diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,36428 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 5198,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00019238168526356292,
+      "grad_norm": 21.455511021876493,
+      "learning_rate": 1.282051282051282e-07,
+      "loss": 1.3371,
+      "step": 1
+    },
+    {
+      "epoch": 0.00038476337052712584,
+      "grad_norm": 12.011949794989873,
+      "learning_rate": 2.564102564102564e-07,
+      "loss": 1.4733,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005771450557906887,
+      "grad_norm": 21.672558063071218,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 1.4107,
+      "step": 3
+    },
+    {
+      "epoch": 0.0007695267410542517,
+      "grad_norm": 13.253985177581358,
+      "learning_rate": 5.128205128205128e-07,
+      "loss": 1.3312,
+      "step": 4
+    },
+    {
+      "epoch": 0.0009619084263178145,
+      "grad_norm": 5.573452151499016,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 1.3945,
+      "step": 5
+    },
+    {
+      "epoch": 0.0011542901115813775,
+      "grad_norm": 5.585404966582236,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 1.3861,
+      "step": 6
+    },
+    {
+      "epoch": 0.0013466717968449403,
+      "grad_norm": 5.803961834305012,
+      "learning_rate": 8.974358974358975e-07,
+      "loss": 1.3595,
+      "step": 7
+    },
+    {
+      "epoch": 0.0015390534821085034,
+      "grad_norm": 5.561422143770103,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 1.3418,
+      "step": 8
+    },
+    {
+      "epoch": 0.0017314351673720662,
+      "grad_norm": 63.25454536893955,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 1.4226,
+      "step": 9
+    },
+    {
+      "epoch": 0.001923816852635629,
+      "grad_norm": 4.437562529559971,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 1.3083,
+      "step": 10
+    },
+    {
+      "epoch": 0.002116198537899192,
+      "grad_norm": 4.439038068663812,
+      "learning_rate": 1.4102564102564104e-06,
+      "loss": 1.2915,
+      "step": 11
+    },
+    {
+      "epoch": 0.002308580223162755,
+      "grad_norm": 5.158091059734221,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 1.3661,
+      "step": 12
+    },
+    {
+      "epoch": 0.002500961908426318,
+      "grad_norm": 42.320438066060355,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.3375,
+      "step": 13
+    },
+    {
+      "epoch": 0.0026933435936898806,
+      "grad_norm": 5.745353264448665,
+      "learning_rate": 1.794871794871795e-06,
+      "loss": 1.2422,
+      "step": 14
+    },
+    {
+      "epoch": 0.0028857252789534437,
+      "grad_norm": 17.45917864560197,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 1.272,
+      "step": 15
+    },
+    {
+      "epoch": 0.0030781069642170067,
+      "grad_norm": 62.47752633019166,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 1.3202,
+      "step": 16
+    },
+    {
+      "epoch": 0.0032704886494805694,
+      "grad_norm": 0.7078173143024034,
+      "learning_rate": 2.1794871794871797e-06,
+      "loss": 0.2838,
+      "step": 17
+    },
+    {
+      "epoch": 0.0034628703347441324,
+      "grad_norm": 2.113296431200322,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 1.1816,
+      "step": 18
+    },
+    {
+      "epoch": 0.0036552520200076955,
+      "grad_norm": 5.289067163245759,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 1.2172,
+      "step": 19
+    },
+    {
+      "epoch": 0.003847633705271258,
+      "grad_norm": 6.231348620699606,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 1.238,
+      "step": 20
+    },
+    {
+      "epoch": 0.004040015390534821,
+      "grad_norm": 5.579588170874156,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 1.1757,
+      "step": 21
+    },
+    {
+      "epoch": 0.004232397075798384,
+      "grad_norm": 4.7647167424056835,
+      "learning_rate": 2.8205128205128207e-06,
+      "loss": 1.2201,
+      "step": 22
+    },
+    {
+      "epoch": 0.004424778761061947,
+      "grad_norm": 5.78987072552975,
+      "learning_rate": 2.948717948717949e-06,
+      "loss": 1.122,
+      "step": 23
+    },
+    {
+      "epoch": 0.00461716044632551,
+      "grad_norm": 30.870427546458384,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 1.1878,
+      "step": 24
+    },
+    {
+      "epoch": 0.004809542131589073,
+      "grad_norm": 2.735160461268804,
+      "learning_rate": 3.205128205128206e-06,
+      "loss": 1.1734,
+      "step": 25
+    },
+    {
+      "epoch": 0.005001923816852636,
+      "grad_norm": 2.0497589262748237,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.1416,
+      "step": 26
+    },
+    {
+      "epoch": 0.005194305502116198,
+      "grad_norm": 2.395602174951531,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 1.2011,
+      "step": 27
+    },
+    {
+      "epoch": 0.005386687187379761,
+      "grad_norm": 2.0028604206203715,
+      "learning_rate": 3.58974358974359e-06,
+      "loss": 1.1294,
+      "step": 28
+    },
+    {
+      "epoch": 0.005579068872643324,
+      "grad_norm": 1.7914400202699532,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 1.0771,
+      "step": 29
+    },
+    {
+      "epoch": 0.005771450557906887,
+      "grad_norm": 12.887043799173515,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 1.1455,
+      "step": 30
+    },
+    {
+      "epoch": 0.00596383224317045,
+      "grad_norm": 3.010841134015081,
+      "learning_rate": 3.974358974358974e-06,
+      "loss": 1.1053,
+      "step": 31
+    },
+    {
+      "epoch": 0.0061562139284340135,
+      "grad_norm": 85.94330802175908,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 1.1526,
+      "step": 32
+    },
+    {
+      "epoch": 0.006348595613697576,
+      "grad_norm": 2.0537641904102273,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 1.0991,
+      "step": 33
+    },
+    {
+      "epoch": 0.006540977298961139,
+      "grad_norm": 2.5766970021639466,
+      "learning_rate": 4.358974358974359e-06,
+      "loss": 1.0502,
+      "step": 34
+    },
+    {
+      "epoch": 0.006733358984224702,
+      "grad_norm": 1.9783013921154213,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 1.0522,
+      "step": 35
+    },
+    {
+      "epoch": 0.006925740669488265,
+      "grad_norm": 5.4283215230434445,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 1.0515,
+      "step": 36
+    },
+    {
+      "epoch": 0.007118122354751828,
+      "grad_norm": 2.662528918448472,
+      "learning_rate": 4.743589743589744e-06,
+      "loss": 1.1229,
+      "step": 37
+    },
+    {
+      "epoch": 0.007310504040015391,
+      "grad_norm": 1.696171983152647,
+      "learning_rate": 4.871794871794872e-06,
+      "loss": 1.0596,
+      "step": 38
+    },
+    {
+      "epoch": 0.007502885725278953,
+      "grad_norm": 3.957370035511471,
+      "learning_rate": 5e-06,
+      "loss": 1.0698,
+      "step": 39
+    },
+    {
+      "epoch": 0.007695267410542516,
+      "grad_norm": 15.30349963423196,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 1.0282,
+      "step": 40
+    },
+    {
+      "epoch": 0.007887649095806078,
+      "grad_norm": 12.248396830541529,
+      "learning_rate": 5.256410256410257e-06,
+      "loss": 1.0564,
+      "step": 41
+    },
+    {
+      "epoch": 0.008080030781069641,
+      "grad_norm": 1.5376467120444262,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 1.0043,
+      "step": 42
+    },
+    {
+      "epoch": 0.008272412466333205,
+      "grad_norm": 1.7599170750801088,
+      "learning_rate": 5.512820512820514e-06,
+      "loss": 1.0549,
+      "step": 43
+    },
+    {
+      "epoch": 0.008464794151596768,
+      "grad_norm": 1.51304054938919,
+      "learning_rate": 5.641025641025641e-06,
+      "loss": 1.0087,
+      "step": 44
+    },
+    {
+      "epoch": 0.00865717583686033,
+      "grad_norm": 1.5320399398575675,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 1.0241,
+      "step": 45
+    },
+    {
+      "epoch": 0.008849557522123894,
+      "grad_norm": 1.9939912848436403,
+      "learning_rate": 5.897435897435898e-06,
+      "loss": 1.0392,
+      "step": 46
+    },
+    {
+      "epoch": 0.009041939207387457,
+      "grad_norm": 1.7319482598451845,
+      "learning_rate": 6.025641025641026e-06,
+      "loss": 1.0884,
+      "step": 47
+    },
+    {
+      "epoch": 0.00923432089265102,
+      "grad_norm": 1.6283744561478977,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.9857,
+      "step": 48
+    },
+    {
+      "epoch": 0.009426702577914583,
+      "grad_norm": 2.1410056409644085,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 1.0604,
+      "step": 49
+    },
+    {
+      "epoch": 0.009619084263178146,
+      "grad_norm": 1.6734898815201478,
+      "learning_rate": 6.410256410256412e-06,
+      "loss": 1.059,
+      "step": 50
+    },
+    {
+      "epoch": 0.009811465948441709,
+      "grad_norm": 1.7152301160920311,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 1.0744,
+      "step": 51
+    },
+    {
+      "epoch": 0.010003847633705272,
+      "grad_norm": 1.6106519108074826,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.0352,
+      "step": 52
+    },
+    {
+      "epoch": 0.010196229318968833,
+      "grad_norm": 1.4895444136984546,
+      "learning_rate": 6.794871794871796e-06,
+      "loss": 1.0323,
+      "step": 53
+    },
+    {
+      "epoch": 0.010388611004232396,
+      "grad_norm": 15.155774258026407,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 1.0282,
+      "step": 54
+    },
+    {
+      "epoch": 0.01058099268949596,
+      "grad_norm": 1.882333615043608,
+      "learning_rate": 7.051282051282053e-06,
+      "loss": 1.1403,
+      "step": 55
+    },
+    {
+      "epoch": 0.010773374374759523,
+      "grad_norm": 1.8311983000074952,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 1.0848,
+      "step": 56
+    },
+    {
+      "epoch": 0.010965756060023086,
+      "grad_norm": 2.6469640380180848,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 1.0284,
+      "step": 57
+    },
+    {
+      "epoch": 0.011158137745286649,
+      "grad_norm": 3.0868968920425344,
+      "learning_rate": 7.435897435897437e-06,
+      "loss": 1.0563,
+      "step": 58
+    },
+    {
+      "epoch": 0.011350519430550212,
+      "grad_norm": 1.4879884305079736,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 1.0422,
+      "step": 59
+    },
+    {
+      "epoch": 0.011542901115813775,
+      "grad_norm": 15.628801053612683,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 1.0209,
+      "step": 60
+    },
+    {
+      "epoch": 0.011735282801077338,
+      "grad_norm": 1.5348608733748232,
+      "learning_rate": 7.820512820512822e-06,
+      "loss": 1.0449,
+      "step": 61
+    },
+    {
+      "epoch": 0.0119276644863409,
+      "grad_norm": 3.4237725844105382,
+      "learning_rate": 7.948717948717949e-06,
+      "loss": 1.0682,
+      "step": 62
+    },
+    {
+      "epoch": 0.012120046171604464,
+      "grad_norm": 1.6819303928899418,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 1.0894,
+      "step": 63
+    },
+    {
+      "epoch": 0.012312427856868027,
+      "grad_norm": 1.7467803382758034,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 1.0124,
+      "step": 64
+    },
+    {
+      "epoch": 0.012504809542131588,
+      "grad_norm": 2.4101532804642685,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.0181,
+      "step": 65
+    },
+    {
+      "epoch": 0.012697191227395151,
+      "grad_norm": 1.8902090354461247,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 1.0221,
+      "step": 66
+    },
+    {
+      "epoch": 0.012889572912658714,
+      "grad_norm": 3.441345525008632,
+      "learning_rate": 8.58974358974359e-06,
+      "loss": 1.0241,
+      "step": 67
+    },
+    {
+      "epoch": 0.013081954597922277,
+      "grad_norm": 1.661702919384361,
+      "learning_rate": 8.717948717948719e-06,
+      "loss": 1.0164,
+      "step": 68
+    },
+    {
+      "epoch": 0.01327433628318584,
+      "grad_norm": 2.5798490845240103,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 1.0636,
+      "step": 69
+    },
+    {
+      "epoch": 0.013466717968449404,
+      "grad_norm": 2.8572390097570795,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 0.9998,
+      "step": 70
+    },
+    {
+      "epoch": 0.013659099653712967,
+      "grad_norm": 2.2942443433836917,
+      "learning_rate": 9.102564102564104e-06,
+      "loss": 0.989,
+      "step": 71
+    },
+    {
+      "epoch": 0.01385148133897653,
+      "grad_norm": 15.070434173077697,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 1.0279,
+      "step": 72
+    },
+    {
+      "epoch": 0.014043863024240093,
+      "grad_norm": 1.8147288148614154,
+      "learning_rate": 9.358974358974359e-06,
+      "loss": 1.0275,
+      "step": 73
+    },
+    {
+      "epoch": 0.014236244709503656,
+      "grad_norm": 3.4175144851582484,
+      "learning_rate": 9.487179487179487e-06,
+      "loss": 1.0263,
+      "step": 74
+    },
+    {
+      "epoch": 0.014428626394767219,
+      "grad_norm": 2.7037386140405855,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 1.0058,
+      "step": 75
+    },
+    {
+      "epoch": 0.014621008080030782,
+      "grad_norm": 1.8987029558334259,
+      "learning_rate": 9.743589743589744e-06,
+      "loss": 1.0155,
+      "step": 76
+    },
+    {
+      "epoch": 0.014813389765294343,
+      "grad_norm": 1.742720018190699,
+      "learning_rate": 9.871794871794872e-06,
+      "loss": 0.9935,
+      "step": 77
+    },
+    {
+      "epoch": 0.015005771450557906,
+      "grad_norm": 17.747996411048685,
+      "learning_rate": 1e-05,
+      "loss": 1.0588,
+      "step": 78
+    },
+    {
+      "epoch": 0.01519815313582147,
+      "grad_norm": 12.94735591234143,
+      "learning_rate": 1.012820512820513e-05,
+      "loss": 1.0197,
+      "step": 79
+    },
+    {
+      "epoch": 0.015390534821085032,
+      "grad_norm": 5.709808047430847,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 1.008,
+      "step": 80
+    },
+    {
+      "epoch": 0.015582916506348595,
+      "grad_norm": 9.127252883503418,
+      "learning_rate": 1.0384615384615386e-05,
+      "loss": 1.0198,
+      "step": 81
+    },
+    {
+      "epoch": 0.015775298191612157,
+      "grad_norm": 1.8139132318151248,
+      "learning_rate": 1.0512820512820514e-05,
+      "loss": 0.9885,
+      "step": 82
+    },
+    {
+      "epoch": 0.01596767987687572,
+      "grad_norm": 9.7521081446334,
+      "learning_rate": 1.0641025641025643e-05,
+      "loss": 0.9765,
+      "step": 83
+    },
+    {
+      "epoch": 0.016160061562139283,
+      "grad_norm": 1.9214651628119763,
+      "learning_rate": 1.076923076923077e-05,
+      "loss": 1.0211,
+      "step": 84
+    },
+    {
+      "epoch": 0.016352443247402846,
+      "grad_norm": 1.4334731740898417,
+      "learning_rate": 1.0897435897435898e-05,
+      "loss": 0.2957,
+      "step": 85
+    },
+    {
+      "epoch": 0.01654482493266641,
+      "grad_norm": 2.8474852754012256,
+      "learning_rate": 1.1025641025641028e-05,
+      "loss": 1.0057,
+      "step": 86
+    },
+    {
+      "epoch": 0.016737206617929972,
+      "grad_norm": 2.2289878435528165,
+      "learning_rate": 1.1153846153846154e-05,
+      "loss": 0.9867,
+      "step": 87
+    },
+    {
+      "epoch": 0.016929588303193535,
+      "grad_norm": 2.4267117697897977,
+      "learning_rate": 1.1282051282051283e-05,
+      "loss": 1.0239,
+      "step": 88
+    },
+    {
+      "epoch": 0.017121969988457098,
+      "grad_norm": 1.9232077087805939,
+      "learning_rate": 1.1410256410256411e-05,
+      "loss": 1.0149,
+      "step": 89
+    },
+    {
+      "epoch": 0.01731435167372066,
+      "grad_norm": 4.329727123771674,
+      "learning_rate": 1.1538461538461538e-05,
+      "loss": 1.0543,
+      "step": 90
+    },
+    {
+      "epoch": 0.017506733358984224,
+      "grad_norm": 2.6186923922184744,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 1.0261,
+      "step": 91
+    },
+    {
+      "epoch": 0.017699115044247787,
+      "grad_norm": 1.8905363563726585,
+      "learning_rate": 1.1794871794871796e-05,
+      "loss": 1.0324,
+      "step": 92
+    },
+    {
+      "epoch": 0.01789149672951135,
+      "grad_norm": 1.8439010223391261,
+      "learning_rate": 1.1923076923076925e-05,
+      "loss": 1.0484,
+      "step": 93
+    },
+    {
+      "epoch": 0.018083878414774913,
+      "grad_norm": 2.0155963179755036,
+      "learning_rate": 1.2051282051282051e-05,
+      "loss": 0.993,
+      "step": 94
+    },
+    {
+      "epoch": 0.018276260100038477,
+      "grad_norm": 2.1136757514694837,
+      "learning_rate": 1.217948717948718e-05,
+      "loss": 1.0276,
+      "step": 95
+    },
+    {
+      "epoch": 0.01846864178530204,
+      "grad_norm": 2.282761304350824,
+      "learning_rate": 1.230769230769231e-05,
+      "loss": 0.9962,
+      "step": 96
+    },
+    {
+      "epoch": 0.018661023470565603,
+      "grad_norm": 1.7209178849786484,
+      "learning_rate": 1.2435897435897436e-05,
+      "loss": 0.9858,
+      "step": 97
+    },
+    {
+      "epoch": 0.018853405155829166,
+      "grad_norm": 1.99942752320992,
+      "learning_rate": 1.2564102564102565e-05,
+      "loss": 0.99,
+      "step": 98
+    },
+    {
+      "epoch": 0.01904578684109273,
+      "grad_norm": 3.911309158364588,
+      "learning_rate": 1.2692307692307693e-05,
+      "loss": 0.9829,
+      "step": 99
+    },
+    {
+      "epoch": 0.019238168526356292,
+      "grad_norm": 2.1734757339864537,
+      "learning_rate": 1.2820512820512823e-05,
+      "loss": 1.026,
+      "step": 100
+    },
+    {
+      "epoch": 0.019430550211619855,
+      "grad_norm": 1.9163493064128443,
+      "learning_rate": 1.294871794871795e-05,
+      "loss": 1.0824,
+      "step": 101
+    },
+    {
+      "epoch": 0.019622931896883418,
+      "grad_norm": 2.6879791570625913,
+      "learning_rate": 1.3076923076923078e-05,
+      "loss": 1.0171,
+      "step": 102
+    },
+    {
+      "epoch": 0.01981531358214698,
+      "grad_norm": 15.307502796670391,
+      "learning_rate": 1.3205128205128207e-05,
+      "loss": 1.0117,
+      "step": 103
+    },
+    {
+      "epoch": 0.020007695267410544,
+      "grad_norm": 2.082940756242923,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.937,
+      "step": 104
+    },
+    {
+      "epoch": 0.020200076952674107,
+      "grad_norm": 1.9313604360440337,
+      "learning_rate": 1.3461538461538463e-05,
+      "loss": 1.0406,
+      "step": 105
+    },
+    {
+      "epoch": 0.020392458637937667,
+      "grad_norm": 1.638278370486506,
+      "learning_rate": 1.3589743589743592e-05,
+      "loss": 1.0009,
+      "step": 106
+    },
+    {
+      "epoch": 0.02058484032320123,
+      "grad_norm": 17.455841689836713,
+      "learning_rate": 1.3717948717948718e-05,
+      "loss": 1.038,
+      "step": 107
+    },
+    {
+      "epoch": 0.020777222008464793,
+      "grad_norm": 1.6359536649002515,
+      "learning_rate": 1.3846153846153847e-05,
+      "loss": 0.9578,
+      "step": 108
+    },
+    {
+      "epoch": 0.020969603693728356,
+      "grad_norm": 2.9321482388080518,
+      "learning_rate": 1.3974358974358975e-05,
+      "loss": 1.0205,
+      "step": 109
+    },
+    {
+      "epoch": 0.02116198537899192,
+      "grad_norm": 2.3156095573698363,
+      "learning_rate": 1.4102564102564105e-05,
+      "loss": 0.9769,
+      "step": 110
+    },
+    {
+      "epoch": 0.021354367064255482,
+      "grad_norm": 2.249292855056779,
+      "learning_rate": 1.4230769230769232e-05,
+      "loss": 1.0397,
+      "step": 111
+    },
+    {
+      "epoch": 0.021546748749519045,
+      "grad_norm": 1.2927498074962478,
+      "learning_rate": 1.435897435897436e-05,
+      "loss": 0.9931,
+      "step": 112
+    },
+    {
+      "epoch": 0.021739130434782608,
+      "grad_norm": 1.7049509039510924,
+      "learning_rate": 1.4487179487179489e-05,
+      "loss": 0.9718,
+      "step": 113
+    },
+    {
+      "epoch": 0.02193151212004617,
+      "grad_norm": 1.665300119503636,
+      "learning_rate": 1.4615384615384615e-05,
+      "loss": 1.0379,
+      "step": 114
+    },
+    {
+      "epoch": 0.022123893805309734,
+      "grad_norm": 39.395646143583946,
+      "learning_rate": 1.4743589743589745e-05,
+      "loss": 0.9769,
+      "step": 115
+    },
+    {
+      "epoch": 0.022316275490573297,
+      "grad_norm": 1.4371387365489776,
+      "learning_rate": 1.4871794871794874e-05,
+      "loss": 1.0038,
+      "step": 116
+    },
+    {
+      "epoch": 0.02250865717583686,
+      "grad_norm": 1.8132318748478653,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.9417,
+      "step": 117
+    },
+    {
+      "epoch": 0.022701038861100423,
+      "grad_norm": 2.050456412126595,
+      "learning_rate": 1.5128205128205129e-05,
+      "loss": 0.9965,
+      "step": 118
+    },
+    {
+      "epoch": 0.022893420546363986,
+      "grad_norm": 1.9625719002470017,
+      "learning_rate": 1.5256410256410257e-05,
+      "loss": 1.0413,
+      "step": 119
+    },
+    {
+      "epoch": 0.02308580223162755,
+      "grad_norm": 3.0540794374623115,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 0.9994,
+      "step": 120
+    },
+    {
+      "epoch": 0.023278183916891113,
+      "grad_norm": 14.342154706810971,
+      "learning_rate": 1.5512820512820516e-05,
+      "loss": 1.0142,
+      "step": 121
+    },
+    {
+      "epoch": 0.023470565602154676,
+      "grad_norm": 20.74715224623993,
+      "learning_rate": 1.5641025641025644e-05,
+      "loss": 0.9867,
+      "step": 122
+    },
+    {
+      "epoch": 0.02366294728741824,
+      "grad_norm": 17.150962268802076,
+      "learning_rate": 1.576923076923077e-05,
+      "loss": 1.0377,
+      "step": 123
+    },
+    {
+      "epoch": 0.0238553289726818,
+      "grad_norm": 11.481066888369885,
+      "learning_rate": 1.5897435897435897e-05,
+      "loss": 1.0248,
+      "step": 124
+    },
+    {
+      "epoch": 0.024047710657945365,
+      "grad_norm": 5.399927167148925,
+      "learning_rate": 1.602564102564103e-05,
+      "loss": 1.0,
+      "step": 125
+    },
+    {
+      "epoch": 0.024240092343208928,
+      "grad_norm": 1.9856407037742843,
+      "learning_rate": 1.6153846153846154e-05,
+      "loss": 1.0317,
+      "step": 126
+    },
+    {
+      "epoch": 0.02443247402847249,
+      "grad_norm": 2.927463936349562,
+      "learning_rate": 1.6282051282051282e-05,
+      "loss": 0.9981,
+      "step": 127
+    },
+    {
+      "epoch": 0.024624855713736054,
+      "grad_norm": 3.2978273937536593,
+      "learning_rate": 1.641025641025641e-05,
+      "loss": 0.9755,
+      "step": 128
+    },
+    {
+      "epoch": 0.024817237398999614,
+      "grad_norm": 5.382545768269041,
+      "learning_rate": 1.653846153846154e-05,
+      "loss": 1.001,
+      "step": 129
+    },
+    {
+      "epoch": 0.025009619084263177,
+      "grad_norm": 26.33939337988933,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 1.0296,
+      "step": 130
+    },
+    {
+      "epoch": 0.02520200076952674,
+      "grad_norm": 2.908954261952205,
+      "learning_rate": 1.6794871794871796e-05,
+      "loss": 0.9329,
+      "step": 131
+    },
+    {
+      "epoch": 0.025394382454790303,
+      "grad_norm": 1.9574261649362443,
+      "learning_rate": 1.6923076923076924e-05,
+      "loss": 1.0607,
+      "step": 132
+    },
+    {
+      "epoch": 0.025586764140053866,
+      "grad_norm": 2.911584447550824,
+      "learning_rate": 1.7051282051282053e-05,
+      "loss": 0.9891,
+      "step": 133
+    },
+    {
+      "epoch": 0.02577914582531743,
+      "grad_norm": 24.375468190836568,
+      "learning_rate": 1.717948717948718e-05,
+      "loss": 1.0195,
+      "step": 134
+    },
+    {
+      "epoch": 0.025971527510580992,
+      "grad_norm": 3.0265523300169623,
+      "learning_rate": 1.730769230769231e-05,
+      "loss": 1.0612,
+      "step": 135
+    },
+    {
+      "epoch": 0.026163909195844555,
+      "grad_norm": 1.2511939422604599,
+      "learning_rate": 1.7435897435897438e-05,
+      "loss": 0.3383,
+      "step": 136
+    },
+    {
+      "epoch": 0.026356290881108118,
+      "grad_norm": 1.046598396731987,
+      "learning_rate": 1.7564102564102566e-05,
+      "loss": 0.3275,
+      "step": 137
+    },
+    {
+      "epoch": 0.02654867256637168,
+      "grad_norm": 2.5077586873251136,
+      "learning_rate": 1.7692307692307694e-05,
+      "loss": 1.008,
+      "step": 138
+    },
+    {
+      "epoch": 0.026741054251635244,
+      "grad_norm": 4.1920303504683165,
+      "learning_rate": 1.7820512820512823e-05,
+      "loss": 1.0429,
+      "step": 139
+    },
+    {
+      "epoch": 0.026933435936898807,
+      "grad_norm": 14.081849334270945,
+      "learning_rate": 1.794871794871795e-05,
+      "loss": 1.0419,
+      "step": 140
+    },
+    {
+      "epoch": 0.02712581762216237,
+      "grad_norm": 2.6543834248965066,
+      "learning_rate": 1.807692307692308e-05,
+      "loss": 0.9694,
+      "step": 141
+    },
+    {
+      "epoch": 0.027318199307425933,
+      "grad_norm": 3.798731712950747,
+      "learning_rate": 1.8205128205128208e-05,
+      "loss": 1.0074,
+      "step": 142
+    },
+    {
+      "epoch": 0.027510580992689496,
+      "grad_norm": 2.6748524302097225,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 0.9615,
+      "step": 143
+    },
+    {
+      "epoch": 0.02770296267795306,
+      "grad_norm": 4.857848768377247,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 0.9709,
+      "step": 144
+    },
+    {
+      "epoch": 0.027895344363216622,
+      "grad_norm": 3.753810808397975,
+      "learning_rate": 1.8589743589743593e-05,
+      "loss": 0.9486,
+      "step": 145
+    },
+    {
+      "epoch": 0.028087726048480186,
+      "grad_norm": 26.7333773227877,
+      "learning_rate": 1.8717948717948718e-05,
+      "loss": 0.9607,
+      "step": 146
+    },
+    {
+      "epoch": 0.02828010773374375,
+      "grad_norm": 10.372037640427239,
+      "learning_rate": 1.8846153846153846e-05,
+      "loss": 1.018,
+      "step": 147
+    },
+    {
+      "epoch": 0.02847248941900731,
+      "grad_norm": 3.4756342501130675,
+      "learning_rate": 1.8974358974358975e-05,
+      "loss": 0.9948,
+      "step": 148
+    },
+    {
+      "epoch": 0.028664871104270875,
+      "grad_norm": 16.112727646098847,
+      "learning_rate": 1.9102564102564106e-05,
+      "loss": 1.0556,
+      "step": 149
+    },
+    {
+      "epoch": 0.028857252789534438,
+      "grad_norm": 13.936844964617539,
+      "learning_rate": 1.923076923076923e-05,
+      "loss": 1.0569,
+      "step": 150
+    },
+    {
+      "epoch": 0.029049634474798,
+      "grad_norm": 24.98685746996389,
+      "learning_rate": 1.935897435897436e-05,
+      "loss": 1.0014,
+      "step": 151
+    },
+    {
+      "epoch": 0.029242016160061564,
+      "grad_norm": 4.819929405308127,
+      "learning_rate": 1.9487179487179488e-05,
+      "loss": 0.9971,
+      "step": 152
+    },
+    {
+      "epoch": 0.029434397845325123,
+      "grad_norm": 5.586009401609174,
+      "learning_rate": 1.9615384615384617e-05,
+      "loss": 1.0053,
+      "step": 153
+    },
+    {
+      "epoch": 0.029626779530588686,
+      "grad_norm": 2.4938708039774036,
+      "learning_rate": 1.9743589743589745e-05,
+      "loss": 1.0018,
+      "step": 154
+    },
+    {
+      "epoch": 0.02981916121585225,
+      "grad_norm": 2.0322333398662864,
+      "learning_rate": 1.9871794871794873e-05,
+      "loss": 1.0106,
+      "step": 155
+    },
+    {
+      "epoch": 0.030011542901115813,
+      "grad_norm": 2.2508717302127565,
+      "learning_rate": 2e-05,
+      "loss": 1.0258,
+      "step": 156
+    },
+    {
+      "epoch": 0.030203924586379376,
+      "grad_norm": 6.062399575832744,
+      "learning_rate": 1.9999998058827844e-05,
+      "loss": 1.0051,
+      "step": 157
+    },
+    {
+      "epoch": 0.03039630627164294,
+      "grad_norm": 5.570807421899651,
+      "learning_rate": 1.9999992235312136e-05,
+      "loss": 1.0102,
+      "step": 158
+    },
+    {
+      "epoch": 0.030588687956906502,
+      "grad_norm": 2.4594075968701645,
+      "learning_rate": 1.9999982529455127e-05,
+      "loss": 0.9477,
+      "step": 159
+    },
+    {
+      "epoch": 0.030781069642170065,
+      "grad_norm": 2.3250883400151947,
+      "learning_rate": 1.9999968941260596e-05,
+      "loss": 1.04,
+      "step": 160
+    },
+    {
+      "epoch": 0.030973451327433628,
+      "grad_norm": 2.9409247878371345,
+      "learning_rate": 1.9999951470733808e-05,
+      "loss": 0.941,
+      "step": 161
+    },
+    {
+      "epoch": 0.03116583301269719,
+      "grad_norm": 13.197188161282128,
+      "learning_rate": 1.9999930117881548e-05,
+      "loss": 0.9354,
+      "step": 162
+    },
+    {
+      "epoch": 0.031358214697960754,
+      "grad_norm": 5.70928588640988,
+      "learning_rate": 1.9999904882712115e-05,
+      "loss": 0.9957,
+      "step": 163
+    },
+    {
+      "epoch": 0.031550596383224314,
+      "grad_norm": 1.9303626339599533,
+      "learning_rate": 1.99998757652353e-05,
+      "loss": 1.0106,
+      "step": 164
+    },
+    {
+      "epoch": 0.03174297806848788,
+      "grad_norm": 2.045921515446836,
+      "learning_rate": 1.9999842765462403e-05,
+      "loss": 1.0298,
+      "step": 165
+    },
+    {
+      "epoch": 0.03193535975375144,
+      "grad_norm": 1.9624261432498917,
+      "learning_rate": 1.999980588340624e-05,
+      "loss": 0.9769,
+      "step": 166
+    },
+    {
+      "epoch": 0.032127741439015006,
+      "grad_norm": 3.34493090728931,
+      "learning_rate": 1.9999765119081132e-05,
+      "loss": 1.0215,
+      "step": 167
+    },
+    {
+      "epoch": 0.032320123124278566,
+      "grad_norm": 3.350648459423529,
+      "learning_rate": 1.9999720472502902e-05,
+      "loss": 1.0265,
+      "step": 168
+    },
+    {
+      "epoch": 0.03251250480954213,
+      "grad_norm": 2.2942581944219103,
+      "learning_rate": 1.9999671943688885e-05,
+      "loss": 0.3639,
+      "step": 169
+    },
+    {
+      "epoch": 0.03270488649480569,
+      "grad_norm": 8.963614605452253,
+      "learning_rate": 1.9999619532657915e-05,
+      "loss": 0.983,
+      "step": 170
+    },
+    {
+      "epoch": 0.03289726818006926,
+      "grad_norm": 8.16395752104468,
+      "learning_rate": 1.9999563239430352e-05,
+      "loss": 1.0125,
+      "step": 171
+    },
+    {
+      "epoch": 0.03308964986533282,
+      "grad_norm": 3.853703073062293,
+      "learning_rate": 1.9999503064028043e-05,
+      "loss": 1.008,
+      "step": 172
+    },
+    {
+      "epoch": 0.033282031550596385,
+      "grad_norm": 3.6431030999008227,
+      "learning_rate": 1.999943900647435e-05,
+      "loss": 0.9825,
+      "step": 173
+    },
+    {
+      "epoch": 0.033474413235859944,
+      "grad_norm": 12.437748457798014,
+      "learning_rate": 1.9999371066794146e-05,
+      "loss": 1.0827,
+      "step": 174
+    },
+    {
+      "epoch": 0.03366679492112351,
+      "grad_norm": 3.801859997696396,
+      "learning_rate": 1.9999299245013805e-05,
+      "loss": 0.9873,
+      "step": 175
+    },
+    {
+      "epoch": 0.03385917660638707,
+      "grad_norm": 1.8697249192273286,
+      "learning_rate": 1.999922354116121e-05,
+      "loss": 1.0103,
+      "step": 176
+    },
+    {
+      "epoch": 0.03405155829165064,
+      "grad_norm": 9.381256016814906,
+      "learning_rate": 1.999914395526575e-05,
+      "loss": 0.998,
+      "step": 177
+    },
+    {
+      "epoch": 0.034243939976914196,
+      "grad_norm": 2.544738052294976,
+      "learning_rate": 1.9999060487358333e-05,
+      "loss": 1.0305,
+      "step": 178
+    },
+    {
+      "epoch": 0.03443632166217776,
+      "grad_norm": 1.5444058781161245,
+      "learning_rate": 1.9998973137471352e-05,
+      "loss": 0.9965,
+      "step": 179
+    },
+    {
+      "epoch": 0.03462870334744132,
+      "grad_norm": 4.28185294436897,
+      "learning_rate": 1.9998881905638727e-05,
+      "loss": 0.9921,
+      "step": 180
+    },
+    {
+      "epoch": 0.03482108503270489,
+      "grad_norm": 4.331549123887,
+      "learning_rate": 1.9998786791895874e-05,
+      "loss": 0.9829,
+      "step": 181
+    },
+    {
+      "epoch": 0.03501346671796845,
+      "grad_norm": 3.971348292452636,
+      "learning_rate": 1.999868779627972e-05,
+      "loss": 1.0009,
+      "step": 182
+    },
+    {
+      "epoch": 0.035205848403232015,
+      "grad_norm": 3.079710076426067,
+      "learning_rate": 1.9998584918828695e-05,
+      "loss": 1.0255,
+      "step": 183
+    },
+    {
+      "epoch": 0.035398230088495575,
+      "grad_norm": 1.9183087671907009,
+      "learning_rate": 1.9998478159582747e-05,
+      "loss": 1.0543,
+      "step": 184
+    },
+    {
+      "epoch": 0.03559061177375914,
+      "grad_norm": 4.263279154514553,
+      "learning_rate": 1.999836751858332e-05,
+      "loss": 0.9822,
+      "step": 185
+    },
+    {
+      "epoch": 0.0357829934590227,
+      "grad_norm": 1.8771042852057906,
+      "learning_rate": 1.9998252995873367e-05,
+      "loss": 0.987,
+      "step": 186
+    },
+    {
+      "epoch": 0.03597537514428627,
+      "grad_norm": 2.467810664089958,
+      "learning_rate": 1.999813459149735e-05,
+      "loss": 1.0335,
+      "step": 187
+    },
+    {
+      "epoch": 0.03616775682954983,
+      "grad_norm": 1.8617108858068712,
+      "learning_rate": 1.9998012305501243e-05,
+      "loss": 0.9991,
+      "step": 188
+    },
+    {
+      "epoch": 0.03636013851481339,
+      "grad_norm": 4.482244245669265,
+      "learning_rate": 1.999788613793251e-05,
+      "loss": 1.0585,
+      "step": 189
+    },
+    {
+      "epoch": 0.03655252020007695,
+      "grad_norm": 1.9508672412135009,
+      "learning_rate": 1.999775608884015e-05,
+      "loss": 1.0047,
+      "step": 190
+    },
+    {
+      "epoch": 0.03674490188534051,
+      "grad_norm": 1.6840976503550298,
+      "learning_rate": 1.9997622158274635e-05,
+      "loss": 0.9788,
+      "step": 191
+    },
+    {
+      "epoch": 0.03693728357060408,
+      "grad_norm": 1.776747059914446,
+      "learning_rate": 1.9997484346287973e-05,
+      "loss": 0.9902,
+      "step": 192
+    },
+    {
+      "epoch": 0.03712966525586764,
+      "grad_norm": 2.3503726275637504,
+      "learning_rate": 1.9997342652933668e-05,
+      "loss": 0.9879,
+      "step": 193
+    },
+    {
+      "epoch": 0.037322046941131205,
+      "grad_norm": 2.6381796289258435,
+      "learning_rate": 1.9997197078266723e-05,
+      "loss": 0.9891,
+      "step": 194
+    },
+    {
+      "epoch": 0.037514428626394765,
+      "grad_norm": 1.7922523932877217,
+      "learning_rate": 1.999704762234366e-05,
+      "loss": 1.0086,
+      "step": 195
+    },
+    {
+      "epoch": 0.03770681031165833,
+      "grad_norm": 1.9457925333141992,
+      "learning_rate": 1.99968942852225e-05,
+      "loss": 0.9971,
+      "step": 196
+    },
+    {
+      "epoch": 0.03789919199692189,
+      "grad_norm": 1.9738787558210773,
+      "learning_rate": 1.9996737066962778e-05,
+      "loss": 0.9556,
+      "step": 197
+    },
+    {
+      "epoch": 0.03809157368218546,
+      "grad_norm": 2.011099349435828,
+      "learning_rate": 1.9996575967625525e-05,
+      "loss": 0.9944,
+      "step": 198
+    },
+    {
+      "epoch": 0.03828395536744902,
+      "grad_norm": 2.2622710677112936,
+      "learning_rate": 1.999641098727329e-05,
+      "loss": 1.0196,
+      "step": 199
+    },
+    {
+      "epoch": 0.038476337052712584,
+      "grad_norm": 2.4035361137917635,
+      "learning_rate": 1.999624212597013e-05,
+      "loss": 0.9545,
+      "step": 200
+    },
+    {
+      "epoch": 0.03866871873797614,
+      "grad_norm": 2.8793207178042923,
+      "learning_rate": 1.9996069383781587e-05,
+      "loss": 0.9918,
+      "step": 201
+    },
+    {
+      "epoch": 0.03886110042323971,
+      "grad_norm": 1.9487606346319537,
+      "learning_rate": 1.9995892760774738e-05,
+      "loss": 1.0031,
+      "step": 202
+    },
+    {
+      "epoch": 0.03905348210850327,
+      "grad_norm": 2.7921934683614524,
+      "learning_rate": 1.9995712257018153e-05,
+      "loss": 1.0309,
+      "step": 203
+    },
+    {
+      "epoch": 0.039245863793766836,
+      "grad_norm": 2.1900452319553447,
+      "learning_rate": 1.9995527872581903e-05,
+      "loss": 0.9906,
+      "step": 204
+    },
+    {
+      "epoch": 0.039438245479030395,
+      "grad_norm": 2.401266120072187,
+      "learning_rate": 1.9995339607537578e-05,
+      "loss": 1.0203,
+      "step": 205
+    },
+    {
+      "epoch": 0.03963062716429396,
+      "grad_norm": 1.7092393451023629,
+      "learning_rate": 1.9995147461958267e-05,
+      "loss": 0.9994,
+      "step": 206
+    },
+    {
+      "epoch": 0.03982300884955752,
+      "grad_norm": 1.6180836003132446,
+      "learning_rate": 1.999495143591857e-05,
+      "loss": 0.3733,
+      "step": 207
+    },
+    {
+      "epoch": 0.04001539053482109,
+      "grad_norm": 2.160465724928752,
+      "learning_rate": 1.999475152949459e-05,
+      "loss": 0.9922,
+      "step": 208
+    },
+    {
+      "epoch": 0.04020777222008465,
+      "grad_norm": 2.0655401663510027,
+      "learning_rate": 1.9994547742763935e-05,
+      "loss": 0.9693,
+      "step": 209
+    },
+    {
+      "epoch": 0.040400153905348214,
+      "grad_norm": 1.6677465033142602,
+      "learning_rate": 1.9994340075805724e-05,
+      "loss": 1.0378,
+      "step": 210
+    },
+    {
+      "epoch": 0.040592535590611774,
+      "grad_norm": 1.6920698266877157,
+      "learning_rate": 1.9994128528700583e-05,
+      "loss": 1.0532,
+      "step": 211
+    },
+    {
+      "epoch": 0.04078491727587533,
+      "grad_norm": 1.9087677109941978,
+      "learning_rate": 1.9993913101530635e-05,
+      "loss": 0.9768,
+      "step": 212
+    },
+    {
+      "epoch": 0.0409772989611389,
+      "grad_norm": 3.6773714056320026,
+      "learning_rate": 1.9993693794379525e-05,
+      "loss": 0.9251,
+      "step": 213
+    },
+    {
+      "epoch": 0.04116968064640246,
+      "grad_norm": 1.6851067347316617,
+      "learning_rate": 1.9993470607332387e-05,
+      "loss": 1.0032,
+      "step": 214
+    },
+    {
+      "epoch": 0.041362062331666026,
+      "grad_norm": 2.011324582840973,
+      "learning_rate": 1.999324354047588e-05,
+      "loss": 1.0243,
+      "step": 215
+    },
+    {
+      "epoch": 0.041554444016929586,
+      "grad_norm": 1.5347597969076157,
+      "learning_rate": 1.9993012593898146e-05,
+      "loss": 0.9732,
+      "step": 216
+    },
+    {
+      "epoch": 0.04174682570219315,
+      "grad_norm": 1.8010279457261764,
+      "learning_rate": 1.9992777767688857e-05,
+      "loss": 0.9704,
+      "step": 217
+    },
+    {
+      "epoch": 0.04193920738745671,
+      "grad_norm": 1.6582074082231524,
+      "learning_rate": 1.9992539061939175e-05,
+      "loss": 1.0261,
+      "step": 218
+    },
+    {
+      "epoch": 0.04213158907272028,
+      "grad_norm": 2.6922995350897145,
+      "learning_rate": 1.999229647674178e-05,
+      "loss": 0.9933,
+      "step": 219
+    },
+    {
+      "epoch": 0.04232397075798384,
+      "grad_norm": 5.021267730925554,
+      "learning_rate": 1.9992050012190845e-05,
+      "loss": 0.979,
+      "step": 220
+    },
+    {
+      "epoch": 0.042516352443247404,
+      "grad_norm": 0.9150999872794174,
+      "learning_rate": 1.9991799668382058e-05,
+      "loss": 0.3415,
+      "step": 221
+    },
+    {
+      "epoch": 0.042708734128510964,
+      "grad_norm": 1.9639282527853914,
+      "learning_rate": 1.9991545445412614e-05,
+      "loss": 0.9748,
+      "step": 222
+    },
+    {
+      "epoch": 0.04290111581377453,
+      "grad_norm": 3.9929144032002286,
+      "learning_rate": 1.9991287343381208e-05,
+      "loss": 1.0892,
+      "step": 223
+    },
+    {
+      "epoch": 0.04309349749903809,
+      "grad_norm": 3.2327380860885735,
+      "learning_rate": 1.9991025362388044e-05,
+      "loss": 0.9952,
+      "step": 224
+    },
+    {
+      "epoch": 0.04328587918430166,
+      "grad_norm": 2.0799345706353787,
+      "learning_rate": 1.9990759502534835e-05,
+      "loss": 0.9661,
+      "step": 225
+    },
+    {
+      "epoch": 0.043478260869565216,
+      "grad_norm": 2.2348440788760873,
+      "learning_rate": 1.9990489763924796e-05,
+      "loss": 0.9733,
+      "step": 226
+    },
+    {
+      "epoch": 0.04367064255482878,
+      "grad_norm": 2.322157280980084,
+      "learning_rate": 1.9990216146662648e-05,
+      "loss": 0.985,
+      "step": 227
+    },
+    {
+      "epoch": 0.04386302424009234,
+      "grad_norm": 1.625634805825751,
+      "learning_rate": 1.9989938650854618e-05,
+      "loss": 1.0254,
+      "step": 228
+    },
+    {
+      "epoch": 0.04405540592535591,
+      "grad_norm": 1.9282926460016017,
+      "learning_rate": 1.998965727660844e-05,
+      "loss": 0.9957,
+      "step": 229
+    },
+    {
+      "epoch": 0.04424778761061947,
+      "grad_norm": 1.7778516310689925,
+      "learning_rate": 1.9989372024033352e-05,
+      "loss": 1.0078,
+      "step": 230
+    },
+    {
+      "epoch": 0.044440169295883035,
+      "grad_norm": 2.0941966016535494,
+      "learning_rate": 1.99890828932401e-05,
+      "loss": 1.0255,
+      "step": 231
+    },
+    {
+      "epoch": 0.044632550981146595,
+      "grad_norm": 1.7172684617684904,
+      "learning_rate": 1.9988789884340938e-05,
+      "loss": 0.9381,
+      "step": 232
+    },
+    {
+      "epoch": 0.04482493266641016,
+      "grad_norm": 2.5302100272382324,
+      "learning_rate": 1.9988492997449615e-05,
+      "loss": 0.9632,
+      "step": 233
+    },
+    {
+      "epoch": 0.04501731435167372,
+      "grad_norm": 2.782867430082819,
+      "learning_rate": 1.9988192232681398e-05,
+      "loss": 0.9473,
+      "step": 234
+    },
+    {
+      "epoch": 0.04520969603693728,
+      "grad_norm": 3.729481843682031,
+      "learning_rate": 1.9987887590153055e-05,
+      "loss": 1.0093,
+      "step": 235
+    },
+    {
+      "epoch": 0.04540207772220085,
+      "grad_norm": 2.7175787521895054,
+      "learning_rate": 1.9987579069982856e-05,
+      "loss": 0.9979,
+      "step": 236
+    },
+    {
+      "epoch": 0.045594459407464406,
+      "grad_norm": 1.9970701988598305,
+      "learning_rate": 1.9987266672290577e-05,
+      "loss": 0.9685,
+      "step": 237
+    },
+    {
+      "epoch": 0.04578684109272797,
+      "grad_norm": 2.4521037449389667,
+      "learning_rate": 1.9986950397197503e-05,
+      "loss": 1.0204,
+      "step": 238
+    },
+    {
+      "epoch": 0.04597922277799153,
+      "grad_norm": 4.063683537372398,
+      "learning_rate": 1.9986630244826425e-05,
+      "loss": 0.9712,
+      "step": 239
+    },
+    {
+      "epoch": 0.0461716044632551,
+      "grad_norm": 2.0164014235584644,
+      "learning_rate": 1.998630621530164e-05,
+      "loss": 0.9903,
+      "step": 240
+    },
+    {
+      "epoch": 0.04636398614851866,
+      "grad_norm": 1.7125281420857839,
+      "learning_rate": 1.998597830874894e-05,
+      "loss": 0.9646,
+      "step": 241
+    },
+    {
+      "epoch": 0.046556367833782225,
+      "grad_norm": 1.9654758039824962,
+      "learning_rate": 1.9985646525295634e-05,
+      "loss": 1.0289,
+      "step": 242
+    },
+    {
+      "epoch": 0.046748749519045785,
+      "grad_norm": 1.5416492823124428,
+      "learning_rate": 1.998531086507053e-05,
+      "loss": 0.9789,
+      "step": 243
+    },
+    {
+      "epoch": 0.04694113120430935,
+      "grad_norm": 1.9984785227519435,
+      "learning_rate": 1.9984971328203945e-05,
+      "loss": 1.0106,
+      "step": 244
+    },
+    {
+      "epoch": 0.04713351288957291,
+      "grad_norm": 1.7855433361696575,
+      "learning_rate": 1.9984627914827698e-05,
+      "loss": 0.9998,
+      "step": 245
+    },
+    {
+      "epoch": 0.04732589457483648,
+      "grad_norm": 3.040720214031602,
+      "learning_rate": 1.9984280625075115e-05,
+      "loss": 0.9977,
+      "step": 246
+    },
+    {
+      "epoch": 0.04751827626010004,
+      "grad_norm": 1.778843619236305,
+      "learning_rate": 1.9983929459081022e-05,
+      "loss": 0.9465,
+      "step": 247
+    },
+    {
+      "epoch": 0.0477106579453636,
+      "grad_norm": 0.775528832930659,
+      "learning_rate": 1.998357441698176e-05,
+      "loss": 0.3366,
+      "step": 248
+    },
+    {
+      "epoch": 0.04790303963062716,
+      "grad_norm": 4.027113971423746,
+      "learning_rate": 1.998321549891516e-05,
+      "loss": 0.988,
+      "step": 249
+    },
+    {
+      "epoch": 0.04809542131589073,
+      "grad_norm": 1.4962262005500082,
+      "learning_rate": 1.9982852705020572e-05,
+      "loss": 1.0064,
+      "step": 250
+    },
+    {
+      "epoch": 0.04828780300115429,
+      "grad_norm": 1.6749044587677224,
+      "learning_rate": 1.9982486035438848e-05,
+      "loss": 1.0028,
+      "step": 251
+    },
+    {
+      "epoch": 0.048480184686417856,
+      "grad_norm": 1.3574072089977003,
+      "learning_rate": 1.9982115490312334e-05,
+      "loss": 0.9368,
+      "step": 252
+    },
+    {
+      "epoch": 0.048672566371681415,
+      "grad_norm": 0.7322708745113847,
+      "learning_rate": 1.9981741069784894e-05,
+      "loss": 0.3317,
+      "step": 253
+    },
+    {
+      "epoch": 0.04886494805694498,
+      "grad_norm": 4.614949514895051,
+      "learning_rate": 1.9981362774001886e-05,
+      "loss": 1.0303,
+      "step": 254
+    },
+    {
+      "epoch": 0.04905732974220854,
+      "grad_norm": 1.5361675602149718,
+      "learning_rate": 1.9980980603110185e-05,
+      "loss": 0.9967,
+      "step": 255
+    },
+    {
+      "epoch": 0.04924971142747211,
+      "grad_norm": 0.6605019419571164,
+      "learning_rate": 1.9980594557258158e-05,
+      "loss": 0.3393,
+      "step": 256
+    },
+    {
+      "epoch": 0.04944209311273567,
+      "grad_norm": 0.65995524234701,
+      "learning_rate": 1.9980204636595682e-05,
+      "loss": 0.339,
+      "step": 257
+    },
+    {
+      "epoch": 0.04963447479799923,
+      "grad_norm": 0.6001034129719862,
+      "learning_rate": 1.9979810841274135e-05,
+      "loss": 0.3078,
+      "step": 258
+    },
+    {
+      "epoch": 0.049826856483262794,
+      "grad_norm": 1.6562121838550425,
+      "learning_rate": 1.9979413171446403e-05,
+      "loss": 0.9435,
+      "step": 259
+    },
+    {
+      "epoch": 0.05001923816852635,
+      "grad_norm": 2.148320197947554,
+      "learning_rate": 1.9979011627266884e-05,
+      "loss": 0.9863,
+      "step": 260
+    },
+    {
+      "epoch": 0.05021161985378992,
+      "grad_norm": 1.7055535601330953,
+      "learning_rate": 1.997860620889146e-05,
+      "loss": 0.9929,
+      "step": 261
+    },
+    {
+      "epoch": 0.05040400153905348,
+      "grad_norm": 1.6910479692110731,
+      "learning_rate": 1.997819691647753e-05,
+      "loss": 0.9741,
+      "step": 262
+    },
+    {
+      "epoch": 0.050596383224317046,
+      "grad_norm": 1.6025517782252436,
+      "learning_rate": 1.9977783750184e-05,
+      "loss": 0.9803,
+      "step": 263
+    },
+    {
+      "epoch": 0.050788764909580605,
+      "grad_norm": 1.8036545137620363,
+      "learning_rate": 1.9977366710171274e-05,
+      "loss": 0.9197,
+      "step": 264
+    },
+    {
+      "epoch": 0.05098114659484417,
+      "grad_norm": 2.695640319064733,
+      "learning_rate": 1.9976945796601258e-05,
+      "loss": 1.0078,
+      "step": 265
+    },
+    {
+      "epoch": 0.05117352828010773,
+      "grad_norm": 1.5987358596214722,
+      "learning_rate": 1.9976521009637366e-05,
+      "loss": 0.9375,
+      "step": 266
+    },
+    {
+      "epoch": 0.0513659099653713,
+      "grad_norm": 1.5239812655135472,
+      "learning_rate": 1.997609234944452e-05,
+      "loss": 0.9937,
+      "step": 267
+    },
+    {
+      "epoch": 0.05155829165063486,
+      "grad_norm": 4.6037447683462664,
+      "learning_rate": 1.9975659816189137e-05,
+      "loss": 0.9302,
+      "step": 268
+    },
+    {
+      "epoch": 0.051750673335898424,
+      "grad_norm": 3.0503808182424934,
+      "learning_rate": 1.997522341003914e-05,
+      "loss": 1.0388,
+      "step": 269
+    },
+    {
+      "epoch": 0.051943055021161984,
+      "grad_norm": 4.257926413002911,
+      "learning_rate": 1.9974783131163957e-05,
+      "loss": 0.9971,
+      "step": 270
+    },
+    {
+      "epoch": 0.05213543670642555,
+      "grad_norm": 2.407807686883654,
+      "learning_rate": 1.9974338979734523e-05,
+      "loss": 0.9379,
+      "step": 271
+    },
+    {
+      "epoch": 0.05232781839168911,
+      "grad_norm": 2.303629669358429,
+      "learning_rate": 1.997389095592327e-05,
+      "loss": 0.9444,
+      "step": 272
+    },
+    {
+      "epoch": 0.052520200076952676,
+      "grad_norm": 1.4127821583740012,
+      "learning_rate": 1.9973439059904133e-05,
+      "loss": 1.0135,
+      "step": 273
+    },
+    {
+      "epoch": 0.052712581762216236,
+      "grad_norm": 1.6405443600231049,
+      "learning_rate": 1.9972983291852565e-05,
+      "loss": 0.9897,
+      "step": 274
+    },
+    {
+      "epoch": 0.0529049634474798,
+      "grad_norm": 1.9283838913201787,
+      "learning_rate": 1.9972523651945496e-05,
+      "loss": 0.9667,
+      "step": 275
+    },
+    {
+      "epoch": 0.05309734513274336,
+      "grad_norm": 1.4722333332425688,
+      "learning_rate": 1.9972060140361384e-05,
+      "loss": 1.0118,
+      "step": 276
+    },
+    {
+      "epoch": 0.05328972681800693,
+      "grad_norm": 1.553459848014345,
+      "learning_rate": 1.997159275728018e-05,
+      "loss": 0.982,
+      "step": 277
+    },
+    {
+      "epoch": 0.05348210850327049,
+      "grad_norm": 1.8846297321649041,
+      "learning_rate": 1.9971121502883332e-05,
+      "loss": 0.979,
+      "step": 278
+    },
+    {
+      "epoch": 0.053674490188534055,
+      "grad_norm": 11.860610666955402,
+      "learning_rate": 1.9970646377353802e-05,
+      "loss": 0.9931,
+      "step": 279
+    },
+    {
+      "epoch": 0.053866871873797614,
+      "grad_norm": 2.0852461061774075,
+      "learning_rate": 1.997016738087605e-05,
+      "loss": 1.0057,
+      "step": 280
+    },
+    {
+      "epoch": 0.054059253559061174,
+      "grad_norm": 1.3941939199077902,
+      "learning_rate": 1.9969684513636035e-05,
+      "loss": 0.9069,
+      "step": 281
+    },
+    {
+      "epoch": 0.05425163524432474,
+      "grad_norm": 2.555163749945864,
+      "learning_rate": 1.9969197775821227e-05,
+      "loss": 0.997,
+      "step": 282
+    },
+    {
+      "epoch": 0.0544440169295883,
+      "grad_norm": 1.5992366398734477,
+      "learning_rate": 1.9968707167620593e-05,
+      "loss": 0.969,
+      "step": 283
+    },
+    {
+      "epoch": 0.05463639861485187,
+      "grad_norm": 1.7251447418075532,
+      "learning_rate": 1.9968212689224603e-05,
+      "loss": 0.9694,
+      "step": 284
+    },
+    {
+      "epoch": 0.054828780300115426,
+      "grad_norm": 1.513794328350471,
+      "learning_rate": 1.996771434082523e-05,
+      "loss": 0.944,
+      "step": 285
+    },
+    {
+      "epoch": 0.05502116198537899,
+      "grad_norm": 1.4138951840295282,
+      "learning_rate": 1.9967212122615958e-05,
+      "loss": 0.9568,
+      "step": 286
+    },
+    {
+      "epoch": 0.05521354367064255,
+      "grad_norm": 0.8175578003625622,
+      "learning_rate": 1.9966706034791752e-05,
+      "loss": 0.3133,
+      "step": 287
+    },
+    {
+      "epoch": 0.05540592535590612,
+      "grad_norm": 1.990426516385094,
+      "learning_rate": 1.9966196077549106e-05,
+      "loss": 0.9924,
+      "step": 288
+    },
+    {
+      "epoch": 0.05559830704116968,
+      "grad_norm": 0.6582001560792683,
+      "learning_rate": 1.996568225108599e-05,
+      "loss": 0.3131,
+      "step": 289
+    },
+    {
+      "epoch": 0.055790688726433245,
+      "grad_norm": 2.0008644713870796,
+      "learning_rate": 1.99651645556019e-05,
+      "loss": 0.9578,
+      "step": 290
+    },
+    {
+      "epoch": 0.055983070411696804,
+      "grad_norm": 1.7980932362910047,
+      "learning_rate": 1.9964642991297817e-05,
+      "loss": 1.0309,
+      "step": 291
+    },
+    {
+      "epoch": 0.05617545209696037,
+      "grad_norm": 2.0864619511615046,
+      "learning_rate": 1.996411755837623e-05,
+      "loss": 0.9728,
+      "step": 292
+    },
+    {
+      "epoch": 0.05636783378222393,
+      "grad_norm": 1.6834938753246855,
+      "learning_rate": 1.9963588257041137e-05,
+      "loss": 1.0378,
+      "step": 293
+    },
+    {
+      "epoch": 0.0565602154674875,
+      "grad_norm": 1.5408774983618967,
+      "learning_rate": 1.996305508749802e-05,
+      "loss": 1.0123,
+      "step": 294
+    },
+    {
+      "epoch": 0.05675259715275106,
+      "grad_norm": 14.27538190311774,
+      "learning_rate": 1.9962518049953887e-05,
+      "loss": 0.9674,
+      "step": 295
+    },
+    {
+      "epoch": 0.05694497883801462,
+      "grad_norm": 2.069178460186291,
+      "learning_rate": 1.9961977144617225e-05,
+      "loss": 0.9408,
+      "step": 296
+    },
+    {
+      "epoch": 0.05713736052327818,
+      "grad_norm": 2.327012273353525,
+      "learning_rate": 1.996143237169803e-05,
+      "loss": 0.9825,
+      "step": 297
+    },
+    {
+      "epoch": 0.05732974220854175,
+      "grad_norm": 1.5353781768255277,
+      "learning_rate": 1.996088373140781e-05,
+      "loss": 0.9481,
+      "step": 298
+    },
+    {
+      "epoch": 0.05752212389380531,
+      "grad_norm": 2.844138585041771,
+      "learning_rate": 1.9960331223959564e-05,
+      "loss": 0.9897,
+      "step": 299
+    },
+    {
+      "epoch": 0.057714505579068875,
+      "grad_norm": 1.2746164274426561,
+      "learning_rate": 1.995977484956779e-05,
+      "loss": 0.9212,
+      "step": 300
+    },
+    {
+      "epoch": 0.057906887264332435,
+      "grad_norm": 1.7169363274227265,
+      "learning_rate": 1.9959214608448495e-05,
+      "loss": 0.9424,
+      "step": 301
+    },
+    {
+      "epoch": 0.058099268949596,
+      "grad_norm": 4.692321079021565,
+      "learning_rate": 1.9958650500819183e-05,
+      "loss": 0.9796,
+      "step": 302
+    },
+    {
+      "epoch": 0.05829165063485956,
+      "grad_norm": 2.081407693886344,
+      "learning_rate": 1.995808252689886e-05,
+      "loss": 1.027,
+      "step": 303
+    },
+    {
+      "epoch": 0.05848403232012313,
+      "grad_norm": 1.7647087222263194,
+      "learning_rate": 1.9957510686908034e-05,
+      "loss": 0.9837,
+      "step": 304
+    },
+    {
+      "epoch": 0.05867641400538669,
+      "grad_norm": 1.96305065139417,
+      "learning_rate": 1.9956934981068713e-05,
+      "loss": 0.9869,
+      "step": 305
+    },
+    {
+      "epoch": 0.05886879569065025,
+      "grad_norm": 1.7340753771880726,
+      "learning_rate": 1.9956355409604402e-05,
+      "loss": 0.9779,
+      "step": 306
+    },
+    {
+      "epoch": 0.05906117737591381,
+      "grad_norm": 1.6464355868552554,
+      "learning_rate": 1.9955771972740118e-05,
+      "loss": 0.9742,
+      "step": 307
+    },
+    {
+      "epoch": 0.05925355906117737,
+      "grad_norm": 2.4501012476373285,
+      "learning_rate": 1.9955184670702363e-05,
+      "loss": 0.9448,
+      "step": 308
+    },
+    {
+      "epoch": 0.05944594074644094,
+      "grad_norm": 2.1917429933427774,
+      "learning_rate": 1.995459350371915e-05,
+      "loss": 0.9883,
+      "step": 309
+    },
+    {
+      "epoch": 0.0596383224317045,
+      "grad_norm": 1.7875270850895397,
+      "learning_rate": 1.9953998472019996e-05,
+      "loss": 1.0542,
+      "step": 310
+    },
+    {
+      "epoch": 0.059830704116968066,
+      "grad_norm": 3.737643505062909,
+      "learning_rate": 1.995339957583591e-05,
+      "loss": 1.031,
+      "step": 311
+    },
+    {
+      "epoch": 0.060023085802231625,
+      "grad_norm": 1.0698252315589796,
+      "learning_rate": 1.9952796815399403e-05,
+      "loss": 0.3365,
+      "step": 312
+    },
+    {
+      "epoch": 0.06021546748749519,
+      "grad_norm": 1.5425608415873402,
+      "learning_rate": 1.9952190190944484e-05,
+      "loss": 0.9384,
+      "step": 313
+    },
+    {
+      "epoch": 0.06040784917275875,
+      "grad_norm": 2.6424292547881336,
+      "learning_rate": 1.9951579702706668e-05,
+      "loss": 0.9699,
+      "step": 314
+    },
+    {
+      "epoch": 0.06060023085802232,
+      "grad_norm": 2.077749829204852,
+      "learning_rate": 1.9950965350922975e-05,
+      "loss": 0.9749,
+      "step": 315
+    },
+    {
+      "epoch": 0.06079261254328588,
+      "grad_norm": 1.8289497351348887,
+      "learning_rate": 1.9950347135831907e-05,
+      "loss": 0.9656,
+      "step": 316
+    },
+    {
+      "epoch": 0.060984994228549444,
+      "grad_norm": 2.5929293338127315,
+      "learning_rate": 1.994972505767348e-05,
+      "loss": 0.8682,
+      "step": 317
+    },
+    {
+      "epoch": 0.061177375913813004,
+      "grad_norm": 1.4605392865840916,
+      "learning_rate": 1.994909911668921e-05,
+      "loss": 0.9527,
+      "step": 318
+    },
+    {
+      "epoch": 0.06136975759907657,
+      "grad_norm": 1.6597390076110299,
+      "learning_rate": 1.99484693131221e-05,
+      "loss": 1.0381,
+      "step": 319
+    },
+    {
+      "epoch": 0.06156213928434013,
+      "grad_norm": 2.3934758319903935,
+      "learning_rate": 1.994783564721667e-05,
+      "loss": 1.0145,
+      "step": 320
+    },
+    {
+      "epoch": 0.061754520969603696,
+      "grad_norm": 1.6617033711473772,
+      "learning_rate": 1.9947198119218924e-05,
+      "loss": 0.9946,
+      "step": 321
+    },
+    {
+      "epoch": 0.061946902654867256,
+      "grad_norm": 1.7139166291598449,
+      "learning_rate": 1.994655672937638e-05,
+      "loss": 0.9584,
+      "step": 322
+    },
+    {
+      "epoch": 0.06213928434013082,
+      "grad_norm": 1.5056565412136098,
+      "learning_rate": 1.9945911477938044e-05,
+      "loss": 0.9929,
+      "step": 323
+    },
+    {
+      "epoch": 0.06233166602539438,
+      "grad_norm": 0.6995930709576063,
+      "learning_rate": 1.994526236515442e-05,
+      "loss": 0.3326,
+      "step": 324
+    },
+    {
+      "epoch": 0.06252404771065795,
+      "grad_norm": 0.6669359690589866,
+      "learning_rate": 1.994460939127753e-05,
+      "loss": 0.3441,
+      "step": 325
+    },
+    {
+      "epoch": 0.06271642939592151,
+      "grad_norm": 1.5996547779157864,
+      "learning_rate": 1.9943952556560863e-05,
+      "loss": 0.926,
+      "step": 326
+    },
+    {
+      "epoch": 0.06290881108118507,
+      "grad_norm": 1.5559493559186053,
+      "learning_rate": 1.9943291861259433e-05,
+      "loss": 0.8865,
+      "step": 327
+    },
+    {
+      "epoch": 0.06310119276644863,
+      "grad_norm": 3.022966407305599,
+      "learning_rate": 1.9942627305629747e-05,
+      "loss": 1.011,
+      "step": 328
+    },
+    {
+      "epoch": 0.0632935744517122,
+      "grad_norm": 1.6521214019732477,
+      "learning_rate": 1.9941958889929808e-05,
+      "loss": 0.9795,
+      "step": 329
+    },
+    {
+      "epoch": 0.06348595613697576,
+      "grad_norm": 1.6398590851769006,
+      "learning_rate": 1.9941286614419113e-05,
+      "loss": 1.0072,
+      "step": 330
+    },
+    {
+      "epoch": 0.06367833782223932,
+      "grad_norm": 1.6831967655493982,
+      "learning_rate": 1.994061047935867e-05,
+      "loss": 0.951,
+      "step": 331
+    },
+    {
+      "epoch": 0.06387071950750288,
+      "grad_norm": 2.0561755610010604,
+      "learning_rate": 1.9939930485010968e-05,
+      "loss": 1.0101,
+      "step": 332
+    },
+    {
+      "epoch": 0.06406310119276645,
+      "grad_norm": 1.533524913632678,
+      "learning_rate": 1.9939246631640014e-05,
+      "loss": 0.9618,
+      "step": 333
+    },
+    {
+      "epoch": 0.06425548287803001,
+      "grad_norm": 2.439106235238485,
+      "learning_rate": 1.99385589195113e-05,
+      "loss": 0.9623,
+      "step": 334
+    },
+    {
+      "epoch": 0.06444786456329357,
+      "grad_norm": 1.6901873062169606,
+      "learning_rate": 1.9937867348891815e-05,
+      "loss": 0.9306,
+      "step": 335
+    },
+    {
+      "epoch": 0.06464024624855713,
+      "grad_norm": 1.639326166211691,
+      "learning_rate": 1.9937171920050057e-05,
+      "loss": 0.9683,
+      "step": 336
+    },
+    {
+      "epoch": 0.0648326279338207,
+      "grad_norm": 0.6985872071018121,
+      "learning_rate": 1.9936472633256012e-05,
+      "loss": 0.3109,
+      "step": 337
+    },
+    {
+      "epoch": 0.06502500961908426,
+      "grad_norm": 1.8978047636926219,
+      "learning_rate": 1.9935769488781167e-05,
+      "loss": 0.9531,
+      "step": 338
+    },
+    {
+      "epoch": 0.06521739130434782,
+      "grad_norm": 1.7088475228191373,
+      "learning_rate": 1.993506248689851e-05,
+      "loss": 0.9628,
+      "step": 339
+    },
+    {
+      "epoch": 0.06540977298961138,
+      "grad_norm": 3.888127445475316,
+      "learning_rate": 1.993435162788252e-05,
+      "loss": 1.0309,
+      "step": 340
+    },
+    {
+      "epoch": 0.06560215467487496,
+      "grad_norm": 3.6127853020677696,
+      "learning_rate": 1.993363691200918e-05,
+      "loss": 1.0155,
+      "step": 341
+    },
+    {
+      "epoch": 0.06579453636013852,
+      "grad_norm": 6.277231835486991,
+      "learning_rate": 1.9932918339555965e-05,
+      "loss": 0.9436,
+      "step": 342
+    },
+    {
+      "epoch": 0.06598691804540208,
+      "grad_norm": 1.6572407530733466,
+      "learning_rate": 1.9932195910801848e-05,
+      "loss": 1.0429,
+      "step": 343
+    },
+    {
+      "epoch": 0.06617929973066564,
+      "grad_norm": 1.77559823786002,
+      "learning_rate": 1.9931469626027305e-05,
+      "loss": 0.9897,
+      "step": 344
+    },
+    {
+      "epoch": 0.06637168141592921,
+      "grad_norm": 1.6651353396760569,
+      "learning_rate": 1.9930739485514304e-05,
+      "loss": 1.0064,
+      "step": 345
+    },
+    {
+      "epoch": 0.06656406310119277,
+      "grad_norm": 11.901897524050584,
+      "learning_rate": 1.9930005489546308e-05,
+      "loss": 0.9743,
+      "step": 346
+    },
+    {
+      "epoch": 0.06675644478645633,
+      "grad_norm": 7.2701344461502115,
+      "learning_rate": 1.9929267638408277e-05,
+      "loss": 0.9609,
+      "step": 347
+    },
+    {
+      "epoch": 0.06694882647171989,
+      "grad_norm": 0.6254717203261694,
+      "learning_rate": 1.9928525932386678e-05,
+      "loss": 0.3432,
+      "step": 348
+    },
+    {
+      "epoch": 0.06714120815698346,
+      "grad_norm": 0.5763079324072748,
+      "learning_rate": 1.9927780371769463e-05,
+      "loss": 0.3202,
+      "step": 349
+    },
+    {
+      "epoch": 0.06733358984224702,
+      "grad_norm": 3.175573100264352,
+      "learning_rate": 1.9927030956846083e-05,
+      "loss": 0.9497,
+      "step": 350
+    },
+    {
+      "epoch": 0.06752597152751058,
+      "grad_norm": 1.6742872575263503,
+      "learning_rate": 1.992627768790749e-05,
+      "loss": 0.9271,
+      "step": 351
+    },
+    {
+      "epoch": 0.06771835321277414,
+      "grad_norm": 1.988081573857575,
+      "learning_rate": 1.9925520565246125e-05,
+      "loss": 0.9915,
+      "step": 352
+    },
+    {
+      "epoch": 0.0679107348980377,
+      "grad_norm": 2.465501883737824,
+      "learning_rate": 1.9924759589155932e-05,
+      "loss": 0.9984,
+      "step": 353
+    },
+    {
+      "epoch": 0.06810311658330127,
+      "grad_norm": 5.339505405429253,
+      "learning_rate": 1.9923994759932344e-05,
+      "loss": 0.9557,
+      "step": 354
+    },
+    {
+      "epoch": 0.06829549826856483,
+      "grad_norm": 3.832039580182098,
+      "learning_rate": 1.9923226077872296e-05,
+      "loss": 1.0098,
+      "step": 355
+    },
+    {
+      "epoch": 0.06848787995382839,
+      "grad_norm": 5.972332569490575,
+      "learning_rate": 1.9922453543274223e-05,
+      "loss": 0.9621,
+      "step": 356
+    },
+    {
+      "epoch": 0.06868026163909195,
+      "grad_norm": 3.6301536948482656,
+      "learning_rate": 1.9921677156438044e-05,
+      "loss": 1.0042,
+      "step": 357
+    },
+    {
+      "epoch": 0.06887264332435553,
+      "grad_norm": 1.8332876671732574,
+      "learning_rate": 1.9920896917665178e-05,
+      "loss": 1.0062,
+      "step": 358
+    },
+    {
+      "epoch": 0.06906502500961909,
+      "grad_norm": 16.9979906629001,
+      "learning_rate": 1.992011282725854e-05,
+      "loss": 0.9681,
+      "step": 359
+    },
+    {
+      "epoch": 0.06925740669488265,
+      "grad_norm": 3.4352254399632067,
+      "learning_rate": 1.9919324885522548e-05,
+      "loss": 1.0245,
+      "step": 360
+    },
+    {
+      "epoch": 0.0694497883801462,
+      "grad_norm": 1.5890617241118368,
+      "learning_rate": 1.99185330927631e-05,
+      "loss": 0.9519,
+      "step": 361
+    },
+    {
+      "epoch": 0.06964217006540978,
+      "grad_norm": 1.8428948038702744,
+      "learning_rate": 1.99177374492876e-05,
+      "loss": 0.946,
+      "step": 362
+    },
+    {
+      "epoch": 0.06983455175067334,
+      "grad_norm": 2.2062052010826654,
+      "learning_rate": 1.991693795540494e-05,
+      "loss": 1.0383,
+      "step": 363
+    },
+    {
+      "epoch": 0.0700269334359369,
+      "grad_norm": 0.6171918232559819,
+      "learning_rate": 1.9916134611425522e-05,
+      "loss": 0.3092,
+      "step": 364
+    },
+    {
+      "epoch": 0.07021931512120046,
+      "grad_norm": 1.8281451035641134,
+      "learning_rate": 1.9915327417661226e-05,
+      "loss": 0.9348,
+      "step": 365
+    },
+    {
+      "epoch": 0.07041169680646403,
+      "grad_norm": 4.353395102486739,
+      "learning_rate": 1.991451637442543e-05,
+      "loss": 0.9878,
+      "step": 366
+    },
+    {
+      "epoch": 0.07060407849172759,
+      "grad_norm": 0.6849432086758035,
+      "learning_rate": 1.9913701482033008e-05,
+      "loss": 0.3105,
+      "step": 367
+    },
+    {
+      "epoch": 0.07079646017699115,
+      "grad_norm": 3.241697043581032,
+      "learning_rate": 1.9912882740800336e-05,
+      "loss": 0.9112,
+      "step": 368
+    },
+    {
+      "epoch": 0.07098884186225471,
+      "grad_norm": 2.1816688993296554,
+      "learning_rate": 1.9912060151045273e-05,
+      "loss": 0.9912,
+      "step": 369
+    },
+    {
+      "epoch": 0.07118122354751828,
+      "grad_norm": 2.0412976099562736,
+      "learning_rate": 1.9911233713087172e-05,
+      "loss": 0.9725,
+      "step": 370
+    },
+    {
+      "epoch": 0.07137360523278184,
+      "grad_norm": 2.3381122912733248,
+      "learning_rate": 1.9910403427246895e-05,
+      "loss": 0.891,
+      "step": 371
+    },
+    {
+      "epoch": 0.0715659869180454,
+      "grad_norm": 0.6328879463830739,
+      "learning_rate": 1.990956929384678e-05,
+      "loss": 0.3272,
+      "step": 372
+    },
+    {
+      "epoch": 0.07175836860330896,
+      "grad_norm": 2.87613279803751,
+      "learning_rate": 1.990873131321067e-05,
+      "loss": 1.0264,
+      "step": 373
+    },
+    {
+      "epoch": 0.07195075028857253,
+      "grad_norm": 6.206970320226217,
+      "learning_rate": 1.9907889485663897e-05,
+      "loss": 0.9965,
+      "step": 374
+    },
+    {
+      "epoch": 0.0721431319738361,
+      "grad_norm": 2.4054901010144096,
+      "learning_rate": 1.9907043811533283e-05,
+      "loss": 0.9984,
+      "step": 375
+    },
+    {
+      "epoch": 0.07233551365909965,
+      "grad_norm": 1.7136796708493285,
+      "learning_rate": 1.9906194291147155e-05,
+      "loss": 1.0097,
+      "step": 376
+    },
+    {
+      "epoch": 0.07252789534436321,
+      "grad_norm": 2.537172622681187,
+      "learning_rate": 1.9905340924835322e-05,
+      "loss": 0.9815,
+      "step": 377
+    },
+    {
+      "epoch": 0.07272027702962677,
+      "grad_norm": 1.6832299532488413,
+      "learning_rate": 1.9904483712929094e-05,
+      "loss": 0.9647,
+      "step": 378
+    },
+    {
+      "epoch": 0.07291265871489035,
+      "grad_norm": 2.5267544000971793,
+      "learning_rate": 1.9903622655761267e-05,
+      "loss": 1.0301,
+      "step": 379
+    },
+    {
+      "epoch": 0.0731050404001539,
+      "grad_norm": 2.498689489563929,
+      "learning_rate": 1.990275775366613e-05,
+      "loss": 0.9887,
+      "step": 380
+    },
+    {
+      "epoch": 0.07329742208541747,
+      "grad_norm": 3.1848182529738045,
+      "learning_rate": 1.9901889006979473e-05,
+      "loss": 1.0261,
+      "step": 381
+    },
+    {
+      "epoch": 0.07348980377068103,
+      "grad_norm": 5.916564491967754,
+      "learning_rate": 1.990101641603857e-05,
+      "loss": 0.9797,
+      "step": 382
+    },
+    {
+      "epoch": 0.0736821854559446,
+      "grad_norm": 1.6084700455262597,
+      "learning_rate": 1.9900139981182193e-05,
+      "loss": 1.0212,
+      "step": 383
+    },
+    {
+      "epoch": 0.07387456714120816,
+      "grad_norm": 0.6907096029736617,
+      "learning_rate": 1.9899259702750604e-05,
+      "loss": 0.2801,
+      "step": 384
+    },
+    {
+      "epoch": 0.07406694882647172,
+      "grad_norm": 4.580310839104569,
+      "learning_rate": 1.9898375581085555e-05,
+      "loss": 0.9541,
+      "step": 385
+    },
+    {
+      "epoch": 0.07425933051173528,
+      "grad_norm": 6.043655899370393,
+      "learning_rate": 1.9897487616530296e-05,
+      "loss": 0.9202,
+      "step": 386
+    },
+    {
+      "epoch": 0.07445171219699885,
+      "grad_norm": 7.471854037002716,
+      "learning_rate": 1.9896595809429565e-05,
+      "loss": 1.0125,
+      "step": 387
+    },
+    {
+      "epoch": 0.07464409388226241,
+      "grad_norm": 1.3030701949916372,
+      "learning_rate": 1.9895700160129593e-05,
+      "loss": 0.9449,
+      "step": 388
+    },
+    {
+      "epoch": 0.07483647556752597,
+      "grad_norm": 4.505141319123394,
+      "learning_rate": 1.9894800668978095e-05,
+      "loss": 0.9712,
+      "step": 389
+    },
+    {
+      "epoch": 0.07502885725278953,
+      "grad_norm": 1.8398340135026856,
+      "learning_rate": 1.9893897336324292e-05,
+      "loss": 0.9075,
+      "step": 390
+    },
+    {
+      "epoch": 0.0752212389380531,
+      "grad_norm": 4.0342903577032505,
+      "learning_rate": 1.9892990162518884e-05,
+      "loss": 0.9245,
+      "step": 391
+    },
+    {
+      "epoch": 0.07541362062331666,
+      "grad_norm": 24.14231905950537,
+      "learning_rate": 1.9892079147914072e-05,
+      "loss": 0.9664,
+      "step": 392
+    },
+    {
+      "epoch": 0.07560600230858022,
+      "grad_norm": 6.857953789121692,
+      "learning_rate": 1.9891164292863537e-05,
+      "loss": 0.9356,
+      "step": 393
+    },
+    {
+      "epoch": 0.07579838399384378,
+      "grad_norm": 1.3867095339020457,
+      "learning_rate": 1.9890245597722465e-05,
+      "loss": 0.9508,
+      "step": 394
+    },
+    {
+      "epoch": 0.07599076567910736,
+      "grad_norm": 5.004132606626806,
+      "learning_rate": 1.9889323062847516e-05,
+      "loss": 0.9693,
+      "step": 395
+    },
+    {
+      "epoch": 0.07618314736437092,
+      "grad_norm": 1.5632631200192422,
+      "learning_rate": 1.988839668859686e-05,
+      "loss": 0.9597,
+      "step": 396
+    },
+    {
+      "epoch": 0.07637552904963447,
+      "grad_norm": 2.0227880604139177,
+      "learning_rate": 1.988746647533014e-05,
+      "loss": 0.9908,
+      "step": 397
+    },
+    {
+      "epoch": 0.07656791073489803,
+      "grad_norm": 1.4090621066765685,
+      "learning_rate": 1.9886532423408495e-05,
+      "loss": 0.9925,
+      "step": 398
+    },
+    {
+      "epoch": 0.0767602924201616,
+      "grad_norm": 2.4756524330077676,
+      "learning_rate": 1.9885594533194564e-05,
+      "loss": 0.9933,
+      "step": 399
+    },
+    {
+      "epoch": 0.07695267410542517,
+      "grad_norm": 3.224444090047456,
+      "learning_rate": 1.9884652805052465e-05,
+      "loss": 0.9639,
+      "step": 400
+    },
+    {
+      "epoch": 0.07714505579068873,
+      "grad_norm": 2.427774263088068,
+      "learning_rate": 1.9883707239347804e-05,
+      "loss": 0.9685,
+      "step": 401
+    },
+    {
+      "epoch": 0.07733743747595229,
+      "grad_norm": 27.62032127383575,
+      "learning_rate": 1.988275783644769e-05,
+      "loss": 0.9799,
+      "step": 402
+    },
+    {
+      "epoch": 0.07752981916121585,
+      "grad_norm": 2.2049588117673826,
+      "learning_rate": 1.988180459672071e-05,
+      "loss": 0.9381,
+      "step": 403
+    },
+    {
+      "epoch": 0.07772220084647942,
+      "grad_norm": 2.3229228933562207,
+      "learning_rate": 1.988084752053695e-05,
+      "loss": 1.0227,
+      "step": 404
+    },
+    {
+      "epoch": 0.07791458253174298,
+      "grad_norm": 2.1495522486740875,
+      "learning_rate": 1.9879886608267967e-05,
+      "loss": 0.9335,
+      "step": 405
+    },
+    {
+      "epoch": 0.07810696421700654,
+      "grad_norm": 2.2909579166258087,
+      "learning_rate": 1.9878921860286832e-05,
+      "loss": 1.021,
+      "step": 406
+    },
+    {
+      "epoch": 0.0782993459022701,
+      "grad_norm": 2.0305719728287386,
+      "learning_rate": 1.9877953276968088e-05,
+      "loss": 0.9928,
+      "step": 407
+    },
+    {
+      "epoch": 0.07849172758753367,
+      "grad_norm": 2.4142061452580648,
+      "learning_rate": 1.9876980858687777e-05,
+      "loss": 0.9596,
+      "step": 408
+    },
+    {
+      "epoch": 0.07868410927279723,
+      "grad_norm": 2.567508504146908,
+      "learning_rate": 1.9876004605823417e-05,
+      "loss": 0.9523,
+      "step": 409
+    },
+    {
+      "epoch": 0.07887649095806079,
+      "grad_norm": 2.9227670959337475,
+      "learning_rate": 1.987502451875403e-05,
+      "loss": 0.9766,
+      "step": 410
+    },
+    {
+      "epoch": 0.07906887264332435,
+      "grad_norm": 2.0776131686278068,
+      "learning_rate": 1.987404059786012e-05,
+      "loss": 0.9813,
+      "step": 411
+    },
+    {
+      "epoch": 0.07926125432858792,
+      "grad_norm": 2.679537991463165,
+      "learning_rate": 1.9873052843523676e-05,
+      "loss": 0.956,
+      "step": 412
+    },
+    {
+      "epoch": 0.07945363601385148,
+      "grad_norm": 2.73326437430649,
+      "learning_rate": 1.987206125612818e-05,
+      "loss": 1.0067,
+      "step": 413
+    },
+    {
+      "epoch": 0.07964601769911504,
+      "grad_norm": 4.602573192236841,
+      "learning_rate": 1.98710658360586e-05,
+      "loss": 1.0221,
+      "step": 414
+    },
+    {
+      "epoch": 0.0798383993843786,
+      "grad_norm": 2.346417479918876,
+      "learning_rate": 1.987006658370139e-05,
+      "loss": 0.979,
+      "step": 415
+    },
+    {
+      "epoch": 0.08003078106964218,
+      "grad_norm": 2.5201438770435463,
+      "learning_rate": 1.9869063499444495e-05,
+      "loss": 0.9767,
+      "step": 416
+    },
+    {
+      "epoch": 0.08022316275490574,
+      "grad_norm": 2.5554859937814025,
+      "learning_rate": 1.9868056583677346e-05,
+      "loss": 0.977,
+      "step": 417
+    },
+    {
+      "epoch": 0.0804155444401693,
+      "grad_norm": 10.782178407864738,
+      "learning_rate": 1.9867045836790867e-05,
+      "loss": 0.9306,
+      "step": 418
+    },
+    {
+      "epoch": 0.08060792612543285,
+      "grad_norm": 0.6398162183371509,
+      "learning_rate": 1.9866031259177463e-05,
+      "loss": 0.3019,
+      "step": 419
+    },
+    {
+      "epoch": 0.08080030781069643,
+      "grad_norm": 13.01546945241854,
+      "learning_rate": 1.9865012851231022e-05,
+      "loss": 1.0989,
+      "step": 420
+    },
+    {
+      "epoch": 0.08099268949595999,
+      "grad_norm": 2.280181908967689,
+      "learning_rate": 1.9863990613346936e-05,
+      "loss": 0.9873,
+      "step": 421
+    },
+    {
+      "epoch": 0.08118507118122355,
+      "grad_norm": 1.7543003222453515,
+      "learning_rate": 1.986296454592206e-05,
+      "loss": 1.0102,
+      "step": 422
+    },
+    {
+      "epoch": 0.08137745286648711,
+      "grad_norm": 1.5510468762538718,
+      "learning_rate": 1.9861934649354763e-05,
+      "loss": 0.9175,
+      "step": 423
+    },
+    {
+      "epoch": 0.08156983455175067,
+      "grad_norm": 2.0777956179560872,
+      "learning_rate": 1.9860900924044873e-05,
+      "loss": 0.9472,
+      "step": 424
+    },
+    {
+      "epoch": 0.08176221623701424,
+      "grad_norm": 0.6488596813572075,
+      "learning_rate": 1.9859863370393726e-05,
+      "loss": 0.3257,
+      "step": 425
+    },
+    {
+      "epoch": 0.0819545979222778,
+      "grad_norm": 2.546728531046623,
+      "learning_rate": 1.9858821988804132e-05,
+      "loss": 0.9567,
+      "step": 426
+    },
+    {
+      "epoch": 0.08214697960754136,
+      "grad_norm": 4.268118232754378,
+      "learning_rate": 1.9857776779680393e-05,
+      "loss": 1.0126,
+      "step": 427
+    },
+    {
+      "epoch": 0.08233936129280492,
+      "grad_norm": 5.322938362352782,
+      "learning_rate": 1.98567277434283e-05,
+      "loss": 1.011,
+      "step": 428
+    },
+    {
+      "epoch": 0.08253174297806849,
+      "grad_norm": 2.9719405812333637,
+      "learning_rate": 1.9855674880455115e-05,
+      "loss": 0.9647,
+      "step": 429
+    },
+    {
+      "epoch": 0.08272412466333205,
+      "grad_norm": 1.5564646872302321,
+      "learning_rate": 1.98546181911696e-05,
+      "loss": 0.9316,
+      "step": 430
+    },
+    {
+      "epoch": 0.08291650634859561,
+      "grad_norm": 6.457403614501553,
+      "learning_rate": 1.9853557675982e-05,
+      "loss": 1.0056,
+      "step": 431
+    },
+    {
+      "epoch": 0.08310888803385917,
+      "grad_norm": 2.4602515198047525,
+      "learning_rate": 1.985249333530404e-05,
+      "loss": 0.9992,
+      "step": 432
+    },
+    {
+      "epoch": 0.08330126971912274,
+      "grad_norm": 10.482545631067339,
+      "learning_rate": 1.9851425169548938e-05,
+      "loss": 1.0042,
+      "step": 433
+    },
+    {
+      "epoch": 0.0834936514043863,
+      "grad_norm": 1.8799258352319557,
+      "learning_rate": 1.9850353179131392e-05,
+      "loss": 0.9282,
+      "step": 434
+    },
+    {
+      "epoch": 0.08368603308964986,
+      "grad_norm": 3.133846103280131,
+      "learning_rate": 1.9849277364467585e-05,
+      "loss": 1.0229,
+      "step": 435
+    },
+    {
+      "epoch": 0.08387841477491342,
+      "grad_norm": 0.6689791731856818,
+      "learning_rate": 1.984819772597518e-05,
+      "loss": 0.3324,
+      "step": 436
+    },
+    {
+      "epoch": 0.084070796460177,
+      "grad_norm": 2.2763207808844634,
+      "learning_rate": 1.9847114264073336e-05,
+      "loss": 0.9322,
+      "step": 437
+    },
+    {
+      "epoch": 0.08426317814544056,
+      "grad_norm": 5.630328677976771,
+      "learning_rate": 1.984602697918269e-05,
+      "loss": 0.9676,
+      "step": 438
+    },
+    {
+      "epoch": 0.08445555983070412,
+      "grad_norm": 2.241725147102341,
+      "learning_rate": 1.9844935871725363e-05,
+      "loss": 0.9585,
+      "step": 439
+    },
+    {
+      "epoch": 0.08464794151596768,
+      "grad_norm": 3.517758064906209,
+      "learning_rate": 1.9843840942124956e-05,
+      "loss": 0.9869,
+      "step": 440
+    },
+    {
+      "epoch": 0.08484032320123125,
+      "grad_norm": 1.93977266753099,
+      "learning_rate": 1.9842742190806566e-05,
+      "loss": 0.9448,
+      "step": 441
+    },
+    {
+      "epoch": 0.08503270488649481,
+      "grad_norm": 1.632866700270635,
+      "learning_rate": 1.984163961819676e-05,
+      "loss": 1.0086,
+      "step": 442
+    },
+    {
+      "epoch": 0.08522508657175837,
+      "grad_norm": 1.94709446184891,
+      "learning_rate": 1.9840533224723595e-05,
+      "loss": 0.9592,
+      "step": 443
+    },
+    {
+      "epoch": 0.08541746825702193,
+      "grad_norm": 1.9861644624655832,
+      "learning_rate": 1.9839423010816616e-05,
+      "loss": 0.9541,
+      "step": 444
+    },
+    {
+      "epoch": 0.08560984994228549,
+      "grad_norm": 2.2048874689098845,
+      "learning_rate": 1.983830897690684e-05,
+      "loss": 0.9312,
+      "step": 445
+    },
+    {
+      "epoch": 0.08580223162754906,
+      "grad_norm": 3.7131596100848947,
+      "learning_rate": 1.9837191123426777e-05,
+      "loss": 1.0162,
+      "step": 446
+    },
+    {
+      "epoch": 0.08599461331281262,
+      "grad_norm": 1.5995945095674273,
+      "learning_rate": 1.983606945081042e-05,
+      "loss": 1.0226,
+      "step": 447
+    },
+    {
+      "epoch": 0.08618699499807618,
+      "grad_norm": 2.1502573773401155,
+      "learning_rate": 1.983494395949323e-05,
+      "loss": 0.9867,
+      "step": 448
+    },
+    {
+      "epoch": 0.08637937668333974,
+      "grad_norm": 2.5712789593662606,
+      "learning_rate": 1.983381464991217e-05,
+      "loss": 0.9965,
+      "step": 449
+    },
+    {
+      "epoch": 0.08657175836860331,
+      "grad_norm": 1.3826173370840509,
+      "learning_rate": 1.9832681522505676e-05,
+      "loss": 0.9475,
+      "step": 450
+    },
+    {
+      "epoch": 0.08676414005386687,
+      "grad_norm": 1.6744078880786275,
+      "learning_rate": 1.9831544577713663e-05,
+      "loss": 0.9094,
+      "step": 451
+    },
+    {
+      "epoch": 0.08695652173913043,
+      "grad_norm": 2.415699923711429,
+      "learning_rate": 1.983040381597754e-05,
+      "loss": 0.9442,
+      "step": 452
+    },
+    {
+      "epoch": 0.08714890342439399,
+      "grad_norm": 2.0622157669633725,
+      "learning_rate": 1.982925923774018e-05,
+      "loss": 1.0205,
+      "step": 453
+    },
+    {
+      "epoch": 0.08734128510965757,
+      "grad_norm": 0.7048842603104774,
+      "learning_rate": 1.9828110843445954e-05,
+      "loss": 0.3131,
+      "step": 454
+    },
+    {
+      "epoch": 0.08753366679492113,
+      "grad_norm": 0.6490267678623377,
+      "learning_rate": 1.982695863354071e-05,
+      "loss": 0.3543,
+      "step": 455
+    },
+    {
+      "epoch": 0.08772604848018468,
+      "grad_norm": 1.6090755543968194,
+      "learning_rate": 1.9825802608471767e-05,
+      "loss": 0.9628,
+      "step": 456
+    },
+    {
+      "epoch": 0.08791843016544824,
+      "grad_norm": 1.8783570353618793,
+      "learning_rate": 1.982464276868794e-05,
+      "loss": 0.9197,
+      "step": 457
+    },
+    {
+      "epoch": 0.08811081185071182,
+      "grad_norm": 3.111579337895232,
+      "learning_rate": 1.982347911463952e-05,
+      "loss": 1.0101,
+      "step": 458
+    },
+    {
+      "epoch": 0.08830319353597538,
+      "grad_norm": 2.385480613476829,
+      "learning_rate": 1.9822311646778277e-05,
+      "loss": 1.0503,
+      "step": 459
+    },
+    {
+      "epoch": 0.08849557522123894,
+      "grad_norm": 4.490637315377798,
+      "learning_rate": 1.982114036555746e-05,
+      "loss": 0.9906,
+      "step": 460
+    },
+    {
+      "epoch": 0.0886879569065025,
+      "grad_norm": 1.7115375952974123,
+      "learning_rate": 1.9819965271431797e-05,
+      "loss": 1.0044,
+      "step": 461
+    },
+    {
+      "epoch": 0.08888033859176607,
+      "grad_norm": 7.7422022494036815,
+      "learning_rate": 1.9818786364857506e-05,
+      "loss": 0.9986,
+      "step": 462
+    },
+    {
+      "epoch": 0.08907272027702963,
+      "grad_norm": 1.8065460667058943,
+      "learning_rate": 1.9817603646292278e-05,
+      "loss": 0.9547,
+      "step": 463
+    },
+    {
+      "epoch": 0.08926510196229319,
+      "grad_norm": 2.257803562228419,
+      "learning_rate": 1.9816417116195287e-05,
+      "loss": 0.9304,
+      "step": 464
+    },
+    {
+      "epoch": 0.08945748364755675,
+      "grad_norm": 1.4876369116956412,
+      "learning_rate": 1.9815226775027182e-05,
+      "loss": 0.9487,
+      "step": 465
+    },
+    {
+      "epoch": 0.08964986533282032,
+      "grad_norm": 0.6891251185934115,
+      "learning_rate": 1.9814032623250093e-05,
+      "loss": 0.3171,
+      "step": 466
+    },
+    {
+      "epoch": 0.08984224701808388,
+      "grad_norm": 2.161236445679132,
+      "learning_rate": 1.9812834661327632e-05,
+      "loss": 1.0063,
+      "step": 467
+    },
+    {
+      "epoch": 0.09003462870334744,
+      "grad_norm": 1.696635921387977,
+      "learning_rate": 1.9811632889724888e-05,
+      "loss": 0.9519,
+      "step": 468
+    },
+    {
+      "epoch": 0.090227010388611,
+      "grad_norm": 1.5526740341147631,
+      "learning_rate": 1.9810427308908437e-05,
+      "loss": 0.9407,
+      "step": 469
+    },
+    {
+      "epoch": 0.09041939207387456,
+      "grad_norm": 2.169802740771616,
+      "learning_rate": 1.9809217919346318e-05,
+      "loss": 0.9984,
+      "step": 470
+    },
+    {
+      "epoch": 0.09061177375913813,
+      "grad_norm": 2.5442890541249263,
+      "learning_rate": 1.980800472150806e-05,
+      "loss": 0.9856,
+      "step": 471
+    },
+    {
+      "epoch": 0.0908041554444017,
+      "grad_norm": 1.9482593971158253,
+      "learning_rate": 1.9806787715864674e-05,
+      "loss": 0.9835,
+      "step": 472
+    },
+    {
+      "epoch": 0.09099653712966525,
+      "grad_norm": 1.7896248936681614,
+      "learning_rate": 1.9805566902888637e-05,
+      "loss": 0.9624,
+      "step": 473
+    },
+    {
+      "epoch": 0.09118891881492881,
+      "grad_norm": 1.736846148533495,
+      "learning_rate": 1.9804342283053916e-05,
+      "loss": 0.9964,
+      "step": 474
+    },
+    {
+      "epoch": 0.09138130050019239,
+      "grad_norm": 1.888660255316938,
+      "learning_rate": 1.980311385683594e-05,
+      "loss": 0.873,
+      "step": 475
+    },
+    {
+      "epoch": 0.09157368218545595,
+      "grad_norm": 1.8255395144473543,
+      "learning_rate": 1.980188162471164e-05,
+      "loss": 0.9782,
+      "step": 476
+    },
+    {
+      "epoch": 0.0917660638707195,
+      "grad_norm": 2.640829571863251,
+      "learning_rate": 1.98006455871594e-05,
+      "loss": 0.9471,
+      "step": 477
+    },
+    {
+      "epoch": 0.09195844555598306,
+      "grad_norm": 8.520883023835419,
+      "learning_rate": 1.97994057446591e-05,
+      "loss": 1.0017,
+      "step": 478
+    },
+    {
+      "epoch": 0.09215082724124664,
+      "grad_norm": 1.9739046257002755,
+      "learning_rate": 1.979816209769209e-05,
+      "loss": 0.945,
+      "step": 479
+    },
+    {
+      "epoch": 0.0923432089265102,
+      "grad_norm": 2.138439843962102,
+      "learning_rate": 1.9796914646741187e-05,
+      "loss": 1.03,
+      "step": 480
+    },
+    {
+      "epoch": 0.09253559061177376,
+      "grad_norm": 4.107770431061617,
+      "learning_rate": 1.9795663392290702e-05,
+      "loss": 0.9787,
+      "step": 481
+    },
+    {
+      "epoch": 0.09272797229703732,
+      "grad_norm": 0.8018275783984364,
+      "learning_rate": 1.9794408334826415e-05,
+      "loss": 0.3312,
+      "step": 482
+    },
+    {
+      "epoch": 0.09292035398230089,
+      "grad_norm": 1.6138240834075674,
+      "learning_rate": 1.979314947483558e-05,
+      "loss": 0.9794,
+      "step": 483
+    },
+    {
+      "epoch": 0.09311273566756445,
+      "grad_norm": 2.468776738157773,
+      "learning_rate": 1.9791886812806932e-05,
+      "loss": 0.9667,
+      "step": 484
+    },
+    {
+      "epoch": 0.09330511735282801,
+      "grad_norm": 2.5890608097179206,
+      "learning_rate": 1.9790620349230676e-05,
+      "loss": 0.9297,
+      "step": 485
+    },
+    {
+      "epoch": 0.09349749903809157,
+      "grad_norm": 0.6848484280605575,
+      "learning_rate": 1.9789350084598504e-05,
+      "loss": 0.3175,
+      "step": 486
+    },
+    {
+      "epoch": 0.09368988072335514,
+      "grad_norm": 2.3955133068554395,
+      "learning_rate": 1.9788076019403565e-05,
+      "loss": 0.9775,
+      "step": 487
+    },
+    {
+      "epoch": 0.0938822624086187,
+      "grad_norm": 1.8956335506312585,
+      "learning_rate": 1.9786798154140507e-05,
+      "loss": 0.9794,
+      "step": 488
+    },
+    {
+      "epoch": 0.09407464409388226,
+      "grad_norm": 8.644656686486078,
+      "learning_rate": 1.9785516489305437e-05,
+      "loss": 0.9093,
+      "step": 489
+    },
+    {
+      "epoch": 0.09426702577914582,
+      "grad_norm": 12.96017742126758,
+      "learning_rate": 1.9784231025395936e-05,
+      "loss": 1.016,
+      "step": 490
+    },
+    {
+      "epoch": 0.0944594074644094,
+      "grad_norm": 1.9224726357517838,
+      "learning_rate": 1.9782941762911075e-05,
+      "loss": 0.9942,
+      "step": 491
+    },
+    {
+      "epoch": 0.09465178914967295,
+      "grad_norm": 2.2821306038702756,
+      "learning_rate": 1.9781648702351383e-05,
+      "loss": 0.968,
+      "step": 492
+    },
+    {
+      "epoch": 0.09484417083493651,
+      "grad_norm": 19.856835092752757,
+      "learning_rate": 1.9780351844218874e-05,
+      "loss": 1.0068,
+      "step": 493
+    },
+    {
+      "epoch": 0.09503655252020007,
+      "grad_norm": 2.0309890955001717,
+      "learning_rate": 1.977905118901703e-05,
+      "loss": 0.9801,
+      "step": 494
+    },
+    {
+      "epoch": 0.09522893420546363,
+      "grad_norm": 2.1744530156946325,
+      "learning_rate": 1.977774673725081e-05,
+      "loss": 0.9253,
+      "step": 495
+    },
+    {
+      "epoch": 0.0954213158907272,
+      "grad_norm": 1.8286036675106048,
+      "learning_rate": 1.977643848942665e-05,
+      "loss": 0.9904,
+      "step": 496
+    },
+    {
+      "epoch": 0.09561369757599077,
+      "grad_norm": 2.143871345041799,
+      "learning_rate": 1.977512644605246e-05,
+      "loss": 0.9647,
+      "step": 497
+    },
+    {
+      "epoch": 0.09580607926125433,
+      "grad_norm": 0.7314875800639394,
+      "learning_rate": 1.9773810607637612e-05,
+      "loss": 0.3366,
+      "step": 498
+    },
+    {
+      "epoch": 0.09599846094651789,
+      "grad_norm": 4.846187976082955,
+      "learning_rate": 1.9772490974692962e-05,
+      "loss": 0.9786,
+      "step": 499
+    },
+    {
+      "epoch": 0.09619084263178146,
+      "grad_norm": 1.8320331321013335,
+      "learning_rate": 1.9771167547730844e-05,
+      "loss": 0.963,
+      "step": 500
+    },
+    {
+      "epoch": 0.09638322431704502,
+      "grad_norm": 1.7866658300192384,
+      "learning_rate": 1.976984032726505e-05,
+      "loss": 0.9555,
+      "step": 501
+    },
+    {
+      "epoch": 0.09657560600230858,
+      "grad_norm": 2.094318410911561,
+      "learning_rate": 1.976850931381086e-05,
+      "loss": 0.9713,
+      "step": 502
+    },
+    {
+      "epoch": 0.09676798768757214,
+      "grad_norm": 2.289061208795748,
+      "learning_rate": 1.976717450788501e-05,
+      "loss": 0.9884,
+      "step": 503
+    },
+    {
+      "epoch": 0.09696036937283571,
+      "grad_norm": 1.7977697616606239,
+      "learning_rate": 1.9765835910005726e-05,
+      "loss": 1.0156,
+      "step": 504
+    },
+    {
+      "epoch": 0.09715275105809927,
+      "grad_norm": 2.982983010749027,
+      "learning_rate": 1.9764493520692685e-05,
+      "loss": 0.9394,
+      "step": 505
+    },
+    {
+      "epoch": 0.09734513274336283,
+      "grad_norm": 2.2574832283562367,
+      "learning_rate": 1.9763147340467067e-05,
+      "loss": 1.018,
+      "step": 506
+    },
+    {
+      "epoch": 0.09753751442862639,
+      "grad_norm": 5.313080638593996,
+      "learning_rate": 1.9761797369851498e-05,
+      "loss": 0.9481,
+      "step": 507
+    },
+    {
+      "epoch": 0.09772989611388996,
+      "grad_norm": 2.038833183128186,
+      "learning_rate": 1.9760443609370074e-05,
+      "loss": 0.9281,
+      "step": 508
+    },
+    {
+      "epoch": 0.09792227779915352,
+      "grad_norm": 0.8918037472675211,
+      "learning_rate": 1.975908605954838e-05,
+      "loss": 0.3411,
+      "step": 509
+    },
+    {
+      "epoch": 0.09811465948441708,
+      "grad_norm": 9.673157376112226,
+      "learning_rate": 1.9757724720913466e-05,
+      "loss": 1.0046,
+      "step": 510
+    },
+    {
+      "epoch": 0.09830704116968064,
+      "grad_norm": 1.8176861526145351,
+      "learning_rate": 1.9756359593993845e-05,
+      "loss": 0.922,
+      "step": 511
+    },
+    {
+      "epoch": 0.09849942285494422,
+      "grad_norm": 1.395681931142589,
+      "learning_rate": 1.975499067931951e-05,
+      "loss": 0.9647,
+      "step": 512
+    },
+    {
+      "epoch": 0.09869180454020778,
+      "grad_norm": 3.3824413063853584,
+      "learning_rate": 1.975361797742192e-05,
+      "loss": 0.9108,
+      "step": 513
+    },
+    {
+      "epoch": 0.09888418622547133,
+      "grad_norm": 2.466590455488595,
+      "learning_rate": 1.9752241488834002e-05,
+      "loss": 0.9305,
+      "step": 514
+    },
+    {
+      "epoch": 0.0990765679107349,
+      "grad_norm": 1.980516036912761,
+      "learning_rate": 1.975086121409016e-05,
+      "loss": 1.0077,
+      "step": 515
+    },
+    {
+      "epoch": 0.09926894959599845,
+      "grad_norm": 1.4186799462500568,
+      "learning_rate": 1.974947715372626e-05,
+      "loss": 0.9523,
+      "step": 516
+    },
+    {
+      "epoch": 0.09946133128126203,
+      "grad_norm": 2.071100306531761,
+      "learning_rate": 1.974808930827965e-05,
+      "loss": 1.0039,
+      "step": 517
+    },
+    {
+      "epoch": 0.09965371296652559,
+      "grad_norm": 1.7386027533225437,
+      "learning_rate": 1.9746697678289128e-05,
+      "loss": 0.9798,
+      "step": 518
+    },
+    {
+      "epoch": 0.09984609465178915,
+      "grad_norm": 2.027300873644433,
+      "learning_rate": 1.9745302264294982e-05,
+      "loss": 1.0197,
+      "step": 519
+    },
+    {
+      "epoch": 0.1000384763370527,
+      "grad_norm": 2.330490175063221,
+      "learning_rate": 1.9743903066838954e-05,
+      "loss": 1.027,
+      "step": 520
+    },
+    {
+      "epoch": 0.10023085802231628,
+      "grad_norm": 1.4945265046358427,
+      "learning_rate": 1.9742500086464266e-05,
+      "loss": 0.9487,
+      "step": 521
+    },
+    {
+      "epoch": 0.10042323970757984,
+      "grad_norm": 1.7121306665703804,
+      "learning_rate": 1.9741093323715597e-05,
+      "loss": 0.9212,
+      "step": 522
+    },
+    {
+      "epoch": 0.1006156213928434,
+      "grad_norm": 2.344750627051672,
+      "learning_rate": 1.9739682779139107e-05,
+      "loss": 1.0277,
+      "step": 523
+    },
+    {
+      "epoch": 0.10080800307810696,
+      "grad_norm": 1.6315191668676494,
+      "learning_rate": 1.9738268453282414e-05,
+      "loss": 0.9211,
+      "step": 524
+    },
+    {
+      "epoch": 0.10100038476337053,
+      "grad_norm": 2.7303767543489563,
+      "learning_rate": 1.9736850346694608e-05,
+      "loss": 0.9581,
+      "step": 525
+    },
+    {
+      "epoch": 0.10119276644863409,
+      "grad_norm": 1.4383501519512996,
+      "learning_rate": 1.973542845992625e-05,
+      "loss": 1.0223,
+      "step": 526
+    },
+    {
+      "epoch": 0.10138514813389765,
+      "grad_norm": 1.6865861860011835,
+      "learning_rate": 1.9734002793529362e-05,
+      "loss": 0.9806,
+      "step": 527
+    },
+    {
+      "epoch": 0.10157752981916121,
+      "grad_norm": 3.9378493179922067,
+      "learning_rate": 1.9732573348057437e-05,
+      "loss": 0.9515,
+      "step": 528
+    },
+    {
+      "epoch": 0.10176991150442478,
+      "grad_norm": 2.2168484112058096,
+      "learning_rate": 1.973114012406544e-05,
+      "loss": 0.921,
+      "step": 529
+    },
+    {
+      "epoch": 0.10196229318968834,
+      "grad_norm": 1.2523966880805453,
+      "learning_rate": 1.9729703122109788e-05,
+      "loss": 0.9577,
+      "step": 530
+    },
+    {
+      "epoch": 0.1021546748749519,
+      "grad_norm": 1.5923230283430676,
+      "learning_rate": 1.9728262342748384e-05,
+      "loss": 0.9742,
+      "step": 531
+    },
+    {
+      "epoch": 0.10234705656021546,
+      "grad_norm": 1.9161888888604428,
+      "learning_rate": 1.9726817786540584e-05,
+      "loss": 0.9584,
+      "step": 532
+    },
+    {
+      "epoch": 0.10253943824547904,
+      "grad_norm": 2.6547011532204263,
+      "learning_rate": 1.9725369454047215e-05,
+      "loss": 0.9195,
+      "step": 533
+    },
+    {
+      "epoch": 0.1027318199307426,
+      "grad_norm": 7.190511530342431,
+      "learning_rate": 1.9723917345830568e-05,
+      "loss": 0.9452,
+      "step": 534
+    },
+    {
+      "epoch": 0.10292420161600616,
+      "grad_norm": 1.4431699551160937,
+      "learning_rate": 1.9722461462454405e-05,
+      "loss": 0.9448,
+      "step": 535
+    },
+    {
+      "epoch": 0.10311658330126972,
+      "grad_norm": 1.4422348426261284,
+      "learning_rate": 1.9721001804483947e-05,
+      "loss": 0.9546,
+      "step": 536
+    },
+    {
+      "epoch": 0.10330896498653329,
+      "grad_norm": 1.7003173821572266,
+      "learning_rate": 1.9719538372485887e-05,
+      "loss": 1.0001,
+      "step": 537
+    },
+    {
+      "epoch": 0.10350134667179685,
+      "grad_norm": 1.8357082492204688,
+      "learning_rate": 1.9718071167028376e-05,
+      "loss": 0.932,
+      "step": 538
+    },
+    {
+      "epoch": 0.10369372835706041,
+      "grad_norm": 1.9710258503944644,
+      "learning_rate": 1.9716600188681038e-05,
+      "loss": 0.9814,
+      "step": 539
+    },
+    {
+      "epoch": 0.10388611004232397,
+      "grad_norm": 1.7077762812771666,
+      "learning_rate": 1.971512543801495e-05,
+      "loss": 0.912,
+      "step": 540
+    },
+    {
+      "epoch": 0.10407849172758753,
+      "grad_norm": 4.020843941842493,
+      "learning_rate": 1.9713646915602663e-05,
+      "loss": 0.9291,
+      "step": 541
+    },
+    {
+      "epoch": 0.1042708734128511,
+      "grad_norm": 1.7202959590199909,
+      "learning_rate": 1.9712164622018197e-05,
+      "loss": 1.0056,
+      "step": 542
+    },
+    {
+      "epoch": 0.10446325509811466,
+      "grad_norm": 1.386462184501969,
+      "learning_rate": 1.9710678557837024e-05,
+      "loss": 1.0022,
+      "step": 543
+    },
+    {
+      "epoch": 0.10465563678337822,
+      "grad_norm": 1.6084852510006593,
+      "learning_rate": 1.9709188723636088e-05,
+      "loss": 0.9961,
+      "step": 544
+    },
+    {
+      "epoch": 0.10484801846864178,
+      "grad_norm": 1.7321878676948104,
+      "learning_rate": 1.970769511999379e-05,
+      "loss": 0.9755,
+      "step": 545
+    },
+    {
+      "epoch": 0.10504040015390535,
+      "grad_norm": 1.6411762041167541,
+      "learning_rate": 1.9706197747490004e-05,
+      "loss": 0.9902,
+      "step": 546
+    },
+    {
+      "epoch": 0.10523278183916891,
+      "grad_norm": 1.4261489648092727,
+      "learning_rate": 1.9704696606706055e-05,
+      "loss": 0.9583,
+      "step": 547
+    },
+    {
+      "epoch": 0.10542516352443247,
+      "grad_norm": 1.3813228989483242,
+      "learning_rate": 1.9703191698224742e-05,
+      "loss": 0.88,
+      "step": 548
+    },
+    {
+      "epoch": 0.10561754520969603,
+      "grad_norm": 1.659016981286994,
+      "learning_rate": 1.9701683022630323e-05,
+      "loss": 0.9592,
+      "step": 549
+    },
+    {
+      "epoch": 0.1058099268949596,
+      "grad_norm": 1.7899269209409499,
+      "learning_rate": 1.9700170580508514e-05,
+      "loss": 0.9765,
+      "step": 550
+    },
+    {
+      "epoch": 0.10600230858022316,
+      "grad_norm": 1.7153479673949419,
+      "learning_rate": 1.9698654372446495e-05,
+      "loss": 0.9371,
+      "step": 551
+    },
+    {
+      "epoch": 0.10619469026548672,
+      "grad_norm": 1.7213424179964079,
+      "learning_rate": 1.969713439903292e-05,
+      "loss": 1.0006,
+      "step": 552
+    },
+    {
+      "epoch": 0.10638707195075028,
+      "grad_norm": 9.781912801077166,
+      "learning_rate": 1.9695610660857886e-05,
+      "loss": 0.9711,
+      "step": 553
+    },
+    {
+      "epoch": 0.10657945363601386,
+      "grad_norm": 5.8531526877325115,
+      "learning_rate": 1.9694083158512965e-05,
+      "loss": 1.0277,
+      "step": 554
+    },
+    {
+      "epoch": 0.10677183532127742,
+      "grad_norm": 2.6872867108563963,
+      "learning_rate": 1.9692551892591185e-05,
+      "loss": 0.997,
+      "step": 555
+    },
+    {
+      "epoch": 0.10696421700654098,
+      "grad_norm": 1.267573279125936,
+      "learning_rate": 1.9691016863687037e-05,
+      "loss": 0.9246,
+      "step": 556
+    },
+    {
+      "epoch": 0.10715659869180454,
+      "grad_norm": 1.6711074028302961,
+      "learning_rate": 1.968947807239647e-05,
+      "loss": 0.9469,
+      "step": 557
+    },
+    {
+      "epoch": 0.10734898037706811,
+      "grad_norm": 1.4238747457384777,
+      "learning_rate": 1.9687935519316897e-05,
+      "loss": 1.0015,
+      "step": 558
+    },
+    {
+      "epoch": 0.10754136206233167,
+      "grad_norm": 1.1785075387880062,
+      "learning_rate": 1.9686389205047186e-05,
+      "loss": 0.3265,
+      "step": 559
+    },
+    {
+      "epoch": 0.10773374374759523,
+      "grad_norm": 1.527467705235616,
+      "learning_rate": 1.9684839130187678e-05,
+      "loss": 1.0168,
+      "step": 560
+    },
+    {
+      "epoch": 0.10792612543285879,
+      "grad_norm": 1.649555885933623,
+      "learning_rate": 1.968328529534016e-05,
+      "loss": 0.9927,
+      "step": 561
+    },
+    {
+      "epoch": 0.10811850711812235,
+      "grad_norm": 1.9920360224563674,
+      "learning_rate": 1.9681727701107885e-05,
+      "loss": 0.9305,
+      "step": 562
+    },
+    {
+      "epoch": 0.10831088880338592,
+      "grad_norm": 2.35558875562952,
+      "learning_rate": 1.9680166348095568e-05,
+      "loss": 0.9723,
+      "step": 563
+    },
+    {
+      "epoch": 0.10850327048864948,
+      "grad_norm": 4.2006912579159845,
+      "learning_rate": 1.967860123690937e-05,
+      "loss": 0.9389,
+      "step": 564
+    },
+    {
+      "epoch": 0.10869565217391304,
+      "grad_norm": 2.4528986602132266,
+      "learning_rate": 1.9677032368156934e-05,
+      "loss": 0.9704,
+      "step": 565
+    },
+    {
+      "epoch": 0.1088880338591766,
+      "grad_norm": 3.484704853164341,
+      "learning_rate": 1.967545974244734e-05,
+      "loss": 0.9275,
+      "step": 566
+    },
+    {
+      "epoch": 0.10908041554444017,
+      "grad_norm": 1.581348241116543,
+      "learning_rate": 1.9673883360391138e-05,
+      "loss": 0.9635,
+      "step": 567
+    },
+    {
+      "epoch": 0.10927279722970373,
+      "grad_norm": 1.8286350924610968,
+      "learning_rate": 1.9672303222600333e-05,
+      "loss": 0.9187,
+      "step": 568
+    },
+    {
+      "epoch": 0.10946517891496729,
+      "grad_norm": 1.896206166749727,
+      "learning_rate": 1.967071932968839e-05,
+      "loss": 1.0018,
+      "step": 569
+    },
+    {
+      "epoch": 0.10965756060023085,
+      "grad_norm": 0.9818685923704381,
+      "learning_rate": 1.9669131682270232e-05,
+      "loss": 0.3257,
+      "step": 570
+    },
+    {
+      "epoch": 0.10984994228549443,
+      "grad_norm": 6.194742814936529,
+      "learning_rate": 1.9667540280962235e-05,
+      "loss": 0.9678,
+      "step": 571
+    },
+    {
+      "epoch": 0.11004232397075799,
+      "grad_norm": 2.524439375513444,
+      "learning_rate": 1.966594512638224e-05,
+      "loss": 0.8716,
+      "step": 572
+    },
+    {
+      "epoch": 0.11023470565602154,
+      "grad_norm": 1.5315027522178883,
+      "learning_rate": 1.9664346219149538e-05,
+      "loss": 0.9443,
+      "step": 573
+    },
+    {
+      "epoch": 0.1104270873412851,
+      "grad_norm": 1.5142317619006171,
+      "learning_rate": 1.966274355988488e-05,
+      "loss": 0.9815,
+      "step": 574
+    },
+    {
+      "epoch": 0.11061946902654868,
+      "grad_norm": 2.3961037813726493,
+      "learning_rate": 1.9661137149210473e-05,
+      "loss": 0.9326,
+      "step": 575
+    },
+    {
+      "epoch": 0.11081185071181224,
+      "grad_norm": 2.0259430973774957,
+      "learning_rate": 1.9659526987749987e-05,
+      "loss": 0.945,
+      "step": 576
+    },
+    {
+      "epoch": 0.1110042323970758,
+      "grad_norm": 1.3495050576258103,
+      "learning_rate": 1.9657913076128532e-05,
+      "loss": 0.9874,
+      "step": 577
+    },
+    {
+      "epoch": 0.11119661408233936,
+      "grad_norm": 1.6042814132252567,
+      "learning_rate": 1.965629541497269e-05,
+      "loss": 0.9599,
+      "step": 578
+    },
+    {
+      "epoch": 0.11138899576760293,
+      "grad_norm": 2.243931335054226,
+      "learning_rate": 1.9654674004910493e-05,
+      "loss": 0.9366,
+      "step": 579
+    },
+    {
+      "epoch": 0.11158137745286649,
+      "grad_norm": 1.3162323144537078,
+      "learning_rate": 1.9653048846571427e-05,
+      "loss": 0.9404,
+      "step": 580
+    },
+    {
+      "epoch": 0.11177375913813005,
+      "grad_norm": 1.919523134706591,
+      "learning_rate": 1.9651419940586437e-05,
+      "loss": 0.9529,
+      "step": 581
+    },
+    {
+      "epoch": 0.11196614082339361,
+      "grad_norm": 3.5273939400967738,
+      "learning_rate": 1.964978728758791e-05,
+      "loss": 1.01,
+      "step": 582
+    },
+    {
+      "epoch": 0.11215852250865718,
+      "grad_norm": 5.214485773314805,
+      "learning_rate": 1.9648150888209715e-05,
+      "loss": 0.9533,
+      "step": 583
+    },
+    {
+      "epoch": 0.11235090419392074,
+      "grad_norm": 1.3774343194139977,
+      "learning_rate": 1.9646510743087144e-05,
+      "loss": 0.9871,
+      "step": 584
+    },
+    {
+      "epoch": 0.1125432858791843,
+      "grad_norm": 1.6515205184408763,
+      "learning_rate": 1.964486685285697e-05,
+      "loss": 1.001,
+      "step": 585
+    },
+    {
+      "epoch": 0.11273566756444786,
+      "grad_norm": 1.4888079433757064,
+      "learning_rate": 1.9643219218157395e-05,
+      "loss": 0.911,
+      "step": 586
+    },
+    {
+      "epoch": 0.11292804924971142,
+      "grad_norm": 1.2048717127814297,
+      "learning_rate": 1.9641567839628092e-05,
+      "loss": 0.9392,
+      "step": 587
+    },
+    {
+      "epoch": 0.113120430934975,
+      "grad_norm": 1.6852635572900814,
+      "learning_rate": 1.963991271791019e-05,
+      "loss": 0.8947,
+      "step": 588
+    },
+    {
+      "epoch": 0.11331281262023855,
+      "grad_norm": 1.416739653820038,
+      "learning_rate": 1.9638253853646255e-05,
+      "loss": 0.9188,
+      "step": 589
+    },
+    {
+      "epoch": 0.11350519430550211,
+      "grad_norm": 1.5035762070302587,
+      "learning_rate": 1.9636591247480323e-05,
+      "loss": 0.933,
+      "step": 590
+    },
+    {
+      "epoch": 0.11369757599076567,
+      "grad_norm": 1.6989360223186245,
+      "learning_rate": 1.9634924900057867e-05,
+      "loss": 1.0466,
+      "step": 591
+    },
+    {
+      "epoch": 0.11388995767602925,
+      "grad_norm": 2.2744110557560897,
+      "learning_rate": 1.963325481202583e-05,
+      "loss": 0.926,
+      "step": 592
+    },
+    {
+      "epoch": 0.1140823393612928,
+      "grad_norm": 1.2929331405214373,
+      "learning_rate": 1.963158098403259e-05,
+      "loss": 0.9124,
+      "step": 593
+    },
+    {
+      "epoch": 0.11427472104655637,
+      "grad_norm": 1.4028981436964303,
+      "learning_rate": 1.9629903416727987e-05,
+      "loss": 0.9311,
+      "step": 594
+    },
+    {
+      "epoch": 0.11446710273181993,
+      "grad_norm": 0.861642167393819,
+      "learning_rate": 1.962822211076331e-05,
+      "loss": 0.3357,
+      "step": 595
+    },
+    {
+      "epoch": 0.1146594844170835,
+      "grad_norm": 1.520999615439995,
+      "learning_rate": 1.96265370667913e-05,
+      "loss": 0.9308,
+      "step": 596
+    },
+    {
+      "epoch": 0.11485186610234706,
+      "grad_norm": 2.697891616828697,
+      "learning_rate": 1.9624848285466146e-05,
+      "loss": 0.9338,
+      "step": 597
+    },
+    {
+      "epoch": 0.11504424778761062,
+      "grad_norm": 1.3220668845005745,
+      "learning_rate": 1.9623155767443498e-05,
+      "loss": 0.9204,
+      "step": 598
+    },
+    {
+      "epoch": 0.11523662947287418,
+      "grad_norm": 1.3566158194272702,
+      "learning_rate": 1.9621459513380445e-05,
+      "loss": 0.9507,
+      "step": 599
+    },
+    {
+      "epoch": 0.11542901115813775,
+      "grad_norm": 1.3928424912897632,
+      "learning_rate": 1.9619759523935532e-05,
+      "loss": 0.973,
+      "step": 600
+    },
+    {
+      "epoch": 0.11562139284340131,
+      "grad_norm": 1.5116491219045591,
+      "learning_rate": 1.9618055799768757e-05,
+      "loss": 0.9366,
+      "step": 601
+    },
+    {
+      "epoch": 0.11581377452866487,
+      "grad_norm": 1.2626681830520872,
+      "learning_rate": 1.961634834154156e-05,
+      "loss": 0.9379,
+      "step": 602
+    },
+    {
+      "epoch": 0.11600615621392843,
+      "grad_norm": 0.6596019382507137,
+      "learning_rate": 1.9614637149916834e-05,
+      "loss": 0.3637,
+      "step": 603
+    },
+    {
+      "epoch": 0.116198537899192,
+      "grad_norm": 1.2471841475156307,
+      "learning_rate": 1.9612922225558924e-05,
+      "loss": 0.9998,
+      "step": 604
+    },
+    {
+      "epoch": 0.11639091958445556,
+      "grad_norm": 1.5625761384188244,
+      "learning_rate": 1.961120356913363e-05,
+      "loss": 0.9535,
+      "step": 605
+    },
+    {
+      "epoch": 0.11658330126971912,
+      "grad_norm": 1.6365838790757472,
+      "learning_rate": 1.960948118130818e-05,
+      "loss": 1.0199,
+      "step": 606
+    },
+    {
+      "epoch": 0.11677568295498268,
+      "grad_norm": 1.5044227696330326,
+      "learning_rate": 1.9607755062751273e-05,
+      "loss": 0.9598,
+      "step": 607
+    },
+    {
+      "epoch": 0.11696806464024626,
+      "grad_norm": 1.3207099325589717,
+      "learning_rate": 1.9606025214133046e-05,
+      "loss": 0.9739,
+      "step": 608
+    },
+    {
+      "epoch": 0.11716044632550981,
+      "grad_norm": 1.315318669656008,
+      "learning_rate": 1.9604291636125084e-05,
+      "loss": 0.9475,
+      "step": 609
+    },
+    {
+      "epoch": 0.11735282801077337,
+      "grad_norm": 1.3243049174357604,
+      "learning_rate": 1.960255432940043e-05,
+      "loss": 0.987,
+      "step": 610
+    },
+    {
+      "epoch": 0.11754520969603693,
+      "grad_norm": 1.409663504374695,
+      "learning_rate": 1.9600813294633552e-05,
+      "loss": 0.9637,
+      "step": 611
+    },
+    {
+      "epoch": 0.1177375913813005,
+      "grad_norm": 0.6913381744713764,
+      "learning_rate": 1.9599068532500394e-05,
+      "loss": 0.3093,
+      "step": 612
+    },
+    {
+      "epoch": 0.11792997306656407,
+      "grad_norm": 1.3346874528410104,
+      "learning_rate": 1.9597320043678322e-05,
+      "loss": 0.9343,
+      "step": 613
+    },
+    {
+      "epoch": 0.11812235475182763,
+      "grad_norm": 1.490790743313668,
+      "learning_rate": 1.9595567828846166e-05,
+      "loss": 0.9577,
+      "step": 614
+    },
+    {
+      "epoch": 0.11831473643709119,
+      "grad_norm": 1.5649342629349645,
+      "learning_rate": 1.9593811888684192e-05,
+      "loss": 0.9023,
+      "step": 615
+    },
+    {
+      "epoch": 0.11850711812235475,
+      "grad_norm": 1.3480948694931163,
+      "learning_rate": 1.9592052223874115e-05,
+      "loss": 0.9528,
+      "step": 616
+    },
+    {
+      "epoch": 0.11869949980761832,
+      "grad_norm": 1.4690402820282027,
+      "learning_rate": 1.959028883509911e-05,
+      "loss": 0.9319,
+      "step": 617
+    },
+    {
+      "epoch": 0.11889188149288188,
+      "grad_norm": 1.8038151565059986,
+      "learning_rate": 1.9588521723043764e-05,
+      "loss": 0.9537,
+      "step": 618
+    },
+    {
+      "epoch": 0.11908426317814544,
+      "grad_norm": 1.4095200845292788,
+      "learning_rate": 1.958675088839415e-05,
+      "loss": 0.9546,
+      "step": 619
+    },
+    {
+      "epoch": 0.119276644863409,
+      "grad_norm": 2.960537018340571,
+      "learning_rate": 1.9584976331837758e-05,
+      "loss": 0.9244,
+      "step": 620
+    },
+    {
+      "epoch": 0.11946902654867257,
+      "grad_norm": 1.4149383901840544,
+      "learning_rate": 1.9583198054063535e-05,
+      "loss": 0.9271,
+      "step": 621
+    },
+    {
+      "epoch": 0.11966140823393613,
+      "grad_norm": 2.7270657774990408,
+      "learning_rate": 1.9581416055761865e-05,
+      "loss": 0.938,
+      "step": 622
+    },
+    {
+      "epoch": 0.11985378991919969,
+      "grad_norm": 0.7740265403097105,
+      "learning_rate": 1.9579630337624585e-05,
+      "loss": 0.3219,
+      "step": 623
+    },
+    {
+      "epoch": 0.12004617160446325,
+      "grad_norm": 1.7047296673403374,
+      "learning_rate": 1.9577840900344974e-05,
+      "loss": 0.9904,
+      "step": 624
+    },
+    {
+      "epoch": 0.12023855328972682,
+      "grad_norm": 1.3636331478478405,
+      "learning_rate": 1.9576047744617752e-05,
+      "loss": 0.9681,
+      "step": 625
+    },
+    {
+      "epoch": 0.12043093497499038,
+      "grad_norm": 0.5947682785400626,
+      "learning_rate": 1.957425087113908e-05,
+      "loss": 0.329,
+      "step": 626
+    },
+    {
+      "epoch": 0.12062331666025394,
+      "grad_norm": 1.5028364203058224,
+      "learning_rate": 1.9572450280606568e-05,
+      "loss": 0.9964,
+      "step": 627
+    },
+    {
+      "epoch": 0.1208156983455175,
+      "grad_norm": 1.1746954305890382,
+      "learning_rate": 1.9570645973719273e-05,
+      "loss": 0.9549,
+      "step": 628
+    },
+    {
+      "epoch": 0.12100808003078108,
+      "grad_norm": 1.5820440481633813,
+      "learning_rate": 1.9568837951177677e-05,
+      "loss": 0.9976,
+      "step": 629
+    },
+    {
+      "epoch": 0.12120046171604464,
+      "grad_norm": 1.2817732222853635,
+      "learning_rate": 1.9567026213683728e-05,
+      "loss": 0.9952,
+      "step": 630
+    },
+    {
+      "epoch": 0.1213928434013082,
+      "grad_norm": 0.8647748310861977,
+      "learning_rate": 1.9565210761940798e-05,
+      "loss": 0.3527,
+      "step": 631
+    },
+    {
+      "epoch": 0.12158522508657175,
+      "grad_norm": 7.233708434283594,
+      "learning_rate": 1.956339159665371e-05,
+      "loss": 0.9308,
+      "step": 632
+    },
+    {
+      "epoch": 0.12177760677183531,
+      "grad_norm": 1.329210331273863,
+      "learning_rate": 1.956156871852873e-05,
+      "loss": 0.9312,
+      "step": 633
+    },
+    {
+      "epoch": 0.12196998845709889,
+      "grad_norm": 0.6225168905955182,
+      "learning_rate": 1.9559742128273558e-05,
+      "loss": 0.3179,
+      "step": 634
+    },
+    {
+      "epoch": 0.12216237014236245,
+      "grad_norm": 1.2356977352252072,
+      "learning_rate": 1.9557911826597337e-05,
+      "loss": 0.9211,
+      "step": 635
+    },
+    {
+      "epoch": 0.12235475182762601,
+      "grad_norm": 1.42510427832234,
+      "learning_rate": 1.9556077814210662e-05,
+      "loss": 0.9233,
+      "step": 636
+    },
+    {
+      "epoch": 0.12254713351288957,
+      "grad_norm": 1.3530583908914826,
+      "learning_rate": 1.955424009182555e-05,
+      "loss": 1.0031,
+      "step": 637
+    },
+    {
+      "epoch": 0.12273951519815314,
+      "grad_norm": 1.4709130016322258,
+      "learning_rate": 1.955239866015547e-05,
+      "loss": 0.9897,
+      "step": 638
+    },
+    {
+      "epoch": 0.1229318968834167,
+      "grad_norm": 1.5205802374678457,
+      "learning_rate": 1.9550553519915335e-05,
+      "loss": 0.9989,
+      "step": 639
+    },
+    {
+      "epoch": 0.12312427856868026,
+      "grad_norm": 2.182518617792874,
+      "learning_rate": 1.954870467182149e-05,
+      "loss": 0.9445,
+      "step": 640
+    },
+    {
+      "epoch": 0.12331666025394382,
+      "grad_norm": 1.28097280141891,
+      "learning_rate": 1.954685211659172e-05,
+      "loss": 0.9739,
+      "step": 641
+    },
+    {
+      "epoch": 0.12350904193920739,
+      "grad_norm": 1.316286129767769,
+      "learning_rate": 1.9544995854945248e-05,
+      "loss": 0.9045,
+      "step": 642
+    },
+    {
+      "epoch": 0.12370142362447095,
+      "grad_norm": 1.6418976674953007,
+      "learning_rate": 1.954313588760274e-05,
+      "loss": 0.9363,
+      "step": 643
+    },
+    {
+      "epoch": 0.12389380530973451,
+      "grad_norm": 1.3159380310016804,
+      "learning_rate": 1.9541272215286304e-05,
+      "loss": 0.9798,
+      "step": 644
+    },
+    {
+      "epoch": 0.12408618699499807,
+      "grad_norm": 1.4426947374587622,
+      "learning_rate": 1.9539404838719477e-05,
+      "loss": 0.9207,
+      "step": 645
+    },
+    {
+      "epoch": 0.12427856868026164,
+      "grad_norm": 1.2929170057547281,
+      "learning_rate": 1.9537533758627242e-05,
+      "loss": 0.9414,
+      "step": 646
+    },
+    {
+      "epoch": 0.1244709503655252,
+      "grad_norm": 1.6652753032277068,
+      "learning_rate": 1.953565897573601e-05,
+      "loss": 0.9137,
+      "step": 647
+    },
+    {
+      "epoch": 0.12466333205078876,
+      "grad_norm": 2.6735558050789825,
+      "learning_rate": 1.9533780490773645e-05,
+      "loss": 0.9511,
+      "step": 648
+    },
+    {
+      "epoch": 0.12485571373605232,
+      "grad_norm": 1.5268009743458828,
+      "learning_rate": 1.9531898304469435e-05,
+      "loss": 0.9576,
+      "step": 649
+    },
+    {
+      "epoch": 0.1250480954213159,
+      "grad_norm": 0.8713704826098291,
+      "learning_rate": 1.953001241755411e-05,
+      "loss": 0.3132,
+      "step": 650
+    },
+    {
+      "epoch": 0.12524047710657946,
+      "grad_norm": 4.693372016554596,
+      "learning_rate": 1.952812283075984e-05,
+      "loss": 0.8927,
+      "step": 651
+    },
+    {
+      "epoch": 0.12543285879184302,
+      "grad_norm": 1.5223064742153187,
+      "learning_rate": 1.952622954482022e-05,
