning_rate": 4.757486038727086e-06,
+      "loss": 0.7931,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5744073476480079,
+      "grad_norm": 0.135231376685361,
+      "learning_rate": 4.757065945725816e-06,
+      "loss": 0.7154,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5749038103512474,
+      "grad_norm": 0.13684138959608425,
+      "learning_rate": 4.756645507768207e-06,
+      "loss": 0.7383,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5754002730544868,
+      "grad_norm": 0.1460330953509727,
+      "learning_rate": 4.756224724918513e-06,
+      "loss": 0.7872,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5758967357577262,
+      "grad_norm": 0.1318765034689836,
+      "learning_rate": 4.755803597241047e-06,
+      "loss": 0.768,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5763931984609656,
+      "grad_norm": 0.13719986250467792,
+      "learning_rate": 4.755382124800169e-06,
+      "loss": 0.7453,
+      "step": 1161
+    },
+    {
+      "epoch": 0.576889661164205,
+      "grad_norm": 0.14235485272473405,
+      "learning_rate": 4.754960307660296e-06,
+      "loss": 0.7821,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5773861238674445,
+      "grad_norm": 0.1356872066401077,
+      "learning_rate": 4.754538145885896e-06,
+      "loss": 0.7501,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5778825865706839,
+      "grad_norm": 0.13978700563928276,
+      "learning_rate": 4.754115639541489e-06,
+      "loss": 0.7891,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5783790492739233,
+      "grad_norm": 0.1386484989822081,
+      "learning_rate": 4.7536927886916486e-06,
+      "loss": 0.8111,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5788755119771627,
+      "grad_norm": 0.1352803729379626,
+      "learning_rate": 4.753269593401e-06,
+      "loss": 0.7986,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5793719746804021,
+      "grad_norm": 0.14338323635053893,
+      "learning_rate": 4.752846053734223e-06,
+      "loss": 0.8083,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5798684373836416,
+      "grad_norm": 0.14290830558032838,
+      "learning_rate": 4.752422169756048e-06,
+      "loss": 0.7245,
+      "step": 1168
+    },
+    {
+      "epoch": 0.580364900086881,
+      "grad_norm": 0.14435662586254663,
+      "learning_rate": 4.7519979415312595e-06,
+      "loss": 0.7937,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5808613627901204,
+      "grad_norm": 0.1301445915669164,
+      "learning_rate": 4.751573369124693e-06,
+      "loss": 0.7629,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5813578254933598,
+      "grad_norm": 0.14135646579813632,
+      "learning_rate": 4.751148452601239e-06,
+      "loss": 0.8281,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5818542881965992,
+      "grad_norm": 0.14137941871842752,
+      "learning_rate": 4.750723192025839e-06,
+      "loss": 0.751,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5823507508998387,
+      "grad_norm": 0.13846670183254522,
+      "learning_rate": 4.750297587463486e-06,
+      "loss": 0.8034,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5828472136030781,
+      "grad_norm": 0.13325672340874672,
+      "learning_rate": 4.749871638979227e-06,
+      "loss": 0.6927,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5833436763063174,
+      "grad_norm": 0.125743329419156,
+      "learning_rate": 4.749445346638163e-06,
+      "loss": 0.7256,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5838401390095569,
+      "grad_norm": 0.1417089701135048,
+      "learning_rate": 4.749018710505444e-06,
+      "loss": 0.7831,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5843366017127963,
+      "grad_norm": 0.13605194542345714,
+      "learning_rate": 4.748591730646276e-06,
+      "loss": 0.8015,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5848330644160358,
+      "grad_norm": 0.13481999719592982,
+      "learning_rate": 4.748164407125915e-06,
+      "loss": 0.7615,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5853295271192752,
+      "grad_norm": 0.14186151175015785,
+      "learning_rate": 4.747736740009671e-06,
+      "loss": 0.764,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5858259898225145,
+      "grad_norm": 0.14184400720811477,
+      "learning_rate": 4.747308729362906e-06,
+      "loss": 0.7587,
+      "step": 1180
+    },
+    {
+      "epoch": 0.586322452525754,
+      "grad_norm": 0.1385901962211324,
+      "learning_rate": 4.746880375251034e-06,
+      "loss": 0.7336,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5868189152289934,
+      "grad_norm": 0.13451407782163222,
+      "learning_rate": 4.7464516777395234e-06,
+      "loss": 0.7429,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5873153779322329,
+      "grad_norm": 0.13022215035798407,
+      "learning_rate": 4.746022636893894e-06,
+      "loss": 0.7643,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5878118406354722,
+      "grad_norm": 0.13690715708703163,
+      "learning_rate": 4.745593252779715e-06,
+      "loss": 0.8217,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5883083033387116,
+      "grad_norm": 0.13748468118288174,
+      "learning_rate": 4.745163525462613e-06,
+      "loss": 0.7703,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5888047660419511,
+      "grad_norm": 0.1405973365721223,
+      "learning_rate": 4.744733455008265e-06,
+      "loss": 0.7383,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5893012287451905,
+      "grad_norm": 0.13717194937849955,
+      "learning_rate": 4.7443030414824e-06,
+      "loss": 0.7589,
+      "step": 1187
+    },
+    {
+      "epoch": 0.58979769144843,
+      "grad_norm": 0.13528229246468063,
+      "learning_rate": 4.743872284950799e-06,
+      "loss": 0.75,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5902941541516693,
+      "grad_norm": 0.133425574497451,
+      "learning_rate": 4.743441185479298e-06,
+      "loss": 0.7411,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5907906168549087,
+      "grad_norm": 0.1324925987986178,
+      "learning_rate": 4.743009743133782e-06,
+      "loss": 0.7685,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5912870795581482,
+      "grad_norm": 0.1369066542388308,
+      "learning_rate": 4.742577957980191e-06,
+      "loss": 0.7832,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5917835422613876,
+      "grad_norm": 0.13896137808156944,
+      "learning_rate": 4.7421458300845156e-06,
+      "loss": 0.7763,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5922800049646271,
+      "grad_norm": 0.13714450419418717,
+      "learning_rate": 4.7417133595128e-06,
+      "loss": 0.7755,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5927764676678664,
+      "grad_norm": 0.13866795158105188,
+      "learning_rate": 4.741280546331142e-06,
+      "loss": 0.7525,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5932729303711058,
+      "grad_norm": 0.13821511176421178,
+      "learning_rate": 4.740847390605688e-06,
+      "loss": 0.7808,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5937693930743453,
+      "grad_norm": 0.1380634473432959,
+      "learning_rate": 4.740413892402639e-06,
+      "loss": 0.7547,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5942658557775847,
+      "grad_norm": 0.138297189072512,
+      "learning_rate": 4.73998005178825e-06,
+      "loss": 0.7983,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5947623184808242,
+      "grad_norm": 0.13758909195503097,
+      "learning_rate": 4.739545868828824e-06,
+      "loss": 0.7437,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5952587811840635,
+      "grad_norm": 0.13643426919261928,
+      "learning_rate": 4.739111343590722e-06,
+      "loss": 0.7617,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5957552438873029,
+      "grad_norm": 0.13419264066031278,
+      "learning_rate": 4.7386764761403515e-06,
+      "loss": 0.7577,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5962517065905424,
+      "grad_norm": 0.1458962881146663,
+      "learning_rate": 4.738241266544176e-06,
+      "loss": 0.7542,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5967481692937818,
+      "grad_norm": 0.13326276596473177,
+      "learning_rate": 4.737805714868711e-06,
+      "loss": 0.7639,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5972446319970213,
+      "grad_norm": 0.1327575012050733,
+      "learning_rate": 4.737369821180522e-06,
+      "loss": 0.7815,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5977410947002606,
+      "grad_norm": 0.13976077843344958,
+      "learning_rate": 4.736933585546229e-06,
+      "loss": 0.771,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5982375574035,
+      "grad_norm": 0.1448357997256506,
+      "learning_rate": 4.736497008032505e-06,
+      "loss": 0.797,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5987340201067395,
+      "grad_norm": 0.14152292832314567,
+      "learning_rate": 4.7360600887060735e-06,
+      "loss": 0.8046,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5992304828099789,
+      "grad_norm": 0.13983648625283432,
+      "learning_rate": 4.735622827633709e-06,
+      "loss": 0.7767,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5997269455132184,
+      "grad_norm": 0.13723445239630844,
+      "learning_rate": 4.7351852248822405e-06,
+      "loss": 0.7898,
+      "step": 1208
+    },
+    {
+      "epoch": 0.6002234082164577,
+      "grad_norm": 0.14040576284527934,
+      "learning_rate": 4.734747280518549e-06,
+      "loss": 0.7507,
+      "step": 1209
+    },
+    {
+      "epoch": 0.6007198709196971,
+      "grad_norm": 0.13281149215833118,
+      "learning_rate": 4.734308994609568e-06,
+      "loss": 0.7344,
+      "step": 1210
+    },
+    {
+      "epoch": 0.6012163336229366,
+      "grad_norm": 0.14696680910440754,
+      "learning_rate": 4.73387036722228e-06,
+      "loss": 0.8057,
+      "step": 1211
+    },
+    {
+      "epoch": 0.601712796326176,
+      "grad_norm": 0.14656238080300293,
+      "learning_rate": 4.733431398423725e-06,
+      "loss": 0.7715,
+      "step": 1212
+    },
+    {
+      "epoch": 0.6022092590294155,
+      "grad_norm": 0.13640813199119411,
+      "learning_rate": 4.732992088280991e-06,
+      "loss": 0.7887,
+      "step": 1213
+    },
+    {
+      "epoch": 0.6027057217326548,
+      "grad_norm": 0.14416087368058042,
+      "learning_rate": 4.73255243686122e-06,
+      "loss": 0.7845,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6032021844358942,
+      "grad_norm": 0.1448460344463032,
+      "learning_rate": 4.732112444231604e-06,
+      "loss": 0.7858,
+      "step": 1215
+    },
+    {
+      "epoch": 0.6036986471391337,
+      "grad_norm": 0.14284369655429804,
+      "learning_rate": 4.731672110459391e-06,
+      "loss": 0.7595,
+      "step": 1216
+    },
+    {
+      "epoch": 0.6041951098423731,
+      "grad_norm": 0.13333186060643346,
+      "learning_rate": 4.7312314356118774e-06,
+      "loss": 0.7787,
+      "step": 1217
+    },
+    {
+      "epoch": 0.6046915725456126,
+      "grad_norm": 0.13448997562242113,
+      "learning_rate": 4.7307904197564146e-06,
+      "loss": 0.7907,
+      "step": 1218
+    },
+    {
+      "epoch": 0.6051880352488519,
+      "grad_norm": 0.13856754170442995,
+      "learning_rate": 4.730349062960405e-06,
+      "loss": 0.703,
+      "step": 1219
+    },
+    {
+      "epoch": 0.6056844979520913,
+      "grad_norm": 0.13197537382149402,
+      "learning_rate": 4.7299073652912995e-06,
+      "loss": 0.7698,
+      "step": 1220
+    },
+    {
+      "epoch": 0.6061809606553308,
+      "grad_norm": 0.1308991610627273,
+      "learning_rate": 4.729465326816609e-06,
+      "loss": 0.7331,
+      "step": 1221
+    },
+    {
+      "epoch": 0.6066774233585702,
+      "grad_norm": 0.13827465076211878,
+      "learning_rate": 4.72902294760389e-06,
+      "loss": 0.8255,
+      "step": 1222
+    },
+    {
+      "epoch": 0.6071738860618096,
+      "grad_norm": 0.14152077690894366,
+      "learning_rate": 4.7285802277207525e-06,
+      "loss": 0.7757,
+      "step": 1223
+    },
+    {
+      "epoch": 0.607670348765049,
+      "grad_norm": 0.13327976935901228,
+      "learning_rate": 4.72813716723486e-06,
+      "loss": 0.7834,
+      "step": 1224
+    },
+    {
+      "epoch": 0.6081668114682884,
+      "grad_norm": 0.14560592025153601,
+      "learning_rate": 4.727693766213927e-06,
+      "loss": 0.8553,
+      "step": 1225
+    },
+    {
+      "epoch": 0.6086632741715279,
+      "grad_norm": 0.139357875506195,
+      "learning_rate": 4.72725002472572e-06,
+      "loss": 0.7334,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6091597368747673,
+      "grad_norm": 0.1414789031418474,
+      "learning_rate": 4.726805942838058e-06,
+      "loss": 0.7975,
+      "step": 1227
+    },
+    {
+      "epoch": 0.6096561995780067,
+      "grad_norm": 0.13771948814756751,
+      "learning_rate": 4.726361520618812e-06,
+      "loss": 0.738,
+      "step": 1228
+    },
+    {
+      "epoch": 0.6101526622812461,
+      "grad_norm": 0.12954626047367315,
+      "learning_rate": 4.725916758135905e-06,
+      "loss": 0.7446,
+      "step": 1229
+    },
+    {
+      "epoch": 0.6106491249844855,
+      "grad_norm": 0.1321303236295978,
+      "learning_rate": 4.72547165545731e-06,
+      "loss": 0.7225,
+      "step": 1230
+    },
+    {
+      "epoch": 0.611145587687725,
+      "grad_norm": 0.13934783137333412,
+      "learning_rate": 4.725026212651056e-06,
+      "loss": 0.7504,
+      "step": 1231
+    },
+    {
+      "epoch": 0.6116420503909644,
+      "grad_norm": 0.1387209464213516,
+      "learning_rate": 4.72458042978522e-06,
+      "loss": 0.779,
+      "step": 1232
+    },
+    {
+      "epoch": 0.6121385130942038,
+      "grad_norm": 0.1417563704806412,
+      "learning_rate": 4.7241343069279355e-06,
+      "loss": 0.7727,
+      "step": 1233
+    },
+    {
+      "epoch": 0.6126349757974432,
+      "grad_norm": 0.1420634383737698,
+      "learning_rate": 4.723687844147383e-06,
+      "loss": 0.8045,
+      "step": 1234
+    },
+    {
+      "epoch": 0.6131314385006826,
+      "grad_norm": 0.142217432211696,
+      "learning_rate": 4.723241041511797e-06,
+      "loss": 0.7966,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6136279012039221,
+      "grad_norm": 0.13955237043987861,
+      "learning_rate": 4.722793899089465e-06,
+      "loss": 0.763,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6141243639071615,
+      "grad_norm": 0.1383129773036087,
+      "learning_rate": 4.7223464169487255e-06,
+      "loss": 0.7865,
+      "step": 1237
+    },
+    {
+      "epoch": 0.6146208266104008,
+      "grad_norm": 0.1318995281229908,
+      "learning_rate": 4.721898595157969e-06,
+      "loss": 0.7592,
+      "step": 1238
+    },
+    {
+      "epoch": 0.6151172893136403,
+      "grad_norm": 0.13409755823637834,
+      "learning_rate": 4.721450433785637e-06,
+      "loss": 0.7349,
+      "step": 1239
+    },
+    {
+      "epoch": 0.6156137520168797,
+      "grad_norm": 0.13477979687672653,
+      "learning_rate": 4.721001932900224e-06,
+      "loss": 0.7168,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6161102147201192,
+      "grad_norm": 0.1339375435013451,
+      "learning_rate": 4.720553092570278e-06,
+      "loss": 0.7679,
+      "step": 1241
+    },
+    {
+      "epoch": 0.6166066774233586,
+      "grad_norm": 0.13714226350401248,
+      "learning_rate": 4.720103912864395e-06,
+      "loss": 0.8179,
+      "step": 1242
+    },
+    {
+      "epoch": 0.617103140126598,
+      "grad_norm": 0.1495165323744844,
+      "learning_rate": 4.719654393851225e-06,
+      "loss": 0.8059,
+      "step": 1243
+    },
+    {
+      "epoch": 0.6175996028298374,
+      "grad_norm": 0.13210522151583382,
+      "learning_rate": 4.719204535599472e-06,
+      "loss": 0.7327,
+      "step": 1244
+    },
+    {
+      "epoch": 0.6180960655330768,
+      "grad_norm": 0.135327563201833,
+      "learning_rate": 4.718754338177887e-06,
+      "loss": 0.7884,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6185925282363163,
+      "grad_norm": 0.1493290671655753,
+      "learning_rate": 4.7183038016552765e-06,
+      "loss": 0.8058,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6190889909395557,
+      "grad_norm": 0.13731834920199315,
+      "learning_rate": 4.717852926100497e-06,
+      "loss": 0.7525,
+      "step": 1247
+    },
+    {
+      "epoch": 0.619585453642795,
+      "grad_norm": 0.13121903632834106,
+      "learning_rate": 4.717401711582459e-06,
+      "loss": 0.748,
+      "step": 1248
+    },
+    {
+      "epoch": 0.6200819163460345,
+      "grad_norm": 0.13357953738588796,
+      "learning_rate": 4.716950158170123e-06,
+      "loss": 0.7191,
+      "step": 1249
+    },
+    {
+      "epoch": 0.6205783790492739,
+      "grad_norm": 0.1333298023226559,
+      "learning_rate": 4.716498265932501e-06,
+      "loss": 0.7536,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6210748417525134,
+      "grad_norm": 0.13600946084924598,
+      "learning_rate": 4.7160460349386575e-06,
+      "loss": 0.7327,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6215713044557528,
+      "grad_norm": 0.15789147253069744,
+      "learning_rate": 4.7155934652577095e-06,
+      "loss": 0.7542,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6220677671589921,
+      "grad_norm": 0.14117884230963118,
+      "learning_rate": 4.7151405569588245e-06,
+      "loss": 0.7566,
+      "step": 1253
+    },
+    {
+      "epoch": 0.6225642298622316,
+      "grad_norm": 0.1434269696587236,
+      "learning_rate": 4.714687310111224e-06,
+      "loss": 0.7661,
+      "step": 1254
+    },
+    {
+      "epoch": 0.623060692565471,
+      "grad_norm": 0.13233837304493712,
+      "learning_rate": 4.714233724784176e-06,
+      "loss": 0.7387,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6235571552687105,
+      "grad_norm": 0.13451446763955208,
+      "learning_rate": 4.713779801047006e-06,
+      "loss": 0.7593,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6240536179719498,
+      "grad_norm": 0.14045733263477267,
+      "learning_rate": 4.713325538969089e-06,
+      "loss": 0.7588,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6245500806751892,
+      "grad_norm": 0.13811994049159718,
+      "learning_rate": 4.7128709386198516e-06,
+      "loss": 0.7849,
+      "step": 1258
+    },
+    {
+      "epoch": 0.6250465433784287,
+      "grad_norm": 0.13661650047767832,
+      "learning_rate": 4.712416000068771e-06,
+      "loss": 0.7871,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6255430060816681,
+      "grad_norm": 0.13356087506731615,
+      "learning_rate": 4.7119607233853795e-06,
+      "loss": 0.6997,
+      "step": 1260
+    },
+    {
+      "epoch": 0.6260394687849076,
+      "grad_norm": 0.14326475290125062,
+      "learning_rate": 4.7115051086392575e-06,
+      "loss": 0.7913,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6265359314881469,
+      "grad_norm": 0.14111338297951223,
+      "learning_rate": 4.711049155900037e-06,
+      "loss": 0.775,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6270323941913863,
+      "grad_norm": 0.1384448111488101,
+      "learning_rate": 4.710592865237406e-06,
+      "loss": 0.7453,
+      "step": 1263
+    },
+    {
+      "epoch": 0.6275288568946258,
+      "grad_norm": 0.13688056535276805,
+      "learning_rate": 4.710136236721099e-06,
+      "loss": 0.7835,
+      "step": 1264
+    },
+    {
+      "epoch": 0.6280253195978652,
+      "grad_norm": 0.1369014193152322,
+      "learning_rate": 4.709679270420905e-06,
+      "loss": 0.7708,
+      "step": 1265
+    },
+    {
+      "epoch": 0.6285217823011047,
+      "grad_norm": 0.13619728251726942,
+      "learning_rate": 4.709221966406664e-06,
+      "loss": 0.7593,
+      "step": 1266
+    },
+    {
+      "epoch": 0.629018245004344,
+      "grad_norm": 0.13341268255654185,
+      "learning_rate": 4.7087643247482675e-06,
+      "loss": 0.7857,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6295147077075834,
+      "grad_norm": 0.1396936626784199,
+      "learning_rate": 4.70830634551566e-06,
+      "loss": 0.7934,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6300111704108229,
+      "grad_norm": 0.13784257345636344,
+      "learning_rate": 4.7078480287788335e-06,
+      "loss": 0.8105,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6305076331140623,
+      "grad_norm": 0.14024550205782157,
+      "learning_rate": 4.707389374607837e-06,
+      "loss": 0.7744,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6310040958173018,
+      "grad_norm": 0.1331894256857835,
+      "learning_rate": 4.7069303830727665e-06,
+      "loss": 0.7454,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6315005585205411,
+      "grad_norm": 0.13774853065612155,
+      "learning_rate": 4.706471054243773e-06,
+      "loss": 0.7572,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6319970212237805,
+      "grad_norm": 0.13274604497668133,
+      "learning_rate": 4.7060113881910565e-06,
+      "loss": 0.8029,
+      "step": 1273
+    },
+    {
+      "epoch": 0.63249348392702,
+      "grad_norm": 0.13007067612592235,
+      "learning_rate": 4.705551384984871e-06,
+      "loss": 0.7329,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6329899466302594,
+      "grad_norm": 0.1348605302563776,
+      "learning_rate": 4.705091044695519e-06,
+      "loss": 0.7854,
+      "step": 1275
+    },
+    {
+      "epoch": 0.6334864093334989,
+      "grad_norm": 0.13943981892113053,
+      "learning_rate": 4.7046303673933566e-06,
+      "loss": 0.7621,
+      "step": 1276
+    },
+    {
+      "epoch": 0.6339828720367382,
+      "grad_norm": 0.13080686345224168,
+      "learning_rate": 4.7041693531487905e-06,
+      "loss": 0.7511,
+      "step": 1277
+    },
+    {
+      "epoch": 0.6344793347399776,
+      "grad_norm": 0.13367213440770978,
+      "learning_rate": 4.703708002032281e-06,
+      "loss": 0.8208,
+      "step": 1278
+    },
+    {
+      "epoch": 0.6349757974432171,
+      "grad_norm": 0.1338922082718588,
+      "learning_rate": 4.703246314114337e-06,
+      "loss": 0.7583,
+      "step": 1279
+    },
+    {
+      "epoch": 0.6354722601464565,
+      "grad_norm": 0.13384133583383315,
+      "learning_rate": 4.702784289465521e-06,
+      "loss": 0.7247,
+      "step": 1280
+    },
+    {
+      "epoch": 0.635968722849696,
+      "grad_norm": 0.14316701050668065,
+      "learning_rate": 4.702321928156443e-06,
+      "loss": 0.7769,
+      "step": 1281
+    },
+    {
+      "epoch": 0.6364651855529353,
+      "grad_norm": 0.1363830874610169,
+      "learning_rate": 4.701859230257772e-06,
+      "loss": 0.7508,
+      "step": 1282
+    },
+    {
+      "epoch": 0.6369616482561747,
+      "grad_norm": 0.13044310663712724,
+      "learning_rate": 4.701396195840221e-06,
+      "loss": 0.6918,
+      "step": 1283
+    },
+    {
+      "epoch": 0.6374581109594142,
+      "grad_norm": 0.14799662097345562,
+      "learning_rate": 4.700932824974558e-06,
+      "loss": 0.8122,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6379545736626536,
+      "grad_norm": 0.13719379503307672,
+      "learning_rate": 4.700469117731602e-06,
+      "loss": 0.7589,
+      "step": 1285
+    },
+    {
+      "epoch": 0.6384510363658931,
+      "grad_norm": 0.15005736439386755,
+      "learning_rate": 4.700005074182223e-06,
+      "loss": 0.7942,
+      "step": 1286
+    },
+    {
+      "epoch": 0.6389474990691324,
+      "grad_norm": 0.13361765649422278,
+      "learning_rate": 4.699540694397344e-06,
+      "loss": 0.7418,
+      "step": 1287
+    },
+    {
+      "epoch": 0.6394439617723718,
+      "grad_norm": 0.12927575638530414,
+      "learning_rate": 4.699075978447936e-06,
+      "loss": 0.7314,
+      "step": 1288
+    },
+    {
+      "epoch": 0.6399404244756113,
+      "grad_norm": 0.13136950718329457,
+      "learning_rate": 4.698610926405024e-06,
+      "loss": 0.7591,
+      "step": 1289
+    },
+    {
+      "epoch": 0.6404368871788507,
+      "grad_norm": 0.1404280522487197,
+      "learning_rate": 4.6981455383396845e-06,
+      "loss": 0.8238,
+      "step": 1290
+    },
+    {
+      "epoch": 0.64093334988209,
+      "grad_norm": 0.1303989293917867,
+      "learning_rate": 4.697679814323044e-06,
+      "loss": 0.7489,
+      "step": 1291
+    },
+    {
+      "epoch": 0.6414298125853295,
+      "grad_norm": 0.13338423096237687,
+      "learning_rate": 4.69721375442628e-06,
+      "loss": 0.7306,
+      "step": 1292
+    },
+    {
+      "epoch": 0.6419262752885689,
+      "grad_norm": 0.13305102973746538,
+      "learning_rate": 4.696747358720624e-06,
+      "loss": 0.7886,
+      "step": 1293
+    },
+    {
+      "epoch": 0.6424227379918084,
+      "grad_norm": 0.13820820437935433,
+      "learning_rate": 4.696280627277356e-06,
+      "loss": 0.7081,
+      "step": 1294
+    },
+    {
+      "epoch": 0.6429192006950478,
+      "grad_norm": 0.13801203340059798,
+      "learning_rate": 4.695813560167809e-06,
+      "loss": 0.7402,
+      "step": 1295
+    },
+    {
+      "epoch": 0.6434156633982872,
+      "grad_norm": 0.13519623515883733,
+      "learning_rate": 4.695346157463367e-06,
+      "loss": 0.7847,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6439121261015266,
+      "grad_norm": 0.13214986252403507,
+      "learning_rate": 4.6948784192354645e-06,
+      "loss": 0.7734,
+      "step": 1297
+    },
+    {
+      "epoch": 0.644408588804766,
+      "grad_norm": 0.1378462366811087,
+      "learning_rate": 4.694410345555588e-06,
+      "loss": 0.6981,
+      "step": 1298
+    },
+    {
+      "epoch": 0.6449050515080055,
+      "grad_norm": 0.1366789326165814,
+      "learning_rate": 4.6939419364952734e-06,
+      "loss": 0.7557,
+      "step": 1299
+    },
+    {
+      "epoch": 0.6454015142112449,
+      "grad_norm": 0.14607757273947794,
+      "learning_rate": 4.693473192126112e-06,
+      "loss": 0.7782,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6458979769144843,
+      "grad_norm": 0.13866618330491343,
+      "learning_rate": 4.693004112519743e-06,
+      "loss": 0.7827,
+      "step": 1301
+    },
+    {
+      "epoch": 0.6463944396177237,
+      "grad_norm": 0.13908697250264154,
+      "learning_rate": 4.692534697747858e-06,
+      "loss": 0.7945,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6468909023209631,
+      "grad_norm": 0.13233872664417434,
+      "learning_rate": 4.692064947882198e-06,
+      "loss": 0.786,
+      "step": 1303
+    },
+    {
+      "epoch": 0.6473873650242026,
+      "grad_norm": 0.13187870756426126,
+      "learning_rate": 4.6915948629945585e-06,
+      "loss": 0.6941,
+      "step": 1304
+    },
+    {
+      "epoch": 0.647883827727442,
+      "grad_norm": 0.13132502035875618,
+      "learning_rate": 4.691124443156784e-06,
+      "loss": 0.7943,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6483802904306813,
+      "grad_norm": 0.1473223707369921,
+      "learning_rate": 4.690653688440769e-06,
+      "loss": 0.758,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6488767531339208,
+      "grad_norm": 0.1374434003262379,
+      "learning_rate": 4.6901825989184634e-06,
+      "loss": 0.7477,
+      "step": 1307
+    },
+    {
+      "epoch": 0.6493732158371602,
+      "grad_norm": 0.13383271780435643,
+      "learning_rate": 4.689711174661864e-06,
+      "loss": 0.7669,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6498696785403997,
+      "grad_norm": 0.13444066067651259,
+      "learning_rate": 4.689239415743021e-06,
+      "loss": 0.7757,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6503661412436391,
+      "grad_norm": 0.12857578926705554,
+      "learning_rate": 4.688767322234035e-06,
+      "loss": 0.7442,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6508626039468784,
+      "grad_norm": 0.13406381840115994,
+      "learning_rate": 4.688294894207058e-06,
+      "loss": 0.7598,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6513590666501179,
+      "grad_norm": 0.14039516549020117,
+      "learning_rate": 4.687822131734293e-06,
+      "loss": 0.7842,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6518555293533573,
+      "grad_norm": 0.1374250686491604,
+      "learning_rate": 4.687349034887994e-06,
+      "loss": 0.7636,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6523519920565968,
+      "grad_norm": 0.13288394681388876,
+      "learning_rate": 4.686875603740467e-06,
+      "loss": 0.7648,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6528484547598362,
+      "grad_norm": 0.15493904800962308,
+      "learning_rate": 4.686401838364069e-06,
+      "loss": 0.7857,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6533449174630755,
+      "grad_norm": 0.1592114663190787,
+      "learning_rate": 4.685927738831204e-06,
+      "loss": 0.7933,
+      "step": 1316
+    },
+    {
+      "epoch": 0.653841380166315,
+      "grad_norm": 0.138465643640585,
+      "learning_rate": 4.685453305214335e-06,
+      "loss": 0.7698,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6543378428695544,
+      "grad_norm": 0.1379981991943177,
+      "learning_rate": 4.684978537585968e-06,
+      "loss": 0.7459,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6548343055727939,
+      "grad_norm": 0.14481568247936927,
+      "learning_rate": 4.684503436018664e-06,
+      "loss": 0.7588,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6553307682760333,
+      "grad_norm": 0.14210563691826714,
+      "learning_rate": 4.684028000585038e-06,
+      "loss": 0.7329,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6558272309792726,
+      "grad_norm": 0.12867626236605553,
+      "learning_rate": 4.683552231357749e-06,
+      "loss": 0.7337,
+      "step": 1321
+    },
+    {
+      "epoch": 0.6563236936825121,
+      "grad_norm": 0.1376861133072269,
+      "learning_rate": 4.683076128409512e-06,
+      "loss": 0.7972,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6568201563857515,
+      "grad_norm": 0.14125134793520855,
+      "learning_rate": 4.682599691813092e-06,
+      "loss": 0.7393,
+      "step": 1323
+    },
+    {
+      "epoch": 0.657316619088991,
+      "grad_norm": 0.14276487162583956,
+      "learning_rate": 4.682122921641305e-06,
+      "loss": 0.7879,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6578130817922304,
+      "grad_norm": 0.13468493472362367,
+      "learning_rate": 4.681645817967017e-06,
+      "loss": 0.7335,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6583095444954697,
+      "grad_norm": 0.13943793893639297,
+      "learning_rate": 4.681168380863145e-06,
+      "loss": 0.758,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6588060071987092,
+      "grad_norm": 0.13951539561696433,
+      "learning_rate": 4.680690610402659e-06,
+      "loss": 0.8151,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6593024699019486,
+      "grad_norm": 0.13406494700163638,
+      "learning_rate": 4.6802125066585765e-06,
+      "loss": 0.7013,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6597989326051881,
+      "grad_norm": 0.14160309961844755,
+      "learning_rate": 4.679734069703971e-06,
+      "loss": 0.7574,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6602953953084274,
+      "grad_norm": 0.13446373370464942,
+      "learning_rate": 4.679255299611961e-06,
+      "loss": 0.7507,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6607918580116668,
+      "grad_norm": 0.12942976744487117,
+      "learning_rate": 4.678776196455722e-06,
+      "loss": 0.7311,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6612883207149063,
+      "grad_norm": 0.1391321723122554,
+      "learning_rate": 4.678296760308474e-06,
+      "loss": 0.7981,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6617847834181457,
+      "grad_norm": 0.14428783420102687,
+      "learning_rate": 4.677816991243493e-06,
+      "loss": 0.8215,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6622812461213852,
+      "grad_norm": 0.1394878539696028,
+      "learning_rate": 4.677336889334103e-06,
+      "loss": 0.7721,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6627777088246245,
+      "grad_norm": 0.15266983844158868,
+      "learning_rate": 4.676856454653681e-06,
+      "loss": 0.7336,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6632741715278639,
+      "grad_norm": 0.14455346704491276,
+      "learning_rate": 4.676375687275653e-06,
+      "loss": 0.7278,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6637706342311034,
+      "grad_norm": 0.1316229836144885,
+      "learning_rate": 4.675894587273496e-06,
+      "loss": 0.7647,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6642670969343428,
+      "grad_norm": 0.12941666247842312,
+      "learning_rate": 4.6754131547207406e-06,
+      "loss": 0.7125,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6647635596375823,
+      "grad_norm": 0.14046373964383968,
+      "learning_rate": 4.674931389690963e-06,
+      "loss": 0.7341,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6652600223408216,
+      "grad_norm": 0.13645194981403033,
+      "learning_rate": 4.674449292257796e-06,
+      "loss": 0.7676,
+      "step": 1340
+    },
+    {
+      "epoch": 0.665756485044061,
+      "grad_norm": 0.14519132705834006,
+      "learning_rate": 4.6739668624949196e-06,
+      "loss": 0.7952,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6662529477473005,
+      "grad_norm": 0.13696033048575837,
+      "learning_rate": 4.6734841004760644e-06,
+      "loss": 0.7269,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6667494104505399,
+      "grad_norm": 0.134676078421444,
+      "learning_rate": 4.673001006275013e-06,
+      "loss": 0.7453,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6672458731537794,
+      "grad_norm": 0.14578953183512341,
+      "learning_rate": 4.672517579965601e-06,
+      "loss": 0.8015,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6677423358570187,
+      "grad_norm": 0.13466978301237612,
+      "learning_rate": 4.6720338216217096e-06,
+      "loss": 0.7375,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6682387985602581,
+      "grad_norm": 0.13431133998664943,
+      "learning_rate": 4.671549731317274e-06,
+      "loss": 0.7931,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6687352612634976,
+      "grad_norm": 0.14300934227788512,
+      "learning_rate": 4.67106530912628e-06,
+      "loss": 0.8223,
+      "step": 1347
+    },
+    {
+      "epoch": 0.669231723966737,
+      "grad_norm": 0.1335079638925206,
+      "learning_rate": 4.670580555122765e-06,
+      "loss": 0.7081,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6697281866699765,
+      "grad_norm": 0.13925505067882116,
+      "learning_rate": 4.670095469380814e-06,
+      "loss": 0.7556,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6702246493732158,
+      "grad_norm": 0.1385402073066566,
+      "learning_rate": 4.669610051974566e-06,
+      "loss": 0.7862,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6707211120764552,
+      "grad_norm": 0.14168635102257782,
+      "learning_rate": 4.669124302978208e-06,
+      "loss": 0.7532,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6712175747796947,
+      "grad_norm": 0.1319645107321289,
+      "learning_rate": 4.6686382224659795e-06,
+      "loss": 0.721,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6717140374829341,
+      "grad_norm": 0.14159994069821713,
+      "learning_rate": 4.66815181051217e-06,
+      "loss": 0.7324,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6722105001861736,
+      "grad_norm": 0.13302384063587522,
+      "learning_rate": 4.667665067191121e-06,
+      "loss": 0.7914,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6727069628894129,
+      "grad_norm": 0.14520590994816845,
+      "learning_rate": 4.667177992577222e-06,
+      "loss": 0.803,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6732034255926523,
+      "grad_norm": 0.13258982595212945,
+      "learning_rate": 4.666690586744914e-06,
+      "loss": 0.7802,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6736998882958918,
+      "grad_norm": 0.14243373706488904,
+      "learning_rate": 4.666202849768691e-06,
+      "loss": 0.7915,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6741963509991312,
+      "grad_norm": 0.14197795066279506,
+      "learning_rate": 4.6657147817230945e-06,
+      "loss": 0.7584,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6746928137023707,
+      "grad_norm": 0.13496558186041357,
+      "learning_rate": 4.665226382682718e-06,
+      "loss": 0.735,
+      "step": 1359
+    },
+    {
+      "epoch": 0.67518927640561,
+      "grad_norm": 0.13833577662827512,
+      "learning_rate": 4.664737652722205e-06,
+      "loss": 0.7653,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6756857391088494,
+      "grad_norm": 0.13481054586272334,
+      "learning_rate": 4.664248591916252e-06,
+      "loss": 0.7518,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6761822018120889,
+      "grad_norm": 0.13380123421346032,
+      "learning_rate": 4.663759200339603e-06,
+      "loss": 0.742,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6766786645153283,
+      "grad_norm": 0.13210724903338833,
+      "learning_rate": 4.663269478067053e-06,
+      "loss": 0.7546,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6771751272185677,
+      "grad_norm": 0.13236042360384812,
+      "learning_rate": 4.6627794251734485e-06,
+      "loss": 0.7404,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6776715899218071,
+      "grad_norm": 0.13755480780239387,
+      "learning_rate": 4.662289041733686e-06,
+      "loss": 0.7639,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6781680526250465,
+      "grad_norm": 0.13277407792473037,
+      "learning_rate": 4.661798327822713e-06,
+      "loss": 0.78,
+      "step": 1366
+    },
+    {
+      "epoch": 0.678664515328286,
+      "grad_norm": 0.13349951689576497,
+      "learning_rate": 4.661307283515528e-06,
+      "loss": 0.7362,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6791609780315254,
+      "grad_norm": 0.1454330089242611,
+      "learning_rate": 4.660815908887179e-06,
+      "loss": 0.786,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6796574407347648,
+      "grad_norm": 0.13805559122664302,
+      "learning_rate": 4.660324204012764e-06,
+      "loss": 0.7168,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6801539034380042,
+      "grad_norm": 0.1390252829777148,
+      "learning_rate": 4.659832168967432e-06,
+      "loss": 0.736,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6806503661412436,
+      "grad_norm": 0.13618470081629033,
+      "learning_rate": 4.659339803826384e-06,
+      "loss": 0.7786,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6811468288444831,
+      "grad_norm": 0.13932005962814825,
+      "learning_rate": 4.658847108664869e-06,
+      "loss": 0.7333,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6816432915477225,
+      "grad_norm": 0.1357691607502142,
+      "learning_rate": 4.6583540835581885e-06,
+      "loss": 0.7899,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6821397542509618,
+      "grad_norm": 0.1464322972528331,
+      "learning_rate": 4.657860728581692e-06,
+      "loss": 0.8177,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6826362169542013,
+      "grad_norm": 0.13218140620942265,
+      "learning_rate": 4.657367043810783e-06,
+      "loss": 0.7308,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6831326796574407,
+      "grad_norm": 0.1461276762453906,
+      "learning_rate": 4.656873029320911e-06,
+      "loss": 0.7628,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6836291423606802,
+      "grad_norm": 0.13682220089172875,
+      "learning_rate": 4.656378685187579e-06,
+      "loss": 0.8118,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6841256050639196,
+      "grad_norm": 0.13808113028870317,
+      "learning_rate": 4.655884011486341e-06,
+      "loss": 0.778,
+      "step": 1378
+    },
+    {
+      "epoch": 0.684622067767159,
+      "grad_norm": 0.13141352149483548,
+      "learning_rate": 4.655389008292798e-06,
+      "loss": 0.7884,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6851185304703984,
+      "grad_norm": 0.13773832084927776,
+      "learning_rate": 4.654893675682605e-06,
+      "loss": 0.7103,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6856149931736378,
+      "grad_norm": 0.139138670836966,
+      "learning_rate": 4.654398013731464e-06,
+      "loss": 0.7643,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6861114558768773,
+      "grad_norm": 0.13544047173805013,
+      "learning_rate": 4.65390202251513e-06,
+      "loss": 0.7806,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6866079185801167,
+      "grad_norm": 0.13728555567097042,
+      "learning_rate": 4.653405702109407e-06,
+      "loss": 0.7961,
+      "step": 1383
+    },
+    {
+      "epoch": 0.687104381283356,
+      "grad_norm": 0.13134136227546692,
+      "learning_rate": 4.65290905259015e-06,
+      "loss": 0.7708,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6876008439865955,
+      "grad_norm": 0.14072716822510162,
+      "learning_rate": 4.652412074033263e-06,
+      "loss": 0.7598,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6880973066898349,
+      "grad_norm": 0.18508181188821365,
+      "learning_rate": 4.651914766514703e-06,
+      "loss": 0.7481,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6885937693930744,
+      "grad_norm": 0.13447285883304055,
+      "learning_rate": 4.651417130110473e-06,
+      "loss": 0.7048,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6890902320963138,
+      "grad_norm": 0.13797950483454202,
+      "learning_rate": 4.65091916489663e-06,
+      "loss": 0.7958,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6895866947995531,
+      "grad_norm": 0.14969109022454788,
+      "learning_rate": 4.65042087094928e-06,
+      "loss": 0.8282,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6900831575027926,
+      "grad_norm": 0.13393294593783833,
+      "learning_rate": 4.64992224834458e-06,
+      "loss": 0.6999,
+      "step": 1390
+    },
+    {
+      "epoch": 0.690579620206032,
+      "grad_norm": 0.13957414144835012,
+      "learning_rate": 4.649423297158736e-06,
+      "loss": 0.8215,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6910760829092715,
+      "grad_norm": 0.13264027359412842,
+      "learning_rate": 4.648924017468003e-06,
+      "loss": 0.7282,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6915725456125109,
+      "grad_norm": 0.13318683074105853,
+      "learning_rate": 4.648424409348691e-06,
+      "loss": 0.7237,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6920690083157502,
+      "grad_norm": 0.14079961861355597,
+      "learning_rate": 4.647924472877154e-06,
+      "loss": 0.7589,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6925654710189897,
+      "grad_norm": 0.13173357758319654,
+      "learning_rate": 4.647424208129801e-06,
+      "loss": 0.7327,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6930619337222291,
+      "grad_norm": 0.13990361296789605,
+      "learning_rate": 4.646923615183089e-06,
+      "loss": 0.7952,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6935583964254686,
+      "grad_norm": 0.13139251687364273,
+      "learning_rate": 4.646422694113526e-06,
+      "loss": 0.7317,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6940548591287079,
+      "grad_norm": 0.14426243953892712,
+      "learning_rate": 4.645921444997669e-06,
+      "loss": 0.807,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6945513218319473,
+      "grad_norm": 0.13812024833123682,
+      "learning_rate": 4.645419867912127e-06,
+      "loss": 0.7351,
+      "step": 1399
+    },
+    {
+      "epoch": 0.6950477845351868,
+      "grad_norm": 0.1382819086009079,
+      "learning_rate": 4.644917962933558e-06,
+      "loss": 0.7554,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6955442472384262,
+      "grad_norm": 0.13000958747168673,
+      "learning_rate": 4.644415730138669e-06,
+      "loss": 0.7277,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6960407099416657,
+      "grad_norm": 0.14153919488215605,
+      "learning_rate": 4.643913169604218e-06,
+      "loss": 0.8176,
+      "step": 1402
+    },
+    {
+      "epoch": 0.696537172644905,
+      "grad_norm": 0.13514290656512232,
+      "learning_rate": 4.643410281407014e-06,
+      "loss": 0.7458,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6970336353481444,
+      "grad_norm": 0.1323956462686169,
+      "learning_rate": 4.642907065623916e-06,
+      "loss": 0.7801,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6975300980513839,
+      "grad_norm": 0.13517969708499833,
+      "learning_rate": 4.642403522331832e-06,
+      "loss": 0.7493,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6980265607546233,
+      "grad_norm": 0.14025387703072958,
+      "learning_rate": 4.641899651607721e-06,
+      "loss": 0.7722,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6985230234578628,
+      "grad_norm": 0.13173070675094867,
+      "learning_rate": 4.64139545352859e-06,
+      "loss": 0.7185,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6990194861611021,
+      "grad_norm": 0.13136540290209375,
+      "learning_rate": 4.640890928171499e-06,
+      "loss": 0.7754,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6995159488643415,
+      "grad_norm": 0.13827112688996815,
+      "learning_rate": 4.640386075613556e-06,
+      "loss": 0.792,
+      "step": 1409
+    },
+    {
+      "epoch": 0.700012411567581,
+      "grad_norm": 0.13975488734948466,
+      "learning_rate": 4.639880895931919e-06,
+      "loss": 0.7988,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7005088742708204,
+      "grad_norm": 0.1336382331212796,
+      "learning_rate": 4.6393753892038e-06,
+      "loss": 0.7215,
+      "step": 1411
+    },
+    {
+      "epoch": 0.7010053369740599,
+      "grad_norm": 0.13724481212117196,
+      "learning_rate": 4.638869555506452e-06,
+      "loss": 0.7844,
+      "step": 1412
+    },
+    {
+      "epoch": 0.7015017996772992,
+      "grad_norm": 0.13404812402428848,
+      "learning_rate": 4.638363394917189e-06,
+      "loss": 0.7224,
+      "step": 1413
+    },
+    {
+      "epoch": 0.7019982623805386,
+      "grad_norm": 0.1407590590379482,
+      "learning_rate": 4.637856907513366e-06,
+      "loss": 0.7599,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7024947250837781,
+      "grad_norm": 0.14010062720805114,
+      "learning_rate": 4.637350093372393e-06,
+      "loss": 0.7773,
+      "step": 1415
+    },
+    {
+      "epoch": 0.7029911877870175,
+      "grad_norm": 0.13348329573150353,
+      "learning_rate": 4.636842952571727e-06,
+      "loss": 0.761,
+      "step": 1416
+    },
+    {
+      "epoch": 0.703487650490257,
+      "grad_norm": 0.14369061438866643,
+      "learning_rate": 4.636335485188879e-06,
+      "loss": 0.7699,
+      "step": 1417
+    },
+    {
+      "epoch": 0.7039841131934963,
+      "grad_norm": 0.13433232422419997,
+      "learning_rate": 4.635827691301404e-06,
+      "loss": 0.7661,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7044805758967357,
+      "grad_norm": 0.14032937963835693,
+      "learning_rate": 4.635319570986913e-06,
+      "loss": 0.7695,
+      "step": 1419
+    },
+    {
+      "epoch": 0.7049770385999752,
+      "grad_norm": 0.13443137094696417,
+      "learning_rate": 4.634811124323062e-06,
+      "loss": 0.7334,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7054735013032146,
+      "grad_norm": 0.1342398609304696,
+      "learning_rate": 4.63430235138756e-06,
+      "loss": 0.7465,
+      "step": 1421
+    },
+    {
+      "epoch": 0.7059699640064541,
+      "grad_norm": 0.13174620950915217,
+      "learning_rate": 4.6337932522581656e-06,
+      "loss": 0.7551,
+      "step": 1422
+    },
+    {
+      "epoch": 0.7064664267096934,
+      "grad_norm": 0.1439021802463562,
+      "learning_rate": 4.633283827012684e-06,
+      "loss": 0.7224,
+      "step": 1423
+    },
+    {
+      "epoch": 0.7069628894129328,
+      "grad_norm": 0.1431482149278234,
+      "learning_rate": 4.632774075728974e-06,
+      "loss": 0.8265,
+      "step": 1424
+    },
+    {
+      "epoch": 0.7074593521161723,
+      "grad_norm": 0.13651431312093584,
+      "learning_rate": 4.632263998484944e-06,
+      "loss": 0.7448,
+      "step": 1425
+    },
+    {
+      "epoch": 0.7079558148194117,
+      "grad_norm": 0.13562176642096896,
+      "learning_rate": 4.63175359535855e-06,
+      "loss": 0.731,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7084522775226512,
+      "grad_norm": 0.1331841089582722,
+      "learning_rate": 4.631242866427798e-06,
+      "loss": 0.7785,
+      "step": 1427
+    },
+    {
+      "epoch": 0.7089487402258905,
+      "grad_norm": 0.13572962352737827,
+      "learning_rate": 4.6307318117707465e-06,
+      "loss": 0.7472,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7094452029291299,
+      "grad_norm": 0.1337025002321428,
+      "learning_rate": 4.630220431465501e-06,
+      "loss": 0.7177,
+      "step": 1429
+    },
+    {
+      "epoch": 0.7099416656323694,
+      "grad_norm": 0.13714690654454087,
+      "learning_rate": 4.629708725590219e-06,
+      "loss": 0.7595,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7104381283356088,
+      "grad_norm": 0.13237056679093584,
+      "learning_rate": 4.629196694223104e-06,
+      "loss": 0.7719,
+      "step": 1431
+    },
+    {
+      "epoch": 0.7109345910388482,
+      "grad_norm": 0.1347783497383712,
+      "learning_rate": 4.628684337442414e-06,
+      "loss": 0.7559,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7114310537420876,
+      "grad_norm": 0.14212353203573067,
+      "learning_rate": 4.6281716553264535e-06,
+      "loss": 0.7647,
+      "step": 1433
+    },
+    {
+      "epoch": 0.711927516445327,
+      "grad_norm": 0.1391353138825361,
+      "learning_rate": 4.627658647953579e-06,
+      "loss": 0.7869,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7124239791485665,
+      "grad_norm": 0.1367605845380993,
+      "learning_rate": 4.6271453154021936e-06,
+      "loss": 0.7652,
+      "step": 1435
+    },
+    {
+      "epoch": 0.7129204418518059,
+      "grad_norm": 0.13776225189007474,
+      "learning_rate": 4.626631657750754e-06,
+      "loss": 0.7352,
+      "step": 1436
+    },
+    {
+      "epoch": 0.7134169045550453,
+      "grad_norm": 0.15623278977035873,
+      "learning_rate": 4.626117675077762e-06,
+      "loss": 0.772,
+      "step": 1437
+    },
+    {
+      "epoch": 0.7139133672582847,
+      "grad_norm": 0.13542364624954573,
+      "learning_rate": 4.625603367461775e-06,
+      "loss": 0.755,
+      "step": 1438
+    },
+    {
+      "epoch": 0.7144098299615241,
+      "grad_norm": 0.1363585075610717,
+      "learning_rate": 4.6250887349813935e-06,
+      "loss": 0.7923,
+      "step": 1439
+    },
+    {
+      "epoch": 0.7149062926647636,
+      "grad_norm": 0.1349317484055486,
+      "learning_rate": 4.6245737777152725e-06,
+      "loss": 0.737,
+      "step": 1440
+    },
+    {
+      "epoch": 0.715402755368003,
+      "grad_norm": 0.12840691438132237,
+      "learning_rate": 4.624058495742115e-06,
+      "loss": 0.7329,
+      "step": 1441
+    },
+    {
+      "epoch": 0.7158992180712423,
+      "grad_norm": 0.14115995132355305,
+      "learning_rate": 4.623542889140671e-06,
+      "loss": 0.7799,
+      "step": 1442
+    },
+    {
+      "epoch": 0.7163956807744818,
+      "grad_norm": 0.1357743633796934,
+      "learning_rate": 4.623026957989746e-06,
+      "loss": 0.7492,
+      "step": 1443
+    },
+    {
+      "epoch": 0.7168921434777212,
+      "grad_norm": 0.140747920382236,
+      "learning_rate": 4.622510702368191e-06,
+      "loss": 0.7822,
+      "step": 1444
+    },
+    {
+      "epoch": 0.7173886061809607,
+      "grad_norm": 0.13586217424188604,
+      "learning_rate": 4.621994122354907e-06,
+      "loss": 0.737,
+      "step": 1445
+    },
+    {
+      "epoch": 0.7178850688842001,
+      "grad_norm": 0.1319421489214161,
+      "learning_rate": 4.621477218028845e-06,
+      "loss": 0.7439,
+      "step": 1446
+    },
+    {
+      "epoch": 0.7183815315874394,
+      "grad_norm": 0.1467289582504143,
+      "learning_rate": 4.620959989469005e-06,
+      "loss": 0.7969,
+      "step": 1447
+    },
+    {
+      "epoch": 0.7188779942906789,
+      "grad_norm": 0.13593369108799191,
+      "learning_rate": 4.620442436754438e-06,
+      "loss": 0.7792,
+      "step": 1448
+    },
+    {
+      "epoch": 0.7193744569939183,
+      "grad_norm": 0.13667469147949424,
+      "learning_rate": 4.619924559964243e-06,
+      "loss": 0.7416,
+      "step": 1449
+    },
+    {
+      "epoch": 0.7198709196971578,
+      "grad_norm": 0.13941735212726125,
+      "learning_rate": 4.61940635917757e-06,
+      "loss": 0.7836,
+      "step": 1450
+    },
+    {
+      "epoch": 0.7203673824003972,
+      "grad_norm": 0.14522609940766362,
+      "learning_rate": 4.618887834473616e-06,
+      "loss": 0.8191,
+      "step": 1451
+    },
+    {
+      "epoch": 0.7208638451036365,
+      "grad_norm": 0.1347623385534868,
+      "learning_rate": 4.618368985931631e-06,
+      "loss": 0.72,
+      "step": 1452
+    },
+    {
+      "epoch": 0.721360307806876,
+      "grad_norm": 0.13229916497360666,
+      "learning_rate": 4.617849813630913e-06,
+      "loss": 0.7019,
+      "step": 1453
+    },
+    {
+      "epoch": 0.7218567705101154,
+      "grad_norm": 0.13890901285175117,
+      "learning_rate": 4.617330317650806e-06,
+      "loss": 0.7371,
+      "step": 1454
+    },
+    {
+      "epoch": 0.7223532332133549,
+      "grad_norm": 0.13269598761541726,
+      "learning_rate": 4.6168104980707105e-06,
+      "loss": 0.7291,
+      "step": 1455
+    },
+    {
+      "epoch": 0.7228496959165943,
+      "grad_norm": 0.1348717930791648,
+      "learning_rate": 4.61629035497007e-06,
+      "loss": 0.7328,
+      "step": 1456
+    },
+    {
+      "epoch": 0.7233461586198336,
+      "grad_norm": 0.13798017679082805,
+      "learning_rate": 4.615769888428382e-06,
+      "loss": 0.7664,
+      "step": 1457
+    },
+    {
+      "epoch": 0.7238426213230731,
+      "grad_norm": 0.13280778392582876,
+      "learning_rate": 4.615249098525189e-06,
+      "loss": 0.73,
+      "step": 1458
+    },
+    {
+      "epoch": 0.7243390840263125,
+      "grad_norm": 0.13189279813229704,
+      "learning_rate": 4.614727985340087e-06,
+      "loss": 0.7065,
+      "step": 1459
+    },
+    {
+      "epoch": 0.724835546729552,
+      "grad_norm": 0.1378715254587552,
+      "learning_rate": 4.61420654895272e-06,
+      "loss": 0.7774,
+      "step": 1460
+    },
+    {
+      "epoch": 0.7253320094327914,
+      "grad_norm": 0.1311464168604436,
+      "learning_rate": 4.613684789442781e-06,
+      "loss": 0.7273,
+      "step": 1461
+    },
+    {
+      "epoch": 0.7258284721360307,
+      "grad_norm": 0.14345074089926546,
+      "learning_rate": 4.613162706890011e-06,
+      "loss": 0.7582,
+      "step": 1462
+    },
+    {
+      "epoch": 0.7263249348392702,
+      "grad_norm": 0.13588405827285918,
+      "learning_rate": 4.612640301374204e-06,
+      "loss": 0.7507,
+      "step": 1463
+    },
+    {
+      "epoch": 0.7268213975425096,
+      "grad_norm": 0.13152235282939082,
+      "learning_rate": 4.6121175729752e-06,
+      "loss": 0.7425,
+      "step": 1464
+    },
+    {
+      "epoch": 0.7273178602457491,
+      "grad_norm": 0.1387161072572674,
+      "learning_rate": 4.611594521772891e-06,
+      "loss": 0.7184,
+      "step": 1465
+    },
+    {
+      "epoch": 0.7278143229489885,
+      "grad_norm": 0.13801900811363219,
+      "learning_rate": 4.611071147847216e-06,
+      "loss": 0.7633,
+      "step": 1466
+    },
+    {
+      "epoch": 0.7283107856522278,
+      "grad_norm": 0.2567131236080573,
+      "learning_rate": 4.610547451278164e-06,
+      "loss": 0.7701,
+      "step": 1467
+    },
+    {
+      "epoch": 0.7288072483554673,
+      "grad_norm": 0.14345292271439838,
+      "learning_rate": 4.6100234321457746e-06,
+      "loss": 0.749,
+      "step": 1468
+    },
+    {
+      "epoch": 0.7293037110587067,
+      "grad_norm": 0.13240897000188626,
+      "learning_rate": 4.6094990905301354e-06,
+      "loss": 0.7646,
+      "step": 1469
+    },
+    {
+      "epoch": 0.7298001737619462,
+      "grad_norm": 0.13348526628448198,
+      "learning_rate": 4.608974426511383e-06,
+      "loss": 0.7496,
+      "step": 1470
+    },
+    {
+      "epoch": 0.7302966364651855,
+      "grad_norm": 0.14970955112386783,
+      "learning_rate": 4.608449440169705e-06,
+      "loss": 0.7601,
+      "step": 1471
+    },
+    {
+      "epoch": 0.7307930991684249,
+      "grad_norm": 0.1407703027493505,
+      "learning_rate": 4.607924131585336e-06,
+      "loss": 0.7591,
+      "step": 1472
+    },
+    {
+      "epoch": 0.7312895618716644,
+      "grad_norm": 0.1318082753134835,
+      "learning_rate": 4.607398500838561e-06,
+      "loss": 0.7051,
+      "step": 1473
+    },
+    {
+      "epoch": 0.7317860245749038,
+      "grad_norm": 0.34847062208945645,
+      "learning_rate": 4.606872548009716e-06,
+      "loss": 0.7823,
+      "step": 1474
+    },
+    {
+      "epoch": 0.7322824872781433,
+      "grad_norm": 0.13028506291705924,
+      "learning_rate": 4.606346273179182e-06,
+      "loss": 0.7068,
+      "step": 1475
+    },
+    {
+      "epoch": 0.7327789499813826,
+      "grad_norm": 0.1452834099987651,
+      "learning_rate": 4.605819676427393e-06,
+      "loss": 0.7924,
+      "step": 1476
+    },
+    {
+      "epoch": 0.733275412684622,
+      "grad_norm": 0.13237347766328517,
+      "learning_rate": 4.605292757834832e-06,
+      "loss": 0.7094,
+      "step": 1477
+    },
+    {
+      "epoch": 0.7337718753878615,
+      "grad_norm": 0.14517976496072124,
+      "learning_rate": 4.6047655174820275e-06,
+      "loss": 0.7906,
+      "step": 1478
+    },
+    {
+      "epoch": 0.7342683380911009,
+      "grad_norm": 0.14057990814296611,
+      "learning_rate": 4.604237955449561e-06,
+      "loss": 0.7671,
+      "step": 1479
+    },
+    {
+      "epoch": 0.7347648007943404,
+      "grad_norm": 0.13415285581754496,
+      "learning_rate": 4.603710071818062e-06,
+      "loss": 0.6983,
+      "step": 1480
+    },
+    {
+      "epoch": 0.7352612634975797,
+      "grad_norm": 0.13508721935027235,
+      "learning_rate": 4.603181866668209e-06,
+      "loss": 0.7999,
+      "step": 1481
+    },
+    {
+      "epoch": 0.7357577262008191,
+      "grad_norm": 0.19074847415713053,
+      "learning_rate": 4.60265334008073e-06,
+      "loss": 0.7689,
+      "step": 1482
+    },
+    {
+      "epoch": 0.7362541889040586,
+      "grad_norm": 0.1359795575688059,
+      "learning_rate": 4.602124492136401e-06,
+      "loss": 0.7653,
+      "step": 1483
+    },
+    {
+      "epoch": 0.736750651607298,
+      "grad_norm": 0.15392110510905885,
+      "learning_rate": 4.601595322916049e-06,
+      "loss": 0.7726,
+      "step": 1484
+    },
+    {
+      "epoch": 0.7372471143105375,
+      "grad_norm": 0.1400383089128481,
+      "learning_rate": 4.601065832500548e-06,
+      "loss": 0.7744,
+      "step": 1485
+    },
+    {
+      "epoch": 0.7377435770137768,
+      "grad_norm": 0.14190989377886226,
+      "learning_rate": 4.600536020970822e-06,
+      "loss": 0.779,
+      "step": 1486
+    },
+    {
+      "epoch": 0.7382400397170162,
+      "grad_norm": 0.14143141374332557,
+      "learning_rate": 4.600005888407846e-06,
+      "loss": 0.8004,
+      "step": 1487
+    },
+    {
+      "epoch": 0.7387365024202557,
+      "grad_norm": 0.1448247491292055,
+      "learning_rate": 4.59947543489264e-06,
+      "loss": 0.7369,
+      "step": 1488
+    },
+    {
+      "epoch": 0.7392329651234951,
+      "grad_norm": 0.14353791020576437,
+      "learning_rate": 4.598944660506276e-06,
+      "loss": 0.7517,
+      "step": 1489
+    },
+    {
+      "epoch": 0.7397294278267346,
+      "grad_norm": 0.1397411403947714,
+      "learning_rate": 4.598413565329876e-06,
+      "loss": 0.7872,
+      "step": 1490
+    },
+    {
+      "epoch": 0.7402258905299739,
+      "grad_norm": 0.14686545557180947,
+      "learning_rate": 4.597882149444607e-06,
+      "loss": 0.7655,
+      "step": 1491
+    },
+    {
+      "epoch": 0.7407223532332133,
+      "grad_norm": 0.13888886824301397,
+      "learning_rate": 4.597350412931688e-06,
+      "loss": 0.7759,
+      "step": 1492
+    },
+    {
+      "epoch": 0.7412188159364528,
+      "grad_norm": 0.13368575317341758,
+      "learning_rate": 4.5968183558723876e-06,
+      "loss": 0.7642,
+      "step": 1493
+    },
+    {
+      "epoch": 0.7417152786396922,
+      "grad_norm": 0.13671963488130195,
+      "learning_rate": 4.596285978348022e-06,
+      "loss": 0.8001,
+      "step": 1494
+    },
+    {
+      "epoch": 0.7422117413429317,
+      "grad_norm": 0.13436579260102113,
+      "learning_rate": 4.595753280439955e-06,
+      "loss": 0.7504,
+      "step": 1495
+    },
+    {
+      "epoch": 0.742708204046171,
+      "grad_norm": 0.1350714284302843,
+      "learning_rate": 4.5952202622296015e-06,
+      "loss": 0.7258,
+      "step": 1496
+    },
+    {
+      "epoch": 0.7432046667494104,
+      "grad_norm": 0.14306485237289734,
+      "learning_rate": 4.594686923798426e-06,
+      "loss": 0.7651,
+      "step": 1497
+    },
+    {
+      "epoch": 0.7437011294526499,
+      "grad_norm": 0.13844557571686955,
+      "learning_rate": 4.594153265227941e-06,
+      "loss": 0.7099,
+      "step": 1498
+    },
+    {
+      "epoch": 0.7441975921558893,
+      "grad_norm": 0.13813214807550506,
+      "learning_rate": 4.5936192865997055e-06,
+      "loss": 0.7654,
+      "step": 1499
+    },
+    {
+      "epoch": 0.7446940548591288,
+      "grad_norm": 0.14890702452844934,
+      "learning_rate": 4.59308498799533e-06,
+      "loss": 0.7431,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7451905175623681,
+      "grad_norm": 0.13792041924804616,
+      "learning_rate": 4.592550369496475e-06,
+      "loss": 0.7929,
+      "step": 1501
+    },
+    {
+      "epoch": 0.7456869802656075,
+      "grad_norm": 0.14078699310970763,
+      "learning_rate": 4.592015431184847e-06,
+      "loss": 0.7453,
+      "step": 1502
+    },
+    {
+      "epoch": 0.746183442968847,
+      "grad_norm": 0.14047870303022741,
+      "learning_rate": 4.591480173142204e-06,
+      "loss": 0.7569,
+      "step": 1503
+    },
+    {
+      "epoch": 0.7466799056720864,
+      "grad_norm": 0.1306675590393236,
+      "learning_rate": 4.590944595450351e-06,
+      "loss": 0.7193,
+      "step": 1504
+    },
+    {
+      "epoch": 0.7471763683753258,
+      "grad_norm": 0.1360716461251853,
+      "learning_rate": 4.590408698191142e-06,
+      "loss": 0.7554,
+      "step": 1505
+    },
+    {
+      "epoch": 0.7476728310785652,
+      "grad_norm": 0.13401655261545067,
+      "learning_rate": 4.58987248144648e-06,
+      "loss": 0.7505,
+      "step": 1506
+    },
+    {
+      "epoch": 0.7481692937818046,
+      "grad_norm": 0.13656970997298545,
+      "learning_rate": 4.589335945298318e-06,
+      "loss": 0.7943,
+      "step": 1507
+    },
+    {
+      "epoch": 0.7486657564850441,
+      "grad_norm": 0.13808934356621622,
+      "learning_rate": 4.588799089828657e-06,
+      "loss": 0.8049,
+      "step": 1508
+    },
+    {
+      "epoch": 0.7491622191882835,
+      "grad_norm": 0.14038587528162197,
+      "learning_rate": 4.588261915119547e-06,
+      "loss": 0.7568,
+      "step": 1509
+    },
+    {
+      "epoch": 0.7496586818915228,
+      "grad_norm": 0.13223604221937782,
+      "learning_rate": 4.587724421253085e-06,
+      "loss": 0.7214,
+      "step": 1510
+    },
+    {
+      "epoch": 0.7501551445947623,
+      "grad_norm": 0.13312725764514682,
+      "learning_rate": 4.5871866083114206e-06,
+      "loss": 0.7061,
+      "step": 1511
+    },
+    {
+      "epoch": 0.7506516072980017,
+      "grad_norm": 0.13511067606997035,
+      "learning_rate": 4.586648476376747e-06,
+      "loss": 0.7755,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7506516072980017,
+      "eval_loss": 0.7611762285232544,
+      "eval_runtime": 135.5972,
+      "eval_samples_per_second": 223.847,
+      "eval_steps_per_second": 27.987,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7511480700012412,
+      "grad_norm": 0.1379636114182231,
+      "learning_rate": 4.586110025531312e-06,
+      "loss": 0.7179,
+      "step": 1513
+    },
+    {
+      "epoch": 0.7516445327044806,
+      "grad_norm": 0.13856904591246538,
+      "learning_rate": 4.585571255857408e-06,
+      "loss": 0.7782,
+      "step": 1514
+    },
+    {
+      "epoch": 0.75214099540772,
+      "grad_norm": 0.1338118052517091,
+      "learning_rate": 4.585032167437375e-06,
+      "loss": 0.7429,
+      "step": 1515
+    },
+    {
+      "epoch": 0.7526374581109594,
+      "grad_norm": 0.12995673319277454,
+      "learning_rate": 4.584492760353607e-06,
+      "loss": 0.7371,
+      "step": 1516
+    },
+    {
+      "epoch": 0.7531339208141988,
+      "grad_norm": 0.13175985923680122,
+      "learning_rate": 4.583953034688544e-06,
+      "loss": 0.7269,
+      "step": 1517
+    },
+    {
+      "epoch": 0.7536303835174383,
+      "grad_norm": 0.13337482313493468,
+      "learning_rate": 4.5834129905246725e-06,
+      "loss": 0.7407,
+      "step": 1518
+    },
+    {
+      "epoch": 0.7541268462206777,
+      "grad_norm": 0.13450828654016309,
+      "learning_rate": 4.582872627944531e-06,
+      "loss": 0.7411,
+      "step": 1519
+    },
+    {
+      "epoch": 0.754623308923917,
+      "grad_norm": 0.13585453398292363,
+      "learning_rate": 4.582331947030704e-06,
+      "loss": 0.7709,
+      "step": 1520
+    },
+    {
+      "epoch": 0.7551197716271565,
+      "grad_norm": 0.16174753448144683,
+      "learning_rate": 4.581790947865827e-06,
+      "loss": 0.7486,
+      "step": 1521
+    },
+    {
+      "epoch": 0.7556162343303959,
+      "grad_norm": 0.13477064890424165,
+      "learning_rate": 4.581249630532582e-06,
+      "loss": 0.7218,
+      "step": 1522
+    },
+    {
+      "epoch": 0.7561126970336354,
+      "grad_norm": 0.14221417784622448,
+      "learning_rate": 4.580707995113703e-06,
+      "loss": 0.7464,
+      "step": 1523
+    },
+    {
+      "epoch": 0.7566091597368748,
+      "grad_norm": 0.1471611979618392,
+      "learning_rate": 4.580166041691966e-06,
+      "loss": 0.7741,
+      "step": 1524
+    },
+    {
+      "epoch": 0.7571056224401141,
+      "grad_norm": 0.13565970403363753,
+      "learning_rate": 4.579623770350205e-06,
+      "loss": 0.7506,
+      "step": 1525
+    },
+    {
+      "epoch": 0.7576020851433536,
+      "grad_norm": 0.14016327819488653,
+      "learning_rate": 4.579081181171292e-06,
+      "loss": 0.7694,
+      "step": 1526
+    },
+    {
+      "epoch": 0.758098547846593,
+      "grad_norm": 0.13870513168528667,
+      "learning_rate": 4.5785382742381586e-06,
+      "loss": 0.7733,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7585950105498325,
+      "grad_norm": 0.14061847589896362,
+      "learning_rate": 4.577995049633776e-06,
+      "loss": 0.708,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7590914732530719,
+      "grad_norm": 0.12927094131164107,
+      "learning_rate": 4.577451507441167e-06,
+      "loss": 0.7336,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7595879359563112,
+      "grad_norm": 0.13010758376806392,
+      "learning_rate": 4.576907647743406e-06,
+      "loss": 0.7284,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7600843986595507,
+      "grad_norm": 0.13474374832851801,
+      "learning_rate": 4.576363470623612e-06,
+      "loss": 0.7211,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7605808613627901,
+      "grad_norm": 0.13588121835212055,
+      "learning_rate": 4.575818976164952e-06,
+      "loss": 0.7767,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7610773240660296,
+      "grad_norm": 0.13581473484517131,
+      "learning_rate": 4.575274164450645e-06,
+      "loss": 0.7967,
+      "step": 1533
+    },
+    {
+      "epoch": 0.761573786769269,
+      "grad_norm": 0.1319619163329429,
+      "learning_rate": 4.574729035563957e-06,
+      "loss": 0.7489,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7620702494725083,
+      "grad_norm": 0.13070368454070141,
+      "learning_rate": 4.574183589588202e-06,
+      "loss": 0.7371,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7625667121757478,
+      "grad_norm": 0.14405906001792207,
+      "learning_rate": 4.573637826606742e-06,
+      "loss": 0.8019,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7630631748789872,
+      "grad_norm": 0.13830904744938607,
+      "learning_rate": 4.573091746702988e-06,
+      "loss": 0.8099,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7635596375822267,
+      "grad_norm": 0.1308369875528322,
+      "learning_rate": 4.572545349960401e-06,
+      "loss": 0.7099,
+      "step": 1538
+    },
+    {
+      "epoch": 0.764056100285466,
+      "grad_norm": 0.13746899245422908,
+      "learning_rate": 4.571998636462487e-06,
+      "loss": 0.7422,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7645525629887054,
+      "grad_norm": 0.13435052506307477,
+      "learning_rate": 4.571451606292803e-06,
+      "loss": 0.7412,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7650490256919449,
+      "grad_norm": 0.14575768487845997,
+      "learning_rate": 4.570904259534955e-06,
+      "loss": 0.799,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7655454883951843,
+      "grad_norm": 0.13296485164519237,
+      "learning_rate": 4.570356596272596e-06,
+      "loss": 0.7189,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7660419510984238,
+      "grad_norm": 0.13484403439852938,
+      "learning_rate": 4.569808616589426e-06,
+      "loss": 0.7768,
+      "step": 1543
+    },
+    {
+      "epoch": 0.7665384138016631,
+      "grad_norm": 0.1332711897109076,
+      "learning_rate": 4.569260320569196e-06,
+      "loss": 0.756,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7670348765049025,
+      "grad_norm": 0.1419358962518124,
+      "learning_rate": 4.568711708295704e-06,
+      "loss": 0.7668,
+      "step": 1545
+    },
+    {
+      "epoch": 0.767531339208142,
+      "grad_norm": 0.13596711085205604,
+      "learning_rate": 4.5681627798527965e-06,
+      "loss": 0.7775,
+      "step": 1546
+    },
+    {
+      "epoch": 0.7680278019113814,
+      "grad_norm": 0.13788837250657413,
+      "learning_rate": 4.5676135353243685e-06,
+      "loss": 0.724,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7685242646146209,
+      "grad_norm": 0.13462158663256213,
+      "learning_rate": 4.567063974794363e-06,
+      "loss": 0.7606,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7690207273178602,
+      "grad_norm": 0.14640713050413612,
+      "learning_rate": 4.566514098346774e-06,
+      "loss": 0.75,
+      "step": 1549
+    },
+    {
+      "epoch": 0.7695171900210996,
+      "grad_norm": 0.14008125739494734,
+      "learning_rate": 4.565963906065637e-06,
+      "loss": 0.7769,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7700136527243391,
+      "grad_norm": 0.13457373398967837,
+      "learning_rate": 4.565413398035043e-06,
+      "loss": 0.7223,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7705101154275785,
+      "grad_norm": 0.12909038799878353,
+      "learning_rate": 4.564862574339126e-06,
+      "loss": 0.7108,
+      "step": 1552
+    },
+    {
+      "epoch": 0.771006578130818,
+      "grad_norm": 0.15132140494001553,
+      "learning_rate": 4.564311435062074e-06,
+      "loss": 0.7578,
+      "step": 1553
+    },
+    {
+      "epoch": 0.7715030408340573,
+      "grad_norm": 0.13338663206299184,
+      "learning_rate": 4.563759980288117e-06,
+      "loss": 0.7337,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7719995035372967,
+      "grad_norm": 0.13671337018894522,
+      "learning_rate": 4.563208210101536e-06,
+      "loss": 0.7435,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7724959662405362,
+      "grad_norm": 0.1339512782588895,
+      "learning_rate": 4.562656124586663e-06,
+      "loss": 0.734,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7729924289437756,
+      "grad_norm": 0.13189544572945966,
+      "learning_rate": 4.562103723827872e-06,
+      "loss": 0.7639,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7734888916470151,
+      "grad_norm": 0.13412150276620746,
+      "learning_rate": 4.561551007909592e-06,
+      "loss": 0.7409,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7739853543502544,
+      "grad_norm": 0.13352894722705605,
+      "learning_rate": 4.560997976916293e-06,
+      "loss": 0.7254,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7744818170534938,
+      "grad_norm": 0.13385714130707613,
+      "learning_rate": 4.560444630932499e-06,
+      "loss": 0.7355,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7749782797567333,
+      "grad_norm": 0.14383353757373613,
+      "learning_rate": 4.5598909700427805e-06,
+      "loss": 0.7632,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7754747424599727,
+      "grad_norm": 0.13057620170286605,
+      "learning_rate": 4.559336994331755e-06,
+      "loss": 0.74,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7759712051632122,
+      "grad_norm": 0.22403616831916165,
+      "learning_rate": 4.558782703884089e-06,
+      "loss": 0.7815,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7764676678664515,
+      "grad_norm": 0.13988728351688395,
+      "learning_rate": 4.558228098784496e-06,
+      "loss": 0.8111,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7769641305696909,
+      "grad_norm": 0.12851052315735476,
+      "learning_rate": 4.55767317911774e-06,
+      "loss": 0.7076,
+      "step": 1565
+    },
+    {
+      "epoch": 0.7774605932729304,
+      "grad_norm": 0.13445147845150343,
+      "learning_rate": 4.557117944968631e-06,
+      "loss": 0.7776,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7779570559761698,
+      "grad_norm": 0.13626735223788075,
+      "learning_rate": 4.556562396422027e-06,
+      "loss": 0.7418,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7784535186794093,
+      "grad_norm": 0.13902885288869005,
+      "learning_rate": 4.5560065335628356e-06,
+      "loss": 0.7585,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7789499813826486,
+      "grad_norm": 0.1347101759097276,
+      "learning_rate": 4.55545035647601e-06,
+      "loss": 0.7497,
+      "step": 1569
+    },
+    {
+      "epoch": 0.779446444085888,
+      "grad_norm": 0.14453358307933684,
+      "learning_rate": 4.5548938652465555e-06,
+      "loss": 0.7764,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7799429067891275,
+      "grad_norm": 0.13507717127592933,
+      "learning_rate": 4.55433705995952e-06,
+      "loss": 0.7964,
+      "step": 1571
+    },
+    {
+      "epoch": 0.7804393694923669,
+      "grad_norm": 0.13356730248477877,
+      "learning_rate": 4.553779940700005e-06,
+      "loss": 0.7519,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7809358321956062,
+      "grad_norm": 0.14113785981123475,
+      "learning_rate": 4.553222507553155e-06,
+      "loss": 0.6896,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7814322948988457,
+      "grad_norm": 0.13798920017452704,
+      "learning_rate": 4.552664760604167e-06,
+      "loss": 0.753,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7819287576020851,
+      "grad_norm": 0.13634927023321175,
+      "learning_rate": 4.552106699938281e-06,
+      "loss": 0.7779,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7824252203053246,
+      "grad_norm": 0.13862438111784472,
+      "learning_rate": 4.551548325640789e-06,
+      "loss": 0.7653,
+      "step": 1576
+    },
+    {
+      "epoch": 0.782921683008564,
+      "grad_norm": 0.13229593912177584,
+      "learning_rate": 4.550989637797031e-06,
+      "loss": 0.7289,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7834181457118033,
+      "grad_norm": 0.13125159194957922,
+      "learning_rate": 4.55043063649239e-06,
+      "loss": 0.7287,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7839146084150428,
+      "grad_norm": 0.1417064775776418,
+      "learning_rate": 4.549871321812304e-06,
+      "loss": 0.7426,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7844110711182822,
+      "grad_norm": 0.16883124210144626,
+      "learning_rate": 4.549311693842252e-06,
+      "loss": 0.8431,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7849075338215217,
+      "grad_norm": 0.13414566338046124,
+      "learning_rate": 4.548751752667767e-06,
+      "loss": 0.8359,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7854039965247611,
+      "grad_norm": 0.13725844757239852,
+      "learning_rate": 4.548191498374425e-06,
+      "loss": 0.755,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7859004592280004,
+      "grad_norm": 0.1361745571866967,
+      "learning_rate": 4.547630931047853e-06,
+      "loss": 0.7442,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7863969219312399,
+      "grad_norm": 0.13750113879876855,
+      "learning_rate": 4.547070050773725e-06,
+      "loss": 0.7196,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7868933846344793,
+      "grad_norm": 0.13909283760787944,
+      "learning_rate": 4.5465088576377614e-06,
+      "loss": 0.77,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7873898473377188,
+      "grad_norm": 0.13324992099967756,
+      "learning_rate": 4.545947351725732e-06,
+      "loss": 0.7228,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7878863100409582,
+      "grad_norm": 0.13413613523960244,
+      "learning_rate": 4.5453855331234555e-06,
+      "loss": 0.7636,
+      "step": 1587
+    },
+    {
+      "epoch": 0.7883827727441975,
+      "grad_norm": 0.13214222144352114,
+      "learning_rate": 4.544823401916794e-06,
+      "loss": 0.7119,
+      "step": 1588
+    },
+    {
+      "epoch": 0.788879235447437,
+      "grad_norm": 0.1347073970533374,
+      "learning_rate": 4.544260958191663e-06,
+      "loss": 0.7696,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7893756981506764,
+      "grad_norm": 0.1322596137845652,
+      "learning_rate": 4.543698202034021e-06,
+      "loss": 0.7173,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7898721608539159,
+      "grad_norm": 0.13812558700775102,
+      "learning_rate": 4.543135133529878e-06,
+      "loss": 0.7625,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7903686235571553,
+      "grad_norm": 0.13570198883227674,
+      "learning_rate": 4.542571752765288e-06,
+      "loss": 0.8061,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7908650862603946,
+      "grad_norm": 0.1502291763200914,
+      "learning_rate": 4.542008059826356e-06,
+      "loss": 0.7479,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7913615489636341,
+      "grad_norm": 0.14188040098609095,
+      "learning_rate": 4.5414440547992325e-06,
+      "loss": 0.7453,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7918580116668735,
+      "grad_norm": 0.12979186868674533,
+      "learning_rate": 4.540879737770118e-06,
+      "loss": 0.7411,
+      "step": 1595
+    },
+    {
+      "epoch": 0.792354474370113,
+      "grad_norm": 0.13210637876846007,
+      "learning_rate": 4.540315108825258e-06,
+      "loss": 0.7322,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7928509370733524,
+      "grad_norm": 0.13744920964363747,
+      "learning_rate": 4.539750168050949e-06,
+      "loss": 0.7314,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7933473997765917,
+      "grad_norm": 0.13854762530674888,
+      "learning_rate": 4.539184915533531e-06,
+      "loss": 0.7586,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7938438624798312,
+      "grad_norm": 0.14216681342036058,
+      "learning_rate": 4.538619351359393e-06,
+      "loss": 0.7504,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7943403251830706,
+      "grad_norm": 0.13158698944804878,
+      "learning_rate": 4.538053475614976e-06,
+      "loss": 0.7344,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7948367878863101,
+      "grad_norm": 0.133052297447574,
+      "learning_rate": 4.537487288386763e-06,
+      "loss": 0.7265,
+      "step": 1601
+    },
+    {
+      "epoch": 0.7953332505895495,
+      "grad_norm": 0.14026215243529064,
+      "learning_rate": 4.536920789761286e-06,
+      "loss": 0.7422,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7958297132927888,
+      "grad_norm": 0.13205252140563034,
+      "learning_rate": 4.536353979825125e-06,
+      "loss": 0.7223,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7963261759960283,
+      "grad_norm": 0.12972258505388326,
+      "learning_rate": 4.535786858664909e-06,
+      "loss": 0.7183,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7968226386992677,
+      "grad_norm": 0.14485812514741278,
+      "learning_rate": 4.5352194263673135e-06,
+      "loss": 0.7563,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7973191014025072,
+      "grad_norm": 0.14507460323696492,
+      "learning_rate": 4.534651683019061e-06,
+      "loss": 0.7467,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7978155641057466,
+      "grad_norm": 0.13781694651061116,
+      "learning_rate": 4.534083628706921e-06,
+      "loss": 0.7728,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7983120268089859,
+      "grad_norm": 0.13742843647821426,
+      "learning_rate": 4.533515263517713e-06,
+      "loss": 0.7739,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7988084895122254,
+      "grad_norm": 0.14505009227353202,
+      "learning_rate": 4.532946587538302e-06,
+      "loss": 0.7943,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7993049522154648,
+      "grad_norm": 0.13478499327557095,
+      "learning_rate": 4.532377600855601e-06,
+      "loss": 0.744,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7998014149187043,
+      "grad_norm": 0.13368137981304712,
+      "learning_rate": 4.53180830355657e-06,
+      "loss": 0.7207,
+      "step": 1611
+    },
+    {
+      "epoch": 0.8002978776219436,
+      "grad_norm": 0.1367615628975349,
+      "learning_rate": 4.531238695728218e-06,
+      "loss": 0.7439,
+      "step": 1612
+    },
+    {
+      "epoch": 0.800794340325183,
+      "grad_norm": 0.1414644058349886,
+      "learning_rate": 4.5306687774576e-06,
+      "loss": 0.8015,
+      "step": 1613
+    },
+    {
+      "epoch": 0.8012908030284225,
+      "grad_norm": 0.1367196701213475,
+      "learning_rate": 4.530098548831817e-06,
+      "loss": 0.7596,
+      "step": 1614
+    },
+    {
+      "epoch": 0.8017872657316619,
+      "grad_norm": 0.13522642664685247,
+      "learning_rate": 4.529528009938022e-06,
+      "loss": 0.7498,
+      "step": 1615
+    },
+    {
+      "epoch": 0.8022837284349014,
+      "grad_norm": 0.13889758013185657,
+      "learning_rate": 4.528957160863412e-06,
+      "loss": 0.7589,
+      "step": 1616
+    },
+    {
+      "epoch": 0.8027801911381407,
+      "grad_norm": 0.13673089456510074,
+      "learning_rate": 4.528386001695232e-06,
+      "loss": 0.7202,
+      "step": 1617
+    },
+    {
+      "epoch": 0.8032766538413801,
+      "grad_norm": 0.13642986188283496,
+      "learning_rate": 4.5278145325207735e-06,
+      "loss": 0.7115,
+      "step": 1618
+    },
+    {
+      "epoch": 0.8037731165446196,
+      "grad_norm": 0.13872017776841988,
+      "learning_rate": 4.527242753427378e-06,
+      "loss": 0.7473,
+      "step": 1619
+    },
+    {
+      "epoch": 0.804269579247859,
+      "grad_norm": 0.13515176865610115,
+      "learning_rate": 4.526670664502432e-06,
+      "loss": 0.7328,
+      "step": 1620
+    },
+    {
+      "epoch": 0.8047660419510985,
+      "grad_norm": 0.13643944196282173,
+      "learning_rate": 4.52609826583337e-06,
+      "loss": 0.7182,
+      "step": 1621
+    },
+    {
+      "epoch": 0.8052625046543378,
+      "grad_norm": 0.13670018877356768,
+      "learning_rate": 4.525525557507673e-06,
+      "loss": 0.7556,
+      "step": 1622
+    },
+    {
+      "epoch": 0.8057589673575772,
+      "grad_norm": 0.13504339939566765,
+      "learning_rate": 4.524952539612872e-06,
+      "loss": 0.7584,
+      "step": 1623
+    },
+    {
+      "epoch": 0.8062554300608167,
+      "grad_norm": 0.13368086807436755,
+      "learning_rate": 4.524379212236544e-06,
+      "loss": 0.7667,
+      "step": 1624
+    },
+    {
+      "epoch": 0.8067518927640561,
+      "grad_norm": 0.14428096858814662,
+      "learning_rate": 4.5238055754663105e-06,
+      "loss": 0.8251,
+      "step": 1625
+    },
+    {
+      "epoch": 0.8072483554672956,
+      "grad_norm": 0.14219994124125016,
+      "learning_rate": 4.523231629389845e-06,
+      "loss": 0.739,
+      "step": 1626
+    },
+    {
+      "epoch": 0.8077448181705349,
+      "grad_norm": 0.1326897765805104,
+      "learning_rate": 4.522657374094864e-06,
+      "loss": 0.7558,
+      "step": 1627
+    },
+    {
+      "epoch": 0.8082412808737743,
+      "grad_norm": 0.13799541492346612,
+      "learning_rate": 4.522082809669135e-06,
+      "loss": 0.7563,
+      "step": 1628
+    },
+    {
+      "epoch": 0.8087377435770138,
+      "grad_norm": 0.13141846846705796,
+      "learning_rate": 4.52150793620047e-06,
+      "loss": 0.6994,
+      "step": 1629
+    },
+    {
+      "epoch": 0.8092342062802532,
+      "grad_norm": 0.14133853029523705,
+      "learning_rate": 4.520932753776729e-06,
+      "loss": 0.7406,
+      "step": 1630
+    },
+    {
+      "epoch": 0.8097306689834927,
+      "grad_norm": 0.14262392679819444,
+      "learning_rate": 4.52035726248582e-06,
+      "loss": 0.7419,
+      "step": 1631
+    },
+    {
+      "epoch": 0.810227131686732,
+      "grad_norm": 0.13946667026700096,
+      "learning_rate": 4.519781462415698e-06,
+      "loss": 0.7623,
+      "step": 1632
+    },
+    {
+      "epoch": 0.8107235943899714,
+      "grad_norm": 0.14076729287492634,
+      "learning_rate": 4.5192053536543636e-06,
+      "loss": 0.7801,
+      "step": 1633
+    },
+    {
+      "epoch": 0.8112200570932109,
+      "grad_norm": 0.132725847901949,
+      "learning_rate": 4.518628936289867e-06,
+      "loss": 0.7443,
+      "step": 1634
+    },
+    {
+      "epoch": 0.8117165197964503,
+      "grad_norm": 0.13691408254882964,
+      "learning_rate": 4.518052210410303e-06,
+      "loss": 0.7553,
+      "step": 1635
+    },
+    {
+      "epoch": 0.8122129824996898,
+      "grad_norm": 0.13655790072644422,
+      "learning_rate": 4.517475176103816e-06,
+      "loss": 0.7799,
+      "step": 1636
+    },
+    {
+      "epoch": 0.8127094452029291,
+      "grad_norm": 0.13537469163429455,
+      "learning_rate": 4.5168978334585955e-06,
+      "loss": 0.7422,
+      "step": 1637
+    },
+    {
+      "epoch": 0.8132059079061685,
+      "grad_norm": 0.13341240046444236,
+      "learning_rate": 4.5163201825628805e-06,
+      "loss": 0.7611,
+      "step": 1638
+    },
+    {
+      "epoch": 0.813702370609408,
+      "grad_norm": 0.1330300275048859,
+      "learning_rate": 4.515742223504954e-06,
+      "loss": 0.7526,
+      "step": 1639
+    },
+    {
+      "epoch": 0.8141988333126474,
+      "grad_norm": 0.1289270904071567,
+      "learning_rate": 4.51516395637315e-06,
+      "loss": 0.6907,
+      "step": 1640
+    },
+    {
+      "epoch": 0.8146952960158869,
+      "grad_norm": 0.13653838617553693,
+      "learning_rate": 4.514585381255845e-06,
+      "loss": 0.7272,
+      "step": 1641
+    },
+    {
+      "epoch": 0.8151917587191262,
+      "grad_norm": 0.14220864005729456,
+      "learning_rate": 4.514006498241465e-06,
+      "loss": 0.7533,
+      "step": 1642
+    },
+    {
+      "epoch": 0.8156882214223656,
+      "grad_norm": 0.1334721980787763,
+      "learning_rate": 4.513427307418485e-06,
+      "loss": 0.7727,
+      "step": 1643
+    },
+    {
+      "epoch": 0.8161846841256051,
+      "grad_norm": 0.13759063653846024,
+      "learning_rate": 4.512847808875424e-06,
+      "loss": 0.7862,
+      "step": 1644
+    },
+    {
+      "epoch": 0.8166811468288445,
+      "grad_norm": 0.13689081776996725,
+      "learning_rate": 4.512268002700848e-06,
+      "loss": 0.7217,
+      "step": 1645
+    },
+    {
+      "epoch": 0.8171776095320838,
+      "grad_norm": 0.1447472368767314,
+      "learning_rate": 4.5116878889833735e-06,
+      "loss": 0.7671,
+      "step": 1646
+    },
+    {
+      "epoch": 0.8176740722353233,
+      "grad_norm": 0.13925674131676116,
+      "learning_rate": 4.511107467811659e-06,
+      "loss": 0.7461,
+      "step": 1647
+    },
+    {
+      "epoch": 0.8181705349385627,
+      "grad_norm": 0.12956181530658847,
+      "learning_rate": 4.510526739274415e-06,
+      "loss": 0.7147,
+      "step": 1648
+    },
+    {
+      "epoch": 0.8186669976418022,
+      "grad_norm": 0.14156922012695827,
+      "learning_rate": 4.509945703460394e-06,
+      "loss": 0.7724,
+      "step": 1649
+    },
+    {
+      "epoch": 0.8191634603450416,
+      "grad_norm": 0.1337498799210342,
+      "learning_rate": 4.509364360458399e-06,
+      "loss": 0.7494,
+      "step": 1650
+    },
+    {
+      "epoch": 0.819659923048281,
+      "grad_norm": 0.14328558587229834,
+      "learning_rate": 4.50878271035728e-06,
+      "loss": 0.7505,
+      "step": 1651
+    },
+    {
+      "epoch": 0.8201563857515204,
+      "grad_norm": 0.13902637442350718,
+      "learning_rate": 4.508200753245932e-06,
+      "loss": 0.7722,
+      "step": 1652
+    },
+    {
+      "epoch": 0.8206528484547598,
+      "grad_norm": 0.13220848269342542,
+      "learning_rate": 4.507618489213298e-06,
+      "loss": 0.7764,
+      "step": 1653
+    },
+    {
+      "epoch": 0.8211493111579993,
+      "grad_norm": 0.1327753210553815,
+      "learning_rate": 4.507035918348367e-06,
+      "loss": 0.7135,
+      "step": 1654
+    },
+    {
+      "epoch": 0.8216457738612387,
+      "grad_norm": 0.1406321284403388,
+      "learning_rate": 4.506453040740177e-06,
+      "loss": 0.7467,
+      "step": 1655
+    },
+    {
+      "epoch": 0.822142236564478,
+      "grad_norm": 0.13632541689225586,
+      "learning_rate": 4.505869856477811e-06,
+      "loss": 0.749,
+      "step": 1656
+    },
+    {
+      "epoch": 0.8226386992677175,
+      "grad_norm": 0.14246540537344085,
+      "learning_rate": 4.505286365650398e-06,
+      "loss": 0.7677,
+      "step": 1657
+    },
+    {
+      "epoch": 0.8231351619709569,
+      "grad_norm": 0.13667688867874997,
+      "learning_rate": 4.504702568347117e-06,
+      "loss": 0.7307,
+      "step": 1658
+    },
+    {
+      "epoch": 0.8236316246741964,
+      "grad_norm": 0.1431423731753008,
+      "learning_rate": 4.5041184646571915e-06,
+      "loss": 0.7707,
+      "step": 1659
+    },
+    {
+      "epoch": 0.8241280873774358,
+      "grad_norm": 0.14005994273435252,
+      "learning_rate": 4.5035340546698915e-06,
+      "loss": 0.7612,
+      "step": 1660
+    },
+    {
+      "epoch": 0.8246245500806751,
+      "grad_norm": 0.13515045815804208,
+      "learning_rate": 4.502949338474536e-06,
+      "loss": 0.7746,
+      "step": 1661
+    },
+    {
+      "epoch": 0.8251210127839146,
+      "grad_norm": 0.13147659327086783,
+      "learning_rate": 4.50236431616049e-06,
+      "loss": 0.7014,
+      "step": 1662
+    },
+    {
+      "epoch": 0.825617475487154,
+      "grad_norm": 0.13900795898461954,
+      "learning_rate": 4.501778987817162e-06,
+      "loss": 0.7509,
+      "step": 1663
+    },
+    {
+      "epoch": 0.8261139381903935,
+      "grad_norm": 0.139057514709019,
+      "learning_rate": 4.501193353534013e-06,
+      "loss": 0.7598,
+      "step": 1664
+    },
+    {
+      "epoch": 0.8266104008936329,
+      "grad_norm": 0.1431266182489607,
+      "learning_rate": 4.500607413400546e-06,
+      "loss": 0.7914,
+      "step": 1665
+    },
+    {
+      "epoch": 0.8271068635968722,
+      "grad_norm": 0.1334091878020073,
+      "learning_rate": 4.5000211675063134e-06,
+      "loss": 0.7663,
+      "step": 1666
+    },
+    {
+      "epoch": 0.8276033263001117,
+      "grad_norm": 0.13230619594672582,
+      "learning_rate": 4.499434615940913e-06,
+      "loss": 0.751,
+      "step": 1667
+    },
+    {
+      "epoch": 0.8280997890033511,
+      "grad_norm": 0.14337035164539053,
+      "learning_rate": 4.498847758793991e-06,
+      "loss": 0.7694,
+      "step": 1668
+    },
+    {
+      "epoch": 0.8285962517065906,
+      "grad_norm": 0.13666058128401123,
+      "learning_rate": 4.498260596155237e-06,
+      "loss": 0.7791,
+      "step": 1669
+    },
+    {
+      "epoch": 0.82909271440983,
+      "grad_norm": 0.14465851413099565,
+      "learning_rate": 4.497673128114391e-06,
+      "loss": 0.7803,
+      "step": 1670
+    },
+    {
+      "epoch": 0.8295891771130693,
+      "grad_norm": 0.14652649532294804,
+      "learning_rate": 4.497085354761237e-06,
+      "loss": 0.7799,
+      "step": 1671
+    },
+    {
+      "epoch": 0.8300856398163088,
+      "grad_norm": 0.13131787380504173,
+      "learning_rate": 4.4964972761856086e-06,
+      "loss": 0.7452,
+      "step": 1672
+    },
+    {
+      "epoch": 0.8305821025195482,
+      "grad_norm": 0.14145954525931556,
+      "learning_rate": 4.495908892477382e-06,
+      "loss": 0.7794,
+      "step": 1673
+    },
+    {
+      "epoch": 0.8310785652227877,
+      "grad_norm": 0.1384021322282935,
+      "learning_rate": 4.495320203726483e-06,
+      "loss": 0.7108,
+      "step": 1674
+    },
+    {
+      "epoch": 0.8315750279260271,
+      "grad_norm": 0.1415771904329555,
+      "learning_rate": 4.494731210022884e-06,
+      "loss": 0.7311,
+      "step": 1675
+    },
+    {
+      "epoch": 0.8320714906292664,
+      "grad_norm": 0.1438025629578128,
+      "learning_rate": 4.494141911456602e-06,
+      "loss": 0.7992,
+      "step": 1676
+    },
+    {
+      "epoch": 0.8325679533325059,
+      "grad_norm": 0.1414714443428166,
+      "learning_rate": 4.4935523081177035e-06,
+      "loss": 0.7469,
+      "step": 1677
+    },
+    {
+      "epoch": 0.8330644160357453,
+      "grad_norm": 0.14271477816026024,
+      "learning_rate": 4.492962400096298e-06,
+      "loss": 0.8226,
+      "step": 1678
+    },
+    {
+      "epoch": 0.8335608787389848,
+      "grad_norm": 0.1443375820025856,
+      "learning_rate": 4.492372187482545e-06,
+      "loss": 0.8516,
+      "step": 1679
+    },
+    {
+      "epoch": 0.8340573414422241,
+      "grad_norm": 0.13815933437015357,
+      "learning_rate": 4.491781670366648e-06,
+      "loss": 0.8077,
+      "step": 1680
+    },
+    {
+      "epoch": 0.8345538041454635,
+      "grad_norm": 0.1388616626139753,
+      "learning_rate": 4.491190848838858e-06,
+      "loss": 0.7864,
+      "step": 1681
+    },
+    {
+      "epoch": 0.835050266848703,
+      "grad_norm": 0.13570857271523626,
+      "learning_rate": 4.490599722989474e-06,
+      "loss": 0.7412,
+      "step": 1682
+    },
+    {
+      "epoch": 0.8355467295519424,
+      "grad_norm": 0.1438978989274344,
+      "learning_rate": 4.490008292908839e-06,
+      "loss": 0.7208,
+      "step": 1683
+    },
+    {
+      "epoch": 0.8360431922551819,
+      "grad_norm": 0.143547018622754,
+      "learning_rate": 4.4894165586873426e-06,
+      "loss": 0.7816,
+      "step": 1684
+    },
+    {
+      "epoch": 0.8365396549584212,
+      "grad_norm": 0.13635442176917034,
+      "learning_rate": 4.488824520415425e-06,
+      "loss": 0.7406,
+      "step": 1685
+    },
+    {
+      "epoch": 0.8370361176616606,
+      "grad_norm": 0.14562544936887342,
+      "learning_rate": 4.4882321781835666e-06,
+      "loss": 0.7833,
+      "step": 1686
+    },
+    {
+      "epoch": 0.8375325803649001,
+      "grad_norm": 0.14484628055837453,
+      "learning_rate": 4.4876395320822984e-06,
+      "loss": 0.7634,
+      "step": 1687
+    },
+    {
+      "epoch": 0.8380290430681395,
+      "grad_norm": 0.13816444882771123,
+      "learning_rate": 4.487046582202198e-06,
+      "loss": 0.7416,
+      "step": 1688
+    },
+    {
+      "epoch": 0.838525505771379,
+      "grad_norm": 0.14266194907778323,
+      "learning_rate": 4.486453328633887e-06,
+      "loss": 0.7377,
+      "step": 1689
+    },
+    {
+      "epoch": 0.8390219684746183,
+      "grad_norm": 0.14297267750408402,
+      "learning_rate": 4.485859771468035e-06,
+      "loss": 0.8454,
+      "step": 1690
+    },
+    {
+      "epoch": 0.8395184311778577,
+      "grad_norm": 0.1403189883914047,
+      "learning_rate": 4.4852659107953574e-06,
+      "loss": 0.7529,
+      "step": 1691
+    },
+    {
+      "epoch": 0.8400148938810972,
+      "grad_norm": 0.13495181072938314,
+      "learning_rate": 4.484671746706617e-06,
+      "loss": 0.7426,
+      "step": 1692
+    },
+    {
+      "epoch": 0.8405113565843366,
+      "grad_norm": 0.13845259745230812,
+      "learning_rate": 4.484077279292622e-06,
+      "loss": 0.768,
+      "step": 1693
+    },
+    {
+      "epoch": 0.8410078192875761,
+      "grad_norm": 0.1386205743503297,
+      "learning_rate": 4.483482508644228e-06,
+      "loss": 0.7478,
+      "step": 1694
+    },
+    {
+      "epoch": 0.8415042819908154,
+      "grad_norm": 0.14075629580146526,
+      "learning_rate": 4.482887434852334e-06,
+      "loss": 0.7795,
+      "step": 1695
+    },
+    {
+      "epoch": 0.8420007446940548,
+      "grad_norm": 0.13730753156141765,
+      "learning_rate": 4.4822920580078885e-06,
+      "loss": 0.7747,
+      "step": 1696
+    },
+    {
+      "epoch": 0.8424972073972943,
+      "grad_norm": 0.13372905034092983,
+      "learning_rate": 4.481696378201887e-06,
+      "loss": 0.7619,
+      "step": 1697
+    },
+    {
+      "epoch": 0.8429936701005337,
+      "grad_norm": 0.13367135782867837,
+      "learning_rate": 4.481100395525367e-06,
+      "loss": 0.7271,
+      "step": 1698
+    },
+    {
+      "epoch": 0.8434901328037732,
+      "grad_norm": 0.15581577968511073,
+      "learning_rate": 4.4805041100694165e-06,
+      "loss": 0.7648,
+      "step": 1699
+    },
+    {
+      "epoch": 0.8439865955070125,
+      "grad_norm": 0.1401851479710573,
+      "learning_rate": 4.479907521925168e-06,
+      "loss": 0.7723,
+      "step": 1700
+    },
+    {
+      "epoch": 0.8444830582102519,
+      "grad_norm": 0.13725236828476695,
+      "learning_rate": 4.4793106311838e-06,
+      "loss": 0.7257,
+      "step": 1701
+    },
+    {
+      "epoch": 0.8449795209134914,
+      "grad_norm": 0.1356076078010008,
+      "learning_rate": 4.478713437936538e-06,
+      "loss": 0.7374,
+      "step": 1702
+    },
+    {
+      "epoch": 0.8454759836167308,
+      "grad_norm": 0.13425339120243374,
+      "learning_rate": 4.478115942274652e-06,
+      "loss": 0.7241,
+      "step": 1703
+    },
+    {
+      "epoch": 0.8459724463199703,
+      "grad_norm": 0.13679058651123055,
+      "learning_rate": 4.477518144289462e-06,
+      "loss": 0.766,
+      "step": 1704
+    },
+    {
+      "epoch": 0.8464689090232096,
+      "grad_norm": 0.13984700650107668,
+      "learning_rate": 4.476920044072331e-06,
+      "loss": 0.7733,
+      "step": 1705
+    },
+    {
+      "epoch": 0.846965371726449,
+      "grad_norm": 0.13407682285557562,
+      "learning_rate": 4.476321641714669e-06,
+      "loss": 0.7057,
+      "step": 1706
+    },
+    {
+      "epoch": 0.8474618344296885,
+      "grad_norm": 0.13417781616310068,
+      "learning_rate": 4.475722937307931e-06,
+      "loss": 0.7953,
+      "step": 1707
+    },
+    {
+      "epoch": 0.8479582971329279,
+      "grad_norm": 0.13220445140764078,
+      "learning_rate": 4.475123930943621e-06,
+      "loss": 0.6939,
+      "step": 1708
+    },
+    {
+      "epoch": 0.8484547598361674,
+      "grad_norm": 0.13757038634337526,
+      "learning_rate": 4.474524622713286e-06,
+      "loss": 0.7181,
+      "step": 1709
+    },
+    {
+      "epoch": 0.8489512225394067,
+      "grad_norm": 0.1405832726304441,
+      "learning_rate": 4.473925012708522e-06,
+      "loss": 0.7676,
+      "step": 1710
+    },
+    {
+      "epoch": 0.8494476852426461,
+      "grad_norm": 0.14268436213062238,
+      "learning_rate": 4.47332510102097e-06,
+      "loss": 0.7663,
+      "step": 1711
+    },
+    {
+      "epoch": 0.8499441479458856,
+      "grad_norm": 0.13994019393605722,
+      "learning_rate": 4.472724887742316e-06,
+      "loss": 0.7941,
+      "step": 1712
+    },
+    {
+      "epoch": 0.850440610649125,
+      "grad_norm": 0.13811450219788193,
+      "learning_rate": 4.472124372964292e-06,
+      "loss": 0.734,
+      "step": 1713
+    },
+    {
+      "epoch": 0.8509370733523643,
+      "grad_norm": 0.1318935825701855,
+      "learning_rate": 4.471523556778679e-06,
+      "loss": 0.739,
+      "step": 1714
+    },
+    {
+      "epoch": 0.8514335360556038,
+      "grad_norm": 0.1419282323412836,
+      "learning_rate": 4.470922439277301e-06,
+      "loss": 0.7729,
+      "step": 1715
+    },
+    {
+      "epoch": 0.8519299987588432,
+      "grad_norm": 0.1393041197530227,
+      "learning_rate": 4.47032102055203e-06,
+      "loss": 0.7483,
+      "step": 1716
+    },
+    {
+      "epoch": 0.8524264614620827,
+      "grad_norm": 0.13626087249042654,
+      "learning_rate": 4.469719300694783e-06,
+      "loss": 0.7349,
+      "step": 1717
+    },
+    {
+      "epoch": 0.8529229241653221,
+      "grad_norm": 0.1464856140158772,
+      "learning_rate": 4.469117279797522e-06,
+      "loss": 0.836,
+      "step": 1718
+    },
+    {
+      "epoch": 0.8534193868685614,
+      "grad_norm": 0.1409940881870901,
+      "learning_rate": 4.468514957952258e-06,
+      "loss": 0.7472,
+      "step": 1719
+    },
+    {
+      "epoch": 0.8539158495718009,
+      "grad_norm": 0.14285146112331779,
+      "learning_rate": 4.467912335251045e-06,
+      "loss": 0.8371,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8544123122750403,
+      "grad_norm": 0.13806293901326697,
+      "learning_rate": 4.467309411785984e-06,
+      "loss": 0.8033,
+      "step": 1721
+    },
+    {
+      "epoch": 0.8549087749782798,
+      "grad_norm": 0.1361187990693462,
+      "learning_rate": 4.466706187649223e-06,
+      "loss": 0.7249,
+      "step": 1722
+    },
+    {
+      "epoch": 0.8554052376815192,
+      "grad_norm": 0.1392046645207596,
+      "learning_rate": 4.466102662932956e-06,
+      "loss": 0.7949,
+      "step": 1723
+    },
+    {
+      "epoch": 0.8559017003847585,
+      "grad_norm": 0.1402331299424571,
+      "learning_rate": 4.46549883772942e-06,
+      "loss": 0.8077,
+      "step": 1724
+    },
+    {
+      "epoch": 0.856398163087998,
+      "grad_norm": 0.1350641709477664,
+      "learning_rate": 4.464894712130902e-06,
+      "loss": 0.7314,
+      "step": 1725
+    },
+    {
+      "epoch": 0.8568946257912374,
+      "grad_norm": 0.13330313530654128,
+      "learning_rate": 4.464290286229731e-06,
+      "loss": 0.7255,
+      "step": 1726
+    },
+    {
+      "epoch": 0.8573910884944769,
+      "grad_norm": 0.14647673202761563,
+      "learning_rate": 4.463685560118285e-06,
+      "loss": 0.7336,
+      "step": 1727
+    },
+    {
+      "epoch": 0.8578875511977163,
+      "grad_norm": 0.13358377643813468,
+      "learning_rate": 4.463080533888987e-06,
+      "loss": 0.7289,
+      "step": 1728
+    },
+    {
+      "epoch": 0.8583840139009556,
+      "grad_norm": 0.16223228182246507,
+      "learning_rate": 4.4624752076343044e-06,
+      "loss": 0.8098,
+      "step": 1729
+    },
+    {
+      "epoch": 0.8588804766041951,
+      "grad_norm": 0.1362449815913255,
+      "learning_rate": 4.461869581446752e-06,
+      "loss": 0.7671,
+      "step": 1730
+    },
+    {
+      "epoch": 0.8593769393074345,
+      "grad_norm": 0.14081968859441232,
+      "learning_rate": 4.461263655418891e-06,
+      "loss": 0.6943,
+      "step": 1731
+    },
+    {
+      "epoch": 0.859873402010674,
+      "grad_norm": 0.13606349832132275,
+      "learning_rate": 4.460657429643326e-06,
+      "loss": 0.7334,
+      "step": 1732
+    },
+    {
+      "epoch": 0.8603698647139134,
+      "grad_norm": 0.1369487135112441,
+      "learning_rate": 4.460050904212711e-06,
+      "loss": 0.7225,
+      "step": 1733
+    },
+    {
+      "epoch": 0.8608663274171527,
+      "grad_norm": 0.14809100220976384,
+      "learning_rate": 4.45944407921974e-06,
+      "loss": 0.7686,
+      "step": 1734
+    },
+    {
+      "epoch": 0.8613627901203922,
+      "grad_norm": 0.13703527982602742,
+      "learning_rate": 4.458836954757161e-06,
+      "loss": 0.7545,
+      "step": 1735
+    },
+    {
+      "epoch": 0.8618592528236316,
+      "grad_norm": 0.1351821399394306,
+      "learning_rate": 4.4582295309177595e-06,
+      "loss": 0.7797,
+      "step": 1736
+    },
+    {
+      "epoch": 0.8623557155268711,
+      "grad_norm": 0.1354673204843339,
+      "learning_rate": 4.457621807794372e-06,
+      "loss": 0.8066,
+      "step": 1737
+    },
+    {
+      "epoch": 0.8628521782301105,
+      "grad_norm": 0.13862706776026357,
+      "learning_rate": 4.457013785479881e-06,
+      "loss": 0.7333,
+      "step": 1738
+    },
+    {
+      "epoch": 0.8633486409333498,
+      "grad_norm": 0.1389346115440188,
+      "learning_rate": 4.45640546406721e-06,
+      "loss": 0.764,
+      "step": 1739
+    },
+    {
+      "epoch": 0.8638451036365893,
+      "grad_norm": 0.15558347243623133,
+      "learning_rate": 4.455796843649332e-06,
+      "loss": 0.7593,
+      "step": 1740
+    },
+    {
+      "epoch": 0.8643415663398287,
+      "grad_norm": 0.14353099958802945,
+      "learning_rate": 4.455187924319266e-06,
+      "loss": 0.7507,
+      "step": 1741
+    },
+    {
+      "epoch": 0.8648380290430682,
+      "grad_norm": 0.14119369358264125,
+      "learning_rate": 4.454578706170075e-06,
+      "loss": 0.7599,
+      "step": 1742
+    },
+    {
+      "epoch": 0.8653344917463076,
+      "grad_norm": 0.13360839981041983,
+      "learning_rate": 4.453969189294867e-06,
+      "loss": 0.7387,
+      "step": 1743
+    },
+    {
+      "epoch": 0.8658309544495469,
+      "grad_norm": 0.14010427578913046,
+      "learning_rate": 4.453359373786799e-06,
+      "loss": 0.754,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8663274171527864,
+      "grad_norm": 0.13347970173643106,
+      "learning_rate": 4.45274925973907e-06,
+      "loss": 0.7363,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8668238798560258,
+      "grad_norm": 0.143214337774515,
+      "learning_rate": 4.4521388472449265e-06,
+      "loss": 0.74,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8673203425592653,
+      "grad_norm": 0.14170555954343675,
+      "learning_rate": 4.451528136397661e-06,
+      "loss": 0.74,
+      "step": 1747
+    },
+    {
+      "epoch": 0.8678168052625047,
+      "grad_norm": 0.13910626879867338,
+      "learning_rate": 4.450917127290609e-06,
+      "loss": 0.7129,
+      "step": 1748
+    },
+    {
+      "epoch": 0.868313267965744,
+      "grad_norm": 0.14451162598319245,
+      "learning_rate": 4.4503058200171566e-06,
+      "loss": 0.7683,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8688097306689835,
+      "grad_norm": 0.13670177426956756,
+      "learning_rate": 4.449694214670729e-06,
+      "loss": 0.7196,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8693061933722229,
+      "grad_norm": 0.1425374537551891,
+      "learning_rate": 4.449082311344802e-06,
+      "loss": 0.713,
+      "step": 1751
+    },
+    {
+      "epoch": 0.8698026560754624,
+      "grad_norm": 0.13684146510608788,
+      "learning_rate": 4.4484701101328944e-06,
+      "loss": 0.6968,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8702991187787017,
+      "grad_norm": 0.13936638023848308,
+      "learning_rate": 4.447857611128572e-06,
+      "loss": 0.7731,
+      "step": 1753
+    },
+    {
+      "epoch": 0.8707955814819411,
+      "grad_norm": 0.14653517321071075,
+      "learning_rate": 4.447244814425446e-06,
+      "loss": 0.7236,
+      "step": 1754
+    },
+    {
+      "epoch": 0.8712920441851806,
+      "grad_norm": 0.14331030940384665,
+      "learning_rate": 4.446631720117171e-06,
+      "loss": 0.7306,
+      "step": 1755
+    },
+    {
+      "epoch": 0.87178850688842,
+      "grad_norm": 0.14673685332839945,
+      "learning_rate": 4.446018328297449e-06,
+      "loss": 0.7242,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8722849695916595,
+      "grad_norm": 0.1347449091302609,
+      "learning_rate": 4.445404639060028e-06,
+      "loss": 0.7709,
+      "step": 1757
+    },
+    {
+      "epoch": 0.8727814322948988,
+      "grad_norm": 0.1411266352695579,
+      "learning_rate": 4.4447906524987006e-06,
+      "loss": 0.8022,
+      "step": 1758
+    },
+    {
+      "epoch": 0.8732778949981382,
+      "grad_norm": 0.14208496504174006,
+      "learning_rate": 4.444176368707305e-06,
+      "loss": 0.7066,
+      "step": 1759
+    },
+    {
+      "epoch": 0.8737743577013777,
+      "grad_norm": 0.13674759170668294,
+      "learning_rate": 4.443561787779722e-06,
+      "loss": 0.7553,
+      "step": 1760
+    },
+    {
+      "epoch": 0.8742708204046171,
+      "grad_norm": 0.14090838321939578,
+      "learning_rate": 4.442946909809884e-06,
+      "loss": 0.704,
+      "step": 1761
+    },
+    {
+      "epoch": 0.8747672831078566,
+      "grad_norm": 0.14668605063528445,
+      "learning_rate": 4.442331734891763e-06,
+      "loss": 0.7503,
+      "step": 1762
+    },
+    {
+      "epoch": 0.8752637458110959,
+      "grad_norm": 0.1397458175530683,
+      "learning_rate": 4.441716263119379e-06,
+      "loss": 0.7832,
+      "step": 1763
+    },
+    {
+      "epoch": 0.8757602085143353,
+      "grad_norm": 0.138318109593277,
+      "learning_rate": 4.441100494586797e-06,
+      "loss": 0.7535,
+      "step": 1764
+    },
+    {
+      "epoch": 0.8762566712175748,
+      "grad_norm": 0.14189795702675218,
+      "learning_rate": 4.4404844293881285e-06,
+      "loss": 0.7643,
+      "step": 1765
+    },
+    {
+      "epoch": 0.8767531339208142,
+      "grad_norm": 0.13684604258689606,
+      "learning_rate": 4.439868067617528e-06,
+      "loss": 0.7602,
+      "step": 1766
+    },
+    {
+      "epoch": 0.8772495966240537,
+      "grad_norm": 0.13657716615346063,
+      "learning_rate": 4.4392514093691965e-06,
+      "loss": 0.7519,
+      "step": 1767
+    },
+    {
+      "epoch": 0.877746059327293,
+      "grad_norm": 0.16720007998528186,
+      "learning_rate": 4.43863445473738e-06,
+      "loss": 0.7311,
+      "step": 1768
+    },
+    {
+      "epoch": 0.8782425220305324,
