diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,11779 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 1678,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005959475566150178,
+      "grad_norm": Infinity,
+      "learning_rate": 2e-05,
+      "loss": 13.3308,
+      "step": 1
+    },
+    {
+      "epoch": 0.0011918951132300357,
+      "grad_norm": Infinity,
+      "learning_rate": 2e-05,
+      "loss": 13.3851,
+      "step": 2
+    },
+    {
+      "epoch": 0.0017878426698450535,
+      "grad_norm": Infinity,
+      "learning_rate": 2e-05,
+      "loss": 13.2555,
+      "step": 3
+    },
+    {
+      "epoch": 0.0023837902264600714,
+      "grad_norm": Infinity,
+      "learning_rate": 2e-05,
+      "loss": 12.794,
+      "step": 4
+    },
+    {
+      "epoch": 0.0029797377830750892,
+      "grad_norm": Infinity,
+      "learning_rate": 2e-05,
+      "loss": 12.693,
+      "step": 5
+    },
+    {
+      "epoch": 0.003575685339690107,
+      "grad_norm": 184.77955627441406,
+      "learning_rate": 1.99880810488677e-05,
+      "loss": 13.384,
+      "step": 6
+    },
+    {
+      "epoch": 0.004171632896305125,
+      "grad_norm": Infinity,
+      "learning_rate": 1.99880810488677e-05,
+      "loss": 7.6746,
+      "step": 7
+    },
+    {
+      "epoch": 0.004767580452920143,
+      "grad_norm": 137.8367156982422,
+      "learning_rate": 1.99761620977354e-05,
+      "loss": 7.4832,
+      "step": 8
+    },
+    {
+      "epoch": 0.005363528009535161,
+      "grad_norm": 129.59103393554688,
+      "learning_rate": 1.99642431466031e-05,
+      "loss": 6.2899,
+      "step": 9
+    },
+    {
+      "epoch": 0.0059594755661501785,
+      "grad_norm": 135.12991333007812,
+      "learning_rate": 1.99523241954708e-05,
+      "loss": 5.787,
+      "step": 10
+    },
+    {
+      "epoch": 0.006555423122765197,
+      "grad_norm": NaN,
+      "learning_rate": 1.99523241954708e-05,
+      "loss": 5.6186,
+      "step": 11
+    },
+    {
+      "epoch": 0.007151370679380214,
+      "grad_norm": 146.142822265625,
+      "learning_rate": 1.99404052443385e-05,
+      "loss": 5.4871,
+      "step": 12
+    },
+    {
+      "epoch": 0.0077473182359952325,
+      "grad_norm": 140.1932373046875,
+      "learning_rate": 1.99284862932062e-05,
+      "loss": 5.2155,
+      "step": 13
+    },
+    {
+      "epoch": 0.00834326579261025,
+      "grad_norm": 141.83511352539062,
+      "learning_rate": 1.9916567342073898e-05,
+      "loss": 4.9468,
+      "step": 14
+    },
+    {
+      "epoch": 0.008939213349225268,
+      "grad_norm": 147.3586883544922,
+      "learning_rate": 1.99046483909416e-05,
+      "loss": 4.7433,
+      "step": 15
+    },
+    {
+      "epoch": 0.009535160905840286,
+      "grad_norm": 139.62564086914062,
+      "learning_rate": 1.9892729439809298e-05,
+      "loss": 4.6783,
+      "step": 16
+    },
+    {
+      "epoch": 0.010131108462455305,
+      "grad_norm": 144.9489288330078,
+      "learning_rate": 1.9880810488677e-05,
+      "loss": 4.2209,
+      "step": 17
+    },
+    {
+      "epoch": 0.010727056019070322,
+      "grad_norm": 141.47108459472656,
+      "learning_rate": 1.9868891537544697e-05,
+      "loss": 3.9897,
+      "step": 18
+    },
+    {
+      "epoch": 0.01132300357568534,
+      "grad_norm": 142.5516357421875,
+      "learning_rate": 1.98569725864124e-05,
+      "loss": 3.7936,
+      "step": 19
+    },
+    {
+      "epoch": 0.011918951132300357,
+      "grad_norm": 145.7139434814453,
+      "learning_rate": 1.9845053635280096e-05,
+      "loss": 3.6945,
+      "step": 20
+    },
+    {
+      "epoch": 0.012514898688915376,
+      "grad_norm": 139.25314331054688,
+      "learning_rate": 1.9833134684147798e-05,
+      "loss": 3.9907,
+      "step": 21
+    },
+    {
+      "epoch": 0.013110846245530394,
+      "grad_norm": 135.04376220703125,
+      "learning_rate": 1.9821215733015496e-05,
+      "loss": 3.7992,
+      "step": 22
+    },
+    {
+      "epoch": 0.013706793802145411,
+      "grad_norm": 131.8486785888672,
+      "learning_rate": 1.9809296781883197e-05,
+      "loss": 3.7508,
+      "step": 23
+    },
+    {
+      "epoch": 0.014302741358760428,
+      "grad_norm": 140.6162109375,
+      "learning_rate": 1.9797377830750895e-05,
+      "loss": 3.1134,
+      "step": 24
+    },
+    {
+      "epoch": 0.014898688915375448,
+      "grad_norm": 138.89697265625,
+      "learning_rate": 1.9785458879618597e-05,
+      "loss": 2.9985,
+      "step": 25
+    },
+    {
+      "epoch": 0.015494636471990465,
+      "grad_norm": 129.2012481689453,
+      "learning_rate": 1.9773539928486295e-05,
+      "loss": 3.1729,
+      "step": 26
+    },
+    {
+      "epoch": 0.016090584028605484,
+      "grad_norm": 132.99954223632812,
+      "learning_rate": 1.9761620977353996e-05,
+      "loss": 2.9702,
+      "step": 27
+    },
+    {
+      "epoch": 0.0166865315852205,
+      "grad_norm": 127.68590545654297,
+      "learning_rate": 1.9749702026221694e-05,
+      "loss": 2.7236,
+      "step": 28
+    },
+    {
+      "epoch": 0.01728247914183552,
+      "grad_norm": 135.4009246826172,
+      "learning_rate": 1.9737783075089392e-05,
+      "loss": 2.3383,
+      "step": 29
+    },
+    {
+      "epoch": 0.017878426698450536,
+      "grad_norm": 135.87203979492188,
+      "learning_rate": 1.9725864123957094e-05,
+      "loss": 2.2006,
+      "step": 30
+    },
+    {
+      "epoch": 0.018474374255065554,
+      "grad_norm": 131.64749145507812,
+      "learning_rate": 1.9713945172824792e-05,
+      "loss": 2.101,
+      "step": 31
+    },
+    {
+      "epoch": 0.01907032181168057,
+      "grad_norm": 128.7794952392578,
+      "learning_rate": 1.9702026221692493e-05,
+      "loss": 2.0455,
+      "step": 32
+    },
+    {
+      "epoch": 0.01966626936829559,
+      "grad_norm": 120.35737609863281,
+      "learning_rate": 1.969010727056019e-05,
+      "loss": 2.0152,
+      "step": 33
+    },
+    {
+      "epoch": 0.02026221692491061,
+      "grad_norm": 113.17239379882812,
+      "learning_rate": 1.9678188319427893e-05,
+      "loss": 2.2596,
+      "step": 34
+    },
+    {
+      "epoch": 0.020858164481525627,
+      "grad_norm": 120.55252075195312,
+      "learning_rate": 1.966626936829559e-05,
+      "loss": 1.5437,
+      "step": 35
+    },
+    {
+      "epoch": 0.021454112038140644,
+      "grad_norm": 116.4307632446289,
+      "learning_rate": 1.965435041716329e-05,
+      "loss": 1.519,
+      "step": 36
+    },
+    {
+      "epoch": 0.02205005959475566,
+      "grad_norm": 110.38825988769531,
+      "learning_rate": 1.964243146603099e-05,
+      "loss": 1.2784,
+      "step": 37
+    },
+    {
+      "epoch": 0.02264600715137068,
+      "grad_norm": 104.32449340820312,
+      "learning_rate": 1.9630512514898688e-05,
+      "loss": 1.188,
+      "step": 38
+    },
+    {
+      "epoch": 0.023241954707985697,
+      "grad_norm": 97.6529312133789,
+      "learning_rate": 1.961859356376639e-05,
+      "loss": 1.4237,
+      "step": 39
+    },
+    {
+      "epoch": 0.023837902264600714,
+      "grad_norm": 91.67023468017578,
+      "learning_rate": 1.9606674612634088e-05,
+      "loss": 1.0271,
+      "step": 40
+    },
+    {
+      "epoch": 0.02443384982121573,
+      "grad_norm": 88.21671295166016,
+      "learning_rate": 1.959475566150179e-05,
+      "loss": 0.9279,
+      "step": 41
+    },
+    {
+      "epoch": 0.025029797377830752,
+      "grad_norm": 80.54320526123047,
+      "learning_rate": 1.9582836710369487e-05,
+      "loss": 0.8975,
+      "step": 42
+    },
+    {
+      "epoch": 0.02562574493444577,
+      "grad_norm": 74.2779312133789,
+      "learning_rate": 1.957091775923719e-05,
+      "loss": 0.7555,
+      "step": 43
+    },
+    {
+      "epoch": 0.026221692491060787,
+      "grad_norm": 66.96721649169922,
+      "learning_rate": 1.9558998808104887e-05,
+      "loss": 0.8921,
+      "step": 44
+    },
+    {
+      "epoch": 0.026817640047675805,
+      "grad_norm": 61.11864471435547,
+      "learning_rate": 1.9547079856972588e-05,
+      "loss": 0.615,
+      "step": 45
+    },
+    {
+      "epoch": 0.027413587604290822,
+      "grad_norm": 54.41999053955078,
+      "learning_rate": 1.9535160905840286e-05,
+      "loss": 0.6077,
+      "step": 46
+    },
+    {
+      "epoch": 0.02800953516090584,
+      "grad_norm": 46.97095489501953,
+      "learning_rate": 1.9523241954707987e-05,
+      "loss": 0.8338,
+      "step": 47
+    },
+    {
+      "epoch": 0.028605482717520857,
+      "grad_norm": 42.371009826660156,
+      "learning_rate": 1.9511323003575685e-05,
+      "loss": 0.49,
+      "step": 48
+    },
+    {
+      "epoch": 0.029201430274135878,
+      "grad_norm": 37.706153869628906,
+      "learning_rate": 1.9499404052443387e-05,
+      "loss": 0.4607,
+      "step": 49
+    },
+    {
+      "epoch": 0.029797377830750895,
+      "grad_norm": 32.81333541870117,
+      "learning_rate": 1.9487485101311085e-05,
+      "loss": 0.7072,
+      "step": 50
+    },
+    {
+      "epoch": 0.030393325387365912,
+      "grad_norm": 29.02756118774414,
+      "learning_rate": 1.9475566150178786e-05,
+      "loss": 0.4015,
+      "step": 51
+    },
+    {
+      "epoch": 0.03098927294398093,
+      "grad_norm": 24.758953094482422,
+      "learning_rate": 1.9463647199046484e-05,
+      "loss": 0.4546,
+      "step": 52
+    },
+    {
+      "epoch": 0.03158522050059595,
+      "grad_norm": 21.313373565673828,
+      "learning_rate": 1.9451728247914186e-05,
+      "loss": 0.5707,
+      "step": 53
+    },
+    {
+      "epoch": 0.03218116805721097,
+      "grad_norm": 18.881160736083984,
+      "learning_rate": 1.9439809296781884e-05,
+      "loss": 0.453,
+      "step": 54
+    },
+    {
+      "epoch": 0.032777115613825986,
+      "grad_norm": 16.438549041748047,
+      "learning_rate": 1.9427890345649585e-05,
+      "loss": 0.3477,
+      "step": 55
+    },
+    {
+      "epoch": 0.033373063170441,
+      "grad_norm": 14.138507843017578,
+      "learning_rate": 1.9415971394517283e-05,
+      "loss": 0.2178,
+      "step": 56
+    },
+    {
+      "epoch": 0.03396901072705602,
+      "grad_norm": 12.81236743927002,
+      "learning_rate": 1.9404052443384985e-05,
+      "loss": 0.5035,
+      "step": 57
+    },
+    {
+      "epoch": 0.03456495828367104,
+      "grad_norm": 10.96316146850586,
+      "learning_rate": 1.9392133492252683e-05,
+      "loss": 0.226,
+      "step": 58
+    },
+    {
+      "epoch": 0.035160905840286055,
+      "grad_norm": 9.160541534423828,
+      "learning_rate": 1.9380214541120384e-05,
+      "loss": 0.3757,
+      "step": 59
+    },
+    {
+      "epoch": 0.03575685339690107,
+      "grad_norm": 9.226065635681152,
+      "learning_rate": 1.9368295589988082e-05,
+      "loss": 0.4098,
+      "step": 60
+    },
+    {
+      "epoch": 0.03635280095351609,
+      "grad_norm": 8.672224044799805,
+      "learning_rate": 1.9356376638855784e-05,
+      "loss": 0.2866,
+      "step": 61
+    },
+    {
+      "epoch": 0.03694874851013111,
+      "grad_norm": 11.421109199523926,
+      "learning_rate": 1.9344457687723482e-05,
+      "loss": 0.9299,
+      "step": 62
+    },
+    {
+      "epoch": 0.037544696066746125,
+      "grad_norm": 9.902897834777832,
+      "learning_rate": 1.9332538736591183e-05,
+      "loss": 0.9075,
+      "step": 63
+    },
+    {
+      "epoch": 0.03814064362336114,
+      "grad_norm": 7.028914451599121,
+      "learning_rate": 1.932061978545888e-05,
+      "loss": 0.311,
+      "step": 64
+    },
+    {
+      "epoch": 0.03873659117997616,
+      "grad_norm": 5.583148002624512,
+      "learning_rate": 1.9308700834326583e-05,
+      "loss": 0.1797,
+      "step": 65
+    },
+    {
+      "epoch": 0.03933253873659118,
+      "grad_norm": 8.053176879882812,
+      "learning_rate": 1.929678188319428e-05,
+      "loss": 0.6665,
+      "step": 66
+    },
+    {
+      "epoch": 0.039928486293206195,
+      "grad_norm": 6.080432415008545,
+      "learning_rate": 1.9284862932061982e-05,
+      "loss": 0.3453,
+      "step": 67
+    },
+    {
+      "epoch": 0.04052443384982122,
+      "grad_norm": 5.835549354553223,
+      "learning_rate": 1.927294398092968e-05,
+      "loss": 0.1982,
+      "step": 68
+    },
+    {
+      "epoch": 0.041120381406436236,
+      "grad_norm": 5.077045917510986,
+      "learning_rate": 1.926102502979738e-05,
+      "loss": 0.1953,
+      "step": 69
+    },
+    {
+      "epoch": 0.041716328963051254,
+      "grad_norm": 3.7127647399902344,
+      "learning_rate": 1.924910607866508e-05,
+      "loss": 0.0752,
+      "step": 70
+    },
+    {
+      "epoch": 0.04231227651966627,
+      "grad_norm": 3.248577117919922,
+      "learning_rate": 1.9237187127532778e-05,
+      "loss": 0.1377,
+      "step": 71
+    },
+    {
+      "epoch": 0.04290822407628129,
+      "grad_norm": 8.281804084777832,
+      "learning_rate": 1.922526817640048e-05,
+      "loss": 0.3446,
+      "step": 72
+    },
+    {
+      "epoch": 0.043504171632896306,
+      "grad_norm": 8.818718910217285,
+      "learning_rate": 1.9213349225268177e-05,
+      "loss": 0.6061,
+      "step": 73
+    },
+    {
+      "epoch": 0.04410011918951132,
+      "grad_norm": 6.09969425201416,
+      "learning_rate": 1.920143027413588e-05,
+      "loss": 0.2747,
+      "step": 74
+    },
+    {
+      "epoch": 0.04469606674612634,
+      "grad_norm": 7.3518805503845215,
+      "learning_rate": 1.9189511323003577e-05,
+      "loss": 0.3024,
+      "step": 75
+    },
+    {
+      "epoch": 0.04529201430274136,
+      "grad_norm": 9.40757942199707,
+      "learning_rate": 1.9177592371871278e-05,
+      "loss": 0.7653,
+      "step": 76
+    },
+    {
+      "epoch": 0.045887961859356376,
+      "grad_norm": 4.576373100280762,
+      "learning_rate": 1.9165673420738976e-05,
+      "loss": 0.187,
+      "step": 77
+    },
+    {
+      "epoch": 0.04648390941597139,
+      "grad_norm": 7.151569843292236,
+      "learning_rate": 1.9153754469606677e-05,
+      "loss": 0.4414,
+      "step": 78
+    },
+    {
+      "epoch": 0.04707985697258641,
+      "grad_norm": 3.414809465408325,
+      "learning_rate": 1.9141835518474375e-05,
+      "loss": 0.1401,
+      "step": 79
+    },
+    {
+      "epoch": 0.04767580452920143,
+      "grad_norm": 6.369510650634766,
+      "learning_rate": 1.9129916567342073e-05,
+      "loss": 0.4594,
+      "step": 80
+    },
+    {
+      "epoch": 0.048271752085816445,
+      "grad_norm": 3.743135929107666,
+      "learning_rate": 1.9117997616209775e-05,
+      "loss": 0.1178,
+      "step": 81
+    },
+    {
+      "epoch": 0.04886769964243146,
+      "grad_norm": 4.639786720275879,
+      "learning_rate": 1.9106078665077473e-05,
+      "loss": 0.1813,
+      "step": 82
+    },
+    {
+      "epoch": 0.04946364719904649,
+      "grad_norm": 4.444789409637451,
+      "learning_rate": 1.9094159713945174e-05,
+      "loss": 0.1138,
+      "step": 83
+    },
+    {
+      "epoch": 0.050059594755661505,
+      "grad_norm": 4.018325328826904,
+      "learning_rate": 1.9082240762812872e-05,
+      "loss": 0.1074,
+      "step": 84
+    },
+    {
+      "epoch": 0.05065554231227652,
+      "grad_norm": 5.398996353149414,
+      "learning_rate": 1.9070321811680574e-05,
+      "loss": 0.417,
+      "step": 85
+    },
+    {
+      "epoch": 0.05125148986889154,
+      "grad_norm": 6.999519348144531,
+      "learning_rate": 1.9058402860548272e-05,
+      "loss": 0.4248,
+      "step": 86
+    },
+    {
+      "epoch": 0.05184743742550656,
+      "grad_norm": 5.825213432312012,
+      "learning_rate": 1.9046483909415973e-05,
+      "loss": 0.3249,
+      "step": 87
+    },
+    {
+      "epoch": 0.052443384982121574,