+      "loss": 0.9531,
+      "step": 652
+    },
+    {
+      "epoch": 0.12562524047710658,
+      "grad_norm": 1.2359376369400743,
+      "learning_rate": 1.9524332560470293e-05,
+      "loss": 0.9129,
+      "step": 653
+    },
+    {
+      "epoch": 0.12581762216237014,
+      "grad_norm": 1.3778718847036584,
+      "learning_rate": 1.9522431878446536e-05,
+      "loss": 0.9583,
+      "step": 654
+    },
+    {
+      "epoch": 0.1260100038476337,
+      "grad_norm": 1.757626376325446,
+      "learning_rate": 1.9520527499486856e-05,
+      "loss": 0.9488,
+      "step": 655
+    },
+    {
+      "epoch": 0.12620238553289725,
+      "grad_norm": 1.3384979461235145,
+      "learning_rate": 1.95186194243306e-05,
+      "loss": 0.9246,
+      "step": 656
+    },
+    {
+      "epoch": 0.12639476721816084,
+      "grad_norm": 1.3276983265767917,
+      "learning_rate": 1.9516707653718546e-05,
+      "loss": 0.9822,
+      "step": 657
+    },
+    {
+      "epoch": 0.1265871489034244,
+      "grad_norm": 1.8147516920872029,
+      "learning_rate": 1.9514792188392914e-05,
+      "loss": 0.9409,
+      "step": 658
+    },
+    {
+      "epoch": 0.12677953058868796,
+      "grad_norm": 1.2884876601461204,
+      "learning_rate": 1.9512873029097347e-05,
+      "loss": 0.8959,
+      "step": 659
+    },
+    {
+      "epoch": 0.12697191227395152,
+      "grad_norm": 1.320833636979032,
+      "learning_rate": 1.9510950176576933e-05,
+      "loss": 0.9338,
+      "step": 660
+    },
+    {
+      "epoch": 0.12716429395921508,
+      "grad_norm": 1.4942268897821616,
+      "learning_rate": 1.950902363157819e-05,
+      "loss": 0.9574,
+      "step": 661
+    },
+    {
+      "epoch": 0.12735667564447864,
+      "grad_norm": 1.7369517043205571,
+      "learning_rate": 1.950709339484907e-05,
+      "loss": 0.9406,
+      "step": 662
+    },
+    {
+      "epoch": 0.1275490573297422,
+      "grad_norm": 1.718616035802018,
+      "learning_rate": 1.9505159467138954e-05,
+      "loss": 0.9577,
+      "step": 663
+    },
+    {
+      "epoch": 0.12774143901500576,
+      "grad_norm": 1.018692784194588,
+      "learning_rate": 1.9503221849198655e-05,
+      "loss": 0.3278,
+      "step": 664
+    },
+    {
+      "epoch": 0.12793382070026935,
+      "grad_norm": 1.4679768199680194,
+      "learning_rate": 1.9501280541780435e-05,
+      "loss": 0.9696,
+      "step": 665
+    },
+    {
+      "epoch": 0.1281262023855329,
+      "grad_norm": 2.5144341322596215,
+      "learning_rate": 1.9499335545637968e-05,
+      "loss": 1.0068,
+      "step": 666
+    },
+    {
+      "epoch": 0.12831858407079647,
+      "grad_norm": 0.6486739239238266,
+      "learning_rate": 1.949738686152637e-05,
+      "loss": 0.3387,
+      "step": 667
+    },
+    {
+      "epoch": 0.12851096575606002,
+      "grad_norm": 1.464659760905128,
+      "learning_rate": 1.9495434490202188e-05,
+      "loss": 1.0074,
+      "step": 668
+    },
+    {
+      "epoch": 0.12870334744132358,
+      "grad_norm": 1.3633797236104535,
+      "learning_rate": 1.94934784324234e-05,
+      "loss": 0.8933,
+      "step": 669
+    },
+    {
+      "epoch": 0.12889572912658714,
+      "grad_norm": 1.2891706530366498,
+      "learning_rate": 1.9491518688949417e-05,
+      "loss": 0.9564,
+      "step": 670
+    },
+    {
+      "epoch": 0.1290881108118507,
+      "grad_norm": 1.5721734434613672,
+      "learning_rate": 1.9489555260541074e-05,
+      "loss": 0.9611,
+      "step": 671
+    },
+    {
+      "epoch": 0.12928049249711426,
+      "grad_norm": 1.540418190101427,
+      "learning_rate": 1.948758814796064e-05,
+      "loss": 1.0026,
+      "step": 672
+    },
+    {
+      "epoch": 0.12947287418237785,
+      "grad_norm": 1.8230493169444362,
+      "learning_rate": 1.9485617351971827e-05,
+      "loss": 0.877,
+      "step": 673
+    },
+    {
+      "epoch": 0.1296652558676414,
+      "grad_norm": 1.4264321632481436,
+      "learning_rate": 1.9483642873339753e-05,
+      "loss": 0.9582,
+      "step": 674
+    },
+    {
+      "epoch": 0.12985763755290497,
+      "grad_norm": 1.4748965137829024,
+      "learning_rate": 1.9481664712830987e-05,
+      "loss": 1.0147,
+      "step": 675
+    },
+    {
+      "epoch": 0.13005001923816853,
+      "grad_norm": 1.628983782183953,
+      "learning_rate": 1.9479682871213515e-05,
+      "loss": 0.9596,
+      "step": 676
+    },
+    {
+      "epoch": 0.1302424009234321,
+      "grad_norm": 1.5077543344795448,
+      "learning_rate": 1.9477697349256756e-05,
+      "loss": 0.9559,
+      "step": 677
+    },
+    {
+      "epoch": 0.13043478260869565,
+      "grad_norm": 1.2794498896869588,
+      "learning_rate": 1.947570814773156e-05,
+      "loss": 0.9603,
+      "step": 678
+    },
+    {
+      "epoch": 0.1306271642939592,
+      "grad_norm": 1.4864325430362975,
+      "learning_rate": 1.9473715267410206e-05,
+      "loss": 0.9465,
+      "step": 679
+    },
+    {
+      "epoch": 0.13081954597922277,
+      "grad_norm": 2.270304174675666,
+      "learning_rate": 1.9471718709066392e-05,
+      "loss": 0.9845,
+      "step": 680
+    },
+    {
+      "epoch": 0.13101192766448633,
+      "grad_norm": 1.5795312935251842,
+      "learning_rate": 1.9469718473475256e-05,
+      "loss": 1.0053,
+      "step": 681
+    },
+    {
+      "epoch": 0.13120430934974991,
+      "grad_norm": 1.2867775615739112,
+      "learning_rate": 1.9467714561413358e-05,
+      "loss": 0.9763,
+      "step": 682
+    },
+    {
+      "epoch": 0.13139669103501347,
+      "grad_norm": 1.4471358020665661,
+      "learning_rate": 1.9465706973658683e-05,
+      "loss": 0.8766,
+      "step": 683
+    },
+    {
+      "epoch": 0.13158907272027703,
+      "grad_norm": 1.0182313252410757,
+      "learning_rate": 1.9463695710990648e-05,
+      "loss": 0.3533,
+      "step": 684
+    },
+    {
+      "epoch": 0.1317814544055406,
+      "grad_norm": 1.4251050513296244,
+      "learning_rate": 1.946168077419009e-05,
+      "loss": 0.897,
+      "step": 685
+    },
+    {
+      "epoch": 0.13197383609080415,
+      "grad_norm": 2.0570146592942096,
+      "learning_rate": 1.9459662164039283e-05,
+      "loss": 0.9841,
+      "step": 686
+    },
+    {
+      "epoch": 0.1321662177760677,
+      "grad_norm": 1.2924932589911413,
+      "learning_rate": 1.9457639881321917e-05,
+      "loss": 0.9142,
+      "step": 687
+    },
+    {
+      "epoch": 0.13235859946133127,
+      "grad_norm": 1.7022283542957861,
+      "learning_rate": 1.9455613926823115e-05,
+      "loss": 0.9326,
+      "step": 688
+    },
+    {
+      "epoch": 0.13255098114659483,
+      "grad_norm": 4.606602523703765,
+      "learning_rate": 1.945358430132942e-05,
+      "loss": 0.9477,
+      "step": 689
+    },
+    {
+      "epoch": 0.13274336283185842,
+      "grad_norm": 1.5464198209663604,
+      "learning_rate": 1.9451551005628803e-05,
+      "loss": 1.0668,
+      "step": 690
+    },
+    {
+      "epoch": 0.13293574451712198,
+      "grad_norm": 1.4141544459023223,
+      "learning_rate": 1.9449514040510654e-05,
+      "loss": 0.9261,
+      "step": 691
+    },
+    {
+      "epoch": 0.13312812620238554,
+      "grad_norm": 1.5515757993994195,
+      "learning_rate": 1.9447473406765803e-05,
+      "loss": 0.9172,
+      "step": 692
+    },
+    {
+      "epoch": 0.1333205078876491,
+      "grad_norm": 1.2741000040666364,
+      "learning_rate": 1.9445429105186487e-05,
+      "loss": 0.8616,
+      "step": 693
+    },
+    {
+      "epoch": 0.13351288957291266,
+      "grad_norm": 1.9614807090541895,
+      "learning_rate": 1.9443381136566382e-05,
+      "loss": 1.0042,
+      "step": 694
+    },
+    {
+      "epoch": 0.13370527125817622,
+      "grad_norm": 1.4370761860169174,
+      "learning_rate": 1.9441329501700568e-05,
+      "loss": 0.9214,
+      "step": 695
+    },
+    {
+      "epoch": 0.13389765294343978,
+      "grad_norm": 1.669687141517634,
+      "learning_rate": 1.943927420138557e-05,
+      "loss": 0.9788,
+      "step": 696
+    },
+    {
+      "epoch": 0.13409003462870334,
+      "grad_norm": 3.1334438001806535,
+      "learning_rate": 1.9437215236419322e-05,
+      "loss": 0.8748,
+      "step": 697
+    },
+    {
+      "epoch": 0.13428241631396692,
+      "grad_norm": 1.240998371168093,
+      "learning_rate": 1.9435152607601187e-05,
+      "loss": 0.9233,
+      "step": 698
+    },
+    {
+      "epoch": 0.13447479799923048,
+      "grad_norm": 1.5728797819370877,
+      "learning_rate": 1.943308631573195e-05,
+      "loss": 0.9227,
+      "step": 699
+    },
+    {
+      "epoch": 0.13466717968449404,
+      "grad_norm": 2.6128374709389144,
+      "learning_rate": 1.9431016361613816e-05,
+      "loss": 0.8918,
+      "step": 700
+    },
+    {
+      "epoch": 0.1348595613697576,
+      "grad_norm": 1.194159231239783,
+      "learning_rate": 1.9428942746050406e-05,
+      "loss": 0.9773,
+      "step": 701
+    },
+    {
+      "epoch": 0.13505194305502116,
+      "grad_norm": 2.0422134608753266,
+      "learning_rate": 1.9426865469846773e-05,
+      "loss": 0.9602,
+      "step": 702
+    },
+    {
+      "epoch": 0.13524432474028472,
+      "grad_norm": 1.5631182075889705,
+      "learning_rate": 1.9424784533809393e-05,
+      "loss": 0.9429,
+      "step": 703
+    },
+    {
+      "epoch": 0.13543670642554828,
+      "grad_norm": 1.5475389145958292,
+      "learning_rate": 1.942269993874615e-05,
+      "loss": 0.9235,
+      "step": 704
+    },
+    {
+      "epoch": 0.13562908811081184,
+      "grad_norm": 1.7055836034294685,
+      "learning_rate": 1.9420611685466358e-05,
+      "loss": 0.8841,
+      "step": 705
+    },
+    {
+      "epoch": 0.1358214697960754,
+      "grad_norm": 1.4685392506473265,
+      "learning_rate": 1.9418519774780748e-05,
+      "loss": 0.9309,
+      "step": 706
+    },
+    {
+      "epoch": 0.136013851481339,
+      "grad_norm": 1.60693875781274,
+      "learning_rate": 1.9416424207501474e-05,
+      "loss": 0.9498,
+      "step": 707
+    },
+    {
+      "epoch": 0.13620623316660255,
+      "grad_norm": 1.430989704893107,
+      "learning_rate": 1.9414324984442102e-05,
+      "loss": 0.9154,
+      "step": 708
+    },
+    {
+      "epoch": 0.1363986148518661,
+      "grad_norm": 2.225040685749258,
+      "learning_rate": 1.9412222106417632e-05,
+      "loss": 0.9723,
+      "step": 709
+    },
+    {
+      "epoch": 0.13659099653712967,
+      "grad_norm": 2.024343841514712,
+      "learning_rate": 1.9410115574244462e-05,
+      "loss": 0.9049,
+      "step": 710
+    },
+    {
+      "epoch": 0.13678337822239323,
+      "grad_norm": 3.098850356723002,
+      "learning_rate": 1.9408005388740433e-05,
+      "loss": 0.9517,
+      "step": 711
+    },
+    {
+      "epoch": 0.13697575990765679,
+      "grad_norm": 1.5263206084690204,
+      "learning_rate": 1.9405891550724778e-05,
+      "loss": 0.9441,
+      "step": 712
+    },
+    {
+      "epoch": 0.13716814159292035,
+      "grad_norm": 1.988636050920499,
+      "learning_rate": 1.940377406101817e-05,
+      "loss": 0.9476,
+      "step": 713
+    },
+    {
+      "epoch": 0.1373605232781839,
+      "grad_norm": 2.268578801497943,
+      "learning_rate": 1.9401652920442694e-05,
+      "loss": 0.9952,
+      "step": 714
+    },
+    {
+      "epoch": 0.1375529049634475,
+      "grad_norm": 1.526561977513663,
+      "learning_rate": 1.9399528129821842e-05,
+      "loss": 1.0147,
+      "step": 715
+    },
+    {
+      "epoch": 0.13774528664871105,
+      "grad_norm": 1.5927237805404029,
+      "learning_rate": 1.939739968998054e-05,
+      "loss": 0.921,
+      "step": 716
+    },
+    {
+      "epoch": 0.1379376683339746,
+      "grad_norm": 1.3376727634413266,
+      "learning_rate": 1.939526760174511e-05,
+      "loss": 0.931,
+      "step": 717
+    },
+    {
+      "epoch": 0.13813005001923817,
+      "grad_norm": 2.418001445410859,
+      "learning_rate": 1.939313186594331e-05,
+      "loss": 0.9802,
+      "step": 718
+    },
+    {
+      "epoch": 0.13832243170450173,
+      "grad_norm": 2.829413331167143,
+      "learning_rate": 1.9390992483404308e-05,
+      "loss": 0.9728,
+      "step": 719
+    },
+    {
+      "epoch": 0.1385148133897653,
+      "grad_norm": 1.3903399014097961,
+      "learning_rate": 1.938884945495868e-05,
+      "loss": 0.9189,
+      "step": 720
+    },
+    {
+      "epoch": 0.13870719507502885,
+      "grad_norm": 1.5804159293734603,
+      "learning_rate": 1.9386702781438425e-05,
+      "loss": 0.9358,
+      "step": 721
+    },
+    {
+      "epoch": 0.1388995767602924,
+      "grad_norm": 1.3401438885205041,
+      "learning_rate": 1.938455246367696e-05,
+      "loss": 0.9524,
+      "step": 722
+    },
+    {
+      "epoch": 0.139091958445556,
+      "grad_norm": 2.7800546072508086,
+      "learning_rate": 1.9382398502509107e-05,
+      "loss": 0.9579,
+      "step": 723
+    },
+    {
+      "epoch": 0.13928434013081956,
+      "grad_norm": 1.3814474925051372,
+      "learning_rate": 1.938024089877111e-05,
+      "loss": 0.8869,
+      "step": 724
+    },
+    {
+      "epoch": 0.13947672181608312,
+      "grad_norm": 1.3013972268606158,
+      "learning_rate": 1.9378079653300624e-05,
+      "loss": 0.9804,
+      "step": 725
+    },
+    {
+      "epoch": 0.13966910350134668,
+      "grad_norm": 0.8781245968912329,
+      "learning_rate": 1.9375914766936723e-05,
+      "loss": 0.3482,
+      "step": 726
+    },
+    {
+      "epoch": 0.13986148518661023,
+      "grad_norm": 1.5472132578859605,
+      "learning_rate": 1.9373746240519884e-05,
+      "loss": 0.9739,
+      "step": 727
+    },
+    {
+      "epoch": 0.1400538668718738,
+      "grad_norm": 1.155010190120765,
+      "learning_rate": 1.937157407489201e-05,
+      "loss": 0.922,
+      "step": 728
+    },
+    {
+      "epoch": 0.14024624855713735,
+      "grad_norm": 1.7283240666048663,
+      "learning_rate": 1.9369398270896403e-05,
+      "loss": 0.94,
+      "step": 729
+    },
+    {
+      "epoch": 0.1404386302424009,
+      "grad_norm": 1.2579334025068054,
+      "learning_rate": 1.936721882937779e-05,
+      "loss": 0.9662,
+      "step": 730
+    },
+    {
+      "epoch": 0.14063101192766447,
+      "grad_norm": 1.3866043883324333,
+      "learning_rate": 1.9365035751182307e-05,
+      "loss": 0.9403,
+      "step": 731
+    },
+    {
+      "epoch": 0.14082339361292806,
+      "grad_norm": 1.2313422078142786,
+      "learning_rate": 1.93628490371575e-05,
+      "loss": 0.9061,
+      "step": 732
+    },
+    {
+      "epoch": 0.14101577529819162,
+      "grad_norm": 1.3686372913074647,
+      "learning_rate": 1.9360658688152322e-05,
+      "loss": 0.9527,
+      "step": 733
+    },
+    {
+      "epoch": 0.14120815698345518,
+      "grad_norm": 1.3908399758799583,
+      "learning_rate": 1.9358464705017143e-05,
+      "loss": 0.8833,
+      "step": 734
+    },
+    {
+      "epoch": 0.14140053866871874,
+      "grad_norm": 2.9338615499615823,
+      "learning_rate": 1.9356267088603745e-05,
+      "loss": 0.8645,
+      "step": 735
+    },
+    {
+      "epoch": 0.1415929203539823,
+      "grad_norm": 1.130620322530018,
+      "learning_rate": 1.9354065839765316e-05,
+      "loss": 0.9857,
+      "step": 736
+    },
+    {
+      "epoch": 0.14178530203924586,
+      "grad_norm": 1.5923115304456943,
+      "learning_rate": 1.9351860959356462e-05,
+      "loss": 0.9556,
+      "step": 737
+    },
+    {
+      "epoch": 0.14197768372450942,
+      "grad_norm": 1.2791701539347429,
+      "learning_rate": 1.9349652448233187e-05,
+      "loss": 0.9376,
+      "step": 738
+    },
+    {
+      "epoch": 0.14217006540977298,
+      "grad_norm": 1.3694379395866716,
+      "learning_rate": 1.934744030725291e-05,
+      "loss": 0.9355,
+      "step": 739
+    },
+    {
+      "epoch": 0.14236244709503657,
+      "grad_norm": 1.3314322907707943,
+      "learning_rate": 1.934522453727447e-05,
+      "loss": 0.925,
+      "step": 740
+    },
+    {
+      "epoch": 0.14255482878030012,
+      "grad_norm": 1.5839871200921445,
+      "learning_rate": 1.93430051391581e-05,
+      "loss": 0.9237,
+      "step": 741
+    },
+    {
+      "epoch": 0.14274721046556368,
+      "grad_norm": 1.284672795763551,
+      "learning_rate": 1.934078211376544e-05,
+      "loss": 0.9736,
+      "step": 742
+    },
+    {
+      "epoch": 0.14293959215082724,
+      "grad_norm": 1.3159965810547973,
+      "learning_rate": 1.9338555461959554e-05,
+      "loss": 0.9321,
+      "step": 743
+    },
+    {
+      "epoch": 0.1431319738360908,
+      "grad_norm": 1.3160979364661174,
+      "learning_rate": 1.93363251846049e-05,
+      "loss": 0.9189,
+      "step": 744
+    },
+    {
+      "epoch": 0.14332435552135436,
+      "grad_norm": 1.339977433449074,
+      "learning_rate": 1.9334091282567352e-05,
+      "loss": 0.9324,
+      "step": 745
+    },
+    {
+      "epoch": 0.14351673720661792,
+      "grad_norm": 1.3839287194460883,
+      "learning_rate": 1.9331853756714185e-05,
+      "loss": 1.006,
+      "step": 746
+    },
+    {
+      "epoch": 0.14370911889188148,
+      "grad_norm": 1.2272426558750362,
+      "learning_rate": 1.9329612607914088e-05,
+      "loss": 0.9614,
+      "step": 747
+    },
+    {
+      "epoch": 0.14390150057714507,
+      "grad_norm": 1.5302186940451483,
+      "learning_rate": 1.9327367837037142e-05,
+      "loss": 0.949,
+      "step": 748
+    },
+    {
+      "epoch": 0.14409388226240863,
+      "grad_norm": 1.4367313790566902,
+      "learning_rate": 1.9325119444954855e-05,
+      "loss": 0.945,
+      "step": 749
+    },
+    {
+      "epoch": 0.1442862639476722,
+      "grad_norm": 0.8286568713093568,
+      "learning_rate": 1.9322867432540126e-05,
+      "loss": 0.3505,
+      "step": 750
+    },
+    {
+      "epoch": 0.14447864563293575,
+      "grad_norm": 1.43815885767332,
+      "learning_rate": 1.9320611800667268e-05,
+      "loss": 0.9631,
+      "step": 751
+    },
+    {
+      "epoch": 0.1446710273181993,
+      "grad_norm": 1.4171320242480796,
+      "learning_rate": 1.9318352550211986e-05,
+      "loss": 0.9004,
+      "step": 752
+    },
+    {
+      "epoch": 0.14486340900346287,
+      "grad_norm": 1.6141042652422246,
+      "learning_rate": 1.9316089682051403e-05,
+      "loss": 0.9073,
+      "step": 753
+    },
+    {
+      "epoch": 0.14505579068872643,
+      "grad_norm": 1.3333355773271969,
+      "learning_rate": 1.9313823197064042e-05,
+      "loss": 0.8707,
+      "step": 754
+    },
+    {
+      "epoch": 0.14524817237399,
+      "grad_norm": 1.2534226214202226,
+      "learning_rate": 1.9311553096129835e-05,
+      "loss": 0.9798,
+      "step": 755
+    },
+    {
+      "epoch": 0.14544055405925355,
+      "grad_norm": 1.3221839045557355,
+      "learning_rate": 1.9309279380130112e-05,
+      "loss": 0.9253,
+      "step": 756
+    },
+    {
+      "epoch": 0.14563293574451713,
+      "grad_norm": 1.401150202951637,
+      "learning_rate": 1.93070020499476e-05,
+      "loss": 0.9295,
+      "step": 757
+    },
+    {
+      "epoch": 0.1458253174297807,
+      "grad_norm": 1.463020124367182,
+      "learning_rate": 1.930472110646645e-05,
+      "loss": 0.9125,
+      "step": 758
+    },
+    {
+      "epoch": 0.14601769911504425,
+      "grad_norm": 0.6792141637465359,
+      "learning_rate": 1.9302436550572187e-05,
+      "loss": 0.3042,
+      "step": 759
+    },
+    {
+      "epoch": 0.1462100808003078,
+      "grad_norm": 1.1745726617940904,
+      "learning_rate": 1.930014838315177e-05,
+      "loss": 0.8871,
+      "step": 760
+    },
+    {
+      "epoch": 0.14640246248557137,
+      "grad_norm": 1.3779996500219882,
+      "learning_rate": 1.9297856605093534e-05,
+      "loss": 0.9512,
+      "step": 761
+    },
+    {
+      "epoch": 0.14659484417083493,
+      "grad_norm": 1.5479398933252488,
+      "learning_rate": 1.9295561217287226e-05,
+      "loss": 0.9482,
+      "step": 762
+    },
+    {
+      "epoch": 0.1467872258560985,
+      "grad_norm": 1.4026244434838289,
+      "learning_rate": 1.9293262220624002e-05,
+      "loss": 0.9557,
+      "step": 763
+    },
+    {
+      "epoch": 0.14697960754136205,
+      "grad_norm": 1.121231609327139,
+      "learning_rate": 1.9290959615996407e-05,
+      "loss": 0.9453,
+      "step": 764
+    },
+    {
+      "epoch": 0.14717198922662564,
+      "grad_norm": 1.8729768737057098,
+      "learning_rate": 1.9288653404298392e-05,
+      "loss": 0.9634,
+      "step": 765
+    },
+    {
+      "epoch": 0.1473643709118892,
+      "grad_norm": 2.4191850095394636,
+      "learning_rate": 1.9286343586425307e-05,
+      "loss": 0.9382,
+      "step": 766
+    },
+    {
+      "epoch": 0.14755675259715276,
+      "grad_norm": 1.5186724055997487,
+      "learning_rate": 1.9284030163273907e-05,
+      "loss": 0.9678,
+      "step": 767
+    },
+    {
+      "epoch": 0.14774913428241632,
+      "grad_norm": 1.2905720428531529,
+      "learning_rate": 1.9281713135742333e-05,
+      "loss": 0.9177,
+      "step": 768
+    },
+    {
+      "epoch": 0.14794151596767988,
+      "grad_norm": 2.909792631374713,
+      "learning_rate": 1.9279392504730147e-05,
+      "loss": 0.9753,
+      "step": 769
+    },
+    {
+      "epoch": 0.14813389765294344,
+      "grad_norm": 1.2187822727115285,
+      "learning_rate": 1.9277068271138287e-05,
+      "loss": 0.8861,
+      "step": 770
+    },
+    {
+      "epoch": 0.148326279338207,
+      "grad_norm": 1.255417958029271,
+      "learning_rate": 1.9274740435869107e-05,
+      "loss": 0.9789,
+      "step": 771
+    },
+    {
+      "epoch": 0.14851866102347056,
+      "grad_norm": 1.1979305040858732,
+      "learning_rate": 1.927240899982635e-05,
+      "loss": 0.8899,
+      "step": 772
+    },
+    {
+      "epoch": 0.14871104270873411,
+      "grad_norm": 1.1500524437012163,
+      "learning_rate": 1.9270073963915162e-05,
+      "loss": 0.9439,
+      "step": 773
+    },
+    {
+      "epoch": 0.1489034243939977,
+      "grad_norm": 1.2798364307466623,
+      "learning_rate": 1.9267735329042086e-05,
+      "loss": 0.9656,
+      "step": 774
+    },
+    {
+      "epoch": 0.14909580607926126,
+      "grad_norm": 1.4854625069562029,
+      "learning_rate": 1.9265393096115056e-05,
+      "loss": 0.9189,
+      "step": 775
+    },
+    {
+      "epoch": 0.14928818776452482,
+      "grad_norm": 1.4527002124382726,
+      "learning_rate": 1.926304726604341e-05,
+      "loss": 0.9179,
+      "step": 776
+    },
+    {
+      "epoch": 0.14948056944978838,
+      "grad_norm": 1.2498672374169604,
+      "learning_rate": 1.9260697839737875e-05,
+      "loss": 0.9591,
+      "step": 777
+    },
+    {
+      "epoch": 0.14967295113505194,
+      "grad_norm": 1.5980444174108426,
+      "learning_rate": 1.925834481811059e-05,
+      "loss": 0.8975,
+      "step": 778
+    },
+    {
+      "epoch": 0.1498653328203155,
+      "grad_norm": 1.2249011290442635,
+      "learning_rate": 1.9255988202075065e-05,
+      "loss": 0.9154,
+      "step": 779
+    },
+    {
+      "epoch": 0.15005771450557906,
+      "grad_norm": 1.3092633291967184,
+      "learning_rate": 1.925362799254623e-05,
+      "loss": 0.9624,
+      "step": 780
+    },
+    {
+      "epoch": 0.15025009619084262,
+      "grad_norm": 1.2792722432211494,
+      "learning_rate": 1.9251264190440398e-05,
+      "loss": 0.939,
+      "step": 781
+    },
+    {
+      "epoch": 0.1504424778761062,
+      "grad_norm": 1.2824915880258942,
+      "learning_rate": 1.9248896796675277e-05,
+      "loss": 0.9455,
+      "step": 782
+    },
+    {
+      "epoch": 0.15063485956136977,
+      "grad_norm": 1.1899690160352354,
+      "learning_rate": 1.924652581216997e-05,
+      "loss": 0.9352,
+      "step": 783
+    },
+    {
+      "epoch": 0.15082724124663333,
+      "grad_norm": 1.1786815119531544,
+      "learning_rate": 1.9244151237844975e-05,
+      "loss": 0.9668,
+      "step": 784
+    },
+    {
+      "epoch": 0.15101962293189689,
+      "grad_norm": 2.0422789613990733,
+      "learning_rate": 1.9241773074622182e-05,
+      "loss": 0.9314,
+      "step": 785
+    },
+    {
+      "epoch": 0.15121200461716044,
+      "grad_norm": 2.260709960142809,
+      "learning_rate": 1.923939132342488e-05,
+      "loss": 0.9221,
+      "step": 786
+    },
+    {
+      "epoch": 0.151404386302424,
+      "grad_norm": 2.248618226097601,
+      "learning_rate": 1.923700598517775e-05,
+      "loss": 0.9777,
+      "step": 787
+    },
+    {
+      "epoch": 0.15159676798768756,
+      "grad_norm": 1.298234260788327,
+      "learning_rate": 1.923461706080685e-05,
+      "loss": 0.9451,
+      "step": 788
+    },
+    {
+      "epoch": 0.15178914967295112,
+      "grad_norm": 1.2198719412397148,
+      "learning_rate": 1.923222455123965e-05,
+      "loss": 0.9333,
+      "step": 789
+    },
+    {
+      "epoch": 0.1519815313582147,
+      "grad_norm": 1.1718973550731944,
+      "learning_rate": 1.9229828457405005e-05,
+      "loss": 0.8797,
+      "step": 790
+    },
+    {
+      "epoch": 0.15217391304347827,
+      "grad_norm": 1.3433894998588622,
+      "learning_rate": 1.9227428780233162e-05,
+      "loss": 0.9484,
+      "step": 791
+    },
+    {
+      "epoch": 0.15236629472874183,
+      "grad_norm": 1.2646941831164977,
+      "learning_rate": 1.922502552065576e-05,
+      "loss": 0.9127,
+      "step": 792
+    },
+    {
+      "epoch": 0.1525586764140054,
+      "grad_norm": 1.5850921653684604,
+      "learning_rate": 1.922261867960582e-05,
+      "loss": 0.9499,
+      "step": 793
+    },
+    {
+      "epoch": 0.15275105809926895,
+      "grad_norm": 1.2496108854027588,
+      "learning_rate": 1.9220208258017763e-05,
+      "loss": 0.9281,
+      "step": 794
+    },
+    {
+      "epoch": 0.1529434397845325,
+      "grad_norm": 7.954644532113613,
+      "learning_rate": 1.92177942568274e-05,
+      "loss": 0.9636,
+      "step": 795
+    },
+    {
+      "epoch": 0.15313582146979607,
+      "grad_norm": 1.4116720905473457,
+      "learning_rate": 1.921537667697193e-05,
+      "loss": 0.9331,
+      "step": 796
+    },
+    {
+      "epoch": 0.15332820315505963,
+      "grad_norm": 1.1573136223021976,
+      "learning_rate": 1.9212955519389938e-05,
+      "loss": 0.9701,
+      "step": 797
+    },
+    {
+      "epoch": 0.1535205848403232,
+      "grad_norm": 1.4316035215061613,
+      "learning_rate": 1.9210530785021405e-05,
+      "loss": 0.9517,
+      "step": 798
+    },
+    {
+      "epoch": 0.15371296652558677,
+      "grad_norm": 1.2245741616304413,
+      "learning_rate": 1.9208102474807692e-05,
+      "loss": 0.9641,
+      "step": 799
+    },
+    {
+      "epoch": 0.15390534821085033,
+      "grad_norm": 1.2183652698956193,
+      "learning_rate": 1.920567058969155e-05,
+      "loss": 0.9148,
+      "step": 800
+    },
+    {
+      "epoch": 0.1540977298961139,
+      "grad_norm": 1.3730258735383685,
+      "learning_rate": 1.920323513061713e-05,
+      "loss": 0.9857,
+      "step": 801
+    },
+    {
+      "epoch": 0.15429011158137745,
+      "grad_norm": 1.1960335207449806,
+      "learning_rate": 1.9200796098529956e-05,
+      "loss": 0.9141,
+      "step": 802
+    },
+    {
+      "epoch": 0.154482493266641,
+      "grad_norm": 1.4760500240764196,
+      "learning_rate": 1.919835349437694e-05,
+      "loss": 0.9806,
+      "step": 803
+    },
+    {
+      "epoch": 0.15467487495190457,
+      "grad_norm": 1.302560278184582,
+      "learning_rate": 1.9195907319106394e-05,
+      "loss": 0.8874,
+      "step": 804
+    },
+    {
+      "epoch": 0.15486725663716813,
+      "grad_norm": 1.4373621578409037,
+      "learning_rate": 1.9193457573667996e-05,
+      "loss": 0.9599,
+      "step": 805
+    },
+    {
+      "epoch": 0.1550596383224317,
+      "grad_norm": 1.6101207935475736,
+      "learning_rate": 1.919100425901283e-05,
+      "loss": 0.9545,
+      "step": 806
+    },
+    {
+      "epoch": 0.15525202000769528,
+      "grad_norm": 1.415056441412526,
+      "learning_rate": 1.9188547376093355e-05,
+      "loss": 0.9011,
+      "step": 807
+    },
+    {
+      "epoch": 0.15544440169295884,
+      "grad_norm": 1.566878349130002,
+      "learning_rate": 1.918608692586342e-05,
+      "loss": 0.9366,
+      "step": 808
+    },
+    {
+      "epoch": 0.1556367833782224,
+      "grad_norm": 1.389006659527061,
+      "learning_rate": 1.918362290927825e-05,
+      "loss": 0.9493,
+      "step": 809
+    },
+    {
+      "epoch": 0.15582916506348596,
+      "grad_norm": 1.4125370681000586,
+      "learning_rate": 1.9181155327294468e-05,
+      "loss": 0.9346,
+      "step": 810
+    },
+    {
+      "epoch": 0.15602154674874952,
+      "grad_norm": 2.745422152109825,
+      "learning_rate": 1.9178684180870072e-05,
+      "loss": 0.9273,
+      "step": 811
+    },
+    {
+      "epoch": 0.15621392843401308,
+      "grad_norm": 1.3954357206946169,
+      "learning_rate": 1.9176209470964446e-05,
+      "loss": 0.9354,
+      "step": 812
+    },
+    {
+      "epoch": 0.15640631011927664,
+      "grad_norm": 1.615217295937984,
+      "learning_rate": 1.9173731198538354e-05,
+      "loss": 0.9446,
+      "step": 813
+    },
+    {
+      "epoch": 0.1565986918045402,
+      "grad_norm": 1.3858337919378503,
+      "learning_rate": 1.9171249364553956e-05,
+      "loss": 0.9425,
+      "step": 814
+    },
+    {
+      "epoch": 0.15679107348980378,
+      "grad_norm": 3.1230297558487385,
+      "learning_rate": 1.9168763969974773e-05,
+      "loss": 0.9395,
+      "step": 815
+    },
+    {
+      "epoch": 0.15698345517506734,
+      "grad_norm": 1.2086298311222745,
+      "learning_rate": 1.916627501576573e-05,
+      "loss": 0.9509,
+      "step": 816
+    },
+    {
+      "epoch": 0.1571758368603309,
+      "grad_norm": 1.5734806238005654,
+      "learning_rate": 1.916378250289312e-05,
+      "loss": 0.8856,
+      "step": 817
+    },
+    {
+      "epoch": 0.15736821854559446,
+      "grad_norm": 1.4077958218868056,
+      "learning_rate": 1.9161286432324628e-05,
+      "loss": 0.9674,
+      "step": 818
+    },
+    {
+      "epoch": 0.15756060023085802,
+      "grad_norm": 1.7072550950389194,
+      "learning_rate": 1.9158786805029307e-05,
+      "loss": 0.8888,
+      "step": 819
+    },
+    {
+      "epoch": 0.15775298191612158,
+      "grad_norm": 1.3567982279194994,
+      "learning_rate": 1.9156283621977603e-05,
+      "loss": 0.962,
+      "step": 820
+    },
+    {
+      "epoch": 0.15794536360138514,
+      "grad_norm": 2.1839650251177307,
+      "learning_rate": 1.9153776884141336e-05,
+      "loss": 0.9776,
+      "step": 821
+    },
+    {
+      "epoch": 0.1581377452866487,
+      "grad_norm": 2.5089387153978766,
+      "learning_rate": 1.915126659249371e-05,
+      "loss": 0.9854,
+      "step": 822
+    },
+    {
+      "epoch": 0.15833012697191226,
+      "grad_norm": 1.3396296582339788,
+      "learning_rate": 1.9148752748009304e-05,
+      "loss": 0.8939,
+      "step": 823
+    },
+    {
+      "epoch": 0.15852250865717585,
+      "grad_norm": 1.5114134645364747,
+      "learning_rate": 1.914623535166408e-05,
+      "loss": 0.8655,
+      "step": 824
+    },
+    {
+      "epoch": 0.1587148903424394,
+      "grad_norm": 0.6753973136350012,
+      "learning_rate": 1.9143714404435382e-05,
+      "loss": 0.359,
+      "step": 825
+    },
+    {
+      "epoch": 0.15890727202770297,
+      "grad_norm": 2.2914887131400667,
+      "learning_rate": 1.9141189907301922e-05,
+      "loss": 0.9812,
+      "step": 826
+    },
+    {
+      "epoch": 0.15909965371296653,
+      "grad_norm": 2.540521855782035,
+      "learning_rate": 1.9138661861243802e-05,
+      "loss": 0.8844,
+      "step": 827
+    },
+    {
+      "epoch": 0.1592920353982301,
+      "grad_norm": 1.2950570671902897,
+      "learning_rate": 1.913613026724249e-05,
+      "loss": 0.957,
+      "step": 828
+    },
+    {
+      "epoch": 0.15948441708349365,
+      "grad_norm": 1.616375722709864,
+      "learning_rate": 1.9133595126280848e-05,
+      "loss": 0.9217,
+      "step": 829
+    },
+    {
+      "epoch": 0.1596767987687572,
+      "grad_norm": 1.3271886405368483,
+      "learning_rate": 1.9131056439343095e-05,
+      "loss": 0.9219,
+      "step": 830
+    },
+    {
+      "epoch": 0.15986918045402077,
+      "grad_norm": 12.13693199267405,
+      "learning_rate": 1.9128514207414838e-05,
+      "loss": 0.902,
+      "step": 831
+    },
+    {
+      "epoch": 0.16006156213928435,
+      "grad_norm": 1.3250623768136696,
+      "learning_rate": 1.9125968431483068e-05,
+      "loss": 0.9515,
+      "step": 832
+    },
+    {
+      "epoch": 0.1602539438245479,
+      "grad_norm": 1.9785819524591886,
+      "learning_rate": 1.9123419112536132e-05,
+      "loss": 0.9003,
+      "step": 833
+    },
+    {
+      "epoch": 0.16044632550981147,
+      "grad_norm": 1.2828648252059576,
+      "learning_rate": 1.912086625156377e-05,
+      "loss": 0.9982,
+      "step": 834
+    },
+    {
+      "epoch": 0.16063870719507503,
+      "grad_norm": 1.8341230642795054,
+      "learning_rate": 1.911830984955709e-05,
+      "loss": 1.0124,
+      "step": 835
+    },
+    {
+      "epoch": 0.1608310888803386,
+      "grad_norm": 1.5118979800968764,
+      "learning_rate": 1.911574990750857e-05,
+      "loss": 0.9159,
+      "step": 836
+    },
+    {
+      "epoch": 0.16102347056560215,
+      "grad_norm": 1.6377407696177686,
+      "learning_rate": 1.9113186426412073e-05,
+      "loss": 0.9008,
+      "step": 837
+    },
+    {
+      "epoch": 0.1612158522508657,
+      "grad_norm": 1.3312536566734143,
+      "learning_rate": 1.9110619407262828e-05,
+      "loss": 0.9337,
+      "step": 838
+    },
+    {
+      "epoch": 0.16140823393612927,
+      "grad_norm": 1.5894459204016238,
+      "learning_rate": 1.9108048851057447e-05,
+      "loss": 0.9789,
+      "step": 839
+    },
+    {
+      "epoch": 0.16160061562139286,
+      "grad_norm": 3.7390923034982606,
+      "learning_rate": 1.9105474758793897e-05,
+      "loss": 0.9479,
+      "step": 840
+    },
+    {
+      "epoch": 0.16179299730665642,
+      "grad_norm": 1.590585432719347,
+      "learning_rate": 1.9102897131471536e-05,
+      "loss": 0.9455,
+      "step": 841
+    },
+    {
+      "epoch": 0.16198537899191998,
+      "grad_norm": 1.4052582065352688,
+      "learning_rate": 1.9100315970091088e-05,
+      "loss": 0.9459,
+      "step": 842
+    },
+    {
+      "epoch": 0.16217776067718354,
+      "grad_norm": 2.40634620155746,
+      "learning_rate": 1.9097731275654645e-05,
+      "loss": 0.9284,
+      "step": 843
+    },
+    {
+      "epoch": 0.1623701423624471,
+      "grad_norm": 1.6724936117447937,
+      "learning_rate": 1.909514304916568e-05,
+      "loss": 0.9461,
+      "step": 844
+    },
+    {
+      "epoch": 0.16256252404771065,
+      "grad_norm": 1.4382337331923878,
+      "learning_rate": 1.9092551291629026e-05,
+      "loss": 0.9615,
+      "step": 845
+    },
+    {
+      "epoch": 0.16275490573297421,
+      "grad_norm": 1.4926502325848336,
+      "learning_rate": 1.9089956004050893e-05,
+      "loss": 0.9639,
+      "step": 846
+    },
+    {
+      "epoch": 0.16294728741823777,
+      "grad_norm": 1.3533272282557758,
+      "learning_rate": 1.908735718743887e-05,
+      "loss": 0.9453,
+      "step": 847
+    },
+    {
+      "epoch": 0.16313966910350133,
+      "grad_norm": 1.4339441047961443,
+      "learning_rate": 1.908475484280189e-05,
+      "loss": 0.9391,
+      "step": 848
+    },
+    {
+      "epoch": 0.16333205078876492,
+      "grad_norm": 1.6259749333010665,
+      "learning_rate": 1.908214897115029e-05,
+      "loss": 0.9208,
+      "step": 849
+    },
+    {
+      "epoch": 0.16352443247402848,
+      "grad_norm": 1.4689998748467226,
+      "learning_rate": 1.907953957349575e-05,
+      "loss": 0.9867,
+      "step": 850
+    },
+    {
+      "epoch": 0.16371681415929204,
+      "grad_norm": 1.689105244167669,
+      "learning_rate": 1.907692665085133e-05,
+      "loss": 0.8907,
+      "step": 851
+    },
+    {
+      "epoch": 0.1639091958445556,
+      "grad_norm": 2.0081772984429147,
+      "learning_rate": 1.9074310204231457e-05,
+      "loss": 0.9277,
+      "step": 852
+    },
+    {
+      "epoch": 0.16410157752981916,
+      "grad_norm": 1.3243961291339292,
+      "learning_rate": 1.9071690234651923e-05,
+      "loss": 0.9035,
+      "step": 853
+    },
+    {
+      "epoch": 0.16429395921508272,
+      "grad_norm": 1.3085419993923069,
+      "learning_rate": 1.9069066743129893e-05,
+      "loss": 0.9877,
+      "step": 854
+    },
+    {
+      "epoch": 0.16448634090034628,
+      "grad_norm": 1.8128055199071498,
+      "learning_rate": 1.90664397306839e-05,
+      "loss": 0.9721,
+      "step": 855
+    },
+    {
+      "epoch": 0.16467872258560984,
+      "grad_norm": 1.4506312313514618,
+      "learning_rate": 1.9063809198333832e-05,
+      "loss": 0.9299,
+      "step": 856
+    },
+    {
+      "epoch": 0.16487110427087343,
+      "grad_norm": 1.472230001909075,
+      "learning_rate": 1.9061175147100957e-05,
+      "loss": 0.9193,
+      "step": 857
+    },
+    {
+      "epoch": 0.16506348595613698,
+      "grad_norm": 0.6348687138570442,
+      "learning_rate": 1.905853757800791e-05,
+      "loss": 0.3613,
+      "step": 858
+    },
+    {
+      "epoch": 0.16525586764140054,
+      "grad_norm": 1.3320344631584335,
+      "learning_rate": 1.9055896492078675e-05,
+      "loss": 0.9597,
+      "step": 859
+    },
+    {
+      "epoch": 0.1654482493266641,
+      "grad_norm": 1.4508166396403501,
+      "learning_rate": 1.905325189033862e-05,
+      "loss": 0.9244,
+      "step": 860
+    },
+    {
+      "epoch": 0.16564063101192766,
+      "grad_norm": 1.4108412061634772,
+      "learning_rate": 1.905060377381447e-05,
+      "loss": 0.9215,
+      "step": 861
+    },
+    {
+      "epoch": 0.16583301269719122,
+      "grad_norm": 1.3150109862819899,
+      "learning_rate": 1.904795214353431e-05,
+      "loss": 0.9628,
+      "step": 862
+    },
+    {
+      "epoch": 0.16602539438245478,
+      "grad_norm": 1.5093424921442522,
+      "learning_rate": 1.90452970005276e-05,
+      "loss": 0.9758,
+      "step": 863
+    },
+    {
+      "epoch": 0.16621777606771834,
+      "grad_norm": 1.4906416551888908,
+      "learning_rate": 1.9042638345825155e-05,
+      "loss": 0.9567,
+      "step": 864
+    },
+    {
+      "epoch": 0.16641015775298193,
+      "grad_norm": 1.316320447902619,
+      "learning_rate": 1.9039976180459158e-05,
+      "loss": 1.0269,
+      "step": 865
+    },
+    {
+      "epoch": 0.1666025394382455,
+      "grad_norm": 1.382785315162819,
+      "learning_rate": 1.9037310505463153e-05,
+      "loss": 0.9417,
+      "step": 866
+    },
+    {
+      "epoch": 0.16679492112350905,
+      "grad_norm": 1.2348748436415176,
+      "learning_rate": 1.9034641321872043e-05,
+      "loss": 0.9735,
+      "step": 867
+    },
+    {
+      "epoch": 0.1669873028087726,
+      "grad_norm": 1.5140667242297199,
+      "learning_rate": 1.9031968630722104e-05,
+      "loss": 0.9045,
+      "step": 868
+    },
+    {
+      "epoch": 0.16717968449403617,
+      "grad_norm": 2.0035339238220446,
+      "learning_rate": 1.902929243305096e-05,
+      "loss": 0.9276,
+      "step": 869
+    },
+    {
+      "epoch": 0.16737206617929973,
+      "grad_norm": 1.3372224733384088,
+      "learning_rate": 1.902661272989761e-05,
+      "loss": 0.8476,
+      "step": 870
+    },
+    {
+      "epoch": 0.1675644478645633,
+      "grad_norm": 1.3183291365201655,
+      "learning_rate": 1.9023929522302394e-05,
+      "loss": 0.9388,
+      "step": 871
+    },
+    {
+      "epoch": 0.16775682954982685,
+      "grad_norm": 1.3375521352507702,
+      "learning_rate": 1.9021242811307044e-05,
+      "loss": 0.979,
+      "step": 872
+    },
+    {
+      "epoch": 0.1679492112350904,
+      "grad_norm": 1.297606868979052,
+      "learning_rate": 1.901855259795462e-05,
+      "loss": 0.9101,
+      "step": 873
+    },
+    {
+      "epoch": 0.168141592920354,
+      "grad_norm": 1.422111396339113,
+      "learning_rate": 1.9015858883289556e-05,
+      "loss": 0.9126,
+      "step": 874
+    },
+    {
+      "epoch": 0.16833397460561755,
+      "grad_norm": 1.4488393051447908,
+      "learning_rate": 1.9013161668357655e-05,
+      "loss": 0.9051,
+      "step": 875
+    },
+    {
+      "epoch": 0.1685263562908811,
+      "grad_norm": 1.2831525078512906,
+      "learning_rate": 1.901046095420606e-05,
+      "loss": 0.9801,
+      "step": 876
+    },
+    {
+      "epoch": 0.16871873797614467,
+      "grad_norm": 1.9024507677081783,
+      "learning_rate": 1.9007756741883284e-05,
+      "loss": 0.8753,
+      "step": 877
+    },
+    {
+      "epoch": 0.16891111966140823,
+      "grad_norm": 1.6275372841660762,
+      "learning_rate": 1.9005049032439193e-05,
+      "loss": 0.9428,
+      "step": 878
+    },
+    {
+      "epoch": 0.1691035013466718,
+      "grad_norm": 6.438908722357357,
+      "learning_rate": 1.9002337826925012e-05,
+      "loss": 0.9621,
+      "step": 879
+    },
+    {
+      "epoch": 0.16929588303193535,
+      "grad_norm": 0.662684508916197,
+      "learning_rate": 1.899962312639333e-05,
+      "loss": 0.341,
+      "step": 880
+    },
+    {
+      "epoch": 0.1694882647171989,
+      "grad_norm": 1.3097459006727423,
+      "learning_rate": 1.8996904931898085e-05,
+      "loss": 0.9096,
+      "step": 881
+    },
+    {
+      "epoch": 0.1696806464024625,
+      "grad_norm": 1.3120966028576637,
+      "learning_rate": 1.899418324449457e-05,
+      "loss": 0.9577,
+      "step": 882
+    },
+    {
+      "epoch": 0.16987302808772606,
+      "grad_norm": 1.8407876083844978,
+      "learning_rate": 1.8991458065239444e-05,
+      "loss": 1.0078,
+      "step": 883
+    },
+    {
+      "epoch": 0.17006540977298962,
+      "grad_norm": 1.5072989692273462,
+      "learning_rate": 1.8988729395190712e-05,
+      "loss": 0.897,
+      "step": 884
+    },
+    {
+      "epoch": 0.17025779145825318,
+      "grad_norm": 1.4144871066565607,
+      "learning_rate": 1.8985997235407735e-05,
+      "loss": 0.9512,
+      "step": 885
+    },
+    {
+      "epoch": 0.17045017314351674,
+      "grad_norm": 1.4179685281081917,
+      "learning_rate": 1.898326158695124e-05,
+      "loss": 0.9252,
+      "step": 886
+    },
+    {
+      "epoch": 0.1706425548287803,
+      "grad_norm": 1.4958346907408933,
+      "learning_rate": 1.8980522450883287e-05,
+      "loss": 0.9254,
+      "step": 887
+    },
+    {
+      "epoch": 0.17083493651404386,
+      "grad_norm": 4.035385025964132,
+      "learning_rate": 1.8977779828267314e-05,
+      "loss": 0.9734,
+      "step": 888
+    },
+    {
+      "epoch": 0.17102731819930742,
+      "grad_norm": 1.2367421098353681,
+      "learning_rate": 1.8975033720168094e-05,
+      "loss": 0.9379,
+      "step": 889
+    },
+    {
+      "epoch": 0.17121969988457098,
+      "grad_norm": 1.5342569317685004,
+      "learning_rate": 1.897228412765177e-05,
+      "loss": 0.9319,
+      "step": 890
+    },
+    {
+      "epoch": 0.17141208156983456,
+      "grad_norm": 0.6114381756717626,
+      "learning_rate": 1.896953105178582e-05,
+      "loss": 0.3607,
+      "step": 891
+    },
+    {
+      "epoch": 0.17160446325509812,
+      "grad_norm": 81.79999997087403,
+      "learning_rate": 1.8966774493639084e-05,
+      "loss": 0.9503,
+      "step": 892
+    },
+    {
+      "epoch": 0.17179684494036168,
+      "grad_norm": 0.6824147464058579,
+      "learning_rate": 1.896401445428176e-05,
+      "loss": 0.3409,
+      "step": 893
+    },
+    {
+      "epoch": 0.17198922662562524,
+      "grad_norm": 4.394892609809291,
+      "learning_rate": 1.896125093478538e-05,
+      "loss": 0.9519,
+      "step": 894
+    },
+    {
+      "epoch": 0.1721816083108888,
+      "grad_norm": 0.6429298310226318,
+      "learning_rate": 1.895848393622284e-05,
+      "loss": 0.3396,
+      "step": 895
+    },
+    {
+      "epoch": 0.17237398999615236,
+      "grad_norm": 8.743891108811711,
+      "learning_rate": 1.895571345966839e-05,
+      "loss": 0.9153,
+      "step": 896
+    },
+    {
+      "epoch": 0.17256637168141592,
+      "grad_norm": 5.826561225438077,
+      "learning_rate": 1.8952939506197622e-05,
+      "loss": 0.9874,
+      "step": 897
+    },
+    {
+      "epoch": 0.17275875336667948,
+      "grad_norm": 8.16387264314585,
+      "learning_rate": 1.8950162076887477e-05,
+      "loss": 0.9336,
+      "step": 898
+    },
+    {
+      "epoch": 0.17295113505194307,
+      "grad_norm": 7.413546975126775,
+      "learning_rate": 1.894738117281625e-05,
+      "loss": 0.9541,
+      "step": 899
+    },
+    {
+      "epoch": 0.17314351673720663,
+      "grad_norm": 5.671774863268572,
+      "learning_rate": 1.8944596795063584e-05,
+      "loss": 0.95,
+      "step": 900
+    },
+    {
+      "epoch": 0.17333589842247019,
+      "grad_norm": 4.812582471713461,
+      "learning_rate": 1.894180894471047e-05,
+      "loss": 0.9503,
+      "step": 901
+    },
+    {
+      "epoch": 0.17352828010773375,
+      "grad_norm": 4.557049182820924,
+      "learning_rate": 1.8939017622839253e-05,
+      "loss": 0.8979,
+      "step": 902
+    },
+    {
+      "epoch": 0.1737206617929973,
+      "grad_norm": 0.6814202842876347,
+      "learning_rate": 1.8936222830533613e-05,
+      "loss": 0.3285,
+      "step": 903
+    },
+    {
+      "epoch": 0.17391304347826086,
+      "grad_norm": 3.626422556944087,
+      "learning_rate": 1.8933424568878586e-05,
+      "loss": 0.9111,
+      "step": 904
+    },
+    {
+      "epoch": 0.17410542516352442,
+      "grad_norm": 2.283521050332702,
+      "learning_rate": 1.8930622838960555e-05,
+      "loss": 1.0058,
+      "step": 905
+    },
+    {
+      "epoch": 0.17429780684878798,
+      "grad_norm": 2.7669348274181806,
+      "learning_rate": 1.8927817641867244e-05,
+      "loss": 0.9561,
+      "step": 906
+    },
+    {
+      "epoch": 0.17449018853405157,
+      "grad_norm": 2.48771854864518,
+      "learning_rate": 1.8925008978687737e-05,
+      "loss": 0.9064,
+      "step": 907
+    },
+    {
+      "epoch": 0.17468257021931513,
+      "grad_norm": 2.3165236536998326,
+      "learning_rate": 1.8922196850512446e-05,
+      "loss": 0.9895,
+      "step": 908
+    },
+    {
+      "epoch": 0.1748749519045787,
+      "grad_norm": 2.099855178158897,
+      "learning_rate": 1.8919381258433135e-05,
+      "loss": 0.9156,
+      "step": 909
+    },
+    {
+      "epoch": 0.17506733358984225,
+      "grad_norm": 2.521845899108205,
+      "learning_rate": 1.8916562203542916e-05,
+      "loss": 0.9409,
+      "step": 910
+    },
+    {
+      "epoch": 0.1752597152751058,
+      "grad_norm": 2.596695338338218,
+      "learning_rate": 1.8913739686936244e-05,
+      "loss": 0.8866,
+      "step": 911
+    },
+    {
+      "epoch": 0.17545209696036937,
+      "grad_norm": 3.216563387712872,
+      "learning_rate": 1.8910913709708918e-05,
+      "loss": 0.886,
+      "step": 912
+    },
+    {
+      "epoch": 0.17564447864563293,
+      "grad_norm": 2.553744982463681,
+      "learning_rate": 1.8908084272958077e-05,
+      "loss": 0.9112,
+      "step": 913
+    },
+    {
+      "epoch": 0.1758368603308965,
+      "grad_norm": 2.474256812694684,
+      "learning_rate": 1.8905251377782206e-05,
+      "loss": 0.8789,
+      "step": 914
+    },
+    {
+      "epoch": 0.17602924201616005,
+      "grad_norm": 0.7373537538761477,
+      "learning_rate": 1.8902415025281136e-05,
+      "loss": 0.3634,
+      "step": 915
+    },
+    {
+      "epoch": 0.17622162370142364,
+      "grad_norm": 1.9624632907385893,
+      "learning_rate": 1.889957521655603e-05,
+      "loss": 1.0162,
+      "step": 916
+    },
+    {
+      "epoch": 0.1764140053866872,
+      "grad_norm": 1.6436058641958113,
+      "learning_rate": 1.8896731952709408e-05,
+      "loss": 0.998,
+      "step": 917
+    },
+    {
+      "epoch": 0.17660638707195075,
+      "grad_norm": 4.323326561723672,
+      "learning_rate": 1.8893885234845117e-05,
+      "loss": 0.9804,
+      "step": 918
+    },
+    {
+      "epoch": 0.17679876875721431,
+      "grad_norm": 2.2209054357172024,
+      "learning_rate": 1.8891035064068354e-05,
+      "loss": 0.9155,
+      "step": 919
+    },
+    {
+      "epoch": 0.17699115044247787,
+      "grad_norm": 1.8786985985723017,
+      "learning_rate": 1.888818144148565e-05,
+      "loss": 0.9607,
+      "step": 920
+    },
+    {
+      "epoch": 0.17718353212774143,
+      "grad_norm": 8.046448143976951,
+      "learning_rate": 1.888532436820488e-05,
+      "loss": 0.9407,
+      "step": 921
+    },
+    {
+      "epoch": 0.177375913813005,
+      "grad_norm": 1.5516378620402926,
+      "learning_rate": 1.8882463845335263e-05,
+      "loss": 0.904,
+      "step": 922
+    },
+    {
+      "epoch": 0.17756829549826855,
+      "grad_norm": 1.5114247539230699,
+      "learning_rate": 1.8879599873987343e-05,
+      "loss": 0.917,
+      "step": 923
+    },
+    {
+      "epoch": 0.17776067718353214,
+      "grad_norm": 1.608545859148326,
+      "learning_rate": 1.8876732455273022e-05,
+      "loss": 0.9232,
+      "step": 924
+    },
+    {
+      "epoch": 0.1779530588687957,
+      "grad_norm": 2.328940098297717,
+      "learning_rate": 1.8873861590305527e-05,
+      "loss": 0.9492,
+      "step": 925
+    },
+    {
+      "epoch": 0.17814544055405926,
+      "grad_norm": 1.660489675461689,
+      "learning_rate": 1.8870987280199428e-05,
+      "loss": 0.9159,
+      "step": 926
+    },
+    {
+      "epoch": 0.17833782223932282,
+      "grad_norm": 3.3016622506757636,
+      "learning_rate": 1.886810952607063e-05,
+      "loss": 0.9005,
+      "step": 927
+    },
+    {
+      "epoch": 0.17853020392458638,
+      "grad_norm": 1.316729042662133,
+      "learning_rate": 1.8865228329036372e-05,
+      "loss": 0.9002,
+      "step": 928
+    },
+    {
+      "epoch": 0.17872258560984994,
+      "grad_norm": 1.6949685471997151,
+      "learning_rate": 1.886234369021524e-05,
+      "loss": 0.9334,
+      "step": 929
+    },
+    {
+      "epoch": 0.1789149672951135,
+      "grad_norm": 0.6699498768407273,
+      "learning_rate": 1.885945561072715e-05,
+      "loss": 0.3472,
+      "step": 930
+    },
+    {
+      "epoch": 0.17910734898037706,
+      "grad_norm": 1.4031236269017486,
+      "learning_rate": 1.885656409169335e-05,
+      "loss": 0.9415,
+      "step": 931
+    },
+    {
+      "epoch": 0.17929973066564064,
+      "grad_norm": 1.240184949564046,
+      "learning_rate": 1.885366913423643e-05,
+      "loss": 0.9432,
+      "step": 932
+    },
+    {
+      "epoch": 0.1794921123509042,
+      "grad_norm": 1.534614252794812,
+      "learning_rate": 1.8850770739480312e-05,
+      "loss": 0.9008,
+      "step": 933
+    },
+    {
+      "epoch": 0.17968449403616776,
+      "grad_norm": 1.6314743875022824,
+      "learning_rate": 1.8847868908550252e-05,
+      "loss": 0.9202,
+      "step": 934
+    },
+    {
+      "epoch": 0.17987687572143132,
+      "grad_norm": 1.5738755533077835,
+      "learning_rate": 1.8844963642572837e-05,
+      "loss": 0.9852,
+      "step": 935
+    },
+    {
+      "epoch": 0.18006925740669488,
+      "grad_norm": 1.8981068104899046,
+      "learning_rate": 1.8842054942676e-05,
+      "loss": 0.9567,
+      "step": 936
+    },
+    {
+      "epoch": 0.18026163909195844,
+      "grad_norm": 1.6943998091767027,
+      "learning_rate": 1.8839142809988987e-05,
+      "loss": 0.9113,
+      "step": 937
+    },
+    {
+      "epoch": 0.180454020777222,
+      "grad_norm": 2.0695744437434844,
+      "learning_rate": 1.88362272456424e-05,
+      "loss": 0.9799,
+      "step": 938
+    },
+    {
+      "epoch": 0.18064640246248556,
+      "grad_norm": 1.760141240096835,
+      "learning_rate": 1.8833308250768153e-05,
+      "loss": 0.9271,
+      "step": 939
+    },
+    {
+      "epoch": 0.18083878414774912,
+      "grad_norm": 0.6867871195490912,
+      "learning_rate": 1.8830385826499507e-05,
+      "loss": 0.3282,
+      "step": 940
+    },
+    {
+      "epoch": 0.1810311658330127,
+      "grad_norm": 1.2657172184960517,
+      "learning_rate": 1.882745997397104e-05,
+      "loss": 0.9156,
+      "step": 941
+    },
+    {
+      "epoch": 0.18122354751827627,
+      "grad_norm": 1.9494461434213062,
+      "learning_rate": 1.8824530694318675e-05,
+      "loss": 0.9124,
+      "step": 942
+    },
+    {
+      "epoch": 0.18141592920353983,
+      "grad_norm": 1.8237401558861572,
+      "learning_rate": 1.882159798867966e-05,
+      "loss": 0.9282,
+      "step": 943
+    },
+    {
+      "epoch": 0.1816083108888034,
+      "grad_norm": 2.7623464610178647,
+      "learning_rate": 1.8818661858192562e-05,
+      "loss": 0.93,
+      "step": 944
+    },
+    {
+      "epoch": 0.18180069257406695,
+      "grad_norm": 1.5166025658732623,
+      "learning_rate": 1.88157223039973e-05,
+      "loss": 0.8978,
+      "step": 945
+    },
+    {
+      "epoch": 0.1819930742593305,
+      "grad_norm": 1.3751714937671227,
+      "learning_rate": 1.8812779327235106e-05,
+      "loss": 0.9165,
+      "step": 946
+    },
+    {
+      "epoch": 0.18218545594459407,
+      "grad_norm": 2.7901709650379796,
+      "learning_rate": 1.880983292904854e-05,
+      "loss": 0.9081,
+      "step": 947
+    },
+    {
+      "epoch": 0.18237783762985763,
+      "grad_norm": 1.6213828895025786,
+      "learning_rate": 1.88068831105815e-05,
+      "loss": 0.9964,
+      "step": 948
+    },
+    {
+      "epoch": 0.1825702193151212,
+      "grad_norm": 4.041457023676932,
+      "learning_rate": 1.8803929872979214e-05,
+      "loss": 0.9469,
+      "step": 949
+    },
+    {
+      "epoch": 0.18276260100038477,
+      "grad_norm": 1.5634988931489373,
+      "learning_rate": 1.8800973217388215e-05,
+      "loss": 0.8851,
+      "step": 950
+    },
+    {
+      "epoch": 0.18295498268564833,
+      "grad_norm": 1.3040048180214143,
+      "learning_rate": 1.879801314495639e-05,
+      "loss": 0.9525,
+      "step": 951
+    },
+    {
+      "epoch": 0.1831473643709119,
+      "grad_norm": 1.5760578830002991,
+      "learning_rate": 1.879504965683294e-05,
+      "loss": 0.9455,
+      "step": 952
+    },
+    {
+      "epoch": 0.18333974605617545,
+      "grad_norm": 1.3763979698867583,
+      "learning_rate": 1.8792082754168385e-05,
+      "loss": 1.0075,
+      "step": 953
+    },
+    {
+      "epoch": 0.183532127741439,
+      "grad_norm": 1.616759213675038,
+      "learning_rate": 1.878911243811459e-05,
+      "loss": 0.9631,
+      "step": 954
+    },
+    {
+      "epoch": 0.18372450942670257,
+      "grad_norm": 1.4916666687347477,
+      "learning_rate": 1.8786138709824726e-05,
+      "loss": 0.9096,
+      "step": 955
+    },
+    {
+      "epoch": 0.18391689111196613,
+      "grad_norm": 1.5112191854592696,
+      "learning_rate": 1.8783161570453295e-05,
+      "loss": 0.8575,
+      "step": 956
+    },
+    {
+      "epoch": 0.18410927279722972,
+      "grad_norm": 0.6689178001053664,
+      "learning_rate": 1.878018102115614e-05,
+      "loss": 0.361,
+      "step": 957
+    },
+    {
+      "epoch": 0.18430165448249328,
+      "grad_norm": 1.3899056579064526,
+      "learning_rate": 1.8777197063090394e-05,
+      "loss": 0.9299,
+      "step": 958
+    },
+    {
+      "epoch": 0.18449403616775684,
+      "grad_norm": 1.4243133581354284,
+      "learning_rate": 1.877420969741454e-05,
+      "loss": 0.9041,
+      "step": 959
+    },
+    {
+      "epoch": 0.1846864178530204,
+      "grad_norm": 1.4940036408690676,
+      "learning_rate": 1.877121892528838e-05,
+      "loss": 0.9374,
+      "step": 960
+    },
+    {
+      "epoch": 0.18487879953828396,
+      "grad_norm": 1.9687601922830693,
+      "learning_rate": 1.876822474787303e-05,
+      "loss": 0.9369,
+      "step": 961
+    },
+    {
+      "epoch": 0.18507118122354752,
+      "grad_norm": 1.3379862705641488,
+      "learning_rate": 1.8765227166330933e-05,
+      "loss": 0.9946,
+      "step": 962
+    },
+    {
+      "epoch": 0.18526356290881107,
+      "grad_norm": 1.5176146016800125,
+      "learning_rate": 1.8762226181825857e-05,
+      "loss": 0.9939,
+      "step": 963
+    },
+    {
+      "epoch": 0.18545594459407463,
+      "grad_norm": 1.5224555086773377,
+      "learning_rate": 1.875922179552288e-05,
+      "loss": 0.9483,
+      "step": 964
+    },
+    {
+      "epoch": 0.1856483262793382,
+      "grad_norm": 3.97045709899023,
+      "learning_rate": 1.875621400858842e-05,
+      "loss": 0.91,
+      "step": 965
+    },
+    {
+      "epoch": 0.18584070796460178,
+      "grad_norm": 1.517406690027942,
+      "learning_rate": 1.875320282219019e-05,
+      "loss": 0.9732,
+      "step": 966
+    },
+    {
+      "epoch": 0.18603308964986534,
+      "grad_norm": 1.410401150610644,
+      "learning_rate": 1.8750188237497247e-05,
+      "loss": 0.9022,
+      "step": 967
+    },
+    {
+      "epoch": 0.1862254713351289,
+      "grad_norm": 1.4908325175567911,
+      "learning_rate": 1.874717025567995e-05,
+      "loss": 0.9615,
+      "step": 968
+    },
+    {
+      "epoch": 0.18641785302039246,
+      "grad_norm": 17.82896829043053,
+      "learning_rate": 1.874414887790999e-05,
+      "loss": 0.9618,
+      "step": 969
+    },
+    {
+      "epoch": 0.18661023470565602,
+      "grad_norm": 1.578840063274061,
+      "learning_rate": 1.8741124105360363e-05,
+      "loss": 0.9588,
+      "step": 970
+    },
+    {
+      "epoch": 0.18680261639091958,
+      "grad_norm": 1.5322305271650434,
+      "learning_rate": 1.873809593920539e-05,
+      "loss": 0.9351,
+      "step": 971
+    },
+    {
+      "epoch": 0.18699499807618314,
+      "grad_norm": 1.8703539531805342,
+      "learning_rate": 1.8735064380620717e-05,
+      "loss": 0.9473,
+      "step": 972
+    },
+    {
+      "epoch": 0.1871873797614467,
+      "grad_norm": 1.7697583939341972,
+      "learning_rate": 1.873202943078329e-05,
+      "loss": 0.9712,
+      "step": 973
+    },
+    {
+      "epoch": 0.18737976144671029,
+      "grad_norm": 1.1758716282554333,
+      "learning_rate": 1.8728991090871387e-05,
+      "loss": 0.8753,
+      "step": 974
+    },
+    {
+      "epoch": 0.18757214313197385,
+      "grad_norm": 5.813767891632461,
+      "learning_rate": 1.8725949362064596e-05,
+      "loss": 0.9492,
+      "step": 975
+    },
+    {
+      "epoch": 0.1877645248172374,
+      "grad_norm": 1.5382290209448044,
+      "learning_rate": 1.8722904245543817e-05,
+      "loss": 0.9604,
+      "step": 976
+    },
+    {
+      "epoch": 0.18795690650250096,
+      "grad_norm": 2.1148418361657315,
+      "learning_rate": 1.871985574249127e-05,
+      "loss": 0.9005,
+      "step": 977
+    },
+    {
+      "epoch": 0.18814928818776452,
+      "grad_norm": 1.9942739396925306,
+      "learning_rate": 1.8716803854090495e-05,
+      "loss": 0.9468,
+      "step": 978
+    },
+    {
+      "epoch": 0.18834166987302808,
+      "grad_norm": 1.4758868385923691,
+      "learning_rate": 1.8713748581526334e-05,
+      "loss": 0.9651,
+      "step": 979
+    },
+    {
+      "epoch": 0.18853405155829164,
+      "grad_norm": 2.417085702685872,
+      "learning_rate": 1.871068992598495e-05,
+      "loss": 0.9527,
+      "step": 980
+    },
+    {
+      "epoch": 0.1887264332435552,
+      "grad_norm": 2.226736703160175,
+      "learning_rate": 1.8707627888653816e-05,
+      "loss": 0.9429,
+      "step": 981
+    },
+    {
+      "epoch": 0.1889188149288188,
+      "grad_norm": 1.779735106009623,
+      "learning_rate": 1.8704562470721728e-05,
+      "loss": 0.9067,
+      "step": 982
+    },
+    {
+      "epoch": 0.18911119661408235,
+      "grad_norm": 1.5533353807167647,
+      "learning_rate": 1.870149367337878e-05,
+      "loss": 0.8886,
+      "step": 983
+    },
+    {
+      "epoch": 0.1893035782993459,
+      "grad_norm": 0.8059444351633595,
+      "learning_rate": 1.8698421497816386e-05,
+      "loss": 0.369,
+      "step": 984
+    },
+    {
+      "epoch": 0.18949595998460947,
+      "grad_norm": 2.7065463938835554,
+      "learning_rate": 1.869534594522727e-05,
+      "loss": 0.9354,
+      "step": 985
+    },
+    {
+      "epoch": 0.18968834166987303,
+      "grad_norm": 1.6243997244381199,
+      "learning_rate": 1.8692267016805473e-05,
+      "loss": 0.9152,
+      "step": 986
+    },
+    {
+      "epoch": 0.1898807233551366,
+      "grad_norm": 1.7866278783101672,
+      "learning_rate": 1.8689184713746333e-05,
+      "loss": 0.9261,
+      "step": 987
+    },
+    {
+      "epoch": 0.19007310504040015,
+      "grad_norm": 2.788527286381285,
+      "learning_rate": 1.868609903724651e-05,
+      "loss": 0.8412,
+      "step": 988
+    },
+    {
+      "epoch": 0.1902654867256637,
+      "grad_norm": 1.494087369590009,
+      "learning_rate": 1.8683009988503972e-05,
+      "loss": 0.8937,
+      "step": 989
+    },
+    {
+      "epoch": 0.19045786841092727,
+      "grad_norm": 0.612285091570073,
+      "learning_rate": 1.867991756871799e-05,
+      "loss": 0.3684,
+      "step": 990
+    },
+    {
+      "epoch": 0.19065025009619085,
+      "grad_norm": 1.4486239432294967,
+      "learning_rate": 1.867682177908915e-05,
+      "loss": 0.8756,
+      "step": 991
+    },
+    {
+      "epoch": 0.1908426317814544,
+      "grad_norm": 1.6890072029942935,
+      "learning_rate": 1.867372262081934e-05,
+      "loss": 0.9676,
+      "step": 992
+    },
+    {
+      "epoch": 0.19103501346671797,
+      "grad_norm": 1.693292271413478,
+      "learning_rate": 1.8670620095111766e-05,
+      "loss": 0.9551,
+      "step": 993
+    },
+    {
+      "epoch": 0.19122739515198153,
+      "grad_norm": 1.775086563902635,
+      "learning_rate": 1.8667514203170934e-05,
+      "loss": 0.9746,
+      "step": 994
+    },
+    {
+      "epoch": 0.1914197768372451,
+      "grad_norm": 2.0102308585702713,
+      "learning_rate": 1.8664404946202658e-05,
+      "loss": 0.9763,
+      "step": 995
+    },
+    {
+      "epoch": 0.19161215852250865,
+      "grad_norm": 1.520788133395406,
+      "learning_rate": 1.8661292325414058e-05,
+      "loss": 0.8959,
+      "step": 996
+    },
+    {
+      "epoch": 0.1918045402077722,
+      "grad_norm": 1.6244228830796021,
+      "learning_rate": 1.865817634201356e-05,
+      "loss": 0.9105,
+      "step": 997
+    },
+    {
+      "epoch": 0.19199692189303577,
+      "grad_norm": 1.5348198598994962,
+      "learning_rate": 1.8655056997210893e-05,
+      "loss": 0.9327,
+      "step": 998
+    },
+    {
+      "epoch": 0.19218930357829936,
+      "grad_norm": 1.3918134039935715,
+      "learning_rate": 1.8651934292217097e-05,
+      "loss": 0.9247,
+      "step": 999
+    },
+    {
+      "epoch": 0.19238168526356292,
+      "grad_norm": 1.4649600047154412,
+      "learning_rate": 1.864880822824452e-05,
+      "loss": 0.9466,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19257406694882648,
+      "grad_norm": 1.4259148029278501,
+      "learning_rate": 1.8645678806506795e-05,
+      "loss": 0.9102,
+      "step": 1001
+    },
+    {
+      "epoch": 0.19276644863409004,
+      "grad_norm": 1.7223214255125667,
+      "learning_rate": 1.864254602821888e-05,
+      "loss": 0.9398,
+      "step": 1002
+    },
+    {
+      "epoch": 0.1929588303193536,
+      "grad_norm": 1.2323275205681963,
+      "learning_rate": 1.8639409894597026e-05,
+      "loss": 0.886,
+      "step": 1003
+    },
+    {
+      "epoch": 0.19315121200461716,
+      "grad_norm": 1.724291510780539,
+      "learning_rate": 1.8636270406858786e-05,
+      "loss": 0.9271,
+      "step": 1004
+    },
+    {
+      "epoch": 0.19334359368988072,
+      "grad_norm": 1.4789951886145836,
+      "learning_rate": 1.8633127566223023e-05,
+      "loss": 0.9238,
+      "step": 1005
+    },
+    {
+      "epoch": 0.19353597537514428,
+      "grad_norm": 1.6078207258310684,
+      "learning_rate": 1.862998137390989e-05,
+      "loss": 0.8916,
+      "step": 1006
+    },
+    {
+      "epoch": 0.19372835706040784,
+      "grad_norm": 2.8296055634478696,
+      "learning_rate": 1.8626831831140845e-05,
+      "loss": 0.8669,
+      "step": 1007
+    },
+    {
+      "epoch": 0.19392073874567142,
+      "grad_norm": 2.7093621008854853,
+      "learning_rate": 1.8623678939138652e-05,
+      "loss": 0.9857,
+      "step": 1008
+    },
+    {
+      "epoch": 0.19411312043093498,
+      "grad_norm": 1.4490961236969275,
+      "learning_rate": 1.8620522699127374e-05,
+      "loss": 0.9383,
+      "step": 1009
+    },
+    {
+      "epoch": 0.19430550211619854,
+      "grad_norm": 1.2571487261806926,
+      "learning_rate": 1.8617363112332376e-05,
+      "loss": 0.9127,
+      "step": 1010
+    },
+    {
+      "epoch": 0.1944978838014621,
+      "grad_norm": 1.919749434996529,
+      "learning_rate": 1.8614200179980307e-05,
+      "loss": 0.9767,
+      "step": 1011
+    },
+    {
+      "epoch": 0.19469026548672566,
+      "grad_norm": 1.424542227158111,
+      "learning_rate": 1.8611033903299136e-05,
+      "loss": 0.9202,
+      "step": 1012
+    },
+    {
+      "epoch": 0.19488264717198922,
+      "grad_norm": 1.361258952293407,
+      "learning_rate": 1.8607864283518116e-05,
+      "loss": 0.9241,
+      "step": 1013
+    },
+    {
+      "epoch": 0.19507502885725278,
+      "grad_norm": 1.454536659395299,
+      "learning_rate": 1.8604691321867804e-05,
+      "loss": 0.8883,
+      "step": 1014
+    },
+    {
+      "epoch": 0.19526741054251634,
+      "grad_norm": 1.2797927510736071,
+      "learning_rate": 1.8601515019580053e-05,
+      "loss": 0.9217,
+      "step": 1015
+    },
+    {
+      "epoch": 0.19545979222777993,
+      "grad_norm": 1.5166520328389925,
+      "learning_rate": 1.8598335377888012e-05,
+      "loss": 0.9859,
+      "step": 1016
+    },
+    {
+      "epoch": 0.1956521739130435,
+      "grad_norm": 2.547696628986193,
+      "learning_rate": 1.8595152398026128e-05,
+      "loss": 0.8902,
+      "step": 1017
+    },
+    {
+      "epoch": 0.19584455559830705,
+      "grad_norm": 0.7500357284209249,
+      "learning_rate": 1.8591966081230142e-05,
+      "loss": 0.3166,
+      "step": 1018
+    },
+    {
+      "epoch": 0.1960369372835706,
+      "grad_norm": 0.7074381979844122,
+      "learning_rate": 1.8588776428737095e-05,
+      "loss": 0.337,
+      "step": 1019
+    },
+    {
+      "epoch": 0.19622931896883417,
+      "grad_norm": 1.249100408636832,
+      "learning_rate": 1.858558344178532e-05,
+      "loss": 0.9109,
+      "step": 1020
+    },
+    {
+      "epoch": 0.19642170065409773,
+      "grad_norm": 1.510077373056751,
+      "learning_rate": 1.8582387121614437e-05,
+      "loss": 0.9596,
+      "step": 1021
+    },
+    {
+      "epoch": 0.19661408233936128,
+      "grad_norm": 1.4873008529252274,
+      "learning_rate": 1.857918746946538e-05,
+      "loss": 0.9444,
+      "step": 1022
+    },
+    {
+      "epoch": 0.19680646402462484,
+      "grad_norm": 1.6925800916231892,
+      "learning_rate": 1.8575984486580353e-05,
+      "loss": 1.0021,
+      "step": 1023
+    },
+    {
+      "epoch": 0.19699884570988843,
+      "grad_norm": 1.422316945637719,
+      "learning_rate": 1.857277817420287e-05,
+      "loss": 0.9844,
+      "step": 1024
+    },
+    {
+      "epoch": 0.197191227395152,
+      "grad_norm": 1.5846427770290625,
+      "learning_rate": 1.8569568533577727e-05,
+      "loss": 0.9549,
+      "step": 1025
+    },
+    {
+      "epoch": 0.19738360908041555,
+      "grad_norm": 1.3135859131359509,
+      "learning_rate": 1.8566355565951023e-05,
+      "loss": 0.9848,
+      "step": 1026
+    },
+    {
+      "epoch": 0.1975759907656791,
+      "grad_norm": 5.278988174024068,
+      "learning_rate": 1.8563139272570142e-05,
+      "loss": 0.9499,
+      "step": 1027
+    },
+    {
+      "epoch": 0.19776837245094267,
+      "grad_norm": 1.5761390589150068,
+      "learning_rate": 1.8559919654683756e-05,
+      "loss": 0.9367,
+      "step": 1028
+    },
+    {
+      "epoch": 0.19796075413620623,
+      "grad_norm": 1.4195119800204359,
+      "learning_rate": 1.8556696713541833e-05,
+      "loss": 0.9525,
+      "step": 1029
+    },
+    {
+      "epoch": 0.1981531358214698,
+      "grad_norm": 1.9028056010058068,
+      "learning_rate": 1.855347045039563e-05,
+      "loss": 0.9429,
+      "step": 1030
+    },
+    {
+      "epoch": 0.19834551750673335,
+      "grad_norm": 1.2371372697145793,
+      "learning_rate": 1.8550240866497697e-05,
+      "loss": 0.9332,
+      "step": 1031
+    },
+    {
+      "epoch": 0.1985378991919969,
+      "grad_norm": 1.4888498872738378,
+      "learning_rate": 1.854700796310186e-05,
+      "loss": 0.9207,
+      "step": 1032
+    },
+    {
+      "epoch": 0.1987302808772605,
+      "grad_norm": 1.6005251012594846,
+      "learning_rate": 1.8543771741463254e-05,
+      "loss": 0.9294,
+      "step": 1033
+    },
+    {
+      "epoch": 0.19892266256252406,
+      "grad_norm": 2.004217789112375,
+      "learning_rate": 1.8540532202838286e-05,
+      "loss": 0.9653,
+      "step": 1034
+    },
+    {
+      "epoch": 0.19911504424778761,
+      "grad_norm": 1.3100253372868307,
+      "learning_rate": 1.8537289348484658e-05,
+      "loss": 0.9059,
+      "step": 1035
+    },
+    {
+      "epoch": 0.19930742593305117,
+      "grad_norm": 3.5067833994875044,
+      "learning_rate": 1.8534043179661357e-05,
+      "loss": 0.8886,
+      "step": 1036
+    },
+    {
+      "epoch": 0.19949980761831473,
+      "grad_norm": 2.010025535605026,
+      "learning_rate": 1.8530793697628658e-05,
+      "loss": 0.3633,
+      "step": 1037
+    },
+    {
+      "epoch": 0.1996921893035783,
+      "grad_norm": 1.7921992423970592,
+      "learning_rate": 1.8527540903648122e-05,
+      "loss": 0.9943,
+      "step": 1038
+    },
+    {
+      "epoch": 0.19988457098884185,
+      "grad_norm": 1.6755558142352192,
+      "learning_rate": 1.8524284798982595e-05,
+      "loss": 0.978,
+      "step": 1039
+    },
+    {
+      "epoch": 0.2000769526741054,
+      "grad_norm": 1.8103120625552271,
+      "learning_rate": 1.852102538489621e-05,
+      "loss": 0.9533,
+      "step": 1040
+    },
+    {
+      "epoch": 0.200269334359369,
+      "grad_norm": 1.2514249989107553,
+      "learning_rate": 1.8517762662654383e-05,
+      "loss": 0.9224,
+      "step": 1041
+    },
+    {
+      "epoch": 0.20046171604463256,
+      "grad_norm": 1.1746506863511406,
+      "learning_rate": 1.851449663352381e-05,
+      "loss": 0.3744,
+      "step": 1042
+    },
+    {
+      "epoch": 0.20065409772989612,
+      "grad_norm": 1.6053212718311387,
+      "learning_rate": 1.851122729877249e-05,
+      "loss": 0.8966,
+      "step": 1043
+    },
+    {
+      "epoch": 0.20084647941515968,
+      "grad_norm": 1.3514847301027242,
+      "learning_rate": 1.8507954659669677e-05,
+      "loss": 0.8617,
+      "step": 1044
+    },
+    {
+      "epoch": 0.20103886110042324,
+      "grad_norm": 1.400060494837672,
+      "learning_rate": 1.850467871748593e-05,
+      "loss": 0.9549,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2012312427856868,
+      "grad_norm": 1.374405691992397,
+      "learning_rate": 1.850139947349308e-05,
+      "loss": 0.9086,
+      "step": 1046
+    },
+    {
+      "epoch": 0.20142362447095036,
+      "grad_norm": 1.2273497198874446,
+      "learning_rate": 1.8498116928964244e-05,
+      "loss": 0.9433,
+      "step": 1047
+    },
+    {
+      "epoch": 0.20161600615621392,
+      "grad_norm": 1.4601898482323454,
+      "learning_rate": 1.849483108517381e-05,
+      "loss": 0.9265,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2018083878414775,
+      "grad_norm": 1.087842901315358,
+      "learning_rate": 1.849154194339747e-05,
+      "loss": 0.9196,
+      "step": 1049
+    },
+    {
+      "epoch": 0.20200076952674106,
+      "grad_norm": 1.3903694881779574,
+      "learning_rate": 1.8488249504912173e-05,
+      "loss": 0.9388,
+      "step": 1050
+    },
+    {
+      "epoch": 0.20219315121200462,
+      "grad_norm": 2.4395364364793326,
+      "learning_rate": 1.8484953770996163e-05,
+      "loss": 0.917,
+      "step": 1051
+    },
+    {
+      "epoch": 0.20238553289726818,
+      "grad_norm": 1.4589321824399653,
+      "learning_rate": 1.848165474292895e-05,
+      "loss": 0.8711,
+      "step": 1052
+    },
+    {
+      "epoch": 0.20257791458253174,
+      "grad_norm": 1.4489690956431909,
+      "learning_rate": 1.8478352421991334e-05,
+      "loss": 0.9076,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2027702962677953,
+      "grad_norm": 1.292801918566612,
+      "learning_rate": 1.847504680946539e-05,
+      "loss": 1.0156,
+      "step": 1054
+    },
+    {
+      "epoch": 0.20296267795305886,
+      "grad_norm": 1.2970977390302652,
+      "learning_rate": 1.847173790663447e-05,
+      "loss": 0.9246,
+      "step": 1055
+    },
+    {
+      "epoch": 0.20315505963832242,
+      "grad_norm": 2.09691533161905,
+      "learning_rate": 1.8468425714783206e-05,
+      "loss": 0.9121,
+      "step": 1056
+    },
+    {
+      "epoch": 0.20334744132358598,
+      "grad_norm": 1.2971114934183277,
+      "learning_rate": 1.84651102351975e-05,
+      "loss": 0.9459,
+      "step": 1057
+    },
+    {
+      "epoch": 0.20353982300884957,
+      "grad_norm": 1.2811844633652112,
+      "learning_rate": 1.846179146916454e-05,
+      "loss": 0.94,
+      "step": 1058
+    },
+    {
+      "epoch": 0.20373220469411313,
+      "grad_norm": 1.2996477751687816,
+      "learning_rate": 1.8458469417972783e-05,
+      "loss": 0.9156,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2039245863793767,
+      "grad_norm": 0.933798785029116,
+      "learning_rate": 1.8455144082911965e-05,
+      "loss": 0.3567,
+      "step": 1060
+    },
+    {
+      "epoch": 0.20411696806464025,
+      "grad_norm": 1.3458001627499392,
+      "learning_rate": 1.8451815465273097e-05,
+      "loss": 0.9045,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2043093497499038,
+      "grad_norm": 1.3737659564710107,
+      "learning_rate": 1.8448483566348456e-05,
+      "loss": 0.9399,
+      "step": 1062
+    },
+    {
+      "epoch": 0.20450173143516737,
+      "grad_norm": 1.2293144416085717,
+      "learning_rate": 1.8445148387431605e-05,
+      "loss": 0.8989,
+      "step": 1063
+    },
+    {
+      "epoch": 0.20469411312043093,
+      "grad_norm": 4.302380218689566,
+      "learning_rate": 1.8441809929817382e-05,
+      "loss": 0.9384,
+      "step": 1064
+    },
+    {
+      "epoch": 0.20488649480569449,
+      "grad_norm": 1.1740666480547899,
+      "learning_rate": 1.8438468194801876e-05,
+      "loss": 0.9501,
+      "step": 1065
+    },
+    {
+      "epoch": 0.20507887649095807,
+      "grad_norm": 1.350961990740879,
+      "learning_rate": 1.8435123183682475e-05,
+      "loss": 0.9089,
+      "step": 1066
+    },
+    {
+      "epoch": 0.20527125817622163,
+      "grad_norm": 1.4627569500005748,
+      "learning_rate": 1.8431774897757824e-05,
+      "loss": 0.8601,
+      "step": 1067
+    },
+    {
+      "epoch": 0.2054636398614852,
+      "grad_norm": 1.9843927999159918,
+      "learning_rate": 1.8428423338327847e-05,
+      "loss": 0.9152,
+      "step": 1068
+    },
+    {
+      "epoch": 0.20565602154674875,
+      "grad_norm": 1.311965827502017,
+      "learning_rate": 1.8425068506693727e-05,
+      "loss": 0.9865,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2058484032320123,
+      "grad_norm": 2.523246474624472,
+      "learning_rate": 1.842171040415793e-05,
+      "loss": 0.8808,
+      "step": 1070
+    },
+    {
+      "epoch": 0.20604078491727587,
+      "grad_norm": 1.3246553857869892,
+      "learning_rate": 1.8418349032024185e-05,
+      "loss": 0.9185,
+      "step": 1071
+    },
+    {
+      "epoch": 0.20623316660253943,
+      "grad_norm": 1.3506880308754432,
+      "learning_rate": 1.8414984391597492e-05,
+      "loss": 0.98,
+      "step": 1072
+    },
+    {
+      "epoch": 0.206425548287803,
+      "grad_norm": 1.3038727909642922,
+      "learning_rate": 1.8411616484184126e-05,
+      "loss": 0.9172,
+      "step": 1073
+    },
+    {
+      "epoch": 0.20661792997306658,
+      "grad_norm": 1.290225647019821,
+      "learning_rate": 1.8408245311091618e-05,
+      "loss": 0.9149,
+      "step": 1074
+    },
+    {
+      "epoch": 0.20681031165833014,
+      "grad_norm": 1.2073437191552778,
+      "learning_rate": 1.8404870873628774e-05,
+      "loss": 0.8656,
+      "step": 1075
+    },
+    {
+      "epoch": 0.2070026933435937,
+      "grad_norm": 1.3970876814648214,
+      "learning_rate": 1.8401493173105675e-05,
+      "loss": 0.967,
+      "step": 1076
+    },
+    {
+      "epoch": 0.20719507502885726,
+      "grad_norm": 1.5830792387564114,
+      "learning_rate": 1.8398112210833648e-05,
+      "loss": 0.8886,
+      "step": 1077
+    },
+    {
+      "epoch": 0.20738745671412082,
+      "grad_norm": 1.454108438616329,
+      "learning_rate": 1.8394727988125308e-05,
+      "loss": 0.9048,
+      "step": 1078
+    },
+    {
+      "epoch": 0.20757983839938438,
+      "grad_norm": 1.3187536931899475,
+      "learning_rate": 1.8391340506294524e-05,
+      "loss": 0.8948,
+      "step": 1079
+    },
+    {
+      "epoch": 0.20777222008464794,
+      "grad_norm": 1.3365705224386186,
+      "learning_rate": 1.8387949766656434e-05,
+      "loss": 0.8753,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2079646017699115,
+      "grad_norm": 1.3380467455519764,
+      "learning_rate": 1.8384555770527438e-05,
+      "loss": 0.9374,
+      "step": 1081
+    },
+    {
+      "epoch": 0.20815698345517505,
+      "grad_norm": 1.39124111426851,
+      "learning_rate": 1.8381158519225204e-05,
+      "loss": 0.9309,
+      "step": 1082
+    },
+    {
+      "epoch": 0.20834936514043864,
+      "grad_norm": 5.576799111678552,
+      "learning_rate": 1.8377758014068662e-05,
+      "loss": 0.9339,
+      "step": 1083
+    },
+    {
+      "epoch": 0.2085417468257022,
+      "grad_norm": 1.3618563689441945,
+      "learning_rate": 1.8374354256378e-05,
+      "loss": 0.9413,
+      "step": 1084
+    },
+    {
+      "epoch": 0.20873412851096576,
+      "grad_norm": 0.7147114323026161,
+      "learning_rate": 1.837094724747468e-05,
+      "loss": 0.374,
+      "step": 1085
+    },
+    {
+      "epoch": 0.20892651019622932,
+      "grad_norm": 1.5022874183839172,
+      "learning_rate": 1.8367536988681422e-05,
+      "loss": 0.8895,
+      "step": 1086
+    },
+    {
+      "epoch": 0.20911889188149288,
+      "grad_norm": 1.4747293632654443,
+      "learning_rate": 1.83641234813222e-05,
+      "loss": 0.9716,
+      "step": 1087
+    },
+    {
+      "epoch": 0.20931127356675644,
+      "grad_norm": 1.2207536613191994,
+      "learning_rate": 1.8360706726722253e-05,
+      "loss": 0.9124,
+      "step": 1088
+    },
+    {
+      "epoch": 0.20950365525202,
+      "grad_norm": 1.6160466286598982,
+      "learning_rate": 1.835728672620809e-05,
+      "loss": 0.9101,
+      "step": 1089
+    },
+    {
+      "epoch": 0.20969603693728356,
+      "grad_norm": 2.370371010718713,
+      "learning_rate": 1.8353863481107473e-05,
+      "loss": 0.9953,
+      "step": 1090
+    },
+    {
+      "epoch": 0.20988841862254715,
+      "grad_norm": 1.751185535164767,
+      "learning_rate": 1.835043699274942e-05,
+      "loss": 0.9033,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2100808003078107,
+      "grad_norm": 1.4900833196076224,
+      "learning_rate": 1.8347007262464206e-05,
+      "loss": 0.9256,
+      "step": 1092
+    },
+    {
+      "epoch": 0.21027318199307427,
+      "grad_norm": 1.8030381360985517,
+      "learning_rate": 1.8343574291583385e-05,
+      "loss": 0.9227,
+      "step": 1093
+    },
+    {
+      "epoch": 0.21046556367833782,
+      "grad_norm": 1.5574702099645608,
+      "learning_rate": 1.8340138081439743e-05,
+      "loss": 0.9085,
+      "step": 1094
+    },
+    {
+      "epoch": 0.21065794536360138,
+      "grad_norm": 1.395985735092629,
+      "learning_rate": 1.833669863336734e-05,
+      "loss": 0.9302,
+      "step": 1095
+    },
+    {
+      "epoch": 0.21085032704886494,
+      "grad_norm": 1.2744592674335717,
+      "learning_rate": 1.833325594870148e-05,
+      "loss": 0.8941,
+      "step": 1096
+    },
+    {
+      "epoch": 0.2110427087341285,
+      "grad_norm": 1.7708945056281744,
+      "learning_rate": 1.8329810028778747e-05,
+      "loss": 0.9715,
+      "step": 1097
+    },
+    {
+      "epoch": 0.21123509041939206,
+      "grad_norm": 1.3661068630879194,
+      "learning_rate": 1.8326360874936952e-05,
+      "loss": 0.9473,
+      "step": 1098
+    },
+    {
+      "epoch": 0.21142747210465565,
+      "grad_norm": 1.6248068577082324,
+      "learning_rate": 1.8322908488515182e-05,
+      "loss": 0.9288,
+      "step": 1099
+    },
+    {
+      "epoch": 0.2116198537899192,
+      "grad_norm": 2.651675520606779,
+      "learning_rate": 1.8319452870853772e-05,
+      "loss": 0.9798,
+      "step": 1100
+    },
+    {
+      "epoch": 0.21181223547518277,
+      "grad_norm": 1.4279318847224438,
+      "learning_rate": 1.8315994023294306e-05,
+      "loss": 0.9666,
+      "step": 1101
+    },
+    {
+      "epoch": 0.21200461716044633,
+      "grad_norm": 1.194866418283679,
+      "learning_rate": 1.8312531947179634e-05,
+      "loss": 0.927,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2121969988457099,
+      "grad_norm": 1.674210583678688,
+      "learning_rate": 1.8309066643853854e-05,
+      "loss": 0.964,
+      "step": 1103
+    },
+    {
+      "epoch": 0.21238938053097345,
+      "grad_norm": 1.228358816783259,
+      "learning_rate": 1.8305598114662312e-05,
+      "loss": 0.9311,
+      "step": 1104
+    },
+    {
+      "epoch": 0.212581762216237,
+      "grad_norm": 1.876831640940696,
+      "learning_rate": 1.830212636095161e-05,
+      "loss": 0.9582,
+      "step": 1105
+    },
+    {
+      "epoch": 0.21277414390150057,
+      "grad_norm": 2.3628373066353956,
+      "learning_rate": 1.8298651384069605e-05,
+      "loss": 0.9105,
+      "step": 1106
+    },
+    {
+      "epoch": 0.21296652558676413,
+      "grad_norm": 0.6748134035711201,
+      "learning_rate": 1.8295173185365405e-05,
+      "loss": 0.3391,
+      "step": 1107
+    },
+    {
+      "epoch": 0.21315890727202771,
+      "grad_norm": 3.5051086556366604,
+      "learning_rate": 1.829169176618936e-05,
+      "loss": 0.9308,
+      "step": 1108
+    },
+    {
+      "epoch": 0.21335128895729127,
+      "grad_norm": 1.332353748427702,
+      "learning_rate": 1.828820712789308e-05,
+      "loss": 0.9034,
+      "step": 1109
+    },
+    {
+      "epoch": 0.21354367064255483,
+      "grad_norm": 1.3114607268008769,
+      "learning_rate": 1.828471927182942e-05,
+      "loss": 0.8923,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2137360523278184,
+      "grad_norm": 1.6261084893389817,
+      "learning_rate": 1.828122819935249e-05,
+      "loss": 0.916,
+      "step": 1111
+    },
+    {
+      "epoch": 0.21392843401308195,
+      "grad_norm": 1.3257440443494428,
+      "learning_rate": 1.8277733911817642e-05,
+      "loss": 0.8731,
+      "step": 1112
+    },
+    {
+      "epoch": 0.2141208156983455,
+      "grad_norm": 1.3073986535004594,
+      "learning_rate": 1.8274236410581478e-05,
+      "loss": 0.8581,
+      "step": 1113
+    },
+    {
+      "epoch": 0.21431319738360907,
+      "grad_norm": 1.9386419565725066,
+      "learning_rate": 1.827073569700185e-05,
+      "loss": 0.9581,
+      "step": 1114
+    },
+    {
+      "epoch": 0.21450557906887263,
+      "grad_norm": 1.8576748797536486,
+      "learning_rate": 1.8267231772437854e-05,
+      "loss": 0.9344,
+      "step": 1115
+    },
+    {
+      "epoch": 0.21469796075413622,
+      "grad_norm": 1.8346922054925239,
+      "learning_rate": 1.8263724638249834e-05,
+      "loss": 0.9806,
+      "step": 1116
+    },
+    {
+      "epoch": 0.21489034243939978,
+      "grad_norm": 1.714885672206539,
+      "learning_rate": 1.8260214295799382e-05,
+      "loss": 0.8984,
+      "step": 1117
+    },
+    {
+      "epoch": 0.21508272412466334,
+      "grad_norm": 1.4220478851657945,
+      "learning_rate": 1.825670074644933e-05,
+      "loss": 0.9249,
+      "step": 1118
+    },
+    {
+      "epoch": 0.2152751058099269,
+      "grad_norm": 0.6765836836296114,
+      "learning_rate": 1.8253183991563768e-05,
+      "loss": 0.3482,
+      "step": 1119
+    },
+    {
+      "epoch": 0.21546748749519046,
+      "grad_norm": 1.3319743544415512,
+      "learning_rate": 1.824966403250801e-05,
+      "loss": 0.9256,
+      "step": 1120
+    },
+    {
+      "epoch": 0.21565986918045402,
+      "grad_norm": 1.8841846997610145,
+      "learning_rate": 1.8246140870648633e-05,
+      "loss": 0.9676,
+      "step": 1121
+    },
+    {
+      "epoch": 0.21585225086571758,
+      "grad_norm": 0.6216095252539734,
+      "learning_rate": 1.8242614507353446e-05,
+      "loss": 0.3656,
+      "step": 1122
+    },
+    {
+      "epoch": 0.21604463255098114,
+      "grad_norm": 1.2184491543102556,
+      "learning_rate": 1.8239084943991507e-05,
+      "loss": 0.9129,
+      "step": 1123
+    },
+    {
+      "epoch": 0.2162370142362447,
+      "grad_norm": 1.475685808997508,
+      "learning_rate": 1.823555218193311e-05,
+      "loss": 0.8836,
+      "step": 1124
+    },
+    {
+      "epoch": 0.21642939592150828,
+      "grad_norm": 1.6319915321219953,
+      "learning_rate": 1.8232016222549797e-05,
+      "loss": 0.9548,
+      "step": 1125
+    },
+    {
+      "epoch": 0.21662177760677184,
+      "grad_norm": 0.6462189388295263,
+      "learning_rate": 1.8228477067214352e-05,
+      "loss": 0.3883,
+      "step": 1126
+    },
+    {
+      "epoch": 0.2168141592920354,
+      "grad_norm": 1.4428514156777572,
+      "learning_rate": 1.8224934717300794e-05,
+      "loss": 0.905,
+      "step": 1127
+    },
+    {
+      "epoch": 0.21700654097729896,
+      "grad_norm": 1.1954999059660618,
+      "learning_rate": 1.8221389174184385e-05,
+      "loss": 0.9313,
+      "step": 1128
+    },
+    {
+      "epoch": 0.21719892266256252,
+      "grad_norm": 1.4206305994535442,
+      "learning_rate": 1.8217840439241633e-05,
+      "loss": 0.9004,
+      "step": 1129
+    },
+    {
+      "epoch": 0.21739130434782608,
+      "grad_norm": 1.348187898406821,
+      "learning_rate": 1.8214288513850267e-05,
+      "loss": 0.9348,
+      "step": 1130
+    },
+    {
+      "epoch": 0.21758368603308964,
+      "grad_norm": 1.660212074355041,
+      "learning_rate": 1.8210733399389277e-05,
+      "loss": 0.8833,
+      "step": 1131
+    },
+    {
+      "epoch": 0.2177760677183532,
+      "grad_norm": 1.2691569614874127,
+      "learning_rate": 1.820717509723888e-05,
+      "loss": 0.8766,
+      "step": 1132
+    },
+    {
+      "epoch": 0.2179684494036168,
+      "grad_norm": 1.322716062821615,
+      "learning_rate": 1.8203613608780525e-05,
+      "loss": 0.9326,
+      "step": 1133
+    },
+    {
+      "epoch": 0.21816083108888035,
+      "grad_norm": 1.5071426978498794,
+      "learning_rate": 1.8200048935396908e-05,
+      "loss": 0.9639,
+      "step": 1134
+    },
+    {
+      "epoch": 0.2183532127741439,
+      "grad_norm": 1.4283042347217678,
+      "learning_rate": 1.819648107847196e-05,
+      "loss": 0.899,
+      "step": 1135
+    },
+    {
+      "epoch": 0.21854559445940747,
+      "grad_norm": 1.4290228784421357,
+      "learning_rate": 1.8192910039390844e-05,
+      "loss": 0.9217,
+      "step": 1136
+    },
+    {
+      "epoch": 0.21873797614467103,
+      "grad_norm": 2.86226953360701,
+      "learning_rate": 1.8189335819539963e-05,
+      "loss": 0.9687,
+      "step": 1137
+    },
+    {
+      "epoch": 0.21893035782993459,
+      "grad_norm": 1.2761617408871024,
+      "learning_rate": 1.8185758420306947e-05,
+      "loss": 0.8558,
+      "step": 1138
+    },
+    {
+      "epoch": 0.21912273951519814,
+      "grad_norm": 1.30060170914067,
+      "learning_rate": 1.818217784308067e-05,
+      "loss": 0.9551,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2193151212004617,
+      "grad_norm": 0.6457349339455908,
+      "learning_rate": 1.817859408925123e-05,
+      "loss": 0.3507,
+      "step": 1140
+    },
+    {
+      "epoch": 0.2195075028857253,
+      "grad_norm": 1.4666117645010999,
+      "learning_rate": 1.817500716020997e-05,
+      "loss": 0.9248,
+      "step": 1141
+    },
+    {
+      "epoch": 0.21969988457098885,
+      "grad_norm": 1.3993239194166667,
+      "learning_rate": 1.8171417057349457e-05,
+      "loss": 0.8652,
+      "step": 1142
+    },
+    {
+      "epoch": 0.2198922662562524,
+      "grad_norm": 2.928978699932161,
+      "learning_rate": 1.816782378206349e-05,
+      "loss": 0.9418,
+      "step": 1143
+    },
+    {
+      "epoch": 0.22008464794151597,
+      "grad_norm": 1.354523054702284,
+      "learning_rate": 1.8164227335747108e-05,
+      "loss": 1.0106,
+      "step": 1144
+    },
+    {
+      "epoch": 0.22027702962677953,
+      "grad_norm": 1.7861113865763831,
+      "learning_rate": 1.8160627719796568e-05,
+      "loss": 0.9824,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2204694113120431,
+      "grad_norm": 1.6545161876269119,
+      "learning_rate": 1.815702493560937e-05,
+      "loss": 0.9835,
+      "step": 1146
+    },
+    {
+      "epoch": 0.22066179299730665,
+      "grad_norm": 3.0966897488685814,
+      "learning_rate": 1.8153418984584238e-05,
+      "loss": 0.9436,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2208541746825702,
+      "grad_norm": 1.4437914410183852,
+      "learning_rate": 1.8149809868121125e-05,
+      "loss": 0.9359,
+      "step": 1148
+    },
+    {
+      "epoch": 0.22104655636783377,
+      "grad_norm": 1.608795419672404,
+      "learning_rate": 1.8146197587621217e-05,
+      "loss": 0.902,
+      "step": 1149
+    },
+    {
+      "epoch": 0.22123893805309736,
+      "grad_norm": 1.4321855057952109,
+      "learning_rate": 1.814258214448692e-05,
+      "loss": 0.942,
+      "step": 1150
+    },
+    {
+      "epoch": 0.22143131973836092,
+      "grad_norm": 1.3226786340753092,
+      "learning_rate": 1.8138963540121878e-05,
+      "loss": 0.931,
+      "step": 1151
+    },
+    {
+      "epoch": 0.22162370142362448,
+      "grad_norm": 2.8577412018491213,
+      "learning_rate": 1.813534177593096e-05,
+      "loss": 0.8938,
+      "step": 1152
+    },
+    {
+      "epoch": 0.22181608310888803,
+      "grad_norm": 1.3374221088581184,
+      "learning_rate": 1.8131716853320254e-05,
+      "loss": 0.9169,
+      "step": 1153
+    },
+    {
+      "epoch": 0.2220084647941516,
+      "grad_norm": 1.5369261538026955,
+      "learning_rate": 1.8128088773697086e-05,
+      "loss": 0.8936,
+      "step": 1154
+    },
+    {
+      "epoch": 0.22220084647941515,
+      "grad_norm": 1.601617269193923,
+      "learning_rate": 1.8124457538469996e-05,
+      "loss": 0.9505,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2223932281646787,
+      "grad_norm": 1.4685525312909997,
+      "learning_rate": 1.8120823149048753e-05,
+      "loss": 0.8817,
+      "step": 1156
+    },
+    {
+      "epoch": 0.22258560984994227,
+      "grad_norm": 1.3723236109766423,
+      "learning_rate": 1.811718560684436e-05,
+      "loss": 0.9579,
+      "step": 1157
+    },
+    {
+      "epoch": 0.22277799153520586,
+      "grad_norm": 0.6610736935065331,
+      "learning_rate": 1.8113544913269025e-05,
+      "loss": 0.3602,
+      "step": 1158
+    },
+    {
+      "epoch": 0.22297037322046942,
+      "grad_norm": 1.261935498999892,
+      "learning_rate": 1.8109901069736202e-05,
+      "loss": 0.9232,
+      "step": 1159
+    },
+    {
+      "epoch": 0.22316275490573298,
+      "grad_norm": 1.3627691428783215,
+      "learning_rate": 1.8106254077660552e-05,
+      "loss": 0.908,
+      "step": 1160
+    },
+    {
+      "epoch": 0.22335513659099654,
+      "grad_norm": 1.5065260282960622,
+      "learning_rate": 1.810260393845796e-05,
+      "loss": 0.9218,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2235475182762601,
+      "grad_norm": 1.5708510942669078,
+      "learning_rate": 1.809895065354554e-05,
+      "loss": 0.9259,
+      "step": 1162
+    },
+    {
+      "epoch": 0.22373989996152366,
+      "grad_norm": 1.5772565297296453,
+      "learning_rate": 1.8095294224341622e-05,
+      "loss": 0.8951,
+      "step": 1163
+    },
+    {
+      "epoch": 0.22393228164678722,
+      "grad_norm": 1.5123018224284166,
+      "learning_rate": 1.8091634652265755e-05,
+      "loss": 0.9537,
+      "step": 1164
+    },
+    {
+      "epoch": 0.22412466333205078,
+      "grad_norm": 1.4078791330428047,
+      "learning_rate": 1.8087971938738715e-05,
+      "loss": 0.9123,
+      "step": 1165
+    },
+    {
+      "epoch": 0.22431704501731436,
+      "grad_norm": 1.231709323297761,
+      "learning_rate": 1.808430608518249e-05,
+      "loss": 0.9091,
+      "step": 1166
+    },
+    {
+      "epoch": 0.22450942670257792,
+      "grad_norm": 1.4900137214413556,
+      "learning_rate": 1.808063709302029e-05,
+      "loss": 0.9182,
+      "step": 1167
+    },
+    {
+      "epoch": 0.22470180838784148,
+      "grad_norm": 1.59506036029931,
+      "learning_rate": 1.807696496367655e-05,
+      "loss": 0.9225,
+      "step": 1168
+    },
+    {
+      "epoch": 0.22489419007310504,
+      "grad_norm": 1.244775707750116,
+      "learning_rate": 1.8073289698576913e-05,
+      "loss": 0.9093,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2250865717583686,
+      "grad_norm": 4.954434955746153,
+      "learning_rate": 1.8069611299148236e-05,
+      "loss": 0.9185,
+      "step": 1170
+    },
+    {
+      "epoch": 0.22527895344363216,
+      "grad_norm": 1.4739019405471876,
+      "learning_rate": 1.8065929766818617e-05,
+      "loss": 0.8889,
+      "step": 1171
+    },
+    {
+      "epoch": 0.22547133512889572,
+      "grad_norm": 0.6439791865350821,
+      "learning_rate": 1.806224510301734e-05,
+      "loss": 0.3215,
+      "step": 1172
+    },
+    {
+      "epoch": 0.22566371681415928,
+      "grad_norm": 1.6107194467528712,
+      "learning_rate": 1.8058557309174926e-05,
+      "loss": 0.9344,
+      "step": 1173
+    },
+    {
+      "epoch": 0.22585609849942284,
+      "grad_norm": 2.3340891313681436,
+      "learning_rate": 1.8054866386723096e-05,
+      "loss": 0.8954,
+      "step": 1174
+    },
+    {
+      "epoch": 0.22604848018468643,
+      "grad_norm": 1.4126793124084362,
+      "learning_rate": 1.80511723370948e-05,
+      "loss": 0.9471,
+      "step": 1175
+    },
+    {
+      "epoch": 0.22624086186995,
+      "grad_norm": 3.008222655954071,
+      "learning_rate": 1.804747516172419e-05,
+      "loss": 0.9804,
+      "step": 1176
+    },
+    {
+      "epoch": 0.22643324355521355,
+      "grad_norm": 1.2669056193459258,
+      "learning_rate": 1.8043774862046644e-05,
+      "loss": 0.8707,
+      "step": 1177
+    },
+    {
+      "epoch": 0.2266256252404771,
+      "grad_norm": 1.4304318435140437,
+      "learning_rate": 1.804007143949874e-05,
+      "loss": 0.9153,
+      "step": 1178
+    },
+    {
+      "epoch": 0.22681800692574067,
+      "grad_norm": 1.5055162251307106,
+      "learning_rate": 1.8036364895518272e-05,
+      "loss": 0.8729,
+      "step": 1179
+    },
+    {
+      "epoch": 0.22701038861100423,
+      "grad_norm": 1.4815691916071023,
+      "learning_rate": 1.8032655231544253e-05,
+      "loss": 0.9,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2272027702962678,
+      "grad_norm": 2.008922104536506,
+      "learning_rate": 1.8028942449016903e-05,
+      "loss": 0.9447,
+      "step": 1181
+    },
+    {
+      "epoch": 0.22739515198153135,
+      "grad_norm": 1.8841327222854816,
+      "learning_rate": 1.8025226549377647e-05,
+      "loss": 0.9109,
+      "step": 1182
+    },
+    {
+      "epoch": 0.22758753366679493,
+      "grad_norm": 1.395942085814252,
+      "learning_rate": 1.8021507534069133e-05,
+      "loss": 0.9647,
+      "step": 1183
+    },
+    {
+      "epoch": 0.2277799153520585,
+      "grad_norm": 2.33028191982058,
+      "learning_rate": 1.8017785404535198e-05,
+      "loss": 0.8771,
+      "step": 1184
+    },
+    {
+      "epoch": 0.22797229703732205,
+      "grad_norm": 1.6366392671505077,
+      "learning_rate": 1.8014060162220916e-05,
+      "loss": 0.9609,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2281646787225856,
+      "grad_norm": 1.4094125148336936,
+      "learning_rate": 1.801033180857254e-05,
+      "loss": 0.9105,
+      "step": 1186
+    },
+    {
+      "epoch": 0.22835706040784917,
+      "grad_norm": 1.4142571186653174,
+      "learning_rate": 1.8006600345037558e-05,
+      "loss": 0.939,
+      "step": 1187
+    },
+    {
+      "epoch": 0.22854944209311273,
+      "grad_norm": 1.371751902881487,
+      "learning_rate": 1.8002865773064644e-05,
+      "loss": 0.9166,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2287418237783763,
+      "grad_norm": 1.4812488504604657,
+      "learning_rate": 1.799912809410369e-05,
+      "loss": 0.9394,
+      "step": 1189
+    },
+    {
+      "epoch": 0.22893420546363985,
+      "grad_norm": 1.3942247854945535,
+      "learning_rate": 1.799538730960579e-05,
+      "loss": 0.9238,
+      "step": 1190
+    },
+    {
+      "epoch": 0.22912658714890344,
+      "grad_norm": 1.4487514923889637,
+      "learning_rate": 1.799164342102325e-05,
+      "loss": 0.9249,
+      "step": 1191
+    },
+    {
+      "epoch": 0.229318968834167,
+      "grad_norm": 1.4178562798772794,
+      "learning_rate": 1.7987896429809573e-05,
+      "loss": 0.8947,
+      "step": 1192
+    },
+    {
+      "epoch": 0.22951135051943056,
+      "grad_norm": 0.7255698841819923,
+      "learning_rate": 1.798414633741947e-05,
+      "loss": 0.3443,
+      "step": 1193
+    },
+    {
+      "epoch": 0.22970373220469412,
+      "grad_norm": 1.5982228136365777,
+      "learning_rate": 1.7980393145308857e-05,
+      "loss": 0.9013,
+      "step": 1194
+    },
+    {
+      "epoch": 0.22989611388995768,
+      "grad_norm": 1.1395409338075007,
+      "learning_rate": 1.797663685493485e-05,
+      "loss": 0.9505,
+      "step": 1195
+    },
+    {
+      "epoch": 0.23008849557522124,
+      "grad_norm": 1.4174444013212326,
+      "learning_rate": 1.7972877467755777e-05,
+      "loss": 0.9406,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2302808772604848,
+      "grad_norm": 1.1955953395852041,
+      "learning_rate": 1.7969114985231152e-05,
+      "loss": 0.8933,
+      "step": 1197
+    },
+    {
+      "epoch": 0.23047325894574835,
+      "grad_norm": 2.245673461049665,
+      "learning_rate": 1.796534940882171e-05,
+      "loss": 0.9613,
+      "step": 1198
+    },
+    {
+      "epoch": 0.23066564063101191,
+      "grad_norm": 1.1677524036337534,
+      "learning_rate": 1.7961580739989365e-05,
+      "loss": 0.8299,
+      "step": 1199
+    },
+    {
+      "epoch": 0.2308580223162755,
+      "grad_norm": 1.3666018905367783,
+      "learning_rate": 1.795780898019726e-05,
+      "loss": 0.9326,
+      "step": 1200
+    },
+    {
+      "epoch": 0.23105040400153906,
+      "grad_norm": 1.450349038175281,
+      "learning_rate": 1.795403413090971e-05,
+      "loss": 0.9373,
+      "step": 1201
+    },
+    {
+      "epoch": 0.23124278568680262,
+      "grad_norm": 1.6006425793684438,
+      "learning_rate": 1.7950256193592243e-05,
+      "loss": 0.8539,
+      "step": 1202
+    },
+    {
+      "epoch": 0.23143516737206618,
+      "grad_norm": 1.3829513654005006,
+      "learning_rate": 1.794647516971159e-05,
+      "loss": 0.9087,
+      "step": 1203
+    },
+    {
+      "epoch": 0.23162754905732974,
+      "grad_norm": 1.431814873579557,
+      "learning_rate": 1.7942691060735666e-05,
+      "loss": 0.878,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2318199307425933,
+      "grad_norm": 1.5551297535584998,
+      "learning_rate": 1.79389038681336e-05,
+      "loss": 0.9129,
+      "step": 1205
+    },
+    {
+      "epoch": 0.23201231242785686,
+      "grad_norm": 1.6828811968511708,
+      "learning_rate": 1.7935113593375707e-05,
+      "loss": 0.8897,
+      "step": 1206
+    },
+    {
+      "epoch": 0.23220469411312042,
+      "grad_norm": 1.392016281421234,
+      "learning_rate": 1.7931320237933503e-05,
+      "loss": 0.9037,
+      "step": 1207
+    },
+    {
+      "epoch": 0.232397075798384,
+      "grad_norm": 1.2625211523553166,
+      "learning_rate": 1.79275238032797e-05,
+      "loss": 0.936,
+      "step": 1208
+    },
+    {
+      "epoch": 0.23258945748364757,
+      "grad_norm": 1.489308941133539,
+      "learning_rate": 1.7923724290888205e-05,
+      "loss": 0.9107,
+      "step": 1209
+    },
+    {
+      "epoch": 0.23278183916891113,
+      "grad_norm": 1.5629446058432255,
+      "learning_rate": 1.791992170223412e-05,
+      "loss": 0.9335,
+      "step": 1210
+    },
+    {
+      "epoch": 0.23297422085417469,
+      "grad_norm": 1.4006743414108902,
+      "learning_rate": 1.791611603879374e-05,
+      "loss": 0.9188,
+      "step": 1211
+    },
+    {
+      "epoch": 0.23316660253943824,
+      "grad_norm": 2.1328017380935416,
+      "learning_rate": 1.791230730204455e-05,
+      "loss": 0.9412,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2333589842247018,
+      "grad_norm": 1.311529852241483,
+      "learning_rate": 1.7908495493465236e-05,
+      "loss": 0.9163,
+      "step": 1213
+    },
+    {
+      "epoch": 0.23355136590996536,
+      "grad_norm": 1.1392715735777932,
+      "learning_rate": 1.7904680614535675e-05,
+      "loss": 0.9738,
+      "step": 1214
+    },
+    {
+      "epoch": 0.23374374759522892,
+      "grad_norm": 1.3395939159764316,
+      "learning_rate": 1.7900862666736935e-05,
+      "loss": 0.8835,
+      "step": 1215
+    },
+    {
+      "epoch": 0.2339361292804925,
+      "grad_norm": 1.5623927304121163,
+      "learning_rate": 1.789704165155127e-05,
+      "loss": 0.907,
+      "step": 1216
+    },
+    {
+      "epoch": 0.23412851096575607,
+      "grad_norm": 1.5338738011564015,
+      "learning_rate": 1.7893217570462134e-05,
+      "loss": 0.8905,
+      "step": 1217
+    },
+    {
+      "epoch": 0.23432089265101963,
+      "grad_norm": 1.4045961300843826,
+      "learning_rate": 1.7889390424954168e-05,
+      "loss": 0.9019,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2345132743362832,
+      "grad_norm": 1.1899113590417805,
+      "learning_rate": 1.78855602165132e-05,
+      "loss": 0.8893,
+      "step": 1219
+    },
+    {
+      "epoch": 0.23470565602154675,
+      "grad_norm": 1.6092922447638809,
+      "learning_rate": 1.7881726946626244e-05,
+      "loss": 0.9318,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2348980377068103,
+      "grad_norm": 1.6829474793540162,
+      "learning_rate": 1.787789061678151e-05,
+      "loss": 0.9196,
+      "step": 1221
+    },
+    {
+      "epoch": 0.23509041939207387,
+      "grad_norm": 1.2491233213628865,
+      "learning_rate": 1.78740512284684e-05,
+      "loss": 0.9364,
+      "step": 1222
+    },
+    {
+      "epoch": 0.23528280107733743,
+      "grad_norm": 1.5938014658972706,
+      "learning_rate": 1.787020878317749e-05,
+      "loss": 0.9234,
+      "step": 1223
+    },
+    {
+      "epoch": 0.235475182762601,
+      "grad_norm": 1.2253320965910257,
+      "learning_rate": 1.7866363282400555e-05,
+      "loss": 0.9039,
+      "step": 1224
+    },
+    {
+      "epoch": 0.23566756444786457,
+      "grad_norm": 1.5128276763062911,
+      "learning_rate": 1.7862514727630543e-05,
+      "loss": 0.9193,
+      "step": 1225
+    },
+    {
+      "epoch": 0.23585994613312813,
+      "grad_norm": 1.368942671830943,
+      "learning_rate": 1.7858663120361597e-05,
+      "loss": 0.9495,
+      "step": 1226
+    },
+    {
+      "epoch": 0.2360523278183917,
+      "grad_norm": 1.439455666765798,
+      "learning_rate": 1.785480846208905e-05,
+      "loss": 0.9528,
+      "step": 1227
+    },
+    {
+      "epoch": 0.23624470950365525,
+      "grad_norm": 1.4436470770631344,
+      "learning_rate": 1.7850950754309405e-05,
+      "loss": 0.8717,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2364370911889188,
+      "grad_norm": 2.4940187135404264,
+      "learning_rate": 1.7847089998520365e-05,
+      "loss": 0.8983,
+      "step": 1229
+    },
+    {
+      "epoch": 0.23662947287418237,
+      "grad_norm": 1.4423931469992612,
+      "learning_rate": 1.7843226196220803e-05,
+      "loss": 0.9739,
+      "step": 1230
+    },
+    {
+      "epoch": 0.23682185455944593,
+      "grad_norm": 1.2098626095747729,
+      "learning_rate": 1.783935934891078e-05,
+      "loss": 0.893,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2370142362447095,
+      "grad_norm": 1.07345755917059,
+      "learning_rate": 1.7835489458091544e-05,
+      "loss": 0.9461,
+      "step": 1232
+    },
+    {
+      "epoch": 0.23720661792997308,
+      "grad_norm": 1.3872845455260447,
+      "learning_rate": 1.7831616525265515e-05,
+      "loss": 0.9108,
+      "step": 1233
+    },
+    {
+      "epoch": 0.23739899961523664,
+      "grad_norm": 2.94098640157213,
+      "learning_rate": 1.7827740551936296e-05,
+      "loss": 0.97,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2375913813005002,
+      "grad_norm": 1.3654621465800467,
+      "learning_rate": 1.7823861539608686e-05,
+      "loss": 0.9304,
+      "step": 1235
+    },
+    {
+      "epoch": 0.23778376298576376,
+      "grad_norm": 1.1353223355223285,
+      "learning_rate": 1.7819979489788638e-05,
+      "loss": 0.9813,
+      "step": 1236
+    },
+    {
+      "epoch": 0.23797614467102732,
+      "grad_norm": 1.1792491333850932,
+      "learning_rate": 1.7816094403983298e-05,
+      "loss": 0.8818,
+      "step": 1237
+    },
+    {
+      "epoch": 0.23816852635629088,
+      "grad_norm": 1.2715167090830664,
+      "learning_rate": 1.7812206283701002e-05,
+      "loss": 0.9088,
+      "step": 1238
+    },
+    {
+      "epoch": 0.23836090804155444,
+      "grad_norm": 1.314293468850535,
+      "learning_rate": 1.7808315130451244e-05,
+      "loss": 0.9248,
+      "step": 1239
+    },
+    {
+      "epoch": 0.238553289726818,
+      "grad_norm": 1.4614868123567932,
+      "learning_rate": 1.78044209457447e-05,
+      "loss": 0.8783,
+      "step": 1240
+    },
+    {
+      "epoch": 0.23874567141208156,
+      "grad_norm": 12.267411907257847,
+      "learning_rate": 1.7800523731093232e-05,
+      "loss": 0.9642,
+      "step": 1241
+    },
+    {
+      "epoch": 0.23893805309734514,
+      "grad_norm": 1.5246646506132633,
+      "learning_rate": 1.7796623488009875e-05,
+      "loss": 0.938,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2391304347826087,
+      "grad_norm": 1.3427676269337545,
+      "learning_rate": 1.7792720218008826e-05,
+      "loss": 0.934,
+      "step": 1243
+    },
+    {
+      "epoch": 0.23932281646787226,
+      "grad_norm": 1.3850346849065618,
+      "learning_rate": 1.7788813922605488e-05,
+      "loss": 0.8834,
+      "step": 1244
+    },
+    {
+      "epoch": 0.23951519815313582,
+      "grad_norm": 1.4348687925688173,
+      "learning_rate": 1.7784904603316402e-05,
+      "loss": 0.8846,
+      "step": 1245
+    },
+    {
+      "epoch": 0.23970757983839938,
+      "grad_norm": 1.5966458970086754,
+      "learning_rate": 1.7780992261659305e-05,
+      "loss": 0.9119,
+      "step": 1246
+    },
+    {
+      "epoch": 0.23989996152366294,
+      "grad_norm": 1.7029222945931266,
+      "learning_rate": 1.777707689915311e-05,
+      "loss": 0.9143,
+      "step": 1247
+    },
+    {
+      "epoch": 0.2400923432089265,
+      "grad_norm": 8.487057742228059,
+      "learning_rate": 1.777315851731789e-05,
+      "loss": 0.9077,
+      "step": 1248
+    },
+    {
+      "epoch": 0.24028472489419006,
+      "grad_norm": 1.6894325063793447,
+      "learning_rate": 1.7769237117674893e-05,
+      "loss": 0.9054,
+      "step": 1249
+    },
+    {
+      "epoch": 0.24047710657945365,
+      "grad_norm": 1.596905584077774,
+      "learning_rate": 1.7765312701746543e-05,
+      "loss": 0.9169,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2406694882647172,
+      "grad_norm": 4.249281109824366,
+      "learning_rate": 1.7761385271056436e-05,
+      "loss": 0.9643,
+      "step": 1251
+    },
+    {
+      "epoch": 0.24086186994998077,
+      "grad_norm": 2.502960741789566,
+      "learning_rate": 1.7757454827129338e-05,
+      "loss": 0.9365,
+      "step": 1252
+    },
+    {
+      "epoch": 0.24105425163524433,
+      "grad_norm": 0.808219032402478,
+      "learning_rate": 1.7753521371491174e-05,
+      "loss": 0.3463,
+      "step": 1253
+    },
+    {
+      "epoch": 0.24124663332050789,
+      "grad_norm": 2.7357685393692397,
+      "learning_rate": 1.7749584905669057e-05,
+      "loss": 0.8544,
+      "step": 1254
+    },
+    {
+      "epoch": 0.24143901500577145,
+      "grad_norm": 3.046287234103758,
+      "learning_rate": 1.774564543119125e-05,
+      "loss": 0.9648,
+      "step": 1255
+    },
+    {
+      "epoch": 0.241631396691035,
+      "grad_norm": 2.356285522390031,
+      "learning_rate": 1.7741702949587196e-05,
+      "loss": 0.9734,
+      "step": 1256
+    },
+    {
+      "epoch": 0.24182377837629856,
+      "grad_norm": 3.7778933778716737,
+      "learning_rate": 1.7737757462387507e-05,
+      "loss": 0.9559,
+      "step": 1257
+    },
+    {
+      "epoch": 0.24201616006156215,
+      "grad_norm": 1.476351682370277,
+      "learning_rate": 1.7733808971123946e-05,
+      "loss": 0.9491,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2422085417468257,
+      "grad_norm": 1.7808738564256348,
+      "learning_rate": 1.7729857477329463e-05,
+      "loss": 0.913,
+      "step": 1259
+    },
+    {
+      "epoch": 0.24240092343208927,
+      "grad_norm": 1.9654131178041585,
+      "learning_rate": 1.7725902982538162e-05,
+      "loss": 0.9104,
+      "step": 1260
+    },
+    {
+      "epoch": 0.24259330511735283,
+      "grad_norm": 2.3235152039954277,
+      "learning_rate": 1.772194548828531e-05,
+      "loss": 0.8894,
+      "step": 1261
+    },
+    {
+      "epoch": 0.2427856868026164,
+      "grad_norm": 2.221777078638393,
+      "learning_rate": 1.7717984996107346e-05,
+      "loss": 0.8863,
+      "step": 1262
+    },
+    {
+      "epoch": 0.24297806848787995,
+      "grad_norm": 0.6817046373575161,
+      "learning_rate": 1.771402150754187e-05,
+      "loss": 0.361,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2431704501731435,
+      "grad_norm": 4.1107042908266855,
+      "learning_rate": 1.7710055024127637e-05,
+      "loss": 0.8767,
+      "step": 1264
+    },
+    {
+      "epoch": 0.24336283185840707,
+      "grad_norm": 2.072430641418533,
+      "learning_rate": 1.7706085547404582e-05,
+      "loss": 0.9221,
+      "step": 1265
+    },
+    {
+      "epoch": 0.24355521354367063,
+      "grad_norm": 1.4488889945187386,
+      "learning_rate": 1.770211307891379e-05,
+      "loss": 0.948,
+      "step": 1266
+    },
+    {
+      "epoch": 0.24374759522893422,
+      "grad_norm": 1.9872983991265571,
+      "learning_rate": 1.769813762019751e-05,
+      "loss": 0.8839,
+      "step": 1267
+    },
+    {
+      "epoch": 0.24393997691419778,
+      "grad_norm": 1.7608519566255012,
+      "learning_rate": 1.769415917279915e-05,
+      "loss": 0.8795,
+      "step": 1268
+    },
+    {
+      "epoch": 0.24413235859946134,
+      "grad_norm": 1.6804646955626954,
+      "learning_rate": 1.7690177738263284e-05,
+      "loss": 0.9172,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2443247402847249,
+      "grad_norm": 1.6977172011251558,
+      "learning_rate": 1.7686193318135635e-05,
+      "loss": 0.9751,
+      "step": 1270
+    },
+    {
+      "epoch": 0.24451712196998845,
+      "grad_norm": 2.5644450579254805,
+      "learning_rate": 1.76822059139631e-05,
+      "loss": 0.9044,
+      "step": 1271
+    },
+    {
+      "epoch": 0.24470950365525201,
+      "grad_norm": 1.9817623124618122,
+      "learning_rate": 1.7678215527293724e-05,
+      "loss": 0.9334,
+      "step": 1272
+    },
+    {
+      "epoch": 0.24490188534051557,
+      "grad_norm": 1.545126753336433,
+      "learning_rate": 1.767422215967671e-05,
+      "loss": 0.9616,
+      "step": 1273
+    },
+    {
+      "epoch": 0.24509426702577913,
+      "grad_norm": 2.028454841235243,
+      "learning_rate": 1.767022581266242e-05,
+      "loss": 0.9488,
+      "step": 1274
+    },
+    {
+      "epoch": 0.24528664871104272,
+      "grad_norm": 1.4238665761240543,
+      "learning_rate": 1.766622648780238e-05,
+      "loss": 0.9662,
+      "step": 1275
+    },
+    {
+      "epoch": 0.24547903039630628,
+      "grad_norm": 2.191530836286052,
+      "learning_rate": 1.766222418664926e-05,
+      "loss": 0.9081,
+      "step": 1276
+    },
+    {
+      "epoch": 0.24567141208156984,
+      "grad_norm": 1.5280079230681363,
+      "learning_rate": 1.765821891075689e-05,
+      "loss": 0.965,
+      "step": 1277
+    },
+    {
+      "epoch": 0.2458637937668334,
+      "grad_norm": 1.6835763079637498,
+      "learning_rate": 1.7654210661680263e-05,
+      "loss": 0.9092,
+      "step": 1278
+    },
+    {
+      "epoch": 0.24605617545209696,
+      "grad_norm": 1.3849799691545168,
+      "learning_rate": 1.765019944097551e-05,
+      "loss": 0.9649,
+      "step": 1279
+    },
+    {
+      "epoch": 0.24624855713736052,
+      "grad_norm": 0.6553854177804136,
+      "learning_rate": 1.7646185250199936e-05,
+      "loss": 0.3451,
+      "step": 1280
+    },
+    {
+      "epoch": 0.24644093882262408,
+      "grad_norm": 0.688250040480236,
+      "learning_rate": 1.7642168090911976e-05,
+      "loss": 0.3233,
+      "step": 1281
+    },
+    {
+      "epoch": 0.24663332050788764,
+      "grad_norm": 1.8755609348803794,
+      "learning_rate": 1.763814796467124e-05,
+      "loss": 0.9699,
+      "step": 1282
+    },
+    {
+      "epoch": 0.24682570219315123,
+      "grad_norm": 1.5227271000633715,
+      "learning_rate": 1.763412487303847e-05,
+      "loss": 0.8904,
+      "step": 1283
+    },
+    {
+      "epoch": 0.24701808387841478,
+      "grad_norm": 1.4006580146455028,
+      "learning_rate": 1.7630098817575578e-05,
+      "loss": 0.9972,
+      "step": 1284
+    },
+    {
+      "epoch": 0.24721046556367834,
+      "grad_norm": 1.8896685891678637,
+      "learning_rate": 1.762606979984561e-05,
+      "loss": 0.9414,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2474028472489419,
+      "grad_norm": 1.7832225519105651,
+      "learning_rate": 1.7622037821412775e-05,
+      "loss": 0.9542,
+      "step": 1286
+    },
+    {
+      "epoch": 0.24759522893420546,
+      "grad_norm": 1.6368865314835228,
+      "learning_rate": 1.7618002883842426e-05,
+      "loss": 0.9249,
+      "step": 1287
+    },
+    {
+      "epoch": 0.24778761061946902,
+      "grad_norm": 1.3725078515452414,
+      "learning_rate": 1.7613964988701057e-05,
+      "loss": 0.9556,
+      "step": 1288
+    },
+    {
+      "epoch": 0.24797999230473258,
+      "grad_norm": 1.435184575169829,
+      "learning_rate": 1.7609924137556326e-05,
+      "loss": 0.9719,
+      "step": 1289
+    },
+    {
+      "epoch": 0.24817237398999614,
+      "grad_norm": 1.7260531956061564,
+      "learning_rate": 1.7605880331977022e-05,
+      "loss": 0.8895,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2483647556752597,
+      "grad_norm": 1.896739544985368,
+      "learning_rate": 1.76018335735331e-05,
+      "loss": 0.9422,
+      "step": 1291
+    },
+    {
+      "epoch": 0.2485571373605233,
+      "grad_norm": 1.4605609835913962,
+      "learning_rate": 1.7597783863795644e-05,
+      "loss": 0.9563,
+      "step": 1292
+    },
+    {
+      "epoch": 0.24874951904578685,
+      "grad_norm": 1.6592291973133289,
+      "learning_rate": 1.7593731204336895e-05,
+      "loss": 0.893,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2489419007310504,
+      "grad_norm": 1.600020855399038,
+      "learning_rate": 1.7589675596730233e-05,
+      "loss": 0.9224,
+      "step": 1294
+    },
+    {
+      "epoch": 0.24913428241631397,
+      "grad_norm": 1.3127767642539077,
+      "learning_rate": 1.758561704255018e-05,
+      "loss": 0.9538,
+      "step": 1295
+    },
+    {
+      "epoch": 0.24932666410157753,
+      "grad_norm": 1.4615528929210553,
+      "learning_rate": 1.7581555543372413e-05,
+      "loss": 0.9365,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2495190457868411,
+      "grad_norm": 1.317031321851893,
+      "learning_rate": 1.7577491100773744e-05,
+      "loss": 0.9017,
+      "step": 1297
+    },
+    {
+      "epoch": 0.24971142747210465,
+      "grad_norm": 1.4956917503387284,
+      "learning_rate": 1.7573423716332128e-05,
+      "loss": 0.8547,
+      "step": 1298
+    },
+    {
+      "epoch": 0.2499038091573682,
+      "grad_norm": 2.1866682901210406,
+      "learning_rate": 1.7569353391626665e-05,
+      "loss": 0.8979,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2500961908426318,
+      "grad_norm": 1.3375283753214997,
+      "learning_rate": 1.7565280128237595e-05,
+      "loss": 0.9248,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2502885725278953,
+      "grad_norm": 0.8234640914441792,