+      "grad_norm": 0.13527051039700383,
+      "learning_rate": 4.4380172038163716e-06,
+      "loss": 0.727,
+      "step": 1769
+    },
+    {
+      "epoch": 0.8787389847337719,
+      "grad_norm": 0.13977003542487368,
+      "learning_rate": 4.437399656700507e-06,
+      "loss": 0.7412,
+      "step": 1770
+    },
+    {
+      "epoch": 0.8792354474370113,
+      "grad_norm": 0.14066430808137664,
+      "learning_rate": 4.436781813484169e-06,
+      "loss": 0.7626,
+      "step": 1771
+    },
+    {
+      "epoch": 0.8797319101402508,
+      "grad_norm": 0.14316718597333672,
+      "learning_rate": 4.436163674261785e-06,
+      "loss": 0.783,
+      "step": 1772
+    },
+    {
+      "epoch": 0.8802283728434901,
+      "grad_norm": 0.14181967035602686,
+      "learning_rate": 4.435545239127827e-06,
+      "loss": 0.7452,
+      "step": 1773
+    },
+    {
+      "epoch": 0.8807248355467295,
+      "grad_norm": 0.13364947209125708,
+      "learning_rate": 4.434926508176814e-06,
+      "loss": 0.7446,
+      "step": 1774
+    },
+    {
+      "epoch": 0.881221298249969,
+      "grad_norm": 0.1351435150199207,
+      "learning_rate": 4.434307481503307e-06,
+      "loss": 0.7263,
+      "step": 1775
+    },
+    {
+      "epoch": 0.8817177609532084,
+      "grad_norm": 0.13242522480531707,
+      "learning_rate": 4.433688159201917e-06,
+      "loss": 0.6855,
+      "step": 1776
+    },
+    {
+      "epoch": 0.8822142236564479,
+      "grad_norm": 0.1392138600504658,
+      "learning_rate": 4.433068541367296e-06,
+      "loss": 0.7762,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8827106863596872,
+      "grad_norm": 0.13595036312344588,
+      "learning_rate": 4.432448628094142e-06,
+      "loss": 0.7368,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8832071490629266,
+      "grad_norm": 0.14043982715211084,
+      "learning_rate": 4.4318284194772e-06,
+      "loss": 0.7809,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8837036117661661,
+      "grad_norm": 0.13928983515854973,
+      "learning_rate": 4.431207915611259e-06,
+      "loss": 0.7963,
+      "step": 1780
+    },
+    {
+      "epoch": 0.8842000744694055,
+      "grad_norm": 0.15323129087107526,
+      "learning_rate": 4.430587116591152e-06,
+      "loss": 0.7638,
+      "step": 1781
+    },
+    {
+      "epoch": 0.884696537172645,
+      "grad_norm": 0.13672127380953056,
+      "learning_rate": 4.429966022511759e-06,
+      "loss": 0.7521,
+      "step": 1782
+    },
+    {
+      "epoch": 0.8851929998758843,
+      "grad_norm": 0.13675982862960748,
+      "learning_rate": 4.429344633468005e-06,
+      "loss": 0.735,
+      "step": 1783
+    },
+    {
+      "epoch": 0.8856894625791237,
+      "grad_norm": 0.14468317226179433,
+      "learning_rate": 4.428722949554858e-06,
+      "loss": 0.7728,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8861859252823632,
+      "grad_norm": 0.1392240635874152,
+      "learning_rate": 4.428100970867332e-06,
+      "loss": 0.7631,
+      "step": 1785
+    },
+    {
+      "epoch": 0.8866823879856026,
+      "grad_norm": 0.13905551197842414,
+      "learning_rate": 4.4274786975004886e-06,
+      "loss": 0.7664,
+      "step": 1786
+    },
+    {
+      "epoch": 0.887178850688842,
+      "grad_norm": 0.13218082313555352,
+      "learning_rate": 4.426856129549431e-06,
+      "loss": 0.7664,
+      "step": 1787
+    },
+    {
+      "epoch": 0.8876753133920814,
+      "grad_norm": 0.14542219733133013,
+      "learning_rate": 4.426233267109308e-06,
+      "loss": 0.7537,
+      "step": 1788
+    },
+    {
+      "epoch": 0.8881717760953208,
+      "grad_norm": 0.13937675043101774,
+      "learning_rate": 4.4256101102753154e-06,
+      "loss": 0.7624,
+      "step": 1789
+    },
+    {
+      "epoch": 0.8886682387985603,
+      "grad_norm": 0.14213592075320755,
+      "learning_rate": 4.424986659142691e-06,
+      "loss": 0.7745,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8891647015017997,
+      "grad_norm": 0.13222515034482818,
+      "learning_rate": 4.424362913806723e-06,
+      "loss": 0.7704,
+      "step": 1791
+    },
+    {
+      "epoch": 0.889661164205039,
+      "grad_norm": 0.12997849285210467,
+      "learning_rate": 4.423738874362737e-06,
+      "loss": 0.7473,
+      "step": 1792
+    },
+    {
+      "epoch": 0.8901576269082785,
+      "grad_norm": 0.13865947461800895,
+      "learning_rate": 4.423114540906108e-06,
+      "loss": 0.7922,
+      "step": 1793
+    },
+    {
+      "epoch": 0.8906540896115179,
+      "grad_norm": 0.1400478476922931,
+      "learning_rate": 4.422489913532258e-06,
+      "loss": 0.7136,
+      "step": 1794
+    },
+    {
+      "epoch": 0.8911505523147574,
+      "grad_norm": 0.13354935512552799,
+      "learning_rate": 4.421864992336648e-06,
+      "loss": 0.7358,
+      "step": 1795
+    },
+    {
+      "epoch": 0.8916470150179968,
+      "grad_norm": 0.1503327640511868,
+      "learning_rate": 4.421239777414789e-06,
+      "loss": 0.73,
+      "step": 1796
+    },
+    {
+      "epoch": 0.8921434777212361,
+      "grad_norm": 0.1378043512733515,
+      "learning_rate": 4.4206142688622365e-06,
+      "loss": 0.7978,
+      "step": 1797
+    },
+    {
+      "epoch": 0.8926399404244756,
+      "grad_norm": 0.14901127137101272,
+      "learning_rate": 4.4199884667745866e-06,
+      "loss": 0.8067,
+      "step": 1798
+    },
+    {
+      "epoch": 0.893136403127715,
+      "grad_norm": 0.14348536864057324,
+      "learning_rate": 4.419362371247483e-06,
+      "loss": 0.7077,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8936328658309545,
+      "grad_norm": 0.14175662380934076,
+      "learning_rate": 4.418735982376617e-06,
+      "loss": 0.7913,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8941293285341939,
+      "grad_norm": 0.1358379514515194,
+      "learning_rate": 4.41810930025772e-06,
+      "loss": 0.7867,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8946257912374332,
+      "grad_norm": 0.13785444500689084,
+      "learning_rate": 4.417482324986572e-06,
+      "loss": 0.8063,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8951222539406727,
+      "grad_norm": 0.13967910977720893,
+      "learning_rate": 4.416855056658994e-06,
+      "loss": 0.7678,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8956187166439121,
+      "grad_norm": 0.13147269887815224,
+      "learning_rate": 4.416227495370855e-06,
+      "loss": 0.7064,
+      "step": 1804
+    },
+    {
+      "epoch": 0.8961151793471516,
+      "grad_norm": 0.13960191918704914,
+      "learning_rate": 4.415599641218068e-06,
+      "loss": 0.7231,
+      "step": 1805
+    },
+    {
+      "epoch": 0.896611642050391,
+      "grad_norm": 0.15042506597615363,
+      "learning_rate": 4.4149714942965905e-06,
+      "loss": 0.7629,
+      "step": 1806
+    },
+    {
+      "epoch": 0.8971081047536303,
+      "grad_norm": 0.15012721023542636,
+      "learning_rate": 4.414343054702424e-06,
+      "loss": 0.8244,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8976045674568698,
+      "grad_norm": 0.13677278853363833,
+      "learning_rate": 4.4137143225316155e-06,
+      "loss": 0.7664,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8981010301601092,
+      "grad_norm": 0.13593716948768078,
+      "learning_rate": 4.4130852978802575e-06,
+      "loss": 0.7305,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8985974928633487,
+      "grad_norm": 0.13364129430992383,
+      "learning_rate": 4.412455980844486e-06,
+      "loss": 0.7374,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8990939555665881,
+      "grad_norm": 0.1383082700694781,
+      "learning_rate": 4.4118263715204805e-06,
+      "loss": 0.7336,
+      "step": 1811
+    },
+    {
+      "epoch": 0.8995904182698274,
+      "grad_norm": 0.13423570328137593,
+      "learning_rate": 4.4111964700044684e-06,
+      "loss": 0.7759,
+      "step": 1812
+    },
+    {
+      "epoch": 0.9000868809730669,
+      "grad_norm": 0.1427047436168598,
+      "learning_rate": 4.41056627639272e-06,
+      "loss": 0.7461,
+      "step": 1813
+    },
+    {
+      "epoch": 0.9005833436763063,
+      "grad_norm": 0.14114806673952404,
+      "learning_rate": 4.409935790781549e-06,
+      "loss": 0.8111,
+      "step": 1814
+    },
+    {
+      "epoch": 0.9010798063795458,
+      "grad_norm": 0.13948474917166498,
+      "learning_rate": 4.4093050132673166e-06,
+      "loss": 0.7487,
+      "step": 1815
+    },
+    {
+      "epoch": 0.9015762690827852,
+      "grad_norm": 0.13399238294763033,
+      "learning_rate": 4.408673943946426e-06,
+      "loss": 0.7499,
+      "step": 1816
+    },
+    {
+      "epoch": 0.9020727317860245,
+      "grad_norm": 0.13587971531472334,
+      "learning_rate": 4.408042582915327e-06,
+      "loss": 0.7214,
+      "step": 1817
+    },
+    {
+      "epoch": 0.902569194489264,
+      "grad_norm": 0.14273283326406191,
+      "learning_rate": 4.407410930270512e-06,
+      "loss": 0.726,
+      "step": 1818
+    },
+    {
+      "epoch": 0.9030656571925034,
+      "grad_norm": 0.1377272028925061,
+      "learning_rate": 4.406778986108519e-06,
+      "loss": 0.7314,
+      "step": 1819
+    },
+    {
+      "epoch": 0.9035621198957429,
+      "grad_norm": 0.14415656316000652,
+      "learning_rate": 4.406146750525931e-06,
+      "loss": 0.8112,
+      "step": 1820
+    },
+    {
+      "epoch": 0.9040585825989822,
+      "grad_norm": 0.13299537098888492,
+      "learning_rate": 4.405514223619375e-06,
+      "loss": 0.7636,
+      "step": 1821
+    },
+    {
+      "epoch": 0.9045550453022216,
+      "grad_norm": 0.14030718986113996,
+      "learning_rate": 4.404881405485522e-06,
+      "loss": 0.7662,
+      "step": 1822
+    },
+    {
+      "epoch": 0.9050515080054611,
+      "grad_norm": 0.1285002264716999,
+      "learning_rate": 4.404248296221089e-06,
+      "loss": 0.7578,
+      "step": 1823
+    },
+    {
+      "epoch": 0.9055479707087005,
+      "grad_norm": 0.1357249456374853,
+      "learning_rate": 4.4036148959228365e-06,
+      "loss": 0.785,
+      "step": 1824
+    },
+    {
+      "epoch": 0.90604443341194,
+      "grad_norm": 0.13335345428150927,
+      "learning_rate": 4.402981204687569e-06,
+      "loss": 0.7519,
+      "step": 1825
+    },
+    {
+      "epoch": 0.9065408961151793,
+      "grad_norm": 0.13323854464179258,
+      "learning_rate": 4.402347222612137e-06,
+      "loss": 0.7363,
+      "step": 1826
+    },
+    {
+      "epoch": 0.9070373588184187,
+      "grad_norm": 0.1480468732609157,
+      "learning_rate": 4.401712949793433e-06,
+      "loss": 0.8581,
+      "step": 1827
+    },
+    {
+      "epoch": 0.9075338215216582,
+      "grad_norm": 0.13471889336102866,
+      "learning_rate": 4.401078386328397e-06,
+      "loss": 0.7886,
+      "step": 1828
+    },
+    {
+      "epoch": 0.9080302842248976,
+      "grad_norm": 0.1330688142356741,
+      "learning_rate": 4.40044353231401e-06,
+      "loss": 0.7205,
+      "step": 1829
+    },
+    {
+      "epoch": 0.9085267469281371,
+      "grad_norm": 0.14615671078844134,
+      "learning_rate": 4.3998083878473004e-06,
+      "loss": 0.7106,
+      "step": 1830
+    },
+    {
+      "epoch": 0.9090232096313764,
+      "grad_norm": 0.13549720868989765,
+      "learning_rate": 4.399172953025341e-06,
+      "loss": 0.7009,
+      "step": 1831
+    },
+    {
+      "epoch": 0.9095196723346158,
+      "grad_norm": 0.13185385773550817,
+      "learning_rate": 4.398537227945246e-06,
+      "loss": 0.7404,
+      "step": 1832
+    },
+    {
+      "epoch": 0.9100161350378553,
+      "grad_norm": 0.13266084243229623,
+      "learning_rate": 4.397901212704176e-06,
+      "loss": 0.762,
+      "step": 1833
+    },
+    {
+      "epoch": 0.9105125977410947,
+      "grad_norm": 0.1359129457194815,
+      "learning_rate": 4.397264907399337e-06,
+      "loss": 0.7739,
+      "step": 1834
+    },
+    {
+      "epoch": 0.9110090604443342,
+      "grad_norm": 0.13102612310251532,
+      "learning_rate": 4.396628312127977e-06,
+      "loss": 0.7081,
+      "step": 1835
+    },
+    {
+      "epoch": 0.9115055231475735,
+      "grad_norm": 0.1352565063423689,
+      "learning_rate": 4.39599142698739e-06,
+      "loss": 0.7689,
+      "step": 1836
+    },
+    {
+      "epoch": 0.9120019858508129,
+      "grad_norm": 0.1358199459420324,
+      "learning_rate": 4.395354252074912e-06,
+      "loss": 0.7418,
+      "step": 1837
+    },
+    {
+      "epoch": 0.9124984485540524,
+      "grad_norm": 0.1340420997123764,
+      "learning_rate": 4.394716787487928e-06,
+      "loss": 0.7555,
+      "step": 1838
+    },
+    {
+      "epoch": 0.9129949112572918,
+      "grad_norm": 0.14029534457477436,
+      "learning_rate": 4.394079033323862e-06,
+      "loss": 0.759,
+      "step": 1839
+    },
+    {
+      "epoch": 0.9134913739605313,
+      "grad_norm": 0.14168093185036393,
+      "learning_rate": 4.393440989680184e-06,
+      "loss": 0.793,
+      "step": 1840
+    },
+    {
+      "epoch": 0.9139878366637706,
+      "grad_norm": 0.13613343802332856,
+      "learning_rate": 4.39280265665441e-06,
+      "loss": 0.7622,
+      "step": 1841
+    },
+    {
+      "epoch": 0.91448429936701,
+      "grad_norm": 0.134141801326751,
+      "learning_rate": 4.392164034344099e-06,
+      "loss": 0.7318,
+      "step": 1842
+    },
+    {
+      "epoch": 0.9149807620702495,
+      "grad_norm": 0.13523796345465022,
+      "learning_rate": 4.391525122846855e-06,
+      "loss": 0.7306,
+      "step": 1843
+    },
+    {
+      "epoch": 0.9154772247734889,
+      "grad_norm": 0.1404574112739095,
+      "learning_rate": 4.390885922260323e-06,
+      "loss": 0.7078,
+      "step": 1844
+    },
+    {
+      "epoch": 0.9159736874767284,
+      "grad_norm": 0.1336817191525083,
+      "learning_rate": 4.390246432682196e-06,
+      "loss": 0.7827,
+      "step": 1845
+    },
+    {
+      "epoch": 0.9164701501799677,
+      "grad_norm": 0.13488164789434137,
+      "learning_rate": 4.389606654210209e-06,
+      "loss": 0.7829,
+      "step": 1846
+    },
+    {
+      "epoch": 0.9169666128832071,
+      "grad_norm": 0.13753302958393473,
+      "learning_rate": 4.388966586942144e-06,
+      "loss": 0.7592,
+      "step": 1847
+    },
+    {
+      "epoch": 0.9174630755864466,
+      "grad_norm": 0.13535888688211714,
+      "learning_rate": 4.388326230975822e-06,
+      "loss": 0.7047,
+      "step": 1848
+    },
+    {
+      "epoch": 0.917959538289686,
+      "grad_norm": 0.15778012406959804,
+      "learning_rate": 4.387685586409113e-06,
+      "loss": 0.8187,
+      "step": 1849
+    },
+    {
+      "epoch": 0.9184560009929255,
+      "grad_norm": 0.1421612541209436,
+      "learning_rate": 4.387044653339929e-06,
+      "loss": 0.7156,
+      "step": 1850
+    },
+    {
+      "epoch": 0.9189524636961648,
+      "grad_norm": 0.13379900149196902,
+      "learning_rate": 4.386403431866227e-06,
+      "loss": 0.7026,
+      "step": 1851
+    },
+    {
+      "epoch": 0.9194489263994042,
+      "grad_norm": 0.13738592358777368,
+      "learning_rate": 4.385761922086006e-06,
+      "loss": 0.7395,
+      "step": 1852
+    },
+    {
+      "epoch": 0.9199453891026437,
+      "grad_norm": 0.13390600462410002,
+      "learning_rate": 4.385120124097311e-06,
+      "loss": 0.7409,
+      "step": 1853
+    },
+    {
+      "epoch": 0.9204418518058831,
+      "grad_norm": 0.16039084660182953,
+      "learning_rate": 4.38447803799823e-06,
+      "loss": 0.7908,
+      "step": 1854
+    },
+    {
+      "epoch": 0.9209383145091226,
+      "grad_norm": 0.14623098390188546,
+      "learning_rate": 4.383835663886897e-06,
+      "loss": 0.8183,
+      "step": 1855
+    },
+    {
+      "epoch": 0.9214347772123619,
+      "grad_norm": 0.13634108193684458,
+      "learning_rate": 4.383193001861488e-06,
+      "loss": 0.7497,
+      "step": 1856
+    },
+    {
+      "epoch": 0.9219312399156013,
+      "grad_norm": 0.14386362828828383,
+      "learning_rate": 4.382550052020223e-06,
+      "loss": 0.7665,
+      "step": 1857
+    },
+    {
+      "epoch": 0.9224277026188408,
+      "grad_norm": 0.14077375377005916,
+      "learning_rate": 4.381906814461366e-06,
+      "loss": 0.7573,
+      "step": 1858
+    },
+    {
+      "epoch": 0.9229241653220802,
+      "grad_norm": 0.13040249185121516,
+      "learning_rate": 4.3812632892832275e-06,
+      "loss": 0.7399,
+      "step": 1859
+    },
+    {
+      "epoch": 0.9234206280253195,
+      "grad_norm": 0.14035305401909165,
+      "learning_rate": 4.38061947658416e-06,
+      "loss": 0.7425,
+      "step": 1860
+    },
+    {
+      "epoch": 0.923917090728559,
+      "grad_norm": 0.13852311964581557,
+      "learning_rate": 4.379975376462557e-06,
+      "loss": 0.7366,
+      "step": 1861
+    },
+    {
+      "epoch": 0.9244135534317984,
+      "grad_norm": 0.14569169577045035,
+      "learning_rate": 4.379330989016861e-06,
+      "loss": 0.7598,
+      "step": 1862
+    },
+    {
+      "epoch": 0.9249100161350379,
+      "grad_norm": 0.1420209523517496,
+      "learning_rate": 4.3786863143455575e-06,
+      "loss": 0.7468,
+      "step": 1863
+    },
+    {
+      "epoch": 0.9254064788382773,
+      "grad_norm": 0.13850510279377587,
+      "learning_rate": 4.378041352547172e-06,
+      "loss": 0.7813,
+      "step": 1864
+    },
+    {
+      "epoch": 0.9259029415415166,
+      "grad_norm": 0.13571804405986027,
+      "learning_rate": 4.3773961037202784e-06,
+      "loss": 0.7678,
+      "step": 1865
+    },
+    {
+      "epoch": 0.9263994042447561,
+      "grad_norm": 0.15050220230033115,
+      "learning_rate": 4.3767505679634925e-06,
+      "loss": 0.7502,
+      "step": 1866
+    },
+    {
+      "epoch": 0.9268958669479955,
+      "grad_norm": 0.13968131538073705,
+      "learning_rate": 4.376104745375474e-06,
+      "loss": 0.7503,
+      "step": 1867
+    },
+    {
+      "epoch": 0.927392329651235,
+      "grad_norm": 0.15025837625476948,
+      "learning_rate": 4.375458636054924e-06,
+      "loss": 0.7845,
+      "step": 1868
+    },
+    {
+      "epoch": 0.9278887923544744,
+      "grad_norm": 0.14063528342813086,
+      "learning_rate": 4.374812240100594e-06,
+      "loss": 0.7252,
+      "step": 1869
+    },
+    {
+      "epoch": 0.9283852550577137,
+      "grad_norm": 0.13713965606972475,
+      "learning_rate": 4.374165557611273e-06,
+      "loss": 0.7426,
+      "step": 1870
+    },
+    {
+      "epoch": 0.9288817177609532,
+      "grad_norm": 0.13695944367515048,
+      "learning_rate": 4.373518588685797e-06,
+      "loss": 0.7409,
+      "step": 1871
+    },
+    {
+      "epoch": 0.9293781804641926,
+      "grad_norm": 0.1422748389549011,
+      "learning_rate": 4.372871333423044e-06,
+      "loss": 0.7377,
+      "step": 1872
+    },
+    {
+      "epoch": 0.9298746431674321,
+      "grad_norm": 0.15953047007232668,
+      "learning_rate": 4.372223791921937e-06,
+      "loss": 0.794,
+      "step": 1873
+    },
+    {
+      "epoch": 0.9303711058706715,
+      "grad_norm": 0.14027429808379158,
+      "learning_rate": 4.371575964281441e-06,
+      "loss": 0.7338,
+      "step": 1874
+    },
+    {
+      "epoch": 0.9308675685739108,
+      "grad_norm": 0.1402138448734584,
+      "learning_rate": 4.370927850600569e-06,
+      "loss": 0.8293,
+      "step": 1875
+    },
+    {
+      "epoch": 0.9313640312771503,
+      "grad_norm": 0.14210713241359957,
+      "learning_rate": 4.370279450978372e-06,
+      "loss": 0.724,
+      "step": 1876
+    },
+    {
+      "epoch": 0.9318604939803897,
+      "grad_norm": 0.13815508146885405,
+      "learning_rate": 4.369630765513949e-06,
+      "loss": 0.7071,
+      "step": 1877
+    },
+    {
+      "epoch": 0.9323569566836292,
+      "grad_norm": 0.13952059008632328,
+      "learning_rate": 4.368981794306441e-06,
+      "loss": 0.7539,
+      "step": 1878
+    },
+    {
+      "epoch": 0.9328534193868686,
+      "grad_norm": 0.13939971283229288,
+      "learning_rate": 4.368332537455032e-06,
+      "loss": 0.7861,
+      "step": 1879
+    },
+    {
+      "epoch": 0.9333498820901079,
+      "grad_norm": 0.1419171606522094,
+      "learning_rate": 4.367682995058952e-06,
+      "loss": 0.7177,
+      "step": 1880
+    },
+    {
+      "epoch": 0.9338463447933474,
+      "grad_norm": 0.1368329170972017,
+      "learning_rate": 4.367033167217472e-06,
+      "loss": 0.6944,
+      "step": 1881
+    },
+    {
+      "epoch": 0.9343428074965868,
+      "grad_norm": 0.13781096600616807,
+      "learning_rate": 4.366383054029907e-06,
+      "loss": 0.7615,
+      "step": 1882
+    },
+    {
+      "epoch": 0.9348392701998263,
+      "grad_norm": 0.13708727022446426,
+      "learning_rate": 4.365732655595618e-06,
+      "loss": 0.743,
+      "step": 1883
+    },
+    {
+      "epoch": 0.9353357329030657,
+      "grad_norm": 0.13672878592574392,
+      "learning_rate": 4.365081972014007e-06,
+      "loss": 0.7352,
+      "step": 1884
+    },
+    {
+      "epoch": 0.935832195606305,
+      "grad_norm": 0.1367173447197288,
+      "learning_rate": 4.364431003384522e-06,
+      "loss": 0.7197,
+      "step": 1885
+    },
+    {
+      "epoch": 0.9363286583095445,
+      "grad_norm": 0.13990056273075885,
+      "learning_rate": 4.36377974980665e-06,
+      "loss": 0.7777,
+      "step": 1886
+    },
+    {
+      "epoch": 0.9368251210127839,
+      "grad_norm": 0.1330138557768458,
+      "learning_rate": 4.363128211379929e-06,
+      "loss": 0.7605,
+      "step": 1887
+    },
+    {
+      "epoch": 0.9373215837160234,
+      "grad_norm": 0.13245661584732554,
+      "learning_rate": 4.362476388203932e-06,
+      "loss": 0.6882,
+      "step": 1888
+    },
+    {
+      "epoch": 0.9378180464192628,
+      "grad_norm": 0.13413118235673072,
+      "learning_rate": 4.361824280378283e-06,
+      "loss": 0.7746,
+      "step": 1889
+    },
+    {
+      "epoch": 0.9383145091225021,
+      "grad_norm": 0.1482744865905161,
+      "learning_rate": 4.361171888002644e-06,
+      "loss": 0.6962,
+      "step": 1890
+    },
+    {
+      "epoch": 0.9388109718257416,
+      "grad_norm": 0.13948550276160163,
+      "learning_rate": 4.360519211176724e-06,
+      "loss": 0.7621,
+      "step": 1891
+    },
+    {
+      "epoch": 0.939307434528981,
+      "grad_norm": 0.13306011715981336,
+      "learning_rate": 4.359866250000273e-06,
+      "loss": 0.7034,
+      "step": 1892
+    },
+    {
+      "epoch": 0.9398038972322205,
+      "grad_norm": 0.13596845635437071,
+      "learning_rate": 4.359213004573087e-06,
+      "loss": 0.743,
+      "step": 1893
+    },
+    {
+      "epoch": 0.9403003599354598,
+      "grad_norm": 0.1353452240695457,
+      "learning_rate": 4.358559474995003e-06,
+      "loss": 0.7207,
+      "step": 1894
+    },
+    {
+      "epoch": 0.9407968226386992,
+      "grad_norm": 0.14241184989931713,
+      "learning_rate": 4.357905661365904e-06,
+      "loss": 0.7474,
+      "step": 1895
+    },
+    {
+      "epoch": 0.9412932853419387,
+      "grad_norm": 0.13833487675781403,
+      "learning_rate": 4.357251563785712e-06,
+      "loss": 0.783,
+      "step": 1896
+    },
+    {
+      "epoch": 0.9417897480451781,
+      "grad_norm": 0.14544580018979297,
+      "learning_rate": 4.3565971823543995e-06,
+      "loss": 0.7439,
+      "step": 1897
+    },
+    {
+      "epoch": 0.9422862107484176,
+      "grad_norm": 0.13830700272336566,
+      "learning_rate": 4.355942517171975e-06,
+      "loss": 0.7761,
+      "step": 1898
+    },
+    {
+      "epoch": 0.9427826734516569,
+      "grad_norm": 0.13022704197917237,
+      "learning_rate": 4.355287568338494e-06,
+      "loss": 0.7591,
+      "step": 1899
+    },
+    {
+      "epoch": 0.9432791361548963,
+      "grad_norm": 0.14112497002640229,
+      "learning_rate": 4.354632335954056e-06,
+      "loss": 0.7541,
+      "step": 1900
+    },
+    {
+      "epoch": 0.9437755988581358,
+      "grad_norm": 0.13978038311901106,
+      "learning_rate": 4.353976820118803e-06,
+      "loss": 0.7811,
+      "step": 1901
+    },
+    {
+      "epoch": 0.9442720615613752,
+      "grad_norm": 0.14646514863621188,
+      "learning_rate": 4.353321020932918e-06,
+      "loss": 0.77,
+      "step": 1902
+    },
+    {
+      "epoch": 0.9447685242646147,
+      "grad_norm": 0.14275771666539322,
+      "learning_rate": 4.352664938496631e-06,
+      "loss": 0.7156,
+      "step": 1903
+    },
+    {
+      "epoch": 0.945264986967854,
+      "grad_norm": 0.15199585201866644,
+      "learning_rate": 4.352008572910213e-06,
+      "loss": 0.7803,
+      "step": 1904
+    },
+    {
+      "epoch": 0.9457614496710934,
+      "grad_norm": 0.15012570466464772,
+      "learning_rate": 4.35135192427398e-06,
+      "loss": 0.8116,
+      "step": 1905
+    },
+    {
+      "epoch": 0.9462579123743329,
+      "grad_norm": 0.14798566149846193,
+      "learning_rate": 4.350694992688289e-06,
+      "loss": 0.7983,
+      "step": 1906
+    },
+    {
+      "epoch": 0.9467543750775723,
+      "grad_norm": 0.13120199262792004,
+      "learning_rate": 4.350037778253543e-06,
+      "loss": 0.6915,
+      "step": 1907
+    },
+    {
+      "epoch": 0.9472508377808118,
+      "grad_norm": 0.14145729405767365,
+      "learning_rate": 4.3493802810701845e-06,
+      "loss": 0.7309,
+      "step": 1908
+    },
+    {
+      "epoch": 0.9477473004840511,
+      "grad_norm": 0.15086944052092097,
+      "learning_rate": 4.348722501238704e-06,
+      "loss": 0.744,
+      "step": 1909
+    },
+    {
+      "epoch": 0.9482437631872905,
+      "grad_norm": 0.14048210428493602,
+      "learning_rate": 4.348064438859629e-06,
+      "loss": 0.8235,
+      "step": 1910
+    },
+    {
+      "epoch": 0.94874022589053,
+      "grad_norm": 0.14254823425879515,
+      "learning_rate": 4.347406094033539e-06,
+      "loss": 0.7813,
+      "step": 1911
+    },
+    {
+      "epoch": 0.9492366885937694,
+      "grad_norm": 0.1346422118252251,
+      "learning_rate": 4.346747466861046e-06,
+      "loss": 0.7329,
+      "step": 1912
+    },
+    {
+      "epoch": 0.9497331512970089,
+      "grad_norm": 0.13446205456154095,
+      "learning_rate": 4.346088557442813e-06,
+      "loss": 0.7271,
+      "step": 1913
+    },
+    {
+      "epoch": 0.9502296140002482,
+      "grad_norm": 0.13403300687387085,
+      "learning_rate": 4.345429365879545e-06,
+      "loss": 0.7879,
+      "step": 1914
+    },
+    {
+      "epoch": 0.9507260767034876,
+      "grad_norm": 0.13602743859277094,
+      "learning_rate": 4.344769892271987e-06,
+      "loss": 0.7527,
+      "step": 1915
+    },
+    {
+      "epoch": 0.9512225394067271,
+      "grad_norm": 0.13274614412032776,
+      "learning_rate": 4.34411013672093e-06,
+      "loss": 0.7,
+      "step": 1916
+    },
+    {
+      "epoch": 0.9517190021099665,
+      "grad_norm": 0.1417039402628044,
+      "learning_rate": 4.343450099327207e-06,
+      "loss": 0.7745,
+      "step": 1917
+    },
+    {
+      "epoch": 0.952215464813206,
+      "grad_norm": 0.13859569069194763,
+      "learning_rate": 4.342789780191693e-06,
+      "loss": 0.8001,
+      "step": 1918
+    },
+    {
+      "epoch": 0.9527119275164453,
+      "grad_norm": 0.14105331622265738,
+      "learning_rate": 4.342129179415308e-06,
+      "loss": 0.7568,
+      "step": 1919
+    },
+    {
+      "epoch": 0.9532083902196847,
+      "grad_norm": 0.13279996199472438,
+      "learning_rate": 4.341468297099014e-06,
+      "loss": 0.7257,
+      "step": 1920
+    },
+    {
+      "epoch": 0.9537048529229242,
+      "grad_norm": 0.14360632974813572,
+      "learning_rate": 4.340807133343817e-06,
+      "loss": 0.7325,
+      "step": 1921
+    },
+    {
+      "epoch": 0.9542013156261636,
+      "grad_norm": 0.1388585347465505,
+      "learning_rate": 4.340145688250766e-06,
+      "loss": 0.7798,
+      "step": 1922
+    },
+    {
+      "epoch": 0.9546977783294031,
+      "grad_norm": 0.1432237090077402,
+      "learning_rate": 4.339483961920949e-06,
+      "loss": 0.7478,
+      "step": 1923
+    },
+    {
+      "epoch": 0.9551942410326424,
+      "grad_norm": 0.13728709678976309,
+      "learning_rate": 4.3388219544555035e-06,
+      "loss": 0.7208,
+      "step": 1924
+    },
+    {
+      "epoch": 0.9556907037358818,
+      "grad_norm": 0.138040404421968,
+      "learning_rate": 4.338159665955605e-06,
+      "loss": 0.7527,
+      "step": 1925
+    },
+    {
+      "epoch": 0.9561871664391213,
+      "grad_norm": 0.13539548550777614,
+      "learning_rate": 4.337497096522474e-06,
+      "loss": 0.759,
+      "step": 1926
+    },
+    {
+      "epoch": 0.9566836291423607,
+      "grad_norm": 0.14360137088724848,
+      "learning_rate": 4.336834246257374e-06,
+      "loss": 0.7671,
+      "step": 1927
+    },
+    {
+      "epoch": 0.9571800918456,
+      "grad_norm": 0.13929702076400688,
+      "learning_rate": 4.336171115261611e-06,
+      "loss": 0.7567,
+      "step": 1928
+    },
+    {
+      "epoch": 0.9576765545488395,
+      "grad_norm": 0.13535723942569114,
+      "learning_rate": 4.335507703636533e-06,
+      "loss": 0.7778,
+      "step": 1929
+    },
+    {
+      "epoch": 0.9581730172520789,
+      "grad_norm": 0.1388756684946805,
+      "learning_rate": 4.334844011483534e-06,
+      "loss": 0.7776,
+      "step": 1930
+    },
+    {
+      "epoch": 0.9586694799553184,
+      "grad_norm": 0.1483955001430376,
+      "learning_rate": 4.3341800389040465e-06,
+      "loss": 0.791,
+      "step": 1931
+    },
+    {
+      "epoch": 0.9591659426585578,
+      "grad_norm": 0.13859104875120595,
+      "learning_rate": 4.333515785999549e-06,
+      "loss": 0.7262,
+      "step": 1932
+    },
+    {
+      "epoch": 0.9596624053617971,
+      "grad_norm": 0.14028837341254094,
+      "learning_rate": 4.3328512528715624e-06,
+      "loss": 0.7514,
+      "step": 1933
+    },
+    {
+      "epoch": 0.9601588680650366,
+      "grad_norm": 0.13979863885496932,
+      "learning_rate": 4.332186439621649e-06,
+      "loss": 0.7499,
+      "step": 1934
+    },
+    {
+      "epoch": 0.960655330768276,
+      "grad_norm": 0.1364911188988977,
+      "learning_rate": 4.331521346351415e-06,
+      "loss": 0.7733,
+      "step": 1935
+    },
+    {
+      "epoch": 0.9611517934715155,
+      "grad_norm": 0.13731108508080173,
+      "learning_rate": 4.330855973162509e-06,
+      "loss": 0.7811,
+      "step": 1936
+    },
+    {
+      "epoch": 0.9616482561747549,
+      "grad_norm": 0.13586218904438216,
+      "learning_rate": 4.330190320156623e-06,
+      "loss": 0.7125,
+      "step": 1937
+    },
+    {
+      "epoch": 0.9621447188779942,
+      "grad_norm": 0.14068018909196964,
+      "learning_rate": 4.329524387435493e-06,
+      "loss": 0.7356,
+      "step": 1938
+    },
+    {
+      "epoch": 0.9626411815812337,
+      "grad_norm": 0.13818425959230285,
+      "learning_rate": 4.328858175100893e-06,
+      "loss": 0.803,
+      "step": 1939
+    },
+    {
+      "epoch": 0.9631376442844731,
+      "grad_norm": 0.13682184061425626,
+      "learning_rate": 4.328191683254646e-06,
+      "loss": 0.7409,
+      "step": 1940
+    },
+    {
+      "epoch": 0.9636341069877126,
+      "grad_norm": 0.13139663700635784,
+      "learning_rate": 4.327524911998611e-06,
+      "loss": 0.7198,
+      "step": 1941
+    },
+    {
+      "epoch": 0.964130569690952,
+      "grad_norm": 0.13247474416573432,
+      "learning_rate": 4.326857861434697e-06,
+      "loss": 0.7251,
+      "step": 1942
+    },
+    {
+      "epoch": 0.9646270323941913,
+      "grad_norm": 0.13908703323534935,
+      "learning_rate": 4.326190531664849e-06,
+      "loss": 0.7511,
+      "step": 1943
+    },
+    {
+      "epoch": 0.9651234950974308,
+      "grad_norm": 0.13675826493632884,
+      "learning_rate": 4.32552292279106e-06,
+      "loss": 0.7485,
+      "step": 1944
+    },
+    {
+      "epoch": 0.9656199578006702,
+      "grad_norm": 0.13134376978813925,
+      "learning_rate": 4.3248550349153614e-06,
+      "loss": 0.6937,
+      "step": 1945
+    },
+    {
+      "epoch": 0.9661164205039097,
+      "grad_norm": 0.13386771511134407,
+      "learning_rate": 4.324186868139831e-06,
+      "loss": 0.7163,
+      "step": 1946
+    },
+    {
+      "epoch": 0.9666128832071491,
+      "grad_norm": 0.14332833450437155,
+      "learning_rate": 4.323518422566586e-06,
+      "loss": 0.7213,
+      "step": 1947
+    },
+    {
+      "epoch": 0.9671093459103884,
+      "grad_norm": 0.1420284438393322,
+      "learning_rate": 4.322849698297787e-06,
+      "loss": 0.7476,
+      "step": 1948
+    },
+    {
+      "epoch": 0.9676058086136279,
+      "grad_norm": 0.13709550107289373,
+      "learning_rate": 4.322180695435641e-06,
+      "loss": 0.7763,
+      "step": 1949
+    },
+    {
+      "epoch": 0.9681022713168673,
+      "grad_norm": 0.14649727566446094,
+      "learning_rate": 4.32151141408239e-06,
+      "loss": 0.7886,
+      "step": 1950
+    },
+    {
+      "epoch": 0.9685987340201068,
+      "grad_norm": 0.13158072122538214,
+      "learning_rate": 4.320841854340327e-06,
+      "loss": 0.7063,
+      "step": 1951
+    },
+    {
+      "epoch": 0.9690951967233462,
+      "grad_norm": 0.13262016397878115,
+      "learning_rate": 4.32017201631178e-06,
+      "loss": 0.6991,
+      "step": 1952
+    },
+    {
+      "epoch": 0.9695916594265855,
+      "grad_norm": 0.14641301035110524,
+      "learning_rate": 4.319501900099125e-06,
+      "loss": 0.7397,
+      "step": 1953
+    },
+    {
+      "epoch": 0.970088122129825,
+      "grad_norm": 0.1372723043098086,
+      "learning_rate": 4.318831505804778e-06,
+      "loss": 0.7251,
+      "step": 1954
+    },
+    {
+      "epoch": 0.9705845848330644,
+      "grad_norm": 0.13338014846503463,
+      "learning_rate": 4.318160833531199e-06,
+      "loss": 0.6907,
+      "step": 1955
+    },
+    {
+      "epoch": 0.9710810475363039,
+      "grad_norm": 0.15322411141967948,
+      "learning_rate": 4.317489883380887e-06,
+      "loss": 0.7292,
+      "step": 1956
+    },
+    {
+      "epoch": 0.9715775102395433,
+      "grad_norm": 0.14101937816845833,
+      "learning_rate": 4.3168186554563885e-06,
+      "loss": 0.7558,
+      "step": 1957
+    },
+    {
+      "epoch": 0.9720739729427826,
+      "grad_norm": 0.14044388792373283,
+      "learning_rate": 4.316147149860289e-06,
+      "loss": 0.7488,
+      "step": 1958
+    },
+    {
+      "epoch": 0.9725704356460221,
+      "grad_norm": 0.13087099733521357,
+      "learning_rate": 4.315475366695217e-06,
+      "loss": 0.7315,
+      "step": 1959
+    },
+    {
+      "epoch": 0.9730668983492615,
+      "grad_norm": 0.13481714845223355,
+      "learning_rate": 4.314803306063845e-06,
+      "loss": 0.7712,
+      "step": 1960
+    },
+    {
+      "epoch": 0.973563361052501,
+      "grad_norm": 0.14067827771924268,
+      "learning_rate": 4.3141309680688845e-06,
+      "loss": 0.768,
+      "step": 1961
+    },
+    {
+      "epoch": 0.9740598237557403,
+      "grad_norm": 0.15976691205185806,
+      "learning_rate": 4.313458352813093e-06,
+      "loss": 0.7287,
+      "step": 1962
+    },
+    {
+      "epoch": 0.9745562864589797,
+      "grad_norm": 0.15233463929685187,
+      "learning_rate": 4.31278546039927e-06,
+      "loss": 0.7577,
+      "step": 1963
+    },
+    {
+      "epoch": 0.9750527491622192,
+      "grad_norm": 0.13881182089614136,
+      "learning_rate": 4.312112290930255e-06,
+      "loss": 0.7525,
+      "step": 1964
+    },
+    {
+      "epoch": 0.9755492118654586,
+      "grad_norm": 0.13613919026600058,
+      "learning_rate": 4.31143884450893e-06,
+      "loss": 0.7338,
+      "step": 1965
+    },
+    {
+      "epoch": 0.9760456745686981,
+      "grad_norm": 0.14009747972178313,
+      "learning_rate": 4.310765121238223e-06,
+      "loss": 0.7685,
+      "step": 1966
+    },
+    {
+      "epoch": 0.9765421372719374,
+      "grad_norm": 0.1398735823592921,
+      "learning_rate": 4.310091121221101e-06,
+      "loss": 0.7142,
+      "step": 1967
+    },
+    {
+      "epoch": 0.9770385999751768,
+      "grad_norm": 0.15370727403582576,
+      "learning_rate": 4.3094168445605735e-06,
+      "loss": 0.7518,
+      "step": 1968
+    },
+    {
+      "epoch": 0.9775350626784163,
+      "grad_norm": 0.13221826861513192,
+      "learning_rate": 4.308742291359692e-06,
+      "loss": 0.7045,
+      "step": 1969
+    },
+    {
+      "epoch": 0.9780315253816557,
+      "grad_norm": 0.13298617913810548,
+      "learning_rate": 4.308067461721553e-06,
+      "loss": 0.7197,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9785279880848952,
+      "grad_norm": 0.13310197028851603,
+      "learning_rate": 4.307392355749293e-06,
+      "loss": 0.7276,
+      "step": 1971
+    },
+    {
+      "epoch": 0.9790244507881345,
+      "grad_norm": 0.13441131799419243,
+      "learning_rate": 4.30671697354609e-06,
+      "loss": 0.7461,
+      "step": 1972
+    },
+    {
+      "epoch": 0.9795209134913739,
+      "grad_norm": 0.13598707003248428,
+      "learning_rate": 4.306041315215167e-06,
+      "loss": 0.751,
+      "step": 1973
+    },
+    {
+      "epoch": 0.9800173761946134,
+      "grad_norm": 0.13020590326743284,
+      "learning_rate": 4.305365380859786e-06,
+      "loss": 0.7471,
+      "step": 1974
+    },
+    {
+      "epoch": 0.9805138388978528,
+      "grad_norm": 0.13548633327801685,
+      "learning_rate": 4.304689170583254e-06,
+      "loss": 0.7304,
+      "step": 1975
+    },
+    {
+      "epoch": 0.9810103016010923,
+      "grad_norm": 0.14147295011534075,
+      "learning_rate": 4.304012684488917e-06,
+      "loss": 0.7926,
+      "step": 1976
+    },
+    {
+      "epoch": 0.9815067643043316,
+      "grad_norm": 0.14704693808375974,
+      "learning_rate": 4.303335922680167e-06,
+      "loss": 0.7629,
+      "step": 1977
+    },
+    {
+      "epoch": 0.982003227007571,
+      "grad_norm": 0.135220427389001,
+      "learning_rate": 4.302658885260436e-06,
+      "loss": 0.7305,
+      "step": 1978
+    },
+    {
+      "epoch": 0.9824996897108105,
+      "grad_norm": 0.13439053824847158,
+      "learning_rate": 4.301981572333197e-06,
+      "loss": 0.7752,
+      "step": 1979
+    },
+    {
+      "epoch": 0.9829961524140499,
+      "grad_norm": 0.13771634609127179,
+      "learning_rate": 4.3013039840019675e-06,
+      "loss": 0.7251,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9834926151172894,
+      "grad_norm": 0.1326406234325998,
+      "learning_rate": 4.300626120370306e-06,
+      "loss": 0.7595,
+      "step": 1981
+    },
+    {
+      "epoch": 0.9839890778205287,
+      "grad_norm": 0.14246037641515524,
+      "learning_rate": 4.2999479815418135e-06,
+      "loss": 0.7618,
+      "step": 1982
+    },
+    {
+      "epoch": 0.9844855405237681,
+      "grad_norm": 0.14003774333863,
+      "learning_rate": 4.299269567620131e-06,
+      "loss": 0.7133,
+      "step": 1983
+    },
+    {
+      "epoch": 0.9849820032270076,
+      "grad_norm": 0.13443584919833715,
+      "learning_rate": 4.2985908787089445e-06,
+      "loss": 0.7722,
+      "step": 1984
+    },
+    {
+      "epoch": 0.985478465930247,
+      "grad_norm": 0.13668575653481171,
+      "learning_rate": 4.29791191491198e-06,
+      "loss": 0.7583,
+      "step": 1985
+    },
+    {
+      "epoch": 0.9859749286334865,
+      "grad_norm": 0.14279910896798906,
+      "learning_rate": 4.297232676333007e-06,
+      "loss": 0.7981,
+      "step": 1986
+    },
+    {
+      "epoch": 0.9864713913367258,
+      "grad_norm": 0.14161222593397266,
+      "learning_rate": 4.296553163075836e-06,
+      "loss": 0.7594,
+      "step": 1987
+    },
+    {
+      "epoch": 0.9869678540399652,
+      "grad_norm": 0.13831563618388676,
+      "learning_rate": 4.295873375244319e-06,
+      "loss": 0.6993,
+      "step": 1988
+    },
+    {
+      "epoch": 0.9874643167432047,
+      "grad_norm": 0.13787046565146302,
+      "learning_rate": 4.2951933129423515e-06,
+      "loss": 0.7358,
+      "step": 1989
+    },
+    {
+      "epoch": 0.9879607794464441,
+      "grad_norm": 0.13779555768214863,
+      "learning_rate": 4.29451297627387e-06,
+      "loss": 0.7178,
+      "step": 1990
+    },
+    {
+      "epoch": 0.9884572421496836,
+      "grad_norm": 0.13217653339807073,
+      "learning_rate": 4.293832365342853e-06,
+      "loss": 0.7439,
+      "step": 1991
+    },
+    {
+      "epoch": 0.9889537048529229,
+      "grad_norm": 0.15651907445083008,
+      "learning_rate": 4.293151480253321e-06,
+      "loss": 0.7852,
+      "step": 1992
+    },
+    {
+      "epoch": 0.9894501675561623,
+      "grad_norm": 0.14035753112824711,
+      "learning_rate": 4.292470321109336e-06,
+      "loss": 0.8147,
+      "step": 1993
+    },
+    {
+      "epoch": 0.9899466302594018,
+      "grad_norm": 0.1344265290214628,
+      "learning_rate": 4.291788888015002e-06,
+      "loss": 0.7576,
+      "step": 1994
+    },
+    {
+      "epoch": 0.9904430929626412,
+      "grad_norm": 0.13379707071517158,
+      "learning_rate": 4.291107181074466e-06,
+      "loss": 0.7163,
+      "step": 1995
+    },
+    {
+      "epoch": 0.9909395556658807,
+      "grad_norm": 0.1399492515733686,
+      "learning_rate": 4.290425200391917e-06,
+      "loss": 0.7922,
+      "step": 1996
+    },
+    {
+      "epoch": 0.99143601836912,
+      "grad_norm": 0.14344554049611216,
+      "learning_rate": 4.289742946071581e-06,
+      "loss": 0.7665,
+      "step": 1997
+    },
+    {
+      "epoch": 0.9919324810723594,
+      "grad_norm": 0.1382864134304432,
+      "learning_rate": 4.2890604182177336e-06,
+      "loss": 0.7557,
+      "step": 1998
+    },
+    {
+      "epoch": 0.9924289437755989,
+      "grad_norm": 0.14309914226914516,
+      "learning_rate": 4.288377616934686e-06,
+      "loss": 0.7213,
+      "step": 1999
+    },
+    {
+      "epoch": 0.9929254064788383,
+      "grad_norm": 0.13717535026025166,
+      "learning_rate": 4.287694542326795e-06,
+      "loss": 0.7113,
+      "step": 2000
+    },
+    {
+      "epoch": 0.9934218691820776,
+      "grad_norm": 0.13746827797324848,
+      "learning_rate": 4.287011194498456e-06,
+      "loss": 0.7891,
+      "step": 2001
+    },
+    {
+      "epoch": 0.9939183318853171,
+      "grad_norm": 0.1421486270249576,
+      "learning_rate": 4.2863275735541085e-06,
+      "loss": 0.737,
+      "step": 2002
+    },
+    {
+      "epoch": 0.9944147945885565,
+      "grad_norm": 0.13682547516347832,
+      "learning_rate": 4.285643679598233e-06,
+      "loss": 0.7373,
+      "step": 2003
+    },
+    {
+      "epoch": 0.994911257291796,
+      "grad_norm": 0.1266137711385144,
+      "learning_rate": 4.284959512735352e-06,
+      "loss": 0.6983,
+      "step": 2004
+    },
+    {
+      "epoch": 0.9954077199950354,
+      "grad_norm": 0.14377205813881055,
+      "learning_rate": 4.284275073070028e-06,
+      "loss": 0.7426,
+      "step": 2005
+    },
+    {
+      "epoch": 0.9959041826982747,
+      "grad_norm": 0.1350794657949948,
+      "learning_rate": 4.283590360706868e-06,
+      "loss": 0.7668,
+      "step": 2006
+    },
+    {
+      "epoch": 0.9964006454015142,
+      "grad_norm": 0.14250458148212633,
+      "learning_rate": 4.28290537575052e-06,
+      "loss": 0.8081,
+      "step": 2007
+    },
+    {
+      "epoch": 0.9968971081047536,
+      "grad_norm": 0.13575371037098183,
+      "learning_rate": 4.282220118305672e-06,
+      "loss": 0.7055,
+      "step": 2008
+    },
+    {
+      "epoch": 0.9973935708079931,
+      "grad_norm": 0.13942020771733923,
+      "learning_rate": 4.281534588477054e-06,
+      "loss": 0.7798,
+      "step": 2009
+    },
+    {
+      "epoch": 0.9978900335112325,
+      "grad_norm": 0.13808731679650627,
+      "learning_rate": 4.280848786369439e-06,
+      "loss": 0.7241,
+      "step": 2010
+    },
+    {
+      "epoch": 0.9983864962144718,
+      "grad_norm": 0.14200950616113964,
+      "learning_rate": 4.280162712087641e-06,
+      "loss": 0.7211,
+      "step": 2011
+    },
+    {
+      "epoch": 0.9988829589177113,
+      "grad_norm": 0.13227471958984183,
+      "learning_rate": 4.2794763657365155e-06,
+      "loss": 0.7434,
+      "step": 2012
+    },
+    {
+      "epoch": 0.9993794216209507,
+      "grad_norm": 0.1506748510159815,
+      "learning_rate": 4.278789747420959e-06,
+      "loss": 0.772,
+      "step": 2013
+    },
+    {
+      "epoch": 0.9998758843241902,
+      "grad_norm": 0.1692117125114903,
+      "learning_rate": 4.27810285724591e-06,
+      "loss": 0.8515,
+      "step": 2014
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1692117125114903,
+      "learning_rate": 4.27741569531635e-06,
+      "loss": 0.1869,
+      "step": 2015
+    },
+    {
+      "epoch": 1.0003723470274295,
+      "grad_norm": 0.13641143847300066,
+      "learning_rate": 4.276728261737298e-06,
+      "loss": 0.5693,
+      "step": 2016
+    },
+    {
+      "epoch": 1.0003723470274295,
+      "eval_loss": 0.7493109107017517,
+      "eval_runtime": 135.5139,
+      "eval_samples_per_second": 223.984,
+      "eval_steps_per_second": 28.005,
+      "step": 2016
+    },
+    {
+      "epoch": 1.0004964627032393,
+      "grad_norm": 0.14993340674591726,
+      "learning_rate": 4.27604055661382e-06,
+      "loss": 0.741,
+      "step": 2017
+    },
+    {
+      "epoch": 1.0009929254064789,
+      "grad_norm": 0.15136956647535096,
+      "learning_rate": 4.275352580051019e-06,
+      "loss": 0.7305,
+      "step": 2018
+    },
+    {
+      "epoch": 1.0014893881097182,
+      "grad_norm": 0.13556367994060173,
+      "learning_rate": 4.274664332154042e-06,
+      "loss": 0.6991,
+      "step": 2019
+    },
+    {
+      "epoch": 1.0019858508129578,
+      "grad_norm": 0.13474503555709075,
+      "learning_rate": 4.273975813028076e-06,
+      "loss": 0.7061,
+      "step": 2020
+    },
+    {
+      "epoch": 1.002482313516197,
+      "grad_norm": 0.13926546524904157,
+      "learning_rate": 4.273287022778351e-06,
+      "loss": 0.7501,
+      "step": 2021
+    },
+    {
+      "epoch": 1.0029787762194364,
+      "grad_norm": 0.1417176043167176,
+      "learning_rate": 4.272597961510137e-06,
+      "loss": 0.7204,
+      "step": 2022
+    },
+    {
+      "epoch": 1.003475238922676,
+      "grad_norm": 0.1419815927815163,
+      "learning_rate": 4.271908629328747e-06,
+      "loss": 0.7269,
+      "step": 2023
+    },
+    {
+      "epoch": 1.0039717016259153,
+      "grad_norm": 0.14837462280399044,
+      "learning_rate": 4.2712190263395315e-06,
+      "loss": 0.7593,
+      "step": 2024
+    },
+    {
+      "epoch": 1.0044681643291549,
+      "grad_norm": 0.1424958004520062,
+      "learning_rate": 4.270529152647889e-06,
+      "loss": 0.7987,
+      "step": 2025
+    },
+    {
+      "epoch": 1.0049646270323942,
+      "grad_norm": 0.13557327739425565,
+      "learning_rate": 4.269839008359252e-06,
+      "loss": 0.7095,
+      "step": 2026
+    },
+    {
+      "epoch": 1.0054610897356335,
+      "grad_norm": 0.1350892189012179,
+      "learning_rate": 4.269148593579101e-06,
+      "loss": 0.7297,
+      "step": 2027
+    },
+    {
+      "epoch": 1.005957552438873,
+      "grad_norm": 0.1431149142071869,
+      "learning_rate": 4.268457908412953e-06,
+      "loss": 0.7074,
+      "step": 2028
+    },
+    {
+      "epoch": 1.0064540151421124,
+      "grad_norm": 0.13418266947538876,
+      "learning_rate": 4.267766952966369e-06,
+      "loss": 0.7235,
+      "step": 2029
+    },
+    {
+      "epoch": 1.006950477845352,
+      "grad_norm": 0.131551107709248,
+      "learning_rate": 4.267075727344951e-06,
+      "loss": 0.7286,
+      "step": 2030
+    },
+    {
+      "epoch": 1.0074469405485913,
+      "grad_norm": 0.13666318482495718,
+      "learning_rate": 4.266384231654339e-06,
+      "loss": 0.7225,
+      "step": 2031
+    },
+    {
+      "epoch": 1.0079434032518306,
+      "grad_norm": 0.14957531887872869,
+      "learning_rate": 4.265692466000221e-06,
+      "loss": 0.7452,
+      "step": 2032
+    },
+    {
+      "epoch": 1.0084398659550702,
+      "grad_norm": 0.1352458457312148,
+      "learning_rate": 4.2650004304883195e-06,
+      "loss": 0.7488,
+      "step": 2033
+    },
+    {
+      "epoch": 1.0089363286583095,
+      "grad_norm": 0.13431771924046446,
+      "learning_rate": 4.2643081252244024e-06,
+      "loss": 0.7408,
+      "step": 2034
+    },
+    {
+      "epoch": 1.009432791361549,
+      "grad_norm": 0.1404702073381516,
+      "learning_rate": 4.263615550314276e-06,
+      "loss": 0.7136,
+      "step": 2035
+    },
+    {
+      "epoch": 1.0099292540647884,
+      "grad_norm": 0.14333737075703962,
+      "learning_rate": 4.262922705863791e-06,
+      "loss": 0.7315,
+      "step": 2036
+    },
+    {
+      "epoch": 1.0104257167680277,
+      "grad_norm": 0.13931858496722177,
+      "learning_rate": 4.262229591978836e-06,
+      "loss": 0.7515,
+      "step": 2037
+    },
+    {
+      "epoch": 1.0109221794712673,
+      "grad_norm": 0.14111176709272,
+      "learning_rate": 4.261536208765343e-06,
+      "loss": 0.685,
+      "step": 2038
+    },
+    {
+      "epoch": 1.0114186421745066,
+      "grad_norm": 0.13873104450186644,
+      "learning_rate": 4.260842556329285e-06,
+      "loss": 0.7419,
+      "step": 2039
+    },
+    {
+      "epoch": 1.0119151048777462,
+      "grad_norm": 0.14733619687470917,
+      "learning_rate": 4.260148634776675e-06,
+      "loss": 0.745,
+      "step": 2040
+    },
+    {
+      "epoch": 1.0124115675809855,
+      "grad_norm": 0.13985004223191186,
+      "learning_rate": 4.259454444213568e-06,
+      "loss": 0.7246,
+      "step": 2041
+    },
+    {
+      "epoch": 1.0129080302842248,
+      "grad_norm": 0.13995194992924537,
+      "learning_rate": 4.25875998474606e-06,
+      "loss": 0.7642,
+      "step": 2042
+    },
+    {
+      "epoch": 1.0134044929874644,
+      "grad_norm": 0.1362055125185364,
+      "learning_rate": 4.258065256480288e-06,
+      "loss": 0.7341,
+      "step": 2043
+    },
+    {
+      "epoch": 1.0139009556907037,
+      "grad_norm": 0.1422465441519774,
+      "learning_rate": 4.25737025952243e-06,
+      "loss": 0.7266,
+      "step": 2044
+    },
+    {
+      "epoch": 1.0143974183939433,
+      "grad_norm": 0.13512192671393297,
+      "learning_rate": 4.2566749939787056e-06,
+      "loss": 0.7098,
+      "step": 2045
+    },
+    {
+      "epoch": 1.0148938810971826,
+      "grad_norm": 0.13453521708159794,
+      "learning_rate": 4.255979459955374e-06,
+      "loss": 0.6598,
+      "step": 2046
+    },
+    {
+      "epoch": 1.015390343800422,
+      "grad_norm": 0.1458759220859774,
+      "learning_rate": 4.255283657558736e-06,
+      "loss": 0.7255,
+      "step": 2047
+    },
+    {
+      "epoch": 1.0158868065036615,
+      "grad_norm": 0.13826399594709918,
+      "learning_rate": 4.2545875868951355e-06,
+      "loss": 0.6747,
+      "step": 2048
+    },
+    {
+      "epoch": 1.0163832692069008,
+      "grad_norm": 0.15055189765595503,
+      "learning_rate": 4.253891248070956e-06,
+      "loss": 0.7719,
+      "step": 2049
+    },
+    {
+      "epoch": 1.0168797319101404,
+      "grad_norm": 0.13416248781345586,
+      "learning_rate": 4.253194641192621e-06,
+      "loss": 0.6815,
+      "step": 2050
+    },
+    {
+      "epoch": 1.0173761946133797,
+      "grad_norm": 0.13559489599334676,
+      "learning_rate": 4.252497766366593e-06,
+      "loss": 0.7037,
+      "step": 2051
+    },
+    {
+      "epoch": 1.017872657316619,
+      "grad_norm": 0.172228917141983,
+      "learning_rate": 4.251800623699382e-06,
+      "loss": 0.7434,
+      "step": 2052
+    },
+    {
+      "epoch": 1.0183691200198586,
+      "grad_norm": 0.1370557981462678,
+      "learning_rate": 4.251103213297534e-06,
+      "loss": 0.724,
+      "step": 2053
+    },
+    {
+      "epoch": 1.018865582723098,
+      "grad_norm": 0.13629385144601702,
+      "learning_rate": 4.250405535267636e-06,
+      "loss": 0.753,
+      "step": 2054
+    },
+    {
+      "epoch": 1.0193620454263372,
+      "grad_norm": 0.13309228573779064,
+      "learning_rate": 4.249707589716318e-06,
+      "loss": 0.7319,
+      "step": 2055
+    },
+    {
+      "epoch": 1.0198585081295768,
+      "grad_norm": 0.13715514685240124,
+      "learning_rate": 4.24900937675025e-06,
+      "loss": 0.7897,
+      "step": 2056
+    },
+    {
+      "epoch": 1.0203549708328161,
+      "grad_norm": 0.1357669090684218,
+      "learning_rate": 4.24831089647614e-06,
+      "loss": 0.7267,
+      "step": 2057
+    },
+    {
+      "epoch": 1.0208514335360557,
+      "grad_norm": 0.1498727514671048,
+      "learning_rate": 4.247612149000743e-06,
+      "loss": 0.7297,
+      "step": 2058
+    },
+    {
+      "epoch": 1.021347896239295,
+      "grad_norm": 0.14049047303313542,
+      "learning_rate": 4.24691313443085e-06,
+      "loss": 0.7304,
+      "step": 2059
+    },
+    {
+      "epoch": 1.0218443589425343,
+      "grad_norm": 0.13909453659414045,
+      "learning_rate": 4.2462138528732935e-06,
+      "loss": 0.7048,
+      "step": 2060
+    },
+    {
+      "epoch": 1.022340821645774,
+      "grad_norm": 0.13412019436547604,
+      "learning_rate": 4.245514304434948e-06,
+      "loss": 0.6914,
+      "step": 2061
+    },
+    {
+      "epoch": 1.0228372843490132,
+      "grad_norm": 0.1391340532915626,
+      "learning_rate": 4.244814489222728e-06,
+      "loss": 0.7017,
+      "step": 2062
+    },
+    {
+      "epoch": 1.0233337470522528,
+      "grad_norm": 0.14295513021033426,
+      "learning_rate": 4.244114407343589e-06,
+      "loss": 0.7457,
+      "step": 2063
+    },
+    {
+      "epoch": 1.023830209755492,
+      "grad_norm": 0.13900941541232267,
+      "learning_rate": 4.2434140589045286e-06,
+      "loss": 0.7401,
+      "step": 2064
+    },
+    {
+      "epoch": 1.0243266724587314,
+      "grad_norm": 0.13944892230939118,
+      "learning_rate": 4.242713444012583e-06,
+      "loss": 0.7874,
+      "step": 2065
+    },
+    {
+      "epoch": 1.024823135161971,
+      "grad_norm": 0.13593208488424774,
+      "learning_rate": 4.242012562774829e-06,
+      "loss": 0.7239,
+      "step": 2066
+    },
+    {
+      "epoch": 1.0253195978652103,
+      "grad_norm": 0.13931000983919048,
+      "learning_rate": 4.241311415298386e-06,
+      "loss": 0.7454,
+      "step": 2067
+    },
+    {
+      "epoch": 1.0258160605684499,
+      "grad_norm": 0.1330003858692133,
+      "learning_rate": 4.240610001690413e-06,
+      "loss": 0.734,
+      "step": 2068
+    },
+    {
+      "epoch": 1.0263125232716892,
+      "grad_norm": 0.14108224227175165,
+      "learning_rate": 4.239908322058109e-06,
+      "loss": 0.6962,
+      "step": 2069
+    },
+    {
+      "epoch": 1.0268089859749285,
+      "grad_norm": 0.1391861450047936,
+      "learning_rate": 4.239206376508716e-06,
+      "loss": 0.7008,
+      "step": 2070
+    },
+    {
+      "epoch": 1.027305448678168,
+      "grad_norm": 0.1370473901313178,
+      "learning_rate": 4.238504165149515e-06,
+      "loss": 0.7544,
+      "step": 2071
+    },
+    {
+      "epoch": 1.0278019113814074,
+      "grad_norm": 0.13536037419786456,
+      "learning_rate": 4.237801688087827e-06,
+      "loss": 0.7369,
+      "step": 2072
+    },
+    {
+      "epoch": 1.028298374084647,
+      "grad_norm": 0.14978518639680113,
+      "learning_rate": 4.237098945431014e-06,
+      "loss": 0.7679,
+      "step": 2073
+    },
+    {
+      "epoch": 1.0287948367878863,
+      "grad_norm": 0.13912767164252948,
+      "learning_rate": 4.236395937286479e-06,
+      "loss": 0.7166,
+      "step": 2074
+    },
+    {
+      "epoch": 1.0292912994911256,
+      "grad_norm": 0.1386818465495855,
+      "learning_rate": 4.2356926637616665e-06,
+      "loss": 0.7385,
+      "step": 2075
+    },
+    {
+      "epoch": 1.0297877621943652,
+      "grad_norm": 0.1411775085091878,
+      "learning_rate": 4.234989124964061e-06,
+      "loss": 0.7256,
+      "step": 2076
+    },
+    {
+      "epoch": 1.0302842248976045,
+      "grad_norm": 0.14419739467783685,
+      "learning_rate": 4.234285321001185e-06,
+      "loss": 0.7371,
+      "step": 2077
+    },
+    {
+      "epoch": 1.030780687600844,
+      "grad_norm": 0.13803664666766124,
+      "learning_rate": 4.233581251980604e-06,
+      "loss": 0.7746,
+      "step": 2078
+    },
+    {
+      "epoch": 1.0312771503040834,
+      "grad_norm": 0.14388287736016128,
+      "learning_rate": 4.2328769180099265e-06,
+      "loss": 0.7547,
+      "step": 2079
+    },
+    {
+      "epoch": 1.0317736130073227,
+      "grad_norm": 0.14306598998666248,
+      "learning_rate": 4.232172319196795e-06,
+      "loss": 0.7117,
+      "step": 2080
+    },
+    {
+      "epoch": 1.0322700757105623,
+      "grad_norm": 0.14932389713396255,
+      "learning_rate": 4.231467455648899e-06,
+      "loss": 0.7507,
+      "step": 2081
+    },
+    {
+      "epoch": 1.0327665384138016,
+      "grad_norm": 0.14803637284334528,
+      "learning_rate": 4.230762327473964e-06,
+      "loss": 0.7168,
+      "step": 2082
+    },
+    {
+      "epoch": 1.0332630011170412,
+      "grad_norm": 0.14111484516048733,
+      "learning_rate": 4.2300569347797584e-06,
+      "loss": 0.73,
+      "step": 2083
+    },
+    {
+      "epoch": 1.0337594638202805,
+      "grad_norm": 0.13730617408120754,
+      "learning_rate": 4.229351277674088e-06,
+      "loss": 0.7295,
+      "step": 2084
+    },
+    {
+      "epoch": 1.0342559265235198,
+      "grad_norm": 0.14834889202516877,
+      "learning_rate": 4.228645356264805e-06,
+      "loss": 0.8029,
+      "step": 2085
+    },
+    {
+      "epoch": 1.0347523892267594,
+      "grad_norm": 0.14599972266177927,
+      "learning_rate": 4.227939170659795e-06,
+      "loss": 0.7629,
+      "step": 2086
+    },
+    {
+      "epoch": 1.0352488519299987,
+      "grad_norm": 0.14092123214072694,
+      "learning_rate": 4.227232720966988e-06,
+      "loss": 0.7107,
+      "step": 2087
+    },
+    {
+      "epoch": 1.0357453146332383,
+      "grad_norm": 0.14122940427509462,
+      "learning_rate": 4.226526007294353e-06,
+      "loss": 0.7213,
+      "step": 2088
+    },
+    {
+      "epoch": 1.0362417773364776,
+      "grad_norm": 0.13842815838317993,
+      "learning_rate": 4.225819029749902e-06,
+      "loss": 0.7574,
+      "step": 2089
+    },
+    {
+      "epoch": 1.036738240039717,
+      "grad_norm": 0.13969707819774496,
+      "learning_rate": 4.225111788441682e-06,
+      "loss": 0.7901,
+      "step": 2090
+    },
+    {
+      "epoch": 1.0372347027429565,
+      "grad_norm": 0.14454482563554347,
+      "learning_rate": 4.224404283477788e-06,
+      "loss": 0.7426,
+      "step": 2091
+    },
+    {
+      "epoch": 1.0377311654461958,
+      "grad_norm": 0.1589968114041004,
+      "learning_rate": 4.223696514966346e-06,
+      "loss": 0.757,
+      "step": 2092
+    },
+    {
+      "epoch": 1.0382276281494354,
+      "grad_norm": 0.14630132012052294,
+      "learning_rate": 4.2229884830155285e-06,
+      "loss": 0.7052,
+      "step": 2093
+    },
+    {
+      "epoch": 1.0387240908526747,
+      "grad_norm": 0.13974708317715842,
+      "learning_rate": 4.222280187733549e-06,
+      "loss": 0.7687,
+      "step": 2094
+    },
+    {
+      "epoch": 1.039220553555914,
+      "grad_norm": 0.1383464362903483,
+      "learning_rate": 4.2215716292286555e-06,
+      "loss": 0.7048,
+      "step": 2095
+    },
+    {
+      "epoch": 1.0397170162591536,
+      "grad_norm": 0.14009397553084701,
+      "learning_rate": 4.220862807609144e-06,
+      "loss": 0.6834,
+      "step": 2096
+    },
+    {
+      "epoch": 1.040213478962393,
+      "grad_norm": 0.14079336307045506,
+      "learning_rate": 4.220153722983342e-06,
+      "loss": 0.7392,
+      "step": 2097
+    },
+    {
+      "epoch": 1.0407099416656325,
+      "grad_norm": 0.13787553354880905,
+      "learning_rate": 4.219444375459626e-06,
+      "loss": 0.7496,
+      "step": 2098
+    },
+    {
+      "epoch": 1.0412064043688718,
+      "grad_norm": 0.1400054541103246,
+      "learning_rate": 4.218734765146405e-06,
+      "loss": 0.7155,
+      "step": 2099
+    },
+    {
+      "epoch": 1.0417028670721111,
+      "grad_norm": 0.13722004560528184,
+      "learning_rate": 4.218024892152134e-06,
+      "loss": 0.736,
+      "step": 2100
+    },
+    {
+      "epoch": 1.0421993297753507,
+      "grad_norm": 0.13237548806271895,
+      "learning_rate": 4.217314756585305e-06,
+      "loss": 0.6985,
+      "step": 2101
+    },
+    {
+      "epoch": 1.04269579247859,
+      "grad_norm": 0.13814683970889546,
+      "learning_rate": 4.2166043585544495e-06,
+      "loss": 0.742,
+      "step": 2102
+    },
+    {
+      "epoch": 1.0431922551818296,
+      "grad_norm": 0.1519709587964635,
+      "learning_rate": 4.2158936981681415e-06,
+      "loss": 0.7286,
+      "step": 2103
+    },
+    {
+      "epoch": 1.043688717885069,
+      "grad_norm": 0.13343117788795786,
+      "learning_rate": 4.215182775534994e-06,
+      "loss": 0.7769,
+      "step": 2104
+    },
+    {
+      "epoch": 1.0441851805883082,
+      "grad_norm": 0.14132688675642163,
+      "learning_rate": 4.21447159076366e-06,
+      "loss": 0.7645,
+      "step": 2105
+    },
+    {
+      "epoch": 1.0446816432915478,
+      "grad_norm": 0.142210221906363,
+      "learning_rate": 4.213760143962834e-06,
+      "loss": 0.7481,
+      "step": 2106
+    },
+    {
+      "epoch": 1.045178105994787,
+      "grad_norm": 0.1374760586531246,
+      "learning_rate": 4.2130484352412475e-06,
+      "loss": 0.6909,
+      "step": 2107
+    },
+    {
+      "epoch": 1.0456745686980267,
+      "grad_norm": 0.1309015889916558,
+      "learning_rate": 4.212336464707674e-06,
+      "loss": 0.7001,
+      "step": 2108
+    },
+    {
+      "epoch": 1.046171031401266,
+      "grad_norm": 0.13676033591829806,
+      "learning_rate": 4.211624232470927e-06,
+      "loss": 0.7778,
+      "step": 2109
+    },
+    {
+      "epoch": 1.0466674941045053,
+      "grad_norm": 0.13760414687569777,
+      "learning_rate": 4.2109117386398595e-06,
+      "loss": 0.7199,
+      "step": 2110
+    },
+    {
+      "epoch": 1.0471639568077449,
+      "grad_norm": 0.14075911332828087,
+      "learning_rate": 4.210198983323366e-06,
+      "loss": 0.7741,
+      "step": 2111
+    },
+    {
+      "epoch": 1.0476604195109842,
+      "grad_norm": 0.13356672723592797,
+      "learning_rate": 4.209485966630377e-06,
+      "loss": 0.7253,
+      "step": 2112
+    },
+    {
+      "epoch": 1.0481568822142238,
+      "grad_norm": 0.14150649799028853,
+      "learning_rate": 4.2087726886698695e-06,
+      "loss": 0.7779,
+      "step": 2113
+    },
+    {
+      "epoch": 1.048653344917463,
+      "grad_norm": 0.13823807245765327,
+      "learning_rate": 4.208059149550855e-06,
+      "loss": 0.741,
+      "step": 2114
+    },
+    {
+      "epoch": 1.0491498076207024,
+      "grad_norm": 0.14366242195045892,
+      "learning_rate": 4.207345349382385e-06,
+      "loss": 0.7521,
+      "step": 2115
+    },
+    {
+      "epoch": 1.049646270323942,
+      "grad_norm": 0.13980640222600257,
+      "learning_rate": 4.206631288273554e-06,
+      "loss": 0.7132,
+      "step": 2116
+    },
+    {
+      "epoch": 1.0501427330271813,
+      "grad_norm": 0.14845761168836935,
+      "learning_rate": 4.205916966333494e-06,
+      "loss": 0.7072,
+      "step": 2117
+    },
+    {
+      "epoch": 1.0506391957304206,
+      "grad_norm": 0.14139479483348766,
+      "learning_rate": 4.2052023836713785e-06,
+      "loss": 0.7411,
+      "step": 2118
+    },
+    {
+      "epoch": 1.0511356584336602,
+      "grad_norm": 0.14059968450149515,
+      "learning_rate": 4.204487540396419e-06,
+      "loss": 0.691,
+      "step": 2119
+    },
+    {
+      "epoch": 1.0516321211368995,
+      "grad_norm": 0.1362156286798585,
+      "learning_rate": 4.203772436617868e-06,
+      "loss": 0.7444,
+      "step": 2120
+    },
+    {
+      "epoch": 1.052128583840139,
+      "grad_norm": 0.13831235259558317,
+      "learning_rate": 4.203057072445019e-06,
+      "loss": 0.7317,
+      "step": 2121
+    },
+    {
+      "epoch": 1.0526250465433784,
+      "grad_norm": 0.14207834501053565,
+      "learning_rate": 4.202341447987202e-06,
+      "loss": 0.7359,
+      "step": 2122
+    },
+    {
+      "epoch": 1.0531215092466177,
+      "grad_norm": 0.13778286488933011,
+      "learning_rate": 4.201625563353791e-06,
+      "loss": 0.7756,
+      "step": 2123
+    },
+    {
+      "epoch": 1.0536179719498573,
+      "grad_norm": 0.13980048084090257,
+      "learning_rate": 4.200909418654194e-06,
+      "loss": 0.7331,
+      "step": 2124
+    },
+    {
+      "epoch": 1.0541144346530966,
+      "grad_norm": 0.13997426427628695,
+      "learning_rate": 4.200193013997866e-06,
+      "loss": 0.6935,
+      "step": 2125
+    },
+    {
+      "epoch": 1.0546108973563362,
+      "grad_norm": 0.1459212095265043,
+      "learning_rate": 4.199476349494296e-06,
+      "loss": 0.7204,
+      "step": 2126
+    },
+    {
+      "epoch": 1.0551073600595755,
+      "grad_norm": 0.14329178565231715,
+      "learning_rate": 4.198759425253015e-06,
+      "loss": 0.7464,
+      "step": 2127
+    },
+    {
+      "epoch": 1.0556038227628148,
+      "grad_norm": 0.1422075716131861,
+      "learning_rate": 4.198042241383593e-06,
+      "loss": 0.7486,
+      "step": 2128
+    },
+    {
+      "epoch": 1.0561002854660544,
+      "grad_norm": 0.14092996648947848,
+      "learning_rate": 4.19732479799564e-06,
+      "loss": 0.7074,
+      "step": 2129
+    },
+    {
+      "epoch": 1.0565967481692937,
+      "grad_norm": 0.14086274951921257,
+      "learning_rate": 4.196607095198808e-06,
+      "loss": 0.7577,
+      "step": 2130
+    },
+    {
+      "epoch": 1.0570932108725333,
+      "grad_norm": 0.13776776723760562,
+      "learning_rate": 4.1958891331027826e-06,
+      "loss": 0.7324,
+      "step": 2131
+    },
+    {
+      "epoch": 1.0575896735757726,
+      "grad_norm": 0.13566064448922902,
+      "learning_rate": 4.1951709118172954e-06,
+      "loss": 0.7626,
+      "step": 2132
+    },
+    {
+      "epoch": 1.058086136279012,
+      "grad_norm": 0.13763795010324914,
+      "learning_rate": 4.194452431452115e-06,
+      "loss": 0.7537,
+      "step": 2133
+    },
+    {
+      "epoch": 1.0585825989822515,
+      "grad_norm": 0.1459065764808873,
+      "learning_rate": 4.193733692117048e-06,
+      "loss": 0.7355,
+      "step": 2134
+    },
+    {
+      "epoch": 1.0590790616854908,
+      "grad_norm": 0.13884201088984038,
+      "learning_rate": 4.193014693921944e-06,
+      "loss": 0.7535,
+      "step": 2135
+    },
+    {
+      "epoch": 1.0595755243887304,
+      "grad_norm": 0.14001522032769478,
+      "learning_rate": 4.192295436976688e-06,
+      "loss": 0.7447,
+      "step": 2136
+    },
+    {
+      "epoch": 1.0600719870919697,
+      "grad_norm": 0.15172505034245246,
+      "learning_rate": 4.19157592139121e-06,
+      "loss": 0.7522,
+      "step": 2137
+    },
+    {
+      "epoch": 1.060568449795209,
+      "grad_norm": 0.13794840697062816,
+      "learning_rate": 4.190856147275474e-06,
+      "loss": 0.6883,
+      "step": 2138
+    },
+    {
+      "epoch": 1.0610649124984486,
+      "grad_norm": 0.14217691316180742,
+      "learning_rate": 4.190136114739487e-06,
+      "loss": 0.727,
+      "step": 2139
+    },
+    {
+      "epoch": 1.061561375201688,
+      "grad_norm": 0.13583861770554806,
+      "learning_rate": 4.189415823893293e-06,
+      "loss": 0.704,
+      "step": 2140
+    },
+    {
+      "epoch": 1.0620578379049275,
+      "grad_norm": 0.1363645606259115,
+      "learning_rate": 4.188695274846979e-06,
+      "loss": 0.7381,
+      "step": 2141
+    },
+    {
+      "epoch": 1.0625543006081668,
+      "grad_norm": 0.14814722988133758,
+      "learning_rate": 4.1879744677106685e-06,
+      "loss": 0.7346,
+      "step": 2142
+    },
+    {
+      "epoch": 1.0630507633114061,
+      "grad_norm": 0.16059324736898764,
+      "learning_rate": 4.1872534025945255e-06,
+      "loss": 0.7899,
+      "step": 2143
+    },
+    {
+      "epoch": 1.0635472260146457,
+      "grad_norm": 0.14014168016673342,
+      "learning_rate": 4.186532079608753e-06,
+      "loss": 0.7599,
+      "step": 2144
+    },
+    {
+      "epoch": 1.064043688717885,
+      "grad_norm": 0.14674448113708913,
+      "learning_rate": 4.185810498863592e-06,
+      "loss": 0.7054,
+      "step": 2145
+    },
+    {
+      "epoch": 1.0645401514211246,
+      "grad_norm": 0.13552371087730036,
+      "learning_rate": 4.185088660469328e-06,
+      "loss": 0.7602,
+      "step": 2146
+    },
+    {
+      "epoch": 1.065036614124364,
+      "grad_norm": 0.1336555213075988,
+      "learning_rate": 4.18436656453628e-06,
+      "loss": 0.7583,
+      "step": 2147
+    },
+    {
+      "epoch": 1.0655330768276032,
+      "grad_norm": 0.13559730592832767,
+      "learning_rate": 4.1836442111748086e-06,
+      "loss": 0.7456,
+      "step": 2148
+    },
+    {
+      "epoch": 1.0660295395308428,
+      "grad_norm": 0.14992961750694342,
+      "learning_rate": 4.182921600495316e-06,
+      "loss": 0.7379,
+      "step": 2149
+    },
+    {
+      "epoch": 1.0665260022340821,
+      "grad_norm": 0.13935556972523674,
+      "learning_rate": 4.1821987326082396e-06,
+      "loss": 0.696,
+      "step": 2150
+    },
+    {
+      "epoch": 1.0670224649373217,
+      "grad_norm": 0.14083946307481723,
+      "learning_rate": 4.181475607624059e-06,
+      "loss": 0.7851,
+      "step": 2151
+    },
+    {
+      "epoch": 1.067518927640561,
+      "grad_norm": 0.1421578052366813,
+      "learning_rate": 4.1807522256532925e-06,
+      "loss": 0.7259,
+      "step": 2152
+    },
+    {
+      "epoch": 1.0680153903438003,
+      "grad_norm": 0.13548843386534212,
+      "learning_rate": 4.1800285868064964e-06,
+      "loss": 0.7304,
+      "step": 2153
+    },
+    {
+      "epoch": 1.0685118530470399,
+      "grad_norm": 0.13702431747439558,
+      "learning_rate": 4.179304691194269e-06,
+      "loss": 0.758,
+      "step": 2154
+    },
+    {
+      "epoch": 1.0690083157502792,
+      "grad_norm": 0.13318149795074,
+      "learning_rate": 4.178580538927245e-06,
+      "loss": 0.6844,
+      "step": 2155
+    },
+    {
+      "epoch": 1.0695047784535188,
+      "grad_norm": 0.1407081020046659,
+      "learning_rate": 4.177856130116099e-06,
+      "loss": 0.7639,
+      "step": 2156
+    },
+    {
+      "epoch": 1.070001241156758,
+      "grad_norm": 0.1438833118655734,
+      "learning_rate": 4.177131464871545e-06,
+      "loss": 0.7006,
+      "step": 2157
+    },
+    {
+      "epoch": 1.0704977038599974,
+      "grad_norm": 0.1461487574061599,
+      "learning_rate": 4.176406543304339e-06,
+      "loss": 0.738,
+      "step": 2158
+    },
+    {
+      "epoch": 1.070994166563237,
+      "grad_norm": 0.1398230912718051,
+      "learning_rate": 4.175681365525271e-06,
+      "loss": 0.6985,
+      "step": 2159
+    },
+    {
+      "epoch": 1.0714906292664763,
+      "grad_norm": 0.13817741460521907,
+      "learning_rate": 4.174955931645175e-06,
+      "loss": 0.7285,
+      "step": 2160
+    },
+    {
+      "epoch": 1.0719870919697159,
+      "grad_norm": 0.14224093463734003,
+      "learning_rate": 4.1742302417749205e-06,
+      "loss": 0.7586,
+      "step": 2161
+    },
+    {
+      "epoch": 1.0724835546729552,
+      "grad_norm": 0.13835662156431366,
+      "learning_rate": 4.173504296025417e-06,
+      "loss": 0.7539,
+      "step": 2162
+    },
+    {
+      "epoch": 1.0729800173761945,
+      "grad_norm": 0.13861535686149404,
+      "learning_rate": 4.1727780945076155e-06,
+      "loss": 0.654,
+      "step": 2163
+    },
+    {
+      "epoch": 1.073476480079434,
+      "grad_norm": 0.13428993721206128,
+      "learning_rate": 4.172051637332501e-06,
+      "loss": 0.7285,
+      "step": 2164
+    },
+    {
+      "epoch": 1.0739729427826734,
+      "grad_norm": 0.13538857395884993,
+      "learning_rate": 4.171324924611105e-06,
+      "loss": 0.7218,
+      "step": 2165
+    },
+    {
+      "epoch": 1.074469405485913,
+      "grad_norm": 0.1452550536947396,
+      "learning_rate": 4.170597956454492e-06,
+      "loss": 0.8168,
+      "step": 2166
+    },
+    {
+      "epoch": 1.0749658681891523,
+      "grad_norm": 0.13624518253158813,
+      "learning_rate": 4.169870732973767e-06,
+      "loss": 0.7141,
+      "step": 2167
+    },
+    {
+      "epoch": 1.0754623308923916,
+      "grad_norm": 0.14029339027722842,
+      "learning_rate": 4.169143254280074e-06,
+      "loss": 0.7285,
+      "step": 2168
+    },
+    {
+      "epoch": 1.0759587935956312,
+      "grad_norm": 0.14109579501178382,
+      "learning_rate": 4.168415520484598e-06,
+      "loss": 0.7896,
+      "step": 2169
+    },
+    {
+      "epoch": 1.0764552562988705,
+      "grad_norm": 0.144954703610336,
+      "learning_rate": 4.167687531698561e-06,
+      "loss": 0.7255,
+      "step": 2170
+    },
+    {
+      "epoch": 1.07695171900211,
+      "grad_norm": 0.13685269108041861,
+      "learning_rate": 4.166959288033223e-06,
+      "loss": 0.7257,
+      "step": 2171
+    },
+    {
+      "epoch": 1.0774481817053494,
+      "grad_norm": 0.14271875588892513,
+      "learning_rate": 4.166230789599886e-06,
+      "loss": 0.7159,
+      "step": 2172
+    },
+    {
+      "epoch": 1.0779446444085887,
+      "grad_norm": 0.13783479927505968,
+      "learning_rate": 4.16550203650989e-06,
+      "loss": 0.7129,
+      "step": 2173
+    },
+    {
+      "epoch": 1.0784411071118283,
+      "grad_norm": 0.14283091206105697,
+      "learning_rate": 4.16477302887461e-06,
+      "loss": 0.6965,
+      "step": 2174
+    },
+    {
+      "epoch": 1.0789375698150676,
+      "grad_norm": 0.14010606156996824,
+      "learning_rate": 4.1640437668054665e-06,
+      "loss": 0.7179,
+      "step": 2175
+    },
+    {
+      "epoch": 1.0794340325183072,
+      "grad_norm": 0.14730576733153716,
+      "learning_rate": 4.163314250413913e-06,
+      "loss": 0.7082,
+      "step": 2176
+    },
+    {
+      "epoch": 1.0799304952215465,
+      "grad_norm": 0.1335227779819202,
+      "learning_rate": 4.1625844798114464e-06,
+      "loss": 0.7183,
+      "step": 2177
+    },
+    {
+      "epoch": 1.0804269579247858,
+      "grad_norm": 0.13601772614825755,
+      "learning_rate": 4.1618544551096005e-06,
+      "loss": 0.6847,
+      "step": 2178
+    },
+    {
+      "epoch": 1.0809234206280254,
+      "grad_norm": 0.13698194445852674,
+      "learning_rate": 4.1611241764199465e-06,
+      "loss": 0.7435,
+      "step": 2179
+    },
+    {
+      "epoch": 1.0814198833312647,
+      "grad_norm": 0.14062893743840316,
+      "learning_rate": 4.160393643854095e-06,
+      "loss": 0.7363,
+      "step": 2180
+    },
+    {
+      "epoch": 1.0819163460345043,
+      "grad_norm": 0.13799273854449803,
+      "learning_rate": 4.159662857523697e-06,
+      "loss": 0.7243,
+      "step": 2181
+    },
+    {
+      "epoch": 1.0824128087377436,
+      "grad_norm": 0.13716660868793465,
+      "learning_rate": 4.158931817540443e-06,
+      "loss": 0.7577,
+      "step": 2182
+    },
+    {
+      "epoch": 1.082909271440983,
+      "grad_norm": 0.1466433823136944,
+      "learning_rate": 4.158200524016061e-06,
+      "loss": 0.7054,
+      "step": 2183
+    },
+    {
+      "epoch": 1.0834057341442225,
+      "grad_norm": 0.1382394309059659,
+      "learning_rate": 4.1574689770623145e-06,
+      "loss": 0.7566,
+      "step": 2184
+    },
+    {
+      "epoch": 1.0839021968474618,
+      "grad_norm": 0.13654874611387244,
+      "learning_rate": 4.15673717679101e-06,
+      "loss": 0.7251,
+      "step": 2185
+    },
+    {
+      "epoch": 1.0843986595507014,
+      "grad_norm": 0.1367869405798336,
+      "learning_rate": 4.156005123313993e-06,
+      "loss": 0.719,
+      "step": 2186
+    },
+    {
+      "epoch": 1.0848951222539407,
+      "grad_norm": 0.1317568061970169,
+      "learning_rate": 4.155272816743145e-06,
+      "loss": 0.6876,
+      "step": 2187
+    },
+    {
+      "epoch": 1.08539158495718,
+      "grad_norm": 0.13762978347128463,
+      "learning_rate": 4.1545402571903855e-06,
+      "loss": 0.7022,
+      "step": 2188
+    },
+    {
+      "epoch": 1.0858880476604196,
+      "grad_norm": 0.1362143217786678,
+      "learning_rate": 4.153807444767677e-06,
+      "loss": 0.7125,
+      "step": 2189
+    },
+    {
+      "epoch": 1.086384510363659,
+      "grad_norm": 0.1516046995027248,
+      "learning_rate": 4.153074379587018e-06,
+      "loss": 0.809,
+      "step": 2190
+    },
+    {
+      "epoch": 1.0868809730668985,
+      "grad_norm": 0.13562138385068911,
+      "learning_rate": 4.152341061760445e-06,
+      "loss": 0.7172,
+      "step": 2191
+    },
+    {
+      "epoch": 1.0873774357701378,
+      "grad_norm": 0.14008342460158815,
+      "learning_rate": 4.151607491400034e-06,
+      "loss": 0.7781,
+      "step": 2192
+    },
+    {
+      "epoch": 1.0878738984733771,
+      "grad_norm": 0.13999792205054443,
+      "learning_rate": 4.150873668617899e-06,
+      "loss": 0.6984,
+      "step": 2193
+    },
+    {
+      "epoch": 1.0883703611766167,
+      "grad_norm": 0.12918569307833325,
+      "learning_rate": 4.150139593526193e-06,
+      "loss": 0.707,
+      "step": 2194
+    },
+    {
+      "epoch": 1.088866823879856,
+      "grad_norm": 0.1373176298653365,
+      "learning_rate": 4.149405266237109e-06,
+      "loss": 0.7073,
+      "step": 2195
+    },
+    {
+      "epoch": 1.0893632865830956,
+      "grad_norm": 0.12770538449262095,
+      "learning_rate": 4.148670686862877e-06,
+      "loss": 0.6754,
+      "step": 2196
+    },
+    {
+      "epoch": 1.089859749286335,
+      "grad_norm": 0.14027683877629007,
+      "learning_rate": 4.147935855515763e-06,
+      "loss": 0.7618,
+      "step": 2197
+    },
+    {
+      "epoch": 1.0903562119895742,
+      "grad_norm": 0.13474412755523313,
+      "learning_rate": 4.1472007723080774e-06,
+      "loss": 0.7486,
+      "step": 2198
+    },
+    {
+      "epoch": 1.0908526746928138,
+      "grad_norm": 0.13436431752038755,
+      "learning_rate": 4.146465437352164e-06,
+      "loss": 0.7237,
+      "step": 2199
+    },
+    {
+      "epoch": 1.091349137396053,
+      "grad_norm": 0.13682091282171036,
+      "learning_rate": 4.145729850760408e-06,
+      "loss": 0.7377,
+      "step": 2200
+    },
+    {
+      "epoch": 1.0918456000992927,
+      "grad_norm": 0.13924706240826093,
+      "learning_rate": 4.144994012645232e-06,
+      "loss": 0.7402,
+      "step": 2201
+    },
+    {
+      "epoch": 1.092342062802532,
+      "grad_norm": 0.14103770773255148,
+      "learning_rate": 4.1442579231190964e-06,
+      "loss": 0.69,
+      "step": 2202
+    },
+    {
+      "epoch": 1.0928385255057713,
+      "grad_norm": 0.1394654587729206,
+      "learning_rate": 4.143521582294501e-06,
+      "loss": 0.7393,
+      "step": 2203
+    },
+    {
+      "epoch": 1.0933349882090109,
+      "grad_norm": 0.13644960570918493,
+      "learning_rate": 4.1427849902839826e-06,
+      "loss": 0.7141,
+      "step": 2204
+    },
+    {
+      "epoch": 1.0938314509122502,
+      "grad_norm": 0.13692258907444782,
+      "learning_rate": 4.142048147200119e-06,
+      "loss": 0.7393,
+      "step": 2205
+    },
+    {
+      "epoch": 1.0943279136154895,
+      "grad_norm": 0.1373142702334565,
+      "learning_rate": 4.141311053155524e-06,
+      "loss": 0.755,
+      "step": 2206
+    },
+    {
+      "epoch": 1.094824376318729,
+      "grad_norm": 0.13763207554959367,
+      "learning_rate": 4.140573708262852e-06,
+      "loss": 0.7546,
+      "step": 2207
+    },
+    {
+      "epoch": 1.0953208390219684,
+      "grad_norm": 0.14478103422831126,
+      "learning_rate": 4.139836112634792e-06,
+      "loss": 0.7679,
+      "step": 2208
+    },
+    {
+      "epoch": 1.095817301725208,
+      "grad_norm": 0.1390259770870048,
+      "learning_rate": 4.139098266384076e-06,
+      "loss": 0.743,
+      "step": 2209
+    },
+    {
+      "epoch": 1.0963137644284473,
+      "grad_norm": 0.14390923525858076,
+      "learning_rate": 4.138360169623471e-06,
+      "loss": 0.7453,
+      "step": 2210
+    },
+    {
+      "epoch": 1.0968102271316866,
+      "grad_norm": 0.14517278952992427,
+      "learning_rate": 4.137621822465782e-06,
+      "loss": 0.7426,
+      "step": 2211
+    },
+    {
+      "epoch": 1.0973066898349262,
+      "grad_norm": 0.15525845893147422,
+      "learning_rate": 4.1368832250238564e-06,
+      "loss": 0.7313,
+      "step": 2212
+    },
+    {
+      "epoch": 1.0978031525381655,
+      "grad_norm": 0.14066801820933797,
+      "learning_rate": 4.136144377410574e-06,
+      "loss": 0.7877,
+      "step": 2213
+    },
+    {
+      "epoch": 1.098299615241405,
+      "grad_norm": 0.1424752647162409,
+      "learning_rate": 4.135405279738858e-06,
+      "loss": 0.7418,
+      "step": 2214
+    },
+    {
+      "epoch": 1.0987960779446444,
+      "grad_norm": 0.15190627011179594,
+      "learning_rate": 4.134665932121665e-06,
+      "loss": 0.7654,
+      "step": 2215
+    },
+    {
+      "epoch": 1.0992925406478837,
+      "grad_norm": 0.1412407950996624,
+      "learning_rate": 4.133926334671996e-06,
+      "loss": 0.6707,
+      "step": 2216
+    },
+    {
+      "epoch": 1.0997890033511233,
+      "grad_norm": 0.14456861330667964,
+      "learning_rate": 4.133186487502884e-06,
+      "loss": 0.7841,
+      "step": 2217
+    },
+    {
+      "epoch": 1.1002854660543626,
+      "grad_norm": 0.15407876229352044,
+      "learning_rate": 4.1324463907274035e-06,
+      "loss": 0.7367,
+      "step": 2218
+    },
+    {
+      "epoch": 1.1007819287576022,
+      "grad_norm": 0.14216692584554558,
+      "learning_rate": 4.131706044458667e-06,
+      "loss": 0.7149,
+      "step": 2219
+    },
+    {
+      "epoch": 1.1012783914608415,
+      "grad_norm": 0.1324859332684942,
+      "learning_rate": 4.130965448809824e-06,
+      "loss": 0.6563,
+      "step": 2220
+    },
+    {
+      "epoch": 1.1017748541640808,
+      "grad_norm": 0.13983336523306109,
+      "learning_rate": 4.130224603894062e-06,
+      "loss": 0.7078,
+      "step": 2221
+    },
+    {
+      "epoch": 1.1022713168673204,
+      "grad_norm": 0.14405406563289364,
+      "learning_rate": 4.129483509824608e-06,
+      "loss": 0.7529,
+      "step": 2222
+    },
+    {
+      "epoch": 1.1027677795705597,
+      "grad_norm": 0.14158523055382277,
+      "learning_rate": 4.128742166714726e-06,
+      "loss": 0.7113,
+      "step": 2223
+    },
+    {
+      "epoch": 1.1032642422737993,
+      "grad_norm": 0.13671925424224868,
+      "learning_rate": 4.128000574677719e-06,
+      "loss": 0.746,
+      "step": 2224
+    },
+    {
+      "epoch": 1.1037607049770386,
+      "grad_norm": 0.13505116983586582,
+      "learning_rate": 4.127258733826929e-06,
+      "loss": 0.7181,
+      "step": 2225
+    },
+    {
+      "epoch": 1.104257167680278,
+      "grad_norm": 0.1374278474736183,
+      "learning_rate": 4.126516644275731e-06,
+      "loss": 0.7529,
+      "step": 2226
+    },
+    {
+      "epoch": 1.1047536303835175,
+      "grad_norm": 0.14165368952037294,
+      "learning_rate": 4.125774306137543e-06,
+      "loss": 0.7476,
+      "step": 2227
+    },
+    {
+      "epoch": 1.1052500930867568,
+      "grad_norm": 0.15143713359026406,
+      "learning_rate": 4.12503171952582e-06,
+      "loss": 0.7326,
+      "step": 2228
+    },
+    {
+      "epoch": 1.1057465557899964,
+      "grad_norm": 0.137643913086052,
+      "learning_rate": 4.124288884554053e-06,
+      "loss": 0.7351,
+      "step": 2229
+    },
+    {
+      "epoch": 1.1062430184932357,
+      "grad_norm": 0.13550077923391268,
+      "learning_rate": 4.123545801335776e-06,
+      "loss": 0.7181,
+      "step": 2230
+    },
+    {
+      "epoch": 1.106739481196475,
+      "grad_norm": 0.13907847760418357,
+      "learning_rate": 4.122802469984552e-06,
+      "loss": 0.7113,
+      "step": 2231
+    },
+    {
+      "epoch": 1.1072359438997146,
+      "grad_norm": 0.1409638418684744,
+      "learning_rate": 4.122058890613991e-06,
+      "loss": 0.7532,
+      "step": 2232
+    },
+    {
+      "epoch": 1.107732406602954,
+      "grad_norm": 0.13839521717196338,
+      "learning_rate": 4.121315063337737e-06,
+      "loss": 0.7593,
+      "step": 2233
+    },
+    {
+      "epoch": 1.1082288693061935,
+      "grad_norm": 0.14464848496201313,
+      "learning_rate": 4.120570988269472e-06,
+      "loss": 0.798,
+      "step": 2234
+    },
+    {
+      "epoch": 1.1087253320094328,
+      "grad_norm": 0.14170615709306067,
+      "learning_rate": 4.119826665522914e-06,
+      "loss": 0.7214,
+      "step": 2235
+    },
+    {
+      "epoch": 1.1092217947126721,
+      "grad_norm": 0.13750992725040465,
+      "learning_rate": 4.119082095211823e-06,
+      "loss": 0.7025,
+      "step": 2236
+    },
+    {
+      "epoch": 1.1097182574159117,
+      "grad_norm": 0.13524887733070357,
+      "learning_rate": 4.118337277449993e-06,
+      "loss": 0.7299,
+      "step": 2237
+    },
+    {
+      "epoch": 1.110214720119151,
+      "grad_norm": 0.13742212811395596,
+      "learning_rate": 4.117592212351258e-06,
+      "loss": 0.7152,
+      "step": 2238
+    },
+    {
+      "epoch": 1.1107111828223906,
+      "grad_norm": 0.13549859577744783,
+      "learning_rate": 4.11684690002949e-06,
+      "loss": 0.7939,
+      "step": 2239
+    },
+    {
+      "epoch": 1.11120764552563,
+      "grad_norm": 0.13072436304815396,
+      "learning_rate": 4.116101340598597e-06,
+      "loss": 0.6907,
+      "step": 2240
+    },
+    {
+      "epoch": 1.1117041082288692,
+      "grad_norm": 0.13708530876777447,
+      "learning_rate": 4.115355534172527e-06,
+      "loss": 0.7668,
+      "step": 2241
+    },
+    {
+      "epoch": 1.1122005709321088,
+      "grad_norm": 0.1331260842138928,
+      "learning_rate": 4.114609480865264e-06,
+      "loss": 0.7315,
+      "step": 2242
+    },
+    {
+      "epoch": 1.112697033635348,
+      "grad_norm": 0.1345957058229882,
+      "learning_rate": 4.113863180790829e-06,
+      "loss": 0.7249,
+      "step": 2243
+    },
+    {
+      "epoch": 1.1131934963385877,
+      "grad_norm": 0.13896142397489988,
+      "learning_rate": 4.113116634063285e-06,
+      "loss": 0.7174,
+      "step": 2244
+    },
+    {
+      "epoch": 1.113689959041827,
+      "grad_norm": 0.13616762035864557,
+      "learning_rate": 4.1123698407967265e-06,
+      "loss": 0.7206,
+      "step": 2245
+    },
+    {
+      "epoch": 1.1141864217450663,
+      "grad_norm": 0.1428487363329108,
+      "learning_rate": 4.11162280110529e-06,
+      "loss": 0.7447,
+      "step": 2246
+    },
+    {
+      "epoch": 1.1146828844483059,
+      "grad_norm": 0.1381247261812151,
+      "learning_rate": 4.110875515103148e-06,
+      "loss": 0.7262,
+      "step": 2247
+    },
+    {
+      "epoch": 1.1151793471515452,
+      "grad_norm": 0.13985941699766852,
+      "learning_rate": 4.110127982904513e-06,
+      "loss": 0.737,
+      "step": 2248
+    },
+    {
+      "epoch": 1.1156758098547848,
+      "grad_norm": 0.1371407690195463,
+      "learning_rate": 4.109380204623634e-06,
+      "loss": 0.7367,
+      "step": 2249
+    },
+    {
+      "epoch": 1.116172272558024,
+      "grad_norm": 0.1410923264807246,
+      "learning_rate": 4.108632180374794e-06,
+      "loss": 0.685,
+      "step": 2250
+    },
+    {
+      "epoch": 1.1166687352612634,
+      "grad_norm": 0.13607561024834763,
+      "learning_rate": 4.107883910272316e-06,
+      "loss": 0.7416,
+      "step": 2251
+    },
+    {
+      "epoch": 1.117165197964503,
+      "grad_norm": 0.14401553966823152,
+      "learning_rate": 4.107135394430565e-06,
+      "loss": 0.7741,
+      "step": 2252
+    },
+    {
+      "epoch": 1.1176616606677423,
+      "grad_norm": 0.1347979188420261,
+      "learning_rate": 4.106386632963935e-06,
+      "loss": 0.7289,
+      "step": 2253
+    },
+    {
+      "epoch": 1.1181581233709816,
+      "grad_norm": 0.13563125984475435,
+      "learning_rate": 4.105637625986867e-06,
+      "loss": 0.7241,
+      "step": 2254
+    },
+    {
+      "epoch": 1.1186545860742212,
+      "grad_norm": 0.13554783899608014,
+      "learning_rate": 4.104888373613832e-06,
+      "loss": 0.7051,
+      "step": 2255
+    },
+    {
+      "epoch": 1.1191510487774605,
+      "grad_norm": 0.13902713235820974,
+      "learning_rate": 4.10413887595934e-06,
+      "loss": 0.7442,
+      "step": 2256
+    },
+    {
+      "epoch": 1.1196475114807,
+      "grad_norm": 0.1416398097741949,
+      "learning_rate": 4.1033891331379425e-06,
+      "loss": 0.7234,
+      "step": 2257
+    },
+    {
+      "epoch": 1.1201439741839394,
+      "grad_norm": 0.13151231872027344,
+      "learning_rate": 4.102639145264223e-06,
+      "loss": 0.6901,
+      "step": 2258
+    },
+    {
+      "epoch": 1.1206404368871787,
+      "grad_norm": 0.1310560747375517,
+      "learning_rate": 4.101888912452809e-06,
+      "loss": 0.7034,
+      "step": 2259
+    },
+    {
+      "epoch": 1.1211368995904183,
+      "grad_norm": 0.13368674747757894,
+      "learning_rate": 4.1011384348183565e-06,
+      "loss": 0.7369,
+      "step": 2260
+    },
+    {
+      "epoch": 1.1216333622936576,
+      "grad_norm": 0.14030562550210587,
+      "learning_rate": 4.100387712475568e-06,
+      "loss": 0.7473,
+      "step": 2261
+    },
+    {
+      "epoch": 1.1221298249968972,
+      "grad_norm": 0.13783403570651975,
+      "learning_rate": 4.0996367455391774e-06,
+      "loss": 0.6827,
+      "step": 2262
+    },
+    {
+      "epoch": 1.1226262877001365,
+      "grad_norm": 0.13794119866048046,
+      "learning_rate": 4.098885534123958e-06,
+      "loss": 0.7144,
+      "step": 2263
+    },
+    {
+      "epoch": 1.1231227504033758,
+      "grad_norm": 0.1403803932613888,
+      "learning_rate": 4.098134078344722e-06,
+      "loss": 0.7752,
+      "step": 2264
+    },
+    {
+      "epoch": 1.1236192131066154,
+      "grad_norm": 0.1393102973681631,
+      "learning_rate": 4.097382378316316e-06,
+      "loss": 0.7241,
+      "step": 2265
+    },
+    {
+      "epoch": 1.1241156758098547,
+      "grad_norm": 0.13730972644736142,
+      "learning_rate": 4.0966304341536255e-06,
+      "loss": 0.7399,
+      "step": 2266
+    },
+    {
+      "epoch": 1.1246121385130943,
+      "grad_norm": 0.14795637490908276,
+      "learning_rate": 4.095878245971573e-06,
+      "loss": 0.7947,
+      "step": 2267
+    },
+    {
+      "epoch": 1.1251086012163336,
+      "grad_norm": 0.13800836804454775,
+      "learning_rate": 4.095125813885118e-06,
+      "loss": 0.7454,
+      "step": 2268
+    },
+    {
+      "epoch": 1.125605063919573,
+      "grad_norm": 0.13510837243563636,
+      "learning_rate": 4.094373138009259e-06,
+      "loss": 0.7126,
+      "step": 2269
+    },
+    {
+      "epoch": 1.1261015266228125,
+      "grad_norm": 0.14183611387295958,
+      "learning_rate": 4.093620218459029e-06,
+      "loss": 0.7305,
+      "step": 2270
+    },
+    {
+      "epoch": 1.1265979893260518,
+      "grad_norm": 0.14547694629965133,
+      "learning_rate": 4.092867055349501e-06,
+      "loss": 0.7342,
+      "step": 2271
+    },
+    {
+      "epoch": 1.1270944520292914,
+      "grad_norm": 0.13272000346114335,
+      "learning_rate": 4.092113648795784e-06,
+      "loss": 0.7244,
+      "step": 2272
+    },
+    {
+      "epoch": 1.1275909147325307,
+      "grad_norm": 0.13445748304111405,
+      "learning_rate": 4.091359998913021e-06,
+      "loss": 0.7135,
+      "step": 2273
+    },
+    {
+      "epoch": 1.12808737743577,
+      "grad_norm": 0.13823114531273367,
+      "learning_rate": 4.0906061058164e-06,
+      "loss": 0.7296,
+      "step": 2274
+    },
+    {
+      "epoch": 1.1285838401390096,
+      "grad_norm": 0.1331655131810508,
+      "learning_rate": 4.089851969621138e-06,
+      "loss": 0.7039,
+      "step": 2275
+    },
+    {
+      "epoch": 1.129080302842249,
+      "grad_norm": 0.13502020601948467,
+      "learning_rate": 4.0890975904424946e-06,
+      "loss": 0.7227,
+      "step": 2276
+    },
+    {
+      "epoch": 1.1295767655454885,
+      "grad_norm": 0.13554787632722626,
+      "learning_rate": 4.088342968395763e-06,
+      "loss": 0.7166,
+      "step": 2277
+    },
+    {
+      "epoch": 1.1300732282487278,
+      "grad_norm": 0.15351257577564006,
+      "learning_rate": 4.087588103596276e-06,
+      "loss": 0.7515,
+      "step": 2278
+    },
+    {
+      "epoch": 1.1305696909519671,
+      "grad_norm": 0.13285248996749063,
+      "learning_rate": 4.0868329961594025e-06,
+      "loss": 0.7398,
+      "step": 2279
+    },
+    {
+      "epoch": 1.1310661536552067,
+      "grad_norm": 0.14474300166445284,
+      "learning_rate": 4.086077646200548e-06,
+      "loss": 0.765,
+      "step": 2280
+    },
+    {
+      "epoch": 1.131562616358446,
+      "grad_norm": 0.136914338868158,
+      "learning_rate": 4.085322053835157e-06,
+      "loss": 0.7387,
+      "step": 2281
+    },
+    {
+      "epoch": 1.1320590790616856,
+      "grad_norm": 0.13701247994557483,
+      "learning_rate": 4.084566219178708e-06,
+      "loss": 0.7484,
+      "step": 2282
+    },
+    {
+      "epoch": 1.132555541764925,
+      "grad_norm": 0.14213309578391173,
+      "learning_rate": 4.08381014234672e-06,
+      "loss": 0.734,
+      "step": 2283
+    },
+    {
+      "epoch": 1.1330520044681642,
+      "grad_norm": 0.1413403485403749,
+      "learning_rate": 4.0830538234547455e-06,
+      "loss": 0.739,
+      "step": 2284
+    },
+    {
+      "epoch": 1.1335484671714038,
+      "grad_norm": 0.14178436767190955,
+      "learning_rate": 4.082297262618376e-06,
+      "loss": 0.7338,
+      "step": 2285
+    },
+    {
+      "epoch": 1.1340449298746431,
+      "grad_norm": 0.1384665628085734,
+      "learning_rate": 4.081540459953241e-06,
+      "loss": 0.7293,
+      "step": 2286
+    },
+    {
+      "epoch": 1.1345413925778827,
+      "grad_norm": 0.14117953096422015,
+      "learning_rate": 4.080783415575004e-06,
+      "loss": 0.7241,
+      "step": 2287
+    },
+    {
+      "epoch": 1.135037855281122,
+      "grad_norm": 0.13841249272824596,
+      "learning_rate": 4.080026129599368e-06,
+      "loss": 0.7129,
+      "step": 2288
+    },
+    {
+      "epoch": 1.1355343179843613,
+      "grad_norm": 0.1389416393419498,
+      "learning_rate": 4.079268602142072e-06,
+      "loss": 0.7196,
+      "step": 2289
+    },
+    {
+      "epoch": 1.1360307806876009,
+      "grad_norm": 0.13525972832589087,
+      "learning_rate": 4.078510833318892e-06,
+      "loss": 0.6799,
+      "step": 2290
+    },
+    {
+      "epoch": 1.1365272433908402,
+      "grad_norm": 0.14116131185065883,
+      "learning_rate": 4.077752823245641e-06,
+      "loss": 0.7207,
+      "step": 2291
+    },
+    {
+      "epoch": 1.1370237060940798,
+      "grad_norm": 0.13868074387182236,
+      "learning_rate": 4.076994572038168e-06,
+      "loss": 0.7414,
+      "step": 2292
+    },
+    {
+      "epoch": 1.137520168797319,
+      "grad_norm": 0.13837495741874617,
+      "learning_rate": 4.076236079812359e-06,
+      "loss": 0.7518,
+      "step": 2293
+    },
+    {
+      "epoch": 1.1380166315005584,
+      "grad_norm": 0.13909330761790814,
+      "learning_rate": 4.075477346684139e-06,
+      "loss": 0.7353,
+      "step": 2294
+    },
+    {
+      "epoch": 1.138513094203798,
+      "grad_norm": 0.15340107106559084,
+      "learning_rate": 4.074718372769467e-06,
+      "loss": 0.7625,
+      "step": 2295
+    },
+    {
+      "epoch": 1.1390095569070373,
+      "grad_norm": 0.14086941509084788,
+      "learning_rate": 4.07395915818434e-06,
+      "loss": 0.7586,
+      "step": 2296
+    },
+    {
+      "epoch": 1.1395060196102769,
+      "grad_norm": 0.13446534552526093,
+      "learning_rate": 4.073199703044793e-06,
+      "loss": 0.7621,
+      "step": 2297
+    },
+    {
+      "epoch": 1.1400024823135162,
+      "grad_norm": 0.1373097144881244,
+      "learning_rate": 4.072440007466896e-06,
+      "loss": 0.6873,
+      "step": 2298
+    },
+    {
+      "epoch": 1.1404989450167555,
+      "grad_norm": 0.13928210663269355,
+      "learning_rate": 4.071680071566756e-06,
+      "loss": 0.7514,
+      "step": 2299
+    },
+    {
+      "epoch": 1.140995407719995,
+      "grad_norm": 0.1421392811034866,
+      "learning_rate": 4.070919895460517e-06,
+      "loss": 0.7295,
+      "step": 2300
+    },
+    {
+      "epoch": 1.1414918704232344,
+      "grad_norm": 0.13241332535161587,
+      "learning_rate": 4.070159479264359e-06,
+      "loss": 0.7118,
+      "step": 2301
+    },
+    {
+      "epoch": 1.141988333126474,
+      "grad_norm": 0.1302518945245061,
+      "learning_rate": 4.0693988230945e-06,
+      "loss": 0.7399,
+      "step": 2302
+    },
+    {
+      "epoch": 1.1424847958297133,
+      "grad_norm": 0.14324912590708974,
+      "learning_rate": 4.0686379270671955e-06,
+      "loss": 0.768,
+      "step": 2303
+    },
+    {
+      "epoch": 1.1429812585329526,
+      "grad_norm": 0.14029305482378046,
+      "learning_rate": 4.067876791298734e-06,
+      "loss": 0.7787,
+      "step": 2304
+    },
+    {
+      "epoch": 1.1434777212361922,
+      "grad_norm": 0.15155995127547886,
+      "learning_rate": 4.067115415905445e-06,
+      "loss": 0.7777,
+      "step": 2305
+    },
+    {
+      "epoch": 1.1439741839394315,
+      "grad_norm": 0.13840654029220392,
+      "learning_rate": 4.066353801003691e-06,
+      "loss": 0.7449,
+      "step": 2306
+    },
+    {
+      "epoch": 1.144470646642671,
+      "grad_norm": 0.1379028297317787,
+      "learning_rate": 4.065591946709873e-06,
+      "loss": 0.7468,
+      "step": 2307
+    },
+    {
+      "epoch": 1.1449671093459104,
+      "grad_norm": 0.13315214722190571,
+      "learning_rate": 4.064829853140428e-06,
+      "loss": 0.7142,
+      "step": 2308
+    },
+    {
+      "epoch": 1.1454635720491497,
+      "grad_norm": 0.13949231615359325,
+      "learning_rate": 4.0640675204118305e-06,
+      "loss": 0.6793,
+      "step": 2309
+    },
+    {
+      "epoch": 1.1459600347523893,
+      "grad_norm": 0.14382388332090082,
+      "learning_rate": 4.06330494864059e-06,
+      "loss": 0.7437,
+      "step": 2310
+    },
+    {
+      "epoch": 1.1464564974556286,
+      "grad_norm": 0.14590984361270626,
+      "learning_rate": 4.062542137943254e-06,
+      "loss": 0.8103,
+      "step": 2311
+    },
+    {
+      "epoch": 1.1469529601588682,
+      "grad_norm": 0.1609583000360408,
+      "learning_rate": 4.061779088436406e-06,
+      "loss": 0.7311,
+      "step": 2312
+    },
+    {
+      "epoch": 1.1474494228621075,
+      "grad_norm": 0.13805487885883164,
+      "learning_rate": 4.061015800236665e-06,
+      "loss": 0.7186,
+      "step": 2313
+    },
+    {
+      "epoch": 1.1479458855653468,
+      "grad_norm": 0.13893797112646772,
+      "learning_rate": 4.060252273460688e-06,
+      "loss": 0.7381,
+      "step": 2314
+    },
+    {
+      "epoch": 1.1484423482685864,
+      "grad_norm": 0.13923921175817758,
+      "learning_rate": 4.0594885082251685e-06,
+      "loss": 0.7233,
+      "step": 2315
+    },
+    {
+      "epoch": 1.1489388109718257,
+      "grad_norm": 0.13902288626720571,
+      "learning_rate": 4.058724504646834e-06,
+      "loss": 0.7134,
+      "step": 2316
+    },
+    {
+      "epoch": 1.1494352736750653,
+      "grad_norm": 0.13827739981540818,
+      "learning_rate": 4.057960262842452e-06,
+      "loss": 0.7263,
+      "step": 2317
+    },
+    {
+      "epoch": 1.1499317363783046,
+      "grad_norm": 0.14118983269415886,
+      "learning_rate": 4.057195782928823e-06,
+      "loss": 0.7381,
+      "step": 2318
+    },
+    {
+      "epoch": 1.150428199081544,
+      "grad_norm": 0.1322019940248954,
+      "learning_rate": 4.056431065022787e-06,
+      "loss": 0.7327,
+      "step": 2319
+    },
+    {
+      "epoch": 1.1509246617847835,
+      "grad_norm": 0.1313464348071736,
+      "learning_rate": 4.055666109241218e-06,
+      "loss": 0.6943,
+      "step": 2320
+    },
+    {
+      "epoch": 1.1514211244880228,
+      "grad_norm": 0.13797475553246952,
+      "learning_rate": 4.054900915701028e-06,
+      "loss": 0.7245,
+      "step": 2321
+    },
+    {
+      "epoch": 1.1519175871912624,
+      "grad_norm": 0.15720508745478762,
+      "learning_rate": 4.054135484519163e-06,
+      "loss": 0.6975,
+      "step": 2322
+    },
+    {
+      "epoch": 1.1524140498945017,
+      "grad_norm": 0.14150123947563276,
+      "learning_rate": 4.053369815812608e-06,
+      "loss": 0.7432,
+      "step": 2323
+    },
+    {
+      "epoch": 1.152910512597741,
+      "grad_norm": 0.13546580390502527,
+      "learning_rate": 4.052603909698383e-06,
+      "loss": 0.6945,
+      "step": 2324
+    },
+    {
+      "epoch": 1.1534069753009806,
+      "grad_norm": 0.13797133959426647,
+      "learning_rate": 4.051837766293545e-06,
+      "loss": 0.7659,
+      "step": 2325
+    },
+    {
+      "epoch": 1.15390343800422,
+      "grad_norm": 0.13678662933843044,
+      "learning_rate": 4.051071385715186e-06,
+      "loss": 0.7662,
+      "step": 2326
+    },
+    {
+      "epoch": 1.1543999007074595,
+      "grad_norm": 0.1456527753758712,
+      "learning_rate": 4.0503047680804354e-06,
+      "loss": 0.7451,
+      "step": 2327
+    },
+    {
+      "epoch": 1.1548963634106988,
+      "grad_norm": 0.13498395892098344,
+      "learning_rate": 4.0495379135064585e-06,
+      "loss": 0.7201,
+      "step": 2328
+    },
+    {
+      "epoch": 1.1553928261139381,
+      "grad_norm": 0.13613972964983925,
+      "learning_rate": 4.048770822110456e-06,
+      "loss": 0.692,
+      "step": 2329
+    },
+    {
+      "epoch": 1.1558892888171777,
+      "grad_norm": 0.13981046419867807,
+      "learning_rate": 4.048003494009666e-06,
+      "loss": 0.7372,
+      "step": 2330
+    },
+    {
+      "epoch": 1.156385751520417,
+      "grad_norm": 0.13473584929200108,
+      "learning_rate": 4.047235929321363e-06,
+      "loss": 0.7036,
+      "step": 2331
+    },
+    {
+      "epoch": 1.1568822142236566,
+      "grad_norm": 0.13805910028899931,
+      "learning_rate": 4.0464681281628545e-06,
+      "loss": 0.6842,
+      "step": 2332
+    },
+    {
+      "epoch": 1.1573786769268959,
+      "grad_norm": 0.139073362785338,
+      "learning_rate": 4.04570009065149e-06,
+      "loss": 0.7541,
+      "step": 2333
+    },
+    {
+      "epoch": 1.1578751396301352,
+      "grad_norm": 0.13908629112696577,
+      "learning_rate": 4.044931816904649e-06,
+      "loss": 0.6851,
+      "step": 2334
+    },
+    {
+      "epoch": 1.1583716023333748,
+      "grad_norm": 0.13330790586530183,
+      "learning_rate": 4.044163307039751e-06,
+      "loss": 0.7023,
+      "step": 2335
+    },
+    {
+      "epoch": 1.158868065036614,
+      "grad_norm": 0.1516948361883498,
+      "learning_rate": 4.043394561174252e-06,
+      "loss": 0.7247,
+      "step": 2336
+    },
+    {
+      "epoch": 1.1593645277398537,
+      "grad_norm": 0.13561857587166704,
+      "learning_rate": 4.042625579425639e-06,
+      "loss": 0.6929,
+      "step": 2337
+    },
+    {
+      "epoch": 1.159860990443093,
+      "grad_norm": 0.1381505731871968,
+      "learning_rate": 4.04185636191144e-06,
+      "loss": 0.7877,
+      "step": 2338
+    },
+    {
+      "epoch": 1.1603574531463323,
+      "grad_norm": 0.14465367702750925,
+      "learning_rate": 4.04108690874922e-06,
+      "loss": 0.7335,
+      "step": 2339
+    },
+    {
+      "epoch": 1.1608539158495719,
+      "grad_norm": 0.13856535592329788,
+      "learning_rate": 4.040317220056574e-06,
+      "loss": 0.7261,
+      "step": 2340
+    },
+    {
+      "epoch": 1.1613503785528112,
+      "grad_norm": 0.1359117640924008,
+      "learning_rate": 4.039547295951138e-06,
+      "loss": 0.7154,
+      "step": 2341
+    },
+    {
+      "epoch": 1.1618468412560508,
+      "grad_norm": 0.14408892510554583,
+      "learning_rate": 4.038777136550583e-06,
+      "loss": 0.7449,
+      "step": 2342
+    },
+    {
+      "epoch": 1.16234330395929,
+      "grad_norm": 0.1392596970150696,
+      "learning_rate": 4.038006741972614e-06,
+      "loss": 0.7677,
+      "step": 2343
+    },
+    {
+      "epoch": 1.1628397666625294,
+      "grad_norm": 0.13340381888036884,
+      "learning_rate": 4.037236112334976e-06,
+      "loss": 0.6938,
+      "step": 2344
+    },
+    {
+      "epoch": 1.163336229365769,
+      "grad_norm": 0.138524100141458,
+      "learning_rate": 4.036465247755444e-06,
+      "loss": 0.678,
+      "step": 2345
+    },
+    {
+      "epoch": 1.1638326920690083,
+      "grad_norm": 0.13602968738361657,
+      "learning_rate": 4.035694148351836e-06,
+      "loss": 0.7189,
+      "step": 2346
+    },
+    {
+      "epoch": 1.1643291547722479,
+      "grad_norm": 0.13592578354190615,
+      "learning_rate": 4.034922814242001e-06,
+      "loss": 0.6993,
+      "step": 2347
+    },
+    {
+      "epoch": 1.1648256174754872,
+      "grad_norm": 0.13795965217373635,
+      "learning_rate": 4.034151245543823e-06,
+      "loss": 0.7135,
+      "step": 2348
+    },
+    {
+      "epoch": 1.1653220801787265,
+      "grad_norm": 0.13946922199269032,
+      "learning_rate": 4.033379442375225e-06,
+      "loss": 0.7757,
+      "step": 2349
+    },
+    {
+      "epoch": 1.165818542881966,
+      "grad_norm": 0.14045917203188657,
+      "learning_rate": 4.032607404854166e-06,
+      "loss": 0.7284,
+      "step": 2350
+    },
+    {
+      "epoch": 1.1663150055852054,
+      "grad_norm": 0.1440756518140952,
+      "learning_rate": 4.031835133098639e-06,
+      "loss": 0.7247,
+      "step": 2351
+    },
+    {
+      "epoch": 1.166811468288445,
+      "grad_norm": 0.13724789791500103,
+      "learning_rate": 4.031062627226671e-06,
+      "loss": 0.6963,
+      "step": 2352
+    },
+    {
+      "epoch": 1.1673079309916843,
+      "grad_norm": 0.13553830052897844,
+      "learning_rate": 4.030289887356332e-06,
+      "loss": 0.6995,
+      "step": 2353
+    },
+    {
+      "epoch": 1.1678043936949236,
+      "grad_norm": 0.13522250185478063,
+      "learning_rate": 4.029516913605719e-06,
+      "loss": 0.7359,
+      "step": 2354
+    },
+    {
+      "epoch": 1.1683008563981632,
+      "grad_norm": 0.1386403051237985,
+      "learning_rate": 4.028743706092969e-06,
+      "loss": 0.7783,
+      "step": 2355
+    },
+    {
+      "epoch": 1.1687973191014025,
+      "grad_norm": 0.14135795928242792,
+      "learning_rate": 4.027970264936256e-06,
+      "loss": 0.7393,
+      "step": 2356
+    },
+    {
+      "epoch": 1.1692937818046418,
+      "grad_norm": 0.15298427203204537,
+      "learning_rate": 4.027196590253786e-06,
+      "loss": 0.7416,
+      "step": 2357
+    },
+    {
+      "epoch": 1.1697902445078814,
+      "grad_norm": 0.13356310495808743,
+      "learning_rate": 4.026422682163804e-06,
+      "loss": 0.7174,
+      "step": 2358
+    },
+    {
+      "epoch": 1.1702867072111207,
+      "grad_norm": 0.14258979851062023,
+      "learning_rate": 4.02564854078459e-06,
+      "loss": 0.7025,
+      "step": 2359
+    },
+    {
+      "epoch": 1.1707831699143603,
+      "grad_norm": 0.1394231832343289,
+      "learning_rate": 4.024874166234459e-06,
+      "loss": 0.7084,
+      "step": 2360
+    },
+    {
+      "epoch": 1.1712796326175996,
+      "grad_norm": 0.1311008710936396,
+      "learning_rate": 4.02409955863176e-06,
+      "loss": 0.7364,
+      "step": 2361
+    },
+    {
+      "epoch": 1.171776095320839,
+      "grad_norm": 0.13638386404046124,
+      "learning_rate": 4.023324718094881e-06,
+      "loss": 0.7396,
+      "step": 2362
+    },
+    {
+      "epoch": 1.1722725580240785,
+      "grad_norm": 0.14537907916628248,
+      "learning_rate": 4.022549644742244e-06,
+      "loss": 0.6914,
+      "step": 2363
+    },
+    {
+      "epoch": 1.1727690207273178,
+      "grad_norm": 0.14040298917938324,
+      "learning_rate": 4.0217743386923055e-06,
+      "loss": 0.7228,
+      "step": 2364
+    },
+    {
+      "epoch": 1.1732654834305574,
+      "grad_norm": 0.14446607932799382,
+      "learning_rate": 4.020998800063559e-06,
+      "loss": 0.7441,
+      "step": 2365
+    },
+    {
+      "epoch": 1.1737619461337967,
+      "grad_norm": 0.13665124704375234,
+      "learning_rate": 4.020223028974534e-06,
+      "loss": 0.7516,
+      "step": 2366
+    },
+    {
+      "epoch": 1.174258408837036,
+      "grad_norm": 0.13399912735468633,
+      "learning_rate": 4.019447025543793e-06,
+      "loss": 0.7183,
+      "step": 2367
+    },
+    {
+      "epoch": 1.1747548715402756,
+      "grad_norm": 0.1464865537624882,
+      "learning_rate": 4.018670789889938e-06,
+      "loss": 0.7937,
+      "step": 2368
+    },
+    {
+      "epoch": 1.175251334243515,
+      "grad_norm": 0.1420811329454974,
+      "learning_rate": 4.0178943221316014e-06,
+      "loss": 0.728,
+      "step": 2369
+    },
+    {
+      "epoch": 1.1757477969467545,
+      "grad_norm": 0.1432300356176437,
+      "learning_rate": 4.0171176223874555e-06,
+      "loss": 0.657,
+      "step": 2370
+    },
+    {
+      "epoch": 1.1762442596499938,
+      "grad_norm": 0.1432930911399873,
+      "learning_rate": 4.0163406907762074e-06,
+      "loss": 0.76,
+      "step": 2371
+    },
+    {
+      "epoch": 1.1767407223532331,
+      "grad_norm": 0.14260952105955038,
+      "learning_rate": 4.015563527416596e-06,
+      "loss": 0.75,
+      "step": 2372
+    },
+    {
+      "epoch": 1.1772371850564727,
+      "grad_norm": 0.14070002945171584,
+      "learning_rate": 4.0147861324273994e-06,
+      "loss": 0.7111,
+      "step": 2373
+    },
+    {
+      "epoch": 1.177733647759712,
+      "grad_norm": 0.13824362597359197,
+      "learning_rate": 4.014008505927431e-06,
+      "loss": 0.7146,
+      "step": 2374
+    },
+    {
+      "epoch": 1.1782301104629516,
+      "grad_norm": 0.13615039910533921,
+      "learning_rate": 4.0132306480355374e-06,
+      "loss": 0.6898,
+      "step": 2375
+    },
+    {
+      "epoch": 1.178726573166191,
+      "grad_norm": 0.13640593448421545,
+      "learning_rate": 4.012452558870602e-06,
+      "loss": 0.7966,
+      "step": 2376
+    },
+    {
+      "epoch": 1.1792230358694302,
+      "grad_norm": 0.13629910190444364,
+      "learning_rate": 4.011674238551544e-06,
+      "loss": 0.751,
+      "step": 2377
+    },
+    {
+      "epoch": 1.1797194985726698,
+      "grad_norm": 0.13855859156200617,
+      "learning_rate": 4.010895687197316e-06,
+      "loss": 0.7481,
+      "step": 2378
+    },
+    {
+      "epoch": 1.180215961275909,
+      "grad_norm": 0.14195565031842594,
+      "learning_rate": 4.010116904926908e-06,
+      "loss": 0.7214,
+      "step": 2379
+    },
+    {
+      "epoch": 1.1807124239791487,
+      "grad_norm": 0.14168331391237854,
+      "learning_rate": 4.009337891859343e-06,
+      "loss": 0.7386,
+      "step": 2380
