diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,28053 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6809669731018045,
+  "eval_steps": 1000,
+  "global_step": 4000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00017024174327545113,
+      "grad_norm": 5.9435296058654785,
+      "learning_rate": 0.0,
+      "loss": 0.9299,
+      "step": 1
+    },
+    {
+      "epoch": 0.00034048348655090226,
+      "grad_norm": 4.946147918701172,
+      "learning_rate": 1.3391173292339812e-07,
+      "loss": 0.8753,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005107252298263534,
+      "grad_norm": 4.654350280761719,
+      "learning_rate": 2.1224507509017273e-07,
+      "loss": 0.8961,
+      "step": 3
+    },
+    {
+      "epoch": 0.0006809669731018045,
+      "grad_norm": 4.624459266662598,
+      "learning_rate": 2.6782346584679625e-07,
+      "loss": 0.8702,
+      "step": 4
+    },
+    {
+      "epoch": 0.0008512087163772557,
+      "grad_norm": 4.840507984161377,
+      "learning_rate": 3.109334149098911e-07,
+      "loss": 0.9089,
+      "step": 5
+    },
+    {
+      "epoch": 0.0010214504596527069,
+      "grad_norm": 4.609108924865723,
+      "learning_rate": 3.4615680801357083e-07,
+      "loss": 0.8922,
+      "step": 6
+    },
+    {
+      "epoch": 0.001191692202928158,
+      "grad_norm": 4.153110980987549,
+      "learning_rate": 3.759377625437651e-07,
+      "loss": 0.8303,
+      "step": 7
+    },
+    {
+      "epoch": 0.001361933946203609,
+      "grad_norm": 3.97541880607605,
+      "learning_rate": 4.017351987701944e-07,
+      "loss": 0.8569,
+      "step": 8
+    },
+    {
+      "epoch": 0.0015321756894790602,
+      "grad_norm": 3.4270944595336914,
+      "learning_rate": 4.2449015018034546e-07,
+      "loss": 0.7777,
+      "step": 9
+    },
+    {
+      "epoch": 0.0017024174327545114,
+      "grad_norm": 3.244053840637207,
+      "learning_rate": 4.4484514783328935e-07,
+      "loss": 0.7431,
+      "step": 10
+    },
+    {
+      "epoch": 0.0018726591760299626,
+      "grad_norm": 3.140742778778076,
+      "learning_rate": 4.632584829817167e-07,
+      "loss": 0.7605,
+      "step": 11
+    },
+    {
+      "epoch": 0.0020429009193054137,
+      "grad_norm": 3.1321463584899902,
+      "learning_rate": 4.80068540936969e-07,
+      "loss": 0.7243,
+      "step": 12
+    },
+    {
+      "epoch": 0.0022131426625808647,
+      "grad_norm": 2.890293836593628,
+      "learning_rate": 4.955322952348447e-07,
+      "loss": 0.6453,
+      "step": 13
+    },
+    {
+      "epoch": 0.002383384405856316,
+      "grad_norm": 2.773414134979248,
+      "learning_rate": 5.098494954671633e-07,
+      "loss": 0.6187,
+      "step": 14
+    },
+    {
+      "epoch": 0.002553626149131767,
+      "grad_norm": 2.7988698482513428,
+      "learning_rate": 5.231784900000639e-07,
+      "loss": 0.6197,
+      "step": 15
+    },
+    {
+      "epoch": 0.002723867892407218,
+      "grad_norm": 2.7417197227478027,
+      "learning_rate": 5.356469316935925e-07,
+      "loss": 0.6139,
+      "step": 16
+    },
+    {
+      "epoch": 0.0028941096356826694,
+      "grad_norm": 2.771998167037964,
+      "learning_rate": 5.473592323318297e-07,
+      "loss": 0.6009,
+      "step": 17
+    },
+    {
+      "epoch": 0.0030643513789581204,
+      "grad_norm": 3.031147003173828,
+      "learning_rate": 5.584018831037436e-07,
+      "loss": 0.562,
+      "step": 18
+    },
+    {
+      "epoch": 0.003234593122233572,
+      "grad_norm": 2.6269404888153076,
+      "learning_rate": 5.688473346582122e-07,
+      "loss": 0.4996,
+      "step": 19
+    },
+    {
+      "epoch": 0.0034048348655090228,
+      "grad_norm": 2.2291667461395264,
+      "learning_rate": 5.787568807566874e-07,
+      "loss": 0.5131,
+      "step": 20
+    },
+    {
+      "epoch": 0.003575076608784474,
+      "grad_norm": 2.112966299057007,
+      "learning_rate": 5.881828376339378e-07,
+      "loss": 0.4314,
+      "step": 21
+    },
+    {
+      "epoch": 0.003745318352059925,
+      "grad_norm": 2.252114772796631,
+      "learning_rate": 5.971702159051149e-07,
+      "loss": 0.4716,
+      "step": 22
+    },
+    {
+      "epoch": 0.003915560095335376,
+      "grad_norm": 1.981942057609558,
+      "learning_rate": 6.057580205219512e-07,
+      "loss": 0.4206,
+      "step": 23
+    },
+    {
+      "epoch": 0.0040858018386108275,
+      "grad_norm": 1.7979234457015991,
+      "learning_rate": 6.139802738603672e-07,
+      "loss": 0.4061,
+      "step": 24
+    },
+    {
+      "epoch": 0.004256043581886279,
+      "grad_norm": 1.8551833629608154,
+      "learning_rate": 6.218668298197822e-07,
+      "loss": 0.427,
+      "step": 25
+    },
+    {
+      "epoch": 0.004426285325161729,
+      "grad_norm": 1.7879689931869507,
+      "learning_rate": 6.294440281582428e-07,
+      "loss": 0.4368,
+      "step": 26
+    },
+    {
+      "epoch": 0.004596527068437181,
+      "grad_norm": 1.7607426643371582,
+      "learning_rate": 6.367352252705181e-07,
+      "loss": 0.3931,
+      "step": 27
+    },
+    {
+      "epoch": 0.004766768811712632,
+      "grad_norm": 1.5092447996139526,
+      "learning_rate": 6.437612283905613e-07,
+      "loss": 0.3544,
+      "step": 28
+    },
+    {
+      "epoch": 0.004937010554988083,
+      "grad_norm": 2.1652700901031494,
+      "learning_rate": 6.505406535664675e-07,
+      "loss": 0.3799,
+      "step": 29
+    },
+    {
+      "epoch": 0.005107252298263534,
+      "grad_norm": 1.570780634880066,
+      "learning_rate": 6.57090222923462e-07,
+      "loss": 0.3741,
+      "step": 30
+    },
+    {
+      "epoch": 0.0052774940415389856,
+      "grad_norm": 1.539757251739502,
+      "learning_rate": 6.634250131666118e-07,
+      "loss": 0.3894,
+      "step": 31
+    },
+    {
+      "epoch": 0.005447735784814436,
+      "grad_norm": 1.5071591138839722,
+      "learning_rate": 6.695586646169908e-07,
+      "loss": 0.3559,
+      "step": 32
+    },
+    {
+      "epoch": 0.0056179775280898875,
+      "grad_norm": 1.2861617803573608,
+      "learning_rate": 6.755035580718894e-07,
+      "loss": 0.2831,
+      "step": 33
+    },
+    {
+      "epoch": 0.005788219271365339,
+      "grad_norm": 1.2855749130249023,
+      "learning_rate": 6.812709652552277e-07,
+      "loss": 0.3348,
+      "step": 34
+    },
+    {
+      "epoch": 0.00595846101464079,
+      "grad_norm": 1.649814248085022,
+      "learning_rate": 6.868711774536562e-07,
+      "loss": 0.3381,
+      "step": 35
+    },
+    {
+      "epoch": 0.006128702757916241,
+      "grad_norm": 1.4062033891677856,
+      "learning_rate": 6.923136160271417e-07,
+      "loss": 0.343,
+      "step": 36
+    },
+    {
+      "epoch": 0.006298944501191692,
+      "grad_norm": 1.3435696363449097,
+      "learning_rate": 6.976069277750015e-07,
+      "loss": 0.3239,
+      "step": 37
+    },
+    {
+      "epoch": 0.006469186244467144,
+      "grad_norm": 1.419482946395874,
+      "learning_rate": 7.027590675816104e-07,
+      "loss": 0.3034,
+      "step": 38
+    },
+    {
+      "epoch": 0.006639427987742594,
+      "grad_norm": 1.3096425533294678,
+      "learning_rate": 7.077773703250174e-07,
+      "loss": 0.3282,
+      "step": 39
+    },
+    {
+      "epoch": 0.0068096697310180455,
+      "grad_norm": 1.3494988679885864,
+      "learning_rate": 7.126686136800855e-07,
+      "loss": 0.3264,
+      "step": 40
+    },
+    {
+      "epoch": 0.006979911474293497,
+      "grad_norm": 1.2407349348068237,
+      "learning_rate": 7.174390731656332e-07,
+      "loss": 0.3306,
+      "step": 41
+    },
+    {
+      "epoch": 0.007150153217568948,
+      "grad_norm": 1.4115320444107056,
+      "learning_rate": 7.220945705573361e-07,
+      "loss": 0.3093,
+      "step": 42
+    },
+    {
+      "epoch": 0.007320394960844399,
+      "grad_norm": 1.4195406436920166,
+      "learning_rate": 7.266405166033159e-07,
+      "loss": 0.3102,
+      "step": 43
+    },
+    {
+      "epoch": 0.00749063670411985,
+      "grad_norm": 1.1780062913894653,
+      "learning_rate": 7.31081948828513e-07,
+      "loss": 0.3072,
+      "step": 44
+    },
+    {
+      "epoch": 0.007660878447395302,
+      "grad_norm": 1.3273332118988037,
+      "learning_rate": 7.354235650902366e-07,
+      "loss": 0.3161,
+      "step": 45
+    },
+    {
+      "epoch": 0.007831120190670752,
+      "grad_norm": 1.2345938682556152,
+      "learning_rate": 7.396697534453494e-07,
+      "loss": 0.312,
+      "step": 46
+    },
+    {
+      "epoch": 0.008001361933946204,
+      "grad_norm": 1.327564001083374,
+      "learning_rate": 7.438246188051406e-07,
+      "loss": 0.3177,
+      "step": 47
+    },
+    {
+      "epoch": 0.008171603677221655,
+      "grad_norm": 1.195912480354309,
+      "learning_rate": 7.478920067837654e-07,
+      "loss": 0.3031,
+      "step": 48
+    },
+    {
+      "epoch": 0.008341845420497106,
+      "grad_norm": 1.3030645847320557,
+      "learning_rate": 7.518755250875302e-07,
+      "loss": 0.3009,
+      "step": 49
+    },
+    {
+      "epoch": 0.008512087163772558,
+      "grad_norm": 1.2128993272781372,
+      "learning_rate": 7.557785627431804e-07,
+      "loss": 0.3139,
+      "step": 50
+    },
+    {
+      "epoch": 0.008682328907048007,
+      "grad_norm": 1.1837964057922363,
+      "learning_rate": 7.596043074220024e-07,
+      "loss": 0.2816,
+      "step": 51
+    },
+    {
+      "epoch": 0.008852570650323459,
+      "grad_norm": 1.1090530157089233,
+      "learning_rate": 7.633557610816411e-07,
+      "loss": 0.2451,
+      "step": 52
+    },
+    {
+      "epoch": 0.00902281239359891,
+      "grad_norm": 1.2700027227401733,
+      "learning_rate": 7.670357541179365e-07,
+      "loss": 0.2702,
+      "step": 53
+    },
+    {
+      "epoch": 0.009193054136874362,
+      "grad_norm": 1.1058295965194702,
+      "learning_rate": 7.706469581939163e-07,
+      "loss": 0.2632,
+      "step": 54
+    },
+    {
+      "epoch": 0.009363295880149813,
+      "grad_norm": 1.152600884437561,
+      "learning_rate": 7.741918978916079e-07,
+      "loss": 0.2693,
+      "step": 55
+    },
+    {
+      "epoch": 0.009533537623425264,
+      "grad_norm": 1.1038074493408203,
+      "learning_rate": 7.776729613139597e-07,
+      "loss": 0.2361,
+      "step": 56
+    },
+    {
+      "epoch": 0.009703779366700716,
+      "grad_norm": 1.298447847366333,
+      "learning_rate": 7.81092409748385e-07,
+      "loss": 0.2814,
+      "step": 57
+    },
+    {
+      "epoch": 0.009874021109976166,
+      "grad_norm": 1.1311403512954712,
+      "learning_rate": 7.844523864898656e-07,
+      "loss": 0.2706,
+      "step": 58
+    },
+    {
+      "epoch": 0.010044262853251617,
+      "grad_norm": 1.2915065288543701,
+      "learning_rate": 7.877549249098274e-07,
+      "loss": 0.2774,
+      "step": 59
+    },
+    {
+      "epoch": 0.010214504596527068,
+      "grad_norm": 1.0150870084762573,
+      "learning_rate": 7.910019558468602e-07,
+      "loss": 0.2325,
+      "step": 60
+    },
+    {
+      "epoch": 0.01038474633980252,
+      "grad_norm": 1.2452411651611328,
+      "learning_rate": 7.941953143865467e-07,
+      "loss": 0.2873,
+      "step": 61
+    },
+    {
+      "epoch": 0.010554988083077971,
+      "grad_norm": 1.3862929344177246,
+      "learning_rate": 7.973367460900099e-07,
+      "loss": 0.3041,
+      "step": 62
+    },
+    {
+      "epoch": 0.010725229826353423,
+      "grad_norm": 1.0656795501708984,
+      "learning_rate": 8.004279127241105e-07,
+      "loss": 0.2215,
+      "step": 63
+    },
+    {
+      "epoch": 0.010895471569628872,
+      "grad_norm": 1.3162792921066284,
+      "learning_rate": 8.034703975403888e-07,
+      "loss": 0.3019,
+      "step": 64
+    },
+    {
+      "epoch": 0.011065713312904324,
+      "grad_norm": 1.0575990676879883,
+      "learning_rate": 8.064657101447357e-07,
+      "loss": 0.2394,
+      "step": 65
+    },
+    {
+      "epoch": 0.011235955056179775,
+      "grad_norm": 1.0472103357315063,
+      "learning_rate": 8.094152909952874e-07,
+      "loss": 0.246,
+      "step": 66
+    },
+    {
+      "epoch": 0.011406196799455226,
+      "grad_norm": 1.1870509386062622,
+      "learning_rate": 8.123205155620937e-07,
+      "loss": 0.2323,
+      "step": 67
+    },
+    {
+      "epoch": 0.011576438542730678,
+      "grad_norm": 1.0447919368743896,
+      "learning_rate": 8.15182698178626e-07,
+      "loss": 0.2372,
+      "step": 68
+    },
+    {
+      "epoch": 0.01174668028600613,
+      "grad_norm": 1.3799355030059814,
+      "learning_rate": 8.18003095612124e-07,
+      "loss": 0.2981,
+      "step": 69
+    },
+    {
+      "epoch": 0.01191692202928158,
+      "grad_norm": 1.3175947666168213,
+      "learning_rate": 8.207829103770545e-07,
+      "loss": 0.2427,
+      "step": 70
+    },
+    {
+      "epoch": 0.01208716377255703,
+      "grad_norm": 1.8826605081558228,
+      "learning_rate": 8.235232938135481e-07,
+      "loss": 0.2829,
+      "step": 71
+    },
+    {
+      "epoch": 0.012257405515832482,
+      "grad_norm": 1.0391643047332764,
+      "learning_rate": 8.262253489505398e-07,
+      "loss": 0.2228,
+      "step": 72
+    },
+    {
+      "epoch": 0.012427647259107933,
+      "grad_norm": 0.9970806241035461,
+      "learning_rate": 8.288901331714316e-07,
+      "loss": 0.2154,
+      "step": 73
+    },
+    {
+      "epoch": 0.012597889002383384,
+      "grad_norm": 1.1637834310531616,
+      "learning_rate": 8.315186606983998e-07,
+      "loss": 0.2596,
+      "step": 74
+    },
+    {
+      "epoch": 0.012768130745658836,
+      "grad_norm": 1.374963402748108,
+      "learning_rate": 8.34111904909955e-07,
+      "loss": 0.2749,
+      "step": 75
+    },
+    {
+      "epoch": 0.012938372488934287,
+      "grad_norm": 1.0730386972427368,
+      "learning_rate": 8.366708005050085e-07,
+      "loss": 0.2227,
+      "step": 76
+    },
+    {
+      "epoch": 0.013108614232209739,
+      "grad_norm": 1.1448426246643066,
+      "learning_rate": 8.391962455254819e-07,
+      "loss": 0.2361,
+      "step": 77
+    },
+    {
+      "epoch": 0.013278855975485188,
+      "grad_norm": 1.2940086126327515,
+      "learning_rate": 8.416891032484155e-07,
+      "loss": 0.3044,
+      "step": 78
+    },
+    {
+      "epoch": 0.01344909771876064,
+      "grad_norm": 1.3453335762023926,
+      "learning_rate": 8.441502039575513e-07,
+      "loss": 0.3157,
+      "step": 79
+    },
+    {
+      "epoch": 0.013619339462036091,
+      "grad_norm": 1.1563291549682617,
+      "learning_rate": 8.465803466034837e-07,
+      "loss": 0.2501,
+      "step": 80
+    },
+    {
+      "epoch": 0.013789581205311542,
+      "grad_norm": 1.1329809427261353,
+      "learning_rate": 8.489803003606909e-07,
+      "loss": 0.2429,
+      "step": 81
+    },
+    {
+      "epoch": 0.013959822948586994,
+      "grad_norm": 1.0661827325820923,
+      "learning_rate": 8.513508060890314e-07,
+      "loss": 0.2232,
+      "step": 82
+    },
+    {
+      "epoch": 0.014130064691862445,
+      "grad_norm": 1.1219416856765747,
+      "learning_rate": 8.536925777066614e-07,
+      "loss": 0.2142,
+      "step": 83
+    },
+    {
+      "epoch": 0.014300306435137897,
+      "grad_norm": 1.2325749397277832,
+      "learning_rate": 8.560063034807341e-07,
+      "loss": 0.2122,
+      "step": 84
+    },
+    {
+      "epoch": 0.014470548178413346,
+      "grad_norm": 1.0120998620986938,
+      "learning_rate": 8.582926472417208e-07,
+      "loss": 0.2147,
+      "step": 85
+    },
+    {
+      "epoch": 0.014640789921688798,
+      "grad_norm": 1.0740280151367188,
+      "learning_rate": 8.605522495267141e-07,
+      "loss": 0.2271,
+      "step": 86
+    },
+    {
+      "epoch": 0.01481103166496425,
+      "grad_norm": 1.309480905532837,
+      "learning_rate": 8.627857286566401e-07,
+      "loss": 0.2563,
+      "step": 87
+    },
+    {
+      "epoch": 0.0149812734082397,
+      "grad_norm": 1.108237385749817,
+      "learning_rate": 8.649936817519112e-07,
+      "loss": 0.2231,
+      "step": 88
+    },
+    {
+      "epoch": 0.015151515151515152,
+      "grad_norm": 1.1064398288726807,
+      "learning_rate": 8.671766856906931e-07,
+      "loss": 0.2516,
+      "step": 89
+    },
+    {
+      "epoch": 0.015321756894790603,
+      "grad_norm": 1.2427948713302612,
+      "learning_rate": 8.693352980136347e-07,
+      "loss": 0.2604,
+      "step": 90
+    },
+    {
+      "epoch": 0.015491998638066053,
+      "grad_norm": 1.1325477361679077,
+      "learning_rate": 8.714700577786097e-07,
+      "loss": 0.2314,
+      "step": 91
+    },
+    {
+      "epoch": 0.015662240381341504,
+      "grad_norm": 0.9738770127296448,
+      "learning_rate": 8.735814863687475e-07,
+      "loss": 0.2126,
+      "step": 92
+    },
+    {
+      "epoch": 0.015832482124616958,
+      "grad_norm": 1.1820428371429443,
+      "learning_rate": 8.756700882567846e-07,
+      "loss": 0.2116,
+      "step": 93
+    },
+    {
+      "epoch": 0.016002723867892407,
+      "grad_norm": 1.0578908920288086,
+      "learning_rate": 8.777363517285388e-07,
+      "loss": 0.2284,
+      "step": 94
+    },
+    {
+      "epoch": 0.016172965611167857,
+      "grad_norm": 1.2253124713897705,
+      "learning_rate": 8.797807495681034e-07,
+      "loss": 0.2492,
+      "step": 95
+    },
+    {
+      "epoch": 0.01634320735444331,
+      "grad_norm": 1.4362291097640991,
+      "learning_rate": 8.818037397071634e-07,
+      "loss": 0.2468,
+      "step": 96
+    },
+    {
+      "epoch": 0.01651344909771876,
+      "grad_norm": 1.2666908502578735,
+      "learning_rate": 8.838057658406682e-07,
+      "loss": 0.2349,
+      "step": 97
+    },
+    {
+      "epoch": 0.016683690840994213,
+      "grad_norm": 1.0576752424240112,
+      "learning_rate": 8.857872580109284e-07,
+      "loss": 0.2142,
+      "step": 98
+    },
+    {
+      "epoch": 0.016853932584269662,
+      "grad_norm": 0.937912106513977,
+      "learning_rate": 8.877486331620622e-07,
+      "loss": 0.1829,
+      "step": 99
+    },
+    {
+      "epoch": 0.017024174327545116,
+      "grad_norm": 1.2614257335662842,
+      "learning_rate": 8.896902956665787e-07,
+      "loss": 0.2405,
+      "step": 100
+    },
+    {
+      "epoch": 0.017194416070820565,
+      "grad_norm": 1.2120734453201294,
+      "learning_rate": 8.916126378257612e-07,
+      "loss": 0.2347,
+      "step": 101
+    },
+    {
+      "epoch": 0.017364657814096015,
+      "grad_norm": 1.108526349067688,
+      "learning_rate": 8.935160403454004e-07,
+      "loss": 0.1874,
+      "step": 102
+    },
+    {
+      "epoch": 0.017534899557371468,
+      "grad_norm": 1.0109800100326538,
+      "learning_rate": 8.954008727883201e-07,
+      "loss": 0.2053,
+      "step": 103
+    },
+    {
+      "epoch": 0.017705141300646918,
+      "grad_norm": 1.0737659931182861,
+      "learning_rate": 8.972674940050391e-07,
+      "loss": 0.2015,
+      "step": 104
+    },
+    {
+      "epoch": 0.01787538304392237,
+      "grad_norm": 1.0084854364395142,
+      "learning_rate": 8.991162525438289e-07,
+      "loss": 0.178,
+      "step": 105
+    },
+    {
+      "epoch": 0.01804562478719782,
+      "grad_norm": 1.0313851833343506,
+      "learning_rate": 9.009474870413346e-07,
+      "loss": 0.1977,
+      "step": 106
+    },
+    {
+      "epoch": 0.018215866530473274,
+      "grad_norm": 1.1986593008041382,
+      "learning_rate": 9.02761526594856e-07,
+      "loss": 0.2481,
+      "step": 107
+    },
+    {
+      "epoch": 0.018386108273748723,
+      "grad_norm": 1.0787945985794067,
+      "learning_rate": 9.045586911173146e-07,
+      "loss": 0.2128,
+      "step": 108
+    },
+    {
+      "epoch": 0.018556350017024173,
+      "grad_norm": 1.36089289188385,
+      "learning_rate": 9.063392916758576e-07,
+      "loss": 0.2232,
+      "step": 109
+    },
+    {
+      "epoch": 0.018726591760299626,
+      "grad_norm": 1.1465134620666504,
+      "learning_rate": 9.081036308150061e-07,
+      "loss": 0.2174,
+      "step": 110
+    },
+    {
+      "epoch": 0.018896833503575076,
+      "grad_norm": 1.1831120252609253,
+      "learning_rate": 9.098520028651742e-07,
+      "loss": 0.2266,
+      "step": 111
+    },
+    {
+      "epoch": 0.01906707524685053,
+      "grad_norm": 1.2039859294891357,
+      "learning_rate": 9.115846942373576e-07,
+      "loss": 0.1919,
+      "step": 112
+    },
+    {
+      "epoch": 0.01923731699012598,
+      "grad_norm": 1.119031548500061,
+      "learning_rate": 9.133019837047214e-07,
+      "loss": 0.2067,
+      "step": 113
+    },
+    {
+      "epoch": 0.01940755873340143,
+      "grad_norm": 1.1545226573944092,
+      "learning_rate": 9.150041426717831e-07,
+      "loss": 0.2066,
+      "step": 114
+    },
+    {
+      "epoch": 0.01957780047667688,
+      "grad_norm": 1.141434669494629,
+      "learning_rate": 9.166914354318424e-07,
+      "loss": 0.2301,
+      "step": 115
+    },
+    {
+      "epoch": 0.01974804221995233,
+      "grad_norm": 0.9759832620620728,
+      "learning_rate": 9.183641194132636e-07,
+      "loss": 0.1774,
+      "step": 116
+    },
+    {
+      "epoch": 0.019918283963227784,
+      "grad_norm": 1.1361424922943115,
+      "learning_rate": 9.200224454151901e-07,
+      "loss": 0.2169,
+      "step": 117
+    },
+    {
+      "epoch": 0.020088525706503234,
+      "grad_norm": 1.0914556980133057,
+      "learning_rate": 9.216666578332256e-07,
+      "loss": 0.188,
+      "step": 118
+    },
+    {
+      "epoch": 0.020258767449778687,
+      "grad_norm": 1.4587888717651367,
+      "learning_rate": 9.232969948755948e-07,
+      "loss": 0.2582,
+      "step": 119
+    },
+    {
+      "epoch": 0.020429009193054137,
+      "grad_norm": 1.1200069189071655,
+      "learning_rate": 9.249136887702583e-07,
+      "loss": 0.2058,
+      "step": 120
+    },
+    {
+      "epoch": 0.020599250936329586,
+      "grad_norm": 1.232216715812683,
+      "learning_rate": 9.265169659634334e-07,
+      "loss": 0.2383,
+      "step": 121
+    },
+    {
+      "epoch": 0.02076949267960504,
+      "grad_norm": 1.1444164514541626,
+      "learning_rate": 9.281070473099448e-07,
+      "loss": 0.2153,
+      "step": 122
+    },
+    {
+      "epoch": 0.02093973442288049,
+      "grad_norm": 1.0245447158813477,
+      "learning_rate": 9.296841482558059e-07,
+      "loss": 0.1701,
+      "step": 123
+    },
+    {
+      "epoch": 0.021109976166155942,
+      "grad_norm": 1.1672013998031616,
+      "learning_rate": 9.312484790134081e-07,
+      "loss": 0.2171,
+      "step": 124
+    },
+    {
+      "epoch": 0.021280217909431392,
+      "grad_norm": 1.001262903213501,
+      "learning_rate": 9.328002447296736e-07,
+      "loss": 0.1813,
+      "step": 125
+    },
+    {
+      "epoch": 0.021450459652706845,
+      "grad_norm": 1.345151662826538,
+      "learning_rate": 9.343396456475086e-07,
+      "loss": 0.2214,
+      "step": 126
+    },
+    {
+      "epoch": 0.021620701395982295,
+      "grad_norm": 1.1059603691101074,
+      "learning_rate": 9.358668772608768e-07,
+      "loss": 0.2258,
+      "step": 127
+    },
+    {
+      "epoch": 0.021790943139257744,
+      "grad_norm": 1.2178932428359985,
+      "learning_rate": 9.37382130463787e-07,
+      "loss": 0.1938,
+      "step": 128
+    },
+    {
+      "epoch": 0.021961184882533197,
+      "grad_norm": 1.2222176790237427,
+      "learning_rate": 9.388855916934886e-07,
+      "loss": 0.2357,
+      "step": 129
+    },
+    {
+      "epoch": 0.022131426625808647,
+      "grad_norm": 1.1123160123825073,
+      "learning_rate": 9.403774430681339e-07,
+      "loss": 0.1898,
+      "step": 130
+    },
+    {
+      "epoch": 0.0223016683690841,
+      "grad_norm": 1.3277643918991089,
+      "learning_rate": 9.418578625191684e-07,
+      "loss": 0.2397,
+      "step": 131
+    },
+    {
+      "epoch": 0.02247191011235955,
+      "grad_norm": 1.0972932577133179,
+      "learning_rate": 9.433270239186857e-07,
+      "loss": 0.187,
+      "step": 132
+    },
+    {
+      "epoch": 0.022642151855635003,
+      "grad_norm": 1.1459786891937256,
+      "learning_rate": 9.447850972019773e-07,
+      "loss": 0.1931,
+      "step": 133
+    },
+    {
+      "epoch": 0.022812393598910453,
+      "grad_norm": 1.1958963871002197,
+      "learning_rate": 9.462322484854918e-07,
+      "loss": 0.2066,
+      "step": 134
+    },
+    {
+      "epoch": 0.022982635342185902,
+      "grad_norm": 1.410596251487732,
+      "learning_rate": 9.476686401804093e-07,
+      "loss": 0.1863,
+      "step": 135
+    },
+    {
+      "epoch": 0.023152877085461356,
+      "grad_norm": 1.1559172868728638,
+      "learning_rate": 9.490944311020241e-07,
+      "loss": 0.1947,
+      "step": 136
+    },
+    {
+      "epoch": 0.023323118828736805,
+      "grad_norm": 1.2141169309616089,
+      "learning_rate": 9.505097765751215e-07,
+      "loss": 0.2312,
+      "step": 137
+    },
+    {
+      "epoch": 0.02349336057201226,
+      "grad_norm": 1.081640362739563,
+      "learning_rate": 9.519148285355222e-07,
+      "loss": 0.204,
+      "step": 138
+    },
+    {
+      "epoch": 0.023663602315287708,
+      "grad_norm": 1.086338996887207,
+      "learning_rate": 9.533097356279598e-07,
+      "loss": 0.1895,
+      "step": 139
+    },
+    {
+      "epoch": 0.02383384405856316,
+      "grad_norm": 1.0459181070327759,
+      "learning_rate": 9.546946433004524e-07,
+      "loss": 0.1944,
+      "step": 140
+    },
+    {
+      "epoch": 0.02400408580183861,
+      "grad_norm": 1.192455530166626,
+      "learning_rate": 9.560696938953133e-07,
+      "loss": 0.1916,
+      "step": 141
+    },
+    {
+      "epoch": 0.02417432754511406,
+      "grad_norm": 1.0824813842773438,
+      "learning_rate": 9.574350267369461e-07,
+      "loss": 0.1585,
+      "step": 142
+    },
+    {
+      "epoch": 0.024344569288389514,
+      "grad_norm": 1.211179494857788,
+      "learning_rate": 9.587907782165614e-07,
+      "loss": 0.1878,
+      "step": 143
+    },
+    {
+      "epoch": 0.024514811031664963,
+      "grad_norm": 1.272672176361084,
+      "learning_rate": 9.60137081873938e-07,
+      "loss": 0.1899,
+      "step": 144
+    },
+    {
+      "epoch": 0.024685052774940416,
+      "grad_norm": 1.2296862602233887,
+      "learning_rate": 9.614740684763584e-07,
+      "loss": 0.1851,
+      "step": 145
+    },
+    {
+      "epoch": 0.024855294518215866,
+      "grad_norm": 1.2672345638275146,
+      "learning_rate": 9.628018660948297e-07,
+      "loss": 0.2202,
+      "step": 146
+    },
+    {
+      "epoch": 0.02502553626149132,
+      "grad_norm": 1.336018443107605,
+      "learning_rate": 9.641206001777028e-07,
+      "loss": 0.2084,
+      "step": 147
+    },
+    {
+      "epoch": 0.02519577800476677,
+      "grad_norm": 1.0895048379898071,
+      "learning_rate": 9.654303936217977e-07,
+      "loss": 0.1939,
+      "step": 148
+    },
+    {
+      "epoch": 0.02536601974804222,
+      "grad_norm": 1.124503493309021,
+      "learning_rate": 9.667313668411324e-07,
+      "loss": 0.186,
+      "step": 149
+    },
+    {
+      "epoch": 0.02553626149131767,
+      "grad_norm": 1.1752759218215942,
+      "learning_rate": 9.680236378333531e-07,
+      "loss": 0.1991,
+      "step": 150
+    },
+    {
+      "epoch": 0.02570650323459312,
+      "grad_norm": 1.1944782733917236,
+      "learning_rate": 9.693073222439593e-07,
+      "loss": 0.2166,
+      "step": 151
+    },
+    {
+      "epoch": 0.025876744977868574,
+      "grad_norm": 1.1967934370040894,
+      "learning_rate": 9.705825334284067e-07,
+      "loss": 0.2066,
+      "step": 152
+    },
+    {
+      "epoch": 0.026046986721144024,
+      "grad_norm": 1.2898709774017334,
+      "learning_rate": 9.71849382512175e-07,
+      "loss": 0.2167,
+      "step": 153
+    },
+    {
+      "epoch": 0.026217228464419477,
+      "grad_norm": 1.0959877967834473,
+      "learning_rate": 9.7310797844888e-07,
+      "loss": 0.1884,
+      "step": 154
+    },
+    {
+      "epoch": 0.026387470207694927,
+      "grad_norm": 1.192084789276123,
+      "learning_rate": 9.743584280765029e-07,
+      "loss": 0.2111,
+      "step": 155
+    },
+    {
+      "epoch": 0.026557711950970377,
+      "grad_norm": 1.4502787590026855,
+      "learning_rate": 9.756008361718137e-07,
+      "loss": 0.2387,
+      "step": 156
+    },
+    {
+      "epoch": 0.02672795369424583,
+      "grad_norm": 1.3559253215789795,
+      "learning_rate": 9.76835305503054e-07,
+      "loss": 0.2201,
+      "step": 157
+    },
+    {
+      "epoch": 0.02689819543752128,
+      "grad_norm": 1.1862843036651611,
+      "learning_rate": 9.780619368809492e-07,
+      "loss": 0.1881,
+      "step": 158
+    },
+    {
+      "epoch": 0.027068437180796732,
+      "grad_norm": 1.2608420848846436,
+      "learning_rate": 9.792808292081091e-07,
+      "loss": 0.2009,
+      "step": 159
+    },
+    {
+      "epoch": 0.027238678924072182,
+      "grad_norm": 1.3061561584472656,
+      "learning_rate": 9.804920795268817e-07,
+      "loss": 0.2324,
+      "step": 160
+    },
+    {
+      "epoch": 0.027408920667347635,
+      "grad_norm": 1.3358960151672363,
+      "learning_rate": 9.816957830657163e-07,
+      "loss": 0.2116,
+      "step": 161
+    },
+    {
+      "epoch": 0.027579162410623085,
+      "grad_norm": 1.0722428560256958,
+      "learning_rate": 9.828920332840889e-07,
+      "loss": 0.1641,
+      "step": 162
+    },
+    {
+      "epoch": 0.027749404153898535,
+      "grad_norm": 1.2872867584228516,
+      "learning_rate": 9.840809219160487e-07,
+      "loss": 0.1849,
+      "step": 163
+    },
+    {
+      "epoch": 0.027919645897173988,
+      "grad_norm": 1.3819077014923096,
+      "learning_rate": 9.852625390124294e-07,
+      "loss": 0.1933,
+      "step": 164
+    },
+    {
+      "epoch": 0.028089887640449437,
+      "grad_norm": 1.278441309928894,
+      "learning_rate": 9.864369729817805e-07,
+      "loss": 0.1835,
+      "step": 165
+    },
+    {
+      "epoch": 0.02826012938372489,
+      "grad_norm": 1.1808550357818604,
+      "learning_rate": 9.876043106300596e-07,
+      "loss": 0.166,
+      "step": 166
+    },
+    {
+      "epoch": 0.02843037112700034,
+      "grad_norm": 1.233665108680725,
+      "learning_rate": 9.887646371991337e-07,
+      "loss": 0.1918,
+      "step": 167
+    },
+    {
+      "epoch": 0.028600612870275793,
+      "grad_norm": 1.1117044687271118,
+      "learning_rate": 9.899180364041324e-07,
+      "loss": 0.1699,
+      "step": 168
+    },
+    {
+      "epoch": 0.028770854613551243,
+      "grad_norm": 1.112967848777771,
+      "learning_rate": 9.910645904696893e-07,
+      "loss": 0.1629,
+      "step": 169
+    },
+    {
+      "epoch": 0.028941096356826693,
+      "grad_norm": 1.1570054292678833,
+      "learning_rate": 9.92204380165119e-07,
+      "loss": 0.1729,
+      "step": 170
+    },
+    {
+      "epoch": 0.029111338100102146,
+      "grad_norm": 1.1414086818695068,
+      "learning_rate": 9.933374848385576e-07,
+      "loss": 0.1659,
+      "step": 171
+    },
+    {
+      "epoch": 0.029281579843377595,
+      "grad_norm": 1.2004859447479248,
+      "learning_rate": 9.944639824501122e-07,
+      "loss": 0.1835,
+      "step": 172
+    },
+    {
+      "epoch": 0.02945182158665305,
+      "grad_norm": 1.2365108728408813,
+      "learning_rate": 9.95583949604046e-07,
+      "loss": 0.2133,
+      "step": 173
+    },
+    {
+      "epoch": 0.0296220633299285,
+      "grad_norm": 1.2216784954071045,
+      "learning_rate": 9.966974615800383e-07,
+      "loss": 0.1848,
+      "step": 174
+    },
+    {
+      "epoch": 0.029792305073203948,
+      "grad_norm": 1.2305302619934082,
+      "learning_rate": 9.978045923635475e-07,
+      "loss": 0.2026,
+      "step": 175
+    },
+    {
+      "epoch": 0.0299625468164794,
+      "grad_norm": 1.3286123275756836,
+      "learning_rate": 9.989054146753091e-07,
+      "loss": 0.1817,
+      "step": 176
+    },
+    {
+      "epoch": 0.03013278855975485,
+      "grad_norm": 1.3320600986480713,
+      "learning_rate": 1e-06,
+      "loss": 0.2147,
+      "step": 177
+    },
+    {
+      "epoch": 0.030303030303030304,
+      "grad_norm": 1.2844750881195068,
+      "learning_rate": 1e-06,
+      "loss": 0.1893,
+      "step": 178
+    },
+    {
+      "epoch": 0.030473272046305754,
+      "grad_norm": 1.1248033046722412,
+      "learning_rate": 1e-06,
+      "loss": 0.1617,
+      "step": 179
+    },
+    {
+      "epoch": 0.030643513789581207,
+      "grad_norm": 1.3026221990585327,
+      "learning_rate": 1e-06,
+      "loss": 0.1815,
+      "step": 180
+    },
+    {
+      "epoch": 0.030813755532856656,
+      "grad_norm": 1.2536060810089111,
+      "learning_rate": 1e-06,
+      "loss": 0.1852,
+      "step": 181
+    },
+    {
+      "epoch": 0.030983997276132106,
+      "grad_norm": 1.1810585260391235,
+      "learning_rate": 1e-06,
+      "loss": 0.1712,
+      "step": 182
+    },
+    {
+      "epoch": 0.03115423901940756,
+      "grad_norm": 1.4999784231185913,
+      "learning_rate": 1e-06,
+      "loss": 0.2461,
+      "step": 183
+    },
+    {
+      "epoch": 0.03132448076268301,
+      "grad_norm": 1.1074563264846802,
+      "learning_rate": 1e-06,
+      "loss": 0.1796,
+      "step": 184
+    },
+    {
+      "epoch": 0.03149472250595846,
+      "grad_norm": 1.0583021640777588,
+      "learning_rate": 1e-06,
+      "loss": 0.1603,
+      "step": 185
+    },
+    {
+      "epoch": 0.031664964249233915,
+      "grad_norm": 1.1220263242721558,
+      "learning_rate": 1e-06,
+      "loss": 0.1732,
+      "step": 186
+    },
+    {
+      "epoch": 0.031835205992509365,
+      "grad_norm": 1.1038565635681152,
+      "learning_rate": 1e-06,
+      "loss": 0.1766,
+      "step": 187
+    },
+    {
+      "epoch": 0.032005447735784814,
+      "grad_norm": 1.2731438875198364,
+      "learning_rate": 1e-06,
+      "loss": 0.1993,
+      "step": 188
+    },
+    {
+      "epoch": 0.032175689479060264,
+      "grad_norm": 1.569389820098877,
+      "learning_rate": 1e-06,
+      "loss": 0.1862,
+      "step": 189
+    },
+    {
+      "epoch": 0.032345931222335714,
+      "grad_norm": 1.1800955533981323,
+      "learning_rate": 1e-06,
+      "loss": 0.1411,
+      "step": 190
+    },
+    {
+      "epoch": 0.03251617296561117,
+      "grad_norm": 1.2948437929153442,
+      "learning_rate": 1e-06,
+      "loss": 0.186,
+      "step": 191
+    },
+    {
+      "epoch": 0.03268641470888662,
+      "grad_norm": 1.4020929336547852,
+      "learning_rate": 1e-06,
+      "loss": 0.1837,
+      "step": 192
+    },
+    {
+      "epoch": 0.03285665645216207,
+      "grad_norm": 1.25234055519104,
+      "learning_rate": 1e-06,
+      "loss": 0.1645,
+      "step": 193
+    },
+    {
+      "epoch": 0.03302689819543752,
+      "grad_norm": 1.349990725517273,
+      "learning_rate": 1e-06,
+      "loss": 0.2217,
+      "step": 194
+    },
+    {
+      "epoch": 0.03319713993871297,
+      "grad_norm": 1.2276535034179688,
+      "learning_rate": 1e-06,
+      "loss": 0.1825,
+      "step": 195
+    },
+    {
+      "epoch": 0.033367381681988426,
+      "grad_norm": 1.2954577207565308,
+      "learning_rate": 1e-06,
+      "loss": 0.1811,
+      "step": 196
+    },
+    {
+      "epoch": 0.033537623425263875,
+      "grad_norm": 1.2191767692565918,
+      "learning_rate": 1e-06,
+      "loss": 0.195,
+      "step": 197
+    },
+    {
+      "epoch": 0.033707865168539325,
+      "grad_norm": 1.1188750267028809,
+      "learning_rate": 1e-06,
+      "loss": 0.1505,
+      "step": 198
+    },
+    {
+      "epoch": 0.033878106911814775,
+      "grad_norm": 1.246447205543518,
+      "learning_rate": 1e-06,
+      "loss": 0.1615,
+      "step": 199
+    },
+    {
+      "epoch": 0.03404834865509023,
+      "grad_norm": 1.1524237394332886,
+      "learning_rate": 1e-06,
+      "loss": 0.1858,
+      "step": 200
+    },
+    {
+      "epoch": 0.03421859039836568,
+      "grad_norm": 1.1863471269607544,
+      "learning_rate": 1e-06,
+      "loss": 0.142,
+      "step": 201
+    },
+    {
+      "epoch": 0.03438883214164113,
+      "grad_norm": 1.2720561027526855,
+      "learning_rate": 1e-06,
+      "loss": 0.1689,
+      "step": 202
+    },
+    {
+      "epoch": 0.03455907388491658,
+      "grad_norm": 1.3305280208587646,
+      "learning_rate": 1e-06,
+      "loss": 0.1679,
+      "step": 203
+    },
+    {
+      "epoch": 0.03472931562819203,
+      "grad_norm": 1.2643027305603027,
+      "learning_rate": 1e-06,
+      "loss": 0.1906,
+      "step": 204
+    },
+    {
+      "epoch": 0.034899557371467486,
+      "grad_norm": 1.2647238969802856,
+      "learning_rate": 1e-06,
+      "loss": 0.192,
+      "step": 205
+    },
+    {
+      "epoch": 0.035069799114742936,
+      "grad_norm": 1.4146193265914917,
+      "learning_rate": 1e-06,
+      "loss": 0.1747,
+      "step": 206
+    },
+    {
+      "epoch": 0.035240040858018386,
+      "grad_norm": 1.366545557975769,
+      "learning_rate": 1e-06,
+      "loss": 0.182,
+      "step": 207
+    },
+    {
+      "epoch": 0.035410282601293835,
+      "grad_norm": 1.150009036064148,
+      "learning_rate": 1e-06,
+      "loss": 0.1536,
+      "step": 208
+    },
+    {
+      "epoch": 0.035580524344569285,
+      "grad_norm": 1.2122458219528198,
+      "learning_rate": 1e-06,
+      "loss": 0.1717,
+      "step": 209
+    },
+    {
+      "epoch": 0.03575076608784474,
+      "grad_norm": 1.1708459854125977,
+      "learning_rate": 1e-06,
+      "loss": 0.1685,
+      "step": 210
+    },
+    {
+      "epoch": 0.03592100783112019,
+      "grad_norm": 1.1359509229660034,
+      "learning_rate": 1e-06,
+      "loss": 0.1628,
+      "step": 211
+    },
+    {
+      "epoch": 0.03609124957439564,
+      "grad_norm": 2.6132564544677734,
+      "learning_rate": 1e-06,
+      "loss": 0.2673,
+      "step": 212
+    },
+    {
+      "epoch": 0.03626149131767109,
+      "grad_norm": 1.2932703495025635,
+      "learning_rate": 1e-06,
+      "loss": 0.1568,
+      "step": 213
+    },
+    {
+      "epoch": 0.03643173306094655,
+      "grad_norm": 1.6617991924285889,
+      "learning_rate": 1e-06,
+      "loss": 0.1944,
+      "step": 214
+    },
+    {
+      "epoch": 0.036601974804222,
+      "grad_norm": 2.0360565185546875,
+      "learning_rate": 1e-06,
+      "loss": 0.2282,
+      "step": 215
+    },
+    {
+      "epoch": 0.03677221654749745,
+      "grad_norm": 1.231723666191101,
+      "learning_rate": 1e-06,
+      "loss": 0.1452,
+      "step": 216
+    },
+    {
+      "epoch": 0.036942458290772896,
+      "grad_norm": 1.3259527683258057,
+      "learning_rate": 1e-06,
+      "loss": 0.167,
+      "step": 217
+    },
+    {
+      "epoch": 0.037112700034048346,
+      "grad_norm": 1.054499626159668,
+      "learning_rate": 1e-06,
+      "loss": 0.1342,
+      "step": 218
+    },
+    {
+      "epoch": 0.0372829417773238,
+      "grad_norm": 1.4665144681930542,
+      "learning_rate": 1e-06,
+      "loss": 0.1978,
+      "step": 219
+    },
+    {
+      "epoch": 0.03745318352059925,
+      "grad_norm": 1.3940861225128174,
+      "learning_rate": 1e-06,
+      "loss": 0.1971,
+      "step": 220
+    },
+    {
+      "epoch": 0.0376234252638747,
+      "grad_norm": 1.164638638496399,
+      "learning_rate": 1e-06,
+      "loss": 0.1727,
+      "step": 221
+    },
+    {
+      "epoch": 0.03779366700715015,
+      "grad_norm": 1.1719894409179688,
+      "learning_rate": 1e-06,
+      "loss": 0.1426,
+      "step": 222
+    },
+    {
+      "epoch": 0.0379639087504256,
+      "grad_norm": 1.1395282745361328,
+      "learning_rate": 1e-06,
+      "loss": 0.1544,
+      "step": 223
+    },
+    {
+      "epoch": 0.03813415049370106,
+      "grad_norm": 1.2896770238876343,
+      "learning_rate": 1e-06,
+      "loss": 0.1709,
+      "step": 224
+    },
+    {
+      "epoch": 0.03830439223697651,
+      "grad_norm": 1.1964484453201294,
+      "learning_rate": 1e-06,
+      "loss": 0.1723,
+      "step": 225
+    },
+    {
+      "epoch": 0.03847463398025196,
+      "grad_norm": 1.2355952262878418,
+      "learning_rate": 1e-06,
+      "loss": 0.1388,
+      "step": 226
+    },
+    {
+      "epoch": 0.03864487572352741,
+      "grad_norm": 1.3132867813110352,
+      "learning_rate": 1e-06,
+      "loss": 0.174,
+      "step": 227
+    },
+    {
+      "epoch": 0.03881511746680286,
+      "grad_norm": 1.175132393836975,
+      "learning_rate": 1e-06,
+      "loss": 0.1533,
+      "step": 228
+    },
+    {
+      "epoch": 0.03898535921007831,
+      "grad_norm": 1.4419723749160767,
+      "learning_rate": 1e-06,
+      "loss": 0.1839,
+      "step": 229
+    },
+    {
+      "epoch": 0.03915560095335376,
+      "grad_norm": 1.1911609172821045,
+      "learning_rate": 1e-06,
+      "loss": 0.1473,
+      "step": 230
+    },
+    {
+      "epoch": 0.03932584269662921,
+      "grad_norm": 1.047421932220459,
+      "learning_rate": 1e-06,
+      "loss": 0.1437,
+      "step": 231
+    },
+    {
+      "epoch": 0.03949608443990466,
+      "grad_norm": 1.2604156732559204,
+      "learning_rate": 1e-06,
+      "loss": 0.1651,
+      "step": 232
+    },
+    {
+      "epoch": 0.03966632618318012,
+      "grad_norm": 1.2047826051712036,
+      "learning_rate": 1e-06,
+      "loss": 0.1587,
+      "step": 233
+    },
+    {
+      "epoch": 0.03983656792645557,
+      "grad_norm": 1.0610707998275757,
+      "learning_rate": 1e-06,
+      "loss": 0.1317,
+      "step": 234
+    },
+    {
+      "epoch": 0.04000680966973102,
+      "grad_norm": 1.1626613140106201,
+      "learning_rate": 1e-06,
+      "loss": 0.1543,
+      "step": 235
+    },
+    {
+      "epoch": 0.04017705141300647,
+      "grad_norm": 1.0833077430725098,
+      "learning_rate": 1e-06,
+      "loss": 0.1316,
+      "step": 236
+    },
+    {
+      "epoch": 0.04034729315628192,
+      "grad_norm": 1.4806079864501953,
+      "learning_rate": 1e-06,
+      "loss": 0.1699,
+      "step": 237
+    },
+    {
+      "epoch": 0.040517534899557374,
+      "grad_norm": 1.360552191734314,
+      "learning_rate": 1e-06,
+      "loss": 0.1559,
+      "step": 238
+    },
+    {
+      "epoch": 0.040687776642832824,
+      "grad_norm": 1.2109061479568481,
+      "learning_rate": 1e-06,
+      "loss": 0.1507,
+      "step": 239
+    },
+    {
+      "epoch": 0.04085801838610827,
+      "grad_norm": 1.1271363496780396,
+      "learning_rate": 1e-06,
+      "loss": 0.1522,
+      "step": 240
+    },
+    {
+      "epoch": 0.04102826012938372,
+      "grad_norm": 1.3026715517044067,
+      "learning_rate": 1e-06,
+      "loss": 0.1586,
+      "step": 241
+    },
+    {
+      "epoch": 0.04119850187265917,
+      "grad_norm": 1.589646816253662,
+      "learning_rate": 1e-06,
+      "loss": 0.2006,
+      "step": 242
+    },
+    {
+      "epoch": 0.04136874361593463,
+      "grad_norm": 1.1894760131835938,
+      "learning_rate": 1e-06,
+      "loss": 0.1547,
+      "step": 243
+    },
+    {
+      "epoch": 0.04153898535921008,
+      "grad_norm": 1.1556522846221924,
+      "learning_rate": 1e-06,
+      "loss": 0.1509,
+      "step": 244
+    },
+    {
+      "epoch": 0.04170922710248553,
+      "grad_norm": 2.4897043704986572,
+      "learning_rate": 1e-06,
+      "loss": 0.2587,
+      "step": 245
+    },
+    {
+      "epoch": 0.04187946884576098,
+      "grad_norm": 1.3815762996673584,
+      "learning_rate": 1e-06,
+      "loss": 0.1604,
+      "step": 246
+    },
+    {
+      "epoch": 0.042049710589036435,
+      "grad_norm": 1.4068046808242798,
+      "learning_rate": 1e-06,
+      "loss": 0.1628,
+      "step": 247
+    },
+    {
+      "epoch": 0.042219952332311884,
+      "grad_norm": 1.4070250988006592,
+      "learning_rate": 1e-06,
+      "loss": 0.1631,
+      "step": 248
+    },
+    {
+      "epoch": 0.042390194075587334,
+      "grad_norm": 1.1561863422393799,
+      "learning_rate": 1e-06,
+      "loss": 0.1488,
+      "step": 249
+    },
+    {
+      "epoch": 0.042560435818862784,
+      "grad_norm": 1.0689971446990967,
+      "learning_rate": 1e-06,
+      "loss": 0.1318,
+      "step": 250
+    },
+    {
+      "epoch": 0.04273067756213823,
+      "grad_norm": 1.1282958984375,
+      "learning_rate": 1e-06,
+      "loss": 0.1243,
+      "step": 251
+    },
+    {
+      "epoch": 0.04290091930541369,
+      "grad_norm": 1.2809712886810303,
+      "learning_rate": 1e-06,
+      "loss": 0.1609,
+      "step": 252
+    },
+    {
+      "epoch": 0.04307116104868914,
+      "grad_norm": 1.1898773908615112,
+      "learning_rate": 1e-06,
+      "loss": 0.1653,
+      "step": 253
+    },
+    {
+      "epoch": 0.04324140279196459,
+      "grad_norm": 1.215122938156128,
+      "learning_rate": 1e-06,
+      "loss": 0.1459,
+      "step": 254
+    },
+    {
+      "epoch": 0.04341164453524004,
+      "grad_norm": 1.4629710912704468,
+      "learning_rate": 1e-06,
+      "loss": 0.1945,
+      "step": 255
+    },
+    {
+      "epoch": 0.04358188627851549,
+      "grad_norm": 1.206394910812378,
+      "learning_rate": 1e-06,
+      "loss": 0.149,
+      "step": 256
+    },
+    {
+      "epoch": 0.043752128021790945,
+      "grad_norm": 1.3022671937942505,
+      "learning_rate": 1e-06,
+      "loss": 0.1529,
+      "step": 257
+    },
+    {
+      "epoch": 0.043922369765066395,
+      "grad_norm": 1.2597123384475708,
+      "learning_rate": 1e-06,
+      "loss": 0.1411,
+      "step": 258
+    },
+    {
+      "epoch": 0.044092611508341845,
+      "grad_norm": 1.3716976642608643,
+      "learning_rate": 1e-06,
+      "loss": 0.1552,
+      "step": 259
+    },
+    {
+      "epoch": 0.044262853251617294,
+      "grad_norm": 1.2082346677780151,
+      "learning_rate": 1e-06,
+      "loss": 0.1583,
+      "step": 260
+    },
+    {
+      "epoch": 0.04443309499489275,
+      "grad_norm": 1.1413241624832153,
+      "learning_rate": 1e-06,
+      "loss": 0.1197,
+      "step": 261
+    },
+    {
+      "epoch": 0.0446033367381682,
+      "grad_norm": 1.346548080444336,
+      "learning_rate": 1e-06,
+      "loss": 0.175,
+      "step": 262
+    },
+    {
+      "epoch": 0.04477357848144365,
+      "grad_norm": 1.326627254486084,
+      "learning_rate": 1e-06,
+      "loss": 0.168,
+      "step": 263
+    },
+    {
+      "epoch": 0.0449438202247191,
+      "grad_norm": 1.259684681892395,
+      "learning_rate": 1e-06,
+      "loss": 0.1541,
+      "step": 264
+    },
+    {
+      "epoch": 0.04511406196799455,
+      "grad_norm": 1.3728045225143433,
+      "learning_rate": 1e-06,
+      "loss": 0.1653,
+      "step": 265
+    },
+    {
+      "epoch": 0.045284303711270006,
+      "grad_norm": 1.1758908033370972,
+      "learning_rate": 1e-06,
+      "loss": 0.1519,
+      "step": 266
+    },
+    {
+      "epoch": 0.045454545454545456,
+      "grad_norm": 1.4522403478622437,
+      "learning_rate": 1e-06,
+      "loss": 0.1541,
+      "step": 267
+    },
+    {
+      "epoch": 0.045624787197820905,
+      "grad_norm": 1.3297243118286133,
+      "learning_rate": 1e-06,
+      "loss": 0.1647,
+      "step": 268
+    },
+    {
+      "epoch": 0.045795028941096355,
+      "grad_norm": 1.1865285634994507,
+      "learning_rate": 1e-06,
+      "loss": 0.1415,
+      "step": 269
+    },
+    {
+      "epoch": 0.045965270684371805,
+      "grad_norm": 1.2091939449310303,
+      "learning_rate": 1e-06,
+      "loss": 0.1386,
+      "step": 270
+    },
+    {
+      "epoch": 0.04613551242764726,
+      "grad_norm": 1.2446978092193604,
+      "learning_rate": 1e-06,
+      "loss": 0.1285,
+      "step": 271
+    },
+    {
+      "epoch": 0.04630575417092271,
+      "grad_norm": 1.1907007694244385,
+      "learning_rate": 1e-06,
+      "loss": 0.1607,
+      "step": 272
+    },
+    {
+      "epoch": 0.04647599591419816,
+      "grad_norm": 1.2765071392059326,
+      "learning_rate": 1e-06,
+      "loss": 0.1539,
+      "step": 273
+    },
+    {
+      "epoch": 0.04664623765747361,
+      "grad_norm": 1.364173173904419,
+      "learning_rate": 1e-06,
+      "loss": 0.1434,
+      "step": 274
+    },
+    {
+      "epoch": 0.04681647940074907,
+      "grad_norm": 1.212433934211731,
+      "learning_rate": 1e-06,
+      "loss": 0.1221,
+      "step": 275
+    },
+    {
+      "epoch": 0.04698672114402452,
+      "grad_norm": 1.203960657119751,
+      "learning_rate": 1e-06,
+      "loss": 0.1292,
+      "step": 276
+    },
+    {
+      "epoch": 0.047156962887299966,
+      "grad_norm": 1.4043571949005127,
+      "learning_rate": 1e-06,
+      "loss": 0.1496,
+      "step": 277
+    },
+    {
+      "epoch": 0.047327204630575416,
+      "grad_norm": 1.4482816457748413,
+      "learning_rate": 1e-06,
+      "loss": 0.1747,
+      "step": 278
+    },
+    {
+      "epoch": 0.047497446373850866,
+      "grad_norm": 1.1347603797912598,
+      "learning_rate": 1e-06,
+      "loss": 0.1268,
+      "step": 279
+    },
+    {
+      "epoch": 0.04766768811712632,
+      "grad_norm": 1.1415354013442993,
+      "learning_rate": 1e-06,
+      "loss": 0.1448,
+      "step": 280
+    },
+    {
+      "epoch": 0.04783792986040177,
+      "grad_norm": 1.4771915674209595,
+      "learning_rate": 1e-06,
+      "loss": 0.1596,
+      "step": 281
+    },
+    {
+      "epoch": 0.04800817160367722,
+      "grad_norm": 1.176045298576355,
+      "learning_rate": 1e-06,
+      "loss": 0.1207,
+      "step": 282
+    },
+    {
+      "epoch": 0.04817841334695267,
+      "grad_norm": 1.3680100440979004,
+      "learning_rate": 1e-06,
+      "loss": 0.1413,
+      "step": 283
+    },
+    {
+      "epoch": 0.04834865509022812,
+      "grad_norm": 1.3888130187988281,
+      "learning_rate": 1e-06,
+      "loss": 0.1395,
+      "step": 284
+    },
+    {
+      "epoch": 0.04851889683350358,
+      "grad_norm": 1.4310017824172974,
+      "learning_rate": 1e-06,
+      "loss": 0.1395,
+      "step": 285
+    },
+    {
+      "epoch": 0.04868913857677903,
+      "grad_norm": 1.4663360118865967,
+      "learning_rate": 1e-06,
+      "loss": 0.1674,
+      "step": 286
+    },
+    {
+      "epoch": 0.04885938032005448,
+      "grad_norm": 1.2074915170669556,
+      "learning_rate": 1e-06,
+      "loss": 0.1384,
+      "step": 287
+    },
+    {
+      "epoch": 0.049029622063329927,
+      "grad_norm": 1.3467768430709839,
+      "learning_rate": 1e-06,
+      "loss": 0.1702,
+      "step": 288
+    },
+    {
+      "epoch": 0.049199863806605376,
+      "grad_norm": 1.192862629890442,
+      "learning_rate": 1e-06,
+      "loss": 0.1347,
+      "step": 289
+    },
+    {
+      "epoch": 0.04937010554988083,
+      "grad_norm": 1.304807424545288,
+      "learning_rate": 1e-06,
+      "loss": 0.1459,
+      "step": 290
+    },
+    {
+      "epoch": 0.04954034729315628,
+      "grad_norm": 1.309431791305542,
+      "learning_rate": 1e-06,
+      "loss": 0.1651,
+      "step": 291
+    },
+    {
+      "epoch": 0.04971058903643173,
+      "grad_norm": 1.534616231918335,
+      "learning_rate": 1e-06,
+      "loss": 0.1481,
+      "step": 292
+    },
+    {
+      "epoch": 0.04988083077970718,
+      "grad_norm": 1.2383191585540771,
+      "learning_rate": 1e-06,
+      "loss": 0.1238,
+      "step": 293
+    },
+    {
+      "epoch": 0.05005107252298264,
+      "grad_norm": 1.1695579290390015,
+      "learning_rate": 1e-06,
+      "loss": 0.1144,
+      "step": 294
+    },
+    {
+      "epoch": 0.05022131426625809,
+      "grad_norm": 1.3590017557144165,
+      "learning_rate": 1e-06,
+      "loss": 0.1376,
+      "step": 295
+    },
+    {
+      "epoch": 0.05039155600953354,
+      "grad_norm": 1.186447024345398,
+      "learning_rate": 1e-06,
+      "loss": 0.1336,
+      "step": 296
+    },
+    {
+      "epoch": 0.05056179775280899,
+      "grad_norm": 1.2510392665863037,
+      "learning_rate": 1e-06,
+      "loss": 0.1363,
+      "step": 297
+    },
+    {
+      "epoch": 0.05073203949608444,
+      "grad_norm": 1.3094967603683472,
+      "learning_rate": 1e-06,
+      "loss": 0.1339,
+      "step": 298
+    },
+    {
+      "epoch": 0.050902281239359894,
+      "grad_norm": 1.4106348752975464,
+      "learning_rate": 1e-06,
+      "loss": 0.1527,
+      "step": 299
+    },
+    {
+      "epoch": 0.05107252298263534,
+      "grad_norm": 1.2612876892089844,
+      "learning_rate": 1e-06,
+      "loss": 0.1343,
+      "step": 300
+    },
+    {
+      "epoch": 0.05124276472591079,
+      "grad_norm": 1.3075928688049316,
+      "learning_rate": 1e-06,
+      "loss": 0.1572,
+      "step": 301
+    },
+    {
+      "epoch": 0.05141300646918624,
+      "grad_norm": 1.3252586126327515,
+      "learning_rate": 1e-06,
+      "loss": 0.1491,
+      "step": 302
+    },
+    {
+      "epoch": 0.05158324821246169,
+      "grad_norm": 1.3578503131866455,
+      "learning_rate": 1e-06,
+      "loss": 0.1526,
+      "step": 303
+    },
+    {
+      "epoch": 0.05175348995573715,
+      "grad_norm": 1.383754849433899,
+      "learning_rate": 1e-06,
+      "loss": 0.1547,
+      "step": 304
+    },
+    {
+      "epoch": 0.0519237316990126,
+      "grad_norm": 1.463248372077942,
+      "learning_rate": 1e-06,
+      "loss": 0.171,
+      "step": 305
+    },
+    {
+      "epoch": 0.05209397344228805,
+      "grad_norm": 1.3403490781784058,
+      "learning_rate": 1e-06,
+      "loss": 0.1738,
+      "step": 306
+    },
+    {
+      "epoch": 0.0522642151855635,
+      "grad_norm": 1.3475086688995361,
+      "learning_rate": 1e-06,
+      "loss": 0.1434,
+      "step": 307
+    },
+    {
+      "epoch": 0.052434456928838954,
+      "grad_norm": 1.1605323553085327,
+      "learning_rate": 1e-06,
+      "loss": 0.1362,
+      "step": 308
+    },
+    {
+      "epoch": 0.052604698672114404,
+      "grad_norm": 1.2388291358947754,
+      "learning_rate": 1e-06,
+      "loss": 0.1377,
+      "step": 309
+    },
+    {
+      "epoch": 0.052774940415389854,
+      "grad_norm": 1.6835265159606934,
+      "learning_rate": 1e-06,
+      "loss": 0.1976,
+      "step": 310
+    },
+    {
+      "epoch": 0.052945182158665303,
+      "grad_norm": 1.2333765029907227,
+      "learning_rate": 1e-06,
+      "loss": 0.1284,
+      "step": 311
+    },
+    {
+      "epoch": 0.05311542390194075,
+      "grad_norm": 1.581054925918579,
+      "learning_rate": 1e-06,
+      "loss": 0.1709,
+      "step": 312
+    },
+    {
+      "epoch": 0.05328566564521621,
+      "grad_norm": 1.165010690689087,
+      "learning_rate": 1e-06,
+      "loss": 0.1258,
+      "step": 313
+    },
+    {
+      "epoch": 0.05345590738849166,
+      "grad_norm": 1.209064245223999,
+      "learning_rate": 1e-06,
+      "loss": 0.1223,
+      "step": 314
+    },
+    {
+      "epoch": 0.05362614913176711,
+      "grad_norm": 1.3305888175964355,
+      "learning_rate": 1e-06,
+      "loss": 0.1425,
+      "step": 315
+    },
+    {
+      "epoch": 0.05379639087504256,
+      "grad_norm": 1.828088402748108,
+      "learning_rate": 1e-06,
+      "loss": 0.1634,
+      "step": 316
+    },
+    {
+      "epoch": 0.05396663261831801,
+      "grad_norm": 1.332383394241333,
+      "learning_rate": 1e-06,
+      "loss": 0.1268,
+      "step": 317
+    },
+    {
+      "epoch": 0.054136874361593465,
+      "grad_norm": 1.339270830154419,
+      "learning_rate": 1e-06,
+      "loss": 0.1428,
+      "step": 318
+    },
+    {
+      "epoch": 0.054307116104868915,
+      "grad_norm": 1.2108736038208008,
+      "learning_rate": 1e-06,
+      "loss": 0.1345,
+      "step": 319
+    },
+    {
+      "epoch": 0.054477357848144364,
+      "grad_norm": 1.2508689165115356,
+      "learning_rate": 1e-06,
+      "loss": 0.1251,
+      "step": 320
+    },
+    {
+      "epoch": 0.054647599591419814,
+      "grad_norm": 1.0894383192062378,
+      "learning_rate": 1e-06,
+      "loss": 0.1272,
+      "step": 321
+    },
+    {
+      "epoch": 0.05481784133469527,
+      "grad_norm": 1.2453547716140747,
+      "learning_rate": 1e-06,
+      "loss": 0.1315,
+      "step": 322
+    },
+    {
+      "epoch": 0.05498808307797072,
+      "grad_norm": 1.3076040744781494,
+      "learning_rate": 1e-06,
+      "loss": 0.1178,
+      "step": 323
+    },
+    {
+      "epoch": 0.05515832482124617,
+      "grad_norm": 2.6135568618774414,
+      "learning_rate": 1e-06,
+      "loss": 0.1613,
+      "step": 324
+    },
+    {
+      "epoch": 0.05532856656452162,
+      "grad_norm": 1.5091561079025269,
+      "learning_rate": 1e-06,
+      "loss": 0.1126,
+      "step": 325
+    },
+    {
+      "epoch": 0.05549880830779707,
+      "grad_norm": 1.7307822704315186,
+      "learning_rate": 1e-06,
+      "loss": 0.171,
+      "step": 326
+    },
+    {
+      "epoch": 0.055669050051072526,
+      "grad_norm": 1.2100858688354492,
+      "learning_rate": 1e-06,
+      "loss": 0.1094,
+      "step": 327
+    },
+    {
+      "epoch": 0.055839291794347976,
+      "grad_norm": 1.2510110139846802,
+      "learning_rate": 1e-06,
+      "loss": 0.1162,
+      "step": 328
+    },
+    {
+      "epoch": 0.056009533537623425,
+      "grad_norm": 1.3461787700653076,
+      "learning_rate": 1e-06,
+      "loss": 0.1427,
+      "step": 329
+    },
+    {
+      "epoch": 0.056179775280898875,
+      "grad_norm": 1.6088210344314575,
+      "learning_rate": 1e-06,
+      "loss": 0.1648,
+      "step": 330
+    },
+    {
+      "epoch": 0.056350017024174325,
+      "grad_norm": 1.4088263511657715,
+      "learning_rate": 1e-06,
+      "loss": 0.1334,
+      "step": 331
+    },
+    {
+      "epoch": 0.05652025876744978,
+      "grad_norm": 1.4224536418914795,
+      "learning_rate": 1e-06,
+      "loss": 0.1414,
+      "step": 332
+    },
+    {
+      "epoch": 0.05669050051072523,
+      "grad_norm": 1.1860085725784302,
+      "learning_rate": 1e-06,
+      "loss": 0.132,
+      "step": 333
+    },
+    {
+      "epoch": 0.05686074225400068,
+      "grad_norm": 1.3678163290023804,
+      "learning_rate": 1e-06,
+      "loss": 0.1496,
+      "step": 334
+    },
+    {
+      "epoch": 0.05703098399727613,
+      "grad_norm": 1.4010056257247925,
+      "learning_rate": 1e-06,
+      "loss": 0.1444,
+      "step": 335
+    },
+    {
+      "epoch": 0.05720122574055159,
+      "grad_norm": 1.2456624507904053,
+      "learning_rate": 1e-06,
+      "loss": 0.1222,
+      "step": 336
+    },
+    {
+      "epoch": 0.057371467483827036,
+      "grad_norm": 1.5460395812988281,
+      "learning_rate": 1e-06,
+      "loss": 0.1385,
+      "step": 337
+    },
+    {
+      "epoch": 0.057541709227102486,
+      "grad_norm": 1.2900766134262085,
+      "learning_rate": 1e-06,
+      "loss": 0.1096,
+      "step": 338
+    },
+    {
+      "epoch": 0.057711950970377936,
+      "grad_norm": 1.5039972066879272,
+      "learning_rate": 1e-06,
+      "loss": 0.1393,
+      "step": 339
+    },
+    {
+      "epoch": 0.057882192713653385,
+      "grad_norm": 1.2665271759033203,
+      "learning_rate": 1e-06,
+      "loss": 0.1264,
+      "step": 340
+    },
+    {
+      "epoch": 0.05805243445692884,
+      "grad_norm": 1.4474594593048096,
+      "learning_rate": 1e-06,
+      "loss": 0.1308,
+      "step": 341
+    },
+    {
+      "epoch": 0.05822267620020429,
+      "grad_norm": 1.3482589721679688,
+      "learning_rate": 1e-06,
+      "loss": 0.1473,
+      "step": 342
+    },
+    {
+      "epoch": 0.05839291794347974,
+      "grad_norm": 1.3561499118804932,
+      "learning_rate": 1e-06,
+      "loss": 0.1227,
+      "step": 343
+    },
+    {
+      "epoch": 0.05856315968675519,
+      "grad_norm": 1.194254755973816,
+      "learning_rate": 1e-06,
+      "loss": 0.116,
+      "step": 344
+    },
+    {
+      "epoch": 0.05873340143003064,
+      "grad_norm": 1.2828115224838257,
+      "learning_rate": 1e-06,
+      "loss": 0.1218,
+      "step": 345
+    },
+    {
+      "epoch": 0.0589036431733061,
+      "grad_norm": 2.596560001373291,
+      "learning_rate": 1e-06,
+      "loss": 0.2237,
+      "step": 346
+    },
+    {
+      "epoch": 0.05907388491658155,
+      "grad_norm": 1.3315943479537964,
+      "learning_rate": 1e-06,
+      "loss": 0.1271,
+      "step": 347
+    },
+    {
+      "epoch": 0.059244126659857,
+      "grad_norm": 1.2427936792373657,
+      "learning_rate": 1e-06,
+      "loss": 0.1181,
+      "step": 348
+    },
+    {
+      "epoch": 0.059414368403132446,
+      "grad_norm": 1.223074197769165,
+      "learning_rate": 1e-06,
+      "loss": 0.1073,
+      "step": 349
+    },
+    {
+      "epoch": 0.059584610146407896,
+      "grad_norm": 1.4834375381469727,
+      "learning_rate": 1e-06,
+      "loss": 0.1435,
+      "step": 350
+    },
+    {
+      "epoch": 0.05975485188968335,
+      "grad_norm": 1.4834375381469727,
+      "learning_rate": 1e-06,
+      "loss": 0.1392,
+      "step": 351
+    },
+    {
+      "epoch": 0.0599250936329588,
+      "grad_norm": 1.2152388095855713,
+      "learning_rate": 1e-06,
+      "loss": 0.1158,
+      "step": 352
+    },
+    {
+      "epoch": 0.06009533537623425,
+      "grad_norm": 1.3129734992980957,
+      "learning_rate": 1e-06,
+      "loss": 0.1272,
+      "step": 353
+    },
+    {
+      "epoch": 0.0602655771195097,
+      "grad_norm": 1.3754154443740845,
+      "learning_rate": 1e-06,
+      "loss": 0.1416,
+      "step": 354
+    },
+    {
+      "epoch": 0.06043581886278516,
+      "grad_norm": 1.2157810926437378,
+      "learning_rate": 1e-06,
+      "loss": 0.1207,
+      "step": 355
+    },
+    {
+      "epoch": 0.06060606060606061,
+      "grad_norm": 1.3933988809585571,
+      "learning_rate": 1e-06,
+      "loss": 0.1504,
+      "step": 356
+    },
+    {
+      "epoch": 0.06077630234933606,
+      "grad_norm": 1.3696870803833008,
+      "learning_rate": 1e-06,
+      "loss": 0.1303,
+      "step": 357
+    },
+    {
+      "epoch": 0.06094654409261151,
+      "grad_norm": 1.324859380722046,
+      "learning_rate": 1e-06,
+      "loss": 0.1344,
+      "step": 358
+    },
+    {
+      "epoch": 0.06111678583588696,
+      "grad_norm": 1.1997361183166504,
+      "learning_rate": 1e-06,
+      "loss": 0.1165,
+      "step": 359
+    },
+    {
+      "epoch": 0.06128702757916241,
+      "grad_norm": 1.3134782314300537,
+      "learning_rate": 1e-06,
+      "loss": 0.1269,
+      "step": 360
+    },
+    {
+      "epoch": 0.06145726932243786,
+      "grad_norm": 2.7401535511016846,
+      "learning_rate": 1e-06,
+      "loss": 0.253,
+      "step": 361
+    },
+    {
+      "epoch": 0.06162751106571331,
+      "grad_norm": 1.3712646961212158,
+      "learning_rate": 1e-06,
+      "loss": 0.138,
+      "step": 362
+    },
+    {
+      "epoch": 0.06179775280898876,
+      "grad_norm": 1.2538585662841797,
+      "learning_rate": 1e-06,
+      "loss": 0.1225,
+      "step": 363
+    },
+    {
+      "epoch": 0.06196799455226421,
+      "grad_norm": 1.2842880487442017,
+      "learning_rate": 1e-06,
+      "loss": 0.1301,
+      "step": 364
+    },
+    {
+      "epoch": 0.06213823629553967,
+      "grad_norm": 1.1870968341827393,
+      "learning_rate": 1e-06,
+      "loss": 0.1116,
+      "step": 365
+    },
+    {
+      "epoch": 0.06230847803881512,
+      "grad_norm": 1.3063998222351074,
+      "learning_rate": 1e-06,
+      "loss": 0.124,
+      "step": 366
+    },
+    {
+      "epoch": 0.06247871978209057,
+      "grad_norm": 1.3379669189453125,
+      "learning_rate": 1e-06,
+      "loss": 0.1396,
+      "step": 367
+    },
+    {
+      "epoch": 0.06264896152536602,
+      "grad_norm": 1.3882386684417725,
+      "learning_rate": 1e-06,
+      "loss": 0.1274,
+      "step": 368
+    },
+    {
+      "epoch": 0.06281920326864147,
+      "grad_norm": 1.38962984085083,
+      "learning_rate": 1e-06,
+      "loss": 0.1192,
+      "step": 369
+    },
+    {
+      "epoch": 0.06298944501191692,
+      "grad_norm": 1.5751211643218994,
+      "learning_rate": 1e-06,
+      "loss": 0.1465,
+      "step": 370
+    },
+    {
+      "epoch": 0.06315968675519237,
+      "grad_norm": 1.4705424308776855,
+      "learning_rate": 1e-06,
+      "loss": 0.1428,
+      "step": 371
+    },
+    {
+      "epoch": 0.06332992849846783,
+      "grad_norm": 1.27107834815979,
+      "learning_rate": 1e-06,
+      "loss": 0.1309,
+      "step": 372
+    },
+    {
+      "epoch": 0.06350017024174327,
+      "grad_norm": 1.3142486810684204,
+      "learning_rate": 1e-06,
+      "loss": 0.1407,
+      "step": 373
+    },
+    {
+      "epoch": 0.06367041198501873,
+      "grad_norm": 1.4944632053375244,
+      "learning_rate": 1e-06,
+      "loss": 0.139,
+      "step": 374
+    },
+    {
+      "epoch": 0.06384065372829417,
+      "grad_norm": 1.5329262018203735,
+      "learning_rate": 1e-06,
+      "loss": 0.1533,
+      "step": 375
+    },
+    {
+      "epoch": 0.06401089547156963,
+      "grad_norm": 1.682967185974121,
+      "learning_rate": 1e-06,
+      "loss": 0.1446,
+      "step": 376
+    },
+    {
+      "epoch": 0.06418113721484509,
+      "grad_norm": 1.3599567413330078,
+      "learning_rate": 1e-06,
+      "loss": 0.117,
+      "step": 377
+    },
+    {
+      "epoch": 0.06435137895812053,
+      "grad_norm": 1.41896653175354,
+      "learning_rate": 1e-06,
+      "loss": 0.1316,
+      "step": 378
+    },
+    {
+      "epoch": 0.06452162070139598,
+      "grad_norm": 1.2187316417694092,
+      "learning_rate": 1e-06,
+      "loss": 0.1104,
+      "step": 379
+    },
+    {
+      "epoch": 0.06469186244467143,
+      "grad_norm": 1.3118098974227905,
+      "learning_rate": 1e-06,
+      "loss": 0.1173,
+      "step": 380
+    },
+    {
+      "epoch": 0.06486210418794688,
+      "grad_norm": 1.2350298166275024,
+      "learning_rate": 1e-06,
+      "loss": 0.1188,
+      "step": 381
+    },
+    {
+      "epoch": 0.06503234593122234,
+      "grad_norm": 1.3306002616882324,
+      "learning_rate": 1e-06,
+      "loss": 0.1153,
+      "step": 382
+    },
+    {
+      "epoch": 0.06520258767449778,
+      "grad_norm": 2.9044129848480225,
+      "learning_rate": 1e-06,
+      "loss": 0.1784,
+      "step": 383
+    },
+    {
+      "epoch": 0.06537282941777324,
+      "grad_norm": 1.6895009279251099,
+      "learning_rate": 1e-06,
+      "loss": 0.1522,
+      "step": 384
+    },
+    {
+      "epoch": 0.06554307116104868,
+      "grad_norm": 1.2503180503845215,
+      "learning_rate": 1e-06,
+      "loss": 0.1048,
+      "step": 385
+    },
+    {
+      "epoch": 0.06571331290432414,
+      "grad_norm": 1.4215972423553467,
+      "learning_rate": 1e-06,
+      "loss": 0.135,
+      "step": 386
+    },
+    {
+      "epoch": 0.0658835546475996,
+      "grad_norm": 1.4079679250717163,
+      "learning_rate": 1e-06,
+      "loss": 0.1266,
+      "step": 387
+    },
+    {
+      "epoch": 0.06605379639087504,
+      "grad_norm": 1.29921293258667,
+      "learning_rate": 1e-06,
+      "loss": 0.1122,
+      "step": 388
+    },
+    {
+      "epoch": 0.0662240381341505,
+      "grad_norm": 1.442528486251831,
+      "learning_rate": 1e-06,
+      "loss": 0.1333,
+      "step": 389
+    },
+    {
+      "epoch": 0.06639427987742594,
+      "grad_norm": 1.8958500623703003,
+      "learning_rate": 1e-06,
+      "loss": 0.1928,
+      "step": 390
+    },
+    {
+      "epoch": 0.0665645216207014,
+      "grad_norm": 1.1162981986999512,
+      "learning_rate": 1e-06,
+      "loss": 0.0949,
+      "step": 391
+    },
+    {
+      "epoch": 0.06673476336397685,
+      "grad_norm": 1.3155403137207031,
+      "learning_rate": 1e-06,
+      "loss": 0.1148,
+      "step": 392
+    },
+    {
+      "epoch": 0.0669050051072523,
+      "grad_norm": 1.2109743356704712,
+      "learning_rate": 1e-06,
+      "loss": 0.0965,
+      "step": 393
+    },
+    {
+      "epoch": 0.06707524685052775,
+      "grad_norm": 1.2797634601593018,
+      "learning_rate": 1e-06,
+      "loss": 0.108,
+      "step": 394
+    },
+    {
+      "epoch": 0.06724548859380321,
+      "grad_norm": 1.2622194290161133,
+      "learning_rate": 1e-06,
+      "loss": 0.1205,
+      "step": 395
+    },
+    {
+      "epoch": 0.06741573033707865,
+      "grad_norm": 1.4360229969024658,
+      "learning_rate": 1e-06,
+      "loss": 0.1337,
+      "step": 396
+    },
+    {
+      "epoch": 0.0675859720803541,
+      "grad_norm": 1.1915457248687744,
+      "learning_rate": 1e-06,
+      "loss": 0.1186,
+      "step": 397
+    },
+    {
+      "epoch": 0.06775621382362955,
+      "grad_norm": 1.2633016109466553,
+      "learning_rate": 1e-06,
+      "loss": 0.1055,
+      "step": 398
+    },
+    {
+      "epoch": 0.067926455566905,
+      "grad_norm": 1.318787932395935,
+      "learning_rate": 1e-06,
+      "loss": 0.1247,
+      "step": 399
+    },
+    {
+      "epoch": 0.06809669731018046,
+      "grad_norm": 1.0899081230163574,
+      "learning_rate": 1e-06,
+      "loss": 0.0921,
+      "step": 400
+    },
+    {
+      "epoch": 0.0682669390534559,
+      "grad_norm": 1.4745516777038574,
+      "learning_rate": 1e-06,
+      "loss": 0.134,
+      "step": 401
+    },
+    {
+      "epoch": 0.06843718079673136,
+      "grad_norm": 1.2535583972930908,
+      "learning_rate": 1e-06,
+      "loss": 0.1183,
+      "step": 402
+    },
+    {
+      "epoch": 0.0686074225400068,
+      "grad_norm": 1.5067999362945557,
+      "learning_rate": 1e-06,
+      "loss": 0.1411,
+      "step": 403
+    },
+    {
+      "epoch": 0.06877766428328226,
+      "grad_norm": 1.5428158044815063,
+      "learning_rate": 1e-06,
+      "loss": 0.1325,
+      "step": 404
+    },
+    {
+      "epoch": 0.06894790602655772,
+      "grad_norm": 1.7841521501541138,
+      "learning_rate": 1e-06,
+      "loss": 0.1571,
+      "step": 405
+    },
+    {
+      "epoch": 0.06911814776983316,
+      "grad_norm": 1.1209527254104614,
+      "learning_rate": 1e-06,
+      "loss": 0.094,
+      "step": 406
+    },
+    {
+      "epoch": 0.06928838951310862,
+      "grad_norm": 1.2920805215835571,
+      "learning_rate": 1e-06,
+      "loss": 0.1211,
+      "step": 407
+    },
+    {
+      "epoch": 0.06945863125638406,
+      "grad_norm": 1.410902500152588,
+      "learning_rate": 1e-06,
+      "loss": 0.1163,
+      "step": 408
+    },
+    {
+      "epoch": 0.06962887299965952,
+      "grad_norm": 1.3287694454193115,
+      "learning_rate": 1e-06,
+      "loss": 0.1061,
+      "step": 409
+    },
+    {
+      "epoch": 0.06979911474293497,
+      "grad_norm": 1.3526452779769897,
+      "learning_rate": 1e-06,
+      "loss": 0.1176,
+      "step": 410
+    },
+    {
+      "epoch": 0.06996935648621042,
+      "grad_norm": 1.6113089323043823,
+      "learning_rate": 1e-06,
+      "loss": 0.1318,
+      "step": 411
+    },
+    {
+      "epoch": 0.07013959822948587,
+      "grad_norm": 1.4161698818206787,
+      "learning_rate": 1e-06,
+      "loss": 0.1344,
+      "step": 412
+    },
+    {
+      "epoch": 0.07030983997276131,
+      "grad_norm": 1.4419130086898804,
+      "learning_rate": 1e-06,
+      "loss": 0.137,
+      "step": 413
+    },
+    {
+      "epoch": 0.07048008171603677,
+      "grad_norm": 1.2640091180801392,
+      "learning_rate": 1e-06,
+      "loss": 0.1227,
+      "step": 414
+    },
+    {
+      "epoch": 0.07065032345931223,
+      "grad_norm": 1.23671555519104,
+      "learning_rate": 1e-06,
+      "loss": 0.121,
+      "step": 415
+    },
+    {
+      "epoch": 0.07082056520258767,
+      "grad_norm": 1.3423279523849487,
+      "learning_rate": 1e-06,
+      "loss": 0.0975,
+      "step": 416
+    },
+    {
+      "epoch": 0.07099080694586313,
+      "grad_norm": 1.2852997779846191,
+      "learning_rate": 1e-06,
+      "loss": 0.1246,
+      "step": 417
+    },
+    {
+      "epoch": 0.07116104868913857,
+      "grad_norm": 1.4518605470657349,
+      "learning_rate": 1e-06,
+      "loss": 0.1341,
+      "step": 418
+    },
+    {
+      "epoch": 0.07133129043241403,
+      "grad_norm": 1.4686856269836426,
+      "learning_rate": 1e-06,
+      "loss": 0.1259,
+      "step": 419
+    },
+    {
+      "epoch": 0.07150153217568948,
+      "grad_norm": 1.3495961427688599,
+      "learning_rate": 1e-06,
+      "loss": 0.122,
+      "step": 420
+    },
+    {
+      "epoch": 0.07167177391896493,
+      "grad_norm": 1.5383343696594238,
+      "learning_rate": 1e-06,
+      "loss": 0.1229,
+      "step": 421
+    },
+    {
+      "epoch": 0.07184201566224038,
+      "grad_norm": 1.2680972814559937,
+      "learning_rate": 1e-06,
+      "loss": 0.105,
+      "step": 422
+    },
+    {
+      "epoch": 0.07201225740551584,
+      "grad_norm": 1.4354145526885986,
+      "learning_rate": 1e-06,
+      "loss": 0.1247,
+      "step": 423
+    },
+    {
+      "epoch": 0.07218249914879128,
+      "grad_norm": 1.4354145526885986,
+      "learning_rate": 1e-06,
+      "loss": 0.2006,
+      "step": 424
+    },
+    {
+      "epoch": 0.07235274089206674,
+      "grad_norm": 1.7208505868911743,
+      "learning_rate": 1e-06,
+      "loss": 0.1541,
+      "step": 425
+    },
+    {
+      "epoch": 0.07252298263534218,
+      "grad_norm": 1.2794008255004883,
+      "learning_rate": 1e-06,
+      "loss": 0.1088,
+      "step": 426
+    },
+    {
+      "epoch": 0.07269322437861764,
+      "grad_norm": 1.3522515296936035,
+      "learning_rate": 1e-06,
+      "loss": 0.1139,
+      "step": 427
+    },
+    {
+      "epoch": 0.0728634661218931,
+      "grad_norm": 1.4536526203155518,
+      "learning_rate": 1e-06,
+      "loss": 0.1293,
+      "step": 428
+    },
+    {
+      "epoch": 0.07303370786516854,
+      "grad_norm": 1.2702877521514893,
+      "learning_rate": 1e-06,
+      "loss": 0.0999,
+      "step": 429
+    },
+    {
+      "epoch": 0.073203949608444,
+      "grad_norm": 1.3044495582580566,
+      "learning_rate": 1e-06,
+      "loss": 0.1225,
+      "step": 430
+    },
+    {
+      "epoch": 0.07337419135171944,
+      "grad_norm": 1.5265694856643677,
+      "learning_rate": 1e-06,
+      "loss": 0.1321,
+      "step": 431
+    },
+    {
+      "epoch": 0.0735444330949949,
+      "grad_norm": 1.3504886627197266,
+      "learning_rate": 1e-06,
+      "loss": 0.0999,
+      "step": 432
+    },
+    {
+      "epoch": 0.07371467483827035,
+      "grad_norm": 1.9963023662567139,
+      "learning_rate": 1e-06,
+      "loss": 0.1559,
+      "step": 433
+    },
+    {
+      "epoch": 0.07388491658154579,
+      "grad_norm": 1.6677991151809692,
+      "learning_rate": 1e-06,
+      "loss": 0.1331,
+      "step": 434
+    },
+    {
+      "epoch": 0.07405515832482125,
+      "grad_norm": 1.486445665359497,
+      "learning_rate": 1e-06,
+      "loss": 0.1224,
+      "step": 435
+    },
+    {
+      "epoch": 0.07422540006809669,
+      "grad_norm": 1.301155686378479,
+      "learning_rate": 1e-06,
+      "loss": 0.1045,
+      "step": 436
+    },
+    {
+      "epoch": 0.07439564181137215,
+      "grad_norm": 1.8198158740997314,
+      "learning_rate": 1e-06,
+      "loss": 0.1324,
+      "step": 437
+    },
+    {
+      "epoch": 0.0745658835546476,
+      "grad_norm": 1.3100183010101318,
+      "learning_rate": 1e-06,
+      "loss": 0.1107,
+      "step": 438
+    },
+    {
+      "epoch": 0.07473612529792305,
+      "grad_norm": 1.3502755165100098,
+      "learning_rate": 1e-06,
+      "loss": 0.1082,
+      "step": 439
+    },
+    {
+      "epoch": 0.0749063670411985,
+      "grad_norm": 1.5028979778289795,
+      "learning_rate": 1e-06,
+      "loss": 0.1325,
+      "step": 440
+    },
+    {
+      "epoch": 0.07507660878447395,
+      "grad_norm": 1.3118284940719604,
+      "learning_rate": 1e-06,
+      "loss": 0.0964,
+      "step": 441
+    },
+    {
+      "epoch": 0.0752468505277494,
+      "grad_norm": 1.4086387157440186,
+      "learning_rate": 1e-06,
+      "loss": 0.094,
+      "step": 442
+    },
+    {
+      "epoch": 0.07541709227102486,
+      "grad_norm": 1.5664148330688477,
+      "learning_rate": 1e-06,
+      "loss": 0.1195,
+      "step": 443
+    },
+    {
+      "epoch": 0.0755873340143003,
+      "grad_norm": 1.4086772203445435,
+      "learning_rate": 1e-06,
+      "loss": 0.1104,
+      "step": 444
+    },
+    {
+      "epoch": 0.07575757575757576,
+      "grad_norm": 1.4743391275405884,
+      "learning_rate": 1e-06,
+      "loss": 0.1163,
+      "step": 445
+    },
+    {
+      "epoch": 0.0759278175008512,
+      "grad_norm": 1.600606083869934,
+      "learning_rate": 1e-06,
+      "loss": 0.1199,
+      "step": 446
+    },
+    {
+      "epoch": 0.07609805924412666,
+      "grad_norm": 1.3386887311935425,
+      "learning_rate": 1e-06,
+      "loss": 0.1372,
+      "step": 447
+    },
+    {
+      "epoch": 0.07626830098740212,
+      "grad_norm": 1.4845508337020874,
+      "learning_rate": 1e-06,
+      "loss": 0.1201,
+      "step": 448
+    },
+    {
+      "epoch": 0.07643854273067756,
+      "grad_norm": 1.503592610359192,
+      "learning_rate": 1e-06,
+      "loss": 0.1243,
+      "step": 449
+    },
+    {
+      "epoch": 0.07660878447395301,
+      "grad_norm": 1.3089518547058105,
+      "learning_rate": 1e-06,
+      "loss": 0.0884,
+      "step": 450
+    },
+    {
+      "epoch": 0.07677902621722846,
+      "grad_norm": 1.3585859537124634,
+      "learning_rate": 1e-06,
+      "loss": 0.0991,
+      "step": 451
+    },
+    {
+      "epoch": 0.07694926796050391,
+      "grad_norm": 3.5047550201416016,
+      "learning_rate": 1e-06,
+      "loss": 0.2257,
+      "step": 452
+    },
+    {
+      "epoch": 0.07711950970377937,
+      "grad_norm": 2.3448007106781006,
+      "learning_rate": 1e-06,
+      "loss": 0.1543,
+      "step": 453
+    },
+    {
+      "epoch": 0.07728975144705481,
+      "grad_norm": 1.8489949703216553,
+      "learning_rate": 1e-06,
+      "loss": 0.1335,
+      "step": 454
+    },
+    {
+      "epoch": 0.07745999319033027,
+      "grad_norm": 1.5891618728637695,
+      "learning_rate": 1e-06,
+      "loss": 0.108,
+      "step": 455
+    },
+    {
+      "epoch": 0.07763023493360573,
+      "grad_norm": 1.9977794885635376,
+      "learning_rate": 1e-06,
+      "loss": 0.1622,
+      "step": 456
+    },
+    {
+      "epoch": 0.07780047667688117,
+      "grad_norm": 1.131685495376587,
+      "learning_rate": 1e-06,
+      "loss": 0.0874,
+      "step": 457
+    },
+    {
+      "epoch": 0.07797071842015663,
+      "grad_norm": 1.4111319780349731,
+      "learning_rate": 1e-06,
+      "loss": 0.122,
+      "step": 458
+    },
+    {
+      "epoch": 0.07814096016343207,
+      "grad_norm": 1.3373247385025024,
+      "learning_rate": 1e-06,
+      "loss": 0.0998,
+      "step": 459
+    },
+    {
+      "epoch": 0.07831120190670753,
+      "grad_norm": 1.3671199083328247,
+      "learning_rate": 1e-06,
+      "loss": 0.1006,
+      "step": 460
+    },
+    {
+      "epoch": 0.07848144364998298,
+      "grad_norm": 1.4334213733673096,
+      "learning_rate": 1e-06,
+      "loss": 0.1115,
+      "step": 461
+    },
+    {
+      "epoch": 0.07865168539325842,
+      "grad_norm": 1.6302123069763184,
+      "learning_rate": 1e-06,
+      "loss": 0.1225,
+      "step": 462
+    },
+    {
+      "epoch": 0.07882192713653388,
+      "grad_norm": 1.5899648666381836,
+      "learning_rate": 1e-06,
+      "loss": 0.1196,
+      "step": 463
+    },
+    {
+      "epoch": 0.07899216887980932,
+      "grad_norm": 1.3022098541259766,
+      "learning_rate": 1e-06,
+      "loss": 0.0863,
+      "step": 464
+    },
+    {
+      "epoch": 0.07916241062308478,
+      "grad_norm": 1.5681778192520142,
+      "learning_rate": 1e-06,
+      "loss": 0.1,
+      "step": 465
+    },
+    {
+      "epoch": 0.07933265236636024,
+      "grad_norm": 1.5576342344284058,
+      "learning_rate": 1e-06,
+      "loss": 0.1098,
+      "step": 466
+    },
+    {
+      "epoch": 0.07950289410963568,
+      "grad_norm": 1.1382691860198975,
+      "learning_rate": 1e-06,
+      "loss": 0.0896,
+      "step": 467
+    },
+    {
+      "epoch": 0.07967313585291114,
+      "grad_norm": 1.551830768585205,
+      "learning_rate": 1e-06,
+      "loss": 0.1475,
+      "step": 468
+    },
+    {
+      "epoch": 0.07984337759618658,
+      "grad_norm": 1.4834508895874023,
+      "learning_rate": 1e-06,
+      "loss": 0.1331,
+      "step": 469
+    },
+    {
+      "epoch": 0.08001361933946204,
+      "grad_norm": 1.7695828676223755,
+      "learning_rate": 1e-06,
+      "loss": 0.1544,
+      "step": 470
+    },
+    {
+      "epoch": 0.08018386108273749,
+      "grad_norm": 2.0113699436187744,
+      "learning_rate": 1e-06,
+      "loss": 0.1643,
+      "step": 471
+    },
+    {
+      "epoch": 0.08035410282601294,
+      "grad_norm": 1.7527496814727783,
+      "learning_rate": 1e-06,
+      "loss": 0.1529,
+      "step": 472
+    },
+    {
+      "epoch": 0.08052434456928839,
+      "grad_norm": 1.2186193466186523,
+      "learning_rate": 1e-06,
+      "loss": 0.0892,
+      "step": 473
+    },
+    {
+      "epoch": 0.08069458631256383,
+      "grad_norm": 1.2762824296951294,
+      "learning_rate": 1e-06,
+      "loss": 0.0839,
+      "step": 474
+    },
+    {
+      "epoch": 0.08086482805583929,
+      "grad_norm": 1.3743529319763184,
+      "learning_rate": 1e-06,
+      "loss": 0.1033,
+      "step": 475
+    },
+    {
+      "epoch": 0.08103506979911475,
+      "grad_norm": 1.5724990367889404,
+      "learning_rate": 1e-06,
+      "loss": 0.104,
+      "step": 476
+    },
+    {
+      "epoch": 0.08120531154239019,
+      "grad_norm": 1.5919311046600342,