+      "grad_norm": 12.55892562866211,
+      "learning_rate": 1.903456495828367e-05,
+      "loss": 0.8146,
+      "step": 88
+    },
+    {
+      "epoch": 0.05303933253873659,
+      "grad_norm": 3.8351542949676514,
+      "learning_rate": 1.9022646007151373e-05,
+      "loss": 0.1325,
+      "step": 89
+    },
+    {
+      "epoch": 0.05363528009535161,
+      "grad_norm": 2.609816074371338,
+      "learning_rate": 1.901072705601907e-05,
+      "loss": 0.1198,
+      "step": 90
+    },
+    {
+      "epoch": 0.054231227651966626,
+      "grad_norm": 5.644624710083008,
+      "learning_rate": 1.8998808104886772e-05,
+      "loss": 0.3787,
+      "step": 91
+    },
+    {
+      "epoch": 0.054827175208581644,
+      "grad_norm": 12.737455368041992,
+      "learning_rate": 1.898688915375447e-05,
+      "loss": 1.0291,
+      "step": 92
+    },
+    {
+      "epoch": 0.05542312276519666,
+      "grad_norm": 4.7286858558654785,
+      "learning_rate": 1.897497020262217e-05,
+      "loss": 0.2583,
+      "step": 93
+    },
+    {
+      "epoch": 0.05601907032181168,
+      "grad_norm": 5.505688667297363,
+      "learning_rate": 1.896305125148987e-05,
+      "loss": 0.2768,
+      "step": 94
+    },
+    {
+      "epoch": 0.056615017878426696,
+      "grad_norm": 7.857344150543213,
+      "learning_rate": 1.895113230035757e-05,
+      "loss": 0.4292,
+      "step": 95
+    },
+    {
+      "epoch": 0.057210965435041714,
+      "grad_norm": 5.854008674621582,
+      "learning_rate": 1.893921334922527e-05,
+      "loss": 0.2735,
+      "step": 96
+    },
+    {
+      "epoch": 0.05780691299165673,
+      "grad_norm": 7.622176647186279,
+      "learning_rate": 1.892729439809297e-05,
+      "loss": 0.5842,
+      "step": 97
+    },
+    {
+      "epoch": 0.058402860548271755,
+      "grad_norm": 5.666799545288086,
+      "learning_rate": 1.891537544696067e-05,
+      "loss": 0.388,
+      "step": 98
+    },
+    {
+      "epoch": 0.05899880810488677,
+      "grad_norm": 4.963497638702393,
+      "learning_rate": 1.890345649582837e-05,
+      "loss": 0.2795,
+      "step": 99
+    },
+    {
+      "epoch": 0.05959475566150179,
+      "grad_norm": 3.273045539855957,
+      "learning_rate": 1.8891537544696068e-05,
+      "loss": 0.1235,
+      "step": 100
+    },
+    {
+      "epoch": 0.06019070321811681,
+      "grad_norm": 5.198390960693359,
+      "learning_rate": 1.887961859356377e-05,
+      "loss": 0.2454,
+      "step": 101
+    },
+    {
+      "epoch": 0.060786650774731825,
+      "grad_norm": 3.798640012741089,
+      "learning_rate": 1.8867699642431468e-05,
+      "loss": 0.1865,
+      "step": 102
+    },
+    {
+      "epoch": 0.06138259833134684,
+      "grad_norm": 3.564096689224243,
+      "learning_rate": 1.885578069129917e-05,
+      "loss": 0.1452,
+      "step": 103
+    },
+    {
+      "epoch": 0.06197854588796186,
+      "grad_norm": 3.4654436111450195,
+      "learning_rate": 1.8843861740166867e-05,
+      "loss": 0.1698,
+      "step": 104
+    },
+    {
+      "epoch": 0.06257449344457688,
+      "grad_norm": 5.602324485778809,
+      "learning_rate": 1.883194278903457e-05,
+      "loss": 0.3947,
+      "step": 105
+    },
+    {
+      "epoch": 0.0631704410011919,
+      "grad_norm": 4.1538543701171875,
+      "learning_rate": 1.8820023837902266e-05,
+      "loss": 0.1222,
+      "step": 106
+    },
+    {
+      "epoch": 0.06376638855780692,
+      "grad_norm": 5.4469757080078125,
+      "learning_rate": 1.8808104886769968e-05,
+      "loss": 0.1916,
+      "step": 107
+    },
+    {
+      "epoch": 0.06436233611442194,
+      "grad_norm": 3.6765644550323486,
+      "learning_rate": 1.8796185935637666e-05,
+      "loss": 0.187,
+      "step": 108
+    },
+    {
+      "epoch": 0.06495828367103695,
+      "grad_norm": 4.824008464813232,
+      "learning_rate": 1.8784266984505367e-05,
+      "loss": 0.2379,
+      "step": 109
+    },
+    {
+      "epoch": 0.06555423122765197,
+      "grad_norm": 4.435067176818848,
+      "learning_rate": 1.8772348033373065e-05,
+      "loss": 0.1973,
+      "step": 110
+    },
+    {
+      "epoch": 0.06615017878426699,
+      "grad_norm": 4.906465530395508,
+      "learning_rate": 1.8760429082240767e-05,
+      "loss": 0.2473,
+      "step": 111
+    },
+    {
+      "epoch": 0.066746126340882,
+      "grad_norm": 7.3632731437683105,
+      "learning_rate": 1.8748510131108465e-05,
+      "loss": 0.6298,
+      "step": 112
+    },
+    {
+      "epoch": 0.06734207389749702,
+      "grad_norm": 4.159063339233398,
+      "learning_rate": 1.8736591179976166e-05,
+      "loss": 0.1748,
+      "step": 113
+    },
+    {
+      "epoch": 0.06793802145411204,
+      "grad_norm": 3.355159282684326,
+      "learning_rate": 1.8724672228843864e-05,
+      "loss": 0.1008,
+      "step": 114
+    },
+    {
+      "epoch": 0.06853396901072706,
+      "grad_norm": 10.36339282989502,
+      "learning_rate": 1.8712753277711562e-05,
+      "loss": 0.906,
+      "step": 115
+    },
+    {
+      "epoch": 0.06912991656734208,
+      "grad_norm": 4.6436309814453125,
+      "learning_rate": 1.8700834326579264e-05,
+      "loss": 0.2151,
+      "step": 116
+    },
+    {
+      "epoch": 0.0697258641239571,
+      "grad_norm": 4.008418560028076,
+      "learning_rate": 1.8688915375446962e-05,
+      "loss": 0.1666,
+      "step": 117
+    },
+    {
+      "epoch": 0.07032181168057211,
+      "grad_norm": 3.397493362426758,
+      "learning_rate": 1.8676996424314663e-05,
+      "loss": 0.1279,
+      "step": 118
+    },
+    {
+      "epoch": 0.07091775923718713,
+      "grad_norm": 5.331108570098877,
+      "learning_rate": 1.866507747318236e-05,
+      "loss": 0.2709,
+      "step": 119
+    },
+    {
+      "epoch": 0.07151370679380215,
+      "grad_norm": 3.3081233501434326,
+      "learning_rate": 1.8653158522050063e-05,
+      "loss": 0.072,
+      "step": 120
+    },
+    {
+      "epoch": 0.07210965435041716,
+      "grad_norm": 4.1846394538879395,
+      "learning_rate": 1.864123957091776e-05,
+      "loss": 0.0644,
+      "step": 121
+    },
+    {
+      "epoch": 0.07270560190703218,
+      "grad_norm": 4.554184913635254,
+      "learning_rate": 1.862932061978546e-05,
+      "loss": 0.258,
+      "step": 122
+    },
+    {
+      "epoch": 0.0733015494636472,
+      "grad_norm": 4.762580394744873,
+      "learning_rate": 1.861740166865316e-05,
+      "loss": 0.1486,
+      "step": 123
+    },
+    {
+      "epoch": 0.07389749702026222,
+      "grad_norm": 4.100523471832275,
+      "learning_rate": 1.8605482717520858e-05,
+      "loss": 0.0663,
+      "step": 124
+    },
+    {
+      "epoch": 0.07449344457687723,
+      "grad_norm": 13.981008529663086,
+      "learning_rate": 1.859356376638856e-05,
+      "loss": 0.2176,
+      "step": 125
+    },
+    {
+      "epoch": 0.07508939213349225,
+      "grad_norm": 3.96645188331604,
+      "learning_rate": 1.8581644815256258e-05,
+      "loss": 0.1608,
+      "step": 126
+    },
+    {
+      "epoch": 0.07568533969010727,
+      "grad_norm": 2.609151601791382,
+      "learning_rate": 1.856972586412396e-05,
+      "loss": 0.0701,
+      "step": 127
+    },
+    {
+      "epoch": 0.07628128724672228,
+      "grad_norm": 4.328754425048828,
+      "learning_rate": 1.8557806912991657e-05,
+      "loss": 0.2382,
+      "step": 128
+    },
+    {
+      "epoch": 0.0768772348033373,
+      "grad_norm": 1.5854657888412476,
+      "learning_rate": 1.8545887961859355e-05,
+      "loss": 0.0309,
+      "step": 129
+    },
+    {
+      "epoch": 0.07747318235995232,
+      "grad_norm": 3.3432695865631104,
+      "learning_rate": 1.8533969010727057e-05,
+      "loss": 0.0794,
+      "step": 130
+    },
+    {
+      "epoch": 0.07806912991656734,
+      "grad_norm": 3.9022483825683594,
+      "learning_rate": 1.8522050059594755e-05,
+      "loss": 0.1531,
+      "step": 131
+    },
+    {
+      "epoch": 0.07866507747318235,
+      "grad_norm": 3.414245843887329,
+      "learning_rate": 1.8510131108462456e-05,
+      "loss": 0.1827,
+      "step": 132
+    },
+    {
+      "epoch": 0.07926102502979737,
+      "grad_norm": 8.38995361328125,
+      "learning_rate": 1.8498212157330154e-05,
+      "loss": 0.6837,
+      "step": 133
+    },
+    {
+      "epoch": 0.07985697258641239,
+      "grad_norm": 3.4881913661956787,
+      "learning_rate": 1.8486293206197856e-05,
+      "loss": 0.0845,
+      "step": 134
+    },
+    {
+      "epoch": 0.08045292014302742,
+      "grad_norm": 6.429633617401123,
+      "learning_rate": 1.8474374255065554e-05,
+      "loss": 0.3727,
+      "step": 135
+    },
+    {
+      "epoch": 0.08104886769964244,
+      "grad_norm": 7.204514026641846,
+      "learning_rate": 1.8462455303933255e-05,
+      "loss": 0.4328,
+      "step": 136
+    },
+    {
+      "epoch": 0.08164481525625746,
+      "grad_norm": 4.075184345245361,
+      "learning_rate": 1.8450536352800953e-05,
+      "loss": 0.0691,
+      "step": 137
+    },
+    {
+      "epoch": 0.08224076281287247,
+      "grad_norm": 3.96445631980896,
+      "learning_rate": 1.8438617401668654e-05,
+      "loss": 0.1986,
+      "step": 138
+    },
+    {
+      "epoch": 0.08283671036948749,
+      "grad_norm": 5.069427490234375,
+      "learning_rate": 1.8426698450536352e-05,
+      "loss": 0.2399,
+      "step": 139
+    },
+    {
+      "epoch": 0.08343265792610251,
+      "grad_norm": 6.1384501457214355,
+      "learning_rate": 1.8414779499404054e-05,
+      "loss": 0.5938,
+      "step": 140
+    },
+    {
+      "epoch": 0.08402860548271752,
+      "grad_norm": 6.480672836303711,
+      "learning_rate": 1.8402860548271752e-05,
+      "loss": 0.5124,
+      "step": 141
+    },
+    {
+      "epoch": 0.08462455303933254,
+      "grad_norm": 8.202727317810059,
+      "learning_rate": 1.8390941597139453e-05,
+      "loss": 0.5156,
+      "step": 142
+    },
+    {
+      "epoch": 0.08522050059594756,
+      "grad_norm": 6.866783142089844,
+      "learning_rate": 1.837902264600715e-05,
+      "loss": 0.5801,
+      "step": 143
+    },
+    {
+      "epoch": 0.08581644815256258,
+      "grad_norm": 2.4662835597991943,
+      "learning_rate": 1.8367103694874853e-05,
+      "loss": 0.0752,
+      "step": 144
+    },
+    {
+      "epoch": 0.0864123957091776,
+      "grad_norm": 4.357907772064209,
+      "learning_rate": 1.835518474374255e-05,
+      "loss": 0.1749,
+      "step": 145
+    },
+    {
+      "epoch": 0.08700834326579261,
+      "grad_norm": 33.758792877197266,
+      "learning_rate": 1.8343265792610252e-05,
+      "loss": 1.0844,
+      "step": 146
+    },
+    {
+      "epoch": 0.08760429082240763,
+      "grad_norm": 5.2374138832092285,
+      "learning_rate": 1.833134684147795e-05,
+      "loss": 0.2972,
+      "step": 147
+    },
+    {
+      "epoch": 0.08820023837902265,
+      "grad_norm": 4.944085121154785,
+      "learning_rate": 1.8319427890345652e-05,
+      "loss": 0.2445,
+      "step": 148
+    },
+    {
+      "epoch": 0.08879618593563766,
+      "grad_norm": 5.903630256652832,
+      "learning_rate": 1.830750893921335e-05,
+      "loss": 0.2985,
+      "step": 149
+    },
+    {
+      "epoch": 0.08939213349225268,
+      "grad_norm": 7.153575897216797,
+      "learning_rate": 1.829558998808105e-05,
+      "loss": 0.5676,
+      "step": 150
+    },
+    {
+      "epoch": 0.0899880810488677,
+      "grad_norm": 2.368042230606079,
+      "learning_rate": 1.828367103694875e-05,
+      "loss": 0.0666,
+      "step": 151
+    },
+    {
+      "epoch": 0.09058402860548272,
+      "grad_norm": 3.9073240756988525,
+      "learning_rate": 1.827175208581645e-05,
+      "loss": 0.1956,
+      "step": 152
+    },
+    {
+      "epoch": 0.09117997616209773,
+      "grad_norm": 6.607059955596924,
+      "learning_rate": 1.825983313468415e-05,
+      "loss": 0.3595,
+      "step": 153
+    },
+    {
+      "epoch": 0.09177592371871275,
+      "grad_norm": 2.6297554969787598,
+      "learning_rate": 1.824791418355185e-05,
+      "loss": 0.0574,
+      "step": 154
+    },
+    {
+      "epoch": 0.09237187127532777,
+      "grad_norm": 3.3877627849578857,
+      "learning_rate": 1.8235995232419548e-05,
+      "loss": 0.0985,
+      "step": 155
+    },
+    {
+      "epoch": 0.09296781883194279,
+      "grad_norm": 3.0368971824645996,
+      "learning_rate": 1.822407628128725e-05,
+      "loss": 0.1247,
+      "step": 156
+    },
+    {
+      "epoch": 0.0935637663885578,
+      "grad_norm": 3.5671985149383545,
+      "learning_rate": 1.8212157330154948e-05,
+      "loss": 0.2132,
+      "step": 157
+    },
+    {
+      "epoch": 0.09415971394517282,
+      "grad_norm": 3.48213791847229,
+      "learning_rate": 1.820023837902265e-05,
+      "loss": 0.1161,
+      "step": 158
+    },
+    {
+      "epoch": 0.09475566150178784,
+      "grad_norm": 3.135429859161377,
+      "learning_rate": 1.8188319427890347e-05,
+      "loss": 0.1238,
+      "step": 159
+    },
+    {
+      "epoch": 0.09535160905840286,
+      "grad_norm": 3.5636098384857178,
+      "learning_rate": 1.817640047675805e-05,
+      "loss": 0.1446,
+      "step": 160
+    },
+    {
+      "epoch": 0.09594755661501787,
+      "grad_norm": 3.3986358642578125,
+      "learning_rate": 1.8164481525625747e-05,
+      "loss": 0.0779,
+      "step": 161
+    },
+    {
+      "epoch": 0.09654350417163289,
+      "grad_norm": 5.139955043792725,
+      "learning_rate": 1.8152562574493448e-05,
+      "loss": 0.3341,
+      "step": 162
+    },
+    {
+      "epoch": 0.09713945172824791,
+      "grad_norm": 4.62930154800415,
+      "learning_rate": 1.8140643623361146e-05,
+      "loss": 0.2878,
+      "step": 163
+    },
+    {
+      "epoch": 0.09773539928486293,
+      "grad_norm": 4.725478649139404,
+      "learning_rate": 1.8128724672228844e-05,
+      "loss": 0.1353,
+      "step": 164
+    },
+    {
+      "epoch": 0.09833134684147796,
+      "grad_norm": 2.8831610679626465,
+      "learning_rate": 1.8116805721096545e-05,
+      "loss": 0.1311,
+      "step": 165
+    },
+    {
+      "epoch": 0.09892729439809297,
+      "grad_norm": 2.89089035987854,
+      "learning_rate": 1.8104886769964244e-05,
+      "loss": 0.1227,
+      "step": 166
+    },
+    {
+      "epoch": 0.09952324195470799,
+      "grad_norm": 4.159611701965332,
+      "learning_rate": 1.8092967818831945e-05,
+      "loss": 0.2113,
+      "step": 167
+    },
+    {
+      "epoch": 0.10011918951132301,
+      "grad_norm": 5.738676071166992,
+      "learning_rate": 1.8081048867699643e-05,
+      "loss": 0.0882,
+      "step": 168
+    },
+    {
+      "epoch": 0.10071513706793803,
+      "grad_norm": 4.988555431365967,
+      "learning_rate": 1.8069129916567344e-05,
+      "loss": 0.3135,
+      "step": 169
+    },
+    {
+      "epoch": 0.10131108462455304,
+      "grad_norm": 5.939578533172607,
+      "learning_rate": 1.8057210965435042e-05,
+      "loss": 0.1234,
+      "step": 170
+    },
+    {
+      "epoch": 0.10190703218116806,
+      "grad_norm": 6.8424201011657715,
+      "learning_rate": 1.804529201430274e-05,
+      "loss": 0.4079,
+      "step": 171
+    },
+    {
+      "epoch": 0.10250297973778308,
+      "grad_norm": 4.633545875549316,
+      "learning_rate": 1.8033373063170442e-05,
+      "loss": 0.2347,
+      "step": 172
+    },
+    {
+      "epoch": 0.1030989272943981,
+      "grad_norm": 2.521296501159668,
+      "learning_rate": 1.802145411203814e-05,
+      "loss": 0.0631,
+      "step": 173
+    },
+    {
+      "epoch": 0.10369487485101311,
+      "grad_norm": 3.530196189880371,
+      "learning_rate": 1.800953516090584e-05,
+      "loss": 0.1376,
+      "step": 174
+    },
+    {
+      "epoch": 0.10429082240762813,