+    },
+    {
+      "epoch": 1.181208886682388,
+      "grad_norm": 0.13835896324890185,
+      "learning_rate": 4.008558648113682e-06,
+      "loss": 0.7219,
+      "step": 2381
+    },
+    {
+      "epoch": 1.1817053493856273,
+      "grad_norm": 0.14190577391365433,
+      "learning_rate": 4.0077791738090195e-06,
+      "loss": 0.7628,
+      "step": 2382
+    },
+    {
+      "epoch": 1.1822018120888669,
+      "grad_norm": 0.1341911874683844,
+      "learning_rate": 4.006999469064487e-06,
+      "loss": 0.6969,
+      "step": 2383
+    },
+    {
+      "epoch": 1.1826982747921062,
+      "grad_norm": 0.13969845313519358,
+      "learning_rate": 4.006219533999247e-06,
+      "loss": 0.7244,
+      "step": 2384
+    },
+    {
+      "epoch": 1.1831947374953455,
+      "grad_norm": 0.13979366267677537,
+      "learning_rate": 4.005439368732502e-06,
+      "loss": 0.7645,
+      "step": 2385
+    },
+    {
+      "epoch": 1.183691200198585,
+      "grad_norm": 0.1335214336393622,
+      "learning_rate": 4.0046589733834875e-06,
+      "loss": 0.688,
+      "step": 2386
+    },
+    {
+      "epoch": 1.1841876629018244,
+      "grad_norm": 0.14780650963457725,
+      "learning_rate": 4.003878348071474e-06,
+      "loss": 0.7289,
+      "step": 2387
+    },
+    {
+      "epoch": 1.184684125605064,
+      "grad_norm": 0.1340780920943342,
+      "learning_rate": 4.0030974929157685e-06,
+      "loss": 0.6878,
+      "step": 2388
+    },
+    {
+      "epoch": 1.1851805883083033,
+      "grad_norm": 0.1359899436643017,
+      "learning_rate": 4.002316408035711e-06,
+      "loss": 0.7322,
+      "step": 2389
+    },
+    {
+      "epoch": 1.1856770510115426,
+      "grad_norm": 0.13282726421917596,
+      "learning_rate": 4.001535093550678e-06,
+      "loss": 0.7032,
+      "step": 2390
+    },
+    {
+      "epoch": 1.1861735137147822,
+      "grad_norm": 0.13903709744169873,
+      "learning_rate": 4.000753549580082e-06,
+      "loss": 0.6988,
+      "step": 2391
+    },
+    {
+      "epoch": 1.1866699764180215,
+      "grad_norm": 0.1375603444876677,
+      "learning_rate": 3.999971776243369e-06,
+      "loss": 0.7292,
+      "step": 2392
+    },
+    {
+      "epoch": 1.187166439121261,
+      "grad_norm": 0.13837931077084456,
+      "learning_rate": 3.999189773660019e-06,
+      "loss": 0.7046,
+      "step": 2393
+    },
+    {
+      "epoch": 1.1876629018245004,
+      "grad_norm": 0.13869342809783666,
+      "learning_rate": 3.998407541949551e-06,
+      "loss": 0.6902,
+      "step": 2394
+    },
+    {
+      "epoch": 1.1881593645277397,
+      "grad_norm": 0.13979740548264746,
+      "learning_rate": 3.997625081231514e-06,
+      "loss": 0.7353,
+      "step": 2395
+    },
+    {
+      "epoch": 1.1886558272309793,
+      "grad_norm": 0.13527705164090786,
+      "learning_rate": 3.996842391625497e-06,
+      "loss": 0.726,
+      "step": 2396
+    },
+    {
+      "epoch": 1.1891522899342186,
+      "grad_norm": 0.14174483460790632,
+      "learning_rate": 3.99605947325112e-06,
+      "loss": 0.7005,
+      "step": 2397
+    },
+    {
+      "epoch": 1.1896487526374582,
+      "grad_norm": 0.13548057476768477,
+      "learning_rate": 3.99527632622804e-06,
+      "loss": 0.7066,
+      "step": 2398
+    },
+    {
+      "epoch": 1.1901452153406975,
+      "grad_norm": 0.13722530722455226,
+      "learning_rate": 3.994492950675949e-06,
+      "loss": 0.7293,
+      "step": 2399
+    },
+    {
+      "epoch": 1.1906416780439368,
+      "grad_norm": 0.14238980858246564,
+      "learning_rate": 3.9937093467145725e-06,
+      "loss": 0.7513,
+      "step": 2400
+    },
+    {
+      "epoch": 1.1911381407471764,
+      "grad_norm": 0.14077263746379465,
+      "learning_rate": 3.992925514463672e-06,
+      "loss": 0.6797,
+      "step": 2401
+    },
+    {
+      "epoch": 1.1916346034504157,
+      "grad_norm": 0.13906663141907152,
+      "learning_rate": 3.992141454043045e-06,
+      "loss": 0.705,
+      "step": 2402
+    },
+    {
+      "epoch": 1.1921310661536553,
+      "grad_norm": 0.14643513675700848,
+      "learning_rate": 3.99135716557252e-06,
+      "loss": 0.7516,
+      "step": 2403
+    },
+    {
+      "epoch": 1.1926275288568946,
+      "grad_norm": 0.1356991566400479,
+      "learning_rate": 3.990572649171964e-06,
+      "loss": 0.7296,
+      "step": 2404
+    },
+    {
+      "epoch": 1.193123991560134,
+      "grad_norm": 0.13520242104672095,
+      "learning_rate": 3.9897879049612795e-06,
+      "loss": 0.7248,
+      "step": 2405
+    },
+    {
+      "epoch": 1.1936204542633735,
+      "grad_norm": 0.14080671561882466,
+      "learning_rate": 3.9890029330604e-06,
+      "loss": 0.7011,
+      "step": 2406
+    },
+    {
+      "epoch": 1.1941169169666128,
+      "grad_norm": 0.14134376264347104,
+      "learning_rate": 3.988217733589296e-06,
+      "loss": 0.7477,
+      "step": 2407
+    },
+    {
+      "epoch": 1.1946133796698524,
+      "grad_norm": 0.13870900196918226,
+      "learning_rate": 3.987432306667975e-06,
+      "loss": 0.7606,
+      "step": 2408
+    },
+    {
+      "epoch": 1.1951098423730917,
+      "grad_norm": 0.1498135326898326,
+      "learning_rate": 3.986646652416473e-06,
+      "loss": 0.7166,
+      "step": 2409
+    },
+    {
+      "epoch": 1.195606305076331,
+      "grad_norm": 0.13666015376761742,
+      "learning_rate": 3.985860770954867e-06,
+      "loss": 0.6994,
+      "step": 2410
+    },
+    {
+      "epoch": 1.1961027677795706,
+      "grad_norm": 0.14270391448274597,
+      "learning_rate": 3.985074662403267e-06,
+      "loss": 0.7084,
+      "step": 2411
+    },
+    {
+      "epoch": 1.19659923048281,
+      "grad_norm": 0.13549585466845787,
+      "learning_rate": 3.984288326881817e-06,
+      "loss": 0.7065,
+      "step": 2412
+    },
+    {
+      "epoch": 1.1970956931860495,
+      "grad_norm": 0.13648442428428606,
+      "learning_rate": 3.983501764510693e-06,
+      "loss": 0.7372,
+      "step": 2413
+    },
+    {
+      "epoch": 1.1975921558892888,
+      "grad_norm": 0.14086948426447465,
+      "learning_rate": 3.982714975410111e-06,
+      "loss": 0.7205,
+      "step": 2414
+    },
+    {
+      "epoch": 1.1980886185925281,
+      "grad_norm": 0.1325423947366576,
+      "learning_rate": 3.9819279597003195e-06,
+      "loss": 0.6983,
+      "step": 2415
+    },
+    {
+      "epoch": 1.1985850812957677,
+      "grad_norm": 0.13634160260964182,
+      "learning_rate": 3.9811407175015995e-06,
+      "loss": 0.6579,
+      "step": 2416
+    },
+    {
+      "epoch": 1.199081543999007,
+      "grad_norm": 0.13874737933942757,
+      "learning_rate": 3.980353248934269e-06,
+      "loss": 0.7378,
+      "step": 2417
+    },
+    {
+      "epoch": 1.1995780067022466,
+      "grad_norm": 0.13788946383443576,
+      "learning_rate": 3.9795655541186805e-06,
+      "loss": 0.7417,
+      "step": 2418
+    },
+    {
+      "epoch": 1.200074469405486,
+      "grad_norm": 0.13647093705243393,
+      "learning_rate": 3.97877763317522e-06,
+      "loss": 0.7199,
+      "step": 2419
+    },
+    {
+      "epoch": 1.2005709321087252,
+      "grad_norm": 0.14052612906793216,
+      "learning_rate": 3.977989486224309e-06,
+      "loss": 0.7305,
+      "step": 2420
+    },
+    {
+      "epoch": 1.2010673948119648,
+      "grad_norm": 0.14559651220194383,
+      "learning_rate": 3.977201113386402e-06,
+      "loss": 0.7112,
+      "step": 2421
+    },
+    {
+      "epoch": 1.2015638575152041,
+      "grad_norm": 0.14825330957397817,
+      "learning_rate": 3.97641251478199e-06,
+      "loss": 0.798,
+      "step": 2422
+    },
+    {
+      "epoch": 1.2020603202184437,
+      "grad_norm": 0.1411793849816043,
+      "learning_rate": 3.975623690531598e-06,
+      "loss": 0.775,
+      "step": 2423
+    },
+    {
+      "epoch": 1.202556782921683,
+      "grad_norm": 0.13329544412504718,
+      "learning_rate": 3.9748346407557845e-06,
+      "loss": 0.7021,
+      "step": 2424
+    },
+    {
+      "epoch": 1.2030532456249223,
+      "grad_norm": 0.1322912568134602,
+      "learning_rate": 3.9740453655751435e-06,
+      "loss": 0.7112,
+      "step": 2425
+    },
+    {
+      "epoch": 1.2035497083281619,
+      "grad_norm": 0.1387834596545233,
+      "learning_rate": 3.973255865110302e-06,
+      "loss": 0.698,
+      "step": 2426
+    },
+    {
+      "epoch": 1.2040461710314012,
+      "grad_norm": 0.14956785941966386,
+      "learning_rate": 3.972466139481925e-06,
+      "loss": 0.7206,
+      "step": 2427
+    },
+    {
+      "epoch": 1.2045426337346408,
+      "grad_norm": 0.1322262204706779,
+      "learning_rate": 3.971676188810707e-06,
+      "loss": 0.7204,
+      "step": 2428
+    },
+    {
+      "epoch": 1.20503909643788,
+      "grad_norm": 0.12896570477009364,
+      "learning_rate": 3.9708860132173795e-06,
+      "loss": 0.695,
+      "step": 2429
+    },
+    {
+      "epoch": 1.2055355591411194,
+      "grad_norm": 0.13536258893402728,
+      "learning_rate": 3.97009561282271e-06,
+      "loss": 0.7283,
+      "step": 2430
+    },
+    {
+      "epoch": 1.206032021844359,
+      "grad_norm": 0.13492371176463214,
+      "learning_rate": 3.969304987747496e-06,
+      "loss": 0.7073,
+      "step": 2431
+    },
+    {
+      "epoch": 1.2065284845475983,
+      "grad_norm": 0.13134753297863244,
+      "learning_rate": 3.9685141381125745e-06,
+      "loss": 0.7089,
+      "step": 2432
+    },
+    {
+      "epoch": 1.2070249472508379,
+      "grad_norm": 0.13554849626488005,
+      "learning_rate": 3.967723064038812e-06,
+      "loss": 0.735,
+      "step": 2433
+    },
+    {
+      "epoch": 1.2075214099540772,
+      "grad_norm": 0.14349704386244447,
+      "learning_rate": 3.966931765647112e-06,
+      "loss": 0.7232,
+      "step": 2434
+    },
+    {
+      "epoch": 1.2080178726573165,
+      "grad_norm": 0.1385163792731118,
+      "learning_rate": 3.966140243058413e-06,
+      "loss": 0.7218,
+      "step": 2435
+    },
+    {
+      "epoch": 1.208514335360556,
+      "grad_norm": 0.14279978031188895,
+      "learning_rate": 3.965348496393685e-06,
+      "loss": 0.8045,
+      "step": 2436
+    },
+    {
+      "epoch": 1.2090107980637954,
+      "grad_norm": 0.13858682489268342,
+      "learning_rate": 3.964556525773935e-06,
+      "loss": 0.7069,
+      "step": 2437
+    },
+    {
+      "epoch": 1.209507260767035,
+      "grad_norm": 0.1275555575643711,
+      "learning_rate": 3.963764331320201e-06,
+      "loss": 0.7232,
+      "step": 2438
+    },
+    {
+      "epoch": 1.2100037234702743,
+      "grad_norm": 0.13319214619249178,
+      "learning_rate": 3.9629719131535595e-06,
+      "loss": 0.7551,
+      "step": 2439
+    },
+    {
+      "epoch": 1.2105001861735136,
+      "grad_norm": 0.1387567942220467,
+      "learning_rate": 3.962179271395118e-06,
+      "loss": 0.7367,
+      "step": 2440
+    },
+    {
+      "epoch": 1.2109966488767532,
+      "grad_norm": 0.1373104804870825,
+      "learning_rate": 3.961386406166019e-06,
+      "loss": 0.7157,
+      "step": 2441
+    },
+    {
+      "epoch": 1.2114931115799925,
+      "grad_norm": 0.13779972795149475,
+      "learning_rate": 3.96059331758744e-06,
+      "loss": 0.7441,
+      "step": 2442
+    },
+    {
+      "epoch": 1.211989574283232,
+      "grad_norm": 0.13902570136362163,
+      "learning_rate": 3.9598000057805905e-06,
+      "loss": 0.7297,
+      "step": 2443
+    },
+    {
+      "epoch": 1.2124860369864714,
+      "grad_norm": 0.14200167374497805,
+      "learning_rate": 3.959006470866717e-06,
+      "loss": 0.7664,
+      "step": 2444
+    },
+    {
+      "epoch": 1.2129824996897107,
+      "grad_norm": 0.13182479218295906,
+      "learning_rate": 3.958212712967097e-06,
+      "loss": 0.7333,
+      "step": 2445
+    },
+    {
+      "epoch": 1.2134789623929503,
+      "grad_norm": 0.13500957253720283,
+      "learning_rate": 3.957418732203045e-06,
+      "loss": 0.7128,
+      "step": 2446
+    },
+    {
+      "epoch": 1.2139754250961896,
+      "grad_norm": 0.1364082757123999,
+      "learning_rate": 3.9566245286959074e-06,
+      "loss": 0.7516,
+      "step": 2447
+    },
+    {
+      "epoch": 1.2144718877994292,
+      "grad_norm": 0.15584714093121874,
+      "learning_rate": 3.955830102567066e-06,
+      "loss": 0.7542,
+      "step": 2448
+    },
+    {
+      "epoch": 1.2149683505026685,
+      "grad_norm": 0.13446548054487187,
+      "learning_rate": 3.955035453937936e-06,
+      "loss": 0.7358,
+      "step": 2449
+    },
+    {
+      "epoch": 1.2154648132059078,
+      "grad_norm": 0.13145486309075377,
+      "learning_rate": 3.954240582929965e-06,
+      "loss": 0.7025,
+      "step": 2450
+    },
+    {
+      "epoch": 1.2159612759091474,
+      "grad_norm": 0.13574735830422388,
+      "learning_rate": 3.953445489664641e-06,
+      "loss": 0.7401,
+      "step": 2451
+    },
+    {
+      "epoch": 1.2164577386123867,
+      "grad_norm": 0.1366039565699845,
+      "learning_rate": 3.952650174263476e-06,
+      "loss": 0.7242,
+      "step": 2452
+    },
+    {
+      "epoch": 1.2169542013156263,
+      "grad_norm": 0.13329175729167014,
+      "learning_rate": 3.9518546368480235e-06,
+      "loss": 0.701,
+      "step": 2453
+    },
+    {
+      "epoch": 1.2174506640188656,
+      "grad_norm": 0.13635014090129313,
+      "learning_rate": 3.951058877539869e-06,
+      "loss": 0.7085,
+      "step": 2454
+    },
+    {
+      "epoch": 1.217947126722105,
+      "grad_norm": 0.14550135826438018,
+      "learning_rate": 3.95026289646063e-06,
+      "loss": 0.7563,
+      "step": 2455
+    },
+    {
+      "epoch": 1.2184435894253445,
+      "grad_norm": 0.13708956139933903,
+      "learning_rate": 3.949466693731962e-06,
+      "loss": 0.7396,
+      "step": 2456
+    },
+    {
+      "epoch": 1.2189400521285838,
+      "grad_norm": 0.13788736852218864,
+      "learning_rate": 3.948670269475549e-06,
+      "loss": 0.7099,
+      "step": 2457
+    },
+    {
+      "epoch": 1.2194365148318234,
+      "grad_norm": 0.1472672823984794,
+      "learning_rate": 3.947873623813114e-06,
+      "loss": 0.7551,
+      "step": 2458
+    },
+    {
+      "epoch": 1.2199329775350627,
+      "grad_norm": 0.13567982397574574,
+      "learning_rate": 3.94707675686641e-06,
+      "loss": 0.7215,
+      "step": 2459
+    },
+    {
+      "epoch": 1.220429440238302,
+      "grad_norm": 0.12785953951320136,
+      "learning_rate": 3.946279668757226e-06,
+      "loss": 0.6893,
+      "step": 2460
+    },
+    {
+      "epoch": 1.2209259029415416,
+      "grad_norm": 0.139042104896544,
+      "learning_rate": 3.945482359607383e-06,
+      "loss": 0.7116,
+      "step": 2461
+    },
+    {
+      "epoch": 1.221422365644781,
+      "grad_norm": 0.1310146478620868,
+      "learning_rate": 3.94468482953874e-06,
+      "loss": 0.6821,
+      "step": 2462
+    },
+    {
+      "epoch": 1.2219188283480205,
+      "grad_norm": 0.14080053665645972,
+      "learning_rate": 3.943887078673182e-06,
+      "loss": 0.72,
+      "step": 2463
+    },
+    {
+      "epoch": 1.2224152910512598,
+      "grad_norm": 0.1374193970967945,
+      "learning_rate": 3.943089107132637e-06,
+      "loss": 0.7277,
+      "step": 2464
+    },
+    {
+      "epoch": 1.2229117537544991,
+      "grad_norm": 0.13142687505073025,
+      "learning_rate": 3.942290915039059e-06,
+      "loss": 0.7756,
+      "step": 2465
+    },
+    {
+      "epoch": 1.2234082164577387,
+      "grad_norm": 0.1383208661807758,
+      "learning_rate": 3.94149250251444e-06,
+      "loss": 0.7427,
+      "step": 2466
+    },
+    {
+      "epoch": 1.223904679160978,
+      "grad_norm": 0.13946709784881417,
+      "learning_rate": 3.940693869680805e-06,
+      "loss": 0.7091,
+      "step": 2467
+    },
+    {
+      "epoch": 1.2244011418642176,
+      "grad_norm": 0.13594769238575358,
+      "learning_rate": 3.939895016660212e-06,
+      "loss": 0.7248,
+      "step": 2468
+    },
+    {
+      "epoch": 1.2248976045674569,
+      "grad_norm": 0.13618933752020215,
+      "learning_rate": 3.939095943574752e-06,
+      "loss": 0.7535,
+      "step": 2469
+    },
+    {
+      "epoch": 1.2253940672706962,
+      "grad_norm": 0.14274958557643477,
+      "learning_rate": 3.938296650546552e-06,
+      "loss": 0.7819,
+      "step": 2470
+    },
+    {
+      "epoch": 1.2258905299739358,
+      "grad_norm": 0.13553582376705678,
+      "learning_rate": 3.93749713769777e-06,
+      "loss": 0.7561,
+      "step": 2471
+    },
+    {
+      "epoch": 1.226386992677175,
+      "grad_norm": 0.14247983844696746,
+      "learning_rate": 3.936697405150599e-06,
+      "loss": 0.7169,
+      "step": 2472
+    },
+    {
+      "epoch": 1.2268834553804147,
+      "grad_norm": 0.13401472437678552,
+      "learning_rate": 3.935897453027265e-06,
+      "loss": 0.7331,
+      "step": 2473
+    },
+    {
+      "epoch": 1.227379918083654,
+      "grad_norm": 0.13467273748712041,
+      "learning_rate": 3.935097281450027e-06,
+      "loss": 0.7224,
+      "step": 2474
+    },
+    {
+      "epoch": 1.2278763807868933,
+      "grad_norm": 0.13626379999904872,
+      "learning_rate": 3.934296890541182e-06,
+      "loss": 0.7443,
+      "step": 2475
+    },
+    {
+      "epoch": 1.2283728434901329,
+      "grad_norm": 0.13003861644898487,
+      "learning_rate": 3.933496280423054e-06,
+      "loss": 0.6546,
+      "step": 2476
+    },
+    {
+      "epoch": 1.2288693061933722,
+      "grad_norm": 0.1425981221162396,
+      "learning_rate": 3.932695451218003e-06,
+      "loss": 0.7141,
+      "step": 2477
+    },
+    {
+      "epoch": 1.2293657688966118,
+      "grad_norm": 0.13560701136791695,
+      "learning_rate": 3.931894403048424e-06,
+      "loss": 0.7208,
+      "step": 2478
+    },
+    {
+      "epoch": 1.229862231599851,
+      "grad_norm": 0.1401873943379598,
+      "learning_rate": 3.931093136036744e-06,
+      "loss": 0.7266,
+      "step": 2479
+    },
+    {
+      "epoch": 1.2303586943030904,
+      "grad_norm": 0.13728694609128989,
+      "learning_rate": 3.930291650305424e-06,
+      "loss": 0.7134,
+      "step": 2480
+    },
+    {
+      "epoch": 1.23085515700633,
+      "grad_norm": 0.13218011780198854,
+      "learning_rate": 3.929489945976959e-06,
+      "loss": 0.7239,
+      "step": 2481
+    },
+    {
+      "epoch": 1.2313516197095693,
+      "grad_norm": 0.1410943689944911,
+      "learning_rate": 3.928688023173875e-06,
+      "loss": 0.7284,
+      "step": 2482
+    },
+    {
+      "epoch": 1.2318480824128089,
+      "grad_norm": 0.1361830396710614,
+      "learning_rate": 3.927885882018735e-06,
+      "loss": 0.7211,
+      "step": 2483
+    },
+    {
+      "epoch": 1.2323445451160482,
+      "grad_norm": 0.1351899227297423,
+      "learning_rate": 3.927083522634132e-06,
+      "loss": 0.7422,
+      "step": 2484
+    },
+    {
+      "epoch": 1.2328410078192875,
+      "grad_norm": 0.1406973288937524,
+      "learning_rate": 3.926280945142693e-06,
+      "loss": 0.7229,
+      "step": 2485
+    },
+    {
+      "epoch": 1.233337470522527,
+      "grad_norm": 0.13754991570942163,
+      "learning_rate": 3.925478149667081e-06,
+      "loss": 0.7677,
+      "step": 2486
+    },
+    {
+      "epoch": 1.2338339332257664,
+      "grad_norm": 0.1384634788260933,
+      "learning_rate": 3.92467513632999e-06,
+      "loss": 0.7408,
+      "step": 2487
+    },
+    {
+      "epoch": 1.234330395929006,
+      "grad_norm": 0.14392311337089933,
+      "learning_rate": 3.923871905254146e-06,
+      "loss": 0.7471,
+      "step": 2488
+    },
+    {
+      "epoch": 1.2348268586322453,
+      "grad_norm": 0.1335035566492651,
+      "learning_rate": 3.9230684565623135e-06,
+      "loss": 0.6966,
+      "step": 2489
+    },
+    {
+      "epoch": 1.2353233213354846,
+      "grad_norm": 0.132634223213086,
+      "learning_rate": 3.922264790377283e-06,
+      "loss": 0.6934,
+      "step": 2490
+    },
+    {
+      "epoch": 1.2358197840387242,
+      "grad_norm": 0.13827031271642745,
+      "learning_rate": 3.921460906821884e-06,
+      "loss": 0.7098,
+      "step": 2491
+    },
+    {
+      "epoch": 1.2363162467419635,
+      "grad_norm": 0.14451954689789367,
+      "learning_rate": 3.920656806018977e-06,
+      "loss": 0.788,
+      "step": 2492
+    },
+    {
+      "epoch": 1.236812709445203,
+      "grad_norm": 0.13561380554064417,
+      "learning_rate": 3.919852488091455e-06,
+      "loss": 0.7527,
+      "step": 2493
+    },
+    {
+      "epoch": 1.2373091721484424,
+      "grad_norm": 0.13501364970911942,
+      "learning_rate": 3.9190479531622465e-06,
+      "loss": 0.7612,
+      "step": 2494
+    },
+    {
+      "epoch": 1.2378056348516817,
+      "grad_norm": 0.12818687643314722,
+      "learning_rate": 3.91824320135431e-06,
+      "loss": 0.6963,
+      "step": 2495
+    },
+    {
+      "epoch": 1.2383020975549213,
+      "grad_norm": 0.13312040030918668,
+      "learning_rate": 3.9174382327906415e-06,
+      "loss": 0.7064,
+      "step": 2496
+    },
+    {
+      "epoch": 1.2387985602581606,
+      "grad_norm": 0.13300761423052482,
+      "learning_rate": 3.916633047594265e-06,
+      "loss": 0.7077,
+      "step": 2497
+    },
+    {
+      "epoch": 1.2392950229614,
+      "grad_norm": 0.13620840402675596,
+      "learning_rate": 3.915827645888242e-06,
+      "loss": 0.7187,
+      "step": 2498
+    },
+    {
+      "epoch": 1.2397914856646395,
+      "grad_norm": 0.13426305401154814,
+      "learning_rate": 3.915022027795663e-06,
+      "loss": 0.69,
+      "step": 2499
+    },
+    {
+      "epoch": 1.2402879483678788,
+      "grad_norm": 0.1464897445532207,
+      "learning_rate": 3.914216193439657e-06,
+      "loss": 0.8251,
+      "step": 2500
+    },
+    {
+      "epoch": 1.2407844110711184,
+      "grad_norm": 0.13743582313648897,
+      "learning_rate": 3.91341014294338e-06,
+      "loss": 0.7444,
+      "step": 2501
+    },
+    {
+      "epoch": 1.2412808737743577,
+      "grad_norm": 0.13874526596758197,
+      "learning_rate": 3.912603876430025e-06,
+      "loss": 0.708,
+      "step": 2502
+    },
+    {
+      "epoch": 1.241777336477597,
+      "grad_norm": 0.13008996722051902,
+      "learning_rate": 3.9117973940228166e-06,
+      "loss": 0.7709,
+      "step": 2503
+    },
+    {
+      "epoch": 1.2422737991808366,
+      "grad_norm": 0.1355480166714516,
+      "learning_rate": 3.910990695845013e-06,
+      "loss": 0.6943,
+      "step": 2504
+    },
+    {
+      "epoch": 1.242770261884076,
+      "grad_norm": 0.14539960501531302,
+      "learning_rate": 3.910183782019905e-06,
+      "loss": 0.7631,
+      "step": 2505
+    },
+    {
+      "epoch": 1.2432667245873155,
+      "grad_norm": 0.14083590125770834,
+      "learning_rate": 3.909376652670818e-06,
+      "loss": 0.7338,
+      "step": 2506
+    },
+    {
+      "epoch": 1.2437631872905548,
+      "grad_norm": 0.1431658480547782,
+      "learning_rate": 3.908569307921106e-06,
+      "loss": 0.8085,
+      "step": 2507
+    },
+    {
+      "epoch": 1.2442596499937941,
+      "grad_norm": 0.14578702074830277,
+      "learning_rate": 3.90776174789416e-06,
+      "loss": 0.7466,
+      "step": 2508
+    },
+    {
+      "epoch": 1.2447561126970337,
+      "grad_norm": 0.13316331073208876,
+      "learning_rate": 3.906953972713403e-06,
+      "loss": 0.7069,
+      "step": 2509
+    },
+    {
+      "epoch": 1.245252575400273,
+      "grad_norm": 0.13867830074059864,
+      "learning_rate": 3.90614598250229e-06,
+      "loss": 0.7267,
+      "step": 2510
+    },
+    {
+      "epoch": 1.2457490381035126,
+      "grad_norm": 0.13076874574140407,
+      "learning_rate": 3.905337777384308e-06,
+      "loss": 0.698,
+      "step": 2511
+    },
+    {
+      "epoch": 1.246245500806752,
+      "grad_norm": 0.13962139881766297,
+      "learning_rate": 3.904529357482981e-06,
+      "loss": 0.7147,
+      "step": 2512
+    },
+    {
+      "epoch": 1.2467419635099912,
+      "grad_norm": 0.1462052190977223,
+      "learning_rate": 3.9037207229218615e-06,
+      "loss": 0.728,
+      "step": 2513
+    },
+    {
+      "epoch": 1.2472384262132308,
+      "grad_norm": 0.1418049627962229,
+      "learning_rate": 3.902911873824536e-06,
+      "loss": 0.7354,
+      "step": 2514
+    },
+    {
+      "epoch": 1.24773488891647,
+      "grad_norm": 0.14139015990729262,
+      "learning_rate": 3.902102810314625e-06,
+      "loss": 0.736,
+      "step": 2515
+    },
+    {
+      "epoch": 1.2482313516197097,
+      "grad_norm": 0.13974548545259946,
+      "learning_rate": 3.9012935325157805e-06,
+      "loss": 0.735,
+      "step": 2516
+    },
+    {
+      "epoch": 1.248727814322949,
+      "grad_norm": 0.14137472030628268,
+      "learning_rate": 3.900484040551688e-06,
+      "loss": 0.7538,
+      "step": 2517
+    },
+    {
+      "epoch": 1.2492242770261883,
+      "grad_norm": 0.13631373740449088,
+      "learning_rate": 3.899674334546064e-06,
+      "loss": 0.6965,
+      "step": 2518
+    },
+    {
+      "epoch": 1.2497207397294279,
+      "grad_norm": 0.13581041061864152,
+      "learning_rate": 3.898864414622661e-06,
+      "loss": 0.7229,
+      "step": 2519
+    },
+    {
+      "epoch": 1.2502172024326672,
+      "grad_norm": 0.1351856956359364,
+      "learning_rate": 3.89805428090526e-06,
+      "loss": 0.6548,
+      "step": 2520
+    },
+    {
+      "epoch": 1.2502172024326672,
+      "eval_loss": 0.742116391658783,
+      "eval_runtime": 135.9502,
+      "eval_samples_per_second": 223.266,
+      "eval_steps_per_second": 27.915,
+      "step": 2520
+    },
+    {
+      "epoch": 1.2507136651359065,
+      "grad_norm": 0.1384586104535219,
+      "learning_rate": 3.897243933517679e-06,
+      "loss": 0.7376,
+      "step": 2521
+    },
+    {
+      "epoch": 1.251210127839146,
+      "grad_norm": 0.16315907780591418,
+      "learning_rate": 3.896433372583766e-06,
+      "loss": 0.7179,
+      "step": 2522
+    },
+    {
+      "epoch": 1.2517065905423854,
+      "grad_norm": 0.13165553758104281,
+      "learning_rate": 3.895622598227402e-06,
+      "loss": 0.6968,
+      "step": 2523
+    },
+    {
+      "epoch": 1.252203053245625,
+      "grad_norm": 0.1468700776665726,
+      "learning_rate": 3.894811610572501e-06,
+      "loss": 0.7961,
+      "step": 2524
+    },
+    {
+      "epoch": 1.2526995159488643,
+      "grad_norm": 0.13484457344159942,
+      "learning_rate": 3.894000409743009e-06,
+      "loss": 0.6972,
+      "step": 2525
+    },
+    {
+      "epoch": 1.2531959786521036,
+      "grad_norm": 0.14298803461222828,
+      "learning_rate": 3.893188995862907e-06,
+      "loss": 0.7434,
+      "step": 2526
+    },
+    {
+      "epoch": 1.2536924413553432,
+      "grad_norm": 0.13247384088857173,
+      "learning_rate": 3.892377369056203e-06,
+      "loss": 0.697,
+      "step": 2527
+    },
+    {
+      "epoch": 1.2541889040585825,
+      "grad_norm": 0.12855401841158223,
+      "learning_rate": 3.8915655294469445e-06,
+      "loss": 0.7368,
+      "step": 2528
+    },
+    {
+      "epoch": 1.254685366761822,
+      "grad_norm": 0.1399262509376029,
+      "learning_rate": 3.890753477159206e-06,
+      "loss": 0.7499,
+      "step": 2529
+    },
+    {
+      "epoch": 1.2551818294650614,
+      "grad_norm": 0.1532990862064815,
+      "learning_rate": 3.8899412123170984e-06,
+      "loss": 0.7841,
+      "step": 2530
+    },
+    {
+      "epoch": 1.2556782921683007,
+      "grad_norm": 0.13912282728279968,
+      "learning_rate": 3.889128735044762e-06,
+      "loss": 0.7458,
+      "step": 2531
+    },
+    {
+      "epoch": 1.2561747548715403,
+      "grad_norm": 0.1369771636022506,
+      "learning_rate": 3.888316045466372e-06,
+      "loss": 0.7154,
+      "step": 2532
+    },
+    {
+      "epoch": 1.2566712175747796,
+      "grad_norm": 0.13198426469797134,
+      "learning_rate": 3.887503143706134e-06,
+      "loss": 0.674,
+      "step": 2533
+    },
+    {
+      "epoch": 1.2571676802780192,
+      "grad_norm": 0.14611135628302505,
+      "learning_rate": 3.886690029888287e-06,
+      "loss": 0.7117,
+      "step": 2534
+    },
+    {
+      "epoch": 1.2576641429812585,
+      "grad_norm": 0.1385484671417284,
+      "learning_rate": 3.885876704137104e-06,
+      "loss": 0.7399,
+      "step": 2535
+    },
+    {
+      "epoch": 1.2581606056844978,
+      "grad_norm": 0.13166836823073533,
+      "learning_rate": 3.885063166576886e-06,
+      "loss": 0.7609,
+      "step": 2536
+    },
+    {
+      "epoch": 1.2586570683877374,
+      "grad_norm": 0.1387003911974032,
+      "learning_rate": 3.8842494173319726e-06,
+      "loss": 0.725,
+      "step": 2537
+    },
+    {
+      "epoch": 1.2591535310909767,
+      "grad_norm": 0.13305768328582457,
+      "learning_rate": 3.883435456526728e-06,
+      "loss": 0.7375,
+      "step": 2538
+    },
+    {
+      "epoch": 1.2596499937942163,
+      "grad_norm": 0.13978466433671136,
+      "learning_rate": 3.882621284285558e-06,
+      "loss": 0.7691,
+      "step": 2539
+    },
+    {
+      "epoch": 1.2601464564974556,
+      "grad_norm": 0.13757446247537838,
+      "learning_rate": 3.881806900732893e-06,
+      "loss": 0.7069,
+      "step": 2540
+    },
+    {
+      "epoch": 1.260642919200695,
+      "grad_norm": 0.14032437189289912,
+      "learning_rate": 3.880992305993198e-06,
+      "loss": 0.7281,
+      "step": 2541
+    },
+    {
+      "epoch": 1.2611393819039345,
+      "grad_norm": 0.1377868179608081,
+      "learning_rate": 3.880177500190971e-06,
+      "loss": 0.7587,
+      "step": 2542
+    },
+    {
+      "epoch": 1.2616358446071738,
+      "grad_norm": 0.14861646208837478,
+      "learning_rate": 3.8793624834507435e-06,
+      "loss": 0.7251,
+      "step": 2543
+    },
+    {
+      "epoch": 1.2621323073104134,
+      "grad_norm": 0.13517554466674267,
+      "learning_rate": 3.8785472558970776e-06,
+      "loss": 0.6974,
+      "step": 2544
+    },
+    {
+      "epoch": 1.2626287700136527,
+      "grad_norm": 0.1309887066083378,
+      "learning_rate": 3.877731817654566e-06,
+      "loss": 0.7402,
+      "step": 2545
+    },
+    {
+      "epoch": 1.263125232716892,
+      "grad_norm": 0.1396787950926557,
+      "learning_rate": 3.876916168847836e-06,
+      "loss": 0.7621,
+      "step": 2546
+    },
+    {
+      "epoch": 1.2636216954201316,
+      "grad_norm": 0.14129831083302125,
+      "learning_rate": 3.876100309601547e-06,
+      "loss": 0.8199,
+      "step": 2547
+    },
+    {
+      "epoch": 1.264118158123371,
+      "grad_norm": 0.14102270939055808,
+      "learning_rate": 3.87528424004039e-06,
+      "loss": 0.7137,
+      "step": 2548
+    },
+    {
+      "epoch": 1.2646146208266105,
+      "grad_norm": 0.13270563152474793,
+      "learning_rate": 3.874467960289088e-06,
+      "loss": 0.6974,
+      "step": 2549
+    },
+    {
+      "epoch": 1.2651110835298498,
+      "grad_norm": 0.14077175671585102,
+      "learning_rate": 3.8736514704723956e-06,
+      "loss": 0.7488,
+      "step": 2550
+    },
+    {
+      "epoch": 1.2656075462330891,
+      "grad_norm": 0.14294951439526327,
+      "learning_rate": 3.872834770715102e-06,
+      "loss": 0.7407,
+      "step": 2551
+    },
+    {
+      "epoch": 1.2661040089363287,
+      "grad_norm": 0.14108160173034903,
+      "learning_rate": 3.872017861142024e-06,
+      "loss": 0.7285,
+      "step": 2552
+    },
+    {
+      "epoch": 1.266600471639568,
+      "grad_norm": 0.14172910962233856,
+      "learning_rate": 3.871200741878015e-06,
+      "loss": 0.7434,
+      "step": 2553
+    },
+    {
+      "epoch": 1.2670969343428076,
+      "grad_norm": 0.1418472981583555,
+      "learning_rate": 3.870383413047959e-06,
+      "loss": 0.693,
+      "step": 2554
+    },
+    {
+      "epoch": 1.267593397046047,
+      "grad_norm": 0.14824626937177632,
+      "learning_rate": 3.86956587477677e-06,
+      "loss": 0.7304,
+      "step": 2555
+    },
+    {
+      "epoch": 1.2680898597492862,
+      "grad_norm": 0.14417792095901208,
+      "learning_rate": 3.868748127189397e-06,
+      "loss": 0.689,
+      "step": 2556
+    },
+    {
+      "epoch": 1.2685863224525258,
+      "grad_norm": 0.13628181036537815,
+      "learning_rate": 3.8679301704108176e-06,
+      "loss": 0.7269,
+      "step": 2557
+    },
+    {
+      "epoch": 1.2690827851557651,
+      "grad_norm": 0.1432671252492769,
+      "learning_rate": 3.8671120045660456e-06,
+      "loss": 0.7273,
+      "step": 2558
+    },
+    {
+      "epoch": 1.2695792478590047,
+      "grad_norm": 0.13729463835110303,
+      "learning_rate": 3.8662936297801235e-06,
+      "loss": 0.7118,
+      "step": 2559
+    },
+    {
+      "epoch": 1.270075710562244,
+      "grad_norm": 0.1458195417043056,
+      "learning_rate": 3.865475046178127e-06,
+      "loss": 0.7171,
+      "step": 2560
+    },
+    {
+      "epoch": 1.2705721732654833,
+      "grad_norm": 0.13880022855949167,
+      "learning_rate": 3.864656253885163e-06,
+      "loss": 0.6906,
+      "step": 2561
+    },
+    {
+      "epoch": 1.2710686359687229,
+      "grad_norm": 0.1349359231670376,
+      "learning_rate": 3.863837253026372e-06,
+      "loss": 0.7689,
+      "step": 2562
+    },
+    {
+      "epoch": 1.2715650986719622,
+      "grad_norm": 0.14164412974060797,
+      "learning_rate": 3.863018043726924e-06,
+      "loss": 0.6878,
+      "step": 2563
+    },
+    {
+      "epoch": 1.2720615613752018,
+      "grad_norm": 0.14323827765065936,
+      "learning_rate": 3.862198626112023e-06,
+      "loss": 0.7606,
+      "step": 2564
+    },
+    {
+      "epoch": 1.272558024078441,
+      "grad_norm": 0.1473424236959722,
+      "learning_rate": 3.861379000306902e-06,
+      "loss": 0.7605,
+      "step": 2565
+    },
+    {
+      "epoch": 1.2730544867816804,
+      "grad_norm": 0.1292191998705604,
+      "learning_rate": 3.8605591664368295e-06,
+      "loss": 0.7022,
+      "step": 2566
+    },
+    {
+      "epoch": 1.27355094948492,
+      "grad_norm": 0.13465698019970399,
+      "learning_rate": 3.859739124627103e-06,
+      "loss": 0.6971,
+      "step": 2567
+    },
+    {
+      "epoch": 1.2740474121881593,
+      "grad_norm": 0.13654253074049172,
+      "learning_rate": 3.858918875003053e-06,
+      "loss": 0.6928,
+      "step": 2568
+    },
+    {
+      "epoch": 1.2745438748913989,
+      "grad_norm": 0.1379719697081152,
+      "learning_rate": 3.858098417690042e-06,
+      "loss": 0.717,
+      "step": 2569
+    },
+    {
+      "epoch": 1.2750403375946382,
+      "grad_norm": 0.14175199937004399,
+      "learning_rate": 3.857277752813463e-06,
+      "loss": 0.7616,
+      "step": 2570
+    },
+    {
+      "epoch": 1.2755368002978775,
+      "grad_norm": 0.13443150257007153,
+      "learning_rate": 3.856456880498742e-06,
+      "loss": 0.6973,
+      "step": 2571
+    },
+    {
+      "epoch": 1.276033263001117,
+      "grad_norm": 0.14593077908065585,
+      "learning_rate": 3.855635800871335e-06,
+      "loss": 0.7569,
+      "step": 2572
+    },
+    {
+      "epoch": 1.2765297257043564,
+      "grad_norm": 0.15100942741983306,
+      "learning_rate": 3.854814514056734e-06,
+      "loss": 0.7642,
+      "step": 2573
+    },
+    {
+      "epoch": 1.277026188407596,
+      "grad_norm": 0.13642851129916278,
+      "learning_rate": 3.853993020180456e-06,
+      "loss": 0.7406,
+      "step": 2574
+    },
+    {
+      "epoch": 1.2775226511108353,
+      "grad_norm": 0.1400663992803376,
+      "learning_rate": 3.853171319368054e-06,
+      "loss": 0.7454,
+      "step": 2575
+    },
+    {
+      "epoch": 1.2780191138140746,
+      "grad_norm": 0.1319083373021114,
+      "learning_rate": 3.852349411745113e-06,
+      "loss": 0.74,
+      "step": 2576
+    },
+    {
+      "epoch": 1.2785155765173142,
+      "grad_norm": 0.14724655880988557,
+      "learning_rate": 3.851527297437247e-06,
+      "loss": 0.7528,
+      "step": 2577
+    },
+    {
+      "epoch": 1.2790120392205535,
+      "grad_norm": 0.13674211565816105,
+      "learning_rate": 3.8507049765701045e-06,
+      "loss": 0.6716,
+      "step": 2578
+    },
+    {
+      "epoch": 1.279508501923793,
+      "grad_norm": 0.1382868822761622,
+      "learning_rate": 3.849882449269363e-06,
+      "loss": 0.746,
+      "step": 2579
+    },
+    {
+      "epoch": 1.2800049646270324,
+      "grad_norm": 0.1390259790054548,
+      "learning_rate": 3.849059715660732e-06,
+      "loss": 0.7088,
+      "step": 2580
+    },
+    {
+      "epoch": 1.2805014273302717,
+      "grad_norm": 0.1369367923225582,
+      "learning_rate": 3.848236775869955e-06,
+      "loss": 0.7699,
+      "step": 2581
+    },
+    {
+      "epoch": 1.2809978900335113,
+      "grad_norm": 0.15971172103749032,
+      "learning_rate": 3.847413630022804e-06,
+      "loss": 0.746,
+      "step": 2582
+    },
+    {
+      "epoch": 1.2814943527367506,
+      "grad_norm": 0.14277522047325522,
+      "learning_rate": 3.846590278245083e-06,
+      "loss": 0.7567,
+      "step": 2583
+    },
+    {
+      "epoch": 1.2819908154399902,
+      "grad_norm": 0.14699081523439286,
+      "learning_rate": 3.8457667206626306e-06,
+      "loss": 0.7873,
+      "step": 2584
+    },
+    {
+      "epoch": 1.2824872781432295,
+      "grad_norm": 0.14366187562768565,
+      "learning_rate": 3.844942957401311e-06,
+      "loss": 0.6949,
+      "step": 2585
+    },
+    {
+      "epoch": 1.2829837408464688,
+      "grad_norm": 0.14181909585351757,
+      "learning_rate": 3.844118988587025e-06,
+      "loss": 0.7135,
+      "step": 2586
+    },
+    {
+      "epoch": 1.2834802035497084,
+      "grad_norm": 0.13109014026406526,
+      "learning_rate": 3.843294814345705e-06,
+      "loss": 0.6893,
+      "step": 2587
+    },
+    {
+      "epoch": 1.2839766662529477,
+      "grad_norm": 0.13263413654791315,
+      "learning_rate": 3.8424704348033084e-06,
+      "loss": 0.6963,
+      "step": 2588
+    },
+    {
+      "epoch": 1.2844731289561873,
+      "grad_norm": 0.13552375577647977,
+      "learning_rate": 3.841645850085831e-06,
+      "loss": 0.718,
+      "step": 2589
+    },
+    {
+      "epoch": 1.2849695916594266,
+      "grad_norm": 0.138899051483369,
+      "learning_rate": 3.840821060319298e-06,
+      "loss": 0.754,
+      "step": 2590
+    },
+    {
+      "epoch": 1.285466054362666,
+      "grad_norm": 0.14600374894252893,
+      "learning_rate": 3.839996065629764e-06,
+      "loss": 0.7421,
+      "step": 2591
+    },
+    {
+      "epoch": 1.2859625170659055,
+      "grad_norm": 0.14495501404251815,
+      "learning_rate": 3.839170866143317e-06,
+      "loss": 0.6984,
+      "step": 2592
+    },
+    {
+      "epoch": 1.2864589797691448,
+      "grad_norm": 0.13341183715508106,
+      "learning_rate": 3.838345461986074e-06,
+      "loss": 0.7205,
+      "step": 2593
+    },
+    {
+      "epoch": 1.2869554424723844,
+      "grad_norm": 0.13169443537805922,
+      "learning_rate": 3.837519853284186e-06,
+      "loss": 0.691,
+      "step": 2594
+    },
+    {
+      "epoch": 1.2874519051756237,
+      "grad_norm": 0.13790138352749554,
+      "learning_rate": 3.836694040163834e-06,
+      "loss": 0.6926,
+      "step": 2595
+    },
+    {
+      "epoch": 1.287948367878863,
+      "grad_norm": 0.1437529575160023,
+      "learning_rate": 3.835868022751231e-06,
+      "loss": 0.7934,
+      "step": 2596
+    },
+    {
+      "epoch": 1.2884448305821026,
+      "grad_norm": 0.13306894590543736,
+      "learning_rate": 3.835041801172619e-06,
+      "loss": 0.6854,
+      "step": 2597
+    },
+    {
+      "epoch": 1.288941293285342,
+      "grad_norm": 0.14395437206135459,
+      "learning_rate": 3.834215375554275e-06,
+      "loss": 0.7379,
+      "step": 2598
+    },
+    {
+      "epoch": 1.2894377559885815,
+      "grad_norm": 0.1378707356978414,
+      "learning_rate": 3.8333887460225015e-06,
+      "loss": 0.7178,
+      "step": 2599
+    },
+    {
+      "epoch": 1.2899342186918208,
+      "grad_norm": 0.13148423330705314,
+      "learning_rate": 3.832561912703638e-06,
+      "loss": 0.6924,
+      "step": 2600
+    },
+    {
+      "epoch": 1.2904306813950601,
+      "grad_norm": 0.13730143809759457,
+      "learning_rate": 3.831734875724052e-06,
+      "loss": 0.7503,
+      "step": 2601
+    },
+    {
+      "epoch": 1.2909271440982997,
+      "grad_norm": 0.13498579164231725,
+      "learning_rate": 3.830907635210143e-06,
+      "loss": 0.738,
+      "step": 2602
+    },
+    {
+      "epoch": 1.291423606801539,
+      "grad_norm": 0.13053407240787798,
+      "learning_rate": 3.830080191288342e-06,
+      "loss": 0.7039,
+      "step": 2603
+    },
+    {
+      "epoch": 1.2919200695047786,
+      "grad_norm": 0.13619974831779605,
+      "learning_rate": 3.82925254408511e-06,
+      "loss": 0.7462,
+      "step": 2604
+    },
+    {
+      "epoch": 1.2924165322080179,
+      "grad_norm": 0.1415670740958942,
+      "learning_rate": 3.828424693726939e-06,
+      "loss": 0.7302,
+      "step": 2605
+    },
+    {
+      "epoch": 1.2929129949112572,
+      "grad_norm": 0.14173779929791688,
+      "learning_rate": 3.827596640340353e-06,
+      "loss": 0.7732,
+      "step": 2606
+    },
+    {
+      "epoch": 1.2934094576144968,
+      "grad_norm": 0.1359458995344357,
+      "learning_rate": 3.826768384051907e-06,
+      "loss": 0.7372,
+      "step": 2607
+    },
+    {
+      "epoch": 1.293905920317736,
+      "grad_norm": 0.13416195193113473,
+      "learning_rate": 3.825939924988187e-06,
+      "loss": 0.7289,
+      "step": 2608
+    },
+    {
+      "epoch": 1.2944023830209757,
+      "grad_norm": 0.1337826798946644,
+      "learning_rate": 3.825111263275809e-06,
+      "loss": 0.6944,
+      "step": 2609
+    },
+    {
+      "epoch": 1.294898845724215,
+      "grad_norm": 0.13167566643050344,
+      "learning_rate": 3.824282399041421e-06,
+      "loss": 0.677,
+      "step": 2610
+    },
+    {
+      "epoch": 1.2953953084274543,
+      "grad_norm": 0.1432765949213179,
+      "learning_rate": 3.823453332411702e-06,
+      "loss": 0.7226,
+      "step": 2611
+    },
+    {
+      "epoch": 1.2958917711306939,
+      "grad_norm": 0.13409930276110446,
+      "learning_rate": 3.8226240635133615e-06,
+      "loss": 0.7633,
+      "step": 2612
+    },
+    {
+      "epoch": 1.2963882338339332,
+      "grad_norm": 0.14108249887977412,
+      "learning_rate": 3.8217945924731385e-06,
+      "loss": 0.7139,
+      "step": 2613
+    },
+    {
+      "epoch": 1.2968846965371728,
+      "grad_norm": 0.13605604217842482,
+      "learning_rate": 3.8209649194178065e-06,
+      "loss": 0.7486,
+      "step": 2614
+    },
+    {
+      "epoch": 1.297381159240412,
+      "grad_norm": 0.13562858110835124,
+      "learning_rate": 3.820135044474166e-06,
+      "loss": 0.7403,
+      "step": 2615
+    },
+    {
+      "epoch": 1.2978776219436514,
+      "grad_norm": 0.13982122188113263,
+      "learning_rate": 3.81930496776905e-06,
+      "loss": 0.7424,
+      "step": 2616
+    },
+    {
+      "epoch": 1.298374084646891,
+      "grad_norm": 0.13772514732208074,
+      "learning_rate": 3.818474689429324e-06,
+      "loss": 0.7531,
+      "step": 2617
+    },
+    {
+      "epoch": 1.2988705473501303,
+      "grad_norm": 0.13901119411689639,
+      "learning_rate": 3.817644209581881e-06,
+      "loss": 0.7395,
+      "step": 2618
+    },
+    {
+      "epoch": 1.2993670100533699,
+      "grad_norm": 0.1344248019109414,
+      "learning_rate": 3.8168135283536485e-06,
+      "loss": 0.732,
+      "step": 2619
+    },
+    {
+      "epoch": 1.2998634727566092,
+      "grad_norm": 0.13540832289022223,
+      "learning_rate": 3.815982645871582e-06,
+      "loss": 0.7063,
+      "step": 2620
+    },
+    {
+      "epoch": 1.3003599354598485,
+      "grad_norm": 0.13397167892750497,
+      "learning_rate": 3.815151562262666e-06,
+      "loss": 0.7302,
+      "step": 2621
+    },
+    {
+      "epoch": 1.300856398163088,
+      "grad_norm": 0.13514380521337327,
+      "learning_rate": 3.8143202776539224e-06,
+      "loss": 0.7267,
+      "step": 2622
+    },
+    {
+      "epoch": 1.3013528608663274,
+      "grad_norm": 0.1334082478010419,
+      "learning_rate": 3.8134887921723975e-06,
+      "loss": 0.7107,
+      "step": 2623
+    },
+    {
+      "epoch": 1.301849323569567,
+      "grad_norm": 0.1359740311554934,
+      "learning_rate": 3.812657105945171e-06,
+      "loss": 0.6509,
+      "step": 2624
+    },
+    {
+      "epoch": 1.3023457862728063,
+      "grad_norm": 0.1445356509888893,
+      "learning_rate": 3.8118252190993533e-06,
+      "loss": 0.7013,
+      "step": 2625
+    },
+    {
+      "epoch": 1.3028422489760456,
+      "grad_norm": 0.13788799234754298,
+      "learning_rate": 3.810993131762083e-06,
+      "loss": 0.7414,
+      "step": 2626
+    },
+    {
+      "epoch": 1.3033387116792852,
+      "grad_norm": 0.13253611567442633,
+      "learning_rate": 3.810160844060533e-06,
+      "loss": 0.7793,
+      "step": 2627
+    },
+    {
+      "epoch": 1.3038351743825245,
+      "grad_norm": 0.133375725717547,
+      "learning_rate": 3.8093283561219063e-06,
+      "loss": 0.7065,
+      "step": 2628
+    },
+    {
+      "epoch": 1.304331637085764,
+      "grad_norm": 0.14059616928874114,
+      "learning_rate": 3.808495668073432e-06,
+      "loss": 0.7164,
+      "step": 2629
+    },
+    {
+      "epoch": 1.3048280997890034,
+      "grad_norm": 0.1307516991478391,
+      "learning_rate": 3.807662780042376e-06,
+      "loss": 0.6775,
+      "step": 2630
+    },
+    {
+      "epoch": 1.3053245624922427,
+      "grad_norm": 0.14152325120856557,
+      "learning_rate": 3.806829692156031e-06,
+      "loss": 0.7759,
+      "step": 2631
+    },
+    {
+      "epoch": 1.3058210251954823,
+      "grad_norm": 0.12924559643345673,
+      "learning_rate": 3.8059964045417196e-06,
+      "loss": 0.6981,
+      "step": 2632
+    },
+    {
+      "epoch": 1.3063174878987216,
+      "grad_norm": 0.1370915470558916,
+      "learning_rate": 3.805162917326799e-06,
+      "loss": 0.7678,
+      "step": 2633
+    },
+    {
+      "epoch": 1.3068139506019611,
+      "grad_norm": 0.14767292411211091,
+      "learning_rate": 3.8043292306386528e-06,
+      "loss": 0.6807,
+      "step": 2634
+    },
+    {
+      "epoch": 1.3073104133052005,
+      "grad_norm": 0.13677942779385033,
+      "learning_rate": 3.8034953446046974e-06,
+      "loss": 0.7533,
+      "step": 2635
+    },
+    {
+      "epoch": 1.3078068760084398,
+      "grad_norm": 0.13632882033348248,
+      "learning_rate": 3.8026612593523795e-06,
+      "loss": 0.7246,
+      "step": 2636
+    },
+    {
+      "epoch": 1.3083033387116794,
+      "grad_norm": 0.13657887532594618,
+      "learning_rate": 3.801826975009173e-06,
+      "loss": 0.7122,
+      "step": 2637
+    },
+    {
+      "epoch": 1.3087998014149187,
+      "grad_norm": 0.13144574952305246,
+      "learning_rate": 3.8009924917025864e-06,
+      "loss": 0.727,
+      "step": 2638
+    },
+    {
+      "epoch": 1.3092962641181582,
+      "grad_norm": 0.13554336516404342,
+      "learning_rate": 3.8001578095601583e-06,
+      "loss": 0.6958,
+      "step": 2639
+    },
+    {
+      "epoch": 1.3097927268213976,
+      "grad_norm": 0.13859859321388554,
+      "learning_rate": 3.799322928709455e-06,
+      "loss": 0.7051,
+      "step": 2640
+    },
+    {
+      "epoch": 1.310289189524637,
+      "grad_norm": 0.1389305915074386,
+      "learning_rate": 3.7984878492780754e-06,
+      "loss": 0.7113,
+      "step": 2641
+    },
+    {
+      "epoch": 1.3107856522278765,
+      "grad_norm": 0.13940935351453346,
+      "learning_rate": 3.797652571393647e-06,
+      "loss": 0.7732,
+      "step": 2642
+    },
+    {
+      "epoch": 1.3112821149311158,
+      "grad_norm": 0.13336829577684228,
+      "learning_rate": 3.79681709518383e-06,
+      "loss": 0.7421,
+      "step": 2643
+    },
+    {
+      "epoch": 1.3117785776343553,
+      "grad_norm": 0.1338637968337137,
+      "learning_rate": 3.7959814207763134e-06,
+      "loss": 0.7025,
+      "step": 2644
+    },
+    {
+      "epoch": 1.3122750403375947,
+      "grad_norm": 0.1322361693764335,
+      "learning_rate": 3.7951455482988154e-06,
+      "loss": 0.6567,
+      "step": 2645
+    },
+    {
+      "epoch": 1.312771503040834,
+      "grad_norm": 0.13690317157186413,
+      "learning_rate": 3.7943094778790866e-06,
+      "loss": 0.7347,
+      "step": 2646
+    },
+    {
+      "epoch": 1.3132679657440733,
+      "grad_norm": 0.13886094412827465,
+      "learning_rate": 3.7934732096449066e-06,
+      "loss": 0.7281,
+      "step": 2647
+    },
+    {
+      "epoch": 1.313764428447313,
+      "grad_norm": 0.1433387530701494,
+      "learning_rate": 3.792636743724085e-06,
+      "loss": 0.7451,
+      "step": 2648
+    },
+    {
+      "epoch": 1.3142608911505524,
+      "grad_norm": 0.13856198601421107,
+      "learning_rate": 3.7918000802444644e-06,
+      "loss": 0.7467,
+      "step": 2649
+    },
+    {
+      "epoch": 1.3147573538537918,
+      "grad_norm": 0.13420789080751208,
+      "learning_rate": 3.790963219333913e-06,
+      "loss": 0.7276,
+      "step": 2650
+    },
+    {
+      "epoch": 1.315253816557031,
+      "grad_norm": 0.1291136871280018,
+      "learning_rate": 3.790126161120333e-06,
+      "loss": 0.6971,
+      "step": 2651
+    },
+    {
+      "epoch": 1.3157502792602704,
+      "grad_norm": 0.1290712942967362,
+      "learning_rate": 3.789288905731655e-06,
+      "loss": 0.7456,
+      "step": 2652
+    },
+    {
+      "epoch": 1.31624674196351,
+      "grad_norm": 0.13062887909116458,
+      "learning_rate": 3.78845145329584e-06,
+      "loss": 0.7069,
+      "step": 2653
+    },
+    {
+      "epoch": 1.3167432046667495,
+      "grad_norm": 0.13036053551686685,
+      "learning_rate": 3.7876138039408784e-06,
+      "loss": 0.7227,
+      "step": 2654
+    },
+    {
+      "epoch": 1.3172396673699889,
+      "grad_norm": 0.13222431704901577,
+      "learning_rate": 3.786775957794793e-06,
+      "loss": 0.7432,
+      "step": 2655
+    },
+    {
+      "epoch": 1.3177361300732282,
+      "grad_norm": 0.13104899861223418,
+      "learning_rate": 3.7859379149856335e-06,
+      "loss": 0.7501,
+      "step": 2656
+    },
+    {
+      "epoch": 1.3182325927764675,
+      "grad_norm": 0.13397270135551584,
+      "learning_rate": 3.7850996756414832e-06,
+      "loss": 0.7326,
+      "step": 2657
+    },
+    {
+      "epoch": 1.318729055479707,
+      "grad_norm": 0.13613150488476347,
+      "learning_rate": 3.7842612398904515e-06,
+      "loss": 0.7147,
+      "step": 2658
+    },
+    {
+      "epoch": 1.3192255181829466,
+      "grad_norm": 0.14339490468621394,
+      "learning_rate": 3.783422607860681e-06,
+      "loss": 0.7324,
+      "step": 2659
+    },
+    {
+      "epoch": 1.319721980886186,
+      "grad_norm": 0.13378917893428685,
+      "learning_rate": 3.7825837796803438e-06,
+      "loss": 0.7268,
+      "step": 2660
+    },
+    {
+      "epoch": 1.3202184435894253,
+      "grad_norm": 0.13360282147099303,
+      "learning_rate": 3.7817447554776397e-06,
+      "loss": 0.7304,
+      "step": 2661
+    },
+    {
+      "epoch": 1.3207149062926646,
+      "grad_norm": 0.13904280037330674,
+      "learning_rate": 3.780905535380801e-06,
+      "loss": 0.7572,
+      "step": 2662
+    },
+    {
+      "epoch": 1.3212113689959042,
+      "grad_norm": 0.13011808148888576,
+      "learning_rate": 3.78006611951809e-06,
+      "loss": 0.6836,
+      "step": 2663
+    },
+    {
+      "epoch": 1.3217078316991435,
+      "grad_norm": 0.13510584692748015,
+      "learning_rate": 3.779226508017796e-06,
+      "loss": 0.7061,
+      "step": 2664
+    },
+    {
+      "epoch": 1.322204294402383,
+      "grad_norm": 0.1371054013890361,
+      "learning_rate": 3.778386701008241e-06,
+      "loss": 0.751,
+      "step": 2665
+    },
+    {
+      "epoch": 1.3227007571056224,
+      "grad_norm": 0.13439056933714283,
+      "learning_rate": 3.7775466986177763e-06,
+      "loss": 0.7382,
+      "step": 2666
+    },
+    {
+      "epoch": 1.3231972198088617,
+      "grad_norm": 0.13676259742243488,
+      "learning_rate": 3.776706500974783e-06,
+      "loss": 0.7439,
+      "step": 2667
+    },
+    {
+      "epoch": 1.3236936825121013,
+      "grad_norm": 0.1384039615266504,
+      "learning_rate": 3.775866108207671e-06,
+      "loss": 0.7215,
+      "step": 2668
+    },
+    {
+      "epoch": 1.3241901452153406,
+      "grad_norm": 0.14031483696025457,
+      "learning_rate": 3.7750255204448817e-06,
+      "loss": 0.7596,
+      "step": 2669
+    },
+    {
+      "epoch": 1.3246866079185802,
+      "grad_norm": 0.13734515606529832,
+      "learning_rate": 3.7741847378148845e-06,
+      "loss": 0.7511,
+      "step": 2670
+    },
+    {
+      "epoch": 1.3251830706218195,
+      "grad_norm": 0.130610747808502,
+      "learning_rate": 3.7733437604461804e-06,
+      "loss": 0.675,
+      "step": 2671
+    },
+    {
+      "epoch": 1.3256795333250588,
+      "grad_norm": 0.13229432283223813,
+      "learning_rate": 3.7725025884672987e-06,
+      "loss": 0.7204,
+      "step": 2672
+    },
+    {
+      "epoch": 1.3261759960282984,
+      "grad_norm": 0.13459703835228942,
+      "learning_rate": 3.7716612220068004e-06,
+      "loss": 0.6899,
+      "step": 2673
+    },
+    {
+      "epoch": 1.3266724587315377,
+      "grad_norm": 0.13416054723437704,
+      "learning_rate": 3.770819661193273e-06,
+      "loss": 0.6853,
+      "step": 2674
+    },
+    {
+      "epoch": 1.3271689214347773,
+      "grad_norm": 0.1361667362155641,
+      "learning_rate": 3.7699779061553365e-06,
+      "loss": 0.7009,
+      "step": 2675
+    },
+    {
+      "epoch": 1.3276653841380166,
+      "grad_norm": 0.13873053615880002,
+      "learning_rate": 3.7691359570216404e-06,
+      "loss": 0.6775,
+      "step": 2676
+    },
+    {
+      "epoch": 1.328161846841256,
+      "grad_norm": 0.13977475275303883,
+      "learning_rate": 3.7682938139208615e-06,
+      "loss": 0.7638,
+      "step": 2677
+    },
+    {
+      "epoch": 1.3286583095444955,
+      "grad_norm": 0.13414664404632431,
+      "learning_rate": 3.7674514769817083e-06,
+      "loss": 0.7539,
+      "step": 2678
+    },
+    {
+      "epoch": 1.3291547722477348,
+      "grad_norm": 0.14091362715741365,
+      "learning_rate": 3.7666089463329196e-06,
+      "loss": 0.746,
+      "step": 2679
+    },
+    {
+      "epoch": 1.3296512349509744,
+      "grad_norm": 0.1470455702766219,
+      "learning_rate": 3.765766222103262e-06,
+      "loss": 0.7381,
+      "step": 2680
+    },
+    {
+      "epoch": 1.3301476976542137,
+      "grad_norm": 0.13534175218735514,
+      "learning_rate": 3.7649233044215314e-06,
+      "loss": 0.7117,
+      "step": 2681
+    },
+    {
+      "epoch": 1.330644160357453,
+      "grad_norm": 0.13478773220973986,
+      "learning_rate": 3.764080193416556e-06,
+      "loss": 0.7127,
+      "step": 2682
+    },
+    {
+      "epoch": 1.3311406230606926,
+      "grad_norm": 0.14594060460245392,
+      "learning_rate": 3.7632368892171916e-06,
+      "loss": 0.7218,
+      "step": 2683
+    },
+    {
+      "epoch": 1.331637085763932,
+      "grad_norm": 0.13227694228798864,
+      "learning_rate": 3.7623933919523226e-06,
+      "loss": 0.6883,
+      "step": 2684
+    },
+    {
+      "epoch": 1.3321335484671715,
+      "grad_norm": 0.1331372361004289,
+      "learning_rate": 3.761549701750865e-06,
+      "loss": 0.7251,
+      "step": 2685
+    },
+    {
+      "epoch": 1.3326300111704108,
+      "grad_norm": 0.13633213773374722,
+      "learning_rate": 3.7607058187417624e-06,
+      "loss": 0.6906,
+      "step": 2686
+    },
+    {
+      "epoch": 1.3331264738736501,
+      "grad_norm": 0.14075233741360202,
+      "learning_rate": 3.7598617430539886e-06,
+      "loss": 0.7518,
+      "step": 2687
+    },
+    {
+      "epoch": 1.3336229365768897,
+      "grad_norm": 0.14126613421524764,
+      "learning_rate": 3.7590174748165487e-06,
+      "loss": 0.7323,
+      "step": 2688
+    },
+    {
+      "epoch": 1.334119399280129,
+      "grad_norm": 0.13817784028859825,
+      "learning_rate": 3.758173014158475e-06,
+      "loss": 0.7186,
+      "step": 2689
+    },
+    {
+      "epoch": 1.3346158619833686,
+      "grad_norm": 0.14776049205061048,
+      "learning_rate": 3.757328361208828e-06,
+      "loss": 0.7606,
+      "step": 2690
+    },
+    {
+      "epoch": 1.335112324686608,
+      "grad_norm": 0.1335821510798814,
+      "learning_rate": 3.756483516096702e-06,
+      "loss": 0.7053,
+      "step": 2691
+    },
+    {
+      "epoch": 1.3356087873898472,
+      "grad_norm": 0.1383320848994581,
+      "learning_rate": 3.7556384789512156e-06,
+      "loss": 0.7501,
+      "step": 2692
+    },
+    {
+      "epoch": 1.3361052500930868,
+      "grad_norm": 0.14093892215142187,
+      "learning_rate": 3.754793249901521e-06,
+      "loss": 0.7381,
+      "step": 2693
+    },
+    {
+      "epoch": 1.3366017127963261,
+      "grad_norm": 0.13669736220023862,
+      "learning_rate": 3.753947829076797e-06,
+      "loss": 0.6916,
+      "step": 2694
+    },
+    {
+      "epoch": 1.3370981754995657,
+      "grad_norm": 0.13362283897482466,
+      "learning_rate": 3.7531022166062538e-06,
+      "loss": 0.7439,
+      "step": 2695
+    },
+    {
+      "epoch": 1.337594638202805,
+      "grad_norm": 0.13205258206921125,
+      "learning_rate": 3.7522564126191276e-06,
+      "loss": 0.6928,
+      "step": 2696
+    },
+    {
+      "epoch": 1.3380911009060443,
+      "grad_norm": 0.13796484578752577,
+      "learning_rate": 3.751410417244687e-06,
+      "loss": 0.8034,
+      "step": 2697
+    },
+    {
+      "epoch": 1.3385875636092839,
+      "grad_norm": 0.13414239398576636,
+      "learning_rate": 3.7505642306122293e-06,
+      "loss": 0.7181,
+      "step": 2698
+    },
+    {
+      "epoch": 1.3390840263125232,
+      "grad_norm": 0.14145403843670953,
+      "learning_rate": 3.7497178528510803e-06,
+      "loss": 0.691,
+      "step": 2699
+    },
+    {
+      "epoch": 1.3395804890157628,
+      "grad_norm": 0.1379866463002123,
+      "learning_rate": 3.7488712840905955e-06,
+      "loss": 0.7399,
+      "step": 2700
+    },
+    {
+      "epoch": 1.340076951719002,
+      "grad_norm": 0.13204289710931297,
+      "learning_rate": 3.7480245244601587e-06,
+      "loss": 0.7253,
+      "step": 2701
+    },
+    {
+      "epoch": 1.3405734144222414,
+      "grad_norm": 0.1397985609990718,
+      "learning_rate": 3.747177574089184e-06,
+      "loss": 0.7438,
+      "step": 2702
+    },
+    {
+      "epoch": 1.341069877125481,
+      "grad_norm": 0.14161488684867624,
+      "learning_rate": 3.746330433107114e-06,
+      "loss": 0.7537,
+      "step": 2703
+    },
+    {
+      "epoch": 1.3415663398287203,
+      "grad_norm": 0.14437747804639947,
+      "learning_rate": 3.7454831016434206e-06,
+      "loss": 0.7241,
+      "step": 2704
+    },
+    {
+      "epoch": 1.3420628025319599,
+      "grad_norm": 0.13861952176789427,
+      "learning_rate": 3.744635579827606e-06,
+      "loss": 0.7398,
+      "step": 2705
+    },
+    {
+      "epoch": 1.3425592652351992,
+      "grad_norm": 0.1399777307376627,
+      "learning_rate": 3.7437878677891977e-06,
+      "loss": 0.7055,
+      "step": 2706
+    },
+    {
+      "epoch": 1.3430557279384385,
+      "grad_norm": 0.1370426735312947,
+      "learning_rate": 3.742939965657757e-06,
+      "loss": 0.7354,
+      "step": 2707
+    },
+    {
+      "epoch": 1.343552190641678,
+      "grad_norm": 0.1457622692943864,
+      "learning_rate": 3.7420918735628714e-06,
+      "loss": 0.7592,
+      "step": 2708
+    },
+    {
+      "epoch": 1.3440486533449174,
+      "grad_norm": 0.1301536298141118,
+      "learning_rate": 3.741243591634159e-06,
+      "loss": 0.7435,
+      "step": 2709
+    },
+    {
+      "epoch": 1.344545116048157,
+      "grad_norm": 0.13359769838918073,
+      "learning_rate": 3.7403951200012645e-06,
+      "loss": 0.7271,
+      "step": 2710
+    },
+    {
+      "epoch": 1.3450415787513963,
+      "grad_norm": 0.1361918708494249,
+      "learning_rate": 3.7395464587938652e-06,
+      "loss": 0.7421,
+      "step": 2711
+    },
+    {
+      "epoch": 1.3455380414546356,
+      "grad_norm": 0.13334017361026135,
+      "learning_rate": 3.738697608141664e-06,
+      "loss": 0.6886,
+      "step": 2712
+    },
+    {
+      "epoch": 1.3460345041578752,
+      "grad_norm": 0.1387171887689253,
+      "learning_rate": 3.7378485681743935e-06,
+      "loss": 0.6873,
+      "step": 2713
+    },
+    {
+      "epoch": 1.3465309668611145,
+      "grad_norm": 0.14983174029057572,
+      "learning_rate": 3.7369993390218172e-06,
+      "loss": 0.7513,
+      "step": 2714
+    },
+    {
+      "epoch": 1.347027429564354,
+      "grad_norm": 0.1424875756746005,
+      "learning_rate": 3.736149920813726e-06,
+      "loss": 0.6993,
+      "step": 2715
+    },
+    {
+      "epoch": 1.3475238922675934,
+      "grad_norm": 0.13814327332815962,
+      "learning_rate": 3.7353003136799394e-06,
+      "loss": 0.7383,
+      "step": 2716
+    },
+    {
+      "epoch": 1.3480203549708327,
+      "grad_norm": 0.13291744742349632,
+      "learning_rate": 3.7344505177503064e-06,
+      "loss": 0.6902,
+      "step": 2717
+    },
+    {
+      "epoch": 1.3485168176740723,
+      "grad_norm": 0.1294966956204128,
+      "learning_rate": 3.733600533154705e-06,
+      "loss": 0.6753,
+      "step": 2718
+    },
+    {
+      "epoch": 1.3490132803773116,
+      "grad_norm": 0.13384401858376338,
+      "learning_rate": 3.732750360023041e-06,
+      "loss": 0.7194,
+      "step": 2719
+    },
+    {
+      "epoch": 1.3495097430805512,
+      "grad_norm": 0.1510726665603985,
+      "learning_rate": 3.73189999848525e-06,
+      "loss": 0.7286,
+      "step": 2720
+    },
+    {
+      "epoch": 1.3500062057837905,
+      "grad_norm": 0.17389508520864652,
+      "learning_rate": 3.7310494486712966e-06,
+      "loss": 0.7334,
+      "step": 2721
+    },
+    {
+      "epoch": 1.3505026684870298,
+      "grad_norm": 0.14457583333245747,
+      "learning_rate": 3.730198710711173e-06,
+      "loss": 0.7759,
+      "step": 2722
+    },
+    {
+      "epoch": 1.3509991311902694,
+      "grad_norm": 0.1304606464207145,
+      "learning_rate": 3.729347784734901e-06,
+      "loss": 0.7004,
+      "step": 2723
+    },
+    {
+      "epoch": 1.3514955938935087,
+      "grad_norm": 0.13664237040256322,
+      "learning_rate": 3.7284966708725316e-06,
+      "loss": 0.7148,
+      "step": 2724
+    },
+    {
+      "epoch": 1.3519920565967483,
+      "grad_norm": 0.1319353306336256,
+      "learning_rate": 3.727645369254144e-06,
+      "loss": 0.7099,
+      "step": 2725
+    },
+    {
+      "epoch": 1.3524885192999876,
+      "grad_norm": 0.13747258866291462,
+      "learning_rate": 3.7267938800098454e-06,
+      "loss": 0.7371,
+      "step": 2726
+    },
+    {
+      "epoch": 1.352984982003227,
+      "grad_norm": 0.13461034808749692,
+      "learning_rate": 3.7259422032697724e-06,
+      "loss": 0.7092,
+      "step": 2727
+    },
+    {
+      "epoch": 1.3534814447064665,
+      "grad_norm": 0.13212884182845225,
+      "learning_rate": 3.7250903391640893e-06,
+      "loss": 0.69,
+      "step": 2728
+    },
+    {
+      "epoch": 1.3539779074097058,
+      "grad_norm": 0.14143560310569692,
+      "learning_rate": 3.724238287822991e-06,
+      "loss": 0.7601,
+      "step": 2729
+    },
+    {
+      "epoch": 1.3544743701129454,
+      "grad_norm": 0.13608277738602498,
+      "learning_rate": 3.723386049376699e-06,
+      "loss": 0.7386,
+      "step": 2730
+    },
+    {
+      "epoch": 1.3549708328161847,
+      "grad_norm": 0.13308627959339595,
+      "learning_rate": 3.7225336239554655e-06,
+      "loss": 0.6834,
+      "step": 2731
+    },
+    {
+      "epoch": 1.355467295519424,
+      "grad_norm": 0.13859104882727302,
+      "learning_rate": 3.72168101168957e-06,
+      "loss": 0.7224,
+      "step": 2732
+    },
+    {
+      "epoch": 1.3559637582226636,
+      "grad_norm": 0.13567729073653412,
+      "learning_rate": 3.7208282127093197e-06,
+      "loss": 0.6887,
+      "step": 2733
+    },
+    {
+      "epoch": 1.356460220925903,
+      "grad_norm": 0.13365628178016073,
+      "learning_rate": 3.7199752271450514e-06,
+      "loss": 0.7077,
+      "step": 2734
+    },
+    {
+      "epoch": 1.3569566836291425,
+      "grad_norm": 0.13570402215298025,
+      "learning_rate": 3.71912205512713e-06,
+      "loss": 0.709,
+      "step": 2735
+    },
+    {
+      "epoch": 1.3574531463323818,
+      "grad_norm": 0.13867332477147382,
+      "learning_rate": 3.71826869678595e-06,
+      "loss": 0.7279,
+      "step": 2736
+    },
+    {
+      "epoch": 1.3579496090356211,
+      "grad_norm": 0.14117778740626497,
+      "learning_rate": 3.717415152251933e-06,
+      "loss": 0.7077,
+      "step": 2737
+    },
+    {
+      "epoch": 1.3584460717388607,
+      "grad_norm": 0.13314975615492863,
+      "learning_rate": 3.71656142165553e-06,
+      "loss": 0.7517,
+      "step": 2738
+    },
+    {
+      "epoch": 1.3589425344421,
+      "grad_norm": 0.13579612596374233,
+      "learning_rate": 3.7157075051272196e-06,
+      "loss": 0.7127,
+      "step": 2739
+    },
+    {
+      "epoch": 1.3594389971453396,
+      "grad_norm": 0.13878702089854905,
+      "learning_rate": 3.71485340279751e-06,
+      "loss": 0.7418,
+      "step": 2740
+    },
+    {
+      "epoch": 1.3599354598485789,
+      "grad_norm": 0.14239282194169628,
+      "learning_rate": 3.7139991147969363e-06,
+      "loss": 0.7374,
+      "step": 2741
+    },
+    {
+      "epoch": 1.3604319225518182,
+      "grad_norm": 0.14056420662523464,
+      "learning_rate": 3.7131446412560624e-06,
+      "loss": 0.7458,
+      "step": 2742
+    },
+    {
+      "epoch": 1.3609283852550578,
+      "grad_norm": 0.13228428795282451,
+      "learning_rate": 3.7122899823054815e-06,
+      "loss": 0.7236,
+      "step": 2743
+    },
+    {
+      "epoch": 1.361424847958297,
+      "grad_norm": 0.13281946270383405,
+      "learning_rate": 3.7114351380758145e-06,
+      "loss": 0.6975,
+      "step": 2744
+    },
+    {
+      "epoch": 1.3619213106615367,
+      "grad_norm": 0.13670672603836637,
+      "learning_rate": 3.71058010869771e-06,
+      "loss": 0.7332,
+      "step": 2745
+    },
+    {
+      "epoch": 1.362417773364776,
+      "grad_norm": 0.14071995024167427,
+      "learning_rate": 3.7097248943018467e-06,
+      "loss": 0.7082,
+      "step": 2746
+    },
+    {
+      "epoch": 1.3629142360680153,
+      "grad_norm": 0.13414480406046128,
+      "learning_rate": 3.7088694950189297e-06,
+      "loss": 0.7126,
+      "step": 2747
+    },
+    {
+      "epoch": 1.3634106987712549,
+      "grad_norm": 0.14516674208748181,
+      "learning_rate": 3.7080139109796933e-06,
+      "loss": 0.7322,
+      "step": 2748
+    },
+    {
+      "epoch": 1.3639071614744942,
+      "grad_norm": 0.13482497758832893,
+      "learning_rate": 3.7071581423148996e-06,
+      "loss": 0.7323,
+      "step": 2749
+    },
+    {
+      "epoch": 1.3644036241777338,
+      "grad_norm": 0.13236613491037902,
+      "learning_rate": 3.7063021891553384e-06,
+      "loss": 0.7301,
+      "step": 2750
+    },
+    {
+      "epoch": 1.364900086880973,
+      "grad_norm": 0.13863145388334375,
+      "learning_rate": 3.7054460516318302e-06,
+      "loss": 0.6891,
+      "step": 2751
+    },
+    {
+      "epoch": 1.3653965495842124,
+      "grad_norm": 0.1305802200578501,
+      "learning_rate": 3.7045897298752196e-06,
+      "loss": 0.754,
+      "step": 2752
+    },
+    {
+      "epoch": 1.365893012287452,
+      "grad_norm": 0.1320898504801369,
+      "learning_rate": 3.703733224016384e-06,
+      "loss": 0.7232,
+      "step": 2753
+    },
+    {
+      "epoch": 1.3663894749906913,
+      "grad_norm": 0.1382071992088785,
+      "learning_rate": 3.7028765341862256e-06,
+      "loss": 0.6633,
+      "step": 2754
+    },
+    {
+      "epoch": 1.3668859376939309,
+      "grad_norm": 0.13239411486453712,
+      "learning_rate": 3.702019660515675e-06,
+      "loss": 0.6814,
+      "step": 2755
+    },
+    {
+      "epoch": 1.3673824003971702,
+      "grad_norm": 0.13222359579600135,
+      "learning_rate": 3.7011626031356924e-06,
+      "loss": 0.7755,
+      "step": 2756
+    },
+    {
+      "epoch": 1.3678788631004095,
+      "grad_norm": 0.15096634032752865,
+      "learning_rate": 3.7003053621772655e-06,
+      "loss": 0.7288,
+      "step": 2757
+    },
+    {
+      "epoch": 1.368375325803649,
+      "grad_norm": 0.13367062633330973,
+      "learning_rate": 3.699447937771409e-06,
+      "loss": 0.7281,
+      "step": 2758
+    },
+    {
+      "epoch": 1.3688717885068884,
+      "grad_norm": 0.13930251553734874,
+      "learning_rate": 3.698590330049167e-06,
+      "loss": 0.7536,
+      "step": 2759
+    },
+    {
+      "epoch": 1.369368251210128,
+      "grad_norm": 0.1333947524333695,
+      "learning_rate": 3.697732539141611e-06,
+      "loss": 0.7267,
+      "step": 2760
+    },
+    {
+      "epoch": 1.3698647139133673,
+      "grad_norm": 0.13373105854598036,
+      "learning_rate": 3.6968745651798404e-06,
+      "loss": 0.7416,
+      "step": 2761
+    },
+    {
+      "epoch": 1.3703611766166066,
+      "grad_norm": 0.15126475875952416,
+      "learning_rate": 3.6960164082949827e-06,
+      "loss": 0.755,
+      "step": 2762
+    },
+    {
+      "epoch": 1.3708576393198462,
+      "grad_norm": 0.32186904582209985,
+      "learning_rate": 3.6951580686181944e-06,
+      "loss": 0.7316,
+      "step": 2763
+    },
+    {
+      "epoch": 1.3713541020230855,
+      "grad_norm": 0.1458070721124448,
+      "learning_rate": 3.6942995462806574e-06,
+      "loss": 0.7678,
+      "step": 2764
+    },
+    {
+      "epoch": 1.371850564726325,
+      "grad_norm": 0.13289186710198236,
+      "learning_rate": 3.693440841413585e-06,
+      "loss": 0.7255,
+      "step": 2765
+    },
+    {
+      "epoch": 1.3723470274295644,
+      "grad_norm": 0.14607683472265695,
+      "learning_rate": 3.6925819541482142e-06,
+      "loss": 0.7241,
+      "step": 2766
+    },
+    {
+      "epoch": 1.3728434901328037,
+      "grad_norm": 0.13565639692027404,
+      "learning_rate": 3.691722884615814e-06,
+      "loss": 0.7772,
+      "step": 2767
+    },
+    {
+      "epoch": 1.3733399528360433,
+      "grad_norm": 0.1395165105118244,
+      "learning_rate": 3.690863632947678e-06,
+      "loss": 0.7336,
+      "step": 2768
+    },
+    {
+      "epoch": 1.3738364155392826,
+      "grad_norm": 0.13691536096706877,
+      "learning_rate": 3.69000419927513e-06,
+      "loss": 0.7331,
+      "step": 2769
+    },
+    {
+      "epoch": 1.3743328782425221,
+      "grad_norm": 0.13120428601535297,
+      "learning_rate": 3.6891445837295215e-06,
+      "loss": 0.7361,
+      "step": 2770
+    },
+    {
+      "epoch": 1.3748293409457615,
+      "grad_norm": 0.13462822014791612,
+      "learning_rate": 3.6882847864422287e-06,
+      "loss": 0.7201,
+      "step": 2771
+    },
+    {
+      "epoch": 1.3753258036490008,
+      "grad_norm": 0.15260067191313678,
+      "learning_rate": 3.687424807544659e-06,
+      "loss": 0.7355,
+      "step": 2772
+    },
+    {
+      "epoch": 1.3758222663522404,
+      "grad_norm": 0.1356268994256984,
+      "learning_rate": 3.686564647168247e-06,
+      "loss": 0.6913,
+      "step": 2773
+    },
+    {
+      "epoch": 1.3763187290554797,
+      "grad_norm": 0.1317137965843151,
+      "learning_rate": 3.6857043054444534e-06,
+      "loss": 0.6801,
+      "step": 2774
+    },
+    {
+      "epoch": 1.3768151917587192,
+      "grad_norm": 0.13155836504147503,
+      "learning_rate": 3.6848437825047678e-06,
+      "loss": 0.701,
+      "step": 2775
+    },
+    {
+      "epoch": 1.3773116544619586,
+      "grad_norm": 0.1271126456023588,
+      "learning_rate": 3.6839830784807086e-06,
+      "loss": 0.6839,
+      "step": 2776
+    },
+    {
+      "epoch": 1.377808117165198,
+      "grad_norm": 0.14089310593605184,
+      "learning_rate": 3.6831221935038185e-06,
+      "loss": 0.7237,
+      "step": 2777
+    },
+    {
+      "epoch": 1.3783045798684375,
+      "grad_norm": 0.13973014329965902,
+      "learning_rate": 3.682261127705671e-06,
+      "loss": 0.6941,
+      "step": 2778
+    },
+    {
+      "epoch": 1.3788010425716768,
+      "grad_norm": 0.12910171833031905,
+      "learning_rate": 3.6813998812178665e-06,
+      "loss": 0.6799,
+      "step": 2779
+    },
+    {
+      "epoch": 1.3792975052749163,
+      "grad_norm": 0.13356773763189392,
+      "learning_rate": 3.680538454172033e-06,
+      "loss": 0.6942,
+      "step": 2780
+    },
+    {
+      "epoch": 1.3797939679781557,
+      "grad_norm": 0.14461559430138288,
+      "learning_rate": 3.6796768466998256e-06,
+      "loss": 0.728,
+      "step": 2781
+    },
+    {
+      "epoch": 1.380290430681395,
+      "grad_norm": 0.14044495895998144,
+      "learning_rate": 3.678815058932926e-06,
+      "loss": 0.7567,
+      "step": 2782
+    },
+    {
+      "epoch": 1.3807868933846346,
+      "grad_norm": 0.13731082360762506,
+      "learning_rate": 3.6779530910030455e-06,
+      "loss": 0.715,
+      "step": 2783
+    },
+    {
+      "epoch": 1.381283356087874,
+      "grad_norm": 0.1445938769341219,
+      "learning_rate": 3.6770909430419216e-06,
+      "loss": 0.7439,
+      "step": 2784
+    },
+    {
+      "epoch": 1.3817798187911134,
+      "grad_norm": 0.13607123879339897,
+      "learning_rate": 3.6762286151813207e-06,
+      "loss": 0.7089,
+      "step": 2785
+    },
+    {
+      "epoch": 1.3822762814943528,
+      "grad_norm": 0.1425234259681637,
+      "learning_rate": 3.6753661075530363e-06,
+      "loss": 0.7555,
+      "step": 2786
+    },
+    {
+      "epoch": 1.382772744197592,
+      "grad_norm": 0.13691972240341396,
+      "learning_rate": 3.6745034202888868e-06,
+      "loss": 0.6762,
+      "step": 2787
+    },
+    {
+      "epoch": 1.3832692069008314,
+      "grad_norm": 0.1359578557968317,
+      "learning_rate": 3.6736405535207215e-06,
+      "loss": 0.6988,
+      "step": 2788
+    },
+    {
+      "epoch": 1.383765669604071,
+      "grad_norm": 0.14251525701942244,
+      "learning_rate": 3.672777507380416e-06,
+      "loss": 0.8063,
+      "step": 2789
+    },
+    {
+      "epoch": 1.3842621323073105,
+      "grad_norm": 0.1361191613459699,
+      "learning_rate": 3.671914281999872e-06,
+      "loss": 0.7197,
+      "step": 2790
+    },
+    {
+      "epoch": 1.3847585950105499,
+      "grad_norm": 0.13112963174458994,
+      "learning_rate": 3.6710508775110204e-06,
+      "loss": 0.7006,
+      "step": 2791
+    },
+    {
+      "epoch": 1.3852550577137892,
+      "grad_norm": 0.13997501686830185,
+      "learning_rate": 3.670187294045819e-06,
+      "loss": 0.7215,
+      "step": 2792
+    },
+    {
+      "epoch": 1.3857515204170285,
+      "grad_norm": 0.13735338515895265,
+      "learning_rate": 3.6693235317362513e-06,
+      "loss": 0.6828,
+      "step": 2793
+    },
+    {
+      "epoch": 1.386247983120268,
+      "grad_norm": 0.13804648308059947,
+      "learning_rate": 3.6684595907143307e-06,
+      "loss": 0.6705,
+      "step": 2794
+    },
+    {
+      "epoch": 1.3867444458235076,
+      "grad_norm": 0.14859762749120925,
+      "learning_rate": 3.6675954711120964e-06,
+      "loss": 0.816,
+      "step": 2795
+    },
+    {
+      "epoch": 1.387240908526747,
+      "grad_norm": 0.13538015684912,
+      "learning_rate": 3.666731173061616e-06,
+      "loss": 0.7397,
+      "step": 2796
+    },
+    {
+      "epoch": 1.3877373712299863,
+      "grad_norm": 0.1310415765785277,
+      "learning_rate": 3.6658666966949823e-06,
+      "loss": 0.6973,
+      "step": 2797
+    },
+    {
+      "epoch": 1.3882338339332256,
+      "grad_norm": 0.13540679229302185,
+      "learning_rate": 3.665002042144318e-06,
+      "loss": 0.7459,
+      "step": 2798
+    },
+    {
+      "epoch": 1.3887302966364652,
+      "grad_norm": 0.14371652911821897,
+      "learning_rate": 3.6641372095417703e-06,
+      "loss": 0.7883,
+      "step": 2799
+    },
+    {
+      "epoch": 1.3892267593397047,
+      "grad_norm": 0.13703857411739195,
+      "learning_rate": 3.663272199019516e-06,
+      "loss": 0.6883,
+      "step": 2800
+    },
+    {
+      "epoch": 1.389723222042944,
+      "grad_norm": 0.13173661216385477,
+      "learning_rate": 3.662407010709757e-06,
+      "loss": 0.6969,
+      "step": 2801
+    },
+    {
+      "epoch": 1.3902196847461834,
+      "grad_norm": 0.12981624688845955,
+      "learning_rate": 3.661541644744725e-06,
+      "loss": 0.6822,
+      "step": 2802
+    },
+    {
+      "epoch": 1.3907161474494227,
+      "grad_norm": 0.13538269173616568,
+      "learning_rate": 3.660676101256676e-06,
+      "loss": 0.6703,
+      "step": 2803
+    },
+    {
+      "epoch": 1.3912126101526623,
+      "grad_norm": 0.13513544679451667,
+      "learning_rate": 3.659810380377895e-06,
+      "loss": 0.7126,
+      "step": 2804
+    },
+    {
+      "epoch": 1.3917090728559016,
+      "grad_norm": 0.13634117970671397,
+      "learning_rate": 3.6589444822406938e-06,
+      "loss": 0.7628,
+      "step": 2805
+    },
+    {
+      "epoch": 1.3922055355591412,
+      "grad_norm": 0.13469088579151892,
+      "learning_rate": 3.6580784069774104e-06,
+      "loss": 0.6827,
+      "step": 2806
+    },
+    {
+      "epoch": 1.3927019982623805,
+      "grad_norm": 0.14097471598925498,
+      "learning_rate": 3.657212154720411e-06,
+      "loss": 0.762,
+      "step": 2807
+    },
+    {
+      "epoch": 1.3931984609656198,
+      "grad_norm": 0.13688984690687683,
+      "learning_rate": 3.656345725602089e-06,
+      "loss": 0.723,
+      "step": 2808
+    },
+    {
+      "epoch": 1.3936949236688594,
+      "grad_norm": 0.13703297547170337,
+      "learning_rate": 3.6554791197548624e-06,
+      "loss": 0.7123,
+      "step": 2809
+    },
+    {
+      "epoch": 1.3941913863720987,
+      "grad_norm": 0.13555651922489106,
+      "learning_rate": 3.654612337311179e-06,
+      "loss": 0.7277,
+      "step": 2810
+    },
+    {
+      "epoch": 1.3946878490753383,
+      "grad_norm": 0.13430071149047185,
+      "learning_rate": 3.6537453784035133e-06,
+      "loss": 0.7369,
+      "step": 2811
+    },
+    {
+      "epoch": 1.3951843117785776,
+      "grad_norm": 0.1342596803538824,
+      "learning_rate": 3.6528782431643652e-06,
+      "loss": 0.7032,
+      "step": 2812
+    },
+    {
+      "epoch": 1.395680774481817,
+      "grad_norm": 0.14006820605549658,
+      "learning_rate": 3.6520109317262624e-06,
+      "loss": 0.711,
+      "step": 2813
+    },
+    {
+      "epoch": 1.3961772371850565,
+      "grad_norm": 0.13345080058461264,
+      "learning_rate": 3.65114344422176e-06,
+      "loss": 0.7336,
+      "step": 2814
+    },
+    {
+      "epoch": 1.3966736998882958,
+      "grad_norm": 0.13603896308716434,
+      "learning_rate": 3.6502757807834392e-06,
+      "loss": 0.7395,
+      "step": 2815
+    },
+    {
+      "epoch": 1.3971701625915354,
+      "grad_norm": 0.15103712063012376,
+      "learning_rate": 3.6494079415439087e-06,
+      "loss": 0.7604,
+      "step": 2816
+    },
+    {
+      "epoch": 1.3976666252947747,
+      "grad_norm": 0.14018161536599755,
+      "learning_rate": 3.6485399266358033e-06,
+      "loss": 0.748,
+      "step": 2817
+    },
+    {
+      "epoch": 1.398163087998014,
+      "grad_norm": 0.13994970770354836,
+      "learning_rate": 3.6476717361917867e-06,
+      "loss": 0.7203,
+      "step": 2818
+    },
+    {
+      "epoch": 1.3986595507012536,
+      "grad_norm": 0.1316527442596197,
+      "learning_rate": 3.6468033703445456e-06,
+      "loss": 0.7076,
+      "step": 2819
+    },
+    {
+      "epoch": 1.399156013404493,
+      "grad_norm": 0.1369640741315148,
+      "learning_rate": 3.645934829226797e-06,
+      "loss": 0.7325,
+      "step": 2820
+    },
+    {
+      "epoch": 1.3996524761077325,
+      "grad_norm": 0.12996187838882917,
+      "learning_rate": 3.6450661129712837e-06,
+      "loss": 0.7624,
+      "step": 2821
+    },
+    {
+      "epoch": 1.4001489388109718,
+      "grad_norm": 0.1481187395661545,
+      "learning_rate": 3.644197221710775e-06,
+      "loss": 0.7542,
+      "step": 2822
+    },
+    {
+      "epoch": 1.4006454015142111,
+      "grad_norm": 0.13864790575291003,
+      "learning_rate": 3.6433281555780666e-06,
+      "loss": 0.7592,
+      "step": 2823
+    },
+    {
+      "epoch": 1.4011418642174507,
+      "grad_norm": 0.13723959229827493,
+      "learning_rate": 3.6424589147059817e-06,
+      "loss": 0.7052,
+      "step": 2824
+    },
+    {
+      "epoch": 1.40163832692069,
+      "grad_norm": 0.14601293853858663,
+      "learning_rate": 3.641589499227369e-06,
+      "loss": 0.7408,
+      "step": 2825
+    },
+    {
+      "epoch": 1.4021347896239296,
+      "grad_norm": 0.1381724745946492,
+      "learning_rate": 3.6407199092751055e-06,
+      "loss": 0.7931,
+      "step": 2826
+    },
+    {
+      "epoch": 1.402631252327169,
+      "grad_norm": 0.14149071737598315,
+      "learning_rate": 3.6398501449820937e-06,
+      "loss": 0.6858,
+      "step": 2827
+    },
+    {
+      "epoch": 1.4031277150304082,
+      "grad_norm": 0.13297864017352057,
+      "learning_rate": 3.638980206481264e-06,
+      "loss": 0.7066,
+      "step": 2828
+    },
+    {
+      "epoch": 1.4036241777336478,
+      "grad_norm": 0.13269406137410125,
+      "learning_rate": 3.638110093905572e-06,
+      "loss": 0.7504,
+      "step": 2829
+    },
+    {
+      "epoch": 1.4041206404368871,
+      "grad_norm": 0.13910245211978362,
+      "learning_rate": 3.6372398073880006e-06,
+      "loss": 0.7451,
+      "step": 2830
+    },
+    {
+      "epoch": 1.4046171031401267,
+      "grad_norm": 0.13944456861859922,
+      "learning_rate": 3.636369347061558e-06,
+      "loss": 0.7568,
+      "step": 2831
+    },
+    {
+      "epoch": 1.405113565843366,
+      "grad_norm": 0.13566144602099214,
+      "learning_rate": 3.6354987130592814e-06,
+      "loss": 0.7047,
+      "step": 2832
+    },
+    {
+      "epoch": 1.4056100285466053,
+      "grad_norm": 0.13220561100847655,
+      "learning_rate": 3.634627905514232e-06,
+      "loss": 0.7132,
+      "step": 2833
+    },
+    {
+      "epoch": 1.4061064912498449,
+      "grad_norm": 0.13332119323744976,
+      "learning_rate": 3.6337569245595007e-06,
+      "loss": 0.6862,
+      "step": 2834
+    },
+    {
+      "epoch": 1.4066029539530842,
+      "grad_norm": 0.14000698566512068,
+      "learning_rate": 3.632885770328202e-06,
+      "loss": 0.7582,
+      "step": 2835
+    },
+    {
+      "epoch": 1.4070994166563238,
+      "grad_norm": 0.1366951831127539,
+      "learning_rate": 3.6320144429534764e-06,
+      "loss": 0.7625,
+      "step": 2836
+    },
+    {
+      "epoch": 1.407595879359563,
+      "grad_norm": 0.13797275616542223,
+      "learning_rate": 3.631142942568495e-06,
+      "loss": 0.715,
+      "step": 2837
+    },
+    {
+      "epoch": 1.4080923420628024,
+      "grad_norm": 0.13373290920899406,
+      "learning_rate": 3.630271269306451e-06,
+      "loss": 0.6955,
+      "step": 2838
+    },
+    {
+      "epoch": 1.408588804766042,
+      "grad_norm": 0.13580255325284915,
+      "learning_rate": 3.629399423300566e-06,
+      "loss": 0.7029,
+      "step": 2839
+    },
+    {
+      "epoch": 1.4090852674692813,
+      "grad_norm": 0.13617036873069288,
+      "learning_rate": 3.628527404684088e-06,
+      "loss": 0.7225,
+      "step": 2840
+    },
+    {
+      "epoch": 1.4095817301725209,
+      "grad_norm": 0.1303825802583835,
+      "learning_rate": 3.6276552135902897e-06,
+      "loss": 0.7193,
+      "step": 2841
+    },
+    {
+      "epoch": 1.4100781928757602,
+      "grad_norm": 0.13392854436162474,
+      "learning_rate": 3.626782850152473e-06,
+      "loss": 0.6943,
+      "step": 2842
+    },
+    {
+      "epoch": 1.4105746555789995,
+      "grad_norm": 0.13806544562439935,
+      "learning_rate": 3.625910314503965e-06,
+      "loss": 0.7319,
+      "step": 2843
+    },
+    {
+      "epoch": 1.411071118282239,
+      "grad_norm": 0.14112019047589364,
+      "learning_rate": 3.625037606778117e-06,
+      "loss": 0.7376,
+      "step": 2844
+    },
+    {
+      "epoch": 1.4115675809854784,
+      "grad_norm": 0.1293965423295043,
+      "learning_rate": 3.62416472710831e-06,
+      "loss": 0.6752,
+      "step": 2845
+    },
+    {
+      "epoch": 1.412064043688718,
+      "grad_norm": 0.13458946982919687,
+      "learning_rate": 3.6232916756279497e-06,
+      "loss": 0.7362,
+      "step": 2846
+    },
+    {
+      "epoch": 1.4125605063919573,
+      "grad_norm": 0.13524959671009265,
+      "learning_rate": 3.6224184524704665e-06,
+      "loss": 0.7239,
+      "step": 2847
+    },
+    {
+      "epoch": 1.4130569690951966,
+      "grad_norm": 0.13614099418088066,
+      "learning_rate": 3.6215450577693196e-06,
+      "loss": 0.6924,
+      "step": 2848
+    },
+    {
+      "epoch": 1.4135534317984362,
+      "grad_norm": 0.1392998440162914,
+      "learning_rate": 3.6206714916579925e-06,
+      "loss": 0.7753,
+      "step": 2849
+    },
+    {
+      "epoch": 1.4140498945016755,
+      "grad_norm": 0.1391830902646412,
+      "learning_rate": 3.6197977542699974e-06,
+      "loss": 0.7125,
+      "step": 2850
+    },
+    {
+      "epoch": 1.414546357204915,
+      "grad_norm": 0.13496515125842937,
+      "learning_rate": 3.6189238457388704e-06,
+      "loss": 0.7242,
+      "step": 2851
+    },
+    {
+      "epoch": 1.4150428199081544,
+      "grad_norm": 0.13540843902758287,
+      "learning_rate": 3.6180497661981733e-06,
+      "loss": 0.7189,
+      "step": 2852
+    },
+    {
+      "epoch": 1.4155392826113937,
+      "grad_norm": 0.1387877546863783,
+      "learning_rate": 3.617175515781497e-06,
+      "loss": 0.7011,
+      "step": 2853
+    },
+    {
+      "epoch": 1.4160357453146333,
+      "grad_norm": 0.132229382194975,
+      "learning_rate": 3.6163010946224552e-06,
+      "loss": 0.7071,
+      "step": 2854
+    },
+    {
+      "epoch": 1.4165322080178726,
+      "grad_norm": 0.13602711318145774,
+      "learning_rate": 3.615426502854689e-06,
+      "loss": 0.6891,
+      "step": 2855
+    },
+    {
+      "epoch": 1.4170286707211122,
+      "grad_norm": 0.13600373841927585,
+      "learning_rate": 3.6145517406118673e-06,
+      "loss": 0.7299,
+      "step": 2856
+    },
+    {
+      "epoch": 1.4175251334243515,
+      "grad_norm": 0.1324714432015326,
+      "learning_rate": 3.613676808027682e-06,
+      "loss": 0.715,
+      "step": 2857
+    },
+    {
+      "epoch": 1.4180215961275908,
+      "grad_norm": 0.132452376428356,
+      "learning_rate": 3.6128017052358535e-06,
+      "loss": 0.7385,
+      "step": 2858
+    },
+    {
+      "epoch": 1.4185180588308304,
+      "grad_norm": 0.13456443333348478,
+      "learning_rate": 3.6119264323701257e-06,
+      "loss": 0.7005,
+      "step": 2859
+    },
+    {
+      "epoch": 1.4190145215340697,
+      "grad_norm": 0.13156032063720458,
+      "learning_rate": 3.611050989564272e-06,
+      "loss": 0.7155,
+      "step": 2860
+    },
+    {
+      "epoch": 1.4195109842373093,
+      "grad_norm": 0.1342914875531962,
+      "learning_rate": 3.6101753769520885e-06,
+      "loss": 0.7072,
+      "step": 2861
+    },
+    {
+      "epoch": 1.4200074469405486,
+      "grad_norm": 0.13305702646687323,
+      "learning_rate": 3.6092995946673996e-06,
+      "loss": 0.7399,
+      "step": 2862
+    },
+    {
+      "epoch": 1.420503909643788,
+      "grad_norm": 0.1429132617790171,
+      "learning_rate": 3.608423642844053e-06,
+      "loss": 0.7405,
+      "step": 2863
+    },
+    {
+      "epoch": 1.4210003723470275,
+      "grad_norm": 0.14321697351882565,
+      "learning_rate": 3.607547521615926e-06,
+      "loss": 0.7462,
+      "step": 2864
+    },
+    {
+      "epoch": 1.4214968350502668,
+      "grad_norm": 0.13562736080521035,
+      "learning_rate": 3.6066712311169173e-06,
+      "loss": 0.7409,
+      "step": 2865
+    },
+    {
+      "epoch": 1.4219932977535064,
+      "grad_norm": 0.1411773380972186,
+      "learning_rate": 3.6057947714809555e-06,
+      "loss": 0.7401,
+      "step": 2866
+    },
+    {
+      "epoch": 1.4224897604567457,
+      "grad_norm": 0.1339410402597349,
+      "learning_rate": 3.6049181428419935e-06,
+      "loss": 0.7163,
+      "step": 2867
+    },
+    {
+      "epoch": 1.422986223159985,
+      "grad_norm": 0.13416362073643048,
+      "learning_rate": 3.6040413453340085e-06,
+      "loss": 0.6948,
+      "step": 2868
+    },
+    {
+      "epoch": 1.4234826858632246,
+      "grad_norm": 0.13676566176317306,
+      "learning_rate": 3.6031643790910066e-06,
+      "loss": 0.6875,
+      "step": 2869
+    },
+    {
+      "epoch": 1.423979148566464,
+      "grad_norm": 0.1313168534405185,
+      "learning_rate": 3.602287244247017e-06,
+      "loss": 0.6881,
+      "step": 2870
+    },
+    {
+      "epoch": 1.4244756112697035,
+      "grad_norm": 0.13677303706245855,
+      "learning_rate": 3.6014099409360955e-06,
+      "loss": 0.7393,
+      "step": 2871
+    },
+    {
+      "epoch": 1.4249720739729428,
+      "grad_norm": 0.14114326056602117,
+      "learning_rate": 3.6005324692923242e-06,
+      "loss": 0.6949,
+      "step": 2872
+    },
+    {
+      "epoch": 1.4254685366761821,
+      "grad_norm": 0.13688796725400607,
+      "learning_rate": 3.5996548294498113e-06,
+      "loss": 0.7323,
+      "step": 2873
+    },
+    {
+      "epoch": 1.4259649993794217,
+      "grad_norm": 0.13491954912768386,
+      "learning_rate": 3.598777021542689e-06,
+      "loss": 0.7144,
+      "step": 2874
+    },
+    {
+      "epoch": 1.426461462082661,
+      "grad_norm": 0.13401282593444483,
+      "learning_rate": 3.5978990457051165e-06,
+      "loss": 0.6912,
+      "step": 2875
+    },
+    {
+      "epoch": 1.4269579247859006,
+      "grad_norm": 0.1402341000199709,
+      "learning_rate": 3.597020902071278e-06,
+      "loss": 0.7434,
+      "step": 2876
+    },
+    {
+      "epoch": 1.4274543874891399,
+      "grad_norm": 0.14610852875970756,
+      "learning_rate": 3.596142590775385e-06,
+      "loss": 0.7442,
+      "step": 2877
+    },
+    {
+      "epoch": 1.4279508501923792,
+      "grad_norm": 0.1395377838118118,
+      "learning_rate": 3.5952641119516725e-06,
+      "loss": 0.7266,
+      "step": 2878
+    },
+    {
+      "epoch": 1.4284473128956188,
+      "grad_norm": 0.13617309441688394,
+      "learning_rate": 3.594385465734401e-06,
+      "loss": 0.7328,
+      "step": 2879
+    },
+    {
+      "epoch": 1.428943775598858,
+      "grad_norm": 0.14331534212615055,
+      "learning_rate": 3.5935066522578576e-06,
+      "loss": 0.7336,
+      "step": 2880
+    },
+    {
+      "epoch": 1.4294402383020977,
+      "grad_norm": 0.13574693488844447,
+      "learning_rate": 3.592627671656356e-06,
+      "loss": 0.6804,
+      "step": 2881
+    },
+    {
+      "epoch": 1.429936701005337,
+      "grad_norm": 0.14742713291670548,
+      "learning_rate": 3.5917485240642336e-06,
+      "loss": 0.7212,
+      "step": 2882
+    },
+    {
+      "epoch": 1.4304331637085763,
+      "grad_norm": 0.14811510498423192,
+      "learning_rate": 3.590869209615854e-06,
+      "loss": 0.7231,
+      "step": 2883
+    },
+    {
+      "epoch": 1.4309296264118159,
+      "grad_norm": 0.13396810414200716,
+      "learning_rate": 3.589989728445607e-06,
+      "loss": 0.7543,
+      "step": 2884
+    },
+    {
+      "epoch": 1.4314260891150552,
+      "grad_norm": 0.13821113307135227,
+      "learning_rate": 3.589110080687907e-06,
+      "loss": 0.7313,
+      "step": 2885
+    },
+    {
+      "epoch": 1.4319225518182948,
+      "grad_norm": 0.1403140258389397,
+      "learning_rate": 3.588230266477193e-06,
+      "loss": 0.7028,
+      "step": 2886
+    },
+    {
+      "epoch": 1.432419014521534,
+      "grad_norm": 0.13675009572376795,
+      "learning_rate": 3.5873502859479316e-06,
+      "loss": 0.7236,
+      "step": 2887
+    },
+    {
+      "epoch": 1.4329154772247734,
+      "grad_norm": 0.129493586039917,
+      "learning_rate": 3.5864701392346125e-06,
+      "loss": 0.7098,
+      "step": 2888
+    },
+    {
+      "epoch": 1.433411939928013,
+      "grad_norm": 0.13454419198703996,
+      "learning_rate": 3.5855898264717535e-06,
+      "loss": 0.6686,
+      "step": 2889
+    },
+    {
+      "epoch": 1.4339084026312523,
+      "grad_norm": 0.1347116060515282,
+      "learning_rate": 3.5847093477938955e-06,
+      "loss": 0.7205,
+      "step": 2890
+    },
+    {
+      "epoch": 1.4344048653344919,
+      "grad_norm": 0.13371886505328962,
+      "learning_rate": 3.583828703335606e-06,
+      "loss": 0.7308,
+      "step": 2891
+    },
+    {
+      "epoch": 1.4349013280377312,
+      "grad_norm": 0.1312328405818565,
+      "learning_rate": 3.5829478932314763e-06,
+      "loss": 0.7005,
+      "step": 2892
+    },
+    {
+      "epoch": 1.4353977907409705,
+      "grad_norm": 0.13178342567206527,
+      "learning_rate": 3.582066917616126e-06,
+      "loss": 0.7287,
+      "step": 2893
+    },
+    {
+      "epoch": 1.43589425344421,
+      "grad_norm": 0.13275451628062787,
+      "learning_rate": 3.5811857766241966e-06,
+      "loss": 0.6984,
+      "step": 2894
+    },
+    {
+      "epoch": 1.4363907161474494,
+      "grad_norm": 0.1341420790665092,
+      "learning_rate": 3.5803044703903566e-06,
+      "loss": 0.6854,
+      "step": 2895
+    },
+    {
+      "epoch": 1.436887178850689,
+      "grad_norm": 0.13503560969815714,
+      "learning_rate": 3.579422999049299e-06,
+      "loss": 0.7089,
+      "step": 2896
+    },
+    {
+      "epoch": 1.4373836415539283,
+      "grad_norm": 0.14368925529103932,
+      "learning_rate": 3.578541362735744e-06,
+      "loss": 0.7332,
+      "step": 2897
+    },
+    {
+      "epoch": 1.4378801042571676,
+      "grad_norm": 0.13993715166267706,
+      "learning_rate": 3.5776595615844343e-06,
+      "loss": 0.7407,
+      "step": 2898
+    },
+    {
+      "epoch": 1.4383765669604072,
+      "grad_norm": 0.13262872626301542,
+      "learning_rate": 3.5767775957301402e-06,
+      "loss": 0.7408,
+      "step": 2899
+    },
+    {
+      "epoch": 1.4388730296636465,
+      "grad_norm": 0.12963084057116644,
+      "learning_rate": 3.575895465307655e-06,
+      "loss": 0.7321,
+      "step": 2900
+    },
+    {
+      "epoch": 1.439369492366886,
+      "grad_norm": 0.13223114216083978,
+      "learning_rate": 3.5750131704517987e-06,
+      "loss": 0.7259,
+      "step": 2901
+    },
+    {
+      "epoch": 1.4398659550701254,
+      "grad_norm": 0.13545769906603244,
+      "learning_rate": 3.574130711297416e-06,
+      "loss": 0.6985,
+      "step": 2902
+    },
+    {
+      "epoch": 1.4403624177733647,
+      "grad_norm": 0.14269902147683536,
+      "learning_rate": 3.5732480879793763e-06,
+      "loss": 0.7081,
+      "step": 2903
+    },
+    {
+      "epoch": 1.4408588804766043,
+      "grad_norm": 0.12939717891557706,
+      "learning_rate": 3.572365300632574e-06,
+      "loss": 0.7087,
+      "step": 2904
+    },
+    {
+      "epoch": 1.4413553431798436,
+      "grad_norm": 0.13198711044968622,
+      "learning_rate": 3.5714823493919305e-06,
+      "loss": 0.6999,
+      "step": 2905
+    },
+    {
+      "epoch": 1.4418518058830831,
+      "grad_norm": 0.14765136086907135,
+      "learning_rate": 3.570599234392389e-06,
+      "loss": 0.7115,
+      "step": 2906
+    },
+    {
+      "epoch": 1.4423482685863225,
+      "grad_norm": 0.13571957777801233,
+      "learning_rate": 3.569715955768921e-06,
+      "loss": 0.6962,
+      "step": 2907
+    },
+    {
+      "epoch": 1.4428447312895618,
+      "grad_norm": 0.13574730889325312,
+      "learning_rate": 3.56883251365652e-06,
+      "loss": 0.7506,
+      "step": 2908
+    },
+    {
+      "epoch": 1.4433411939928014,
+      "grad_norm": 0.13592878384629192,
+      "learning_rate": 3.5679489081902073e-06,
+      "loss": 0.7426,
+      "step": 2909
+    },
+    {
+      "epoch": 1.4438376566960407,
+      "grad_norm": 0.12847140272775823,
+      "learning_rate": 3.5670651395050273e-06,
+      "loss": 0.6982,
+      "step": 2910
+    },
+    {
+      "epoch": 1.4443341193992802,
+      "grad_norm": 0.13102794138380155,
+      "learning_rate": 3.5661812077360496e-06,
+      "loss": 0.7805,
+      "step": 2911
+    },
+    {
+      "epoch": 1.4448305821025196,
+      "grad_norm": 0.13544839042143206,
+      "learning_rate": 3.5652971130183696e-06,
+      "loss": 0.7568,
+      "step": 2912
+    },
+    {
+      "epoch": 1.445327044805759,
+      "grad_norm": 0.12813227644476155,
+      "learning_rate": 3.564412855487106e-06,
+      "loss": 0.7082,
+      "step": 2913
+    },
+    {
+      "epoch": 1.4458235075089985,
+      "grad_norm": 0.14071968051452866,
+      "learning_rate": 3.5635284352774035e-06,
+      "loss": 0.789,
+      "step": 2914
+    },
+    {
+      "epoch": 1.4463199702122378,
+      "grad_norm": 0.13520452750443218,
+      "learning_rate": 3.5626438525244335e-06,
+      "loss": 0.7401,
+      "step": 2915
+    },
+    {
+      "epoch": 1.4468164329154773,
+      "grad_norm": 0.142978764949799,
+      "learning_rate": 3.5617591073633877e-06,
+      "loss": 0.7518,
+      "step": 2916
+    },
+    {
+      "epoch": 1.4473128956187167,
+      "grad_norm": 0.13847892621534574,
+      "learning_rate": 3.560874199929487e-06,
+      "loss": 0.7487,
+      "step": 2917
+    },
+    {
+      "epoch": 1.447809358321956,
+      "grad_norm": 0.13593385078802972,
+      "learning_rate": 3.5599891303579747e-06,
+      "loss": 0.6684,
+      "step": 2918
+    },
+    {
+      "epoch": 1.4483058210251956,
+      "grad_norm": 0.13672639592332414,
+      "learning_rate": 3.559103898784119e-06,
+      "loss": 0.7517,
+      "step": 2919
+    },
+    {
+      "epoch": 1.448802283728435,
+      "grad_norm": 0.13336311824180638,
+      "learning_rate": 3.5582185053432137e-06,
+      "loss": 0.7068,
+      "step": 2920
+    },
+    {
+      "epoch": 1.4492987464316744,
+      "grad_norm": 0.13281209912714664,
+      "learning_rate": 3.5573329501705777e-06,
+      "loss": 0.7424,
+      "step": 2921
+    },
+    {
+      "epoch": 1.4497952091349138,
+      "grad_norm": 0.13043603194523082,
+      "learning_rate": 3.556447233401553e-06,
+      "loss": 0.7195,
+      "step": 2922
+    },
+    {
+      "epoch": 1.450291671838153,
+      "grad_norm": 0.138625753926843,
+      "learning_rate": 3.5555613551715072e-06,
+      "loss": 0.7468,
+      "step": 2923
+    },
+    {
+      "epoch": 1.4507881345413927,
+      "grad_norm": 0.14020220328471075,
+      "learning_rate": 3.554675315615833e-06,
+      "loss": 0.731,
+      "step": 2924
+    },
+    {
+      "epoch": 1.451284597244632,
+      "grad_norm": 0.13627514864142962,
+      "learning_rate": 3.5537891148699476e-06,
+      "loss": 0.7456,
+      "step": 2925
+    },
+    {
+      "epoch": 1.4517810599478715,
+      "grad_norm": 0.12926303570402706,
+      "learning_rate": 3.552902753069293e-06,
+      "loss": 0.7147,
+      "step": 2926
+    },
+    {
+      "epoch": 1.4522775226511109,
+      "grad_norm": 0.14055188911284178,
+      "learning_rate": 3.552016230349334e-06,
+      "loss": 0.7487,
+      "step": 2927
+    },
+    {
+      "epoch": 1.4527739853543502,
+      "grad_norm": 0.14061422002541768,
+      "learning_rate": 3.551129546845561e-06,
+      "loss": 0.7029,
+      "step": 2928
+    },
+    {
+      "epoch": 1.4532704480575895,
+      "grad_norm": 0.13754407040631086,
+      "learning_rate": 3.550242702693491e-06,
+      "loss": 0.7129,
+      "step": 2929
+    },
+    {
+      "epoch": 1.453766910760829,
+      "grad_norm": 0.1326488941997002,
+      "learning_rate": 3.549355698028663e-06,
+      "loss": 0.7334,
+      "step": 2930
+    },
+    {
+      "epoch": 1.4542633734640686,
+      "grad_norm": 0.14133250168355752,
+      "learning_rate": 3.5484685329866424e-06,
+      "loss": 0.7082,
+      "step": 2931
+    },
+    {
+      "epoch": 1.454759836167308,
+      "grad_norm": 0.13403821729170973,
+      "learning_rate": 3.547581207703017e-06,
+      "loss": 0.7646,
+      "step": 2932
+    },
+    {
+      "epoch": 1.4552562988705473,
+      "grad_norm": 0.13637800149605267,
+      "learning_rate": 3.5466937223134007e-06,
+      "loss": 0.7302,
+      "step": 2933
+    },
+    {
+      "epoch": 1.4557527615737866,
+      "grad_norm": 0.13875702141124502,
+      "learning_rate": 3.5458060769534317e-06,
+      "loss": 0.6984,
+      "step": 2934
+    },
+    {
+      "epoch": 1.4562492242770262,
+      "grad_norm": 0.13654925956282576,
+      "learning_rate": 3.5449182717587717e-06,
+      "loss": 0.7609,
+      "step": 2935
+    },
+    {
+      "epoch": 1.4567456869802657,
+      "grad_norm": 0.13693125562691924,
+      "learning_rate": 3.5440303068651077e-06,
+      "loss": 0.7749,
+      "step": 2936
+    },
+    {
+      "epoch": 1.457242149683505,
+      "grad_norm": 0.13851101958423012,
+      "learning_rate": 3.5431421824081512e-06,
+      "loss": 0.7404,
+      "step": 2937
+    },
+    {
+      "epoch": 1.4577386123867444,
+      "grad_norm": 0.1439809260358722,
+      "learning_rate": 3.542253898523638e-06,
+      "loss": 0.7923,
+      "step": 2938
+    },
+    {
+      "epoch": 1.4582350750899837,
+      "grad_norm": 0.13805419183906129,
+      "learning_rate": 3.5413654553473274e-06,
+      "loss": 0.7017,
+      "step": 2939
+    },
+    {
+      "epoch": 1.4587315377932233,
+      "grad_norm": 0.13835157639154178,
+      "learning_rate": 3.5404768530150035e-06,
+      "loss": 0.7551,
+      "step": 2940
+    },
+    {
+      "epoch": 1.4592280004964628,
+      "grad_norm": 0.13287184038076372,
+      "learning_rate": 3.539588091662476e-06,
+      "loss": 0.712,
+      "step": 2941
+    },
+    {
+      "epoch": 1.4597244631997022,
+      "grad_norm": 0.1452095077228309,
+      "learning_rate": 3.5386991714255775e-06,
+      "loss": 0.7842,
+      "step": 2942
+    },
+    {
+      "epoch": 1.4602209259029415,
+      "grad_norm": 0.1314235216478171,
+      "learning_rate": 3.537810092440165e-06,
+      "loss": 0.6679,
+      "step": 2943
+    },
+    {
+      "epoch": 1.4607173886061808,
+      "grad_norm": 0.13906999072318538,
+      "learning_rate": 3.536920854842119e-06,
+      "loss": 0.7524,
+      "step": 2944
+    },
+    {
+      "epoch": 1.4612138513094204,
+      "grad_norm": 0.13429900107538834,
+      "learning_rate": 3.5360314587673463e-06,
+      "loss": 0.7309,
+      "step": 2945
+    },
+    {
+      "epoch": 1.4617103140126597,
+      "grad_norm": 0.141994678409815,
+      "learning_rate": 3.5351419043517764e-06,
+      "loss": 0.7338,
+      "step": 2946
+    },
+    {
+      "epoch": 1.4622067767158993,
+      "grad_norm": 0.13636727486493982,
+      "learning_rate": 3.5342521917313643e-06,
+      "loss": 0.6874,
+      "step": 2947
+    },
+    {
+      "epoch": 1.4627032394191386,
+      "grad_norm": 0.14177928006977158,
+      "learning_rate": 3.5333623210420877e-06,
+      "loss": 0.7831,
+      "step": 2948
+    },
+    {
+      "epoch": 1.463199702122378,
+      "grad_norm": 0.13586560334697423,
+      "learning_rate": 3.532472292419949e-06,
+      "loss": 0.7399,
+      "step": 2949
+    },
+    {
+      "epoch": 1.4636961648256175,
+      "grad_norm": 0.1395384541698917,
+      "learning_rate": 3.531582106000975e-06,
+      "loss": 0.7093,
+      "step": 2950
+    },
+    {
+      "epoch": 1.4641926275288568,
+      "grad_norm": 0.1302253949626591,
+      "learning_rate": 3.5306917619212157e-06,
+      "loss": 0.7005,
+      "step": 2951
+    },
+    {
+      "epoch": 1.4646890902320964,
+      "grad_norm": 0.13335922133133257,
+      "learning_rate": 3.5298012603167463e-06,
+      "loss": 0.7291,
+      "step": 2952
+    },
+    {
+      "epoch": 1.4651855529353357,
+      "grad_norm": 0.1377977708057461,
+      "learning_rate": 3.528910601323666e-06,
+      "loss": 0.7324,
+      "step": 2953
+    },
+    {
+      "epoch": 1.465682015638575,
+      "grad_norm": 0.13572427845757098,
+      "learning_rate": 3.5280197850780986e-06,
+      "loss": 0.723,
+      "step": 2954
+    },
+    {
+      "epoch": 1.4661784783418146,
+      "grad_norm": 0.13536174132092055,
+      "learning_rate": 3.527128811716189e-06,
+      "loss": 0.7521,
+      "step": 2955
+    },
+    {
+      "epoch": 1.466674941045054,
+      "grad_norm": 0.13617464941765656,
+      "learning_rate": 3.5262376813741095e-06,
+      "loss": 0.6934,
+      "step": 2956
+    },
+    {
+      "epoch": 1.4671714037482935,
+      "grad_norm": 0.12812292562730554,
+      "learning_rate": 3.525346394188055e-06,
+      "loss": 0.7028,
+      "step": 2957
+    },
+    {
+      "epoch": 1.4676678664515328,
+      "grad_norm": 0.13368438695645662,
+      "learning_rate": 3.524454950294244e-06,
+      "loss": 0.6785,
+      "step": 2958
+    },
+    {
+      "epoch": 1.4681643291547721,
+      "grad_norm": 0.14014375660987655,
+      "learning_rate": 3.523563349828921e-06,
+      "loss": 0.7574,
+      "step": 2959
+    },
+    {
+      "epoch": 1.4686607918580117,
+      "grad_norm": 0.13236384889227554,
+      "learning_rate": 3.5226715929283507e-06,
+      "loss": 0.7025,
+      "step": 2960
+    },
+    {
+      "epoch": 1.469157254561251,
+      "grad_norm": 0.13657143100492303,
+      "learning_rate": 3.521779679728824e-06,
+      "loss": 0.7412,
+      "step": 2961
+    },
+    {
+      "epoch": 1.4696537172644906,
+      "grad_norm": 0.14598345092076667,
+      "learning_rate": 3.5208876103666566e-06,
+      "loss": 0.7279,
+      "step": 2962
+    },
+    {
+      "epoch": 1.47015017996773,
+      "grad_norm": 0.13362251541331926,
+      "learning_rate": 3.519995384978187e-06,
+      "loss": 0.7256,
+      "step": 2963
+    },
+    {
+      "epoch": 1.4706466426709692,
+      "grad_norm": 0.1405944370949059,
+      "learning_rate": 3.5191030036997774e-06,
+      "loss": 0.7497,
+      "step": 2964
+    },
+    {
+      "epoch": 1.4711431053742088,
+      "grad_norm": 0.13130473785305005,
+      "learning_rate": 3.5182104666678136e-06,
+      "loss": 0.7074,
+      "step": 2965
+    },
+    {
+      "epoch": 1.4716395680774481,
+      "grad_norm": 0.1382341947573506,
+      "learning_rate": 3.517317774018706e-06,
+      "loss": 0.7273,
+      "step": 2966
+    },
+    {
+      "epoch": 1.4721360307806877,
+      "grad_norm": 0.1398909288694717,
+      "learning_rate": 3.516424925888887e-06,
+      "loss": 0.7298,
+      "step": 2967
+    },
+    {
+      "epoch": 1.472632493483927,
+      "grad_norm": 0.13083337585357746,
+      "learning_rate": 3.515531922414816e-06,
+      "loss": 0.6801,
+      "step": 2968
+    },
+    {
+      "epoch": 1.4731289561871663,
+      "grad_norm": 0.13314911288815096,
+      "learning_rate": 3.514638763732974e-06,
+      "loss": 0.6965,
+      "step": 2969
+    },
+    {
+      "epoch": 1.4736254188904059,
+      "grad_norm": 0.14032556414048164,
+      "learning_rate": 3.5137454499798646e-06,
+      "loss": 0.7974,
+      "step": 2970
+    },
+    {
+      "epoch": 1.4741218815936452,
+      "grad_norm": 0.13813913057905594,
+      "learning_rate": 3.5128519812920176e-06,
+      "loss": 0.7583,
+      "step": 2971
+    },
+    {
+      "epoch": 1.4746183442968848,
+      "grad_norm": 0.13702265237609512,
+      "learning_rate": 3.5119583578059845e-06,
+      "loss": 0.703,
+      "step": 2972
+    },
+    {
+      "epoch": 1.475114807000124,
+      "grad_norm": 0.13982120405929221,
+      "learning_rate": 3.5110645796583425e-06,
+      "loss": 0.7253,
+      "step": 2973
+    },
+    {
+      "epoch": 1.4756112697033634,
+      "grad_norm": 0.1364392260235214,
+      "learning_rate": 3.5101706469856913e-06,
+      "loss": 0.748,
+      "step": 2974
+    },
+    {
+      "epoch": 1.476107732406603,
+      "grad_norm": 0.140027409645103,
+      "learning_rate": 3.509276559924653e-06,
+      "loss": 0.7313,
+      "step": 2975
+    },
+    {
+      "epoch": 1.4766041951098423,
+      "grad_norm": 0.1320770359244831,
+      "learning_rate": 3.5083823186118748e-06,
+      "loss": 0.6947,
+      "step": 2976
+    },
+    {
+      "epoch": 1.4771006578130819,
+      "grad_norm": 0.12822200062993674,
+      "learning_rate": 3.5074879231840274e-06,
+      "loss": 0.6715,
+      "step": 2977
+    },
+    {
+      "epoch": 1.4775971205163212,
+      "grad_norm": 0.13286535524254892,
+      "learning_rate": 3.5065933737778054e-06,
+      "loss": 0.7157,
+      "step": 2978
+    },
+    {
+      "epoch": 1.4780935832195605,
+      "grad_norm": 0.13858890088438325,
+      "learning_rate": 3.505698670529925e-06,
+      "loss": 0.7208,
+      "step": 2979
+    },
+    {
+      "epoch": 1.4785900459228,
+      "grad_norm": 0.13027070434072718,
+      "learning_rate": 3.5048038135771285e-06,
+      "loss": 0.7217,
+      "step": 2980
+    },
+    {
+      "epoch": 1.4790865086260394,
+      "grad_norm": 0.1413449153673528,
+      "learning_rate": 3.5039088030561798e-06,
+      "loss": 0.7565,
+      "step": 2981
+    },
+    {
+      "epoch": 1.479582971329279,
+      "grad_norm": 0.1400056680073501,
+      "learning_rate": 3.503013639103867e-06,
+      "loss": 0.7495,
+      "step": 2982
+    },
+    {
+      "epoch": 1.4800794340325183,
+      "grad_norm": 0.13165330464949065,
+      "learning_rate": 3.502118321857001e-06,
+      "loss": 0.7036,
+      "step": 2983
+    },
+    {
+      "epoch": 1.4805758967357576,
+      "grad_norm": 0.13774638206096473,
+      "learning_rate": 3.5012228514524177e-06,
+      "loss": 0.7311,
+      "step": 2984
+    },
+    {
+      "epoch": 1.4810723594389972,
+      "grad_norm": 0.13581207084716526,
+      "learning_rate": 3.5003272280269745e-06,
+      "loss": 0.682,
+      "step": 2985
+    },
+    {
+      "epoch": 1.4815688221422365,
+      "grad_norm": 0.13429634165776128,
+      "learning_rate": 3.4994314517175544e-06,
+      "loss": 0.6948,
+      "step": 2986
+    },
+    {
+      "epoch": 1.482065284845476,
+      "grad_norm": 0.13400596652450936,
+      "learning_rate": 3.4985355226610613e-06,
+      "loss": 0.7803,
+      "step": 2987
+    },
+    {
+      "epoch": 1.4825617475487154,
+      "grad_norm": 0.13658283943813354,
+      "learning_rate": 3.4976394409944236e-06,
+      "loss": 0.6891,
+      "step": 2988
+    },
+    {
+      "epoch": 1.4830582102519547,
+      "grad_norm": 0.1434351594908228,
+      "learning_rate": 3.4967432068545933e-06,
+      "loss": 0.706,
+      "step": 2989
+    },
+    {
+      "epoch": 1.4835546729551943,
+      "grad_norm": 0.14046517002064254,
+      "learning_rate": 3.4958468203785454e-06,
+      "loss": 0.7323,
+      "step": 2990
+    },
+    {
+      "epoch": 1.4840511356584336,
+      "grad_norm": 0.14214435255887042,
+      "learning_rate": 3.4949502817032787e-06,