+      "learning_rate": 1e-06,
+      "loss": 0.0943,
+      "step": 477
+    },
+    {
+      "epoch": 0.08137555328566565,
+      "grad_norm": 1.5781210660934448,
+      "learning_rate": 1e-06,
+      "loss": 0.1012,
+      "step": 478
+    },
+    {
+      "epoch": 0.08154579502894109,
+      "grad_norm": 1.8504365682601929,
+      "learning_rate": 1e-06,
+      "loss": 0.1361,
+      "step": 479
+    },
+    {
+      "epoch": 0.08171603677221655,
+      "grad_norm": 1.4634361267089844,
+      "learning_rate": 1e-06,
+      "loss": 0.1075,
+      "step": 480
+    },
+    {
+      "epoch": 0.081886278515492,
+      "grad_norm": 1.2929086685180664,
+      "learning_rate": 1e-06,
+      "loss": 0.0966,
+      "step": 481
+    },
+    {
+      "epoch": 0.08205652025876745,
+      "grad_norm": 3.413325786590576,
+      "learning_rate": 1e-06,
+      "loss": 0.2179,
+      "step": 482
+    },
+    {
+      "epoch": 0.0822267620020429,
+      "grad_norm": 1.2101235389709473,
+      "learning_rate": 1e-06,
+      "loss": 0.0915,
+      "step": 483
+    },
+    {
+      "epoch": 0.08239700374531835,
+      "grad_norm": 1.7361688613891602,
+      "learning_rate": 1e-06,
+      "loss": 0.121,
+      "step": 484
+    },
+    {
+      "epoch": 0.0825672454885938,
+      "grad_norm": 1.4495158195495605,
+      "learning_rate": 1e-06,
+      "loss": 0.1387,
+      "step": 485
+    },
+    {
+      "epoch": 0.08273748723186926,
+      "grad_norm": 1.3097033500671387,
+      "learning_rate": 1e-06,
+      "loss": 0.0994,
+      "step": 486
+    },
+    {
+      "epoch": 0.0829077289751447,
+      "grad_norm": 1.6425195932388306,
+      "learning_rate": 1e-06,
+      "loss": 0.1233,
+      "step": 487
+    },
+    {
+      "epoch": 0.08307797071842016,
+      "grad_norm": 1.344718098640442,
+      "learning_rate": 1e-06,
+      "loss": 0.107,
+      "step": 488
+    },
+    {
+      "epoch": 0.08324821246169561,
+      "grad_norm": 1.1989573240280151,
+      "learning_rate": 1e-06,
+      "loss": 0.0888,
+      "step": 489
+    },
+    {
+      "epoch": 0.08341845420497106,
+      "grad_norm": 1.4948607683181763,
+      "learning_rate": 1e-06,
+      "loss": 0.1105,
+      "step": 490
+    },
+    {
+      "epoch": 0.08358869594824651,
+      "grad_norm": 1.4378379583358765,
+      "learning_rate": 1e-06,
+      "loss": 0.106,
+      "step": 491
+    },
+    {
+      "epoch": 0.08375893769152196,
+      "grad_norm": 1.2525907754898071,
+      "learning_rate": 1e-06,
+      "loss": 0.0968,
+      "step": 492
+    },
+    {
+      "epoch": 0.08392917943479741,
+      "grad_norm": 1.479597806930542,
+      "learning_rate": 1e-06,
+      "loss": 0.1217,
+      "step": 493
+    },
+    {
+      "epoch": 0.08409942117807287,
+      "grad_norm": 1.3180419206619263,
+      "learning_rate": 1e-06,
+      "loss": 0.1011,
+      "step": 494
+    },
+    {
+      "epoch": 0.08426966292134831,
+      "grad_norm": 1.3772739171981812,
+      "learning_rate": 1e-06,
+      "loss": 0.1049,
+      "step": 495
+    },
+    {
+      "epoch": 0.08443990466462377,
+      "grad_norm": 1.814520001411438,
+      "learning_rate": 1e-06,
+      "loss": 0.1293,
+      "step": 496
+    },
+    {
+      "epoch": 0.08461014640789921,
+      "grad_norm": 1.5137838125228882,
+      "learning_rate": 1e-06,
+      "loss": 0.0999,
+      "step": 497
+    },
+    {
+      "epoch": 0.08478038815117467,
+      "grad_norm": 1.4192203283309937,
+      "learning_rate": 1e-06,
+      "loss": 0.114,
+      "step": 498
+    },
+    {
+      "epoch": 0.08495062989445012,
+      "grad_norm": 1.3078948259353638,
+      "learning_rate": 1e-06,
+      "loss": 0.1158,
+      "step": 499
+    },
+    {
+      "epoch": 0.08512087163772557,
+      "grad_norm": 1.5848225355148315,
+      "learning_rate": 1e-06,
+      "loss": 0.1163,
+      "step": 500
+    },
+    {
+      "epoch": 0.08529111338100102,
+      "grad_norm": 1.3920193910598755,
+      "learning_rate": 1e-06,
+      "loss": 0.0995,
+      "step": 501
+    },
+    {
+      "epoch": 0.08546135512427647,
+      "grad_norm": 1.3783040046691895,
+      "learning_rate": 1e-06,
+      "loss": 0.0955,
+      "step": 502
+    },
+    {
+      "epoch": 0.08563159686755192,
+      "grad_norm": 1.8510295152664185,
+      "learning_rate": 1e-06,
+      "loss": 0.1328,
+      "step": 503
+    },
+    {
+      "epoch": 0.08580183861082738,
+      "grad_norm": 3.0699658393859863,
+      "learning_rate": 1e-06,
+      "loss": 0.1774,
+      "step": 504
+    },
+    {
+      "epoch": 0.08597208035410282,
+      "grad_norm": 1.5914885997772217,
+      "learning_rate": 1e-06,
+      "loss": 0.1055,
+      "step": 505
+    },
+    {
+      "epoch": 0.08614232209737828,
+      "grad_norm": 1.7265233993530273,
+      "learning_rate": 1e-06,
+      "loss": 0.1271,
+      "step": 506
+    },
+    {
+      "epoch": 0.08631256384065372,
+      "grad_norm": 1.5482544898986816,
+      "learning_rate": 1e-06,
+      "loss": 0.1279,
+      "step": 507
+    },
+    {
+      "epoch": 0.08648280558392918,
+      "grad_norm": 1.321733832359314,
+      "learning_rate": 1e-06,
+      "loss": 0.0937,
+      "step": 508
+    },
+    {
+      "epoch": 0.08665304732720464,
+      "grad_norm": 1.4876199960708618,
+      "learning_rate": 1e-06,
+      "loss": 0.1192,
+      "step": 509
+    },
+    {
+      "epoch": 0.08682328907048008,
+      "grad_norm": 1.8378245830535889,
+      "learning_rate": 1e-06,
+      "loss": 0.1157,
+      "step": 510
+    },
+    {
+      "epoch": 0.08699353081375553,
+      "grad_norm": 1.625184178352356,
+      "learning_rate": 1e-06,
+      "loss": 0.1206,
+      "step": 511
+    },
+    {
+      "epoch": 0.08716377255703098,
+      "grad_norm": 1.4779977798461914,
+      "learning_rate": 1e-06,
+      "loss": 0.1103,
+      "step": 512
+    },
+    {
+      "epoch": 0.08733401430030643,
+      "grad_norm": 1.4066017866134644,
+      "learning_rate": 1e-06,
+      "loss": 0.1011,
+      "step": 513
+    },
+    {
+      "epoch": 0.08750425604358189,
+      "grad_norm": 1.525978922843933,
+      "learning_rate": 1e-06,
+      "loss": 0.096,
+      "step": 514
+    },
+    {
+      "epoch": 0.08767449778685733,
+      "grad_norm": 1.49380362033844,
+      "learning_rate": 1e-06,
+      "loss": 0.1135,
+      "step": 515
+    },
+    {
+      "epoch": 0.08784473953013279,
+      "grad_norm": 1.74233078956604,
+      "learning_rate": 1e-06,
+      "loss": 0.1114,
+      "step": 516
+    },
+    {
+      "epoch": 0.08801498127340825,
+      "grad_norm": 1.3099770545959473,
+      "learning_rate": 1e-06,
+      "loss": 0.0907,
+      "step": 517
+    },
+    {
+      "epoch": 0.08818522301668369,
+      "grad_norm": 1.7866302728652954,
+      "learning_rate": 1e-06,
+      "loss": 0.1178,
+      "step": 518
+    },
+    {
+      "epoch": 0.08835546475995915,
+      "grad_norm": 1.6818584203720093,
+      "learning_rate": 1e-06,
+      "loss": 0.1252,
+      "step": 519
+    },
+    {
+      "epoch": 0.08852570650323459,
+      "grad_norm": 1.628466248512268,
+      "learning_rate": 1e-06,
+      "loss": 0.1105,
+      "step": 520
+    },
+    {
+      "epoch": 0.08869594824651005,
+      "grad_norm": 1.681505799293518,
+      "learning_rate": 1e-06,
+      "loss": 0.1177,
+      "step": 521
+    },
+    {
+      "epoch": 0.0888661899897855,
+      "grad_norm": 1.2766761779785156,
+      "learning_rate": 1e-06,
+      "loss": 0.0799,
+      "step": 522
+    },
+    {
+      "epoch": 0.08903643173306094,
+      "grad_norm": 1.575498342514038,
+      "learning_rate": 1e-06,
+      "loss": 0.1068,
+      "step": 523
+    },
+    {
+      "epoch": 0.0892066734763364,
+      "grad_norm": 1.4087481498718262,
+      "learning_rate": 1e-06,
+      "loss": 0.0935,
+      "step": 524
+    },
+    {
+      "epoch": 0.08937691521961184,
+      "grad_norm": 1.4377367496490479,
+      "learning_rate": 1e-06,
+      "loss": 0.0968,
+      "step": 525
+    },
+    {
+      "epoch": 0.0895471569628873,
+      "grad_norm": 1.7580372095108032,
+      "learning_rate": 1e-06,
+      "loss": 0.1101,
+      "step": 526
+    },
+    {
+      "epoch": 0.08971739870616276,
+      "grad_norm": 1.5135084390640259,
+      "learning_rate": 1e-06,
+      "loss": 0.1005,
+      "step": 527
+    },
+    {
+      "epoch": 0.0898876404494382,
+      "grad_norm": 1.4128092527389526,
+      "learning_rate": 1e-06,
+      "loss": 0.0858,
+      "step": 528
+    },
+    {
+      "epoch": 0.09005788219271366,
+      "grad_norm": 1.4026193618774414,
+      "learning_rate": 1e-06,
+      "loss": 0.1002,
+      "step": 529
+    },
+    {
+      "epoch": 0.0902281239359891,
+      "grad_norm": 1.3924407958984375,
+      "learning_rate": 1e-06,
+      "loss": 0.0895,
+      "step": 530
+    },
+    {
+      "epoch": 0.09039836567926456,
+      "grad_norm": 1.4779433012008667,
+      "learning_rate": 1e-06,
+      "loss": 0.1033,
+      "step": 531
+    },
+    {
+      "epoch": 0.09056860742254001,
+      "grad_norm": 1.9175746440887451,
+      "learning_rate": 1e-06,
+      "loss": 0.1302,
+      "step": 532
+    },
+    {
+      "epoch": 0.09073884916581546,
+      "grad_norm": 1.7925680875778198,
+      "learning_rate": 1e-06,
+      "loss": 0.1162,
+      "step": 533
+    },
+    {
+      "epoch": 0.09090909090909091,
+      "grad_norm": 1.3707115650177002,
+      "learning_rate": 1e-06,
+      "loss": 0.0872,
+      "step": 534
+    },
+    {
+      "epoch": 0.09107933265236635,
+      "grad_norm": 1.3779429197311401,
+      "learning_rate": 1e-06,
+      "loss": 0.0982,
+      "step": 535
+    },
+    {
+      "epoch": 0.09124957439564181,
+      "grad_norm": 1.5113381147384644,
+      "learning_rate": 1e-06,
+      "loss": 0.1036,
+      "step": 536
+    },
+    {
+      "epoch": 0.09141981613891727,
+      "grad_norm": 1.5571949481964111,
+      "learning_rate": 1e-06,
+      "loss": 0.0933,
+      "step": 537
+    },
+    {
+      "epoch": 0.09159005788219271,
+      "grad_norm": 1.3230814933776855,
+      "learning_rate": 1e-06,
+      "loss": 0.1019,
+      "step": 538
+    },
+    {
+      "epoch": 0.09176029962546817,
+      "grad_norm": 1.257027506828308,
+      "learning_rate": 1e-06,
+      "loss": 0.0775,
+      "step": 539
+    },
+    {
+      "epoch": 0.09193054136874361,
+      "grad_norm": 1.3978290557861328,
+      "learning_rate": 1e-06,
+      "loss": 0.0942,
+      "step": 540
+    },
+    {
+      "epoch": 0.09210078311201907,
+      "grad_norm": 1.290602207183838,
+      "learning_rate": 1e-06,
+      "loss": 0.0809,
+      "step": 541
+    },
+    {
+      "epoch": 0.09227102485529452,
+      "grad_norm": 1.3179094791412354,
+      "learning_rate": 1e-06,
+      "loss": 0.0963,
+      "step": 542
+    },
+    {
+      "epoch": 0.09244126659856997,
+      "grad_norm": 1.6563512086868286,
+      "learning_rate": 1e-06,
+      "loss": 0.111,
+      "step": 543
+    },
+    {
+      "epoch": 0.09261150834184542,
+      "grad_norm": 1.4956837892532349,
+      "learning_rate": 1e-06,
+      "loss": 0.1165,
+      "step": 544
+    },
+    {
+      "epoch": 0.09278175008512086,
+      "grad_norm": 1.3757632970809937,
+      "learning_rate": 1e-06,
+      "loss": 0.0897,
+      "step": 545
+    },
+    {
+      "epoch": 0.09295199182839632,
+      "grad_norm": 1.205857276916504,
+      "learning_rate": 1e-06,
+      "loss": 0.0787,
+      "step": 546
+    },
+    {
+      "epoch": 0.09312223357167178,
+      "grad_norm": 1.7391396760940552,
+      "learning_rate": 1e-06,
+      "loss": 0.1286,
+      "step": 547
+    },
+    {
+      "epoch": 0.09329247531494722,
+      "grad_norm": 1.492804765701294,
+      "learning_rate": 1e-06,
+      "loss": 0.0916,
+      "step": 548
+    },
+    {
+      "epoch": 0.09346271705822268,
+      "grad_norm": 1.6261539459228516,
+      "learning_rate": 1e-06,
+      "loss": 0.0995,
+      "step": 549
+    },
+    {
+      "epoch": 0.09363295880149813,
+      "grad_norm": 1.5201900005340576,
+      "learning_rate": 1e-06,
+      "loss": 0.0987,
+      "step": 550
+    },
+    {
+      "epoch": 0.09380320054477358,
+      "grad_norm": 1.3725306987762451,
+      "learning_rate": 1e-06,
+      "loss": 0.0958,
+      "step": 551
+    },
+    {
+      "epoch": 0.09397344228804903,
+      "grad_norm": 1.601542592048645,
+      "learning_rate": 1e-06,
+      "loss": 0.1168,
+      "step": 552
+    },
+    {
+      "epoch": 0.09414368403132448,
+      "grad_norm": 1.4362895488739014,
+      "learning_rate": 1e-06,
+      "loss": 0.0896,
+      "step": 553
+    },
+    {
+      "epoch": 0.09431392577459993,
+      "grad_norm": 1.3668473958969116,
+      "learning_rate": 1e-06,
+      "loss": 0.1029,
+      "step": 554
+    },
+    {
+      "epoch": 0.09448416751787539,
+      "grad_norm": 1.6272999048233032,
+      "learning_rate": 1e-06,
+      "loss": 0.0938,
+      "step": 555
+    },
+    {
+      "epoch": 0.09465440926115083,
+      "grad_norm": 1.8201227188110352,
+      "learning_rate": 1e-06,
+      "loss": 0.1186,
+      "step": 556
+    },
+    {
+      "epoch": 0.09482465100442629,
+      "grad_norm": 1.4684869050979614,
+      "learning_rate": 1e-06,
+      "loss": 0.1029,
+      "step": 557
+    },
+    {
+      "epoch": 0.09499489274770173,
+      "grad_norm": 1.4213141202926636,
+      "learning_rate": 1e-06,
+      "loss": 0.0875,
+      "step": 558
+    },
+    {
+      "epoch": 0.09516513449097719,
+      "grad_norm": 1.4483574628829956,
+      "learning_rate": 1e-06,
+      "loss": 0.0996,
+      "step": 559
+    },
+    {
+      "epoch": 0.09533537623425264,
+      "grad_norm": 1.4041225910186768,
+      "learning_rate": 1e-06,
+      "loss": 0.0769,
+      "step": 560
+    },
+    {
+      "epoch": 0.09550561797752809,
+      "grad_norm": 1.3655953407287598,
+      "learning_rate": 1e-06,
+      "loss": 0.0927,
+      "step": 561
+    },
+    {
+      "epoch": 0.09567585972080354,
+      "grad_norm": 1.2655277252197266,
+      "learning_rate": 1e-06,
+      "loss": 0.078,
+      "step": 562
+    },
+    {
+      "epoch": 0.09584610146407899,
+      "grad_norm": 1.6605805158615112,
+      "learning_rate": 1e-06,
+      "loss": 0.1101,
+      "step": 563
+    },
+    {
+      "epoch": 0.09601634320735444,
+      "grad_norm": 1.6565370559692383,
+      "learning_rate": 1e-06,
+      "loss": 0.1107,
+      "step": 564
+    },
+    {
+      "epoch": 0.0961865849506299,
+      "grad_norm": 1.5182013511657715,
+      "learning_rate": 1e-06,
+      "loss": 0.0992,
+      "step": 565
+    },
+    {
+      "epoch": 0.09635682669390534,
+      "grad_norm": 1.3900082111358643,
+      "learning_rate": 1e-06,
+      "loss": 0.0999,
+      "step": 566
+    },
+    {
+      "epoch": 0.0965270684371808,
+      "grad_norm": 1.2898858785629272,
+      "learning_rate": 1e-06,
+      "loss": 0.0906,
+      "step": 567
+    },
+    {
+      "epoch": 0.09669731018045624,
+      "grad_norm": 3.67594838142395,
+      "learning_rate": 1e-06,
+      "loss": 0.1686,
+      "step": 568
+    },
+    {
+      "epoch": 0.0968675519237317,
+      "grad_norm": 1.313101887702942,
+      "learning_rate": 1e-06,
+      "loss": 0.0799,
+      "step": 569
+    },
+    {
+      "epoch": 0.09703779366700716,
+      "grad_norm": 3.672680139541626,
+      "learning_rate": 1e-06,
+      "loss": 0.2146,
+      "step": 570
+    },
+    {
+      "epoch": 0.0972080354102826,
+      "grad_norm": 1.770923376083374,
+      "learning_rate": 1e-06,
+      "loss": 0.0999,
+      "step": 571
+    },
+    {
+      "epoch": 0.09737827715355805,
+      "grad_norm": 1.533212661743164,
+      "learning_rate": 1e-06,
+      "loss": 0.1004,
+      "step": 572
+    },
+    {
+      "epoch": 0.0975485188968335,
+      "grad_norm": 1.6698973178863525,
+      "learning_rate": 1e-06,
+      "loss": 0.0945,
+      "step": 573
+    },
+    {
+      "epoch": 0.09771876064010895,
+      "grad_norm": 1.4515860080718994,
+      "learning_rate": 1e-06,
+      "loss": 0.1065,
+      "step": 574
+    },
+    {
+      "epoch": 0.09788900238338441,
+      "grad_norm": 1.332733154296875,
+      "learning_rate": 1e-06,
+      "loss": 0.0875,
+      "step": 575
+    },
+    {
+      "epoch": 0.09805924412665985,
+      "grad_norm": 1.3322087526321411,
+      "learning_rate": 1e-06,
+      "loss": 0.0872,
+      "step": 576
+    },
+    {
+      "epoch": 0.09822948586993531,
+      "grad_norm": 1.320599913597107,
+      "learning_rate": 1e-06,
+      "loss": 0.0832,
+      "step": 577
+    },
+    {
+      "epoch": 0.09839972761321075,
+      "grad_norm": 1.4648524522781372,
+      "learning_rate": 1e-06,
+      "loss": 0.0963,
+      "step": 578
+    },
+    {
+      "epoch": 0.09856996935648621,
+      "grad_norm": 1.6072378158569336,
+      "learning_rate": 1e-06,
+      "loss": 0.09,
+      "step": 579
+    },
+    {
+      "epoch": 0.09874021109976167,
+      "grad_norm": 2.066091537475586,
+      "learning_rate": 1e-06,
+      "loss": 0.1273,
+      "step": 580
+    },
+    {
+      "epoch": 0.09891045284303711,
+      "grad_norm": 1.3935089111328125,
+      "learning_rate": 1e-06,
+      "loss": 0.0938,
+      "step": 581
+    },
+    {
+      "epoch": 0.09908069458631256,
+      "grad_norm": 1.4537439346313477,
+      "learning_rate": 1e-06,
+      "loss": 0.0908,
+      "step": 582
+    },
+    {
+      "epoch": 0.09925093632958802,
+      "grad_norm": 1.4111661911010742,
+      "learning_rate": 1e-06,
+      "loss": 0.0826,
+      "step": 583
+    },
+    {
+      "epoch": 0.09942117807286346,
+      "grad_norm": 1.417158603668213,
+      "learning_rate": 1e-06,
+      "loss": 0.0916,
+      "step": 584
+    },
+    {
+      "epoch": 0.09959141981613892,
+      "grad_norm": 1.5110013484954834,
+      "learning_rate": 1e-06,
+      "loss": 0.0842,
+      "step": 585
+    },
+    {
+      "epoch": 0.09976166155941436,
+      "grad_norm": 1.3743090629577637,
+      "learning_rate": 1e-06,
+      "loss": 0.0891,
+      "step": 586
+    },
+    {
+      "epoch": 0.09993190330268982,
+      "grad_norm": 1.4649263620376587,
+      "learning_rate": 1e-06,
+      "loss": 0.0791,
+      "step": 587
+    },
+    {
+      "epoch": 0.10010214504596528,
+      "grad_norm": 1.6102653741836548,
+      "learning_rate": 1e-06,
+      "loss": 0.0856,
+      "step": 588
+    },
+    {
+      "epoch": 0.10027238678924072,
+      "grad_norm": 1.7673293352127075,
+      "learning_rate": 1e-06,
+      "loss": 0.1013,
+      "step": 589
+    },
+    {
+      "epoch": 0.10044262853251618,
+      "grad_norm": 1.4860748052597046,
+      "learning_rate": 1e-06,
+      "loss": 0.0854,
+      "step": 590
+    },
+    {
+      "epoch": 0.10061287027579162,
+      "grad_norm": 1.5368369817733765,
+      "learning_rate": 1e-06,
+      "loss": 0.089,
+      "step": 591
+    },
+    {
+      "epoch": 0.10078311201906708,
+      "grad_norm": 1.6211143732070923,
+      "learning_rate": 1e-06,
+      "loss": 0.1087,
+      "step": 592
+    },
+    {
+      "epoch": 0.10095335376234253,
+      "grad_norm": 1.432572603225708,
+      "learning_rate": 1e-06,
+      "loss": 0.0934,
+      "step": 593
+    },
+    {
+      "epoch": 0.10112359550561797,
+      "grad_norm": 1.351190209388733,
+      "learning_rate": 1e-06,
+      "loss": 0.0816,
+      "step": 594
+    },
+    {
+      "epoch": 0.10129383724889343,
+      "grad_norm": 1.5220497846603394,
+      "learning_rate": 1e-06,
+      "loss": 0.1014,
+      "step": 595
+    },
+    {
+      "epoch": 0.10146407899216887,
+      "grad_norm": 3.933367967605591,
+      "learning_rate": 1e-06,
+      "loss": 0.2032,
+      "step": 596
+    },
+    {
+      "epoch": 0.10163432073544433,
+      "grad_norm": 1.5819555521011353,
+      "learning_rate": 1e-06,
+      "loss": 0.1125,
+      "step": 597
+    },
+    {
+      "epoch": 0.10180456247871979,
+      "grad_norm": 1.5007481575012207,
+      "learning_rate": 1e-06,
+      "loss": 0.0991,
+      "step": 598
+    },
+    {
+      "epoch": 0.10197480422199523,
+      "grad_norm": 1.6869536638259888,
+      "learning_rate": 1e-06,
+      "loss": 0.1113,
+      "step": 599
+    },
+    {
+      "epoch": 0.10214504596527069,
+      "grad_norm": 1.5548107624053955,
+      "learning_rate": 1e-06,
+      "loss": 0.1018,
+      "step": 600
+    },
+    {
+      "epoch": 0.10231528770854613,
+      "grad_norm": 1.7281802892684937,
+      "learning_rate": 1e-06,
+      "loss": 0.0962,
+      "step": 601
+    },
+    {
+      "epoch": 0.10248552945182159,
+      "grad_norm": 1.603119969367981,
+      "learning_rate": 1e-06,
+      "loss": 0.0764,
+      "step": 602
+    },
+    {
+      "epoch": 0.10265577119509704,
+      "grad_norm": 1.6384819746017456,
+      "learning_rate": 1e-06,
+      "loss": 0.0842,
+      "step": 603
+    },
+    {
+      "epoch": 0.10282601293837249,
+      "grad_norm": 1.245998501777649,
+      "learning_rate": 1e-06,
+      "loss": 0.086,
+      "step": 604
+    },
+    {
+      "epoch": 0.10299625468164794,
+      "grad_norm": 1.5214192867279053,
+      "learning_rate": 1e-06,
+      "loss": 0.0834,
+      "step": 605
+    },
+    {
+      "epoch": 0.10316649642492338,
+      "grad_norm": 1.345001459121704,
+      "learning_rate": 1e-06,
+      "loss": 0.0869,
+      "step": 606
+    },
+    {
+      "epoch": 0.10333673816819884,
+      "grad_norm": 3.7370526790618896,
+      "learning_rate": 1e-06,
+      "loss": 0.1794,
+      "step": 607
+    },
+    {
+      "epoch": 0.1035069799114743,
+      "grad_norm": 1.3892709016799927,
+      "learning_rate": 1e-06,
+      "loss": 0.0912,
+      "step": 608
+    },
+    {
+      "epoch": 0.10367722165474974,
+      "grad_norm": 1.4373610019683838,
+      "learning_rate": 1e-06,
+      "loss": 0.0914,
+      "step": 609
+    },
+    {
+      "epoch": 0.1038474633980252,
+      "grad_norm": 1.4336578845977783,
+      "learning_rate": 1e-06,
+      "loss": 0.0732,
+      "step": 610
+    },
+    {
+      "epoch": 0.10401770514130065,
+      "grad_norm": 1.7176496982574463,
+      "learning_rate": 1e-06,
+      "loss": 0.1008,
+      "step": 611
+    },
+    {
+      "epoch": 0.1041879468845761,
+      "grad_norm": 1.9851466417312622,
+      "learning_rate": 1e-06,
+      "loss": 0.1067,
+      "step": 612
+    },
+    {
+      "epoch": 0.10435818862785155,
+      "grad_norm": 1.8875466585159302,
+      "learning_rate": 1e-06,
+      "loss": 0.1106,
+      "step": 613
+    },
+    {
+      "epoch": 0.104528430371127,
+      "grad_norm": 1.458789348602295,
+      "learning_rate": 1e-06,
+      "loss": 0.0843,
+      "step": 614
+    },
+    {
+      "epoch": 0.10469867211440245,
+      "grad_norm": 1.6216390132904053,
+      "learning_rate": 1e-06,
+      "loss": 0.09,
+      "step": 615
+    },
+    {
+      "epoch": 0.10486891385767791,
+      "grad_norm": 1.2169405221939087,
+      "learning_rate": 1e-06,
+      "loss": 0.0626,
+      "step": 616
+    },
+    {
+      "epoch": 0.10503915560095335,
+      "grad_norm": 1.3988337516784668,
+      "learning_rate": 1e-06,
+      "loss": 0.0891,
+      "step": 617
+    },
+    {
+      "epoch": 0.10520939734422881,
+      "grad_norm": 1.5023198127746582,
+      "learning_rate": 1e-06,
+      "loss": 0.0968,
+      "step": 618
+    },
+    {
+      "epoch": 0.10537963908750425,
+      "grad_norm": 1.4495683908462524,
+      "learning_rate": 1e-06,
+      "loss": 0.0865,
+      "step": 619
+    },
+    {
+      "epoch": 0.10554988083077971,
+      "grad_norm": 1.256415843963623,
+      "learning_rate": 1e-06,
+      "loss": 0.0713,
+      "step": 620
+    },
+    {
+      "epoch": 0.10572012257405516,
+      "grad_norm": 1.7719178199768066,
+      "learning_rate": 1e-06,
+      "loss": 0.0998,
+      "step": 621
+    },
+    {
+      "epoch": 0.10589036431733061,
+      "grad_norm": 1.5151079893112183,
+      "learning_rate": 1e-06,
+      "loss": 0.0812,
+      "step": 622
+    },
+    {
+      "epoch": 0.10606060606060606,
+      "grad_norm": 4.284844875335693,
+      "learning_rate": 1e-06,
+      "loss": 0.178,
+      "step": 623
+    },
+    {
+      "epoch": 0.1062308478038815,
+      "grad_norm": 1.978816032409668,
+      "learning_rate": 1e-06,
+      "loss": 0.0872,
+      "step": 624
+    },
+    {
+      "epoch": 0.10640108954715696,
+      "grad_norm": 1.5746681690216064,
+      "learning_rate": 1e-06,
+      "loss": 0.0849,
+      "step": 625
+    },
+    {
+      "epoch": 0.10657133129043242,
+      "grad_norm": 1.4485297203063965,
+      "learning_rate": 1e-06,
+      "loss": 0.081,
+      "step": 626
+    },
+    {
+      "epoch": 0.10674157303370786,
+      "grad_norm": 1.581984281539917,
+      "learning_rate": 1e-06,
+      "loss": 0.1037,
+      "step": 627
+    },
+    {
+      "epoch": 0.10691181477698332,
+      "grad_norm": 1.3287845849990845,
+      "learning_rate": 1e-06,
+      "loss": 0.0722,
+      "step": 628
+    },
+    {
+      "epoch": 0.10708205652025876,
+      "grad_norm": 1.2169119119644165,
+      "learning_rate": 1e-06,
+      "loss": 0.0588,
+      "step": 629
+    },
+    {
+      "epoch": 0.10725229826353422,
+      "grad_norm": 1.2675771713256836,
+      "learning_rate": 1e-06,
+      "loss": 0.0792,
+      "step": 630
+    },
+    {
+      "epoch": 0.10742254000680967,
+      "grad_norm": 1.3391927480697632,
+      "learning_rate": 1e-06,
+      "loss": 0.0884,
+      "step": 631
+    },
+    {
+      "epoch": 0.10759278175008512,
+      "grad_norm": 1.7200751304626465,
+      "learning_rate": 1e-06,
+      "loss": 0.0922,
+      "step": 632
+    },
+    {
+      "epoch": 0.10776302349336057,
+      "grad_norm": 1.4246052503585815,
+      "learning_rate": 1e-06,
+      "loss": 0.08,
+      "step": 633
+    },
+    {
+      "epoch": 0.10793326523663602,
+      "grad_norm": 1.4974547624588013,
+      "learning_rate": 1e-06,
+      "loss": 0.0738,
+      "step": 634
+    },
+    {
+      "epoch": 0.10810350697991147,
+      "grad_norm": 1.7298548221588135,
+      "learning_rate": 1e-06,
+      "loss": 0.0869,
+      "step": 635
+    },
+    {
+      "epoch": 0.10827374872318693,
+      "grad_norm": 1.4626245498657227,
+      "learning_rate": 1e-06,
+      "loss": 0.0745,
+      "step": 636
+    },
+    {
+      "epoch": 0.10844399046646237,
+      "grad_norm": 1.5561753511428833,
+      "learning_rate": 1e-06,
+      "loss": 0.0807,
+      "step": 637
+    },
+    {
+      "epoch": 0.10861423220973783,
+      "grad_norm": 1.6584053039550781,
+      "learning_rate": 1e-06,
+      "loss": 0.0951,
+      "step": 638
+    },
+    {
+      "epoch": 0.10878447395301327,
+      "grad_norm": 1.3954371213912964,
+      "learning_rate": 1e-06,
+      "loss": 0.0754,
+      "step": 639
+    },
+    {
+      "epoch": 0.10895471569628873,
+      "grad_norm": 3.7084250450134277,
+      "learning_rate": 1e-06,
+      "loss": 0.1604,
+      "step": 640
+    },
+    {
+      "epoch": 0.10912495743956419,
+      "grad_norm": 1.594537377357483,
+      "learning_rate": 1e-06,
+      "loss": 0.1014,
+      "step": 641
+    },
+    {
+      "epoch": 0.10929519918283963,
+      "grad_norm": 3.0633435249328613,
+      "learning_rate": 1e-06,
+      "loss": 0.1342,
+      "step": 642
+    },
+    {
+      "epoch": 0.10946544092611508,
+      "grad_norm": 1.6951473951339722,
+      "learning_rate": 1e-06,
+      "loss": 0.0925,
+      "step": 643
+    },
+    {
+      "epoch": 0.10963568266939054,
+      "grad_norm": 1.4243108034133911,
+      "learning_rate": 1e-06,
+      "loss": 0.0696,
+      "step": 644
+    },
+    {
+      "epoch": 0.10980592441266598,
+      "grad_norm": 1.480381965637207,
+      "learning_rate": 1e-06,
+      "loss": 0.0768,
+      "step": 645
+    },
+    {
+      "epoch": 0.10997616615594144,
+      "grad_norm": 1.3805941343307495,
+      "learning_rate": 1e-06,
+      "loss": 0.0848,
+      "step": 646
+    },
+    {
+      "epoch": 0.11014640789921688,
+      "grad_norm": 1.4682902097702026,
+      "learning_rate": 1e-06,
+      "loss": 0.0722,
+      "step": 647
+    },
+    {
+      "epoch": 0.11031664964249234,
+      "grad_norm": 1.408152461051941,
+      "learning_rate": 1e-06,
+      "loss": 0.0734,
+      "step": 648
+    },
+    {
+      "epoch": 0.1104868913857678,
+      "grad_norm": 1.8382225036621094,
+      "learning_rate": 1e-06,
+      "loss": 0.1047,
+      "step": 649
+    },
+    {
+      "epoch": 0.11065713312904324,
+      "grad_norm": 1.4985864162445068,
+      "learning_rate": 1e-06,
+      "loss": 0.0845,
+      "step": 650
+    },
+    {
+      "epoch": 0.1108273748723187,
+      "grad_norm": 1.340008020401001,
+      "learning_rate": 1e-06,
+      "loss": 0.0741,
+      "step": 651
+    },
+    {
+      "epoch": 0.11099761661559414,
+      "grad_norm": 1.65091073513031,
+      "learning_rate": 1e-06,
+      "loss": 0.0914,
+      "step": 652
+    },
+    {
+      "epoch": 0.1111678583588696,
+      "grad_norm": 1.6151150465011597,
+      "learning_rate": 1e-06,
+      "loss": 0.084,
+      "step": 653
+    },
+    {
+      "epoch": 0.11133810010214505,
+      "grad_norm": 1.8154164552688599,
+      "learning_rate": 1e-06,
+      "loss": 0.0988,
+      "step": 654
+    },
+    {
+      "epoch": 0.1115083418454205,
+      "grad_norm": 1.4415311813354492,
+      "learning_rate": 1e-06,
+      "loss": 0.0841,
+      "step": 655
+    },
+    {
+      "epoch": 0.11167858358869595,
+      "grad_norm": 2.2858686447143555,
+      "learning_rate": 1e-06,
+      "loss": 0.1037,
+      "step": 656
+    },
+    {
+      "epoch": 0.1118488253319714,
+      "grad_norm": 1.6448941230773926,
+      "learning_rate": 1e-06,
+      "loss": 0.082,
+      "step": 657
+    },
+    {
+      "epoch": 0.11201906707524685,
+      "grad_norm": 1.4810398817062378,
+      "learning_rate": 1e-06,
+      "loss": 0.0806,
+      "step": 658
+    },
+    {
+      "epoch": 0.11218930881852231,
+      "grad_norm": 1.6105705499649048,
+      "learning_rate": 1e-06,
+      "loss": 0.0962,
+      "step": 659
+    },
+    {
+      "epoch": 0.11235955056179775,
+      "grad_norm": 1.5040578842163086,
+      "learning_rate": 1e-06,
+      "loss": 0.0842,
+      "step": 660
+    },
+    {
+      "epoch": 0.1125297923050732,
+      "grad_norm": 1.6204015016555786,
+      "learning_rate": 1e-06,
+      "loss": 0.082,
+      "step": 661
+    },
+    {
+      "epoch": 0.11270003404834865,
+      "grad_norm": 1.3563892841339111,
+      "learning_rate": 1e-06,
+      "loss": 0.061,
+      "step": 662
+    },
+    {
+      "epoch": 0.1128702757916241,
+      "grad_norm": 2.079106330871582,
+      "learning_rate": 1e-06,
+      "loss": 0.0947,
+      "step": 663
+    },
+    {
+      "epoch": 0.11304051753489956,
+      "grad_norm": 1.3969217538833618,
+      "learning_rate": 1e-06,
+      "loss": 0.0723,
+      "step": 664
+    },
+    {
+      "epoch": 0.113210759278175,
+      "grad_norm": 1.3509082794189453,
+      "learning_rate": 1e-06,
+      "loss": 0.0811,
+      "step": 665
+    },
+    {
+      "epoch": 0.11338100102145046,
+      "grad_norm": 1.4950233697891235,
+      "learning_rate": 1e-06,
+      "loss": 0.089,
+      "step": 666
+    },
+    {
+      "epoch": 0.1135512427647259,
+      "grad_norm": 1.6716737747192383,
+      "learning_rate": 1e-06,
+      "loss": 0.0861,
+      "step": 667
+    },
+    {
+      "epoch": 0.11372148450800136,
+      "grad_norm": 1.8707127571105957,
+      "learning_rate": 1e-06,
+      "loss": 0.0924,
+      "step": 668
+    },
+    {
+      "epoch": 0.11389172625127682,
+      "grad_norm": 1.5277262926101685,
+      "learning_rate": 1e-06,
+      "loss": 0.079,
+      "step": 669
+    },
+    {
+      "epoch": 0.11406196799455226,
+      "grad_norm": 1.7219480276107788,
+      "learning_rate": 1e-06,
+      "loss": 0.0928,
+      "step": 670
+    },
+    {
+      "epoch": 0.11423220973782772,
+      "grad_norm": 1.4405816793441772,
+      "learning_rate": 1e-06,
+      "loss": 0.0767,
+      "step": 671
+    },
+    {
+      "epoch": 0.11440245148110317,
+      "grad_norm": 1.7040899991989136,
+      "learning_rate": 1e-06,
+      "loss": 0.0855,
+      "step": 672
+    },
+    {
+      "epoch": 0.11457269322437862,
+      "grad_norm": 1.588266134262085,
+      "learning_rate": 1e-06,
+      "loss": 0.0772,
+      "step": 673
+    },
+    {
+      "epoch": 0.11474293496765407,
+      "grad_norm": 1.8362873792648315,
+      "learning_rate": 1e-06,
+      "loss": 0.0978,
+      "step": 674
+    },
+    {
+      "epoch": 0.11491317671092952,
+      "grad_norm": 1.6758869886398315,
+      "learning_rate": 1e-06,
+      "loss": 0.0928,
+      "step": 675
+    },
+    {
+      "epoch": 0.11508341845420497,
+      "grad_norm": 1.7752368450164795,
+      "learning_rate": 1e-06,
+      "loss": 0.0879,
+      "step": 676
+    },
+    {
+      "epoch": 0.11525366019748043,
+      "grad_norm": 1.6252251863479614,
+      "learning_rate": 1e-06,
+      "loss": 0.096,
+      "step": 677
+    },
+    {
+      "epoch": 0.11542390194075587,
+      "grad_norm": 1.7671598196029663,
+      "learning_rate": 1e-06,
+      "loss": 0.0975,
+      "step": 678
+    },
+    {
+      "epoch": 0.11559414368403133,
+      "grad_norm": 1.3313266038894653,
+      "learning_rate": 1e-06,
+      "loss": 0.0782,
+      "step": 679
+    },
+    {
+      "epoch": 0.11576438542730677,
+      "grad_norm": 1.298072338104248,
+      "learning_rate": 1e-06,
+      "loss": 0.0788,
+      "step": 680
+    },
+    {
+      "epoch": 0.11593462717058223,
+      "grad_norm": 1.4549241065979004,
+      "learning_rate": 1e-06,
+      "loss": 0.0827,
+      "step": 681
+    },
+    {
+      "epoch": 0.11610486891385768,
+      "grad_norm": 1.5572149753570557,
+      "learning_rate": 1e-06,
+      "loss": 0.0866,
+      "step": 682
+    },
+    {
+      "epoch": 0.11627511065713313,
+      "grad_norm": 1.3730863332748413,
+      "learning_rate": 1e-06,
+      "loss": 0.0714,
+      "step": 683
+    },
+    {
+      "epoch": 0.11644535240040858,
+      "grad_norm": 1.7244811058044434,
+      "learning_rate": 1e-06,
+      "loss": 0.0768,
+      "step": 684
+    },
+    {
+      "epoch": 0.11661559414368403,
+      "grad_norm": 1.879199743270874,
+      "learning_rate": 1e-06,
+      "loss": 0.0863,
+      "step": 685
+    },
+    {
+      "epoch": 0.11678583588695948,
+      "grad_norm": 1.3578203916549683,
+      "learning_rate": 1e-06,
+      "loss": 0.0623,
+      "step": 686
+    },
+    {
+      "epoch": 0.11695607763023494,
+      "grad_norm": 1.5125126838684082,
+      "learning_rate": 1e-06,
+      "loss": 0.0872,
+      "step": 687
+    },
+    {
+      "epoch": 0.11712631937351038,
+      "grad_norm": 2.28521728515625,
+      "learning_rate": 1e-06,
+      "loss": 0.1136,
+      "step": 688
+    },
+    {
+      "epoch": 0.11729656111678584,
+      "grad_norm": 1.5524616241455078,
+      "learning_rate": 1e-06,
+      "loss": 0.0815,
+      "step": 689
+    },
+    {
+      "epoch": 0.11746680286006128,
+      "grad_norm": 1.5951274633407593,
+      "learning_rate": 1e-06,
+      "loss": 0.0779,
+      "step": 690
+    },
+    {
+      "epoch": 0.11763704460333674,
+      "grad_norm": 1.1863946914672852,
+      "learning_rate": 1e-06,
+      "loss": 0.0611,
+      "step": 691
+    },
+    {
+      "epoch": 0.1178072863466122,
+      "grad_norm": 1.413318395614624,
+      "learning_rate": 1e-06,
+      "loss": 0.0735,
+      "step": 692
+    },
+    {
+      "epoch": 0.11797752808988764,
+      "grad_norm": 1.5461151599884033,
+      "learning_rate": 1e-06,
+      "loss": 0.0794,
+      "step": 693
+    },
+    {
+      "epoch": 0.1181477698331631,
+      "grad_norm": 1.5801036357879639,
+      "learning_rate": 1e-06,
+      "loss": 0.0805,
+      "step": 694
+    },
+    {
+      "epoch": 0.11831801157643854,
+      "grad_norm": 1.341295838356018,
+      "learning_rate": 1e-06,
+      "loss": 0.07,
+      "step": 695
+    },
+    {
+      "epoch": 0.118488253319714,
+      "grad_norm": 1.561644196510315,
+      "learning_rate": 1e-06,
+      "loss": 0.0756,
+      "step": 696
+    },
+    {
+      "epoch": 0.11865849506298945,
+      "grad_norm": 1.3806345462799072,
+      "learning_rate": 1e-06,
+      "loss": 0.0646,
+      "step": 697
+    },
+    {
+      "epoch": 0.11882873680626489,
+      "grad_norm": 1.5367329120635986,
+      "learning_rate": 1e-06,
+      "loss": 0.0775,
+      "step": 698
+    },
+    {
+      "epoch": 0.11899897854954035,
+      "grad_norm": 1.8795855045318604,
+      "learning_rate": 1e-06,
+      "loss": 0.0936,
+      "step": 699
+    },
+    {
+      "epoch": 0.11916922029281579,
+      "grad_norm": 1.3554261922836304,
+      "learning_rate": 1e-06,
+      "loss": 0.0624,
+      "step": 700
+    },
+    {
+      "epoch": 0.11933946203609125,
+      "grad_norm": 1.4295129776000977,
+      "learning_rate": 1e-06,
+      "loss": 0.0817,
+      "step": 701
+    },
+    {
+      "epoch": 0.1195097037793667,
+      "grad_norm": 2.1465654373168945,
+      "learning_rate": 1e-06,
+      "loss": 0.1037,
+      "step": 702
+    },
+    {
+      "epoch": 0.11967994552264215,
+      "grad_norm": 1.2775219678878784,
+      "learning_rate": 1e-06,
+      "loss": 0.0698,
+      "step": 703
+    },
+    {
+      "epoch": 0.1198501872659176,
+      "grad_norm": 1.6362903118133545,
+      "learning_rate": 1e-06,
+      "loss": 0.0867,
+      "step": 704
+    },
+    {
+      "epoch": 0.12002042900919306,
+      "grad_norm": 1.531864047050476,
+      "learning_rate": 1e-06,
+      "loss": 0.075,
+      "step": 705
+    },
+    {
+      "epoch": 0.1201906707524685,
+      "grad_norm": 1.4992605447769165,
+      "learning_rate": 1e-06,
+      "loss": 0.0909,
+      "step": 706
+    },
+    {
+      "epoch": 0.12036091249574396,
+      "grad_norm": 1.376063346862793,
+      "learning_rate": 1e-06,
+      "loss": 0.0639,
+      "step": 707
+    },
+    {
+      "epoch": 0.1205311542390194,
+      "grad_norm": 1.4128174781799316,
+      "learning_rate": 1e-06,
+      "loss": 0.0725,
+      "step": 708
+    },
+    {
+      "epoch": 0.12070139598229486,
+      "grad_norm": 1.9290549755096436,
+      "learning_rate": 1e-06,
+      "loss": 0.1059,
+      "step": 709
+    },
+    {
+      "epoch": 0.12087163772557032,
+      "grad_norm": 1.6471437215805054,
+      "learning_rate": 1e-06,
+      "loss": 0.0768,
+      "step": 710
+    },
+    {
+      "epoch": 0.12104187946884576,
+      "grad_norm": 1.5400367975234985,
+      "learning_rate": 1e-06,
+      "loss": 0.0828,
+      "step": 711
+    },
+    {
+      "epoch": 0.12121212121212122,
+      "grad_norm": 1.3457374572753906,
+      "learning_rate": 1e-06,
+      "loss": 0.0738,
+      "step": 712
+    },
+    {
+      "epoch": 0.12138236295539666,
+      "grad_norm": 1.8186038732528687,
+      "learning_rate": 1e-06,
+      "loss": 0.0968,
+      "step": 713
+    },
+    {
+      "epoch": 0.12155260469867211,
+      "grad_norm": 1.650877833366394,
+      "learning_rate": 1e-06,
+      "loss": 0.076,
+      "step": 714
+    },
+    {
+      "epoch": 0.12172284644194757,
+      "grad_norm": 1.4798765182495117,
+      "learning_rate": 1e-06,
+      "loss": 0.0787,
+      "step": 715
+    },
+    {
+      "epoch": 0.12189308818522301,
+      "grad_norm": 1.2624869346618652,
+      "learning_rate": 1e-06,
+      "loss": 0.0592,
+      "step": 716
+    },
+    {
+      "epoch": 0.12206332992849847,
+      "grad_norm": 1.596134901046753,
+      "learning_rate": 1e-06,
+      "loss": 0.0761,
+      "step": 717
+    },
+    {
+      "epoch": 0.12223357167177391,
+      "grad_norm": 1.538567304611206,
+      "learning_rate": 1e-06,
+      "loss": 0.0723,
+      "step": 718
+    },
+    {
+      "epoch": 0.12240381341504937,
+      "grad_norm": 5.429737567901611,
+      "learning_rate": 1e-06,
+      "loss": 0.1379,
+      "step": 719
+    },
+    {
+      "epoch": 0.12257405515832483,
+      "grad_norm": 1.6454511880874634,
+      "learning_rate": 1e-06,
+      "loss": 0.0871,
+      "step": 720
+    },
+    {
+      "epoch": 0.12274429690160027,
+      "grad_norm": 1.4540003538131714,
+      "learning_rate": 1e-06,
+      "loss": 0.0788,
+      "step": 721
+    },
+    {
+      "epoch": 0.12291453864487573,
+      "grad_norm": 1.774742603302002,
+      "learning_rate": 1e-06,
+      "loss": 0.0824,
+      "step": 722
+    },
+    {
+      "epoch": 0.12308478038815117,
+      "grad_norm": 1.881834626197815,
+      "learning_rate": 1e-06,
+      "loss": 0.1062,
+      "step": 723
+    },
+    {
+      "epoch": 0.12325502213142663,
+      "grad_norm": 1.543564796447754,
+      "learning_rate": 1e-06,
+      "loss": 0.0949,
+      "step": 724
+    },
+    {
+      "epoch": 0.12342526387470208,
+      "grad_norm": 1.6501184701919556,
+      "learning_rate": 1e-06,
+      "loss": 0.0775,
+      "step": 725
+    },
+    {
+      "epoch": 0.12359550561797752,
+      "grad_norm": 1.3438044786453247,
+      "learning_rate": 1e-06,
+      "loss": 0.0764,
+      "step": 726
+    },
+    {
+      "epoch": 0.12376574736125298,
+      "grad_norm": 1.2549158334732056,
+      "learning_rate": 1e-06,
+      "loss": 0.0575,
+      "step": 727
+    },
+    {
+      "epoch": 0.12393598910452842,
+      "grad_norm": 1.8060505390167236,
+      "learning_rate": 1e-06,
+      "loss": 0.0829,
+      "step": 728
+    },
+    {
+      "epoch": 0.12410623084780388,
+      "grad_norm": 1.6881535053253174,
+      "learning_rate": 1e-06,
+      "loss": 0.0849,
+      "step": 729
+    },
+    {
+      "epoch": 0.12427647259107934,
+      "grad_norm": 1.4681129455566406,
+      "learning_rate": 1e-06,
+      "loss": 0.054,
+      "step": 730
+    },
+    {
+      "epoch": 0.12444671433435478,
+      "grad_norm": 1.806591510772705,
+      "learning_rate": 1e-06,
+      "loss": 0.081,
+      "step": 731
+    },
+    {
+      "epoch": 0.12461695607763024,
+      "grad_norm": 1.6089766025543213,
+      "learning_rate": 1e-06,
+      "loss": 0.0773,
+      "step": 732
+    },
+    {
+      "epoch": 0.12478719782090568,
+      "grad_norm": 1.6091641187667847,
+      "learning_rate": 1e-06,
+      "loss": 0.0804,
+      "step": 733
+    },
+    {
+      "epoch": 0.12495743956418114,
+      "grad_norm": 1.669281005859375,
+      "learning_rate": 1e-06,
+      "loss": 0.0771,
+      "step": 734
+    },
+    {
+      "epoch": 0.12512768130745658,
+      "grad_norm": 1.7492468357086182,
+      "learning_rate": 1e-06,
+      "loss": 0.0777,
+      "step": 735
+    },
+    {
+      "epoch": 0.12529792305073204,
+      "grad_norm": 1.7161555290222168,
+      "learning_rate": 1e-06,
+      "loss": 0.0898,
+      "step": 736
+    },
+    {
+      "epoch": 0.1254681647940075,
+      "grad_norm": 1.7574137449264526,
+      "learning_rate": 1e-06,
+      "loss": 0.107,
+      "step": 737
+    },
+    {
+      "epoch": 0.12563840653728295,
+      "grad_norm": 1.3838268518447876,
+      "learning_rate": 1e-06,
+      "loss": 0.0699,
+      "step": 738
+    },
+    {
+      "epoch": 0.1258086482805584,
+      "grad_norm": 1.6737573146820068,
+      "learning_rate": 1e-06,
+      "loss": 0.0846,
+      "step": 739
+    },
+    {
+      "epoch": 0.12597889002383383,
+      "grad_norm": 2.306833028793335,
+      "learning_rate": 1e-06,
+      "loss": 0.1018,
+      "step": 740
+    },
+    {
+      "epoch": 0.1261491317671093,
+      "grad_norm": 2.311218738555908,
+      "learning_rate": 1e-06,
+      "loss": 0.1155,
+      "step": 741
+    },
+    {
+      "epoch": 0.12631937351038475,
+      "grad_norm": 1.696584701538086,
+      "learning_rate": 1e-06,
+      "loss": 0.069,
+      "step": 742
+    },
+    {
+      "epoch": 0.1264896152536602,
+      "grad_norm": 1.3152414560317993,
+      "learning_rate": 1e-06,
+      "loss": 0.0606,
+      "step": 743
+    },
+    {
+      "epoch": 0.12665985699693566,
+      "grad_norm": 1.6591945886611938,
+      "learning_rate": 1e-06,
+      "loss": 0.0859,
+      "step": 744
+    },
+    {
+      "epoch": 0.1268300987402111,
+      "grad_norm": 1.4438961744308472,
+      "learning_rate": 1e-06,
+      "loss": 0.0702,
+      "step": 745
+    },
+    {
+      "epoch": 0.12700034048348655,
+      "grad_norm": 1.289958119392395,
+      "learning_rate": 1e-06,
+      "loss": 0.067,
+      "step": 746
+    },
+    {
+      "epoch": 0.127170582226762,
+      "grad_norm": 2.529083251953125,
+      "learning_rate": 1e-06,
+      "loss": 0.1006,
+      "step": 747
+    },
+    {
+      "epoch": 0.12734082397003746,
+      "grad_norm": 1.5549131631851196,
+      "learning_rate": 1e-06,
+      "loss": 0.0815,
+      "step": 748
+    },
+    {
+      "epoch": 0.12751106571331292,
+      "grad_norm": 1.5461150407791138,
+      "learning_rate": 1e-06,
+      "loss": 0.0835,
+      "step": 749
+    },
+    {
+      "epoch": 0.12768130745658834,
+      "grad_norm": 1.4177753925323486,
+      "learning_rate": 1e-06,
+      "loss": 0.072,
+      "step": 750
+    },
+    {
+      "epoch": 0.1278515491998638,
+      "grad_norm": 1.6254005432128906,
+      "learning_rate": 1e-06,
+      "loss": 0.0788,
+      "step": 751
+    },
+    {
+      "epoch": 0.12802179094313926,
+      "grad_norm": 1.7862988710403442,
+      "learning_rate": 1e-06,
+      "loss": 0.0813,
+      "step": 752
+    },
+    {
+      "epoch": 0.12819203268641471,
+      "grad_norm": 1.7907599210739136,
+      "learning_rate": 1e-06,
+      "loss": 0.0889,
+      "step": 753
+    },
+    {
+      "epoch": 0.12836227442969017,
+      "grad_norm": 1.9308052062988281,
+      "learning_rate": 1e-06,
+      "loss": 0.0851,
+      "step": 754
+    },
+    {
+      "epoch": 0.1285325161729656,
+      "grad_norm": 1.512603998184204,
+      "learning_rate": 1e-06,
+      "loss": 0.0737,
+      "step": 755
+    },
+    {
+      "epoch": 0.12870275791624106,
+      "grad_norm": 1.6706660985946655,
+      "learning_rate": 1e-06,
+      "loss": 0.0857,
+      "step": 756
+    },
+    {
+      "epoch": 0.1288729996595165,
+      "grad_norm": 1.6297451257705688,
+      "learning_rate": 1e-06,
+      "loss": 0.0796,
+      "step": 757
+    },
+    {
+      "epoch": 0.12904324140279197,
+      "grad_norm": 1.4156213998794556,
+      "learning_rate": 1e-06,
+      "loss": 0.0762,
+      "step": 758
+    },
+    {
+      "epoch": 0.12921348314606743,
+      "grad_norm": 1.7496767044067383,
+      "learning_rate": 1e-06,
+      "loss": 0.0802,
+      "step": 759
+    },
+    {
+      "epoch": 0.12938372488934285,
+      "grad_norm": 1.380969524383545,
+      "learning_rate": 1e-06,
+      "loss": 0.0757,
+      "step": 760
+    },
+    {
+      "epoch": 0.1295539666326183,
+      "grad_norm": 1.8817622661590576,
+      "learning_rate": 1e-06,
+      "loss": 0.0812,
+      "step": 761
+    },
+    {
+      "epoch": 0.12972420837589377,
+      "grad_norm": 1.3597335815429688,
+      "learning_rate": 1e-06,
+      "loss": 0.063,
+      "step": 762
+    },
+    {
+      "epoch": 0.12989445011916922,
+      "grad_norm": 1.5356440544128418,
+      "learning_rate": 1e-06,
+      "loss": 0.077,
+      "step": 763
+    },
+    {
+      "epoch": 0.13006469186244468,
+      "grad_norm": 1.3353973627090454,
+      "learning_rate": 1e-06,
+      "loss": 0.0552,
+      "step": 764
+    },
+    {
+      "epoch": 0.1302349336057201,
+      "grad_norm": 1.738538384437561,
+      "learning_rate": 1e-06,
+      "loss": 0.066,
+      "step": 765
+    },
+    {
+      "epoch": 0.13040517534899557,
+      "grad_norm": 1.3719581365585327,
+      "learning_rate": 1e-06,
+      "loss": 0.0606,
+      "step": 766
+    },
+    {
+      "epoch": 0.13057541709227102,
+      "grad_norm": 2.158234119415283,
+      "learning_rate": 1e-06,
+      "loss": 0.0843,
+      "step": 767
+    },
+    {
+      "epoch": 0.13074565883554648,
+      "grad_norm": 1.713789701461792,
+      "learning_rate": 1e-06,
+      "loss": 0.0808,
+      "step": 768
+    },
+    {
+      "epoch": 0.13091590057882194,
+      "grad_norm": 1.5063124895095825,
+      "learning_rate": 1e-06,
+      "loss": 0.0769,
+      "step": 769
+    },
+    {
+      "epoch": 0.13108614232209737,
+      "grad_norm": 1.3353931903839111,
+      "learning_rate": 1e-06,
+      "loss": 0.0698,
+      "step": 770
+    },
+    {
+      "epoch": 0.13125638406537282,
+      "grad_norm": 1.8870093822479248,
+      "learning_rate": 1e-06,
+      "loss": 0.091,
+      "step": 771
+    },
+    {
+      "epoch": 0.13142662580864828,
+      "grad_norm": 1.8463548421859741,
+      "learning_rate": 1e-06,
+      "loss": 0.0668,
+      "step": 772
+    },
+    {
+      "epoch": 0.13159686755192374,
+      "grad_norm": 1.3892488479614258,
+      "learning_rate": 1e-06,
+      "loss": 0.0718,
+      "step": 773
+    },
+    {
+      "epoch": 0.1317671092951992,
+      "grad_norm": 1.5559097528457642,
+      "learning_rate": 1e-06,
+      "loss": 0.0706,
+      "step": 774
+    },
+    {
+      "epoch": 0.13193735103847462,
+      "grad_norm": 2.101165533065796,
+      "learning_rate": 1e-06,
+      "loss": 0.1133,
+      "step": 775
+    },
+    {
+      "epoch": 0.13210759278175008,
+      "grad_norm": 1.7307274341583252,
+      "learning_rate": 1e-06,
+      "loss": 0.0663,
+      "step": 776
+    },
+    {
+      "epoch": 0.13227783452502553,
+      "grad_norm": 1.909165382385254,
+      "learning_rate": 1e-06,
+      "loss": 0.0817,
+      "step": 777
+    },
+    {
+      "epoch": 0.132448076268301,
+      "grad_norm": 1.6024906635284424,
+      "learning_rate": 1e-06,
+      "loss": 0.0757,
+      "step": 778
+    },
+    {
+      "epoch": 0.13261831801157645,
+      "grad_norm": 1.6666183471679688,
+      "learning_rate": 1e-06,
+      "loss": 0.0841,
+      "step": 779
+    },
+    {
+      "epoch": 0.13278855975485188,
+      "grad_norm": 2.6020991802215576,
+      "learning_rate": 1e-06,
+      "loss": 0.1077,
+      "step": 780
+    },
+    {
+      "epoch": 0.13295880149812733,
+      "grad_norm": 1.5541698932647705,
+      "learning_rate": 1e-06,
+      "loss": 0.0802,
+      "step": 781
+    },
+    {
+      "epoch": 0.1331290432414028,
+      "grad_norm": 1.538192868232727,
+      "learning_rate": 1e-06,
+      "loss": 0.0685,
+      "step": 782
+    },
+    {
+      "epoch": 0.13329928498467825,
+      "grad_norm": 1.704047679901123,
+      "learning_rate": 1e-06,
+      "loss": 0.0652,
+      "step": 783
+    },
+    {
+      "epoch": 0.1334695267279537,
+      "grad_norm": 1.5509108304977417,
+      "learning_rate": 1e-06,
+      "loss": 0.0736,
+      "step": 784
+    },
+    {
+      "epoch": 0.13363976847122916,
+      "grad_norm": 1.4850906133651733,
+      "learning_rate": 1e-06,
+      "loss": 0.0599,
+      "step": 785
+    },
+    {
+      "epoch": 0.1338100102145046,
+      "grad_norm": 1.46503484249115,
+      "learning_rate": 1e-06,
+      "loss": 0.0594,
+      "step": 786
+    },
+    {
+      "epoch": 0.13398025195778004,
+      "grad_norm": 1.633912205696106,
+      "learning_rate": 1e-06,
+      "loss": 0.0699,
+      "step": 787
+    },
+    {
+      "epoch": 0.1341504937010555,
+      "grad_norm": 1.7929328680038452,
+      "learning_rate": 1e-06,
+      "loss": 0.0616,
+      "step": 788
+    },
+    {
+      "epoch": 0.13432073544433096,
+      "grad_norm": 1.710610270500183,
+      "learning_rate": 1e-06,
+      "loss": 0.0724,
+      "step": 789
+    },
+    {
+      "epoch": 0.13449097718760641,
+      "grad_norm": 1.7763170003890991,
+      "learning_rate": 1e-06,
+      "loss": 0.0858,
+      "step": 790
+    },
+    {
+      "epoch": 0.13466121893088184,
+      "grad_norm": 1.5168254375457764,
+      "learning_rate": 1e-06,
+      "loss": 0.0769,
+      "step": 791
+    },
+    {
+      "epoch": 0.1348314606741573,
+      "grad_norm": 1.3891587257385254,
+      "learning_rate": 1e-06,
+      "loss": 0.0615,
+      "step": 792
+    },
+    {
+      "epoch": 0.13500170241743276,
+      "grad_norm": 1.804847002029419,
+      "learning_rate": 1e-06,
+      "loss": 0.0925,
+      "step": 793
+    },
+    {
+      "epoch": 0.1351719441607082,
+      "grad_norm": 1.5551631450653076,
+      "learning_rate": 1e-06,
+      "loss": 0.0702,
+      "step": 794
+    },
+    {
+      "epoch": 0.13534218590398367,
+      "grad_norm": 1.4586671590805054,
+      "learning_rate": 1e-06,
+      "loss": 0.0745,
+      "step": 795
+    },
+    {
+      "epoch": 0.1355124276472591,
+      "grad_norm": 1.7194534540176392,
+      "learning_rate": 1e-06,
+      "loss": 0.0944,
+      "step": 796
+    },
+    {
+      "epoch": 0.13568266939053455,
+      "grad_norm": 1.5851624011993408,
+      "learning_rate": 1e-06,
+      "loss": 0.0713,
+      "step": 797
+    },
+    {
+      "epoch": 0.13585291113381,
+      "grad_norm": 2.052922010421753,
+      "learning_rate": 1e-06,
+      "loss": 0.0922,
+      "step": 798
+    },
+    {
+      "epoch": 0.13602315287708547,
+      "grad_norm": 1.4071455001831055,
+      "learning_rate": 1e-06,
+      "loss": 0.0655,
+      "step": 799
+    },
+    {
+      "epoch": 0.13619339462036092,
+      "grad_norm": 1.8842096328735352,
+      "learning_rate": 1e-06,
+      "loss": 0.0725,
+      "step": 800
+    },
+    {
+      "epoch": 0.13636363636363635,
+      "grad_norm": 1.692454218864441,
+      "learning_rate": 1e-06,
+      "loss": 0.0718,
+      "step": 801
+    },
+    {
+      "epoch": 0.1365338781069118,
+      "grad_norm": 1.4987190961837769,
+      "learning_rate": 1e-06,
+      "loss": 0.0563,
+      "step": 802
+    },
+    {
+      "epoch": 0.13670411985018727,
+      "grad_norm": 1.3979241847991943,
+      "learning_rate": 1e-06,
+      "loss": 0.0658,
+      "step": 803
+    },
+    {
+      "epoch": 0.13687436159346272,
+      "grad_norm": 1.5798490047454834,
+      "learning_rate": 1e-06,
+      "loss": 0.0739,
+      "step": 804
+    },
+    {
+      "epoch": 0.13704460333673818,
+      "grad_norm": 1.990831732749939,
+      "learning_rate": 1e-06,
+      "loss": 0.0759,
+      "step": 805
+    },
+    {
+      "epoch": 0.1372148450800136,
+      "grad_norm": 1.9039353132247925,
+      "learning_rate": 1e-06,
+      "loss": 0.0583,
+      "step": 806
+    },
+    {
+      "epoch": 0.13738508682328907,
+      "grad_norm": 1.76191246509552,
+      "learning_rate": 1e-06,
+      "loss": 0.0836,
+      "step": 807
+    },
+    {
+      "epoch": 0.13755532856656452,
+      "grad_norm": 1.5676465034484863,
+      "learning_rate": 1e-06,
+      "loss": 0.062,
+      "step": 808
+    },
+    {
+      "epoch": 0.13772557030983998,
+      "grad_norm": 1.9201511144638062,
+      "learning_rate": 1e-06,
+      "loss": 0.0705,
+      "step": 809
+    },
+    {
+      "epoch": 0.13789581205311544,
+      "grad_norm": 1.5441217422485352,
+      "learning_rate": 1e-06,
+      "loss": 0.0594,
+      "step": 810
+    },
+    {
+      "epoch": 0.13806605379639086,
+      "grad_norm": 1.9137436151504517,
+      "learning_rate": 1e-06,
+      "loss": 0.0745,
+      "step": 811
+    },
+    {
+      "epoch": 0.13823629553966632,
+      "grad_norm": 1.3507276773452759,
+      "learning_rate": 1e-06,
+      "loss": 0.0656,
+      "step": 812
+    },
+    {
+      "epoch": 0.13840653728294178,
+      "grad_norm": 1.6629807949066162,
+      "learning_rate": 1e-06,
+      "loss": 0.0721,
+      "step": 813
+    },
+    {
+      "epoch": 0.13857677902621723,
+      "grad_norm": 1.6247719526290894,
+      "learning_rate": 1e-06,
+      "loss": 0.0717,
+      "step": 814
+    },
+    {
+      "epoch": 0.1387470207694927,
+      "grad_norm": 1.6012877225875854,
+      "learning_rate": 1e-06,
+      "loss": 0.072,
+      "step": 815
+    },
+    {
+      "epoch": 0.13891726251276812,
+      "grad_norm": 1.6676996946334839,
+      "learning_rate": 1e-06,
+      "loss": 0.0812,
+      "step": 816
+    },
+    {
+      "epoch": 0.13908750425604358,
+      "grad_norm": 1.6991463899612427,
+      "learning_rate": 1e-06,
+      "loss": 0.0729,
+      "step": 817
+    },
+    {
+      "epoch": 0.13925774599931903,
+      "grad_norm": 1.4762986898422241,
+      "learning_rate": 1e-06,
+      "loss": 0.0686,
+      "step": 818
+    },
+    {
+      "epoch": 0.1394279877425945,
+      "grad_norm": 1.65053391456604,
+      "learning_rate": 1e-06,
+      "loss": 0.0795,
+      "step": 819
+    },
+    {
+      "epoch": 0.13959822948586995,
+      "grad_norm": 1.437187910079956,
+      "learning_rate": 1e-06,
+      "loss": 0.0638,
+      "step": 820
+    },
+    {
+      "epoch": 0.13976847122914537,
+      "grad_norm": 1.4831326007843018,
+      "learning_rate": 1e-06,
+      "loss": 0.075,
+      "step": 821
+    },
+    {
+      "epoch": 0.13993871297242083,
+      "grad_norm": 1.3216532468795776,
+      "learning_rate": 1e-06,
+      "loss": 0.0532,
+      "step": 822
+    },
+    {
+      "epoch": 0.1401089547156963,
+      "grad_norm": 1.9568979740142822,
+      "learning_rate": 1e-06,
+      "loss": 0.0817,
+      "step": 823
+    },
+    {
+      "epoch": 0.14027919645897174,
+      "grad_norm": 1.7128760814666748,
+      "learning_rate": 1e-06,
+      "loss": 0.0717,
+      "step": 824
+    },
+    {
+      "epoch": 0.1404494382022472,
+      "grad_norm": 1.5816516876220703,
+      "learning_rate": 1e-06,
+      "loss": 0.0662,
+      "step": 825
+    },
+    {
+      "epoch": 0.14061967994552263,
+      "grad_norm": 1.5961077213287354,
+      "learning_rate": 1e-06,
+      "loss": 0.0748,
+      "step": 826
+    },
+    {
+      "epoch": 0.1407899216887981,
+      "grad_norm": 1.8750134706497192,
+      "learning_rate": 1e-06,
+      "loss": 0.0588,
+      "step": 827
+    },
+    {
+      "epoch": 0.14096016343207354,
+      "grad_norm": 1.8028770685195923,
+      "learning_rate": 1e-06,
+      "loss": 0.0777,
+      "step": 828
+    },
+    {
+      "epoch": 0.141130405175349,
+      "grad_norm": 1.7617617845535278,
+      "learning_rate": 1e-06,
+      "loss": 0.0753,
+      "step": 829
+    },
+    {
+      "epoch": 0.14130064691862446,
+      "grad_norm": 1.6493316888809204,
+      "learning_rate": 1e-06,
+      "loss": 0.0638,
+      "step": 830
+    },
+    {
+      "epoch": 0.14147088866189989,
+      "grad_norm": 4.191693305969238,
+      "learning_rate": 1e-06,
+      "loss": 0.1091,
+      "step": 831
+    },
+    {
+      "epoch": 0.14164113040517534,
+      "grad_norm": 1.787387728691101,
+      "learning_rate": 1e-06,
+      "loss": 0.0766,
+      "step": 832
+    },
+    {
+      "epoch": 0.1418113721484508,
+      "grad_norm": 1.9080184698104858,
+      "learning_rate": 1e-06,
+      "loss": 0.0744,
+      "step": 833
+    },
+    {
+      "epoch": 0.14198161389172625,
+      "grad_norm": 2.04909086227417,
+      "learning_rate": 1e-06,
+      "loss": 0.0637,
+      "step": 834
+    },
+    {
+      "epoch": 0.1421518556350017,
+      "grad_norm": 1.588629126548767,
+      "learning_rate": 1e-06,
+      "loss": 0.0513,
+      "step": 835
+    },
+    {
+      "epoch": 0.14232209737827714,
+      "grad_norm": 1.3033459186553955,
+      "learning_rate": 1e-06,
+      "loss": 0.0605,
+      "step": 836
+    },
+    {
+      "epoch": 0.1424923391215526,
+      "grad_norm": 1.4227380752563477,
+      "learning_rate": 1e-06,
+      "loss": 0.0583,
+      "step": 837
+    },
+    {
+      "epoch": 0.14266258086482805,
+      "grad_norm": 2.39058780670166,
+      "learning_rate": 1e-06,
+      "loss": 0.0931,
+      "step": 838
+    },
+    {
+      "epoch": 0.1428328226081035,
+      "grad_norm": 1.6543277502059937,
+      "learning_rate": 1e-06,
+      "loss": 0.0752,
+      "step": 839
+    },
+    {
+      "epoch": 0.14300306435137897,
+      "grad_norm": 1.3659149408340454,
+      "learning_rate": 1e-06,
+      "loss": 0.0439,
+      "step": 840
+    },
+    {
+      "epoch": 0.1431733060946544,
+      "grad_norm": 1.9133633375167847,
+      "learning_rate": 1e-06,
+      "loss": 0.0874,
+      "step": 841
+    },
+    {
+      "epoch": 0.14334354783792985,
+      "grad_norm": 2.2252583503723145,
+      "learning_rate": 1e-06,
+      "loss": 0.0994,
+      "step": 842
+    },
+    {
+      "epoch": 0.1435137895812053,
+      "grad_norm": 1.5215318202972412,
+      "learning_rate": 1e-06,
+      "loss": 0.065,
+      "step": 843
+    },
+    {
+      "epoch": 0.14368403132448077,
+      "grad_norm": 1.6508617401123047,
+      "learning_rate": 1e-06,
+      "loss": 0.0641,
+      "step": 844
+    },
+    {
+      "epoch": 0.14385427306775622,
+      "grad_norm": 1.4455782175064087,
+      "learning_rate": 1e-06,
+      "loss": 0.0459,
+      "step": 845
+    },
+    {
+      "epoch": 0.14402451481103168,
+      "grad_norm": 1.5651944875717163,
+      "learning_rate": 1e-06,
+      "loss": 0.0639,
+      "step": 846
+    },
+    {
+      "epoch": 0.1441947565543071,
+      "grad_norm": 1.6415917873382568,
+      "learning_rate": 1e-06,
+      "loss": 0.057,
+      "step": 847
+    },
+    {
+      "epoch": 0.14436499829758256,
+      "grad_norm": 1.7259793281555176,
+      "learning_rate": 1e-06,
+      "loss": 0.0627,
+      "step": 848
+    },
+    {
+      "epoch": 0.14453524004085802,
+      "grad_norm": 1.8741511106491089,
+      "learning_rate": 1e-06,
+      "loss": 0.0695,
+      "step": 849
+    },
+    {
+      "epoch": 0.14470548178413348,
+      "grad_norm": 1.6153074502944946,
+      "learning_rate": 1e-06,
+      "loss": 0.0753,
+      "step": 850
+    },
+    {
+      "epoch": 0.14487572352740893,
+      "grad_norm": 1.6454153060913086,
+      "learning_rate": 1e-06,
+      "loss": 0.0708,
+      "step": 851
+    },
+    {
+      "epoch": 0.14504596527068436,
+      "grad_norm": 2.058832883834839,
+      "learning_rate": 1e-06,
+      "loss": 0.0723,
+      "step": 852
+    },
+    {
+      "epoch": 0.14521620701395982,
+      "grad_norm": 1.3613487482070923,
+      "learning_rate": 1e-06,
+      "loss": 0.0551,
+      "step": 853
+    },
+    {
+      "epoch": 0.14538644875723528,
+      "grad_norm": 1.6068713665008545,
+      "learning_rate": 1e-06,
+      "loss": 0.0554,
+      "step": 854
+    },
+    {
+      "epoch": 0.14555669050051073,
+      "grad_norm": 1.7198082208633423,
+      "learning_rate": 1e-06,
+      "loss": 0.071,
+      "step": 855
+    },
+    {
+      "epoch": 0.1457269322437862,
+      "grad_norm": 1.3624064922332764,
+      "learning_rate": 1e-06,
+      "loss": 0.0504,
+      "step": 856
+    },
+    {
+      "epoch": 0.14589717398706162,
+      "grad_norm": 1.5122796297073364,
+      "learning_rate": 1e-06,
+      "loss": 0.0606,
+      "step": 857
+    },
+    {
+      "epoch": 0.14606741573033707,
+      "grad_norm": 1.4191750288009644,
+      "learning_rate": 1e-06,
+      "loss": 0.0609,
+      "step": 858
+    },
+    {
+      "epoch": 0.14623765747361253,
+      "grad_norm": 1.983462929725647,
+      "learning_rate": 1e-06,
+      "loss": 0.0647,
+      "step": 859
+    },
+    {
+      "epoch": 0.146407899216888,
+      "grad_norm": 1.5732680559158325,
+      "learning_rate": 1e-06,
+      "loss": 0.0594,
+      "step": 860
+    },
+    {
+      "epoch": 0.14657814096016344,
+      "grad_norm": 1.5888450145721436,
+      "learning_rate": 1e-06,
+      "loss": 0.0609,
+      "step": 861
+    },
+    {
+      "epoch": 0.14674838270343887,
+      "grad_norm": 1.727030873298645,
+      "learning_rate": 1e-06,
+      "loss": 0.0762,
+      "step": 862
+    },
+    {
+      "epoch": 0.14691862444671433,
+      "grad_norm": 1.599660038948059,
+      "learning_rate": 1e-06,
+      "loss": 0.0706,
+      "step": 863
+    },
+    {
+      "epoch": 0.1470888661899898,
+      "grad_norm": 1.5314996242523193,
+      "learning_rate": 1e-06,
+      "loss": 0.0706,
+      "step": 864
+    },
+    {
+      "epoch": 0.14725910793326524,
+      "grad_norm": 1.532114863395691,
+      "learning_rate": 1e-06,
+      "loss": 0.0531,
+      "step": 865
+    },
+    {
+      "epoch": 0.1474293496765407,
+      "grad_norm": 1.5699235200881958,
+      "learning_rate": 1e-06,
+      "loss": 0.0575,
+      "step": 866
+    },
+    {
+      "epoch": 0.14759959141981613,
+      "grad_norm": 1.4962087869644165,
+      "learning_rate": 1e-06,
+      "loss": 0.048,
+      "step": 867
+    },
+    {
+      "epoch": 0.14776983316309159,
+      "grad_norm": 1.6223524808883667,
+      "learning_rate": 1e-06,
+      "loss": 0.0557,
+      "step": 868
+    },
+    {
+      "epoch": 0.14794007490636704,
+      "grad_norm": 1.379461407661438,
+      "learning_rate": 1e-06,
+      "loss": 0.0488,
+      "step": 869
+    },
+    {
+      "epoch": 0.1481103166496425,
+      "grad_norm": 1.6299303770065308,
+      "learning_rate": 1e-06,
+      "loss": 0.0712,
+      "step": 870
+    },
+    {
+      "epoch": 0.14828055839291795,
+      "grad_norm": 1.5951275825500488,
+      "learning_rate": 1e-06,
+      "loss": 0.0633,
+      "step": 871
+    },
+    {
+      "epoch": 0.14845080013619338,
+      "grad_norm": 1.6737635135650635,
+      "learning_rate": 1e-06,
+      "loss": 0.0576,
+      "step": 872
+    },
+    {
+      "epoch": 0.14862104187946884,
+      "grad_norm": 2.7894175052642822,
+      "learning_rate": 1e-06,
+      "loss": 0.0916,
+      "step": 873
+    },
+    {
+      "epoch": 0.1487912836227443,
+      "grad_norm": 1.7583144903182983,
+      "learning_rate": 1e-06,
+      "loss": 0.0556,
+      "step": 874
+    },
+    {
+      "epoch": 0.14896152536601975,
+      "grad_norm": 1.6645101308822632,
+      "learning_rate": 1e-06,
+      "loss": 0.0673,
+      "step": 875
+    },
+    {
+      "epoch": 0.1491317671092952,
+      "grad_norm": 1.6808711290359497,
+      "learning_rate": 1e-06,
+      "loss": 0.0541,
+      "step": 876
+    },
+    {
+      "epoch": 0.14930200885257064,
+      "grad_norm": 1.5654128789901733,
+      "learning_rate": 1e-06,
+      "loss": 0.0636,
+      "step": 877
+    },
+    {
+      "epoch": 0.1494722505958461,
+      "grad_norm": 1.5253329277038574,
+      "learning_rate": 1e-06,
+      "loss": 0.0561,
+      "step": 878
+    },
+    {
+      "epoch": 0.14964249233912155,
+      "grad_norm": 1.4865022897720337,
+      "learning_rate": 1e-06,
+      "loss": 0.0564,
+      "step": 879
+    },
+    {
+      "epoch": 0.149812734082397,
+      "grad_norm": 1.4917443990707397,
+      "learning_rate": 1e-06,
+      "loss": 0.0575,
+      "step": 880
+    },
+    {
+      "epoch": 0.14998297582567247,
+      "grad_norm": 1.7758963108062744,
+      "learning_rate": 1e-06,
+      "loss": 0.0681,
+      "step": 881
+    },
+    {
+      "epoch": 0.1501532175689479,
+      "grad_norm": 1.369234323501587,
+      "learning_rate": 1e-06,
+      "loss": 0.0596,
+      "step": 882
+    },
+    {
+      "epoch": 0.15032345931222335,
+      "grad_norm": 2.3506531715393066,
+      "learning_rate": 1e-06,
+      "loss": 0.083,
+      "step": 883
+    },
+    {
+      "epoch": 0.1504937010554988,
+      "grad_norm": 1.8270442485809326,
+      "learning_rate": 1e-06,
+      "loss": 0.0722,
+      "step": 884
+    },
+    {
+      "epoch": 0.15066394279877426,
+      "grad_norm": 1.7770593166351318,
+      "learning_rate": 1e-06,
+      "loss": 0.0672,
+      "step": 885
+    },
+    {
+      "epoch": 0.15083418454204972,
+      "grad_norm": 1.7954052686691284,
+      "learning_rate": 1e-06,
+      "loss": 0.068,
+      "step": 886
+    },
+    {
+      "epoch": 0.15100442628532515,
+      "grad_norm": 1.5579285621643066,
+      "learning_rate": 1e-06,
+      "loss": 0.0665,
+      "step": 887
+    },
+    {
+      "epoch": 0.1511746680286006,
+      "grad_norm": 1.6267443895339966,
+      "learning_rate": 1e-06,
+      "loss": 0.0616,
+      "step": 888
+    },
+    {
+      "epoch": 0.15134490977187606,
+      "grad_norm": 1.9561134576797485,
+      "learning_rate": 1e-06,
+      "loss": 0.0672,
+      "step": 889
+    },
+    {
+      "epoch": 0.15151515151515152,
+      "grad_norm": 1.592724084854126,
+      "learning_rate": 1e-06,
+      "loss": 0.0641,
+      "step": 890
+    },
+    {
+      "epoch": 0.15168539325842698,
+      "grad_norm": 1.448255181312561,
+      "learning_rate": 1e-06,
+      "loss": 0.0499,
+      "step": 891
+    },
+    {
+      "epoch": 0.1518556350017024,
+      "grad_norm": 1.7166950702667236,
+      "learning_rate": 1e-06,
+      "loss": 0.0724,
+      "step": 892
+    },
+    {
+      "epoch": 0.15202587674497786,
+      "grad_norm": 2.7402710914611816,
+      "learning_rate": 1e-06,
+      "loss": 0.0981,
+      "step": 893
+    },
+    {
+      "epoch": 0.15219611848825332,
+      "grad_norm": 1.9034618139266968,
+      "learning_rate": 1e-06,
+      "loss": 0.0655,
+      "step": 894
+    },
+    {
+      "epoch": 0.15236636023152877,
+      "grad_norm": 1.6078972816467285,
+      "learning_rate": 1e-06,
+      "loss": 0.0545,
+      "step": 895
+    },
+    {
+      "epoch": 0.15253660197480423,
+      "grad_norm": 3.4988811016082764,
+      "learning_rate": 1e-06,
+      "loss": 0.0725,
+      "step": 896
+    },
+    {
+      "epoch": 0.15270684371807966,
+      "grad_norm": 1.6912193298339844,
+      "learning_rate": 1e-06,
+      "loss": 0.0617,
+      "step": 897
+    },
+    {
+      "epoch": 0.15287708546135512,
+      "grad_norm": 1.5782009363174438,
+      "learning_rate": 1e-06,
+      "loss": 0.0454,
+      "step": 898
+    },
+    {
+      "epoch": 0.15304732720463057,
+      "grad_norm": 2.1376571655273438,
+      "learning_rate": 1e-06,
+      "loss": 0.1007,
+      "step": 899
+    },
+    {
+      "epoch": 0.15321756894790603,
+      "grad_norm": 1.8803651332855225,
+      "learning_rate": 1e-06,
+      "loss": 0.0655,
+      "step": 900
+    },
+    {
+      "epoch": 0.1533878106911815,
+      "grad_norm": 1.6837856769561768,
+      "learning_rate": 1e-06,
+      "loss": 0.07,
+      "step": 901
+    },
+    {
+      "epoch": 0.15355805243445692,
+      "grad_norm": 1.5627678632736206,
+      "learning_rate": 1e-06,
+      "loss": 0.0583,
+      "step": 902
+    },
+    {
+      "epoch": 0.15372829417773237,
+      "grad_norm": 1.413833498954773,
+      "learning_rate": 1e-06,
+      "loss": 0.05,
+      "step": 903
+    },
+    {
+      "epoch": 0.15389853592100783,
+      "grad_norm": 2.4595329761505127,
+      "learning_rate": 1e-06,
+      "loss": 0.1166,
+      "step": 904
+    },
+    {
+      "epoch": 0.15406877766428329,
+      "grad_norm": 4.622979640960693,
+      "learning_rate": 1e-06,
+      "loss": 0.102,
+      "step": 905
+    },
+    {
+      "epoch": 0.15423901940755874,
+      "grad_norm": 1.9842865467071533,
+      "learning_rate": 1e-06,
+      "loss": 0.0704,
+      "step": 906
+    },
+    {
+      "epoch": 0.1544092611508342,
+      "grad_norm": 1.6348425149917603,
+      "learning_rate": 1e-06,
+      "loss": 0.0519,
+      "step": 907
+    },
+    {
+      "epoch": 0.15457950289410963,
+      "grad_norm": 1.920792579650879,
+      "learning_rate": 1e-06,
+      "loss": 0.0644,
+      "step": 908
+    },
+    {
+      "epoch": 0.15474974463738508,
+      "grad_norm": 2.1553070545196533,
+      "learning_rate": 1e-06,
+      "loss": 0.0698,
+      "step": 909
+    },
+    {
+      "epoch": 0.15491998638066054,
+      "grad_norm": 1.422676920890808,
+      "learning_rate": 1e-06,
+      "loss": 0.0542,
+      "step": 910
+    },
+    {
+      "epoch": 0.155090228123936,
+      "grad_norm": 1.8286123275756836,
+      "learning_rate": 1e-06,
+      "loss": 0.0608,
+      "step": 911
+    },
+    {
+      "epoch": 0.15526046986721145,
+      "grad_norm": 1.6634122133255005,
+      "learning_rate": 1e-06,
+      "loss": 0.0581,
+      "step": 912
+    },
+    {
+      "epoch": 0.15543071161048688,
+      "grad_norm": 1.5610778331756592,
+      "learning_rate": 1e-06,
+      "loss": 0.0667,
+      "step": 913
+    },
+    {
+      "epoch": 0.15560095335376234,
+      "grad_norm": 1.7075679302215576,
+      "learning_rate": 1e-06,
+      "loss": 0.0587,
+      "step": 914
+    },
+    {
+      "epoch": 0.1557711950970378,
+      "grad_norm": 1.6121772527694702,
+      "learning_rate": 1e-06,
+      "loss": 0.0511,
+      "step": 915
+    },
+    {
+      "epoch": 0.15594143684031325,
+      "grad_norm": 1.7952123880386353,
+      "learning_rate": 1e-06,
+      "loss": 0.0716,
+      "step": 916
+    },
+    {
+      "epoch": 0.1561116785835887,
+      "grad_norm": 1.8294349908828735,
+      "learning_rate": 1e-06,
+      "loss": 0.0611,
+      "step": 917
+    },
+    {
+      "epoch": 0.15628192032686414,
+      "grad_norm": 1.8930115699768066,
+      "learning_rate": 1e-06,
+      "loss": 0.083,
+      "step": 918
+    },
+    {
+      "epoch": 0.1564521620701396,
+      "grad_norm": 1.7695558071136475,
+      "learning_rate": 1e-06,
+      "loss": 0.0718,
+      "step": 919
+    },
+    {
+      "epoch": 0.15662240381341505,
+      "grad_norm": 1.6812986135482788,
+      "learning_rate": 1e-06,
+      "loss": 0.0709,
+      "step": 920
+    },
+    {
+      "epoch": 0.1567926455566905,
+      "grad_norm": 1.6664036512374878,
+      "learning_rate": 1e-06,
+      "loss": 0.0609,
+      "step": 921
+    },
+    {
+      "epoch": 0.15696288729996596,
+      "grad_norm": 1.7315361499786377,
+      "learning_rate": 1e-06,
+      "loss": 0.0638,
+      "step": 922
+    },
+    {
+      "epoch": 0.1571331290432414,
+      "grad_norm": 1.693820595741272,
+      "learning_rate": 1e-06,
+      "loss": 0.0604,
+      "step": 923
+    },
+    {
+      "epoch": 0.15730337078651685,
+      "grad_norm": 1.9926408529281616,
+      "learning_rate": 1e-06,
+      "loss": 0.065,
+      "step": 924
+    },
+    {
+      "epoch": 0.1574736125297923,
+      "grad_norm": 1.3008970022201538,
+      "learning_rate": 1e-06,
+      "loss": 0.0424,
+      "step": 925
+    },
+    {
+      "epoch": 0.15764385427306776,
+      "grad_norm": 1.7588164806365967,
+      "learning_rate": 1e-06,
+      "loss": 0.0569,
+      "step": 926
+    },
+    {
+      "epoch": 0.15781409601634322,
+      "grad_norm": 1.7521356344223022,
+      "learning_rate": 1e-06,
+      "loss": 0.057,
+      "step": 927
+    },
+    {
+      "epoch": 0.15798433775961865,
+      "grad_norm": 1.6399370431900024,
+      "learning_rate": 1e-06,
+      "loss": 0.055,
+      "step": 928
+    },
+    {
+      "epoch": 0.1581545795028941,
+      "grad_norm": 1.6132686138153076,
+      "learning_rate": 1e-06,
+      "loss": 0.062,
+      "step": 929
+    },
+    {
+      "epoch": 0.15832482124616956,
+      "grad_norm": 1.8911019563674927,
+      "learning_rate": 1e-06,
+      "loss": 0.0673,
+      "step": 930
+    },
+    {
+      "epoch": 0.15849506298944502,
+      "grad_norm": 1.8688396215438843,
+      "learning_rate": 1e-06,
+      "loss": 0.0593,
+      "step": 931
+    },
+    {
+      "epoch": 0.15866530473272047,
+      "grad_norm": 1.948891282081604,
+      "learning_rate": 1e-06,
+      "loss": 0.0717,
+      "step": 932
+    },
+    {
+      "epoch": 0.1588355464759959,
+      "grad_norm": 1.5135903358459473,
+      "learning_rate": 1e-06,
+      "loss": 0.0538,
+      "step": 933
+    },
+    {
+      "epoch": 0.15900578821927136,
+      "grad_norm": 1.983309030532837,
+      "learning_rate": 1e-06,
+      "loss": 0.0656,
+      "step": 934
+    },
+    {
+      "epoch": 0.15917602996254682,
+      "grad_norm": 2.004861354827881,
+      "learning_rate": 1e-06,
+      "loss": 0.0662,
+      "step": 935
+    },
+    {
+      "epoch": 0.15934627170582227,
+      "grad_norm": 1.3474713563919067,
+      "learning_rate": 1e-06,
+      "loss": 0.0607,
+      "step": 936
+    },
+    {
+      "epoch": 0.15951651344909773,
+      "grad_norm": 1.5540618896484375,
+      "learning_rate": 1e-06,
+      "loss": 0.0607,
+      "step": 937
+    },
+    {
+      "epoch": 0.15968675519237316,
+      "grad_norm": 1.8663222789764404,
+      "learning_rate": 1e-06,
+      "loss": 0.0694,
+      "step": 938
+    },
+    {
+      "epoch": 0.15985699693564862,
+      "grad_norm": 1.5832575559616089,
+      "learning_rate": 1e-06,
+      "loss": 0.0722,
+      "step": 939
+    },
+    {
+      "epoch": 0.16002723867892407,
+      "grad_norm": 1.6708149909973145,
+      "learning_rate": 1e-06,
+      "loss": 0.0715,
+      "step": 940
+    },
+    {
+      "epoch": 0.16019748042219953,
+      "grad_norm": 1.7680753469467163,
+      "learning_rate": 1e-06,
+      "loss": 0.0586,
+      "step": 941
+    },
+    {
+      "epoch": 0.16036772216547499,
+      "grad_norm": 1.8743613958358765,
+      "learning_rate": 1e-06,
+      "loss": 0.0568,
+      "step": 942
+    },
+    {
+      "epoch": 0.16053796390875041,
+      "grad_norm": 4.353957176208496,
+      "learning_rate": 1e-06,
+      "loss": 0.1204,
+      "step": 943
+    },
+    {
+      "epoch": 0.16070820565202587,
+      "grad_norm": 1.36697518825531,
+      "learning_rate": 1e-06,
+      "loss": 0.0462,
+      "step": 944
+    },
+    {
+      "epoch": 0.16087844739530133,
+      "grad_norm": 1.474086046218872,
+      "learning_rate": 1e-06,
+      "loss": 0.0508,
+      "step": 945
+    },
+    {
+      "epoch": 0.16104868913857678,
+      "grad_norm": 1.7142692804336548,
+      "learning_rate": 1e-06,
+      "loss": 0.0528,
+      "step": 946
+    },
+    {
+      "epoch": 0.16121893088185224,
+      "grad_norm": 2.086660623550415,
+      "learning_rate": 1e-06,
+      "loss": 0.055,
+      "step": 947
+    },
+    {
+      "epoch": 0.16138917262512767,
+      "grad_norm": 2.088574171066284,
+      "learning_rate": 1e-06,
+      "loss": 0.0755,
+      "step": 948
+    },
+    {
+      "epoch": 0.16155941436840313,
+      "grad_norm": 1.7473136186599731,
+      "learning_rate": 1e-06,
+      "loss": 0.0598,
+      "step": 949
+    },
+    {
+      "epoch": 0.16172965611167858,
+      "grad_norm": 1.3327255249023438,
+      "learning_rate": 1e-06,
+      "loss": 0.049,
+      "step": 950
+    },
+    {
+      "epoch": 0.16189989785495404,
+      "grad_norm": 3.631272792816162,
+      "learning_rate": 1e-06,
+      "loss": 0.084,
+      "step": 951
+    },
+    {
+      "epoch": 0.1620701395982295,
+      "grad_norm": 1.5777565240859985,
+      "learning_rate": 1e-06,
+      "loss": 0.0448,
+      "step": 952
+    },
+    {
+      "epoch": 0.16224038134150492,
+      "grad_norm": 2.02554988861084,
+      "learning_rate": 1e-06,
+      "loss": 0.0661,
+      "step": 953
+    },
+    {
+      "epoch": 0.16241062308478038,
+      "grad_norm": 1.5341172218322754,
+      "learning_rate": 1e-06,
+      "loss": 0.0508,
+      "step": 954
+    },
+    {
+      "epoch": 0.16258086482805584,
+      "grad_norm": 1.34352707862854,
+      "learning_rate": 1e-06,
+      "loss": 0.0397,
+      "step": 955
+    },
+    {
+      "epoch": 0.1627511065713313,
+      "grad_norm": 1.5781581401824951,
+      "learning_rate": 1e-06,
+      "loss": 0.0613,
+      "step": 956
+    },
+    {
+      "epoch": 0.16292134831460675,
+      "grad_norm": 1.7862120866775513,
+      "learning_rate": 1e-06,
+      "loss": 0.0512,
+      "step": 957
+    },
+    {
+      "epoch": 0.16309159005788218,
+      "grad_norm": 2.01899790763855,
+      "learning_rate": 1e-06,
+      "loss": 0.0689,
+      "step": 958
+    },
+    {
+      "epoch": 0.16326183180115764,
+      "grad_norm": 1.6229596138000488,
+      "learning_rate": 1e-06,
+      "loss": 0.0539,
+      "step": 959
+    },
+    {
+      "epoch": 0.1634320735444331,
+      "grad_norm": 1.663812279701233,
+      "learning_rate": 1e-06,
+      "loss": 0.0629,
+      "step": 960
+    },
+    {
+      "epoch": 0.16360231528770855,
+      "grad_norm": 1.2906622886657715,
+      "learning_rate": 1e-06,
+      "loss": 0.0541,
+      "step": 961
+    },
+    {
+      "epoch": 0.163772557030984,
+      "grad_norm": 1.674870491027832,
+      "learning_rate": 1e-06,
+      "loss": 0.0665,
+      "step": 962
+    },
+    {
+      "epoch": 0.16394279877425944,
+      "grad_norm": 1.584000587463379,
+      "learning_rate": 1e-06,
+      "loss": 0.0559,
+      "step": 963
+    },
+    {
+      "epoch": 0.1641130405175349,
+      "grad_norm": 1.8097106218338013,
+      "learning_rate": 1e-06,
+      "loss": 0.0663,
+      "step": 964
+    },
+    {
+      "epoch": 0.16428328226081035,
+      "grad_norm": 1.6824593544006348,
+      "learning_rate": 1e-06,
+      "loss": 0.0592,
+      "step": 965
+    },
+    {
+      "epoch": 0.1644535240040858,
+      "grad_norm": 1.6757420301437378,
+      "learning_rate": 1e-06,
+      "loss": 0.0565,
+      "step": 966
+    },
+    {
+      "epoch": 0.16462376574736126,
+      "grad_norm": 1.91703462600708,
+      "learning_rate": 1e-06,
+      "loss": 0.0617,
+      "step": 967
+    },
+    {
+      "epoch": 0.1647940074906367,
+      "grad_norm": 1.8160761594772339,
+      "learning_rate": 1e-06,
+      "loss": 0.0579,
+      "step": 968
+    },
+    {
+      "epoch": 0.16496424923391215,
+      "grad_norm": 1.3976374864578247,
+      "learning_rate": 1e-06,
+      "loss": 0.0428,
+      "step": 969
+    },
+    {
+      "epoch": 0.1651344909771876,
+      "grad_norm": 2.016111373901367,
+      "learning_rate": 1e-06,
+      "loss": 0.054,
+      "step": 970
+    },
+    {
+      "epoch": 0.16530473272046306,
+      "grad_norm": 1.7259122133255005,
+      "learning_rate": 1e-06,
+      "loss": 0.0488,
+      "step": 971
+    },
+    {
+      "epoch": 0.16547497446373852,
+      "grad_norm": 1.5745218992233276,
+      "learning_rate": 1e-06,
+      "loss": 0.0569,
+      "step": 972
+    },
+    {
+      "epoch": 0.16564521620701397,
+      "grad_norm": 1.3785955905914307,