+      "learning_rate": 1.75612039277463e-05,
+      "loss": 0.3509,
+      "step": 1301
+    },
+    {
+      "epoch": 0.2504809542131589,
+      "grad_norm": 1.5493285659562477,
+      "learning_rate": 1.75571247917353e-05,
+      "loss": 0.9416,
+      "step": 1302
+    },
+    {
+      "epoch": 0.25067333589842244,
+      "grad_norm": 1.5271175845259628,
+      "learning_rate": 1.7553042721788255e-05,
+      "loss": 0.8843,
+      "step": 1303
+    },
+    {
+      "epoch": 0.25086571758368603,
+      "grad_norm": 0.64010563824874,
+      "learning_rate": 1.754895771948997e-05,
+      "loss": 0.3776,
+      "step": 1304
+    },
+    {
+      "epoch": 0.2510580992689496,
+      "grad_norm": 1.3849759438023934,
+      "learning_rate": 1.754486978642637e-05,
+      "loss": 0.9575,
+      "step": 1305
+    },
+    {
+      "epoch": 0.25125048095421315,
+      "grad_norm": 1.377004519446963,
+      "learning_rate": 1.7540778924184553e-05,
+      "loss": 0.9226,
+      "step": 1306
+    },
+    {
+      "epoch": 0.25144286263947674,
+      "grad_norm": 1.3940821341956227,
+      "learning_rate": 1.7536685134352717e-05,
+      "loss": 0.9254,
+      "step": 1307
+    },
+    {
+      "epoch": 0.25163524432474027,
+      "grad_norm": 1.422511197517404,
+      "learning_rate": 1.7532588418520215e-05,
+      "loss": 0.9889,
+      "step": 1308
+    },
+    {
+      "epoch": 0.25182762601000386,
+      "grad_norm": 1.5826713213610315,
+      "learning_rate": 1.7528488778277535e-05,
+      "loss": 1.0002,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2520200076952674,
+      "grad_norm": 1.707816037814451,
+      "learning_rate": 1.75243862152163e-05,
+      "loss": 0.8989,
+      "step": 1310
+    },
+    {
+      "epoch": 0.252212389380531,
+      "grad_norm": 2.0280634587157422,
+      "learning_rate": 1.752028073092926e-05,
+      "loss": 0.8712,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2524047710657945,
+      "grad_norm": 1.9286014013343244,
+      "learning_rate": 1.7516172327010314e-05,
+      "loss": 0.9601,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2525971527510581,
+      "grad_norm": 1.4432047377581612,
+      "learning_rate": 1.751206100505448e-05,
+      "loss": 0.9719,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2527895344363217,
+      "grad_norm": 1.6012382639219322,
+      "learning_rate": 1.7507946766657914e-05,
+      "loss": 0.9813,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2529819161215852,
+      "grad_norm": 1.4605100043260226,
+      "learning_rate": 1.7503829613417905e-05,
+      "loss": 0.9111,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2531742978068488,
+      "grad_norm": 1.8393567196306455,
+      "learning_rate": 1.749970954693288e-05,
+      "loss": 0.8659,
+      "step": 1316
+    },
+    {
+      "epoch": 0.25336667949211233,
+      "grad_norm": 1.3419159958059022,
+      "learning_rate": 1.7495586568802384e-05,
+      "loss": 0.8895,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2535590611773759,
+      "grad_norm": 1.5608754872849546,
+      "learning_rate": 1.7491460680627105e-05,
+      "loss": 0.9268,
+      "step": 1318
+    },
+    {
+      "epoch": 0.25375144286263945,
+      "grad_norm": 1.4915132256247017,
+      "learning_rate": 1.7487331884008845e-05,
+      "loss": 0.9172,
+      "step": 1319
+    },
+    {
+      "epoch": 0.25394382454790304,
+      "grad_norm": 1.4250726994888643,
+      "learning_rate": 1.7483200180550554e-05,
+      "loss": 0.9973,
+      "step": 1320
+    },
+    {
+      "epoch": 0.25413620623316663,
+      "grad_norm": 1.5298094256318757,
+      "learning_rate": 1.74790655718563e-05,
+      "loss": 0.9389,
+      "step": 1321
+    },
+    {
+      "epoch": 0.25432858791843016,
+      "grad_norm": 0.8223950121543493,
+      "learning_rate": 1.747492805953128e-05,
+      "loss": 0.3392,
+      "step": 1322
+    },
+    {
+      "epoch": 0.25452096960369375,
+      "grad_norm": 1.5452150698395615,
+      "learning_rate": 1.7470787645181818e-05,
+      "loss": 0.9013,
+      "step": 1323
+    },
+    {
+      "epoch": 0.2547133512889573,
+      "grad_norm": 1.5496722238557536,
+      "learning_rate": 1.7466644330415362e-05,
+      "loss": 0.9248,
+      "step": 1324
+    },
+    {
+      "epoch": 0.25490573297422087,
+      "grad_norm": 1.397429532889201,
+      "learning_rate": 1.7462498116840496e-05,
+      "loss": 0.9024,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2550981146594844,
+      "grad_norm": 1.6351982068171056,
+      "learning_rate": 1.745834900606692e-05,
+      "loss": 0.9812,
+      "step": 1326
+    },
+    {
+      "epoch": 0.255290496344748,
+      "grad_norm": 1.6675430939190936,
+      "learning_rate": 1.7454196999705458e-05,
+      "loss": 0.8968,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2554828780300115,
+      "grad_norm": 1.4098835531757217,
+      "learning_rate": 1.7450042099368066e-05,
+      "loss": 0.9413,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2556752597152751,
+      "grad_norm": 2.2997520920506687,
+      "learning_rate": 1.7445884306667823e-05,
+      "loss": 0.8993,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2558676414005387,
+      "grad_norm": 1.3758782054580718,
+      "learning_rate": 1.7441723623218917e-05,
+      "loss": 0.8729,
+      "step": 1330
+    },
+    {
+      "epoch": 0.2560600230858022,
+      "grad_norm": 1.2780774557514019,
+      "learning_rate": 1.7437560050636678e-05,
+      "loss": 0.8958,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2562524047710658,
+      "grad_norm": 1.4523994615010865,
+      "learning_rate": 1.7433393590537543e-05,
+      "loss": 0.8602,
+      "step": 1332
+    },
+    {
+      "epoch": 0.25644478645632934,
+      "grad_norm": 1.42900148773113,
+      "learning_rate": 1.7429224244539077e-05,
+      "loss": 0.9415,
+      "step": 1333
+    },
+    {
+      "epoch": 0.25663716814159293,
+      "grad_norm": 2.1752259767975857,
+      "learning_rate": 1.7425052014259965e-05,
+      "loss": 0.9656,
+      "step": 1334
+    },
+    {
+      "epoch": 0.25682954982685646,
+      "grad_norm": 1.5016327187638492,
+      "learning_rate": 1.7420876901320006e-05,
+      "loss": 0.957,
+      "step": 1335
+    },
+    {
+      "epoch": 0.25702193151212005,
+      "grad_norm": 4.6788318416728805,
+      "learning_rate": 1.7416698907340128e-05,
+      "loss": 0.9633,
+      "step": 1336
+    },
+    {
+      "epoch": 0.2572143131973836,
+      "grad_norm": 1.3307789194229787,
+      "learning_rate": 1.741251803394237e-05,
+      "loss": 0.9741,
+      "step": 1337
+    },
+    {
+      "epoch": 0.25740669488264717,
+      "grad_norm": 1.8902700860538122,
+      "learning_rate": 1.740833428274989e-05,
+      "loss": 0.9343,
+      "step": 1338
+    },
+    {
+      "epoch": 0.25759907656791076,
+      "grad_norm": 1.3932137627204082,
+      "learning_rate": 1.7404147655386966e-05,
+      "loss": 0.9204,
+      "step": 1339
+    },
+    {
+      "epoch": 0.2577914582531743,
+      "grad_norm": 1.3124937527739946,
+      "learning_rate": 1.739995815347899e-05,
+      "loss": 0.9317,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2579838399384379,
+      "grad_norm": 1.69779015221463,
+      "learning_rate": 1.739576577865247e-05,
+      "loss": 0.941,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2581762216237014,
+      "grad_norm": 1.3675301657586585,
+      "learning_rate": 1.739157053253503e-05,
+      "loss": 0.9182,
+      "step": 1342
+    },
+    {
+      "epoch": 0.258368603308965,
+      "grad_norm": 1.8600315949902388,
+      "learning_rate": 1.738737241675541e-05,
+      "loss": 0.9064,
+      "step": 1343
+    },
+    {
+      "epoch": 0.2585609849942285,
+      "grad_norm": 1.5217593249959738,
+      "learning_rate": 1.7383171432943466e-05,
+      "loss": 0.9255,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2587533666794921,
+      "grad_norm": 1.7533281681800412,
+      "learning_rate": 1.737896758273016e-05,
+      "loss": 0.9193,
+      "step": 1345
+    },
+    {
+      "epoch": 0.2589457483647557,
+      "grad_norm": 1.24290773759755,
+      "learning_rate": 1.7374760867747574e-05,
+      "loss": 0.978,
+      "step": 1346
+    },
+    {
+      "epoch": 0.25913813005001923,
+      "grad_norm": 1.632111382054558,
+      "learning_rate": 1.7370551289628895e-05,
+      "loss": 0.9275,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2593305117352828,
+      "grad_norm": 1.5398640776096992,
+      "learning_rate": 1.7366338850008432e-05,
+      "loss": 0.894,
+      "step": 1348
+    },
+    {
+      "epoch": 0.25952289342054635,
+      "grad_norm": 2.2517284491235654,
+      "learning_rate": 1.73621235505216e-05,
+      "loss": 0.9244,
+      "step": 1349
+    },
+    {
+      "epoch": 0.25971527510580994,
+      "grad_norm": 1.4260650472285905,
+      "learning_rate": 1.7357905392804918e-05,
+      "loss": 0.981,
+      "step": 1350
+    },
+    {
+      "epoch": 0.25990765679107347,
+      "grad_norm": 1.2655133296430388,
+      "learning_rate": 1.735368437849602e-05,
+      "loss": 0.9135,
+      "step": 1351
+    },
+    {
+      "epoch": 0.26010003847633706,
+      "grad_norm": 0.7549479948063662,
+      "learning_rate": 1.7349460509233654e-05,
+      "loss": 0.3513,
+      "step": 1352
+    },
+    {
+      "epoch": 0.2602924201616006,
+      "grad_norm": 1.5203884152513478,
+      "learning_rate": 1.734523378665767e-05,
+      "loss": 0.9204,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2604848018468642,
+      "grad_norm": 1.4930056171329111,
+      "learning_rate": 1.7341004212409026e-05,
+      "loss": 0.9688,
+      "step": 1354
+    },
+    {
+      "epoch": 0.26067718353212777,
+      "grad_norm": 1.8572648087503978,
+      "learning_rate": 1.7336771788129785e-05,
+      "loss": 0.8972,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2608695652173913,
+      "grad_norm": 1.3163915330435498,
+      "learning_rate": 1.7332536515463126e-05,
+      "loss": 0.949,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2610619469026549,
+      "grad_norm": 1.9417133431184022,
+      "learning_rate": 1.7328298396053324e-05,
+      "loss": 0.96,
+      "step": 1357
+    },
+    {
+      "epoch": 0.2612543285879184,
+      "grad_norm": 1.5267718640441448,
+      "learning_rate": 1.7324057431545768e-05,
+      "loss": 0.9825,
+      "step": 1358
+    },
+    {
+      "epoch": 0.261446710273182,
+      "grad_norm": 1.4584675994888885,
+      "learning_rate": 1.7319813623586935e-05,
+      "loss": 0.9281,
+      "step": 1359
+    },
+    {
+      "epoch": 0.26163909195844554,
+      "grad_norm": 1.2876691145455907,
+      "learning_rate": 1.7315566973824433e-05,
+      "loss": 0.9843,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2618314736437091,
+      "grad_norm": 1.9310689373529535,
+      "learning_rate": 1.7311317483906946e-05,
+      "loss": 0.9808,
+      "step": 1361
+    },
+    {
+      "epoch": 0.26202385532897265,
+      "grad_norm": 1.3554030247510158,
+      "learning_rate": 1.730706515548427e-05,
+      "loss": 0.9205,
+      "step": 1362
+    },
+    {
+      "epoch": 0.26221623701423624,
+      "grad_norm": 1.5863554137324272,
+      "learning_rate": 1.730280999020732e-05,
+      "loss": 0.9223,
+      "step": 1363
+    },
+    {
+      "epoch": 0.26240861869949983,
+      "grad_norm": 1.3245713253566962,
+      "learning_rate": 1.729855198972808e-05,
+      "loss": 0.9141,
+      "step": 1364
+    },
+    {
+      "epoch": 0.26260100038476336,
+      "grad_norm": 1.4231214037215225,
+      "learning_rate": 1.729429115569967e-05,
+      "loss": 0.9239,
+      "step": 1365
+    },
+    {
+      "epoch": 0.26279338207002695,
+      "grad_norm": 1.2414906867373223,
+      "learning_rate": 1.729002748977628e-05,
+      "loss": 0.8917,
+      "step": 1366
+    },
+    {
+      "epoch": 0.2629857637552905,
+      "grad_norm": 1.373048063379256,
+      "learning_rate": 1.7285760993613215e-05,
+      "loss": 0.9265,
+      "step": 1367
+    },
+    {
+      "epoch": 0.26317814544055407,
+      "grad_norm": 1.3610074172135032,
+      "learning_rate": 1.7281491668866874e-05,
+      "loss": 0.9076,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2633705271258176,
+      "grad_norm": 1.324703843832564,
+      "learning_rate": 1.727721951719476e-05,
+      "loss": 0.9329,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2635629088110812,
+      "grad_norm": 1.320067572409108,
+      "learning_rate": 1.7272944540255468e-05,
+      "loss": 0.9483,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2637552904963448,
+      "grad_norm": 1.537416772840183,
+      "learning_rate": 1.726866673970869e-05,
+      "loss": 0.9766,
+      "step": 1371
+    },
+    {
+      "epoch": 0.2639476721816083,
+      "grad_norm": 1.3097253716918884,
+      "learning_rate": 1.7264386117215216e-05,
+      "loss": 0.9286,
+      "step": 1372
+    },
+    {
+      "epoch": 0.2641400538668719,
+      "grad_norm": 1.589834021198922,
+      "learning_rate": 1.7260102674436933e-05,
+      "loss": 0.9154,
+      "step": 1373
+    },
+    {
+      "epoch": 0.2643324355521354,
+      "grad_norm": 1.4085970377254389,
+      "learning_rate": 1.7255816413036818e-05,
+      "loss": 0.88,
+      "step": 1374
+    },
+    {
+      "epoch": 0.264524817237399,
+      "grad_norm": 0.7232419541627513,
+      "learning_rate": 1.7251527334678946e-05,
+      "loss": 0.3484,
+      "step": 1375
+    },
+    {
+      "epoch": 0.26471719892266254,
+      "grad_norm": 1.2124908690644465,
+      "learning_rate": 1.7247235441028486e-05,
+      "loss": 0.965,
+      "step": 1376
+    },
+    {
+      "epoch": 0.26490958060792613,
+      "grad_norm": 1.6126444582740695,
+      "learning_rate": 1.7242940733751696e-05,
+      "loss": 0.9415,
+      "step": 1377
+    },
+    {
+      "epoch": 0.26510196229318966,
+      "grad_norm": 1.4381236426069808,
+      "learning_rate": 1.7238643214515934e-05,
+      "loss": 0.9296,
+      "step": 1378
+    },
+    {
+      "epoch": 0.26529434397845325,
+      "grad_norm": 1.3658851705889148,
+      "learning_rate": 1.7234342884989642e-05,
+      "loss": 0.8767,
+      "step": 1379
+    },
+    {
+      "epoch": 0.26548672566371684,
+      "grad_norm": 1.2876222007557543,
+      "learning_rate": 1.7230039746842352e-05,
+      "loss": 0.9389,
+      "step": 1380
+    },
+    {
+      "epoch": 0.26567910734898037,
+      "grad_norm": 1.366597798162509,
+      "learning_rate": 1.7225733801744698e-05,
+      "loss": 0.9274,
+      "step": 1381
+    },
+    {
+      "epoch": 0.26587148903424396,
+      "grad_norm": 1.1982312584165213,
+      "learning_rate": 1.7221425051368394e-05,
+      "loss": 0.9508,
+      "step": 1382
+    },
+    {
+      "epoch": 0.2660638707195075,
+      "grad_norm": 2.3761300879733716,
+      "learning_rate": 1.7217113497386245e-05,
+      "loss": 0.9245,
+      "step": 1383
+    },
+    {
+      "epoch": 0.2662562524047711,
+      "grad_norm": 1.301126093503405,
+      "learning_rate": 1.721279914147214e-05,
+      "loss": 0.95,
+      "step": 1384
+    },
+    {
+      "epoch": 0.2664486340900346,
+      "grad_norm": 1.3581776099916234,
+      "learning_rate": 1.7208481985301065e-05,
+      "loss": 0.9398,
+      "step": 1385
+    },
+    {
+      "epoch": 0.2666410157752982,
+      "grad_norm": 1.2967906060356642,
+      "learning_rate": 1.7204162030549093e-05,
+      "loss": 0.9359,
+      "step": 1386
+    },
+    {
+      "epoch": 0.2668333974605617,
+      "grad_norm": 1.1156192724149347,
+      "learning_rate": 1.7199839278893368e-05,
+      "loss": 0.8764,
+      "step": 1387
+    },
+    {
+      "epoch": 0.2670257791458253,
+      "grad_norm": 1.6398098315643694,
+      "learning_rate": 1.719551373201214e-05,
+      "loss": 0.892,
+      "step": 1388
+    },
+    {
+      "epoch": 0.2672181608310889,
+      "grad_norm": 1.392864820270682,
+      "learning_rate": 1.7191185391584736e-05,
+      "loss": 0.9365,
+      "step": 1389
+    },
+    {
+      "epoch": 0.26741054251635243,
+      "grad_norm": 1.4046202500898075,
+      "learning_rate": 1.7186854259291558e-05,
+      "loss": 0.9852,
+      "step": 1390
+    },
+    {
+      "epoch": 0.267602924201616,
+      "grad_norm": 1.5861642995406724,
+      "learning_rate": 1.7182520336814105e-05,
+      "loss": 0.9045,
+      "step": 1391
+    },
+    {
+      "epoch": 0.26779530588687955,
+      "grad_norm": 1.2910430024798667,
+      "learning_rate": 1.717818362583496e-05,
+      "loss": 0.9001,
+      "step": 1392
+    },
+    {
+      "epoch": 0.26798768757214314,
+      "grad_norm": 1.32803285273187,
+      "learning_rate": 1.7173844128037777e-05,
+      "loss": 0.9088,
+      "step": 1393
+    },
+    {
+      "epoch": 0.2681800692574067,
+      "grad_norm": 1.3284331312002833,
+      "learning_rate": 1.71695018451073e-05,
+      "loss": 0.9513,
+      "step": 1394
+    },
+    {
+      "epoch": 0.26837245094267026,
+      "grad_norm": 1.3213744443278592,
+      "learning_rate": 1.7165156778729355e-05,
+      "loss": 0.8873,
+      "step": 1395
+    },
+    {
+      "epoch": 0.26856483262793385,
+      "grad_norm": 1.1621315766345894,
+      "learning_rate": 1.7160808930590845e-05,
+      "loss": 0.9577,
+      "step": 1396
+    },
+    {
+      "epoch": 0.2687572143131974,
+      "grad_norm": 1.870223367433301,
+      "learning_rate": 1.7156458302379753e-05,
+      "loss": 0.9669,
+      "step": 1397
+    },
+    {
+      "epoch": 0.26894959599846097,
+      "grad_norm": 1.5253797068999728,
+      "learning_rate": 1.7152104895785147e-05,
+      "loss": 0.8738,
+      "step": 1398
+    },
+    {
+      "epoch": 0.2691419776837245,
+      "grad_norm": 1.3548185901193512,
+      "learning_rate": 1.7147748712497162e-05,
+      "loss": 0.9228,
+      "step": 1399
+    },
+    {
+      "epoch": 0.2693343593689881,
+      "grad_norm": 1.2000546833097268,
+      "learning_rate": 1.7143389754207026e-05,
+      "loss": 0.9455,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2695267410542516,
+      "grad_norm": 1.3436377617559718,
+      "learning_rate": 1.713902802260703e-05,
+      "loss": 0.938,
+      "step": 1401
+    },
+    {
+      "epoch": 0.2697191227395152,
+      "grad_norm": 1.35030990854481,
+      "learning_rate": 1.7134663519390557e-05,
+      "loss": 0.9326,
+      "step": 1402
+    },
+    {
+      "epoch": 0.26991150442477874,
+      "grad_norm": 1.2391690607539603,
+      "learning_rate": 1.7130296246252048e-05,
+      "loss": 0.9356,
+      "step": 1403
+    },
+    {
+      "epoch": 0.2701038861100423,
+      "grad_norm": 1.1939128052794266,
+      "learning_rate": 1.7125926204887034e-05,
+      "loss": 0.9035,
+      "step": 1404
+    },
+    {
+      "epoch": 0.2702962677953059,
+      "grad_norm": 1.5740541813747606,
+      "learning_rate": 1.712155339699211e-05,
+      "loss": 0.9302,
+      "step": 1405
+    },
+    {
+      "epoch": 0.27048864948056944,
+      "grad_norm": 1.427110853651075,
+      "learning_rate": 1.7117177824264962e-05,
+      "loss": 0.9539,
+      "step": 1406
+    },
+    {
+      "epoch": 0.27068103116583303,
+      "grad_norm": 0.7724013861210297,
+      "learning_rate": 1.7112799488404327e-05,
+      "loss": 0.3597,
+      "step": 1407
+    },
+    {
+      "epoch": 0.27087341285109656,
+      "grad_norm": 1.4230061227300654,
+      "learning_rate": 1.7108418391110033e-05,
+      "loss": 0.8862,
+      "step": 1408
+    },
+    {
+      "epoch": 0.27106579453636015,
+      "grad_norm": 1.597992451919525,
+      "learning_rate": 1.7104034534082968e-05,
+      "loss": 0.9478,
+      "step": 1409
+    },
+    {
+      "epoch": 0.2712581762216237,
+      "grad_norm": 1.5843401080159567,
+      "learning_rate": 1.7099647919025096e-05,
+      "loss": 0.869,
+      "step": 1410
+    },
+    {
+      "epoch": 0.27145055790688727,
+      "grad_norm": 1.1843082313200077,
+      "learning_rate": 1.7095258547639456e-05,
+      "loss": 0.9271,
+      "step": 1411
+    },
+    {
+      "epoch": 0.2716429395921508,
+      "grad_norm": 1.5431677543401905,
+      "learning_rate": 1.709086642163015e-05,
+      "loss": 0.8978,
+      "step": 1412
+    },
+    {
+      "epoch": 0.2718353212774144,
+      "grad_norm": 1.1192200087181559,
+      "learning_rate": 1.7086471542702355e-05,
+      "loss": 0.9232,
+      "step": 1413
+    },
+    {
+      "epoch": 0.272027702962678,
+      "grad_norm": 4.604447111608582,
+      "learning_rate": 1.708207391256231e-05,
+      "loss": 0.8804,
+      "step": 1414
+    },
+    {
+      "epoch": 0.2722200846479415,
+      "grad_norm": 1.435412239263252,
+      "learning_rate": 1.707767353291733e-05,
+      "loss": 0.8793,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2724124663332051,
+      "grad_norm": 1.5036549358083544,
+      "learning_rate": 1.7073270405475796e-05,
+      "loss": 0.9098,
+      "step": 1416
+    },
+    {
+      "epoch": 0.2726048480184686,
+      "grad_norm": 2.3238629824600707,
+      "learning_rate": 1.7068864531947147e-05,
+      "loss": 0.9077,
+      "step": 1417
+    },
+    {
+      "epoch": 0.2727972297037322,
+      "grad_norm": 1.5151298056474556,
+      "learning_rate": 1.70644559140419e-05,
+      "loss": 0.9438,
+      "step": 1418
+    },
+    {
+      "epoch": 0.27298961138899575,
+      "grad_norm": 2.764204933446641,
+      "learning_rate": 1.706004455347163e-05,
+      "loss": 0.8981,
+      "step": 1419
+    },
+    {
+      "epoch": 0.27318199307425933,
+      "grad_norm": 1.2027495818782394,
+      "learning_rate": 1.705563045194898e-05,
+      "loss": 0.9034,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2733743747595229,
+      "grad_norm": 1.7034947342910125,
+      "learning_rate": 1.7051213611187657e-05,
+      "loss": 0.9266,
+      "step": 1421
+    },
+    {
+      "epoch": 0.27356675644478645,
+      "grad_norm": 1.4043766323077924,
+      "learning_rate": 1.704679403290243e-05,
+      "loss": 0.8768,
+      "step": 1422
+    },
+    {
+      "epoch": 0.27375913813005004,
+      "grad_norm": 1.3270913761816363,
+      "learning_rate": 1.7042371718809132e-05,
+      "loss": 0.9133,
+      "step": 1423
+    },
+    {
+      "epoch": 0.27395151981531357,
+      "grad_norm": 1.3319820950664485,
+      "learning_rate": 1.7037946670624652e-05,
+      "loss": 0.949,
+      "step": 1424
+    },
+    {
+      "epoch": 0.27414390150057716,
+      "grad_norm": 1.4814211846142438,
+      "learning_rate": 1.7033518890066956e-05,
+      "loss": 0.9262,
+      "step": 1425
+    },
+    {
+      "epoch": 0.2743362831858407,
+      "grad_norm": 1.5709313592760252,
+      "learning_rate": 1.7029088378855055e-05,
+      "loss": 0.9319,
+      "step": 1426
+    },
+    {
+      "epoch": 0.2745286648711043,
+      "grad_norm": 1.3388836577540373,
+      "learning_rate": 1.7024655138709025e-05,
+      "loss": 0.9183,
+      "step": 1427
+    },
+    {
+      "epoch": 0.2747210465563678,
+      "grad_norm": 1.3226005067977222,
+      "learning_rate": 1.7020219171350004e-05,
+      "loss": 0.8999,
+      "step": 1428
+    },
+    {
+      "epoch": 0.2749134282416314,
+      "grad_norm": 0.7965553594685139,
+      "learning_rate": 1.7015780478500187e-05,
+      "loss": 0.3555,
+      "step": 1429
+    },
+    {
+      "epoch": 0.275105809926895,
+      "grad_norm": 3.013619660498979,
+      "learning_rate": 1.701133906188283e-05,
+      "loss": 0.8824,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2752981916121585,
+      "grad_norm": 1.4430259837219945,
+      "learning_rate": 1.700689492322224e-05,
+      "loss": 0.8888,
+      "step": 1431
+    },
+    {
+      "epoch": 0.2754905732974221,
+      "grad_norm": 1.3923174117034562,
+      "learning_rate": 1.700244806424379e-05,
+      "loss": 0.9324,
+      "step": 1432
+    },
+    {
+      "epoch": 0.27568295498268564,
+      "grad_norm": 1.4709641897794354,
+      "learning_rate": 1.6997998486673893e-05,
+      "loss": 0.8647,
+      "step": 1433
+    },
+    {
+      "epoch": 0.2758753366679492,
+      "grad_norm": 1.3377934565899,
+      "learning_rate": 1.699354619224004e-05,
+      "loss": 0.9281,
+      "step": 1434
+    },
+    {
+      "epoch": 0.27606771835321275,
+      "grad_norm": 1.5904392569965256,
+      "learning_rate": 1.698909118267076e-05,
+      "loss": 0.9017,
+      "step": 1435
+    },
+    {
+      "epoch": 0.27626010003847634,
+      "grad_norm": 1.1002712701934745,
+      "learning_rate": 1.6984633459695646e-05,
+      "loss": 0.9146,
+      "step": 1436
+    },
+    {
+      "epoch": 0.2764524817237399,
+      "grad_norm": 1.3248080201113563,
+      "learning_rate": 1.6980173025045328e-05,
+      "loss": 0.8506,
+      "step": 1437
+    },
+    {
+      "epoch": 0.27664486340900346,
+      "grad_norm": 1.414387663034322,
+      "learning_rate": 1.697570988045151e-05,
+      "loss": 0.944,
+      "step": 1438
+    },
+    {
+      "epoch": 0.27683724509426705,
+      "grad_norm": 1.557900709807223,
+      "learning_rate": 1.6971244027646937e-05,
+      "loss": 0.9565,
+      "step": 1439
+    },
+    {
+      "epoch": 0.2770296267795306,
+      "grad_norm": 1.5611306593106047,
+      "learning_rate": 1.69667754683654e-05,
+      "loss": 0.8977,
+      "step": 1440
+    },
+    {
+      "epoch": 0.27722200846479417,
+      "grad_norm": 1.4809730221675212,
+      "learning_rate": 1.6962304204341758e-05,
+      "loss": 0.9041,
+      "step": 1441
+    },
+    {
+      "epoch": 0.2774143901500577,
+      "grad_norm": 1.4437759091264175,
+      "learning_rate": 1.6957830237311904e-05,
+      "loss": 0.9012,
+      "step": 1442
+    },
+    {
+      "epoch": 0.2776067718353213,
+      "grad_norm": 1.4282549402132318,
+      "learning_rate": 1.6953353569012784e-05,
+      "loss": 0.8919,
+      "step": 1443
+    },
+    {
+      "epoch": 0.2777991535205848,
+      "grad_norm": 1.9392282609438207,
+      "learning_rate": 1.6948874201182402e-05,
+      "loss": 0.9122,
+      "step": 1444
+    },
+    {
+      "epoch": 0.2779915352058484,
+      "grad_norm": 1.572027444553582,
+      "learning_rate": 1.6944392135559798e-05,
+      "loss": 0.9225,
+      "step": 1445
+    },
+    {
+      "epoch": 0.278183916891112,
+      "grad_norm": 1.7272836773458689,
+      "learning_rate": 1.6939907373885062e-05,
+      "loss": 0.9127,
+      "step": 1446
+    },
+    {
+      "epoch": 0.2783762985763755,
+      "grad_norm": 1.3094104587853765,
+      "learning_rate": 1.6935419917899335e-05,
+      "loss": 0.8913,
+      "step": 1447
+    },
+    {
+      "epoch": 0.2785686802616391,
+      "grad_norm": 1.782558441319251,
+      "learning_rate": 1.6930929769344807e-05,
+      "loss": 0.9218,
+      "step": 1448
+    },
+    {
+      "epoch": 0.27876106194690264,
+      "grad_norm": 1.2789955701913798,
+      "learning_rate": 1.69264369299647e-05,
+      "loss": 0.9063,
+      "step": 1449
+    },
+    {
+      "epoch": 0.27895344363216623,
+      "grad_norm": 1.3876108310094828,
+      "learning_rate": 1.692194140150329e-05,
+      "loss": 0.9237,
+      "step": 1450
+    },
+    {
+      "epoch": 0.27914582531742976,
+      "grad_norm": 1.2043258913112795,
+      "learning_rate": 1.69174431857059e-05,
+      "loss": 0.9289,
+      "step": 1451
+    },
+    {
+      "epoch": 0.27933820700269335,
+      "grad_norm": 0.7358708774638161,
+      "learning_rate": 1.6912942284318898e-05,
+      "loss": 0.3693,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2795305886879569,
+      "grad_norm": 1.3071111463351153,
+      "learning_rate": 1.6908438699089674e-05,
+      "loss": 0.9261,
+      "step": 1453
+    },
+    {
+      "epoch": 0.27972297037322047,
+      "grad_norm": 1.258586018170485,
+      "learning_rate": 1.690393243176668e-05,
+      "loss": 0.8791,
+      "step": 1454
+    },
+    {
+      "epoch": 0.27991535205848406,
+      "grad_norm": 1.286172161686376,
+      "learning_rate": 1.6899423484099413e-05,
+      "loss": 0.8496,
+      "step": 1455
+    },
+    {
+      "epoch": 0.2801077337437476,
+      "grad_norm": 1.5749061563829099,
+      "learning_rate": 1.6894911857838394e-05,
+      "loss": 0.9014,
+      "step": 1456
+    },
+    {
+      "epoch": 0.2803001154290112,
+      "grad_norm": 1.1588739744623644,
+      "learning_rate": 1.689039755473519e-05,
+      "loss": 0.9442,
+      "step": 1457
+    },
+    {
+      "epoch": 0.2804924971142747,
+      "grad_norm": 1.3668739619707413,
+      "learning_rate": 1.6885880576542417e-05,
+      "loss": 0.873,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2806848787995383,
+      "grad_norm": 1.2445390419719427,
+      "learning_rate": 1.6881360925013712e-05,
+      "loss": 0.9221,
+      "step": 1459
+    },
+    {
+      "epoch": 0.2808772604848018,
+      "grad_norm": 1.375472502889773,
+      "learning_rate": 1.6876838601903765e-05,
+      "loss": 0.9142,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2810696421700654,
+      "grad_norm": 1.4743872728956473,
+      "learning_rate": 1.6872313608968296e-05,
+      "loss": 0.9219,
+      "step": 1461
+    },
+    {
+      "epoch": 0.28126202385532895,
+      "grad_norm": 2.074945308592774,
+      "learning_rate": 1.6867785947964065e-05,
+      "loss": 0.9531,
+      "step": 1462
+    },
+    {
+      "epoch": 0.28145440554059253,
+      "grad_norm": 1.573526189624601,
+      "learning_rate": 1.6863255620648866e-05,
+      "loss": 0.9217,
+      "step": 1463
+    },
+    {
+      "epoch": 0.2816467872258561,
+      "grad_norm": 1.2516813876473583,
+      "learning_rate": 1.685872262878152e-05,
+      "loss": 0.8801,
+      "step": 1464
+    },
+    {
+      "epoch": 0.28183916891111965,
+      "grad_norm": 1.6223204689270057,
+      "learning_rate": 1.6854186974121903e-05,
+      "loss": 0.9417,
+      "step": 1465
+    },
+    {
+      "epoch": 0.28203155059638324,
+      "grad_norm": 1.8560064496439324,
+      "learning_rate": 1.68496486584309e-05,
+      "loss": 0.904,
+      "step": 1466
+    },
+    {
+      "epoch": 0.28222393228164677,
+      "grad_norm": 1.6148298626699544,
+      "learning_rate": 1.6845107683470453e-05,
+      "loss": 0.9223,
+      "step": 1467
+    },
+    {
+      "epoch": 0.28241631396691036,
+      "grad_norm": 1.3768019275054304,
+      "learning_rate": 1.6840564051003517e-05,
+      "loss": 0.8804,
+      "step": 1468
+    },
+    {
+      "epoch": 0.2826086956521739,
+      "grad_norm": 1.241473150766179,
+      "learning_rate": 1.6836017762794087e-05,
+      "loss": 0.8947,
+      "step": 1469
+    },
+    {
+      "epoch": 0.2828010773374375,
+      "grad_norm": 1.2394293068664384,
+      "learning_rate": 1.6831468820607192e-05,
+      "loss": 0.9063,
+      "step": 1470
+    },
+    {
+      "epoch": 0.28299345902270107,
+      "grad_norm": 1.2255573633193253,
+      "learning_rate": 1.6826917226208886e-05,
+      "loss": 0.9535,
+      "step": 1471
+    },
+    {
+      "epoch": 0.2831858407079646,
+      "grad_norm": 1.1885630605263278,
+      "learning_rate": 1.6822362981366257e-05,
+      "loss": 0.8895,
+      "step": 1472
+    },
+    {
+      "epoch": 0.2833782223932282,
+      "grad_norm": 1.2655517193375743,
+      "learning_rate": 1.6817806087847417e-05,
+      "loss": 0.9122,
+      "step": 1473
+    },
+    {
+      "epoch": 0.2835706040784917,
+      "grad_norm": 1.5080111108051981,
+      "learning_rate": 1.681324654742151e-05,
+      "loss": 0.8887,
+      "step": 1474
+    },
+    {
+      "epoch": 0.2837629857637553,
+      "grad_norm": 1.8021834141389552,
+      "learning_rate": 1.6808684361858706e-05,
+      "loss": 0.8806,
+      "step": 1475
+    },
+    {
+      "epoch": 0.28395536744901884,
+      "grad_norm": 1.4331074502717698,
+      "learning_rate": 1.6804119532930202e-05,
+      "loss": 0.9025,
+      "step": 1476
+    },
+    {
+      "epoch": 0.2841477491342824,
+      "grad_norm": 1.4610680876540767,
+      "learning_rate": 1.6799552062408225e-05,
+      "loss": 0.9093,
+      "step": 1477
+    },
+    {
+      "epoch": 0.28434013081954596,
+      "grad_norm": 1.4905453100806256,
+      "learning_rate": 1.6794981952066018e-05,
+      "loss": 0.9143,
+      "step": 1478
+    },
+    {
+      "epoch": 0.28453251250480954,
+      "grad_norm": 0.7721013481957981,
+      "learning_rate": 1.6790409203677863e-05,
+      "loss": 0.3741,
+      "step": 1479
+    },
+    {
+      "epoch": 0.28472489419007313,
+      "grad_norm": 1.1274409876207165,
+      "learning_rate": 1.6785833819019052e-05,
+      "loss": 0.8855,
+      "step": 1480
+    },
+    {
+      "epoch": 0.28491727587533666,
+      "grad_norm": 1.3051368499086724,
+      "learning_rate": 1.678125579986591e-05,
+      "loss": 0.8985,
+      "step": 1481
+    },
+    {
+      "epoch": 0.28510965756060025,
+      "grad_norm": 1.5810397169572548,
+      "learning_rate": 1.677667514799578e-05,
+      "loss": 0.9295,
+      "step": 1482
+    },
+    {
+      "epoch": 0.2853020392458638,
+      "grad_norm": 1.8425534424652676,
+      "learning_rate": 1.6772091865187032e-05,
+      "loss": 0.8769,
+      "step": 1483
+    },
+    {
+      "epoch": 0.28549442093112737,
+      "grad_norm": 1.2799762963593992,
+      "learning_rate": 1.676750595321905e-05,
+      "loss": 0.9174,
+      "step": 1484
+    },
+    {
+      "epoch": 0.2856868026163909,
+      "grad_norm": 1.2668102139778348,
+      "learning_rate": 1.6762917413872246e-05,
+      "loss": 0.8797,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2858791843016545,
+      "grad_norm": 1.2865482781215454,
+      "learning_rate": 1.675832624892805e-05,
+      "loss": 0.928,
+      "step": 1486
+    },
+    {
+      "epoch": 0.286071565986918,
+      "grad_norm": 1.4292142271270083,
+      "learning_rate": 1.6753732460168907e-05,
+      "loss": 0.8203,
+      "step": 1487
+    },
+    {
+      "epoch": 0.2862639476721816,
+      "grad_norm": 1.523793122778115,
+      "learning_rate": 1.674913604937828e-05,
+      "loss": 0.9168,
+      "step": 1488
+    },
+    {
+      "epoch": 0.2864563293574452,
+      "grad_norm": 1.2499118482210803,
+      "learning_rate": 1.6744537018340662e-05,
+      "loss": 0.9283,
+      "step": 1489
+    },
+    {
+      "epoch": 0.2866487110427087,
+      "grad_norm": 1.6222609791137976,
+      "learning_rate": 1.6739935368841555e-05,
+      "loss": 0.9092,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2868410927279723,
+      "grad_norm": 1.2299512204717644,
+      "learning_rate": 1.6735331102667475e-05,
+      "loss": 0.8883,
+      "step": 1491
+    },
+    {
+      "epoch": 0.28703347441323585,
+      "grad_norm": 1.317152821010321,
+      "learning_rate": 1.6730724221605955e-05,
+      "loss": 0.9352,
+      "step": 1492
+    },
+    {
+      "epoch": 0.28722585609849943,
+      "grad_norm": 1.2539449999267036,
+      "learning_rate": 1.6726114727445547e-05,
+      "loss": 0.9058,
+      "step": 1493
+    },
+    {
+      "epoch": 0.28741823778376296,
+      "grad_norm": 1.4049500327808275,
+      "learning_rate": 1.6721502621975813e-05,
+      "loss": 0.9225,
+      "step": 1494
+    },
+    {
+      "epoch": 0.28761061946902655,
+      "grad_norm": 1.2352429804992886,
+      "learning_rate": 1.6716887906987332e-05,
+      "loss": 0.9198,
+      "step": 1495
+    },
+    {
+      "epoch": 0.28780300115429014,
+      "grad_norm": 1.516361043117736,
+      "learning_rate": 1.6712270584271703e-05,
+      "loss": 0.9299,
+      "step": 1496
+    },
+    {
+      "epoch": 0.28799538283955367,
+      "grad_norm": 1.6085212160706408,
+      "learning_rate": 1.670765065562152e-05,
+      "loss": 0.97,
+      "step": 1497
+    },
+    {
+      "epoch": 0.28818776452481726,
+      "grad_norm": 1.5587269813494011,
+      "learning_rate": 1.67030281228304e-05,
+      "loss": 0.8912,
+      "step": 1498
+    },
+    {
+      "epoch": 0.2883801462100808,
+      "grad_norm": 1.939156690425274,
+      "learning_rate": 1.6698402987692968e-05,
+      "loss": 0.9295,
+      "step": 1499
+    },
+    {
+      "epoch": 0.2885725278953444,
+      "grad_norm": 1.497799733895884,
+      "learning_rate": 1.6693775252004866e-05,
+      "loss": 0.9335,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2887649095806079,
+      "grad_norm": 1.368526390362278,
+      "learning_rate": 1.668914491756274e-05,
+      "loss": 0.9135,
+      "step": 1501
+    },
+    {
+      "epoch": 0.2889572912658715,
+      "grad_norm": 1.320076062722364,
+      "learning_rate": 1.668451198616424e-05,
+      "loss": 0.8636,
+      "step": 1502
+    },
+    {
+      "epoch": 0.28914967295113503,
+      "grad_norm": 1.3123049708047279,
+      "learning_rate": 1.6679876459608033e-05,
+      "loss": 0.9086,
+      "step": 1503
+    },
+    {
+      "epoch": 0.2893420546363986,
+      "grad_norm": 0.6797085775363699,
+      "learning_rate": 1.667523833969379e-05,
+      "loss": 0.3273,
+      "step": 1504
+    },
+    {
+      "epoch": 0.2895344363216622,
+      "grad_norm": 1.8462334990837639,
+      "learning_rate": 1.667059762822219e-05,
+      "loss": 0.9153,
+      "step": 1505
+    },
+    {
+      "epoch": 0.28972681800692573,
+      "grad_norm": 1.2576309601649356,
+      "learning_rate": 1.666595432699491e-05,
+      "loss": 0.9232,
+      "step": 1506
+    },
+    {
+      "epoch": 0.2899191996921893,
+      "grad_norm": 1.2615956893448883,
+      "learning_rate": 1.6661308437814652e-05,
+      "loss": 0.8915,
+      "step": 1507
+    },
+    {
+      "epoch": 0.29011158137745285,
+      "grad_norm": 1.2103358197985399,
+      "learning_rate": 1.6656659962485097e-05,
+      "loss": 0.8922,
+      "step": 1508
+    },
+    {
+      "epoch": 0.29030396306271644,
+      "grad_norm": 1.6909765128437302,
+      "learning_rate": 1.6652008902810952e-05,
+      "loss": 0.9317,
+      "step": 1509
+    },
+    {
+      "epoch": 0.29049634474798,
+      "grad_norm": 1.4229720488533952,
+      "learning_rate": 1.6647355260597915e-05,
+      "loss": 0.8813,
+      "step": 1510
+    },
+    {
+      "epoch": 0.29068872643324356,
+      "grad_norm": 1.2442620388033423,
+      "learning_rate": 1.664269903765269e-05,
+      "loss": 0.8901,
+      "step": 1511
+    },
+    {
+      "epoch": 0.2908811081185071,
+      "grad_norm": 1.6191155967958533,
+      "learning_rate": 1.6638040235782983e-05,
+      "loss": 0.9949,
+      "step": 1512
+    },
+    {
+      "epoch": 0.2910734898037707,
+      "grad_norm": 1.3262620694732026,
+      "learning_rate": 1.6633378856797505e-05,
+      "loss": 0.9477,
+      "step": 1513
+    },
+    {
+      "epoch": 0.29126587148903427,
+      "grad_norm": 1.2639292802460573,
+      "learning_rate": 1.662871490250596e-05,
+      "loss": 0.9247,
+      "step": 1514
+    },
+    {
+      "epoch": 0.2914582531742978,
+      "grad_norm": 1.4206952489464602,
+      "learning_rate": 1.662404837471905e-05,
+      "loss": 0.9383,
+      "step": 1515
+    },
+    {
+      "epoch": 0.2916506348595614,
+      "grad_norm": 1.3687708000111398,
+      "learning_rate": 1.66193792752485e-05,
+      "loss": 0.9147,
+      "step": 1516
+    },
+    {
+      "epoch": 0.2918430165448249,
+      "grad_norm": 1.1686776491933681,
+      "learning_rate": 1.6614707605906995e-05,
+      "loss": 0.9006,
+      "step": 1517
+    },
+    {
+      "epoch": 0.2920353982300885,
+      "grad_norm": 1.4155347790173787,
+      "learning_rate": 1.661003336850825e-05,
+      "loss": 0.9479,
+      "step": 1518
+    },
+    {
+      "epoch": 0.29222777991535204,
+      "grad_norm": 1.1599701975681376,
+      "learning_rate": 1.660535656486696e-05,
+      "loss": 0.9401,
+      "step": 1519
+    },
+    {
+      "epoch": 0.2924201616006156,
+      "grad_norm": 1.4091695702020428,
+      "learning_rate": 1.660067719679882e-05,
+      "loss": 0.8944,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2926125432858792,
+      "grad_norm": 1.2152328773371996,
+      "learning_rate": 1.6595995266120528e-05,
+      "loss": 0.9856,
+      "step": 1521
+    },
+    {
+      "epoch": 0.29280492497114274,
+      "grad_norm": 1.525306702587365,
+      "learning_rate": 1.6591310774649766e-05,
+      "loss": 0.9179,
+      "step": 1522
+    },
+    {
+      "epoch": 0.29299730665640633,
+      "grad_norm": 1.3338736125021393,
+      "learning_rate": 1.6586623724205216e-05,
+      "loss": 0.8947,
+      "step": 1523
+    },
+    {
+      "epoch": 0.29318968834166986,
+      "grad_norm": 1.3052655847097783,
+      "learning_rate": 1.6581934116606554e-05,
+      "loss": 0.9488,
+      "step": 1524
+    },
+    {
+      "epoch": 0.29338207002693345,
+      "grad_norm": 1.2753992836993095,
+      "learning_rate": 1.657724195367444e-05,
+      "loss": 0.9046,
+      "step": 1525
+    },
+    {
+      "epoch": 0.293574451712197,
+      "grad_norm": 1.1498796986675779,
+      "learning_rate": 1.657254723723054e-05,
+      "loss": 0.8703,
+      "step": 1526
+    },
+    {
+      "epoch": 0.29376683339746057,
+      "grad_norm": 1.3819870049418546,
+      "learning_rate": 1.6567849969097505e-05,
+      "loss": 0.8907,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2939592150827241,
+      "grad_norm": 1.3426046983309194,
+      "learning_rate": 1.6563150151098973e-05,
+      "loss": 0.8765,
+      "step": 1528
+    },
+    {
+      "epoch": 0.2941515967679877,
+      "grad_norm": 0.7280681233525133,
+      "learning_rate": 1.6558447785059577e-05,
+      "loss": 0.3195,
+      "step": 1529
+    },
+    {
+      "epoch": 0.2943439784532513,
+      "grad_norm": 1.4967653988386806,
+      "learning_rate": 1.655374287280494e-05,
+      "loss": 0.94,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2945363601385148,
+      "grad_norm": 1.4599521306879282,
+      "learning_rate": 1.6549035416161662e-05,
+      "loss": 0.9297,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2947287418237784,
+      "grad_norm": 1.21670216813166,
+      "learning_rate": 1.654432541695735e-05,
+      "loss": 0.8841,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2949211235090419,
+      "grad_norm": 1.2759412770252103,
+      "learning_rate": 1.653961287702058e-05,
+      "loss": 0.9242,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2951135051943055,
+      "grad_norm": 1.4357471029212545,
+      "learning_rate": 1.653489779818093e-05,
+      "loss": 0.8816,
+      "step": 1534
+    },
+    {
+      "epoch": 0.29530588687956905,
+      "grad_norm": 1.3020172234117047,
+      "learning_rate": 1.6530180182268946e-05,
+      "loss": 0.8434,
+      "step": 1535
+    },
+    {
+      "epoch": 0.29549826856483263,
+      "grad_norm": 1.368167965368425,
+      "learning_rate": 1.652546003111618e-05,
+      "loss": 0.8528,
+      "step": 1536
+    },
+    {
+      "epoch": 0.29569065025009617,
+      "grad_norm": 1.2769379577895827,
+      "learning_rate": 1.652073734655515e-05,
+      "loss": 0.8898,
+      "step": 1537
+    },
+    {
+      "epoch": 0.29588303193535975,
+      "grad_norm": 1.265469692471149,
+      "learning_rate": 1.6516012130419366e-05,
+      "loss": 0.89,
+      "step": 1538
+    },
+    {
+      "epoch": 0.29607541362062334,
+      "grad_norm": 1.2509704243666322,
+      "learning_rate": 1.6511284384543317e-05,
+      "loss": 0.8753,
+      "step": 1539
+    },
+    {
+      "epoch": 0.29626779530588687,
+      "grad_norm": 1.5425690357186144,
+      "learning_rate": 1.6506554110762483e-05,
+      "loss": 0.9109,
+      "step": 1540
+    },
+    {
+      "epoch": 0.29646017699115046,
+      "grad_norm": 1.6985262544932356,
+      "learning_rate": 1.650182131091332e-05,
+      "loss": 0.9313,
+      "step": 1541
+    },
+    {
+      "epoch": 0.296652558676414,
+      "grad_norm": 1.359951146268466,
+      "learning_rate": 1.6497085986833252e-05,
+      "loss": 0.8941,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2968449403616776,
+      "grad_norm": 1.2940133013835218,
+      "learning_rate": 1.6492348140360704e-05,
+      "loss": 0.9135,
+      "step": 1543
+    },
+    {
+      "epoch": 0.2970373220469411,
+      "grad_norm": 1.4675132167914793,
+      "learning_rate": 1.6487607773335074e-05,
+      "loss": 0.8941,
+      "step": 1544
+    },
+    {
+      "epoch": 0.2972297037322047,
+      "grad_norm": 1.6463893599509423,
+      "learning_rate": 1.648286488759673e-05,
+      "loss": 0.8446,
+      "step": 1545
+    },
+    {
+      "epoch": 0.29742208541746823,
+      "grad_norm": 1.1117743921177952,
+      "learning_rate": 1.6478119484987026e-05,
+      "loss": 0.8988,
+      "step": 1546
+    },
+    {
+      "epoch": 0.2976144671027318,
+      "grad_norm": 1.484288162276531,
+      "learning_rate": 1.6473371567348287e-05,
+      "loss": 0.9065,
+      "step": 1547
+    },
+    {
+      "epoch": 0.2978068487879954,
+      "grad_norm": 0.679399367648244,
+      "learning_rate": 1.6468621136523823e-05,
+      "loss": 0.3289,
+      "step": 1548
+    },
+    {
+      "epoch": 0.29799923047325894,
+      "grad_norm": 1.5510784599312921,
+      "learning_rate": 1.646386819435791e-05,
+      "loss": 0.8671,
+      "step": 1549
+    },
+    {
+      "epoch": 0.2981916121585225,
+      "grad_norm": 1.1987760304779131,
+      "learning_rate": 1.6459112742695807e-05,
+      "loss": 0.9186,
+      "step": 1550
+    },
+    {
+      "epoch": 0.29838399384378606,
+      "grad_norm": 1.5950130080726923,
+      "learning_rate": 1.6454354783383748e-05,
+      "loss": 0.8767,
+      "step": 1551
+    },
+    {
+      "epoch": 0.29857637552904964,
+      "grad_norm": 1.4528587184635995,
+      "learning_rate": 1.644959431826893e-05,
+      "loss": 0.9686,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2987687572143132,
+      "grad_norm": 1.224340230320657,
+      "learning_rate": 1.6444831349199528e-05,
+      "loss": 0.8577,
+      "step": 1553
+    },
+    {
+      "epoch": 0.29896113889957676,
+      "grad_norm": 1.4996683618316693,
+      "learning_rate": 1.6440065878024697e-05,
+      "loss": 0.8899,
+      "step": 1554
+    },
+    {
+      "epoch": 0.29915352058484035,
+      "grad_norm": 1.2508398689909903,
+      "learning_rate": 1.6435297906594553e-05,
+      "loss": 0.9287,
+      "step": 1555
+    },
+    {
+      "epoch": 0.2993459022701039,
+      "grad_norm": 1.6737672821023846,
+      "learning_rate": 1.643052743676019e-05,
+      "loss": 0.833,
+      "step": 1556
+    },
+    {
+      "epoch": 0.29953828395536747,
+      "grad_norm": 1.4326721466990688,
+      "learning_rate": 1.6425754470373667e-05,
+      "loss": 0.9361,
+      "step": 1557
+    },
+    {
+      "epoch": 0.299730665640631,
+      "grad_norm": 0.6400258202408518,
+      "learning_rate": 1.642097900928801e-05,
+      "loss": 0.3706,
+      "step": 1558
+    },
+    {
+      "epoch": 0.2999230473258946,
+      "grad_norm": 1.4157155617857924,
+      "learning_rate": 1.6416201055357225e-05,
+      "loss": 0.8832,
+      "step": 1559
+    },
+    {
+      "epoch": 0.3001154290111581,
+      "grad_norm": 8.348136048356652,
+      "learning_rate": 1.641142061043627e-05,
+      "loss": 0.8169,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3003078106964217,
+      "grad_norm": 1.5468825204659846,
+      "learning_rate": 1.640663767638108e-05,
+      "loss": 0.951,
+      "step": 1561
+    },
+    {
+      "epoch": 0.30050019238168524,
+      "grad_norm": 1.5517377065729554,
+      "learning_rate": 1.6401852255048564e-05,
+      "loss": 0.9367,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3006925740669488,
+      "grad_norm": 1.2889721465604345,
+      "learning_rate": 1.6397064348296578e-05,
+      "loss": 0.9018,
+      "step": 1563
+    },
+    {
+      "epoch": 0.3008849557522124,
+      "grad_norm": 0.6301393287680018,
+      "learning_rate": 1.6392273957983955e-05,
+      "loss": 0.3721,
+      "step": 1564
+    },
+    {
+      "epoch": 0.30107733743747594,
+      "grad_norm": 1.7775791812570203,
+      "learning_rate": 1.638748108597049e-05,
+      "loss": 0.9544,
+      "step": 1565
+    },
+    {
+      "epoch": 0.30126971912273953,
+      "grad_norm": 1.2609601710569445,
+      "learning_rate": 1.6382685734116934e-05,
+      "loss": 0.9167,
+      "step": 1566
+    },
+    {
+      "epoch": 0.30146210080800306,
+      "grad_norm": 1.2677026858701335,
+      "learning_rate": 1.6377887904285018e-05,
+      "loss": 0.8971,
+      "step": 1567
+    },
+    {
+      "epoch": 0.30165448249326665,
+      "grad_norm": 1.981410562671747,
+      "learning_rate": 1.637308759833742e-05,
+      "loss": 0.8885,
+      "step": 1568
+    },
+    {
+      "epoch": 0.3018468641785302,
+      "grad_norm": 1.8016318713375508,
+      "learning_rate": 1.6368284818137787e-05,
+      "loss": 0.9438,
+      "step": 1569
+    },
+    {
+      "epoch": 0.30203924586379377,
+      "grad_norm": 1.2225503871466803,
+      "learning_rate": 1.636347956555072e-05,
+      "loss": 0.9101,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3022316275490573,
+      "grad_norm": 2.2728330532573766,
+      "learning_rate": 1.635867184244178e-05,
+      "loss": 0.8827,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3024240092343209,
+      "grad_norm": 1.8121360899022918,
+      "learning_rate": 1.63538616506775e-05,
+      "loss": 0.8704,
+      "step": 1572
+    },
+    {
+      "epoch": 0.3026163909195845,
+      "grad_norm": 1.5065552448234194,
+      "learning_rate": 1.6349048992125358e-05,
+      "loss": 0.9374,
+      "step": 1573
+    },
+    {
+      "epoch": 0.302808772604848,
+      "grad_norm": 1.9177666976876688,
+      "learning_rate": 1.634423386865379e-05,
+      "loss": 0.9037,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3030011542901116,
+      "grad_norm": 1.7653946343448024,
+      "learning_rate": 1.6339416282132196e-05,
+      "loss": 0.9064,
+      "step": 1575
+    },
+    {
+      "epoch": 0.30319353597537513,
+      "grad_norm": 1.6193930540611057,
+      "learning_rate": 1.633459623443093e-05,
+      "loss": 0.9572,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3033859176606387,
+      "grad_norm": 1.4847570253907494,
+      "learning_rate": 1.6329773727421297e-05,
+      "loss": 0.9365,
+      "step": 1577
+    },
+    {
+      "epoch": 0.30357829934590225,
+      "grad_norm": 1.554495802584951,
+      "learning_rate": 1.6324948762975567e-05,
+      "loss": 0.8781,
+      "step": 1578
+    },
+    {
+      "epoch": 0.30377068103116583,
+      "grad_norm": 1.5192476603529044,
+      "learning_rate": 1.632012134296695e-05,
+      "loss": 0.9073,
+      "step": 1579
+    },
+    {
+      "epoch": 0.3039630627164294,
+      "grad_norm": 1.8330217400847895,
+      "learning_rate": 1.6315291469269617e-05,
+      "loss": 0.8723,
+      "step": 1580
+    },
+    {
+      "epoch": 0.30415544440169295,
+      "grad_norm": 0.7830947957246119,
+      "learning_rate": 1.63104591437587e-05,
+      "loss": 0.3797,
+      "step": 1581
+    },
+    {
+      "epoch": 0.30434782608695654,
+      "grad_norm": 0.6496747746837818,
+      "learning_rate": 1.6305624368310265e-05,
+      "loss": 0.3428,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3045402077722201,
+      "grad_norm": 1.3984606140966218,
+      "learning_rate": 1.630078714480134e-05,
+      "loss": 0.8943,
+      "step": 1583
+    },
+    {
+      "epoch": 0.30473258945748366,
+      "grad_norm": 1.3452431372747375,
+      "learning_rate": 1.6295947475109904e-05,
+      "loss": 0.9056,
+      "step": 1584
+    },
+    {
+      "epoch": 0.3049249711427472,
+      "grad_norm": 1.6430766592078216,
+      "learning_rate": 1.629110536111488e-05,
+      "loss": 0.9377,
+      "step": 1585
+    },
+    {
+      "epoch": 0.3051173528280108,
+      "grad_norm": 1.8623513872655162,
+      "learning_rate": 1.628626080469615e-05,
+      "loss": 0.9277,
+      "step": 1586
+    },
+    {
+      "epoch": 0.3053097345132743,
+      "grad_norm": 1.8376967020725552,
+      "learning_rate": 1.628141380773453e-05,
+      "loss": 0.8519,
+      "step": 1587
+    },
+    {
+      "epoch": 0.3055021161985379,
+      "grad_norm": 1.5270445527477612,
+      "learning_rate": 1.6276564372111797e-05,
+      "loss": 0.9859,
+      "step": 1588
+    },
+    {
+      "epoch": 0.3056944978838015,
+      "grad_norm": 1.8940351961509727,
+      "learning_rate": 1.6271712499710663e-05,
+      "loss": 0.8654,
+      "step": 1589
+    },
+    {
+      "epoch": 0.305886879569065,
+      "grad_norm": 1.505992068706958,
+      "learning_rate": 1.62668581924148e-05,
+      "loss": 0.9442,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3060792612543286,
+      "grad_norm": 0.9689755953729494,
+      "learning_rate": 1.6262001452108807e-05,
+      "loss": 0.3532,
+      "step": 1591
+    },
+    {
+      "epoch": 0.30627164293959214,
+      "grad_norm": 1.2456071809553553,
+      "learning_rate": 1.6257142280678247e-05,
+      "loss": 0.9197,
+      "step": 1592
+    },
+    {
+      "epoch": 0.3064640246248557,
+      "grad_norm": 1.565158087030123,
+      "learning_rate": 1.6252280680009613e-05,
+      "loss": 0.9758,
+      "step": 1593
+    },
+    {
+      "epoch": 0.30665640631011926,
+      "grad_norm": 1.398766983102164,
+      "learning_rate": 1.6247416651990343e-05,
+      "loss": 0.8981,
+      "step": 1594
+    },
+    {
+      "epoch": 0.30684878799538284,
+      "grad_norm": 1.3423152736015633,
+      "learning_rate": 1.624255019850883e-05,
+      "loss": 0.8536,
+      "step": 1595
+    },
+    {
+      "epoch": 0.3070411696806464,
+      "grad_norm": 1.2048023827175143,
+      "learning_rate": 1.6237681321454387e-05,
+      "loss": 0.9282,
+      "step": 1596
+    },
+    {
+      "epoch": 0.30723355136590996,
+      "grad_norm": 1.4266846036063339,
+      "learning_rate": 1.623281002271729e-05,
+      "loss": 0.9252,
+      "step": 1597
+    },
+    {
+      "epoch": 0.30742593305117355,
+      "grad_norm": 1.2397647537705412,
+      "learning_rate": 1.6227936304188738e-05,
+      "loss": 0.8477,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3076183147364371,
+      "grad_norm": 1.5717953370943163,
+      "learning_rate": 1.622306016776088e-05,
+      "loss": 0.8818,
+      "step": 1599
+    },
+    {
+      "epoch": 0.30781069642170067,
+      "grad_norm": 1.496772345299462,
+      "learning_rate": 1.6218181615326795e-05,
+      "loss": 0.9309,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3080030781069642,
+      "grad_norm": 1.2628514046881478,
+      "learning_rate": 1.6213300648780515e-05,
+      "loss": 0.9322,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3081954597922278,
+      "grad_norm": 1.3299075454770928,
+      "learning_rate": 1.620841727001699e-05,
+      "loss": 0.8545,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3083878414774913,
+      "grad_norm": 1.4310992511733422,
+      "learning_rate": 1.6203531480932114e-05,
+      "loss": 0.9016,
+      "step": 1603
+    },
+    {
+      "epoch": 0.3085802231627549,
+      "grad_norm": 2.2516508205107124,
+      "learning_rate": 1.619864328342273e-05,
+      "loss": 0.9146,
+      "step": 1604
+    },
+    {
+      "epoch": 0.3087726048480185,
+      "grad_norm": 1.2972412189064062,
+      "learning_rate": 1.6193752679386593e-05,
+      "loss": 0.8897,
+      "step": 1605
+    },
+    {
+      "epoch": 0.308964986533282,
+      "grad_norm": 1.4553044941068691,
+      "learning_rate": 1.6188859670722414e-05,
+      "loss": 0.8775,
+      "step": 1606
+    },
+    {
+      "epoch": 0.3091573682185456,
+      "grad_norm": 1.3830632570309747,
+      "learning_rate": 1.6183964259329817e-05,
+      "loss": 0.9116,
+      "step": 1607
+    },
+    {
+      "epoch": 0.30934974990380915,
+      "grad_norm": 7.370708822331498,
+      "learning_rate": 1.6179066447109376e-05,
+      "loss": 0.901,
+      "step": 1608
+    },
+    {
+      "epoch": 0.30954213158907273,
+      "grad_norm": 1.830511905656211,
+      "learning_rate": 1.6174166235962588e-05,
+      "loss": 0.8898,
+      "step": 1609
+    },
+    {
+      "epoch": 0.30973451327433627,
+      "grad_norm": 1.3775387849867002,
+      "learning_rate": 1.6169263627791886e-05,
+      "loss": 0.9557,
+      "step": 1610
+    },
+    {
+      "epoch": 0.30992689495959985,
+      "grad_norm": 1.4455011582618253,
+      "learning_rate": 1.616435862450063e-05,
+      "loss": 0.9107,
+      "step": 1611
+    },
+    {
+      "epoch": 0.3101192766448634,
+      "grad_norm": 0.8424644738278244,
+      "learning_rate": 1.615945122799311e-05,
+      "loss": 0.3738,
+      "step": 1612
+    },
+    {
+      "epoch": 0.31031165833012697,
+      "grad_norm": 0.698247470349564,
+      "learning_rate": 1.6154541440174547e-05,
+      "loss": 0.3547,
+      "step": 1613
+    },
+    {
+      "epoch": 0.31050404001539056,
+      "grad_norm": 1.3756545757741043,
+      "learning_rate": 1.614962926295109e-05,
+      "loss": 0.8794,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3106964217006541,
+      "grad_norm": 2.2508988620231376,
+      "learning_rate": 1.6144714698229814e-05,
+      "loss": 0.8715,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3108888033859177,
+      "grad_norm": 1.2444253391848927,
+      "learning_rate": 1.6139797747918725e-05,
+      "loss": 0.937,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3110811850711812,
+      "grad_norm": 1.6315768224295262,
+      "learning_rate": 1.613487841392675e-05,
+      "loss": 0.9147,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3112735667564448,
+      "grad_norm": 1.8334196792951845,
+      "learning_rate": 1.612995669816375e-05,
+      "loss": 0.9021,
+      "step": 1618
+    },
+    {
+      "epoch": 0.31146594844170833,
+      "grad_norm": 1.422020215396686,
+      "learning_rate": 1.6125032602540492e-05,
+      "loss": 0.9451,
+      "step": 1619
+    },
+    {
+      "epoch": 0.3116583301269719,
+      "grad_norm": 1.2921042771522637,
+      "learning_rate": 1.6120106128968686e-05,
+      "loss": 0.9327,
+      "step": 1620
+    },
+    {
+      "epoch": 0.31185071181223545,
+      "grad_norm": 1.4122433578065148,
+      "learning_rate": 1.6115177279360965e-05,
+      "loss": 0.9134,
+      "step": 1621
+    },
+    {
+      "epoch": 0.31204309349749904,
+      "grad_norm": 1.9958961549716863,
+      "learning_rate": 1.611024605563087e-05,
+      "loss": 0.89,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3122354751827626,
+      "grad_norm": 1.4934267925836693,
+      "learning_rate": 1.610531245969287e-05,
+      "loss": 0.9,
+      "step": 1623
+    },
+    {
+      "epoch": 0.31242785686802615,
+      "grad_norm": 1.3737669376233466,
+      "learning_rate": 1.6100376493462368e-05,
+      "loss": 0.9569,
+      "step": 1624
+    },
+    {
+      "epoch": 0.31262023855328974,
+      "grad_norm": 1.5709251854173467,
+      "learning_rate": 1.6095438158855668e-05,
+      "loss": 0.8524,
+      "step": 1625
+    },
+    {
+      "epoch": 0.3128126202385533,
+      "grad_norm": 1.0070964292779518,
+      "learning_rate": 1.609049745779e-05,
+      "loss": 0.3236,
+      "step": 1626
+    },
+    {
+      "epoch": 0.31300500192381686,
+      "grad_norm": 1.6618799046263677,
+      "learning_rate": 1.6085554392183517e-05,
+      "loss": 0.9517,
+      "step": 1627
+    },
+    {
+      "epoch": 0.3131973836090804,
+      "grad_norm": 1.2996481046227812,
+      "learning_rate": 1.608060896395529e-05,
+      "loss": 0.8867,
+      "step": 1628
+    },
+    {
+      "epoch": 0.313389765294344,
+      "grad_norm": 1.3465064414692751,
+      "learning_rate": 1.60756611750253e-05,
+      "loss": 0.9316,
+      "step": 1629
+    },
+    {
+      "epoch": 0.31358214697960757,
+      "grad_norm": 1.50912233735626,
+      "learning_rate": 1.6070711027314446e-05,
+      "loss": 0.8664,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3137745286648711,
+      "grad_norm": 1.2950745838922573,
+      "learning_rate": 1.606575852274456e-05,
+      "loss": 1.0031,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3139669103501347,
+      "grad_norm": 0.6845007190947715,
+      "learning_rate": 1.6060803663238357e-05,
+      "loss": 0.3569,
+      "step": 1632
+    },
+    {
+      "epoch": 0.3141592920353982,
+      "grad_norm": 2.0225849698142686,
+      "learning_rate": 1.6055846450719498e-05,
+      "loss": 0.8942,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3143516737206618,
+      "grad_norm": 1.3706029209222745,
+      "learning_rate": 1.6050886887112535e-05,
+      "loss": 0.929,
+      "step": 1634
+    },
+    {
+      "epoch": 0.31454405540592534,
+      "grad_norm": 1.4612056153581192,
+      "learning_rate": 1.6045924974342945e-05,
+      "loss": 0.9149,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3147364370911889,
+      "grad_norm": 1.4303528770376726,
+      "learning_rate": 1.604096071433711e-05,
+      "loss": 0.8714,
+      "step": 1636
+    },
+    {
+      "epoch": 0.31492881877645246,
+      "grad_norm": 1.2586968182682219,
+      "learning_rate": 1.6035994109022333e-05,
+      "loss": 0.9107,
+      "step": 1637
+    },
+    {
+      "epoch": 0.31512120046171604,
+      "grad_norm": 1.3256290745584558,
+      "learning_rate": 1.6031025160326814e-05,
+      "loss": 0.9106,
+      "step": 1638
+    },
+    {
+      "epoch": 0.31531358214697963,
+      "grad_norm": 1.400940957668645,
+      "learning_rate": 1.6026053870179678e-05,
+      "loss": 0.8702,
+      "step": 1639
+    },
+    {
+      "epoch": 0.31550596383224316,
+      "grad_norm": 1.2624930903787122,
+      "learning_rate": 1.6021080240510943e-05,
+      "loss": 0.8754,
+      "step": 1640
+    },
+    {
+      "epoch": 0.31569834551750675,
+      "grad_norm": 1.5286991270640289,
+      "learning_rate": 1.601610427325155e-05,
+      "loss": 0.9147,
+      "step": 1641
+    },
+    {
+      "epoch": 0.3158907272027703,
+      "grad_norm": 1.1147209282933015,
+      "learning_rate": 1.6011125970333333e-05,
+      "loss": 0.9,
+      "step": 1642
+    },
+    {
+      "epoch": 0.31608310888803387,
+      "grad_norm": 1.639063056153239,
+      "learning_rate": 1.600614533368905e-05,
+      "loss": 0.8829,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3162754905732974,
+      "grad_norm": 1.4115137225661174,
+      "learning_rate": 1.6001162365252348e-05,
+      "loss": 0.852,
+      "step": 1644
+    },
+    {
+      "epoch": 0.316467872258561,
+      "grad_norm": 1.8987921301488244,
+      "learning_rate": 1.5996177066957787e-05,
+      "loss": 0.888,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3166602539438245,
+      "grad_norm": 1.5484117404070192,
+      "learning_rate": 1.5991189440740838e-05,
+      "loss": 0.8919,
+      "step": 1646
+    },
+    {
+      "epoch": 0.3168526356290881,
+      "grad_norm": 1.216709517219925,
+      "learning_rate": 1.5986199488537867e-05,
+      "loss": 0.9246,
+      "step": 1647
+    },
+    {
+      "epoch": 0.3170450173143517,
+      "grad_norm": 1.8074810519962194,
+      "learning_rate": 1.598120721228614e-05,
+      "loss": 0.9029,
+      "step": 1648
+    },
+    {
+      "epoch": 0.3172373989996152,
+      "grad_norm": 1.5410983842053714,
+      "learning_rate": 1.5976212613923836e-05,
+      "loss": 0.902,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3174297806848788,
+      "grad_norm": 1.3314024368229942,
+      "learning_rate": 1.5971215695390026e-05,
+      "loss": 0.8902,
+      "step": 1650
+    },
+    {
+      "epoch": 0.31762216237014235,
+      "grad_norm": 1.2296481139530793,
+      "learning_rate": 1.5966216458624692e-05,
+      "loss": 0.8744,
+      "step": 1651
+    },
+    {
+      "epoch": 0.31781454405540593,
+      "grad_norm": 1.57613905831354,
+      "learning_rate": 1.5961214905568705e-05,
+      "loss": 0.9084,
+      "step": 1652
+    },
+    {
+      "epoch": 0.31800692574066947,
+      "grad_norm": 1.332525997343529,
+      "learning_rate": 1.595621103816384e-05,
+      "loss": 0.9649,
+      "step": 1653
+    },
+    {
+      "epoch": 0.31819930742593305,
+      "grad_norm": 1.7102605250257592,
+      "learning_rate": 1.5951204858352772e-05,
+      "loss": 0.9567,
+      "step": 1654
+    },
+    {
+      "epoch": 0.31839168911119664,
+      "grad_norm": 1.1780486627090296,
+      "learning_rate": 1.594619636807907e-05,
+      "loss": 0.8734,
+      "step": 1655
+    },
+    {
+      "epoch": 0.3185840707964602,
+      "grad_norm": 1.3602393063952811,
+      "learning_rate": 1.5941185569287206e-05,
+      "loss": 0.9385,
+      "step": 1656
+    },
+    {
+      "epoch": 0.31877645248172376,
+      "grad_norm": 2.815401417112556,
+      "learning_rate": 1.5936172463922542e-05,
+      "loss": 0.9056,
+      "step": 1657
+    },
+    {
+      "epoch": 0.3189688341669873,
+      "grad_norm": 1.518325916217633,
+      "learning_rate": 1.593115705393134e-05,
+      "loss": 0.9209,
+      "step": 1658
+    },
+    {
+      "epoch": 0.3191612158522509,
+      "grad_norm": 1.3071600946715678,
+      "learning_rate": 1.5926139341260755e-05,
+      "loss": 0.9269,
+      "step": 1659
+    },
+    {
+      "epoch": 0.3193535975375144,
+      "grad_norm": 1.6231505616699582,
+      "learning_rate": 1.5921119327858835e-05,
+      "loss": 0.9465,
+      "step": 1660
+    },
+    {
+      "epoch": 0.319545979222778,
+      "grad_norm": 1.5799284659580766,
+      "learning_rate": 1.5916097015674518e-05,
+      "loss": 0.9333,
+      "step": 1661
+    },
+    {
+      "epoch": 0.31973836090804153,
+      "grad_norm": 1.4089545914212152,
+      "learning_rate": 1.5911072406657646e-05,
+      "loss": 0.9572,
+      "step": 1662
+    },
+    {
+      "epoch": 0.3199307425933051,
+      "grad_norm": 1.2451430256445326,
+      "learning_rate": 1.5906045502758943e-05,
+      "loss": 0.8718,
+      "step": 1663
+    },
+    {
+      "epoch": 0.3201231242785687,
+      "grad_norm": 1.3084688674999838,
+      "learning_rate": 1.590101630593002e-05,
+      "loss": 0.8909,
+      "step": 1664
+    },
+    {
+      "epoch": 0.32031550596383224,
+      "grad_norm": 1.1134318270784402,
+      "learning_rate": 1.5895984818123392e-05,
+      "loss": 0.9648,
+      "step": 1665
+    },
+    {
+      "epoch": 0.3205078876490958,
+      "grad_norm": 1.3051919856853462,
+      "learning_rate": 1.5890951041292453e-05,
+      "loss": 0.9304,
+      "step": 1666
+    },
+    {
+      "epoch": 0.32070026933435936,
+      "grad_norm": 1.4551552416855367,
+      "learning_rate": 1.588591497739149e-05,
+      "loss": 0.8968,
+      "step": 1667
+    },
+    {
+      "epoch": 0.32089265101962294,
+      "grad_norm": 1.257658112188447,
+      "learning_rate": 1.5880876628375668e-05,
+      "loss": 0.8448,
+      "step": 1668
+    },
+    {
+      "epoch": 0.3210850327048865,
+      "grad_norm": 1.3291642340674583,
+      "learning_rate": 1.587583599620106e-05,
+      "loss": 0.8516,
+      "step": 1669
+    },
+    {
+      "epoch": 0.32127741439015006,
+      "grad_norm": 1.3942461244952655,
+      "learning_rate": 1.5870793082824604e-05,
+      "loss": 0.8907,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3214697960754136,
+      "grad_norm": 1.340214644371293,
+      "learning_rate": 1.5865747890204138e-05,
+      "loss": 0.9221,
+      "step": 1671
+    },
+    {
+      "epoch": 0.3216621777606772,
+      "grad_norm": 0.9320803672728799,
+      "learning_rate": 1.5860700420298377e-05,
+      "loss": 0.3653,
+      "step": 1672
+    },
+    {
+      "epoch": 0.32185455944594077,
+      "grad_norm": 1.1749675323416613,
+      "learning_rate": 1.5855650675066924e-05,
+      "loss": 0.9106,
+      "step": 1673
+    },
+    {
+      "epoch": 0.3220469411312043,
+      "grad_norm": 1.945678610409696,
+      "learning_rate": 1.5850598656470265e-05,
+      "loss": 0.9278,
+      "step": 1674
+    },
+    {
+      "epoch": 0.3222393228164679,
+      "grad_norm": 1.3103326851853736,
+      "learning_rate": 1.584554436646976e-05,
+      "loss": 0.9306,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3224317045017314,
+      "grad_norm": 1.418726525684453,
+      "learning_rate": 1.5840487807027665e-05,
+      "loss": 0.9094,
+      "step": 1676
+    },
+    {
+      "epoch": 0.322624086186995,
+      "grad_norm": 1.7384318092509448,
+      "learning_rate": 1.5835428980107113e-05,
+      "loss": 0.917,
+      "step": 1677
+    },
+    {
+      "epoch": 0.32281646787225854,
+      "grad_norm": 1.290361938710432,
+      "learning_rate": 1.583036788767211e-05,
+      "loss": 0.9063,
+      "step": 1678
+    },
+    {
+      "epoch": 0.3230088495575221,
+      "grad_norm": 1.433554636041544,
+      "learning_rate": 1.5825304531687548e-05,
+      "loss": 0.8922,
+      "step": 1679
+    },
+    {
+      "epoch": 0.3232012312427857,
+      "grad_norm": 1.1865136280807553,
+      "learning_rate": 1.5820238914119195e-05,
+      "loss": 0.8504,
+      "step": 1680
+    },
+    {
+      "epoch": 0.32339361292804925,
+      "grad_norm": 1.4211701099965526,
+      "learning_rate": 1.5815171036933697e-05,
+      "loss": 0.8733,
+      "step": 1681
+    },
+    {
+      "epoch": 0.32358599461331283,
+      "grad_norm": 1.492516934925325,
+      "learning_rate": 1.5810100902098582e-05,
+      "loss": 0.9667,
+      "step": 1682
+    },
+    {
+      "epoch": 0.32377837629857636,
+      "grad_norm": 1.3897607658096967,
+      "learning_rate": 1.580502851158225e-05,
+      "loss": 0.9484,
+      "step": 1683
+    },
+    {
+      "epoch": 0.32397075798383995,
+      "grad_norm": 1.2679125935268443,
+      "learning_rate": 1.5799953867353975e-05,
+      "loss": 0.9085,
+      "step": 1684
+    },
+    {
+      "epoch": 0.3241631396691035,
+      "grad_norm": 1.2816968657611842,
+      "learning_rate": 1.579487697138391e-05,
+      "loss": 0.9177,
+      "step": 1685
+    },
+    {
+      "epoch": 0.32435552135436707,
+      "grad_norm": 2.502821926446338,
+      "learning_rate": 1.5789797825643086e-05,
+      "loss": 0.9011,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3245479030396306,
+      "grad_norm": 1.2473955225926636,
+      "learning_rate": 1.5784716432103394e-05,
+      "loss": 0.926,
+      "step": 1687
+    },
+    {
+      "epoch": 0.3247402847248942,
+      "grad_norm": 1.3752209700464548,
+      "learning_rate": 1.5779632792737608e-05,
+      "loss": 0.9759,
+      "step": 1688
+    },
+    {
+      "epoch": 0.3249326664101578,
+      "grad_norm": 1.3250134780865974,
+      "learning_rate": 1.5774546909519376e-05,
+      "loss": 0.9191,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3251250480954213,
+      "grad_norm": 1.1247072684784292,
+      "learning_rate": 1.5769458784423206e-05,
+      "loss": 0.8919,
+      "step": 1690
+    },
+    {
+      "epoch": 0.3253174297806849,
+      "grad_norm": 1.5191179233132681,
+      "learning_rate": 1.5764368419424488e-05,
+      "loss": 0.9256,
+      "step": 1691
+    },
+    {
+      "epoch": 0.32550981146594843,
+      "grad_norm": 1.2226272424482774,
+      "learning_rate": 1.575927581649948e-05,
+      "loss": 0.9085,
+      "step": 1692
+    },
+    {
+      "epoch": 0.325702193151212,
+      "grad_norm": 1.2787281600356366,
+      "learning_rate": 1.5754180977625303e-05,
+      "loss": 0.9418,
+      "step": 1693
+    },
+    {
+      "epoch": 0.32589457483647555,
+      "grad_norm": 1.949602715240439,
+      "learning_rate": 1.574908390477995e-05,
+      "loss": 0.8746,
+      "step": 1694
+    },
+    {
+      "epoch": 0.32608695652173914,
+      "grad_norm": 1.290878618068371,
+      "learning_rate": 1.5743984599942273e-05,
+      "loss": 0.8953,
+      "step": 1695
+    },
+    {
+      "epoch": 0.32627933820700267,
+      "grad_norm": 1.1890287908255814,
+      "learning_rate": 1.5738883065092005e-05,
+      "loss": 0.8716,
+      "step": 1696
+    },
+    {
+      "epoch": 0.32647171989226625,
+      "grad_norm": 1.2981761007221835,
+      "learning_rate": 1.5733779302209735e-05,
+      "loss": 0.8415,
+      "step": 1697
+    },
+    {
+      "epoch": 0.32666410157752984,
+      "grad_norm": 1.3860949502710087,
+      "learning_rate": 1.572867331327692e-05,
+      "loss": 0.8948,
+      "step": 1698
+    },
+    {
+      "epoch": 0.3268564832627934,
+      "grad_norm": 1.2928462232390505,
+      "learning_rate": 1.5723565100275884e-05,
+      "loss": 0.9297,
+      "step": 1699
+    },
+    {
+      "epoch": 0.32704886494805696,
+      "grad_norm": 1.2089368212550362,
+      "learning_rate": 1.5718454665189806e-05,
+      "loss": 0.9154,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3272412466333205,
+      "grad_norm": 1.5481415695918037,
+      "learning_rate": 1.5713342010002733e-05,
+      "loss": 0.8865,
+      "step": 1701
+    },
+    {
+      "epoch": 0.3274336283185841,
+      "grad_norm": 1.3482874850156303,
+      "learning_rate": 1.5708227136699578e-05,
+      "loss": 0.8709,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3276260100038476,
+      "grad_norm": 1.266525086328612,
+      "learning_rate": 1.5703110047266105e-05,
+      "loss": 0.8522,
+      "step": 1703
+    },
+    {
+      "epoch": 0.3278183916891112,
+      "grad_norm": 1.3352403603764638,
+      "learning_rate": 1.569799074368895e-05,
+      "loss": 0.8736,
+      "step": 1704
+    },
+    {
+      "epoch": 0.3280107733743748,
+      "grad_norm": 1.32013500549489,
+      "learning_rate": 1.5692869227955603e-05,
+      "loss": 0.8761,
+      "step": 1705
+    },
+    {
+      "epoch": 0.3282031550596383,
+      "grad_norm": 1.4108876258207108,
+      "learning_rate": 1.5687745502054407e-05,
+      "loss": 0.8298,
+      "step": 1706
+    },
+    {
+      "epoch": 0.3283955367449019,
+      "grad_norm": 1.2007258902194604,
+      "learning_rate": 1.5682619567974575e-05,
+      "loss": 0.9245,
+      "step": 1707
+    },
+    {
+      "epoch": 0.32858791843016544,
+      "grad_norm": 1.415003550796323,
+      "learning_rate": 1.567749142770617e-05,
+      "loss": 0.9579,
+      "step": 1708
+    },
+    {
+      "epoch": 0.328780300115429,
+      "grad_norm": 1.3068859536423223,
+      "learning_rate": 1.5672361083240106e-05,
+      "loss": 0.8986,
+      "step": 1709
+    },
+    {
+      "epoch": 0.32897268180069256,
+      "grad_norm": 1.3393368130526402,
+      "learning_rate": 1.5667228536568167e-05,
+      "loss": 0.9369,
+      "step": 1710
+    },
+    {
+      "epoch": 0.32916506348595614,
+      "grad_norm": 0.8616126760337359,
+      "learning_rate": 1.566209378968298e-05,
+      "loss": 0.3617,
+      "step": 1711
+    },
+    {
+      "epoch": 0.3293574451712197,
+      "grad_norm": 1.6715064147044505,
+      "learning_rate": 1.565695684457803e-05,
+      "loss": 0.8247,
+      "step": 1712
+    },
+    {
+      "epoch": 0.32954982685648326,
+      "grad_norm": 1.2063877560101197,
+      "learning_rate": 1.5651817703247666e-05,
+      "loss": 0.9619,
+      "step": 1713
+    },
+    {
+      "epoch": 0.32974220854174685,
+      "grad_norm": 1.4588831708291468,
+      "learning_rate": 1.5646676367687067e-05,
+      "loss": 0.926,
+      "step": 1714
+    },
+    {
+      "epoch": 0.3299345902270104,
+      "grad_norm": 1.3951254062464347,
+      "learning_rate": 1.564153283989228e-05,
+      "loss": 0.8702,
+      "step": 1715
+    },
+    {
+      "epoch": 0.33012697191227397,
+      "grad_norm": 0.7058073292143646,
+      "learning_rate": 1.5636387121860207e-05,
+      "loss": 0.3562,
+      "step": 1716
+    },
+    {
+      "epoch": 0.3303193535975375,
+      "grad_norm": 1.4031405254798526,
+      "learning_rate": 1.5631239215588578e-05,
+      "loss": 0.836,
+      "step": 1717
+    },
+    {
+      "epoch": 0.3305117352828011,
+      "grad_norm": 1.4913351637214123,
+      "learning_rate": 1.5626089123076004e-05,
+      "loss": 0.8983,
+      "step": 1718
+    },
+    {
+      "epoch": 0.3307041169680646,
+      "grad_norm": 1.2654393794525776,
+      "learning_rate": 1.5620936846321917e-05,
+      "loss": 0.9494,
+      "step": 1719
+    },
+    {
+      "epoch": 0.3308964986533282,
+      "grad_norm": 1.2290870989986304,
+      "learning_rate": 1.561578238732661e-05,
+      "loss": 0.8529,
+      "step": 1720
+    },
+    {
+      "epoch": 0.33108888033859174,
+      "grad_norm": 1.329163581865413,
+      "learning_rate": 1.561062574809123e-05,
+      "loss": 0.8772,
+      "step": 1721
+    },
+    {
+      "epoch": 0.3312812620238553,
+      "grad_norm": 1.2778956142352107,
+      "learning_rate": 1.5605466930617747e-05,
+      "loss": 0.9174,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3314736437091189,
+      "grad_norm": 1.338830300014514,
+      "learning_rate": 1.5600305936909005e-05,
+      "loss": 0.8972,
+      "step": 1723
+    },
+    {
+      "epoch": 0.33166602539438245,
+      "grad_norm": 1.0875296814453828,
+      "learning_rate": 1.559514276896867e-05,
+      "loss": 0.9384,
+      "step": 1724
+    },
+    {
+      "epoch": 0.33185840707964603,
+      "grad_norm": 1.412905228017399,
+      "learning_rate": 1.558997742880127e-05,
+      "loss": 0.9669,
+      "step": 1725
+    },
+    {
+      "epoch": 0.33205078876490957,
+      "grad_norm": 1.2136315963743967,
+      "learning_rate": 1.5584809918412158e-05,
+      "loss": 0.9109,
+      "step": 1726
+    },
+    {
+      "epoch": 0.33224317045017315,
+      "grad_norm": 1.1091600129793762,
+      "learning_rate": 1.557964023980755e-05,
+      "loss": 0.864,
+      "step": 1727
+    },
+    {
+      "epoch": 0.3324355521354367,
+      "grad_norm": 0.8878417388514263,
+      "learning_rate": 1.5574468394994486e-05,
+      "loss": 0.3568,
+      "step": 1728
+    },
+    {
+      "epoch": 0.3326279338207003,
+      "grad_norm": 1.2415381658538314,
+      "learning_rate": 1.5569294385980856e-05,
+      "loss": 0.8892,
+      "step": 1729
+    },
+    {
+      "epoch": 0.33282031550596386,
+      "grad_norm": 1.1779126659951458,
+      "learning_rate": 1.556411821477539e-05,
+      "loss": 0.842,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3330126971912274,
+      "grad_norm": 1.3069263166728107,
+      "learning_rate": 1.5558939883387657e-05,
+      "loss": 0.9201,
+      "step": 1731
+    },
+    {
+      "epoch": 0.333205078876491,
+      "grad_norm": 2.302025010722525,
+      "learning_rate": 1.5553759393828058e-05,
+      "loss": 0.8363,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3333974605617545,
+      "grad_norm": 1.268185816755074,
+      "learning_rate": 1.554857674810784e-05,
+      "loss": 0.944,
+      "step": 1733
+    },
+    {
+      "epoch": 0.3335898422470181,
+      "grad_norm": 1.3361509981029487,
+      "learning_rate": 1.554339194823909e-05,
+      "loss": 0.9324,
+      "step": 1734
+    },
+    {
+      "epoch": 0.33378222393228163,
+      "grad_norm": 1.2582322114471556,
+      "learning_rate": 1.553820499623472e-05,
+      "loss": 0.8115,
+      "step": 1735
+    },
+    {
+      "epoch": 0.3339746056175452,
+      "grad_norm": 1.2618594158724914,
+      "learning_rate": 1.553301589410848e-05,
+      "loss": 0.8543,
+      "step": 1736
+    },
+    {
+      "epoch": 0.33416698730280875,
+      "grad_norm": 1.1902133521859535,
+      "learning_rate": 1.5527824643874968e-05,
+      "loss": 0.903,
+      "step": 1737
+    },
+    {
+      "epoch": 0.33435936898807234,
+      "grad_norm": 1.4193013197111775,
+      "learning_rate": 1.5522631247549598e-05,
+      "loss": 0.8821,
+      "step": 1738
+    },
+    {
+      "epoch": 0.3345517506733359,
+      "grad_norm": 1.246888171922412,
+      "learning_rate": 1.5517435707148628e-05,
+      "loss": 0.9199,
+      "step": 1739
+    },
+    {
+      "epoch": 0.33474413235859946,
+      "grad_norm": 1.435059433552667,
+      "learning_rate": 1.5512238024689144e-05,
+      "loss": 0.9452,
+      "step": 1740
+    },
+    {
+      "epoch": 0.33493651404386304,
+      "grad_norm": 1.3178018006706282,
+      "learning_rate": 1.550703820218907e-05,
+      "loss": 0.9115,
+      "step": 1741
+    },
+    {
+      "epoch": 0.3351288957291266,
+      "grad_norm": 1.428088379350606,
+      "learning_rate": 1.550183624166715e-05,
+      "loss": 0.9088,
+      "step": 1742
+    },
+    {
+      "epoch": 0.33532127741439016,
+      "grad_norm": 1.3016225550314948,
+      "learning_rate": 1.549663214514297e-05,
+      "loss": 0.926,
+      "step": 1743
+    },
+    {
+      "epoch": 0.3355136590996537,
+      "grad_norm": 1.6714885568109195,
+      "learning_rate": 1.5491425914636934e-05,
+      "loss": 0.8298,
+      "step": 1744
+    },
+    {
+      "epoch": 0.3357060407849173,
+      "grad_norm": 2.028546168395615,
+      "learning_rate": 1.5486217552170283e-05,
+      "loss": 0.9075,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3358984224701808,
+      "grad_norm": 1.366444666574281,
+      "learning_rate": 1.548100705976508e-05,
+      "loss": 0.8931,
+      "step": 1746
+    },
+    {
+      "epoch": 0.3360908041554444,
+      "grad_norm": 1.259259834128046,
+      "learning_rate": 1.5475794439444226e-05,
+      "loss": 0.9739,
+      "step": 1747
+    },
+    {
+      "epoch": 0.336283185840708,
+      "grad_norm": 1.336087169446166,
+      "learning_rate": 1.5470579693231432e-05,
+      "loss": 0.8849,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3364755675259715,
+      "grad_norm": 2.061600528611185,
+      "learning_rate": 1.5465362823151245e-05,
+      "loss": 0.88,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3366679492112351,
+      "grad_norm": 1.1674879068308937,
+      "learning_rate": 1.5460143831229026e-05,
+      "loss": 0.9144,
+      "step": 1750
+    },
+    {
+      "epoch": 0.33686033089649864,
+      "grad_norm": 1.5766970255866983,
+      "learning_rate": 1.545492271949098e-05,
+      "loss": 0.9293,
+      "step": 1751
+    },
+    {
+      "epoch": 0.3370527125817622,
+      "grad_norm": 1.365557557601289,
+      "learning_rate": 1.544969948996411e-05,
+      "loss": 0.9053,
+      "step": 1752
+    },
+    {
+      "epoch": 0.33724509426702576,
+      "grad_norm": 1.1005592880468793,
+      "learning_rate": 1.544447414467626e-05,
+      "loss": 0.8577,
+      "step": 1753
+    },
+    {
+      "epoch": 0.33743747595228935,
+      "grad_norm": 1.4596581932820785,
+      "learning_rate": 1.5439246685656093e-05,
+      "loss": 0.9309,
+      "step": 1754
+    },
+    {
+      "epoch": 0.33762985763755293,
+      "grad_norm": 1.4289447929000791,
+      "learning_rate": 1.5434017114933082e-05,
+      "loss": 0.8894,
+      "step": 1755
+    },
+    {
+      "epoch": 0.33782223932281646,
+      "grad_norm": 1.680976812234918,
+      "learning_rate": 1.5428785434537527e-05,
+      "loss": 0.9292,
+      "step": 1756
+    },
+    {
+      "epoch": 0.33801462100808005,
+      "grad_norm": 1.2121606593222218,
+      "learning_rate": 1.542355164650055e-05,
+      "loss": 0.9128,
+      "step": 1757
+    },
+    {
+      "epoch": 0.3382070026933436,
+      "grad_norm": 1.2144882525717922,
+      "learning_rate": 1.541831575285408e-05,
+      "loss": 0.8836,
+      "step": 1758
+    },
+    {
+      "epoch": 0.33839938437860717,
+      "grad_norm": 1.3146297751996348,
+      "learning_rate": 1.541307775563088e-05,
+      "loss": 0.9257,
+      "step": 1759
+    },
+    {
+      "epoch": 0.3385917660638707,
+      "grad_norm": 1.3023281050726003,
+      "learning_rate": 1.540783765686452e-05,
+      "loss": 0.8773,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3387841477491343,
+      "grad_norm": 1.7529700708455258,
+      "learning_rate": 1.540259545858938e-05,
+      "loss": 0.931,
+      "step": 1761
+    },
+    {
+      "epoch": 0.3389765294343978,
+      "grad_norm": 1.546347004749628,
+      "learning_rate": 1.539735116284067e-05,
+      "loss": 0.9552,
+      "step": 1762
+    },
+    {
+      "epoch": 0.3391689111196614,
+      "grad_norm": 1.4647576824967494,
+      "learning_rate": 1.53921047716544e-05,
+      "loss": 0.8555,
+      "step": 1763
+    },
+    {
+      "epoch": 0.339361292804925,
+      "grad_norm": 1.3369725408708333,
+      "learning_rate": 1.53868562870674e-05,
+      "loss": 0.8353,
+      "step": 1764
+    },
+    {
+      "epoch": 0.33955367449018853,
+      "grad_norm": 1.5529574991246706,
+      "learning_rate": 1.5381605711117318e-05,
+      "loss": 0.8997,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3397460561754521,
+      "grad_norm": 0.711181622211266,
+      "learning_rate": 1.5376353045842604e-05,
+      "loss": 0.3587,
+      "step": 1766
+    },
+    {
+      "epoch": 0.33993843786071565,
+      "grad_norm": 2.083955842029109,
+      "learning_rate": 1.5371098293282526e-05,
+      "loss": 0.9839,
+      "step": 1767
+    },
+    {
+      "epoch": 0.34013081954597923,
+      "grad_norm": 1.1497645476977862,
+      "learning_rate": 1.5365841455477158e-05,
+      "loss": 0.9066,
+      "step": 1768
+    },
+    {
+      "epoch": 0.34032320123124277,
+      "grad_norm": 1.9135261119995464,
+      "learning_rate": 1.5360582534467382e-05,
+      "loss": 0.8802,
+      "step": 1769
+    },
+    {
+      "epoch": 0.34051558291650635,
+      "grad_norm": 1.3833693812101067,
+      "learning_rate": 1.5355321532294897e-05,
+      "loss": 0.9001,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3407079646017699,
+      "grad_norm": 0.5786101461754086,
+      "learning_rate": 1.5350058451002204e-05,
+      "loss": 0.3565,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3409003462870335,
+      "grad_norm": 1.493365663595381,
+      "learning_rate": 1.5344793292632614e-05,
+      "loss": 0.8781,
+      "step": 1772
+    },
+    {
+      "epoch": 0.34109272797229706,
+      "grad_norm": 1.7202817805123785,
+      "learning_rate": 1.533952605923024e-05,
+      "loss": 0.8569,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3412851096575606,
+      "grad_norm": 1.6548456401691398,
+      "learning_rate": 1.5334256752840007e-05,
+      "loss": 0.8792,
+      "step": 1774
+    },
+    {
+      "epoch": 0.3414774913428242,
+      "grad_norm": 3.5760198912492647,
+      "learning_rate": 1.532898537550764e-05,
+      "loss": 0.8924,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3416698730280877,
+      "grad_norm": 1.3021724147874583,
+      "learning_rate": 1.532371192927966e-05,
+      "loss": 0.8754,
+      "step": 1776
+    },
+    {
+      "epoch": 0.3418622547133513,
+      "grad_norm": 1.4130166066892906,
+      "learning_rate": 1.5318436416203412e-05,
+      "loss": 0.8287,
+      "step": 1777
+    },
+    {
+      "epoch": 0.34205463639861483,
+      "grad_norm": 1.1771332859557948,
+      "learning_rate": 1.531315883832703e-05,
+      "loss": 0.8853,
+      "step": 1778
+    },
+    {
+      "epoch": 0.3422470180838784,
+      "grad_norm": 1.3255865061442078,
+      "learning_rate": 1.530787919769945e-05,
+      "loss": 0.8441,
+      "step": 1779
+    },
+    {
+      "epoch": 0.34243939976914195,
+      "grad_norm": 0.7132700163149693,
+      "learning_rate": 1.5302597496370408e-05,
+      "loss": 0.3794,
+      "step": 1780
+    },
+    {
+      "epoch": 0.34263178145440554,
+      "grad_norm": 1.1978865741951563,
+      "learning_rate": 1.5297313736390447e-05,
+      "loss": 0.864,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3428241631396691,
+      "grad_norm": 1.1383033581155866,
+      "learning_rate": 1.5292027919810898e-05,
+      "loss": 0.9082,
+      "step": 1782
+    },
+    {
+      "epoch": 0.34301654482493266,
+      "grad_norm": 1.2842985620136074,
+      "learning_rate": 1.52867400486839e-05,
+      "loss": 0.854,
+      "step": 1783
+    },
+    {
+      "epoch": 0.34320892651019624,
+      "grad_norm": 1.305908653655962,
+      "learning_rate": 1.528145012506239e-05,
+      "loss": 0.8752,
+      "step": 1784
+    },
+    {
+      "epoch": 0.3434013081954598,
+      "grad_norm": 1.3662882699020293,
+      "learning_rate": 1.5276158151000096e-05,
+      "loss": 0.8955,
+      "step": 1785
+    },
+    {
+      "epoch": 0.34359368988072336,
+      "grad_norm": 0.6292663210655275,
+      "learning_rate": 1.5270864128551542e-05,
+      "loss": 0.3462,
+      "step": 1786
+    },
+    {
+      "epoch": 0.3437860715659869,
+      "grad_norm": 1.2544301661224246,
+      "learning_rate": 1.5265568059772053e-05,
+      "loss": 0.868,
+      "step": 1787
+    },
+    {
+      "epoch": 0.3439784532512505,
+      "grad_norm": 1.178320014627109,
+      "learning_rate": 1.5260269946717746e-05,
+      "loss": 0.9034,
+      "step": 1788
+    },
+    {
+      "epoch": 0.34417083493651407,
+      "grad_norm": 1.5464699320536885,
+      "learning_rate": 1.5254969791445526e-05,
+      "loss": 0.8916,
+      "step": 1789
+    },
+    {
+      "epoch": 0.3443632166217776,
+      "grad_norm": 1.6422847121786739,
+      "learning_rate": 1.5249667596013102e-05,
+      "loss": 0.851,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3445555983070412,
+      "grad_norm": 1.269424415664039,
+      "learning_rate": 1.5244363362478967e-05,
+      "loss": 0.9048,
+      "step": 1791
+    },
+    {
+      "epoch": 0.3447479799923047,
+      "grad_norm": 1.343696629498384,
+      "learning_rate": 1.5239057092902404e-05,
+      "loss": 0.8946,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3449403616775683,
+      "grad_norm": 1.189929153740215,
+      "learning_rate": 1.523374878934349e-05,
+      "loss": 0.8863,
+      "step": 1793
+    },
+    {
+      "epoch": 0.34513274336283184,
+      "grad_norm": 1.4539970867093985,
+      "learning_rate": 1.5228438453863095e-05,
+      "loss": 0.9277,
+      "step": 1794
+    },
+    {
+      "epoch": 0.3453251250480954,
+      "grad_norm": 1.1977544361851509,
+      "learning_rate": 1.522312608852287e-05,
+      "loss": 0.8698,
+      "step": 1795
+    },
+    {
+      "epoch": 0.34551750673335896,
+      "grad_norm": 1.4050946731071239,
+      "learning_rate": 1.5217811695385263e-05,
+      "loss": 0.9514,
+      "step": 1796
+    },
+    {
+      "epoch": 0.34570988841862255,
+      "grad_norm": 1.1478370685320876,
+      "learning_rate": 1.52124952765135e-05,
+      "loss": 0.906,
+      "step": 1797
+    },
+    {
+      "epoch": 0.34590227010388613,
+      "grad_norm": 0.6503252624843171,
+      "learning_rate": 1.5207176833971598e-05,
+      "loss": 0.3771,
+      "step": 1798
+    },
+    {
+      "epoch": 0.34609465178914967,
+      "grad_norm": 0.6521309179479103,
+      "learning_rate": 1.520185636982436e-05,
+      "loss": 0.3516,
+      "step": 1799
+    },
+    {
+      "epoch": 0.34628703347441325,
+      "grad_norm": 1.210121064699872,
+      "learning_rate": 1.5196533886137376e-05,
+      "loss": 0.9282,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3464794151596768,
+      "grad_norm": 1.3748189321739517,
+      "learning_rate": 1.5191209384977014e-05,
+      "loss": 0.8713,
+      "step": 1801
+    },
+    {
+      "epoch": 0.34667179684494037,
+      "grad_norm": 1.2818470670976534,
+      "learning_rate": 1.5185882868410431e-05,
+      "loss": 0.8989,
+      "step": 1802
+    },
+    {
+      "epoch": 0.3468641785302039,
+      "grad_norm": 0.6873718219023365,
+      "learning_rate": 1.5180554338505564e-05,
+      "loss": 0.3504,
+      "step": 1803
+    },
+    {
+      "epoch": 0.3470565602154675,
+      "grad_norm": 1.391548961767055,
+      "learning_rate": 1.517522379733113e-05,
+      "loss": 0.8799,
+      "step": 1804
+    },
+    {
+      "epoch": 0.347248941900731,
+      "grad_norm": 1.2377593715239237,
+      "learning_rate": 1.5169891246956629e-05,
+      "loss": 0.8724,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3474413235859946,
+      "grad_norm": 1.5177898944999697,
+      "learning_rate": 1.5164556689452346e-05,
+      "loss": 0.8625,
+      "step": 1806
+    },
+    {
+      "epoch": 0.3476337052712582,
+      "grad_norm": 1.5539357346747944,
+      "learning_rate": 1.5159220126889329e-05,
+      "loss": 0.8296,
+      "step": 1807
+    },
+    {
+      "epoch": 0.34782608695652173,
+      "grad_norm": 1.3776153438596395,
+      "learning_rate": 1.5153881561339426e-05,
+      "loss": 0.9402,
+      "step": 1808
+    },
+    {
+      "epoch": 0.3480184686417853,
+      "grad_norm": 1.6487143487030504,
+      "learning_rate": 1.5148540994875242e-05,
+      "loss": 0.8649,
+      "step": 1809
+    },
+    {
+      "epoch": 0.34821085032704885,
+      "grad_norm": 1.270470226800381,
+      "learning_rate": 1.5143198429570181e-05,
+      "loss": 0.91,
+      "step": 1810
+    },
+    {
+      "epoch": 0.34840323201231244,
+      "grad_norm": 0.6741578468078233,
+      "learning_rate": 1.5137853867498403e-05,
+      "loss": 0.3528,
+      "step": 1811
+    },
+    {
+      "epoch": 0.34859561369757597,
+      "grad_norm": 1.316297072814937,
+      "learning_rate": 1.5132507310734847e-05,
+      "loss": 0.9318,
+      "step": 1812
+    },
+    {
+      "epoch": 0.34878799538283956,
+      "grad_norm": 1.2965625251327875,
+      "learning_rate": 1.5127158761355241e-05,
+      "loss": 0.945,
+      "step": 1813
+    },
+    {
+      "epoch": 0.34898037706810314,
+      "grad_norm": 1.3488398787378693,
+      "learning_rate": 1.512180822143607e-05,
+      "loss": 0.8679,
+      "step": 1814
+    },
+    {
+      "epoch": 0.3491727587533667,
+      "grad_norm": 1.3265177792431981,
+      "learning_rate": 1.5116455693054594e-05,
+      "loss": 0.9116,
+      "step": 1815
+    },
+    {
+      "epoch": 0.34936514043863026,
+      "grad_norm": 1.4516182587010658,
+      "learning_rate": 1.5111101178288858e-05,
+      "loss": 0.889,
+      "step": 1816
+    },
+    {
+      "epoch": 0.3495575221238938,
+      "grad_norm": 1.621769450959846,
+      "learning_rate": 1.510574467921766e-05,
+      "loss": 0.8723,
+      "step": 1817
+    },
+    {
+      "epoch": 0.3497499038091574,
+      "grad_norm": 1.419066063768393,
+      "learning_rate": 1.5100386197920585e-05,
+      "loss": 0.844,
+      "step": 1818
+    },
+    {
+      "epoch": 0.3499422854944209,
+      "grad_norm": 1.366828260274715,
+      "learning_rate": 1.5095025736477977e-05,
+      "loss": 0.8956,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3501346671796845,
+      "grad_norm": 1.430287198804004,
+      "learning_rate": 1.5089663296970952e-05,
+      "loss": 0.8869,
+      "step": 1820
+    },
+    {
+      "epoch": 0.35032704886494803,
+      "grad_norm": 1.3019954981605328,
+      "learning_rate": 1.5084298881481388e-05,
+      "loss": 0.8709,
+      "step": 1821
+    },
+    {
+      "epoch": 0.3505194305502116,
+      "grad_norm": 1.7373963923395703,
+      "learning_rate": 1.5078932492091942e-05,
+      "loss": 0.9051,
+      "step": 1822
+    },
+    {
+      "epoch": 0.3507118122354752,
+      "grad_norm": 1.356091696251858,
+      "learning_rate": 1.5073564130886032e-05,
+      "loss": 0.9246,
+      "step": 1823
+    },
+    {
+      "epoch": 0.35090419392073874,
+      "grad_norm": 2.776117448542096,
+      "learning_rate": 1.506819379994784e-05,
+      "loss": 0.8509,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3510965756060023,
+      "grad_norm": 1.4148902833015702,
+      "learning_rate": 1.5062821501362308e-05,
+      "loss": 0.8476,
+      "step": 1825
+    },
+    {
+      "epoch": 0.35128895729126586,
+      "grad_norm": 1.5699261752252565,
+      "learning_rate": 1.5057447237215152e-05,
+      "loss": 0.9293,
+      "step": 1826
+    },
+    {
+      "epoch": 0.35148133897652944,
+      "grad_norm": 1.5058361147577066,
+      "learning_rate": 1.5052071009592846e-05,
+      "loss": 0.8531,
+      "step": 1827
+    },
+    {
+      "epoch": 0.351673720661793,
+      "grad_norm": 1.514001464673424,
+      "learning_rate": 1.5046692820582625e-05,
+      "loss": 0.9117,
+      "step": 1828
+    },
+    {
+      "epoch": 0.35186610234705656,
+      "grad_norm": 1.1532846730381539,
+      "learning_rate": 1.504131267227249e-05,
+      "loss": 0.8747,
+      "step": 1829
+    },
+    {
+      "epoch": 0.3520584840323201,
+      "grad_norm": 1.61841961173823,
+      "learning_rate": 1.5035930566751198e-05,
+      "loss": 0.9184,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3522508657175837,
+      "grad_norm": 1.2863280664969896,
+      "learning_rate": 1.5030546506108268e-05,
+      "loss": 0.9302,
+      "step": 1831
+    },
+    {
+      "epoch": 0.35244324740284727,
+      "grad_norm": 2.058716084608437,
+      "learning_rate": 1.5025160492433976e-05,
+      "loss": 0.885,
+      "step": 1832
+    },
+    {
+      "epoch": 0.3526356290881108,
+      "grad_norm": 1.399716994471872,
+      "learning_rate": 1.501977252781936e-05,
+      "loss": 0.882,
+      "step": 1833
+    },
+    {
+      "epoch": 0.3528280107733744,
+      "grad_norm": 1.2206073453425588,
+      "learning_rate": 1.5014382614356213e-05,
+      "loss": 0.8979,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3530203924586379,
+      "grad_norm": 1.6131593787820504,
+      "learning_rate": 1.5008990754137088e-05,
+      "loss": 0.8008,
+      "step": 1835
+    },
+    {
+      "epoch": 0.3532127741439015,
+      "grad_norm": 1.3195154710494195,
+      "learning_rate": 1.5003596949255284e-05,
+      "loss": 0.8663,
+      "step": 1836
+    },
+    {
+      "epoch": 0.35340515582916504,
+      "grad_norm": 1.3561366713876932,
+      "learning_rate": 1.4998201201804867e-05,
+      "loss": 0.8941,
+      "step": 1837
+    },
+    {
+      "epoch": 0.35359753751442863,
+      "grad_norm": 1.3498614140169032,
+      "learning_rate": 1.499280351388065e-05,
+      "loss": 0.8834,
+      "step": 1838
+    },
+    {
+      "epoch": 0.3537899191996922,
+      "grad_norm": 1.3003329123960732,
+      "learning_rate": 1.49874038875782e-05,
+      "loss": 0.8617,
+      "step": 1839
+    },
+    {
+      "epoch": 0.35398230088495575,
+      "grad_norm": 1.4417848179176624,
+      "learning_rate": 1.498200232499384e-05,
+      "loss": 0.8923,
+      "step": 1840
+    },
+    {
+      "epoch": 0.35417468257021933,
+      "grad_norm": 1.4876567432323315,
+      "learning_rate": 1.4976598828224643e-05,
+      "loss": 0.9449,
+      "step": 1841
+    },
+    {
+      "epoch": 0.35436706425548287,
+      "grad_norm": 1.3511918957575073,
+      "learning_rate": 1.497119339936843e-05,
+      "loss": 0.8374,
+      "step": 1842
+    },
+    {
+      "epoch": 0.35455944594074645,
+      "grad_norm": 1.6788540162198824,
+      "learning_rate": 1.4965786040523779e-05,
+      "loss": 0.9106,
+      "step": 1843
+    },
+    {
+      "epoch": 0.35475182762601,
+      "grad_norm": 1.2816450673124617,
+      "learning_rate": 1.496037675379001e-05,
+      "loss": 0.8967,
+      "step": 1844
+    },
+    {
+      "epoch": 0.3549442093112736,
+      "grad_norm": 0.7801330139686229,
+      "learning_rate": 1.4954965541267192e-05,
+      "loss": 0.3298,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3551365909965371,
+      "grad_norm": 1.4872880349184834,
+      "learning_rate": 1.494955240505615e-05,
+      "loss": 0.8318,
+      "step": 1846
+    },
+    {
+      "epoch": 0.3553289726818007,
+      "grad_norm": 1.6103180304658387,
+      "learning_rate": 1.494413734725844e-05,
+      "loss": 0.8656,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3555213543670643,
+      "grad_norm": 1.2118554180815406,
+      "learning_rate": 1.4938720369976385e-05,
+      "loss": 0.8705,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3557137360523278,
+      "grad_norm": 1.198650162255159,
+      "learning_rate": 1.4933301475313036e-05,
+      "loss": 0.9046,
+      "step": 1849
+    },
+    {
+      "epoch": 0.3559061177375914,
+      "grad_norm": 1.611031262638704,
+      "learning_rate": 1.4927880665372197e-05,
+      "loss": 0.9476,
+      "step": 1850
+    },
+    {
+      "epoch": 0.35609849942285493,
+      "grad_norm": 1.2660228448508848,
+      "learning_rate": 1.4922457942258411e-05,
+      "loss": 0.8816,
+      "step": 1851
+    },
+    {
+      "epoch": 0.3562908811081185,
+      "grad_norm": 1.2571625345978847,
+      "learning_rate": 1.4917033308076967e-05,
+      "loss": 0.9117,
+      "step": 1852
+    },
+    {
+      "epoch": 0.35648326279338205,
+      "grad_norm": 1.5599131164036721,
+      "learning_rate": 1.4911606764933892e-05,
+      "loss": 0.9404,
+      "step": 1853
+    },
+    {
+      "epoch": 0.35667564447864564,
+      "grad_norm": 1.38962953628217,
+      "learning_rate": 1.490617831493596e-05,
+      "loss": 0.8993,
+      "step": 1854
+    },
+    {
+      "epoch": 0.35686802616390917,
+      "grad_norm": 1.1359118106401038,
+      "learning_rate": 1.4900747960190682e-05,
+      "loss": 0.8684,
+      "step": 1855
+    },
+    {
+      "epoch": 0.35706040784917276,
+      "grad_norm": 1.349192572623237,
+      "learning_rate": 1.489531570280631e-05,
+      "loss": 0.9303,
+      "step": 1856
+    },
+    {
+      "epoch": 0.35725278953443634,
+      "grad_norm": 1.5152652958354764,
+      "learning_rate": 1.488988154489183e-05,
+      "loss": 0.8665,
+      "step": 1857
+    },
+    {
+      "epoch": 0.3574451712196999,
+      "grad_norm": 1.4524476884899484,
+      "learning_rate": 1.4884445488556972e-05,
+      "loss": 0.8606,
+      "step": 1858
+    },
+    {
+      "epoch": 0.35763755290496346,
+      "grad_norm": 1.2274671359187745,
+      "learning_rate": 1.4879007535912198e-05,
+      "loss": 0.9107,
+      "step": 1859
+    },
+    {
+      "epoch": 0.357829934590227,
+      "grad_norm": 0.7321762734120114,
+      "learning_rate": 1.4873567689068708e-05,
+      "loss": 0.3475,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3580223162754906,
+      "grad_norm": 1.0795576494543784,
+      "learning_rate": 1.4868125950138442e-05,
+      "loss": 0.9044,
+      "step": 1861
+    },
+    {
+      "epoch": 0.3582146979607541,
+      "grad_norm": 1.410055113698886,
+      "learning_rate": 1.4862682321234064e-05,
+      "loss": 0.8889,
+      "step": 1862
+    },
+    {
+      "epoch": 0.3584070796460177,
+      "grad_norm": 1.114971005447042,
+      "learning_rate": 1.4857236804468983e-05,
+      "loss": 0.9254,
+      "step": 1863
+    },
+    {
+      "epoch": 0.3585994613312813,
+      "grad_norm": 1.2648955663317565,
+      "learning_rate": 1.4851789401957338e-05,
+      "loss": 0.8676,
+      "step": 1864
+    },
+    {
+      "epoch": 0.3587918430165448,
+      "grad_norm": 1.276391526420769,
+      "learning_rate": 1.4846340115813993e-05,
+      "loss": 0.849,
+      "step": 1865
+    },
+    {
+      "epoch": 0.3589842247018084,
+      "grad_norm": 1.2786109036551678,
+      "learning_rate": 1.484088894815455e-05,
+      "loss": 0.9144,
+      "step": 1866
+    },
+    {
+      "epoch": 0.35917660638707194,
+      "grad_norm": 1.451117481825918,
+      "learning_rate": 1.4835435901095341e-05,
+      "loss": 0.9829,
+      "step": 1867
+    },
+    {
+      "epoch": 0.3593689880723355,
+      "grad_norm": 0.7383372770868302,
+      "learning_rate": 1.4829980976753426e-05,
+      "loss": 0.3595,
+      "step": 1868
+    },
+    {
+      "epoch": 0.35956136975759906,
+      "grad_norm": 1.407865122456153,
+      "learning_rate": 1.4824524177246597e-05,
+      "loss": 0.8627,
+      "step": 1869
+    },
+    {
+      "epoch": 0.35975375144286265,
+      "grad_norm": 1.478913367257557,
+      "learning_rate": 1.4819065504693365e-05,
+      "loss": 0.8304,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3599461331281262,
+      "grad_norm": 1.1979273178906618,
+      "learning_rate": 1.4813604961212984e-05,
+      "loss": 0.9334,
+      "step": 1871
+    },
+    {
+      "epoch": 0.36013851481338977,
+      "grad_norm": 2.3643210481448507,
+      "learning_rate": 1.4808142548925417e-05,
+      "loss": 0.9388,
+      "step": 1872
+    },
+    {
+      "epoch": 0.36033089649865335,
+      "grad_norm": 1.5128978023652755,
+      "learning_rate": 1.4802678269951365e-05,
+      "loss": 0.891,
+      "step": 1873
+    },
+    {
+      "epoch": 0.3605232781839169,
+      "grad_norm": 1.6298654226373421,
+      "learning_rate": 1.4797212126412243e-05,
+      "loss": 0.8979,
+      "step": 1874
+    },
+    {
+      "epoch": 0.36071565986918047,
+      "grad_norm": 1.4920841455572604,
+      "learning_rate": 1.4791744120430202e-05,
+      "loss": 0.8742,
+      "step": 1875
+    },
+    {
+      "epoch": 0.360908041554444,
+      "grad_norm": 1.4144723587444774,
+      "learning_rate": 1.4786274254128112e-05,
+      "loss": 0.9322,
+      "step": 1876
+    },
+    {
+      "epoch": 0.3611004232397076,
+      "grad_norm": 1.6957568358256154,
+      "learning_rate": 1.4780802529629559e-05,
+      "loss": 0.9107,
+      "step": 1877
+    },
+    {
+      "epoch": 0.3612928049249711,
+      "grad_norm": 1.6921328409161338,
+      "learning_rate": 1.4775328949058856e-05,
+      "loss": 0.9033,
+      "step": 1878
+    },
+    {
+      "epoch": 0.3614851866102347,
+      "grad_norm": 4.9458250124612295,
+      "learning_rate": 1.4769853514541037e-05,
+      "loss": 0.9866,
+      "step": 1879
+    },
+    {
+      "epoch": 0.36167756829549824,
+      "grad_norm": 1.2483448515200326,
+      "learning_rate": 1.4764376228201848e-05,
+      "loss": 0.8358,
+      "step": 1880
+    },
+    {
+      "epoch": 0.36186994998076183,
+      "grad_norm": 1.5589741690405727,
+      "learning_rate": 1.475889709216777e-05,
+      "loss": 0.8824,
+      "step": 1881
+    },
+    {
+      "epoch": 0.3620623316660254,
+      "grad_norm": 1.1939398037958011,
+      "learning_rate": 1.4753416108565985e-05,
+      "loss": 0.9267,
+      "step": 1882
+    },
+    {
+      "epoch": 0.36225471335128895,
+      "grad_norm": 1.2102226512701573,
+      "learning_rate": 1.47479332795244e-05,
+      "loss": 0.876,
+      "step": 1883
+    },
+    {
+      "epoch": 0.36244709503655254,
+      "grad_norm": 1.637249315363224,
+      "learning_rate": 1.4742448607171644e-05,
+      "loss": 0.9462,
+      "step": 1884
+    },
+    {
+      "epoch": 0.36263947672181607,
+      "grad_norm": 0.6855968906872206,
+      "learning_rate": 1.473696209363705e-05,
+      "loss": 0.3901,
+      "step": 1885
+    },
+    {
+      "epoch": 0.36283185840707965,
+      "grad_norm": 1.4261722547802331,
+      "learning_rate": 1.4731473741050673e-05,
+      "loss": 0.8827,
+      "step": 1886
+    },
+    {
+      "epoch": 0.3630242400923432,
+      "grad_norm": 2.2651878236823286,
+      "learning_rate": 1.4725983551543279e-05,
+      "loss": 0.874,
+      "step": 1887
+    },
+    {
+      "epoch": 0.3632166217776068,
+      "grad_norm": 2.038955619676474,
+      "learning_rate": 1.472049152724635e-05,
+      "loss": 0.8803,
+      "step": 1888
+    },
+    {
+      "epoch": 0.36340900346287036,
+      "grad_norm": 1.5027237670149365,
+      "learning_rate": 1.471499767029208e-05,
+      "loss": 0.9262,
+      "step": 1889
+    },
+    {
+      "epoch": 0.3636013851481339,
+      "grad_norm": 1.563242631671576,
+      "learning_rate": 1.470950198281337e-05,
+      "loss": 0.9302,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3637937668333975,
+      "grad_norm": 1.564323333447396,
+      "learning_rate": 1.470400446694384e-05,
+      "loss": 0.865,
+      "step": 1891
+    },
+    {
+      "epoch": 0.363986148518661,
+      "grad_norm": 5.243018068552054,
+      "learning_rate": 1.4698505124817811e-05,
+      "loss": 0.8832,
+      "step": 1892
+    },
+    {
+      "epoch": 0.3641785302039246,
+      "grad_norm": 1.8972104703887231,
+      "learning_rate": 1.4693003958570318e-05,
+      "loss": 0.8736,
+      "step": 1893
+    },
+    {
+      "epoch": 0.36437091188918813,
+      "grad_norm": 0.6754161547735122,
+      "learning_rate": 1.4687500970337103e-05,
+      "loss": 0.3415,
+      "step": 1894
+    },
+    {
+      "epoch": 0.3645632935744517,
+      "grad_norm": 1.257300617281897,
+      "learning_rate": 1.4681996162254618e-05,
+      "loss": 0.9191,
+      "step": 1895
+    },
+    {
+      "epoch": 0.36475567525971525,
+      "grad_norm": 1.4590241806281314,
+      "learning_rate": 1.4676489536460015e-05,
+      "loss": 0.9504,
+      "step": 1896
+    },
+    {
+      "epoch": 0.36494805694497884,
+      "grad_norm": 1.383816358781125,
+      "learning_rate": 1.467098109509116e-05,
+      "loss": 0.9259,
+      "step": 1897
+    },
+    {
+      "epoch": 0.3651404386302424,
+      "grad_norm": 1.332760467299209,
+      "learning_rate": 1.4665470840286614e-05,
+      "loss": 0.8827,
+      "step": 1898
+    },
+    {
+      "epoch": 0.36533282031550596,
+      "grad_norm": 1.5962195492638371,
+      "learning_rate": 1.4659958774185654e-05,
+      "loss": 0.9252,
+      "step": 1899
+    },
+    {
+      "epoch": 0.36552520200076954,
+      "grad_norm": 1.2259198591930769,
+      "learning_rate": 1.4654444898928249e-05,
+      "loss": 0.8825,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3657175836860331,
+      "grad_norm": 1.2826758370764568,
+      "learning_rate": 1.4648929216655077e-05,
+      "loss": 0.9379,
+      "step": 1901
+    },
+    {
+      "epoch": 0.36590996537129666,
+      "grad_norm": 2.4564436412217496,
+      "learning_rate": 1.4643411729507517e-05,
+      "loss": 0.9115,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3661023470565602,
+      "grad_norm": 1.5956017456688403,
+      "learning_rate": 1.4637892439627644e-05,
+      "loss": 0.8981,
+      "step": 1903
+    },
+    {
+      "epoch": 0.3662947287418238,
+      "grad_norm": 1.414482377458824,
+      "learning_rate": 1.4632371349158241e-05,
+      "loss": 0.9309,
+      "step": 1904
+    },
+    {
+      "epoch": 0.3664871104270873,
+      "grad_norm": 1.3146698375582215,
+      "learning_rate": 1.4626848460242782e-05,
+      "loss": 0.9397,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3666794921123509,
+      "grad_norm": 1.470312470807216,
+      "learning_rate": 1.4621323775025444e-05,
+      "loss": 0.8641,
+      "step": 1906
+    },
+    {
+      "epoch": 0.3668718737976145,
+      "grad_norm": 1.3870641170237419,
+      "learning_rate": 1.4615797295651099e-05,
+      "loss": 0.8852,
+      "step": 1907
+    },
+    {
+      "epoch": 0.367064255482878,
+      "grad_norm": 1.2947707670675788,
+      "learning_rate": 1.4610269024265317e-05,
+      "loss": 0.9565,
+      "step": 1908
+    },
+    {
+      "epoch": 0.3672566371681416,
+      "grad_norm": 1.9319624273824314,
+      "learning_rate": 1.4604738963014365e-05,
+      "loss": 0.8936,
+      "step": 1909
+    },
+    {
+      "epoch": 0.36744901885340514,
+      "grad_norm": 0.7058505718348735,
+      "learning_rate": 1.4599207114045202e-05,
+      "loss": 0.3855,
+      "step": 1910
+    },
+    {
+      "epoch": 0.36764140053866873,
+      "grad_norm": 1.2791184740336352,
+      "learning_rate": 1.4593673479505482e-05,
+      "loss": 0.8924,
+      "step": 1911
+    },
+    {
+      "epoch": 0.36783378222393226,
+      "grad_norm": 1.4904159367932077,
+      "learning_rate": 1.4588138061543551e-05,
+      "loss": 0.8731,
+      "step": 1912
+    },
+    {
+      "epoch": 0.36802616390919585,
+      "grad_norm": 1.6049272729074708,
+      "learning_rate": 1.458260086230845e-05,
+      "loss": 0.9501,
+      "step": 1913
+    },
+    {
+      "epoch": 0.36821854559445943,
+      "grad_norm": 0.5736561048679438,
+      "learning_rate": 1.4577061883949912e-05,
+      "loss": 0.334,
+      "step": 1914
+    },
+    {
+      "epoch": 0.36841092727972297,
+      "grad_norm": 1.3919814672119775,
+      "learning_rate": 1.4571521128618358e-05,
+      "loss": 0.8405,
+      "step": 1915
+    },
+    {
+      "epoch": 0.36860330896498655,
+      "grad_norm": 1.5995324751082896,
+      "learning_rate": 1.4565978598464895e-05,
+      "loss": 0.911,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3687956906502501,
+      "grad_norm": 1.5754023727711584,
+      "learning_rate": 1.4560434295641338e-05,
+      "loss": 0.8758,
+      "step": 1917
+    },
+    {
+      "epoch": 0.3689880723355137,
+      "grad_norm": 1.4256115041703046,
+      "learning_rate": 1.455488822230016e-05,
+      "loss": 0.8544,
+      "step": 1918
+    },
+    {
+      "epoch": 0.3691804540207772,
+      "grad_norm": 1.5770284424801246,
+      "learning_rate": 1.4549340380594545e-05,
+      "loss": 0.8598,
+      "step": 1919
+    },
+    {
+      "epoch": 0.3693728357060408,
+      "grad_norm": 1.2385011694513344,
+      "learning_rate": 1.454379077267836e-05,
+      "loss": 0.8992,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3695652173913043,
+      "grad_norm": 1.7169854424841546,
+      "learning_rate": 1.4538239400706147e-05,
+      "loss": 0.8802,
+      "step": 1921
+    },
+    {
+      "epoch": 0.3697575990765679,
+      "grad_norm": 1.9179809900240614,
+      "learning_rate": 1.4532686266833143e-05,
+      "loss": 0.8871,
+      "step": 1922
+    },
+    {
+      "epoch": 0.3699499807618315,
+      "grad_norm": 0.7499271755269523,
+      "learning_rate": 1.4527131373215265e-05,
+      "loss": 0.3542,
+      "step": 1923
+    },
+    {
+      "epoch": 0.37014236244709503,
+      "grad_norm": 1.811189103189731,
+      "learning_rate": 1.4521574722009115e-05,
+      "loss": 0.9237,
+      "step": 1924
+    },
+    {
+      "epoch": 0.3703347441323586,
+      "grad_norm": 0.6085356343705514,
+      "learning_rate": 1.4516016315371974e-05,
+      "loss": 0.3269,
+      "step": 1925
+    },
+    {
+      "epoch": 0.37052712581762215,
+      "grad_norm": 2.28191332627133,
+      "learning_rate": 1.4510456155461807e-05,
+      "loss": 0.9307,
+      "step": 1926
+    },
+    {
+      "epoch": 0.37071950750288574,
+      "grad_norm": 1.3517270507937489,
+      "learning_rate": 1.4504894244437264e-05,
+      "loss": 0.881,
+      "step": 1927
+    },
+    {
+      "epoch": 0.37091188918814927,
+      "grad_norm": 1.6273825515886342,
+      "learning_rate": 1.4499330584457667e-05,
+      "loss": 0.9363,
+      "step": 1928
+    },
+    {
+      "epoch": 0.37110427087341286,
+      "grad_norm": 1.2390942021279368,
+      "learning_rate": 1.4493765177683017e-05,
+      "loss": 0.8926,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3712966525586764,
+      "grad_norm": 1.4328253008461043,
+      "learning_rate": 1.4488198026274007e-05,
+      "loss": 0.8981,
+      "step": 1930
+    },
+    {
+      "epoch": 0.37148903424394,
+      "grad_norm": 1.3982166183300195,
+      "learning_rate": 1.4482629132391985e-05,
+      "loss": 0.9175,
+      "step": 1931
+    },
+    {
+      "epoch": 0.37168141592920356,
+      "grad_norm": 1.3068870492426419,
+      "learning_rate": 1.4477058498198993e-05,
+      "loss": 0.8987,
+      "step": 1932
+    },
+    {
+      "epoch": 0.3718737976144671,
+      "grad_norm": 1.0308438655613066,
+      "learning_rate": 1.4471486125857743e-05,
+      "loss": 0.3568,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3720661792997307,
+      "grad_norm": 3.9777034044297315,
+      "learning_rate": 1.446591201753162e-05,
+      "loss": 0.8668,
+      "step": 1934
+    },
+    {
+      "epoch": 0.3722585609849942,
+      "grad_norm": 1.3862447798410487,
+      "learning_rate": 1.4460336175384688e-05,
+      "loss": 0.9205,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3724509426702578,
+      "grad_norm": 1.6922888243720389,
+      "learning_rate": 1.4454758601581675e-05,
+      "loss": 0.8809,
+      "step": 1936
+    },
+    {
+      "epoch": 0.37264332435552133,
+      "grad_norm": 1.5659559284945928,
+      "learning_rate": 1.4449179298287999e-05,
+      "loss": 0.9327,
+      "step": 1937
+    },
+    {
+      "epoch": 0.3728357060407849,
+      "grad_norm": 1.5060453666589764,
+      "learning_rate": 1.4443598267669723e-05,
+      "loss": 0.8817,
+      "step": 1938
+    },
+    {
+      "epoch": 0.3730280877260485,
+      "grad_norm": 2.5416760377054044,
+      "learning_rate": 1.4438015511893602e-05,
+      "loss": 0.8822,
+      "step": 1939
+    },
+    {
+      "epoch": 0.37322046941131204,
+      "grad_norm": 1.4073299655674156,
+      "learning_rate": 1.4432431033127056e-05,
+      "loss": 0.9428,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3734128510965756,
+      "grad_norm": 1.523629606851072,
+      "learning_rate": 1.442684483353817e-05,
+      "loss": 0.9024,
+      "step": 1941
+    },
+    {
+      "epoch": 0.37360523278183916,
+      "grad_norm": 1.8348738605425412,
+      "learning_rate": 1.4421256915295697e-05,
+      "loss": 0.8019,
+      "step": 1942
+    },
+    {
+      "epoch": 0.37379761446710275,
+      "grad_norm": 1.5079043257087204,
+      "learning_rate": 1.4415667280569064e-05,
+      "loss": 0.8253,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3739899961523663,
+      "grad_norm": 1.2282286465788097,
+      "learning_rate": 1.4410075931528356e-05,
+      "loss": 0.901,
+      "step": 1944
+    },
+    {
+      "epoch": 0.37418237783762986,
+      "grad_norm": 1.9347680095052222,
+      "learning_rate": 1.4404482870344322e-05,
+      "loss": 0.8911,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3743747595228934,
+      "grad_norm": 1.5893259342637545,
+      "learning_rate": 1.4398888099188396e-05,
+      "loss": 0.9532,
+      "step": 1946
+    },
+    {
+      "epoch": 0.374567141208157,
+      "grad_norm": 1.5642222215320782,
+      "learning_rate": 1.4393291620232646e-05,
+      "loss": 0.8934,
+      "step": 1947
+    },
+    {
+      "epoch": 0.37475952289342057,
+      "grad_norm": 1.4252561835002329,
+      "learning_rate": 1.4387693435649826e-05,
+      "loss": 0.9063,
+      "step": 1948
+    },
+    {
+      "epoch": 0.3749519045786841,
+      "grad_norm": 1.3198633133118272,
+      "learning_rate": 1.4382093547613338e-05,