+      "loss": 0.7598,
+      "step": 2991
+    },
+    {
+      "epoch": 1.4845475983616732,
+      "grad_norm": 0.13754977790132406,
+      "learning_rate": 3.4940535909658134e-06,
+      "loss": 0.7143,
+      "step": 2992
+    },
+    {
+      "epoch": 1.4850440610649125,
+      "grad_norm": 0.14453237890736265,
+      "learning_rate": 3.493156748303196e-06,
+      "loss": 0.7702,
+      "step": 2993
+    },
+    {
+      "epoch": 1.4855405237681518,
+      "grad_norm": 0.1373786250860733,
+      "learning_rate": 3.4922597538524925e-06,
+      "loss": 0.732,
+      "step": 2994
+    },
+    {
+      "epoch": 1.4860369864713914,
+      "grad_norm": 0.1325135135773225,
+      "learning_rate": 3.491362607750796e-06,
+      "loss": 0.6744,
+      "step": 2995
+    },
+    {
+      "epoch": 1.4865334491746307,
+      "grad_norm": 0.13625614276317374,
+      "learning_rate": 3.4904653101352204e-06,
+      "loss": 0.7523,
+      "step": 2996
+    },
+    {
+      "epoch": 1.4870299118778703,
+      "grad_norm": 0.13518539082417194,
+      "learning_rate": 3.4895678611429027e-06,
+      "loss": 0.7093,
+      "step": 2997
+    },
+    {
+      "epoch": 1.4875263745811096,
+      "grad_norm": 0.13657072146153515,
+      "learning_rate": 3.4886702609110045e-06,
+      "loss": 0.717,
+      "step": 2998
+    },
+    {
+      "epoch": 1.488022837284349,
+      "grad_norm": 0.13163907731177207,
+      "learning_rate": 3.487772509576709e-06,
+      "loss": 0.7427,
+      "step": 2999
+    },
+    {
+      "epoch": 1.4885192999875885,
+      "grad_norm": 0.1352649941460766,
+      "learning_rate": 3.4868746072772224e-06,
+      "loss": 0.7275,
+      "step": 3000
+    },
+    {
+      "epoch": 1.4890157626908278,
+      "grad_norm": 0.13558793748009995,
+      "learning_rate": 3.4859765541497758e-06,
+      "loss": 0.7283,
+      "step": 3001
+    },
+    {
+      "epoch": 1.4895122253940674,
+      "grad_norm": 0.1364033885783727,
+      "learning_rate": 3.485078350331622e-06,
+      "loss": 0.7014,
+      "step": 3002
+    },
+    {
+      "epoch": 1.4900086880973067,
+      "grad_norm": 0.13933690940861296,
+      "learning_rate": 3.4841799959600364e-06,
+      "loss": 0.744,
+      "step": 3003
+    },
+    {
+      "epoch": 1.490505150800546,
+      "grad_norm": 0.14140035289104919,
+      "learning_rate": 3.4832814911723187e-06,
+      "loss": 0.7211,
+      "step": 3004
+    },
+    {
+      "epoch": 1.4910016135037856,
+      "grad_norm": 0.13562997010154254,
+      "learning_rate": 3.482382836105791e-06,
+      "loss": 0.7335,
+      "step": 3005
+    },
+    {
+      "epoch": 1.491498076207025,
+      "grad_norm": 0.13945900902072886,
+      "learning_rate": 3.481484030897798e-06,
+      "loss": 0.8175,
+      "step": 3006
+    },
+    {
+      "epoch": 1.4919945389102645,
+      "grad_norm": 0.13803655176116866,
+      "learning_rate": 3.4805850756857083e-06,
+      "loss": 0.7692,
+      "step": 3007
+    },
+    {
+      "epoch": 1.4924910016135038,
+      "grad_norm": 0.14038571658779134,
+      "learning_rate": 3.479685970606912e-06,
+      "loss": 0.7348,
+      "step": 3008
+    },
+    {
+      "epoch": 1.4929874643167431,
+      "grad_norm": 0.13668694894227656,
+      "learning_rate": 3.478786715798823e-06,
+      "loss": 0.7535,
+      "step": 3009
+    },
+    {
+      "epoch": 1.4934839270199827,
+      "grad_norm": 0.12831253400925277,
+      "learning_rate": 3.4778873113988776e-06,
+      "loss": 0.7013,
+      "step": 3010
+    },
+    {
+      "epoch": 1.493980389723222,
+      "grad_norm": 0.13510790471247616,
+      "learning_rate": 3.4769877575445366e-06,
+      "loss": 0.7007,
+      "step": 3011
+    },
+    {
+      "epoch": 1.4944768524264616,
+      "grad_norm": 0.13836794462260757,
+      "learning_rate": 3.4760880543732816e-06,
+      "loss": 0.6941,
+      "step": 3012
+    },
+    {
+      "epoch": 1.4949733151297009,
+      "grad_norm": 0.13929710909188023,
+      "learning_rate": 3.4751882020226174e-06,
+      "loss": 0.7835,
+      "step": 3013
+    },
+    {
+      "epoch": 1.4954697778329402,
+      "grad_norm": 0.13277890430638015,
+      "learning_rate": 3.4742882006300734e-06,
+      "loss": 0.6666,
+      "step": 3014
+    },
+    {
+      "epoch": 1.4959662405361798,
+      "grad_norm": 0.13390371855032038,
+      "learning_rate": 3.4733880503331983e-06,
+      "loss": 0.7052,
+      "step": 3015
+    },
+    {
+      "epoch": 1.496462703239419,
+      "grad_norm": 0.13663110970590517,
+      "learning_rate": 3.4724877512695677e-06,
+      "loss": 0.7179,
+      "step": 3016
+    },
+    {
+      "epoch": 1.4969591659426587,
+      "grad_norm": 0.13773347718243625,
+      "learning_rate": 3.4715873035767766e-06,
+      "loss": 0.7086,
+      "step": 3017
+    },
+    {
+      "epoch": 1.497455628645898,
+      "grad_norm": 0.1363564450553434,
+      "learning_rate": 3.4706867073924446e-06,
+      "loss": 0.7173,
+      "step": 3018
+    },
+    {
+      "epoch": 1.4979520913491373,
+      "grad_norm": 0.13186824374836328,
+      "learning_rate": 3.469785962854213e-06,
+      "loss": 0.7346,
+      "step": 3019
+    },
+    {
+      "epoch": 1.4984485540523769,
+      "grad_norm": 0.13838548862045663,
+      "learning_rate": 3.4688850700997455e-06,
+      "loss": 0.723,
+      "step": 3020
+    },
+    {
+      "epoch": 1.4989450167556162,
+      "grad_norm": 0.13121107819663974,
+      "learning_rate": 3.467984029266731e-06,
+      "loss": 0.7178,
+      "step": 3021
+    },
+    {
+      "epoch": 1.4994414794588558,
+      "grad_norm": 0.1359428642141061,
+      "learning_rate": 3.467082840492878e-06,
+      "loss": 0.731,
+      "step": 3022
+    },
+    {
+      "epoch": 1.499937942162095,
+      "grad_norm": 0.13804903069761093,
+      "learning_rate": 3.4661815039159186e-06,
+      "loss": 0.6986,
+      "step": 3023
+    },
+    {
+      "epoch": 1.5004344048653344,
+      "grad_norm": 0.12815687067739065,
+      "learning_rate": 3.465280019673608e-06,
+      "loss": 0.7029,
+      "step": 3024
+    },
+    {
+      "epoch": 1.5004344048653344,
+      "eval_loss": 0.7363704442977905,
+      "eval_runtime": 135.8597,
+      "eval_samples_per_second": 223.414,
+      "eval_steps_per_second": 27.933,
+      "step": 3024
+    },
+    {
+      "epoch": 1.5009308675685737,
+      "grad_norm": 0.1470067433495751,
+      "learning_rate": 3.4643783879037235e-06,
+      "loss": 0.7213,
+      "step": 3025
+    },
+    {
+      "epoch": 1.5014273302718133,
+      "grad_norm": 0.21715999444549708,
+      "learning_rate": 3.4634766087440645e-06,
+      "loss": 0.7743,
+      "step": 3026
+    },
+    {
+      "epoch": 1.5019237929750529,
+      "grad_norm": 0.13601624552016667,
+      "learning_rate": 3.4625746823324545e-06,
+      "loss": 0.7205,
+      "step": 3027
+    },
+    {
+      "epoch": 1.5024202556782922,
+      "grad_norm": 0.1323109309363181,
+      "learning_rate": 3.4616726088067383e-06,
+      "loss": 0.7679,
+      "step": 3028
+    },
+    {
+      "epoch": 1.5029167183815315,
+      "grad_norm": 0.13051111527820897,
+      "learning_rate": 3.460770388304782e-06,
+      "loss": 0.7117,
+      "step": 3029
+    },
+    {
+      "epoch": 1.5034131810847708,
+      "grad_norm": 0.1354078149158313,
+      "learning_rate": 3.459868020964478e-06,
+      "loss": 0.7352,
+      "step": 3030
+    },
+    {
+      "epoch": 1.5039096437880104,
+      "grad_norm": 0.14053113861350988,
+      "learning_rate": 3.4589655069237367e-06,
+      "loss": 0.7575,
+      "step": 3031
+    },
+    {
+      "epoch": 1.50440610649125,
+      "grad_norm": 0.13146810822091834,
+      "learning_rate": 3.4580628463204936e-06,
+      "loss": 0.7172,
+      "step": 3032
+    },
+    {
+      "epoch": 1.5049025691944893,
+      "grad_norm": 0.13001154124097078,
+      "learning_rate": 3.457160039292705e-06,
+      "loss": 0.6829,
+      "step": 3033
+    },
+    {
+      "epoch": 1.5053990318977286,
+      "grad_norm": 0.13574315079485202,
+      "learning_rate": 3.456257085978352e-06,
+      "loss": 0.7401,
+      "step": 3034
+    },
+    {
+      "epoch": 1.505895494600968,
+      "grad_norm": 0.13467407064352518,
+      "learning_rate": 3.4553539865154362e-06,
+      "loss": 0.7276,
+      "step": 3035
+    },
+    {
+      "epoch": 1.5063919573042075,
+      "grad_norm": 0.16132459867477508,
+      "learning_rate": 3.4544507410419807e-06,
+      "loss": 0.7567,
+      "step": 3036
+    },
+    {
+      "epoch": 1.506888420007447,
+      "grad_norm": 0.13133038324487403,
+      "learning_rate": 3.453547349696033e-06,
+      "loss": 0.7089,
+      "step": 3037
+    },
+    {
+      "epoch": 1.5073848827106864,
+      "grad_norm": 0.13434120206254171,
+      "learning_rate": 3.4526438126156624e-06,
+      "loss": 0.7183,
+      "step": 3038
+    },
+    {
+      "epoch": 1.5078813454139257,
+      "grad_norm": 0.133739491071614,
+      "learning_rate": 3.45174012993896e-06,
+      "loss": 0.6883,
+      "step": 3039
+    },
+    {
+      "epoch": 1.508377808117165,
+      "grad_norm": 0.13292833480778746,
+      "learning_rate": 3.450836301804038e-06,
+      "loss": 0.7354,
+      "step": 3040
+    },
+    {
+      "epoch": 1.5088742708204046,
+      "grad_norm": 0.1310771402662206,
+      "learning_rate": 3.449932328349033e-06,
+      "loss": 0.7078,
+      "step": 3041
+    },
+    {
+      "epoch": 1.5093707335236441,
+      "grad_norm": 0.13464111175584303,
+      "learning_rate": 3.449028209712102e-06,
+      "loss": 0.6777,
+      "step": 3042
+    },
+    {
+      "epoch": 1.5098671962268835,
+      "grad_norm": 0.13372159063346198,
+      "learning_rate": 3.4481239460314252e-06,
+      "loss": 0.7078,
+      "step": 3043
+    },
+    {
+      "epoch": 1.5103636589301228,
+      "grad_norm": 0.14328077784905502,
+      "learning_rate": 3.4472195374452067e-06,
+      "loss": 0.7541,
+      "step": 3044
+    },
+    {
+      "epoch": 1.5108601216333621,
+      "grad_norm": 0.14108614619902352,
+      "learning_rate": 3.446314984091669e-06,
+      "loss": 0.7537,
+      "step": 3045
+    },
+    {
+      "epoch": 1.5113565843366017,
+      "grad_norm": 0.1387413397672071,
+      "learning_rate": 3.445410286109059e-06,
+      "loss": 0.7313,
+      "step": 3046
+    },
+    {
+      "epoch": 1.5118530470398412,
+      "grad_norm": 0.1437298602735984,
+      "learning_rate": 3.444505443635645e-06,
+      "loss": 0.7198,
+      "step": 3047
+    },
+    {
+      "epoch": 1.5123495097430806,
+      "grad_norm": 0.14561460770522325,
+      "learning_rate": 3.4436004568097177e-06,
+      "loss": 0.7763,
+      "step": 3048
+    },
+    {
+      "epoch": 1.51284597244632,
+      "grad_norm": 0.1437698294167037,
+      "learning_rate": 3.4426953257695904e-06,
+      "loss": 0.7282,
+      "step": 3049
+    },
+    {
+      "epoch": 1.5133424351495592,
+      "grad_norm": 0.13602577897519388,
+      "learning_rate": 3.4417900506535977e-06,
+      "loss": 0.7059,
+      "step": 3050
+    },
+    {
+      "epoch": 1.5138388978527988,
+      "grad_norm": 0.13627196775475553,
+      "learning_rate": 3.440884631600096e-06,
+      "loss": 0.6885,
+      "step": 3051
+    },
+    {
+      "epoch": 1.5143353605560383,
+      "grad_norm": 0.13016712398288627,
+      "learning_rate": 3.4399790687474638e-06,
+      "loss": 0.7099,
+      "step": 3052
+    },
+    {
+      "epoch": 1.5148318232592777,
+      "grad_norm": 0.1420534555060904,
+      "learning_rate": 3.4390733622341025e-06,
+      "loss": 0.7624,
+      "step": 3053
+    },
+    {
+      "epoch": 1.515328285962517,
+      "grad_norm": 0.13710013215332392,
+      "learning_rate": 3.438167512198436e-06,
+      "loss": 0.7549,
+      "step": 3054
+    },
+    {
+      "epoch": 1.5158247486657563,
+      "grad_norm": 0.14905485277450392,
+      "learning_rate": 3.437261518778906e-06,
+      "loss": 0.7153,
+      "step": 3055
+    },
+    {
+      "epoch": 1.516321211368996,
+      "grad_norm": 0.13714440652108204,
+      "learning_rate": 3.4363553821139822e-06,
+      "loss": 0.7061,
+      "step": 3056
+    },
+    {
+      "epoch": 1.5168176740722354,
+      "grad_norm": 0.1448142948924188,
+      "learning_rate": 3.4354491023421503e-06,
+      "loss": 0.7118,
+      "step": 3057
+    },
+    {
+      "epoch": 1.5173141367754748,
+      "grad_norm": 0.13221928060116708,
+      "learning_rate": 3.434542679601922e-06,
+      "loss": 0.6995,
+      "step": 3058
+    },
+    {
+      "epoch": 1.517810599478714,
+      "grad_norm": 0.1406755984922586,
+      "learning_rate": 3.4336361140318298e-06,
+      "loss": 0.7159,
+      "step": 3059
+    },
+    {
+      "epoch": 1.5183070621819534,
+      "grad_norm": 0.13490595450975446,
+      "learning_rate": 3.432729405770427e-06,
+      "loss": 0.7348,
+      "step": 3060
+    },
+    {
+      "epoch": 1.518803524885193,
+      "grad_norm": 0.1331655447314572,
+      "learning_rate": 3.4318225549562907e-06,
+      "loss": 0.6946,
+      "step": 3061
+    },
+    {
+      "epoch": 1.5192999875884325,
+      "grad_norm": 0.1376254608523503,
+      "learning_rate": 3.4309155617280164e-06,
+      "loss": 0.7276,
+      "step": 3062
+    },
+    {
+      "epoch": 1.5197964502916719,
+      "grad_norm": 0.131419562336429,
+      "learning_rate": 3.430008426224225e-06,
+      "loss": 0.7282,
+      "step": 3063
+    },
+    {
+      "epoch": 1.5202929129949112,
+      "grad_norm": 0.13982511067051417,
+      "learning_rate": 3.429101148583557e-06,
+      "loss": 0.7345,
+      "step": 3064
+    },
+    {
+      "epoch": 1.5207893756981505,
+      "grad_norm": 0.13655012597581628,
+      "learning_rate": 3.4281937289446753e-06,
+      "loss": 0.7052,
+      "step": 3065
+    },
+    {
+      "epoch": 1.52128583840139,
+      "grad_norm": 0.13360080661036083,
+      "learning_rate": 3.4272861674462653e-06,
+      "loss": 0.6868,
+      "step": 3066
+    },
+    {
+      "epoch": 1.5217823011046296,
+      "grad_norm": 0.12748341437271118,
+      "learning_rate": 3.426378464227032e-06,
+      "loss": 0.6866,
+      "step": 3067
+    },
+    {
+      "epoch": 1.522278763807869,
+      "grad_norm": 0.13955206652734053,
+      "learning_rate": 3.425470619425704e-06,
+      "loss": 0.723,
+      "step": 3068
+    },
+    {
+      "epoch": 1.5227752265111083,
+      "grad_norm": 0.1419812477049388,
+      "learning_rate": 3.42456263318103e-06,
+      "loss": 0.7752,
+      "step": 3069
+    },
+    {
+      "epoch": 1.5232716892143476,
+      "grad_norm": 0.1301543894648006,
+      "learning_rate": 3.4236545056317828e-06,
+      "loss": 0.7152,
+      "step": 3070
+    },
+    {
+      "epoch": 1.5237681519175872,
+      "grad_norm": 0.13531435292463873,
+      "learning_rate": 3.422746236916753e-06,
+      "loss": 0.7006,
+      "step": 3071
+    },
+    {
+      "epoch": 1.5242646146208267,
+      "grad_norm": 0.1316850500436868,
+      "learning_rate": 3.4218378271747566e-06,
+      "loss": 0.706,
+      "step": 3072
+    },
+    {
+      "epoch": 1.524761077324066,
+      "grad_norm": 0.13593788426927872,
+      "learning_rate": 3.420929276544629e-06,
+      "loss": 0.6914,
+      "step": 3073
+    },
+    {
+      "epoch": 1.5252575400273054,
+      "grad_norm": 0.13157963484148716,
+      "learning_rate": 3.420020585165227e-06,
+      "loss": 0.6814,
+      "step": 3074
+    },
+    {
+      "epoch": 1.5257540027305447,
+      "grad_norm": 0.13002115412677262,
+      "learning_rate": 3.4191117531754296e-06,
+      "loss": 0.7078,
+      "step": 3075
+    },
+    {
+      "epoch": 1.5262504654337843,
+      "grad_norm": 0.13527130246956548,
+      "learning_rate": 3.418202780714138e-06,
+      "loss": 0.7036,
+      "step": 3076
+    },
+    {
+      "epoch": 1.5267469281370238,
+      "grad_norm": 0.13262041773411234,
+      "learning_rate": 3.4172936679202745e-06,
+      "loss": 0.7757,
+      "step": 3077
+    },
+    {
+      "epoch": 1.5272433908402632,
+      "grad_norm": 0.13085661822705708,
+      "learning_rate": 3.4163844149327807e-06,
+      "loss": 0.737,
+      "step": 3078
+    },
+    {
+      "epoch": 1.5277398535435025,
+      "grad_norm": 0.13709896836317348,
+      "learning_rate": 3.4154750218906226e-06,
+      "loss": 0.7503,
+      "step": 3079
+    },
+    {
+      "epoch": 1.5282363162467418,
+      "grad_norm": 0.15213123399903336,
+      "learning_rate": 3.414565488932785e-06,
+      "loss": 0.6988,
+      "step": 3080
+    },
+    {
+      "epoch": 1.5287327789499814,
+      "grad_norm": 0.13278188283664377,
+      "learning_rate": 3.4136558161982767e-06,
+      "loss": 0.6908,
+      "step": 3081
+    },
+    {
+      "epoch": 1.529229241653221,
+      "grad_norm": 0.1414124845528321,
+      "learning_rate": 3.4127460038261274e-06,
+      "loss": 0.7579,
+      "step": 3082
+    },
+    {
+      "epoch": 1.5297257043564603,
+      "grad_norm": 0.1327347365759449,
+      "learning_rate": 3.411836051955385e-06,
+      "loss": 0.7004,
+      "step": 3083
+    },
+    {
+      "epoch": 1.5302221670596996,
+      "grad_norm": 0.13080333502954117,
+      "learning_rate": 3.410925960725123e-06,
+      "loss": 0.6939,
+      "step": 3084
+    },
+    {
+      "epoch": 1.530718629762939,
+      "grad_norm": 0.13257078672902534,
+      "learning_rate": 3.4100157302744324e-06,
+      "loss": 0.6948,
+      "step": 3085
+    },
+    {
+      "epoch": 1.5312150924661785,
+      "grad_norm": 0.1322459345068561,
+      "learning_rate": 3.4091053607424295e-06,
+      "loss": 0.7182,
+      "step": 3086
+    },
+    {
+      "epoch": 1.531711555169418,
+      "grad_norm": 0.14457430144423755,
+      "learning_rate": 3.408194852268248e-06,
+      "loss": 0.7577,
+      "step": 3087
+    },
+    {
+      "epoch": 1.5322080178726574,
+      "grad_norm": 0.13598763735719102,
+      "learning_rate": 3.4072842049910458e-06,
+      "loss": 0.7082,
+      "step": 3088
+    },
+    {
+      "epoch": 1.5327044805758967,
+      "grad_norm": 0.13563192470184163,
+      "learning_rate": 3.406373419049999e-06,
+      "loss": 0.7036,
+      "step": 3089
+    },
+    {
+      "epoch": 1.533200943279136,
+      "grad_norm": 0.13408648564933517,
+      "learning_rate": 3.4054624945843083e-06,
+      "loss": 0.7111,
+      "step": 3090
+    },
+    {
+      "epoch": 1.5336974059823756,
+      "grad_norm": 0.13291448319089896,
+      "learning_rate": 3.404551431733193e-06,
+      "loss": 0.692,
+      "step": 3091
+    },
+    {
+      "epoch": 1.5341938686856151,
+      "grad_norm": 0.1357471952977748,
+      "learning_rate": 3.403640230635895e-06,
+      "loss": 0.7497,
+      "step": 3092
+    },
+    {
+      "epoch": 1.5346903313888545,
+      "grad_norm": 0.13026474673855643,
+      "learning_rate": 3.402728891431677e-06,
+      "loss": 0.7322,
+      "step": 3093
+    },
+    {
+      "epoch": 1.5351867940920938,
+      "grad_norm": 0.1344691776866103,
+      "learning_rate": 3.4018174142598215e-06,
+      "loss": 0.6748,
+      "step": 3094
+    },
+    {
+      "epoch": 1.5356832567953331,
+      "grad_norm": 0.1275077838340116,
+      "learning_rate": 3.400905799259634e-06,
+      "loss": 0.6254,
+      "step": 3095
+    },
+    {
+      "epoch": 1.5361797194985727,
+      "grad_norm": 0.13405200474821982,
+      "learning_rate": 3.3999940465704397e-06,
+      "loss": 0.7132,
+      "step": 3096
+    },
+    {
+      "epoch": 1.5366761822018122,
+      "grad_norm": 0.137409082156887,
+      "learning_rate": 3.3990821563315857e-06,
+      "loss": 0.7572,
+      "step": 3097
+    },
+    {
+      "epoch": 1.5371726449050516,
+      "grad_norm": 0.13337458486706275,
+      "learning_rate": 3.3981701286824396e-06,
+      "loss": 0.7658,
+      "step": 3098
+    },
+    {
+      "epoch": 1.537669107608291,
+      "grad_norm": 0.14360779977581084,
+      "learning_rate": 3.3972579637623916e-06,
+      "loss": 0.7088,
+      "step": 3099
+    },
+    {
+      "epoch": 1.5381655703115302,
+      "grad_norm": 0.1298485707521375,
+      "learning_rate": 3.396345661710849e-06,
+      "loss": 0.6769,
+      "step": 3100
+    },
+    {
+      "epoch": 1.5386620330147698,
+      "grad_norm": 0.13722224767161476,
+      "learning_rate": 3.3954332226672444e-06,
+      "loss": 0.7531,
+      "step": 3101
+    },
+    {
+      "epoch": 1.5391584957180093,
+      "grad_norm": 0.13819578335401508,
+      "learning_rate": 3.394520646771029e-06,
+      "loss": 0.764,
+      "step": 3102
+    },
+    {
+      "epoch": 1.5396549584212487,
+      "grad_norm": 0.13810759011008472,
+      "learning_rate": 3.3936079341616746e-06,
+      "loss": 0.7278,
+      "step": 3103
+    },
+    {
+      "epoch": 1.540151421124488,
+      "grad_norm": 0.13089937223200337,
+      "learning_rate": 3.392695084978677e-06,
+      "loss": 0.7161,
+      "step": 3104
+    },
+    {
+      "epoch": 1.5406478838277273,
+      "grad_norm": 0.13279609037952766,
+      "learning_rate": 3.391782099361547e-06,
+      "loss": 0.7043,
+      "step": 3105
+    },
+    {
+      "epoch": 1.5411443465309669,
+      "grad_norm": 0.1398313704093573,
+      "learning_rate": 3.390868977449822e-06,
+      "loss": 0.7615,
+      "step": 3106
+    },
+    {
+      "epoch": 1.5416408092342064,
+      "grad_norm": 0.13318449528960383,
+      "learning_rate": 3.3899557193830585e-06,
+      "loss": 0.6915,
+      "step": 3107
+    },
+    {
+      "epoch": 1.5421372719374458,
+      "grad_norm": 0.13626189356645965,
+      "learning_rate": 3.389042325300832e-06,
+      "loss": 0.7317,
+      "step": 3108
+    },
+    {
+      "epoch": 1.542633734640685,
+      "grad_norm": 0.1354355510287649,
+      "learning_rate": 3.3881287953427423e-06,
+      "loss": 0.684,
+      "step": 3109
+    },
+    {
+      "epoch": 1.5431301973439244,
+      "grad_norm": 0.13700779510437183,
+      "learning_rate": 3.387215129648405e-06,
+      "loss": 0.7707,
+      "step": 3110
+    },
+    {
+      "epoch": 1.543626660047164,
+      "grad_norm": 0.13797047616141508,
+      "learning_rate": 3.386301328357461e-06,
+      "loss": 0.7168,
+      "step": 3111
+    },
+    {
+      "epoch": 1.5441231227504035,
+      "grad_norm": 0.14189987866803816,
+      "learning_rate": 3.3853873916095693e-06,
+      "loss": 0.6903,
+      "step": 3112
+    },
+    {
+      "epoch": 1.5446195854536429,
+      "grad_norm": 0.1373537496866223,
+      "learning_rate": 3.3844733195444108e-06,
+      "loss": 0.7168,
+      "step": 3113
+    },
+    {
+      "epoch": 1.5451160481568822,
+      "grad_norm": 0.13354628494592605,
+      "learning_rate": 3.383559112301687e-06,
+      "loss": 0.7073,
+      "step": 3114
+    },
+    {
+      "epoch": 1.5456125108601215,
+      "grad_norm": 0.134865493027683,
+      "learning_rate": 3.38264477002112e-06,
+      "loss": 0.7345,
+      "step": 3115
+    },
+    {
+      "epoch": 1.546108973563361,
+      "grad_norm": 0.13796959049690674,
+      "learning_rate": 3.3817302928424517e-06,
+      "loss": 0.7315,
+      "step": 3116
+    },
+    {
+      "epoch": 1.5466054362666006,
+      "grad_norm": 0.1321638322323823,
+      "learning_rate": 3.380815680905446e-06,
+      "loss": 0.6957,
+      "step": 3117
+    },
+    {
+      "epoch": 1.54710189896984,
+      "grad_norm": 0.13809644723238076,
+      "learning_rate": 3.3799009343498863e-06,
+      "loss": 0.7913,
+      "step": 3118
+    },
+    {
+      "epoch": 1.5475983616730793,
+      "grad_norm": 0.14219819418095606,
+      "learning_rate": 3.3789860533155764e-06,
+      "loss": 0.738,
+      "step": 3119
+    },
+    {
+      "epoch": 1.5480948243763186,
+      "grad_norm": 0.1405366068876724,
+      "learning_rate": 3.3780710379423425e-06,
+      "loss": 0.7318,
+      "step": 3120
+    },
+    {
+      "epoch": 1.5485912870795582,
+      "grad_norm": 0.12933097234296537,
+      "learning_rate": 3.3771558883700286e-06,
+      "loss": 0.7257,
+      "step": 3121
+    },
+    {
+      "epoch": 1.5490877497827977,
+      "grad_norm": 0.13488996517159482,
+      "learning_rate": 3.3762406047385012e-06,
+      "loss": 0.6776,
+      "step": 3122
+    },
+    {
+      "epoch": 1.549584212486037,
+      "grad_norm": 0.1393952179684061,
+      "learning_rate": 3.3753251871876467e-06,
+      "loss": 0.7573,
+      "step": 3123
+    },
+    {
+      "epoch": 1.5500806751892764,
+      "grad_norm": 0.1386055398625771,
+      "learning_rate": 3.3744096358573724e-06,
+      "loss": 0.7472,
+      "step": 3124
+    },
+    {
+      "epoch": 1.5505771378925157,
+      "grad_norm": 0.1360908123346939,
+      "learning_rate": 3.3734939508876057e-06,
+      "loss": 0.7527,
+      "step": 3125
+    },
+    {
+      "epoch": 1.5510736005957553,
+      "grad_norm": 0.13138200688447002,
+      "learning_rate": 3.3725781324182945e-06,
+      "loss": 0.6892,
+      "step": 3126
+    },
+    {
+      "epoch": 1.5515700632989948,
+      "grad_norm": 0.13489013837525454,
+      "learning_rate": 3.3716621805894056e-06,
+      "loss": 0.7711,
+      "step": 3127
+    },
+    {
+      "epoch": 1.5520665260022342,
+      "grad_norm": 0.13712844460423768,
+      "learning_rate": 3.370746095540928e-06,
+      "loss": 0.7255,
+      "step": 3128
+    },
+    {
+      "epoch": 1.5525629887054735,
+      "grad_norm": 0.13677616322194255,
+      "learning_rate": 3.369829877412871e-06,
+      "loss": 0.7349,
+      "step": 3129
+    },
+    {
+      "epoch": 1.5530594514087128,
+      "grad_norm": 0.13423740883277105,
+      "learning_rate": 3.368913526345265e-06,
+      "loss": 0.7294,
+      "step": 3130
+    },
+    {
+      "epoch": 1.5535559141119524,
+      "grad_norm": 0.1328411152620118,
+      "learning_rate": 3.3679970424781584e-06,
+      "loss": 0.709,
+      "step": 3131
+    },
+    {
+      "epoch": 1.5540523768151917,
+      "grad_norm": 0.14012966529834792,
+      "learning_rate": 3.3670804259516203e-06,
+      "loss": 0.7004,
+      "step": 3132
+    },
+    {
+      "epoch": 1.5545488395184313,
+      "grad_norm": 0.13291614080398312,
+      "learning_rate": 3.366163676905742e-06,
+      "loss": 0.7144,
+      "step": 3133
+    },
+    {
+      "epoch": 1.5550453022216706,
+      "grad_norm": 0.13260952957552452,
+      "learning_rate": 3.365246795480634e-06,
+      "loss": 0.7659,
+      "step": 3134
+    },
+    {
+      "epoch": 1.55554176492491,
+      "grad_norm": 0.135098720690184,
+      "learning_rate": 3.3643297818164263e-06,
+      "loss": 0.6968,
+      "step": 3135
+    },
+    {
+      "epoch": 1.5560382276281495,
+      "grad_norm": 0.1337151472755321,
+      "learning_rate": 3.3634126360532694e-06,
+      "loss": 0.7579,
+      "step": 3136
+    },
+    {
+      "epoch": 1.5565346903313888,
+      "grad_norm": 0.13823809792997804,
+      "learning_rate": 3.362495358331336e-06,
+      "loss": 0.7366,
+      "step": 3137
+    },
+    {
+      "epoch": 1.5570311530346284,
+      "grad_norm": 0.13540112581495553,
+      "learning_rate": 3.3615779487908147e-06,
+      "loss": 0.778,
+      "step": 3138
+    },
+    {
+      "epoch": 1.5575276157378677,
+      "grad_norm": 0.12755996566178962,
+      "learning_rate": 3.3606604075719187e-06,
+      "loss": 0.6915,
+      "step": 3139
+    },
+    {
+      "epoch": 1.558024078441107,
+      "grad_norm": 0.131901346948097,
+      "learning_rate": 3.359742734814879e-06,
+      "loss": 0.6949,
+      "step": 3140
+    },
+    {
+      "epoch": 1.5585205411443466,
+      "grad_norm": 0.13117046874293734,
+      "learning_rate": 3.358824930659948e-06,
+      "loss": 0.6904,
+      "step": 3141
+    },
+    {
+      "epoch": 1.559017003847586,
+      "grad_norm": 0.13765163137556422,
+      "learning_rate": 3.3579069952473964e-06,
+      "loss": 0.7074,
+      "step": 3142
+    },
+    {
+      "epoch": 1.5595134665508255,
+      "grad_norm": 0.143697911811937,
+      "learning_rate": 3.3569889287175155e-06,
+      "loss": 0.71,
+      "step": 3143
+    },
+    {
+      "epoch": 1.5600099292540648,
+      "grad_norm": 0.13255298712174185,
+      "learning_rate": 3.356070731210618e-06,
+      "loss": 0.7144,
+      "step": 3144
+    },
+    {
+      "epoch": 1.5605063919573041,
+      "grad_norm": 0.1390079209743311,
+      "learning_rate": 3.3551524028670348e-06,
+      "loss": 0.7387,
+      "step": 3145
+    },
+    {
+      "epoch": 1.5610028546605437,
+      "grad_norm": 0.15837144226586045,
+      "learning_rate": 3.3542339438271184e-06,
+      "loss": 0.8018,
+      "step": 3146
+    },
+    {
+      "epoch": 1.561499317363783,
+      "grad_norm": 0.13346525456733854,
+      "learning_rate": 3.35331535423124e-06,
+      "loss": 0.7318,
+      "step": 3147
+    },
+    {
+      "epoch": 1.5619957800670226,
+      "grad_norm": 0.1303144364718177,
+      "learning_rate": 3.352396634219792e-06,
+      "loss": 0.691,
+      "step": 3148
+    },
+    {
+      "epoch": 1.5624922427702619,
+      "grad_norm": 0.13463555403818014,
+      "learning_rate": 3.351477783933186e-06,
+      "loss": 0.7917,
+      "step": 3149
+    },
+    {
+      "epoch": 1.5629887054735012,
+      "grad_norm": 0.13535091861217688,
+      "learning_rate": 3.3505588035118517e-06,
+      "loss": 0.7196,
+      "step": 3150
+    },
+    {
+      "epoch": 1.5634851681767408,
+      "grad_norm": 0.1263572778973972,
+      "learning_rate": 3.3496396930962437e-06,
+      "loss": 0.6855,
+      "step": 3151
+    },
+    {
+      "epoch": 1.56398163087998,
+      "grad_norm": 0.18384206339467457,
+      "learning_rate": 3.3487204528268302e-06,
+      "loss": 0.7055,
+      "step": 3152
+    },
+    {
+      "epoch": 1.5644780935832197,
+      "grad_norm": 0.14081009970447547,
+      "learning_rate": 3.347801082844105e-06,
+      "loss": 0.7067,
+      "step": 3153
+    },
+    {
+      "epoch": 1.564974556286459,
+      "grad_norm": 0.13263091004679445,
+      "learning_rate": 3.3468815832885772e-06,
+      "loss": 0.6618,
+      "step": 3154
+    },
+    {
+      "epoch": 1.5654710189896983,
+      "grad_norm": 0.1440081195073848,
+      "learning_rate": 3.3459619543007772e-06,
+      "loss": 0.7223,
+      "step": 3155
+    },
+    {
+      "epoch": 1.5659674816929379,
+      "grad_norm": 0.13271016860764132,
+      "learning_rate": 3.345042196021257e-06,
+      "loss": 0.7639,
+      "step": 3156
+    },
+    {
+      "epoch": 1.5664639443961772,
+      "grad_norm": 0.13353122154660912,
+      "learning_rate": 3.3441223085905873e-06,
+      "loss": 0.7087,
+      "step": 3157
+    },
+    {
+      "epoch": 1.5669604070994168,
+      "grad_norm": 0.13011358405310236,
+      "learning_rate": 3.3432022921493555e-06,
+      "loss": 0.723,
+      "step": 3158
+    },
+    {
+      "epoch": 1.567456869802656,
+      "grad_norm": 0.13401305719243067,
+      "learning_rate": 3.342282146838175e-06,
+      "loss": 0.6941,
+      "step": 3159
+    },
+    {
+      "epoch": 1.5679533325058954,
+      "grad_norm": 0.13781349363786358,
+      "learning_rate": 3.3413618727976718e-06,
+      "loss": 0.7399,
+      "step": 3160
+    },
+    {
+      "epoch": 1.568449795209135,
+      "grad_norm": 0.1374012136941711,
+      "learning_rate": 3.3404414701684966e-06,
+      "loss": 0.6933,
+      "step": 3161
+    },
+    {
+      "epoch": 1.5689462579123743,
+      "grad_norm": 0.1387539946595465,
+      "learning_rate": 3.3395209390913184e-06,
+      "loss": 0.7569,
+      "step": 3162
+    },
+    {
+      "epoch": 1.5694427206156139,
+      "grad_norm": 0.13114107541423803,
+      "learning_rate": 3.338600279706826e-06,
+      "loss": 0.7204,
+      "step": 3163
+    },
+    {
+      "epoch": 1.5699391833188532,
+      "grad_norm": 0.14071169984330964,
+      "learning_rate": 3.3376794921557266e-06,
+      "loss": 0.7178,
+      "step": 3164
+    },
+    {
+      "epoch": 1.5704356460220925,
+      "grad_norm": 0.13592040074130213,
+      "learning_rate": 3.3367585765787476e-06,
+      "loss": 0.688,
+      "step": 3165
+    },
+    {
+      "epoch": 1.5709321087253318,
+      "grad_norm": 0.14138854126933367,
+      "learning_rate": 3.3358375331166364e-06,
+      "loss": 0.7637,
+      "step": 3166
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.13989023590057628,
+      "learning_rate": 3.3349163619101606e-06,
+      "loss": 0.7293,
+      "step": 3167
+    },
+    {
+      "epoch": 1.571925034131811,
+      "grad_norm": 0.14193624544675174,
+      "learning_rate": 3.333995063100105e-06,
+      "loss": 0.6964,
+      "step": 3168
+    },
+    {
+      "epoch": 1.5724214968350503,
+      "grad_norm": 0.12940528095135467,
+      "learning_rate": 3.333073636827277e-06,
+      "loss": 0.7047,
+      "step": 3169
+    },
+    {
+      "epoch": 1.5729179595382896,
+      "grad_norm": 0.13154882534410767,
+      "learning_rate": 3.3321520832325e-06,
+      "loss": 0.7167,
+      "step": 3170
+    },
+    {
+      "epoch": 1.573414422241529,
+      "grad_norm": 0.13169987947643613,
+      "learning_rate": 3.33123040245662e-06,
+      "loss": 0.6984,
+      "step": 3171
+    },
+    {
+      "epoch": 1.5739108849447685,
+      "grad_norm": 0.12901406346373048,
+      "learning_rate": 3.3303085946405006e-06,
+      "loss": 0.7369,
+      "step": 3172
+    },
+    {
+      "epoch": 1.574407347648008,
+      "grad_norm": 0.1390711912220589,
+      "learning_rate": 3.329386659925025e-06,
+      "loss": 0.7334,
+      "step": 3173
+    },
+    {
+      "epoch": 1.5749038103512474,
+      "grad_norm": 0.13614690824410902,
+      "learning_rate": 3.3284645984510977e-06,
+      "loss": 0.7191,
+      "step": 3174
+    },
+    {
+      "epoch": 1.5754002730544867,
+      "grad_norm": 0.1354970816221801,
+      "learning_rate": 3.32754241035964e-06,
+      "loss": 0.7187,
+      "step": 3175
+    },
+    {
+      "epoch": 1.575896735757726,
+      "grad_norm": 0.13457360527562395,
+      "learning_rate": 3.3266200957915927e-06,
+      "loss": 0.713,
+      "step": 3176
+    },
+    {
+      "epoch": 1.5763931984609656,
+      "grad_norm": 0.1348370258972551,
+      "learning_rate": 3.3256976548879183e-06,
+      "loss": 0.7231,
+      "step": 3177
+    },
+    {
+      "epoch": 1.5768896611642051,
+      "grad_norm": 0.13666530116217873,
+      "learning_rate": 3.3247750877895955e-06,
+      "loss": 0.7021,
+      "step": 3178
+    },
+    {
+      "epoch": 1.5773861238674445,
+      "grad_norm": 0.13309005504337973,
+      "learning_rate": 3.3238523946376256e-06,
+      "loss": 0.6904,
+      "step": 3179
+    },
+    {
+      "epoch": 1.5778825865706838,
+      "grad_norm": 0.14080349679221404,
+      "learning_rate": 3.322929575573026e-06,
+      "loss": 0.7235,
+      "step": 3180
+    },
+    {
+      "epoch": 1.5783790492739231,
+      "grad_norm": 0.13251312434856302,
+      "learning_rate": 3.322006630736836e-06,
+      "loss": 0.701,
+      "step": 3181
+    },
+    {
+      "epoch": 1.5788755119771627,
+      "grad_norm": 0.13612624607173604,
+      "learning_rate": 3.321083560270112e-06,
+      "loss": 0.7359,
+      "step": 3182
+    },
+    {
+      "epoch": 1.5793719746804022,
+      "grad_norm": 0.13678666091572295,
+      "learning_rate": 3.3201603643139314e-06,
+      "loss": 0.7353,
+      "step": 3183
+    },
+    {
+      "epoch": 1.5798684373836416,
+      "grad_norm": 0.1369846073186433,
+      "learning_rate": 3.319237043009389e-06,
+      "loss": 0.7178,
+      "step": 3184
+    },
+    {
+      "epoch": 1.580364900086881,
+      "grad_norm": 0.13713704995068668,
+      "learning_rate": 3.3183135964976003e-06,
+      "loss": 0.7221,
+      "step": 3185
+    },
+    {
+      "epoch": 1.5808613627901202,
+      "grad_norm": 0.13486305893810832,
+      "learning_rate": 3.3173900249196986e-06,
+      "loss": 0.7516,
+      "step": 3186
+    },
+    {
+      "epoch": 1.5813578254933598,
+      "grad_norm": 0.13246788290751252,
+      "learning_rate": 3.3164663284168382e-06,
+      "loss": 0.6897,
+      "step": 3187
+    },
+    {
+      "epoch": 1.5818542881965993,
+      "grad_norm": 0.1338514865785925,
+      "learning_rate": 3.3155425071301894e-06,
+      "loss": 0.7198,
+      "step": 3188
+    },
+    {
+      "epoch": 1.5823507508998387,
+      "grad_norm": 0.1408007323666459,
+      "learning_rate": 3.3146185612009453e-06,
+      "loss": 0.695,
+      "step": 3189
+    },
+    {
+      "epoch": 1.582847213603078,
+      "grad_norm": 0.131260440618678,
+      "learning_rate": 3.313694490770316e-06,
+      "loss": 0.7185,
+      "step": 3190
+    },
+    {
+      "epoch": 1.5833436763063173,
+      "grad_norm": 0.13685270748076497,
+      "learning_rate": 3.31277029597953e-06,
+      "loss": 0.7291,
+      "step": 3191
+    },
+    {
+      "epoch": 1.583840139009557,
+      "grad_norm": 0.13649871749420334,
+      "learning_rate": 3.311845976969836e-06,
+      "loss": 0.7442,
+      "step": 3192
+    },
+    {
+      "epoch": 1.5843366017127964,
+      "grad_norm": 0.14178510091564836,
+      "learning_rate": 3.3109215338825008e-06,
+      "loss": 0.7721,
+      "step": 3193
+    },
+    {
+      "epoch": 1.5848330644160358,
+      "grad_norm": 0.13415549660130519,
+      "learning_rate": 3.3099969668588117e-06,
+      "loss": 0.7347,
+      "step": 3194
+    },
+    {
+      "epoch": 1.585329527119275,
+      "grad_norm": 0.1391423380776298,
+      "learning_rate": 3.309072276040074e-06,
+      "loss": 0.7648,
+      "step": 3195
+    },
+    {
+      "epoch": 1.5858259898225144,
+      "grad_norm": 0.13047621822203795,
+      "learning_rate": 3.3081474615676106e-06,
+      "loss": 0.7229,
+      "step": 3196
+    },
+    {
+      "epoch": 1.586322452525754,
+      "grad_norm": 0.14349775839749804,
+      "learning_rate": 3.307222523582766e-06,
+      "loss": 0.7207,
+      "step": 3197
+    },
+    {
+      "epoch": 1.5868189152289935,
+      "grad_norm": 0.1333296259536943,
+      "learning_rate": 3.3062974622269006e-06,
+      "loss": 0.7344,
+      "step": 3198
+    },
+    {
+      "epoch": 1.5873153779322329,
+      "grad_norm": 0.1336681573376393,
+      "learning_rate": 3.3053722776413978e-06,
+      "loss": 0.6781,
+      "step": 3199
+    },
+    {
+      "epoch": 1.5878118406354722,
+      "grad_norm": 0.13336081958471946,
+      "learning_rate": 3.304446969967654e-06,
+      "loss": 0.6875,
+      "step": 3200
+    },
+    {
+      "epoch": 1.5883083033387115,
+      "grad_norm": 0.1301759528310653,
+      "learning_rate": 3.3035215393470896e-06,
+      "loss": 0.6865,
+      "step": 3201
+    },
+    {
+      "epoch": 1.588804766041951,
+      "grad_norm": 0.13281805769994326,
+      "learning_rate": 3.3025959859211416e-06,
+      "loss": 0.7184,
+      "step": 3202
+    },
+    {
+      "epoch": 1.5893012287451906,
+      "grad_norm": 0.1350945524620567,
+      "learning_rate": 3.3016703098312653e-06,
+      "loss": 0.7637,
+      "step": 3203
+    },
+    {
+      "epoch": 1.58979769144843,
+      "grad_norm": 0.13780227036057782,
+      "learning_rate": 3.3007445112189362e-06,
+      "loss": 0.6973,
+      "step": 3204
+    },
+    {
+      "epoch": 1.5902941541516693,
+      "grad_norm": 0.1367630069339855,
+      "learning_rate": 3.2998185902256475e-06,
+      "loss": 0.724,
+      "step": 3205
+    },
+    {
+      "epoch": 1.5907906168549086,
+      "grad_norm": 0.13924121360508623,
+      "learning_rate": 3.298892546992912e-06,
+      "loss": 0.7636,
+      "step": 3206
+    },
+    {
+      "epoch": 1.5912870795581482,
+      "grad_norm": 0.13353110373070412,
+      "learning_rate": 3.29796638166226e-06,
+      "loss": 0.7253,
+      "step": 3207
+    },
+    {
+      "epoch": 1.5917835422613877,
+      "grad_norm": 0.13363257815429952,
+      "learning_rate": 3.29704009437524e-06,
+      "loss": 0.709,
+      "step": 3208
+    },
+    {
+      "epoch": 1.592280004964627,
+      "grad_norm": 0.1323634598863784,
+      "learning_rate": 3.2961136852734215e-06,
+      "loss": 0.7379,
+      "step": 3209
+    },
+    {
+      "epoch": 1.5927764676678664,
+      "grad_norm": 0.1397078632401545,
+      "learning_rate": 3.295187154498391e-06,
+      "loss": 0.7352,
+      "step": 3210
+    },
+    {
+      "epoch": 1.5932729303711057,
+      "grad_norm": 0.13431457529442403,
+      "learning_rate": 3.2942605021917535e-06,
+      "loss": 0.7324,
+      "step": 3211
+    },
+    {
+      "epoch": 1.5937693930743453,
+      "grad_norm": 0.1365499136783772,
+      "learning_rate": 3.2933337284951338e-06,
+      "loss": 0.7463,
+      "step": 3212
+    },
+    {
+      "epoch": 1.5942658557775848,
+      "grad_norm": 0.13865059264043936,
+      "learning_rate": 3.2924068335501734e-06,
+      "loss": 0.7572,
+      "step": 3213
+    },
+    {
+      "epoch": 1.5947623184808242,
+      "grad_norm": 0.13306898436599435,
+      "learning_rate": 3.291479817498534e-06,
+      "loss": 0.6667,
+      "step": 3214
+    },
+    {
+      "epoch": 1.5952587811840635,
+      "grad_norm": 0.12929290122854256,
+      "learning_rate": 3.2905526804818954e-06,
+      "loss": 0.7137,
+      "step": 3215
+    },
+    {
+      "epoch": 1.5957552438873028,
+      "grad_norm": 0.1372438954972776,
+      "learning_rate": 3.2896254226419543e-06,
+      "loss": 0.7392,
+      "step": 3216
+    },
+    {
+      "epoch": 1.5962517065905424,
+      "grad_norm": 0.13554630669851628,
+      "learning_rate": 3.2886980441204287e-06,
+      "loss": 0.7278,
+      "step": 3217
+    },
+    {
+      "epoch": 1.596748169293782,
+      "grad_norm": 0.1304930159730209,
+      "learning_rate": 3.2877705450590525e-06,
+      "loss": 0.6764,
+      "step": 3218
+    },
+    {
+      "epoch": 1.5972446319970213,
+      "grad_norm": 0.13425983285516158,
+      "learning_rate": 3.2868429255995792e-06,
+      "loss": 0.7731,
+      "step": 3219
+    },
+    {
+      "epoch": 1.5977410947002606,
+      "grad_norm": 0.13109350786854748,
+      "learning_rate": 3.2859151858837807e-06,
+      "loss": 0.7339,
+      "step": 3220
+    },
+    {
+      "epoch": 1.5982375574035,
+      "grad_norm": 0.13818419724327707,
+      "learning_rate": 3.2849873260534472e-06,
+      "loss": 0.7421,
+      "step": 3221
+    },
+    {
+      "epoch": 1.5987340201067395,
+      "grad_norm": 0.13351674846911626,
+      "learning_rate": 3.2840593462503878e-06,
+      "loss": 0.7282,
+      "step": 3222
+    },
+    {
+      "epoch": 1.599230482809979,
+      "grad_norm": 0.1375585476076764,
+      "learning_rate": 3.283131246616429e-06,
+      "loss": 0.6996,
+      "step": 3223
+    },
+    {
+      "epoch": 1.5997269455132184,
+      "grad_norm": 0.13392787493113698,
+      "learning_rate": 3.2822030272934154e-06,
+      "loss": 0.7567,
+      "step": 3224
+    },
+    {
+      "epoch": 1.6002234082164577,
+      "grad_norm": 0.1319504916436644,
+      "learning_rate": 3.2812746884232106e-06,
+      "loss": 0.6982,
+      "step": 3225
+    },
+    {
+      "epoch": 1.600719870919697,
+      "grad_norm": 0.13878346494318192,
+      "learning_rate": 3.2803462301476962e-06,
+      "loss": 0.7176,
+      "step": 3226
+    },
+    {
+      "epoch": 1.6012163336229366,
+      "grad_norm": 0.13949329751448514,
+      "learning_rate": 3.2794176526087723e-06,
+      "loss": 0.7316,
+      "step": 3227
+    },
+    {
+      "epoch": 1.6017127963261761,
+      "grad_norm": 0.13219100886136206,
+      "learning_rate": 3.278488955948358e-06,
+      "loss": 0.6974,
+      "step": 3228
+    },
+    {
+      "epoch": 1.6022092590294155,
+      "grad_norm": 0.12840250780955081,
+      "learning_rate": 3.2775601403083886e-06,
+      "loss": 0.7009,
+      "step": 3229
+    },
+    {
+      "epoch": 1.6027057217326548,
+      "grad_norm": 0.13490025484599139,
+      "learning_rate": 3.276631205830818e-06,
+      "loss": 0.6955,
+      "step": 3230
+    },
+    {
+      "epoch": 1.6032021844358941,
+      "grad_norm": 0.1433135689367546,
+      "learning_rate": 3.2757021526576216e-06,
+      "loss": 0.7346,
+      "step": 3231
+    },
+    {
+      "epoch": 1.6036986471391337,
+      "grad_norm": 0.1594064448899915,
+      "learning_rate": 3.2747729809307878e-06,
+      "loss": 0.7767,
+      "step": 3232
+    },
+    {
+      "epoch": 1.6041951098423732,
+      "grad_norm": 0.13512124092823508,
+      "learning_rate": 3.2738436907923267e-06,
+      "loss": 0.732,
+      "step": 3233
+    },
+    {
+      "epoch": 1.6046915725456126,
+      "grad_norm": 0.13435293487480612,
+      "learning_rate": 3.2729142823842645e-06,
+      "loss": 0.7113,
+      "step": 3234
+    },
+    {
+      "epoch": 1.605188035248852,
+      "grad_norm": 0.1319079246448511,
+      "learning_rate": 3.2719847558486474e-06,
+      "loss": 0.7377,
+      "step": 3235
+    },
+    {
+      "epoch": 1.6056844979520912,
+      "grad_norm": 0.1279027378230854,
+      "learning_rate": 3.271055111327538e-06,
+      "loss": 0.7124,
+      "step": 3236
+    },
+    {
+      "epoch": 1.6061809606553308,
+      "grad_norm": 0.13463254886388712,
+      "learning_rate": 3.2701253489630175e-06,
+      "loss": 0.699,
+      "step": 3237
+    },
+    {
+      "epoch": 1.6066774233585703,
+      "grad_norm": 0.1366952695028082,
+      "learning_rate": 3.2691954688971862e-06,
+      "loss": 0.7035,
+      "step": 3238
+    },
+    {
+      "epoch": 1.6071738860618097,
+      "grad_norm": 0.13669325319793554,
+      "learning_rate": 3.2682654712721605e-06,
+      "loss": 0.7521,
+      "step": 3239
+    },
+    {
+      "epoch": 1.607670348765049,
+      "grad_norm": 0.13083810989888964,
+      "learning_rate": 3.2673353562300753e-06,
+      "loss": 0.7283,
+      "step": 3240
+    },
+    {
+      "epoch": 1.6081668114682883,
+      "grad_norm": 0.14043885513177584,
+      "learning_rate": 3.266405123913084e-06,
+      "loss": 0.7589,
+      "step": 3241
+    },
+    {
+      "epoch": 1.6086632741715279,
+      "grad_norm": 0.13900658899664461,
+      "learning_rate": 3.265474774463358e-06,
+      "loss": 0.7308,
+      "step": 3242
+    },
+    {
+      "epoch": 1.6091597368747674,
+      "grad_norm": 0.14121156080517872,
+      "learning_rate": 3.264544308023086e-06,
+      "loss": 0.7011,
+      "step": 3243
+    },
+    {
+      "epoch": 1.6096561995780068,
+      "grad_norm": 0.13303533153675115,
+      "learning_rate": 3.2636137247344756e-06,
+      "loss": 0.6789,
+      "step": 3244
+    },
+    {
+      "epoch": 1.610152662281246,
+      "grad_norm": 0.1345546966620587,
+      "learning_rate": 3.2626830247397507e-06,
+      "loss": 0.7209,
+      "step": 3245
+    },
+    {
+      "epoch": 1.6106491249844854,
+      "grad_norm": 0.13372939850048826,
+      "learning_rate": 3.261752208181154e-06,
+      "loss": 0.7131,
+      "step": 3246
+    },
+    {
+      "epoch": 1.611145587687725,
+      "grad_norm": 0.13637214131020378,
+      "learning_rate": 3.260821275200947e-06,
+      "loss": 0.7006,
+      "step": 3247
+    },
+    {
+      "epoch": 1.6116420503909645,
+      "grad_norm": 0.133959325485992,
+      "learning_rate": 3.2598902259414055e-06,
+      "loss": 0.71,
+      "step": 3248
+    },
+    {
+      "epoch": 1.6121385130942039,
+      "grad_norm": 0.1317569091674975,
+      "learning_rate": 3.258959060544828e-06,
+      "loss": 0.6934,
+      "step": 3249
+    },
+    {
+      "epoch": 1.6126349757974432,
+      "grad_norm": 0.13549906434028006,
+      "learning_rate": 3.258027779153527e-06,
+      "loss": 0.7516,
+      "step": 3250
+    },
+    {
+      "epoch": 1.6131314385006825,
+      "grad_norm": 0.13646477270544233,
+      "learning_rate": 3.2570963819098333e-06,
+      "loss": 0.7309,
+      "step": 3251
+    },
+    {
+      "epoch": 1.613627901203922,
+      "grad_norm": 0.13531221765549428,
+      "learning_rate": 3.2561648689560972e-06,
+      "loss": 0.7276,
+      "step": 3252
+    },
+    {
+      "epoch": 1.6141243639071616,
+      "grad_norm": 0.13706437632515092,
+      "learning_rate": 3.2552332404346847e-06,
+      "loss": 0.7446,
+      "step": 3253
+    },
+    {
+      "epoch": 1.614620826610401,
+      "grad_norm": 0.12982692323683104,
+      "learning_rate": 3.2543014964879814e-06,
+      "loss": 0.6706,
+      "step": 3254
+    },
+    {
+      "epoch": 1.6151172893136403,
+      "grad_norm": 0.1417142856105007,
+      "learning_rate": 3.2533696372583886e-06,
+      "loss": 0.7691,
+      "step": 3255
+    },
+    {
+      "epoch": 1.6156137520168796,
+      "grad_norm": 0.1338299722054298,
+      "learning_rate": 3.2524376628883254e-06,
+      "loss": 0.7318,
+      "step": 3256
+    },
+    {
+      "epoch": 1.6161102147201192,
+      "grad_norm": 0.13244784626065542,
+      "learning_rate": 3.25150557352023e-06,
+      "loss": 0.6839,
+      "step": 3257
+    },
+    {
+      "epoch": 1.6166066774233587,
+      "grad_norm": 0.1373652541161793,
+      "learning_rate": 3.250573369296557e-06,
+      "loss": 0.7263,
+      "step": 3258
+    },
+    {
+      "epoch": 1.617103140126598,
+      "grad_norm": 0.13733039620980356,
+      "learning_rate": 3.249641050359779e-06,
+      "loss": 0.7403,
+      "step": 3259
+    },
+    {
+      "epoch": 1.6175996028298374,
+      "grad_norm": 0.14578749260154306,
+      "learning_rate": 3.248708616852387e-06,
+      "loss": 0.789,
+      "step": 3260
+    },
+    {
+      "epoch": 1.6180960655330767,
+      "grad_norm": 0.137192295052107,
+      "learning_rate": 3.247776068916887e-06,
+      "loss": 0.7158,
+      "step": 3261
+    },
+    {
+      "epoch": 1.6185925282363163,
+      "grad_norm": 0.1343518895157817,
+      "learning_rate": 3.246843406695804e-06,
+      "loss": 0.6879,
+      "step": 3262
+    },
+    {
+      "epoch": 1.6190889909395558,
+      "grad_norm": 0.13371425918634516,
+      "learning_rate": 3.245910630331682e-06,
+      "loss": 0.7908,
+      "step": 3263
+    },
+    {
+      "epoch": 1.6195854536427952,
+      "grad_norm": 0.13354797872889265,
+      "learning_rate": 3.2449777399670788e-06,
+      "loss": 0.6948,
+      "step": 3264
+    },
+    {
+      "epoch": 1.6200819163460345,
+      "grad_norm": 0.13151382649489246,
+      "learning_rate": 3.2440447357445737e-06,
+      "loss": 0.7567,
+      "step": 3265
+    },
+    {
+      "epoch": 1.6205783790492738,
+      "grad_norm": 0.13595368534380006,
+      "learning_rate": 3.243111617806761e-06,
+      "loss": 0.7589,
+      "step": 3266
+    },
+    {
+      "epoch": 1.6210748417525134,
+      "grad_norm": 0.1365032790773031,
+      "learning_rate": 3.2421783862962513e-06,
+      "loss": 0.7405,
+      "step": 3267
+    },
+    {
+      "epoch": 1.621571304455753,
+      "grad_norm": 0.13435638512861106,
+      "learning_rate": 3.2412450413556753e-06,
+      "loss": 0.7066,
+      "step": 3268
+    },
+    {
+      "epoch": 1.6220677671589923,
+      "grad_norm": 0.13947202661774769,
+      "learning_rate": 3.24031158312768e-06,
+      "loss": 0.7538,
+      "step": 3269
+    },
+    {
+      "epoch": 1.6225642298622316,
+      "grad_norm": 0.13439183598377127,
+      "learning_rate": 3.2393780117549294e-06,
+      "loss": 0.7072,
+      "step": 3270
+    },
+    {
+      "epoch": 1.623060692565471,
+      "grad_norm": 0.13125548459747396,
+      "learning_rate": 3.238444327380105e-06,
+      "loss": 0.7033,
+      "step": 3271
+    },
+    {
+      "epoch": 1.6235571552687105,
+      "grad_norm": 0.13335877452614855,
+      "learning_rate": 3.2375105301459046e-06,
+      "loss": 0.7374,
+      "step": 3272
+    },
+    {
+      "epoch": 1.6240536179719498,
+      "grad_norm": 0.13797996283169242,
+      "learning_rate": 3.236576620195045e-06,
+      "loss": 0.7282,
+      "step": 3273
+    },
+    {
+      "epoch": 1.6245500806751894,
+      "grad_norm": 0.1346903208785757,
+      "learning_rate": 3.2356425976702587e-06,
+      "loss": 0.7005,
+      "step": 3274
+    },
+    {
+      "epoch": 1.6250465433784287,
+      "grad_norm": 0.13164913588291022,
+      "learning_rate": 3.2347084627142967e-06,
+      "loss": 0.712,
+      "step": 3275
+    },
+    {
+      "epoch": 1.625543006081668,
+      "grad_norm": 0.13395069471794036,
+      "learning_rate": 3.233774215469927e-06,
+      "loss": 0.6553,
+      "step": 3276
+    },
+    {
+      "epoch": 1.6260394687849076,
+      "grad_norm": 0.12946118362950973,
+      "learning_rate": 3.2328398560799327e-06,
+      "loss": 0.7068,
+      "step": 3277
+    },
+    {
+      "epoch": 1.626535931488147,
+      "grad_norm": 0.13651890789151064,
+      "learning_rate": 3.231905384687117e-06,
+      "loss": 0.7263,
+      "step": 3278
+    },
+    {
+      "epoch": 1.6270323941913865,
+      "grad_norm": 0.13320408589478813,
+      "learning_rate": 3.2309708014342987e-06,
+      "loss": 0.659,
+      "step": 3279
+    },
+    {
+      "epoch": 1.6275288568946258,
+      "grad_norm": 0.13232211632916635,
+      "learning_rate": 3.2300361064643133e-06,
+      "loss": 0.695,
+      "step": 3280
+    },
+    {
+      "epoch": 1.6280253195978651,
+      "grad_norm": 0.13567576909886284,
+      "learning_rate": 3.229101299920014e-06,
+      "loss": 0.7111,
+      "step": 3281
+    },
+    {
+      "epoch": 1.6285217823011047,
+      "grad_norm": 0.13518794018443017,
+      "learning_rate": 3.228166381944272e-06,
+      "loss": 0.7415,
+      "step": 3282
+    },
+    {
+      "epoch": 1.629018245004344,
+      "grad_norm": 0.13083493674868937,
+      "learning_rate": 3.227231352679973e-06,
+      "loss": 0.7146,
+      "step": 3283
+    },
+    {
+      "epoch": 1.6295147077075836,
+      "grad_norm": 0.14297032801648918,
+      "learning_rate": 3.2262962122700224e-06,
+      "loss": 0.7503,
+      "step": 3284
+    },
+    {
+      "epoch": 1.6300111704108229,
+      "grad_norm": 0.13219562864852205,
+      "learning_rate": 3.2253609608573412e-06,
+      "loss": 0.7098,
+      "step": 3285
+    },
+    {
+      "epoch": 1.6305076331140622,
+      "grad_norm": 0.1356006220123195,
+      "learning_rate": 3.2244255985848664e-06,
+      "loss": 0.7158,
+      "step": 3286
+    },
+    {
+      "epoch": 1.6310040958173018,
+      "grad_norm": 0.13516073020936664,
+      "learning_rate": 3.2234901255955554e-06,
+      "loss": 0.7443,
+      "step": 3287
+    },
+    {
+      "epoch": 1.631500558520541,
+      "grad_norm": 0.13223315868550364,
+      "learning_rate": 3.222554542032379e-06,
+      "loss": 0.7013,
+      "step": 3288
+    },
+    {
+      "epoch": 1.6319970212237807,
+      "grad_norm": 0.1291224602102518,
+      "learning_rate": 3.221618848038326e-06,
+      "loss": 0.725,
+      "step": 3289
+    },
+    {
+      "epoch": 1.63249348392702,
+      "grad_norm": 0.13959827562469576,
+      "learning_rate": 3.220683043756402e-06,
+      "loss": 0.7296,
+      "step": 3290
+    },
+    {
+      "epoch": 1.6329899466302593,
+      "grad_norm": 0.14159509303723317,
+      "learning_rate": 3.2197471293296297e-06,
+      "loss": 0.7486,
+      "step": 3291
+    },
+    {
+      "epoch": 1.6334864093334989,
+      "grad_norm": 0.13293711515989248,
+      "learning_rate": 3.2188111049010496e-06,
+      "loss": 0.7258,
+      "step": 3292
+    },
+    {
+      "epoch": 1.6339828720367382,
+      "grad_norm": 0.1399399476886841,
+      "learning_rate": 3.217874970613717e-06,
+      "loss": 0.7762,
+      "step": 3293
+    },
+    {
+      "epoch": 1.6344793347399778,
+      "grad_norm": 0.13066912497887148,
+      "learning_rate": 3.216938726610705e-06,
+      "loss": 0.682,
+      "step": 3294
+    },
+    {
+      "epoch": 1.634975797443217,
+      "grad_norm": 0.13206917222232278,
+      "learning_rate": 3.2160023730351047e-06,
+      "loss": 0.6876,
+      "step": 3295
+    },
+    {
+      "epoch": 1.6354722601464564,
+      "grad_norm": 0.12881195914477797,
+      "learning_rate": 3.2150659100300215e-06,
+      "loss": 0.7009,
+      "step": 3296
+    },
+    {
+      "epoch": 1.635968722849696,
+      "grad_norm": 0.1354978102327594,
+      "learning_rate": 3.214129337738579e-06,
+      "loss": 0.7216,
+      "step": 3297
+    },
+    {
+      "epoch": 1.6364651855529353,
+      "grad_norm": 0.13272278700868193,
+      "learning_rate": 3.2131926563039184e-06,
+      "loss": 0.7671,
+      "step": 3298
+    },
+    {
+      "epoch": 1.6369616482561749,
+      "grad_norm": 0.13268702786871522,
+      "learning_rate": 3.2122558658691943e-06,
+      "loss": 0.7306,
+      "step": 3299
+    },
+    {
+      "epoch": 1.6374581109594142,
+      "grad_norm": 0.14693293257842704,
+      "learning_rate": 3.2113189665775812e-06,
+      "loss": 0.7262,
+      "step": 3300
+    },
+    {
+      "epoch": 1.6379545736626535,
+      "grad_norm": 0.13647387480975237,
+      "learning_rate": 3.2103819585722696e-06,
+      "loss": 0.7228,
+      "step": 3301
+    },
+    {
+      "epoch": 1.638451036365893,
+      "grad_norm": 0.13489368002148763,
+      "learning_rate": 3.2094448419964657e-06,
+      "loss": 0.72,
+      "step": 3302
+    },
+    {
+      "epoch": 1.6389474990691324,
+      "grad_norm": 0.1334332663332396,
+      "learning_rate": 3.208507616993393e-06,
+      "loss": 0.6933,
+      "step": 3303
+    },
+    {
+      "epoch": 1.639443961772372,
+      "grad_norm": 0.13159985839475669,
+      "learning_rate": 3.2075702837062915e-06,
+      "loss": 0.6915,
+      "step": 3304
+    },
+    {
+      "epoch": 1.6399404244756113,
+      "grad_norm": 0.13257761470120857,
+      "learning_rate": 3.2066328422784166e-06,
+      "loss": 0.733,
+      "step": 3305
+    },
+    {
+      "epoch": 1.6404368871788506,
+      "grad_norm": 0.13649241457619013,
+      "learning_rate": 3.2056952928530417e-06,
+      "loss": 0.7765,
+      "step": 3306
+    },
+    {
+      "epoch": 1.64093334988209,
+      "grad_norm": 0.12816321098365577,
+      "learning_rate": 3.204757635573456e-06,
+      "loss": 0.6676,
+      "step": 3307
+    },
+    {
+      "epoch": 1.6414298125853295,
+      "grad_norm": 0.1357189860674346,
+      "learning_rate": 3.2038198705829658e-06,
+      "loss": 0.7106,
+      "step": 3308
+    },
+    {
+      "epoch": 1.641926275288569,
+      "grad_norm": 0.12905046251866562,
+      "learning_rate": 3.202881998024894e-06,
+      "loss": 0.6678,
+      "step": 3309
+    },
+    {
+      "epoch": 1.6424227379918084,
+      "grad_norm": 0.13424725400492285,
+      "learning_rate": 3.2019440180425774e-06,
+      "loss": 0.6833,
+      "step": 3310
+    },
+    {
+      "epoch": 1.6429192006950477,
+      "grad_norm": 0.1310598204136392,
+      "learning_rate": 3.201005930779374e-06,
+      "loss": 0.7341,
+      "step": 3311
+    },
+    {
+      "epoch": 1.643415663398287,
+      "grad_norm": 0.14192929179095823,
+      "learning_rate": 3.2000677363786525e-06,
+      "loss": 0.6982,
+      "step": 3312
+    },
+    {
+      "epoch": 1.6439121261015266,
+      "grad_norm": 0.12782211475807093,
+      "learning_rate": 3.199129434983803e-06,
+      "loss": 0.6782,
+      "step": 3313
+    },
+    {
+      "epoch": 1.6444085888047661,
+      "grad_norm": 0.13517571601842548,
+      "learning_rate": 3.1981910267382294e-06,
+      "loss": 0.7526,
+      "step": 3314
+    },
+    {
+      "epoch": 1.6449050515080055,
+      "grad_norm": 0.13283442693929196,
+      "learning_rate": 3.197252511785351e-06,
+      "loss": 0.7385,
+      "step": 3315
+    },
+    {
+      "epoch": 1.6454015142112448,
+      "grad_norm": 0.129952438407659,
+      "learning_rate": 3.196313890268606e-06,
+      "loss": 0.6833,
+      "step": 3316
+    },
+    {
+      "epoch": 1.6458979769144841,
+      "grad_norm": 0.13333153615133936,
+      "learning_rate": 3.1953751623314476e-06,
+      "loss": 0.7165,
+      "step": 3317
+    },
+    {
+      "epoch": 1.6463944396177237,
+      "grad_norm": 0.139666486016696,
+      "learning_rate": 3.1944363281173445e-06,
+      "loss": 0.7806,
+      "step": 3318
+    },
+    {
+      "epoch": 1.6468909023209632,
+      "grad_norm": 0.13280138772226624,
+      "learning_rate": 3.1934973877697843e-06,
+      "loss": 0.7155,
+      "step": 3319
+    },
+    {
+      "epoch": 1.6473873650242026,
+      "grad_norm": 0.13553005489923173,
+      "learning_rate": 3.1925583414322677e-06,
+      "loss": 0.7518,
+      "step": 3320
+    },
+    {
+      "epoch": 1.647883827727442,
+      "grad_norm": 0.13507802161305021,
+      "learning_rate": 3.191619189248312e-06,
+      "loss": 0.7639,
+      "step": 3321
+    },
+    {
+      "epoch": 1.6483802904306812,
+      "grad_norm": 0.13564619513742512,
+      "learning_rate": 3.1906799313614526e-06,
+      "loss": 0.7453,
+      "step": 3322
+    },
+    {
+      "epoch": 1.6488767531339208,
+      "grad_norm": 0.13874442942300055,
+      "learning_rate": 3.1897405679152403e-06,
+      "loss": 0.6741,
+      "step": 3323
+    },
+    {
+      "epoch": 1.6493732158371603,
+      "grad_norm": 0.14111170898428813,
+      "learning_rate": 3.1888010990532412e-06,
+      "loss": 0.7146,
+      "step": 3324
+    },
+    {
+      "epoch": 1.6498696785403997,
+      "grad_norm": 0.1284924018943679,
+      "learning_rate": 3.1878615249190386e-06,
+      "loss": 0.6811,
+      "step": 3325
+    },
+    {
+      "epoch": 1.650366141243639,
+      "grad_norm": 0.14995165389634305,
+      "learning_rate": 3.18692184565623e-06,
+      "loss": 0.6953,
+      "step": 3326
+    },
+    {
+      "epoch": 1.6508626039468783,
+      "grad_norm": 0.13386502577631518,
+      "learning_rate": 3.1859820614084326e-06,
+      "loss": 0.7294,
+      "step": 3327
+    },
+    {
+      "epoch": 1.651359066650118,
+      "grad_norm": 0.13753075751861904,
+      "learning_rate": 3.1850421723192748e-06,
+      "loss": 0.7097,
+      "step": 3328
+    },
+    {
+      "epoch": 1.6518555293533574,
+      "grad_norm": 0.1372523640506848,
+      "learning_rate": 3.1841021785324057e-06,
+      "loss": 0.7533,
+      "step": 3329
+    },
+    {
+      "epoch": 1.6523519920565968,
+      "grad_norm": 0.13132741612801554,
+      "learning_rate": 3.183162080191487e-06,
+      "loss": 0.7468,
+      "step": 3330
+    },
+    {
+      "epoch": 1.652848454759836,
+      "grad_norm": 0.14093059894911422,
+      "learning_rate": 3.1822218774401974e-06,
+      "loss": 0.7054,
+      "step": 3331
+    },
+    {
+      "epoch": 1.6533449174630754,
+      "grad_norm": 0.13503393548333384,
+      "learning_rate": 3.1812815704222326e-06,
+      "loss": 0.7092,
+      "step": 3332
+    },
+    {
+      "epoch": 1.653841380166315,
+      "grad_norm": 0.13795156477927883,
+      "learning_rate": 3.1803411592813036e-06,
+      "loss": 0.7342,
+      "step": 3333
+    },
+    {
+      "epoch": 1.6543378428695545,
+      "grad_norm": 0.14715484138992083,
+      "learning_rate": 3.179400644161137e-06,
+      "loss": 0.7639,
+      "step": 3334
+    },
+    {
+      "epoch": 1.6548343055727939,
+      "grad_norm": 0.13505583893408066,
+      "learning_rate": 3.1784600252054752e-06,
+      "loss": 0.7639,
+      "step": 3335
+    },
+    {
+      "epoch": 1.6553307682760332,
+      "grad_norm": 0.13398674733013766,
+      "learning_rate": 3.1775193025580774e-06,
+      "loss": 0.7302,
+      "step": 3336
+    },
+    {
+      "epoch": 1.6558272309792725,
+      "grad_norm": 0.13122340668034163,
+      "learning_rate": 3.176578476362716e-06,
+      "loss": 0.6994,
+      "step": 3337
+    },
+    {
+      "epoch": 1.656323693682512,
+      "grad_norm": 0.1392905054739376,
+      "learning_rate": 3.1756375467631832e-06,
+      "loss": 0.743,
+      "step": 3338
+    },
+    {
+      "epoch": 1.6568201563857516,
+      "grad_norm": 0.13237002846348273,
+      "learning_rate": 3.1746965139032846e-06,
+      "loss": 0.7088,
+      "step": 3339
+    },
+    {
+      "epoch": 1.657316619088991,
+      "grad_norm": 0.1342680889751068,
+      "learning_rate": 3.1737553779268415e-06,
+      "loss": 0.742,
+      "step": 3340
+    },
+    {
+      "epoch": 1.6578130817922303,
+      "grad_norm": 0.13437620364569894,
+      "learning_rate": 3.172814138977692e-06,
+      "loss": 0.7017,
+      "step": 3341
+    },
+    {
+      "epoch": 1.6583095444954696,
+      "grad_norm": 0.13152990564641498,
+      "learning_rate": 3.1718727971996893e-06,
+      "loss": 0.6658,
+      "step": 3342
+    },
+    {
+      "epoch": 1.6588060071987092,
+      "grad_norm": 0.13578569658737435,
+      "learning_rate": 3.1709313527367025e-06,
+      "loss": 0.7434,
+      "step": 3343
+    },
+    {
+      "epoch": 1.6593024699019487,
+      "grad_norm": 0.13282327497087398,
+      "learning_rate": 3.1699898057326156e-06,
+      "loss": 0.684,
+      "step": 3344
+    },
+    {
+      "epoch": 1.659798932605188,
+      "grad_norm": 0.1266967797135504,
+      "learning_rate": 3.169048156331329e-06,
+      "loss": 0.7131,
+      "step": 3345
+    },
+    {
+      "epoch": 1.6602953953084274,
+      "grad_norm": 0.14302577618653753,
+      "learning_rate": 3.168106404676759e-06,
+      "loss": 0.7456,
+      "step": 3346
+    },
+    {
+      "epoch": 1.6607918580116667,
+      "grad_norm": 0.1366996921041571,
+      "learning_rate": 3.167164550912838e-06,
+      "loss": 0.7763,
+      "step": 3347
+    },
+    {
+      "epoch": 1.6612883207149063,
+      "grad_norm": 0.13073297302259157,
+      "learning_rate": 3.1662225951835124e-06,
+      "loss": 0.7044,
+      "step": 3348
+    },
+    {
+      "epoch": 1.6617847834181458,
+      "grad_norm": 0.12656041564587794,
+      "learning_rate": 3.1652805376327446e-06,
+      "loss": 0.6735,
+      "step": 3349
+    },
+    {
+      "epoch": 1.6622812461213852,
+      "grad_norm": 0.13902637524276387,
+      "learning_rate": 3.164338378404514e-06,
+      "loss": 0.7789,
+      "step": 3350
+    },
+    {
+      "epoch": 1.6627777088246245,
+      "grad_norm": 0.13433714640026365,
+      "learning_rate": 3.163396117642814e-06,
+      "loss": 0.7237,
+      "step": 3351
+    },
+    {
+      "epoch": 1.6632741715278638,
+      "grad_norm": 0.13306259899090125,
+      "learning_rate": 3.162453755491655e-06,
+      "loss": 0.6549,
+      "step": 3352
+    },
+    {
+      "epoch": 1.6637706342311034,
+      "grad_norm": 0.14000124931833177,
+      "learning_rate": 3.1615112920950604e-06,
+      "loss": 0.7011,
+      "step": 3353
+    },
+    {
+      "epoch": 1.664267096934343,
+      "grad_norm": 0.13277684994304378,
+      "learning_rate": 3.160568727597071e-06,
+      "loss": 0.7056,
+      "step": 3354
+    },
+    {
+      "epoch": 1.6647635596375823,
+      "grad_norm": 0.13574274974818243,
+      "learning_rate": 3.159626062141743e-06,
+      "loss": 0.6961,
+      "step": 3355
+    },
+    {
+      "epoch": 1.6652600223408216,
+      "grad_norm": 0.13515486910318286,
+      "learning_rate": 3.158683295873148e-06,
+      "loss": 0.7505,
+      "step": 3356
+    },
+    {
+      "epoch": 1.665756485044061,
+      "grad_norm": 0.13117466652194615,
+      "learning_rate": 3.157740428935373e-06,
+      "loss": 0.6759,
+      "step": 3357
+    },
+    {
+      "epoch": 1.6662529477473005,
+      "grad_norm": 0.13734801985031647,
+      "learning_rate": 3.156797461472518e-06,
+      "loss": 0.7327,
+      "step": 3358
+    },
+    {
+      "epoch": 1.66674941045054,
+      "grad_norm": 0.12921730687145758,
+      "learning_rate": 3.155854393628704e-06,
+      "loss": 0.7142,
+      "step": 3359
+    },
+    {
+      "epoch": 1.6672458731537794,
+      "grad_norm": 0.13410715617644764,
+      "learning_rate": 3.15491122554806e-06,
+      "loss": 0.6928,
+      "step": 3360
+    },
+    {
+      "epoch": 1.6677423358570187,
+      "grad_norm": 0.13115756337495957,
+      "learning_rate": 3.1539679573747366e-06,
+      "loss": 0.7292,
+      "step": 3361
+    },
+    {
+      "epoch": 1.668238798560258,
+      "grad_norm": 0.13270364362867004,
+      "learning_rate": 3.1530245892528964e-06,
+      "loss": 0.6981,
+      "step": 3362
+    },
+    {
+      "epoch": 1.6687352612634976,
+      "grad_norm": 0.1312122707358505,
+      "learning_rate": 3.1520811213267187e-06,
+      "loss": 0.7039,
+      "step": 3363
+    },
+    {
+      "epoch": 1.6692317239667371,
+      "grad_norm": 0.13612808788926317,
+      "learning_rate": 3.151137553740396e-06,
+      "loss": 0.7128,
+      "step": 3364
+    },
+    {
+      "epoch": 1.6697281866699765,
+      "grad_norm": 0.1356721052844853,
+      "learning_rate": 3.150193886638139e-06,
+      "loss": 0.7021,
+      "step": 3365
+    },
+    {
+      "epoch": 1.6702246493732158,
+      "grad_norm": 0.1334498879831998,
+      "learning_rate": 3.149250120164171e-06,
+      "loss": 0.6959,
+      "step": 3366
+    },
+    {
+      "epoch": 1.6707211120764551,
+      "grad_norm": 0.12993420392929775,
+      "learning_rate": 3.148306254462733e-06,
+      "loss": 0.7035,
+      "step": 3367
+    },
+    {
+      "epoch": 1.6712175747796947,
+      "grad_norm": 0.12953639610818973,
+      "learning_rate": 3.1473622896780787e-06,
+      "loss": 0.6884,
+      "step": 3368
+    },
+    {
+      "epoch": 1.6717140374829342,
+      "grad_norm": 0.13434437759235057,
+      "learning_rate": 3.1464182259544774e-06,
+      "loss": 0.7411,
+      "step": 3369
+    },
+    {
+      "epoch": 1.6722105001861736,
+      "grad_norm": 0.1351221908451605,
+      "learning_rate": 3.1454740634362146e-06,
+      "loss": 0.7518,
+      "step": 3370
+    },
+    {
+      "epoch": 1.672706962889413,
+      "grad_norm": 0.13151338326848652,
+      "learning_rate": 3.144529802267591e-06,
+      "loss": 0.7119,
+      "step": 3371
+    },
+    {
+      "epoch": 1.6732034255926522,
+      "grad_norm": 0.1330736698831485,
+      "learning_rate": 3.1435854425929207e-06,
+      "loss": 0.7364,
+      "step": 3372
+    },
+    {
+      "epoch": 1.6736998882958918,
+      "grad_norm": 0.13120786063832604,
+      "learning_rate": 3.142640984556536e-06,
+      "loss": 0.6817,
+      "step": 3373
+    },
+    {
+      "epoch": 1.6741963509991313,
+      "grad_norm": 0.13199790092665806,
+      "learning_rate": 3.1416964283027796e-06,
+      "loss": 0.6751,
+      "step": 3374
+    },
+    {
+      "epoch": 1.6746928137023707,
+      "grad_norm": 0.13247768384510822,
+      "learning_rate": 3.1407517739760135e-06,
+      "loss": 0.7246,
+      "step": 3375
+    },
+    {
+      "epoch": 1.67518927640561,
+      "grad_norm": 0.15212250506043967,
+      "learning_rate": 3.139807021720613e-06,
+      "loss": 0.7784,
+      "step": 3376
+    },
+    {
+      "epoch": 1.6756857391088493,
+      "grad_norm": 0.13118756683065375,
+      "learning_rate": 3.1388621716809664e-06,
+      "loss": 0.6729,
+      "step": 3377
+    },
+    {
+      "epoch": 1.6761822018120889,
+      "grad_norm": 0.1288392412087169,
+      "learning_rate": 3.137917224001481e-06,
+      "loss": 0.6989,
+      "step": 3378
+    },
+    {
+      "epoch": 1.6766786645153284,
+      "grad_norm": 0.1315769406124471,
+      "learning_rate": 3.136972178826576e-06,
+      "loss": 0.6975,
+      "step": 3379
+    },
+    {
+      "epoch": 1.6771751272185678,
+      "grad_norm": 0.13147429747221198,
+      "learning_rate": 3.136027036300687e-06,
+      "loss": 0.733,
+      "step": 3380
+    },
+    {
+      "epoch": 1.677671589921807,
+      "grad_norm": 0.12989489332738152,
+      "learning_rate": 3.1350817965682624e-06,
+      "loss": 0.7395,
+      "step": 3381
+    },
+    {
+      "epoch": 1.6781680526250464,
+      "grad_norm": 0.1352912050224614,
+      "learning_rate": 3.1341364597737684e-06,
+      "loss": 0.7523,
+      "step": 3382
+    },
+    {
+      "epoch": 1.678664515328286,
+      "grad_norm": 0.13840497131463406,
+      "learning_rate": 3.1331910260616845e-06,
+      "loss": 0.7422,
+      "step": 3383
+    },
+    {
+      "epoch": 1.6791609780315255,
+      "grad_norm": 0.132998592069353,
+      "learning_rate": 3.132245495576505e-06,
+      "loss": 0.7312,
+      "step": 3384
+    },
+    {
+      "epoch": 1.6796574407347649,
+      "grad_norm": 0.12905667482053407,
+      "learning_rate": 3.1312998684627383e-06,
+      "loss": 0.721,
+      "step": 3385
+    },
+    {
+      "epoch": 1.6801539034380042,
+      "grad_norm": 0.1399236151235144,
+      "learning_rate": 3.130354144864909e-06,
+      "loss": 0.733,
+      "step": 3386
+    },
+    {
+      "epoch": 1.6806503661412435,
+      "grad_norm": 0.13157662254402422,
+      "learning_rate": 3.1294083249275546e-06,
+      "loss": 0.727,
+      "step": 3387
+    },
+    {
+      "epoch": 1.681146828844483,
+      "grad_norm": 0.13230069517151366,
+      "learning_rate": 3.1284624087952307e-06,
+      "loss": 0.7103,
+      "step": 3388
+    },
+    {
+      "epoch": 1.6816432915477226,
+      "grad_norm": 0.134244822451022,
+      "learning_rate": 3.1275163966125042e-06,
+      "loss": 0.7265,
+      "step": 3389
+    },
+    {
+      "epoch": 1.682139754250962,
+      "grad_norm": 0.134846550069988,
+      "learning_rate": 3.1265702885239573e-06,
+      "loss": 0.6948,
+      "step": 3390
+    },
+    {
+      "epoch": 1.6826362169542013,
+      "grad_norm": 0.1312183196880194,
+      "learning_rate": 3.125624084674188e-06,
+      "loss": 0.7065,
+      "step": 3391
+    },
+    {
+      "epoch": 1.6831326796574406,
+      "grad_norm": 0.13279036282716666,
+      "learning_rate": 3.12467778520781e-06,
+      "loss": 0.7109,
+      "step": 3392
+    },
+    {
+      "epoch": 1.6836291423606802,
+      "grad_norm": 0.13146686432062818,
+      "learning_rate": 3.1237313902694467e-06,
+      "loss": 0.6808,
+      "step": 3393
+    },
+    {
+      "epoch": 1.6841256050639197,
+      "grad_norm": 0.13576130928931848,
+      "learning_rate": 3.122784900003742e-06,
+      "loss": 0.7479,
+      "step": 3394
+    },
+    {
+      "epoch": 1.684622067767159,
+      "grad_norm": 0.1337831193034485,
+      "learning_rate": 3.121838314555351e-06,
+      "loss": 0.8127,
+      "step": 3395
+    },
+    {
+      "epoch": 1.6851185304703984,
+      "grad_norm": 0.1388930824488273,
+      "learning_rate": 3.120891634068944e-06,
+      "loss": 0.7343,
+      "step": 3396
+    },
+    {
+      "epoch": 1.6856149931736377,
+      "grad_norm": 0.13614435991184165,
+      "learning_rate": 3.1199448586892056e-06,
+      "loss": 0.7719,
+      "step": 3397
+    },
+    {
+      "epoch": 1.6861114558768773,
+      "grad_norm": 0.1361801580234787,
+      "learning_rate": 3.1189979885608358e-06,
+      "loss": 0.7374,
+      "step": 3398
+    },
+    {
+      "epoch": 1.6866079185801168,
+      "grad_norm": 0.13142547998982623,
+      "learning_rate": 3.118051023828548e-06,
+      "loss": 0.738,
+      "step": 3399
+    },
+    {
+      "epoch": 1.6871043812833562,
+      "grad_norm": 0.1357277571483877,
+      "learning_rate": 3.1171039646370714e-06,
+      "loss": 0.7191,
+      "step": 3400
+    },
+    {
+      "epoch": 1.6876008439865955,
+      "grad_norm": 0.12800485874343553,
+      "learning_rate": 3.1161568111311487e-06,
+      "loss": 0.7144,
+      "step": 3401
+    },
+    {
+      "epoch": 1.6880973066898348,
+      "grad_norm": 0.12514438054459623,
+      "learning_rate": 3.115209563455536e-06,
+      "loss": 0.6717,
+      "step": 3402
+    },
+    {
+      "epoch": 1.6885937693930744,
+      "grad_norm": 0.13334147860764942,
+      "learning_rate": 3.1142622217550054e-06,
+      "loss": 0.7519,
+      "step": 3403
+    },
+    {
+      "epoch": 1.689090232096314,
+      "grad_norm": 0.13393223913614705,
+      "learning_rate": 3.1133147861743435e-06,
+      "loss": 0.6909,
+      "step": 3404
+    },
+    {
+      "epoch": 1.6895866947995533,
+      "grad_norm": 0.1430945896299099,
+      "learning_rate": 3.112367256858351e-06,
+      "loss": 0.7288,
+      "step": 3405
+    },
+    {
+      "epoch": 1.6900831575027926,
+      "grad_norm": 0.13874258590546748,
+      "learning_rate": 3.111419633951841e-06,
+      "loss": 0.7006,
+      "step": 3406
+    },
+    {
+      "epoch": 1.690579620206032,
+      "grad_norm": 0.12943087282029672,
+      "learning_rate": 3.110471917599643e-06,
+      "loss": 0.6984,
+      "step": 3407
+    },
+    {
+      "epoch": 1.6910760829092715,
+      "grad_norm": 0.12968650358300085,
+      "learning_rate": 3.109524107946602e-06,
+      "loss": 0.6686,
+      "step": 3408
+    },
+    {
+      "epoch": 1.691572545612511,
+      "grad_norm": 0.1273962713358661,
+      "learning_rate": 3.1085762051375727e-06,
+      "loss": 0.6858,
+      "step": 3409
+    },
+    {
+      "epoch": 1.6920690083157504,
+      "grad_norm": 0.13601298863571853,
+      "learning_rate": 3.107628209317429e-06,
+      "loss": 0.7531,
+      "step": 3410
+    },
+    {
+      "epoch": 1.6925654710189897,
+      "grad_norm": 0.13179660551493383,
+      "learning_rate": 3.1066801206310558e-06,
+      "loss": 0.7169,
+      "step": 3411
+    },
+    {
+      "epoch": 1.693061933722229,
+      "grad_norm": 0.1262797410728671,
+      "learning_rate": 3.105731939223354e-06,
+      "loss": 0.6924,
+      "step": 3412
+    },
+    {
+      "epoch": 1.6935583964254686,
+      "grad_norm": 0.12881341815960923,
+      "learning_rate": 3.1047836652392367e-06,
+      "loss": 0.704,
+      "step": 3413
+    },
+    {
+      "epoch": 1.694054859128708,
+      "grad_norm": 0.13333244805830557,
+      "learning_rate": 3.103835298823633e-06,
+      "loss": 0.7076,
+      "step": 3414
+    },
+    {
+      "epoch": 1.6945513218319475,
+      "grad_norm": 0.13343352452172727,
+      "learning_rate": 3.1028868401214862e-06,
+      "loss": 0.7251,
+      "step": 3415
+    },
+    {
+      "epoch": 1.6950477845351868,
+      "grad_norm": 0.1310649697161129,
+      "learning_rate": 3.101938289277753e-06,
+      "loss": 0.7048,
+      "step": 3416
+    },
+    {
+      "epoch": 1.6955442472384261,
+      "grad_norm": 0.13399637642806747,
+      "learning_rate": 3.1009896464374027e-06,
+      "loss": 0.7047,
+      "step": 3417
+    },
+    {
+      "epoch": 1.6960407099416657,
+      "grad_norm": 0.14811556928768213,
+      "learning_rate": 3.100040911745421e-06,
+      "loss": 0.7407,
+      "step": 3418
+    },
+    {
+      "epoch": 1.696537172644905,
+      "grad_norm": 0.13730816941238383,
+      "learning_rate": 3.0990920853468076e-06,
+      "loss": 0.7355,
+      "step": 3419
+    },
+    {
+      "epoch": 1.6970336353481446,
+      "grad_norm": 0.13592416626522716,
+      "learning_rate": 3.098143167386574e-06,
+      "loss": 0.672,
+      "step": 3420
+    },
+    {
+      "epoch": 1.6975300980513839,
+      "grad_norm": 0.13560020957046776,
+      "learning_rate": 3.0971941580097476e-06,
+      "loss": 0.7311,
+      "step": 3421
+    },
+    {
+      "epoch": 1.6980265607546232,
+      "grad_norm": 0.12963432619444307,
+      "learning_rate": 3.0962450573613705e-06,
+      "loss": 0.7721,
+      "step": 3422
+    },
+    {
+      "epoch": 1.6985230234578628,
+      "grad_norm": 0.13067867842531036,
+      "learning_rate": 3.0952958655864957e-06,
+      "loss": 0.7049,
+      "step": 3423
+    },
+    {
+      "epoch": 1.699019486161102,
+      "grad_norm": 0.13236415877160046,
+      "learning_rate": 3.0943465828301935e-06,
+      "loss": 0.7503,
+      "step": 3424
+    },
+    {
+      "epoch": 1.6995159488643417,
+      "grad_norm": 0.1347895104957468,
+      "learning_rate": 3.0933972092375447e-06,
+      "loss": 0.7277,
+      "step": 3425
+    },
+    {
+      "epoch": 1.700012411567581,
+      "grad_norm": 0.13020223927968744,
+      "learning_rate": 3.0924477449536467e-06,
+      "loss": 0.6961,
+      "step": 3426
+    },
+    {
+      "epoch": 1.7005088742708203,
+      "grad_norm": 0.12869798181832637,
+      "learning_rate": 3.0914981901236113e-06,
+      "loss": 0.7275,
+      "step": 3427
+    },
+    {
+      "epoch": 1.7010053369740599,
+      "grad_norm": 0.1285881035433599,
+      "learning_rate": 3.0905485448925603e-06,
+      "loss": 0.6905,
+      "step": 3428
+    },
+    {
+      "epoch": 1.7015017996772992,
+      "grad_norm": 0.13590269430618224,
+      "learning_rate": 3.0895988094056333e-06,
+      "loss": 0.7392,
+      "step": 3429
+    },
+    {
+      "epoch": 1.7019982623805388,
+      "grad_norm": 0.1321536527714624,
+      "learning_rate": 3.088648983807982e-06,
+      "loss": 0.7341,
+      "step": 3430
+    },
+    {
+      "epoch": 1.702494725083778,
+      "grad_norm": 0.13823633479008848,
+      "learning_rate": 3.087699068244771e-06,
+      "loss": 0.7409,
+      "step": 3431
+    },
+    {
+      "epoch": 1.7029911877870174,
+      "grad_norm": 0.1342374430839832,
+      "learning_rate": 3.0867490628611817e-06,
+      "loss": 0.7252,
+      "step": 3432
+    },
+    {
+      "epoch": 1.703487650490257,
+      "grad_norm": 0.13031168815094007,
+      "learning_rate": 3.0857989678024057e-06,
+      "loss": 0.7556,
+      "step": 3433
+    },
+    {
+      "epoch": 1.7039841131934963,
+      "grad_norm": 0.132695870959402,
+      "learning_rate": 3.08484878321365e-06,
+      "loss": 0.7524,
+      "step": 3434
+    },
+    {
+      "epoch": 1.7044805758967358,
+      "grad_norm": 0.13160792019629425,
+      "learning_rate": 3.083898509240134e-06,
+      "loss": 0.6997,
+      "step": 3435
+    },
+    {
+      "epoch": 1.7049770385999752,
+      "grad_norm": 0.13528254474745396,
+      "learning_rate": 3.0829481460270937e-06,
+      "loss": 0.7547,
+      "step": 3436
+    },
+    {
+      "epoch": 1.7054735013032145,
+      "grad_norm": 0.12969198957971417,
+      "learning_rate": 3.0819976937197767e-06,
+      "loss": 0.6953,
+      "step": 3437
+    },
+    {
+      "epoch": 1.705969964006454,
+      "grad_norm": 0.13427548997210034,
+      "learning_rate": 3.0810471524634432e-06,
+      "loss": 0.7189,
+      "step": 3438
+    },
+    {
+      "epoch": 1.7064664267096934,
+      "grad_norm": 0.1380792625189435,
+      "learning_rate": 3.080096522403369e-06,
+      "loss": 0.757,
+      "step": 3439
+    },
+    {
+      "epoch": 1.706962889412933,
+      "grad_norm": 0.12798109709025035,
+      "learning_rate": 3.079145803684843e-06,
+      "loss": 0.731,
+      "step": 3440
+    },
+    {
+      "epoch": 1.7074593521161723,
+      "grad_norm": 0.13775152334069862,
+      "learning_rate": 3.078194996453166e-06,
+      "loss": 0.7227,
+      "step": 3441
+    },
+    {
+      "epoch": 1.7079558148194116,
+      "grad_norm": 0.15389555123946358,
+      "learning_rate": 3.0772441008536545e-06,
+      "loss": 0.7067,
+      "step": 3442
+    },
+    {
+      "epoch": 1.7084522775226512,
+      "grad_norm": 0.13139328016325508,
+      "learning_rate": 3.0762931170316384e-06,
+      "loss": 0.7064,
+      "step": 3443
+    },
+    {
+      "epoch": 1.7089487402258905,
+      "grad_norm": 0.14093931277261854,
+      "learning_rate": 3.075342045132459e-06,
+      "loss": 0.7493,
+      "step": 3444
+    },
+    {
+      "epoch": 1.70944520292913,
+      "grad_norm": 0.1339425299662374,
+      "learning_rate": 3.0743908853014726e-06,
+      "loss": 0.7333,
+      "step": 3445
+    },
+    {
+      "epoch": 1.7099416656323694,
+      "grad_norm": 0.13334111715937227,
+      "learning_rate": 3.073439637684049e-06,
+      "loss": 0.6911,
+      "step": 3446
+    },
+    {
+      "epoch": 1.7104381283356087,
+      "grad_norm": 0.13485035304461898,
+      "learning_rate": 3.0724883024255713e-06,
+      "loss": 0.7105,
+      "step": 3447
+    },
+    {
+      "epoch": 1.710934591038848,
+      "grad_norm": 0.13875744459588266,
+      "learning_rate": 3.071536879671436e-06,
+      "loss": 0.6832,
+      "step": 3448
+    },
+    {
+      "epoch": 1.7114310537420876,
+      "grad_norm": 0.12935014733112105,
+      "learning_rate": 3.0705853695670535e-06,
+      "loss": 0.6948,
+      "step": 3449
+    },
+    {
+      "epoch": 1.7119275164453271,
+      "grad_norm": 0.13892908729449366,
+      "learning_rate": 3.0696337722578444e-06,
+      "loss": 0.7187,
+      "step": 3450
+    },
+    {
+      "epoch": 1.7124239791485665,
+      "grad_norm": 0.1311918745816494,
+      "learning_rate": 3.0686820878892472e-06,
+      "loss": 0.698,
+      "step": 3451
+    },
+    {
+      "epoch": 1.7129204418518058,
+      "grad_norm": 0.13310007177839275,
+      "learning_rate": 3.067730316606711e-06,
+      "loss": 0.7454,
+      "step": 3452
+    },
+    {
+      "epoch": 1.7134169045550451,
+      "grad_norm": 0.1407962461572666,
+      "learning_rate": 3.066778458555698e-06,
+      "loss": 0.7591,
+      "step": 3453
+    },
+    {
+      "epoch": 1.7139133672582847,
+      "grad_norm": 0.13091646981705454,
+      "learning_rate": 3.065826513881686e-06,
+      "loss": 0.7167,
+      "step": 3454
+    },
+    {
+      "epoch": 1.7144098299615242,
+      "grad_norm": 0.13923461390544403,
+      "learning_rate": 3.0648744827301636e-06,
+      "loss": 0.7779,
+      "step": 3455
+    },
+    {
+      "epoch": 1.7149062926647636,
+      "grad_norm": 0.1344779145205169,
+      "learning_rate": 3.0639223652466337e-06,
+      "loss": 0.7465,
+      "step": 3456
+    },
+    {
+      "epoch": 1.715402755368003,
+      "grad_norm": 0.12825800482719335,
+      "learning_rate": 3.0629701615766116e-06,
+      "loss": 0.6914,
+      "step": 3457
+    },
+    {
+      "epoch": 1.7158992180712422,
+      "grad_norm": 0.13450044556444432,
+      "learning_rate": 3.0620178718656272e-06,
+      "loss": 0.7268,
+      "step": 3458
+    },
+    {
+      "epoch": 1.7163956807744818,
+      "grad_norm": 0.13374849327239205,
+      "learning_rate": 3.061065496259222e-06,
+      "loss": 0.6687,
+      "step": 3459
+    },
+    {
+      "epoch": 1.7168921434777213,
+      "grad_norm": 0.1350801681751466,
+      "learning_rate": 3.060113034902952e-06,
+      "loss": 0.6999,
+      "step": 3460
+    },
+    {
+      "epoch": 1.7173886061809607,
+      "grad_norm": 0.13712402938215143,
+      "learning_rate": 3.0591604879423858e-06,
+      "loss": 0.7099,
+      "step": 3461
+    },
+    {
+      "epoch": 1.7178850688842,
+      "grad_norm": 0.1305628583173919,
+      "learning_rate": 3.058207855523104e-06,
+      "loss": 0.6846,
+      "step": 3462
+    },
+    {
+      "epoch": 1.7183815315874393,
+      "grad_norm": 0.13419917510098936,
+      "learning_rate": 3.057255137790701e-06,
+      "loss": 0.7718,
+      "step": 3463
+    },
+    {
+      "epoch": 1.718877994290679,
+      "grad_norm": 0.13680810835604765,
+      "learning_rate": 3.056302334890786e-06,
+      "loss": 0.7589,
+      "step": 3464
+    },
+    {
+      "epoch": 1.7193744569939184,
+      "grad_norm": 0.1336240292540668,
+      "learning_rate": 3.0553494469689792e-06,
+      "loss": 0.6608,
+      "step": 3465
+    },
+    {
+      "epoch": 1.7198709196971578,
+      "grad_norm": 0.1396188299452781,
+      "learning_rate": 3.054396474170913e-06,
+      "loss": 0.749,
+      "step": 3466
+    },
+    {
+      "epoch": 1.720367382400397,
+      "grad_norm": 0.13274972442712532,
+      "learning_rate": 3.0534434166422346e-06,
+      "loss": 0.7062,
+      "step": 3467
+    },
+    {
+      "epoch": 1.7208638451036364,
+      "grad_norm": 0.1381067587560184,
+      "learning_rate": 3.052490274528604e-06,
+      "loss": 0.7976,
+      "step": 3468
+    },
+    {
+      "epoch": 1.721360307806876,
+      "grad_norm": 0.1415613666037084,
+      "learning_rate": 3.0515370479756936e-06,
+      "loss": 0.6759,
+      "step": 3469
+    },
+    {
+      "epoch": 1.7218567705101155,
+      "grad_norm": 0.136114676951742,
+      "learning_rate": 3.0505837371291885e-06,
+      "loss": 0.6921,
+      "step": 3470
+    },
+    {
+      "epoch": 1.7223532332133549,
+      "grad_norm": 0.1356442686679228,
+      "learning_rate": 3.0496303421347872e-06,
+      "loss": 0.6642,
+      "step": 3471
+    },
+    {
+      "epoch": 1.7228496959165942,
+      "grad_norm": 0.1376358359049435,
+      "learning_rate": 3.0486768631382012e-06,
+      "loss": 0.7658,
+      "step": 3472
+    },
+    {
+      "epoch": 1.7233461586198335,
+      "grad_norm": 0.1392338301585101,
+      "learning_rate": 3.047723300285153e-06,
+      "loss": 0.7241,
+      "step": 3473
+    },
+    {
+      "epoch": 1.723842621323073,
+      "grad_norm": 0.1366470624963273,
+      "learning_rate": 3.046769653721381e-06,
+      "loss": 0.6675,
+      "step": 3474
+    },
+    {
+      "epoch": 1.7243390840263126,
+      "grad_norm": 0.13307292368981993,
+      "learning_rate": 3.045815923592634e-06,
+      "loss": 0.6673,
+      "step": 3475
+    },
+    {
+      "epoch": 1.724835546729552,
+      "grad_norm": 0.1357966319706256,
+      "learning_rate": 3.0448621100446753e-06,
+      "loss": 0.7633,
+      "step": 3476
+    },
+    {
+      "epoch": 1.7253320094327913,
+      "grad_norm": 0.14547912410392153,
+      "learning_rate": 3.0439082132232785e-06,
+      "loss": 0.7257,
+      "step": 3477
+    },
+    {
+      "epoch": 1.7258284721360306,
+      "grad_norm": 0.13325165665610728,
+      "learning_rate": 3.0429542332742322e-06,
+      "loss": 0.7046,
+      "step": 3478
+    },
+    {
+      "epoch": 1.7263249348392702,
+      "grad_norm": 0.1455990009531011,
+      "learning_rate": 3.042000170343337e-06,
+      "loss": 0.7195,
+      "step": 3479
+    },
+    {
+      "epoch": 1.7268213975425097,
+      "grad_norm": 0.13234995456293744,
+      "learning_rate": 3.0410460245764066e-06,
+      "loss": 0.6786,
+      "step": 3480
+    },
+    {
+      "epoch": 1.727317860245749,
+      "grad_norm": 0.13558320267676813,
+      "learning_rate": 3.0400917961192667e-06,
+      "loss": 0.7187,
+      "step": 3481
+    },
+    {
+      "epoch": 1.7278143229489884,
+      "grad_norm": 0.1309367275550789,
+      "learning_rate": 3.0391374851177547e-06,
+      "loss": 0.724,
+      "step": 3482
+    },
+    {
+      "epoch": 1.7283107856522277,
+      "grad_norm": 0.12849728447833228,
+      "learning_rate": 3.0381830917177225e-06,
+      "loss": 0.6885,
+      "step": 3483
+    },
+    {
+      "epoch": 1.7288072483554673,
+      "grad_norm": 0.13668857277562277,
+      "learning_rate": 3.0372286160650334e-06,
+      "loss": 0.7577,
+      "step": 3484
+    },
+    {
+      "epoch": 1.7293037110587068,
+      "grad_norm": 0.13650280495720618,
+      "learning_rate": 3.036274058305565e-06,
+      "loss": 0.7419,
+      "step": 3485
+    },
+    {
+      "epoch": 1.7298001737619462,
+      "grad_norm": 0.1294860482203652,
+      "learning_rate": 3.0353194185852052e-06,
+      "loss": 0.7442,
+      "step": 3486
+    },
+    {
+      "epoch": 1.7302966364651855,
+      "grad_norm": 0.13841166664415844,
+      "learning_rate": 3.0343646970498554e-06,
+      "loss": 0.7568,
+      "step": 3487
+    },
+    {
+      "epoch": 1.7307930991684248,
+      "grad_norm": 0.1293180189453792,
+      "learning_rate": 3.0334098938454293e-06,
+      "loss": 0.7047,
+      "step": 3488
+    },
+    {
+      "epoch": 1.7312895618716644,
+      "grad_norm": 0.13258057280123206,
+      "learning_rate": 3.0324550091178536e-06,
+      "loss": 0.7315,
+      "step": 3489
+    },
+    {
+      "epoch": 1.731786024574904,
+      "grad_norm": 0.13503628372669735,
+      "learning_rate": 3.031500043013067e-06,
+      "loss": 0.7302,
+      "step": 3490
+    },
+    {
+      "epoch": 1.7322824872781433,
+      "grad_norm": 0.13290187674269438,
+      "learning_rate": 3.0305449956770206e-06,
+      "loss": 0.7343,
+      "step": 3491
+    },
+    {
+      "epoch": 1.7327789499813826,
+      "grad_norm": 0.13540856252868388,
+      "learning_rate": 3.0295898672556785e-06,
+      "loss": 0.7448,
+      "step": 3492
+    },
+    {
+      "epoch": 1.733275412684622,
+      "grad_norm": 0.13564514354413926,
+      "learning_rate": 3.0286346578950165e-06,
+      "loss": 0.703,
+      "step": 3493
+    },
+    {
+      "epoch": 1.7337718753878615,
+      "grad_norm": 0.1377806848578677,
+      "learning_rate": 3.0276793677410226e-06,
+      "loss": 0.7445,
+      "step": 3494
+    },
+    {
+      "epoch": 1.734268338091101,
+      "grad_norm": 0.1357491981149147,
+      "learning_rate": 3.0267239969396983e-06,
+      "loss": 0.7689,
+      "step": 3495
+    },
+    {
+      "epoch": 1.7347648007943404,
+      "grad_norm": 0.1278410172706573,
+      "learning_rate": 3.0257685456370573e-06,
+      "loss": 0.7004,
+      "step": 3496
+    },
+    {
+      "epoch": 1.7352612634975797,
+      "grad_norm": 0.1314164181234785,
+      "learning_rate": 3.024813013979123e-06,
+      "loss": 0.6591,
+      "step": 3497
+    },
+    {
+      "epoch": 1.735757726200819,
+      "grad_norm": 0.1455338423625045,
+      "learning_rate": 3.0238574021119356e-06,
+      "loss": 0.74,
+      "step": 3498
+    },
+    {
+      "epoch": 1.7362541889040586,
+      "grad_norm": 0.1334665037915578,
+      "learning_rate": 3.0229017101815424e-06,
+      "loss": 0.7069,
+      "step": 3499
+    },
+    {
+      "epoch": 1.7367506516072981,
+      "grad_norm": 0.13544177303837263,
+      "learning_rate": 3.0219459383340073e-06,
+      "loss": 0.7632,
+      "step": 3500
+    },
+    {
+      "epoch": 1.7372471143105375,
+      "grad_norm": 0.13969483295131455,
+      "learning_rate": 3.0209900867154042e-06,
+      "loss": 0.7542,
+      "step": 3501
+    },
+    {
+      "epoch": 1.7377435770137768,
+      "grad_norm": 0.13416029469080804,
+      "learning_rate": 3.02003415547182e-06,
+      "loss": 0.6914,
+      "step": 3502
+    },
+    {
+      "epoch": 1.7382400397170161,
+      "grad_norm": 0.1342512889947578,
+      "learning_rate": 3.0190781447493526e-06,
+      "loss": 0.7125,
+      "step": 3503
+    },
+    {
+      "epoch": 1.7387365024202557,
+      "grad_norm": 0.14318589298841014,
+      "learning_rate": 3.018122054694115e-06,
+      "loss": 0.7625,
+      "step": 3504
+    },
+    {
+      "epoch": 1.7392329651234952,
+      "grad_norm": 0.13889004826208431,
+      "learning_rate": 3.0171658854522274e-06,
+      "loss": 0.7114,
+      "step": 3505
+    },
+    {
+      "epoch": 1.7397294278267346,
+      "grad_norm": 0.13605943240627727,
+      "learning_rate": 3.016209637169827e-06,
+      "loss": 0.7414,
+      "step": 3506
+    },
+    {
+      "epoch": 1.740225890529974,
+      "grad_norm": 0.13235701067519334,
+      "learning_rate": 3.01525330999306e-06,
+      "loss": 0.7208,
+      "step": 3507
+    },
+    {
+      "epoch": 1.7407223532332132,
+      "grad_norm": 0.135326876994932,
+      "learning_rate": 3.0142969040680865e-06,
+      "loss": 0.7344,
+      "step": 3508
+    },
+    {
+      "epoch": 1.7412188159364528,
+      "grad_norm": 0.13477736579977354,
+      "learning_rate": 3.0133404195410764e-06,
+      "loss": 0.7326,
+      "step": 3509
+    },
+    {
+      "epoch": 1.7417152786396923,
+      "grad_norm": 0.13284231516066178,
+      "learning_rate": 3.0123838565582147e-06,
+      "loss": 0.7119,
+      "step": 3510
+    },
+    {
+      "epoch": 1.7422117413429317,
+      "grad_norm": 0.13238135868096979,
+      "learning_rate": 3.011427215265696e-06,
+      "loss": 0.7018,
+      "step": 3511
+    },
+    {
+      "epoch": 1.742708204046171,
+      "grad_norm": 0.13474513462013873,
+      "learning_rate": 3.0104704958097275e-06,
+      "loss": 0.7405,
+      "step": 3512
+    },
+    {
+      "epoch": 1.7432046667494103,
+      "grad_norm": 0.1302603109994265,
+      "learning_rate": 3.009513698336529e-06,
+      "loss": 0.7194,
+      "step": 3513
+    },
+    {
+      "epoch": 1.7437011294526499,
+      "grad_norm": 0.13946924173551373,
+      "learning_rate": 3.00855682299233e-06,
+      "loss": 0.7454,
+      "step": 3514
+    },
+    {
+      "epoch": 1.7441975921558894,
+      "grad_norm": 0.14022806898060922,
+      "learning_rate": 3.0075998699233757e-06,
+      "loss": 0.7258,
+      "step": 3515
+    },
+    {
+      "epoch": 1.7446940548591288,
+      "grad_norm": 0.14276302517939646,
+      "learning_rate": 3.006642839275919e-06,
+      "loss": 0.7353,
+      "step": 3516
+    },
+    {
+      "epoch": 1.745190517562368,
+      "grad_norm": 0.13440824467755944,
+      "learning_rate": 3.0056857311962286e-06,
+      "loss": 0.6975,
+      "step": 3517
+    },
+    {
+      "epoch": 1.7456869802656074,
+      "grad_norm": 0.13069918155058813,
+      "learning_rate": 3.0047285458305818e-06,
+      "loss": 0.7307,
+      "step": 3518
+    },
+    {
+      "epoch": 1.746183442968847,
+      "grad_norm": 0.13013866848398856,
+      "learning_rate": 3.0037712833252696e-06,
+      "loss": 0.7075,
+      "step": 3519
+    },
+    {
+      "epoch": 1.7466799056720865,
+      "grad_norm": 0.13588085761716664,
+      "learning_rate": 3.0028139438265946e-06,
+      "loss": 0.7028,
+      "step": 3520
+    },
+    {
+      "epoch": 1.7471763683753259,
+      "grad_norm": 0.1379722426469276,
+      "learning_rate": 3.00185652748087e-06,
+      "loss": 0.7558,
+      "step": 3521
+    },
+    {
+      "epoch": 1.7476728310785652,
+      "grad_norm": 0.13440744873614507,
+      "learning_rate": 3.000899034434421e-06,
+      "loss": 0.6893,
+      "step": 3522
+    },
+    {
+      "epoch": 1.7481692937818045,
+      "grad_norm": 0.13573972148872435,
+      "learning_rate": 2.9999414648335866e-06,
+      "loss": 0.7284,
+      "step": 3523
+    },
+    {
+      "epoch": 1.748665756485044,
+      "grad_norm": 0.12921771068329435,
+      "learning_rate": 2.9989838188247157e-06,
+      "loss": 0.7153,
+      "step": 3524
+    },
+    {
+      "epoch": 1.7491622191882836,
+      "grad_norm": 0.128812429516214,
+      "learning_rate": 2.998026096554168e-06,
+      "loss": 0.7018,
+      "step": 3525
+    },
+    {
+      "epoch": 1.749658681891523,
+      "grad_norm": 0.13351254077010827,
+      "learning_rate": 2.9970682981683165e-06,
+      "loss": 0.7398,
+      "step": 3526
+    },
+    {
+      "epoch": 1.7501551445947623,
+      "grad_norm": 0.13254352974014932,
+      "learning_rate": 2.9961104238135457e-06,
+      "loss": 0.7104,
+      "step": 3527
+    },
+    {
+      "epoch": 1.7506516072980016,
+      "grad_norm": 0.13040410881238124,
+      "learning_rate": 2.9951524736362513e-06,
+      "loss": 0.6791,
+      "step": 3528
+    },
+    {
+      "epoch": 1.7506516072980016,
+      "eval_loss": 0.732168436050415,
+      "eval_runtime": 135.7964,
+      "eval_samples_per_second": 223.519,
+      "eval_steps_per_second": 27.946,
+      "step": 3528
+    },
+    {
+      "epoch": 1.7511480700012412,
+      "grad_norm": 0.1454158991301188,
+      "learning_rate": 2.9941944477828405e-06,
+      "loss": 0.729,
+      "step": 3529
+    },
+    {
+      "epoch": 1.7516445327044807,
+      "grad_norm": 0.14264344607881654,
+      "learning_rate": 2.9932363463997325e-06,
+      "loss": 0.761,
+      "step": 3530
+    },
+    {
+      "epoch": 1.75214099540772,
+      "grad_norm": 0.13191391058050855,
+      "learning_rate": 2.992278169633357e-06,
+      "loss": 0.6719,
+      "step": 3531
+    },
+    {
+      "epoch": 1.7526374581109594,
+      "grad_norm": 0.13456024392134855,
+      "learning_rate": 2.9913199176301567e-06,
+      "loss": 0.7191,
+      "step": 3532
+    },
+    {
+      "epoch": 1.7531339208141987,
+      "grad_norm": 0.12996595723949278,
+      "learning_rate": 2.990361590536584e-06,
+      "loss": 0.6824,
+      "step": 3533
+    },
+    {
+      "epoch": 1.7536303835174383,
+      "grad_norm": 0.1328036765145108,
+      "learning_rate": 2.9894031884991047e-06,
+      "loss": 0.7057,
+      "step": 3534
+    },
+    {
+      "epoch": 1.7541268462206778,
+      "grad_norm": 0.13831131399589613,
+      "learning_rate": 2.9884447116641955e-06,
+      "loss": 0.697,
+      "step": 3535
+    },
+    {
+      "epoch": 1.7546233089239172,
+      "grad_norm": 0.1421695480209381,
+      "learning_rate": 2.987486160178344e-06,
+      "loss": 0.7423,
+      "step": 3536
+    },
+    {
+      "epoch": 1.7551197716271565,
+      "grad_norm": 0.12821462370317024,
+      "learning_rate": 2.9865275341880484e-06,
+      "loss": 0.7165,
+      "step": 3537
+    },
+    {
+      "epoch": 1.7556162343303958,
+      "grad_norm": 0.1420145829606162,
+      "learning_rate": 2.9855688338398202e-06,
+      "loss": 0.6948,
+      "step": 3538
+    },
+    {
+      "epoch": 1.7561126970336354,
+      "grad_norm": 0.13455255604118072,
+      "learning_rate": 2.9846100592801815e-06,
+      "loss": 0.7232,
+      "step": 3539
+    },
+    {
+      "epoch": 1.756609159736875,
+      "grad_norm": 0.13103602177339743,
+      "learning_rate": 2.9836512106556655e-06,
+      "loss": 0.7248,
+      "step": 3540
+    },
+    {
+      "epoch": 1.7571056224401143,
+      "grad_norm": 0.1384857291962389,
+      "learning_rate": 2.982692288112816e-06,
+      "loss": 0.7303,
+      "step": 3541
+    },
+    {
+      "epoch": 1.7576020851433536,
+      "grad_norm": 0.16056289613315078,
+      "learning_rate": 2.98173329179819e-06,
+      "loss": 0.7049,
+      "step": 3542
+    },
+    {
+      "epoch": 1.758098547846593,
+      "grad_norm": 0.1344941324110072,
+      "learning_rate": 2.9807742218583547e-06,
+      "loss": 0.691,
+      "step": 3543
+    },
+    {
+      "epoch": 1.7585950105498325,
+      "grad_norm": 0.1354748157954558,
+      "learning_rate": 2.9798150784398885e-06,
+      "loss": 0.7215,
+      "step": 3544
+    },
+    {
+      "epoch": 1.759091473253072,
+      "grad_norm": 0.14122504918334078,
+      "learning_rate": 2.9788558616893796e-06,
+      "loss": 0.7167,
+      "step": 3545
+    },
+    {
+      "epoch": 1.7595879359563114,
+      "grad_norm": 0.14019815087178897,
+      "learning_rate": 2.9778965717534314e-06,
+      "loss": 0.7125,
+      "step": 3546
+    },
+    {
+      "epoch": 1.7600843986595507,
+      "grad_norm": 0.13688281809385325,
+      "learning_rate": 2.976937208778654e-06,
+      "loss": 0.715,
+      "step": 3547
+    },
+    {
+      "epoch": 1.76058086136279,
+      "grad_norm": 0.14612296075479175,
+      "learning_rate": 2.975977772911671e-06,
+      "loss": 0.7693,
+      "step": 3548
+    },
+    {
+      "epoch": 1.7610773240660296,
+      "grad_norm": 0.13373675241491112,
+      "learning_rate": 2.9750182642991175e-06,
+      "loss": 0.7149,
+      "step": 3549
+    },
+    {
+      "epoch": 1.7615737867692691,
+      "grad_norm": 0.13249339511074687,
+      "learning_rate": 2.974058683087639e-06,
+      "loss": 0.7176,
+      "step": 3550
+    },
+    {
+      "epoch": 1.7620702494725085,
+      "grad_norm": 0.1445155372590999,
+      "learning_rate": 2.9730990294238914e-06,
+      "loss": 0.7565,
+      "step": 3551
+    },
+    {
+      "epoch": 1.7625667121757478,
+      "grad_norm": 0.1301812819229513,
+      "learning_rate": 2.972139303454543e-06,
+      "loss": 0.7337,
+      "step": 3552
+    },
+    {
+      "epoch": 1.7630631748789871,
+      "grad_norm": 0.13813116954578447,
+      "learning_rate": 2.971179505326272e-06,
+      "loss": 0.7344,
+      "step": 3553
+    },
+    {
+      "epoch": 1.7635596375822267,
+      "grad_norm": 0.13748336922899862,
+      "learning_rate": 2.970219635185769e-06,
+      "loss": 0.781,
+      "step": 3554
+    },
+    {
+      "epoch": 1.764056100285466,
+      "grad_norm": 0.13434117881943183,
+      "learning_rate": 2.969259693179733e-06,
+      "loss": 0.7186,
+      "step": 3555
+    },
+    {
+      "epoch": 1.7645525629887056,
+      "grad_norm": 0.13228896105397034,
+      "learning_rate": 2.968299679454878e-06,
+      "loss": 0.7308,
+      "step": 3556
+    },
+    {
+      "epoch": 1.7650490256919449,
+      "grad_norm": 0.1279445994151255,
+      "learning_rate": 2.967339594157925e-06,
+      "loss": 0.7249,
+      "step": 3557
+    },
+    {
+      "epoch": 1.7655454883951842,
+      "grad_norm": 0.13875252416949538,
+      "learning_rate": 2.9663794374356082e-06,
+      "loss": 0.7315,
+      "step": 3558
+    },
+    {
+      "epoch": 1.7660419510984238,
+      "grad_norm": 0.13828288612620065,
+      "learning_rate": 2.9654192094346716e-06,
+      "loss": 0.6718,
+      "step": 3559
+    },
+    {
+      "epoch": 1.766538413801663,
+      "grad_norm": 0.13503162566112942,
+      "learning_rate": 2.9644589103018728e-06,
+      "loss": 0.7089,
+      "step": 3560
+    },
+    {
+      "epoch": 1.7670348765049027,
+      "grad_norm": 0.12961214402076365,
+      "learning_rate": 2.9634985401839754e-06,
+      "loss": 0.7265,
+      "step": 3561
+    },
+    {
+      "epoch": 1.767531339208142,
+      "grad_norm": 0.1376599301401424,
+      "learning_rate": 2.9625380992277585e-06,
+      "loss": 0.7361,
+      "step": 3562
+    },
+    {
+      "epoch": 1.7680278019113813,
+      "grad_norm": 0.1315003737365444,
+      "learning_rate": 2.961577587580009e-06,
+      "loss": 0.7222,
+      "step": 3563
+    },
+    {
+      "epoch": 1.7685242646146209,
+      "grad_norm": 0.13997334248150592,
+      "learning_rate": 2.960617005387526e-06,
+      "loss": 0.7678,
+      "step": 3564
+    },
+    {
+      "epoch": 1.7690207273178602,
+      "grad_norm": 0.1377189418395456,
+      "learning_rate": 2.9596563527971197e-06,
+      "loss": 0.7488,
+      "step": 3565
+    },
+    {
+      "epoch": 1.7695171900210998,
+      "grad_norm": 0.13949649187035196,
+      "learning_rate": 2.9586956299556093e-06,
+      "loss": 0.7356,
+      "step": 3566
+    },
+    {
+      "epoch": 1.770013652724339,
+      "grad_norm": 0.13072149028094537,
+      "learning_rate": 2.957734837009827e-06,
+      "loss": 0.7439,
+      "step": 3567
+    },
+    {
+      "epoch": 1.7705101154275784,
+      "grad_norm": 0.13314492915960163,
+      "learning_rate": 2.9567739741066135e-06,
+      "loss": 0.7131,
+      "step": 3568
+    },
+    {
+      "epoch": 1.771006578130818,
+      "grad_norm": 0.138259104473977,
+      "learning_rate": 2.955813041392822e-06,
+      "loss": 0.7083,
+      "step": 3569
+    },
+    {
+      "epoch": 1.7715030408340573,
+      "grad_norm": 0.13460619412681396,
+      "learning_rate": 2.9548520390153157e-06,
+      "loss": 0.7247,
+      "step": 3570
+    },
+    {
+      "epoch": 1.7719995035372968,
+      "grad_norm": 0.13188160151808176,
+      "learning_rate": 2.9538909671209683e-06,
+      "loss": 0.7327,
+      "step": 3571
+    },
+    {
+      "epoch": 1.7724959662405362,
+      "grad_norm": 0.13915943835734876,
+      "learning_rate": 2.952929825856664e-06,
+      "loss": 0.7245,
+      "step": 3572
+    },
+    {
+      "epoch": 1.7729924289437755,
+      "grad_norm": 0.13113442999553518,
+      "learning_rate": 2.9519686153692984e-06,
+      "loss": 0.6766,
+      "step": 3573
+    },
+    {
+      "epoch": 1.773488891647015,
+      "grad_norm": 0.15563993017589076,
+      "learning_rate": 2.9510073358057763e-06,
+      "loss": 0.778,
+      "step": 3574
+    },
+    {
+      "epoch": 1.7739853543502544,
+      "grad_norm": 0.13605674921389552,
+      "learning_rate": 2.950045987313014e-06,
+      "loss": 0.7889,
+      "step": 3575
+    },
+    {
+      "epoch": 1.774481817053494,
+      "grad_norm": 0.13255856624671072,
+      "learning_rate": 2.949084570037939e-06,
+      "loss": 0.7268,
+      "step": 3576
+    },
+    {
+      "epoch": 1.7749782797567333,
+      "grad_norm": 0.1425167455509219,
+      "learning_rate": 2.948123084127488e-06,
+      "loss": 0.779,
+      "step": 3577
+    },
+    {
+      "epoch": 1.7754747424599726,
+      "grad_norm": 0.13366479508025797,
+      "learning_rate": 2.947161529728609e-06,
+      "loss": 0.6817,
+      "step": 3578
+    },
+    {
+      "epoch": 1.7759712051632122,
+      "grad_norm": 0.13634858914868714,
+      "learning_rate": 2.946199906988259e-06,
+      "loss": 0.7171,
+      "step": 3579
+    },
+    {
+      "epoch": 1.7764676678664515,
+      "grad_norm": 0.1339820519944257,
+      "learning_rate": 2.9452382160534075e-06,
+      "loss": 0.6727,
+      "step": 3580
+    },
+    {
+      "epoch": 1.776964130569691,
+      "grad_norm": 0.13103750913100626,
+      "learning_rate": 2.9442764570710343e-06,
+      "loss": 0.694,
+      "step": 3581
+    },
+    {
+      "epoch": 1.7774605932729304,
+      "grad_norm": 0.1313014217590884,
+      "learning_rate": 2.943314630188127e-06,
+      "loss": 0.7572,
+      "step": 3582
+    },
+    {
+      "epoch": 1.7779570559761697,
+      "grad_norm": 0.13018315808842246,
+      "learning_rate": 2.942352735551688e-06,
+      "loss": 0.7372,
+      "step": 3583
+    },
+    {
+      "epoch": 1.7784535186794093,
+      "grad_norm": 0.1317368490396673,
+      "learning_rate": 2.9413907733087255e-06,
+      "loss": 0.7442,
+      "step": 3584
+    },
+    {
+      "epoch": 1.7789499813826486,
+      "grad_norm": 0.14221916855971456,
+      "learning_rate": 2.9404287436062596e-06,
+      "loss": 0.7101,
+      "step": 3585
+    },
+    {
+      "epoch": 1.7794464440858881,
+      "grad_norm": 0.13279652857342705,
+      "learning_rate": 2.9394666465913225e-06,
+      "loss": 0.7166,
+      "step": 3586
+    },
+    {
+      "epoch": 1.7799429067891275,
+      "grad_norm": 0.13104555294939943,
+      "learning_rate": 2.9385044824109544e-06,
+      "loss": 0.6961,
+      "step": 3587
+    },
+    {
+      "epoch": 1.7804393694923668,
+      "grad_norm": 0.13924421190691488,
+      "learning_rate": 2.937542251212207e-06,
+      "loss": 0.7798,
+      "step": 3588
+    },
+    {
+      "epoch": 1.7809358321956061,
+      "grad_norm": 0.13257146284425395,
+      "learning_rate": 2.936579953142143e-06,
+      "loss": 0.7222,
+      "step": 3589
+    },
+    {
+      "epoch": 1.7814322948988457,
+      "grad_norm": 0.13510842196780357,
+      "learning_rate": 2.935617588347832e-06,
+      "loss": 0.6802,
+      "step": 3590
+    },
+    {
+      "epoch": 1.7819287576020852,
+      "grad_norm": 0.1370255691430651,
+      "learning_rate": 2.9346551569763584e-06,
+      "loss": 0.7058,
+      "step": 3591
+    },
+    {
+      "epoch": 1.7824252203053246,
+      "grad_norm": 0.13103573445504976,
+      "learning_rate": 2.9336926591748116e-06,
+      "loss": 0.7076,
+      "step": 3592
+    },
+    {
+      "epoch": 1.782921683008564,
+      "grad_norm": 0.1332868798807041,
+      "learning_rate": 2.932730095090297e-06,
+      "loss": 0.7525,
+      "step": 3593
+    },
+    {
+      "epoch": 1.7834181457118032,
+      "grad_norm": 0.13239816333094517,
+      "learning_rate": 2.931767464869926e-06,
+      "loss": 0.7118,
+      "step": 3594
+    },
+    {
+      "epoch": 1.7839146084150428,
+      "grad_norm": 0.13806112422697028,
+      "learning_rate": 2.93080476866082e-06,
+      "loss": 0.6911,
+      "step": 3595
+    },
+    {
+      "epoch": 1.7844110711182823,
+      "grad_norm": 0.13143745347123123,
+      "learning_rate": 2.9298420066101123e-06,
+      "loss": 0.7114,
+      "step": 3596
+    },
+    {
+      "epoch": 1.7849075338215217,
+      "grad_norm": 0.1303545454261341,
+      "learning_rate": 2.928879178864946e-06,
+      "loss": 0.7026,
+      "step": 3597
+    },
+    {
+      "epoch": 1.785403996524761,
+      "grad_norm": 0.13284060964480524,
+      "learning_rate": 2.9279162855724736e-06,
+      "loss": 0.7719,
+      "step": 3598
+    },
+    {
+      "epoch": 1.7859004592280003,
+      "grad_norm": 0.13126700923262716,
+      "learning_rate": 2.926953326879859e-06,
+      "loss": 0.745,
+      "step": 3599
+    },
+    {
+      "epoch": 1.78639692193124,
+      "grad_norm": 0.13808740540622605,
+      "learning_rate": 2.9259903029342733e-06,