+      "learning_rate": 1e-06,
+      "loss": 0.0412,
+      "step": 973
+    },
+    {
+      "epoch": 0.1658154579502894,
+      "grad_norm": 1.76412034034729,
+      "learning_rate": 1e-06,
+      "loss": 0.0533,
+      "step": 974
+    },
+    {
+      "epoch": 0.16598569969356486,
+      "grad_norm": 1.6779992580413818,
+      "learning_rate": 1e-06,
+      "loss": 0.0636,
+      "step": 975
+    },
+    {
+      "epoch": 0.16615594143684032,
+      "grad_norm": 1.5417025089263916,
+      "learning_rate": 1e-06,
+      "loss": 0.0477,
+      "step": 976
+    },
+    {
+      "epoch": 0.16632618318011577,
+      "grad_norm": 1.7664419412612915,
+      "learning_rate": 1e-06,
+      "loss": 0.0628,
+      "step": 977
+    },
+    {
+      "epoch": 0.16649642492339123,
+      "grad_norm": 1.905269980430603,
+      "learning_rate": 1e-06,
+      "loss": 0.0582,
+      "step": 978
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 2.259796619415283,
+      "learning_rate": 1e-06,
+      "loss": 0.0721,
+      "step": 979
+    },
+    {
+      "epoch": 0.16683690840994211,
+      "grad_norm": 1.6967008113861084,
+      "learning_rate": 1e-06,
+      "loss": 0.0517,
+      "step": 980
+    },
+    {
+      "epoch": 0.16700715015321757,
+      "grad_norm": 2.18546986579895,
+      "learning_rate": 1e-06,
+      "loss": 0.0645,
+      "step": 981
+    },
+    {
+      "epoch": 0.16717739189649303,
+      "grad_norm": 1.5567885637283325,
+      "learning_rate": 1e-06,
+      "loss": 0.0471,
+      "step": 982
+    },
+    {
+      "epoch": 0.16734763363976848,
+      "grad_norm": 1.516837477684021,
+      "learning_rate": 1e-06,
+      "loss": 0.0442,
+      "step": 983
+    },
+    {
+      "epoch": 0.1675178753830439,
+      "grad_norm": 1.4908742904663086,
+      "learning_rate": 1e-06,
+      "loss": 0.0524,
+      "step": 984
+    },
+    {
+      "epoch": 0.16768811712631937,
+      "grad_norm": 1.570236086845398,
+      "learning_rate": 1e-06,
+      "loss": 0.054,
+      "step": 985
+    },
+    {
+      "epoch": 0.16785835886959483,
+      "grad_norm": 1.672652006149292,
+      "learning_rate": 1e-06,
+      "loss": 0.0613,
+      "step": 986
+    },
+    {
+      "epoch": 0.16802860061287028,
+      "grad_norm": 1.6871075630187988,
+      "learning_rate": 1e-06,
+      "loss": 0.0683,
+      "step": 987
+    },
+    {
+      "epoch": 0.16819884235614574,
+      "grad_norm": 1.6366732120513916,
+      "learning_rate": 1e-06,
+      "loss": 0.0542,
+      "step": 988
+    },
+    {
+      "epoch": 0.16836908409942117,
+      "grad_norm": 1.6381199359893799,
+      "learning_rate": 1e-06,
+      "loss": 0.0522,
+      "step": 989
+    },
+    {
+      "epoch": 0.16853932584269662,
+      "grad_norm": 1.6033238172531128,
+      "learning_rate": 1e-06,
+      "loss": 0.0519,
+      "step": 990
+    },
+    {
+      "epoch": 0.16870956758597208,
+      "grad_norm": 2.074885606765747,
+      "learning_rate": 1e-06,
+      "loss": 0.0668,
+      "step": 991
+    },
+    {
+      "epoch": 0.16887980932924754,
+      "grad_norm": 1.8530831336975098,
+      "learning_rate": 1e-06,
+      "loss": 0.0569,
+      "step": 992
+    },
+    {
+      "epoch": 0.169050051072523,
+      "grad_norm": 1.8304858207702637,
+      "learning_rate": 1e-06,
+      "loss": 0.0613,
+      "step": 993
+    },
+    {
+      "epoch": 0.16922029281579842,
+      "grad_norm": 2.9935953617095947,
+      "learning_rate": 1e-06,
+      "loss": 0.0868,
+      "step": 994
+    },
+    {
+      "epoch": 0.16939053455907388,
+      "grad_norm": 1.593682050704956,
+      "learning_rate": 1e-06,
+      "loss": 0.0525,
+      "step": 995
+    },
+    {
+      "epoch": 0.16956077630234934,
+      "grad_norm": 1.4647144079208374,
+      "learning_rate": 1e-06,
+      "loss": 0.0539,
+      "step": 996
+    },
+    {
+      "epoch": 0.1697310180456248,
+      "grad_norm": 1.6193076372146606,
+      "learning_rate": 1e-06,
+      "loss": 0.0556,
+      "step": 997
+    },
+    {
+      "epoch": 0.16990125978890025,
+      "grad_norm": 1.3825421333312988,
+      "learning_rate": 1e-06,
+      "loss": 0.0485,
+      "step": 998
+    },
+    {
+      "epoch": 0.17007150153217568,
+      "grad_norm": 1.6311777830123901,
+      "learning_rate": 1e-06,
+      "loss": 0.0604,
+      "step": 999
+    },
+    {
+      "epoch": 0.17024174327545114,
+      "grad_norm": 1.859731912612915,
+      "learning_rate": 1e-06,
+      "loss": 0.0722,
+      "step": 1000
+    },
+    {
+      "epoch": 0.17024174327545114,
+      "eval_loss": 0.262949138879776,
+      "eval_runtime": 21.3587,
+      "eval_samples_per_second": 14.046,
+      "eval_steps_per_second": 0.375,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1704119850187266,
+      "grad_norm": 1.5789490938186646,
+      "learning_rate": 1e-06,
+      "loss": 0.0589,
+      "step": 1001
+    },
+    {
+      "epoch": 0.17058222676200205,
+      "grad_norm": 1.7059580087661743,
+      "learning_rate": 1e-06,
+      "loss": 0.0505,
+      "step": 1002
+    },
+    {
+      "epoch": 0.1707524685052775,
+      "grad_norm": 1.5325723886489868,
+      "learning_rate": 1e-06,
+      "loss": 0.0557,
+      "step": 1003
+    },
+    {
+      "epoch": 0.17092271024855293,
+      "grad_norm": 1.3633655309677124,
+      "learning_rate": 1e-06,
+      "loss": 0.0467,
+      "step": 1004
+    },
+    {
+      "epoch": 0.1710929519918284,
+      "grad_norm": 2.3447840213775635,
+      "learning_rate": 1e-06,
+      "loss": 0.0671,
+      "step": 1005
+    },
+    {
+      "epoch": 0.17126319373510385,
+      "grad_norm": 1.9885766506195068,
+      "learning_rate": 1e-06,
+      "loss": 0.0615,
+      "step": 1006
+    },
+    {
+      "epoch": 0.1714334354783793,
+      "grad_norm": 1.5275179147720337,
+      "learning_rate": 1e-06,
+      "loss": 0.0505,
+      "step": 1007
+    },
+    {
+      "epoch": 0.17160367722165476,
+      "grad_norm": 1.6165765523910522,
+      "learning_rate": 1e-06,
+      "loss": 0.0465,
+      "step": 1008
+    },
+    {
+      "epoch": 0.1717739189649302,
+      "grad_norm": 2.4540200233459473,
+      "learning_rate": 1e-06,
+      "loss": 0.0682,
+      "step": 1009
+    },
+    {
+      "epoch": 0.17194416070820565,
+      "grad_norm": 1.5132991075515747,
+      "learning_rate": 1e-06,
+      "loss": 0.0498,
+      "step": 1010
+    },
+    {
+      "epoch": 0.1721144024514811,
+      "grad_norm": 1.4574193954467773,
+      "learning_rate": 1e-06,
+      "loss": 0.0515,
+      "step": 1011
+    },
+    {
+      "epoch": 0.17228464419475656,
+      "grad_norm": 2.0926284790039062,
+      "learning_rate": 1e-06,
+      "loss": 0.0676,
+      "step": 1012
+    },
+    {
+      "epoch": 0.17245488593803202,
+      "grad_norm": 1.316934585571289,
+      "learning_rate": 1e-06,
+      "loss": 0.049,
+      "step": 1013
+    },
+    {
+      "epoch": 0.17262512768130744,
+      "grad_norm": 1.4470518827438354,
+      "learning_rate": 1e-06,
+      "loss": 0.0527,
+      "step": 1014
+    },
+    {
+      "epoch": 0.1727953694245829,
+      "grad_norm": 1.254742980003357,
+      "learning_rate": 1e-06,
+      "loss": 0.04,
+      "step": 1015
+    },
+    {
+      "epoch": 0.17296561116785836,
+      "grad_norm": 1.6093502044677734,
+      "learning_rate": 1e-06,
+      "loss": 0.0575,
+      "step": 1016
+    },
+    {
+      "epoch": 0.17313585291113381,
+      "grad_norm": 1.9037551879882812,
+      "learning_rate": 1e-06,
+      "loss": 0.0602,
+      "step": 1017
+    },
+    {
+      "epoch": 0.17330609465440927,
+      "grad_norm": 1.4383158683776855,
+      "learning_rate": 1e-06,
+      "loss": 0.0481,
+      "step": 1018
+    },
+    {
+      "epoch": 0.1734763363976847,
+      "grad_norm": 1.7174170017242432,
+      "learning_rate": 1e-06,
+      "loss": 0.0677,
+      "step": 1019
+    },
+    {
+      "epoch": 0.17364657814096016,
+      "grad_norm": 1.838952660560608,
+      "learning_rate": 1e-06,
+      "loss": 0.0555,
+      "step": 1020
+    },
+    {
+      "epoch": 0.1738168198842356,
+      "grad_norm": 1.9162826538085938,
+      "learning_rate": 1e-06,
+      "loss": 0.0619,
+      "step": 1021
+    },
+    {
+      "epoch": 0.17398706162751107,
+      "grad_norm": 1.5161962509155273,
+      "learning_rate": 1e-06,
+      "loss": 0.048,
+      "step": 1022
+    },
+    {
+      "epoch": 0.17415730337078653,
+      "grad_norm": 1.8026893138885498,
+      "learning_rate": 1e-06,
+      "loss": 0.0615,
+      "step": 1023
+    },
+    {
+      "epoch": 0.17432754511406195,
+      "grad_norm": 1.637364387512207,
+      "learning_rate": 1e-06,
+      "loss": 0.0497,
+      "step": 1024
+    },
+    {
+      "epoch": 0.1744977868573374,
+      "grad_norm": 1.7427301406860352,
+      "learning_rate": 1e-06,
+      "loss": 0.0467,
+      "step": 1025
+    },
+    {
+      "epoch": 0.17466802860061287,
+      "grad_norm": 1.6015710830688477,
+      "learning_rate": 1e-06,
+      "loss": 0.0456,
+      "step": 1026
+    },
+    {
+      "epoch": 0.17483827034388832,
+      "grad_norm": 1.5106914043426514,
+      "learning_rate": 1e-06,
+      "loss": 0.0583,
+      "step": 1027
+    },
+    {
+      "epoch": 0.17500851208716378,
+      "grad_norm": 1.7485862970352173,
+      "learning_rate": 1e-06,
+      "loss": 0.0743,
+      "step": 1028
+    },
+    {
+      "epoch": 0.1751787538304392,
+      "grad_norm": 1.5311837196350098,
+      "learning_rate": 1e-06,
+      "loss": 0.0555,
+      "step": 1029
+    },
+    {
+      "epoch": 0.17534899557371467,
+      "grad_norm": 1.6525341272354126,
+      "learning_rate": 1e-06,
+      "loss": 0.0566,
+      "step": 1030
+    },
+    {
+      "epoch": 0.17551923731699012,
+      "grad_norm": 1.3050777912139893,
+      "learning_rate": 1e-06,
+      "loss": 0.0454,
+      "step": 1031
+    },
+    {
+      "epoch": 0.17568947906026558,
+      "grad_norm": 1.8630820512771606,
+      "learning_rate": 1e-06,
+      "loss": 0.0615,
+      "step": 1032
+    },
+    {
+      "epoch": 0.17585972080354104,
+      "grad_norm": 1.974867582321167,
+      "learning_rate": 1e-06,
+      "loss": 0.0574,
+      "step": 1033
+    },
+    {
+      "epoch": 0.1760299625468165,
+      "grad_norm": 1.6042343378067017,
+      "learning_rate": 1e-06,
+      "loss": 0.0469,
+      "step": 1034
+    },
+    {
+      "epoch": 0.17620020429009192,
+      "grad_norm": 1.5605472326278687,
+      "learning_rate": 1e-06,
+      "loss": 0.0607,
+      "step": 1035
+    },
+    {
+      "epoch": 0.17637044603336738,
+      "grad_norm": 1.6991108655929565,
+      "learning_rate": 1e-06,
+      "loss": 0.0458,
+      "step": 1036
+    },
+    {
+      "epoch": 0.17654068777664284,
+      "grad_norm": 2.3888437747955322,
+      "learning_rate": 1e-06,
+      "loss": 0.0646,
+      "step": 1037
+    },
+    {
+      "epoch": 0.1767109295199183,
+      "grad_norm": 1.9789478778839111,
+      "learning_rate": 1e-06,
+      "loss": 0.0657,
+      "step": 1038
+    },
+    {
+      "epoch": 0.17688117126319375,
+      "grad_norm": 1.5386762619018555,
+      "learning_rate": 1e-06,
+      "loss": 0.0591,
+      "step": 1039
+    },
+    {
+      "epoch": 0.17705141300646918,
+      "grad_norm": 1.657543659210205,
+      "learning_rate": 1e-06,
+      "loss": 0.0504,
+      "step": 1040
+    },
+    {
+      "epoch": 0.17722165474974463,
+      "grad_norm": 1.5504556894302368,
+      "learning_rate": 1e-06,
+      "loss": 0.0467,
+      "step": 1041
+    },
+    {
+      "epoch": 0.1773918964930201,
+      "grad_norm": 1.718420147895813,
+      "learning_rate": 1e-06,
+      "loss": 0.0559,
+      "step": 1042
+    },
+    {
+      "epoch": 0.17756213823629555,
+      "grad_norm": 2.331953525543213,
+      "learning_rate": 1e-06,
+      "loss": 0.056,
+      "step": 1043
+    },
+    {
+      "epoch": 0.177732379979571,
+      "grad_norm": 1.6991020441055298,
+      "learning_rate": 1e-06,
+      "loss": 0.0474,
+      "step": 1044
+    },
+    {
+      "epoch": 0.17790262172284643,
+      "grad_norm": 1.8949966430664062,
+      "learning_rate": 1e-06,
+      "loss": 0.043,
+      "step": 1045
+    },
+    {
+      "epoch": 0.1780728634661219,
+      "grad_norm": 1.8064547777175903,
+      "learning_rate": 1e-06,
+      "loss": 0.0436,
+      "step": 1046
+    },
+    {
+      "epoch": 0.17824310520939735,
+      "grad_norm": 1.4450627565383911,
+      "learning_rate": 1e-06,
+      "loss": 0.0453,
+      "step": 1047
+    },
+    {
+      "epoch": 0.1784133469526728,
+      "grad_norm": 1.9606854915618896,
+      "learning_rate": 1e-06,
+      "loss": 0.052,
+      "step": 1048
+    },
+    {
+      "epoch": 0.17858358869594826,
+      "grad_norm": 1.8185573816299438,
+      "learning_rate": 1e-06,
+      "loss": 0.0438,
+      "step": 1049
+    },
+    {
+      "epoch": 0.1787538304392237,
+      "grad_norm": 1.8913239240646362,
+      "learning_rate": 1e-06,
+      "loss": 0.0456,
+      "step": 1050
+    },
+    {
+      "epoch": 0.17892407218249914,
+      "grad_norm": 1.6773005723953247,
+      "learning_rate": 1e-06,
+      "loss": 0.0502,
+      "step": 1051
+    },
+    {
+      "epoch": 0.1790943139257746,
+      "grad_norm": 1.6672918796539307,
+      "learning_rate": 1e-06,
+      "loss": 0.0471,
+      "step": 1052
+    },
+    {
+      "epoch": 0.17926455566905006,
+      "grad_norm": 1.8581830263137817,
+      "learning_rate": 1e-06,
+      "loss": 0.0539,
+      "step": 1053
+    },
+    {
+      "epoch": 0.17943479741232551,
+      "grad_norm": 1.4864681959152222,
+      "learning_rate": 1e-06,
+      "loss": 0.0546,
+      "step": 1054
+    },
+    {
+      "epoch": 0.17960503915560094,
+      "grad_norm": 1.527038812637329,
+      "learning_rate": 1e-06,
+      "loss": 0.0417,
+      "step": 1055
+    },
+    {
+      "epoch": 0.1797752808988764,
+      "grad_norm": 1.6635370254516602,
+      "learning_rate": 1e-06,
+      "loss": 0.0457,
+      "step": 1056
+    },
+    {
+      "epoch": 0.17994552264215186,
+      "grad_norm": 1.5412622690200806,
+      "learning_rate": 1e-06,
+      "loss": 0.0437,
+      "step": 1057
+    },
+    {
+      "epoch": 0.1801157643854273,
+      "grad_norm": 1.4829204082489014,
+      "learning_rate": 1e-06,
+      "loss": 0.0493,
+      "step": 1058
+    },
+    {
+      "epoch": 0.18028600612870277,
+      "grad_norm": 1.4343328475952148,
+      "learning_rate": 1e-06,
+      "loss": 0.0405,
+      "step": 1059
+    },
+    {
+      "epoch": 0.1804562478719782,
+      "grad_norm": 1.7797609567642212,
+      "learning_rate": 1e-06,
+      "loss": 0.0464,
+      "step": 1060
+    },
+    {
+      "epoch": 0.18062648961525365,
+      "grad_norm": 1.488383412361145,
+      "learning_rate": 1e-06,
+      "loss": 0.0355,
+      "step": 1061
+    },
+    {
+      "epoch": 0.1807967313585291,
+      "grad_norm": 1.3695396184921265,
+      "learning_rate": 1e-06,
+      "loss": 0.048,
+      "step": 1062
+    },
+    {
+      "epoch": 0.18096697310180457,
+      "grad_norm": 1.8657076358795166,
+      "learning_rate": 1e-06,
+      "loss": 0.0478,
+      "step": 1063
+    },
+    {
+      "epoch": 0.18113721484508002,
+      "grad_norm": 1.7366667985916138,
+      "learning_rate": 1e-06,
+      "loss": 0.0429,
+      "step": 1064
+    },
+    {
+      "epoch": 0.18130745658835545,
+      "grad_norm": 1.4432430267333984,
+      "learning_rate": 1e-06,
+      "loss": 0.0451,
+      "step": 1065
+    },
+    {
+      "epoch": 0.1814776983316309,
+      "grad_norm": 1.6333496570587158,
+      "learning_rate": 1e-06,
+      "loss": 0.0612,
+      "step": 1066
+    },
+    {
+      "epoch": 0.18164794007490637,
+      "grad_norm": 2.0347399711608887,
+      "learning_rate": 1e-06,
+      "loss": 0.0544,
+      "step": 1067
+    },
+    {
+      "epoch": 0.18181818181818182,
+      "grad_norm": 2.103520393371582,
+      "learning_rate": 1e-06,
+      "loss": 0.0528,
+      "step": 1068
+    },
+    {
+      "epoch": 0.18198842356145728,
+      "grad_norm": 2.289806604385376,
+      "learning_rate": 1e-06,
+      "loss": 0.0714,
+      "step": 1069
+    },
+    {
+      "epoch": 0.1821586653047327,
+      "grad_norm": 1.8326396942138672,
+      "learning_rate": 1e-06,
+      "loss": 0.0477,
+      "step": 1070
+    },
+    {
+      "epoch": 0.18232890704800817,
+      "grad_norm": 1.6226252317428589,
+      "learning_rate": 1e-06,
+      "loss": 0.0596,
+      "step": 1071
+    },
+    {
+      "epoch": 0.18249914879128362,
+      "grad_norm": 1.6972850561141968,
+      "learning_rate": 1e-06,
+      "loss": 0.0555,
+      "step": 1072
+    },
+    {
+      "epoch": 0.18266939053455908,
+      "grad_norm": 1.7216845750808716,
+      "learning_rate": 1e-06,
+      "loss": 0.0727,
+      "step": 1073
+    },
+    {
+      "epoch": 0.18283963227783454,
+      "grad_norm": 1.396399974822998,
+      "learning_rate": 1e-06,
+      "loss": 0.0451,
+      "step": 1074
+    },
+    {
+      "epoch": 0.18300987402110996,
+      "grad_norm": 1.5802775621414185,
+      "learning_rate": 1e-06,
+      "loss": 0.0523,
+      "step": 1075
+    },
+    {
+      "epoch": 0.18318011576438542,
+      "grad_norm": 1.5000909566879272,
+      "learning_rate": 1e-06,
+      "loss": 0.0503,
+      "step": 1076
+    },
+    {
+      "epoch": 0.18335035750766088,
+      "grad_norm": 1.5903329849243164,
+      "learning_rate": 1e-06,
+      "loss": 0.0453,
+      "step": 1077
+    },
+    {
+      "epoch": 0.18352059925093633,
+      "grad_norm": 1.3907889127731323,
+      "learning_rate": 1e-06,
+      "loss": 0.0482,
+      "step": 1078
+    },
+    {
+      "epoch": 0.1836908409942118,
+      "grad_norm": 1.4356728792190552,
+      "learning_rate": 1e-06,
+      "loss": 0.0358,
+      "step": 1079
+    },
+    {
+      "epoch": 0.18386108273748722,
+      "grad_norm": 1.425837755203247,
+      "learning_rate": 1e-06,
+      "loss": 0.0481,
+      "step": 1080
+    },
+    {
+      "epoch": 0.18403132448076268,
+      "grad_norm": 1.793095588684082,
+      "learning_rate": 1e-06,
+      "loss": 0.0577,
+      "step": 1081
+    },
+    {
+      "epoch": 0.18420156622403813,
+      "grad_norm": 1.5674705505371094,
+      "learning_rate": 1e-06,
+      "loss": 0.0405,
+      "step": 1082
+    },
+    {
+      "epoch": 0.1843718079673136,
+      "grad_norm": 1.527916431427002,
+      "learning_rate": 1e-06,
+      "loss": 0.0394,
+      "step": 1083
+    },
+    {
+      "epoch": 0.18454204971058905,
+      "grad_norm": 1.8252404928207397,
+      "learning_rate": 1e-06,
+      "loss": 0.0506,
+      "step": 1084
+    },
+    {
+      "epoch": 0.18471229145386447,
+      "grad_norm": 1.6396743059158325,
+      "learning_rate": 1e-06,
+      "loss": 0.0462,
+      "step": 1085
+    },
+    {
+      "epoch": 0.18488253319713993,
+      "grad_norm": 1.7540291547775269,
+      "learning_rate": 1e-06,
+      "loss": 0.051,
+      "step": 1086
+    },
+    {
+      "epoch": 0.1850527749404154,
+      "grad_norm": 1.7107782363891602,
+      "learning_rate": 1e-06,
+      "loss": 0.0552,
+      "step": 1087
+    },
+    {
+      "epoch": 0.18522301668369084,
+      "grad_norm": 1.7277930974960327,
+      "learning_rate": 1e-06,
+      "loss": 0.0558,
+      "step": 1088
+    },
+    {
+      "epoch": 0.1853932584269663,
+      "grad_norm": 1.531584620475769,
+      "learning_rate": 1e-06,
+      "loss": 0.0436,
+      "step": 1089
+    },
+    {
+      "epoch": 0.18556350017024173,
+      "grad_norm": 1.708498477935791,
+      "learning_rate": 1e-06,
+      "loss": 0.0599,
+      "step": 1090
+    },
+    {
+      "epoch": 0.1857337419135172,
+      "grad_norm": 1.5215576887130737,
+      "learning_rate": 1e-06,
+      "loss": 0.0413,
+      "step": 1091
+    },
+    {
+      "epoch": 0.18590398365679264,
+      "grad_norm": 1.6072683334350586,
+      "learning_rate": 1e-06,
+      "loss": 0.0472,
+      "step": 1092
+    },
+    {
+      "epoch": 0.1860742254000681,
+      "grad_norm": 2.161273717880249,
+      "learning_rate": 1e-06,
+      "loss": 0.053,
+      "step": 1093
+    },
+    {
+      "epoch": 0.18624446714334356,
+      "grad_norm": 2.1874208450317383,
+      "learning_rate": 1e-06,
+      "loss": 0.0676,
+      "step": 1094
+    },
+    {
+      "epoch": 0.186414708886619,
+      "grad_norm": 1.7869967222213745,
+      "learning_rate": 1e-06,
+      "loss": 0.0566,
+      "step": 1095
+    },
+    {
+      "epoch": 0.18658495062989444,
+      "grad_norm": 1.7522093057632446,
+      "learning_rate": 1e-06,
+      "loss": 0.0482,
+      "step": 1096
+    },
+    {
+      "epoch": 0.1867551923731699,
+      "grad_norm": 1.5887916088104248,
+      "learning_rate": 1e-06,
+      "loss": 0.0439,
+      "step": 1097
+    },
+    {
+      "epoch": 0.18692543411644535,
+      "grad_norm": 1.3974252939224243,
+      "learning_rate": 1e-06,
+      "loss": 0.0413,
+      "step": 1098
+    },
+    {
+      "epoch": 0.1870956758597208,
+      "grad_norm": 1.798572301864624,
+      "learning_rate": 1e-06,
+      "loss": 0.0612,
+      "step": 1099
+    },
+    {
+      "epoch": 0.18726591760299627,
+      "grad_norm": 2.0376672744750977,
+      "learning_rate": 1e-06,
+      "loss": 0.0512,
+      "step": 1100
+    },
+    {
+      "epoch": 0.1874361593462717,
+      "grad_norm": 1.7600435018539429,
+      "learning_rate": 1e-06,
+      "loss": 0.0516,
+      "step": 1101
+    },
+    {
+      "epoch": 0.18760640108954715,
+      "grad_norm": 1.638627529144287,
+      "learning_rate": 1e-06,
+      "loss": 0.0438,
+      "step": 1102
+    },
+    {
+      "epoch": 0.1877766428328226,
+      "grad_norm": 1.4002454280853271,
+      "learning_rate": 1e-06,
+      "loss": 0.0406,
+      "step": 1103
+    },
+    {
+      "epoch": 0.18794688457609807,
+      "grad_norm": 1.3637229204177856,
+      "learning_rate": 1e-06,
+      "loss": 0.0363,
+      "step": 1104
+    },
+    {
+      "epoch": 0.18811712631937352,
+      "grad_norm": 1.7648390531539917,
+      "learning_rate": 1e-06,
+      "loss": 0.0491,
+      "step": 1105
+    },
+    {
+      "epoch": 0.18828736806264895,
+      "grad_norm": 1.8793604373931885,
+      "learning_rate": 1e-06,
+      "loss": 0.0658,
+      "step": 1106
+    },
+    {
+      "epoch": 0.1884576098059244,
+      "grad_norm": 1.87686288356781,
+      "learning_rate": 1e-06,
+      "loss": 0.0495,
+      "step": 1107
+    },
+    {
+      "epoch": 0.18862785154919987,
+      "grad_norm": 1.5583778619766235,
+      "learning_rate": 1e-06,
+      "loss": 0.047,
+      "step": 1108
+    },
+    {
+      "epoch": 0.18879809329247532,
+      "grad_norm": 1.7888959646224976,
+      "learning_rate": 1e-06,
+      "loss": 0.0373,
+      "step": 1109
+    },
+    {
+      "epoch": 0.18896833503575078,
+      "grad_norm": 1.628745436668396,
+      "learning_rate": 1e-06,
+      "loss": 0.0405,
+      "step": 1110
+    },
+    {
+      "epoch": 0.1891385767790262,
+      "grad_norm": 2.113985776901245,
+      "learning_rate": 1e-06,
+      "loss": 0.0488,
+      "step": 1111
+    },
+    {
+      "epoch": 0.18930881852230166,
+      "grad_norm": 2.1266591548919678,
+      "learning_rate": 1e-06,
+      "loss": 0.058,
+      "step": 1112
+    },
+    {
+      "epoch": 0.18947906026557712,
+      "grad_norm": 1.9797148704528809,
+      "learning_rate": 1e-06,
+      "loss": 0.0585,
+      "step": 1113
+    },
+    {
+      "epoch": 0.18964930200885258,
+      "grad_norm": 1.6259199380874634,
+      "learning_rate": 1e-06,
+      "loss": 0.0519,
+      "step": 1114
+    },
+    {
+      "epoch": 0.18981954375212803,
+      "grad_norm": 1.7262063026428223,
+      "learning_rate": 1e-06,
+      "loss": 0.038,
+      "step": 1115
+    },
+    {
+      "epoch": 0.18998978549540346,
+      "grad_norm": 1.5812803506851196,
+      "learning_rate": 1e-06,
+      "loss": 0.0487,
+      "step": 1116
+    },
+    {
+      "epoch": 0.19016002723867892,
+      "grad_norm": 2.021449327468872,
+      "learning_rate": 1e-06,
+      "loss": 0.057,
+      "step": 1117
+    },
+    {
+      "epoch": 0.19033026898195438,
+      "grad_norm": 1.7292096614837646,
+      "learning_rate": 1e-06,
+      "loss": 0.0379,
+      "step": 1118
+    },
+    {
+      "epoch": 0.19050051072522983,
+      "grad_norm": 1.4616914987564087,
+      "learning_rate": 1e-06,
+      "loss": 0.0401,
+      "step": 1119
+    },
+    {
+      "epoch": 0.1906707524685053,
+      "grad_norm": 1.4097241163253784,
+      "learning_rate": 1e-06,
+      "loss": 0.0463,
+      "step": 1120
+    },
+    {
+      "epoch": 0.19084099421178072,
+      "grad_norm": 1.8328516483306885,
+      "learning_rate": 1e-06,
+      "loss": 0.0454,
+      "step": 1121
+    },
+    {
+      "epoch": 0.19101123595505617,
+      "grad_norm": 2.055018663406372,
+      "learning_rate": 1e-06,
+      "loss": 0.0606,
+      "step": 1122
+    },
+    {
+      "epoch": 0.19118147769833163,
+      "grad_norm": 1.4977216720581055,
+      "learning_rate": 1e-06,
+      "loss": 0.0347,
+      "step": 1123
+    },
+    {
+      "epoch": 0.1913517194416071,
+      "grad_norm": 1.8649815320968628,
+      "learning_rate": 1e-06,
+      "loss": 0.0484,
+      "step": 1124
+    },
+    {
+      "epoch": 0.19152196118488254,
+      "grad_norm": 1.5965502262115479,
+      "learning_rate": 1e-06,
+      "loss": 0.0413,
+      "step": 1125
+    },
+    {
+      "epoch": 0.19169220292815797,
+      "grad_norm": 1.7801200151443481,
+      "learning_rate": 1e-06,
+      "loss": 0.0462,
+      "step": 1126
+    },
+    {
+      "epoch": 0.19186244467143343,
+      "grad_norm": 5.706264495849609,
+      "learning_rate": 1e-06,
+      "loss": 0.1093,
+      "step": 1127
+    },
+    {
+      "epoch": 0.1920326864147089,
+      "grad_norm": 1.6867637634277344,
+      "learning_rate": 1e-06,
+      "loss": 0.0417,
+      "step": 1128
+    },
+    {
+      "epoch": 0.19220292815798434,
+      "grad_norm": 1.7014727592468262,
+      "learning_rate": 1e-06,
+      "loss": 0.0461,
+      "step": 1129
+    },
+    {
+      "epoch": 0.1923731699012598,
+      "grad_norm": 1.6920650005340576,
+      "learning_rate": 1e-06,
+      "loss": 0.0513,
+      "step": 1130
+    },
+    {
+      "epoch": 0.19254341164453523,
+      "grad_norm": 1.7450604438781738,
+      "learning_rate": 1e-06,
+      "loss": 0.0562,
+      "step": 1131
+    },
+    {
+      "epoch": 0.19271365338781068,
+      "grad_norm": 1.501325011253357,
+      "learning_rate": 1e-06,
+      "loss": 0.0447,
+      "step": 1132
+    },
+    {
+      "epoch": 0.19288389513108614,
+      "grad_norm": 1.8626445531845093,
+      "learning_rate": 1e-06,
+      "loss": 0.0755,
+      "step": 1133
+    },
+    {
+      "epoch": 0.1930541368743616,
+      "grad_norm": 1.3214149475097656,
+      "learning_rate": 1e-06,
+      "loss": 0.0405,
+      "step": 1134
+    },
+    {
+      "epoch": 0.19322437861763705,
+      "grad_norm": 1.639206886291504,
+      "learning_rate": 1e-06,
+      "loss": 0.0486,
+      "step": 1135
+    },
+    {
+      "epoch": 0.19339462036091248,
+      "grad_norm": 1.6951420307159424,
+      "learning_rate": 1e-06,
+      "loss": 0.0458,
+      "step": 1136
+    },
+    {
+      "epoch": 0.19356486210418794,
+      "grad_norm": 1.5768959522247314,
+      "learning_rate": 1e-06,
+      "loss": 0.0466,
+      "step": 1137
+    },
+    {
+      "epoch": 0.1937351038474634,
+      "grad_norm": 1.6302675008773804,
+      "learning_rate": 1e-06,
+      "loss": 0.0413,
+      "step": 1138
+    },
+    {
+      "epoch": 0.19390534559073885,
+      "grad_norm": 1.4018832445144653,
+      "learning_rate": 1e-06,
+      "loss": 0.0429,
+      "step": 1139
+    },
+    {
+      "epoch": 0.1940755873340143,
+      "grad_norm": 1.482530951499939,
+      "learning_rate": 1e-06,
+      "loss": 0.0342,
+      "step": 1140
+    },
+    {
+      "epoch": 0.19424582907728974,
+      "grad_norm": 1.4372907876968384,
+      "learning_rate": 1e-06,
+      "loss": 0.0384,
+      "step": 1141
+    },
+    {
+      "epoch": 0.1944160708205652,
+      "grad_norm": 2.328382730484009,
+      "learning_rate": 1e-06,
+      "loss": 0.0549,
+      "step": 1142
+    },
+    {
+      "epoch": 0.19458631256384065,
+      "grad_norm": 1.780719518661499,
+      "learning_rate": 1e-06,
+      "loss": 0.0468,
+      "step": 1143
+    },
+    {
+      "epoch": 0.1947565543071161,
+      "grad_norm": 1.8703736066818237,
+      "learning_rate": 1e-06,
+      "loss": 0.044,
+      "step": 1144
+    },
+    {
+      "epoch": 0.19492679605039157,
+      "grad_norm": 1.92543625831604,
+      "learning_rate": 1e-06,
+      "loss": 0.0583,
+      "step": 1145
+    },
+    {
+      "epoch": 0.195097037793667,
+      "grad_norm": 1.7063385248184204,
+      "learning_rate": 1e-06,
+      "loss": 0.0481,
+      "step": 1146
+    },
+    {
+      "epoch": 0.19526727953694245,
+      "grad_norm": 1.2378170490264893,
+      "learning_rate": 1e-06,
+      "loss": 0.0309,
+      "step": 1147
+    },
+    {
+      "epoch": 0.1954375212802179,
+      "grad_norm": 1.4159069061279297,
+      "learning_rate": 1e-06,
+      "loss": 0.0413,
+      "step": 1148
+    },
+    {
+      "epoch": 0.19560776302349336,
+      "grad_norm": 1.816462516784668,
+      "learning_rate": 1e-06,
+      "loss": 0.0421,
+      "step": 1149
+    },
+    {
+      "epoch": 0.19577800476676882,
+      "grad_norm": 1.6749504804611206,
+      "learning_rate": 1e-06,
+      "loss": 0.0514,
+      "step": 1150
+    },
+    {
+      "epoch": 0.19594824651004425,
+      "grad_norm": 1.813492774963379,
+      "learning_rate": 1e-06,
+      "loss": 0.0387,
+      "step": 1151
+    },
+    {
+      "epoch": 0.1961184882533197,
+      "grad_norm": 1.8329745531082153,
+      "learning_rate": 1e-06,
+      "loss": 0.0493,
+      "step": 1152
+    },
+    {
+      "epoch": 0.19628872999659516,
+      "grad_norm": 2.6087090969085693,
+      "learning_rate": 1e-06,
+      "loss": 0.0479,
+      "step": 1153
+    },
+    {
+      "epoch": 0.19645897173987062,
+      "grad_norm": 2.178959369659424,
+      "learning_rate": 1e-06,
+      "loss": 0.0515,
+      "step": 1154
+    },
+    {
+      "epoch": 0.19662921348314608,
+      "grad_norm": 1.7682230472564697,
+      "learning_rate": 1e-06,
+      "loss": 0.0426,
+      "step": 1155
+    },
+    {
+      "epoch": 0.1967994552264215,
+      "grad_norm": 1.9333738088607788,
+      "learning_rate": 1e-06,
+      "loss": 0.0395,
+      "step": 1156
+    },
+    {
+      "epoch": 0.19696969696969696,
+      "grad_norm": 2.0627245903015137,
+      "learning_rate": 1e-06,
+      "loss": 0.0678,
+      "step": 1157
+    },
+    {
+      "epoch": 0.19713993871297242,
+      "grad_norm": 1.649121642112732,
+      "learning_rate": 1e-06,
+      "loss": 0.0548,
+      "step": 1158
+    },
+    {
+      "epoch": 0.19731018045624787,
+      "grad_norm": 1.3847073316574097,
+      "learning_rate": 1e-06,
+      "loss": 0.0336,
+      "step": 1159
+    },
+    {
+      "epoch": 0.19748042219952333,
+      "grad_norm": 1.4557738304138184,
+      "learning_rate": 1e-06,
+      "loss": 0.0334,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1976506639427988,
+      "grad_norm": 1.7297767400741577,
+      "learning_rate": 1e-06,
+      "loss": 0.0404,
+      "step": 1161
+    },
+    {
+      "epoch": 0.19782090568607422,
+      "grad_norm": 1.6031403541564941,
+      "learning_rate": 1e-06,
+      "loss": 0.0457,
+      "step": 1162
+    },
+    {
+      "epoch": 0.19799114742934967,
+      "grad_norm": 1.4427376985549927,
+      "learning_rate": 1e-06,
+      "loss": 0.0394,
+      "step": 1163
+    },
+    {
+      "epoch": 0.19816138917262513,
+      "grad_norm": 1.751774787902832,
+      "learning_rate": 1e-06,
+      "loss": 0.0394,
+      "step": 1164
+    },
+    {
+      "epoch": 0.1983316309159006,
+      "grad_norm": 1.5437021255493164,
+      "learning_rate": 1e-06,
+      "loss": 0.0405,
+      "step": 1165
+    },
+    {
+      "epoch": 0.19850187265917604,
+      "grad_norm": 1.793131947517395,
+      "learning_rate": 1e-06,
+      "loss": 0.0521,
+      "step": 1166
+    },
+    {
+      "epoch": 0.19867211440245147,
+      "grad_norm": 1.4936705827713013,
+      "learning_rate": 1e-06,
+      "loss": 0.0365,
+      "step": 1167
+    },
+    {
+      "epoch": 0.19884235614572693,
+      "grad_norm": 1.961836814880371,
+      "learning_rate": 1e-06,
+      "loss": 0.0466,
+      "step": 1168
+    },
+    {
+      "epoch": 0.19901259788900239,
+      "grad_norm": 1.4718447923660278,
+      "learning_rate": 1e-06,
+      "loss": 0.0377,
+      "step": 1169
+    },
+    {
+      "epoch": 0.19918283963227784,
+      "grad_norm": 1.747491717338562,
+      "learning_rate": 1e-06,
+      "loss": 0.0542,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1993530813755533,
+      "grad_norm": 1.4570775032043457,
+      "learning_rate": 1e-06,
+      "loss": 0.0378,
+      "step": 1171
+    },
+    {
+      "epoch": 0.19952332311882873,
+      "grad_norm": 1.637116551399231,
+      "learning_rate": 1e-06,
+      "loss": 0.0397,
+      "step": 1172
+    },
+    {
+      "epoch": 0.19969356486210418,
+      "grad_norm": 1.496801733970642,
+      "learning_rate": 1e-06,
+      "loss": 0.0526,
+      "step": 1173
+    },
+    {
+      "epoch": 0.19986380660537964,
+      "grad_norm": 1.7981535196304321,
+      "learning_rate": 1e-06,
+      "loss": 0.0393,
+      "step": 1174
+    },
+    {
+      "epoch": 0.2000340483486551,
+      "grad_norm": 1.7196568250656128,
+      "learning_rate": 1e-06,
+      "loss": 0.0409,
+      "step": 1175
+    },
+    {
+      "epoch": 0.20020429009193055,
+      "grad_norm": 2.1747894287109375,
+      "learning_rate": 1e-06,
+      "loss": 0.0571,
+      "step": 1176
+    },
+    {
+      "epoch": 0.20037453183520598,
+      "grad_norm": 1.8108248710632324,
+      "learning_rate": 1e-06,
+      "loss": 0.0495,
+      "step": 1177
+    },
+    {
+      "epoch": 0.20054477357848144,
+      "grad_norm": 1.820467233657837,
+      "learning_rate": 1e-06,
+      "loss": 0.0341,
+      "step": 1178
+    },
+    {
+      "epoch": 0.2007150153217569,
+      "grad_norm": 1.7773289680480957,
+      "learning_rate": 1e-06,
+      "loss": 0.0464,
+      "step": 1179
+    },
+    {
+      "epoch": 0.20088525706503235,
+      "grad_norm": 1.433434247970581,
+      "learning_rate": 1e-06,
+      "loss": 0.0351,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2010554988083078,
+      "grad_norm": 1.836634874343872,
+      "learning_rate": 1e-06,
+      "loss": 0.0588,
+      "step": 1181
+    },
+    {
+      "epoch": 0.20122574055158324,
+      "grad_norm": 1.8555299043655396,
+      "learning_rate": 1e-06,
+      "loss": 0.0517,
+      "step": 1182
+    },
+    {
+      "epoch": 0.2013959822948587,
+      "grad_norm": 1.853201985359192,
+      "learning_rate": 1e-06,
+      "loss": 0.0396,
+      "step": 1183
+    },
+    {
+      "epoch": 0.20156622403813415,
+      "grad_norm": 1.4348528385162354,
+      "learning_rate": 1e-06,
+      "loss": 0.0393,
+      "step": 1184
+    },
+    {
+      "epoch": 0.2017364657814096,
+      "grad_norm": 1.7525590658187866,
+      "learning_rate": 1e-06,
+      "loss": 0.043,
+      "step": 1185
+    },
+    {
+      "epoch": 0.20190670752468506,
+      "grad_norm": 1.9075602293014526,
+      "learning_rate": 1e-06,
+      "loss": 0.0399,
+      "step": 1186
+    },
+    {
+      "epoch": 0.2020769492679605,
+      "grad_norm": 1.6843738555908203,
+      "learning_rate": 1e-06,
+      "loss": 0.0434,
+      "step": 1187
+    },
+    {
+      "epoch": 0.20224719101123595,
+      "grad_norm": 1.3292150497436523,
+      "learning_rate": 1e-06,
+      "loss": 0.0267,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2024174327545114,
+      "grad_norm": 1.8276677131652832,
+      "learning_rate": 1e-06,
+      "loss": 0.0394,
+      "step": 1189
+    },
+    {
+      "epoch": 0.20258767449778686,
+      "grad_norm": 1.8090132474899292,
+      "learning_rate": 1e-06,
+      "loss": 0.0448,
+      "step": 1190
+    },
+    {
+      "epoch": 0.20275791624106232,
+      "grad_norm": 1.500030755996704,
+      "learning_rate": 1e-06,
+      "loss": 0.0251,
+      "step": 1191
+    },
+    {
+      "epoch": 0.20292815798433775,
+      "grad_norm": 1.9649564027786255,
+      "learning_rate": 1e-06,
+      "loss": 0.0413,
+      "step": 1192
+    },
+    {
+      "epoch": 0.2030983997276132,
+      "grad_norm": 1.6460673809051514,
+      "learning_rate": 1e-06,
+      "loss": 0.0322,
+      "step": 1193
+    },
+    {
+      "epoch": 0.20326864147088866,
+      "grad_norm": 2.098881959915161,
+      "learning_rate": 1e-06,
+      "loss": 0.0381,
+      "step": 1194
+    },
+    {
+      "epoch": 0.20343888321416412,
+      "grad_norm": 1.7853596210479736,
+      "learning_rate": 1e-06,
+      "loss": 0.0428,
+      "step": 1195
+    },
+    {
+      "epoch": 0.20360912495743957,
+      "grad_norm": 2.36584210395813,
+      "learning_rate": 1e-06,
+      "loss": 0.0648,
+      "step": 1196
+    },
+    {
+      "epoch": 0.203779366700715,
+      "grad_norm": 2.340277671813965,
+      "learning_rate": 1e-06,
+      "loss": 0.0431,
+      "step": 1197
+    },
+    {
+      "epoch": 0.20394960844399046,
+      "grad_norm": 1.4898146390914917,
+      "learning_rate": 1e-06,
+      "loss": 0.0441,
+      "step": 1198
+    },
+    {
+      "epoch": 0.20411985018726592,
+      "grad_norm": 1.8527204990386963,
+      "learning_rate": 1e-06,
+      "loss": 0.0483,
+      "step": 1199
+    },
+    {
+      "epoch": 0.20429009193054137,
+      "grad_norm": 1.7014769315719604,
+      "learning_rate": 1e-06,
+      "loss": 0.0329,
+      "step": 1200
+    },
+    {
+      "epoch": 0.20446033367381683,
+      "grad_norm": 1.4059642553329468,
+      "learning_rate": 1e-06,
+      "loss": 0.0355,
+      "step": 1201
+    },
+    {
+      "epoch": 0.20463057541709226,
+      "grad_norm": 2.0433590412139893,
+      "learning_rate": 1e-06,
+      "loss": 0.0547,
+      "step": 1202
+    },
+    {
+      "epoch": 0.20480081716036772,
+      "grad_norm": 1.3744386434555054,
+      "learning_rate": 1e-06,
+      "loss": 0.0304,
+      "step": 1203
+    },
+    {
+      "epoch": 0.20497105890364317,
+      "grad_norm": 1.6091663837432861,
+      "learning_rate": 1e-06,
+      "loss": 0.0469,
+      "step": 1204
+    },
+    {
+      "epoch": 0.20514130064691863,
+      "grad_norm": 1.5374170541763306,
+      "learning_rate": 1e-06,
+      "loss": 0.0378,
+      "step": 1205
+    },
+    {
+      "epoch": 0.20531154239019409,
+      "grad_norm": 1.6638165712356567,
+      "learning_rate": 1e-06,
+      "loss": 0.04,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2054817841334695,
+      "grad_norm": 1.4747909307479858,
+      "learning_rate": 1e-06,
+      "loss": 0.05,
+      "step": 1207
+    },
+    {
+      "epoch": 0.20565202587674497,
+      "grad_norm": 1.9081274271011353,
+      "learning_rate": 1e-06,
+      "loss": 0.0428,
+      "step": 1208
+    },
+    {
+      "epoch": 0.20582226762002043,
+      "grad_norm": 1.4791818857192993,
+      "learning_rate": 1e-06,
+      "loss": 0.0359,
+      "step": 1209
+    },
+    {
+      "epoch": 0.20599250936329588,
+      "grad_norm": 1.411188006401062,
+      "learning_rate": 1e-06,
+      "loss": 0.0325,
+      "step": 1210
+    },
+    {
+      "epoch": 0.20616275110657134,
+      "grad_norm": 2.2119061946868896,
+      "learning_rate": 1e-06,
+      "loss": 0.0473,
+      "step": 1211
+    },
+    {
+      "epoch": 0.20633299284984677,
+      "grad_norm": 1.4180293083190918,
+      "learning_rate": 1e-06,
+      "loss": 0.0399,
+      "step": 1212
+    },
+    {
+      "epoch": 0.20650323459312223,
+      "grad_norm": 1.3908727169036865,
+      "learning_rate": 1e-06,
+      "loss": 0.0362,
+      "step": 1213
+    },
+    {
+      "epoch": 0.20667347633639768,
+      "grad_norm": 2.12986421585083,
+      "learning_rate": 1e-06,
+      "loss": 0.0596,
+      "step": 1214
+    },
+    {
+      "epoch": 0.20684371807967314,
+      "grad_norm": 1.529612421989441,
+      "learning_rate": 1e-06,
+      "loss": 0.0486,
+      "step": 1215
+    },
+    {
+      "epoch": 0.2070139598229486,
+      "grad_norm": 1.533721685409546,
+      "learning_rate": 1e-06,
+      "loss": 0.0343,
+      "step": 1216
+    },
+    {
+      "epoch": 0.20718420156622402,
+      "grad_norm": 1.1491042375564575,
+      "learning_rate": 1e-06,
+      "loss": 0.0301,
+      "step": 1217
+    },
+    {
+      "epoch": 0.20735444330949948,
+      "grad_norm": 1.4477553367614746,
+      "learning_rate": 1e-06,
+      "loss": 0.0401,
+      "step": 1218
+    },
+    {
+      "epoch": 0.20752468505277494,
+      "grad_norm": 1.9513334035873413,
+      "learning_rate": 1e-06,
+      "loss": 0.0453,
+      "step": 1219
+    },
+    {
+      "epoch": 0.2076949267960504,
+      "grad_norm": 1.6228469610214233,
+      "learning_rate": 1e-06,
+      "loss": 0.0335,
+      "step": 1220
+    },
+    {
+      "epoch": 0.20786516853932585,
+      "grad_norm": 1.736630916595459,
+      "learning_rate": 1e-06,
+      "loss": 0.0431,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2080354102826013,
+      "grad_norm": 1.7524003982543945,
+      "learning_rate": 1e-06,
+      "loss": 0.0354,
+      "step": 1222
+    },
+    {
+      "epoch": 0.20820565202587674,
+      "grad_norm": 1.9606053829193115,
+      "learning_rate": 1e-06,
+      "loss": 0.0413,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2083758937691522,
+      "grad_norm": 1.4940640926361084,
+      "learning_rate": 1e-06,
+      "loss": 0.0351,
+      "step": 1224
+    },
+    {
+      "epoch": 0.20854613551242765,
+      "grad_norm": 1.542036533355713,
+      "learning_rate": 1e-06,
+      "loss": 0.0445,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2087163772557031,
+      "grad_norm": 1.5856529474258423,
+      "learning_rate": 1e-06,
+      "loss": 0.0389,
+      "step": 1226
+    },
+    {
+      "epoch": 0.20888661899897856,
+      "grad_norm": 1.537645697593689,
+      "learning_rate": 1e-06,
+      "loss": 0.0402,
+      "step": 1227
+    },
+    {
+      "epoch": 0.209056860742254,
+      "grad_norm": 1.4162458181381226,
+      "learning_rate": 1e-06,
+      "loss": 0.0421,
+      "step": 1228
+    },
+    {
+      "epoch": 0.20922710248552945,
+      "grad_norm": 2.1916866302490234,
+      "learning_rate": 1e-06,
+      "loss": 0.038,
+      "step": 1229
+    },
+    {
+      "epoch": 0.2093973442288049,
+      "grad_norm": 1.5257288217544556,
+      "learning_rate": 1e-06,
+      "loss": 0.0417,
+      "step": 1230
+    },
+    {
+      "epoch": 0.20956758597208036,
+      "grad_norm": 1.9720427989959717,
+      "learning_rate": 1e-06,
+      "loss": 0.0439,
+      "step": 1231
+    },
+    {
+      "epoch": 0.20973782771535582,
+      "grad_norm": 1.6400502920150757,
+      "learning_rate": 1e-06,
+      "loss": 0.0373,
+      "step": 1232
+    },
+    {
+      "epoch": 0.20990806945863125,
+      "grad_norm": 1.5405869483947754,
+      "learning_rate": 1e-06,
+      "loss": 0.0459,
+      "step": 1233
+    },
+    {
+      "epoch": 0.2100783112019067,
+      "grad_norm": 1.7419536113739014,
+      "learning_rate": 1e-06,
+      "loss": 0.0407,
+      "step": 1234
+    },
+    {
+      "epoch": 0.21024855294518216,
+      "grad_norm": 1.5766996145248413,
+      "learning_rate": 1e-06,
+      "loss": 0.0347,
+      "step": 1235
+    },
+    {
+      "epoch": 0.21041879468845762,
+      "grad_norm": 2.0112199783325195,
+      "learning_rate": 1e-06,
+      "loss": 0.058,
+      "step": 1236
+    },
+    {
+      "epoch": 0.21058903643173307,
+      "grad_norm": 2.1589255332946777,
+      "learning_rate": 1e-06,
+      "loss": 0.0524,
+      "step": 1237
+    },
+    {
+      "epoch": 0.2107592781750085,
+      "grad_norm": 1.5398106575012207,
+      "learning_rate": 1e-06,
+      "loss": 0.0405,
+      "step": 1238
+    },
+    {
+      "epoch": 0.21092951991828396,
+      "grad_norm": 5.148733139038086,
+      "learning_rate": 1e-06,
+      "loss": 0.0627,
+      "step": 1239
+    },
+    {
+      "epoch": 0.21109976166155942,
+      "grad_norm": 1.5265625715255737,
+      "learning_rate": 1e-06,
+      "loss": 0.0405,
+      "step": 1240
+    },
+    {
+      "epoch": 0.21127000340483487,
+      "grad_norm": 2.0388882160186768,
+      "learning_rate": 1e-06,
+      "loss": 0.0422,
+      "step": 1241
+    },
+    {
+      "epoch": 0.21144024514811033,
+      "grad_norm": 1.5845919847488403,
+      "learning_rate": 1e-06,
+      "loss": 0.0388,
+      "step": 1242
+    },
+    {
+      "epoch": 0.21161048689138576,
+      "grad_norm": 1.984523057937622,
+      "learning_rate": 1e-06,
+      "loss": 0.0507,
+      "step": 1243
+    },
+    {
+      "epoch": 0.21178072863466121,
+      "grad_norm": 1.4896509647369385,
+      "learning_rate": 1e-06,
+      "loss": 0.0357,
+      "step": 1244
+    },
+    {
+      "epoch": 0.21195097037793667,
+      "grad_norm": 1.240064024925232,
+      "learning_rate": 1e-06,
+      "loss": 0.0345,
+      "step": 1245
+    },
+    {
+      "epoch": 0.21212121212121213,
+      "grad_norm": 1.753086805343628,
+      "learning_rate": 1e-06,
+      "loss": 0.0492,
+      "step": 1246
+    },
+    {
+      "epoch": 0.21229145386448758,
+      "grad_norm": 1.9514403343200684,
+      "learning_rate": 1e-06,
+      "loss": 0.038,
+      "step": 1247
+    },
+    {
+      "epoch": 0.212461695607763,
+      "grad_norm": 1.8083099126815796,
+      "learning_rate": 1e-06,
+      "loss": 0.0425,
+      "step": 1248
+    },
+    {
+      "epoch": 0.21263193735103847,
+      "grad_norm": 1.6861008405685425,
+      "learning_rate": 1e-06,
+      "loss": 0.043,
+      "step": 1249
+    },
+    {
+      "epoch": 0.21280217909431393,
+      "grad_norm": 1.4214813709259033,
+      "learning_rate": 1e-06,
+      "loss": 0.037,
+      "step": 1250
+    },
+    {
+      "epoch": 0.21297242083758938,
+      "grad_norm": 1.8322423696517944,
+      "learning_rate": 1e-06,
+      "loss": 0.0427,
+      "step": 1251
+    },
+    {
+      "epoch": 0.21314266258086484,
+      "grad_norm": 4.470928192138672,
+      "learning_rate": 1e-06,
+      "loss": 0.0992,
+      "step": 1252
+    },
+    {
+      "epoch": 0.21331290432414027,
+      "grad_norm": 1.9335957765579224,
+      "learning_rate": 1e-06,
+      "loss": 0.0402,
+      "step": 1253
+    },
+    {
+      "epoch": 0.21348314606741572,
+      "grad_norm": 1.5598244667053223,
+      "learning_rate": 1e-06,
+      "loss": 0.0402,
+      "step": 1254
+    },
+    {
+      "epoch": 0.21365338781069118,
+      "grad_norm": 1.5040446519851685,
+      "learning_rate": 1e-06,
+      "loss": 0.0361,
+      "step": 1255
+    },
+    {
+      "epoch": 0.21382362955396664,
+      "grad_norm": 2.0225491523742676,
+      "learning_rate": 1e-06,
+      "loss": 0.0518,
+      "step": 1256
+    },
+    {
+      "epoch": 0.2139938712972421,
+      "grad_norm": 1.7718695402145386,
+      "learning_rate": 1e-06,
+      "loss": 0.0487,
+      "step": 1257
+    },
+    {
+      "epoch": 0.21416411304051752,
+      "grad_norm": 2.1986560821533203,
+      "learning_rate": 1e-06,
+      "loss": 0.0538,
+      "step": 1258
+    },
+    {
+      "epoch": 0.21433435478379298,
+      "grad_norm": 1.7088544368743896,
+      "learning_rate": 1e-06,
+      "loss": 0.0452,
+      "step": 1259
+    },
+    {
+      "epoch": 0.21450459652706844,
+      "grad_norm": 1.5693634748458862,
+      "learning_rate": 1e-06,
+      "loss": 0.036,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2146748382703439,
+      "grad_norm": 1.8983958959579468,
+      "learning_rate": 1e-06,
+      "loss": 0.0543,
+      "step": 1261
+    },
+    {
+      "epoch": 0.21484508001361935,
+      "grad_norm": 1.3244619369506836,
+      "learning_rate": 1e-06,
+      "loss": 0.032,
+      "step": 1262
+    },
+    {
+      "epoch": 0.21501532175689478,
+      "grad_norm": 1.7973686456680298,
+      "learning_rate": 1e-06,
+      "loss": 0.0458,
+      "step": 1263
+    },
+    {
+      "epoch": 0.21518556350017023,
+      "grad_norm": 1.7030102014541626,
+      "learning_rate": 1e-06,
+      "loss": 0.041,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2153558052434457,
+      "grad_norm": 1.5380405187606812,
+      "learning_rate": 1e-06,
+      "loss": 0.0382,
+      "step": 1265
+    },
+    {
+      "epoch": 0.21552604698672115,
+      "grad_norm": 2.168999433517456,
+      "learning_rate": 1e-06,
+      "loss": 0.0585,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2156962887299966,
+      "grad_norm": 2.1720099449157715,
+      "learning_rate": 1e-06,
+      "loss": 0.0363,
+      "step": 1267
+    },
+    {
+      "epoch": 0.21586653047327203,
+      "grad_norm": 2.3076796531677246,
+      "learning_rate": 1e-06,
+      "loss": 0.0679,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2160367722165475,
+      "grad_norm": 1.696959376335144,
+      "learning_rate": 1e-06,
+      "loss": 0.0331,
+      "step": 1269
+    },
+    {
+      "epoch": 0.21620701395982295,
+      "grad_norm": 1.8012551069259644,
+      "learning_rate": 1e-06,
+      "loss": 0.0399,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2163772557030984,
+      "grad_norm": 1.706807017326355,
+      "learning_rate": 1e-06,
+      "loss": 0.044,
+      "step": 1271
+    },
+    {
+      "epoch": 0.21654749744637386,
+      "grad_norm": 1.3523311614990234,
+      "learning_rate": 1e-06,
+      "loss": 0.04,
+      "step": 1272
+    },
+    {
+      "epoch": 0.2167177391896493,
+      "grad_norm": 1.5000497102737427,
+      "learning_rate": 1e-06,
+      "loss": 0.0383,
+      "step": 1273
+    },
+    {
+      "epoch": 0.21688798093292475,
+      "grad_norm": 1.4352927207946777,
+      "learning_rate": 1e-06,
+      "loss": 0.0327,
+      "step": 1274
+    },
+    {
+      "epoch": 0.2170582226762002,
+      "grad_norm": 1.4274100065231323,
+      "learning_rate": 1e-06,
+      "loss": 0.0302,
+      "step": 1275
+    },
+    {
+      "epoch": 0.21722846441947566,
+      "grad_norm": 1.4734169244766235,
+      "learning_rate": 1e-06,
+      "loss": 0.0291,
+      "step": 1276
+    },
+    {
+      "epoch": 0.21739870616275112,
+      "grad_norm": 1.532433271408081,
+      "learning_rate": 1e-06,
+      "loss": 0.0424,
+      "step": 1277
+    },
+    {
+      "epoch": 0.21756894790602654,
+      "grad_norm": 1.4756892919540405,
+      "learning_rate": 1e-06,
+      "loss": 0.0291,
+      "step": 1278
+    },
+    {
+      "epoch": 0.217739189649302,
+      "grad_norm": 1.4791312217712402,
+      "learning_rate": 1e-06,
+      "loss": 0.0428,
+      "step": 1279
+    },
+    {
+      "epoch": 0.21790943139257746,
+      "grad_norm": 1.201405644416809,
+      "learning_rate": 1e-06,
+      "loss": 0.0272,
+      "step": 1280
+    },
+    {
+      "epoch": 0.21807967313585291,
+      "grad_norm": 3.065424680709839,
+      "learning_rate": 1e-06,
+      "loss": 0.0406,
+      "step": 1281
+    },
+    {
+      "epoch": 0.21824991487912837,
+      "grad_norm": 1.7154473066329956,
+      "learning_rate": 1e-06,
+      "loss": 0.0307,
+      "step": 1282
+    },
+    {
+      "epoch": 0.21842015662240383,
+      "grad_norm": 1.5745826959609985,
+      "learning_rate": 1e-06,
+      "loss": 0.0295,
+      "step": 1283
+    },
+    {
+      "epoch": 0.21859039836567926,
+      "grad_norm": 1.9603149890899658,
+      "learning_rate": 1e-06,
+      "loss": 0.0469,
+      "step": 1284
+    },
+    {
+      "epoch": 0.2187606401089547,
+      "grad_norm": 1.6602104902267456,
+      "learning_rate": 1e-06,
+      "loss": 0.0433,
+      "step": 1285
+    },
+    {
+      "epoch": 0.21893088185223017,
+      "grad_norm": 1.7477741241455078,
+      "learning_rate": 1e-06,
+      "loss": 0.0291,
+      "step": 1286
+    },
+    {
+      "epoch": 0.21910112359550563,
+      "grad_norm": 1.8926868438720703,
+      "learning_rate": 1e-06,
+      "loss": 0.0392,
+      "step": 1287
+    },
+    {
+      "epoch": 0.21927136533878108,
+      "grad_norm": 1.508943796157837,
+      "learning_rate": 1e-06,
+      "loss": 0.0286,
+      "step": 1288
+    },
+    {
+      "epoch": 0.2194416070820565,
+      "grad_norm": 1.6563001871109009,
+      "learning_rate": 1e-06,
+      "loss": 0.0359,
+      "step": 1289
+    },
+    {
+      "epoch": 0.21961184882533197,
+      "grad_norm": 1.6094985008239746,
+      "learning_rate": 1e-06,
+      "loss": 0.0357,
+      "step": 1290
+    },
+    {
+      "epoch": 0.21978209056860742,
+      "grad_norm": 1.9460690021514893,
+      "learning_rate": 1e-06,
+      "loss": 0.0515,
+      "step": 1291
+    },
+    {
+      "epoch": 0.21995233231188288,
+      "grad_norm": 1.3542574644088745,
+      "learning_rate": 1e-06,
+      "loss": 0.0389,
+      "step": 1292
+    },
+    {
+      "epoch": 0.22012257405515834,
+      "grad_norm": 1.3899677991867065,
+      "learning_rate": 1e-06,
+      "loss": 0.0318,
+      "step": 1293
+    },
+    {
+      "epoch": 0.22029281579843377,
+      "grad_norm": 1.9105228185653687,
+      "learning_rate": 1e-06,
+      "loss": 0.049,
+      "step": 1294
+    },
+    {
+      "epoch": 0.22046305754170922,
+      "grad_norm": 1.5152599811553955,
+      "learning_rate": 1e-06,
+      "loss": 0.0416,
+      "step": 1295
+    },
+    {
+      "epoch": 0.22063329928498468,
+      "grad_norm": 1.2295457124710083,
+      "learning_rate": 1e-06,
+      "loss": 0.031,
+      "step": 1296
+    },
+    {
+      "epoch": 0.22080354102826014,
+      "grad_norm": 1.8313056230545044,
+      "learning_rate": 1e-06,
+      "loss": 0.0451,
+      "step": 1297
+    },
+    {
+      "epoch": 0.2209737827715356,
+      "grad_norm": 1.5762004852294922,
+      "learning_rate": 1e-06,
+      "loss": 0.0437,
+      "step": 1298
+    },
+    {
+      "epoch": 0.22114402451481102,
+      "grad_norm": 1.603684663772583,
+      "learning_rate": 1e-06,
+      "loss": 0.0372,
+      "step": 1299
+    },
+    {
+      "epoch": 0.22131426625808648,
+      "grad_norm": 1.809328317642212,
+      "learning_rate": 1e-06,
+      "loss": 0.0328,
+      "step": 1300
+    },
+    {
+      "epoch": 0.22148450800136193,
+      "grad_norm": 1.6629222631454468,
+      "learning_rate": 1e-06,
+      "loss": 0.0421,
+      "step": 1301
+    },
+    {
+      "epoch": 0.2216547497446374,
+      "grad_norm": 1.5856209993362427,
+      "learning_rate": 1e-06,
+      "loss": 0.0377,
+      "step": 1302
+    },
+    {
+      "epoch": 0.22182499148791285,
+      "grad_norm": 1.8281983137130737,
+      "learning_rate": 1e-06,
+      "loss": 0.0339,
+      "step": 1303
+    },
+    {
+      "epoch": 0.22199523323118828,
+      "grad_norm": 1.8064937591552734,
+      "learning_rate": 1e-06,
+      "loss": 0.0359,
+      "step": 1304
+    },
+    {
+      "epoch": 0.22216547497446373,
+      "grad_norm": 1.6382505893707275,
+      "learning_rate": 1e-06,
+      "loss": 0.0376,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2223357167177392,
+      "grad_norm": 1.8759084939956665,
+      "learning_rate": 1e-06,
+      "loss": 0.0397,
+      "step": 1306
+    },
+    {
+      "epoch": 0.22250595846101465,
+      "grad_norm": 1.587537407875061,
+      "learning_rate": 1e-06,
+      "loss": 0.0361,
+      "step": 1307
+    },
+    {
+      "epoch": 0.2226762002042901,
+      "grad_norm": 1.6119834184646606,
+      "learning_rate": 1e-06,
+      "loss": 0.0368,
+      "step": 1308
+    },
+    {
+      "epoch": 0.22284644194756553,
+      "grad_norm": 1.4815311431884766,
+      "learning_rate": 1e-06,
+      "loss": 0.0318,
+      "step": 1309
+    },
+    {
+      "epoch": 0.223016683690841,
+      "grad_norm": 1.4200611114501953,
+      "learning_rate": 1e-06,
+      "loss": 0.0282,
+      "step": 1310
+    },
+    {
+      "epoch": 0.22318692543411645,
+      "grad_norm": 1.3545717000961304,
+      "learning_rate": 1e-06,
+      "loss": 0.0374,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2233571671773919,
+      "grad_norm": 1.5961992740631104,
+      "learning_rate": 1e-06,
+      "loss": 0.0402,
+      "step": 1312
+    },
+    {
+      "epoch": 0.22352740892066736,
+      "grad_norm": 1.420921802520752,
+      "learning_rate": 1e-06,
+      "loss": 0.0443,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2236976506639428,
+      "grad_norm": 1.7957236766815186,
+      "learning_rate": 1e-06,
+      "loss": 0.0354,
+      "step": 1314
+    },
+    {
+      "epoch": 0.22386789240721824,
+      "grad_norm": 1.4843519926071167,
+      "learning_rate": 1e-06,
+      "loss": 0.0337,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2240381341504937,
+      "grad_norm": 1.5852254629135132,
+      "learning_rate": 1e-06,
+      "loss": 0.0366,
+      "step": 1316
+    },
+    {
+      "epoch": 0.22420837589376916,
+      "grad_norm": 1.654274344444275,
+      "learning_rate": 1e-06,
+      "loss": 0.0407,
+      "step": 1317
+    },
+    {
+      "epoch": 0.22437861763704461,
+      "grad_norm": 1.884421467781067,
+      "learning_rate": 1e-06,
+      "loss": 0.0379,
+      "step": 1318
+    },
+    {
+      "epoch": 0.22454885938032004,
+      "grad_norm": 1.8321963548660278,
+      "learning_rate": 1e-06,
+      "loss": 0.0404,
+      "step": 1319
+    },
+    {
+      "epoch": 0.2247191011235955,
+      "grad_norm": 1.735929012298584,
+      "learning_rate": 1e-06,
+      "loss": 0.0477,
+      "step": 1320
+    },
+    {
+      "epoch": 0.22488934286687096,
+      "grad_norm": 1.5793464183807373,
+      "learning_rate": 1e-06,
+      "loss": 0.0268,
+      "step": 1321
+    },
+    {
+      "epoch": 0.2250595846101464,
+      "grad_norm": 1.6962897777557373,
+      "learning_rate": 1e-06,
+      "loss": 0.0387,
+      "step": 1322
+    },
+    {
+      "epoch": 0.22522982635342187,
+      "grad_norm": 1.4622056484222412,
+      "learning_rate": 1e-06,
+      "loss": 0.0315,
+      "step": 1323
+    },
+    {
+      "epoch": 0.2254000680966973,
+      "grad_norm": 1.4477473497390747,
+      "learning_rate": 1e-06,
+      "loss": 0.0419,
+      "step": 1324
+    },
+    {
+      "epoch": 0.22557030983997275,
+      "grad_norm": 1.508447289466858,
+      "learning_rate": 1e-06,
+      "loss": 0.0317,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2257405515832482,
+      "grad_norm": 1.3146283626556396,
+      "learning_rate": 1e-06,
+      "loss": 0.0271,
+      "step": 1326
+    },
+    {
+      "epoch": 0.22591079332652367,
+      "grad_norm": 1.3582558631896973,
+      "learning_rate": 1e-06,
+      "loss": 0.0304,
+      "step": 1327
+    },
+    {
+      "epoch": 0.22608103506979912,
+      "grad_norm": 2.6392621994018555,
+      "learning_rate": 1e-06,
+      "loss": 0.0521,
+      "step": 1328
+    },
+    {
+      "epoch": 0.22625127681307455,
+      "grad_norm": 1.992223858833313,
+      "learning_rate": 1e-06,
+      "loss": 0.0608,
+      "step": 1329
+    },
+    {
+      "epoch": 0.22642151855635,
+      "grad_norm": 1.6414374113082886,
+      "learning_rate": 1e-06,
+      "loss": 0.0394,
+      "step": 1330
+    },
+    {
+      "epoch": 0.22659176029962547,
+      "grad_norm": 1.3846744298934937,
+      "learning_rate": 1e-06,
+      "loss": 0.0291,
+      "step": 1331
+    },
+    {
+      "epoch": 0.22676200204290092,
+      "grad_norm": 1.5988634824752808,
+      "learning_rate": 1e-06,
+      "loss": 0.0307,
+      "step": 1332
+    },
+    {
+      "epoch": 0.22693224378617638,
+      "grad_norm": 1.4296313524246216,
+      "learning_rate": 1e-06,
+      "loss": 0.0322,
+      "step": 1333
+    },
+    {
+      "epoch": 0.2271024855294518,
+      "grad_norm": 1.8813896179199219,
+      "learning_rate": 1e-06,
+      "loss": 0.0487,
+      "step": 1334
+    },
+    {
+      "epoch": 0.22727272727272727,
+      "grad_norm": 1.5750256776809692,
+      "learning_rate": 1e-06,
+      "loss": 0.0335,
+      "step": 1335
+    },
+    {
+      "epoch": 0.22744296901600272,
+      "grad_norm": 1.7783371210098267,
+      "learning_rate": 1e-06,
+      "loss": 0.0428,
+      "step": 1336
+    },
+    {
+      "epoch": 0.22761321075927818,
+      "grad_norm": 1.2634700536727905,
+      "learning_rate": 1e-06,
+      "loss": 0.0248,
+      "step": 1337
+    },
+    {
+      "epoch": 0.22778345250255363,
+      "grad_norm": 1.2207298278808594,
+      "learning_rate": 1e-06,
+      "loss": 0.0283,
+      "step": 1338
+    },
+    {
+      "epoch": 0.22795369424582906,
+      "grad_norm": 1.94607412815094,
+      "learning_rate": 1e-06,
+      "loss": 0.0345,
+      "step": 1339
+    },
+    {
+      "epoch": 0.22812393598910452,
+      "grad_norm": 3.4443957805633545,
+      "learning_rate": 1e-06,
+      "loss": 0.0546,
+      "step": 1340
+    },
+    {
+      "epoch": 0.22829417773237998,
+      "grad_norm": 2.1095688343048096,
+      "learning_rate": 1e-06,
+      "loss": 0.0488,
+      "step": 1341
+    },
+    {
+      "epoch": 0.22846441947565543,
+      "grad_norm": 5.005306720733643,
+      "learning_rate": 1e-06,
+      "loss": 0.079,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2286346612189309,
+      "grad_norm": 1.6204818487167358,
+      "learning_rate": 1e-06,
+      "loss": 0.034,
+      "step": 1343
+    },
+    {
+      "epoch": 0.22880490296220635,
+      "grad_norm": 1.8814785480499268,
+      "learning_rate": 1e-06,
+      "loss": 0.0365,
+      "step": 1344
+    },
+    {
+      "epoch": 0.22897514470548178,
+      "grad_norm": 1.5905134677886963,
+      "learning_rate": 1e-06,
+      "loss": 0.0328,
+      "step": 1345
+    },
+    {
+      "epoch": 0.22914538644875723,
+      "grad_norm": 1.4241366386413574,
+      "learning_rate": 1e-06,
+      "loss": 0.0307,
+      "step": 1346
+    },
+    {
+      "epoch": 0.2293156281920327,
+      "grad_norm": 1.5406595468521118,
+      "learning_rate": 1e-06,
+      "loss": 0.032,
+      "step": 1347
+    },
+    {
+      "epoch": 0.22948586993530815,
+      "grad_norm": 1.3714312314987183,
+      "learning_rate": 1e-06,
+      "loss": 0.0271,
+      "step": 1348
+    },
+    {
+      "epoch": 0.2296561116785836,
+      "grad_norm": 1.919778823852539,
+      "learning_rate": 1e-06,
+      "loss": 0.0372,
+      "step": 1349
+    },
+    {
+      "epoch": 0.22982635342185903,
+      "grad_norm": 1.4447976350784302,
+      "learning_rate": 1e-06,
+      "loss": 0.0289,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2299965951651345,
+      "grad_norm": 2.128904104232788,
+      "learning_rate": 1e-06,
+      "loss": 0.0554,
+      "step": 1351
+    },
+    {
+      "epoch": 0.23016683690840994,
+      "grad_norm": 1.2440881729125977,
+      "learning_rate": 1e-06,
+      "loss": 0.0365,
+      "step": 1352
+    },
+    {
+      "epoch": 0.2303370786516854,
+      "grad_norm": 2.688145637512207,
+      "learning_rate": 1e-06,
+      "loss": 0.0558,
+      "step": 1353
+    },
+    {
+      "epoch": 0.23050732039496086,
+      "grad_norm": 1.7107890844345093,
+      "learning_rate": 1e-06,
+      "loss": 0.0364,
+      "step": 1354
+    },
+    {
+      "epoch": 0.2306775621382363,
+      "grad_norm": 1.3972139358520508,
+      "learning_rate": 1e-06,
+      "loss": 0.0306,
+      "step": 1355
+    },
+    {
+      "epoch": 0.23084780388151174,
+      "grad_norm": 1.6243966817855835,
+      "learning_rate": 1e-06,
+      "loss": 0.0385,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2310180456247872,
+      "grad_norm": 1.5047791004180908,
+      "learning_rate": 1e-06,
+      "loss": 0.0344,
+      "step": 1357
+    },
+    {
+      "epoch": 0.23118828736806266,
+      "grad_norm": 1.5549938678741455,
+      "learning_rate": 1e-06,
+      "loss": 0.0321,
+      "step": 1358
+    },
+    {
+      "epoch": 0.2313585291113381,
+      "grad_norm": 1.5315052270889282,
+      "learning_rate": 1e-06,
+      "loss": 0.0272,
+      "step": 1359
+    },
+    {
+      "epoch": 0.23152877085461354,
+      "grad_norm": 1.6653809547424316,
+      "learning_rate": 1e-06,
+      "loss": 0.0239,
+      "step": 1360
+    },
+    {
+      "epoch": 0.231699012597889,
+      "grad_norm": 1.8215115070343018,
+      "learning_rate": 1e-06,
+      "loss": 0.034,
+      "step": 1361
+    },
+    {
+      "epoch": 0.23186925434116445,
+      "grad_norm": 1.3523900508880615,
+      "learning_rate": 1e-06,
+      "loss": 0.0282,
+      "step": 1362
+    },
+    {
+      "epoch": 0.2320394960844399,
+      "grad_norm": 1.6606805324554443,
+      "learning_rate": 1e-06,
+      "loss": 0.0462,
+      "step": 1363
+    },
+    {
+      "epoch": 0.23220973782771537,
+      "grad_norm": 1.6801470518112183,
+      "learning_rate": 1e-06,
+      "loss": 0.0428,
+      "step": 1364
+    },
+    {
+      "epoch": 0.2323799795709908,
+      "grad_norm": 1.6056513786315918,
+      "learning_rate": 1e-06,
+      "loss": 0.037,
+      "step": 1365
+    },
+    {
+      "epoch": 0.23255022131426625,
+      "grad_norm": 1.549023151397705,
+      "learning_rate": 1e-06,
+      "loss": 0.0362,
+      "step": 1366
+    },
+    {
+      "epoch": 0.2327204630575417,
+      "grad_norm": 1.8244285583496094,
+      "learning_rate": 1e-06,
+      "loss": 0.0469,
+      "step": 1367
+    },
+    {
+      "epoch": 0.23289070480081717,
+      "grad_norm": 1.7030590772628784,
+      "learning_rate": 1e-06,
+      "loss": 0.0396,
+      "step": 1368
+    },
+    {
+      "epoch": 0.23306094654409262,
+      "grad_norm": 1.9018099308013916,
+      "learning_rate": 1e-06,
+      "loss": 0.0462,
+      "step": 1369
+    },
+    {
+      "epoch": 0.23323118828736805,
+      "grad_norm": 1.8366174697875977,
+      "learning_rate": 1e-06,
+      "loss": 0.0431,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2334014300306435,
+      "grad_norm": 1.7807953357696533,
+      "learning_rate": 1e-06,
+      "loss": 0.0397,
+      "step": 1371
+    },
+    {
+      "epoch": 0.23357167177391897,
+      "grad_norm": 3.430758476257324,
+      "learning_rate": 1e-06,
+      "loss": 0.0504,
+      "step": 1372
+    },
+    {
+      "epoch": 0.23374191351719442,
+      "grad_norm": 1.6039738655090332,
+      "learning_rate": 1e-06,
+      "loss": 0.0326,
+      "step": 1373
+    },
+    {
+      "epoch": 0.23391215526046988,
+      "grad_norm": 2.001668691635132,
+      "learning_rate": 1e-06,
+      "loss": 0.0322,
+      "step": 1374
+    },
+    {
+      "epoch": 0.2340823970037453,
+      "grad_norm": 1.9030237197875977,
+      "learning_rate": 1e-06,
+      "loss": 0.0322,
+      "step": 1375
+    },
+    {
+      "epoch": 0.23425263874702076,
+      "grad_norm": 1.4003009796142578,
+      "learning_rate": 1e-06,
+      "loss": 0.0317,
+      "step": 1376
+    },
+    {
+      "epoch": 0.23442288049029622,
+      "grad_norm": 1.6500396728515625,
+      "learning_rate": 1e-06,
+      "loss": 0.0332,
+      "step": 1377
+    },
+    {
+      "epoch": 0.23459312223357168,
+      "grad_norm": 4.6141438484191895,
+      "learning_rate": 1e-06,
+      "loss": 0.0547,
+      "step": 1378
+    },
+    {
+      "epoch": 0.23476336397684713,
+      "grad_norm": 1.4108147621154785,
+      "learning_rate": 1e-06,
+      "loss": 0.0233,
+      "step": 1379
+    },
+    {
+      "epoch": 0.23493360572012256,
+      "grad_norm": 1.68300461769104,
+      "learning_rate": 1e-06,
+      "loss": 0.0265,
+      "step": 1380
+    },
+    {
+      "epoch": 0.23510384746339802,
+      "grad_norm": 1.4702105522155762,
+      "learning_rate": 1e-06,
+      "loss": 0.0311,
+      "step": 1381
+    },
+    {
+      "epoch": 0.23527408920667348,
+      "grad_norm": 1.669421911239624,
+      "learning_rate": 1e-06,
+      "loss": 0.0459,
+      "step": 1382
+    },
+    {
+      "epoch": 0.23544433094994893,
+      "grad_norm": 1.987108588218689,
+      "learning_rate": 1e-06,
+      "loss": 0.0317,
+      "step": 1383
+    },
+    {
+      "epoch": 0.2356145726932244,
+      "grad_norm": 1.5817410945892334,
+      "learning_rate": 1e-06,
+      "loss": 0.0297,
+      "step": 1384
+    },
+    {
+      "epoch": 0.23578481443649982,
+      "grad_norm": 1.524600863456726,
+      "learning_rate": 1e-06,
+      "loss": 0.027,
+      "step": 1385
+    },
+    {
+      "epoch": 0.23595505617977527,
+      "grad_norm": 1.9512219429016113,
+      "learning_rate": 1e-06,
+      "loss": 0.0426,
+      "step": 1386
+    },
+    {
+      "epoch": 0.23612529792305073,
+      "grad_norm": 1.6807881593704224,
+      "learning_rate": 1e-06,
+      "loss": 0.0332,
+      "step": 1387
+    },
+    {
+      "epoch": 0.2362955396663262,
+      "grad_norm": 1.5827664136886597,
+      "learning_rate": 1e-06,
+      "loss": 0.0393,
+      "step": 1388
+    },
+    {
+      "epoch": 0.23646578140960164,
+      "grad_norm": 1.4349877834320068,
+      "learning_rate": 1e-06,
+      "loss": 0.0316,
+      "step": 1389
+    },
+    {
+      "epoch": 0.23663602315287707,
+      "grad_norm": 1.8205857276916504,
+      "learning_rate": 1e-06,
+      "loss": 0.0422,
+      "step": 1390
+    },
+    {
+      "epoch": 0.23680626489615253,
+      "grad_norm": 2.0506768226623535,
+      "learning_rate": 1e-06,
+      "loss": 0.0371,
+      "step": 1391
+    },
+    {
+      "epoch": 0.236976506639428,
+      "grad_norm": 3.028653860092163,
+      "learning_rate": 1e-06,
+      "loss": 0.043,
+      "step": 1392
+    },
+    {
+      "epoch": 0.23714674838270344,
+      "grad_norm": 1.7093989849090576,
+      "learning_rate": 1e-06,
+      "loss": 0.0338,
+      "step": 1393
+    },
+    {
+      "epoch": 0.2373169901259789,
+      "grad_norm": 1.727501630783081,
+      "learning_rate": 1e-06,
+      "loss": 0.0279,
+      "step": 1394
+    },
+    {
+      "epoch": 0.23748723186925433,
+      "grad_norm": 1.9987183809280396,
+      "learning_rate": 1e-06,
+      "loss": 0.0358,
+      "step": 1395
+    },
+    {
+      "epoch": 0.23765747361252978,
+      "grad_norm": 1.4499446153640747,
+      "learning_rate": 1e-06,
+      "loss": 0.0228,
+      "step": 1396
+    },
+    {
+      "epoch": 0.23782771535580524,
+      "grad_norm": 2.0619688034057617,
+      "learning_rate": 1e-06,
+      "loss": 0.0451,
+      "step": 1397
+    },
+    {
+      "epoch": 0.2379979570990807,
+      "grad_norm": 1.4072680473327637,
+      "learning_rate": 1e-06,
+      "loss": 0.0331,
+      "step": 1398
+    },
+    {
+      "epoch": 0.23816819884235615,
+      "grad_norm": 1.9472742080688477,
+      "learning_rate": 1e-06,
+      "loss": 0.0357,
+      "step": 1399
+    },
+    {
+      "epoch": 0.23833844058563158,
+      "grad_norm": 1.682454228401184,
+      "learning_rate": 1e-06,
+      "loss": 0.0349,
+      "step": 1400
+    },
+    {
+      "epoch": 0.23850868232890704,
+      "grad_norm": 1.4352442026138306,
+      "learning_rate": 1e-06,
+      "loss": 0.0222,
+      "step": 1401
+    },
+    {
+      "epoch": 0.2386789240721825,
+      "grad_norm": 1.4091311693191528,
+      "learning_rate": 1e-06,
+      "loss": 0.0328,
+      "step": 1402
+    },
+    {
+      "epoch": 0.23884916581545795,
+      "grad_norm": 1.3856940269470215,
+      "learning_rate": 1e-06,
+      "loss": 0.0245,
+      "step": 1403
+    },
+    {
+      "epoch": 0.2390194075587334,
+      "grad_norm": 3.6680867671966553,
+      "learning_rate": 1e-06,
+      "loss": 0.049,
+      "step": 1404
+    },
+    {
+      "epoch": 0.23918964930200884,
+      "grad_norm": 1.7768093347549438,
+      "learning_rate": 1e-06,
+      "loss": 0.0396,
+      "step": 1405
+    },
+    {
+      "epoch": 0.2393598910452843,
+      "grad_norm": 1.83253014087677,
+      "learning_rate": 1e-06,
+      "loss": 0.0303,
+      "step": 1406
+    },
+    {
+      "epoch": 0.23953013278855975,
+      "grad_norm": 3.3626549243927,
+      "learning_rate": 1e-06,
+      "loss": 0.0739,
+      "step": 1407
+    },
+    {
+      "epoch": 0.2397003745318352,
+      "grad_norm": 1.9892669916152954,
+      "learning_rate": 1e-06,
+      "loss": 0.0423,
+      "step": 1408
+    },
+    {
+      "epoch": 0.23987061627511067,
+      "grad_norm": 1.340533971786499,
+      "learning_rate": 1e-06,
+      "loss": 0.0245,
+      "step": 1409
+    },
+    {
+      "epoch": 0.24004085801838612,
+      "grad_norm": 1.6560157537460327,
+      "learning_rate": 1e-06,
+      "loss": 0.0401,
+      "step": 1410
+    },
+    {
+      "epoch": 0.24021109976166155,
+      "grad_norm": 1.6459312438964844,
+      "learning_rate": 1e-06,
+      "loss": 0.0447,
+      "step": 1411
+    },
+    {
+      "epoch": 0.240381341504937,
+      "grad_norm": 2.4004392623901367,
+      "learning_rate": 1e-06,
+      "loss": 0.0525,
+      "step": 1412
+    },
+    {
+      "epoch": 0.24055158324821246,
+      "grad_norm": 1.7633525133132935,
+      "learning_rate": 1e-06,
+      "loss": 0.0334,
+      "step": 1413
+    },
+    {
+      "epoch": 0.24072182499148792,
+      "grad_norm": 1.6249415874481201,
+      "learning_rate": 1e-06,
+      "loss": 0.039,
+      "step": 1414
+    },
+    {
+      "epoch": 0.24089206673476338,
+      "grad_norm": 1.817467212677002,
+      "learning_rate": 1e-06,
+      "loss": 0.0329,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2410623084780388,
+      "grad_norm": 1.592246174812317,
+      "learning_rate": 1e-06,
+      "loss": 0.0493,
+      "step": 1416
+    },
+    {
+      "epoch": 0.24123255022131426,
+      "grad_norm": 2.0346999168395996,
+      "learning_rate": 1e-06,
+      "loss": 0.0353,
+      "step": 1417
+    },
+    {
+      "epoch": 0.24140279196458972,
+      "grad_norm": 1.3118454217910767,
+      "learning_rate": 1e-06,
+      "loss": 0.0321,
+      "step": 1418
+    },
+    {
+      "epoch": 0.24157303370786518,
+      "grad_norm": 1.5990475416183472,
+      "learning_rate": 1e-06,
+      "loss": 0.0401,
+      "step": 1419
+    },
+    {
+      "epoch": 0.24174327545114063,
+      "grad_norm": 1.3188730478286743,
+      "learning_rate": 1e-06,
+      "loss": 0.0188,
+      "step": 1420
+    },
+    {
+      "epoch": 0.24191351719441606,
+      "grad_norm": 1.5186814069747925,
+      "learning_rate": 1e-06,
+      "loss": 0.0286,
+      "step": 1421
+    },
+    {
+      "epoch": 0.24208375893769152,
+      "grad_norm": 3.969571352005005,
+      "learning_rate": 1e-06,
+      "loss": 0.0573,
+      "step": 1422
+    },
+    {
+      "epoch": 0.24225400068096697,
+      "grad_norm": 1.6175631284713745,
+      "learning_rate": 1e-06,
+      "loss": 0.0308,
+      "step": 1423
+    },
+    {
+      "epoch": 0.24242424242424243,
+      "grad_norm": 1.3874751329421997,
+      "learning_rate": 1e-06,
+      "loss": 0.0298,
+      "step": 1424
+    },
+    {
+      "epoch": 0.2425944841675179,
+      "grad_norm": 2.3302395343780518,
+      "learning_rate": 1e-06,
+      "loss": 0.0483,
+      "step": 1425
+    },
+    {
+      "epoch": 0.24276472591079332,
+      "grad_norm": 2.0167653560638428,
+      "learning_rate": 1e-06,
+      "loss": 0.0356,
+      "step": 1426
+    },
+    {
+      "epoch": 0.24293496765406877,
+      "grad_norm": 1.5584750175476074,
+      "learning_rate": 1e-06,
+      "loss": 0.0331,
+      "step": 1427
+    },
+    {
+      "epoch": 0.24310520939734423,
+      "grad_norm": 1.7811205387115479,
+      "learning_rate": 1e-06,
+      "loss": 0.0296,
+      "step": 1428
+    },
+    {
+      "epoch": 0.2432754511406197,
+      "grad_norm": 1.6993045806884766,
+      "learning_rate": 1e-06,
+      "loss": 0.0341,
+      "step": 1429
+    },
+    {
+      "epoch": 0.24344569288389514,
+      "grad_norm": 1.6028584241867065,
+      "learning_rate": 1e-06,
+      "loss": 0.03,
+      "step": 1430
+    },
+    {
+      "epoch": 0.24361593462717057,
+      "grad_norm": 2.2071762084960938,
+      "learning_rate": 1e-06,
+      "loss": 0.0418,
+      "step": 1431
+    },
+    {
+      "epoch": 0.24378617637044603,
+      "grad_norm": 1.5499199628829956,
+      "learning_rate": 1e-06,
+      "loss": 0.037,
+      "step": 1432
+    },
+    {
+      "epoch": 0.24395641811372148,
+      "grad_norm": 1.59584641456604,
+      "learning_rate": 1e-06,
+      "loss": 0.0312,
+      "step": 1433
+    },
+    {
+      "epoch": 0.24412665985699694,
+      "grad_norm": 1.5070613622665405,
+      "learning_rate": 1e-06,
+      "loss": 0.0274,
+      "step": 1434
+    },
+    {
+      "epoch": 0.2442969016002724,
+      "grad_norm": 1.6811244487762451,
+      "learning_rate": 1e-06,
+      "loss": 0.0277,
+      "step": 1435
+    },
+    {
+      "epoch": 0.24446714334354783,
+      "grad_norm": 1.631415843963623,
+      "learning_rate": 1e-06,
+      "loss": 0.0363,
+      "step": 1436
+    },
+    {
+      "epoch": 0.24463738508682328,
+      "grad_norm": 1.6105327606201172,
+      "learning_rate": 1e-06,
+      "loss": 0.0279,
+      "step": 1437
+    },
+    {
+      "epoch": 0.24480762683009874,
+      "grad_norm": 1.791298508644104,
+      "learning_rate": 1e-06,
+      "loss": 0.0494,
+      "step": 1438
+    },
+    {
+      "epoch": 0.2449778685733742,
+      "grad_norm": 1.1787437200546265,
+      "learning_rate": 1e-06,
+      "loss": 0.0278,
+      "step": 1439
+    },
+    {
+      "epoch": 0.24514811031664965,
+      "grad_norm": 1.3481965065002441,
+      "learning_rate": 1e-06,
+      "loss": 0.031,
+      "step": 1440
+    },
+    {
+      "epoch": 0.24531835205992508,
+      "grad_norm": 1.495886206626892,
+      "learning_rate": 1e-06,
+      "loss": 0.0419,
+      "step": 1441
+    },
+    {
+      "epoch": 0.24548859380320054,
+      "grad_norm": 1.6901342868804932,
+      "learning_rate": 1e-06,
+      "loss": 0.0329,
+      "step": 1442
+    },
+    {
+      "epoch": 0.245658835546476,
+      "grad_norm": 1.443678379058838,
+      "learning_rate": 1e-06,
+      "loss": 0.0412,
+      "step": 1443
+    },
+    {
+      "epoch": 0.24582907728975145,
+      "grad_norm": 1.8046194314956665,
+      "learning_rate": 1e-06,
+      "loss": 0.0384,
+      "step": 1444
+    },
+    {
+      "epoch": 0.2459993190330269,
+      "grad_norm": 1.5077475309371948,
+      "learning_rate": 1e-06,
+      "loss": 0.0321,
+      "step": 1445
+    },
+    {
+      "epoch": 0.24616956077630234,
+      "grad_norm": 1.7492755651474,
+      "learning_rate": 1e-06,
+      "loss": 0.0361,
+      "step": 1446
+    },
+    {
+      "epoch": 0.2463398025195778,
+      "grad_norm": 1.5502599477767944,
+      "learning_rate": 1e-06,
+      "loss": 0.0273,
+      "step": 1447
+    },
+    {
+      "epoch": 0.24651004426285325,
+      "grad_norm": 1.9191864728927612,
+      "learning_rate": 1e-06,
+      "loss": 0.0369,
+      "step": 1448
+    },
+    {
+      "epoch": 0.2466802860061287,
+      "grad_norm": 1.9377306699752808,
+      "learning_rate": 1e-06,
+      "loss": 0.0401,
+      "step": 1449
+    },
+    {
+      "epoch": 0.24685052774940416,
+      "grad_norm": 2.082916259765625,
+      "learning_rate": 1e-06,
+      "loss": 0.0402,
+      "step": 1450
+    },
+    {
+      "epoch": 0.2470207694926796,
+      "grad_norm": 1.3470356464385986,
+      "learning_rate": 1e-06,
+      "loss": 0.0214,
+      "step": 1451
+    },
+    {
+      "epoch": 0.24719101123595505,
+      "grad_norm": 1.5550161600112915,
+      "learning_rate": 1e-06,
+      "loss": 0.0257,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2473612529792305,
+      "grad_norm": 1.9106121063232422,
+      "learning_rate": 1e-06,
+      "loss": 0.0399,
+      "step": 1453
+    },
+    {
+      "epoch": 0.24753149472250596,
+      "grad_norm": 1.2977889776229858,
+      "learning_rate": 1e-06,
+      "loss": 0.0238,
+      "step": 1454
+    },
+    {
+      "epoch": 0.24770173646578142,
+      "grad_norm": 1.8015820980072021,
+      "learning_rate": 1e-06,
+      "loss": 0.0337,
+      "step": 1455
+    },
+    {
+      "epoch": 0.24787197820905685,
+      "grad_norm": 1.515289545059204,
+      "learning_rate": 1e-06,
+      "loss": 0.0266,
+      "step": 1456
+    },
+    {
+      "epoch": 0.2480422199523323,
+      "grad_norm": 2.0699691772460938,
+      "learning_rate": 1e-06,
+      "loss": 0.0408,
+      "step": 1457
+    },
+    {
+      "epoch": 0.24821246169560776,
+      "grad_norm": 1.9157500267028809,
+      "learning_rate": 1e-06,
+      "loss": 0.0431,
+      "step": 1458
+    },
+    {
+      "epoch": 0.24838270343888322,
+      "grad_norm": 1.7912602424621582,
+      "learning_rate": 1e-06,
+      "loss": 0.0362,
+      "step": 1459
+    },
+    {
+      "epoch": 0.24855294518215867,
+      "grad_norm": 1.5445823669433594,
+      "learning_rate": 1e-06,
+      "loss": 0.0337,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2487231869254341,
+      "grad_norm": 1.5842821598052979,
+      "learning_rate": 1e-06,
+      "loss": 0.0283,
+      "step": 1461
+    },
+    {
+      "epoch": 0.24889342866870956,
+      "grad_norm": 2.117060899734497,
+      "learning_rate": 1e-06,
+      "loss": 0.033,
+      "step": 1462
+    },
+    {
+      "epoch": 0.24906367041198502,
+      "grad_norm": 1.1993916034698486,
+      "learning_rate": 1e-06,
+      "loss": 0.0265,
+      "step": 1463
+    },
+    {
+      "epoch": 0.24923391215526047,
+      "grad_norm": 1.9758274555206299,
+      "learning_rate": 1e-06,
+      "loss": 0.0448,
+      "step": 1464
+    },
+    {
+      "epoch": 0.24940415389853593,
+      "grad_norm": 1.6750165224075317,
+      "learning_rate": 1e-06,
+      "loss": 0.0371,