+      "grad_norm": 4.016501426696777,
+      "learning_rate": 1.799761620977354e-05,
+      "loss": 0.1158,
+      "step": 175
+    },
+    {
+      "epoch": 0.10488676996424315,
+      "grad_norm": 5.288716793060303,
+      "learning_rate": 1.798569725864124e-05,
+      "loss": 0.2934,
+      "step": 176
+    },
+    {
+      "epoch": 0.10548271752085817,
+      "grad_norm": 3.0140902996063232,
+      "learning_rate": 1.797377830750894e-05,
+      "loss": 0.1485,
+      "step": 177
+    },
+    {
+      "epoch": 0.10607866507747318,
+      "grad_norm": 1.881548285484314,
+      "learning_rate": 1.796185935637664e-05,
+      "loss": 0.0396,
+      "step": 178
+    },
+    {
+      "epoch": 0.1066746126340882,
+      "grad_norm": 3.816077947616577,
+      "learning_rate": 1.794994040524434e-05,
+      "loss": 0.1751,
+      "step": 179
+    },
+    {
+      "epoch": 0.10727056019070322,
+      "grad_norm": 3.3778727054595947,
+      "learning_rate": 1.793802145411204e-05,
+      "loss": 0.0682,
+      "step": 180
+    },
+    {
+      "epoch": 0.10786650774731824,
+      "grad_norm": 3.352740526199341,
+      "learning_rate": 1.7926102502979738e-05,
+      "loss": 0.1384,
+      "step": 181
+    },
+    {
+      "epoch": 0.10846245530393325,
+      "grad_norm": 5.727471828460693,
+      "learning_rate": 1.791418355184744e-05,
+      "loss": 0.3424,
+      "step": 182
+    },
+    {
+      "epoch": 0.10905840286054827,
+      "grad_norm": 4.831298351287842,
+      "learning_rate": 1.7902264600715137e-05,
+      "loss": 0.289,
+      "step": 183
+    },
+    {
+      "epoch": 0.10965435041716329,
+      "grad_norm": 6.837507247924805,
+      "learning_rate": 1.789034564958284e-05,
+      "loss": 0.4744,
+      "step": 184
+    },
+    {
+      "epoch": 0.1102502979737783,
+      "grad_norm": 3.6191353797912598,
+      "learning_rate": 1.7878426698450537e-05,
+      "loss": 0.1138,
+      "step": 185
+    },
+    {
+      "epoch": 0.11084624553039332,
+      "grad_norm": 3.563807487487793,
+      "learning_rate": 1.7866507747318238e-05,
+      "loss": 0.0697,
+      "step": 186
+    },
+    {
+      "epoch": 0.11144219308700834,
+      "grad_norm": 6.018884181976318,
+      "learning_rate": 1.7854588796185936e-05,
+      "loss": 0.4043,
+      "step": 187
+    },
+    {
+      "epoch": 0.11203814064362336,
+      "grad_norm": 2.367751359939575,
+      "learning_rate": 1.7842669845053638e-05,
+      "loss": 0.0571,
+      "step": 188
+    },
+    {
+      "epoch": 0.11263408820023837,
+      "grad_norm": 4.860586166381836,
+      "learning_rate": 1.7830750893921336e-05,
+      "loss": 0.101,
+      "step": 189
+    },
+    {
+      "epoch": 0.11323003575685339,
+      "grad_norm": 7.2254862785339355,
+      "learning_rate": 1.7818831942789037e-05,
+      "loss": 0.3463,
+      "step": 190
+    },
+    {
+      "epoch": 0.11382598331346841,
+      "grad_norm": 5.238307476043701,
+      "learning_rate": 1.7806912991656735e-05,
+      "loss": 0.2587,
+      "step": 191
+    },
+    {
+      "epoch": 0.11442193087008343,
+      "grad_norm": 3.523691177368164,
+      "learning_rate": 1.7794994040524436e-05,
+      "loss": 0.0414,
+      "step": 192
+    },
+    {
+      "epoch": 0.11501787842669844,
+      "grad_norm": 5.268319606781006,
+      "learning_rate": 1.7783075089392135e-05,
+      "loss": 0.2459,
+      "step": 193
+    },
+    {
+      "epoch": 0.11561382598331346,
+      "grad_norm": 3.1465065479278564,
+      "learning_rate": 1.7771156138259836e-05,
+      "loss": 0.0846,
+      "step": 194
+    },
+    {
+      "epoch": 0.1162097735399285,
+      "grad_norm": 3.5932445526123047,
+      "learning_rate": 1.7759237187127534e-05,
+      "loss": 0.0655,
+      "step": 195
+    },
+    {
+      "epoch": 0.11680572109654351,
+      "grad_norm": 9.544511795043945,
+      "learning_rate": 1.7747318235995235e-05,
+      "loss": 0.9057,
+      "step": 196
+    },
+    {
+      "epoch": 0.11740166865315853,
+      "grad_norm": 5.303696632385254,
+      "learning_rate": 1.7735399284862933e-05,
+      "loss": 0.1309,
+      "step": 197
+    },
+    {
+      "epoch": 0.11799761620977355,
+      "grad_norm": 3.789337396621704,
+      "learning_rate": 1.7723480333730635e-05,
+      "loss": 0.1435,
+      "step": 198
+    },
+    {
+      "epoch": 0.11859356376638856,
+      "grad_norm": 5.966098785400391,
+      "learning_rate": 1.7711561382598333e-05,
+      "loss": 0.3302,
+      "step": 199
+    },
+    {
+      "epoch": 0.11918951132300358,
+      "grad_norm": 3.6656033992767334,
+      "learning_rate": 1.7699642431466034e-05,
+      "loss": 0.1324,
+      "step": 200
+    },
+    {
+      "epoch": 0.1197854588796186,
+      "grad_norm": 1.9895102977752686,
+      "learning_rate": 1.7687723480333732e-05,
+      "loss": 0.0246,
+      "step": 201
+    },
+    {
+      "epoch": 0.12038140643623362,
+      "grad_norm": 8.061299324035645,
+      "learning_rate": 1.7675804529201434e-05,
+      "loss": 0.1567,
+      "step": 202
+    },
+    {
+      "epoch": 0.12097735399284863,
+      "grad_norm": 7.891781806945801,
+      "learning_rate": 1.7663885578069132e-05,
+      "loss": 0.3643,
+      "step": 203
+    },
+    {
+      "epoch": 0.12157330154946365,
+      "grad_norm": 4.830395221710205,
+      "learning_rate": 1.7651966626936833e-05,
+      "loss": 0.2363,
+      "step": 204
+    },
+    {
+      "epoch": 0.12216924910607867,
+      "grad_norm": 5.611730098724365,
+      "learning_rate": 1.764004767580453e-05,
+      "loss": 0.2592,
+      "step": 205
+    },
+    {
+      "epoch": 0.12276519666269368,
+      "grad_norm": 6.569658279418945,
+      "learning_rate": 1.762812872467223e-05,
+      "loss": 0.3077,
+      "step": 206
+    },
+    {
+      "epoch": 0.1233611442193087,
+      "grad_norm": 4.949154376983643,
+      "learning_rate": 1.761620977353993e-05,
+      "loss": 0.332,
+      "step": 207
+    },
+    {
+      "epoch": 0.12395709177592372,
+      "grad_norm": 5.548079967498779,
+      "learning_rate": 1.760429082240763e-05,
+      "loss": 0.161,
+      "step": 208
+    },
+    {
+      "epoch": 0.12455303933253874,
+      "grad_norm": 2.7452847957611084,
+      "learning_rate": 1.759237187127533e-05,
+      "loss": 0.0668,
+      "step": 209
+    },
+    {
+      "epoch": 0.12514898688915377,
+      "grad_norm": 6.824844837188721,
+      "learning_rate": 1.7580452920143028e-05,
+      "loss": 0.317,
+      "step": 210
+    },
+    {
+      "epoch": 0.12574493444576879,
+      "grad_norm": 4.535915374755859,
+      "learning_rate": 1.756853396901073e-05,
+      "loss": 0.2102,
+      "step": 211
+    },
+    {
+      "epoch": 0.1263408820023838,
+      "grad_norm": 3.1857800483703613,
+      "learning_rate": 1.7556615017878428e-05,
+      "loss": 0.1054,
+      "step": 212
+    },
+    {
+      "epoch": 0.12693682955899882,
+      "grad_norm": 3.0451090335845947,
+      "learning_rate": 1.754469606674613e-05,
+      "loss": 0.0605,
+      "step": 213
+    },
+    {
+      "epoch": 0.12753277711561384,
+      "grad_norm": 2.5186240673065186,
+      "learning_rate": 1.7532777115613827e-05,
+      "loss": 0.0543,
+      "step": 214
+    },
+    {
+      "epoch": 0.12812872467222886,
+      "grad_norm": 8.574667930603027,
+      "learning_rate": 1.7520858164481525e-05,
+      "loss": 0.779,
+      "step": 215
+    },
+    {
+      "epoch": 0.12872467222884387,
+      "grad_norm": 11.198761940002441,
+      "learning_rate": 1.7508939213349227e-05,
+      "loss": 0.2995,
+      "step": 216
+    },
+    {
+      "epoch": 0.1293206197854589,
+      "grad_norm": 6.322816371917725,
+      "learning_rate": 1.7497020262216925e-05,
+      "loss": 0.2521,
+      "step": 217
+    },
+    {
+      "epoch": 0.1299165673420739,
+      "grad_norm": 4.039135456085205,
+      "learning_rate": 1.7485101311084626e-05,
+      "loss": 0.1411,
+      "step": 218
+    },
+    {
+      "epoch": 0.13051251489868892,
+      "grad_norm": 3.707063913345337,
+      "learning_rate": 1.7473182359952324e-05,
+      "loss": 0.112,
+      "step": 219
+    },
+    {
+      "epoch": 0.13110846245530394,
+      "grad_norm": 2.106912851333618,
+      "learning_rate": 1.7461263408820026e-05,
+      "loss": 0.0525,
+      "step": 220
+    },
+    {
+      "epoch": 0.13170441001191896,
+      "grad_norm": 4.139005184173584,
+      "learning_rate": 1.7449344457687724e-05,
+      "loss": 0.1409,
+      "step": 221
+    },
+    {
+      "epoch": 0.13230035756853398,
+      "grad_norm": 15.226531028747559,
+      "learning_rate": 1.7437425506555425e-05,
+      "loss": 0.6601,
+      "step": 222
+    },
+    {
+      "epoch": 0.132896305125149,
+      "grad_norm": 6.510365962982178,
+      "learning_rate": 1.7425506555423123e-05,
+      "loss": 0.4981,
+      "step": 223
+    },
+    {
+      "epoch": 0.133492252681764,
+      "grad_norm": 3.3074169158935547,
+      "learning_rate": 1.7413587604290824e-05,
+      "loss": 0.0498,
+      "step": 224
+    },
+    {
+      "epoch": 0.13408820023837903,
+      "grad_norm": 4.977347373962402,
+      "learning_rate": 1.7401668653158523e-05,
+      "loss": 0.1616,
+      "step": 225
+    },
+    {
+      "epoch": 0.13468414779499405,
+      "grad_norm": 2.7214295864105225,
+      "learning_rate": 1.7389749702026224e-05,
+      "loss": 0.0972,
+      "step": 226
+    },
+    {
+      "epoch": 0.13528009535160906,
+      "grad_norm": 2.533388614654541,
+      "learning_rate": 1.7377830750893922e-05,
+      "loss": 0.0451,
+      "step": 227
+    },
+    {
+      "epoch": 0.13587604290822408,
+      "grad_norm": 8.205904006958008,
+      "learning_rate": 1.7365911799761623e-05,
+      "loss": 0.7512,
+      "step": 228
+    },
+    {
+      "epoch": 0.1364719904648391,
+      "grad_norm": 3.2948498725891113,
+      "learning_rate": 1.735399284862932e-05,
+      "loss": 0.0687,
+      "step": 229
+    },
+    {
+      "epoch": 0.13706793802145412,
+      "grad_norm": 2.03399658203125,
+      "learning_rate": 1.7342073897497023e-05,
+      "loss": 0.043,
+      "step": 230
+    },
+    {
+      "epoch": 0.13766388557806913,
+      "grad_norm": 4.314394950866699,
+      "learning_rate": 1.733015494636472e-05,
+      "loss": 0.1456,
+      "step": 231
+    },
+    {
+      "epoch": 0.13825983313468415,
+      "grad_norm": 2.9395751953125,
+      "learning_rate": 1.7318235995232422e-05,
+      "loss": 0.0614,
+      "step": 232
+    },
+    {
+      "epoch": 0.13885578069129917,
+      "grad_norm": 2.627901315689087,
+      "learning_rate": 1.730631704410012e-05,
+      "loss": 0.0511,
+      "step": 233
+    },
+    {
+      "epoch": 0.1394517282479142,
+      "grad_norm": 4.270862102508545,
+      "learning_rate": 1.7294398092967822e-05,
+      "loss": 0.2473,
+      "step": 234
+    },
+    {
+      "epoch": 0.1400476758045292,
+      "grad_norm": 8.463306427001953,
+      "learning_rate": 1.728247914183552e-05,
+      "loss": 0.4872,
+      "step": 235
+    },
+    {
+      "epoch": 0.14064362336114422,
+      "grad_norm": 4.905069351196289,
+      "learning_rate": 1.727056019070322e-05,
+      "loss": 0.2226,
+      "step": 236
+    },
+    {
+      "epoch": 0.14123957091775924,
+      "grad_norm": 3.16874098777771,
+      "learning_rate": 1.725864123957092e-05,
+      "loss": 0.1025,
+      "step": 237
+    },
+    {
+      "epoch": 0.14183551847437426,
+      "grad_norm": 2.76073956489563,
+      "learning_rate": 1.724672228843862e-05,
+      "loss": 0.0752,
+      "step": 238
+    },
+    {
+      "epoch": 0.14243146603098927,
+      "grad_norm": 13.412144660949707,
+      "learning_rate": 1.723480333730632e-05,
+      "loss": 1.2795,
+      "step": 239
+    },
+    {
+      "epoch": 0.1430274135876043,
+      "grad_norm": 4.42921781539917,
+      "learning_rate": 1.722288438617402e-05,
+      "loss": 0.2094,
+      "step": 240
+    },
+    {
+      "epoch": 0.1436233611442193,
+      "grad_norm": 5.295729637145996,
+      "learning_rate": 1.7210965435041718e-05,
+      "loss": 0.3328,
+      "step": 241
+    },
+    {
+      "epoch": 0.14421930870083433,
+      "grad_norm": 2.8549704551696777,
+      "learning_rate": 1.719904648390942e-05,
+      "loss": 0.0888,
+      "step": 242
+    },
+    {
+      "epoch": 0.14481525625744934,
+      "grad_norm": 3.3968183994293213,
+      "learning_rate": 1.7187127532777118e-05,
+      "loss": 0.2034,
+      "step": 243
+    },
+    {
+      "epoch": 0.14541120381406436,
+      "grad_norm": 6.6438398361206055,
+      "learning_rate": 1.717520858164482e-05,
+      "loss": 0.0845,
+      "step": 244
+    },
+    {
+      "epoch": 0.14600715137067938,
+      "grad_norm": 7.65512228012085,
+      "learning_rate": 1.7163289630512517e-05,
+      "loss": 0.667,
+      "step": 245
+    },
+    {
+      "epoch": 0.1466030989272944,
+      "grad_norm": 3.6990160942077637,
+      "learning_rate": 1.715137067938022e-05,
+      "loss": 0.1571,
+      "step": 246
+    },
+    {
+      "epoch": 0.1471990464839094,
+      "grad_norm": 5.922319412231445,
+      "learning_rate": 1.7139451728247917e-05,
+      "loss": 0.4082,
+      "step": 247
+    },
+    {
+      "epoch": 0.14779499404052443,
+      "grad_norm": 7.498687267303467,
+      "learning_rate": 1.7127532777115615e-05,
+      "loss": 0.6622,
+      "step": 248
+    },
+    {
+      "epoch": 0.14839094159713945,
+      "grad_norm": 1.7127811908721924,
+      "learning_rate": 1.7115613825983316e-05,
+      "loss": 0.0316,
+      "step": 249
+    },
+    {
+      "epoch": 0.14898688915375446,
+      "grad_norm": 2.6821420192718506,
+      "learning_rate": 1.7103694874851014e-05,
+      "loss": 0.0673,
+      "step": 250
+    },
+    {
+      "epoch": 0.14958283671036948,
+      "grad_norm": 5.412259101867676,
+      "learning_rate": 1.7091775923718716e-05,
+      "loss": 0.4101,
+      "step": 251
+    },
+    {
+      "epoch": 0.1501787842669845,
+      "grad_norm": 2.6168012619018555,
+      "learning_rate": 1.7079856972586414e-05,
+      "loss": 0.0671,
+      "step": 252
+    },
+    {
+      "epoch": 0.15077473182359952,
+      "grad_norm": 5.003539562225342,
+      "learning_rate": 1.7067938021454115e-05,
+      "loss": 0.1967,
+      "step": 253
+    },
+    {
+      "epoch": 0.15137067938021453,
+      "grad_norm": 6.902320384979248,
+      "learning_rate": 1.7056019070321813e-05,
+      "loss": 0.6078,
+      "step": 254
+    },
+    {
+      "epoch": 0.15196662693682955,
+      "grad_norm": 4.822064399719238,
+      "learning_rate": 1.7044100119189514e-05,
+      "loss": 0.2449,
+      "step": 255
+    },
+    {
+      "epoch": 0.15256257449344457,
+      "grad_norm": 8.10243034362793,
+      "learning_rate": 1.7032181168057212e-05,
+      "loss": 0.7341,
+      "step": 256
+    },
+    {
+      "epoch": 0.1531585220500596,
+      "grad_norm": 3.8869712352752686,
+      "learning_rate": 1.702026221692491e-05,
+      "loss": 0.2025,
+      "step": 257
+    },
+    {
+      "epoch": 0.1537544696066746,
+      "grad_norm": 4.3090105056762695,
+      "learning_rate": 1.7008343265792612e-05,
+      "loss": 0.2487,
+      "step": 258
+    },
+    {
+      "epoch": 0.15435041716328962,
+      "grad_norm": 3.206827402114868,
+      "learning_rate": 1.699642431466031e-05,
+      "loss": 0.1675,
+      "step": 259
+    },
+    {
+      "epoch": 0.15494636471990464,
+      "grad_norm": 3.7907142639160156,
+      "learning_rate": 1.698450536352801e-05,
+      "loss": 0.1764,
+      "step": 260
+    },
+    {
+      "epoch": 0.15554231227651966,
+      "grad_norm": 6.753353118896484,
+      "learning_rate": 1.697258641239571e-05,
+      "loss": 0.383,
+      "step": 261
+    },
+    {
+      "epoch": 0.15613825983313467,