+      "loss": 0.7492,
+      "step": 3600
+    },
+    {
+      "epoch": 1.7868933846344794,
+      "grad_norm": 0.13223911487658072,
+      "learning_rate": 2.9250272138829004e-06,
+      "loss": 0.7115,
+      "step": 3601
+    },
+    {
+      "epoch": 1.7873898473377188,
+      "grad_norm": 0.13756497219754027,
+      "learning_rate": 2.9240640598729325e-06,
+      "loss": 0.6888,
+      "step": 3602
+    },
+    {
+      "epoch": 1.787886310040958,
+      "grad_norm": 0.12838540835989878,
+      "learning_rate": 2.923100841051572e-06,
+      "loss": 0.695,
+      "step": 3603
+    },
+    {
+      "epoch": 1.7883827727441974,
+      "grad_norm": 0.1345217155143832,
+      "learning_rate": 2.922137557566032e-06,
+      "loss": 0.7352,
+      "step": 3604
+    },
+    {
+      "epoch": 1.788879235447437,
+      "grad_norm": 0.13011116824419314,
+      "learning_rate": 2.9211742095635353e-06,
+      "loss": 0.715,
+      "step": 3605
+    },
+    {
+      "epoch": 1.7893756981506765,
+      "grad_norm": 0.13345770601068715,
+      "learning_rate": 2.9202107971913135e-06,
+      "loss": 0.6526,
+      "step": 3606
+    },
+    {
+      "epoch": 1.7898721608539159,
+      "grad_norm": 0.13058633473314155,
+      "learning_rate": 2.9192473205966086e-06,
+      "loss": 0.7298,
+      "step": 3607
+    },
+    {
+      "epoch": 1.7903686235571552,
+      "grad_norm": 0.13576890050935692,
+      "learning_rate": 2.918283779926673e-06,
+      "loss": 0.7379,
+      "step": 3608
+    },
+    {
+      "epoch": 1.7908650862603945,
+      "grad_norm": 0.1357470033215107,
+      "learning_rate": 2.917320175328769e-06,
+      "loss": 0.7684,
+      "step": 3609
+    },
+    {
+      "epoch": 1.791361548963634,
+      "grad_norm": 0.12842283154879058,
+      "learning_rate": 2.9163565069501665e-06,
+      "loss": 0.7074,
+      "step": 3610
+    },
+    {
+      "epoch": 1.7918580116668736,
+      "grad_norm": 0.1315525180161788,
+      "learning_rate": 2.915392774938148e-06,
+      "loss": 0.7179,
+      "step": 3611
+    },
+    {
+      "epoch": 1.792354474370113,
+      "grad_norm": 0.13320229301947734,
+      "learning_rate": 2.914428979440004e-06,
+      "loss": 0.7488,
+      "step": 3612
+    },
+    {
+      "epoch": 1.7928509370733523,
+      "grad_norm": 0.17742426766768907,
+      "learning_rate": 2.9134651206030356e-06,
+      "loss": 0.6971,
+      "step": 3613
+    },
+    {
+      "epoch": 1.7933473997765916,
+      "grad_norm": 0.13687894210660118,
+      "learning_rate": 2.9125011985745526e-06,
+      "loss": 0.7579,
+      "step": 3614
+    },
+    {
+      "epoch": 1.7938438624798312,
+      "grad_norm": 0.14356923426142504,
+      "learning_rate": 2.911537213501876e-06,
+      "loss": 0.8023,
+      "step": 3615
+    },
+    {
+      "epoch": 1.7943403251830707,
+      "grad_norm": 0.13255693124637588,
+      "learning_rate": 2.9105731655323345e-06,
+      "loss": 0.7161,
+      "step": 3616
+    },
+    {
+      "epoch": 1.79483678788631,
+      "grad_norm": 0.1340919080282845,
+      "learning_rate": 2.9096090548132678e-06,
+      "loss": 0.698,
+      "step": 3617
+    },
+    {
+      "epoch": 1.7953332505895494,
+      "grad_norm": 0.13702138442017256,
+      "learning_rate": 2.908644881492024e-06,
+      "loss": 0.7523,
+      "step": 3618
+    },
+    {
+      "epoch": 1.7958297132927887,
+      "grad_norm": 0.1353149372278838,
+      "learning_rate": 2.9076806457159628e-06,
+      "loss": 0.7048,
+      "step": 3619
+    },
+    {
+      "epoch": 1.7963261759960283,
+      "grad_norm": 0.13799204504298163,
+      "learning_rate": 2.9067163476324513e-06,
+      "loss": 0.7481,
+      "step": 3620
+    },
+    {
+      "epoch": 1.7968226386992678,
+      "grad_norm": 0.12790377926586102,
+      "learning_rate": 2.905751987388868e-06,
+      "loss": 0.6927,
+      "step": 3621
+    },
+    {
+      "epoch": 1.7973191014025072,
+      "grad_norm": 0.13965087986871189,
+      "learning_rate": 2.904787565132598e-06,
+      "loss": 0.7361,
+      "step": 3622
+    },
+    {
+      "epoch": 1.7978155641057465,
+      "grad_norm": 0.1336271147899048,
+      "learning_rate": 2.903823081011039e-06,
+      "loss": 0.7112,
+      "step": 3623
+    },
+    {
+      "epoch": 1.7983120268089858,
+      "grad_norm": 0.13734805652624563,
+      "learning_rate": 2.9028585351715977e-06,
+      "loss": 0.7405,
+      "step": 3624
+    },
+    {
+      "epoch": 1.7988084895122254,
+      "grad_norm": 0.13407436084230553,
+      "learning_rate": 2.9018939277616887e-06,
+      "loss": 0.7503,
+      "step": 3625
+    },
+    {
+      "epoch": 1.799304952215465,
+      "grad_norm": 0.13291972705380428,
+      "learning_rate": 2.9009292589287357e-06,
+      "loss": 0.7089,
+      "step": 3626
+    },
+    {
+      "epoch": 1.7998014149187043,
+      "grad_norm": 0.13331187297256578,
+      "learning_rate": 2.899964528820175e-06,
+      "loss": 0.7445,
+      "step": 3627
+    },
+    {
+      "epoch": 1.8002978776219436,
+      "grad_norm": 0.13026121253634557,
+      "learning_rate": 2.8989997375834485e-06,
+      "loss": 0.732,
+      "step": 3628
+    },
+    {
+      "epoch": 1.800794340325183,
+      "grad_norm": 0.12960005121948162,
+      "learning_rate": 2.8980348853660096e-06,
+      "loss": 0.7023,
+      "step": 3629
+    },
+    {
+      "epoch": 1.8012908030284225,
+      "grad_norm": 0.13361582398506516,
+      "learning_rate": 2.89706997231532e-06,
+      "loss": 0.7045,
+      "step": 3630
+    },
+    {
+      "epoch": 1.801787265731662,
+      "grad_norm": 0.13469473169085272,
+      "learning_rate": 2.8961049985788524e-06,
+      "loss": 0.7225,
+      "step": 3631
+    },
+    {
+      "epoch": 1.8022837284349014,
+      "grad_norm": 0.13319255321122753,
+      "learning_rate": 2.8951399643040867e-06,
+      "loss": 0.7368,
+      "step": 3632
+    },
+    {
+      "epoch": 1.8027801911381407,
+      "grad_norm": 0.12936117726098809,
+      "learning_rate": 2.894174869638513e-06,
+      "loss": 0.7553,
+      "step": 3633
+    },
+    {
+      "epoch": 1.80327665384138,
+      "grad_norm": 0.12979563347690065,
+      "learning_rate": 2.8932097147296308e-06,
+      "loss": 0.7155,
+      "step": 3634
+    },
+    {
+      "epoch": 1.8037731165446196,
+      "grad_norm": 0.1309079661022396,
+      "learning_rate": 2.8922444997249477e-06,
+      "loss": 0.7244,
+      "step": 3635
+    },
+    {
+      "epoch": 1.8042695792478591,
+      "grad_norm": 0.1427263829925681,
+      "learning_rate": 2.891279224771982e-06,
+      "loss": 0.7333,
+      "step": 3636
+    },
+    {
+      "epoch": 1.8047660419510985,
+      "grad_norm": 0.1407202293176257,
+      "learning_rate": 2.8903138900182615e-06,
+      "loss": 0.7456,
+      "step": 3637
+    },
+    {
+      "epoch": 1.8052625046543378,
+      "grad_norm": 0.16815294681624532,
+      "learning_rate": 2.88934849561132e-06,
+      "loss": 0.6916,
+      "step": 3638
+    },
+    {
+      "epoch": 1.8057589673575771,
+      "grad_norm": 0.1424105389357212,
+      "learning_rate": 2.8883830416987043e-06,
+      "loss": 0.7531,
+      "step": 3639
+    },
+    {
+      "epoch": 1.8062554300608167,
+      "grad_norm": 0.13301046552150622,
+      "learning_rate": 2.887417528427967e-06,
+      "loss": 0.7107,
+      "step": 3640
+    },
+    {
+      "epoch": 1.8067518927640562,
+      "grad_norm": 0.13697780218782096,
+      "learning_rate": 2.8864519559466738e-06,
+      "loss": 0.7519,
+      "step": 3641
+    },
+    {
+      "epoch": 1.8072483554672956,
+      "grad_norm": 0.13703156295639382,
+      "learning_rate": 2.8854863244023945e-06,
+      "loss": 0.7769,
+      "step": 3642
+    },
+    {
+      "epoch": 1.807744818170535,
+      "grad_norm": 0.1322398802236882,
+      "learning_rate": 2.884520633942712e-06,
+      "loss": 0.6999,
+      "step": 3643
+    },
+    {
+      "epoch": 1.8082412808737742,
+      "grad_norm": 0.13492661813080623,
+      "learning_rate": 2.8835548847152143e-06,
+      "loss": 0.7147,
+      "step": 3644
+    },
+    {
+      "epoch": 1.8087377435770138,
+      "grad_norm": 0.13231443100813342,
+      "learning_rate": 2.8825890768675035e-06,
+      "loss": 0.6931,
+      "step": 3645
+    },
+    {
+      "epoch": 1.8092342062802533,
+      "grad_norm": 0.13342027348327531,
+      "learning_rate": 2.8816232105471864e-06,
+      "loss": 0.7234,
+      "step": 3646
+    },
+    {
+      "epoch": 1.8097306689834927,
+      "grad_norm": 0.13502995129426054,
+      "learning_rate": 2.8806572859018806e-06,
+      "loss": 0.7344,
+      "step": 3647
+    },
+    {
+      "epoch": 1.810227131686732,
+      "grad_norm": 0.13216879878124957,
+      "learning_rate": 2.8796913030792116e-06,
+      "loss": 0.7127,
+      "step": 3648
+    },
+    {
+      "epoch": 1.8107235943899713,
+      "grad_norm": 0.1389960465198068,
+      "learning_rate": 2.878725262226816e-06,
+      "loss": 0.7331,
+      "step": 3649
+    },
+    {
+      "epoch": 1.8112200570932109,
+      "grad_norm": 0.133130506783979,
+      "learning_rate": 2.8777591634923353e-06,
+      "loss": 0.7085,
+      "step": 3650
+    },
+    {
+      "epoch": 1.8117165197964504,
+      "grad_norm": 0.12953910081868406,
+      "learning_rate": 2.8767930070234233e-06,
+      "loss": 0.7059,
+      "step": 3651
+    },
+    {
+      "epoch": 1.8122129824996898,
+      "grad_norm": 0.13304246175165207,
+      "learning_rate": 2.8758267929677418e-06,
+      "loss": 0.7506,
+      "step": 3652
+    },
+    {
+      "epoch": 1.812709445202929,
+      "grad_norm": 0.13171896309203177,
+      "learning_rate": 2.874860521472962e-06,
+      "loss": 0.704,
+      "step": 3653
+    },
+    {
+      "epoch": 1.8132059079061684,
+      "grad_norm": 0.12867716679784078,
+      "learning_rate": 2.873894192686761e-06,
+      "loss": 0.7,
+      "step": 3654
+    },
+    {
+      "epoch": 1.813702370609408,
+      "grad_norm": 0.13335540704857535,
+      "learning_rate": 2.8729278067568272e-06,
+      "loss": 0.7251,
+      "step": 3655
+    },
+    {
+      "epoch": 1.8141988333126475,
+      "grad_norm": 0.1335222309537191,
+      "learning_rate": 2.871961363830858e-06,
+      "loss": 0.753,
+      "step": 3656
+    },
+    {
+      "epoch": 1.8146952960158869,
+      "grad_norm": 0.13606044352762497,
+      "learning_rate": 2.8709948640565582e-06,
+      "loss": 0.7602,
+      "step": 3657
+    },
+    {
+      "epoch": 1.8151917587191262,
+      "grad_norm": 0.13248054985778066,
+      "learning_rate": 2.870028307581642e-06,
+      "loss": 0.6717,
+      "step": 3658
+    },
+    {
+      "epoch": 1.8156882214223655,
+      "grad_norm": 0.12524368516267864,
+      "learning_rate": 2.8690616945538324e-06,
+      "loss": 0.6575,
+      "step": 3659
+    },
+    {
+      "epoch": 1.816184684125605,
+      "grad_norm": 0.12893746506812964,
+      "learning_rate": 2.8680950251208595e-06,
+      "loss": 0.6987,
+      "step": 3660
+    },
+    {
+      "epoch": 1.8166811468288446,
+      "grad_norm": 0.13668691367646468,
+      "learning_rate": 2.8671282994304637e-06,
+      "loss": 0.6763,
+      "step": 3661
+    },
+    {
+      "epoch": 1.817177609532084,
+      "grad_norm": 0.13337887117478803,
+      "learning_rate": 2.8661615176303944e-06,
+      "loss": 0.721,
+      "step": 3662
+    },
+    {
+      "epoch": 1.8176740722353233,
+      "grad_norm": 0.137232017944419,
+      "learning_rate": 2.865194679868408e-06,
+      "loss": 0.7203,
+      "step": 3663
+    },
+    {
+      "epoch": 1.8181705349385626,
+      "grad_norm": 0.1283303529499583,
+      "learning_rate": 2.8642277862922702e-06,
+      "loss": 0.7107,
+      "step": 3664
+    },
+    {
+      "epoch": 1.8186669976418022,
+      "grad_norm": 0.14113427375297083,
+      "learning_rate": 2.8632608370497555e-06,
+      "loss": 0.7671,
+      "step": 3665
+    },
+    {
+      "epoch": 1.8191634603450417,
+      "grad_norm": 0.1337368727892662,
+      "learning_rate": 2.862293832288646e-06,
+      "loss": 0.6931,
+      "step": 3666
+    },
+    {
+      "epoch": 1.819659923048281,
+      "grad_norm": 0.1305593466225775,
+      "learning_rate": 2.8613267721567334e-06,
+      "loss": 0.7555,
+      "step": 3667
+    },
+    {
+      "epoch": 1.8201563857515204,
+      "grad_norm": 0.13071527307271635,
+      "learning_rate": 2.8603596568018166e-06,
+      "loss": 0.7269,
+      "step": 3668
+    },
+    {
+      "epoch": 1.8206528484547597,
+      "grad_norm": 0.13020862347403772,
+      "learning_rate": 2.859392486371705e-06,
+      "loss": 0.7363,
+      "step": 3669
+    },
+    {
+      "epoch": 1.8211493111579993,
+      "grad_norm": 0.13695112485546784,
+      "learning_rate": 2.8584252610142133e-06,
+      "loss": 0.763,
+      "step": 3670
+    },
+    {
+      "epoch": 1.8216457738612388,
+      "grad_norm": 0.13490142562797158,
+      "learning_rate": 2.8574579808771676e-06,
+      "loss": 0.6992,
+      "step": 3671
+    },
+    {
+      "epoch": 1.8221422365644782,
+      "grad_norm": 0.12665466211192675,
+      "learning_rate": 2.8564906461084014e-06,
+      "loss": 0.6498,
+      "step": 3672
+    },
+    {
+      "epoch": 1.8226386992677175,
+      "grad_norm": 0.13516985312128005,
+      "learning_rate": 2.855523256855756e-06,
+      "loss": 0.7067,
+      "step": 3673
+    },
+    {
+      "epoch": 1.8231351619709568,
+      "grad_norm": 0.1288573931245718,
+      "learning_rate": 2.8545558132670804e-06,
+      "loss": 0.6909,
+      "step": 3674
+    },
+    {
+      "epoch": 1.8236316246741964,
+      "grad_norm": 0.12713861994971612,
+      "learning_rate": 2.8535883154902345e-06,
+      "loss": 0.6958,
+      "step": 3675
+    },
+    {
+      "epoch": 1.824128087377436,
+      "grad_norm": 0.13381135557162596,
+      "learning_rate": 2.852620763673083e-06,
+      "loss": 0.7299,
+      "step": 3676
+    },
+    {
+      "epoch": 1.8246245500806753,
+      "grad_norm": 0.1319129422115045,
+      "learning_rate": 2.851653157963502e-06,
+      "loss": 0.6831,
+      "step": 3677
+    },
+    {
+      "epoch": 1.8251210127839146,
+      "grad_norm": 0.1394166318448852,
+      "learning_rate": 2.850685498509374e-06,
+      "loss": 0.7536,
+      "step": 3678
+    },
+    {
+      "epoch": 1.825617475487154,
+      "grad_norm": 0.13103025516203937,
+      "learning_rate": 2.84971778545859e-06,
+      "loss": 0.6629,
+      "step": 3679
+    },
+    {
+      "epoch": 1.8261139381903935,
+      "grad_norm": 0.1309333120946672,
+      "learning_rate": 2.8487500189590513e-06,
+      "loss": 0.7068,
+      "step": 3680
+    },
+    {
+      "epoch": 1.826610400893633,
+      "grad_norm": 0.12983925391000833,
+      "learning_rate": 2.847782199158663e-06,
+      "loss": 0.7237,
+      "step": 3681
+    },
+    {
+      "epoch": 1.8271068635968724,
+      "grad_norm": 0.13732026044139703,
+      "learning_rate": 2.8468143262053416e-06,
+      "loss": 0.7156,
+      "step": 3682
+    },
+    {
+      "epoch": 1.8276033263001117,
+      "grad_norm": 0.1364540970178744,
+      "learning_rate": 2.8458464002470114e-06,
+      "loss": 0.7403,
+      "step": 3683
+    },
+    {
+      "epoch": 1.828099789003351,
+      "grad_norm": 0.13692850761642453,
+      "learning_rate": 2.844878421431604e-06,
+      "loss": 0.6859,
+      "step": 3684
+    },
+    {
+      "epoch": 1.8285962517065906,
+      "grad_norm": 0.13088408167168533,
+      "learning_rate": 2.8439103899070595e-06,
+      "loss": 0.6676,
+      "step": 3685
+    },
+    {
+      "epoch": 1.8290927144098301,
+      "grad_norm": 0.12989265875764777,
+      "learning_rate": 2.8429423058213267e-06,
+      "loss": 0.706,
+      "step": 3686
+    },
+    {
+      "epoch": 1.8295891771130695,
+      "grad_norm": 0.1291389182253454,
+      "learning_rate": 2.8419741693223607e-06,
+      "loss": 0.7156,
+      "step": 3687
+    },
+    {
+      "epoch": 1.8300856398163088,
+      "grad_norm": 0.13918286355290396,
+      "learning_rate": 2.8410059805581258e-06,
+      "loss": 0.7665,
+      "step": 3688
+    },
+    {
+      "epoch": 1.8305821025195481,
+      "grad_norm": 0.13286704143055944,
+      "learning_rate": 2.840037739676595e-06,
+      "loss": 0.6953,
+      "step": 3689
+    },
+    {
+      "epoch": 1.8310785652227877,
+      "grad_norm": 0.13464853923111927,
+      "learning_rate": 2.8390694468257474e-06,
+      "loss": 0.7558,
+      "step": 3690
+    },
+    {
+      "epoch": 1.8315750279260272,
+      "grad_norm": 0.1338921743633815,
+      "learning_rate": 2.838101102153572e-06,
+      "loss": 0.7216,
+      "step": 3691
+    },
+    {
+      "epoch": 1.8320714906292666,
+      "grad_norm": 0.1332544592794539,
+      "learning_rate": 2.8371327058080634e-06,
+      "loss": 0.7237,
+      "step": 3692
+    },
+    {
+      "epoch": 1.8325679533325059,
+      "grad_norm": 0.13859871403436683,
+      "learning_rate": 2.836164257937226e-06,
+      "loss": 0.6647,
+      "step": 3693
+    },
+    {
+      "epoch": 1.8330644160357452,
+      "grad_norm": 0.13773496966704152,
+      "learning_rate": 2.8351957586890724e-06,
+      "loss": 0.7558,
+      "step": 3694
+    },
+    {
+      "epoch": 1.8335608787389848,
+      "grad_norm": 0.1363040721397413,
+      "learning_rate": 2.8342272082116214e-06,
+      "loss": 0.7176,
+      "step": 3695
+    },
+    {
+      "epoch": 1.834057341442224,
+      "grad_norm": 0.13238261911259094,
+      "learning_rate": 2.833258606652901e-06,
+      "loss": 0.7428,
+      "step": 3696
+    },
+    {
+      "epoch": 1.8345538041454637,
+      "grad_norm": 0.1289742090498221,
+      "learning_rate": 2.8322899541609457e-06,
+      "loss": 0.7474,
+      "step": 3697
+    },
+    {
+      "epoch": 1.835050266848703,
+      "grad_norm": 0.13229223790941982,
+      "learning_rate": 2.8313212508837985e-06,
+      "loss": 0.6888,
+      "step": 3698
+    },
+    {
+      "epoch": 1.8355467295519423,
+      "grad_norm": 0.1405977314818644,
+      "learning_rate": 2.83035249696951e-06,
+      "loss": 0.7118,
+      "step": 3699
+    },
+    {
+      "epoch": 1.8360431922551819,
+      "grad_norm": 0.1322368134123888,
+      "learning_rate": 2.82938369256614e-06,
+      "loss": 0.7254,
+      "step": 3700
+    },
+    {
+      "epoch": 1.8365396549584212,
+      "grad_norm": 0.13239550569763764,
+      "learning_rate": 2.828414837821753e-06,
+      "loss": 0.7031,
+      "step": 3701
+    },
+    {
+      "epoch": 1.8370361176616608,
+      "grad_norm": 0.12443941139663224,
+      "learning_rate": 2.827445932884425e-06,
+      "loss": 0.6938,
+      "step": 3702
+    },
+    {
+      "epoch": 1.8375325803649,
+      "grad_norm": 0.13533098431151683,
+      "learning_rate": 2.8264769779022355e-06,
+      "loss": 0.715,
+      "step": 3703
+    },
+    {
+      "epoch": 1.8380290430681394,
+      "grad_norm": 0.13336648970273562,
+      "learning_rate": 2.8255079730232742e-06,
+      "loss": 0.7551,
+      "step": 3704
+    },
+    {
+      "epoch": 1.838525505771379,
+      "grad_norm": 0.12584694627355517,
+      "learning_rate": 2.8245389183956395e-06,
+      "loss": 0.686,
+      "step": 3705
+    },
+    {
+      "epoch": 1.8390219684746183,
+      "grad_norm": 0.1345026841457402,
+      "learning_rate": 2.8235698141674338e-06,
+      "loss": 0.7361,
+      "step": 3706
+    },
+    {
+      "epoch": 1.8395184311778578,
+      "grad_norm": 0.1303560374635088,
+      "learning_rate": 2.8226006604867705e-06,
+      "loss": 0.7158,
+      "step": 3707
+    },
+    {
+      "epoch": 1.8400148938810972,
+      "grad_norm": 0.13250398945920666,
+      "learning_rate": 2.821631457501769e-06,
+      "loss": 0.7667,
+      "step": 3708
+    },
+    {
+      "epoch": 1.8405113565843365,
+      "grad_norm": 0.13580387586544665,
+      "learning_rate": 2.820662205360555e-06,
+      "loss": 0.7681,
+      "step": 3709
+    },
+    {
+      "epoch": 1.841007819287576,
+      "grad_norm": 0.13301684404493566,
+      "learning_rate": 2.8196929042112652e-06,
+      "loss": 0.7285,
+      "step": 3710
+    },
+    {
+      "epoch": 1.8415042819908154,
+      "grad_norm": 0.13473692403172938,
+      "learning_rate": 2.818723554202041e-06,
+      "loss": 0.721,
+      "step": 3711
+    },
+    {
+      "epoch": 1.842000744694055,
+      "grad_norm": 0.12587991568709622,
+      "learning_rate": 2.817754155481032e-06,
+      "loss": 0.7165,
+      "step": 3712
+    },
+    {
+      "epoch": 1.8424972073972943,
+      "grad_norm": 0.13035348271632663,
+      "learning_rate": 2.816784708196395e-06,
+      "loss": 0.7185,
+      "step": 3713
+    },
+    {
+      "epoch": 1.8429936701005336,
+      "grad_norm": 0.15215075104364917,
+      "learning_rate": 2.815815212496294e-06,
+      "loss": 0.763,
+      "step": 3714
+    },
+    {
+      "epoch": 1.8434901328037732,
+      "grad_norm": 0.1373937835941121,
+      "learning_rate": 2.8148456685289016e-06,
+      "loss": 0.7423,
+      "step": 3715
+    },
+    {
+      "epoch": 1.8439865955070125,
+      "grad_norm": 0.12962189019847445,
+      "learning_rate": 2.813876076442397e-06,
+      "loss": 0.7229,
+      "step": 3716
+    },
+    {
+      "epoch": 1.844483058210252,
+      "grad_norm": 0.12465566308505892,
+      "learning_rate": 2.8129064363849674e-06,
+      "loss": 0.664,
+      "step": 3717
+    },
+    {
+      "epoch": 1.8449795209134914,
+      "grad_norm": 0.1320992278623956,
+      "learning_rate": 2.811936748504806e-06,
+      "loss": 0.7538,
+      "step": 3718
+    },
+    {
+      "epoch": 1.8454759836167307,
+      "grad_norm": 0.1332828335116305,
+      "learning_rate": 2.810967012950113e-06,
+      "loss": 0.7239,
+      "step": 3719
+    },
+    {
+      "epoch": 1.8459724463199703,
+      "grad_norm": 0.12797528135715117,
+      "learning_rate": 2.809997229869099e-06,
+      "loss": 0.7122,
+      "step": 3720
+    },
+    {
+      "epoch": 1.8464689090232096,
+      "grad_norm": 0.1301117350778891,
+      "learning_rate": 2.8090273994099793e-06,
+      "loss": 0.72,
+      "step": 3721
+    },
+    {
+      "epoch": 1.8469653717264491,
+      "grad_norm": 0.1300300122398293,
+      "learning_rate": 2.8080575217209756e-06,
+      "loss": 0.6907,
+      "step": 3722
+    },
+    {
+      "epoch": 1.8474618344296885,
+      "grad_norm": 0.1331300565652563,
+      "learning_rate": 2.8070875969503194e-06,
+      "loss": 0.7219,
+      "step": 3723
+    },
+    {
+      "epoch": 1.8479582971329278,
+      "grad_norm": 0.13587506563623994,
+      "learning_rate": 2.8061176252462473e-06,
+      "loss": 0.7145,
+      "step": 3724
+    },
+    {
+      "epoch": 1.8484547598361674,
+      "grad_norm": 0.12978747829512013,
+      "learning_rate": 2.805147606757005e-06,
+      "loss": 0.7188,
+      "step": 3725
+    },
+    {
+      "epoch": 1.8489512225394067,
+      "grad_norm": 0.12938008759831762,
+      "learning_rate": 2.804177541630843e-06,
+      "loss": 0.69,
+      "step": 3726
+    },
+    {
+      "epoch": 1.8494476852426462,
+      "grad_norm": 0.13496840351576017,
+      "learning_rate": 2.803207430016021e-06,
+      "loss": 0.7572,
+      "step": 3727
+    },
+    {
+      "epoch": 1.8499441479458856,
+      "grad_norm": 0.13064015668880996,
+      "learning_rate": 2.802237272060806e-06,
+      "loss": 0.7055,
+      "step": 3728
+    },
+    {
+      "epoch": 1.850440610649125,
+      "grad_norm": 0.1310236785495075,
+      "learning_rate": 2.8012670679134694e-06,
+      "loss": 0.6944,
+      "step": 3729
+    },
+    {
+      "epoch": 1.8509370733523642,
+      "grad_norm": 0.13058022532167907,
+      "learning_rate": 2.8002968177222916e-06,
+      "loss": 0.7276,
+      "step": 3730
+    },
+    {
+      "epoch": 1.8514335360556038,
+      "grad_norm": 0.1399741355065743,
+      "learning_rate": 2.7993265216355597e-06,
+      "loss": 0.7019,
+      "step": 3731
+    },
+    {
+      "epoch": 1.8519299987588433,
+      "grad_norm": 0.1309873590126447,
+      "learning_rate": 2.798356179801569e-06,
+      "loss": 0.7323,
+      "step": 3732
+    },
+    {
+      "epoch": 1.8524264614620827,
+      "grad_norm": 0.12823473606854824,
+      "learning_rate": 2.7973857923686192e-06,
+      "loss": 0.6766,
+      "step": 3733
+    },
+    {
+      "epoch": 1.852922924165322,
+      "grad_norm": 0.12599315475512168,
+      "learning_rate": 2.7964153594850207e-06,
+      "loss": 0.6611,
+      "step": 3734
+    },
+    {
+      "epoch": 1.8534193868685613,
+      "grad_norm": 0.13218616368397323,
+      "learning_rate": 2.7954448812990857e-06,
+      "loss": 0.7666,
+      "step": 3735
+    },
+    {
+      "epoch": 1.853915849571801,
+      "grad_norm": 0.13258167830517983,
+      "learning_rate": 2.7944743579591383e-06,
+      "loss": 0.7688,
+      "step": 3736
+    },
+    {
+      "epoch": 1.8544123122750404,
+      "grad_norm": 0.13636167071644384,
+      "learning_rate": 2.793503789613507e-06,
+      "loss": 0.7893,
+      "step": 3737
+    },
+    {
+      "epoch": 1.8549087749782798,
+      "grad_norm": 0.1353506593944718,
+      "learning_rate": 2.7925331764105272e-06,
+      "loss": 0.718,
+      "step": 3738
+    },
+    {
+      "epoch": 1.855405237681519,
+      "grad_norm": 0.13351328738260573,
+      "learning_rate": 2.791562518498542e-06,
+      "loss": 0.72,
+      "step": 3739
+    },
+    {
+      "epoch": 1.8559017003847584,
+      "grad_norm": 0.12916911002443757,
+      "learning_rate": 2.7905918160259005e-06,
+      "loss": 0.6834,
+      "step": 3740
+    },
+    {
+      "epoch": 1.856398163087998,
+      "grad_norm": 0.13034694869758517,
+      "learning_rate": 2.789621069140959e-06,
+      "loss": 0.7152,
+      "step": 3741
+    },
+    {
+      "epoch": 1.8568946257912375,
+      "grad_norm": 0.1358558324112095,
+      "learning_rate": 2.788650277992081e-06,
+      "loss": 0.7477,
+      "step": 3742
+    },
+    {
+      "epoch": 1.8573910884944769,
+      "grad_norm": 0.13072120609339014,
+      "learning_rate": 2.7876794427276362e-06,
+      "loss": 0.7296,
+      "step": 3743
+    },
+    {
+      "epoch": 1.8578875511977162,
+      "grad_norm": 0.12977265231141502,
+      "learning_rate": 2.786708563496002e-06,
+      "loss": 0.689,
+      "step": 3744
+    },
+    {
+      "epoch": 1.8583840139009555,
+      "grad_norm": 0.14176411747850756,
+      "learning_rate": 2.78573764044556e-06,
+      "loss": 0.7371,
+      "step": 3745
+    },
+    {
+      "epoch": 1.858880476604195,
+      "grad_norm": 0.13585066242740843,
+      "learning_rate": 2.7847666737247008e-06,
+      "loss": 0.7051,
+      "step": 3746
+    },
+    {
+      "epoch": 1.8593769393074346,
+      "grad_norm": 0.12837037510490845,
+      "learning_rate": 2.783795663481822e-06,
+      "loss": 0.7081,
+      "step": 3747
+    },
+    {
+      "epoch": 1.859873402010674,
+      "grad_norm": 0.13389138484666943,
+      "learning_rate": 2.7828246098653255e-06,
+      "loss": 0.7607,
+      "step": 3748
+    },
+    {
+      "epoch": 1.8603698647139133,
+      "grad_norm": 0.13565169731031765,
+      "learning_rate": 2.781853513023623e-06,
+      "loss": 0.7119,
+      "step": 3749
+    },
+    {
+      "epoch": 1.8608663274171526,
+      "grad_norm": 0.13181731856838821,
+      "learning_rate": 2.7808823731051306e-06,
+      "loss": 0.7253,
+      "step": 3750
+    },
+    {
+      "epoch": 1.8613627901203922,
+      "grad_norm": 0.1341896017701526,
+      "learning_rate": 2.7799111902582697e-06,
+      "loss": 0.6954,
+      "step": 3751
+    },
+    {
+      "epoch": 1.8618592528236317,
+      "grad_norm": 0.13618587231085355,
+      "learning_rate": 2.7789399646314723e-06,
+      "loss": 0.7345,
+      "step": 3752
+    },
+    {
+      "epoch": 1.862355715526871,
+      "grad_norm": 0.1369640920886463,
+      "learning_rate": 2.7779686963731738e-06,
+      "loss": 0.7393,
+      "step": 3753
+    },
+    {
+      "epoch": 1.8628521782301104,
+      "grad_norm": 0.13390573696400448,
+      "learning_rate": 2.7769973856318167e-06,
+      "loss": 0.725,
+      "step": 3754
+    },
+    {
+      "epoch": 1.8633486409333497,
+      "grad_norm": 0.1336592460566132,
+      "learning_rate": 2.7760260325558507e-06,
+      "loss": 0.6977,
+      "step": 3755
+    },
+    {
+      "epoch": 1.8638451036365893,
+      "grad_norm": 0.13604007586743042,
+      "learning_rate": 2.7750546372937315e-06,
+      "loss": 0.7183,
+      "step": 3756
+    },
+    {
+      "epoch": 1.8643415663398288,
+      "grad_norm": 0.1357385191072151,
+      "learning_rate": 2.774083199993921e-06,
+      "loss": 0.7517,
+      "step": 3757
+    },
+    {
+      "epoch": 1.8648380290430682,
+      "grad_norm": 0.1301840034604384,
+      "learning_rate": 2.7731117208048875e-06,
+      "loss": 0.6671,
+      "step": 3758
+    },
+    {
+      "epoch": 1.8653344917463075,
+      "grad_norm": 0.13571137517897333,
+      "learning_rate": 2.772140199875107e-06,
+      "loss": 0.7193,
+      "step": 3759
+    },
+    {
+      "epoch": 1.8658309544495468,
+      "grad_norm": 0.13744700754172479,
+      "learning_rate": 2.77116863735306e-06,
+      "loss": 0.7301,
+      "step": 3760
+    },
+    {
+      "epoch": 1.8663274171527864,
+      "grad_norm": 0.1338974951951887,
+      "learning_rate": 2.7701970333872354e-06,
+      "loss": 0.6994,
+      "step": 3761
+    },
+    {
+      "epoch": 1.866823879856026,
+      "grad_norm": 0.12864049194839253,
+      "learning_rate": 2.769225388126126e-06,
+      "loss": 0.693,
+      "step": 3762
+    },
+    {
+      "epoch": 1.8673203425592653,
+      "grad_norm": 0.13155044309844505,
+      "learning_rate": 2.7682537017182326e-06,
+      "loss": 0.721,
+      "step": 3763
+    },
+    {
+      "epoch": 1.8678168052625046,
+      "grad_norm": 0.1347463327038866,
+      "learning_rate": 2.767281974312062e-06,
+      "loss": 0.712,
+      "step": 3764
+    },
+    {
+      "epoch": 1.868313267965744,
+      "grad_norm": 0.12971495960250434,
+      "learning_rate": 2.7663102060561274e-06,
+      "loss": 0.7143,
+      "step": 3765
+    },
+    {
+      "epoch": 1.8688097306689835,
+      "grad_norm": 0.13510904366559584,
+      "learning_rate": 2.7653383970989477e-06,
+      "loss": 0.7057,
+      "step": 3766
+    },
+    {
+      "epoch": 1.869306193372223,
+      "grad_norm": 0.13401673104563622,
+      "learning_rate": 2.7643665475890484e-06,
+      "loss": 0.6864,
+      "step": 3767
+    },
+    {
+      "epoch": 1.8698026560754624,
+      "grad_norm": 0.13210648198821992,
+      "learning_rate": 2.763394657674961e-06,
+      "loss": 0.7038,
+      "step": 3768
+    },
+    {
+      "epoch": 1.8702991187787017,
+      "grad_norm": 0.1318377078194905,
+      "learning_rate": 2.762422727505224e-06,
+      "loss": 0.7596,
+      "step": 3769
+    },
+    {
+      "epoch": 1.870795581481941,
+      "grad_norm": 0.13025033485200868,
+      "learning_rate": 2.76145075722838e-06,
+      "loss": 0.7144,
+      "step": 3770
+    },
+    {
+      "epoch": 1.8712920441851806,
+      "grad_norm": 0.14491613825665298,
+      "learning_rate": 2.76047874699298e-06,
+      "loss": 0.7712,
+      "step": 3771
+    },
+    {
+      "epoch": 1.8717885068884201,
+      "grad_norm": 0.13563190298000607,
+      "learning_rate": 2.75950669694758e-06,
+      "loss": 0.7374,
+      "step": 3772
+    },
+    {
+      "epoch": 1.8722849695916595,
+      "grad_norm": 0.1320789962394144,
+      "learning_rate": 2.7585346072407422e-06,
+      "loss": 0.7426,
+      "step": 3773
+    },
+    {
+      "epoch": 1.8727814322948988,
+      "grad_norm": 0.13057853212026468,
+      "learning_rate": 2.757562478021035e-06,
+      "loss": 0.7262,
+      "step": 3774
+    },
+    {
+      "epoch": 1.8732778949981381,
+      "grad_norm": 0.13102963708589455,
+      "learning_rate": 2.756590309437033e-06,
+      "loss": 0.7011,
+      "step": 3775
+    },
+    {
+      "epoch": 1.8737743577013777,
+      "grad_norm": 0.13375423324978467,
+      "learning_rate": 2.755618101637315e-06,
+      "loss": 0.718,
+      "step": 3776
+    },
+    {
+      "epoch": 1.8742708204046172,
+      "grad_norm": 0.12830373485814348,
+      "learning_rate": 2.754645854770471e-06,
+      "loss": 0.7265,
+      "step": 3777
+    },
+    {
+      "epoch": 1.8747672831078566,
+      "grad_norm": 0.13336641643595246,
+      "learning_rate": 2.75367356898509e-06,
+      "loss": 0.6742,
+      "step": 3778
+    },
+    {
+      "epoch": 1.875263745811096,
+      "grad_norm": 0.13157972272739404,
+      "learning_rate": 2.7527012444297707e-06,
+      "loss": 0.7321,
+      "step": 3779
+    },
+    {
+      "epoch": 1.8757602085143352,
+      "grad_norm": 0.14208051021228643,
+      "learning_rate": 2.751728881253118e-06,
+      "loss": 0.7672,
+      "step": 3780
+    },
+    {
+      "epoch": 1.8762566712175748,
+      "grad_norm": 0.1371097076190842,
+      "learning_rate": 2.7507564796037424e-06,
+      "loss": 0.7512,
+      "step": 3781
+    },
+    {
+      "epoch": 1.8767531339208143,
+      "grad_norm": 0.13188702665160387,
+      "learning_rate": 2.7497840396302596e-06,
+      "loss": 0.7401,
+      "step": 3782
+    },
+    {
+      "epoch": 1.8772495966240537,
+      "grad_norm": 0.13500026817249772,
+      "learning_rate": 2.748811561481291e-06,
+      "loss": 0.7295,
+      "step": 3783
+    },
+    {
+      "epoch": 1.877746059327293,
+      "grad_norm": 0.13273560547309662,
+      "learning_rate": 2.7478390453054645e-06,
+      "loss": 0.7283,
+      "step": 3784
+    },
+    {
+      "epoch": 1.8782425220305323,
+      "grad_norm": 0.13872675617089372,
+      "learning_rate": 2.746866491251414e-06,
+      "loss": 0.7602,
+      "step": 3785
+    },
+    {
+      "epoch": 1.8787389847337719,
+      "grad_norm": 0.12924106818518485,
+      "learning_rate": 2.7458938994677784e-06,
+      "loss": 0.7546,
+      "step": 3786
+    },
+    {
+      "epoch": 1.8792354474370114,
+      "grad_norm": 0.136941377040484,
+      "learning_rate": 2.744921270103203e-06,
+      "loss": 0.7579,
+      "step": 3787
+    },
+    {
+      "epoch": 1.8797319101402508,
+      "grad_norm": 0.1298041641175948,
+      "learning_rate": 2.743948603306339e-06,
+      "loss": 0.7051,
+      "step": 3788
+    },
+    {
+      "epoch": 1.88022837284349,
+      "grad_norm": 0.13832860111903247,
+      "learning_rate": 2.7429758992258416e-06,
+      "loss": 0.7243,
+      "step": 3789
+    },
+    {
+      "epoch": 1.8807248355467294,
+      "grad_norm": 0.13402102344704708,
+      "learning_rate": 2.7420031580103736e-06,
+      "loss": 0.7225,
+      "step": 3790
+    },
+    {
+      "epoch": 1.881221298249969,
+      "grad_norm": 0.13987575442514208,
+      "learning_rate": 2.7410303798086034e-06,
+      "loss": 0.7264,
+      "step": 3791
+    },
+    {
+      "epoch": 1.8817177609532085,
+      "grad_norm": 0.13290550293386685,
+      "learning_rate": 2.7400575647692046e-06,
+      "loss": 0.7677,
+      "step": 3792
+    },
+    {
+      "epoch": 1.8822142236564479,
+      "grad_norm": 0.13083520400866375,
+      "learning_rate": 2.739084713040856e-06,
+      "loss": 0.7193,
+      "step": 3793
+    },
+    {
+      "epoch": 1.8827106863596872,
+      "grad_norm": 0.1282525167355114,
+      "learning_rate": 2.7381118247722427e-06,
+      "loss": 0.6802,
+      "step": 3794
+    },
+    {
+      "epoch": 1.8832071490629265,
+      "grad_norm": 0.13627486143137021,
+      "learning_rate": 2.7371389001120545e-06,
+      "loss": 0.722,
+      "step": 3795
+    },
+    {
+      "epoch": 1.883703611766166,
+      "grad_norm": 0.1277896117884158,
+      "learning_rate": 2.736165939208987e-06,
+      "loss": 0.6816,
+      "step": 3796
+    },
+    {
+      "epoch": 1.8842000744694056,
+      "grad_norm": 0.12853176516695636,
+      "learning_rate": 2.735192942211743e-06,
+      "loss": 0.7553,
+      "step": 3797
+    },
+    {
+      "epoch": 1.884696537172645,
+      "grad_norm": 0.13015691759857087,
+      "learning_rate": 2.7342199092690284e-06,
+      "loss": 0.7142,
+      "step": 3798
+    },
+    {
+      "epoch": 1.8851929998758843,
+      "grad_norm": 0.13097419516180478,
+      "learning_rate": 2.733246840529557e-06,
+      "loss": 0.7132,
+      "step": 3799
+    },
+    {
+      "epoch": 1.8856894625791236,
+      "grad_norm": 0.12724020311608158,
+      "learning_rate": 2.7322737361420454e-06,
+      "loss": 0.6748,
+      "step": 3800
+    },
+    {
+      "epoch": 1.8861859252823632,
+      "grad_norm": 0.1270428675400565,
+      "learning_rate": 2.7313005962552174e-06,
+      "loss": 0.7363,
+      "step": 3801
+    },
+    {
+      "epoch": 1.8866823879856027,
+      "grad_norm": 0.13592361098777284,
+      "learning_rate": 2.7303274210178023e-06,
+      "loss": 0.7309,
+      "step": 3802
+    },
+    {
+      "epoch": 1.887178850688842,
+      "grad_norm": 0.13079941165628042,
+      "learning_rate": 2.729354210578533e-06,
+      "loss": 0.7352,
+      "step": 3803
+    },
+    {
+      "epoch": 1.8876753133920814,
+      "grad_norm": 0.1340097748828976,
+      "learning_rate": 2.7283809650861508e-06,
+      "loss": 0.7285,
+      "step": 3804
+    },
+    {
+      "epoch": 1.8881717760953207,
+      "grad_norm": 0.13773420298973918,
+      "learning_rate": 2.727407684689399e-06,
+      "loss": 0.7649,
+      "step": 3805
+    },
+    {
+      "epoch": 1.8886682387985603,
+      "grad_norm": 0.133353759836882,
+      "learning_rate": 2.7264343695370294e-06,
+      "loss": 0.7464,
+      "step": 3806
+    },
+    {
+      "epoch": 1.8891647015017998,
+      "grad_norm": 0.131033603827758,
+      "learning_rate": 2.725461019777797e-06,
+      "loss": 0.6845,
+      "step": 3807
+    },
+    {
+      "epoch": 1.8896611642050392,
+      "grad_norm": 0.15770720974803792,
+      "learning_rate": 2.7244876355604627e-06,
+      "loss": 0.7314,
+      "step": 3808
+    },
+    {
+      "epoch": 1.8901576269082785,
+      "grad_norm": 0.1304863297812739,
+      "learning_rate": 2.723514217033793e-06,
+      "loss": 0.7471,
+      "step": 3809
+    },
+    {
+      "epoch": 1.8906540896115178,
+      "grad_norm": 0.13132658095809105,
+      "learning_rate": 2.722540764346559e-06,
+      "loss": 0.7382,
+      "step": 3810
+    },
+    {
+      "epoch": 1.8911505523147574,
+      "grad_norm": 0.13948668806141548,
+      "learning_rate": 2.7215672776475373e-06,
+      "loss": 0.7345,
+      "step": 3811
+    },
+    {
+      "epoch": 1.891647015017997,
+      "grad_norm": 0.14071216008468773,
+      "learning_rate": 2.720593757085509e-06,
+      "loss": 0.7849,
+      "step": 3812
+    },
+    {
+      "epoch": 1.8921434777212363,
+      "grad_norm": 0.1328645102350295,
+      "learning_rate": 2.719620202809262e-06,
+      "loss": 0.7345,
+      "step": 3813
+    },
+    {
+      "epoch": 1.8926399404244756,
+      "grad_norm": 0.1323492318218001,
+      "learning_rate": 2.718646614967589e-06,
+      "loss": 0.7121,
+      "step": 3814
+    },
+    {
+      "epoch": 1.893136403127715,
+      "grad_norm": 0.13311000745662738,
+      "learning_rate": 2.7176729937092868e-06,
+      "loss": 0.6856,
+      "step": 3815
+    },
+    {
+      "epoch": 1.8936328658309545,
+      "grad_norm": 0.14096801189996547,
+      "learning_rate": 2.716699339183157e-06,
+      "loss": 0.6663,
+      "step": 3816
+    },
+    {
+      "epoch": 1.894129328534194,
+      "grad_norm": 0.13592293501407796,
+      "learning_rate": 2.7157256515380075e-06,
+      "loss": 0.7389,
+      "step": 3817
+    },
+    {
+      "epoch": 1.8946257912374334,
+      "grad_norm": 0.13440467247038096,
+      "learning_rate": 2.7147519309226524e-06,
+      "loss": 0.7136,
+      "step": 3818
+    },
+    {
+      "epoch": 1.8951222539406727,
+      "grad_norm": 0.13477075791277746,
+      "learning_rate": 2.713778177485906e-06,
+      "loss": 0.7187,
+      "step": 3819
+    },
+    {
+      "epoch": 1.895618716643912,
+      "grad_norm": 0.1273414392999846,
+      "learning_rate": 2.712804391376594e-06,
+      "loss": 0.742,
+      "step": 3820
+    },
+    {
+      "epoch": 1.8961151793471516,
+      "grad_norm": 0.13460725669842077,
+      "learning_rate": 2.7118305727435433e-06,
+      "loss": 0.7034,
+      "step": 3821
+    },
+    {
+      "epoch": 1.8966116420503911,
+      "grad_norm": 0.13175344309138737,
+      "learning_rate": 2.710856721735585e-06,
+      "loss": 0.6716,
+      "step": 3822
+    },
+    {
+      "epoch": 1.8971081047536305,
+      "grad_norm": 0.12573529615472806,
+      "learning_rate": 2.709882838501558e-06,
+      "loss": 0.6647,
+      "step": 3823
+    },
+    {
+      "epoch": 1.8976045674568698,
+      "grad_norm": 0.1464775291640466,
+      "learning_rate": 2.7089089231903045e-06,
+      "loss": 0.7118,
+      "step": 3824
+    },
+    {
+      "epoch": 1.8981010301601091,
+      "grad_norm": 0.14724481437959286,
+      "learning_rate": 2.707934975950672e-06,
+      "loss": 0.7267,
+      "step": 3825
+    },
+    {
+      "epoch": 1.8985974928633487,
+      "grad_norm": 0.12701179330406961,
+      "learning_rate": 2.706960996931512e-06,
+      "loss": 0.7076,
+      "step": 3826
+    },
+    {
+      "epoch": 1.8990939555665882,
+      "grad_norm": 0.13539179556079792,
+      "learning_rate": 2.7059869862816817e-06,
+      "loss": 0.7247,
+      "step": 3827
+    },
+    {
+      "epoch": 1.8995904182698276,
+      "grad_norm": 0.13682853094889003,
+      "learning_rate": 2.7050129441500437e-06,
+      "loss": 0.7146,
+      "step": 3828
+    },
+    {
+      "epoch": 1.9000868809730669,
+      "grad_norm": 0.13008454352233842,
+      "learning_rate": 2.7040388706854636e-06,
+      "loss": 0.7115,
+      "step": 3829
+    },
+    {
+      "epoch": 1.9005833436763062,
+      "grad_norm": 0.13324370142235686,
+      "learning_rate": 2.703064766036814e-06,
+      "loss": 0.6941,
+      "step": 3830
+    },
+    {
+      "epoch": 1.9010798063795458,
+      "grad_norm": 0.1379864973843644,
+      "learning_rate": 2.7020906303529722e-06,
+      "loss": 0.7258,
+      "step": 3831
+    },
+    {
+      "epoch": 1.9015762690827853,
+      "grad_norm": 0.13430922945541224,
+      "learning_rate": 2.701116463782816e-06,
+      "loss": 0.717,
+      "step": 3832
+    },
+    {
+      "epoch": 1.9020727317860247,
+      "grad_norm": 0.13535178628287625,
+      "learning_rate": 2.7001422664752338e-06,
+      "loss": 0.7187,
+      "step": 3833
+    },
+    {
+      "epoch": 1.902569194489264,
+      "grad_norm": 0.13401050040469997,
+      "learning_rate": 2.6991680385791154e-06,
+      "loss": 0.6976,
+      "step": 3834
+    },
+    {
+      "epoch": 1.9030656571925033,
+      "grad_norm": 0.13110828507944752,
+      "learning_rate": 2.698193780243355e-06,
+      "loss": 0.7308,
+      "step": 3835
+    },
+    {
+      "epoch": 1.9035621198957429,
+      "grad_norm": 0.1330174198944991,
+      "learning_rate": 2.6972194916168533e-06,
+      "loss": 0.68,
+      "step": 3836
+    },
+    {
+      "epoch": 1.9040585825989822,
+      "grad_norm": 0.129123889581125,
+      "learning_rate": 2.696245172848515e-06,
+      "loss": 0.6816,
+      "step": 3837
+    },
+    {
+      "epoch": 1.9045550453022217,
+      "grad_norm": 0.1323170000848133,
+      "learning_rate": 2.6952708240872477e-06,
+      "loss": 0.74,
+      "step": 3838
+    },
+    {
+      "epoch": 1.905051508005461,
+      "grad_norm": 0.1281877570650443,
+      "learning_rate": 2.6942964454819663e-06,
+      "loss": 0.6716,
+      "step": 3839
+    },
+    {
+      "epoch": 1.9055479707087004,
+      "grad_norm": 0.1273111925333337,
+      "learning_rate": 2.693322037181588e-06,
+      "loss": 0.7161,
+      "step": 3840
+    },
+    {
+      "epoch": 1.90604443341194,
+      "grad_norm": 0.13411942656602813,
+      "learning_rate": 2.692347599335037e-06,
+      "loss": 0.7475,
+      "step": 3841
+    },
+    {
+      "epoch": 1.9065408961151793,
+      "grad_norm": 0.13465714368570397,
+      "learning_rate": 2.69137313209124e-06,
+      "loss": 0.6818,
+      "step": 3842
+    },
+    {
+      "epoch": 1.9070373588184188,
+      "grad_norm": 0.13693810595779668,
+      "learning_rate": 2.6903986355991267e-06,
+      "loss": 0.7217,
+      "step": 3843
+    },
+    {
+      "epoch": 1.9075338215216582,
+      "grad_norm": 0.13227383220776676,
+      "learning_rate": 2.6894241100076356e-06,
+      "loss": 0.7046,
+      "step": 3844
+    },
+    {
+      "epoch": 1.9080302842248975,
+      "grad_norm": 0.13024783516033986,
+      "learning_rate": 2.6884495554657057e-06,
+      "loss": 0.7341,
+      "step": 3845
+    },
+    {
+      "epoch": 1.908526746928137,
+      "grad_norm": 0.13571948071256487,
+      "learning_rate": 2.687474972122283e-06,
+      "loss": 0.7179,
+      "step": 3846
+    },
+    {
+      "epoch": 1.9090232096313764,
+      "grad_norm": 0.14165402752990086,
+      "learning_rate": 2.6865003601263177e-06,
+      "loss": 0.7488,
+      "step": 3847
+    },
+    {
+      "epoch": 1.909519672334616,
+      "grad_norm": 0.13217107325979766,
+      "learning_rate": 2.685525719626762e-06,
+      "loss": 0.7316,
+      "step": 3848
+    },
+    {
+      "epoch": 1.9100161350378553,
+      "grad_norm": 0.12981859327701908,
+      "learning_rate": 2.6845510507725747e-06,
+      "loss": 0.6838,
+      "step": 3849
+    },
+    {
+      "epoch": 1.9105125977410946,
+      "grad_norm": 0.13278332867245013,
+      "learning_rate": 2.6835763537127186e-06,
+      "loss": 0.698,
+      "step": 3850
+    },
+    {
+      "epoch": 1.9110090604443342,
+      "grad_norm": 0.12972846902968782,
+      "learning_rate": 2.68260162859616e-06,
+      "loss": 0.6937,
+      "step": 3851
+    },
+    {
+      "epoch": 1.9115055231475735,
+      "grad_norm": 0.13499670331712021,
+      "learning_rate": 2.68162687557187e-06,
+      "loss": 0.7544,
+      "step": 3852
+    },
+    {
+      "epoch": 1.912001985850813,
+      "grad_norm": 0.13281988396086394,
+      "learning_rate": 2.680652094788825e-06,
+      "loss": 0.7012,
+      "step": 3853
+    },
+    {
+      "epoch": 1.9124984485540524,
+      "grad_norm": 0.13640084140042943,
+      "learning_rate": 2.679677286396003e-06,
+      "loss": 0.7506,
+      "step": 3854
+    },
+    {
+      "epoch": 1.9129949112572917,
+      "grad_norm": 0.1385667059889805,
+      "learning_rate": 2.678702450542389e-06,
+      "loss": 0.7315,
+      "step": 3855
+    },
+    {
+      "epoch": 1.9134913739605313,
+      "grad_norm": 0.1340058683822914,
+      "learning_rate": 2.6777275873769703e-06,
+      "loss": 0.7702,
+      "step": 3856
+    },
+    {
+      "epoch": 1.9139878366637706,
+      "grad_norm": 0.13229770495593288,
+      "learning_rate": 2.67675269704874e-06,
+      "loss": 0.688,
+      "step": 3857
+    },
+    {
+      "epoch": 1.9144842993670101,
+      "grad_norm": 0.13565711456891463,
+      "learning_rate": 2.6757777797066947e-06,
+      "loss": 0.7322,
+      "step": 3858
+    },
+    {
+      "epoch": 1.9149807620702495,
+      "grad_norm": 0.12998843981576316,
+      "learning_rate": 2.6748028354998333e-06,
+      "loss": 0.7108,
+      "step": 3859
+    },
+    {
+      "epoch": 1.9154772247734888,
+      "grad_norm": 0.13608274463655146,
+      "learning_rate": 2.6738278645771615e-06,
+      "loss": 0.7161,
+      "step": 3860
+    },
+    {
+      "epoch": 1.9159736874767284,
+      "grad_norm": 0.13224537124708974,
+      "learning_rate": 2.6728528670876875e-06,
+      "loss": 0.6993,
+      "step": 3861
+    },
+    {
+      "epoch": 1.9164701501799677,
+      "grad_norm": 0.13440646467554213,
+      "learning_rate": 2.6718778431804243e-06,
+      "loss": 0.7041,
+      "step": 3862
+    },
+    {
+      "epoch": 1.9169666128832072,
+      "grad_norm": 0.12745811657851805,
+      "learning_rate": 2.670902793004389e-06,
+      "loss": 0.6779,
+      "step": 3863
+    },
+    {
+      "epoch": 1.9174630755864466,
+      "grad_norm": 0.13050712516771976,
+      "learning_rate": 2.6699277167086013e-06,
+      "loss": 0.7166,
+      "step": 3864
+    },
+    {
+      "epoch": 1.917959538289686,
+      "grad_norm": 0.12713091659389295,
+      "learning_rate": 2.668952614442087e-06,
+      "loss": 0.7017,
+      "step": 3865
+    },
+    {
+      "epoch": 1.9184560009929255,
+      "grad_norm": 0.135043325139347,
+      "learning_rate": 2.6679774863538747e-06,
+      "loss": 0.7295,
+      "step": 3866
+    },
+    {
+      "epoch": 1.9189524636961648,
+      "grad_norm": 0.13202671430433874,
+      "learning_rate": 2.667002332592997e-06,
+      "loss": 0.7315,
+      "step": 3867
+    },
+    {
+      "epoch": 1.9194489263994043,
+      "grad_norm": 0.13505500052720468,
+      "learning_rate": 2.6660271533084895e-06,
+      "loss": 0.7183,
+      "step": 3868
+    },
+    {
+      "epoch": 1.9199453891026437,
+      "grad_norm": 0.13089106796090635,
+      "learning_rate": 2.6650519486493955e-06,
+      "loss": 0.7629,
+      "step": 3869
+    },
+    {
+      "epoch": 1.920441851805883,
+      "grad_norm": 0.13514599448403475,
+      "learning_rate": 2.664076718764756e-06,
+      "loss": 0.7292,
+      "step": 3870
+    },
+    {
+      "epoch": 1.9209383145091226,
+      "grad_norm": 0.13758072791357281,
+      "learning_rate": 2.663101463803621e-06,
+      "loss": 0.7058,
+      "step": 3871
+    },
+    {
+      "epoch": 1.9214347772123619,
+      "grad_norm": 0.13451491052620126,
+      "learning_rate": 2.6621261839150426e-06,
+      "loss": 0.764,
+      "step": 3872
+    },
+    {
+      "epoch": 1.9219312399156014,
+      "grad_norm": 0.13120349619972196,
+      "learning_rate": 2.6611508792480763e-06,
+      "loss": 0.7192,
+      "step": 3873
+    },
+    {
+      "epoch": 1.9224277026188408,
+      "grad_norm": 0.1343506319273604,
+      "learning_rate": 2.6601755499517826e-06,
+      "loss": 0.7515,
+      "step": 3874
+    },
+    {
+      "epoch": 1.92292416532208,
+      "grad_norm": 0.13155915623807082,
+      "learning_rate": 2.6592001961752246e-06,
+      "loss": 0.695,
+      "step": 3875
+    },
+    {
+      "epoch": 1.9234206280253194,
+      "grad_norm": 0.12537341810788755,
+      "learning_rate": 2.658224818067468e-06,
+      "loss": 0.6914,
+      "step": 3876
+    },
+    {
+      "epoch": 1.923917090728559,
+      "grad_norm": 0.1298224456392821,
+      "learning_rate": 2.657249415777585e-06,
+      "loss": 0.7031,
+      "step": 3877
+    },
+    {
+      "epoch": 1.9244135534317985,
+      "grad_norm": 0.12831397033755135,
+      "learning_rate": 2.6562739894546507e-06,
+      "loss": 0.6985,
+      "step": 3878
+    },
+    {
+      "epoch": 1.9249100161350379,
+      "grad_norm": 0.13463509794079495,
+      "learning_rate": 2.6552985392477424e-06,
+      "loss": 0.7578,
+      "step": 3879
+    },
+    {
+      "epoch": 1.9254064788382772,
+      "grad_norm": 0.13466504296387913,
+      "learning_rate": 2.6543230653059427e-06,
+      "loss": 0.7357,
+      "step": 3880
+    },
+    {
+      "epoch": 1.9259029415415165,
+      "grad_norm": 0.13590866765903148,
+      "learning_rate": 2.6533475677783364e-06,
+      "loss": 0.7367,
+      "step": 3881
+    },
+    {
+      "epoch": 1.926399404244756,
+      "grad_norm": 0.13302016457236643,
+      "learning_rate": 2.652372046814014e-06,
+      "loss": 0.7737,
+      "step": 3882
+    },
+    {
+      "epoch": 1.9268958669479956,
+      "grad_norm": 0.13348758747050615,
+      "learning_rate": 2.651396502562067e-06,
+      "loss": 0.8016,
+      "step": 3883
+    },
+    {
+      "epoch": 1.927392329651235,
+      "grad_norm": 0.1327244870657118,
+      "learning_rate": 2.6504209351715914e-06,
+      "loss": 0.7173,
+      "step": 3884
+    },
+    {
+      "epoch": 1.9278887923544743,
+      "grad_norm": 0.12927521222084867,
+      "learning_rate": 2.6494453447916884e-06,
+      "loss": 0.6832,
+      "step": 3885
+    },
+    {
+      "epoch": 1.9283852550577136,
+      "grad_norm": 0.13386451158693466,
+      "learning_rate": 2.6484697315714602e-06,
+      "loss": 0.7336,
+      "step": 3886
+    },
+    {
+      "epoch": 1.9288817177609532,
+      "grad_norm": 0.1264106819143579,
+      "learning_rate": 2.6474940956600143e-06,
+      "loss": 0.6716,
+      "step": 3887
+    },
+    {
+      "epoch": 1.9293781804641927,
+      "grad_norm": 0.1280344281232393,
+      "learning_rate": 2.64651843720646e-06,
+      "loss": 0.7613,
+      "step": 3888
+    },
+    {
+      "epoch": 1.929874643167432,
+      "grad_norm": 0.13685165914605946,
+      "learning_rate": 2.6455427563599128e-06,
+      "loss": 0.7595,
+      "step": 3889
+    },
+    {
+      "epoch": 1.9303711058706714,
+      "grad_norm": 0.12802262711212847,
+      "learning_rate": 2.644567053269489e-06,
+      "loss": 0.7105,
+      "step": 3890
+    },
+    {
+      "epoch": 1.9308675685739107,
+      "grad_norm": 0.1401396120332125,
+      "learning_rate": 2.643591328084309e-06,
+      "loss": 0.7363,
+      "step": 3891
+    },
+    {
+      "epoch": 1.9313640312771503,
+      "grad_norm": 0.13423489071016823,
+      "learning_rate": 2.6426155809534958e-06,
+      "loss": 0.6799,
+      "step": 3892
+    },
+    {
+      "epoch": 1.9318604939803898,
+      "grad_norm": 0.125010992787907,
+      "learning_rate": 2.6416398120261782e-06,
+      "loss": 0.6502,
+      "step": 3893
+    },
+    {
+      "epoch": 1.9323569566836292,
+      "grad_norm": 0.1378318453962945,
+      "learning_rate": 2.6406640214514866e-06,
+      "loss": 0.7189,
+      "step": 3894
+    },
+    {
+      "epoch": 1.9328534193868685,
+      "grad_norm": 0.12766766267445206,
+      "learning_rate": 2.639688209378554e-06,
+      "loss": 0.6806,
+      "step": 3895
+    },
+    {
+      "epoch": 1.9333498820901078,
+      "grad_norm": 0.1328404296148707,
+      "learning_rate": 2.63871237595652e-06,
+      "loss": 0.716,
+      "step": 3896
+    },
+    {
+      "epoch": 1.9338463447933474,
+      "grad_norm": 0.13020525639573502,
+      "learning_rate": 2.6377365213345217e-06,
+      "loss": 0.6673,
+      "step": 3897
+    },
+    {
+      "epoch": 1.934342807496587,
+      "grad_norm": 0.12842653922306596,
+      "learning_rate": 2.6367606456617057e-06,
+      "loss": 0.6998,
+      "step": 3898
+    },
+    {
+      "epoch": 1.9348392701998263,
+      "grad_norm": 0.1363023679071453,
+      "learning_rate": 2.6357847490872176e-06,
+      "loss": 0.7224,
+      "step": 3899
+    },
+    {
+      "epoch": 1.9353357329030656,
+      "grad_norm": 0.1371628916129143,
+      "learning_rate": 2.634808831760207e-06,
+      "loss": 0.7227,
+      "step": 3900
+    },
+    {
+      "epoch": 1.935832195606305,
+      "grad_norm": 0.13510099814120166,
+      "learning_rate": 2.6338328938298287e-06,
+      "loss": 0.7311,
+      "step": 3901
+    },
+    {
+      "epoch": 1.9363286583095445,
+      "grad_norm": 0.1298081070212144,
+      "learning_rate": 2.632856935445238e-06,
+      "loss": 0.6655,
+      "step": 3902
+    },
+    {
+      "epoch": 1.936825121012784,
+      "grad_norm": 0.1289538644890483,
+      "learning_rate": 2.6318809567555946e-06,
+      "loss": 0.7062,
+      "step": 3903
+    },
+    {
+      "epoch": 1.9373215837160234,
+      "grad_norm": 0.13439353563363904,
+      "learning_rate": 2.630904957910062e-06,
+      "loss": 0.6991,
+      "step": 3904
+    },
+    {
+      "epoch": 1.9378180464192627,
+      "grad_norm": 0.13576155759044287,
+      "learning_rate": 2.6299289390578054e-06,
+      "loss": 0.7344,
+      "step": 3905
+    },
+    {
+      "epoch": 1.938314509122502,
+      "grad_norm": 0.1284387703639709,
+      "learning_rate": 2.628952900347994e-06,
+      "loss": 0.7029,
+      "step": 3906
+    },
+    {
+      "epoch": 1.9388109718257416,
+      "grad_norm": 0.12937400679523783,
+      "learning_rate": 2.6279768419297997e-06,
+      "loss": 0.6859,
+      "step": 3907
+    },
+    {
+      "epoch": 1.9393074345289811,
+      "grad_norm": 0.1317149945336735,
+      "learning_rate": 2.6270007639523966e-06,
+      "loss": 0.679,
+      "step": 3908
+    },
+    {
+      "epoch": 1.9398038972322205,
+      "grad_norm": 0.1306880273284504,
+      "learning_rate": 2.6260246665649623e-06,
+      "loss": 0.6934,
+      "step": 3909
+    },
+    {
+      "epoch": 1.9403003599354598,
+      "grad_norm": 0.13275523795140065,
+      "learning_rate": 2.625048549916679e-06,
+      "loss": 0.71,
+      "step": 3910
+    },
+    {
+      "epoch": 1.9407968226386991,
+      "grad_norm": 0.13269576003308478,
+      "learning_rate": 2.6240724141567296e-06,
+      "loss": 0.7458,
+      "step": 3911
+    },
+    {
+      "epoch": 1.9412932853419387,
+      "grad_norm": 0.12913224685659597,
+      "learning_rate": 2.6230962594343018e-06,
+      "loss": 0.639,
+      "step": 3912
+    },
+    {
+      "epoch": 1.9417897480451782,
+      "grad_norm": 0.13621237829923233,
+      "learning_rate": 2.622120085898584e-06,
+      "loss": 0.7167,
+      "step": 3913
+    },
+    {
+      "epoch": 1.9422862107484176,
+      "grad_norm": 0.13555011416798202,
+      "learning_rate": 2.6211438936987692e-06,
+      "loss": 0.7019,
+      "step": 3914
+    },
+    {
+      "epoch": 1.942782673451657,
+      "grad_norm": 0.13015371567538989,
+      "learning_rate": 2.620167682984052e-06,
+      "loss": 0.7581,
+      "step": 3915
+    },
+    {
+      "epoch": 1.9432791361548962,
+      "grad_norm": 0.14513175483219928,
+      "learning_rate": 2.6191914539036318e-06,
+      "loss": 0.7588,
+      "step": 3916
+    },
+    {
+      "epoch": 1.9437755988581358,
+      "grad_norm": 0.13652288010038582,
+      "learning_rate": 2.6182152066067095e-06,
+      "loss": 0.7355,
+      "step": 3917
+    },
+    {
+      "epoch": 1.9442720615613753,
+      "grad_norm": 0.14112063768896954,
+      "learning_rate": 2.6172389412424876e-06,
+      "loss": 0.8433,
+      "step": 3918
+    },
+    {
+      "epoch": 1.9447685242646147,
+      "grad_norm": 0.13310261496502976,
+      "learning_rate": 2.616262657960173e-06,
+      "loss": 0.749,
+      "step": 3919
+    },
+    {
+      "epoch": 1.945264986967854,
+      "grad_norm": 0.1298763438679353,
+      "learning_rate": 2.6152863569089754e-06,
+      "loss": 0.7002,
+      "step": 3920
+    },
+    {
+      "epoch": 1.9457614496710933,
+      "grad_norm": 0.13406943600819382,
+      "learning_rate": 2.614310038238107e-06,
+      "loss": 0.7198,
+      "step": 3921
+    },
+    {
+      "epoch": 1.9462579123743329,
+      "grad_norm": 0.1388915294171547,
+      "learning_rate": 2.613333702096782e-06,
+      "loss": 0.732,
+      "step": 3922
+    },
+    {
+      "epoch": 1.9467543750775724,
+      "grad_norm": 0.13098745963258238,
+      "learning_rate": 2.6123573486342185e-06,
+      "loss": 0.694,
+      "step": 3923
+    },
+    {
+      "epoch": 1.9472508377808118,
+      "grad_norm": 0.13673305400029112,
+      "learning_rate": 2.6113809779996344e-06,
+      "loss": 0.709,
+      "step": 3924
+    },
+    {
+      "epoch": 1.947747300484051,
+      "grad_norm": 0.13672731547591427,
+      "learning_rate": 2.610404590342254e-06,
+      "loss": 0.7119,
+      "step": 3925
+    },
+    {
+      "epoch": 1.9482437631872904,
+      "grad_norm": 0.12667903565039704,
+      "learning_rate": 2.6094281858113026e-06,
+      "loss": 0.7128,
+      "step": 3926
+    },
+    {
+      "epoch": 1.94874022589053,
+      "grad_norm": 0.1295518629610594,
+      "learning_rate": 2.6084517645560077e-06,
+      "loss": 0.7045,
+      "step": 3927
+    },
+    {
+      "epoch": 1.9492366885937695,
+      "grad_norm": 0.13912368621627139,
+      "learning_rate": 2.6074753267255994e-06,
+      "loss": 0.7286,
+      "step": 3928
+    },
+    {
+      "epoch": 1.9497331512970089,
+      "grad_norm": 0.1408437562224136,
+      "learning_rate": 2.60649887246931e-06,
+      "loss": 0.7566,
+      "step": 3929
+    },
+    {
+      "epoch": 1.9502296140002482,
+      "grad_norm": 0.1343158417044777,
+      "learning_rate": 2.605522401936376e-06,
+      "loss": 0.7252,
+      "step": 3930
+    },
+    {
+      "epoch": 1.9507260767034875,
+      "grad_norm": 0.1312175766834046,
+      "learning_rate": 2.604545915276035e-06,
+      "loss": 0.718,
+      "step": 3931
+    },
+    {
+      "epoch": 1.951222539406727,
+      "grad_norm": 0.1375316196872564,
+      "learning_rate": 2.6035694126375265e-06,
+      "loss": 0.7681,
+      "step": 3932
+    },
+    {
+      "epoch": 1.9517190021099666,
+      "grad_norm": 0.1347425466729892,
+      "learning_rate": 2.6025928941700945e-06,
+      "loss": 0.6754,
+      "step": 3933
+    },
+    {
+      "epoch": 1.952215464813206,
+      "grad_norm": 0.12903423135149714,
+      "learning_rate": 2.6016163600229832e-06,
+      "loss": 0.7674,
+      "step": 3934
+    },
+    {
+      "epoch": 1.9527119275164453,
+      "grad_norm": 0.13070824473355627,
+      "learning_rate": 2.6006398103454407e-06,
+      "loss": 0.6968,
+      "step": 3935
+    },
+    {
+      "epoch": 1.9532083902196846,
+      "grad_norm": 0.12638361456974517,
+      "learning_rate": 2.5996632452867167e-06,
+      "loss": 0.6966,
+      "step": 3936
+    },
+    {
+      "epoch": 1.9537048529229242,
+      "grad_norm": 0.13442945386349076,
+      "learning_rate": 2.5986866649960634e-06,
+      "loss": 0.7475,
+      "step": 3937
+    },
+    {
+      "epoch": 1.9542013156261637,
+      "grad_norm": 0.13647786271821394,
+      "learning_rate": 2.597710069622736e-06,
+      "loss": 0.6832,
+      "step": 3938
+    },
+    {
+      "epoch": 1.954697778329403,
+      "grad_norm": 0.1303452381637909,
+      "learning_rate": 2.596733459315992e-06,
+      "loss": 0.7067,
+      "step": 3939
+    },
+    {
+      "epoch": 1.9551942410326424,
+      "grad_norm": 0.1272985129582747,
+      "learning_rate": 2.595756834225089e-06,
+      "loss": 0.6699,
+      "step": 3940
+    },
+    {
+      "epoch": 1.9556907037358817,
+      "grad_norm": 0.12900166946504296,
+      "learning_rate": 2.594780194499289e-06,
+      "loss": 0.6893,
+      "step": 3941
+    },
+    {
+      "epoch": 1.9561871664391213,
+      "grad_norm": 0.13345773544468753,
+      "learning_rate": 2.593803540287856e-06,
+      "loss": 0.6645,
+      "step": 3942
+    },
+    {
+      "epoch": 1.9566836291423608,
+      "grad_norm": 0.12499172467109242,
+      "learning_rate": 2.592826871740056e-06,
+      "loss": 0.6568,
+      "step": 3943
+    },
+    {
+      "epoch": 1.9571800918456002,
+      "grad_norm": 0.1369899943691365,
+      "learning_rate": 2.5918501890051573e-06,
+      "loss": 0.7235,
+      "step": 3944
+    },
+    {
+      "epoch": 1.9576765545488395,
+      "grad_norm": 0.13445906303874938,
+      "learning_rate": 2.5908734922324293e-06,
+      "loss": 0.725,
+      "step": 3945
+    },
+    {
+      "epoch": 1.9581730172520788,
+      "grad_norm": 0.1620759841591545,
+      "learning_rate": 2.5898967815711455e-06,
+      "loss": 0.6952,
+      "step": 3946
+    },
+    {
+      "epoch": 1.9586694799553184,
+      "grad_norm": 0.1398600398265006,
+      "learning_rate": 2.58892005717058e-06,
+      "loss": 0.766,
+      "step": 3947
+    },
+    {
+      "epoch": 1.959165942658558,
+      "grad_norm": 0.13096623784090528,
+      "learning_rate": 2.5879433191800093e-06,
+      "loss": 0.7198,
+      "step": 3948
+    },
+    {
+      "epoch": 1.9596624053617973,
+      "grad_norm": 0.12657285136649368,
+      "learning_rate": 2.5869665677487122e-06,
+      "loss": 0.6872,
+      "step": 3949
+    },
+    {
+      "epoch": 1.9601588680650366,
+      "grad_norm": 0.13897203572470015,
+      "learning_rate": 2.58598980302597e-06,
+      "loss": 0.7355,
+      "step": 3950
+    },
+    {
+      "epoch": 1.960655330768276,
+      "grad_norm": 0.12963711745676562,
+      "learning_rate": 2.585013025161065e-06,
+      "loss": 0.6979,
+      "step": 3951
+    },
+    {
+      "epoch": 1.9611517934715155,
+      "grad_norm": 0.13783927250748948,
+      "learning_rate": 2.584036234303282e-06,
+      "loss": 0.7381,
+      "step": 3952
+    },
+    {
+      "epoch": 1.961648256174755,
+      "grad_norm": 0.13448371125196593,
+      "learning_rate": 2.583059430601908e-06,
+      "loss": 0.699,
+      "step": 3953
+    },
+    {
+      "epoch": 1.9621447188779944,
+      "grad_norm": 0.13318126568011363,
+      "learning_rate": 2.5820826142062323e-06,
+      "loss": 0.6785,
+      "step": 3954
+    },
+    {
+      "epoch": 1.9626411815812337,
+      "grad_norm": 0.13076288936075522,
+      "learning_rate": 2.581105785265545e-06,
+      "loss": 0.7204,
+      "step": 3955
+    },
+    {
+      "epoch": 1.963137644284473,
+      "grad_norm": 0.13117966572328443,
+      "learning_rate": 2.580128943929139e-06,
+      "loss": 0.7089,
+      "step": 3956
+    },
+    {
+      "epoch": 1.9636341069877126,
+      "grad_norm": 0.14643814504267247,
+      "learning_rate": 2.5791520903463076e-06,
+      "loss": 0.7698,
+      "step": 3957
+    },
+    {
+      "epoch": 1.9641305696909521,
+      "grad_norm": 0.13110955970838375,
+      "learning_rate": 2.578175224666349e-06,
+      "loss": 0.702,
+      "step": 3958
+    },
+    {
+      "epoch": 1.9646270323941915,
+      "grad_norm": 0.13211603472359895,
+      "learning_rate": 2.5771983470385604e-06,
+      "loss": 0.691,
+      "step": 3959
+    },
+    {
+      "epoch": 1.9651234950974308,
+      "grad_norm": 0.13433138088355598,
+      "learning_rate": 2.576221457612243e-06,
+      "loss": 0.7116,
+      "step": 3960
+    },
+    {
+      "epoch": 1.9656199578006701,
+      "grad_norm": 0.1329409392764817,
+      "learning_rate": 2.575244556536697e-06,
+      "loss": 0.7445,
+      "step": 3961
+    },
+    {
+      "epoch": 1.9661164205039097,
+      "grad_norm": 0.1314186532703457,
+      "learning_rate": 2.5742676439612283e-06,
+      "loss": 0.7041,
+      "step": 3962
+    },
+    {
+      "epoch": 1.9666128832071492,
+      "grad_norm": 0.13197553608649243,
+      "learning_rate": 2.5732907200351402e-06,
+      "loss": 0.7109,
+      "step": 3963
+    },
+    {
+      "epoch": 1.9671093459103886,
+      "grad_norm": 0.139383832666028,
+      "learning_rate": 2.5723137849077406e-06,
+      "loss": 0.7219,
+      "step": 3964
+    },
+    {
+      "epoch": 1.9676058086136279,
+      "grad_norm": 0.12890251354191545,
+      "learning_rate": 2.571336838728338e-06,
+      "loss": 0.7209,
+      "step": 3965
+    },
+    {
+      "epoch": 1.9681022713168672,
+      "grad_norm": 0.13068185566410134,
+      "learning_rate": 2.5703598816462443e-06,
+      "loss": 0.7415,
+      "step": 3966
+    },
+    {
+      "epoch": 1.9685987340201068,
+      "grad_norm": 0.13578478265192953,
+      "learning_rate": 2.5693829138107707e-06,
+      "loss": 0.7458,
+      "step": 3967
+    },
+    {
+      "epoch": 1.9690951967233463,
+      "grad_norm": 0.13252065296864216,
+      "learning_rate": 2.568405935371231e-06,
+      "loss": 0.7234,
+      "step": 3968
+    },
+    {
+      "epoch": 1.9695916594265857,
+      "grad_norm": 0.13323778860939306,
+      "learning_rate": 2.5674289464769405e-06,
+      "loss": 0.7013,
+      "step": 3969
+    },
+    {
+      "epoch": 1.970088122129825,
+      "grad_norm": 0.1287653472052357,
+      "learning_rate": 2.566451947277217e-06,
+      "loss": 0.6984,
+      "step": 3970
+    },
+    {
+      "epoch": 1.9705845848330643,
+      "grad_norm": 0.12933587377045058,
+      "learning_rate": 2.565474937921379e-06,
+      "loss": 0.7015,
+      "step": 3971
+    },
+    {
+      "epoch": 1.9710810475363039,
+      "grad_norm": 0.13330915382752193,
+      "learning_rate": 2.5644979185587466e-06,
+      "loss": 0.7413,
+      "step": 3972
+    },
+    {
+      "epoch": 1.9715775102395434,
+      "grad_norm": 0.1266644141116552,
+      "learning_rate": 2.5635208893386416e-06,
+      "loss": 0.6668,
+      "step": 3973
+    },
+    {
+      "epoch": 1.9720739729427827,
+      "grad_norm": 0.1394923566595831,
+      "learning_rate": 2.5625438504103863e-06,
+      "loss": 0.725,
+      "step": 3974
+    },
+    {
+      "epoch": 1.972570435646022,
+      "grad_norm": 0.1333230212803404,
+      "learning_rate": 2.5615668019233065e-06,
+      "loss": 0.7066,
+      "step": 3975
+    },
+    {
+      "epoch": 1.9730668983492614,
+      "grad_norm": 0.13786776101291032,
+      "learning_rate": 2.560589744026729e-06,
+      "loss": 0.7136,
+      "step": 3976
+    },
+    {
+      "epoch": 1.973563361052501,
+      "grad_norm": 0.13850183160275445,
+      "learning_rate": 2.5596126768699798e-06,
+      "loss": 0.6687,
+      "step": 3977
+    },
+    {
+      "epoch": 1.9740598237557403,
+      "grad_norm": 0.13732643590313048,
+      "learning_rate": 2.5586356006023894e-06,
+      "loss": 0.7616,
+      "step": 3978
+    },
+    {
+      "epoch": 1.9745562864589798,
+      "grad_norm": 0.13057844494817822,
+      "learning_rate": 2.5576585153732875e-06,
+      "loss": 0.6654,
+      "step": 3979
+    },
+    {
+      "epoch": 1.9750527491622192,
+      "grad_norm": 0.13328648064830395,
+      "learning_rate": 2.556681421332005e-06,
+      "loss": 0.7171,
+      "step": 3980
+    },
+    {
+      "epoch": 1.9755492118654585,
+      "grad_norm": 0.13209382988878132,
+      "learning_rate": 2.555704318627877e-06,
+      "loss": 0.7245,
+      "step": 3981
+    },
+    {
+      "epoch": 1.976045674568698,
+      "grad_norm": 0.1255898846757212,
+      "learning_rate": 2.5547272074102375e-06,
+      "loss": 0.6884,
+      "step": 3982
+    },
+    {
+      "epoch": 1.9765421372719374,
+      "grad_norm": 0.13542305325011644,
+      "learning_rate": 2.553750087828421e-06,
+      "loss": 0.7229,
+      "step": 3983
+    },
+    {
+      "epoch": 1.977038599975177,
+      "grad_norm": 0.13274636935908424,
+      "learning_rate": 2.552772960031765e-06,
+      "loss": 0.727,
+      "step": 3984
+    },
+    {
+      "epoch": 1.9775350626784163,
+      "grad_norm": 0.13087815752389365,
+      "learning_rate": 2.551795824169609e-06,
+      "loss": 0.6981,
+      "step": 3985
+    },
+    {
+      "epoch": 1.9780315253816556,
+      "grad_norm": 0.13311849040345575,
+      "learning_rate": 2.550818680391292e-06,
+      "loss": 0.7227,
+      "step": 3986
+    },
+    {
+      "epoch": 1.9785279880848952,
+      "grad_norm": 0.13708649225810834,
+      "learning_rate": 2.5498415288461537e-06,
+      "loss": 0.7279,
+      "step": 3987
+    },
+    {
+      "epoch": 1.9790244507881345,
+      "grad_norm": 0.13033779988678035,
+      "learning_rate": 2.548864369683538e-06,
+      "loss": 0.6814,
+      "step": 3988
+    },
+    {
+      "epoch": 1.979520913491374,
+      "grad_norm": 0.13116550526012039,
+      "learning_rate": 2.547887203052786e-06,
+      "loss": 0.7825,
+      "step": 3989
+    },
+    {
+      "epoch": 1.9800173761946134,
+      "grad_norm": 0.13081602543099036,
+      "learning_rate": 2.5469100291032423e-06,
+      "loss": 0.7239,
+      "step": 3990
+    },
+    {
+      "epoch": 1.9805138388978527,
+      "grad_norm": 0.13072879071258814,
+      "learning_rate": 2.545932847984254e-06,
+      "loss": 0.7104,
+      "step": 3991
+    },
+    {
+      "epoch": 1.9810103016010923,
+      "grad_norm": 0.13258224731686477,
+      "learning_rate": 2.5449556598451656e-06,
+      "loss": 0.6974,
+      "step": 3992
+    },
+    {
+      "epoch": 1.9815067643043316,
+      "grad_norm": 0.1351940871833718,
+      "learning_rate": 2.5439784648353256e-06,
+      "loss": 0.7572,
+      "step": 3993
+    },
+    {
+      "epoch": 1.9820032270075711,
+      "grad_norm": 0.14461540568622883,
+      "learning_rate": 2.543001263104083e-06,
+      "loss": 0.7281,
+      "step": 3994
+    },
+    {
+      "epoch": 1.9824996897108105,
+      "grad_norm": 0.13342177790858398,
+      "learning_rate": 2.5420240548007856e-06,
+      "loss": 0.7461,
+      "step": 3995
+    },
+    {
+      "epoch": 1.9829961524140498,
+      "grad_norm": 0.13543936141725377,
+      "learning_rate": 2.5410468400747858e-06,
+      "loss": 0.7591,
+      "step": 3996
+    },
+    {
+      "epoch": 1.9834926151172894,
+      "grad_norm": 0.1292242889450314,
+      "learning_rate": 2.5400696190754347e-06,
+      "loss": 0.7125,
+      "step": 3997
+    },
+    {
+      "epoch": 1.9839890778205287,
+      "grad_norm": 0.13366333141114367,
+      "learning_rate": 2.539092391952085e-06,
+      "loss": 0.7215,
+      "step": 3998
+    },
+    {
+      "epoch": 1.9844855405237682,
+      "grad_norm": 0.13947805335489766,
+      "learning_rate": 2.5381151588540896e-06,
+      "loss": 0.7721,
+      "step": 3999
+    },
+    {
+      "epoch": 1.9849820032270076,
+      "grad_norm": 0.13011957215342748,
+      "learning_rate": 2.537137919930803e-06,
+      "loss": 0.7054,
+      "step": 4000
+    },
+    {
+      "epoch": 1.985478465930247,
+      "grad_norm": 0.1350409911144854,
+      "learning_rate": 2.5361606753315814e-06,
+      "loss": 0.687,
+      "step": 4001
+    },
+    {
+      "epoch": 1.9859749286334865,
+      "grad_norm": 0.129620679858614,
+      "learning_rate": 2.5351834252057805e-06,
+      "loss": 0.7008,
+      "step": 4002
+    },
+    {
+      "epoch": 1.9864713913367258,
+      "grad_norm": 0.12678966363438363,
+      "learning_rate": 2.534206169702757e-06,
+      "loss": 0.6995,
+      "step": 4003
+    },
+    {
+      "epoch": 1.9869678540399653,
+      "grad_norm": 0.1318913991769552,
+      "learning_rate": 2.533228908971869e-06,
+      "loss": 0.734,
+      "step": 4004
+    },
+    {
+      "epoch": 1.9874643167432047,
+      "grad_norm": 0.13831865321707615,
+      "learning_rate": 2.532251643162475e-06,
+      "loss": 0.7618,
+      "step": 4005
+    },
+    {
+      "epoch": 1.987960779446444,
+      "grad_norm": 0.1349051666242995,
+      "learning_rate": 2.5312743724239336e-06,
+      "loss": 0.7522,
+      "step": 4006
+    },
+    {
+      "epoch": 1.9884572421496836,
+      "grad_norm": 0.13052404435468878,
+      "learning_rate": 2.5302970969056068e-06,
+      "loss": 0.6867,
+      "step": 4007
+    },
+    {
+      "epoch": 1.9889537048529229,
+      "grad_norm": 0.13818539051178105,
+      "learning_rate": 2.529319816756854e-06,
+      "loss": 0.733,
+      "step": 4008
+    },
+    {
+      "epoch": 1.9894501675561624,
+      "grad_norm": 0.12800876536122638,
+      "learning_rate": 2.5283425321270377e-06,
+      "loss": 0.6684,
+      "step": 4009
+    },
+    {
+      "epoch": 1.9899466302594018,
+      "grad_norm": 0.12765205661654766,
+      "learning_rate": 2.5273652431655204e-06,
+      "loss": 0.7032,
+      "step": 4010
+    },
+    {
+      "epoch": 1.990443092962641,
+      "grad_norm": 0.1338564783816033,
+      "learning_rate": 2.526387950021663e-06,
+      "loss": 0.7496,
+      "step": 4011
+    },
+    {
+      "epoch": 1.9909395556658807,
+      "grad_norm": 0.13415496043164832,
+      "learning_rate": 2.525410652844831e-06,
+      "loss": 0.7157,
+      "step": 4012
+    },
+    {
+      "epoch": 1.99143601836912,
+      "grad_norm": 0.1282775301010404,
+      "learning_rate": 2.524433351784389e-06,
+      "loss": 0.6783,
+      "step": 4013
+    },
+    {
+      "epoch": 1.9919324810723595,
+      "grad_norm": 0.13209341786586526,
+      "learning_rate": 2.523456046989701e-06,
+      "loss": 0.7353,
+      "step": 4014
+    },
+    {
+      "epoch": 1.9924289437755989,
+      "grad_norm": 0.1287222818009373,
+      "learning_rate": 2.5224787386101307e-06,
+      "loss": 0.6862,
+      "step": 4015
+    },
+    {
+      "epoch": 1.9929254064788382,
+      "grad_norm": 0.13035922850407525,
+      "learning_rate": 2.5215014267950465e-06,
+      "loss": 0.7004,
+      "step": 4016
+    },
+    {
+      "epoch": 1.9934218691820775,
+      "grad_norm": 0.131542341490382,
+      "learning_rate": 2.5205241116938137e-06,
+      "loss": 0.7147,
+      "step": 4017
+    },
+    {
+      "epoch": 1.993918331885317,
+      "grad_norm": 0.136610862731398,
+      "learning_rate": 2.5195467934558003e-06,
+      "loss": 0.7341,
+      "step": 4018
+    },
+    {
+      "epoch": 1.9944147945885566,
+      "grad_norm": 0.13012603782449922,
+      "learning_rate": 2.5185694722303728e-06,
+      "loss": 0.6945,
+      "step": 4019
+    },
+    {
+      "epoch": 1.994911257291796,
+      "grad_norm": 0.1325429269300131,
+      "learning_rate": 2.517592148166899e-06,
+      "loss": 0.7291,
+      "step": 4020
+    },
+    {
+      "epoch": 1.9954077199950353,
+      "grad_norm": 0.1297801158635372,
+      "learning_rate": 2.516614821414747e-06,
+      "loss": 0.6867,
+      "step": 4021
+    },
+    {
+      "epoch": 1.9959041826982746,
+      "grad_norm": 0.13181943517502306,
+      "learning_rate": 2.5156374921232862e-06,
+      "loss": 0.6634,
+      "step": 4022
+    },
+    {
+      "epoch": 1.9964006454015142,
+      "grad_norm": 0.13355813171588157,
+      "learning_rate": 2.5146601604418854e-06,
+      "loss": 0.7556,
+      "step": 4023
+    },
+    {
+      "epoch": 1.9968971081047537,
+      "grad_norm": 0.1310512552198657,
+      "learning_rate": 2.5136828265199143e-06,
+      "loss": 0.7198,
+      "step": 4024
+    },
+    {
+      "epoch": 1.997393570807993,
+      "grad_norm": 0.1272191857000951,
+      "learning_rate": 2.512705490506743e-06,
+      "loss": 0.7172,
+      "step": 4025
+    },
+    {
+      "epoch": 1.9978900335112324,
+      "grad_norm": 0.12766117521597195,
+      "learning_rate": 2.511728152551741e-06,
+      "loss": 0.724,
+      "step": 4026
+    },
+    {
+      "epoch": 1.9983864962144717,
+      "grad_norm": 0.13122724706133568,
+      "learning_rate": 2.5107508128042786e-06,
+      "loss": 0.7186,
+      "step": 4027
+    },
+    {
+      "epoch": 1.9988829589177113,
+      "grad_norm": 0.13170560779745644,
+      "learning_rate": 2.509773471413726e-06,
+      "loss": 0.7394,
+      "step": 4028
+    },
+    {
+      "epoch": 1.9993794216209508,
+      "grad_norm": 0.1279325800617962,
+      "learning_rate": 2.508796128529456e-06,
+      "loss": 0.6678,
+      "step": 4029
+    },
+    {
+      "epoch": 1.9998758843241902,
+      "grad_norm": 0.12867108507101366,
+      "learning_rate": 2.507818784300839e-06,
+      "loss": 0.6981,
+      "step": 4030
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.12867108507101366,
+      "learning_rate": 2.5068414388772454e-06,
+      "loss": 0.191,
+      "step": 4031
+    },
+    {
+      "epoch": 2.0003723470274295,
+      "grad_norm": 0.13238445630417517,
+      "learning_rate": 2.505864092408048e-06,
+      "loss": 0.5359,
+      "step": 4032
+    },
+    {
+      "epoch": 2.0003723470274295,
+      "eval_loss": 0.7288655042648315,
+      "eval_runtime": 135.8088,
+      "eval_samples_per_second": 223.498,
+      "eval_steps_per_second": 27.944,
+      "step": 4032
+    },
+    {
+      "epoch": 2.0004964627032393,
+      "grad_norm": 0.13212599455112428,
+      "learning_rate": 2.504886745042618e-06,
+      "loss": 0.724,
+      "step": 4033
+    },
+    {
+      "epoch": 2.0009929254064787,
+      "grad_norm": 0.1289165900817434,
+      "learning_rate": 2.503909396930328e-06,
+      "loss": 0.6931,
+      "step": 4034
+    },
+    {
+      "epoch": 2.0014893881097184,
+      "grad_norm": 0.12817979511727562,
+      "learning_rate": 2.5029320482205487e-06,
+      "loss": 0.7045,
+      "step": 4035
+    },
+    {
+      "epoch": 2.0019858508129578,
+      "grad_norm": 0.13305215189225175,
+      "learning_rate": 2.501954699062653e-06,
+      "loss": 0.6943,
+      "step": 4036
+    },
+    {
+      "epoch": 2.002482313516197,
+      "grad_norm": 0.1293844522970027,
+      "learning_rate": 2.500977349606013e-06,
+      "loss": 0.7171,
+      "step": 4037
+    },
+    {
+      "epoch": 2.0029787762194364,
+      "grad_norm": 0.14606041921842416,
+      "learning_rate": 2.5e-06,
+      "loss": 0.759,
+      "step": 4038
+    },
+    {
+      "epoch": 2.0034752389226758,
+      "grad_norm": 0.12805929586454654,
+      "learning_rate": 2.499022650393988e-06,
+      "loss": 0.6875,
+      "step": 4039
+    },
+    {
+      "epoch": 2.0039717016259155,
+      "grad_norm": 0.13380230251599434,
+      "learning_rate": 2.4980453009373475e-06,
+      "loss": 0.7062,
+      "step": 4040
+    },
+    {
+      "epoch": 2.004468164329155,
+      "grad_norm": 0.12965303996516353,
+      "learning_rate": 2.497067951779452e-06,
+      "loss": 0.6996,
+      "step": 4041
+    },
+    {
+      "epoch": 2.004964627032394,
+      "grad_norm": 0.1295766633898341,
+      "learning_rate": 2.4960906030696727e-06,
+      "loss": 0.6684,
+      "step": 4042
+    },
+    {
+      "epoch": 2.0054610897356335,
+      "grad_norm": 0.1312131877586875,
+      "learning_rate": 2.495113254957382e-06,
+      "loss": 0.7207,
+      "step": 4043
+    },
+    {
+      "epoch": 2.005957552438873,
+      "grad_norm": 0.13817927580280212,
+      "learning_rate": 2.4941359075919523e-06,
+      "loss": 0.7428,
+      "step": 4044
+    },
+    {
+      "epoch": 2.0064540151421126,
+      "grad_norm": 0.136455142677692,
+      "learning_rate": 2.493158561122754e-06,
+      "loss": 0.6922,
+      "step": 4045
+    },
+    {
+      "epoch": 2.006950477845352,
+      "grad_norm": 0.13906822012805292,
+      "learning_rate": 2.492181215699162e-06,
+      "loss": 0.7097,
+      "step": 4046
+    },
+    {
+      "epoch": 2.0074469405485913,
+      "grad_norm": 0.12826980150415948,
+      "learning_rate": 2.4912038714705447e-06,
+      "loss": 0.6902,
+      "step": 4047
+    },
+    {
+      "epoch": 2.0079434032518306,
+      "grad_norm": 0.1388317530263215,
+      "learning_rate": 2.490226528586275e-06,
+      "loss": 0.7149,
+      "step": 4048
+    },
+    {
+      "epoch": 2.00843986595507,
+      "grad_norm": 0.13268448735817925,
+      "learning_rate": 2.489249187195723e-06,
+      "loss": 0.7098,
+      "step": 4049
+    },
+    {
+      "epoch": 2.0089363286583097,
+      "grad_norm": 0.1302092877003093,
+      "learning_rate": 2.4882718474482604e-06,
+      "loss": 0.7251,
+      "step": 4050
+    },
+    {
+      "epoch": 2.009432791361549,
+      "grad_norm": 0.13047139016142711,
+      "learning_rate": 2.487294509493258e-06,
+      "loss": 0.7105,
+      "step": 4051
+    },
+    {
+      "epoch": 2.0099292540647884,
+      "grad_norm": 0.13056451140753167,
+      "learning_rate": 2.4863171734800866e-06,
+      "loss": 0.6717,
+      "step": 4052
+    },
+    {
+      "epoch": 2.0104257167680277,
+      "grad_norm": 0.1272813781903442,
+      "learning_rate": 2.485339839558115e-06,
+      "loss": 0.6912,
+      "step": 4053
+    },
+    {
+      "epoch": 2.010922179471267,
+      "grad_norm": 0.1293306773850289,
+      "learning_rate": 2.4843625078767146e-06,
+      "loss": 0.6625,
+      "step": 4054
+    },
+    {
+      "epoch": 2.011418642174507,
+      "grad_norm": 0.13026414160725963,
+      "learning_rate": 2.4833851785852536e-06,
+      "loss": 0.7192,
+      "step": 4055
+    },
+    {
+      "epoch": 2.011915104877746,
+      "grad_norm": 0.13154534592494477,
+      "learning_rate": 2.4824078518331017e-06,
+      "loss": 0.7098,
+      "step": 4056
+    },
+    {
+      "epoch": 2.0124115675809855,
+      "grad_norm": 0.13413230046665645,
+      "learning_rate": 2.481430527769628e-06,
+      "loss": 0.7306,
+      "step": 4057
+    },
+    {
+      "epoch": 2.012908030284225,
+      "grad_norm": 0.13231043126436018,
+      "learning_rate": 2.4804532065442e-06,
+      "loss": 0.7263,
+      "step": 4058
+    },
+    {
+      "epoch": 2.013404492987464,
+      "grad_norm": 0.13114069790750638,
+      "learning_rate": 2.4794758883061862e-06,
+      "loss": 0.6813,
+      "step": 4059
+    },
+    {
+      "epoch": 2.013900955690704,
+      "grad_norm": 0.1330470419586042,
+      "learning_rate": 2.4784985732049535e-06,
+      "loss": 0.7061,
+      "step": 4060
+    },
+    {
+      "epoch": 2.0143974183939433,
+      "grad_norm": 0.1345085568471715,
+      "learning_rate": 2.4775212613898693e-06,
+      "loss": 0.7379,
+      "step": 4061
+    },
+    {
+      "epoch": 2.0148938810971826,
+      "grad_norm": 0.12782673695827584,
+      "learning_rate": 2.4765439530103004e-06,
+      "loss": 0.7046,
+      "step": 4062
+    },
+    {
+      "epoch": 2.015390343800422,
+      "grad_norm": 0.13254520200468667,
+      "learning_rate": 2.475566648215612e-06,
+      "loss": 0.7034,
+      "step": 4063
+    },
+    {
+      "epoch": 2.0158868065036613,
+      "grad_norm": 0.13807047247892,
+      "learning_rate": 2.4745893471551697e-06,
+      "loss": 0.6927,
+      "step": 4064
+    },
+    {
+      "epoch": 2.016383269206901,
+      "grad_norm": 0.12831463879417368,
+      "learning_rate": 2.4736120499783378e-06,
+      "loss": 0.6694,
+      "step": 4065
+    },
+    {
+      "epoch": 2.0168797319101404,
+      "grad_norm": 0.13082807641735877,
+      "learning_rate": 2.472634756834481e-06,
+      "loss": 0.6968,
+      "step": 4066
+    },
+    {
+      "epoch": 2.0173761946133797,
+      "grad_norm": 0.1274033272524088,
+      "learning_rate": 2.4716574678729627e-06,
+      "loss": 0.6998,
+      "step": 4067
+    },
+    {
+      "epoch": 2.017872657316619,
+      "grad_norm": 0.12742003523402126,
+      "learning_rate": 2.470680183243147e-06,
+      "loss": 0.6828,
+      "step": 4068
+    },
+    {
+      "epoch": 2.0183691200198584,
+      "grad_norm": 0.13490291668223267,
+      "learning_rate": 2.469702903094394e-06,
+      "loss": 0.7621,
+      "step": 4069
+    },
+    {
+      "epoch": 2.018865582723098,
+      "grad_norm": 0.13385281497853518,
+      "learning_rate": 2.4687256275760668e-06,
+      "loss": 0.7237,
+      "step": 4070
+    },
+    {
+      "epoch": 2.0193620454263375,
+      "grad_norm": 0.1360671850549657,
+      "learning_rate": 2.467748356837526e-06,
+      "loss": 0.7226,
+      "step": 4071
+    },
+    {
+      "epoch": 2.019858508129577,
+      "grad_norm": 0.13234055538841305,
+      "learning_rate": 2.4667710910281318e-06,
+      "loss": 0.7682,
+      "step": 4072
+    },
+    {
+      "epoch": 2.020354970832816,
+      "grad_norm": 0.12771711139534347,
+      "learning_rate": 2.465793830297244e-06,
+      "loss": 0.6369,
+      "step": 4073
+    },
+    {
+      "epoch": 2.0208514335360555,
+      "grad_norm": 0.13312672540262874,
+      "learning_rate": 2.4648165747942203e-06,
+      "loss": 0.7368,
+      "step": 4074
+    },
+    {
+      "epoch": 2.0213478962392952,
+      "grad_norm": 0.13105724105229263,
+      "learning_rate": 2.463839324668419e-06,
+      "loss": 0.7299,
+      "step": 4075
+    },
+    {
+      "epoch": 2.0218443589425346,
+      "grad_norm": 0.13275204495209755,
+      "learning_rate": 2.462862080069197e-06,
+      "loss": 0.7286,
+      "step": 4076
+    },
+    {
+      "epoch": 2.022340821645774,
+      "grad_norm": 0.13761097722270485,
+      "learning_rate": 2.461884841145911e-06,
+      "loss": 0.725,
+      "step": 4077
+    },
+    {
+      "epoch": 2.022837284349013,
+      "grad_norm": 0.13174698392561143,
+      "learning_rate": 2.460907608047916e-06,
+      "loss": 0.6977,
+      "step": 4078
+    },
+    {
+      "epoch": 2.0233337470522526,
+      "grad_norm": 0.14004069015911763,
+      "learning_rate": 2.459930380924566e-06,
+      "loss": 0.7458,
+      "step": 4079
+    },
+    {
+      "epoch": 2.0238302097554923,
+      "grad_norm": 0.129016892174802,
+      "learning_rate": 2.4589531599252155e-06,
+      "loss": 0.6907,
+      "step": 4080
+    },
+    {
+      "epoch": 2.0243266724587317,
+      "grad_norm": 0.13159681156179226,
+      "learning_rate": 2.4579759451992157e-06,
+      "loss": 0.7164,
+      "step": 4081
+    },
+    {
+      "epoch": 2.024823135161971,
+      "grad_norm": 0.13167520844662808,
+      "learning_rate": 2.4569987368959186e-06,
+      "loss": 0.7421,
+      "step": 4082
+    },
+    {
+      "epoch": 2.0253195978652103,
+      "grad_norm": 0.1363803013638871,
+      "learning_rate": 2.4560215351646752e-06,
+      "loss": 0.7367,
+      "step": 4083
+    },
+    {
+      "epoch": 2.0258160605684497,
+      "grad_norm": 0.12838138545946867,
+      "learning_rate": 2.4550443401548348e-06,
+      "loss": 0.7207,
+      "step": 4084
+    },
+    {
+      "epoch": 2.0263125232716894,
+      "grad_norm": 0.1316322999014349,
+      "learning_rate": 2.4540671520157474e-06,
+      "loss": 0.7001,
+      "step": 4085
+    },
+    {
+      "epoch": 2.0268089859749288,
+      "grad_norm": 0.13525722121983028,
+      "learning_rate": 2.453089970896758e-06,
+      "loss": 0.7593,
+      "step": 4086
+    },
+    {
+      "epoch": 2.027305448678168,
+      "grad_norm": 0.1305004559151505,
+      "learning_rate": 2.452112796947215e-06,
+      "loss": 0.6767,
+      "step": 4087
+    },
+    {
+      "epoch": 2.0278019113814074,
+      "grad_norm": 0.13465746531659445,
+      "learning_rate": 2.451135630316463e-06,
+      "loss": 0.7024,
+      "step": 4088
+    },
+    {
+      "epoch": 2.0282983740846467,
+      "grad_norm": 0.12955552855480887,
+      "learning_rate": 2.4501584711538467e-06,
+      "loss": 0.6726,
+      "step": 4089
+    },
+    {
+      "epoch": 2.0287948367878865,
+      "grad_norm": 0.1320046748116342,
+      "learning_rate": 2.4491813196087087e-06,
+      "loss": 0.7145,
+      "step": 4090
+    },
+    {
+      "epoch": 2.029291299491126,
+      "grad_norm": 0.1426680668898575,
+      "learning_rate": 2.4482041758303914e-06,
+      "loss": 0.7211,
+      "step": 4091
+    },
+    {
+      "epoch": 2.029787762194365,
+      "grad_norm": 0.13091969364968287,
+      "learning_rate": 2.447227039968235e-06,
+      "loss": 0.6937,
+      "step": 4092
+    },
+    {
+      "epoch": 2.0302842248976045,
+      "grad_norm": 0.13053571724955956,
+      "learning_rate": 2.4462499121715794e-06,
+      "loss": 0.7057,
+      "step": 4093
+    },
+    {
+      "epoch": 2.030780687600844,
+      "grad_norm": 0.13244995057683245,
+      "learning_rate": 2.4452727925897633e-06,
+      "loss": 0.715,
+      "step": 4094
+    },
+    {
+      "epoch": 2.0312771503040836,
+      "grad_norm": 0.12980740024168078,
+      "learning_rate": 2.4442956813721235e-06,
+      "loss": 0.6907,
+      "step": 4095
+    },
+    {
+      "epoch": 2.031773613007323,
+      "grad_norm": 0.1320072156478787,
+      "learning_rate": 2.443318578667996e-06,
+      "loss": 0.7048,
+      "step": 4096
+    },
+    {
+      "epoch": 2.0322700757105623,
+      "grad_norm": 0.13288927327614064,
+      "learning_rate": 2.442341484626714e-06,
+      "loss": 0.6985,
+      "step": 4097
+    },
+    {
+      "epoch": 2.0327665384138016,
+      "grad_norm": 0.13044804570837723,
+      "learning_rate": 2.4413643993976114e-06,
+      "loss": 0.7046,
+      "step": 4098
+    },
+    {
+      "epoch": 2.033263001117041,
+      "grad_norm": 0.12860176814635504,
+      "learning_rate": 2.4403873231300206e-06,
+      "loss": 0.6514,
+      "step": 4099
+    },
+    {
+      "epoch": 2.0337594638202807,
+      "grad_norm": 0.12841049842685956,
+      "learning_rate": 2.4394102559732717e-06,
+      "loss": 0.7143,
+      "step": 4100
+    },
+    {
+      "epoch": 2.03425592652352,
+      "grad_norm": 0.1383324791180656,
+      "learning_rate": 2.438433198076694e-06,
+      "loss": 0.7236,
+      "step": 4101
+    },
+    {
+      "epoch": 2.0347523892267594,
+      "grad_norm": 0.1334274425685486,
+      "learning_rate": 2.437456149589614e-06,
+      "loss": 0.717,
+      "step": 4102
+    },
+    {
+      "epoch": 2.0352488519299987,
+      "grad_norm": 0.1328619959886049,
+      "learning_rate": 2.4364791106613596e-06,
+      "loss": 0.7392,
+      "step": 4103
+    },
+    {
+      "epoch": 2.035745314633238,
+      "grad_norm": 0.12837963819339782,
+      "learning_rate": 2.435502081441254e-06,
+      "loss": 0.6838,
+      "step": 4104
+    },
+    {
+      "epoch": 2.0362417773364774,
+      "grad_norm": 0.1314972595674822,
+      "learning_rate": 2.434525062078622e-06,
+      "loss": 0.7104,
+      "step": 4105
+    },
+    {
+      "epoch": 2.036738240039717,
+      "grad_norm": 0.1376872036291365,
+      "learning_rate": 2.4335480527227833e-06,
+      "loss": 0.7023,
+      "step": 4106
+    },
+    {
+      "epoch": 2.0372347027429565,
+      "grad_norm": 0.12746910204547873,
+      "learning_rate": 2.43257105352306e-06,
+      "loss": 0.7065,
+      "step": 4107
+    },
+    {
+      "epoch": 2.037731165446196,
+      "grad_norm": 0.12941269428669738,
+      "learning_rate": 2.4315940646287693e-06,
+      "loss": 0.6786,
+      "step": 4108
+    },
+    {
+      "epoch": 2.038227628149435,
+      "grad_norm": 0.1334515747366998,
+      "learning_rate": 2.4306170861892293e-06,
+      "loss": 0.7258,
+      "step": 4109
+    },
+    {
+      "epoch": 2.0387240908526745,
+      "grad_norm": 0.13098506524365233,
+      "learning_rate": 2.429640118353756e-06,
+      "loss": 0.6948,
+      "step": 4110
+    },
+    {
+      "epoch": 2.0392205535559143,
+      "grad_norm": 0.1314973323181186,
+      "learning_rate": 2.4286631612716623e-06,
+      "loss": 0.6943,
+      "step": 4111
+    },
+    {
+      "epoch": 2.0397170162591536,
+      "grad_norm": 0.12965431662554677,
+      "learning_rate": 2.427686215092261e-06,
+      "loss": 0.6807,
+      "step": 4112
+    },
+    {
+      "epoch": 2.040213478962393,
+      "grad_norm": 0.13024307058284337,
+      "learning_rate": 2.426709279964861e-06,
+      "loss": 0.7308,
+      "step": 4113
+    },
+    {
+      "epoch": 2.0407099416656322,
+      "grad_norm": 0.13033943720971095,
+      "learning_rate": 2.425732356038773e-06,
+      "loss": 0.6952,
+      "step": 4114
+    },
+    {
+      "epoch": 2.0412064043688716,
+      "grad_norm": 0.12922563196874387,
+      "learning_rate": 2.424755443463303e-06,
+      "loss": 0.7066,
+      "step": 4115
+    },
+    {
+      "epoch": 2.0417028670721113,
+      "grad_norm": 0.13643925819128516,
+      "learning_rate": 2.4237785423877576e-06,
+      "loss": 0.7388,
+      "step": 4116
+    },
+    {
+      "epoch": 2.0421993297753507,
+      "grad_norm": 0.1285865306223446,
+      "learning_rate": 2.42280165296144e-06,
+      "loss": 0.753,
+      "step": 4117
+    },
+    {
+      "epoch": 2.04269579247859,
+      "grad_norm": 0.12885979273302448,
+      "learning_rate": 2.421824775333652e-06,
+      "loss": 0.6723,
+      "step": 4118
+    },
+    {
+      "epoch": 2.0431922551818293,
+      "grad_norm": 0.13298912364217827,
+      "learning_rate": 2.420847909653693e-06,
+      "loss": 0.7522,
+      "step": 4119
+    },
+    {
+      "epoch": 2.0436887178850687,
+      "grad_norm": 0.12918140250053928,
+      "learning_rate": 2.4198710560708623e-06,
+      "loss": 0.6887,
+      "step": 4120
+    },
+    {
+      "epoch": 2.0441851805883084,
+      "grad_norm": 0.1298629305623038,
+      "learning_rate": 2.4188942147344557e-06,
+      "loss": 0.6918,
+      "step": 4121
+    },
+    {
+      "epoch": 2.044681643291548,
+      "grad_norm": 0.1329165561864215,
+      "learning_rate": 2.4179173857937686e-06,
+      "loss": 0.7004,
+      "step": 4122
+    },
+    {
+      "epoch": 2.045178105994787,
+      "grad_norm": 0.13353203604650155,
+      "learning_rate": 2.4169405693980926e-06,
+      "loss": 0.6686,
+      "step": 4123
+    },
+    {
+      "epoch": 2.0456745686980264,
+      "grad_norm": 0.12807440054179015,
+      "learning_rate": 2.4159637656967185e-06,
+      "loss": 0.6834,
+      "step": 4124
+    },
+    {
+      "epoch": 2.0461710314012658,
+      "grad_norm": 0.13435476970975527,
+      "learning_rate": 2.4149869748389355e-06,
+      "loss": 0.713,
+      "step": 4125
+    },
+    {
+      "epoch": 2.0466674941045055,
+      "grad_norm": 0.13449883152953462,
+      "learning_rate": 2.4140101969740305e-06,
+      "loss": 0.7422,
+      "step": 4126
+    },
+    {
+      "epoch": 2.047163956807745,
+      "grad_norm": 0.12884574289334033,
+      "learning_rate": 2.413033432251289e-06,
+      "loss": 0.7106,
+      "step": 4127
+    },
+    {
+      "epoch": 2.047660419510984,
+      "grad_norm": 0.13241382070645977,
+      "learning_rate": 2.412056680819992e-06,
+      "loss": 0.7156,
+      "step": 4128
+    },
+    {
+      "epoch": 2.0481568822142235,
+      "grad_norm": 0.13077843297811131,
+      "learning_rate": 2.4110799428294214e-06,
+      "loss": 0.669,
+      "step": 4129
+    },
+    {
+      "epoch": 2.048653344917463,
+      "grad_norm": 0.12749052274478226,
+      "learning_rate": 2.4101032184288558e-06,
+      "loss": 0.6908,
+      "step": 4130
+    },
+    {
+      "epoch": 2.0491498076207026,
+      "grad_norm": 0.12836437459659322,
+      "learning_rate": 2.4091265077675716e-06,
+      "loss": 0.7328,
+      "step": 4131
+    },
+    {
+      "epoch": 2.049646270323942,
+      "grad_norm": 0.13191584290268923,
+      "learning_rate": 2.4081498109948435e-06,
+      "loss": 0.7269,
+      "step": 4132
+    },
+    {
+      "epoch": 2.0501427330271813,
+      "grad_norm": 0.1336022422875839,
+      "learning_rate": 2.407173128259945e-06,
+      "loss": 0.746,
+      "step": 4133
+    },
+    {
+      "epoch": 2.0506391957304206,
+      "grad_norm": 0.12850600170285384,
+      "learning_rate": 2.406196459712145e-06,
+      "loss": 0.6999,
+      "step": 4134
+    },
+    {
+      "epoch": 2.05113565843366,
+      "grad_norm": 0.1280500632235552,
+      "learning_rate": 2.4052198055007117e-06,
+      "loss": 0.7095,
+      "step": 4135
+    },
+    {
+      "epoch": 2.0516321211368997,
+      "grad_norm": 0.13308799322438067,
+      "learning_rate": 2.404243165774912e-06,
+      "loss": 0.7411,
+      "step": 4136
+    },
+    {
+      "epoch": 2.052128583840139,
+      "grad_norm": 0.12920055964205324,
+      "learning_rate": 2.4032665406840084e-06,
+      "loss": 0.6944,
+      "step": 4137
+    },
+    {
+      "epoch": 2.0526250465433784,
+      "grad_norm": 0.1349853449657708,
+      "learning_rate": 2.402289930377264e-06,
+      "loss": 0.7414,
+      "step": 4138
+    },
+    {
+      "epoch": 2.0531215092466177,
+      "grad_norm": 0.12839001848947515,
+      "learning_rate": 2.4013133350039366e-06,
+      "loss": 0.6611,
+      "step": 4139
+    },
+    {
+      "epoch": 2.053617971949857,
+      "grad_norm": 0.12960386307340557,
+      "learning_rate": 2.4003367547132833e-06,
+      "loss": 0.6837,
+      "step": 4140
+    },
+    {
+      "epoch": 2.054114434653097,
+      "grad_norm": 0.12992168154334086,
+      "learning_rate": 2.3993601896545593e-06,
+      "loss": 0.7583,
+      "step": 4141
+    },
+    {
+      "epoch": 2.054610897356336,
+      "grad_norm": 0.13162250535061173,
+      "learning_rate": 2.398383639977017e-06,
+      "loss": 0.6762,
+      "step": 4142
+    },
+    {
+      "epoch": 2.0551073600595755,
+      "grad_norm": 0.1322234940484418,
+      "learning_rate": 2.3974071058299063e-06,
+      "loss": 0.6959,
+      "step": 4143
+    },
+    {
+      "epoch": 2.055603822762815,
+      "grad_norm": 0.13363673998654405,
+      "learning_rate": 2.3964305873624748e-06,
+      "loss": 0.6872,
+      "step": 4144
+    },
+    {
+      "epoch": 2.056100285466054,
+      "grad_norm": 0.1315605491232442,
+      "learning_rate": 2.3954540847239663e-06,
+      "loss": 0.7403,
+      "step": 4145
+    },
+    {
+      "epoch": 2.056596748169294,
+      "grad_norm": 0.13173975211915448,
+      "learning_rate": 2.394477598063625e-06,
+      "loss": 0.7021,
+      "step": 4146
+    },
+    {
+      "epoch": 2.0570932108725333,
+      "grad_norm": 0.12934541651577286,
+      "learning_rate": 2.3935011275306907e-06,
+      "loss": 0.6995,
+      "step": 4147
+    },
+    {
+      "epoch": 2.0575896735757726,
+      "grad_norm": 0.1337787954167388,
+      "learning_rate": 2.3925246732744014e-06,
+      "loss": 0.6788,
+      "step": 4148
+    },
+    {
+      "epoch": 2.058086136279012,
+      "grad_norm": 0.1409884703911019,
+      "learning_rate": 2.3915482354439935e-06,
+      "loss": 0.6618,
+      "step": 4149
+    },
+    {
+      "epoch": 2.0585825989822513,
+      "grad_norm": 0.13863027742696146,
+      "learning_rate": 2.390571814188698e-06,
+      "loss": 0.6907,
+      "step": 4150
+    },
+    {
+      "epoch": 2.059079061685491,
+      "grad_norm": 0.1264859190593734,
+      "learning_rate": 2.3895954096577466e-06,
+      "loss": 0.715,
+      "step": 4151
+    },
+    {
+      "epoch": 2.0595755243887304,
+      "grad_norm": 0.12838817948703266,
+      "learning_rate": 2.388619022000366e-06,
+      "loss": 0.6928,
+      "step": 4152
+    },
+    {
+      "epoch": 2.0600719870919697,
+      "grad_norm": 0.1382264776368354,
+      "learning_rate": 2.3876426513657823e-06,
+      "loss": 0.7077,
+      "step": 4153
+    },
+    {
+      "epoch": 2.060568449795209,
+      "grad_norm": 0.13252726490660227,
+      "learning_rate": 2.3866662979032183e-06,
+      "loss": 0.6868,
+      "step": 4154
+    },
+    {
+      "epoch": 2.0610649124984484,
+      "grad_norm": 0.13082652966537198,
+      "learning_rate": 2.385689961761893e-06,
+      "loss": 0.684,
+      "step": 4155
+    },
+    {
+      "epoch": 2.061561375201688,
+      "grad_norm": 0.1330519426707675,
+      "learning_rate": 2.384713643091025e-06,
+      "loss": 0.6788,
+      "step": 4156
+    },
+    {
+      "epoch": 2.0620578379049275,
+      "grad_norm": 0.12951338154381573,
+      "learning_rate": 2.3837373420398274e-06,
+      "loss": 0.7167,
+      "step": 4157
+    },
+    {
+      "epoch": 2.062554300608167,
+      "grad_norm": 0.12959226185363074,
+      "learning_rate": 2.382761058757513e-06,
+      "loss": 0.6929,
+      "step": 4158
+    },
+    {
+      "epoch": 2.063050763311406,
+      "grad_norm": 0.13289167957939557,
+      "learning_rate": 2.381784793393292e-06,
+      "loss": 0.7145,
+      "step": 4159
+    },
+    {
+      "epoch": 2.0635472260146455,
+      "grad_norm": 0.13354521709268397,
+      "learning_rate": 2.3808085460963686e-06,
+      "loss": 0.6988,
+      "step": 4160
+    },
+    {
+      "epoch": 2.0640436887178852,
+      "grad_norm": 0.13397190629908656,
+      "learning_rate": 2.3798323170159487e-06,
+      "loss": 0.7283,
+      "step": 4161
+    },
+    {
+      "epoch": 2.0645401514211246,
+      "grad_norm": 0.13299819686134604,
+      "learning_rate": 2.378856106301232e-06,
+      "loss": 0.7049,
+      "step": 4162
+    },
+    {
+      "epoch": 2.065036614124364,
+      "grad_norm": 0.1288123646274204,
+      "learning_rate": 2.377879914101417e-06,
+      "loss": 0.7245,
+      "step": 4163
+    },
+    {
+      "epoch": 2.0655330768276032,
+      "grad_norm": 0.1393615916081668,
+      "learning_rate": 2.376903740565699e-06,
+      "loss": 0.6912,
+      "step": 4164
+    },
+    {
+      "epoch": 2.0660295395308426,
+      "grad_norm": 0.12457000357374992,
+      "learning_rate": 2.375927585843271e-06,
+      "loss": 0.6801,
+      "step": 4165
+    },
+    {
+      "epoch": 2.0665260022340823,
+      "grad_norm": 0.12940223030790443,
+      "learning_rate": 2.3749514500833218e-06,
+      "loss": 0.7121,
+      "step": 4166
+    },
+    {
+      "epoch": 2.0670224649373217,
+      "grad_norm": 0.1382124046186019,
+      "learning_rate": 2.373975333435038e-06,
+      "loss": 0.7042,
+      "step": 4167
+    },
+    {
+      "epoch": 2.067518927640561,
+      "grad_norm": 0.13543223125625606,
+      "learning_rate": 2.3729992360476047e-06,
+      "loss": 0.6908,
+      "step": 4168
+    },
+    {
+      "epoch": 2.0680153903438003,
+      "grad_norm": 0.1367570277824597,
+      "learning_rate": 2.372023158070201e-06,
+      "loss": 0.7385,
+      "step": 4169
+    },
+    {
+      "epoch": 2.0685118530470397,
+      "grad_norm": 0.12658280910237,
+      "learning_rate": 2.3710470996520067e-06,
+      "loss": 0.6485,
+      "step": 4170
+    },
+    {
+      "epoch": 2.0690083157502794,
+      "grad_norm": 0.1318360238221834,
+      "learning_rate": 2.370071060942195e-06,
+      "loss": 0.7073,
+      "step": 4171
+    },
+    {
+      "epoch": 2.0695047784535188,
+      "grad_norm": 0.1355974621244026,
+      "learning_rate": 2.369095042089938e-06,
+      "loss": 0.7383,
+      "step": 4172
+    },
+    {
+      "epoch": 2.070001241156758,
+      "grad_norm": 0.1354835108761855,
+      "learning_rate": 2.368119043244405e-06,
+      "loss": 0.7249,
+      "step": 4173
+    },
+    {
+      "epoch": 2.0704977038599974,
+      "grad_norm": 0.1316834160537684,
+      "learning_rate": 2.3671430645547622e-06,
+      "loss": 0.718,
+      "step": 4174
+    },
+    {
+      "epoch": 2.0709941665632368,
+      "grad_norm": 0.1337398224105994,
+      "learning_rate": 2.3661671061701725e-06,
+      "loss": 0.7617,
+      "step": 4175
+    },
+    {
+      "epoch": 2.0714906292664765,
+      "grad_norm": 0.13071098977935075,
+      "learning_rate": 2.3651911682397937e-06,
+      "loss": 0.7205,
+      "step": 4176
+    },
+    {
+      "epoch": 2.071987091969716,
+      "grad_norm": 0.13108332634784578,
+      "learning_rate": 2.3642152509127837e-06,
+      "loss": 0.7222,
+      "step": 4177
+    },
+    {
+      "epoch": 2.072483554672955,
+      "grad_norm": 0.12917550227439295,
+      "learning_rate": 2.363239354338295e-06,
+      "loss": 0.6863,
+      "step": 4178
+    },
+    {
+      "epoch": 2.0729800173761945,
+      "grad_norm": 0.13058903950875422,
+      "learning_rate": 2.3622634786654787e-06,
+      "loss": 0.741,
+      "step": 4179
+    },
+    {
+      "epoch": 2.073476480079434,
+      "grad_norm": 0.1317101832126491,
+      "learning_rate": 2.361287624043481e-06,
+      "loss": 0.69,
+      "step": 4180
+    },
+    {
+      "epoch": 2.0739729427826736,
+      "grad_norm": 0.13179285865877882,
+      "learning_rate": 2.3603117906214463e-06,
+      "loss": 0.6874,
+      "step": 4181
+    },
+    {
+      "epoch": 2.074469405485913,
+      "grad_norm": 0.13074448118330892,
+      "learning_rate": 2.3593359785485143e-06,
+      "loss": 0.7324,
+      "step": 4182
+    },
+    {
+      "epoch": 2.0749658681891523,
+      "grad_norm": 0.13408000673376183,
+      "learning_rate": 2.358360187973822e-06,
+      "loss": 0.7242,
+      "step": 4183
+    },
+    {
+      "epoch": 2.0754623308923916,
+      "grad_norm": 0.13011307891286955,
+      "learning_rate": 2.3573844190465046e-06,
+      "loss": 0.6941,
+      "step": 4184
+    },
+    {
+      "epoch": 2.075958793595631,
+      "grad_norm": 0.12572911313821467,
+      "learning_rate": 2.356408671915692e-06,
+      "loss": 0.6861,
+      "step": 4185
+    },
+    {
+      "epoch": 2.0764552562988707,
+      "grad_norm": 0.13279434671974327,
+      "learning_rate": 2.355432946730512e-06,
+      "loss": 0.7343,
+      "step": 4186
+    },
+    {
+      "epoch": 2.07695171900211,
+      "grad_norm": 0.1375180906665279,
+      "learning_rate": 2.3544572436400876e-06,
+      "loss": 0.7285,
+      "step": 4187
+    },
+    {
+      "epoch": 2.0774481817053494,
+      "grad_norm": 0.13344354731066305,
+      "learning_rate": 2.3534815627935397e-06,
+      "loss": 0.7002,
+      "step": 4188
+    },
+    {
+      "epoch": 2.0779446444085887,
+      "grad_norm": 0.12763100455109364,
+      "learning_rate": 2.352505904339986e-06,
+      "loss": 0.6759,
+      "step": 4189
+    },
+    {
+      "epoch": 2.078441107111828,
+      "grad_norm": 0.12628191599075542,
+      "learning_rate": 2.3515302684285398e-06,
+      "loss": 0.6725,
+      "step": 4190
+    },
+    {
+      "epoch": 2.078937569815068,
+      "grad_norm": 0.13183736319992664,
+      "learning_rate": 2.350554655208313e-06,
+      "loss": 0.6876,
+      "step": 4191
+    },
+    {
+      "epoch": 2.079434032518307,
+      "grad_norm": 0.1326348427538915,
+      "learning_rate": 2.349579064828409e-06,
+      "loss": 0.7015,
+      "step": 4192
+    },
+    {
+      "epoch": 2.0799304952215465,
+      "grad_norm": 0.12900392788556156,
+      "learning_rate": 2.3486034974379344e-06,
+      "loss": 0.6592,
+      "step": 4193
+    },
+    {
+      "epoch": 2.080426957924786,
+      "grad_norm": 0.12916628410416808,
+      "learning_rate": 2.347627953185987e-06,
+      "loss": 0.6965,
+      "step": 4194
+    },
+    {
+      "epoch": 2.080923420628025,
+      "grad_norm": 0.13911006307022702,
+      "learning_rate": 2.346652432221664e-06,
+      "loss": 0.7477,
+      "step": 4195
+    },
+    {
+      "epoch": 2.081419883331265,
+      "grad_norm": 0.12991109186278407,
+      "learning_rate": 2.3456769346940578e-06,
+      "loss": 0.7108,
+      "step": 4196
+    },
+    {
+      "epoch": 2.0819163460345043,
+      "grad_norm": 0.1310175516830794,
+      "learning_rate": 2.344701460752258e-06,
+      "loss": 0.72,
+      "step": 4197
+    },
+    {
+      "epoch": 2.0824128087377436,
+      "grad_norm": 0.1302438861387863,
+      "learning_rate": 2.34372601054535e-06,
+      "loss": 0.7045,
+      "step": 4198
+    },
+    {
+      "epoch": 2.082909271440983,
+      "grad_norm": 0.131287589350185,
+      "learning_rate": 2.3427505842224153e-06,
+      "loss": 0.6859,
+      "step": 4199
+    },
+    {
+      "epoch": 2.0834057341442223,
+      "grad_norm": 0.13636020184165526,
+      "learning_rate": 2.3417751819325327e-06,
+      "loss": 0.7343,
+      "step": 4200
+    },
+    {
+      "epoch": 2.083902196847462,
+      "grad_norm": 0.13036105078338428,
+      "learning_rate": 2.3407998038247767e-06,
+      "loss": 0.6583,
+      "step": 4201
+    },
+    {
+      "epoch": 2.0843986595507014,
+      "grad_norm": 0.13703412523752168,
+      "learning_rate": 2.339824450048218e-06,
+      "loss": 0.6875,
+      "step": 4202
+    },
+    {
+      "epoch": 2.0848951222539407,
+      "grad_norm": 0.1273783968662836,
+      "learning_rate": 2.3388491207519237e-06,
+      "loss": 0.6739,
+      "step": 4203
+    },
+    {
+      "epoch": 2.08539158495718,
+      "grad_norm": 0.1293141936432381,
+      "learning_rate": 2.337873816084958e-06,
+      "loss": 0.6807,
+      "step": 4204
+    },
+    {
+      "epoch": 2.0858880476604194,
+      "grad_norm": 0.13247637384139838,
+      "learning_rate": 2.3368985361963793e-06,
+      "loss": 0.7367,
+      "step": 4205
+    },
+    {
+      "epoch": 2.086384510363659,
+      "grad_norm": 0.12888079453701196,
+      "learning_rate": 2.3359232812352444e-06,
+      "loss": 0.6811,
+      "step": 4206
+    },
+    {
+      "epoch": 2.0868809730668985,
+      "grad_norm": 0.14305169239978224,
+      "learning_rate": 2.334948051350606e-06,
+      "loss": 0.6632,
+      "step": 4207
+    },
+    {
+      "epoch": 2.087377435770138,
+      "grad_norm": 0.13761847185157933,
+      "learning_rate": 2.333972846691511e-06,
+      "loss": 0.7505,
+      "step": 4208
+    },
+    {
+      "epoch": 2.087873898473377,
+      "grad_norm": 0.13472353385622,
+      "learning_rate": 2.3329976674070045e-06,
+      "loss": 0.7507,
+      "step": 4209
+    },
+    {
+      "epoch": 2.0883703611766165,
+      "grad_norm": 0.1332250195890825,
+      "learning_rate": 2.332022513646126e-06,
+      "loss": 0.686,
+      "step": 4210
+    },
+    {
+      "epoch": 2.0888668238798562,
+      "grad_norm": 0.13094514564456525,