+      "step": 1465
+    },
+    {
+      "epoch": 0.24957439564181136,
+      "grad_norm": 1.9693377017974854,
+      "learning_rate": 1e-06,
+      "loss": 0.0462,
+      "step": 1466
+    },
+    {
+      "epoch": 0.24974463738508682,
+      "grad_norm": 1.2334027290344238,
+      "learning_rate": 1e-06,
+      "loss": 0.0242,
+      "step": 1467
+    },
+    {
+      "epoch": 0.24991487912836227,
+      "grad_norm": 1.2368199825286865,
+      "learning_rate": 1e-06,
+      "loss": 0.0276,
+      "step": 1468
+    },
+    {
+      "epoch": 0.2500851208716377,
+      "grad_norm": 1.7857489585876465,
+      "learning_rate": 1e-06,
+      "loss": 0.0408,
+      "step": 1469
+    },
+    {
+      "epoch": 0.25025536261491316,
+      "grad_norm": 1.4412367343902588,
+      "learning_rate": 1e-06,
+      "loss": 0.0289,
+      "step": 1470
+    },
+    {
+      "epoch": 0.2504256043581886,
+      "grad_norm": 1.598808765411377,
+      "learning_rate": 1e-06,
+      "loss": 0.0255,
+      "step": 1471
+    },
+    {
+      "epoch": 0.25059584610146407,
+      "grad_norm": 1.440071940422058,
+      "learning_rate": 1e-06,
+      "loss": 0.0343,
+      "step": 1472
+    },
+    {
+      "epoch": 0.2507660878447395,
+      "grad_norm": 1.5133793354034424,
+      "learning_rate": 1e-06,
+      "loss": 0.0295,
+      "step": 1473
+    },
+    {
+      "epoch": 0.250936329588015,
+      "grad_norm": 1.5876449346542358,
+      "learning_rate": 1e-06,
+      "loss": 0.0282,
+      "step": 1474
+    },
+    {
+      "epoch": 0.25110657133129044,
+      "grad_norm": 1.636086106300354,
+      "learning_rate": 1e-06,
+      "loss": 0.0178,
+      "step": 1475
+    },
+    {
+      "epoch": 0.2512768130745659,
+      "grad_norm": 1.7871562242507935,
+      "learning_rate": 1e-06,
+      "loss": 0.0362,
+      "step": 1476
+    },
+    {
+      "epoch": 0.25144705481784135,
+      "grad_norm": 2.1147830486297607,
+      "learning_rate": 1e-06,
+      "loss": 0.043,
+      "step": 1477
+    },
+    {
+      "epoch": 0.2516172965611168,
+      "grad_norm": 2.0674445629119873,
+      "learning_rate": 1e-06,
+      "loss": 0.0411,
+      "step": 1478
+    },
+    {
+      "epoch": 0.2517875383043922,
+      "grad_norm": 1.684304118156433,
+      "learning_rate": 1e-06,
+      "loss": 0.0261,
+      "step": 1479
+    },
+    {
+      "epoch": 0.25195778004766767,
+      "grad_norm": 1.8230916261672974,
+      "learning_rate": 1e-06,
+      "loss": 0.0328,
+      "step": 1480
+    },
+    {
+      "epoch": 0.2521280217909431,
+      "grad_norm": 1.3470852375030518,
+      "learning_rate": 1e-06,
+      "loss": 0.0276,
+      "step": 1481
+    },
+    {
+      "epoch": 0.2522982635342186,
+      "grad_norm": 1.640721321105957,
+      "learning_rate": 1e-06,
+      "loss": 0.037,
+      "step": 1482
+    },
+    {
+      "epoch": 0.25246850527749404,
+      "grad_norm": 1.4035189151763916,
+      "learning_rate": 1e-06,
+      "loss": 0.0286,
+      "step": 1483
+    },
+    {
+      "epoch": 0.2526387470207695,
+      "grad_norm": 1.6055939197540283,
+      "learning_rate": 1e-06,
+      "loss": 0.0269,
+      "step": 1484
+    },
+    {
+      "epoch": 0.25280898876404495,
+      "grad_norm": 1.5646830797195435,
+      "learning_rate": 1e-06,
+      "loss": 0.0357,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2529792305073204,
+      "grad_norm": 1.7002915143966675,
+      "learning_rate": 1e-06,
+      "loss": 0.0351,
+      "step": 1486
+    },
+    {
+      "epoch": 0.25314947225059586,
+      "grad_norm": 1.1237139701843262,
+      "learning_rate": 1e-06,
+      "loss": 0.0229,
+      "step": 1487
+    },
+    {
+      "epoch": 0.2533197139938713,
+      "grad_norm": 1.8360260725021362,
+      "learning_rate": 1e-06,
+      "loss": 0.0251,
+      "step": 1488
+    },
+    {
+      "epoch": 0.2534899557371467,
+      "grad_norm": 1.483035683631897,
+      "learning_rate": 1e-06,
+      "loss": 0.0387,
+      "step": 1489
+    },
+    {
+      "epoch": 0.2536601974804222,
+      "grad_norm": 1.544597864151001,
+      "learning_rate": 1e-06,
+      "loss": 0.0308,
+      "step": 1490
+    },
+    {
+      "epoch": 0.25383043922369763,
+      "grad_norm": 1.5810545682907104,
+      "learning_rate": 1e-06,
+      "loss": 0.0358,
+      "step": 1491
+    },
+    {
+      "epoch": 0.2540006809669731,
+      "grad_norm": 1.6289693117141724,
+      "learning_rate": 1e-06,
+      "loss": 0.032,
+      "step": 1492
+    },
+    {
+      "epoch": 0.25417092271024855,
+      "grad_norm": 1.4735667705535889,
+      "learning_rate": 1e-06,
+      "loss": 0.035,
+      "step": 1493
+    },
+    {
+      "epoch": 0.254341164453524,
+      "grad_norm": 1.8274134397506714,
+      "learning_rate": 1e-06,
+      "loss": 0.026,
+      "step": 1494
+    },
+    {
+      "epoch": 0.25451140619679946,
+      "grad_norm": 1.5539690256118774,
+      "learning_rate": 1e-06,
+      "loss": 0.0302,
+      "step": 1495
+    },
+    {
+      "epoch": 0.2546816479400749,
+      "grad_norm": 1.251301884651184,
+      "learning_rate": 1e-06,
+      "loss": 0.0325,
+      "step": 1496
+    },
+    {
+      "epoch": 0.2548518896833504,
+      "grad_norm": 2.4516842365264893,
+      "learning_rate": 1e-06,
+      "loss": 0.033,
+      "step": 1497
+    },
+    {
+      "epoch": 0.25502213142662583,
+      "grad_norm": 1.677709698677063,
+      "learning_rate": 1e-06,
+      "loss": 0.0331,
+      "step": 1498
+    },
+    {
+      "epoch": 0.25519237316990123,
+      "grad_norm": 1.771200180053711,
+      "learning_rate": 1e-06,
+      "loss": 0.0336,
+      "step": 1499
+    },
+    {
+      "epoch": 0.2553626149131767,
+      "grad_norm": 1.574262022972107,
+      "learning_rate": 1e-06,
+      "loss": 0.0306,
+      "step": 1500
+    },
+    {
+      "epoch": 0.25553285665645215,
+      "grad_norm": 1.3920859098434448,
+      "learning_rate": 1e-06,
+      "loss": 0.0261,
+      "step": 1501
+    },
+    {
+      "epoch": 0.2557030983997276,
+      "grad_norm": 2.0418004989624023,
+      "learning_rate": 1e-06,
+      "loss": 0.036,
+      "step": 1502
+    },
+    {
+      "epoch": 0.25587334014300306,
+      "grad_norm": 1.639630913734436,
+      "learning_rate": 1e-06,
+      "loss": 0.0239,
+      "step": 1503
+    },
+    {
+      "epoch": 0.2560435818862785,
+      "grad_norm": 3.079418659210205,
+      "learning_rate": 1e-06,
+      "loss": 0.0548,
+      "step": 1504
+    },
+    {
+      "epoch": 0.25621382362955397,
+      "grad_norm": 1.5250223875045776,
+      "learning_rate": 1e-06,
+      "loss": 0.025,
+      "step": 1505
+    },
+    {
+      "epoch": 0.25638406537282943,
+      "grad_norm": 1.6824452877044678,
+      "learning_rate": 1e-06,
+      "loss": 0.034,
+      "step": 1506
+    },
+    {
+      "epoch": 0.2565543071161049,
+      "grad_norm": 1.9331268072128296,
+      "learning_rate": 1e-06,
+      "loss": 0.0261,
+      "step": 1507
+    },
+    {
+      "epoch": 0.25672454885938034,
+      "grad_norm": 1.964712142944336,
+      "learning_rate": 1e-06,
+      "loss": 0.0385,
+      "step": 1508
+    },
+    {
+      "epoch": 0.2568947906026558,
+      "grad_norm": 1.5585107803344727,
+      "learning_rate": 1e-06,
+      "loss": 0.0267,
+      "step": 1509
+    },
+    {
+      "epoch": 0.2570650323459312,
+      "grad_norm": 1.9522138833999634,
+      "learning_rate": 1e-06,
+      "loss": 0.0481,
+      "step": 1510
+    },
+    {
+      "epoch": 0.25723527408920666,
+      "grad_norm": 2.522395372390747,
+      "learning_rate": 1e-06,
+      "loss": 0.0503,
+      "step": 1511
+    },
+    {
+      "epoch": 0.2574055158324821,
+      "grad_norm": 1.4549627304077148,
+      "learning_rate": 1e-06,
+      "loss": 0.0349,
+      "step": 1512
+    },
+    {
+      "epoch": 0.25757575757575757,
+      "grad_norm": 1.0665477514266968,
+      "learning_rate": 1e-06,
+      "loss": 0.0216,
+      "step": 1513
+    },
+    {
+      "epoch": 0.257745999319033,
+      "grad_norm": 1.1988400220870972,
+      "learning_rate": 1e-06,
+      "loss": 0.0261,
+      "step": 1514
+    },
+    {
+      "epoch": 0.2579162410623085,
+      "grad_norm": 1.5212563276290894,
+      "learning_rate": 1e-06,
+      "loss": 0.0319,
+      "step": 1515
+    },
+    {
+      "epoch": 0.25808648280558394,
+      "grad_norm": 1.6229534149169922,
+      "learning_rate": 1e-06,
+      "loss": 0.0387,
+      "step": 1516
+    },
+    {
+      "epoch": 0.2582567245488594,
+      "grad_norm": 1.9084266424179077,
+      "learning_rate": 1e-06,
+      "loss": 0.0354,
+      "step": 1517
+    },
+    {
+      "epoch": 0.25842696629213485,
+      "grad_norm": 1.2355334758758545,
+      "learning_rate": 1e-06,
+      "loss": 0.0266,
+      "step": 1518
+    },
+    {
+      "epoch": 0.2585972080354103,
+      "grad_norm": 1.6511914730072021,
+      "learning_rate": 1e-06,
+      "loss": 0.0238,
+      "step": 1519
+    },
+    {
+      "epoch": 0.2587674497786857,
+      "grad_norm": 1.7273814678192139,
+      "learning_rate": 1e-06,
+      "loss": 0.0369,
+      "step": 1520
+    },
+    {
+      "epoch": 0.25893769152196117,
+      "grad_norm": 1.715128779411316,
+      "learning_rate": 1e-06,
+      "loss": 0.0337,
+      "step": 1521
+    },
+    {
+      "epoch": 0.2591079332652366,
+      "grad_norm": 1.704223871231079,
+      "learning_rate": 1e-06,
+      "loss": 0.029,
+      "step": 1522
+    },
+    {
+      "epoch": 0.2592781750085121,
+      "grad_norm": 1.452068567276001,
+      "learning_rate": 1e-06,
+      "loss": 0.0258,
+      "step": 1523
+    },
+    {
+      "epoch": 0.25944841675178754,
+      "grad_norm": 2.0453882217407227,
+      "learning_rate": 1e-06,
+      "loss": 0.0342,
+      "step": 1524
+    },
+    {
+      "epoch": 0.259618658495063,
+      "grad_norm": 2.1579129695892334,
+      "learning_rate": 1e-06,
+      "loss": 0.0409,
+      "step": 1525
+    },
+    {
+      "epoch": 0.25978890023833845,
+      "grad_norm": 1.6939774751663208,
+      "learning_rate": 1e-06,
+      "loss": 0.0228,
+      "step": 1526
+    },
+    {
+      "epoch": 0.2599591419816139,
+      "grad_norm": 1.9167999029159546,
+      "learning_rate": 1e-06,
+      "loss": 0.0269,
+      "step": 1527
+    },
+    {
+      "epoch": 0.26012938372488936,
+      "grad_norm": 1.4578943252563477,
+      "learning_rate": 1e-06,
+      "loss": 0.0269,
+      "step": 1528
+    },
+    {
+      "epoch": 0.2602996254681648,
+      "grad_norm": 1.426521897315979,
+      "learning_rate": 1e-06,
+      "loss": 0.0268,
+      "step": 1529
+    },
+    {
+      "epoch": 0.2604698672114402,
+      "grad_norm": 1.7971490621566772,
+      "learning_rate": 1e-06,
+      "loss": 0.0337,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2606401089547157,
+      "grad_norm": 1.6107807159423828,
+      "learning_rate": 1e-06,
+      "loss": 0.0434,
+      "step": 1531
+    },
+    {
+      "epoch": 0.26081035069799113,
+      "grad_norm": 1.8623952865600586,
+      "learning_rate": 1e-06,
+      "loss": 0.0319,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2609805924412666,
+      "grad_norm": 1.3411558866500854,
+      "learning_rate": 1e-06,
+      "loss": 0.0244,
+      "step": 1533
+    },
+    {
+      "epoch": 0.26115083418454205,
+      "grad_norm": 1.5986648797988892,
+      "learning_rate": 1e-06,
+      "loss": 0.023,
+      "step": 1534
+    },
+    {
+      "epoch": 0.2613210759278175,
+      "grad_norm": 1.803185224533081,
+      "learning_rate": 1e-06,
+      "loss": 0.034,
+      "step": 1535
+    },
+    {
+      "epoch": 0.26149131767109296,
+      "grad_norm": 1.4397742748260498,
+      "learning_rate": 1e-06,
+      "loss": 0.0408,
+      "step": 1536
+    },
+    {
+      "epoch": 0.2616615594143684,
+      "grad_norm": 1.5375397205352783,
+      "learning_rate": 1e-06,
+      "loss": 0.0277,
+      "step": 1537
+    },
+    {
+      "epoch": 0.2618318011576439,
+      "grad_norm": 1.6096251010894775,
+      "learning_rate": 1e-06,
+      "loss": 0.0284,
+      "step": 1538
+    },
+    {
+      "epoch": 0.26200204290091933,
+      "grad_norm": 1.9255061149597168,
+      "learning_rate": 1e-06,
+      "loss": 0.0266,
+      "step": 1539
+    },
+    {
+      "epoch": 0.26217228464419473,
+      "grad_norm": 1.6099532842636108,
+      "learning_rate": 1e-06,
+      "loss": 0.0248,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2623425263874702,
+      "grad_norm": 2.104907989501953,
+      "learning_rate": 1e-06,
+      "loss": 0.0265,
+      "step": 1541
+    },
+    {
+      "epoch": 0.26251276813074564,
+      "grad_norm": 1.6234943866729736,
+      "learning_rate": 1e-06,
+      "loss": 0.0365,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2626830098740211,
+      "grad_norm": 1.7157036066055298,
+      "learning_rate": 1e-06,
+      "loss": 0.0321,
+      "step": 1543
+    },
+    {
+      "epoch": 0.26285325161729656,
+      "grad_norm": 1.219995141029358,
+      "learning_rate": 1e-06,
+      "loss": 0.0201,
+      "step": 1544
+    },
+    {
+      "epoch": 0.263023493360572,
+      "grad_norm": 1.5472625494003296,
+      "learning_rate": 1e-06,
+      "loss": 0.0327,
+      "step": 1545
+    },
+    {
+      "epoch": 0.26319373510384747,
+      "grad_norm": 1.6542850732803345,
+      "learning_rate": 1e-06,
+      "loss": 0.0285,
+      "step": 1546
+    },
+    {
+      "epoch": 0.2633639768471229,
+      "grad_norm": 1.4218086004257202,
+      "learning_rate": 1e-06,
+      "loss": 0.0267,
+      "step": 1547
+    },
+    {
+      "epoch": 0.2635342185903984,
+      "grad_norm": 1.3034826517105103,
+      "learning_rate": 1e-06,
+      "loss": 0.0257,
+      "step": 1548
+    },
+    {
+      "epoch": 0.26370446033367384,
+      "grad_norm": 2.216113805770874,
+      "learning_rate": 1e-06,
+      "loss": 0.0568,
+      "step": 1549
+    },
+    {
+      "epoch": 0.26387470207694924,
+      "grad_norm": 1.4958362579345703,
+      "learning_rate": 1e-06,
+      "loss": 0.0278,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2640449438202247,
+      "grad_norm": 1.4572584629058838,
+      "learning_rate": 1e-06,
+      "loss": 0.0256,
+      "step": 1551
+    },
+    {
+      "epoch": 0.26421518556350015,
+      "grad_norm": 1.7182279825210571,
+      "learning_rate": 1e-06,
+      "loss": 0.0406,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2643854273067756,
+      "grad_norm": 2.1565287113189697,
+      "learning_rate": 1e-06,
+      "loss": 0.0321,
+      "step": 1553
+    },
+    {
+      "epoch": 0.26455566905005107,
+      "grad_norm": 1.421190619468689,
+      "learning_rate": 1e-06,
+      "loss": 0.031,
+      "step": 1554
+    },
+    {
+      "epoch": 0.2647259107933265,
+      "grad_norm": 1.4493217468261719,
+      "learning_rate": 1e-06,
+      "loss": 0.0334,
+      "step": 1555
+    },
+    {
+      "epoch": 0.264896152536602,
+      "grad_norm": 1.5228818655014038,
+      "learning_rate": 1e-06,
+      "loss": 0.0305,
+      "step": 1556
+    },
+    {
+      "epoch": 0.26506639427987744,
+      "grad_norm": 1.5019917488098145,
+      "learning_rate": 1e-06,
+      "loss": 0.0254,
+      "step": 1557
+    },
+    {
+      "epoch": 0.2652366360231529,
+      "grad_norm": 1.8647435903549194,
+      "learning_rate": 1e-06,
+      "loss": 0.0295,
+      "step": 1558
+    },
+    {
+      "epoch": 0.26540687776642835,
+      "grad_norm": 1.5852400064468384,
+      "learning_rate": 1e-06,
+      "loss": 0.0235,
+      "step": 1559
+    },
+    {
+      "epoch": 0.26557711950970375,
+      "grad_norm": 1.5699732303619385,
+      "learning_rate": 1e-06,
+      "loss": 0.0369,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2657473612529792,
+      "grad_norm": 1.5314867496490479,
+      "learning_rate": 1e-06,
+      "loss": 0.0231,
+      "step": 1561
+    },
+    {
+      "epoch": 0.26591760299625467,
+      "grad_norm": 1.3057242631912231,
+      "learning_rate": 1e-06,
+      "loss": 0.0248,
+      "step": 1562
+    },
+    {
+      "epoch": 0.2660878447395301,
+      "grad_norm": 1.5217634439468384,
+      "learning_rate": 1e-06,
+      "loss": 0.0383,
+      "step": 1563
+    },
+    {
+      "epoch": 0.2662580864828056,
+      "grad_norm": 1.9693313837051392,
+      "learning_rate": 1e-06,
+      "loss": 0.0274,
+      "step": 1564
+    },
+    {
+      "epoch": 0.26642832822608103,
+      "grad_norm": 2.246293544769287,
+      "learning_rate": 1e-06,
+      "loss": 0.0421,
+      "step": 1565
+    },
+    {
+      "epoch": 0.2665985699693565,
+      "grad_norm": 1.439656376838684,
+      "learning_rate": 1e-06,
+      "loss": 0.0319,
+      "step": 1566
+    },
+    {
+      "epoch": 0.26676881171263195,
+      "grad_norm": 1.522289752960205,
+      "learning_rate": 1e-06,
+      "loss": 0.0309,
+      "step": 1567
+    },
+    {
+      "epoch": 0.2669390534559074,
+      "grad_norm": 1.6061466932296753,
+      "learning_rate": 1e-06,
+      "loss": 0.0264,
+      "step": 1568
+    },
+    {
+      "epoch": 0.26710929519918286,
+      "grad_norm": 1.4138331413269043,
+      "learning_rate": 1e-06,
+      "loss": 0.026,
+      "step": 1569
+    },
+    {
+      "epoch": 0.2672795369424583,
+      "grad_norm": 1.924535870552063,
+      "learning_rate": 1e-06,
+      "loss": 0.0253,
+      "step": 1570
+    },
+    {
+      "epoch": 0.2674497786857337,
+      "grad_norm": 1.5255926847457886,
+      "learning_rate": 1e-06,
+      "loss": 0.0285,
+      "step": 1571
+    },
+    {
+      "epoch": 0.2676200204290092,
+      "grad_norm": 1.752744197845459,
+      "learning_rate": 1e-06,
+      "loss": 0.0223,
+      "step": 1572
+    },
+    {
+      "epoch": 0.26779026217228463,
+      "grad_norm": 1.2975589036941528,
+      "learning_rate": 1e-06,
+      "loss": 0.029,
+      "step": 1573
+    },
+    {
+      "epoch": 0.2679605039155601,
+      "grad_norm": 1.4539504051208496,
+      "learning_rate": 1e-06,
+      "loss": 0.0303,
+      "step": 1574
+    },
+    {
+      "epoch": 0.26813074565883555,
+      "grad_norm": 1.4913675785064697,
+      "learning_rate": 1e-06,
+      "loss": 0.0337,
+      "step": 1575
+    },
+    {
+      "epoch": 0.268300987402111,
+      "grad_norm": 1.5786329507827759,
+      "learning_rate": 1e-06,
+      "loss": 0.0317,
+      "step": 1576
+    },
+    {
+      "epoch": 0.26847122914538646,
+      "grad_norm": 1.719687581062317,
+      "learning_rate": 1e-06,
+      "loss": 0.0419,
+      "step": 1577
+    },
+    {
+      "epoch": 0.2686414708886619,
+      "grad_norm": 1.5540446043014526,
+      "learning_rate": 1e-06,
+      "loss": 0.0295,
+      "step": 1578
+    },
+    {
+      "epoch": 0.26881171263193737,
+      "grad_norm": 1.3455661535263062,
+      "learning_rate": 1e-06,
+      "loss": 0.0216,
+      "step": 1579
+    },
+    {
+      "epoch": 0.26898195437521283,
+      "grad_norm": 1.514863133430481,
+      "learning_rate": 1e-06,
+      "loss": 0.0309,
+      "step": 1580
+    },
+    {
+      "epoch": 0.26915219611848823,
+      "grad_norm": 1.63846755027771,
+      "learning_rate": 1e-06,
+      "loss": 0.0317,
+      "step": 1581
+    },
+    {
+      "epoch": 0.2693224378617637,
+      "grad_norm": 1.4479107856750488,
+      "learning_rate": 1e-06,
+      "loss": 0.03,
+      "step": 1582
+    },
+    {
+      "epoch": 0.26949267960503914,
+      "grad_norm": 1.504742980003357,
+      "learning_rate": 1e-06,
+      "loss": 0.0204,
+      "step": 1583
+    },
+    {
+      "epoch": 0.2696629213483146,
+      "grad_norm": 1.895027995109558,
+      "learning_rate": 1e-06,
+      "loss": 0.0383,
+      "step": 1584
+    },
+    {
+      "epoch": 0.26983316309159006,
+      "grad_norm": 1.5659292936325073,
+      "learning_rate": 1e-06,
+      "loss": 0.0269,
+      "step": 1585
+    },
+    {
+      "epoch": 0.2700034048348655,
+      "grad_norm": 1.855080246925354,
+      "learning_rate": 1e-06,
+      "loss": 0.0341,
+      "step": 1586
+    },
+    {
+      "epoch": 0.27017364657814097,
+      "grad_norm": 2.0592076778411865,
+      "learning_rate": 1e-06,
+      "loss": 0.0367,
+      "step": 1587
+    },
+    {
+      "epoch": 0.2703438883214164,
+      "grad_norm": 1.626511812210083,
+      "learning_rate": 1e-06,
+      "loss": 0.0286,
+      "step": 1588
+    },
+    {
+      "epoch": 0.2705141300646919,
+      "grad_norm": 1.715602159500122,
+      "learning_rate": 1e-06,
+      "loss": 0.0267,
+      "step": 1589
+    },
+    {
+      "epoch": 0.27068437180796734,
+      "grad_norm": 1.7395516633987427,
+      "learning_rate": 1e-06,
+      "loss": 0.0306,
+      "step": 1590
+    },
+    {
+      "epoch": 0.27085461355124274,
+      "grad_norm": 2.016375780105591,
+      "learning_rate": 1e-06,
+      "loss": 0.0316,
+      "step": 1591
+    },
+    {
+      "epoch": 0.2710248552945182,
+      "grad_norm": 1.7519330978393555,
+      "learning_rate": 1e-06,
+      "loss": 0.0323,
+      "step": 1592
+    },
+    {
+      "epoch": 0.27119509703779365,
+      "grad_norm": 1.3459405899047852,
+      "learning_rate": 1e-06,
+      "loss": 0.025,
+      "step": 1593
+    },
+    {
+      "epoch": 0.2713653387810691,
+      "grad_norm": 1.7252540588378906,
+      "learning_rate": 1e-06,
+      "loss": 0.0313,
+      "step": 1594
+    },
+    {
+      "epoch": 0.27153558052434457,
+      "grad_norm": 1.3951072692871094,
+      "learning_rate": 1e-06,
+      "loss": 0.0341,
+      "step": 1595
+    },
+    {
+      "epoch": 0.27170582226762,
+      "grad_norm": 1.615648627281189,
+      "learning_rate": 1e-06,
+      "loss": 0.0302,
+      "step": 1596
+    },
+    {
+      "epoch": 0.2718760640108955,
+      "grad_norm": 1.938193678855896,
+      "learning_rate": 1e-06,
+      "loss": 0.0267,
+      "step": 1597
+    },
+    {
+      "epoch": 0.27204630575417094,
+      "grad_norm": 1.4162940979003906,
+      "learning_rate": 1e-06,
+      "loss": 0.0267,
+      "step": 1598
+    },
+    {
+      "epoch": 0.2722165474974464,
+      "grad_norm": 1.7857416868209839,
+      "learning_rate": 1e-06,
+      "loss": 0.0287,
+      "step": 1599
+    },
+    {
+      "epoch": 0.27238678924072185,
+      "grad_norm": 2.2122995853424072,
+      "learning_rate": 1e-06,
+      "loss": 0.0339,
+      "step": 1600
+    },
+    {
+      "epoch": 0.27255703098399725,
+      "grad_norm": 1.2226827144622803,
+      "learning_rate": 1e-06,
+      "loss": 0.0224,
+      "step": 1601
+    },
+    {
+      "epoch": 0.2727272727272727,
+      "grad_norm": 1.366214394569397,
+      "learning_rate": 1e-06,
+      "loss": 0.0268,
+      "step": 1602
+    },
+    {
+      "epoch": 0.27289751447054816,
+      "grad_norm": 1.472756266593933,
+      "learning_rate": 1e-06,
+      "loss": 0.0251,
+      "step": 1603
+    },
+    {
+      "epoch": 0.2730677562138236,
+      "grad_norm": 1.4528722763061523,
+      "learning_rate": 1e-06,
+      "loss": 0.0245,
+      "step": 1604
+    },
+    {
+      "epoch": 0.2732379979570991,
+      "grad_norm": 1.2808316946029663,
+      "learning_rate": 1e-06,
+      "loss": 0.0231,
+      "step": 1605
+    },
+    {
+      "epoch": 0.27340823970037453,
+      "grad_norm": 1.5016905069351196,
+      "learning_rate": 1e-06,
+      "loss": 0.0338,
+      "step": 1606
+    },
+    {
+      "epoch": 0.27357848144365,
+      "grad_norm": 1.743894100189209,
+      "learning_rate": 1e-06,
+      "loss": 0.029,
+      "step": 1607
+    },
+    {
+      "epoch": 0.27374872318692545,
+      "grad_norm": 1.6557116508483887,
+      "learning_rate": 1e-06,
+      "loss": 0.0288,
+      "step": 1608
+    },
+    {
+      "epoch": 0.2739189649302009,
+      "grad_norm": 1.2677494287490845,
+      "learning_rate": 1e-06,
+      "loss": 0.0229,
+      "step": 1609
+    },
+    {
+      "epoch": 0.27408920667347636,
+      "grad_norm": 1.6573220491409302,
+      "learning_rate": 1e-06,
+      "loss": 0.0307,
+      "step": 1610
+    },
+    {
+      "epoch": 0.27425944841675176,
+      "grad_norm": 1.8028841018676758,
+      "learning_rate": 1e-06,
+      "loss": 0.0289,
+      "step": 1611
+    },
+    {
+      "epoch": 0.2744296901600272,
+      "grad_norm": 1.5891400575637817,
+      "learning_rate": 1e-06,
+      "loss": 0.0344,
+      "step": 1612
+    },
+    {
+      "epoch": 0.2745999319033027,
+      "grad_norm": 1.492483377456665,
+      "learning_rate": 1e-06,
+      "loss": 0.0214,
+      "step": 1613
+    },
+    {
+      "epoch": 0.27477017364657813,
+      "grad_norm": 1.5204499959945679,
+      "learning_rate": 1e-06,
+      "loss": 0.0257,
+      "step": 1614
+    },
+    {
+      "epoch": 0.2749404153898536,
+      "grad_norm": 2.0463969707489014,
+      "learning_rate": 1e-06,
+      "loss": 0.0287,
+      "step": 1615
+    },
+    {
+      "epoch": 0.27511065713312904,
+      "grad_norm": 1.5479378700256348,
+      "learning_rate": 1e-06,
+      "loss": 0.0291,
+      "step": 1616
+    },
+    {
+      "epoch": 0.2752808988764045,
+      "grad_norm": 2.649272918701172,
+      "learning_rate": 1e-06,
+      "loss": 0.0455,
+      "step": 1617
+    },
+    {
+      "epoch": 0.27545114061967996,
+      "grad_norm": 1.8164010047912598,
+      "learning_rate": 1e-06,
+      "loss": 0.0291,
+      "step": 1618
+    },
+    {
+      "epoch": 0.2756213823629554,
+      "grad_norm": 1.2867664098739624,
+      "learning_rate": 1e-06,
+      "loss": 0.0197,
+      "step": 1619
+    },
+    {
+      "epoch": 0.27579162410623087,
+      "grad_norm": 1.3096405267715454,
+      "learning_rate": 1e-06,
+      "loss": 0.0262,
+      "step": 1620
+    },
+    {
+      "epoch": 0.27596186584950627,
+      "grad_norm": 1.3673827648162842,
+      "learning_rate": 1e-06,
+      "loss": 0.0177,
+      "step": 1621
+    },
+    {
+      "epoch": 0.27613210759278173,
+      "grad_norm": 1.6427737474441528,
+      "learning_rate": 1e-06,
+      "loss": 0.0302,
+      "step": 1622
+    },
+    {
+      "epoch": 0.2763023493360572,
+      "grad_norm": 1.4813086986541748,
+      "learning_rate": 1e-06,
+      "loss": 0.0307,
+      "step": 1623
+    },
+    {
+      "epoch": 0.27647259107933264,
+      "grad_norm": 1.318463921546936,
+      "learning_rate": 1e-06,
+      "loss": 0.0221,
+      "step": 1624
+    },
+    {
+      "epoch": 0.2766428328226081,
+      "grad_norm": 1.6629798412322998,
+      "learning_rate": 1e-06,
+      "loss": 0.0289,
+      "step": 1625
+    },
+    {
+      "epoch": 0.27681307456588355,
+      "grad_norm": 1.4706048965454102,
+      "learning_rate": 1e-06,
+      "loss": 0.0241,
+      "step": 1626
+    },
+    {
+      "epoch": 0.276983316309159,
+      "grad_norm": 1.6562780141830444,
+      "learning_rate": 1e-06,
+      "loss": 0.027,
+      "step": 1627
+    },
+    {
+      "epoch": 0.27715355805243447,
+      "grad_norm": 1.7971525192260742,
+      "learning_rate": 1e-06,
+      "loss": 0.0322,
+      "step": 1628
+    },
+    {
+      "epoch": 0.2773237997957099,
+      "grad_norm": 1.9127774238586426,
+      "learning_rate": 1e-06,
+      "loss": 0.0382,
+      "step": 1629
+    },
+    {
+      "epoch": 0.2774940415389854,
+      "grad_norm": 1.661810278892517,
+      "learning_rate": 1e-06,
+      "loss": 0.0334,
+      "step": 1630
+    },
+    {
+      "epoch": 0.27766428328226084,
+      "grad_norm": 2.6759703159332275,
+      "learning_rate": 1e-06,
+      "loss": 0.0297,
+      "step": 1631
+    },
+    {
+      "epoch": 0.27783452502553624,
+      "grad_norm": 1.6530686616897583,
+      "learning_rate": 1e-06,
+      "loss": 0.0353,
+      "step": 1632
+    },
+    {
+      "epoch": 0.2780047667688117,
+      "grad_norm": 1.8694590330123901,
+      "learning_rate": 1e-06,
+      "loss": 0.0461,
+      "step": 1633
+    },
+    {
+      "epoch": 0.27817500851208715,
+      "grad_norm": 1.3104816675186157,
+      "learning_rate": 1e-06,
+      "loss": 0.0233,
+      "step": 1634
+    },
+    {
+      "epoch": 0.2783452502553626,
+      "grad_norm": 1.3213906288146973,
+      "learning_rate": 1e-06,
+      "loss": 0.0212,
+      "step": 1635
+    },
+    {
+      "epoch": 0.27851549199863807,
+      "grad_norm": 1.9300750494003296,
+      "learning_rate": 1e-06,
+      "loss": 0.0359,
+      "step": 1636
+    },
+    {
+      "epoch": 0.2786857337419135,
+      "grad_norm": 1.7006559371948242,
+      "learning_rate": 1e-06,
+      "loss": 0.0276,
+      "step": 1637
+    },
+    {
+      "epoch": 0.278855975485189,
+      "grad_norm": 1.3280925750732422,
+      "learning_rate": 1e-06,
+      "loss": 0.0228,
+      "step": 1638
+    },
+    {
+      "epoch": 0.27902621722846443,
+      "grad_norm": 1.5057625770568848,
+      "learning_rate": 1e-06,
+      "loss": 0.0228,
+      "step": 1639
+    },
+    {
+      "epoch": 0.2791964589717399,
+      "grad_norm": 1.3928744792938232,
+      "learning_rate": 1e-06,
+      "loss": 0.0194,
+      "step": 1640
+    },
+    {
+      "epoch": 0.27936670071501535,
+      "grad_norm": 1.3840206861495972,
+      "learning_rate": 1e-06,
+      "loss": 0.0209,
+      "step": 1641
+    },
+    {
+      "epoch": 0.27953694245829075,
+      "grad_norm": 1.7165454626083374,
+      "learning_rate": 1e-06,
+      "loss": 0.0301,
+      "step": 1642
+    },
+    {
+      "epoch": 0.2797071842015662,
+      "grad_norm": 2.1626944541931152,
+      "learning_rate": 1e-06,
+      "loss": 0.0354,
+      "step": 1643
+    },
+    {
+      "epoch": 0.27987742594484166,
+      "grad_norm": 1.480241060256958,
+      "learning_rate": 1e-06,
+      "loss": 0.0251,
+      "step": 1644
+    },
+    {
+      "epoch": 0.2800476676881171,
+      "grad_norm": 1.88970947265625,
+      "learning_rate": 1e-06,
+      "loss": 0.0319,
+      "step": 1645
+    },
+    {
+      "epoch": 0.2802179094313926,
+      "grad_norm": 1.82145357131958,
+      "learning_rate": 1e-06,
+      "loss": 0.033,
+      "step": 1646
+    },
+    {
+      "epoch": 0.28038815117466803,
+      "grad_norm": 1.4362519979476929,
+      "learning_rate": 1e-06,
+      "loss": 0.0279,
+      "step": 1647
+    },
+    {
+      "epoch": 0.2805583929179435,
+      "grad_norm": 1.5925521850585938,
+      "learning_rate": 1e-06,
+      "loss": 0.0274,
+      "step": 1648
+    },
+    {
+      "epoch": 0.28072863466121895,
+      "grad_norm": 1.46236252784729,
+      "learning_rate": 1e-06,
+      "loss": 0.0245,
+      "step": 1649
+    },
+    {
+      "epoch": 0.2808988764044944,
+      "grad_norm": 2.9916434288024902,
+      "learning_rate": 1e-06,
+      "loss": 0.0428,
+      "step": 1650
+    },
+    {
+      "epoch": 0.28106911814776986,
+      "grad_norm": 1.4459913969039917,
+      "learning_rate": 1e-06,
+      "loss": 0.0264,
+      "step": 1651
+    },
+    {
+      "epoch": 0.28123935989104526,
+      "grad_norm": 1.4686808586120605,
+      "learning_rate": 1e-06,
+      "loss": 0.0319,
+      "step": 1652
+    },
+    {
+      "epoch": 0.2814096016343207,
+      "grad_norm": 1.5071327686309814,
+      "learning_rate": 1e-06,
+      "loss": 0.0234,
+      "step": 1653
+    },
+    {
+      "epoch": 0.2815798433775962,
+      "grad_norm": 1.4609589576721191,
+      "learning_rate": 1e-06,
+      "loss": 0.0255,
+      "step": 1654
+    },
+    {
+      "epoch": 0.28175008512087163,
+      "grad_norm": 1.7668788433074951,
+      "learning_rate": 1e-06,
+      "loss": 0.0235,
+      "step": 1655
+    },
+    {
+      "epoch": 0.2819203268641471,
+      "grad_norm": 1.9806582927703857,
+      "learning_rate": 1e-06,
+      "loss": 0.0307,
+      "step": 1656
+    },
+    {
+      "epoch": 0.28209056860742254,
+      "grad_norm": 1.548424243927002,
+      "learning_rate": 1e-06,
+      "loss": 0.0177,
+      "step": 1657
+    },
+    {
+      "epoch": 0.282260810350698,
+      "grad_norm": 1.318804144859314,
+      "learning_rate": 1e-06,
+      "loss": 0.0148,
+      "step": 1658
+    },
+    {
+      "epoch": 0.28243105209397346,
+      "grad_norm": 1.7616188526153564,
+      "learning_rate": 1e-06,
+      "loss": 0.0308,
+      "step": 1659
+    },
+    {
+      "epoch": 0.2826012938372489,
+      "grad_norm": 1.4536751508712769,
+      "learning_rate": 1e-06,
+      "loss": 0.0226,
+      "step": 1660
+    },
+    {
+      "epoch": 0.28277153558052437,
+      "grad_norm": 1.8000961542129517,
+      "learning_rate": 1e-06,
+      "loss": 0.0287,
+      "step": 1661
+    },
+    {
+      "epoch": 0.28294177732379977,
+      "grad_norm": 1.4188201427459717,
+      "learning_rate": 1e-06,
+      "loss": 0.0202,
+      "step": 1662
+    },
+    {
+      "epoch": 0.2831120190670752,
+      "grad_norm": 1.6086786985397339,
+      "learning_rate": 1e-06,
+      "loss": 0.0248,
+      "step": 1663
+    },
+    {
+      "epoch": 0.2832822608103507,
+      "grad_norm": 1.3086551427841187,
+      "learning_rate": 1e-06,
+      "loss": 0.0232,
+      "step": 1664
+    },
+    {
+      "epoch": 0.28345250255362614,
+      "grad_norm": 1.5026829242706299,
+      "learning_rate": 1e-06,
+      "loss": 0.0203,
+      "step": 1665
+    },
+    {
+      "epoch": 0.2836227442969016,
+      "grad_norm": 1.9546946287155151,
+      "learning_rate": 1e-06,
+      "loss": 0.0257,
+      "step": 1666
+    },
+    {
+      "epoch": 0.28379298604017705,
+      "grad_norm": 1.1366575956344604,
+      "learning_rate": 1e-06,
+      "loss": 0.0228,
+      "step": 1667
+    },
+    {
+      "epoch": 0.2839632277834525,
+      "grad_norm": 1.8793129920959473,
+      "learning_rate": 1e-06,
+      "loss": 0.0443,
+      "step": 1668
+    },
+    {
+      "epoch": 0.28413346952672797,
+      "grad_norm": 1.4484946727752686,
+      "learning_rate": 1e-06,
+      "loss": 0.0312,
+      "step": 1669
+    },
+    {
+      "epoch": 0.2843037112700034,
+      "grad_norm": 1.7814072370529175,
+      "learning_rate": 1e-06,
+      "loss": 0.0346,
+      "step": 1670
+    },
+    {
+      "epoch": 0.2844739530132789,
+      "grad_norm": 1.5171910524368286,
+      "learning_rate": 1e-06,
+      "loss": 0.0231,
+      "step": 1671
+    },
+    {
+      "epoch": 0.2846441947565543,
+      "grad_norm": 1.7551907300949097,
+      "learning_rate": 1e-06,
+      "loss": 0.0196,
+      "step": 1672
+    },
+    {
+      "epoch": 0.28481443649982974,
+      "grad_norm": 1.8677160739898682,
+      "learning_rate": 1e-06,
+      "loss": 0.0276,
+      "step": 1673
+    },
+    {
+      "epoch": 0.2849846782431052,
+      "grad_norm": 1.342197299003601,
+      "learning_rate": 1e-06,
+      "loss": 0.0183,
+      "step": 1674
+    },
+    {
+      "epoch": 0.28515491998638065,
+      "grad_norm": 1.656657338142395,
+      "learning_rate": 1e-06,
+      "loss": 0.0262,
+      "step": 1675
+    },
+    {
+      "epoch": 0.2853251617296561,
+      "grad_norm": 1.6236470937728882,
+      "learning_rate": 1e-06,
+      "loss": 0.0234,
+      "step": 1676
+    },
+    {
+      "epoch": 0.28549540347293156,
+      "grad_norm": 1.7253212928771973,
+      "learning_rate": 1e-06,
+      "loss": 0.0274,
+      "step": 1677
+    },
+    {
+      "epoch": 0.285665645216207,
+      "grad_norm": 1.9276634454727173,
+      "learning_rate": 1e-06,
+      "loss": 0.0295,
+      "step": 1678
+    },
+    {
+      "epoch": 0.2858358869594825,
+      "grad_norm": 1.4296467304229736,
+      "learning_rate": 1e-06,
+      "loss": 0.0208,
+      "step": 1679
+    },
+    {
+      "epoch": 0.28600612870275793,
+      "grad_norm": 2.423842430114746,
+      "learning_rate": 1e-06,
+      "loss": 0.0329,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2861763704460334,
+      "grad_norm": 1.48721444606781,
+      "learning_rate": 1e-06,
+      "loss": 0.0237,
+      "step": 1681
+    },
+    {
+      "epoch": 0.2863466121893088,
+      "grad_norm": 1.7897123098373413,
+      "learning_rate": 1e-06,
+      "loss": 0.0281,
+      "step": 1682
+    },
+    {
+      "epoch": 0.28651685393258425,
+      "grad_norm": 1.6738888025283813,
+      "learning_rate": 1e-06,
+      "loss": 0.0332,
+      "step": 1683
+    },
+    {
+      "epoch": 0.2866870956758597,
+      "grad_norm": 1.615144968032837,
+      "learning_rate": 1e-06,
+      "loss": 0.025,
+      "step": 1684
+    },
+    {
+      "epoch": 0.28685733741913516,
+      "grad_norm": 1.4232816696166992,
+      "learning_rate": 1e-06,
+      "loss": 0.0209,
+      "step": 1685
+    },
+    {
+      "epoch": 0.2870275791624106,
+      "grad_norm": 1.6573175191879272,
+      "learning_rate": 1e-06,
+      "loss": 0.0269,
+      "step": 1686
+    },
+    {
+      "epoch": 0.2871978209056861,
+      "grad_norm": 1.7181310653686523,
+      "learning_rate": 1e-06,
+      "loss": 0.0267,
+      "step": 1687
+    },
+    {
+      "epoch": 0.28736806264896153,
+      "grad_norm": 1.8144760131835938,
+      "learning_rate": 1e-06,
+      "loss": 0.0373,
+      "step": 1688
+    },
+    {
+      "epoch": 0.287538304392237,
+      "grad_norm": 1.360830545425415,
+      "learning_rate": 1e-06,
+      "loss": 0.0241,
+      "step": 1689
+    },
+    {
+      "epoch": 0.28770854613551244,
+      "grad_norm": 1.4611048698425293,
+      "learning_rate": 1e-06,
+      "loss": 0.0262,
+      "step": 1690
+    },
+    {
+      "epoch": 0.2878787878787879,
+      "grad_norm": 1.525647759437561,
+      "learning_rate": 1e-06,
+      "loss": 0.0283,
+      "step": 1691
+    },
+    {
+      "epoch": 0.28804902962206336,
+      "grad_norm": 1.4439351558685303,
+      "learning_rate": 1e-06,
+      "loss": 0.0274,
+      "step": 1692
+    },
+    {
+      "epoch": 0.28821927136533876,
+      "grad_norm": 1.6593124866485596,
+      "learning_rate": 1e-06,
+      "loss": 0.0256,
+      "step": 1693
+    },
+    {
+      "epoch": 0.2883895131086142,
+      "grad_norm": 1.6211251020431519,
+      "learning_rate": 1e-06,
+      "loss": 0.0388,
+      "step": 1694
+    },
+    {
+      "epoch": 0.28855975485188967,
+      "grad_norm": 1.6730520725250244,
+      "learning_rate": 1e-06,
+      "loss": 0.0229,
+      "step": 1695
+    },
+    {
+      "epoch": 0.28872999659516513,
+      "grad_norm": 1.850048542022705,
+      "learning_rate": 1e-06,
+      "loss": 0.0202,
+      "step": 1696
+    },
+    {
+      "epoch": 0.2889002383384406,
+      "grad_norm": 1.4730114936828613,
+      "learning_rate": 1e-06,
+      "loss": 0.0235,
+      "step": 1697
+    },
+    {
+      "epoch": 0.28907048008171604,
+      "grad_norm": 1.8073022365570068,
+      "learning_rate": 1e-06,
+      "loss": 0.04,
+      "step": 1698
+    },
+    {
+      "epoch": 0.2892407218249915,
+      "grad_norm": 1.8779622316360474,
+      "learning_rate": 1e-06,
+      "loss": 0.0437,
+      "step": 1699
+    },
+    {
+      "epoch": 0.28941096356826695,
+      "grad_norm": 1.5742496252059937,
+      "learning_rate": 1e-06,
+      "loss": 0.0293,
+      "step": 1700
+    },
+    {
+      "epoch": 0.2895812053115424,
+      "grad_norm": 1.4416614770889282,
+      "learning_rate": 1e-06,
+      "loss": 0.023,
+      "step": 1701
+    },
+    {
+      "epoch": 0.28975144705481787,
+      "grad_norm": 1.4189447164535522,
+      "learning_rate": 1e-06,
+      "loss": 0.0281,
+      "step": 1702
+    },
+    {
+      "epoch": 0.28992168879809327,
+      "grad_norm": 1.4489812850952148,
+      "learning_rate": 1e-06,
+      "loss": 0.0269,
+      "step": 1703
+    },
+    {
+      "epoch": 0.2900919305413687,
+      "grad_norm": 1.590919852256775,
+      "learning_rate": 1e-06,
+      "loss": 0.0261,
+      "step": 1704
+    },
+    {
+      "epoch": 0.2902621722846442,
+      "grad_norm": 1.528601884841919,
+      "learning_rate": 1e-06,
+      "loss": 0.0246,
+      "step": 1705
+    },
+    {
+      "epoch": 0.29043241402791964,
+      "grad_norm": 1.3978060483932495,
+      "learning_rate": 1e-06,
+      "loss": 0.0202,
+      "step": 1706
+    },
+    {
+      "epoch": 0.2906026557711951,
+      "grad_norm": 1.4895422458648682,
+      "learning_rate": 1e-06,
+      "loss": 0.0263,
+      "step": 1707
+    },
+    {
+      "epoch": 0.29077289751447055,
+      "grad_norm": 1.5982104539871216,
+      "learning_rate": 1e-06,
+      "loss": 0.0143,
+      "step": 1708
+    },
+    {
+      "epoch": 0.290943139257746,
+      "grad_norm": 1.5358942747116089,
+      "learning_rate": 1e-06,
+      "loss": 0.0223,
+      "step": 1709
+    },
+    {
+      "epoch": 0.29111338100102147,
+      "grad_norm": 1.491877794265747,
+      "learning_rate": 1e-06,
+      "loss": 0.0252,
+      "step": 1710
+    },
+    {
+      "epoch": 0.2912836227442969,
+      "grad_norm": 1.6114052534103394,
+      "learning_rate": 1e-06,
+      "loss": 0.0226,
+      "step": 1711
+    },
+    {
+      "epoch": 0.2914538644875724,
+      "grad_norm": 1.6193110942840576,
+      "learning_rate": 1e-06,
+      "loss": 0.0221,
+      "step": 1712
+    },
+    {
+      "epoch": 0.2916241062308478,
+      "grad_norm": 1.40981924533844,
+      "learning_rate": 1e-06,
+      "loss": 0.0209,
+      "step": 1713
+    },
+    {
+      "epoch": 0.29179434797412324,
+      "grad_norm": 2.308316469192505,
+      "learning_rate": 1e-06,
+      "loss": 0.0314,
+      "step": 1714
+    },
+    {
+      "epoch": 0.2919645897173987,
+      "grad_norm": 1.631408929824829,
+      "learning_rate": 1e-06,
+      "loss": 0.0307,
+      "step": 1715
+    },
+    {
+      "epoch": 0.29213483146067415,
+      "grad_norm": 1.4540927410125732,
+      "learning_rate": 1e-06,
+      "loss": 0.0258,
+      "step": 1716
+    },
+    {
+      "epoch": 0.2923050732039496,
+      "grad_norm": 1.7663581371307373,
+      "learning_rate": 1e-06,
+      "loss": 0.0332,
+      "step": 1717
+    },
+    {
+      "epoch": 0.29247531494722506,
+      "grad_norm": 1.2516728639602661,
+      "learning_rate": 1e-06,
+      "loss": 0.021,
+      "step": 1718
+    },
+    {
+      "epoch": 0.2926455566905005,
+      "grad_norm": 1.9072999954223633,
+      "learning_rate": 1e-06,
+      "loss": 0.0308,
+      "step": 1719
+    },
+    {
+      "epoch": 0.292815798433776,
+      "grad_norm": 1.8111470937728882,
+      "learning_rate": 1e-06,
+      "loss": 0.0261,
+      "step": 1720
+    },
+    {
+      "epoch": 0.29298604017705143,
+      "grad_norm": 1.5507558584213257,
+      "learning_rate": 1e-06,
+      "loss": 0.022,
+      "step": 1721
+    },
+    {
+      "epoch": 0.2931562819203269,
+      "grad_norm": 1.2865614891052246,
+      "learning_rate": 1e-06,
+      "loss": 0.0196,
+      "step": 1722
+    },
+    {
+      "epoch": 0.2933265236636023,
+      "grad_norm": 1.6672788858413696,
+      "learning_rate": 1e-06,
+      "loss": 0.0191,
+      "step": 1723
+    },
+    {
+      "epoch": 0.29349676540687775,
+      "grad_norm": 1.670101284980774,
+      "learning_rate": 1e-06,
+      "loss": 0.03,
+      "step": 1724
+    },
+    {
+      "epoch": 0.2936670071501532,
+      "grad_norm": 1.2135088443756104,
+      "learning_rate": 1e-06,
+      "loss": 0.0219,
+      "step": 1725
+    },
+    {
+      "epoch": 0.29383724889342866,
+      "grad_norm": 1.2077770233154297,
+      "learning_rate": 1e-06,
+      "loss": 0.0179,
+      "step": 1726
+    },
+    {
+      "epoch": 0.2940074906367041,
+      "grad_norm": 1.3592512607574463,
+      "learning_rate": 1e-06,
+      "loss": 0.0235,
+      "step": 1727
+    },
+    {
+      "epoch": 0.2941777323799796,
+      "grad_norm": 1.661252737045288,
+      "learning_rate": 1e-06,
+      "loss": 0.0247,
+      "step": 1728
+    },
+    {
+      "epoch": 0.29434797412325503,
+      "grad_norm": 1.8278785943984985,
+      "learning_rate": 1e-06,
+      "loss": 0.0403,
+      "step": 1729
+    },
+    {
+      "epoch": 0.2945182158665305,
+      "grad_norm": 1.3359884023666382,
+      "learning_rate": 1e-06,
+      "loss": 0.0206,
+      "step": 1730
+    },
+    {
+      "epoch": 0.29468845760980594,
+      "grad_norm": 1.8488993644714355,
+      "learning_rate": 1e-06,
+      "loss": 0.0247,
+      "step": 1731
+    },
+    {
+      "epoch": 0.2948586993530814,
+      "grad_norm": 1.6495951414108276,
+      "learning_rate": 1e-06,
+      "loss": 0.0268,
+      "step": 1732
+    },
+    {
+      "epoch": 0.2950289410963568,
+      "grad_norm": 2.0835161209106445,
+      "learning_rate": 1e-06,
+      "loss": 0.0373,
+      "step": 1733
+    },
+    {
+      "epoch": 0.29519918283963226,
+      "grad_norm": 1.2408726215362549,
+      "learning_rate": 1e-06,
+      "loss": 0.0212,
+      "step": 1734
+    },
+    {
+      "epoch": 0.2953694245829077,
+      "grad_norm": 1.3629958629608154,
+      "learning_rate": 1e-06,
+      "loss": 0.0263,
+      "step": 1735
+    },
+    {
+      "epoch": 0.29553966632618317,
+      "grad_norm": 1.1778775453567505,
+      "learning_rate": 1e-06,
+      "loss": 0.0185,
+      "step": 1736
+    },
+    {
+      "epoch": 0.2957099080694586,
+      "grad_norm": 1.1877045631408691,
+      "learning_rate": 1e-06,
+      "loss": 0.0216,
+      "step": 1737
+    },
+    {
+      "epoch": 0.2958801498127341,
+      "grad_norm": 1.8210134506225586,
+      "learning_rate": 1e-06,
+      "loss": 0.0264,
+      "step": 1738
+    },
+    {
+      "epoch": 0.29605039155600954,
+      "grad_norm": 1.5456937551498413,
+      "learning_rate": 1e-06,
+      "loss": 0.0188,
+      "step": 1739
+    },
+    {
+      "epoch": 0.296220633299285,
+      "grad_norm": 1.523476004600525,
+      "learning_rate": 1e-06,
+      "loss": 0.0195,
+      "step": 1740
+    },
+    {
+      "epoch": 0.29639087504256045,
+      "grad_norm": 1.5979241132736206,
+      "learning_rate": 1e-06,
+      "loss": 0.0226,
+      "step": 1741
+    },
+    {
+      "epoch": 0.2965611167858359,
+      "grad_norm": 1.4582843780517578,
+      "learning_rate": 1e-06,
+      "loss": 0.0227,
+      "step": 1742
+    },
+    {
+      "epoch": 0.2967313585291113,
+      "grad_norm": 1.4900776147842407,
+      "learning_rate": 1e-06,
+      "loss": 0.0283,
+      "step": 1743
+    },
+    {
+      "epoch": 0.29690160027238677,
+      "grad_norm": 1.4199949502944946,
+      "learning_rate": 1e-06,
+      "loss": 0.031,
+      "step": 1744
+    },
+    {
+      "epoch": 0.2970718420156622,
+      "grad_norm": 1.9383891820907593,
+      "learning_rate": 1e-06,
+      "loss": 0.0351,
+      "step": 1745
+    },
+    {
+      "epoch": 0.2972420837589377,
+      "grad_norm": 1.6513484716415405,
+      "learning_rate": 1e-06,
+      "loss": 0.0283,
+      "step": 1746
+    },
+    {
+      "epoch": 0.29741232550221314,
+      "grad_norm": 1.4408584833145142,
+      "learning_rate": 1e-06,
+      "loss": 0.0253,
+      "step": 1747
+    },
+    {
+      "epoch": 0.2975825672454886,
+      "grad_norm": 1.4601348638534546,
+      "learning_rate": 1e-06,
+      "loss": 0.0246,
+      "step": 1748
+    },
+    {
+      "epoch": 0.29775280898876405,
+      "grad_norm": 2.014059066772461,
+      "learning_rate": 1e-06,
+      "loss": 0.0239,
+      "step": 1749
+    },
+    {
+      "epoch": 0.2979230507320395,
+      "grad_norm": 1.4093255996704102,
+      "learning_rate": 1e-06,
+      "loss": 0.0242,
+      "step": 1750
+    },
+    {
+      "epoch": 0.29809329247531496,
+      "grad_norm": 1.8597747087478638,
+      "learning_rate": 1e-06,
+      "loss": 0.0305,
+      "step": 1751
+    },
+    {
+      "epoch": 0.2982635342185904,
+      "grad_norm": 1.4685695171356201,
+      "learning_rate": 1e-06,
+      "loss": 0.0226,
+      "step": 1752
+    },
+    {
+      "epoch": 0.2984337759618659,
+      "grad_norm": 1.4107317924499512,
+      "learning_rate": 1e-06,
+      "loss": 0.0278,
+      "step": 1753
+    },
+    {
+      "epoch": 0.2986040177051413,
+      "grad_norm": 1.1390619277954102,
+      "learning_rate": 1e-06,
+      "loss": 0.0272,
+      "step": 1754
+    },
+    {
+      "epoch": 0.29877425944841673,
+      "grad_norm": 1.5630130767822266,
+      "learning_rate": 1e-06,
+      "loss": 0.03,
+      "step": 1755
+    },
+    {
+      "epoch": 0.2989445011916922,
+      "grad_norm": 1.6469208002090454,
+      "learning_rate": 1e-06,
+      "loss": 0.02,
+      "step": 1756
+    },
+    {
+      "epoch": 0.29911474293496765,
+      "grad_norm": 1.4508461952209473,
+      "learning_rate": 1e-06,
+      "loss": 0.028,
+      "step": 1757
+    },
+    {
+      "epoch": 0.2992849846782431,
+      "grad_norm": 1.506699800491333,
+      "learning_rate": 1e-06,
+      "loss": 0.0292,
+      "step": 1758
+    },
+    {
+      "epoch": 0.29945522642151856,
+      "grad_norm": 1.0725475549697876,
+      "learning_rate": 1e-06,
+      "loss": 0.0173,
+      "step": 1759
+    },
+    {
+      "epoch": 0.299625468164794,
+      "grad_norm": 1.733446717262268,
+      "learning_rate": 1e-06,
+      "loss": 0.0375,
+      "step": 1760
+    },
+    {
+      "epoch": 0.2997957099080695,
+      "grad_norm": 1.4227663278579712,
+      "learning_rate": 1e-06,
+      "loss": 0.0203,
+      "step": 1761
+    },
+    {
+      "epoch": 0.29996595165134493,
+      "grad_norm": 1.4566940069198608,
+      "learning_rate": 1e-06,
+      "loss": 0.0264,
+      "step": 1762
+    },
+    {
+      "epoch": 0.3001361933946204,
+      "grad_norm": 1.6807788610458374,
+      "learning_rate": 1e-06,
+      "loss": 0.0281,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3003064351378958,
+      "grad_norm": 1.8362802267074585,
+      "learning_rate": 1e-06,
+      "loss": 0.0251,
+      "step": 1764
+    },
+    {
+      "epoch": 0.30047667688117125,
+      "grad_norm": 1.8242303133010864,
+      "learning_rate": 1e-06,
+      "loss": 0.0274,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3006469186244467,
+      "grad_norm": 1.639448881149292,
+      "learning_rate": 1e-06,
+      "loss": 0.0282,
+      "step": 1766
+    },
+    {
+      "epoch": 0.30081716036772216,
+      "grad_norm": 1.406978726387024,
+      "learning_rate": 1e-06,
+      "loss": 0.0238,
+      "step": 1767
+    },
+    {
+      "epoch": 0.3009874021109976,
+      "grad_norm": 1.5948950052261353,
+      "learning_rate": 1e-06,
+      "loss": 0.0266,
+      "step": 1768
+    },
+    {
+      "epoch": 0.30115764385427307,
+      "grad_norm": 1.1911594867706299,
+      "learning_rate": 1e-06,
+      "loss": 0.016,
+      "step": 1769
+    },
+    {
+      "epoch": 0.30132788559754853,
+      "grad_norm": 1.3455018997192383,
+      "learning_rate": 1e-06,
+      "loss": 0.0196,
+      "step": 1770
+    },
+    {
+      "epoch": 0.301498127340824,
+      "grad_norm": 1.5457820892333984,
+      "learning_rate": 1e-06,
+      "loss": 0.021,
+      "step": 1771
+    },
+    {
+      "epoch": 0.30166836908409944,
+      "grad_norm": 1.7167928218841553,
+      "learning_rate": 1e-06,
+      "loss": 0.0282,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3018386108273749,
+      "grad_norm": 1.1246789693832397,
+      "learning_rate": 1e-06,
+      "loss": 0.0225,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3020088525706503,
+      "grad_norm": 1.7185958623886108,
+      "learning_rate": 1e-06,
+      "loss": 0.0319,
+      "step": 1774
+    },
+    {
+      "epoch": 0.30217909431392576,
+      "grad_norm": 1.2617838382720947,
+      "learning_rate": 1e-06,
+      "loss": 0.0219,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3023493360572012,
+      "grad_norm": 1.5254855155944824,
+      "learning_rate": 1e-06,
+      "loss": 0.0202,
+      "step": 1776
+    },
+    {
+      "epoch": 0.30251957780047667,
+      "grad_norm": 1.1870532035827637,
+      "learning_rate": 1e-06,
+      "loss": 0.0131,
+      "step": 1777
+    },
+    {
+      "epoch": 0.3026898195437521,
+      "grad_norm": 1.7242926359176636,
+      "learning_rate": 1e-06,
+      "loss": 0.0253,
+      "step": 1778
+    },
+    {
+      "epoch": 0.3028600612870276,
+      "grad_norm": 1.8268846273422241,
+      "learning_rate": 1e-06,
+      "loss": 0.0249,
+      "step": 1779
+    },
+    {
+      "epoch": 0.30303030303030304,
+      "grad_norm": 1.4591509103775024,
+      "learning_rate": 1e-06,
+      "loss": 0.0193,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3032005447735785,
+      "grad_norm": 1.7261550426483154,
+      "learning_rate": 1e-06,
+      "loss": 0.0299,
+      "step": 1781
+    },
+    {
+      "epoch": 0.30337078651685395,
+      "grad_norm": 1.9499040842056274,
+      "learning_rate": 1e-06,
+      "loss": 0.0374,
+      "step": 1782
+    },
+    {
+      "epoch": 0.3035410282601294,
+      "grad_norm": 1.2761621475219727,
+      "learning_rate": 1e-06,
+      "loss": 0.0215,
+      "step": 1783
+    },
+    {
+      "epoch": 0.3037112700034048,
+      "grad_norm": 1.402917742729187,
+      "learning_rate": 1e-06,
+      "loss": 0.0223,
+      "step": 1784
+    },
+    {
+      "epoch": 0.30388151174668027,
+      "grad_norm": 3.2693288326263428,
+      "learning_rate": 1e-06,
+      "loss": 0.0334,
+      "step": 1785
+    },
+    {
+      "epoch": 0.3040517534899557,
+      "grad_norm": 1.3207221031188965,
+      "learning_rate": 1e-06,
+      "loss": 0.0174,
+      "step": 1786
+    },
+    {
+      "epoch": 0.3042219952332312,
+      "grad_norm": 2.6473004817962646,
+      "learning_rate": 1e-06,
+      "loss": 0.0406,
+      "step": 1787
+    },
+    {
+      "epoch": 0.30439223697650664,
+      "grad_norm": 1.6894813776016235,
+      "learning_rate": 1e-06,
+      "loss": 0.0283,
+      "step": 1788
+    },
+    {
+      "epoch": 0.3045624787197821,
+      "grad_norm": 1.3546245098114014,
+      "learning_rate": 1e-06,
+      "loss": 0.0238,
+      "step": 1789
+    },
+    {
+      "epoch": 0.30473272046305755,
+      "grad_norm": 1.6987773180007935,
+      "learning_rate": 1e-06,
+      "loss": 0.0301,
+      "step": 1790
+    },
+    {
+      "epoch": 0.304902962206333,
+      "grad_norm": 1.6079025268554688,
+      "learning_rate": 1e-06,
+      "loss": 0.0222,
+      "step": 1791
+    },
+    {
+      "epoch": 0.30507320394960846,
+      "grad_norm": 1.5735406875610352,
+      "learning_rate": 1e-06,
+      "loss": 0.0222,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3052434456928839,
+      "grad_norm": 1.862619400024414,
+      "learning_rate": 1e-06,
+      "loss": 0.0432,
+      "step": 1793
+    },
+    {
+      "epoch": 0.3054136874361593,
+      "grad_norm": 1.18687903881073,
+      "learning_rate": 1e-06,
+      "loss": 0.0145,
+      "step": 1794
+    },
+    {
+      "epoch": 0.3055839291794348,
+      "grad_norm": 1.4821864366531372,
+      "learning_rate": 1e-06,
+      "loss": 0.0234,
+      "step": 1795
+    },
+    {
+      "epoch": 0.30575417092271023,
+      "grad_norm": 1.6538077592849731,
+      "learning_rate": 1e-06,
+      "loss": 0.0225,
+      "step": 1796
+    },
+    {
+      "epoch": 0.3059244126659857,
+      "grad_norm": 1.546225666999817,
+      "learning_rate": 1e-06,
+      "loss": 0.0261,
+      "step": 1797
+    },
+    {
+      "epoch": 0.30609465440926115,
+      "grad_norm": 1.4453809261322021,
+      "learning_rate": 1e-06,
+      "loss": 0.0216,
+      "step": 1798
+    },
+    {
+      "epoch": 0.3062648961525366,
+      "grad_norm": 2.1485629081726074,
+      "learning_rate": 1e-06,
+      "loss": 0.0324,
+      "step": 1799
+    },
+    {
+      "epoch": 0.30643513789581206,
+      "grad_norm": 1.6263507604599,
+      "learning_rate": 1e-06,
+      "loss": 0.0212,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3066053796390875,
+      "grad_norm": 1.6911835670471191,
+      "learning_rate": 1e-06,
+      "loss": 0.0241,
+      "step": 1801
+    },
+    {
+      "epoch": 0.306775621382363,
+      "grad_norm": 1.4971544742584229,
+      "learning_rate": 1e-06,
+      "loss": 0.0215,
+      "step": 1802
+    },
+    {
+      "epoch": 0.30694586312563843,
+      "grad_norm": 1.3353301286697388,
+      "learning_rate": 1e-06,
+      "loss": 0.0167,
+      "step": 1803
+    },
+    {
+      "epoch": 0.30711610486891383,
+      "grad_norm": 1.5353763103485107,
+      "learning_rate": 1e-06,
+      "loss": 0.021,
+      "step": 1804
+    },
+    {
+      "epoch": 0.3072863466121893,
+      "grad_norm": 1.2348135709762573,
+      "learning_rate": 1e-06,
+      "loss": 0.0203,
+      "step": 1805
+    },
+    {
+      "epoch": 0.30745658835546474,
+      "grad_norm": 1.4471068382263184,
+      "learning_rate": 1e-06,
+      "loss": 0.0204,
+      "step": 1806
+    },
+    {
+      "epoch": 0.3076268300987402,
+      "grad_norm": 1.7940800189971924,
+      "learning_rate": 1e-06,
+      "loss": 0.0204,
+      "step": 1807
+    },
+    {
+      "epoch": 0.30779707184201566,
+      "grad_norm": 1.4490859508514404,
+      "learning_rate": 1e-06,
+      "loss": 0.0186,
+      "step": 1808
+    },
+    {
+      "epoch": 0.3079673135852911,
+      "grad_norm": 1.3849128484725952,
+      "learning_rate": 1e-06,
+      "loss": 0.0229,
+      "step": 1809
+    },
+    {
+      "epoch": 0.30813755532856657,
+      "grad_norm": 1.6811866760253906,
+      "learning_rate": 1e-06,
+      "loss": 0.0159,
+      "step": 1810
+    },
+    {
+      "epoch": 0.308307797071842,
+      "grad_norm": 1.5585503578186035,
+      "learning_rate": 1e-06,
+      "loss": 0.0178,
+      "step": 1811
+    },
+    {
+      "epoch": 0.3084780388151175,
+      "grad_norm": 1.498989462852478,
+      "learning_rate": 1e-06,
+      "loss": 0.0215,
+      "step": 1812
+    },
+    {
+      "epoch": 0.30864828055839294,
+      "grad_norm": 1.5518637895584106,
+      "learning_rate": 1e-06,
+      "loss": 0.0145,
+      "step": 1813
+    },
+    {
+      "epoch": 0.3088185223016684,
+      "grad_norm": 1.9985408782958984,
+      "learning_rate": 1e-06,
+      "loss": 0.0384,
+      "step": 1814
+    },
+    {
+      "epoch": 0.3089887640449438,
+      "grad_norm": 1.9101368188858032,
+      "learning_rate": 1e-06,
+      "loss": 0.022,
+      "step": 1815
+    },
+    {
+      "epoch": 0.30915900578821925,
+      "grad_norm": 1.438533902168274,
+      "learning_rate": 1e-06,
+      "loss": 0.0206,
+      "step": 1816
+    },
+    {
+      "epoch": 0.3093292475314947,
+      "grad_norm": 1.6751413345336914,
+      "learning_rate": 1e-06,
+      "loss": 0.0245,
+      "step": 1817
+    },
+    {
+      "epoch": 0.30949948927477017,
+      "grad_norm": 1.1888127326965332,
+      "learning_rate": 1e-06,
+      "loss": 0.0207,
+      "step": 1818
+    },
+    {
+      "epoch": 0.3096697310180456,
+      "grad_norm": 2.00555157661438,
+      "learning_rate": 1e-06,
+      "loss": 0.0188,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3098399727613211,
+      "grad_norm": 1.2455949783325195,
+      "learning_rate": 1e-06,
+      "loss": 0.021,
+      "step": 1820
+    },
+    {
+      "epoch": 0.31001021450459654,
+      "grad_norm": 1.2500693798065186,
+      "learning_rate": 1e-06,
+      "loss": 0.0188,
+      "step": 1821
+    },
+    {
+      "epoch": 0.310180456247872,
+      "grad_norm": 1.3195619583129883,
+      "learning_rate": 1e-06,
+      "loss": 0.022,
+      "step": 1822
+    },
+    {
+      "epoch": 0.31035069799114745,
+      "grad_norm": 1.429261326789856,
+      "learning_rate": 1e-06,
+      "loss": 0.0294,
+      "step": 1823
+    },
+    {
+      "epoch": 0.3105209397344229,
+      "grad_norm": 1.3425854444503784,
+      "learning_rate": 1e-06,
+      "loss": 0.0209,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3106911814776983,
+      "grad_norm": 1.5613363981246948,
+      "learning_rate": 1e-06,
+      "loss": 0.0249,
+      "step": 1825
+    },
+    {
+      "epoch": 0.31086142322097376,
+      "grad_norm": 1.9861626625061035,
+      "learning_rate": 1e-06,
+      "loss": 0.0237,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3110316649642492,
+      "grad_norm": 1.774733304977417,
+      "learning_rate": 1e-06,
+      "loss": 0.0295,
+      "step": 1827
+    },
+    {
+      "epoch": 0.3112019067075247,
+      "grad_norm": 1.4516446590423584,
+      "learning_rate": 1e-06,
+      "loss": 0.0247,
+      "step": 1828
+    },
+    {
+      "epoch": 0.31137214845080013,
+      "grad_norm": 1.5618529319763184,
+      "learning_rate": 1e-06,
+      "loss": 0.0227,
+      "step": 1829
+    },
+    {
+      "epoch": 0.3115423901940756,
+      "grad_norm": 1.410374641418457,
+      "learning_rate": 1e-06,
+      "loss": 0.019,
+      "step": 1830
+    },
+    {
+      "epoch": 0.31171263193735105,
+      "grad_norm": 1.3269094228744507,
+      "learning_rate": 1e-06,
+      "loss": 0.0232,
+      "step": 1831
+    },
+    {
+      "epoch": 0.3118828736806265,
+      "grad_norm": 2.6260929107666016,
+      "learning_rate": 1e-06,
+      "loss": 0.0334,
+      "step": 1832
+    },
+    {
+      "epoch": 0.31205311542390196,
+      "grad_norm": 1.4052633047103882,
+      "learning_rate": 1e-06,
+      "loss": 0.0147,
+      "step": 1833
+    },
+    {
+      "epoch": 0.3122233571671774,
+      "grad_norm": 1.3996262550354004,
+      "learning_rate": 1e-06,
+      "loss": 0.0231,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3123935989104528,
+      "grad_norm": 1.7314420938491821,
+      "learning_rate": 1e-06,
+      "loss": 0.0289,
+      "step": 1835
+    },
+    {
+      "epoch": 0.3125638406537283,
+      "grad_norm": 1.2537227869033813,
+      "learning_rate": 1e-06,
+      "loss": 0.0209,
+      "step": 1836
+    },
+    {
+      "epoch": 0.31273408239700373,
+      "grad_norm": 1.3801499605178833,
+      "learning_rate": 1e-06,
+      "loss": 0.023,
+      "step": 1837
+    },
+    {
+      "epoch": 0.3129043241402792,
+      "grad_norm": 1.6672334671020508,
+      "learning_rate": 1e-06,
+      "loss": 0.0261,
+      "step": 1838
+    },
+    {
+      "epoch": 0.31307456588355465,
+      "grad_norm": 1.7810288667678833,
+      "learning_rate": 1e-06,
+      "loss": 0.031,
+      "step": 1839
+    },
+    {
+      "epoch": 0.3132448076268301,
+      "grad_norm": 1.435265064239502,
+      "learning_rate": 1e-06,
+      "loss": 0.0178,
+      "step": 1840
+    },
+    {
+      "epoch": 0.31341504937010556,
+      "grad_norm": 1.537177562713623,
+      "learning_rate": 1e-06,
+      "loss": 0.0229,
+      "step": 1841
+    },
+    {
+      "epoch": 0.313585291113381,
+      "grad_norm": 1.7948952913284302,
+      "learning_rate": 1e-06,
+      "loss": 0.0289,
+      "step": 1842
+    },
+    {
+      "epoch": 0.31375553285665647,
+      "grad_norm": 1.2194470167160034,
+      "learning_rate": 1e-06,
+      "loss": 0.0172,
+      "step": 1843
+    },
+    {
+      "epoch": 0.31392577459993193,
+      "grad_norm": 1.6376681327819824,
+      "learning_rate": 1e-06,
+      "loss": 0.0291,
+      "step": 1844
+    },
+    {
+      "epoch": 0.31409601634320733,
+      "grad_norm": 1.389352798461914,
+      "learning_rate": 1e-06,
+      "loss": 0.015,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3142662580864828,
+      "grad_norm": 1.5804452896118164,
+      "learning_rate": 1e-06,
+      "loss": 0.0386,
+      "step": 1846
+    },
+    {
+      "epoch": 0.31443649982975824,
+      "grad_norm": 1.9733364582061768,
+      "learning_rate": 1e-06,
+      "loss": 0.0261,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3146067415730337,
+      "grad_norm": 1.4925074577331543,
+      "learning_rate": 1e-06,
+      "loss": 0.0318,
+      "step": 1848
+    },
+    {
+      "epoch": 0.31477698331630916,
+      "grad_norm": 1.4996342658996582,
+      "learning_rate": 1e-06,
+      "loss": 0.0203,
+      "step": 1849
+    },
+    {
+      "epoch": 0.3149472250595846,
+      "grad_norm": 1.4963217973709106,
+      "learning_rate": 1e-06,
+      "loss": 0.0268,
+      "step": 1850
+    },
+    {
+      "epoch": 0.31511746680286007,
+      "grad_norm": 1.6725646257400513,
+      "learning_rate": 1e-06,
+      "loss": 0.0176,
+      "step": 1851
+    },
+    {
+      "epoch": 0.3152877085461355,
+      "grad_norm": 1.3835301399230957,
+      "learning_rate": 1e-06,
+      "loss": 0.0256,
+      "step": 1852
+    },
+    {
+      "epoch": 0.315457950289411,
+      "grad_norm": 1.7713634967803955,
+      "learning_rate": 1e-06,
+      "loss": 0.0255,
+      "step": 1853
+    },
+    {
+      "epoch": 0.31562819203268644,
+      "grad_norm": 1.1334712505340576,
+      "learning_rate": 1e-06,
+      "loss": 0.0175,
+      "step": 1854
+    },
+    {
+      "epoch": 0.31579843377596184,
+      "grad_norm": 1.6713346242904663,
+      "learning_rate": 1e-06,
+      "loss": 0.0192,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3159686755192373,
+      "grad_norm": 1.463097095489502,
+      "learning_rate": 1e-06,
+      "loss": 0.0243,
+      "step": 1856
+    },
+    {
+      "epoch": 0.31613891726251275,
+      "grad_norm": 1.6081440448760986,
+      "learning_rate": 1e-06,
+      "loss": 0.0262,
+      "step": 1857
+    },
+    {
+      "epoch": 0.3163091590057882,
+      "grad_norm": 1.7677350044250488,
+      "learning_rate": 1e-06,
+      "loss": 0.022,
+      "step": 1858
+    },
+    {
+      "epoch": 0.31647940074906367,
+      "grad_norm": 2.0143535137176514,
+      "learning_rate": 1e-06,
+      "loss": 0.0252,
+      "step": 1859
+    },
+    {
+      "epoch": 0.3166496424923391,
+      "grad_norm": 1.4883443117141724,
+      "learning_rate": 1e-06,
+      "loss": 0.0303,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3168198842356146,
+      "grad_norm": 1.5226455926895142,
+      "learning_rate": 1e-06,
+      "loss": 0.0199,
+      "step": 1861
+    },
+    {
+      "epoch": 0.31699012597889004,
+      "grad_norm": 1.593501091003418,
+      "learning_rate": 1e-06,
+      "loss": 0.0232,
+      "step": 1862
+    },
+    {
+      "epoch": 0.3171603677221655,
+      "grad_norm": 1.3263167142868042,
+      "learning_rate": 1e-06,
+      "loss": 0.024,
+      "step": 1863
+    },
+    {
+      "epoch": 0.31733060946544095,
+      "grad_norm": 1.4621020555496216,
+      "learning_rate": 1e-06,
+      "loss": 0.024,
+      "step": 1864
+    },
+    {
+      "epoch": 0.31750085120871635,
+      "grad_norm": 1.8614875078201294,
+      "learning_rate": 1e-06,
+      "loss": 0.0263,
+      "step": 1865
+    },
+    {
+      "epoch": 0.3176710929519918,
+      "grad_norm": 1.5570732355117798,
+      "learning_rate": 1e-06,
+      "loss": 0.0178,
+      "step": 1866
+    },
+    {
+      "epoch": 0.31784133469526726,
+      "grad_norm": 1.5920957326889038,
+      "learning_rate": 1e-06,
+      "loss": 0.0283,
+      "step": 1867
+    },
+    {
+      "epoch": 0.3180115764385427,
+      "grad_norm": 1.3949881792068481,
+      "learning_rate": 1e-06,
+      "loss": 0.0262,
+      "step": 1868
+    },
+    {
+      "epoch": 0.3181818181818182,
+      "grad_norm": 1.364362359046936,
+      "learning_rate": 1e-06,
+      "loss": 0.0292,
+      "step": 1869
+    },
+    {
+      "epoch": 0.31835205992509363,
+      "grad_norm": 1.4598324298858643,
+      "learning_rate": 1e-06,
+      "loss": 0.0244,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3185223016683691,
+      "grad_norm": 1.320860505104065,
+      "learning_rate": 1e-06,
+      "loss": 0.0198,
+      "step": 1871
+    },
+    {
+      "epoch": 0.31869254341164455,
+      "grad_norm": 1.9381211996078491,
+      "learning_rate": 1e-06,
+      "loss": 0.0263,
+      "step": 1872
+    },
+    {
+      "epoch": 0.31886278515492,
+      "grad_norm": 1.4324305057525635,
+      "learning_rate": 1e-06,
+      "loss": 0.029,
+      "step": 1873
+    },
+    {
+      "epoch": 0.31903302689819546,
+      "grad_norm": 1.2118055820465088,
+      "learning_rate": 1e-06,
+      "loss": 0.0144,
+      "step": 1874
+    },
+    {
+      "epoch": 0.3192032686414709,
+      "grad_norm": 1.200643539428711,
+      "learning_rate": 1e-06,
+      "loss": 0.0174,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3193735103847463,
+      "grad_norm": 1.5699355602264404,
+      "learning_rate": 1e-06,
+      "loss": 0.0288,
+      "step": 1876
+    },
+    {
+      "epoch": 0.3195437521280218,
+      "grad_norm": 1.7179347276687622,
+      "learning_rate": 1e-06,
+      "loss": 0.0267,
+      "step": 1877
+    },
+    {
+      "epoch": 0.31971399387129723,
+      "grad_norm": 1.8065325021743774,
+      "learning_rate": 1e-06,
+      "loss": 0.0256,
+      "step": 1878
+    },
+    {
+      "epoch": 0.3198842356145727,
+      "grad_norm": 2.4595632553100586,
+      "learning_rate": 1e-06,
+      "loss": 0.0239,
+      "step": 1879
+    },
+    {
+      "epoch": 0.32005447735784814,
+      "grad_norm": 1.140232801437378,
+      "learning_rate": 1e-06,
+      "loss": 0.0174,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3202247191011236,
+      "grad_norm": 1.3284320831298828,
+      "learning_rate": 1e-06,
+      "loss": 0.0212,
+      "step": 1881
+    },
+    {
+      "epoch": 0.32039496084439906,
+      "grad_norm": 1.7007908821105957,
+      "learning_rate": 1e-06,
+      "loss": 0.0252,
+      "step": 1882
+    },
+    {
+      "epoch": 0.3205652025876745,
+      "grad_norm": 1.8130091428756714,
+      "learning_rate": 1e-06,
+      "loss": 0.0322,
+      "step": 1883
+    },
+    {
+      "epoch": 0.32073544433094997,
+      "grad_norm": 1.434629201889038,
+      "learning_rate": 1e-06,
+      "loss": 0.0176,
+      "step": 1884
+    },
+    {
+      "epoch": 0.3209056860742254,
+      "grad_norm": 1.0741052627563477,
+      "learning_rate": 1e-06,
+      "loss": 0.0174,
+      "step": 1885
+    },
+    {
+      "epoch": 0.32107592781750083,
+      "grad_norm": 1.5076146125793457,
+      "learning_rate": 1e-06,
+      "loss": 0.0187,
+      "step": 1886
+    },
+    {
+      "epoch": 0.3212461695607763,
+      "grad_norm": 1.420843243598938,
+      "learning_rate": 1e-06,
+      "loss": 0.0247,
+      "step": 1887
+    },
+    {
+      "epoch": 0.32141641130405174,
+      "grad_norm": 1.4577027559280396,
+      "learning_rate": 1e-06,
+      "loss": 0.019,
+      "step": 1888
+    },
+    {
+      "epoch": 0.3215866530473272,
+      "grad_norm": 1.7815074920654297,
+      "learning_rate": 1e-06,
+      "loss": 0.0256,
+      "step": 1889
+    },
+    {
+      "epoch": 0.32175689479060265,
+      "grad_norm": 1.2624595165252686,
+      "learning_rate": 1e-06,
+      "loss": 0.023,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3219271365338781,
+      "grad_norm": 1.2014528512954712,
+      "learning_rate": 1e-06,
+      "loss": 0.0191,
+      "step": 1891
+    },
+    {
+      "epoch": 0.32209737827715357,
+      "grad_norm": 1.948778510093689,
+      "learning_rate": 1e-06,
+      "loss": 0.0239,
+      "step": 1892
+    },
+    {
+      "epoch": 0.322267620020429,
+      "grad_norm": 1.3787339925765991,
+      "learning_rate": 1e-06,
+      "loss": 0.0231,
+      "step": 1893
+    },
+    {
+      "epoch": 0.3224378617637045,
+      "grad_norm": 1.5238713026046753,
+      "learning_rate": 1e-06,
+      "loss": 0.0195,
+      "step": 1894
+    },
+    {
+      "epoch": 0.32260810350697994,
+      "grad_norm": 1.623833417892456,
+      "learning_rate": 1e-06,
+      "loss": 0.0222,
+      "step": 1895
+    },
+    {
+      "epoch": 0.32277834525025534,
+      "grad_norm": 1.7165459394454956,
+      "learning_rate": 1e-06,
+      "loss": 0.0294,
+      "step": 1896
+    },
+    {
+      "epoch": 0.3229485869935308,
+      "grad_norm": 1.7165459394454956,
+      "learning_rate": 1e-06,
+      "loss": 0.0261,
+      "step": 1897
+    },
+    {
+      "epoch": 0.32311882873680625,
+      "grad_norm": 1.6914831399917603,
+      "learning_rate": 1e-06,
+      "loss": 0.0212,
+      "step": 1898
+    },
+    {
+      "epoch": 0.3232890704800817,
+      "grad_norm": 1.4278634786605835,
+      "learning_rate": 1e-06,
+      "loss": 0.0173,
+      "step": 1899
+    },
+    {
+      "epoch": 0.32345931222335716,
+      "grad_norm": 1.3899396657943726,
+      "learning_rate": 1e-06,
+      "loss": 0.0216,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3236295539666326,
+      "grad_norm": 1.8707932233810425,
+      "learning_rate": 1e-06,
+      "loss": 0.0296,
+      "step": 1901
+    },
+    {
+      "epoch": 0.3237997957099081,
+      "grad_norm": 1.5746642351150513,
+      "learning_rate": 1e-06,
+      "loss": 0.0202,
+      "step": 1902
+    },
+    {
+      "epoch": 0.32397003745318353,
+      "grad_norm": 1.5746642351150513,
+      "learning_rate": 1e-06,
+      "loss": 0.0459,
+      "step": 1903
+    },
+    {
+      "epoch": 0.324140279196459,
+      "grad_norm": 1.444451928138733,
+      "learning_rate": 1e-06,
+      "loss": 0.0239,
+      "step": 1904
+    },
+    {
+      "epoch": 0.32431052093973445,
+      "grad_norm": 1.6318027973175049,
+      "learning_rate": 1e-06,
+      "loss": 0.0168,
+      "step": 1905
+    },
+    {
+      "epoch": 0.32448076268300985,
+      "grad_norm": 1.5491737127304077,
+      "learning_rate": 1e-06,
+      "loss": 0.0229,
+      "step": 1906
+    },
+    {
+      "epoch": 0.3246510044262853,
+      "grad_norm": 1.1921838521957397,
+      "learning_rate": 1e-06,
+      "loss": 0.0217,
+      "step": 1907
+    },
+    {
+      "epoch": 0.32482124616956076,
+      "grad_norm": 1.1826146841049194,
+      "learning_rate": 1e-06,
+      "loss": 0.0153,
+      "step": 1908
+    },
+    {
+      "epoch": 0.3249914879128362,
+      "grad_norm": 1.4094597101211548,
+      "learning_rate": 1e-06,
+      "loss": 0.019,
+      "step": 1909
+    },
+    {
+      "epoch": 0.3251617296561117,
+      "grad_norm": 1.6654062271118164,
+      "learning_rate": 1e-06,
+      "loss": 0.0239,
+      "step": 1910
+    },
+    {
+      "epoch": 0.32533197139938713,
+      "grad_norm": 1.497878909111023,
+      "learning_rate": 1e-06,
+      "loss": 0.0234,
+      "step": 1911
+    },
+    {
+      "epoch": 0.3255022131426626,
+      "grad_norm": 1.8206701278686523,
+      "learning_rate": 1e-06,
+      "loss": 0.0277,
+      "step": 1912
+    },
+    {
+      "epoch": 0.32567245488593805,
+      "grad_norm": 1.2228989601135254,
+      "learning_rate": 1e-06,
+      "loss": 0.017,
+      "step": 1913
+    },
+    {
+      "epoch": 0.3258426966292135,
+      "grad_norm": 1.4604421854019165,
+      "learning_rate": 1e-06,
+      "loss": 0.0266,
+      "step": 1914
+    },
+    {
+      "epoch": 0.32601293837248896,
+      "grad_norm": 1.377022624015808,
+      "learning_rate": 1e-06,
+      "loss": 0.0206,
+      "step": 1915
+    },
+    {
+      "epoch": 0.32618318011576436,
+      "grad_norm": 3.5428032875061035,
+      "learning_rate": 1e-06,
+      "loss": 0.0489,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3263534218590398,
+      "grad_norm": 1.4174351692199707,
+      "learning_rate": 1e-06,
+      "loss": 0.0177,
+      "step": 1917
+    },
+    {
+      "epoch": 0.3265236636023153,
+      "grad_norm": 1.2907001972198486,
+      "learning_rate": 1e-06,
+      "loss": 0.0171,
+      "step": 1918
+    },
+    {
+      "epoch": 0.32669390534559073,
+      "grad_norm": 1.2209053039550781,
+      "learning_rate": 1e-06,
+      "loss": 0.019,
+      "step": 1919
+    },
+    {
+      "epoch": 0.3268641470888662,
+      "grad_norm": 1.1538865566253662,
+      "learning_rate": 1e-06,
+      "loss": 0.0144,
+      "step": 1920
+    },
+    {
+      "epoch": 0.32703438883214164,
+      "grad_norm": 1.5341880321502686,
+      "learning_rate": 1e-06,
+      "loss": 0.019,
+      "step": 1921
+    },
+    {
+      "epoch": 0.3272046305754171,
+      "grad_norm": 1.3872793912887573,
+      "learning_rate": 1e-06,
+      "loss": 0.0181,
+      "step": 1922
+    },
+    {
+      "epoch": 0.32737487231869256,
+      "grad_norm": 1.6427528858184814,
+      "learning_rate": 1e-06,
+      "loss": 0.0374,
+      "step": 1923
+    },
+    {
+      "epoch": 0.327545114061968,
+      "grad_norm": 1.2254210710525513,
+      "learning_rate": 1e-06,
+      "loss": 0.0155,
+      "step": 1924
+    },
+    {
+      "epoch": 0.32771535580524347,
+      "grad_norm": 1.4030553102493286,
+      "learning_rate": 1e-06,
+      "loss": 0.0207,
+      "step": 1925
+    },
+    {
+      "epoch": 0.32788559754851887,
+      "grad_norm": 1.9961392879486084,
+      "learning_rate": 1e-06,
+      "loss": 0.0262,
+      "step": 1926
+    },
+    {
+      "epoch": 0.3280558392917943,
+      "grad_norm": 1.2701325416564941,
+      "learning_rate": 1e-06,
+      "loss": 0.0194,
+      "step": 1927
+    },
+    {
+      "epoch": 0.3282260810350698,
+      "grad_norm": 1.761595606803894,
+      "learning_rate": 1e-06,
+      "loss": 0.0242,
+      "step": 1928
+    },
+    {
+      "epoch": 0.32839632277834524,
+      "grad_norm": 1.5316126346588135,
+      "learning_rate": 1e-06,
+      "loss": 0.0224,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3285665645216207,
+      "grad_norm": 1.808524489402771,
+      "learning_rate": 1e-06,
+      "loss": 0.0289,
+      "step": 1930
+    },
+    {
+      "epoch": 0.32873680626489615,
+      "grad_norm": 1.5288134813308716,
+      "learning_rate": 1e-06,
+      "loss": 0.023,
+      "step": 1931
+    },
+    {
+      "epoch": 0.3289070480081716,
+      "grad_norm": 1.089560627937317,
+      "learning_rate": 1e-06,
+      "loss": 0.0148,
+      "step": 1932
+    },
+    {
+      "epoch": 0.32907728975144707,
+      "grad_norm": 1.5515779256820679,
+      "learning_rate": 1e-06,
+      "loss": 0.0187,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3292475314947225,
+      "grad_norm": 1.5481963157653809,
+      "learning_rate": 1e-06,
+      "loss": 0.0249,
+      "step": 1934
+    },
+    {
+      "epoch": 0.329417773237998,
+      "grad_norm": 1.1329618692398071,
+      "learning_rate": 1e-06,
+      "loss": 0.0175,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3295880149812734,
+      "grad_norm": 1.2369213104248047,
+      "learning_rate": 1e-06,
+      "loss": 0.0196,
+      "step": 1936
+    },
+    {
+      "epoch": 0.32975825672454884,
+      "grad_norm": 1.7791813611984253,
+      "learning_rate": 1e-06,
+      "loss": 0.028,
+      "step": 1937
+    },
+    {
+      "epoch": 0.3299284984678243,
+      "grad_norm": 1.6558544635772705,
+      "learning_rate": 1e-06,
+      "loss": 0.0208,
+      "step": 1938
+    },
+    {
+      "epoch": 0.33009874021109975,
+      "grad_norm": 1.5044019222259521,
+      "learning_rate": 1e-06,
+      "loss": 0.0194,
+      "step": 1939
+    },
+    {
+      "epoch": 0.3302689819543752,
+      "grad_norm": 1.4375008344650269,
+      "learning_rate": 1e-06,
+      "loss": 0.021,
+      "step": 1940
+    },
+    {
+      "epoch": 0.33043922369765066,
+      "grad_norm": 1.3709217309951782,
+      "learning_rate": 1e-06,
+      "loss": 0.0155,
+      "step": 1941
+    },
+    {
+      "epoch": 0.3306094654409261,
+      "grad_norm": 1.812625765800476,
+      "learning_rate": 1e-06,
+      "loss": 0.0261,
+      "step": 1942
+    },
+    {
+      "epoch": 0.3307797071842016,
+      "grad_norm": 1.2849189043045044,
+      "learning_rate": 1e-06,
+      "loss": 0.0187,
+      "step": 1943
+    },
+    {
+      "epoch": 0.33094994892747703,
+      "grad_norm": 1.6139867305755615,
+      "learning_rate": 1e-06,
+      "loss": 0.0269,
+      "step": 1944
+    },
+    {
+      "epoch": 0.3311201906707525,
+      "grad_norm": 1.5526776313781738,
+      "learning_rate": 1e-06,
+      "loss": 0.0236,
+      "step": 1945
+    },
+    {
+      "epoch": 0.33129043241402795,
+      "grad_norm": 1.4651141166687012,
+      "learning_rate": 1e-06,
+      "loss": 0.0247,
+      "step": 1946
+    },
+    {
+      "epoch": 0.33146067415730335,
+      "grad_norm": 2.1171090602874756,
+      "learning_rate": 1e-06,
+      "loss": 0.0345,
+      "step": 1947
+    },
+    {
+      "epoch": 0.3316309159005788,
+      "grad_norm": 1.532902717590332,
+      "learning_rate": 1e-06,
+      "loss": 0.0234,
+      "step": 1948
+    },
+    {
+      "epoch": 0.33180115764385426,
+      "grad_norm": 1.4364941120147705,
+      "learning_rate": 1e-06,
+      "loss": 0.0224,
+      "step": 1949
+    },
+    {
+      "epoch": 0.3319713993871297,
+      "grad_norm": 1.2812119722366333,
+      "learning_rate": 1e-06,
+      "loss": 0.0178,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3321416411304052,
+      "grad_norm": 1.3794339895248413,
+      "learning_rate": 1e-06,
+      "loss": 0.02,
+      "step": 1951
+    },
+    {
+      "epoch": 0.33231188287368063,
+      "grad_norm": 1.5946048498153687,
+      "learning_rate": 1e-06,
+      "loss": 0.0179,
+      "step": 1952
+    },
+    {
+      "epoch": 0.3324821246169561,
+      "grad_norm": 2.0443038940429688,
+      "learning_rate": 1e-06,
+      "loss": 0.0215,
+      "step": 1953
+    },
+    {
+      "epoch": 0.33265236636023154,
+      "grad_norm": 1.368597149848938,
+      "learning_rate": 1e-06,
+      "loss": 0.0158,
+      "step": 1954
+    },
+    {
+      "epoch": 0.332822608103507,
+      "grad_norm": 1.441415786743164,
+      "learning_rate": 1e-06,
+      "loss": 0.0197,
+      "step": 1955
+    },
+    {
+      "epoch": 0.33299284984678246,
+      "grad_norm": 1.6092345714569092,
+      "learning_rate": 1e-06,
+      "loss": 0.025,
+      "step": 1956
+    },
+    {
+      "epoch": 0.33316309159005786,
+      "grad_norm": 1.4012231826782227,
+      "learning_rate": 1e-06,
+      "loss": 0.0217,
+      "step": 1957
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 1.6141374111175537,
+      "learning_rate": 1e-06,
+      "loss": 0.0241,
+      "step": 1958
+    },
+    {
+      "epoch": 0.33350357507660877,
+      "grad_norm": 1.4171791076660156,
+      "learning_rate": 1e-06,
+      "loss": 0.0178,
+      "step": 1959
+    },
+    {
+      "epoch": 0.33367381681988423,