+      "loss": 0.9395,
+      "step": 1949
+    },
+    {
+      "epoch": 0.3751442862639477,
+      "grad_norm": 1.6701751194516723,
+      "learning_rate": 1.4376491958297263e-05,
+      "loss": 0.9005,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3753366679492112,
+      "grad_norm": 1.4167844142135295,
+      "learning_rate": 1.4370888669876317e-05,
+      "loss": 0.9582,
+      "step": 1951
+    },
+    {
+      "epoch": 0.3755290496344748,
+      "grad_norm": 1.5729550599690052,
+      "learning_rate": 1.4365283684525895e-05,
+      "loss": 0.8844,
+      "step": 1952
+    },
+    {
+      "epoch": 0.37572143131973834,
+      "grad_norm": 1.6054864204746127,
+      "learning_rate": 1.4359677004422045e-05,
+      "loss": 0.8352,
+      "step": 1953
+    },
+    {
+      "epoch": 0.37591381300500193,
+      "grad_norm": 2.054152018120309,
+      "learning_rate": 1.4354068631741476e-05,
+      "loss": 0.9053,
+      "step": 1954
+    },
+    {
+      "epoch": 0.37610619469026546,
+      "grad_norm": 1.2758062592193775,
+      "learning_rate": 1.4348458568661548e-05,
+      "loss": 0.9886,
+      "step": 1955
+    },
+    {
+      "epoch": 0.37629857637552905,
+      "grad_norm": 1.4761282003785892,
+      "learning_rate": 1.434284681736028e-05,
+      "loss": 0.9048,
+      "step": 1956
+    },
+    {
+      "epoch": 0.37649095806079264,
+      "grad_norm": 1.277104071210093,
+      "learning_rate": 1.4337233380016354e-05,
+      "loss": 0.9162,
+      "step": 1957
+    },
+    {
+      "epoch": 0.37668333974605617,
+      "grad_norm": 1.2799262642735985,
+      "learning_rate": 1.433161825880909e-05,
+      "loss": 0.8883,
+      "step": 1958
+    },
+    {
+      "epoch": 0.37687572143131975,
+      "grad_norm": 1.5363652339306877,
+      "learning_rate": 1.432600145591848e-05,
+      "loss": 0.8984,
+      "step": 1959
+    },
+    {
+      "epoch": 0.3770681031165833,
+      "grad_norm": 0.7929370128491571,
+      "learning_rate": 1.4320382973525151e-05,
+      "loss": 0.3631,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3772604848018469,
+      "grad_norm": 1.3200279417403762,
+      "learning_rate": 1.43147628138104e-05,
+      "loss": 0.8967,
+      "step": 1961
+    },
+    {
+      "epoch": 0.3774528664871104,
+      "grad_norm": 1.2457605904142162,
+      "learning_rate": 1.4309140978956161e-05,
+      "loss": 0.9165,
+      "step": 1962
+    },
+    {
+      "epoch": 0.377645248172374,
+      "grad_norm": 1.1809577784747778,
+      "learning_rate": 1.430351747114503e-05,
+      "loss": 0.8947,
+      "step": 1963
+    },
+    {
+      "epoch": 0.3778376298576376,
+      "grad_norm": 1.3167461575997783,
+      "learning_rate": 1.429789229256024e-05,
+      "loss": 0.8667,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3780300115429011,
+      "grad_norm": 0.62012046332152,
+      "learning_rate": 1.429226544538568e-05,
+      "loss": 0.2785,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3782223932281647,
+      "grad_norm": 0.6331493355101543,
+      "learning_rate": 1.4286636931805887e-05,
+      "loss": 0.3014,
+      "step": 1966
+    },
+    {
+      "epoch": 0.37841477491342823,
+      "grad_norm": 1.2304507060538417,
+      "learning_rate": 1.4281006754006045e-05,
+      "loss": 0.8779,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3786071565986918,
+      "grad_norm": 1.2587860566803906,
+      "learning_rate": 1.427537491417198e-05,
+      "loss": 0.9076,
+      "step": 1968
+    },
+    {
+      "epoch": 0.37879953828395535,
+      "grad_norm": 1.2936413852769317,
+      "learning_rate": 1.426974141449017e-05,
+      "loss": 0.8297,
+      "step": 1969
+    },
+    {
+      "epoch": 0.37899191996921894,
+      "grad_norm": 1.4023596917572205,
+      "learning_rate": 1.4264106257147732e-05,
+      "loss": 0.9103,
+      "step": 1970
+    },
+    {
+      "epoch": 0.37918430165448247,
+      "grad_norm": 1.4956924793445112,
+      "learning_rate": 1.4258469444332423e-05,
+      "loss": 0.894,
+      "step": 1971
+    },
+    {
+      "epoch": 0.37937668333974606,
+      "grad_norm": 1.3804349828982128,
+      "learning_rate": 1.4252830978232658e-05,
+      "loss": 0.9416,
+      "step": 1972
+    },
+    {
+      "epoch": 0.37956906502500964,
+      "grad_norm": 1.2314510090265682,
+      "learning_rate": 1.4247190861037474e-05,
+      "loss": 0.8746,
+      "step": 1973
+    },
+    {
+      "epoch": 0.3797614467102732,
+      "grad_norm": 1.4824153788156924,
+      "learning_rate": 1.4241549094936567e-05,
+      "loss": 0.9268,
+      "step": 1974
+    },
+    {
+      "epoch": 0.37995382839553676,
+      "grad_norm": 1.3517294893112375,
+      "learning_rate": 1.4235905682120255e-05,
+      "loss": 0.8718,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3801462100808003,
+      "grad_norm": 1.3823151492815804,
+      "learning_rate": 1.4230260624779512e-05,
+      "loss": 0.9061,
+      "step": 1976
+    },
+    {
+      "epoch": 0.3803385917660639,
+      "grad_norm": 1.2007727305750977,
+      "learning_rate": 1.4224613925105947e-05,
+      "loss": 0.8642,
+      "step": 1977
+    },
+    {
+      "epoch": 0.3805309734513274,
+      "grad_norm": 1.2996283514927767,
+      "learning_rate": 1.4218965585291792e-05,
+      "loss": 0.8912,
+      "step": 1978
+    },
+    {
+      "epoch": 0.380723355136591,
+      "grad_norm": 1.3217820115600663,
+      "learning_rate": 1.4213315607529939e-05,
+      "loss": 0.896,
+      "step": 1979
+    },
+    {
+      "epoch": 0.38091573682185453,
+      "grad_norm": 2.3469584285742595,
+      "learning_rate": 1.4207663994013896e-05,
+      "loss": 0.8891,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3811081185071181,
+      "grad_norm": 1.4249986323603105,
+      "learning_rate": 1.4202010746937815e-05,
+      "loss": 0.8988,
+      "step": 1981
+    },
+    {
+      "epoch": 0.3813005001923817,
+      "grad_norm": 1.2405360633497635,
+      "learning_rate": 1.4196355868496485e-05,
+      "loss": 0.9153,
+      "step": 1982
+    },
+    {
+      "epoch": 0.38149288187764524,
+      "grad_norm": 1.3686078083706685,
+      "learning_rate": 1.4190699360885323e-05,
+      "loss": 0.9203,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3816852635629088,
+      "grad_norm": 1.3188433541795308,
+      "learning_rate": 1.4185041226300376e-05,
+      "loss": 0.8246,
+      "step": 1984
+    },
+    {
+      "epoch": 0.38187764524817236,
+      "grad_norm": 1.5750607963542544,
+      "learning_rate": 1.4179381466938332e-05,
+      "loss": 0.8466,
+      "step": 1985
+    },
+    {
+      "epoch": 0.38207002693343595,
+      "grad_norm": 2.2305825945192224,
+      "learning_rate": 1.4173720084996501e-05,
+      "loss": 0.8804,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3822624086186995,
+      "grad_norm": 1.4164246490769858,
+      "learning_rate": 1.4168057082672828e-05,
+      "loss": 0.9558,
+      "step": 1987
+    },
+    {
+      "epoch": 0.38245479030396307,
+      "grad_norm": 1.4104211042736776,
+      "learning_rate": 1.4162392462165884e-05,
+      "loss": 0.9103,
+      "step": 1988
+    },
+    {
+      "epoch": 0.38264717198922665,
+      "grad_norm": 1.3140510227339528,
+      "learning_rate": 1.4156726225674874e-05,
+      "loss": 0.8565,
+      "step": 1989
+    },
+    {
+      "epoch": 0.3828395536744902,
+      "grad_norm": 1.4611595273877827,
+      "learning_rate": 1.415105837539962e-05,
+      "loss": 0.9033,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3830319353597538,
+      "grad_norm": 1.3585857555842602,
+      "learning_rate": 1.414538891354058e-05,
+      "loss": 0.8832,
+      "step": 1991
+    },
+    {
+      "epoch": 0.3832243170450173,
+      "grad_norm": 1.224793765921781,
+      "learning_rate": 1.4139717842298835e-05,
+      "loss": 0.9015,
+      "step": 1992
+    },
+    {
+      "epoch": 0.3834166987302809,
+      "grad_norm": 1.1463457706614024,
+      "learning_rate": 1.4134045163876086e-05,
+      "loss": 0.912,
+      "step": 1993
+    },
+    {
+      "epoch": 0.3836090804155444,
+      "grad_norm": 1.3936237802863363,
+      "learning_rate": 1.4128370880474667e-05,
+      "loss": 0.8911,
+      "step": 1994
+    },
+    {
+      "epoch": 0.383801462100808,
+      "grad_norm": 1.2947431624449537,
+      "learning_rate": 1.412269499429753e-05,
+      "loss": 0.8781,
+      "step": 1995
+    },
+    {
+      "epoch": 0.38399384378607154,
+      "grad_norm": 1.3546834895252364,
+      "learning_rate": 1.4117017507548244e-05,
+      "loss": 0.9338,
+      "step": 1996
+    },
+    {
+      "epoch": 0.38418622547133513,
+      "grad_norm": 5.207971788231757,
+      "learning_rate": 1.4111338422431013e-05,
+      "loss": 0.9213,
+      "step": 1997
+    },
+    {
+      "epoch": 0.3843786071565987,
+      "grad_norm": 1.4288010224649883,
+      "learning_rate": 1.4105657741150648e-05,
+      "loss": 0.852,
+      "step": 1998
+    },
+    {
+      "epoch": 0.38457098884186225,
+      "grad_norm": 1.2295656167269189,
+      "learning_rate": 1.4099975465912584e-05,
+      "loss": 0.9027,
+      "step": 1999
+    },
+    {
+      "epoch": 0.38476337052712584,
+      "grad_norm": 1.7324199955623272,
+      "learning_rate": 1.4094291598922877e-05,
+      "loss": 0.8761,
+      "step": 2000
+    },
+    {
+      "epoch": 0.38495575221238937,
+      "grad_norm": 1.8374153682303218,
+      "learning_rate": 1.40886061423882e-05,
+      "loss": 0.8762,
+      "step": 2001
+    },
+    {
+      "epoch": 0.38514813389765296,
+      "grad_norm": 1.2670987959981812,
+      "learning_rate": 1.4082919098515846e-05,
+      "loss": 0.8653,
+      "step": 2002
+    },
+    {
+      "epoch": 0.3853405155829165,
+      "grad_norm": 1.3001713826963495,
+      "learning_rate": 1.407723046951372e-05,
+      "loss": 0.8362,
+      "step": 2003
+    },
+    {
+      "epoch": 0.3855328972681801,
+      "grad_norm": 1.522664154820561,
+      "learning_rate": 1.4071540257590341e-05,
+      "loss": 0.8988,
+      "step": 2004
+    },
+    {
+      "epoch": 0.3857252789534436,
+      "grad_norm": 1.498459335348674,
+      "learning_rate": 1.4065848464954848e-05,
+      "loss": 0.9339,
+      "step": 2005
+    },
+    {
+      "epoch": 0.3859176606387072,
+      "grad_norm": 1.5503847535266189,
+      "learning_rate": 1.4060155093816988e-05,
+      "loss": 0.8568,
+      "step": 2006
+    },
+    {
+      "epoch": 0.3861100423239708,
+      "grad_norm": 1.0994414090154603,
+      "learning_rate": 1.4054460146387124e-05,
+      "loss": 0.346,
+      "step": 2007
+    },
+    {
+      "epoch": 0.3863024240092343,
+      "grad_norm": 1.6068243368003714,
+      "learning_rate": 1.4048763624876233e-05,
+      "loss": 0.8353,
+      "step": 2008
+    },
+    {
+      "epoch": 0.3864948056944979,
+      "grad_norm": 2.5655003351302748,
+      "learning_rate": 1.4043065531495904e-05,
+      "loss": 0.8641,
+      "step": 2009
+    },
+    {
+      "epoch": 0.38668718737976143,
+      "grad_norm": 1.4796770036338243,
+      "learning_rate": 1.4037365868458325e-05,
+      "loss": 0.8681,
+      "step": 2010
+    },
+    {
+      "epoch": 0.386879569065025,
+      "grad_norm": 1.4354756604049064,
+      "learning_rate": 1.4031664637976305e-05,
+      "loss": 0.8509,
+      "step": 2011
+    },
+    {
+      "epoch": 0.38707195075028855,
+      "grad_norm": 0.6444972664549111,
+      "learning_rate": 1.402596184226326e-05,
+      "loss": 0.3537,
+      "step": 2012
+    },
+    {
+      "epoch": 0.38726433243555214,
+      "grad_norm": 1.206378841912582,
+      "learning_rate": 1.4020257483533208e-05,
+      "loss": 0.8853,
+      "step": 2013
+    },
+    {
+      "epoch": 0.38745671412081567,
+      "grad_norm": 1.7955630538888556,
+      "learning_rate": 1.401455156400078e-05,
+      "loss": 0.8844,
+      "step": 2014
+    },
+    {
+      "epoch": 0.38764909580607926,
+      "grad_norm": 1.7109830854303283,
+      "learning_rate": 1.400884408588121e-05,
+      "loss": 0.8904,
+      "step": 2015
+    },
+    {
+      "epoch": 0.38784147749134285,
+      "grad_norm": 1.3947176891020034,
+      "learning_rate": 1.400313505139034e-05,
+      "loss": 0.9352,
+      "step": 2016
+    },
+    {
+      "epoch": 0.3880338591766064,
+      "grad_norm": 1.2613205986640454,
+      "learning_rate": 1.3997424462744607e-05,
+      "loss": 0.8919,
+      "step": 2017
+    },
+    {
+      "epoch": 0.38822624086186996,
+      "grad_norm": 1.2608614384381032,
+      "learning_rate": 1.3991712322161065e-05,
+      "loss": 0.9187,
+      "step": 2018
+    },
+    {
+      "epoch": 0.3884186225471335,
+      "grad_norm": 1.7939594810426236,
+      "learning_rate": 1.3985998631857359e-05,
+      "loss": 0.905,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3886110042323971,
+      "grad_norm": 1.384236385375243,
+      "learning_rate": 1.398028339405174e-05,
+      "loss": 0.9103,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3888033859176606,
+      "grad_norm": 0.8053585444222687,
+      "learning_rate": 1.3974566610963068e-05,
+      "loss": 0.3661,
+      "step": 2021
+    },
+    {
+      "epoch": 0.3889957676029242,
+      "grad_norm": 3.008057263317829,
+      "learning_rate": 1.3968848284810785e-05,
+      "loss": 0.8843,
+      "step": 2022
+    },
+    {
+      "epoch": 0.3891881492881878,
+      "grad_norm": 1.4744297866160323,
+      "learning_rate": 1.3963128417814951e-05,
+      "loss": 0.8634,
+      "step": 2023
+    },
+    {
+      "epoch": 0.3893805309734513,
+      "grad_norm": 1.4563144460874502,
+      "learning_rate": 1.3957407012196204e-05,
+      "loss": 0.9781,
+      "step": 2024
+    },
+    {
+      "epoch": 0.3895729126587149,
+      "grad_norm": 1.3131922981046724,
+      "learning_rate": 1.3951684070175802e-05,
+      "loss": 0.9376,
+      "step": 2025
+    },
+    {
+      "epoch": 0.38976529434397844,
+      "grad_norm": 1.581190372888405,
+      "learning_rate": 1.3945959593975582e-05,
+      "loss": 0.952,
+      "step": 2026
+    },
+    {
+      "epoch": 0.38995767602924203,
+      "grad_norm": 1.4248169998034352,
+      "learning_rate": 1.3940233585817984e-05,
+      "loss": 0.8632,
+      "step": 2027
+    },
+    {
+      "epoch": 0.39015005771450556,
+      "grad_norm": 1.5327615258233696,
+      "learning_rate": 1.3934506047926042e-05,
+      "loss": 0.8482,
+      "step": 2028
+    },
+    {
+      "epoch": 0.39034243939976915,
+      "grad_norm": 3.8205710947364206,
+      "learning_rate": 1.3928776982523384e-05,
+      "loss": 0.914,
+      "step": 2029
+    },
+    {
+      "epoch": 0.3905348210850327,
+      "grad_norm": 1.3799403502340155,
+      "learning_rate": 1.3923046391834229e-05,
+      "loss": 0.9378,
+      "step": 2030
+    },
+    {
+      "epoch": 0.39072720277029627,
+      "grad_norm": 1.2725860083313856,
+      "learning_rate": 1.3917314278083391e-05,
+      "loss": 0.839,
+      "step": 2031
+    },
+    {
+      "epoch": 0.39091958445555985,
+      "grad_norm": 0.6695120438692134,
+      "learning_rate": 1.3911580643496272e-05,
+      "loss": 0.3641,
+      "step": 2032
+    },
+    {
+      "epoch": 0.3911119661408234,
+      "grad_norm": 1.4115460215153948,
+      "learning_rate": 1.3905845490298867e-05,
+      "loss": 0.8327,
+      "step": 2033
+    },
+    {
+      "epoch": 0.391304347826087,
+      "grad_norm": 1.5876843180157694,
+      "learning_rate": 1.390010882071776e-05,
+      "loss": 0.936,
+      "step": 2034
+    },
+    {
+      "epoch": 0.3914967295113505,
+      "grad_norm": 0.6494951495829865,
+      "learning_rate": 1.3894370636980128e-05,
+      "loss": 0.3614,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3916891111966141,
+      "grad_norm": 1.1087591451022922,
+      "learning_rate": 1.3888630941313728e-05,
+      "loss": 0.8611,
+      "step": 2036
+    },
+    {
+      "epoch": 0.3918814928818776,
+      "grad_norm": 1.749623886640572,
+      "learning_rate": 1.3882889735946901e-05,
+      "loss": 0.9852,
+      "step": 2037
+    },
+    {
+      "epoch": 0.3920738745671412,
+      "grad_norm": 0.6609089565686378,
+      "learning_rate": 1.3877147023108592e-05,
+      "loss": 0.3226,
+      "step": 2038
+    },
+    {
+      "epoch": 0.39226625625240474,
+      "grad_norm": 1.3659289405215547,
+      "learning_rate": 1.3871402805028314e-05,
+      "loss": 0.9473,
+      "step": 2039
+    },
+    {
+      "epoch": 0.39245863793766833,
+      "grad_norm": 0.6371456701304344,
+      "learning_rate": 1.3865657083936167e-05,
+      "loss": 0.353,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3926510196229319,
+      "grad_norm": 1.3027794779172965,
+      "learning_rate": 1.3859909862062844e-05,
+      "loss": 0.9009,
+      "step": 2041
+    },
+    {
+      "epoch": 0.39284340130819545,
+      "grad_norm": 1.219533885984043,
+      "learning_rate": 1.385416114163961e-05,
+      "loss": 0.9061,
+      "step": 2042
+    },
+    {
+      "epoch": 0.39303578299345904,
+      "grad_norm": 1.8925959854275642,
+      "learning_rate": 1.3848410924898321e-05,
+      "loss": 0.8838,
+      "step": 2043
+    },
+    {
+      "epoch": 0.39322816467872257,
+      "grad_norm": 1.9323305264516677,
+      "learning_rate": 1.3842659214071406e-05,
+      "loss": 0.8652,
+      "step": 2044
+    },
+    {
+      "epoch": 0.39342054636398616,
+      "grad_norm": 1.285889527440245,
+      "learning_rate": 1.3836906011391878e-05,
+      "loss": 0.9091,
+      "step": 2045
+    },
+    {
+      "epoch": 0.3936129280492497,
+      "grad_norm": 1.2528554386604684,
+      "learning_rate": 1.3831151319093323e-05,
+      "loss": 0.911,
+      "step": 2046
+    },
+    {
+      "epoch": 0.3938053097345133,
+      "grad_norm": 1.9608255074244703,
+      "learning_rate": 1.382539513940992e-05,
+      "loss": 0.8446,
+      "step": 2047
+    },
+    {
+      "epoch": 0.39399769141977686,
+      "grad_norm": 1.4381419069886257,
+      "learning_rate": 1.3819637474576411e-05,
+      "loss": 0.9286,
+      "step": 2048
+    },
+    {
+      "epoch": 0.3941900731050404,
+      "grad_norm": 3.0253115979626055,
+      "learning_rate": 1.381387832682812e-05,
+      "loss": 0.8534,
+      "step": 2049
+    },
+    {
+      "epoch": 0.394382454790304,
+      "grad_norm": 1.2606544038254077,
+      "learning_rate": 1.380811769840095e-05,
+      "loss": 0.8628,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3945748364755675,
+      "grad_norm": 1.416749761754615,
+      "learning_rate": 1.3802355591531366e-05,
+      "loss": 0.8886,
+      "step": 2051
+    },
+    {
+      "epoch": 0.3947672181608311,
+      "grad_norm": 1.9899547813232414,
+      "learning_rate": 1.3796592008456427e-05,
+      "loss": 0.8731,
+      "step": 2052
+    },
+    {
+      "epoch": 0.39495959984609463,
+      "grad_norm": 0.6712613280128646,
+      "learning_rate": 1.3790826951413747e-05,
+      "loss": 0.3555,
+      "step": 2053
+    },
+    {
+      "epoch": 0.3951519815313582,
+      "grad_norm": 1.3506057881770916,
+      "learning_rate": 1.3785060422641526e-05,
+      "loss": 0.8574,
+      "step": 2054
+    },
+    {
+      "epoch": 0.39534436321662175,
+      "grad_norm": 1.6208677403557492,
+      "learning_rate": 1.3779292424378521e-05,
+      "loss": 0.8623,
+      "step": 2055
+    },
+    {
+      "epoch": 0.39553674490188534,
+      "grad_norm": 2.28053491634738,
+      "learning_rate": 1.3773522958864076e-05,
+      "loss": 0.8632,
+      "step": 2056
+    },
+    {
+      "epoch": 0.3957291265871489,
+      "grad_norm": 1.340807114196148,
+      "learning_rate": 1.3767752028338091e-05,
+      "loss": 0.9073,
+      "step": 2057
+    },
+    {
+      "epoch": 0.39592150827241246,
+      "grad_norm": 1.2860649363401373,
+      "learning_rate": 1.376197963504104e-05,
+      "loss": 0.9494,
+      "step": 2058
+    },
+    {
+      "epoch": 0.39611388995767605,
+      "grad_norm": 1.155569501209257,
+      "learning_rate": 1.3756205781213965e-05,
+      "loss": 0.8794,
+      "step": 2059
+    },
+    {
+      "epoch": 0.3963062716429396,
+      "grad_norm": 1.8344863639357862,
+      "learning_rate": 1.375043046909848e-05,
+      "loss": 0.8717,
+      "step": 2060
+    },
+    {
+      "epoch": 0.39649865332820317,
+      "grad_norm": 1.3298337060008356,
+      "learning_rate": 1.3744653700936752e-05,
+      "loss": 0.8961,
+      "step": 2061
+    },
+    {
+      "epoch": 0.3966910350134667,
+      "grad_norm": 1.7798715659418216,
+      "learning_rate": 1.3738875478971526e-05,
+      "loss": 0.8793,
+      "step": 2062
+    },
+    {
+      "epoch": 0.3968834166987303,
+      "grad_norm": 1.2425765720848025,
+      "learning_rate": 1.3733095805446107e-05,
+      "loss": 0.9666,
+      "step": 2063
+    },
+    {
+      "epoch": 0.3970757983839938,
+      "grad_norm": 1.8988115575134705,
+      "learning_rate": 1.372731468260436e-05,
+      "loss": 0.8526,
+      "step": 2064
+    },
+    {
+      "epoch": 0.3972681800692574,
+      "grad_norm": 1.692609214190637,
+      "learning_rate": 1.372153211269072e-05,
+      "loss": 0.8957,
+      "step": 2065
+    },
+    {
+      "epoch": 0.397460561754521,
+      "grad_norm": 1.5051043337575596,
+      "learning_rate": 1.3715748097950176e-05,
+      "loss": 0.8837,
+      "step": 2066
+    },
+    {
+      "epoch": 0.3976529434397845,
+      "grad_norm": 1.4444672121438287,
+      "learning_rate": 1.3709962640628284e-05,
+      "loss": 0.9278,
+      "step": 2067
+    },
+    {
+      "epoch": 0.3978453251250481,
+      "grad_norm": 1.5218728739850902,
+      "learning_rate": 1.3704175742971158e-05,
+      "loss": 0.9175,
+      "step": 2068
+    },
+    {
+      "epoch": 0.39803770681031164,
+      "grad_norm": 1.5340968432825217,
+      "learning_rate": 1.369838740722547e-05,
+      "loss": 0.862,
+      "step": 2069
+    },
+    {
+      "epoch": 0.39823008849557523,
+      "grad_norm": 1.46329632505005,
+      "learning_rate": 1.3692597635638452e-05,
+      "loss": 0.822,
+      "step": 2070
+    },
+    {
+      "epoch": 0.39842247018083876,
+      "grad_norm": 1.9384230950768706,
+      "learning_rate": 1.368680643045789e-05,
+      "loss": 0.8671,
+      "step": 2071
+    },
+    {
+      "epoch": 0.39861485186610235,
+      "grad_norm": 1.2493223896439973,
+      "learning_rate": 1.3681013793932132e-05,
+      "loss": 0.8788,
+      "step": 2072
+    },
+    {
+      "epoch": 0.39880723355136594,
+      "grad_norm": 1.4398123039287627,
+      "learning_rate": 1.3675219728310076e-05,
+      "loss": 0.8991,
+      "step": 2073
+    },
+    {
+      "epoch": 0.39899961523662947,
+      "grad_norm": 1.3159293098818021,
+      "learning_rate": 1.3669424235841185e-05,
+      "loss": 0.8786,
+      "step": 2074
+    },
+    {
+      "epoch": 0.39919199692189306,
+      "grad_norm": 1.658472358486151,
+      "learning_rate": 1.3663627318775459e-05,
+      "loss": 0.8815,
+      "step": 2075
+    },
+    {
+      "epoch": 0.3993843786071566,
+      "grad_norm": 1.2524779047982129,
+      "learning_rate": 1.3657828979363468e-05,
+      "loss": 0.8988,
+      "step": 2076
+    },
+    {
+      "epoch": 0.3995767602924202,
+      "grad_norm": 1.434688610723458,
+      "learning_rate": 1.3652029219856324e-05,
+      "loss": 0.8965,
+      "step": 2077
+    },
+    {
+      "epoch": 0.3997691419776837,
+      "grad_norm": 0.677787214539511,
+      "learning_rate": 1.3646228042505694e-05,
+      "loss": 0.382,
+      "step": 2078
+    },
+    {
+      "epoch": 0.3999615236629473,
+      "grad_norm": 1.3651407161853006,
+      "learning_rate": 1.3640425449563793e-05,
+      "loss": 0.8438,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4001539053482108,
+      "grad_norm": 1.3794626522609685,
+      "learning_rate": 1.3634621443283389e-05,
+      "loss": 0.8641,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4003462870334744,
+      "grad_norm": 1.471105537189052,
+      "learning_rate": 1.36288160259178e-05,
+      "loss": 0.8919,
+      "step": 2081
+    },
+    {
+      "epoch": 0.400538668718738,
+      "grad_norm": 1.4388899533807413,
+      "learning_rate": 1.3623009199720882e-05,
+      "loss": 0.8902,
+      "step": 2082
+    },
+    {
+      "epoch": 0.40073105040400153,
+      "grad_norm": 1.0497019729275936,
+      "learning_rate": 1.3617200966947053e-05,
+      "loss": 0.9074,
+      "step": 2083
+    },
+    {
+      "epoch": 0.4009234320892651,
+      "grad_norm": 1.0825678839201744,
+      "learning_rate": 1.3611391329851262e-05,
+      "loss": 0.8961,
+      "step": 2084
+    },
+    {
+      "epoch": 0.40111581377452865,
+      "grad_norm": 1.3764334099528348,
+      "learning_rate": 1.3605580290689013e-05,
+      "loss": 0.8849,
+      "step": 2085
+    },
+    {
+      "epoch": 0.40130819545979224,
+      "grad_norm": 1.3197976297461058,
+      "learning_rate": 1.3599767851716353e-05,
+      "loss": 0.9117,
+      "step": 2086
+    },
+    {
+      "epoch": 0.40150057714505577,
+      "grad_norm": 1.3276693253380616,
+      "learning_rate": 1.3593954015189867e-05,
+      "loss": 0.8965,
+      "step": 2087
+    },
+    {
+      "epoch": 0.40169295883031936,
+      "grad_norm": 1.163163728774223,
+      "learning_rate": 1.3588138783366692e-05,
+      "loss": 0.9088,
+      "step": 2088
+    },
+    {
+      "epoch": 0.4018853405155829,
+      "grad_norm": 1.4578564899897737,
+      "learning_rate": 1.3582322158504495e-05,
+      "loss": 0.8886,
+      "step": 2089
+    },
+    {
+      "epoch": 0.4020777222008465,
+      "grad_norm": 1.2481314150985396,
+      "learning_rate": 1.3576504142861496e-05,
+      "loss": 0.8785,
+      "step": 2090
+    },
+    {
+      "epoch": 0.40227010388611006,
+      "grad_norm": 1.179059088596955,
+      "learning_rate": 1.3570684738696444e-05,
+      "loss": 0.9502,
+      "step": 2091
+    },
+    {
+      "epoch": 0.4024624855713736,
+      "grad_norm": 2.3205693079889653,
+      "learning_rate": 1.3564863948268631e-05,
+      "loss": 0.8543,
+      "step": 2092
+    },
+    {
+      "epoch": 0.4026548672566372,
+      "grad_norm": 2.2105641767612556,
+      "learning_rate": 1.3559041773837898e-05,
+      "loss": 0.9628,
+      "step": 2093
+    },
+    {
+      "epoch": 0.4028472489419007,
+      "grad_norm": 1.198827909995869,
+      "learning_rate": 1.3553218217664603e-05,
+      "loss": 0.8483,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4030396306271643,
+      "grad_norm": 1.3382260058380104,
+      "learning_rate": 1.3547393282009656e-05,
+      "loss": 0.8852,
+      "step": 2095
+    },
+    {
+      "epoch": 0.40323201231242783,
+      "grad_norm": 3.708542537552998,
+      "learning_rate": 1.3541566969134496e-05,
+      "loss": 0.778,
+      "step": 2096
+    },
+    {
+      "epoch": 0.4034243939976914,
+      "grad_norm": 1.214292285746447,
+      "learning_rate": 1.3535739281301102e-05,
+      "loss": 0.903,
+      "step": 2097
+    },
+    {
+      "epoch": 0.403616775682955,
+      "grad_norm": 1.7651274029658797,
+      "learning_rate": 1.3529910220771975e-05,
+      "loss": 0.8316,
+      "step": 2098
+    },
+    {
+      "epoch": 0.40380915736821854,
+      "grad_norm": 1.4089205332244634,
+      "learning_rate": 1.3524079789810163e-05,
+      "loss": 0.8431,
+      "step": 2099
+    },
+    {
+      "epoch": 0.40400153905348213,
+      "grad_norm": 1.3868637045508558,
+      "learning_rate": 1.3518247990679241e-05,
+      "loss": 0.8681,
+      "step": 2100
+    },
+    {
+      "epoch": 0.40419392073874566,
+      "grad_norm": 1.0708339892496472,
+      "learning_rate": 1.3512414825643312e-05,
+      "loss": 0.8923,
+      "step": 2101
+    },
+    {
+      "epoch": 0.40438630242400925,
+      "grad_norm": 2.4472478001273186,
+      "learning_rate": 1.3506580296967011e-05,
+      "loss": 0.8807,
+      "step": 2102
+    },
+    {
+      "epoch": 0.4045786841092728,
+      "grad_norm": 1.3174929354540135,
+      "learning_rate": 1.3500744406915505e-05,
+      "loss": 0.8188,
+      "step": 2103
+    },
+    {
+      "epoch": 0.40477106579453637,
+      "grad_norm": 1.2856724755864097,
+      "learning_rate": 1.3494907157754485e-05,
+      "loss": 0.91,
+      "step": 2104
+    },
+    {
+      "epoch": 0.4049634474797999,
+      "grad_norm": 1.6953519257148408,
+      "learning_rate": 1.348906855175017e-05,
+      "loss": 0.8627,
+      "step": 2105
+    },
+    {
+      "epoch": 0.4051558291650635,
+      "grad_norm": 1.6667115523455522,
+      "learning_rate": 1.3483228591169315e-05,
+      "loss": 0.8789,
+      "step": 2106
+    },
+    {
+      "epoch": 0.4053482108503271,
+      "grad_norm": 1.5594723629354608,
+      "learning_rate": 1.347738727827919e-05,
+      "loss": 0.8092,
+      "step": 2107
+    },
+    {
+      "epoch": 0.4055405925355906,
+      "grad_norm": 1.4699928309247843,
+      "learning_rate": 1.3471544615347591e-05,
+      "loss": 0.8383,
+      "step": 2108
+    },
+    {
+      "epoch": 0.4057329742208542,
+      "grad_norm": 1.343097426101094,
+      "learning_rate": 1.3465700604642847e-05,
+      "loss": 0.8588,
+      "step": 2109
+    },
+    {
+      "epoch": 0.4059253559061177,
+      "grad_norm": 1.8261735784896596,
+      "learning_rate": 1.34598552484338e-05,
+      "loss": 0.8809,
+      "step": 2110
+    },
+    {
+      "epoch": 0.4061177375913813,
+      "grad_norm": 0.6424698067969271,
+      "learning_rate": 1.3454008548989816e-05,
+      "loss": 0.3324,
+      "step": 2111
+    },
+    {
+      "epoch": 0.40631011927664484,
+      "grad_norm": 1.493109295273804,
+      "learning_rate": 1.3448160508580789e-05,
+      "loss": 0.8494,
+      "step": 2112
+    },
+    {
+      "epoch": 0.40650250096190843,
+      "grad_norm": 1.3756732748743181,
+      "learning_rate": 1.3442311129477133e-05,
+      "loss": 0.9392,
+      "step": 2113
+    },
+    {
+      "epoch": 0.40669488264717196,
+      "grad_norm": 0.6529561839363538,
+      "learning_rate": 1.343646041394977e-05,
+      "loss": 0.3325,
+      "step": 2114
+    },
+    {
+      "epoch": 0.40688726433243555,
+      "grad_norm": 1.88212512795477,
+      "learning_rate": 1.3430608364270156e-05,
+      "loss": 0.9171,
+      "step": 2115
+    },
+    {
+      "epoch": 0.40707964601769914,
+      "grad_norm": 1.5014329048633264,
+      "learning_rate": 1.3424754982710256e-05,
+      "loss": 0.9382,
+      "step": 2116
+    },
+    {
+      "epoch": 0.40727202770296267,
+      "grad_norm": 1.6423507337435355,
+      "learning_rate": 1.3418900271542552e-05,
+      "loss": 0.8648,
+      "step": 2117
+    },
+    {
+      "epoch": 0.40746440938822626,
+      "grad_norm": 1.1745773178497203,
+      "learning_rate": 1.3413044233040045e-05,
+      "loss": 0.9003,
+      "step": 2118
+    },
+    {
+      "epoch": 0.4076567910734898,
+      "grad_norm": 1.5977576518833394,
+      "learning_rate": 1.3407186869476253e-05,
+      "loss": 0.9114,
+      "step": 2119
+    },
+    {
+      "epoch": 0.4078491727587534,
+      "grad_norm": 1.437367132975582,
+      "learning_rate": 1.3401328183125208e-05,
+      "loss": 0.8282,
+      "step": 2120
+    },
+    {
+      "epoch": 0.4080415544440169,
+      "grad_norm": 2.293044906883491,
+      "learning_rate": 1.339546817626145e-05,
+      "loss": 0.8768,
+      "step": 2121
+    },
+    {
+      "epoch": 0.4082339361292805,
+      "grad_norm": 1.5018586042724833,
+      "learning_rate": 1.3389606851160037e-05,
+      "loss": 0.9079,
+      "step": 2122
+    },
+    {
+      "epoch": 0.4084263178145441,
+      "grad_norm": 1.7165295065553214,
+      "learning_rate": 1.3383744210096537e-05,
+      "loss": 0.8903,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4086186994998076,
+      "grad_norm": 1.288121941851918,
+      "learning_rate": 1.3377880255347026e-05,
+      "loss": 0.9005,
+      "step": 2124
+    },
+    {
+      "epoch": 0.4088110811850712,
+      "grad_norm": 1.1748278736978377,
+      "learning_rate": 1.3372014989188098e-05,
+      "loss": 0.8952,
+      "step": 2125
+    },
+    {
+      "epoch": 0.40900346287033473,
+      "grad_norm": 1.5463410608570642,
+      "learning_rate": 1.3366148413896851e-05,
+      "loss": 0.8893,
+      "step": 2126
+    },
+    {
+      "epoch": 0.4091958445555983,
+      "grad_norm": 2.566029971354862,
+      "learning_rate": 1.3360280531750886e-05,
+      "loss": 0.8619,
+      "step": 2127
+    },
+    {
+      "epoch": 0.40938822624086185,
+      "grad_norm": 1.3986524793092237,
+      "learning_rate": 1.3354411345028324e-05,
+      "loss": 0.993,
+      "step": 2128
+    },
+    {
+      "epoch": 0.40958060792612544,
+      "grad_norm": 1.1245785998338915,
+      "learning_rate": 1.3348540856007782e-05,
+      "loss": 0.9143,
+      "step": 2129
+    },
+    {
+      "epoch": 0.40977298961138897,
+      "grad_norm": 1.4944322832644654,
+      "learning_rate": 1.3342669066968385e-05,
+      "loss": 0.8706,
+      "step": 2130
+    },
+    {
+      "epoch": 0.40996537129665256,
+      "grad_norm": 1.2314058274236819,
+      "learning_rate": 1.3336795980189763e-05,
+      "loss": 0.8895,
+      "step": 2131
+    },
+    {
+      "epoch": 0.41015775298191615,
+      "grad_norm": 7.755077282283412,
+      "learning_rate": 1.3330921597952056e-05,
+      "loss": 0.884,
+      "step": 2132
+    },
+    {
+      "epoch": 0.4103501346671797,
+      "grad_norm": 1.317038057123304,
+      "learning_rate": 1.3325045922535896e-05,
+      "loss": 0.8596,
+      "step": 2133
+    },
+    {
+      "epoch": 0.41054251635244327,
+      "grad_norm": 1.5531787578474932,
+      "learning_rate": 1.3319168956222423e-05,
+      "loss": 0.9033,
+      "step": 2134
+    },
+    {
+      "epoch": 0.4107348980377068,
+      "grad_norm": 1.5932519142357977,
+      "learning_rate": 1.331329070129328e-05,
+      "loss": 0.9117,
+      "step": 2135
+    },
+    {
+      "epoch": 0.4109272797229704,
+      "grad_norm": 1.4643176433790541,
+      "learning_rate": 1.3307411160030608e-05,
+      "loss": 0.8787,
+      "step": 2136
+    },
+    {
+      "epoch": 0.4111196614082339,
+      "grad_norm": 1.4000007116616364,
+      "learning_rate": 1.3301530334717046e-05,
+      "loss": 0.9081,
+      "step": 2137
+    },
+    {
+      "epoch": 0.4113120430934975,
+      "grad_norm": 1.7589505846017006,
+      "learning_rate": 1.3295648227635729e-05,
+      "loss": 0.9291,
+      "step": 2138
+    },
+    {
+      "epoch": 0.41150442477876104,
+      "grad_norm": 2.790075271278033,
+      "learning_rate": 1.32897648410703e-05,
+      "loss": 0.8729,
+      "step": 2139
+    },
+    {
+      "epoch": 0.4116968064640246,
+      "grad_norm": 1.8241445247377435,
+      "learning_rate": 1.328388017730489e-05,
+      "loss": 0.8667,
+      "step": 2140
+    },
+    {
+      "epoch": 0.4118891881492882,
+      "grad_norm": 1.395566278134103,
+      "learning_rate": 1.327799423862413e-05,
+      "loss": 0.8673,
+      "step": 2141
+    },
+    {
+      "epoch": 0.41208156983455174,
+      "grad_norm": 0.6450886029087679,
+      "learning_rate": 1.3272107027313142e-05,
+      "loss": 0.3343,
+      "step": 2142
+    },
+    {
+      "epoch": 0.41227395151981533,
+      "grad_norm": 2.1937113934373684,
+      "learning_rate": 1.3266218545657541e-05,
+      "loss": 0.8992,
+      "step": 2143
+    },
+    {
+      "epoch": 0.41246633320507886,
+      "grad_norm": 1.5692739858314,
+      "learning_rate": 1.326032879594344e-05,
+      "loss": 0.9272,
+      "step": 2144
+    },
+    {
+      "epoch": 0.41265871489034245,
+      "grad_norm": 1.5746295387743192,
+      "learning_rate": 1.3254437780457448e-05,
+      "loss": 0.8506,
+      "step": 2145
+    },
+    {
+      "epoch": 0.412851096575606,
+      "grad_norm": 1.8859759293633744,
+      "learning_rate": 1.3248545501486654e-05,
+      "loss": 0.8817,
+      "step": 2146
+    },
+    {
+      "epoch": 0.41304347826086957,
+      "grad_norm": 1.972391466350968,
+      "learning_rate": 1.3242651961318646e-05,
+      "loss": 0.9041,
+      "step": 2147
+    },
+    {
+      "epoch": 0.41323585994613315,
+      "grad_norm": 2.2734913837649473,
+      "learning_rate": 1.32367571622415e-05,
+      "loss": 0.852,
+      "step": 2148
+    },
+    {
+      "epoch": 0.4134282416313967,
+      "grad_norm": 1.557962815801456,
+      "learning_rate": 1.3230861106543777e-05,
+      "loss": 0.9497,
+      "step": 2149
+    },
+    {
+      "epoch": 0.4136206233166603,
+      "grad_norm": 1.4267032764527496,
+      "learning_rate": 1.3224963796514532e-05,
+      "loss": 0.9079,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4138130050019238,
+      "grad_norm": 1.9338305757786105,
+      "learning_rate": 1.32190652344433e-05,
+      "loss": 0.8851,
+      "step": 2151
+    },
+    {
+      "epoch": 0.4140053866871874,
+      "grad_norm": 1.625231965096056,
+      "learning_rate": 1.3213165422620111e-05,
+      "loss": 0.8447,
+      "step": 2152
+    },
+    {
+      "epoch": 0.4141977683724509,
+      "grad_norm": 1.4458228445513215,
+      "learning_rate": 1.3207264363335472e-05,
+      "loss": 0.8765,
+      "step": 2153
+    },
+    {
+      "epoch": 0.4143901500577145,
+      "grad_norm": 0.6490443664250084,
+      "learning_rate": 1.3201362058880375e-05,
+      "loss": 0.3339,
+      "step": 2154
+    },
+    {
+      "epoch": 0.41458253174297804,
+      "grad_norm": 1.635966945565565,
+      "learning_rate": 1.3195458511546307e-05,
+      "loss": 0.8508,
+      "step": 2155
+    },
+    {
+      "epoch": 0.41477491342824163,
+      "grad_norm": 1.4138440773006287,
+      "learning_rate": 1.3189553723625217e-05,
+      "loss": 0.8562,
+      "step": 2156
+    },
+    {
+      "epoch": 0.4149672951135052,
+      "grad_norm": 1.2252004927560733,
+      "learning_rate": 1.318364769740955e-05,
+      "loss": 0.93,
+      "step": 2157
+    },
+    {
+      "epoch": 0.41515967679876875,
+      "grad_norm": 2.1857249771102683,
+      "learning_rate": 1.3177740435192235e-05,
+      "loss": 0.8533,
+      "step": 2158
+    },
+    {
+      "epoch": 0.41535205848403234,
+      "grad_norm": 1.5708143998090376,
+      "learning_rate": 1.3171831939266668e-05,
+      "loss": 0.9064,
+      "step": 2159
+    },
+    {
+      "epoch": 0.41554444016929587,
+      "grad_norm": 1.7076204383472102,
+      "learning_rate": 1.3165922211926734e-05,
+      "loss": 0.9392,
+      "step": 2160
+    },
+    {
+      "epoch": 0.41573682185455946,
+      "grad_norm": 1.3403669084746046,
+      "learning_rate": 1.3160011255466791e-05,
+      "loss": 0.8317,
+      "step": 2161
+    },
+    {
+      "epoch": 0.415929203539823,
+      "grad_norm": 2.5163926648897017,
+      "learning_rate": 1.3154099072181677e-05,
+      "loss": 0.8867,
+      "step": 2162
+    },
+    {
+      "epoch": 0.4161215852250866,
+      "grad_norm": 1.4798079459639486,
+      "learning_rate": 1.3148185664366704e-05,
+      "loss": 0.9078,
+      "step": 2163
+    },
+    {
+      "epoch": 0.4163139669103501,
+      "grad_norm": 2.0973361995180007,
+      "learning_rate": 1.314227103431766e-05,
+      "loss": 0.8493,
+      "step": 2164
+    },
+    {
+      "epoch": 0.4165063485956137,
+      "grad_norm": 0.6839444046119258,
+      "learning_rate": 1.3136355184330809e-05,
+      "loss": 0.322,
+      "step": 2165
+    },
+    {
+      "epoch": 0.4166987302808773,
+      "grad_norm": 1.5236235506463285,
+      "learning_rate": 1.3130438116702888e-05,
+      "loss": 0.8764,
+      "step": 2166
+    },
+    {
+      "epoch": 0.4168911119661408,
+      "grad_norm": 1.4641355591819054,
+      "learning_rate": 1.3124519833731106e-05,
+      "loss": 0.8766,
+      "step": 2167
+    },
+    {
+      "epoch": 0.4170834936514044,
+      "grad_norm": 1.3024302366810987,
+      "learning_rate": 1.3118600337713146e-05,
+      "loss": 0.8528,
+      "step": 2168
+    },
+    {
+      "epoch": 0.41727587533666793,
+      "grad_norm": 1.514719826191195,
+      "learning_rate": 1.3112679630947156e-05,
+      "loss": 0.8797,
+      "step": 2169
+    },
+    {
+      "epoch": 0.4174682570219315,
+      "grad_norm": 1.5657071618765253,
+      "learning_rate": 1.310675771573176e-05,
+      "loss": 0.9248,
+      "step": 2170
+    },
+    {
+      "epoch": 0.41766063870719505,
+      "grad_norm": 1.7388443341517252,
+      "learning_rate": 1.310083459436605e-05,
+      "loss": 0.8882,
+      "step": 2171
+    },
+    {
+      "epoch": 0.41785302039245864,
+      "grad_norm": 1.4273159087822656,
+      "learning_rate": 1.3094910269149587e-05,
+      "loss": 0.8381,
+      "step": 2172
+    },
+    {
+      "epoch": 0.41804540207772223,
+      "grad_norm": 1.2007701233505612,
+      "learning_rate": 1.3088984742382395e-05,
+      "loss": 0.8767,
+      "step": 2173
+    },
+    {
+      "epoch": 0.41823778376298576,
+      "grad_norm": 1.6521500054306724,
+      "learning_rate": 1.3083058016364972e-05,
+      "loss": 0.9006,
+      "step": 2174
+    },
+    {
+      "epoch": 0.41843016544824935,
+      "grad_norm": 1.5901903494688532,
+      "learning_rate": 1.3077130093398274e-05,
+      "loss": 0.9008,
+      "step": 2175
+    },
+    {
+      "epoch": 0.4186225471335129,
+      "grad_norm": 1.840047355367906,
+      "learning_rate": 1.3071200975783725e-05,
+      "loss": 0.8723,
+      "step": 2176
+    },
+    {
+      "epoch": 0.41881492881877647,
+      "grad_norm": 1.314719102128274,
+      "learning_rate": 1.3065270665823206e-05,
+      "loss": 0.9049,
+      "step": 2177
+    },
+    {
+      "epoch": 0.41900731050404,
+      "grad_norm": 1.5870267133064075,
+      "learning_rate": 1.3059339165819082e-05,
+      "loss": 0.863,
+      "step": 2178
+    },
+    {
+      "epoch": 0.4191996921893036,
+      "grad_norm": 1.4065263094919815,
+      "learning_rate": 1.3053406478074155e-05,
+      "loss": 0.9066,
+      "step": 2179
+    },
+    {
+      "epoch": 0.4193920738745671,
+      "grad_norm": 1.4871705060005638,
+      "learning_rate": 1.3047472604891701e-05,
+      "loss": 0.9224,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4195844555598307,
+      "grad_norm": 1.6165550135232662,
+      "learning_rate": 1.3041537548575455e-05,
+      "loss": 0.8519,
+      "step": 2181
+    },
+    {
+      "epoch": 0.4197768372450943,
+      "grad_norm": 1.4299107935390323,
+      "learning_rate": 1.303560131142961e-05,
+      "loss": 0.9105,
+      "step": 2182
+    },
+    {
+      "epoch": 0.4199692189303578,
+      "grad_norm": 1.7502204131169696,
+      "learning_rate": 1.3029663895758814e-05,
+      "loss": 0.9166,
+      "step": 2183
+    },
+    {
+      "epoch": 0.4201616006156214,
+      "grad_norm": 1.4050282930014455,
+      "learning_rate": 1.3023725303868183e-05,
+      "loss": 0.8453,
+      "step": 2184
+    },
+    {
+      "epoch": 0.42035398230088494,
+      "grad_norm": 1.5038088653437993,
+      "learning_rate": 1.3017785538063277e-05,
+      "loss": 0.8767,
+      "step": 2185
+    },
+    {
+      "epoch": 0.42054636398614853,
+      "grad_norm": 1.4755152274150602,
+      "learning_rate": 1.3011844600650121e-05,
+      "loss": 0.9001,
+      "step": 2186
+    },
+    {
+      "epoch": 0.42073874567141206,
+      "grad_norm": 1.3481148211377503,
+      "learning_rate": 1.300590249393519e-05,
+      "loss": 0.8999,
+      "step": 2187
+    },
+    {
+      "epoch": 0.42093112735667565,
+      "grad_norm": 1.3503004176437952,
+      "learning_rate": 1.2999959220225416e-05,
+      "loss": 0.9029,
+      "step": 2188
+    },
+    {
+      "epoch": 0.4211235090419392,
+      "grad_norm": 1.3202073412162245,
+      "learning_rate": 1.299401478182818e-05,
+      "loss": 0.8919,
+      "step": 2189
+    },
+    {
+      "epoch": 0.42131589072720277,
+      "grad_norm": 0.6370460644933824,
+      "learning_rate": 1.2988069181051314e-05,
+      "loss": 0.3435,
+      "step": 2190
+    },
+    {
+      "epoch": 0.42150827241246636,
+      "grad_norm": 1.138986221917855,
+      "learning_rate": 1.2982122420203114e-05,
+      "loss": 0.8706,
+      "step": 2191
+    },
+    {
+      "epoch": 0.4217006540977299,
+      "grad_norm": 1.3434251390671215,
+      "learning_rate": 1.2976174501592313e-05,
+      "loss": 0.8732,
+      "step": 2192
+    },
+    {
+      "epoch": 0.4218930357829935,
+      "grad_norm": 1.741812961162895,
+      "learning_rate": 1.2970225427528098e-05,
+      "loss": 0.8606,
+      "step": 2193
+    },
+    {
+      "epoch": 0.422085417468257,
+      "grad_norm": 1.2020494700916287,
+      "learning_rate": 1.2964275200320104e-05,
+      "loss": 0.8831,
+      "step": 2194
+    },
+    {
+      "epoch": 0.4222777991535206,
+      "grad_norm": 1.1402676244650696,
+      "learning_rate": 1.2958323822278413e-05,
+      "loss": 0.8753,
+      "step": 2195
+    },
+    {
+      "epoch": 0.4224701808387841,
+      "grad_norm": 2.3524681933416045,
+      "learning_rate": 1.2952371295713558e-05,
+      "loss": 0.9132,
+      "step": 2196
+    },
+    {
+      "epoch": 0.4226625625240477,
+      "grad_norm": 1.2758320397567096,
+      "learning_rate": 1.2946417622936512e-05,
+      "loss": 0.8726,
+      "step": 2197
+    },
+    {
+      "epoch": 0.4228549442093113,
+      "grad_norm": 1.328840046638587,
+      "learning_rate": 1.2940462806258696e-05,
+      "loss": 0.8843,
+      "step": 2198
+    },
+    {
+      "epoch": 0.42304732589457483,
+      "grad_norm": 1.2206755409346508,
+      "learning_rate": 1.2934506847991976e-05,
+      "loss": 0.8621,
+      "step": 2199
+    },
+    {
+      "epoch": 0.4232397075798384,
+      "grad_norm": 1.490245352916321,
+      "learning_rate": 1.2928549750448661e-05,
+      "loss": 0.8178,
+      "step": 2200
+    },
+    {
+      "epoch": 0.42343208926510195,
+      "grad_norm": 1.386315271404229,
+      "learning_rate": 1.2922591515941498e-05,
+      "loss": 0.8758,
+      "step": 2201
+    },
+    {
+      "epoch": 0.42362447095036554,
+      "grad_norm": 1.4411330725144234,
+      "learning_rate": 1.2916632146783683e-05,
+      "loss": 0.8792,
+      "step": 2202
+    },
+    {
+      "epoch": 0.42381685263562907,
+      "grad_norm": 4.339847219877638,
+      "learning_rate": 1.2910671645288841e-05,
+      "loss": 0.9239,
+      "step": 2203
+    },
+    {
+      "epoch": 0.42400923432089266,
+      "grad_norm": 3.5103682529594495,
+      "learning_rate": 1.2904710013771054e-05,
+      "loss": 0.826,
+      "step": 2204
+    },
+    {
+      "epoch": 0.4242016160061562,
+      "grad_norm": 3.4233191663764226,
+      "learning_rate": 1.2898747254544826e-05,
+      "loss": 0.9501,
+      "step": 2205
+    },
+    {
+      "epoch": 0.4243939976914198,
+      "grad_norm": 1.6051053677727434,
+      "learning_rate": 1.2892783369925105e-05,
+      "loss": 0.8672,
+      "step": 2206
+    },
+    {
+      "epoch": 0.42458637937668336,
+      "grad_norm": 2.1554042685089985,
+      "learning_rate": 1.2886818362227283e-05,
+      "loss": 0.8661,
+      "step": 2207
+    },
+    {
+      "epoch": 0.4247787610619469,
+      "grad_norm": 1.6882461437762153,
+      "learning_rate": 1.2880852233767174e-05,
+      "loss": 0.9075,
+      "step": 2208
+    },
+    {
+      "epoch": 0.4249711427472105,
+      "grad_norm": 1.2525291192564392,
+      "learning_rate": 1.2874884986861038e-05,
+      "loss": 0.8953,
+      "step": 2209
+    },
+    {
+      "epoch": 0.425163524432474,
+      "grad_norm": 1.6481621783646918,
+      "learning_rate": 1.2868916623825561e-05,
+      "loss": 0.9161,
+      "step": 2210
+    },
+    {
+      "epoch": 0.4253559061177376,
+      "grad_norm": 2.075638478499381,
+      "learning_rate": 1.2862947146977876e-05,
+      "loss": 0.8731,
+      "step": 2211
+    },
+    {
+      "epoch": 0.42554828780300114,
+      "grad_norm": 1.5289295904440732,
+      "learning_rate": 1.2856976558635532e-05,
+      "loss": 0.9045,
+      "step": 2212
+    },
+    {
+      "epoch": 0.4257406694882647,
+      "grad_norm": 1.3084040038008342,
+      "learning_rate": 1.2851004861116519e-05,
+      "loss": 0.9022,
+      "step": 2213
+    },
+    {
+      "epoch": 0.42593305117352825,
+      "grad_norm": 1.421522617904019,
+      "learning_rate": 1.2845032056739257e-05,
+      "loss": 0.9001,
+      "step": 2214
+    },
+    {
+      "epoch": 0.42612543285879184,
+      "grad_norm": 1.328292527940839,
+      "learning_rate": 1.2839058147822595e-05,
+      "loss": 0.8911,
+      "step": 2215
+    },
+    {
+      "epoch": 0.42631781454405543,
+      "grad_norm": 1.5977532562876025,
+      "learning_rate": 1.2833083136685803e-05,
+      "loss": 0.8758,
+      "step": 2216
+    },
+    {
+      "epoch": 0.42651019622931896,
+      "grad_norm": 1.123899875475584,
+      "learning_rate": 1.2827107025648595e-05,
+      "loss": 0.8681,
+      "step": 2217
+    },
+    {
+      "epoch": 0.42670257791458255,
+      "grad_norm": 1.5404908213745174,
+      "learning_rate": 1.2821129817031099e-05,
+      "loss": 0.8945,
+      "step": 2218
+    },
+    {
+      "epoch": 0.4268949595998461,
+      "grad_norm": 1.4793051488379587,
+      "learning_rate": 1.2815151513153874e-05,
+      "loss": 0.9163,
+      "step": 2219
+    },
+    {
+      "epoch": 0.42708734128510967,
+      "grad_norm": 1.3039514540302393,
+      "learning_rate": 1.2809172116337903e-05,
+      "loss": 0.9197,
+      "step": 2220
+    },
+    {
+      "epoch": 0.4272797229703732,
+      "grad_norm": 1.3672635158062265,
+      "learning_rate": 1.2803191628904594e-05,
+      "loss": 0.9056,
+      "step": 2221
+    },
+    {
+      "epoch": 0.4274721046556368,
+      "grad_norm": 1.2974903261077524,
+      "learning_rate": 1.2797210053175779e-05,
+      "loss": 0.8489,
+      "step": 2222
+    },
+    {
+      "epoch": 0.4276644863409004,
+      "grad_norm": 1.3125846053144272,
+      "learning_rate": 1.2791227391473706e-05,
+      "loss": 0.8805,
+      "step": 2223
+    },
+    {
+      "epoch": 0.4278568680261639,
+      "grad_norm": 1.5765260451183827,
+      "learning_rate": 1.2785243646121059e-05,
+      "loss": 0.8629,
+      "step": 2224
+    },
+    {
+      "epoch": 0.4280492497114275,
+      "grad_norm": 2.6081369216639207,
+      "learning_rate": 1.277925881944093e-05,
+      "loss": 0.9158,
+      "step": 2225
+    },
+    {
+      "epoch": 0.428241631396691,
+      "grad_norm": 1.2859194144498436,
+      "learning_rate": 1.2773272913756833e-05,
+      "loss": 0.9167,
+      "step": 2226
+    },
+    {
+      "epoch": 0.4284340130819546,
+      "grad_norm": 1.496719038583247,
+      "learning_rate": 1.2767285931392705e-05,
+      "loss": 0.9447,
+      "step": 2227
+    },
+    {
+      "epoch": 0.42862639476721814,
+      "grad_norm": 1.5051680153675357,
+      "learning_rate": 1.27612978746729e-05,
+      "loss": 0.8961,
+      "step": 2228
+    },
+    {
+      "epoch": 0.42881877645248173,
+      "grad_norm": 1.3027489627889863,
+      "learning_rate": 1.2755308745922182e-05,
+      "loss": 0.9378,
+      "step": 2229
+    },
+    {
+      "epoch": 0.42901115813774526,
+      "grad_norm": 1.3901632700871538,
+      "learning_rate": 1.2749318547465742e-05,
+      "loss": 0.8834,
+      "step": 2230
+    },
+    {
+      "epoch": 0.42920353982300885,
+      "grad_norm": 1.765635333446556,
+      "learning_rate": 1.2743327281629181e-05,
+      "loss": 0.8402,
+      "step": 2231
+    },
+    {
+      "epoch": 0.42939592150827244,
+      "grad_norm": 1.4467960632536694,
+      "learning_rate": 1.2737334950738512e-05,
+      "loss": 0.8794,
+      "step": 2232
+    },
+    {
+      "epoch": 0.42958830319353597,
+      "grad_norm": 0.6295823555122928,
+      "learning_rate": 1.273134155712017e-05,
+      "loss": 0.346,
+      "step": 2233
+    },
+    {
+      "epoch": 0.42978068487879956,
+      "grad_norm": 0.6547378133372389,
+      "learning_rate": 1.272534710310099e-05,
+      "loss": 0.3714,
+      "step": 2234
+    },
+    {
+      "epoch": 0.4299730665640631,
+      "grad_norm": 1.5721053899650959,
+      "learning_rate": 1.2719351591008228e-05,
+      "loss": 0.9027,
+      "step": 2235
+    },
+    {
+      "epoch": 0.4301654482493267,
+      "grad_norm": 1.6573723773734346,
+      "learning_rate": 1.2713355023169547e-05,
+      "loss": 0.8024,
+      "step": 2236
+    },
+    {
+      "epoch": 0.4303578299345902,
+      "grad_norm": 2.47122010967055,
+      "learning_rate": 1.2707357401913022e-05,
+      "loss": 0.9075,
+      "step": 2237
+    },
+    {
+      "epoch": 0.4305502116198538,
+      "grad_norm": 1.257109475798645,
+      "learning_rate": 1.270135872956714e-05,
+      "loss": 0.8669,
+      "step": 2238
+    },
+    {
+      "epoch": 0.4307425933051173,
+      "grad_norm": 1.339606700913974,
+      "learning_rate": 1.2695359008460785e-05,
+      "loss": 0.8792,
+      "step": 2239
+    },
+    {
+      "epoch": 0.4309349749903809,
+      "grad_norm": 1.4493702609988548,
+      "learning_rate": 1.2689358240923264e-05,
+      "loss": 0.8408,
+      "step": 2240
+    },
+    {
+      "epoch": 0.4311273566756445,
+      "grad_norm": 1.5178488775882952,
+      "learning_rate": 1.2683356429284273e-05,
+      "loss": 0.8794,
+      "step": 2241
+    },
+    {
+      "epoch": 0.43131973836090803,
+      "grad_norm": 1.3680995184881815,
+      "learning_rate": 1.2677353575873926e-05,
+      "loss": 0.8674,
+      "step": 2242
+    },
+    {
+      "epoch": 0.4315121200461716,
+      "grad_norm": 1.4056910055142902,
+      "learning_rate": 1.2671349683022736e-05,
+      "loss": 0.8897,
+      "step": 2243
+    },
+    {
+      "epoch": 0.43170450173143515,
+      "grad_norm": 0.6973943100968918,
+      "learning_rate": 1.2665344753061622e-05,
+      "loss": 0.3669,
+      "step": 2244
+    },
+    {
+      "epoch": 0.43189688341669874,
+      "grad_norm": 0.6816466474230318,
+      "learning_rate": 1.2659338788321904e-05,
+      "loss": 0.3694,
+      "step": 2245
+    },
+    {
+      "epoch": 0.43208926510196227,
+      "grad_norm": 1.3246180763914892,
+      "learning_rate": 1.2653331791135308e-05,
+      "loss": 0.8754,
+      "step": 2246
+    },
+    {
+      "epoch": 0.43228164678722586,
+      "grad_norm": 1.8970380705563845,
+      "learning_rate": 1.2647323763833952e-05,
+      "loss": 0.9335,
+      "step": 2247
+    },
+    {
+      "epoch": 0.4324740284724894,
+      "grad_norm": 1.6796681432440084,
+      "learning_rate": 1.264131470875036e-05,
+      "loss": 0.8501,
+      "step": 2248
+    },
+    {
+      "epoch": 0.432666410157753,
+      "grad_norm": 1.2692536467802347,
+      "learning_rate": 1.2635304628217452e-05,
+      "loss": 0.9247,
+      "step": 2249
+    },
+    {
+      "epoch": 0.43285879184301657,
+      "grad_norm": 3.3792251789375154,
+      "learning_rate": 1.2629293524568555e-05,
+      "loss": 0.9437,
+      "step": 2250
+    },
+    {
+      "epoch": 0.4330511735282801,
+      "grad_norm": 0.7040679851593047,
+      "learning_rate": 1.2623281400137383e-05,
+      "loss": 0.3494,
+      "step": 2251
+    },
+    {
+      "epoch": 0.4332435552135437,
+      "grad_norm": 1.372620250941867,
+      "learning_rate": 1.2617268257258051e-05,
+      "loss": 0.8275,
+      "step": 2252
+    },
+    {
+      "epoch": 0.4334359368988072,
+      "grad_norm": 1.5633913685077307,
+      "learning_rate": 1.2611254098265063e-05,
+      "loss": 0.9406,
+      "step": 2253
+    },
+    {
+      "epoch": 0.4336283185840708,
+      "grad_norm": 1.3721947079178178,
+      "learning_rate": 1.2605238925493326e-05,
+      "loss": 0.8971,
+      "step": 2254
+    },
+    {
+      "epoch": 0.43382070026933434,
+      "grad_norm": 1.411694994202765,
+      "learning_rate": 1.2599222741278136e-05,
+      "loss": 0.9182,
+      "step": 2255
+    },
+    {
+      "epoch": 0.4340130819545979,
+      "grad_norm": 2.0348615928272134,
+      "learning_rate": 1.2593205547955185e-05,
+      "loss": 0.8367,
+      "step": 2256
+    },
+    {
+      "epoch": 0.4342054636398615,
+      "grad_norm": 1.2573899406472373,
+      "learning_rate": 1.2587187347860554e-05,
+      "loss": 0.9156,
+      "step": 2257
+    },
+    {
+      "epoch": 0.43439784532512504,
+      "grad_norm": 1.2301703834703603,
+      "learning_rate": 1.2581168143330716e-05,
+      "loss": 0.8221,
+      "step": 2258
+    },
+    {
+      "epoch": 0.43459022701038863,
+      "grad_norm": 1.3161598027492327,
+      "learning_rate": 1.2575147936702531e-05,
+      "loss": 0.8707,
+      "step": 2259
+    },
+    {
+      "epoch": 0.43478260869565216,
+      "grad_norm": 1.5312794857940204,
+      "learning_rate": 1.2569126730313255e-05,
+      "loss": 0.8809,
+      "step": 2260
+    },
+    {
+      "epoch": 0.43497499038091575,
+      "grad_norm": 2.8780518902409424,
+      "learning_rate": 1.2563104526500523e-05,
+      "loss": 0.9366,
+      "step": 2261
+    },
+    {
+      "epoch": 0.4351673720661793,
+      "grad_norm": 0.6469202837580672,
+      "learning_rate": 1.2557081327602361e-05,
+      "loss": 0.3636,
+      "step": 2262
+    },
+    {
+      "epoch": 0.43535975375144287,
+      "grad_norm": 1.4561955417187027,
+      "learning_rate": 1.2551057135957187e-05,
+      "loss": 0.8605,
+      "step": 2263
+    },
+    {
+      "epoch": 0.4355521354367064,
+      "grad_norm": 1.6225340812718536,
+      "learning_rate": 1.2545031953903796e-05,
+      "loss": 0.8869,
+      "step": 2264
+    },
+    {
+      "epoch": 0.43574451712197,
+      "grad_norm": 1.1819273916941369,
+      "learning_rate": 1.2539005783781374e-05,
+      "loss": 0.8424,
+      "step": 2265
+    },
+    {
+      "epoch": 0.4359368988072336,
+      "grad_norm": 1.0964904868233791,
+      "learning_rate": 1.2532978627929486e-05,
+      "loss": 0.9052,
+      "step": 2266
+    },
+    {
+      "epoch": 0.4361292804924971,
+      "grad_norm": 1.232344886907517,
+      "learning_rate": 1.2526950488688083e-05,
+      "loss": 0.9144,
+      "step": 2267
+    },
+    {
+      "epoch": 0.4363216621777607,
+      "grad_norm": 1.5296915595183158,
+      "learning_rate": 1.2520921368397492e-05,
+      "loss": 0.9,
+      "step": 2268
+    },
+    {
+      "epoch": 0.4365140438630242,
+      "grad_norm": 1.388204710785612,
+      "learning_rate": 1.2514891269398429e-05,
+      "loss": 0.8784,
+      "step": 2269
+    },
+    {
+      "epoch": 0.4367064255482878,
+      "grad_norm": 1.3900343105598352,
+      "learning_rate": 1.2508860194031986e-05,
+      "loss": 0.8726,
+      "step": 2270
+    },
+    {
+      "epoch": 0.43689880723355135,
+      "grad_norm": 0.6535934715528349,
+      "learning_rate": 1.2502828144639629e-05,
+      "loss": 0.354,
+      "step": 2271
+    },
+    {
+      "epoch": 0.43709118891881493,
+      "grad_norm": 1.1662983321656906,
+      "learning_rate": 1.2496795123563218e-05,
+      "loss": 0.8669,
+      "step": 2272
+    },
+    {
+      "epoch": 0.43728357060407846,
+      "grad_norm": 1.5300098427845699,
+      "learning_rate": 1.249076113314497e-05,
+      "loss": 0.9214,
+      "step": 2273
+    },
+    {
+      "epoch": 0.43747595228934205,
+      "grad_norm": 1.552549854280064,
+      "learning_rate": 1.248472617572749e-05,
+      "loss": 0.7865,
+      "step": 2274
+    },
+    {
+      "epoch": 0.43766833397460564,
+      "grad_norm": 1.2803926648811377,
+      "learning_rate": 1.2478690253653756e-05,
+      "loss": 0.9105,
+      "step": 2275
+    },
+    {
+      "epoch": 0.43786071565986917,
+      "grad_norm": 1.8816770208996096,
+      "learning_rate": 1.2472653369267122e-05,
+      "loss": 0.9002,
+      "step": 2276
+    },
+    {
+      "epoch": 0.43805309734513276,
+      "grad_norm": 1.599301129521268,
+      "learning_rate": 1.2466615524911316e-05,
+      "loss": 0.7968,
+      "step": 2277
+    },
+    {
+      "epoch": 0.4382454790303963,
+      "grad_norm": 1.343294991692372,
+      "learning_rate": 1.2460576722930432e-05,
+      "loss": 0.8787,
+      "step": 2278
+    },
+    {
+      "epoch": 0.4384378607156599,
+      "grad_norm": 1.4047980495774868,
+      "learning_rate": 1.2454536965668949e-05,
+      "loss": 0.9102,
+      "step": 2279
+    },
+    {
+      "epoch": 0.4386302424009234,
+      "grad_norm": 1.2913116002498786,
+      "learning_rate": 1.24484962554717e-05,
+      "loss": 0.9518,
+      "step": 2280
+    },
+    {
+      "epoch": 0.438822624086187,
+      "grad_norm": 1.8515026027011596,
+      "learning_rate": 1.24424545946839e-05,
+      "loss": 0.8405,
+      "step": 2281
+    },
+    {
+      "epoch": 0.4390150057714506,
+      "grad_norm": 1.1874984227524985,
+      "learning_rate": 1.2436411985651131e-05,
+      "loss": 0.9164,
+      "step": 2282
+    },
+    {
+      "epoch": 0.4392073874567141,
+      "grad_norm": 1.261528741288549,
+      "learning_rate": 1.2430368430719342e-05,
+      "loss": 0.8366,
+      "step": 2283
+    },
+    {
+      "epoch": 0.4393997691419777,
+      "grad_norm": 1.4008388390877224,
+      "learning_rate": 1.242432393223485e-05,
+      "loss": 0.9155,
+      "step": 2284
+    },
+    {
+      "epoch": 0.43959215082724123,
+      "grad_norm": 1.6119634996852874,
+      "learning_rate": 1.2418278492544328e-05,
+      "loss": 0.878,
+      "step": 2285
+    },
+    {
+      "epoch": 0.4397845325125048,
+      "grad_norm": 1.482535377596559,
+      "learning_rate": 1.2412232113994841e-05,
+      "loss": 0.8923,
+      "step": 2286
+    },
+    {
+      "epoch": 0.43997691419776835,
+      "grad_norm": 1.988429233896276,
+      "learning_rate": 1.2406184798933786e-05,
+      "loss": 0.8875,
+      "step": 2287
+    },
+    {
+      "epoch": 0.44016929588303194,
+      "grad_norm": 1.1897553969813486,
+      "learning_rate": 1.2400136549708945e-05,
+      "loss": 0.8707,
+      "step": 2288
+    },
+    {
+      "epoch": 0.4403616775682955,
+      "grad_norm": 1.0915104961587956,
+      "learning_rate": 1.239408736866846e-05,
+      "loss": 0.8841,
+      "step": 2289
+    },
+    {
+      "epoch": 0.44055405925355906,
+      "grad_norm": 1.3097717561307791,
+      "learning_rate": 1.2388037258160823e-05,
+      "loss": 0.8515,
+      "step": 2290
+    },
+    {
+      "epoch": 0.44074644093882265,
+      "grad_norm": 1.1880205566702666,
+      "learning_rate": 1.23819862205349e-05,
+      "loss": 0.8563,
+      "step": 2291
+    },
+    {
+      "epoch": 0.4409388226240862,
+      "grad_norm": 1.4196987942395551,
+      "learning_rate": 1.2375934258139917e-05,
+      "loss": 0.887,
+      "step": 2292
+    },
+    {
+      "epoch": 0.44113120430934977,
+      "grad_norm": 1.1559550939567227,
+      "learning_rate": 1.2369881373325448e-05,
+      "loss": 0.9007,
+      "step": 2293
+    },
+    {
+      "epoch": 0.4413235859946133,
+      "grad_norm": 1.3218128625625019,
+      "learning_rate": 1.236382756844143e-05,
+      "loss": 0.9395,
+      "step": 2294
+    },
+    {
+      "epoch": 0.4415159676798769,
+      "grad_norm": 0.6771022515220492,
+      "learning_rate": 1.2357772845838159e-05,
+      "loss": 0.3539,
+      "step": 2295
+    },
+    {
+      "epoch": 0.4417083493651404,
+      "grad_norm": 1.1738783281545502,
+      "learning_rate": 1.2351717207866292e-05,
+      "loss": 0.8496,
+      "step": 2296
+    },
+    {
+      "epoch": 0.441900731050404,
+      "grad_norm": 1.2982548641514113,
+      "learning_rate": 1.2345660656876832e-05,
+      "loss": 0.8668,
+      "step": 2297
+    },
+    {
+      "epoch": 0.44209311273566754,
+      "grad_norm": 1.2234809705646443,
+      "learning_rate": 1.233960319522114e-05,
+      "loss": 0.9079,
+      "step": 2298
+    },
+    {
+      "epoch": 0.4422854944209311,
+      "grad_norm": 1.423960822997122,
+      "learning_rate": 1.2333544825250938e-05,
+      "loss": 0.9126,
+      "step": 2299
+    },
+    {
+      "epoch": 0.4424778761061947,
+      "grad_norm": 2.260872978518875,
+      "learning_rate": 1.2327485549318285e-05,
+      "loss": 0.8605,
+      "step": 2300
+    },
+    {
+      "epoch": 0.44267025779145824,
+      "grad_norm": 1.220186524853864,
+      "learning_rate": 1.2321425369775601e-05,
+      "loss": 0.8754,
+      "step": 2301
+    },
+    {
+      "epoch": 0.44286263947672183,
+      "grad_norm": 1.2971255286939065,
+      "learning_rate": 1.2315364288975665e-05,
+      "loss": 0.8582,
+      "step": 2302
+    },
+    {
+      "epoch": 0.44305502116198536,
+      "grad_norm": 1.2258509201073762,
+      "learning_rate": 1.2309302309271587e-05,
+      "loss": 0.8551,
+      "step": 2303
+    },
+    {
+      "epoch": 0.44324740284724895,
+      "grad_norm": 1.3814907315770204,
+      "learning_rate": 1.2303239433016842e-05,
+      "loss": 0.89,
+      "step": 2304
+    },
+    {
+      "epoch": 0.4434397845325125,
+      "grad_norm": 1.573059319701258,
+      "learning_rate": 1.2297175662565248e-05,
+      "loss": 0.8801,
+      "step": 2305
+    },
+    {
+      "epoch": 0.44363216621777607,
+      "grad_norm": 0.6271844125556482,
+      "learning_rate": 1.229111100027097e-05,
+      "loss": 0.3406,
+      "step": 2306
+    },
+    {
+      "epoch": 0.44382454790303966,
+      "grad_norm": 1.6888903792062326,
+      "learning_rate": 1.228504544848851e-05,
+      "loss": 0.9383,
+      "step": 2307
+    },
+    {
+      "epoch": 0.4440169295883032,
+      "grad_norm": 0.618608845411503,
+      "learning_rate": 1.2278979009572736e-05,
+      "loss": 0.3497,
+      "step": 2308
+    },
+    {
+      "epoch": 0.4442093112735668,
+      "grad_norm": 1.3515016169342413,
+      "learning_rate": 1.2272911685878841e-05,
+      "loss": 0.8193,
+      "step": 2309
+    },
+    {
+      "epoch": 0.4444016929588303,
+      "grad_norm": 1.3088485434149082,
+      "learning_rate": 1.2266843479762372e-05,
+      "loss": 0.8765,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4445940746440939,
+      "grad_norm": 1.4619697564468779,
+      "learning_rate": 1.2260774393579209e-05,
+      "loss": 0.8661,
+      "step": 2311
+    },
+    {
+      "epoch": 0.4447864563293574,
+      "grad_norm": 1.1578390096684823,
+      "learning_rate": 1.2254704429685593e-05,
+      "loss": 0.885,
+      "step": 2312
+    },
+    {
+      "epoch": 0.444978838014621,
+      "grad_norm": 1.2762421241972906,
+      "learning_rate": 1.2248633590438084e-05,
+      "loss": 0.9179,
+      "step": 2313
+    },
+    {
+      "epoch": 0.44517121969988455,
+      "grad_norm": 1.0679407210381284,
+      "learning_rate": 1.2242561878193589e-05,
+      "loss": 0.8352,
+      "step": 2314
+    },
+    {
+      "epoch": 0.44536360138514813,
+      "grad_norm": 1.5793163017432665,
+      "learning_rate": 1.2236489295309362e-05,
+      "loss": 0.9107,
+      "step": 2315
+    },
+    {
+      "epoch": 0.4455559830704117,
+      "grad_norm": 0.6567504889110719,
+      "learning_rate": 1.2230415844142984e-05,
+      "loss": 0.3158,
+      "step": 2316
+    },
+    {
+      "epoch": 0.44574836475567525,
+      "grad_norm": 1.2509541766197352,
+      "learning_rate": 1.2224341527052378e-05,
+      "loss": 0.8897,
+      "step": 2317
+    },
+    {
+      "epoch": 0.44594074644093884,
+      "grad_norm": 1.186637171019848,
+      "learning_rate": 1.2218266346395811e-05,
+      "loss": 0.8647,
+      "step": 2318
+    },
+    {
+      "epoch": 0.44613312812620237,
+      "grad_norm": 1.3110992531229815,
+      "learning_rate": 1.221219030453187e-05,
+      "loss": 0.9068,
+      "step": 2319
+    },
+    {
+      "epoch": 0.44632550981146596,
+      "grad_norm": 1.1193107344686684,
+      "learning_rate": 1.220611340381948e-05,
+      "loss": 0.8917,
+      "step": 2320
+    },
+    {
+      "epoch": 0.4465178914967295,
+      "grad_norm": 1.073916036524892,
+      "learning_rate": 1.2200035646617912e-05,
+      "loss": 0.8625,
+      "step": 2321
+    },
+    {
+      "epoch": 0.4467102731819931,
+      "grad_norm": 1.3872581177847563,
+      "learning_rate": 1.2193957035286757e-05,
+      "loss": 0.8543,
+      "step": 2322
+    },
+    {
+      "epoch": 0.4469026548672566,
+      "grad_norm": 1.4348202431786756,
+      "learning_rate": 1.2187877572185937e-05,
+      "loss": 0.8939,
+      "step": 2323
+    },
+    {
+      "epoch": 0.4470950365525202,
+      "grad_norm": 1.3916512831909822,
+      "learning_rate": 1.2181797259675713e-05,
+      "loss": 0.9325,
+      "step": 2324
+    },
+    {
+      "epoch": 0.4472874182377838,
+      "grad_norm": 1.2988966134941748,
+      "learning_rate": 1.2175716100116677e-05,
+      "loss": 0.9179,
+      "step": 2325
+    },
+    {
+      "epoch": 0.4474797999230473,
+      "grad_norm": 1.1882563680887055,
+      "learning_rate": 1.2169634095869736e-05,
+      "loss": 0.8465,
+      "step": 2326
+    },
+    {
+      "epoch": 0.4476721816083109,
+      "grad_norm": 1.4671583052761452,
+      "learning_rate": 1.2163551249296132e-05,
+      "loss": 0.9373,
+      "step": 2327
+    },
+    {
+      "epoch": 0.44786456329357444,
+      "grad_norm": 2.086551686072085,
+      "learning_rate": 1.2157467562757443e-05,
+      "loss": 0.8683,
+      "step": 2328
+    },
+    {
+      "epoch": 0.448056944978838,
+      "grad_norm": 1.2179741633555015,
+      "learning_rate": 1.2151383038615563e-05,
+      "loss": 0.8474,
+      "step": 2329
+    },
+    {
+      "epoch": 0.44824932666410156,
+      "grad_norm": 1.5000313092348114,
+      "learning_rate": 1.214529767923271e-05,
+      "loss": 0.8591,
+      "step": 2330
+    },
+    {
+      "epoch": 0.44844170834936514,
+      "grad_norm": 1.4693659119857778,
+      "learning_rate": 1.2139211486971436e-05,
+      "loss": 0.8769,
+      "step": 2331
+    },
+    {
+      "epoch": 0.44863409003462873,
+      "grad_norm": 1.1318574351622277,
+      "learning_rate": 1.213312446419461e-05,
+      "loss": 0.8788,
+      "step": 2332
+    },
+    {
+      "epoch": 0.44882647171989226,
+      "grad_norm": 1.2437713953697036,
+      "learning_rate": 1.2127036613265418e-05,
+      "loss": 0.9061,
+      "step": 2333
+    },
+    {
+      "epoch": 0.44901885340515585,
+      "grad_norm": 1.2633263260309098,
+      "learning_rate": 1.2120947936547375e-05,
+      "loss": 0.8509,
+      "step": 2334
+    },
+    {
+      "epoch": 0.4492112350904194,
+      "grad_norm": 0.7297037225429532,
+      "learning_rate": 1.2114858436404322e-05,
+      "loss": 0.3169,
+      "step": 2335
+    },
+    {
+      "epoch": 0.44940361677568297,
+      "grad_norm": 0.7023376159004668,
+      "learning_rate": 1.2108768115200405e-05,
+      "loss": 0.3659,
+      "step": 2336
+    },
+    {
+      "epoch": 0.4495959984609465,
+      "grad_norm": 1.475696081660852,
+      "learning_rate": 1.2102676975300095e-05,
+      "loss": 0.917,
+      "step": 2337
+    },
+    {
+      "epoch": 0.4497883801462101,
+      "grad_norm": 1.208575127194329,
+      "learning_rate": 1.209658501906819e-05,
+      "loss": 0.9228,
+      "step": 2338
+    },
+    {
+      "epoch": 0.4499807618314736,
+      "grad_norm": 1.509655180108531,
+      "learning_rate": 1.2090492248869795e-05,
+      "loss": 0.8336,
+      "step": 2339
+    },
+    {
+      "epoch": 0.4501731435167372,
+      "grad_norm": 1.6300190827056225,
+      "learning_rate": 1.2084398667070325e-05,
+      "loss": 0.9249,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4503655252020008,
+      "grad_norm": 1.1458670981939598,
+      "learning_rate": 1.2078304276035527e-05,
+      "loss": 0.8597,
+      "step": 2341
+    },
+    {
+      "epoch": 0.4505579068872643,
+      "grad_norm": 1.2389371534641915,
+      "learning_rate": 1.2072209078131451e-05,
+      "loss": 0.8649,
+      "step": 2342
+    },
+    {
+      "epoch": 0.4507502885725279,
+      "grad_norm": 1.3470177209990521,
+      "learning_rate": 1.2066113075724461e-05,
+      "loss": 0.8833,
+      "step": 2343
+    },
+    {
+      "epoch": 0.45094267025779144,
+      "grad_norm": 1.3521562612893554,
+      "learning_rate": 1.206001627118124e-05,
+      "loss": 0.9044,
+      "step": 2344
+    },
+    {
+      "epoch": 0.45113505194305503,
+      "grad_norm": 1.147228804668784,
+      "learning_rate": 1.2053918666868776e-05,
+      "loss": 0.8296,
+      "step": 2345
+    },
+    {
+      "epoch": 0.45132743362831856,
+      "grad_norm": 1.1311873080333932,
+      "learning_rate": 1.2047820265154362e-05,
+      "loss": 0.8341,
+      "step": 2346
+    },
+    {
+      "epoch": 0.45151981531358215,
+      "grad_norm": 1.3234876040099923,
+      "learning_rate": 1.2041721068405614e-05,
+      "loss": 0.8999,
+      "step": 2347
+    },
+    {
+      "epoch": 0.4517121969988457,
+      "grad_norm": 1.2015331675592837,
+      "learning_rate": 1.203562107899045e-05,
+      "loss": 0.8413,
+      "step": 2348
+    },
+    {
+      "epoch": 0.45190457868410927,
+      "grad_norm": 1.356504285409257,
+      "learning_rate": 1.2029520299277095e-05,
+      "loss": 0.8607,
+      "step": 2349
+    },
+    {
+      "epoch": 0.45209696036937286,
+      "grad_norm": 1.3886940547481808,
+      "learning_rate": 1.2023418731634078e-05,
+      "loss": 0.9326,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4522893420546364,
+      "grad_norm": 1.7142259041862473,
+      "learning_rate": 1.2017316378430244e-05,
+      "loss": 0.8667,
+      "step": 2351
+    },
+    {
+      "epoch": 0.4524817237399,
+      "grad_norm": 1.5671540941240538,
+      "learning_rate": 1.2011213242034733e-05,
+      "loss": 0.885,
+      "step": 2352
+    },
+    {
+      "epoch": 0.4526741054251635,
+      "grad_norm": 1.3725761762008346,
+      "learning_rate": 1.2005109324816992e-05,
+      "loss": 0.846,
+      "step": 2353
+    },
+    {
+      "epoch": 0.4528664871104271,
+      "grad_norm": 1.9171908683225825,
+      "learning_rate": 1.1999004629146775e-05,
+      "loss": 0.9725,
+      "step": 2354
+    },
+    {
+      "epoch": 0.45305886879569063,
+      "grad_norm": 1.4808620125307945,
+      "learning_rate": 1.1992899157394133e-05,
+      "loss": 0.9057,
+      "step": 2355
+    },
+    {
+      "epoch": 0.4532512504809542,
+      "grad_norm": 1.2813221632661242,
+      "learning_rate": 1.1986792911929418e-05,
+      "loss": 0.8468,
+      "step": 2356
+    },
+    {
+      "epoch": 0.4534436321662178,
+      "grad_norm": 1.1875519159866192,
+      "learning_rate": 1.198068589512329e-05,
+      "loss": 0.8622,
+      "step": 2357
+    },
+    {
+      "epoch": 0.45363601385148133,
+      "grad_norm": 1.4537671616720274,
+      "learning_rate": 1.1974578109346702e-05,
+      "loss": 0.8822,
+      "step": 2358
+    },
+    {
+      "epoch": 0.4538283955367449,
+      "grad_norm": 0.9151771036607663,
+      "learning_rate": 1.1968469556970905e-05,
+      "loss": 0.3228,
+      "step": 2359
+    },
+    {
+      "epoch": 0.45402077722200845,
+      "grad_norm": 1.6520799003268476,
+      "learning_rate": 1.1962360240367445e-05,
+      "loss": 0.9581,
+      "step": 2360
+    },
+    {
+      "epoch": 0.45421315890727204,
+      "grad_norm": 1.2094077750053864,
+      "learning_rate": 1.1956250161908179e-05,
+      "loss": 0.8925,
+      "step": 2361
+    },
+    {
+      "epoch": 0.4544055405925356,
+      "grad_norm": 2.0157029742699906,
+      "learning_rate": 1.195013932396524e-05,
+      "loss": 0.8164,
+      "step": 2362
+    },
+    {
+      "epoch": 0.45459792227779916,
+      "grad_norm": 1.57018532050978,
+      "learning_rate": 1.1944027728911072e-05,
+      "loss": 0.9015,
+      "step": 2363
+    },
+    {
+      "epoch": 0.4547903039630627,
+      "grad_norm": 1.9477506970167473,
+      "learning_rate": 1.1937915379118406e-05,
+      "loss": 0.8835,
+      "step": 2364
+    },
+    {
+      "epoch": 0.4549826856483263,
+      "grad_norm": 1.2390946942454122,
+      "learning_rate": 1.1931802276960265e-05,
+      "loss": 0.948,
+      "step": 2365
+    },
+    {
+      "epoch": 0.45517506733358987,
+      "grad_norm": 1.386069134922011,
+      "learning_rate": 1.1925688424809965e-05,
+      "loss": 0.9202,
+      "step": 2366
+    },
+    {
+      "epoch": 0.4553674490188534,
+      "grad_norm": 1.144974559471121,
+      "learning_rate": 1.1919573825041115e-05,
+      "loss": 0.9004,
+      "step": 2367
+    },
+    {
+      "epoch": 0.455559830704117,
+      "grad_norm": 1.1948164174030385,
+      "learning_rate": 1.1913458480027614e-05,
+      "loss": 0.8793,
+      "step": 2368
+    },
+    {
+      "epoch": 0.4557522123893805,
+      "grad_norm": 1.1622348574444552,
+      "learning_rate": 1.1907342392143646e-05,
+      "loss": 0.924,
+      "step": 2369
+    },
+    {
+      "epoch": 0.4559445940746441,
+      "grad_norm": 1.5031571219304936,
+      "learning_rate": 1.1901225563763694e-05,
+      "loss": 0.885,
+      "step": 2370
+    },
+    {
+      "epoch": 0.45613697575990764,
+      "grad_norm": 2.164891101175655,
+      "learning_rate": 1.1895107997262516e-05,
+      "loss": 0.8318,
+      "step": 2371
+    },
+    {
+      "epoch": 0.4563293574451712,
+      "grad_norm": 1.190745898027612,
+      "learning_rate": 1.1888989695015166e-05,
+      "loss": 0.822,
+      "step": 2372
+    },
+    {
+      "epoch": 0.45652173913043476,
+      "grad_norm": 1.404395485550701,
+      "learning_rate": 1.1882870659396968e-05,
+      "loss": 0.8778,
+      "step": 2373
+    },
+    {
+      "epoch": 0.45671412081569834,
+      "grad_norm": 1.2168393338377879,
+      "learning_rate": 1.1876750892783558e-05,
+      "loss": 0.8821,
+      "step": 2374
+    },
+    {
+      "epoch": 0.45690650250096193,
+      "grad_norm": 1.2264818057895666,
+      "learning_rate": 1.1870630397550831e-05,
+      "loss": 0.9351,
+      "step": 2375
+    },
+    {
+      "epoch": 0.45709888418622546,
+      "grad_norm": 1.3327459572893146,
+      "learning_rate": 1.1864509176074974e-05,
+      "loss": 0.9035,
+      "step": 2376
+    },
+    {
+      "epoch": 0.45729126587148905,
+      "grad_norm": 1.7234941673702175,
+      "learning_rate": 1.185838723073246e-05,
+      "loss": 0.9067,
+      "step": 2377
+    },
+    {
+      "epoch": 0.4574836475567526,
+      "grad_norm": 1.18626934767387,
+      "learning_rate": 1.1852264563900038e-05,
+      "loss": 0.91,
+      "step": 2378
+    },
+    {
+      "epoch": 0.45767602924201617,
+      "grad_norm": 3.773335192310924,
+      "learning_rate": 1.1846141177954733e-05,
+      "loss": 0.9028,
+      "step": 2379
+    },
+    {
+      "epoch": 0.4578684109272797,
+      "grad_norm": 1.3170188939343286,
+      "learning_rate": 1.1840017075273861e-05,
+      "loss": 0.8867,
+      "step": 2380
+    },
+    {
+      "epoch": 0.4580607926125433,
+      "grad_norm": 1.4257182662236545,
+      "learning_rate": 1.1833892258235008e-05,
+      "loss": 0.8957,
+      "step": 2381
+    },
+    {
+      "epoch": 0.4582531742978069,
+      "grad_norm": 2.0019295846087357,
+      "learning_rate": 1.1827766729216035e-05,
+      "loss": 0.8593,
+      "step": 2382
+    },
+    {
+      "epoch": 0.4584455559830704,
+      "grad_norm": 1.2833865382716374,
+      "learning_rate": 1.1821640490595086e-05,
+      "loss": 0.9086,
+      "step": 2383
+    },
+    {
+      "epoch": 0.458637937668334,
+      "grad_norm": 1.5068947791887333,
+      "learning_rate": 1.181551354475058e-05,
+      "loss": 0.8455,
+      "step": 2384
+    },
+    {
+      "epoch": 0.4588303193535975,
+      "grad_norm": 1.6727921089924516,
+      "learning_rate": 1.1809385894061206e-05,
+      "loss": 0.8507,
+      "step": 2385
+    },
+    {
+      "epoch": 0.4590227010388611,
+      "grad_norm": 1.5622451832548576,
+      "learning_rate": 1.1803257540905926e-05,
+      "loss": 0.9604,
+      "step": 2386
+    },
+    {
+      "epoch": 0.45921508272412465,
+      "grad_norm": 1.2292484082867474,
+      "learning_rate": 1.1797128487663982e-05,
+      "loss": 0.9112,
+      "step": 2387
+    },
+    {
+      "epoch": 0.45940746440938823,
+      "grad_norm": 1.6529620696687786,
+      "learning_rate": 1.1790998736714882e-05,
+      "loss": 0.8937,
+      "step": 2388
+    },
+    {
+      "epoch": 0.45959984609465176,
+      "grad_norm": 1.3313980725645937,
+      "learning_rate": 1.1784868290438404e-05,
+      "loss": 0.8298,
+      "step": 2389
+    },
+    {
+      "epoch": 0.45979222777991535,
+      "grad_norm": 1.5952143506399505,
+      "learning_rate": 1.1778737151214606e-05,
+      "loss": 0.8679,
+      "step": 2390
+    },
+    {
+      "epoch": 0.45998460946517894,
+      "grad_norm": 1.4602708110272344,
+      "learning_rate": 1.17726053214238e-05,
+      "loss": 0.8292,
+      "step": 2391
+    },
+    {
+      "epoch": 0.46017699115044247,
+      "grad_norm": 1.3719002155121143,
+      "learning_rate": 1.1766472803446577e-05,
+      "loss": 0.8314,
+      "step": 2392
+    },
+    {
+      "epoch": 0.46036937283570606,
+      "grad_norm": 1.6891928719688925,
+      "learning_rate": 1.1760339599663788e-05,
+      "loss": 0.8612,
+      "step": 2393
+    },
+    {
+      "epoch": 0.4605617545209696,
+      "grad_norm": 1.3457706287559186,
+      "learning_rate": 1.1754205712456556e-05,
+      "loss": 0.8835,
+      "step": 2394
+    },
+    {
+      "epoch": 0.4607541362062332,
+      "grad_norm": 1.433044980697689,
+      "learning_rate": 1.1748071144206266e-05,
+      "loss": 0.9047,
+      "step": 2395
+    },
+    {
+      "epoch": 0.4609465178914967,
+      "grad_norm": 1.8538368633479558,
+      "learning_rate": 1.1741935897294572e-05,
+      "loss": 0.9004,
+      "step": 2396
+    },
+    {
+      "epoch": 0.4611388995767603,
+      "grad_norm": 0.6958197206962649,
+      "learning_rate": 1.1735799974103388e-05,
+      "loss": 0.3404,
+      "step": 2397
+    },
+    {
+      "epoch": 0.46133128126202383,
+      "grad_norm": 1.1782244109390576,
+      "learning_rate": 1.1729663377014888e-05,
+      "loss": 0.896,
+      "step": 2398
+    },
+    {
+      "epoch": 0.4615236629472874,
+      "grad_norm": 1.4241957081052214,
+      "learning_rate": 1.172352610841151e-05,
+      "loss": 0.7999,
+      "step": 2399
+    },
+    {
+      "epoch": 0.461716044632551,
+      "grad_norm": 1.7743599064873652,
+      "learning_rate": 1.1717388170675954e-05,
+      "loss": 0.8797,
+      "step": 2400
+    },
+    {
+      "epoch": 0.46190842631781454,
+      "grad_norm": 2.7589051484032896,
+      "learning_rate": 1.1711249566191179e-05,
+      "loss": 0.9042,
+      "step": 2401
+    },
+    {
+      "epoch": 0.4621008080030781,
+      "grad_norm": 1.5452148513662967,
+      "learning_rate": 1.17051102973404e-05,
+      "loss": 0.8617,
+      "step": 2402
+    },
+    {
+      "epoch": 0.46229318968834165,
+      "grad_norm": 1.76716146332395,
+      "learning_rate": 1.1698970366507096e-05,
+      "loss": 0.8814,
+      "step": 2403
+    },
+    {
+      "epoch": 0.46248557137360524,
+      "grad_norm": 1.166325165215305,
+      "learning_rate": 1.1692829776074999e-05,
+      "loss": 0.9091,
+      "step": 2404
+    },
+    {
+      "epoch": 0.4626779530588688,
+      "grad_norm": 1.3591560583646376,
+      "learning_rate": 1.1686688528428099e-05,
+      "loss": 0.8691,
+      "step": 2405
+    },
+    {
+      "epoch": 0.46287033474413236,
+      "grad_norm": 1.5500475021559323,
+      "learning_rate": 1.1680546625950635e-05,
+      "loss": 0.878,
+      "step": 2406
+    },
+    {
+      "epoch": 0.46306271642939595,
+      "grad_norm": 1.2501196361207618,
+      "learning_rate": 1.167440407102711e-05,
+      "loss": 0.9399,
+      "step": 2407
+    },
+    {
+      "epoch": 0.4632550981146595,
+      "grad_norm": 1.2017669615733608,
+      "learning_rate": 1.1668260866042271e-05,
+      "loss": 0.8549,
+      "step": 2408
+    },
+    {
+      "epoch": 0.46344747979992307,
+      "grad_norm": 1.2469385585704915,
+      "learning_rate": 1.1662117013381126e-05,
+      "loss": 0.7902,
+      "step": 2409
+    },
+    {
+      "epoch": 0.4636398614851866,
+      "grad_norm": 2.5254655274445685,
+      "learning_rate": 1.1655972515428928e-05,
+      "loss": 0.8711,
+      "step": 2410
+    },
+    {
+      "epoch": 0.4638322431704502,
+      "grad_norm": 1.3859744964734475,
+      "learning_rate": 1.1649827374571182e-05,
+      "loss": 0.9182,
+      "step": 2411
+    },
+    {
+      "epoch": 0.4640246248557137,
+      "grad_norm": 1.3763081759415845,
+      "learning_rate": 1.1643681593193642e-05,
+      "loss": 0.8679,
+      "step": 2412
+    },
+    {
+      "epoch": 0.4642170065409773,
+      "grad_norm": 1.59263286382257,
+      "learning_rate": 1.1637535173682318e-05,
+      "loss": 0.851,
+      "step": 2413
+    },
+    {
+      "epoch": 0.46440938822624084,
+      "grad_norm": 1.2901909699855454,
+      "learning_rate": 1.1631388118423457e-05,
+      "loss": 0.8929,
+      "step": 2414
+    },
+    {
+      "epoch": 0.4646017699115044,
+      "grad_norm": 0.6558565925558054,
+      "learning_rate": 1.1625240429803553e-05,
+      "loss": 0.337,
+      "step": 2415
+    },
+    {
+      "epoch": 0.464794151596768,
+      "grad_norm": 0.651151809165468,
+      "learning_rate": 1.1619092110209361e-05,
+      "loss": 0.3665,
+      "step": 2416
+    },
+    {
+      "epoch": 0.46498653328203154,
+      "grad_norm": 1.5422251695432274,
+      "learning_rate": 1.1612943162027863e-05,
+      "loss": 0.9103,
+      "step": 2417
+    },
+    {
+      "epoch": 0.46517891496729513,
+      "grad_norm": 1.455639354039091,
+      "learning_rate": 1.1606793587646295e-05,
+      "loss": 0.8583,
+      "step": 2418
+    },
+    {
+      "epoch": 0.46537129665255866,
+      "grad_norm": 1.424393629903972,
+      "learning_rate": 1.160064338945213e-05,
+      "loss": 0.9246,
+      "step": 2419
+    },
+    {
+      "epoch": 0.46556367833782225,
+      "grad_norm": 1.6658994463833459,
+      "learning_rate": 1.1594492569833093e-05,
+      "loss": 0.9238,
+      "step": 2420
+    },
+    {
+      "epoch": 0.4657560600230858,
+      "grad_norm": 0.6063527955414815,
+      "learning_rate": 1.1588341131177137e-05,
+      "loss": 0.3594,
+      "step": 2421
+    },
+    {
+      "epoch": 0.46594844170834937,
+      "grad_norm": 1.3377138836174234,
+      "learning_rate": 1.1582189075872467e-05,
+      "loss": 0.9109,
+      "step": 2422
+    },
+    {
+      "epoch": 0.4661408233936129,
+      "grad_norm": 1.3396335870467238,
+      "learning_rate": 1.1576036406307523e-05,
+      "loss": 0.8627,
+      "step": 2423
+    },
+    {
+      "epoch": 0.4663332050788765,
+      "grad_norm": 1.3867773691735772,
+      "learning_rate": 1.156988312487098e-05,
+      "loss": 0.8693,
+      "step": 2424
+    },
+    {
+      "epoch": 0.4665255867641401,
+      "grad_norm": 1.2727251840538993,
+      "learning_rate": 1.1563729233951757e-05,
+      "loss": 0.9255,
+      "step": 2425
+    },
+    {
+      "epoch": 0.4667179684494036,
+      "grad_norm": 1.3468997781822558,
+      "learning_rate": 1.1557574735939003e-05,
+      "loss": 0.886,
+      "step": 2426
+    },
+    {
+      "epoch": 0.4669103501346672,
+      "grad_norm": 1.2568505728261807,
+      "learning_rate": 1.1551419633222107e-05,
+      "loss": 0.8474,
+      "step": 2427
+    },
+    {
+      "epoch": 0.4671027318199307,
+      "grad_norm": 1.1599233747364674,
+      "learning_rate": 1.1545263928190692e-05,
+      "loss": 0.8749,
+      "step": 2428
+    },
+    {
+      "epoch": 0.4672951135051943,
+      "grad_norm": 1.2577515510141866,
+      "learning_rate": 1.1539107623234618e-05,
+      "loss": 0.8867,
+      "step": 2429
+    },
+    {
+      "epoch": 0.46748749519045785,
+      "grad_norm": 1.2959659088623776,
+      "learning_rate": 1.153295072074397e-05,
+      "loss": 0.8921,
+      "step": 2430
+    },
+    {
+      "epoch": 0.46767987687572143,
+      "grad_norm": 1.3177243295614713,
+      "learning_rate": 1.1526793223109072e-05,
+      "loss": 0.8772,
+      "step": 2431
+    },
+    {
+      "epoch": 0.467872258560985,
+      "grad_norm": 1.4597841385237924,
+      "learning_rate": 1.1520635132720475e-05,
+      "loss": 0.8579,
+      "step": 2432
+    },
+    {
+      "epoch": 0.46806464024624855,
+      "grad_norm": 1.2955914481059385,
+      "learning_rate": 1.1514476451968961e-05,
+      "loss": 0.911,
+      "step": 2433
+    },
+    {
+      "epoch": 0.46825702193151214,
+      "grad_norm": 1.2670131363990864,
+      "learning_rate": 1.1508317183245545e-05,
+      "loss": 0.8399,
+      "step": 2434
+    },
+    {
+      "epoch": 0.4684494036167757,
+      "grad_norm": 1.4885709534281208,
+      "learning_rate": 1.1502157328941466e-05,
+      "loss": 0.8862,
+      "step": 2435
+    },
+    {
+      "epoch": 0.46864178530203926,
+      "grad_norm": 1.2348395392978038,
+      "learning_rate": 1.149599689144819e-05,
+      "loss": 0.8954,
+      "step": 2436
+    },
+    {
+      "epoch": 0.4688341669873028,
+      "grad_norm": 1.2956092651254427,
+      "learning_rate": 1.1489835873157414e-05,
+      "loss": 0.9107,
+      "step": 2437
+    },
+    {
+      "epoch": 0.4690265486725664,
+      "grad_norm": 1.0757183018120406,
+      "learning_rate": 1.1483674276461053e-05,
+      "loss": 0.8558,
+      "step": 2438
+    },
+    {
+      "epoch": 0.4692189303578299,
+      "grad_norm": 1.8025980493466771,
+      "learning_rate": 1.1477512103751254e-05,
+      "loss": 0.9093,
+      "step": 2439
+    },
+    {
+      "epoch": 0.4694113120430935,
+      "grad_norm": 1.3724256071700227,
+      "learning_rate": 1.1471349357420384e-05,
+      "loss": 0.9004,
+      "step": 2440
+    },
+    {
+      "epoch": 0.4696036937283571,
+      "grad_norm": 1.453051437217561,
+      "learning_rate": 1.1465186039861033e-05,
+      "loss": 0.8691,
+      "step": 2441
+    },
+    {
+      "epoch": 0.4697960754136206,
+      "grad_norm": 1.2876697567435536,
+      "learning_rate": 1.1459022153466016e-05,
+      "loss": 0.9091,
+      "step": 2442
+    },
+    {
+      "epoch": 0.4699884570988842,
+      "grad_norm": 1.3651900642387533,
+      "learning_rate": 1.1452857700628362e-05,
+      "loss": 0.9088,
+      "step": 2443
+    },
+    {
+      "epoch": 0.47018083878414774,
+      "grad_norm": 1.3032200388002275,
+      "learning_rate": 1.1446692683741326e-05,
+      "loss": 0.866,
+      "step": 2444
+    },
+    {
+      "epoch": 0.4703732204694113,
+      "grad_norm": 1.5516961954708168,
+      "learning_rate": 1.1440527105198377e-05,
+      "loss": 0.9049,
+      "step": 2445
+    },
+    {
+      "epoch": 0.47056560215467486,
+      "grad_norm": 1.5557540786597273,
+      "learning_rate": 1.143436096739321e-05,
+      "loss": 0.8537,
+      "step": 2446
+    },
+    {
+      "epoch": 0.47075798383993844,
+      "grad_norm": 1.4608685007428481,
+      "learning_rate": 1.1428194272719729e-05,
+      "loss": 0.8844,
+      "step": 2447
+    },
+    {
+      "epoch": 0.470950365525202,
+      "grad_norm": 1.1708228494427395,
+      "learning_rate": 1.1422027023572052e-05,
+      "loss": 0.9,
+      "step": 2448
+    },
+    {
+      "epoch": 0.47114274721046556,
+      "grad_norm": 1.2236791317055302,
+      "learning_rate": 1.1415859222344525e-05,
+      "loss": 0.8728,
+      "step": 2449
+    },
+    {
+      "epoch": 0.47133512889572915,
+      "grad_norm": 1.4482721577774529,
+      "learning_rate": 1.14096908714317e-05,
+      "loss": 0.8713,
+      "step": 2450
+    },
+    {
+      "epoch": 0.4715275105809927,
+      "grad_norm": 5.291970154344779,
+      "learning_rate": 1.1403521973228342e-05,
+      "loss": 0.8579,
+      "step": 2451
+    },
+    {
+      "epoch": 0.47171989226625627,
+      "grad_norm": 1.2008876054317783,
+      "learning_rate": 1.1397352530129428e-05,
+      "loss": 0.9055,
+      "step": 2452
+    },
+    {
+      "epoch": 0.4719122739515198,
+      "grad_norm": 1.380122951891391,
+      "learning_rate": 1.139118254453015e-05,
+      "loss": 0.923,
+      "step": 2453
+    },
+    {
+      "epoch": 0.4721046556367834,
+      "grad_norm": 1.2775924626465986,
+      "learning_rate": 1.1385012018825907e-05,
+      "loss": 0.859,
+      "step": 2454
+    },
+    {
+      "epoch": 0.4722970373220469,
+      "grad_norm": 0.7116998048951882,
+      "learning_rate": 1.1378840955412313e-05,
+      "loss": 0.3487,
+      "step": 2455
+    },
+    {
+      "epoch": 0.4724894190073105,
+      "grad_norm": 1.2992222188950577,
+      "learning_rate": 1.1372669356685185e-05,
+      "loss": 0.9106,
+      "step": 2456
+    },
+    {
+      "epoch": 0.4726818006925741,
+      "grad_norm": 1.4912845713138947,
+      "learning_rate": 1.1366497225040549e-05,
+      "loss": 0.8538,
+      "step": 2457
+    },
+    {
+      "epoch": 0.4728741823778376,
+      "grad_norm": 1.3866317137201503,
+      "learning_rate": 1.1360324562874643e-05,
+      "loss": 0.8971,
+      "step": 2458
+    },
+    {
+      "epoch": 0.4730665640631012,
+      "grad_norm": 1.384870268821325,
+      "learning_rate": 1.1354151372583901e-05,
+      "loss": 0.9153,
+      "step": 2459
+    },
+    {
+      "epoch": 0.47325894574836475,
+      "grad_norm": 1.478451015632857,
+      "learning_rate": 1.1347977656564974e-05,
+      "loss": 0.8641,
+      "step": 2460
+    },
+    {
+      "epoch": 0.47345132743362833,
+      "grad_norm": 1.3766175132593759,
+      "learning_rate": 1.1341803417214705e-05,
+      "loss": 0.8604,
+      "step": 2461
+    },
+    {
+      "epoch": 0.47364370911889186,
+      "grad_norm": 1.5541309260496983,
+      "learning_rate": 1.1335628656930153e-05,
+      "loss": 0.9049,
+      "step": 2462
+    },
+    {
+      "epoch": 0.47383609080415545,
+      "grad_norm": 1.3857296297759318,
+      "learning_rate": 1.132945337810857e-05,
+      "loss": 0.8459,
+      "step": 2463
+    },
+    {
+      "epoch": 0.474028472489419,
+      "grad_norm": 1.3188698064664872,
+      "learning_rate": 1.132327758314741e-05,
+      "loss": 0.8284,
+      "step": 2464
+    },
+    {
+      "epoch": 0.47422085417468257,
+      "grad_norm": 1.665668132475729,
+      "learning_rate": 1.131710127444433e-05,
+      "loss": 0.8981,
+      "step": 2465
+    },
+    {
+      "epoch": 0.47441323585994616,
+      "grad_norm": 1.5332969974086712,
+      "learning_rate": 1.1310924454397187e-05,
+      "loss": 0.8707,
+      "step": 2466
+    },
+    {
+      "epoch": 0.4746056175452097,
+      "grad_norm": 1.218751967061097,
+      "learning_rate": 1.1304747125404031e-05,
+      "loss": 0.8768,
+      "step": 2467
+    },
+    {
+      "epoch": 0.4747979992304733,
+      "grad_norm": 1.4263389700089502,
+      "learning_rate": 1.129856928986312e-05,
+      "loss": 0.9613,
+      "step": 2468
+    },
+    {
+      "epoch": 0.4749903809157368,
+      "grad_norm": 1.2995116456385234,
+      "learning_rate": 1.12923909501729e-05,
+      "loss": 0.8677,
+      "step": 2469
+    },
+    {
+      "epoch": 0.4751827626010004,
+      "grad_norm": 1.4538081341810285,
+      "learning_rate": 1.1286212108732015e-05,
+      "loss": 0.8976,
+      "step": 2470
+    },
+    {
+      "epoch": 0.47537514428626393,
+      "grad_norm": 1.5721290281899418,
+      "learning_rate": 1.1280032767939302e-05,
+      "loss": 0.8328,
+      "step": 2471
+    },
+    {
+      "epoch": 0.4755675259715275,
+      "grad_norm": 1.675484484002741,
+      "learning_rate": 1.1273852930193798e-05,
+      "loss": 0.8485,
+      "step": 2472
+    },
+    {
+      "epoch": 0.47575990765679105,
+      "grad_norm": 1.3972474754868724,
+      "learning_rate": 1.1267672597894725e-05,
+      "loss": 0.9168,
+      "step": 2473
+    },
+    {
+      "epoch": 0.47595228934205464,
+      "grad_norm": 1.3765035514448043,
+      "learning_rate": 1.12614917734415e-05,
+      "loss": 0.7519,
+      "step": 2474
+    },
+    {
+      "epoch": 0.4761446710273182,
+      "grad_norm": 1.188334416719753,
+      "learning_rate": 1.1255310459233737e-05,
+      "loss": 0.8595,
+      "step": 2475
+    },
+    {
+      "epoch": 0.47633705271258175,
+      "grad_norm": 1.1568614303854958,
+      "learning_rate": 1.1249128657671233e-05,
+      "loss": 0.9303,
+      "step": 2476
+    },
+    {
+      "epoch": 0.47652943439784534,
+      "grad_norm": 1.9746035440486511,
+      "learning_rate": 1.1242946371153974e-05,
+      "loss": 0.884,
+      "step": 2477
+    },
+    {
+      "epoch": 0.4767218160831089,
+      "grad_norm": 1.5591156236711852,
+      "learning_rate": 1.1236763602082136e-05,
+      "loss": 0.9142,
+      "step": 2478
+    },
+    {
+      "epoch": 0.47691419776837246,
+      "grad_norm": 1.477653664614633,
+      "learning_rate": 1.1230580352856088e-05,
+      "loss": 0.8912,
+      "step": 2479
+    },
+    {
+      "epoch": 0.477106579453636,
+      "grad_norm": 1.3047597567835165,
+      "learning_rate": 1.1224396625876375e-05,
+      "loss": 0.877,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4772989611388996,
+      "grad_norm": 1.2861543432895053,
+      "learning_rate": 1.1218212423543734e-05,
+      "loss": 0.8473,
+      "step": 2481
+    },
+    {
+      "epoch": 0.4774913428241631,
+      "grad_norm": 1.6430109195336124,
+      "learning_rate": 1.1212027748259086e-05,
+      "loss": 0.9446,
+      "step": 2482
+    },
+    {
+      "epoch": 0.4776837245094267,
+      "grad_norm": 1.3350825727197237,
+      "learning_rate": 1.1205842602423537e-05,
+      "loss": 0.8496,
+      "step": 2483
+    },
+    {
+      "epoch": 0.4778761061946903,
+      "grad_norm": 1.629120436918724,
+      "learning_rate": 1.1199656988438373e-05,
+      "loss": 0.8427,
+      "step": 2484
+    },
+    {
+      "epoch": 0.4780684878799538,
+      "grad_norm": 2.0594556441836103,
+      "learning_rate": 1.1193470908705055e-05,
+      "loss": 0.9139,
+      "step": 2485
+    },
+    {
+      "epoch": 0.4782608695652174,
+      "grad_norm": 1.8972144382307294,
+      "learning_rate": 1.1187284365625241e-05,
+      "loss": 0.9123,
+      "step": 2486
+    },
+    {
+      "epoch": 0.47845325125048094,
+      "grad_norm": 1.5157510677218495,
+      "learning_rate": 1.1181097361600754e-05,
+      "loss": 0.8873,
+      "step": 2487
+    },
+    {
+      "epoch": 0.4786456329357445,
+      "grad_norm": 1.3943545413865102,
+      "learning_rate": 1.1174909899033608e-05,
+      "loss": 0.7931,
+      "step": 2488
+    },
+    {
+      "epoch": 0.47883801462100806,
+      "grad_norm": 1.4263967187641726,
+      "learning_rate": 1.1168721980325987e-05,
+      "loss": 0.9392,
+      "step": 2489
+    },
+    {
+      "epoch": 0.47903039630627164,
+      "grad_norm": 1.9317885005531596,
+      "learning_rate": 1.1162533607880251e-05,
+      "loss": 0.866,
+      "step": 2490
+    },
+    {
+      "epoch": 0.47922277799153523,
+      "grad_norm": 1.2631185222043222,
+      "learning_rate": 1.1156344784098942e-05,
+      "loss": 0.9213,
+      "step": 2491
+    },
+    {
+      "epoch": 0.47941515967679876,
+      "grad_norm": 1.4676704070188675,
+      "learning_rate": 1.1150155511384772e-05,
+      "loss": 0.8767,
+      "step": 2492
+    },
+    {
+      "epoch": 0.47960754136206235,
+      "grad_norm": 1.3024770433969826,
+      "learning_rate": 1.1143965792140631e-05,
+      "loss": 0.8903,
+      "step": 2493
+    },
+    {
+      "epoch": 0.4797999230473259,
+      "grad_norm": 1.4142390354628591,
+      "learning_rate": 1.1137775628769584e-05,
+      "loss": 0.9457,
+      "step": 2494
+    },
+    {
+      "epoch": 0.47999230473258947,
+      "grad_norm": 1.2065668138350312,
+      "learning_rate": 1.1131585023674863e-05,
+      "loss": 0.9233,
+      "step": 2495
+    },
+    {
+      "epoch": 0.480184686417853,
+      "grad_norm": 3.1624670481343826,
+      "learning_rate": 1.1125393979259874e-05,
+      "loss": 0.8423,
+      "step": 2496
+    },
+    {
+      "epoch": 0.4803770681031166,
+      "grad_norm": 1.5986995125990635,
+      "learning_rate": 1.1119202497928192e-05,
+      "loss": 0.8499,
+      "step": 2497
+    },
+    {
+      "epoch": 0.4805694497883801,
+      "grad_norm": 1.2702968562154133,
+      "learning_rate": 1.1113010582083568e-05,
+      "loss": 0.8896,
+      "step": 2498
+    },
+    {
+      "epoch": 0.4807618314736437,
+      "grad_norm": 1.2987249459808028,
+      "learning_rate": 1.1106818234129913e-05,
+      "loss": 0.8817,
+      "step": 2499
+    },
+    {
+      "epoch": 0.4809542131589073,
+      "grad_norm": 1.2725973247074867,
+      "learning_rate": 1.1100625456471307e-05,
+      "loss": 0.8802,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4811465948441708,
+      "grad_norm": 1.4402513337332434,
+      "learning_rate": 1.1094432251512006e-05,
+      "loss": 0.8747,
+      "step": 2501
+    },
+    {
+      "epoch": 0.4813389765294344,
+      "grad_norm": 1.3323240832447882,
+      "learning_rate": 1.1088238621656422e-05,
+      "loss": 0.8641,
+      "step": 2502
+    },
+    {
+      "epoch": 0.48153135821469795,
+      "grad_norm": 1.8958058994050562,
+      "learning_rate": 1.1082044569309138e-05,
+      "loss": 0.922,
+      "step": 2503
+    },
+    {
+      "epoch": 0.48172373989996153,
+      "grad_norm": 1.7003433126671097,
+      "learning_rate": 1.1075850096874894e-05,
+      "loss": 0.8905,
+      "step": 2504
+    },
+    {
+      "epoch": 0.48191612158522507,
+      "grad_norm": 0.7240490273656374,
+      "learning_rate": 1.1069655206758603e-05,
+      "loss": 0.3416,
+      "step": 2505
+    },
+    {
+      "epoch": 0.48210850327048865,
+      "grad_norm": 1.2248792202805072,
+      "learning_rate": 1.1063459901365325e-05,
+      "loss": 0.9282,
+      "step": 2506
+    },
+    {
+      "epoch": 0.4823008849557522,
+      "grad_norm": 2.017787884542109,
+      "learning_rate": 1.1057264183100303e-05,
+      "loss": 0.9091,
+      "step": 2507
+    },
+    {
+      "epoch": 0.48249326664101577,
+      "grad_norm": 1.2478888329596203,
+      "learning_rate": 1.1051068054368921e-05,
+      "loss": 0.8449,
+      "step": 2508
+    },
+    {
+      "epoch": 0.48268564832627936,
+      "grad_norm": 1.1371701269222578,
+      "learning_rate": 1.104487151757673e-05,
+      "loss": 0.8481,
+      "step": 2509
+    },
+    {
+      "epoch": 0.4828780300115429,
+      "grad_norm": 1.3236796712902124,
+      "learning_rate": 1.1038674575129442e-05,
+      "loss": 0.8464,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4830704116968065,
+      "grad_norm": 1.3925832471569313,
+      "learning_rate": 1.1032477229432921e-05,
+      "loss": 0.872,
+      "step": 2511
+    },
+    {
+      "epoch": 0.48326279338207,
+      "grad_norm": 0.6320617084029329,
+      "learning_rate": 1.1026279482893187e-05,
+      "loss": 0.3573,
+      "step": 2512
+    },
+    {
+      "epoch": 0.4834551750673336,
+      "grad_norm": 1.1881131509096503,
+      "learning_rate": 1.1020081337916425e-05,
+      "loss": 0.912,
+      "step": 2513
+    },
+    {
+      "epoch": 0.48364755675259713,
+      "grad_norm": 0.626304144819442,
+      "learning_rate": 1.1013882796908963e-05,
+      "loss": 0.3675,
+      "step": 2514
+    },
+    {
+      "epoch": 0.4838399384378607,
+      "grad_norm": 1.2081816912348387,
+      "learning_rate": 1.1007683862277292e-05,
+      "loss": 0.8643,
+      "step": 2515
+    },
+    {
+      "epoch": 0.4840323201231243,
+      "grad_norm": 1.3953854683201559,
+      "learning_rate": 1.1001484536428052e-05,
+      "loss": 0.8621,
+      "step": 2516
+    },
+    {
+      "epoch": 0.48422470180838784,
+      "grad_norm": 1.8824810198259705,
+      "learning_rate": 1.0995284821768029e-05,
+      "loss": 0.8706,
+      "step": 2517
+    },
+    {
+      "epoch": 0.4844170834936514,
+      "grad_norm": 1.153783936877306,
+      "learning_rate": 1.098908472070417e-05,
+      "loss": 0.8809,
+      "step": 2518
+    },
+    {
+      "epoch": 0.48460946517891496,
+      "grad_norm": 1.4684892921933943,
+      "learning_rate": 1.0982884235643567e-05,
+      "loss": 0.9052,
+      "step": 2519
+    },
+    {
+      "epoch": 0.48480184686417854,
+      "grad_norm": 2.334756739471636,
+      "learning_rate": 1.0976683368993464e-05,
+      "loss": 0.9325,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4849942285494421,
+      "grad_norm": 1.145497457194484,
+      "learning_rate": 1.0970482123161249e-05,
+      "loss": 0.9489,
+      "step": 2521
+    },
+    {
+      "epoch": 0.48518661023470566,
+      "grad_norm": 1.2219099380809377,
+      "learning_rate": 1.0964280500554459e-05,
+      "loss": 0.8974,
+      "step": 2522
+    },
+    {
+      "epoch": 0.4853789919199692,
+      "grad_norm": 1.249648762301962,
+      "learning_rate": 1.0958078503580776e-05,
+      "loss": 0.9203,
+      "step": 2523
+    },
+    {
+      "epoch": 0.4855713736052328,
+      "grad_norm": 1.478907078156249,
+      "learning_rate": 1.0951876134648032e-05,
+      "loss": 0.881,
+      "step": 2524
+    },
+    {
+      "epoch": 0.48576375529049637,
+      "grad_norm": 1.2898213968924455,
+      "learning_rate": 1.0945673396164198e-05,
+      "loss": 0.8779,
+      "step": 2525
+    },
+    {
+      "epoch": 0.4859561369757599,
+      "grad_norm": 1.2348231821249072,
+      "learning_rate": 1.0939470290537389e-05,
+      "loss": 0.9015,
+      "step": 2526
+    },
+    {
+      "epoch": 0.4861485186610235,
+      "grad_norm": 1.3015576081008158,
+      "learning_rate": 1.0933266820175868e-05,
+      "loss": 0.8401,
+      "step": 2527
+    },
+    {
+      "epoch": 0.486340900346287,
+      "grad_norm": 0.620057932041844,
+      "learning_rate": 1.0927062987488035e-05,
+      "loss": 0.3431,
+      "step": 2528
+    },
+    {
+      "epoch": 0.4865332820315506,
+      "grad_norm": 1.1867102198492956,
+      "learning_rate": 1.0920858794882429e-05,
+      "loss": 0.8852,
+      "step": 2529
+    },
+    {
+      "epoch": 0.48672566371681414,
+      "grad_norm": 1.3060238824452788,
+      "learning_rate": 1.0914654244767736e-05,
+      "loss": 0.9039,
+      "step": 2530
+    },
+    {
+      "epoch": 0.4869180454020777,
+      "grad_norm": 0.6313275183234095,
+      "learning_rate": 1.0908449339552769e-05,
+      "loss": 0.3168,
+      "step": 2531
+    },
+    {
+      "epoch": 0.48711042708734126,
+      "grad_norm": 1.4148713278622789,
+      "learning_rate": 1.0902244081646489e-05,
+      "loss": 0.8203,
+      "step": 2532
+    },
+    {
+      "epoch": 0.48730280877260485,
+      "grad_norm": 1.1972594078440943,
+      "learning_rate": 1.0896038473457993e-05,
+      "loss": 0.8307,
+      "step": 2533
+    },
+    {
+      "epoch": 0.48749519045786843,
+      "grad_norm": 1.2767237979759685,
+      "learning_rate": 1.0889832517396511e-05,
+      "loss": 0.8695,
+      "step": 2534
+    },
+    {
+      "epoch": 0.48768757214313196,
+      "grad_norm": 1.26199061081444,
+      "learning_rate": 1.0883626215871408e-05,
+      "loss": 0.8869,
+      "step": 2535
+    },
+    {
+      "epoch": 0.48787995382839555,
+      "grad_norm": 1.208412242290822,
+      "learning_rate": 1.0877419571292183e-05,
+      "loss": 0.9003,
+      "step": 2536
+    },
+    {
+      "epoch": 0.4880723355136591,
+      "grad_norm": 1.2957489133899442,
+      "learning_rate": 1.0871212586068469e-05,
+      "loss": 0.9152,
+      "step": 2537
+    },
+    {
+      "epoch": 0.48826471719892267,
+      "grad_norm": 1.3976364501412146,
+      "learning_rate": 1.0865005262610033e-05,
+      "loss": 0.8567,
+      "step": 2538
+    },
+    {
+      "epoch": 0.4884570988841862,
+      "grad_norm": 1.4134228425913793,
+      "learning_rate": 1.085879760332677e-05,
+      "loss": 0.8557,
+      "step": 2539
+    },
+    {
+      "epoch": 0.4886494805694498,
+      "grad_norm": 1.5610071701402721,
+      "learning_rate": 1.085258961062871e-05,
+      "loss": 0.8735,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4888418622547134,
+      "grad_norm": 1.1816303869902596,
+      "learning_rate": 1.0846381286926007e-05,
+      "loss": 0.8375,
+      "step": 2541
+    },
+    {
+      "epoch": 0.4890342439399769,
+      "grad_norm": 1.1930093107125361,
+      "learning_rate": 1.0840172634628948e-05,
+      "loss": 0.8505,
+      "step": 2542
+    },
+    {
+      "epoch": 0.4892266256252405,
+      "grad_norm": 1.2951490674799433,
+      "learning_rate": 1.0833963656147944e-05,
+      "loss": 0.8521,
+      "step": 2543
+    },
+    {
+      "epoch": 0.48941900731050403,
+      "grad_norm": 0.6935601229508069,
+      "learning_rate": 1.082775435389353e-05,
+      "loss": 0.3395,
+      "step": 2544
+    },
+    {
+      "epoch": 0.4896113889957676,
+      "grad_norm": 1.7787048826160805,
+      "learning_rate": 1.0821544730276379e-05,
+      "loss": 0.8802,
+      "step": 2545
+    },
+    {
+      "epoch": 0.48980377068103115,
+      "grad_norm": 1.1840778731028307,
+      "learning_rate": 1.0815334787707277e-05,
+      "loss": 0.834,
+      "step": 2546
+    },
+    {
+      "epoch": 0.48999615236629473,
+      "grad_norm": 1.172282628787275,
+      "learning_rate": 1.0809124528597138e-05,
+      "loss": 0.8673,
+      "step": 2547
+    },
+    {
+      "epoch": 0.49018853405155827,
+      "grad_norm": 1.2954956287600776,
+      "learning_rate": 1.0802913955356998e-05,
+      "loss": 0.9118,
+      "step": 2548
+    },
+    {
+      "epoch": 0.49038091573682185,
+      "grad_norm": 1.3494198400190514,
+      "learning_rate": 1.0796703070398016e-05,
+      "loss": 0.8708,
+      "step": 2549
+    },
+    {
+      "epoch": 0.49057329742208544,
+      "grad_norm": 1.6042991439282308,
+      "learning_rate": 1.079049187613147e-05,
+      "loss": 0.8867,
+      "step": 2550
+    },
+    {
+      "epoch": 0.490765679107349,
+      "grad_norm": 1.4206487739385238,
+      "learning_rate": 1.0784280374968761e-05,
+      "loss": 0.901,
+      "step": 2551
+    },
+    {
+      "epoch": 0.49095806079261256,
+      "grad_norm": 1.1979676867882318,
+      "learning_rate": 1.0778068569321403e-05,
+      "loss": 0.9238,
+      "step": 2552
+    },
+    {
+      "epoch": 0.4911504424778761,
+      "grad_norm": 1.5212649202700768,
+      "learning_rate": 1.077185646160104e-05,
+      "loss": 0.907,
+      "step": 2553
+    },
+    {
+      "epoch": 0.4913428241631397,
+      "grad_norm": 1.471556701440831,
+      "learning_rate": 1.0765644054219422e-05,
+      "loss": 0.8928,
+      "step": 2554
+    },
+    {
+      "epoch": 0.4915352058484032,
+      "grad_norm": 1.3913828751059285,
+      "learning_rate": 1.0759431349588421e-05,
+      "loss": 0.8439,
+      "step": 2555
+    },
+    {
+      "epoch": 0.4917275875336668,
+      "grad_norm": 1.2783991858303432,
+      "learning_rate": 1.0753218350120023e-05,
+      "loss": 0.9108,
+      "step": 2556
+    },
+    {
+      "epoch": 0.49191996921893033,
+      "grad_norm": 1.2487003137631123,
+      "learning_rate": 1.0747005058226325e-05,
+      "loss": 0.8558,
+      "step": 2557
+    },
+    {
+      "epoch": 0.4921123509041939,
+      "grad_norm": 1.429763243445838,
+      "learning_rate": 1.0740791476319543e-05,
+      "loss": 0.828,
+      "step": 2558
+    },
+    {
+      "epoch": 0.4923047325894575,
+      "grad_norm": 1.436344470895534,
+      "learning_rate": 1.0734577606812007e-05,
+      "loss": 0.8534,
+      "step": 2559
+    },
+    {
+      "epoch": 0.49249711427472104,
+      "grad_norm": 1.4186722910472411,
+      "learning_rate": 1.0728363452116149e-05,
+      "loss": 0.8782,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4926894959599846,
+      "grad_norm": 1.3318190283165618,
+      "learning_rate": 1.0722149014644523e-05,
+      "loss": 0.8501,
+      "step": 2561
+    },
+    {
+      "epoch": 0.49288187764524816,
+      "grad_norm": 1.4698167881108208,
+      "learning_rate": 1.0715934296809782e-05,
+      "loss": 0.8593,
+      "step": 2562
+    },
+    {
+      "epoch": 0.49307425933051174,
+      "grad_norm": 1.2945539164040079,
+      "learning_rate": 1.0709719301024698e-05,
+      "loss": 0.8873,
+      "step": 2563
+    },
+    {
+      "epoch": 0.4932666410157753,
+      "grad_norm": 1.558458146976519,
+      "learning_rate": 1.0703504029702148e-05,
+      "loss": 0.9079,
+      "step": 2564
+    },
+    {
+      "epoch": 0.49345902270103886,
+      "grad_norm": 1.4985314845466329,
+      "learning_rate": 1.0697288485255107e-05,
+      "loss": 0.9,
+      "step": 2565
+    },
+    {
+      "epoch": 0.49365140438630245,
+      "grad_norm": 0.6701808316017512,
+      "learning_rate": 1.0691072670096669e-05,
+      "loss": 0.3429,
+      "step": 2566
+    },
+    {
+      "epoch": 0.493843786071566,
+      "grad_norm": 1.2354588174758683,
+      "learning_rate": 1.0684856586640026e-05,
+      "loss": 0.832,
+      "step": 2567
+    },
+    {
+      "epoch": 0.49403616775682957,
+      "grad_norm": 1.955894037270844,
+      "learning_rate": 1.0678640237298476e-05,
+      "loss": 0.8741,
+      "step": 2568
+    },
+    {
+      "epoch": 0.4942285494420931,
+      "grad_norm": 1.15448587211266,
+      "learning_rate": 1.0672423624485423e-05,
+      "loss": 0.8725,
+      "step": 2569
+    },
+    {
+      "epoch": 0.4944209311273567,
+      "grad_norm": 1.3863619717370927,
+      "learning_rate": 1.0666206750614363e-05,
+      "loss": 0.8672,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4946133128126202,
+      "grad_norm": 6.3596614223631,
+      "learning_rate": 1.0659989618098904e-05,
+      "loss": 0.8888,
+      "step": 2571
+    },
+    {
+      "epoch": 0.4948056944978838,
+      "grad_norm": 1.347260926959854,
+      "learning_rate": 1.065377222935275e-05,
+      "loss": 0.9183,
+      "step": 2572
+    },
+    {
+      "epoch": 0.49499807618314734,
+      "grad_norm": 1.5129883811311078,
+      "learning_rate": 1.0647554586789708e-05,
+      "loss": 0.9237,
+      "step": 2573
+    },
+    {
+      "epoch": 0.4951904578684109,
+      "grad_norm": 1.162480340616802,
+      "learning_rate": 1.064133669282368e-05,
+      "loss": 0.8648,
+      "step": 2574
+    },
+    {
+      "epoch": 0.4953828395536745,
+      "grad_norm": 1.3820542055895668,
+      "learning_rate": 1.0635118549868668e-05,
+      "loss": 0.8868,
+      "step": 2575
+    },
+    {
+      "epoch": 0.49557522123893805,
+      "grad_norm": 1.3338358551704212,
+      "learning_rate": 1.0628900160338764e-05,
+      "loss": 0.844,
+      "step": 2576
+    },
+    {
+      "epoch": 0.49576760292420163,
+      "grad_norm": 2.0175104424746007,
+      "learning_rate": 1.0622681526648167e-05,
+      "loss": 0.817,
+      "step": 2577
+    },
+    {
+      "epoch": 0.49595998460946517,
+      "grad_norm": 1.4837560054787071,
+      "learning_rate": 1.0616462651211156e-05,
+      "loss": 0.8526,
+      "step": 2578
+    },
+    {
+      "epoch": 0.49615236629472875,
+      "grad_norm": 1.0894353454285515,
+      "learning_rate": 1.0610243536442125e-05,
+      "loss": 0.7964,
+      "step": 2579
+    },
+    {
+      "epoch": 0.4963447479799923,
+      "grad_norm": 0.6574634759346526,
+      "learning_rate": 1.0604024184755539e-05,
+      "loss": 0.3504,
+      "step": 2580
+    },
+    {
+      "epoch": 0.49653712966525587,
+      "grad_norm": 1.3092662224276759,
+      "learning_rate": 1.0597804598565969e-05,
+      "loss": 0.9427,
+      "step": 2581
+    },
+    {
+      "epoch": 0.4967295113505194,
+      "grad_norm": 1.3260763440576313,
+      "learning_rate": 1.0591584780288069e-05,
+      "loss": 0.8819,
+      "step": 2582
+    },
+    {
+      "epoch": 0.496921893035783,
+      "grad_norm": 1.3087628428376397,
+      "learning_rate": 1.0585364732336587e-05,
+      "loss": 0.8847,
+      "step": 2583
+    },
+    {
+      "epoch": 0.4971142747210466,
+      "grad_norm": 1.596409175519866,
+      "learning_rate": 1.0579144457126365e-05,
+      "loss": 0.8573,
+      "step": 2584
+    },
+    {
+      "epoch": 0.4973066564063101,
+      "grad_norm": 1.1814659045304783,
+      "learning_rate": 1.057292395707232e-05,
+      "loss": 0.8506,
+      "step": 2585
+    },
+    {
+      "epoch": 0.4974990380915737,
+      "grad_norm": 1.381608952848615,
+      "learning_rate": 1.0566703234589471e-05,
+      "loss": 0.8636,
+      "step": 2586
+    },
+    {
+      "epoch": 0.49769141977683723,
+      "grad_norm": 0.6164284094376703,
+      "learning_rate": 1.0560482292092912e-05,
+      "loss": 0.3727,
+      "step": 2587
+    },
+    {
+      "epoch": 0.4978838014621008,
+      "grad_norm": 1.8956420029198238,
+      "learning_rate": 1.0554261131997833e-05,
+      "loss": 0.9143,
+      "step": 2588
+    },
+    {
+      "epoch": 0.49807618314736435,
+      "grad_norm": 1.3040005198740459,
+      "learning_rate": 1.0548039756719497e-05,
+      "loss": 0.8516,
+      "step": 2589
+    },
+    {
+      "epoch": 0.49826856483262794,
+      "grad_norm": 1.626411191645724,
+      "learning_rate": 1.054181816867326e-05,
+      "loss": 0.9353,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4984609465178915,
+      "grad_norm": 1.5325790966611672,
+      "learning_rate": 1.053559637027455e-05,
+      "loss": 0.9109,
+      "step": 2591
+    },
+    {
+      "epoch": 0.49865332820315506,
+      "grad_norm": 1.9735768696737304,
+      "learning_rate": 1.0529374363938888e-05,
+      "loss": 0.9398,
+      "step": 2592
+    },
+    {
+      "epoch": 0.49884570988841864,
+      "grad_norm": 1.5314390302273162,
+      "learning_rate": 1.0523152152081875e-05,
+      "loss": 0.8222,
+      "step": 2593
+    },
+    {
+      "epoch": 0.4990380915736822,
+      "grad_norm": 1.4230147650677496,
+      "learning_rate": 1.051692973711918e-05,
+      "loss": 0.8932,
+      "step": 2594
+    },
+    {
+      "epoch": 0.49923047325894576,
+      "grad_norm": 1.2838780253428501,
+      "learning_rate": 1.0510707121466568e-05,
+      "loss": 0.9056,
+      "step": 2595
+    },
+    {
+      "epoch": 0.4994228549442093,
+      "grad_norm": 0.6321346157639602,
+      "learning_rate": 1.0504484307539864e-05,
+      "loss": 0.3179,