+      "learning_rate": 2.331047385557914e-06,
+      "loss": 0.6997,
+      "step": 4211
+    },
+    {
+      "epoch": 2.0893632865830956,
+      "grad_norm": 0.13537482547078766,
+      "learning_rate": 2.3300722832913995e-06,
+      "loss": 0.6934,
+      "step": 4212
+    },
+    {
+      "epoch": 2.089859749286335,
+      "grad_norm": 0.13110966344634223,
+      "learning_rate": 2.329097206995612e-06,
+      "loss": 0.701,
+      "step": 4213
+    },
+    {
+      "epoch": 2.090356211989574,
+      "grad_norm": 0.13289972655864118,
+      "learning_rate": 2.3281221568195765e-06,
+      "loss": 0.6511,
+      "step": 4214
+    },
+    {
+      "epoch": 2.0908526746928136,
+      "grad_norm": 0.13256649712938065,
+      "learning_rate": 2.3271471329123134e-06,
+      "loss": 0.7757,
+      "step": 4215
+    },
+    {
+      "epoch": 2.0913491373960533,
+      "grad_norm": 0.12718273334445213,
+      "learning_rate": 2.3261721354228394e-06,
+      "loss": 0.6746,
+      "step": 4216
+    },
+    {
+      "epoch": 2.0918456000992927,
+      "grad_norm": 0.1279366898208473,
+      "learning_rate": 2.325197164500167e-06,
+      "loss": 0.7287,
+      "step": 4217
+    },
+    {
+      "epoch": 2.092342062802532,
+      "grad_norm": 0.1267772446319626,
+      "learning_rate": 2.324222220293306e-06,
+      "loss": 0.7235,
+      "step": 4218
+    },
+    {
+      "epoch": 2.0928385255057713,
+      "grad_norm": 0.12657392759783875,
+      "learning_rate": 2.3232473029512603e-06,
+      "loss": 0.67,
+      "step": 4219
+    },
+    {
+      "epoch": 2.0933349882090106,
+      "grad_norm": 0.13143006626346396,
+      "learning_rate": 2.3222724126230296e-06,
+      "loss": 0.6967,
+      "step": 4220
+    },
+    {
+      "epoch": 2.0938314509122504,
+      "grad_norm": 0.13250048306785056,
+      "learning_rate": 2.3212975494576113e-06,
+      "loss": 0.6659,
+      "step": 4221
+    },
+    {
+      "epoch": 2.0943279136154898,
+      "grad_norm": 0.1304492685665225,
+      "learning_rate": 2.3203227136039973e-06,
+      "loss": 0.754,
+      "step": 4222
+    },
+    {
+      "epoch": 2.094824376318729,
+      "grad_norm": 0.14113302529355157,
+      "learning_rate": 2.3193479052111763e-06,
+      "loss": 0.7788,
+      "step": 4223
+    },
+    {
+      "epoch": 2.0953208390219684,
+      "grad_norm": 0.12947346258330397,
+      "learning_rate": 2.3183731244281307e-06,
+      "loss": 0.6974,
+      "step": 4224
+    },
+    {
+      "epoch": 2.0958173017252077,
+      "grad_norm": 0.12969243044770545,
+      "learning_rate": 2.3173983714038413e-06,
+      "loss": 0.6881,
+      "step": 4225
+    },
+    {
+      "epoch": 2.0963137644284475,
+      "grad_norm": 0.12805474797723002,
+      "learning_rate": 2.3164236462872826e-06,
+      "loss": 0.6631,
+      "step": 4226
+    },
+    {
+      "epoch": 2.096810227131687,
+      "grad_norm": 0.12991846472801588,
+      "learning_rate": 2.315448949227426e-06,
+      "loss": 0.7117,
+      "step": 4227
+    },
+    {
+      "epoch": 2.097306689834926,
+      "grad_norm": 0.12737529203692569,
+      "learning_rate": 2.3144742803732386e-06,
+      "loss": 0.6619,
+      "step": 4228
+    },
+    {
+      "epoch": 2.0978031525381655,
+      "grad_norm": 0.1348388296052507,
+      "learning_rate": 2.3134996398736827e-06,
+      "loss": 0.6702,
+      "step": 4229
+    },
+    {
+      "epoch": 2.098299615241405,
+      "grad_norm": 0.12949314939302115,
+      "learning_rate": 2.3125250278777173e-06,
+      "loss": 0.6692,
+      "step": 4230
+    },
+    {
+      "epoch": 2.0987960779446446,
+      "grad_norm": 0.1319653430303982,
+      "learning_rate": 2.3115504445342947e-06,
+      "loss": 0.7186,
+      "step": 4231
+    },
+    {
+      "epoch": 2.099292540647884,
+      "grad_norm": 0.12754394160183852,
+      "learning_rate": 2.3105758899923652e-06,
+      "loss": 0.6983,
+      "step": 4232
+    },
+    {
+      "epoch": 2.0997890033511233,
+      "grad_norm": 0.13105593757086795,
+      "learning_rate": 2.3096013644008738e-06,
+      "loss": 0.7449,
+      "step": 4233
+    },
+    {
+      "epoch": 2.1002854660543626,
+      "grad_norm": 0.1352001782140455,
+      "learning_rate": 2.308626867908761e-06,
+      "loss": 0.7227,
+      "step": 4234
+    },
+    {
+      "epoch": 2.100781928757602,
+      "grad_norm": 0.1291811312002174,
+      "learning_rate": 2.3076524006649633e-06,
+      "loss": 0.6668,
+      "step": 4235
+    },
+    {
+      "epoch": 2.1012783914608413,
+      "grad_norm": 0.13013537752074916,
+      "learning_rate": 2.3066779628184115e-06,
+      "loss": 0.6691,
+      "step": 4236
+    },
+    {
+      "epoch": 2.101774854164081,
+      "grad_norm": 0.13054657602863368,
+      "learning_rate": 2.3057035545180337e-06,
+      "loss": 0.6978,
+      "step": 4237
+    },
+    {
+      "epoch": 2.1022713168673204,
+      "grad_norm": 0.12839328024574606,
+      "learning_rate": 2.3047291759127523e-06,
+      "loss": 0.6554,
+      "step": 4238
+    },
+    {
+      "epoch": 2.1027677795705597,
+      "grad_norm": 0.12996764440677216,
+      "learning_rate": 2.3037548271514863e-06,
+      "loss": 0.7042,
+      "step": 4239
+    },
+    {
+      "epoch": 2.103264242273799,
+      "grad_norm": 0.13523019374688494,
+      "learning_rate": 2.3027805083831475e-06,
+      "loss": 0.7477,
+      "step": 4240
+    },
+    {
+      "epoch": 2.103760704977039,
+      "grad_norm": 0.13283035054495967,
+      "learning_rate": 2.3018062197566464e-06,
+      "loss": 0.7262,
+      "step": 4241
+    },
+    {
+      "epoch": 2.104257167680278,
+      "grad_norm": 0.12953670758307892,
+      "learning_rate": 2.3008319614208863e-06,
+      "loss": 0.723,
+      "step": 4242
+    },
+    {
+      "epoch": 2.1047536303835175,
+      "grad_norm": 0.13144160332075827,
+      "learning_rate": 2.299857733524767e-06,
+      "loss": 0.7333,
+      "step": 4243
+    },
+    {
+      "epoch": 2.105250093086757,
+      "grad_norm": 0.12885131945661563,
+      "learning_rate": 2.2988835362171845e-06,
+      "loss": 0.736,
+      "step": 4244
+    },
+    {
+      "epoch": 2.105746555789996,
+      "grad_norm": 0.13082507084952819,
+      "learning_rate": 2.297909369647029e-06,
+      "loss": 0.7058,
+      "step": 4245
+    },
+    {
+      "epoch": 2.1062430184932355,
+      "grad_norm": 0.1308644150778887,
+      "learning_rate": 2.2969352339631864e-06,
+      "loss": 0.7137,
+      "step": 4246
+    },
+    {
+      "epoch": 2.1067394811964752,
+      "grad_norm": 0.13130767461578424,
+      "learning_rate": 2.295961129314537e-06,
+      "loss": 0.6985,
+      "step": 4247
+    },
+    {
+      "epoch": 2.1072359438997146,
+      "grad_norm": 0.13689384112058248,
+      "learning_rate": 2.294987055849957e-06,
+      "loss": 0.7249,
+      "step": 4248
+    },
+    {
+      "epoch": 2.107732406602954,
+      "grad_norm": 0.13528534527442918,
+      "learning_rate": 2.2940130137183187e-06,
+      "loss": 0.7141,
+      "step": 4249
+    },
+    {
+      "epoch": 2.1082288693061932,
+      "grad_norm": 0.1367594631782608,
+      "learning_rate": 2.2930390030684886e-06,
+      "loss": 0.7424,
+      "step": 4250
+    },
+    {
+      "epoch": 2.108725332009433,
+      "grad_norm": 0.13326510053921575,
+      "learning_rate": 2.292065024049329e-06,
+      "loss": 0.732,
+      "step": 4251
+    },
+    {
+      "epoch": 2.1092217947126723,
+      "grad_norm": 0.13486667743727,
+      "learning_rate": 2.291091076809696e-06,
+      "loss": 0.7336,
+      "step": 4252
+    },
+    {
+      "epoch": 2.1097182574159117,
+      "grad_norm": 0.1342722797696259,
+      "learning_rate": 2.2901171614984422e-06,
+      "loss": 0.7149,
+      "step": 4253
+    },
+    {
+      "epoch": 2.110214720119151,
+      "grad_norm": 0.12916009534662892,
+      "learning_rate": 2.2891432782644148e-06,
+      "loss": 0.6769,
+      "step": 4254
+    },
+    {
+      "epoch": 2.1107111828223903,
+      "grad_norm": 0.12895780049088768,
+      "learning_rate": 2.288169427256458e-06,
+      "loss": 0.6941,
+      "step": 4255
+    },
+    {
+      "epoch": 2.1112076455256297,
+      "grad_norm": 0.13336065989780396,
+      "learning_rate": 2.2871956086234066e-06,
+      "loss": 0.7312,
+      "step": 4256
+    },
+    {
+      "epoch": 2.1117041082288694,
+      "grad_norm": 0.1311287040733774,
+      "learning_rate": 2.2862218225140948e-06,
+      "loss": 0.6849,
+      "step": 4257
+    },
+    {
+      "epoch": 2.112200570932109,
+      "grad_norm": 0.1326013584545713,
+      "learning_rate": 2.2852480690773493e-06,
+      "loss": 0.7127,
+      "step": 4258
+    },
+    {
+      "epoch": 2.112697033635348,
+      "grad_norm": 0.1293700176479527,
+      "learning_rate": 2.284274348461993e-06,
+      "loss": 0.6787,
+      "step": 4259
+    },
+    {
+      "epoch": 2.1131934963385874,
+      "grad_norm": 0.1270172759070305,
+      "learning_rate": 2.283300660816844e-06,
+      "loss": 0.7109,
+      "step": 4260
+    },
+    {
+      "epoch": 2.1136899590418268,
+      "grad_norm": 0.12987078013128955,
+      "learning_rate": 2.282327006290714e-06,
+      "loss": 0.6754,
+      "step": 4261
+    },
+    {
+      "epoch": 2.1141864217450665,
+      "grad_norm": 0.1311168878775037,
+      "learning_rate": 2.281353385032412e-06,
+      "loss": 0.6938,
+      "step": 4262
+    },
+    {
+      "epoch": 2.114682884448306,
+      "grad_norm": 0.13308422505081596,
+      "learning_rate": 2.2803797971907384e-06,
+      "loss": 0.7151,
+      "step": 4263
+    },
+    {
+      "epoch": 2.115179347151545,
+      "grad_norm": 0.1305774573301847,
+      "learning_rate": 2.2794062429144916e-06,
+      "loss": 0.7384,
+      "step": 4264
+    },
+    {
+      "epoch": 2.1156758098547845,
+      "grad_norm": 0.13262558018414852,
+      "learning_rate": 2.278432722352464e-06,
+      "loss": 0.7158,
+      "step": 4265
+    },
+    {
+      "epoch": 2.116172272558024,
+      "grad_norm": 0.1321143377532212,
+      "learning_rate": 2.2774592356534417e-06,
+      "loss": 0.644,
+      "step": 4266
+    },
+    {
+      "epoch": 2.1166687352612636,
+      "grad_norm": 0.13040322905026952,
+      "learning_rate": 2.276485782966208e-06,
+      "loss": 0.6894,
+      "step": 4267
+    },
+    {
+      "epoch": 2.117165197964503,
+      "grad_norm": 0.14208837528045287,
+      "learning_rate": 2.2755123644395377e-06,
+      "loss": 0.727,
+      "step": 4268
+    },
+    {
+      "epoch": 2.1176616606677423,
+      "grad_norm": 0.14092145083010946,
+      "learning_rate": 2.2745389802222034e-06,
+      "loss": 0.7016,
+      "step": 4269
+    },
+    {
+      "epoch": 2.1181581233709816,
+      "grad_norm": 0.12913632924933782,
+      "learning_rate": 2.2735656304629706e-06,
+      "loss": 0.7004,
+      "step": 4270
+    },
+    {
+      "epoch": 2.118654586074221,
+      "grad_norm": 0.1334111014407336,
+      "learning_rate": 2.272592315310601e-06,
+      "loss": 0.7373,
+      "step": 4271
+    },
+    {
+      "epoch": 2.1191510487774607,
+      "grad_norm": 0.13363016697183605,
+      "learning_rate": 2.2716190349138505e-06,
+      "loss": 0.7042,
+      "step": 4272
+    },
+    {
+      "epoch": 2.1196475114807,
+      "grad_norm": 0.13210258534803404,
+      "learning_rate": 2.270645789421468e-06,
+      "loss": 0.7621,
+      "step": 4273
+    },
+    {
+      "epoch": 2.1201439741839394,
+      "grad_norm": 0.13777719715311174,
+      "learning_rate": 2.2696725789821994e-06,
+      "loss": 0.7225,
+      "step": 4274
+    },
+    {
+      "epoch": 2.1206404368871787,
+      "grad_norm": 0.1308204005740167,
+      "learning_rate": 2.268699403744784e-06,
+      "loss": 0.7062,
+      "step": 4275
+    },
+    {
+      "epoch": 2.121136899590418,
+      "grad_norm": 0.12957459770630428,
+      "learning_rate": 2.2677262638579554e-06,
+      "loss": 0.6877,
+      "step": 4276
+    },
+    {
+      "epoch": 2.121633362293658,
+      "grad_norm": 0.13016124393616152,
+      "learning_rate": 2.266753159470444e-06,
+      "loss": 0.6826,
+      "step": 4277
+    },
+    {
+      "epoch": 2.122129824996897,
+      "grad_norm": 0.12806205702427592,
+      "learning_rate": 2.265780090730972e-06,
+      "loss": 0.645,
+      "step": 4278
+    },
+    {
+      "epoch": 2.1226262877001365,
+      "grad_norm": 0.1337185042729098,
+      "learning_rate": 2.2648070577882573e-06,
+      "loss": 0.7274,
+      "step": 4279
+    },
+    {
+      "epoch": 2.123122750403376,
+      "grad_norm": 0.13067486282767551,
+      "learning_rate": 2.2638340607910133e-06,
+      "loss": 0.7087,
+      "step": 4280
+    },
+    {
+      "epoch": 2.123619213106615,
+      "grad_norm": 0.13386479780042845,
+      "learning_rate": 2.2628610998879463e-06,
+      "loss": 0.7024,
+      "step": 4281
+    },
+    {
+      "epoch": 2.124115675809855,
+      "grad_norm": 0.13272729071166328,
+      "learning_rate": 2.261888175227758e-06,
+      "loss": 0.7484,
+      "step": 4282
+    },
+    {
+      "epoch": 2.1246121385130943,
+      "grad_norm": 0.1310493554217813,
+      "learning_rate": 2.2609152869591445e-06,
+      "loss": 0.651,
+      "step": 4283
+    },
+    {
+      "epoch": 2.1251086012163336,
+      "grad_norm": 0.1316640668680429,
+      "learning_rate": 2.2599424352307958e-06,
+      "loss": 0.755,
+      "step": 4284
+    },
+    {
+      "epoch": 2.125605063919573,
+      "grad_norm": 0.12948769043340413,
+      "learning_rate": 2.2589696201913966e-06,
+      "loss": 0.7538,
+      "step": 4285
+    },
+    {
+      "epoch": 2.1261015266228123,
+      "grad_norm": 0.1259209524836695,
+      "learning_rate": 2.2579968419896264e-06,
+      "loss": 0.7165,
+      "step": 4286
+    },
+    {
+      "epoch": 2.126597989326052,
+      "grad_norm": 0.1274147611602142,
+      "learning_rate": 2.257024100774159e-06,
+      "loss": 0.7053,
+      "step": 4287
+    },
+    {
+      "epoch": 2.1270944520292914,
+      "grad_norm": 0.1273913753461934,
+      "learning_rate": 2.2560513966936626e-06,
+      "loss": 0.6987,
+      "step": 4288
+    },
+    {
+      "epoch": 2.1275909147325307,
+      "grad_norm": 0.17368980947168566,
+      "learning_rate": 2.255078729896798e-06,
+      "loss": 0.6387,
+      "step": 4289
+    },
+    {
+      "epoch": 2.12808737743577,
+      "grad_norm": 0.12393658684142564,
+      "learning_rate": 2.254106100532223e-06,
+      "loss": 0.673,
+      "step": 4290
+    },
+    {
+      "epoch": 2.1285838401390094,
+      "grad_norm": 0.12933453592109587,
+      "learning_rate": 2.253133508748587e-06,
+      "loss": 0.7271,
+      "step": 4291
+    },
+    {
+      "epoch": 2.129080302842249,
+      "grad_norm": 0.1343441282973702,
+      "learning_rate": 2.252160954694536e-06,
+      "loss": 0.6757,
+      "step": 4292
+    },
+    {
+      "epoch": 2.1295767655454885,
+      "grad_norm": 0.13289063203957274,
+      "learning_rate": 2.2511884385187098e-06,
+      "loss": 0.735,
+      "step": 4293
+    },
+    {
+      "epoch": 2.130073228248728,
+      "grad_norm": 0.12926554854547287,
+      "learning_rate": 2.250215960369741e-06,
+      "loss": 0.6849,
+      "step": 4294
+    },
+    {
+      "epoch": 2.130569690951967,
+      "grad_norm": 0.13121955352587697,
+      "learning_rate": 2.2492435203962584e-06,
+      "loss": 0.6844,
+      "step": 4295
+    },
+    {
+      "epoch": 2.1310661536552065,
+      "grad_norm": 0.13785478772610635,
+      "learning_rate": 2.2482711187468825e-06,
+      "loss": 0.6656,
+      "step": 4296
+    },
+    {
+      "epoch": 2.1315626163584462,
+      "grad_norm": 0.13675030797419835,
+      "learning_rate": 2.24729875557023e-06,
+      "loss": 0.7755,
+      "step": 4297
+    },
+    {
+      "epoch": 2.1320590790616856,
+      "grad_norm": 0.13218540295978803,
+      "learning_rate": 2.246326431014911e-06,
+      "loss": 0.7865,
+      "step": 4298
+    },
+    {
+      "epoch": 2.132555541764925,
+      "grad_norm": 0.13058917038281245,
+      "learning_rate": 2.2453541452295304e-06,
+      "loss": 0.6822,
+      "step": 4299
+    },
+    {
+      "epoch": 2.1330520044681642,
+      "grad_norm": 0.1356025414078166,
+      "learning_rate": 2.2443818983626845e-06,
+      "loss": 0.7416,
+      "step": 4300
+    },
+    {
+      "epoch": 2.1335484671714036,
+      "grad_norm": 0.13259643816943026,
+      "learning_rate": 2.2434096905629675e-06,
+      "loss": 0.7331,
+      "step": 4301
+    },
+    {
+      "epoch": 2.1340449298746433,
+      "grad_norm": 0.13178426485776873,
+      "learning_rate": 2.242437521978965e-06,
+      "loss": 0.7416,
+      "step": 4302
+    },
+    {
+      "epoch": 2.1345413925778827,
+      "grad_norm": 0.12712500731582294,
+      "learning_rate": 2.2414653927592578e-06,
+      "loss": 0.6676,
+      "step": 4303
+    },
+    {
+      "epoch": 2.135037855281122,
+      "grad_norm": 0.132762631536009,
+      "learning_rate": 2.240493303052421e-06,
+      "loss": 0.7412,
+      "step": 4304
+    },
+    {
+      "epoch": 2.1355343179843613,
+      "grad_norm": 0.13927370521334112,
+      "learning_rate": 2.239521253007021e-06,
+      "loss": 0.7256,
+      "step": 4305
+    },
+    {
+      "epoch": 2.1360307806876007,
+      "grad_norm": 0.1298913459257255,
+      "learning_rate": 2.238549242771621e-06,
+      "loss": 0.726,
+      "step": 4306
+    },
+    {
+      "epoch": 2.1365272433908404,
+      "grad_norm": 0.13103490851669664,
+      "learning_rate": 2.237577272494777e-06,
+      "loss": 0.7101,
+      "step": 4307
+    },
+    {
+      "epoch": 2.1370237060940798,
+      "grad_norm": 0.13052143851126424,
+      "learning_rate": 2.2366053423250396e-06,
+      "loss": 0.7063,
+      "step": 4308
+    },
+    {
+      "epoch": 2.137520168797319,
+      "grad_norm": 0.1325952941973161,
+      "learning_rate": 2.235633452410952e-06,
+      "loss": 0.7102,
+      "step": 4309
+    },
+    {
+      "epoch": 2.1380166315005584,
+      "grad_norm": 0.13151130162140853,
+      "learning_rate": 2.2346616029010527e-06,
+      "loss": 0.72,
+      "step": 4310
+    },
+    {
+      "epoch": 2.1385130942037978,
+      "grad_norm": 0.13684162828630525,
+      "learning_rate": 2.2336897939438734e-06,
+      "loss": 0.7484,
+      "step": 4311
+    },
+    {
+      "epoch": 2.1390095569070375,
+      "grad_norm": 0.13016064599020155,
+      "learning_rate": 2.2327180256879384e-06,
+      "loss": 0.7219,
+      "step": 4312
+    },
+    {
+      "epoch": 2.139506019610277,
+      "grad_norm": 0.13271882669694315,
+      "learning_rate": 2.231746298281768e-06,
+      "loss": 0.7285,
+      "step": 4313
+    },
+    {
+      "epoch": 2.140002482313516,
+      "grad_norm": 0.12493929654212456,
+      "learning_rate": 2.230774611873875e-06,
+      "loss": 0.6597,
+      "step": 4314
+    },
+    {
+      "epoch": 2.1404989450167555,
+      "grad_norm": 0.1270152459435393,
+      "learning_rate": 2.2298029666127654e-06,
+      "loss": 0.6268,
+      "step": 4315
+    },
+    {
+      "epoch": 2.140995407719995,
+      "grad_norm": 0.13120756963636904,
+      "learning_rate": 2.2288313626469403e-06,
+      "loss": 0.7737,
+      "step": 4316
+    },
+    {
+      "epoch": 2.1414918704232346,
+      "grad_norm": 0.13093118304542808,
+      "learning_rate": 2.2278598001248935e-06,
+      "loss": 0.7182,
+      "step": 4317
+    },
+    {
+      "epoch": 2.141988333126474,
+      "grad_norm": 0.13406739346906132,
+      "learning_rate": 2.2268882791951125e-06,
+      "loss": 0.6923,
+      "step": 4318
+    },
+    {
+      "epoch": 2.1424847958297133,
+      "grad_norm": 0.1321265210500075,
+      "learning_rate": 2.2259168000060793e-06,
+      "loss": 0.7341,
+      "step": 4319
+    },
+    {
+      "epoch": 2.1429812585329526,
+      "grad_norm": 0.1297032816599765,
+      "learning_rate": 2.2249453627062697e-06,
+      "loss": 0.6591,
+      "step": 4320
+    },
+    {
+      "epoch": 2.143477721236192,
+      "grad_norm": 0.13306620569232747,
+      "learning_rate": 2.22397396744415e-06,
+      "loss": 0.7009,
+      "step": 4321
+    },
+    {
+      "epoch": 2.1439741839394317,
+      "grad_norm": 0.1315841117210727,
+      "learning_rate": 2.223002614368184e-06,
+      "loss": 0.693,
+      "step": 4322
+    },
+    {
+      "epoch": 2.144470646642671,
+      "grad_norm": 0.13522931404154925,
+      "learning_rate": 2.2220313036268275e-06,
+      "loss": 0.7885,
+      "step": 4323
+    },
+    {
+      "epoch": 2.1449671093459104,
+      "grad_norm": 0.13000494547408423,
+      "learning_rate": 2.2210600353685286e-06,
+      "loss": 0.7126,
+      "step": 4324
+    },
+    {
+      "epoch": 2.1454635720491497,
+      "grad_norm": 0.1331174145801834,
+      "learning_rate": 2.2200888097417308e-06,
+      "loss": 0.7385,
+      "step": 4325
+    },
+    {
+      "epoch": 2.145960034752389,
+      "grad_norm": 0.13199478030805042,
+      "learning_rate": 2.2191176268948707e-06,
+      "loss": 0.6859,
+      "step": 4326
+    },
+    {
+      "epoch": 2.146456497455629,
+      "grad_norm": 0.12924843473690692,
+      "learning_rate": 2.2181464869763777e-06,
+      "loss": 0.7077,
+      "step": 4327
+    },
+    {
+      "epoch": 2.146952960158868,
+      "grad_norm": 0.13383047347606078,
+      "learning_rate": 2.217175390134675e-06,
+      "loss": 0.7301,
+      "step": 4328
+    },
+    {
+      "epoch": 2.1474494228621075,
+      "grad_norm": 0.1301216163563912,
+      "learning_rate": 2.216204336518179e-06,
+      "loss": 0.7075,
+      "step": 4329
+    },
+    {
+      "epoch": 2.147945885565347,
+      "grad_norm": 0.13513976833741134,
+      "learning_rate": 2.2152333262752996e-06,
+      "loss": 0.7069,
+      "step": 4330
+    },
+    {
+      "epoch": 2.148442348268586,
+      "grad_norm": 0.12921645151422304,
+      "learning_rate": 2.2142623595544407e-06,
+      "loss": 0.7121,
+      "step": 4331
+    },
+    {
+      "epoch": 2.148938810971826,
+      "grad_norm": 0.134789032200641,
+      "learning_rate": 2.2132914365039993e-06,
+      "loss": 0.7279,
+      "step": 4332
+    },
+    {
+      "epoch": 2.1494352736750653,
+      "grad_norm": 0.12767095954972132,
+      "learning_rate": 2.2123205572723638e-06,
+      "loss": 0.6819,
+      "step": 4333
+    },
+    {
+      "epoch": 2.1499317363783046,
+      "grad_norm": 0.132607654917447,
+      "learning_rate": 2.211349722007919e-06,
+      "loss": 0.7035,
+      "step": 4334
+    },
+    {
+      "epoch": 2.150428199081544,
+      "grad_norm": 0.13081966486435517,
+      "learning_rate": 2.210378930859041e-06,
+      "loss": 0.7125,
+      "step": 4335
+    },
+    {
+      "epoch": 2.1509246617847833,
+      "grad_norm": 0.13392118968950856,
+      "learning_rate": 2.2094081839741007e-06,
+      "loss": 0.6874,
+      "step": 4336
+    },
+    {
+      "epoch": 2.151421124488023,
+      "grad_norm": 0.13563928410684517,
+      "learning_rate": 2.208437481501459e-06,
+      "loss": 0.7077,
+      "step": 4337
+    },
+    {
+      "epoch": 2.1519175871912624,
+      "grad_norm": 0.13102290237210468,
+      "learning_rate": 2.207466823589474e-06,
+      "loss": 0.7314,
+      "step": 4338
+    },
+    {
+      "epoch": 2.1524140498945017,
+      "grad_norm": 0.1308681684431675,
+      "learning_rate": 2.206496210386494e-06,
+      "loss": 0.6915,
+      "step": 4339
+    },
+    {
+      "epoch": 2.152910512597741,
+      "grad_norm": 0.13474477124109288,
+      "learning_rate": 2.2055256420408625e-06,
+      "loss": 0.7038,
+      "step": 4340
+    },
+    {
+      "epoch": 2.1534069753009804,
+      "grad_norm": 0.13030093914916455,
+      "learning_rate": 2.2045551187009147e-06,
+      "loss": 0.7066,
+      "step": 4341
+    },
+    {
+      "epoch": 2.15390343800422,
+      "grad_norm": 0.12838709690946015,
+      "learning_rate": 2.20358464051498e-06,
+      "loss": 0.6839,
+      "step": 4342
+    },
+    {
+      "epoch": 2.1543999007074595,
+      "grad_norm": 0.13054218481985586,
+      "learning_rate": 2.2026142076313816e-06,
+      "loss": 0.7189,
+      "step": 4343
+    },
+    {
+      "epoch": 2.154896363410699,
+      "grad_norm": 0.1329355796495672,
+      "learning_rate": 2.201643820198432e-06,
+      "loss": 0.7484,
+      "step": 4344
+    },
+    {
+      "epoch": 2.155392826113938,
+      "grad_norm": 0.12685915768729691,
+      "learning_rate": 2.2006734783644407e-06,
+      "loss": 0.6744,
+      "step": 4345
+    },
+    {
+      "epoch": 2.1558892888171775,
+      "grad_norm": 0.12969948887272129,
+      "learning_rate": 2.1997031822777093e-06,
+      "loss": 0.6559,
+      "step": 4346
+    },
+    {
+      "epoch": 2.1563857515204172,
+      "grad_norm": 0.12690777845885462,
+      "learning_rate": 2.1987329320865314e-06,
+      "loss": 0.6719,
+      "step": 4347
+    },
+    {
+      "epoch": 2.1568822142236566,
+      "grad_norm": 0.13746314892957628,
+      "learning_rate": 2.197762727939195e-06,
+      "loss": 0.7068,
+      "step": 4348
+    },
+    {
+      "epoch": 2.157378676926896,
+      "grad_norm": 0.13974129132766783,
+      "learning_rate": 2.196792569983979e-06,
+      "loss": 0.7702,
+      "step": 4349
+    },
+    {
+      "epoch": 2.157875139630135,
+      "grad_norm": 0.13205901471380596,
+      "learning_rate": 2.195822458369157e-06,
+      "loss": 0.6789,
+      "step": 4350
+    },
+    {
+      "epoch": 2.1583716023333746,
+      "grad_norm": 0.13151573181044662,
+      "learning_rate": 2.194852393242995e-06,
+      "loss": 0.6997,
+      "step": 4351
+    },
+    {
+      "epoch": 2.1588680650366143,
+      "grad_norm": 0.1311550041985863,
+      "learning_rate": 2.1938823747537535e-06,
+      "loss": 0.7186,
+      "step": 4352
+    },
+    {
+      "epoch": 2.1593645277398537,
+      "grad_norm": 0.12785154545568428,
+      "learning_rate": 2.192912403049682e-06,
+      "loss": 0.7337,
+      "step": 4353
+    },
+    {
+      "epoch": 2.159860990443093,
+      "grad_norm": 0.1384165497626903,
+      "learning_rate": 2.1919424782790256e-06,
+      "loss": 0.7423,
+      "step": 4354
+    },
+    {
+      "epoch": 2.1603574531463323,
+      "grad_norm": 0.1326394817355734,
+      "learning_rate": 2.1909726005900224e-06,
+      "loss": 0.7139,
+      "step": 4355
+    },
+    {
+      "epoch": 2.1608539158495716,
+      "grad_norm": 0.1332787692360546,
+      "learning_rate": 2.1900027701309016e-06,
+      "loss": 0.7198,
+      "step": 4356
+    },
+    {
+      "epoch": 2.1613503785528114,
+      "grad_norm": 0.13058676686648546,
+      "learning_rate": 2.1890329870498873e-06,
+      "loss": 0.6919,
+      "step": 4357
+    },
+    {
+      "epoch": 2.1618468412560508,
+      "grad_norm": 0.12920400653725772,
+      "learning_rate": 2.188063251495195e-06,
+      "loss": 0.7204,
+      "step": 4358
+    },
+    {
+      "epoch": 2.16234330395929,
+      "grad_norm": 0.12727242356432697,
+      "learning_rate": 2.187093563615034e-06,
+      "loss": 0.7072,
+      "step": 4359
+    },
+    {
+      "epoch": 2.1628397666625294,
+      "grad_norm": 0.12536367743791044,
+      "learning_rate": 2.1861239235576033e-06,
+      "loss": 0.6552,
+      "step": 4360
+    },
+    {
+      "epoch": 2.1633362293657687,
+      "grad_norm": 0.12786174220800478,
+      "learning_rate": 2.185154331471099e-06,
+      "loss": 0.6844,
+      "step": 4361
+    },
+    {
+      "epoch": 2.1638326920690085,
+      "grad_norm": 0.13320070492782934,
+      "learning_rate": 2.1841847875037065e-06,
+      "loss": 0.7398,
+      "step": 4362
+    },
+    {
+      "epoch": 2.164329154772248,
+      "grad_norm": 0.1388558108961887,
+      "learning_rate": 2.1832152918036058e-06,
+      "loss": 0.7204,
+      "step": 4363
+    },
+    {
+      "epoch": 2.164825617475487,
+      "grad_norm": 0.12693695099985525,
+      "learning_rate": 2.182245844518969e-06,
+      "loss": 0.677,
+      "step": 4364
+    },
+    {
+      "epoch": 2.1653220801787265,
+      "grad_norm": 0.12924014986291152,
+      "learning_rate": 2.1812764457979594e-06,
+      "loss": 0.6683,
+      "step": 4365
+    },
+    {
+      "epoch": 2.165818542881966,
+      "grad_norm": 0.13159270443065568,
+      "learning_rate": 2.1803070957887348e-06,
+      "loss": 0.7281,
+      "step": 4366
+    },
+    {
+      "epoch": 2.166315005585205,
+      "grad_norm": 0.13127772161024706,
+      "learning_rate": 2.1793377946394448e-06,
+      "loss": 0.6727,
+      "step": 4367
+    },
+    {
+      "epoch": 2.166811468288445,
+      "grad_norm": 0.12854307859962671,
+      "learning_rate": 2.1783685424982326e-06,
+      "loss": 0.6921,
+      "step": 4368
+    },
+    {
+      "epoch": 2.1673079309916843,
+      "grad_norm": 0.13019316986612073,
+      "learning_rate": 2.1773993395132303e-06,
+      "loss": 0.6875,
+      "step": 4369
+    },
+    {
+      "epoch": 2.1678043936949236,
+      "grad_norm": 0.1353274833386413,
+      "learning_rate": 2.176430185832567e-06,
+      "loss": 0.7027,
+      "step": 4370
+    },
+    {
+      "epoch": 2.168300856398163,
+      "grad_norm": 0.13265862920248359,
+      "learning_rate": 2.175461081604362e-06,
+      "loss": 0.6899,
+      "step": 4371
+    },
+    {
+      "epoch": 2.1687973191014027,
+      "grad_norm": 0.13334516546543757,
+      "learning_rate": 2.1744920269767266e-06,
+      "loss": 0.6968,
+      "step": 4372
+    },
+    {
+      "epoch": 2.169293781804642,
+      "grad_norm": 0.13704956210912356,
+      "learning_rate": 2.1735230220977653e-06,
+      "loss": 0.7129,
+      "step": 4373
+    },
+    {
+      "epoch": 2.1697902445078814,
+      "grad_norm": 0.13014552774663893,
+      "learning_rate": 2.172554067115576e-06,
+      "loss": 0.7324,
+      "step": 4374
+    },
+    {
+      "epoch": 2.1702867072111207,
+      "grad_norm": 0.1281661360302656,
+      "learning_rate": 2.1715851621782473e-06,
+      "loss": 0.6973,
+      "step": 4375
+    },
+    {
+      "epoch": 2.17078316991436,
+      "grad_norm": 0.13330711985579916,
+      "learning_rate": 2.170616307433861e-06,
+      "loss": 0.7038,
+      "step": 4376
+    },
+    {
+      "epoch": 2.1712796326175994,
+      "grad_norm": 0.13056351010723308,
+      "learning_rate": 2.1696475030304902e-06,
+      "loss": 0.6866,
+      "step": 4377
+    },
+    {
+      "epoch": 2.171776095320839,
+      "grad_norm": 0.13268943046062295,
+      "learning_rate": 2.1686787491162023e-06,
+      "loss": 0.6891,
+      "step": 4378
+    },
+    {
+      "epoch": 2.1722725580240785,
+      "grad_norm": 0.13924146096251458,
+      "learning_rate": 2.1677100458390547e-06,
+      "loss": 0.6916,
+      "step": 4379
+    },
+    {
+      "epoch": 2.172769020727318,
+      "grad_norm": 0.13972180628069247,
+      "learning_rate": 2.1667413933470998e-06,
+      "loss": 0.7463,
+      "step": 4380
+    },
+    {
+      "epoch": 2.173265483430557,
+      "grad_norm": 0.13661608957760332,
+      "learning_rate": 2.165772791788379e-06,
+      "loss": 0.7453,
+      "step": 4381
+    },
+    {
+      "epoch": 2.173761946133797,
+      "grad_norm": 0.1282782445633921,
+      "learning_rate": 2.1648042413109276e-06,
+      "loss": 0.7236,
+      "step": 4382
+    },
+    {
+      "epoch": 2.1742584088370362,
+      "grad_norm": 0.12716164432242413,
+      "learning_rate": 2.163835742062774e-06,
+      "loss": 0.687,
+      "step": 4383
+    },
+    {
+      "epoch": 2.1747548715402756,
+      "grad_norm": 0.13109851282826201,
+      "learning_rate": 2.162867294191938e-06,
+      "loss": 0.7189,
+      "step": 4384
+    },
+    {
+      "epoch": 2.175251334243515,
+      "grad_norm": 0.14106981155374845,
+      "learning_rate": 2.1618988978464296e-06,
+      "loss": 0.7263,
+      "step": 4385
+    },
+    {
+      "epoch": 2.1757477969467542,
+      "grad_norm": 0.1287290868986218,
+      "learning_rate": 2.1609305531742534e-06,
+      "loss": 0.6981,
+      "step": 4386
+    },
+    {
+      "epoch": 2.1762442596499936,
+      "grad_norm": 0.1294869956830014,
+      "learning_rate": 2.159962260323406e-06,
+      "loss": 0.7166,
+      "step": 4387
+    },
+    {
+      "epoch": 2.1767407223532333,
+      "grad_norm": 0.13558488604187074,
+      "learning_rate": 2.158994019441875e-06,
+      "loss": 0.6741,
+      "step": 4388
+    },
+    {
+      "epoch": 2.1772371850564727,
+      "grad_norm": 0.1337077290344641,
+      "learning_rate": 2.15802583067764e-06,
+      "loss": 0.712,
+      "step": 4389
+    },
+    {
+      "epoch": 2.177733647759712,
+      "grad_norm": 0.13200934226771094,
+      "learning_rate": 2.157057694178674e-06,
+      "loss": 0.6919,
+      "step": 4390
+    },
+    {
+      "epoch": 2.1782301104629513,
+      "grad_norm": 0.12792142983023513,
+      "learning_rate": 2.1560896100929413e-06,
+      "loss": 0.672,
+      "step": 4391
+    },
+    {
+      "epoch": 2.178726573166191,
+      "grad_norm": 0.14201553304959288,
+      "learning_rate": 2.155121578568397e-06,
+      "loss": 0.7635,
+      "step": 4392
+    },
+    {
+      "epoch": 2.1792230358694304,
+      "grad_norm": 0.12863093955613852,
+      "learning_rate": 2.1541535997529894e-06,
+      "loss": 0.6829,
+      "step": 4393
+    },
+    {
+      "epoch": 2.17971949857267,
+      "grad_norm": 0.13051647735684252,
+      "learning_rate": 2.153185673794659e-06,
+      "loss": 0.6927,
+      "step": 4394
+    },
+    {
+      "epoch": 2.180215961275909,
+      "grad_norm": 0.1378899333922816,
+      "learning_rate": 2.1522178008413376e-06,
+      "loss": 0.7242,
+      "step": 4395
+    },
+    {
+      "epoch": 2.1807124239791484,
+      "grad_norm": 0.12915706282038747,
+      "learning_rate": 2.15124998104095e-06,
+      "loss": 0.7223,
+      "step": 4396
+    },
+    {
+      "epoch": 2.1812088866823878,
+      "grad_norm": 0.13756383241184975,
+      "learning_rate": 2.15028221454141e-06,
+      "loss": 0.7269,
+      "step": 4397
+    },
+    {
+      "epoch": 2.1817053493856275,
+      "grad_norm": 0.13566981965904096,
+      "learning_rate": 2.1493145014906264e-06,
+      "loss": 0.7178,
+      "step": 4398
+    },
+    {
+      "epoch": 2.182201812088867,
+      "grad_norm": 0.13120681413105775,
+      "learning_rate": 2.1483468420364984e-06,
+      "loss": 0.6638,
+      "step": 4399
+    },
+    {
+      "epoch": 2.182698274792106,
+      "grad_norm": 0.13010531035199235,
+      "learning_rate": 2.1473792363269183e-06,
+      "loss": 0.7361,
+      "step": 4400
+    },
+    {
+      "epoch": 2.1831947374953455,
+      "grad_norm": 0.13102941992333803,
+      "learning_rate": 2.1464116845097672e-06,
+      "loss": 0.6625,
+      "step": 4401
+    },
+    {
+      "epoch": 2.1836912001985853,
+      "grad_norm": 0.13143047623874493,
+      "learning_rate": 2.1454441867329205e-06,
+      "loss": 0.7456,
+      "step": 4402
+    },
+    {
+      "epoch": 2.1841876629018246,
+      "grad_norm": 0.12849792172954375,
+      "learning_rate": 2.1444767431442455e-06,
+      "loss": 0.6448,
+      "step": 4403
+    },
+    {
+      "epoch": 2.184684125605064,
+      "grad_norm": 0.1339908605256051,
+      "learning_rate": 2.1435093538916e-06,
+      "loss": 0.7366,
+      "step": 4404
+    },
+    {
+      "epoch": 2.1851805883083033,
+      "grad_norm": 0.1355852126115786,
+      "learning_rate": 2.1425420191228328e-06,
+      "loss": 0.7442,
+      "step": 4405
+    },
+    {
+      "epoch": 2.1856770510115426,
+      "grad_norm": 0.13537532518922524,
+      "learning_rate": 2.1415747389857875e-06,
+      "loss": 0.7584,
+      "step": 4406
+    },
+    {
+      "epoch": 2.186173513714782,
+      "grad_norm": 0.1376705768549547,
+      "learning_rate": 2.140607513628296e-06,
+      "loss": 0.6996,
+      "step": 4407
+    },
+    {
+      "epoch": 2.1866699764180217,
+      "grad_norm": 0.13019400686207325,
+      "learning_rate": 2.1396403431981843e-06,
+      "loss": 0.6652,
+      "step": 4408
+    },
+    {
+      "epoch": 2.187166439121261,
+      "grad_norm": 0.1322767474800185,
+      "learning_rate": 2.1386732278432674e-06,
+      "loss": 0.6934,
+      "step": 4409
+    },
+    {
+      "epoch": 2.1876629018245004,
+      "grad_norm": 0.13168416201103167,
+      "learning_rate": 2.1377061677113547e-06,
+      "loss": 0.6779,
+      "step": 4410
+    },
+    {
+      "epoch": 2.1881593645277397,
+      "grad_norm": 0.12841437119394158,
+      "learning_rate": 2.136739162950245e-06,
+      "loss": 0.6991,
+      "step": 4411
+    },
+    {
+      "epoch": 2.188655827230979,
+      "grad_norm": 0.13341291386169005,
+      "learning_rate": 2.13577221370773e-06,
+      "loss": 0.6849,
+      "step": 4412
+    },
+    {
+      "epoch": 2.189152289934219,
+      "grad_norm": 0.13556040065635502,
+      "learning_rate": 2.1348053201315926e-06,
+      "loss": 0.7101,
+      "step": 4413
+    },
+    {
+      "epoch": 2.189648752637458,
+      "grad_norm": 0.1347896034578669,
+      "learning_rate": 2.1338384823696056e-06,
+      "loss": 0.7173,
+      "step": 4414
+    },
+    {
+      "epoch": 2.1901452153406975,
+      "grad_norm": 0.13265640243012952,
+      "learning_rate": 2.1328717005695363e-06,
+      "loss": 0.6666,
+      "step": 4415
+    },
+    {
+      "epoch": 2.190641678043937,
+      "grad_norm": 0.1283246868922253,
+      "learning_rate": 2.1319049748791418e-06,
+      "loss": 0.7144,
+      "step": 4416
+    },
+    {
+      "epoch": 2.191138140747176,
+      "grad_norm": 0.13569707723504326,
+      "learning_rate": 2.1309383054461692e-06,
+      "loss": 0.7276,
+      "step": 4417
+    },
+    {
+      "epoch": 2.191634603450416,
+      "grad_norm": 0.1350932272158915,
+      "learning_rate": 2.1299716924183586e-06,
+      "loss": 0.7084,
+      "step": 4418
+    },
+    {
+      "epoch": 2.1921310661536553,
+      "grad_norm": 0.1293789781280897,
+      "learning_rate": 2.1290051359434426e-06,
+      "loss": 0.7082,
+      "step": 4419
+    },
+    {
+      "epoch": 2.1926275288568946,
+      "grad_norm": 0.1334911228325809,
+      "learning_rate": 2.128038636169143e-06,
+      "loss": 0.694,
+      "step": 4420
+    },
+    {
+      "epoch": 2.193123991560134,
+      "grad_norm": 0.13202766897134072,
+      "learning_rate": 2.1270721932431736e-06,
+      "loss": 0.7453,
+      "step": 4421
+    },
+    {
+      "epoch": 2.1936204542633733,
+      "grad_norm": 0.132585729157581,
+      "learning_rate": 2.1261058073132403e-06,
+      "loss": 0.7075,
+      "step": 4422
+    },
+    {
+      "epoch": 2.194116916966613,
+      "grad_norm": 0.13273752133950062,
+      "learning_rate": 2.1251394785270388e-06,
+      "loss": 0.7006,
+      "step": 4423
+    },
+    {
+      "epoch": 2.1946133796698524,
+      "grad_norm": 0.13508600941080354,
+      "learning_rate": 2.1241732070322586e-06,
+      "loss": 0.7252,
+      "step": 4424
+    },
+    {
+      "epoch": 2.1951098423730917,
+      "grad_norm": 0.12963329955825967,
+      "learning_rate": 2.123206992976577e-06,
+      "loss": 0.7375,
+      "step": 4425
+    },
+    {
+      "epoch": 2.195606305076331,
+      "grad_norm": 0.12468851759067956,
+      "learning_rate": 2.122240836507665e-06,
+      "loss": 0.6814,
+      "step": 4426
+    },
+    {
+      "epoch": 2.1961027677795704,
+      "grad_norm": 0.13146278221760682,
+      "learning_rate": 2.1212747377731845e-06,
+      "loss": 0.721,
+      "step": 4427
+    },
+    {
+      "epoch": 2.19659923048281,
+      "grad_norm": 0.14207414385746342,
+      "learning_rate": 2.120308696920789e-06,
+      "loss": 0.749,
+      "step": 4428
+    },
+    {
+      "epoch": 2.1970956931860495,
+      "grad_norm": 0.1299724476761059,
+      "learning_rate": 2.11934271409812e-06,
+      "loss": 0.6992,
+      "step": 4429
+    },
+    {
+      "epoch": 2.197592155889289,
+      "grad_norm": 0.13750736020293242,
+      "learning_rate": 2.1183767894528135e-06,
+      "loss": 0.7027,
+      "step": 4430
+    },
+    {
+      "epoch": 2.198088618592528,
+      "grad_norm": 0.1327599147789671,
+      "learning_rate": 2.1174109231324965e-06,
+      "loss": 0.661,
+      "step": 4431
+    },
+    {
+      "epoch": 2.1985850812957675,
+      "grad_norm": 0.13320281277199114,
+      "learning_rate": 2.1164451152847865e-06,
+      "loss": 0.7928,
+      "step": 4432
+    },
+    {
+      "epoch": 2.1990815439990072,
+      "grad_norm": 0.12945660859813715,
+      "learning_rate": 2.1154793660572897e-06,
+      "loss": 0.7051,
+      "step": 4433
+    },
+    {
+      "epoch": 2.1995780067022466,
+      "grad_norm": 0.1404344217727835,
+      "learning_rate": 2.1145136755976063e-06,
+      "loss": 0.7047,
+      "step": 4434
+    },
+    {
+      "epoch": 2.200074469405486,
+      "grad_norm": 0.13253286798252542,
+      "learning_rate": 2.1135480440533275e-06,
+      "loss": 0.687,
+      "step": 4435
+    },
+    {
+      "epoch": 2.2005709321087252,
+      "grad_norm": 0.1315732384032246,
+      "learning_rate": 2.1125824715720335e-06,
+      "loss": 0.7066,
+      "step": 4436
+    },
+    {
+      "epoch": 2.2010673948119646,
+      "grad_norm": 0.12983170486233375,
+      "learning_rate": 2.1116169583012965e-06,
+      "loss": 0.7092,
+      "step": 4437
+    },
+    {
+      "epoch": 2.2015638575152043,
+      "grad_norm": 0.12962669211533243,
+      "learning_rate": 2.1106515043886804e-06,
+      "loss": 0.7187,
+      "step": 4438
+    },
+    {
+      "epoch": 2.2020603202184437,
+      "grad_norm": 0.13258204247552738,
+      "learning_rate": 2.1096861099817394e-06,
+      "loss": 0.7111,
+      "step": 4439
+    },
+    {
+      "epoch": 2.202556782921683,
+      "grad_norm": 0.13733375443222062,
+      "learning_rate": 2.1087207752280186e-06,
+      "loss": 0.7446,
+      "step": 4440
+    },
+    {
+      "epoch": 2.2030532456249223,
+      "grad_norm": 0.1335139785372661,
+      "learning_rate": 2.107755500275053e-06,
+      "loss": 0.733,
+      "step": 4441
+    },
+    {
+      "epoch": 2.2035497083281617,
+      "grad_norm": 0.12985134152635452,
+      "learning_rate": 2.10679028527037e-06,
+      "loss": 0.7306,
+      "step": 4442
+    },
+    {
+      "epoch": 2.2040461710314014,
+      "grad_norm": 0.12811473834247125,
+      "learning_rate": 2.1058251303614875e-06,
+      "loss": 0.7068,
+      "step": 4443
+    },
+    {
+      "epoch": 2.2045426337346408,
+      "grad_norm": 0.12645032456670896,
+      "learning_rate": 2.1048600356959133e-06,
+      "loss": 0.6633,
+      "step": 4444
+    },
+    {
+      "epoch": 2.20503909643788,
+      "grad_norm": 0.13171507130184576,
+      "learning_rate": 2.103895001421148e-06,
+      "loss": 0.6973,
+      "step": 4445
+    },
+    {
+      "epoch": 2.2055355591411194,
+      "grad_norm": 0.1302019612259144,
+      "learning_rate": 2.10293002768468e-06,
+      "loss": 0.68,
+      "step": 4446
+    },
+    {
+      "epoch": 2.2060320218443588,
+      "grad_norm": 0.12865346316964513,
+      "learning_rate": 2.101965114633991e-06,
+      "loss": 0.7241,
+      "step": 4447
+    },
+    {
+      "epoch": 2.2065284845475985,
+      "grad_norm": 0.12938944539941274,
+      "learning_rate": 2.1010002624165528e-06,
+      "loss": 0.7068,
+      "step": 4448
+    },
+    {
+      "epoch": 2.207024947250838,
+      "grad_norm": 0.1315311127735872,
+      "learning_rate": 2.1000354711798258e-06,
+      "loss": 0.7197,
+      "step": 4449
+    },
+    {
+      "epoch": 2.207521409954077,
+      "grad_norm": 0.1304635101926576,
+      "learning_rate": 2.0990707410712647e-06,
+      "loss": 0.7177,
+      "step": 4450
+    },
+    {
+      "epoch": 2.2080178726573165,
+      "grad_norm": 0.12566986860796162,
+      "learning_rate": 2.098106072238313e-06,
+      "loss": 0.7147,
+      "step": 4451
+    },
+    {
+      "epoch": 2.208514335360556,
+      "grad_norm": 0.1291854407145579,
+      "learning_rate": 2.097141464828403e-06,
+      "loss": 0.6987,
+      "step": 4452
+    },
+    {
+      "epoch": 2.2090107980637956,
+      "grad_norm": 0.12932256073625428,
+      "learning_rate": 2.0961769189889612e-06,
+      "loss": 0.6682,
+      "step": 4453
+    },
+    {
+      "epoch": 2.209507260767035,
+      "grad_norm": 0.13534326258562565,
+      "learning_rate": 2.0952124348674027e-06,
+      "loss": 0.7585,
+      "step": 4454
+    },
+    {
+      "epoch": 2.2100037234702743,
+      "grad_norm": 0.13139570296537265,
+      "learning_rate": 2.094248012611133e-06,
+      "loss": 0.7238,
+      "step": 4455
+    },
+    {
+      "epoch": 2.2105001861735136,
+      "grad_norm": 0.12736205425045874,
+      "learning_rate": 2.0932836523675495e-06,
+      "loss": 0.7331,
+      "step": 4456
+    },
+    {
+      "epoch": 2.210996648876753,
+      "grad_norm": 0.1285738140080393,
+      "learning_rate": 2.0923193542840376e-06,
+      "loss": 0.6831,
+      "step": 4457
+    },
+    {
+      "epoch": 2.2114931115799927,
+      "grad_norm": 0.1373192184817093,
+      "learning_rate": 2.0913551185079763e-06,
+      "loss": 0.7511,
+      "step": 4458
+    },
+    {
+      "epoch": 2.211989574283232,
+      "grad_norm": 0.1315083921749297,
+      "learning_rate": 2.0903909451867327e-06,
+      "loss": 0.7186,
+      "step": 4459
+    },
+    {
+      "epoch": 2.2124860369864714,
+      "grad_norm": 0.13197214005113778,
+      "learning_rate": 2.089426834467666e-06,
+      "loss": 0.7278,
+      "step": 4460
+    },
+    {
+      "epoch": 2.2129824996897107,
+      "grad_norm": 0.13117095945392646,
+      "learning_rate": 2.0884627864981247e-06,
+      "loss": 0.7158,
+      "step": 4461
+    },
+    {
+      "epoch": 2.21347896239295,
+      "grad_norm": 0.14366920819680415,
+      "learning_rate": 2.0874988014254474e-06,
+      "loss": 0.697,
+      "step": 4462
+    },
+    {
+      "epoch": 2.21397542509619,
+      "grad_norm": 0.13103101512448223,
+      "learning_rate": 2.0865348793969644e-06,
+      "loss": 0.7099,
+      "step": 4463
+    },
+    {
+      "epoch": 2.214471887799429,
+      "grad_norm": 0.13241402045221123,
+      "learning_rate": 2.085571020559997e-06,
+      "loss": 0.6921,
+      "step": 4464
+    },
+    {
+      "epoch": 2.2149683505026685,
+      "grad_norm": 0.12588765448059275,
+      "learning_rate": 2.084607225061853e-06,
+      "loss": 0.6612,
+      "step": 4465
+    },
+    {
+      "epoch": 2.215464813205908,
+      "grad_norm": 0.1310167292666887,
+      "learning_rate": 2.0836434930498343e-06,
+      "loss": 0.6417,
+      "step": 4466
+    },
+    {
+      "epoch": 2.215961275909147,
+      "grad_norm": 0.13194158154041208,
+      "learning_rate": 2.082679824671232e-06,
+      "loss": 0.7063,
+      "step": 4467
+    },
+    {
+      "epoch": 2.216457738612387,
+      "grad_norm": 0.13066331982217858,
+      "learning_rate": 2.0817162200733275e-06,
+      "loss": 0.6948,
+      "step": 4468
+    },
+    {
+      "epoch": 2.2169542013156263,
+      "grad_norm": 0.1318788924378777,
+      "learning_rate": 2.080752679403392e-06,
+      "loss": 0.7012,
+      "step": 4469
+    },
+    {
+      "epoch": 2.2174506640188656,
+      "grad_norm": 0.13209070722048075,
+      "learning_rate": 2.0797892028086873e-06,
+      "loss": 0.7201,
+      "step": 4470
+    },
+    {
+      "epoch": 2.217947126722105,
+      "grad_norm": 0.1274600384445204,
+      "learning_rate": 2.078825790436465e-06,
+      "loss": 0.7131,
+      "step": 4471
+    },
+    {
+      "epoch": 2.2184435894253443,
+      "grad_norm": 0.13371229036895962,
+      "learning_rate": 2.0778624424339684e-06,
+      "loss": 0.7382,
+      "step": 4472
+    },
+    {
+      "epoch": 2.218940052128584,
+      "grad_norm": 0.13862248032289534,
+      "learning_rate": 2.0768991589484284e-06,
+      "loss": 0.7566,
+      "step": 4473
+    },
+    {
+      "epoch": 2.2194365148318234,
+      "grad_norm": 0.13523476510615928,
+      "learning_rate": 2.0759359401270683e-06,
+      "loss": 0.7046,
+      "step": 4474
+    },
+    {
+      "epoch": 2.2199329775350627,
+      "grad_norm": 0.13349052506342798,
+      "learning_rate": 2.0749727861171e-06,
+      "loss": 0.7179,
+      "step": 4475
+    },
+    {
+      "epoch": 2.220429440238302,
+      "grad_norm": 0.12706375105277193,
+      "learning_rate": 2.074009697065727e-06,
+      "loss": 0.7001,
+      "step": 4476
+    },
+    {
+      "epoch": 2.2209259029415414,
+      "grad_norm": 0.13175995019806763,
+      "learning_rate": 2.073046673120142e-06,
+      "loss": 0.7124,
+      "step": 4477
+    },
+    {
+      "epoch": 2.221422365644781,
+      "grad_norm": 0.12839478228159643,
+      "learning_rate": 2.0720837144275264e-06,
+      "loss": 0.6965,
+      "step": 4478
+    },
+    {
+      "epoch": 2.2219188283480205,
+      "grad_norm": 0.13013639445767175,
+      "learning_rate": 2.0711208211350543e-06,
+      "loss": 0.6632,
+      "step": 4479
+    },
+    {
+      "epoch": 2.22241529105126,
+      "grad_norm": 0.13013755734088792,
+      "learning_rate": 2.070157993389889e-06,
+      "loss": 0.6993,
+      "step": 4480
+    },
+    {
+      "epoch": 2.222911753754499,
+      "grad_norm": 0.12451994734244691,
+      "learning_rate": 2.069195231339182e-06,
+      "loss": 0.6995,
+      "step": 4481
+    },
+    {
+      "epoch": 2.2234082164577385,
+      "grad_norm": 0.13477779261267323,
+      "learning_rate": 2.0682325351300754e-06,
+      "loss": 0.6804,
+      "step": 4482
+    },
+    {
+      "epoch": 2.2239046791609782,
+      "grad_norm": 0.12668543781608235,
+      "learning_rate": 2.0672699049097034e-06,
+      "loss": 0.7061,
+      "step": 4483
+    },
+    {
+      "epoch": 2.2244011418642176,
+      "grad_norm": 0.13314313496622696,
+      "learning_rate": 2.0663073408251888e-06,
+      "loss": 0.7014,
+      "step": 4484
+    },
+    {
+      "epoch": 2.224897604567457,
+      "grad_norm": 0.12970701923177527,
+      "learning_rate": 2.065344843023643e-06,
+      "loss": 0.68,
+      "step": 4485
+    },
+    {
+      "epoch": 2.225394067270696,
+      "grad_norm": 0.12387800818585266,
+      "learning_rate": 2.0643824116521683e-06,
+      "loss": 0.6338,
+      "step": 4486
+    },
+    {
+      "epoch": 2.2258905299739355,
+      "grad_norm": 0.13553410886475062,
+      "learning_rate": 2.0634200468578577e-06,
+      "loss": 0.6881,
+      "step": 4487
+    },
+    {
+      "epoch": 2.2263869926771753,
+      "grad_norm": 0.1316309439657593,
+      "learning_rate": 2.0624577487877934e-06,
+      "loss": 0.7049,
+      "step": 4488
+    },
+    {
+      "epoch": 2.2268834553804147,
+      "grad_norm": 0.12922645687790163,
+      "learning_rate": 2.0614955175890464e-06,
+      "loss": 0.7002,
+      "step": 4489
+    },
+    {
+      "epoch": 2.227379918083654,
+      "grad_norm": 0.13080644734923877,
+      "learning_rate": 2.0605333534086783e-06,
+      "loss": 0.7158,
+      "step": 4490
+    },
+    {
+      "epoch": 2.2278763807868933,
+      "grad_norm": 0.13245492757085336,
+      "learning_rate": 2.0595712563937412e-06,
+      "loss": 0.749,
+      "step": 4491
+    },
+    {
+      "epoch": 2.2283728434901326,
+      "grad_norm": 0.13270571932773106,
+      "learning_rate": 2.0586092266912753e-06,
+      "loss": 0.7214,
+      "step": 4492
+    },
+    {
+      "epoch": 2.2288693061933724,
+      "grad_norm": 0.13643011762485555,
+      "learning_rate": 2.0576472644483133e-06,
+      "loss": 0.7247,
+      "step": 4493
+    },
+    {
+      "epoch": 2.2293657688966118,
+      "grad_norm": 0.13005640680659797,
+      "learning_rate": 2.056685369811873e-06,
+      "loss": 0.7014,
+      "step": 4494
+    },
+    {
+      "epoch": 2.229862231599851,
+      "grad_norm": 0.13129470666359272,
+      "learning_rate": 2.055723542928966e-06,
+      "loss": 0.7518,
+      "step": 4495
+    },
+    {
+      "epoch": 2.2303586943030904,
+      "grad_norm": 0.13311828367670825,
+      "learning_rate": 2.0547617839465925e-06,
+      "loss": 0.7229,
+      "step": 4496
+    },
+    {
+      "epoch": 2.2308551570063297,
+      "grad_norm": 0.12371038963722301,
+      "learning_rate": 2.0538000930117424e-06,
+      "loss": 0.6783,
+      "step": 4497
+    },
+    {
+      "epoch": 2.2313516197095695,
+      "grad_norm": 0.1292453879856836,
+      "learning_rate": 2.0528384702713924e-06,
+      "loss": 0.6828,
+      "step": 4498
+    },
+    {
+      "epoch": 2.231848082412809,
+      "grad_norm": 0.13392931788982132,
+      "learning_rate": 2.0518769158725126e-06,
+      "loss": 0.7158,
+      "step": 4499
+    },
+    {
+      "epoch": 2.232344545116048,
+      "grad_norm": 0.13236176803456814,
+      "learning_rate": 2.0509154299620622e-06,
+      "loss": 0.7292,
+      "step": 4500
+    },
+    {
+      "epoch": 2.2328410078192875,
+      "grad_norm": 0.12728431676565705,
+      "learning_rate": 2.0499540126869864e-06,
+      "loss": 0.709,
+      "step": 4501
+    },
+    {
+      "epoch": 2.233337470522527,
+      "grad_norm": 0.13394951954145282,
+      "learning_rate": 2.0489926641942245e-06,
+      "loss": 0.7129,
+      "step": 4502
+    },
+    {
+      "epoch": 2.2338339332257666,
+      "grad_norm": 0.12837119042814552,
+      "learning_rate": 2.0480313846307025e-06,
+      "loss": 0.6862,
+      "step": 4503
+    },
+    {
+      "epoch": 2.234330395929006,
+      "grad_norm": 0.13885619880171285,
+      "learning_rate": 2.047070174143337e-06,
+      "loss": 0.7047,
+      "step": 4504
+    },
+    {
+      "epoch": 2.2348268586322453,
+      "grad_norm": 0.13053786492237074,
+      "learning_rate": 2.0461090328790325e-06,
+      "loss": 0.741,
+      "step": 4505
+    },
+    {
+      "epoch": 2.2353233213354846,
+      "grad_norm": 0.12630369808853714,
+      "learning_rate": 2.0451479609846847e-06,
+      "loss": 0.6303,
+      "step": 4506
+    },
+    {
+      "epoch": 2.235819784038724,
+      "grad_norm": 0.1334775735012497,
+      "learning_rate": 2.0441869586071784e-06,
+      "loss": 0.7161,
+      "step": 4507
+    },
+    {
+      "epoch": 2.2363162467419633,
+      "grad_norm": 0.1243537069117441,
+      "learning_rate": 2.043226025893387e-06,
+      "loss": 0.7083,
+      "step": 4508
+    },
+    {
+      "epoch": 2.236812709445203,
+      "grad_norm": 0.1362055506738069,
+      "learning_rate": 2.0422651629901743e-06,
+      "loss": 0.7073,
+      "step": 4509
+    },
+    {
+      "epoch": 2.2373091721484424,
+      "grad_norm": 0.13381191349240082,
+      "learning_rate": 2.041304370044391e-06,
+      "loss": 0.715,
+      "step": 4510
+    },
+    {
+      "epoch": 2.2378056348516817,
+      "grad_norm": 0.13008180169114836,
+      "learning_rate": 2.0403436472028807e-06,
+      "loss": 0.6912,
+      "step": 4511
+    },
+    {
+      "epoch": 2.238302097554921,
+      "grad_norm": 0.13665438512065606,
+      "learning_rate": 2.0393829946124737e-06,
+      "loss": 0.7122,
+      "step": 4512
+    },
+    {
+      "epoch": 2.238798560258161,
+      "grad_norm": 0.1306034467966443,
+      "learning_rate": 2.0384224124199918e-06,
+      "loss": 0.6518,
+      "step": 4513
+    },
+    {
+      "epoch": 2.2392950229614,
+      "grad_norm": 0.14766441708462488,
+      "learning_rate": 2.0374619007722423e-06,
+      "loss": 0.75,
+      "step": 4514
+    },
+    {
+      "epoch": 2.2397914856646395,
+      "grad_norm": 0.14199023614242443,
+      "learning_rate": 2.036501459816025e-06,
+      "loss": 0.7057,
+      "step": 4515
+    },
+    {
+      "epoch": 2.240287948367879,
+      "grad_norm": 0.13183647650911845,
+      "learning_rate": 2.0355410896981285e-06,
+      "loss": 0.7167,
+      "step": 4516
+    },
+    {
+      "epoch": 2.240784411071118,
+      "grad_norm": 0.12935480831259677,
+      "learning_rate": 2.034580790565329e-06,
+      "loss": 0.6945,
+      "step": 4517
+    },
+    {
+      "epoch": 2.2412808737743575,
+      "grad_norm": 0.123741363188653,
+      "learning_rate": 2.033620562564393e-06,
+      "loss": 0.6593,
+      "step": 4518
+    },
+    {
+      "epoch": 2.2417773364775972,
+      "grad_norm": 0.1324876830108556,
+      "learning_rate": 2.032660405842076e-06,
+      "loss": 0.7795,
+      "step": 4519
+    },
+    {
+      "epoch": 2.2422737991808366,
+      "grad_norm": 0.13082736167618114,
+      "learning_rate": 2.031700320545123e-06,
+      "loss": 0.6666,
+      "step": 4520
+    },
+    {
+      "epoch": 2.242770261884076,
+      "grad_norm": 0.13454831662707822,
+      "learning_rate": 2.0307403068202677e-06,
+      "loss": 0.6773,
+      "step": 4521
+    },
+    {
+      "epoch": 2.2432667245873152,
+      "grad_norm": 0.13082635983605484,
+      "learning_rate": 2.0297803648142324e-06,
+      "loss": 0.6761,
+      "step": 4522
+    },
+    {
+      "epoch": 2.243763187290555,
+      "grad_norm": 0.1300274112305257,
+      "learning_rate": 2.0288204946737283e-06,
+      "loss": 0.6994,
+      "step": 4523
+    },
+    {
+      "epoch": 2.2442596499937943,
+      "grad_norm": 0.13137498359275115,
+      "learning_rate": 2.0278606965454573e-06,
+      "loss": 0.7014,
+      "step": 4524
+    },
+    {
+      "epoch": 2.2447561126970337,
+      "grad_norm": 0.1380783519062254,
+      "learning_rate": 2.026900970576109e-06,
+      "loss": 0.7264,
+      "step": 4525
+    },
+    {
+      "epoch": 2.245252575400273,
+      "grad_norm": 0.1334871254288439,
+      "learning_rate": 2.0259413169123615e-06,
+      "loss": 0.7035,
+      "step": 4526
+    },
+    {
+      "epoch": 2.2457490381035123,
+      "grad_norm": 0.13467702526258202,
+      "learning_rate": 2.0249817357008825e-06,
+      "loss": 0.7435,
+      "step": 4527
+    },
+    {
+      "epoch": 2.2462455008067517,
+      "grad_norm": 0.13343539696190834,
+      "learning_rate": 2.024022227088329e-06,
+      "loss": 0.766,
+      "step": 4528
+    },
+    {
+      "epoch": 2.2467419635099914,
+      "grad_norm": 0.1336433964166616,
+      "learning_rate": 2.0230627912213475e-06,
+      "loss": 0.7278,
+      "step": 4529
+    },
+    {
+      "epoch": 2.2472384262132308,
+      "grad_norm": 0.13069507120960716,
+      "learning_rate": 2.02210342824657e-06,
+      "loss": 0.6983,
+      "step": 4530
+    },
+    {
+      "epoch": 2.24773488891647,
+      "grad_norm": 0.13107224337831874,
+      "learning_rate": 2.0211441383106208e-06,
+      "loss": 0.677,
+      "step": 4531
+    },
+    {
+      "epoch": 2.2482313516197094,
+      "grad_norm": 0.13776105234734956,
+      "learning_rate": 2.020184921560113e-06,
+      "loss": 0.7045,
+      "step": 4532
+    },
+    {
+      "epoch": 2.248727814322949,
+      "grad_norm": 0.12525227813051776,
+      "learning_rate": 2.019225778141646e-06,
+      "loss": 0.6817,
+      "step": 4533
+    },
+    {
+      "epoch": 2.2492242770261885,
+      "grad_norm": 0.13625075248045249,
+      "learning_rate": 2.0182667082018104e-06,
+      "loss": 0.7373,
+      "step": 4534
+    },
+    {
+      "epoch": 2.249720739729428,
+      "grad_norm": 0.13175367979094055,
+      "learning_rate": 2.0173077118871847e-06,
+      "loss": 0.7571,
+      "step": 4535
+    },
+    {
+      "epoch": 2.250217202432667,
+      "grad_norm": 0.1313325217650813,
+      "learning_rate": 2.016348789344335e-06,
+      "loss": 0.6996,
+      "step": 4536
+    },
+    {
+      "epoch": 2.250217202432667,
+      "eval_loss": 0.7270435690879822,
+      "eval_runtime": 141.8878,
+      "eval_samples_per_second": 213.923,
+      "eval_steps_per_second": 26.746,
+      "step": 4536
+    },
+    {
+      "epoch": 2.2507136651359065,
+      "grad_norm": 0.130395751964868,
+      "learning_rate": 2.0153899407198193e-06,
+      "loss": 0.7365,
+      "step": 4537
+    },
+    {
+      "epoch": 2.251210127839146,
+      "grad_norm": 0.133483489650425,
+      "learning_rate": 2.01443116616018e-06,
+      "loss": 0.7063,
+      "step": 4538
+    },
+    {
+      "epoch": 2.2517065905423856,
+      "grad_norm": 0.12958543122285368,
+      "learning_rate": 2.0134724658119525e-06,
+      "loss": 0.7304,
+      "step": 4539
+    },
+    {
+      "epoch": 2.252203053245625,
+      "grad_norm": 0.13021997063309637,
+      "learning_rate": 2.012513839821657e-06,
+      "loss": 0.6835,
+      "step": 4540
+    },
+    {
+      "epoch": 2.2526995159488643,
+      "grad_norm": 0.1327109634903606,
+      "learning_rate": 2.011555288335805e-06,
+      "loss": 0.7004,
+      "step": 4541
+    },
+    {
+      "epoch": 2.2531959786521036,
+      "grad_norm": 0.13418960633883067,
+      "learning_rate": 2.0105968115008957e-06,
+      "loss": 0.6589,
+      "step": 4542
+    },
+    {
+      "epoch": 2.2536924413553434,
+      "grad_norm": 0.134302800194372,
+      "learning_rate": 2.0096384094634165e-06,
+      "loss": 0.7305,
+      "step": 4543
+    },
+    {
+      "epoch": 2.2541889040585827,
+      "grad_norm": 0.1313733378397713,
+      "learning_rate": 2.0086800823698437e-06,
+      "loss": 0.7205,
+      "step": 4544
+    },
+    {
+      "epoch": 2.254685366761822,
+      "grad_norm": 0.13297045310339728,
+      "learning_rate": 2.007721830366644e-06,
+      "loss": 0.7046,
+      "step": 4545
+    },
+    {
+      "epoch": 2.2551818294650614,
+      "grad_norm": 0.1336572863632231,
+      "learning_rate": 2.0067636536002687e-06,
+      "loss": 0.6846,
+      "step": 4546
+    },
+    {
+      "epoch": 2.2556782921683007,
+      "grad_norm": 0.13982170058768661,
+      "learning_rate": 2.00580555221716e-06,
+      "loss": 0.7349,
+      "step": 4547
+    },
+    {
+      "epoch": 2.25617475487154,
+      "grad_norm": 0.13758235182829331,
+      "learning_rate": 2.0048475263637495e-06,
+      "loss": 0.7076,
+      "step": 4548
+    },
+    {
+      "epoch": 2.25667121757478,
+      "grad_norm": 0.13023416929452136,
+      "learning_rate": 2.003889576186455e-06,
+      "loss": 0.7703,
+      "step": 4549
+    },
+    {
+      "epoch": 2.257167680278019,
+      "grad_norm": 0.12888207678478222,
+      "learning_rate": 2.002931701831684e-06,
+      "loss": 0.7119,
+      "step": 4550
+    },
+    {
+      "epoch": 2.2576641429812585,
+      "grad_norm": 0.12816010732415842,
+      "learning_rate": 2.0019739034458328e-06,
+      "loss": 0.725,
+      "step": 4551
+    },
+    {
+      "epoch": 2.258160605684498,
+      "grad_norm": 0.14248392567472656,
+      "learning_rate": 2.0010161811752856e-06,
+      "loss": 0.7778,
+      "step": 4552
+    },
+    {
+      "epoch": 2.2586570683877376,
+      "grad_norm": 0.13798620466349681,
+      "learning_rate": 2.000058535166414e-06,
+      "loss": 0.6932,
+      "step": 4553
+    },
+    {
+      "epoch": 2.259153531090977,
+      "grad_norm": 0.1329997986135555,
+      "learning_rate": 1.9991009655655796e-06,
+      "loss": 0.6758,
+      "step": 4554
+    },
+    {
+      "epoch": 2.2596499937942163,
+      "grad_norm": 0.13463323485659834,
+      "learning_rate": 1.9981434725191314e-06,
+      "loss": 0.7121,
+      "step": 4555
+    },
+    {
+      "epoch": 2.2601464564974556,
+      "grad_norm": 0.12854539155868666,
+      "learning_rate": 1.9971860561734062e-06,
+      "loss": 0.6769,
+      "step": 4556
+    },
+    {
+      "epoch": 2.260642919200695,
+      "grad_norm": 0.1282085935483695,
+      "learning_rate": 1.9962287166747304e-06,
+      "loss": 0.6794,
+      "step": 4557
+    },
+    {
+      "epoch": 2.2611393819039343,
+      "grad_norm": 0.1294147737679884,
+      "learning_rate": 1.9952714541694186e-06,
+      "loss": 0.7308,
+      "step": 4558
+    },
+    {
+      "epoch": 2.261635844607174,
+      "grad_norm": 0.13473387061029163,
+      "learning_rate": 1.994314268803772e-06,
+      "loss": 0.7255,
+      "step": 4559
+    },
+    {
+      "epoch": 2.2621323073104134,
+      "grad_norm": 0.1325041645843652,
+      "learning_rate": 1.993357160724081e-06,
+      "loss": 0.697,
+      "step": 4560
+    },
+    {
+      "epoch": 2.2626287700136527,
+      "grad_norm": 0.1342544420441968,
+      "learning_rate": 1.9924001300766256e-06,
+      "loss": 0.7283,
+      "step": 4561
+    },
+    {
+      "epoch": 2.263125232716892,
+      "grad_norm": 0.13379243253594394,
+      "learning_rate": 1.9914431770076707e-06,
+      "loss": 0.7161,
+      "step": 4562
+    },
+    {
+      "epoch": 2.263621695420132,
+      "grad_norm": 0.1319578992058116,
+      "learning_rate": 1.9904863016634724e-06,
+      "loss": 0.7133,
+      "step": 4563
+    },
+    {
+      "epoch": 2.264118158123371,
+      "grad_norm": 0.13825130015364384,
+      "learning_rate": 1.9895295041902733e-06,
+      "loss": 0.7064,
+      "step": 4564
+    },
+    {
+      "epoch": 2.2646146208266105,
+      "grad_norm": 0.1261519713290881,
+      "learning_rate": 1.988572784734305e-06,
+      "loss": 0.7139,
+      "step": 4565
+    },
+    {
+      "epoch": 2.26511108352985,
+      "grad_norm": 0.13132799813077256,
+      "learning_rate": 1.9876161434417857e-06,
+      "loss": 0.6944,
+      "step": 4566
+    },
+    {
+      "epoch": 2.265607546233089,
+      "grad_norm": 0.13312127226666143,
+      "learning_rate": 1.986659580458924e-06,
+      "loss": 0.7534,
+      "step": 4567
+    },
+    {
+      "epoch": 2.2661040089363285,
+      "grad_norm": 0.13300727555740405,
+      "learning_rate": 1.9857030959319143e-06,
+      "loss": 0.6843,
+      "step": 4568
+    },
+    {
+      "epoch": 2.2666004716395682,
+      "grad_norm": 0.12803205729134937,
+      "learning_rate": 1.984746690006941e-06,
+      "loss": 0.6661,
+      "step": 4569
+    },
+    {
+      "epoch": 2.2670969343428076,
+      "grad_norm": 0.13134108543944287,
+      "learning_rate": 1.983790362830174e-06,
+      "loss": 0.715,
+      "step": 4570
+    },
+    {
+      "epoch": 2.267593397046047,
+      "grad_norm": 0.1310214195071098,
+      "learning_rate": 1.982834114547773e-06,
+      "loss": 0.6808,
+      "step": 4571
+    },
+    {
+      "epoch": 2.2680898597492862,
+      "grad_norm": 0.12868216227851317,
+      "learning_rate": 1.981877945305886e-06,
+      "loss": 0.6859,
+      "step": 4572
+    },
+    {
+      "epoch": 2.2685863224525256,
+      "grad_norm": 0.1308307615447908,
+      "learning_rate": 1.980921855250647e-06,
+      "loss": 0.706,
+      "step": 4573
+    },
+    {
+      "epoch": 2.2690827851557653,
+      "grad_norm": 0.13070485382531827,
+      "learning_rate": 1.979965844528181e-06,
+      "loss": 0.7233,
+      "step": 4574
+    },
+    {
+      "epoch": 2.2695792478590047,
+      "grad_norm": 0.13488405502043913,
+      "learning_rate": 1.979009913284596e-06,
+      "loss": 0.6707,
+      "step": 4575
+    },
+    {
+      "epoch": 2.270075710562244,
+      "grad_norm": 0.1357162099459169,
+      "learning_rate": 1.978054061665993e-06,
+      "loss": 0.7506,
+      "step": 4576
+    },
+    {
+      "epoch": 2.2705721732654833,
+      "grad_norm": 0.1324174585525689,
+      "learning_rate": 1.977098289818459e-06,
+      "loss": 0.7365,
+      "step": 4577
+    },
+    {
+      "epoch": 2.2710686359687227,
+      "grad_norm": 0.13116743586662247,
+      "learning_rate": 1.976142597888066e-06,
+      "loss": 0.6689,
+      "step": 4578
+    },
+    {
+      "epoch": 2.2715650986719624,
+      "grad_norm": 0.13835608552718603,
+      "learning_rate": 1.9751869860208774e-06,
+      "loss": 0.706,
+      "step": 4579
+    },
+    {
+      "epoch": 2.2720615613752018,
+      "grad_norm": 0.13412470397613815,
+      "learning_rate": 1.974231454362944e-06,
+      "loss": 0.7028,
+      "step": 4580
+    },
+    {
+      "epoch": 2.272558024078441,
+      "grad_norm": 0.12973612445352137,
+      "learning_rate": 1.973276003060302e-06,
+      "loss": 0.7142,
+      "step": 4581
+    },
+    {
+      "epoch": 2.2730544867816804,
+      "grad_norm": 0.12782910807584713,
+      "learning_rate": 1.972320632258978e-06,
+      "loss": 0.6797,
+      "step": 4582
+    },
+    {
+      "epoch": 2.2735509494849198,
+      "grad_norm": 0.12805462086856045,
+      "learning_rate": 1.9713653421049844e-06,
+      "loss": 0.6942,
+      "step": 4583
+    },
+    {
+      "epoch": 2.2740474121881595,
+      "grad_norm": 0.13882829796209556,
+      "learning_rate": 1.970410132744322e-06,
+      "loss": 0.6987,
+      "step": 4584
+    },
+    {
+      "epoch": 2.274543874891399,
+      "grad_norm": 0.13480442463376907,
+      "learning_rate": 1.9694550043229802e-06,
+      "loss": 0.7136,
+      "step": 4585
+    },
+    {
+      "epoch": 2.275040337594638,
+      "grad_norm": 0.12813305978151937,
+      "learning_rate": 1.968499956986934e-06,
+      "loss": 0.6925,
+      "step": 4586
+    },
+    {
+      "epoch": 2.2755368002978775,
+      "grad_norm": 0.13565940409117114,
+      "learning_rate": 1.9675449908821473e-06,
+      "loss": 0.7073,
+      "step": 4587
+    },
+    {
+      "epoch": 2.276033263001117,
+      "grad_norm": 0.13297564484436258,
+      "learning_rate": 1.9665901061545715e-06,
+      "loss": 0.7146,
+      "step": 4588
+    },
+    {
+      "epoch": 2.2765297257043566,
+      "grad_norm": 0.1292833330312642,
+      "learning_rate": 1.965635302950145e-06,
+      "loss": 0.6681,
+      "step": 4589
+    },
+    {
+      "epoch": 2.277026188407596,
+      "grad_norm": 0.13161230553838985,
+      "learning_rate": 1.9646805814147956e-06,
+      "loss": 0.7131,
+      "step": 4590
+    },
+    {
+      "epoch": 2.2775226511108353,
+      "grad_norm": 0.13259185019815198,
+      "learning_rate": 1.9637259416944352e-06,
+      "loss": 0.7156,
+      "step": 4591
+    },
+    {
+      "epoch": 2.2780191138140746,
+      "grad_norm": 0.13109212966194678,
+      "learning_rate": 1.9627713839349665e-06,
+      "loss": 0.7343,
+      "step": 4592
+    },
+    {
+      "epoch": 2.278515576517314,
+      "grad_norm": 0.13887226544683762,
+      "learning_rate": 1.961816908282279e-06,
+      "loss": 0.7579,
+      "step": 4593
+    },
+    {
+      "epoch": 2.2790120392205537,
+      "grad_norm": 0.13379712126091625,
+      "learning_rate": 1.960862514882247e-06,
+      "loss": 0.741,
+      "step": 4594
+    },
+    {
+      "epoch": 2.279508501923793,
+      "grad_norm": 0.12731411318156655,
+      "learning_rate": 1.959908203880735e-06,
+      "loss": 0.6717,
+      "step": 4595
+    },
+    {
+      "epoch": 2.2800049646270324,
+      "grad_norm": 0.13191309058764128,
+      "learning_rate": 1.9589539754235938e-06,
+      "loss": 0.6887,
+      "step": 4596
+    },
+    {
+      "epoch": 2.2805014273302717,
+      "grad_norm": 0.13477829801291258,
+      "learning_rate": 1.957999829656664e-06,
+      "loss": 0.7024,
+      "step": 4597
+    },
+    {
+      "epoch": 2.280997890033511,
+      "grad_norm": 0.1408048819606926,
+      "learning_rate": 1.9570457667257686e-06,
+      "loss": 0.736,
+      "step": 4598
+    },
+    {
+      "epoch": 2.281494352736751,
+      "grad_norm": 0.13409166712691484,
+      "learning_rate": 1.9560917867767223e-06,
+      "loss": 0.7182,
+      "step": 4599
+    },
+    {
+      "epoch": 2.28199081543999,
+      "grad_norm": 0.1323033246480032,
+      "learning_rate": 1.9551378899553255e-06,
+      "loss": 0.6757,
+      "step": 4600
+    },
+    {
+      "epoch": 2.2824872781432295,
+      "grad_norm": 0.1279575053978906,
+      "learning_rate": 1.9541840764073666e-06,
+      "loss": 0.6639,
+      "step": 4601
+    },
+    {
+      "epoch": 2.282983740846469,
+      "grad_norm": 0.1284865860394091,
+      "learning_rate": 1.9532303462786196e-06,
+      "loss": 0.7076,
+      "step": 4602
+    },
+    {
+      "epoch": 2.283480203549708,
+      "grad_norm": 0.13236730912378264,
+      "learning_rate": 1.9522766997148473e-06,
+      "loss": 0.7177,
+      "step": 4603
+    },
+    {
+      "epoch": 2.283976666252948,
+      "grad_norm": 0.13334477268216768,
+      "learning_rate": 1.9513231368617996e-06,
+      "loss": 0.7152,
+      "step": 4604
+    },
+    {
+      "epoch": 2.2844731289561873,
+      "grad_norm": 0.129554016821733,
+      "learning_rate": 1.950369657865213e-06,
+      "loss": 0.7274,
+      "step": 4605
+    },
+    {
+      "epoch": 2.2849695916594266,
+      "grad_norm": 0.12809936242000544,
+      "learning_rate": 1.949416262870812e-06,
+      "loss": 0.6868,
+      "step": 4606
+    },
+    {
+      "epoch": 2.285466054362666,
+      "grad_norm": 0.12619361467512472,
+      "learning_rate": 1.948462952024307e-06,
+      "loss": 0.6536,
+      "step": 4607
+    },
+    {
+      "epoch": 2.2859625170659053,
+      "grad_norm": 0.13019492596083582,
+      "learning_rate": 1.9475097254713963e-06,
+      "loss": 0.6921,
+      "step": 4608
+    },
+    {
+      "epoch": 2.286458979769145,
+      "grad_norm": 0.12712933900402623,
+      "learning_rate": 1.9465565833577667e-06,
+      "loss": 0.7071,
+      "step": 4609
+    },
+    {
+      "epoch": 2.2869554424723844,
+      "grad_norm": 0.13073678098550098,
+      "learning_rate": 1.9456035258290886e-06,
+      "loss": 0.7274,
+      "step": 4610
+    },
+    {
+      "epoch": 2.2874519051756237,
+      "grad_norm": 0.13476232848198935,
+      "learning_rate": 1.944650553031022e-06,
+      "loss": 0.7651,
+      "step": 4611
+    },
+    {
+      "epoch": 2.287948367878863,
+      "grad_norm": 0.12940407399602424,
+      "learning_rate": 1.9436976651092143e-06,
+      "loss": 0.7052,
+      "step": 4612
+    },
+    {
+      "epoch": 2.2884448305821024,
+      "grad_norm": 0.12521819355840508,
+      "learning_rate": 1.9427448622092997e-06,
+      "loss": 0.6643,
+      "step": 4613
+    },
+    {
+      "epoch": 2.288941293285342,
+      "grad_norm": 0.1364514191356369,
+      "learning_rate": 1.941792144476897e-06,
+      "loss": 0.6871,
+      "step": 4614
+    },
+    {
+      "epoch": 2.2894377559885815,
+      "grad_norm": 0.12501502708363635,
+      "learning_rate": 1.940839512057615e-06,
+      "loss": 0.6713,
+      "step": 4615
+    },
+    {
+      "epoch": 2.289934218691821,
+      "grad_norm": 0.13288530127304785,
+      "learning_rate": 1.9398869650970483e-06,
+      "loss": 0.7355,
+      "step": 4616
+    },
+    {
+      "epoch": 2.29043068139506,
+      "grad_norm": 0.1356585375859948,
+      "learning_rate": 1.9389345037407787e-06,
+      "loss": 0.7225,
+      "step": 4617
+    },
+    {
+      "epoch": 2.2909271440982995,
+      "grad_norm": 0.13053848389346934,
+      "learning_rate": 1.9379821281343736e-06,
+      "loss": 0.7053,
+      "step": 4618
+    },
+    {
+      "epoch": 2.2914236068015392,
+      "grad_norm": 0.13397921868792434,
+      "learning_rate": 1.937029838423389e-06,
+      "loss": 0.7573,
+      "step": 4619
+    },
+    {
+      "epoch": 2.2919200695047786,
+      "grad_norm": 0.13057634121314804,
+      "learning_rate": 1.9360776347533667e-06,
+      "loss": 0.7348,
+      "step": 4620
+    },
+    {
+      "epoch": 2.292416532208018,
+      "grad_norm": 0.12556161832630441,
+      "learning_rate": 1.9351255172698368e-06,
+      "loss": 0.6724,
+      "step": 4621
+    },
+    {
+      "epoch": 2.292912994911257,
+      "grad_norm": 0.13237700928498475,
+      "learning_rate": 1.9341734861183146e-06,
+      "loss": 0.7015,
+      "step": 4622
+    },
+    {
+      "epoch": 2.2934094576144965,
+      "grad_norm": 0.12996320795796523,
+      "learning_rate": 1.9332215414443023e-06,
+      "loss": 0.6838,
+      "step": 4623
+    },
+    {
+      "epoch": 2.2939059203177363,
+      "grad_norm": 0.1264053811415136,
+      "learning_rate": 1.9322696833932896e-06,
+      "loss": 0.672,
+      "step": 4624
+    },
+    {
+      "epoch": 2.2944023830209757,
+      "grad_norm": 0.13696458123198366,
+      "learning_rate": 1.931317912110754e-06,
+      "loss": 0.7194,
+      "step": 4625
+    },
+    {
+      "epoch": 2.294898845724215,
+      "grad_norm": 0.12935669893699142,
+      "learning_rate": 1.930366227742157e-06,
+      "loss": 0.7196,
+      "step": 4626
+    },
+    {
+      "epoch": 2.2953953084274543,
+      "grad_norm": 0.13011349995194715,
+      "learning_rate": 1.9294146304329482e-06,
+      "loss": 0.7146,
+      "step": 4627
+    },
+    {
+      "epoch": 2.2958917711306936,
+      "grad_norm": 0.132257641760658,
+      "learning_rate": 1.9284631203285644e-06,
+      "loss": 0.7539,
+      "step": 4628
+    },
+    {
+      "epoch": 2.296388233833933,
+      "grad_norm": 0.12618209390437157,
+      "learning_rate": 1.927511697574429e-06,
+      "loss": 0.6542,
+      "step": 4629
+    },
+    {
+      "epoch": 2.2968846965371728,
+      "grad_norm": 0.13076964633199217,
+      "learning_rate": 1.9265603623159517e-06,
+      "loss": 0.736,
+      "step": 4630
+    },
+    {
+      "epoch": 2.297381159240412,
+      "grad_norm": 0.1278597050601475,
+      "learning_rate": 1.9256091146985282e-06,
+      "loss": 0.7204,
+      "step": 4631
+    },
+    {
+      "epoch": 2.2978776219436514,
+      "grad_norm": 0.12737458954653222,
+      "learning_rate": 1.924657954867542e-06,
+      "loss": 0.6605,
+      "step": 4632
+    },
+    {
+      "epoch": 2.2983740846468907,
+      "grad_norm": 0.12824874951751916,
+      "learning_rate": 1.923706882968362e-06,
+      "loss": 0.6611,
+      "step": 4633
+    },
+    {
+      "epoch": 2.2988705473501305,
+      "grad_norm": 0.13788032532644687,
+      "learning_rate": 1.922755899146346e-06,
+      "loss": 0.7261,
+      "step": 4634
+    },
+    {
+      "epoch": 2.29936701005337,
+      "grad_norm": 0.13925532124868073,
+      "learning_rate": 1.9218050035468343e-06,
+      "loss": 0.6746,
+      "step": 4635
+    },
+    {
+      "epoch": 2.299863472756609,
+      "grad_norm": 0.13354164221145085,
+      "learning_rate": 1.9208541963151576e-06,
+      "loss": 0.6896,
+      "step": 4636
+    },
+    {
+      "epoch": 2.3003599354598485,
+      "grad_norm": 0.13161140898185,
+      "learning_rate": 1.919903477596631e-06,
+      "loss": 0.6969,
+      "step": 4637
+    },
+    {
+      "epoch": 2.300856398163088,
+      "grad_norm": 0.13142493984817077,
+      "learning_rate": 1.918952847536557e-06,
+      "loss": 0.6899,
+      "step": 4638
+    },
+    {
+      "epoch": 2.301352860866327,
+      "grad_norm": 0.12827685841442193,
+      "learning_rate": 1.9180023062802237e-06,
+      "loss": 0.6423,
+      "step": 4639
+    },
+    {
+      "epoch": 2.301849323569567,
+      "grad_norm": 0.13137125453834245,
+      "learning_rate": 1.9170518539729063e-06,
+      "loss": 0.6769,
+      "step": 4640
+    },
+    {
+      "epoch": 2.3023457862728063,
+      "grad_norm": 0.1256889202189166,
+      "learning_rate": 1.9161014907598668e-06,
+      "loss": 0.6618,
+      "step": 4641
+    },
+    {
+      "epoch": 2.3028422489760456,
+      "grad_norm": 0.12999406754723933,
+      "learning_rate": 1.915151216786352e-06,
+      "loss": 0.7263,
+      "step": 4642
+    },
+    {
+      "epoch": 2.303338711679285,
+      "grad_norm": 0.12802875535724298,
+      "learning_rate": 1.9142010321975956e-06,
+      "loss": 0.6686,
+      "step": 4643
+    },
+    {
+      "epoch": 2.3038351743825247,
+      "grad_norm": 0.13208418783479883,
+      "learning_rate": 1.9132509371388187e-06,
+      "loss": 0.7076,
+      "step": 4644
+    },
+    {
+      "epoch": 2.304331637085764,
+      "grad_norm": 0.13301405076047154,
+      "learning_rate": 1.9123009317552294e-06,
+      "loss": 0.7435,
+      "step": 4645
+    },
+    {
+      "epoch": 2.3048280997890034,
+      "grad_norm": 0.13096465156829554,
+      "learning_rate": 1.911351016192019e-06,
+      "loss": 0.6746,
+      "step": 4646
+    },
+    {
+      "epoch": 2.3053245624922427,
+      "grad_norm": 0.1282318000984796,
+      "learning_rate": 1.9104011905943675e-06,
+      "loss": 0.6925,
+      "step": 4647
+    },
+    {
+      "epoch": 2.305821025195482,
+      "grad_norm": 0.1374326520419337,
+      "learning_rate": 1.9094514551074405e-06,
+      "loss": 0.721,
+      "step": 4648
+    },
+    {
+      "epoch": 2.3063174878987214,
+      "grad_norm": 0.13296090301658972,
+      "learning_rate": 1.9085018098763895e-06,
+      "loss": 0.7127,
+      "step": 4649
+    },
+    {
+      "epoch": 2.306813950601961,
+      "grad_norm": 0.12863266739621676,
+      "learning_rate": 1.9075522550463538e-06,
+      "loss": 0.6811,
+      "step": 4650
+    },
+    {
+      "epoch": 2.3073104133052005,
+      "grad_norm": 0.1292808442724112,
+      "learning_rate": 1.9066027907624563e-06,
+      "loss": 0.6921,
+      "step": 4651
+    },
+    {
+      "epoch": 2.30780687600844,
+      "grad_norm": 0.13198793172057285,
+      "learning_rate": 1.9056534171698076e-06,
+      "loss": 0.6872,
+      "step": 4652
+    },
+    {
+      "epoch": 2.308303338711679,
+      "grad_norm": 0.1308224331561388,
+      "learning_rate": 1.9047041344135045e-06,
+      "loss": 0.6865,
+      "step": 4653
+    },
+    {
+      "epoch": 2.308799801414919,
+      "grad_norm": 0.130383003378401,
+      "learning_rate": 1.9037549426386304e-06,
+      "loss": 0.7286,
+      "step": 4654
+    },
+    {
+      "epoch": 2.3092962641181582,
+      "grad_norm": 0.13067038581766985,
+      "learning_rate": 1.9028058419902524e-06,
+      "loss": 0.6786,
+      "step": 4655
+    },
+    {
+      "epoch": 2.3097927268213976,
+      "grad_norm": 0.1312758611395035,
+      "learning_rate": 1.9018568326134262e-06,
+      "loss": 0.6942,
+      "step": 4656
+    },
+    {
+      "epoch": 2.310289189524637,
+      "grad_norm": 0.13012405788871972,
+      "learning_rate": 1.900907914653194e-06,
+      "loss": 0.6997,
+      "step": 4657
+    },
+    {
+      "epoch": 2.3107856522278762,
+      "grad_norm": 0.13238886769700214,
+      "learning_rate": 1.8999590882545798e-06,
+      "loss": 0.7042,
+      "step": 4658
+    },
+    {
+      "epoch": 2.3112821149311156,
+      "grad_norm": 0.13003113363469201,
+      "learning_rate": 1.8990103535625983e-06,
+      "loss": 0.7026,
+      "step": 4659
+    },
+    {
+      "epoch": 2.3117785776343553,
+      "grad_norm": 0.12888802864649773,
+      "learning_rate": 1.8980617107222482e-06,
+      "loss": 0.6841,
+      "step": 4660
+    },
+    {
+      "epoch": 2.3122750403375947,
+      "grad_norm": 0.13431432633019263,
+      "learning_rate": 1.8971131598785148e-06,
+      "loss": 0.7537,
+      "step": 4661
+    },
+    {
+      "epoch": 2.312771503040834,
+      "grad_norm": 0.12849233342145183,
+      "learning_rate": 1.8961647011763676e-06,
+      "loss": 0.6768,
+      "step": 4662
+    },
+    {
+      "epoch": 2.3132679657440733,
+      "grad_norm": 0.13345147046633807,
+      "learning_rate": 1.8952163347607642e-06,
+      "loss": 0.7059,
+      "step": 4663
+    },
+    {
+      "epoch": 2.313764428447313,
+      "grad_norm": 0.13053349945942708,
+      "learning_rate": 1.894268060776647e-06,
+      "loss": 0.7291,
+      "step": 4664
+    },
+    {
+      "epoch": 2.3142608911505524,
+      "grad_norm": 0.1343964907723944,
+      "learning_rate": 1.8933198793689444e-06,
+      "loss": 0.7698,
+      "step": 4665
+    },
+    {
+      "epoch": 2.3147573538537918,
+      "grad_norm": 0.14226824478362488,
+      "learning_rate": 1.8923717906825718e-06,
+      "loss": 0.7851,
+      "step": 4666
+    },
+    {
+      "epoch": 2.315253816557031,
+      "grad_norm": 0.12925409867242452,
+      "learning_rate": 1.8914237948624275e-06,
+      "loss": 0.655,
+      "step": 4667
+    },
+    {
+      "epoch": 2.3157502792602704,
+      "grad_norm": 0.12816123194872267,
+      "learning_rate": 1.8904758920533988e-06,
+      "loss": 0.6711,
+      "step": 4668
+    },
+    {
+      "epoch": 2.3162467419635098,
+      "grad_norm": 0.1294204482164869,
+      "learning_rate": 1.8895280824003568e-06,
+      "loss": 0.6621,
+      "step": 4669
+    },
+    {
+      "epoch": 2.3167432046667495,
+      "grad_norm": 0.12776210547198066,
+      "learning_rate": 1.888580366048159e-06,
+      "loss": 0.7061,
+      "step": 4670
+    },
+    {
+      "epoch": 2.317239667369989,
+      "grad_norm": 0.12704629242381982,
+      "learning_rate": 1.8876327431416498e-06,
+      "loss": 0.7007,
+      "step": 4671
+    },
+    {
+      "epoch": 2.317736130073228,
+      "grad_norm": 0.1316157139453224,
+      "learning_rate": 1.8866852138256565e-06,
+      "loss": 0.7049,
+      "step": 4672
+    },
+    {
+      "epoch": 2.3182325927764675,
+      "grad_norm": 0.13375751981024753,
+      "learning_rate": 1.8857377782449956e-06,
+      "loss": 0.7209,
+      "step": 4673
+    },
+    {
+      "epoch": 2.3187290554797073,
+      "grad_norm": 0.13344576229065264,
+      "learning_rate": 1.8847904365444653e-06,
+      "loss": 0.7304,
+      "step": 4674
+    },
+    {
+      "epoch": 2.3192255181829466,
+      "grad_norm": 0.13003917367562956,
+      "learning_rate": 1.8838431888688528e-06,
+      "loss": 0.6972,
+      "step": 4675
+    },
+    {
+      "epoch": 2.319721980886186,
+      "grad_norm": 0.14062745708919816,
+      "learning_rate": 1.8828960353629294e-06,
+      "loss": 0.738,
+      "step": 4676
+    },
+    {
+      "epoch": 2.3202184435894253,
+      "grad_norm": 0.12901976125458806,
+      "learning_rate": 1.881948976171453e-06,
+      "loss": 0.6849,
+      "step": 4677
+    },
+    {
+      "epoch": 2.3207149062926646,
+      "grad_norm": 0.1277269250608724,
+      "learning_rate": 1.8810020114391653e-06,
+      "loss": 0.6756,
+      "step": 4678
+    },
+    {
+      "epoch": 2.321211368995904,
+      "grad_norm": 0.1291372202878994,
+      "learning_rate": 1.8800551413107955e-06,
+      "loss": 0.6776,
+      "step": 4679
+    },
+    {
+      "epoch": 2.3217078316991437,
+      "grad_norm": 0.12757692002915727,
+      "learning_rate": 1.8791083659310568e-06,
+      "loss": 0.6778,
+      "step": 4680
+    },
+    {
+      "epoch": 2.322204294402383,
+      "grad_norm": 0.1396567920307466,
+      "learning_rate": 1.8781616854446496e-06,
+      "loss": 0.7325,
+      "step": 4681
+    },
+    {
+      "epoch": 2.3227007571056224,
+      "grad_norm": 0.13455039781184047,
+      "learning_rate": 1.8772150999962588e-06,
+      "loss": 0.7038,
+      "step": 4682
+    },
+    {
+      "epoch": 2.3231972198088617,
+      "grad_norm": 0.13252407557746912,
+      "learning_rate": 1.8762686097305537e-06,
+      "loss": 0.6989,
+      "step": 4683
+    },
+    {
+      "epoch": 2.3236936825121015,
+      "grad_norm": 0.12477814641502248,
+      "learning_rate": 1.875322214792191e-06,
+      "loss": 0.6746,
+      "step": 4684
+    },
+    {
+      "epoch": 2.324190145215341,
+      "grad_norm": 0.1295558006823391,
+      "learning_rate": 1.8743759153258118e-06,
+      "loss": 0.7017,
+      "step": 4685
+    },
+    {
+      "epoch": 2.32468660791858,
+      "grad_norm": 0.13196900667132647,
+      "learning_rate": 1.8734297114760427e-06,
+      "loss": 0.7403,
+      "step": 4686
+    },
+    {
+      "epoch": 2.3251830706218195,
+      "grad_norm": 0.13187534712299806,
+      "learning_rate": 1.8724836033874966e-06,
+      "loss": 0.7402,
+      "step": 4687
+    },
+    {
+      "epoch": 2.325679533325059,
+      "grad_norm": 0.12450825043027855,
+      "learning_rate": 1.8715375912047695e-06,
+      "loss": 0.6719,
+      "step": 4688
+    },
+    {
+      "epoch": 2.326175996028298,
+      "grad_norm": 0.1264511706987313,
+      "learning_rate": 1.8705916750724462e-06,
+      "loss": 0.7006,
+      "step": 4689
+    },
+    {
+      "epoch": 2.326672458731538,
+      "grad_norm": 0.13114832252468836,
+      "learning_rate": 1.8696458551350927e-06,
+      "loss": 0.7447,
+      "step": 4690
+    },
+    {
+      "epoch": 2.3271689214347773,
+      "grad_norm": 0.1354210750717035,
+      "learning_rate": 1.868700131537263e-06,
+      "loss": 0.7688,
+      "step": 4691
+    },
+    {
+      "epoch": 2.3276653841380166,
+      "grad_norm": 0.1301186261261761,
+      "learning_rate": 1.8677545044234962e-06,
+      "loss": 0.7138,
+      "step": 4692
+    },
+    {
+      "epoch": 2.328161846841256,
+      "grad_norm": 0.13060257839716047,
+      "learning_rate": 1.8668089739383165e-06,
+      "loss": 0.7208,
+      "step": 4693
+    },
+    {
+      "epoch": 2.3286583095444957,
+      "grad_norm": 0.12633300107009157,
+      "learning_rate": 1.865863540226232e-06,
+      "loss": 0.6773,
+      "step": 4694
+    },
+    {
+      "epoch": 2.329154772247735,
+      "grad_norm": 0.12498631011162595,
+      "learning_rate": 1.8649182034317382e-06,
+      "loss": 0.693,
+      "step": 4695
+    },
+    {
+      "epoch": 2.3296512349509744,
+      "grad_norm": 0.13096791761906182,
+      "learning_rate": 1.863972963699314e-06,
+      "loss": 0.7156,
+      "step": 4696
+    },
+    {
+      "epoch": 2.3301476976542137,
+      "grad_norm": 0.1302613925721332,
+      "learning_rate": 1.8630278211734243e-06,
+      "loss": 0.7099,
+      "step": 4697
+    },
+    {
+      "epoch": 2.330644160357453,
+      "grad_norm": 0.13213767324984668,
+      "learning_rate": 1.86208277599852e-06,
+      "loss": 0.7447,
+      "step": 4698
+    },
+    {
+      "epoch": 2.3311406230606924,
+      "grad_norm": 0.1307800597646881,
+      "learning_rate": 1.861137828319034e-06,
+      "loss": 0.7238,
+      "step": 4699
+    },
+    {
+      "epoch": 2.331637085763932,
+      "grad_norm": 0.12668332112462782,
+      "learning_rate": 1.8601929782793882e-06,
+      "loss": 0.663,
+      "step": 4700
+    },
+    {
+      "epoch": 2.3321335484671715,
+      "grad_norm": 0.1299102574958521,
+      "learning_rate": 1.8592482260239869e-06,
+      "loss": 0.7101,
+      "step": 4701
+    },
+    {
+      "epoch": 2.332630011170411,
+      "grad_norm": 0.1263931699115363,
+      "learning_rate": 1.8583035716972203e-06,
+      "loss": 0.6929,
+      "step": 4702
+    },
+    {
+      "epoch": 2.33312647387365,
+      "grad_norm": 0.12904444206327081,
+      "learning_rate": 1.8573590154434648e-06,
+      "loss": 0.7177,
+      "step": 4703
+    },
+    {
+      "epoch": 2.33362293657689,
+      "grad_norm": 0.12940234345889146,
+      "learning_rate": 1.856414557407079e-06,
+      "loss": 0.6601,
+      "step": 4704
+    },
+    {
+      "epoch": 2.3341193992801292,
+      "grad_norm": 0.12912909817222648,
+      "learning_rate": 1.8554701977324104e-06,
+      "loss": 0.7058,
+      "step": 4705
+    },
+    {
+      "epoch": 2.3346158619833686,
+      "grad_norm": 0.13166188626901182,
+      "learning_rate": 1.8545259365637869e-06,
+      "loss": 0.6971,
+      "step": 4706
+    },
+    {
+      "epoch": 2.335112324686608,
+      "grad_norm": 0.13412107252812872,
+      "learning_rate": 1.8535817740455243e-06,
+      "loss": 0.7281,
+      "step": 4707
+    },
+    {
+      "epoch": 2.3356087873898472,
+      "grad_norm": 0.14219099357241585,
+      "learning_rate": 1.8526377103219228e-06,
+      "loss": 0.7635,
+      "step": 4708
+    },
+    {
+      "epoch": 2.3361052500930866,
+      "grad_norm": 0.12785200595648355,
+      "learning_rate": 1.851693745537268e-06,
+      "loss": 0.6974,
+      "step": 4709
+    },
+    {
+      "epoch": 2.3366017127963263,
+      "grad_norm": 0.1281849252246767,
+      "learning_rate": 1.8507498798358298e-06,
+      "loss": 0.7275,
+      "step": 4710
+    },
+    {
+      "epoch": 2.3370981754995657,
+      "grad_norm": 0.13346008136791995,
+      "learning_rate": 1.849806113361862e-06,
+      "loss": 0.7424,
+      "step": 4711
+    },
+    {
+      "epoch": 2.337594638202805,
+      "grad_norm": 0.12589143302285735,
+      "learning_rate": 1.8488624462596045e-06,
+      "loss": 0.7088,
+      "step": 4712
+    },
+    {
+      "epoch": 2.3380911009060443,
+      "grad_norm": 0.12714926877182972,
+      "learning_rate": 1.8479188786732821e-06,
+      "loss": 0.7187,
+      "step": 4713
+    },
+    {
+      "epoch": 2.3385875636092837,
+      "grad_norm": 0.1279972258903848,
+      "learning_rate": 1.8469754107471045e-06,
+      "loss": 0.6718,
+      "step": 4714
+    },
+    {
+      "epoch": 2.3390840263125234,
+      "grad_norm": 0.1312317942984953,
+      "learning_rate": 1.846032042625264e-06,
+      "loss": 0.7033,
+      "step": 4715
+    },
+    {
+      "epoch": 2.3395804890157628,
+      "grad_norm": 0.12873451062526245,
+      "learning_rate": 1.8450887744519402e-06,
+      "loss": 0.702,
+      "step": 4716
+    },
+    {
+      "epoch": 2.340076951719002,
+      "grad_norm": 0.12882507141223737,
+      "learning_rate": 1.844145606371297e-06,
+      "loss": 0.7278,
+      "step": 4717
+    },
+    {
+      "epoch": 2.3405734144222414,
+      "grad_norm": 0.12746855121924552,
+      "learning_rate": 1.8432025385274816e-06,
+      "loss": 0.6702,
+      "step": 4718
+    },
+    {
+      "epoch": 2.3410698771254808,
+      "grad_norm": 0.1279698618848009,
+      "learning_rate": 1.8422595710646279e-06,
+      "loss": 0.6851,
+      "step": 4719
+    },
+    {
+      "epoch": 2.3415663398287205,
+      "grad_norm": 0.13439282233196687,
+      "learning_rate": 1.841316704126852e-06,
+      "loss": 0.7056,
+      "step": 4720
+    },
+    {
+      "epoch": 2.34206280253196,
+      "grad_norm": 0.13075250402139188,
+      "learning_rate": 1.840373937858257e-06,
+      "loss": 0.7028,
+      "step": 4721
+    },
+    {
+      "epoch": 2.342559265235199,
+      "grad_norm": 0.13576682060101108,
+      "learning_rate": 1.8394312724029302e-06,
+      "loss": 0.715,
+      "step": 4722
+    },
+    {
+      "epoch": 2.3430557279384385,
+      "grad_norm": 0.13095475394049344,
+      "learning_rate": 1.838488707904941e-06,
+      "loss": 0.7263,
+      "step": 4723
+    },
+    {
+      "epoch": 2.343552190641678,
+      "grad_norm": 0.12252662103848144,
+      "learning_rate": 1.8375462445083464e-06,
+      "loss": 0.6212,
+      "step": 4724
+    },
+    {
+      "epoch": 2.3440486533449176,
+      "grad_norm": 0.1321014567093405,
+      "learning_rate": 1.8366038823571864e-06,
+      "loss": 0.6869,
+      "step": 4725
+    },
+    {
+      "epoch": 2.344545116048157,
+      "grad_norm": 0.12839430799368412,
+      "learning_rate": 1.835661621595487e-06,
+      "loss": 0.7022,
+      "step": 4726
+    },
+    {
+      "epoch": 2.3450415787513963,
+      "grad_norm": 0.13572608728421834,
+      "learning_rate": 1.834719462367256e-06,
+      "loss": 0.7124,
+      "step": 4727
+    },
+    {
+      "epoch": 2.3455380414546356,
+      "grad_norm": 0.1327198212125845,
+      "learning_rate": 1.8337774048164886e-06,
+      "loss": 0.7145,
+      "step": 4728
+    },
+    {
+      "epoch": 2.346034504157875,
+      "grad_norm": 0.13233453890145816,
+      "learning_rate": 1.8328354490871624e-06,
+      "loss": 0.723,
+      "step": 4729
+    },
+    {
+      "epoch": 2.3465309668611147,
+      "grad_norm": 0.1275819850450933,
+      "learning_rate": 1.8318935953232417e-06,
+      "loss": 0.6562,
+      "step": 4730
+    },
+    {
+      "epoch": 2.347027429564354,
+      "grad_norm": 0.12862371806632422,
+      "learning_rate": 1.8309518436686718e-06,
+      "loss": 0.6557,
+      "step": 4731
+    },
+    {
+      "epoch": 2.3475238922675934,
+      "grad_norm": 0.13020123965864694,
+      "learning_rate": 1.8300101942673854e-06,
+      "loss": 0.7,
+      "step": 4732
+    },
+    {
+      "epoch": 2.3480203549708327,
+      "grad_norm": 0.13526821353378768,
+      "learning_rate": 1.829068647263298e-06,
+      "loss": 0.7129,
+      "step": 4733
+    },
+    {
+      "epoch": 2.348516817674072,
+      "grad_norm": 0.12778442163510337,
+      "learning_rate": 1.8281272028003105e-06,
+      "loss": 0.6919,
+      "step": 4734
+    },
+    {
+      "epoch": 2.349013280377312,
+      "grad_norm": 0.12730442338186904,
+      "learning_rate": 1.827185861022308e-06,
+      "loss": 0.6737,
+      "step": 4735
+    },
+    {
+      "epoch": 2.349509743080551,
+      "grad_norm": 0.13135300464883143,
+      "learning_rate": 1.8262446220731583e-06,
+      "loss": 0.6902,
+      "step": 4736
+    },
+    {
+      "epoch": 2.3500062057837905,
+      "grad_norm": 0.13633030408853347,
+      "learning_rate": 1.8253034860967156e-06,
+      "loss": 0.7665,
+      "step": 4737
+    },
+    {
+      "epoch": 2.35050266848703,
+      "grad_norm": 0.12967623071202264,
+      "learning_rate": 1.8243624532368176e-06,
+      "loss": 0.6925,
+      "step": 4738
+    },
+    {
+      "epoch": 2.350999131190269,
+      "grad_norm": 0.12823110926047526,
+      "learning_rate": 1.823421523637285e-06,
+      "loss": 0.6446,
+      "step": 4739
+    },
+    {
+      "epoch": 2.351495593893509,
+      "grad_norm": 0.13140489998321872,
+      "learning_rate": 1.822480697441924e-06,
+      "loss": 0.736,
+      "step": 4740
+    },
+    {
+      "epoch": 2.3519920565967483,
+      "grad_norm": 0.12898773935905025,
+      "learning_rate": 1.8215399747945256e-06,
+      "loss": 0.7298,
+      "step": 4741
+    },
+    {
+      "epoch": 2.3524885192999876,
+      "grad_norm": 0.139264930199382,
+      "learning_rate": 1.820599355838864e-06,
+      "loss": 0.716,
+      "step": 4742
+    },
+    {
+      "epoch": 2.352984982003227,
+      "grad_norm": 0.12812055474218081,
+      "learning_rate": 1.8196588407186972e-06,
+      "loss": 0.7316,
+      "step": 4743
+    },
+    {
+      "epoch": 2.3534814447064663,
+      "grad_norm": 0.1305997310787121,
+      "learning_rate": 1.8187184295777676e-06,
+      "loss": 0.6861,
+      "step": 4744
+    },
+    {
+      "epoch": 2.353977907409706,
+      "grad_norm": 0.1353100470985911,
+      "learning_rate": 1.8177781225598033e-06,
+      "loss": 0.7478,
+      "step": 4745
+    },
+    {
+      "epoch": 2.3544743701129454,
+      "grad_norm": 0.13303661646904097,
+      "learning_rate": 1.816837919808514e-06,
+      "loss": 0.7119,
+      "step": 4746
+    },
+    {
+      "epoch": 2.3549708328161847,
+      "grad_norm": 0.13363770660484806,
+      "learning_rate": 1.8158978214675953e-06,
+      "loss": 0.673,
+      "step": 4747
+    },
+    {
+      "epoch": 2.355467295519424,
+      "grad_norm": 0.13089630550011988,
+      "learning_rate": 1.8149578276807257e-06,
+      "loss": 0.6883,
+      "step": 4748
+    },
+    {
+      "epoch": 2.3559637582226634,
+      "grad_norm": 0.13168358638427374,
+      "learning_rate": 1.814017938591568e-06,
+      "loss": 0.6873,
+      "step": 4749
+    },
+    {
+      "epoch": 2.356460220925903,
+      "grad_norm": 0.13484692782487573,
+      "learning_rate": 1.8130781543437698e-06,
+      "loss": 0.7583,
+      "step": 4750
+    },
+    {
+      "epoch": 2.3569566836291425,
+      "grad_norm": 0.1306126749166303,
+      "learning_rate": 1.8121384750809623e-06,
+      "loss": 0.6905,
+      "step": 4751
+    },
+    {
+      "epoch": 2.357453146332382,
+      "grad_norm": 0.12784958982539274,
+      "learning_rate": 1.811198900946759e-06,
+      "loss": 0.6999,
+      "step": 4752
+    },
+    {
+      "epoch": 2.357949609035621,
+      "grad_norm": 0.13515705011245394,
+      "learning_rate": 1.81025943208476e-06,
+      "loss": 0.7064,
+      "step": 4753
+    },
+    {
+      "epoch": 2.3584460717388605,
+      "grad_norm": 0.13046415278522383,
+      "learning_rate": 1.8093200686385484e-06,
+      "loss": 0.7217,
+      "step": 4754
+    },
+    {
+      "epoch": 2.3589425344421002,
+      "grad_norm": 0.12991942766461034,
+      "learning_rate": 1.8083808107516892e-06,
+      "loss": 0.71,
+      "step": 4755
+    },
+    {
+      "epoch": 2.3594389971453396,
+      "grad_norm": 0.13073051110349773,
+      "learning_rate": 1.8074416585677335e-06,
+      "loss": 0.6924,
+      "step": 4756
+    },
+    {
+      "epoch": 2.359935459848579,
+      "grad_norm": 0.13719409511910485,
+      "learning_rate": 1.8065026122302165e-06,
+      "loss": 0.7551,
+      "step": 4757
+    },
+    {
+      "epoch": 2.360431922551818,
+      "grad_norm": 0.12947943146773364,
+      "learning_rate": 1.8055636718826557e-06,
+      "loss": 0.6755,
+      "step": 4758
+    },
+    {
+      "epoch": 2.3609283852550575,
+      "grad_norm": 0.12841550971350973,
+      "learning_rate": 1.8046248376685532e-06,
+      "loss": 0.703,
+      "step": 4759
+    },
+    {
+      "epoch": 2.3614248479582973,
+      "grad_norm": 0.1311479945588234,
+      "learning_rate": 1.8036861097313946e-06,
+      "loss": 0.7219,
+      "step": 4760
+    },
+    {
+      "epoch": 2.3619213106615367,
+      "grad_norm": 0.12849688100852175,
+      "learning_rate": 1.8027474882146498e-06,
+      "loss": 0.6653,
+      "step": 4761
+    },
+    {
+      "epoch": 2.362417773364776,
+      "grad_norm": 0.13113427313547274,
+      "learning_rate": 1.8018089732617716e-06,
+      "loss": 0.7302,
+      "step": 4762
+    },
+    {
+      "epoch": 2.3629142360680153,
+      "grad_norm": 0.12964668689183,
+      "learning_rate": 1.8008705650161979e-06,
+      "loss": 0.7078,
+      "step": 4763
+    },
+    {
+      "epoch": 2.3634106987712546,
+      "grad_norm": 0.13260160889826755,
+      "learning_rate": 1.799932263621348e-06,
+      "loss": 0.7268,
+      "step": 4764
+    },
+    {
+      "epoch": 2.3639071614744944,
+      "grad_norm": 0.13006599121273307,
+      "learning_rate": 1.7989940692206267e-06,
+      "loss": 0.7147,
+      "step": 4765
+    },
+    {
+      "epoch": 2.3644036241777338,
+      "grad_norm": 0.13478322856677002,
+      "learning_rate": 1.7980559819574222e-06,
+      "loss": 0.7086,
+      "step": 4766
+    },
+    {
+      "epoch": 2.364900086880973,
+      "grad_norm": 0.1278201414701354,
+      "learning_rate": 1.7971180019751068e-06,
+      "loss": 0.6734,
+      "step": 4767
+    },
+    {
+      "epoch": 2.3653965495842124,
+      "grad_norm": 0.12854256872198694,
+      "learning_rate": 1.7961801294170344e-06,
+      "loss": 0.7037,
+      "step": 4768
+    },
+    {
+      "epoch": 2.3658930122874517,
+      "grad_norm": 0.1326157442892243,
+      "learning_rate": 1.7952423644265441e-06,
+      "loss": 0.7093,
+      "step": 4769
+    },
+    {
+      "epoch": 2.366389474990691,
+      "grad_norm": 0.13759009456859725,
+      "learning_rate": 1.7943047071469597e-06,
+      "loss": 0.7401,
+      "step": 4770
+    },
+    {
+      "epoch": 2.366885937693931,
+      "grad_norm": 0.13497019322133594,
+      "learning_rate": 1.7933671577215846e-06,
+      "loss": 0.7085,
+      "step": 4771
+    },
+    {
+      "epoch": 2.36738240039717,
+      "grad_norm": 0.12878040498696178,
+      "learning_rate": 1.7924297162937095e-06,
+      "loss": 0.6994,
+      "step": 4772
+    },
+    {
+      "epoch": 2.3678788631004095,
+      "grad_norm": 0.12913566113987018,
+      "learning_rate": 1.7914923830066074e-06,
+      "loss": 0.7142,
+      "step": 4773
+    },
+    {
+      "epoch": 2.368375325803649,
+      "grad_norm": 0.13029016007788516,
+      "learning_rate": 1.790555158003535e-06,
+      "loss": 0.6907,
+      "step": 4774
+    },
+    {
+      "epoch": 2.3688717885068886,
+      "grad_norm": 0.1317576623479519,
+      "learning_rate": 1.7896180414277308e-06,
+      "loss": 0.7187,
+      "step": 4775
+    },
+    {
+      "epoch": 2.369368251210128,
+      "grad_norm": 0.13402471815823666,
+      "learning_rate": 1.7886810334224192e-06,
+      "loss": 0.6727,
+      "step": 4776
+    },
+    {
+      "epoch": 2.3698647139133673,
+      "grad_norm": 0.12910099553816476,
+      "learning_rate": 1.7877441341308065e-06,
+      "loss": 0.6883,
+      "step": 4777
+    },
+    {
+      "epoch": 2.3703611766166066,
+      "grad_norm": 0.12857345243673593,
+      "learning_rate": 1.7868073436960826e-06,
+      "loss": 0.7037,
+      "step": 4778
+    },
+    {
+      "epoch": 2.370857639319846,
+      "grad_norm": 0.13050987325503174,
+      "learning_rate": 1.7858706622614217e-06,
+      "loss": 0.7294,
+      "step": 4779
+    },
+    {
+      "epoch": 2.3713541020230853,
+      "grad_norm": 0.12825871824045287,
+      "learning_rate": 1.784934089969979e-06,
+      "loss": 0.6861,
+      "step": 4780
+    },
+    {
+      "epoch": 2.371850564726325,
+      "grad_norm": 0.13089580339165302,
+      "learning_rate": 1.783997626964896e-06,
+      "loss": 0.7541,
+      "step": 4781
+    },
+    {
+      "epoch": 2.3723470274295644,
+      "grad_norm": 0.1334628464244765,
+      "learning_rate": 1.783061273389295e-06,
+      "loss": 0.7078,
+      "step": 4782
+    },
+    {
+      "epoch": 2.3728434901328037,
+      "grad_norm": 0.1340782941366848,
+      "learning_rate": 1.7821250293862836e-06,
+      "loss": 0.7718,
+      "step": 4783
+    },
+    {
+      "epoch": 2.373339952836043,
+      "grad_norm": 0.13228953216571118,
+      "learning_rate": 1.7811888950989515e-06,
+      "loss": 0.7161,
+      "step": 4784
+    },
+    {
+      "epoch": 2.373836415539283,
+      "grad_norm": 0.1351301578501817,
+      "learning_rate": 1.780252870670371e-06,
+      "loss": 0.6803,
+      "step": 4785
+    },
+    {
+      "epoch": 2.374332878242522,
+      "grad_norm": 0.12686143379738357,
+      "learning_rate": 1.7793169562435996e-06,
+      "loss": 0.6786,
+      "step": 4786
+    },
+    {
+      "epoch": 2.3748293409457615,
+      "grad_norm": 0.12977295444856893,
+      "learning_rate": 1.7783811519616756e-06,
+      "loss": 0.7315,
+      "step": 4787
+    },
+    {
+      "epoch": 2.375325803649001,
+      "grad_norm": 0.13124111123654278,
+      "learning_rate": 1.777445457967622e-06,
+      "loss": 0.7298,
+      "step": 4788
+    },
+    {
+      "epoch": 2.37582226635224,
+      "grad_norm": 0.13452592431133958,
+      "learning_rate": 1.7765098744044452e-06,
+      "loss": 0.6963,
+      "step": 4789
+    },
+    {
+      "epoch": 2.3763187290554795,
+      "grad_norm": 0.13429719254562902,
+      "learning_rate": 1.7755744014151338e-06,
+      "loss": 0.6756,
+      "step": 4790
+    },
+    {
+      "epoch": 2.3768151917587192,
+      "grad_norm": 0.12721035894910274,
+      "learning_rate": 1.7746390391426598e-06,
+      "loss": 0.6617,
+      "step": 4791
+    },
+    {
+      "epoch": 2.3773116544619586,
+      "grad_norm": 0.12936319272862273,
+      "learning_rate": 1.7737037877299784e-06,
+      "loss": 0.683,
+      "step": 4792
+    },
+    {
+      "epoch": 2.377808117165198,
+      "grad_norm": 0.13813116770259098,
+      "learning_rate": 1.7727686473200276e-06,
+      "loss": 0.7706,
+      "step": 4793
+    },
+    {
+      "epoch": 2.3783045798684372,
+      "grad_norm": 0.13184854493613757,
+      "learning_rate": 1.771833618055729e-06,
+      "loss": 0.7047,
+      "step": 4794
+    },
+    {
+      "epoch": 2.378801042571677,
+      "grad_norm": 0.1326136257727315,
+      "learning_rate": 1.7708987000799866e-06,
+      "loss": 0.6991,
+      "step": 4795
+    },
+    {
+      "epoch": 2.3792975052749163,
+      "grad_norm": 0.13583743329061757,
+      "learning_rate": 1.7699638935356876e-06,
+      "loss": 0.7719,
+      "step": 4796
+    },
+    {
+      "epoch": 2.3797939679781557,
+      "grad_norm": 0.13008368245130267,
+      "learning_rate": 1.7690291985657021e-06,
+      "loss": 0.6925,
+      "step": 4797
+    },
+    {
+      "epoch": 2.380290430681395,
+      "grad_norm": 0.12928279258704853,
+      "learning_rate": 1.7680946153128833e-06,
+      "loss": 0.6964,
+      "step": 4798
+    },
+    {
+      "epoch": 2.3807868933846343,
+      "grad_norm": 0.126896632250804,
+      "learning_rate": 1.7671601439200675e-06,
+      "loss": 0.7119,
+      "step": 4799
+    },
+    {
+      "epoch": 2.3812833560878737,
+      "grad_norm": 0.1343279347693961,
+      "learning_rate": 1.766225784530074e-06,
+      "loss": 0.7165,
+      "step": 4800
+    },
+    {
+      "epoch": 2.3817798187911134,
+      "grad_norm": 0.12647325466363368,
+      "learning_rate": 1.7652915372857035e-06,
+      "loss": 0.6949,
+      "step": 4801
+    },
+    {
+      "epoch": 2.3822762814943528,
+      "grad_norm": 0.1300158284756839,
+      "learning_rate": 1.7643574023297424e-06,
+      "loss": 0.7007,
+      "step": 4802
+    },
+    {
+      "epoch": 2.382772744197592,
+      "grad_norm": 0.1303069401818913,
+      "learning_rate": 1.7634233798049563e-06,
+      "loss": 0.6973,
+      "step": 4803
+    },
+    {
+      "epoch": 2.3832692069008314,
+      "grad_norm": 0.1250185249073528,
+      "learning_rate": 1.7624894698540964e-06,
+      "loss": 0.6759,
+      "step": 4804
+    },
+    {
+      "epoch": 2.383765669604071,
+      "grad_norm": 0.12683133498041957,
+      "learning_rate": 1.7615556726198963e-06,
+      "loss": 0.6577,
+      "step": 4805
+    },
+    {
+      "epoch": 2.3842621323073105,
+      "grad_norm": 0.13065274070747337,
+      "learning_rate": 1.7606219882450713e-06,
+      "loss": 0.7088,
+      "step": 4806
+    },
+    {
+      "epoch": 2.38475859501055,
+      "grad_norm": 0.13340077639503192,
+      "learning_rate": 1.7596884168723209e-06,
+      "loss": 0.7117,
+      "step": 4807
+    },
+    {
+      "epoch": 2.385255057713789,
+      "grad_norm": 0.12846892823423328,
+      "learning_rate": 1.7587549586443253e-06,
+      "loss": 0.7508,
+      "step": 4808
+    },
+    {
+      "epoch": 2.3857515204170285,
+      "grad_norm": 0.133668088002894,
+      "learning_rate": 1.7578216137037496e-06,
+      "loss": 0.7036,
+      "step": 4809
+    },
+    {
+      "epoch": 2.386247983120268,
+      "grad_norm": 0.1297358762173912,
+      "learning_rate": 1.75688838219324e-06,
+      "loss": 0.6972,
+      "step": 4810
+    },
+    {
+      "epoch": 2.3867444458235076,
+      "grad_norm": 0.12907419259927128,
+      "learning_rate": 1.755955264255427e-06,
+      "loss": 0.6749,
+      "step": 4811
+    },
+    {
+      "epoch": 2.387240908526747,
+      "grad_norm": 0.14320560575199254,
+      "learning_rate": 1.7550222600329214e-06,
+      "loss": 0.7493,
+      "step": 4812
+    },
+    {
+      "epoch": 2.3877373712299863,
+      "grad_norm": 0.13107914192089937,
+      "learning_rate": 1.7540893696683187e-06,
+      "loss": 0.6801,
+      "step": 4813
+    },
+    {
+      "epoch": 2.3882338339332256,
+      "grad_norm": 0.1299399190627306,
+      "learning_rate": 1.753156593304196e-06,
+      "loss": 0.6749,
+      "step": 4814
+    },
+    {
+      "epoch": 2.3887302966364654,
+      "grad_norm": 0.12936932954561534,
+      "learning_rate": 1.7522239310831134e-06,
+      "loss": 0.7269,
+      "step": 4815
+    },
+    {
+      "epoch": 2.3892267593397047,
+      "grad_norm": 0.13265135552744697,
+      "learning_rate": 1.7512913831476135e-06,
+      "loss": 0.7383,
+      "step": 4816
+    },
+    {
+      "epoch": 2.389723222042944,
+      "grad_norm": 0.13764593191752025,
+      "learning_rate": 1.750358949640221e-06,
+      "loss": 0.7176,
+      "step": 4817
+    },
+    {
+      "epoch": 2.3902196847461834,
+      "grad_norm": 0.12796750916884678,
+      "learning_rate": 1.749426630703444e-06,
+      "loss": 0.6821,
+      "step": 4818
+    },
+    {
+      "epoch": 2.3907161474494227,
+      "grad_norm": 0.13180747695825185,
+      "learning_rate": 1.7484944264797713e-06,
+      "loss": 0.7342,
+      "step": 4819
+    },
+    {
+      "epoch": 2.391212610152662,
+      "grad_norm": 0.14126658198697672,
+      "learning_rate": 1.7475623371116759e-06,
+      "loss": 0.7135,
+      "step": 4820
+    },
+    {
+      "epoch": 2.391709072855902,
+      "grad_norm": 0.1305260276629293,
+      "learning_rate": 1.7466303627416129e-06,
+      "loss": 0.7204,
+      "step": 4821
+    },
+    {
+      "epoch": 2.392205535559141,