+      "grad_norm": 1.7821341753005981,
+      "learning_rate": 1e-06,
+      "loss": 0.0288,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3338440585631597,
+      "grad_norm": 1.6299926042556763,
+      "learning_rate": 1e-06,
+      "loss": 0.0269,
+      "step": 1961
+    },
+    {
+      "epoch": 0.33401430030643514,
+      "grad_norm": 1.5912959575653076,
+      "learning_rate": 1e-06,
+      "loss": 0.0196,
+      "step": 1962
+    },
+    {
+      "epoch": 0.3341845420497106,
+      "grad_norm": 1.282433271408081,
+      "learning_rate": 1e-06,
+      "loss": 0.0169,
+      "step": 1963
+    },
+    {
+      "epoch": 0.33435478379298605,
+      "grad_norm": 1.5464577674865723,
+      "learning_rate": 1e-06,
+      "loss": 0.027,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3345250255362615,
+      "grad_norm": 1.4779020547866821,
+      "learning_rate": 1e-06,
+      "loss": 0.0301,
+      "step": 1965
+    },
+    {
+      "epoch": 0.33469526727953697,
+      "grad_norm": 1.4697922468185425,
+      "learning_rate": 1e-06,
+      "loss": 0.0234,
+      "step": 1966
+    },
+    {
+      "epoch": 0.33486550902281237,
+      "grad_norm": 1.5022433996200562,
+      "learning_rate": 1e-06,
+      "loss": 0.0189,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3350357507660878,
+      "grad_norm": 1.5426301956176758,
+      "learning_rate": 1e-06,
+      "loss": 0.0221,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3352059925093633,
+      "grad_norm": 1.9386368989944458,
+      "learning_rate": 1e-06,
+      "loss": 0.0269,
+      "step": 1969
+    },
+    {
+      "epoch": 0.33537623425263874,
+      "grad_norm": 1.8475077152252197,
+      "learning_rate": 1e-06,
+      "loss": 0.0194,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3355464759959142,
+      "grad_norm": 1.6919015645980835,
+      "learning_rate": 1e-06,
+      "loss": 0.0259,
+      "step": 1971
+    },
+    {
+      "epoch": 0.33571671773918965,
+      "grad_norm": 1.3801902532577515,
+      "learning_rate": 1e-06,
+      "loss": 0.0143,
+      "step": 1972
+    },
+    {
+      "epoch": 0.3358869594824651,
+      "grad_norm": 1.8841779232025146,
+      "learning_rate": 1e-06,
+      "loss": 0.0382,
+      "step": 1973
+    },
+    {
+      "epoch": 0.33605720122574056,
+      "grad_norm": 1.4700556993484497,
+      "learning_rate": 1e-06,
+      "loss": 0.0222,
+      "step": 1974
+    },
+    {
+      "epoch": 0.336227442969016,
+      "grad_norm": 1.4399062395095825,
+      "learning_rate": 1e-06,
+      "loss": 0.0233,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3363976847122915,
+      "grad_norm": 1.2615573406219482,
+      "learning_rate": 1e-06,
+      "loss": 0.0149,
+      "step": 1976
+    },
+    {
+      "epoch": 0.3365679264555669,
+      "grad_norm": 1.5497113466262817,
+      "learning_rate": 1e-06,
+      "loss": 0.0234,
+      "step": 1977
+    },
+    {
+      "epoch": 0.33673816819884234,
+      "grad_norm": 1.5951131582260132,
+      "learning_rate": 1e-06,
+      "loss": 0.0189,
+      "step": 1978
+    },
+    {
+      "epoch": 0.3369084099421178,
+      "grad_norm": 1.7845149040222168,
+      "learning_rate": 1e-06,
+      "loss": 0.026,
+      "step": 1979
+    },
+    {
+      "epoch": 0.33707865168539325,
+      "grad_norm": 1.608124017715454,
+      "learning_rate": 1e-06,
+      "loss": 0.0247,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3372488934286687,
+      "grad_norm": 1.1790499687194824,
+      "learning_rate": 1e-06,
+      "loss": 0.0163,
+      "step": 1981
+    },
+    {
+      "epoch": 0.33741913517194416,
+      "grad_norm": 1.192211627960205,
+      "learning_rate": 1e-06,
+      "loss": 0.0215,
+      "step": 1982
+    },
+    {
+      "epoch": 0.3375893769152196,
+      "grad_norm": 1.3114814758300781,
+      "learning_rate": 1e-06,
+      "loss": 0.0229,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3377596186584951,
+      "grad_norm": 1.1542383432388306,
+      "learning_rate": 1e-06,
+      "loss": 0.0124,
+      "step": 1984
+    },
+    {
+      "epoch": 0.33792986040177053,
+      "grad_norm": 1.2791386842727661,
+      "learning_rate": 1e-06,
+      "loss": 0.0155,
+      "step": 1985
+    },
+    {
+      "epoch": 0.338100102145046,
+      "grad_norm": 1.5622920989990234,
+      "learning_rate": 1e-06,
+      "loss": 0.0191,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3382703438883214,
+      "grad_norm": 1.3029829263687134,
+      "learning_rate": 1e-06,
+      "loss": 0.0188,
+      "step": 1987
+    },
+    {
+      "epoch": 0.33844058563159685,
+      "grad_norm": 1.5938515663146973,
+      "learning_rate": 1e-06,
+      "loss": 0.0261,
+      "step": 1988
+    },
+    {
+      "epoch": 0.3386108273748723,
+      "grad_norm": 1.4531913995742798,
+      "learning_rate": 1e-06,
+      "loss": 0.0189,
+      "step": 1989
+    },
+    {
+      "epoch": 0.33878106911814776,
+      "grad_norm": 1.8602267503738403,
+      "learning_rate": 1e-06,
+      "loss": 0.021,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3389513108614232,
+      "grad_norm": 1.388229489326477,
+      "learning_rate": 1e-06,
+      "loss": 0.0118,
+      "step": 1991
+    },
+    {
+      "epoch": 0.3391215526046987,
+      "grad_norm": 1.3557769060134888,
+      "learning_rate": 1e-06,
+      "loss": 0.0178,
+      "step": 1992
+    },
+    {
+      "epoch": 0.33929179434797413,
+      "grad_norm": 1.5094513893127441,
+      "learning_rate": 1e-06,
+      "loss": 0.0186,
+      "step": 1993
+    },
+    {
+      "epoch": 0.3394620360912496,
+      "grad_norm": 1.5209040641784668,
+      "learning_rate": 1e-06,
+      "loss": 0.0234,
+      "step": 1994
+    },
+    {
+      "epoch": 0.33963227783452504,
+      "grad_norm": 2.0958139896392822,
+      "learning_rate": 1e-06,
+      "loss": 0.024,
+      "step": 1995
+    },
+    {
+      "epoch": 0.3398025195778005,
+      "grad_norm": 1.7948811054229736,
+      "learning_rate": 1e-06,
+      "loss": 0.0195,
+      "step": 1996
+    },
+    {
+      "epoch": 0.3399727613210759,
+      "grad_norm": 1.6881929636001587,
+      "learning_rate": 1e-06,
+      "loss": 0.0259,
+      "step": 1997
+    },
+    {
+      "epoch": 0.34014300306435136,
+      "grad_norm": 1.1193768978118896,
+      "learning_rate": 1e-06,
+      "loss": 0.0157,
+      "step": 1998
+    },
+    {
+      "epoch": 0.3403132448076268,
+      "grad_norm": 2.3566431999206543,
+      "learning_rate": 1e-06,
+      "loss": 0.0536,
+      "step": 1999
+    },
+    {
+      "epoch": 0.34048348655090227,
+      "grad_norm": 1.4195446968078613,
+      "learning_rate": 1e-06,
+      "loss": 0.0161,
+      "step": 2000
+    },
+    {
+      "epoch": 0.34048348655090227,
+      "eval_loss": 0.27453577518463135,
+      "eval_runtime": 21.0487,
+      "eval_samples_per_second": 14.253,
+      "eval_steps_per_second": 0.38,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3406537282941777,
+      "grad_norm": 1.61619234085083,
+      "learning_rate": 1e-06,
+      "loss": 0.0193,
+      "step": 2001
+    },
+    {
+      "epoch": 0.3408239700374532,
+      "grad_norm": 1.2892980575561523,
+      "learning_rate": 1e-06,
+      "loss": 0.0173,
+      "step": 2002
+    },
+    {
+      "epoch": 0.34099421178072864,
+      "grad_norm": 1.3476999998092651,
+      "learning_rate": 1e-06,
+      "loss": 0.0194,
+      "step": 2003
+    },
+    {
+      "epoch": 0.3411644535240041,
+      "grad_norm": 1.6146029233932495,
+      "learning_rate": 1e-06,
+      "loss": 0.0154,
+      "step": 2004
+    },
+    {
+      "epoch": 0.34133469526727955,
+      "grad_norm": 1.2237156629562378,
+      "learning_rate": 1e-06,
+      "loss": 0.0159,
+      "step": 2005
+    },
+    {
+      "epoch": 0.341504937010555,
+      "grad_norm": 1.6599453687667847,
+      "learning_rate": 1e-06,
+      "loss": 0.0228,
+      "step": 2006
+    },
+    {
+      "epoch": 0.34167517875383047,
+      "grad_norm": 4.322335720062256,
+      "learning_rate": 1e-06,
+      "loss": 0.0276,
+      "step": 2007
+    },
+    {
+      "epoch": 0.34184542049710587,
+      "grad_norm": 1.4886971712112427,
+      "learning_rate": 1e-06,
+      "loss": 0.0307,
+      "step": 2008
+    },
+    {
+      "epoch": 0.3420156622403813,
+      "grad_norm": 1.504891037940979,
+      "learning_rate": 1e-06,
+      "loss": 0.0144,
+      "step": 2009
+    },
+    {
+      "epoch": 0.3421859039836568,
+      "grad_norm": 1.3250677585601807,
+      "learning_rate": 1e-06,
+      "loss": 0.0178,
+      "step": 2010
+    },
+    {
+      "epoch": 0.34235614572693224,
+      "grad_norm": 1.3704164028167725,
+      "learning_rate": 1e-06,
+      "loss": 0.0175,
+      "step": 2011
+    },
+    {
+      "epoch": 0.3425263874702077,
+      "grad_norm": 1.270844578742981,
+      "learning_rate": 1e-06,
+      "loss": 0.0188,
+      "step": 2012
+    },
+    {
+      "epoch": 0.34269662921348315,
+      "grad_norm": 1.2841618061065674,
+      "learning_rate": 1e-06,
+      "loss": 0.0162,
+      "step": 2013
+    },
+    {
+      "epoch": 0.3428668709567586,
+      "grad_norm": 2.089329242706299,
+      "learning_rate": 1e-06,
+      "loss": 0.0362,
+      "step": 2014
+    },
+    {
+      "epoch": 0.34303711270003406,
+      "grad_norm": 2.0143887996673584,
+      "learning_rate": 1e-06,
+      "loss": 0.0259,
+      "step": 2015
+    },
+    {
+      "epoch": 0.3432073544433095,
+      "grad_norm": 1.7846940755844116,
+      "learning_rate": 1e-06,
+      "loss": 0.0268,
+      "step": 2016
+    },
+    {
+      "epoch": 0.343377596186585,
+      "grad_norm": 1.5742454528808594,
+      "learning_rate": 1e-06,
+      "loss": 0.0262,
+      "step": 2017
+    },
+    {
+      "epoch": 0.3435478379298604,
+      "grad_norm": 1.1249366998672485,
+      "learning_rate": 1e-06,
+      "loss": 0.0133,
+      "step": 2018
+    },
+    {
+      "epoch": 0.34371807967313583,
+      "grad_norm": 1.3523153066635132,
+      "learning_rate": 1e-06,
+      "loss": 0.0234,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3438883214164113,
+      "grad_norm": 1.611651062965393,
+      "learning_rate": 1e-06,
+      "loss": 0.0179,
+      "step": 2020
+    },
+    {
+      "epoch": 0.34405856315968675,
+      "grad_norm": 1.476097583770752,
+      "learning_rate": 1e-06,
+      "loss": 0.017,
+      "step": 2021
+    },
+    {
+      "epoch": 0.3442288049029622,
+      "grad_norm": 1.2952362298965454,
+      "learning_rate": 1e-06,
+      "loss": 0.0255,
+      "step": 2022
+    },
+    {
+      "epoch": 0.34439904664623766,
+      "grad_norm": 1.3099634647369385,
+      "learning_rate": 1e-06,
+      "loss": 0.0153,
+      "step": 2023
+    },
+    {
+      "epoch": 0.3445692883895131,
+      "grad_norm": 1.4053715467453003,
+      "learning_rate": 1e-06,
+      "loss": 0.0159,
+      "step": 2024
+    },
+    {
+      "epoch": 0.3447395301327886,
+      "grad_norm": 1.7385302782058716,
+      "learning_rate": 1e-06,
+      "loss": 0.022,
+      "step": 2025
+    },
+    {
+      "epoch": 0.34490977187606403,
+      "grad_norm": 1.365162968635559,
+      "learning_rate": 1e-06,
+      "loss": 0.0171,
+      "step": 2026
+    },
+    {
+      "epoch": 0.3450800136193395,
+      "grad_norm": 1.7785168886184692,
+      "learning_rate": 1e-06,
+      "loss": 0.0202,
+      "step": 2027
+    },
+    {
+      "epoch": 0.3452502553626149,
+      "grad_norm": 1.1140302419662476,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 2028
+    },
+    {
+      "epoch": 0.34542049710589035,
+      "grad_norm": 1.29301917552948,
+      "learning_rate": 1e-06,
+      "loss": 0.0212,
+      "step": 2029
+    },
+    {
+      "epoch": 0.3455907388491658,
+      "grad_norm": 1.9448950290679932,
+      "learning_rate": 1e-06,
+      "loss": 0.0285,
+      "step": 2030
+    },
+    {
+      "epoch": 0.34576098059244126,
+      "grad_norm": 1.5493046045303345,
+      "learning_rate": 1e-06,
+      "loss": 0.0268,
+      "step": 2031
+    },
+    {
+      "epoch": 0.3459312223357167,
+      "grad_norm": 1.561806321144104,
+      "learning_rate": 1e-06,
+      "loss": 0.0258,
+      "step": 2032
+    },
+    {
+      "epoch": 0.34610146407899217,
+      "grad_norm": 1.5834723711013794,
+      "learning_rate": 1e-06,
+      "loss": 0.0336,
+      "step": 2033
+    },
+    {
+      "epoch": 0.34627170582226763,
+      "grad_norm": 1.1366455554962158,
+      "learning_rate": 1e-06,
+      "loss": 0.0182,
+      "step": 2034
+    },
+    {
+      "epoch": 0.3464419475655431,
+      "grad_norm": 1.152408242225647,
+      "learning_rate": 1e-06,
+      "loss": 0.0166,
+      "step": 2035
+    },
+    {
+      "epoch": 0.34661218930881854,
+      "grad_norm": 1.2690989971160889,
+      "learning_rate": 1e-06,
+      "loss": 0.014,
+      "step": 2036
+    },
+    {
+      "epoch": 0.346782431052094,
+      "grad_norm": 1.0005087852478027,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 2037
+    },
+    {
+      "epoch": 0.3469526727953694,
+      "grad_norm": 1.1189073324203491,
+      "learning_rate": 1e-06,
+      "loss": 0.0143,
+      "step": 2038
+    },
+    {
+      "epoch": 0.34712291453864486,
+      "grad_norm": 1.6101375818252563,
+      "learning_rate": 1e-06,
+      "loss": 0.024,
+      "step": 2039
+    },
+    {
+      "epoch": 0.3472931562819203,
+      "grad_norm": 1.4370726346969604,
+      "learning_rate": 1e-06,
+      "loss": 0.0193,
+      "step": 2040
+    },
+    {
+      "epoch": 0.34746339802519577,
+      "grad_norm": 1.6193634271621704,
+      "learning_rate": 1e-06,
+      "loss": 0.0249,
+      "step": 2041
+    },
+    {
+      "epoch": 0.3476336397684712,
+      "grad_norm": 1.6027393341064453,
+      "learning_rate": 1e-06,
+      "loss": 0.0178,
+      "step": 2042
+    },
+    {
+      "epoch": 0.3478038815117467,
+      "grad_norm": 1.3268249034881592,
+      "learning_rate": 1e-06,
+      "loss": 0.0194,
+      "step": 2043
+    },
+    {
+      "epoch": 0.34797412325502214,
+      "grad_norm": 1.6447010040283203,
+      "learning_rate": 1e-06,
+      "loss": 0.016,
+      "step": 2044
+    },
+    {
+      "epoch": 0.3481443649982976,
+      "grad_norm": 1.3561208248138428,
+      "learning_rate": 1e-06,
+      "loss": 0.0166,
+      "step": 2045
+    },
+    {
+      "epoch": 0.34831460674157305,
+      "grad_norm": 1.494844913482666,
+      "learning_rate": 1e-06,
+      "loss": 0.0288,
+      "step": 2046
+    },
+    {
+      "epoch": 0.3484848484848485,
+      "grad_norm": 1.2338416576385498,
+      "learning_rate": 1e-06,
+      "loss": 0.0196,
+      "step": 2047
+    },
+    {
+      "epoch": 0.3486550902281239,
+      "grad_norm": 1.3532445430755615,
+      "learning_rate": 1e-06,
+      "loss": 0.0208,
+      "step": 2048
+    },
+    {
+      "epoch": 0.34882533197139937,
+      "grad_norm": 1.3149325847625732,
+      "learning_rate": 1e-06,
+      "loss": 0.02,
+      "step": 2049
+    },
+    {
+      "epoch": 0.3489955737146748,
+      "grad_norm": 1.2256778478622437,
+      "learning_rate": 1e-06,
+      "loss": 0.0188,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3491658154579503,
+      "grad_norm": 1.6831778287887573,
+      "learning_rate": 1e-06,
+      "loss": 0.0164,
+      "step": 2051
+    },
+    {
+      "epoch": 0.34933605720122574,
+      "grad_norm": 1.2866861820220947,
+      "learning_rate": 1e-06,
+      "loss": 0.0197,
+      "step": 2052
+    },
+    {
+      "epoch": 0.3495062989445012,
+      "grad_norm": 1.3010034561157227,
+      "learning_rate": 1e-06,
+      "loss": 0.0218,
+      "step": 2053
+    },
+    {
+      "epoch": 0.34967654068777665,
+      "grad_norm": 1.460037350654602,
+      "learning_rate": 1e-06,
+      "loss": 0.021,
+      "step": 2054
+    },
+    {
+      "epoch": 0.3498467824310521,
+      "grad_norm": 1.2930116653442383,
+      "learning_rate": 1e-06,
+      "loss": 0.0189,
+      "step": 2055
+    },
+    {
+      "epoch": 0.35001702417432756,
+      "grad_norm": 1.4779536724090576,
+      "learning_rate": 1e-06,
+      "loss": 0.0195,
+      "step": 2056
+    },
+    {
+      "epoch": 0.350187265917603,
+      "grad_norm": 1.1668182611465454,
+      "learning_rate": 1e-06,
+      "loss": 0.0168,
+      "step": 2057
+    },
+    {
+      "epoch": 0.3503575076608784,
+      "grad_norm": 1.5473500490188599,
+      "learning_rate": 1e-06,
+      "loss": 0.0245,
+      "step": 2058
+    },
+    {
+      "epoch": 0.3505277494041539,
+      "grad_norm": 1.5026984214782715,
+      "learning_rate": 1e-06,
+      "loss": 0.0263,
+      "step": 2059
+    },
+    {
+      "epoch": 0.35069799114742933,
+      "grad_norm": 1.4979209899902344,
+      "learning_rate": 1e-06,
+      "loss": 0.0192,
+      "step": 2060
+    },
+    {
+      "epoch": 0.3508682328907048,
+      "grad_norm": 1.6945948600769043,
+      "learning_rate": 1e-06,
+      "loss": 0.0215,
+      "step": 2061
+    },
+    {
+      "epoch": 0.35103847463398025,
+      "grad_norm": 1.4094072580337524,
+      "learning_rate": 1e-06,
+      "loss": 0.0159,
+      "step": 2062
+    },
+    {
+      "epoch": 0.3512087163772557,
+      "grad_norm": 1.5919791460037231,
+      "learning_rate": 1e-06,
+      "loss": 0.017,
+      "step": 2063
+    },
+    {
+      "epoch": 0.35137895812053116,
+      "grad_norm": 1.1257418394088745,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 2064
+    },
+    {
+      "epoch": 0.3515491998638066,
+      "grad_norm": 1.4687309265136719,
+      "learning_rate": 1e-06,
+      "loss": 0.0161,
+      "step": 2065
+    },
+    {
+      "epoch": 0.3517194416070821,
+      "grad_norm": 1.5042623281478882,
+      "learning_rate": 1e-06,
+      "loss": 0.0206,
+      "step": 2066
+    },
+    {
+      "epoch": 0.35188968335035753,
+      "grad_norm": 1.5676453113555908,
+      "learning_rate": 1e-06,
+      "loss": 0.0191,
+      "step": 2067
+    },
+    {
+      "epoch": 0.352059925093633,
+      "grad_norm": 1.368834137916565,
+      "learning_rate": 1e-06,
+      "loss": 0.0177,
+      "step": 2068
+    },
+    {
+      "epoch": 0.3522301668369084,
+      "grad_norm": 1.1905148029327393,
+      "learning_rate": 1e-06,
+      "loss": 0.0159,
+      "step": 2069
+    },
+    {
+      "epoch": 0.35240040858018384,
+      "grad_norm": 1.3578230142593384,
+      "learning_rate": 1e-06,
+      "loss": 0.0146,
+      "step": 2070
+    },
+    {
+      "epoch": 0.3525706503234593,
+      "grad_norm": 1.6715102195739746,
+      "learning_rate": 1e-06,
+      "loss": 0.0239,
+      "step": 2071
+    },
+    {
+      "epoch": 0.35274089206673476,
+      "grad_norm": 1.2917587757110596,
+      "learning_rate": 1e-06,
+      "loss": 0.0231,
+      "step": 2072
+    },
+    {
+      "epoch": 0.3529111338100102,
+      "grad_norm": 1.7715479135513306,
+      "learning_rate": 1e-06,
+      "loss": 0.0151,
+      "step": 2073
+    },
+    {
+      "epoch": 0.35308137555328567,
+      "grad_norm": 1.1060619354248047,
+      "learning_rate": 1e-06,
+      "loss": 0.0133,
+      "step": 2074
+    },
+    {
+      "epoch": 0.3532516172965611,
+      "grad_norm": 2.2120845317840576,
+      "learning_rate": 1e-06,
+      "loss": 0.0233,
+      "step": 2075
+    },
+    {
+      "epoch": 0.3534218590398366,
+      "grad_norm": 1.4091529846191406,
+      "learning_rate": 1e-06,
+      "loss": 0.027,
+      "step": 2076
+    },
+    {
+      "epoch": 0.35359210078311204,
+      "grad_norm": 1.3464224338531494,
+      "learning_rate": 1e-06,
+      "loss": 0.0183,
+      "step": 2077
+    },
+    {
+      "epoch": 0.3537623425263875,
+      "grad_norm": 1.6281346082687378,
+      "learning_rate": 1e-06,
+      "loss": 0.0253,
+      "step": 2078
+    },
+    {
+      "epoch": 0.3539325842696629,
+      "grad_norm": 1.2976142168045044,
+      "learning_rate": 1e-06,
+      "loss": 0.0157,
+      "step": 2079
+    },
+    {
+      "epoch": 0.35410282601293835,
+      "grad_norm": 1.4444133043289185,
+      "learning_rate": 1e-06,
+      "loss": 0.0186,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3542730677562138,
+      "grad_norm": 1.5797218084335327,
+      "learning_rate": 1e-06,
+      "loss": 0.0183,
+      "step": 2081
+    },
+    {
+      "epoch": 0.35444330949948927,
+      "grad_norm": 1.4098842144012451,
+      "learning_rate": 1e-06,
+      "loss": 0.0211,
+      "step": 2082
+    },
+    {
+      "epoch": 0.3546135512427647,
+      "grad_norm": 1.34474778175354,
+      "learning_rate": 1e-06,
+      "loss": 0.0163,
+      "step": 2083
+    },
+    {
+      "epoch": 0.3547837929860402,
+      "grad_norm": 1.0983127355575562,
+      "learning_rate": 1e-06,
+      "loss": 0.0163,
+      "step": 2084
+    },
+    {
+      "epoch": 0.35495403472931564,
+      "grad_norm": 1.8528269529342651,
+      "learning_rate": 1e-06,
+      "loss": 0.0211,
+      "step": 2085
+    },
+    {
+      "epoch": 0.3551242764725911,
+      "grad_norm": 1.4813817739486694,
+      "learning_rate": 1e-06,
+      "loss": 0.0148,
+      "step": 2086
+    },
+    {
+      "epoch": 0.35529451821586655,
+      "grad_norm": 1.7108453512191772,
+      "learning_rate": 1e-06,
+      "loss": 0.0221,
+      "step": 2087
+    },
+    {
+      "epoch": 0.355464759959142,
+      "grad_norm": 1.5667725801467896,
+      "learning_rate": 1e-06,
+      "loss": 0.0196,
+      "step": 2088
+    },
+    {
+      "epoch": 0.3556350017024174,
+      "grad_norm": 1.5677913427352905,
+      "learning_rate": 1e-06,
+      "loss": 0.0221,
+      "step": 2089
+    },
+    {
+      "epoch": 0.35580524344569286,
+      "grad_norm": 0.9959020614624023,
+      "learning_rate": 1e-06,
+      "loss": 0.0119,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3559754851889683,
+      "grad_norm": 1.4427545070648193,
+      "learning_rate": 1e-06,
+      "loss": 0.018,
+      "step": 2091
+    },
+    {
+      "epoch": 0.3561457269322438,
+      "grad_norm": 1.162044644355774,
+      "learning_rate": 1e-06,
+      "loss": 0.0193,
+      "step": 2092
+    },
+    {
+      "epoch": 0.35631596867551923,
+      "grad_norm": 1.3956273794174194,
+      "learning_rate": 1e-06,
+      "loss": 0.0219,
+      "step": 2093
+    },
+    {
+      "epoch": 0.3564862104187947,
+      "grad_norm": 1.4350906610488892,
+      "learning_rate": 1e-06,
+      "loss": 0.0204,
+      "step": 2094
+    },
+    {
+      "epoch": 0.35665645216207015,
+      "grad_norm": 1.2975064516067505,
+      "learning_rate": 1e-06,
+      "loss": 0.0147,
+      "step": 2095
+    },
+    {
+      "epoch": 0.3568266939053456,
+      "grad_norm": 2.6082117557525635,
+      "learning_rate": 1e-06,
+      "loss": 0.0259,
+      "step": 2096
+    },
+    {
+      "epoch": 0.35699693564862106,
+      "grad_norm": 1.2373225688934326,
+      "learning_rate": 1e-06,
+      "loss": 0.0158,
+      "step": 2097
+    },
+    {
+      "epoch": 0.3571671773918965,
+      "grad_norm": 1.4951549768447876,
+      "learning_rate": 1e-06,
+      "loss": 0.0154,
+      "step": 2098
+    },
+    {
+      "epoch": 0.3573374191351719,
+      "grad_norm": 1.339722752571106,
+      "learning_rate": 1e-06,
+      "loss": 0.012,
+      "step": 2099
+    },
+    {
+      "epoch": 0.3575076608784474,
+      "grad_norm": 1.1775256395339966,
+      "learning_rate": 1e-06,
+      "loss": 0.0141,
+      "step": 2100
+    },
+    {
+      "epoch": 0.35767790262172283,
+      "grad_norm": 1.7469888925552368,
+      "learning_rate": 1e-06,
+      "loss": 0.0147,
+      "step": 2101
+    },
+    {
+      "epoch": 0.3578481443649983,
+      "grad_norm": 1.4239497184753418,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 2102
+    },
+    {
+      "epoch": 0.35801838610827375,
+      "grad_norm": 1.3292256593704224,
+      "learning_rate": 1e-06,
+      "loss": 0.0162,
+      "step": 2103
+    },
+    {
+      "epoch": 0.3581886278515492,
+      "grad_norm": 1.902808427810669,
+      "learning_rate": 1e-06,
+      "loss": 0.0274,
+      "step": 2104
+    },
+    {
+      "epoch": 0.35835886959482466,
+      "grad_norm": 1.5837531089782715,
+      "learning_rate": 1e-06,
+      "loss": 0.0222,
+      "step": 2105
+    },
+    {
+      "epoch": 0.3585291113381001,
+      "grad_norm": 1.3720051050186157,
+      "learning_rate": 1e-06,
+      "loss": 0.0207,
+      "step": 2106
+    },
+    {
+      "epoch": 0.35869935308137557,
+      "grad_norm": 1.7276686429977417,
+      "learning_rate": 1e-06,
+      "loss": 0.0251,
+      "step": 2107
+    },
+    {
+      "epoch": 0.35886959482465103,
+      "grad_norm": 1.7324738502502441,
+      "learning_rate": 1e-06,
+      "loss": 0.0256,
+      "step": 2108
+    },
+    {
+      "epoch": 0.35903983656792643,
+      "grad_norm": 1.5119861364364624,
+      "learning_rate": 1e-06,
+      "loss": 0.0177,
+      "step": 2109
+    },
+    {
+      "epoch": 0.3592100783112019,
+      "grad_norm": 1.672288417816162,
+      "learning_rate": 1e-06,
+      "loss": 0.0249,
+      "step": 2110
+    },
+    {
+      "epoch": 0.35938032005447734,
+      "grad_norm": 3.63991379737854,
+      "learning_rate": 1e-06,
+      "loss": 0.0522,
+      "step": 2111
+    },
+    {
+      "epoch": 0.3595505617977528,
+      "grad_norm": 1.3657325506210327,
+      "learning_rate": 1e-06,
+      "loss": 0.0194,
+      "step": 2112
+    },
+    {
+      "epoch": 0.35972080354102826,
+      "grad_norm": 1.196782112121582,
+      "learning_rate": 1e-06,
+      "loss": 0.0213,
+      "step": 2113
+    },
+    {
+      "epoch": 0.3598910452843037,
+      "grad_norm": 1.248388648033142,
+      "learning_rate": 1e-06,
+      "loss": 0.0125,
+      "step": 2114
+    },
+    {
+      "epoch": 0.36006128702757917,
+      "grad_norm": 1.267297387123108,
+      "learning_rate": 1e-06,
+      "loss": 0.0182,
+      "step": 2115
+    },
+    {
+      "epoch": 0.3602315287708546,
+      "grad_norm": 1.4637548923492432,
+      "learning_rate": 1e-06,
+      "loss": 0.0212,
+      "step": 2116
+    },
+    {
+      "epoch": 0.3604017705141301,
+      "grad_norm": 1.5229331254959106,
+      "learning_rate": 1e-06,
+      "loss": 0.0254,
+      "step": 2117
+    },
+    {
+      "epoch": 0.36057201225740554,
+      "grad_norm": 1.511386752128601,
+      "learning_rate": 1e-06,
+      "loss": 0.0193,
+      "step": 2118
+    },
+    {
+      "epoch": 0.36074225400068094,
+      "grad_norm": 1.6698894500732422,
+      "learning_rate": 1e-06,
+      "loss": 0.0272,
+      "step": 2119
+    },
+    {
+      "epoch": 0.3609124957439564,
+      "grad_norm": 2.0563039779663086,
+      "learning_rate": 1e-06,
+      "loss": 0.0234,
+      "step": 2120
+    },
+    {
+      "epoch": 0.36108273748723185,
+      "grad_norm": 2.0152013301849365,
+      "learning_rate": 1e-06,
+      "loss": 0.0169,
+      "step": 2121
+    },
+    {
+      "epoch": 0.3612529792305073,
+      "grad_norm": 1.2313683032989502,
+      "learning_rate": 1e-06,
+      "loss": 0.0146,
+      "step": 2122
+    },
+    {
+      "epoch": 0.36142322097378277,
+      "grad_norm": 1.654374361038208,
+      "learning_rate": 1e-06,
+      "loss": 0.0238,
+      "step": 2123
+    },
+    {
+      "epoch": 0.3615934627170582,
+      "grad_norm": 1.3135615587234497,
+      "learning_rate": 1e-06,
+      "loss": 0.013,
+      "step": 2124
+    },
+    {
+      "epoch": 0.3617637044603337,
+      "grad_norm": 1.285529613494873,
+      "learning_rate": 1e-06,
+      "loss": 0.0205,
+      "step": 2125
+    },
+    {
+      "epoch": 0.36193394620360914,
+      "grad_norm": 1.392561912536621,
+      "learning_rate": 1e-06,
+      "loss": 0.0174,
+      "step": 2126
+    },
+    {
+      "epoch": 0.3621041879468846,
+      "grad_norm": 1.8578132390975952,
+      "learning_rate": 1e-06,
+      "loss": 0.0177,
+      "step": 2127
+    },
+    {
+      "epoch": 0.36227442969016005,
+      "grad_norm": 1.368056297302246,
+      "learning_rate": 1e-06,
+      "loss": 0.0215,
+      "step": 2128
+    },
+    {
+      "epoch": 0.3624446714334355,
+      "grad_norm": 1.1833387613296509,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 2129
+    },
+    {
+      "epoch": 0.3626149131767109,
+      "grad_norm": 1.4502012729644775,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 2130
+    },
+    {
+      "epoch": 0.36278515491998636,
+      "grad_norm": 1.522831678390503,
+      "learning_rate": 1e-06,
+      "loss": 0.0261,
+      "step": 2131
+    },
+    {
+      "epoch": 0.3629553966632618,
+      "grad_norm": 1.4958367347717285,
+      "learning_rate": 1e-06,
+      "loss": 0.0168,
+      "step": 2132
+    },
+    {
+      "epoch": 0.3631256384065373,
+      "grad_norm": 1.1259353160858154,
+      "learning_rate": 1e-06,
+      "loss": 0.0168,
+      "step": 2133
+    },
+    {
+      "epoch": 0.36329588014981273,
+      "grad_norm": 2.0761754512786865,
+      "learning_rate": 1e-06,
+      "loss": 0.0311,
+      "step": 2134
+    },
+    {
+      "epoch": 0.3634661218930882,
+      "grad_norm": 1.4105058908462524,
+      "learning_rate": 1e-06,
+      "loss": 0.0211,
+      "step": 2135
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 1.2559282779693604,
+      "learning_rate": 1e-06,
+      "loss": 0.0168,
+      "step": 2136
+    },
+    {
+      "epoch": 0.3638066053796391,
+      "grad_norm": 1.5729042291641235,
+      "learning_rate": 1e-06,
+      "loss": 0.0292,
+      "step": 2137
+    },
+    {
+      "epoch": 0.36397684712291456,
+      "grad_norm": 1.7037521600723267,
+      "learning_rate": 1e-06,
+      "loss": 0.0228,
+      "step": 2138
+    },
+    {
+      "epoch": 0.36414708886619,
+      "grad_norm": 1.3164565563201904,
+      "learning_rate": 1e-06,
+      "loss": 0.0178,
+      "step": 2139
+    },
+    {
+      "epoch": 0.3643173306094654,
+      "grad_norm": 1.6225502490997314,
+      "learning_rate": 1e-06,
+      "loss": 0.0195,
+      "step": 2140
+    },
+    {
+      "epoch": 0.3644875723527409,
+      "grad_norm": 1.2100952863693237,
+      "learning_rate": 1e-06,
+      "loss": 0.0165,
+      "step": 2141
+    },
+    {
+      "epoch": 0.36465781409601633,
+      "grad_norm": 1.2184245586395264,
+      "learning_rate": 1e-06,
+      "loss": 0.0136,
+      "step": 2142
+    },
+    {
+      "epoch": 0.3648280558392918,
+      "grad_norm": 1.6735187768936157,
+      "learning_rate": 1e-06,
+      "loss": 0.0227,
+      "step": 2143
+    },
+    {
+      "epoch": 0.36499829758256724,
+      "grad_norm": 1.3809057474136353,
+      "learning_rate": 1e-06,
+      "loss": 0.0211,
+      "step": 2144
+    },
+    {
+      "epoch": 0.3651685393258427,
+      "grad_norm": 1.2712934017181396,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 2145
+    },
+    {
+      "epoch": 0.36533878106911816,
+      "grad_norm": 1.435024619102478,
+      "learning_rate": 1e-06,
+      "loss": 0.0146,
+      "step": 2146
+    },
+    {
+      "epoch": 0.3655090228123936,
+      "grad_norm": 1.45426344871521,
+      "learning_rate": 1e-06,
+      "loss": 0.0159,
+      "step": 2147
+    },
+    {
+      "epoch": 0.36567926455566907,
+      "grad_norm": 1.9181464910507202,
+      "learning_rate": 1e-06,
+      "loss": 0.0285,
+      "step": 2148
+    },
+    {
+      "epoch": 0.3658495062989445,
+      "grad_norm": 1.4425302743911743,
+      "learning_rate": 1e-06,
+      "loss": 0.0196,
+      "step": 2149
+    },
+    {
+      "epoch": 0.36601974804221993,
+      "grad_norm": 1.8058220148086548,
+      "learning_rate": 1e-06,
+      "loss": 0.0204,
+      "step": 2150
+    },
+    {
+      "epoch": 0.3661899897854954,
+      "grad_norm": 1.3857768774032593,
+      "learning_rate": 1e-06,
+      "loss": 0.0152,
+      "step": 2151
+    },
+    {
+      "epoch": 0.36636023152877084,
+      "grad_norm": 1.4955687522888184,
+      "learning_rate": 1e-06,
+      "loss": 0.0243,
+      "step": 2152
+    },
+    {
+      "epoch": 0.3665304732720463,
+      "grad_norm": 1.6004747152328491,
+      "learning_rate": 1e-06,
+      "loss": 0.0177,
+      "step": 2153
+    },
+    {
+      "epoch": 0.36670071501532175,
+      "grad_norm": 1.3243181705474854,
+      "learning_rate": 1e-06,
+      "loss": 0.0145,
+      "step": 2154
+    },
+    {
+      "epoch": 0.3668709567585972,
+      "grad_norm": 1.8293629884719849,
+      "learning_rate": 1e-06,
+      "loss": 0.0236,
+      "step": 2155
+    },
+    {
+      "epoch": 0.36704119850187267,
+      "grad_norm": 1.2558562755584717,
+      "learning_rate": 1e-06,
+      "loss": 0.0163,
+      "step": 2156
+    },
+    {
+      "epoch": 0.3672114402451481,
+      "grad_norm": 1.2633802890777588,
+      "learning_rate": 1e-06,
+      "loss": 0.0144,
+      "step": 2157
+    },
+    {
+      "epoch": 0.3673816819884236,
+      "grad_norm": 1.4996285438537598,
+      "learning_rate": 1e-06,
+      "loss": 0.0229,
+      "step": 2158
+    },
+    {
+      "epoch": 0.36755192373169904,
+      "grad_norm": 1.1702207326889038,
+      "learning_rate": 1e-06,
+      "loss": 0.0154,
+      "step": 2159
+    },
+    {
+      "epoch": 0.36772216547497444,
+      "grad_norm": 1.2386761903762817,
+      "learning_rate": 1e-06,
+      "loss": 0.0167,
+      "step": 2160
+    },
+    {
+      "epoch": 0.3678924072182499,
+      "grad_norm": 1.7300621271133423,
+      "learning_rate": 1e-06,
+      "loss": 0.0228,
+      "step": 2161
+    },
+    {
+      "epoch": 0.36806264896152535,
+      "grad_norm": 1.3940644264221191,
+      "learning_rate": 1e-06,
+      "loss": 0.022,
+      "step": 2162
+    },
+    {
+      "epoch": 0.3682328907048008,
+      "grad_norm": 1.4714255332946777,
+      "learning_rate": 1e-06,
+      "loss": 0.019,
+      "step": 2163
+    },
+    {
+      "epoch": 0.36840313244807626,
+      "grad_norm": 1.099491834640503,
+      "learning_rate": 1e-06,
+      "loss": 0.012,
+      "step": 2164
+    },
+    {
+      "epoch": 0.3685733741913517,
+      "grad_norm": 1.4436107873916626,
+      "learning_rate": 1e-06,
+      "loss": 0.0184,
+      "step": 2165
+    },
+    {
+      "epoch": 0.3687436159346272,
+      "grad_norm": 1.4421783685684204,
+      "learning_rate": 1e-06,
+      "loss": 0.0158,
+      "step": 2166
+    },
+    {
+      "epoch": 0.36891385767790263,
+      "grad_norm": 1.221912145614624,
+      "learning_rate": 1e-06,
+      "loss": 0.0153,
+      "step": 2167
+    },
+    {
+      "epoch": 0.3690840994211781,
+      "grad_norm": 1.3377691507339478,
+      "learning_rate": 1e-06,
+      "loss": 0.0196,
+      "step": 2168
+    },
+    {
+      "epoch": 0.36925434116445355,
+      "grad_norm": 1.5522762537002563,
+      "learning_rate": 1e-06,
+      "loss": 0.0217,
+      "step": 2169
+    },
+    {
+      "epoch": 0.36942458290772895,
+      "grad_norm": 1.452081561088562,
+      "learning_rate": 1e-06,
+      "loss": 0.0174,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3695948246510044,
+      "grad_norm": 1.293049931526184,
+      "learning_rate": 1e-06,
+      "loss": 0.0131,
+      "step": 2171
+    },
+    {
+      "epoch": 0.36976506639427986,
+      "grad_norm": 1.621479868888855,
+      "learning_rate": 1e-06,
+      "loss": 0.0177,
+      "step": 2172
+    },
+    {
+      "epoch": 0.3699353081375553,
+      "grad_norm": 1.6249638795852661,
+      "learning_rate": 1e-06,
+      "loss": 0.0203,
+      "step": 2173
+    },
+    {
+      "epoch": 0.3701055498808308,
+      "grad_norm": 1.5299254655838013,
+      "learning_rate": 1e-06,
+      "loss": 0.0205,
+      "step": 2174
+    },
+    {
+      "epoch": 0.37027579162410623,
+      "grad_norm": 1.5919313430786133,
+      "learning_rate": 1e-06,
+      "loss": 0.0191,
+      "step": 2175
+    },
+    {
+      "epoch": 0.3704460333673817,
+      "grad_norm": 1.651296615600586,
+      "learning_rate": 1e-06,
+      "loss": 0.0235,
+      "step": 2176
+    },
+    {
+      "epoch": 0.37061627511065715,
+      "grad_norm": 1.3494011163711548,
+      "learning_rate": 1e-06,
+      "loss": 0.0202,
+      "step": 2177
+    },
+    {
+      "epoch": 0.3707865168539326,
+      "grad_norm": 1.4936208724975586,
+      "learning_rate": 1e-06,
+      "loss": 0.0174,
+      "step": 2178
+    },
+    {
+      "epoch": 0.37095675859720806,
+      "grad_norm": 1.3437141180038452,
+      "learning_rate": 1e-06,
+      "loss": 0.0166,
+      "step": 2179
+    },
+    {
+      "epoch": 0.37112700034048346,
+      "grad_norm": 1.137948751449585,
+      "learning_rate": 1e-06,
+      "loss": 0.0134,
+      "step": 2180
+    },
+    {
+      "epoch": 0.3712972420837589,
+      "grad_norm": 1.409578800201416,
+      "learning_rate": 1e-06,
+      "loss": 0.0187,
+      "step": 2181
+    },
+    {
+      "epoch": 0.3714674838270344,
+      "grad_norm": 1.4696379899978638,
+      "learning_rate": 1e-06,
+      "loss": 0.0212,
+      "step": 2182
+    },
+    {
+      "epoch": 0.37163772557030983,
+      "grad_norm": 1.8238104581832886,
+      "learning_rate": 1e-06,
+      "loss": 0.0156,
+      "step": 2183
+    },
+    {
+      "epoch": 0.3718079673135853,
+      "grad_norm": 1.643831729888916,
+      "learning_rate": 1e-06,
+      "loss": 0.0176,
+      "step": 2184
+    },
+    {
+      "epoch": 0.37197820905686074,
+      "grad_norm": 1.536838173866272,
+      "learning_rate": 1e-06,
+      "loss": 0.017,
+      "step": 2185
+    },
+    {
+      "epoch": 0.3721484508001362,
+      "grad_norm": 1.572147011756897,
+      "learning_rate": 1e-06,
+      "loss": 0.0185,
+      "step": 2186
+    },
+    {
+      "epoch": 0.37231869254341166,
+      "grad_norm": 1.0589184761047363,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 2187
+    },
+    {
+      "epoch": 0.3724889342866871,
+      "grad_norm": 2.404597043991089,
+      "learning_rate": 1e-06,
+      "loss": 0.0245,
+      "step": 2188
+    },
+    {
+      "epoch": 0.37265917602996257,
+      "grad_norm": 1.3876649141311646,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 2189
+    },
+    {
+      "epoch": 0.372829417773238,
+      "grad_norm": 1.3182275295257568,
+      "learning_rate": 1e-06,
+      "loss": 0.0153,
+      "step": 2190
+    },
+    {
+      "epoch": 0.3729996595165134,
+      "grad_norm": 1.2899410724639893,
+      "learning_rate": 1e-06,
+      "loss": 0.0233,
+      "step": 2191
+    },
+    {
+      "epoch": 0.3731699012597889,
+      "grad_norm": 1.5908676385879517,
+      "learning_rate": 1e-06,
+      "loss": 0.0156,
+      "step": 2192
+    },
+    {
+      "epoch": 0.37334014300306434,
+      "grad_norm": 1.2450308799743652,
+      "learning_rate": 1e-06,
+      "loss": 0.018,
+      "step": 2193
+    },
+    {
+      "epoch": 0.3735103847463398,
+      "grad_norm": 1.247621774673462,
+      "learning_rate": 1e-06,
+      "loss": 0.0134,
+      "step": 2194
+    },
+    {
+      "epoch": 0.37368062648961525,
+      "grad_norm": 1.1101661920547485,
+      "learning_rate": 1e-06,
+      "loss": 0.013,
+      "step": 2195
+    },
+    {
+      "epoch": 0.3738508682328907,
+      "grad_norm": 1.9619181156158447,
+      "learning_rate": 1e-06,
+      "loss": 0.0247,
+      "step": 2196
+    },
+    {
+      "epoch": 0.37402110997616617,
+      "grad_norm": 1.3228130340576172,
+      "learning_rate": 1e-06,
+      "loss": 0.0272,
+      "step": 2197
+    },
+    {
+      "epoch": 0.3741913517194416,
+      "grad_norm": 1.2315998077392578,
+      "learning_rate": 1e-06,
+      "loss": 0.0153,
+      "step": 2198
+    },
+    {
+      "epoch": 0.3743615934627171,
+      "grad_norm": 1.2987167835235596,
+      "learning_rate": 1e-06,
+      "loss": 0.0151,
+      "step": 2199
+    },
+    {
+      "epoch": 0.37453183520599254,
+      "grad_norm": 1.2436332702636719,
+      "learning_rate": 1e-06,
+      "loss": 0.0175,
+      "step": 2200
+    },
+    {
+      "epoch": 0.37470207694926794,
+      "grad_norm": 1.6267950534820557,
+      "learning_rate": 1e-06,
+      "loss": 0.0196,
+      "step": 2201
+    },
+    {
+      "epoch": 0.3748723186925434,
+      "grad_norm": 1.679490327835083,
+      "learning_rate": 1e-06,
+      "loss": 0.0191,
+      "step": 2202
+    },
+    {
+      "epoch": 0.37504256043581885,
+      "grad_norm": 1.6547735929489136,
+      "learning_rate": 1e-06,
+      "loss": 0.0159,
+      "step": 2203
+    },
+    {
+      "epoch": 0.3752128021790943,
+      "grad_norm": 1.4193183183670044,
+      "learning_rate": 1e-06,
+      "loss": 0.0184,
+      "step": 2204
+    },
+    {
+      "epoch": 0.37538304392236976,
+      "grad_norm": 1.39267098903656,
+      "learning_rate": 1e-06,
+      "loss": 0.025,
+      "step": 2205
+    },
+    {
+      "epoch": 0.3755532856656452,
+      "grad_norm": 1.2222113609313965,
+      "learning_rate": 1e-06,
+      "loss": 0.0129,
+      "step": 2206
+    },
+    {
+      "epoch": 0.3757235274089207,
+      "grad_norm": 1.31162428855896,
+      "learning_rate": 1e-06,
+      "loss": 0.0147,
+      "step": 2207
+    },
+    {
+      "epoch": 0.37589376915219613,
+      "grad_norm": 1.1816409826278687,
+      "learning_rate": 1e-06,
+      "loss": 0.0152,
+      "step": 2208
+    },
+    {
+      "epoch": 0.3760640108954716,
+      "grad_norm": 1.3292574882507324,
+      "learning_rate": 1e-06,
+      "loss": 0.0146,
+      "step": 2209
+    },
+    {
+      "epoch": 0.37623425263874705,
+      "grad_norm": 1.1428711414337158,
+      "learning_rate": 1e-06,
+      "loss": 0.013,
+      "step": 2210
+    },
+    {
+      "epoch": 0.37640449438202245,
+      "grad_norm": 1.3439229726791382,
+      "learning_rate": 1e-06,
+      "loss": 0.0159,
+      "step": 2211
+    },
+    {
+      "epoch": 0.3765747361252979,
+      "grad_norm": 1.5189659595489502,
+      "learning_rate": 1e-06,
+      "loss": 0.0231,
+      "step": 2212
+    },
+    {
+      "epoch": 0.37674497786857336,
+      "grad_norm": 1.3098925352096558,
+      "learning_rate": 1e-06,
+      "loss": 0.015,
+      "step": 2213
+    },
+    {
+      "epoch": 0.3769152196118488,
+      "grad_norm": 1.3993158340454102,
+      "learning_rate": 1e-06,
+      "loss": 0.0225,
+      "step": 2214
+    },
+    {
+      "epoch": 0.3770854613551243,
+      "grad_norm": 1.4516868591308594,
+      "learning_rate": 1e-06,
+      "loss": 0.0204,
+      "step": 2215
+    },
+    {
+      "epoch": 0.37725570309839973,
+      "grad_norm": 1.124986171722412,
+      "learning_rate": 1e-06,
+      "loss": 0.0119,
+      "step": 2216
+    },
+    {
+      "epoch": 0.3774259448416752,
+      "grad_norm": 1.4332947731018066,
+      "learning_rate": 1e-06,
+      "loss": 0.023,
+      "step": 2217
+    },
+    {
+      "epoch": 0.37759618658495064,
+      "grad_norm": 1.2420907020568848,
+      "learning_rate": 1e-06,
+      "loss": 0.015,
+      "step": 2218
+    },
+    {
+      "epoch": 0.3777664283282261,
+      "grad_norm": 1.521194577217102,
+      "learning_rate": 1e-06,
+      "loss": 0.0173,
+      "step": 2219
+    },
+    {
+      "epoch": 0.37793667007150156,
+      "grad_norm": 1.1087653636932373,
+      "learning_rate": 1e-06,
+      "loss": 0.0148,
+      "step": 2220
+    },
+    {
+      "epoch": 0.37810691181477696,
+      "grad_norm": 1.2441989183425903,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 2221
+    },
+    {
+      "epoch": 0.3782771535580524,
+      "grad_norm": 1.500753402709961,
+      "learning_rate": 1e-06,
+      "loss": 0.0223,
+      "step": 2222
+    },
+    {
+      "epoch": 0.37844739530132787,
+      "grad_norm": 1.2610942125320435,
+      "learning_rate": 1e-06,
+      "loss": 0.0132,
+      "step": 2223
+    },
+    {
+      "epoch": 0.37861763704460333,
+      "grad_norm": 1.7212361097335815,
+      "learning_rate": 1e-06,
+      "loss": 0.0143,
+      "step": 2224
+    },
+    {
+      "epoch": 0.3787878787878788,
+      "grad_norm": 5.513698101043701,
+      "learning_rate": 1e-06,
+      "loss": 0.0828,
+      "step": 2225
+    },
+    {
+      "epoch": 0.37895812053115424,
+      "grad_norm": 1.726049542427063,
+      "learning_rate": 1e-06,
+      "loss": 0.018,
+      "step": 2226
+    },
+    {
+      "epoch": 0.3791283622744297,
+      "grad_norm": 1.936808466911316,
+      "learning_rate": 1e-06,
+      "loss": 0.0195,
+      "step": 2227
+    },
+    {
+      "epoch": 0.37929860401770515,
+      "grad_norm": 1.879072666168213,
+      "learning_rate": 1e-06,
+      "loss": 0.0226,
+      "step": 2228
+    },
+    {
+      "epoch": 0.3794688457609806,
+      "grad_norm": 1.6716440916061401,
+      "learning_rate": 1e-06,
+      "loss": 0.0294,
+      "step": 2229
+    },
+    {
+      "epoch": 0.37963908750425607,
+      "grad_norm": 1.3907865285873413,
+      "learning_rate": 1e-06,
+      "loss": 0.0205,
+      "step": 2230
+    },
+    {
+      "epoch": 0.37980932924753147,
+      "grad_norm": 1.4366625547409058,
+      "learning_rate": 1e-06,
+      "loss": 0.022,
+      "step": 2231
+    },
+    {
+      "epoch": 0.3799795709908069,
+      "grad_norm": 1.3679766654968262,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 2232
+    },
+    {
+      "epoch": 0.3801498127340824,
+      "grad_norm": 1.318230390548706,
+      "learning_rate": 1e-06,
+      "loss": 0.0253,
+      "step": 2233
+    },
+    {
+      "epoch": 0.38032005447735784,
+      "grad_norm": 1.0809030532836914,
+      "learning_rate": 1e-06,
+      "loss": 0.0159,
+      "step": 2234
+    },
+    {
+      "epoch": 0.3804902962206333,
+      "grad_norm": 1.47706937789917,
+      "learning_rate": 1e-06,
+      "loss": 0.0191,
+      "step": 2235
+    },
+    {
+      "epoch": 0.38066053796390875,
+      "grad_norm": 1.1768239736557007,
+      "learning_rate": 1e-06,
+      "loss": 0.0145,
+      "step": 2236
+    },
+    {
+      "epoch": 0.3808307797071842,
+      "grad_norm": 1.2666648626327515,
+      "learning_rate": 1e-06,
+      "loss": 0.0175,
+      "step": 2237
+    },
+    {
+      "epoch": 0.38100102145045966,
+      "grad_norm": 1.3947467803955078,
+      "learning_rate": 1e-06,
+      "loss": 0.0177,
+      "step": 2238
+    },
+    {
+      "epoch": 0.3811712631937351,
+      "grad_norm": 1.308648943901062,
+      "learning_rate": 1e-06,
+      "loss": 0.0174,
+      "step": 2239
+    },
+    {
+      "epoch": 0.3813415049370106,
+      "grad_norm": 1.5134379863739014,
+      "learning_rate": 1e-06,
+      "loss": 0.0133,
+      "step": 2240
+    },
+    {
+      "epoch": 0.381511746680286,
+      "grad_norm": 1.6946591138839722,
+      "learning_rate": 1e-06,
+      "loss": 0.0219,
+      "step": 2241
+    },
+    {
+      "epoch": 0.38168198842356144,
+      "grad_norm": 1.2524915933609009,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 2242
+    },
+    {
+      "epoch": 0.3818522301668369,
+      "grad_norm": 1.4766790866851807,
+      "learning_rate": 1e-06,
+      "loss": 0.0228,
+      "step": 2243
+    },
+    {
+      "epoch": 0.38202247191011235,
+      "grad_norm": 1.1624605655670166,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 2244
+    },
+    {
+      "epoch": 0.3821927136533878,
+      "grad_norm": 1.6786658763885498,
+      "learning_rate": 1e-06,
+      "loss": 0.019,
+      "step": 2245
+    },
+    {
+      "epoch": 0.38236295539666326,
+      "grad_norm": 1.1682196855545044,
+      "learning_rate": 1e-06,
+      "loss": 0.0136,
+      "step": 2246
+    },
+    {
+      "epoch": 0.3825331971399387,
+      "grad_norm": 1.470716953277588,
+      "learning_rate": 1e-06,
+      "loss": 0.016,
+      "step": 2247
+    },
+    {
+      "epoch": 0.3827034388832142,
+      "grad_norm": 1.309889793395996,
+      "learning_rate": 1e-06,
+      "loss": 0.0106,
+      "step": 2248
+    },
+    {
+      "epoch": 0.38287368062648963,
+      "grad_norm": 1.0708969831466675,
+      "learning_rate": 1e-06,
+      "loss": 0.0138,
+      "step": 2249
+    },
+    {
+      "epoch": 0.3830439223697651,
+      "grad_norm": 1.1626148223876953,
+      "learning_rate": 1e-06,
+      "loss": 0.0162,
+      "step": 2250
+    },
+    {
+      "epoch": 0.38321416411304055,
+      "grad_norm": 1.4477424621582031,
+      "learning_rate": 1e-06,
+      "loss": 0.0168,
+      "step": 2251
+    },
+    {
+      "epoch": 0.38338440585631595,
+      "grad_norm": 1.3030651807785034,
+      "learning_rate": 1e-06,
+      "loss": 0.0153,
+      "step": 2252
+    },
+    {
+      "epoch": 0.3835546475995914,
+      "grad_norm": 1.7228481769561768,
+      "learning_rate": 1e-06,
+      "loss": 0.0206,
+      "step": 2253
+    },
+    {
+      "epoch": 0.38372488934286686,
+      "grad_norm": 1.1536338329315186,
+      "learning_rate": 1e-06,
+      "loss": 0.0184,
+      "step": 2254
+    },
+    {
+      "epoch": 0.3838951310861423,
+      "grad_norm": 1.347286581993103,
+      "learning_rate": 1e-06,
+      "loss": 0.0106,
+      "step": 2255
+    },
+    {
+      "epoch": 0.3840653728294178,
+      "grad_norm": 1.1855061054229736,
+      "learning_rate": 1e-06,
+      "loss": 0.0186,
+      "step": 2256
+    },
+    {
+      "epoch": 0.38423561457269323,
+      "grad_norm": 1.283276915550232,
+      "learning_rate": 1e-06,
+      "loss": 0.0133,
+      "step": 2257
+    },
+    {
+      "epoch": 0.3844058563159687,
+      "grad_norm": 1.4037493467330933,
+      "learning_rate": 1e-06,
+      "loss": 0.0184,
+      "step": 2258
+    },
+    {
+      "epoch": 0.38457609805924414,
+      "grad_norm": 1.4328378438949585,
+      "learning_rate": 1e-06,
+      "loss": 0.022,
+      "step": 2259
+    },
+    {
+      "epoch": 0.3847463398025196,
+      "grad_norm": 1.7087554931640625,
+      "learning_rate": 1e-06,
+      "loss": 0.02,
+      "step": 2260
+    },
+    {
+      "epoch": 0.38491658154579506,
+      "grad_norm": 1.0174990892410278,
+      "learning_rate": 1e-06,
+      "loss": 0.012,
+      "step": 2261
+    },
+    {
+      "epoch": 0.38508682328907046,
+      "grad_norm": 1.4762239456176758,
+      "learning_rate": 1e-06,
+      "loss": 0.0168,
+      "step": 2262
+    },
+    {
+      "epoch": 0.3852570650323459,
+      "grad_norm": 1.040967345237732,
+      "learning_rate": 1e-06,
+      "loss": 0.0114,
+      "step": 2263
+    },
+    {
+      "epoch": 0.38542730677562137,
+      "grad_norm": 1.2842293977737427,
+      "learning_rate": 1e-06,
+      "loss": 0.0144,
+      "step": 2264
+    },
+    {
+      "epoch": 0.3855975485188968,
+      "grad_norm": 1.1241962909698486,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 2265
+    },
+    {
+      "epoch": 0.3857677902621723,
+      "grad_norm": 1.4264283180236816,
+      "learning_rate": 1e-06,
+      "loss": 0.0184,
+      "step": 2266
+    },
+    {
+      "epoch": 0.38593803200544774,
+      "grad_norm": 1.4640333652496338,
+      "learning_rate": 1e-06,
+      "loss": 0.0226,
+      "step": 2267
+    },
+    {
+      "epoch": 0.3861082737487232,
+      "grad_norm": 1.8459786176681519,
+      "learning_rate": 1e-06,
+      "loss": 0.0183,
+      "step": 2268
+    },
+    {
+      "epoch": 0.38627851549199865,
+      "grad_norm": 1.619437575340271,
+      "learning_rate": 1e-06,
+      "loss": 0.0322,
+      "step": 2269
+    },
+    {
+      "epoch": 0.3864487572352741,
+      "grad_norm": 1.2061455249786377,
+      "learning_rate": 1e-06,
+      "loss": 0.0152,
+      "step": 2270
+    },
+    {
+      "epoch": 0.38661899897854957,
+      "grad_norm": 2.0068230628967285,
+      "learning_rate": 1e-06,
+      "loss": 0.0161,
+      "step": 2271
+    },
+    {
+      "epoch": 0.38678924072182497,
+      "grad_norm": 1.225205659866333,
+      "learning_rate": 1e-06,
+      "loss": 0.0123,
+      "step": 2272
+    },
+    {
+      "epoch": 0.3869594824651004,
+      "grad_norm": 1.4834448099136353,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 2273
+    },
+    {
+      "epoch": 0.3871297242083759,
+      "grad_norm": 1.3730425834655762,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 2274
+    },
+    {
+      "epoch": 0.38729996595165134,
+      "grad_norm": 1.3718072175979614,
+      "learning_rate": 1e-06,
+      "loss": 0.0144,
+      "step": 2275
+    },
+    {
+      "epoch": 0.3874702076949268,
+      "grad_norm": 1.5890707969665527,
+      "learning_rate": 1e-06,
+      "loss": 0.018,
+      "step": 2276
+    },
+    {
+      "epoch": 0.38764044943820225,
+      "grad_norm": 1.437637209892273,
+      "learning_rate": 1e-06,
+      "loss": 0.0266,
+      "step": 2277
+    },
+    {
+      "epoch": 0.3878106911814777,
+      "grad_norm": 1.2185838222503662,
+      "learning_rate": 1e-06,
+      "loss": 0.0118,
+      "step": 2278
+    },
+    {
+      "epoch": 0.38798093292475316,
+      "grad_norm": 1.7744706869125366,
+      "learning_rate": 1e-06,
+      "loss": 0.0246,
+      "step": 2279
+    },
+    {
+      "epoch": 0.3881511746680286,
+      "grad_norm": 1.0997908115386963,
+      "learning_rate": 1e-06,
+      "loss": 0.0128,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3883214164113041,
+      "grad_norm": 1.7149277925491333,
+      "learning_rate": 1e-06,
+      "loss": 0.0189,
+      "step": 2281
+    },
+    {
+      "epoch": 0.3884916581545795,
+      "grad_norm": 4.9614338874816895,
+      "learning_rate": 1e-06,
+      "loss": 0.0606,
+      "step": 2282
+    },
+    {
+      "epoch": 0.38866189989785493,
+      "grad_norm": 1.2618991136550903,
+      "learning_rate": 1e-06,
+      "loss": 0.0162,
+      "step": 2283
+    },
+    {
+      "epoch": 0.3888321416411304,
+      "grad_norm": 1.7077900171279907,
+      "learning_rate": 1e-06,
+      "loss": 0.0212,
+      "step": 2284
+    },
+    {
+      "epoch": 0.38900238338440585,
+      "grad_norm": 1.330156683921814,
+      "learning_rate": 1e-06,
+      "loss": 0.0153,
+      "step": 2285
+    },
+    {
+      "epoch": 0.3891726251276813,
+      "grad_norm": 1.5688471794128418,
+      "learning_rate": 1e-06,
+      "loss": 0.0196,
+      "step": 2286
+    },
+    {
+      "epoch": 0.38934286687095676,
+      "grad_norm": 1.4680328369140625,
+      "learning_rate": 1e-06,
+      "loss": 0.0162,
+      "step": 2287
+    },
+    {
+      "epoch": 0.3895131086142322,
+      "grad_norm": 1.4984068870544434,
+      "learning_rate": 1e-06,
+      "loss": 0.0226,
+      "step": 2288
+    },
+    {
+      "epoch": 0.3896833503575077,
+      "grad_norm": 1.8151462078094482,
+      "learning_rate": 1e-06,
+      "loss": 0.0177,
+      "step": 2289
+    },
+    {
+      "epoch": 0.38985359210078313,
+      "grad_norm": 1.1406170129776,
+      "learning_rate": 1e-06,
+      "loss": 0.0151,
+      "step": 2290
+    },
+    {
+      "epoch": 0.3900238338440586,
+      "grad_norm": 1.425500512123108,
+      "learning_rate": 1e-06,
+      "loss": 0.022,
+      "step": 2291
+    },
+    {
+      "epoch": 0.390194075587334,
+      "grad_norm": 1.5256712436676025,
+      "learning_rate": 1e-06,
+      "loss": 0.0152,
+      "step": 2292
+    },
+    {
+      "epoch": 0.39036431733060944,
+      "grad_norm": 1.2895482778549194,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 2293
+    },
+    {
+      "epoch": 0.3905345590738849,
+      "grad_norm": 1.3082205057144165,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 2294
+    },
+    {
+      "epoch": 0.39070480081716036,
+      "grad_norm": 1.8497793674468994,
+      "learning_rate": 1e-06,
+      "loss": 0.0213,
+      "step": 2295
+    },
+    {
+      "epoch": 0.3908750425604358,
+      "grad_norm": 1.3272291421890259,
+      "learning_rate": 1e-06,
+      "loss": 0.0154,
+      "step": 2296
+    },
+    {
+      "epoch": 0.39104528430371127,
+      "grad_norm": 1.159996747970581,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 2297
+    },
+    {
+      "epoch": 0.39121552604698673,
+      "grad_norm": 1.58273184299469,
+      "learning_rate": 1e-06,
+      "loss": 0.0124,
+      "step": 2298
+    },
+    {
+      "epoch": 0.3913857677902622,
+      "grad_norm": 1.2419120073318481,
+      "learning_rate": 1e-06,
+      "loss": 0.0163,
+      "step": 2299
+    },
+    {
+      "epoch": 0.39155600953353764,
+      "grad_norm": 1.4475661516189575,
+      "learning_rate": 1e-06,
+      "loss": 0.0133,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3917262512768131,
+      "grad_norm": 1.270992636680603,
+      "learning_rate": 1e-06,
+      "loss": 0.0209,
+      "step": 2301
+    },
+    {
+      "epoch": 0.3918964930200885,
+      "grad_norm": 1.5242903232574463,
+      "learning_rate": 1e-06,
+      "loss": 0.0181,
+      "step": 2302
+    },
+    {
+      "epoch": 0.39206673476336396,
+      "grad_norm": 1.34809410572052,
+      "learning_rate": 1e-06,
+      "loss": 0.0148,
+      "step": 2303
+    },
+    {
+      "epoch": 0.3922369765066394,
+      "grad_norm": 1.6840362548828125,
+      "learning_rate": 1e-06,
+      "loss": 0.019,
+      "step": 2304
+    },
+    {
+      "epoch": 0.39240721824991487,
+      "grad_norm": 1.350677728652954,
+      "learning_rate": 1e-06,
+      "loss": 0.0169,
+      "step": 2305
+    },
+    {
+      "epoch": 0.3925774599931903,
+      "grad_norm": 1.4453243017196655,
+      "learning_rate": 1e-06,
+      "loss": 0.0215,
+      "step": 2306
+    },
+    {
+      "epoch": 0.3927477017364658,
+      "grad_norm": 1.3742339611053467,
+      "learning_rate": 1e-06,
+      "loss": 0.0211,
+      "step": 2307
+    },
+    {
+      "epoch": 0.39291794347974124,
+      "grad_norm": 1.471127986907959,
+      "learning_rate": 1e-06,
+      "loss": 0.0149,
+      "step": 2308
+    },
+    {
+      "epoch": 0.3930881852230167,
+      "grad_norm": 1.0254688262939453,
+      "learning_rate": 1e-06,
+      "loss": 0.0131,
+      "step": 2309
+    },
+    {
+      "epoch": 0.39325842696629215,
+      "grad_norm": 1.232068419456482,
+      "learning_rate": 1e-06,
+      "loss": 0.0158,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3934286687095676,
+      "grad_norm": 1.4690749645233154,
+      "learning_rate": 1e-06,
+      "loss": 0.02,
+      "step": 2311
+    },
+    {
+      "epoch": 0.393598910452843,
+      "grad_norm": 1.1098586320877075,
+      "learning_rate": 1e-06,
+      "loss": 0.0141,
+      "step": 2312
+    },
+    {
+      "epoch": 0.39376915219611847,
+      "grad_norm": 1.5079904794692993,
+      "learning_rate": 1e-06,
+      "loss": 0.0203,
+      "step": 2313
+    },
+    {
+      "epoch": 0.3939393939393939,
+      "grad_norm": 0.9211795926094055,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 2314
+    },
+    {
+      "epoch": 0.3941096356826694,
+      "grad_norm": 1.6116832494735718,
+      "learning_rate": 1e-06,
+      "loss": 0.0187,
+      "step": 2315
+    },
+    {
+      "epoch": 0.39427987742594484,
+      "grad_norm": 1.5355298519134521,
+      "learning_rate": 1e-06,
+      "loss": 0.0154,
+      "step": 2316
+    },
+    {
+      "epoch": 0.3944501191692203,
+      "grad_norm": 1.1545178890228271,
+      "learning_rate": 1e-06,
+      "loss": 0.0164,
+      "step": 2317
+    },
+    {
+      "epoch": 0.39462036091249575,
+      "grad_norm": 1.8506436347961426,
+      "learning_rate": 1e-06,
+      "loss": 0.0176,
+      "step": 2318
+    },
+    {
+      "epoch": 0.3947906026557712,
+      "grad_norm": 2.0721189975738525,
+      "learning_rate": 1e-06,
+      "loss": 0.0176,
+      "step": 2319
+    },
+    {
+      "epoch": 0.39496084439904666,
+      "grad_norm": 1.3172030448913574,
+      "learning_rate": 1e-06,
+      "loss": 0.0173,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3951310861423221,
+      "grad_norm": 1.3635270595550537,
+      "learning_rate": 1e-06,
+      "loss": 0.0142,
+      "step": 2321
+    },
+    {
+      "epoch": 0.3953013278855976,
+      "grad_norm": 1.1138346195220947,
+      "learning_rate": 1e-06,
+      "loss": 0.0189,
+      "step": 2322
+    },
+    {
+      "epoch": 0.395471569628873,
+      "grad_norm": 1.483282446861267,
+      "learning_rate": 1e-06,
+      "loss": 0.0185,
+      "step": 2323
+    },
+    {
+      "epoch": 0.39564181137214843,
+      "grad_norm": 1.542601466178894,
+      "learning_rate": 1e-06,
+      "loss": 0.023,
+      "step": 2324
+    },
+    {
+      "epoch": 0.3958120531154239,
+      "grad_norm": 1.5426020622253418,
+      "learning_rate": 1e-06,
+      "loss": 0.0302,
+      "step": 2325
+    },
+    {
+      "epoch": 0.39598229485869935,
+      "grad_norm": 1.148687720298767,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 2326
+    },
+    {
+      "epoch": 0.3961525366019748,
+      "grad_norm": 1.3703385591506958,
+      "learning_rate": 1e-06,
+      "loss": 0.0179,
+      "step": 2327
+    },
+    {
+      "epoch": 0.39632277834525026,
+      "grad_norm": 1.4654871225357056,
+      "learning_rate": 1e-06,
+      "loss": 0.0146,
+      "step": 2328
+    },
+    {
+      "epoch": 0.3964930200885257,
+      "grad_norm": 2.1327321529388428,
+      "learning_rate": 1e-06,
+      "loss": 0.028,
+      "step": 2329
+    },
+    {
+      "epoch": 0.3966632618318012,
+      "grad_norm": 1.7279863357543945,
+      "learning_rate": 1e-06,
+      "loss": 0.0238,
+      "step": 2330
+    },
+    {
+      "epoch": 0.39683350357507663,
+      "grad_norm": 1.1947439908981323,
+      "learning_rate": 1e-06,
+      "loss": 0.0125,
+      "step": 2331
+    },
+    {
+      "epoch": 0.3970037453183521,
+      "grad_norm": 1.0099948644638062,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 2332
+    },
+    {
+      "epoch": 0.3971739870616275,
+      "grad_norm": 1.3921846151351929,
+      "learning_rate": 1e-06,
+      "loss": 0.0162,
+      "step": 2333
+    },
+    {
+      "epoch": 0.39734422880490294,
+      "grad_norm": 1.3162715435028076,
+      "learning_rate": 1e-06,
+      "loss": 0.0158,
+      "step": 2334
+    },
+    {
+      "epoch": 0.3975144705481784,
+      "grad_norm": 1.7758504152297974,
+      "learning_rate": 1e-06,
+      "loss": 0.018,
+      "step": 2335
+    },
+    {
+      "epoch": 0.39768471229145386,
+      "grad_norm": 1.2216438055038452,
+      "learning_rate": 1e-06,
+      "loss": 0.0212,
+      "step": 2336
+    },
+    {
+      "epoch": 0.3978549540347293,
+      "grad_norm": 1.6880277395248413,
+      "learning_rate": 1e-06,
+      "loss": 0.0149,
+      "step": 2337
+    },
+    {
+      "epoch": 0.39802519577800477,
+      "grad_norm": 1.6719337701797485,
+      "learning_rate": 1e-06,
+      "loss": 0.0175,
+      "step": 2338
+    },
+    {
+      "epoch": 0.3981954375212802,
+      "grad_norm": 1.5387353897094727,
+      "learning_rate": 1e-06,
+      "loss": 0.0261,
+      "step": 2339
+    },
+    {
+      "epoch": 0.3983656792645557,
+      "grad_norm": 1.6653772592544556,
+      "learning_rate": 1e-06,
+      "loss": 0.0161,
+      "step": 2340
+    },
+    {
+      "epoch": 0.39853592100783114,
+      "grad_norm": 1.325972318649292,
+      "learning_rate": 1e-06,
+      "loss": 0.0202,
+      "step": 2341
+    },
+    {
+      "epoch": 0.3987061627511066,
+      "grad_norm": 1.40695321559906,
+      "learning_rate": 1e-06,
+      "loss": 0.0182,
+      "step": 2342
+    },
+    {
+      "epoch": 0.398876404494382,
+      "grad_norm": 1.6472898721694946,
+      "learning_rate": 1e-06,
+      "loss": 0.0167,
+      "step": 2343
+    },
+    {
+      "epoch": 0.39904664623765745,
+      "grad_norm": 1.2183064222335815,
+      "learning_rate": 1e-06,
+      "loss": 0.0104,
+      "step": 2344
+    },
+    {
+      "epoch": 0.3992168879809329,
+      "grad_norm": 1.193183183670044,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 2345
+    },
+    {
+      "epoch": 0.39938712972420837,
+      "grad_norm": 1.4041932821273804,
+      "learning_rate": 1e-06,
+      "loss": 0.0156,
+      "step": 2346
+    },
+    {
+      "epoch": 0.3995573714674838,
+      "grad_norm": 1.3237617015838623,
+      "learning_rate": 1e-06,
+      "loss": 0.0157,
+      "step": 2347
+    },
+    {
+      "epoch": 0.3997276132107593,
+      "grad_norm": 1.5075606107711792,
+      "learning_rate": 1e-06,
+      "loss": 0.0167,
+      "step": 2348
+    },
+    {
+      "epoch": 0.39989785495403474,
+      "grad_norm": 1.4453775882720947,
+      "learning_rate": 1e-06,
+      "loss": 0.0241,
+      "step": 2349
+    },
+    {
+      "epoch": 0.4000680966973102,
+      "grad_norm": 1.4384607076644897,
+      "learning_rate": 1e-06,
+      "loss": 0.0171,
+      "step": 2350
+    },
+    {
+      "epoch": 0.40023833844058565,
+      "grad_norm": 1.3807544708251953,
+      "learning_rate": 1e-06,
+      "loss": 0.0131,
+      "step": 2351
+    },
+    {
+      "epoch": 0.4004085801838611,
+      "grad_norm": 1.7280919551849365,
+      "learning_rate": 1e-06,
+      "loss": 0.0254,
+      "step": 2352
+    },
+    {
+      "epoch": 0.4005788219271365,
+      "grad_norm": 1.2561084032058716,
+      "learning_rate": 1e-06,
+      "loss": 0.0089,
+      "step": 2353
+    },
+    {
+      "epoch": 0.40074906367041196,
+      "grad_norm": 1.4211997985839844,
+      "learning_rate": 1e-06,
+      "loss": 0.0174,
+      "step": 2354
+    },
+    {
+      "epoch": 0.4009193054136874,
+      "grad_norm": 1.79752516746521,
+      "learning_rate": 1e-06,
+      "loss": 0.0177,
+      "step": 2355
+    },
+    {
+      "epoch": 0.4010895471569629,
+      "grad_norm": 1.746137022972107,
+      "learning_rate": 1e-06,
+      "loss": 0.0208,
+      "step": 2356
+    },
+    {
+      "epoch": 0.40125978890023833,
+      "grad_norm": 3.5860726833343506,
+      "learning_rate": 1e-06,
+      "loss": 0.0709,
+      "step": 2357
+    },
+    {
+      "epoch": 0.4014300306435138,
+      "grad_norm": 1.5278226137161255,
+      "learning_rate": 1e-06,
+      "loss": 0.0161,
+      "step": 2358
+    },
+    {
+      "epoch": 0.40160027238678925,
+      "grad_norm": 1.536190390586853,
+      "learning_rate": 1e-06,
+      "loss": 0.0217,
+      "step": 2359
+    },
+    {
+      "epoch": 0.4017705141300647,
+      "grad_norm": 1.5676721334457397,
+      "learning_rate": 1e-06,
+      "loss": 0.0166,
+      "step": 2360
+    },
+    {
+      "epoch": 0.40194075587334016,
+      "grad_norm": 1.004481554031372,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 2361
+    },
+    {
+      "epoch": 0.4021109976166156,
+      "grad_norm": 1.576690673828125,
+      "learning_rate": 1e-06,
+      "loss": 0.0267,
+      "step": 2362
+    },
+    {
+      "epoch": 0.402281239359891,
+      "grad_norm": 1.8953304290771484,
+      "learning_rate": 1e-06,
+      "loss": 0.0205,
+      "step": 2363
+    },
+    {
+      "epoch": 0.4024514811031665,
+      "grad_norm": 1.3143856525421143,
+      "learning_rate": 1e-06,
+      "loss": 0.013,
+      "step": 2364
+    },
+    {
+      "epoch": 0.40262172284644193,
+      "grad_norm": 1.4957778453826904,
+      "learning_rate": 1e-06,
+      "loss": 0.0285,
+      "step": 2365
+    },
+    {
+      "epoch": 0.4027919645897174,
+      "grad_norm": 1.0621047019958496,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 2366
+    },
+    {
+      "epoch": 0.40296220633299284,
+      "grad_norm": 1.4584861993789673,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 2367
+    },
+    {
+      "epoch": 0.4031324480762683,
+      "grad_norm": 1.5141526460647583,
+      "learning_rate": 1e-06,
+      "loss": 0.0149,
+      "step": 2368
+    },
+    {
+      "epoch": 0.40330268981954376,
+      "grad_norm": 1.0880855321884155,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 2369
+    },
+    {
+      "epoch": 0.4034729315628192,
+      "grad_norm": 1.1734309196472168,
+      "learning_rate": 1e-06,
+      "loss": 0.0158,
+      "step": 2370
+    },
+    {
+      "epoch": 0.40364317330609467,
+      "grad_norm": 1.3398406505584717,
+      "learning_rate": 1e-06,
+      "loss": 0.0172,
+      "step": 2371
+    },
+    {
+      "epoch": 0.40381341504937013,
+      "grad_norm": 1.1609718799591064,
+      "learning_rate": 1e-06,
+      "loss": 0.0157,
+      "step": 2372
+    },
+    {
+      "epoch": 0.40398365679264553,
+      "grad_norm": 4.927487373352051,
+      "learning_rate": 1e-06,
+      "loss": 0.0723,
+      "step": 2373
+    },
+    {
+      "epoch": 0.404153898535921,
+      "grad_norm": 1.258873462677002,
+      "learning_rate": 1e-06,
+      "loss": 0.018,
+      "step": 2374
+    },
+    {
+      "epoch": 0.40432414027919644,
+      "grad_norm": 1.1295454502105713,
+      "learning_rate": 1e-06,
+      "loss": 0.0153,
+      "step": 2375
+    },
+    {
+      "epoch": 0.4044943820224719,
+      "grad_norm": 1.213400959968567,
+      "learning_rate": 1e-06,
+      "loss": 0.0137,
+      "step": 2376
+    },
+    {
+      "epoch": 0.40466462376574736,
+      "grad_norm": 1.0463217496871948,
+      "learning_rate": 1e-06,
+      "loss": 0.013,
+      "step": 2377
+    },
+    {
+      "epoch": 0.4048348655090228,
+      "grad_norm": 1.1880186796188354,
+      "learning_rate": 1e-06,
+      "loss": 0.0145,
+      "step": 2378
+    },
+    {
+      "epoch": 0.40500510725229827,
+      "grad_norm": 1.3919751644134521,
+      "learning_rate": 1e-06,
+      "loss": 0.015,
+      "step": 2379
+    },
+    {
+      "epoch": 0.4051753489955737,
+      "grad_norm": 1.2092021703720093,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 2380
+    },
+    {
+      "epoch": 0.4053455907388492,
+      "grad_norm": 1.5249980688095093,
+      "learning_rate": 1e-06,
+      "loss": 0.016,
+      "step": 2381
+    },
+    {
+      "epoch": 0.40551583248212464,
+      "grad_norm": 1.3709051609039307,
+      "learning_rate": 1e-06,
+      "loss": 0.0167,
+      "step": 2382
+    },
+    {
+      "epoch": 0.4056860742254001,
+      "grad_norm": 1.3142911195755005,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 2383
+    },
+    {
+      "epoch": 0.4058563159686755,
+      "grad_norm": 1.423117995262146,
+      "learning_rate": 1e-06,
+      "loss": 0.0144,
+      "step": 2384
+    },
+    {
+      "epoch": 0.40602655771195095,
+      "grad_norm": 1.7531969547271729,
+      "learning_rate": 1e-06,
+      "loss": 0.0164,
+      "step": 2385
+    },
+    {
+      "epoch": 0.4061967994552264,
+      "grad_norm": 1.1912846565246582,
+      "learning_rate": 1e-06,
+      "loss": 0.0177,
+      "step": 2386
+    },
+    {
+      "epoch": 0.40636704119850187,
+      "grad_norm": 1.3569473028182983,
+      "learning_rate": 1e-06,
+      "loss": 0.0152,
+      "step": 2387
+    },
+    {
+      "epoch": 0.4065372829417773,
+      "grad_norm": 1.3196473121643066,
+      "learning_rate": 1e-06,
+      "loss": 0.0183,
+      "step": 2388
+    },
+    {
+      "epoch": 0.4067075246850528,
+      "grad_norm": 1.8479810953140259,
+      "learning_rate": 1e-06,
+      "loss": 0.0224,
+      "step": 2389
+    },
+    {
+      "epoch": 0.40687776642832824,
+      "grad_norm": 1.048545479774475,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 2390
+    },
+    {
+      "epoch": 0.4070480081716037,
+      "grad_norm": 1.090034008026123,
+      "learning_rate": 1e-06,
+      "loss": 0.0124,
+      "step": 2391
+    },
+    {
+      "epoch": 0.40721824991487915,
+      "grad_norm": 1.0449838638305664,
+      "learning_rate": 1e-06,
+      "loss": 0.0124,
+      "step": 2392
+    },
+    {
+      "epoch": 0.4073884916581546,
+      "grad_norm": 1.3128901720046997,
+      "learning_rate": 1e-06,
+      "loss": 0.0118,
+      "step": 2393
+    },
+    {
+      "epoch": 0.40755873340143,
+      "grad_norm": 1.3702256679534912,
+      "learning_rate": 1e-06,
+      "loss": 0.0171,
+      "step": 2394
+    },
+    {
+      "epoch": 0.40772897514470546,
+      "grad_norm": 1.4796867370605469,
+      "learning_rate": 1e-06,
+      "loss": 0.0204,
+      "step": 2395
+    },
+    {
+      "epoch": 0.4078992168879809,
+      "grad_norm": 1.1662784814834595,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 2396
+    },
+    {
+      "epoch": 0.4080694586312564,
+      "grad_norm": 2.216245651245117,
+      "learning_rate": 1e-06,
+      "loss": 0.0212,
+      "step": 2397
+    },
+    {
+      "epoch": 0.40823970037453183,
+      "grad_norm": 1.5245473384857178,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 2398
+    },
+    {
+      "epoch": 0.4084099421178073,
+      "grad_norm": 1.6484607458114624,
+      "learning_rate": 1e-06,
+      "loss": 0.0132,
+      "step": 2399
+    },
+    {
+      "epoch": 0.40858018386108275,
+      "grad_norm": 1.326431155204773,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 2400
+    },
+    {
+      "epoch": 0.4087504256043582,
+      "grad_norm": 1.8550671339035034,
+      "learning_rate": 1e-06,
+      "loss": 0.0198,
+      "step": 2401
+    },
+    {
+      "epoch": 0.40892066734763366,
+      "grad_norm": 1.1143696308135986,
+      "learning_rate": 1e-06,
+      "loss": 0.0141,
+      "step": 2402
+    },
+    {
+      "epoch": 0.4090909090909091,
+      "grad_norm": 1.190245509147644,
+      "learning_rate": 1e-06,
+      "loss": 0.012,
+      "step": 2403
+    },
+    {
+      "epoch": 0.4092611508341845,
+      "grad_norm": 1.6202400922775269,
+      "learning_rate": 1e-06,
+      "loss": 0.019,
+      "step": 2404
+    },
+    {
+      "epoch": 0.40943139257746,
+      "grad_norm": 1.4784173965454102,
+      "learning_rate": 1e-06,
+      "loss": 0.0213,
+      "step": 2405
+    },
+    {
+      "epoch": 0.40960163432073543,
+      "grad_norm": 1.2796027660369873,
+      "learning_rate": 1e-06,
+      "loss": 0.0161,
+      "step": 2406
+    },
+    {
+      "epoch": 0.4097718760640109,
+      "grad_norm": 1.0544267892837524,
+      "learning_rate": 1e-06,
+      "loss": 0.0147,
+      "step": 2407
+    },
+    {
+      "epoch": 0.40994211780728634,
+      "grad_norm": 1.6966776847839355,
+      "learning_rate": 1e-06,
+      "loss": 0.0223,
+      "step": 2408
+    },
+    {
+      "epoch": 0.4101123595505618,
+      "grad_norm": 1.3701565265655518,
+      "learning_rate": 1e-06,
+      "loss": 0.0169,
+      "step": 2409
+    },
+    {
+      "epoch": 0.41028260129383726,
+      "grad_norm": 1.4099042415618896,
+      "learning_rate": 1e-06,
+      "loss": 0.0157,
+      "step": 2410
+    },
+    {
+      "epoch": 0.4104528430371127,
+      "grad_norm": 1.2321785688400269,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 2411
+    },
+    {
+      "epoch": 0.41062308478038817,
+      "grad_norm": 1.3855117559432983,
+      "learning_rate": 1e-06,
+      "loss": 0.0216,
+      "step": 2412
+    },
+    {
+      "epoch": 0.4107933265236636,
+      "grad_norm": 1.1128058433532715,
+      "learning_rate": 1e-06,
+      "loss": 0.0124,
+      "step": 2413
+    },
+    {
+      "epoch": 0.410963568266939,
+      "grad_norm": 1.4113258123397827,
+      "learning_rate": 1e-06,
+      "loss": 0.0144,
+      "step": 2414
+    },
+    {
+      "epoch": 0.4111338100102145,
+      "grad_norm": 1.4246609210968018,
+      "learning_rate": 1e-06,
+      "loss": 0.0156,
+      "step": 2415
+    },
+    {
+      "epoch": 0.41130405175348994,
+      "grad_norm": 1.2959715127944946,
+      "learning_rate": 1e-06,
+      "loss": 0.017,
+      "step": 2416
+    },
+    {
+      "epoch": 0.4114742934967654,
+      "grad_norm": 1.3845857381820679,
+      "learning_rate": 1e-06,
+      "loss": 0.0153,
+      "step": 2417
+    },
+    {
+      "epoch": 0.41164453524004085,
+      "grad_norm": 1.1727241277694702,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 2418
+    },
+    {
+      "epoch": 0.4118147769833163,
+      "grad_norm": 1.3593071699142456,
+      "learning_rate": 1e-06,
+      "loss": 0.0189,
+      "step": 2419
+    },
+    {
+      "epoch": 0.41198501872659177,
+      "grad_norm": 1.650676965713501,
+      "learning_rate": 1e-06,
+      "loss": 0.0179,
+      "step": 2420
+    },
+    {
+      "epoch": 0.4121552604698672,
+      "grad_norm": 1.345276117324829,
+      "learning_rate": 1e-06,
+      "loss": 0.015,
+      "step": 2421
+    },
+    {
+      "epoch": 0.4123255022131427,
+      "grad_norm": 1.3404996395111084,
+      "learning_rate": 1e-06,
+      "loss": 0.0145,
+      "step": 2422
+    },
+    {
+      "epoch": 0.41249574395641814,
+      "grad_norm": 1.4839191436767578,
+      "learning_rate": 1e-06,
+      "loss": 0.0176,
+      "step": 2423
+    },
+    {
+      "epoch": 0.41266598569969354,
+      "grad_norm": 1.3546916246414185,
+      "learning_rate": 1e-06,
+      "loss": 0.0196,
+      "step": 2424
+    },
+    {
+      "epoch": 0.412836227442969,
+      "grad_norm": 1.5656921863555908,
+      "learning_rate": 1e-06,
+      "loss": 0.0242,
+      "step": 2425
+    },
+    {
+      "epoch": 0.41300646918624445,
+      "grad_norm": 1.348214030265808,
+      "learning_rate": 1e-06,
+      "loss": 0.0189,
+      "step": 2426
+    },
+    {
+      "epoch": 0.4131767109295199,
+      "grad_norm": 1.2620658874511719,
+      "learning_rate": 1e-06,
+      "loss": 0.0154,
+      "step": 2427
+    },
+    {
+      "epoch": 0.41334695267279536,
+      "grad_norm": 1.1423382759094238,
+      "learning_rate": 1e-06,
+      "loss": 0.0155,
+      "step": 2428
+    },
+    {
+      "epoch": 0.4135171944160708,
+      "grad_norm": 1.179055094718933,
+      "learning_rate": 1e-06,
+      "loss": 0.0144,
+      "step": 2429
+    },
+    {
+      "epoch": 0.4136874361593463,
+      "grad_norm": 1.2674864530563354,
+      "learning_rate": 1e-06,
+      "loss": 0.0133,
+      "step": 2430
+    },
+    {
+      "epoch": 0.41385767790262173,
+      "grad_norm": 1.2309482097625732,
+      "learning_rate": 1e-06,
+      "loss": 0.0177,
+      "step": 2431
+    },
+    {
+      "epoch": 0.4140279196458972,
+      "grad_norm": 1.4680554866790771,
+      "learning_rate": 1e-06,
+      "loss": 0.0156,
+      "step": 2432
+    },
+    {
+      "epoch": 0.41419816138917265,
+      "grad_norm": 1.4172992706298828,
+      "learning_rate": 1e-06,
+      "loss": 0.021,
+      "step": 2433
+    },
+    {
+      "epoch": 0.41436840313244805,
+      "grad_norm": 1.537314534187317,
+      "learning_rate": 1e-06,
+      "loss": 0.0196,
+      "step": 2434
+    },
+    {
+      "epoch": 0.4145386448757235,
+      "grad_norm": 1.3401296138763428,
+      "learning_rate": 1e-06,
+      "loss": 0.0106,
+      "step": 2435
+    },
+    {
+      "epoch": 0.41470888661899896,
+      "grad_norm": 1.5095027685165405,
+      "learning_rate": 1e-06,
+      "loss": 0.0266,
+      "step": 2436
+    },
+    {
+      "epoch": 0.4148791283622744,
+      "grad_norm": 1.1366171836853027,
+      "learning_rate": 1e-06,
+      "loss": 0.0137,
+      "step": 2437
+    },
+    {
+      "epoch": 0.4150493701055499,
+      "grad_norm": 1.1384146213531494,
+      "learning_rate": 1e-06,
+      "loss": 0.0137,
+      "step": 2438
+    },
+    {
+      "epoch": 0.41521961184882533,
+      "grad_norm": 0.9028575420379639,
+      "learning_rate": 1e-06,
+      "loss": 0.0084,
+      "step": 2439
+    },
+    {
+      "epoch": 0.4153898535921008,
+      "grad_norm": 1.199471116065979,
+      "learning_rate": 1e-06,
+      "loss": 0.0122,
+      "step": 2440
+    },
+    {
+      "epoch": 0.41556009533537625,
+      "grad_norm": 1.3193436861038208,
+      "learning_rate": 1e-06,
+      "loss": 0.017,
+      "step": 2441
+    },
+    {
+      "epoch": 0.4157303370786517,
+      "grad_norm": 1.3699177503585815,
+      "learning_rate": 1e-06,
+      "loss": 0.016,
+      "step": 2442
+    },
+    {
+      "epoch": 0.41590057882192716,
+      "grad_norm": 1.3745206594467163,
+      "learning_rate": 1e-06,
+      "loss": 0.0154,
+      "step": 2443
+    },
+    {
+      "epoch": 0.4160708205652026,
+      "grad_norm": 1.289341926574707,
+      "learning_rate": 1e-06,
+      "loss": 0.0152,
+      "step": 2444
+    },
+    {
+      "epoch": 0.416241062308478,
+      "grad_norm": 1.4632568359375,
+      "learning_rate": 1e-06,
+      "loss": 0.0169,
+      "step": 2445
+    },
+    {
+      "epoch": 0.4164113040517535,
+      "grad_norm": 1.4885531663894653,
+      "learning_rate": 1e-06,
+      "loss": 0.0147,
+      "step": 2446
+    },
+    {
+      "epoch": 0.41658154579502893,
+      "grad_norm": 1.2330007553100586,
+      "learning_rate": 1e-06,
+      "loss": 0.0126,
+      "step": 2447
+    },
+    {
+      "epoch": 0.4167517875383044,
+      "grad_norm": 2.102605104446411,
+      "learning_rate": 1e-06,
+      "loss": 0.0458,
+      "step": 2448
+    },
+    {
+      "epoch": 0.41692202928157984,
+      "grad_norm": 1.2682759761810303,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 2449
+    },
+    {
+      "epoch": 0.4170922710248553,
+      "grad_norm": 1.3345555067062378,
+      "learning_rate": 1e-06,
+      "loss": 0.013,
+      "step": 2450
+    },
+    {
+      "epoch": 0.41726251276813076,
+      "grad_norm": 1.390181064605713,
+      "learning_rate": 1e-06,
+      "loss": 0.0203,
+      "step": 2451
+    },
+    {
+      "epoch": 0.4174327545114062,
+      "grad_norm": 1.3475115299224854,
+      "learning_rate": 1e-06,
+      "loss": 0.0199,
+      "step": 2452
+    },
+    {
+      "epoch": 0.41760299625468167,
+      "grad_norm": 1.5461567640304565,
+      "learning_rate": 1e-06,
+      "loss": 0.0179,
+      "step": 2453
+    },
+    {
+      "epoch": 0.4177732379979571,
+      "grad_norm": 1.3287405967712402,
+      "learning_rate": 1e-06,
+      "loss": 0.0156,
+      "step": 2454
+    },
+    {
+      "epoch": 0.4179434797412325,
+      "grad_norm": 1.3914709091186523,
+      "learning_rate": 1e-06,
+      "loss": 0.0141,
+      "step": 2455
+    },
+    {
+      "epoch": 0.418113721484508,
+      "grad_norm": 1.5743072032928467,
+      "learning_rate": 1e-06,