+      "step": 2596
+    },
+    {
+      "epoch": 0.4996152366294729,
+      "grad_norm": 1.3337739847497507,
+      "learning_rate": 1.0498261297754984e-05,
+      "loss": 0.8829,
+      "step": 2597
+    },
+    {
+      "epoch": 0.4998076183147364,
+      "grad_norm": 1.6619519326126622,
+      "learning_rate": 1.0492038094527907e-05,
+      "loss": 0.8925,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5170877061530397,
+      "learning_rate": 1.0485814700274706e-05,
+      "loss": 0.8559,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5001923816852636,
+      "grad_norm": 1.4588364693558489,
+      "learning_rate": 1.047959111741151e-05,
+      "loss": 0.824,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5003847633705272,
+      "grad_norm": 1.2837806489758623,
+      "learning_rate": 1.0473367348354529e-05,
+      "loss": 0.9111,
+      "step": 2601
+    },
+    {
+      "epoch": 0.5005771450557907,
+      "grad_norm": 1.3755171621862714,
+      "learning_rate": 1.0467143395520044e-05,
+      "loss": 0.8963,
+      "step": 2602
+    },
+    {
+      "epoch": 0.5007695267410542,
+      "grad_norm": 1.2774633571093752,
+      "learning_rate": 1.046091926132441e-05,
+      "loss": 0.8726,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5009619084263178,
+      "grad_norm": 1.5217446446898228,
+      "learning_rate": 1.0454694948184045e-05,
+      "loss": 0.8679,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5011542901115814,
+      "grad_norm": 1.1642167527549572,
+      "learning_rate": 1.044847045851545e-05,
+      "loss": 0.8713,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5013466717968449,
+      "grad_norm": 1.2336757449186588,
+      "learning_rate": 1.044224579473518e-05,
+      "loss": 0.8949,
+      "step": 2606
+    },
+    {
+      "epoch": 0.5015390534821085,
+      "grad_norm": 1.4463740476778333,
+      "learning_rate": 1.0436020959259862e-05,
+      "loss": 0.8492,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5017314351673721,
+      "grad_norm": 2.2225408952428736,
+      "learning_rate": 1.0429795954506203e-05,
+      "loss": 0.8594,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5019238168526357,
+      "grad_norm": 2.3481593543651114,
+      "learning_rate": 1.0423570782890951e-05,
+      "loss": 0.8661,
+      "step": 2609
+    },
+    {
+      "epoch": 0.5021161985378992,
+      "grad_norm": 1.222867640352079,
+      "learning_rate": 1.0417345446830938e-05,
+      "loss": 0.8351,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5023085802231627,
+      "grad_norm": 1.4673178089745091,
+      "learning_rate": 1.0411119948743052e-05,
+      "loss": 0.8544,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5025009619084263,
+      "grad_norm": 1.4814853118815903,
+      "learning_rate": 1.0404894291044247e-05,
+      "loss": 0.8557,
+      "step": 2612
+    },
+    {
+      "epoch": 0.5026933435936899,
+      "grad_norm": 1.232886449434337,
+      "learning_rate": 1.0398668476151538e-05,
+      "loss": 0.8424,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5028857252789535,
+      "grad_norm": 2.0781913958478357,
+      "learning_rate": 1.0392442506482e-05,
+      "loss": 0.9465,
+      "step": 2614
+    },
+    {
+      "epoch": 0.503078106964217,
+      "grad_norm": 1.2929513140431155,
+      "learning_rate": 1.038621638445277e-05,
+      "loss": 0.8629,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5032704886494805,
+      "grad_norm": 1.1427674509486516,
+      "learning_rate": 1.037999011248104e-05,
+      "loss": 0.9244,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5034628703347441,
+      "grad_norm": 1.3524797637773511,
+      "learning_rate": 1.0373763692984062e-05,
+      "loss": 0.8836,
+      "step": 2617
+    },
+    {
+      "epoch": 0.5036552520200077,
+      "grad_norm": 1.2535259357843545,
+      "learning_rate": 1.0367537128379154e-05,
+      "loss": 0.8755,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5038476337052713,
+      "grad_norm": 1.3465501759941578,
+      "learning_rate": 1.0361310421083677e-05,
+      "loss": 0.852,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5040400153905348,
+      "grad_norm": 1.3599838320017203,
+      "learning_rate": 1.0355083573515052e-05,
+      "loss": 0.8969,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5042323970757984,
+      "grad_norm": 1.1680425700301542,
+      "learning_rate": 1.0348856588090764e-05,
+      "loss": 0.8762,
+      "step": 2621
+    },
+    {
+      "epoch": 0.504424778761062,
+      "grad_norm": 1.3212276135500711,
+      "learning_rate": 1.0342629467228331e-05,
+      "loss": 0.8987,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5046171604463255,
+      "grad_norm": 1.351068586222023,
+      "learning_rate": 1.0336402213345345e-05,
+      "loss": 0.835,
+      "step": 2623
+    },
+    {
+      "epoch": 0.504809542131589,
+      "grad_norm": 1.3613844581740915,
+      "learning_rate": 1.0330174828859434e-05,
+      "loss": 0.8669,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5050019238168526,
+      "grad_norm": 1.2691611258573525,
+      "learning_rate": 1.0323947316188288e-05,
+      "loss": 0.8569,
+      "step": 2625
+    },
+    {
+      "epoch": 0.5051943055021162,
+      "grad_norm": 1.1268940058182841,
+      "learning_rate": 1.031771967774964e-05,
+      "loss": 0.8415,
+      "step": 2626
+    },
+    {
+      "epoch": 0.5053866871873798,
+      "grad_norm": 1.528919546374715,
+      "learning_rate": 1.0311491915961271e-05,
+      "loss": 0.8957,
+      "step": 2627
+    },
+    {
+      "epoch": 0.5055790688726434,
+      "grad_norm": 1.8413180759855419,
+      "learning_rate": 1.030526403324102e-05,
+      "loss": 0.9128,
+      "step": 2628
+    },
+    {
+      "epoch": 0.5057714505579068,
+      "grad_norm": 2.42570273726718,
+      "learning_rate": 1.0299036032006759e-05,
+      "loss": 0.8698,
+      "step": 2629
+    },
+    {
+      "epoch": 0.5059638322431704,
+      "grad_norm": 1.356141586021958,
+      "learning_rate": 1.0292807914676412e-05,
+      "loss": 0.8485,
+      "step": 2630
+    },
+    {
+      "epoch": 0.506156213928434,
+      "grad_norm": 4.561564145918174,
+      "learning_rate": 1.0286579683667952e-05,
+      "loss": 0.8706,
+      "step": 2631
+    },
+    {
+      "epoch": 0.5063485956136976,
+      "grad_norm": 1.7699738967965772,
+      "learning_rate": 1.0280351341399392e-05,
+      "loss": 0.8783,
+      "step": 2632
+    },
+    {
+      "epoch": 0.5065409772989612,
+      "grad_norm": 1.1931651583407938,
+      "learning_rate": 1.027412289028879e-05,
+      "loss": 0.871,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5067333589842247,
+      "grad_norm": 1.8428659423394431,
+      "learning_rate": 1.0267894332754243e-05,
+      "loss": 0.9096,
+      "step": 2634
+    },
+    {
+      "epoch": 0.5069257406694883,
+      "grad_norm": 0.6266452330869582,
+      "learning_rate": 1.0261665671213891e-05,
+      "loss": 0.36,
+      "step": 2635
+    },
+    {
+      "epoch": 0.5071181223547518,
+      "grad_norm": 1.3241168207391,
+      "learning_rate": 1.0255436908085919e-05,
+      "loss": 0.8505,
+      "step": 2636
+    },
+    {
+      "epoch": 0.5073105040400154,
+      "grad_norm": 2.835619098633698,
+      "learning_rate": 1.024920804578854e-05,
+      "loss": 0.8748,
+      "step": 2637
+    },
+    {
+      "epoch": 0.5075028857252789,
+      "grad_norm": 0.6286729544446905,
+      "learning_rate": 1.0242979086740019e-05,
+      "loss": 0.3602,
+      "step": 2638
+    },
+    {
+      "epoch": 0.5076952674105425,
+      "grad_norm": 1.2726414705837976,
+      "learning_rate": 1.023675003335865e-05,
+      "loss": 0.8563,
+      "step": 2639
+    },
+    {
+      "epoch": 0.5078876490958061,
+      "grad_norm": 1.707819518947446,
+      "learning_rate": 1.0230520888062765e-05,
+      "loss": 0.8551,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5080800307810697,
+      "grad_norm": 0.6294253039846727,
+      "learning_rate": 1.0224291653270739e-05,
+      "loss": 0.3455,
+      "step": 2641
+    },
+    {
+      "epoch": 0.5082724124663333,
+      "grad_norm": 1.630618758641001,
+      "learning_rate": 1.0218062331400969e-05,
+      "loss": 0.9336,
+      "step": 2642
+    },
+    {
+      "epoch": 0.5084647941515967,
+      "grad_norm": 1.514699494338305,
+      "learning_rate": 1.0211832924871889e-05,
+      "loss": 0.8699,
+      "step": 2643
+    },
+    {
+      "epoch": 0.5086571758368603,
+      "grad_norm": 1.309245754570072,
+      "learning_rate": 1.0205603436101978e-05,
+      "loss": 0.8716,
+      "step": 2644
+    },
+    {
+      "epoch": 0.5088495575221239,
+      "grad_norm": 1.5118388649984946,
+      "learning_rate": 1.0199373867509734e-05,
+      "loss": 0.9305,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5090419392073875,
+      "grad_norm": 1.5227612098559367,
+      "learning_rate": 1.019314422151369e-05,
+      "loss": 0.8879,
+      "step": 2646
+    },
+    {
+      "epoch": 0.509234320892651,
+      "grad_norm": 1.4029759436566036,
+      "learning_rate": 1.0186914500532408e-05,
+      "loss": 0.897,
+      "step": 2647
+    },
+    {
+      "epoch": 0.5094267025779146,
+      "grad_norm": 2.110829651116134,
+      "learning_rate": 1.0180684706984483e-05,
+      "loss": 0.8889,
+      "step": 2648
+    },
+    {
+      "epoch": 0.5096190842631781,
+      "grad_norm": 1.3733322519869722,
+      "learning_rate": 1.0174454843288533e-05,
+      "loss": 0.8976,
+      "step": 2649
+    },
+    {
+      "epoch": 0.5098114659484417,
+      "grad_norm": 1.4859888044103238,
+      "learning_rate": 1.0168224911863205e-05,
+      "loss": 0.91,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5100038476337053,
+      "grad_norm": 1.77645710533982,
+      "learning_rate": 1.0161994915127173e-05,
+      "loss": 0.9001,
+      "step": 2651
+    },
+    {
+      "epoch": 0.5101962293189688,
+      "grad_norm": 1.5904852749213547,
+      "learning_rate": 1.015576485549914e-05,
+      "loss": 0.8297,
+      "step": 2652
+    },
+    {
+      "epoch": 0.5103886110042324,
+      "grad_norm": 1.7551193144553467,
+      "learning_rate": 1.0149534735397823e-05,
+      "loss": 0.9146,
+      "step": 2653
+    },
+    {
+      "epoch": 0.510580992689496,
+      "grad_norm": 1.2525904877705343,
+      "learning_rate": 1.0143304557241979e-05,
+      "loss": 0.868,
+      "step": 2654
+    },
+    {
+      "epoch": 0.5107733743747596,
+      "grad_norm": 2.934919583710128,
+      "learning_rate": 1.0137074323450372e-05,
+      "loss": 0.8778,
+      "step": 2655
+    },
+    {
+      "epoch": 0.510965756060023,
+      "grad_norm": 1.8497992308267326,
+      "learning_rate": 1.0130844036441787e-05,
+      "loss": 0.9264,
+      "step": 2656
+    },
+    {
+      "epoch": 0.5111581377452866,
+      "grad_norm": 1.3876877601387931,
+      "learning_rate": 1.0124613698635043e-05,
+      "loss": 0.8678,
+      "step": 2657
+    },
+    {
+      "epoch": 0.5113505194305502,
+      "grad_norm": 1.2909345621932296,
+      "learning_rate": 1.0118383312448973e-05,
+      "loss": 0.8667,
+      "step": 2658
+    },
+    {
+      "epoch": 0.5115429011158138,
+      "grad_norm": 1.1319755554929942,
+      "learning_rate": 1.0112152880302426e-05,
+      "loss": 0.9176,
+      "step": 2659
+    },
+    {
+      "epoch": 0.5117352828010774,
+      "grad_norm": 0.6858774047688313,
+      "learning_rate": 1.0105922404614265e-05,
+      "loss": 0.345,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5119276644863409,
+      "grad_norm": 2.57824470858322,
+      "learning_rate": 1.0099691887803385e-05,
+      "loss": 0.908,
+      "step": 2661
+    },
+    {
+      "epoch": 0.5121200461716044,
+      "grad_norm": 1.2982556418563131,
+      "learning_rate": 1.0093461332288678e-05,
+      "loss": 0.9096,
+      "step": 2662
+    },
+    {
+      "epoch": 0.512312427856868,
+      "grad_norm": 1.2132699095856532,
+      "learning_rate": 1.0087230740489065e-05,
+      "loss": 0.9302,
+      "step": 2663
+    },
+    {
+      "epoch": 0.5125048095421316,
+      "grad_norm": 0.6607842663911435,
+      "learning_rate": 1.0081000114823473e-05,
+      "loss": 0.3467,
+      "step": 2664
+    },
+    {
+      "epoch": 0.5126971912273951,
+      "grad_norm": 1.3728223232836207,
+      "learning_rate": 1.007476945771085e-05,
+      "loss": 0.8692,
+      "step": 2665
+    },
+    {
+      "epoch": 0.5128895729126587,
+      "grad_norm": 1.6360021654391743,
+      "learning_rate": 1.006853877157015e-05,
+      "loss": 0.8964,
+      "step": 2666
+    },
+    {
+      "epoch": 0.5130819545979223,
+      "grad_norm": 1.4156183633323751,
+      "learning_rate": 1.0062308058820337e-05,
+      "loss": 0.8862,
+      "step": 2667
+    },
+    {
+      "epoch": 0.5132743362831859,
+      "grad_norm": 1.6076170734934594,
+      "learning_rate": 1.0056077321880393e-05,
+      "loss": 0.9176,
+      "step": 2668
+    },
+    {
+      "epoch": 0.5134667179684494,
+      "grad_norm": 1.5260853044428495,
+      "learning_rate": 1.0049846563169297e-05,
+      "loss": 0.8822,
+      "step": 2669
+    },
+    {
+      "epoch": 0.5136590996537129,
+      "grad_norm": 1.3298420735493743,
+      "learning_rate": 1.0043615785106051e-05,
+      "loss": 0.856,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5138514813389765,
+      "grad_norm": 1.306097074381049,
+      "learning_rate": 1.0037384990109658e-05,
+      "loss": 0.8768,
+      "step": 2671
+    },
+    {
+      "epoch": 0.5140438630242401,
+      "grad_norm": 1.7664525584125081,
+      "learning_rate": 1.0031154180599123e-05,
+      "loss": 0.8106,
+      "step": 2672
+    },
+    {
+      "epoch": 0.5142362447095037,
+      "grad_norm": 1.259910883535816,
+      "learning_rate": 1.0024923358993458e-05,
+      "loss": 0.8727,
+      "step": 2673
+    },
+    {
+      "epoch": 0.5144286263947672,
+      "grad_norm": 1.5133013314909556,
+      "learning_rate": 1.0018692527711695e-05,
+      "loss": 0.882,
+      "step": 2674
+    },
+    {
+      "epoch": 0.5146210080800308,
+      "grad_norm": 6.440751519920302,
+      "learning_rate": 1.0012461689172846e-05,
+      "loss": 0.8578,
+      "step": 2675
+    },
+    {
+      "epoch": 0.5148133897652943,
+      "grad_norm": 1.655726853179403,
+      "learning_rate": 1.0006230845795937e-05,
+      "loss": 0.8201,
+      "step": 2676
+    },
+    {
+      "epoch": 0.5150057714505579,
+      "grad_norm": 1.682499255896431,
+      "learning_rate": 1e-05,
+      "loss": 0.8927,
+      "step": 2677
+    },
+    {
+      "epoch": 0.5151981531358215,
+      "grad_norm": 1.4318515304540704,
+      "learning_rate": 9.993769154204063e-06,
+      "loss": 0.8383,
+      "step": 2678
+    },
+    {
+      "epoch": 0.515390534821085,
+      "grad_norm": 1.249520061886214,
+      "learning_rate": 9.987538310827159e-06,
+      "loss": 0.8989,
+      "step": 2679
+    },
+    {
+      "epoch": 0.5155829165063486,
+      "grad_norm": 1.4908946062083388,
+      "learning_rate": 9.981307472288308e-06,
+      "loss": 0.8426,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5157752981916122,
+      "grad_norm": 3.6548003726022475,
+      "learning_rate": 9.975076641006542e-06,
+      "loss": 0.8958,
+      "step": 2681
+    },
+    {
+      "epoch": 0.5159676798768758,
+      "grad_norm": 1.3242628738219797,
+      "learning_rate": 9.968845819400883e-06,
+      "loss": 0.864,
+      "step": 2682
+    },
+    {
+      "epoch": 0.5161600615621393,
+      "grad_norm": 1.4050165726756125,
+      "learning_rate": 9.962615009890346e-06,
+      "loss": 0.8592,
+      "step": 2683
+    },
+    {
+      "epoch": 0.5163524432474028,
+      "grad_norm": 1.2383172398643885,
+      "learning_rate": 9.956384214893949e-06,
+      "loss": 0.9245,
+      "step": 2684
+    },
+    {
+      "epoch": 0.5165448249326664,
+      "grad_norm": 2.247027113488797,
+      "learning_rate": 9.950153436830707e-06,
+      "loss": 0.88,
+      "step": 2685
+    },
+    {
+      "epoch": 0.51673720661793,
+      "grad_norm": 1.5075995362473233,
+      "learning_rate": 9.94392267811961e-06,
+      "loss": 0.9364,
+      "step": 2686
+    },
+    {
+      "epoch": 0.5169295883031936,
+      "grad_norm": 1.5385713951132403,
+      "learning_rate": 9.937691941179665e-06,
+      "loss": 0.8828,
+      "step": 2687
+    },
+    {
+      "epoch": 0.517121969988457,
+      "grad_norm": 2.335667906112092,
+      "learning_rate": 9.931461228429856e-06,
+      "loss": 0.9103,
+      "step": 2688
+    },
+    {
+      "epoch": 0.5173143516737206,
+      "grad_norm": 1.280591795697595,
+      "learning_rate": 9.925230542289151e-06,
+      "loss": 0.8881,
+      "step": 2689
+    },
+    {
+      "epoch": 0.5175067333589842,
+      "grad_norm": 1.4558843097638992,
+      "learning_rate": 9.91899988517653e-06,
+      "loss": 0.8848,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5176991150442478,
+      "grad_norm": 0.6446567989224011,
+      "learning_rate": 9.912769259510938e-06,
+      "loss": 0.3529,
+      "step": 2691
+    },
+    {
+      "epoch": 0.5178914967295114,
+      "grad_norm": 1.6463633273007816,
+      "learning_rate": 9.906538667711324e-06,
+      "loss": 0.9023,
+      "step": 2692
+    },
+    {
+      "epoch": 0.5180838784147749,
+      "grad_norm": 1.488015891851259,
+      "learning_rate": 9.90030811219662e-06,
+      "loss": 0.8517,
+      "step": 2693
+    },
+    {
+      "epoch": 0.5182762601000385,
+      "grad_norm": 1.343764793590325,
+      "learning_rate": 9.894077595385736e-06,
+      "loss": 0.8824,
+      "step": 2694
+    },
+    {
+      "epoch": 0.518468641785302,
+      "grad_norm": 1.3702918175700591,
+      "learning_rate": 9.887847119697577e-06,
+      "loss": 0.8454,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5186610234705656,
+      "grad_norm": 1.1592254971839893,
+      "learning_rate": 9.881616687551032e-06,
+      "loss": 0.9299,
+      "step": 2696
+    },
+    {
+      "epoch": 0.5188534051558291,
+      "grad_norm": 1.6835923609707264,
+      "learning_rate": 9.875386301364958e-06,
+      "loss": 0.8585,
+      "step": 2697
+    },
+    {
+      "epoch": 0.5190457868410927,
+      "grad_norm": 1.6404852146855888,
+      "learning_rate": 9.869155963558215e-06,
+      "loss": 0.8018,
+      "step": 2698
+    },
+    {
+      "epoch": 0.5192381685263563,
+      "grad_norm": 1.206335972188918,
+      "learning_rate": 9.862925676549635e-06,
+      "loss": 0.8528,
+      "step": 2699
+    },
+    {
+      "epoch": 0.5194305502116199,
+      "grad_norm": 2.086822541853569,
+      "learning_rate": 9.856695442758023e-06,
+      "loss": 0.8903,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5196229318968835,
+      "grad_norm": 1.4575688738184034,
+      "learning_rate": 9.850465264602175e-06,
+      "loss": 0.8756,
+      "step": 2701
+    },
+    {
+      "epoch": 0.5198153135821469,
+      "grad_norm": 1.1794153378778243,
+      "learning_rate": 9.844235144500865e-06,
+      "loss": 0.817,
+      "step": 2702
+    },
+    {
+      "epoch": 0.5200076952674105,
+      "grad_norm": 1.308710328988097,
+      "learning_rate": 9.83800508487283e-06,
+      "loss": 0.821,
+      "step": 2703
+    },
+    {
+      "epoch": 0.5202000769526741,
+      "grad_norm": 1.4307726269182182,
+      "learning_rate": 9.831775088136797e-06,
+      "loss": 0.8457,
+      "step": 2704
+    },
+    {
+      "epoch": 0.5203924586379377,
+      "grad_norm": 1.3702168750732304,
+      "learning_rate": 9.82554515671147e-06,
+      "loss": 0.8354,
+      "step": 2705
+    },
+    {
+      "epoch": 0.5205848403232012,
+      "grad_norm": 1.4553682349232924,
+      "learning_rate": 9.819315293015519e-06,
+      "loss": 0.8648,
+      "step": 2706
+    },
+    {
+      "epoch": 0.5207772220084648,
+      "grad_norm": 1.2624578364405379,
+      "learning_rate": 9.813085499467594e-06,
+      "loss": 0.8743,
+      "step": 2707
+    },
+    {
+      "epoch": 0.5209696036937284,
+      "grad_norm": 1.1260797810959235,
+      "learning_rate": 9.806855778486314e-06,
+      "loss": 0.8958,
+      "step": 2708
+    },
+    {
+      "epoch": 0.5211619853789919,
+      "grad_norm": 1.3071598648858869,
+      "learning_rate": 9.800626132490268e-06,
+      "loss": 0.8944,
+      "step": 2709
+    },
+    {
+      "epoch": 0.5213543670642555,
+      "grad_norm": 1.6354334356828963,
+      "learning_rate": 9.794396563898022e-06,
+      "loss": 0.8255,
+      "step": 2710
+    },
+    {
+      "epoch": 0.521546748749519,
+      "grad_norm": 1.3402771992311773,
+      "learning_rate": 9.788167075128113e-06,
+      "loss": 0.8206,
+      "step": 2711
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 1.4668532458112278,
+      "learning_rate": 9.781937668599035e-06,
+      "loss": 0.8145,
+      "step": 2712
+    },
+    {
+      "epoch": 0.5219315121200462,
+      "grad_norm": 1.4022544270354216,
+      "learning_rate": 9.775708346729263e-06,
+      "loss": 0.8416,
+      "step": 2713
+    },
+    {
+      "epoch": 0.5221238938053098,
+      "grad_norm": 1.2213560568759454,
+      "learning_rate": 9.769479111937238e-06,
+      "loss": 0.8379,
+      "step": 2714
+    },
+    {
+      "epoch": 0.5223162754905732,
+      "grad_norm": 0.6884960380185703,
+      "learning_rate": 9.763249966641352e-06,
+      "loss": 0.3399,
+      "step": 2715
+    },
+    {
+      "epoch": 0.5225086571758368,
+      "grad_norm": 1.5019079263160395,
+      "learning_rate": 9.757020913259986e-06,
+      "loss": 0.9173,
+      "step": 2716
+    },
+    {
+      "epoch": 0.5227010388611004,
+      "grad_norm": 1.353314585722105,
+      "learning_rate": 9.750791954211464e-06,
+      "loss": 0.9023,
+      "step": 2717
+    },
+    {
+      "epoch": 0.522893420546364,
+      "grad_norm": 1.3084838359298112,
+      "learning_rate": 9.744563091914085e-06,
+      "loss": 0.8625,
+      "step": 2718
+    },
+    {
+      "epoch": 0.5230858022316276,
+      "grad_norm": 1.3383196863119542,
+      "learning_rate": 9.738334328786114e-06,
+      "loss": 0.8909,
+      "step": 2719
+    },
+    {
+      "epoch": 0.5232781839168911,
+      "grad_norm": 1.1536949067942723,
+      "learning_rate": 9.732105667245759e-06,
+      "loss": 0.8924,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5234705656021547,
+      "grad_norm": 1.1537567779211928,
+      "learning_rate": 9.725877109711212e-06,
+      "loss": 0.9189,
+      "step": 2721
+    },
+    {
+      "epoch": 0.5236629472874182,
+      "grad_norm": 1.54159954283006,
+      "learning_rate": 9.719648658600611e-06,
+      "loss": 0.9037,
+      "step": 2722
+    },
+    {
+      "epoch": 0.5238553289726818,
+      "grad_norm": 1.2619354776009915,
+      "learning_rate": 9.71342031633205e-06,
+      "loss": 0.8531,
+      "step": 2723
+    },
+    {
+      "epoch": 0.5240477106579453,
+      "grad_norm": 1.317562655004245,
+      "learning_rate": 9.70719208532359e-06,
+      "loss": 0.8949,
+      "step": 2724
+    },
+    {
+      "epoch": 0.5242400923432089,
+      "grad_norm": 1.2710887026050308,
+      "learning_rate": 9.700963967993246e-06,
+      "loss": 0.8442,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5244324740284725,
+      "grad_norm": 1.348393065721359,
+      "learning_rate": 9.694735966758982e-06,
+      "loss": 0.9233,
+      "step": 2726
+    },
+    {
+      "epoch": 0.5246248557137361,
+      "grad_norm": 1.2308762772982016,
+      "learning_rate": 9.688508084038729e-06,
+      "loss": 0.8199,
+      "step": 2727
+    },
+    {
+      "epoch": 0.5248172373989997,
+      "grad_norm": 1.4733874661694297,
+      "learning_rate": 9.682280322250365e-06,
+      "loss": 0.8339,
+      "step": 2728
+    },
+    {
+      "epoch": 0.5250096190842631,
+      "grad_norm": 1.2249992116065938,
+      "learning_rate": 9.676052683811715e-06,
+      "loss": 0.8581,
+      "step": 2729
+    },
+    {
+      "epoch": 0.5252020007695267,
+      "grad_norm": 1.231379096172116,
+      "learning_rate": 9.669825171140568e-06,
+      "loss": 0.8771,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5253943824547903,
+      "grad_norm": 1.1599760451178356,
+      "learning_rate": 9.66359778665466e-06,
+      "loss": 0.8301,
+      "step": 2731
+    },
+    {
+      "epoch": 0.5255867641400539,
+      "grad_norm": 1.2431782027280678,
+      "learning_rate": 9.657370532771672e-06,
+      "loss": 0.8549,
+      "step": 2732
+    },
+    {
+      "epoch": 0.5257791458253175,
+      "grad_norm": 1.2312127443193,
+      "learning_rate": 9.651143411909241e-06,
+      "loss": 0.8572,
+      "step": 2733
+    },
+    {
+      "epoch": 0.525971527510581,
+      "grad_norm": 1.1872674448351237,
+      "learning_rate": 9.64491642648495e-06,
+      "loss": 0.9001,
+      "step": 2734
+    },
+    {
+      "epoch": 0.5261639091958445,
+      "grad_norm": 1.0451000510724977,
+      "learning_rate": 9.638689578916326e-06,
+      "loss": 0.885,
+      "step": 2735
+    },
+    {
+      "epoch": 0.5263562908811081,
+      "grad_norm": 1.2756944616638124,
+      "learning_rate": 9.632462871620847e-06,
+      "loss": 0.9121,
+      "step": 2736
+    },
+    {
+      "epoch": 0.5265486725663717,
+      "grad_norm": 1.4515805044474936,
+      "learning_rate": 9.62623630701594e-06,
+      "loss": 0.9026,
+      "step": 2737
+    },
+    {
+      "epoch": 0.5267410542516352,
+      "grad_norm": 0.6716636939226714,
+      "learning_rate": 9.620009887518963e-06,
+      "loss": 0.3325,
+      "step": 2738
+    },
+    {
+      "epoch": 0.5269334359368988,
+      "grad_norm": 1.2416146090248166,
+      "learning_rate": 9.613783615547233e-06,
+      "loss": 0.8741,
+      "step": 2739
+    },
+    {
+      "epoch": 0.5271258176221624,
+      "grad_norm": 1.4072061830968232,
+      "learning_rate": 9.607557493518006e-06,
+      "loss": 0.8368,
+      "step": 2740
+    },
+    {
+      "epoch": 0.527318199307426,
+      "grad_norm": 1.2190386195950524,
+      "learning_rate": 9.601331523848464e-06,
+      "loss": 0.8352,
+      "step": 2741
+    },
+    {
+      "epoch": 0.5275105809926895,
+      "grad_norm": 1.2859672706135423,
+      "learning_rate": 9.595105708955758e-06,
+      "loss": 0.8362,
+      "step": 2742
+    },
+    {
+      "epoch": 0.527702962677953,
+      "grad_norm": 1.4129299767749706,
+      "learning_rate": 9.588880051256951e-06,
+      "loss": 0.9068,
+      "step": 2743
+    },
+    {
+      "epoch": 0.5278953443632166,
+      "grad_norm": 1.358260364085071,
+      "learning_rate": 9.582654553169064e-06,
+      "loss": 0.8333,
+      "step": 2744
+    },
+    {
+      "epoch": 0.5280877260484802,
+      "grad_norm": 1.2112907850527779,
+      "learning_rate": 9.576429217109054e-06,
+      "loss": 0.8638,
+      "step": 2745
+    },
+    {
+      "epoch": 0.5282801077337438,
+      "grad_norm": 0.6301019068463461,
+      "learning_rate": 9.5702040454938e-06,
+      "loss": 0.3284,
+      "step": 2746
+    },
+    {
+      "epoch": 0.5284724894190073,
+      "grad_norm": 1.5285145415272097,
+      "learning_rate": 9.563979040740138e-06,
+      "loss": 0.8317,
+      "step": 2747
+    },
+    {
+      "epoch": 0.5286648711042709,
+      "grad_norm": 0.6880350537630698,
+      "learning_rate": 9.557754205264826e-06,
+      "loss": 0.3441,
+      "step": 2748
+    },
+    {
+      "epoch": 0.5288572527895344,
+      "grad_norm": 1.273101713651106,
+      "learning_rate": 9.551529541484554e-06,
+      "loss": 0.8757,
+      "step": 2749
+    },
+    {
+      "epoch": 0.529049634474798,
+      "grad_norm": 1.336843180277952,
+      "learning_rate": 9.545305051815957e-06,
+      "loss": 0.9369,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5292420161600616,
+      "grad_norm": 1.3630679758704691,
+      "learning_rate": 9.539080738675597e-06,
+      "loss": 0.9349,
+      "step": 2751
+    },
+    {
+      "epoch": 0.5294343978453251,
+      "grad_norm": 1.2690762226740693,
+      "learning_rate": 9.53285660447996e-06,
+      "loss": 0.9031,
+      "step": 2752
+    },
+    {
+      "epoch": 0.5296267795305887,
+      "grad_norm": 1.2229019280363862,
+      "learning_rate": 9.526632651645476e-06,
+      "loss": 0.8327,
+      "step": 2753
+    },
+    {
+      "epoch": 0.5298191612158523,
+      "grad_norm": 1.274881869197217,
+      "learning_rate": 9.520408882588497e-06,
+      "loss": 0.9037,
+      "step": 2754
+    },
+    {
+      "epoch": 0.5300115429011159,
+      "grad_norm": 1.3769175751884437,
+      "learning_rate": 9.514185299725299e-06,
+      "loss": 0.8996,
+      "step": 2755
+    },
+    {
+      "epoch": 0.5302039245863793,
+      "grad_norm": 1.0961969886420555,
+      "learning_rate": 9.507961905472093e-06,
+      "loss": 0.8357,
+      "step": 2756
+    },
+    {
+      "epoch": 0.5303963062716429,
+      "grad_norm": 1.2760071023822142,
+      "learning_rate": 9.501738702245023e-06,
+      "loss": 0.8607,
+      "step": 2757
+    },
+    {
+      "epoch": 0.5305886879569065,
+      "grad_norm": 1.5244846124938556,
+      "learning_rate": 9.495515692460138e-06,
+      "loss": 0.858,
+      "step": 2758
+    },
+    {
+      "epoch": 0.5307810696421701,
+      "grad_norm": 1.2354090771093513,
+      "learning_rate": 9.489292878533436e-06,
+      "loss": 0.8473,
+      "step": 2759
+    },
+    {
+      "epoch": 0.5309734513274337,
+      "grad_norm": 1.3827942871323715,
+      "learning_rate": 9.483070262880823e-06,
+      "loss": 0.9522,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5311658330126972,
+      "grad_norm": 1.3220558925476285,
+      "learning_rate": 9.476847847918126e-06,
+      "loss": 0.9159,
+      "step": 2761
+    },
+    {
+      "epoch": 0.5313582146979607,
+      "grad_norm": 1.2925416400711878,
+      "learning_rate": 9.47062563606111e-06,
+      "loss": 0.9228,
+      "step": 2762
+    },
+    {
+      "epoch": 0.5315505963832243,
+      "grad_norm": 1.3874193449845207,
+      "learning_rate": 9.464403629725454e-06,
+      "loss": 0.827,
+      "step": 2763
+    },
+    {
+      "epoch": 0.5317429780684879,
+      "grad_norm": 1.1822302401054785,
+      "learning_rate": 9.458181831326744e-06,
+      "loss": 0.8553,
+      "step": 2764
+    },
+    {
+      "epoch": 0.5319353597537514,
+      "grad_norm": 1.3737318519011157,
+      "learning_rate": 9.451960243280506e-06,
+      "loss": 0.8706,
+      "step": 2765
+    },
+    {
+      "epoch": 0.532127741439015,
+      "grad_norm": 1.3579754509887436,
+      "learning_rate": 9.44573886800217e-06,
+      "loss": 0.8758,
+      "step": 2766
+    },
+    {
+      "epoch": 0.5323201231242786,
+      "grad_norm": 1.3221179654143604,
+      "learning_rate": 9.43951770790709e-06,
+      "loss": 0.8977,
+      "step": 2767
+    },
+    {
+      "epoch": 0.5325125048095422,
+      "grad_norm": 1.595009950953803,
+      "learning_rate": 9.433296765410534e-06,
+      "loss": 0.9147,
+      "step": 2768
+    },
+    {
+      "epoch": 0.5327048864948057,
+      "grad_norm": 0.6269665068936161,
+      "learning_rate": 9.427076042927683e-06,
+      "loss": 0.3352,
+      "step": 2769
+    },
+    {
+      "epoch": 0.5328972681800692,
+      "grad_norm": 1.2734060468096904,
+      "learning_rate": 9.420855542873638e-06,
+      "loss": 0.8754,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5330896498653328,
+      "grad_norm": 1.1547138978862235,
+      "learning_rate": 9.414635267663416e-06,
+      "loss": 0.8506,
+      "step": 2771
+    },
+    {
+      "epoch": 0.5332820315505964,
+      "grad_norm": 1.2578851765745198,
+      "learning_rate": 9.408415219711934e-06,
+      "loss": 0.8623,
+      "step": 2772
+    },
+    {
+      "epoch": 0.53347441323586,
+      "grad_norm": 1.3603150222544549,
+      "learning_rate": 9.402195401434036e-06,
+      "loss": 0.8121,
+      "step": 2773
+    },
+    {
+      "epoch": 0.5336667949211235,
+      "grad_norm": 1.3613192962936527,
+      "learning_rate": 9.395975815244468e-06,
+      "loss": 0.8408,
+      "step": 2774
+    },
+    {
+      "epoch": 0.533859176606387,
+      "grad_norm": 1.367361183309026,
+      "learning_rate": 9.389756463557878e-06,
+      "loss": 0.8689,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5340515582916506,
+      "grad_norm": 1.2506939788556952,
+      "learning_rate": 9.383537348788844e-06,
+      "loss": 0.8788,
+      "step": 2776
+    },
+    {
+      "epoch": 0.5342439399769142,
+      "grad_norm": 0.706568471121143,
+      "learning_rate": 9.377318473351838e-06,
+      "loss": 0.3601,
+      "step": 2777
+    },
+    {
+      "epoch": 0.5344363216621778,
+      "grad_norm": 0.6314399909819033,
+      "learning_rate": 9.371099839661238e-06,
+      "loss": 0.3779,
+      "step": 2778
+    },
+    {
+      "epoch": 0.5346287033474413,
+      "grad_norm": 1.2371670373425883,
+      "learning_rate": 9.364881450131335e-06,
+      "loss": 0.8474,
+      "step": 2779
+    },
+    {
+      "epoch": 0.5348210850327049,
+      "grad_norm": 1.2966480648939611,
+      "learning_rate": 9.358663307176323e-06,
+      "loss": 0.8481,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5350134667179685,
+      "grad_norm": 1.2523952066450494,
+      "learning_rate": 9.352445413210294e-06,
+      "loss": 0.8805,
+      "step": 2781
+    },
+    {
+      "epoch": 0.535205848403232,
+      "grad_norm": 1.2346887759385623,
+      "learning_rate": 9.346227770647251e-06,
+      "loss": 0.8892,
+      "step": 2782
+    },
+    {
+      "epoch": 0.5353982300884956,
+      "grad_norm": 1.7408697783679918,
+      "learning_rate": 9.3400103819011e-06,
+      "loss": 0.8279,
+      "step": 2783
+    },
+    {
+      "epoch": 0.5355906117737591,
+      "grad_norm": 1.3770496305056203,
+      "learning_rate": 9.33379324938564e-06,
+      "loss": 0.8916,
+      "step": 2784
+    },
+    {
+      "epoch": 0.5357829934590227,
+      "grad_norm": 1.2756630395446542,
+      "learning_rate": 9.327576375514582e-06,
+      "loss": 0.8974,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5359753751442863,
+      "grad_norm": 1.1364964547382876,
+      "learning_rate": 9.321359762701527e-06,
+      "loss": 0.8606,
+      "step": 2786
+    },
+    {
+      "epoch": 0.5361677568295499,
+      "grad_norm": 0.640516898785129,
+      "learning_rate": 9.315143413359975e-06,
+      "loss": 0.3672,
+      "step": 2787
+    },
+    {
+      "epoch": 0.5363601385148133,
+      "grad_norm": 1.2480561125710747,
+      "learning_rate": 9.308927329903333e-06,
+      "loss": 0.8147,
+      "step": 2788
+    },
+    {
+      "epoch": 0.5365525202000769,
+      "grad_norm": 1.4570206123344802,
+      "learning_rate": 9.302711514744897e-06,
+      "loss": 0.854,
+      "step": 2789
+    },
+    {
+      "epoch": 0.5367449018853405,
+      "grad_norm": 1.1552551397163315,
+      "learning_rate": 9.296495970297855e-06,
+      "loss": 0.8977,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5369372835706041,
+      "grad_norm": 1.3256320347470534,
+      "learning_rate": 9.290280698975307e-06,
+      "loss": 0.8204,
+      "step": 2791
+    },
+    {
+      "epoch": 0.5371296652558677,
+      "grad_norm": 1.245194989855332,
+      "learning_rate": 9.284065703190221e-06,
+      "loss": 0.8912,
+      "step": 2792
+    },
+    {
+      "epoch": 0.5373220469411312,
+      "grad_norm": 1.1758743540119012,
+      "learning_rate": 9.27785098535548e-06,
+      "loss": 0.8935,
+      "step": 2793
+    },
+    {
+      "epoch": 0.5375144286263948,
+      "grad_norm": 1.271157717670915,
+      "learning_rate": 9.271636547883856e-06,
+      "loss": 0.8203,
+      "step": 2794
+    },
+    {
+      "epoch": 0.5377068103116583,
+      "grad_norm": 1.7485899548961068,
+      "learning_rate": 9.265422393187998e-06,
+      "loss": 0.8805,
+      "step": 2795
+    },
+    {
+      "epoch": 0.5378991919969219,
+      "grad_norm": 1.2341990101816098,
+      "learning_rate": 9.259208523680457e-06,
+      "loss": 0.8741,
+      "step": 2796
+    },
+    {
+      "epoch": 0.5380915736821854,
+      "grad_norm": 1.1496509379052613,
+      "learning_rate": 9.252994941773679e-06,
+      "loss": 0.8042,
+      "step": 2797
+    },
+    {
+      "epoch": 0.538283955367449,
+      "grad_norm": 1.4649865153269455,
+      "learning_rate": 9.24678164987998e-06,
+      "loss": 0.86,
+      "step": 2798
+    },
+    {
+      "epoch": 0.5384763370527126,
+      "grad_norm": 1.1889886187858387,
+      "learning_rate": 9.24056865041158e-06,
+      "loss": 0.8109,
+      "step": 2799
+    },
+    {
+      "epoch": 0.5386687187379762,
+      "grad_norm": 1.319401357395712,
+      "learning_rate": 9.234355945780581e-06,
+      "loss": 0.8609,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5388611004232398,
+      "grad_norm": 1.7550594800805002,
+      "learning_rate": 9.228143538398963e-06,
+      "loss": 0.9373,
+      "step": 2801
+    },
+    {
+      "epoch": 0.5390534821085032,
+      "grad_norm": 1.3052562758464195,
+      "learning_rate": 9.221931430678598e-06,
+      "loss": 0.8782,
+      "step": 2802
+    },
+    {
+      "epoch": 0.5392458637937668,
+      "grad_norm": 1.347894369991165,
+      "learning_rate": 9.215719625031245e-06,
+      "loss": 0.9269,
+      "step": 2803
+    },
+    {
+      "epoch": 0.5394382454790304,
+      "grad_norm": 1.4538730130129853,
+      "learning_rate": 9.209508123868534e-06,
+      "loss": 0.9111,
+      "step": 2804
+    },
+    {
+      "epoch": 0.539630627164294,
+      "grad_norm": 1.2204294866276684,
+      "learning_rate": 9.203296929601986e-06,
+      "loss": 0.8209,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5398230088495575,
+      "grad_norm": 1.3198458201969372,
+      "learning_rate": 9.197086044643004e-06,
+      "loss": 0.8234,
+      "step": 2806
+    },
+    {
+      "epoch": 0.5400153905348211,
+      "grad_norm": 1.2196846560558012,
+      "learning_rate": 9.190875471402865e-06,
+      "loss": 0.8441,
+      "step": 2807
+    },
+    {
+      "epoch": 0.5402077722200846,
+      "grad_norm": 5.43514561971885,
+      "learning_rate": 9.184665212292723e-06,
+      "loss": 0.8636,
+      "step": 2808
+    },
+    {
+      "epoch": 0.5404001539053482,
+      "grad_norm": 1.2469684016747669,
+      "learning_rate": 9.178455269723623e-06,
+      "loss": 0.8424,
+      "step": 2809
+    },
+    {
+      "epoch": 0.5405925355906118,
+      "grad_norm": 0.6319538526697849,
+      "learning_rate": 9.172245646106471e-06,
+      "loss": 0.347,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5407849172758753,
+      "grad_norm": 1.541244103553251,
+      "learning_rate": 9.166036343852061e-06,
+      "loss": 0.8381,
+      "step": 2811
+    },
+    {
+      "epoch": 0.5409772989611389,
+      "grad_norm": 0.6362169763380151,
+      "learning_rate": 9.159827365371055e-06,
+      "loss": 0.3665,
+      "step": 2812
+    },
+    {
+      "epoch": 0.5411696806464025,
+      "grad_norm": 1.3496433104394072,
+      "learning_rate": 9.153618713073995e-06,
+      "loss": 0.9162,
+      "step": 2813
+    },
+    {
+      "epoch": 0.5413620623316661,
+      "grad_norm": 1.3324065423596128,
+      "learning_rate": 9.14741038937129e-06,
+      "loss": 0.8783,
+      "step": 2814
+    },
+    {
+      "epoch": 0.5415544440169295,
+      "grad_norm": 1.422758178172481,
+      "learning_rate": 9.141202396673232e-06,
+      "loss": 0.8898,
+      "step": 2815
+    },
+    {
+      "epoch": 0.5417468257021931,
+      "grad_norm": 1.8537831790946424,
+      "learning_rate": 9.13499473738997e-06,
+      "loss": 0.8975,
+      "step": 2816
+    },
+    {
+      "epoch": 0.5419392073874567,
+      "grad_norm": 1.2700902581893279,
+      "learning_rate": 9.128787413931536e-06,
+      "loss": 0.8896,
+      "step": 2817
+    },
+    {
+      "epoch": 0.5421315890727203,
+      "grad_norm": 1.8000005185487298,
+      "learning_rate": 9.122580428707822e-06,
+      "loss": 0.863,
+      "step": 2818
+    },
+    {
+      "epoch": 0.5423239707579839,
+      "grad_norm": 1.508834491514525,
+      "learning_rate": 9.116373784128597e-06,
+      "loss": 0.843,
+      "step": 2819
+    },
+    {
+      "epoch": 0.5425163524432474,
+      "grad_norm": 1.4190058974172934,
+      "learning_rate": 9.110167482603494e-06,
+      "loss": 0.8286,
+      "step": 2820
+    },
+    {
+      "epoch": 0.542708734128511,
+      "grad_norm": 1.9313599573750289,
+      "learning_rate": 9.10396152654201e-06,
+      "loss": 0.8719,
+      "step": 2821
+    },
+    {
+      "epoch": 0.5429011158137745,
+      "grad_norm": 1.2452277472262179,
+      "learning_rate": 9.097755918353513e-06,
+      "loss": 0.9022,
+      "step": 2822
+    },
+    {
+      "epoch": 0.5430934974990381,
+      "grad_norm": 1.4192644192146542,
+      "learning_rate": 9.091550660447236e-06,
+      "loss": 0.8906,
+      "step": 2823
+    },
+    {
+      "epoch": 0.5432858791843016,
+      "grad_norm": 1.4563341520887214,
+      "learning_rate": 9.08534575523227e-06,
+      "loss": 0.8742,
+      "step": 2824
+    },
+    {
+      "epoch": 0.5434782608695652,
+      "grad_norm": 1.3524281419034996,
+      "learning_rate": 9.079141205117573e-06,
+      "loss": 0.8644,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5436706425548288,
+      "grad_norm": 1.3306682566832737,
+      "learning_rate": 9.072937012511968e-06,
+      "loss": 0.8442,
+      "step": 2826
+    },
+    {
+      "epoch": 0.5438630242400924,
+      "grad_norm": 1.5911461566262735,
+      "learning_rate": 9.066733179824134e-06,
+      "loss": 0.9274,
+      "step": 2827
+    },
+    {
+      "epoch": 0.544055405925356,
+      "grad_norm": 1.316200306819621,
+      "learning_rate": 9.060529709462613e-06,
+      "loss": 0.771,
+      "step": 2828
+    },
+    {
+      "epoch": 0.5442477876106194,
+      "grad_norm": 1.4367702404053029,
+      "learning_rate": 9.054326603835807e-06,
+      "loss": 0.9021,
+      "step": 2829
+    },
+    {
+      "epoch": 0.544440169295883,
+      "grad_norm": 1.5825982633862032,
+      "learning_rate": 9.048123865351971e-06,
+      "loss": 0.9155,
+      "step": 2830
+    },
+    {
+      "epoch": 0.5446325509811466,
+      "grad_norm": 1.9784239112343145,
+      "learning_rate": 9.041921496419225e-06,
+      "loss": 0.9178,
+      "step": 2831
+    },
+    {
+      "epoch": 0.5448249326664102,
+      "grad_norm": 1.3818819844051817,
+      "learning_rate": 9.035719499445545e-06,
+      "loss": 0.8511,
+      "step": 2832
+    },
+    {
+      "epoch": 0.5450173143516738,
+      "grad_norm": 1.3822364000484708,
+      "learning_rate": 9.029517876838755e-06,
+      "loss": 0.8322,
+      "step": 2833
+    },
+    {
+      "epoch": 0.5452096960369373,
+      "grad_norm": 0.6625929648898058,
+      "learning_rate": 9.023316631006536e-06,
+      "loss": 0.3372,
+      "step": 2834
+    },
+    {
+      "epoch": 0.5454020777222008,
+      "grad_norm": 1.2330941410899956,
+      "learning_rate": 9.017115764356436e-06,
+      "loss": 0.8703,
+      "step": 2835
+    },
+    {
+      "epoch": 0.5455944594074644,
+      "grad_norm": 0.6139386600663134,
+      "learning_rate": 9.010915279295833e-06,
+      "loss": 0.3507,
+      "step": 2836
+    },
+    {
+      "epoch": 0.545786841092728,
+      "grad_norm": 1.1522348338296327,
+      "learning_rate": 9.004715178231975e-06,
+      "loss": 0.8312,
+      "step": 2837
+    },
+    {
+      "epoch": 0.5459792227779915,
+      "grad_norm": 0.6557242542599349,
+      "learning_rate": 8.998515463571953e-06,
+      "loss": 0.3314,
+      "step": 2838
+    },
+    {
+      "epoch": 0.5461716044632551,
+      "grad_norm": 1.314653572236811,
+      "learning_rate": 8.992316137722711e-06,
+      "loss": 0.8121,
+      "step": 2839
+    },
+    {
+      "epoch": 0.5463639861485187,
+      "grad_norm": 1.5692752399167689,
+      "learning_rate": 8.986117203091042e-06,
+      "loss": 0.8546,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5465563678337823,
+      "grad_norm": 1.3139094229467596,
+      "learning_rate": 8.97991866208358e-06,
+      "loss": 0.9064,
+      "step": 2841
+    },
+    {
+      "epoch": 0.5467487495190458,
+      "grad_norm": 1.291469090651672,
+      "learning_rate": 8.973720517106814e-06,
+      "loss": 0.8214,
+      "step": 2842
+    },
+    {
+      "epoch": 0.5469411312043093,
+      "grad_norm": 1.3386940497284199,
+      "learning_rate": 8.967522770567086e-06,
+      "loss": 0.8927,
+      "step": 2843
+    },
+    {
+      "epoch": 0.5471335128895729,
+      "grad_norm": 1.2307945890069785,
+      "learning_rate": 8.961325424870561e-06,
+      "loss": 0.8604,
+      "step": 2844
+    },
+    {
+      "epoch": 0.5473258945748365,
+      "grad_norm": 1.2477298143335669,
+      "learning_rate": 8.955128482423271e-06,
+      "loss": 0.9184,
+      "step": 2845
+    },
+    {
+      "epoch": 0.5475182762601001,
+      "grad_norm": 1.3183564952331945,
+      "learning_rate": 8.948931945631082e-06,
+      "loss": 0.9049,
+      "step": 2846
+    },
+    {
+      "epoch": 0.5477106579453636,
+      "grad_norm": 2.1922504571368173,
+      "learning_rate": 8.9427358168997e-06,
+      "loss": 0.8851,
+      "step": 2847
+    },
+    {
+      "epoch": 0.5479030396306271,
+      "grad_norm": 1.1753232274288934,
+      "learning_rate": 8.936540098634675e-06,
+      "loss": 0.8243,
+      "step": 2848
+    },
+    {
+      "epoch": 0.5480954213158907,
+      "grad_norm": 1.4000716899026722,
+      "learning_rate": 8.930344793241404e-06,
+      "loss": 0.8455,
+      "step": 2849
+    },
+    {
+      "epoch": 0.5482878030011543,
+      "grad_norm": 1.2936945833799096,
+      "learning_rate": 8.924149903125108e-06,
+      "loss": 0.8144,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5484801846864179,
+      "grad_norm": 1.429526450275071,
+      "learning_rate": 8.917955430690865e-06,
+      "loss": 0.8824,
+      "step": 2851
+    },
+    {
+      "epoch": 0.5486725663716814,
+      "grad_norm": 1.2477463734707834,
+      "learning_rate": 8.91176137834358e-06,
+      "loss": 0.8493,
+      "step": 2852
+    },
+    {
+      "epoch": 0.548864948056945,
+      "grad_norm": 1.0545244295265357,
+      "learning_rate": 8.905567748487997e-06,
+      "loss": 0.8271,
+      "step": 2853
+    },
+    {
+      "epoch": 0.5490573297422086,
+      "grad_norm": 1.3425061182370295,
+      "learning_rate": 8.899374543528695e-06,
+      "loss": 0.8766,
+      "step": 2854
+    },
+    {
+      "epoch": 0.5492497114274721,
+      "grad_norm": 0.6310722169385636,
+      "learning_rate": 8.893181765870094e-06,
+      "loss": 0.3646,
+      "step": 2855
+    },
+    {
+      "epoch": 0.5494420931127356,
+      "grad_norm": 1.2819421089567615,
+      "learning_rate": 8.886989417916435e-06,
+      "loss": 0.8317,
+      "step": 2856
+    },
+    {
+      "epoch": 0.5496344747979992,
+      "grad_norm": 1.1602193708026158,
+      "learning_rate": 8.88079750207181e-06,
+      "loss": 0.8297,
+      "step": 2857
+    },
+    {
+      "epoch": 0.5498268564832628,
+      "grad_norm": 1.3140189431691713,
+      "learning_rate": 8.87460602074013e-06,
+      "loss": 0.8688,
+      "step": 2858
+    },
+    {
+      "epoch": 0.5500192381685264,
+      "grad_norm": 1.2225771877100533,
+      "learning_rate": 8.86841497632514e-06,
+      "loss": 0.7964,
+      "step": 2859
+    },
+    {
+      "epoch": 0.55021161985379,
+      "grad_norm": 1.2410813607142441,
+      "learning_rate": 8.862224371230418e-06,
+      "loss": 0.8392,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5504040015390534,
+      "grad_norm": 1.1819482398432561,
+      "learning_rate": 8.85603420785937e-06,
+      "loss": 0.8195,
+      "step": 2861
+    },
+    {
+      "epoch": 0.550596383224317,
+      "grad_norm": 1.4059095047843715,
+      "learning_rate": 8.84984448861523e-06,
+      "loss": 0.907,
+      "step": 2862
+    },
+    {
+      "epoch": 0.5507887649095806,
+      "grad_norm": 1.4198807266380762,
+      "learning_rate": 8.84365521590106e-06,
+      "loss": 0.8412,
+      "step": 2863
+    },
+    {
+      "epoch": 0.5509811465948442,
+      "grad_norm": 1.2878482190099485,
+      "learning_rate": 8.837466392119752e-06,
+      "loss": 0.8001,
+      "step": 2864
+    },
+    {
+      "epoch": 0.5511735282801077,
+      "grad_norm": 1.0687353487087796,
+      "learning_rate": 8.831278019674017e-06,
+      "loss": 0.8259,
+      "step": 2865
+    },
+    {
+      "epoch": 0.5513659099653713,
+      "grad_norm": 1.217950180596804,
+      "learning_rate": 8.825090100966396e-06,
+      "loss": 0.8401,
+      "step": 2866
+    },
+    {
+      "epoch": 0.5515582916506349,
+      "grad_norm": 1.5136390702717848,
+      "learning_rate": 8.818902638399247e-06,
+      "loss": 0.8647,
+      "step": 2867
+    },
+    {
+      "epoch": 0.5517506733358984,
+      "grad_norm": 1.3154361691665257,
+      "learning_rate": 8.81271563437476e-06,
+      "loss": 0.8738,
+      "step": 2868
+    },
+    {
+      "epoch": 0.551943055021162,
+      "grad_norm": 1.0683729165270084,
+      "learning_rate": 8.806529091294948e-06,
+      "loss": 0.8227,
+      "step": 2869
+    },
+    {
+      "epoch": 0.5521354367064255,
+      "grad_norm": 1.2493808768574746,
+      "learning_rate": 8.800343011561633e-06,
+      "loss": 0.8769,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5523278183916891,
+      "grad_norm": 1.4905630858847956,
+      "learning_rate": 8.794157397576464e-06,
+      "loss": 0.8275,
+      "step": 2871
+    },
+    {
+      "epoch": 0.5525202000769527,
+      "grad_norm": 0.7111114338749177,
+      "learning_rate": 8.787972251740916e-06,
+      "loss": 0.3495,
+      "step": 2872
+    },
+    {
+      "epoch": 0.5527125817622163,
+      "grad_norm": 1.2577360178576964,
+      "learning_rate": 8.781787576456269e-06,
+      "loss": 0.8193,
+      "step": 2873
+    },
+    {
+      "epoch": 0.5529049634474797,
+      "grad_norm": 1.5290236218484623,
+      "learning_rate": 8.775603374123627e-06,
+      "loss": 0.8578,
+      "step": 2874
+    },
+    {
+      "epoch": 0.5530973451327433,
+      "grad_norm": 1.2816842847364696,
+      "learning_rate": 8.769419647143917e-06,
+      "loss": 0.8456,
+      "step": 2875
+    },
+    {
+      "epoch": 0.5532897268180069,
+      "grad_norm": 1.197394455707531,
+      "learning_rate": 8.763236397917865e-06,
+      "loss": 0.8398,
+      "step": 2876
+    },
+    {
+      "epoch": 0.5534821085032705,
+      "grad_norm": 1.2661402542488571,
+      "learning_rate": 8.757053628846028e-06,
+      "loss": 0.8688,
+      "step": 2877
+    },
+    {
+      "epoch": 0.5536744901885341,
+      "grad_norm": 1.7151060575222483,
+      "learning_rate": 8.75087134232877e-06,
+      "loss": 0.8526,
+      "step": 2878
+    },
+    {
+      "epoch": 0.5538668718737976,
+      "grad_norm": 1.1768726870417918,
+      "learning_rate": 8.744689540766265e-06,
+      "loss": 0.8502,
+      "step": 2879
+    },
+    {
+      "epoch": 0.5540592535590612,
+      "grad_norm": 1.174868885571107,
+      "learning_rate": 8.738508226558499e-06,
+      "loss": 0.8526,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5542516352443247,
+      "grad_norm": 1.3205764937313142,
+      "learning_rate": 8.73232740210528e-06,
+      "loss": 0.8659,
+      "step": 2881
+    },
+    {
+      "epoch": 0.5544440169295883,
+      "grad_norm": 1.3770858679050733,
+      "learning_rate": 8.726147069806206e-06,
+      "loss": 0.8754,
+      "step": 2882
+    },
+    {
+      "epoch": 0.5546363986148518,
+      "grad_norm": 1.1693867876149628,
+      "learning_rate": 8.719967232060698e-06,
+      "loss": 0.8482,
+      "step": 2883
+    },
+    {
+      "epoch": 0.5548287803001154,
+      "grad_norm": 1.2259478316830406,
+      "learning_rate": 8.713787891267988e-06,
+      "loss": 0.7782,
+      "step": 2884
+    },
+    {
+      "epoch": 0.555021161985379,
+      "grad_norm": 1.3913317339604518,
+      "learning_rate": 8.707609049827102e-06,
+      "loss": 0.8455,
+      "step": 2885
+    },
+    {
+      "epoch": 0.5552135436706426,
+      "grad_norm": 1.473291590008908,
+      "learning_rate": 8.70143071013688e-06,
+      "loss": 0.8549,
+      "step": 2886
+    },
+    {
+      "epoch": 0.5554059253559062,
+      "grad_norm": 1.4826854962338012,
+      "learning_rate": 8.695252874595972e-06,
+      "loss": 0.855,
+      "step": 2887
+    },
+    {
+      "epoch": 0.5555983070411696,
+      "grad_norm": 1.3145403917537521,
+      "learning_rate": 8.689075545602816e-06,
+      "loss": 0.8252,
+      "step": 2888
+    },
+    {
+      "epoch": 0.5557906887264332,
+      "grad_norm": 0.7396965957393117,
+      "learning_rate": 8.68289872555567e-06,
+      "loss": 0.3644,
+      "step": 2889
+    },
+    {
+      "epoch": 0.5559830704116968,
+      "grad_norm": 1.353871417352411,
+      "learning_rate": 8.676722416852594e-06,
+      "loss": 0.8651,
+      "step": 2890
+    },
+    {
+      "epoch": 0.5561754520969604,
+      "grad_norm": 1.1850092810822461,
+      "learning_rate": 8.670546621891434e-06,
+      "loss": 0.8482,
+      "step": 2891
+    },
+    {
+      "epoch": 0.556367833782224,
+      "grad_norm": 1.2632967530861638,
+      "learning_rate": 8.66437134306985e-06,
+      "loss": 0.8577,
+      "step": 2892
+    },
+    {
+      "epoch": 0.5565602154674875,
+      "grad_norm": 1.0805349909147446,
+      "learning_rate": 8.658196582785297e-06,
+      "loss": 0.824,
+      "step": 2893
+    },
+    {
+      "epoch": 0.556752597152751,
+      "grad_norm": 1.2912761076626336,
+      "learning_rate": 8.652022343435027e-06,
+      "loss": 0.8801,
+      "step": 2894
+    },
+    {
+      "epoch": 0.5569449788380146,
+      "grad_norm": 1.3074286321691018,
+      "learning_rate": 8.645848627416102e-06,
+      "loss": 0.8888,
+      "step": 2895
+    },
+    {
+      "epoch": 0.5571373605232782,
+      "grad_norm": 1.3435246470783133,
+      "learning_rate": 8.63967543712536e-06,
+      "loss": 0.8506,
+      "step": 2896
+    },
+    {
+      "epoch": 0.5573297422085417,
+      "grad_norm": 1.274959452821429,
+      "learning_rate": 8.633502774959453e-06,
+      "loss": 0.8874,
+      "step": 2897
+    },
+    {
+      "epoch": 0.5575221238938053,
+      "grad_norm": 1.3320415063845792,
+      "learning_rate": 8.627330643314818e-06,
+      "loss": 0.8647,
+      "step": 2898
+    },
+    {
+      "epoch": 0.5577145055790689,
+      "grad_norm": 1.258892149896094,
+      "learning_rate": 8.62115904458769e-06,
+      "loss": 0.8497,
+      "step": 2899
+    },
+    {
+      "epoch": 0.5579068872643325,
+      "grad_norm": 1.4596742459775764,
+      "learning_rate": 8.614987981174093e-06,
+      "loss": 0.8721,
+      "step": 2900
+    },
+    {
+      "epoch": 0.558099268949596,
+      "grad_norm": 1.383435399163714,
+      "learning_rate": 8.608817455469854e-06,
+      "loss": 0.8482,
+      "step": 2901
+    },
+    {
+      "epoch": 0.5582916506348595,
+      "grad_norm": 1.161384849968683,
+      "learning_rate": 8.602647469870573e-06,
+      "loss": 0.9077,
+      "step": 2902
+    },
+    {
+      "epoch": 0.5584840323201231,
+      "grad_norm": 1.3682266431319896,
+      "learning_rate": 8.596478026771658e-06,
+      "loss": 0.9022,
+      "step": 2903
+    },
+    {
+      "epoch": 0.5586764140053867,
+      "grad_norm": 1.1061269785318393,
+      "learning_rate": 8.590309128568303e-06,
+      "loss": 0.8888,
+      "step": 2904
+    },
+    {
+      "epoch": 0.5588687956906503,
+      "grad_norm": 1.2477670946176993,
+      "learning_rate": 8.584140777655476e-06,
+      "loss": 0.8848,
+      "step": 2905
+    },
+    {
+      "epoch": 0.5590611773759138,
+      "grad_norm": 1.3079203512471687,
+      "learning_rate": 8.57797297642795e-06,
+      "loss": 0.8247,
+      "step": 2906
+    },
+    {
+      "epoch": 0.5592535590611774,
+      "grad_norm": 1.2418618274093511,
+      "learning_rate": 8.571805727280278e-06,
+      "loss": 0.861,
+      "step": 2907
+    },
+    {
+      "epoch": 0.5594459407464409,
+      "grad_norm": 1.089040807739321,
+      "learning_rate": 8.565639032606794e-06,
+      "loss": 0.8772,
+      "step": 2908
+    },
+    {
+      "epoch": 0.5596383224317045,
+      "grad_norm": 1.2851048032954484,
+      "learning_rate": 8.559472894801623e-06,
+      "loss": 0.8265,
+      "step": 2909
+    },
+    {
+      "epoch": 0.5598307041169681,
+      "grad_norm": 1.2949781279626855,
+      "learning_rate": 8.553307316258678e-06,
+      "loss": 0.8283,
+      "step": 2910
+    },
+    {
+      "epoch": 0.5600230858022316,
+      "grad_norm": 1.4087468180895961,
+      "learning_rate": 8.547142299371642e-06,
+      "loss": 0.8326,
+      "step": 2911
+    },
+    {
+      "epoch": 0.5602154674874952,
+      "grad_norm": 1.186116838488124,
+      "learning_rate": 8.540977846533986e-06,
+      "loss": 0.8625,
+      "step": 2912
+    },
+    {
+      "epoch": 0.5604078491727588,
+      "grad_norm": 1.2724065472847732,
+      "learning_rate": 8.534813960138968e-06,
+      "loss": 0.8474,
+      "step": 2913
+    },
+    {
+      "epoch": 0.5606002308580224,
+      "grad_norm": 1.2321558886973432,
+      "learning_rate": 8.528650642579618e-06,
+      "loss": 0.8413,
+      "step": 2914
+    },
+    {
+      "epoch": 0.5607926125432858,
+      "grad_norm": 1.2312849679379971,
+      "learning_rate": 8.52248789624875e-06,
+      "loss": 0.8929,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5609849942285494,
+      "grad_norm": 1.312669924244407,
+      "learning_rate": 8.516325723538949e-06,
+      "loss": 0.9399,
+      "step": 2916
+    },
+    {
+      "epoch": 0.561177375913813,
+      "grad_norm": 1.2656614419133585,
+      "learning_rate": 8.510164126842591e-06,
+      "loss": 0.8391,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5613697575990766,
+      "grad_norm": 1.2552821496770816,
+      "learning_rate": 8.504003108551814e-06,
+      "loss": 0.8795,
+      "step": 2918
+    },
+    {
+      "epoch": 0.5615621392843402,
+      "grad_norm": 1.3014180453931925,
+      "learning_rate": 8.497842671058539e-06,
+      "loss": 0.8967,
+      "step": 2919
+    },
+    {
+      "epoch": 0.5617545209696037,
+      "grad_norm": 1.5240835299595061,
+      "learning_rate": 8.491682816754456e-06,
+      "loss": 0.8248,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5619469026548672,
+      "grad_norm": 1.1670078859767916,
+      "learning_rate": 8.485523548031044e-06,
+      "loss": 0.8911,
+      "step": 2921
+    },
+    {
+      "epoch": 0.5621392843401308,
+      "grad_norm": 1.234139354852884,
+      "learning_rate": 8.479364867279529e-06,
+      "loss": 0.8915,
+      "step": 2922
+    },
+    {
+      "epoch": 0.5623316660253944,
+      "grad_norm": 1.2522366306136146,
+      "learning_rate": 8.47320677689093e-06,
+      "loss": 0.8805,
+      "step": 2923
+    },
+    {
+      "epoch": 0.5625240477106579,
+      "grad_norm": 1.2103899517815344,
+      "learning_rate": 8.467049279256034e-06,
+      "loss": 0.8305,
+      "step": 2924
+    },
+    {
+      "epoch": 0.5627164293959215,
+      "grad_norm": 1.172410457656157,
+      "learning_rate": 8.460892376765387e-06,
+      "loss": 0.9436,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5629088110811851,
+      "grad_norm": 1.2518935575451822,
+      "learning_rate": 8.45473607180931e-06,
+      "loss": 0.8364,
+      "step": 2926
+    },
+    {
+      "epoch": 0.5631011927664487,
+      "grad_norm": 1.4920999056215214,
+      "learning_rate": 8.448580366777898e-06,
+      "loss": 0.8851,
+      "step": 2927
+    },
+    {
+      "epoch": 0.5632935744517122,
+      "grad_norm": 1.3508212276054647,
+      "learning_rate": 8.442425264061e-06,
+      "loss": 0.8925,
+      "step": 2928
+    },
+    {
+      "epoch": 0.5634859561369757,
+      "grad_norm": 1.3453892484253285,
+      "learning_rate": 8.436270766048245e-06,
+      "loss": 0.8549,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5636783378222393,
+      "grad_norm": 1.8719080400450594,
+      "learning_rate": 8.430116875129023e-06,
+      "loss": 0.8795,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5638707195075029,
+      "grad_norm": 1.105751384681539,
+      "learning_rate": 8.42396359369248e-06,
+      "loss": 0.8609,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5640631011927665,
+      "grad_norm": 1.27935924603711,
+      "learning_rate": 8.417810924127533e-06,
+      "loss": 0.9171,
+      "step": 2932
+    },
+    {
+      "epoch": 0.56425548287803,
+      "grad_norm": 1.1705215926630832,
+      "learning_rate": 8.411658868822866e-06,
+      "loss": 0.8717,
+      "step": 2933
+    },
+    {
+      "epoch": 0.5644478645632935,
+      "grad_norm": 1.2828432941430015,
+      "learning_rate": 8.40550743016691e-06,
+      "loss": 0.806,
+      "step": 2934
+    },
+    {
+      "epoch": 0.5646402462485571,
+      "grad_norm": 1.7228801101490696,
+      "learning_rate": 8.39935661054787e-06,
+      "loss": 0.8335,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5648326279338207,
+      "grad_norm": 1.1769369108844425,
+      "learning_rate": 8.393206412353709e-06,
+      "loss": 0.8665,
+      "step": 2936
+    },
+    {
+      "epoch": 0.5650250096190843,
+      "grad_norm": 1.1645037093616808,
+      "learning_rate": 8.38705683797214e-06,
+      "loss": 0.8501,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5652173913043478,
+      "grad_norm": 1.3842310289918205,
+      "learning_rate": 8.38090788979064e-06,
+      "loss": 0.822,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5654097729896114,
+      "grad_norm": 1.3009412358003498,
+      "learning_rate": 8.374759570196448e-06,
+      "loss": 0.8165,
+      "step": 2939
+    },
+    {
+      "epoch": 0.565602154674875,
+      "grad_norm": 1.316138008730748,
+      "learning_rate": 8.368611881576547e-06,
+      "loss": 0.863,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5657945363601385,
+      "grad_norm": 1.2923958956920567,
+      "learning_rate": 8.362464826317687e-06,
+      "loss": 0.8709,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5659869180454021,
+      "grad_norm": 1.2954180391373062,
+      "learning_rate": 8.35631840680636e-06,
+      "loss": 0.8701,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5661792997306656,
+      "grad_norm": 1.2779659751844288,
+      "learning_rate": 8.35017262542882e-06,
+      "loss": 0.8537,
+      "step": 2943
+    },
+    {
+      "epoch": 0.5663716814159292,
+      "grad_norm": 1.2795115621990618,
+      "learning_rate": 8.344027484571075e-06,
+      "loss": 0.8761,
+      "step": 2944
+    },
+    {
+      "epoch": 0.5665640631011928,
+      "grad_norm": 1.3546031221899881,
+      "learning_rate": 8.337882986618877e-06,
+      "loss": 0.8141,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5667564447864564,
+      "grad_norm": 1.3425648400953125,
+      "learning_rate": 8.331739133957729e-06,
+      "loss": 0.8445,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5669488264717198,
+      "grad_norm": 1.5258530901358551,
+      "learning_rate": 8.325595928972894e-06,
+      "loss": 0.857,
+      "step": 2947
+    },
+    {
+      "epoch": 0.5671412081569834,
+      "grad_norm": 1.0781711268704286,
+      "learning_rate": 8.319453374049367e-06,
+      "loss": 0.8319,
+      "step": 2948
+    },
+    {
+      "epoch": 0.567333589842247,
+      "grad_norm": 1.2821270740352482,
+      "learning_rate": 8.313311471571903e-06,
+      "loss": 0.8227,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5675259715275106,
+      "grad_norm": 2.1658414382336577,
+      "learning_rate": 8.307170223925003e-06,
+      "loss": 0.828,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5677183532127742,
+      "grad_norm": 1.3897685873543706,
+      "learning_rate": 8.301029633492907e-06,
+      "loss": 0.9038,
+      "step": 2951
+    },
+    {
+      "epoch": 0.5679107348980377,
+      "grad_norm": 1.1983424089620336,
+      "learning_rate": 8.294889702659602e-06,
+      "loss": 0.8635,
+      "step": 2952
+    },
+    {
+      "epoch": 0.5681031165833013,
+      "grad_norm": 1.2442620877497947,
+      "learning_rate": 8.288750433808828e-06,
+      "loss": 0.8596,
+      "step": 2953
+    },
+    {
+      "epoch": 0.5682954982685648,
+      "grad_norm": 1.1268806440363577,
+      "learning_rate": 8.282611829324049e-06,
+      "loss": 0.8839,
+      "step": 2954
+    },
+    {
+      "epoch": 0.5684878799538284,
+      "grad_norm": 1.30010055268822,
+      "learning_rate": 8.276473891588492e-06,
+      "loss": 0.821,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5686802616390919,
+      "grad_norm": 1.33028394096368,
+      "learning_rate": 8.270336622985116e-06,
+      "loss": 0.8846,
+      "step": 2956
+    },
+    {
+      "epoch": 0.5688726433243555,
+      "grad_norm": 1.4656354970571184,
+      "learning_rate": 8.264200025896616e-06,
+      "loss": 0.8815,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5690650250096191,
+      "grad_norm": 1.3053238105479665,
+      "learning_rate": 8.258064102705428e-06,
+      "loss": 0.8619,
+      "step": 2958
+    },
+    {
+      "epoch": 0.5692574066948827,
+      "grad_norm": 1.2756507049088017,
+      "learning_rate": 8.251928855793736e-06,
+      "loss": 0.8769,
+      "step": 2959
+    },
+    {
+      "epoch": 0.5694497883801463,
+      "grad_norm": 1.384218298911865,
+      "learning_rate": 8.245794287543447e-06,
+      "loss": 0.8311,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5696421700654097,
+      "grad_norm": 1.266746353707903,
+      "learning_rate": 8.239660400336213e-06,
+      "loss": 0.8406,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5698345517506733,
+      "grad_norm": 1.3226918982081588,
+      "learning_rate": 8.233527196553428e-06,
+      "loss": 0.9102,
+      "step": 2962
+    },
+    {
+      "epoch": 0.5700269334359369,
+      "grad_norm": 1.378733343708527,
+      "learning_rate": 8.227394678576204e-06,
+      "loss": 0.8286,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5702193151212005,
+      "grad_norm": 1.369708062736868,
+      "learning_rate": 8.221262848785395e-06,
+      "loss": 0.8441,
+      "step": 2964
+    },
+    {
+      "epoch": 0.570411696806464,
+      "grad_norm": 1.4191047555506051,
+      "learning_rate": 8.215131709561597e-06,
+      "loss": 0.912,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5706040784917276,
+      "grad_norm": 11.00916460637629,
+      "learning_rate": 8.20900126328512e-06,
+      "loss": 0.8416,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5707964601769911,
+      "grad_norm": 1.5281769461090244,
+      "learning_rate": 8.202871512336023e-06,
+      "loss": 0.8692,
+      "step": 2967
+    },
+    {
+      "epoch": 0.5709888418622547,
+      "grad_norm": 1.2897087211422769,
+      "learning_rate": 8.196742459094079e-06,
+      "loss": 0.9013,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5711812235475183,
+      "grad_norm": 0.6835190285353269,
+      "learning_rate": 8.190614105938796e-06,
+      "loss": 0.3494,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5713736052327818,
+      "grad_norm": 1.8351735665011482,
+      "learning_rate": 8.184486455249424e-06,
+      "loss": 0.825,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5715659869180454,
+      "grad_norm": 1.1732232502849627,
+      "learning_rate": 8.178359509404916e-06,
+      "loss": 0.8701,
+      "step": 2971
+    },
+    {
+      "epoch": 0.571758368603309,
+      "grad_norm": 1.3493592557539549,
+      "learning_rate": 8.172233270783966e-06,
+      "loss": 0.8386,
+      "step": 2972
+    },
+    {
+      "epoch": 0.5719507502885726,
+      "grad_norm": 1.2664154575990998,
+      "learning_rate": 8.166107741764997e-06,
+      "loss": 0.829,
+      "step": 2973
+    },
+    {
+      "epoch": 0.572143131973836,
+      "grad_norm": 1.734481787886353,
+      "learning_rate": 8.15998292472614e-06,
+      "loss": 0.838,
+      "step": 2974
+    },
+    {
+      "epoch": 0.5723355136590996,
+      "grad_norm": 1.5552670183553612,
+      "learning_rate": 8.153858822045267e-06,
+      "loss": 0.8787,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5725278953443632,
+      "grad_norm": 1.3579003807465941,
+      "learning_rate": 8.147735436099967e-06,
+      "loss": 0.8522,
+      "step": 2976
+    },
+    {
+      "epoch": 0.5727202770296268,
+      "grad_norm": 1.8463835873508838,
+      "learning_rate": 8.141612769267543e-06,
+      "loss": 0.857,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5729126587148904,
+      "grad_norm": 1.4376417010906892,
+      "learning_rate": 8.135490823925027e-06,
+      "loss": 0.8519,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5731050404001539,
+      "grad_norm": 2.393151692097556,
+      "learning_rate": 8.129369602449176e-06,
+      "loss": 0.8642,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5732974220854175,
+      "grad_norm": 1.2319765533937004,
+      "learning_rate": 8.123249107216446e-06,
+      "loss": 0.8245,
+      "step": 2980
+    },
+    {
+      "epoch": 0.573489803770681,
+      "grad_norm": 1.1931669845058073,
+      "learning_rate": 8.117129340603032e-06,
+      "loss": 0.8473,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5736821854559446,
+      "grad_norm": 1.554566759185833,
+      "learning_rate": 8.111010304984841e-06,
+      "loss": 0.848,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5738745671412081,
+      "grad_norm": 1.398381002911154,
+      "learning_rate": 8.104892002737488e-06,
+      "loss": 0.7946,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5740669488264717,
+      "grad_norm": 1.4329328632511682,
+      "learning_rate": 8.098774436236308e-06,
+      "loss": 0.8486,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5742593305117353,
+      "grad_norm": 1.7912377691592676,
+      "learning_rate": 8.092657607856356e-06,
+      "loss": 0.9031,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5744517121969989,
+      "grad_norm": 0.6540864905134162,
+      "learning_rate": 8.086541519972388e-06,
+      "loss": 0.3228,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5746440938822625,
+      "grad_norm": 1.2164386057804475,
+      "learning_rate": 8.080426174958886e-06,
+      "loss": 0.8634,
+      "step": 2987
+    },
+    {
+      "epoch": 0.5748364755675259,
+      "grad_norm": 1.2498173160204387,
+      "learning_rate": 8.074311575190039e-06,
+      "loss": 0.8599,
+      "step": 2988
+    },
+    {
+      "epoch": 0.5750288572527895,
+      "grad_norm": 1.2777195502348684,
+      "learning_rate": 8.068197723039738e-06,
+      "loss": 0.8157,
+      "step": 2989
+    },
+    {
+      "epoch": 0.5752212389380531,
+      "grad_norm": 0.647907519612668,
+      "learning_rate": 8.062084620881598e-06,
+      "loss": 0.3654,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5754136206233167,
+      "grad_norm": 0.6522015982515701,
+      "learning_rate": 8.055972271088933e-06,
+      "loss": 0.3356,
+      "step": 2991
+    },
+    {
+      "epoch": 0.5756060023085803,
+      "grad_norm": 1.1938761918809113,
+      "learning_rate": 8.049860676034762e-06,
+      "loss": 0.812,
+      "step": 2992
+    },
+    {
+      "epoch": 0.5757983839938438,
+      "grad_norm": 1.7367886904069325,
+      "learning_rate": 8.043749838091828e-06,
+      "loss": 0.8559,
+      "step": 2993
+    },
+    {
+      "epoch": 0.5759907656791073,
+      "grad_norm": 1.2375701414969977,
+      "learning_rate": 8.037639759632558e-06,
+      "loss": 0.8804,
+      "step": 2994
+    },
+    {
+      "epoch": 0.5761831473643709,
+      "grad_norm": 1.278739548106202,
+      "learning_rate": 8.031530443029099e-06,
+      "loss": 0.8929,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5763755290496345,
+      "grad_norm": 1.2421280042809446,
+      "learning_rate": 8.025421890653303e-06,
+      "loss": 0.8523,
+      "step": 2996
+    },
+    {
+      "epoch": 0.576567910734898,
+      "grad_norm": 1.3128497363908216,
+      "learning_rate": 8.019314104876712e-06,
+      "loss": 0.8369,
+      "step": 2997
+    },
+    {
+      "epoch": 0.5767602924201616,
+      "grad_norm": 1.3385446976611366,
+      "learning_rate": 8.013207088070582e-06,
+      "loss": 0.8669,
+      "step": 2998
+    },
+    {
+      "epoch": 0.5769526741054252,
+      "grad_norm": 0.6023841559551208,
+      "learning_rate": 8.007100842605872e-06,
+      "loss": 0.3445,
+      "step": 2999
+    },
+    {
+      "epoch": 0.5771450557906888,
+      "grad_norm": 1.5457532633706765,
+      "learning_rate": 8.000995370853227e-06,
+      "loss": 0.8169,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5773374374759523,
+      "grad_norm": 1.4982734356461314,
+      "learning_rate": 7.994890675183008e-06,
+      "loss": 0.8609,
+      "step": 3001
+    },
+    {
+      "epoch": 0.5775298191612158,
+      "grad_norm": 1.268461714652907,
+      "learning_rate": 7.98878675796527e-06,
+      "loss": 0.7553,
+      "step": 3002
+    },
+    {
+      "epoch": 0.5777222008464794,
+      "grad_norm": 0.6314564565823114,
+      "learning_rate": 7.98268362156976e-06,
+      "loss": 0.3453,
+      "step": 3003
+    },
+    {
+      "epoch": 0.577914582531743,
+      "grad_norm": 1.5269727211536113,
+      "learning_rate": 7.976581268365924e-06,
+      "loss": 0.9385,
+      "step": 3004
+    },
+    {
+      "epoch": 0.5781069642170066,
+      "grad_norm": 1.2960929879676493,
+      "learning_rate": 7.97047970072291e-06,
+      "loss": 0.841,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5782993459022701,
+      "grad_norm": 1.5638290176025682,
+      "learning_rate": 7.964378921009552e-06,
+      "loss": 0.8637,
+      "step": 3006
+    },
+    {
+      "epoch": 0.5784917275875336,
+      "grad_norm": 1.4034021902817706,
+      "learning_rate": 7.958278931594385e-06,
+      "loss": 0.8433,
+      "step": 3007
+    },
+    {
+      "epoch": 0.5786841092727972,
+      "grad_norm": 1.3625265179978967,
+      "learning_rate": 7.952179734845642e-06,
+      "loss": 0.8679,
+      "step": 3008
+    },
+    {
+      "epoch": 0.5788764909580608,
+      "grad_norm": 1.1806313116843636,
+      "learning_rate": 7.946081333131227e-06,
+      "loss": 0.8268,
+      "step": 3009
+    },
+    {
+      "epoch": 0.5790688726433244,
+      "grad_norm": 1.2724260127528257,
+      "learning_rate": 7.93998372881876e-06,
+      "loss": 0.8575,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5792612543285879,
+      "grad_norm": 4.7132245946988345,
+      "learning_rate": 7.93388692427554e-06,
+      "loss": 0.9124,
+      "step": 3011
+    },
+    {
+      "epoch": 0.5794536360138515,
+      "grad_norm": 1.2452939061923927,
+      "learning_rate": 7.92779092186855e-06,
+      "loss": 0.8335,
+      "step": 3012
+    },
+    {
+      "epoch": 0.5796460176991151,
+      "grad_norm": 1.4181359380715965,
+      "learning_rate": 7.921695723964473e-06,
+      "loss": 0.8964,
+      "step": 3013
+    },
+    {
+      "epoch": 0.5798383993843786,
+      "grad_norm": 1.2579254546778995,
+      "learning_rate": 7.915601332929678e-06,
+      "loss": 0.8381,
+      "step": 3014
+    },
+    {
+      "epoch": 0.5800307810696421,
+      "grad_norm": 1.3840632382596214,
+      "learning_rate": 7.90950775113021e-06,
+      "loss": 0.842,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5802231627549057,
+      "grad_norm": 3.7859285683353354,
+      "learning_rate": 7.903414980931813e-06,
+      "loss": 0.8399,
+      "step": 3016
+    },
+    {
+      "epoch": 0.5804155444401693,
+      "grad_norm": 1.183361023414518,
+      "learning_rate": 7.897323024699907e-06,
+      "loss": 0.9029,
+      "step": 3017
+    },
+    {
+      "epoch": 0.5806079261254329,
+      "grad_norm": 1.194209371160122,
+      "learning_rate": 7.8912318847996e-06,
+      "loss": 0.8999,
+      "step": 3018
+    },
+    {
+      "epoch": 0.5808003078106965,
+      "grad_norm": 1.1408203235140397,
+      "learning_rate": 7.885141563595685e-06,
+      "loss": 0.9037,
+      "step": 3019
+    },
+    {
+      "epoch": 0.58099268949596,
+      "grad_norm": 2.0020829584398787,
+      "learning_rate": 7.879052063452626e-06,
+      "loss": 0.869,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5811850711812235,
+      "grad_norm": 1.3087485809908985,
+      "learning_rate": 7.872963386734584e-06,
+      "loss": 0.88,
+      "step": 3021
+    },
+    {
+      "epoch": 0.5813774528664871,
+      "grad_norm": 1.433538047525204,
+      "learning_rate": 7.866875535805394e-06,
+      "loss": 0.9447,
+      "step": 3022
+    },
+    {
+      "epoch": 0.5815698345517507,
+      "grad_norm": 1.2471185516769627,
+      "learning_rate": 7.860788513028566e-06,
+      "loss": 0.8286,
+      "step": 3023
+    },
+    {
+      "epoch": 0.5817622162370142,
+      "grad_norm": 1.1166967085017898,
+      "learning_rate": 7.85470232076729e-06,
+      "loss": 0.8061,
+      "step": 3024
+    },
+    {
+      "epoch": 0.5819545979222778,
+      "grad_norm": 1.2048764500644256,
+      "learning_rate": 7.848616961384442e-06,
+      "loss": 0.8063,
+      "step": 3025
+    },
+    {
+      "epoch": 0.5821469796075414,
+      "grad_norm": 1.3383828960405475,
+      "learning_rate": 7.842532437242559e-06,
+      "loss": 0.8553,
+      "step": 3026
+    },
+    {
+      "epoch": 0.582339361292805,
+      "grad_norm": 1.2408719561978852,
+      "learning_rate": 7.83644875070387e-06,
+      "loss": 0.8092,
+      "step": 3027
+    },
+    {
+      "epoch": 0.5825317429780685,
+      "grad_norm": 1.4868387524522904,
+      "learning_rate": 7.83036590413027e-06,
+      "loss": 0.828,
+      "step": 3028
+    },
+    {
+      "epoch": 0.582724124663332,
+      "grad_norm": 1.125558212168135,
+      "learning_rate": 7.824283899883327e-06,
+      "loss": 0.8769,
+      "step": 3029
+    },
+    {
+      "epoch": 0.5829165063485956,
+      "grad_norm": 1.3361218556541565,
+      "learning_rate": 7.818202740324287e-06,
+      "loss": 0.8952,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5831088880338592,
+      "grad_norm": 0.6107843425809081,
+      "learning_rate": 7.812122427814068e-06,
+      "loss": 0.3326,
+      "step": 3031
+    },
+    {
+      "epoch": 0.5833012697191228,
+      "grad_norm": 1.1210366737260031,
+      "learning_rate": 7.806042964713248e-06,
+      "loss": 0.8447,
+      "step": 3032
+    },
+    {
+      "epoch": 0.5834936514043862,
+      "grad_norm": 1.6290284386299234,
+      "learning_rate": 7.79996435338209e-06,
+      "loss": 0.8655,
+      "step": 3033
+    },
+    {
+      "epoch": 0.5836860330896498,
+      "grad_norm": 1.3011618187677578,
+      "learning_rate": 7.793886596180521e-06,
+      "loss": 0.8781,
+      "step": 3034
+    },
+    {
+      "epoch": 0.5838784147749134,
+      "grad_norm": 0.6031034051436145,
+      "learning_rate": 7.787809695468134e-06,
+      "loss": 0.3542,
+      "step": 3035
+    },
+    {
+      "epoch": 0.584070796460177,
+      "grad_norm": 1.2085293450321186,
+      "learning_rate": 7.78173365360419e-06,
+      "loss": 0.8469,
+      "step": 3036
+    },
+    {
+      "epoch": 0.5842631781454406,
+      "grad_norm": 1.3288158454012247,
+      "learning_rate": 7.775658472947623e-06,
+      "loss": 0.8662,
+      "step": 3037
+    },
+    {
+      "epoch": 0.5844555598307041,
+      "grad_norm": 2.145963080470843,
+      "learning_rate": 7.769584155857019e-06,
+      "loss": 0.9182,
+      "step": 3038
+    },
+    {
+      "epoch": 0.5846479415159677,
+      "grad_norm": 1.3861777149019032,
+      "learning_rate": 7.763510704690645e-06,
+      "loss": 0.7882,
+      "step": 3039
+    },
+    {
+      "epoch": 0.5848403232012312,
+      "grad_norm": 1.6925190354480457,
+      "learning_rate": 7.757438121806414e-06,
+      "loss": 0.9003,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5850327048864948,
+      "grad_norm": 17.281588442963944,
+      "learning_rate": 7.75136640956192e-06,
+      "loss": 0.874,
+      "step": 3041
+    },
+    {
+      "epoch": 0.5852250865717584,
+      "grad_norm": 1.316845708295264,
+      "learning_rate": 7.745295570314412e-06,
+      "loss": 0.8702,
+      "step": 3042
+    },
+    {
+      "epoch": 0.5854174682570219,
+      "grad_norm": 1.57125965504669,
+      "learning_rate": 7.739225606420793e-06,
+      "loss": 0.8235,
+      "step": 3043
+    },
+    {
+      "epoch": 0.5856098499422855,
+      "grad_norm": 3.3647553497233393,
+      "learning_rate": 7.733156520237633e-06,
+      "loss": 0.8614,
+      "step": 3044
+    },
+    {
+      "epoch": 0.5858022316275491,
+      "grad_norm": 1.5089679535857763,
+      "learning_rate": 7.727088314121165e-06,
+      "loss": 0.8633,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5859946133128127,
+      "grad_norm": 4.0832065081170805,
+      "learning_rate": 7.721020990427268e-06,
+      "loss": 0.8472,
+      "step": 3046
+    },
+    {
+      "epoch": 0.5861869949980761,
+      "grad_norm": 2.0817740954989854,
+      "learning_rate": 7.714954551511489e-06,
+      "loss": 0.8188,
+      "step": 3047
+    },
+    {
+      "epoch": 0.5863793766833397,
+      "grad_norm": 2.178636434657909,
+      "learning_rate": 7.708888999729036e-06,
+      "loss": 0.8758,
+      "step": 3048
+    },
+    {
+      "epoch": 0.5865717583686033,
+      "grad_norm": 8.350360881125061,
+      "learning_rate": 7.702824337434756e-06,
+      "loss": 0.863,
+      "step": 3049
+    },
+    {
+      "epoch": 0.5867641400538669,
+      "grad_norm": 1.633132861264127,
+      "learning_rate": 7.69676056698316e-06,
+      "loss": 0.8564,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5869565217391305,
+      "grad_norm": 1.3326922236117917,
+      "learning_rate": 7.690697690728417e-06,
+      "loss": 0.8155,
+      "step": 3051
+    },
+    {
+      "epoch": 0.587148903424394,
+      "grad_norm": 1.8353892214402054,
+      "learning_rate": 7.68463571102434e-06,
+      "loss": 0.9204,
+      "step": 3052
+    },
+    {
+      "epoch": 0.5873412851096576,
+      "grad_norm": 1.77859449712826,
+      "learning_rate": 7.678574630224399e-06,
+      "loss": 0.9023,
+      "step": 3053
+    },
+    {
+      "epoch": 0.5875336667949211,
+      "grad_norm": 1.8883572644854254,
+      "learning_rate": 7.672514450681721e-06,
+      "loss": 0.8434,
+      "step": 3054
+    },
+    {
+      "epoch": 0.5877260484801847,
+      "grad_norm": 1.6273112322416037,
+      "learning_rate": 7.666455174749066e-06,
+      "loss": 0.8737,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5879184301654482,
+      "grad_norm": 1.7228856537570196,
+      "learning_rate": 7.66039680477886e-06,
+      "loss": 0.877,
+      "step": 3056
+    },
+    {
+      "epoch": 0.5881108118507118,
+      "grad_norm": 0.6226017369052277,
+      "learning_rate": 7.654339343123173e-06,
+      "loss": 0.3726,
+      "step": 3057
+    },
+    {
+      "epoch": 0.5883031935359754,
+      "grad_norm": 1.999295743300961,
+      "learning_rate": 7.648282792133711e-06,
+      "loss": 0.8631,
+      "step": 3058
+    },
+    {
+      "epoch": 0.588495575221239,
+      "grad_norm": 1.703615377374469,
+      "learning_rate": 7.642227154161841e-06,
+      "loss": 0.8789,
+      "step": 3059
+    },
+    {
+      "epoch": 0.5886879569065026,
+      "grad_norm": 2.0377825225879915,
+      "learning_rate": 7.636172431558575e-06,
+      "loss": 0.8557,
+      "step": 3060
+    },
+    {
+      "epoch": 0.588880338591766,
+      "grad_norm": 1.5330680044002658,
+      "learning_rate": 7.630118626674557e-06,
+      "loss": 0.8394,
+      "step": 3061
+    },
+    {
+      "epoch": 0.5890727202770296,
+      "grad_norm": 2.347143387267931,
+      "learning_rate": 7.6240657418600846e-06,
+      "loss": 0.8165,
+      "step": 3062
+    },
+    {
+      "epoch": 0.5892651019622932,
+      "grad_norm": 1.4654079483116635,
+      "learning_rate": 7.618013779465101e-06,
+      "loss": 0.8625,
+      "step": 3063
+    },
+    {
+      "epoch": 0.5894574836475568,
+      "grad_norm": 1.514069275026124,
+      "learning_rate": 7.611962741839178e-06,
+      "loss": 0.8987,
+      "step": 3064
+    },
+    {
+      "epoch": 0.5896498653328203,
+      "grad_norm": 1.6219915885823342,
+      "learning_rate": 7.6059126313315466e-06,
+      "loss": 0.878,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5898422470180839,
+      "grad_norm": 1.6058654972550315,
+      "learning_rate": 7.599863450291056e-06,
+      "loss": 0.8538,
+      "step": 3066
+    },
+    {
+      "epoch": 0.5900346287033474,
+      "grad_norm": 1.4607208759206936,
+      "learning_rate": 7.593815201066215e-06,
+      "loss": 0.8754,
+      "step": 3067
+    },
+    {
+      "epoch": 0.590227010388611,
+      "grad_norm": 1.9162526623513503,
+      "learning_rate": 7.587767886005164e-06,
+      "loss": 0.9262,
+      "step": 3068
+    },
+    {
+      "epoch": 0.5904193920738746,
+      "grad_norm": 1.910754197191317,
+      "learning_rate": 7.581721507455672e-06,
+      "loss": 0.868,
+      "step": 3069
+    },
+    {
+      "epoch": 0.5906117737591381,
+      "grad_norm": 1.4506249157513567,
+      "learning_rate": 7.575676067765154e-06,
+      "loss": 0.8757,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5908041554444017,
+      "grad_norm": 1.2602732805405115,
+      "learning_rate": 7.569631569280662e-06,
+      "loss": 0.8355,
+      "step": 3071
+    },
+    {
+      "epoch": 0.5909965371296653,
+      "grad_norm": 1.3180245405812376,
+      "learning_rate": 7.563588014348871e-06,
+      "loss": 0.8434,
+      "step": 3072
+    },
+    {
+      "epoch": 0.5911889188149289,
+      "grad_norm": 0.6257220401049861,
+      "learning_rate": 7.5575454053161e-06,
+      "loss": 0.3227,
+      "step": 3073
+    },
+    {
+      "epoch": 0.5913813005001923,
+      "grad_norm": 1.2083096539189844,
+      "learning_rate": 7.551503744528304e-06,
+      "loss": 0.8355,
+      "step": 3074
+    },
+    {
+      "epoch": 0.5915736821854559,
+      "grad_norm": 2.066402834028911,
+      "learning_rate": 7.545463034331054e-06,
+      "loss": 0.8919,
+      "step": 3075
+    },
+    {
+      "epoch": 0.5917660638707195,
+      "grad_norm": 1.7593576152701422,
+      "learning_rate": 7.539423277069568e-06,
+      "loss": 0.8698,
+      "step": 3076
+    },
+    {
+      "epoch": 0.5919584455559831,
+      "grad_norm": 1.7718186389446848,
+      "learning_rate": 7.53338447508869e-06,
+      "loss": 0.8345,
+      "step": 3077
+    },
+    {
+      "epoch": 0.5921508272412467,
+      "grad_norm": 1.3986167123504634,
+      "learning_rate": 7.52734663073288e-06,
+      "loss": 0.8213,
+      "step": 3078
+    },
+    {
+      "epoch": 0.5923432089265102,
+      "grad_norm": 1.7233870864266945,
+      "learning_rate": 7.521309746346246e-06,
+      "loss": 0.8417,
+      "step": 3079
+    },
+    {
+      "epoch": 0.5925355906117737,
+      "grad_norm": 1.2529844163860553,
+      "learning_rate": 7.515273824272516e-06,
+      "loss": 0.87,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5927279722970373,
+      "grad_norm": 1.3972147463104947,
+      "learning_rate": 7.509238866855033e-06,
+      "loss": 0.8449,
+      "step": 3081
+    },
+    {
+      "epoch": 0.5929203539823009,
+      "grad_norm": 1.6792159507559576,
+      "learning_rate": 7.503204876436785e-06,
+      "loss": 0.8242,
+      "step": 3082
+    },
+    {
+      "epoch": 0.5931127356675644,
+      "grad_norm": 1.6063475039529946,
+      "learning_rate": 7.497171855360372e-06,
+      "loss": 0.8468,
+      "step": 3083
+    },
+    {
+      "epoch": 0.593305117352828,
+      "grad_norm": 1.4244284646530314,
+      "learning_rate": 7.491139805968018e-06,
+      "loss": 0.8383,
+      "step": 3084
+    },
+    {
+      "epoch": 0.5934974990380916,
+      "grad_norm": 1.2298816577880918,
+      "learning_rate": 7.485108730601571e-06,
+      "loss": 0.8515,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5936898807233552,
+      "grad_norm": 1.2401783183705402,
+      "learning_rate": 7.4790786316025125e-06,
+      "loss": 0.7967,
+      "step": 3086
+    },
+    {
+      "epoch": 0.5938822624086187,
+      "grad_norm": 1.5306843733179971,
+      "learning_rate": 7.473049511311921e-06,
+      "loss": 0.9239,
+      "step": 3087
+    },
+    {
+      "epoch": 0.5940746440938822,
+      "grad_norm": 1.6291132452215604,
+      "learning_rate": 7.467021372070515e-06,
+      "loss": 0.8899,
+      "step": 3088
+    },
+    {
+      "epoch": 0.5942670257791458,
+      "grad_norm": 1.449126424344609,
+      "learning_rate": 7.46099421621863e-06,
+      "loss": 0.8844,
+      "step": 3089
+    },
+    {
+      "epoch": 0.5944594074644094,
+      "grad_norm": 1.37648811837837,
+      "learning_rate": 7.4549680460962044e-06,
+      "loss": 0.8268,
+      "step": 3090
+    },
+    {
+      "epoch": 0.594651789149673,
+      "grad_norm": 1.372732475016938,
+      "learning_rate": 7.448942864042819e-06,
+      "loss": 0.8037,
+      "step": 3091
+    },
+    {
+      "epoch": 0.5948441708349365,
+      "grad_norm": 1.2501995088623024,
+      "learning_rate": 7.4429186723976425e-06,
+      "loss": 0.8596,
+      "step": 3092
+    },
+    {
+      "epoch": 0.5950365525202,
+      "grad_norm": 1.6322899280787726,
+      "learning_rate": 7.43689547349948e-06,
+      "loss": 0.8558,
+      "step": 3093
+    },
+    {
+      "epoch": 0.5952289342054636,
+      "grad_norm": 2.006737784001712,
+      "learning_rate": 7.43087326968675e-06,
+      "loss": 0.8321,
+      "step": 3094
+    },
+    {
+      "epoch": 0.5954213158907272,
+      "grad_norm": 1.215648890109573,
+      "learning_rate": 7.42485206329747e-06,
+      "loss": 0.8481,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5956136975759908,
+      "grad_norm": 1.2806569770940046,
+      "learning_rate": 7.418831856669286e-06,
+      "loss": 0.9287,
+      "step": 3096
+    },
+    {
+      "epoch": 0.5958060792612543,
+      "grad_norm": 1.4108083501249358,
+      "learning_rate": 7.41281265213945e-06,
+      "loss": 0.8414,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5959984609465179,
+      "grad_norm": 1.3641443144312302,
+      "learning_rate": 7.406794452044816e-06,
+      "loss": 0.8122,
+      "step": 3098
+    },
+    {
+      "epoch": 0.5961908426317815,
+      "grad_norm": 0.6722476951570593,
+      "learning_rate": 7.400777258721865e-06,
+      "loss": 0.3654,
+      "step": 3099
+    },
+    {
+      "epoch": 0.596383224317045,
+      "grad_norm": 1.3942888594908254,
+      "learning_rate": 7.394761074506679e-06,
+      "loss": 0.8536,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5965756060023086,
+      "grad_norm": 1.5624033285497319,
+      "learning_rate": 7.3887459017349405e-06,
+      "loss": 0.8823,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5967679876875721,
+      "grad_norm": 1.547913486195677,
+      "learning_rate": 7.382731742741953e-06,
+      "loss": 0.8004,
+      "step": 3102
+    },
+    {
+      "epoch": 0.5969603693728357,
+      "grad_norm": 1.5672347633541641,
+      "learning_rate": 7.376718599862621e-06,
+      "loss": 0.8822,
+      "step": 3103
+    },
+    {
+      "epoch": 0.5971527510580993,
+      "grad_norm": 1.4211441822548796,
+      "learning_rate": 7.370706475431446e-06,
+      "loss": 0.8008,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5973451327433629,
+      "grad_norm": 1.3871814717045163,
+      "learning_rate": 7.364695371782547e-06,
+      "loss": 0.8793,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5975375144286263,
+      "grad_norm": 1.2891446637878743,
+      "learning_rate": 7.358685291249644e-06,
+      "loss": 0.8578,
+      "step": 3106
+    },
+    {
+      "epoch": 0.5977298961138899,
+      "grad_norm": 1.2800874944143008,
+      "learning_rate": 7.352676236166051e-06,
+      "loss": 0.844,
+      "step": 3107
+    },
+    {
+      "epoch": 0.5979222777991535,
+      "grad_norm": 1.3859555586253498,
+      "learning_rate": 7.346668208864695e-06,
+      "loss": 0.8206,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5981146594844171,
+      "grad_norm": 1.4064657401605327,
+      "learning_rate": 7.3406612116781e-06,
+      "loss": 0.8446,
+      "step": 3109
+    },
+    {
+      "epoch": 0.5983070411696807,
+      "grad_norm": 1.7731314744290614,
+      "learning_rate": 7.33465524693838e-06,
+      "loss": 0.7903,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5984994228549442,
+      "grad_norm": 1.3278779101264255,
+      "learning_rate": 7.328650316977265e-06,
+      "loss": 0.8326,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5986918045402078,
+      "grad_norm": 1.2612589415265523,
+      "learning_rate": 7.322646424126079e-06,
+      "loss": 0.8377,
+      "step": 3112
+    },
+    {
+      "epoch": 0.5988841862254713,
+      "grad_norm": 1.5463784022805398,
+      "learning_rate": 7.316643570715729e-06,
+      "loss": 0.8493,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5990765679107349,
+      "grad_norm": 1.2738346490701185,
+      "learning_rate": 7.310641759076742e-06,
+      "loss": 0.8013,
+      "step": 3114
+    },
+    {
+      "epoch": 0.5992689495959984,
+      "grad_norm": 1.362346232222819,
+      "learning_rate": 7.304640991539216e-06,
+      "loss": 0.9116,
+      "step": 3115
+    },
+    {
+      "epoch": 0.599461331281262,
+      "grad_norm": 1.2474529282726536,
+      "learning_rate": 7.2986412704328625e-06,
+      "loss": 0.8319,
+      "step": 3116
+    },
+    {
+      "epoch": 0.5996537129665256,
+      "grad_norm": 1.1743836555517948,
+      "learning_rate": 7.292642598086982e-06,
+      "loss": 0.8636,
+      "step": 3117
+    },
+    {
+      "epoch": 0.5998460946517892,
+      "grad_norm": 1.2723786927658636,
+      "learning_rate": 7.286644976830457e-06,
+      "loss": 0.8018,
+      "step": 3118
+    },
+    {
+      "epoch": 0.6000384763370528,
+      "grad_norm": 1.5103674986176197,
+      "learning_rate": 7.280648408991775e-06,
+      "loss": 0.8233,
+      "step": 3119
+    },
+    {
+      "epoch": 0.6002308580223162,
+      "grad_norm": 1.3517354012111364,
+      "learning_rate": 7.274652896899015e-06,
+      "loss": 0.9496,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6004232397075798,
+      "grad_norm": 1.2984368706780822,
+      "learning_rate": 7.268658442879834e-06,
+      "loss": 0.8987,
+      "step": 3121
+    },
+    {
+      "epoch": 0.6006156213928434,
+      "grad_norm": 1.3473642174975076,
+      "learning_rate": 7.262665049261489e-06,
+      "loss": 0.8213,
+      "step": 3122
+    },
+    {
+      "epoch": 0.600808003078107,
+      "grad_norm": 1.547050793279068,
+      "learning_rate": 7.256672718370824e-06,
+      "loss": 0.913,
+      "step": 3123
+    },
+    {
+      "epoch": 0.6010003847633705,
+      "grad_norm": 1.3727242629972793,
+      "learning_rate": 7.250681452534261e-06,
+      "loss": 0.7639,
+      "step": 3124
+    },
+    {
+      "epoch": 0.6011927664486341,
+      "grad_norm": 1.2492418160909786,
+      "learning_rate": 7.2446912540778196e-06,
+      "loss": 0.8434,
+      "step": 3125
+    },
+    {
+      "epoch": 0.6013851481338977,
+      "grad_norm": 1.2998476658365647,
+      "learning_rate": 7.238702125327106e-06,
+      "loss": 0.8507,
+      "step": 3126
+    },
+    {
+      "epoch": 0.6015775298191612,
+      "grad_norm": 1.3418569888827436,
+      "learning_rate": 7.232714068607296e-06,
+      "loss": 0.8422,
+      "step": 3127
+    },
+    {
+      "epoch": 0.6017699115044248,
+      "grad_norm": 1.1281541916652835,
+      "learning_rate": 7.226727086243168e-06,
+      "loss": 0.8114,
+      "step": 3128
+    },
+    {
+      "epoch": 0.6019622931896883,
+      "grad_norm": 1.1918088675923808,
+      "learning_rate": 7.220741180559074e-06,
+      "loss": 0.8619,
+      "step": 3129
+    },
+    {
+      "epoch": 0.6021546748749519,
+      "grad_norm": 1.8082974876906757,
+      "learning_rate": 7.214756353878942e-06,
+      "loss": 0.844,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6023470565602155,
+      "grad_norm": 1.3689098493585832,
+      "learning_rate": 7.208772608526293e-06,
+      "loss": 0.8429,
+      "step": 3131
+    },
+    {
+      "epoch": 0.6025394382454791,
+      "grad_norm": 1.1879842737317963,
+      "learning_rate": 7.202789946824227e-06,
+      "loss": 0.7823,
+      "step": 3132
+    },
+    {
+      "epoch": 0.6027318199307425,
+      "grad_norm": 1.3957475507641626,
+      "learning_rate": 7.1968083710954075e-06,
+      "loss": 0.8463,
+      "step": 3133
+    },
+    {
+      "epoch": 0.6029242016160061,
+      "grad_norm": 0.5802560593836943,
+      "learning_rate": 7.1908278836621e-06,
+      "loss": 0.3677,
+      "step": 3134
+    },
+    {
+      "epoch": 0.6031165833012697,
+      "grad_norm": 1.157229095753932,
+      "learning_rate": 7.184848486846128e-06,
+      "loss": 0.832,
+      "step": 3135
+    },
+    {
+      "epoch": 0.6033089649865333,
+      "grad_norm": 1.2986088057292995,
+      "learning_rate": 7.178870182968904e-06,
+      "loss": 0.8185,
+      "step": 3136
+    },
+    {
+      "epoch": 0.6035013466717969,
+      "grad_norm": 1.3669208320612507,
+      "learning_rate": 7.1728929743514065e-06,
+      "loss": 0.9055,
+      "step": 3137
+    },
+    {
+      "epoch": 0.6036937283570604,
+      "grad_norm": 1.2278576925109574,
+      "learning_rate": 7.166916863314199e-06,
+      "loss": 0.882,
+      "step": 3138
+    },
+    {
+      "epoch": 0.603886110042324,
+      "grad_norm": 1.459814793659577,
+      "learning_rate": 7.1609418521774095e-06,
+      "loss": 0.8555,
+      "step": 3139
+    },
+    {
+      "epoch": 0.6040784917275875,
+      "grad_norm": 1.1609661347588482,
+      "learning_rate": 7.154967943260748e-06,
+      "loss": 0.9279,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6042708734128511,
+      "grad_norm": 1.3305200157547568,
+      "learning_rate": 7.148995138883483e-06,
+      "loss": 0.8801,
+      "step": 3141
+    },
+    {
+      "epoch": 0.6044632550981146,
+      "grad_norm": 1.3887504035635265,
+      "learning_rate": 7.143023441364471e-06,
+      "loss": 0.8545,
+      "step": 3142
+    },
+    {
+      "epoch": 0.6046556367833782,
+      "grad_norm": 1.4800576466118578,
+      "learning_rate": 7.13705285302213e-06,
+      "loss": 0.8441,
+      "step": 3143
+    },
+    {
+      "epoch": 0.6048480184686418,
+      "grad_norm": 1.4856580420379113,
+      "learning_rate": 7.131083376174441e-06,
+      "loss": 0.8486,
+      "step": 3144
+    },
+    {
+      "epoch": 0.6050404001539054,
+      "grad_norm": 1.2301668374314754,
+      "learning_rate": 7.125115013138966e-06,
+      "loss": 0.8493,
+      "step": 3145
+    },
+    {
+      "epoch": 0.605232781839169,
+      "grad_norm": 1.2376263399164946,
+      "learning_rate": 7.119147766232832e-06,
+      "loss": 0.8201,
+      "step": 3146
+    },
+    {
+      "epoch": 0.6054251635244324,
+      "grad_norm": 1.1387427518865774,
+      "learning_rate": 7.113181637772721e-06,
+      "loss": 0.8682,
+      "step": 3147
+    },
+    {
+      "epoch": 0.605617545209696,
+      "grad_norm": 1.233586132115944,
+      "learning_rate": 7.107216630074895e-06,
+      "loss": 0.7924,
+      "step": 3148
+    },
+    {
+      "epoch": 0.6058099268949596,
+      "grad_norm": 0.6075143453800479,
+      "learning_rate": 7.1012527454551795e-06,
+      "loss": 0.3261,
+      "step": 3149
+    },
+    {
+      "epoch": 0.6060023085802232,
+      "grad_norm": 1.1626586513274138,
+      "learning_rate": 7.09528998622895e-06,
+      "loss": 0.8963,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6061946902654868,
+      "grad_norm": 1.3341157634532226,
+      "learning_rate": 7.089328354711159e-06,
+      "loss": 0.8105,
+      "step": 3151
+    },
+    {
+      "epoch": 0.6063870719507503,
+      "grad_norm": 1.1956465647832721,
+      "learning_rate": 7.083367853216323e-06,
+      "loss": 0.8938,
+      "step": 3152
+    },
+    {
+      "epoch": 0.6065794536360138,
+      "grad_norm": 1.4103670672110125,
+      "learning_rate": 7.077408484058505e-06,
+      "loss": 0.8779,
+      "step": 3153
+    },
+    {
+      "epoch": 0.6067718353212774,
+      "grad_norm": 1.3704695242543,
+      "learning_rate": 7.071450249551342e-06,
+      "loss": 0.8062,
+      "step": 3154
+    },
+    {
+      "epoch": 0.606964217006541,
+      "grad_norm": 1.2597355516052262,
+      "learning_rate": 7.065493152008026e-06,
+      "loss": 0.912,
+      "step": 3155
+    },
+    {
+      "epoch": 0.6071565986918045,
+      "grad_norm": 1.3853255486107061,
+      "learning_rate": 7.059537193741306e-06,
+      "loss": 0.8765,
+      "step": 3156
+    },
+    {
+      "epoch": 0.6073489803770681,
+      "grad_norm": 1.3682072641939214,
+      "learning_rate": 7.053582377063489e-06,
+      "loss": 0.882,
+      "step": 3157
+    },
+    {
+      "epoch": 0.6075413620623317,
+      "grad_norm": 1.1190803258571025,
+      "learning_rate": 7.047628704286446e-06,
+      "loss": 0.8376,
+      "step": 3158
+    },
+    {
+      "epoch": 0.6077337437475953,
+      "grad_norm": 1.6182269955549822,
+      "learning_rate": 7.041676177721588e-06,
+      "loss": 0.9463,
+      "step": 3159
+    },
+    {
+      "epoch": 0.6079261254328588,
+      "grad_norm": 0.6393273497834492,
+      "learning_rate": 7.035724799679898e-06,
+      "loss": 0.3573,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6081185071181223,
+      "grad_norm": 0.6503326344186384,
+      "learning_rate": 7.029774572471904e-06,
+      "loss": 0.3617,
+      "step": 3161
+    },
+    {
+      "epoch": 0.6083108888033859,
+      "grad_norm": 1.1422978229265348,
+      "learning_rate": 7.023825498407689e-06,
+      "loss": 0.8503,
+      "step": 3162
+    },
+    {
+      "epoch": 0.6085032704886495,
+      "grad_norm": 1.8288135614735381,
+      "learning_rate": 7.0178775797968855e-06,
+      "loss": 0.8461,
+      "step": 3163
+    },
+    {
+      "epoch": 0.6086956521739131,
+      "grad_norm": 1.4605566624414872,
+      "learning_rate": 7.011930818948688e-06,
+      "loss": 0.872,
+      "step": 3164
+    },
+    {
+      "epoch": 0.6088880338591766,
+      "grad_norm": 1.3536719409555495,
+      "learning_rate": 7.005985218171825e-06,
+      "loss": 0.8914,
+      "step": 3165
+    },
+    {
+      "epoch": 0.6090804155444401,
+      "grad_norm": 1.3859297136251025,
+      "learning_rate": 7.000040779774591e-06,
+      "loss": 0.8568,
+      "step": 3166
+    },
+    {
+      "epoch": 0.6092727972297037,
+      "grad_norm": 1.7403365490861853,
+      "learning_rate": 6.994097506064812e-06,
+      "loss": 0.8871,
+      "step": 3167
+    },
+    {
+      "epoch": 0.6094651789149673,
+      "grad_norm": 1.6847445532772694,
+      "learning_rate": 6.9881553993498805e-06,
+      "loss": 0.8206,
+      "step": 3168
+    },
+    {
+      "epoch": 0.6096575606002309,
+      "grad_norm": 1.2168217889605195,
+      "learning_rate": 6.9822144619367275e-06,
+      "loss": 0.8933,
+      "step": 3169
+    },
+    {
+      "epoch": 0.6098499422854944,
+      "grad_norm": 1.1225883599746482,
+      "learning_rate": 6.97627469613182e-06,
+      "loss": 0.8479,
+      "step": 3170
+    },
+    {
+      "epoch": 0.610042323970758,
+      "grad_norm": 1.6842609111700542,
+      "learning_rate": 6.970336104241186e-06,
+      "loss": 0.8412,
+      "step": 3171
+    },
+    {
+      "epoch": 0.6102347056560216,
+      "grad_norm": 1.9339842017165838,
+      "learning_rate": 6.9643986885703955e-06,
+      "loss": 0.8389,
+      "step": 3172
+    },
+    {
+      "epoch": 0.6104270873412851,
+      "grad_norm": 1.42503562066315,
+      "learning_rate": 6.958462451424547e-06,
+      "loss": 0.8505,
+      "step": 3173
+    },
+    {
+      "epoch": 0.6106194690265486,
+      "grad_norm": 1.3373691870758098,
+      "learning_rate": 6.952527395108302e-06,
+      "loss": 0.8597,
+      "step": 3174
+    },
+    {
+      "epoch": 0.6108118507118122,
+      "grad_norm": 1.381720454421873,
+      "learning_rate": 6.9465935219258504e-06,
+      "loss": 0.844,
+      "step": 3175
+    },
+    {
+      "epoch": 0.6110042323970758,
+      "grad_norm": 0.7016437948072054,
+      "learning_rate": 6.9406608341809215e-06,
+      "loss": 0.335,
+      "step": 3176
+    },
+    {
+      "epoch": 0.6111966140823394,
+      "grad_norm": 1.3716457314592103,
+      "learning_rate": 6.934729334176793e-06,
+      "loss": 0.8443,
+      "step": 3177
+    },
+    {
+      "epoch": 0.611388995767603,
+      "grad_norm": 0.6021485642074837,
+      "learning_rate": 6.928799024216282e-06,
+      "loss": 0.3408,
+      "step": 3178
+    },
+    {
+      "epoch": 0.6115813774528664,
+      "grad_norm": 1.219818396142967,
+      "learning_rate": 6.92286990660173e-06,
+      "loss": 0.847,
+      "step": 3179
+    },
+    {
+      "epoch": 0.61177375913813,
+      "grad_norm": 1.3750423027146697,
+      "learning_rate": 6.91694198363503e-06,
+      "loss": 0.8224,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6119661408233936,
+      "grad_norm": 1.4920822274877714,
+      "learning_rate": 6.911015257617606e-06,
+      "loss": 0.8738,
+      "step": 3181
+    },
+    {
+      "epoch": 0.6121585225086572,
+      "grad_norm": 1.6848057939461987,
+      "learning_rate": 6.905089730850416e-06,
+      "loss": 0.8803,
+      "step": 3182
+    },
+    {
+      "epoch": 0.6123509041939207,
+      "grad_norm": 1.35240337284888,
+      "learning_rate": 6.8991654056339505e-06,
+      "loss": 0.8597,
+      "step": 3183
+    },
+    {
+      "epoch": 0.6125432858791843,
+      "grad_norm": 1.2600255054618508,
+      "learning_rate": 6.893242284268244e-06,
+      "loss": 0.8374,
+      "step": 3184
+    },
+    {
+      "epoch": 0.6127356675644479,
+      "grad_norm": 1.9529550138814051,
+      "learning_rate": 6.887320369052848e-06,
+      "loss": 0.8764,
+      "step": 3185
+    },
+    {
+      "epoch": 0.6129280492497114,
+      "grad_norm": 1.3136450055865618,
+      "learning_rate": 6.8813996622868584e-06,
+      "loss": 0.9054,
+      "step": 3186
+    },
+    {
+      "epoch": 0.613120430934975,
+      "grad_norm": 1.137269386120344,
+      "learning_rate": 6.8754801662688964e-06,
+      "loss": 0.8586,
+      "step": 3187
+    },
+    {
+      "epoch": 0.6133128126202385,
+      "grad_norm": 1.4473447845641128,
+      "learning_rate": 6.869561883297116e-06,
+      "loss": 0.8469,
+      "step": 3188
+    },
+    {
+      "epoch": 0.6135051943055021,
+      "grad_norm": 1.1578858918411983,
+      "learning_rate": 6.863644815669197e-06,
+      "loss": 0.8338,
+      "step": 3189
+    },
+    {
+      "epoch": 0.6136975759907657,
+      "grad_norm": 2.0530765774642425,
+      "learning_rate": 6.857728965682344e-06,
+      "loss": 0.9043,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6138899576760293,
+      "grad_norm": 1.4158907173780346,
+      "learning_rate": 6.851814335633298e-06,
+      "loss": 0.8151,
+      "step": 3191
+    },
+    {
+      "epoch": 0.6140823393612928,
+      "grad_norm": 1.9358926939898957,
+      "learning_rate": 6.8459009278183275e-06,
+      "loss": 0.8379,
+      "step": 3192
+    },
+    {
+      "epoch": 0.6142747210465563,
+      "grad_norm": 1.24814954652076,
+      "learning_rate": 6.839988744533211e-06,
+      "loss": 0.8361,
+      "step": 3193
+    },
+    {
+      "epoch": 0.6144671027318199,
+      "grad_norm": 1.2369459537755867,
+      "learning_rate": 6.834077788073268e-06,
+      "loss": 0.8237,
+      "step": 3194
+    },
+    {
+      "epoch": 0.6146594844170835,
+      "grad_norm": 1.1642822759083689,
+      "learning_rate": 6.8281680607333364e-06,
+      "loss": 0.8404,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6148518661023471,
+      "grad_norm": 1.2732278736339702,
+      "learning_rate": 6.822259564807768e-06,
+      "loss": 0.8642,
+      "step": 3196
+    },
+    {
+      "epoch": 0.6150442477876106,
+      "grad_norm": 1.5114553967342477,
+      "learning_rate": 6.81635230259045e-06,
+      "loss": 0.8957,
+      "step": 3197
+    },
+    {
+      "epoch": 0.6152366294728742,
+      "grad_norm": 1.4846034944264672,
+      "learning_rate": 6.810446276374789e-06,
+      "loss": 0.8165,
+      "step": 3198
+    },
+    {
+      "epoch": 0.6154290111581378,
+      "grad_norm": 1.3448412685066096,
+      "learning_rate": 6.8045414884536975e-06,
+      "loss": 0.8679,
+      "step": 3199
+    },
+    {
+      "epoch": 0.6156213928434013,
+      "grad_norm": 1.4439903193149524,
+      "learning_rate": 6.7986379411196255e-06,
+      "loss": 0.843,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6158137745286649,
+      "grad_norm": 1.20080828999065,
+      "learning_rate": 6.7927356366645315e-06,
+      "loss": 0.8546,
+      "step": 3201
+    },
+    {
+      "epoch": 0.6160061562139284,
+      "grad_norm": 1.0525575460150056,
+      "learning_rate": 6.786834577379893e-06,
+      "loss": 0.8163,
+      "step": 3202
+    },
+    {
+      "epoch": 0.616198537899192,
+      "grad_norm": 1.3822196087776037,
+      "learning_rate": 6.780934765556702e-06,
+      "loss": 0.7846,
+      "step": 3203
+    },
+    {
+      "epoch": 0.6163909195844556,
+      "grad_norm": 1.2850139854741018,
+      "learning_rate": 6.775036203485472e-06,
+      "loss": 0.833,
+      "step": 3204
+    },
+    {
+      "epoch": 0.6165833012697192,
+      "grad_norm": 1.4142954512519679,
+      "learning_rate": 6.769138893456225e-06,
+      "loss": 0.8344,
+      "step": 3205
+    },
+    {
+      "epoch": 0.6167756829549826,
+      "grad_norm": 1.2627183350498679,
+      "learning_rate": 6.763242837758504e-06,
+      "loss": 0.8462,
+      "step": 3206
+    },
+    {
+      "epoch": 0.6169680646402462,
+      "grad_norm": 1.3898451385025763,
+      "learning_rate": 6.757348038681357e-06,
+      "loss": 0.8767,
+      "step": 3207
+    },
+    {
+      "epoch": 0.6171604463255098,
+      "grad_norm": 1.2867314787087765,
+      "learning_rate": 6.751454498513349e-06,
+      "loss": 0.8299,
+      "step": 3208
+    },
+    {
+      "epoch": 0.6173528280107734,
+      "grad_norm": 1.4065356358115209,
+      "learning_rate": 6.745562219542554e-06,
+      "loss": 0.9098,
+      "step": 3209
+    },
+    {
+      "epoch": 0.617545209696037,
+      "grad_norm": 1.2628618268896328,
+      "learning_rate": 6.7396712040565625e-06,
+      "loss": 0.7932,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6177375913813005,
+      "grad_norm": 1.3184706453947577,
+      "learning_rate": 6.733781454342463e-06,
+      "loss": 0.8281,
+      "step": 3211
+    },
+    {
+      "epoch": 0.617929973066564,
+      "grad_norm": 1.762808215346498,
+      "learning_rate": 6.727892972686861e-06,
+      "loss": 0.7984,
+      "step": 3212
+    },
+    {
+      "epoch": 0.6181223547518276,
+      "grad_norm": 0.8160510292083915,
+      "learning_rate": 6.722005761375873e-06,
+      "loss": 0.351,
+      "step": 3213
+    },
+    {
+      "epoch": 0.6183147364370912,
+      "grad_norm": 1.2052020663836724,
+      "learning_rate": 6.716119822695111e-06,
+      "loss": 0.8329,
+      "step": 3214
+    },
+    {
+      "epoch": 0.6185071181223547,
+      "grad_norm": 1.196037055737055,
+      "learning_rate": 6.710235158929703e-06,
+      "loss": 0.861,
+      "step": 3215
+    },
+    {
+      "epoch": 0.6186994998076183,
+      "grad_norm": 1.1815197137140834,
+      "learning_rate": 6.704351772364274e-06,
+      "loss": 0.8481,
+      "step": 3216
+    },
+    {
+      "epoch": 0.6188918814928819,
+      "grad_norm": 1.3144471901099744,
+      "learning_rate": 6.698469665282958e-06,
+      "loss": 0.9132,
+      "step": 3217
+    },
+    {
+      "epoch": 0.6190842631781455,
+      "grad_norm": 2.835302131610055,
+      "learning_rate": 6.692588839969397e-06,
+      "loss": 0.84,
+      "step": 3218
+    },
+    {
+      "epoch": 0.619276644863409,
+      "grad_norm": 1.5234237966968545,
+      "learning_rate": 6.6867092987067214e-06,
+      "loss": 0.8832,
+      "step": 3219
+    },
+    {
+      "epoch": 0.6194690265486725,
+      "grad_norm": 1.2784677877657933,
+      "learning_rate": 6.680831043777579e-06,
+      "loss": 0.8647,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6196614082339361,
+      "grad_norm": 1.1033841884914788,
+      "learning_rate": 6.674954077464108e-06,
+      "loss": 0.8818,
+      "step": 3221
+    },
+    {
+      "epoch": 0.6198537899191997,
+      "grad_norm": 1.1454066840944133,
+      "learning_rate": 6.6690784020479484e-06,
+      "loss": 0.8592,
+      "step": 3222
+    },
+    {
+      "epoch": 0.6200461716044633,
+      "grad_norm": 1.2513506336195106,
+      "learning_rate": 6.6632040198102364e-06,
+      "loss": 0.8225,
+      "step": 3223
+    },
+    {
+      "epoch": 0.6202385532897268,
+      "grad_norm": 1.1777136553554142,
+      "learning_rate": 6.657330933031619e-06,
+      "loss": 0.8496,
+      "step": 3224
+    },
+    {
+      "epoch": 0.6204309349749904,
+      "grad_norm": 1.4897171342613265,
+      "learning_rate": 6.651459143992221e-06,
+      "loss": 0.912,
+      "step": 3225
+    },
+    {
+      "epoch": 0.6206233166602539,
+      "grad_norm": 1.2136931999007476,
+      "learning_rate": 6.645588654971677e-06,
+      "loss": 0.8775,
+      "step": 3226
+    },
+    {
+      "epoch": 0.6208156983455175,
+      "grad_norm": 1.179321828206374,
+      "learning_rate": 6.639719468249115e-06,
+      "loss": 0.8681,
+      "step": 3227
+    },
+    {
+      "epoch": 0.6210080800307811,
+      "grad_norm": 1.2366856233153471,
+      "learning_rate": 6.633851586103153e-06,
+      "loss": 0.8491,
+      "step": 3228
+    },
+    {
+      "epoch": 0.6212004617160446,
+      "grad_norm": 1.442071290515258,
+      "learning_rate": 6.627985010811903e-06,
+      "loss": 0.8366,
+      "step": 3229
+    },
+    {
+      "epoch": 0.6213928434013082,
+      "grad_norm": 1.216081231512137,
+      "learning_rate": 6.622119744652977e-06,
+      "loss": 0.8349,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6215852250865718,
+      "grad_norm": 1.6643864480186916,
+      "learning_rate": 6.616255789903467e-06,
+      "loss": 0.8084,
+      "step": 3231
+    },
+    {
+      "epoch": 0.6217776067718354,
+      "grad_norm": 1.5248927014958356,
+      "learning_rate": 6.610393148839964e-06,
+      "loss": 0.9099,
+      "step": 3232
+    },
+    {
+      "epoch": 0.6219699884570988,
+      "grad_norm": 1.130656708977688,
+      "learning_rate": 6.6045318237385526e-06,
+      "loss": 0.9295,
+      "step": 3233
+    },
+    {
+      "epoch": 0.6221623701423624,
+      "grad_norm": 1.2147805276954258,
+      "learning_rate": 6.598671816874794e-06,
+      "loss": 0.8588,
+      "step": 3234
+    },
+    {
+      "epoch": 0.622354751827626,
+      "grad_norm": 1.5082281872694605,
+      "learning_rate": 6.5928131305237465e-06,
+      "loss": 0.8519,
+      "step": 3235
+    },
+    {
+      "epoch": 0.6225471335128896,
+      "grad_norm": 1.197114305619179,
+      "learning_rate": 6.586955766959958e-06,
+      "loss": 0.8707,
+      "step": 3236
+    },
+    {
+      "epoch": 0.6227395151981532,
+      "grad_norm": 1.6382929085171238,
+      "learning_rate": 6.581099728457451e-06,
+      "loss": 0.8409,
+      "step": 3237
+    },
+    {
+      "epoch": 0.6229318968834167,
+      "grad_norm": 1.5671764382695819,
+      "learning_rate": 6.5752450172897466e-06,
+      "loss": 0.897,
+      "step": 3238
+    },
+    {
+      "epoch": 0.6231242785686802,
+      "grad_norm": 1.3591317350964356,
+      "learning_rate": 6.569391635729847e-06,
+      "loss": 0.889,
+      "step": 3239
+    },
+    {
+      "epoch": 0.6233166602539438,
+      "grad_norm": 1.5019329255833018,
+      "learning_rate": 6.563539586050233e-06,
+      "loss": 0.8781,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6235090419392074,
+      "grad_norm": 1.438823873450202,
+      "learning_rate": 6.557688870522871e-06,
+      "loss": 0.9276,
+      "step": 3241
+    },
+    {
+      "epoch": 0.6237014236244709,
+      "grad_norm": 1.4182611107008491,
+      "learning_rate": 6.551839491419213e-06,
+      "loss": 0.8275,
+      "step": 3242
+    },
+    {
+      "epoch": 0.6238938053097345,
+      "grad_norm": 1.1848664985002917,
+      "learning_rate": 6.545991451010185e-06,
+      "loss": 0.8336,
+      "step": 3243
+    },
+    {
+      "epoch": 0.6240861869949981,
+      "grad_norm": 1.3279602814812326,
+      "learning_rate": 6.5401447515662065e-06,
+      "loss": 0.8146,
+      "step": 3244
+    },
+    {
+      "epoch": 0.6242785686802617,
+      "grad_norm": 1.323334647982992,
+      "learning_rate": 6.5342993953571556e-06,
+      "loss": 0.8086,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6244709503655252,
+      "grad_norm": 1.1628584125420522,
+      "learning_rate": 6.52845538465241e-06,
+      "loss": 0.7779,
+      "step": 3246
+    },
+    {
+      "epoch": 0.6246633320507887,
+      "grad_norm": 1.4673525817015776,
+      "learning_rate": 6.522612721720813e-06,
+      "loss": 0.9271,
+      "step": 3247
+    },
+    {
+      "epoch": 0.6248557137360523,
+      "grad_norm": 1.382233224878452,
+      "learning_rate": 6.5167714088306865e-06,
+      "loss": 0.8559,
+      "step": 3248
+    },
+    {
+      "epoch": 0.6250480954213159,
+      "grad_norm": 1.0762341970320242,
+      "learning_rate": 6.51093144824983e-06,