+      "grad_norm": 4.451046943664551,
+      "learning_rate": 1.696066746126341e-05,
+      "loss": 0.1984,
+      "step": 262
+    },
+    {
+      "epoch": 0.1567342073897497,
+      "grad_norm": 2.6580910682678223,
+      "learning_rate": 1.694874851013111e-05,
+      "loss": 0.0853,
+      "step": 263
+    },
+    {
+      "epoch": 0.1573301549463647,
+      "grad_norm": 2.880986452102661,
+      "learning_rate": 1.6936829558998807e-05,
+      "loss": 0.0681,
+      "step": 264
+    },
+    {
+      "epoch": 0.15792610250297973,
+      "grad_norm": 6.549500942230225,
+      "learning_rate": 1.692491060786651e-05,
+      "loss": 0.5962,
+      "step": 265
+    },
+    {
+      "epoch": 0.15852205005959474,
+      "grad_norm": 4.922301769256592,
+      "learning_rate": 1.6912991656734206e-05,
+      "loss": 0.1673,
+      "step": 266
+    },
+    {
+      "epoch": 0.15911799761620976,
+      "grad_norm": 3.1822335720062256,
+      "learning_rate": 1.6901072705601908e-05,
+      "loss": 0.1098,
+      "step": 267
+    },
+    {
+      "epoch": 0.15971394517282478,
+      "grad_norm": 3.4277942180633545,
+      "learning_rate": 1.6889153754469606e-05,
+      "loss": 0.102,
+      "step": 268
+    },
+    {
+      "epoch": 0.1603098927294398,
+      "grad_norm": 3.9560163021087646,
+      "learning_rate": 1.6877234803337307e-05,
+      "loss": 0.1196,
+      "step": 269
+    },
+    {
+      "epoch": 0.16090584028605484,
+      "grad_norm": 3.7436366081237793,
+      "learning_rate": 1.6865315852205005e-05,
+      "loss": 0.1042,
+      "step": 270
+    },
+    {
+      "epoch": 0.16150178784266986,
+      "grad_norm": 1.723879337310791,
+      "learning_rate": 1.6853396901072707e-05,
+      "loss": 0.0306,
+      "step": 271
+    },
+    {
+      "epoch": 0.16209773539928488,
+      "grad_norm": 8.534758567810059,
+      "learning_rate": 1.6841477949940405e-05,
+      "loss": 0.8245,
+      "step": 272
+    },
+    {
+      "epoch": 0.1626936829558999,
+      "grad_norm": 4.638415336608887,
+      "learning_rate": 1.6829558998808106e-05,
+      "loss": 0.2451,
+      "step": 273
+    },
+    {
+      "epoch": 0.1632896305125149,
+      "grad_norm": 2.4885685443878174,
+      "learning_rate": 1.6817640047675804e-05,
+      "loss": 0.057,
+      "step": 274
+    },
+    {
+      "epoch": 0.16388557806912993,
+      "grad_norm": 3.576721429824829,
+      "learning_rate": 1.6805721096543506e-05,
+      "loss": 0.1561,
+      "step": 275
+    },
+    {
+      "epoch": 0.16448152562574495,
+      "grad_norm": 5.622363090515137,
+      "learning_rate": 1.6793802145411204e-05,
+      "loss": 0.1761,
+      "step": 276
+    },
+    {
+      "epoch": 0.16507747318235996,
+      "grad_norm": 6.871002197265625,
+      "learning_rate": 1.6781883194278905e-05,
+      "loss": 0.2061,
+      "step": 277
+    },
+    {
+      "epoch": 0.16567342073897498,
+      "grad_norm": 4.65151309967041,
+      "learning_rate": 1.6769964243146603e-05,
+      "loss": 0.161,
+      "step": 278
+    },
+    {
+      "epoch": 0.16626936829559,
+      "grad_norm": 5.821991920471191,
+      "learning_rate": 1.6758045292014305e-05,
+      "loss": 0.0845,
+      "step": 279
+    },
+    {
+      "epoch": 0.16686531585220502,
+      "grad_norm": 3.0261967182159424,
+      "learning_rate": 1.6746126340882003e-05,
+      "loss": 0.0693,
+      "step": 280
+    },
+    {
+      "epoch": 0.16746126340882003,
+      "grad_norm": 3.069511651992798,
+      "learning_rate": 1.6734207389749704e-05,
+      "loss": 0.0841,
+      "step": 281
+    },
+    {
+      "epoch": 0.16805721096543505,
+      "grad_norm": 3.86348557472229,
+      "learning_rate": 1.6722288438617402e-05,
+      "loss": 0.0947,
+      "step": 282
+    },
+    {
+      "epoch": 0.16865315852205007,
+      "grad_norm": 3.1852807998657227,
+      "learning_rate": 1.6710369487485103e-05,
+      "loss": 0.123,
+      "step": 283
+    },
+    {
+      "epoch": 0.16924910607866508,
+      "grad_norm": 2.9607508182525635,
+      "learning_rate": 1.66984505363528e-05,
+      "loss": 0.0469,
+      "step": 284
+    },
+    {
+      "epoch": 0.1698450536352801,
+      "grad_norm": 3.2114453315734863,
+      "learning_rate": 1.6686531585220503e-05,
+      "loss": 0.1238,
+      "step": 285
+    },
+    {
+      "epoch": 0.17044100119189512,
+      "grad_norm": 4.657351016998291,
+      "learning_rate": 1.66746126340882e-05,
+      "loss": 0.2781,
+      "step": 286
+    },
+    {
+      "epoch": 0.17103694874851014,
+      "grad_norm": 3.1530210971832275,
+      "learning_rate": 1.6662693682955902e-05,
+      "loss": 0.1059,
+      "step": 287
+    },
+    {
+      "epoch": 0.17163289630512515,
+      "grad_norm": 6.596444129943848,
+      "learning_rate": 1.66507747318236e-05,
+      "loss": 0.3324,
+      "step": 288
+    },
+    {
+      "epoch": 0.17222884386174017,
+      "grad_norm": 3.261715888977051,
+      "learning_rate": 1.6638855780691302e-05,
+      "loss": 0.0665,
+      "step": 289
+    },
+    {
+      "epoch": 0.1728247914183552,
+      "grad_norm": 5.4474663734436035,
+      "learning_rate": 1.6626936829559e-05,
+      "loss": 0.3848,
+      "step": 290
+    },
+    {
+      "epoch": 0.1734207389749702,
+      "grad_norm": 8.177742958068848,
+      "learning_rate": 1.66150178784267e-05,
+      "loss": 0.6142,
+      "step": 291
+    },
+    {
+      "epoch": 0.17401668653158522,
+      "grad_norm": 5.673367023468018,
+      "learning_rate": 1.66030989272944e-05,
+      "loss": 0.2728,
+      "step": 292
+    },
+    {
+      "epoch": 0.17461263408820024,
+      "grad_norm": 8.51845932006836,
+      "learning_rate": 1.65911799761621e-05,
+      "loss": 0.2913,
+      "step": 293
+    },
+    {
+      "epoch": 0.17520858164481526,
+      "grad_norm": 2.9813098907470703,
+      "learning_rate": 1.65792610250298e-05,
+      "loss": 0.0625,
+      "step": 294
+    },
+    {
+      "epoch": 0.17580452920143028,
+      "grad_norm": 5.829262733459473,
+      "learning_rate": 1.65673420738975e-05,
+      "loss": 0.3867,
+      "step": 295
+    },
+    {
+      "epoch": 0.1764004767580453,
+      "grad_norm": 2.68039870262146,
+      "learning_rate": 1.6555423122765198e-05,
+      "loss": 0.0703,
+      "step": 296
+    },
+    {
+      "epoch": 0.1769964243146603,
+      "grad_norm": 3.08335542678833,
+      "learning_rate": 1.65435041716329e-05,
+      "loss": 0.116,
+      "step": 297
+    },
+    {
+      "epoch": 0.17759237187127533,
+      "grad_norm": 4.591975688934326,
+      "learning_rate": 1.6531585220500598e-05,
+      "loss": 0.2351,
+      "step": 298
+    },
+    {
+      "epoch": 0.17818831942789035,
+      "grad_norm": 6.592544078826904,
+      "learning_rate": 1.6519666269368296e-05,
+      "loss": 0.4242,
+      "step": 299
+    },
+    {
+      "epoch": 0.17878426698450536,
+      "grad_norm": 8.587732315063477,
+      "learning_rate": 1.6507747318235997e-05,
+      "loss": 0.7287,
+      "step": 300
+    },
+    {
+      "epoch": 0.17938021454112038,
+      "grad_norm": 3.927314281463623,
+      "learning_rate": 1.6495828367103695e-05,
+      "loss": 0.1341,
+      "step": 301
+    },
+    {
+      "epoch": 0.1799761620977354,
+      "grad_norm": 3.4898624420166016,
+      "learning_rate": 1.6483909415971397e-05,
+      "loss": 0.0632,
+      "step": 302
+    },
+    {
+      "epoch": 0.18057210965435042,
+      "grad_norm": 6.549103260040283,
+      "learning_rate": 1.6471990464839095e-05,
+      "loss": 0.2933,
+      "step": 303
+    },
+    {
+      "epoch": 0.18116805721096543,
+      "grad_norm": 4.804233551025391,
+      "learning_rate": 1.6460071513706796e-05,
+      "loss": 0.2664,
+      "step": 304
+    },
+    {
+      "epoch": 0.18176400476758045,
+      "grad_norm": 3.087395668029785,
+      "learning_rate": 1.6448152562574494e-05,
+      "loss": 0.1211,
+      "step": 305
+    },
+    {
+      "epoch": 0.18235995232419547,
+      "grad_norm": 1.5339934825897217,
+      "learning_rate": 1.6436233611442192e-05,
+      "loss": 0.0329,
+      "step": 306
+    },
+    {
+      "epoch": 0.18295589988081049,
+      "grad_norm": 6.292341709136963,
+      "learning_rate": 1.6424314660309894e-05,
+      "loss": 0.4439,
+      "step": 307
+    },
+    {
+      "epoch": 0.1835518474374255,
+      "grad_norm": 2.708216667175293,
+      "learning_rate": 1.641239570917759e-05,
+      "loss": 0.0996,
+      "step": 308
+    },
+    {
+      "epoch": 0.18414779499404052,
+      "grad_norm": 2.943902015686035,
+      "learning_rate": 1.6400476758045293e-05,
+      "loss": 0.0601,
+      "step": 309
+    },
+    {
+      "epoch": 0.18474374255065554,
+      "grad_norm": 3.500183582305908,
+      "learning_rate": 1.638855780691299e-05,
+      "loss": 0.1622,
+      "step": 310
+    },
+    {
+      "epoch": 0.18533969010727055,
+      "grad_norm": 3.2134599685668945,
+      "learning_rate": 1.6376638855780693e-05,
+      "loss": 0.0506,
+      "step": 311
+    },
+    {
+      "epoch": 0.18593563766388557,
+      "grad_norm": 3.5828495025634766,
+      "learning_rate": 1.636471990464839e-05,
+      "loss": 0.2037,
+      "step": 312
+    },
+    {
+      "epoch": 0.1865315852205006,
+      "grad_norm": 3.4499359130859375,
+      "learning_rate": 1.6352800953516092e-05,
+      "loss": 0.0645,
+      "step": 313
+    },
+    {
+      "epoch": 0.1871275327771156,
+      "grad_norm": 2.617431402206421,
+      "learning_rate": 1.634088200238379e-05,
+      "loss": 0.0577,
+      "step": 314
+    },
+    {
+      "epoch": 0.18772348033373062,
+      "grad_norm": 3.131751537322998,
+      "learning_rate": 1.632896305125149e-05,
+      "loss": 0.0587,
+      "step": 315
+    },
+    {
+      "epoch": 0.18831942789034564,
+      "grad_norm": 3.507444143295288,
+      "learning_rate": 1.631704410011919e-05,
+      "loss": 0.1607,
+      "step": 316
+    },
+    {
+      "epoch": 0.18891537544696066,
+      "grad_norm": 3.848618984222412,
+      "learning_rate": 1.630512514898689e-05,
+      "loss": 0.1019,
+      "step": 317
+    },
+    {
+      "epoch": 0.18951132300357568,
+      "grad_norm": 2.550381660461426,
+      "learning_rate": 1.629320619785459e-05,
+      "loss": 0.0925,
+      "step": 318
+    },
+    {
+      "epoch": 0.1901072705601907,
+      "grad_norm": 7.010753154754639,
+      "learning_rate": 1.628128724672229e-05,
+      "loss": 0.5568,
+      "step": 319
+    },
+    {
+      "epoch": 0.1907032181168057,
+      "grad_norm": 5.270625114440918,
+      "learning_rate": 1.626936829558999e-05,
+      "loss": 0.2054,
+      "step": 320
+    },
+    {
+      "epoch": 0.19129916567342073,
+      "grad_norm": 5.401447296142578,
+      "learning_rate": 1.625744934445769e-05,
+      "loss": 0.2385,
+      "step": 321
+    },
+    {
+      "epoch": 0.19189511323003575,
+      "grad_norm": 3.1718950271606445,
+      "learning_rate": 1.6245530393325388e-05,
+      "loss": 0.1073,
+      "step": 322
+    },
+    {
+      "epoch": 0.19249106078665076,
+      "grad_norm": 2.0048439502716064,
+      "learning_rate": 1.623361144219309e-05,
+      "loss": 0.023,
+      "step": 323
+    },
+    {
+      "epoch": 0.19308700834326578,
+      "grad_norm": 1.6031941175460815,
+      "learning_rate": 1.6221692491060787e-05,
+      "loss": 0.0332,
+      "step": 324
+    },
+    {
+      "epoch": 0.1936829558998808,
+      "grad_norm": 3.8607981204986572,
+      "learning_rate": 1.620977353992849e-05,
+      "loss": 0.1854,
+      "step": 325
+    },
+    {
+      "epoch": 0.19427890345649582,
+      "grad_norm": 4.986361980438232,
+      "learning_rate": 1.6197854588796187e-05,
+      "loss": 0.35,
+      "step": 326
+    },
+    {
+      "epoch": 0.19487485101311083,
+      "grad_norm": 4.2986159324646,
+      "learning_rate": 1.6185935637663888e-05,
+      "loss": 0.2181,
+      "step": 327
+    },
+    {
+      "epoch": 0.19547079856972585,
+      "grad_norm": 1.7152146100997925,
+      "learning_rate": 1.6174016686531586e-05,
+      "loss": 0.0208,
+      "step": 328
+    },
+    {
+      "epoch": 0.19606674612634087,
+      "grad_norm": 5.286261558532715,
+      "learning_rate": 1.6162097735399288e-05,
+      "loss": 0.2484,
+      "step": 329
+    },
+    {
+      "epoch": 0.1966626936829559,
+      "grad_norm": 2.809368133544922,
+      "learning_rate": 1.6150178784266986e-05,
+      "loss": 0.0878,
+      "step": 330
+    },
+    {
+      "epoch": 0.19725864123957093,
+      "grad_norm": 2.3276145458221436,
+      "learning_rate": 1.6138259833134687e-05,
+      "loss": 0.0673,
+      "step": 331
+    },
+    {
+      "epoch": 0.19785458879618595,
+      "grad_norm": 7.617786884307861,
+      "learning_rate": 1.6126340882002385e-05,
+      "loss": 0.6302,
+      "step": 332
+    },
+    {
+      "epoch": 0.19845053635280097,
+      "grad_norm": 4.943353652954102,
+      "learning_rate": 1.6114421930870087e-05,
+      "loss": 0.3376,
+      "step": 333
+    },
+    {
+      "epoch": 0.19904648390941598,
+      "grad_norm": 5.739307403564453,
+      "learning_rate": 1.6102502979737785e-05,
+      "loss": 0.2804,
+      "step": 334
+    },
+    {
+      "epoch": 0.199642431466031,
+      "grad_norm": 4.681702613830566,
+      "learning_rate": 1.6090584028605486e-05,
+      "loss": 0.2931,
+      "step": 335
+    },
+    {
+      "epoch": 0.20023837902264602,
+      "grad_norm": 6.099559783935547,
+      "learning_rate": 1.6078665077473184e-05,
+      "loss": 0.1712,
+      "step": 336
+    },
+    {
+      "epoch": 0.20083432657926104,
+      "grad_norm": 2.805178642272949,
+      "learning_rate": 1.6066746126340886e-05,
+      "loss": 0.1238,
+      "step": 337
+    },
+    {
+      "epoch": 0.20143027413587605,
+      "grad_norm": 4.062296390533447,
+      "learning_rate": 1.6054827175208584e-05,
+      "loss": 0.216,
+      "step": 338
+    },
+    {
+      "epoch": 0.20202622169249107,
+      "grad_norm": 2.606431245803833,
+      "learning_rate": 1.6042908224076285e-05,
+      "loss": 0.0715,
+      "step": 339
+    },
+    {
+      "epoch": 0.2026221692491061,
+      "grad_norm": 4.935174465179443,
+      "learning_rate": 1.6030989272943983e-05,
+      "loss": 0.1455,
+      "step": 340
+    },
+    {
+      "epoch": 0.2032181168057211,
+      "grad_norm": 5.954669952392578,
+      "learning_rate": 1.601907032181168e-05,
+      "loss": 0.359,
+      "step": 341
+    },
+    {
+      "epoch": 0.20381406436233612,
+      "grad_norm": 3.1365551948547363,
+      "learning_rate": 1.6007151370679382e-05,
+      "loss": 0.0711,
+      "step": 342
+    },
+    {
+      "epoch": 0.20441001191895114,
+      "grad_norm": 3.0025413036346436,
+      "learning_rate": 1.599523241954708e-05,
+      "loss": 0.0733,
+      "step": 343
+    },
+    {
+      "epoch": 0.20500595947556616,
+      "grad_norm": 1.2385191917419434,
+      "learning_rate": 1.5983313468414782e-05,
+      "loss": 0.0349,
+      "step": 344
+    },
+    {
+      "epoch": 0.20560190703218117,
+      "grad_norm": 2.8543715476989746,
+      "learning_rate": 1.597139451728248e-05,
+      "loss": 0.0283,
+      "step": 345
+    },
+    {
+      "epoch": 0.2061978545887962,
+      "grad_norm": 5.443460464477539,
+      "learning_rate": 1.595947556615018e-05,
+      "loss": 0.3397,
+      "step": 346
+    },
+    {
+      "epoch": 0.2067938021454112,
+      "grad_norm": 5.841041564941406,
+      "learning_rate": 1.594755661501788e-05,
+      "loss": 0.1587,
+      "step": 347
+    },
+    {
+      "epoch": 0.20738974970202623,
+      "grad_norm": 3.164992570877075,
+      "learning_rate": 1.593563766388558e-05,
+      "loss": 0.1204,
+      "step": 348
+    },
+    {
+      "epoch": 0.20798569725864124,
+      "grad_norm": 5.2390923500061035,