+      "loss": 0.0166,
+      "step": 2456
+    },
+    {
+      "epoch": 0.41828396322778344,
+      "grad_norm": 1.3226776123046875,
+      "learning_rate": 1e-06,
+      "loss": 0.0176,
+      "step": 2457
+    },
+    {
+      "epoch": 0.4184542049710589,
+      "grad_norm": 1.3625043630599976,
+      "learning_rate": 1e-06,
+      "loss": 0.0209,
+      "step": 2458
+    },
+    {
+      "epoch": 0.41862444671433435,
+      "grad_norm": 0.9081085920333862,
+      "learning_rate": 1e-06,
+      "loss": 0.0076,
+      "step": 2459
+    },
+    {
+      "epoch": 0.4187946884576098,
+      "grad_norm": 1.608961820602417,
+      "learning_rate": 1e-06,
+      "loss": 0.0182,
+      "step": 2460
+    },
+    {
+      "epoch": 0.41896493020088527,
+      "grad_norm": 1.2214875221252441,
+      "learning_rate": 1e-06,
+      "loss": 0.0147,
+      "step": 2461
+    },
+    {
+      "epoch": 0.4191351719441607,
+      "grad_norm": 1.1907516717910767,
+      "learning_rate": 1e-06,
+      "loss": 0.0166,
+      "step": 2462
+    },
+    {
+      "epoch": 0.4193054136874362,
+      "grad_norm": 1.0932464599609375,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 2463
+    },
+    {
+      "epoch": 0.41947565543071164,
+      "grad_norm": 1.2685447931289673,
+      "learning_rate": 1e-06,
+      "loss": 0.0141,
+      "step": 2464
+    },
+    {
+      "epoch": 0.41964589717398704,
+      "grad_norm": 1.4846066236495972,
+      "learning_rate": 1e-06,
+      "loss": 0.0146,
+      "step": 2465
+    },
+    {
+      "epoch": 0.4198161389172625,
+      "grad_norm": 1.1014434099197388,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 2466
+    },
+    {
+      "epoch": 0.41998638066053795,
+      "grad_norm": 1.5888346433639526,
+      "learning_rate": 1e-06,
+      "loss": 0.0195,
+      "step": 2467
+    },
+    {
+      "epoch": 0.4201566224038134,
+      "grad_norm": 1.507070779800415,
+      "learning_rate": 1e-06,
+      "loss": 0.0171,
+      "step": 2468
+    },
+    {
+      "epoch": 0.42032686414708886,
+      "grad_norm": 1.330121636390686,
+      "learning_rate": 1e-06,
+      "loss": 0.013,
+      "step": 2469
+    },
+    {
+      "epoch": 0.4204971058903643,
+      "grad_norm": 1.4843699932098389,
+      "learning_rate": 1e-06,
+      "loss": 0.0232,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4206673476336398,
+      "grad_norm": 1.1940176486968994,
+      "learning_rate": 1e-06,
+      "loss": 0.0112,
+      "step": 2471
+    },
+    {
+      "epoch": 0.42083758937691523,
+      "grad_norm": 1.3040800094604492,
+      "learning_rate": 1e-06,
+      "loss": 0.0164,
+      "step": 2472
+    },
+    {
+      "epoch": 0.4210078311201907,
+      "grad_norm": 1.4341161251068115,
+      "learning_rate": 1e-06,
+      "loss": 0.0146,
+      "step": 2473
+    },
+    {
+      "epoch": 0.42117807286346615,
+      "grad_norm": 1.1169599294662476,
+      "learning_rate": 1e-06,
+      "loss": 0.0105,
+      "step": 2474
+    },
+    {
+      "epoch": 0.42134831460674155,
+      "grad_norm": 1.2243717908859253,
+      "learning_rate": 1e-06,
+      "loss": 0.0121,
+      "step": 2475
+    },
+    {
+      "epoch": 0.421518556350017,
+      "grad_norm": 1.5474375486373901,
+      "learning_rate": 1e-06,
+      "loss": 0.0188,
+      "step": 2476
+    },
+    {
+      "epoch": 0.42168879809329246,
+      "grad_norm": 1.487414002418518,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 2477
+    },
+    {
+      "epoch": 0.4218590398365679,
+      "grad_norm": 1.5315251350402832,
+      "learning_rate": 1e-06,
+      "loss": 0.0153,
+      "step": 2478
+    },
+    {
+      "epoch": 0.4220292815798434,
+      "grad_norm": 4.279947280883789,
+      "learning_rate": 1e-06,
+      "loss": 0.0537,
+      "step": 2479
+    },
+    {
+      "epoch": 0.42219952332311883,
+      "grad_norm": 1.396422266960144,
+      "learning_rate": 1e-06,
+      "loss": 0.0219,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4223697650663943,
+      "grad_norm": 1.583313226699829,
+      "learning_rate": 1e-06,
+      "loss": 0.0171,
+      "step": 2481
+    },
+    {
+      "epoch": 0.42254000680966974,
+      "grad_norm": 1.1320841312408447,
+      "learning_rate": 1e-06,
+      "loss": 0.013,
+      "step": 2482
+    },
+    {
+      "epoch": 0.4227102485529452,
+      "grad_norm": 1.1860630512237549,
+      "learning_rate": 1e-06,
+      "loss": 0.015,
+      "step": 2483
+    },
+    {
+      "epoch": 0.42288049029622066,
+      "grad_norm": 1.2758294343948364,
+      "learning_rate": 1e-06,
+      "loss": 0.0168,
+      "step": 2484
+    },
+    {
+      "epoch": 0.42305073203949606,
+      "grad_norm": 1.084784984588623,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 2485
+    },
+    {
+      "epoch": 0.4232209737827715,
+      "grad_norm": 1.0579909086227417,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 2486
+    },
+    {
+      "epoch": 0.42339121552604697,
+      "grad_norm": 1.5031887292861938,
+      "learning_rate": 1e-06,
+      "loss": 0.0159,
+      "step": 2487
+    },
+    {
+      "epoch": 0.42356145726932243,
+      "grad_norm": 1.2599958181381226,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 2488
+    },
+    {
+      "epoch": 0.4237316990125979,
+      "grad_norm": 1.1589609384536743,
+      "learning_rate": 1e-06,
+      "loss": 0.01,
+      "step": 2489
+    },
+    {
+      "epoch": 0.42390194075587334,
+      "grad_norm": 1.139289140701294,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 2490
+    },
+    {
+      "epoch": 0.4240721824991488,
+      "grad_norm": 1.1821390390396118,
+      "learning_rate": 1e-06,
+      "loss": 0.0129,
+      "step": 2491
+    },
+    {
+      "epoch": 0.42424242424242425,
+      "grad_norm": 1.2557470798492432,
+      "learning_rate": 1e-06,
+      "loss": 0.0114,
+      "step": 2492
+    },
+    {
+      "epoch": 0.4244126659856997,
+      "grad_norm": 1.3615427017211914,
+      "learning_rate": 1e-06,
+      "loss": 0.0142,
+      "step": 2493
+    },
+    {
+      "epoch": 0.42458290772897517,
+      "grad_norm": 1.4991490840911865,
+      "learning_rate": 1e-06,
+      "loss": 0.0142,
+      "step": 2494
+    },
+    {
+      "epoch": 0.42475314947225057,
+      "grad_norm": 1.1081904172897339,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 2495
+    },
+    {
+      "epoch": 0.424923391215526,
+      "grad_norm": 1.3824467658996582,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 2496
+    },
+    {
+      "epoch": 0.4250936329588015,
+      "grad_norm": 1.4526976346969604,
+      "learning_rate": 1e-06,
+      "loss": 0.0229,
+      "step": 2497
+    },
+    {
+      "epoch": 0.42526387470207694,
+      "grad_norm": 1.1925878524780273,
+      "learning_rate": 1e-06,
+      "loss": 0.014,
+      "step": 2498
+    },
+    {
+      "epoch": 0.4254341164453524,
+      "grad_norm": 1.7539995908737183,
+      "learning_rate": 1e-06,
+      "loss": 0.0295,
+      "step": 2499
+    },
+    {
+      "epoch": 0.42560435818862785,
+      "grad_norm": 1.2335952520370483,
+      "learning_rate": 1e-06,
+      "loss": 0.016,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4257745999319033,
+      "grad_norm": 1.1143063306808472,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 2501
+    },
+    {
+      "epoch": 0.42594484167517876,
+      "grad_norm": 1.0467058420181274,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 2502
+    },
+    {
+      "epoch": 0.4261150834184542,
+      "grad_norm": 1.118386149406433,
+      "learning_rate": 1e-06,
+      "loss": 0.0124,
+      "step": 2503
+    },
+    {
+      "epoch": 0.4262853251617297,
+      "grad_norm": 1.4646353721618652,
+      "learning_rate": 1e-06,
+      "loss": 0.0148,
+      "step": 2504
+    },
+    {
+      "epoch": 0.42645556690500513,
+      "grad_norm": 1.2250324487686157,
+      "learning_rate": 1e-06,
+      "loss": 0.0132,
+      "step": 2505
+    },
+    {
+      "epoch": 0.42662580864828054,
+      "grad_norm": 1.757466435432434,
+      "learning_rate": 1e-06,
+      "loss": 0.0219,
+      "step": 2506
+    },
+    {
+      "epoch": 0.426796050391556,
+      "grad_norm": 1.2650184631347656,
+      "learning_rate": 1e-06,
+      "loss": 0.0137,
+      "step": 2507
+    },
+    {
+      "epoch": 0.42696629213483145,
+      "grad_norm": 1.2739180326461792,
+      "learning_rate": 1e-06,
+      "loss": 0.011,
+      "step": 2508
+    },
+    {
+      "epoch": 0.4271365338781069,
+      "grad_norm": 1.3488551378250122,
+      "learning_rate": 1e-06,
+      "loss": 0.0144,
+      "step": 2509
+    },
+    {
+      "epoch": 0.42730677562138236,
+      "grad_norm": 1.4845232963562012,
+      "learning_rate": 1e-06,
+      "loss": 0.0196,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4274770173646578,
+      "grad_norm": 1.5139720439910889,
+      "learning_rate": 1e-06,
+      "loss": 0.0178,
+      "step": 2511
+    },
+    {
+      "epoch": 0.4276472591079333,
+      "grad_norm": 1.3799965381622314,
+      "learning_rate": 1e-06,
+      "loss": 0.0142,
+      "step": 2512
+    },
+    {
+      "epoch": 0.42781750085120873,
+      "grad_norm": 1.0673723220825195,
+      "learning_rate": 1e-06,
+      "loss": 0.0093,
+      "step": 2513
+    },
+    {
+      "epoch": 0.4279877425944842,
+      "grad_norm": 1.4284578561782837,
+      "learning_rate": 1e-06,
+      "loss": 0.0185,
+      "step": 2514
+    },
+    {
+      "epoch": 0.42815798433775965,
+      "grad_norm": 0.9315198659896851,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 2515
+    },
+    {
+      "epoch": 0.42832822608103505,
+      "grad_norm": 1.5946029424667358,
+      "learning_rate": 1e-06,
+      "loss": 0.0183,
+      "step": 2516
+    },
+    {
+      "epoch": 0.4284984678243105,
+      "grad_norm": 1.6120253801345825,
+      "learning_rate": 1e-06,
+      "loss": 0.0203,
+      "step": 2517
+    },
+    {
+      "epoch": 0.42866870956758596,
+      "grad_norm": 1.2495324611663818,
+      "learning_rate": 1e-06,
+      "loss": 0.0138,
+      "step": 2518
+    },
+    {
+      "epoch": 0.4288389513108614,
+      "grad_norm": 1.4058576822280884,
+      "learning_rate": 1e-06,
+      "loss": 0.0137,
+      "step": 2519
+    },
+    {
+      "epoch": 0.4290091930541369,
+      "grad_norm": 1.436219573020935,
+      "learning_rate": 1e-06,
+      "loss": 0.0151,
+      "step": 2520
+    },
+    {
+      "epoch": 0.42917943479741233,
+      "grad_norm": 1.9257749319076538,
+      "learning_rate": 1e-06,
+      "loss": 0.0199,
+      "step": 2521
+    },
+    {
+      "epoch": 0.4293496765406878,
+      "grad_norm": 1.195271611213684,
+      "learning_rate": 1e-06,
+      "loss": 0.0126,
+      "step": 2522
+    },
+    {
+      "epoch": 0.42951991828396324,
+      "grad_norm": 2.288235664367676,
+      "learning_rate": 1e-06,
+      "loss": 0.0229,
+      "step": 2523
+    },
+    {
+      "epoch": 0.4296901600272387,
+      "grad_norm": 1.20006263256073,
+      "learning_rate": 1e-06,
+      "loss": 0.011,
+      "step": 2524
+    },
+    {
+      "epoch": 0.42986040177051416,
+      "grad_norm": 1.8085074424743652,
+      "learning_rate": 1e-06,
+      "loss": 0.0246,
+      "step": 2525
+    },
+    {
+      "epoch": 0.43003064351378956,
+      "grad_norm": 1.3439559936523438,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 2526
+    },
+    {
+      "epoch": 0.430200885257065,
+      "grad_norm": 1.2932634353637695,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 2527
+    },
+    {
+      "epoch": 0.43037112700034047,
+      "grad_norm": 1.3444640636444092,
+      "learning_rate": 1e-06,
+      "loss": 0.0134,
+      "step": 2528
+    },
+    {
+      "epoch": 0.4305413687436159,
+      "grad_norm": 1.4149837493896484,
+      "learning_rate": 1e-06,
+      "loss": 0.0173,
+      "step": 2529
+    },
+    {
+      "epoch": 0.4307116104868914,
+      "grad_norm": 1.1546047925949097,
+      "learning_rate": 1e-06,
+      "loss": 0.0105,
+      "step": 2530
+    },
+    {
+      "epoch": 0.43088185223016684,
+      "grad_norm": 1.6338205337524414,
+      "learning_rate": 1e-06,
+      "loss": 0.0204,
+      "step": 2531
+    },
+    {
+      "epoch": 0.4310520939734423,
+      "grad_norm": 1.8675543069839478,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 2532
+    },
+    {
+      "epoch": 0.43122233571671775,
+      "grad_norm": 1.2775949239730835,
+      "learning_rate": 1e-06,
+      "loss": 0.0137,
+      "step": 2533
+    },
+    {
+      "epoch": 0.4313925774599932,
+      "grad_norm": 1.2907770872116089,
+      "learning_rate": 1e-06,
+      "loss": 0.0156,
+      "step": 2534
+    },
+    {
+      "epoch": 0.43156281920326867,
+      "grad_norm": 1.519423484802246,
+      "learning_rate": 1e-06,
+      "loss": 0.0142,
+      "step": 2535
+    },
+    {
+      "epoch": 0.43173306094654407,
+      "grad_norm": 1.4875999689102173,
+      "learning_rate": 1e-06,
+      "loss": 0.0177,
+      "step": 2536
+    },
+    {
+      "epoch": 0.4319033026898195,
+      "grad_norm": 1.0595074892044067,
+      "learning_rate": 1e-06,
+      "loss": 0.013,
+      "step": 2537
+    },
+    {
+      "epoch": 0.432073544433095,
+      "grad_norm": 1.461539626121521,
+      "learning_rate": 1e-06,
+      "loss": 0.0192,
+      "step": 2538
+    },
+    {
+      "epoch": 0.43224378617637044,
+      "grad_norm": 1.2863537073135376,
+      "learning_rate": 1e-06,
+      "loss": 0.0142,
+      "step": 2539
+    },
+    {
+      "epoch": 0.4324140279196459,
+      "grad_norm": 1.447748064994812,
+      "learning_rate": 1e-06,
+      "loss": 0.0147,
+      "step": 2540
+    },
+    {
+      "epoch": 0.43258426966292135,
+      "grad_norm": 1.0105310678482056,
+      "learning_rate": 1e-06,
+      "loss": 0.012,
+      "step": 2541
+    },
+    {
+      "epoch": 0.4327545114061968,
+      "grad_norm": 1.367261290550232,
+      "learning_rate": 1e-06,
+      "loss": 0.0161,
+      "step": 2542
+    },
+    {
+      "epoch": 0.43292475314947226,
+      "grad_norm": 1.2634276151657104,
+      "learning_rate": 1e-06,
+      "loss": 0.0121,
+      "step": 2543
+    },
+    {
+      "epoch": 0.4330949948927477,
+      "grad_norm": 1.2317240238189697,
+      "learning_rate": 1e-06,
+      "loss": 0.0136,
+      "step": 2544
+    },
+    {
+      "epoch": 0.4332652366360232,
+      "grad_norm": 1.2238634824752808,
+      "learning_rate": 1e-06,
+      "loss": 0.0106,
+      "step": 2545
+    },
+    {
+      "epoch": 0.4334354783792986,
+      "grad_norm": 1.3249047994613647,
+      "learning_rate": 1e-06,
+      "loss": 0.0133,
+      "step": 2546
+    },
+    {
+      "epoch": 0.43360572012257403,
+      "grad_norm": 1.049531102180481,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 2547
+    },
+    {
+      "epoch": 0.4337759618658495,
+      "grad_norm": 1.623108983039856,
+      "learning_rate": 1e-06,
+      "loss": 0.0204,
+      "step": 2548
+    },
+    {
+      "epoch": 0.43394620360912495,
+      "grad_norm": 1.2683053016662598,
+      "learning_rate": 1e-06,
+      "loss": 0.0173,
+      "step": 2549
+    },
+    {
+      "epoch": 0.4341164453524004,
+      "grad_norm": 1.6255425214767456,
+      "learning_rate": 1e-06,
+      "loss": 0.0163,
+      "step": 2550
+    },
+    {
+      "epoch": 0.43428668709567586,
+      "grad_norm": 1.1563000679016113,
+      "learning_rate": 1e-06,
+      "loss": 0.0145,
+      "step": 2551
+    },
+    {
+      "epoch": 0.4344569288389513,
+      "grad_norm": 1.3939322233200073,
+      "learning_rate": 1e-06,
+      "loss": 0.02,
+      "step": 2552
+    },
+    {
+      "epoch": 0.4346271705822268,
+      "grad_norm": 1.2949892282485962,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 2553
+    },
+    {
+      "epoch": 0.43479741232550223,
+      "grad_norm": 1.4135187864303589,
+      "learning_rate": 1e-06,
+      "loss": 0.017,
+      "step": 2554
+    },
+    {
+      "epoch": 0.4349676540687777,
+      "grad_norm": 1.1989250183105469,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 2555
+    },
+    {
+      "epoch": 0.4351378958120531,
+      "grad_norm": 1.180803894996643,
+      "learning_rate": 1e-06,
+      "loss": 0.0145,
+      "step": 2556
+    },
+    {
+      "epoch": 0.43530813755532854,
+      "grad_norm": 0.8694745302200317,
+      "learning_rate": 1e-06,
+      "loss": 0.0084,
+      "step": 2557
+    },
+    {
+      "epoch": 0.435478379298604,
+      "grad_norm": 1.489912986755371,
+      "learning_rate": 1e-06,
+      "loss": 0.0158,
+      "step": 2558
+    },
+    {
+      "epoch": 0.43564862104187946,
+      "grad_norm": 1.1703554391860962,
+      "learning_rate": 1e-06,
+      "loss": 0.011,
+      "step": 2559
+    },
+    {
+      "epoch": 0.4358188627851549,
+      "grad_norm": 1.3352656364440918,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 2560
+    },
+    {
+      "epoch": 0.43598910452843037,
+      "grad_norm": 2.0167603492736816,
+      "learning_rate": 1e-06,
+      "loss": 0.0132,
+      "step": 2561
+    },
+    {
+      "epoch": 0.43615934627170583,
+      "grad_norm": 1.5169191360473633,
+      "learning_rate": 1e-06,
+      "loss": 0.0266,
+      "step": 2562
+    },
+    {
+      "epoch": 0.4363295880149813,
+      "grad_norm": 1.1504062414169312,
+      "learning_rate": 1e-06,
+      "loss": 0.0124,
+      "step": 2563
+    },
+    {
+      "epoch": 0.43649982975825674,
+      "grad_norm": 1.3212984800338745,
+      "learning_rate": 1e-06,
+      "loss": 0.0156,
+      "step": 2564
+    },
+    {
+      "epoch": 0.4366700715015322,
+      "grad_norm": 1.149043321609497,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 2565
+    },
+    {
+      "epoch": 0.43684031324480765,
+      "grad_norm": 1.2085059881210327,
+      "learning_rate": 1e-06,
+      "loss": 0.012,
+      "step": 2566
+    },
+    {
+      "epoch": 0.43701055498808306,
+      "grad_norm": 1.2474228143692017,
+      "learning_rate": 1e-06,
+      "loss": 0.0123,
+      "step": 2567
+    },
+    {
+      "epoch": 0.4371807967313585,
+      "grad_norm": 1.2562228441238403,
+      "learning_rate": 1e-06,
+      "loss": 0.0118,
+      "step": 2568
+    },
+    {
+      "epoch": 0.43735103847463397,
+      "grad_norm": 1.5228354930877686,
+      "learning_rate": 1e-06,
+      "loss": 0.0136,
+      "step": 2569
+    },
+    {
+      "epoch": 0.4375212802179094,
+      "grad_norm": 1.6287275552749634,
+      "learning_rate": 1e-06,
+      "loss": 0.0179,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4376915219611849,
+      "grad_norm": 1.3553366661071777,
+      "learning_rate": 1e-06,
+      "loss": 0.0154,
+      "step": 2571
+    },
+    {
+      "epoch": 0.43786176370446034,
+      "grad_norm": 1.4978290796279907,
+      "learning_rate": 1e-06,
+      "loss": 0.0142,
+      "step": 2572
+    },
+    {
+      "epoch": 0.4380320054477358,
+      "grad_norm": 1.4585553407669067,
+      "learning_rate": 1e-06,
+      "loss": 0.0153,
+      "step": 2573
+    },
+    {
+      "epoch": 0.43820224719101125,
+      "grad_norm": 1.175268530845642,
+      "learning_rate": 1e-06,
+      "loss": 0.0151,
+      "step": 2574
+    },
+    {
+      "epoch": 0.4383724889342867,
+      "grad_norm": 1.2472771406173706,
+      "learning_rate": 1e-06,
+      "loss": 0.0172,
+      "step": 2575
+    },
+    {
+      "epoch": 0.43854273067756216,
+      "grad_norm": 1.109674096107483,
+      "learning_rate": 1e-06,
+      "loss": 0.0094,
+      "step": 2576
+    },
+    {
+      "epoch": 0.43871297242083757,
+      "grad_norm": 0.974772572517395,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 2577
+    },
+    {
+      "epoch": 0.438883214164113,
+      "grad_norm": 1.1081093549728394,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 2578
+    },
+    {
+      "epoch": 0.4390534559073885,
+      "grad_norm": 1.1672372817993164,
+      "learning_rate": 1e-06,
+      "loss": 0.0125,
+      "step": 2579
+    },
+    {
+      "epoch": 0.43922369765066394,
+      "grad_norm": 1.2188431024551392,
+      "learning_rate": 1e-06,
+      "loss": 0.0147,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4393939393939394,
+      "grad_norm": 1.0513874292373657,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 2581
+    },
+    {
+      "epoch": 0.43956418113721485,
+      "grad_norm": 1.2327890396118164,
+      "learning_rate": 1e-06,
+      "loss": 0.0129,
+      "step": 2582
+    },
+    {
+      "epoch": 0.4397344228804903,
+      "grad_norm": 2.443394422531128,
+      "learning_rate": 1e-06,
+      "loss": 0.0233,
+      "step": 2583
+    },
+    {
+      "epoch": 0.43990466462376576,
+      "grad_norm": 1.189650535583496,
+      "learning_rate": 1e-06,
+      "loss": 0.0211,
+      "step": 2584
+    },
+    {
+      "epoch": 0.4400749063670412,
+      "grad_norm": 1.22821843624115,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 2585
+    },
+    {
+      "epoch": 0.4402451481103167,
+      "grad_norm": 1.223922848701477,
+      "learning_rate": 1e-06,
+      "loss": 0.0112,
+      "step": 2586
+    },
+    {
+      "epoch": 0.4404153898535921,
+      "grad_norm": 1.3211407661437988,
+      "learning_rate": 1e-06,
+      "loss": 0.0141,
+      "step": 2587
+    },
+    {
+      "epoch": 0.44058563159686753,
+      "grad_norm": 1.2087403535842896,
+      "learning_rate": 1e-06,
+      "loss": 0.0152,
+      "step": 2588
+    },
+    {
+      "epoch": 0.440755873340143,
+      "grad_norm": 1.6649106740951538,
+      "learning_rate": 1e-06,
+      "loss": 0.0225,
+      "step": 2589
+    },
+    {
+      "epoch": 0.44092611508341845,
+      "grad_norm": 1.302064061164856,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4410963568266939,
+      "grad_norm": 1.2485460042953491,
+      "learning_rate": 1e-06,
+      "loss": 0.014,
+      "step": 2591
+    },
+    {
+      "epoch": 0.44126659856996936,
+      "grad_norm": 1.1560567617416382,
+      "learning_rate": 1e-06,
+      "loss": 0.0134,
+      "step": 2592
+    },
+    {
+      "epoch": 0.4414368403132448,
+      "grad_norm": 1.436091661453247,
+      "learning_rate": 1e-06,
+      "loss": 0.0132,
+      "step": 2593
+    },
+    {
+      "epoch": 0.4416070820565203,
+      "grad_norm": 1.1101467609405518,
+      "learning_rate": 1e-06,
+      "loss": 0.0131,
+      "step": 2594
+    },
+    {
+      "epoch": 0.44177732379979573,
+      "grad_norm": 1.2781106233596802,
+      "learning_rate": 1e-06,
+      "loss": 0.0217,
+      "step": 2595
+    },
+    {
+      "epoch": 0.4419475655430712,
+      "grad_norm": 1.1032869815826416,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 2596
+    },
+    {
+      "epoch": 0.4421178072863466,
+      "grad_norm": 1.4114645719528198,
+      "learning_rate": 1e-06,
+      "loss": 0.0118,
+      "step": 2597
+    },
+    {
+      "epoch": 0.44228804902962204,
+      "grad_norm": 1.307861566543579,
+      "learning_rate": 1e-06,
+      "loss": 0.0106,
+      "step": 2598
+    },
+    {
+      "epoch": 0.4424582907728975,
+      "grad_norm": 1.3546010255813599,
+      "learning_rate": 1e-06,
+      "loss": 0.0124,
+      "step": 2599
+    },
+    {
+      "epoch": 0.44262853251617296,
+      "grad_norm": 1.1858961582183838,
+      "learning_rate": 1e-06,
+      "loss": 0.0181,
+      "step": 2600
+    },
+    {
+      "epoch": 0.4427987742594484,
+      "grad_norm": 1.7051303386688232,
+      "learning_rate": 1e-06,
+      "loss": 0.0155,
+      "step": 2601
+    },
+    {
+      "epoch": 0.44296901600272387,
+      "grad_norm": 1.2881419658660889,
+      "learning_rate": 1e-06,
+      "loss": 0.015,
+      "step": 2602
+    },
+    {
+      "epoch": 0.4431392577459993,
+      "grad_norm": 1.2599304914474487,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 2603
+    },
+    {
+      "epoch": 0.4433094994892748,
+      "grad_norm": 1.2731688022613525,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 2604
+    },
+    {
+      "epoch": 0.44347974123255024,
+      "grad_norm": 1.2084245681762695,
+      "learning_rate": 1e-06,
+      "loss": 0.0129,
+      "step": 2605
+    },
+    {
+      "epoch": 0.4436499829758257,
+      "grad_norm": 1.2445482015609741,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 2606
+    },
+    {
+      "epoch": 0.4438202247191011,
+      "grad_norm": 1.8854734897613525,
+      "learning_rate": 1e-06,
+      "loss": 0.0193,
+      "step": 2607
+    },
+    {
+      "epoch": 0.44399046646237655,
+      "grad_norm": 1.4180908203125,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 2608
+    },
+    {
+      "epoch": 0.444160708205652,
+      "grad_norm": 1.7596639394760132,
+      "learning_rate": 1e-06,
+      "loss": 0.0215,
+      "step": 2609
+    },
+    {
+      "epoch": 0.44433094994892747,
+      "grad_norm": 1.8333537578582764,
+      "learning_rate": 1e-06,
+      "loss": 0.0171,
+      "step": 2610
+    },
+    {
+      "epoch": 0.4445011916922029,
+      "grad_norm": 1.5384416580200195,
+      "learning_rate": 1e-06,
+      "loss": 0.0121,
+      "step": 2611
+    },
+    {
+      "epoch": 0.4446714334354784,
+      "grad_norm": 1.6996245384216309,
+      "learning_rate": 1e-06,
+      "loss": 0.0296,
+      "step": 2612
+    },
+    {
+      "epoch": 0.44484167517875384,
+      "grad_norm": 1.2037662267684937,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 2613
+    },
+    {
+      "epoch": 0.4450119169220293,
+      "grad_norm": 1.2231796979904175,
+      "learning_rate": 1e-06,
+      "loss": 0.0118,
+      "step": 2614
+    },
+    {
+      "epoch": 0.44518215866530475,
+      "grad_norm": 1.1597298383712769,
+      "learning_rate": 1e-06,
+      "loss": 0.0147,
+      "step": 2615
+    },
+    {
+      "epoch": 0.4453524004085802,
+      "grad_norm": 1.1221568584442139,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 2616
+    },
+    {
+      "epoch": 0.4455226421518556,
+      "grad_norm": 3.223909378051758,
+      "learning_rate": 1e-06,
+      "loss": 0.0231,
+      "step": 2617
+    },
+    {
+      "epoch": 0.44569288389513106,
+      "grad_norm": 1.6201415061950684,
+      "learning_rate": 1e-06,
+      "loss": 0.0174,
+      "step": 2618
+    },
+    {
+      "epoch": 0.4458631256384065,
+      "grad_norm": 1.248681902885437,
+      "learning_rate": 1e-06,
+      "loss": 0.0129,
+      "step": 2619
+    },
+    {
+      "epoch": 0.446033367381682,
+      "grad_norm": 1.1186449527740479,
+      "learning_rate": 1e-06,
+      "loss": 0.0093,
+      "step": 2620
+    },
+    {
+      "epoch": 0.44620360912495743,
+      "grad_norm": 2.316514253616333,
+      "learning_rate": 1e-06,
+      "loss": 0.0223,
+      "step": 2621
+    },
+    {
+      "epoch": 0.4463738508682329,
+      "grad_norm": 1.214141607284546,
+      "learning_rate": 1e-06,
+      "loss": 0.01,
+      "step": 2622
+    },
+    {
+      "epoch": 0.44654409261150835,
+      "grad_norm": 1.4486373662948608,
+      "learning_rate": 1e-06,
+      "loss": 0.0292,
+      "step": 2623
+    },
+    {
+      "epoch": 0.4467143343547838,
+      "grad_norm": 1.2913445234298706,
+      "learning_rate": 1e-06,
+      "loss": 0.0123,
+      "step": 2624
+    },
+    {
+      "epoch": 0.44688457609805926,
+      "grad_norm": 1.2563960552215576,
+      "learning_rate": 1e-06,
+      "loss": 0.0222,
+      "step": 2625
+    },
+    {
+      "epoch": 0.4470548178413347,
+      "grad_norm": 1.4476072788238525,
+      "learning_rate": 1e-06,
+      "loss": 0.0188,
+      "step": 2626
+    },
+    {
+      "epoch": 0.4472250595846102,
+      "grad_norm": 1.5944808721542358,
+      "learning_rate": 1e-06,
+      "loss": 0.0161,
+      "step": 2627
+    },
+    {
+      "epoch": 0.4473953013278856,
+      "grad_norm": 1.1989669799804688,
+      "learning_rate": 1e-06,
+      "loss": 0.0134,
+      "step": 2628
+    },
+    {
+      "epoch": 0.44756554307116103,
+      "grad_norm": 1.5682440996170044,
+      "learning_rate": 1e-06,
+      "loss": 0.021,
+      "step": 2629
+    },
+    {
+      "epoch": 0.4477357848144365,
+      "grad_norm": 1.1739362478256226,
+      "learning_rate": 1e-06,
+      "loss": 0.0105,
+      "step": 2630
+    },
+    {
+      "epoch": 0.44790602655771194,
+      "grad_norm": 1.2502210140228271,
+      "learning_rate": 1e-06,
+      "loss": 0.019,
+      "step": 2631
+    },
+    {
+      "epoch": 0.4480762683009874,
+      "grad_norm": 1.1751649379730225,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 2632
+    },
+    {
+      "epoch": 0.44824651004426286,
+      "grad_norm": 1.1609954833984375,
+      "learning_rate": 1e-06,
+      "loss": 0.0186,
+      "step": 2633
+    },
+    {
+      "epoch": 0.4484167517875383,
+      "grad_norm": 1.4711358547210693,
+      "learning_rate": 1e-06,
+      "loss": 0.0197,
+      "step": 2634
+    },
+    {
+      "epoch": 0.44858699353081377,
+      "grad_norm": 1.154991865158081,
+      "learning_rate": 1e-06,
+      "loss": 0.0182,
+      "step": 2635
+    },
+    {
+      "epoch": 0.44875723527408923,
+      "grad_norm": 1.0405231714248657,
+      "learning_rate": 1e-06,
+      "loss": 0.0104,
+      "step": 2636
+    },
+    {
+      "epoch": 0.4489274770173647,
+      "grad_norm": 1.2849845886230469,
+      "learning_rate": 1e-06,
+      "loss": 0.0148,
+      "step": 2637
+    },
+    {
+      "epoch": 0.4490977187606401,
+      "grad_norm": 1.4233804941177368,
+      "learning_rate": 1e-06,
+      "loss": 0.0149,
+      "step": 2638
+    },
+    {
+      "epoch": 0.44926796050391554,
+      "grad_norm": 1.2203373908996582,
+      "learning_rate": 1e-06,
+      "loss": 0.0156,
+      "step": 2639
+    },
+    {
+      "epoch": 0.449438202247191,
+      "grad_norm": 1.1618257761001587,
+      "learning_rate": 1e-06,
+      "loss": 0.0114,
+      "step": 2640
+    },
+    {
+      "epoch": 0.44960844399046646,
+      "grad_norm": 1.307241439819336,
+      "learning_rate": 1e-06,
+      "loss": 0.0121,
+      "step": 2641
+    },
+    {
+      "epoch": 0.4497786857337419,
+      "grad_norm": 1.3096569776535034,
+      "learning_rate": 1e-06,
+      "loss": 0.016,
+      "step": 2642
+    },
+    {
+      "epoch": 0.44994892747701737,
+      "grad_norm": 2.0565989017486572,
+      "learning_rate": 1e-06,
+      "loss": 0.0164,
+      "step": 2643
+    },
+    {
+      "epoch": 0.4501191692202928,
+      "grad_norm": 1.3968826532363892,
+      "learning_rate": 1e-06,
+      "loss": 0.0118,
+      "step": 2644
+    },
+    {
+      "epoch": 0.4502894109635683,
+      "grad_norm": 1.3210657835006714,
+      "learning_rate": 1e-06,
+      "loss": 0.0126,
+      "step": 2645
+    },
+    {
+      "epoch": 0.45045965270684374,
+      "grad_norm": 1.1253306865692139,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 2646
+    },
+    {
+      "epoch": 0.4506298944501192,
+      "grad_norm": 1.6668893098831177,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 2647
+    },
+    {
+      "epoch": 0.4508001361933946,
+      "grad_norm": 1.3010493516921997,
+      "learning_rate": 1e-06,
+      "loss": 0.0151,
+      "step": 2648
+    },
+    {
+      "epoch": 0.45097037793667005,
+      "grad_norm": 1.4834797382354736,
+      "learning_rate": 1e-06,
+      "loss": 0.0105,
+      "step": 2649
+    },
+    {
+      "epoch": 0.4511406196799455,
+      "grad_norm": 1.3294862508773804,
+      "learning_rate": 1e-06,
+      "loss": 0.0119,
+      "step": 2650
+    },
+    {
+      "epoch": 0.45131086142322097,
+      "grad_norm": 1.2867186069488525,
+      "learning_rate": 1e-06,
+      "loss": 0.0124,
+      "step": 2651
+    },
+    {
+      "epoch": 0.4514811031664964,
+      "grad_norm": 1.1677923202514648,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 2652
+    },
+    {
+      "epoch": 0.4516513449097719,
+      "grad_norm": 1.1350395679473877,
+      "learning_rate": 1e-06,
+      "loss": 0.0131,
+      "step": 2653
+    },
+    {
+      "epoch": 0.45182158665304734,
+      "grad_norm": 1.4789958000183105,
+      "learning_rate": 1e-06,
+      "loss": 0.0141,
+      "step": 2654
+    },
+    {
+      "epoch": 0.4519918283963228,
+      "grad_norm": 1.325280785560608,
+      "learning_rate": 1e-06,
+      "loss": 0.0194,
+      "step": 2655
+    },
+    {
+      "epoch": 0.45216207013959825,
+      "grad_norm": 1.2727978229522705,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 2656
+    },
+    {
+      "epoch": 0.4523323118828737,
+      "grad_norm": 1.313781976699829,
+      "learning_rate": 1e-06,
+      "loss": 0.0177,
+      "step": 2657
+    },
+    {
+      "epoch": 0.4525025536261491,
+      "grad_norm": 1.6075435876846313,
+      "learning_rate": 1e-06,
+      "loss": 0.0196,
+      "step": 2658
+    },
+    {
+      "epoch": 0.45267279536942456,
+      "grad_norm": 1.2357505559921265,
+      "learning_rate": 1e-06,
+      "loss": 0.0122,
+      "step": 2659
+    },
+    {
+      "epoch": 0.4528430371127,
+      "grad_norm": 1.182496428489685,
+      "learning_rate": 1e-06,
+      "loss": 0.0181,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4530132788559755,
+      "grad_norm": 0.940984845161438,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 2661
+    },
+    {
+      "epoch": 0.45318352059925093,
+      "grad_norm": 1.304612636566162,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 2662
+    },
+    {
+      "epoch": 0.4533537623425264,
+      "grad_norm": 1.482349157333374,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 2663
+    },
+    {
+      "epoch": 0.45352400408580185,
+      "grad_norm": 1.1997534036636353,
+      "learning_rate": 1e-06,
+      "loss": 0.0157,
+      "step": 2664
+    },
+    {
+      "epoch": 0.4536942458290773,
+      "grad_norm": 1.2524585723876953,
+      "learning_rate": 1e-06,
+      "loss": 0.0124,
+      "step": 2665
+    },
+    {
+      "epoch": 0.45386448757235276,
+      "grad_norm": 1.139270544052124,
+      "learning_rate": 1e-06,
+      "loss": 0.0141,
+      "step": 2666
+    },
+    {
+      "epoch": 0.4540347293156282,
+      "grad_norm": 1.2429019212722778,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 2667
+    },
+    {
+      "epoch": 0.4542049710589036,
+      "grad_norm": 1.0963373184204102,
+      "learning_rate": 1e-06,
+      "loss": 0.0112,
+      "step": 2668
+    },
+    {
+      "epoch": 0.4543752128021791,
+      "grad_norm": 1.0817755460739136,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 2669
+    },
+    {
+      "epoch": 0.45454545454545453,
+      "grad_norm": 1.246390700340271,
+      "learning_rate": 1e-06,
+      "loss": 0.013,
+      "step": 2670
+    },
+    {
+      "epoch": 0.45471569628873,
+      "grad_norm": 1.3719927072525024,
+      "learning_rate": 1e-06,
+      "loss": 0.0196,
+      "step": 2671
+    },
+    {
+      "epoch": 0.45488593803200544,
+      "grad_norm": 1.0424156188964844,
+      "learning_rate": 1e-06,
+      "loss": 0.0122,
+      "step": 2672
+    },
+    {
+      "epoch": 0.4550561797752809,
+      "grad_norm": 1.1549937725067139,
+      "learning_rate": 1e-06,
+      "loss": 0.015,
+      "step": 2673
+    },
+    {
+      "epoch": 0.45522642151855636,
+      "grad_norm": 1.3112993240356445,
+      "learning_rate": 1e-06,
+      "loss": 0.0137,
+      "step": 2674
+    },
+    {
+      "epoch": 0.4553966632618318,
+      "grad_norm": 1.9749765396118164,
+      "learning_rate": 1e-06,
+      "loss": 0.0143,
+      "step": 2675
+    },
+    {
+      "epoch": 0.45556690500510727,
+      "grad_norm": 1.4129223823547363,
+      "learning_rate": 1e-06,
+      "loss": 0.0171,
+      "step": 2676
+    },
+    {
+      "epoch": 0.4557371467483827,
+      "grad_norm": 1.2779748439788818,
+      "learning_rate": 1e-06,
+      "loss": 0.0153,
+      "step": 2677
+    },
+    {
+      "epoch": 0.4559073884916581,
+      "grad_norm": 4.037950038909912,
+      "learning_rate": 1e-06,
+      "loss": 0.0177,
+      "step": 2678
+    },
+    {
+      "epoch": 0.4560776302349336,
+      "grad_norm": 1.1884809732437134,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 2679
+    },
+    {
+      "epoch": 0.45624787197820904,
+      "grad_norm": 1.5136353969573975,
+      "learning_rate": 1e-06,
+      "loss": 0.0176,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4564181137214845,
+      "grad_norm": 5.290273189544678,
+      "learning_rate": 1e-06,
+      "loss": 0.0812,
+      "step": 2681
+    },
+    {
+      "epoch": 0.45658835546475995,
+      "grad_norm": 1.0113693475723267,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 2682
+    },
+    {
+      "epoch": 0.4567585972080354,
+      "grad_norm": 1.279430627822876,
+      "learning_rate": 1e-06,
+      "loss": 0.0182,
+      "step": 2683
+    },
+    {
+      "epoch": 0.45692883895131087,
+      "grad_norm": 1.4505813121795654,
+      "learning_rate": 1e-06,
+      "loss": 0.0156,
+      "step": 2684
+    },
+    {
+      "epoch": 0.4570990806945863,
+      "grad_norm": 1.6365201473236084,
+      "learning_rate": 1e-06,
+      "loss": 0.017,
+      "step": 2685
+    },
+    {
+      "epoch": 0.4572693224378618,
+      "grad_norm": 1.1942694187164307,
+      "learning_rate": 1e-06,
+      "loss": 0.0119,
+      "step": 2686
+    },
+    {
+      "epoch": 0.45743956418113724,
+      "grad_norm": 1.1934460401535034,
+      "learning_rate": 1e-06,
+      "loss": 0.0172,
+      "step": 2687
+    },
+    {
+      "epoch": 0.4576098059244127,
+      "grad_norm": 1.3042709827423096,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 2688
+    },
+    {
+      "epoch": 0.4577800476676881,
+      "grad_norm": 1.2046012878417969,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 2689
+    },
+    {
+      "epoch": 0.45795028941096355,
+      "grad_norm": 1.5803378820419312,
+      "learning_rate": 1e-06,
+      "loss": 0.0146,
+      "step": 2690
+    },
+    {
+      "epoch": 0.458120531154239,
+      "grad_norm": 1.2354884147644043,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 2691
+    },
+    {
+      "epoch": 0.45829077289751446,
+      "grad_norm": 2.588909864425659,
+      "learning_rate": 1e-06,
+      "loss": 0.026,
+      "step": 2692
+    },
+    {
+      "epoch": 0.4584610146407899,
+      "grad_norm": 1.2800123691558838,
+      "learning_rate": 1e-06,
+      "loss": 0.012,
+      "step": 2693
+    },
+    {
+      "epoch": 0.4586312563840654,
+      "grad_norm": 1.6324779987335205,
+      "learning_rate": 1e-06,
+      "loss": 0.0157,
+      "step": 2694
+    },
+    {
+      "epoch": 0.45880149812734083,
+      "grad_norm": 1.0287292003631592,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 2695
+    },
+    {
+      "epoch": 0.4589717398706163,
+      "grad_norm": 1.4863289594650269,
+      "learning_rate": 1e-06,
+      "loss": 0.0166,
+      "step": 2696
+    },
+    {
+      "epoch": 0.45914198161389175,
+      "grad_norm": 1.7314716577529907,
+      "learning_rate": 1e-06,
+      "loss": 0.0222,
+      "step": 2697
+    },
+    {
+      "epoch": 0.4593122233571672,
+      "grad_norm": 1.315182089805603,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 2698
+    },
+    {
+      "epoch": 0.4594824651004426,
+      "grad_norm": 1.8149510622024536,
+      "learning_rate": 1e-06,
+      "loss": 0.0246,
+      "step": 2699
+    },
+    {
+      "epoch": 0.45965270684371806,
+      "grad_norm": 1.528834581375122,
+      "learning_rate": 1e-06,
+      "loss": 0.0249,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4598229485869935,
+      "grad_norm": 1.3489001989364624,
+      "learning_rate": 1e-06,
+      "loss": 0.021,
+      "step": 2701
+    },
+    {
+      "epoch": 0.459993190330269,
+      "grad_norm": 1.1386735439300537,
+      "learning_rate": 1e-06,
+      "loss": 0.0146,
+      "step": 2702
+    },
+    {
+      "epoch": 0.46016343207354443,
+      "grad_norm": 0.9807332754135132,
+      "learning_rate": 1e-06,
+      "loss": 0.0097,
+      "step": 2703
+    },
+    {
+      "epoch": 0.4603336738168199,
+      "grad_norm": 1.237453579902649,
+      "learning_rate": 1e-06,
+      "loss": 0.018,
+      "step": 2704
+    },
+    {
+      "epoch": 0.46050391556009534,
+      "grad_norm": 1.3836872577667236,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 2705
+    },
+    {
+      "epoch": 0.4606741573033708,
+      "grad_norm": 1.1257848739624023,
+      "learning_rate": 1e-06,
+      "loss": 0.0132,
+      "step": 2706
+    },
+    {
+      "epoch": 0.46084439904664626,
+      "grad_norm": 1.412534475326538,
+      "learning_rate": 1e-06,
+      "loss": 0.0153,
+      "step": 2707
+    },
+    {
+      "epoch": 0.4610146407899217,
+      "grad_norm": 1.4209263324737549,
+      "learning_rate": 1e-06,
+      "loss": 0.0144,
+      "step": 2708
+    },
+    {
+      "epoch": 0.4611848825331971,
+      "grad_norm": 1.466731309890747,
+      "learning_rate": 1e-06,
+      "loss": 0.0163,
+      "step": 2709
+    },
+    {
+      "epoch": 0.4613551242764726,
+      "grad_norm": 2.445279121398926,
+      "learning_rate": 1e-06,
+      "loss": 0.0225,
+      "step": 2710
+    },
+    {
+      "epoch": 0.46152536601974803,
+      "grad_norm": 1.6007717847824097,
+      "learning_rate": 1e-06,
+      "loss": 0.0136,
+      "step": 2711
+    },
+    {
+      "epoch": 0.4616956077630235,
+      "grad_norm": 4.2332844734191895,
+      "learning_rate": 1e-06,
+      "loss": 0.0242,
+      "step": 2712
+    },
+    {
+      "epoch": 0.46186584950629894,
+      "grad_norm": 1.627395749092102,
+      "learning_rate": 1e-06,
+      "loss": 0.0187,
+      "step": 2713
+    },
+    {
+      "epoch": 0.4620360912495744,
+      "grad_norm": 1.2684941291809082,
+      "learning_rate": 1e-06,
+      "loss": 0.0114,
+      "step": 2714
+    },
+    {
+      "epoch": 0.46220633299284986,
+      "grad_norm": 1.287532925605774,
+      "learning_rate": 1e-06,
+      "loss": 0.012,
+      "step": 2715
+    },
+    {
+      "epoch": 0.4623765747361253,
+      "grad_norm": 1.6071957349777222,
+      "learning_rate": 1e-06,
+      "loss": 0.0171,
+      "step": 2716
+    },
+    {
+      "epoch": 0.46254681647940077,
+      "grad_norm": 1.3653993606567383,
+      "learning_rate": 1e-06,
+      "loss": 0.0153,
+      "step": 2717
+    },
+    {
+      "epoch": 0.4627170582226762,
+      "grad_norm": 0.9938444495201111,
+      "learning_rate": 1e-06,
+      "loss": 0.0084,
+      "step": 2718
+    },
+    {
+      "epoch": 0.4628872999659516,
+      "grad_norm": 1.3033769130706787,
+      "learning_rate": 1e-06,
+      "loss": 0.0105,
+      "step": 2719
+    },
+    {
+      "epoch": 0.4630575417092271,
+      "grad_norm": 1.2116801738739014,
+      "learning_rate": 1e-06,
+      "loss": 0.0147,
+      "step": 2720
+    },
+    {
+      "epoch": 0.46322778345250254,
+      "grad_norm": 0.969497561454773,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 2721
+    },
+    {
+      "epoch": 0.463398025195778,
+      "grad_norm": 1.1007535457611084,
+      "learning_rate": 1e-06,
+      "loss": 0.0151,
+      "step": 2722
+    },
+    {
+      "epoch": 0.46356826693905345,
+      "grad_norm": 1.2838865518569946,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 2723
+    },
+    {
+      "epoch": 0.4637385086823289,
+      "grad_norm": 1.1375126838684082,
+      "learning_rate": 1e-06,
+      "loss": 0.0142,
+      "step": 2724
+    },
+    {
+      "epoch": 0.46390875042560437,
+      "grad_norm": 0.9952174425125122,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 2725
+    },
+    {
+      "epoch": 0.4640789921688798,
+      "grad_norm": 1.0825743675231934,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 2726
+    },
+    {
+      "epoch": 0.4642492339121553,
+      "grad_norm": 1.658443808555603,
+      "learning_rate": 1e-06,
+      "loss": 0.0286,
+      "step": 2727
+    },
+    {
+      "epoch": 0.46441947565543074,
+      "grad_norm": 1.2224160432815552,
+      "learning_rate": 1e-06,
+      "loss": 0.0171,
+      "step": 2728
+    },
+    {
+      "epoch": 0.46458971739870614,
+      "grad_norm": 1.210042119026184,
+      "learning_rate": 1e-06,
+      "loss": 0.0128,
+      "step": 2729
+    },
+    {
+      "epoch": 0.4647599591419816,
+      "grad_norm": 1.4320639371871948,
+      "learning_rate": 1e-06,
+      "loss": 0.0099,
+      "step": 2730
+    },
+    {
+      "epoch": 0.46493020088525705,
+      "grad_norm": 1.2222720384597778,
+      "learning_rate": 1e-06,
+      "loss": 0.0134,
+      "step": 2731
+    },
+    {
+      "epoch": 0.4651004426285325,
+      "grad_norm": 1.5248254537582397,
+      "learning_rate": 1e-06,
+      "loss": 0.0136,
+      "step": 2732
+    },
+    {
+      "epoch": 0.46527068437180796,
+      "grad_norm": 1.599933385848999,
+      "learning_rate": 1e-06,
+      "loss": 0.0173,
+      "step": 2733
+    },
+    {
+      "epoch": 0.4654409261150834,
+      "grad_norm": 1.824179768562317,
+      "learning_rate": 1e-06,
+      "loss": 0.0158,
+      "step": 2734
+    },
+    {
+      "epoch": 0.4656111678583589,
+      "grad_norm": 1.516343355178833,
+      "learning_rate": 1e-06,
+      "loss": 0.0209,
+      "step": 2735
+    },
+    {
+      "epoch": 0.46578140960163433,
+      "grad_norm": 1.768630862236023,
+      "learning_rate": 1e-06,
+      "loss": 0.0191,
+      "step": 2736
+    },
+    {
+      "epoch": 0.4659516513449098,
+      "grad_norm": 1.7268173694610596,
+      "learning_rate": 1e-06,
+      "loss": 0.0312,
+      "step": 2737
+    },
+    {
+      "epoch": 0.46612189308818525,
+      "grad_norm": 1.7336803674697876,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 2738
+    },
+    {
+      "epoch": 0.46629213483146065,
+      "grad_norm": 1.4071462154388428,
+      "learning_rate": 1e-06,
+      "loss": 0.0165,
+      "step": 2739
+    },
+    {
+      "epoch": 0.4664623765747361,
+      "grad_norm": 1.6221487522125244,
+      "learning_rate": 1e-06,
+      "loss": 0.0204,
+      "step": 2740
+    },
+    {
+      "epoch": 0.46663261831801156,
+      "grad_norm": 1.0625100135803223,
+      "learning_rate": 1e-06,
+      "loss": 0.018,
+      "step": 2741
+    },
+    {
+      "epoch": 0.466802860061287,
+      "grad_norm": 3.000713348388672,
+      "learning_rate": 1e-06,
+      "loss": 0.0445,
+      "step": 2742
+    },
+    {
+      "epoch": 0.4669731018045625,
+      "grad_norm": 1.5852463245391846,
+      "learning_rate": 1e-06,
+      "loss": 0.0152,
+      "step": 2743
+    },
+    {
+      "epoch": 0.46714334354783793,
+      "grad_norm": 1.2370667457580566,
+      "learning_rate": 1e-06,
+      "loss": 0.0136,
+      "step": 2744
+    },
+    {
+      "epoch": 0.4673135852911134,
+      "grad_norm": 1.1240363121032715,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 2745
+    },
+    {
+      "epoch": 0.46748382703438884,
+      "grad_norm": 1.7612371444702148,
+      "learning_rate": 1e-06,
+      "loss": 0.0312,
+      "step": 2746
+    },
+    {
+      "epoch": 0.4676540687776643,
+      "grad_norm": 1.0981045961380005,
+      "learning_rate": 1e-06,
+      "loss": 0.016,
+      "step": 2747
+    },
+    {
+      "epoch": 0.46782431052093976,
+      "grad_norm": 1.3208484649658203,
+      "learning_rate": 1e-06,
+      "loss": 0.0167,
+      "step": 2748
+    },
+    {
+      "epoch": 0.46799455226421516,
+      "grad_norm": 1.1048396825790405,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 2749
+    },
+    {
+      "epoch": 0.4681647940074906,
+      "grad_norm": 1.2207963466644287,
+      "learning_rate": 1e-06,
+      "loss": 0.0188,
+      "step": 2750
+    },
+    {
+      "epoch": 0.46833503575076607,
+      "grad_norm": 1.3159321546554565,
+      "learning_rate": 1e-06,
+      "loss": 0.0136,
+      "step": 2751
+    },
+    {
+      "epoch": 0.4685052774940415,
+      "grad_norm": 1.4273492097854614,
+      "learning_rate": 1e-06,
+      "loss": 0.013,
+      "step": 2752
+    },
+    {
+      "epoch": 0.468675519237317,
+      "grad_norm": 1.4401754140853882,
+      "learning_rate": 1e-06,
+      "loss": 0.0151,
+      "step": 2753
+    },
+    {
+      "epoch": 0.46884576098059244,
+      "grad_norm": 1.262450098991394,
+      "learning_rate": 1e-06,
+      "loss": 0.0145,
+      "step": 2754
+    },
+    {
+      "epoch": 0.4690160027238679,
+      "grad_norm": 1.0684226751327515,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 2755
+    },
+    {
+      "epoch": 0.46918624446714335,
+      "grad_norm": 1.1081675291061401,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 2756
+    },
+    {
+      "epoch": 0.4693564862104188,
+      "grad_norm": 1.0820542573928833,
+      "learning_rate": 1e-06,
+      "loss": 0.0112,
+      "step": 2757
+    },
+    {
+      "epoch": 0.46952672795369427,
+      "grad_norm": 1.4552348852157593,
+      "learning_rate": 1e-06,
+      "loss": 0.0123,
+      "step": 2758
+    },
+    {
+      "epoch": 0.4696969696969697,
+      "grad_norm": 1.2042700052261353,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 2759
+    },
+    {
+      "epoch": 0.4698672114402451,
+      "grad_norm": 1.2401795387268066,
+      "learning_rate": 1e-06,
+      "loss": 0.0119,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4700374531835206,
+      "grad_norm": 1.3959617614746094,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 2761
+    },
+    {
+      "epoch": 0.47020769492679604,
+      "grad_norm": 1.220746636390686,
+      "learning_rate": 1e-06,
+      "loss": 0.0172,
+      "step": 2762
+    },
+    {
+      "epoch": 0.4703779366700715,
+      "grad_norm": 1.241463303565979,
+      "learning_rate": 1e-06,
+      "loss": 0.0131,
+      "step": 2763
+    },
+    {
+      "epoch": 0.47054817841334695,
+      "grad_norm": 1.2252508401870728,
+      "learning_rate": 1e-06,
+      "loss": 0.013,
+      "step": 2764
+    },
+    {
+      "epoch": 0.4707184201566224,
+      "grad_norm": 1.9142897129058838,
+      "learning_rate": 1e-06,
+      "loss": 0.028,
+      "step": 2765
+    },
+    {
+      "epoch": 0.47088866189989786,
+      "grad_norm": 1.1304916143417358,
+      "learning_rate": 1e-06,
+      "loss": 0.0119,
+      "step": 2766
+    },
+    {
+      "epoch": 0.4710589036431733,
+      "grad_norm": 1.247910499572754,
+      "learning_rate": 1e-06,
+      "loss": 0.0179,
+      "step": 2767
+    },
+    {
+      "epoch": 0.4712291453864488,
+      "grad_norm": 1.3291040658950806,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 2768
+    },
+    {
+      "epoch": 0.47139938712972423,
+      "grad_norm": 1.139250636100769,
+      "learning_rate": 1e-06,
+      "loss": 0.0146,
+      "step": 2769
+    },
+    {
+      "epoch": 0.47156962887299964,
+      "grad_norm": 1.6703858375549316,
+      "learning_rate": 1e-06,
+      "loss": 0.0184,
+      "step": 2770
+    },
+    {
+      "epoch": 0.4717398706162751,
+      "grad_norm": 1.6940549612045288,
+      "learning_rate": 1e-06,
+      "loss": 0.0164,
+      "step": 2771
+    },
+    {
+      "epoch": 0.47191011235955055,
+      "grad_norm": 1.24887216091156,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 2772
+    },
+    {
+      "epoch": 0.472080354102826,
+      "grad_norm": 1.0762343406677246,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 2773
+    },
+    {
+      "epoch": 0.47225059584610146,
+      "grad_norm": 1.1544651985168457,
+      "learning_rate": 1e-06,
+      "loss": 0.0129,
+      "step": 2774
+    },
+    {
+      "epoch": 0.4724208375893769,
+      "grad_norm": 1.3130018711090088,
+      "learning_rate": 1e-06,
+      "loss": 0.0144,
+      "step": 2775
+    },
+    {
+      "epoch": 0.4725910793326524,
+      "grad_norm": 1.620080590248108,
+      "learning_rate": 1e-06,
+      "loss": 0.0181,
+      "step": 2776
+    },
+    {
+      "epoch": 0.47276132107592783,
+      "grad_norm": 1.6613308191299438,
+      "learning_rate": 1e-06,
+      "loss": 0.0118,
+      "step": 2777
+    },
+    {
+      "epoch": 0.4729315628192033,
+      "grad_norm": 1.066480040550232,
+      "learning_rate": 1e-06,
+      "loss": 0.0077,
+      "step": 2778
+    },
+    {
+      "epoch": 0.47310180456247874,
+      "grad_norm": 1.109659194946289,
+      "learning_rate": 1e-06,
+      "loss": 0.0123,
+      "step": 2779
+    },
+    {
+      "epoch": 0.47327204630575415,
+      "grad_norm": 1.551084041595459,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4734422880490296,
+      "grad_norm": 1.2061805725097656,
+      "learning_rate": 1e-06,
+      "loss": 0.0166,
+      "step": 2781
+    },
+    {
+      "epoch": 0.47361252979230506,
+      "grad_norm": 1.9850165843963623,
+      "learning_rate": 1e-06,
+      "loss": 0.0128,
+      "step": 2782
+    },
+    {
+      "epoch": 0.4737827715355805,
+      "grad_norm": 1.3873836994171143,
+      "learning_rate": 1e-06,
+      "loss": 0.0146,
+      "step": 2783
+    },
+    {
+      "epoch": 0.473953013278856,
+      "grad_norm": 1.0683759450912476,
+      "learning_rate": 1e-06,
+      "loss": 0.0106,
+      "step": 2784
+    },
+    {
+      "epoch": 0.47412325502213143,
+      "grad_norm": 0.9684468507766724,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 2785
+    },
+    {
+      "epoch": 0.4742934967654069,
+      "grad_norm": 1.323331594467163,
+      "learning_rate": 1e-06,
+      "loss": 0.0156,
+      "step": 2786
+    },
+    {
+      "epoch": 0.47446373850868234,
+      "grad_norm": 1.1693027019500732,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 2787
+    },
+    {
+      "epoch": 0.4746339802519578,
+      "grad_norm": 0.9231655597686768,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 2788
+    },
+    {
+      "epoch": 0.47480422199523326,
+      "grad_norm": 1.9782781600952148,
+      "learning_rate": 1e-06,
+      "loss": 0.0199,
+      "step": 2789
+    },
+    {
+      "epoch": 0.47497446373850866,
+      "grad_norm": 1.5112477540969849,
+      "learning_rate": 1e-06,
+      "loss": 0.0166,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4751447054817841,
+      "grad_norm": 1.178739309310913,
+      "learning_rate": 1e-06,
+      "loss": 0.012,
+      "step": 2791
+    },
+    {
+      "epoch": 0.47531494722505957,
+      "grad_norm": 2.3978826999664307,
+      "learning_rate": 1e-06,
+      "loss": 0.0155,
+      "step": 2792
+    },
+    {
+      "epoch": 0.475485188968335,
+      "grad_norm": 1.3973369598388672,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 2793
+    },
+    {
+      "epoch": 0.4756554307116105,
+      "grad_norm": 1.1762423515319824,
+      "learning_rate": 1e-06,
+      "loss": 0.0124,
+      "step": 2794
+    },
+    {
+      "epoch": 0.47582567245488594,
+      "grad_norm": 1.2299197912216187,
+      "learning_rate": 1e-06,
+      "loss": 0.0118,
+      "step": 2795
+    },
+    {
+      "epoch": 0.4759959141981614,
+      "grad_norm": 1.3088619709014893,
+      "learning_rate": 1e-06,
+      "loss": 0.0174,
+      "step": 2796
+    },
+    {
+      "epoch": 0.47616615594143685,
+      "grad_norm": 1.0132900476455688,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 2797
+    },
+    {
+      "epoch": 0.4763363976847123,
+      "grad_norm": 1.129879355430603,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 2798
+    },
+    {
+      "epoch": 0.47650663942798777,
+      "grad_norm": 0.9757645130157471,
+      "learning_rate": 1e-06,
+      "loss": 0.0132,
+      "step": 2799
+    },
+    {
+      "epoch": 0.47667688117126317,
+      "grad_norm": 1.4585630893707275,
+      "learning_rate": 1e-06,
+      "loss": 0.017,
+      "step": 2800
+    },
+    {
+      "epoch": 0.4768471229145386,
+      "grad_norm": 1.0814754962921143,
+      "learning_rate": 1e-06,
+      "loss": 0.0162,
+      "step": 2801
+    },
+    {
+      "epoch": 0.4770173646578141,
+      "grad_norm": 1.095276117324829,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 2802
+    },
+    {
+      "epoch": 0.47718760640108954,
+      "grad_norm": 1.1605823040008545,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 2803
+    },
+    {
+      "epoch": 0.477357848144365,
+      "grad_norm": 0.9512302279472351,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 2804
+    },
+    {
+      "epoch": 0.47752808988764045,
+      "grad_norm": 0.8988835215568542,
+      "learning_rate": 1e-06,
+      "loss": 0.0097,
+      "step": 2805
+    },
+    {
+      "epoch": 0.4776983316309159,
+      "grad_norm": 1.4442946910858154,
+      "learning_rate": 1e-06,
+      "loss": 0.017,
+      "step": 2806
+    },
+    {
+      "epoch": 0.47786857337419136,
+      "grad_norm": 1.1615439653396606,
+      "learning_rate": 1e-06,
+      "loss": 0.0118,
+      "step": 2807
+    },
+    {
+      "epoch": 0.4780388151174668,
+      "grad_norm": 1.2446208000183105,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 2808
+    },
+    {
+      "epoch": 0.4782090568607423,
+      "grad_norm": 1.1819733381271362,
+      "learning_rate": 1e-06,
+      "loss": 0.0126,
+      "step": 2809
+    },
+    {
+      "epoch": 0.4783792986040177,
+      "grad_norm": 1.3760933876037598,
+      "learning_rate": 1e-06,
+      "loss": 0.011,
+      "step": 2810
+    },
+    {
+      "epoch": 0.47854954034729313,
+      "grad_norm": 1.2199891805648804,
+      "learning_rate": 1e-06,
+      "loss": 0.011,
+      "step": 2811
+    },
+    {
+      "epoch": 0.4787197820905686,
+      "grad_norm": 1.2610541582107544,
+      "learning_rate": 1e-06,
+      "loss": 0.0119,
+      "step": 2812
+    },
+    {
+      "epoch": 0.47889002383384405,
+      "grad_norm": 1.4175294637680054,
+      "learning_rate": 1e-06,
+      "loss": 0.0105,
+      "step": 2813
+    },
+    {
+      "epoch": 0.4790602655771195,
+      "grad_norm": 1.0820069313049316,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 2814
+    },
+    {
+      "epoch": 0.47923050732039496,
+      "grad_norm": 1.236606478691101,
+      "learning_rate": 1e-06,
+      "loss": 0.015,
+      "step": 2815
+    },
+    {
+      "epoch": 0.4794007490636704,
+      "grad_norm": 1.2256829738616943,
+      "learning_rate": 1e-06,
+      "loss": 0.0121,
+      "step": 2816
+    },
+    {
+      "epoch": 0.4795709908069459,
+      "grad_norm": 1.4265694618225098,
+      "learning_rate": 1e-06,
+      "loss": 0.0137,
+      "step": 2817
+    },
+    {
+      "epoch": 0.47974123255022133,
+      "grad_norm": 1.2129721641540527,
+      "learning_rate": 1e-06,
+      "loss": 0.0104,
+      "step": 2818
+    },
+    {
+      "epoch": 0.4799114742934968,
+      "grad_norm": 1.3204402923583984,
+      "learning_rate": 1e-06,
+      "loss": 0.0143,
+      "step": 2819
+    },
+    {
+      "epoch": 0.48008171603677224,
+      "grad_norm": 2.682424783706665,
+      "learning_rate": 1e-06,
+      "loss": 0.0244,
+      "step": 2820
+    },
+    {
+      "epoch": 0.48025195778004764,
+      "grad_norm": 1.4030735492706299,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 2821
+    },
+    {
+      "epoch": 0.4804221995233231,
+      "grad_norm": 2.13521671295166,
+      "learning_rate": 1e-06,
+      "loss": 0.02,
+      "step": 2822
+    },
+    {
+      "epoch": 0.48059244126659856,
+      "grad_norm": 1.1667373180389404,
+      "learning_rate": 1e-06,
+      "loss": 0.0134,
+      "step": 2823
+    },
+    {
+      "epoch": 0.480762683009874,
+      "grad_norm": 1.3074268102645874,
+      "learning_rate": 1e-06,
+      "loss": 0.0114,
+      "step": 2824
+    },
+    {
+      "epoch": 0.48093292475314947,
+      "grad_norm": 1.321334719657898,
+      "learning_rate": 1e-06,
+      "loss": 0.0118,
+      "step": 2825
+    },
+    {
+      "epoch": 0.4811031664964249,
+      "grad_norm": 0.9559478759765625,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 2826
+    },
+    {
+      "epoch": 0.4812734082397004,
+      "grad_norm": 0.8174387812614441,
+      "learning_rate": 1e-06,
+      "loss": 0.007,
+      "step": 2827
+    },
+    {
+      "epoch": 0.48144364998297584,
+      "grad_norm": 1.3920475244522095,
+      "learning_rate": 1e-06,
+      "loss": 0.0136,
+      "step": 2828
+    },
+    {
+      "epoch": 0.4816138917262513,
+      "grad_norm": 1.5686707496643066,
+      "learning_rate": 1e-06,
+      "loss": 0.0187,
+      "step": 2829
+    },
+    {
+      "epoch": 0.48178413346952675,
+      "grad_norm": 1.0519627332687378,
+      "learning_rate": 1e-06,
+      "loss": 0.0105,
+      "step": 2830
+    },
+    {
+      "epoch": 0.48195437521280216,
+      "grad_norm": 1.2499573230743408,
+      "learning_rate": 1e-06,
+      "loss": 0.0158,
+      "step": 2831
+    },
+    {
+      "epoch": 0.4821246169560776,
+      "grad_norm": 1.9623509645462036,
+      "learning_rate": 1e-06,
+      "loss": 0.0192,
+      "step": 2832
+    },
+    {
+      "epoch": 0.48229485869935307,
+      "grad_norm": 1.0774935483932495,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 2833
+    },
+    {
+      "epoch": 0.4824651004426285,
+      "grad_norm": 1.0941082239151,
+      "learning_rate": 1e-06,
+      "loss": 0.0097,
+      "step": 2834
+    },
+    {
+      "epoch": 0.482635342185904,
+      "grad_norm": 1.3845523595809937,
+      "learning_rate": 1e-06,
+      "loss": 0.0154,
+      "step": 2835
+    },
+    {
+      "epoch": 0.48280558392917944,
+      "grad_norm": 1.569908857345581,
+      "learning_rate": 1e-06,
+      "loss": 0.018,
+      "step": 2836
+    },
+    {
+      "epoch": 0.4829758256724549,
+      "grad_norm": 1.1683586835861206,
+      "learning_rate": 1e-06,
+      "loss": 0.0151,
+      "step": 2837
+    },
+    {
+      "epoch": 0.48314606741573035,
+      "grad_norm": 1.277343511581421,
+      "learning_rate": 1e-06,
+      "loss": 0.0166,
+      "step": 2838
+    },
+    {
+      "epoch": 0.4833163091590058,
+      "grad_norm": 1.2092841863632202,
+      "learning_rate": 1e-06,
+      "loss": 0.0105,
+      "step": 2839
+    },
+    {
+      "epoch": 0.48348655090228126,
+      "grad_norm": 1.1916158199310303,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 2840
+    },
+    {
+      "epoch": 0.48365679264555667,
+      "grad_norm": 1.280978798866272,
+      "learning_rate": 1e-06,
+      "loss": 0.0123,
+      "step": 2841
+    },
+    {
+      "epoch": 0.4838270343888321,
+      "grad_norm": 1.2461068630218506,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 2842
+    },
+    {
+      "epoch": 0.4839972761321076,
+      "grad_norm": 1.471252202987671,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 2843
+    },
+    {
+      "epoch": 0.48416751787538304,
+      "grad_norm": 1.4099452495574951,
+      "learning_rate": 1e-06,
+      "loss": 0.0151,
+      "step": 2844
+    },
+    {
+      "epoch": 0.4843377596186585,
+      "grad_norm": 1.3784700632095337,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 2845
+    },
+    {
+      "epoch": 0.48450800136193395,
+      "grad_norm": 1.322296142578125,
+      "learning_rate": 1e-06,
+      "loss": 0.0154,
+      "step": 2846
+    },
+    {
+      "epoch": 0.4846782431052094,
+      "grad_norm": 1.3607927560806274,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 2847
+    },
+    {
+      "epoch": 0.48484848484848486,
+      "grad_norm": 0.7826800346374512,
+      "learning_rate": 1e-06,
+      "loss": 0.0089,
+      "step": 2848
+    },
+    {
+      "epoch": 0.4850187265917603,
+      "grad_norm": 1.2559735774993896,
+      "learning_rate": 1e-06,
+      "loss": 0.0118,
+      "step": 2849
+    },
+    {
+      "epoch": 0.4851889683350358,
+      "grad_norm": 1.423887848854065,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 2850
+    },
+    {
+      "epoch": 0.4853592100783112,
+      "grad_norm": 0.9544010758399963,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 2851
+    },
+    {
+      "epoch": 0.48552945182158663,
+      "grad_norm": 1.8954381942749023,
+      "learning_rate": 1e-06,
+      "loss": 0.0208,
+      "step": 2852
+    },
+    {
+      "epoch": 0.4856996935648621,
+      "grad_norm": 1.5164066553115845,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 2853
+    },
+    {
+      "epoch": 0.48586993530813755,
+      "grad_norm": 1.3428789377212524,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 2854
+    },
+    {
+      "epoch": 0.486040177051413,
+      "grad_norm": 1.2199963331222534,
+      "learning_rate": 1e-06,
+      "loss": 0.0168,
+      "step": 2855
+    },
+    {
+      "epoch": 0.48621041879468846,
+      "grad_norm": 1.103547215461731,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 2856
+    },
+    {
+      "epoch": 0.4863806605379639,
+      "grad_norm": 1.19355309009552,
+      "learning_rate": 1e-06,
+      "loss": 0.0137,
+      "step": 2857
+    },
+    {
+      "epoch": 0.4865509022812394,
+      "grad_norm": 1.5257407426834106,
+      "learning_rate": 1e-06,
+      "loss": 0.0125,
+      "step": 2858
+    },
+    {
+      "epoch": 0.48672114402451483,
+      "grad_norm": 1.4707660675048828,
+      "learning_rate": 1e-06,
+      "loss": 0.0142,
+      "step": 2859
+    },
+    {
+      "epoch": 0.4868913857677903,
+      "grad_norm": 1.3224432468414307,
+      "learning_rate": 1e-06,
+      "loss": 0.0141,
+      "step": 2860
+    },
+    {
+      "epoch": 0.4870616275110657,
+      "grad_norm": 1.5693126916885376,
+      "learning_rate": 1e-06,
+      "loss": 0.0144,
+      "step": 2861
+    },
+    {
+      "epoch": 0.48723186925434114,
+      "grad_norm": 1.315860629081726,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 2862
+    },
+    {
+      "epoch": 0.4874021109976166,
+      "grad_norm": 1.2941914796829224,
+      "learning_rate": 1e-06,
+      "loss": 0.0125,
+      "step": 2863
+    },
+    {
+      "epoch": 0.48757235274089206,
+      "grad_norm": 1.343846321105957,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 2864
+    },
+    {
+      "epoch": 0.4877425944841675,
+      "grad_norm": 1.0334171056747437,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 2865
+    },
+    {
+      "epoch": 0.48791283622744297,
+      "grad_norm": 1.295588493347168,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 2866
+    },
+    {
+      "epoch": 0.4880830779707184,
+      "grad_norm": 1.4225810766220093,
+      "learning_rate": 1e-06,
+      "loss": 0.0131,
+      "step": 2867
+    },
+    {
+      "epoch": 0.4882533197139939,
+      "grad_norm": 1.4328070878982544,
+      "learning_rate": 1e-06,
+      "loss": 0.0186,
+      "step": 2868
+    },
+    {
+      "epoch": 0.48842356145726934,
+      "grad_norm": 1.6347346305847168,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 2869
+    },
+    {
+      "epoch": 0.4885938032005448,
+      "grad_norm": 1.3069541454315186,
+      "learning_rate": 1e-06,
+      "loss": 0.0124,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4887640449438202,
+      "grad_norm": 1.1802916526794434,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 2871
+    },
+    {
+      "epoch": 0.48893428668709565,
+      "grad_norm": 1.2617202997207642,
+      "learning_rate": 1e-06,
+      "loss": 0.0106,
+      "step": 2872
+    },
+    {
+      "epoch": 0.4891045284303711,
+      "grad_norm": 1.1735727787017822,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 2873
+    },
+    {
+      "epoch": 0.48927477017364657,
+      "grad_norm": 1.4026036262512207,
+      "learning_rate": 1e-06,
+      "loss": 0.0163,
+      "step": 2874
+    },
+    {
+      "epoch": 0.489445011916922,
+      "grad_norm": 1.503218173980713,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 2875
+    },
+    {
+      "epoch": 0.4896152536601975,
+      "grad_norm": 1.212363362312317,
+      "learning_rate": 1e-06,
+      "loss": 0.0138,
+      "step": 2876
+    },
+    {
+      "epoch": 0.48978549540347294,
+      "grad_norm": 1.4121812582015991,
+      "learning_rate": 1e-06,
+      "loss": 0.0141,
+      "step": 2877
+    },
+    {
+      "epoch": 0.4899557371467484,
+      "grad_norm": 1.3481481075286865,
+      "learning_rate": 1e-06,
+      "loss": 0.0136,
+      "step": 2878
+    },
+    {
+      "epoch": 0.49012597889002385,
+      "grad_norm": 1.1686102151870728,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 2879
+    },
+    {
+      "epoch": 0.4902962206332993,
+      "grad_norm": 1.311177372932434,
+      "learning_rate": 1e-06,
+      "loss": 0.0189,
+      "step": 2880
+    },
+    {
+      "epoch": 0.49046646237657476,
+      "grad_norm": 1.1763383150100708,
+      "learning_rate": 1e-06,
+      "loss": 0.0133,
+      "step": 2881
+    },
+    {
+      "epoch": 0.49063670411985016,
+      "grad_norm": 1.164952039718628,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 2882
+    },
+    {
+      "epoch": 0.4908069458631256,
+      "grad_norm": 1.2506550550460815,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 2883
+    },
+    {
+      "epoch": 0.4909771876064011,
+      "grad_norm": 1.4701337814331055,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 2884
+    },
+    {
+      "epoch": 0.49114742934967653,
+      "grad_norm": 1.318747639656067,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 2885
+    },
+    {
+      "epoch": 0.491317671092952,
+      "grad_norm": 1.1562262773513794,
+      "learning_rate": 1e-06,
+      "loss": 0.0121,
+      "step": 2886
+    },
+    {
+      "epoch": 0.49148791283622745,
+      "grad_norm": 1.6450200080871582,
+      "learning_rate": 1e-06,
+      "loss": 0.0184,
+      "step": 2887
+    },
+    {
+      "epoch": 0.4916581545795029,
+      "grad_norm": 1.0631816387176514,
+      "learning_rate": 1e-06,
+      "loss": 0.0121,
+      "step": 2888
+    },
+    {
+      "epoch": 0.49182839632277836,
+      "grad_norm": 0.9874283671379089,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 2889
+    },
+    {
+      "epoch": 0.4919986380660538,
+      "grad_norm": 1.4185420274734497,
+      "learning_rate": 1e-06,
+      "loss": 0.0162,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4921688798093293,
+      "grad_norm": 0.9347524642944336,
+      "learning_rate": 1e-06,
+      "loss": 0.0069,
+      "step": 2891
+    },
+    {
+      "epoch": 0.4923391215526047,
+      "grad_norm": 0.9645169973373413,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 2892
+    },
+    {
+      "epoch": 0.49250936329588013,
+      "grad_norm": 1.236021876335144,
+      "learning_rate": 1e-06,
+      "loss": 0.0136,
+      "step": 2893
+    },
+    {
+      "epoch": 0.4926796050391556,
+      "grad_norm": 1.1450927257537842,
+      "learning_rate": 1e-06,
+      "loss": 0.0149,
+      "step": 2894
+    },
+    {
+      "epoch": 0.49284984678243104,
+      "grad_norm": 1.5991291999816895,
+      "learning_rate": 1e-06,
+      "loss": 0.0155,
+      "step": 2895
+    },
+    {
+      "epoch": 0.4930200885257065,
+      "grad_norm": 1.1790274381637573,
+      "learning_rate": 1e-06,
+      "loss": 0.0134,
+      "step": 2896
+    },
+    {
+      "epoch": 0.49319033026898196,
+      "grad_norm": 1.1705747842788696,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 2897
+    },
+    {
+      "epoch": 0.4933605720122574,
+      "grad_norm": 0.9381265640258789,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 2898
+    },
+    {
+      "epoch": 0.49353081375553287,
+      "grad_norm": 1.261146903038025,
+      "learning_rate": 1e-06,
+      "loss": 0.014,
+      "step": 2899
+    },
+    {
+      "epoch": 0.4937010554988083,
+      "grad_norm": 1.0365086793899536,
+      "learning_rate": 1e-06,
+      "loss": 0.0125,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4938712972420838,
+      "grad_norm": 1.20415198802948,
+      "learning_rate": 1e-06,
+      "loss": 0.0063,
+      "step": 2901
+    },
+    {
+      "epoch": 0.4940415389853592,
+      "grad_norm": 1.1824898719787598,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 2902
+    },
+    {
+      "epoch": 0.49421178072863464,
+      "grad_norm": 1.3033336400985718,
+      "learning_rate": 1e-06,
+      "loss": 0.0153,
+      "step": 2903
+    },
+    {
+      "epoch": 0.4943820224719101,
+      "grad_norm": 1.490906834602356,
+      "learning_rate": 1e-06,
+      "loss": 0.0129,
+      "step": 2904
+    },
+    {
+      "epoch": 0.49455226421518556,
+      "grad_norm": 1.298890471458435,
+      "learning_rate": 1e-06,
+      "loss": 0.0121,
+      "step": 2905
+    },
+    {
+      "epoch": 0.494722505958461,
+      "grad_norm": 1.805892825126648,
+      "learning_rate": 1e-06,
+      "loss": 0.028,
+      "step": 2906
+    },
+    {
+      "epoch": 0.49489274770173647,
+      "grad_norm": 1.0890419483184814,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 2907
+    },
+    {
+      "epoch": 0.4950629894450119,
+      "grad_norm": 1.0342925786972046,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 2908
+    },
+    {
+      "epoch": 0.4952332311882874,
+      "grad_norm": 1.5081183910369873,
+      "learning_rate": 1e-06,
+      "loss": 0.0134,
+      "step": 2909
+    },
+    {
+      "epoch": 0.49540347293156284,
+      "grad_norm": 2.6579811573028564,
+      "learning_rate": 1e-06,
+      "loss": 0.0148,
+      "step": 2910
+    },
+    {
+      "epoch": 0.4955737146748383,
+      "grad_norm": 1.3652353286743164,
+      "learning_rate": 1e-06,
+      "loss": 0.0125,
+      "step": 2911
+    },
+    {
+      "epoch": 0.4957439564181137,
+      "grad_norm": 1.6202207803726196,
+      "learning_rate": 1e-06,
+      "loss": 0.0186,
+      "step": 2912
+    },
+    {
+      "epoch": 0.49591419816138915,
+      "grad_norm": 1.1754422187805176,
+      "learning_rate": 1e-06,
+      "loss": 0.0154,
+      "step": 2913
+    },
+    {
+      "epoch": 0.4960844399046646,
+      "grad_norm": 1.2409707307815552,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 2914
+    },
+    {
+      "epoch": 0.49625468164794007,
+      "grad_norm": 1.7688099145889282,
+      "learning_rate": 1e-06,
+      "loss": 0.0294,
+      "step": 2915
+    },
+    {
+      "epoch": 0.4964249233912155,
+      "grad_norm": 1.4038777351379395,
+      "learning_rate": 1e-06,
+      "loss": 0.0121,
+      "step": 2916
+    },
+    {
+      "epoch": 0.496595165134491,
+      "grad_norm": 0.8886030316352844,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 2917
+    },
+    {
+      "epoch": 0.49676540687776644,
+      "grad_norm": 2.1861026287078857,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 2918
+    },
+    {
+      "epoch": 0.4969356486210419,
+      "grad_norm": 1.338416576385498,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 2919
+    },
+    {
+      "epoch": 0.49710589036431735,
+      "grad_norm": 1.1682435274124146,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 2920
+    },
+    {
+      "epoch": 0.4972761321075928,
+      "grad_norm": 1.538461446762085,
+      "learning_rate": 1e-06,
+      "loss": 0.0213,
+      "step": 2921
+    },
+    {
+      "epoch": 0.4974463738508682,
+      "grad_norm": 1.5773953199386597,
+      "learning_rate": 1e-06,
+      "loss": 0.0114,
+      "step": 2922
+    },
+    {
+      "epoch": 0.49761661559414366,
+      "grad_norm": 1.3020201921463013,
+      "learning_rate": 1e-06,
+      "loss": 0.0145,
+      "step": 2923
+    },
+    {
+      "epoch": 0.4977868573374191,
+      "grad_norm": 1.1434447765350342,
+      "learning_rate": 1e-06,
+      "loss": 0.01,
+      "step": 2924
+    },
+    {
+      "epoch": 0.4979570990806946,
+      "grad_norm": 1.2094712257385254,
+      "learning_rate": 1e-06,
+      "loss": 0.011,
+      "step": 2925
+    },
+    {
+      "epoch": 0.49812734082397003,
+      "grad_norm": 1.0923223495483398,
+      "learning_rate": 1e-06,
+      "loss": 0.0066,
+      "step": 2926
+    },
+    {
+      "epoch": 0.4982975825672455,
+      "grad_norm": 1.6678919792175293,
+      "learning_rate": 1e-06,
+      "loss": 0.0123,
+      "step": 2927
+    },
+    {
+      "epoch": 0.49846782431052095,
+      "grad_norm": 1.2051362991333008,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 2928
+    },
+    {
+      "epoch": 0.4986380660537964,
+      "grad_norm": 1.1779026985168457,
+      "learning_rate": 1e-06,
+      "loss": 0.0148,
+      "step": 2929
+    },
+    {
+      "epoch": 0.49880830779707186,
+      "grad_norm": 1.1191192865371704,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 2930
+    },
+    {
+      "epoch": 0.4989785495403473,
+      "grad_norm": 1.4694077968597412,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 2931
+    },
+    {
+      "epoch": 0.4991487912836227,
+      "grad_norm": 1.1187613010406494,
+      "learning_rate": 1e-06,
+      "loss": 0.0126,
+      "step": 2932
+    },
+    {
+      "epoch": 0.4993190330268982,
+      "grad_norm": 1.3665251731872559,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 2933
+    },
+    {
+      "epoch": 0.49948927477017363,
+      "grad_norm": 2.8328680992126465,
+      "learning_rate": 1e-06,
+      "loss": 0.0335,
+      "step": 2934
+    },
+    {
+      "epoch": 0.4996595165134491,
+      "grad_norm": 1.5214513540267944,
+      "learning_rate": 1e-06,
+      "loss": 0.0125,
+      "step": 2935
+    },
+    {
+      "epoch": 0.49982975825672454,
+      "grad_norm": 1.1730064153671265,
+      "learning_rate": 1e-06,
+      "loss": 0.013,
+      "step": 2936
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2666690349578857,
+      "learning_rate": 1e-06,
+      "loss": 0.0125,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5001702417432754,
+      "grad_norm": 1.3443002700805664,
+      "learning_rate": 1e-06,
+      "loss": 0.0124,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5003404834865509,
+      "grad_norm": 1.225641131401062,
+      "learning_rate": 1e-06,
+      "loss": 0.0089,
+      "step": 2939
+    },
+    {
+      "epoch": 0.5005107252298263,
+      "grad_norm": 1.4913207292556763,
+      "learning_rate": 1e-06,
+      "loss": 0.0155,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5006809669731018,
+      "grad_norm": 1.65010666847229,
+      "learning_rate": 1e-06,
+      "loss": 0.0235,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5008512087163772,
+      "grad_norm": 0.9611451625823975,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5010214504596527,
+      "grad_norm": 1.4015928506851196,
+      "learning_rate": 1e-06,
+      "loss": 0.0099,
+      "step": 2943
+    },
+    {
+      "epoch": 0.5011916922029281,
+      "grad_norm": 0.9811087250709534,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 2944
+    },
+    {
+      "epoch": 0.5013619339462037,
+      "grad_norm": 1.1719774007797241,
+      "learning_rate": 1e-06,
+      "loss": 0.0122,
+      "step": 2945
+    },
+    {
+      "epoch": 0.501532175689479,
+      "grad_norm": 1.2584707736968994,
+      "learning_rate": 1e-06,
+      "loss": 0.011,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5017024174327546,
+      "grad_norm": 2.9327192306518555,
+      "learning_rate": 1e-06,
+      "loss": 0.0499,
+      "step": 2947
+    },
+    {
+      "epoch": 0.50187265917603,
+      "grad_norm": 1.3452236652374268,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 2948
+    },
+    {
+      "epoch": 0.5020429009193054,
+      "grad_norm": 1.2244794368743896,
+      "learning_rate": 1e-06,
+      "loss": 0.0104,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5022131426625809,
+      "grad_norm": 1.3074755668640137,
+      "learning_rate": 1e-06,
+      "loss": 0.0099,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5023833844058563,
+      "grad_norm": 1.3225749731063843,
+      "learning_rate": 1e-06,
+      "loss": 0.0242,
+      "step": 2951
+    },
+    {
+      "epoch": 0.5025536261491318,
+      "grad_norm": 1.5591984987258911,
+      "learning_rate": 1e-06,
+      "loss": 0.0122,
+      "step": 2952
+    },
+    {
+      "epoch": 0.5027238678924072,
+      "grad_norm": 1.4079065322875977,
+      "learning_rate": 1e-06,
+      "loss": 0.01,
+      "step": 2953
+    },
+    {
+      "epoch": 0.5028941096356827,
+      "grad_norm": 1.699751377105713,
+      "learning_rate": 1e-06,
+      "loss": 0.0138,
+      "step": 2954
+    },
+    {
+      "epoch": 0.5030643513789581,
+      "grad_norm": 1.5341709852218628,
+      "learning_rate": 1e-06,
+      "loss": 0.0247,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5032345931222336,
+      "grad_norm": 1.3966187238693237,
+      "learning_rate": 1e-06,
+      "loss": 0.0099,
+      "step": 2956
+    },
+    {
+      "epoch": 0.503404834865509,
+      "grad_norm": 1.1383638381958008,
+      "learning_rate": 1e-06,
+      "loss": 0.0119,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5035750766087844,
+      "grad_norm": 1.2133922576904297,
+      "learning_rate": 1e-06,
+      "loss": 0.0134,
+      "step": 2958
+    },
+    {
+      "epoch": 0.5037453183520599,
+      "grad_norm": 1.335353970527649,
+      "learning_rate": 1e-06,
+      "loss": 0.013,
+      "step": 2959
+    },
+    {
+      "epoch": 0.5039155600953353,
+      "grad_norm": 1.509978175163269,
+      "learning_rate": 1e-06,
+      "loss": 0.0143,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5040858018386108,
+      "grad_norm": 1.110413670539856,
+      "learning_rate": 1e-06,
+      "loss": 0.0104,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5042560435818862,
+      "grad_norm": 1.325568437576294,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 2962
+    },
+    {
+      "epoch": 0.5044262853251618,
+      "grad_norm": 1.420128345489502,
+      "learning_rate": 1e-06,