+      "grad_norm": 0.12966913916891723,
+      "learning_rate": 1.7456985035120194e-06,
+      "loss": 0.6982,
+      "step": 4822
+    },
+    {
+      "epoch": 2.3927019982623805,
+      "grad_norm": 0.1254585082749373,
+      "learning_rate": 1.7447667595653162e-06,
+      "loss": 0.6747,
+      "step": 4823
+    },
+    {
+      "epoch": 2.39319846096562,
+      "grad_norm": 0.15094163056194543,
+      "learning_rate": 1.7438351310439036e-06,
+      "loss": 0.7751,
+      "step": 4824
+    },
+    {
+      "epoch": 2.3936949236688596,
+      "grad_norm": 0.12776045500839703,
+      "learning_rate": 1.7429036180901674e-06,
+      "loss": 0.7159,
+      "step": 4825
+    },
+    {
+      "epoch": 2.394191386372099,
+      "grad_norm": 0.13208737547487745,
+      "learning_rate": 1.741972220846474e-06,
+      "loss": 0.67,
+      "step": 4826
+    },
+    {
+      "epoch": 2.3946878490753383,
+      "grad_norm": 0.12814416225344802,
+      "learning_rate": 1.741040939455173e-06,
+      "loss": 0.6881,
+      "step": 4827
+    },
+    {
+      "epoch": 2.3951843117785776,
+      "grad_norm": 0.1329316175712077,
+      "learning_rate": 1.7401097740585947e-06,
+      "loss": 0.7058,
+      "step": 4828
+    },
+    {
+      "epoch": 2.395680774481817,
+      "grad_norm": 0.1314293899989232,
+      "learning_rate": 1.7391787247990537e-06,
+      "loss": 0.6882,
+      "step": 4829
+    },
+    {
+      "epoch": 2.3961772371850563,
+      "grad_norm": 0.13031161109059494,
+      "learning_rate": 1.7382477918188462e-06,
+      "loss": 0.7541,
+      "step": 4830
+    },
+    {
+      "epoch": 2.396673699888296,
+      "grad_norm": 0.12930814709648858,
+      "learning_rate": 1.7373169752602493e-06,
+      "loss": 0.7058,
+      "step": 4831
+    },
+    {
+      "epoch": 2.3971701625915354,
+      "grad_norm": 0.1291944574432241,
+      "learning_rate": 1.7363862752655248e-06,
+      "loss": 0.6674,
+      "step": 4832
+    },
+    {
+      "epoch": 2.3976666252947747,
+      "grad_norm": 0.1304229911398871,
+      "learning_rate": 1.735455691976914e-06,
+      "loss": 0.6796,
+      "step": 4833
+    },
+    {
+      "epoch": 2.398163087998014,
+      "grad_norm": 0.13562249887328157,
+      "learning_rate": 1.7345252255366434e-06,
+      "loss": 0.7102,
+      "step": 4834
+    },
+    {
+      "epoch": 2.398659550701254,
+      "grad_norm": 0.13155960598330008,
+      "learning_rate": 1.7335948760869175e-06,
+      "loss": 0.6952,
+      "step": 4835
+    },
+    {
+      "epoch": 2.399156013404493,
+      "grad_norm": 0.12956566494989916,
+      "learning_rate": 1.7326646437699262e-06,
+      "loss": 0.7399,
+      "step": 4836
+    },
+    {
+      "epoch": 2.3996524761077325,
+      "grad_norm": 0.129991409065886,
+      "learning_rate": 1.7317345287278408e-06,
+      "loss": 0.6942,
+      "step": 4837
+    },
+    {
+      "epoch": 2.400148938810972,
+      "grad_norm": 0.13550941541419204,
+      "learning_rate": 1.7308045311028148e-06,
+      "loss": 0.7386,
+      "step": 4838
+    },
+    {
+      "epoch": 2.400645401514211,
+      "grad_norm": 0.12897072490235445,
+      "learning_rate": 1.7298746510369836e-06,
+      "loss": 0.6888,
+      "step": 4839
+    },
+    {
+      "epoch": 2.4011418642174505,
+      "grad_norm": 0.12791367362789313,
+      "learning_rate": 1.728944888672463e-06,
+      "loss": 0.6921,
+      "step": 4840
+    },
+    {
+      "epoch": 2.4016383269206902,
+      "grad_norm": 0.12980510973949744,
+      "learning_rate": 1.7280152441513536e-06,
+      "loss": 0.6914,
+      "step": 4841
+    },
+    {
+      "epoch": 2.4021347896239296,
+      "grad_norm": 0.12698243358765066,
+      "learning_rate": 1.7270857176157363e-06,
+      "loss": 0.7243,
+      "step": 4842
+    },
+    {
+      "epoch": 2.402631252327169,
+      "grad_norm": 0.12943217000030752,
+      "learning_rate": 1.7261563092076742e-06,
+      "loss": 0.6894,
+      "step": 4843
+    },
+    {
+      "epoch": 2.4031277150304082,
+      "grad_norm": 0.14555273425787743,
+      "learning_rate": 1.7252270190692133e-06,
+      "loss": 0.7761,
+      "step": 4844
+    },
+    {
+      "epoch": 2.403624177733648,
+      "grad_norm": 0.13293572550265437,
+      "learning_rate": 1.724297847342379e-06,
+      "loss": 0.7029,
+      "step": 4845
+    },
+    {
+      "epoch": 2.4041206404368873,
+      "grad_norm": 0.12653010232213735,
+      "learning_rate": 1.7233687941691819e-06,
+      "loss": 0.6353,
+      "step": 4846
+    },
+    {
+      "epoch": 2.4046171031401267,
+      "grad_norm": 0.1314378535643785,
+      "learning_rate": 1.722439859691612e-06,
+      "loss": 0.698,
+      "step": 4847
+    },
+    {
+      "epoch": 2.405113565843366,
+      "grad_norm": 0.12696993067028972,
+      "learning_rate": 1.7215110440516427e-06,
+      "loss": 0.6477,
+      "step": 4848
+    },
+    {
+      "epoch": 2.4056100285466053,
+      "grad_norm": 0.1293761765904413,
+      "learning_rate": 1.720582347391228e-06,
+      "loss": 0.6863,
+      "step": 4849
+    },
+    {
+      "epoch": 2.4061064912498447,
+      "grad_norm": 0.12977487418429212,
+      "learning_rate": 1.7196537698523052e-06,
+      "loss": 0.6806,
+      "step": 4850
+    },
+    {
+      "epoch": 2.4066029539530844,
+      "grad_norm": 0.13188762511672084,
+      "learning_rate": 1.718725311576791e-06,
+      "loss": 0.7124,
+      "step": 4851
+    },
+    {
+      "epoch": 2.4070994166563238,
+      "grad_norm": 0.1290870421059128,
+      "learning_rate": 1.7177969727065861e-06,
+      "loss": 0.7406,
+      "step": 4852
+    },
+    {
+      "epoch": 2.407595879359563,
+      "grad_norm": 0.1324826867302518,
+      "learning_rate": 1.716868753383572e-06,
+      "loss": 0.7023,
+      "step": 4853
+    },
+    {
+      "epoch": 2.4080923420628024,
+      "grad_norm": 0.13190718413667324,
+      "learning_rate": 1.7159406537496127e-06,
+      "loss": 0.698,
+      "step": 4854
+    },
+    {
+      "epoch": 2.4085888047660418,
+      "grad_norm": 0.12929265660157932,
+      "learning_rate": 1.7150126739465534e-06,
+      "loss": 0.6906,
+      "step": 4855
+    },
+    {
+      "epoch": 2.4090852674692815,
+      "grad_norm": 0.12998790137705218,
+      "learning_rate": 1.7140848141162201e-06,
+      "loss": 0.663,
+      "step": 4856
+    },
+    {
+      "epoch": 2.409581730172521,
+      "grad_norm": 0.13078521035118826,
+      "learning_rate": 1.7131570744004216e-06,
+      "loss": 0.6851,
+      "step": 4857
+    },
+    {
+      "epoch": 2.41007819287576,
+      "grad_norm": 0.12843707540780416,
+      "learning_rate": 1.7122294549409486e-06,
+      "loss": 0.6911,
+      "step": 4858
+    },
+    {
+      "epoch": 2.4105746555789995,
+      "grad_norm": 0.13192234560191793,
+      "learning_rate": 1.7113019558795722e-06,
+      "loss": 0.7878,
+      "step": 4859
+    },
+    {
+      "epoch": 2.411071118282239,
+      "grad_norm": 0.1319340861365238,
+      "learning_rate": 1.7103745773580465e-06,
+      "loss": 0.7045,
+      "step": 4860
+    },
+    {
+      "epoch": 2.4115675809854786,
+      "grad_norm": 0.13307582097518272,
+      "learning_rate": 1.7094473195181054e-06,
+      "loss": 0.7248,
+      "step": 4861
+    },
+    {
+      "epoch": 2.412064043688718,
+      "grad_norm": 0.13305909648819517,
+      "learning_rate": 1.7085201825014663e-06,
+      "loss": 0.7069,
+      "step": 4862
+    },
+    {
+      "epoch": 2.4125605063919573,
+      "grad_norm": 0.12564507225345678,
+      "learning_rate": 1.7075931664498265e-06,
+      "loss": 0.6419,
+      "step": 4863
+    },
+    {
+      "epoch": 2.4130569690951966,
+      "grad_norm": 0.1353169682104955,
+      "learning_rate": 1.7066662715048668e-06,
+      "loss": 0.74,
+      "step": 4864
+    },
+    {
+      "epoch": 2.413553431798436,
+      "grad_norm": 0.12985183063208408,
+      "learning_rate": 1.7057394978082465e-06,
+      "loss": 0.6654,
+      "step": 4865
+    },
+    {
+      "epoch": 2.4140498945016757,
+      "grad_norm": 0.13177829235442678,
+      "learning_rate": 1.7048128455016105e-06,
+      "loss": 0.7362,
+      "step": 4866
+    },
+    {
+      "epoch": 2.414546357204915,
+      "grad_norm": 0.13043257360310376,
+      "learning_rate": 1.7038863147265796e-06,
+      "loss": 0.7091,
+      "step": 4867
+    },
+    {
+      "epoch": 2.4150428199081544,
+      "grad_norm": 0.12897077992222986,
+      "learning_rate": 1.702959905624761e-06,
+      "loss": 0.7194,
+      "step": 4868
+    },
+    {
+      "epoch": 2.4155392826113937,
+      "grad_norm": 0.129946311887381,
+      "learning_rate": 1.7020336183377414e-06,
+      "loss": 0.7313,
+      "step": 4869
+    },
+    {
+      "epoch": 2.416035745314633,
+      "grad_norm": 0.13239014442468938,
+      "learning_rate": 1.7011074530070888e-06,
+      "loss": 0.7355,
+      "step": 4870
+    },
+    {
+      "epoch": 2.416532208017873,
+      "grad_norm": 0.13601224242524987,
+      "learning_rate": 1.700181409774353e-06,
+      "loss": 0.7279,
+      "step": 4871
+    },
+    {
+      "epoch": 2.417028670721112,
+      "grad_norm": 0.13833920681618717,
+      "learning_rate": 1.6992554887810642e-06,
+      "loss": 0.7629,
+      "step": 4872
+    },
+    {
+      "epoch": 2.4175251334243515,
+      "grad_norm": 0.13419638193226116,
+      "learning_rate": 1.6983296901687351e-06,
+      "loss": 0.7244,
+      "step": 4873
+    },
+    {
+      "epoch": 2.418021596127591,
+      "grad_norm": 0.12689180227839011,
+      "learning_rate": 1.697404014078859e-06,
+      "loss": 0.6595,
+      "step": 4874
+    },
+    {
+      "epoch": 2.41851805883083,
+      "grad_norm": 0.13100363032562112,
+      "learning_rate": 1.6964784606529106e-06,
+      "loss": 0.6761,
+      "step": 4875
+    },
+    {
+      "epoch": 2.41901452153407,
+      "grad_norm": 0.13288360240949995,
+      "learning_rate": 1.6955530300323467e-06,
+      "loss": 0.7103,
+      "step": 4876
+    },
+    {
+      "epoch": 2.4195109842373093,
+      "grad_norm": 0.13073699217986937,
+      "learning_rate": 1.6946277223586033e-06,
+      "loss": 0.688,
+      "step": 4877
+    },
+    {
+      "epoch": 2.4200074469405486,
+      "grad_norm": 0.13046075666698376,
+      "learning_rate": 1.6937025377730992e-06,
+      "loss": 0.6997,
+      "step": 4878
+    },
+    {
+      "epoch": 2.420503909643788,
+      "grad_norm": 0.1375868918604195,
+      "learning_rate": 1.6927774764172344e-06,
+      "loss": 0.7279,
+      "step": 4879
+    },
+    {
+      "epoch": 2.4210003723470273,
+      "grad_norm": 0.1290611075411292,
+      "learning_rate": 1.6918525384323892e-06,
+      "loss": 0.6923,
+      "step": 4880
+    },
+    {
+      "epoch": 2.421496835050267,
+      "grad_norm": 0.12719530065201395,
+      "learning_rate": 1.6909277239599266e-06,
+      "loss": 0.7026,
+      "step": 4881
+    },
+    {
+      "epoch": 2.4219932977535064,
+      "grad_norm": 0.12404152026479529,
+      "learning_rate": 1.690003033141189e-06,
+      "loss": 0.6793,
+      "step": 4882
+    },
+    {
+      "epoch": 2.4224897604567457,
+      "grad_norm": 0.1281425281209562,
+      "learning_rate": 1.6890784661175003e-06,
+      "loss": 0.6684,
+      "step": 4883
+    },
+    {
+      "epoch": 2.422986223159985,
+      "grad_norm": 0.13530606073217435,
+      "learning_rate": 1.6881540230301654e-06,
+      "loss": 0.7593,
+      "step": 4884
+    },
+    {
+      "epoch": 2.4234826858632244,
+      "grad_norm": 0.13469127543994983,
+      "learning_rate": 1.687229704020471e-06,
+      "loss": 0.7071,
+      "step": 4885
+    },
+    {
+      "epoch": 2.423979148566464,
+      "grad_norm": 0.13832168876093476,
+      "learning_rate": 1.6863055092296849e-06,
+      "loss": 0.7108,
+      "step": 4886
+    },
+    {
+      "epoch": 2.4244756112697035,
+      "grad_norm": 0.13226395995413887,
+      "learning_rate": 1.6853814387990553e-06,
+      "loss": 0.7019,
+      "step": 4887
+    },
+    {
+      "epoch": 2.424972073972943,
+      "grad_norm": 0.13201504974645803,
+      "learning_rate": 1.684457492869811e-06,
+      "loss": 0.7436,
+      "step": 4888
+    },
+    {
+      "epoch": 2.425468536676182,
+      "grad_norm": 0.12931429802413458,
+      "learning_rate": 1.683533671583163e-06,
+      "loss": 0.7012,
+      "step": 4889
+    },
+    {
+      "epoch": 2.4259649993794214,
+      "grad_norm": 0.13156592559267707,
+      "learning_rate": 1.6826099750803015e-06,
+      "loss": 0.7484,
+      "step": 4890
+    },
+    {
+      "epoch": 2.4264614620826612,
+      "grad_norm": 0.12950517255649976,
+      "learning_rate": 1.6816864035024e-06,
+      "loss": 0.7049,
+      "step": 4891
+    },
+    {
+      "epoch": 2.4269579247859006,
+      "grad_norm": 0.12953295412900734,
+      "learning_rate": 1.6807629569906113e-06,
+      "loss": 0.7068,
+      "step": 4892
+    },
+    {
+      "epoch": 2.42745438748914,
+      "grad_norm": 0.13090036376760572,
+      "learning_rate": 1.679839635686069e-06,
+      "loss": 0.705,
+      "step": 4893
+    },
+    {
+      "epoch": 2.427950850192379,
+      "grad_norm": 0.12975357279758154,
+      "learning_rate": 1.678916439729888e-06,
+      "loss": 0.6817,
+      "step": 4894
+    },
+    {
+      "epoch": 2.4284473128956185,
+      "grad_norm": 0.13799080774336994,
+      "learning_rate": 1.6779933692631639e-06,
+      "loss": 0.7595,
+      "step": 4895
+    },
+    {
+      "epoch": 2.4289437755988583,
+      "grad_norm": 0.1350060724395047,
+      "learning_rate": 1.6770704244269737e-06,
+      "loss": 0.7284,
+      "step": 4896
+    },
+    {
+      "epoch": 2.4294402383020977,
+      "grad_norm": 0.1303139802763453,
+      "learning_rate": 1.6761476053623748e-06,
+      "loss": 0.7219,
+      "step": 4897
+    },
+    {
+      "epoch": 2.429936701005337,
+      "grad_norm": 0.1273561658716899,
+      "learning_rate": 1.675224912210405e-06,
+      "loss": 0.7254,
+      "step": 4898
+    },
+    {
+      "epoch": 2.4304331637085763,
+      "grad_norm": 0.13457224415207078,
+      "learning_rate": 1.6743023451120831e-06,
+      "loss": 0.7421,
+      "step": 4899
+    },
+    {
+      "epoch": 2.4309296264118156,
+      "grad_norm": 0.13264396885754928,
+      "learning_rate": 1.673379904208408e-06,
+      "loss": 0.7332,
+      "step": 4900
+    },
+    {
+      "epoch": 2.4314260891150554,
+      "grad_norm": 0.1349589055432028,
+      "learning_rate": 1.672457589640361e-06,
+      "loss": 0.7287,
+      "step": 4901
+    },
+    {
+      "epoch": 2.4319225518182948,
+      "grad_norm": 0.13262022984457753,
+      "learning_rate": 1.6715354015489028e-06,
+      "loss": 0.7355,
+      "step": 4902
+    },
+    {
+      "epoch": 2.432419014521534,
+      "grad_norm": 0.13029145762137417,
+      "learning_rate": 1.6706133400749752e-06,
+      "loss": 0.7185,
+      "step": 4903
+    },
+    {
+      "epoch": 2.4329154772247734,
+      "grad_norm": 0.12961039350323814,
+      "learning_rate": 1.6696914053595004e-06,
+      "loss": 0.7257,
+      "step": 4904
+    },
+    {
+      "epoch": 2.4334119399280127,
+      "grad_norm": 0.13202507745720699,
+      "learning_rate": 1.668769597543381e-06,
+      "loss": 0.6851,
+      "step": 4905
+    },
+    {
+      "epoch": 2.4339084026312525,
+      "grad_norm": 0.13315852105173265,
+      "learning_rate": 1.6678479167675005e-06,
+      "loss": 0.7019,
+      "step": 4906
+    },
+    {
+      "epoch": 2.434404865334492,
+      "grad_norm": 0.1407425970112104,
+      "learning_rate": 1.6669263631727239e-06,
+      "loss": 0.7146,
+      "step": 4907
+    },
+    {
+      "epoch": 2.434901328037731,
+      "grad_norm": 0.13143176907968773,
+      "learning_rate": 1.6660049368998958e-06,
+      "loss": 0.7233,
+      "step": 4908
+    },
+    {
+      "epoch": 2.4353977907409705,
+      "grad_norm": 0.12643932478055675,
+      "learning_rate": 1.6650836380898402e-06,
+      "loss": 0.6687,
+      "step": 4909
+    },
+    {
+      "epoch": 2.43589425344421,
+      "grad_norm": 0.12854404310842657,
+      "learning_rate": 1.664162466883364e-06,
+      "loss": 0.7138,
+      "step": 4910
+    },
+    {
+      "epoch": 2.436390716147449,
+      "grad_norm": 0.1334591427590283,
+      "learning_rate": 1.663241423421253e-06,
+      "loss": 0.6816,
+      "step": 4911
+    },
+    {
+      "epoch": 2.436887178850689,
+      "grad_norm": 0.1293428646101715,
+      "learning_rate": 1.6623205078442739e-06,
+      "loss": 0.7063,
+      "step": 4912
+    },
+    {
+      "epoch": 2.4373836415539283,
+      "grad_norm": 0.13844369579181617,
+      "learning_rate": 1.6613997202931745e-06,
+      "loss": 0.7307,
+      "step": 4913
+    },
+    {
+      "epoch": 2.4378801042571676,
+      "grad_norm": 0.13252378360013214,
+      "learning_rate": 1.6604790609086818e-06,
+      "loss": 0.6987,
+      "step": 4914
+    },
+    {
+      "epoch": 2.438376566960407,
+      "grad_norm": 0.13126753203447522,
+      "learning_rate": 1.6595585298315043e-06,
+      "loss": 0.6512,
+      "step": 4915
+    },
+    {
+      "epoch": 2.4388730296636467,
+      "grad_norm": 0.13183951523709572,
+      "learning_rate": 1.6586381272023295e-06,
+      "loss": 0.7062,
+      "step": 4916
+    },
+    {
+      "epoch": 2.439369492366886,
+      "grad_norm": 0.13000918489742638,
+      "learning_rate": 1.6577178531618266e-06,
+      "loss": 0.6794,
+      "step": 4917
+    },
+    {
+      "epoch": 2.4398659550701254,
+      "grad_norm": 0.13272390835266695,
+      "learning_rate": 1.6567977078506447e-06,
+      "loss": 0.6733,
+      "step": 4918
+    },
+    {
+      "epoch": 2.4403624177733647,
+      "grad_norm": 0.13196802532399263,
+      "learning_rate": 1.6558776914094138e-06,
+      "loss": 0.6988,
+      "step": 4919
+    },
+    {
+      "epoch": 2.440858880476604,
+      "grad_norm": 0.13711894757545962,
+      "learning_rate": 1.6549578039787436e-06,
+      "loss": 0.7248,
+      "step": 4920
+    },
+    {
+      "epoch": 2.4413553431798434,
+      "grad_norm": 0.13100192331000934,
+      "learning_rate": 1.6540380456992234e-06,
+      "loss": 0.7116,
+      "step": 4921
+    },
+    {
+      "epoch": 2.441851805883083,
+      "grad_norm": 0.13106084542326454,
+      "learning_rate": 1.653118416711424e-06,
+      "loss": 0.6839,
+      "step": 4922
+    },
+    {
+      "epoch": 2.4423482685863225,
+      "grad_norm": 0.13151486257585226,
+      "learning_rate": 1.6521989171558958e-06,
+      "loss": 0.7221,
+      "step": 4923
+    },
+    {
+      "epoch": 2.442844731289562,
+      "grad_norm": 0.13197614090048165,
+      "learning_rate": 1.6512795471731702e-06,
+      "loss": 0.7504,
+      "step": 4924
+    },
+    {
+      "epoch": 2.443341193992801,
+      "grad_norm": 0.12679261823216995,
+      "learning_rate": 1.6503603069037572e-06,
+      "loss": 0.6952,
+      "step": 4925
+    },
+    {
+      "epoch": 2.443837656696041,
+      "grad_norm": 0.12936331957688477,
+      "learning_rate": 1.6494411964881482e-06,
+      "loss": 0.6724,
+      "step": 4926
+    },
+    {
+      "epoch": 2.4443341193992802,
+      "grad_norm": 0.1342595649174414,
+      "learning_rate": 1.6485222160668147e-06,
+      "loss": 0.721,
+      "step": 4927
+    },
+    {
+      "epoch": 2.4448305821025196,
+      "grad_norm": 0.13228023991573945,
+      "learning_rate": 1.6476033657802081e-06,
+      "loss": 0.7077,
+      "step": 4928
+    },
+    {
+      "epoch": 2.445327044805759,
+      "grad_norm": 0.13098966056504466,
+      "learning_rate": 1.6466846457687603e-06,
+      "loss": 0.6618,
+      "step": 4929
+    },
+    {
+      "epoch": 2.4458235075089982,
+      "grad_norm": 0.1292930748407539,
+      "learning_rate": 1.6457660561728827e-06,
+      "loss": 0.6647,
+      "step": 4930
+    },
+    {
+      "epoch": 2.4463199702122376,
+      "grad_norm": 0.13636772124500665,
+      "learning_rate": 1.6448475971329667e-06,
+      "loss": 0.7434,
+      "step": 4931
+    },
+    {
+      "epoch": 2.4468164329154773,
+      "grad_norm": 0.13511041258014103,
+      "learning_rate": 1.6439292687893838e-06,
+      "loss": 0.723,
+      "step": 4932
+    },
+    {
+      "epoch": 2.4473128956187167,
+      "grad_norm": 0.13305991113837243,
+      "learning_rate": 1.6430110712824857e-06,
+      "loss": 0.7419,
+      "step": 4933
+    },
+    {
+      "epoch": 2.447809358321956,
+      "grad_norm": 0.1284621762342019,
+      "learning_rate": 1.6420930047526048e-06,
+      "loss": 0.703,
+      "step": 4934
+    },
+    {
+      "epoch": 2.4483058210251953,
+      "grad_norm": 0.13208141055635828,
+      "learning_rate": 1.6411750693400527e-06,
+      "loss": 0.7114,
+      "step": 4935
+    },
+    {
+      "epoch": 2.448802283728435,
+      "grad_norm": 0.1334427283256501,
+      "learning_rate": 1.6402572651851217e-06,
+      "loss": 0.7029,
+      "step": 4936
+    },
+    {
+      "epoch": 2.4492987464316744,
+      "grad_norm": 0.13331062356506063,
+      "learning_rate": 1.639339592428082e-06,
+      "loss": 0.7153,
+      "step": 4937
+    },
+    {
+      "epoch": 2.4497952091349138,
+      "grad_norm": 0.13759074892380946,
+      "learning_rate": 1.638422051209186e-06,
+      "loss": 0.7888,
+      "step": 4938
+    },
+    {
+      "epoch": 2.450291671838153,
+      "grad_norm": 0.1283084309158128,
+      "learning_rate": 1.6375046416686652e-06,
+      "loss": 0.6724,
+      "step": 4939
+    },
+    {
+      "epoch": 2.4507881345413924,
+      "grad_norm": 0.13133438114077706,
+      "learning_rate": 1.6365873639467314e-06,
+      "loss": 0.7411,
+      "step": 4940
+    },
+    {
+      "epoch": 2.4512845972446318,
+      "grad_norm": 0.13217904256740817,
+      "learning_rate": 1.635670218183575e-06,
+      "loss": 0.685,
+      "step": 4941
+    },
+    {
+      "epoch": 2.4517810599478715,
+      "grad_norm": 0.13190452013229165,
+      "learning_rate": 1.6347532045193664e-06,
+      "loss": 0.7293,
+      "step": 4942
+    },
+    {
+      "epoch": 2.452277522651111,
+      "grad_norm": 0.13055191375790906,
+      "learning_rate": 1.6338363230942583e-06,
+      "loss": 0.7185,
+      "step": 4943
+    },
+    {
+      "epoch": 2.45277398535435,
+      "grad_norm": 0.13277302356989085,
+      "learning_rate": 1.6329195740483797e-06,
+      "loss": 0.746,
+      "step": 4944
+    },
+    {
+      "epoch": 2.4532704480575895,
+      "grad_norm": 0.13175605359588108,
+      "learning_rate": 1.6320029575218424e-06,
+      "loss": 0.7421,
+      "step": 4945
+    },
+    {
+      "epoch": 2.4537669107608293,
+      "grad_norm": 0.13186597595104096,
+      "learning_rate": 1.6310864736547352e-06,
+      "loss": 0.707,
+      "step": 4946
+    },
+    {
+      "epoch": 2.4542633734640686,
+      "grad_norm": 0.12766107544270525,
+      "learning_rate": 1.6301701225871297e-06,
+      "loss": 0.6824,
+      "step": 4947
+    },
+    {
+      "epoch": 2.454759836167308,
+      "grad_norm": 0.13352247929313615,
+      "learning_rate": 1.629253904459073e-06,
+      "loss": 0.716,
+      "step": 4948
+    },
+    {
+      "epoch": 2.4552562988705473,
+      "grad_norm": 0.12517152600846954,
+      "learning_rate": 1.6283378194105958e-06,
+      "loss": 0.688,
+      "step": 4949
+    },
+    {
+      "epoch": 2.4557527615737866,
+      "grad_norm": 0.1281114060367298,
+      "learning_rate": 1.627421867581707e-06,
+      "loss": 0.6663,
+      "step": 4950
+    },
+    {
+      "epoch": 2.456249224277026,
+      "grad_norm": 0.12627705318778099,
+      "learning_rate": 1.6265060491123945e-06,
+      "loss": 0.7013,
+      "step": 4951
+    },
+    {
+      "epoch": 2.4567456869802657,
+      "grad_norm": 0.1324199896128721,
+      "learning_rate": 1.6255903641426282e-06,
+      "loss": 0.725,
+      "step": 4952
+    },
+    {
+      "epoch": 2.457242149683505,
+      "grad_norm": 0.13004158500704271,
+      "learning_rate": 1.6246748128123537e-06,
+      "loss": 0.6722,
+      "step": 4953
+    },
+    {
+      "epoch": 2.4577386123867444,
+      "grad_norm": 0.13174097008700347,
+      "learning_rate": 1.6237593952614994e-06,
+      "loss": 0.6889,
+      "step": 4954
+    },
+    {
+      "epoch": 2.4582350750899837,
+      "grad_norm": 0.13318866205780763,
+      "learning_rate": 1.622844111629972e-06,
+      "loss": 0.7379,
+      "step": 4955
+    },
+    {
+      "epoch": 2.4587315377932235,
+      "grad_norm": 0.1337606359354983,
+      "learning_rate": 1.6219289620576583e-06,
+      "loss": 0.7546,
+      "step": 4956
+    },
+    {
+      "epoch": 2.459228000496463,
+      "grad_norm": 0.12844627082050522,
+      "learning_rate": 1.6210139466844244e-06,
+      "loss": 0.6996,
+      "step": 4957
+    },
+    {
+      "epoch": 2.459724463199702,
+      "grad_norm": 0.12849347327276872,
+      "learning_rate": 1.6200990656501146e-06,
+      "loss": 0.7006,
+      "step": 4958
+    },
+    {
+      "epoch": 2.4602209259029415,
+      "grad_norm": 0.12716771704033275,
+      "learning_rate": 1.6191843190945544e-06,
+      "loss": 0.7079,
+      "step": 4959
+    },
+    {
+      "epoch": 2.460717388606181,
+      "grad_norm": 0.1313359040978842,
+      "learning_rate": 1.618269707157548e-06,
+      "loss": 0.705,
+      "step": 4960
+    },
+    {
+      "epoch": 2.46121385130942,
+      "grad_norm": 0.1332264821711248,
+      "learning_rate": 1.6173552299788805e-06,
+      "loss": 0.7429,
+      "step": 4961
+    },
+    {
+      "epoch": 2.46171031401266,
+      "grad_norm": 0.12976033946371668,
+      "learning_rate": 1.616440887698313e-06,
+      "loss": 0.6886,
+      "step": 4962
+    },
+    {
+      "epoch": 2.4622067767158993,
+      "grad_norm": 0.1275893351046771,
+      "learning_rate": 1.6155266804555903e-06,
+      "loss": 0.7263,
+      "step": 4963
+    },
+    {
+      "epoch": 2.4627032394191386,
+      "grad_norm": 0.13229798618783215,
+      "learning_rate": 1.614612608390432e-06,
+      "loss": 0.7365,
+      "step": 4964
+    },
+    {
+      "epoch": 2.463199702122378,
+      "grad_norm": 0.12896939980843639,
+      "learning_rate": 1.6136986716425404e-06,
+      "loss": 0.6556,
+      "step": 4965
+    },
+    {
+      "epoch": 2.4636961648256177,
+      "grad_norm": 0.1288248754058537,
+      "learning_rate": 1.6127848703515962e-06,
+      "loss": 0.6835,
+      "step": 4966
+    },
+    {
+      "epoch": 2.464192627528857,
+      "grad_norm": 0.13426674389004808,
+      "learning_rate": 1.6118712046572587e-06,
+      "loss": 0.7412,
+      "step": 4967
+    },
+    {
+      "epoch": 2.4646890902320964,
+      "grad_norm": 0.13038176699779896,
+      "learning_rate": 1.6109576746991684e-06,
+      "loss": 0.6967,
+      "step": 4968
+    },
+    {
+      "epoch": 2.4651855529353357,
+      "grad_norm": 0.13301186253518696,
+      "learning_rate": 1.6100442806169423e-06,
+      "loss": 0.7605,
+      "step": 4969
+    },
+    {
+      "epoch": 2.465682015638575,
+      "grad_norm": 0.13134857714155998,
+      "learning_rate": 1.6091310225501782e-06,
+      "loss": 0.7556,
+      "step": 4970
+    },
+    {
+      "epoch": 2.4661784783418144,
+      "grad_norm": 0.1332594834907538,
+      "learning_rate": 1.6082179006384535e-06,
+      "loss": 0.7556,
+      "step": 4971
+    },
+    {
+      "epoch": 2.466674941045054,
+      "grad_norm": 0.1250597053615197,
+      "learning_rate": 1.6073049150213243e-06,
+      "loss": 0.6891,
+      "step": 4972
+    },
+    {
+      "epoch": 2.4671714037482935,
+      "grad_norm": 0.1321067771404992,
+      "learning_rate": 1.6063920658383258e-06,
+      "loss": 0.6851,
+      "step": 4973
+    },
+    {
+      "epoch": 2.467667866451533,
+      "grad_norm": 0.12621805146652318,
+      "learning_rate": 1.605479353228972e-06,
+      "loss": 0.6641,
+      "step": 4974
+    },
+    {
+      "epoch": 2.468164329154772,
+      "grad_norm": 0.12940412380904856,
+      "learning_rate": 1.6045667773327562e-06,
+      "loss": 0.6976,
+      "step": 4975
+    },
+    {
+      "epoch": 2.468660791858012,
+      "grad_norm": 0.13197779767436546,
+      "learning_rate": 1.6036543382891512e-06,
+      "loss": 0.7368,
+      "step": 4976
+    },
+    {
+      "epoch": 2.4691572545612512,
+      "grad_norm": 0.1322531297862413,
+      "learning_rate": 1.6027420362376092e-06,
+      "loss": 0.7481,
+      "step": 4977
+    },
+    {
+      "epoch": 2.4696537172644906,
+      "grad_norm": 0.1269834035874267,
+      "learning_rate": 1.6018298713175602e-06,
+      "loss": 0.6829,
+      "step": 4978
+    },
+    {
+      "epoch": 2.47015017996773,
+      "grad_norm": 0.13354786856232534,
+      "learning_rate": 1.6009178436684153e-06,
+      "loss": 0.7016,
+      "step": 4979
+    },
+    {
+      "epoch": 2.4706466426709692,
+      "grad_norm": 0.13062340464990232,
+      "learning_rate": 1.6000059534295614e-06,
+      "loss": 0.7193,
+      "step": 4980
+    },
+    {
+      "epoch": 2.4711431053742086,
+      "grad_norm": 0.1306702771269536,
+      "learning_rate": 1.599094200740367e-06,
+      "loss": 0.7059,
+      "step": 4981
+    },
+    {
+      "epoch": 2.4716395680774483,
+      "grad_norm": 0.13005823715313075,
+      "learning_rate": 1.5981825857401796e-06,
+      "loss": 0.6991,
+      "step": 4982
+    },
+    {
+      "epoch": 2.4721360307806877,
+      "grad_norm": 0.12974971867119356,
+      "learning_rate": 1.5972711085683241e-06,
+      "loss": 0.7102,
+      "step": 4983
+    },
+    {
+      "epoch": 2.472632493483927,
+      "grad_norm": 0.12960690860093224,
+      "learning_rate": 1.596359769364106e-06,
+      "loss": 0.6814,
+      "step": 4984
+    },
+    {
+      "epoch": 2.4731289561871663,
+      "grad_norm": 0.1264988539550879,
+      "learning_rate": 1.5954485682668075e-06,
+      "loss": 0.6807,
+      "step": 4985
+    },
+    {
+      "epoch": 2.473625418890406,
+      "grad_norm": 0.13011994137414026,
+      "learning_rate": 1.5945375054156926e-06,
+      "loss": 0.7104,
+      "step": 4986
+    },
+    {
+      "epoch": 2.4741218815936454,
+      "grad_norm": 0.14027792550347323,
+      "learning_rate": 1.5936265809500012e-06,
+      "loss": 0.7041,
+      "step": 4987
+    },
+    {
+      "epoch": 2.4746183442968848,
+      "grad_norm": 0.13203696744805918,
+      "learning_rate": 1.592715795008955e-06,
+      "loss": 0.7102,
+      "step": 4988
+    },
+    {
+      "epoch": 2.475114807000124,
+      "grad_norm": 0.1318975498818065,
+      "learning_rate": 1.5918051477317525e-06,
+      "loss": 0.7065,
+      "step": 4989
+    },
+    {
+      "epoch": 2.4756112697033634,
+      "grad_norm": 0.13322263973824197,
+      "learning_rate": 1.5908946392575713e-06,
+      "loss": 0.7667,
+      "step": 4990
+    },
+    {
+      "epoch": 2.4761077324066028,
+      "grad_norm": 0.1275784020608823,
+      "learning_rate": 1.5899842697255676e-06,
+      "loss": 0.7045,
+      "step": 4991
+    },
+    {
+      "epoch": 2.4766041951098425,
+      "grad_norm": 0.12621490198653326,
+      "learning_rate": 1.5890740392748778e-06,
+      "loss": 0.6955,
+      "step": 4992
+    },
+    {
+      "epoch": 2.477100657813082,
+      "grad_norm": 0.13147433530074085,
+      "learning_rate": 1.588163948044615e-06,
+      "loss": 0.6624,
+      "step": 4993
+    },
+    {
+      "epoch": 2.477597120516321,
+      "grad_norm": 0.12864338251047736,
+      "learning_rate": 1.587253996173873e-06,
+      "loss": 0.6439,
+      "step": 4994
+    },
+    {
+      "epoch": 2.4780935832195605,
+      "grad_norm": 0.12915557397838578,
+      "learning_rate": 1.5863441838017235e-06,
+      "loss": 0.6478,
+      "step": 4995
+    },
+    {
+      "epoch": 2.4785900459228,
+      "grad_norm": 0.13124819867045107,
+      "learning_rate": 1.585434511067216e-06,
+      "loss": 0.763,
+      "step": 4996
+    },
+    {
+      "epoch": 2.4790865086260396,
+      "grad_norm": 0.13444097971021127,
+      "learning_rate": 1.5845249781093786e-06,
+      "loss": 0.7284,
+      "step": 4997
+    },
+    {
+      "epoch": 2.479582971329279,
+      "grad_norm": 0.12597972778964714,
+      "learning_rate": 1.5836155850672202e-06,
+      "loss": 0.6775,
+      "step": 4998
+    },
+    {
+      "epoch": 2.4800794340325183,
+      "grad_norm": 0.13275658755510444,
+      "learning_rate": 1.5827063320797266e-06,
+      "loss": 0.7051,
+      "step": 4999
+    },
+    {
+      "epoch": 2.4805758967357576,
+      "grad_norm": 0.12509208961914398,
+      "learning_rate": 1.5817972192858624e-06,
+      "loss": 0.7111,
+      "step": 5000
+    },
+    {
+      "epoch": 2.481072359438997,
+      "grad_norm": 0.12842493642077443,
+      "learning_rate": 1.5808882468245706e-06,
+      "loss": 0.6731,
+      "step": 5001
+    },
+    {
+      "epoch": 2.4815688221422367,
+      "grad_norm": 0.13994016275679041,
+      "learning_rate": 1.5799794148347738e-06,
+      "loss": 0.7332,
+      "step": 5002
+    },
+    {
+      "epoch": 2.482065284845476,
+      "grad_norm": 0.13160007453858913,
+      "learning_rate": 1.579070723455372e-06,
+      "loss": 0.7184,
+      "step": 5003
+    },
+    {
+      "epoch": 2.4825617475487154,
+      "grad_norm": 0.12476420674542148,
+      "learning_rate": 1.5781621728252439e-06,
+      "loss": 0.7323,
+      "step": 5004
+    },
+    {
+      "epoch": 2.4830582102519547,
+      "grad_norm": 0.13342172172903713,
+      "learning_rate": 1.5772537630832477e-06,
+      "loss": 0.6979,
+      "step": 5005
+    },
+    {
+      "epoch": 2.483554672955194,
+      "grad_norm": 0.13261519543470868,
+      "learning_rate": 1.5763454943682183e-06,
+      "loss": 0.6798,
+      "step": 5006
+    },
+    {
+      "epoch": 2.484051135658434,
+      "grad_norm": 0.13196930690487815,
+      "learning_rate": 1.5754373668189703e-06,
+      "loss": 0.6926,
+      "step": 5007
+    },
+    {
+      "epoch": 2.484547598361673,
+      "grad_norm": 0.1284437605438022,
+      "learning_rate": 1.5745293805742968e-06,
+      "loss": 0.7104,
+      "step": 5008
+    },
+    {
+      "epoch": 2.4850440610649125,
+      "grad_norm": 0.1271584061825038,
+      "learning_rate": 1.5736215357729683e-06,
+      "loss": 0.7149,
+      "step": 5009
+    },
+    {
+      "epoch": 2.485540523768152,
+      "grad_norm": 0.12986340829886878,
+      "learning_rate": 1.5727138325537355e-06,
+      "loss": 0.7077,
+      "step": 5010
+    },
+    {
+      "epoch": 2.486036986471391,
+      "grad_norm": 0.13788400838359663,
+      "learning_rate": 1.5718062710553253e-06,
+      "loss": 0.751,
+      "step": 5011
+    },
+    {
+      "epoch": 2.486533449174631,
+      "grad_norm": 0.1348639652151765,
+      "learning_rate": 1.5708988514164442e-06,
+      "loss": 0.7636,
+      "step": 5012
+    },
+    {
+      "epoch": 2.4870299118778703,
+      "grad_norm": 0.13129936769800069,
+      "learning_rate": 1.569991573775776e-06,
+      "loss": 0.6946,
+      "step": 5013
+    },
+    {
+      "epoch": 2.4875263745811096,
+      "grad_norm": 0.131016067710159,
+      "learning_rate": 1.5690844382719844e-06,
+      "loss": 0.7299,
+      "step": 5014
+    },
+    {
+      "epoch": 2.488022837284349,
+      "grad_norm": 0.13469638317810959,
+      "learning_rate": 1.5681774450437104e-06,
+      "loss": 0.7123,
+      "step": 5015
+    },
+    {
+      "epoch": 2.4885192999875883,
+      "grad_norm": 0.12704079453926578,
+      "learning_rate": 1.5672705942295735e-06,
+      "loss": 0.6537,
+      "step": 5016
+    },
+    {
+      "epoch": 2.489015762690828,
+      "grad_norm": 0.12753576550951762,
+      "learning_rate": 1.5663638859681706e-06,
+      "loss": 0.7195,
+      "step": 5017
+    },
+    {
+      "epoch": 2.4895122253940674,
+      "grad_norm": 0.13049316211920853,
+      "learning_rate": 1.5654573203980782e-06,
+      "loss": 0.7029,
+      "step": 5018
+    },
+    {
+      "epoch": 2.4900086880973067,
+      "grad_norm": 0.12674348845812947,
+      "learning_rate": 1.5645508976578501e-06,
+      "loss": 0.6749,
+      "step": 5019
+    },
+    {
+      "epoch": 2.490505150800546,
+      "grad_norm": 0.13119282880613511,
+      "learning_rate": 1.5636446178860188e-06,
+      "loss": 0.7071,
+      "step": 5020
+    },
+    {
+      "epoch": 2.4910016135037854,
+      "grad_norm": 0.12857653431579513,
+      "learning_rate": 1.5627384812210945e-06,
+      "loss": 0.7098,
+      "step": 5021
+    },
+    {
+      "epoch": 2.491498076207025,
+      "grad_norm": 0.1299396609011369,
+      "learning_rate": 1.561832487801565e-06,
+      "loss": 0.7165,
+      "step": 5022
+    },
+    {
+      "epoch": 2.4919945389102645,
+      "grad_norm": 0.13187761577665433,
+      "learning_rate": 1.560926637765897e-06,
+      "loss": 0.6788,
+      "step": 5023
+    },
+    {
+      "epoch": 2.492491001613504,
+      "grad_norm": 0.13676658429929037,
+      "learning_rate": 1.560020931252536e-06,
+      "loss": 0.774,
+      "step": 5024
+    },
+    {
+      "epoch": 2.492987464316743,
+      "grad_norm": 0.2391281093395671,
+      "learning_rate": 1.5591153683999045e-06,
+      "loss": 0.7162,
+      "step": 5025
+    },
+    {
+      "epoch": 2.4934839270199824,
+      "grad_norm": 0.1321280807778348,
+      "learning_rate": 1.5582099493464032e-06,
+      "loss": 0.7193,
+      "step": 5026
+    },
+    {
+      "epoch": 2.4939803897232222,
+      "grad_norm": 0.12836015141870605,
+      "learning_rate": 1.5573046742304104e-06,
+      "loss": 0.7098,
+      "step": 5027
+    },
+    {
+      "epoch": 2.4944768524264616,
+      "grad_norm": 0.12731919945775716,
+      "learning_rate": 1.5563995431902834e-06,
+      "loss": 0.7006,
+      "step": 5028
+    },
+    {
+      "epoch": 2.494973315129701,
+      "grad_norm": 0.129683691750222,
+      "learning_rate": 1.5554945563643564e-06,
+      "loss": 0.7271,
+      "step": 5029
+    },
+    {
+      "epoch": 2.49546977783294,
+      "grad_norm": 0.12619265071805652,
+      "learning_rate": 1.5545897138909423e-06,
+      "loss": 0.7269,
+      "step": 5030
+    },
+    {
+      "epoch": 2.4959662405361795,
+      "grad_norm": 0.12864219160660795,
+      "learning_rate": 1.5536850159083319e-06,
+      "loss": 0.7173,
+      "step": 5031
+    },
+    {
+      "epoch": 2.4964627032394193,
+      "grad_norm": 0.12745661281840148,
+      "learning_rate": 1.5527804625547937e-06,
+      "loss": 0.6913,
+      "step": 5032
+    },
+    {
+      "epoch": 2.4969591659426587,
+      "grad_norm": 0.12858809922709444,
+      "learning_rate": 1.5518760539685752e-06,
+      "loss": 0.7134,
+      "step": 5033
+    },
+    {
+      "epoch": 2.497455628645898,
+      "grad_norm": 0.12827890445554876,
+      "learning_rate": 1.550971790287899e-06,
+      "loss": 0.6648,
+      "step": 5034
+    },
+    {
+      "epoch": 2.4979520913491373,
+      "grad_norm": 0.13300762899595023,
+      "learning_rate": 1.5500676716509683e-06,
+      "loss": 0.7096,
+      "step": 5035
+    },
+    {
+      "epoch": 2.4984485540523766,
+      "grad_norm": 0.13329585467541116,
+      "learning_rate": 1.549163698195963e-06,
+      "loss": 0.739,
+      "step": 5036
+    },
+    {
+      "epoch": 2.4989450167556164,
+      "grad_norm": 0.13351068526158444,
+      "learning_rate": 1.5482598700610412e-06,
+      "loss": 0.7359,
+      "step": 5037
+    },
+    {
+      "epoch": 2.4994414794588558,
+      "grad_norm": 0.13746974435710943,
+      "learning_rate": 1.5473561873843378e-06,
+      "loss": 0.7429,
+      "step": 5038
+    },
+    {
+      "epoch": 2.499937942162095,
+      "grad_norm": 0.1311946484491049,
+      "learning_rate": 1.5464526503039668e-06,
+      "loss": 0.7333,
+      "step": 5039
+    },
+    {
+      "epoch": 2.5004344048653344,
+      "grad_norm": 0.13003936186444456,
+      "learning_rate": 1.5455492589580195e-06,
+      "loss": 0.6862,
+      "step": 5040
+    },
+    {
+      "epoch": 2.5004344048653344,
+      "eval_loss": 0.7251803278923035,
+      "eval_runtime": 135.8229,
+      "eval_samples_per_second": 223.475,
+      "eval_steps_per_second": 27.941,
+      "step": 5040
+    },
+    {
+      "epoch": 2.5009308675685737,
+      "grad_norm": 0.12908028509507846,
+      "learning_rate": 1.5446460134845642e-06,
+      "loss": 0.6778,
+      "step": 5041
+    },
+    {
+      "epoch": 2.501427330271813,
+      "grad_norm": 0.13166843080491383,
+      "learning_rate": 1.543742914021648e-06,
+      "loss": 0.7092,
+      "step": 5042
+    },
+    {
+      "epoch": 2.501923792975053,
+      "grad_norm": 0.13088243606617905,
+      "learning_rate": 1.5428399607072956e-06,
+      "loss": 0.7373,
+      "step": 5043
+    },
+    {
+      "epoch": 2.502420255678292,
+      "grad_norm": 0.12629275270854762,
+      "learning_rate": 1.541937153679508e-06,
+      "loss": 0.708,
+      "step": 5044
+    },
+    {
+      "epoch": 2.5029167183815315,
+      "grad_norm": 0.13225143684011295,
+      "learning_rate": 1.5410344930762648e-06,
+      "loss": 0.701,
+      "step": 5045
+    },
+    {
+      "epoch": 2.503413181084771,
+      "grad_norm": 0.13340939597155863,
+      "learning_rate": 1.5401319790355232e-06,
+      "loss": 0.7098,
+      "step": 5046
+    },
+    {
+      "epoch": 2.5039096437880106,
+      "grad_norm": 0.13043769011930573,
+      "learning_rate": 1.5392296116952181e-06,
+      "loss": 0.6952,
+      "step": 5047
+    },
+    {
+      "epoch": 2.50440610649125,
+      "grad_norm": 0.13682071863988435,
+      "learning_rate": 1.5383273911932627e-06,
+      "loss": 0.7542,
+      "step": 5048
+    },
+    {
+      "epoch": 2.5049025691944893,
+      "grad_norm": 0.13032047556352128,
+      "learning_rate": 1.5374253176675464e-06,
+      "loss": 0.6926,
+      "step": 5049
+    },
+    {
+      "epoch": 2.5053990318977286,
+      "grad_norm": 0.13244019431180845,
+      "learning_rate": 1.5365233912559364e-06,
+      "loss": 0.7175,
+      "step": 5050
+    },
+    {
+      "epoch": 2.505895494600968,
+      "grad_norm": 0.13817001240133622,
+      "learning_rate": 1.5356216120962774e-06,
+      "loss": 0.7282,
+      "step": 5051
+    },
+    {
+      "epoch": 2.5063919573042073,
+      "grad_norm": 0.13140272145786022,
+      "learning_rate": 1.5347199803263927e-06,
+      "loss": 0.6855,
+      "step": 5052
+    },
+    {
+      "epoch": 2.506888420007447,
+      "grad_norm": 0.1360933004526906,
+      "learning_rate": 1.5338184960840824e-06,
+      "loss": 0.7409,
+      "step": 5053
+    },
+    {
+      "epoch": 2.5073848827106864,
+      "grad_norm": 0.13091513330256985,
+      "learning_rate": 1.5329171595071227e-06,
+      "loss": 0.6919,
+      "step": 5054
+    },
+    {
+      "epoch": 2.5078813454139257,
+      "grad_norm": 0.1265533074222129,
+      "learning_rate": 1.5320159707332695e-06,
+      "loss": 0.6243,
+      "step": 5055
+    },
+    {
+      "epoch": 2.508377808117165,
+      "grad_norm": 0.1252099518570422,
+      "learning_rate": 1.5311149299002542e-06,
+      "loss": 0.6715,
+      "step": 5056
+    },
+    {
+      "epoch": 2.508874270820405,
+      "grad_norm": 0.13609272644415313,
+      "learning_rate": 1.5302140371457875e-06,
+      "loss": 0.732,
+      "step": 5057
+    },
+    {
+      "epoch": 2.509370733523644,
+      "grad_norm": 0.13323577694421218,
+      "learning_rate": 1.529313292607556e-06,
+      "loss": 0.6702,
+      "step": 5058
+    },
+    {
+      "epoch": 2.5098671962268835,
+      "grad_norm": 0.13041020040639237,
+      "learning_rate": 1.5284126964232244e-06,
+      "loss": 0.7071,
+      "step": 5059
+    },
+    {
+      "epoch": 2.510363658930123,
+      "grad_norm": 0.1304162413787588,
+      "learning_rate": 1.5275122487304337e-06,
+      "loss": 0.7312,
+      "step": 5060
+    },
+    {
+      "epoch": 2.510860121633362,
+      "grad_norm": 0.13021570150444117,
+      "learning_rate": 1.5266119496668025e-06,
+      "loss": 0.7283,
+      "step": 5061
+    },
+    {
+      "epoch": 2.5113565843366015,
+      "grad_norm": 0.1279339703409767,
+      "learning_rate": 1.5257117993699276e-06,
+      "loss": 0.696,
+      "step": 5062
+    },
+    {
+      "epoch": 2.5118530470398412,
+      "grad_norm": 0.13240644628046436,
+      "learning_rate": 1.524811797977383e-06,
+      "loss": 0.6864,
+      "step": 5063
+    },
+    {
+      "epoch": 2.5123495097430806,
+      "grad_norm": 0.1400762626467057,
+      "learning_rate": 1.523911945626719e-06,
+      "loss": 0.7678,
+      "step": 5064
+    },
+    {
+      "epoch": 2.51284597244632,
+      "grad_norm": 0.13014151678535776,
+      "learning_rate": 1.5230122424554644e-06,
+      "loss": 0.7036,
+      "step": 5065
+    },
+    {
+      "epoch": 2.5133424351495592,
+      "grad_norm": 0.12985990176866047,
+      "learning_rate": 1.5221126886011228e-06,
+      "loss": 0.7224,
+      "step": 5066
+    },
+    {
+      "epoch": 2.513838897852799,
+      "grad_norm": 0.1317976170535492,
+      "learning_rate": 1.5212132842011778e-06,
+      "loss": 0.6924,
+      "step": 5067
+    },
+    {
+      "epoch": 2.5143353605560383,
+      "grad_norm": 0.12970897949094737,
+      "learning_rate": 1.5203140293930888e-06,
+      "loss": 0.6995,
+      "step": 5068
+    },
+    {
+      "epoch": 2.5148318232592777,
+      "grad_norm": 0.1303319284929338,
+      "learning_rate": 1.519414924314292e-06,
+      "loss": 0.7018,
+      "step": 5069
+    },
+    {
+      "epoch": 2.515328285962517,
+      "grad_norm": 0.12850755690334908,
+      "learning_rate": 1.5185159691022023e-06,
+      "loss": 0.7186,
+      "step": 5070
+    },
+    {
+      "epoch": 2.5158247486657563,
+      "grad_norm": 0.12718389440933345,
+      "learning_rate": 1.517617163894209e-06,
+      "loss": 0.6397,
+      "step": 5071
+    },
+    {
+      "epoch": 2.5163212113689957,
+      "grad_norm": 0.13884217395526113,
+      "learning_rate": 1.5167185088276815e-06,
+      "loss": 0.665,
+      "step": 5072
+    },
+    {
+      "epoch": 2.5168176740722354,
+      "grad_norm": 0.12603973119926584,
+      "learning_rate": 1.5158200040399635e-06,
+      "loss": 0.6631,
+      "step": 5073
+    },
+    {
+      "epoch": 2.5173141367754748,
+      "grad_norm": 0.1293877404461229,
+      "learning_rate": 1.5149216496683788e-06,
+      "loss": 0.7018,
+      "step": 5074
+    },
+    {
+      "epoch": 2.517810599478714,
+      "grad_norm": 0.12959977324157554,
+      "learning_rate": 1.514023445850225e-06,
+      "loss": 0.6999,
+      "step": 5075
+    },
+    {
+      "epoch": 2.5183070621819534,
+      "grad_norm": 0.13078603859075733,
+      "learning_rate": 1.513125392722779e-06,
+      "loss": 0.6854,
+      "step": 5076
+    },
+    {
+      "epoch": 2.518803524885193,
+      "grad_norm": 0.13366544771971153,
+      "learning_rate": 1.5122274904232925e-06,
+      "loss": 0.7107,
+      "step": 5077
+    },
+    {
+      "epoch": 2.5192999875884325,
+      "grad_norm": 0.1332349279647867,
+      "learning_rate": 1.5113297390889963e-06,
+      "loss": 0.7199,
+      "step": 5078
+    },
+    {
+      "epoch": 2.519796450291672,
+      "grad_norm": 0.12791951008145022,
+      "learning_rate": 1.5104321388570977e-06,
+      "loss": 0.6795,
+      "step": 5079
+    },
+    {
+      "epoch": 2.520292912994911,
+      "grad_norm": 0.1356457836131563,
+      "learning_rate": 1.50953468986478e-06,
+      "loss": 0.6902,
+      "step": 5080
+    },
+    {
+      "epoch": 2.5207893756981505,
+      "grad_norm": 0.12588538763738047,
+      "learning_rate": 1.5086373922492049e-06,
+      "loss": 0.6544,
+      "step": 5081
+    },
+    {
+      "epoch": 2.52128583840139,
+      "grad_norm": 0.1330568928734858,
+      "learning_rate": 1.5077402461475083e-06,
+      "loss": 0.6666,
+      "step": 5082
+    },
+    {
+      "epoch": 2.5217823011046296,
+      "grad_norm": 0.1225633830068647,
+      "learning_rate": 1.506843251696805e-06,
+      "loss": 0.6217,
+      "step": 5083
+    },
+    {
+      "epoch": 2.522278763807869,
+      "grad_norm": 0.12995958837521615,
+      "learning_rate": 1.505946409034187e-06,
+      "loss": 0.6922,
+      "step": 5084
+    },
+    {
+      "epoch": 2.5227752265111083,
+      "grad_norm": 0.12735849858705797,
+      "learning_rate": 1.5050497182967221e-06,
+      "loss": 0.6844,
+      "step": 5085
+    },
+    {
+      "epoch": 2.5232716892143476,
+      "grad_norm": 0.128231127592946,
+      "learning_rate": 1.5041531796214554e-06,
+      "loss": 0.7062,
+      "step": 5086
+    },
+    {
+      "epoch": 2.5237681519175874,
+      "grad_norm": 0.12711845362503602,
+      "learning_rate": 1.5032567931454073e-06,
+      "loss": 0.6898,
+      "step": 5087
+    },
+    {
+      "epoch": 2.5242646146208267,
+      "grad_norm": 0.13608076036992192,
+      "learning_rate": 1.5023605590055768e-06,
+      "loss": 0.691,
+      "step": 5088
+    },
+    {
+      "epoch": 2.524761077324066,
+      "grad_norm": 0.1301368936095168,
+      "learning_rate": 1.5014644773389391e-06,
+      "loss": 0.7156,
+      "step": 5089
+    },
+    {
+      "epoch": 2.5252575400273054,
+      "grad_norm": 0.13293944449573997,
+      "learning_rate": 1.5005685482824462e-06,
+      "loss": 0.7482,
+      "step": 5090
+    },
+    {
+      "epoch": 2.5257540027305447,
+      "grad_norm": 0.12733134898732534,
+      "learning_rate": 1.499672771973026e-06,
+      "loss": 0.6956,
+      "step": 5091
+    },
+    {
+      "epoch": 2.526250465433784,
+      "grad_norm": 0.12354114386385891,
+      "learning_rate": 1.4987771485475836e-06,
+      "loss": 0.6575,
+      "step": 5092
+    },
+    {
+      "epoch": 2.526746928137024,
+      "grad_norm": 0.1256909063925675,
+      "learning_rate": 1.497881678143e-06,
+      "loss": 0.6983,
+      "step": 5093
+    },
+    {
+      "epoch": 2.527243390840263,
+      "grad_norm": 0.12995383983928524,
+      "learning_rate": 1.4969863608961343e-06,
+      "loss": 0.7182,
+      "step": 5094
+    },
+    {
+      "epoch": 2.5277398535435025,
+      "grad_norm": 0.13104868294903146,
+      "learning_rate": 1.4960911969438213e-06,
+      "loss": 0.7503,
+      "step": 5095
+    },
+    {
+      "epoch": 2.528236316246742,
+      "grad_norm": 0.13569502975102957,
+      "learning_rate": 1.495196186422872e-06,
+      "loss": 0.6806,
+      "step": 5096
+    },
+    {
+      "epoch": 2.5287327789499816,
+      "grad_norm": 0.13339083267038424,
+      "learning_rate": 1.4943013294700758e-06,
+      "loss": 0.7056,
+      "step": 5097
+    },
+    {
+      "epoch": 2.529229241653221,
+      "grad_norm": 0.13361409013339606,
+      "learning_rate": 1.4934066262221954e-06,
+      "loss": 0.7391,
+      "step": 5098
+    },
+    {
+      "epoch": 2.5297257043564603,
+      "grad_norm": 0.12680487158505377,
+      "learning_rate": 1.492512076815973e-06,
+      "loss": 0.6482,
+      "step": 5099
+    },
+    {
+      "epoch": 2.5302221670596996,
+      "grad_norm": 0.12746892302880983,
+      "learning_rate": 1.4916176813881257e-06,
+      "loss": 0.6522,
+      "step": 5100
+    },
+    {
+      "epoch": 2.530718629762939,
+      "grad_norm": 0.13139044491664464,
+      "learning_rate": 1.4907234400753473e-06,
+      "loss": 0.712,
+      "step": 5101
+    },
+    {
+      "epoch": 2.5312150924661783,
+      "grad_norm": 0.12799076572044088,
+      "learning_rate": 1.4898293530143095e-06,
+      "loss": 0.6722,
+      "step": 5102
+    },
+    {
+      "epoch": 2.531711555169418,
+      "grad_norm": 0.12797705381984992,
+      "learning_rate": 1.4889354203416575e-06,
+      "loss": 0.624,
+      "step": 5103
+    },
+    {
+      "epoch": 2.5322080178726574,
+      "grad_norm": 0.1355225972612351,
+      "learning_rate": 1.4880416421940155e-06,
+      "loss": 0.7376,
+      "step": 5104
+    },
+    {
+      "epoch": 2.5327044805758967,
+      "grad_norm": 0.1294579579051531,
+      "learning_rate": 1.4871480187079828e-06,
+      "loss": 0.7096,
+      "step": 5105
+    },
+    {
+      "epoch": 2.533200943279136,
+      "grad_norm": 0.13381112874931977,
+      "learning_rate": 1.4862545500201358e-06,
+      "loss": 0.7277,
+      "step": 5106
+    },
+    {
+      "epoch": 2.533697405982376,
+      "grad_norm": 0.1246715734533751,
+      "learning_rate": 1.4853612362670271e-06,
+      "loss": 0.6442,
+      "step": 5107
+    },
+    {
+      "epoch": 2.534193868685615,
+      "grad_norm": 0.12830907556578108,
+      "learning_rate": 1.4844680775851846e-06,
+      "loss": 0.6337,
+      "step": 5108
+    },
+    {
+      "epoch": 2.5346903313888545,
+      "grad_norm": 0.1278448286390337,
+      "learning_rate": 1.4835750741111139e-06,
+      "loss": 0.7132,
+      "step": 5109
+    },
+    {
+      "epoch": 2.535186794092094,
+      "grad_norm": 0.1335432788354155,
+      "learning_rate": 1.4826822259812957e-06,
+      "loss": 0.6988,
+      "step": 5110
+    },
+    {
+      "epoch": 2.535683256795333,
+      "grad_norm": 0.134293152215964,
+      "learning_rate": 1.4817895333321875e-06,
+      "loss": 0.7336,
+      "step": 5111
+    },
+    {
+      "epoch": 2.5361797194985725,
+      "grad_norm": 0.13777460933447475,
+      "learning_rate": 1.4808969963002234e-06,
+      "loss": 0.7112,
+      "step": 5112
+    },
+    {
+      "epoch": 2.5366761822018122,
+      "grad_norm": 0.12897145890973283,
+      "learning_rate": 1.480004615021814e-06,
+      "loss": 0.6908,
+      "step": 5113
+    },
+    {
+      "epoch": 2.5371726449050516,
+      "grad_norm": 0.13400133466081912,
+      "learning_rate": 1.4791123896333438e-06,
+      "loss": 0.7064,
+      "step": 5114
+    },
+    {
+      "epoch": 2.537669107608291,
+      "grad_norm": 0.13211791906308365,
+      "learning_rate": 1.4782203202711764e-06,
+      "loss": 0.7026,
+      "step": 5115
+    },
+    {
+      "epoch": 2.5381655703115302,
+      "grad_norm": 0.13139779720721778,
+      "learning_rate": 1.4773284070716504e-06,
+      "loss": 0.7138,
+      "step": 5116
+    },
+    {
+      "epoch": 2.53866203301477,
+      "grad_norm": 0.13308362459519374,
+      "learning_rate": 1.4764366501710798e-06,
+      "loss": 0.7263,
+      "step": 5117
+    },
+    {
+      "epoch": 2.5391584957180093,
+      "grad_norm": 0.13141118126729773,
+      "learning_rate": 1.4755450497057563e-06,
+      "loss": 0.678,
+      "step": 5118
+    },
+    {
+      "epoch": 2.5396549584212487,
+      "grad_norm": 0.13186502701805516,
+      "learning_rate": 1.4746536058119454e-06,
+      "loss": 0.7602,
+      "step": 5119
+    },
+    {
+      "epoch": 2.540151421124488,
+      "grad_norm": 0.12869479855496058,
+      "learning_rate": 1.473762318625891e-06,
+      "loss": 0.7195,
+      "step": 5120
+    },
+    {
+      "epoch": 2.5406478838277273,
+      "grad_norm": 0.13445753218214146,
+      "learning_rate": 1.4728711882838115e-06,
+      "loss": 0.7307,
+      "step": 5121
+    },
+    {
+      "epoch": 2.5411443465309667,
+      "grad_norm": 0.12705563245150678,
+      "learning_rate": 1.471980214921902e-06,
+      "loss": 0.6385,
+      "step": 5122
+    },
+    {
+      "epoch": 2.5416408092342064,
+      "grad_norm": 0.13028590027975537,
+      "learning_rate": 1.4710893986763347e-06,
+      "loss": 0.705,
+      "step": 5123
+    },
+    {
+      "epoch": 2.5421372719374458,
+      "grad_norm": 0.12817908251977725,
+      "learning_rate": 1.4701987396832546e-06,
+      "loss": 0.7177,
+      "step": 5124
+    },
+    {
+      "epoch": 2.542633734640685,
+      "grad_norm": 0.12496892532709507,
+      "learning_rate": 1.4693082380787858e-06,
+      "loss": 0.6705,
+      "step": 5125
+    },
+    {
+      "epoch": 2.5431301973439244,
+      "grad_norm": 0.13059250706457962,
+      "learning_rate": 1.4684178939990264e-06,
+      "loss": 0.691,
+      "step": 5126
+    },
+    {
+      "epoch": 2.543626660047164,
+      "grad_norm": 0.1292904883407009,
+      "learning_rate": 1.467527707580052e-06,
+      "loss": 0.7285,
+      "step": 5127
+    },
+    {
+      "epoch": 2.5441231227504035,
+      "grad_norm": 0.12910095210438902,
+      "learning_rate": 1.466637678957913e-06,
+      "loss": 0.6785,
+      "step": 5128
+    },
+    {
+      "epoch": 2.544619585453643,
+      "grad_norm": 0.13231894500519423,
+      "learning_rate": 1.4657478082686363e-06,
+      "loss": 0.734,
+      "step": 5129
+    },
+    {
+      "epoch": 2.545116048156882,
+      "grad_norm": 0.13354137515536624,
+      "learning_rate": 1.4648580956482238e-06,
+      "loss": 0.6792,
+      "step": 5130
+    },
+    {
+      "epoch": 2.5456125108601215,
+      "grad_norm": 0.13437657602685885,
+      "learning_rate": 1.4639685412326543e-06,
+      "loss": 0.6869,
+      "step": 5131
+    },
+    {
+      "epoch": 2.546108973563361,
+      "grad_norm": 0.13419591499955405,
+      "learning_rate": 1.463079145157882e-06,
+      "loss": 0.7086,
+      "step": 5132
+    },
+    {
+      "epoch": 2.5466054362666006,
+      "grad_norm": 0.1317550238055129,
+      "learning_rate": 1.462189907559836e-06,
+      "loss": 0.7011,
+      "step": 5133
+    },
+    {
+      "epoch": 2.54710189896984,
+      "grad_norm": 0.12919424179545982,
+      "learning_rate": 1.4613008285744234e-06,
+      "loss": 0.7132,
+      "step": 5134
+    },
+    {
+      "epoch": 2.5475983616730793,
+      "grad_norm": 0.13156727464529705,
+      "learning_rate": 1.4604119083375242e-06,
+      "loss": 0.6697,
+      "step": 5135
+    },
+    {
+      "epoch": 2.5480948243763186,
+      "grad_norm": 0.13198996730417278,
+      "learning_rate": 1.4595231469849963e-06,
+      "loss": 0.6547,
+      "step": 5136
+    },
+    {
+      "epoch": 2.5485912870795584,
+      "grad_norm": 0.13504489395934224,
+      "learning_rate": 1.4586345446526735e-06,
+      "loss": 0.7082,
+      "step": 5137
+    },
+    {
+      "epoch": 2.5490877497827977,
+      "grad_norm": 0.13159429183393376,
+      "learning_rate": 1.457746101476362e-06,
+      "loss": 0.7077,
+      "step": 5138
+    },
+    {
+      "epoch": 2.549584212486037,
+      "grad_norm": 0.1324983434375749,
+      "learning_rate": 1.4568578175918502e-06,
+      "loss": 0.6928,
+      "step": 5139
+    },
+    {
+      "epoch": 2.5500806751892764,
+      "grad_norm": 0.13854603478769614,
+      "learning_rate": 1.455969693134893e-06,
+      "loss": 0.7396,
+      "step": 5140
+    },
+    {
+      "epoch": 2.5505771378925157,
+      "grad_norm": 0.12842830146725112,
+      "learning_rate": 1.4550817282412293e-06,
+      "loss": 0.7549,
+      "step": 5141
+    },
+    {
+      "epoch": 2.551073600595755,
+      "grad_norm": 0.129204728717719,
+      "learning_rate": 1.454193923046569e-06,
+      "loss": 0.7016,
+      "step": 5142
+    },
+    {
+      "epoch": 2.551570063298995,
+      "grad_norm": 0.12750868562867726,
+      "learning_rate": 1.4533062776866002e-06,
+      "loss": 0.6743,
+      "step": 5143
+    },
+    {
+      "epoch": 2.552066526002234,
+      "grad_norm": 0.12654769018231987,
+      "learning_rate": 1.452418792296984e-06,
+      "loss": 0.7008,
+      "step": 5144
+    },
+    {
+      "epoch": 2.5525629887054735,
+      "grad_norm": 0.13141605016356414,
+      "learning_rate": 1.4515314670133582e-06,
+      "loss": 0.7169,
+      "step": 5145
+    },
+    {
+      "epoch": 2.553059451408713,
+      "grad_norm": 0.13379502140501737,
+      "learning_rate": 1.4506443019713374e-06,
+      "loss": 0.6779,
+      "step": 5146
+    },
+    {
+      "epoch": 2.5535559141119526,
+      "grad_norm": 0.13256467196234833,
+      "learning_rate": 1.4497572973065091e-06,
+      "loss": 0.7206,
+      "step": 5147
+    },
+    {
+      "epoch": 2.554052376815192,
+      "grad_norm": 0.13184270552834443,
+      "learning_rate": 1.4488704531544396e-06,
+      "loss": 0.6414,
+      "step": 5148
+    },
+    {
+      "epoch": 2.5545488395184313,
+      "grad_norm": 0.12861101671614525,
+      "learning_rate": 1.4479837696506677e-06,
+      "loss": 0.7362,
+      "step": 5149
+    },
+    {
+      "epoch": 2.5550453022216706,
+      "grad_norm": 0.1262895343147098,
+      "learning_rate": 1.4470972469307076e-06,
+      "loss": 0.7061,
+      "step": 5150
+    },
+    {
+      "epoch": 2.55554176492491,
+      "grad_norm": 0.12872834020768845,
+      "learning_rate": 1.4462108851300524e-06,
+      "loss": 0.7172,
+      "step": 5151
+    },
+    {
+      "epoch": 2.5560382276281493,
+      "grad_norm": 0.13269323839535846,
+      "learning_rate": 1.4453246843841662e-06,
+      "loss": 0.7117,
+      "step": 5152
+    },
+    {
+      "epoch": 2.5565346903313886,
+      "grad_norm": 0.12885571431373338,
+      "learning_rate": 1.4444386448284925e-06,
+      "loss": 0.6573,
+      "step": 5153
+    },
+    {
+      "epoch": 2.5570311530346284,
+      "grad_norm": 0.13280071499296228,
+      "learning_rate": 1.4435527665984474e-06,
+      "loss": 0.6934,
+      "step": 5154
+    },
+    {
+      "epoch": 2.5575276157378677,
+      "grad_norm": 0.13454797554586437,
+      "learning_rate": 1.442667049829422e-06,
+      "loss": 0.7246,
+      "step": 5155
+    },
+    {
+      "epoch": 2.558024078441107,
+      "grad_norm": 0.12697947714667507,
+      "learning_rate": 1.4417814946567873e-06,
+      "loss": 0.7333,
+      "step": 5156
+    },
+    {
+      "epoch": 2.558520541144347,
+      "grad_norm": 0.12516155676293964,
+      "learning_rate": 1.4408961012158818e-06,
+      "loss": 0.6683,
+      "step": 5157
+    },
+    {
+      "epoch": 2.559017003847586,
+      "grad_norm": 0.1302587030682515,
+      "learning_rate": 1.4400108696420265e-06,
+      "loss": 0.6953,
+      "step": 5158
+    },
+    {
+      "epoch": 2.5595134665508255,
+      "grad_norm": 0.12649071492393144,
+      "learning_rate": 1.4391258000705143e-06,
+      "loss": 0.6955,
+      "step": 5159
+    },
+    {
+      "epoch": 2.560009929254065,
+      "grad_norm": 0.1319214562476525,
+      "learning_rate": 1.4382408926366125e-06,
+      "loss": 0.7035,
+      "step": 5160
+    },
+    {
+      "epoch": 2.560506391957304,
+      "grad_norm": 0.12479301266102043,
+      "learning_rate": 1.4373561474755675e-06,
+      "loss": 0.6584,
+      "step": 5161
+    },
+    {
+      "epoch": 2.5610028546605434,
+      "grad_norm": 0.12302337495146001,
+      "learning_rate": 1.4364715647225963e-06,
+      "loss": 0.6872,
+      "step": 5162
+    },
+    {
+      "epoch": 2.561499317363783,
+      "grad_norm": 0.1313391656279115,
+      "learning_rate": 1.435587144512895e-06,
+      "loss": 0.7152,
+      "step": 5163
+    },
+    {
+      "epoch": 2.5619957800670226,
+      "grad_norm": 0.13419103025976656,
+      "learning_rate": 1.434702886981632e-06,
+      "loss": 0.707,
+      "step": 5164
+    },
+    {
+      "epoch": 2.562492242770262,
+      "grad_norm": 0.12781224085601264,
+      "learning_rate": 1.4338187922639506e-06,
+      "loss": 0.6512,
+      "step": 5165
+    },
+    {
+      "epoch": 2.562988705473501,
+      "grad_norm": 0.1269467241970245,
+      "learning_rate": 1.4329348604949733e-06,
+      "loss": 0.7062,
+      "step": 5166
+    },
+    {
+      "epoch": 2.563485168176741,
+      "grad_norm": 0.13011475251943433,
+      "learning_rate": 1.4320510918097927e-06,
+      "loss": 0.7054,
+      "step": 5167
+    },
+    {
+      "epoch": 2.5639816308799803,
+      "grad_norm": 0.12562228736537093,
+      "learning_rate": 1.4311674863434803e-06,
+      "loss": 0.682,
+      "step": 5168
+    },
+    {
+      "epoch": 2.5644780935832197,
+      "grad_norm": 0.1280244371960763,
+      "learning_rate": 1.43028404423108e-06,
+      "loss": 0.701,
+      "step": 5169
+    },
+    {
+      "epoch": 2.564974556286459,
+      "grad_norm": 0.12628433888008067,
+      "learning_rate": 1.4294007656076108e-06,
+      "loss": 0.6687,
+      "step": 5170
+    },
+    {
+      "epoch": 2.5654710189896983,
+      "grad_norm": 0.12632914581609597,
+      "learning_rate": 1.42851765060807e-06,
+      "loss": 0.6794,
+      "step": 5171
+    },
+    {
+      "epoch": 2.5659674816929376,
+      "grad_norm": 0.12626167379165484,
+      "learning_rate": 1.4276346993674267e-06,
+      "loss": 0.6737,
+      "step": 5172
+    },
+    {
+      "epoch": 2.566463944396177,
+      "grad_norm": 0.13881256752440968,
+      "learning_rate": 1.4267519120206251e-06,
+      "loss": 0.7359,
+      "step": 5173
+    },
+    {
+      "epoch": 2.5669604070994168,
+      "grad_norm": 0.13129891241921918,
+      "learning_rate": 1.425869288702585e-06,
+      "loss": 0.72,
+      "step": 5174
+    },
+    {
+      "epoch": 2.567456869802656,
+      "grad_norm": 0.13010534065898433,
+      "learning_rate": 1.4249868295482021e-06,
+      "loss": 0.6765,
+      "step": 5175
+    },
+    {
+      "epoch": 2.5679533325058954,
+      "grad_norm": 0.13578139920958562,
+      "learning_rate": 1.4241045346923463e-06,
+      "loss": 0.7659,
+      "step": 5176
+    },
+    {
+      "epoch": 2.568449795209135,
+      "grad_norm": 0.12966367090684447,
+      "learning_rate": 1.4232224042698606e-06,
+      "loss": 0.6878,
+      "step": 5177
+    },
+    {
+      "epoch": 2.5689462579123745,
+      "grad_norm": 0.1295589151537278,
+      "learning_rate": 1.4223404384155665e-06,
+      "loss": 0.6967,
+      "step": 5178
+    },
+    {
+      "epoch": 2.569442720615614,
+      "grad_norm": 0.12606606232317147,
+      "learning_rate": 1.4214586372642563e-06,
+      "loss": 0.6885,
+      "step": 5179
+    },
+    {
+      "epoch": 2.569939183318853,
+      "grad_norm": 0.12952004740829784,
+      "learning_rate": 1.4205770009507013e-06,
+      "loss": 0.7703,
+      "step": 5180
+    },
+    {
+      "epoch": 2.5704356460220925,
+      "grad_norm": 0.13061476133482602,
+      "learning_rate": 1.4196955296096449e-06,
+      "loss": 0.7473,
+      "step": 5181
+    },
+    {
+      "epoch": 2.570932108725332,
+      "grad_norm": 0.12905289024380742,
+      "learning_rate": 1.418814223375804e-06,
+      "loss": 0.7198,
+      "step": 5182
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.13521590335808284,
+      "learning_rate": 1.4179330823838749e-06,
+      "loss": 0.6663,
+      "step": 5183
+    },
+    {
+      "epoch": 2.571925034131811,
+      "grad_norm": 0.1357147047734266,
+      "learning_rate": 1.4170521067685234e-06,
+      "loss": 0.7672,
+      "step": 5184
+    },
+    {
+      "epoch": 2.5724214968350503,
+      "grad_norm": 0.1420237272434203,
+      "learning_rate": 1.4161712966643942e-06,
+      "loss": 0.768,
+      "step": 5185
+    },
+    {
+      "epoch": 2.5729179595382896,
+      "grad_norm": 0.14090038663886145,
+      "learning_rate": 1.415290652206105e-06,
+      "loss": 0.7108,
+      "step": 5186
+    },
+    {
+      "epoch": 2.573414422241529,
+      "grad_norm": 0.130067394360935,
+      "learning_rate": 1.4144101735282465e-06,
+      "loss": 0.6818,
+      "step": 5187
+    },
+    {
+      "epoch": 2.5739108849447687,
+      "grad_norm": 0.13055899063930992,
+      "learning_rate": 1.4135298607653885e-06,
+      "loss": 0.7142,
+      "step": 5188
+    },
+    {
+      "epoch": 2.574407347648008,
+      "grad_norm": 0.12598384203770444,
+      "learning_rate": 1.4126497140520696e-06,
+      "loss": 0.6844,
+      "step": 5189
+    },
+    {
+      "epoch": 2.5749038103512474,
+      "grad_norm": 0.13099481254462542,
+      "learning_rate": 1.4117697335228082e-06,
+      "loss": 0.6484,
+      "step": 5190
+    },
+    {
+      "epoch": 2.5754002730544867,
+      "grad_norm": 0.12722812757473498,
+      "learning_rate": 1.410889919312094e-06,
+      "loss": 0.6863,
+      "step": 5191
+    },
+    {
+      "epoch": 2.575896735757726,
+      "grad_norm": 0.13021667807315976,
+      "learning_rate": 1.4100102715543934e-06,
+      "loss": 0.7062,
+      "step": 5192
+    },
+    {
+      "epoch": 2.5763931984609654,
+      "grad_norm": 0.13204286958959552,
+      "learning_rate": 1.4091307903841467e-06,
+      "loss": 0.7369,
+      "step": 5193
+    },
+    {
+      "epoch": 2.576889661164205,
+      "grad_norm": 0.12825087423564113,
+      "learning_rate": 1.4082514759357668e-06,
+      "loss": 0.6818,
+      "step": 5194
+    },
+    {
+      "epoch": 2.5773861238674445,
+      "grad_norm": 0.13195934566675152,
+      "learning_rate": 1.4073723283436447e-06,
+      "loss": 0.7205,
+      "step": 5195
+    },
+    {
+      "epoch": 2.577882586570684,
+      "grad_norm": 0.13298051618730106,
+      "learning_rate": 1.4064933477421435e-06,
+      "loss": 0.7092,
+      "step": 5196
+    },
+    {
+      "epoch": 2.578379049273923,
+      "grad_norm": 0.13045643130916357,
+      "learning_rate": 1.4056145342656002e-06,
+      "loss": 0.721,
+      "step": 5197
+    },
+    {
+      "epoch": 2.578875511977163,
+      "grad_norm": 0.1287770798648141,
+      "learning_rate": 1.4047358880483292e-06,
+      "loss": 0.6623,
+      "step": 5198
+    },
+    {
+      "epoch": 2.5793719746804022,
+      "grad_norm": 0.12703038628187402,
+      "learning_rate": 1.403857409224615e-06,
+      "loss": 0.6748,
+      "step": 5199
+    },
+    {
+      "epoch": 2.5798684373836416,
+      "grad_norm": 0.13086879765206233,
+      "learning_rate": 1.4029790979287217e-06,
+      "loss": 0.6923,
+      "step": 5200
+    },
+    {
+      "epoch": 2.580364900086881,
+      "grad_norm": 0.13490815408873744,
+      "learning_rate": 1.402100954294884e-06,
+      "loss": 0.7387,
+      "step": 5201
+    },
+    {
+      "epoch": 2.5808613627901202,
+      "grad_norm": 0.13403369332196752,
+      "learning_rate": 1.4012229784573111e-06,
+      "loss": 0.7194,
+      "step": 5202
+    },
+    {
+      "epoch": 2.5813578254933596,
+      "grad_norm": 0.13057102934580292,
+      "learning_rate": 1.400345170550189e-06,
+      "loss": 0.6975,
+      "step": 5203
+    },
+    {
+      "epoch": 2.5818542881965993,
+      "grad_norm": 0.13217032680650126,
+      "learning_rate": 1.3994675307076766e-06,
+      "loss": 0.6857,
+      "step": 5204
+    },
+    {
+      "epoch": 2.5823507508998387,
+      "grad_norm": 0.12870495240664498,
+      "learning_rate": 1.3985900590639058e-06,
+      "loss": 0.6732,
+      "step": 5205
+    },
+    {
+      "epoch": 2.582847213603078,
+      "grad_norm": 0.1248568293868238,
+      "learning_rate": 1.397712755752984e-06,
+      "loss": 0.6753,
+      "step": 5206
+    },
+    {
+      "epoch": 2.5833436763063173,
+      "grad_norm": 0.1286211773697058,
+      "learning_rate": 1.3968356209089944e-06,
+      "loss": 0.698,
+      "step": 5207
+    },
+    {
+      "epoch": 2.583840139009557,
+      "grad_norm": 0.12476162094736613,
+      "learning_rate": 1.3959586546659926e-06,
+      "loss": 0.6571,
+      "step": 5208
+    },
+    {
+      "epoch": 2.5843366017127964,
+      "grad_norm": 0.13256034900710464,
+      "learning_rate": 1.3950818571580071e-06,
+      "loss": 0.7052,
+      "step": 5209
+    },
+    {
+      "epoch": 2.5848330644160358,
+      "grad_norm": 0.14170229500379294,
+      "learning_rate": 1.3942052285190453e-06,
+      "loss": 0.7395,
+      "step": 5210
+    },
+    {
+      "epoch": 2.585329527119275,
+      "grad_norm": 0.13306092777696724,
+      "learning_rate": 1.3933287688830827e-06,
+      "loss": 0.7503,
+      "step": 5211
+    },
+    {
+      "epoch": 2.5858259898225144,
+      "grad_norm": 0.12983382658565437,
+      "learning_rate": 1.3924524783840748e-06,
+      "loss": 0.6702,
+      "step": 5212
+    },
+    {
+      "epoch": 2.5863224525257538,
+      "grad_norm": 0.12964608984313164,
+      "learning_rate": 1.3915763571559477e-06,
+      "loss": 0.7302,
+      "step": 5213
+    },
+    {
+      "epoch": 2.5868189152289935,
+      "grad_norm": 0.130514060454796,
+      "learning_rate": 1.3907004053326006e-06,
+      "loss": 0.7262,
+      "step": 5214
+    },
+    {
+      "epoch": 2.587315377932233,
+      "grad_norm": 0.1339908897788451,
+      "learning_rate": 1.3898246230479119e-06,
+      "loss": 0.7146,
+      "step": 5215
+    },
+    {
+      "epoch": 2.587811840635472,
+      "grad_norm": 0.12515165316595245,
+      "learning_rate": 1.3889490104357278e-06,
+      "loss": 0.6528,
+      "step": 5216
+    },
+    {
+      "epoch": 2.5883083033387115,
+      "grad_norm": 0.13700532883591948,
+      "learning_rate": 1.3880735676298743e-06,
+      "loss": 0.7071,
+      "step": 5217
+    },
+    {
+      "epoch": 2.5888047660419513,
+      "grad_norm": 0.130354785696436,
+      "learning_rate": 1.3871982947641478e-06,
+      "loss": 0.6935,
+      "step": 5218
+    },
+    {
+      "epoch": 2.5893012287451906,
+      "grad_norm": 0.13701897641660335,
+      "learning_rate": 1.386323191972318e-06,
+      "loss": 0.7427,
+      "step": 5219
+    },
+    {
+      "epoch": 2.58979769144843,
+      "grad_norm": 0.1320735550699063,
+      "learning_rate": 1.3854482593881342e-06,
+      "loss": 0.7064,
+      "step": 5220
+    },
+    {
+      "epoch": 2.5902941541516693,
+      "grad_norm": 0.12715230958238605,
+      "learning_rate": 1.3845734971453114e-06,
+      "loss": 0.6718,
+      "step": 5221
+    },
+    {
+      "epoch": 2.5907906168549086,
+      "grad_norm": 0.13018988366015433,
+      "learning_rate": 1.3836989053775462e-06,
+      "loss": 0.7509,
+      "step": 5222
+    },
+    {
+      "epoch": 2.591287079558148,
+      "grad_norm": 0.13221945422649725,
+      "learning_rate": 1.3828244842185034e-06,
+      "loss": 0.7166,
+      "step": 5223
+    },
+    {
+      "epoch": 2.5917835422613877,
+      "grad_norm": 0.1304668288729789,
+      "learning_rate": 1.381950233801827e-06,
+      "loss": 0.6849,
+      "step": 5224
+    },
+    {
+      "epoch": 2.592280004964627,
+      "grad_norm": 0.12573646819731626,
+      "learning_rate": 1.3810761542611306e-06,
+      "loss": 0.6589,
+      "step": 5225
+    },
+    {
+      "epoch": 2.5927764676678664,
+      "grad_norm": 0.13280192854756245,
+      "learning_rate": 1.380202245730003e-06,
+      "loss": 0.7207,
+      "step": 5226
+    },
+    {
+      "epoch": 2.5932729303711057,
+      "grad_norm": 0.13071113755387934,
+      "learning_rate": 1.3793285083420077e-06,
+      "loss": 0.7014,
+      "step": 5227
+    },
+    {
+      "epoch": 2.5937693930743455,
+      "grad_norm": 0.12860260319733735,
+      "learning_rate": 1.3784549422306808e-06,
+      "loss": 0.6795,
+      "step": 5228
+    },
+    {
+      "epoch": 2.594265855777585,
+      "grad_norm": 0.1270646298734641,
+      "learning_rate": 1.3775815475295343e-06,
+      "loss": 0.6719,
+      "step": 5229
+    },
+    {
+      "epoch": 2.594762318480824,
+      "grad_norm": 0.131427270856042,
+      "learning_rate": 1.3767083243720516e-06,
+      "loss": 0.7102,
+      "step": 5230
+    },
+    {
+      "epoch": 2.5952587811840635,
+      "grad_norm": 0.1291160109583672,
+      "learning_rate": 1.37583527289169e-06,
+      "loss": 0.6879,
+      "step": 5231
+    },
+    {
+      "epoch": 2.595755243887303,
+      "grad_norm": 0.12483862070084152,
+      "learning_rate": 1.374962393221883e-06,
+      "loss": 0.6806,
+      "step": 5232
+    },
+    {
+      "epoch": 2.596251706590542,
+      "grad_norm": 0.1265687299366481,
+      "learning_rate": 1.3740896854960361e-06,
+      "loss": 0.6648,
+      "step": 5233
+    },
+    {
+      "epoch": 2.596748169293782,
+      "grad_norm": 0.12551276631561484,
+      "learning_rate": 1.3732171498475269e-06,
+      "loss": 0.6796,
+      "step": 5234
+    },
+    {
+      "epoch": 2.5972446319970213,
+      "grad_norm": 0.1919668140078233,
+      "learning_rate": 1.3723447864097105e-06,
+      "loss": 0.7173,
+      "step": 5235
+    },
+    {
+      "epoch": 2.5977410947002606,
+      "grad_norm": 0.13071121709395,
+      "learning_rate": 1.3714725953159136e-06,
+      "loss": 0.725,
+      "step": 5236
+    },
+    {
+      "epoch": 2.5982375574035,
+      "grad_norm": 0.12628300917307875,
+      "learning_rate": 1.3706005766994354e-06,
+      "loss": 0.6583,
+      "step": 5237
+    },
+    {
+      "epoch": 2.5987340201067397,
+      "grad_norm": 0.1324112740619143,
+      "learning_rate": 1.3697287306935498e-06,
+      "loss": 0.7062,
+      "step": 5238
+    },
+    {
+      "epoch": 2.599230482809979,
+      "grad_norm": 0.13328148315750796,
+      "learning_rate": 1.3688570574315058e-06,
+      "loss": 0.7195,
+      "step": 5239
+    },
+    {
+      "epoch": 2.5997269455132184,
+      "grad_norm": 0.1392294677729809,
+      "learning_rate": 1.3679855570465244e-06,
+      "loss": 0.7264,
+      "step": 5240
+    },
+    {
+      "epoch": 2.6002234082164577,
+      "grad_norm": 0.1257272610807176,
+      "learning_rate": 1.367114229671799e-06,
+      "loss": 0.6915,
+      "step": 5241
+    },
+    {
+      "epoch": 2.600719870919697,
+      "grad_norm": 0.12811010695255046,
+      "learning_rate": 1.3662430754405004e-06,
+      "loss": 0.7125,
+      "step": 5242
+    },
+    {
+      "epoch": 2.6012163336229364,
+      "grad_norm": 0.1329892421572713,
+      "learning_rate": 1.365372094485768e-06,
+      "loss": 0.7393,
+      "step": 5243
+    },
+    {
+      "epoch": 2.601712796326176,
+      "grad_norm": 0.1293006710207882,
+      "learning_rate": 1.36450128694072e-06,
+      "loss": 0.6771,
+      "step": 5244
+    },
+    {
+      "epoch": 2.6022092590294155,
+      "grad_norm": 0.13954401243263015,
+      "learning_rate": 1.3636306529384432e-06,
+      "loss": 0.7291,
+      "step": 5245
+    },
+    {
+      "epoch": 2.602705721732655,
+      "grad_norm": 0.12768761269525258,
+      "learning_rate": 1.3627601926120005e-06,
+      "loss": 0.6687,
+      "step": 5246
+    },
+    {
+      "epoch": 2.603202184435894,
+      "grad_norm": 0.1329893303483065,
+      "learning_rate": 1.3618899060944286e-06,
+      "loss": 0.7366,
+      "step": 5247
+    },
+    {
+      "epoch": 2.603698647139134,
+      "grad_norm": 0.13009174728461212,
+      "learning_rate": 1.3610197935187358e-06,
+      "loss": 0.6713,
+      "step": 5248
+    },
+    {
+      "epoch": 2.6041951098423732,
+      "grad_norm": 0.12802975892687476,
+      "learning_rate": 1.3601498550179059e-06,
+      "loss": 0.7355,
+      "step": 5249
+    },
+    {
+      "epoch": 2.6046915725456126,
+      "grad_norm": 0.1297273904569148,
+      "learning_rate": 1.3592800907248949e-06,
+      "loss": 0.7098,
+      "step": 5250
+    },
+    {
+      "epoch": 2.605188035248852,
+      "grad_norm": 0.12806246500929,
+      "learning_rate": 1.3584105007726312e-06,
+      "loss": 0.6429,
+      "step": 5251
+    },
+    {
+      "epoch": 2.6056844979520912,
+      "grad_norm": 0.13041176910308638,
+      "learning_rate": 1.3575410852940202e-06,
+      "loss": 0.6807,
+      "step": 5252
+    },
+    {
+      "epoch": 2.6061809606553306,
+      "grad_norm": 0.12886597784723972,
+      "learning_rate": 1.3566718444219342e-06,
+      "loss": 0.7237,
+      "step": 5253
+    },
+    {
+      "epoch": 2.6066774233585703,
+      "grad_norm": 0.12916633911425887,
+      "learning_rate": 1.355802778289226e-06,
+      "loss": 0.6739,
+      "step": 5254
+    },
+    {
+      "epoch": 2.6071738860618097,
+      "grad_norm": 0.1287678914063443,
+      "learning_rate": 1.3549338870287165e-06,
+      "loss": 0.6716,
+      "step": 5255
+    },
+    {
+      "epoch": 2.607670348765049,
+      "grad_norm": 0.13212685032768234,
+      "learning_rate": 1.3540651707732036e-06,
+      "loss": 0.7382,
+      "step": 5256
+    },
+    {
+      "epoch": 2.6081668114682883,
+      "grad_norm": 0.12905546375739468,
+      "learning_rate": 1.3531966296554555e-06,
+      "loss": 0.6551,
+      "step": 5257
+    },
+    {
+      "epoch": 2.608663274171528,
+      "grad_norm": 0.12906147644682286,
+      "learning_rate": 1.3523282638082142e-06,
+      "loss": 0.7328,
+      "step": 5258
+    },
+    {
+      "epoch": 2.6091597368747674,
+      "grad_norm": 0.1313108855974763,
+      "learning_rate": 1.3514600733641969e-06,
+      "loss": 0.6417,
+      "step": 5259
+    },
+    {
+      "epoch": 2.6096561995780068,
+      "grad_norm": 0.12974355187493028,
+      "learning_rate": 1.3505920584560913e-06,
+      "loss": 0.6928,
+      "step": 5260
+    },
+    {
+      "epoch": 2.610152662281246,
+      "grad_norm": 0.12727154967902357,
+      "learning_rate": 1.349724219216561e-06,
+      "loss": 0.669,
+      "step": 5261
+    },
+    {
+      "epoch": 2.6106491249844854,
+      "grad_norm": 0.12535216696134888,
+      "learning_rate": 1.3488565557782407e-06,
+      "loss": 0.682,
+      "step": 5262
+    },
+    {
+      "epoch": 2.6111455876877248,
+      "grad_norm": 0.13133855825746554,
+      "learning_rate": 1.347989068273738e-06,
+      "loss": 0.7084,
+      "step": 5263
+    },
+    {
+      "epoch": 2.6116420503909645,
+      "grad_norm": 0.13110682521576933,
+      "learning_rate": 1.3471217568356354e-06,
+      "loss": 0.7168,
+      "step": 5264
+    },
+    {
+      "epoch": 2.612138513094204,
+      "grad_norm": 0.12162732379351464,
+      "learning_rate": 1.3462546215964867e-06,
+      "loss": 0.6539,
+      "step": 5265
+    },
+    {
+      "epoch": 2.612634975797443,
+      "grad_norm": 0.13230036317983876,
+      "learning_rate": 1.345387662688821e-06,
+      "loss": 0.7189,
+      "step": 5266
+    },
+    {
+      "epoch": 2.6131314385006825,
+      "grad_norm": 0.12335108832614605,
+      "learning_rate": 1.3445208802451383e-06,
+      "loss": 0.6612,
+      "step": 5267
+    },
+    {
+      "epoch": 2.6136279012039223,
+      "grad_norm": 0.13204613627760084,
+      "learning_rate": 1.3436542743979125e-06,
+      "loss": 0.7287,
+      "step": 5268
+    },
+    {
+      "epoch": 2.6141243639071616,
+      "grad_norm": 0.12765603511898485,
+      "learning_rate": 1.34278784527959e-06,
+      "loss": 0.6788,
+      "step": 5269
+    },
+    {
+      "epoch": 2.614620826610401,
+      "grad_norm": 0.13326553938549146,
+      "learning_rate": 1.3419215930225898e-06,
+      "loss": 0.6903,
+      "step": 5270
+    },
+    {
+      "epoch": 2.6151172893136403,
+      "grad_norm": 0.13006724151910573,
+      "learning_rate": 1.341055517759307e-06,
+      "loss": 0.6968,
+      "step": 5271
+    },
+    {
+      "epoch": 2.6156137520168796,
+      "grad_norm": 0.12782169012677588,
+      "learning_rate": 1.3401896196221061e-06,
+      "loss": 0.6402,
+      "step": 5272
+    },
+    {
+      "epoch": 2.616110214720119,
+      "grad_norm": 0.1298644050131224,
+      "learning_rate": 1.3393238987433247e-06,
+      "loss": 0.6786,
+      "step": 5273
+    },
+    {
+      "epoch": 2.6166066774233587,
+      "grad_norm": 0.13093853621883342,
+      "learning_rate": 1.338458355255276e-06,
+      "loss": 0.7008,
+      "step": 5274
+    },
+    {
+      "epoch": 2.617103140126598,
+      "grad_norm": 0.1251589926214746,
+      "learning_rate": 1.3375929892902435e-06,
+      "loss": 0.6642,
+      "step": 5275
+    },
+    {
+      "epoch": 2.6175996028298374,
+      "grad_norm": 0.1332356566343829,
+      "learning_rate": 1.3367278009804852e-06,
+      "loss": 0.6859,
+      "step": 5276
+    },
+    {
+      "epoch": 2.6180960655330767,
+      "grad_norm": 0.1351996093141888,
+      "learning_rate": 1.3358627904582308e-06,
+      "loss": 0.6886,
+      "step": 5277
+    },
+    {
+      "epoch": 2.6185925282363165,
+      "grad_norm": 0.12760964101813727,
+      "learning_rate": 1.3349979578556827e-06,
+      "loss": 0.6775,
+      "step": 5278
+    },
+    {
+      "epoch": 2.619088990939556,
+      "grad_norm": 0.13492506375570365,
+      "learning_rate": 1.334133303305018e-06,
+      "loss": 0.7233,
+      "step": 5279
+    },
+    {
+      "epoch": 2.619585453642795,
+      "grad_norm": 0.13103285327539846,
+      "learning_rate": 1.3332688269383842e-06,
+      "loss": 0.729,
+      "step": 5280
+    },
+    {
+      "epoch": 2.6200819163460345,
+      "grad_norm": 0.12874709415836544,
+      "learning_rate": 1.3324045288879034e-06,
+      "loss": 0.7383,
+      "step": 5281
+    },
+    {
+      "epoch": 2.620578379049274,
+      "grad_norm": 0.12537613283777926,
+      "learning_rate": 1.33154040928567e-06,
+      "loss": 0.6694,
+      "step": 5282
+    },
+    {
+      "epoch": 2.621074841752513,
+      "grad_norm": 0.1229814161755244,
+      "learning_rate": 1.3306764682637487e-06,
+      "loss": 0.642,
+      "step": 5283
+    },
+    {
+      "epoch": 2.621571304455753,
+      "grad_norm": 0.13180449979729172,
+      "learning_rate": 1.329812705954183e-06,
+      "loss": 0.7413,
+      "step": 5284
+    },
+    {
+      "epoch": 2.6220677671589923,
+      "grad_norm": 0.1299850828781537,
+      "learning_rate": 1.3289491224889805e-06,
+      "loss": 0.7245,
+      "step": 5285
+    },
+    {
+      "epoch": 2.6225642298622316,
+      "grad_norm": 0.12657003185254417,
+      "learning_rate": 1.328085718000129e-06,
+      "loss": 0.691,
+      "step": 5286
+    },
+    {
+      "epoch": 2.623060692565471,
+      "grad_norm": 0.12873747042196176,
+      "learning_rate": 1.3272224926195847e-06,
+      "loss": 0.6719,
+      "step": 5287
+    },
+    {
+      "epoch": 2.6235571552687107,
+      "grad_norm": 0.12986302258803353,
+      "learning_rate": 1.326359446479279e-06,
+      "loss": 0.6856,
+      "step": 5288
+    },
+    {
+      "epoch": 2.62405361797195,
+      "grad_norm": 0.13177548514946316,
+      "learning_rate": 1.325496579711114e-06,
+      "loss": 0.7408,
+      "step": 5289
+    },
+    {
+      "epoch": 2.6245500806751894,
+      "grad_norm": 0.13080981754895,
+      "learning_rate": 1.3246338924469646e-06,
+      "loss": 0.6891,
+      "step": 5290
+    },
+    {
+      "epoch": 2.6250465433784287,
+      "grad_norm": 0.1319670698468296,
+      "learning_rate": 1.3237713848186799e-06,
+      "loss": 0.7171,
+      "step": 5291
+    },
+    {
+      "epoch": 2.625543006081668,
+      "grad_norm": 0.1268286576276871,
+      "learning_rate": 1.3229090569580782e-06,
+      "loss": 0.7105,
+      "step": 5292
+    },
+    {
+      "epoch": 2.6260394687849073,
+      "grad_norm": 0.13031273872848315,
+      "learning_rate": 1.3220469089969556e-06,
+      "loss": 0.6647,
+      "step": 5293
+    },
+    {
+      "epoch": 2.6265359314881467,
+      "grad_norm": 0.12705509580073604,
+      "learning_rate": 1.3211849410670755e-06,
+      "loss": 0.6708,
+      "step": 5294
+    },
+    {
+      "epoch": 2.6270323941913865,
+      "grad_norm": 0.12662469871597498,
+      "learning_rate": 1.3203231533001753e-06,
+      "loss": 0.7167,
+      "step": 5295
+    },
+    {
+      "epoch": 2.627528856894626,
+      "grad_norm": 0.1351319977157379,
+      "learning_rate": 1.3194615458279675e-06,
+      "loss": 0.7027,
+      "step": 5296
+    },
+    {
+      "epoch": 2.628025319597865,
+      "grad_norm": 0.1293833244334666,
+      "learning_rate": 1.3186001187821328e-06,
+      "loss": 0.737,
+      "step": 5297
+    },
+    {
+      "epoch": 2.628521782301105,
+      "grad_norm": 0.12681644363417152,
+      "learning_rate": 1.317738872294329e-06,
+      "loss": 0.687,
+      "step": 5298
+    },
+    {
+      "epoch": 2.6290182450043442,
+      "grad_norm": 0.12782685543955466,
+      "learning_rate": 1.3168778064961824e-06,
+      "loss": 0.6944,
+      "step": 5299
+    },
+    {
+      "epoch": 2.6295147077075836,
+      "grad_norm": 0.1283732539871523,
+      "learning_rate": 1.3160169215192929e-06,
+      "loss": 0.6593,
+      "step": 5300
+    },
+    {
+      "epoch": 2.630011170410823,
+      "grad_norm": 0.12786728676839182,
+      "learning_rate": 1.315156217495233e-06,
+      "loss": 0.6914,
+      "step": 5301
+    },
+    {
+      "epoch": 2.630507633114062,
+      "grad_norm": 0.12640922642050675,
+      "learning_rate": 1.3142956945555474e-06,
+      "loss": 0.6319,
+      "step": 5302
+    },
+    {
+      "epoch": 2.6310040958173015,
+      "grad_norm": 0.12770955937417205,
+      "learning_rate": 1.3134353528317539e-06,
+      "loss": 0.6829,
+      "step": 5303
+    },
+    {
+      "epoch": 2.631500558520541,
+      "grad_norm": 0.12759140923876178,
+      "learning_rate": 1.312575192455341e-06,
+      "loss": 0.7206,
+      "step": 5304
+    },
+    {
+      "epoch": 2.6319970212237807,
+      "grad_norm": 0.13233654333615097,
+      "learning_rate": 1.3117152135577721e-06,
+      "loss": 0.7442,
+      "step": 5305
+    },
+    {
+      "epoch": 2.63249348392702,
+      "grad_norm": 0.13201851791859284,
+      "learning_rate": 1.3108554162704797e-06,
+      "loss": 0.7381,
+      "step": 5306
+    },
+    {
+      "epoch": 2.6329899466302593,
+      "grad_norm": 0.13006021868404868,
+      "learning_rate": 1.3099958007248698e-06,
+      "loss": 0.685,
+      "step": 5307
+    },
+    {
+      "epoch": 2.633486409333499,
+      "grad_norm": 0.1308979224400638,
+      "learning_rate": 1.3091363670523225e-06,
+      "loss": 0.6923,
+      "step": 5308
+    },
+    {
+      "epoch": 2.6339828720367384,
+      "grad_norm": 0.13246572766620834,
+      "learning_rate": 1.3082771153841872e-06,
+      "loss": 0.7535,
+      "step": 5309
+    },
+    {
+      "epoch": 2.6344793347399778,
+      "grad_norm": 0.13028781856699762,
+      "learning_rate": 1.307418045851786e-06,
+      "loss": 0.6802,
+      "step": 5310
+    },
+    {
+      "epoch": 2.634975797443217,
+      "grad_norm": 0.1291552670036102,
+      "learning_rate": 1.3065591585864161e-06,
+      "loss": 0.7137,
+      "step": 5311
+    },
+    {
+      "epoch": 2.6354722601464564,
+      "grad_norm": 0.1297984636190616,
+      "learning_rate": 1.3057004537193424e-06,
+      "loss": 0.6697,
+      "step": 5312
+    },
+    {
+      "epoch": 2.6359687228496957,
+      "grad_norm": 0.12833411800480857,
+      "learning_rate": 1.3048419313818062e-06,
+      "loss": 0.6836,
+      "step": 5313
+    },
+    {
+      "epoch": 2.636465185552935,
+      "grad_norm": 0.13042366934420158,
+      "learning_rate": 1.3039835917050177e-06,
+      "loss": 0.6944,
+      "step": 5314
+    },
+    {
+      "epoch": 2.636961648256175,
+      "grad_norm": 0.13642883453725888,
+      "learning_rate": 1.30312543482016e-06,
+      "loss": 0.7166,
+      "step": 5315
+    },
+    {
+      "epoch": 2.637458110959414,
+      "grad_norm": 0.130255734346025,
+      "learning_rate": 1.3022674608583907e-06,
+      "loss": 0.7537,
+      "step": 5316
+    },
+    {
+      "epoch": 2.6379545736626535,
+      "grad_norm": 0.13076117414943525,
+      "learning_rate": 1.3014096699508338e-06,
+      "loss": 0.7065,
+      "step": 5317
+    },
+    {
+      "epoch": 2.6384510363658933,
+      "grad_norm": 0.13062995849439213,
+      "learning_rate": 1.3005520622285922e-06,
+      "loss": 0.6623,
+      "step": 5318
+    },
+    {
+      "epoch": 2.6389474990691326,
+      "grad_norm": 0.12821638120486448,
+      "learning_rate": 1.2996946378227351e-06,
+      "loss": 0.6664,
+      "step": 5319
+    },
+    {
+      "epoch": 2.639443961772372,
+      "grad_norm": 0.133639093888335,
+      "learning_rate": 1.298837396864308e-06,
+      "loss": 0.75,
+      "step": 5320
+    },
+    {
+      "epoch": 2.6399404244756113,
+      "grad_norm": 0.12479238215751826,
+      "learning_rate": 1.297980339484326e-06,
+      "loss": 0.6877,
+      "step": 5321
+    },
+    {
+      "epoch": 2.6404368871788506,
+      "grad_norm": 0.13212736303956885,
+      "learning_rate": 1.297123465813775e-06,
+      "loss": 0.7034,
+      "step": 5322
+    },
+    {
+      "epoch": 2.64093334988209,
+      "grad_norm": 0.13499589181985006,
+      "learning_rate": 1.2962667759836166e-06,
+      "loss": 0.721,
+      "step": 5323
+    },
+    {
+      "epoch": 2.6414298125853293,
+      "grad_norm": 0.12490032994276366,
+      "learning_rate": 1.2954102701247801e-06,
+      "loss": 0.66,
+      "step": 5324
+    },
+    {
+      "epoch": 2.641926275288569,
+      "grad_norm": 0.14001714407558724,
+      "learning_rate": 1.2945539483681708e-06,
+      "loss": 0.6848,
+      "step": 5325
+    },
+    {
+      "epoch": 2.6424227379918084,
+      "grad_norm": 0.13601215073821113,
+      "learning_rate": 1.2936978108446624e-06,
+      "loss": 0.7245,
+      "step": 5326
+    },
+    {
+      "epoch": 2.6429192006950477,
+      "grad_norm": 0.12605012170458385,
+      "learning_rate": 1.292841857685101e-06,
+      "loss": 0.6667,
+      "step": 5327
+    },
+    {
+      "epoch": 2.643415663398287,
+      "grad_norm": 0.12872625874119725,
+      "learning_rate": 1.2919860890203073e-06,
+      "loss": 0.6868,
+      "step": 5328
+    },
+    {
+      "epoch": 2.643912126101527,
+      "grad_norm": 0.13008341539653537,
+      "learning_rate": 1.2911305049810701e-06,
+      "loss": 0.6629,
+      "step": 5329
+    },
+    {
+      "epoch": 2.644408588804766,
+      "grad_norm": 0.12691494726221766,
+      "learning_rate": 1.2902751056981533e-06,
+      "loss": 0.6905,
+      "step": 5330
+    },
+    {
+      "epoch": 2.6449050515080055,
+      "grad_norm": 0.1238746354104558,
+      "learning_rate": 1.2894198913022903e-06,
+      "loss": 0.6731,
+      "step": 5331
+    },
+    {
+      "epoch": 2.645401514211245,
+      "grad_norm": 0.13315315264555727,
+      "learning_rate": 1.2885648619241866e-06,
+      "loss": 0.699,
+      "step": 5332
+    },
+    {
+      "epoch": 2.645897976914484,
+      "grad_norm": 0.1313601573492306,
+      "learning_rate": 1.28771001769452e-06,
+      "loss": 0.6966,
+      "step": 5333
+    },
+    {
+      "epoch": 2.6463944396177235,
+      "grad_norm": 0.12904285233884105,
+      "learning_rate": 1.2868553587439386e-06,
+      "loss": 0.6657,
+      "step": 5334
+    },
+    {
+      "epoch": 2.6468909023209632,
+      "grad_norm": 0.1276652699366476,
+      "learning_rate": 1.2860008852030653e-06,
+      "loss": 0.7054,
+      "step": 5335
+    },
+    {
+      "epoch": 2.6473873650242026,
+      "grad_norm": 0.12726233644290474,
+      "learning_rate": 1.2851465972024908e-06,
+      "loss": 0.6983,
+      "step": 5336
+    },
+    {
+      "epoch": 2.647883827727442,
+      "grad_norm": 0.128524573267683,
+      "learning_rate": 1.2842924948727809e-06,
+      "loss": 0.6639,
+      "step": 5337
+    },
+    {
+      "epoch": 2.6483802904306812,
+      "grad_norm": 0.1354082209207073,
+      "learning_rate": 1.2834385783444708e-06,
+      "loss": 0.7063,
+      "step": 5338
+    },
+    {
+      "epoch": 2.648876753133921,
+      "grad_norm": 0.1286859619680868,
+      "learning_rate": 1.282584847748067e-06,
+      "loss": 0.6669,
+      "step": 5339
+    },
+    {
+      "epoch": 2.6493732158371603,
+      "grad_norm": 0.12527601100039995,
+      "learning_rate": 1.2817313032140504e-06,
+      "loss": 0.6949,
+      "step": 5340
+    },
+    {
+      "epoch": 2.6498696785403997,
+      "grad_norm": 0.13317900586790732,
+      "learning_rate": 1.2808779448728701e-06,
+      "loss": 0.6903,
+      "step": 5341
+    },
+    {
+      "epoch": 2.650366141243639,
+      "grad_norm": 0.1339289905433541,
+      "learning_rate": 1.2800247728549492e-06,
+      "loss": 0.707,
+      "step": 5342
+    },
+    {
+      "epoch": 2.6508626039468783,
+      "grad_norm": 0.13121373346042195,
+      "learning_rate": 1.2791717872906812e-06,
+      "loss": 0.7315,
+      "step": 5343
+    },
+    {
+      "epoch": 2.6513590666501177,
+      "grad_norm": 0.13089098565705848,
+      "learning_rate": 1.2783189883104301e-06,
+      "loss": 0.6792,
+      "step": 5344
+    },
+    {
+      "epoch": 2.6518555293533574,
+      "grad_norm": 0.12878111758105115,
+      "learning_rate": 1.2774663760445343e-06,
+      "loss": 0.6792,
+      "step": 5345
+    },
+    {
+      "epoch": 2.6523519920565968,
+      "grad_norm": 0.14725422361310098,
+      "learning_rate": 1.2766139506233012e-06,
+      "loss": 0.6749,
+      "step": 5346
+    },
+    {
+      "epoch": 2.652848454759836,
+      "grad_norm": 0.1301170341633137,
+      "learning_rate": 1.2757617121770093e-06,
+      "loss": 0.6623,
+      "step": 5347
+    },
+    {
+      "epoch": 2.6533449174630754,
+      "grad_norm": 0.1317723532692615,
+      "learning_rate": 1.2749096608359124e-06,
+      "loss": 0.7481,
+      "step": 5348
+    },
+    {
+      "epoch": 2.653841380166315,
+      "grad_norm": 0.12676960816569183,
+      "learning_rate": 1.2740577967302292e-06,
+      "loss": 0.6932,
+      "step": 5349
+    },
+    {
+      "epoch": 2.6543378428695545,
+      "grad_norm": 0.12671113631004585,
+      "learning_rate": 1.2732061199901563e-06,
+      "loss": 0.6807,
+      "step": 5350
+    },
+    {
+      "epoch": 2.654834305572794,
+      "grad_norm": 0.13126321692463977,
+      "learning_rate": 1.2723546307458564e-06,
+      "loss": 0.7329,
+      "step": 5351
+    },
+    {
+      "epoch": 2.655330768276033,
+      "grad_norm": 0.1264003258300388,
+      "learning_rate": 1.2715033291274686e-06,
+      "loss": 0.7149,
+      "step": 5352
+    },
+    {
+      "epoch": 2.6558272309792725,
+      "grad_norm": 0.13089334621865706,
+      "learning_rate": 1.2706522152650997e-06,
+      "loss": 0.7176,
+      "step": 5353
+    },
+    {
+      "epoch": 2.656323693682512,
+      "grad_norm": 0.13062906649658093,
+      "learning_rate": 1.2698012892888272e-06,
+      "loss": 0.6898,
+      "step": 5354
+    },
+    {
+      "epoch": 2.6568201563857516,
+      "grad_norm": 0.12660598434158638,
+      "learning_rate": 1.2689505513287042e-06,
+      "loss": 0.6998,
+      "step": 5355
+    },
+    {
+      "epoch": 2.657316619088991,
+      "grad_norm": 0.13240501373271096,
+      "learning_rate": 1.2681000015147505e-06,
+      "loss": 0.7185,
+      "step": 5356
+    },
+    {
+      "epoch": 2.6578130817922303,
+      "grad_norm": 0.13214361476854886,
+      "learning_rate": 1.2672496399769596e-06,
+      "loss": 0.717,
+      "step": 5357
+    },
+    {
+      "epoch": 2.6583095444954696,
+      "grad_norm": 0.12662065221064372,
+      "learning_rate": 1.2663994668452961e-06,
+      "loss": 0.6901,
+      "step": 5358
+    },
+    {
+      "epoch": 2.6588060071987094,
+      "grad_norm": 0.12807260004668625,
+      "learning_rate": 1.2655494822496938e-06,
+      "loss": 0.7518,
+      "step": 5359
+    },
+    {
+      "epoch": 2.6593024699019487,
+      "grad_norm": 0.12843060423085104,
+      "learning_rate": 1.2646996863200612e-06,
+      "loss": 0.6849,
+      "step": 5360
+    },
+    {
+      "epoch": 2.659798932605188,
+      "grad_norm": 0.13206054308179124,
+      "learning_rate": 1.263850079186274e-06,
+      "loss": 0.7003,
+      "step": 5361
+    },
+    {
+      "epoch": 2.6602953953084274,
+      "grad_norm": 0.1374628315367967,
+      "learning_rate": 1.2630006609781832e-06,
+      "loss": 0.7265,
+      "step": 5362
+    },
+    {
+      "epoch": 2.6607918580116667,
+      "grad_norm": 0.133031557785691,
+      "learning_rate": 1.2621514318256073e-06,
+      "loss": 0.728,
+      "step": 5363
+    },
+    {
+      "epoch": 2.661288320714906,
+      "grad_norm": 0.13243919879974533,
+      "learning_rate": 1.2613023918583379e-06,
+      "loss": 0.699,
+      "step": 5364
+    },
+    {
+      "epoch": 2.661784783418146,
+      "grad_norm": 0.13056450670934303,
+      "learning_rate": 1.2604535412061367e-06,
+      "loss": 0.6934,
+      "step": 5365
+    },
+    {
+      "epoch": 2.662281246121385,
+      "grad_norm": 0.1327556834023822,
+      "learning_rate": 1.259604879998736e-06,
+      "loss": 0.6961,
+      "step": 5366
+    },
+    {
+      "epoch": 2.6627777088246245,
+      "grad_norm": 0.12818630552200058,
+      "learning_rate": 1.2587564083658424e-06,
+      "loss": 0.7106,
+      "step": 5367
+    },
+    {
+      "epoch": 2.663274171527864,
+      "grad_norm": 0.1310766749031488,
+      "learning_rate": 1.257908126437129e-06,
+      "loss": 0.7148,
+      "step": 5368
+    },
+    {
+      "epoch": 2.6637706342311036,
+      "grad_norm": 0.127265232183585,
+      "learning_rate": 1.257060034342244e-06,
+      "loss": 0.6936,
+      "step": 5369
+    },
+    {
+      "epoch": 2.664267096934343,
+      "grad_norm": 0.13015485116988296,
+      "learning_rate": 1.2562121322108033e-06,
+      "loss": 0.6881,
+      "step": 5370
+    },
+    {
+      "epoch": 2.6647635596375823,
+      "grad_norm": 0.12987369619590036,
+      "learning_rate": 1.2553644201723953e-06,
+      "loss": 0.6782,
+      "step": 5371
+    },
+    {
+      "epoch": 2.6652600223408216,
+      "grad_norm": 0.1286930218309498,
+      "learning_rate": 1.25451689835658e-06,
+      "loss": 0.7157,
+      "step": 5372
+    },
+    {
+      "epoch": 2.665756485044061,
+      "grad_norm": 0.13613822958082458,
+      "learning_rate": 1.2536695668928861e-06,
+      "loss": 0.7231,
+      "step": 5373
+    },
+    {
+      "epoch": 2.6662529477473003,
+      "grad_norm": 0.13210551836350637,
+      "learning_rate": 1.2528224259108165e-06,
+      "loss": 0.7344,
+      "step": 5374
+    },
+    {
+      "epoch": 2.66674941045054,
+      "grad_norm": 0.12948567861798826,
+      "learning_rate": 1.2519754755398422e-06,
+      "loss": 0.7031,
+      "step": 5375
+    },
+    {
+      "epoch": 2.6672458731537794,
+      "grad_norm": 0.1267189214191555,
+      "learning_rate": 1.251128715909405e-06,
+      "loss": 0.7144,
+      "step": 5376
+    },
+    {
+      "epoch": 2.6677423358570187,
+      "grad_norm": 0.12858865733135458,
+      "learning_rate": 1.25028214714892e-06,
+      "loss": 0.6898,
+      "step": 5377
+    },
+    {
+      "epoch": 2.668238798560258,
+      "grad_norm": 0.1259741786554077,
+      "learning_rate": 1.2494357693877707e-06,
+      "loss": 0.6758,
+      "step": 5378
+    },
+    {
+      "epoch": 2.668735261263498,
+      "grad_norm": 0.13563486865602353,
+      "learning_rate": 1.2485895827553132e-06,
+      "loss": 0.7236,
+      "step": 5379
+    },
+    {
+      "epoch": 2.669231723966737,
+      "grad_norm": 0.13321717377807327,
+      "learning_rate": 1.2477435873808736e-06,
+      "loss": 0.7188,
+      "step": 5380
+    },
+    {
+      "epoch": 2.6697281866699765,
+      "grad_norm": 0.13087033023193448,
+      "learning_rate": 1.246897783393748e-06,
+      "loss": 0.7512,
+      "step": 5381
+    },
+    {
+      "epoch": 2.670224649373216,
+      "grad_norm": 0.13658879188594283,
+      "learning_rate": 1.2460521709232042e-06,
+      "loss": 0.6728,
+      "step": 5382
+    },
+    {
+      "epoch": 2.670721112076455,
+      "grad_norm": 0.13628616113751904,
+      "learning_rate": 1.2452067500984797e-06,
+      "loss": 0.6972,
+      "step": 5383
+    },
+    {
+      "epoch": 2.6712175747796945,
+      "grad_norm": 0.13793409627064301,
+      "learning_rate": 1.2443615210487853e-06,
+      "loss": 0.7301,
+      "step": 5384
+    },
+    {
+      "epoch": 2.6717140374829342,
+      "grad_norm": 0.12840778349061024,
+      "learning_rate": 1.2435164839032999e-06,
+      "loss": 0.6916,
+      "step": 5385
+    },
+    {
+      "epoch": 2.6722105001861736,
+      "grad_norm": 0.13033494926277364,
+      "learning_rate": 1.2426716387911728e-06,
+      "loss": 0.6811,
+      "step": 5386
+    },
+    {
+      "epoch": 2.672706962889413,
+      "grad_norm": 0.13130268215437454,
+      "learning_rate": 1.2418269858415267e-06,
+      "loss": 0.6996,
+      "step": 5387
+    },
+    {
+      "epoch": 2.6732034255926522,
+      "grad_norm": 0.12904504134003447,
+      "learning_rate": 1.2409825251834518e-06,
+      "loss": 0.6958,
+      "step": 5388
+    },
+    {
+      "epoch": 2.673699888295892,
+      "grad_norm": 0.12717110744137422,
+      "learning_rate": 1.2401382569460118e-06,
+      "loss": 0.6495,
+      "step": 5389
+    },
+    {
+      "epoch": 2.6741963509991313,
+      "grad_norm": 0.1318115465132516,
+      "learning_rate": 1.239294181258239e-06,
+      "loss": 0.6998,
+      "step": 5390
+    },
+    {
+      "epoch": 2.6746928137023707,
+      "grad_norm": 0.133476545289,
+      "learning_rate": 1.2384502982491359e-06,
+      "loss": 0.7364,
+      "step": 5391
+    },
+    {
+      "epoch": 2.67518927640561,
+      "grad_norm": 0.13316247822909816,
+      "learning_rate": 1.237606608047678e-06,
+      "loss": 0.7539,
+      "step": 5392
+    },
+    {
+      "epoch": 2.6756857391088493,
+      "grad_norm": 0.13762117719884678,
+      "learning_rate": 1.2367631107828086e-06,
+      "loss": 0.7731,
+      "step": 5393
+    },
+    {
+      "epoch": 2.6761822018120887,
+      "grad_norm": 0.1294653958734037,
+      "learning_rate": 1.2359198065834439e-06,
+      "loss": 0.7375,
+      "step": 5394
+    },
+    {
+      "epoch": 2.6766786645153284,
+      "grad_norm": 0.13185956190100087,
+      "learning_rate": 1.2350766955784688e-06,
+      "loss": 0.7211,
+      "step": 5395
+    },
+    {
+      "epoch": 2.6771751272185678,
+      "grad_norm": 0.13191116919803744,
+      "learning_rate": 1.2342337778967383e-06,
+      "loss": 0.6744,
+      "step": 5396
+    },
+    {
+      "epoch": 2.677671589921807,
+      "grad_norm": 0.12866861106797114,
+      "learning_rate": 1.2333910536670818e-06,
+      "loss": 0.7133,
+      "step": 5397
+    },
+    {
+      "epoch": 2.6781680526250464,
+      "grad_norm": 0.13443812133789382,
+      "learning_rate": 1.2325485230182923e-06,
+      "loss": 0.6642,
+      "step": 5398
+    },
+    {
+      "epoch": 2.678664515328286,
+      "grad_norm": 0.13194917442407558,
+      "learning_rate": 1.2317061860791402e-06,
+      "loss": 0.7635,
+      "step": 5399
+    },
+    {
+      "epoch": 2.6791609780315255,
+      "grad_norm": 0.12949501549978698,
+      "learning_rate": 1.230864042978361e-06,
+      "loss": 0.708,
+      "step": 5400
+    },
+    {
+      "epoch": 2.679657440734765,
+      "grad_norm": 0.12914002116089188,
+      "learning_rate": 1.230022093844664e-06,
+      "loss": 0.6673,
+      "step": 5401
+    },
+    {
+      "epoch": 2.680153903438004,
+      "grad_norm": 0.1288450806713303,
+      "learning_rate": 1.2291803388067284e-06,
+      "loss": 0.6895,
+      "step": 5402
+    },
+    {
+      "epoch": 2.6806503661412435,
+      "grad_norm": 0.13187999417119586,
+      "learning_rate": 1.2283387779932005e-06,
+      "loss": 0.7022,
+      "step": 5403
+    },
+    {
+      "epoch": 2.681146828844483,
+      "grad_norm": 0.12707167814369877,
+      "learning_rate": 1.2274974115327017e-06,
+      "loss": 0.7049,
+      "step": 5404
+    },
+    {
+      "epoch": 2.6816432915477226,
+      "grad_norm": 0.12903769384971894,
+      "learning_rate": 1.2266562395538198e-06,
+      "loss": 0.7194,
+      "step": 5405
+    },
+    {
+      "epoch": 2.682139754250962,
+      "grad_norm": 0.13367318620502683,
+      "learning_rate": 1.225815262185116e-06,
+      "loss": 0.6971,
+      "step": 5406
+    },
+    {
+      "epoch": 2.6826362169542013,
+      "grad_norm": 0.12972052003617227,
+      "learning_rate": 1.2249744795551198e-06,
+      "loss": 0.7434,
+      "step": 5407
+    },
+    {
+      "epoch": 2.6831326796574406,
+      "grad_norm": 0.1270398790488789,
+      "learning_rate": 1.2241338917923295e-06,
+      "loss": 0.6903,
+      "step": 5408
+    },
+    {
+      "epoch": 2.6836291423606804,
+      "grad_norm": 0.1273000126272403,
+      "learning_rate": 1.223293499025218e-06,
+      "loss": 0.6947,
+      "step": 5409
+    },
+    {
+      "epoch": 2.6841256050639197,
+      "grad_norm": 0.12789714754456988,
+      "learning_rate": 1.2224533013822237e-06,
+      "loss": 0.7394,
+      "step": 5410
+    },
+    {
+      "epoch": 2.684622067767159,
+      "grad_norm": 0.1284746698224543,
+      "learning_rate": 1.2216132989917592e-06,
+      "loss": 0.6935,
+      "step": 5411
+    },
+    {
+      "epoch": 2.6851185304703984,
+      "grad_norm": 0.128337900001136,
+      "learning_rate": 1.2207734919822047e-06,
+      "loss": 0.6772,
+      "step": 5412
+    },
+    {
+      "epoch": 2.6856149931736377,
+      "grad_norm": 0.1286919735265176,
+      "learning_rate": 1.2199338804819114e-06,
+      "loss": 0.696,
+      "step": 5413
+    },
+    {
+      "epoch": 2.686111455876877,
+      "grad_norm": 0.12513034231984738,
+      "learning_rate": 1.2190944646191999e-06,
+      "loss": 0.6521,
+      "step": 5414
+    },
+    {
+      "epoch": 2.686607918580117,
+      "grad_norm": 0.1371942316709031,
+      "learning_rate": 1.2182552445223609e-06,
+      "loss": 0.7441,
+      "step": 5415
+    },
+    {
+      "epoch": 2.687104381283356,
+      "grad_norm": 0.12633983584561082,
+      "learning_rate": 1.2174162203196575e-06,
+      "loss": 0.6826,
+      "step": 5416
+    },
+    {
+      "epoch": 2.6876008439865955,
+      "grad_norm": 0.1321860232959971,
+      "learning_rate": 1.216577392139319e-06,
+      "loss": 0.7607,
+      "step": 5417
+    },
+    {
+      "epoch": 2.688097306689835,
+      "grad_norm": 0.12852572536801632,
+      "learning_rate": 1.2157387601095492e-06,
+      "loss": 0.692,
+      "step": 5418
+    },
+    {
+      "epoch": 2.6885937693930746,
+      "grad_norm": 0.12900473130966916,
+      "learning_rate": 1.214900324358518e-06,
+      "loss": 0.7091,
+      "step": 5419
+    },
+    {
+      "epoch": 2.689090232096314,
+      "grad_norm": 0.12971154741512667,
+      "learning_rate": 1.2140620850143667e-06,
+      "loss": 0.7053,
+      "step": 5420
+    },
+    {
+      "epoch": 2.6895866947995533,
+      "grad_norm": 0.13521129476773464,
+      "learning_rate": 1.213224042205208e-06,
+      "loss": 0.7227,
+      "step": 5421
+    },
+    {
+      "epoch": 2.6900831575027926,
+      "grad_norm": 0.13226811800646973,
+      "learning_rate": 1.2123861960591224e-06,
+      "loss": 0.7147,
+      "step": 5422
+    },
+    {
+      "epoch": 2.690579620206032,
+      "grad_norm": 0.131271672560679,
+      "learning_rate": 1.2115485467041608e-06,
+      "loss": 0.7435,
+      "step": 5423
+    },
+    {
+      "epoch": 2.6910760829092712,
+      "grad_norm": 0.12939692136576958,
+      "learning_rate": 1.2107110942683459e-06,
+      "loss": 0.6961,
+      "step": 5424
+    },
+    {
+      "epoch": 2.691572545612511,
+      "grad_norm": 0.13543860120642587,
+      "learning_rate": 1.2098738388796668e-06,
+      "loss": 0.6976,
+      "step": 5425
+    },
+    {
+      "epoch": 2.6920690083157504,
+      "grad_norm": 0.1270996570143555,
+      "learning_rate": 1.2090367806660872e-06,
+      "loss": 0.693,
+      "step": 5426
+    },
+    {
+      "epoch": 2.6925654710189897,
+      "grad_norm": 0.13061886754156452,
+      "learning_rate": 1.2081999197555366e-06,
+      "loss": 0.6917,
+      "step": 5427
+    },
+    {
+      "epoch": 2.693061933722229,
+      "grad_norm": 0.13261693716322412,
+      "learning_rate": 1.2073632562759146e-06,
+      "loss": 0.7223,
+      "step": 5428
+    },
+    {
+      "epoch": 2.693558396425469,
+      "grad_norm": 0.12738184225408514,
+      "learning_rate": 1.2065267903550953e-06,
+      "loss": 0.6637,
+      "step": 5429
+    },
+    {
+      "epoch": 2.694054859128708,
+      "grad_norm": 0.12686598161535245,
+      "learning_rate": 1.2056905221209147e-06,
+      "loss": 0.6801,
+      "step": 5430
+    },
+    {
+      "epoch": 2.6945513218319475,
+      "grad_norm": 0.12651761385982638,
+      "learning_rate": 1.2048544517011863e-06,
+      "loss": 0.7168,
+      "step": 5431
+    },
+    {
+      "epoch": 2.695047784535187,
+      "grad_norm": 0.12731010103172005,