+      "learning_rate": 1.592371871275328e-05,
+      "loss": 0.1196,
+      "step": 349
+    },
+    {
+      "epoch": 0.20858164481525626,
+      "grad_norm": 2.1591460704803467,
+      "learning_rate": 1.5911799761620977e-05,
+      "loss": 0.0946,
+      "step": 350
+    },
+    {
+      "epoch": 0.20917759237187128,
+      "grad_norm": 2.6391801834106445,
+      "learning_rate": 1.589988081048868e-05,
+      "loss": 0.0284,
+      "step": 351
+    },
+    {
+      "epoch": 0.2097735399284863,
+      "grad_norm": 1.8103128671646118,
+      "learning_rate": 1.5887961859356376e-05,
+      "loss": 0.0452,
+      "step": 352
+    },
+    {
+      "epoch": 0.21036948748510131,
+      "grad_norm": 4.232999801635742,
+      "learning_rate": 1.5876042908224078e-05,
+      "loss": 0.1085,
+      "step": 353
+    },
+    {
+      "epoch": 0.21096543504171633,
+      "grad_norm": 3.733548164367676,
+      "learning_rate": 1.5864123957091776e-05,
+      "loss": 0.1124,
+      "step": 354
+    },
+    {
+      "epoch": 0.21156138259833135,
+      "grad_norm": 3.6490635871887207,
+      "learning_rate": 1.5852205005959477e-05,
+      "loss": 0.1191,
+      "step": 355
+    },
+    {
+      "epoch": 0.21215733015494637,
+      "grad_norm": 1.9155492782592773,
+      "learning_rate": 1.5840286054827175e-05,
+      "loss": 0.0189,
+      "step": 356
+    },
+    {
+      "epoch": 0.21275327771156138,
+      "grad_norm": 4.024199485778809,
+      "learning_rate": 1.5828367103694877e-05,
+      "loss": 0.1968,
+      "step": 357
+    },
+    {
+      "epoch": 0.2133492252681764,
+      "grad_norm": 2.118933916091919,
+      "learning_rate": 1.5816448152562575e-05,
+      "loss": 0.0337,
+      "step": 358
+    },
+    {
+      "epoch": 0.21394517282479142,
+      "grad_norm": 2.5160293579101562,
+      "learning_rate": 1.5804529201430276e-05,
+      "loss": 0.0572,
+      "step": 359
+    },
+    {
+      "epoch": 0.21454112038140644,
+      "grad_norm": 2.6285560131073,
+      "learning_rate": 1.5792610250297974e-05,
+      "loss": 0.0985,
+      "step": 360
+    },
+    {
+      "epoch": 0.21513706793802145,
+      "grad_norm": 4.790384769439697,
+      "learning_rate": 1.5780691299165676e-05,
+      "loss": 0.2031,
+      "step": 361
+    },
+    {
+      "epoch": 0.21573301549463647,
+      "grad_norm": 5.043698787689209,
+      "learning_rate": 1.5768772348033374e-05,
+      "loss": 0.102,
+      "step": 362
+    },
+    {
+      "epoch": 0.2163289630512515,
+      "grad_norm": 3.676065444946289,
+      "learning_rate": 1.5756853396901075e-05,
+      "loss": 0.064,
+      "step": 363
+    },
+    {
+      "epoch": 0.2169249106078665,
+      "grad_norm": 4.249757766723633,
+      "learning_rate": 1.5744934445768773e-05,
+      "loss": 0.1559,
+      "step": 364
+    },
+    {
+      "epoch": 0.21752085816448152,
+      "grad_norm": 1.6513293981552124,
+      "learning_rate": 1.5733015494636475e-05,
+      "loss": 0.0179,
+      "step": 365
+    },
+    {
+      "epoch": 0.21811680572109654,
+      "grad_norm": 5.160050392150879,
+      "learning_rate": 1.5721096543504173e-05,
+      "loss": 0.2222,
+      "step": 366
+    },
+    {
+      "epoch": 0.21871275327771156,
+      "grad_norm": 5.062220573425293,
+      "learning_rate": 1.570917759237187e-05,
+      "loss": 0.2263,
+      "step": 367
+    },
+    {
+      "epoch": 0.21930870083432658,
+      "grad_norm": 7.087697505950928,
+      "learning_rate": 1.5697258641239572e-05,
+      "loss": 0.3948,
+      "step": 368
+    },
+    {
+      "epoch": 0.2199046483909416,
+      "grad_norm": 3.613529682159424,
+      "learning_rate": 1.568533969010727e-05,
+      "loss": 0.0595,
+      "step": 369
+    },
+    {
+      "epoch": 0.2205005959475566,
+      "grad_norm": 2.503635883331299,
+      "learning_rate": 1.567342073897497e-05,
+      "loss": 0.0432,
+      "step": 370
+    },
+    {
+      "epoch": 0.22109654350417163,
+      "grad_norm": 4.606808662414551,
+      "learning_rate": 1.566150178784267e-05,
+      "loss": 0.1261,
+      "step": 371
+    },
+    {
+      "epoch": 0.22169249106078665,
+      "grad_norm": 4.654660224914551,
+      "learning_rate": 1.564958283671037e-05,
+      "loss": 0.2559,
+      "step": 372
+    },
+    {
+      "epoch": 0.22228843861740166,
+      "grad_norm": 3.2481532096862793,
+      "learning_rate": 1.563766388557807e-05,
+      "loss": 0.0853,
+      "step": 373
+    },
+    {
+      "epoch": 0.22288438617401668,
+      "grad_norm": 3.7485756874084473,
+      "learning_rate": 1.562574493444577e-05,
+      "loss": 0.1417,
+      "step": 374
+    },
+    {
+      "epoch": 0.2234803337306317,
+      "grad_norm": 3.3564679622650146,
+      "learning_rate": 1.561382598331347e-05,
+      "loss": 0.0816,
+      "step": 375
+    },
+    {
+      "epoch": 0.22407628128724671,
+      "grad_norm": 3.2078781127929688,
+      "learning_rate": 1.560190703218117e-05,
+      "loss": 0.1494,
+      "step": 376
+    },
+    {
+      "epoch": 0.22467222884386173,
+      "grad_norm": 2.247511148452759,
+      "learning_rate": 1.5589988081048868e-05,
+      "loss": 0.0497,
+      "step": 377
+    },
+    {
+      "epoch": 0.22526817640047675,
+      "grad_norm": 7.090639114379883,
+      "learning_rate": 1.557806912991657e-05,
+      "loss": 0.3702,
+      "step": 378
+    },
+    {
+      "epoch": 0.22586412395709177,
+      "grad_norm": 3.4250786304473877,
+      "learning_rate": 1.5566150178784267e-05,
+      "loss": 0.0874,
+      "step": 379
+    },
+    {
+      "epoch": 0.22646007151370678,
+      "grad_norm": 9.891451835632324,
+      "learning_rate": 1.555423122765197e-05,
+      "loss": 0.814,
+      "step": 380
+    },
+    {
+      "epoch": 0.2270560190703218,
+      "grad_norm": 4.847662925720215,
+      "learning_rate": 1.5542312276519667e-05,
+      "loss": 0.2187,
+      "step": 381
+    },
+    {
+      "epoch": 0.22765196662693682,
+      "grad_norm": 2.520387649536133,
+      "learning_rate": 1.553039332538737e-05,
+      "loss": 0.0561,
+      "step": 382
+    },
+    {
+      "epoch": 0.22824791418355184,
+      "grad_norm": 2.512922525405884,
+      "learning_rate": 1.5518474374255066e-05,
+      "loss": 0.0449,
+      "step": 383
+    },
+    {
+      "epoch": 0.22884386174016685,
+      "grad_norm": 2.806347370147705,
+      "learning_rate": 1.5506555423122768e-05,
+      "loss": 0.0598,
+      "step": 384
+    },
+    {
+      "epoch": 0.22943980929678187,
+      "grad_norm": 2.7087254524230957,
+      "learning_rate": 1.5494636471990466e-05,
+      "loss": 0.0741,
+      "step": 385
+    },
+    {
+      "epoch": 0.2300357568533969,
+      "grad_norm": 3.98553466796875,
+      "learning_rate": 1.5482717520858167e-05,
+      "loss": 0.09,
+      "step": 386
+    },
+    {
+      "epoch": 0.2306317044100119,
+      "grad_norm": 6.685116767883301,
+      "learning_rate": 1.5470798569725865e-05,
+      "loss": 0.5057,
+      "step": 387
+    },
+    {
+      "epoch": 0.23122765196662692,
+      "grad_norm": 4.080874919891357,
+      "learning_rate": 1.5458879618593567e-05,
+      "loss": 0.1506,
+      "step": 388
+    },
+    {
+      "epoch": 0.23182359952324194,
+      "grad_norm": 4.778779029846191,
+      "learning_rate": 1.5446960667461265e-05,
+      "loss": 0.2171,
+      "step": 389
+    },
+    {
+      "epoch": 0.232419547079857,
+      "grad_norm": 2.3364665508270264,
+      "learning_rate": 1.5435041716328966e-05,
+      "loss": 0.0443,
+      "step": 390
+    },
+    {
+      "epoch": 0.233015494636472,
+      "grad_norm": 2.1844589710235596,
+      "learning_rate": 1.5423122765196664e-05,
+      "loss": 0.0659,
+      "step": 391
+    },
+    {
+      "epoch": 0.23361144219308702,
+      "grad_norm": 2.4924731254577637,
+      "learning_rate": 1.5411203814064362e-05,
+      "loss": 0.0607,
+      "step": 392
+    },
+    {
+      "epoch": 0.23420738974970204,
+      "grad_norm": 2.9716289043426514,
+      "learning_rate": 1.5399284862932064e-05,
+      "loss": 0.0641,
+      "step": 393
+    },
+    {
+      "epoch": 0.23480333730631706,
+      "grad_norm": 5.301799297332764,
+      "learning_rate": 1.538736591179976e-05,
+      "loss": 0.2746,
+      "step": 394
+    },
+    {
+      "epoch": 0.23539928486293207,
+      "grad_norm": 3.8295562267303467,
+      "learning_rate": 1.5375446960667463e-05,
+      "loss": 0.1274,
+      "step": 395
+    },
+    {
+      "epoch": 0.2359952324195471,
+      "grad_norm": 2.7859420776367188,
+      "learning_rate": 1.536352800953516e-05,
+      "loss": 0.0879,
+      "step": 396
+    },
+    {
+      "epoch": 0.2365911799761621,
+      "grad_norm": 6.524483680725098,
+      "learning_rate": 1.5351609058402863e-05,
+      "loss": 0.4559,
+      "step": 397
+    },
+    {
+      "epoch": 0.23718712753277713,
+      "grad_norm": 4.256045818328857,
+      "learning_rate": 1.533969010727056e-05,
+      "loss": 0.3722,
+      "step": 398
+    },
+    {
+      "epoch": 0.23778307508939214,
+      "grad_norm": 2.237961769104004,
+      "learning_rate": 1.532777115613826e-05,
+      "loss": 0.0371,
+      "step": 399
+    },
+    {
+      "epoch": 0.23837902264600716,
+      "grad_norm": 3.8318328857421875,
+      "learning_rate": 1.531585220500596e-05,
+      "loss": 0.2026,
+      "step": 400
+    },
+    {
+      "epoch": 0.23897497020262218,
+      "grad_norm": 5.003171443939209,
+      "learning_rate": 1.5303933253873658e-05,
+      "loss": 0.1505,
+      "step": 401
+    },
+    {
+      "epoch": 0.2395709177592372,
+      "grad_norm": 1.857621431350708,
+      "learning_rate": 1.529201430274136e-05,
+      "loss": 0.0397,
+      "step": 402
+    },
+    {
+      "epoch": 0.2401668653158522,
+      "grad_norm": 1.6185299158096313,
+      "learning_rate": 1.5280095351609058e-05,
+      "loss": 0.0312,
+      "step": 403
+    },
+    {
+      "epoch": 0.24076281287246723,
+      "grad_norm": 4.245564937591553,
+      "learning_rate": 1.526817640047676e-05,
+      "loss": 0.1104,
+      "step": 404
+    },
+    {
+      "epoch": 0.24135876042908225,
+      "grad_norm": 2.735276699066162,
+      "learning_rate": 1.5256257449344459e-05,
+      "loss": 0.0386,
+      "step": 405
+    },
+    {
+      "epoch": 0.24195470798569726,
+      "grad_norm": 2.736358165740967,
+      "learning_rate": 1.5244338498212158e-05,
+      "loss": 0.0392,
+      "step": 406
+    },
+    {
+      "epoch": 0.24255065554231228,
+      "grad_norm": 3.210637092590332,
+      "learning_rate": 1.5232419547079858e-05,
+      "loss": 0.0713,
+      "step": 407
+    },
+    {
+      "epoch": 0.2431466030989273,
+      "grad_norm": 2.781184673309326,
+      "learning_rate": 1.5220500595947558e-05,
+      "loss": 0.0492,
+      "step": 408
+    },
+    {
+      "epoch": 0.24374255065554232,
+      "grad_norm": 4.172142028808594,
+      "learning_rate": 1.5208581644815256e-05,
+      "loss": 0.2239,
+      "step": 409
+    },
+    {
+      "epoch": 0.24433849821215733,
+      "grad_norm": 3.2288448810577393,
+      "learning_rate": 1.5196662693682957e-05,
+      "loss": 0.1304,
+      "step": 410
+    },
+    {
+      "epoch": 0.24493444576877235,
+      "grad_norm": 3.3766136169433594,
+      "learning_rate": 1.5184743742550655e-05,
+      "loss": 0.2009,
+      "step": 411
+    },
+    {
+      "epoch": 0.24553039332538737,
+      "grad_norm": 4.9276442527771,
+      "learning_rate": 1.5172824791418357e-05,
+      "loss": 0.1941,
+      "step": 412
+    },
+    {
+      "epoch": 0.2461263408820024,
+      "grad_norm": 4.262429714202881,
+      "learning_rate": 1.5160905840286055e-05,
+      "loss": 0.1838,
+      "step": 413
+    },
+    {
+      "epoch": 0.2467222884386174,
+      "grad_norm": 3.2646408081054688,
+      "learning_rate": 1.5148986889153756e-05,
+      "loss": 0.09,
+      "step": 414
+    },
+    {
+      "epoch": 0.24731823599523242,
+      "grad_norm": 3.8027186393737793,
+      "learning_rate": 1.5137067938021454e-05,
+      "loss": 0.0614,
+      "step": 415
+    },
+    {
+      "epoch": 0.24791418355184744,
+      "grad_norm": 3.8104536533355713,
+      "learning_rate": 1.5125148986889156e-05,
+      "loss": 0.1438,
+      "step": 416
+    },
+    {
+      "epoch": 0.24851013110846246,
+      "grad_norm": 4.466651439666748,
+      "learning_rate": 1.5113230035756854e-05,
+      "loss": 0.1947,
+      "step": 417
+    },
+    {
+      "epoch": 0.24910607866507747,
+      "grad_norm": 4.47767448425293,
+      "learning_rate": 1.5101311084624555e-05,
+      "loss": 0.2192,
+      "step": 418
+    },
+    {
+      "epoch": 0.2497020262216925,
+      "grad_norm": 4.857601165771484,
+      "learning_rate": 1.5089392133492253e-05,
+      "loss": 0.2937,
+      "step": 419
+    },
+    {
+      "epoch": 0.25029797377830754,
+      "grad_norm": 7.460949420928955,
+      "learning_rate": 1.5077473182359955e-05,
+      "loss": 0.5518,
+      "step": 420
+    },
+    {
+      "epoch": 0.25089392133492255,
+      "grad_norm": 3.0366694927215576,
+      "learning_rate": 1.5065554231227653e-05,
+      "loss": 0.0382,
+      "step": 421
+    },
+    {
+      "epoch": 0.25148986889153757,
+      "grad_norm": 4.94278621673584,
+      "learning_rate": 1.5053635280095354e-05,
+      "loss": 0.1588,
+      "step": 422
+    },
+    {
+      "epoch": 0.2520858164481526,
+      "grad_norm": 5.939173698425293,
+      "learning_rate": 1.5041716328963052e-05,
+      "loss": 0.238,
+      "step": 423
+    },
+    {
+      "epoch": 0.2526817640047676,
+      "grad_norm": 2.0417304039001465,
+      "learning_rate": 1.5029797377830752e-05,
+      "loss": 0.0349,
+      "step": 424
+    },
+    {
+      "epoch": 0.2532777115613826,
+      "grad_norm": 4.990805149078369,
+      "learning_rate": 1.5017878426698452e-05,
+      "loss": 0.2312,
+      "step": 425
+    },
+    {
+      "epoch": 0.25387365911799764,
+      "grad_norm": 3.167367935180664,
+      "learning_rate": 1.5005959475566151e-05,
+      "loss": 0.0806,
+      "step": 426
+    },
+    {
+      "epoch": 0.25446960667461266,
+      "grad_norm": 3.880845308303833,
+      "learning_rate": 1.4994040524433851e-05,
+      "loss": 0.088,
+      "step": 427
+    },
+    {
+      "epoch": 0.2550655542312277,
+      "grad_norm": 3.7935373783111572,
+      "learning_rate": 1.498212157330155e-05,
+      "loss": 0.2711,
+      "step": 428
+    },
+    {
+      "epoch": 0.2556615017878427,
+      "grad_norm": 3.723200798034668,
+      "learning_rate": 1.497020262216925e-05,
+      "loss": 0.0728,
+      "step": 429
+    },
+    {
+      "epoch": 0.2562574493444577,
+      "grad_norm": 2.968971014022827,
+      "learning_rate": 1.495828367103695e-05,
+      "loss": 0.0862,
+      "step": 430
+    },
+    {
+      "epoch": 0.25685339690107273,
+      "grad_norm": 4.175726890563965,
+      "learning_rate": 1.4946364719904648e-05,
+      "loss": 0.3469,
+      "step": 431
+    },
+    {
+      "epoch": 0.25744934445768775,
+      "grad_norm": 2.940873622894287,
+      "learning_rate": 1.493444576877235e-05,
+      "loss": 0.1179,
+      "step": 432
+    },
+    {
+      "epoch": 0.25804529201430276,
+      "grad_norm": 3.014033794403076,
+      "learning_rate": 1.4922526817640048e-05,
+      "loss": 0.0745,
+      "step": 433
+    },
+    {
+      "epoch": 0.2586412395709178,
+      "grad_norm": 2.1003010272979736,
+      "learning_rate": 1.491060786650775e-05,
+      "loss": 0.0283,
+      "step": 434
+    },
+    {
+      "epoch": 0.2592371871275328,
+      "grad_norm": 3.102855682373047,
+      "learning_rate": 1.4898688915375447e-05,
+      "loss": 0.1083,
+      "step": 435
+    },
+    {
+      "epoch": 0.2598331346841478,
+      "grad_norm": 3.0520870685577393,