+      "loss": 0.015,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5045965270684372,
+      "grad_norm": 0.9586927890777588,
+      "learning_rate": 1e-06,
+      "loss": 0.0074,
+      "step": 2964
+    },
+    {
+      "epoch": 0.5047667688117127,
+      "grad_norm": 1.4876079559326172,
+      "learning_rate": 1e-06,
+      "loss": 0.0169,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5049370105549881,
+      "grad_norm": 1.4042977094650269,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5051072522982636,
+      "grad_norm": 1.723920226097107,
+      "learning_rate": 1e-06,
+      "loss": 0.0132,
+      "step": 2967
+    },
+    {
+      "epoch": 0.505277494041539,
+      "grad_norm": 1.1755249500274658,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5054477357848144,
+      "grad_norm": 1.0757211446762085,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5056179775280899,
+      "grad_norm": 1.336470603942871,
+      "learning_rate": 1e-06,
+      "loss": 0.0131,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5057882192713653,
+      "grad_norm": 1.0691769123077393,
+      "learning_rate": 1e-06,
+      "loss": 0.0112,
+      "step": 2971
+    },
+    {
+      "epoch": 0.5059584610146408,
+      "grad_norm": 1.1425819396972656,
+      "learning_rate": 1e-06,
+      "loss": 0.0089,
+      "step": 2972
+    },
+    {
+      "epoch": 0.5061287027579162,
+      "grad_norm": 1.1680139303207397,
+      "learning_rate": 1e-06,
+      "loss": 0.0128,
+      "step": 2973
+    },
+    {
+      "epoch": 0.5062989445011917,
+      "grad_norm": 0.9332813024520874,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 2974
+    },
+    {
+      "epoch": 0.5064691862444671,
+      "grad_norm": 0.983639121055603,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5066394279877426,
+      "grad_norm": 1.32431161403656,
+      "learning_rate": 1e-06,
+      "loss": 0.0165,
+      "step": 2976
+    },
+    {
+      "epoch": 0.506809669731018,
+      "grad_norm": 1.2261784076690674,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5069799114742934,
+      "grad_norm": 3.250417470932007,
+      "learning_rate": 1e-06,
+      "loss": 0.0345,
+      "step": 2978
+    },
+    {
+      "epoch": 0.507150153217569,
+      "grad_norm": 1.4886819124221802,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5073203949608444,
+      "grad_norm": 1.391330361366272,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5074906367041199,
+      "grad_norm": 1.0295233726501465,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5076608784473953,
+      "grad_norm": 1.6181305646896362,
+      "learning_rate": 1e-06,
+      "loss": 0.0126,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5078311201906708,
+      "grad_norm": 1.2847509384155273,
+      "learning_rate": 1e-06,
+      "loss": 0.0105,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5080013619339462,
+      "grad_norm": 1.1932363510131836,
+      "learning_rate": 1e-06,
+      "loss": 0.0119,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5081716036772217,
+      "grad_norm": 1.5506092309951782,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5083418454204971,
+      "grad_norm": 1.11466646194458,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5085120871637726,
+      "grad_norm": 1.4007675647735596,
+      "learning_rate": 1e-06,
+      "loss": 0.0114,
+      "step": 2987
+    },
+    {
+      "epoch": 0.508682328907048,
+      "grad_norm": 1.0601606369018555,
+      "learning_rate": 1e-06,
+      "loss": 0.0084,
+      "step": 2988
+    },
+    {
+      "epoch": 0.5088525706503234,
+      "grad_norm": 1.1195155382156372,
+      "learning_rate": 1e-06,
+      "loss": 0.0104,
+      "step": 2989
+    },
+    {
+      "epoch": 0.5090228123935989,
+      "grad_norm": 1.2109330892562866,
+      "learning_rate": 1e-06,
+      "loss": 0.0176,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5091930541368743,
+      "grad_norm": 1.20180082321167,
+      "learning_rate": 1e-06,
+      "loss": 0.0119,
+      "step": 2991
+    },
+    {
+      "epoch": 0.5093632958801498,
+      "grad_norm": 1.3377587795257568,
+      "learning_rate": 1e-06,
+      "loss": 0.0121,
+      "step": 2992
+    },
+    {
+      "epoch": 0.5095335376234252,
+      "grad_norm": 1.0854805707931519,
+      "learning_rate": 1e-06,
+      "loss": 0.0081,
+      "step": 2993
+    },
+    {
+      "epoch": 0.5097037793667007,
+      "grad_norm": 1.8077919483184814,
+      "learning_rate": 1e-06,
+      "loss": 0.0248,
+      "step": 2994
+    },
+    {
+      "epoch": 0.5098740211099761,
+      "grad_norm": 1.4350749254226685,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5100442628532517,
+      "grad_norm": 1.6307591199874878,
+      "learning_rate": 1e-06,
+      "loss": 0.0142,
+      "step": 2996
+    },
+    {
+      "epoch": 0.5102145045965271,
+      "grad_norm": 1.2001097202301025,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 2997
+    },
+    {
+      "epoch": 0.5103847463398025,
+      "grad_norm": 0.9367316365242004,
+      "learning_rate": 1e-06,
+      "loss": 0.0073,
+      "step": 2998
+    },
+    {
+      "epoch": 0.510554988083078,
+      "grad_norm": 1.4046683311462402,
+      "learning_rate": 1e-06,
+      "loss": 0.0142,
+      "step": 2999
+    },
+    {
+      "epoch": 0.5107252298263534,
+      "grad_norm": 1.2139997482299805,
+      "learning_rate": 1e-06,
+      "loss": 0.0213,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5107252298263534,
+      "eval_loss": 0.31468451023101807,
+      "eval_runtime": 21.1588,
+      "eval_samples_per_second": 14.178,
+      "eval_steps_per_second": 0.378,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5108954715696289,
+      "grad_norm": 1.300648808479309,
+      "learning_rate": 1e-06,
+      "loss": 0.0112,
+      "step": 3001
+    },
+    {
+      "epoch": 0.5110657133129043,
+      "grad_norm": 1.0096200704574585,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 3002
+    },
+    {
+      "epoch": 0.5112359550561798,
+      "grad_norm": 1.1899932622909546,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 3003
+    },
+    {
+      "epoch": 0.5114061967994552,
+      "grad_norm": 0.9526229500770569,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 3004
+    },
+    {
+      "epoch": 0.5115764385427307,
+      "grad_norm": 1.8259762525558472,
+      "learning_rate": 1e-06,
+      "loss": 0.0231,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5117466802860061,
+      "grad_norm": 1.3105390071868896,
+      "learning_rate": 1e-06,
+      "loss": 0.0119,
+      "step": 3006
+    },
+    {
+      "epoch": 0.5119169220292816,
+      "grad_norm": 1.3401422500610352,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 3007
+    },
+    {
+      "epoch": 0.512087163772557,
+      "grad_norm": 1.4094361066818237,
+      "learning_rate": 1e-06,
+      "loss": 0.0152,
+      "step": 3008
+    },
+    {
+      "epoch": 0.5122574055158324,
+      "grad_norm": 1.5171374082565308,
+      "learning_rate": 1e-06,
+      "loss": 0.0166,
+      "step": 3009
+    },
+    {
+      "epoch": 0.5124276472591079,
+      "grad_norm": 1.069426417350769,
+      "learning_rate": 1e-06,
+      "loss": 0.0162,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5125978890023833,
+      "grad_norm": 1.3579994440078735,
+      "learning_rate": 1e-06,
+      "loss": 0.0136,
+      "step": 3011
+    },
+    {
+      "epoch": 0.5127681307456589,
+      "grad_norm": 0.9840942621231079,
+      "learning_rate": 1e-06,
+      "loss": 0.0114,
+      "step": 3012
+    },
+    {
+      "epoch": 0.5129383724889343,
+      "grad_norm": 1.523397445678711,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 3013
+    },
+    {
+      "epoch": 0.5131086142322098,
+      "grad_norm": 1.2539904117584229,
+      "learning_rate": 1e-06,
+      "loss": 0.0112,
+      "step": 3014
+    },
+    {
+      "epoch": 0.5132788559754852,
+      "grad_norm": 1.0365042686462402,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5134490977187607,
+      "grad_norm": 0.8410796523094177,
+      "learning_rate": 1e-06,
+      "loss": 0.0089,
+      "step": 3016
+    },
+    {
+      "epoch": 0.5136193394620361,
+      "grad_norm": 1.3191943168640137,
+      "learning_rate": 1e-06,
+      "loss": 0.0182,
+      "step": 3017
+    },
+    {
+      "epoch": 0.5137895812053116,
+      "grad_norm": 0.8363417387008667,
+      "learning_rate": 1e-06,
+      "loss": 0.0068,
+      "step": 3018
+    },
+    {
+      "epoch": 0.513959822948587,
+      "grad_norm": 1.1895376443862915,
+      "learning_rate": 1e-06,
+      "loss": 0.0141,
+      "step": 3019
+    },
+    {
+      "epoch": 0.5141300646918624,
+      "grad_norm": 1.4793716669082642,
+      "learning_rate": 1e-06,
+      "loss": 0.0191,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5143003064351379,
+      "grad_norm": 1.583125114440918,
+      "learning_rate": 1e-06,
+      "loss": 0.0159,
+      "step": 3021
+    },
+    {
+      "epoch": 0.5144705481784133,
+      "grad_norm": 1.3003768920898438,
+      "learning_rate": 1e-06,
+      "loss": 0.0138,
+      "step": 3022
+    },
+    {
+      "epoch": 0.5146407899216888,
+      "grad_norm": 1.2707653045654297,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 3023
+    },
+    {
+      "epoch": 0.5148110316649642,
+      "grad_norm": 1.094768762588501,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 3024
+    },
+    {
+      "epoch": 0.5149812734082397,
+      "grad_norm": 0.997310221195221,
+      "learning_rate": 1e-06,
+      "loss": 0.0073,
+      "step": 3025
+    },
+    {
+      "epoch": 0.5151515151515151,
+      "grad_norm": 1.2261766195297241,
+      "learning_rate": 1e-06,
+      "loss": 0.0144,
+      "step": 3026
+    },
+    {
+      "epoch": 0.5153217568947907,
+      "grad_norm": 1.513581395149231,
+      "learning_rate": 1e-06,
+      "loss": 0.0145,
+      "step": 3027
+    },
+    {
+      "epoch": 0.515491998638066,
+      "grad_norm": 1.1396069526672363,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 3028
+    },
+    {
+      "epoch": 0.5156622403813415,
+      "grad_norm": 1.0320284366607666,
+      "learning_rate": 1e-06,
+      "loss": 0.0074,
+      "step": 3029
+    },
+    {
+      "epoch": 0.515832482124617,
+      "grad_norm": 1.4781116247177124,
+      "learning_rate": 1e-06,
+      "loss": 0.0167,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5160027238678924,
+      "grad_norm": 1.4949243068695068,
+      "learning_rate": 1e-06,
+      "loss": 0.0176,
+      "step": 3031
+    },
+    {
+      "epoch": 0.5161729656111679,
+      "grad_norm": 1.2848188877105713,
+      "learning_rate": 1e-06,
+      "loss": 0.0142,
+      "step": 3032
+    },
+    {
+      "epoch": 0.5163432073544433,
+      "grad_norm": 1.240247368812561,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 3033
+    },
+    {
+      "epoch": 0.5165134490977188,
+      "grad_norm": 0.9559020400047302,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3034
+    },
+    {
+      "epoch": 0.5166836908409942,
+      "grad_norm": 1.2412127256393433,
+      "learning_rate": 1e-06,
+      "loss": 0.012,
+      "step": 3035
+    },
+    {
+      "epoch": 0.5168539325842697,
+      "grad_norm": 1.1315244436264038,
+      "learning_rate": 1e-06,
+      "loss": 0.0145,
+      "step": 3036
+    },
+    {
+      "epoch": 0.5170241743275451,
+      "grad_norm": 1.2756315469741821,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 3037
+    },
+    {
+      "epoch": 0.5171944160708206,
+      "grad_norm": 1.620142936706543,
+      "learning_rate": 1e-06,
+      "loss": 0.0264,
+      "step": 3038
+    },
+    {
+      "epoch": 0.517364657814096,
+      "grad_norm": 1.0165133476257324,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 3039
+    },
+    {
+      "epoch": 0.5175348995573714,
+      "grad_norm": 1.0977479219436646,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5177051413006469,
+      "grad_norm": 1.153247356414795,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 3041
+    },
+    {
+      "epoch": 0.5178753830439223,
+      "grad_norm": 1.3911783695220947,
+      "learning_rate": 1e-06,
+      "loss": 0.0167,
+      "step": 3042
+    },
+    {
+      "epoch": 0.5180456247871978,
+      "grad_norm": 1.2800657749176025,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 3043
+    },
+    {
+      "epoch": 0.5182158665304732,
+      "grad_norm": 1.216147780418396,
+      "learning_rate": 1e-06,
+      "loss": 0.0126,
+      "step": 3044
+    },
+    {
+      "epoch": 0.5183861082737488,
+      "grad_norm": 1.263554573059082,
+      "learning_rate": 1e-06,
+      "loss": 0.0114,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5185563500170242,
+      "grad_norm": 1.123915672302246,
+      "learning_rate": 1e-06,
+      "loss": 0.0082,
+      "step": 3046
+    },
+    {
+      "epoch": 0.5187265917602997,
+      "grad_norm": 1.1444989442825317,
+      "learning_rate": 1e-06,
+      "loss": 0.0131,
+      "step": 3047
+    },
+    {
+      "epoch": 0.5188968335035751,
+      "grad_norm": 1.2586398124694824,
+      "learning_rate": 1e-06,
+      "loss": 0.0084,
+      "step": 3048
+    },
+    {
+      "epoch": 0.5190670752468505,
+      "grad_norm": 1.3242063522338867,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 3049
+    },
+    {
+      "epoch": 0.519237316990126,
+      "grad_norm": 1.1678466796875,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5194075587334014,
+      "grad_norm": 1.157733678817749,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 3051
+    },
+    {
+      "epoch": 0.5195778004766769,
+      "grad_norm": 1.495124340057373,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3052
+    },
+    {
+      "epoch": 0.5197480422199523,
+      "grad_norm": 1.0295337438583374,
+      "learning_rate": 1e-06,
+      "loss": 0.0097,
+      "step": 3053
+    },
+    {
+      "epoch": 0.5199182839632278,
+      "grad_norm": 0.9237085580825806,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 3054
+    },
+    {
+      "epoch": 0.5200885257065032,
+      "grad_norm": 1.2110072374343872,
+      "learning_rate": 1e-06,
+      "loss": 0.0122,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5202587674497787,
+      "grad_norm": 1.4104846715927124,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 3056
+    },
+    {
+      "epoch": 0.5204290091930541,
+      "grad_norm": 1.0618224143981934,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 3057
+    },
+    {
+      "epoch": 0.5205992509363296,
+      "grad_norm": 1.23426353931427,
+      "learning_rate": 1e-06,
+      "loss": 0.011,
+      "step": 3058
+    },
+    {
+      "epoch": 0.520769492679605,
+      "grad_norm": 1.1079343557357788,
+      "learning_rate": 1e-06,
+      "loss": 0.0089,
+      "step": 3059
+    },
+    {
+      "epoch": 0.5209397344228804,
+      "grad_norm": 1.472641110420227,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 3060
+    },
+    {
+      "epoch": 0.521109976166156,
+      "grad_norm": 1.1031280755996704,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 3061
+    },
+    {
+      "epoch": 0.5212802179094314,
+      "grad_norm": 1.4007097482681274,
+      "learning_rate": 1e-06,
+      "loss": 0.0093,
+      "step": 3062
+    },
+    {
+      "epoch": 0.5214504596527069,
+      "grad_norm": 1.2048598527908325,
+      "learning_rate": 1e-06,
+      "loss": 0.0141,
+      "step": 3063
+    },
+    {
+      "epoch": 0.5216207013959823,
+      "grad_norm": 1.1933079957962036,
+      "learning_rate": 1e-06,
+      "loss": 0.012,
+      "step": 3064
+    },
+    {
+      "epoch": 0.5217909431392578,
+      "grad_norm": 1.139629602432251,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5219611848825332,
+      "grad_norm": 1.330549716949463,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 3066
+    },
+    {
+      "epoch": 0.5221314266258087,
+      "grad_norm": 0.8479977250099182,
+      "learning_rate": 1e-06,
+      "loss": 0.009,
+      "step": 3067
+    },
+    {
+      "epoch": 0.5223016683690841,
+      "grad_norm": 1.093963623046875,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3068
+    },
+    {
+      "epoch": 0.5224719101123596,
+      "grad_norm": 1.0407942533493042,
+      "learning_rate": 1e-06,
+      "loss": 0.0104,
+      "step": 3069
+    },
+    {
+      "epoch": 0.522642151855635,
+      "grad_norm": 1.4160226583480835,
+      "learning_rate": 1e-06,
+      "loss": 0.0178,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5228123935989104,
+      "grad_norm": 1.3179391622543335,
+      "learning_rate": 1e-06,
+      "loss": 0.01,
+      "step": 3071
+    },
+    {
+      "epoch": 0.5229826353421859,
+      "grad_norm": 1.8564910888671875,
+      "learning_rate": 1e-06,
+      "loss": 0.0199,
+      "step": 3072
+    },
+    {
+      "epoch": 0.5231528770854613,
+      "grad_norm": 1.2088946104049683,
+      "learning_rate": 1e-06,
+      "loss": 0.0129,
+      "step": 3073
+    },
+    {
+      "epoch": 0.5233231188287368,
+      "grad_norm": 1.1301778554916382,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 3074
+    },
+    {
+      "epoch": 0.5234933605720122,
+      "grad_norm": 1.4394501447677612,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 3075
+    },
+    {
+      "epoch": 0.5236636023152877,
+      "grad_norm": 1.0562634468078613,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 3076
+    },
+    {
+      "epoch": 0.5238338440585631,
+      "grad_norm": 1.2691398859024048,
+      "learning_rate": 1e-06,
+      "loss": 0.0072,
+      "step": 3077
+    },
+    {
+      "epoch": 0.5240040858018387,
+      "grad_norm": 1.257546305656433,
+      "learning_rate": 1e-06,
+      "loss": 0.013,
+      "step": 3078
+    },
+    {
+      "epoch": 0.5241743275451141,
+      "grad_norm": 1.5691494941711426,
+      "learning_rate": 1e-06,
+      "loss": 0.0152,
+      "step": 3079
+    },
+    {
+      "epoch": 0.5243445692883895,
+      "grad_norm": 1.1722148656845093,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3080
+    },
+    {
+      "epoch": 0.524514811031665,
+      "grad_norm": 0.8400589823722839,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3081
+    },
+    {
+      "epoch": 0.5246850527749404,
+      "grad_norm": 1.016576886177063,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 3082
+    },
+    {
+      "epoch": 0.5248552945182159,
+      "grad_norm": 1.129569411277771,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 3083
+    },
+    {
+      "epoch": 0.5250255362614913,
+      "grad_norm": 1.5929679870605469,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 3084
+    },
+    {
+      "epoch": 0.5251957780047668,
+      "grad_norm": 1.9398027658462524,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5253660197480422,
+      "grad_norm": 1.6181927919387817,
+      "learning_rate": 1e-06,
+      "loss": 0.0182,
+      "step": 3086
+    },
+    {
+      "epoch": 0.5255362614913177,
+      "grad_norm": 1.1063342094421387,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 3087
+    },
+    {
+      "epoch": 0.5257065032345931,
+      "grad_norm": 1.2465970516204834,
+      "learning_rate": 1e-06,
+      "loss": 0.0136,
+      "step": 3088
+    },
+    {
+      "epoch": 0.5258767449778686,
+      "grad_norm": 0.9947730302810669,
+      "learning_rate": 1e-06,
+      "loss": 0.0081,
+      "step": 3089
+    },
+    {
+      "epoch": 0.526046986721144,
+      "grad_norm": 1.0790003538131714,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5262172284644194,
+      "grad_norm": 1.3154680728912354,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3091
+    },
+    {
+      "epoch": 0.5263874702076949,
+      "grad_norm": 1.074992060661316,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 3092
+    },
+    {
+      "epoch": 0.5265577119509703,
+      "grad_norm": 1.3531382083892822,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 3093
+    },
+    {
+      "epoch": 0.5267279536942459,
+      "grad_norm": 1.8065028190612793,
+      "learning_rate": 1e-06,
+      "loss": 0.0149,
+      "step": 3094
+    },
+    {
+      "epoch": 0.5268981954375213,
+      "grad_norm": 1.652100920677185,
+      "learning_rate": 1e-06,
+      "loss": 0.015,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5270684371807968,
+      "grad_norm": 1.073090672492981,
+      "learning_rate": 1e-06,
+      "loss": 0.0123,
+      "step": 3096
+    },
+    {
+      "epoch": 0.5272386789240722,
+      "grad_norm": 1.0769790410995483,
+      "learning_rate": 1e-06,
+      "loss": 0.01,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5274089206673477,
+      "grad_norm": 1.2385014295578003,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3098
+    },
+    {
+      "epoch": 0.5275791624106231,
+      "grad_norm": 1.9769033193588257,
+      "learning_rate": 1e-06,
+      "loss": 0.0126,
+      "step": 3099
+    },
+    {
+      "epoch": 0.5277494041538985,
+      "grad_norm": 1.2743234634399414,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 3100
+    },
+    {
+      "epoch": 0.527919645897174,
+      "grad_norm": 0.9275697469711304,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5280898876404494,
+      "grad_norm": 1.1067229509353638,
+      "learning_rate": 1e-06,
+      "loss": 0.0112,
+      "step": 3102
+    },
+    {
+      "epoch": 0.5282601293837249,
+      "grad_norm": 1.1381607055664062,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 3103
+    },
+    {
+      "epoch": 0.5284303711270003,
+      "grad_norm": 1.1659680604934692,
+      "learning_rate": 1e-06,
+      "loss": 0.0104,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5286006128702758,
+      "grad_norm": 1.089293122291565,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5287708546135512,
+      "grad_norm": 1.252436876296997,
+      "learning_rate": 1e-06,
+      "loss": 0.0097,
+      "step": 3106
+    },
+    {
+      "epoch": 0.5289410963568267,
+      "grad_norm": 1.4552236795425415,
+      "learning_rate": 1e-06,
+      "loss": 0.0137,
+      "step": 3107
+    },
+    {
+      "epoch": 0.5291113381001021,
+      "grad_norm": 1.2212755680084229,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5292815798433776,
+      "grad_norm": 1.3947482109069824,
+      "learning_rate": 1e-06,
+      "loss": 0.0149,
+      "step": 3109
+    },
+    {
+      "epoch": 0.529451821586653,
+      "grad_norm": 1.2081656455993652,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5296220633299284,
+      "grad_norm": 1.2081656455993652,
+      "learning_rate": 1e-06,
+      "loss": 0.0395,
+      "step": 3111
+    },
+    {
+      "epoch": 0.529792305073204,
+      "grad_norm": 1.6731754541397095,
+      "learning_rate": 1e-06,
+      "loss": 0.0147,
+      "step": 3112
+    },
+    {
+      "epoch": 0.5299625468164794,
+      "grad_norm": 1.3174570798873901,
+      "learning_rate": 1e-06,
+      "loss": 0.0143,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5301327885597549,
+      "grad_norm": 1.2845019102096558,
+      "learning_rate": 1e-06,
+      "loss": 0.0094,
+      "step": 3114
+    },
+    {
+      "epoch": 0.5303030303030303,
+      "grad_norm": 1.1987799406051636,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 3115
+    },
+    {
+      "epoch": 0.5304732720463058,
+      "grad_norm": 0.9767221808433533,
+      "learning_rate": 1e-06,
+      "loss": 0.0066,
+      "step": 3116
+    },
+    {
+      "epoch": 0.5306435137895812,
+      "grad_norm": 1.1443259716033936,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 3117
+    },
+    {
+      "epoch": 0.5308137555328567,
+      "grad_norm": 1.4838154315948486,
+      "learning_rate": 1e-06,
+      "loss": 0.0178,
+      "step": 3118
+    },
+    {
+      "epoch": 0.5309839972761321,
+      "grad_norm": 1.5887151956558228,
+      "learning_rate": 1e-06,
+      "loss": 0.0122,
+      "step": 3119
+    },
+    {
+      "epoch": 0.5311542390194075,
+      "grad_norm": 1.2178723812103271,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3120
+    },
+    {
+      "epoch": 0.531324480762683,
+      "grad_norm": 1.0301523208618164,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3121
+    },
+    {
+      "epoch": 0.5314947225059584,
+      "grad_norm": 1.1088286638259888,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 3122
+    },
+    {
+      "epoch": 0.5316649642492339,
+      "grad_norm": 1.2651501893997192,
+      "learning_rate": 1e-06,
+      "loss": 0.0141,
+      "step": 3123
+    },
+    {
+      "epoch": 0.5318352059925093,
+      "grad_norm": 1.6318732500076294,
+      "learning_rate": 1e-06,
+      "loss": 0.0198,
+      "step": 3124
+    },
+    {
+      "epoch": 0.5320054477357848,
+      "grad_norm": 1.1797488927841187,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 3125
+    },
+    {
+      "epoch": 0.5321756894790602,
+      "grad_norm": 1.480616569519043,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3126
+    },
+    {
+      "epoch": 0.5323459312223358,
+      "grad_norm": 1.1608449220657349,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 3127
+    },
+    {
+      "epoch": 0.5325161729656112,
+      "grad_norm": 1.0857185125350952,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 3128
+    },
+    {
+      "epoch": 0.5326864147088867,
+      "grad_norm": 1.1040613651275635,
+      "learning_rate": 1e-06,
+      "loss": 0.0082,
+      "step": 3129
+    },
+    {
+      "epoch": 0.5328566564521621,
+      "grad_norm": 1.3238484859466553,
+      "learning_rate": 1e-06,
+      "loss": 0.01,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5330268981954375,
+      "grad_norm": 1.1844618320465088,
+      "learning_rate": 1e-06,
+      "loss": 0.0153,
+      "step": 3131
+    },
+    {
+      "epoch": 0.533197139938713,
+      "grad_norm": 1.6046870946884155,
+      "learning_rate": 1e-06,
+      "loss": 0.0159,
+      "step": 3132
+    },
+    {
+      "epoch": 0.5333673816819884,
+      "grad_norm": 1.2018859386444092,
+      "learning_rate": 1e-06,
+      "loss": 0.01,
+      "step": 3133
+    },
+    {
+      "epoch": 0.5335376234252639,
+      "grad_norm": 1.0841635465621948,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 3134
+    },
+    {
+      "epoch": 0.5337078651685393,
+      "grad_norm": 1.3309770822525024,
+      "learning_rate": 1e-06,
+      "loss": 0.0162,
+      "step": 3135
+    },
+    {
+      "epoch": 0.5338781069118148,
+      "grad_norm": 1.043618083000183,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 3136
+    },
+    {
+      "epoch": 0.5340483486550902,
+      "grad_norm": 1.2823837995529175,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 3137
+    },
+    {
+      "epoch": 0.5342185903983657,
+      "grad_norm": 1.3309781551361084,
+      "learning_rate": 1e-06,
+      "loss": 0.011,
+      "step": 3138
+    },
+    {
+      "epoch": 0.5343888321416411,
+      "grad_norm": 1.053794264793396,
+      "learning_rate": 1e-06,
+      "loss": 0.0119,
+      "step": 3139
+    },
+    {
+      "epoch": 0.5345590738849166,
+      "grad_norm": 1.136751413345337,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 3140
+    },
+    {
+      "epoch": 0.534729315628192,
+      "grad_norm": 1.0801527500152588,
+      "learning_rate": 1e-06,
+      "loss": 0.0071,
+      "step": 3141
+    },
+    {
+      "epoch": 0.5348995573714674,
+      "grad_norm": 1.2816129922866821,
+      "learning_rate": 1e-06,
+      "loss": 0.0147,
+      "step": 3142
+    },
+    {
+      "epoch": 0.535069799114743,
+      "grad_norm": 1.2644041776657104,
+      "learning_rate": 1e-06,
+      "loss": 0.0076,
+      "step": 3143
+    },
+    {
+      "epoch": 0.5352400408580184,
+      "grad_norm": 1.4873162508010864,
+      "learning_rate": 1e-06,
+      "loss": 0.0158,
+      "step": 3144
+    },
+    {
+      "epoch": 0.5354102826012939,
+      "grad_norm": 1.1576414108276367,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 3145
+    },
+    {
+      "epoch": 0.5355805243445693,
+      "grad_norm": 1.1078191995620728,
+      "learning_rate": 1e-06,
+      "loss": 0.0112,
+      "step": 3146
+    },
+    {
+      "epoch": 0.5357507660878448,
+      "grad_norm": 1.2426562309265137,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 3147
+    },
+    {
+      "epoch": 0.5359210078311202,
+      "grad_norm": 1.977630615234375,
+      "learning_rate": 1e-06,
+      "loss": 0.0157,
+      "step": 3148
+    },
+    {
+      "epoch": 0.5360912495743957,
+      "grad_norm": 1.5125465393066406,
+      "learning_rate": 1e-06,
+      "loss": 0.0126,
+      "step": 3149
+    },
+    {
+      "epoch": 0.5362614913176711,
+      "grad_norm": 1.4426789283752441,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5364317330609465,
+      "grad_norm": 1.0922977924346924,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 3151
+    },
+    {
+      "epoch": 0.536601974804222,
+      "grad_norm": 1.118429183959961,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 3152
+    },
+    {
+      "epoch": 0.5367722165474974,
+      "grad_norm": 1.0450609922409058,
+      "learning_rate": 1e-06,
+      "loss": 0.009,
+      "step": 3153
+    },
+    {
+      "epoch": 0.5369424582907729,
+      "grad_norm": 1.1401301622390747,
+      "learning_rate": 1e-06,
+      "loss": 0.0093,
+      "step": 3154
+    },
+    {
+      "epoch": 0.5371127000340483,
+      "grad_norm": 0.9356278777122498,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 3155
+    },
+    {
+      "epoch": 0.5372829417773238,
+      "grad_norm": 0.9514819979667664,
+      "learning_rate": 1e-06,
+      "loss": 0.0097,
+      "step": 3156
+    },
+    {
+      "epoch": 0.5374531835205992,
+      "grad_norm": 1.2086161375045776,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 3157
+    },
+    {
+      "epoch": 0.5376234252638747,
+      "grad_norm": 1.0937926769256592,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 3158
+    },
+    {
+      "epoch": 0.5377936670071501,
+      "grad_norm": 1.2354090213775635,
+      "learning_rate": 1e-06,
+      "loss": 0.0121,
+      "step": 3159
+    },
+    {
+      "epoch": 0.5379639087504257,
+      "grad_norm": 1.7234115600585938,
+      "learning_rate": 1e-06,
+      "loss": 0.0144,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5381341504937011,
+      "grad_norm": 1.390219807624817,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 3161
+    },
+    {
+      "epoch": 0.5383043922369765,
+      "grad_norm": 1.2615468502044678,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3162
+    },
+    {
+      "epoch": 0.538474633980252,
+      "grad_norm": 1.4214125871658325,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 3163
+    },
+    {
+      "epoch": 0.5386448757235274,
+      "grad_norm": 1.3602176904678345,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 3164
+    },
+    {
+      "epoch": 0.5388151174668029,
+      "grad_norm": 1.4177559614181519,
+      "learning_rate": 1e-06,
+      "loss": 0.0143,
+      "step": 3165
+    },
+    {
+      "epoch": 0.5389853592100783,
+      "grad_norm": 0.868809163570404,
+      "learning_rate": 1e-06,
+      "loss": 0.0094,
+      "step": 3166
+    },
+    {
+      "epoch": 0.5391556009533538,
+      "grad_norm": 1.1446378231048584,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 3167
+    },
+    {
+      "epoch": 0.5393258426966292,
+      "grad_norm": 1.3364777565002441,
+      "learning_rate": 1e-06,
+      "loss": 0.0112,
+      "step": 3168
+    },
+    {
+      "epoch": 0.5394960844399047,
+      "grad_norm": 1.2700159549713135,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3169
+    },
+    {
+      "epoch": 0.5396663261831801,
+      "grad_norm": 1.1539145708084106,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5398365679264555,
+      "grad_norm": 1.2744473218917847,
+      "learning_rate": 1e-06,
+      "loss": 0.0146,
+      "step": 3171
+    },
+    {
+      "epoch": 0.540006809669731,
+      "grad_norm": 1.206419587135315,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 3172
+    },
+    {
+      "epoch": 0.5401770514130064,
+      "grad_norm": 1.7593352794647217,
+      "learning_rate": 1e-06,
+      "loss": 0.0106,
+      "step": 3173
+    },
+    {
+      "epoch": 0.5403472931562819,
+      "grad_norm": 1.1174089908599854,
+      "learning_rate": 1e-06,
+      "loss": 0.013,
+      "step": 3174
+    },
+    {
+      "epoch": 0.5405175348995573,
+      "grad_norm": 1.2210553884506226,
+      "learning_rate": 1e-06,
+      "loss": 0.012,
+      "step": 3175
+    },
+    {
+      "epoch": 0.5406877766428329,
+      "grad_norm": 1.1429721117019653,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3176
+    },
+    {
+      "epoch": 0.5408580183861083,
+      "grad_norm": 1.086243748664856,
+      "learning_rate": 1e-06,
+      "loss": 0.0079,
+      "step": 3177
+    },
+    {
+      "epoch": 0.5410282601293838,
+      "grad_norm": 1.1069355010986328,
+      "learning_rate": 1e-06,
+      "loss": 0.0148,
+      "step": 3178
+    },
+    {
+      "epoch": 0.5411985018726592,
+      "grad_norm": 1.1497927904129028,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 3179
+    },
+    {
+      "epoch": 0.5413687436159347,
+      "grad_norm": 1.2102113962173462,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5415389853592101,
+      "grad_norm": 2.604693651199341,
+      "learning_rate": 1e-06,
+      "loss": 0.0258,
+      "step": 3181
+    },
+    {
+      "epoch": 0.5417092271024855,
+      "grad_norm": 1.2018678188323975,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 3182
+    },
+    {
+      "epoch": 0.541879468845761,
+      "grad_norm": 1.0666275024414062,
+      "learning_rate": 1e-06,
+      "loss": 0.0081,
+      "step": 3183
+    },
+    {
+      "epoch": 0.5420497105890364,
+      "grad_norm": 1.0146986246109009,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 3184
+    },
+    {
+      "epoch": 0.5422199523323119,
+      "grad_norm": 1.2989884614944458,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 3185
+    },
+    {
+      "epoch": 0.5423901940755873,
+      "grad_norm": 1.071435570716858,
+      "learning_rate": 1e-06,
+      "loss": 0.0133,
+      "step": 3186
+    },
+    {
+      "epoch": 0.5425604358188628,
+      "grad_norm": 1.090047836303711,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 3187
+    },
+    {
+      "epoch": 0.5427306775621382,
+      "grad_norm": 1.1832104921340942,
+      "learning_rate": 1e-06,
+      "loss": 0.0131,
+      "step": 3188
+    },
+    {
+      "epoch": 0.5429009193054137,
+      "grad_norm": 1.1112650632858276,
+      "learning_rate": 1e-06,
+      "loss": 0.0112,
+      "step": 3189
+    },
+    {
+      "epoch": 0.5430711610486891,
+      "grad_norm": 1.0457887649536133,
+      "learning_rate": 1e-06,
+      "loss": 0.0158,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5432414027919646,
+      "grad_norm": 1.2125804424285889,
+      "learning_rate": 1e-06,
+      "loss": 0.0165,
+      "step": 3191
+    },
+    {
+      "epoch": 0.54341164453524,
+      "grad_norm": 1.0135105848312378,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 3192
+    },
+    {
+      "epoch": 0.5435818862785154,
+      "grad_norm": 1.0347403287887573,
+      "learning_rate": 1e-06,
+      "loss": 0.0094,
+      "step": 3193
+    },
+    {
+      "epoch": 0.543752128021791,
+      "grad_norm": 1.2791105508804321,
+      "learning_rate": 1e-06,
+      "loss": 0.0093,
+      "step": 3194
+    },
+    {
+      "epoch": 0.5439223697650664,
+      "grad_norm": 1.1392316818237305,
+      "learning_rate": 1e-06,
+      "loss": 0.0121,
+      "step": 3195
+    },
+    {
+      "epoch": 0.5440926115083419,
+      "grad_norm": 1.273457646369934,
+      "learning_rate": 1e-06,
+      "loss": 0.0074,
+      "step": 3196
+    },
+    {
+      "epoch": 0.5442628532516173,
+      "grad_norm": 1.5065120458602905,
+      "learning_rate": 1e-06,
+      "loss": 0.0144,
+      "step": 3197
+    },
+    {
+      "epoch": 0.5444330949948928,
+      "grad_norm": 1.0680369138717651,
+      "learning_rate": 1e-06,
+      "loss": 0.0082,
+      "step": 3198
+    },
+    {
+      "epoch": 0.5446033367381682,
+      "grad_norm": 1.3402657508850098,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 3199
+    },
+    {
+      "epoch": 0.5447735784814437,
+      "grad_norm": 1.0954920053482056,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5449438202247191,
+      "grad_norm": 1.4809321165084839,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 3201
+    },
+    {
+      "epoch": 0.5451140619679945,
+      "grad_norm": 1.1664936542510986,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3202
+    },
+    {
+      "epoch": 0.54528430371127,
+      "grad_norm": 1.4334688186645508,
+      "learning_rate": 1e-06,
+      "loss": 0.0119,
+      "step": 3203
+    },
+    {
+      "epoch": 0.5454545454545454,
+      "grad_norm": 1.2824522256851196,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 3204
+    },
+    {
+      "epoch": 0.5456247871978209,
+      "grad_norm": 1.2238925695419312,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3205
+    },
+    {
+      "epoch": 0.5457950289410963,
+      "grad_norm": 1.2942602634429932,
+      "learning_rate": 1e-06,
+      "loss": 0.0142,
+      "step": 3206
+    },
+    {
+      "epoch": 0.5459652706843718,
+      "grad_norm": 0.9721873998641968,
+      "learning_rate": 1e-06,
+      "loss": 0.0082,
+      "step": 3207
+    },
+    {
+      "epoch": 0.5461355124276472,
+      "grad_norm": 1.3409883975982666,
+      "learning_rate": 1e-06,
+      "loss": 0.0122,
+      "step": 3208
+    },
+    {
+      "epoch": 0.5463057541709228,
+      "grad_norm": 1.283263087272644,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 3209
+    },
+    {
+      "epoch": 0.5464759959141982,
+      "grad_norm": 0.9081169366836548,
+      "learning_rate": 1e-06,
+      "loss": 0.0071,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5466462376574737,
+      "grad_norm": 1.0881484746932983,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3211
+    },
+    {
+      "epoch": 0.5468164794007491,
+      "grad_norm": 1.2378934621810913,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3212
+    },
+    {
+      "epoch": 0.5469867211440245,
+      "grad_norm": 1.277190089225769,
+      "learning_rate": 1e-06,
+      "loss": 0.0152,
+      "step": 3213
+    },
+    {
+      "epoch": 0.5471569628873,
+      "grad_norm": 1.6451539993286133,
+      "learning_rate": 1e-06,
+      "loss": 0.017,
+      "step": 3214
+    },
+    {
+      "epoch": 0.5473272046305754,
+      "grad_norm": 1.5039424896240234,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 3215
+    },
+    {
+      "epoch": 0.5474974463738509,
+      "grad_norm": 0.7711275219917297,
+      "learning_rate": 1e-06,
+      "loss": 0.007,
+      "step": 3216
+    },
+    {
+      "epoch": 0.5476676881171263,
+      "grad_norm": 1.2757947444915771,
+      "learning_rate": 1e-06,
+      "loss": 0.0137,
+      "step": 3217
+    },
+    {
+      "epoch": 0.5478379298604018,
+      "grad_norm": 1.2242053747177124,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3218
+    },
+    {
+      "epoch": 0.5480081716036772,
+      "grad_norm": 1.3338735103607178,
+      "learning_rate": 1e-06,
+      "loss": 0.01,
+      "step": 3219
+    },
+    {
+      "epoch": 0.5481784133469527,
+      "grad_norm": 0.9244385361671448,
+      "learning_rate": 1e-06,
+      "loss": 0.0084,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5483486550902281,
+      "grad_norm": 1.0538643598556519,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3221
+    },
+    {
+      "epoch": 0.5485188968335035,
+      "grad_norm": 1.0866847038269043,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 3222
+    },
+    {
+      "epoch": 0.548689138576779,
+      "grad_norm": 0.8596549034118652,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 3223
+    },
+    {
+      "epoch": 0.5488593803200544,
+      "grad_norm": 1.1239334344863892,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3224
+    },
+    {
+      "epoch": 0.54902962206333,
+      "grad_norm": 1.1236169338226318,
+      "learning_rate": 1e-06,
+      "loss": 0.0094,
+      "step": 3225
+    },
+    {
+      "epoch": 0.5491998638066053,
+      "grad_norm": 1.6141034364700317,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 3226
+    },
+    {
+      "epoch": 0.5493701055498809,
+      "grad_norm": 1.1685870885849,
+      "learning_rate": 1e-06,
+      "loss": 0.0082,
+      "step": 3227
+    },
+    {
+      "epoch": 0.5495403472931563,
+      "grad_norm": 1.2277820110321045,
+      "learning_rate": 1e-06,
+      "loss": 0.0146,
+      "step": 3228
+    },
+    {
+      "epoch": 0.5497105890364318,
+      "grad_norm": 1.1030385494232178,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 3229
+    },
+    {
+      "epoch": 0.5498808307797072,
+      "grad_norm": 1.246597409248352,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5500510725229827,
+      "grad_norm": 1.3571773767471313,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3231
+    },
+    {
+      "epoch": 0.5502213142662581,
+      "grad_norm": 1.283843755722046,
+      "learning_rate": 1e-06,
+      "loss": 0.011,
+      "step": 3232
+    },
+    {
+      "epoch": 0.5503915560095335,
+      "grad_norm": 1.2857335805892944,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 3233
+    },
+    {
+      "epoch": 0.550561797752809,
+      "grad_norm": 1.0135763883590698,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3234
+    },
+    {
+      "epoch": 0.5507320394960844,
+      "grad_norm": 1.3282479047775269,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 3235
+    },
+    {
+      "epoch": 0.5509022812393599,
+      "grad_norm": 1.251345157623291,
+      "learning_rate": 1e-06,
+      "loss": 0.0118,
+      "step": 3236
+    },
+    {
+      "epoch": 0.5510725229826353,
+      "grad_norm": 1.0486385822296143,
+      "learning_rate": 1e-06,
+      "loss": 0.0125,
+      "step": 3237
+    },
+    {
+      "epoch": 0.5512427647259108,
+      "grad_norm": 0.894183874130249,
+      "learning_rate": 1e-06,
+      "loss": 0.007,
+      "step": 3238
+    },
+    {
+      "epoch": 0.5514130064691862,
+      "grad_norm": 1.545824646949768,
+      "learning_rate": 1e-06,
+      "loss": 0.0141,
+      "step": 3239
+    },
+    {
+      "epoch": 0.5515832482124617,
+      "grad_norm": 1.3999972343444824,
+      "learning_rate": 1e-06,
+      "loss": 0.0137,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5517534899557371,
+      "grad_norm": 1.1850388050079346,
+      "learning_rate": 1e-06,
+      "loss": 0.0081,
+      "step": 3241
+    },
+    {
+      "epoch": 0.5519237316990125,
+      "grad_norm": 1.2595492601394653,
+      "learning_rate": 1e-06,
+      "loss": 0.0143,
+      "step": 3242
+    },
+    {
+      "epoch": 0.552093973442288,
+      "grad_norm": 1.0089160203933716,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3243
+    },
+    {
+      "epoch": 0.5522642151855635,
+      "grad_norm": 2.293635845184326,
+      "learning_rate": 1e-06,
+      "loss": 0.0285,
+      "step": 3244
+    },
+    {
+      "epoch": 0.552434456928839,
+      "grad_norm": 1.250726342201233,
+      "learning_rate": 1e-06,
+      "loss": 0.0168,
+      "step": 3245
+    },
+    {
+      "epoch": 0.5526046986721144,
+      "grad_norm": 1.1201728582382202,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 3246
+    },
+    {
+      "epoch": 0.5527749404153899,
+      "grad_norm": 1.013512134552002,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3247
+    },
+    {
+      "epoch": 0.5529451821586653,
+      "grad_norm": 1.111093282699585,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 3248
+    },
+    {
+      "epoch": 0.5531154239019408,
+      "grad_norm": 0.9368535876274109,
+      "learning_rate": 1e-06,
+      "loss": 0.0149,
+      "step": 3249
+    },
+    {
+      "epoch": 0.5532856656452162,
+      "grad_norm": 1.0745536088943481,
+      "learning_rate": 1e-06,
+      "loss": 0.0134,
+      "step": 3250
+    },
+    {
+      "epoch": 0.5534559073884917,
+      "grad_norm": 1.2841395139694214,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3251
+    },
+    {
+      "epoch": 0.5536261491317671,
+      "grad_norm": 1.3827086687088013,
+      "learning_rate": 1e-06,
+      "loss": 0.0094,
+      "step": 3252
+    },
+    {
+      "epoch": 0.5537963908750425,
+      "grad_norm": 1.475622534751892,
+      "learning_rate": 1e-06,
+      "loss": 0.0131,
+      "step": 3253
+    },
+    {
+      "epoch": 0.553966632618318,
+      "grad_norm": 0.9399949312210083,
+      "learning_rate": 1e-06,
+      "loss": 0.0104,
+      "step": 3254
+    },
+    {
+      "epoch": 0.5541368743615934,
+      "grad_norm": 1.1346850395202637,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 3255
+    },
+    {
+      "epoch": 0.5543071161048689,
+      "grad_norm": 1.2634351253509521,
+      "learning_rate": 1e-06,
+      "loss": 0.0105,
+      "step": 3256
+    },
+    {
+      "epoch": 0.5544773578481443,
+      "grad_norm": 0.9699029326438904,
+      "learning_rate": 1e-06,
+      "loss": 0.0081,
+      "step": 3257
+    },
+    {
+      "epoch": 0.5546475995914198,
+      "grad_norm": 1.6691503524780273,
+      "learning_rate": 1e-06,
+      "loss": 0.0188,
+      "step": 3258
+    },
+    {
+      "epoch": 0.5548178413346952,
+      "grad_norm": 1.39431893825531,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 3259
+    },
+    {
+      "epoch": 0.5549880830779708,
+      "grad_norm": 1.5263134241104126,
+      "learning_rate": 1e-06,
+      "loss": 0.0186,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5551583248212462,
+      "grad_norm": 1.1584216356277466,
+      "learning_rate": 1e-06,
+      "loss": 0.0164,
+      "step": 3261
+    },
+    {
+      "epoch": 0.5553285665645217,
+      "grad_norm": 1.0202586650848389,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 3262
+    },
+    {
+      "epoch": 0.5554988083077971,
+      "grad_norm": 1.393701434135437,
+      "learning_rate": 1e-06,
+      "loss": 0.0097,
+      "step": 3263
+    },
+    {
+      "epoch": 0.5556690500510725,
+      "grad_norm": 1.0818877220153809,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3264
+    },
+    {
+      "epoch": 0.555839291794348,
+      "grad_norm": 1.3524203300476074,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 3265
+    },
+    {
+      "epoch": 0.5560095335376234,
+      "grad_norm": 1.1028423309326172,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3266
+    },
+    {
+      "epoch": 0.5561797752808989,
+      "grad_norm": 1.4970299005508423,
+      "learning_rate": 1e-06,
+      "loss": 0.0173,
+      "step": 3267
+    },
+    {
+      "epoch": 0.5563500170241743,
+      "grad_norm": 1.0553059577941895,
+      "learning_rate": 1e-06,
+      "loss": 0.0178,
+      "step": 3268
+    },
+    {
+      "epoch": 0.5565202587674498,
+      "grad_norm": 1.0315494537353516,
+      "learning_rate": 1e-06,
+      "loss": 0.0082,
+      "step": 3269
+    },
+    {
+      "epoch": 0.5566905005107252,
+      "grad_norm": 1.201059341430664,
+      "learning_rate": 1e-06,
+      "loss": 0.0134,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5568607422540007,
+      "grad_norm": 1.472936987876892,
+      "learning_rate": 1e-06,
+      "loss": 0.0179,
+      "step": 3271
+    },
+    {
+      "epoch": 0.5570309839972761,
+      "grad_norm": 1.2015950679779053,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3272
+    },
+    {
+      "epoch": 0.5572012257405515,
+      "grad_norm": 1.715442180633545,
+      "learning_rate": 1e-06,
+      "loss": 0.0126,
+      "step": 3273
+    },
+    {
+      "epoch": 0.557371467483827,
+      "grad_norm": 0.8229348659515381,
+      "learning_rate": 1e-06,
+      "loss": 0.0067,
+      "step": 3274
+    },
+    {
+      "epoch": 0.5575417092271024,
+      "grad_norm": 1.4160443544387817,
+      "learning_rate": 1e-06,
+      "loss": 0.0123,
+      "step": 3275
+    },
+    {
+      "epoch": 0.557711950970378,
+      "grad_norm": 1.3626257181167603,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 3276
+    },
+    {
+      "epoch": 0.5578821927136534,
+      "grad_norm": 0.9082250595092773,
+      "learning_rate": 1e-06,
+      "loss": 0.0081,
+      "step": 3277
+    },
+    {
+      "epoch": 0.5580524344569289,
+      "grad_norm": 1.0986019372940063,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 3278
+    },
+    {
+      "epoch": 0.5582226762002043,
+      "grad_norm": 1.828899621963501,
+      "learning_rate": 1e-06,
+      "loss": 0.0307,
+      "step": 3279
+    },
+    {
+      "epoch": 0.5583929179434798,
+      "grad_norm": 0.9443734288215637,
+      "learning_rate": 1e-06,
+      "loss": 0.0076,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5585631596867552,
+      "grad_norm": 0.9179772138595581,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 3281
+    },
+    {
+      "epoch": 0.5587334014300307,
+      "grad_norm": 1.0771028995513916,
+      "learning_rate": 1e-06,
+      "loss": 0.0071,
+      "step": 3282
+    },
+    {
+      "epoch": 0.5589036431733061,
+      "grad_norm": 1.1734918355941772,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 3283
+    },
+    {
+      "epoch": 0.5590738849165815,
+      "grad_norm": 1.24713134765625,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 3284
+    },
+    {
+      "epoch": 0.559244126659857,
+      "grad_norm": 1.0508451461791992,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 3285
+    },
+    {
+      "epoch": 0.5594143684031324,
+      "grad_norm": 1.0896153450012207,
+      "learning_rate": 1e-06,
+      "loss": 0.011,
+      "step": 3286
+    },
+    {
+      "epoch": 0.5595846101464079,
+      "grad_norm": 1.1460729837417603,
+      "learning_rate": 1e-06,
+      "loss": 0.0112,
+      "step": 3287
+    },
+    {
+      "epoch": 0.5597548518896833,
+      "grad_norm": 1.5633426904678345,
+      "learning_rate": 1e-06,
+      "loss": 0.0148,
+      "step": 3288
+    },
+    {
+      "epoch": 0.5599250936329588,
+      "grad_norm": 1.5234661102294922,
+      "learning_rate": 1e-06,
+      "loss": 0.0132,
+      "step": 3289
+    },
+    {
+      "epoch": 0.5600953353762342,
+      "grad_norm": 0.8329904079437256,
+      "learning_rate": 1e-06,
+      "loss": 0.0059,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5602655771195098,
+      "grad_norm": 1.244648814201355,
+      "learning_rate": 1e-06,
+      "loss": 0.0106,
+      "step": 3291
+    },
+    {
+      "epoch": 0.5604358188627852,
+      "grad_norm": 1.2756433486938477,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 3292
+    },
+    {
+      "epoch": 0.5606060606060606,
+      "grad_norm": 0.9432902932167053,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3293
+    },
+    {
+      "epoch": 0.5607763023493361,
+      "grad_norm": 1.3495780229568481,
+      "learning_rate": 1e-06,
+      "loss": 0.0122,
+      "step": 3294
+    },
+    {
+      "epoch": 0.5609465440926115,
+      "grad_norm": 1.3346593379974365,
+      "learning_rate": 1e-06,
+      "loss": 0.0121,
+      "step": 3295
+    },
+    {
+      "epoch": 0.561116785835887,
+      "grad_norm": 1.12068510055542,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 3296
+    },
+    {
+      "epoch": 0.5612870275791624,
+      "grad_norm": 1.0519905090332031,
+      "learning_rate": 1e-06,
+      "loss": 0.009,
+      "step": 3297
+    },
+    {
+      "epoch": 0.5614572693224379,
+      "grad_norm": 1.1385189294815063,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 3298
+    },
+    {
+      "epoch": 0.5616275110657133,
+      "grad_norm": 1.3988077640533447,
+      "learning_rate": 1e-06,
+      "loss": 0.0183,
+      "step": 3299
+    },
+    {
+      "epoch": 0.5617977528089888,
+      "grad_norm": 1.1090185642242432,
+      "learning_rate": 1e-06,
+      "loss": 0.0072,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5619679945522642,
+      "grad_norm": 1.2562520503997803,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 3301
+    },
+    {
+      "epoch": 0.5621382362955397,
+      "grad_norm": 1.1855425834655762,
+      "learning_rate": 1e-06,
+      "loss": 0.0132,
+      "step": 3302
+    },
+    {
+      "epoch": 0.5623084780388151,
+      "grad_norm": 1.492267370223999,
+      "learning_rate": 1e-06,
+      "loss": 0.0105,
+      "step": 3303
+    },
+    {
+      "epoch": 0.5624787197820905,
+      "grad_norm": 1.1446338891983032,
+      "learning_rate": 1e-06,
+      "loss": 0.0112,
+      "step": 3304
+    },
+    {
+      "epoch": 0.562648961525366,
+      "grad_norm": 0.9587407112121582,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3305
+    },
+    {
+      "epoch": 0.5628192032686414,
+      "grad_norm": 1.0618356466293335,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3306
+    },
+    {
+      "epoch": 0.562989445011917,
+      "grad_norm": 1.6512682437896729,
+      "learning_rate": 1e-06,
+      "loss": 0.0229,
+      "step": 3307
+    },
+    {
+      "epoch": 0.5631596867551923,
+      "grad_norm": 1.2739150524139404,
+      "learning_rate": 1e-06,
+      "loss": 0.0156,
+      "step": 3308
+    },
+    {
+      "epoch": 0.5633299284984679,
+      "grad_norm": 1.4853968620300293,
+      "learning_rate": 1e-06,
+      "loss": 0.0197,
+      "step": 3309
+    },
+    {
+      "epoch": 0.5635001702417433,
+      "grad_norm": 1.269023060798645,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5636704119850188,
+      "grad_norm": 1.162980556488037,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 3311
+    },
+    {
+      "epoch": 0.5638406537282942,
+      "grad_norm": 2.1068434715270996,
+      "learning_rate": 1e-06,
+      "loss": 0.0151,
+      "step": 3312
+    },
+    {
+      "epoch": 0.5640108954715697,
+      "grad_norm": 1.2134860754013062,
+      "learning_rate": 1e-06,
+      "loss": 0.0142,
+      "step": 3313
+    },
+    {
+      "epoch": 0.5641811372148451,
+      "grad_norm": 1.2850688695907593,
+      "learning_rate": 1e-06,
+      "loss": 0.015,
+      "step": 3314
+    },
+    {
+      "epoch": 0.5643513789581205,
+      "grad_norm": 1.0942927598953247,
+      "learning_rate": 1e-06,
+      "loss": 0.0068,
+      "step": 3315
+    },
+    {
+      "epoch": 0.564521620701396,
+      "grad_norm": 1.20088791847229,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 3316
+    },
+    {
+      "epoch": 0.5646918624446714,
+      "grad_norm": 1.0646437406539917,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3317
+    },
+    {
+      "epoch": 0.5648621041879469,
+      "grad_norm": 0.8980295062065125,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3318
+    },
+    {
+      "epoch": 0.5650323459312223,
+      "grad_norm": 1.4501545429229736,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3319
+    },
+    {
+      "epoch": 0.5652025876744978,
+      "grad_norm": 1.4403294324874878,
+      "learning_rate": 1e-06,
+      "loss": 0.0204,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5653728294177732,
+      "grad_norm": 1.6485029458999634,
+      "learning_rate": 1e-06,
+      "loss": 0.0119,
+      "step": 3321
+    },
+    {
+      "epoch": 0.5655430711610487,
+      "grad_norm": 1.0820796489715576,
+      "learning_rate": 1e-06,
+      "loss": 0.0068,
+      "step": 3322
+    },
+    {
+      "epoch": 0.5657133129043241,
+      "grad_norm": 1.0240590572357178,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 3323
+    },
+    {
+      "epoch": 0.5658835546475995,
+      "grad_norm": 1.1677789688110352,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3324
+    },
+    {
+      "epoch": 0.566053796390875,
+      "grad_norm": 1.3193532228469849,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 3325
+    },
+    {
+      "epoch": 0.5662240381341505,
+      "grad_norm": 0.9879570007324219,
+      "learning_rate": 1e-06,
+      "loss": 0.009,
+      "step": 3326
+    },
+    {
+      "epoch": 0.566394279877426,
+      "grad_norm": 1.6975972652435303,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 3327
+    },
+    {
+      "epoch": 0.5665645216207014,
+      "grad_norm": 1.4392386674880981,
+      "learning_rate": 1e-06,
+      "loss": 0.0126,
+      "step": 3328
+    },
+    {
+      "epoch": 0.5667347633639769,
+      "grad_norm": 1.341951847076416,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 3329
+    },
+    {
+      "epoch": 0.5669050051072523,
+      "grad_norm": 1.0540547370910645,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5670752468505278,
+      "grad_norm": 1.0311988592147827,
+      "learning_rate": 1e-06,
+      "loss": 0.0147,
+      "step": 3331
+    },
+    {
+      "epoch": 0.5672454885938032,
+      "grad_norm": 1.1205745935440063,
+      "learning_rate": 1e-06,
+      "loss": 0.0094,
+      "step": 3332
+    },
+    {
+      "epoch": 0.5674157303370787,
+      "grad_norm": 0.9916560649871826,
+      "learning_rate": 1e-06,
+      "loss": 0.01,
+      "step": 3333
+    },
+    {
+      "epoch": 0.5675859720803541,
+      "grad_norm": 1.0641469955444336,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 3334
+    },
+    {
+      "epoch": 0.5677562138236295,
+      "grad_norm": 1.4012588262557983,
+      "learning_rate": 1e-06,
+      "loss": 0.0185,
+      "step": 3335
+    },
+    {
+      "epoch": 0.567926455566905,
+      "grad_norm": 1.0425010919570923,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 3336
+    },
+    {
+      "epoch": 0.5680966973101804,
+      "grad_norm": 1.2166441679000854,
+      "learning_rate": 1e-06,
+      "loss": 0.0123,
+      "step": 3337
+    },
+    {
+      "epoch": 0.5682669390534559,
+      "grad_norm": 1.4279252290725708,
+      "learning_rate": 1e-06,
+      "loss": 0.0118,
+      "step": 3338
+    },
+    {
+      "epoch": 0.5684371807967313,
+      "grad_norm": 1.7417263984680176,
+      "learning_rate": 1e-06,
+      "loss": 0.0167,
+      "step": 3339
+    },
+    {
+      "epoch": 0.5686074225400068,
+      "grad_norm": 1.190510869026184,
+      "learning_rate": 1e-06,
+      "loss": 0.0112,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5687776642832822,
+      "grad_norm": 1.2967720031738281,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 3341
+    },
+    {
+      "epoch": 0.5689479060265578,
+      "grad_norm": 1.2348685264587402,
+      "learning_rate": 1e-06,
+      "loss": 0.0084,
+      "step": 3342
+    },
+    {
+      "epoch": 0.5691181477698332,
+      "grad_norm": 1.24281907081604,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 3343
+    },
+    {
+      "epoch": 0.5692883895131086,
+      "grad_norm": 1.3598650693893433,
+      "learning_rate": 1e-06,
+      "loss": 0.0173,
+      "step": 3344
+    },
+    {
+      "epoch": 0.5694586312563841,
+      "grad_norm": 1.173242211341858,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 3345
+    },
+    {
+      "epoch": 0.5696288729996595,
+      "grad_norm": 1.2969194650650024,
+      "learning_rate": 1e-06,
+      "loss": 0.0114,
+      "step": 3346
+    },
+    {
+      "epoch": 0.569799114742935,
+      "grad_norm": 0.953393280506134,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3347
+    },
+    {
+      "epoch": 0.5699693564862104,
+      "grad_norm": 1.2837368249893188,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3348
+    },
+    {
+      "epoch": 0.5701395982294859,
+      "grad_norm": 0.9645003080368042,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 3349
+    },
+    {
+      "epoch": 0.5703098399727613,
+      "grad_norm": 0.9485358595848083,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5704800817160368,
+      "grad_norm": 1.000497817993164,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 3351
+    },
+    {
+      "epoch": 0.5706503234593122,
+      "grad_norm": 1.262580394744873,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3352
+    },
+    {
+      "epoch": 0.5708205652025877,
+      "grad_norm": 1.2770668268203735,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 3353
+    },
+    {
+      "epoch": 0.5709908069458631,
+      "grad_norm": 0.9978349804878235,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3354
+    },
+    {
+      "epoch": 0.5711610486891385,
+      "grad_norm": 1.1348481178283691,
+      "learning_rate": 1e-06,
+      "loss": 0.0114,
+      "step": 3355
+    },
+    {
+      "epoch": 0.571331290432414,
+      "grad_norm": 1.1138725280761719,
+      "learning_rate": 1e-06,
+      "loss": 0.0077,
+      "step": 3356
+    },
+    {
+      "epoch": 0.5715015321756894,
+      "grad_norm": 1.243876576423645,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 3357
+    },
+    {
+      "epoch": 0.571671773918965,
+      "grad_norm": 2.1010210514068604,
+      "learning_rate": 1e-06,
+      "loss": 0.0229,
+      "step": 3358
+    },
+    {
+      "epoch": 0.5718420156622404,
+      "grad_norm": 1.5689268112182617,
+      "learning_rate": 1e-06,
+      "loss": 0.0199,
+      "step": 3359
+    },
+    {
+      "epoch": 0.5720122574055159,
+      "grad_norm": 1.4927451610565186,
+      "learning_rate": 1e-06,
+      "loss": 0.0145,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5721824991487913,
+      "grad_norm": 1.2088425159454346,
+      "learning_rate": 1e-06,
+      "loss": 0.0084,
+      "step": 3361
+    },
+    {
+      "epoch": 0.5723527408920668,
+      "grad_norm": 3.114051580429077,
+      "learning_rate": 1e-06,
+      "loss": 0.0122,
+      "step": 3362
+    },
+    {
+      "epoch": 0.5725229826353422,
+      "grad_norm": 0.9580630660057068,
+      "learning_rate": 1e-06,
+      "loss": 0.0076,
+      "step": 3363
+    },
+    {
+      "epoch": 0.5726932243786176,
+      "grad_norm": 0.9830418229103088,
+      "learning_rate": 1e-06,
+      "loss": 0.0097,
+      "step": 3364
+    },
+    {
+      "epoch": 0.5728634661218931,
+      "grad_norm": 1.2389250993728638,
+      "learning_rate": 1e-06,
+      "loss": 0.0099,
+      "step": 3365
+    },
+    {
+      "epoch": 0.5730337078651685,
+      "grad_norm": 1.159663438796997,
+      "learning_rate": 1e-06,
+      "loss": 0.0141,
+      "step": 3366
+    },
+    {
+      "epoch": 0.573203949608444,
+      "grad_norm": 1.5765507221221924,
+      "learning_rate": 1e-06,
+      "loss": 0.0123,
+      "step": 3367
+    },
+    {
+      "epoch": 0.5733741913517194,
+      "grad_norm": 1.1403703689575195,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3368
+    },
+    {
+      "epoch": 0.5735444330949949,
+      "grad_norm": 1.376774787902832,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 3369
+    },
+    {
+      "epoch": 0.5737146748382703,
+      "grad_norm": 1.386721134185791,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5738849165815458,
+      "grad_norm": 1.6140387058258057,
+      "learning_rate": 1e-06,
+      "loss": 0.014,
+      "step": 3371
+    },
+    {
+      "epoch": 0.5740551583248212,
+      "grad_norm": 0.9297884702682495,
+      "learning_rate": 1e-06,
+      "loss": 0.0089,
+      "step": 3372
+    },
+    {
+      "epoch": 0.5742254000680967,
+      "grad_norm": 1.196844220161438,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 3373
+    },
+    {
+      "epoch": 0.5743956418113721,
+      "grad_norm": 0.999453604221344,
+      "learning_rate": 1e-06,
+      "loss": 0.0097,
+      "step": 3374
+    },
+    {
+      "epoch": 0.5745658835546475,
+      "grad_norm": 1.378943920135498,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 3375
+    },
+    {
+      "epoch": 0.5747361252979231,
+      "grad_norm": 1.1324682235717773,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3376
+    },
+    {
+      "epoch": 0.5749063670411985,
+      "grad_norm": 1.0931276082992554,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3377
+    },
+    {
+      "epoch": 0.575076608784474,
+      "grad_norm": 1.0368367433547974,
+      "learning_rate": 1e-06,
+      "loss": 0.0073,
+      "step": 3378
+    },
+    {
+      "epoch": 0.5752468505277494,
+      "grad_norm": 1.0412521362304688,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 3379
+    },
+    {
+      "epoch": 0.5754170922710249,
+      "grad_norm": 1.5202747583389282,
+      "learning_rate": 1e-06,
+      "loss": 0.0134,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5755873340143003,
+      "grad_norm": 0.9431686401367188,
+      "learning_rate": 1e-06,
+      "loss": 0.0062,
+      "step": 3381
+    },
+    {
+      "epoch": 0.5757575757575758,
+      "grad_norm": 1.391595482826233,
+      "learning_rate": 1e-06,
+      "loss": 0.0187,
+      "step": 3382
+    },
+    {
+      "epoch": 0.5759278175008512,
+      "grad_norm": 1.1645992994308472,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3383
+    },
+    {
+      "epoch": 0.5760980592441267,
+      "grad_norm": 1.1957941055297852,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 3384
+    },
+    {
+      "epoch": 0.5762683009874021,
+      "grad_norm": 0.9163592457771301,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 3385
+    },
+    {
+      "epoch": 0.5764385427306775,
+      "grad_norm": 1.4192230701446533,
+      "learning_rate": 1e-06,
+      "loss": 0.0132,
+      "step": 3386
+    },
+    {
+      "epoch": 0.576608784473953,
+      "grad_norm": 1.128556728363037,
+      "learning_rate": 1e-06,
+      "loss": 0.0094,
+      "step": 3387
+    },
+    {
+      "epoch": 0.5767790262172284,
+      "grad_norm": 1.0899877548217773,
+      "learning_rate": 1e-06,
+      "loss": 0.0099,
+      "step": 3388
+    },
+    {
+      "epoch": 0.5769492679605039,
+      "grad_norm": 1.5892912149429321,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 3389
+    },
+    {
+      "epoch": 0.5771195097037793,
+      "grad_norm": 0.9154031276702881,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5772897514470549,
+      "grad_norm": 1.380571722984314,
+      "learning_rate": 1e-06,
+      "loss": 0.0131,
+      "step": 3391
+    },
+    {
+      "epoch": 0.5774599931903303,
+      "grad_norm": 1.0964083671569824,
+      "learning_rate": 1e-06,
+      "loss": 0.0084,
+      "step": 3392
+    },
+    {
+      "epoch": 0.5776302349336058,
+      "grad_norm": 1.2165638208389282,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 3393
+    },
+    {
+      "epoch": 0.5778004766768812,
+      "grad_norm": 1.234350323677063,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 3394
+    },
+    {
+      "epoch": 0.5779707184201566,
+      "grad_norm": 1.2220001220703125,
+      "learning_rate": 1e-06,
+      "loss": 0.0143,
+      "step": 3395
+    },
+    {
+      "epoch": 0.5781409601634321,
+      "grad_norm": 1.1645594835281372,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 3396
+    },
+    {
+      "epoch": 0.5783112019067075,
+      "grad_norm": 1.3842546939849854,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 3397
+    },
+    {
+      "epoch": 0.578481443649983,
+      "grad_norm": 0.8232072591781616,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 3398
+    },
+    {
+      "epoch": 0.5786516853932584,
+      "grad_norm": 1.1387802362442017,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 3399
+    },
+    {
+      "epoch": 0.5788219271365339,
+      "grad_norm": 1.1812922954559326,
+      "learning_rate": 1e-06,
+      "loss": 0.0097,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5789921688798093,
+      "grad_norm": 1.0787347555160522,
+      "learning_rate": 1e-06,
+      "loss": 0.0167,
+      "step": 3401
+    },
+    {
+      "epoch": 0.5791624106230848,
+      "grad_norm": 1.2608331441879272,
+      "learning_rate": 1e-06,
+      "loss": 0.0122,
+      "step": 3402
+    },
+    {
+      "epoch": 0.5793326523663602,
+      "grad_norm": 1.1550672054290771,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3403
+    },
+    {
+      "epoch": 0.5795028941096357,
+      "grad_norm": 1.1368712186813354,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 3404
+    },
+    {
+      "epoch": 0.5796731358529111,
+      "grad_norm": 1.4843052625656128,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 3405
+    },
+    {
+      "epoch": 0.5798433775961865,
+      "grad_norm": 1.0383031368255615,
+      "learning_rate": 1e-06,
+      "loss": 0.0069,
+      "step": 3406
+    },
+    {
+      "epoch": 0.580013619339462,
+      "grad_norm": 1.9053559303283691,
+      "learning_rate": 1e-06,
+      "loss": 0.0148,
+      "step": 3407
+    },
+    {
+      "epoch": 0.5801838610827375,
+      "grad_norm": 1.11162531375885,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 3408
+    },
+    {
+      "epoch": 0.580354102826013,
+      "grad_norm": 1.5268166065216064,
+      "learning_rate": 1e-06,
+      "loss": 0.0105,
+      "step": 3409
+    },
+    {
+      "epoch": 0.5805243445692884,
+      "grad_norm": 0.8884620070457458,
+      "learning_rate": 1e-06,
+      "loss": 0.0074,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5806945863125639,
+      "grad_norm": 1.1013774871826172,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 3411
+    },
+    {
+      "epoch": 0.5808648280558393,
+      "grad_norm": 0.822043776512146,
+      "learning_rate": 1e-06,
+      "loss": 0.0082,
+      "step": 3412
+    },
+    {
+      "epoch": 0.5810350697991148,
+      "grad_norm": 0.9275667667388916,
+      "learning_rate": 1e-06,
+      "loss": 0.0093,
+      "step": 3413
+    },
+    {
+      "epoch": 0.5812053115423902,
+      "grad_norm": 1.002583384513855,
+      "learning_rate": 1e-06,
+      "loss": 0.0125,
+      "step": 3414
+    },
+    {
+      "epoch": 0.5813755532856656,
+      "grad_norm": 0.8230078816413879,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3415
+    },
+    {
+      "epoch": 0.5815457950289411,
+      "grad_norm": 1.2379169464111328,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3416
+    },
+    {
+      "epoch": 0.5817160367722165,
+      "grad_norm": 0.9411685466766357,
+      "learning_rate": 1e-06,
+      "loss": 0.0071,
+      "step": 3417
+    },
+    {
+      "epoch": 0.581886278515492,
+      "grad_norm": 1.2261695861816406,
+      "learning_rate": 1e-06,
+      "loss": 0.0106,
+      "step": 3418
+    },
+    {
+      "epoch": 0.5820565202587674,
+      "grad_norm": 1.1066349744796753,
+      "learning_rate": 1e-06,
+      "loss": 0.0084,
+      "step": 3419
+    },
+    {
+      "epoch": 0.5822267620020429,
+      "grad_norm": 1.058130145072937,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5823970037453183,
+      "grad_norm": 2.5851633548736572,
+      "learning_rate": 1e-06,
+      "loss": 0.0253,
+      "step": 3421
+    },
+    {
+      "epoch": 0.5825672454885938,
+      "grad_norm": 1.2730610370635986,
+      "learning_rate": 1e-06,
+      "loss": 0.0077,
+      "step": 3422
+    },
+    {
+      "epoch": 0.5827374872318692,
+      "grad_norm": 1.3514959812164307,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3423
+    },
+    {
+      "epoch": 0.5829077289751448,
+      "grad_norm": 1.4177110195159912,
+      "learning_rate": 1e-06,
+      "loss": 0.0097,
+      "step": 3424
+    },
+    {
+      "epoch": 0.5830779707184202,
+      "grad_norm": 1.2021658420562744,
+      "learning_rate": 1e-06,
+      "loss": 0.0118,
+      "step": 3425
+    },
+    {
+      "epoch": 0.5832482124616956,
+      "grad_norm": 1.4584933519363403,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 3426
+    },
+    {
+      "epoch": 0.5834184542049711,
+      "grad_norm": 1.345293402671814,
+      "learning_rate": 1e-06,
+      "loss": 0.0128,
+      "step": 3427
+    },
+    {
+      "epoch": 0.5835886959482465,
+      "grad_norm": 1.2940164804458618,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 3428
+    },
+    {
+      "epoch": 0.583758937691522,
+      "grad_norm": 1.1534115076065063,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 3429
+    },
+    {
+      "epoch": 0.5839291794347974,
+      "grad_norm": 0.9437724947929382,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5840994211780729,
+      "grad_norm": 1.2992208003997803,
+      "learning_rate": 1e-06,
+      "loss": 0.0164,
+      "step": 3431
+    },
+    {
+      "epoch": 0.5842696629213483,
+      "grad_norm": 0.95462566614151,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3432
+    },
+    {
+      "epoch": 0.5844399046646238,
+      "grad_norm": 1.0334690809249878,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3433
+    },
+    {
+      "epoch": 0.5846101464078992,
+      "grad_norm": 1.707252025604248,
+      "learning_rate": 1e-06,
+      "loss": 0.0143,
+      "step": 3434
+    },
+    {
+      "epoch": 0.5847803881511746,
+      "grad_norm": 0.9804726839065552,
+      "learning_rate": 1e-06,
+      "loss": 0.0077,
+      "step": 3435
+    },
+    {
+      "epoch": 0.5849506298944501,
+      "grad_norm": 1.5533514022827148,
+      "learning_rate": 1e-06,
+      "loss": 0.0143,
+      "step": 3436
+    },
+    {
+      "epoch": 0.5851208716377255,
+      "grad_norm": 1.154433012008667,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 3437
+    },
+    {
+      "epoch": 0.585291113381001,
+      "grad_norm": 1.2386829853057861,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 3438
+    },
+    {
+      "epoch": 0.5854613551242764,
+      "grad_norm": 0.9458797574043274,
+      "learning_rate": 1e-06,
+      "loss": 0.0082,
+      "step": 3439
+    },
+    {
+      "epoch": 0.585631596867552,
+      "grad_norm": 1.1015353202819824,
+      "learning_rate": 1e-06,
+      "loss": 0.0093,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5858018386108274,
+      "grad_norm": 1.14461350440979,
+      "learning_rate": 1e-06,
+      "loss": 0.0105,
+      "step": 3441
+    },
+    {
+      "epoch": 0.5859720803541029,
+      "grad_norm": 1.4417644739151,
+      "learning_rate": 1e-06,
+      "loss": 0.0104,
+      "step": 3442
+    },
+    {
+      "epoch": 0.5861423220973783,
+      "grad_norm": 1.3879674673080444,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 3443
+    },
+    {
+      "epoch": 0.5863125638406538,
+      "grad_norm": 1.0989384651184082,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3444
+    },
+    {
+      "epoch": 0.5864828055839292,
+      "grad_norm": 1.2886449098587036,
+      "learning_rate": 1e-06,
+      "loss": 0.0084,
+      "step": 3445
+    },
+    {
+      "epoch": 0.5866530473272046,
+      "grad_norm": 0.9253196716308594,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3446
+    },
+    {
+      "epoch": 0.5868232890704801,
+      "grad_norm": 1.2055374383926392,
+      "learning_rate": 1e-06,
+      "loss": 0.01,
+      "step": 3447
+    },
+    {
+      "epoch": 0.5869935308137555,
+      "grad_norm": 0.9988220930099487,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 3448
+    },
+    {
+      "epoch": 0.587163772557031,
+      "grad_norm": 0.9941924214363098,
+      "learning_rate": 1e-06,
+      "loss": 0.0072,
+      "step": 3449
+    },
+    {
+      "epoch": 0.5873340143003064,
+      "grad_norm": 1.3294003009796143,
+      "learning_rate": 1e-06,
+      "loss": 0.0099,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5875042560435819,
+      "grad_norm": 1.015142560005188,
+      "learning_rate": 1e-06,
+      "loss": 0.0106,
+      "step": 3451
+    },
+    {
+      "epoch": 0.5876744977868573,
+      "grad_norm": 1.1107553243637085,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 3452
+    },
+    {
+      "epoch": 0.5878447395301328,
+      "grad_norm": 1.1980547904968262,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 3453
+    },
+    {
+      "epoch": 0.5880149812734082,
+      "grad_norm": 1.1775740385055542,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 3454
+    },
+    {
+      "epoch": 0.5881852230166837,
+      "grad_norm": 1.5608429908752441,
+      "learning_rate": 1e-06,
+      "loss": 0.0131,
+      "step": 3455
+    },
+    {
+      "epoch": 0.5883554647599591,
+      "grad_norm": 1.3156591653823853,
+      "learning_rate": 1e-06,
+      "loss": 0.011,
+      "step": 3456
+    },
+    {
+      "epoch": 0.5885257065032345,
+      "grad_norm": 1.0954924821853638,
+      "learning_rate": 1e-06,
+      "loss": 0.0077,
+      "step": 3457
+    },
+    {
+      "epoch": 0.5886959482465101,
+      "grad_norm": 0.908723771572113,
+      "learning_rate": 1e-06,
+      "loss": 0.0067,
+      "step": 3458
+    },
+    {
+      "epoch": 0.5888661899897855,
+      "grad_norm": 0.9518954753875732,
+      "learning_rate": 1e-06,
+      "loss": 0.0065,
+      "step": 3459
+    },
+    {
+      "epoch": 0.589036431733061,
+      "grad_norm": 1.0838017463684082,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5892066734763364,
+      "grad_norm": 1.235297441482544,
+      "learning_rate": 1e-06,
+      "loss": 0.0149,
+      "step": 3461
+    },
+    {
+      "epoch": 0.5893769152196119,
+      "grad_norm": 1.1786282062530518,
+      "learning_rate": 1e-06,
+      "loss": 0.0094,
+      "step": 3462
+    },
+    {
+      "epoch": 0.5895471569628873,
+      "grad_norm": 1.1594127416610718,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3463
+    },
+    {
+      "epoch": 0.5897173987061628,
+      "grad_norm": 0.8134921193122864,
+      "learning_rate": 1e-06,
+      "loss": 0.0079,
+      "step": 3464
+    },
+    {
+      "epoch": 0.5898876404494382,
+      "grad_norm": 1.253074288368225,
+      "learning_rate": 1e-06,
+      "loss": 0.0132,
+      "step": 3465
+    },
+    {
+      "epoch": 0.5900578821927136,
+      "grad_norm": 0.9157419800758362,
+      "learning_rate": 1e-06,
+      "loss": 0.0062,
+      "step": 3466
+    },
+    {
+      "epoch": 0.5902281239359891,
+      "grad_norm": 0.8979124426841736,
+      "learning_rate": 1e-06,
+      "loss": 0.0077,
+      "step": 3467
+    },
+    {
+      "epoch": 0.5903983656792645,
+      "grad_norm": 0.9070771336555481,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3468
+    },
+    {
+      "epoch": 0.59056860742254,
+      "grad_norm": 1.6368584632873535,
+      "learning_rate": 1e-06,
+      "loss": 0.0089,
+      "step": 3469
+    },
+    {
+      "epoch": 0.5907388491658154,
+      "grad_norm": 1.0019363164901733,
+      "learning_rate": 1e-06,
+      "loss": 0.0077,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5909090909090909,
+      "grad_norm": 1.3202464580535889,
+      "learning_rate": 1e-06,
+      "loss": 0.0146,
+      "step": 3471
+    },
+    {
+      "epoch": 0.5910793326523663,
+      "grad_norm": 1.0647854804992676,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3472
+    },
+    {
+      "epoch": 0.5912495743956419,
+      "grad_norm": 1.138110637664795,
+      "learning_rate": 1e-06,
+      "loss": 0.0112,
+      "step": 3473
+    },
+    {
+      "epoch": 0.5914198161389173,
+      "grad_norm": 0.9572812914848328,
+      "learning_rate": 1e-06,
+      "loss": 0.0073,
+      "step": 3474
+    },
+    {
+      "epoch": 0.5915900578821928,
+      "grad_norm": 1.1067378520965576,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 3475
+    },
+    {
+      "epoch": 0.5917602996254682,
+      "grad_norm": 1.1624847650527954,
+      "learning_rate": 1e-06,
+      "loss": 0.015,
+      "step": 3476
+    },
+    {
+      "epoch": 0.5919305413687436,
+      "grad_norm": 1.1068007946014404,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 3477
+    },
+    {
+      "epoch": 0.5921007831120191,
+      "grad_norm": 0.9079290628433228,
+      "learning_rate": 1e-06,
+      "loss": 0.0082,
+      "step": 3478
+    },
+    {
+      "epoch": 0.5922710248552945,
+      "grad_norm": 1.3529939651489258,
+      "learning_rate": 1e-06,
+      "loss": 0.0145,
+      "step": 3479
+    },
+    {
+      "epoch": 0.59244126659857,
+      "grad_norm": 0.914438009262085,
+      "learning_rate": 1e-06,
+      "loss": 0.0073,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5926115083418454,
+      "grad_norm": 0.9735174775123596,
+      "learning_rate": 1e-06,
+      "loss": 0.0068,
+      "step": 3481
+    },
+    {
+      "epoch": 0.5927817500851209,
+      "grad_norm": 1.064338207244873,
+      "learning_rate": 1e-06,
+      "loss": 0.0061,
+      "step": 3482
+    },
+    {
+      "epoch": 0.5929519918283963,
+      "grad_norm": 1.145073413848877,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 3483
+    },
+    {
+      "epoch": 0.5931222335716718,
+      "grad_norm": 0.9396929144859314,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3484
+    },
+    {
+      "epoch": 0.5932924753149472,
+      "grad_norm": 0.8736780881881714,
+      "learning_rate": 1e-06,
+      "loss": 0.0063,
+      "step": 3485
+    },
+    {
+      "epoch": 0.5934627170582226,
+      "grad_norm": 1.157373070716858,
+      "learning_rate": 1e-06,
+      "loss": 0.0094,
+      "step": 3486
+    },
+    {
+      "epoch": 0.5936329588014981,
+      "grad_norm": 1.1017827987670898,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 3487
+    },
+    {
+      "epoch": 0.5938032005447735,
+      "grad_norm": 1.0480536222457886,