+      "loss": 0.8245,
+      "step": 3249
+    },
+    {
+      "epoch": 0.6252404771065795,
+      "grad_norm": 1.3229406814787206,
+      "learning_rate": 6.505092842245519e-06,
+      "loss": 0.8459,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6254328587918431,
+      "grad_norm": 1.3467375461585886,
+      "learning_rate": 6.499255593084498e-06,
+      "loss": 0.8614,
+      "step": 3251
+    },
+    {
+      "epoch": 0.6256252404771065,
+      "grad_norm": 1.2597370706922746,
+      "learning_rate": 6.493419703032991e-06,
+      "loss": 0.8816,
+      "step": 3252
+    },
+    {
+      "epoch": 0.6258176221623701,
+      "grad_norm": 1.3356797964650813,
+      "learning_rate": 6.487585174356691e-06,
+      "loss": 0.8772,
+      "step": 3253
+    },
+    {
+      "epoch": 0.6260100038476337,
+      "grad_norm": 1.6342007990343141,
+      "learning_rate": 6.481752009320761e-06,
+      "loss": 0.8637,
+      "step": 3254
+    },
+    {
+      "epoch": 0.6262023855328973,
+      "grad_norm": 1.228360016251028,
+      "learning_rate": 6.4759202101898366e-06,
+      "loss": 0.8388,
+      "step": 3255
+    },
+    {
+      "epoch": 0.6263947672181608,
+      "grad_norm": 1.492043865629626,
+      "learning_rate": 6.4700897792280285e-06,
+      "loss": 0.8919,
+      "step": 3256
+    },
+    {
+      "epoch": 0.6265871489034244,
+      "grad_norm": 1.33557572627027,
+      "learning_rate": 6.464260718698902e-06,
+      "loss": 0.8657,
+      "step": 3257
+    },
+    {
+      "epoch": 0.626779530588688,
+      "grad_norm": 1.1970345668214684,
+      "learning_rate": 6.458433030865503e-06,
+      "loss": 0.8448,
+      "step": 3258
+    },
+    {
+      "epoch": 0.6269719122739515,
+      "grad_norm": 1.4192052954437315,
+      "learning_rate": 6.452606717990346e-06,
+      "loss": 0.8303,
+      "step": 3259
+    },
+    {
+      "epoch": 0.6271642939592151,
+      "grad_norm": 1.0351028074314357,
+      "learning_rate": 6.4467817823354005e-06,
+      "loss": 0.9094,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6273566756444786,
+      "grad_norm": 1.2413131484261901,
+      "learning_rate": 6.440958226162104e-06,
+      "loss": 0.8103,
+      "step": 3261
+    },
+    {
+      "epoch": 0.6275490573297422,
+      "grad_norm": 1.2348407368010375,
+      "learning_rate": 6.43513605173137e-06,
+      "loss": 0.8253,
+      "step": 3262
+    },
+    {
+      "epoch": 0.6277414390150058,
+      "grad_norm": 1.2508712115218625,
+      "learning_rate": 6.4293152613035594e-06,
+      "loss": 0.8022,
+      "step": 3263
+    },
+    {
+      "epoch": 0.6279338207002694,
+      "grad_norm": 1.154676419587363,
+      "learning_rate": 6.4234958571385095e-06,
+      "loss": 0.8327,
+      "step": 3264
+    },
+    {
+      "epoch": 0.6281262023855329,
+      "grad_norm": 1.0367894072790358,
+      "learning_rate": 6.4176778414955075e-06,
+      "loss": 0.8317,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6283185840707964,
+      "grad_norm": 1.17846090799437,
+      "learning_rate": 6.4118612166333124e-06,
+      "loss": 0.779,
+      "step": 3266
+    },
+    {
+      "epoch": 0.62851096575606,
+      "grad_norm": 1.123265766137819,
+      "learning_rate": 6.4060459848101354e-06,
+      "loss": 0.8602,
+      "step": 3267
+    },
+    {
+      "epoch": 0.6287033474413236,
+      "grad_norm": 1.1745383200699184,
+      "learning_rate": 6.400232148283651e-06,
+      "loss": 0.8516,
+      "step": 3268
+    },
+    {
+      "epoch": 0.6288957291265872,
+      "grad_norm": 1.62873277210222,
+      "learning_rate": 6.3944197093109885e-06,
+      "loss": 0.837,
+      "step": 3269
+    },
+    {
+      "epoch": 0.6290881108118507,
+      "grad_norm": 1.281633978360826,
+      "learning_rate": 6.388608670148741e-06,
+      "loss": 0.7814,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6292804924971143,
+      "grad_norm": 1.3876676962819883,
+      "learning_rate": 6.38279903305295e-06,
+      "loss": 0.9005,
+      "step": 3271
+    },
+    {
+      "epoch": 0.6294728741823779,
+      "grad_norm": 1.1474298799448974,
+      "learning_rate": 6.376990800279119e-06,
+      "loss": 0.8097,
+      "step": 3272
+    },
+    {
+      "epoch": 0.6296652558676414,
+      "grad_norm": 1.313529504895298,
+      "learning_rate": 6.3711839740822035e-06,
+      "loss": 0.8811,
+      "step": 3273
+    },
+    {
+      "epoch": 0.6298576375529049,
+      "grad_norm": 1.1736137587613438,
+      "learning_rate": 6.3653785567166125e-06,
+      "loss": 0.8848,
+      "step": 3274
+    },
+    {
+      "epoch": 0.6300500192381685,
+      "grad_norm": 1.2978026971790748,
+      "learning_rate": 6.359574550436209e-06,
+      "loss": 0.8169,
+      "step": 3275
+    },
+    {
+      "epoch": 0.6302424009234321,
+      "grad_norm": 1.2016690233186458,
+      "learning_rate": 6.3537719574943105e-06,
+      "loss": 0.8547,
+      "step": 3276
+    },
+    {
+      "epoch": 0.6304347826086957,
+      "grad_norm": 1.349709264142455,
+      "learning_rate": 6.347970780143678e-06,
+      "loss": 0.8413,
+      "step": 3277
+    },
+    {
+      "epoch": 0.6306271642939593,
+      "grad_norm": 0.6349237726300795,
+      "learning_rate": 6.342171020636533e-06,
+      "loss": 0.332,
+      "step": 3278
+    },
+    {
+      "epoch": 0.6308195459792227,
+      "grad_norm": 1.2889177320153316,
+      "learning_rate": 6.336372681224543e-06,
+      "loss": 0.8244,
+      "step": 3279
+    },
+    {
+      "epoch": 0.6310119276644863,
+      "grad_norm": 1.21075444325665,
+      "learning_rate": 6.330575764158819e-06,
+      "loss": 0.8746,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6312043093497499,
+      "grad_norm": 1.4626632011669494,
+      "learning_rate": 6.324780271689923e-06,
+      "loss": 0.8739,
+      "step": 3281
+    },
+    {
+      "epoch": 0.6313966910350135,
+      "grad_norm": 1.2458485238813415,
+      "learning_rate": 6.318986206067872e-06,
+      "loss": 0.9189,
+      "step": 3282
+    },
+    {
+      "epoch": 0.631589072720277,
+      "grad_norm": 0.6478859932832178,
+      "learning_rate": 6.313193569542113e-06,
+      "loss": 0.356,
+      "step": 3283
+    },
+    {
+      "epoch": 0.6317814544055406,
+      "grad_norm": 1.2910878106211063,
+      "learning_rate": 6.30740236436155e-06,
+      "loss": 0.84,
+      "step": 3284
+    },
+    {
+      "epoch": 0.6319738360908042,
+      "grad_norm": 1.2059326535224175,
+      "learning_rate": 6.301612592774533e-06,
+      "loss": 0.8382,
+      "step": 3285
+    },
+    {
+      "epoch": 0.6321662177760677,
+      "grad_norm": 1.3459400601035763,
+      "learning_rate": 6.295824257028844e-06,
+      "loss": 0.8422,
+      "step": 3286
+    },
+    {
+      "epoch": 0.6323585994613313,
+      "grad_norm": 1.4647070051592779,
+      "learning_rate": 6.290037359371717e-06,
+      "loss": 0.8858,
+      "step": 3287
+    },
+    {
+      "epoch": 0.6325509811465948,
+      "grad_norm": 1.158048392052511,
+      "learning_rate": 6.284251902049827e-06,
+      "loss": 0.8801,
+      "step": 3288
+    },
+    {
+      "epoch": 0.6327433628318584,
+      "grad_norm": 1.1816346510708482,
+      "learning_rate": 6.278467887309283e-06,
+      "loss": 0.8222,
+      "step": 3289
+    },
+    {
+      "epoch": 0.632935744517122,
+      "grad_norm": 1.3760676773098421,
+      "learning_rate": 6.272685317395644e-06,
+      "loss": 0.8717,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6331281262023856,
+      "grad_norm": 1.3175433019335638,
+      "learning_rate": 6.266904194553896e-06,
+      "loss": 0.8354,
+      "step": 3291
+    },
+    {
+      "epoch": 0.633320507887649,
+      "grad_norm": 1.2378420211550654,
+      "learning_rate": 6.261124521028477e-06,
+      "loss": 0.8689,
+      "step": 3292
+    },
+    {
+      "epoch": 0.6335128895729126,
+      "grad_norm": 1.2630200723809557,
+      "learning_rate": 6.255346299063252e-06,
+      "loss": 0.83,
+      "step": 3293
+    },
+    {
+      "epoch": 0.6337052712581762,
+      "grad_norm": 1.3506124657252774,
+      "learning_rate": 6.249569530901525e-06,
+      "loss": 0.8438,
+      "step": 3294
+    },
+    {
+      "epoch": 0.6338976529434398,
+      "grad_norm": 1.6533161253715116,
+      "learning_rate": 6.243794218786034e-06,
+      "loss": 0.8392,
+      "step": 3295
+    },
+    {
+      "epoch": 0.6340900346287034,
+      "grad_norm": 1.3204699013963688,
+      "learning_rate": 6.238020364958964e-06,
+      "loss": 0.9035,
+      "step": 3296
+    },
+    {
+      "epoch": 0.6342824163139669,
+      "grad_norm": 1.2239301006155237,
+      "learning_rate": 6.232247971661912e-06,
+      "loss": 0.8542,
+      "step": 3297
+    },
+    {
+      "epoch": 0.6344747979992305,
+      "grad_norm": 1.5857436763500203,
+      "learning_rate": 6.2264770411359256e-06,
+      "loss": 0.8493,
+      "step": 3298
+    },
+    {
+      "epoch": 0.634667179684494,
+      "grad_norm": 1.388860906031681,
+      "learning_rate": 6.22070757562148e-06,
+      "loss": 0.8376,
+      "step": 3299
+    },
+    {
+      "epoch": 0.6348595613697576,
+      "grad_norm": 1.4683523262438365,
+      "learning_rate": 6.214939577358479e-06,
+      "loss": 0.8313,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6350519430550212,
+      "grad_norm": 1.3901957735563504,
+      "learning_rate": 6.209173048586253e-06,
+      "loss": 0.8829,
+      "step": 3301
+    },
+    {
+      "epoch": 0.6352443247402847,
+      "grad_norm": 1.1441582803482346,
+      "learning_rate": 6.203407991543577e-06,
+      "loss": 0.8726,
+      "step": 3302
+    },
+    {
+      "epoch": 0.6354367064255483,
+      "grad_norm": 1.186915587928379,
+      "learning_rate": 6.197644408468635e-06,
+      "loss": 0.9114,
+      "step": 3303
+    },
+    {
+      "epoch": 0.6356290881108119,
+      "grad_norm": 1.229041708296009,
+      "learning_rate": 6.191882301599052e-06,
+      "loss": 0.8741,
+      "step": 3304
+    },
+    {
+      "epoch": 0.6358214697960755,
+      "grad_norm": 1.3179759883390751,
+      "learning_rate": 6.186121673171882e-06,
+      "loss": 0.8558,
+      "step": 3305
+    },
+    {
+      "epoch": 0.6360138514813389,
+      "grad_norm": 1.15625897501879,
+      "learning_rate": 6.180362525423591e-06,
+      "loss": 0.8845,
+      "step": 3306
+    },
+    {
+      "epoch": 0.6362062331666025,
+      "grad_norm": 1.460265407482265,
+      "learning_rate": 6.174604860590081e-06,
+      "loss": 0.8789,
+      "step": 3307
+    },
+    {
+      "epoch": 0.6363986148518661,
+      "grad_norm": 1.5044544693922297,
+      "learning_rate": 6.168848680906678e-06,
+      "loss": 0.8576,
+      "step": 3308
+    },
+    {
+      "epoch": 0.6365909965371297,
+      "grad_norm": 1.3252175474996464,
+      "learning_rate": 6.163093988608127e-06,
+      "loss": 0.8351,
+      "step": 3309
+    },
+    {
+      "epoch": 0.6367833782223933,
+      "grad_norm": 1.4896304337891275,
+      "learning_rate": 6.157340785928595e-06,
+      "loss": 0.8202,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6369757599076568,
+      "grad_norm": 1.227345849897146,
+      "learning_rate": 6.151589075101681e-06,
+      "loss": 0.8522,
+      "step": 3311
+    },
+    {
+      "epoch": 0.6371681415929203,
+      "grad_norm": 1.4815805986454995,
+      "learning_rate": 6.145838858360391e-06,
+      "loss": 0.8364,
+      "step": 3312
+    },
+    {
+      "epoch": 0.6373605232781839,
+      "grad_norm": 1.366656274754096,
+      "learning_rate": 6.140090137937158e-06,
+      "loss": 0.8319,
+      "step": 3313
+    },
+    {
+      "epoch": 0.6375529049634475,
+      "grad_norm": 1.2648679341564024,
+      "learning_rate": 6.134342916063838e-06,
+      "loss": 0.8659,
+      "step": 3314
+    },
+    {
+      "epoch": 0.637745286648711,
+      "grad_norm": 1.4404059529583308,
+      "learning_rate": 6.128597194971691e-06,
+      "loss": 0.8751,
+      "step": 3315
+    },
+    {
+      "epoch": 0.6379376683339746,
+      "grad_norm": 1.2667823577629376,
+      "learning_rate": 6.122852976891413e-06,
+      "loss": 0.8834,
+      "step": 3316
+    },
+    {
+      "epoch": 0.6381300500192382,
+      "grad_norm": 1.5075466986770378,
+      "learning_rate": 6.117110264053101e-06,
+      "loss": 0.8226,
+      "step": 3317
+    },
+    {
+      "epoch": 0.6383224317045018,
+      "grad_norm": 1.582979435280943,
+      "learning_rate": 6.111369058686276e-06,
+      "loss": 0.7994,
+      "step": 3318
+    },
+    {
+      "epoch": 0.6385148133897653,
+      "grad_norm": 1.3205309718726486,
+      "learning_rate": 6.105629363019875e-06,
+      "loss": 0.8094,
+      "step": 3319
+    },
+    {
+      "epoch": 0.6387071950750288,
+      "grad_norm": 1.4348049884159078,
+      "learning_rate": 6.099891179282242e-06,
+      "loss": 0.726,
+      "step": 3320
+    },
+    {
+      "epoch": 0.6388995767602924,
+      "grad_norm": 0.6743362843691146,
+      "learning_rate": 6.094154509701133e-06,
+      "loss": 0.3245,
+      "step": 3321
+    },
+    {
+      "epoch": 0.639091958445556,
+      "grad_norm": 1.2921252968935037,
+      "learning_rate": 6.088419356503732e-06,
+      "loss": 0.7907,
+      "step": 3322
+    },
+    {
+      "epoch": 0.6392843401308196,
+      "grad_norm": 1.2025791516689501,
+      "learning_rate": 6.082685721916612e-06,
+      "loss": 0.8388,
+      "step": 3323
+    },
+    {
+      "epoch": 0.6394767218160831,
+      "grad_norm": 1.3322772162002106,
+      "learning_rate": 6.076953608165772e-06,
+      "loss": 0.9395,
+      "step": 3324
+    },
+    {
+      "epoch": 0.6396691035013466,
+      "grad_norm": 1.5386114066166265,
+      "learning_rate": 6.07122301747662e-06,
+      "loss": 0.8285,
+      "step": 3325
+    },
+    {
+      "epoch": 0.6398614851866102,
+      "grad_norm": 2.7422437389300915,
+      "learning_rate": 6.065493952073961e-06,
+      "loss": 0.8866,
+      "step": 3326
+    },
+    {
+      "epoch": 0.6400538668718738,
+      "grad_norm": 1.343071436562389,
+      "learning_rate": 6.0597664141820176e-06,
+      "loss": 0.88,
+      "step": 3327
+    },
+    {
+      "epoch": 0.6402462485571374,
+      "grad_norm": 1.433114104274912,
+      "learning_rate": 6.054040406024422e-06,
+      "loss": 0.8176,
+      "step": 3328
+    },
+    {
+      "epoch": 0.6404386302424009,
+      "grad_norm": 1.3465762050183139,
+      "learning_rate": 6.0483159298242e-06,
+      "loss": 0.9078,
+      "step": 3329
+    },
+    {
+      "epoch": 0.6406310119276645,
+      "grad_norm": 1.4302754132346989,
+      "learning_rate": 6.042592987803796e-06,
+      "loss": 0.8081,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6408233936129281,
+      "grad_norm": 1.2404211683267885,
+      "learning_rate": 6.036871582185054e-06,
+      "loss": 0.8493,
+      "step": 3331
+    },
+    {
+      "epoch": 0.6410157752981916,
+      "grad_norm": 1.2547816589940286,
+      "learning_rate": 6.031151715189217e-06,
+      "loss": 0.8136,
+      "step": 3332
+    },
+    {
+      "epoch": 0.6412081569834551,
+      "grad_norm": 1.6184797139129368,
+      "learning_rate": 6.025433389036935e-06,
+      "loss": 0.8335,
+      "step": 3333
+    },
+    {
+      "epoch": 0.6414005386687187,
+      "grad_norm": 1.3104675628180393,
+      "learning_rate": 6.019716605948261e-06,
+      "loss": 0.8805,
+      "step": 3334
+    },
+    {
+      "epoch": 0.6415929203539823,
+      "grad_norm": 1.472365972068826,
+      "learning_rate": 6.014001368142643e-06,
+      "loss": 0.9227,
+      "step": 3335
+    },
+    {
+      "epoch": 0.6417853020392459,
+      "grad_norm": 2.2927373742786794,
+      "learning_rate": 6.008287677838937e-06,
+      "loss": 0.7514,
+      "step": 3336
+    },
+    {
+      "epoch": 0.6419776837245095,
+      "grad_norm": 1.2133889684695653,
+      "learning_rate": 6.002575537255395e-06,
+      "loss": 0.9089,
+      "step": 3337
+    },
+    {
+      "epoch": 0.642170065409773,
+      "grad_norm": 1.2590572748829294,
+      "learning_rate": 5.996864948609662e-06,
+      "loss": 0.8747,
+      "step": 3338
+    },
+    {
+      "epoch": 0.6423624470950365,
+      "grad_norm": 0.6274985627682709,
+      "learning_rate": 5.9911559141187924e-06,
+      "loss": 0.359,
+      "step": 3339
+    },
+    {
+      "epoch": 0.6425548287803001,
+      "grad_norm": 1.1613398869547653,
+      "learning_rate": 5.9854484359992235e-06,
+      "loss": 0.8714,
+      "step": 3340
+    },
+    {
+      "epoch": 0.6427472104655637,
+      "grad_norm": 1.2470085986601191,
+      "learning_rate": 5.979742516466793e-06,
+      "loss": 0.853,
+      "step": 3341
+    },
+    {
+      "epoch": 0.6429395921508272,
+      "grad_norm": 1.7137083255004317,
+      "learning_rate": 5.974038157736746e-06,
+      "loss": 0.8908,
+      "step": 3342
+    },
+    {
+      "epoch": 0.6431319738360908,
+      "grad_norm": 1.231372983854548,
+      "learning_rate": 5.968335362023697e-06,
+      "loss": 0.8231,
+      "step": 3343
+    },
+    {
+      "epoch": 0.6433243555213544,
+      "grad_norm": 1.2116128667639003,
+      "learning_rate": 5.962634131541676e-06,
+      "loss": 0.8953,
+      "step": 3344
+    },
+    {
+      "epoch": 0.643516737206618,
+      "grad_norm": 1.431656641815401,
+      "learning_rate": 5.956934468504101e-06,
+      "loss": 0.8645,
+      "step": 3345
+    },
+    {
+      "epoch": 0.6437091188918815,
+      "grad_norm": 1.3819646233801355,
+      "learning_rate": 5.951236375123768e-06,
+      "loss": 0.8726,
+      "step": 3346
+    },
+    {
+      "epoch": 0.643901500577145,
+      "grad_norm": 1.37127535154856,
+      "learning_rate": 5.945539853612876e-06,
+      "loss": 0.8709,
+      "step": 3347
+    },
+    {
+      "epoch": 0.6440938822624086,
+      "grad_norm": 1.3950555867771637,
+      "learning_rate": 5.939844906183016e-06,
+      "loss": 0.9258,
+      "step": 3348
+    },
+    {
+      "epoch": 0.6442862639476722,
+      "grad_norm": 1.3566328913375636,
+      "learning_rate": 5.934151535045156e-06,
+      "loss": 0.8716,
+      "step": 3349
+    },
+    {
+      "epoch": 0.6444786456329358,
+      "grad_norm": 1.4470077137973978,
+      "learning_rate": 5.92845974240966e-06,
+      "loss": 0.7959,
+      "step": 3350
+    },
+    {
+      "epoch": 0.6446710273181993,
+      "grad_norm": 1.2768002166875039,
+      "learning_rate": 5.922769530486283e-06,
+      "loss": 0.8497,
+      "step": 3351
+    },
+    {
+      "epoch": 0.6448634090034628,
+      "grad_norm": 1.1106781882950612,
+      "learning_rate": 5.917080901484156e-06,
+      "loss": 0.8026,
+      "step": 3352
+    },
+    {
+      "epoch": 0.6450557906887264,
+      "grad_norm": 1.3742801217762777,
+      "learning_rate": 5.9113938576118e-06,
+      "loss": 0.796,
+      "step": 3353
+    },
+    {
+      "epoch": 0.64524817237399,
+      "grad_norm": 1.303367314185178,
+      "learning_rate": 5.905708401077128e-06,
+      "loss": 0.884,
+      "step": 3354
+    },
+    {
+      "epoch": 0.6454405540592536,
+      "grad_norm": 1.3958976811041288,
+      "learning_rate": 5.900024534087421e-06,
+      "loss": 0.8603,
+      "step": 3355
+    },
+    {
+      "epoch": 0.6456329357445171,
+      "grad_norm": 1.306952724754636,
+      "learning_rate": 5.894342258849355e-06,
+      "loss": 0.8505,
+      "step": 3356
+    },
+    {
+      "epoch": 0.6458253174297807,
+      "grad_norm": 1.5082424506945615,
+      "learning_rate": 5.88866157756899e-06,
+      "loss": 0.8231,
+      "step": 3357
+    },
+    {
+      "epoch": 0.6460176991150443,
+      "grad_norm": 1.1833025101455048,
+      "learning_rate": 5.882982492451757e-06,
+      "loss": 0.8747,
+      "step": 3358
+    },
+    {
+      "epoch": 0.6462100808003078,
+      "grad_norm": 2.426611522710653,
+      "learning_rate": 5.877305005702471e-06,
+      "loss": 0.8276,
+      "step": 3359
+    },
+    {
+      "epoch": 0.6464024624855714,
+      "grad_norm": 1.7767568576547947,
+      "learning_rate": 5.871629119525335e-06,
+      "loss": 0.9361,
+      "step": 3360
+    },
+    {
+      "epoch": 0.6465948441708349,
+      "grad_norm": 1.2575330160867524,
+      "learning_rate": 5.865954836123915e-06,
+      "loss": 0.8033,
+      "step": 3361
+    },
+    {
+      "epoch": 0.6467872258560985,
+      "grad_norm": 1.3728728866830588,
+      "learning_rate": 5.860282157701167e-06,
+      "loss": 0.842,
+      "step": 3362
+    },
+    {
+      "epoch": 0.6469796075413621,
+      "grad_norm": 1.518457052784656,
+      "learning_rate": 5.854611086459423e-06,
+      "loss": 0.8636,
+      "step": 3363
+    },
+    {
+      "epoch": 0.6471719892266257,
+      "grad_norm": 1.4076218092325155,
+      "learning_rate": 5.8489416246003814e-06,
+      "loss": 0.8544,
+      "step": 3364
+    },
+    {
+      "epoch": 0.6473643709118891,
+      "grad_norm": 1.8199850227022403,
+      "learning_rate": 5.8432737743251315e-06,
+      "loss": 0.8589,
+      "step": 3365
+    },
+    {
+      "epoch": 0.6475567525971527,
+      "grad_norm": 1.4466751867180638,
+      "learning_rate": 5.8376075378341194e-06,
+      "loss": 0.8489,
+      "step": 3366
+    },
+    {
+      "epoch": 0.6477491342824163,
+      "grad_norm": 1.556634994952425,
+      "learning_rate": 5.831942917327172e-06,
+      "loss": 0.8525,
+      "step": 3367
+    },
+    {
+      "epoch": 0.6479415159676799,
+      "grad_norm": 1.1106249423936647,
+      "learning_rate": 5.826279915003503e-06,
+      "loss": 0.8436,
+      "step": 3368
+    },
+    {
+      "epoch": 0.6481338976529435,
+      "grad_norm": 1.2573983872940753,
+      "learning_rate": 5.8206185330616725e-06,
+      "loss": 0.8693,
+      "step": 3369
+    },
+    {
+      "epoch": 0.648326279338207,
+      "grad_norm": 1.3048936051511895,
+      "learning_rate": 5.814958773699625e-06,
+      "loss": 0.8859,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6485186610234706,
+      "grad_norm": 1.2630172445398697,
+      "learning_rate": 5.809300639114683e-06,
+      "loss": 0.8,
+      "step": 3371
+    },
+    {
+      "epoch": 0.6487110427087341,
+      "grad_norm": 1.4653176855560237,
+      "learning_rate": 5.803644131503516e-06,
+      "loss": 0.876,
+      "step": 3372
+    },
+    {
+      "epoch": 0.6489034243939977,
+      "grad_norm": 1.8961822560725596,
+      "learning_rate": 5.797989253062186e-06,
+      "loss": 0.867,
+      "step": 3373
+    },
+    {
+      "epoch": 0.6490958060792612,
+      "grad_norm": 1.1806878440011972,
+      "learning_rate": 5.792336005986105e-06,
+      "loss": 0.8839,
+      "step": 3374
+    },
+    {
+      "epoch": 0.6492881877645248,
+      "grad_norm": 1.3210632763821633,
+      "learning_rate": 5.786684392470064e-06,
+      "loss": 0.8109,
+      "step": 3375
+    },
+    {
+      "epoch": 0.6494805694497884,
+      "grad_norm": 1.973998867411895,
+      "learning_rate": 5.781034414708208e-06,
+      "loss": 0.8515,
+      "step": 3376
+    },
+    {
+      "epoch": 0.649672951135052,
+      "grad_norm": 1.7702955525510873,
+      "learning_rate": 5.775386074894058e-06,
+      "loss": 0.8398,
+      "step": 3377
+    },
+    {
+      "epoch": 0.6498653328203156,
+      "grad_norm": 1.4383864531306678,
+      "learning_rate": 5.769739375220489e-06,
+      "loss": 0.85,
+      "step": 3378
+    },
+    {
+      "epoch": 0.650057714505579,
+      "grad_norm": 1.4772656442421752,
+      "learning_rate": 5.7640943178797445e-06,
+      "loss": 0.8763,
+      "step": 3379
+    },
+    {
+      "epoch": 0.6502500961908426,
+      "grad_norm": 1.634472351838974,
+      "learning_rate": 5.7584509050634395e-06,
+      "loss": 0.8189,
+      "step": 3380
+    },
+    {
+      "epoch": 0.6504424778761062,
+      "grad_norm": 1.1934873821001069,
+      "learning_rate": 5.752809138962525e-06,
+      "loss": 0.8303,
+      "step": 3381
+    },
+    {
+      "epoch": 0.6506348595613698,
+      "grad_norm": 1.452538818768777,
+      "learning_rate": 5.747169021767342e-06,
+      "loss": 0.8245,
+      "step": 3382
+    },
+    {
+      "epoch": 0.6508272412466333,
+      "grad_norm": 1.2263320744175057,
+      "learning_rate": 5.7415305556675805e-06,
+      "loss": 0.872,
+      "step": 3383
+    },
+    {
+      "epoch": 0.6510196229318969,
+      "grad_norm": 1.303630527470766,
+      "learning_rate": 5.73589374285227e-06,
+      "loss": 0.8286,
+      "step": 3384
+    },
+    {
+      "epoch": 0.6512120046171604,
+      "grad_norm": 1.844321956889208,
+      "learning_rate": 5.730258585509832e-06,
+      "loss": 0.8125,
+      "step": 3385
+    },
+    {
+      "epoch": 0.651404386302424,
+      "grad_norm": 1.6977390535277022,
+      "learning_rate": 5.724625085828022e-06,
+      "loss": 0.8202,
+      "step": 3386
+    },
+    {
+      "epoch": 0.6515967679876876,
+      "grad_norm": 1.3429858126528587,
+      "learning_rate": 5.718993245993958e-06,
+      "loss": 0.8954,
+      "step": 3387
+    },
+    {
+      "epoch": 0.6517891496729511,
+      "grad_norm": 1.4300427928556894,
+      "learning_rate": 5.713363068194115e-06,
+      "loss": 0.8472,
+      "step": 3388
+    },
+    {
+      "epoch": 0.6519815313582147,
+      "grad_norm": 1.2587964506605027,
+      "learning_rate": 5.7077345546143235e-06,
+      "loss": 0.7994,
+      "step": 3389
+    },
+    {
+      "epoch": 0.6521739130434783,
+      "grad_norm": 1.6040019098130949,
+      "learning_rate": 5.702107707439766e-06,
+      "loss": 0.8667,
+      "step": 3390
+    },
+    {
+      "epoch": 0.6523662947287419,
+      "grad_norm": 1.4287018331888908,
+      "learning_rate": 5.6964825288549745e-06,
+      "loss": 0.8814,
+      "step": 3391
+    },
+    {
+      "epoch": 0.6525586764140053,
+      "grad_norm": 1.5787319863952156,
+      "learning_rate": 5.690859021043842e-06,
+      "loss": 0.7824,
+      "step": 3392
+    },
+    {
+      "epoch": 0.6527510580992689,
+      "grad_norm": 1.430907582204647,
+      "learning_rate": 5.685237186189601e-06,
+      "loss": 0.8012,
+      "step": 3393
+    },
+    {
+      "epoch": 0.6529434397845325,
+      "grad_norm": 1.4946173016405375,
+      "learning_rate": 5.679617026474853e-06,
+      "loss": 0.8215,
+      "step": 3394
+    },
+    {
+      "epoch": 0.6531358214697961,
+      "grad_norm": 1.2529287336377115,
+      "learning_rate": 5.673998544081527e-06,
+      "loss": 0.8382,
+      "step": 3395
+    },
+    {
+      "epoch": 0.6533282031550597,
+      "grad_norm": 1.5331134696323565,
+      "learning_rate": 5.6683817411909114e-06,
+      "loss": 0.8824,
+      "step": 3396
+    },
+    {
+      "epoch": 0.6535205848403232,
+      "grad_norm": 0.6109707258082765,
+      "learning_rate": 5.662766619983653e-06,
+      "loss": 0.3481,
+      "step": 3397
+    },
+    {
+      "epoch": 0.6537129665255867,
+      "grad_norm": 1.4846328535301758,
+      "learning_rate": 5.65715318263972e-06,
+      "loss": 0.8429,
+      "step": 3398
+    },
+    {
+      "epoch": 0.6539053482108503,
+      "grad_norm": 1.3041566591479166,
+      "learning_rate": 5.651541431338454e-06,
+      "loss": 0.8559,
+      "step": 3399
+    },
+    {
+      "epoch": 0.6540977298961139,
+      "grad_norm": 1.3141580514752285,
+      "learning_rate": 5.645931368258527e-06,
+      "loss": 0.8651,
+      "step": 3400
+    },
+    {
+      "epoch": 0.6542901115813774,
+      "grad_norm": 1.6038054969325268,
+      "learning_rate": 5.640322995577958e-06,
+      "loss": 0.8377,
+      "step": 3401
+    },
+    {
+      "epoch": 0.654482493266641,
+      "grad_norm": 1.2058164777700902,
+      "learning_rate": 5.634716315474109e-06,
+      "loss": 0.8737,
+      "step": 3402
+    },
+    {
+      "epoch": 0.6546748749519046,
+      "grad_norm": 1.4524130331668115,
+      "learning_rate": 5.629111330123689e-06,
+      "loss": 0.8033,
+      "step": 3403
+    },
+    {
+      "epoch": 0.6548672566371682,
+      "grad_norm": 1.5008747084734753,
+      "learning_rate": 5.623508041702743e-06,
+      "loss": 0.8172,
+      "step": 3404
+    },
+    {
+      "epoch": 0.6550596383224317,
+      "grad_norm": 1.3082601677378267,
+      "learning_rate": 5.617906452386659e-06,
+      "loss": 0.8521,
+      "step": 3405
+    },
+    {
+      "epoch": 0.6552520200076952,
+      "grad_norm": 1.5531543385381517,
+      "learning_rate": 5.612306564350179e-06,
+      "loss": 0.9085,
+      "step": 3406
+    },
+    {
+      "epoch": 0.6554444016929588,
+      "grad_norm": 1.5034044578269388,
+      "learning_rate": 5.6067083797673535e-06,
+      "loss": 0.8633,
+      "step": 3407
+    },
+    {
+      "epoch": 0.6556367833782224,
+      "grad_norm": 1.1755780358702577,
+      "learning_rate": 5.601111900811607e-06,
+      "loss": 0.8584,
+      "step": 3408
+    },
+    {
+      "epoch": 0.655829165063486,
+      "grad_norm": 1.3943314547640853,
+      "learning_rate": 5.595517129655681e-06,
+      "loss": 0.7897,
+      "step": 3409
+    },
+    {
+      "epoch": 0.6560215467487496,
+      "grad_norm": 1.369790482164943,
+      "learning_rate": 5.589924068471648e-06,
+      "loss": 0.8596,
+      "step": 3410
+    },
+    {
+      "epoch": 0.656213928434013,
+      "grad_norm": 1.4130340296958437,
+      "learning_rate": 5.58433271943094e-06,
+      "loss": 0.7807,
+      "step": 3411
+    },
+    {
+      "epoch": 0.6564063101192766,
+      "grad_norm": 1.4753889322275886,
+      "learning_rate": 5.578743084704306e-06,
+      "loss": 0.8431,
+      "step": 3412
+    },
+    {
+      "epoch": 0.6565986918045402,
+      "grad_norm": 1.349732554294072,
+      "learning_rate": 5.573155166461833e-06,
+      "loss": 0.8561,
+      "step": 3413
+    },
+    {
+      "epoch": 0.6567910734898038,
+      "grad_norm": 1.4451747618894315,
+      "learning_rate": 5.567568966872947e-06,
+      "loss": 0.8731,
+      "step": 3414
+    },
+    {
+      "epoch": 0.6569834551750673,
+      "grad_norm": 1.5195269974450534,
+      "learning_rate": 5.5619844881064e-06,
+      "loss": 0.8773,
+      "step": 3415
+    },
+    {
+      "epoch": 0.6571758368603309,
+      "grad_norm": 1.1979363725866654,
+      "learning_rate": 5.556401732330281e-06,
+      "loss": 0.8342,
+      "step": 3416
+    },
+    {
+      "epoch": 0.6573682185455945,
+      "grad_norm": 1.1611457017667743,
+      "learning_rate": 5.550820701712007e-06,
+      "loss": 0.8865,
+      "step": 3417
+    },
+    {
+      "epoch": 0.657560600230858,
+      "grad_norm": 1.2449614175432582,
+      "learning_rate": 5.545241398418326e-06,
+      "loss": 0.8545,
+      "step": 3418
+    },
+    {
+      "epoch": 0.6577529819161216,
+      "grad_norm": 1.2903873484326818,
+      "learning_rate": 5.539663824615312e-06,
+      "loss": 0.8173,
+      "step": 3419
+    },
+    {
+      "epoch": 0.6579453636013851,
+      "grad_norm": 1.5347532588973603,
+      "learning_rate": 5.534087982468384e-06,
+      "loss": 0.8414,
+      "step": 3420
+    },
+    {
+      "epoch": 0.6581377452866487,
+      "grad_norm": 1.2845830135950302,
+      "learning_rate": 5.5285138741422615e-06,
+      "loss": 0.8777,
+      "step": 3421
+    },
+    {
+      "epoch": 0.6583301269719123,
+      "grad_norm": 1.251172642143527,
+      "learning_rate": 5.522941501801008e-06,
+      "loss": 0.8761,
+      "step": 3422
+    },
+    {
+      "epoch": 0.6585225086571759,
+      "grad_norm": 1.5035461748459829,
+      "learning_rate": 5.517370867608021e-06,
+      "loss": 0.8072,
+      "step": 3423
+    },
+    {
+      "epoch": 0.6587148903424394,
+      "grad_norm": 1.3291696024884572,
+      "learning_rate": 5.511801973725997e-06,
+      "loss": 0.8339,
+      "step": 3424
+    },
+    {
+      "epoch": 0.6589072720277029,
+      "grad_norm": 1.4500191870947312,
+      "learning_rate": 5.506234822316983e-06,
+      "loss": 0.8416,
+      "step": 3425
+    },
+    {
+      "epoch": 0.6590996537129665,
+      "grad_norm": 1.2135615522212808,
+      "learning_rate": 5.500669415542336e-06,
+      "loss": 0.9294,
+      "step": 3426
+    },
+    {
+      "epoch": 0.6592920353982301,
+      "grad_norm": 1.3012850300904883,
+      "learning_rate": 5.495105755562738e-06,
+      "loss": 0.8439,
+      "step": 3427
+    },
+    {
+      "epoch": 0.6594844170834937,
+      "grad_norm": 1.4029424418427283,
+      "learning_rate": 5.4895438445381945e-06,
+      "loss": 0.8539,
+      "step": 3428
+    },
+    {
+      "epoch": 0.6596767987687572,
+      "grad_norm": 1.35908060359272,
+      "learning_rate": 5.48398368462803e-06,
+      "loss": 0.8487,
+      "step": 3429
+    },
+    {
+      "epoch": 0.6598691804540208,
+      "grad_norm": 1.3088306399472525,
+      "learning_rate": 5.4784252779908905e-06,
+      "loss": 0.8874,
+      "step": 3430
+    },
+    {
+      "epoch": 0.6600615621392844,
+      "grad_norm": 1.5081492114657136,
+      "learning_rate": 5.4728686267847354e-06,
+      "loss": 0.8711,
+      "step": 3431
+    },
+    {
+      "epoch": 0.6602539438245479,
+      "grad_norm": 1.2612093859650781,
+      "learning_rate": 5.467313733166863e-06,
+      "loss": 0.8501,
+      "step": 3432
+    },
+    {
+      "epoch": 0.6604463255098114,
+      "grad_norm": 1.5110521170295546,
+      "learning_rate": 5.461760599293855e-06,
+      "loss": 0.8553,
+      "step": 3433
+    },
+    {
+      "epoch": 0.660638707195075,
+      "grad_norm": 1.2059207440025028,
+      "learning_rate": 5.456209227321643e-06,
+      "loss": 0.8655,
+      "step": 3434
+    },
+    {
+      "epoch": 0.6608310888803386,
+      "grad_norm": 1.207082897711135,
+      "learning_rate": 5.450659619405458e-06,
+      "loss": 0.8314,
+      "step": 3435
+    },
+    {
+      "epoch": 0.6610234705656022,
+      "grad_norm": 1.646539278883033,
+      "learning_rate": 5.445111777699842e-06,
+      "loss": 0.8392,
+      "step": 3436
+    },
+    {
+      "epoch": 0.6612158522508658,
+      "grad_norm": 1.275553575004824,
+      "learning_rate": 5.439565704358667e-06,
+      "loss": 0.8238,
+      "step": 3437
+    },
+    {
+      "epoch": 0.6614082339361292,
+      "grad_norm": 1.253864669646419,
+      "learning_rate": 5.434021401535105e-06,
+      "loss": 0.8681,
+      "step": 3438
+    },
+    {
+      "epoch": 0.6616006156213928,
+      "grad_norm": 1.3899238692168103,
+      "learning_rate": 5.428478871381646e-06,
+      "loss": 0.8004,
+      "step": 3439
+    },
+    {
+      "epoch": 0.6617929973066564,
+      "grad_norm": 1.7939998185758024,
+      "learning_rate": 5.422938116050092e-06,
+      "loss": 0.8801,
+      "step": 3440
+    },
+    {
+      "epoch": 0.66198537899192,
+      "grad_norm": 1.2732218221685658,
+      "learning_rate": 5.417399137691552e-06,
+      "loss": 0.8828,
+      "step": 3441
+    },
+    {
+      "epoch": 0.6621777606771835,
+      "grad_norm": 0.6526999033056514,
+      "learning_rate": 5.411861938456453e-06,
+      "loss": 0.3726,
+      "step": 3442
+    },
+    {
+      "epoch": 0.6623701423624471,
+      "grad_norm": 1.4108333200074208,
+      "learning_rate": 5.406326520494522e-06,
+      "loss": 0.8324,
+      "step": 3443
+    },
+    {
+      "epoch": 0.6625625240477107,
+      "grad_norm": 1.5069895332377208,
+      "learning_rate": 5.400792885954802e-06,
+      "loss": 0.8223,
+      "step": 3444
+    },
+    {
+      "epoch": 0.6627549057329742,
+      "grad_norm": 1.3275840470916376,
+      "learning_rate": 5.395261036985635e-06,
+      "loss": 0.8818,
+      "step": 3445
+    },
+    {
+      "epoch": 0.6629472874182378,
+      "grad_norm": 1.4846227105181034,
+      "learning_rate": 5.389730975734686e-06,
+      "loss": 0.8596,
+      "step": 3446
+    },
+    {
+      "epoch": 0.6631396691035013,
+      "grad_norm": 1.2743329240622212,
+      "learning_rate": 5.384202704348902e-06,
+      "loss": 0.8528,
+      "step": 3447
+    },
+    {
+      "epoch": 0.6633320507887649,
+      "grad_norm": 1.2925345947883349,
+      "learning_rate": 5.378676224974557e-06,
+      "loss": 0.875,
+      "step": 3448
+    },
+    {
+      "epoch": 0.6635244324740285,
+      "grad_norm": 1.3477092045106873,
+      "learning_rate": 5.373151539757224e-06,
+      "loss": 0.8283,
+      "step": 3449
+    },
+    {
+      "epoch": 0.6637168141592921,
+      "grad_norm": 1.348031385041221,
+      "learning_rate": 5.367628650841761e-06,
+      "loss": 0.8566,
+      "step": 3450
+    },
+    {
+      "epoch": 0.6639091958445555,
+      "grad_norm": 1.415795841560053,
+      "learning_rate": 5.362107560372358e-06,
+      "loss": 0.8306,
+      "step": 3451
+    },
+    {
+      "epoch": 0.6641015775298191,
+      "grad_norm": 1.4833738582467222,
+      "learning_rate": 5.356588270492487e-06,
+      "loss": 0.8501,
+      "step": 3452
+    },
+    {
+      "epoch": 0.6642939592150827,
+      "grad_norm": 1.3690992819627577,
+      "learning_rate": 5.351070783344926e-06,
+      "loss": 0.8812,
+      "step": 3453
+    },
+    {
+      "epoch": 0.6644863409003463,
+      "grad_norm": 1.2123189690618728,
+      "learning_rate": 5.3455551010717545e-06,
+      "loss": 0.893,
+      "step": 3454
+    },
+    {
+      "epoch": 0.6646787225856099,
+      "grad_norm": 1.2899302448649828,
+      "learning_rate": 5.34004122581435e-06,
+      "loss": 0.885,
+      "step": 3455
+    },
+    {
+      "epoch": 0.6648711042708734,
+      "grad_norm": 1.3387316901105668,
+      "learning_rate": 5.334529159713389e-06,
+      "loss": 0.9237,
+      "step": 3456
+    },
+    {
+      "epoch": 0.665063485956137,
+      "grad_norm": 1.3531870690147738,
+      "learning_rate": 5.329018904908841e-06,
+      "loss": 0.8182,
+      "step": 3457
+    },
+    {
+      "epoch": 0.6652558676414005,
+      "grad_norm": 1.174039310006719,
+      "learning_rate": 5.323510463539989e-06,
+      "loss": 0.7621,
+      "step": 3458
+    },
+    {
+      "epoch": 0.6654482493266641,
+      "grad_norm": 1.2190424263967226,
+      "learning_rate": 5.318003837745382e-06,
+      "loss": 0.8677,
+      "step": 3459
+    },
+    {
+      "epoch": 0.6656406310119277,
+      "grad_norm": 1.6263986988858912,
+      "learning_rate": 5.3124990296628974e-06,
+      "loss": 0.8368,
+      "step": 3460
+    },
+    {
+      "epoch": 0.6658330126971912,
+      "grad_norm": 1.3527985933033106,
+      "learning_rate": 5.306996041429688e-06,
+      "loss": 0.9023,
+      "step": 3461
+    },
+    {
+      "epoch": 0.6660253943824548,
+      "grad_norm": 1.34528444132113,
+      "learning_rate": 5.301494875182192e-06,
+      "loss": 0.756,
+      "step": 3462
+    },
+    {
+      "epoch": 0.6662177760677184,
+      "grad_norm": 0.6612351818227679,
+      "learning_rate": 5.295995533056162e-06,
+      "loss": 0.3643,
+      "step": 3463
+    },
+    {
+      "epoch": 0.666410157752982,
+      "grad_norm": 1.2109143533613058,
+      "learning_rate": 5.290498017186631e-06,
+      "loss": 0.8869,
+      "step": 3464
+    },
+    {
+      "epoch": 0.6666025394382454,
+      "grad_norm": 1.2451593471666744,
+      "learning_rate": 5.2850023297079235e-06,
+      "loss": 0.8982,
+      "step": 3465
+    },
+    {
+      "epoch": 0.666794921123509,
+      "grad_norm": 1.2695503176676228,
+      "learning_rate": 5.279508472753654e-06,
+      "loss": 0.8531,
+      "step": 3466
+    },
+    {
+      "epoch": 0.6669873028087726,
+      "grad_norm": 1.3321327986597606,
+      "learning_rate": 5.274016448456725e-06,
+      "loss": 0.8748,
+      "step": 3467
+    },
+    {
+      "epoch": 0.6671796844940362,
+      "grad_norm": 1.1932506024350817,
+      "learning_rate": 5.2685262589493314e-06,
+      "loss": 0.8608,
+      "step": 3468
+    },
+    {
+      "epoch": 0.6673720661792998,
+      "grad_norm": 1.2936363601652847,
+      "learning_rate": 5.263037906362953e-06,
+      "loss": 0.8262,
+      "step": 3469
+    },
+    {
+      "epoch": 0.6675644478645633,
+      "grad_norm": 0.5780986979935478,
+      "learning_rate": 5.257551392828359e-06,
+      "loss": 0.3391,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6677568295498268,
+      "grad_norm": 1.2374250482271383,
+      "learning_rate": 5.252066720475597e-06,
+      "loss": 0.8491,
+      "step": 3471
+    },
+    {
+      "epoch": 0.6679492112350904,
+      "grad_norm": 1.2596216645410054,
+      "learning_rate": 5.246583891434018e-06,
+      "loss": 0.8726,
+      "step": 3472
+    },
+    {
+      "epoch": 0.668141592920354,
+      "grad_norm": 1.3652404637754776,
+      "learning_rate": 5.241102907832232e-06,
+      "loss": 0.8499,
+      "step": 3473
+    },
+    {
+      "epoch": 0.6683339746056175,
+      "grad_norm": 1.6777083942165611,
+      "learning_rate": 5.235623771798151e-06,
+      "loss": 0.858,
+      "step": 3474
+    },
+    {
+      "epoch": 0.6685263562908811,
+      "grad_norm": 1.1973242388908278,
+      "learning_rate": 5.23014648545897e-06,
+      "loss": 0.8597,
+      "step": 3475
+    },
+    {
+      "epoch": 0.6687187379761447,
+      "grad_norm": 1.3224870734643295,
+      "learning_rate": 5.224671050941146e-06,
+      "loss": 0.8747,
+      "step": 3476
+    },
+    {
+      "epoch": 0.6689111196614083,
+      "grad_norm": 1.361172711190235,
+      "learning_rate": 5.2191974703704425e-06,
+      "loss": 0.7964,
+      "step": 3477
+    },
+    {
+      "epoch": 0.6691035013466718,
+      "grad_norm": 1.660572546426119,
+      "learning_rate": 5.213725745871889e-06,
+      "loss": 0.8599,
+      "step": 3478
+    },
+    {
+      "epoch": 0.6692958830319353,
+      "grad_norm": 1.3203897752038487,
+      "learning_rate": 5.208255879569799e-06,
+      "loss": 0.7777,
+      "step": 3479
+    },
+    {
+      "epoch": 0.6694882647171989,
+      "grad_norm": 1.21805831976094,
+      "learning_rate": 5.20278787358776e-06,
+      "loss": 0.8454,
+      "step": 3480
+    },
+    {
+      "epoch": 0.6696806464024625,
+      "grad_norm": 1.092775105777749,
+      "learning_rate": 5.197321730048641e-06,
+      "loss": 0.8167,
+      "step": 3481
+    },
+    {
+      "epoch": 0.6698730280877261,
+      "grad_norm": 1.29135274605785,
+      "learning_rate": 5.1918574510745865e-06,
+      "loss": 0.8283,
+      "step": 3482
+    },
+    {
+      "epoch": 0.6700654097729896,
+      "grad_norm": 1.3941253007513594,
+      "learning_rate": 5.186395038787017e-06,
+      "loss": 0.8813,
+      "step": 3483
+    },
+    {
+      "epoch": 0.6702577914582531,
+      "grad_norm": 1.2760248993834535,
+      "learning_rate": 5.180934495306638e-06,
+      "loss": 0.8216,
+      "step": 3484
+    },
+    {
+      "epoch": 0.6704501731435167,
+      "grad_norm": 1.2842040420223413,
+      "learning_rate": 5.175475822753404e-06,
+      "loss": 0.8299,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6706425548287803,
+      "grad_norm": 1.5490448073987866,
+      "learning_rate": 5.170019023246574e-06,
+      "loss": 0.8654,
+      "step": 3486
+    },
+    {
+      "epoch": 0.6708349365140439,
+      "grad_norm": 1.2431190517854456,
+      "learning_rate": 5.16456409890466e-06,
+      "loss": 0.7988,
+      "step": 3487
+    },
+    {
+      "epoch": 0.6710273181993074,
+      "grad_norm": 1.2659165068689617,
+      "learning_rate": 5.159111051845451e-06,
+      "loss": 0.8588,
+      "step": 3488
+    },
+    {
+      "epoch": 0.671219699884571,
+      "grad_norm": 1.1441768810638453,
+      "learning_rate": 5.153659884186013e-06,
+      "loss": 0.8594,
+      "step": 3489
+    },
+    {
+      "epoch": 0.6714120815698346,
+      "grad_norm": 1.4998357746776658,
+      "learning_rate": 5.148210598042665e-06,
+      "loss": 0.8628,
+      "step": 3490
+    },
+    {
+      "epoch": 0.6716044632550981,
+      "grad_norm": 1.3083883401988907,
+      "learning_rate": 5.142763195531017e-06,
+      "loss": 0.8545,
+      "step": 3491
+    },
+    {
+      "epoch": 0.6717968449403616,
+      "grad_norm": 1.352945225123064,
+      "learning_rate": 5.137317678765939e-06,
+      "loss": 0.8337,
+      "step": 3492
+    },
+    {
+      "epoch": 0.6719892266256252,
+      "grad_norm": 1.5945894218249128,
+      "learning_rate": 5.131874049861563e-06,
+      "loss": 0.8457,
+      "step": 3493
+    },
+    {
+      "epoch": 0.6721816083108888,
+      "grad_norm": 1.358603784389523,
+      "learning_rate": 5.126432310931295e-06,
+      "loss": 0.8479,
+      "step": 3494
+    },
+    {
+      "epoch": 0.6723739899961524,
+      "grad_norm": 1.2110377816728173,
+      "learning_rate": 5.120992464087807e-06,
+      "loss": 0.8188,
+      "step": 3495
+    },
+    {
+      "epoch": 0.672566371681416,
+      "grad_norm": 1.3243634293528626,
+      "learning_rate": 5.115554511443033e-06,
+      "loss": 0.836,
+      "step": 3496
+    },
+    {
+      "epoch": 0.6727587533666795,
+      "grad_norm": 1.6079161401168445,
+      "learning_rate": 5.1101184551081705e-06,
+      "loss": 0.8451,
+      "step": 3497
+    },
+    {
+      "epoch": 0.672951135051943,
+      "grad_norm": 1.299546303866573,
+      "learning_rate": 5.104684297193694e-06,
+      "loss": 0.7906,
+      "step": 3498
+    },
+    {
+      "epoch": 0.6731435167372066,
+      "grad_norm": 1.4595213328089482,
+      "learning_rate": 5.099252039809317e-06,
+      "loss": 0.8944,
+      "step": 3499
+    },
+    {
+      "epoch": 0.6733358984224702,
+      "grad_norm": 1.1749017501634766,
+      "learning_rate": 5.09382168506404e-06,
+      "loss": 0.8632,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6735282801077337,
+      "grad_norm": 1.3902681883937074,
+      "learning_rate": 5.088393235066114e-06,
+      "loss": 0.8317,
+      "step": 3501
+    },
+    {
+      "epoch": 0.6737206617929973,
+      "grad_norm": 1.3338694639594122,
+      "learning_rate": 5.082966691923037e-06,
+      "loss": 0.8143,
+      "step": 3502
+    },
+    {
+      "epoch": 0.6739130434782609,
+      "grad_norm": 1.340298564085069,
+      "learning_rate": 5.077542057741592e-06,
+      "loss": 0.8047,
+      "step": 3503
+    },
+    {
+      "epoch": 0.6741054251635245,
+      "grad_norm": 1.342778710660999,
+      "learning_rate": 5.0721193346278066e-06,
+      "loss": 0.8458,
+      "step": 3504
+    },
+    {
+      "epoch": 0.674297806848788,
+      "grad_norm": 1.407731335596921,
+      "learning_rate": 5.066698524686966e-06,
+      "loss": 0.9127,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6744901885340515,
+      "grad_norm": 1.438444275129722,
+      "learning_rate": 5.061279630023618e-06,
+      "loss": 0.8527,
+      "step": 3506
+    },
+    {
+      "epoch": 0.6746825702193151,
+      "grad_norm": 1.3476899391795423,
+      "learning_rate": 5.055862652741562e-06,
+      "loss": 0.8576,
+      "step": 3507
+    },
+    {
+      "epoch": 0.6748749519045787,
+      "grad_norm": 1.463487839534693,
+      "learning_rate": 5.050447594943856e-06,
+      "loss": 0.8578,
+      "step": 3508
+    },
+    {
+      "epoch": 0.6750673335898423,
+      "grad_norm": 1.2709655003132132,
+      "learning_rate": 5.045034458732808e-06,
+      "loss": 0.8138,
+      "step": 3509
+    },
+    {
+      "epoch": 0.6752597152751059,
+      "grad_norm": 0.662689858117814,
+      "learning_rate": 5.0396232462099945e-06,
+      "loss": 0.3573,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6754520969603693,
+      "grad_norm": 1.2830271697423197,
+      "learning_rate": 5.034213959476222e-06,
+      "loss": 0.8844,
+      "step": 3511
+    },
+    {
+      "epoch": 0.6756444786456329,
+      "grad_norm": 1.491888691918646,
+      "learning_rate": 5.028806600631569e-06,
+      "loss": 0.8405,
+      "step": 3512
+    },
+    {
+      "epoch": 0.6758368603308965,
+      "grad_norm": 1.3588564124231701,
+      "learning_rate": 5.023401171775357e-06,
+      "loss": 0.805,
+      "step": 3513
+    },
+    {
+      "epoch": 0.6760292420161601,
+      "grad_norm": 1.9377606748456742,
+      "learning_rate": 5.017997675006161e-06,
+      "loss": 0.8833,
+      "step": 3514
+    },
+    {
+      "epoch": 0.6762216237014236,
+      "grad_norm": 1.3891564336453508,
+      "learning_rate": 5.012596112421806e-06,
+      "loss": 0.8292,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6764140053866872,
+      "grad_norm": 1.3955437617284003,
+      "learning_rate": 5.007196486119355e-06,
+      "loss": 0.8659,
+      "step": 3516
+    },
+    {
+      "epoch": 0.6766063870719508,
+      "grad_norm": 1.2593107380849053,
+      "learning_rate": 5.001798798195136e-06,
+      "loss": 0.8113,
+      "step": 3517
+    },
+    {
+      "epoch": 0.6767987687572143,
+      "grad_norm": 1.5090209985746357,
+      "learning_rate": 4.996403050744719e-06,
+      "loss": 0.8831,
+      "step": 3518
+    },
+    {
+      "epoch": 0.6769911504424779,
+      "grad_norm": 1.3713379322022734,
+      "learning_rate": 4.991009245862917e-06,
+      "loss": 0.8674,
+      "step": 3519
+    },
+    {
+      "epoch": 0.6771835321277414,
+      "grad_norm": 1.3647249892198612,
+      "learning_rate": 4.985617385643789e-06,
+      "loss": 0.9156,
+      "step": 3520
+    },
+    {
+      "epoch": 0.677375913813005,
+      "grad_norm": 1.4510791419889972,
+      "learning_rate": 4.980227472180643e-06,
+      "loss": 0.7873,
+      "step": 3521
+    },
+    {
+      "epoch": 0.6775682954982686,
+      "grad_norm": 1.378675982904595,
+      "learning_rate": 4.974839507566027e-06,
+      "loss": 0.8083,
+      "step": 3522
+    },
+    {
+      "epoch": 0.6777606771835322,
+      "grad_norm": 1.868591212362591,
+      "learning_rate": 4.969453493891733e-06,
+      "loss": 0.8374,
+      "step": 3523
+    },
+    {
+      "epoch": 0.6779530588687956,
+      "grad_norm": 1.4030231167165021,
+      "learning_rate": 4.9640694332488075e-06,
+      "loss": 0.7841,
+      "step": 3524
+    },
+    {
+      "epoch": 0.6781454405540592,
+      "grad_norm": 1.3568246416232301,
+      "learning_rate": 4.958687327727511e-06,
+      "loss": 0.8722,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6783378222393228,
+      "grad_norm": 1.227871470331587,
+      "learning_rate": 4.953307179417376e-06,
+      "loss": 0.807,
+      "step": 3526
+    },
+    {
+      "epoch": 0.6785302039245864,
+      "grad_norm": 1.6906108373602753,
+      "learning_rate": 4.947928990407156e-06,
+      "loss": 0.8345,
+      "step": 3527
+    },
+    {
+      "epoch": 0.67872258560985,
+      "grad_norm": 1.2789367949157529,
+      "learning_rate": 4.94255276278485e-06,
+      "loss": 0.8748,
+      "step": 3528
+    },
+    {
+      "epoch": 0.6789149672951135,
+      "grad_norm": 0.5950076061543825,
+      "learning_rate": 4.937178498637696e-06,
+      "loss": 0.3191,
+      "step": 3529
+    },
+    {
+      "epoch": 0.6791073489803771,
+      "grad_norm": 1.4218799614391415,
+      "learning_rate": 4.931806200052165e-06,
+      "loss": 0.8603,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6792997306656406,
+      "grad_norm": 2.2335150815377394,
+      "learning_rate": 4.926435869113971e-06,
+      "loss": 0.8603,
+      "step": 3531
+    },
+    {
+      "epoch": 0.6794921123509042,
+      "grad_norm": 0.614500345782926,
+      "learning_rate": 4.92106750790806e-06,
+      "loss": 0.3268,
+      "step": 3532
+    },
+    {
+      "epoch": 0.6796844940361677,
+      "grad_norm": 1.623088206274375,
+      "learning_rate": 4.915701118518616e-06,
+      "loss": 0.867,
+      "step": 3533
+    },
+    {
+      "epoch": 0.6798768757214313,
+      "grad_norm": 1.2938384095976392,
+      "learning_rate": 4.910336703029055e-06,
+      "loss": 0.8382,
+      "step": 3534
+    },
+    {
+      "epoch": 0.6800692574066949,
+      "grad_norm": 1.2906836829428572,
+      "learning_rate": 4.904974263522025e-06,
+      "loss": 0.8759,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6802616390919585,
+      "grad_norm": 1.3047028080838465,
+      "learning_rate": 4.899613802079419e-06,
+      "loss": 0.8164,
+      "step": 3536
+    },
+    {
+      "epoch": 0.6804540207772221,
+      "grad_norm": 1.5385494492711236,
+      "learning_rate": 4.8942553207823395e-06,
+      "loss": 0.8863,
+      "step": 3537
+    },
+    {
+      "epoch": 0.6806464024624855,
+      "grad_norm": 1.3261792305972995,
+      "learning_rate": 4.888898821711144e-06,
+      "loss": 0.812,
+      "step": 3538
+    },
+    {
+      "epoch": 0.6808387841477491,
+      "grad_norm": 1.2489766641499263,
+      "learning_rate": 4.883544306945407e-06,
+      "loss": 0.8531,
+      "step": 3539
+    },
+    {
+      "epoch": 0.6810311658330127,
+      "grad_norm": 1.5063013888397658,
+      "learning_rate": 4.878191778563934e-06,
+      "loss": 0.8438,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6812235475182763,
+      "grad_norm": 1.181247227132492,
+      "learning_rate": 4.872841238644766e-06,
+      "loss": 0.8264,
+      "step": 3541
+    },
+    {
+      "epoch": 0.6814159292035398,
+      "grad_norm": 1.1961824146074742,
+      "learning_rate": 4.867492689265154e-06,
+      "loss": 0.8617,
+      "step": 3542
+    },
+    {
+      "epoch": 0.6816083108888034,
+      "grad_norm": 1.3835048348073435,
+      "learning_rate": 4.8621461325016015e-06,
+      "loss": 0.8285,
+      "step": 3543
+    },
+    {
+      "epoch": 0.681800692574067,
+      "grad_norm": 1.1469189266294657,
+      "learning_rate": 4.856801570429822e-06,
+      "loss": 0.8923,
+      "step": 3544
+    },
+    {
+      "epoch": 0.6819930742593305,
+      "grad_norm": 1.223543018137829,
+      "learning_rate": 4.851459005124759e-06,
+      "loss": 0.8663,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6821854559445941,
+      "grad_norm": 1.5989212109518252,
+      "learning_rate": 4.846118438660578e-06,
+      "loss": 0.8189,
+      "step": 3546
+    },
+    {
+      "epoch": 0.6823778376298576,
+      "grad_norm": 1.4156899867193,
+      "learning_rate": 4.840779873110675e-06,
+      "loss": 0.8745,
+      "step": 3547
+    },
+    {
+      "epoch": 0.6825702193151212,
+      "grad_norm": 0.6581077917591834,
+      "learning_rate": 4.83544331054766e-06,
+      "loss": 0.3036,
+      "step": 3548
+    },
+    {
+      "epoch": 0.6827626010003848,
+      "grad_norm": 1.304915819354955,
+      "learning_rate": 4.83010875304337e-06,
+      "loss": 0.8476,
+      "step": 3549
+    },
+    {
+      "epoch": 0.6829549826856484,
+      "grad_norm": 1.2620419370274698,
+      "learning_rate": 4.824776202668875e-06,
+      "loss": 0.8492,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6831473643709118,
+      "grad_norm": 1.4696126150706412,
+      "learning_rate": 4.819445661494437e-06,
+      "loss": 0.8123,
+      "step": 3551
+    },
+    {
+      "epoch": 0.6833397460561754,
+      "grad_norm": 1.3835351962385447,
+      "learning_rate": 4.8141171315895694e-06,
+      "loss": 0.8385,
+      "step": 3552
+    },
+    {
+      "epoch": 0.683532127741439,
+      "grad_norm": 1.180767677324277,
+      "learning_rate": 4.808790615022987e-06,
+      "loss": 0.8793,
+      "step": 3553
+    },
+    {
+      "epoch": 0.6837245094267026,
+      "grad_norm": 1.2657180486589261,
+      "learning_rate": 4.803466113862626e-06,
+      "loss": 0.8852,
+      "step": 3554
+    },
+    {
+      "epoch": 0.6839168911119662,
+      "grad_norm": 1.440429554122597,
+      "learning_rate": 4.798143630175642e-06,
+      "loss": 0.8154,
+      "step": 3555
+    },
+    {
+      "epoch": 0.6841092727972297,
+      "grad_norm": 1.3329642157724138,
+      "learning_rate": 4.792823166028405e-06,
+      "loss": 0.8509,
+      "step": 3556
+    },
+    {
+      "epoch": 0.6843016544824932,
+      "grad_norm": 1.3047070979366195,
+      "learning_rate": 4.787504723486505e-06,
+      "loss": 0.8559,
+      "step": 3557
+    },
+    {
+      "epoch": 0.6844940361677568,
+      "grad_norm": 1.4870331955828728,
+      "learning_rate": 4.7821883046147414e-06,
+      "loss": 0.8779,
+      "step": 3558
+    },
+    {
+      "epoch": 0.6846864178530204,
+      "grad_norm": 1.2826442550748989,
+      "learning_rate": 4.776873911477133e-06,
+      "loss": 0.8875,
+      "step": 3559
+    },
+    {
+      "epoch": 0.6848787995382839,
+      "grad_norm": 1.4203355287245332,
+      "learning_rate": 4.771561546136908e-06,
+      "loss": 0.8489,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6850711812235475,
+      "grad_norm": 1.4547290470693706,
+      "learning_rate": 4.766251210656509e-06,
+      "loss": 0.8814,
+      "step": 3561
+    },
+    {
+      "epoch": 0.6852635629088111,
+      "grad_norm": 1.5806080406548766,
+      "learning_rate": 4.760942907097601e-06,
+      "loss": 0.85,
+      "step": 3562
+    },
+    {
+      "epoch": 0.6854559445940747,
+      "grad_norm": 0.6378586677513246,
+      "learning_rate": 4.755636637521035e-06,
+      "loss": 0.343,
+      "step": 3563
+    },
+    {
+      "epoch": 0.6856483262793382,
+      "grad_norm": 1.4274786583615744,
+      "learning_rate": 4.750332403986902e-06,
+      "loss": 0.841,
+      "step": 3564
+    },
+    {
+      "epoch": 0.6858407079646017,
+      "grad_norm": 1.2337161426000913,
+      "learning_rate": 4.7450302085544735e-06,
+      "loss": 0.8119,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6860330896498653,
+      "grad_norm": 1.1964594045485821,
+      "learning_rate": 4.739730053282255e-06,
+      "loss": 0.8372,
+      "step": 3566
+    },
+    {
+      "epoch": 0.6862254713351289,
+      "grad_norm": 1.2209603131146591,
+      "learning_rate": 4.734431940227951e-06,
+      "loss": 0.8362,
+      "step": 3567
+    },
+    {
+      "epoch": 0.6864178530203925,
+      "grad_norm": 1.4900925869204555,
+      "learning_rate": 4.7291358714484594e-06,
+      "loss": 0.8751,
+      "step": 3568
+    },
+    {
+      "epoch": 0.6866102347056561,
+      "grad_norm": 1.5326401476703353,
+      "learning_rate": 4.723841848999907e-06,
+      "loss": 0.8438,
+      "step": 3569
+    },
+    {
+      "epoch": 0.6868026163909196,
+      "grad_norm": 1.7960412575887004,
+      "learning_rate": 4.718549874937612e-06,
+      "loss": 0.8767,
+      "step": 3570
+    },
+    {
+      "epoch": 0.6869949980761831,
+      "grad_norm": 1.2844812095961806,
+      "learning_rate": 4.713259951316103e-06,
+      "loss": 0.7854,
+      "step": 3571
+    },
+    {
+      "epoch": 0.6871873797614467,
+      "grad_norm": 2.8000184346747083,
+      "learning_rate": 4.707972080189106e-06,
+      "loss": 0.8082,
+      "step": 3572
+    },
+    {
+      "epoch": 0.6873797614467103,
+      "grad_norm": 1.338095860460746,
+      "learning_rate": 4.702686263609559e-06,
+      "loss": 0.8367,
+      "step": 3573
+    },
+    {
+      "epoch": 0.6875721431319738,
+      "grad_norm": 1.4300537664602466,
+      "learning_rate": 4.697402503629596e-06,
+      "loss": 0.8257,
+      "step": 3574
+    },
+    {
+      "epoch": 0.6877645248172374,
+      "grad_norm": 1.2764805666012506,
+      "learning_rate": 4.69212080230055e-06,
+      "loss": 0.8708,
+      "step": 3575
+    },
+    {
+      "epoch": 0.687956906502501,
+      "grad_norm": 1.301714729029086,
+      "learning_rate": 4.686841161672974e-06,
+      "loss": 0.8893,
+      "step": 3576
+    },
+    {
+      "epoch": 0.6881492881877646,
+      "grad_norm": 1.4211969118534171,
+      "learning_rate": 4.681563583796587e-06,
+      "loss": 0.8558,
+      "step": 3577
+    },
+    {
+      "epoch": 0.6883416698730281,
+      "grad_norm": 1.251300653377483,
+      "learning_rate": 4.67628807072034e-06,
+      "loss": 0.793,
+      "step": 3578
+    },
+    {
+      "epoch": 0.6885340515582916,
+      "grad_norm": 1.3202363869984823,
+      "learning_rate": 4.6710146244923645e-06,
+      "loss": 0.8899,
+      "step": 3579
+    },
+    {
+      "epoch": 0.6887264332435552,
+      "grad_norm": 1.2698663934398975,
+      "learning_rate": 4.665743247159995e-06,
+      "loss": 0.8439,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6889188149288188,
+      "grad_norm": 1.1290025294916182,
+      "learning_rate": 4.660473940769761e-06,
+      "loss": 0.8264,
+      "step": 3581
+    },
+    {
+      "epoch": 0.6891111966140824,
+      "grad_norm": 1.400729017276321,
+      "learning_rate": 4.655206707367388e-06,
+      "loss": 0.8281,
+      "step": 3582
+    },
+    {
+      "epoch": 0.6893035782993459,
+      "grad_norm": 1.1466931500304167,
+      "learning_rate": 4.649941548997797e-06,
+      "loss": 0.8618,
+      "step": 3583
+    },
+    {
+      "epoch": 0.6894959599846094,
+      "grad_norm": 1.3849653495899714,
+      "learning_rate": 4.644678467705101e-06,
+      "loss": 0.843,
+      "step": 3584
+    },
+    {
+      "epoch": 0.689688341669873,
+      "grad_norm": 1.3703739718577537,
+      "learning_rate": 4.639417465532622e-06,
+      "loss": 0.8209,
+      "step": 3585
+    },
+    {
+      "epoch": 0.6898807233551366,
+      "grad_norm": 1.3512273342108219,
+      "learning_rate": 4.634158544522849e-06,
+      "loss": 0.8439,
+      "step": 3586
+    },
+    {
+      "epoch": 0.6900731050404002,
+      "grad_norm": 1.1330302467317648,
+      "learning_rate": 4.628901706717476e-06,
+      "loss": 0.8307,
+      "step": 3587
+    },
+    {
+      "epoch": 0.6902654867256637,
+      "grad_norm": 1.269167861089902,
+      "learning_rate": 4.623646954157399e-06,
+      "loss": 0.8389,
+      "step": 3588
+    },
+    {
+      "epoch": 0.6904578684109273,
+      "grad_norm": 1.2024129966826298,
+      "learning_rate": 4.618394288882681e-06,
+      "loss": 0.8611,
+      "step": 3589
+    },
+    {
+      "epoch": 0.6906502500961909,
+      "grad_norm": 1.1999285088926008,
+      "learning_rate": 4.613143712932603e-06,
+      "loss": 0.8553,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6908426317814544,
+      "grad_norm": 0.6158453988851019,
+      "learning_rate": 4.607895228345603e-06,
+      "loss": 0.3728,
+      "step": 3591
+    },
+    {
+      "epoch": 0.6910350134667179,
+      "grad_norm": 1.1107602979481155,
+      "learning_rate": 4.602648837159333e-06,
+      "loss": 0.8117,
+      "step": 3592
+    },
+    {
+      "epoch": 0.6912273951519815,
+      "grad_norm": 0.6363155835126323,
+      "learning_rate": 4.597404541410622e-06,
+      "loss": 0.3562,
+      "step": 3593
+    },
+    {
+      "epoch": 0.6914197768372451,
+      "grad_norm": 1.269528003654571,
+      "learning_rate": 4.592162343135483e-06,
+      "loss": 0.8187,
+      "step": 3594
+    },
+    {
+      "epoch": 0.6916121585225087,
+      "grad_norm": 1.1994796709693731,
+      "learning_rate": 4.586922244369122e-06,
+      "loss": 0.834,
+      "step": 3595
+    },
+    {
+      "epoch": 0.6918045402077723,
+      "grad_norm": 1.2383111677051153,
+      "learning_rate": 4.5816842471459224e-06,
+      "loss": 0.8337,
+      "step": 3596
+    },
+    {
+      "epoch": 0.6919969218930357,
+      "grad_norm": 1.1912092842963375,
+      "learning_rate": 4.576448353499457e-06,
+      "loss": 0.8782,
+      "step": 3597
+    },
+    {
+      "epoch": 0.6921893035782993,
+      "grad_norm": 1.7069434409519544,
+      "learning_rate": 4.571214565462477e-06,
+      "loss": 0.8097,
+      "step": 3598
+    },
+    {
+      "epoch": 0.6923816852635629,
+      "grad_norm": 1.263736175853771,
+      "learning_rate": 4.565982885066923e-06,
+      "loss": 0.7677,
+      "step": 3599
+    },
+    {
+      "epoch": 0.6925740669488265,
+      "grad_norm": 1.4066790290651212,
+      "learning_rate": 4.560753314343912e-06,
+      "loss": 0.8354,
+      "step": 3600
+    },
+    {
+      "epoch": 0.69276644863409,
+      "grad_norm": 1.3993734118287038,
+      "learning_rate": 4.555525855323738e-06,
+      "loss": 0.793,
+      "step": 3601
+    },
+    {
+      "epoch": 0.6929588303193536,
+      "grad_norm": 3.14253382815869,
+      "learning_rate": 4.5503005100358945e-06,
+      "loss": 0.7922,
+      "step": 3602
+    },
+    {
+      "epoch": 0.6931512120046172,
+      "grad_norm": 0.6041029847424804,
+      "learning_rate": 4.545077280509022e-06,
+      "loss": 0.3443,
+      "step": 3603
+    },
+    {
+      "epoch": 0.6933435936898807,
+      "grad_norm": 0.5708627879221306,
+      "learning_rate": 4.539856168770974e-06,
+      "loss": 0.3327,
+      "step": 3604
+    },
+    {
+      "epoch": 0.6935359753751443,
+      "grad_norm": 1.228340939469642,
+      "learning_rate": 4.534637176848758e-06,
+      "loss": 0.8685,
+      "step": 3605
+    },
+    {
+      "epoch": 0.6937283570604078,
+      "grad_norm": 1.3404645459764075,
+      "learning_rate": 4.52942030676857e-06,
+      "loss": 0.9075,
+      "step": 3606
+    },
+    {
+      "epoch": 0.6939207387456714,
+      "grad_norm": 1.3372027968798739,
+      "learning_rate": 4.524205560555774e-06,
+      "loss": 0.7836,
+      "step": 3607
+    },
+    {
+      "epoch": 0.694113120430935,
+      "grad_norm": 1.307974563140228,
+      "learning_rate": 4.5189929402349175e-06,
+      "loss": 0.8109,
+      "step": 3608
+    },
+    {
+      "epoch": 0.6943055021161986,
+      "grad_norm": 1.3471548106221176,
+      "learning_rate": 4.513782447829717e-06,
+      "loss": 0.8536,
+      "step": 3609
+    },
+    {
+      "epoch": 0.694497883801462,
+      "grad_norm": 1.6468450187974173,
+      "learning_rate": 4.508574085363065e-06,
+      "loss": 0.8362,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6946902654867256,
+      "grad_norm": 1.3532684945976108,
+      "learning_rate": 4.503367854857035e-06,
+      "loss": 0.8886,
+      "step": 3611
+    },
+    {
+      "epoch": 0.6948826471719892,
+      "grad_norm": 1.423664070897657,
+      "learning_rate": 4.498163758332853e-06,
+      "loss": 0.761,
+      "step": 3612
+    },
+    {
+      "epoch": 0.6950750288572528,
+      "grad_norm": 1.25394169555739,
+      "learning_rate": 4.492961797810932e-06,
+      "loss": 0.871,
+      "step": 3613
+    },
+    {
+      "epoch": 0.6952674105425164,
+      "grad_norm": 1.4350287180743648,
+      "learning_rate": 4.4877619753108605e-06,
+      "loss": 0.7966,
+      "step": 3614
+    },
+    {
+      "epoch": 0.6954597922277799,
+      "grad_norm": 1.4898289736888888,
+      "learning_rate": 4.4825642928513746e-06,
+      "loss": 0.8261,
+      "step": 3615
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "grad_norm": 1.3573690911905945,
+      "learning_rate": 4.477368752450409e-06,
+      "loss": 0.8457,
+      "step": 3616
+    },
+    {
+      "epoch": 0.695844555598307,
+      "grad_norm": 1.3581402362422135,
+      "learning_rate": 4.472175356125036e-06,
+      "loss": 0.8172,
+      "step": 3617
+    },
+    {
+      "epoch": 0.6960369372835706,
+      "grad_norm": 1.2423165565705547,
+      "learning_rate": 4.466984105891521e-06,
+      "loss": 0.8132,
+      "step": 3618
+    },
+    {
+      "epoch": 0.6962293189688342,
+      "grad_norm": 1.109947885051931,
+      "learning_rate": 4.461795003765285e-06,
+      "loss": 0.8892,
+      "step": 3619
+    },
+    {
+      "epoch": 0.6964217006540977,
+      "grad_norm": 1.5116388742962465,
+      "learning_rate": 4.456608051760914e-06,
+      "loss": 0.887,
+      "step": 3620
+    },
+    {
+      "epoch": 0.6966140823393613,
+      "grad_norm": 2.2523441896990946,
+      "learning_rate": 4.45142325189216e-06,
+      "loss": 0.8415,
+      "step": 3621
+    },
+    {
+      "epoch": 0.6968064640246249,
+      "grad_norm": 1.3123929747320917,
+      "learning_rate": 4.446240606171945e-06,
+      "loss": 0.8456,
+      "step": 3622
+    },
+    {
+      "epoch": 0.6969988457098885,
+      "grad_norm": 1.3573289080732889,
+      "learning_rate": 4.4410601166123475e-06,
+      "loss": 0.8689,
+      "step": 3623
+    },
+    {
+      "epoch": 0.6971912273951519,
+      "grad_norm": 0.6246499106629397,
+      "learning_rate": 4.4358817852246124e-06,
+      "loss": 0.3398,
+      "step": 3624
+    },
+    {
+      "epoch": 0.6973836090804155,
+      "grad_norm": 2.237688917398295,
+      "learning_rate": 4.430705614019147e-06,
+      "loss": 0.7918,
+      "step": 3625
+    },
+    {
+      "epoch": 0.6975759907656791,
+      "grad_norm": 1.425815977744903,
+      "learning_rate": 4.425531605005519e-06,
+      "loss": 0.8932,
+      "step": 3626
+    },
+    {
+      "epoch": 0.6977683724509427,
+      "grad_norm": 1.7024709995532719,
+      "learning_rate": 4.420359760192452e-06,
+      "loss": 0.8766,
+      "step": 3627
+    },
+    {
+      "epoch": 0.6979607541362063,
+      "grad_norm": 1.4506162574287327,
+      "learning_rate": 4.4151900815878455e-06,
+      "loss": 0.8307,
+      "step": 3628
+    },
+    {
+      "epoch": 0.6981531358214698,
+      "grad_norm": 1.2767983372415708,
+      "learning_rate": 4.410022571198734e-06,
+      "loss": 0.86,
+      "step": 3629
+    },
+    {
+      "epoch": 0.6983455175067333,
+      "grad_norm": 1.172073862600037,
+      "learning_rate": 4.404857231031332e-06,
+      "loss": 0.8286,
+      "step": 3630
+    },
+    {
+      "epoch": 0.6985378991919969,
+      "grad_norm": 1.1514393624335164,
+      "learning_rate": 4.399694063090999e-06,
+      "loss": 0.8758,
+      "step": 3631
+    },
+    {
+      "epoch": 0.6987302808772605,
+      "grad_norm": 1.3675097503625775,
+      "learning_rate": 4.394533069382255e-06,
+      "loss": 0.8996,
+      "step": 3632
+    },
+    {
+      "epoch": 0.698922662562524,
+      "grad_norm": 1.2516501660677775,
+      "learning_rate": 4.3893742519087754e-06,
+      "loss": 0.7957,
+      "step": 3633
+    },
+    {
+      "epoch": 0.6991150442477876,
+      "grad_norm": 1.141911789404819,
+      "learning_rate": 4.3842176126733914e-06,
+      "loss": 0.8388,
+      "step": 3634
+    },
+    {
+      "epoch": 0.6993074259330512,
+      "grad_norm": 1.2072720247925666,
+      "learning_rate": 4.379063153678087e-06,
+      "loss": 0.8238,
+      "step": 3635
+    },
+    {
+      "epoch": 0.6994998076183148,
+      "grad_norm": 1.0900497854509223,
+      "learning_rate": 4.373910876923997e-06,
+      "loss": 0.8785,
+      "step": 3636
+    },
+    {
+      "epoch": 0.6996921893035783,
+      "grad_norm": 1.34905956687211,
+      "learning_rate": 4.368760784411423e-06,
+      "loss": 0.8233,
+      "step": 3637
+    },
+    {
+      "epoch": 0.6998845709888418,
+      "grad_norm": 1.418658099037477,
+      "learning_rate": 4.363612878139799e-06,
+      "loss": 0.8092,
+      "step": 3638
+    },
+    {
+      "epoch": 0.7000769526741054,
+      "grad_norm": 1.3104902964045315,
+      "learning_rate": 4.3584671601077224e-06,
+      "loss": 0.8713,
+      "step": 3639
+    },
+    {
+      "epoch": 0.700269334359369,
+      "grad_norm": 1.3160926951705696,
+      "learning_rate": 4.353323632312938e-06,
+      "loss": 0.8727,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7004617160446326,
+      "grad_norm": 1.3813926004553712,
+      "learning_rate": 4.348182296752336e-06,
+      "loss": 0.8788,
+      "step": 3641
+    },
+    {
+      "epoch": 0.7006540977298961,
+      "grad_norm": 1.1502475293345733,
+      "learning_rate": 4.343043155421971e-06,
+      "loss": 0.9303,
+      "step": 3642
+    },
+    {
+      "epoch": 0.7008464794151597,
+      "grad_norm": 1.438031595726889,
+      "learning_rate": 4.3379062103170214e-06,
+      "loss": 0.8654,
+      "step": 3643
+    },
+    {
+      "epoch": 0.7010388611004232,
+      "grad_norm": 1.2819862079448465,
+      "learning_rate": 4.332771463431837e-06,
+      "loss": 0.8351,
+      "step": 3644
+    },
+    {
+      "epoch": 0.7012312427856868,
+      "grad_norm": 1.1447187048977925,
+      "learning_rate": 4.327638916759898e-06,
+      "loss": 0.8231,
+      "step": 3645
+    },
+    {
+      "epoch": 0.7014236244709504,
+      "grad_norm": 1.4565906481840198,
+      "learning_rate": 4.322508572293836e-06,
+      "loss": 0.8661,
+      "step": 3646
+    },
+    {
+      "epoch": 0.7016160061562139,
+      "grad_norm": 1.3236815208135173,
+      "learning_rate": 4.317380432025428e-06,
+      "loss": 0.856,
+      "step": 3647
+    },
+    {
+      "epoch": 0.7018083878414775,
+      "grad_norm": 1.5153525927461244,
+      "learning_rate": 4.312254497945595e-06,
+      "loss": 0.7692,
+      "step": 3648
+    },
+    {
+      "epoch": 0.7020007695267411,
+      "grad_norm": 1.2433209112558303,
+      "learning_rate": 4.3071307720444015e-06,
+      "loss": 0.8124,
+      "step": 3649
+    },
+    {
+      "epoch": 0.7021931512120047,
+      "grad_norm": 1.3017913706805602,
+      "learning_rate": 4.3020092563110485e-06,
+      "loss": 0.8387,
+      "step": 3650
+    },
+    {
+      "epoch": 0.7023855328972681,
+      "grad_norm": 1.5659362984887257,
+      "learning_rate": 4.2968899527338984e-06,
+      "loss": 0.8275,
+      "step": 3651
+    },
+    {
+      "epoch": 0.7025779145825317,
+      "grad_norm": 1.1410832739387347,
+      "learning_rate": 4.291772863300428e-06,
+      "loss": 0.8234,
+      "step": 3652
+    },
+    {
+      "epoch": 0.7027702962677953,
+      "grad_norm": 1.34138894016158,
+      "learning_rate": 4.2866579899972686e-06,
+      "loss": 0.8781,
+      "step": 3653
+    },
+    {
+      "epoch": 0.7029626779530589,
+      "grad_norm": 1.6804128686911821,
+      "learning_rate": 4.281545334810201e-06,
+      "loss": 0.813,
+      "step": 3654
+    },
+    {
+      "epoch": 0.7031550596383225,
+      "grad_norm": 1.2113590732903867,
+      "learning_rate": 4.276434899724119e-06,
+      "loss": 0.8793,
+      "step": 3655
+    },
+    {
+      "epoch": 0.703347441323586,
+      "grad_norm": 1.2009093678180587,
+      "learning_rate": 4.27132668672308e-06,
+      "loss": 0.7964,
+      "step": 3656
+    },
+    {
+      "epoch": 0.7035398230088495,
+      "grad_norm": 3.473944400063257,
+      "learning_rate": 4.266220697790266e-06,
+      "loss": 0.8417,
+      "step": 3657
+    },
+    {
+      "epoch": 0.7037322046941131,
+      "grad_norm": 1.2965447030536938,
+      "learning_rate": 4.2611169349079985e-06,
+      "loss": 0.855,
+      "step": 3658
+    },
+    {
+      "epoch": 0.7039245863793767,
+      "grad_norm": 1.1649411677908252,
+      "learning_rate": 4.25601540005773e-06,
+      "loss": 0.8616,
+      "step": 3659
+    },
+    {
+      "epoch": 0.7041169680646402,
+      "grad_norm": 1.3350571853014244,
+      "learning_rate": 4.250916095220056e-06,
+      "loss": 0.8605,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7043093497499038,
+      "grad_norm": 1.3387697679489077,
+      "learning_rate": 4.2458190223747e-06,
+      "loss": 0.8338,
+      "step": 3661
+    },
+    {
+      "epoch": 0.7045017314351674,
+      "grad_norm": 1.1975119717092595,
+      "learning_rate": 4.240724183500518e-06,
+      "loss": 0.8615,
+      "step": 3662
+    },
+    {
+      "epoch": 0.704694113120431,
+      "grad_norm": 1.4349370728748057,
+      "learning_rate": 4.2356315805755135e-06,
+      "loss": 0.9353,
+      "step": 3663
+    },
+    {
+      "epoch": 0.7048864948056945,
+      "grad_norm": 1.190661791659316,
+      "learning_rate": 4.230541215576798e-06,
+      "loss": 0.8716,
+      "step": 3664
+    },
+    {
+      "epoch": 0.705078876490958,
+      "grad_norm": 1.2979688198791672,
+      "learning_rate": 4.225453090480631e-06,
+      "loss": 0.8752,
+      "step": 3665
+    },
+    {
+      "epoch": 0.7052712581762216,
+      "grad_norm": 0.6569505769960408,
+      "learning_rate": 4.220367207262398e-06,
+      "loss": 0.3664,
+      "step": 3666
+    },
+    {
+      "epoch": 0.7054636398614852,
+      "grad_norm": 1.2616156885633663,
+      "learning_rate": 4.21528356789661e-06,
+      "loss": 0.8558,
+      "step": 3667
+    },
+    {
+      "epoch": 0.7056560215467488,
+      "grad_norm": 1.7004406639245007,
+      "learning_rate": 4.210202174356922e-06,
+      "loss": 0.8746,
+      "step": 3668
+    },
+    {
+      "epoch": 0.7058484032320124,
+      "grad_norm": 1.1584972235709812,
+      "learning_rate": 4.20512302861609e-06,
+      "loss": 0.7976,
+      "step": 3669
+    },
+    {
+      "epoch": 0.7060407849172758,
+      "grad_norm": 1.2789352620914909,
+      "learning_rate": 4.2000461326460274e-06,
+      "loss": 0.8297,
+      "step": 3670
+    },
+    {
+      "epoch": 0.7062331666025394,
+      "grad_norm": 1.519968240504878,
+      "learning_rate": 4.194971488417753e-06,
+      "loss": 0.8533,
+      "step": 3671
+    },
+    {
+      "epoch": 0.706425548287803,
+      "grad_norm": 1.7636988610150544,
+      "learning_rate": 4.189899097901421e-06,
+      "loss": 0.8973,
+      "step": 3672
+    },
+    {
+      "epoch": 0.7066179299730666,
+      "grad_norm": 1.3078249669354192,
+      "learning_rate": 4.184828963066305e-06,
+      "loss": 0.8975,
+      "step": 3673
+    },
+    {
+      "epoch": 0.7068103116583301,
+      "grad_norm": 1.2633175222485538,
+      "learning_rate": 4.179761085880809e-06,
+      "loss": 0.8353,
+      "step": 3674
+    },
+    {
+      "epoch": 0.7070026933435937,
+      "grad_norm": 1.6591455065675649,
+      "learning_rate": 4.174695468312456e-06,
+      "loss": 0.8097,
+      "step": 3675
+    },
+    {
+      "epoch": 0.7071950750288573,
+      "grad_norm": 1.2491804505547537,
+      "learning_rate": 4.16963211232789e-06,
+      "loss": 0.8278,
+      "step": 3676
+    },
+    {
+      "epoch": 0.7073874567141208,
+      "grad_norm": 1.258882555552021,
+      "learning_rate": 4.16457101989289e-06,
+      "loss": 0.8681,
+      "step": 3677
+    },
+    {
+      "epoch": 0.7075798383993844,
+      "grad_norm": 1.764360706452338,
+      "learning_rate": 4.159512192972337e-06,
+      "loss": 0.8494,
+      "step": 3678
+    },
+    {
+      "epoch": 0.7077722200846479,
+      "grad_norm": 1.2374379318849422,
+      "learning_rate": 4.15445563353024e-06,
+      "loss": 0.7967,
+      "step": 3679
+    },
+    {
+      "epoch": 0.7079646017699115,
+      "grad_norm": 0.5897939471604849,
+      "learning_rate": 4.149401343529742e-06,
+      "loss": 0.3183,
+      "step": 3680
+    },
+    {
+      "epoch": 0.7081569834551751,
+      "grad_norm": 1.3527066791562201,
+      "learning_rate": 4.144349324933077e-06,
+      "loss": 0.7692,
+      "step": 3681
+    },
+    {
+      "epoch": 0.7083493651404387,
+      "grad_norm": 1.0677942926733384,
+      "learning_rate": 4.139299579701623e-06,
+      "loss": 0.7929,
+      "step": 3682
+    },
+    {
+      "epoch": 0.7085417468257021,
+      "grad_norm": 2.246765443098884,
+      "learning_rate": 4.134252109795863e-06,
+      "loss": 0.7974,
+      "step": 3683
+    },
+    {
+      "epoch": 0.7087341285109657,
+      "grad_norm": 1.3766010829167568,
+      "learning_rate": 4.129206917175397e-06,
+      "loss": 0.8884,
+      "step": 3684
+    },
+    {
+      "epoch": 0.7089265101962293,
+      "grad_norm": 1.1946943975499302,
+      "learning_rate": 4.124164003798944e-06,
+      "loss": 0.8471,
+      "step": 3685
+    },
+    {
+      "epoch": 0.7091188918814929,
+      "grad_norm": 1.5525745701540958,
+      "learning_rate": 4.119123371624335e-06,
+      "loss": 0.8813,
+      "step": 3686
+    },
+    {
+      "epoch": 0.7093112735667565,
+      "grad_norm": 1.15100235176219,
+      "learning_rate": 4.114085022608517e-06,
+      "loss": 0.8299,
+      "step": 3687
+    },
+    {
+      "epoch": 0.70950365525202,
+      "grad_norm": 1.4849106532429601,
+      "learning_rate": 4.109048958707552e-06,
+      "loss": 0.9006,
+      "step": 3688
+    },
+    {
+      "epoch": 0.7096960369372836,
+      "grad_norm": 1.445095391739624,
+      "learning_rate": 4.104015181876613e-06,
+      "loss": 0.8443,
+      "step": 3689
+    },
+    {
+      "epoch": 0.7098884186225471,
+      "grad_norm": 1.1860722419585124,
+      "learning_rate": 4.09898369406998e-06,
+      "loss": 0.8271,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7100808003078107,
+      "grad_norm": 1.486026970071776,
+      "learning_rate": 4.0939544972410636e-06,
+      "loss": 0.8845,
+      "step": 3691
+    },
+    {
+      "epoch": 0.7102731819930742,
+      "grad_norm": 1.2944878635285098,
+      "learning_rate": 4.0889275933423576e-06,
+      "loss": 0.8069,
+      "step": 3692
+    },
+    {
+      "epoch": 0.7104655636783378,
+      "grad_norm": 1.3365260268540382,
+      "learning_rate": 4.0839029843254815e-06,
+      "loss": 0.8071,
+      "step": 3693
+    },
+    {
+      "epoch": 0.7106579453636014,
+      "grad_norm": 1.5897663591551947,
+      "learning_rate": 4.078880672141171e-06,
+      "loss": 0.8036,
+      "step": 3694
+    },
+    {
+      "epoch": 0.710850327048865,
+      "grad_norm": 1.3889272983864591,
+      "learning_rate": 4.073860658739246e-06,
+      "loss": 0.8973,
+      "step": 3695
+    },
+    {
+      "epoch": 0.7110427087341286,
+      "grad_norm": 1.5252341836931416,
+      "learning_rate": 4.068842946068661e-06,
+      "loss": 0.7882,
+      "step": 3696
+    },
+    {
+      "epoch": 0.711235090419392,
+      "grad_norm": 1.1265425510322111,
+      "learning_rate": 4.063827536077459e-06,
+      "loss": 0.7945,
+      "step": 3697
+    },
+    {
+      "epoch": 0.7114274721046556,
+      "grad_norm": 3.433037652588296,
+      "learning_rate": 4.058814430712796e-06,
+      "loss": 0.8275,
+      "step": 3698
+    },
+    {
+      "epoch": 0.7116198537899192,
+      "grad_norm": 1.4832228151610436,
+      "learning_rate": 4.0538036319209325e-06,
+      "loss": 0.8486,
+      "step": 3699
+    },
+    {
+      "epoch": 0.7118122354751828,
+      "grad_norm": 1.2156169815185822,
+      "learning_rate": 4.0487951416472324e-06,
+      "loss": 0.8259,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7120046171604463,
+      "grad_norm": 1.283762361660768,
+      "learning_rate": 4.043788961836164e-06,
+      "loss": 0.866,
+      "step": 3701
+    },
+    {
+      "epoch": 0.7121969988457099,
+      "grad_norm": 1.2645648279756467,
+      "learning_rate": 4.038785094431295e-06,
+      "loss": 0.8434,
+      "step": 3702
+    },
+    {
+      "epoch": 0.7123893805309734,
+      "grad_norm": 0.6002782602133102,
+      "learning_rate": 4.0337835413753116e-06,
+      "loss": 0.307,
+      "step": 3703
+    },
+    {
+      "epoch": 0.712581762216237,
+      "grad_norm": 1.2061296139434066,
+      "learning_rate": 4.0287843046099765e-06,
+      "loss": 0.8407,
+      "step": 3704
+    },
+    {
+      "epoch": 0.7127741439015006,
+      "grad_norm": 1.2727683865935553,
+      "learning_rate": 4.0237873860761645e-06,
+      "loss": 0.8413,
+      "step": 3705
+    },
+    {
+      "epoch": 0.7129665255867641,
+      "grad_norm": 1.4180269980330256,
+      "learning_rate": 4.018792787713865e-06,
+      "loss": 0.8302,
+      "step": 3706
+    },
+    {
+      "epoch": 0.7131589072720277,
+      "grad_norm": 1.2280446184707414,
+      "learning_rate": 4.013800511462135e-06,
+      "loss": 0.8617,
+      "step": 3707
+    },
+    {
+      "epoch": 0.7133512889572913,
+      "grad_norm": 1.3067062862564538,
+      "learning_rate": 4.008810559259162e-06,
+      "loss": 0.7969,
+      "step": 3708
+    },
+    {
+      "epoch": 0.7135436706425549,
+      "grad_norm": 1.2554541735078764,
+      "learning_rate": 4.003822933042213e-06,
+      "loss": 0.8223,
+      "step": 3709
+    },
+    {
+      "epoch": 0.7137360523278183,
+      "grad_norm": 1.3220886433016366,
+      "learning_rate": 3.998837634747655e-06,
+      "loss": 0.7998,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7139284340130819,
+      "grad_norm": 1.280792437635919,
+      "learning_rate": 3.993854666310955e-06,
+      "loss": 0.8513,
+      "step": 3711
+    },
+    {
+      "epoch": 0.7141208156983455,
+      "grad_norm": 1.1748747639582793,
+      "learning_rate": 3.98887402966667e-06,
+      "loss": 0.8771,
+      "step": 3712
+    },
+    {
+      "epoch": 0.7143131973836091,
+      "grad_norm": 1.2642792650263843,
+      "learning_rate": 3.983895726748455e-06,
+      "loss": 0.8326,
+      "step": 3713
+    },
+    {
+      "epoch": 0.7145055790688727,
+      "grad_norm": 1.2927145100186574,
+      "learning_rate": 3.97891975948906e-06,
+      "loss": 0.8806,
+      "step": 3714
+    },
+    {
+      "epoch": 0.7146979607541362,
+      "grad_norm": 1.351505087549575,
+      "learning_rate": 3.973946129820326e-06,
+      "loss": 0.8424,
+      "step": 3715
+    },
+    {
+      "epoch": 0.7148903424393998,
+      "grad_norm": 1.3108691239455013,
+      "learning_rate": 3.968974839673186e-06,
+      "loss": 0.7484,
+      "step": 3716
+    },
+    {
+      "epoch": 0.7150827241246633,
+      "grad_norm": 1.262387269752966,
+      "learning_rate": 3.964005890977672e-06,
+      "loss": 0.8694,
+      "step": 3717
+    },
+    {
+      "epoch": 0.7152751058099269,
+      "grad_norm": 0.5766500312243684,
+      "learning_rate": 3.9590392856628946e-06,
+      "loss": 0.3283,
+      "step": 3718
+    },
+    {
+      "epoch": 0.7154674874951905,
+      "grad_norm": 1.248777488000166,
+      "learning_rate": 3.954075025657058e-06,
+      "loss": 0.8395,
+      "step": 3719
+    },
+    {
+      "epoch": 0.715659869180454,
+      "grad_norm": 1.1995945583610947,
+      "learning_rate": 3.949113112887471e-06,
+      "loss": 0.8689,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7158522508657176,
+      "grad_norm": 1.3348208792496232,
+      "learning_rate": 3.944153549280506e-06,
+      "loss": 0.8962,
+      "step": 3721
+    },
+    {
+      "epoch": 0.7160446325509812,
+      "grad_norm": 1.0872751510113814,
+      "learning_rate": 3.939196336761645e-06,
+      "loss": 0.8036,
+      "step": 3722
+    },
+    {
+      "epoch": 0.7162370142362448,
+      "grad_norm": 1.1653418351055134,
+      "learning_rate": 3.934241477255445e-06,
+      "loss": 0.8561,
+      "step": 3723
+    },
+    {
+      "epoch": 0.7164293959215082,
+      "grad_norm": 1.3969563592458636,
+      "learning_rate": 3.929288972685555e-06,
+      "loss": 0.8011,
+      "step": 3724
+    },
+    {
+      "epoch": 0.7166217776067718,
+      "grad_norm": 1.535578692323379,
+      "learning_rate": 3.924338824974705e-06,
+      "loss": 0.8995,
+      "step": 3725
+    },
+    {
+      "epoch": 0.7168141592920354,
+      "grad_norm": 0.6209023103717433,
+      "learning_rate": 3.919391036044715e-06,
+      "loss": 0.324,
+      "step": 3726
+    },
+    {
+      "epoch": 0.717006540977299,
+      "grad_norm": 1.5438476142458675,
+      "learning_rate": 3.914445607816486e-06,
+      "loss": 0.7821,
+      "step": 3727
+    },
+    {
+      "epoch": 0.7171989226625626,
+      "grad_norm": 1.180167448208778,
+      "learning_rate": 3.909502542210001e-06,
+      "loss": 0.8408,
+      "step": 3728
+    },
+    {
+      "epoch": 0.717391304347826,
+      "grad_norm": 2.0716731942937163,
+      "learning_rate": 3.904561841144338e-06,
+      "loss": 0.8096,
+      "step": 3729
+    },
+    {
+      "epoch": 0.7175836860330896,
+      "grad_norm": 1.3824117598159407,
+      "learning_rate": 3.899623506537635e-06,
+      "loss": 0.839,
+      "step": 3730
+    },
+    {
+      "epoch": 0.7177760677183532,
+      "grad_norm": 1.4571795124896907,
+      "learning_rate": 3.894687540307127e-06,
+      "loss": 0.8479,
+      "step": 3731
+    },
+    {
+      "epoch": 0.7179684494036168,
+      "grad_norm": 1.2045882741552456,
+      "learning_rate": 3.8897539443691355e-06,
+      "loss": 0.9111,
+      "step": 3732
+    },
+    {
+      "epoch": 0.7181608310888803,
+      "grad_norm": 1.3423158060029758,
+      "learning_rate": 3.884822720639036e-06,
+      "loss": 0.8744,
+      "step": 3733
+    },
+    {
+      "epoch": 0.7183532127741439,
+      "grad_norm": 1.785621897173759,
+      "learning_rate": 3.879893871031314e-06,
+      "loss": 0.8342,
+      "step": 3734
+    },
+    {
+      "epoch": 0.7185455944594075,
+      "grad_norm": 1.4859467232981027,
+      "learning_rate": 3.874967397459511e-06,
+      "loss": 0.8362,
+      "step": 3735
+    },
+    {
+      "epoch": 0.718737976144671,
+      "grad_norm": 1.2519044243080784,
+      "learning_rate": 3.870043301836256e-06,
+      "loss": 0.8502,
+      "step": 3736
+    },
+    {
+      "epoch": 0.7189303578299346,
+      "grad_norm": 1.378225890039622,
+      "learning_rate": 3.86512158607325e-06,
+      "loss": 0.8847,
+      "step": 3737
+    },
+    {
+      "epoch": 0.7191227395151981,
+      "grad_norm": 1.563599448966041,
+      "learning_rate": 3.860202252081276e-06,
+      "loss": 0.9204,
+      "step": 3738
+    },
+    {
+      "epoch": 0.7193151212004617,
+      "grad_norm": 1.4044148770882545,
+      "learning_rate": 3.855285301770188e-06,
+      "loss": 0.825,
+      "step": 3739
+    },
+    {
+      "epoch": 0.7195075028857253,
+      "grad_norm": 1.6396503555088893,
+      "learning_rate": 3.850370737048913e-06,
+      "loss": 0.8048,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7196998845709889,
+      "grad_norm": 1.4213014914985385,
+      "learning_rate": 3.8454585598254565e-06,
+      "loss": 0.8325,
+      "step": 3741
+    },
+    {
+      "epoch": 0.7198922662562524,
+      "grad_norm": 1.3430594557924984,
+      "learning_rate": 3.840548772006891e-06,
+      "loss": 0.8257,
+      "step": 3742
+    },
+    {
+      "epoch": 0.7200846479415159,
+      "grad_norm": 1.3311206753979807,
+      "learning_rate": 3.835641375499375e-06,
+      "loss": 0.8428,
+      "step": 3743
+    },
+    {
+      "epoch": 0.7202770296267795,
+      "grad_norm": 1.045298278913022,
+      "learning_rate": 3.830736372208118e-06,
+      "loss": 0.8266,
+      "step": 3744
+    },
+    {
+      "epoch": 0.7204694113120431,
+      "grad_norm": 1.323894797619529,
+      "learning_rate": 3.8258337640374125e-06,
+      "loss": 0.8286,
+      "step": 3745
+    },
+    {
+      "epoch": 0.7206617929973067,
+      "grad_norm": 1.3542778761838052,
+      "learning_rate": 3.820933552890629e-06,
+      "loss": 0.8185,
+      "step": 3746
+    },
+    {
+      "epoch": 0.7208541746825702,
+      "grad_norm": 1.1259570071877258,
+      "learning_rate": 3.816035740670185e-06,
+      "loss": 0.8138,
+      "step": 3747
+    },
+    {
+      "epoch": 0.7210465563678338,
+      "grad_norm": 1.1777748534834753,
+      "learning_rate": 3.811140329277591e-06,
+      "loss": 0.8342,
+      "step": 3748
+    },
+    {
+      "epoch": 0.7212389380530974,
+      "grad_norm": 0.6642500971236356,
+      "learning_rate": 3.8062473206134088e-06,
+      "loss": 0.3293,
+      "step": 3749
+    },
+    {
+      "epoch": 0.7214313197383609,
+      "grad_norm": 1.636633051604331,
+      "learning_rate": 3.8013567165772735e-06,
+      "loss": 0.8596,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7216237014236244,
+      "grad_norm": 1.5773114436972024,
+      "learning_rate": 3.7964685190678874e-06,
+      "loss": 0.8357,
+      "step": 3751
+    },
+    {
+      "epoch": 0.721816083108888,
+      "grad_norm": 1.290273615184994,
+      "learning_rate": 3.7915827299830154e-06,
+      "loss": 0.8444,
+      "step": 3752
+    },
+    {
+      "epoch": 0.7220084647941516,
+      "grad_norm": 1.3575924375725565,
+      "learning_rate": 3.7866993512194895e-06,
+      "loss": 0.8168,
+      "step": 3753
+    },
+    {
+      "epoch": 0.7222008464794152,
+      "grad_norm": 1.4197439414796595,
+      "learning_rate": 3.7818183846732024e-06,
+      "loss": 0.8387,
+      "step": 3754
+    },
+    {
+      "epoch": 0.7223932281646788,
+      "grad_norm": 1.3532819385718893,
+      "learning_rate": 3.776939832239125e-06,
+      "loss": 0.7143,
+      "step": 3755
+    },
+    {
+      "epoch": 0.7225856098499422,
+      "grad_norm": 1.87105161432023,
+      "learning_rate": 3.7720636958112623e-06,
+      "loss": 0.7876,
+      "step": 3756
+    },
+    {
+      "epoch": 0.7227779915352058,
+      "grad_norm": 1.249429443043136,
+      "learning_rate": 3.7671899772827113e-06,
+      "loss": 0.8481,
+      "step": 3757
+    },
+    {
+      "epoch": 0.7229703732204694,
+      "grad_norm": 1.3112881934489158,
+      "learning_rate": 3.7623186785456156e-06,
+      "loss": 0.8155,
+      "step": 3758
+    },
+    {
+      "epoch": 0.723162754905733,
+      "grad_norm": 1.3691697582003879,
+      "learning_rate": 3.757449801491172e-06,
+      "loss": 0.7906,
+      "step": 3759
+    },
+    {
+      "epoch": 0.7233551365909965,
+      "grad_norm": 1.2751983375404552,
+      "learning_rate": 3.7525833480096575e-06,
+      "loss": 0.742,
+      "step": 3760
+    },
+    {
+      "epoch": 0.7235475182762601,
+      "grad_norm": 1.3197901619046157,
+      "learning_rate": 3.7477193199903903e-06,
+      "loss": 0.8649,
+      "step": 3761
+    },
+    {
+      "epoch": 0.7237398999615237,
+      "grad_norm": 1.239089129349937,
+      "learning_rate": 3.7428577193217563e-06,
+      "loss": 0.9018,
+      "step": 3762
+    },
+    {
+      "epoch": 0.7239322816467872,
+      "grad_norm": 1.1295581748951913,
+      "learning_rate": 3.737998547891195e-06,
+      "loss": 0.82,
+      "step": 3763
+    },
+    {
+      "epoch": 0.7241246633320508,
+      "grad_norm": 1.1838889766337892,
+      "learning_rate": 3.7331418075852053e-06,
+      "loss": 0.862,
+      "step": 3764
+    },
+    {
+      "epoch": 0.7243170450173143,
+      "grad_norm": 1.2960801228034977,
+      "learning_rate": 3.728287500289339e-06,
+      "loss": 0.8538,
+      "step": 3765
+    },
+    {
+      "epoch": 0.7245094267025779,
+      "grad_norm": 1.3472668502892522,
+      "learning_rate": 3.7234356278882076e-06,
+      "loss": 0.88,
+      "step": 3766
+    },
+    {
+      "epoch": 0.7247018083878415,
+      "grad_norm": 1.284187985339541,
+      "learning_rate": 3.718586192265473e-06,
+      "loss": 0.8102,
+      "step": 3767
+    },
+    {
+      "epoch": 0.7248941900731051,
+      "grad_norm": 1.387334488794683,
+      "learning_rate": 3.7137391953038516e-06,
+      "loss": 0.9024,
+      "step": 3768
+    },
+    {
+      "epoch": 0.7250865717583687,
+      "grad_norm": 1.5515809821789062,
+      "learning_rate": 3.7088946388851223e-06,
+      "loss": 0.8158,
+      "step": 3769
+    },
+    {
+      "epoch": 0.7252789534436321,
+      "grad_norm": 2.1966850541590803,
+      "learning_rate": 3.7040525248901003e-06,
+      "loss": 0.8141,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7254713351288957,
+      "grad_norm": 1.1616330558045656,
+      "learning_rate": 3.6992128551986617e-06,
+      "loss": 0.8669,
+      "step": 3771
+    },
+    {
+      "epoch": 0.7256637168141593,
+      "grad_norm": 1.521122791565137,
+      "learning_rate": 3.6943756316897406e-06,
+      "loss": 0.8624,
+      "step": 3772
+    },
+    {
+      "epoch": 0.7258560984994229,
+      "grad_norm": 2.3891522663008433,
+      "learning_rate": 3.6895408562413027e-06,
+      "loss": 0.8662,
+      "step": 3773
+    },
+    {
+      "epoch": 0.7260484801846864,
+      "grad_norm": 1.2933478859506728,
+      "learning_rate": 3.684708530730382e-06,
+      "loss": 0.8545,
+      "step": 3774
+    },
+    {
+      "epoch": 0.72624086186995,
+      "grad_norm": 1.398977207998114,
+      "learning_rate": 3.6798786570330526e-06,
+      "loss": 0.8404,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7264332435552135,
+      "grad_norm": 1.3255764300150772,
+      "learning_rate": 3.6750512370244363e-06,
+      "loss": 0.8456,
+      "step": 3776
+    },
+    {
+      "epoch": 0.7266256252404771,
+      "grad_norm": 1.432341704212416,
+      "learning_rate": 3.670226272578704e-06,
+      "loss": 0.8493,
+      "step": 3777
+    },
+    {
+      "epoch": 0.7268180069257407,
+      "grad_norm": 1.3976608545135651,
+      "learning_rate": 3.6654037655690732e-06,
+      "loss": 0.8485,
+      "step": 3778
+    },
+    {
+      "epoch": 0.7270103886110042,
+      "grad_norm": 1.359178644483555,
+      "learning_rate": 3.660583717867807e-06,
+      "loss": 0.8599,
+      "step": 3779
+    },
+    {
+      "epoch": 0.7272027702962678,
+      "grad_norm": 1.178161617046152,
+      "learning_rate": 3.655766131346211e-06,
+      "loss": 0.8996,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7273951519815314,
+      "grad_norm": 1.7436948753804602,
+      "learning_rate": 3.650951007874648e-06,
+      "loss": 0.889,
+      "step": 3781
+    },
+    {
+      "epoch": 0.727587533666795,
+      "grad_norm": 1.2379807163708678,
+      "learning_rate": 3.6461383493225012e-06,
+      "loss": 0.8434,
+      "step": 3782
+    },
+    {
+      "epoch": 0.7277799153520584,
+      "grad_norm": 1.4736548217608731,
+      "learning_rate": 3.6413281575582194e-06,
+      "loss": 0.8077,
+      "step": 3783
+    },
+    {
+      "epoch": 0.727972297037322,
+      "grad_norm": 1.248781886280778,
+      "learning_rate": 3.6365204344492867e-06,
+      "loss": 0.8559,
+      "step": 3784
+    },
+    {
+      "epoch": 0.7281646787225856,
+      "grad_norm": 1.4031536659956676,
+      "learning_rate": 3.6317151818622154e-06,
+      "loss": 0.7629,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7283570604078492,
+      "grad_norm": 1.264503548954264,
+      "learning_rate": 3.62691240166258e-06,
+      "loss": 0.8319,
+      "step": 3786
+    },
+    {
+      "epoch": 0.7285494420931128,
+      "grad_norm": 1.1741276423175597,
+      "learning_rate": 3.6221120957149826e-06,
+      "loss": 0.8789,
+      "step": 3787
+    },
+    {
+      "epoch": 0.7287418237783763,
+      "grad_norm": 1.6672454542374051,
+      "learning_rate": 3.617314265883066e-06,
+      "loss": 0.866,
+      "step": 3788
+    },
+    {
+      "epoch": 0.7289342054636399,
+      "grad_norm": 1.3243261060221176,
+      "learning_rate": 3.612518914029515e-06,
+      "loss": 0.8741,
+      "step": 3789
+    },
+    {
+      "epoch": 0.7291265871489034,
+      "grad_norm": 1.5219625463426272,
+      "learning_rate": 3.6077260420160487e-06,
+      "loss": 0.8421,
+      "step": 3790
+    },
+    {
+      "epoch": 0.729318968834167,
+      "grad_norm": 1.2901842277094215,
+      "learning_rate": 3.602935651703424e-06,
+      "loss": 0.8567,
+      "step": 3791
+    },
+    {
+      "epoch": 0.7295113505194305,
+      "grad_norm": 1.3133914573177528,
+      "learning_rate": 3.598147744951438e-06,
+      "loss": 0.812,
+      "step": 3792
+    },
+    {
+      "epoch": 0.7297037322046941,
+      "grad_norm": 1.1909589257657296,
+      "learning_rate": 3.5933623236189198e-06,
+      "loss": 0.8452,
+      "step": 3793
+    },
+    {
+      "epoch": 0.7298961138899577,
+      "grad_norm": 1.4060536289581655,
+      "learning_rate": 3.58857938956373e-06,
+      "loss": 0.87,
+      "step": 3794
+    },
+    {
+      "epoch": 0.7300884955752213,
+      "grad_norm": 0.6354807195713525,
+      "learning_rate": 3.58379894464278e-06,
+      "loss": 0.3205,
+      "step": 3795
+    },
+    {
+      "epoch": 0.7302808772604849,
+      "grad_norm": 1.3327940095671997,
+      "learning_rate": 3.57902099071199e-06,
+      "loss": 0.8453,
+      "step": 3796
+    },
+    {
+      "epoch": 0.7304732589457483,
+      "grad_norm": 1.5033338025073648,
+      "learning_rate": 3.5742455296263346e-06,
+      "loss": 0.9025,
+      "step": 3797
+    },
+    {
+      "epoch": 0.7306656406310119,
+      "grad_norm": 2.554346545489949,
+      "learning_rate": 3.569472563239814e-06,
+      "loss": 0.8376,
+      "step": 3798
+    },
+    {
+      "epoch": 0.7308580223162755,
+      "grad_norm": 1.4395624899535173,
+      "learning_rate": 3.5647020934054465e-06,
+      "loss": 0.7719,
+      "step": 3799
+    },
+    {
+      "epoch": 0.7310504040015391,
+      "grad_norm": 1.5319958536107865,
+      "learning_rate": 3.559934121975304e-06,
+      "loss": 0.7946,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7312427856868026,
+      "grad_norm": 1.2712975827919126,
+      "learning_rate": 3.5551686508004735e-06,
+      "loss": 0.9032,
+      "step": 3801
+    },
+    {
+      "epoch": 0.7314351673720662,
+      "grad_norm": 1.2399231358403933,
+      "learning_rate": 3.550405681731074e-06,
+      "loss": 0.8583,
+      "step": 3802
+    },
+    {
+      "epoch": 0.7316275490573297,
+      "grad_norm": 1.1610003549851742,
+      "learning_rate": 3.5456452166162547e-06,
+      "loss": 0.8602,
+      "step": 3803
+    },
+    {
+      "epoch": 0.7318199307425933,
+      "grad_norm": 1.2938911588639581,
+      "learning_rate": 3.540887257304193e-06,
+      "loss": 0.8127,
+      "step": 3804
+    },
+    {
+      "epoch": 0.7320123124278569,
+      "grad_norm": 1.3835772178290766,
+      "learning_rate": 3.5361318056420925e-06,
+      "loss": 0.8809,
+      "step": 3805
+    },
+    {
+      "epoch": 0.7322046941131204,
+      "grad_norm": 1.366232365955396,
+      "learning_rate": 3.531378863476178e-06,
+      "loss": 0.8781,
+      "step": 3806
+    },
+    {
+      "epoch": 0.732397075798384,
+      "grad_norm": 0.6041935574679798,
+      "learning_rate": 3.5266284326517165e-06,
+      "loss": 0.3225,
+      "step": 3807
+    },
+    {
+      "epoch": 0.7325894574836476,
+      "grad_norm": 1.6670857508124595,
+      "learning_rate": 3.5218805150129755e-06,
+      "loss": 0.8576,
+      "step": 3808
+    },
+    {
+      "epoch": 0.7327818391689112,
+      "grad_norm": 1.3406232267240314,
+      "learning_rate": 3.5171351124032703e-06,
+      "loss": 0.8221,
+      "step": 3809
+    },
+    {
+      "epoch": 0.7329742208541746,
+      "grad_norm": 0.6072740984856847,
+      "learning_rate": 3.51239222666493e-06,
+      "loss": 0.3396,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7331666025394382,
+      "grad_norm": 1.4724557852111262,
+      "learning_rate": 3.507651859639295e-06,
+      "loss": 0.8811,
+      "step": 3811
+    },
+    {
+      "epoch": 0.7333589842247018,
+      "grad_norm": 1.2645366971531629,
+      "learning_rate": 3.5029140131667493e-06,
+      "loss": 0.8377,
+      "step": 3812
+    },
+    {
+      "epoch": 0.7335513659099654,
+      "grad_norm": 1.4169100036613747,
+      "learning_rate": 3.4981786890866853e-06,
+      "loss": 0.826,
+      "step": 3813
+    },
+    {
+      "epoch": 0.733743747595229,
+      "grad_norm": 1.1832890291562626,
+      "learning_rate": 3.493445889237518e-06,
+      "loss": 0.8276,
+      "step": 3814
+    },
+    {
+      "epoch": 0.7339361292804925,
+      "grad_norm": 1.2224400662814923,
+      "learning_rate": 3.4887156154566847e-06,
+      "loss": 0.8453,
+      "step": 3815
+    },
+    {
+      "epoch": 0.734128510965756,
+      "grad_norm": 1.237787370220006,
+      "learning_rate": 3.4839878695806385e-06,
+      "loss": 0.8477,
+      "step": 3816
+    },
+    {
+      "epoch": 0.7343208926510196,
+      "grad_norm": 1.7810292903735656,
+      "learning_rate": 3.4792626534448547e-06,
+      "loss": 0.8582,
+      "step": 3817
+    },
+    {
+      "epoch": 0.7345132743362832,
+      "grad_norm": 1.33460521851668,
+      "learning_rate": 3.4745399688838243e-06,
+      "loss": 0.902,
+      "step": 3818
+    },
+    {
+      "epoch": 0.7347056560215467,
+      "grad_norm": 1.2560019357529482,
+      "learning_rate": 3.469819817731056e-06,
+      "loss": 0.8438,
+      "step": 3819
+    },
+    {
+      "epoch": 0.7348980377068103,
+      "grad_norm": 1.1905415780887312,
+      "learning_rate": 3.4651022018190715e-06,
+      "loss": 0.8169,
+      "step": 3820
+    },
+    {
+      "epoch": 0.7350904193920739,
+      "grad_norm": 1.1899112491140644,
+      "learning_rate": 3.460387122979423e-06,
+      "loss": 0.8433,
+      "step": 3821
+    },
+    {
+      "epoch": 0.7352828010773375,
+      "grad_norm": 1.0701501510537792,
+      "learning_rate": 3.455674583042652e-06,
+      "loss": 0.8199,
+      "step": 3822
+    },
+    {
+      "epoch": 0.735475182762601,
+      "grad_norm": 0.5982244586460519,
+      "learning_rate": 3.4509645838383386e-06,
+      "loss": 0.3542,
+      "step": 3823
+    },
+    {
+      "epoch": 0.7356675644478645,
+      "grad_norm": 1.4750820393243886,
+      "learning_rate": 3.4462571271950674e-06,
+      "loss": 0.8344,
+      "step": 3824
+    },
+    {
+      "epoch": 0.7358599461331281,
+      "grad_norm": 2.600215823495481,
+      "learning_rate": 3.4415522149404233e-06,
+      "loss": 0.8727,
+      "step": 3825
+    },
+    {
+      "epoch": 0.7360523278183917,
+      "grad_norm": 1.1862851048459857,
+      "learning_rate": 3.436849848901028e-06,
+      "loss": 0.8528,
+      "step": 3826
+    },
+    {
+      "epoch": 0.7362447095036553,
+      "grad_norm": 1.074621724276366,
+      "learning_rate": 3.432150030902497e-06,
+      "loss": 0.8761,
+      "step": 3827
+    },
+    {
+      "epoch": 0.7364370911889189,
+      "grad_norm": 1.3375467748211747,
+      "learning_rate": 3.427452762769462e-06,
+      "loss": 0.8457,
+      "step": 3828
+    },
+    {
+      "epoch": 0.7366294728741823,
+      "grad_norm": 1.3792920216998026,
+      "learning_rate": 3.4227580463255628e-06,
+      "loss": 0.8485,
+      "step": 3829
+    },
+    {
+      "epoch": 0.7368218545594459,
+      "grad_norm": 0.6185821145090196,
+      "learning_rate": 3.4180658833934523e-06,
+      "loss": 0.343,
+      "step": 3830
+    },
+    {
+      "epoch": 0.7370142362447095,
+      "grad_norm": 1.2745500119204343,
+      "learning_rate": 3.4133762757947873e-06,
+      "loss": 0.8612,
+      "step": 3831
+    },
+    {
+      "epoch": 0.7372066179299731,
+      "grad_norm": 1.1941946868232935,
+      "learning_rate": 3.4086892253502344e-06,
+      "loss": 0.8505,
+      "step": 3832
+    },
+    {
+      "epoch": 0.7373989996152366,
+      "grad_norm": 1.468649088837791,
+      "learning_rate": 3.4040047338794756e-06,
+      "loss": 0.7898,
+      "step": 3833
+    },
+    {
+      "epoch": 0.7375913813005002,
+      "grad_norm": 1.8914346218849878,
+      "learning_rate": 3.3993228032011784e-06,
+      "loss": 0.8666,
+      "step": 3834
+    },
+    {
+      "epoch": 0.7377837629857638,
+      "grad_norm": 1.6179720173161782,
+      "learning_rate": 3.3946434351330415e-06,
+      "loss": 0.8157,
+      "step": 3835
+    },
+    {
+      "epoch": 0.7379761446710273,
+      "grad_norm": 0.6184796923568923,
+      "learning_rate": 3.3899666314917512e-06,
+      "loss": 0.334,
+      "step": 3836
+    },
+    {
+      "epoch": 0.7381685263562909,
+      "grad_norm": 1.1999469697838845,
+      "learning_rate": 3.385292394093006e-06,
+      "loss": 0.852,
+      "step": 3837
+    },
+    {
+      "epoch": 0.7383609080415544,
+      "grad_norm": 1.2638290065181477,
+      "learning_rate": 3.3806207247515068e-06,
+      "loss": 0.8402,
+      "step": 3838
+    },
+    {
+      "epoch": 0.738553289726818,
+      "grad_norm": 1.3142510163521093,
+      "learning_rate": 3.375951625280948e-06,
+      "loss": 0.8054,
+      "step": 3839
+    },
+    {
+      "epoch": 0.7387456714120816,
+      "grad_norm": 1.4934761060601314,
+      "learning_rate": 3.3712850974940437e-06,
+      "loss": 0.8369,
+      "step": 3840
+    },
+    {
+      "epoch": 0.7389380530973452,
+      "grad_norm": 1.3563059271026838,
+      "learning_rate": 3.3666211432024974e-06,
+      "loss": 0.8215,
+      "step": 3841
+    },
+    {
+      "epoch": 0.7391304347826086,
+      "grad_norm": 1.3262941851536287,
+      "learning_rate": 3.361959764217018e-06,
+      "loss": 0.874,
+      "step": 3842
+    },
+    {
+      "epoch": 0.7393228164678722,
+      "grad_norm": 1.27096471510716,
+      "learning_rate": 3.357300962347313e-06,
+      "loss": 0.8814,
+      "step": 3843
+    },
+    {
+      "epoch": 0.7395151981531358,
+      "grad_norm": 1.4230443504024037,
+      "learning_rate": 3.3526447394020887e-06,
+      "loss": 0.8397,
+      "step": 3844
+    },
+    {
+      "epoch": 0.7397075798383994,
+      "grad_norm": 1.1881344028224277,
+      "learning_rate": 3.3479910971890516e-06,
+      "loss": 0.8572,
+      "step": 3845
+    },
+    {
+      "epoch": 0.739899961523663,
+      "grad_norm": 1.3149306917511845,
+      "learning_rate": 3.343340037514903e-06,
+      "loss": 0.8695,
+      "step": 3846
+    },
+    {
+      "epoch": 0.7400923432089265,
+      "grad_norm": 1.3144640833852408,
+      "learning_rate": 3.3386915621853533e-06,
+      "loss": 0.8434,
+      "step": 3847
+    },
+    {
+      "epoch": 0.7402847248941901,
+      "grad_norm": 1.5305089075497973,
+      "learning_rate": 3.3340456730050887e-06,
+      "loss": 0.8842,
+      "step": 3848
+    },
+    {
+      "epoch": 0.7404771065794536,
+      "grad_norm": 1.4052374372721474,
+      "learning_rate": 3.3294023717778122e-06,
+      "loss": 0.8685,
+      "step": 3849
+    },
+    {
+      "epoch": 0.7406694882647172,
+      "grad_norm": 1.3124435526302236,
+      "learning_rate": 3.324761660306215e-06,
+      "loss": 0.896,
+      "step": 3850
+    },
+    {
+      "epoch": 0.7408618699499807,
+      "grad_norm": 1.5632905587441803,
+      "learning_rate": 3.3201235403919683e-06,
+      "loss": 0.8441,
+      "step": 3851
+    },
+    {
+      "epoch": 0.7410542516352443,
+      "grad_norm": 0.6285186184429582,
+      "learning_rate": 3.3154880138357626e-06,
+      "loss": 0.3222,
+      "step": 3852
+    },
+    {
+      "epoch": 0.7412466333205079,
+      "grad_norm": 1.1828549129403605,
+      "learning_rate": 3.3108550824372632e-06,
+      "loss": 0.8182,
+      "step": 3853
+    },
+    {
+      "epoch": 0.7414390150057715,
+      "grad_norm": 2.209204668830113,
+      "learning_rate": 3.306224747995136e-06,
+      "loss": 0.8453,
+      "step": 3854
+    },
+    {
+      "epoch": 0.7416313966910351,
+      "grad_norm": 1.5171910842515282,
+      "learning_rate": 3.301597012307034e-06,
+      "loss": 0.7614,
+      "step": 3855
+    },
+    {
+      "epoch": 0.7418237783762985,
+      "grad_norm": 1.4993666257365568,
+      "learning_rate": 3.2969718771696047e-06,
+      "loss": 0.8752,
+      "step": 3856
+    },
+    {
+      "epoch": 0.7420161600615621,
+      "grad_norm": 1.3477331340479741,
+      "learning_rate": 3.292349344378486e-06,
+      "loss": 0.7809,
+      "step": 3857
+    },
+    {
+      "epoch": 0.7422085417468257,
+      "grad_norm": 1.326333752332574,
+      "learning_rate": 3.287729415728298e-06,
+      "loss": 0.8129,
+      "step": 3858
+    },
+    {
+      "epoch": 0.7424009234320893,
+      "grad_norm": 1.277965131789565,
+      "learning_rate": 3.283112093012669e-06,
+      "loss": 0.8319,
+      "step": 3859
+    },
+    {
+      "epoch": 0.7425933051173528,
+      "grad_norm": 1.2197279871774849,
+      "learning_rate": 3.278497378024187e-06,
+      "loss": 0.8095,
+      "step": 3860
+    },
+    {
+      "epoch": 0.7427856868026164,
+      "grad_norm": 1.3202058453926309,
+      "learning_rate": 3.2738852725544547e-06,
+      "loss": 0.8748,
+      "step": 3861
+    },
+    {
+      "epoch": 0.74297806848788,
+      "grad_norm": 1.176689189357264,
+      "learning_rate": 3.2692757783940467e-06,
+      "loss": 0.8247,
+      "step": 3862
+    },
+    {
+      "epoch": 0.7431704501731435,
+      "grad_norm": 1.2864378007795245,
+      "learning_rate": 3.264668897332527e-06,
+      "loss": 0.816,
+      "step": 3863
+    },
+    {
+      "epoch": 0.7433628318584071,
+      "grad_norm": 1.2063505937331451,
+      "learning_rate": 3.2600646311584494e-06,
+      "loss": 0.8354,
+      "step": 3864
+    },
+    {
+      "epoch": 0.7435552135436706,
+      "grad_norm": 1.1979012841415737,
+      "learning_rate": 3.2554629816593375e-06,
+      "loss": 0.87,
+      "step": 3865
+    },
+    {
+      "epoch": 0.7437475952289342,
+      "grad_norm": 1.452194366545032,
+      "learning_rate": 3.250863950621721e-06,
+      "loss": 0.7984,
+      "step": 3866
+    },
+    {
+      "epoch": 0.7439399769141978,
+      "grad_norm": 1.3095924148590317,
+      "learning_rate": 3.2462675398310984e-06,
+      "loss": 0.8132,
+      "step": 3867
+    },
+    {
+      "epoch": 0.7441323585994614,
+      "grad_norm": 1.4176951349277884,
+      "learning_rate": 3.241673751071954e-06,
+      "loss": 0.8084,
+      "step": 3868
+    },
+    {
+      "epoch": 0.7443247402847248,
+      "grad_norm": 1.2127304128304908,
+      "learning_rate": 3.2370825861277567e-06,
+      "loss": 0.806,
+      "step": 3869
+    },
+    {
+      "epoch": 0.7445171219699884,
+      "grad_norm": 1.469332019164403,
+      "learning_rate": 3.2324940467809527e-06,
+      "loss": 0.8331,
+      "step": 3870
+    },
+    {
+      "epoch": 0.744709503655252,
+      "grad_norm": 1.3252400988678035,
+      "learning_rate": 3.2279081348129713e-06,
+      "loss": 0.838,
+      "step": 3871
+    },
+    {
+      "epoch": 0.7449018853405156,
+      "grad_norm": 1.1836958152817305,
+      "learning_rate": 3.223324852004219e-06,
+      "loss": 0.7744,
+      "step": 3872
+    },
+    {
+      "epoch": 0.7450942670257792,
+      "grad_norm": 1.6041571563748964,
+      "learning_rate": 3.2187442001340942e-06,
+      "loss": 0.8613,
+      "step": 3873
+    },
+    {
+      "epoch": 0.7452866487110427,
+      "grad_norm": 3.52318276323214,
+      "learning_rate": 3.21416618098095e-06,
+      "loss": 0.8885,
+      "step": 3874
+    },
+    {
+      "epoch": 0.7454790303963063,
+      "grad_norm": 1.535482960719999,
+      "learning_rate": 3.2095907963221396e-06,
+      "loss": 0.8364,
+      "step": 3875
+    },
+    {
+      "epoch": 0.7456714120815698,
+      "grad_norm": 0.6485202308365516,
+      "learning_rate": 3.2050180479339865e-06,
+      "loss": 0.3412,
+      "step": 3876
+    },
+    {
+      "epoch": 0.7458637937668334,
+      "grad_norm": 1.2394696311334612,
+      "learning_rate": 3.2004479375917783e-06,
+      "loss": 0.8188,
+      "step": 3877
+    },
+    {
+      "epoch": 0.746056175452097,
+      "grad_norm": 0.6166557688014616,
+      "learning_rate": 3.1958804670698008e-06,
+      "loss": 0.3495,
+      "step": 3878
+    },
+    {
+      "epoch": 0.7462485571373605,
+      "grad_norm": 1.306695233974811,
+      "learning_rate": 3.191315638141297e-06,
+      "loss": 0.8169,
+      "step": 3879
+    },
+    {
+      "epoch": 0.7464409388226241,
+      "grad_norm": 1.3219529512014707,
+      "learning_rate": 3.1867534525784937e-06,
+      "loss": 0.8021,
+      "step": 3880
+    },
+    {
+      "epoch": 0.7466333205078877,
+      "grad_norm": 1.4958121536310718,
+      "learning_rate": 3.182193912152586e-06,
+      "loss": 0.8496,
+      "step": 3881
+    },
+    {
+      "epoch": 0.7468257021931513,
+      "grad_norm": 1.1922507292235889,
+      "learning_rate": 3.177637018633746e-06,
+      "loss": 0.8392,
+      "step": 3882
+    },
+    {
+      "epoch": 0.7470180838784147,
+      "grad_norm": 1.3892775857092148,
+      "learning_rate": 3.1730827737911163e-06,
+      "loss": 0.8373,
+      "step": 3883
+    },
+    {
+      "epoch": 0.7472104655636783,
+      "grad_norm": 1.5254037830096367,
+      "learning_rate": 3.1685311793928077e-06,
+      "loss": 0.8334,
+      "step": 3884
+    },
+    {
+      "epoch": 0.7474028472489419,
+      "grad_norm": 1.2992089802263136,
+      "learning_rate": 3.163982237205917e-06,
+      "loss": 0.7623,
+      "step": 3885
+    },
+    {
+      "epoch": 0.7475952289342055,
+      "grad_norm": 1.4684951483323718,
+      "learning_rate": 3.1594359489964853e-06,
+      "loss": 0.8805,
+      "step": 3886
+    },
+    {
+      "epoch": 0.7477876106194691,
+      "grad_norm": 1.5573001986702344,
+      "learning_rate": 3.15489231652955e-06,
+      "loss": 0.7959,
+      "step": 3887
+    },
+    {
+      "epoch": 0.7479799923047326,
+      "grad_norm": 1.382569523408073,
+      "learning_rate": 3.150351341569101e-06,
+      "loss": 0.8358,
+      "step": 3888
+    },
+    {
+      "epoch": 0.7481723739899961,
+      "grad_norm": 1.2189684913802732,
+      "learning_rate": 3.1458130258781006e-06,
+      "loss": 0.8907,
+      "step": 3889
+    },
+    {
+      "epoch": 0.7483647556752597,
+      "grad_norm": 1.498499662222657,
+      "learning_rate": 3.141277371218484e-06,
+      "loss": 0.847,
+      "step": 3890
+    },
+    {
+      "epoch": 0.7485571373605233,
+      "grad_norm": 1.3462200115491825,
+      "learning_rate": 3.136744379351139e-06,
+      "loss": 0.8425,
+      "step": 3891
+    },
+    {
+      "epoch": 0.7487495190457868,
+      "grad_norm": 0.614873215342231,
+      "learning_rate": 3.1322140520359366e-06,
+      "loss": 0.3438,
+      "step": 3892
+    },
+    {
+      "epoch": 0.7489419007310504,
+      "grad_norm": 1.310310805032599,
+      "learning_rate": 3.1276863910317057e-06,
+      "loss": 0.8531,
+      "step": 3893
+    },
+    {
+      "epoch": 0.749134282416314,
+      "grad_norm": 1.1617006974686703,
+      "learning_rate": 3.1231613980962373e-06,
+      "loss": 0.8327,
+      "step": 3894
+    },
+    {
+      "epoch": 0.7493266641015776,
+      "grad_norm": 1.1389463871140295,
+      "learning_rate": 3.1186390749862904e-06,
+      "loss": 0.8634,
+      "step": 3895
+    },
+    {
+      "epoch": 0.7495190457868411,
+      "grad_norm": 1.379523496894811,
+      "learning_rate": 3.1141194234575878e-06,
+      "loss": 0.8122,
+      "step": 3896
+    },
+    {
+      "epoch": 0.7497114274721046,
+      "grad_norm": 1.2969333710994468,
+      "learning_rate": 3.1096024452648123e-06,
+      "loss": 0.7867,
+      "step": 3897
+    },
+    {
+      "epoch": 0.7499038091573682,
+      "grad_norm": 1.266228063118497,
+      "learning_rate": 3.1050881421616076e-06,
+      "loss": 0.8631,
+      "step": 3898
+    },
+    {
+      "epoch": 0.7500961908426318,
+      "grad_norm": 1.3023029777310693,
+      "learning_rate": 3.100576515900591e-06,
+      "loss": 0.8026,
+      "step": 3899
+    },
+    {
+      "epoch": 0.7502885725278954,
+      "grad_norm": 1.2909022168070867,
+      "learning_rate": 3.0960675682333186e-06,
+      "loss": 0.8753,
+      "step": 3900
+    },
+    {