+      "learning_rate": 1.4886769964243149e-05,
+      "loss": 0.065,
+      "step": 436
+    },
+    {
+      "epoch": 0.26042908224076283,
+      "grad_norm": 3.2476065158843994,
+      "learning_rate": 1.4874851013110847e-05,
+      "loss": 0.0585,
+      "step": 437
+    },
+    {
+      "epoch": 0.26102502979737785,
+      "grad_norm": 5.285493850708008,
+      "learning_rate": 1.4862932061978548e-05,
+      "loss": 0.4336,
+      "step": 438
+    },
+    {
+      "epoch": 0.26162097735399287,
+      "grad_norm": 3.071126699447632,
+      "learning_rate": 1.4851013110846246e-05,
+      "loss": 0.0944,
+      "step": 439
+    },
+    {
+      "epoch": 0.2622169249106079,
+      "grad_norm": 2.484213352203369,
+      "learning_rate": 1.4839094159713948e-05,
+      "loss": 0.0418,
+      "step": 440
+    },
+    {
+      "epoch": 0.2628128724672229,
+      "grad_norm": 3.0895721912384033,
+      "learning_rate": 1.4827175208581646e-05,
+      "loss": 0.0706,
+      "step": 441
+    },
+    {
+      "epoch": 0.2634088200238379,
+      "grad_norm": 2.587733268737793,
+      "learning_rate": 1.4815256257449347e-05,
+      "loss": 0.0903,
+      "step": 442
+    },
+    {
+      "epoch": 0.26400476758045294,
+      "grad_norm": 3.7364773750305176,
+      "learning_rate": 1.4803337306317045e-05,
+      "loss": 0.1652,
+      "step": 443
+    },
+    {
+      "epoch": 0.26460071513706795,
+      "grad_norm": 3.819730520248413,
+      "learning_rate": 1.4791418355184747e-05,
+      "loss": 0.1682,
+      "step": 444
+    },
+    {
+      "epoch": 0.26519666269368297,
+      "grad_norm": 2.5532383918762207,
+      "learning_rate": 1.4779499404052445e-05,
+      "loss": 0.0416,
+      "step": 445
+    },
+    {
+      "epoch": 0.265792610250298,
+      "grad_norm": 3.5387840270996094,
+      "learning_rate": 1.4767580452920144e-05,
+      "loss": 0.0779,
+      "step": 446
+    },
+    {
+      "epoch": 0.266388557806913,
+      "grad_norm": 2.9245944023132324,
+      "learning_rate": 1.4755661501787844e-05,
+      "loss": 0.0865,
+      "step": 447
+    },
+    {
+      "epoch": 0.266984505363528,
+      "grad_norm": 1.8668475151062012,
+      "learning_rate": 1.4743742550655544e-05,
+      "loss": 0.0561,
+      "step": 448
+    },
+    {
+      "epoch": 0.26758045292014304,
+      "grad_norm": 3.8985657691955566,
+      "learning_rate": 1.4731823599523243e-05,
+      "loss": 0.1902,
+      "step": 449
+    },
+    {
+      "epoch": 0.26817640047675806,
+      "grad_norm": 4.478765964508057,
+      "learning_rate": 1.4719904648390943e-05,
+      "loss": 0.1871,
+      "step": 450
+    },
+    {
+      "epoch": 0.2687723480333731,
+      "grad_norm": 3.5976247787475586,
+      "learning_rate": 1.4707985697258643e-05,
+      "loss": 0.2091,
+      "step": 451
+    },
+    {
+      "epoch": 0.2693682955899881,
+      "grad_norm": 5.108302116394043,
+      "learning_rate": 1.4696066746126343e-05,
+      "loss": 0.0545,
+      "step": 452
+    },
+    {
+      "epoch": 0.2699642431466031,
+      "grad_norm": 3.5877585411071777,
+      "learning_rate": 1.468414779499404e-05,
+      "loss": 0.0413,
+      "step": 453
+    },
+    {
+      "epoch": 0.27056019070321813,
+      "grad_norm": 2.6002588272094727,
+      "learning_rate": 1.4672228843861742e-05,
+      "loss": 0.1083,
+      "step": 454
+    },
+    {
+      "epoch": 0.27115613825983315,
+      "grad_norm": 2.5038881301879883,
+      "learning_rate": 1.466030989272944e-05,
+      "loss": 0.044,
+      "step": 455
+    },
+    {
+      "epoch": 0.27175208581644816,
+      "grad_norm": 5.461592197418213,
+      "learning_rate": 1.4648390941597142e-05,
+      "loss": 0.4276,
+      "step": 456
+    },
+    {
+      "epoch": 0.2723480333730632,
+      "grad_norm": 4.955808639526367,
+      "learning_rate": 1.463647199046484e-05,
+      "loss": 0.3051,
+      "step": 457
+    },
+    {
+      "epoch": 0.2729439809296782,
+      "grad_norm": 4.2490410804748535,
+      "learning_rate": 1.4624553039332541e-05,
+      "loss": 0.2695,
+      "step": 458
+    },
+    {
+      "epoch": 0.2735399284862932,
+      "grad_norm": 7.013411998748779,
+      "learning_rate": 1.4612634088200239e-05,
+      "loss": 0.5958,
+      "step": 459
+    },
+    {
+      "epoch": 0.27413587604290823,
+      "grad_norm": 7.976939678192139,
+      "learning_rate": 1.460071513706794e-05,
+      "loss": 0.7984,
+      "step": 460
+    },
+    {
+      "epoch": 0.27473182359952325,
+      "grad_norm": 1.7881031036376953,
+      "learning_rate": 1.4588796185935639e-05,
+      "loss": 0.0449,
+      "step": 461
+    },
+    {
+      "epoch": 0.27532777115613827,
+      "grad_norm": 2.7737040519714355,
+      "learning_rate": 1.457687723480334e-05,
+      "loss": 0.0795,
+      "step": 462
+    },
+    {
+      "epoch": 0.2759237187127533,
+      "grad_norm": 1.8559626340866089,
+      "learning_rate": 1.4564958283671038e-05,
+      "loss": 0.0309,
+      "step": 463
+    },
+    {
+      "epoch": 0.2765196662693683,
+      "grad_norm": 6.5311360359191895,
+      "learning_rate": 1.455303933253874e-05,
+      "loss": 0.4182,
+      "step": 464
+    },
+    {
+      "epoch": 0.2771156138259833,
+      "grad_norm": 2.4511115550994873,
+      "learning_rate": 1.4541120381406437e-05,
+      "loss": 0.0571,
+      "step": 465
+    },
+    {
+      "epoch": 0.27771156138259834,
+      "grad_norm": 3.4907400608062744,
+      "learning_rate": 1.4529201430274137e-05,
+      "loss": 0.1101,
+      "step": 466
+    },
+    {
+      "epoch": 0.27830750893921335,
+      "grad_norm": 2.767814874649048,
+      "learning_rate": 1.4517282479141837e-05,
+      "loss": 0.0653,
+      "step": 467
+    },
+    {
+      "epoch": 0.2789034564958284,
+      "grad_norm": 3.12827730178833,
+      "learning_rate": 1.4505363528009537e-05,
+      "loss": 0.0854,
+      "step": 468
+    },
+    {
+      "epoch": 0.2794994040524434,
+      "grad_norm": 2.629626989364624,
+      "learning_rate": 1.4493444576877236e-05,
+      "loss": 0.044,
+      "step": 469
+    },
+    {
+      "epoch": 0.2800953516090584,
+      "grad_norm": 6.199581623077393,
+      "learning_rate": 1.4481525625744936e-05,
+      "loss": 0.5164,
+      "step": 470
+    },
+    {
+      "epoch": 0.2806912991656734,
+      "grad_norm": 8.337032318115234,
+      "learning_rate": 1.4469606674612636e-05,
+      "loss": 0.6672,
+      "step": 471
+    },
+    {
+      "epoch": 0.28128724672228844,
+      "grad_norm": 1.8186225891113281,
+      "learning_rate": 1.4457687723480336e-05,
+      "loss": 0.0588,
+      "step": 472
+    },
+    {
+      "epoch": 0.28188319427890346,
+      "grad_norm": 6.147401809692383,
+      "learning_rate": 1.4445768772348034e-05,
+      "loss": 0.4751,
+      "step": 473
+    },
+    {
+      "epoch": 0.2824791418355185,
+      "grad_norm": 2.7147891521453857,
+      "learning_rate": 1.4433849821215735e-05,
+      "loss": 0.0744,
+      "step": 474
+    },
+    {
+      "epoch": 0.2830750893921335,
+      "grad_norm": 8.214629173278809,
+      "learning_rate": 1.4421930870083433e-05,
+      "loss": 0.5454,
+      "step": 475
+    },
+    {
+      "epoch": 0.2836710369487485,
+      "grad_norm": 2.9637773036956787,
+      "learning_rate": 1.4410011918951134e-05,
+      "loss": 0.0718,
+      "step": 476
+    },
+    {
+      "epoch": 0.28426698450536353,
+      "grad_norm": 4.236182689666748,
+      "learning_rate": 1.4398092967818833e-05,
+      "loss": 0.0474,
+      "step": 477
+    },
+    {
+      "epoch": 0.28486293206197855,
+      "grad_norm": 4.56669807434082,
+      "learning_rate": 1.4386174016686534e-05,
+      "loss": 0.3349,
+      "step": 478
+    },
+    {
+      "epoch": 0.28545887961859356,
+      "grad_norm": 3.86352276802063,
+      "learning_rate": 1.4374255065554232e-05,
+      "loss": 0.2356,
+      "step": 479
+    },
+    {
+      "epoch": 0.2860548271752086,
+      "grad_norm": 3.2093992233276367,
+      "learning_rate": 1.4362336114421933e-05,
+      "loss": 0.0807,
+      "step": 480
+    },
+    {
+      "epoch": 0.2866507747318236,
+      "grad_norm": 2.615388870239258,
+      "learning_rate": 1.4350417163289631e-05,
+      "loss": 0.0408,
+      "step": 481
+    },
+    {
+      "epoch": 0.2872467222884386,
+      "grad_norm": 2.412203073501587,
+      "learning_rate": 1.4338498212157333e-05,
+      "loss": 0.0385,
+      "step": 482
+    },
+    {
+      "epoch": 0.28784266984505363,
+      "grad_norm": 1.1043392419815063,
+      "learning_rate": 1.4326579261025031e-05,
+      "loss": 0.0188,
+      "step": 483
+    },
+    {
+      "epoch": 0.28843861740166865,
+      "grad_norm": 2.307523488998413,
+      "learning_rate": 1.4314660309892732e-05,
+      "loss": 0.0518,
+      "step": 484
+    },
+    {
+      "epoch": 0.28903456495828367,
+      "grad_norm": 2.9132611751556396,
+      "learning_rate": 1.430274135876043e-05,
+      "loss": 0.0377,
+      "step": 485
+    },
+    {
+      "epoch": 0.2896305125148987,
+      "grad_norm": 5.395816802978516,
+      "learning_rate": 1.4290822407628132e-05,
+      "loss": 0.2856,
+      "step": 486
+    },
+    {
+      "epoch": 0.2902264600715137,
+      "grad_norm": 2.5879149436950684,
+      "learning_rate": 1.427890345649583e-05,
+      "loss": 0.0568,
+      "step": 487
+    },
+    {
+      "epoch": 0.2908224076281287,
+      "grad_norm": 4.823842525482178,
+      "learning_rate": 1.4266984505363528e-05,
+      "loss": 0.311,
+      "step": 488
+    },
+    {
+      "epoch": 0.29141835518474374,
+      "grad_norm": 1.8846055269241333,
+      "learning_rate": 1.425506555423123e-05,
+      "loss": 0.0339,
+      "step": 489
+    },
+    {
+      "epoch": 0.29201430274135876,
+      "grad_norm": 4.279256820678711,
+      "learning_rate": 1.4243146603098927e-05,
+      "loss": 0.1958,
+      "step": 490
+    },
+    {
+      "epoch": 0.2926102502979738,
+      "grad_norm": 9.75631332397461,
+      "learning_rate": 1.4231227651966629e-05,
+      "loss": 0.972,
+      "step": 491
+    },
+    {
+      "epoch": 0.2932061978545888,
+      "grad_norm": 2.10233736038208,
+      "learning_rate": 1.4219308700834327e-05,
+      "loss": 0.0664,
+      "step": 492
+    },
+    {
+      "epoch": 0.2938021454112038,
+      "grad_norm": 3.1594085693359375,
+      "learning_rate": 1.4207389749702028e-05,
+      "loss": 0.115,
+      "step": 493
+    },
+    {
+      "epoch": 0.2943980929678188,
+      "grad_norm": 3.892519474029541,
+      "learning_rate": 1.4195470798569726e-05,
+      "loss": 0.1553,
+      "step": 494
+    },
+    {
+      "epoch": 0.29499404052443384,
+      "grad_norm": 24.172910690307617,
+      "learning_rate": 1.4183551847437426e-05,
+      "loss": 0.2514,
+      "step": 495
+    },
+    {
+      "epoch": 0.29558998808104886,
+      "grad_norm": 4.714790344238281,
+      "learning_rate": 1.4171632896305126e-05,
+      "loss": 0.1898,
+      "step": 496
+    },
+    {
+      "epoch": 0.2961859356376639,
+      "grad_norm": 4.153560161590576,
+      "learning_rate": 1.4159713945172825e-05,
+      "loss": 0.2247,
+      "step": 497
+    },
+    {
+      "epoch": 0.2967818831942789,
+      "grad_norm": 3.070568561553955,
+      "learning_rate": 1.4147794994040525e-05,
+      "loss": 0.1211,
+      "step": 498
+    },
+    {
+      "epoch": 0.2973778307508939,
+      "grad_norm": 2.4574832916259766,
+      "learning_rate": 1.4135876042908225e-05,
+      "loss": 0.0503,
+      "step": 499
+    },
+    {
+      "epoch": 0.29797377830750893,
+      "grad_norm": 3.0890557765960693,
+      "learning_rate": 1.4123957091775925e-05,
+      "loss": 0.0807,
+      "step": 500
+    },
+    {
+      "epoch": 0.29856972586412395,
+      "grad_norm": 2.228830575942993,
+      "learning_rate": 1.4112038140643624e-05,
+      "loss": 0.0323,
+      "step": 501
+    },
+    {
+      "epoch": 0.29916567342073896,
+      "grad_norm": 4.063799858093262,
+      "learning_rate": 1.4100119189511322e-05,
+      "loss": 0.1101,
+      "step": 502
+    },
+    {
+      "epoch": 0.299761620977354,
+      "grad_norm": 5.5590386390686035,
+      "learning_rate": 1.4088200238379024e-05,
+      "loss": 0.4582,
+      "step": 503
+    },
+    {
+      "epoch": 0.300357568533969,
+      "grad_norm": 4.587915897369385,
+      "learning_rate": 1.4076281287246722e-05,
+      "loss": 0.2101,
+      "step": 504
+    },
+    {
+      "epoch": 0.300953516090584,
+      "grad_norm": 4.897768497467041,
+      "learning_rate": 1.4064362336114423e-05,
+      "loss": 0.1522,
+      "step": 505
+    },
+    {
+      "epoch": 0.30154946364719903,
+      "grad_norm": 4.087748050689697,
+      "learning_rate": 1.4052443384982121e-05,
+      "loss": 0.1745,
+      "step": 506
+    },
+    {
+      "epoch": 0.30214541120381405,
+      "grad_norm": 2.282811403274536,
+      "learning_rate": 1.4040524433849823e-05,
+      "loss": 0.0229,
+      "step": 507
+    },
+    {
+      "epoch": 0.30274135876042907,
+      "grad_norm": 0.7391116619110107,
+      "learning_rate": 1.402860548271752e-05,
+      "loss": 0.0177,
+      "step": 508
+    },
+    {
+      "epoch": 0.3033373063170441,
+      "grad_norm": 7.224737167358398,
+      "learning_rate": 1.4016686531585222e-05,
+      "loss": 0.588,
+      "step": 509
+    },
+    {
+      "epoch": 0.3039332538736591,
+      "grad_norm": 3.4185519218444824,
+      "learning_rate": 1.400476758045292e-05,
+      "loss": 0.0938,
+      "step": 510
+    },
+    {
+      "epoch": 0.3045292014302741,
+      "grad_norm": 3.6759955883026123,
+      "learning_rate": 1.3992848629320622e-05,
+      "loss": 0.1649,
+      "step": 511
+    },
+    {
+      "epoch": 0.30512514898688914,
+      "grad_norm": 4.904788017272949,
+      "learning_rate": 1.398092967818832e-05,
+      "loss": 0.131,
+      "step": 512
+    },
+    {
+      "epoch": 0.30572109654350416,
+      "grad_norm": 3.3704440593719482,
+      "learning_rate": 1.3969010727056021e-05,
+      "loss": 0.1282,
+      "step": 513
+    },
+    {
+      "epoch": 0.3063170441001192,
+      "grad_norm": 3.9749178886413574,
+      "learning_rate": 1.395709177592372e-05,
+      "loss": 0.1129,
+      "step": 514
+    },
+    {
+      "epoch": 0.3069129916567342,
+      "grad_norm": 3.193791151046753,
+      "learning_rate": 1.394517282479142e-05,
+      "loss": 0.0996,
+      "step": 515
+    },
+    {
+      "epoch": 0.3075089392133492,
+      "grad_norm": 3.1608386039733887,
+      "learning_rate": 1.3933253873659119e-05,
+      "loss": 0.1312,
+      "step": 516
+    },
+    {
+      "epoch": 0.3081048867699642,
+      "grad_norm": 4.956654071807861,
+      "learning_rate": 1.3921334922526818e-05,
+      "loss": 0.2756,
+      "step": 517
+    },
+    {
+      "epoch": 0.30870083432657924,
+      "grad_norm": 3.7328097820281982,
+      "learning_rate": 1.3909415971394518e-05,
+      "loss": 0.2027,
+      "step": 518
+    },
+    {
+      "epoch": 0.30929678188319426,
+      "grad_norm": 2.870579957962036,
+      "learning_rate": 1.3897497020262218e-05,
+      "loss": 0.0937,
+      "step": 519
+    },
+    {
+      "epoch": 0.3098927294398093,
+      "grad_norm": 4.193901538848877,
+      "learning_rate": 1.3885578069129918e-05,
+      "loss": 0.1329,
+      "step": 520
+    },
+    {
+      "epoch": 0.3104886769964243,
+      "grad_norm": 2.3944690227508545,
+      "learning_rate": 1.3873659117997617e-05,
+      "loss": 0.0571,
+      "step": 521
+    },
+    {
+      "epoch": 0.3110846245530393,
+      "grad_norm": 1.8659662008285522,
+      "learning_rate": 1.3861740166865317e-05,
+      "loss": 0.0376,
+      "step": 522
+    },
+    {
+      "epoch": 0.31168057210965433,
+      "grad_norm": 3.642270565032959,