+      "learning_rate": 1e-06,
+      "loss": 0.0093,
+      "step": 3488
+    },
+    {
+      "epoch": 0.593973442288049,
+      "grad_norm": 0.9790979027748108,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3489
+    },
+    {
+      "epoch": 0.5941436840313244,
+      "grad_norm": 0.9106320142745972,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 3490
+    },
+    {
+      "epoch": 0.5943139257746,
+      "grad_norm": 0.946974515914917,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3491
+    },
+    {
+      "epoch": 0.5944841675178754,
+      "grad_norm": 1.0634249448776245,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3492
+    },
+    {
+      "epoch": 0.5946544092611509,
+      "grad_norm": 1.1991939544677734,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 3493
+    },
+    {
+      "epoch": 0.5948246510044263,
+      "grad_norm": 0.9980155229568481,
+      "learning_rate": 1e-06,
+      "loss": 0.0077,
+      "step": 3494
+    },
+    {
+      "epoch": 0.5949948927477018,
+      "grad_norm": 1.0674749612808228,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3495
+    },
+    {
+      "epoch": 0.5951651344909772,
+      "grad_norm": 0.9656354784965515,
+      "learning_rate": 1e-06,
+      "loss": 0.0067,
+      "step": 3496
+    },
+    {
+      "epoch": 0.5953353762342526,
+      "grad_norm": 1.1175800561904907,
+      "learning_rate": 1e-06,
+      "loss": 0.0084,
+      "step": 3497
+    },
+    {
+      "epoch": 0.5955056179775281,
+      "grad_norm": 0.8521952629089355,
+      "learning_rate": 1e-06,
+      "loss": 0.0076,
+      "step": 3498
+    },
+    {
+      "epoch": 0.5956758597208035,
+      "grad_norm": 1.266019344329834,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3499
+    },
+    {
+      "epoch": 0.595846101464079,
+      "grad_norm": 1.155662178993225,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5960163432073544,
+      "grad_norm": 1.200804352760315,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 3501
+    },
+    {
+      "epoch": 0.5961865849506299,
+      "grad_norm": 1.0332434177398682,
+      "learning_rate": 1e-06,
+      "loss": 0.0093,
+      "step": 3502
+    },
+    {
+      "epoch": 0.5963568266939053,
+      "grad_norm": 1.114933729171753,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 3503
+    },
+    {
+      "epoch": 0.5965270684371808,
+      "grad_norm": 1.211915135383606,
+      "learning_rate": 1e-06,
+      "loss": 0.0081,
+      "step": 3504
+    },
+    {
+      "epoch": 0.5966973101804562,
+      "grad_norm": 1.0030415058135986,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 3505
+    },
+    {
+      "epoch": 0.5968675519237318,
+      "grad_norm": 1.073610782623291,
+      "learning_rate": 1e-06,
+      "loss": 0.0077,
+      "step": 3506
+    },
+    {
+      "epoch": 0.5970377936670072,
+      "grad_norm": 0.9142771363258362,
+      "learning_rate": 1e-06,
+      "loss": 0.0065,
+      "step": 3507
+    },
+    {
+      "epoch": 0.5972080354102826,
+      "grad_norm": 1.1004252433776855,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 3508
+    },
+    {
+      "epoch": 0.5973782771535581,
+      "grad_norm": 1.1081154346466064,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3509
+    },
+    {
+      "epoch": 0.5975485188968335,
+      "grad_norm": 1.1693594455718994,
+      "learning_rate": 1e-06,
+      "loss": 0.0106,
+      "step": 3510
+    },
+    {
+      "epoch": 0.597718760640109,
+      "grad_norm": 1.2381548881530762,
+      "learning_rate": 1e-06,
+      "loss": 0.009,
+      "step": 3511
+    },
+    {
+      "epoch": 0.5978890023833844,
+      "grad_norm": 1.1379530429840088,
+      "learning_rate": 1e-06,
+      "loss": 0.0114,
+      "step": 3512
+    },
+    {
+      "epoch": 0.5980592441266599,
+      "grad_norm": 1.3537330627441406,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 3513
+    },
+    {
+      "epoch": 0.5982294858699353,
+      "grad_norm": 0.8830147981643677,
+      "learning_rate": 1e-06,
+      "loss": 0.009,
+      "step": 3514
+    },
+    {
+      "epoch": 0.5983997276132108,
+      "grad_norm": 1.09042489528656,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 3515
+    },
+    {
+      "epoch": 0.5985699693564862,
+      "grad_norm": 1.1855099201202393,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 3516
+    },
+    {
+      "epoch": 0.5987402110997616,
+      "grad_norm": 0.7881447672843933,
+      "learning_rate": 1e-06,
+      "loss": 0.0068,
+      "step": 3517
+    },
+    {
+      "epoch": 0.5989104528430371,
+      "grad_norm": 0.8492823243141174,
+      "learning_rate": 1e-06,
+      "loss": 0.0068,
+      "step": 3518
+    },
+    {
+      "epoch": 0.5990806945863125,
+      "grad_norm": 1.2348202466964722,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3519
+    },
+    {
+      "epoch": 0.599250936329588,
+      "grad_norm": 1.514280080795288,
+      "learning_rate": 1e-06,
+      "loss": 0.0134,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5994211780728634,
+      "grad_norm": 1.5777965784072876,
+      "learning_rate": 1e-06,
+      "loss": 0.0104,
+      "step": 3521
+    },
+    {
+      "epoch": 0.599591419816139,
+      "grad_norm": 1.3330321311950684,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 3522
+    },
+    {
+      "epoch": 0.5997616615594143,
+      "grad_norm": 1.630621314048767,
+      "learning_rate": 1e-06,
+      "loss": 0.0122,
+      "step": 3523
+    },
+    {
+      "epoch": 0.5999319033026899,
+      "grad_norm": 1.2280832529067993,
+      "learning_rate": 1e-06,
+      "loss": 0.0097,
+      "step": 3524
+    },
+    {
+      "epoch": 0.6001021450459653,
+      "grad_norm": 1.187753677368164,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6002723867892408,
+      "grad_norm": 1.6510494947433472,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3526
+    },
+    {
+      "epoch": 0.6004426285325162,
+      "grad_norm": 1.1509414911270142,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 3527
+    },
+    {
+      "epoch": 0.6006128702757916,
+      "grad_norm": 1.300317406654358,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 3528
+    },
+    {
+      "epoch": 0.6007831120190671,
+      "grad_norm": 1.013220191001892,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 3529
+    },
+    {
+      "epoch": 0.6009533537623425,
+      "grad_norm": 0.9694526195526123,
+      "learning_rate": 1e-06,
+      "loss": 0.0076,
+      "step": 3530
+    },
+    {
+      "epoch": 0.601123595505618,
+      "grad_norm": 1.1134835481643677,
+      "learning_rate": 1e-06,
+      "loss": 0.011,
+      "step": 3531
+    },
+    {
+      "epoch": 0.6012938372488934,
+      "grad_norm": 1.0825552940368652,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3532
+    },
+    {
+      "epoch": 0.6014640789921689,
+      "grad_norm": 1.277435064315796,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 3533
+    },
+    {
+      "epoch": 0.6016343207354443,
+      "grad_norm": 1.1507532596588135,
+      "learning_rate": 1e-06,
+      "loss": 0.009,
+      "step": 3534
+    },
+    {
+      "epoch": 0.6018045624787198,
+      "grad_norm": 1.923765778541565,
+      "learning_rate": 1e-06,
+      "loss": 0.0134,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6019748042219952,
+      "grad_norm": 1.156172275543213,
+      "learning_rate": 1e-06,
+      "loss": 0.0123,
+      "step": 3536
+    },
+    {
+      "epoch": 0.6021450459652706,
+      "grad_norm": 0.9838459491729736,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 3537
+    },
+    {
+      "epoch": 0.6023152877085461,
+      "grad_norm": 1.3179664611816406,
+      "learning_rate": 1e-06,
+      "loss": 0.0119,
+      "step": 3538
+    },
+    {
+      "epoch": 0.6024855294518215,
+      "grad_norm": 0.9081220626831055,
+      "learning_rate": 1e-06,
+      "loss": 0.0076,
+      "step": 3539
+    },
+    {
+      "epoch": 0.6026557711950971,
+      "grad_norm": 1.1752504110336304,
+      "learning_rate": 1e-06,
+      "loss": 0.0106,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6028260129383725,
+      "grad_norm": 0.9486705660820007,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 3541
+    },
+    {
+      "epoch": 0.602996254681648,
+      "grad_norm": 0.8370993733406067,
+      "learning_rate": 1e-06,
+      "loss": 0.0059,
+      "step": 3542
+    },
+    {
+      "epoch": 0.6031664964249234,
+      "grad_norm": 0.8764846920967102,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3543
+    },
+    {
+      "epoch": 0.6033367381681989,
+      "grad_norm": 1.19331955909729,
+      "learning_rate": 1e-06,
+      "loss": 0.0097,
+      "step": 3544
+    },
+    {
+      "epoch": 0.6035069799114743,
+      "grad_norm": 0.8320559859275818,
+      "learning_rate": 1e-06,
+      "loss": 0.0074,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6036772216547498,
+      "grad_norm": 1.54590904712677,
+      "learning_rate": 1e-06,
+      "loss": 0.0126,
+      "step": 3546
+    },
+    {
+      "epoch": 0.6038474633980252,
+      "grad_norm": 0.878666877746582,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3547
+    },
+    {
+      "epoch": 0.6040177051413006,
+      "grad_norm": 1.5745292901992798,
+      "learning_rate": 1e-06,
+      "loss": 0.0122,
+      "step": 3548
+    },
+    {
+      "epoch": 0.6041879468845761,
+      "grad_norm": 0.9554889798164368,
+      "learning_rate": 1e-06,
+      "loss": 0.0072,
+      "step": 3549
+    },
+    {
+      "epoch": 0.6043581886278515,
+      "grad_norm": 1.1246119737625122,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3550
+    },
+    {
+      "epoch": 0.604528430371127,
+      "grad_norm": 1.0322120189666748,
+      "learning_rate": 1e-06,
+      "loss": 0.0063,
+      "step": 3551
+    },
+    {
+      "epoch": 0.6046986721144024,
+      "grad_norm": 0.9092923998832703,
+      "learning_rate": 1e-06,
+      "loss": 0.0064,
+      "step": 3552
+    },
+    {
+      "epoch": 0.6048689138576779,
+      "grad_norm": 1.2549725770950317,
+      "learning_rate": 1e-06,
+      "loss": 0.0141,
+      "step": 3553
+    },
+    {
+      "epoch": 0.6050391556009533,
+      "grad_norm": 1.0988869667053223,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 3554
+    },
+    {
+      "epoch": 0.6052093973442288,
+      "grad_norm": 1.0146656036376953,
+      "learning_rate": 1e-06,
+      "loss": 0.0082,
+      "step": 3555
+    },
+    {
+      "epoch": 0.6053796390875043,
+      "grad_norm": 1.2625722885131836,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3556
+    },
+    {
+      "epoch": 0.6055498808307797,
+      "grad_norm": 1.0923871994018555,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3557
+    },
+    {
+      "epoch": 0.6057201225740552,
+      "grad_norm": 0.8863467574119568,
+      "learning_rate": 1e-06,
+      "loss": 0.006,
+      "step": 3558
+    },
+    {
+      "epoch": 0.6058903643173306,
+      "grad_norm": 1.0210617780685425,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3559
+    },
+    {
+      "epoch": 0.6060606060606061,
+      "grad_norm": 2.0322105884552,
+      "learning_rate": 1e-06,
+      "loss": 0.02,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6062308478038815,
+      "grad_norm": 1.0655431747436523,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3561
+    },
+    {
+      "epoch": 0.606401089547157,
+      "grad_norm": 1.0927060842514038,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 3562
+    },
+    {
+      "epoch": 0.6065713312904324,
+      "grad_norm": 1.0595030784606934,
+      "learning_rate": 1e-06,
+      "loss": 0.0082,
+      "step": 3563
+    },
+    {
+      "epoch": 0.6067415730337079,
+      "grad_norm": 1.1079046726226807,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3564
+    },
+    {
+      "epoch": 0.6069118147769833,
+      "grad_norm": 1.0296462774276733,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6070820565202588,
+      "grad_norm": 0.868334949016571,
+      "learning_rate": 1e-06,
+      "loss": 0.0065,
+      "step": 3566
+    },
+    {
+      "epoch": 0.6072522982635342,
+      "grad_norm": 1.4144976139068604,
+      "learning_rate": 1e-06,
+      "loss": 0.0141,
+      "step": 3567
+    },
+    {
+      "epoch": 0.6074225400068096,
+      "grad_norm": 1.1668305397033691,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3568
+    },
+    {
+      "epoch": 0.6075927817500851,
+      "grad_norm": 1.2079553604125977,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3569
+    },
+    {
+      "epoch": 0.6077630234933605,
+      "grad_norm": 1.1177006959915161,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 3570
+    },
+    {
+      "epoch": 0.607933265236636,
+      "grad_norm": 1.1665843725204468,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 3571
+    },
+    {
+      "epoch": 0.6081035069799114,
+      "grad_norm": 1.0102547407150269,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3572
+    },
+    {
+      "epoch": 0.608273748723187,
+      "grad_norm": 0.9635432362556458,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3573
+    },
+    {
+      "epoch": 0.6084439904664624,
+      "grad_norm": 1.3346338272094727,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 3574
+    },
+    {
+      "epoch": 0.6086142322097379,
+      "grad_norm": 1.608765721321106,
+      "learning_rate": 1e-06,
+      "loss": 0.0268,
+      "step": 3575
+    },
+    {
+      "epoch": 0.6087844739530133,
+      "grad_norm": 0.9008107781410217,
+      "learning_rate": 1e-06,
+      "loss": 0.0071,
+      "step": 3576
+    },
+    {
+      "epoch": 0.6089547156962888,
+      "grad_norm": 1.0293365716934204,
+      "learning_rate": 1e-06,
+      "loss": 0.0061,
+      "step": 3577
+    },
+    {
+      "epoch": 0.6091249574395642,
+      "grad_norm": 1.082380771636963,
+      "learning_rate": 1e-06,
+      "loss": 0.011,
+      "step": 3578
+    },
+    {
+      "epoch": 0.6092951991828396,
+      "grad_norm": 0.7760728597640991,
+      "learning_rate": 1e-06,
+      "loss": 0.0069,
+      "step": 3579
+    },
+    {
+      "epoch": 0.6094654409261151,
+      "grad_norm": 0.9977663159370422,
+      "learning_rate": 1e-06,
+      "loss": 0.0057,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6096356826693905,
+      "grad_norm": 0.7965632081031799,
+      "learning_rate": 1e-06,
+      "loss": 0.0066,
+      "step": 3581
+    },
+    {
+      "epoch": 0.609805924412666,
+      "grad_norm": 1.3011534214019775,
+      "learning_rate": 1e-06,
+      "loss": 0.0158,
+      "step": 3582
+    },
+    {
+      "epoch": 0.6099761661559414,
+      "grad_norm": 1.1788239479064941,
+      "learning_rate": 1e-06,
+      "loss": 0.0141,
+      "step": 3583
+    },
+    {
+      "epoch": 0.6101464078992169,
+      "grad_norm": 1.0740551948547363,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 3584
+    },
+    {
+      "epoch": 0.6103166496424923,
+      "grad_norm": 1.2038638591766357,
+      "learning_rate": 1e-06,
+      "loss": 0.0105,
+      "step": 3585
+    },
+    {
+      "epoch": 0.6104868913857678,
+      "grad_norm": 1.0219711065292358,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 3586
+    },
+    {
+      "epoch": 0.6106571331290432,
+      "grad_norm": 1.2618956565856934,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 3587
+    },
+    {
+      "epoch": 0.6108273748723186,
+      "grad_norm": 1.1306350231170654,
+      "learning_rate": 1e-06,
+      "loss": 0.007,
+      "step": 3588
+    },
+    {
+      "epoch": 0.6109976166155942,
+      "grad_norm": 0.8603691458702087,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3589
+    },
+    {
+      "epoch": 0.6111678583588696,
+      "grad_norm": 1.2200685739517212,
+      "learning_rate": 1e-06,
+      "loss": 0.0105,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6113381001021451,
+      "grad_norm": 1.4106483459472656,
+      "learning_rate": 1e-06,
+      "loss": 0.0089,
+      "step": 3591
+    },
+    {
+      "epoch": 0.6115083418454205,
+      "grad_norm": 1.3192464113235474,
+      "learning_rate": 1e-06,
+      "loss": 0.0123,
+      "step": 3592
+    },
+    {
+      "epoch": 0.611678583588696,
+      "grad_norm": 1.1287734508514404,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3593
+    },
+    {
+      "epoch": 0.6118488253319714,
+      "grad_norm": 1.193594217300415,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 3594
+    },
+    {
+      "epoch": 0.6120190670752469,
+      "grad_norm": 1.3210780620574951,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 3595
+    },
+    {
+      "epoch": 0.6121893088185223,
+      "grad_norm": 1.2266240119934082,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3596
+    },
+    {
+      "epoch": 0.6123595505617978,
+      "grad_norm": 1.1135960817337036,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3597
+    },
+    {
+      "epoch": 0.6125297923050732,
+      "grad_norm": 1.2418755292892456,
+      "learning_rate": 1e-06,
+      "loss": 0.0104,
+      "step": 3598
+    },
+    {
+      "epoch": 0.6127000340483486,
+      "grad_norm": 1.2024139165878296,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 3599
+    },
+    {
+      "epoch": 0.6128702757916241,
+      "grad_norm": 0.8348981142044067,
+      "learning_rate": 1e-06,
+      "loss": 0.0071,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6130405175348995,
+      "grad_norm": 1.1524181365966797,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 3601
+    },
+    {
+      "epoch": 0.613210759278175,
+      "grad_norm": 1.0412416458129883,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 3602
+    },
+    {
+      "epoch": 0.6133810010214504,
+      "grad_norm": 0.7285078763961792,
+      "learning_rate": 1e-06,
+      "loss": 0.0063,
+      "step": 3603
+    },
+    {
+      "epoch": 0.613551242764726,
+      "grad_norm": 0.9448837041854858,
+      "learning_rate": 1e-06,
+      "loss": 0.0064,
+      "step": 3604
+    },
+    {
+      "epoch": 0.6137214845080013,
+      "grad_norm": 0.8977110385894775,
+      "learning_rate": 1e-06,
+      "loss": 0.007,
+      "step": 3605
+    },
+    {
+      "epoch": 0.6138917262512769,
+      "grad_norm": 1.5285866260528564,
+      "learning_rate": 1e-06,
+      "loss": 0.0143,
+      "step": 3606
+    },
+    {
+      "epoch": 0.6140619679945523,
+      "grad_norm": 1.3463897705078125,
+      "learning_rate": 1e-06,
+      "loss": 0.0138,
+      "step": 3607
+    },
+    {
+      "epoch": 0.6142322097378277,
+      "grad_norm": 1.2851765155792236,
+      "learning_rate": 1e-06,
+      "loss": 0.0084,
+      "step": 3608
+    },
+    {
+      "epoch": 0.6144024514811032,
+      "grad_norm": 1.5624653100967407,
+      "learning_rate": 1e-06,
+      "loss": 0.0094,
+      "step": 3609
+    },
+    {
+      "epoch": 0.6145726932243786,
+      "grad_norm": 1.6467691659927368,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6147429349676541,
+      "grad_norm": 1.128787636756897,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3611
+    },
+    {
+      "epoch": 0.6149131767109295,
+      "grad_norm": 1.4503066539764404,
+      "learning_rate": 1e-06,
+      "loss": 0.02,
+      "step": 3612
+    },
+    {
+      "epoch": 0.615083418454205,
+      "grad_norm": 0.8577298521995544,
+      "learning_rate": 1e-06,
+      "loss": 0.0079,
+      "step": 3613
+    },
+    {
+      "epoch": 0.6152536601974804,
+      "grad_norm": 1.0738738775253296,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3614
+    },
+    {
+      "epoch": 0.6154239019407559,
+      "grad_norm": 0.769759476184845,
+      "learning_rate": 1e-06,
+      "loss": 0.0058,
+      "step": 3615
+    },
+    {
+      "epoch": 0.6155941436840313,
+      "grad_norm": 0.8557948470115662,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 3616
+    },
+    {
+      "epoch": 0.6157643854273068,
+      "grad_norm": 0.8239853382110596,
+      "learning_rate": 1e-06,
+      "loss": 0.0079,
+      "step": 3617
+    },
+    {
+      "epoch": 0.6159346271705822,
+      "grad_norm": 0.9858947396278381,
+      "learning_rate": 1e-06,
+      "loss": 0.007,
+      "step": 3618
+    },
+    {
+      "epoch": 0.6161048689138576,
+      "grad_norm": 1.341850996017456,
+      "learning_rate": 1e-06,
+      "loss": 0.0122,
+      "step": 3619
+    },
+    {
+      "epoch": 0.6162751106571331,
+      "grad_norm": 0.8130425214767456,
+      "learning_rate": 1e-06,
+      "loss": 0.0072,
+      "step": 3620
+    },
+    {
+      "epoch": 0.6164453524004085,
+      "grad_norm": 0.826560378074646,
+      "learning_rate": 1e-06,
+      "loss": 0.0053,
+      "step": 3621
+    },
+    {
+      "epoch": 0.616615594143684,
+      "grad_norm": 1.0723116397857666,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3622
+    },
+    {
+      "epoch": 0.6167858358869595,
+      "grad_norm": 0.9984582662582397,
+      "learning_rate": 1e-06,
+      "loss": 0.0082,
+      "step": 3623
+    },
+    {
+      "epoch": 0.616956077630235,
+      "grad_norm": 1.1177350282669067,
+      "learning_rate": 1e-06,
+      "loss": 0.0114,
+      "step": 3624
+    },
+    {
+      "epoch": 0.6171263193735104,
+      "grad_norm": 1.006271243095398,
+      "learning_rate": 1e-06,
+      "loss": 0.0061,
+      "step": 3625
+    },
+    {
+      "epoch": 0.6172965611167859,
+      "grad_norm": 1.1038841009140015,
+      "learning_rate": 1e-06,
+      "loss": 0.0068,
+      "step": 3626
+    },
+    {
+      "epoch": 0.6174668028600613,
+      "grad_norm": 1.2689024209976196,
+      "learning_rate": 1e-06,
+      "loss": 0.0097,
+      "step": 3627
+    },
+    {
+      "epoch": 0.6176370446033368,
+      "grad_norm": 1.8578965663909912,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 3628
+    },
+    {
+      "epoch": 0.6178072863466122,
+      "grad_norm": 1.2967028617858887,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3629
+    },
+    {
+      "epoch": 0.6179775280898876,
+      "grad_norm": 0.9618591666221619,
+      "learning_rate": 1e-06,
+      "loss": 0.0069,
+      "step": 3630
+    },
+    {
+      "epoch": 0.6181477698331631,
+      "grad_norm": 0.8689693212509155,
+      "learning_rate": 1e-06,
+      "loss": 0.0073,
+      "step": 3631
+    },
+    {
+      "epoch": 0.6183180115764385,
+      "grad_norm": 1.112473487854004,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 3632
+    },
+    {
+      "epoch": 0.618488253319714,
+      "grad_norm": 1.1240233182907104,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3633
+    },
+    {
+      "epoch": 0.6186584950629894,
+      "grad_norm": 0.7019220590591431,
+      "learning_rate": 1e-06,
+      "loss": 0.0084,
+      "step": 3634
+    },
+    {
+      "epoch": 0.6188287368062649,
+      "grad_norm": 0.8149914145469666,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3635
+    },
+    {
+      "epoch": 0.6189989785495403,
+      "grad_norm": 1.152665376663208,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 3636
+    },
+    {
+      "epoch": 0.6191692202928158,
+      "grad_norm": 1.2790850400924683,
+      "learning_rate": 1e-06,
+      "loss": 0.0134,
+      "step": 3637
+    },
+    {
+      "epoch": 0.6193394620360912,
+      "grad_norm": 1.127205491065979,
+      "learning_rate": 1e-06,
+      "loss": 0.0093,
+      "step": 3638
+    },
+    {
+      "epoch": 0.6195097037793666,
+      "grad_norm": 1.7708073854446411,
+      "learning_rate": 1e-06,
+      "loss": 0.0252,
+      "step": 3639
+    },
+    {
+      "epoch": 0.6196799455226422,
+      "grad_norm": 1.142716646194458,
+      "learning_rate": 1e-06,
+      "loss": 0.0147,
+      "step": 3640
+    },
+    {
+      "epoch": 0.6198501872659176,
+      "grad_norm": 0.881137490272522,
+      "learning_rate": 1e-06,
+      "loss": 0.0069,
+      "step": 3641
+    },
+    {
+      "epoch": 0.6200204290091931,
+      "grad_norm": 1.1176531314849854,
+      "learning_rate": 1e-06,
+      "loss": 0.009,
+      "step": 3642
+    },
+    {
+      "epoch": 0.6201906707524685,
+      "grad_norm": 1.3249547481536865,
+      "learning_rate": 1e-06,
+      "loss": 0.0089,
+      "step": 3643
+    },
+    {
+      "epoch": 0.620360912495744,
+      "grad_norm": 2.152984142303467,
+      "learning_rate": 1e-06,
+      "loss": 0.0188,
+      "step": 3644
+    },
+    {
+      "epoch": 0.6205311542390194,
+      "grad_norm": 1.2621212005615234,
+      "learning_rate": 1e-06,
+      "loss": 0.0153,
+      "step": 3645
+    },
+    {
+      "epoch": 0.6207013959822949,
+      "grad_norm": 1.1430959701538086,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 3646
+    },
+    {
+      "epoch": 0.6208716377255703,
+      "grad_norm": 1.2035990953445435,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3647
+    },
+    {
+      "epoch": 0.6210418794688458,
+      "grad_norm": 1.1001369953155518,
+      "learning_rate": 1e-06,
+      "loss": 0.0081,
+      "step": 3648
+    },
+    {
+      "epoch": 0.6212121212121212,
+      "grad_norm": 1.1591325998306274,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3649
+    },
+    {
+      "epoch": 0.6213823629553966,
+      "grad_norm": 0.8609371781349182,
+      "learning_rate": 1e-06,
+      "loss": 0.0077,
+      "step": 3650
+    },
+    {
+      "epoch": 0.6215526046986721,
+      "grad_norm": 1.3064498901367188,
+      "learning_rate": 1e-06,
+      "loss": 0.0189,
+      "step": 3651
+    },
+    {
+      "epoch": 0.6217228464419475,
+      "grad_norm": 1.1729905605316162,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 3652
+    },
+    {
+      "epoch": 0.621893088185223,
+      "grad_norm": 1.064233660697937,
+      "learning_rate": 1e-06,
+      "loss": 0.0138,
+      "step": 3653
+    },
+    {
+      "epoch": 0.6220633299284984,
+      "grad_norm": 1.0935028791427612,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 3654
+    },
+    {
+      "epoch": 0.622233571671774,
+      "grad_norm": 0.9840522408485413,
+      "learning_rate": 1e-06,
+      "loss": 0.0074,
+      "step": 3655
+    },
+    {
+      "epoch": 0.6224038134150494,
+      "grad_norm": 1.152642846107483,
+      "learning_rate": 1e-06,
+      "loss": 0.0076,
+      "step": 3656
+    },
+    {
+      "epoch": 0.6225740551583249,
+      "grad_norm": 1.3375604152679443,
+      "learning_rate": 1e-06,
+      "loss": 0.011,
+      "step": 3657
+    },
+    {
+      "epoch": 0.6227442969016003,
+      "grad_norm": 1.8616725206375122,
+      "learning_rate": 1e-06,
+      "loss": 0.0119,
+      "step": 3658
+    },
+    {
+      "epoch": 0.6229145386448757,
+      "grad_norm": 1.0592864751815796,
+      "learning_rate": 1e-06,
+      "loss": 0.0077,
+      "step": 3659
+    },
+    {
+      "epoch": 0.6230847803881512,
+      "grad_norm": 1.135156512260437,
+      "learning_rate": 1e-06,
+      "loss": 0.0082,
+      "step": 3660
+    },
+    {
+      "epoch": 0.6232550221314266,
+      "grad_norm": 1.1434532403945923,
+      "learning_rate": 1e-06,
+      "loss": 0.0079,
+      "step": 3661
+    },
+    {
+      "epoch": 0.6234252638747021,
+      "grad_norm": 1.0572514533996582,
+      "learning_rate": 1e-06,
+      "loss": 0.0076,
+      "step": 3662
+    },
+    {
+      "epoch": 0.6235955056179775,
+      "grad_norm": 1.5850212574005127,
+      "learning_rate": 1e-06,
+      "loss": 0.0124,
+      "step": 3663
+    },
+    {
+      "epoch": 0.623765747361253,
+      "grad_norm": 1.4172906875610352,
+      "learning_rate": 1e-06,
+      "loss": 0.0128,
+      "step": 3664
+    },
+    {
+      "epoch": 0.6239359891045284,
+      "grad_norm": 1.1192266941070557,
+      "learning_rate": 1e-06,
+      "loss": 0.0114,
+      "step": 3665
+    },
+    {
+      "epoch": 0.6241062308478039,
+      "grad_norm": 1.0807698965072632,
+      "learning_rate": 1e-06,
+      "loss": 0.0068,
+      "step": 3666
+    },
+    {
+      "epoch": 0.6242764725910793,
+      "grad_norm": 1.0555567741394043,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 3667
+    },
+    {
+      "epoch": 0.6244467143343548,
+      "grad_norm": 1.2566262483596802,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3668
+    },
+    {
+      "epoch": 0.6246169560776302,
+      "grad_norm": 0.9095777273178101,
+      "learning_rate": 1e-06,
+      "loss": 0.0074,
+      "step": 3669
+    },
+    {
+      "epoch": 0.6247871978209056,
+      "grad_norm": 0.8502013683319092,
+      "learning_rate": 1e-06,
+      "loss": 0.0067,
+      "step": 3670
+    },
+    {
+      "epoch": 0.6249574395641811,
+      "grad_norm": 1.1455646753311157,
+      "learning_rate": 1e-06,
+      "loss": 0.0079,
+      "step": 3671
+    },
+    {
+      "epoch": 0.6251276813074566,
+      "grad_norm": 1.1766411066055298,
+      "learning_rate": 1e-06,
+      "loss": 0.0074,
+      "step": 3672
+    },
+    {
+      "epoch": 0.6252979230507321,
+      "grad_norm": 1.0214406251907349,
+      "learning_rate": 1e-06,
+      "loss": 0.007,
+      "step": 3673
+    },
+    {
+      "epoch": 0.6254681647940075,
+      "grad_norm": 1.4715005159378052,
+      "learning_rate": 1e-06,
+      "loss": 0.0187,
+      "step": 3674
+    },
+    {
+      "epoch": 0.625638406537283,
+      "grad_norm": 0.9039010405540466,
+      "learning_rate": 1e-06,
+      "loss": 0.0059,
+      "step": 3675
+    },
+    {
+      "epoch": 0.6258086482805584,
+      "grad_norm": 1.0967904329299927,
+      "learning_rate": 1e-06,
+      "loss": 0.0069,
+      "step": 3676
+    },
+    {
+      "epoch": 0.6259788900238339,
+      "grad_norm": 1.013299822807312,
+      "learning_rate": 1e-06,
+      "loss": 0.0072,
+      "step": 3677
+    },
+    {
+      "epoch": 0.6261491317671093,
+      "grad_norm": 1.229809045791626,
+      "learning_rate": 1e-06,
+      "loss": 0.0094,
+      "step": 3678
+    },
+    {
+      "epoch": 0.6263193735103847,
+      "grad_norm": 1.207761287689209,
+      "learning_rate": 1e-06,
+      "loss": 0.0063,
+      "step": 3679
+    },
+    {
+      "epoch": 0.6264896152536602,
+      "grad_norm": 1.036696195602417,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3680
+    },
+    {
+      "epoch": 0.6266598569969356,
+      "grad_norm": 1.182930827140808,
+      "learning_rate": 1e-06,
+      "loss": 0.0099,
+      "step": 3681
+    },
+    {
+      "epoch": 0.6268300987402111,
+      "grad_norm": 1.3283963203430176,
+      "learning_rate": 1e-06,
+      "loss": 0.0164,
+      "step": 3682
+    },
+    {
+      "epoch": 0.6270003404834865,
+      "grad_norm": 1.021653175354004,
+      "learning_rate": 1e-06,
+      "loss": 0.0093,
+      "step": 3683
+    },
+    {
+      "epoch": 0.627170582226762,
+      "grad_norm": 1.5096207857131958,
+      "learning_rate": 1e-06,
+      "loss": 0.0142,
+      "step": 3684
+    },
+    {
+      "epoch": 0.6273408239700374,
+      "grad_norm": 1.1412389278411865,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 3685
+    },
+    {
+      "epoch": 0.6275110657133129,
+      "grad_norm": 1.141716718673706,
+      "learning_rate": 1e-06,
+      "loss": 0.0082,
+      "step": 3686
+    },
+    {
+      "epoch": 0.6276813074565883,
+      "grad_norm": 1.1129909753799438,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 3687
+    },
+    {
+      "epoch": 0.6278515491998639,
+      "grad_norm": 0.8182403445243835,
+      "learning_rate": 1e-06,
+      "loss": 0.0064,
+      "step": 3688
+    },
+    {
+      "epoch": 0.6280217909431393,
+      "grad_norm": 1.2895714044570923,
+      "learning_rate": 1e-06,
+      "loss": 0.0184,
+      "step": 3689
+    },
+    {
+      "epoch": 0.6281920326864147,
+      "grad_norm": 0.8942720293998718,
+      "learning_rate": 1e-06,
+      "loss": 0.0069,
+      "step": 3690
+    },
+    {
+      "epoch": 0.6283622744296902,
+      "grad_norm": 1.0091437101364136,
+      "learning_rate": 1e-06,
+      "loss": 0.0076,
+      "step": 3691
+    },
+    {
+      "epoch": 0.6285325161729656,
+      "grad_norm": 1.1240427494049072,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 3692
+    },
+    {
+      "epoch": 0.6287027579162411,
+      "grad_norm": 1.328053593635559,
+      "learning_rate": 1e-06,
+      "loss": 0.0125,
+      "step": 3693
+    },
+    {
+      "epoch": 0.6288729996595165,
+      "grad_norm": 0.9419410824775696,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 3694
+    },
+    {
+      "epoch": 0.629043241402792,
+      "grad_norm": 0.7652695178985596,
+      "learning_rate": 1e-06,
+      "loss": 0.0057,
+      "step": 3695
+    },
+    {
+      "epoch": 0.6292134831460674,
+      "grad_norm": 1.2899373769760132,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3696
+    },
+    {
+      "epoch": 0.6293837248893429,
+      "grad_norm": 1.0259873867034912,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 3697
+    },
+    {
+      "epoch": 0.6295539666326183,
+      "grad_norm": 1.1988884210586548,
+      "learning_rate": 1e-06,
+      "loss": 0.0072,
+      "step": 3698
+    },
+    {
+      "epoch": 0.6297242083758938,
+      "grad_norm": 0.9181260466575623,
+      "learning_rate": 1e-06,
+      "loss": 0.0071,
+      "step": 3699
+    },
+    {
+      "epoch": 0.6298944501191692,
+      "grad_norm": 0.9410200119018555,
+      "learning_rate": 1e-06,
+      "loss": 0.0074,
+      "step": 3700
+    },
+    {
+      "epoch": 0.6300646918624446,
+      "grad_norm": 1.4462769031524658,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 3701
+    },
+    {
+      "epoch": 0.6302349336057201,
+      "grad_norm": 1.0774480104446411,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3702
+    },
+    {
+      "epoch": 0.6304051753489955,
+      "grad_norm": 1.004542589187622,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 3703
+    },
+    {
+      "epoch": 0.630575417092271,
+      "grad_norm": 1.9113359451293945,
+      "learning_rate": 1e-06,
+      "loss": 0.0168,
+      "step": 3704
+    },
+    {
+      "epoch": 0.6307456588355465,
+      "grad_norm": 1.03765869140625,
+      "learning_rate": 1e-06,
+      "loss": 0.0076,
+      "step": 3705
+    },
+    {
+      "epoch": 0.630915900578822,
+      "grad_norm": 1.1582138538360596,
+      "learning_rate": 1e-06,
+      "loss": 0.0072,
+      "step": 3706
+    },
+    {
+      "epoch": 0.6310861423220974,
+      "grad_norm": 1.195564866065979,
+      "learning_rate": 1e-06,
+      "loss": 0.0094,
+      "step": 3707
+    },
+    {
+      "epoch": 0.6312563840653729,
+      "grad_norm": 0.9630500674247742,
+      "learning_rate": 1e-06,
+      "loss": 0.0084,
+      "step": 3708
+    },
+    {
+      "epoch": 0.6314266258086483,
+      "grad_norm": 1.4461884498596191,
+      "learning_rate": 1e-06,
+      "loss": 0.0123,
+      "step": 3709
+    },
+    {
+      "epoch": 0.6315968675519237,
+      "grad_norm": 0.9124757051467896,
+      "learning_rate": 1e-06,
+      "loss": 0.0071,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6317671092951992,
+      "grad_norm": 1.134900450706482,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 3711
+    },
+    {
+      "epoch": 0.6319373510384746,
+      "grad_norm": 1.1023823022842407,
+      "learning_rate": 1e-06,
+      "loss": 0.0074,
+      "step": 3712
+    },
+    {
+      "epoch": 0.6321075927817501,
+      "grad_norm": 1.0747976303100586,
+      "learning_rate": 1e-06,
+      "loss": 0.0111,
+      "step": 3713
+    },
+    {
+      "epoch": 0.6322778345250255,
+      "grad_norm": 0.8265736103057861,
+      "learning_rate": 1e-06,
+      "loss": 0.007,
+      "step": 3714
+    },
+    {
+      "epoch": 0.632448076268301,
+      "grad_norm": 1.0795259475708008,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3715
+    },
+    {
+      "epoch": 0.6326183180115764,
+      "grad_norm": 1.0878645181655884,
+      "learning_rate": 1e-06,
+      "loss": 0.0074,
+      "step": 3716
+    },
+    {
+      "epoch": 0.6327885597548519,
+      "grad_norm": 0.9701595306396484,
+      "learning_rate": 1e-06,
+      "loss": 0.0072,
+      "step": 3717
+    },
+    {
+      "epoch": 0.6329588014981273,
+      "grad_norm": 1.2201334238052368,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 3718
+    },
+    {
+      "epoch": 0.6331290432414028,
+      "grad_norm": 1.0813226699829102,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 3719
+    },
+    {
+      "epoch": 0.6332992849846782,
+      "grad_norm": 0.7975080609321594,
+      "learning_rate": 1e-06,
+      "loss": 0.0056,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6334695267279536,
+      "grad_norm": 1.270108938217163,
+      "learning_rate": 1e-06,
+      "loss": 0.0127,
+      "step": 3721
+    },
+    {
+      "epoch": 0.6336397684712292,
+      "grad_norm": 1.3048897981643677,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3722
+    },
+    {
+      "epoch": 0.6338100102145046,
+      "grad_norm": 2.3595523834228516,
+      "learning_rate": 1e-06,
+      "loss": 0.0136,
+      "step": 3723
+    },
+    {
+      "epoch": 0.6339802519577801,
+      "grad_norm": 1.565253734588623,
+      "learning_rate": 1e-06,
+      "loss": 0.0177,
+      "step": 3724
+    },
+    {
+      "epoch": 0.6341504937010555,
+      "grad_norm": 0.9214725494384766,
+      "learning_rate": 1e-06,
+      "loss": 0.0073,
+      "step": 3725
+    },
+    {
+      "epoch": 0.634320735444331,
+      "grad_norm": 0.9782243967056274,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3726
+    },
+    {
+      "epoch": 0.6344909771876064,
+      "grad_norm": 0.9782243967056274,
+      "learning_rate": 1e-06,
+      "loss": 0.0266,
+      "step": 3727
+    },
+    {
+      "epoch": 0.6346612189308819,
+      "grad_norm": 1.1668390035629272,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3728
+    },
+    {
+      "epoch": 0.6348314606741573,
+      "grad_norm": 1.1563256978988647,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 3729
+    },
+    {
+      "epoch": 0.6350017024174327,
+      "grad_norm": 0.9238855242729187,
+      "learning_rate": 1e-06,
+      "loss": 0.0089,
+      "step": 3730
+    },
+    {
+      "epoch": 0.6351719441607082,
+      "grad_norm": 0.9837780594825745,
+      "learning_rate": 1e-06,
+      "loss": 0.0054,
+      "step": 3731
+    },
+    {
+      "epoch": 0.6353421859039836,
+      "grad_norm": 1.0795369148254395,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 3732
+    },
+    {
+      "epoch": 0.6355124276472591,
+      "grad_norm": 0.8233152627944946,
+      "learning_rate": 1e-06,
+      "loss": 0.0056,
+      "step": 3733
+    },
+    {
+      "epoch": 0.6356826693905345,
+      "grad_norm": 0.9861051440238953,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 3734
+    },
+    {
+      "epoch": 0.63585291113381,
+      "grad_norm": 1.1408077478408813,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 3735
+    },
+    {
+      "epoch": 0.6360231528770854,
+      "grad_norm": 1.1144981384277344,
+      "learning_rate": 1e-06,
+      "loss": 0.0089,
+      "step": 3736
+    },
+    {
+      "epoch": 0.636193394620361,
+      "grad_norm": 1.3315359354019165,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 3737
+    },
+    {
+      "epoch": 0.6363636363636364,
+      "grad_norm": 0.9346138834953308,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3738
+    },
+    {
+      "epoch": 0.6365338781069119,
+      "grad_norm": 1.3445444107055664,
+      "learning_rate": 1e-06,
+      "loss": 0.0218,
+      "step": 3739
+    },
+    {
+      "epoch": 0.6367041198501873,
+      "grad_norm": 1.3266631364822388,
+      "learning_rate": 1e-06,
+      "loss": 0.0081,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6368743615934627,
+      "grad_norm": 0.7501433491706848,
+      "learning_rate": 1e-06,
+      "loss": 0.0066,
+      "step": 3741
+    },
+    {
+      "epoch": 0.6370446033367382,
+      "grad_norm": 0.9168145060539246,
+      "learning_rate": 1e-06,
+      "loss": 0.0071,
+      "step": 3742
+    },
+    {
+      "epoch": 0.6372148450800136,
+      "grad_norm": 0.8834933042526245,
+      "learning_rate": 1e-06,
+      "loss": 0.006,
+      "step": 3743
+    },
+    {
+      "epoch": 0.6373850868232891,
+      "grad_norm": 1.115196943283081,
+      "learning_rate": 1e-06,
+      "loss": 0.009,
+      "step": 3744
+    },
+    {
+      "epoch": 0.6375553285665645,
+      "grad_norm": 1.0395985841751099,
+      "learning_rate": 1e-06,
+      "loss": 0.0068,
+      "step": 3745
+    },
+    {
+      "epoch": 0.63772557030984,
+      "grad_norm": 1.1208879947662354,
+      "learning_rate": 1e-06,
+      "loss": 0.0079,
+      "step": 3746
+    },
+    {
+      "epoch": 0.6378958120531154,
+      "grad_norm": 1.482139229774475,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 3747
+    },
+    {
+      "epoch": 0.6380660537963909,
+      "grad_norm": 1.3802002668380737,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 3748
+    },
+    {
+      "epoch": 0.6382362955396663,
+      "grad_norm": 1.1360260248184204,
+      "learning_rate": 1e-06,
+      "loss": 0.0093,
+      "step": 3749
+    },
+    {
+      "epoch": 0.6384065372829418,
+      "grad_norm": 1.353955864906311,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6385767790262172,
+      "grad_norm": 1.0488437414169312,
+      "learning_rate": 1e-06,
+      "loss": 0.0069,
+      "step": 3751
+    },
+    {
+      "epoch": 0.6387470207694926,
+      "grad_norm": 1.547584891319275,
+      "learning_rate": 1e-06,
+      "loss": 0.0152,
+      "step": 3752
+    },
+    {
+      "epoch": 0.6389172625127681,
+      "grad_norm": 1.0610159635543823,
+      "learning_rate": 1e-06,
+      "loss": 0.0074,
+      "step": 3753
+    },
+    {
+      "epoch": 0.6390875042560435,
+      "grad_norm": 1.1896774768829346,
+      "learning_rate": 1e-06,
+      "loss": 0.0072,
+      "step": 3754
+    },
+    {
+      "epoch": 0.6392577459993191,
+      "grad_norm": 1.3751659393310547,
+      "learning_rate": 1e-06,
+      "loss": 0.0152,
+      "step": 3755
+    },
+    {
+      "epoch": 0.6394279877425945,
+      "grad_norm": 1.0783185958862305,
+      "learning_rate": 1e-06,
+      "loss": 0.0087,
+      "step": 3756
+    },
+    {
+      "epoch": 0.63959822948587,
+      "grad_norm": 1.002148985862732,
+      "learning_rate": 1e-06,
+      "loss": 0.0065,
+      "step": 3757
+    },
+    {
+      "epoch": 0.6397684712291454,
+      "grad_norm": 0.9721300005912781,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 3758
+    },
+    {
+      "epoch": 0.6399387129724209,
+      "grad_norm": 1.0367679595947266,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3759
+    },
+    {
+      "epoch": 0.6401089547156963,
+      "grad_norm": 1.3424270153045654,
+      "learning_rate": 1e-06,
+      "loss": 0.0119,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6402791964589717,
+      "grad_norm": 1.1471754312515259,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 3761
+    },
+    {
+      "epoch": 0.6404494382022472,
+      "grad_norm": 1.1951358318328857,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 3762
+    },
+    {
+      "epoch": 0.6406196799455226,
+      "grad_norm": 1.077262282371521,
+      "learning_rate": 1e-06,
+      "loss": 0.0077,
+      "step": 3763
+    },
+    {
+      "epoch": 0.6407899216887981,
+      "grad_norm": 2.4920833110809326,
+      "learning_rate": 1e-06,
+      "loss": 0.0159,
+      "step": 3764
+    },
+    {
+      "epoch": 0.6409601634320735,
+      "grad_norm": 1.1345974206924438,
+      "learning_rate": 1e-06,
+      "loss": 0.01,
+      "step": 3765
+    },
+    {
+      "epoch": 0.641130405175349,
+      "grad_norm": 0.8907074928283691,
+      "learning_rate": 1e-06,
+      "loss": 0.007,
+      "step": 3766
+    },
+    {
+      "epoch": 0.6413006469186244,
+      "grad_norm": 1.1611642837524414,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 3767
+    },
+    {
+      "epoch": 0.6414708886618999,
+      "grad_norm": 1.0396606922149658,
+      "learning_rate": 1e-06,
+      "loss": 0.0104,
+      "step": 3768
+    },
+    {
+      "epoch": 0.6416411304051753,
+      "grad_norm": 0.9746201634407043,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3769
+    },
+    {
+      "epoch": 0.6418113721484509,
+      "grad_norm": 0.9750457406044006,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6419816138917263,
+      "grad_norm": 1.3506128787994385,
+      "learning_rate": 1e-06,
+      "loss": 0.0099,
+      "step": 3771
+    },
+    {
+      "epoch": 0.6421518556350017,
+      "grad_norm": 1.9846067428588867,
+      "learning_rate": 1e-06,
+      "loss": 0.0243,
+      "step": 3772
+    },
+    {
+      "epoch": 0.6423220973782772,
+      "grad_norm": 0.9080547094345093,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3773
+    },
+    {
+      "epoch": 0.6424923391215526,
+      "grad_norm": 1.4013069868087769,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 3774
+    },
+    {
+      "epoch": 0.6426625808648281,
+      "grad_norm": 1.1114245653152466,
+      "learning_rate": 1e-06,
+      "loss": 0.0146,
+      "step": 3775
+    },
+    {
+      "epoch": 0.6428328226081035,
+      "grad_norm": 0.9145612120628357,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 3776
+    },
+    {
+      "epoch": 0.643003064351379,
+      "grad_norm": 0.9096139073371887,
+      "learning_rate": 1e-06,
+      "loss": 0.0089,
+      "step": 3777
+    },
+    {
+      "epoch": 0.6431733060946544,
+      "grad_norm": 2.0348570346832275,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3778
+    },
+    {
+      "epoch": 0.6433435478379299,
+      "grad_norm": 1.3212345838546753,
+      "learning_rate": 1e-06,
+      "loss": 0.0129,
+      "step": 3779
+    },
+    {
+      "epoch": 0.6435137895812053,
+      "grad_norm": 0.8319016098976135,
+      "learning_rate": 1e-06,
+      "loss": 0.0072,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6436840313244807,
+      "grad_norm": 0.9304143786430359,
+      "learning_rate": 1e-06,
+      "loss": 0.0067,
+      "step": 3781
+    },
+    {
+      "epoch": 0.6438542730677562,
+      "grad_norm": 1.048282504081726,
+      "learning_rate": 1e-06,
+      "loss": 0.0089,
+      "step": 3782
+    },
+    {
+      "epoch": 0.6440245148110316,
+      "grad_norm": 1.1110150814056396,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 3783
+    },
+    {
+      "epoch": 0.6441947565543071,
+      "grad_norm": 1.329409122467041,
+      "learning_rate": 1e-06,
+      "loss": 0.0162,
+      "step": 3784
+    },
+    {
+      "epoch": 0.6443649982975825,
+      "grad_norm": 0.838369607925415,
+      "learning_rate": 1e-06,
+      "loss": 0.0064,
+      "step": 3785
+    },
+    {
+      "epoch": 0.644535240040858,
+      "grad_norm": 1.109646201133728,
+      "learning_rate": 1e-06,
+      "loss": 0.0122,
+      "step": 3786
+    },
+    {
+      "epoch": 0.6447054817841334,
+      "grad_norm": 1.0034140348434448,
+      "learning_rate": 1e-06,
+      "loss": 0.0072,
+      "step": 3787
+    },
+    {
+      "epoch": 0.644875723527409,
+      "grad_norm": 1.1564267873764038,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3788
+    },
+    {
+      "epoch": 0.6450459652706844,
+      "grad_norm": 1.3622796535491943,
+      "learning_rate": 1e-06,
+      "loss": 0.0152,
+      "step": 3789
+    },
+    {
+      "epoch": 0.6452162070139599,
+      "grad_norm": 0.9166209697723389,
+      "learning_rate": 1e-06,
+      "loss": 0.0066,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6453864487572353,
+      "grad_norm": 1.0101425647735596,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 3791
+    },
+    {
+      "epoch": 0.6455566905005107,
+      "grad_norm": 1.1960997581481934,
+      "learning_rate": 1e-06,
+      "loss": 0.0143,
+      "step": 3792
+    },
+    {
+      "epoch": 0.6457269322437862,
+      "grad_norm": 1.1836118698120117,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 3793
+    },
+    {
+      "epoch": 0.6458971739870616,
+      "grad_norm": 1.131833553314209,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3794
+    },
+    {
+      "epoch": 0.6460674157303371,
+      "grad_norm": 1.1071782112121582,
+      "learning_rate": 1e-06,
+      "loss": 0.0067,
+      "step": 3795
+    },
+    {
+      "epoch": 0.6462376574736125,
+      "grad_norm": 1.1466583013534546,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3796
+    },
+    {
+      "epoch": 0.646407899216888,
+      "grad_norm": 1.2023851871490479,
+      "learning_rate": 1e-06,
+      "loss": 0.0079,
+      "step": 3797
+    },
+    {
+      "epoch": 0.6465781409601634,
+      "grad_norm": 1.4373587369918823,
+      "learning_rate": 1e-06,
+      "loss": 0.0123,
+      "step": 3798
+    },
+    {
+      "epoch": 0.6467483827034389,
+      "grad_norm": 0.9874997138977051,
+      "learning_rate": 1e-06,
+      "loss": 0.007,
+      "step": 3799
+    },
+    {
+      "epoch": 0.6469186244467143,
+      "grad_norm": 1.1962223052978516,
+      "learning_rate": 1e-06,
+      "loss": 0.0114,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6470888661899897,
+      "grad_norm": 1.095088005065918,
+      "learning_rate": 1e-06,
+      "loss": 0.0089,
+      "step": 3801
+    },
+    {
+      "epoch": 0.6472591079332652,
+      "grad_norm": 1.0901947021484375,
+      "learning_rate": 1e-06,
+      "loss": 0.0074,
+      "step": 3802
+    },
+    {
+      "epoch": 0.6474293496765406,
+      "grad_norm": 1.262510895729065,
+      "learning_rate": 1e-06,
+      "loss": 0.0067,
+      "step": 3803
+    },
+    {
+      "epoch": 0.6475995914198162,
+      "grad_norm": 1.2522022724151611,
+      "learning_rate": 1e-06,
+      "loss": 0.0106,
+      "step": 3804
+    },
+    {
+      "epoch": 0.6477698331630916,
+      "grad_norm": 1.0772358179092407,
+      "learning_rate": 1e-06,
+      "loss": 0.0074,
+      "step": 3805
+    },
+    {
+      "epoch": 0.6479400749063671,
+      "grad_norm": 1.433408498764038,
+      "learning_rate": 1e-06,
+      "loss": 0.011,
+      "step": 3806
+    },
+    {
+      "epoch": 0.6481103166496425,
+      "grad_norm": 1.124001383781433,
+      "learning_rate": 1e-06,
+      "loss": 0.0076,
+      "step": 3807
+    },
+    {
+      "epoch": 0.648280558392918,
+      "grad_norm": 1.0808985233306885,
+      "learning_rate": 1e-06,
+      "loss": 0.0074,
+      "step": 3808
+    },
+    {
+      "epoch": 0.6484508001361934,
+      "grad_norm": 1.0477313995361328,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3809
+    },
+    {
+      "epoch": 0.6486210418794689,
+      "grad_norm": 2.4625258445739746,
+      "learning_rate": 1e-06,
+      "loss": 0.0164,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6487912836227443,
+      "grad_norm": 0.9473697543144226,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3811
+    },
+    {
+      "epoch": 0.6489615253660197,
+      "grad_norm": 0.9401687383651733,
+      "learning_rate": 1e-06,
+      "loss": 0.0076,
+      "step": 3812
+    },
+    {
+      "epoch": 0.6491317671092952,
+      "grad_norm": 1.2510778903961182,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 3813
+    },
+    {
+      "epoch": 0.6493020088525706,
+      "grad_norm": 0.9812240600585938,
+      "learning_rate": 1e-06,
+      "loss": 0.01,
+      "step": 3814
+    },
+    {
+      "epoch": 0.6494722505958461,
+      "grad_norm": 0.99561607837677,
+      "learning_rate": 1e-06,
+      "loss": 0.0077,
+      "step": 3815
+    },
+    {
+      "epoch": 0.6496424923391215,
+      "grad_norm": 1.579348087310791,
+      "learning_rate": 1e-06,
+      "loss": 0.024,
+      "step": 3816
+    },
+    {
+      "epoch": 0.649812734082397,
+      "grad_norm": 0.7100584506988525,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3817
+    },
+    {
+      "epoch": 0.6499829758256724,
+      "grad_norm": 1.0388001203536987,
+      "learning_rate": 1e-06,
+      "loss": 0.0071,
+      "step": 3818
+    },
+    {
+      "epoch": 0.650153217568948,
+      "grad_norm": 1.2186797857284546,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3819
+    },
+    {
+      "epoch": 0.6503234593122234,
+      "grad_norm": 0.8475635647773743,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6504937010554989,
+      "grad_norm": 1.1312545537948608,
+      "learning_rate": 1e-06,
+      "loss": 0.0067,
+      "step": 3821
+    },
+    {
+      "epoch": 0.6506639427987743,
+      "grad_norm": 1.2596800327301025,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 3822
+    },
+    {
+      "epoch": 0.6508341845420497,
+      "grad_norm": 0.8742133378982544,
+      "learning_rate": 1e-06,
+      "loss": 0.01,
+      "step": 3823
+    },
+    {
+      "epoch": 0.6510044262853252,
+      "grad_norm": 0.891171395778656,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 3824
+    },
+    {
+      "epoch": 0.6511746680286006,
+      "grad_norm": 1.0801339149475098,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 3825
+    },
+    {
+      "epoch": 0.6513449097718761,
+      "grad_norm": 0.9043775200843811,
+      "learning_rate": 1e-06,
+      "loss": 0.0094,
+      "step": 3826
+    },
+    {
+      "epoch": 0.6515151515151515,
+      "grad_norm": 1.2618519067764282,
+      "learning_rate": 1e-06,
+      "loss": 0.009,
+      "step": 3827
+    },
+    {
+      "epoch": 0.651685393258427,
+      "grad_norm": 0.9808958172798157,
+      "learning_rate": 1e-06,
+      "loss": 0.0068,
+      "step": 3828
+    },
+    {
+      "epoch": 0.6518556350017024,
+      "grad_norm": 0.9210644364356995,
+      "learning_rate": 1e-06,
+      "loss": 0.0059,
+      "step": 3829
+    },
+    {
+      "epoch": 0.6520258767449779,
+      "grad_norm": 1.0287270545959473,
+      "learning_rate": 1e-06,
+      "loss": 0.0109,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6521961184882533,
+      "grad_norm": 0.9316836595535278,
+      "learning_rate": 1e-06,
+      "loss": 0.0077,
+      "step": 3831
+    },
+    {
+      "epoch": 0.6523663602315287,
+      "grad_norm": 0.9281513094902039,
+      "learning_rate": 1e-06,
+      "loss": 0.0077,
+      "step": 3832
+    },
+    {
+      "epoch": 0.6525366019748042,
+      "grad_norm": 1.0260971784591675,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3833
+    },
+    {
+      "epoch": 0.6527068437180796,
+      "grad_norm": 1.012954831123352,
+      "learning_rate": 1e-06,
+      "loss": 0.0099,
+      "step": 3834
+    },
+    {
+      "epoch": 0.6528770854613551,
+      "grad_norm": 1.0251330137252808,
+      "learning_rate": 1e-06,
+      "loss": 0.0068,
+      "step": 3835
+    },
+    {
+      "epoch": 0.6530473272046305,
+      "grad_norm": 1.228165864944458,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3836
+    },
+    {
+      "epoch": 0.6532175689479061,
+      "grad_norm": 0.8660666346549988,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3837
+    },
+    {
+      "epoch": 0.6533878106911815,
+      "grad_norm": 0.8670229911804199,
+      "learning_rate": 1e-06,
+      "loss": 0.0059,
+      "step": 3838
+    },
+    {
+      "epoch": 0.653558052434457,
+      "grad_norm": 1.1961077451705933,
+      "learning_rate": 1e-06,
+      "loss": 0.0134,
+      "step": 3839
+    },
+    {
+      "epoch": 0.6537282941777324,
+      "grad_norm": 0.9217979311943054,
+      "learning_rate": 1e-06,
+      "loss": 0.0065,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6538985359210079,
+      "grad_norm": 0.9868664741516113,
+      "learning_rate": 1e-06,
+      "loss": 0.0073,
+      "step": 3841
+    },
+    {
+      "epoch": 0.6540687776642833,
+      "grad_norm": 1.3531166315078735,
+      "learning_rate": 1e-06,
+      "loss": 0.0118,
+      "step": 3842
+    },
+    {
+      "epoch": 0.6542390194075587,
+      "grad_norm": 1.6801015138626099,
+      "learning_rate": 1e-06,
+      "loss": 0.012,
+      "step": 3843
+    },
+    {
+      "epoch": 0.6544092611508342,
+      "grad_norm": 1.0095585584640503,
+      "learning_rate": 1e-06,
+      "loss": 0.0065,
+      "step": 3844
+    },
+    {
+      "epoch": 0.6545795028941096,
+      "grad_norm": 0.8943735361099243,
+      "learning_rate": 1e-06,
+      "loss": 0.006,
+      "step": 3845
+    },
+    {
+      "epoch": 0.6547497446373851,
+      "grad_norm": 1.3049546480178833,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 3846
+    },
+    {
+      "epoch": 0.6549199863806605,
+      "grad_norm": 1.1015139818191528,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3847
+    },
+    {
+      "epoch": 0.655090228123936,
+      "grad_norm": 1.2883460521697998,
+      "learning_rate": 1e-06,
+      "loss": 0.0076,
+      "step": 3848
+    },
+    {
+      "epoch": 0.6552604698672114,
+      "grad_norm": 0.9433921575546265,
+      "learning_rate": 1e-06,
+      "loss": 0.007,
+      "step": 3849
+    },
+    {
+      "epoch": 0.6554307116104869,
+      "grad_norm": 1.1686891317367554,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6556009533537623,
+      "grad_norm": 1.2734119892120361,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 3851
+    },
+    {
+      "epoch": 0.6557711950970377,
+      "grad_norm": 1.1788426637649536,
+      "learning_rate": 1e-06,
+      "loss": 0.007,
+      "step": 3852
+    },
+    {
+      "epoch": 0.6559414368403133,
+      "grad_norm": 1.1873812675476074,
+      "learning_rate": 1e-06,
+      "loss": 0.0148,
+      "step": 3853
+    },
+    {
+      "epoch": 0.6561116785835887,
+      "grad_norm": 1.2980464696884155,
+      "learning_rate": 1e-06,
+      "loss": 0.0079,
+      "step": 3854
+    },
+    {
+      "epoch": 0.6562819203268642,
+      "grad_norm": 1.1819038391113281,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 3855
+    },
+    {
+      "epoch": 0.6564521620701396,
+      "grad_norm": 1.5700839757919312,
+      "learning_rate": 1e-06,
+      "loss": 0.0128,
+      "step": 3856
+    },
+    {
+      "epoch": 0.6566224038134151,
+      "grad_norm": 0.907390296459198,
+      "learning_rate": 1e-06,
+      "loss": 0.0081,
+      "step": 3857
+    },
+    {
+      "epoch": 0.6567926455566905,
+      "grad_norm": 1.2444868087768555,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 3858
+    },
+    {
+      "epoch": 0.656962887299966,
+      "grad_norm": 0.8420805931091309,
+      "learning_rate": 1e-06,
+      "loss": 0.007,
+      "step": 3859
+    },
+    {
+      "epoch": 0.6571331290432414,
+      "grad_norm": 0.9762222170829773,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6573033707865169,
+      "grad_norm": 1.3171446323394775,
+      "learning_rate": 1e-06,
+      "loss": 0.0071,
+      "step": 3861
+    },
+    {
+      "epoch": 0.6574736125297923,
+      "grad_norm": 0.9845103025436401,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 3862
+    },
+    {
+      "epoch": 0.6576438542730677,
+      "grad_norm": 1.1617891788482666,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3863
+    },
+    {
+      "epoch": 0.6578140960163432,
+      "grad_norm": 0.9630821943283081,
+      "learning_rate": 1e-06,
+      "loss": 0.0073,
+      "step": 3864
+    },
+    {
+      "epoch": 0.6579843377596186,
+      "grad_norm": 1.274949312210083,
+      "learning_rate": 1e-06,
+      "loss": 0.0079,
+      "step": 3865
+    },
+    {
+      "epoch": 0.6581545795028941,
+      "grad_norm": 1.40157151222229,
+      "learning_rate": 1e-06,
+      "loss": 0.0071,
+      "step": 3866
+    },
+    {
+      "epoch": 0.6583248212461695,
+      "grad_norm": 1.0424250364303589,
+      "learning_rate": 1e-06,
+      "loss": 0.0094,
+      "step": 3867
+    },
+    {
+      "epoch": 0.658495062989445,
+      "grad_norm": 1.3154240846633911,
+      "learning_rate": 1e-06,
+      "loss": 0.0126,
+      "step": 3868
+    },
+    {
+      "epoch": 0.6586653047327204,
+      "grad_norm": 1.1080321073532104,
+      "learning_rate": 1e-06,
+      "loss": 0.0082,
+      "step": 3869
+    },
+    {
+      "epoch": 0.658835546475996,
+      "grad_norm": 1.0510450601577759,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6590057882192714,
+      "grad_norm": 1.3665117025375366,
+      "learning_rate": 1e-06,
+      "loss": 0.0115,
+      "step": 3871
+    },
+    {
+      "epoch": 0.6591760299625468,
+      "grad_norm": 0.914477527141571,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3872
+    },
+    {
+      "epoch": 0.6593462717058223,
+      "grad_norm": 1.2469263076782227,
+      "learning_rate": 1e-06,
+      "loss": 0.0094,
+      "step": 3873
+    },
+    {
+      "epoch": 0.6595165134490977,
+      "grad_norm": 1.1164896488189697,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3874
+    },
+    {
+      "epoch": 0.6596867551923732,
+      "grad_norm": 1.092367172241211,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3875
+    },
+    {
+      "epoch": 0.6598569969356486,
+      "grad_norm": 1.0838600397109985,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 3876
+    },
+    {
+      "epoch": 0.6600272386789241,
+      "grad_norm": 1.1364141702651978,
+      "learning_rate": 1e-06,
+      "loss": 0.0063,
+      "step": 3877
+    },
+    {
+      "epoch": 0.6601974804221995,
+      "grad_norm": 1.0023877620697021,
+      "learning_rate": 1e-06,
+      "loss": 0.0073,
+      "step": 3878
+    },
+    {
+      "epoch": 0.660367722165475,
+      "grad_norm": 0.9350517392158508,
+      "learning_rate": 1e-06,
+      "loss": 0.0076,
+      "step": 3879
+    },
+    {
+      "epoch": 0.6605379639087504,
+      "grad_norm": 0.807680606842041,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6607082056520259,
+      "grad_norm": 0.7163336277008057,
+      "learning_rate": 1e-06,
+      "loss": 0.0057,
+      "step": 3881
+    },
+    {
+      "epoch": 0.6608784473953013,
+      "grad_norm": 1.0191099643707275,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 3882
+    },
+    {
+      "epoch": 0.6610486891385767,
+      "grad_norm": 0.8287161588668823,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 3883
+    },
+    {
+      "epoch": 0.6612189308818522,
+      "grad_norm": 1.2603466510772705,
+      "learning_rate": 1e-06,
+      "loss": 0.0077,
+      "step": 3884
+    },
+    {
+      "epoch": 0.6613891726251276,
+      "grad_norm": 1.042250633239746,
+      "learning_rate": 1e-06,
+      "loss": 0.0123,
+      "step": 3885
+    },
+    {
+      "epoch": 0.6615594143684032,
+      "grad_norm": 1.064152479171753,
+      "learning_rate": 1e-06,
+      "loss": 0.011,
+      "step": 3886
+    },
+    {
+      "epoch": 0.6617296561116786,
+      "grad_norm": 1.2731701135635376,
+      "learning_rate": 1e-06,
+      "loss": 0.0217,
+      "step": 3887
+    },
+    {
+      "epoch": 0.6618998978549541,
+      "grad_norm": 1.1830432415008545,
+      "learning_rate": 1e-06,
+      "loss": 0.0084,
+      "step": 3888
+    },
+    {
+      "epoch": 0.6620701395982295,
+      "grad_norm": 1.008324384689331,
+      "learning_rate": 1e-06,
+      "loss": 0.0064,
+      "step": 3889
+    },
+    {
+      "epoch": 0.662240381341505,
+      "grad_norm": 0.9876424074172974,
+      "learning_rate": 1e-06,
+      "loss": 0.0075,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6624106230847804,
+      "grad_norm": 1.2349107265472412,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 3891
+    },
+    {
+      "epoch": 0.6625808648280559,
+      "grad_norm": 1.1617462635040283,
+      "learning_rate": 1e-06,
+      "loss": 0.0142,
+      "step": 3892
+    },
+    {
+      "epoch": 0.6627511065713313,
+      "grad_norm": 0.8870407938957214,
+      "learning_rate": 1e-06,
+      "loss": 0.0073,
+      "step": 3893
+    },
+    {
+      "epoch": 0.6629213483146067,
+      "grad_norm": 1.0737135410308838,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3894
+    },
+    {
+      "epoch": 0.6630915900578822,
+      "grad_norm": 1.0294846296310425,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 3895
+    },
+    {
+      "epoch": 0.6632618318011576,
+      "grad_norm": 0.8994992971420288,
+      "learning_rate": 1e-06,
+      "loss": 0.0068,
+      "step": 3896
+    },
+    {
+      "epoch": 0.6634320735444331,
+      "grad_norm": 1.0250681638717651,
+      "learning_rate": 1e-06,
+      "loss": 0.0105,
+      "step": 3897
+    },
+    {
+      "epoch": 0.6636023152877085,
+      "grad_norm": 1.2135937213897705,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 3898
+    },
+    {
+      "epoch": 0.663772557030984,
+      "grad_norm": 0.907864511013031,
+      "learning_rate": 1e-06,
+      "loss": 0.0098,
+      "step": 3899
+    },
+    {
+      "epoch": 0.6639427987742594,
+      "grad_norm": 1.0360740423202515,
+      "learning_rate": 1e-06,
+      "loss": 0.0065,
+      "step": 3900
+    },
+    {
+      "epoch": 0.664113040517535,
+      "grad_norm": 1.1412363052368164,
+      "learning_rate": 1e-06,
+      "loss": 0.0126,
+      "step": 3901
+    },
+    {
+      "epoch": 0.6642832822608103,
+      "grad_norm": 1.2569540739059448,
+      "learning_rate": 1e-06,
+      "loss": 0.0121,
+      "step": 3902
+    },
+    {
+      "epoch": 0.6644535240040857,
+      "grad_norm": 1.2806369066238403,
+      "learning_rate": 1e-06,
+      "loss": 0.0082,
+      "step": 3903
+    },
+    {
+      "epoch": 0.6646237657473613,
+      "grad_norm": 1.199313759803772,
+      "learning_rate": 1e-06,
+      "loss": 0.0137,
+      "step": 3904
+    },
+    {
+      "epoch": 0.6647940074906367,
+      "grad_norm": 1.0673506259918213,
+      "learning_rate": 1e-06,
+      "loss": 0.0132,
+      "step": 3905
+    },
+    {
+      "epoch": 0.6649642492339122,
+      "grad_norm": 1.1850353479385376,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 3906
+    },
+    {
+      "epoch": 0.6651344909771876,
+      "grad_norm": 1.9815788269042969,
+      "learning_rate": 1e-06,
+      "loss": 0.0108,
+      "step": 3907
+    },
+    {
+      "epoch": 0.6653047327204631,
+      "grad_norm": 1.181032657623291,
+      "learning_rate": 1e-06,
+      "loss": 0.007,
+      "step": 3908
+    },
+    {
+      "epoch": 0.6654749744637385,
+      "grad_norm": 1.2337028980255127,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3909
+    },
+    {
+      "epoch": 0.665645216207014,
+      "grad_norm": 1.0751506090164185,
+      "learning_rate": 1e-06,
+      "loss": 0.0065,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6658154579502894,
+      "grad_norm": 1.4897148609161377,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3911
+    },
+    {
+      "epoch": 0.6659856996935649,
+      "grad_norm": 1.2287198305130005,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 3912
+    },
+    {
+      "epoch": 0.6661559414368403,
+      "grad_norm": 1.3234024047851562,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3913
+    },
+    {
+      "epoch": 0.6663261831801157,
+      "grad_norm": 1.0900822877883911,
+      "learning_rate": 1e-06,
+      "loss": 0.0114,
+      "step": 3914
+    },
+    {
+      "epoch": 0.6664964249233912,
+      "grad_norm": 0.9296737313270569,
+      "learning_rate": 1e-06,
+      "loss": 0.0063,
+      "step": 3915
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 1.1464951038360596,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 3916
+    },
+    {
+      "epoch": 0.6668369084099421,
+      "grad_norm": 1.5186680555343628,
+      "learning_rate": 1e-06,
+      "loss": 0.013,
+      "step": 3917
+    },
+    {
+      "epoch": 0.6670071501532175,
+      "grad_norm": 1.0801197290420532,
+      "learning_rate": 1e-06,
+      "loss": 0.0095,
+      "step": 3918
+    },
+    {
+      "epoch": 0.667177391896493,
+      "grad_norm": 1.058134913444519,
+      "learning_rate": 1e-06,
+      "loss": 0.0066,
+      "step": 3919
+    },
+    {
+      "epoch": 0.6673476336397685,
+      "grad_norm": 0.9633292555809021,
+      "learning_rate": 1e-06,
+      "loss": 0.0066,
+      "step": 3920
+    },
+    {
+      "epoch": 0.667517875383044,
+      "grad_norm": 1.1550352573394775,
+      "learning_rate": 1e-06,
+      "loss": 0.0091,
+      "step": 3921
+    },
+    {
+      "epoch": 0.6676881171263194,
+      "grad_norm": 1.437263011932373,
+      "learning_rate": 1e-06,
+      "loss": 0.0086,
+      "step": 3922
+    },
+    {
+      "epoch": 0.6678583588695948,
+      "grad_norm": 0.9223193526268005,
+      "learning_rate": 1e-06,
+      "loss": 0.0059,
+      "step": 3923
+    },
+    {
+      "epoch": 0.6680286006128703,
+      "grad_norm": 0.9309805631637573,
+      "learning_rate": 1e-06,
+      "loss": 0.0076,
+      "step": 3924
+    },
+    {
+      "epoch": 0.6681988423561457,
+      "grad_norm": 1.1240042448043823,
+      "learning_rate": 1e-06,
+      "loss": 0.0081,
+      "step": 3925
+    },
+    {
+      "epoch": 0.6683690840994212,
+      "grad_norm": 1.074847936630249,
+      "learning_rate": 1e-06,
+      "loss": 0.0116,
+      "step": 3926
+    },
+    {
+      "epoch": 0.6685393258426966,
+      "grad_norm": 0.9684469103813171,
+      "learning_rate": 1e-06,
+      "loss": 0.0073,
+      "step": 3927
+    },
+    {
+      "epoch": 0.6687095675859721,
+      "grad_norm": 1.2416397333145142,
+      "learning_rate": 1e-06,
+      "loss": 0.0099,
+      "step": 3928
+    },
+    {
+      "epoch": 0.6688798093292475,
+      "grad_norm": 0.9240944981575012,
+      "learning_rate": 1e-06,
+      "loss": 0.0067,
+      "step": 3929
+    },
+    {
+      "epoch": 0.669050051072523,
+      "grad_norm": 0.9790809154510498,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6692202928157984,
+      "grad_norm": 0.9496670961380005,
+      "learning_rate": 1e-06,
+      "loss": 0.0069,
+      "step": 3931
+    },
+    {
+      "epoch": 0.6693905345590739,
+      "grad_norm": 0.8220587968826294,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3932
+    },
+    {
+      "epoch": 0.6695607763023493,
+      "grad_norm": 1.191679835319519,
+      "learning_rate": 1e-06,
+      "loss": 0.0107,
+      "step": 3933
+    },
+    {
+      "epoch": 0.6697310180456247,
+      "grad_norm": 1.0436556339263916,
+      "learning_rate": 1e-06,
+      "loss": 0.0058,
+      "step": 3934
+    },
+    {
+      "epoch": 0.6699012597889002,
+      "grad_norm": 1.3039947748184204,
+      "learning_rate": 1e-06,
+      "loss": 0.0082,
+      "step": 3935
+    },
+    {
+      "epoch": 0.6700715015321757,
+      "grad_norm": 1.2773553133010864,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3936
+    },
+    {
+      "epoch": 0.6702417432754512,
+      "grad_norm": 1.6559447050094604,
+      "learning_rate": 1e-06,
+      "loss": 0.0166,
+      "step": 3937
+    },
+    {
+      "epoch": 0.6704119850187266,
+      "grad_norm": 1.3342622518539429,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 3938
+    },
+    {
+      "epoch": 0.6705822267620021,
+      "grad_norm": 1.0363655090332031,
+      "learning_rate": 1e-06,
+      "loss": 0.0063,
+      "step": 3939
+    },
+    {
+      "epoch": 0.6707524685052775,
+      "grad_norm": 1.0175873041152954,
+      "learning_rate": 1e-06,
+      "loss": 0.0106,
+      "step": 3940
+    },
+    {
+      "epoch": 0.670922710248553,
+      "grad_norm": 0.8500238656997681,
+      "learning_rate": 1e-06,
+      "loss": 0.0099,
+      "step": 3941
+    },
+    {
+      "epoch": 0.6710929519918284,
+      "grad_norm": 4.019010543823242,
+      "learning_rate": 1e-06,
+      "loss": 0.0548,
+      "step": 3942
+    },
+    {
+      "epoch": 0.6712631937351039,
+      "grad_norm": 0.9771483540534973,
+      "learning_rate": 1e-06,
+      "loss": 0.006,
+      "step": 3943
+    },
+    {
+      "epoch": 0.6714334354783793,
+      "grad_norm": 1.0605002641677856,
+      "learning_rate": 1e-06,
+      "loss": 0.0065,
+      "step": 3944
+    },
+    {
+      "epoch": 0.6716036772216547,
+      "grad_norm": 1.0014209747314453,
+      "learning_rate": 1e-06,
+      "loss": 0.0066,
+      "step": 3945
+    },
+    {
+      "epoch": 0.6717739189649302,
+      "grad_norm": 1.0529136657714844,
+      "learning_rate": 1e-06,
+      "loss": 0.0082,
+      "step": 3946
+    },
+    {
+      "epoch": 0.6719441607082056,
+      "grad_norm": 1.1662652492523193,
+      "learning_rate": 1e-06,
+      "loss": 0.0168,
+      "step": 3947
+    },
+    {
+      "epoch": 0.6721144024514811,
+      "grad_norm": 1.425837755203247,
+      "learning_rate": 1e-06,
+      "loss": 0.0101,
+      "step": 3948
+    },
+    {
+      "epoch": 0.6722846441947565,
+      "grad_norm": 0.8996866345405579,
+      "learning_rate": 1e-06,
+      "loss": 0.0063,
+      "step": 3949
+    },
+    {
+      "epoch": 0.672454885938032,
+      "grad_norm": 0.9349283576011658,
+      "learning_rate": 1e-06,
+      "loss": 0.007,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6726251276813074,
+      "grad_norm": 0.9103063344955444,
+      "learning_rate": 1e-06,
+      "loss": 0.0073,
+      "step": 3951
+    },
+    {
+      "epoch": 0.672795369424583,
+      "grad_norm": 1.297703504562378,
+      "learning_rate": 1e-06,
+      "loss": 0.016,
+      "step": 3952
+    },
+    {
+      "epoch": 0.6729656111678584,
+      "grad_norm": 1.0639750957489014,
+      "learning_rate": 1e-06,
+      "loss": 0.0079,
+      "step": 3953
+    },
+    {
+      "epoch": 0.6731358529111338,
+      "grad_norm": 1.0544763803482056,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 3954
+    },
+    {
+      "epoch": 0.6733060946544093,
+      "grad_norm": 1.192960500717163,
+      "learning_rate": 1e-06,
+      "loss": 0.0064,
+      "step": 3955
+    },
+    {
+      "epoch": 0.6734763363976847,
+      "grad_norm": 0.8247693777084351,
+      "learning_rate": 1e-06,
+      "loss": 0.0074,
+      "step": 3956
+    },
+    {
+      "epoch": 0.6736465781409602,
+      "grad_norm": 0.7883464694023132,
+      "learning_rate": 1e-06,
+      "loss": 0.0067,
+      "step": 3957
+    },
+    {
+      "epoch": 0.6738168198842356,
+      "grad_norm": 0.9810028672218323,
+      "learning_rate": 1e-06,
+      "loss": 0.0069,
+      "step": 3958
+    },
+    {
+      "epoch": 0.6739870616275111,
+      "grad_norm": 0.9418751001358032,
+      "learning_rate": 1e-06,
+      "loss": 0.0064,
+      "step": 3959
+    },
+    {
+      "epoch": 0.6741573033707865,
+      "grad_norm": 1.339374303817749,
+      "learning_rate": 1e-06,
+      "loss": 0.0092,
+      "step": 3960
+    },
+    {
+      "epoch": 0.674327545114062,
+      "grad_norm": 1.082369089126587,
+      "learning_rate": 1e-06,
+      "loss": 0.0089,
+      "step": 3961
+    },
+    {
+      "epoch": 0.6744977868573374,
+      "grad_norm": 1.2776871919631958,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3962
+    },
+    {
+      "epoch": 0.6746680286006129,
+      "grad_norm": 1.1005768775939941,
+      "learning_rate": 1e-06,
+      "loss": 0.0081,
+      "step": 3963
+    },
+    {
+      "epoch": 0.6748382703438883,
+      "grad_norm": 1.0241467952728271,
+      "learning_rate": 1e-06,
+      "loss": 0.0056,
+      "step": 3964
+    },
+    {
+      "epoch": 0.6750085120871637,
+      "grad_norm": 0.992150068283081,
+      "learning_rate": 1e-06,
+      "loss": 0.0069,
+      "step": 3965
+    },
+    {
+      "epoch": 0.6751787538304392,
+      "grad_norm": 0.8683270215988159,
+      "learning_rate": 1e-06,
+      "loss": 0.0072,
+      "step": 3966
+    },
+    {
+      "epoch": 0.6753489955737146,
+      "grad_norm": 1.0942341089248657,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 3967
+    },
+    {
+      "epoch": 0.6755192373169902,
+      "grad_norm": 1.0581514835357666,
+      "learning_rate": 1e-06,
+      "loss": 0.0063,
+      "step": 3968
+    },
+    {
+      "epoch": 0.6756894790602656,
+      "grad_norm": 1.0217764377593994,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 3969
+    },
+    {
+      "epoch": 0.6758597208035411,
+      "grad_norm": 1.3109872341156006,
+      "learning_rate": 1e-06,
+      "loss": 0.0068,
+      "step": 3970
+    },
+    {
+      "epoch": 0.6760299625468165,
+      "grad_norm": 0.9485663771629333,
+      "learning_rate": 1e-06,
+      "loss": 0.0077,
+      "step": 3971
+    },
+    {
+      "epoch": 0.676200204290092,
+      "grad_norm": 0.8911463618278503,
+      "learning_rate": 1e-06,
+      "loss": 0.0069,
+      "step": 3972
+    },
+    {
+      "epoch": 0.6763704460333674,
+      "grad_norm": 0.908300518989563,
+      "learning_rate": 1e-06,
+      "loss": 0.0073,
+      "step": 3973
+    },
+    {
+      "epoch": 0.6765406877766428,
+      "grad_norm": 0.9515292644500732,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 3974
+    },
+    {
+      "epoch": 0.6767109295199183,
+      "grad_norm": 1.1426764726638794,
+      "learning_rate": 1e-06,
+      "loss": 0.0085,
+      "step": 3975
+    },
+    {
+      "epoch": 0.6768811712631937,
+      "grad_norm": 1.1612918376922607,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 3976
+    },
+    {
+      "epoch": 0.6770514130064692,
+      "grad_norm": 0.757738471031189,
+      "learning_rate": 1e-06,
+      "loss": 0.0056,
+      "step": 3977
+    },
+    {
+      "epoch": 0.6772216547497446,
+      "grad_norm": 1.1565226316452026,
+      "learning_rate": 1e-06,
+      "loss": 0.0062,
+      "step": 3978
+    },
+    {
+      "epoch": 0.6773918964930201,
+      "grad_norm": 1.3450309038162231,
+      "learning_rate": 1e-06,
+      "loss": 0.0113,
+      "step": 3979
+    },
+    {
+      "epoch": 0.6775621382362955,
+      "grad_norm": 1.1107460260391235,
+      "learning_rate": 1e-06,
+      "loss": 0.0102,
+      "step": 3980
+    },
+    {
+      "epoch": 0.677732379979571,
+      "grad_norm": 1.5839945077896118,
+      "learning_rate": 1e-06,
+      "loss": 0.0118,
+      "step": 3981
+    },
+    {
+      "epoch": 0.6779026217228464,
+      "grad_norm": 1.143868088722229,
+      "learning_rate": 1e-06,
+      "loss": 0.0079,
+      "step": 3982
+    },
+    {
+      "epoch": 0.678072863466122,
+      "grad_norm": 1.0564833879470825,
+      "learning_rate": 1e-06,
+      "loss": 0.0064,
+      "step": 3983
+    },
+    {
+      "epoch": 0.6782431052093973,
+      "grad_norm": 1.1876205205917358,
+      "learning_rate": 1e-06,
+      "loss": 0.0073,
+      "step": 3984
+    },
+    {
+      "epoch": 0.6784133469526727,
+      "grad_norm": 1.2426191568374634,
+      "learning_rate": 1e-06,
+      "loss": 0.0135,
+      "step": 3985
+    },
+    {
+      "epoch": 0.6785835886959483,
+      "grad_norm": 1.0779424905776978,
+      "learning_rate": 1e-06,
+      "loss": 0.0081,
+      "step": 3986
+    },
+    {
+      "epoch": 0.6787538304392237,
+      "grad_norm": 1.3028185367584229,
+      "learning_rate": 1e-06,
+      "loss": 0.0068,
+      "step": 3987
+    },
+    {
+      "epoch": 0.6789240721824992,
+      "grad_norm": 1.04373300075531,
+      "learning_rate": 1e-06,
+      "loss": 0.0089,
+      "step": 3988
+    },
+    {
+      "epoch": 0.6790943139257746,
+      "grad_norm": 0.9343636631965637,
+      "learning_rate": 1e-06,
+      "loss": 0.0084,
+      "step": 3989
+    },
+    {
+      "epoch": 0.6792645556690501,
+      "grad_norm": 1.0329524278640747,
+      "learning_rate": 1e-06,
+      "loss": 0.0081,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6794347974123255,
+      "grad_norm": 0.9664596319198608,
+      "learning_rate": 1e-06,
+      "loss": 0.0077,
+      "step": 3991
+    },
+    {
+      "epoch": 0.679605039155601,
+      "grad_norm": 0.9661188721656799,
+      "learning_rate": 1e-06,
+      "loss": 0.0071,
+      "step": 3992
+    },
+    {
+      "epoch": 0.6797752808988764,
+      "grad_norm": 1.0348271131515503,
+      "learning_rate": 1e-06,
+      "loss": 0.008,
+      "step": 3993
+    },
+    {
+      "epoch": 0.6799455226421518,
+      "grad_norm": 0.9838294386863708,
+      "learning_rate": 1e-06,
+      "loss": 0.0056,
+      "step": 3994
+    },
+    {
+      "epoch": 0.6801157643854273,
+      "grad_norm": 1.4004563093185425,
+      "learning_rate": 1e-06,
+      "loss": 0.01,
+      "step": 3995
+    },
+    {
+      "epoch": 0.6802860061287027,
+      "grad_norm": 0.8205869793891907,
+      "learning_rate": 1e-06,
+      "loss": 0.0053,
+      "step": 3996
+    },
+    {
+      "epoch": 0.6804562478719782,
+      "grad_norm": 0.8853178024291992,
+      "learning_rate": 1e-06,
+      "loss": 0.007,
+      "step": 3997
+    },
+    {
+      "epoch": 0.6806264896152536,
+      "grad_norm": 0.8219318985939026,
+      "learning_rate": 1e-06,
+      "loss": 0.0076,
+      "step": 3998
+    },
+    {
+      "epoch": 0.6807967313585291,
+      "grad_norm": 1.133307695388794,
+      "learning_rate": 1e-06,
+      "loss": 0.0068,
+      "step": 3999
+    },
+    {
+      "epoch": 0.6809669731018045,
+      "grad_norm": 0.9907300472259521,
+      "learning_rate": 1e-06,
+      "loss": 0.0083,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6809669731018045,
+      "eval_loss": 0.34189823269844055,
+      "eval_runtime": 23.8668,
+      "eval_samples_per_second": 12.57,
+      "eval_steps_per_second": 0.335,
+      "step": 4000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 17622,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 2.3577906965334983e+19,
+  "train_batch_size": 5,
+  "trial_name": null,
+  "trial_params": null
+}