+      "epoch": 0.7504809542131589,
+      "grad_norm": 1.3297507456570141,
+      "learning_rate": 3.0915613009103296e-06,
+      "loss": 0.8595,
+      "step": 3901
+    },
+    {
+      "epoch": 0.7506733358984224,
+      "grad_norm": 1.2667941810037222,
+      "learning_rate": 3.0870577156811077e-06,
+      "loss": 0.7809,
+      "step": 3902
+    },
+    {
+      "epoch": 0.750865717583686,
+      "grad_norm": 1.4917322119479952,
+      "learning_rate": 3.0825568142940998e-06,
+      "loss": 0.8087,
+      "step": 3903
+    },
+    {
+      "epoch": 0.7510580992689496,
+      "grad_norm": 3.1771189168901706,
+      "learning_rate": 3.0780585984967113e-06,
+      "loss": 0.8758,
+      "step": 3904
+    },
+    {
+      "epoch": 0.7512504809542132,
+      "grad_norm": 1.1762546478625213,
+      "learning_rate": 3.073563070035305e-06,
+      "loss": 0.8383,
+      "step": 3905
+    },
+    {
+      "epoch": 0.7514428626394767,
+      "grad_norm": 1.2872845629647762,
+      "learning_rate": 3.069070230655198e-06,
+      "loss": 0.8389,
+      "step": 3906
+    },
+    {
+      "epoch": 0.7516352443247403,
+      "grad_norm": 1.339214488734453,
+      "learning_rate": 3.0645800821006667e-06,
+      "loss": 0.8136,
+      "step": 3907
+    },
+    {
+      "epoch": 0.7518276260100039,
+      "grad_norm": 2.4133459294356756,
+      "learning_rate": 3.060092626114941e-06,
+      "loss": 0.7747,
+      "step": 3908
+    },
+    {
+      "epoch": 0.7520200076952674,
+      "grad_norm": 1.179159034488972,
+      "learning_rate": 3.0556078644402066e-06,
+      "loss": 0.8349,
+      "step": 3909
+    },
+    {
+      "epoch": 0.7522123893805309,
+      "grad_norm": 1.2896699297508791,
+      "learning_rate": 3.051125798817598e-06,
+      "loss": 0.7903,
+      "step": 3910
+    },
+    {
+      "epoch": 0.7524047710657945,
+      "grad_norm": 1.7159123935534437,
+      "learning_rate": 3.0466464309872167e-06,
+      "loss": 0.7917,
+      "step": 3911
+    },
+    {
+      "epoch": 0.7525971527510581,
+      "grad_norm": 1.51194026721638,
+      "learning_rate": 3.042169762688096e-06,
+      "loss": 0.7943,
+      "step": 3912
+    },
+    {
+      "epoch": 0.7527895344363217,
+      "grad_norm": 1.222261742807606,
+      "learning_rate": 3.0376957956582452e-06,
+      "loss": 0.8515,
+      "step": 3913
+    },
+    {
+      "epoch": 0.7529819161215853,
+      "grad_norm": 1.565997013744965,
+      "learning_rate": 3.0332245316346e-06,
+      "loss": 0.8961,
+      "step": 3914
+    },
+    {
+      "epoch": 0.7531742978068487,
+      "grad_norm": 1.2359303692764336,
+      "learning_rate": 3.0287559723530667e-06,
+      "loss": 0.8841,
+      "step": 3915
+    },
+    {
+      "epoch": 0.7533666794921123,
+      "grad_norm": 1.866102635662707,
+      "learning_rate": 3.024290119548495e-06,
+      "loss": 0.832,
+      "step": 3916
+    },
+    {
+      "epoch": 0.7535590611773759,
+      "grad_norm": 1.3323546671466644,
+      "learning_rate": 3.019826974954674e-06,
+      "loss": 0.7966,
+      "step": 3917
+    },
+    {
+      "epoch": 0.7537514428626395,
+      "grad_norm": 1.2043067981430924,
+      "learning_rate": 3.0153665403043586e-06,
+      "loss": 0.8094,
+      "step": 3918
+    },
+    {
+      "epoch": 0.753943824547903,
+      "grad_norm": 1.4678641598246467,
+      "learning_rate": 3.01090881732924e-06,
+      "loss": 0.8552,
+      "step": 3919
+    },
+    {
+      "epoch": 0.7541362062331666,
+      "grad_norm": 1.2958897200620005,
+      "learning_rate": 3.0064538077599603e-06,
+      "loss": 0.7986,
+      "step": 3920
+    },
+    {
+      "epoch": 0.7543285879184302,
+      "grad_norm": 1.4634423511470722,
+      "learning_rate": 3.002001513326107e-06,
+      "loss": 0.7634,
+      "step": 3921
+    },
+    {
+      "epoch": 0.7545209696036937,
+      "grad_norm": 1.5539454074245467,
+      "learning_rate": 2.9975519357562155e-06,
+      "loss": 0.8264,
+      "step": 3922
+    },
+    {
+      "epoch": 0.7547133512889573,
+      "grad_norm": 1.2482740033038144,
+      "learning_rate": 2.9931050767777626e-06,
+      "loss": 0.8409,
+      "step": 3923
+    },
+    {
+      "epoch": 0.7549057329742208,
+      "grad_norm": 1.1488982954282143,
+      "learning_rate": 2.9886609381171703e-06,
+      "loss": 0.724,
+      "step": 3924
+    },
+    {
+      "epoch": 0.7550981146594844,
+      "grad_norm": 1.3828412286988585,
+      "learning_rate": 2.984219521499816e-06,
+      "loss": 0.8084,
+      "step": 3925
+    },
+    {
+      "epoch": 0.755290496344748,
+      "grad_norm": 1.3204084673724314,
+      "learning_rate": 2.9797808286499976e-06,
+      "loss": 0.8168,
+      "step": 3926
+    },
+    {
+      "epoch": 0.7554828780300116,
+      "grad_norm": 0.6115999084441758,
+      "learning_rate": 2.9753448612909775e-06,
+      "loss": 0.3056,
+      "step": 3927
+    },
+    {
+      "epoch": 0.7556752597152752,
+      "grad_norm": 1.2958273371771885,
+      "learning_rate": 2.9709116211449484e-06,
+      "loss": 0.8042,
+      "step": 3928
+    },
+    {
+      "epoch": 0.7558676414005386,
+      "grad_norm": 1.3915663794376356,
+      "learning_rate": 2.966481109933047e-06,
+      "loss": 0.8531,
+      "step": 3929
+    },
+    {
+      "epoch": 0.7560600230858022,
+      "grad_norm": 1.5835749163310366,
+      "learning_rate": 2.9620533293753495e-06,
+      "loss": 0.8105,
+      "step": 3930
+    },
+    {
+      "epoch": 0.7562524047710658,
+      "grad_norm": 5.17540975287061,
+      "learning_rate": 2.957628281190873e-06,
+      "loss": 0.7864,
+      "step": 3931
+    },
+    {
+      "epoch": 0.7564447864563294,
+      "grad_norm": 1.4612619601169268,
+      "learning_rate": 2.9532059670975732e-06,
+      "loss": 0.8639,
+      "step": 3932
+    },
+    {
+      "epoch": 0.7566371681415929,
+      "grad_norm": 1.4038481174384245,
+      "learning_rate": 2.948786388812346e-06,
+      "loss": 0.7911,
+      "step": 3933
+    },
+    {
+      "epoch": 0.7568295498268565,
+      "grad_norm": 1.1495777594687804,
+      "learning_rate": 2.9443695480510225e-06,
+      "loss": 0.7866,
+      "step": 3934
+    },
+    {
+      "epoch": 0.75702193151212,
+      "grad_norm": 0.6069335452684202,
+      "learning_rate": 2.9399554465283742e-06,
+      "loss": 0.2985,
+      "step": 3935
+    },
+    {
+      "epoch": 0.7572143131973836,
+      "grad_norm": 1.8679469509942206,
+      "learning_rate": 2.935544085958102e-06,
+      "loss": 0.8241,
+      "step": 3936
+    },
+    {
+      "epoch": 0.7574066948826472,
+      "grad_norm": 1.4432982918430834,
+      "learning_rate": 2.931135468052858e-06,
+      "loss": 0.8586,
+      "step": 3937
+    },
+    {
+      "epoch": 0.7575990765679107,
+      "grad_norm": 1.7123569708050839,
+      "learning_rate": 2.926729594524207e-06,
+      "loss": 0.8648,
+      "step": 3938
+    },
+    {
+      "epoch": 0.7577914582531743,
+      "grad_norm": 1.1802712516681386,
+      "learning_rate": 2.9223264670826746e-06,
+      "loss": 0.8474,
+      "step": 3939
+    },
+    {
+      "epoch": 0.7579838399384379,
+      "grad_norm": 1.1603677163833426,
+      "learning_rate": 2.9179260874376915e-06,
+      "loss": 0.8289,
+      "step": 3940
+    },
+    {
+      "epoch": 0.7581762216237015,
+      "grad_norm": 1.3422609649491895,
+      "learning_rate": 2.9135284572976486e-06,
+      "loss": 0.8157,
+      "step": 3941
+    },
+    {
+      "epoch": 0.7583686033089649,
+      "grad_norm": 1.54796281093976,
+      "learning_rate": 2.9091335783698517e-06,
+      "loss": 0.8159,
+      "step": 3942
+    },
+    {
+      "epoch": 0.7585609849942285,
+      "grad_norm": 1.503114947112578,
+      "learning_rate": 2.9047414523605467e-06,
+      "loss": 0.8385,
+      "step": 3943
+    },
+    {
+      "epoch": 0.7587533666794921,
+      "grad_norm": 1.26090604235219,
+      "learning_rate": 2.9003520809749053e-06,
+      "loss": 0.8405,
+      "step": 3944
+    },
+    {
+      "epoch": 0.7589457483647557,
+      "grad_norm": 1.281598160287887,
+      "learning_rate": 2.8959654659170354e-06,
+      "loss": 0.861,
+      "step": 3945
+    },
+    {
+      "epoch": 0.7591381300500193,
+      "grad_norm": 1.4688597117719595,
+      "learning_rate": 2.8915816088899696e-06,
+      "loss": 0.8583,
+      "step": 3946
+    },
+    {
+      "epoch": 0.7593305117352828,
+      "grad_norm": 1.2325669756218576,
+      "learning_rate": 2.8872005115956746e-06,
+      "loss": 0.8339,
+      "step": 3947
+    },
+    {
+      "epoch": 0.7595228934205464,
+      "grad_norm": 0.5777243850249222,
+      "learning_rate": 2.8828221757350406e-06,
+      "loss": 0.3138,
+      "step": 3948
+    },
+    {
+      "epoch": 0.7597152751058099,
+      "grad_norm": 1.152057788831253,
+      "learning_rate": 2.8784466030078905e-06,
+      "loss": 0.829,
+      "step": 3949
+    },
+    {
+      "epoch": 0.7599076567910735,
+      "grad_norm": 1.6746038916345407,
+      "learning_rate": 2.874073795112967e-06,
+      "loss": 0.7844,
+      "step": 3950
+    },
+    {
+      "epoch": 0.760100038476337,
+      "grad_norm": 1.3439451452549271,
+      "learning_rate": 2.8697037537479565e-06,
+      "loss": 0.8635,
+      "step": 3951
+    },
+    {
+      "epoch": 0.7602924201616006,
+      "grad_norm": 0.5973486044342403,
+      "learning_rate": 2.8653364806094454e-06,
+      "loss": 0.35,
+      "step": 3952
+    },
+    {
+      "epoch": 0.7604848018468642,
+      "grad_norm": 1.2574328446544007,
+      "learning_rate": 2.86097197739297e-06,
+      "loss": 0.8757,
+      "step": 3953
+    },
+    {
+      "epoch": 0.7606771835321278,
+      "grad_norm": 1.1107400329680615,
+      "learning_rate": 2.856610245792976e-06,
+      "loss": 0.8553,
+      "step": 3954
+    },
+    {
+      "epoch": 0.7608695652173914,
+      "grad_norm": 1.2756019228047264,
+      "learning_rate": 2.8522512875028396e-06,
+      "loss": 0.8416,
+      "step": 3955
+    },
+    {
+      "epoch": 0.7610619469026548,
+      "grad_norm": 1.2530790934439238,
+      "learning_rate": 2.847895104214856e-06,
+      "loss": 0.8046,
+      "step": 3956
+    },
+    {
+      "epoch": 0.7612543285879184,
+      "grad_norm": 1.625819162808581,
+      "learning_rate": 2.843541697620249e-06,
+      "loss": 0.772,
+      "step": 3957
+    },
+    {
+      "epoch": 0.761446710273182,
+      "grad_norm": 1.4307736542674332,
+      "learning_rate": 2.8391910694091584e-06,
+      "loss": 0.8016,
+      "step": 3958
+    },
+    {
+      "epoch": 0.7616390919584456,
+      "grad_norm": 1.3657733765680042,
+      "learning_rate": 2.8348432212706443e-06,
+      "loss": 0.8461,
+      "step": 3959
+    },
+    {
+      "epoch": 0.7618314736437091,
+      "grad_norm": 1.6220964410465737,
+      "learning_rate": 2.8304981548927025e-06,
+      "loss": 0.7996,
+      "step": 3960
+    },
+    {
+      "epoch": 0.7620238553289727,
+      "grad_norm": 1.529639295830632,
+      "learning_rate": 2.826155871962227e-06,
+      "loss": 0.8835,
+      "step": 3961
+    },
+    {
+      "epoch": 0.7622162370142362,
+      "grad_norm": 3.4920270576683614,
+      "learning_rate": 2.8218163741650415e-06,
+      "loss": 0.8865,
+      "step": 3962
+    },
+    {
+      "epoch": 0.7624086186994998,
+      "grad_norm": 1.4575036280528293,
+      "learning_rate": 2.817479663185898e-06,
+      "loss": 0.8724,
+      "step": 3963
+    },
+    {
+      "epoch": 0.7626010003847634,
+      "grad_norm": 1.6843845339645993,
+      "learning_rate": 2.813145740708445e-06,
+      "loss": 0.8015,
+      "step": 3964
+    },
+    {
+      "epoch": 0.7627933820700269,
+      "grad_norm": 1.5402561168678115,
+      "learning_rate": 2.808814608415271e-06,
+      "loss": 0.8902,
+      "step": 3965
+    },
+    {
+      "epoch": 0.7629857637552905,
+      "grad_norm": 1.0929418332623966,
+      "learning_rate": 2.8044862679878605e-06,
+      "loss": 0.875,
+      "step": 3966
+    },
+    {
+      "epoch": 0.7631781454405541,
+      "grad_norm": 1.2676421360281471,
+      "learning_rate": 2.800160721106633e-06,
+      "loss": 0.8225,
+      "step": 3967
+    },
+    {
+      "epoch": 0.7633705271258177,
+      "grad_norm": 1.4205162238470015,
+      "learning_rate": 2.7958379694509108e-06,
+      "loss": 0.8339,
+      "step": 3968
+    },
+    {
+      "epoch": 0.7635629088110811,
+      "grad_norm": 1.3436105942006484,
+      "learning_rate": 2.791518014698935e-06,
+      "loss": 0.8217,
+      "step": 3969
+    },
+    {
+      "epoch": 0.7637552904963447,
+      "grad_norm": 1.5544792745173888,
+      "learning_rate": 2.787200858527862e-06,
+      "loss": 0.8211,
+      "step": 3970
+    },
+    {
+      "epoch": 0.7639476721816083,
+      "grad_norm": 0.6281363691611355,
+      "learning_rate": 2.7828865026137584e-06,
+      "loss": 0.3037,
+      "step": 3971
+    },
+    {
+      "epoch": 0.7641400538668719,
+      "grad_norm": 1.1499042329992588,
+      "learning_rate": 2.7785749486316085e-06,
+      "loss": 0.8251,
+      "step": 3972
+    },
+    {
+      "epoch": 0.7643324355521355,
+      "grad_norm": 1.1552666943110448,
+      "learning_rate": 2.774266198255303e-06,
+      "loss": 0.7924,
+      "step": 3973
+    },
+    {
+      "epoch": 0.764524817237399,
+      "grad_norm": 1.3016550620653695,
+      "learning_rate": 2.7699602531576496e-06,
+      "loss": 0.8148,
+      "step": 3974
+    },
+    {
+      "epoch": 0.7647171989226625,
+      "grad_norm": 1.198832428386302,
+      "learning_rate": 2.765657115010364e-06,
+      "loss": 0.8191,
+      "step": 3975
+    },
+    {
+      "epoch": 0.7649095806079261,
+      "grad_norm": 1.890142505849202,
+      "learning_rate": 2.7613567854840685e-06,
+      "loss": 0.8641,
+      "step": 3976
+    },
+    {
+      "epoch": 0.7651019622931897,
+      "grad_norm": 1.1863605780074766,
+      "learning_rate": 2.7570592662483086e-06,
+      "loss": 0.8494,
+      "step": 3977
+    },
+    {
+      "epoch": 0.7652943439784533,
+      "grad_norm": 1.2241130230174648,
+      "learning_rate": 2.752764558971517e-06,
+      "loss": 0.8772,
+      "step": 3978
+    },
+    {
+      "epoch": 0.7654867256637168,
+      "grad_norm": 1.1406327149169972,
+      "learning_rate": 2.748472665321056e-06,
+      "loss": 0.8836,
+      "step": 3979
+    },
+    {
+      "epoch": 0.7656791073489804,
+      "grad_norm": 1.09715000839726,
+      "learning_rate": 2.744183586963185e-06,
+      "loss": 0.7947,
+      "step": 3980
+    },
+    {
+      "epoch": 0.765871489034244,
+      "grad_norm": 1.6244388343444718,
+      "learning_rate": 2.739897325563069e-06,
+      "loss": 0.8684,
+      "step": 3981
+    },
+    {
+      "epoch": 0.7660638707195075,
+      "grad_norm": 1.2306395627014768,
+      "learning_rate": 2.7356138827847856e-06,
+      "loss": 0.8373,
+      "step": 3982
+    },
+    {
+      "epoch": 0.766256252404771,
+      "grad_norm": 1.4067978506860441,
+      "learning_rate": 2.731333260291311e-06,
+      "loss": 0.8309,
+      "step": 3983
+    },
+    {
+      "epoch": 0.7664486340900346,
+      "grad_norm": 1.193131257761191,
+      "learning_rate": 2.7270554597445343e-06,
+      "loss": 0.8497,
+      "step": 3984
+    },
+    {
+      "epoch": 0.7666410157752982,
+      "grad_norm": 1.2904486561405584,
+      "learning_rate": 2.7227804828052384e-06,
+      "loss": 0.783,
+      "step": 3985
+    },
+    {
+      "epoch": 0.7668333974605618,
+      "grad_norm": 1.6943771703704325,
+      "learning_rate": 2.7185083311331283e-06,
+      "loss": 0.8383,
+      "step": 3986
+    },
+    {
+      "epoch": 0.7670257791458254,
+      "grad_norm": 1.5631069307706265,
+      "learning_rate": 2.7142390063867896e-06,
+      "loss": 0.8614,
+      "step": 3987
+    },
+    {
+      "epoch": 0.7672181608310888,
+      "grad_norm": 0.6178407893651355,
+      "learning_rate": 2.709972510223725e-06,
+      "loss": 0.3219,
+      "step": 3988
+    },
+    {
+      "epoch": 0.7674105425163524,
+      "grad_norm": 1.5688678527660234,
+      "learning_rate": 2.7057088443003343e-06,
+      "loss": 0.8798,
+      "step": 3989
+    },
+    {
+      "epoch": 0.767602924201616,
+      "grad_norm": 1.2259063963520018,
+      "learning_rate": 2.7014480102719174e-06,
+      "loss": 0.854,
+      "step": 3990
+    },
+    {
+      "epoch": 0.7677953058868796,
+      "grad_norm": 1.2244195711017771,
+      "learning_rate": 2.697190009792685e-06,
+      "loss": 0.8798,
+      "step": 3991
+    },
+    {
+      "epoch": 0.7679876875721431,
+      "grad_norm": 1.380110270465111,
+      "learning_rate": 2.692934844515729e-06,
+      "loss": 0.8552,
+      "step": 3992
+    },
+    {
+      "epoch": 0.7681800692574067,
+      "grad_norm": 1.3375177797842068,
+      "learning_rate": 2.6886825160930587e-06,
+      "loss": 0.8628,
+      "step": 3993
+    },
+    {
+      "epoch": 0.7683724509426703,
+      "grad_norm": 1.20372389997342,
+      "learning_rate": 2.6844330261755715e-06,
+      "loss": 0.8622,
+      "step": 3994
+    },
+    {
+      "epoch": 0.7685648326279338,
+      "grad_norm": 1.3393389412695649,
+      "learning_rate": 2.6801863764130653e-06,
+      "loss": 0.833,
+      "step": 3995
+    },
+    {
+      "epoch": 0.7687572143131974,
+      "grad_norm": 1.2105707768015486,
+      "learning_rate": 2.675942568454236e-06,
+      "loss": 0.8441,
+      "step": 3996
+    },
+    {
+      "epoch": 0.7689495959984609,
+      "grad_norm": 0.6037436088949284,
+      "learning_rate": 2.671701603946678e-06,
+      "loss": 0.3063,
+      "step": 3997
+    },
+    {
+      "epoch": 0.7691419776837245,
+      "grad_norm": 1.2155376976662862,
+      "learning_rate": 2.667463484536876e-06,
+      "loss": 0.8199,
+      "step": 3998
+    },
+    {
+      "epoch": 0.7693343593689881,
+      "grad_norm": 1.2752229198505265,
+      "learning_rate": 2.6632282118702147e-06,
+      "loss": 0.8672,
+      "step": 3999
+    },
+    {
+      "epoch": 0.7695267410542517,
+      "grad_norm": 1.3908171807669267,
+      "learning_rate": 2.65899578759098e-06,
+      "loss": 0.8466,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7697191227395151,
+      "grad_norm": 1.374385500669684,
+      "learning_rate": 2.654766213342335e-06,
+      "loss": 0.884,
+      "step": 4001
+    },
+    {
+      "epoch": 0.7699115044247787,
+      "grad_norm": 1.142365743074794,
+      "learning_rate": 2.650539490766346e-06,
+      "loss": 0.8224,
+      "step": 4002
+    },
+    {
+      "epoch": 0.7701038861100423,
+      "grad_norm": 2.370307370963418,
+      "learning_rate": 2.646315621503983e-06,
+      "loss": 0.8412,
+      "step": 4003
+    },
+    {
+      "epoch": 0.7702962677953059,
+      "grad_norm": 1.4530514155659846,
+      "learning_rate": 2.642094607195085e-06,
+      "loss": 0.8228,
+      "step": 4004
+    },
+    {
+      "epoch": 0.7704886494805695,
+      "grad_norm": 1.3653824009487756,
+      "learning_rate": 2.6378764494784027e-06,
+      "loss": 0.9138,
+      "step": 4005
+    },
+    {
+      "epoch": 0.770681031165833,
+      "grad_norm": 1.201727957517205,
+      "learning_rate": 2.633661149991569e-06,
+      "loss": 0.849,
+      "step": 4006
+    },
+    {
+      "epoch": 0.7708734128510966,
+      "grad_norm": 1.3104509245641096,
+      "learning_rate": 2.6294487103711064e-06,
+      "loss": 0.8376,
+      "step": 4007
+    },
+    {
+      "epoch": 0.7710657945363601,
+      "grad_norm": 1.1597539601720028,
+      "learning_rate": 2.6252391322524297e-06,
+      "loss": 0.8281,
+      "step": 4008
+    },
+    {
+      "epoch": 0.7712581762216237,
+      "grad_norm": 1.2211370765992462,
+      "learning_rate": 2.6210324172698432e-06,
+      "loss": 0.8582,
+      "step": 4009
+    },
+    {
+      "epoch": 0.7714505579068872,
+      "grad_norm": 1.1504548591965045,
+      "learning_rate": 2.6168285670565374e-06,
+      "loss": 0.8603,
+      "step": 4010
+    },
+    {
+      "epoch": 0.7716429395921508,
+      "grad_norm": 0.6409052995242276,
+      "learning_rate": 2.6126275832445892e-06,
+      "loss": 0.3605,
+      "step": 4011
+    },
+    {
+      "epoch": 0.7718353212774144,
+      "grad_norm": 2.673926424011301,
+      "learning_rate": 2.6084294674649734e-06,
+      "loss": 0.8984,
+      "step": 4012
+    },
+    {
+      "epoch": 0.772027702962678,
+      "grad_norm": 1.3134892461826908,
+      "learning_rate": 2.6042342213475346e-06,
+      "loss": 0.8566,
+      "step": 4013
+    },
+    {
+      "epoch": 0.7722200846479416,
+      "grad_norm": 1.2365052827077452,
+      "learning_rate": 2.6000418465210143e-06,
+      "loss": 0.8248,
+      "step": 4014
+    },
+    {
+      "epoch": 0.772412466333205,
+      "grad_norm": 1.2558309662173812,
+      "learning_rate": 2.595852344613038e-06,
+      "loss": 0.8254,
+      "step": 4015
+    },
+    {
+      "epoch": 0.7726048480184686,
+      "grad_norm": 1.3262615405095795,
+      "learning_rate": 2.5916657172501103e-06,
+      "loss": 0.7898,
+      "step": 4016
+    },
+    {
+      "epoch": 0.7727972297037322,
+      "grad_norm": 1.7737684373508156,
+      "learning_rate": 2.587481966057633e-06,
+      "loss": 0.745,
+      "step": 4017
+    },
+    {
+      "epoch": 0.7729896113889958,
+      "grad_norm": 1.4336750946743737,
+      "learning_rate": 2.583301092659872e-06,
+      "loss": 0.8635,
+      "step": 4018
+    },
+    {
+      "epoch": 0.7731819930742593,
+      "grad_norm": 1.4663734824987706,
+      "learning_rate": 2.5791230986799944e-06,
+      "loss": 0.7919,
+      "step": 4019
+    },
+    {
+      "epoch": 0.7733743747595229,
+      "grad_norm": 1.4445054548086256,
+      "learning_rate": 2.5749479857400383e-06,
+      "loss": 0.8685,
+      "step": 4020
+    },
+    {
+      "epoch": 0.7735667564447865,
+      "grad_norm": 1.4721204302542719,
+      "learning_rate": 2.5707757554609247e-06,
+      "loss": 0.8486,
+      "step": 4021
+    },
+    {
+      "epoch": 0.77375913813005,
+      "grad_norm": 1.1341056390310535,
+      "learning_rate": 2.56660640946246e-06,
+      "loss": 0.8605,
+      "step": 4022
+    },
+    {
+      "epoch": 0.7739515198153136,
+      "grad_norm": 0.604376391386744,
+      "learning_rate": 2.5624399493633257e-06,
+      "loss": 0.3496,
+      "step": 4023
+    },
+    {
+      "epoch": 0.7741439015005771,
+      "grad_norm": 1.2727881621571666,
+      "learning_rate": 2.558276376781086e-06,
+      "loss": 0.8311,
+      "step": 4024
+    },
+    {
+      "epoch": 0.7743362831858407,
+      "grad_norm": 1.2272853086518283,
+      "learning_rate": 2.55411569333218e-06,
+      "loss": 0.8821,
+      "step": 4025
+    },
+    {
+      "epoch": 0.7745286648711043,
+      "grad_norm": 1.2048882295755945,
+      "learning_rate": 2.5499579006319365e-06,
+      "loss": 0.8243,
+      "step": 4026
+    },
+    {
+      "epoch": 0.7747210465563679,
+      "grad_norm": 1.3376513326008943,
+      "learning_rate": 2.5458030002945457e-06,
+      "loss": 0.8387,
+      "step": 4027
+    },
+    {
+      "epoch": 0.7749134282416313,
+      "grad_norm": 1.33636737406699,
+      "learning_rate": 2.5416509939330836e-06,
+      "loss": 0.8339,
+      "step": 4028
+    },
+    {
+      "epoch": 0.7751058099268949,
+      "grad_norm": 1.3881507190899098,
+      "learning_rate": 2.537501883159509e-06,
+      "loss": 0.8791,
+      "step": 4029
+    },
+    {
+      "epoch": 0.7752981916121585,
+      "grad_norm": 1.293642398656404,
+      "learning_rate": 2.5333556695846384e-06,
+      "loss": 0.8473,
+      "step": 4030
+    },
+    {
+      "epoch": 0.7754905732974221,
+      "grad_norm": 1.3652805077901953,
+      "learning_rate": 2.5292123548181847e-06,
+      "loss": 0.8308,
+      "step": 4031
+    },
+    {
+      "epoch": 0.7756829549826857,
+      "grad_norm": 1.236160613172002,
+      "learning_rate": 2.525071940468722e-06,
+      "loss": 0.8931,
+      "step": 4032
+    },
+    {
+      "epoch": 0.7758753366679492,
+      "grad_norm": 1.3420890617193244,
+      "learning_rate": 2.520934428143701e-06,
+      "loss": 0.852,
+      "step": 4033
+    },
+    {
+      "epoch": 0.7760677183532128,
+      "grad_norm": 2.750114641469359,
+      "learning_rate": 2.5167998194494468e-06,
+      "loss": 0.7862,
+      "step": 4034
+    },
+    {
+      "epoch": 0.7762601000384763,
+      "grad_norm": 1.3832592002420758,
+      "learning_rate": 2.5126681159911558e-06,
+      "loss": 0.8128,
+      "step": 4035
+    },
+    {
+      "epoch": 0.7764524817237399,
+      "grad_norm": 1.5690266574145653,
+      "learning_rate": 2.5085393193729e-06,
+      "loss": 0.8406,
+      "step": 4036
+    },
+    {
+      "epoch": 0.7766448634090035,
+      "grad_norm": 1.3733376065272156,
+      "learning_rate": 2.5044134311976156e-06,
+      "loss": 0.8211,
+      "step": 4037
+    },
+    {
+      "epoch": 0.776837245094267,
+      "grad_norm": 1.372889009371797,
+      "learning_rate": 2.5002904530671236e-06,
+      "loss": 0.7746,
+      "step": 4038
+    },
+    {
+      "epoch": 0.7770296267795306,
+      "grad_norm": 1.1843746300940325,
+      "learning_rate": 2.4961703865820974e-06,
+      "loss": 0.8358,
+      "step": 4039
+    },
+    {
+      "epoch": 0.7772220084647942,
+      "grad_norm": 1.3685971303628224,
+      "learning_rate": 2.492053233342091e-06,
+      "loss": 0.8369,
+      "step": 4040
+    },
+    {
+      "epoch": 0.7774143901500578,
+      "grad_norm": 1.2158473081186367,
+      "learning_rate": 2.487938994945527e-06,
+      "loss": 0.7856,
+      "step": 4041
+    },
+    {
+      "epoch": 0.7776067718353212,
+      "grad_norm": 1.4290219781621611,
+      "learning_rate": 2.4838276729896884e-06,
+      "loss": 0.8072,
+      "step": 4042
+    },
+    {
+      "epoch": 0.7777991535205848,
+      "grad_norm": 1.3789418126821622,
+      "learning_rate": 2.479719269070743e-06,
+      "loss": 0.8447,
+      "step": 4043
+    },
+    {
+      "epoch": 0.7779915352058484,
+      "grad_norm": 1.213685352260515,
+      "learning_rate": 2.4756137847837025e-06,
+      "loss": 0.8504,
+      "step": 4044
+    },
+    {
+      "epoch": 0.778183916891112,
+      "grad_norm": 1.3440929224892069,
+      "learning_rate": 2.4715112217224657e-06,
+      "loss": 0.8223,
+      "step": 4045
+    },
+    {
+      "epoch": 0.7783762985763756,
+      "grad_norm": 1.4235396087637802,
+      "learning_rate": 2.467411581479786e-06,
+      "loss": 0.8539,
+      "step": 4046
+    },
+    {
+      "epoch": 0.7785686802616391,
+      "grad_norm": 1.2643414724515936,
+      "learning_rate": 2.463314865647286e-06,
+      "loss": 0.7604,
+      "step": 4047
+    },
+    {
+      "epoch": 0.7787610619469026,
+      "grad_norm": 1.6709516237447424,
+      "learning_rate": 2.45922107581545e-06,
+      "loss": 0.8055,
+      "step": 4048
+    },
+    {
+      "epoch": 0.7789534436321662,
+      "grad_norm": 1.3495516223322785,
+      "learning_rate": 2.4551302135736287e-06,
+      "loss": 0.8844,
+      "step": 4049
+    },
+    {
+      "epoch": 0.7791458253174298,
+      "grad_norm": 1.2189049594803816,
+      "learning_rate": 2.4510422805100366e-06,
+      "loss": 0.8089,
+      "step": 4050
+    },
+    {
+      "epoch": 0.7793382070026933,
+      "grad_norm": 1.4396742681423857,
+      "learning_rate": 2.446957278211746e-06,
+      "loss": 0.7603,
+      "step": 4051
+    },
+    {
+      "epoch": 0.7795305886879569,
+      "grad_norm": 1.7804442445627742,
+      "learning_rate": 2.4428752082647044e-06,
+      "loss": 0.8699,
+      "step": 4052
+    },
+    {
+      "epoch": 0.7797229703732205,
+      "grad_norm": 1.6535112989847545,
+      "learning_rate": 2.438796072253704e-06,
+      "loss": 0.8206,
+      "step": 4053
+    },
+    {
+      "epoch": 0.7799153520584841,
+      "grad_norm": 1.2215853090914195,
+      "learning_rate": 2.4347198717624054e-06,
+      "loss": 0.7941,
+      "step": 4054
+    },
+    {
+      "epoch": 0.7801077337437476,
+      "grad_norm": 1.6652688897894754,
+      "learning_rate": 2.4306466083733392e-06,
+      "loss": 0.8443,
+      "step": 4055
+    },
+    {
+      "epoch": 0.7803001154290111,
+      "grad_norm": 1.2362265002008177,
+      "learning_rate": 2.426576283667873e-06,
+      "loss": 0.8618,
+      "step": 4056
+    },
+    {
+      "epoch": 0.7804924971142747,
+      "grad_norm": 0.6161524336373135,
+      "learning_rate": 2.422508899226258e-06,
+      "loss": 0.3361,
+      "step": 4057
+    },
+    {
+      "epoch": 0.7806848787995383,
+      "grad_norm": 1.3806211676548412,
+      "learning_rate": 2.418444456627589e-06,
+      "loss": 0.7886,
+      "step": 4058
+    },
+    {
+      "epoch": 0.7808772604848019,
+      "grad_norm": 0.660753966612198,
+      "learning_rate": 2.4143829574498224e-06,
+      "loss": 0.3259,
+      "step": 4059
+    },
+    {
+      "epoch": 0.7810696421700654,
+      "grad_norm": 1.1670236873631645,
+      "learning_rate": 2.4103244032697717e-06,
+      "loss": 0.8376,
+      "step": 4060
+    },
+    {
+      "epoch": 0.781262023855329,
+      "grad_norm": 1.340167024217707,
+      "learning_rate": 2.406268795663108e-06,
+      "loss": 0.8332,
+      "step": 4061
+    },
+    {
+      "epoch": 0.7814544055405925,
+      "grad_norm": 2.8487394414664564,
+      "learning_rate": 2.4022161362043574e-06,
+      "loss": 0.8431,
+      "step": 4062
+    },
+    {
+      "epoch": 0.7816467872258561,
+      "grad_norm": 1.4533328706827415,
+      "learning_rate": 2.3981664264669025e-06,
+      "loss": 0.8709,
+      "step": 4063
+    },
+    {
+      "epoch": 0.7818391689111197,
+      "grad_norm": 2.0277900056004756,
+      "learning_rate": 2.3941196680229794e-06,
+      "loss": 0.87,
+      "step": 4064
+    },
+    {
+      "epoch": 0.7820315505963832,
+      "grad_norm": 1.314314359159892,
+      "learning_rate": 2.3900758624436772e-06,
+      "loss": 0.8635,
+      "step": 4065
+    },
+    {
+      "epoch": 0.7822239322816468,
+      "grad_norm": 1.2070933164425452,
+      "learning_rate": 2.3860350112989473e-06,
+      "loss": 0.8616,
+      "step": 4066
+    },
+    {
+      "epoch": 0.7824163139669104,
+      "grad_norm": 1.2143259862369806,
+      "learning_rate": 2.3819971161575807e-06,
+      "loss": 0.7858,
+      "step": 4067
+    },
+    {
+      "epoch": 0.782608695652174,
+      "grad_norm": 1.2820162951030651,
+      "learning_rate": 2.3779621785872252e-06,
+      "loss": 0.835,
+      "step": 4068
+    },
+    {
+      "epoch": 0.7828010773374374,
+      "grad_norm": 1.2779623471117842,
+      "learning_rate": 2.3739302001543918e-06,
+      "loss": 0.803,
+      "step": 4069
+    },
+    {
+      "epoch": 0.782993459022701,
+      "grad_norm": 1.4242331588194677,
+      "learning_rate": 2.3699011824244234e-06,
+      "loss": 0.818,
+      "step": 4070
+    },
+    {
+      "epoch": 0.7831858407079646,
+      "grad_norm": 3.3232788627878227,
+      "learning_rate": 2.365875126961531e-06,
+      "loss": 0.8139,
+      "step": 4071
+    },
+    {
+      "epoch": 0.7833782223932282,
+      "grad_norm": 1.1050870191346422,
+      "learning_rate": 2.3618520353287644e-06,
+      "loss": 0.8239,
+      "step": 4072
+    },
+    {
+      "epoch": 0.7835706040784918,
+      "grad_norm": 1.7601392100587487,
+      "learning_rate": 2.3578319090880263e-06,
+      "loss": 0.8522,
+      "step": 4073
+    },
+    {
+      "epoch": 0.7837629857637552,
+      "grad_norm": 1.4820468024460873,
+      "learning_rate": 2.3538147498000695e-06,
+      "loss": 0.8352,
+      "step": 4074
+    },
+    {
+      "epoch": 0.7839553674490188,
+      "grad_norm": 1.583729716886434,
+      "learning_rate": 2.349800559024492e-06,
+      "loss": 0.7702,
+      "step": 4075
+    },
+    {
+      "epoch": 0.7841477491342824,
+      "grad_norm": 1.2625879425768516,
+      "learning_rate": 2.3457893383197415e-06,
+      "loss": 0.8087,
+      "step": 4076
+    },
+    {
+      "epoch": 0.784340130819546,
+      "grad_norm": 1.2832617419516017,
+      "learning_rate": 2.3417810892431104e-06,
+      "loss": 0.7682,
+      "step": 4077
+    },
+    {
+      "epoch": 0.7845325125048095,
+      "grad_norm": 1.1681883451483148,
+      "learning_rate": 2.3377758133507455e-06,
+      "loss": 0.8193,
+      "step": 4078
+    },
+    {
+      "epoch": 0.7847248941900731,
+      "grad_norm": 1.5124977354343792,
+      "learning_rate": 2.3337735121976247e-06,
+      "loss": 0.8796,
+      "step": 4079
+    },
+    {
+      "epoch": 0.7849172758753367,
+      "grad_norm": 1.2130954603804536,
+      "learning_rate": 2.32977418733758e-06,
+      "loss": 0.8652,
+      "step": 4080
+    },
+    {
+      "epoch": 0.7851096575606002,
+      "grad_norm": 1.4779983562859769,
+      "learning_rate": 2.3257778403232954e-06,
+      "loss": 0.812,
+      "step": 4081
+    },
+    {
+      "epoch": 0.7853020392458638,
+      "grad_norm": 1.4402367556804938,
+      "learning_rate": 2.321784472706279e-06,
+      "loss": 0.827,
+      "step": 4082
+    },
+    {
+      "epoch": 0.7854944209311273,
+      "grad_norm": 1.4312133767111674,
+      "learning_rate": 2.317794086036901e-06,
+      "loss": 0.868,
+      "step": 4083
+    },
+    {
+      "epoch": 0.7856868026163909,
+      "grad_norm": 1.1654696118958008,
+      "learning_rate": 2.3138066818643647e-06,
+      "loss": 0.8255,
+      "step": 4084
+    },
+    {
+      "epoch": 0.7858791843016545,
+      "grad_norm": 1.2605683069080345,
+      "learning_rate": 2.3098222617367184e-06,
+      "loss": 0.7304,
+      "step": 4085
+    },
+    {
+      "epoch": 0.7860715659869181,
+      "grad_norm": 1.4432061102887166,
+      "learning_rate": 2.30584082720085e-06,
+      "loss": 0.8221,
+      "step": 4086
+    },
+    {
+      "epoch": 0.7862639476721817,
+      "grad_norm": 1.477394082619301,
+      "learning_rate": 2.301862379802492e-06,
+      "loss": 0.8623,
+      "step": 4087
+    },
+    {
+      "epoch": 0.7864563293574451,
+      "grad_norm": 1.4277364642305967,
+      "learning_rate": 2.297886921086211e-06,
+      "loss": 0.8086,
+      "step": 4088
+    },
+    {
+      "epoch": 0.7866487110427087,
+      "grad_norm": 2.6617944179446877,
+      "learning_rate": 2.2939144525954194e-06,
+      "loss": 0.849,
+      "step": 4089
+    },
+    {
+      "epoch": 0.7868410927279723,
+      "grad_norm": 1.3354401618985028,
+      "learning_rate": 2.2899449758723657e-06,
+      "loss": 0.8189,
+      "step": 4090
+    },
+    {
+      "epoch": 0.7870334744132359,
+      "grad_norm": 2.0270812805342886,
+      "learning_rate": 2.285978492458134e-06,
+      "loss": 0.7637,
+      "step": 4091
+    },
+    {
+      "epoch": 0.7872258560984994,
+      "grad_norm": 1.4029016641336494,
+      "learning_rate": 2.282015003892659e-06,
+      "loss": 0.8592,
+      "step": 4092
+    },
+    {
+      "epoch": 0.787418237783763,
+      "grad_norm": 1.2051932008377644,
+      "learning_rate": 2.2780545117146947e-06,
+      "loss": 0.7905,
+      "step": 4093
+    },
+    {
+      "epoch": 0.7876106194690266,
+      "grad_norm": 1.2448295340025355,
+      "learning_rate": 2.2740970174618405e-06,
+      "loss": 0.8014,
+      "step": 4094
+    },
+    {
+      "epoch": 0.7878030011542901,
+      "grad_norm": 1.3644194964449512,
+      "learning_rate": 2.270142522670541e-06,
+      "loss": 0.7995,
+      "step": 4095
+    },
+    {
+      "epoch": 0.7879953828395537,
+      "grad_norm": 1.4425329841161365,
+      "learning_rate": 2.2661910288760545e-06,
+      "loss": 0.8432,
+      "step": 4096
+    },
+    {
+      "epoch": 0.7881877645248172,
+      "grad_norm": 1.4074436558627794,
+      "learning_rate": 2.262242537612497e-06,
+      "loss": 0.8675,
+      "step": 4097
+    },
+    {
+      "epoch": 0.7883801462100808,
+      "grad_norm": 1.5419231655928813,
+      "learning_rate": 2.258297050412804e-06,
+      "loss": 0.776,
+      "step": 4098
+    },
+    {
+      "epoch": 0.7885725278953444,
+      "grad_norm": 1.2708817637901633,
+      "learning_rate": 2.254354568808752e-06,
+      "loss": 0.8574,
+      "step": 4099
+    },
+    {
+      "epoch": 0.788764909580608,
+      "grad_norm": 1.4907821775576346,
+      "learning_rate": 2.2504150943309455e-06,
+      "loss": 0.76,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7889572912658714,
+      "grad_norm": 1.301937215304609,
+      "learning_rate": 2.246478628508827e-06,
+      "loss": 0.8157,
+      "step": 4101
+    },
+    {
+      "epoch": 0.789149672951135,
+      "grad_norm": 1.484807573304999,
+      "learning_rate": 2.242545172870665e-06,
+      "loss": 0.8501,
+      "step": 4102
+    },
+    {
+      "epoch": 0.7893420546363986,
+      "grad_norm": 0.597852060125291,
+      "learning_rate": 2.238614728943561e-06,
+      "loss": 0.2929,
+      "step": 4103
+    },
+    {
+      "epoch": 0.7895344363216622,
+      "grad_norm": 1.5374666790096103,
+      "learning_rate": 2.2346872982534584e-06,
+      "loss": 0.8125,
+      "step": 4104
+    },
+    {
+      "epoch": 0.7897268180069258,
+      "grad_norm": 1.3436996635060747,
+      "learning_rate": 2.2307628823251083e-06,
+      "loss": 0.8275,
+      "step": 4105
+    },
+    {
+      "epoch": 0.7899191996921893,
+      "grad_norm": 0.589224016906374,
+      "learning_rate": 2.2268414826821117e-06,
+      "loss": 0.3238,
+      "step": 4106
+    },
+    {
+      "epoch": 0.7901115813774529,
+      "grad_norm": 1.321669383178789,
+      "learning_rate": 2.222923100846893e-06,
+      "loss": 0.8656,
+      "step": 4107
+    },
+    {
+      "epoch": 0.7903039630627164,
+      "grad_norm": 1.270630401356821,
+      "learning_rate": 2.2190077383406938e-06,
+      "loss": 0.8135,
+      "step": 4108
+    },
+    {
+      "epoch": 0.79049634474798,
+      "grad_norm": 1.3899858062306136,
+      "learning_rate": 2.2150953966835996e-06,
+      "loss": 0.8076,
+      "step": 4109
+    },
+    {
+      "epoch": 0.7906887264332435,
+      "grad_norm": 1.287374029802581,
+      "learning_rate": 2.211186077394516e-06,
+      "loss": 0.8106,
+      "step": 4110
+    },
+    {
+      "epoch": 0.7908811081185071,
+      "grad_norm": 1.2040761119516608,
+      "learning_rate": 2.207279781991173e-06,
+      "loss": 0.8273,
+      "step": 4111
+    },
+    {
+      "epoch": 0.7910734898037707,
+      "grad_norm": 1.1645666686136436,
+      "learning_rate": 2.2033765119901294e-06,
+      "loss": 0.822,
+      "step": 4112
+    },
+    {
+      "epoch": 0.7912658714890343,
+      "grad_norm": 1.405697539232348,
+      "learning_rate": 2.1994762689067705e-06,
+      "loss": 0.8177,
+      "step": 4113
+    },
+    {
+      "epoch": 0.7914582531742979,
+      "grad_norm": 1.219585683624778,
+      "learning_rate": 2.1955790542553036e-06,
+      "loss": 0.8201,
+      "step": 4114
+    },
+    {
+      "epoch": 0.7916506348595613,
+      "grad_norm": 1.1892129073905064,
+      "learning_rate": 2.1916848695487615e-06,
+      "loss": 0.8275,
+      "step": 4115
+    },
+    {
+      "epoch": 0.7918430165448249,
+      "grad_norm": 1.2998333235090014,
+      "learning_rate": 2.1877937162990015e-06,
+      "loss": 0.818,
+      "step": 4116
+    },
+    {
+      "epoch": 0.7920353982300885,
+      "grad_norm": 1.5009044495244794,
+      "learning_rate": 2.1839055960167e-06,
+      "loss": 0.7994,
+      "step": 4117
+    },
+    {
+      "epoch": 0.7922277799153521,
+      "grad_norm": 1.286804058239653,
+      "learning_rate": 2.180020510211367e-06,
+      "loss": 0.7765,
+      "step": 4118
+    },
+    {
+      "epoch": 0.7924201616006156,
+      "grad_norm": 1.6984968470610378,
+      "learning_rate": 2.1761384603913203e-06,
+      "loss": 0.8774,
+      "step": 4119
+    },
+    {
+      "epoch": 0.7926125432858792,
+      "grad_norm": 1.5895631692961552,
+      "learning_rate": 2.172259448063704e-06,
+      "loss": 0.8755,
+      "step": 4120
+    },
+    {
+      "epoch": 0.7928049249711427,
+      "grad_norm": 1.1803101154795774,
+      "learning_rate": 2.1683834747344913e-06,
+      "loss": 0.8792,
+      "step": 4121
+    },
+    {
+      "epoch": 0.7929973066564063,
+      "grad_norm": 1.439957758229716,
+      "learning_rate": 2.1645105419084587e-06,
+      "loss": 0.7831,
+      "step": 4122
+    },
+    {
+      "epoch": 0.7931896883416699,
+      "grad_norm": 1.2201521009786982,
+      "learning_rate": 2.160640651089221e-06,
+      "loss": 0.848,
+      "step": 4123
+    },
+    {
+      "epoch": 0.7933820700269334,
+      "grad_norm": 1.4600055975182504,
+      "learning_rate": 2.1567738037791998e-06,
+      "loss": 0.8624,
+      "step": 4124
+    },
+    {
+      "epoch": 0.793574451712197,
+      "grad_norm": 1.3531433728932785,
+      "learning_rate": 2.152910001479638e-06,
+      "loss": 0.8515,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7937668333974606,
+      "grad_norm": 1.2722583579064843,
+      "learning_rate": 2.1490492456905964e-06,
+      "loss": 0.8361,
+      "step": 4126
+    },
+    {
+      "epoch": 0.7939592150827242,
+      "grad_norm": 1.1574025399056216,
+      "learning_rate": 2.1451915379109546e-06,
+      "loss": 0.822,
+      "step": 4127
+    },
+    {
+      "epoch": 0.7941515967679876,
+      "grad_norm": 1.6790536810641228,
+      "learning_rate": 2.141336879638406e-06,
+      "loss": 0.8583,
+      "step": 4128
+    },
+    {
+      "epoch": 0.7943439784532512,
+      "grad_norm": 1.211527488529336,
+      "learning_rate": 2.1374852723694595e-06,
+      "loss": 0.8346,
+      "step": 4129
+    },
+    {
+      "epoch": 0.7945363601385148,
+      "grad_norm": 1.8848315146500316,
+      "learning_rate": 2.133636717599451e-06,
+      "loss": 0.8592,
+      "step": 4130
+    },
+    {
+      "epoch": 0.7947287418237784,
+      "grad_norm": 1.18482770566985,
+      "learning_rate": 2.1297912168225086e-06,
+      "loss": 0.7659,
+      "step": 4131
+    },
+    {
+      "epoch": 0.794921123509042,
+      "grad_norm": 1.3803342044470692,
+      "learning_rate": 2.1259487715316e-06,
+      "loss": 0.797,
+      "step": 4132
+    },
+    {
+      "epoch": 0.7951135051943055,
+      "grad_norm": 0.564862720173882,
+      "learning_rate": 2.1221093832184903e-06,
+      "loss": 0.2778,
+      "step": 4133
+    },
+    {
+      "epoch": 0.795305886879569,
+      "grad_norm": 1.2819845663600398,
+      "learning_rate": 2.118273053373757e-06,
+      "loss": 0.8322,
+      "step": 4134
+    },
+    {
+      "epoch": 0.7954982685648326,
+      "grad_norm": 0.5944444396843248,
+      "learning_rate": 2.1144397834868034e-06,
+      "loss": 0.3225,
+      "step": 4135
+    },
+    {
+      "epoch": 0.7956906502500962,
+      "grad_norm": 1.3330804231659554,
+      "learning_rate": 2.1106095750458332e-06,
+      "loss": 0.8333,
+      "step": 4136
+    },
+    {
+      "epoch": 0.7958830319353598,
+      "grad_norm": 1.1869192863055475,
+      "learning_rate": 2.106782429537866e-06,
+      "loss": 0.8558,
+      "step": 4137
+    },
+    {
+      "epoch": 0.7960754136206233,
+      "grad_norm": 2.066492383985091,
+      "learning_rate": 2.1029583484487315e-06,
+      "loss": 0.8507,
+      "step": 4138
+    },
+    {
+      "epoch": 0.7962677953058869,
+      "grad_norm": 1.3944482983394608,
+      "learning_rate": 2.0991373332630683e-06,
+      "loss": 0.8406,
+      "step": 4139
+    },
+    {
+      "epoch": 0.7964601769911505,
+      "grad_norm": 1.350942751284433,
+      "learning_rate": 2.0953193854643274e-06,
+      "loss": 0.8046,
+      "step": 4140
+    },
+    {
+      "epoch": 0.796652558676414,
+      "grad_norm": 1.0983104749229748,
+      "learning_rate": 2.0915045065347673e-06,
+      "loss": 0.7992,
+      "step": 4141
+    },
+    {
+      "epoch": 0.7968449403616775,
+      "grad_norm": 1.2792025964173996,
+      "learning_rate": 2.0876926979554545e-06,
+      "loss": 0.8263,
+      "step": 4142
+    },
+    {
+      "epoch": 0.7970373220469411,
+      "grad_norm": 1.3207515141905999,
+      "learning_rate": 2.0838839612062633e-06,
+      "loss": 0.8152,
+      "step": 4143
+    },
+    {
+      "epoch": 0.7972297037322047,
+      "grad_norm": 1.150718560323727,
+      "learning_rate": 2.080078297765884e-06,
+      "loss": 0.8414,
+      "step": 4144
+    },
+    {
+      "epoch": 0.7974220854174683,
+      "grad_norm": 1.1896176139691503,
+      "learning_rate": 2.0762757091117937e-06,
+      "loss": 0.875,
+      "step": 4145
+    },
+    {
+      "epoch": 0.7976144671027319,
+      "grad_norm": 1.3261554977911436,
+      "learning_rate": 2.0724761967202987e-06,
+      "loss": 0.8475,
+      "step": 4146
+    },
+    {
+      "epoch": 0.7978068487879953,
+      "grad_norm": 1.0568836783760167,
+      "learning_rate": 2.0686797620664987e-06,
+      "loss": 0.8039,
+      "step": 4147
+    },
+    {
+      "epoch": 0.7979992304732589,
+      "grad_norm": 1.3809498160942968,
+      "learning_rate": 2.0648864066242937e-06,
+      "loss": 0.8612,
+      "step": 4148
+    },
+    {
+      "epoch": 0.7981916121585225,
+      "grad_norm": 1.3775558252119362,
+      "learning_rate": 2.0610961318664013e-06,
+      "loss": 0.8373,
+      "step": 4149
+    },
+    {
+      "epoch": 0.7983839938437861,
+      "grad_norm": 1.2258096509322622,
+      "learning_rate": 2.0573089392643362e-06,
+      "loss": 0.8318,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7985763755290496,
+      "grad_norm": 1.4122912335931477,
+      "learning_rate": 2.0535248302884147e-06,
+      "loss": 0.8403,
+      "step": 4151
+    },
+    {
+      "epoch": 0.7987687572143132,
+      "grad_norm": 1.3235407074862973,
+      "learning_rate": 2.0497438064077603e-06,
+      "loss": 0.8239,
+      "step": 4152
+    },
+    {
+      "epoch": 0.7989611388995768,
+      "grad_norm": 1.4545874649779096,
+      "learning_rate": 2.045965869090295e-06,
+      "loss": 0.8737,
+      "step": 4153
+    },
+    {
+      "epoch": 0.7991535205848403,
+      "grad_norm": 1.258052748081911,
+      "learning_rate": 2.0421910198027452e-06,
+      "loss": 0.8172,
+      "step": 4154
+    },
+    {
+      "epoch": 0.7993459022701039,
+      "grad_norm": 1.575563489373998,
+      "learning_rate": 2.0384192600106335e-06,
+      "loss": 0.8188,
+      "step": 4155
+    },
+    {
+      "epoch": 0.7995382839553674,
+      "grad_norm": 1.1100519618133924,
+      "learning_rate": 2.0346505911782956e-06,
+      "loss": 0.8449,
+      "step": 4156
+    },
+    {
+      "epoch": 0.799730665640631,
+      "grad_norm": 1.4995458985547094,
+      "learning_rate": 2.0308850147688484e-06,
+      "loss": 0.742,
+      "step": 4157
+    },
+    {
+      "epoch": 0.7999230473258946,
+      "grad_norm": 1.2224131932285185,
+      "learning_rate": 2.0271225322442255e-06,
+      "loss": 0.8446,
+      "step": 4158
+    },
+    {
+      "epoch": 0.8001154290111582,
+      "grad_norm": 1.2628513003609727,
+      "learning_rate": 2.0233631450651525e-06,
+      "loss": 0.8535,
+      "step": 4159
+    },
+    {
+      "epoch": 0.8003078106964217,
+      "grad_norm": 1.3523498715680673,
+      "learning_rate": 2.019606854691145e-06,
+      "loss": 0.8192,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8005001923816852,
+      "grad_norm": 1.4287797146693249,
+      "learning_rate": 2.0158536625805325e-06,
+      "loss": 0.8195,
+      "step": 4161
+    },
+    {
+      "epoch": 0.8006925740669488,
+      "grad_norm": 1.1326937231076348,
+      "learning_rate": 2.01210357019043e-06,
+      "loss": 0.8083,
+      "step": 4162
+    },
+    {
+      "epoch": 0.8008849557522124,
+      "grad_norm": 1.31496243289401,
+      "learning_rate": 2.008356578976752e-06,
+      "loss": 0.8175,
+      "step": 4163
+    },
+    {
+      "epoch": 0.801077337437476,
+      "grad_norm": 1.3456481139030412,
+      "learning_rate": 2.004612690394212e-06,
+      "loss": 0.8525,
+      "step": 4164
+    },
+    {
+      "epoch": 0.8012697191227395,
+      "grad_norm": 1.4443566136325054,
+      "learning_rate": 2.0008719058963144e-06,
+      "loss": 0.8391,
+      "step": 4165
+    },
+    {
+      "epoch": 0.8014621008080031,
+      "grad_norm": 1.4054157928554687,
+      "learning_rate": 1.997134226935361e-06,
+      "loss": 0.8416,
+      "step": 4166
+    },
+    {
+      "epoch": 0.8016544824932667,
+      "grad_norm": 1.4417727937279936,
+      "learning_rate": 1.9933996549624468e-06,
+      "loss": 0.8425,
+      "step": 4167
+    },
+    {
+      "epoch": 0.8018468641785302,
+      "grad_norm": 1.8513886119193195,
+      "learning_rate": 1.9896681914274616e-06,
+      "loss": 0.8366,
+      "step": 4168
+    },
+    {
+      "epoch": 0.8020392458637937,
+      "grad_norm": 1.4864748503723773,
+      "learning_rate": 1.9859398377790872e-06,
+      "loss": 0.8373,
+      "step": 4169
+    },
+    {
+      "epoch": 0.8022316275490573,
+      "grad_norm": 0.6167878322511651,
+      "learning_rate": 1.982214595464804e-06,
+      "loss": 0.3411,
+      "step": 4170
+    },
+    {
+      "epoch": 0.8024240092343209,
+      "grad_norm": 1.2658445447433775,
+      "learning_rate": 1.97849246593087e-06,
+      "loss": 0.8537,
+      "step": 4171
+    },
+    {
+      "epoch": 0.8026163909195845,
+      "grad_norm": 1.3447427067515934,
+      "learning_rate": 1.9747734506223525e-06,
+      "loss": 0.8184,
+      "step": 4172
+    },
+    {
+      "epoch": 0.8028087726048481,
+      "grad_norm": 0.6612855567881094,
+      "learning_rate": 1.9710575509831008e-06,
+      "loss": 0.3502,
+      "step": 4173
+    },
+    {
+      "epoch": 0.8030011542901115,
+      "grad_norm": 1.3197835019175161,
+      "learning_rate": 1.967344768455747e-06,
+      "loss": 0.8905,
+      "step": 4174
+    },
+    {
+      "epoch": 0.8031935359753751,
+      "grad_norm": 1.3254119802594415,
+      "learning_rate": 1.9636351044817292e-06,
+      "loss": 0.8006,
+      "step": 4175
+    },
+    {
+      "epoch": 0.8033859176606387,
+      "grad_norm": 1.324528090169969,
+      "learning_rate": 1.9599285605012643e-06,
+      "loss": 0.8514,
+      "step": 4176
+    },
+    {
+      "epoch": 0.8035782993459023,
+      "grad_norm": 1.3281886221047565,
+      "learning_rate": 1.9562251379533593e-06,
+      "loss": 0.8341,
+      "step": 4177
+    },
+    {
+      "epoch": 0.8037706810311658,
+      "grad_norm": 1.7313764385332469,
+      "learning_rate": 1.952524838275811e-06,
+      "loss": 0.8685,
+      "step": 4178
+    },
+    {
+      "epoch": 0.8039630627164294,
+      "grad_norm": 1.3550730248715943,
+      "learning_rate": 1.9488276629052026e-06,
+      "loss": 0.7821,
+      "step": 4179
+    },
+    {
+      "epoch": 0.804155444401693,
+      "grad_norm": 0.6039338120963583,
+      "learning_rate": 1.945133613276907e-06,
+      "loss": 0.3006,
+      "step": 4180
+    },
+    {
+      "epoch": 0.8043478260869565,
+      "grad_norm": 1.3589141904162014,
+      "learning_rate": 1.941442690825076e-06,
+      "loss": 0.8018,
+      "step": 4181
+    },
+    {
+      "epoch": 0.8045402077722201,
+      "grad_norm": 1.2636164528459277,
+      "learning_rate": 1.937754896982663e-06,
+      "loss": 0.8284,
+      "step": 4182
+    },
+    {
+      "epoch": 0.8047325894574836,
+      "grad_norm": 1.4716674462400516,
+      "learning_rate": 1.9340702331813842e-06,
+      "loss": 0.7605,
+      "step": 4183
+    },
+    {
+      "epoch": 0.8049249711427472,
+      "grad_norm": 1.5597135753617908,
+      "learning_rate": 1.9303887008517618e-06,
+      "loss": 0.8209,
+      "step": 4184
+    },
+    {
+      "epoch": 0.8051173528280108,
+      "grad_norm": 1.7117432233742345,
+      "learning_rate": 1.9267103014230935e-06,
+      "loss": 0.8052,
+      "step": 4185
+    },
+    {
+      "epoch": 0.8053097345132744,
+      "grad_norm": 1.1873116702166808,
+      "learning_rate": 1.923035036323452e-06,
+      "loss": 0.78,
+      "step": 4186
+    },
+    {
+      "epoch": 0.805502116198538,
+      "grad_norm": 1.8207457811244172,
+      "learning_rate": 1.91936290697971e-06,
+      "loss": 0.8086,
+      "step": 4187
+    },
+    {
+      "epoch": 0.8056944978838014,
+      "grad_norm": 1.1571536607076212,
+      "learning_rate": 1.9156939148175125e-06,
+      "loss": 0.8681,
+      "step": 4188
+    },
+    {
+      "epoch": 0.805886879569065,
+      "grad_norm": 1.2867259376340476,
+      "learning_rate": 1.9120280612612873e-06,
+      "loss": 0.8096,
+      "step": 4189
+    },
+    {
+      "epoch": 0.8060792612543286,
+      "grad_norm": 1.9457652240407572,
+      "learning_rate": 1.9083653477342467e-06,
+      "loss": 0.8157,
+      "step": 4190
+    },
+    {
+      "epoch": 0.8062716429395922,
+      "grad_norm": 1.2902887180017255,
+      "learning_rate": 1.904705775658381e-06,
+      "loss": 0.8299,
+      "step": 4191
+    },
+    {
+      "epoch": 0.8064640246248557,
+      "grad_norm": 1.5028191099597459,
+      "learning_rate": 1.9010493464544621e-06,
+      "loss": 0.8388,
+      "step": 4192
+    },
+    {
+      "epoch": 0.8066564063101193,
+      "grad_norm": 1.1905549242990152,
+      "learning_rate": 1.8973960615420416e-06,
+      "loss": 0.858,
+      "step": 4193
+    },
+    {
+      "epoch": 0.8068487879953828,
+      "grad_norm": 1.2910002900744197,
+      "learning_rate": 1.8937459223394517e-06,
+      "loss": 0.8292,
+      "step": 4194
+    },
+    {
+      "epoch": 0.8070411696806464,
+      "grad_norm": 1.275063760644768,
+      "learning_rate": 1.8900989302637985e-06,
+      "loss": 0.8159,
+      "step": 4195
+    },
+    {
+      "epoch": 0.80723355136591,
+      "grad_norm": 1.1731916574407457,
+      "learning_rate": 1.8864550867309771e-06,
+      "loss": 0.8125,
+      "step": 4196
+    },
+    {
+      "epoch": 0.8074259330511735,
+      "grad_norm": 1.2566668728177086,
+      "learning_rate": 1.8828143931556442e-06,
+      "loss": 0.861,
+      "step": 4197
+    },
+    {
+      "epoch": 0.8076183147364371,
+      "grad_norm": 1.1454914309999045,
+      "learning_rate": 1.8791768509512487e-06,
+      "loss": 0.8007,
+      "step": 4198
+    },
+    {
+      "epoch": 0.8078106964217007,
+      "grad_norm": 1.6941344467686321,
+      "learning_rate": 1.875542461530011e-06,
+      "loss": 0.8104,
+      "step": 4199
+    },
+    {
+      "epoch": 0.8080030781069643,
+      "grad_norm": 1.3229970483472306,
+      "learning_rate": 1.871911226302917e-06,
+      "loss": 0.8341,
+      "step": 4200
+    },
+    {
+      "epoch": 0.8081954597922277,
+      "grad_norm": 3.3059401724150543,
+      "learning_rate": 1.868283146679747e-06,
+      "loss": 0.8248,
+      "step": 4201
+    },
+    {
+      "epoch": 0.8083878414774913,
+      "grad_norm": 1.3325791811136791,
+      "learning_rate": 1.8646582240690414e-06,
+      "loss": 0.809,
+      "step": 4202
+    },
+    {
+      "epoch": 0.8085802231627549,
+      "grad_norm": 1.3310942345219399,
+      "learning_rate": 1.8610364598781227e-06,
+      "loss": 0.7467,
+      "step": 4203
+    },
+    {
+      "epoch": 0.8087726048480185,
+      "grad_norm": 1.4976622373318929,
+      "learning_rate": 1.8574178555130818e-06,
+      "loss": 0.9037,
+      "step": 4204
+    },
+    {
+      "epoch": 0.8089649865332821,
+      "grad_norm": 1.3331609193052567,
+      "learning_rate": 1.8538024123787868e-06,
+      "loss": 0.8741,
+      "step": 4205
+    },
+    {
+      "epoch": 0.8091573682185456,
+      "grad_norm": 1.5628046669540803,
+      "learning_rate": 1.8501901318788773e-06,
+      "loss": 0.8655,
+      "step": 4206
+    },
+    {
+      "epoch": 0.8093497499038091,
+      "grad_norm": 1.2373110724859793,
+      "learning_rate": 1.8465810154157626e-06,
+      "loss": 0.7928,
+      "step": 4207
+    },
+    {
+      "epoch": 0.8095421315890727,
+      "grad_norm": 1.4417328351056093,
+      "learning_rate": 1.8429750643906331e-06,
+      "loss": 0.7809,
+      "step": 4208
+    },
+    {
+      "epoch": 0.8097345132743363,
+      "grad_norm": 1.3279307085967946,
+      "learning_rate": 1.8393722802034331e-06,
+      "loss": 0.8307,
+      "step": 4209
+    },
+    {
+      "epoch": 0.8099268949595998,
+      "grad_norm": 1.3112677960546018,
+      "learning_rate": 1.835772664252895e-06,
+      "loss": 0.846,
+      "step": 4210
+    },
+    {
+      "epoch": 0.8101192766448634,
+      "grad_norm": 1.2937889261906124,
+      "learning_rate": 1.832176217936511e-06,
+      "loss": 0.7998,
+      "step": 4211
+    },
+    {
+      "epoch": 0.810311658330127,
+      "grad_norm": 1.2992259175514977,
+      "learning_rate": 1.8285829426505453e-06,
+      "loss": 0.8669,
+      "step": 4212
+    },
+    {
+      "epoch": 0.8105040400153906,
+      "grad_norm": 1.1608155113102876,
+      "learning_rate": 1.8249928397900351e-06,
+      "loss": 0.8523,
+      "step": 4213
+    },
+    {
+      "epoch": 0.8106964217006541,
+      "grad_norm": 1.4289038145703552,
+      "learning_rate": 1.8214059107487726e-06,
+      "loss": 0.8797,
+      "step": 4214
+    },
+    {
+      "epoch": 0.8108888033859176,
+      "grad_norm": 1.538889112502739,
+      "learning_rate": 1.8178221569193343e-06,
+      "loss": 0.8332,
+      "step": 4215
+    },
+    {
+      "epoch": 0.8110811850711812,
+      "grad_norm": 1.2932924407276019,
+      "learning_rate": 1.8142415796930568e-06,
+      "loss": 0.8606,
+      "step": 4216
+    },
+    {
+      "epoch": 0.8112735667564448,
+      "grad_norm": 0.6232132672587252,
+      "learning_rate": 1.8106641804600411e-06,
+      "loss": 0.3139,
+      "step": 4217
+    },
+    {
+      "epoch": 0.8114659484417084,
+      "grad_norm": 1.3782479641787517,
+      "learning_rate": 1.8070899606091586e-06,
+      "loss": 0.8357,
+      "step": 4218
+    },
+    {
+      "epoch": 0.8116583301269719,
+      "grad_norm": 1.2413776344752503,
+      "learning_rate": 1.8035189215280423e-06,
+      "loss": 0.8072,
+      "step": 4219
+    },
+    {
+      "epoch": 0.8118507118122354,
+      "grad_norm": 1.2733528313751947,
+      "learning_rate": 1.799951064603095e-06,
+      "loss": 0.817,
+      "step": 4220
+    },
+    {
+      "epoch": 0.812043093497499,
+      "grad_norm": 1.2197514387852104,
+      "learning_rate": 1.7963863912194768e-06,
+      "loss": 0.8768,
+      "step": 4221
+    },
+    {
+      "epoch": 0.8122354751827626,
+      "grad_norm": 1.2572403071706484,
+      "learning_rate": 1.7928249027611255e-06,
+      "loss": 0.8279,
+      "step": 4222
+    },
+    {
+      "epoch": 0.8124278568680262,
+      "grad_norm": 0.6397987243501572,
+      "learning_rate": 1.789266600610724e-06,
+      "loss": 0.3387,
+      "step": 4223
+    },
+    {
+      "epoch": 0.8126202385532897,
+      "grad_norm": 1.4186947952399187,
+      "learning_rate": 1.7857114861497337e-06,
+      "loss": 0.8318,
+      "step": 4224
+    },
+    {
+      "epoch": 0.8128126202385533,
+      "grad_norm": 0.6175029528691907,
+      "learning_rate": 1.782159560758373e-06,
+      "loss": 0.3492,
+      "step": 4225
+    },
+    {
+      "epoch": 0.8130050019238169,
+      "grad_norm": 1.5307598806647231,
+      "learning_rate": 1.7786108258156154e-06,
+      "loss": 0.894,
+      "step": 4226
+    },
+    {
+      "epoch": 0.8131973836090804,
+      "grad_norm": 2.191848968856375,
+      "learning_rate": 1.7750652826992077e-06,
+      "loss": 0.8156,
+      "step": 4227
+    },
+    {
+      "epoch": 0.8133897652943439,
+      "grad_norm": 1.405638178832801,
+      "learning_rate": 1.7715229327856498e-06,
+      "loss": 0.8541,
+      "step": 4228
+    },
+    {
+      "epoch": 0.8135821469796075,
+      "grad_norm": 1.426809674178331,
+      "learning_rate": 1.7679837774502052e-06,
+      "loss": 0.7831,
+      "step": 4229
+    },
+    {
+      "epoch": 0.8137745286648711,
+      "grad_norm": 1.270448876090481,
+      "learning_rate": 1.7644478180668945e-06,
+      "loss": 0.8532,
+      "step": 4230
+    },
+    {
+      "epoch": 0.8139669103501347,
+      "grad_norm": 1.4527341221751375,
+      "learning_rate": 1.7609150560084986e-06,
+      "loss": 0.7807,
+      "step": 4231
+    },
+    {
+      "epoch": 0.8141592920353983,
+      "grad_norm": 1.5689870966595438,
+      "learning_rate": 1.7573854926465582e-06,
+      "loss": 0.8465,
+      "step": 4232
+    },
+    {
+      "epoch": 0.8143516737206618,
+      "grad_norm": 1.4479457752446094,
+      "learning_rate": 1.7538591293513685e-06,
+      "loss": 0.8612,
+      "step": 4233
+    },
+    {
+      "epoch": 0.8145440554059253,
+      "grad_norm": 1.3973973128203614,
+      "learning_rate": 1.7503359674919929e-06,
+      "loss": 0.8397,
+      "step": 4234
+    },
+    {
+      "epoch": 0.8147364370911889,
+      "grad_norm": 1.1316697595740093,
+      "learning_rate": 1.746816008436234e-06,
+      "loss": 0.8539,
+      "step": 4235
+    },
+    {
+      "epoch": 0.8149288187764525,
+      "grad_norm": 1.8259658615705157,
+      "learning_rate": 1.7432992535506687e-06,
+      "loss": 0.8311,
+      "step": 4236
+    },
+    {
+      "epoch": 0.8151212004617161,
+      "grad_norm": 1.2634826217821937,
+      "learning_rate": 1.7397857042006194e-06,
+      "loss": 0.8132,
+      "step": 4237
+    },
+    {
+      "epoch": 0.8153135821469796,
+      "grad_norm": 0.5786776708408803,
+      "learning_rate": 1.736275361750167e-06,
+      "loss": 0.3296,
+      "step": 4238
+    },
+    {
+      "epoch": 0.8155059638322432,
+      "grad_norm": 1.7114682953120701,
+      "learning_rate": 1.7327682275621506e-06,
+      "loss": 0.8463,
+      "step": 4239
+    },
+    {
+      "epoch": 0.8156983455175068,
+      "grad_norm": 1.5805822759970813,
+      "learning_rate": 1.7292643029981525e-06,
+      "loss": 0.8944,
+      "step": 4240
+    },
+    {
+      "epoch": 0.8158907272027703,
+      "grad_norm": 1.370429601765713,
+      "learning_rate": 1.7257635894185232e-06,
+      "loss": 0.8685,
+      "step": 4241
+    },
+    {
+      "epoch": 0.8160831088880338,
+      "grad_norm": 2.4239706789468034,
+      "learning_rate": 1.7222660881823594e-06,
+      "loss": 0.8661,
+      "step": 4242
+    },
+    {
+      "epoch": 0.8162754905732974,
+      "grad_norm": 1.3195958349769,
+      "learning_rate": 1.7187718006475117e-06,
+      "loss": 0.8122,
+      "step": 4243
+    },
+    {
+      "epoch": 0.816467872258561,
+      "grad_norm": 0.5991963275515751,
+      "learning_rate": 1.7152807281705809e-06,
+      "loss": 0.3099,
+      "step": 4244
+    },
+    {
+      "epoch": 0.8166602539438246,
+      "grad_norm": 1.2812851604489055,
+      "learning_rate": 1.7117928721069233e-06,
+      "loss": 0.8807,
+      "step": 4245
+    },
+    {
+      "epoch": 0.8168526356290882,
+      "grad_norm": 0.6115389715973188,
+      "learning_rate": 1.708308233810644e-06,
+      "loss": 0.3178,
+      "step": 4246
+    },
+    {
+      "epoch": 0.8170450173143516,
+      "grad_norm": 1.1230377343859588,
+      "learning_rate": 1.704826814634597e-06,
+      "loss": 0.8413,
+      "step": 4247
+    },
+    {
+      "epoch": 0.8172373989996152,
+      "grad_norm": 1.2199338465974547,
+      "learning_rate": 1.701348615930397e-06,
+      "loss": 0.8387,
+      "step": 4248
+    },
+    {
+      "epoch": 0.8174297806848788,
+      "grad_norm": 1.3148732125994866,
+      "learning_rate": 1.6978736390483896e-06,
+      "loss": 0.7988,
+      "step": 4249
+    },
+    {
+      "epoch": 0.8176221623701424,
+      "grad_norm": 1.1080261552191117,
+      "learning_rate": 1.6944018853376898e-06,
+      "loss": 0.7767,
+      "step": 4250
+    },
+    {
+      "epoch": 0.8178145440554059,
+      "grad_norm": 1.7051311035750827,
+      "learning_rate": 1.6909333561461471e-06,
+      "loss": 0.7972,
+      "step": 4251
+    },
+    {
+      "epoch": 0.8180069257406695,
+      "grad_norm": 1.2326618986004076,
+      "learning_rate": 1.6874680528203657e-06,
+      "loss": 0.8766,
+      "step": 4252
+    },
+    {
+      "epoch": 0.818199307425933,
+      "grad_norm": 1.3722833115088204,
+      "learning_rate": 1.6840059767056949e-06,
+      "loss": 0.8672,
+      "step": 4253
+    },
+    {
+      "epoch": 0.8183916891111966,
+      "grad_norm": 1.282971084631286,
+      "learning_rate": 1.6805471291462316e-06,
+      "loss": 0.8156,
+      "step": 4254
+    },
+    {
+      "epoch": 0.8185840707964602,
+      "grad_norm": 1.36886184650109,
+      "learning_rate": 1.6770915114848197e-06,
+      "loss": 0.7969,
+      "step": 4255
+    },
+    {
+      "epoch": 0.8187764524817237,
+      "grad_norm": 1.3752554029221347,
+      "learning_rate": 1.67363912506305e-06,
+      "loss": 0.8104,
+      "step": 4256
+    },
+    {
+      "epoch": 0.8189688341669873,
+      "grad_norm": 1.2055406331198733,
+      "learning_rate": 1.6701899712212565e-06,
+      "loss": 0.8219,
+      "step": 4257
+    },
+    {
+      "epoch": 0.8191612158522509,
+      "grad_norm": 0.5822470591089137,
+      "learning_rate": 1.66674405129852e-06,
+      "loss": 0.3196,
+      "step": 4258
+    },
+    {
+      "epoch": 0.8193535975375145,
+      "grad_norm": 1.5247610661463094,
+      "learning_rate": 1.6633013666326636e-06,
+      "loss": 0.8659,
+      "step": 4259
+    },
+    {
+      "epoch": 0.8195459792227779,
+      "grad_norm": 1.508398677116466,
+      "learning_rate": 1.6598619185602616e-06,
+      "loss": 0.8597,
+      "step": 4260
+    },
+    {
+      "epoch": 0.8197383609080415,
+      "grad_norm": 1.2753341694314924,
+      "learning_rate": 1.656425708416617e-06,
+      "loss": 0.8135,
+      "step": 4261
+    },
+    {
+      "epoch": 0.8199307425933051,
+      "grad_norm": 1.2755616612980674,
+      "learning_rate": 1.6529927375357957e-06,
+      "loss": 0.8307,
+      "step": 4262
+    },
+    {
+      "epoch": 0.8201231242785687,
+      "grad_norm": 1.3455651722311182,
+      "learning_rate": 1.6495630072505841e-06,
+      "loss": 0.8736,
+      "step": 4263
+    },
+    {
+      "epoch": 0.8203155059638323,
+      "grad_norm": 1.241245652947746,
+      "learning_rate": 1.6461365188925304e-06,
+      "loss": 0.8457,
+      "step": 4264
+    },
+    {
+      "epoch": 0.8205078876490958,
+      "grad_norm": 1.1798879518592937,
+      "learning_rate": 1.642713273791914e-06,
+      "loss": 0.7957,
+      "step": 4265
+    },
+    {
+      "epoch": 0.8207002693343594,
+      "grad_norm": 1.6368889595923823,
+      "learning_rate": 1.6392932732777489e-06,
+      "loss": 0.8017,
+      "step": 4266
+    },
+    {
+      "epoch": 0.8208926510196229,
+      "grad_norm": 1.1826012893896825,
+      "learning_rate": 1.6358765186778057e-06,
+      "loss": 0.8046,
+      "step": 4267
+    },
+    {
+      "epoch": 0.8210850327048865,
+      "grad_norm": 1.3031145763391605,
+      "learning_rate": 1.6324630113185835e-06,
+      "loss": 0.8245,
+      "step": 4268
+    },
+    {
+      "epoch": 0.82127741439015,
+      "grad_norm": 1.3358262418880678,
+      "learning_rate": 1.629052752525323e-06,
+      "loss": 0.8088,
+      "step": 4269
+    },
+    {
+      "epoch": 0.8214697960754136,
+      "grad_norm": 1.495372674432775,
+      "learning_rate": 1.625645743622003e-06,
+      "loss": 0.8474,
+      "step": 4270
+    },
+    {
+      "epoch": 0.8216621777606772,
+      "grad_norm": 1.5003962115322966,
+      "learning_rate": 1.6222419859313443e-06,
+      "loss": 0.9056,
+      "step": 4271
+    },
+    {
+      "epoch": 0.8218545594459408,
+      "grad_norm": 1.438082136366564,
+      "learning_rate": 1.6188414807747999e-06,
+      "loss": 0.8755,
+      "step": 4272
+    },
+    {
+      "epoch": 0.8220469411312044,
+      "grad_norm": 1.1867896364931425,
+      "learning_rate": 1.6154442294725636e-06,
+      "loss": 0.8749,
+      "step": 4273
+    },
+    {
+      "epoch": 0.8222393228164678,
+      "grad_norm": 1.3851004201280317,
+      "learning_rate": 1.6120502333435695e-06,
+      "loss": 0.9181,
+      "step": 4274
+    },
+    {
+      "epoch": 0.8224317045017314,
+      "grad_norm": 1.1950548205252782,
+      "learning_rate": 1.6086594937054767e-06,
+      "loss": 0.8786,
+      "step": 4275
+    },
+    {
+      "epoch": 0.822624086186995,
+      "grad_norm": 1.343140945354784,
+      "learning_rate": 1.6052720118746923e-06,
+      "loss": 0.8331,
+      "step": 4276
+    },
+    {
+      "epoch": 0.8228164678722586,
+      "grad_norm": 1.1392767139547104,
+      "learning_rate": 1.6018877891663521e-06,
+      "loss": 0.8076,
+      "step": 4277
+    },
+    {
+      "epoch": 0.8230088495575221,
+      "grad_norm": 2.0827330799544574,
+      "learning_rate": 1.5985068268943283e-06,
+      "loss": 0.7902,
+      "step": 4278
+    },
+    {
+      "epoch": 0.8232012312427857,
+      "grad_norm": 1.3378434962589962,
+      "learning_rate": 1.5951291263712255e-06,
+      "loss": 0.8649,
+      "step": 4279
+    },
+    {
+      "epoch": 0.8233936129280492,
+      "grad_norm": 1.186443486049424,
+      "learning_rate": 1.5917546889083834e-06,
+      "loss": 0.7474,
+      "step": 4280
+    },
+    {
+      "epoch": 0.8235859946133128,
+      "grad_norm": 1.127976013265005,
+      "learning_rate": 1.5883835158158767e-06,
+      "loss": 0.805,
+      "step": 4281
+    },
+    {
+      "epoch": 0.8237783762985764,
+      "grad_norm": 1.3168349370363461,
+      "learning_rate": 1.5850156084025091e-06,
+      "loss": 0.8202,
+      "step": 4282
+    },
+    {
+      "epoch": 0.8239707579838399,
+      "grad_norm": 0.5612209524937724,
+      "learning_rate": 1.5816509679758185e-06,
+      "loss": 0.3582,
+      "step": 4283
+    },
+    {
+      "epoch": 0.8241631396691035,
+      "grad_norm": 1.2197489843656368,
+      "learning_rate": 1.578289595842074e-06,
+      "loss": 0.7824,
+      "step": 4284
+    },
+    {
+      "epoch": 0.8243555213543671,
+      "grad_norm": 0.5978998912271906,
+      "learning_rate": 1.5749314933062754e-06,
+      "loss": 0.3231,
+      "step": 4285
+    },
+    {
+      "epoch": 0.8245479030396307,
+      "grad_norm": 1.2763482342479897,
+      "learning_rate": 1.5715766616721584e-06,
+      "loss": 0.8296,
+      "step": 4286
+    },
+    {
+      "epoch": 0.8247402847248941,
+      "grad_norm": 1.4116241543868835,
+      "learning_rate": 1.5682251022421757e-06,
+      "loss": 0.7881,
+      "step": 4287
+    },
+    {
+      "epoch": 0.8249326664101577,
+      "grad_norm": 1.1665881397511328,
+      "learning_rate": 1.5648768163175277e-06,
+      "loss": 0.7849,
+      "step": 4288
+    },
+    {
+      "epoch": 0.8251250480954213,
+      "grad_norm": 1.376796916292442,
+      "learning_rate": 1.5615318051981243e-06,
+      "loss": 0.8712,
+      "step": 4289
+    },
+    {
+      "epoch": 0.8253174297806849,
+      "grad_norm": 1.396569382055829,
+      "learning_rate": 1.5581900701826226e-06,
+      "loss": 0.7886,
+      "step": 4290
+    },
+    {
+      "epoch": 0.8255098114659485,
+      "grad_norm": 1.1277150852661486,
+      "learning_rate": 1.5548516125683976e-06,
+      "loss": 0.8593,
+      "step": 4291
+    },
+    {
+      "epoch": 0.825702193151212,
+      "grad_norm": 1.2292183207118745,
+      "learning_rate": 1.5515164336515465e-06,
+      "loss": 0.7822,
+      "step": 4292
+    },
+    {
+      "epoch": 0.8258945748364755,
+      "grad_norm": 1.1494973413483132,
+      "learning_rate": 1.5481845347269077e-06,
+      "loss": 0.8116,
+      "step": 4293
+    },
+    {
+      "epoch": 0.8260869565217391,
+      "grad_norm": 1.2415006509017288,
+      "learning_rate": 1.5448559170880373e-06,
+      "loss": 0.8119,
+      "step": 4294
+    },
+    {
+      "epoch": 0.8262793382070027,
+      "grad_norm": 1.432095360379004,
+      "learning_rate": 1.5415305820272198e-06,
+      "loss": 0.8777,
+      "step": 4295
+    },
+    {
+      "epoch": 0.8264717198922663,
+      "grad_norm": 1.2503687317477314,
+      "learning_rate": 1.5382085308354633e-06,
+      "loss": 0.8462,
+      "step": 4296
+    },
+    {
+      "epoch": 0.8266641015775298,
+      "grad_norm": 1.3028004921388292,
+      "learning_rate": 1.534889764802503e-06,
+      "loss": 0.7724,
+      "step": 4297
+    },
+    {
+      "epoch": 0.8268564832627934,
+      "grad_norm": 1.1826197461056143,
+      "learning_rate": 1.5315742852167992e-06,
+      "loss": 0.8458,
+      "step": 4298
+    },
+    {
+      "epoch": 0.827048864948057,
+      "grad_norm": 1.5406353492946838,
+      "learning_rate": 1.5282620933655312e-06,
+      "loss": 0.851,
+      "step": 4299
+    },
+    {
+      "epoch": 0.8272412466333205,
+      "grad_norm": 1.2573499380110715,
+      "learning_rate": 1.5249531905346138e-06,
+      "loss": 0.8699,
+      "step": 4300
+    },
+    {
+      "epoch": 0.827433628318584,
+      "grad_norm": 1.2147732642489817,
+      "learning_rate": 1.521647578008667e-06,
+      "loss": 0.8162,
+      "step": 4301
+    },
+    {
+      "epoch": 0.8276260100038476,
+      "grad_norm": 1.315168419300271,
+      "learning_rate": 1.5183452570710522e-06,
+      "loss": 0.806,
+      "step": 4302
+    },
+    {
+      "epoch": 0.8278183916891112,
+      "grad_norm": 1.0922537803941155,
+      "learning_rate": 1.5150462290038392e-06,
+      "loss": 0.8396,
+      "step": 4303
+    },
+    {
+      "epoch": 0.8280107733743748,
+      "grad_norm": 1.6060281467277653,
+      "learning_rate": 1.511750495087827e-06,
+      "loss": 0.7819,
+      "step": 4304
+    },
+    {
+      "epoch": 0.8282031550596384,
+      "grad_norm": 1.2764752724526691,
+      "learning_rate": 1.5084580566025309e-06,
+      "loss": 0.8362,
+      "step": 4305
+    },
+    {
+      "epoch": 0.8283955367449019,
+      "grad_norm": 1.2155322743365269,
+      "learning_rate": 1.5051689148261895e-06,
+      "loss": 0.7973,
+      "step": 4306
+    },
+    {
+      "epoch": 0.8285879184301654,
+      "grad_norm": 1.5601801336049712,
+      "learning_rate": 1.5018830710357612e-06,
+      "loss": 0.8201,
+      "step": 4307
+    },
+    {
+      "epoch": 0.828780300115429,
+      "grad_norm": 1.496417040964999,
+      "learning_rate": 1.4986005265069204e-06,
+      "loss": 0.796,
+      "step": 4308
+    },
+    {
+      "epoch": 0.8289726818006926,
+      "grad_norm": 1.3279273309168684,
+      "learning_rate": 1.4953212825140728e-06,
+      "loss": 0.8541,
+      "step": 4309
+    },
+    {
+      "epoch": 0.8291650634859561,
+      "grad_norm": 1.2901991735833307,
+      "learning_rate": 1.4920453403303249e-06,
+      "loss": 0.8368,
+      "step": 4310
+    },
+    {
+      "epoch": 0.8293574451712197,
+      "grad_norm": 1.3473959926839756,
+      "learning_rate": 1.4887727012275112e-06,
+      "loss": 0.8469,
+      "step": 4311
+    },
+    {
+      "epoch": 0.8295498268564833,
+      "grad_norm": 1.223740314961288,
+      "learning_rate": 1.4855033664761898e-06,
+      "loss": 0.8062,
+      "step": 4312
+    },
+    {
+      "epoch": 0.8297422085417469,
+      "grad_norm": 1.056708177769068,
+      "learning_rate": 1.48223733734562e-06,
+      "loss": 0.7774,
+      "step": 4313
+    },
+    {
+      "epoch": 0.8299345902270104,
+      "grad_norm": 1.1778881280282303,
+      "learning_rate": 1.4789746151037942e-06,
+      "loss": 0.7698,
+      "step": 4314
+    },
+    {
+      "epoch": 0.8301269719122739,
+      "grad_norm": 1.274883003668969,
+      "learning_rate": 1.475715201017407e-06,
+      "loss": 0.8433,
+      "step": 4315
+    },
+    {
+      "epoch": 0.8303193535975375,
+      "grad_norm": 1.5578308416007869,
+      "learning_rate": 1.4724590963518803e-06,
+      "loss": 0.8535,
+      "step": 4316
+    },
+    {
+      "epoch": 0.8305117352828011,
+      "grad_norm": 1.3528413493373947,
+      "learning_rate": 1.4692063023713444e-06,
+      "loss": 0.8451,
+      "step": 4317
+    },
+    {
+      "epoch": 0.8307041169680647,
+      "grad_norm": 1.3161050193789328,
+      "learning_rate": 1.4659568203386464e-06,
+      "loss": 0.7954,
+      "step": 4318
+    },
+    {
+      "epoch": 0.8308964986533282,
+      "grad_norm": 1.5150165449900108,
+      "learning_rate": 1.4627106515153456e-06,
+      "loss": 0.8323,
+      "step": 4319
+    },
+    {
+      "epoch": 0.8310888803385917,
+      "grad_norm": 3.4885946052981622,
+      "learning_rate": 1.4594677971617178e-06,
+      "loss": 0.8253,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8312812620238553,
+      "grad_norm": 1.220962147895295,
+      "learning_rate": 1.4562282585367493e-06,
+      "loss": 0.8386,
+      "step": 4321
+    },
+    {
+      "epoch": 0.8314736437091189,
+      "grad_norm": 2.8228923756885917,
+      "learning_rate": 1.452992036898142e-06,
+      "loss": 0.767,
+      "step": 4322
+    },
+    {
+      "epoch": 0.8316660253943825,
+      "grad_norm": 1.2787141542004676,
+      "learning_rate": 1.4497591335023087e-06,
+      "loss": 0.8619,
+      "step": 4323
+    },
+    {
+      "epoch": 0.831858407079646,
+      "grad_norm": 1.2223687484492587,
+      "learning_rate": 1.446529549604373e-06,
+      "loss": 0.7834,
+      "step": 4324
+    },
+    {
+      "epoch": 0.8320507887649096,
+      "grad_norm": 1.7185620469549625,
+      "learning_rate": 1.4433032864581687e-06,
+      "loss": 0.8491,
+      "step": 4325
+    },
+    {
+      "epoch": 0.8322431704501732,
+      "grad_norm": 1.3780450506887203,
+      "learning_rate": 1.4400803453162482e-06,
+      "loss": 0.8284,
+      "step": 4326
+    },
+    {
+      "epoch": 0.8324355521354367,
+      "grad_norm": 1.1442148659567297,
+      "learning_rate": 1.4368607274298596e-06,
+      "loss": 0.8644,
+      "step": 4327
+    },
+    {
+      "epoch": 0.8326279338207002,
+      "grad_norm": 1.2604111700417795,
+      "learning_rate": 1.4336444340489775e-06,
+      "loss": 0.8196,
+      "step": 4328
+    },
+    {
+      "epoch": 0.8328203155059638,
+      "grad_norm": 1.215132377329883,
+      "learning_rate": 1.430431466422273e-06,
+      "loss": 0.8158,
+      "step": 4329
+    },
+    {
+      "epoch": 0.8330126971912274,
+      "grad_norm": 1.2161985134026714,
+      "learning_rate": 1.4272218257971327e-06,
+      "loss": 0.8618,
+      "step": 4330
+    },
+    {
+      "epoch": 0.833205078876491,
+      "grad_norm": 1.3216943284588822,
+      "learning_rate": 1.4240155134196499e-06,
+      "loss": 0.8775,
+      "step": 4331
+    },
+    {
+      "epoch": 0.8333974605617546,
+      "grad_norm": 1.5514399894070818,
+      "learning_rate": 1.4208125305346232e-06,
+      "loss": 0.8521,
+      "step": 4332
+    },
+    {
+      "epoch": 0.833589842247018,
+      "grad_norm": 1.3731913340036075,
+      "learning_rate": 1.4176128783855636e-06,
+      "loss": 0.8376,
+      "step": 4333
+    },
+    {
+      "epoch": 0.8337822239322816,
+      "grad_norm": 1.2641169803039711,
+      "learning_rate": 1.4144165582146819e-06,
+      "loss": 0.8443,
+      "step": 4334
+    },
+    {
+      "epoch": 0.8339746056175452,
+      "grad_norm": 1.3363185330511715,
+      "learning_rate": 1.4112235712629063e-06,
+      "loss": 0.8317,
+      "step": 4335
+    },
+    {
+      "epoch": 0.8341669873028088,
+      "grad_norm": 1.341969159945516,
+      "learning_rate": 1.40803391876986e-06,
+      "loss": 0.8522,
+      "step": 4336
+    },
+    {
+      "epoch": 0.8343593689880723,
+      "grad_norm": 1.3526365906155748,
+      "learning_rate": 1.4048476019738756e-06,
+      "loss": 0.852,
+      "step": 4337
+    },
+    {
+      "epoch": 0.8345517506733359,
+      "grad_norm": 1.2530768331653093,
+      "learning_rate": 1.4016646221119912e-06,
+      "loss": 0.7919,
+      "step": 4338
+    },
+    {
+      "epoch": 0.8347441323585995,
+      "grad_norm": 1.3035948432553242,
+      "learning_rate": 1.3984849804199485e-06,
+      "loss": 0.7883,
+      "step": 4339
+    },
+    {
+      "epoch": 0.834936514043863,
+      "grad_norm": 1.2390382222585958,
+      "learning_rate": 1.395308678132199e-06,
+      "loss": 0.8321,
+      "step": 4340
+    },
+    {
+      "epoch": 0.8351288957291266,
+      "grad_norm": 1.365906915996268,
+      "learning_rate": 1.392135716481885e-06,
+      "loss": 0.8066,
+      "step": 4341
+    },
+    {
+      "epoch": 0.8353212774143901,
+      "grad_norm": 1.3554202267257252,
+      "learning_rate": 1.3889660967008656e-06,
+      "loss": 0.7925,
+      "step": 4342
+    },
+    {
+      "epoch": 0.8355136590996537,
+      "grad_norm": 1.431452684787873,
+      "learning_rate": 1.3857998200196943e-06,
+      "loss": 0.8141,
+      "step": 4343
+    },
+    {
+      "epoch": 0.8357060407849173,
+      "grad_norm": 1.3427020870157398,
+      "learning_rate": 1.3826368876676278e-06,
+      "loss": 0.8307,
+      "step": 4344
+    },
+    {
+      "epoch": 0.8358984224701809,
+      "grad_norm": 1.126380119990642,
+      "learning_rate": 1.379477300872626e-06,
+      "loss": 0.8473,
+      "step": 4345
+    },
+    {
+      "epoch": 0.8360908041554445,
+      "grad_norm": 1.3190044969720713,
+      "learning_rate": 1.3763210608613497e-06,
+      "loss": 0.8324,
+      "step": 4346
+    },
+    {
+      "epoch": 0.8362831858407079,
+      "grad_norm": 1.1644975937923776,
+      "learning_rate": 1.3731681688591593e-06,
+      "loss": 0.827,
+      "step": 4347
+    },
+    {
+      "epoch": 0.8364755675259715,
+      "grad_norm": 2.024967048431252,
+      "learning_rate": 1.370018626090116e-06,
+      "loss": 0.807,
+      "step": 4348
+    },
+    {
+      "epoch": 0.8366679492112351,
+      "grad_norm": 1.5602867814309966,
+      "learning_rate": 1.3668724337769823e-06,
+      "loss": 0.8756,
+      "step": 4349
+    },
+    {
+      "epoch": 0.8368603308964987,
+      "grad_norm": 1.4354710773756245,
+      "learning_rate": 1.3637295931412153e-06,
+      "loss": 0.9121,
+      "step": 4350
+    },
+    {
+      "epoch": 0.8370527125817622,
+      "grad_norm": 1.2305365002856783,
+      "learning_rate": 1.3605901054029746e-06,
+      "loss": 0.8449,
+      "step": 4351
+    },
+    {
+      "epoch": 0.8372450942670258,
+      "grad_norm": 1.2631219124169628,
+      "learning_rate": 1.3574539717811231e-06,
+      "loss": 0.8228,
+      "step": 4352
+    },
+    {
+      "epoch": 0.8374374759522893,
+      "grad_norm": 1.1279193880576983,
+      "learning_rate": 1.3543211934932065e-06,
+      "loss": 0.8273,
+      "step": 4353
+    },
+    {
+      "epoch": 0.8376298576375529,
+      "grad_norm": 1.4259244677387688,
+      "learning_rate": 1.3511917717554846e-06,
+      "loss": 0.8452,
+      "step": 4354
+    },
+    {
+      "epoch": 0.8378222393228165,
+      "grad_norm": 1.294607628833859,
+      "learning_rate": 1.348065707782904e-06,
+      "loss": 0.8356,
+      "step": 4355
+    },
+    {
+      "epoch": 0.83801462100808,
+      "grad_norm": 1.2167795714428333,
+      "learning_rate": 1.3449430027891096e-06,
+      "loss": 0.814,
+      "step": 4356
+    },
+    {
+      "epoch": 0.8382070026933436,
+      "grad_norm": 1.2058272053700498,
+      "learning_rate": 1.3418236579864452e-06,
+      "loss": 0.8957,
+      "step": 4357
+    },
+    {
+      "epoch": 0.8383993843786072,
+      "grad_norm": 1.5229137906843788,
+      "learning_rate": 1.338707674585945e-06,
+      "loss": 0.8578,
+      "step": 4358
+    },
+    {
+      "epoch": 0.8385917660638708,
+      "grad_norm": 1.2765807205336148,
+      "learning_rate": 1.3355950537973438e-06,
+      "loss": 0.8824,
+      "step": 4359
+    },
+    {
+      "epoch": 0.8387841477491342,
+      "grad_norm": 1.2767446615216422,
+      "learning_rate": 1.332485796829065e-06,
+      "loss": 0.8926,
+      "step": 4360
+    },
+    {
+      "epoch": 0.8389765294343978,
+      "grad_norm": 0.5774285432333371,
+      "learning_rate": 1.329379904888235e-06,
+      "loss": 0.3196,
+      "step": 4361
+    },
+    {
+      "epoch": 0.8391689111196614,
+      "grad_norm": 1.2262536642230832,
+      "learning_rate": 1.3262773791806617e-06,
+      "loss": 0.8287,
+      "step": 4362
+    },
+    {
+      "epoch": 0.839361292804925,
+      "grad_norm": 1.471263420805147,
+      "learning_rate": 1.3231782209108546e-06,
+      "loss": 0.8055,
+      "step": 4363
+    },
+    {
+      "epoch": 0.8395536744901886,
+      "grad_norm": 1.3346908603619285,
+      "learning_rate": 1.3200824312820137e-06,
+      "loss": 0.8459,
+      "step": 4364
+    },
+    {
+      "epoch": 0.8397460561754521,
+      "grad_norm": 1.2757401295179407,
+      "learning_rate": 1.3169900114960298e-06,
+      "loss": 0.8548,
+      "step": 4365
+    },
+    {
+      "epoch": 0.8399384378607156,
+      "grad_norm": 1.5474935566397985,
+      "learning_rate": 1.3139009627534927e-06,
+      "loss": 0.8965,
+      "step": 4366
+    },
+    {
+      "epoch": 0.8401308195459792,
+      "grad_norm": 1.1309884344673258,
+      "learning_rate": 1.3108152862536683e-06,
+      "loss": 0.8143,
+      "step": 4367
+    },
+    {
+      "epoch": 0.8403232012312428,
+      "grad_norm": 1.3567006868006746,
+      "learning_rate": 1.3077329831945295e-06,
+      "loss": 0.8302,
+      "step": 4368
+    },
+    {
+      "epoch": 0.8405155829165063,
+      "grad_norm": 1.3934284366029603,
+      "learning_rate": 1.3046540547727305e-06,
+      "loss": 0.8221,
+      "step": 4369
+    },
+    {
+      "epoch": 0.8407079646017699,
+      "grad_norm": 1.3793460758501368,
+      "learning_rate": 1.3015785021836159e-06,
+      "loss": 0.9037,
+      "step": 4370
+    },
+    {
+      "epoch": 0.8409003462870335,
+      "grad_norm": 1.36953159704486,
+      "learning_rate": 1.2985063266212229e-06,
+      "loss": 0.8356,
+      "step": 4371
+    },
+    {
+      "epoch": 0.8410927279722971,
+      "grad_norm": 1.403231012698438,
+      "learning_rate": 1.295437529278275e-06,
+      "loss": 0.8128,
+      "step": 4372
+    },
+    {
+      "epoch": 0.8412851096575606,
+      "grad_norm": 0.6011849643848949,
+      "learning_rate": 1.2923721113461852e-06,
+      "loss": 0.3408,
+      "step": 4373
+    },
+    {
+      "epoch": 0.8414774913428241,
+      "grad_norm": 1.4235821731382676,
+      "learning_rate": 1.2893100740150522e-06,
+      "loss": 0.8074,
+      "step": 4374
+    },
+    {
+      "epoch": 0.8416698730280877,
+      "grad_norm": 1.252786240572184,
+      "learning_rate": 1.2862514184736695e-06,
+      "loss": 0.7927,
+      "step": 4375
+    },
+    {
+      "epoch": 0.8418622547133513,
+      "grad_norm": 1.5467262990742614,
+      "learning_rate": 1.2831961459095088e-06,
+      "loss": 0.851,
+      "step": 4376
+    },
+    {
+      "epoch": 0.8420546363986149,
+      "grad_norm": 1.4409792600785587,
+      "learning_rate": 1.2801442575087296e-06,
+      "loss": 0.7596,
+      "step": 4377
+    },
+    {
+      "epoch": 0.8422470180838784,
+      "grad_norm": 1.2835044406964304,
+      "learning_rate": 1.2770957544561868e-06,
+      "loss": 0.8633,
+      "step": 4378
+    },
+    {
+      "epoch": 0.842439399769142,
+      "grad_norm": 1.494229323111473,
+      "learning_rate": 1.274050637935408e-06,
+      "loss": 0.8513,
+      "step": 4379
+    },
+    {
+      "epoch": 0.8426317814544055,
+      "grad_norm": 1.2359794930777055,
+      "learning_rate": 1.2710089091286148e-06,
+      "loss": 0.8693,
+      "step": 4380
+    },
+    {
+      "epoch": 0.8428241631396691,
+      "grad_norm": 1.564188895472822,
+      "learning_rate": 1.2679705692167122e-06,
+      "loss": 0.7937,
+      "step": 4381
+    },
+    {
+      "epoch": 0.8430165448249327,
+      "grad_norm": 1.1362627411675625,
+      "learning_rate": 1.2649356193792873e-06,
+      "loss": 0.8139,
+      "step": 4382
+    },
+    {
+      "epoch": 0.8432089265101962,
+      "grad_norm": 1.1938596958129393,
+      "learning_rate": 1.261904060794612e-06,
+      "loss": 0.817,
+      "step": 4383
+    },
+    {
+      "epoch": 0.8434013081954598,
+      "grad_norm": 1.3617093584254822,
+      "learning_rate": 1.2588758946396417e-06,
+      "loss": 0.862,
+      "step": 4384
+    },
+    {
+      "epoch": 0.8435936898807234,
+      "grad_norm": 2.2433517436178487,
+      "learning_rate": 1.2558511220900138e-06,
+      "loss": 0.8781,
+      "step": 4385
+    },
+    {
+      "epoch": 0.843786071565987,
+      "grad_norm": 1.4354054255844384,
+      "learning_rate": 1.2528297443200489e-06,
+      "loss": 0.8569,
+      "step": 4386
+    },
+    {
+      "epoch": 0.8439784532512504,
+      "grad_norm": 0.621282802494797,
+      "learning_rate": 1.2498117625027562e-06,
+      "loss": 0.3104,
+      "step": 4387
+    },
+    {
+      "epoch": 0.844170834936514,
+      "grad_norm": 1.4401757885610411,
+      "learning_rate": 1.246797177809812e-06,
+      "loss": 0.8516,
+      "step": 4388
+    },
+    {
+      "epoch": 0.8443632166217776,
+      "grad_norm": 1.5942261989838562,
+      "learning_rate": 1.2437859914115847e-06,
+      "loss": 0.8416,
+      "step": 4389
+    },
+    {
+      "epoch": 0.8445555983070412,
+      "grad_norm": 1.313333201452948,
+      "learning_rate": 1.2407782044771222e-06,
+      "loss": 0.8171,
+      "step": 4390
+    },
+    {
+      "epoch": 0.8447479799923048,
+      "grad_norm": 1.2564360383445832,
+      "learning_rate": 1.237773818174146e-06,
+      "loss": 0.7356,
+      "step": 4391
+    },
+    {
+      "epoch": 0.8449403616775683,
+      "grad_norm": 1.3839191785475051,
+      "learning_rate": 1.23477283366907e-06,
+      "loss": 0.8029,
+      "step": 4392
+    },
+    {
+      "epoch": 0.8451327433628318,
+      "grad_norm": 1.382209791631295,
+      "learning_rate": 1.2317752521269722e-06,
+      "loss": 0.7442,
+      "step": 4393
+    },
+    {
+      "epoch": 0.8453251250480954,
+      "grad_norm": 1.3675350334457756,
+      "learning_rate": 1.2287810747116224e-06,
+      "loss": 0.8374,
+      "step": 4394
+    },
+    {
+      "epoch": 0.845517506733359,
+      "grad_norm": 1.553586116489572,
+      "learning_rate": 1.225790302585461e-06,
+      "loss": 0.7663,
+      "step": 4395
+    },
+    {
+      "epoch": 0.8457098884186226,
+      "grad_norm": 1.3922606646510358,
+      "learning_rate": 1.2228029369096094e-06,
+      "loss": 0.8789,
+      "step": 4396
+    },
+    {
+      "epoch": 0.8459022701038861,
+      "grad_norm": 1.356637279430056,
+      "learning_rate": 1.2198189788438652e-06,
+      "loss": 0.8212,
+      "step": 4397
+    },
+    {
+      "epoch": 0.8460946517891497,
+      "grad_norm": 1.1695159518253162,
+      "learning_rate": 1.216838429546704e-06,
+      "loss": 0.7479,
+      "step": 4398
+    },
+    {
+      "epoch": 0.8462870334744133,
+      "grad_norm": 1.1894289406787706,
+      "learning_rate": 1.2138612901752777e-06,
+      "loss": 0.8123,
+      "step": 4399
+    },
+    {
+      "epoch": 0.8464794151596768,
+      "grad_norm": 0.6471642901951878,
+      "learning_rate": 1.2108875618854122e-06,
+      "loss": 0.3155,
+      "step": 4400
+    },
+    {
+      "epoch": 0.8466717968449403,
+      "grad_norm": 1.3809088125615616,
+      "learning_rate": 1.2079172458316168e-06,
+      "loss": 0.7838,
+      "step": 4401
+    },
+    {
+      "epoch": 0.8468641785302039,
+      "grad_norm": 1.387612004802981,
+      "learning_rate": 1.204950343167065e-06,
+      "loss": 0.8918,
+      "step": 4402
+    },
+    {
+      "epoch": 0.8470565602154675,
+      "grad_norm": 1.3909334978669927,
+      "learning_rate": 1.2019868550436099e-06,
+      "loss": 0.8792,
+      "step": 4403
+    },
+    {
+      "epoch": 0.8472489419007311,
+      "grad_norm": 5.7626170421341465,
+      "learning_rate": 1.1990267826117874e-06,
+      "loss": 0.8513,
+      "step": 4404
+    },
+    {
+      "epoch": 0.8474413235859947,
+      "grad_norm": 1.2133821294603389,
+      "learning_rate": 1.1960701270207885e-06,
+      "loss": 0.79,
+      "step": 4405
+    },
+    {
+      "epoch": 0.8476337052712581,
+      "grad_norm": 1.3255186434903004,
+      "learning_rate": 1.1931168894184974e-06,
+      "loss": 0.8479,
+      "step": 4406
+    },
+    {
+      "epoch": 0.8478260869565217,
+      "grad_norm": 1.3488996801426827,
+      "learning_rate": 1.19016707095146e-06,
+      "loss": 0.8291,
+      "step": 4407
+    },
+    {
+      "epoch": 0.8480184686417853,
+      "grad_norm": 1.3511437990722324,
+      "learning_rate": 1.187220672764897e-06,
+      "loss": 0.8347,
+      "step": 4408
+    },
+    {
+      "epoch": 0.8482108503270489,
+      "grad_norm": 1.1632442408006534,
+      "learning_rate": 1.1842776960027014e-06,
+      "loss": 0.8145,
+      "step": 4409
+    },
+    {
+      "epoch": 0.8484032320123124,
+      "grad_norm": 1.3219492988708559,
+      "learning_rate": 1.1813381418074388e-06,
+      "loss": 0.826,
+      "step": 4410
+    },
+    {
+      "epoch": 0.848595613697576,
+      "grad_norm": 0.6272650338246871,
+      "learning_rate": 1.1784020113203453e-06,
+      "loss": 0.3556,
+      "step": 4411
+    },
+    {
+      "epoch": 0.8487879953828396,
+      "grad_norm": 1.358295593640059,
+      "learning_rate": 1.1754693056813272e-06,
+      "loss": 0.7718,
+      "step": 4412
+    },
+    {
+      "epoch": 0.8489803770681031,
+      "grad_norm": 1.6062370627284117,
+      "learning_rate": 1.172540026028962e-06,
+      "loss": 0.7934,
+      "step": 4413
+    },
+    {
+      "epoch": 0.8491727587533667,
+      "grad_norm": 1.3415547079851082,
+      "learning_rate": 1.169614173500494e-06,
+      "loss": 0.7679,
+      "step": 4414
+    },
+    {
+      "epoch": 0.8493651404386302,
+      "grad_norm": 1.2576317171123759,
+      "learning_rate": 1.1666917492318486e-06,
+      "loss": 0.78,
+      "step": 4415
+    },
+    {
+      "epoch": 0.8495575221238938,
+      "grad_norm": 1.278066720646199,
+      "learning_rate": 1.1637727543576027e-06,
+      "loss": 0.8256,
+      "step": 4416
+    },
+    {
+      "epoch": 0.8497499038091574,
+      "grad_norm": 1.223500017273099,
+      "learning_rate": 1.1608571900110122e-06,
+      "loss": 0.7984,
+      "step": 4417
+    },
+    {
+      "epoch": 0.849942285494421,
+      "grad_norm": 1.39721077863917,
+      "learning_rate": 1.1579450573240058e-06,
+      "loss": 0.7886,
+      "step": 4418
+    },
+    {
+      "epoch": 0.8501346671796844,
+      "grad_norm": 1.5431700839355957,
+      "learning_rate": 1.1550363574271638e-06,
+      "loss": 0.8698,
+      "step": 4419
+    },
+    {
+      "epoch": 0.850327048864948,
+      "grad_norm": 1.2144672900285167,
+      "learning_rate": 1.1521310914497518e-06,
+      "loss": 0.8272,
+      "step": 4420
+    },
+    {
+      "epoch": 0.8505194305502116,
+      "grad_norm": 1.1914675255255784,
+      "learning_rate": 1.149229260519691e-06,
+      "loss": 0.8157,
+      "step": 4421
+    },
+    {
+      "epoch": 0.8507118122354752,
+      "grad_norm": 1.7239864995575556,
+      "learning_rate": 1.1463308657635718e-06,
+      "loss": 0.8854,
+      "step": 4422
+    },
+    {
+      "epoch": 0.8509041939207388,
+      "grad_norm": 1.3040239968878597,
+      "learning_rate": 1.1434359083066515e-06,
+      "loss": 0.8516,
+      "step": 4423
+    },
+    {
+      "epoch": 0.8510965756060023,
+      "grad_norm": 1.7678268885907225,
+      "learning_rate": 1.140544389272853e-06,
+      "loss": 0.8478,
+      "step": 4424
+    },
+    {
+      "epoch": 0.8512889572912659,
+      "grad_norm": 1.4625345215725818,
+      "learning_rate": 1.1376563097847616e-06,
+      "loss": 0.7614,
+      "step": 4425
+    },
+    {
+      "epoch": 0.8514813389765294,
+      "grad_norm": 1.3199530288818981,
+      "learning_rate": 1.1347716709636282e-06,
+      "loss": 0.8018,
+      "step": 4426
+    },
+    {
+      "epoch": 0.851673720661793,
+      "grad_norm": 4.350716699587902,
+      "learning_rate": 1.1318904739293745e-06,
+      "loss": 0.7766,
+      "step": 4427
+    },
+    {
+      "epoch": 0.8518661023470565,
+      "grad_norm": 1.2886754345263296,
+      "learning_rate": 1.129012719800575e-06,
+      "loss": 0.8319,
+      "step": 4428
+    },
+    {
+      "epoch": 0.8520584840323201,
+      "grad_norm": 1.3777170548665634,
+      "learning_rate": 1.1261384096944728e-06,
+      "loss": 0.8233,
+      "step": 4429
+    },
+    {
+      "epoch": 0.8522508657175837,
+      "grad_norm": 1.291755197843562,
+      "learning_rate": 1.1232675447269803e-06,
+      "loss": 0.7815,
+      "step": 4430
+    },
+    {
+      "epoch": 0.8524432474028473,
+      "grad_norm": 1.1082029381176728,
+      "learning_rate": 1.1204001260126574e-06,
+      "loss": 0.7936,
+      "step": 4431
+    },
+    {
+      "epoch": 0.8526356290881109,
+      "grad_norm": 1.7222736580860771,
+      "learning_rate": 1.1175361546647413e-06,
+      "loss": 0.8269,
+      "step": 4432
+    },
+    {
+      "epoch": 0.8528280107733743,
+      "grad_norm": 11.032295991973148,
+      "learning_rate": 1.1146756317951224e-06,
+      "loss": 0.8041,
+      "step": 4433
+    },
+    {
+      "epoch": 0.8530203924586379,
+      "grad_norm": 1.3910216002205504,
+      "learning_rate": 1.1118185585143536e-06,
+      "loss": 0.8431,
+      "step": 4434
+    },
+    {
+      "epoch": 0.8532127741439015,
+      "grad_norm": 1.3279253280631411,
+      "learning_rate": 1.1089649359316501e-06,
+      "loss": 0.8581,
+      "step": 4435
+    },
+    {
+      "epoch": 0.8534051558291651,
+      "grad_norm": 1.2619721945148414,
+      "learning_rate": 1.1061147651548855e-06,
+      "loss": 0.8316,
+      "step": 4436
+    },
+    {
+      "epoch": 0.8535975375144286,
+      "grad_norm": 1.4876541256221463,
+      "learning_rate": 1.1032680472905932e-06,
+      "loss": 0.8123,
+      "step": 4437
+    },
+    {
+      "epoch": 0.8537899191996922,
+      "grad_norm": 1.4454808265340082,
+      "learning_rate": 1.1004247834439697e-06,
+      "loss": 0.8383,
+      "step": 4438
+    },
+    {
+      "epoch": 0.8539823008849557,
+      "grad_norm": 1.3106414991935875,
+      "learning_rate": 1.097584974718866e-06,
+      "loss": 0.8462,
+      "step": 4439
+    },
+    {
+      "epoch": 0.8541746825702193,
+      "grad_norm": 1.738083306068848,
+      "learning_rate": 1.0947486222177928e-06,
+      "loss": 0.8377,
+      "step": 4440
+    },
+    {
+      "epoch": 0.8543670642554829,
+      "grad_norm": 1.7519971527927054,
+      "learning_rate": 1.0919157270419257e-06,
+      "loss": 0.7756,
+      "step": 4441
+    },
+    {
+      "epoch": 0.8545594459407464,
+      "grad_norm": 1.183865721024367,
+      "learning_rate": 1.0890862902910849e-06,
+      "loss": 0.8311,
+      "step": 4442
+    },
+    {
+      "epoch": 0.85475182762601,
+      "grad_norm": 2.297297245176582,
+      "learning_rate": 1.0862603130637562e-06,
+      "loss": 0.8183,
+      "step": 4443
+    },
+    {
+      "epoch": 0.8549442093112736,
+      "grad_norm": 1.4365680932993146,
+      "learning_rate": 1.0834377964570863e-06,
+      "loss": 0.7769,
+      "step": 4444
+    },
+    {
+      "epoch": 0.8551365909965372,
+      "grad_norm": 1.3898840086940227,
+      "learning_rate": 1.0806187415668668e-06,
+      "loss": 0.8126,
+      "step": 4445
+    },
+    {
+      "epoch": 0.8553289726818007,
+      "grad_norm": 1.4589524968627283,
+      "learning_rate": 1.0778031494875574e-06,
+      "loss": 0.819,
+      "step": 4446
+    },
+    {
+      "epoch": 0.8555213543670642,
+      "grad_norm": 1.4553804436741682,
+      "learning_rate": 1.0749910213122649e-06,
+      "loss": 0.7871,
+      "step": 4447
+    },
+    {
+      "epoch": 0.8557137360523278,
+      "grad_norm": 1.2839587078755161,
+      "learning_rate": 1.072182358132755e-06,
+      "loss": 0.8189,
+      "step": 4448
+    },
+    {
+      "epoch": 0.8559061177375914,
+      "grad_norm": 1.2336707447690634,
+      "learning_rate": 1.0693771610394477e-06,
+      "loss": 0.8144,
+      "step": 4449
+    },
+    {
+      "epoch": 0.856098499422855,
+      "grad_norm": 1.6376188669632128,
+      "learning_rate": 1.066575431121417e-06,
+      "loss": 0.8504,
+      "step": 4450
+    },
+    {
+      "epoch": 0.8562908811081185,
+      "grad_norm": 1.4691382520331795,
+      "learning_rate": 1.06377716946639e-06,
+      "loss": 0.8399,
+      "step": 4451
+    },
+    {
+      "epoch": 0.856483262793382,
+      "grad_norm": 1.4722802510590394,
+      "learning_rate": 1.0609823771607487e-06,
+      "loss": 0.8171,
+      "step": 4452
+    },
+    {
+      "epoch": 0.8566756444786456,
+      "grad_norm": 1.1890066495475922,
+      "learning_rate": 1.0581910552895302e-06,
+      "loss": 0.8418,
+      "step": 4453
+    },
+    {
+      "epoch": 0.8568680261639092,
+      "grad_norm": 1.4099070405693257,
+      "learning_rate": 1.055403204936416e-06,
+      "loss": 0.8557,
+      "step": 4454
+    },
+    {
+      "epoch": 0.8570604078491728,
+      "grad_norm": 1.3882781356409972,
+      "learning_rate": 1.0526188271837512e-06,
+      "loss": 0.8264,
+      "step": 4455
+    },
+    {
+      "epoch": 0.8572527895344363,
+      "grad_norm": 1.2461999508018864,
+      "learning_rate": 1.0498379231125278e-06,
+      "loss": 0.8652,
+      "step": 4456
+    },
+    {
+      "epoch": 0.8574451712196999,
+      "grad_norm": 1.1058411657597809,
+      "learning_rate": 1.047060493802381e-06,
+      "loss": 0.8458,
+      "step": 4457
+    },
+    {
+      "epoch": 0.8576375529049635,
+      "grad_norm": 1.1825981816568985,
+      "learning_rate": 1.0442865403316117e-06,
+      "loss": 0.7705,
+      "step": 4458
+    },
+    {
+      "epoch": 0.857829934590227,
+      "grad_norm": 0.6432471163560715,
+      "learning_rate": 1.0415160637771604e-06,
+      "loss": 0.3685,
+      "step": 4459
+    },
+    {
+      "epoch": 0.8580223162754905,
+      "grad_norm": 1.1454958782083433,
+      "learning_rate": 1.0387490652146236e-06,
+      "loss": 0.8825,
+      "step": 4460
+    },
+    {
+      "epoch": 0.8582146979607541,
+      "grad_norm": 1.4597358012565098,
+      "learning_rate": 1.0359855457182455e-06,
+      "loss": 0.8043,
+      "step": 4461
+    },
+    {
+      "epoch": 0.8584070796460177,
+      "grad_norm": 1.5711700920585472,
+      "learning_rate": 1.0332255063609177e-06,
+      "loss": 0.8509,
+      "step": 4462
+    },
+    {
+      "epoch": 0.8585994613312813,
+      "grad_norm": 1.566245041210633,
+      "learning_rate": 1.0304689482141839e-06,
+      "loss": 0.8586,
+      "step": 4463
+    },
+    {
+      "epoch": 0.8587918430165449,
+      "grad_norm": 1.2349951871394615,
+      "learning_rate": 1.027715872348234e-06,
+      "loss": 0.8089,
+      "step": 4464
+    },
+    {
+      "epoch": 0.8589842247018084,
+      "grad_norm": 1.4808706700528655,
+      "learning_rate": 1.0249662798319072e-06,
+      "loss": 0.8041,
+      "step": 4465
+    },
+    {
+      "epoch": 0.8591766063870719,
+      "grad_norm": 1.3783235105362823,
+      "learning_rate": 1.0222201717326885e-06,
+      "loss": 0.8082,
+      "step": 4466
+    },
+    {
+      "epoch": 0.8593689880723355,
+      "grad_norm": 1.3494475762142575,
+      "learning_rate": 1.0194775491167164e-06,
+      "loss": 0.8188,
+      "step": 4467
+    },
+    {
+      "epoch": 0.8595613697575991,
+      "grad_norm": 1.3802713380370886,
+      "learning_rate": 1.0167384130487667e-06,
+      "loss": 0.8393,
+      "step": 4468
+    },
+    {
+      "epoch": 0.8597537514428626,
+      "grad_norm": 2.1641942607728635,
+      "learning_rate": 1.0140027645922656e-06,
+      "loss": 0.7887,
+      "step": 4469
+    },
+    {
+      "epoch": 0.8599461331281262,
+      "grad_norm": 1.9034804026378118,
+      "learning_rate": 1.0112706048092924e-06,
+      "loss": 0.8494,
+      "step": 4470
+    },
+    {
+      "epoch": 0.8601385148133898,
+      "grad_norm": 1.191180995401304,
+      "learning_rate": 1.0085419347605575e-06,
+      "loss": 0.8514,
+      "step": 4471
+    },
+    {
+      "epoch": 0.8603308964986534,
+      "grad_norm": 1.3435140565053052,
+      "learning_rate": 1.00581675550543e-06,
+      "loss": 0.8177,
+      "step": 4472
+    },
+    {
+      "epoch": 0.8605232781839169,
+      "grad_norm": 1.56462045767039,
+      "learning_rate": 1.003095068101917e-06,
+      "loss": 0.7671,
+      "step": 4473
+    },
+    {
+      "epoch": 0.8607156598691804,
+      "grad_norm": 14.952300669199513,
+      "learning_rate": 1.0003768736066722e-06,
+      "loss": 0.7879,
+      "step": 4474
+    },
+    {
+      "epoch": 0.860908041554444,
+      "grad_norm": 1.3294195600124123,
+      "learning_rate": 9.976621730749892e-07,
+      "loss": 0.8315,
+      "step": 4475
+    },
+    {
+      "epoch": 0.8611004232397076,
+      "grad_norm": 1.4651873574499878,
+      "learning_rate": 9.949509675608115e-07,
+      "loss": 0.839,
+      "step": 4476
+    },
+    {
+      "epoch": 0.8612928049249712,
+      "grad_norm": 1.5441778571388909,
+      "learning_rate": 9.922432581167207e-07,
+      "loss": 0.8604,
+      "step": 4477
+    },
+    {
+      "epoch": 0.8614851866102347,
+      "grad_norm": 1.3291722949359843,
+      "learning_rate": 9.895390457939414e-07,
+      "loss": 0.8584,
+      "step": 4478
+    },
+    {
+      "epoch": 0.8616775682954982,
+      "grad_norm": 1.5438298866120748,
+      "learning_rate": 9.86838331642348e-07,
+      "loss": 0.8208,
+      "step": 4479
+    },
+    {
+      "epoch": 0.8618699499807618,
+      "grad_norm": 1.4099137502558285,
+      "learning_rate": 9.84141116710442e-07,
+      "loss": 0.8569,
+      "step": 4480
+    },
+    {
+      "epoch": 0.8620623316660254,
+      "grad_norm": 1.6235415701214326,
+      "learning_rate": 9.814474020453824e-07,
+      "loss": 0.7716,
+      "step": 4481
+    },
+    {
+      "epoch": 0.862254713351289,
+      "grad_norm": 1.3581554600305052,
+      "learning_rate": 9.787571886929604e-07,
+      "loss": 0.8633,
+      "step": 4482
+    },
+    {
+      "epoch": 0.8624470950365525,
+      "grad_norm": 1.154605061379837,
+      "learning_rate": 9.76070477697605e-07,
+      "loss": 0.7811,
+      "step": 4483
+    },
+    {
+      "epoch": 0.8626394767218161,
+      "grad_norm": 1.271985273320863,
+      "learning_rate": 9.733872701023938e-07,
+      "loss": 0.8426,
+      "step": 4484
+    },
+    {
+      "epoch": 0.8628318584070797,
+      "grad_norm": 1.4186671512516797,
+      "learning_rate": 9.707075669490407e-07,
+      "loss": 0.8474,
+      "step": 4485
+    },
+    {
+      "epoch": 0.8630242400923432,
+      "grad_norm": 1.335260500106613,
+      "learning_rate": 9.680313692778976e-07,
+      "loss": 0.7825,
+      "step": 4486
+    },
+    {
+      "epoch": 0.8632166217776067,
+      "grad_norm": 1.5948422022941826,
+      "learning_rate": 9.653586781279567e-07,
+      "loss": 0.7906,
+      "step": 4487
+    },
+    {
+      "epoch": 0.8634090034628703,
+      "grad_norm": 1.5127521835948343,
+      "learning_rate": 9.626894945368492e-07,
+      "loss": 0.8244,
+      "step": 4488
+    },
+    {
+      "epoch": 0.8636013851481339,
+      "grad_norm": 1.2723965723289872,
+      "learning_rate": 9.600238195408428e-07,
+      "loss": 0.8558,
+      "step": 4489
+    },
+    {
+      "epoch": 0.8637937668333975,
+      "grad_norm": 1.2863992989003048,
+      "learning_rate": 9.573616541748464e-07,
+      "loss": 0.8308,
+      "step": 4490
+    },
+    {
+      "epoch": 0.8639861485186611,
+      "grad_norm": 1.1367199083305302,
+      "learning_rate": 9.547029994724023e-07,
+      "loss": 0.7561,
+      "step": 4491
+    },
+    {
+      "epoch": 0.8641785302039245,
+      "grad_norm": 1.3529985790541186,
+      "learning_rate": 9.520478564656898e-07,
+      "loss": 0.779,
+      "step": 4492
+    },
+    {
+      "epoch": 0.8643709118891881,
+      "grad_norm": 1.4480298621245689,
+      "learning_rate": 9.49396226185535e-07,
+      "loss": 0.8218,
+      "step": 4493
+    },
+    {
+      "epoch": 0.8645632935744517,
+      "grad_norm": 1.3535215447488742,
+      "learning_rate": 9.467481096613829e-07,
+      "loss": 0.8114,
+      "step": 4494
+    },
+    {
+      "epoch": 0.8647556752597153,
+      "grad_norm": 1.6481441538245838,
+      "learning_rate": 9.441035079213267e-07,
+      "loss": 0.8743,
+      "step": 4495
+    },
+    {
+      "epoch": 0.8649480569449788,
+      "grad_norm": 1.2451899868782976,
+      "learning_rate": 9.414624219920953e-07,
+      "loss": 0.8204,
+      "step": 4496
+    },
+    {
+      "epoch": 0.8651404386302424,
+      "grad_norm": 1.2028584365761754,
+      "learning_rate": 9.38824852899043e-07,
+      "loss": 0.7803,
+      "step": 4497
+    },
+    {
+      "epoch": 0.865332820315506,
+      "grad_norm": 1.3737265060741903,
+      "learning_rate": 9.361908016661703e-07,
+      "loss": 0.8323,
+      "step": 4498
+    },
+    {
+      "epoch": 0.8655252020007695,
+      "grad_norm": 0.6075101170170294,
+      "learning_rate": 9.335602693161039e-07,
+      "loss": 0.3164,
+      "step": 4499
+    },
+    {
+      "epoch": 0.8657175836860331,
+      "grad_norm": 1.180326270148351,
+      "learning_rate": 9.309332568701079e-07,
+      "loss": 0.835,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8659099653712966,
+      "grad_norm": 1.3354677022953458,
+      "learning_rate": 9.283097653480788e-07,
+      "loss": 0.8261,
+      "step": 4501
+    },
+    {
+      "epoch": 0.8661023470565602,
+      "grad_norm": 1.1522914970260516,
+      "learning_rate": 9.256897957685463e-07,
+      "loss": 0.8907,
+      "step": 4502
+    },
+    {
+      "epoch": 0.8662947287418238,
+      "grad_norm": 1.155531859941561,
+      "learning_rate": 9.230733491486721e-07,
+      "loss": 0.801,
+      "step": 4503
+    },
+    {
+      "epoch": 0.8664871104270874,
+      "grad_norm": 1.111014762032526,
+      "learning_rate": 9.204604265042505e-07,
+      "loss": 0.809,
+      "step": 4504
+    },
+    {
+      "epoch": 0.866679492112351,
+      "grad_norm": 1.3217665866324018,
+      "learning_rate": 9.178510288497123e-07,
+      "loss": 0.7842,
+      "step": 4505
+    },
+    {
+      "epoch": 0.8668718737976144,
+      "grad_norm": 1.7285319738412042,
+      "learning_rate": 9.15245157198108e-07,
+      "loss": 0.8592,
+      "step": 4506
+    },
+    {
+      "epoch": 0.867064255482878,
+      "grad_norm": 1.2543038291749795,
+      "learning_rate": 9.126428125611342e-07,
+      "loss": 0.8604,
+      "step": 4507
+    },
+    {
+      "epoch": 0.8672566371681416,
+      "grad_norm": 1.471907661517946,
+      "learning_rate": 9.10043995949108e-07,
+      "loss": 0.7834,
+      "step": 4508
+    },
+    {
+      "epoch": 0.8674490188534052,
+      "grad_norm": 1.2706726641510624,
+      "learning_rate": 9.074487083709759e-07,
+      "loss": 0.8649,
+      "step": 4509
+    },
+    {
+      "epoch": 0.8676414005386687,
+      "grad_norm": 1.206637875793535,
+      "learning_rate": 9.04856950834323e-07,
+      "loss": 0.8707,
+      "step": 4510
+    },
+    {
+      "epoch": 0.8678337822239323,
+      "grad_norm": 1.2908389209804385,
+      "learning_rate": 9.022687243453554e-07,
+      "loss": 0.8348,
+      "step": 4511
+    },
+    {
+      "epoch": 0.8680261639091958,
+      "grad_norm": 2.1182660073960253,
+      "learning_rate": 8.996840299089149e-07,
+      "loss": 0.8313,
+      "step": 4512
+    },
+    {
+      "epoch": 0.8682185455944594,
+      "grad_norm": 1.182337964412,
+      "learning_rate": 8.971028685284655e-07,
+      "loss": 0.8174,
+      "step": 4513
+    },
+    {
+      "epoch": 0.868410927279723,
+      "grad_norm": 1.306804295668824,
+      "learning_rate": 8.945252412061056e-07,
+      "loss": 0.8295,
+      "step": 4514
+    },
+    {
+      "epoch": 0.8686033089649865,
+      "grad_norm": 2.197963103015757,
+      "learning_rate": 8.91951148942557e-07,
+      "loss": 0.8548,
+      "step": 4515
+    },
+    {
+      "epoch": 0.8687956906502501,
+      "grad_norm": 1.340378723813221,
+      "learning_rate": 8.893805927371724e-07,
+      "loss": 0.841,
+      "step": 4516
+    },
+    {
+      "epoch": 0.8689880723355137,
+      "grad_norm": 1.3661426773799628,
+      "learning_rate": 8.868135735879291e-07,
+      "loss": 0.8611,
+      "step": 4517
+    },
+    {
+      "epoch": 0.8691804540207773,
+      "grad_norm": 1.2781023329098795,
+      "learning_rate": 8.842500924914299e-07,
+      "loss": 0.8699,
+      "step": 4518
+    },
+    {
+      "epoch": 0.8693728357060407,
+      "grad_norm": 1.314818232895854,
+      "learning_rate": 8.816901504429143e-07,
+      "loss": 0.815,
+      "step": 4519
+    },
+    {
+      "epoch": 0.8695652173913043,
+      "grad_norm": 1.5013486499751132,
+      "learning_rate": 8.791337484362305e-07,
+      "loss": 0.8109,
+      "step": 4520
+    },
+    {
+      "epoch": 0.8697575990765679,
+      "grad_norm": 1.4454646219719036,
+      "learning_rate": 8.765808874638682e-07,
+      "loss": 0.7967,
+      "step": 4521
+    },
+    {
+      "epoch": 0.8699499807618315,
+      "grad_norm": 1.2225217414805674,
+      "learning_rate": 8.740315685169364e-07,
+      "loss": 0.8139,
+      "step": 4522
+    },
+    {
+      "epoch": 0.8701423624470951,
+      "grad_norm": 1.3316825063150024,
+      "learning_rate": 8.714857925851617e-07,
+      "loss": 0.8274,
+      "step": 4523
+    },
+    {
+      "epoch": 0.8703347441323586,
+      "grad_norm": 1.3832369257419037,
+      "learning_rate": 8.689435606569086e-07,
+      "loss": 0.8836,
+      "step": 4524
+    },
+    {
+      "epoch": 0.8705271258176221,
+      "grad_norm": 1.4196515648595691,
+      "learning_rate": 8.664048737191566e-07,
+      "loss": 0.8356,
+      "step": 4525
+    },
+    {
+      "epoch": 0.8707195075028857,
+      "grad_norm": 1.1085359544890385,
+      "learning_rate": 8.638697327575108e-07,
+      "loss": 0.8392,
+      "step": 4526
+    },
+    {
+      "epoch": 0.8709118891881493,
+      "grad_norm": 1.2513597180658933,
+      "learning_rate": 8.613381387562015e-07,
+      "loss": 0.851,
+      "step": 4527
+    },
+    {
+      "epoch": 0.8711042708734128,
+      "grad_norm": 1.3885734193403485,
+      "learning_rate": 8.588100926980802e-07,
+      "loss": 0.8269,
+      "step": 4528
+    },
+    {
+      "epoch": 0.8712966525586764,
+      "grad_norm": 1.3408974340419462,
+      "learning_rate": 8.56285595564621e-07,
+      "loss": 0.8452,
+      "step": 4529
+    },
+    {
+      "epoch": 0.87148903424394,
+      "grad_norm": 1.327144261680139,
+      "learning_rate": 8.537646483359185e-07,
+      "loss": 0.7432,
+      "step": 4530
+    },
+    {
+      "epoch": 0.8716814159292036,
+      "grad_norm": 1.181082684847863,
+      "learning_rate": 8.512472519906978e-07,
+      "loss": 0.848,
+      "step": 4531
+    },
+    {
+      "epoch": 0.8718737976144671,
+      "grad_norm": 1.5419781309038294,
+      "learning_rate": 8.487334075062914e-07,
+      "loss": 0.8498,
+      "step": 4532
+    },
+    {
+      "epoch": 0.8720661792997306,
+      "grad_norm": 1.2250896310720467,
+      "learning_rate": 8.462231158586654e-07,
+      "loss": 0.7771,
+      "step": 4533
+    },
+    {
+      "epoch": 0.8722585609849942,
+      "grad_norm": 0.6113527089448731,
+      "learning_rate": 8.437163780224011e-07,
+      "loss": 0.3188,
+      "step": 4534
+    },
+    {
+      "epoch": 0.8724509426702578,
+      "grad_norm": 1.318241598789473,
+      "learning_rate": 8.412131949706958e-07,
+      "loss": 0.8801,
+      "step": 4535
+    },
+    {
+      "epoch": 0.8726433243555214,
+      "grad_norm": 1.9073113521652076,
+      "learning_rate": 8.387135676753755e-07,
+      "loss": 0.8483,
+      "step": 4536
+    },
+    {
+      "epoch": 0.8728357060407849,
+      "grad_norm": 1.7793472022230998,
+      "learning_rate": 8.362174971068804e-07,
+      "loss": 0.8446,
+      "step": 4537
+    },
+    {
+      "epoch": 0.8730280877260485,
+      "grad_norm": 1.1191532783869222,
+      "learning_rate": 8.337249842342721e-07,
+      "loss": 0.8027,
+      "step": 4538
+    },
+    {
+      "epoch": 0.873220469411312,
+      "grad_norm": 1.2240879063320769,
+      "learning_rate": 8.312360300252287e-07,
+      "loss": 0.8223,
+      "step": 4539
+    },
+    {
+      "epoch": 0.8734128510965756,
+      "grad_norm": 1.2799589925899337,
+      "learning_rate": 8.287506354460484e-07,
+      "loss": 0.8531,
+      "step": 4540
+    },
+    {
+      "epoch": 0.8736052327818392,
+      "grad_norm": 2.1238840666320677,
+      "learning_rate": 8.26268801461646e-07,
+      "loss": 0.8245,
+      "step": 4541
+    },
+    {
+      "epoch": 0.8737976144671027,
+      "grad_norm": 1.1953079347161524,
+      "learning_rate": 8.237905290355563e-07,
+      "loss": 0.8018,
+      "step": 4542
+    },
+    {
+      "epoch": 0.8739899961523663,
+      "grad_norm": 1.2464514411936227,
+      "learning_rate": 8.213158191299297e-07,
+      "loss": 0.8208,
+      "step": 4543
+    },
+    {
+      "epoch": 0.8741823778376299,
+      "grad_norm": 1.6065867553754143,
+      "learning_rate": 8.188446727055311e-07,
+      "loss": 0.8058,
+      "step": 4544
+    },
+    {
+      "epoch": 0.8743747595228935,
+      "grad_norm": 1.5406789153331202,
+      "learning_rate": 8.163770907217506e-07,
+      "loss": 0.8275,
+      "step": 4545
+    },
+    {
+      "epoch": 0.8745671412081569,
+      "grad_norm": 1.2522310613274525,
+      "learning_rate": 8.139130741365819e-07,
+      "loss": 0.8466,
+      "step": 4546
+    },
+    {
+      "epoch": 0.8747595228934205,
+      "grad_norm": 1.239653020701151,
+      "learning_rate": 8.114526239066456e-07,
+      "loss": 0.8266,
+      "step": 4547
+    },
+    {
+      "epoch": 0.8749519045786841,
+      "grad_norm": 1.3752471669228898,
+      "learning_rate": 8.08995740987173e-07,
+      "loss": 0.8067,
+      "step": 4548
+    },
+    {
+      "epoch": 0.8751442862639477,
+      "grad_norm": 1.25555659913159,
+      "learning_rate": 8.065424263320054e-07,
+      "loss": 0.7997,
+      "step": 4549
+    },
+    {
+      "epoch": 0.8753366679492113,
+      "grad_norm": 1.236150096527669,
+      "learning_rate": 8.040926808936112e-07,
+      "loss": 0.8161,