+      "learning_rate": 1.2040185792236874e-06,
+      "loss": 0.6676,
+      "step": 5432
+    },
+    {
+      "epoch": 2.695544247238426,
+      "grad_norm": 0.12581732529546605,
+      "learning_rate": 1.2031829048161705e-06,
+      "loss": 0.6796,
+      "step": 5433
+    },
+    {
+      "epoch": 2.6960407099416654,
+      "grad_norm": 0.13000955582734638,
+      "learning_rate": 1.2023474286063538e-06,
+      "loss": 0.708,
+      "step": 5434
+    },
+    {
+      "epoch": 2.696537172644905,
+      "grad_norm": 0.1320803282519354,
+      "learning_rate": 1.2015121507219254e-06,
+      "loss": 0.6949,
+      "step": 5435
+    },
+    {
+      "epoch": 2.6970336353481446,
+      "grad_norm": 0.1317763220856854,
+      "learning_rate": 1.2006770712905458e-06,
+      "loss": 0.6895,
+      "step": 5436
+    },
+    {
+      "epoch": 2.697530098051384,
+      "grad_norm": 0.12990531826150573,
+      "learning_rate": 1.1998421904398423e-06,
+      "loss": 0.7198,
+      "step": 5437
+    },
+    {
+      "epoch": 2.698026560754623,
+      "grad_norm": 0.1383883572643517,
+      "learning_rate": 1.199007508297414e-06,
+      "loss": 0.7659,
+      "step": 5438
+    },
+    {
+      "epoch": 2.698523023457863,
+      "grad_norm": 0.12440590803020435,
+      "learning_rate": 1.1981730249908282e-06,
+      "loss": 0.6478,
+      "step": 5439
+    },
+    {
+      "epoch": 2.6990194861611023,
+      "grad_norm": 0.12784078674498142,
+      "learning_rate": 1.1973387406476216e-06,
+      "loss": 0.6741,
+      "step": 5440
+    },
+    {
+      "epoch": 2.6995159488643417,
+      "grad_norm": 0.13239096818217247,
+      "learning_rate": 1.196504655395303e-06,
+      "loss": 0.7346,
+      "step": 5441
+    },
+    {
+      "epoch": 2.700012411567581,
+      "grad_norm": 0.12753915589030973,
+      "learning_rate": 1.1956707693613468e-06,
+      "loss": 0.6775,
+      "step": 5442
+    },
+    {
+      "epoch": 2.7005088742708203,
+      "grad_norm": 0.12968903089581899,
+      "learning_rate": 1.194837082673201e-06,
+      "loss": 0.7184,
+      "step": 5443
+    },
+    {
+      "epoch": 2.7010053369740596,
+      "grad_norm": 0.12351868593377198,
+      "learning_rate": 1.1940035954582803e-06,
+      "loss": 0.7085,
+      "step": 5444
+    },
+    {
+      "epoch": 2.701501799677299,
+      "grad_norm": 0.13181069411223056,
+      "learning_rate": 1.1931703078439705e-06,
+      "loss": 0.7328,
+      "step": 5445
+    },
+    {
+      "epoch": 2.7019982623805388,
+      "grad_norm": 0.1330437926631701,
+      "learning_rate": 1.1923372199576252e-06,
+      "loss": 0.7267,
+      "step": 5446
+    },
+    {
+      "epoch": 2.702494725083778,
+      "grad_norm": 0.12415515084900948,
+      "learning_rate": 1.1915043319265684e-06,
+      "loss": 0.6483,
+      "step": 5447
+    },
+    {
+      "epoch": 2.7029911877870174,
+      "grad_norm": 0.12773854918997632,
+      "learning_rate": 1.1906716438780952e-06,
+      "loss": 0.6881,
+      "step": 5448
+    },
+    {
+      "epoch": 2.703487650490257,
+      "grad_norm": 0.13125807339891024,
+      "learning_rate": 1.1898391559394668e-06,
+      "loss": 0.708,
+      "step": 5449
+    },
+    {
+      "epoch": 2.7039841131934965,
+      "grad_norm": 0.12144720415007257,
+      "learning_rate": 1.1890068682379175e-06,
+      "loss": 0.6287,
+      "step": 5450
+    },
+    {
+      "epoch": 2.704480575896736,
+      "grad_norm": 0.13122189521060085,
+      "learning_rate": 1.1881747809006483e-06,
+      "loss": 0.7784,
+      "step": 5451
+    },
+    {
+      "epoch": 2.704977038599975,
+      "grad_norm": 0.12593554383340527,
+      "learning_rate": 1.1873428940548293e-06,
+      "loss": 0.7264,
+      "step": 5452
+    },
+    {
+      "epoch": 2.7054735013032145,
+      "grad_norm": 0.13034835041902,
+      "learning_rate": 1.1865112078276032e-06,
+      "loss": 0.6816,
+      "step": 5453
+    },
+    {
+      "epoch": 2.705969964006454,
+      "grad_norm": 0.12793803176190882,
+      "learning_rate": 1.1856797223460776e-06,
+      "loss": 0.6852,
+      "step": 5454
+    },
+    {
+      "epoch": 2.706466426709693,
+      "grad_norm": 0.13001908847017454,
+      "learning_rate": 1.1848484377373336e-06,
+      "loss": 0.7136,
+      "step": 5455
+    },
+    {
+      "epoch": 2.706962889412933,
+      "grad_norm": 0.1244529234238611,
+      "learning_rate": 1.1840173541284198e-06,
+      "loss": 0.6583,
+      "step": 5456
+    },
+    {
+      "epoch": 2.7074593521161723,
+      "grad_norm": 0.12496088308943552,
+      "learning_rate": 1.1831864716463517e-06,
+      "loss": 0.6996,
+      "step": 5457
+    },
+    {
+      "epoch": 2.7079558148194116,
+      "grad_norm": 0.13009009109117492,
+      "learning_rate": 1.182355790418119e-06,
+      "loss": 0.7313,
+      "step": 5458
+    },
+    {
+      "epoch": 2.7084522775226514,
+      "grad_norm": 0.13302558329128752,
+      "learning_rate": 1.181525310570677e-06,
+      "loss": 0.7344,
+      "step": 5459
+    },
+    {
+      "epoch": 2.7089487402258907,
+      "grad_norm": 0.13255724713593725,
+      "learning_rate": 1.1806950322309503e-06,
+      "loss": 0.7073,
+      "step": 5460
+    },
+    {
+      "epoch": 2.70944520292913,
+      "grad_norm": 0.12830942629616623,
+      "learning_rate": 1.1798649555258359e-06,
+      "loss": 0.6774,
+      "step": 5461
+    },
+    {
+      "epoch": 2.7099416656323694,
+      "grad_norm": 0.13135455282779482,
+      "learning_rate": 1.1790350805821948e-06,
+      "loss": 0.6984,
+      "step": 5462
+    },
+    {
+      "epoch": 2.7104381283356087,
+      "grad_norm": 0.12869925272920885,
+      "learning_rate": 1.1782054075268626e-06,
+      "loss": 0.6924,
+      "step": 5463
+    },
+    {
+      "epoch": 2.710934591038848,
+      "grad_norm": 0.12635830105891932,
+      "learning_rate": 1.1773759364866394e-06,
+      "loss": 0.6704,
+      "step": 5464
+    },
+    {
+      "epoch": 2.7114310537420874,
+      "grad_norm": 0.1267057090577105,
+      "learning_rate": 1.1765466675882983e-06,
+      "loss": 0.6603,
+      "step": 5465
+    },
+    {
+      "epoch": 2.711927516445327,
+      "grad_norm": 0.1379598054720685,
+      "learning_rate": 1.1757176009585795e-06,
+      "loss": 0.7162,
+      "step": 5466
+    },
+    {
+      "epoch": 2.7124239791485665,
+      "grad_norm": 0.1267327103751104,
+      "learning_rate": 1.1748887367241913e-06,
+      "loss": 0.6841,
+      "step": 5467
+    },
+    {
+      "epoch": 2.712920441851806,
+      "grad_norm": 0.12581262787382652,
+      "learning_rate": 1.1740600750118136e-06,
+      "loss": 0.7009,
+      "step": 5468
+    },
+    {
+      "epoch": 2.713416904555045,
+      "grad_norm": 0.12485464413297796,
+      "learning_rate": 1.173231615948093e-06,
+      "loss": 0.6748,
+      "step": 5469
+    },
+    {
+      "epoch": 2.713913367258285,
+      "grad_norm": 0.12620028737405145,
+      "learning_rate": 1.1724033596596477e-06,
+      "loss": 0.688,
+      "step": 5470
+    },
+    {
+      "epoch": 2.7144098299615242,
+      "grad_norm": 0.12721872175292415,
+      "learning_rate": 1.1715753062730622e-06,
+      "loss": 0.7245,
+      "step": 5471
+    },
+    {
+      "epoch": 2.7149062926647636,
+      "grad_norm": 0.13289322932822414,
+      "learning_rate": 1.170747455914891e-06,
+      "loss": 0.7097,
+      "step": 5472
+    },
+    {
+      "epoch": 2.715402755368003,
+      "grad_norm": 0.12728705232851031,
+      "learning_rate": 1.169919808711659e-06,
+      "loss": 0.6834,
+      "step": 5473
+    },
+    {
+      "epoch": 2.7158992180712422,
+      "grad_norm": 0.1267143167247784,
+      "learning_rate": 1.169092364789857e-06,
+      "loss": 0.6926,
+      "step": 5474
+    },
+    {
+      "epoch": 2.7163956807744816,
+      "grad_norm": 0.13082421386045198,
+      "learning_rate": 1.1682651242759483e-06,
+      "loss": 0.7155,
+      "step": 5475
+    },
+    {
+      "epoch": 2.7168921434777213,
+      "grad_norm": 0.1313656554797384,
+      "learning_rate": 1.1674380872963629e-06,
+      "loss": 0.689,
+      "step": 5476
+    },
+    {
+      "epoch": 2.7173886061809607,
+      "grad_norm": 0.12776895465581478,
+      "learning_rate": 1.1666112539774998e-06,
+      "loss": 0.6669,
+      "step": 5477
+    },
+    {
+      "epoch": 2.7178850688842,
+      "grad_norm": 0.12847066772569427,
+      "learning_rate": 1.1657846244457272e-06,
+      "loss": 0.7357,
+      "step": 5478
+    },
+    {
+      "epoch": 2.7183815315874393,
+      "grad_norm": 0.13314280685450303,
+      "learning_rate": 1.1649581988273814e-06,
+      "loss": 0.7503,
+      "step": 5479
+    },
+    {
+      "epoch": 2.718877994290679,
+      "grad_norm": 0.13052451777587085,
+      "learning_rate": 1.16413197724877e-06,
+      "loss": 0.7349,
+      "step": 5480
+    },
+    {
+      "epoch": 2.7193744569939184,
+      "grad_norm": 0.1300407125579856,
+      "learning_rate": 1.163305959836166e-06,
+      "loss": 0.7083,
+      "step": 5481
+    },
+    {
+      "epoch": 2.7198709196971578,
+      "grad_norm": 0.1280550112986663,
+      "learning_rate": 1.1624801467158145e-06,
+      "loss": 0.6762,
+      "step": 5482
+    },
+    {
+      "epoch": 2.720367382400397,
+      "grad_norm": 0.12510160831325617,
+      "learning_rate": 1.1616545380139272e-06,
+      "loss": 0.6925,
+      "step": 5483
+    },
+    {
+      "epoch": 2.7208638451036364,
+      "grad_norm": 0.13196997248908202,
+      "learning_rate": 1.1608291338566841e-06,
+      "loss": 0.7083,
+      "step": 5484
+    },
+    {
+      "epoch": 2.7213603078068758,
+      "grad_norm": 0.1340590026668367,
+      "learning_rate": 1.1600039343702368e-06,
+      "loss": 0.6946,
+      "step": 5485
+    },
+    {
+      "epoch": 2.7218567705101155,
+      "grad_norm": 0.12829686594605474,
+      "learning_rate": 1.1591789396807021e-06,
+      "loss": 0.705,
+      "step": 5486
+    },
+    {
+      "epoch": 2.722353233213355,
+      "grad_norm": 0.12881800317843775,
+      "learning_rate": 1.158354149914169e-06,
+      "loss": 0.6883,
+      "step": 5487
+    },
+    {
+      "epoch": 2.722849695916594,
+      "grad_norm": 0.12960535888237937,
+      "learning_rate": 1.1575295651966926e-06,
+      "loss": 0.6959,
+      "step": 5488
+    },
+    {
+      "epoch": 2.7233461586198335,
+      "grad_norm": 0.12426299232658077,
+      "learning_rate": 1.156705185654296e-06,
+      "loss": 0.679,
+      "step": 5489
+    },
+    {
+      "epoch": 2.7238426213230733,
+      "grad_norm": 0.1267642649577807,
+      "learning_rate": 1.1558810114129746e-06,
+      "loss": 0.6957,
+      "step": 5490
+    },
+    {
+      "epoch": 2.7243390840263126,
+      "grad_norm": 0.13057816997613797,
+      "learning_rate": 1.1550570425986884e-06,
+      "loss": 0.6982,
+      "step": 5491
+    },
+    {
+      "epoch": 2.724835546729552,
+      "grad_norm": 0.13036966095415778,
+      "learning_rate": 1.1542332793373699e-06,
+      "loss": 0.7108,
+      "step": 5492
+    },
+    {
+      "epoch": 2.7253320094327913,
+      "grad_norm": 0.12845031666688522,
+      "learning_rate": 1.1534097217549167e-06,
+      "loss": 0.7045,
+      "step": 5493
+    },
+    {
+      "epoch": 2.7258284721360306,
+      "grad_norm": 0.13431116362677278,
+      "learning_rate": 1.1525863699771967e-06,
+      "loss": 0.7211,
+      "step": 5494
+    },
+    {
+      "epoch": 2.72632493483927,
+      "grad_norm": 0.12701692487109137,
+      "learning_rate": 1.1517632241300457e-06,
+      "loss": 0.6845,
+      "step": 5495
+    },
+    {
+      "epoch": 2.7268213975425097,
+      "grad_norm": 0.12491394627771943,
+      "learning_rate": 1.1509402843392681e-06,
+      "loss": 0.658,
+      "step": 5496
+    },
+    {
+      "epoch": 2.727317860245749,
+      "grad_norm": 0.12829490733233098,
+      "learning_rate": 1.150117550730638e-06,
+      "loss": 0.7134,
+      "step": 5497
+    },
+    {
+      "epoch": 2.7278143229489884,
+      "grad_norm": 0.127816041512245,
+      "learning_rate": 1.1492950234298965e-06,
+      "loss": 0.6658,
+      "step": 5498
+    },
+    {
+      "epoch": 2.7283107856522277,
+      "grad_norm": 0.12694117469603244,
+      "learning_rate": 1.1484727025627532e-06,
+      "loss": 0.7073,
+      "step": 5499
+    },
+    {
+      "epoch": 2.7288072483554675,
+      "grad_norm": 0.1273683408910965,
+      "learning_rate": 1.147650588254888e-06,
+      "loss": 0.7044,
+      "step": 5500
+    },
+    {
+      "epoch": 2.729303711058707,
+      "grad_norm": 0.12716535637041657,
+      "learning_rate": 1.1468286806319461e-06,
+      "loss": 0.7027,
+      "step": 5501
+    },
+    {
+      "epoch": 2.729800173761946,
+      "grad_norm": 0.13652878890308245,
+      "learning_rate": 1.146006979819545e-06,
+      "loss": 0.7242,
+      "step": 5502
+    },
+    {
+      "epoch": 2.7302966364651855,
+      "grad_norm": 0.13060886134753424,
+      "learning_rate": 1.1451854859432674e-06,
+      "loss": 0.6823,
+      "step": 5503
+    },
+    {
+      "epoch": 2.730793099168425,
+      "grad_norm": 0.12949358227740704,
+      "learning_rate": 1.1443641991286644e-06,
+      "loss": 0.6812,
+      "step": 5504
+    },
+    {
+      "epoch": 2.731289561871664,
+      "grad_norm": 0.1484208011714529,
+      "learning_rate": 1.1435431195012586e-06,
+      "loss": 0.7169,
+      "step": 5505
+    },
+    {
+      "epoch": 2.731786024574904,
+      "grad_norm": 0.12279528726656044,
+      "learning_rate": 1.1427222471865368e-06,
+      "loss": 0.6499,
+      "step": 5506
+    },
+    {
+      "epoch": 2.7322824872781433,
+      "grad_norm": 0.12686401827050103,
+      "learning_rate": 1.1419015823099582e-06,
+      "loss": 0.6969,
+      "step": 5507
+    },
+    {
+      "epoch": 2.7327789499813826,
+      "grad_norm": 0.1304912606677775,
+      "learning_rate": 1.1410811249969475e-06,
+      "loss": 0.7185,
+      "step": 5508
+    },
+    {
+      "epoch": 2.733275412684622,
+      "grad_norm": 0.1307385183998749,
+      "learning_rate": 1.140260875372898e-06,
+      "loss": 0.7018,
+      "step": 5509
+    },
+    {
+      "epoch": 2.7337718753878617,
+      "grad_norm": 0.12879059404077767,
+      "learning_rate": 1.1394408335631721e-06,
+      "loss": 0.7036,
+      "step": 5510
+    },
+    {
+      "epoch": 2.734268338091101,
+      "grad_norm": 0.13168818270683702,
+      "learning_rate": 1.138620999693099e-06,
+      "loss": 0.733,
+      "step": 5511
+    },
+    {
+      "epoch": 2.7347648007943404,
+      "grad_norm": 0.1302130132221126,
+      "learning_rate": 1.1378013738879787e-06,
+      "loss": 0.7118,
+      "step": 5512
+    },
+    {
+      "epoch": 2.7352612634975797,
+      "grad_norm": 0.12664057418388594,
+      "learning_rate": 1.1369819562730763e-06,
+      "loss": 0.692,
+      "step": 5513
+    },
+    {
+      "epoch": 2.735757726200819,
+      "grad_norm": 0.1787042284319172,
+      "learning_rate": 1.1361627469736286e-06,
+      "loss": 0.6932,
+      "step": 5514
+    },
+    {
+      "epoch": 2.7362541889040584,
+      "grad_norm": 0.13276835251453653,
+      "learning_rate": 1.1353437461148378e-06,
+      "loss": 0.685,
+      "step": 5515
+    },
+    {
+      "epoch": 2.736750651607298,
+      "grad_norm": 0.12840095500860307,
+      "learning_rate": 1.1345249538218736e-06,
+      "loss": 0.6855,
+      "step": 5516
+    },
+    {
+      "epoch": 2.7372471143105375,
+      "grad_norm": 0.13118594667050903,
+      "learning_rate": 1.1337063702198775e-06,
+      "loss": 0.7603,
+      "step": 5517
+    },
+    {
+      "epoch": 2.737743577013777,
+      "grad_norm": 0.12576209372358757,
+      "learning_rate": 1.1328879954339546e-06,
+      "loss": 0.6698,
+      "step": 5518
+    },
+    {
+      "epoch": 2.738240039717016,
+      "grad_norm": 0.133093950748946,
+      "learning_rate": 1.132069829589183e-06,
+      "loss": 0.6886,
+      "step": 5519
+    },
+    {
+      "epoch": 2.738736502420256,
+      "grad_norm": 0.12864037736931044,
+      "learning_rate": 1.1312518728106048e-06,
+      "loss": 0.6766,
+      "step": 5520
+    },
+    {
+      "epoch": 2.7392329651234952,
+      "grad_norm": 0.12945382840659705,
+      "learning_rate": 1.1304341252232307e-06,
+      "loss": 0.6828,
+      "step": 5521
+    },
+    {
+      "epoch": 2.7397294278267346,
+      "grad_norm": 0.1308218826995222,
+      "learning_rate": 1.129616586952042e-06,
+      "loss": 0.6656,
+      "step": 5522
+    },
+    {
+      "epoch": 2.740225890529974,
+      "grad_norm": 0.1314567316421182,
+      "learning_rate": 1.1287992581219846e-06,
+      "loss": 0.7036,
+      "step": 5523
+    },
+    {
+      "epoch": 2.7407223532332132,
+      "grad_norm": 0.1366700099747632,
+      "learning_rate": 1.1279821388579762e-06,
+      "loss": 0.7349,
+      "step": 5524
+    },
+    {
+      "epoch": 2.7412188159364526,
+      "grad_norm": 0.1309681880536747,
+      "learning_rate": 1.1271652292848988e-06,
+      "loss": 0.6795,
+      "step": 5525
+    },
+    {
+      "epoch": 2.7417152786396923,
+      "grad_norm": 0.12837208221450544,
+      "learning_rate": 1.126348529527605e-06,
+      "loss": 0.6809,
+      "step": 5526
+    },
+    {
+      "epoch": 2.7422117413429317,
+      "grad_norm": 0.131214461157075,
+      "learning_rate": 1.1255320397109132e-06,
+      "loss": 0.7304,
+      "step": 5527
+    },
+    {
+      "epoch": 2.742708204046171,
+      "grad_norm": 0.12712062614786912,
+      "learning_rate": 1.1247157599596103e-06,
+      "loss": 0.6985,
+      "step": 5528
+    },
+    {
+      "epoch": 2.7432046667494103,
+      "grad_norm": 0.13592701398261714,
+      "learning_rate": 1.1238996903984537e-06,
+      "loss": 0.763,
+      "step": 5529
+    },
+    {
+      "epoch": 2.74370112945265,
+      "grad_norm": 0.12718223517036215,
+      "learning_rate": 1.1230838311521642e-06,
+      "loss": 0.6984,
+      "step": 5530
+    },
+    {
+      "epoch": 2.7441975921558894,
+      "grad_norm": 0.13534565978463103,
+      "learning_rate": 1.1222681823454349e-06,
+      "loss": 0.7765,
+      "step": 5531
+    },
+    {
+      "epoch": 2.7446940548591288,
+      "grad_norm": 0.13626701847329473,
+      "learning_rate": 1.121452744102924e-06,
+      "loss": 0.7439,
+      "step": 5532
+    },
+    {
+      "epoch": 2.745190517562368,
+      "grad_norm": 0.13284297253885471,
+      "learning_rate": 1.1206375165492564e-06,
+      "loss": 0.7347,
+      "step": 5533
+    },
+    {
+      "epoch": 2.7456869802656074,
+      "grad_norm": 0.13079426764125932,
+      "learning_rate": 1.119822499809029e-06,
+      "loss": 0.6999,
+      "step": 5534
+    },
+    {
+      "epoch": 2.7461834429688468,
+      "grad_norm": 0.12570608638802389,
+      "learning_rate": 1.1190076940068031e-06,
+      "loss": 0.6831,
+      "step": 5535
+    },
+    {
+      "epoch": 2.7466799056720865,
+      "grad_norm": 0.12564077502899773,
+      "learning_rate": 1.1181930992671078e-06,
+      "loss": 0.6883,
+      "step": 5536
+    },
+    {
+      "epoch": 2.747176368375326,
+      "grad_norm": 0.12497737842928656,
+      "learning_rate": 1.1173787157144425e-06,
+      "loss": 0.6551,
+      "step": 5537
+    },
+    {
+      "epoch": 2.747672831078565,
+      "grad_norm": 0.13189163059908987,
+      "learning_rate": 1.116564543473271e-06,
+      "loss": 0.7169,
+      "step": 5538
+    },
+    {
+      "epoch": 2.7481692937818045,
+      "grad_norm": 0.13114812942001947,
+      "learning_rate": 1.1157505826680285e-06,
+      "loss": 0.6959,
+      "step": 5539
+    },
+    {
+      "epoch": 2.7486657564850443,
+      "grad_norm": 0.13908344915755239,
+      "learning_rate": 1.1149368334231146e-06,
+      "loss": 0.7409,
+      "step": 5540
+    },
+    {
+      "epoch": 2.7491622191882836,
+      "grad_norm": 0.13448810312163784,
+      "learning_rate": 1.1141232958628976e-06,
+      "loss": 0.7048,
+      "step": 5541
+    },
+    {
+      "epoch": 2.749658681891523,
+      "grad_norm": 0.13514802207833446,
+      "learning_rate": 1.1133099701117143e-06,
+      "loss": 0.7112,
+      "step": 5542
+    },
+    {
+      "epoch": 2.7501551445947623,
+      "grad_norm": 0.12623353087443057,
+      "learning_rate": 1.112496856293867e-06,
+      "loss": 0.7261,
+      "step": 5543
+    },
+    {
+      "epoch": 2.7506516072980016,
+      "grad_norm": 0.12458653015367169,
+      "learning_rate": 1.1116839545336292e-06,
+      "loss": 0.6775,
+      "step": 5544
+    },
+    {
+      "epoch": 2.7506516072980016,
+      "eval_loss": 0.7236723899841309,
+      "eval_runtime": 135.7629,
+      "eval_samples_per_second": 223.574,
+      "eval_steps_per_second": 27.953,
+      "step": 5544
+    },
+    {
+      "epoch": 2.751148070001241,
+      "grad_norm": 0.1306204466783865,
+      "learning_rate": 1.1108712649552384e-06,
+      "loss": 0.6945,
+      "step": 5545
+    },
+    {
+      "epoch": 2.7516445327044807,
+      "grad_norm": 0.12369576532652772,
+      "learning_rate": 1.1100587876829024e-06,
+      "loss": 0.6289,
+      "step": 5546
+    },
+    {
+      "epoch": 2.75214099540772,
+      "grad_norm": 0.12789165154828827,
+      "learning_rate": 1.1092465228407949e-06,
+      "loss": 0.6195,
+      "step": 5547
+    },
+    {
+      "epoch": 2.7526374581109594,
+      "grad_norm": 0.1266994829459876,
+      "learning_rate": 1.1084344705530561e-06,
+      "loss": 0.7413,
+      "step": 5548
+    },
+    {
+      "epoch": 2.7531339208141987,
+      "grad_norm": 0.1268540066373866,
+      "learning_rate": 1.1076226309437977e-06,
+      "loss": 0.6826,
+      "step": 5549
+    },
+    {
+      "epoch": 2.7536303835174385,
+      "grad_norm": 0.12618417321246705,
+      "learning_rate": 1.1068110041370938e-06,
+      "loss": 0.6762,
+      "step": 5550
+    },
+    {
+      "epoch": 2.754126846220678,
+      "grad_norm": 0.1325579883265857,
+      "learning_rate": 1.1059995902569911e-06,
+      "loss": 0.7399,
+      "step": 5551
+    },
+    {
+      "epoch": 2.754623308923917,
+      "grad_norm": 0.12889480405547812,
+      "learning_rate": 1.1051883894274998e-06,
+      "loss": 0.7098,
+      "step": 5552
+    },
+    {
+      "epoch": 2.7551197716271565,
+      "grad_norm": 0.135934126639656,
+      "learning_rate": 1.104377401772598e-06,
+      "loss": 0.7098,
+      "step": 5553
+    },
+    {
+      "epoch": 2.755616234330396,
+      "grad_norm": 0.1289564414190799,
+      "learning_rate": 1.1035666274162344e-06,
+      "loss": 0.7435,
+      "step": 5554
+    },
+    {
+      "epoch": 2.756112697033635,
+      "grad_norm": 0.12736214807761737,
+      "learning_rate": 1.1027560664823208e-06,
+      "loss": 0.6656,
+      "step": 5555
+    },
+    {
+      "epoch": 2.756609159736875,
+      "grad_norm": 0.12848892854207852,
+      "learning_rate": 1.10194571909474e-06,
+      "loss": 0.7014,
+      "step": 5556
+    },
+    {
+      "epoch": 2.7571056224401143,
+      "grad_norm": 0.1321006717534794,
+      "learning_rate": 1.10113558537734e-06,
+      "loss": 0.7015,
+      "step": 5557
+    },
+    {
+      "epoch": 2.7576020851433536,
+      "grad_norm": 0.12854999523007427,
+      "learning_rate": 1.100325665453937e-06,
+      "loss": 0.6767,
+      "step": 5558
+    },
+    {
+      "epoch": 2.758098547846593,
+      "grad_norm": 0.1279414734388252,
+      "learning_rate": 1.0995159594483138e-06,
+      "loss": 0.696,
+      "step": 5559
+    },
+    {
+      "epoch": 2.7585950105498327,
+      "grad_norm": 0.13005801919260618,
+      "learning_rate": 1.09870646748422e-06,
+      "loss": 0.73,
+      "step": 5560
+    },
+    {
+      "epoch": 2.759091473253072,
+      "grad_norm": 0.1360761488502302,
+      "learning_rate": 1.0978971896853758e-06,
+      "loss": 0.7399,
+      "step": 5561
+    },
+    {
+      "epoch": 2.7595879359563114,
+      "grad_norm": 0.13408584165708903,
+      "learning_rate": 1.0970881261754641e-06,
+      "loss": 0.7369,
+      "step": 5562
+    },
+    {
+      "epoch": 2.7600843986595507,
+      "grad_norm": 0.12871987297525,
+      "learning_rate": 1.096279277078139e-06,
+      "loss": 0.7009,
+      "step": 5563
+    },
+    {
+      "epoch": 2.76058086136279,
+      "grad_norm": 0.12990364606695212,
+      "learning_rate": 1.0954706425170198e-06,
+      "loss": 0.6468,
+      "step": 5564
+    },
+    {
+      "epoch": 2.7610773240660293,
+      "grad_norm": 0.1252919783584692,
+      "learning_rate": 1.094662222615692e-06,
+      "loss": 0.7253,
+      "step": 5565
+    },
+    {
+      "epoch": 2.761573786769269,
+      "grad_norm": 0.1326764590455894,
+      "learning_rate": 1.0938540174977115e-06,
+      "loss": 0.7167,
+      "step": 5566
+    },
+    {
+      "epoch": 2.7620702494725085,
+      "grad_norm": 0.13247387675260933,
+      "learning_rate": 1.0930460272865976e-06,
+      "loss": 0.7151,
+      "step": 5567
+    },
+    {
+      "epoch": 2.762566712175748,
+      "grad_norm": 0.13190057787514423,
+      "learning_rate": 1.0922382521058405e-06,
+      "loss": 0.7643,
+      "step": 5568
+    },
+    {
+      "epoch": 2.763063174878987,
+      "grad_norm": 0.12934627161113424,
+      "learning_rate": 1.091430692078895e-06,
+      "loss": 0.7093,
+      "step": 5569
+    },
+    {
+      "epoch": 2.763559637582227,
+      "grad_norm": 0.1265845815519934,
+      "learning_rate": 1.0906233473291827e-06,
+      "loss": 0.6742,
+      "step": 5570
+    },
+    {
+      "epoch": 2.764056100285466,
+      "grad_norm": 0.1344552360678481,
+      "learning_rate": 1.0898162179800948e-06,
+      "loss": 0.7388,
+      "step": 5571
+    },
+    {
+      "epoch": 2.7645525629887056,
+      "grad_norm": 0.12498891326605503,
+      "learning_rate": 1.0890093041549873e-06,
+      "loss": 0.6614,
+      "step": 5572
+    },
+    {
+      "epoch": 2.765049025691945,
+      "grad_norm": 0.1289570046975174,
+      "learning_rate": 1.0882026059771845e-06,
+      "loss": 0.7288,
+      "step": 5573
+    },
+    {
+      "epoch": 2.765545488395184,
+      "grad_norm": 0.12524071202980924,
+      "learning_rate": 1.0873961235699759e-06,
+      "loss": 0.6666,
+      "step": 5574
+    },
+    {
+      "epoch": 2.7660419510984235,
+      "grad_norm": 0.13322599761967555,
+      "learning_rate": 1.0865898570566212e-06,
+      "loss": 0.7645,
+      "step": 5575
+    },
+    {
+      "epoch": 2.766538413801663,
+      "grad_norm": 0.13519871420996185,
+      "learning_rate": 1.0857838065603447e-06,
+      "loss": 0.7605,
+      "step": 5576
+    },
+    {
+      "epoch": 2.7670348765049027,
+      "grad_norm": 0.13070973410690823,
+      "learning_rate": 1.084977972204337e-06,
+      "loss": 0.6832,
+      "step": 5577
+    },
+    {
+      "epoch": 2.767531339208142,
+      "grad_norm": 0.13143077641170847,
+      "learning_rate": 1.0841723541117594e-06,
+      "loss": 0.7178,
+      "step": 5578
+    },
+    {
+      "epoch": 2.7680278019113813,
+      "grad_norm": 0.13523115464259342,
+      "learning_rate": 1.083366952405736e-06,
+      "loss": 0.7738,
+      "step": 5579
+    },
+    {
+      "epoch": 2.768524264614621,
+      "grad_norm": 0.13073847691026474,
+      "learning_rate": 1.0825617672093592e-06,
+      "loss": 0.7124,
+      "step": 5580
+    },
+    {
+      "epoch": 2.7690207273178604,
+      "grad_norm": 0.12972826652950495,
+      "learning_rate": 1.0817567986456904e-06,
+      "loss": 0.6924,
+      "step": 5581
+    },
+    {
+      "epoch": 2.7695171900210998,
+      "grad_norm": 0.13008197120598594,
+      "learning_rate": 1.0809520468377541e-06,
+      "loss": 0.7238,
+      "step": 5582
+    },
+    {
+      "epoch": 2.770013652724339,
+      "grad_norm": 0.1297808815362704,
+      "learning_rate": 1.0801475119085455e-06,
+      "loss": 0.7345,
+      "step": 5583
+    },
+    {
+      "epoch": 2.7705101154275784,
+      "grad_norm": 0.13507574439437295,
+      "learning_rate": 1.0793431939810243e-06,
+      "loss": 0.7121,
+      "step": 5584
+    },
+    {
+      "epoch": 2.7710065781308177,
+      "grad_norm": 0.1273641822305587,
+      "learning_rate": 1.0785390931781164e-06,
+      "loss": 0.6972,
+      "step": 5585
+    },
+    {
+      "epoch": 2.771503040834057,
+      "grad_norm": 0.13271462621019983,
+      "learning_rate": 1.0777352096227174e-06,
+      "loss": 0.7522,
+      "step": 5586
+    },
+    {
+      "epoch": 2.771999503537297,
+      "grad_norm": 0.13302101809288855,
+      "learning_rate": 1.076931543437687e-06,
+      "loss": 0.7542,
+      "step": 5587
+    },
+    {
+      "epoch": 2.772495966240536,
+      "grad_norm": 0.12899005010841755,
+      "learning_rate": 1.0761280947458536e-06,
+      "loss": 0.6468,
+      "step": 5588
+    },
+    {
+      "epoch": 2.7729924289437755,
+      "grad_norm": 0.12421829548371563,
+      "learning_rate": 1.0753248636700109e-06,
+      "loss": 0.6886,
+      "step": 5589
+    },
+    {
+      "epoch": 2.7734888916470153,
+      "grad_norm": 0.12349016996371331,
+      "learning_rate": 1.0745218503329196e-06,
+      "loss": 0.7158,
+      "step": 5590
+    },
+    {
+      "epoch": 2.7739853543502546,
+      "grad_norm": 0.12644575429599572,
+      "learning_rate": 1.0737190548573082e-06,
+      "loss": 0.6762,
+      "step": 5591
+    },
+    {
+      "epoch": 2.774481817053494,
+      "grad_norm": 0.134593415267307,
+      "learning_rate": 1.0729164773658692e-06,
+      "loss": 0.7065,
+      "step": 5592
+    },
+    {
+      "epoch": 2.7749782797567333,
+      "grad_norm": 0.1260483274577327,
+      "learning_rate": 1.0721141179812664e-06,
+      "loss": 0.6827,
+      "step": 5593
+    },
+    {
+      "epoch": 2.7754747424599726,
+      "grad_norm": 0.12812302335606984,
+      "learning_rate": 1.0713119768261248e-06,
+      "loss": 0.6981,
+      "step": 5594
+    },
+    {
+      "epoch": 2.775971205163212,
+      "grad_norm": 0.12876922838504537,
+      "learning_rate": 1.0705100540230418e-06,
+      "loss": 0.7289,
+      "step": 5595
+    },
+    {
+      "epoch": 2.7764676678664513,
+      "grad_norm": 0.12571235400921374,
+      "learning_rate": 1.0697083496945766e-06,
+      "loss": 0.6929,
+      "step": 5596
+    },
+    {
+      "epoch": 2.776964130569691,
+      "grad_norm": 0.13199476837674784,
+      "learning_rate": 1.0689068639632563e-06,
+      "loss": 0.7835,
+      "step": 5597
+    },
+    {
+      "epoch": 2.7774605932729304,
+      "grad_norm": 0.1273711543359699,
+      "learning_rate": 1.0681055969515769e-06,
+      "loss": 0.7014,
+      "step": 5598
+    },
+    {
+      "epoch": 2.7779570559761697,
+      "grad_norm": 0.1269043028906525,
+      "learning_rate": 1.0673045487819975e-06,
+      "loss": 0.69,
+      "step": 5599
+    },
+    {
+      "epoch": 2.7784535186794095,
+      "grad_norm": 0.1319110742532422,
+      "learning_rate": 1.066503719576947e-06,
+      "loss": 0.7461,
+      "step": 5600
+    },
+    {
+      "epoch": 2.778949981382649,
+      "grad_norm": 0.12579895622801868,
+      "learning_rate": 1.0657031094588184e-06,
+      "loss": 0.6868,
+      "step": 5601
+    },
+    {
+      "epoch": 2.779446444085888,
+      "grad_norm": 0.12882226269354222,
+      "learning_rate": 1.064902718549972e-06,
+      "loss": 0.7045,
+      "step": 5602
+    },
+    {
+      "epoch": 2.7799429067891275,
+      "grad_norm": 0.1270352468548072,
+      "learning_rate": 1.0641025469727356e-06,
+      "loss": 0.7087,
+      "step": 5603
+    },
+    {
+      "epoch": 2.780439369492367,
+      "grad_norm": 0.12900061673731678,
+      "learning_rate": 1.0633025948494014e-06,
+      "loss": 0.7028,
+      "step": 5604
+    },
+    {
+      "epoch": 2.780935832195606,
+      "grad_norm": 0.13016595296281186,
+      "learning_rate": 1.0625028623022305e-06,
+      "loss": 0.692,
+      "step": 5605
+    },
+    {
+      "epoch": 2.7814322948988455,
+      "grad_norm": 0.12885825274574203,
+      "learning_rate": 1.0617033494534486e-06,
+      "loss": 0.6744,
+      "step": 5606
+    },
+    {
+      "epoch": 2.7819287576020852,
+      "grad_norm": 0.12870191436048342,
+      "learning_rate": 1.0609040564252484e-06,
+      "loss": 0.6885,
+      "step": 5607
+    },
+    {
+      "epoch": 2.7824252203053246,
+      "grad_norm": 0.12629534065315853,
+      "learning_rate": 1.0601049833397892e-06,
+      "loss": 0.6859,
+      "step": 5608
+    },
+    {
+      "epoch": 2.782921683008564,
+      "grad_norm": 0.13520278743342204,
+      "learning_rate": 1.0593061303191954e-06,
+      "loss": 0.7816,
+      "step": 5609
+    },
+    {
+      "epoch": 2.7834181457118032,
+      "grad_norm": 0.12636004627171285,
+      "learning_rate": 1.0585074974855605e-06,
+      "loss": 0.7254,
+      "step": 5610
+    },
+    {
+      "epoch": 2.783914608415043,
+      "grad_norm": 0.1258511623485038,
+      "learning_rate": 1.0577090849609415e-06,
+      "loss": 0.668,
+      "step": 5611
+    },
+    {
+      "epoch": 2.7844110711182823,
+      "grad_norm": 0.1268446329679461,
+      "learning_rate": 1.0569108928673642e-06,
+      "loss": 0.6671,
+      "step": 5612
+    },
+    {
+      "epoch": 2.7849075338215217,
+      "grad_norm": 0.13239241247612957,
+      "learning_rate": 1.0561129213268187e-06,
+      "loss": 0.7689,
+      "step": 5613
+    },
+    {
+      "epoch": 2.785403996524761,
+      "grad_norm": 0.1255651588944757,
+      "learning_rate": 1.0553151704612614e-06,
+      "loss": 0.6884,
+      "step": 5614
+    },
+    {
+      "epoch": 2.7859004592280003,
+      "grad_norm": 0.12247730286780059,
+      "learning_rate": 1.0545176403926172e-06,
+      "loss": 0.6955,
+      "step": 5615
+    },
+    {
+      "epoch": 2.7863969219312397,
+      "grad_norm": 0.12958608013612596,
+      "learning_rate": 1.0537203312427752e-06,
+      "loss": 0.6482,
+      "step": 5616
+    },
+    {
+      "epoch": 2.7868933846344794,
+      "grad_norm": 0.1228441492665452,
+      "learning_rate": 1.0529232431335903e-06,
+      "loss": 0.6562,
+      "step": 5617
+    },
+    {
+      "epoch": 2.7873898473377188,
+      "grad_norm": 0.13327477747426705,
+      "learning_rate": 1.0521263761868866e-06,
+      "loss": 0.6961,
+      "step": 5618
+    },
+    {
+      "epoch": 2.787886310040958,
+      "grad_norm": 0.1350215485186969,
+      "learning_rate": 1.0513297305244507e-06,
+      "loss": 0.718,
+      "step": 5619
+    },
+    {
+      "epoch": 2.7883827727441974,
+      "grad_norm": 0.1308450394330969,
+      "learning_rate": 1.0505333062680383e-06,
+      "loss": 0.7255,
+      "step": 5620
+    },
+    {
+      "epoch": 2.788879235447437,
+      "grad_norm": 0.1280745953181878,
+      "learning_rate": 1.04973710353937e-06,
+      "loss": 0.7049,
+      "step": 5621
+    },
+    {
+      "epoch": 2.7893756981506765,
+      "grad_norm": 0.1311328777593363,
+      "learning_rate": 1.048941122460132e-06,
+      "loss": 0.7186,
+      "step": 5622
+    },
+    {
+      "epoch": 2.789872160853916,
+      "grad_norm": 0.12796332817853948,
+      "learning_rate": 1.0481453631519775e-06,
+      "loss": 0.6986,
+      "step": 5623
+    },
+    {
+      "epoch": 2.790368623557155,
+      "grad_norm": 0.13077728648409898,
+      "learning_rate": 1.0473498257365247e-06,
+      "loss": 0.7143,
+      "step": 5624
+    },
+    {
+      "epoch": 2.7908650862603945,
+      "grad_norm": 0.12948794453473453,
+      "learning_rate": 1.0465545103353605e-06,
+      "loss": 0.7077,
+      "step": 5625
+    },
+    {
+      "epoch": 2.791361548963634,
+      "grad_norm": 0.12937599313169448,
+      "learning_rate": 1.0457594170700342e-06,
+      "loss": 0.7617,
+      "step": 5626
+    },
+    {
+      "epoch": 2.7918580116668736,
+      "grad_norm": 0.12552421196790298,
+      "learning_rate": 1.044964546062065e-06,
+      "loss": 0.6971,
+      "step": 5627
+    },
+    {
+      "epoch": 2.792354474370113,
+      "grad_norm": 0.128355301555478,
+      "learning_rate": 1.0441698974329351e-06,
+      "loss": 0.6843,
+      "step": 5628
+    },
+    {
+      "epoch": 2.7928509370733523,
+      "grad_norm": 0.13577852067425666,
+      "learning_rate": 1.043375471304093e-06,
+      "loss": 0.6869,
+      "step": 5629
+    },
+    {
+      "epoch": 2.7933473997765916,
+      "grad_norm": 0.12864716094515163,
+      "learning_rate": 1.0425812677969558e-06,
+      "loss": 0.7449,
+      "step": 5630
+    },
+    {
+      "epoch": 2.7938438624798314,
+      "grad_norm": 0.1293266492985022,
+      "learning_rate": 1.0417872870329029e-06,
+      "loss": 0.6797,
+      "step": 5631
+    },
+    {
+      "epoch": 2.7943403251830707,
+      "grad_norm": 0.1225513439668755,
+      "learning_rate": 1.0409935291332838e-06,
+      "loss": 0.6412,
+      "step": 5632
+    },
+    {
+      "epoch": 2.79483678788631,
+      "grad_norm": 0.12592530737282465,
+      "learning_rate": 1.04019999421941e-06,
+      "loss": 0.6889,
+      "step": 5633
+    },
+    {
+      "epoch": 2.7953332505895494,
+      "grad_norm": 0.12572434567997356,
+      "learning_rate": 1.0394066824125604e-06,
+      "loss": 0.6637,
+      "step": 5634
+    },
+    {
+      "epoch": 2.7958297132927887,
+      "grad_norm": 0.12830761221708809,
+      "learning_rate": 1.0386135938339812e-06,
+      "loss": 0.7142,
+      "step": 5635
+    },
+    {
+      "epoch": 2.796326175996028,
+      "grad_norm": 0.12804406659034892,
+      "learning_rate": 1.0378207286048817e-06,
+      "loss": 0.6682,
+      "step": 5636
+    },
+    {
+      "epoch": 2.796822638699268,
+      "grad_norm": 0.1306245576613395,
+      "learning_rate": 1.0370280868464405e-06,
+      "loss": 0.6909,
+      "step": 5637
+    },
+    {
+      "epoch": 2.797319101402507,
+      "grad_norm": 0.13224450154587467,
+      "learning_rate": 1.0362356686797996e-06,
+      "loss": 0.7285,
+      "step": 5638
+    },
+    {
+      "epoch": 2.7978155641057465,
+      "grad_norm": 0.1265750971979006,
+      "learning_rate": 1.0354434742260665e-06,
+      "loss": 0.6563,
+      "step": 5639
+    },
+    {
+      "epoch": 2.798312026808986,
+      "grad_norm": 0.1255205324161622,
+      "learning_rate": 1.0346515036063165e-06,
+      "loss": 0.6817,
+      "step": 5640
+    },
+    {
+      "epoch": 2.7988084895122256,
+      "grad_norm": 0.12545489772850293,
+      "learning_rate": 1.033859756941588e-06,
+      "loss": 0.6674,
+      "step": 5641
+    },
+    {
+      "epoch": 2.799304952215465,
+      "grad_norm": 0.12709075941299228,
+      "learning_rate": 1.0330682343528886e-06,
+      "loss": 0.6704,
+      "step": 5642
+    },
+    {
+      "epoch": 2.7998014149187043,
+      "grad_norm": 0.1267872987559885,
+      "learning_rate": 1.0322769359611883e-06,
+      "loss": 0.6957,
+      "step": 5643
+    },
+    {
+      "epoch": 2.8002978776219436,
+      "grad_norm": 0.13612904687065247,
+      "learning_rate": 1.0314858618874263e-06,
+      "loss": 0.7679,
+      "step": 5644
+    },
+    {
+      "epoch": 2.800794340325183,
+      "grad_norm": 0.12963249148574502,
+      "learning_rate": 1.0306950122525042e-06,
+      "loss": 0.66,
+      "step": 5645
+    },
+    {
+      "epoch": 2.8012908030284223,
+      "grad_norm": 0.12775287910822225,
+      "learning_rate": 1.0299043871772904e-06,
+      "loss": 0.6721,
+      "step": 5646
+    },
+    {
+      "epoch": 2.801787265731662,
+      "grad_norm": 0.13139858927008313,
+      "learning_rate": 1.0291139867826205e-06,
+      "loss": 0.6754,
+      "step": 5647
+    },
+    {
+      "epoch": 2.8022837284349014,
+      "grad_norm": 0.13065899559643898,
+      "learning_rate": 1.0283238111892929e-06,
+      "loss": 0.6853,
+      "step": 5648
+    },
+    {
+      "epoch": 2.8027801911381407,
+      "grad_norm": 0.13055758162862297,
+      "learning_rate": 1.0275338605180751e-06,
+      "loss": 0.7194,
+      "step": 5649
+    },
+    {
+      "epoch": 2.80327665384138,
+      "grad_norm": 0.13385534399710408,
+      "learning_rate": 1.0267441348896978e-06,
+      "loss": 0.703,
+      "step": 5650
+    },
+    {
+      "epoch": 2.80377311654462,
+      "grad_norm": 0.12413925075970825,
+      "learning_rate": 1.0259546344248567e-06,
+      "loss": 0.6408,
+      "step": 5651
+    },
+    {
+      "epoch": 2.804269579247859,
+      "grad_norm": 0.13054886800472554,
+      "learning_rate": 1.0251653592442157e-06,
+      "loss": 0.7052,
+      "step": 5652
+    },
+    {
+      "epoch": 2.8047660419510985,
+      "grad_norm": 0.13047999060457022,
+      "learning_rate": 1.0243763094684029e-06,
+      "loss": 0.6872,
+      "step": 5653
+    },
+    {
+      "epoch": 2.805262504654338,
+      "grad_norm": 0.1273801577355027,
+      "learning_rate": 1.0235874852180109e-06,
+      "loss": 0.6741,
+      "step": 5654
+    },
+    {
+      "epoch": 2.805758967357577,
+      "grad_norm": 0.1324495467661752,
+      "learning_rate": 1.0227988866135995e-06,
+      "loss": 0.7374,
+      "step": 5655
+    },
+    {
+      "epoch": 2.8062554300608165,
+      "grad_norm": 0.12792054360102523,
+      "learning_rate": 1.0220105137756923e-06,
+      "loss": 0.7069,
+      "step": 5656
+    },
+    {
+      "epoch": 2.8067518927640562,
+      "grad_norm": 0.1351583667186328,
+      "learning_rate": 1.021222366824781e-06,
+      "loss": 0.7174,
+      "step": 5657
+    },
+    {
+      "epoch": 2.8072483554672956,
+      "grad_norm": 0.13071189164436514,
+      "learning_rate": 1.0204344458813201e-06,
+      "loss": 0.7505,
+      "step": 5658
+    },
+    {
+      "epoch": 2.807744818170535,
+      "grad_norm": 0.12840479043760866,
+      "learning_rate": 1.0196467510657315e-06,
+      "loss": 0.7161,
+      "step": 5659
+    },
+    {
+      "epoch": 2.8082412808737742,
+      "grad_norm": 0.12700151293719608,
+      "learning_rate": 1.0188592824984018e-06,
+      "loss": 0.713,
+      "step": 5660
+    },
+    {
+      "epoch": 2.808737743577014,
+      "grad_norm": 0.1246492088887146,
+      "learning_rate": 1.0180720402996814e-06,
+      "loss": 0.698,
+      "step": 5661
+    },
+    {
+      "epoch": 2.8092342062802533,
+      "grad_norm": 0.13557496235898936,
+      "learning_rate": 1.0172850245898893e-06,
+      "loss": 0.7027,
+      "step": 5662
+    },
+    {
+      "epoch": 2.8097306689834927,
+      "grad_norm": 0.13048869432963878,
+      "learning_rate": 1.0164982354893072e-06,
+      "loss": 0.7034,
+      "step": 5663
+    },
+    {
+      "epoch": 2.810227131686732,
+      "grad_norm": 0.1309842813142355,
+      "learning_rate": 1.0157116731181844e-06,
+      "loss": 0.7185,
+      "step": 5664
+    },
+    {
+      "epoch": 2.8107235943899713,
+      "grad_norm": 0.12466311030204848,
+      "learning_rate": 1.0149253375967336e-06,
+      "loss": 0.6622,
+      "step": 5665
+    },
+    {
+      "epoch": 2.8112200570932107,
+      "grad_norm": 0.12962837245694803,
+      "learning_rate": 1.0141392290451327e-06,
+      "loss": 0.6915,
+      "step": 5666
+    },
+    {
+      "epoch": 2.8117165197964504,
+      "grad_norm": 0.12989363238833468,
+      "learning_rate": 1.0133533475835275e-06,
+      "loss": 0.7217,
+      "step": 5667
+    },
+    {
+      "epoch": 2.8122129824996898,
+      "grad_norm": 0.12706569640249818,
+      "learning_rate": 1.0125676933320258e-06,
+      "loss": 0.6702,
+      "step": 5668
+    },
+    {
+      "epoch": 2.812709445202929,
+      "grad_norm": 0.1304053742764276,
+      "learning_rate": 1.0117822664107039e-06,
+      "loss": 0.6888,
+      "step": 5669
+    },
+    {
+      "epoch": 2.8132059079061684,
+      "grad_norm": 0.12731183078943037,
+      "learning_rate": 1.0109970669396008e-06,
+      "loss": 0.7023,
+      "step": 5670
+    },
+    {
+      "epoch": 2.813702370609408,
+      "grad_norm": 0.12732905106726986,
+      "learning_rate": 1.0102120950387217e-06,
+      "loss": 0.7226,
+      "step": 5671
+    },
+    {
+      "epoch": 2.8141988333126475,
+      "grad_norm": 0.1306176241196579,
+      "learning_rate": 1.0094273508280369e-06,
+      "loss": 0.7277,
+      "step": 5672
+    },
+    {
+      "epoch": 2.814695296015887,
+      "grad_norm": 0.12671882523359002,
+      "learning_rate": 1.008642834427481e-06,
+      "loss": 0.7285,
+      "step": 5673
+    },
+    {
+      "epoch": 2.815191758719126,
+      "grad_norm": 0.13283779420386785,
+      "learning_rate": 1.0078585459569568e-06,
+      "loss": 0.7146,
+      "step": 5674
+    },
+    {
+      "epoch": 2.8156882214223655,
+      "grad_norm": 0.12814418090121601,
+      "learning_rate": 1.0070744855363283e-06,
+      "loss": 0.7038,
+      "step": 5675
+    },
+    {
+      "epoch": 2.816184684125605,
+      "grad_norm": 0.12617325806218707,
+      "learning_rate": 1.0062906532854284e-06,
+      "loss": 0.6782,
+      "step": 5676
+    },
+    {
+      "epoch": 2.8166811468288446,
+      "grad_norm": 0.1304282198590571,
+      "learning_rate": 1.0055070493240521e-06,
+      "loss": 0.7038,
+      "step": 5677
+    },
+    {
+      "epoch": 2.817177609532084,
+      "grad_norm": 0.12890829179767882,
+      "learning_rate": 1.00472367377196e-06,
+      "loss": 0.6802,
+      "step": 5678
+    },
+    {
+      "epoch": 2.8176740722353233,
+      "grad_norm": 0.13108770095262734,
+      "learning_rate": 1.0039405267488805e-06,
+      "loss": 0.7088,
+      "step": 5679
+    },
+    {
+      "epoch": 2.8181705349385626,
+      "grad_norm": 0.12760496474052235,
+      "learning_rate": 1.003157608374503e-06,
+      "loss": 0.6967,
+      "step": 5680
+    },
+    {
+      "epoch": 2.8186669976418024,
+      "grad_norm": 0.12724173087136592,
+      "learning_rate": 1.002374918768486e-06,
+      "loss": 0.6888,
+      "step": 5681
+    },
+    {
+      "epoch": 2.8191634603450417,
+      "grad_norm": 0.1281546403873397,
+      "learning_rate": 1.0015924580504502e-06,
+      "loss": 0.6831,
+      "step": 5682
+    },
+    {
+      "epoch": 2.819659923048281,
+      "grad_norm": 0.1306599446069587,
+      "learning_rate": 1.000810226339981e-06,
+      "loss": 0.6949,
+      "step": 5683
+    },
+    {
+      "epoch": 2.8201563857515204,
+      "grad_norm": 0.13275998082707438,
+      "learning_rate": 1.000028223756632e-06,
+      "loss": 0.7333,
+      "step": 5684
+    },
+    {
+      "epoch": 2.8206528484547597,
+      "grad_norm": 0.1259944276365557,
+      "learning_rate": 9.99246450419918e-07,
+      "loss": 0.7038,
+      "step": 5685
+    },
+    {
+      "epoch": 2.821149311157999,
+      "grad_norm": 0.1258776313988055,
+      "learning_rate": 9.984649064493228e-07,
+      "loss": 0.6935,
+      "step": 5686
+    },
+    {
+      "epoch": 2.821645773861239,
+      "grad_norm": 0.1277780519819066,
+      "learning_rate": 9.976835919642897e-07,
+      "loss": 0.6888,
+      "step": 5687
+    },
+    {
+      "epoch": 2.822142236564478,
+      "grad_norm": 0.1255875204317941,
+      "learning_rate": 9.969025070842326e-07,
+      "loss": 0.7692,
+      "step": 5688
+    },
+    {
+      "epoch": 2.8226386992677175,
+      "grad_norm": 0.12833215364936337,
+      "learning_rate": 9.961216519285272e-07,
+      "loss": 0.6744,
+      "step": 5689
+    },
+    {
+      "epoch": 2.823135161970957,
+      "grad_norm": 0.1316740562970518,
+      "learning_rate": 9.953410266165131e-07,
+      "loss": 0.7523,
+      "step": 5690
+    },
+    {
+      "epoch": 2.8236316246741966,
+      "grad_norm": 0.12733567410451435,
+      "learning_rate": 9.945606312674991e-07,
+      "loss": 0.6814,
+      "step": 5691
+    },
+    {
+      "epoch": 2.824128087377436,
+      "grad_norm": 0.13152963950218258,
+      "learning_rate": 9.937804660007545e-07,
+      "loss": 0.7384,
+      "step": 5692
+    },
+    {
+      "epoch": 2.8246245500806753,
+      "grad_norm": 0.12905971140335962,
+      "learning_rate": 9.930005309355143e-07,
+      "loss": 0.6819,
+      "step": 5693
+    },
+    {
+      "epoch": 2.8251210127839146,
+      "grad_norm": 0.1293289412286106,
+      "learning_rate": 9.922208261909811e-07,
+      "loss": 0.6689,
+      "step": 5694
+    },
+    {
+      "epoch": 2.825617475487154,
+      "grad_norm": 0.12656866487417895,
+      "learning_rate": 9.91441351886318e-07,
+      "loss": 0.671,
+      "step": 5695
+    },
+    {
+      "epoch": 2.8261139381903932,
+      "grad_norm": 0.12682407823609218,
+      "learning_rate": 9.906621081406575e-07,
+      "loss": 0.7023,
+      "step": 5696
+    },
+    {
+      "epoch": 2.826610400893633,
+      "grad_norm": 0.12481510953803818,
+      "learning_rate": 9.898830950730935e-07,
+      "loss": 0.7291,
+      "step": 5697
+    },
+    {
+      "epoch": 2.8271068635968724,
+      "grad_norm": 0.13429126569598215,
+      "learning_rate": 9.891043128026846e-07,
+      "loss": 0.75,
+      "step": 5698
+    },
+    {
+      "epoch": 2.8276033263001117,
+      "grad_norm": 0.12713270388437029,
+      "learning_rate": 9.883257614484568e-07,
+      "loss": 0.6696,
+      "step": 5699
+    },
+    {
+      "epoch": 2.828099789003351,
+      "grad_norm": 0.13204195693146775,
+      "learning_rate": 9.875474411293977e-07,
+      "loss": 0.7444,
+      "step": 5700
+    },
+    {
+      "epoch": 2.828596251706591,
+      "grad_norm": 0.12538390670171082,
+      "learning_rate": 9.86769351964463e-07,
+      "loss": 0.6943,
+      "step": 5701
+    },
+    {
+      "epoch": 2.82909271440983,
+      "grad_norm": 0.12942426088477443,
+      "learning_rate": 9.859914940725698e-07,
+      "loss": 0.6697,
+      "step": 5702
+    },
+    {
+      "epoch": 2.8295891771130695,
+      "grad_norm": 0.13073956206544834,
+      "learning_rate": 9.852138675726014e-07,
+      "loss": 0.7027,
+      "step": 5703
+    },
+    {
+      "epoch": 2.830085639816309,
+      "grad_norm": 0.13192378873945426,
+      "learning_rate": 9.844364725834058e-07,
+      "loss": 0.6702,
+      "step": 5704
+    },
+    {
+      "epoch": 2.830582102519548,
+      "grad_norm": 0.13045843127382814,
+      "learning_rate": 9.836593092237942e-07,
+      "loss": 0.7186,
+      "step": 5705
+    },
+    {
+      "epoch": 2.8310785652227874,
+      "grad_norm": 0.13209103372271133,
+      "learning_rate": 9.828823776125455e-07,
+      "loss": 0.6726,
+      "step": 5706
+    },
+    {
+      "epoch": 2.831575027926027,
+      "grad_norm": 0.12836854761747843,
+      "learning_rate": 9.82105677868399e-07,
+      "loss": 0.6604,
+      "step": 5707
+    },
+    {
+      "epoch": 2.8320714906292666,
+      "grad_norm": 0.12440008092199902,
+      "learning_rate": 9.813292101100634e-07,
+      "loss": 0.6939,
+      "step": 5708
+    },
+    {
+      "epoch": 2.832567953332506,
+      "grad_norm": 0.13509346075273426,
+      "learning_rate": 9.805529744562076e-07,
+      "loss": 0.7395,
+      "step": 5709
+    },
+    {
+      "epoch": 2.833064416035745,
+      "grad_norm": 0.12453204951253816,
+      "learning_rate": 9.797769710254664e-07,
+      "loss": 0.6882,
+      "step": 5710
+    },
+    {
+      "epoch": 2.833560878738985,
+      "grad_norm": 0.12697842177012855,
+      "learning_rate": 9.790011999364412e-07,
+      "loss": 0.7063,
+      "step": 5711
+    },
+    {
+      "epoch": 2.8340573414422243,
+      "grad_norm": 0.12754534007734417,
+      "learning_rate": 9.782256613076945e-07,
+      "loss": 0.7482,
+      "step": 5712
+    },
+    {
+      "epoch": 2.8345538041454637,
+      "grad_norm": 0.13069481659575172,
+      "learning_rate": 9.774503552577563e-07,
+      "loss": 0.6936,
+      "step": 5713
+    },
+    {
+      "epoch": 2.835050266848703,
+      "grad_norm": 0.13974772554638237,
+      "learning_rate": 9.766752819051192e-07,
+      "loss": 0.717,
+      "step": 5714
+    },
+    {
+      "epoch": 2.8355467295519423,
+      "grad_norm": 0.1251939515003143,
+      "learning_rate": 9.759004413682396e-07,
+      "loss": 0.6509,
+      "step": 5715
+    },
+    {
+      "epoch": 2.8360431922551816,
+      "grad_norm": 0.1312396524534019,
+      "learning_rate": 9.751258337655418e-07,
+      "loss": 0.7324,
+      "step": 5716
+    },
+    {
+      "epoch": 2.836539654958421,
+      "grad_norm": 0.13060634236111196,
+      "learning_rate": 9.743514592154093e-07,
+      "loss": 0.732,
+      "step": 5717
+    },
+    {
+      "epoch": 2.8370361176616608,
+      "grad_norm": 0.12983746553461947,
+      "learning_rate": 9.735773178361965e-07,
+      "loss": 0.7217,
+      "step": 5718
+    },
+    {
+      "epoch": 2.8375325803649,
+      "grad_norm": 0.1313030207597098,
+      "learning_rate": 9.728034097462144e-07,
+      "loss": 0.7298,
+      "step": 5719
+    },
+    {
+      "epoch": 2.8380290430681394,
+      "grad_norm": 0.1305730115816129,
+      "learning_rate": 9.720297350637453e-07,
+      "loss": 0.73,
+      "step": 5720
+    },
+    {
+      "epoch": 2.838525505771379,
+      "grad_norm": 0.12959419549059026,
+      "learning_rate": 9.712562939070322e-07,
+      "loss": 0.7013,
+      "step": 5721
+    },
+    {
+      "epoch": 2.8390219684746185,
+      "grad_norm": 0.13353606316263741,
+      "learning_rate": 9.704830863942819e-07,
+      "loss": 0.6893,
+      "step": 5722
+    },
+    {
+      "epoch": 2.839518431177858,
+      "grad_norm": 0.12927989012251478,
+      "learning_rate": 9.697101126436689e-07,
+      "loss": 0.6941,
+      "step": 5723
+    },
+    {
+      "epoch": 2.840014893881097,
+      "grad_norm": 0.12460555850658463,
+      "learning_rate": 9.689373727733282e-07,
+      "loss": 0.6749,
+      "step": 5724
+    },
+    {
+      "epoch": 2.8405113565843365,
+      "grad_norm": 0.13017346927566242,
+      "learning_rate": 9.681648669013618e-07,
+      "loss": 0.6884,
+      "step": 5725
+    },
+    {
+      "epoch": 2.841007819287576,
+      "grad_norm": 0.12587233851865584,
+      "learning_rate": 9.673925951458347e-07,
+      "loss": 0.6782,
+      "step": 5726
+    },
+    {
+      "epoch": 2.841504281990815,
+      "grad_norm": 0.12769491177786568,
+      "learning_rate": 9.66620557624775e-07,
+      "loss": 0.7175,
+      "step": 5727
+    },
+    {
+      "epoch": 2.842000744694055,
+      "grad_norm": 0.13179110270409086,
+      "learning_rate": 9.658487544561778e-07,
+      "loss": 0.7183,
+      "step": 5728
+    },
+    {
+      "epoch": 2.8424972073972943,
+      "grad_norm": 0.13106533480871427,
+      "learning_rate": 9.650771857580007e-07,
+      "loss": 0.7449,
+      "step": 5729
+    },
+    {
+      "epoch": 2.8429936701005336,
+      "grad_norm": 0.12932936151209334,
+      "learning_rate": 9.64305851648164e-07,
+      "loss": 0.7064,
+      "step": 5730
+    },
+    {
+      "epoch": 2.8434901328037734,
+      "grad_norm": 0.13312505296398447,
+      "learning_rate": 9.63534752244556e-07,
+      "loss": 0.763,
+      "step": 5731
+    },
+    {
+      "epoch": 2.8439865955070127,
+      "grad_norm": 0.12320453586612351,
+      "learning_rate": 9.627638876650245e-07,
+      "loss": 0.6423,
+      "step": 5732
+    },
+    {
+      "epoch": 2.844483058210252,
+      "grad_norm": 0.12794077575330426,
+      "learning_rate": 9.61993258027386e-07,
+      "loss": 0.6768,
+      "step": 5733
+    },
+    {
+      "epoch": 2.8449795209134914,
+      "grad_norm": 0.1266752168624091,
+      "learning_rate": 9.612228634494184e-07,
+      "loss": 0.6743,
+      "step": 5734
+    },
+    {
+      "epoch": 2.8454759836167307,
+      "grad_norm": 0.12653827532865938,
+      "learning_rate": 9.604527040488631e-07,
+      "loss": 0.6873,
+      "step": 5735
+    },
+    {
+      "epoch": 2.84597244631997,
+      "grad_norm": 0.13105455883016892,
+      "learning_rate": 9.596827799434278e-07,
+      "loss": 0.7279,
+      "step": 5736
+    },
+    {
+      "epoch": 2.8464689090232094,
+      "grad_norm": 0.12824418068445376,
+      "learning_rate": 9.589130912507812e-07,
+      "loss": 0.6876,
+      "step": 5737
+    },
+    {
+      "epoch": 2.846965371726449,
+      "grad_norm": 0.12298579417263769,
+      "learning_rate": 9.581436380885604e-07,
+      "loss": 0.6416,
+      "step": 5738
+    },
+    {
+      "epoch": 2.8474618344296885,
+      "grad_norm": 0.12704521892342857,
+      "learning_rate": 9.573744205743613e-07,
+      "loss": 0.7144,
+      "step": 5739
+    },
+    {
+      "epoch": 2.847958297132928,
+      "grad_norm": 0.13151764091738496,
+      "learning_rate": 9.566054388257492e-07,
+      "loss": 0.6985,
+      "step": 5740
+    },
+    {
+      "epoch": 2.8484547598361676,
+      "grad_norm": 0.12281688304650965,
+      "learning_rate": 9.558366929602492e-07,
+      "loss": 0.6525,
+      "step": 5741
+    },
+    {
+      "epoch": 2.848951222539407,
+      "grad_norm": 0.12985743528766544,
+      "learning_rate": 9.55068183095351e-07,
+      "loss": 0.7118,
+      "step": 5742
+    },
+    {
+      "epoch": 2.8494476852426462,
+      "grad_norm": 0.12697246716589147,
+      "learning_rate": 9.542999093485108e-07,
+      "loss": 0.686,
+      "step": 5743
+    },
+    {
+      "epoch": 2.8499441479458856,
+      "grad_norm": 0.12891175104431518,
+      "learning_rate": 9.535318718371453e-07,
+      "loss": 0.7149,
+      "step": 5744
+    },
+    {
+      "epoch": 2.850440610649125,
+      "grad_norm": 0.12952157865563893,
+      "learning_rate": 9.527640706786381e-07,
+      "loss": 0.6881,
+      "step": 5745
+    },
+    {
+      "epoch": 2.8509370733523642,
+      "grad_norm": 0.13148893826296196,
+      "learning_rate": 9.519965059903349e-07,
+      "loss": 0.6798,
+      "step": 5746
+    },
+    {
+      "epoch": 2.8514335360556036,
+      "grad_norm": 0.13521982035020255,
+      "learning_rate": 9.512291778895444e-07,
+      "loss": 0.7152,
+      "step": 5747
+    },
+    {
+      "epoch": 2.8519299987588433,
+      "grad_norm": 0.13512198272074397,
+      "learning_rate": 9.504620864935421e-07,
+      "loss": 0.7148,
+      "step": 5748
+    },
+    {
+      "epoch": 2.8524264614620827,
+      "grad_norm": 0.1266177814142838,
+      "learning_rate": 9.496952319195643e-07,
+      "loss": 0.6658,
+      "step": 5749
+    },
+    {
+      "epoch": 2.852922924165322,
+      "grad_norm": 0.12979818368870435,
+      "learning_rate": 9.489286142848148e-07,
+      "loss": 0.7075,
+      "step": 5750
+    },
+    {
+      "epoch": 2.8534193868685613,
+      "grad_norm": 0.12554115281420927,
+      "learning_rate": 9.481622337064552e-07,
+      "loss": 0.6843,
+      "step": 5751
+    },
+    {
+      "epoch": 2.853915849571801,
+      "grad_norm": 0.1320343536961316,
+      "learning_rate": 9.473960903016175e-07,
+      "loss": 0.6923,
+      "step": 5752
+    },
+    {
+      "epoch": 2.8544123122750404,
+      "grad_norm": 0.12386913184376706,
+      "learning_rate": 9.466301841873929e-07,
+      "loss": 0.6594,
+      "step": 5753
+    },
+    {
+      "epoch": 2.8549087749782798,
+      "grad_norm": 0.13316531050788066,
+      "learning_rate": 9.458645154808377e-07,
+      "loss": 0.7124,
+      "step": 5754
+    },
+    {
+      "epoch": 2.855405237681519,
+      "grad_norm": 0.1300731124347825,
+      "learning_rate": 9.450990842989732e-07,
+      "loss": 0.709,
+      "step": 5755
+    },
+    {
+      "epoch": 2.8559017003847584,
+      "grad_norm": 0.12945665805477527,
+      "learning_rate": 9.443338907587821e-07,
+      "loss": 0.7232,
+      "step": 5756
+    },
+    {
+      "epoch": 2.8563981630879978,
+      "grad_norm": 0.13305413501745098,
+      "learning_rate": 9.435689349772135e-07,
+      "loss": 0.7081,
+      "step": 5757
+    },
+    {
+      "epoch": 2.8568946257912375,
+      "grad_norm": 0.1283285050578409,
+      "learning_rate": 9.428042170711776e-07,
+      "loss": 0.7441,
+      "step": 5758
+    },
+    {
+      "epoch": 2.857391088494477,
+      "grad_norm": 0.13256787722435986,
+      "learning_rate": 9.420397371575485e-07,
+      "loss": 0.7147,
+      "step": 5759
+    },
+    {
+      "epoch": 2.857887551197716,
+      "grad_norm": 0.13501707551704034,
+      "learning_rate": 9.412754953531664e-07,
+      "loss": 0.7473,
+      "step": 5760
+    },
+    {
+      "epoch": 2.8583840139009555,
+      "grad_norm": 0.12779455860264854,
+      "learning_rate": 9.405114917748318e-07,
+      "loss": 0.6941,
+      "step": 5761
+    },
+    {
+      "epoch": 2.8588804766041953,
+      "grad_norm": 0.13350164440946205,
+      "learning_rate": 9.397477265393121e-07,
+      "loss": 0.7079,
+      "step": 5762
+    },
+    {
+      "epoch": 2.8593769393074346,
+      "grad_norm": 0.12695906616841662,
+      "learning_rate": 9.389841997633356e-07,
+      "loss": 0.7138,
+      "step": 5763
+    },
+    {
+      "epoch": 2.859873402010674,
+      "grad_norm": 0.1339764757838581,
+      "learning_rate": 9.382209115635942e-07,
+      "loss": 0.686,
+      "step": 5764
+    },
+    {
+      "epoch": 2.8603698647139133,
+      "grad_norm": 0.13278274095995127,
+      "learning_rate": 9.374578620567462e-07,
+      "loss": 0.7039,
+      "step": 5765
+    },
+    {
+      "epoch": 2.8608663274171526,
+      "grad_norm": 0.1277777239083202,
+      "learning_rate": 9.366950513594106e-07,
+      "loss": 0.7353,
+      "step": 5766
+    },
+    {
+      "epoch": 2.861362790120392,
+      "grad_norm": 0.12832944909129318,
+      "learning_rate": 9.359324795881708e-07,
+      "loss": 0.726,
+      "step": 5767
+    },
+    {
+      "epoch": 2.8618592528236317,
+      "grad_norm": 0.12594067772100392,
+      "learning_rate": 9.351701468595734e-07,
+      "loss": 0.7546,
+      "step": 5768
+    },
+    {
+      "epoch": 2.862355715526871,
+      "grad_norm": 0.128006328679492,
+      "learning_rate": 9.344080532901282e-07,
+      "loss": 0.708,
+      "step": 5769
+    },
+    {
+      "epoch": 2.8628521782301104,
+      "grad_norm": 0.1275045024635113,
+      "learning_rate": 9.336461989963102e-07,
+      "loss": 0.6734,
+      "step": 5770
+    },
+    {
+      "epoch": 2.8633486409333497,
+      "grad_norm": 0.1414589237121449,
+      "learning_rate": 9.328845840945555e-07,
+      "loss": 0.7276,
+      "step": 5771
+    },
+    {
+      "epoch": 2.8638451036365895,
+      "grad_norm": 0.12748603017529134,
+      "learning_rate": 9.321232087012664e-07,
+      "loss": 0.6894,
+      "step": 5772
+    },
+    {
+      "epoch": 2.864341566339829,
+      "grad_norm": 0.12968727130669763,
+      "learning_rate": 9.313620729328055e-07,
+      "loss": 0.6977,
+      "step": 5773
+    },
+    {
+      "epoch": 2.864838029043068,
+      "grad_norm": 0.12740734697560116,
+      "learning_rate": 9.306011769054999e-07,
+      "loss": 0.702,
+      "step": 5774
+    },
+    {
+      "epoch": 2.8653344917463075,
+      "grad_norm": 0.13202429187909928,
+      "learning_rate": 9.298405207356418e-07,
+      "loss": 0.7235,
+      "step": 5775
+    },
+    {
+      "epoch": 2.865830954449547,
+      "grad_norm": 0.12516118343049698,
+      "learning_rate": 9.290801045394837e-07,
+      "loss": 0.6962,
+      "step": 5776
+    },
+    {
+      "epoch": 2.866327417152786,
+      "grad_norm": 0.12466908791148393,
+      "learning_rate": 9.283199284332448e-07,
+      "loss": 0.6828,
+      "step": 5777
+    },
+    {
+      "epoch": 2.866823879856026,
+      "grad_norm": 0.13168847619221924,
+      "learning_rate": 9.275599925331047e-07,
+      "loss": 0.7276,
+      "step": 5778
+    },
+    {
+      "epoch": 2.8673203425592653,
+      "grad_norm": 0.12380760924529202,
+      "learning_rate": 9.268002969552068e-07,
+      "loss": 0.7109,
+      "step": 5779
+    },
+    {
+      "epoch": 2.8678168052625046,
+      "grad_norm": 0.1273101006035769,
+      "learning_rate": 9.260408418156597e-07,
+      "loss": 0.7053,
+      "step": 5780
+    },
+    {
+      "epoch": 2.868313267965744,
+      "grad_norm": 0.12716786520734547,
+      "learning_rate": 9.252816272305328e-07,
+      "loss": 0.6563,
+      "step": 5781
+    },
+    {
+      "epoch": 2.8688097306689837,
+      "grad_norm": 0.12534292127359717,
+      "learning_rate": 9.245226533158624e-07,
+      "loss": 0.6822,
+      "step": 5782
+    },
+    {
+      "epoch": 2.869306193372223,
+      "grad_norm": 0.12864761504663866,
+      "learning_rate": 9.237639201876417e-07,
+      "loss": 0.7303,
+      "step": 5783
+    },
+    {
+      "epoch": 2.8698026560754624,
+      "grad_norm": 0.12941975850294216,
+      "learning_rate": 9.230054279618336e-07,
+      "loss": 0.6791,
+      "step": 5784
+    },
+    {
+      "epoch": 2.8702991187787017,
+      "grad_norm": 0.12715616195640328,
+      "learning_rate": 9.222471767543608e-07,
+      "loss": 0.677,
+      "step": 5785
+    },
+    {
+      "epoch": 2.870795581481941,
+      "grad_norm": 0.1354882762503265,
+      "learning_rate": 9.214891666811088e-07,
+      "loss": 0.7547,
+      "step": 5786
+    },
+    {
+      "epoch": 2.8712920441851804,
+      "grad_norm": 0.12912527608146743,
+      "learning_rate": 9.207313978579289e-07,
+      "loss": 0.6888,
+      "step": 5787
+    },
+    {
+      "epoch": 2.87178850688842,
+      "grad_norm": 0.1349499512563821,
+      "learning_rate": 9.199738704006322e-07,
+      "loss": 0.7308,
+      "step": 5788
+    },
+    {
+      "epoch": 2.8722849695916595,
+      "grad_norm": 0.1306875820845483,
+      "learning_rate": 9.192165844249967e-07,
+      "loss": 0.6544,
+      "step": 5789
+    },
+    {
+      "epoch": 2.872781432294899,
+      "grad_norm": 0.1284137449808761,
+      "learning_rate": 9.184595400467605e-07,
+      "loss": 0.6812,
+      "step": 5790
+    },
+    {
+      "epoch": 2.873277894998138,
+      "grad_norm": 0.12848694849478118,
+      "learning_rate": 9.177027373816244e-07,
+      "loss": 0.7192,
+      "step": 5791
+    },
+    {
+      "epoch": 2.873774357701378,
+      "grad_norm": 0.12937550210725762,
+      "learning_rate": 9.169461765452556e-07,
+      "loss": 0.6748,
+      "step": 5792
+    },
+    {
+      "epoch": 2.8742708204046172,
+      "grad_norm": 0.12782097735362807,
+      "learning_rate": 9.161898576532805e-07,
+      "loss": 0.6797,
+      "step": 5793
+    },
+    {
+      "epoch": 2.8747672831078566,
+      "grad_norm": 0.1305376367487405,
+      "learning_rate": 9.154337808212921e-07,
+      "loss": 0.6827,
+      "step": 5794
+    },
+    {
+      "epoch": 2.875263745811096,
+      "grad_norm": 0.12880050176675548,
+      "learning_rate": 9.146779461648437e-07,
+      "loss": 0.6894,
+      "step": 5795
+    },
+    {
+      "epoch": 2.8757602085143352,
+      "grad_norm": 0.1259680592660518,
+      "learning_rate": 9.139223537994519e-07,
+      "loss": 0.6695,
+      "step": 5796
+    },
+    {
+      "epoch": 2.8762566712175746,
+      "grad_norm": 0.12642296565251043,
+      "learning_rate": 9.131670038405979e-07,
+      "loss": 0.674,
+      "step": 5797
+    },
+    {
+      "epoch": 2.8767531339208143,
+      "grad_norm": 0.12515161938813915,
+      "learning_rate": 9.124118964037246e-07,
+      "loss": 0.6419,
+      "step": 5798
+    },
+    {
+      "epoch": 2.8772495966240537,
+      "grad_norm": 0.13621267258149164,
+      "learning_rate": 9.11657031604238e-07,
+      "loss": 0.764,
+      "step": 5799
+    },
+    {
+      "epoch": 2.877746059327293,
+      "grad_norm": 0.1267504655200429,
+      "learning_rate": 9.109024095575062e-07,
+      "loss": 0.6827,
+      "step": 5800
+    },
+    {
+      "epoch": 2.8782425220305323,
+      "grad_norm": 0.12834380780687052,
+      "learning_rate": 9.101480303788623e-07,
+      "loss": 0.7056,
+      "step": 5801
+    },
+    {
+      "epoch": 2.878738984733772,
+      "grad_norm": 0.14054019031479334,
+      "learning_rate": 9.093938941836012e-07,
+      "loss": 0.7741,
+      "step": 5802
+    },
+    {
+      "epoch": 2.8792354474370114,
+      "grad_norm": 0.1292406739225414,
+      "learning_rate": 9.086400010869787e-07,
+      "loss": 0.6723,
+      "step": 5803
+    },
+    {
+      "epoch": 2.8797319101402508,
+      "grad_norm": 0.13155167047598984,
+      "learning_rate": 9.078863512042174e-07,
+      "loss": 0.6905,
+      "step": 5804
+    },
+    {
+      "epoch": 2.88022837284349,
+      "grad_norm": 0.128899843641201,
+      "learning_rate": 9.071329446504997e-07,
+      "loss": 0.6903,
+      "step": 5805
+    },
+    {
+      "epoch": 2.8807248355467294,
+      "grad_norm": 0.12938294267895303,
+      "learning_rate": 9.063797815409711e-07,
+      "loss": 0.7468,
+      "step": 5806
+    },
+    {
+      "epoch": 2.8812212982499688,
+      "grad_norm": 0.1302289199669989,
+      "learning_rate": 9.056268619907418e-07,
+      "loss": 0.704,
+      "step": 5807
+    },
+    {
+      "epoch": 2.8817177609532085,
+      "grad_norm": 0.12864445639975267,
+      "learning_rate": 9.048741861148822e-07,
+      "loss": 0.6994,
+      "step": 5808
+    },
+    {
+      "epoch": 2.882214223656448,
+      "grad_norm": 0.12590829953054358,
+      "learning_rate": 9.041217540284278e-07,
+      "loss": 0.6584,
+      "step": 5809
+    },
+    {
+      "epoch": 2.882710686359687,
+      "grad_norm": 0.13423088059524024,
+      "learning_rate": 9.033695658463757e-07,
+      "loss": 0.705,
+      "step": 5810
+    },
+    {
+      "epoch": 2.8832071490629265,
+      "grad_norm": 0.12841940675440747,
+      "learning_rate": 9.026176216836844e-07,
+      "loss": 0.6997,
+      "step": 5811
+    },
+    {
+      "epoch": 2.8837036117661663,
+      "grad_norm": 0.1301253908505875,
+      "learning_rate": 9.018659216552783e-07,
+      "loss": 0.7129,
+      "step": 5812
+    },
+    {
+      "epoch": 2.8842000744694056,
+      "grad_norm": 0.1280984017417436,
+      "learning_rate": 9.011144658760413e-07,
+      "loss": 0.7191,
+      "step": 5813
+    },
+    {
+      "epoch": 2.884696537172645,
+      "grad_norm": 0.12267985245047328,
+      "learning_rate": 9.003632544608235e-07,
+      "loss": 0.6914,
+      "step": 5814
+    },
+    {
+      "epoch": 2.8851929998758843,
+      "grad_norm": 0.1287583871760695,
+      "learning_rate": 8.996122875244325e-07,
+      "loss": 0.728,
+      "step": 5815
+    },
+    {
+      "epoch": 2.8856894625791236,
+      "grad_norm": 0.12382625301443378,
+      "learning_rate": 8.98861565181644e-07,
+      "loss": 0.6676,
+      "step": 5816
+    },
+    {
+      "epoch": 2.886185925282363,
+      "grad_norm": 0.13281271821429838,
+      "learning_rate": 8.981110875471929e-07,
+      "loss": 0.7116,
+      "step": 5817
+    },
+    {
+      "epoch": 2.8866823879856027,
+      "grad_norm": 0.13692553089465404,
+      "learning_rate": 8.973608547357768e-07,
+      "loss": 0.7455,
+      "step": 5818
+    },
+    {
+      "epoch": 2.887178850688842,
+      "grad_norm": 0.1289222222449465,
+      "learning_rate": 8.966108668620585e-07,
+      "loss": 0.7099,
+      "step": 5819
+    },
+    {
+      "epoch": 2.8876753133920814,
+      "grad_norm": 0.1312399620437457,
+      "learning_rate": 8.958611240406601e-07,
+      "loss": 0.7066,
+      "step": 5820
+    },
+    {
+      "epoch": 2.8881717760953207,
+      "grad_norm": 0.13075456926805842,
+      "learning_rate": 8.951116263861693e-07,
+      "loss": 0.6928,
+      "step": 5821
+    },
+    {
+      "epoch": 2.8886682387985605,
+      "grad_norm": 0.128303038865835,
+      "learning_rate": 8.943623740131338e-07,
+      "loss": 0.6977,
+      "step": 5822
+    },
+    {
+      "epoch": 2.8891647015018,
+      "grad_norm": 0.13374921271432447,
+      "learning_rate": 8.936133670360644e-07,
+      "loss": 0.6792,
+      "step": 5823
+    },
+    {
+      "epoch": 2.889661164205039,
+      "grad_norm": 0.13148186697294473,
+      "learning_rate": 8.928646055694362e-07,
+      "loss": 0.7183,
+      "step": 5824
+    },
+    {
+      "epoch": 2.8901576269082785,
+      "grad_norm": 0.1288745562512283,
+      "learning_rate": 8.921160897276837e-07,
+      "loss": 0.707,
+      "step": 5825
+    },
+    {
+      "epoch": 2.890654089611518,
+      "grad_norm": 0.13355621762460396,
+      "learning_rate": 8.913678196252071e-07,
+      "loss": 0.7678,
+      "step": 5826
+    },
+    {
+      "epoch": 2.891150552314757,
+      "grad_norm": 0.1241939860039103,
+      "learning_rate": 8.90619795376367e-07,
+      "loss": 0.6613,
+      "step": 5827
+    },
+    {
+      "epoch": 2.891647015017997,
+      "grad_norm": 0.13480780396182257,
+      "learning_rate": 8.898720170954858e-07,
+      "loss": 0.7461,
+      "step": 5828
+    },
+    {
+      "epoch": 2.8921434777212363,
+      "grad_norm": 0.13949524462587928,
+      "learning_rate": 8.891244848968514e-07,
+      "loss": 0.8129,
+      "step": 5829
+    },
+    {
+      "epoch": 2.8926399404244756,
+      "grad_norm": 0.13174343677934816,
+      "learning_rate": 8.883771988947099e-07,
+      "loss": 0.7369,
+      "step": 5830
+    },
+    {
+      "epoch": 2.893136403127715,
+      "grad_norm": 0.1329364719930937,
+      "learning_rate": 8.876301592032749e-07,
+      "loss": 0.7148,
+      "step": 5831
+    },
+    {
+      "epoch": 2.8936328658309547,
+      "grad_norm": 0.1272203938665379,
+      "learning_rate": 8.868833659367163e-07,
+      "loss": 0.6465,
+      "step": 5832
+    },
+    {
+      "epoch": 2.894129328534194,
+      "grad_norm": 0.1293924264978683,
+      "learning_rate": 8.861368192091713e-07,
+      "loss": 0.6836,
+      "step": 5833
+    },
+    {
+      "epoch": 2.8946257912374334,
+      "grad_norm": 0.13220400859603937,
+      "learning_rate": 8.853905191347373e-07,
+      "loss": 0.7098,
+      "step": 5834
+    },
+    {
+      "epoch": 2.8951222539406727,
+      "grad_norm": 0.13015024295581795,
+      "learning_rate": 8.846444658274733e-07,
+      "loss": 0.6838,
+      "step": 5835
+    },
+    {
+      "epoch": 2.895618716643912,
+      "grad_norm": 0.13216797055404755,
+      "learning_rate": 8.838986594014034e-07,
+      "loss": 0.7106,
+      "step": 5836
+    },
+    {
+      "epoch": 2.8961151793471513,
+      "grad_norm": 0.1278107756500494,
+      "learning_rate": 8.831530999705104e-07,
+      "loss": 0.7313,
+      "step": 5837
+    },
+    {
+      "epoch": 2.896611642050391,
+      "grad_norm": 0.12908014728449954,
+      "learning_rate": 8.824077876487424e-07,
+      "loss": 0.7177,
+      "step": 5838
+    },
+    {
+      "epoch": 2.8971081047536305,
+      "grad_norm": 0.15225280744669853,
+      "learning_rate": 8.816627225500082e-07,
+      "loss": 0.6728,
+      "step": 5839
+    },
+    {
+      "epoch": 2.89760456745687,
+      "grad_norm": 0.12843463953521542,
+      "learning_rate": 8.80917904788178e-07,
+      "loss": 0.6796,
+      "step": 5840
+    },
+    {
+      "epoch": 2.898101030160109,
+      "grad_norm": 0.125053238809787,
+      "learning_rate": 8.801733344770869e-07,
+      "loss": 0.6684,
+      "step": 5841
+    },
+    {
+      "epoch": 2.898597492863349,
+      "grad_norm": 0.12924011340437813,
+      "learning_rate": 8.794290117305296e-07,
+      "loss": 0.6866,
+      "step": 5842
+    },
+    {
+      "epoch": 2.899093955566588,
+      "grad_norm": 0.12939827655850278,
+      "learning_rate": 8.786849366622629e-07,
+      "loss": 0.6713,
+      "step": 5843
+    },
+    {
+      "epoch": 2.8995904182698276,
+      "grad_norm": 0.1369447430816981,
+      "learning_rate": 8.77941109386009e-07,
+      "loss": 0.7056,
+      "step": 5844
+    },
+    {
+      "epoch": 2.900086880973067,
+      "grad_norm": 0.13371285950295925,
+      "learning_rate": 8.771975300154478e-07,
+      "loss": 0.703,
+      "step": 5845
+    },
+    {
+      "epoch": 2.900583343676306,
+      "grad_norm": 0.13319591109450518,
+      "learning_rate": 8.764541986642253e-07,
+      "loss": 0.6762,
+      "step": 5846
+    },
+    {
+      "epoch": 2.9010798063795455,
+      "grad_norm": 0.12748929289799496,
+      "learning_rate": 8.757111154459472e-07,
+      "loss": 0.6883,
+      "step": 5847
+    },
+    {
+      "epoch": 2.9015762690827853,
+      "grad_norm": 0.13341860100028502,
+      "learning_rate": 8.749682804741813e-07,
+      "loss": 0.7439,
+      "step": 5848
+    },
+    {
+      "epoch": 2.9020727317860247,
+      "grad_norm": 0.13140731624898971,
+      "learning_rate": 8.742256938624585e-07,
+      "loss": 0.6919,
+      "step": 5849
+    },
+    {
+      "epoch": 2.902569194489264,
+      "grad_norm": 0.126423779251477,
+      "learning_rate": 8.7348335572427e-07,
+      "loss": 0.6907,
+      "step": 5850
+    },
+    {
+      "epoch": 2.9030656571925033,
+      "grad_norm": 0.12587303898494598,
+      "learning_rate": 8.727412661730724e-07,
+      "loss": 0.6969,
+      "step": 5851
+    },
+    {
+      "epoch": 2.903562119895743,
+      "grad_norm": 0.12579584782238049,
+      "learning_rate": 8.719994253222805e-07,
+      "loss": 0.7108,
+      "step": 5852
+    },
+    {
+      "epoch": 2.9040585825989824,
+      "grad_norm": 0.13036032819321902,
+      "learning_rate": 8.712578332852739e-07,
+      "loss": 0.6665,
+      "step": 5853
+    },
+    {
+      "epoch": 2.9045550453022217,
+      "grad_norm": 0.12737767537759337,
+      "learning_rate": 8.70516490175393e-07,
+      "loss": 0.714,
+      "step": 5854
+    },
+    {
+      "epoch": 2.905051508005461,
+      "grad_norm": 0.12518922341713093,
+      "learning_rate": 8.697753961059385e-07,
+      "loss": 0.6743,
+      "step": 5855
+    },
+    {
+      "epoch": 2.9055479707087004,
+      "grad_norm": 0.13072186624967902,
+      "learning_rate": 8.690345511901771e-07,
+      "loss": 0.7178,
+      "step": 5856
+    },
+    {
+      "epoch": 2.9060444334119397,
+      "grad_norm": 0.12797230065596052,
+      "learning_rate": 8.682939555413334e-07,
+      "loss": 0.6979,
+      "step": 5857
+    },
+    {
+      "epoch": 2.906540896115179,
+      "grad_norm": 0.13125850276158907,
+      "learning_rate": 8.675536092725967e-07,
+      "loss": 0.6864,
+      "step": 5858
+    },
+    {
+      "epoch": 2.907037358818419,
+      "grad_norm": 0.12641599740685236,
+      "learning_rate": 8.668135124971166e-07,
+      "loss": 0.6801,
+      "step": 5859
+    },
+    {
+      "epoch": 2.907533821521658,
+      "grad_norm": 0.127803193061895,
+      "learning_rate": 8.660736653280041e-07,
+      "loss": 0.6713,
+      "step": 5860
+    },
+    {
+      "epoch": 2.9080302842248975,
+      "grad_norm": 0.13027630370244397,
+      "learning_rate": 8.653340678783347e-07,
+      "loss": 0.6756,
+      "step": 5861
+    },
+    {
+      "epoch": 2.9085267469281373,
+      "grad_norm": 0.12973586850281724,
+      "learning_rate": 8.645947202611423e-07,
+      "loss": 0.6992,
+      "step": 5862
+    },
+    {
+      "epoch": 2.9090232096313766,
+      "grad_norm": 0.1325303303913486,
+      "learning_rate": 8.638556225894271e-07,
+      "loss": 0.7145,
+      "step": 5863
+    },
+    {
+      "epoch": 2.909519672334616,
+      "grad_norm": 0.12990059009224683,
+      "learning_rate": 8.631167749761443e-07,
+      "loss": 0.6904,
+      "step": 5864
+    },
+    {
+      "epoch": 2.9100161350378553,
+      "grad_norm": 0.131277660839979,
+      "learning_rate": 8.623781775342183e-07,
+      "loss": 0.7115,
+      "step": 5865
+    },
+    {
+      "epoch": 2.9105125977410946,
+      "grad_norm": 0.1353236422896994,
+      "learning_rate": 8.616398303765303e-07,
+      "loss": 0.7022,
+      "step": 5866
+    },
+    {
+      "epoch": 2.911009060444334,
+      "grad_norm": 0.13235776611917602,
+      "learning_rate": 8.609017336159243e-07,
+      "loss": 0.6841,
+      "step": 5867
+    },
+    {
+      "epoch": 2.9115055231475733,
+      "grad_norm": 0.1352005618433371,
+      "learning_rate": 8.601638873652082e-07,
+      "loss": 0.7363,
+      "step": 5868
+    },
+    {
+      "epoch": 2.912001985850813,
+      "grad_norm": 0.1259476367266435,
+      "learning_rate": 8.594262917371485e-07,
+      "loss": 0.6729,
+      "step": 5869
+    },
+    {
+      "epoch": 2.9124984485540524,
+      "grad_norm": 0.1260089860773777,
+      "learning_rate": 8.586889468444761e-07,
+      "loss": 0.6744,
+      "step": 5870
+    },
+    {
+      "epoch": 2.9129949112572917,
+      "grad_norm": 0.13405412341583484,
+      "learning_rate": 8.57951852799882e-07,
+      "loss": 0.7332,
+      "step": 5871
+    },
+    {
+      "epoch": 2.9134913739605315,
+      "grad_norm": 0.12637262916741504,
+      "learning_rate": 8.572150097160179e-07,
+      "loss": 0.7096,
+      "step": 5872
+    },
+    {
+      "epoch": 2.913987836663771,
+      "grad_norm": 0.12478537511556445,
+      "learning_rate": 8.564784177055005e-07,
+      "loss": 0.6937,
+      "step": 5873
+    },
+    {
+      "epoch": 2.91448429936701,
+      "grad_norm": 0.13209726484500378,
+      "learning_rate": 8.557420768809041e-07,
+      "loss": 0.7309,
+      "step": 5874
+    },
+    {
+      "epoch": 2.9149807620702495,
+      "grad_norm": 0.13508309020119716,
+      "learning_rate": 8.550059873547686e-07,
+      "loss": 0.6698,
+      "step": 5875
+    },
+    {
+      "epoch": 2.915477224773489,
+      "grad_norm": 0.12122947126692105,
+      "learning_rate": 8.542701492395924e-07,
+      "loss": 0.6677,
+      "step": 5876
+    },
+    {
+      "epoch": 2.915973687476728,
+      "grad_norm": 0.12335372065030592,
+      "learning_rate": 8.535345626478355e-07,
+      "loss": 0.6776,
+      "step": 5877
+    },
+    {
+      "epoch": 2.9164701501799675,
+      "grad_norm": 0.12779589751503279,
+      "learning_rate": 8.527992276919228e-07,
+      "loss": 0.6466,
+      "step": 5878
+    },
+    {
+      "epoch": 2.9169666128832072,
+      "grad_norm": 0.12792448170766138,
+      "learning_rate": 8.520641444842373e-07,
+      "loss": 0.6864,
+      "step": 5879
+    },
+    {
+      "epoch": 2.9174630755864466,
+      "grad_norm": 0.12567621559611838,
+      "learning_rate": 8.513293131371245e-07,
+      "loss": 0.7195,
+      "step": 5880
+    },
+    {
+      "epoch": 2.917959538289686,
+      "grad_norm": 0.12846917163481486,
+      "learning_rate": 8.50594733762892e-07,
+      "loss": 0.7167,
+      "step": 5881
+    },
+    {
+      "epoch": 2.9184560009929257,
+      "grad_norm": 0.13195032168838147,
+      "learning_rate": 8.498604064738072e-07,
+      "loss": 0.7381,
+      "step": 5882
+    },
+    {
+      "epoch": 2.918952463696165,
+      "grad_norm": 0.13038989198058165,
+      "learning_rate": 8.491263313821021e-07,
+      "loss": 0.6773,
+      "step": 5883
+    },
+    {
+      "epoch": 2.9194489263994043,
+      "grad_norm": 0.12624725579988283,
+      "learning_rate": 8.483925085999667e-07,
+      "loss": 0.6843,
+      "step": 5884
+    },
+    {
+      "epoch": 2.9199453891026437,
+      "grad_norm": 0.1289543904324568,
+      "learning_rate": 8.476589382395558e-07,
+      "loss": 0.7618,
+      "step": 5885
+    },
+    {
+      "epoch": 2.920441851805883,
+      "grad_norm": 0.12430274931985501,
+      "learning_rate": 8.469256204129827e-07,
+      "loss": 0.6734,
+      "step": 5886
+    },
+    {
+      "epoch": 2.9209383145091223,
+      "grad_norm": 0.13183276150434192,
+      "learning_rate": 8.461925552323231e-07,
+      "loss": 0.6935,
+      "step": 5887
+    },
+    {
+      "epoch": 2.9214347772123617,
+      "grad_norm": 0.12471273296702598,
+      "learning_rate": 8.45459742809615e-07,
+      "loss": 0.6639,
+      "step": 5888
+    },
+    {
+      "epoch": 2.9219312399156014,
+      "grad_norm": 0.13312244162746337,
+      "learning_rate": 8.44727183256856e-07,
+      "loss": 0.6958,
+      "step": 5889
+    },
+    {
+      "epoch": 2.9224277026188408,
+      "grad_norm": 0.12804074982988345,
+      "learning_rate": 8.439948766860076e-07,
+      "loss": 0.7105,
+      "step": 5890
+    },
+    {
+      "epoch": 2.92292416532208,
+      "grad_norm": 0.13266023301103108,
+      "learning_rate": 8.432628232089904e-07,
+      "loss": 0.7281,
+      "step": 5891
+    },
+    {
+      "epoch": 2.9234206280253194,
+      "grad_norm": 0.12978177617631043,
+      "learning_rate": 8.42531022937686e-07,
+      "loss": 0.6977,
+      "step": 5892
+    },
+    {
+      "epoch": 2.923917090728559,
+      "grad_norm": 0.1309078003351543,
+      "learning_rate": 8.417994759839401e-07,
+      "loss": 0.6945,
+      "step": 5893
+    },
+    {
+      "epoch": 2.9244135534317985,
+      "grad_norm": 0.12245231712292412,
+      "learning_rate": 8.41068182459556e-07,
+      "loss": 0.6611,
+      "step": 5894
+    },
+    {
+      "epoch": 2.924910016135038,
+      "grad_norm": 0.1281010559347382,
+      "learning_rate": 8.403371424763032e-07,
+      "loss": 0.6796,
+      "step": 5895
+    },
+    {
+      "epoch": 2.925406478838277,
+      "grad_norm": 0.13285716334194994,
+      "learning_rate": 8.396063561459058e-07,
+      "loss": 0.7713,
+      "step": 5896
+    },
+    {
+      "epoch": 2.9259029415415165,
+      "grad_norm": 0.12525104544597734,
+      "learning_rate": 8.388758235800551e-07,
+      "loss": 0.669,
+      "step": 5897
+    },
+    {
+      "epoch": 2.926399404244756,
+      "grad_norm": 0.12692092156468257,
+      "learning_rate": 8.381455448904008e-07,
+      "loss": 0.716,
+      "step": 5898
+    },
+    {
+      "epoch": 2.9268958669479956,
+      "grad_norm": 0.12577897045335742,
+      "learning_rate": 8.374155201885533e-07,
+      "loss": 0.7094,
+      "step": 5899
+    },
+    {
+      "epoch": 2.927392329651235,
+      "grad_norm": 0.1270829254888127,
+      "learning_rate": 8.366857495860869e-07,
+      "loss": 0.7035,
+      "step": 5900
+    },
+    {
+      "epoch": 2.9278887923544743,
+      "grad_norm": 0.134187041399689,
+      "learning_rate": 8.359562331945337e-07,
+      "loss": 0.7231,
+      "step": 5901
+    },
+    {
+      "epoch": 2.9283852550577136,
+      "grad_norm": 0.13133787410693823,
+      "learning_rate": 8.352269711253902e-07,
+      "loss": 0.7019,
+      "step": 5902
+    },
+    {
+      "epoch": 2.9288817177609534,
+      "grad_norm": 0.13216722658500193,
+      "learning_rate": 8.344979634901115e-07,
+      "loss": 0.7305,
+      "step": 5903
+    },
+    {
+      "epoch": 2.9293781804641927,
+      "grad_norm": 0.12991923696083893,
+      "learning_rate": 8.337692104001138e-07,
+      "loss": 0.72,
+      "step": 5904
+    },
+    {
+      "epoch": 2.929874643167432,
+      "grad_norm": 0.13021761831749384,
+      "learning_rate": 8.330407119667775e-07,
+      "loss": 0.7079,
+      "step": 5905
+    },
+    {
+      "epoch": 2.9303711058706714,
+      "grad_norm": 0.12848310312169373,
+      "learning_rate": 8.323124683014394e-07,
+      "loss": 0.7022,
+      "step": 5906
+    },
+    {
+      "epoch": 2.9308675685739107,
+      "grad_norm": 0.12155773897471595,
+      "learning_rate": 8.315844795154024e-07,
+      "loss": 0.7003,
+      "step": 5907
+    },
+    {
+      "epoch": 2.93136403127715,
+      "grad_norm": 0.12783767747738128,
+      "learning_rate": 8.308567457199265e-07,
+      "loss": 0.6956,
+      "step": 5908
+    },
+    {
+      "epoch": 2.93186049398039,
+      "grad_norm": 0.12517915661300882,
+      "learning_rate": 8.301292670262334e-07,
+      "loss": 0.6805,
+      "step": 5909
+    },
+    {
+      "epoch": 2.932356956683629,
+      "grad_norm": 0.12818781769460139,
+      "learning_rate": 8.294020435455085e-07,
+      "loss": 0.6837,
+      "step": 5910
+    },
+    {
+      "epoch": 2.9328534193868685,
+      "grad_norm": 0.12740869036254676,
+      "learning_rate": 8.286750753888953e-07,
+      "loss": 0.7211,
+      "step": 5911
+    },
+    {
+      "epoch": 2.933349882090108,
+      "grad_norm": 0.12327113825475457,
+      "learning_rate": 8.279483626674992e-07,
+      "loss": 0.6925,
+      "step": 5912
+    },
+    {
+      "epoch": 2.9338463447933476,
+      "grad_norm": 0.12387658865943076,
+      "learning_rate": 8.272219054923855e-07,
+      "loss": 0.6407,
+      "step": 5913
+    },
+    {
+      "epoch": 2.934342807496587,
+      "grad_norm": 0.12861942462803927,
+      "learning_rate": 8.264957039745835e-07,
+      "loss": 0.6996,
+      "step": 5914
+    },
+    {
+      "epoch": 2.9348392701998263,
+      "grad_norm": 0.12836506417919924,
+      "learning_rate": 8.257697582250807e-07,
+      "loss": 0.7062,
+      "step": 5915
+    },
+    {
+      "epoch": 2.9353357329030656,
+      "grad_norm": 0.13217483481431683,
+      "learning_rate": 8.250440683548253e-07,
+      "loss": 0.7346,
+      "step": 5916
+    },
+    {
+      "epoch": 2.935832195606305,
+      "grad_norm": 0.12771709009606644,
+      "learning_rate": 8.243186344747286e-07,
+      "loss": 0.7331,
+      "step": 5917
+    },
+    {
+      "epoch": 2.9363286583095443,
+      "grad_norm": 0.12511348202532716,
+      "learning_rate": 8.235934566956616e-07,
+      "loss": 0.6774,
+      "step": 5918
+    },
+    {
+      "epoch": 2.936825121012784,
+      "grad_norm": 0.1304719690078209,
+      "learning_rate": 8.228685351284547e-07,
+      "loss": 0.7227,
+      "step": 5919
+    },
+    {
+      "epoch": 2.9373215837160234,
+      "grad_norm": 0.12839066350051087,
+      "learning_rate": 8.221438698839021e-07,
+      "loss": 0.6943,
+      "step": 5920
+    },
+    {
+      "epoch": 2.9378180464192627,
+      "grad_norm": 0.12944145552846098,
+      "learning_rate": 8.214194610727557e-07,
+      "loss": 0.7047,
+      "step": 5921
+    },
+    {
+      "epoch": 2.938314509122502,
+      "grad_norm": 0.12851146335009603,
+      "learning_rate": 8.206953088057318e-07,
+      "loss": 0.7166,
+      "step": 5922
+    },
+    {
+      "epoch": 2.938810971825742,
+      "grad_norm": 0.13209306286470085,
+      "learning_rate": 8.199714131935041e-07,
+      "loss": 0.676,
+      "step": 5923
+    },
+    {
+      "epoch": 2.939307434528981,
+      "grad_norm": 0.13328728796070516,
+      "learning_rate": 8.192477743467078e-07,
+      "loss": 0.6941,
+      "step": 5924
+    },
+    {
+      "epoch": 2.9398038972322205,
+      "grad_norm": 0.12676782708503537,
+      "learning_rate": 8.185243923759414e-07,
+      "loss": 0.716,
+      "step": 5925
+    },
+    {
+      "epoch": 2.94030035993546,
+      "grad_norm": 0.12845269590237002,
+      "learning_rate": 8.178012673917604e-07,
+      "loss": 0.6856,
+      "step": 5926
+    },
+    {
+      "epoch": 2.940796822638699,
+      "grad_norm": 0.12791983707358917,
+      "learning_rate": 8.170783995046852e-07,
+      "loss": 0.7007,
+      "step": 5927
+    },
+    {
+      "epoch": 2.9412932853419385,
+      "grad_norm": 0.12653344264597674,
+      "learning_rate": 8.163557888251916e-07,
+      "loss": 0.6654,
+      "step": 5928
+    },
+    {
+      "epoch": 2.9417897480451782,
+      "grad_norm": 0.1311894343517124,
+      "learning_rate": 8.15633435463721e-07,
+      "loss": 0.6941,
+      "step": 5929
+    },
+    {
+      "epoch": 2.9422862107484176,
+      "grad_norm": 0.13045379771212226,
+      "learning_rate": 8.149113395306732e-07,
+      "loss": 0.7029,
+      "step": 5930
+    },
+    {
+      "epoch": 2.942782673451657,
+      "grad_norm": 0.1310448907057462,
+      "learning_rate": 8.141895011364082e-07,
+      "loss": 0.6971,
+      "step": 5931
+    },
+    {
+      "epoch": 2.9432791361548962,
+      "grad_norm": 0.12684756193510152,
+      "learning_rate": 8.134679203912485e-07,
+      "loss": 0.6932,
+      "step": 5932
+    },
+    {
+      "epoch": 2.943775598858136,
+      "grad_norm": 0.13624799129435985,
+      "learning_rate": 8.127465974054749e-07,
+      "loss": 0.7143,
+      "step": 5933
+    },
+    {
+      "epoch": 2.9442720615613753,
+      "grad_norm": 0.13275208970745384,
+      "learning_rate": 8.120255322893319e-07,
+      "loss": 0.7652,
+      "step": 5934
+    },
+    {
+      "epoch": 2.9447685242646147,
+      "grad_norm": 0.13231133707193993,
+      "learning_rate": 8.113047251530215e-07,
+      "loss": 0.7276,
+      "step": 5935
+    },
+    {
+      "epoch": 2.945264986967854,
+      "grad_norm": 0.12564669641658913,
+      "learning_rate": 8.105841761067068e-07,
+      "loss": 0.6761,
+      "step": 5936
+    },
+    {
+      "epoch": 2.9457614496710933,
+      "grad_norm": 0.12956694166664687,
+      "learning_rate": 8.098638852605139e-07,
+      "loss": 0.6975,
+      "step": 5937
+    },
+    {
+      "epoch": 2.9462579123743327,
+      "grad_norm": 0.12673479579130228,
+      "learning_rate": 8.091438527245263e-07,
+      "loss": 0.659,
+      "step": 5938
+    },
+    {
+      "epoch": 2.9467543750775724,
+      "grad_norm": 0.12543122612508772,
+      "learning_rate": 8.084240786087907e-07,
+      "loss": 0.7008,
+      "step": 5939
+    },
+    {
+      "epoch": 2.9472508377808118,
+      "grad_norm": 0.13367896037829807,
+      "learning_rate": 8.077045630233121e-07,
+      "loss": 0.7256,
+      "step": 5940
+    },
+    {
+      "epoch": 2.947747300484051,
+      "grad_norm": 0.13359993578888146,
+      "learning_rate": 8.069853060780566e-07,
+      "loss": 0.6878,
+      "step": 5941
+    },
+    {
+      "epoch": 2.9482437631872904,
+      "grad_norm": 0.12701732001687444,
+      "learning_rate": 8.062663078829524e-07,
+      "loss": 0.7399,
+      "step": 5942
+    },
+    {
+      "epoch": 2.94874022589053,
+      "grad_norm": 0.12747058578699347,
+      "learning_rate": 8.055475685478861e-07,
+      "loss": 0.6999,
+      "step": 5943
+    },
+    {
+      "epoch": 2.9492366885937695,
+      "grad_norm": 0.12657957164210343,
+      "learning_rate": 8.048290881827053e-07,
+      "loss": 0.7354,
+      "step": 5944
+    },
+    {
+      "epoch": 2.949733151297009,
+      "grad_norm": 0.12534187082541948,
+      "learning_rate": 8.041108668972178e-07,
+      "loss": 0.6951,
+      "step": 5945
+    },
+    {
+      "epoch": 2.950229614000248,
+      "grad_norm": 0.12398448643069547,
+      "learning_rate": 8.033929048011934e-07,
+      "loss": 0.6824,
+      "step": 5946
+    },
+    {
+      "epoch": 2.9507260767034875,
+      "grad_norm": 0.13140028368522533,
+      "learning_rate": 8.026752020043604e-07,
+      "loss": 0.6993,
+      "step": 5947
+    },
+    {
+      "epoch": 2.951222539406727,
+      "grad_norm": 0.1339567215216009,
+      "learning_rate": 8.019577586164077e-07,
+      "loss": 0.6883,
+      "step": 5948
+    },
+    {
+      "epoch": 2.9517190021099666,
+      "grad_norm": 0.13254826072565667,
+      "learning_rate": 8.012405747469861e-07,
+      "loss": 0.7031,
+      "step": 5949
+    },
+    {
+      "epoch": 2.952215464813206,
+      "grad_norm": 0.13203437830444636,
+      "learning_rate": 8.005236505057045e-07,
+      "loss": 0.6924,
+      "step": 5950
+    },
+    {
+      "epoch": 2.9527119275164453,
+      "grad_norm": 0.13328318102042208,
+      "learning_rate": 7.998069860021346e-07,
+      "loss": 0.735,
+      "step": 5951
+    },
+    {
+      "epoch": 2.9532083902196846,
+      "grad_norm": 0.12989959931327805,
+      "learning_rate": 7.990905813458063e-07,
+      "loss": 0.7524,
+      "step": 5952
+    },
+    {
+      "epoch": 2.9537048529229244,
+      "grad_norm": 0.12932924465141693,
+      "learning_rate": 7.983744366462101e-07,
+      "loss": 0.7075,
+      "step": 5953
+    },
+    {
+      "epoch": 2.9542013156261637,
+      "grad_norm": 0.1270536630172262,
+      "learning_rate": 7.976585520127982e-07,
+      "loss": 0.6864,
+      "step": 5954
+    },
+    {
+      "epoch": 2.954697778329403,
+      "grad_norm": 0.13045118616961768,
+      "learning_rate": 7.969429275549819e-07,
+      "loss": 0.6788,
+      "step": 5955
+    },
+    {
+      "epoch": 2.9551942410326424,
+      "grad_norm": 0.1273750155772813,
+      "learning_rate": 7.962275633821321e-07,
+      "loss": 0.7083,
+      "step": 5956
+    },
+    {
+      "epoch": 2.9556907037358817,
+      "grad_norm": 0.129306537947579,
+      "learning_rate": 7.955124596035818e-07,
+      "loss": 0.7025,
+      "step": 5957
+    },
+    {
+      "epoch": 2.956187166439121,
+      "grad_norm": 0.13380324003505628,
+      "learning_rate": 7.947976163286219e-07,
+      "loss": 0.7551,
+      "step": 5958
+    },
+    {
+      "epoch": 2.956683629142361,
+      "grad_norm": 0.1316194370579435,
+      "learning_rate": 7.940830336665071e-07,
+      "loss": 0.7246,
+      "step": 5959
+    },
+    {
+      "epoch": 2.9571800918456,
+      "grad_norm": 0.12465311080275596,
+      "learning_rate": 7.933687117264469e-07,
+      "loss": 0.6732,
+      "step": 5960
+    },
+    {
+      "epoch": 2.9576765545488395,
+      "grad_norm": 0.1308457845292269,
+      "learning_rate": 7.926546506176158e-07,
+      "loss": 0.7373,
+      "step": 5961
+    },
+    {
+      "epoch": 2.958173017252079,
+      "grad_norm": 0.13252382874716395,
+      "learning_rate": 7.919408504491464e-07,
+      "loss": 0.7194,
+      "step": 5962
+    },
+    {
+      "epoch": 2.9586694799553186,
+      "grad_norm": 0.12466359487318711,
+      "learning_rate": 7.912273113301306e-07,
+      "loss": 0.6609,
+      "step": 5963
+    },
+    {
+      "epoch": 2.959165942658558,
+      "grad_norm": 0.13237910505477793,
+      "learning_rate": 7.905140333696229e-07,
+      "loss": 0.7391,
+      "step": 5964
+    },
+    {
+      "epoch": 2.9596624053617973,
+      "grad_norm": 0.13205570481341808,
+      "learning_rate": 7.898010166766348e-07,
+      "loss": 0.7033,
+      "step": 5965
+    },
+    {
+      "epoch": 2.9601588680650366,
+      "grad_norm": 0.12754936984318843,
+      "learning_rate": 7.890882613601411e-07,
+      "loss": 0.7007,
+      "step": 5966
+    },
+    {
+      "epoch": 2.960655330768276,
+      "grad_norm": 0.12850522933784753,
+      "learning_rate": 7.883757675290746e-07,
+      "loss": 0.7058,
+      "step": 5967
+    },
+    {
+      "epoch": 2.9611517934715152,
+      "grad_norm": 0.12454681696499198,
+      "learning_rate": 7.87663535292327e-07,
+      "loss": 0.6232,
+      "step": 5968
+    },
+    {
+      "epoch": 2.961648256174755,
+      "grad_norm": 0.12513438738198798,
+      "learning_rate": 7.869515647587539e-07,
+      "loss": 0.6939,
+      "step": 5969
+    },
+    {
+      "epoch": 2.9621447188779944,
+      "grad_norm": 0.12503092799347673,
+      "learning_rate": 7.862398560371665e-07,
+      "loss": 0.6937,
+      "step": 5970
+    },
+    {
+      "epoch": 2.9626411815812337,
+      "grad_norm": 0.1313520263634869,
+      "learning_rate": 7.855284092363399e-07,
+      "loss": 0.7132,
+      "step": 5971
+    },
+    {
+      "epoch": 2.963137644284473,
+      "grad_norm": 0.1324830049247509,
+      "learning_rate": 7.848172244650065e-07,
+      "loss": 0.708,
+      "step": 5972
+    },
+    {
+      "epoch": 2.963634106987713,
+      "grad_norm": 0.13238004154599406,
+      "learning_rate": 7.841063018318587e-07,
+      "loss": 0.7363,
+      "step": 5973
+    },
+    {
+      "epoch": 2.964130569690952,
+      "grad_norm": 0.12851537121562095,
+      "learning_rate": 7.833956414455512e-07,
+      "loss": 0.7088,
+      "step": 5974
+    },
+    {
+      "epoch": 2.9646270323941915,
+      "grad_norm": 0.12750870312778442,
+      "learning_rate": 7.826852434146964e-07,
+      "loss": 0.6739,
+      "step": 5975
+    },
+    {
+      "epoch": 2.965123495097431,
+      "grad_norm": 0.13032488783779714,
+      "learning_rate": 7.819751078478669e-07,
+      "loss": 0.6985,
+      "step": 5976
+    },
+    {
+      "epoch": 2.96561995780067,
+      "grad_norm": 0.1280481984781406,
+      "learning_rate": 7.812652348535948e-07,
+      "loss": 0.7029,
+      "step": 5977
+    },
+    {
+      "epoch": 2.9661164205039094,
+      "grad_norm": 0.12481852362878608,
+      "learning_rate": 7.805556245403748e-07,
+      "loss": 0.6592,
+      "step": 5978
+    },
+    {
+      "epoch": 2.966612883207149,
+      "grad_norm": 0.12694865638365466,
+      "learning_rate": 7.798462770166585e-07,
+      "loss": 0.6451,
+      "step": 5979
+    },
+    {
+      "epoch": 2.9671093459103886,
+      "grad_norm": 0.12968124976820739,
+      "learning_rate": 7.791371923908569e-07,
+      "loss": 0.7499,
+      "step": 5980
+    },
+    {
+      "epoch": 2.967605808613628,
+      "grad_norm": 0.12879024482404638,
+      "learning_rate": 7.784283707713447e-07,
+      "loss": 0.7268,
+      "step": 5981
+    },
+    {
+      "epoch": 2.968102271316867,
+      "grad_norm": 0.13142308258429045,
+      "learning_rate": 7.77719812266452e-07,
+      "loss": 0.7254,
+      "step": 5982
+    },
+    {
+      "epoch": 2.968598734020107,
+      "grad_norm": 0.12715116614343377,
+      "learning_rate": 7.770115169844719e-07,
+      "loss": 0.6715,
+      "step": 5983
+    },
+    {
+      "epoch": 2.9690951967233463,
+      "grad_norm": 0.12867668360826426,
+      "learning_rate": 7.763034850336554e-07,
+      "loss": 0.7037,
+      "step": 5984
+    },
+    {
+      "epoch": 2.9695916594265857,
+      "grad_norm": 0.13174571937124394,
+      "learning_rate": 7.75595716522213e-07,
+      "loss": 0.7087,
+      "step": 5985
+    },
+    {
+      "epoch": 2.970088122129825,
+      "grad_norm": 0.12868392659804886,
+      "learning_rate": 7.748882115583175e-07,
+      "loss": 0.7026,
+      "step": 5986
+    },
+    {
+      "epoch": 2.9705845848330643,
+      "grad_norm": 0.1304279274224821,
+      "learning_rate": 7.741809702500983e-07,
+      "loss": 0.68,
+      "step": 5987
+    },
+    {
+      "epoch": 2.9710810475363036,
+      "grad_norm": 0.13024362648385276,
+      "learning_rate": 7.734739927056467e-07,
+      "loss": 0.7152,
+      "step": 5988
+    },
+    {
+      "epoch": 2.9715775102395434,
+      "grad_norm": 0.13003977622755503,
+      "learning_rate": 7.727672790330129e-07,
+      "loss": 0.6805,
+      "step": 5989
+    },
+    {
+      "epoch": 2.9720739729427827,
+      "grad_norm": 0.1307381982553933,
+      "learning_rate": 7.720608293402055e-07,
+      "loss": 0.7367,
+      "step": 5990
+    },
+    {
+      "epoch": 2.972570435646022,
+      "grad_norm": 0.13211408639458272,
+      "learning_rate": 7.713546437351965e-07,
+      "loss": 0.7141,
+      "step": 5991
+    },
+    {
+      "epoch": 2.9730668983492614,
+      "grad_norm": 0.1247955923407189,
+      "learning_rate": 7.706487223259121e-07,
+      "loss": 0.6826,
+      "step": 5992
+    },
+    {
+      "epoch": 2.973563361052501,
+      "grad_norm": 0.13116304767044074,
+      "learning_rate": 7.69943065220243e-07,
+      "loss": 0.7277,
+      "step": 5993
+    },
+    {
+      "epoch": 2.9740598237557405,
+      "grad_norm": 0.13479680364255306,
+      "learning_rate": 7.692376725260369e-07,
+      "loss": 0.6846,
+      "step": 5994
+    },
+    {
+      "epoch": 2.97455628645898,
+      "grad_norm": 0.12782009680839385,
+      "learning_rate": 7.685325443511015e-07,
+      "loss": 0.6846,
+      "step": 5995
+    },
+    {
+      "epoch": 2.975052749162219,
+      "grad_norm": 0.1336362278759992,
+      "learning_rate": 7.678276808032054e-07,
+      "loss": 0.7207,
+      "step": 5996
+    },
+    {
+      "epoch": 2.9755492118654585,
+      "grad_norm": 0.12954084216492223,
+      "learning_rate": 7.671230819900741e-07,
+      "loss": 0.722,
+      "step": 5997
+    },
+    {
+      "epoch": 2.976045674568698,
+      "grad_norm": 0.12870683674641,
+      "learning_rate": 7.66418748019396e-07,
+      "loss": 0.6928,
+      "step": 5998
+    },
+    {
+      "epoch": 2.976542137271937,
+      "grad_norm": 0.1283984345940394,
+      "learning_rate": 7.657146789988165e-07,
+      "loss": 0.6856,
+      "step": 5999
+    },
+    {
+      "epoch": 2.977038599975177,
+      "grad_norm": 0.12793776246424288,
+      "learning_rate": 7.650108750359403e-07,
+      "loss": 0.6919,
+      "step": 6000
+    },
+    {
+      "epoch": 2.9775350626784163,
+      "grad_norm": 0.12577650631266427,
+      "learning_rate": 7.643073362383341e-07,
+      "loss": 0.6853,
+      "step": 6001
+    },
+    {
+      "epoch": 2.9780315253816556,
+      "grad_norm": 0.12392465457564755,
+      "learning_rate": 7.636040627135211e-07,
+      "loss": 0.722,
+      "step": 6002
+    },
+    {
+      "epoch": 2.9785279880848954,
+      "grad_norm": 0.12418251160738356,
+      "learning_rate": 7.629010545689869e-07,
+      "loss": 0.7056,
+      "step": 6003
+    },
+    {
+      "epoch": 2.9790244507881347,
+      "grad_norm": 0.1271367787777443,
+      "learning_rate": 7.621983119121742e-07,
+      "loss": 0.6494,
+      "step": 6004
+    },
+    {
+      "epoch": 2.979520913491374,
+      "grad_norm": 0.12870235470062621,
+      "learning_rate": 7.614958348504853e-07,
+      "loss": 0.7274,
+      "step": 6005
+    },
+    {
+      "epoch": 2.9800173761946134,
+      "grad_norm": 0.12786870456842656,
+      "learning_rate": 7.607936234912841e-07,
+      "loss": 0.6715,
+      "step": 6006
+    },
+    {
+      "epoch": 2.9805138388978527,
+      "grad_norm": 0.12959870966522002,
+      "learning_rate": 7.600916779418916e-07,
+      "loss": 0.7442,
+      "step": 6007
+    },
+    {
+      "epoch": 2.981010301601092,
+      "grad_norm": 0.12795538478138033,
+      "learning_rate": 7.593899983095884e-07,
+      "loss": 0.6919,
+      "step": 6008
+    },
+    {
+      "epoch": 2.9815067643043314,
+      "grad_norm": 0.14008517233459958,
+      "learning_rate": 7.586885847016148e-07,
+      "loss": 0.7013,
+      "step": 6009
+    },
+    {
+      "epoch": 2.982003227007571,
+      "grad_norm": 0.1273352199473159,
+      "learning_rate": 7.579874372251722e-07,
+      "loss": 0.7015,
+      "step": 6010
+    },
+    {
+      "epoch": 2.9824996897108105,
+      "grad_norm": 0.12663695622291235,
+      "learning_rate": 7.572865559874188e-07,
+      "loss": 0.7045,
+      "step": 6011
+    },
+    {
+      "epoch": 2.98299615241405,
+      "grad_norm": 0.1309777399910531,
+      "learning_rate": 7.565859410954718e-07,
+      "loss": 0.7366,
+      "step": 6012
+    },
+    {
+      "epoch": 2.9834926151172896,
+      "grad_norm": 0.1281630321138167,
+      "learning_rate": 7.558855926564112e-07,
+      "loss": 0.7116,
+      "step": 6013
+    },
+    {
+      "epoch": 2.983989077820529,
+      "grad_norm": 0.1280931136947841,
+      "learning_rate": 7.551855107772724e-07,
+      "loss": 0.7073,
+      "step": 6014
+    },
+    {
+      "epoch": 2.9844855405237682,
+      "grad_norm": 0.12459877040568573,
+      "learning_rate": 7.544856955650532e-07,
+      "loss": 0.6964,
+      "step": 6015
+    },
+    {
+      "epoch": 2.9849820032270076,
+      "grad_norm": 0.12667593238205743,
+      "learning_rate": 7.537861471267077e-07,
+      "loss": 0.6639,
+      "step": 6016
+    },
+    {
+      "epoch": 2.985478465930247,
+      "grad_norm": 0.12902244841475374,
+      "learning_rate": 7.530868655691509e-07,
+      "loss": 0.7196,
+      "step": 6017
+    },
+    {
+      "epoch": 2.9859749286334862,
+      "grad_norm": 0.12779617048573458,
+      "learning_rate": 7.523878509992578e-07,
+      "loss": 0.6979,
+      "step": 6018
+    },
+    {
+      "epoch": 2.9864713913367256,
+      "grad_norm": 0.12907365853546054,
+      "learning_rate": 7.516891035238596e-07,
+      "loss": 0.682,
+      "step": 6019
+    },
+    {
+      "epoch": 2.9869678540399653,
+      "grad_norm": 0.12929753241306882,
+      "learning_rate": 7.509906232497513e-07,
+      "loss": 0.7007,
+      "step": 6020
+    },
+    {
+      "epoch": 2.9874643167432047,
+      "grad_norm": 0.12701020359185758,
+      "learning_rate": 7.502924102836826e-07,
+      "loss": 0.6893,
+      "step": 6021
+    },
+    {
+      "epoch": 2.987960779446444,
+      "grad_norm": 0.1237167641591724,
+      "learning_rate": 7.495944647323639e-07,
+      "loss": 0.709,
+      "step": 6022
+    },
+    {
+      "epoch": 2.988457242149684,
+      "grad_norm": 0.13193601872797872,
+      "learning_rate": 7.488967867024671e-07,
+      "loss": 0.7199,
+      "step": 6023
+    },
+    {
+      "epoch": 2.988953704852923,
+      "grad_norm": 0.12998564461199474,
+      "learning_rate": 7.481993763006184e-07,
+      "loss": 0.6595,
+      "step": 6024
+    },
+    {
+      "epoch": 2.9894501675561624,
+      "grad_norm": 0.13111171647856992,
+      "learning_rate": 7.475022336334075e-07,
+      "loss": 0.6868,
+      "step": 6025
+    },
+    {
+      "epoch": 2.9899466302594018,
+      "grad_norm": 0.12835689645376358,
+      "learning_rate": 7.468053588073803e-07,
+      "loss": 0.7071,
+      "step": 6026
+    },
+    {
+      "epoch": 2.990443092962641,
+      "grad_norm": 0.1259901601242702,
+      "learning_rate": 7.461087519290447e-07,
+      "loss": 0.7059,
+      "step": 6027
+    },
+    {
+      "epoch": 2.9909395556658804,
+      "grad_norm": 0.12268405384296684,
+      "learning_rate": 7.454124131048646e-07,
+      "loss": 0.6555,
+      "step": 6028
+    },
+    {
+      "epoch": 2.9914360183691198,
+      "grad_norm": 0.13427530622463077,
+      "learning_rate": 7.447163424412638e-07,
+      "loss": 0.7329,
+      "step": 6029
+    },
+    {
+      "epoch": 2.9919324810723595,
+      "grad_norm": 0.12630933697427535,
+      "learning_rate": 7.440205400446271e-07,
+      "loss": 0.7285,
+      "step": 6030
+    },
+    {
+      "epoch": 2.992428943775599,
+      "grad_norm": 0.1282172265353785,
+      "learning_rate": 7.433250060212957e-07,
+      "loss": 0.6894,
+      "step": 6031
+    },
+    {
+      "epoch": 2.992925406478838,
+      "grad_norm": 0.1242625795254969,
+      "learning_rate": 7.426297404775701e-07,
+      "loss": 0.6954,
+      "step": 6032
+    },
+    {
+      "epoch": 2.9934218691820775,
+      "grad_norm": 0.12573869185290845,
+      "learning_rate": 7.419347435197125e-07,
+      "loss": 0.6515,
+      "step": 6033
+    },
+    {
+      "epoch": 2.9939183318853173,
+      "grad_norm": 0.12632264327088447,
+      "learning_rate": 7.412400152539398e-07,
+      "loss": 0.6976,
+      "step": 6034
+    },
+    {
+      "epoch": 2.9944147945885566,
+      "grad_norm": 0.12492304063826659,
+      "learning_rate": 7.405455557864322e-07,
+      "loss": 0.6691,
+      "step": 6035
+    },
+    {
+      "epoch": 2.994911257291796,
+      "grad_norm": 0.12637656925192833,
+      "learning_rate": 7.398513652233255e-07,
+      "loss": 0.7189,
+      "step": 6036
+    },
+    {
+      "epoch": 2.9954077199950353,
+      "grad_norm": 0.12460961740826068,
+      "learning_rate": 7.391574436707153e-07,
+      "loss": 0.6838,
+      "step": 6037
+    },
+    {
+      "epoch": 2.9959041826982746,
+      "grad_norm": 0.13406612342755167,
+      "learning_rate": 7.384637912346573e-07,
+      "loss": 0.7276,
+      "step": 6038
+    },
+    {
+      "epoch": 2.996400645401514,
+      "grad_norm": 0.12791763802640216,
+      "learning_rate": 7.377704080211651e-07,
+      "loss": 0.7028,
+      "step": 6039
+    },
+    {
+      "epoch": 2.9968971081047537,
+      "grad_norm": 0.12705616081597926,
+      "learning_rate": 7.370772941362106e-07,
+      "loss": 0.7045,
+      "step": 6040
+    },
+    {
+      "epoch": 2.997393570807993,
+      "grad_norm": 0.13088558569032857,
+      "learning_rate": 7.363844496857247e-07,
+      "loss": 0.6877,
+      "step": 6041
+    },
+    {
+      "epoch": 2.9978900335112324,
+      "grad_norm": 0.12444783921112863,
+      "learning_rate": 7.356918747755989e-07,
+      "loss": 0.688,
+      "step": 6042
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 8056,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 2014,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1358095334768640.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}