+      "learning_rate": 1.3849821215733017e-05,
+      "loss": 0.1404,
+      "step": 523
+    },
+    {
+      "epoch": 0.31227651966626935,
+      "grad_norm": 4.470912456512451,
+      "learning_rate": 1.3837902264600715e-05,
+      "loss": 0.3188,
+      "step": 524
+    },
+    {
+      "epoch": 0.31287246722288437,
+      "grad_norm": 10.376492500305176,
+      "learning_rate": 1.3825983313468416e-05,
+      "loss": 0.6817,
+      "step": 525
+    },
+    {
+      "epoch": 0.3134684147794994,
+      "grad_norm": 6.656085014343262,
+      "learning_rate": 1.3814064362336114e-05,
+      "loss": 0.5184,
+      "step": 526
+    },
+    {
+      "epoch": 0.3140643623361144,
+      "grad_norm": 3.5556893348693848,
+      "learning_rate": 1.3802145411203816e-05,
+      "loss": 0.1825,
+      "step": 527
+    },
+    {
+      "epoch": 0.3146603098927294,
+      "grad_norm": 3.8574345111846924,
+      "learning_rate": 1.3790226460071514e-05,
+      "loss": 0.1193,
+      "step": 528
+    },
+    {
+      "epoch": 0.31525625744934443,
+      "grad_norm": 2.6895792484283447,
+      "learning_rate": 1.3778307508939215e-05,
+      "loss": 0.069,
+      "step": 529
+    },
+    {
+      "epoch": 0.31585220500595945,
+      "grad_norm": 6.012507915496826,
+      "learning_rate": 1.3766388557806913e-05,
+      "loss": 0.4622,
+      "step": 530
+    },
+    {
+      "epoch": 0.31644815256257447,
+      "grad_norm": 3.0153255462646484,
+      "learning_rate": 1.3754469606674615e-05,
+      "loss": 0.0696,
+      "step": 531
+    },
+    {
+      "epoch": 0.3170441001191895,
+      "grad_norm": 2.278900623321533,
+      "learning_rate": 1.3742550655542313e-05,
+      "loss": 0.0251,
+      "step": 532
+    },
+    {
+      "epoch": 0.3176400476758045,
+      "grad_norm": 2.330068826675415,
+      "learning_rate": 1.3730631704410014e-05,
+      "loss": 0.0689,
+      "step": 533
+    },
+    {
+      "epoch": 0.3182359952324195,
+      "grad_norm": 3.925318717956543,
+      "learning_rate": 1.3718712753277712e-05,
+      "loss": 0.1771,
+      "step": 534
+    },
+    {
+      "epoch": 0.31883194278903454,
+      "grad_norm": 3.5365331172943115,
+      "learning_rate": 1.3706793802145413e-05,
+      "loss": 0.2166,
+      "step": 535
+    },
+    {
+      "epoch": 0.31942789034564956,
+      "grad_norm": 2.478240966796875,
+      "learning_rate": 1.3694874851013112e-05,
+      "loss": 0.0321,
+      "step": 536
+    },
+    {
+      "epoch": 0.3200238379022646,
+      "grad_norm": 3.17566180229187,
+      "learning_rate": 1.3682955899880811e-05,
+      "loss": 0.0926,
+      "step": 537
+    },
+    {
+      "epoch": 0.3206197854588796,
+      "grad_norm": 4.212337493896484,
+      "learning_rate": 1.3671036948748511e-05,
+      "loss": 0.1558,
+      "step": 538
+    },
+    {
+      "epoch": 0.3212157330154946,
+      "grad_norm": 3.0905327796936035,
+      "learning_rate": 1.365911799761621e-05,
+      "loss": 0.0651,
+      "step": 539
+    },
+    {
+      "epoch": 0.3218116805721097,
+      "grad_norm": 2.5117387771606445,
+      "learning_rate": 1.364719904648391e-05,
+      "loss": 0.0538,
+      "step": 540
+    },
+    {
+      "epoch": 0.3224076281287247,
+      "grad_norm": 2.924748420715332,
+      "learning_rate": 1.363528009535161e-05,
+      "loss": 0.0838,
+      "step": 541
+    },
+    {
+      "epoch": 0.3230035756853397,
+      "grad_norm": 1.8705695867538452,
+      "learning_rate": 1.362336114421931e-05,
+      "loss": 0.022,
+      "step": 542
+    },
+    {
+      "epoch": 0.32359952324195473,
+      "grad_norm": 3.3179733753204346,
+      "learning_rate": 1.361144219308701e-05,
+      "loss": 0.1072,
+      "step": 543
+    },
+    {
+      "epoch": 0.32419547079856975,
+      "grad_norm": 3.4413230419158936,
+      "learning_rate": 1.3599523241954708e-05,
+      "loss": 0.0709,
+      "step": 544
+    },
+    {
+      "epoch": 0.32479141835518477,
+      "grad_norm": 4.145533561706543,
+      "learning_rate": 1.3587604290822409e-05,
+      "loss": 0.1681,
+      "step": 545
+    },
+    {
+      "epoch": 0.3253873659117998,
+      "grad_norm": 4.72059440612793,
+      "learning_rate": 1.3575685339690107e-05,
+      "loss": 0.2103,
+      "step": 546
+    },
+    {
+      "epoch": 0.3259833134684148,
+      "grad_norm": 2.4660146236419678,
+      "learning_rate": 1.3563766388557809e-05,
+      "loss": 0.0412,
+      "step": 547
+    },
+    {
+      "epoch": 0.3265792610250298,
+      "grad_norm": 2.4961555004119873,
+      "learning_rate": 1.3551847437425507e-05,
+      "loss": 0.0783,
+      "step": 548
+    },
+    {
+      "epoch": 0.32717520858164484,
+      "grad_norm": 1.017722249031067,
+      "learning_rate": 1.3539928486293208e-05,
+      "loss": 0.0167,
+      "step": 549
+    },
+    {
+      "epoch": 0.32777115613825986,
+      "grad_norm": 5.505906581878662,
+      "learning_rate": 1.3528009535160906e-05,
+      "loss": 0.3254,
+      "step": 550
+    },
+    {
+      "epoch": 0.3283671036948749,
+      "grad_norm": 8.321074485778809,
+      "learning_rate": 1.3516090584028607e-05,
+      "loss": 0.798,
+      "step": 551
+    },
+    {
+      "epoch": 0.3289630512514899,
+      "grad_norm": 5.956407070159912,
+      "learning_rate": 1.3504171632896306e-05,
+      "loss": 0.458,
+      "step": 552
+    },
+    {
+      "epoch": 0.3295589988081049,
+      "grad_norm": 1.9115909337997437,
+      "learning_rate": 1.3492252681764007e-05,
+      "loss": 0.0202,
+      "step": 553
+    },
+    {
+      "epoch": 0.3301549463647199,
+      "grad_norm": 2.8708105087280273,
+      "learning_rate": 1.3480333730631705e-05,
+      "loss": 0.0853,
+      "step": 554
+    },
+    {
+      "epoch": 0.33075089392133494,
+      "grad_norm": 6.163847923278809,
+      "learning_rate": 1.3468414779499406e-05,
+      "loss": 0.3321,
+      "step": 555
+    },
+    {
+      "epoch": 0.33134684147794996,
+      "grad_norm": 3.070089101791382,
+      "learning_rate": 1.3456495828367104e-05,
+      "loss": 0.1867,
+      "step": 556
+    },
+    {
+      "epoch": 0.331942789034565,
+      "grad_norm": 7.3529205322265625,
+      "learning_rate": 1.3444576877234806e-05,
+      "loss": 0.646,
+      "step": 557
+    },
+    {
+      "epoch": 0.33253873659118,
+      "grad_norm": 3.590104103088379,
+      "learning_rate": 1.3432657926102504e-05,
+      "loss": 0.1018,
+      "step": 558
+    },
+    {
+      "epoch": 0.333134684147795,
+      "grad_norm": 4.889064788818359,
+      "learning_rate": 1.3420738974970204e-05,
+      "loss": 0.362,
+      "step": 559
+    },
+    {
+      "epoch": 0.33373063170441003,
+      "grad_norm": 3.1876652240753174,
+      "learning_rate": 1.3408820023837903e-05,
+      "loss": 0.164,
+      "step": 560
+    },
+    {
+      "epoch": 0.33432657926102505,
+      "grad_norm": 3.568819999694824,
+      "learning_rate": 1.3396901072705603e-05,
+      "loss": 0.1145,
+      "step": 561
+    },
+    {
+      "epoch": 0.33492252681764006,
+      "grad_norm": 2.5840606689453125,
+      "learning_rate": 1.3384982121573303e-05,
+      "loss": 0.061,
+      "step": 562
+    },
+    {
+      "epoch": 0.3355184743742551,
+      "grad_norm": 4.453753471374512,
+      "learning_rate": 1.3373063170441003e-05,
+      "loss": 0.228,
+      "step": 563
+    },
+    {
+      "epoch": 0.3361144219308701,
+      "grad_norm": 5.986978530883789,
+      "learning_rate": 1.3361144219308702e-05,
+      "loss": 0.1778,
+      "step": 564
+    },
+    {
+      "epoch": 0.3367103694874851,
+      "grad_norm": 3.835988998413086,
+      "learning_rate": 1.3349225268176402e-05,
+      "loss": 0.1167,
+      "step": 565
+    },
+    {
+      "epoch": 0.33730631704410013,
+      "grad_norm": 3.051140785217285,
+      "learning_rate": 1.33373063170441e-05,
+      "loss": 0.0666,
+      "step": 566
+    },
+    {
+      "epoch": 0.33790226460071515,
+      "grad_norm": 3.979266881942749,
+      "learning_rate": 1.3325387365911801e-05,
+      "loss": 0.1625,
+      "step": 567
+    },
+    {
+      "epoch": 0.33849821215733017,
+      "grad_norm": 1.930410385131836,
+      "learning_rate": 1.33134684147795e-05,
+      "loss": 0.0331,
+      "step": 568
+    },
+    {
+      "epoch": 0.3390941597139452,
+      "grad_norm": 2.7404911518096924,
+      "learning_rate": 1.3301549463647201e-05,
+      "loss": 0.0956,
+      "step": 569
+    },
+    {
+      "epoch": 0.3396901072705602,
+      "grad_norm": 4.13883638381958,
+      "learning_rate": 1.3289630512514899e-05,
+      "loss": 0.0582,
+      "step": 570
+    },
+    {
+      "epoch": 0.3402860548271752,
+      "grad_norm": 4.226868152618408,
+      "learning_rate": 1.32777115613826e-05,
+      "loss": 0.1948,
+      "step": 571
+    },
+    {
+      "epoch": 0.34088200238379024,
+      "grad_norm": 3.164144515991211,
+      "learning_rate": 1.3265792610250298e-05,
+      "loss": 0.0719,
+      "step": 572
+    },
+    {
+      "epoch": 0.34147794994040526,
+      "grad_norm": 2.6855525970458984,
+      "learning_rate": 1.3253873659118e-05,
+      "loss": 0.0504,
+      "step": 573
+    },
+    {
+      "epoch": 0.3420738974970203,
+      "grad_norm": 4.396729946136475,
+      "learning_rate": 1.3241954707985698e-05,
+      "loss": 0.1567,
+      "step": 574
+    },
+    {
+      "epoch": 0.3426698450536353,
+      "grad_norm": 4.433279991149902,
+      "learning_rate": 1.32300357568534e-05,
+      "loss": 0.1416,
+      "step": 575
+    },
+    {
+      "epoch": 0.3432657926102503,
+      "grad_norm": 7.948300838470459,
+      "learning_rate": 1.3218116805721097e-05,
+      "loss": 0.697,
+      "step": 576
+    },
+    {
+      "epoch": 0.3438617401668653,
+      "grad_norm": 6.6973795890808105,
+      "learning_rate": 1.3206197854588799e-05,
+      "loss": 0.3286,
+      "step": 577
+    },
+    {
+      "epoch": 0.34445768772348034,
+      "grad_norm": 4.675992012023926,
+      "learning_rate": 1.3194278903456497e-05,
+      "loss": 0.298,
+      "step": 578
+    },
+    {
+      "epoch": 0.34505363528009536,
+      "grad_norm": 2.959054708480835,
+      "learning_rate": 1.3182359952324198e-05,
+      "loss": 0.0574,
+      "step": 579
+    },
+    {
+      "epoch": 0.3456495828367104,
+      "grad_norm": 3.3435616493225098,
+      "learning_rate": 1.3170441001191896e-05,
+      "loss": 0.0784,
+      "step": 580
+    },
+    {
+      "epoch": 0.3462455303933254,
+      "grad_norm": 3.4622182846069336,
+      "learning_rate": 1.3158522050059596e-05,
+      "loss": 0.1813,
+      "step": 581
+    },
+    {
+      "epoch": 0.3468414779499404,
+      "grad_norm": 4.997390270233154,
+      "learning_rate": 1.3146603098927296e-05,
+      "loss": 0.3107,
+      "step": 582
+    },
+    {
+      "epoch": 0.34743742550655543,
+      "grad_norm": 3.0057790279388428,
+      "learning_rate": 1.3134684147794995e-05,
+      "loss": 0.0402,
+      "step": 583
+    },
+    {
+      "epoch": 0.34803337306317045,
+      "grad_norm": 3.4651038646698,
+      "learning_rate": 1.3122765196662695e-05,
+      "loss": 0.0881,
+      "step": 584
+    },
+    {
+      "epoch": 0.34862932061978547,
+      "grad_norm": 1.903336524963379,
+      "learning_rate": 1.3110846245530395e-05,
+      "loss": 0.0298,
+      "step": 585
+    },
+    {
+      "epoch": 0.3492252681764005,
+      "grad_norm": 1.704109787940979,
+      "learning_rate": 1.3098927294398095e-05,
+      "loss": 0.0263,
+      "step": 586
+    },
+    {
+      "epoch": 0.3498212157330155,
+      "grad_norm": 3.7336678504943848,
+      "learning_rate": 1.3087008343265794e-05,
+      "loss": 0.1508,
+      "step": 587
+    },
+    {
+      "epoch": 0.3504171632896305,
+      "grad_norm": 2.1317920684814453,
+      "learning_rate": 1.3075089392133492e-05,
+      "loss": 0.0452,
+      "step": 588
+    },
+    {
+      "epoch": 0.35101311084624554,
+      "grad_norm": 1.7177764177322388,
+      "learning_rate": 1.3063170441001194e-05,
+      "loss": 0.0256,
+      "step": 589
+    },
+    {
+      "epoch": 0.35160905840286055,
+      "grad_norm": 1.4051704406738281,
+      "learning_rate": 1.3051251489868892e-05,
+      "loss": 0.0297,
+      "step": 590
+    },
+    {
+      "epoch": 0.35220500595947557,
+      "grad_norm": 3.51342511177063,
+      "learning_rate": 1.3039332538736593e-05,
+      "loss": 0.072,
+      "step": 591
+    },
+    {
+      "epoch": 0.3528009535160906,
+      "grad_norm": 2.421335220336914,
+      "learning_rate": 1.3027413587604291e-05,
+      "loss": 0.0575,
+      "step": 592
+    },
+    {
+      "epoch": 0.3533969010727056,
+      "grad_norm": 1.837225317955017,
+      "learning_rate": 1.3015494636471993e-05,
+      "loss": 0.0282,
+      "step": 593
+    },
+    {
+      "epoch": 0.3539928486293206,
+      "grad_norm": 3.41684889793396,
+      "learning_rate": 1.300357568533969e-05,
+      "loss": 0.1343,
+      "step": 594
+    },
+    {
+      "epoch": 0.35458879618593564,
+      "grad_norm": 7.998478412628174,
+      "learning_rate": 1.2991656734207392e-05,
+      "loss": 0.5576,
+      "step": 595
+    },
+    {
+      "epoch": 0.35518474374255066,
+      "grad_norm": 3.0062448978424072,
+      "learning_rate": 1.297973778307509e-05,
+      "loss": 0.0684,
+      "step": 596
+    },
+    {
+      "epoch": 0.3557806912991657,
+      "grad_norm": 2.3612401485443115,
+      "learning_rate": 1.2967818831942792e-05,
+      "loss": 0.0251,
+      "step": 597
+    },
+    {
+      "epoch": 0.3563766388557807,
+      "grad_norm": 1.9386613368988037,
+      "learning_rate": 1.295589988081049e-05,
+      "loss": 0.0324,
+      "step": 598
+    },
+    {
+      "epoch": 0.3569725864123957,
+      "grad_norm": 4.650648593902588,
+      "learning_rate": 1.2943980929678191e-05,
+      "loss": 0.2221,
+      "step": 599
+    },
+    {
+      "epoch": 0.3575685339690107,
+      "grad_norm": 3.6872398853302,
+      "learning_rate": 1.293206197854589e-05,
+      "loss": 0.1358,
+      "step": 600
+    },
+    {
+      "epoch": 0.35816448152562574,
+      "grad_norm": 2.4655203819274902,
+      "learning_rate": 1.2920143027413589e-05,
+      "loss": 0.0515,
+      "step": 601
+    },
+    {
+      "epoch": 0.35876042908224076,
+      "grad_norm": 5.462170600891113,
+      "learning_rate": 1.2908224076281289e-05,
+      "loss": 0.1812,
+      "step": 602
+    },
+    {
+      "epoch": 0.3593563766388558,
+      "grad_norm": 1.3872747421264648,
+      "learning_rate": 1.2896305125148988e-05,
+      "loss": 0.0186,
+      "step": 603
+    },
+    {
+      "epoch": 0.3599523241954708,
+      "grad_norm": 2.036639451980591,
+      "learning_rate": 1.2884386174016688e-05,
+      "loss": 0.0137,
+      "step": 604
+    },
+    {
+      "epoch": 0.3605482717520858,
+      "grad_norm": 3.2226405143737793,
+      "learning_rate": 1.2872467222884388e-05,
+      "loss": 0.0809,
+      "step": 605
+    },
+    {
+      "epoch": 0.36114421930870083,
+      "grad_norm": 3.0040385723114014,
+      "learning_rate": 1.2860548271752088e-05,
+      "loss": 0.0932,
+      "step": 606
+    },
+    {
+      "epoch": 0.36174016686531585,
+      "grad_norm": 6.233769416809082,
+      "learning_rate": 1.2848629320619786e-05,
+      "loss": 0.2877,
+      "step": 607
+    },
+    {
+      "epoch": 0.36233611442193087,
+      "grad_norm": 3.302543878555298,
+      "learning_rate": 1.2836710369487485e-05,
+      "loss": 0.1021,
+      "step": 608
+    },
+    {
+      "epoch": 0.3629320619785459,
+      "grad_norm": 5.671337604522705,
+      "learning_rate": 1.2824791418355185e-05,
+      "loss": 0.2357,
+      "step": 609
+    },
+    {
+      "epoch": 0.3635280095351609,
+      "grad_norm": 2.0020344257354736,