+      "step": 4550
+    },
+    {
+      "epoch": 0.8755290496344748,
+      "grad_norm": 1.3069485257782132,
+      "learning_rate": 8.016465056230616e-07,
+      "loss": 0.8217,
+      "step": 4551
+    },
+    {
+      "epoch": 0.8757214313197383,
+      "grad_norm": 1.3485031603274695,
+      "learning_rate": 7.99203901470047e-07,
+      "loss": 0.8446,
+      "step": 4552
+    },
+    {
+      "epoch": 0.8759138130050019,
+      "grad_norm": 1.492725561340997,
+      "learning_rate": 7.967648693828712e-07,
+      "loss": 0.8147,
+      "step": 4553
+    },
+    {
+      "epoch": 0.8761061946902655,
+      "grad_norm": 1.0900791277811943,
+      "learning_rate": 7.943294103084487e-07,
+      "loss": 0.7795,
+      "step": 4554
+    },
+    {
+      "epoch": 0.8762985763755291,
+      "grad_norm": 1.3097314418814954,
+      "learning_rate": 7.9189752519231e-07,
+      "loss": 0.7763,
+      "step": 4555
+    },
+    {
+      "epoch": 0.8764909580607926,
+      "grad_norm": 1.2171048581254307,
+      "learning_rate": 7.894692149785954e-07,
+      "loss": 0.7933,
+      "step": 4556
+    },
+    {
+      "epoch": 0.8766833397460562,
+      "grad_norm": 1.4155510505866948,
+      "learning_rate": 7.870444806100619e-07,
+      "loss": 0.8194,
+      "step": 4557
+    },
+    {
+      "epoch": 0.8768757214313198,
+      "grad_norm": 1.334427442709195,
+      "learning_rate": 7.846233230280698e-07,
+      "loss": 0.7892,
+      "step": 4558
+    },
+    {
+      "epoch": 0.8770681031165833,
+      "grad_norm": 1.3623817997588252,
+      "learning_rate": 7.822057431725994e-07,
+      "loss": 0.8476,
+      "step": 4559
+    },
+    {
+      "epoch": 0.8772604848018468,
+      "grad_norm": 1.3758826258729793,
+      "learning_rate": 7.797917419822377e-07,
+      "loss": 0.8514,
+      "step": 4560
+    },
+    {
+      "epoch": 0.8774528664871104,
+      "grad_norm": 1.8111593214039734,
+      "learning_rate": 7.773813203941827e-07,
+      "loss": 0.8453,
+      "step": 4561
+    },
+    {
+      "epoch": 0.877645248172374,
+      "grad_norm": 1.1925133829596897,
+      "learning_rate": 7.749744793442448e-07,
+      "loss": 0.814,
+      "step": 4562
+    },
+    {
+      "epoch": 0.8778376298576376,
+      "grad_norm": 1.2908966090738683,
+      "learning_rate": 7.725712197668378e-07,
+      "loss": 0.7828,
+      "step": 4563
+    },
+    {
+      "epoch": 0.8780300115429012,
+      "grad_norm": 1.2360362900257778,
+      "learning_rate": 7.701715425949952e-07,
+      "loss": 0.7999,
+      "step": 4564
+    },
+    {
+      "epoch": 0.8782223932281646,
+      "grad_norm": 1.9065637207981991,
+      "learning_rate": 7.677754487603517e-07,
+      "loss": 0.7651,
+      "step": 4565
+    },
+    {
+      "epoch": 0.8784147749134282,
+      "grad_norm": 1.462204493362984,
+      "learning_rate": 7.653829391931533e-07,
+      "loss": 0.7131,
+      "step": 4566
+    },
+    {
+      "epoch": 0.8786071565986918,
+      "grad_norm": 1.9157498290728,
+      "learning_rate": 7.629940148222559e-07,
+      "loss": 0.7829,
+      "step": 4567
+    },
+    {
+      "epoch": 0.8787995382839554,
+      "grad_norm": 1.3871726734355703,
+      "learning_rate": 7.606086765751209e-07,
+      "loss": 0.7757,
+      "step": 4568
+    },
+    {
+      "epoch": 0.8789919199692189,
+      "grad_norm": 1.2840127729163646,
+      "learning_rate": 7.582269253778185e-07,
+      "loss": 0.8895,
+      "step": 4569
+    },
+    {
+      "epoch": 0.8791843016544825,
+      "grad_norm": 1.449360016453471,
+      "learning_rate": 7.55848762155027e-07,
+      "loss": 0.828,
+      "step": 4570
+    },
+    {
+      "epoch": 0.8793766833397461,
+      "grad_norm": 1.1254773785661176,
+      "learning_rate": 7.534741878300333e-07,
+      "loss": 0.7764,
+      "step": 4571
+    },
+    {
+      "epoch": 0.8795690650250096,
+      "grad_norm": 1.8886805892763312,
+      "learning_rate": 7.511032033247256e-07,
+      "loss": 0.8051,
+      "step": 4572
+    },
+    {
+      "epoch": 0.8797614467102732,
+      "grad_norm": 1.566191287549585,
+      "learning_rate": 7.487358095596031e-07,
+      "loss": 0.9092,
+      "step": 4573
+    },
+    {
+      "epoch": 0.8799538283955367,
+      "grad_norm": 1.2460852782363567,
+      "learning_rate": 7.463720074537728e-07,
+      "loss": 0.8622,
+      "step": 4574
+    },
+    {
+      "epoch": 0.8801462100808003,
+      "grad_norm": 2.6120397379520846,
+      "learning_rate": 7.440117979249362e-07,
+      "loss": 0.8367,
+      "step": 4575
+    },
+    {
+      "epoch": 0.8803385917660639,
+      "grad_norm": 1.2319356424870465,
+      "learning_rate": 7.416551818894158e-07,
+      "loss": 0.8541,
+      "step": 4576
+    },
+    {
+      "epoch": 0.8805309734513275,
+      "grad_norm": 1.2856421809862755,
+      "learning_rate": 7.393021602621264e-07,
+      "loss": 0.8049,
+      "step": 4577
+    },
+    {
+      "epoch": 0.880723355136591,
+      "grad_norm": 1.4895555905775582,
+      "learning_rate": 7.369527339565951e-07,
+      "loss": 0.8142,
+      "step": 4578
+    },
+    {
+      "epoch": 0.8809157368218545,
+      "grad_norm": 1.3102619491553376,
+      "learning_rate": 7.346069038849469e-07,
+      "loss": 0.7879,
+      "step": 4579
+    },
+    {
+      "epoch": 0.8811081185071181,
+      "grad_norm": 1.162458322373329,
+      "learning_rate": 7.322646709579173e-07,
+      "loss": 0.7565,
+      "step": 4580
+    },
+    {
+      "epoch": 0.8813005001923817,
+      "grad_norm": 1.3174701416271755,
+      "learning_rate": 7.299260360848382e-07,
+      "loss": 0.8243,
+      "step": 4581
+    },
+    {
+      "epoch": 0.8814928818776453,
+      "grad_norm": 1.3440733112587329,
+      "learning_rate": 7.275910001736497e-07,
+      "loss": 0.8171,
+      "step": 4582
+    },
+    {
+      "epoch": 0.8816852635629088,
+      "grad_norm": 1.1767841763739242,
+      "learning_rate": 7.252595641308957e-07,
+      "loss": 0.8519,
+      "step": 4583
+    },
+    {
+      "epoch": 0.8818776452481724,
+      "grad_norm": 2.3202943089587085,
+      "learning_rate": 7.229317288617144e-07,
+      "loss": 0.8207,
+      "step": 4584
+    },
+    {
+      "epoch": 0.882070026933436,
+      "grad_norm": 1.5527125393237686,
+      "learning_rate": 7.20607495269856e-07,
+      "loss": 0.7687,
+      "step": 4585
+    },
+    {
+      "epoch": 0.8822624086186995,
+      "grad_norm": 1.3063532116499228,
+      "learning_rate": 7.182868642576679e-07,
+      "loss": 0.7901,
+      "step": 4586
+    },
+    {
+      "epoch": 0.882454790303963,
+      "grad_norm": 1.3523151691378739,
+      "learning_rate": 7.15969836726097e-07,
+      "loss": 0.8104,
+      "step": 4587
+    },
+    {
+      "epoch": 0.8826471719892266,
+      "grad_norm": 1.2560272000512167,
+      "learning_rate": 7.13656413574696e-07,
+      "loss": 0.7935,
+      "step": 4588
+    },
+    {
+      "epoch": 0.8828395536744902,
+      "grad_norm": 1.2486897959534307,
+      "learning_rate": 7.113465957016097e-07,
+      "loss": 0.8523,
+      "step": 4589
+    },
+    {
+      "epoch": 0.8830319353597538,
+      "grad_norm": 1.3079714809905376,
+      "learning_rate": 7.090403840035942e-07,
+      "loss": 0.8046,
+      "step": 4590
+    },
+    {
+      "epoch": 0.8832243170450174,
+      "grad_norm": 1.351620812210653,
+      "learning_rate": 7.067377793759999e-07,
+      "loss": 0.8133,
+      "step": 4591
+    },
+    {
+      "epoch": 0.8834166987302808,
+      "grad_norm": 1.4665736021570641,
+      "learning_rate": 7.044387827127752e-07,
+      "loss": 0.8625,
+      "step": 4592
+    },
+    {
+      "epoch": 0.8836090804155444,
+      "grad_norm": 1.2817841271254922,
+      "learning_rate": 7.021433949064704e-07,
+      "loss": 0.7925,
+      "step": 4593
+    },
+    {
+      "epoch": 0.883801462100808,
+      "grad_norm": 1.812456278293631,
+      "learning_rate": 6.99851616848235e-07,
+      "loss": 0.8226,
+      "step": 4594
+    },
+    {
+      "epoch": 0.8839938437860716,
+      "grad_norm": 1.4155437443478416,
+      "learning_rate": 6.975634494278149e-07,
+      "loss": 0.8311,
+      "step": 4595
+    },
+    {
+      "epoch": 0.8841862254713351,
+      "grad_norm": 1.840067480094761,
+      "learning_rate": 6.952788935335541e-07,
+      "loss": 0.815,
+      "step": 4596
+    },
+    {
+      "epoch": 0.8843786071565987,
+      "grad_norm": 1.2270794370540081,
+      "learning_rate": 6.92997950052402e-07,
+      "loss": 0.8274,
+      "step": 4597
+    },
+    {
+      "epoch": 0.8845709888418622,
+      "grad_norm": 1.0870115181802409,
+      "learning_rate": 6.907206198698912e-07,
+      "loss": 0.8038,
+      "step": 4598
+    },
+    {
+      "epoch": 0.8847633705271258,
+      "grad_norm": 1.26768302869436,
+      "learning_rate": 6.884469038701646e-07,
+      "loss": 0.8089,
+      "step": 4599
+    },
+    {
+      "epoch": 0.8849557522123894,
+      "grad_norm": 1.28085112426787,
+      "learning_rate": 6.861768029359595e-07,
+      "loss": 0.8452,
+      "step": 4600
+    },
+    {
+      "epoch": 0.8851481338976529,
+      "grad_norm": 1.1580670607589298,
+      "learning_rate": 6.839103179485995e-07,
+      "loss": 0.8291,
+      "step": 4601
+    },
+    {
+      "epoch": 0.8853405155829165,
+      "grad_norm": 1.4398867661040238,
+      "learning_rate": 6.816474497880177e-07,
+      "loss": 0.8094,
+      "step": 4602
+    },
+    {
+      "epoch": 0.8855328972681801,
+      "grad_norm": 1.3935861857652094,
+      "learning_rate": 6.793881993327366e-07,
+      "loss": 0.8261,
+      "step": 4603
+    },
+    {
+      "epoch": 0.8857252789534437,
+      "grad_norm": 1.312978338551224,
+      "learning_rate": 6.77132567459875e-07,
+      "loss": 0.8252,
+      "step": 4604
+    },
+    {
+      "epoch": 0.8859176606387072,
+      "grad_norm": 1.583864674949652,
+      "learning_rate": 6.748805550451453e-07,
+      "loss": 0.86,
+      "step": 4605
+    },
+    {
+      "epoch": 0.8861100423239707,
+      "grad_norm": 1.17577629836967,
+      "learning_rate": 6.726321629628585e-07,
+      "loss": 0.8467,
+      "step": 4606
+    },
+    {
+      "epoch": 0.8863024240092343,
+      "grad_norm": 1.3165321354860506,
+      "learning_rate": 6.703873920859161e-07,
+      "loss": 0.7997,
+      "step": 4607
+    },
+    {
+      "epoch": 0.8864948056944979,
+      "grad_norm": 1.2371076624755388,
+      "learning_rate": 6.681462432858154e-07,
+      "loss": 0.8427,
+      "step": 4608
+    },
+    {
+      "epoch": 0.8866871873797615,
+      "grad_norm": 0.5862297984481205,
+      "learning_rate": 6.659087174326506e-07,
+      "loss": 0.3261,
+      "step": 4609
+    },
+    {
+      "epoch": 0.886879569065025,
+      "grad_norm": 6.260877774792897,
+      "learning_rate": 6.636748153951e-07,
+      "loss": 0.8246,
+      "step": 4610
+    },
+    {
+      "epoch": 0.8870719507502886,
+      "grad_norm": 0.6110105477943197,
+      "learning_rate": 6.614445380404478e-07,
+      "loss": 0.3427,
+      "step": 4611
+    },
+    {
+      "epoch": 0.8872643324355521,
+      "grad_norm": 1.3131633724240392,
+      "learning_rate": 6.592178862345622e-07,
+      "loss": 0.7608,
+      "step": 4612
+    },
+    {
+      "epoch": 0.8874567141208157,
+      "grad_norm": 1.372808142969139,
+      "learning_rate": 6.569948608419041e-07,
+      "loss": 0.8372,
+      "step": 4613
+    },
+    {
+      "epoch": 0.8876490958060793,
+      "grad_norm": 1.2895568586315758,
+      "learning_rate": 6.547754627255332e-07,
+      "loss": 0.8143,
+      "step": 4614
+    },
+    {
+      "epoch": 0.8878414774913428,
+      "grad_norm": 1.2526264989061495,
+      "learning_rate": 6.52559692747089e-07,
+      "loss": 0.8458,
+      "step": 4615
+    },
+    {
+      "epoch": 0.8880338591766064,
+      "grad_norm": 1.4838819535274794,
+      "learning_rate": 6.503475517668168e-07,
+      "loss": 0.8592,
+      "step": 4616
+    },
+    {
+      "epoch": 0.88822624086187,
+      "grad_norm": 1.5604121435811587,
+      "learning_rate": 6.481390406435417e-07,
+      "loss": 0.8244,
+      "step": 4617
+    },
+    {
+      "epoch": 0.8884186225471336,
+      "grad_norm": 1.5040164577920456,
+      "learning_rate": 6.459341602346858e-07,
+      "loss": 0.8619,
+      "step": 4618
+    },
+    {
+      "epoch": 0.888611004232397,
+      "grad_norm": 0.5755699032790911,
+      "learning_rate": 6.437329113962576e-07,
+      "loss": 0.3563,
+      "step": 4619
+    },
+    {
+      "epoch": 0.8888033859176606,
+      "grad_norm": 1.1624305561609034,
+      "learning_rate": 6.415352949828601e-07,
+      "loss": 0.8268,
+      "step": 4620
+    },
+    {
+      "epoch": 0.8889957676029242,
+      "grad_norm": 1.2826447026368963,
+      "learning_rate": 6.393413118476821e-07,
+      "loss": 0.8418,
+      "step": 4621
+    },
+    {
+      "epoch": 0.8891881492881878,
+      "grad_norm": 1.904015643448291,
+      "learning_rate": 6.371509628425021e-07,
+      "loss": 0.813,
+      "step": 4622
+    },
+    {
+      "epoch": 0.8893805309734514,
+      "grad_norm": 1.2925379921694107,
+      "learning_rate": 6.349642488176943e-07,
+      "loss": 0.8311,
+      "step": 4623
+    },
+    {
+      "epoch": 0.8895729126587149,
+      "grad_norm": 1.2383293660567616,
+      "learning_rate": 6.327811706222097e-07,
+      "loss": 0.83,
+      "step": 4624
+    },
+    {
+      "epoch": 0.8897652943439784,
+      "grad_norm": 1.6353312204148316,
+      "learning_rate": 6.306017291035981e-07,
+      "loss": 0.8225,
+      "step": 4625
+    },
+    {
+      "epoch": 0.889957676029242,
+      "grad_norm": 1.3760586358184475,
+      "learning_rate": 6.284259251079939e-07,
+      "loss": 0.7803,
+      "step": 4626
+    },
+    {
+      "epoch": 0.8901500577145056,
+      "grad_norm": 1.3876952807823115,
+      "learning_rate": 6.262537594801177e-07,
+      "loss": 0.8252,
+      "step": 4627
+    },
+    {
+      "epoch": 0.8903424393997691,
+      "grad_norm": 1.2734857862964322,
+      "learning_rate": 6.240852330632796e-07,
+      "loss": 0.8222,
+      "step": 4628
+    },
+    {
+      "epoch": 0.8905348210850327,
+      "grad_norm": 1.2563972077866021,
+      "learning_rate": 6.219203466993762e-07,
+      "loss": 0.8166,
+      "step": 4629
+    },
+    {
+      "epoch": 0.8907272027702963,
+      "grad_norm": 1.272581000299688,
+      "learning_rate": 6.197591012288918e-07,
+      "loss": 0.8117,
+      "step": 4630
+    },
+    {
+      "epoch": 0.8909195844555599,
+      "grad_norm": 1.4607381801983113,
+      "learning_rate": 6.17601497490895e-07,
+      "loss": 0.8235,
+      "step": 4631
+    },
+    {
+      "epoch": 0.8911119661408234,
+      "grad_norm": 1.197330765787745,
+      "learning_rate": 6.154475363230417e-07,
+      "loss": 0.8489,
+      "step": 4632
+    },
+    {
+      "epoch": 0.8913043478260869,
+      "grad_norm": 1.4644128798888558,
+      "learning_rate": 6.132972185615749e-07,
+      "loss": 0.8617,
+      "step": 4633
+    },
+    {
+      "epoch": 0.8914967295113505,
+      "grad_norm": 1.1729432528595352,
+      "learning_rate": 6.111505450413202e-07,
+      "loss": 0.8685,
+      "step": 4634
+    },
+    {
+      "epoch": 0.8916891111966141,
+      "grad_norm": 1.2706037540525639,
+      "learning_rate": 6.090075165956943e-07,
+      "loss": 0.8404,
+      "step": 4635
+    },
+    {
+      "epoch": 0.8918814928818777,
+      "grad_norm": 1.3630420121453057,
+      "learning_rate": 6.068681340566896e-07,
+      "loss": 0.7815,
+      "step": 4636
+    },
+    {
+      "epoch": 0.8920738745671412,
+      "grad_norm": 1.2403728003790462,
+      "learning_rate": 6.047323982548924e-07,
+      "loss": 0.8216,
+      "step": 4637
+    },
+    {
+      "epoch": 0.8922662562524047,
+      "grad_norm": 1.2664477521145816,
+      "learning_rate": 6.026003100194633e-07,
+      "loss": 0.8604,
+      "step": 4638
+    },
+    {
+      "epoch": 0.8924586379376683,
+      "grad_norm": 1.3060860961142322,
+      "learning_rate": 6.004718701781575e-07,
+      "loss": 0.8267,
+      "step": 4639
+    },
+    {
+      "epoch": 0.8926510196229319,
+      "grad_norm": 1.1063739128041947,
+      "learning_rate": 5.983470795573088e-07,
+      "loss": 0.8101,
+      "step": 4640
+    },
+    {
+      "epoch": 0.8928434013081955,
+      "grad_norm": 1.4711499041800975,
+      "learning_rate": 5.962259389818292e-07,
+      "loss": 0.8188,
+      "step": 4641
+    },
+    {
+      "epoch": 0.893035782993459,
+      "grad_norm": 0.6300655165565238,
+      "learning_rate": 5.941084492752236e-07,
+      "loss": 0.3241,
+      "step": 4642
+    },
+    {
+      "epoch": 0.8932281646787226,
+      "grad_norm": 1.2963620886772784,
+      "learning_rate": 5.91994611259572e-07,
+      "loss": 0.809,
+      "step": 4643
+    },
+    {
+      "epoch": 0.8934205463639862,
+      "grad_norm": 1.2255671023143095,
+      "learning_rate": 5.898844257555392e-07,
+      "loss": 0.8403,
+      "step": 4644
+    },
+    {
+      "epoch": 0.8936129280492497,
+      "grad_norm": 3.2347484182093,
+      "learning_rate": 5.87777893582372e-07,
+      "loss": 0.7985,
+      "step": 4645
+    },
+    {
+      "epoch": 0.8938053097345132,
+      "grad_norm": 1.335196684620126,
+      "learning_rate": 5.856750155578983e-07,
+      "loss": 0.8,
+      "step": 4646
+    },
+    {
+      "epoch": 0.8939976914197768,
+      "grad_norm": 1.2012844169603607,
+      "learning_rate": 5.835757924985286e-07,
+      "loss": 0.8369,
+      "step": 4647
+    },
+    {
+      "epoch": 0.8941900731050404,
+      "grad_norm": 1.607278850917337,
+      "learning_rate": 5.81480225219252e-07,
+      "loss": 0.8453,
+      "step": 4648
+    },
+    {
+      "epoch": 0.894382454790304,
+      "grad_norm": 1.1834106460778513,
+      "learning_rate": 5.793883145336443e-07,
+      "loss": 0.8214,
+      "step": 4649
+    },
+    {
+      "epoch": 0.8945748364755676,
+      "grad_norm": 1.352379209753852,
+      "learning_rate": 5.773000612538505e-07,
+      "loss": 0.8567,
+      "step": 4650
+    },
+    {
+      "epoch": 0.894767218160831,
+      "grad_norm": 1.3636779946772928,
+      "learning_rate": 5.752154661906085e-07,
+      "loss": 0.8504,
+      "step": 4651
+    },
+    {
+      "epoch": 0.8949595998460946,
+      "grad_norm": 1.425078498529117,
+      "learning_rate": 5.731345301532265e-07,
+      "loss": 0.8363,
+      "step": 4652
+    },
+    {
+      "epoch": 0.8951519815313582,
+      "grad_norm": 1.3295496653243526,
+      "learning_rate": 5.710572539495962e-07,
+      "loss": 0.7874,
+      "step": 4653
+    },
+    {
+      "epoch": 0.8953443632166218,
+      "grad_norm": 1.210261131667101,
+      "learning_rate": 5.68983638386188e-07,
+      "loss": 0.8718,
+      "step": 4654
+    },
+    {
+      "epoch": 0.8955367449018854,
+      "grad_norm": 1.351006462575013,
+      "learning_rate": 5.669136842680512e-07,
+      "loss": 0.8377,
+      "step": 4655
+    },
+    {
+      "epoch": 0.8957291265871489,
+      "grad_norm": 1.2424327386726226,
+      "learning_rate": 5.648473923988129e-07,
+      "loss": 0.7885,
+      "step": 4656
+    },
+    {
+      "epoch": 0.8959215082724125,
+      "grad_norm": 1.2879857620735589,
+      "learning_rate": 5.627847635806771e-07,
+      "loss": 0.7771,
+      "step": 4657
+    },
+    {
+      "epoch": 0.896113889957676,
+      "grad_norm": 1.3387135464029118,
+      "learning_rate": 5.607257986144321e-07,
+      "loss": 0.7998,
+      "step": 4658
+    },
+    {
+      "epoch": 0.8963062716429396,
+      "grad_norm": 1.1918427983251114,
+      "learning_rate": 5.58670498299434e-07,
+      "loss": 0.8368,
+      "step": 4659
+    },
+    {
+      "epoch": 0.8964986533282031,
+      "grad_norm": 1.3321311027340428,
+      "learning_rate": 5.566188634336212e-07,
+      "loss": 0.8253,
+      "step": 4660
+    },
+    {
+      "epoch": 0.8966910350134667,
+      "grad_norm": 1.3829070905848773,
+      "learning_rate": 5.545708948135142e-07,
+      "loss": 0.8195,
+      "step": 4661
+    },
+    {
+      "epoch": 0.8968834166987303,
+      "grad_norm": 1.5346955455939184,
+      "learning_rate": 5.525265932341984e-07,
+      "loss": 0.8454,
+      "step": 4662
+    },
+    {
+      "epoch": 0.8970757983839939,
+      "grad_norm": 1.103600812691196,
+      "learning_rate": 5.504859594893475e-07,
+      "loss": 0.8315,
+      "step": 4663
+    },
+    {
+      "epoch": 0.8972681800692575,
+      "grad_norm": 1.846869790759302,
+      "learning_rate": 5.484489943712013e-07,
+      "loss": 0.7918,
+      "step": 4664
+    },
+    {
+      "epoch": 0.8974605617545209,
+      "grad_norm": 1.204820494556778,
+      "learning_rate": 5.464156986705826e-07,
+      "loss": 0.84,
+      "step": 4665
+    },
+    {
+      "epoch": 0.8976529434397845,
+      "grad_norm": 1.5677520071400564,
+      "learning_rate": 5.443860731768869e-07,
+      "loss": 0.8469,
+      "step": 4666
+    },
+    {
+      "epoch": 0.8978453251250481,
+      "grad_norm": 1.8147205647573568,
+      "learning_rate": 5.423601186780836e-07,
+      "loss": 0.8457,
+      "step": 4667
+    },
+    {
+      "epoch": 0.8980377068103117,
+      "grad_norm": 1.21742180187014,
+      "learning_rate": 5.403378359607181e-07,
+      "loss": 0.7968,
+      "step": 4668
+    },
+    {
+      "epoch": 0.8982300884955752,
+      "grad_norm": 1.4729798215226844,
+      "learning_rate": 5.383192258099113e-07,
+      "loss": 0.8515,
+      "step": 4669
+    },
+    {
+      "epoch": 0.8984224701808388,
+      "grad_norm": 2.503985339574427,
+      "learning_rate": 5.36304289009355e-07,
+      "loss": 0.807,
+      "step": 4670
+    },
+    {
+      "epoch": 0.8986148518661023,
+      "grad_norm": 1.3290463859819162,
+      "learning_rate": 5.342930263413193e-07,
+      "loss": 0.776,
+      "step": 4671
+    },
+    {
+      "epoch": 0.8988072335513659,
+      "grad_norm": 1.2857127112871611,
+      "learning_rate": 5.322854385866439e-07,
+      "loss": 0.7782,
+      "step": 4672
+    },
+    {
+      "epoch": 0.8989996152366295,
+      "grad_norm": 1.2345561596727799,
+      "learning_rate": 5.302815265247452e-07,
+      "loss": 0.893,
+      "step": 4673
+    },
+    {
+      "epoch": 0.899191996921893,
+      "grad_norm": 1.4781996665917039,
+      "learning_rate": 5.282812909336077e-07,
+      "loss": 0.8045,
+      "step": 4674
+    },
+    {
+      "epoch": 0.8993843786071566,
+      "grad_norm": 1.8195317945580864,
+      "learning_rate": 5.262847325897968e-07,
+      "loss": 0.7968,
+      "step": 4675
+    },
+    {
+      "epoch": 0.8995767602924202,
+      "grad_norm": 1.2796004552957,
+      "learning_rate": 5.242918522684392e-07,
+      "loss": 0.8461,
+      "step": 4676
+    },
+    {
+      "epoch": 0.8997691419776838,
+      "grad_norm": 1.2499497450218715,
+      "learning_rate": 5.22302650743245e-07,
+      "loss": 0.8447,
+      "step": 4677
+    },
+    {
+      "epoch": 0.8999615236629472,
+      "grad_norm": 1.161958846256562,
+      "learning_rate": 5.203171287864872e-07,
+      "loss": 0.8095,
+      "step": 4678
+    },
+    {
+      "epoch": 0.9001539053482108,
+      "grad_norm": 0.6369884148632012,
+      "learning_rate": 5.183352871690162e-07,
+      "loss": 0.3492,
+      "step": 4679
+    },
+    {
+      "epoch": 0.9003462870334744,
+      "grad_norm": 1.3449126305761419,
+      "learning_rate": 5.163571266602485e-07,
+      "loss": 0.8427,
+      "step": 4680
+    },
+    {
+      "epoch": 0.900538668718738,
+      "grad_norm": 1.297929309408672,
+      "learning_rate": 5.143826480281778e-07,
+      "loss": 0.8109,
+      "step": 4681
+    },
+    {
+      "epoch": 0.9007310504040016,
+      "grad_norm": 1.665538762830539,
+      "learning_rate": 5.124118520393606e-07,
+      "loss": 0.8505,
+      "step": 4682
+    },
+    {
+      "epoch": 0.9009234320892651,
+      "grad_norm": 1.4904194741977927,
+      "learning_rate": 5.104447394589295e-07,
+      "loss": 0.814,
+      "step": 4683
+    },
+    {
+      "epoch": 0.9011158137745287,
+      "grad_norm": 1.2664122761706487,
+      "learning_rate": 5.084813110505871e-07,
+      "loss": 0.8475,
+      "step": 4684
+    },
+    {
+      "epoch": 0.9013081954597922,
+      "grad_norm": 1.2801476255144848,
+      "learning_rate": 5.065215675766023e-07,
+      "loss": 0.8125,
+      "step": 4685
+    },
+    {
+      "epoch": 0.9015005771450558,
+      "grad_norm": 1.1459116856907554,
+      "learning_rate": 5.045655097978131e-07,
+      "loss": 0.8032,
+      "step": 4686
+    },
+    {
+      "epoch": 0.9016929588303193,
+      "grad_norm": 1.2756383418603676,
+      "learning_rate": 5.026131384736321e-07,
+      "loss": 0.823,
+      "step": 4687
+    },
+    {
+      "epoch": 0.9018853405155829,
+      "grad_norm": 1.1151771693244683,
+      "learning_rate": 5.006644543620342e-07,
+      "loss": 0.8665,
+      "step": 4688
+    },
+    {
+      "epoch": 0.9020777222008465,
+      "grad_norm": 1.257508818756427,
+      "learning_rate": 4.987194582195687e-07,
+      "loss": 0.8183,
+      "step": 4689
+    },
+    {
+      "epoch": 0.9022701038861101,
+      "grad_norm": 1.3523853950424356,
+      "learning_rate": 4.967781508013459e-07,
+      "loss": 0.7978,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9024624855713737,
+      "grad_norm": 1.2111198906029566,
+      "learning_rate": 4.948405328610506e-07,
+      "loss": 0.8343,
+      "step": 4691
+    },
+    {
+      "epoch": 0.9026548672566371,
+      "grad_norm": 1.2250579026193604,
+      "learning_rate": 4.929066051509346e-07,
+      "loss": 0.865,
+      "step": 4692
+    },
+    {
+      "epoch": 0.9028472489419007,
+      "grad_norm": 1.4813340680319418,
+      "learning_rate": 4.909763684218116e-07,
+      "loss": 0.7669,
+      "step": 4693
+    },
+    {
+      "epoch": 0.9030396306271643,
+      "grad_norm": 1.3285433142138843,
+      "learning_rate": 4.890498234230689e-07,
+      "loss": 0.8065,
+      "step": 4694
+    },
+    {
+      "epoch": 0.9032320123124279,
+      "grad_norm": 1.3052369056080688,
+      "learning_rate": 4.871269709026561e-07,
+      "loss": 0.8255,
+      "step": 4695
+    },
+    {
+      "epoch": 0.9034243939976914,
+      "grad_norm": 1.4285102208353624,
+      "learning_rate": 4.852078116070902e-07,
+      "loss": 0.7941,
+      "step": 4696
+    },
+    {
+      "epoch": 0.903616775682955,
+      "grad_norm": 1.4124281343470457,
+      "learning_rate": 4.832923462814565e-07,
+      "loss": 0.8161,
+      "step": 4697
+    },
+    {
+      "epoch": 0.9038091573682185,
+      "grad_norm": 1.2837024782340691,
+      "learning_rate": 4.813805756694035e-07,
+      "loss": 0.7857,
+      "step": 4698
+    },
+    {
+      "epoch": 0.9040015390534821,
+      "grad_norm": 2.528884015914698,
+      "learning_rate": 4.794725005131462e-07,
+      "loss": 0.8516,
+      "step": 4699
+    },
+    {
+      "epoch": 0.9041939207387457,
+      "grad_norm": 1.5353283280553416,
+      "learning_rate": 4.775681215534656e-07,
+      "loss": 0.7977,
+      "step": 4700
+    },
+    {
+      "epoch": 0.9043863024240092,
+      "grad_norm": 1.6263212614440306,
+      "learning_rate": 4.7566743952970894e-07,
+      "loss": 0.8123,
+      "step": 4701
+    },
+    {
+      "epoch": 0.9045786841092728,
+      "grad_norm": 0.5763939160450497,
+      "learning_rate": 4.7377045517978173e-07,
+      "loss": 0.3034,
+      "step": 4702
+    },
+    {
+      "epoch": 0.9047710657945364,
+      "grad_norm": 1.4266239716161702,
+      "learning_rate": 4.7187716924016355e-07,
+      "loss": 0.8427,
+      "step": 4703
+    },
+    {
+      "epoch": 0.9049634474798,
+      "grad_norm": 1.337454540425443,
+      "learning_rate": 4.6998758244588995e-07,
+      "loss": 0.7819,
+      "step": 4704
+    },
+    {
+      "epoch": 0.9051558291650635,
+      "grad_norm": 1.709263839741029,
+      "learning_rate": 4.6810169553056616e-07,
+      "loss": 0.7786,
+      "step": 4705
+    },
+    {
+      "epoch": 0.905348210850327,
+      "grad_norm": 1.4898662794577442,
+      "learning_rate": 4.662195092263566e-07,
+      "loss": 0.8013,
+      "step": 4706
+    },
+    {
+      "epoch": 0.9055405925355906,
+      "grad_norm": 1.2286626313087807,
+      "learning_rate": 4.643410242639912e-07,
+      "loss": 0.8248,
+      "step": 4707
+    },
+    {
+      "epoch": 0.9057329742208542,
+      "grad_norm": 1.0842166420301607,
+      "learning_rate": 4.6246624137276206e-07,
+      "loss": 0.8466,
+      "step": 4708
+    },
+    {
+      "epoch": 0.9059253559061178,
+      "grad_norm": 1.2956549873603997,
+      "learning_rate": 4.605951612805237e-07,
+      "loss": 0.8404,
+      "step": 4709
+    },
+    {
+      "epoch": 0.9061177375913813,
+      "grad_norm": 1.4767185520402646,
+      "learning_rate": 4.587277847136984e-07,
+      "loss": 0.8316,
+      "step": 4710
+    },
+    {
+      "epoch": 0.9063101192766448,
+      "grad_norm": 1.4069372398930358,
+      "learning_rate": 4.568641123972606e-07,
+      "loss": 0.8352,
+      "step": 4711
+    },
+    {
+      "epoch": 0.9065025009619084,
+      "grad_norm": 1.3132655305548537,
+      "learning_rate": 4.550041450547549e-07,
+      "loss": 0.8171,
+      "step": 4712
+    },
+    {
+      "epoch": 0.906694882647172,
+      "grad_norm": 1.4392755446531558,
+      "learning_rate": 4.5314788340828365e-07,
+      "loss": 0.8263,
+      "step": 4713
+    },
+    {
+      "epoch": 0.9068872643324356,
+      "grad_norm": 1.2788560136424052,
+      "learning_rate": 4.512953281785104e-07,
+      "loss": 0.836,
+      "step": 4714
+    },
+    {
+      "epoch": 0.9070796460176991,
+      "grad_norm": 1.418940592570974,
+      "learning_rate": 4.494464800846654e-07,
+      "loss": 0.8357,
+      "step": 4715
+    },
+    {
+      "epoch": 0.9072720277029627,
+      "grad_norm": 1.4147171968579448,
+      "learning_rate": 4.476013398445289e-07,
+      "loss": 0.7789,
+      "step": 4716
+    },
+    {
+      "epoch": 0.9074644093882263,
+      "grad_norm": 1.1870612915017709,
+      "learning_rate": 4.4575990817445234e-07,
+      "loss": 0.8753,
+      "step": 4717
+    },
+    {
+      "epoch": 0.9076567910734898,
+      "grad_norm": 1.4355888197696511,
+      "learning_rate": 4.4392218578934164e-07,
+      "loss": 0.85,
+      "step": 4718
+    },
+    {
+      "epoch": 0.9078491727587533,
+      "grad_norm": 1.1342759818092907,
+      "learning_rate": 4.4208817340266385e-07,
+      "loss": 0.8444,
+      "step": 4719
+    },
+    {
+      "epoch": 0.9080415544440169,
+      "grad_norm": 1.237768378018017,
+      "learning_rate": 4.4025787172644495e-07,
+      "loss": 0.8616,
+      "step": 4720
+    },
+    {
+      "epoch": 0.9082339361292805,
+      "grad_norm": 1.398852001145939,
+      "learning_rate": 4.384312814712721e-07,
+      "loss": 0.7889,
+      "step": 4721
+    },
+    {
+      "epoch": 0.9084263178145441,
+      "grad_norm": 1.3033086474766584,
+      "learning_rate": 4.366084033462914e-07,
+      "loss": 0.8557,
+      "step": 4722
+    },
+    {
+      "epoch": 0.9086186994998077,
+      "grad_norm": 2.9416491355445107,
+      "learning_rate": 4.3478923805920335e-07,
+      "loss": 0.8493,
+      "step": 4723
+    },
+    {
+      "epoch": 0.9088110811850711,
+      "grad_norm": 1.2885407430074516,
+      "learning_rate": 4.329737863162753e-07,
+      "loss": 0.7462,
+      "step": 4724
+    },
+    {
+      "epoch": 0.9090034628703347,
+      "grad_norm": 1.2692243536065382,
+      "learning_rate": 4.311620488223256e-07,
+      "loss": 0.769,
+      "step": 4725
+    },
+    {
+      "epoch": 0.9091958445555983,
+      "grad_norm": 1.2123521049560944,
+      "learning_rate": 4.2935402628073166e-07,
+      "loss": 0.8256,
+      "step": 4726
+    },
+    {
+      "epoch": 0.9093882262408619,
+      "grad_norm": 1.3471273550729232,
+      "learning_rate": 4.27549719393433e-07,
+      "loss": 0.8271,
+      "step": 4727
+    },
+    {
+      "epoch": 0.9095806079261254,
+      "grad_norm": 1.5894701838830276,
+      "learning_rate": 4.2574912886092166e-07,
+      "loss": 0.8236,
+      "step": 4728
+    },
+    {
+      "epoch": 0.909772989611389,
+      "grad_norm": 1.4224991278331727,
+      "learning_rate": 4.239522553822495e-07,
+      "loss": 0.8559,
+      "step": 4729
+    },
+    {
+      "epoch": 0.9099653712966526,
+      "grad_norm": 1.2976093504753698,
+      "learning_rate": 4.221590996550251e-07,
+      "loss": 0.7904,
+      "step": 4730
+    },
+    {
+      "epoch": 0.9101577529819161,
+      "grad_norm": 2.615104174869876,
+      "learning_rate": 4.203696623754139e-07,
+      "loss": 0.8043,
+      "step": 4731
+    },
+    {
+      "epoch": 0.9103501346671797,
+      "grad_norm": 1.3916471909361616,
+      "learning_rate": 4.1858394423813563e-07,
+      "loss": 0.8595,
+      "step": 4732
+    },
+    {
+      "epoch": 0.9105425163524432,
+      "grad_norm": 1.4410804786362283,
+      "learning_rate": 4.1680194593646696e-07,
+      "loss": 0.7988,
+      "step": 4733
+    },
+    {
+      "epoch": 0.9107348980377068,
+      "grad_norm": 1.3316627243239985,
+      "learning_rate": 4.1502366816224327e-07,
+      "loss": 0.8631,
+      "step": 4734
+    },
+    {
+      "epoch": 0.9109272797229704,
+      "grad_norm": 1.3975228538422573,
+      "learning_rate": 4.1324911160585014e-07,
+      "loss": 0.8331,
+      "step": 4735
+    },
+    {
+      "epoch": 0.911119661408234,
+      "grad_norm": 1.3963123685413525,
+      "learning_rate": 4.1147827695623643e-07,
+      "loss": 0.8498,
+      "step": 4736
+    },
+    {
+      "epoch": 0.9113120430934974,
+      "grad_norm": 1.4310401702715283,
+      "learning_rate": 4.097111649008967e-07,
+      "loss": 0.8228,
+      "step": 4737
+    },
+    {
+      "epoch": 0.911504424778761,
+      "grad_norm": 1.2937478540672374,
+      "learning_rate": 4.0794777612588543e-07,
+      "loss": 0.8121,
+      "step": 4738
+    },
+    {
+      "epoch": 0.9116968064640246,
+      "grad_norm": 1.1574017503941503,
+      "learning_rate": 4.061881113158117e-07,
+      "loss": 0.8106,
+      "step": 4739
+    },
+    {
+      "epoch": 0.9118891881492882,
+      "grad_norm": 1.166713516523879,
+      "learning_rate": 4.044321711538368e-07,
+      "loss": 0.8174,
+      "step": 4740
+    },
+    {
+      "epoch": 0.9120815698345518,
+      "grad_norm": 1.3813721429960175,
+      "learning_rate": 4.02679956321681e-07,
+      "loss": 0.8082,
+      "step": 4741
+    },
+    {
+      "epoch": 0.9122739515198153,
+      "grad_norm": 1.3107788776102345,
+      "learning_rate": 4.00931467499609e-07,
+      "loss": 0.8542,
+      "step": 4742
+    },
+    {
+      "epoch": 0.9124663332050789,
+      "grad_norm": 1.3399629162908337,
+      "learning_rate": 3.9918670536644776e-07,
+      "loss": 0.8751,
+      "step": 4743
+    },
+    {
+      "epoch": 0.9126587148903424,
+      "grad_norm": 1.2136664594682007,
+      "learning_rate": 3.974456705995733e-07,
+      "loss": 0.824,
+      "step": 4744
+    },
+    {
+      "epoch": 0.912851096575606,
+      "grad_norm": 1.4481801700900352,
+      "learning_rate": 3.9570836387491487e-07,
+      "loss": 0.7957,
+      "step": 4745
+    },
+    {
+      "epoch": 0.9130434782608695,
+      "grad_norm": 1.2853367146008334,
+      "learning_rate": 3.9397478586695513e-07,
+      "loss": 0.845,
+      "step": 4746
+    },
+    {
+      "epoch": 0.9132358599461331,
+      "grad_norm": 1.212046924222265,
+      "learning_rate": 3.9224493724872915e-07,
+      "loss": 0.8471,
+      "step": 4747
+    },
+    {
+      "epoch": 0.9134282416313967,
+      "grad_norm": 4.7868585038045826,
+      "learning_rate": 3.90518818691823e-07,
+      "loss": 0.8009,
+      "step": 4748
+    },
+    {
+      "epoch": 0.9136206233166603,
+      "grad_norm": 1.2587373082882065,
+      "learning_rate": 3.8879643086637384e-07,
+      "loss": 0.8339,
+      "step": 4749
+    },
+    {
+      "epoch": 0.9138130050019239,
+      "grad_norm": 1.3689601787834405,
+      "learning_rate": 3.8707777444107697e-07,
+      "loss": 0.846,
+      "step": 4750
+    },
+    {
+      "epoch": 0.9140053866871873,
+      "grad_norm": 1.2350210513687088,
+      "learning_rate": 3.8536285008316854e-07,
+      "loss": 0.9074,
+      "step": 4751
+    },
+    {
+      "epoch": 0.9141977683724509,
+      "grad_norm": 1.3124118819447663,
+      "learning_rate": 3.8365165845844266e-07,
+      "loss": 0.8576,
+      "step": 4752
+    },
+    {
+      "epoch": 0.9143901500577145,
+      "grad_norm": 1.3318129896684465,
+      "learning_rate": 3.819442002312457e-07,
+      "loss": 0.7843,
+      "step": 4753
+    },
+    {
+      "epoch": 0.9145825317429781,
+      "grad_norm": 1.417116118603243,
+      "learning_rate": 3.8024047606446736e-07,
+      "loss": 0.8081,
+      "step": 4754
+    },
+    {
+      "epoch": 0.9147749134282416,
+      "grad_norm": 1.6140017741585762,
+      "learning_rate": 3.785404866195552e-07,
+      "loss": 0.7833,
+      "step": 4755
+    },
+    {
+      "epoch": 0.9149672951135052,
+      "grad_norm": 1.2430429333444069,
+      "learning_rate": 3.768442325565036e-07,
+      "loss": 0.8507,
+      "step": 4756
+    },
+    {
+      "epoch": 0.9151596767987688,
+      "grad_norm": 1.3173693960227928,
+      "learning_rate": 3.751517145338546e-07,
+      "loss": 0.8247,
+      "step": 4757
+    },
+    {
+      "epoch": 0.9153520584840323,
+      "grad_norm": 1.412757495635814,
+      "learning_rate": 3.7346293320870363e-07,
+      "loss": 0.8033,
+      "step": 4758
+    },
+    {
+      "epoch": 0.9155444401692959,
+      "grad_norm": 1.4390803530675893,
+      "learning_rate": 3.717778892366941e-07,
+      "loss": 0.8349,
+      "step": 4759
+    },
+    {
+      "epoch": 0.9157368218545594,
+      "grad_norm": 0.5928466068422023,
+      "learning_rate": 3.700965832720171e-07,
+      "loss": 0.3162,
+      "step": 4760
+    },
+    {
+      "epoch": 0.915929203539823,
+      "grad_norm": 0.620554228617895,
+      "learning_rate": 3.684190159674117e-07,
+      "loss": 0.3105,
+      "step": 4761
+    },
+    {
+      "epoch": 0.9161215852250866,
+      "grad_norm": 1.3075655907816308,
+      "learning_rate": 3.6674518797417236e-07,
+      "loss": 0.8144,
+      "step": 4762
+    },
+    {
+      "epoch": 0.9163139669103502,
+      "grad_norm": 1.5187127749717675,
+      "learning_rate": 3.6507509994213155e-07,
+      "loss": 0.8203,
+      "step": 4763
+    },
+    {
+      "epoch": 0.9165063485956138,
+      "grad_norm": 1.2120964368318872,
+      "learning_rate": 3.6340875251967946e-07,
+      "loss": 0.8393,
+      "step": 4764
+    },
+    {
+      "epoch": 0.9166987302808772,
+      "grad_norm": 1.244572081995863,
+      "learning_rate": 3.617461463537464e-07,
+      "loss": 0.8077,
+      "step": 4765
+    },
+    {
+      "epoch": 0.9168911119661408,
+      "grad_norm": 1.1370988498152128,
+      "learning_rate": 3.6008728208981157e-07,
+      "loss": 0.8335,
+      "step": 4766
+    },
+    {
+      "epoch": 0.9170834936514044,
+      "grad_norm": 1.2411952494331506,
+      "learning_rate": 3.5843216037190873e-07,
+      "loss": 0.8484,
+      "step": 4767
+    },
+    {
+      "epoch": 0.917275875336668,
+      "grad_norm": 1.3787568732611317,
+      "learning_rate": 3.5678078184260834e-07,
+      "loss": 0.7954,
+      "step": 4768
+    },
+    {
+      "epoch": 0.9174682570219315,
+      "grad_norm": 0.5731309314730335,
+      "learning_rate": 3.5513314714303524e-07,
+      "loss": 0.3255,
+      "step": 4769
+    },
+    {
+      "epoch": 0.917660638707195,
+      "grad_norm": 1.577034884520202,
+      "learning_rate": 3.5348925691285675e-07,
+      "loss": 0.8405,
+      "step": 4770
+    },
+    {
+      "epoch": 0.9178530203924586,
+      "grad_norm": 1.325636304801035,
+      "learning_rate": 3.518491117902878e-07,
+      "loss": 0.8467,
+      "step": 4771
+    },
+    {
+      "epoch": 0.9180454020777222,
+      "grad_norm": 1.2879175901804596,
+      "learning_rate": 3.502127124120891e-07,
+      "loss": 0.8152,
+      "step": 4772
+    },
+    {
+      "epoch": 0.9182377837629858,
+      "grad_norm": 1.1531487979464359,
+      "learning_rate": 3.48580059413568e-07,
+      "loss": 0.8647,
+      "step": 4773
+    },
+    {
+      "epoch": 0.9184301654482493,
+      "grad_norm": 1.6678708521822971,
+      "learning_rate": 3.4695115342857524e-07,
+      "loss": 0.8264,
+      "step": 4774
+    },
+    {
+      "epoch": 0.9186225471335129,
+      "grad_norm": 2.0925538801617387,
+      "learning_rate": 3.4532599508950826e-07,
+      "loss": 0.8195,
+      "step": 4775
+    },
+    {
+      "epoch": 0.9188149288187765,
+      "grad_norm": 1.602056822998003,
+      "learning_rate": 3.437045850273113e-07,
+      "loss": 0.8026,
+      "step": 4776
+    },
+    {
+      "epoch": 0.91900731050404,
+      "grad_norm": 1.2654954291531142,
+      "learning_rate": 3.420869238714708e-07,
+      "loss": 0.8451,
+      "step": 4777
+    },
+    {
+      "epoch": 0.9191996921893035,
+      "grad_norm": 2.1013533828472237,
+      "learning_rate": 3.404730122500155e-07,
+      "loss": 0.8075,
+      "step": 4778
+    },
+    {
+      "epoch": 0.9193920738745671,
+      "grad_norm": 1.268616310287008,
+      "learning_rate": 3.3886285078952753e-07,
+      "loss": 0.7918,
+      "step": 4779
+    },
+    {
+      "epoch": 0.9195844555598307,
+      "grad_norm": 0.6359276916477452,
+      "learning_rate": 3.3725644011512125e-07,
+      "loss": 0.3308,
+      "step": 4780
+    },
+    {
+      "epoch": 0.9197768372450943,
+      "grad_norm": 1.5513457394050192,
+      "learning_rate": 3.356537808504634e-07,
+      "loss": 0.8335,
+      "step": 4781
+    },
+    {
+      "epoch": 0.9199692189303579,
+      "grad_norm": 1.5963041364573547,
+      "learning_rate": 3.3405487361776177e-07,
+      "loss": 0.8429,
+      "step": 4782
+    },
+    {
+      "epoch": 0.9201616006156214,
+      "grad_norm": 1.2681500446203255,
+      "learning_rate": 3.3245971903776654e-07,
+      "loss": 0.7944,
+      "step": 4783
+    },
+    {
+      "epoch": 0.9203539823008849,
+      "grad_norm": 1.244601718994994,
+      "learning_rate": 3.308683177297711e-07,
+      "loss": 0.8849,
+      "step": 4784
+    },
+    {
+      "epoch": 0.9205463639861485,
+      "grad_norm": 1.222367044024693,
+      "learning_rate": 3.292806703116125e-07,
+      "loss": 0.8206,
+      "step": 4785
+    },
+    {
+      "epoch": 0.9207387456714121,
+      "grad_norm": 1.456288164038098,
+      "learning_rate": 3.2769677739966975e-07,
+      "loss": 0.8057,
+      "step": 4786
+    },
+    {
+      "epoch": 0.9209311273566756,
+      "grad_norm": 1.3594431876859514,
+      "learning_rate": 3.2611663960886665e-07,
+      "loss": 0.8165,
+      "step": 4787
+    },
+    {
+      "epoch": 0.9211235090419392,
+      "grad_norm": 1.2756375546511616,
+      "learning_rate": 3.245402575526646e-07,
+      "loss": 0.8313,
+      "step": 4788
+    },
+    {
+      "epoch": 0.9213158907272028,
+      "grad_norm": 1.3032551880745615,
+      "learning_rate": 3.2296763184306965e-07,
+      "loss": 0.8462,
+      "step": 4789
+    },
+    {
+      "epoch": 0.9215082724124664,
+      "grad_norm": 1.3173403176414027,
+      "learning_rate": 3.2139876309063233e-07,
+      "loss": 0.7761,
+      "step": 4790
+    },
+    {
+      "epoch": 0.9217006540977299,
+      "grad_norm": 0.5870754964935675,
+      "learning_rate": 3.198336519044376e-07,
+      "loss": 0.3344,
+      "step": 4791
+    },
+    {
+      "epoch": 0.9218930357829934,
+      "grad_norm": 1.193034471232867,
+      "learning_rate": 3.182722988921161e-07,
+      "loss": 0.7539,
+      "step": 4792
+    },
+    {
+      "epoch": 0.922085417468257,
+      "grad_norm": 1.3683698591437328,
+      "learning_rate": 3.167147046598418e-07,
+      "loss": 0.8326,
+      "step": 4793
+    },
+    {
+      "epoch": 0.9222777991535206,
+      "grad_norm": 1.4935734330331767,
+      "learning_rate": 3.151608698123232e-07,
+      "loss": 0.8687,
+      "step": 4794
+    },
+    {
+      "epoch": 0.9224701808387842,
+      "grad_norm": 1.4552384990671958,
+      "learning_rate": 3.1361079495281443e-07,
+      "loss": 0.8151,
+      "step": 4795
+    },
+    {
+      "epoch": 0.9226625625240477,
+      "grad_norm": 1.6933194117198507,
+      "learning_rate": 3.1206448068310635e-07,
+      "loss": 0.7806,
+      "step": 4796
+    },
+    {
+      "epoch": 0.9228549442093112,
+      "grad_norm": 1.279316969304758,
+      "learning_rate": 3.1052192760353316e-07,
+      "loss": 0.8836,
+      "step": 4797
+    },
+    {
+      "epoch": 0.9230473258945748,
+      "grad_norm": 1.2524474916126713,
+      "learning_rate": 3.0898313631296586e-07,
+      "loss": 0.8504,
+      "step": 4798
+    },
+    {
+      "epoch": 0.9232397075798384,
+      "grad_norm": 1.2127931247640025,
+      "learning_rate": 3.0744810740881646e-07,
+      "loss": 0.7856,
+      "step": 4799
+    },
+    {
+      "epoch": 0.923432089265102,
+      "grad_norm": 1.325274218663552,
+      "learning_rate": 3.0591684148703617e-07,
+      "loss": 0.8459,
+      "step": 4800
+    },
+    {
+      "epoch": 0.9236244709503655,
+      "grad_norm": 1.563503781205002,
+      "learning_rate": 3.043893391421149e-07,
+      "loss": 0.7833,
+      "step": 4801
+    },
+    {
+      "epoch": 0.9238168526356291,
+      "grad_norm": 1.5065441291452262,
+      "learning_rate": 3.0286560096708275e-07,
+      "loss": 0.8446,
+      "step": 4802
+    },
+    {
+      "epoch": 0.9240092343208927,
+      "grad_norm": 1.5817997110948905,
+      "learning_rate": 3.013456275535054e-07,
+      "loss": 0.8311,
+      "step": 4803
+    },
+    {
+      "epoch": 0.9242016160061562,
+      "grad_norm": 1.4963654861660423,
+      "learning_rate": 2.998294194914897e-07,
+      "loss": 0.7829,
+      "step": 4804
+    },
+    {
+      "epoch": 0.9243939976914197,
+      "grad_norm": 1.362668225098311,
+      "learning_rate": 2.983169773696815e-07,
+      "loss": 0.8026,
+      "step": 4805
+    },
+    {
+      "epoch": 0.9245863793766833,
+      "grad_norm": 1.2036544432473402,
+      "learning_rate": 2.968083017752599e-07,
+      "loss": 0.8959,
+      "step": 4806
+    },
+    {
+      "epoch": 0.9247787610619469,
+      "grad_norm": 1.2010534208028003,
+      "learning_rate": 2.953033932939464e-07,
+      "loss": 0.7687,
+      "step": 4807
+    },
+    {
+      "epoch": 0.9249711427472105,
+      "grad_norm": 1.2380984136466804,
+      "learning_rate": 2.938022525099982e-07,
+      "loss": 0.8568,
+      "step": 4808
+    },
+    {
+      "epoch": 0.9251635244324741,
+      "grad_norm": 1.3674937658386033,
+      "learning_rate": 2.9230488000621003e-07,
+      "loss": 0.8694,
+      "step": 4809
+    },
+    {
+      "epoch": 0.9253559061177375,
+      "grad_norm": 1.3134789341422255,
+      "learning_rate": 2.908112763639137e-07,
+      "loss": 0.8282,
+      "step": 4810
+    },
+    {
+      "epoch": 0.9255482878030011,
+      "grad_norm": 0.612202428623619,
+      "learning_rate": 2.8932144216297643e-07,
+      "loss": 0.3022,
+      "step": 4811
+    },
+    {
+      "epoch": 0.9257406694882647,
+      "grad_norm": 1.4458795248037335,
+      "learning_rate": 2.878353779818044e-07,
+      "loss": 0.8257,
+      "step": 4812
+    },
+    {
+      "epoch": 0.9259330511735283,
+      "grad_norm": 1.4672775583375113,
+      "learning_rate": 2.863530843973372e-07,
+      "loss": 0.8611,
+      "step": 4813
+    },
+    {
+      "epoch": 0.9261254328587919,
+      "grad_norm": 1.297952849564289,
+      "learning_rate": 2.848745619850546e-07,
+      "loss": 0.7882,
+      "step": 4814
+    },
+    {
+      "epoch": 0.9263178145440554,
+      "grad_norm": 1.322373823402721,
+      "learning_rate": 2.833998113189662e-07,
+      "loss": 0.8475,
+      "step": 4815
+    },
+    {
+      "epoch": 0.926510196229319,
+      "grad_norm": 1.4965668166321773,
+      "learning_rate": 2.8192883297162634e-07,
+      "loss": 0.8327,
+      "step": 4816
+    },
+    {
+      "epoch": 0.9267025779145825,
+      "grad_norm": 1.3624851214568745,
+      "learning_rate": 2.804616275141148e-07,
+      "loss": 0.8342,
+      "step": 4817
+    },
+    {
+      "epoch": 0.9268949595998461,
+      "grad_norm": 1.261799499246355,
+      "learning_rate": 2.7899819551605256e-07,
+      "loss": 0.8292,
+      "step": 4818
+    },
+    {
+      "epoch": 0.9270873412851096,
+      "grad_norm": 1.2935500926604324,
+      "learning_rate": 2.7753853754559634e-07,
+      "loss": 0.8632,
+      "step": 4819
+    },
+    {
+      "epoch": 0.9272797229703732,
+      "grad_norm": 1.2767728920845636,
+      "learning_rate": 2.760826541694328e-07,
+      "loss": 0.8489,
+      "step": 4820
+    },
+    {
+      "epoch": 0.9274721046556368,
+      "grad_norm": 1.2797763418527648,
+      "learning_rate": 2.746305459527876e-07,
+      "loss": 0.8012,
+      "step": 4821
+    },
+    {
+      "epoch": 0.9276644863409004,
+      "grad_norm": 0.6140926153452283,
+      "learning_rate": 2.7318221345941865e-07,
+      "loss": 0.3502,
+      "step": 4822
+    },
+    {
+      "epoch": 0.927856868026164,
+      "grad_norm": 1.5035851833890486,
+      "learning_rate": 2.717376572516184e-07,
+      "loss": 0.7972,
+      "step": 4823
+    },
+    {
+      "epoch": 0.9280492497114274,
+      "grad_norm": 2.1240690459146245,
+      "learning_rate": 2.7029687789021377e-07,
+      "loss": 0.7762,
+      "step": 4824
+    },
+    {
+      "epoch": 0.928241631396691,
+      "grad_norm": 1.3175983822039632,
+      "learning_rate": 2.688598759345651e-07,
+      "loss": 0.8422,
+      "step": 4825
+    },
+    {
+      "epoch": 0.9284340130819546,
+      "grad_norm": 1.8378745124899407,
+      "learning_rate": 2.67426651942565e-07,
+      "loss": 0.8262,
+      "step": 4826
+    },
+    {
+      "epoch": 0.9286263947672182,
+      "grad_norm": 1.238803590863023,
+      "learning_rate": 2.659972064706406e-07,
+      "loss": 0.8415,
+      "step": 4827
+    },
+    {
+      "epoch": 0.9288187764524817,
+      "grad_norm": 1.3398153574400027,
+      "learning_rate": 2.645715400737536e-07,
+      "loss": 0.8139,
+      "step": 4828
+    },
+    {
+      "epoch": 0.9290111581377453,
+      "grad_norm": 1.2116289935789148,
+      "learning_rate": 2.631496533053934e-07,
+      "loss": 0.8458,
+      "step": 4829
+    },
+    {
+      "epoch": 0.9292035398230089,
+      "grad_norm": 0.5849799278920844,
+      "learning_rate": 2.6173154671758847e-07,
+      "loss": 0.3207,
+      "step": 4830
+    },
+    {
+      "epoch": 0.9293959215082724,
+      "grad_norm": 0.6311517423960075,
+      "learning_rate": 2.603172208608962e-07,
+      "loss": 0.325,
+      "step": 4831
+    },
+    {
+      "epoch": 0.929588303193536,
+      "grad_norm": 1.4710384292676824,
+      "learning_rate": 2.589066762844039e-07,
+      "loss": 0.8447,
+      "step": 4832
+    },
+    {
+      "epoch": 0.9297806848787995,
+      "grad_norm": 1.4094959487463494,
+      "learning_rate": 2.57499913535737e-07,
+      "loss": 0.8662,
+      "step": 4833
+    },
+    {
+      "epoch": 0.9299730665640631,
+      "grad_norm": 1.1869365203767237,
+      "learning_rate": 2.5609693316104745e-07,
+      "loss": 0.8027,
+      "step": 4834
+    },
+    {
+      "epoch": 0.9301654482493267,
+      "grad_norm": 1.2776315437039965,
+      "learning_rate": 2.5469773570502063e-07,
+      "loss": 0.8303,
+      "step": 4835
+    },
+    {
+      "epoch": 0.9303578299345903,
+      "grad_norm": 0.6062901375660581,
+      "learning_rate": 2.5330232171087433e-07,
+      "loss": 0.3708,
+      "step": 4836
+    },
+    {
+      "epoch": 0.9305502116198537,
+      "grad_norm": 1.3181696241071523,
+      "learning_rate": 2.51910691720354e-07,
+      "loss": 0.8145,
+      "step": 4837
+    },
+    {
+      "epoch": 0.9307425933051173,
+      "grad_norm": 1.5700972427012068,
+      "learning_rate": 2.5052284627374077e-07,
+      "loss": 0.8696,
+      "step": 4838
+    },
+    {
+      "epoch": 0.9309349749903809,
+      "grad_norm": 1.382567852799985,
+      "learning_rate": 2.491387859098426e-07,
+      "loss": 0.8792,
+      "step": 4839
+    },
+    {
+      "epoch": 0.9311273566756445,
+      "grad_norm": 1.148878382096043,
+      "learning_rate": 2.477585111659997e-07,
+      "loss": 0.7708,
+      "step": 4840
+    },
+    {
+      "epoch": 0.9313197383609081,
+      "grad_norm": 1.3286313867894308,
+      "learning_rate": 2.463820225780811e-07,
+      "loss": 0.8232,
+      "step": 4841
+    },
+    {
+      "epoch": 0.9315121200461716,
+      "grad_norm": 3.247190174994962,
+      "learning_rate": 2.4500932068049046e-07,
+      "loss": 0.8337,
+      "step": 4842
+    },
+    {
+      "epoch": 0.9317045017314352,
+      "grad_norm": 1.3000028729992728,
+      "learning_rate": 2.4364040600615477e-07,
+      "loss": 0.7733,
+      "step": 4843
+    },
+    {
+      "epoch": 0.9318968834166987,
+      "grad_norm": 1.2403721534999106,
+      "learning_rate": 2.422752790865346e-07,
+      "loss": 0.7927,
+      "step": 4844
+    },
+    {
+      "epoch": 0.9320892651019623,
+      "grad_norm": 1.2786641510190317,
+      "learning_rate": 2.409139404516203e-07,
+      "loss": 0.8229,
+      "step": 4845
+    },
+    {
+      "epoch": 0.9322816467872258,
+      "grad_norm": 1.3549967736610087,
+      "learning_rate": 2.3955639062992696e-07,
+      "loss": 0.7784,
+      "step": 4846
+    },
+    {
+      "epoch": 0.9324740284724894,
+      "grad_norm": 1.4220396578551266,
+      "learning_rate": 2.3820263014850741e-07,
+      "loss": 0.8484,
+      "step": 4847
+    },
+    {
+      "epoch": 0.932666410157753,
+      "grad_norm": 1.351298502746069,
+      "learning_rate": 2.3685265953293345e-07,
+      "loss": 0.8244,
+      "step": 4848
+    },
+    {
+      "epoch": 0.9328587918430166,
+      "grad_norm": 1.371815514883031,
+      "learning_rate": 2.3550647930731362e-07,
+      "loss": 0.7892,
+      "step": 4849
+    },
+    {
+      "epoch": 0.9330511735282802,
+      "grad_norm": 1.5220774681581026,
+      "learning_rate": 2.3416408999427876e-07,
+      "loss": 0.8328,
+      "step": 4850
+    },
+    {
+      "epoch": 0.9332435552135436,
+      "grad_norm": 1.3060120124348915,
+      "learning_rate": 2.3282549211499307e-07,
+      "loss": 0.8674,
+      "step": 4851
+    },
+    {
+      "epoch": 0.9334359368988072,
+      "grad_norm": 1.2329794338327233,
+      "learning_rate": 2.3149068618914417e-07,
+      "loss": 0.8453,
+      "step": 4852
+    },
+    {
+      "epoch": 0.9336283185840708,
+      "grad_norm": 2.3219699451090694,
+      "learning_rate": 2.3015967273494867e-07,
+      "loss": 0.8154,
+      "step": 4853
+    },
+    {
+      "epoch": 0.9338207002693344,
+      "grad_norm": 0.6443822585020794,
+      "learning_rate": 2.2883245226915652e-07,
+      "loss": 0.3403,
+      "step": 4854
+    },
+    {
+      "epoch": 0.9340130819545979,
+      "grad_norm": 1.229531322092596,
+      "learning_rate": 2.2750902530703667e-07,
+      "loss": 0.8147,
+      "step": 4855
+    },
+    {
+      "epoch": 0.9342054636398615,
+      "grad_norm": 1.330172941918002,
+      "learning_rate": 2.2618939236238924e-07,
+      "loss": 0.7803,
+      "step": 4856
+    },
+    {
+      "epoch": 0.934397845325125,
+      "grad_norm": 1.0709486821041054,
+      "learning_rate": 2.2487355394754328e-07,
+      "loss": 0.8527,
+      "step": 4857
+    },
+    {
+      "epoch": 0.9345902270103886,
+      "grad_norm": 1.2302741885829027,
+      "learning_rate": 2.2356151057334908e-07,
+      "loss": 0.8164,
+      "step": 4858
+    },
+    {
+      "epoch": 0.9347826086956522,
+      "grad_norm": 0.622271939290009,
+      "learning_rate": 2.2225326274919135e-07,
+      "loss": 0.3257,
+      "step": 4859
+    },
+    {
+      "epoch": 0.9349749903809157,
+      "grad_norm": 1.1071930141855082,
+      "learning_rate": 2.209488109829727e-07,
+      "loss": 0.8003,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9351673720661793,
+      "grad_norm": 1.4926390116601829,
+      "learning_rate": 2.196481557811303e-07,
+      "loss": 0.7933,
+      "step": 4861
+    },
+    {
+      "epoch": 0.9353597537514429,
+      "grad_norm": 1.3267043978906043,
+      "learning_rate": 2.1835129764861907e-07,
+      "loss": 0.8545,
+      "step": 4862
+    },
+    {
+      "epoch": 0.9355521354367065,
+      "grad_norm": 1.0684106589562508,
+      "learning_rate": 2.1705823708892737e-07,
+      "loss": 0.8223,
+      "step": 4863
+    },
+    {
+      "epoch": 0.93574451712197,
+      "grad_norm": 1.432727178940544,
+      "learning_rate": 2.1576897460406477e-07,
+      "loss": 0.794,
+      "step": 4864
+    },
+    {
+      "epoch": 0.9359368988072335,
+      "grad_norm": 1.8235594682427354,
+      "learning_rate": 2.144835106945664e-07,
+      "loss": 0.7918,
+      "step": 4865
+    },
+    {
+      "epoch": 0.9361292804924971,
+      "grad_norm": 1.4678706172803744,
+      "learning_rate": 2.1320184585949532e-07,
+      "loss": 0.877,
+      "step": 4866
+    },
+    {
+      "epoch": 0.9363216621777607,
+      "grad_norm": 1.2038442077571292,
+      "learning_rate": 2.119239805964357e-07,
+      "loss": 0.8417,
+      "step": 4867
+    },
+    {
+      "epoch": 0.9365140438630243,
+      "grad_norm": 0.6129875251375253,
+      "learning_rate": 2.106499154015018e-07,
+      "loss": 0.3348,
+      "step": 4868
+    },
+    {
+      "epoch": 0.9367064255482878,
+      "grad_norm": 1.2328928603103377,
+      "learning_rate": 2.0937965076932576e-07,
+      "loss": 0.825,
+      "step": 4869
+    },
+    {
+      "epoch": 0.9368988072335513,
+      "grad_norm": 1.496478989879084,
+      "learning_rate": 2.0811318719307194e-07,
+      "loss": 0.8185,
+      "step": 4870
+    },
+    {
+      "epoch": 0.9370911889188149,
+      "grad_norm": 1.2225603395225908,
+      "learning_rate": 2.0685052516442373e-07,
+      "loss": 0.7914,
+      "step": 4871
+    },
+    {
+      "epoch": 0.9372835706040785,
+      "grad_norm": 2.1106777393145815,
+      "learning_rate": 2.0559166517358787e-07,
+      "loss": 0.7892,
+      "step": 4872
+    },
+    {
+      "epoch": 0.9374759522893421,
+      "grad_norm": 1.2619711537194165,
+      "learning_rate": 2.0433660770930009e-07,
+      "loss": 0.8263,
+      "step": 4873
+    },
+    {
+      "epoch": 0.9376683339746056,
+      "grad_norm": 2.8350043319176677,
+      "learning_rate": 2.0308535325881616e-07,
+      "loss": 0.8544,
+      "step": 4874
+    },
+    {
+      "epoch": 0.9378607156598692,
+      "grad_norm": 1.5670130350806857,
+      "learning_rate": 2.0183790230791532e-07,
+      "loss": 0.8634,
+      "step": 4875
+    },
+    {
+      "epoch": 0.9380530973451328,
+      "grad_norm": 2.1967228754153862,
+      "learning_rate": 2.0059425534090128e-07,
+      "loss": 0.7992,
+      "step": 4876
+    },
+    {
+      "epoch": 0.9382454790303963,
+      "grad_norm": 1.2770743777388998,
+      "learning_rate": 1.9935441284059998e-07,
+      "loss": 0.8858,
+      "step": 4877
+    },
+    {
+      "epoch": 0.9384378607156598,
+      "grad_norm": 1.5932029760508069,
+      "learning_rate": 1.981183752883631e-07,
+      "loss": 0.8351,
+      "step": 4878
+    },
+    {
+      "epoch": 0.9386302424009234,
+      "grad_norm": 1.4633281008675376,
+      "learning_rate": 1.9688614316406006e-07,
+      "loss": 0.8203,
+      "step": 4879
+    },
+    {
+      "epoch": 0.938822624086187,
+      "grad_norm": 1.3666195230929081,
+      "learning_rate": 1.9565771694608937e-07,
+      "loss": 0.8246,
+      "step": 4880
+    },
+    {
+      "epoch": 0.9390150057714506,
+      "grad_norm": 1.2860280271736002,
+      "learning_rate": 1.9443309711136393e-07,
+      "loss": 0.7977,
+      "step": 4881
+    },
+    {
+      "epoch": 0.9392073874567142,
+      "grad_norm": 1.3503560419269887,
+      "learning_rate": 1.9321228413532788e-07,
+      "loss": 0.7568,
+      "step": 4882
+    },
+    {
+      "epoch": 0.9393997691419776,
+      "grad_norm": 1.441486543478805,
+      "learning_rate": 1.9199527849194098e-07,
+      "loss": 0.8095,
+      "step": 4883
+    },
+    {
+      "epoch": 0.9395921508272412,
+      "grad_norm": 0.616961363417642,
+      "learning_rate": 1.907820806536842e-07,
+      "loss": 0.3358,
+      "step": 4884
+    },
+    {
+      "epoch": 0.9397845325125048,
+      "grad_norm": 1.2732370723077422,
+      "learning_rate": 1.895726910915663e-07,
+      "loss": 0.7867,
+      "step": 4885
+    },
+    {
+      "epoch": 0.9399769141977684,
+      "grad_norm": 1.4423643087926128,
+      "learning_rate": 1.883671102751128e-07,
+      "loss": 0.8321,
+      "step": 4886
+    },
+    {
+      "epoch": 0.9401692958830319,
+      "grad_norm": 0.6111020391467837,
+      "learning_rate": 1.8716533867237153e-07,
+      "loss": 0.3336,
+      "step": 4887
+    },
+    {
+      "epoch": 0.9403616775682955,
+      "grad_norm": 1.2775828730732353,
+      "learning_rate": 1.859673767499115e-07,
+      "loss": 0.875,
+      "step": 4888
+    },
+    {
+      "epoch": 0.9405540592535591,
+      "grad_norm": 0.6046906021772518,
+      "learning_rate": 1.847732249728218e-07,
+      "loss": 0.3073,
+      "step": 4889
+    },
+    {
+      "epoch": 0.9407464409388226,
+      "grad_norm": 4.4722805311071365,
+      "learning_rate": 1.83582883804716e-07,
+      "loss": 0.8541,
+      "step": 4890
+    },
+    {
+      "epoch": 0.9409388226240862,
+      "grad_norm": 1.1608832246329572,
+      "learning_rate": 1.8239635370772223e-07,
+      "loss": 0.7914,
+      "step": 4891
+    },
+    {
+      "epoch": 0.9411312043093497,
+      "grad_norm": 0.6081541475382002,
+      "learning_rate": 1.8121363514249534e-07,
+      "loss": 0.3272,
+      "step": 4892
+    },
+    {
+      "epoch": 0.9413235859946133,
+      "grad_norm": 1.4119295192241061,
+      "learning_rate": 1.8003472856820469e-07,
+      "loss": 0.8003,
+      "step": 4893
+    },
+    {
+      "epoch": 0.9415159676798769,
+      "grad_norm": 1.3316736625831127,
+      "learning_rate": 1.7885963444254528e-07,
+      "loss": 0.8193,
+      "step": 4894
+    },
+    {
+      "epoch": 0.9417083493651405,
+      "grad_norm": 1.1675107337521626,
+      "learning_rate": 1.7768835322172552e-07,
+      "loss": 0.821,
+      "step": 4895
+    },
+    {
+      "epoch": 0.941900731050404,
+      "grad_norm": 1.2441509194555376,
+      "learning_rate": 1.7652088536048052e-07,
+      "loss": 0.8641,
+      "step": 4896
+    },
+    {
+      "epoch": 0.9420931127356675,
+      "grad_norm": 1.337938664270688,
+      "learning_rate": 1.7535723131206106e-07,
+      "loss": 0.8442,
+      "step": 4897
+    },
+    {
+      "epoch": 0.9422854944209311,
+      "grad_norm": 1.2848385816400754,
+      "learning_rate": 1.7419739152823468e-07,
+      "loss": 0.8767,
+      "step": 4898
+    },
+    {
+      "epoch": 0.9424778761061947,
+      "grad_norm": 1.2068873835765528,
+      "learning_rate": 1.7304136645929448e-07,
+      "loss": 0.8247,
+      "step": 4899
+    },
+    {
+      "epoch": 0.9426702577914583,
+      "grad_norm": 1.3334838578931199,
+      "learning_rate": 1.7188915655404814e-07,
+      "loss": 0.8295,
+      "step": 4900
+    },
+    {
+      "epoch": 0.9428626394767218,
+      "grad_norm": 1.4733446933139516,
+      "learning_rate": 1.707407622598223e-07,
+      "loss": 0.8041,
+      "step": 4901
+    },
+    {
+      "epoch": 0.9430550211619854,
+      "grad_norm": 1.3503886412485264,
+      "learning_rate": 1.695961840224636e-07,
+      "loss": 0.8364,
+      "step": 4902
+    },
+    {
+      "epoch": 0.943247402847249,
+      "grad_norm": 1.403997989727608,
+      "learning_rate": 1.6845542228633772e-07,
+      "loss": 0.7995,
+      "step": 4903
+    },
+    {
+      "epoch": 0.9434397845325125,
+      "grad_norm": 1.513992354703673,
+      "learning_rate": 1.6731847749432705e-07,
+      "loss": 0.8328,
+      "step": 4904
+    },
+    {
+      "epoch": 0.943632166217776,
+      "grad_norm": 1.1907427397911567,
+      "learning_rate": 1.6618535008783075e-07,
+      "loss": 0.816,
+      "step": 4905
+    },
+    {
+      "epoch": 0.9438245479030396,
+      "grad_norm": 1.1492894686790929,
+      "learning_rate": 1.6505604050677249e-07,
+      "loss": 0.8592,
+      "step": 4906
+    },
+    {
+      "epoch": 0.9440169295883032,
+      "grad_norm": 1.3777424602600437,
+      "learning_rate": 1.6393054918958373e-07,
+      "loss": 0.8268,
+      "step": 4907
+    },
+    {
+      "epoch": 0.9442093112735668,
+      "grad_norm": 1.0795680244834758,
+      "learning_rate": 1.6280887657322276e-07,
+      "loss": 0.7485,
+      "step": 4908
+    },
+    {
+      "epoch": 0.9444016929588304,
+      "grad_norm": 1.6480795048459465,
+      "learning_rate": 1.616910230931612e-07,
+      "loss": 0.829,
+      "step": 4909
+    },
+    {
+      "epoch": 0.9445940746440938,
+      "grad_norm": 1.3221291497701195,
+      "learning_rate": 1.6057698918338526e-07,
+      "loss": 0.8544,
+      "step": 4910
+    },
+    {
+      "epoch": 0.9447864563293574,
+      "grad_norm": 1.4281501457799228,
+      "learning_rate": 1.5946677527640563e-07,
+      "loss": 0.8308,
+      "step": 4911
+    },
+    {
+      "epoch": 0.944978838014621,
+      "grad_norm": 1.3139254218138066,
+      "learning_rate": 1.5836038180324198e-07,
+      "loss": 0.8075,
+      "step": 4912
+    },
+    {
+      "epoch": 0.9451712196998846,
+      "grad_norm": 0.578941992212993,
+      "learning_rate": 1.5725780919343624e-07,
+      "loss": 0.3528,
+      "step": 4913
+    },
+    {
+      "epoch": 0.9453636013851482,
+      "grad_norm": 1.193700094131032,
+      "learning_rate": 1.561590578750438e-07,
+      "loss": 0.9023,
+      "step": 4914
+    },
+    {
+      "epoch": 0.9455559830704117,
+      "grad_norm": 0.5877384105408445,
+      "learning_rate": 1.55064128274639e-07,
+      "loss": 0.3516,
+      "step": 4915
+    },
+    {
+      "epoch": 0.9457483647556753,
+      "grad_norm": 1.2750689483687,
+      "learning_rate": 1.5397302081731069e-07,
+      "loss": 0.8422,
+      "step": 4916
+    },
+    {
+      "epoch": 0.9459407464409388,
+      "grad_norm": 1.1615381077772147,
+      "learning_rate": 1.5288573592666445e-07,
+      "loss": 0.8324,
+      "step": 4917
+    },
+    {
+      "epoch": 0.9461331281262024,
+      "grad_norm": 1.2312142158635497,
+      "learning_rate": 1.518022740248215e-07,
+      "loss": 0.8094,
+      "step": 4918
+    },
+    {
+      "epoch": 0.9463255098114659,
+      "grad_norm": 1.3772188173831792,
+      "learning_rate": 1.5072263553241872e-07,
+      "loss": 0.7967,
+      "step": 4919
+    },
+    {
+      "epoch": 0.9465178914967295,
+      "grad_norm": 1.426662954040786,
+      "learning_rate": 1.4964682086861082e-07,
+      "loss": 0.8585,
+      "step": 4920
+    },
+    {
+      "epoch": 0.9467102731819931,
+      "grad_norm": 1.2990992847334402,
+      "learning_rate": 1.4857483045106258e-07,
+      "loss": 0.8815,
+      "step": 4921
+    },
+    {
+      "epoch": 0.9469026548672567,
+      "grad_norm": 1.2780445831737688,
+      "learning_rate": 1.475066646959611e-07,
+      "loss": 0.9059,
+      "step": 4922
+    },
+    {
+      "epoch": 0.9470950365525203,
+      "grad_norm": 1.5161819938245076,
+      "learning_rate": 1.4644232401800352e-07,
+      "loss": 0.8154,
+      "step": 4923
+    },
+    {
+      "epoch": 0.9472874182377837,
+      "grad_norm": 1.2362981099549788,
+      "learning_rate": 1.4538180883040264e-07,
+      "loss": 0.8218,
+      "step": 4924
+    },
+    {
+      "epoch": 0.9474797999230473,
+      "grad_norm": 1.2433655632684648,
+      "learning_rate": 1.4432511954488915e-07,
+      "loss": 0.8136,
+      "step": 4925
+    },
+    {
+      "epoch": 0.9476721816083109,
+      "grad_norm": 1.5910624193206875,
+      "learning_rate": 1.4327225657170485e-07,
+      "loss": 0.915,
+      "step": 4926
+    },
+    {
+      "epoch": 0.9478645632935745,
+      "grad_norm": 1.4061107050365222,
+      "learning_rate": 1.4222322031960723e-07,
+      "loss": 0.8145,
+      "step": 4927
+    },
+    {
+      "epoch": 0.948056944978838,
+      "grad_norm": 1.53810150348699,
+      "learning_rate": 1.411780111958694e-07,
+      "loss": 0.7735,
+      "step": 4928
+    },
+    {
+      "epoch": 0.9482493266641016,
+      "grad_norm": 1.5895230434335996,
+      "learning_rate": 1.4013662960627562e-07,
+      "loss": 0.862,
+      "step": 4929
+    },
+    {
+      "epoch": 0.9484417083493651,
+      "grad_norm": 1.576008081216639,
+      "learning_rate": 1.3909907595512806e-07,
+      "loss": 0.8135,
+      "step": 4930
+    },
+    {
+      "epoch": 0.9486340900346287,
+      "grad_norm": 1.3083109051520088,
+      "learning_rate": 1.3806535064524006e-07,
+      "loss": 0.82,
+      "step": 4931
+    },
+    {
+      "epoch": 0.9488264717198923,
+      "grad_norm": 1.419558054417502,
+      "learning_rate": 1.3703545407793951e-07,
+      "loss": 0.8862,
+      "step": 4932
+    },
+    {
+      "epoch": 0.9490188534051558,
+      "grad_norm": 1.3266202504850855,
+      "learning_rate": 1.360093866530665e-07,
+      "loss": 0.8983,
+      "step": 4933
+    },
+    {
+      "epoch": 0.9492112350904194,
+      "grad_norm": 1.8378837226406703,
+      "learning_rate": 1.34987148768978e-07,
+      "loss": 0.8163,
+      "step": 4934
+    },
+    {
+      "epoch": 0.949403616775683,
+      "grad_norm": 1.2691836105423444,
+      "learning_rate": 1.3396874082253986e-07,
+      "loss": 0.846,
+      "step": 4935
+    },
+    {
+      "epoch": 0.9495959984609466,
+      "grad_norm": 1.260003612976847,
+      "learning_rate": 1.3295416320913357e-07,
+      "loss": 0.8253,
+      "step": 4936
+    },
+    {
+      "epoch": 0.94978838014621,
+      "grad_norm": 1.3576530780541731,
+      "learning_rate": 1.3194341632265518e-07,
+      "loss": 0.8583,
+      "step": 4937
+    },
+    {
+      "epoch": 0.9499807618314736,
+      "grad_norm": 1.1823718097325973,
+      "learning_rate": 1.3093650055550855e-07,
+      "loss": 0.8805,
+      "step": 4938
+    },
+    {
+      "epoch": 0.9501731435167372,
+      "grad_norm": 1.2994292512920629,
+      "learning_rate": 1.2993341629861432e-07,
+      "loss": 0.8227,
+      "step": 4939
+    },
+    {
+      "epoch": 0.9503655252020008,
+      "grad_norm": 1.4501750640884528,
+      "learning_rate": 1.2893416394140323e-07,
+      "loss": 0.8054,
+      "step": 4940
+    },
+    {
+      "epoch": 0.9505579068872644,
+      "grad_norm": 1.4047535944933391,
+      "learning_rate": 1.279387438718216e-07,
+      "loss": 0.8515,
+      "step": 4941
+    },
+    {
+      "epoch": 0.9507502885725279,
+      "grad_norm": 1.4720626973542392,
+      "learning_rate": 1.269471564763247e-07,
+      "loss": 0.8229,
+      "step": 4942
+    },
+    {
+      "epoch": 0.9509426702577914,
+      "grad_norm": 1.3000409787922307,
+      "learning_rate": 1.2595940213988024e-07,
+      "loss": 0.8098,
+      "step": 4943
+    },
+    {
+      "epoch": 0.951135051943055,
+      "grad_norm": 1.1684226553096044,
+      "learning_rate": 1.2497548124597026e-07,
+      "loss": 0.8136,
+      "step": 4944
+    },
+    {
+      "epoch": 0.9513274336283186,
+      "grad_norm": 1.129510843044299,
+      "learning_rate": 1.2399539417658368e-07,
+      "loss": 0.7967,
+      "step": 4945
+    },
+    {
+      "epoch": 0.9515198153135821,
+      "grad_norm": 1.7260062230021354,
+      "learning_rate": 1.2301914131222726e-07,
+      "loss": 0.8218,
+      "step": 4946
+    },
+    {
+      "epoch": 0.9517121969988457,
+      "grad_norm": 1.200111489255699,
+      "learning_rate": 1.2204672303191335e-07,
+      "loss": 0.7627,
+      "step": 4947
+    },
+    {
+      "epoch": 0.9519045786841093,
+      "grad_norm": 2.095917137404938,
+      "learning_rate": 1.2107813971317106e-07,
+      "loss": 0.7735,
+      "step": 4948
+    },
+    {
+      "epoch": 0.9520969603693729,
+      "grad_norm": 1.204164911132609,
+      "learning_rate": 1.201133917320363e-07,
+      "loss": 0.8181,
+      "step": 4949
+    },
+    {
+      "epoch": 0.9522893420546364,
+      "grad_norm": 1.2086757603838953,
+      "learning_rate": 1.1915247946305498e-07,
+      "loss": 0.8593,
+      "step": 4950
+    },
+    {
+      "epoch": 0.9524817237398999,
+      "grad_norm": 1.3180806284958029,
+      "learning_rate": 1.1819540327929092e-07,
+      "loss": 0.8081,
+      "step": 4951
+    },
+    {
+      "epoch": 0.9526741054251635,
+      "grad_norm": 1.4797013843707614,
+      "learning_rate": 1.1724216355231022e-07,
+      "loss": 0.8276,
+      "step": 4952
+    },
+    {
+      "epoch": 0.9528664871104271,
+      "grad_norm": 1.4747876546376493,
+      "learning_rate": 1.1629276065219575e-07,
+      "loss": 0.8115,
+      "step": 4953
+    },
+    {
+      "epoch": 0.9530588687956907,
+      "grad_norm": 1.1930652501062806,
+      "learning_rate": 1.1534719494753821e-07,
+      "loss": 0.863,
+      "step": 4954
+    },
+    {
+      "epoch": 0.9532512504809542,
+      "grad_norm": 1.322801073198137,
+      "learning_rate": 1.144054668054373e-07,
+      "loss": 0.8103,
+      "step": 4955
+    },
+    {
+      "epoch": 0.9534436321662177,
+      "grad_norm": 1.2517962650008254,
+      "learning_rate": 1.1346757659150498e-07,
+      "loss": 0.8659,
+      "step": 4956
+    },
+    {
+      "epoch": 0.9536360138514813,
+      "grad_norm": 1.5289313543927108,
+      "learning_rate": 1.1253352466986334e-07,
+      "loss": 0.8303,
+      "step": 4957
+    },
+    {
+      "epoch": 0.9538283955367449,
+      "grad_norm": 1.2188267244529387,
+      "learning_rate": 1.116033114031434e-07,
+      "loss": 0.8513,
+      "step": 4958
+    },
+    {
+      "epoch": 0.9540207772220085,
+      "grad_norm": 1.0908243726058193,
+      "learning_rate": 1.1067693715248406e-07,
+      "loss": 0.7928,
+      "step": 4959
+    },
+    {
+      "epoch": 0.954213158907272,
+      "grad_norm": 1.5146965431352655,
+      "learning_rate": 1.0975440227753764e-07,
+      "loss": 0.8089,
+      "step": 4960
+    },
+    {
+      "epoch": 0.9544055405925356,
+      "grad_norm": 1.3210003659840865,
+      "learning_rate": 1.0883570713646318e-07,
+      "loss": 0.8521,
+      "step": 4961
+    },
+    {
+      "epoch": 0.9545979222777992,
+      "grad_norm": 1.28799792951912,
+      "learning_rate": 1.0792085208593095e-07,
+      "loss": 0.8439,
+      "step": 4962
+    },
+    {
+      "epoch": 0.9547903039630627,
+      "grad_norm": 1.388844187023427,
+      "learning_rate": 1.0700983748111792e-07,
+      "loss": 0.7957,
+      "step": 4963
+    },
+    {
+      "epoch": 0.9549826856483262,
+      "grad_norm": 1.3618490077082055,
+      "learning_rate": 1.061026636757101e-07,
+      "loss": 0.8395,
+      "step": 4964
+    },
+    {
+      "epoch": 0.9551750673335898,
+      "grad_norm": 1.4108729388732426,
+      "learning_rate": 1.0519933102190682e-07,
+      "loss": 0.7965,
+      "step": 4965
+    },
+    {
+      "epoch": 0.9553674490188534,
+      "grad_norm": 1.2151278505829695,
+      "learning_rate": 1.0429983987041092e-07,
+      "loss": 0.8537,
+      "step": 4966
+    },
+    {
+      "epoch": 0.955559830704117,
+      "grad_norm": 1.3714196452339535,
+      "learning_rate": 1.0340419057043527e-07,
+      "loss": 0.8446,
+      "step": 4967
+    },
+    {
+      "epoch": 0.9557522123893806,
+      "grad_norm": 1.217649725363833,
+      "learning_rate": 1.0251238346970393e-07,
+      "loss": 0.8476,
+      "step": 4968
+    },
+    {
+      "epoch": 0.955944594074644,
+      "grad_norm": 1.1591197636068644,
+      "learning_rate": 1.0162441891444441e-07,
+      "loss": 0.8299,
+      "step": 4969
+    },
+    {
+      "epoch": 0.9561369757599076,
+      "grad_norm": 1.5897697024313813,
+      "learning_rate": 1.007402972493976e-07,
+      "loss": 0.8018,
+      "step": 4970
+    },
+    {
+      "epoch": 0.9563293574451712,
+      "grad_norm": 1.3949853689500828,
+      "learning_rate": 9.986001881780783e-08,
+      "loss": 0.7819,
+      "step": 4971
+    },
+    {
+      "epoch": 0.9565217391304348,
+      "grad_norm": 0.6328760266442559,
+      "learning_rate": 9.898358396143171e-08,
+      "loss": 0.3108,
+      "step": 4972
+    },
+    {
+      "epoch": 0.9567141208156984,
+      "grad_norm": 1.326835701752691,
+      "learning_rate": 9.811099302052928e-08,
+      "loss": 0.7833,
+      "step": 4973
+    },
+    {
+      "epoch": 0.9569065025009619,
+      "grad_norm": 1.7420707413458991,
+      "learning_rate": 9.72422463338718e-08,
+      "loss": 0.804,
+      "step": 4974
+    },
+    {
+      "epoch": 0.9570988841862255,
+      "grad_norm": 1.1654893496142256,
+      "learning_rate": 9.637734423873612e-08,
+      "loss": 0.7912,
+      "step": 4975
+    },
+    {
+      "epoch": 0.957291265871489,
+      "grad_norm": 1.527523332046872,
+      "learning_rate": 9.55162870709081e-08,
+      "loss": 0.7979,
+      "step": 4976
+    },
+    {
+      "epoch": 0.9574836475567526,
+      "grad_norm": 1.4573619874654666,
+      "learning_rate": 9.465907516467698e-08,
+      "loss": 0.8467,
+      "step": 4977
+    },
+    {
+      "epoch": 0.9576760292420161,
+      "grad_norm": 0.6064257742764699,
+      "learning_rate": 9.380570885284546e-08,
+      "loss": 0.3321,
+      "step": 4978
+    },
+    {
+      "epoch": 0.9578684109272797,
+      "grad_norm": 1.2414988885307214,
+      "learning_rate": 9.295618846671739e-08,
+      "loss": 0.866,
+      "step": 4979
+    },
+    {
+      "epoch": 0.9580607926125433,
+      "grad_norm": 1.5841388242961216,
+      "learning_rate": 9.211051433610674e-08,
+      "loss": 0.8722,
+      "step": 4980
+    },
+    {
+      "epoch": 0.9582531742978069,
+      "grad_norm": 1.313342388692367,
+      "learning_rate": 9.126868678933198e-08,
+      "loss": 0.8088,
+      "step": 4981
+    },
+    {
+      "epoch": 0.9584455559830705,
+      "grad_norm": 1.1566726570673709,
+      "learning_rate": 9.04307061532217e-08,
+      "loss": 0.7847,
+      "step": 4982
+    },
+    {
+      "epoch": 0.9586379376683339,
+      "grad_norm": 1.2138670360534,
+      "learning_rate": 8.959657275310674e-08,
+      "loss": 0.819,
+      "step": 4983
+    },
+    {
+      "epoch": 0.9588303193535975,
+      "grad_norm": 1.7541303176908702,
+      "learning_rate": 8.876628691282918e-08,
+      "loss": 0.8791,
+      "step": 4984
+    },
+    {
+      "epoch": 0.9590227010388611,
+      "grad_norm": 5.262859356322104,
+      "learning_rate": 8.793984895473117e-08,
+      "loss": 0.7902,
+      "step": 4985
+    },
+    {
+      "epoch": 0.9592150827241247,
+      "grad_norm": 1.9566183123101082,
+      "learning_rate": 8.711725919966718e-08,
+      "loss": 0.8012,
+      "step": 4986
+    },
+    {
+      "epoch": 0.9594074644093882,
+      "grad_norm": 1.418561128640819,
+      "learning_rate": 8.629851796699284e-08,
+      "loss": 0.7399,
+      "step": 4987
+    },
+    {
+      "epoch": 0.9595998460946518,
+      "grad_norm": 1.2891229404456546,
+      "learning_rate": 8.54836255745728e-08,
+      "loss": 0.8028,
+      "step": 4988
+    },
+    {
+      "epoch": 0.9597922277799154,
+      "grad_norm": 1.3700450008963903,
+      "learning_rate": 8.467258233877728e-08,
+      "loss": 0.7922,
+      "step": 4989
+    },
+    {
+      "epoch": 0.9599846094651789,
+      "grad_norm": 1.5071909781813697,
+      "learning_rate": 8.386538857447779e-08,
+      "loss": 0.8525,
+      "step": 4990
+    },
+    {
+      "epoch": 0.9601769911504425,
+      "grad_norm": 1.3106674944680228,
+      "learning_rate": 8.306204459505807e-08,
+      "loss": 0.802,
+      "step": 4991
+    },
+    {
+      "epoch": 0.960369372835706,
+      "grad_norm": 1.2820820041370158,
+      "learning_rate": 8.226255071240308e-08,
+      "loss": 0.8031,
+      "step": 4992
+    },
+    {
+      "epoch": 0.9605617545209696,
+      "grad_norm": 1.2604583710546895,
+      "learning_rate": 8.146690723690342e-08,
+      "loss": 0.8274,
+      "step": 4993
+    },
+    {
+      "epoch": 0.9607541362062332,
+      "grad_norm": 1.0800737875018551,
+      "learning_rate": 8.067511447745535e-08,
+      "loss": 0.8429,
+      "step": 4994
+    },
+    {
+      "epoch": 0.9609465178914968,
+      "grad_norm": 1.1838076730429552,
+      "learning_rate": 7.988717274146074e-08,
+      "loss": 0.7818,
+      "step": 4995
+    },
+    {
+      "epoch": 0.9611388995767602,
+      "grad_norm": 1.2705944180120334,
+      "learning_rate": 7.910308233482488e-08,
+      "loss": 0.8398,
+      "step": 4996
+    },
+    {
+      "epoch": 0.9613312812620238,
+      "grad_norm": 1.3988105697249398,
+      "learning_rate": 7.832284356195764e-08,
+      "loss": 0.8417,
+      "step": 4997
+    },
+    {
+      "epoch": 0.9615236629472874,
+      "grad_norm": 1.1445077313711756,
+      "learning_rate": 7.754645672577776e-08,
+      "loss": 0.7893,
+      "step": 4998
+    },
+    {
+      "epoch": 0.961716044632551,
+      "grad_norm": 1.3503237918478177,
+      "learning_rate": 7.677392212770196e-08,
+      "loss": 0.7447,
+      "step": 4999
+    },
+    {
+      "epoch": 0.9619084263178146,
+      "grad_norm": 1.433815473512026,
+      "learning_rate": 7.600524006765808e-08,
+      "loss": 0.8173,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9621008080030781,
+      "grad_norm": 1.4632973922652335,
+      "learning_rate": 7.524041084407185e-08,
+      "loss": 0.7552,
+      "step": 5001
+    },
+    {
+      "epoch": 0.9622931896883417,
+      "grad_norm": 1.4614184689040388,
+      "learning_rate": 7.447943475387797e-08,
+      "loss": 0.8152,
+      "step": 5002
+    },
+    {
+      "epoch": 0.9624855713736052,
+      "grad_norm": 1.671353302172112,
+      "learning_rate": 7.372231209251346e-08,
+      "loss": 0.8293,
+      "step": 5003
+    },
+    {
+      "epoch": 0.9626779530588688,
+      "grad_norm": 1.4504547374598782,
+      "learning_rate": 7.296904315391873e-08,
+      "loss": 0.8655,
+      "step": 5004
+    },
+    {
+      "epoch": 0.9628703347441323,
+      "grad_norm": 1.2506500466292498,
+      "learning_rate": 7.221962823053874e-08,
+      "loss": 0.8334,
+      "step": 5005
+    },
+    {
+      "epoch": 0.9630627164293959,
+      "grad_norm": 1.2693138591530229,
+      "learning_rate": 7.147406761332298e-08,
+      "loss": 0.7894,
+      "step": 5006
+    },
+    {
+      "epoch": 0.9632550981146595,
+      "grad_norm": 1.118627382206436,
+      "learning_rate": 7.073236159172325e-08,
+      "loss": 0.8031,
+      "step": 5007
+    },
+    {
+      "epoch": 0.9634474797999231,
+      "grad_norm": 1.2372449992728636,
+      "learning_rate": 6.999451045369587e-08,
+      "loss": 0.8031,
+      "step": 5008
+    },
+    {
+      "epoch": 0.9636398614851867,
+      "grad_norm": 1.352243673102938,
+      "learning_rate": 6.926051448569948e-08,
+      "loss": 0.8661,
+      "step": 5009
+    },
+    {
+      "epoch": 0.9638322431704501,
+      "grad_norm": 1.2683428800976404,
+      "learning_rate": 6.853037397269724e-08,
+      "loss": 0.7788,
+      "step": 5010
+    },
+    {
+      "epoch": 0.9640246248557137,
+      "grad_norm": 1.3617906596702039,
+      "learning_rate": 6.78040891981524e-08,
+      "loss": 0.8365,
+      "step": 5011
+    },
+    {
+      "epoch": 0.9642170065409773,
+      "grad_norm": 1.2657839745507864,
+      "learning_rate": 6.70816604440383e-08,
+      "loss": 0.8653,
+      "step": 5012
+    },
+    {
+      "epoch": 0.9644093882262409,
+      "grad_norm": 1.4861769124393307,
+      "learning_rate": 6.63630879908217e-08,
+      "loss": 0.8438,
+      "step": 5013
+    },
+    {
+      "epoch": 0.9646017699115044,
+      "grad_norm": 1.3640935661094564,
+      "learning_rate": 6.564837211748054e-08,
+      "loss": 0.7889,
+      "step": 5014
+    },
+    {
+      "epoch": 0.964794151596768,
+      "grad_norm": 1.4086504544510365,
+      "learning_rate": 6.493751310149177e-08,
+      "loss": 0.8639,
+      "step": 5015
+    },
+    {
+      "epoch": 0.9649865332820315,
+      "grad_norm": 1.237839865220621,
+      "learning_rate": 6.42305112188335e-08,
+      "loss": 0.8649,
+      "step": 5016
+    },
+    {
+      "epoch": 0.9651789149672951,
+      "grad_norm": 1.4315956861983798,
+      "learning_rate": 6.352736674398951e-08,
+      "loss": 0.8501,
+      "step": 5017
+    },
+    {
+      "epoch": 0.9653712966525587,
+      "grad_norm": 1.509946579700772,
+      "learning_rate": 6.282807994994477e-08,
+      "loss": 0.8335,
+      "step": 5018
+    },
+    {
+      "epoch": 0.9655636783378222,
+      "grad_norm": 1.1109526876163887,
+      "learning_rate": 6.213265110818656e-08,
+      "loss": 0.8437,
+      "step": 5019
+    },
+    {
+      "epoch": 0.9657560600230858,
+      "grad_norm": 1.3043338263198563,
+      "learning_rate": 6.144108048870335e-08,
+      "loss": 0.8318,
+      "step": 5020
+    },
+    {
+      "epoch": 0.9659484417083494,
+      "grad_norm": 1.446280404595158,
+      "learning_rate": 6.075336835998813e-08,
+      "loss": 0.8445,
+      "step": 5021
+    },
+    {
+      "epoch": 0.966140823393613,
+      "grad_norm": 0.6715338684319743,
+      "learning_rate": 6.00695149890329e-08,
+      "loss": 0.338,
+      "step": 5022
+    },
+    {
+      "epoch": 0.9663332050788765,
+      "grad_norm": 1.5804583880442868,
+      "learning_rate": 5.938952064133419e-08,
+      "loss": 0.8218,
+      "step": 5023
+    },
+    {
+      "epoch": 0.96652558676414,
+      "grad_norm": 1.4832818294669472,
+      "learning_rate": 5.871338558088857e-08,
+      "loss": 0.8382,
+      "step": 5024
+    },
+    {
+      "epoch": 0.9667179684494036,
+      "grad_norm": 1.5100522490172432,
+      "learning_rate": 5.8041110070194976e-08,
+      "loss": 0.8577,
+      "step": 5025
+    },
+    {
+      "epoch": 0.9669103501346672,
+      "grad_norm": 1.273127171533179,
+      "learning_rate": 5.7372694370254614e-08,
+      "loss": 0.8472,
+      "step": 5026
+    },
+    {
+      "epoch": 0.9671027318199308,
+      "grad_norm": 1.5024850238542027,
+      "learning_rate": 5.67081387405688e-08,
+      "loss": 0.7842,
+      "step": 5027
+    },
+    {
+      "epoch": 0.9672951135051943,
+      "grad_norm": 1.2891306087782033,
+      "learning_rate": 5.6047443439141146e-08,
+      "loss": 0.8225,
+      "step": 5028
+    },
+    {
+      "epoch": 0.9674874951904578,
+      "grad_norm": 1.3090563162486775,
+      "learning_rate": 5.539060872247537e-08,
+      "loss": 0.8236,
+      "step": 5029
+    },
+    {
+      "epoch": 0.9676798768757214,
+      "grad_norm": 1.0582232141629122,
+      "learning_rate": 5.47376348455797e-08,
+      "loss": 0.8216,
+      "step": 5030
+    },
+    {
+      "epoch": 0.967872258560985,
+      "grad_norm": 0.5958476445119825,
+      "learning_rate": 5.408852206195914e-08,
+      "loss": 0.3305,
+      "step": 5031
+    },
+    {
+      "epoch": 0.9680646402462486,
+      "grad_norm": 1.4221866360255275,
+      "learning_rate": 5.344327062362098e-08,
+      "loss": 0.7851,
+      "step": 5032
+    },
+    {
+      "epoch": 0.9682570219315121,
+      "grad_norm": 0.574781061578418,
+      "learning_rate": 5.2801880781075954e-08,
+      "loss": 0.3004,
+      "step": 5033
+    },
+    {
+      "epoch": 0.9684494036167757,
+      "grad_norm": 1.5670589114993945,
+      "learning_rate": 5.216435278333376e-08,
+      "loss": 0.8036,
+      "step": 5034
+    },
+    {
+      "epoch": 0.9686417853020393,
+      "grad_norm": 1.1541664320969705,
+      "learning_rate": 5.153068687790197e-08,
+      "loss": 0.8369,
+      "step": 5035
+    },
+    {
+      "epoch": 0.9688341669873028,
+      "grad_norm": 1.4931611802145555,
+      "learning_rate": 5.0900883310794903e-08,
+      "loss": 0.8463,
+      "step": 5036
+    },
+    {
+      "epoch": 0.9690265486725663,
+      "grad_norm": 1.387095744380816,
+      "learning_rate": 5.0274942326521414e-08,
+      "loss": 0.8103,
+      "step": 5037
+    },
+    {
+      "epoch": 0.9692189303578299,
+      "grad_norm": 1.3471714176273355,
+      "learning_rate": 4.9652864168096e-08,
+      "loss": 0.783,
+      "step": 5038
+    },
+    {
+      "epoch": 0.9694113120430935,
+      "grad_norm": 1.5549588824483398,
+      "learning_rate": 4.9034649077027706e-08,
+      "loss": 0.7748,
+      "step": 5039
+    },
+    {
+      "epoch": 0.9696036937283571,
+      "grad_norm": 1.2795599904075905,
+      "learning_rate": 4.84202972933312e-08,
+      "loss": 0.8398,
+      "step": 5040
+    },
+    {
+      "epoch": 0.9697960754136207,
+      "grad_norm": 1.6504731073225043,
+      "learning_rate": 4.7809809055517906e-08,
+      "loss": 0.8933,
+      "step": 5041
+    },
+    {
+      "epoch": 0.9699884570988841,
+      "grad_norm": 1.5048205960239711,
+      "learning_rate": 4.720318460060047e-08,
+      "loss": 0.8445,
+      "step": 5042
+    },
+    {
+      "epoch": 0.9701808387841477,
+      "grad_norm": 1.3069773874632133,
+      "learning_rate": 4.6600424164091606e-08,
+      "loss": 0.8194,
+      "step": 5043
+    },
+    {
+      "epoch": 0.9703732204694113,
+      "grad_norm": 1.4646487781827275,
+      "learning_rate": 4.6001527980004125e-08,
+      "loss": 0.7656,
+      "step": 5044
+    },
+    {
+      "epoch": 0.9705656021546749,
+      "grad_norm": 1.8111309797015014,
+      "learning_rate": 4.54064962808487e-08,
+      "loss": 0.8117,
+      "step": 5045
+    },
+    {
+      "epoch": 0.9707579838399384,
+      "grad_norm": 1.2296840158185163,
+      "learning_rate": 4.4815329297639434e-08,
+      "loss": 0.8094,
+      "step": 5046
+    },
+    {
+      "epoch": 0.970950365525202,
+      "grad_norm": 1.2572212699488907,
+      "learning_rate": 4.422802725988606e-08,
+      "loss": 0.7972,
+      "step": 5047
+    },
+    {
+      "epoch": 0.9711427472104656,
+      "grad_norm": 1.5091599164709526,
+      "learning_rate": 4.364459039559843e-08,
+      "loss": 0.8094,
+      "step": 5048
+    },
+    {
+      "epoch": 0.9713351288957291,
+      "grad_norm": 2.8519712624723454,
+      "learning_rate": 4.3065018931289784e-08,
+      "loss": 0.8375,
+      "step": 5049
+    },
+    {
+      "epoch": 0.9715275105809927,
+      "grad_norm": 1.4350961001519655,
+      "learning_rate": 4.248931309196791e-08,
+      "loss": 0.8188,
+      "step": 5050
+    },
+    {
+      "epoch": 0.9717198922662562,
+      "grad_norm": 2.087105565317081,
+      "learning_rate": 4.1917473101140696e-08,
+      "loss": 0.7791,
+      "step": 5051
+    },
+    {
+      "epoch": 0.9719122739515198,
+      "grad_norm": 1.1604446613294321,
+      "learning_rate": 4.134949918081832e-08,
+      "loss": 0.7789,
+      "step": 5052
+    },
+    {
+      "epoch": 0.9721046556367834,
+      "grad_norm": 1.285182991208343,
+      "learning_rate": 4.0785391551506626e-08,
+      "loss": 0.8035,
+      "step": 5053
+    },
+    {
+      "epoch": 0.972297037322047,
+      "grad_norm": 1.2820837405881078,
+      "learning_rate": 4.022515043221154e-08,
+      "loss": 0.7897,
+      "step": 5054
+    },
+    {
+      "epoch": 0.9724894190073105,
+      "grad_norm": 1.5814400542770324,
+      "learning_rate": 3.966877604043795e-08,
+      "loss": 0.826,
+      "step": 5055
+    },
+    {
+      "epoch": 0.972681800692574,
+      "grad_norm": 1.3937433020039622,
+      "learning_rate": 3.9116268592189755e-08,
+      "loss": 0.8377,
+      "step": 5056
+    },
+    {
+      "epoch": 0.9728741823778376,
+      "grad_norm": 1.3251681650814318,
+      "learning_rate": 3.8567628301969806e-08,
+      "loss": 0.833,
+      "step": 5057
+    },
+    {
+      "epoch": 0.9730665640631012,
+      "grad_norm": 1.4125445595312403,
+      "learning_rate": 3.802285538277772e-08,
+      "loss": 0.8518,
+      "step": 5058
+    },
+    {
+      "epoch": 0.9732589457483648,
+      "grad_norm": 1.5312605736291103,
+      "learning_rate": 3.748195004611543e-08,
+      "loss": 0.7939,
+      "step": 5059
+    },
+    {
+      "epoch": 0.9734513274336283,
+      "grad_norm": 1.4211802599377816,
+      "learning_rate": 3.69449125019794e-08,
+      "loss": 0.8335,
+      "step": 5060
+    },
+    {
+      "epoch": 0.9736437091188919,
+      "grad_norm": 1.178759480943146,
+      "learning_rate": 3.6411742958866184e-08,
+      "loss": 0.7864,
+      "step": 5061
+    },
+    {
+      "epoch": 0.9738360908041555,
+      "grad_norm": 1.4400363708125248,
+      "learning_rate": 3.588244162377019e-08,
+      "loss": 0.8724,
+      "step": 5062
+    },
+    {
+      "epoch": 0.974028472489419,
+      "grad_norm": 0.5902144824671413,
+      "learning_rate": 3.5357008702185945e-08,
+      "loss": 0.3211,
+      "step": 5063
+    },
+    {
+      "epoch": 0.9742208541746825,
+      "grad_norm": 0.6061854077151229,
+      "learning_rate": 3.483544439810249e-08,
+      "loss": 0.3946,
+      "step": 5064
+    },
+    {
+      "epoch": 0.9744132358599461,
+      "grad_norm": 1.2261014710751825,
+      "learning_rate": 3.4317748914011187e-08,
+      "loss": 0.7707,
+      "step": 5065
+    },
+    {
+      "epoch": 0.9746056175452097,
+      "grad_norm": 1.1560108842841867,
+      "learning_rate": 3.3803922450897917e-08,
+      "loss": 0.7579,
+      "step": 5066
+    },
+    {
+      "epoch": 0.9747979992304733,
+      "grad_norm": 1.6801498545744051,
+      "learning_rate": 3.329396520824757e-08,
+      "loss": 0.8117,
+      "step": 5067
+    },
+    {
+      "epoch": 0.9749903809157369,
+      "grad_norm": 1.3800288661563749,
+      "learning_rate": 3.2787877384045095e-08,
+      "loss": 0.7931,
+      "step": 5068
+    },
+    {
+      "epoch": 0.9751827626010003,
+      "grad_norm": 1.3730774392498228,
+      "learning_rate": 3.228565917476889e-08,
+      "loss": 0.831,
+      "step": 5069
+    },
+    {
+      "epoch": 0.9753751442862639,
+      "grad_norm": 2.229178427133514,
+      "learning_rate": 3.178731077539743e-08,
+      "loss": 0.8475,
+      "step": 5070
+    },
+    {
+      "epoch": 0.9755675259715275,
+      "grad_norm": 1.179656609388754,
+      "learning_rate": 3.129283237940928e-08,
+      "loss": 0.8162,
+      "step": 5071
+    },
+    {
+      "epoch": 0.9757599076567911,
+      "grad_norm": 1.307925871310655,
+      "learning_rate": 3.080222417877421e-08,
+      "loss": 0.805,
+      "step": 5072
+    },
+    {
+      "epoch": 0.9759522893420547,
+      "grad_norm": 1.4453795163136816,
+      "learning_rate": 3.031548636396764e-08,
+      "loss": 0.8276,
+      "step": 5073
+    },
+    {
+      "epoch": 0.9761446710273182,
+      "grad_norm": 1.4125071431981038,
+      "learning_rate": 2.983261912395397e-08,
+      "loss": 0.8386,
+      "step": 5074
+    },
+    {
+      "epoch": 0.9763370527125818,
+      "grad_norm": 1.5793198281212486,
+      "learning_rate": 2.9353622646199898e-08,
+      "loss": 0.8451,
+      "step": 5075
+    },
+    {
+      "epoch": 0.9765294343978453,
+      "grad_norm": 2.964849589778244,
+      "learning_rate": 2.8878497116671124e-08,
+      "loss": 0.7882,
+      "step": 5076
+    },
+    {
+      "epoch": 0.9767218160831089,
+      "grad_norm": 1.4135986546729395,
+      "learning_rate": 2.8407242719823424e-08,
+      "loss": 0.8771,
+      "step": 5077
+    },
+    {
+      "epoch": 0.9769141977683724,
+      "grad_norm": 1.340757182794578,
+      "learning_rate": 2.7939859638617118e-08,
+      "loss": 0.8297,
+      "step": 5078
+    },
+    {
+      "epoch": 0.977106579453636,
+      "grad_norm": 3.625421179174665,
+      "learning_rate": 2.7476348054504832e-08,
+      "loss": 0.805,
+      "step": 5079
+    },
+    {
+      "epoch": 0.9772989611388996,
+      "grad_norm": 1.3597457974666742,
+      "learning_rate": 2.7016708147439285e-08,
+      "loss": 0.8228,
+      "step": 5080
+    },
+    {
+      "epoch": 0.9774913428241632,
+      "grad_norm": 1.3739794538318182,
+      "learning_rate": 2.6560940095866626e-08,
+      "loss": 0.8574,
+      "step": 5081
+    },
+    {
+      "epoch": 0.9776837245094268,
+      "grad_norm": 0.5935662273765063,
+      "learning_rate": 2.6109044076733092e-08,
+      "loss": 0.3429,
+      "step": 5082
+    },
+    {
+      "epoch": 0.9778761061946902,
+      "grad_norm": 1.4950647298627069,
+      "learning_rate": 2.5661020265479452e-08,
+      "loss": 0.8692,
+      "step": 5083
+    },
+    {
+      "epoch": 0.9780684878799538,
+      "grad_norm": 1.3394540844948708,
+      "learning_rate": 2.5216868836043242e-08,
+      "loss": 0.8155,
+      "step": 5084
+    },
+    {
+      "epoch": 0.9782608695652174,
+      "grad_norm": 1.5820137066901896,
+      "learning_rate": 2.4776589960862074e-08,
+      "loss": 0.8357,
+      "step": 5085
+    },
+    {
+      "epoch": 0.978453251250481,
+      "grad_norm": 2.048177760295367,
+      "learning_rate": 2.434018381086589e-08,
+      "loss": 0.8611,
+      "step": 5086
+    },
+    {
+      "epoch": 0.9786456329357445,
+      "grad_norm": 1.2757748030101692,
+      "learning_rate": 2.3907650555481387e-08,
+      "loss": 0.7692,
+      "step": 5087
+    },
+    {
+      "epoch": 0.9788380146210081,
+      "grad_norm": 1.426843393203853,
+      "learning_rate": 2.3478990362634235e-08,
+      "loss": 0.8339,
+      "step": 5088
+    },
+    {
+      "epoch": 0.9790303963062716,
+      "grad_norm": 1.1962918817616133,
+      "learning_rate": 2.3054203398743537e-08,
+      "loss": 0.8149,
+      "step": 5089
+    },
+    {
+      "epoch": 0.9792227779915352,
+      "grad_norm": 1.800284227488717,
+      "learning_rate": 2.263328982872959e-08,
+      "loss": 0.8958,
+      "step": 5090
+    },
+    {
+      "epoch": 0.9794151596767988,
+      "grad_norm": 1.3355976297550187,
+      "learning_rate": 2.221624981600168e-08,
+      "loss": 0.8415,
+      "step": 5091
+    },
+    {
+      "epoch": 0.9796075413620623,
+      "grad_norm": 1.593368150365925,
+      "learning_rate": 2.1803083522471402e-08,
+      "loss": 0.9,
+      "step": 5092
+    },
+    {
+      "epoch": 0.9797999230473259,
+      "grad_norm": 1.2623250633784533,
+      "learning_rate": 2.1393791108542672e-08,
+      "loss": 0.8484,
+      "step": 5093
+    },
+    {
+      "epoch": 0.9799923047325895,
+      "grad_norm": 1.2405183824279435,
+      "learning_rate": 2.098837273311838e-08,
+      "loss": 0.8148,
+      "step": 5094
+    },
+    {
+      "epoch": 0.9801846864178531,
+      "grad_norm": 1.1592574599388206,
+      "learning_rate": 2.058682855359595e-08,
+      "loss": 0.8326,
+      "step": 5095
+    },
+    {
+      "epoch": 0.9803770681031165,
+      "grad_norm": 1.1966410978476767,
+      "learning_rate": 2.0189158725867353e-08,
+      "loss": 0.7984,
+      "step": 5096
+    },
+    {
+      "epoch": 0.9805694497883801,
+      "grad_norm": 1.2830629323042138,
+      "learning_rate": 1.979536340432131e-08,
+      "loss": 0.7877,
+      "step": 5097
+    },
+    {
+      "epoch": 0.9807618314736437,
+      "grad_norm": 1.3641122014009774,
+      "learning_rate": 1.9405442741844415e-08,
+      "loss": 0.759,
+      "step": 5098
+    },
+    {
+      "epoch": 0.9809542131589073,
+      "grad_norm": 1.3862247050932486,
+      "learning_rate": 1.9019396889816688e-08,
+      "loss": 0.813,
+      "step": 5099
+    },
+    {
+      "epoch": 0.9811465948441709,
+      "grad_norm": 1.3573360576400737,
+      "learning_rate": 1.8637225998114904e-08,
+      "loss": 0.8122,
+      "step": 5100
+    },
+    {
+      "epoch": 0.9813389765294344,
+      "grad_norm": 1.329411423192412,
+      "learning_rate": 1.825893021510927e-08,
+      "loss": 0.8105,
+      "step": 5101
+    },
+    {
+      "epoch": 0.981531358214698,
+      "grad_norm": 1.2401525897333205,
+      "learning_rate": 1.7884509687668972e-08,
+      "loss": 0.8527,
+      "step": 5102
+    },
+    {
+      "epoch": 0.9817237398999615,
+      "grad_norm": 1.3138125117958148,
+      "learning_rate": 1.7513964561156617e-08,
+      "loss": 0.8188,
+      "step": 5103
+    },
+    {
+      "epoch": 0.9819161215852251,
+      "grad_norm": 1.477342661252682,
+      "learning_rate": 1.714729497942935e-08,
+      "loss": 0.847,
+      "step": 5104
+    },
+    {
+      "epoch": 0.9821085032704886,
+      "grad_norm": 1.330345225149823,
+      "learning_rate": 1.6784501084843307e-08,
+      "loss": 0.8611,
+      "step": 5105
+    },
+    {
+      "epoch": 0.9823008849557522,
+      "grad_norm": 1.5501862779230415,
+      "learning_rate": 1.6425583018244706e-08,
+      "loss": 0.8189,
+      "step": 5106
+    },
+    {
+      "epoch": 0.9824932666410158,
+      "grad_norm": 1.4685411834799789,
+      "learning_rate": 1.607054091897986e-08,
+      "loss": 0.7514,
+      "step": 5107
+    },
+    {
+      "epoch": 0.9826856483262794,
+      "grad_norm": 1.2675912704215926,
+      "learning_rate": 1.57193749248874e-08,
+      "loss": 0.8155,
+      "step": 5108
+    },
+    {
+      "epoch": 0.982878030011543,
+      "grad_norm": 1.1357575215686293,
+      "learning_rate": 1.537208517230271e-08,
+      "loss": 0.849,
+      "step": 5109
+    },
+    {
+      "epoch": 0.9830704116968064,
+      "grad_norm": 1.3702945186732136,
+      "learning_rate": 1.5028671796055715e-08,
+      "loss": 0.8394,
+      "step": 5110
+    },
+    {
+      "epoch": 0.98326279338207,
+      "grad_norm": 1.3372517286949583,
+      "learning_rate": 1.4689134929470884e-08,
+      "loss": 0.8416,
+      "step": 5111
+    },
+    {
+      "epoch": 0.9834551750673336,
+      "grad_norm": 1.3240493481074067,
+      "learning_rate": 1.435347470436832e-08,
+      "loss": 0.7949,
+      "step": 5112
+    },
+    {
+      "epoch": 0.9836475567525972,
+      "grad_norm": 1.1749391615275924,
+      "learning_rate": 1.4021691251062675e-08,
+      "loss": 0.7956,
+      "step": 5113
+    },
+    {
+      "epoch": 0.9838399384378607,
+      "grad_norm": 1.5522523972367883,
+      "learning_rate": 1.3693784698363133e-08,
+      "loss": 0.7789,
+      "step": 5114
+    },
+    {
+      "epoch": 0.9840323201231242,
+      "grad_norm": 1.3385585496956978,
+      "learning_rate": 1.3369755173575639e-08,
+      "loss": 0.8226,
+      "step": 5115
+    },
+    {
+      "epoch": 0.9842247018083878,
+      "grad_norm": 2.225372383164197,
+      "learning_rate": 1.3049602802498451e-08,
+      "loss": 0.8674,
+      "step": 5116
+    },
+    {
+      "epoch": 0.9844170834936514,
+      "grad_norm": 1.3451516849679617,
+      "learning_rate": 1.273332770942659e-08,
+      "loss": 0.8304,
+      "step": 5117
+    },
+    {
+      "epoch": 0.984609465178915,
+      "grad_norm": 0.6250611851165766,
+      "learning_rate": 1.2420930017148503e-08,
+      "loss": 0.3336,
+      "step": 5118
+    },
+    {
+      "epoch": 0.9848018468641785,
+      "grad_norm": 1.4743075141037307,
+      "learning_rate": 1.2112409846947171e-08,
+      "loss": 0.8529,
+      "step": 5119
+    },
+    {
+      "epoch": 0.9849942285494421,
+      "grad_norm": 1.2735304613618266,
+      "learning_rate": 1.1807767318602337e-08,
+      "loss": 0.8284,
+      "step": 5120
+    },
+    {
+      "epoch": 0.9851866102347057,
+      "grad_norm": 1.5782973913764564,
+      "learning_rate": 1.150700255038606e-08,
+      "loss": 0.8208,
+      "step": 5121
+    },
+    {
+      "epoch": 0.9853789919199692,
+      "grad_norm": 1.3767809441106615,
+      "learning_rate": 1.1210115659063825e-08,
+      "loss": 0.8039,
+      "step": 5122
+    },
+    {
+      "epoch": 0.9855713736052328,
+      "grad_norm": 1.4120433687931457,
+      "learning_rate": 1.0917106759900097e-08,
+      "loss": 0.7801,
+      "step": 5123
+    },
+    {
+      "epoch": 0.9857637552904963,
+      "grad_norm": 1.8089233781317646,
+      "learning_rate": 1.0627975966649439e-08,
+      "loss": 0.8304,
+      "step": 5124
+    },
+    {
+      "epoch": 0.9859561369757599,
+      "grad_norm": 1.2878455183221413,
+      "learning_rate": 1.034272339156206e-08,
+      "loss": 0.8584,
+      "step": 5125
+    },
+    {
+      "epoch": 0.9861485186610235,
+      "grad_norm": 1.25788793734854,
+      "learning_rate": 1.0061349145383814e-08,
+      "loss": 0.8022,
+      "step": 5126
+    },
+    {
+      "epoch": 0.9863409003462871,
+      "grad_norm": 1.3375530626323318,
+      "learning_rate": 9.783853337353987e-09,
+      "loss": 0.8449,
+      "step": 5127
+    },
+    {
+      "epoch": 0.9865332820315506,
+      "grad_norm": 0.5778930092662605,
+      "learning_rate": 9.510236075205292e-09,
+      "loss": 0.3298,
+      "step": 5128
+    },
+    {
+      "epoch": 0.9867256637168141,
+      "grad_norm": 1.399360844307583,
+      "learning_rate": 9.240497465164978e-09,
+      "loss": 0.7923,
+      "step": 5129
+    },
+    {
+      "epoch": 0.9869180454020777,
+      "grad_norm": 1.4607022959019351,
+      "learning_rate": 8.974637611955939e-09,
+      "loss": 0.8589,
+      "step": 5130
+    },
+    {
+      "epoch": 0.9871104270873413,
+      "grad_norm": 1.575835201200685,
+      "learning_rate": 8.712656618793391e-09,
+      "loss": 0.8228,
+      "step": 5131
+    },
+    {
+      "epoch": 0.9873028087726049,
+      "grad_norm": 1.4747286324219269,
+      "learning_rate": 8.454554587388198e-09,
+      "loss": 0.8226,
+      "step": 5132
+    },
+    {
+      "epoch": 0.9874951904578684,
+      "grad_norm": 1.2124157915776619,
+      "learning_rate": 8.200331617943535e-09,
+      "loss": 0.8546,
+      "step": 5133
+    },
+    {
+      "epoch": 0.987687572143132,
+      "grad_norm": 1.288568842043731,
+      "learning_rate": 7.949987809158232e-09,
+      "loss": 0.8435,
+      "step": 5134
+    },
+    {
+      "epoch": 0.9878799538283956,
+      "grad_norm": 0.6116609856723411,
+      "learning_rate": 7.703523258223433e-09,
+      "loss": 0.3341,
+      "step": 5135
+    },
+    {
+      "epoch": 0.9880723355136591,
+      "grad_norm": 1.220578497215922,
+      "learning_rate": 7.460938060825929e-09,
+      "loss": 0.8014,
+      "step": 5136
+    },
+    {
+      "epoch": 0.9882647171989226,
+      "grad_norm": 1.110499074618718,
+      "learning_rate": 7.222232311145938e-09,
+      "loss": 0.837,
+      "step": 5137
+    },
+    {
+      "epoch": 0.9884570988841862,
+      "grad_norm": 1.4278848537561752,
+      "learning_rate": 6.987406101855998e-09,
+      "loss": 0.8257,
+      "step": 5138
+    },
+    {
+      "epoch": 0.9886494805694498,
+      "grad_norm": 1.5664325290403587,
+      "learning_rate": 6.756459524125403e-09,
+      "loss": 0.8655,
+      "step": 5139
+    },
+    {
+      "epoch": 0.9888418622547134,
+      "grad_norm": 1.2010634789531216,
+      "learning_rate": 6.5293926676135434e-09,
+      "loss": 0.8987,
+      "step": 5140
+    },
+    {
+      "epoch": 0.989034243939977,
+      "grad_norm": 1.369839882170847,
+      "learning_rate": 6.306205620477679e-09,
+      "loss": 0.823,
+      "step": 5141
+    },
+    {
+      "epoch": 0.9892266256252404,
+      "grad_norm": 1.4476163538851818,
+      "learning_rate": 6.086898469365166e-09,
+      "loss": 0.8143,
+      "step": 5142
+    },
+    {
+      "epoch": 0.989419007310504,
+      "grad_norm": 1.2791699816015902,
+      "learning_rate": 5.871471299419007e-09,
+      "loss": 0.8592,
+      "step": 5143
+    },
+    {
+      "epoch": 0.9896113889957676,
+      "grad_norm": 1.3476002280295851,
+      "learning_rate": 5.6599241942767445e-09,
+      "loss": 0.8559,
+      "step": 5144
+    },
+    {
+      "epoch": 0.9898037706810312,
+      "grad_norm": 1.4272509784186076,
+      "learning_rate": 5.452257236066017e-09,
+      "loss": 0.8042,
+      "step": 5145
+    },
+    {
+      "epoch": 0.9899961523662947,
+      "grad_norm": 0.5865033373350239,
+      "learning_rate": 5.248470505412328e-09,
+      "loss": 0.3281,
+      "step": 5146
+    },
+    {
+      "epoch": 0.9901885340515583,
+      "grad_norm": 1.7892921607306886,
+      "learning_rate": 5.0485640814312844e-09,
+      "loss": 0.8805,
+      "step": 5147
+    },
+    {
+      "epoch": 0.9903809157368219,
+      "grad_norm": 1.3373729411529458,
+      "learning_rate": 4.8525380417330234e-09,
+      "loss": 0.8408,
+      "step": 5148
+    },
+    {
+      "epoch": 0.9905732974220854,
+      "grad_norm": 1.662556845179463,
+      "learning_rate": 4.660392462424446e-09,
+      "loss": 0.8207,
+      "step": 5149
+    },
+    {
+      "epoch": 0.990765679107349,
+      "grad_norm": 1.3576529991367168,
+      "learning_rate": 4.472127418099215e-09,
+      "loss": 0.8582,
+      "step": 5150
+    },
+    {
+      "epoch": 0.9909580607926125,
+      "grad_norm": 1.7471868696466155,
+      "learning_rate": 4.287742981851084e-09,
+      "loss": 0.822,
+      "step": 5151
+    },
+    {
+      "epoch": 0.9911504424778761,
+      "grad_norm": 1.261087102318673,
+      "learning_rate": 4.1072392252639034e-09,
+      "loss": 0.7861,
+      "step": 5152
+    },
+    {
+      "epoch": 0.9913428241631397,
+      "grad_norm": 1.2779285471351944,
+      "learning_rate": 3.930616218414951e-09,
+      "loss": 0.8358,
+      "step": 5153
+    },
+    {
+      "epoch": 0.9915352058484033,
+      "grad_norm": 1.3165992901381942,
+      "learning_rate": 3.757874029874931e-09,
+      "loss": 0.8508,
+      "step": 5154
+    },
+    {
+      "epoch": 0.9917275875336667,
+      "grad_norm": 1.2009448699131093,
+      "learning_rate": 3.5890127267090844e-09,
+      "loss": 0.8406,
+      "step": 5155
+    },
+    {
+      "epoch": 0.9919199692189303,
+      "grad_norm": 1.2544844455389965,
+      "learning_rate": 3.424032374476083e-09,
+      "loss": 0.8114,
+      "step": 5156
+    },
+    {
+      "epoch": 0.9921123509041939,
+      "grad_norm": 1.4811471944269665,
+      "learning_rate": 3.2629330372246915e-09,
+      "loss": 0.8389,
+      "step": 5157
+    },
+    {
+      "epoch": 0.9923047325894575,
+      "grad_norm": 1.404955069423165,
+      "learning_rate": 3.105714777501545e-09,
+      "loss": 0.7962,
+      "step": 5158
+    },
+    {
+      "epoch": 0.9924971142747211,
+      "grad_norm": 1.4472104833688317,
+      "learning_rate": 2.9523776563422644e-09,
+      "loss": 0.8227,
+      "step": 5159
+    },
+    {
+      "epoch": 0.9926894959599846,
+      "grad_norm": 1.440167021019008,
+      "learning_rate": 2.802921733278119e-09,
+      "loss": 0.8181,
+      "step": 5160
+    },
+    {
+      "epoch": 0.9928818776452482,
+      "grad_norm": 1.4999415420488402,
+      "learning_rate": 2.657347066333804e-09,
+      "loss": 0.8527,
+      "step": 5161
+    },
+    {
+      "epoch": 0.9930742593305117,
+      "grad_norm": 1.3066584787431275,
+      "learning_rate": 2.5156537120263335e-09,
+      "loss": 0.849,
+      "step": 5162
+    },
+    {
+      "epoch": 0.9932666410157753,
+      "grad_norm": 1.2326674619181472,
+      "learning_rate": 2.3778417253650376e-09,
+      "loss": 0.8687,
+      "step": 5163
+    },
+    {
+      "epoch": 0.9934590227010388,
+      "grad_norm": 1.2162302445111075,
+      "learning_rate": 2.2439111598537844e-09,
+      "loss": 0.8316,
+      "step": 5164
+    },
+    {
+      "epoch": 0.9936514043863024,
+      "grad_norm": 1.205180185901838,
+      "learning_rate": 2.113862067488759e-09,
+      "loss": 0.8658,
+      "step": 5165
+    },
+    {
+      "epoch": 0.993843786071566,
+      "grad_norm": 1.2589211170538492,
+      "learning_rate": 1.987694498760684e-09,
+      "loss": 0.8144,
+      "step": 5166
+    },
+    {
+      "epoch": 0.9940361677568296,
+      "grad_norm": 1.5663414561094566,
+      "learning_rate": 1.865408502650379e-09,
+      "loss": 0.8199,
+      "step": 5167
+    },
+    {
+      "epoch": 0.9942285494420932,
+      "grad_norm": 1.3693296642947863,
+      "learning_rate": 1.747004126635421e-09,
+      "loss": 0.8201,
+      "step": 5168
+    },
+    {
+      "epoch": 0.9944209311273566,
+      "grad_norm": 1.1284837919945288,
+      "learning_rate": 1.6324814166823744e-09,
+      "loss": 0.8109,
+      "step": 5169
+    },
+    {
+      "epoch": 0.9946133128126202,
+      "grad_norm": 1.1654082561922983,
+      "learning_rate": 1.5218404172545609e-09,
+      "loss": 0.8247,
+      "step": 5170
+    },
+    {
+      "epoch": 0.9948056944978838,
+      "grad_norm": 1.3710666498751334,
+      "learning_rate": 1.415081171305399e-09,
+      "loss": 0.8476,
+      "step": 5171
+    },
+    {
+      "epoch": 0.9949980761831474,
+      "grad_norm": 1.2739275953941407,
+      "learning_rate": 1.3122037202828452e-09,
+      "loss": 0.8694,
+      "step": 5172
+    },
+    {
+      "epoch": 0.995190457868411,
+      "grad_norm": 1.1977886656845849,
+      "learning_rate": 1.2132081041282829e-09,
+      "loss": 0.863,
+      "step": 5173
+    },
+    {
+      "epoch": 0.9953828395536745,
+      "grad_norm": 1.6184748821940977,
+      "learning_rate": 1.1180943612754124e-09,
+      "loss": 0.8271,
+      "step": 5174
+    },
+    {
+      "epoch": 0.995575221238938,
+      "grad_norm": 1.4946959082012385,
+      "learning_rate": 1.026862528649142e-09,
+      "loss": 0.8383,
+      "step": 5175
+    },
+    {
+      "epoch": 0.9957676029242016,
+      "grad_norm": 1.680640031111542,
+      "learning_rate": 9.39512641668916e-10,
+      "loss": 0.7965,
+      "step": 5176
+    },
+    {
+      "epoch": 0.9959599846094652,
+      "grad_norm": 1.451374046110849,
+      "learning_rate": 8.560447342487177e-10,
+      "loss": 0.8072,
+      "step": 5177
+    },
+    {
+      "epoch": 0.9961523662947287,
+      "grad_norm": 1.3649215962648944,
+      "learning_rate": 7.764588387915161e-10,
+      "loss": 0.822,
+      "step": 5178
+    },
+    {
+      "epoch": 0.9963447479799923,
+      "grad_norm": 1.2089734462467416,
+      "learning_rate": 7.007549861970387e-10,
+      "loss": 0.7509,
+      "step": 5179
+    },
+    {
+      "epoch": 0.9965371296652559,
+      "grad_norm": 1.3814625334855417,
+      "learning_rate": 6.289332058551089e-10,
+      "loss": 0.8809,
+      "step": 5180
+    },
+    {
+      "epoch": 0.9967295113505195,
+      "grad_norm": 1.2423570229538645,
+      "learning_rate": 5.609935256500887e-10,
+      "loss": 0.8294,
+      "step": 5181
+    },
+    {
+      "epoch": 0.996921893035783,
+      "grad_norm": 1.7269205345439815,
+      "learning_rate": 4.969359719586563e-10,
+      "loss": 0.8013,
+      "step": 5182
+    },
+    {
+      "epoch": 0.9971142747210465,
+      "grad_norm": 1.395636543300398,
+      "learning_rate": 4.3676056964869764e-10,
+      "loss": 0.8236,
+      "step": 5183
+    },
+    {
+      "epoch": 0.9973066564063101,
+      "grad_norm": 1.2717832890401766,
+      "learning_rate": 3.804673420837457e-10,
+      "loss": 0.8129,
+      "step": 5184
+    },
+    {
+      "epoch": 0.9974990380915737,
+      "grad_norm": 1.3632768847788432,
+      "learning_rate": 3.2805631111743064e-10,
+      "loss": 0.8644,
+      "step": 5185
+    },
+    {
+      "epoch": 0.9976914197768373,
+      "grad_norm": 1.1771196861999569,
+      "learning_rate": 2.795274971001405e-10,
+      "loss": 0.7771,
+      "step": 5186
+    },
+    {
+      "epoch": 0.9978838014621008,
+      "grad_norm": 1.4203240473235312,
+      "learning_rate": 2.3488091886902933e-10,
+      "loss": 0.8428,
+      "step": 5187
+    },
+    {
+      "epoch": 0.9980761831473643,
+      "grad_norm": 1.1592635946593879,
+      "learning_rate": 1.941165937602296e-10,
+      "loss": 0.8477,
+      "step": 5188
+    },
+    {
+      "epoch": 0.9982685648326279,
+      "grad_norm": 1.3310307005060529,
+      "learning_rate": 1.5723453759886042e-10,
+      "loss": 0.9138,
+      "step": 5189
+    },
+    {
+      "epoch": 0.9984609465178915,
+      "grad_norm": 1.363171608576605,
+      "learning_rate": 1.2423476470346808e-10,
+      "loss": 0.7951,
+      "step": 5190
+    },
+    {
+      "epoch": 0.9986533282031551,
+      "grad_norm": 1.1710157860772759,
+      "learning_rate": 9.511728788602625e-11,
+      "loss": 0.8324,
+      "step": 5191
+    },
+    {
+      "epoch": 0.9988457098884186,
+      "grad_norm": 1.3561939426996417,
+      "learning_rate": 6.988211845082582e-11,
+      "loss": 0.8773,
+      "step": 5192
+    },
+    {
+      "epoch": 0.9990380915736822,
+      "grad_norm": 1.3438339134388253,
+      "learning_rate": 4.852926619447473e-11,
+      "loss": 0.8837,
+      "step": 5193
+    },
+    {
+      "epoch": 0.9992304732589458,
+      "grad_norm": 1.2623918981145195,
+      "learning_rate": 3.105873940811854e-11,
+      "loss": 0.8623,
+      "step": 5194
+    },
+    {
+      "epoch": 0.9994228549442093,
+      "grad_norm": 1.4561679689864364,
+      "learning_rate": 1.7470544874109706e-11,
+      "loss": 0.82,
+      "step": 5195
+    },
+    {
+      "epoch": 0.9996152366294728,
+      "grad_norm": 1.2449922987160071,
+      "learning_rate": 7.764687866007592e-12,
+      "loss": 0.8269,
+      "step": 5196
+    },
+    {
+      "epoch": 0.9998076183147364,
+      "grad_norm": 2.192954442903564,
+      "learning_rate": 1.9411721552398123e-12,
+      "loss": 0.6743,
+      "step": 5197
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.914099435495328,
+      "learning_rate": 0.0,
+      "loss": 0.4856,
+      "step": 5198
+    },
+    {
+      "epoch": 1.0,
+      "step": 5198,
+      "total_flos": 5135099285733376.0,
+      "train_loss": 0.06068446821160113,
+      "train_runtime": 4892.2125,
+      "train_samples_per_second": 135.991,
+      "train_steps_per_second": 1.063
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5198,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 800,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5135099285733376.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}