+      "learning_rate": 1.2812872467222885e-05,
+      "loss": 0.0459,
+      "step": 610
+    },
+    {
+      "epoch": 0.3641239570917759,
+      "grad_norm": 3.1047258377075195,
+      "learning_rate": 1.2800953516090585e-05,
+      "loss": 0.0927,
+      "step": 611
+    },
+    {
+      "epoch": 0.36471990464839094,
+      "grad_norm": 4.837769985198975,
+      "learning_rate": 1.2789034564958284e-05,
+      "loss": 0.1544,
+      "step": 612
+    },
+    {
+      "epoch": 0.36531585220500595,
+      "grad_norm": 2.6963586807250977,
+      "learning_rate": 1.2777115613825984e-05,
+      "loss": 0.0699,
+      "step": 613
+    },
+    {
+      "epoch": 0.36591179976162097,
+      "grad_norm": 2.21235990524292,
+      "learning_rate": 1.2765196662693684e-05,
+      "loss": 0.0443,
+      "step": 614
+    },
+    {
+      "epoch": 0.366507747318236,
+      "grad_norm": 3.001622200012207,
+      "learning_rate": 1.2753277711561383e-05,
+      "loss": 0.0787,
+      "step": 615
+    },
+    {
+      "epoch": 0.367103694874851,
+      "grad_norm": 3.380511999130249,
+      "learning_rate": 1.2741358760429083e-05,
+      "loss": 0.0642,
+      "step": 616
+    },
+    {
+      "epoch": 0.367699642431466,
+      "grad_norm": 1.8585082292556763,
+      "learning_rate": 1.2729439809296781e-05,
+      "loss": 0.0387,
+      "step": 617
+    },
+    {
+      "epoch": 0.36829558998808104,
+      "grad_norm": 2.275733470916748,
+      "learning_rate": 1.2717520858164483e-05,
+      "loss": 0.0525,
+      "step": 618
+    },
+    {
+      "epoch": 0.36889153754469606,
+      "grad_norm": 2.4696946144104004,
+      "learning_rate": 1.270560190703218e-05,
+      "loss": 0.0436,
+      "step": 619
+    },
+    {
+      "epoch": 0.3694874851013111,
+      "grad_norm": 2.2676820755004883,
+      "learning_rate": 1.2693682955899882e-05,
+      "loss": 0.0494,
+      "step": 620
+    },
+    {
+      "epoch": 0.3700834326579261,
+      "grad_norm": 3.579043388366699,
+      "learning_rate": 1.268176400476758e-05,
+      "loss": 0.1267,
+      "step": 621
+    },
+    {
+      "epoch": 0.3706793802145411,
+      "grad_norm": 5.4460368156433105,
+      "learning_rate": 1.2669845053635282e-05,
+      "loss": 0.2702,
+      "step": 622
+    },
+    {
+      "epoch": 0.3712753277711561,
+      "grad_norm": 3.440387010574341,
+      "learning_rate": 1.265792610250298e-05,
+      "loss": 0.0828,
+      "step": 623
+    },
+    {
+      "epoch": 0.37187127532777114,
+      "grad_norm": 3.8335559368133545,
+      "learning_rate": 1.2646007151370681e-05,
+      "loss": 0.1118,
+      "step": 624
+    },
+    {
+      "epoch": 0.37246722288438616,
+      "grad_norm": 4.149360179901123,
+      "learning_rate": 1.2634088200238379e-05,
+      "loss": 0.118,
+      "step": 625
+    },
+    {
+      "epoch": 0.3730631704410012,
+      "grad_norm": 4.922280788421631,
+      "learning_rate": 1.262216924910608e-05,
+      "loss": 0.1163,
+      "step": 626
+    },
+    {
+      "epoch": 0.3736591179976162,
+      "grad_norm": 2.9473013877868652,
+      "learning_rate": 1.2610250297973779e-05,
+      "loss": 0.166,
+      "step": 627
+    },
+    {
+      "epoch": 0.3742550655542312,
+      "grad_norm": 2.701385498046875,
+      "learning_rate": 1.259833134684148e-05,
+      "loss": 0.0447,
+      "step": 628
+    },
+    {
+      "epoch": 0.37485101311084623,
+      "grad_norm": 2.241863489151001,
+      "learning_rate": 1.2586412395709178e-05,
+      "loss": 0.0561,
+      "step": 629
+    },
+    {
+      "epoch": 0.37544696066746125,
+      "grad_norm": 3.1526219844818115,
+      "learning_rate": 1.2574493444576878e-05,
+      "loss": 0.0521,
+      "step": 630
+    },
+    {
+      "epoch": 0.37604290822407627,
+      "grad_norm": 1.2572112083435059,
+      "learning_rate": 1.2562574493444577e-05,
+      "loss": 0.0212,
+      "step": 631
+    },
+    {
+      "epoch": 0.3766388557806913,
+      "grad_norm": 3.293759822845459,
+      "learning_rate": 1.2550655542312277e-05,
+      "loss": 0.104,
+      "step": 632
+    },
+    {
+      "epoch": 0.3772348033373063,
+      "grad_norm": 1.3019709587097168,
+      "learning_rate": 1.2538736591179977e-05,
+      "loss": 0.021,
+      "step": 633
+    },
+    {
+      "epoch": 0.3778307508939213,
+      "grad_norm": 5.169968128204346,
+      "learning_rate": 1.2526817640047677e-05,
+      "loss": 0.1306,
+      "step": 634
+    },
+    {
+      "epoch": 0.37842669845053634,
+      "grad_norm": 5.304752826690674,
+      "learning_rate": 1.2514898688915376e-05,
+      "loss": 0.3971,
+      "step": 635
+    },
+    {
+      "epoch": 0.37902264600715135,
+      "grad_norm": 2.3723411560058594,
+      "learning_rate": 1.2502979737783076e-05,
+      "loss": 0.0443,
+      "step": 636
+    },
+    {
+      "epoch": 0.37961859356376637,
+      "grad_norm": 3.1224007606506348,
+      "learning_rate": 1.2491060786650774e-05,
+      "loss": 0.0688,
+      "step": 637
+    },
+    {
+      "epoch": 0.3802145411203814,
+      "grad_norm": 1.678567886352539,
+      "learning_rate": 1.2479141835518476e-05,
+      "loss": 0.0336,
+      "step": 638
+    },
+    {
+      "epoch": 0.3808104886769964,
+      "grad_norm": 2.978668212890625,
+      "learning_rate": 1.2467222884386174e-05,
+      "loss": 0.0986,
+      "step": 639
+    },
+    {
+      "epoch": 0.3814064362336114,
+      "grad_norm": 1.542576551437378,
+      "learning_rate": 1.2455303933253875e-05,
+      "loss": 0.0308,
+      "step": 640
+    },
+    {
+      "epoch": 0.38200238379022644,
+      "grad_norm": 2.650332450866699,
+      "learning_rate": 1.2443384982121573e-05,
+      "loss": 0.0564,
+      "step": 641
+    },
+    {
+      "epoch": 0.38259833134684146,
+      "grad_norm": 3.7695212364196777,
+      "learning_rate": 1.2431466030989274e-05,
+      "loss": 0.1411,
+      "step": 642
+    },
+    {
+      "epoch": 0.3831942789034565,
+      "grad_norm": 3.6411709785461426,
+      "learning_rate": 1.2419547079856973e-05,
+      "loss": 0.0622,
+      "step": 643
+    },
+    {
+      "epoch": 0.3837902264600715,
+      "grad_norm": 1.8929977416992188,
+      "learning_rate": 1.2407628128724674e-05,
+      "loss": 0.0371,
+      "step": 644
+    },
+    {
+      "epoch": 0.3843861740166865,
+      "grad_norm": 2.0274276733398438,
+      "learning_rate": 1.2395709177592372e-05,
+      "loss": 0.0426,
+      "step": 645
+    },
+    {
+      "epoch": 0.38498212157330153,
+      "grad_norm": 1.529266357421875,
+      "learning_rate": 1.2383790226460073e-05,
+      "loss": 0.023,
+      "step": 646
+    },
+    {
+      "epoch": 0.38557806912991655,
+      "grad_norm": 3.9683215618133545,
+      "learning_rate": 1.2371871275327771e-05,
+      "loss": 0.1415,
+      "step": 647
+    },
+    {
+      "epoch": 0.38617401668653156,
+      "grad_norm": 2.6790013313293457,
+      "learning_rate": 1.2359952324195473e-05,
+      "loss": 0.0508,
+      "step": 648
+    },
+    {
+      "epoch": 0.3867699642431466,
+      "grad_norm": 3.530477285385132,
+      "learning_rate": 1.2348033373063171e-05,
+      "loss": 0.142,
+      "step": 649
+    },
+    {
+      "epoch": 0.3873659117997616,
+      "grad_norm": 3.1619162559509277,
+      "learning_rate": 1.2336114421930872e-05,
+      "loss": 0.1361,
+      "step": 650
+    },
+    {
+      "epoch": 0.3879618593563766,
+      "grad_norm": 13.816330909729004,
+      "learning_rate": 1.232419547079857e-05,
+      "loss": 1.7421,
+      "step": 651
+    },
+    {
+      "epoch": 0.38855780691299163,
+      "grad_norm": 2.1363441944122314,
+      "learning_rate": 1.231227651966627e-05,
+      "loss": 0.0376,
+      "step": 652
+    },
+    {
+      "epoch": 0.38915375446960665,
+      "grad_norm": 2.9179763793945312,
+      "learning_rate": 1.230035756853397e-05,
+      "loss": 0.0659,
+      "step": 653
+    },
+    {
+      "epoch": 0.38974970202622167,
+      "grad_norm": 3.3663978576660156,
+      "learning_rate": 1.228843861740167e-05,
+      "loss": 0.1284,
+      "step": 654
+    },
+    {
+      "epoch": 0.3903456495828367,
+      "grad_norm": 2.41943621635437,
+      "learning_rate": 1.227651966626937e-05,
+      "loss": 0.0494,
+      "step": 655
+    },
+    {
+      "epoch": 0.3909415971394517,
+      "grad_norm": 5.170660972595215,
+      "learning_rate": 1.2264600715137069e-05,
+      "loss": 0.2181,
+      "step": 656
+    },
+    {
+      "epoch": 0.3915375446960667,
+      "grad_norm": 2.642092704772949,
+      "learning_rate": 1.2252681764004769e-05,
+      "loss": 0.1055,
+      "step": 657
+    },
+    {
+      "epoch": 0.39213349225268174,
+      "grad_norm": 2.909374713897705,
+      "learning_rate": 1.2240762812872468e-05,
+      "loss": 0.1019,
+      "step": 658
+    },
+    {
+      "epoch": 0.39272943980929675,
+      "grad_norm": 2.503941297531128,
+      "learning_rate": 1.2228843861740166e-05,
+      "loss": 0.0391,
+      "step": 659
+    },
+    {
+      "epoch": 0.3933253873659118,
+      "grad_norm": 6.356411933898926,
+      "learning_rate": 1.2216924910607868e-05,
+      "loss": 0.2821,
+      "step": 660
+    },
+    {
+      "epoch": 0.39392133492252684,
+      "grad_norm": 4.524708271026611,
+      "learning_rate": 1.2205005959475566e-05,
+      "loss": 0.094,
+      "step": 661
+    },
+    {
+      "epoch": 0.39451728247914186,
+      "grad_norm": 4.60608434677124,
+      "learning_rate": 1.2193087008343267e-05,
+      "loss": 0.1612,
+      "step": 662
+    },
+    {
+      "epoch": 0.3951132300357569,
+      "grad_norm": 2.713935613632202,
+      "learning_rate": 1.2181168057210965e-05,
+      "loss": 0.0903,
+      "step": 663
+    },
+    {
+      "epoch": 0.3957091775923719,
+      "grad_norm": 2.883127450942993,
+      "learning_rate": 1.2169249106078667e-05,
+      "loss": 0.0662,
+      "step": 664
+    },
+    {
+      "epoch": 0.3963051251489869,
+      "grad_norm": 2.9159200191497803,
+      "learning_rate": 1.2157330154946365e-05,
+      "loss": 0.055,
+      "step": 665
+    },
+    {
+      "epoch": 0.39690107270560193,
+      "grad_norm": 4.793641090393066,
+      "learning_rate": 1.2145411203814066e-05,
+      "loss": 0.2275,
+      "step": 666
+    },
+    {
+      "epoch": 0.39749702026221695,
+      "grad_norm": 3.529507875442505,
+      "learning_rate": 1.2133492252681764e-05,
+      "loss": 0.1909,
+      "step": 667
+    },
+    {
+      "epoch": 0.39809296781883197,
+      "grad_norm": 3.9109435081481934,
+      "learning_rate": 1.2121573301549466e-05,
+      "loss": 0.1014,
+      "step": 668
+    },
+    {
+      "epoch": 0.398688915375447,
+      "grad_norm": 2.9119136333465576,
+      "learning_rate": 1.2109654350417164e-05,
+      "loss": 0.0483,
+      "step": 669
+    },
+    {
+      "epoch": 0.399284862932062,
+      "grad_norm": 1.4599201679229736,
+      "learning_rate": 1.2097735399284865e-05,
+      "loss": 0.0224,
+      "step": 670
+    },
+    {
+      "epoch": 0.399880810488677,
+      "grad_norm": 4.009641647338867,
+      "learning_rate": 1.2085816448152563e-05,
+      "loss": 0.1535,
+      "step": 671
+    },
+    {
+      "epoch": 0.40047675804529204,
+      "grad_norm": 2.6406471729278564,
+      "learning_rate": 1.2073897497020263e-05,
+      "loss": 0.0677,
+      "step": 672
+    },
+    {
+      "epoch": 0.40107270560190705,
+      "grad_norm": 6.773016929626465,
+      "learning_rate": 1.2061978545887963e-05,
+      "loss": 0.4814,
+      "step": 673
+    },
+    {
+      "epoch": 0.40166865315852207,
+      "grad_norm": 7.848520755767822,
+      "learning_rate": 1.2050059594755662e-05,
+      "loss": 0.2966,
+      "step": 674
+    },
+    {
+      "epoch": 0.4022646007151371,
+      "grad_norm": 4.699748992919922,
+      "learning_rate": 1.2038140643623362e-05,
+      "loss": 0.1277,
+      "step": 675
+    },
+    {
+      "epoch": 0.4028605482717521,
+      "grad_norm": 2.9775490760803223,
+      "learning_rate": 1.2026221692491062e-05,
+      "loss": 0.0739,
+      "step": 676
+    },
+    {
+      "epoch": 0.4034564958283671,
+      "grad_norm": 7.156108379364014,
+      "learning_rate": 1.2014302741358762e-05,
+      "loss": 0.553,
+      "step": 677
+    },
+    {
+      "epoch": 0.40405244338498214,
+      "grad_norm": 6.093362331390381,
+      "learning_rate": 1.2002383790226461e-05,
+      "loss": 0.3501,
+      "step": 678
+    },
+    {
+      "epoch": 0.40464839094159716,
+      "grad_norm": 3.155334234237671,
+      "learning_rate": 1.1990464839094161e-05,
+      "loss": 0.132,
+      "step": 679
+    },
+    {
+      "epoch": 0.4052443384982122,
+      "grad_norm": 2.753248453140259,
+      "learning_rate": 1.197854588796186e-05,
+      "loss": 0.0677,
+      "step": 680
+    },
+    {
+      "epoch": 0.4058402860548272,
+      "grad_norm": 6.128621578216553,
+      "learning_rate": 1.1966626936829559e-05,
+      "loss": 0.3333,
+      "step": 681
+    },
+    {
+      "epoch": 0.4064362336114422,
+      "grad_norm": 1.743346929550171,
+      "learning_rate": 1.195470798569726e-05,
+      "loss": 0.033,
+      "step": 682
+    },
+    {
+      "epoch": 0.4070321811680572,
+      "grad_norm": 4.199587821960449,
+      "learning_rate": 1.1942789034564958e-05,
+      "loss": 0.1512,
+      "step": 683
+    },
+    {
+      "epoch": 0.40762812872467225,
+      "grad_norm": 2.238823652267456,
+      "learning_rate": 1.193087008343266e-05,
+      "loss": 0.0469,
+      "step": 684
+    },
+    {
+      "epoch": 0.40822407628128726,
+      "grad_norm": 2.6608726978302,
+      "learning_rate": 1.1918951132300358e-05,
+      "loss": 0.0327,
+      "step": 685
+    },
+    {
+      "epoch": 0.4088200238379023,
+      "grad_norm": 2.6922109127044678,
+      "learning_rate": 1.190703218116806e-05,
+      "loss": 0.0316,
+      "step": 686
+    },
+    {
+      "epoch": 0.4094159713945173,
+      "grad_norm": 3.6056435108184814,
+      "learning_rate": 1.1895113230035757e-05,
+      "loss": 0.1172,
+      "step": 687
+    },
+    {
+      "epoch": 0.4100119189511323,
+      "grad_norm": 3.809875011444092,
+      "learning_rate": 1.1883194278903459e-05,
+      "loss": 0.1522,
+      "step": 688
+    },
+    {
+      "epoch": 0.41060786650774733,
+      "grad_norm": 4.728883266448975,
+      "learning_rate": 1.1871275327771157e-05,
+      "loss": 0.2615,
+      "step": 689
+    },
+    {
+      "epoch": 0.41120381406436235,
+      "grad_norm": 3.74723744392395,
+      "learning_rate": 1.1859356376638858e-05,
+      "loss": 0.1228,
+      "step": 690
+    },
+    {
+      "epoch": 0.41179976162097737,
+      "grad_norm": 1.238860845565796,
+      "learning_rate": 1.1847437425506556e-05,
+      "loss": 0.0227,
+      "step": 691
+    },
+    {
+      "epoch": 0.4123957091775924,
+      "grad_norm": 2.0329248905181885,
+      "learning_rate": 1.1835518474374258e-05,
+      "loss": 0.0384,
+      "step": 692
+    },
+    {
+      "epoch": 0.4129916567342074,
+      "grad_norm": 2.987541675567627,
+      "learning_rate": 1.1823599523241956e-05,
+      "loss": 0.0746,
+      "step": 693
+    },
+    {
+      "epoch": 0.4135876042908224,
+      "grad_norm": 2.707832098007202,
+      "learning_rate": 1.1811680572109655e-05,
+      "loss": 0.037,
+      "step": 694
+    },
+    {
+      "epoch": 0.41418355184743744,
+      "grad_norm": 2.6554768085479736,
+      "learning_rate": 1.1799761620977355e-05,
+      "loss": 0.0585,
+      "step": 695
+    },
+    {
+      "epoch": 0.41477949940405245,
+      "grad_norm": 2.2640035152435303,
+      "learning_rate": 1.1787842669845055e-05,
+      "loss": 0.0424,
+      "step": 696
+    },
+    {
+      "epoch": 0.41537544696066747,
+      "grad_norm": 3.8394651412963867,
+      "learning_rate": 1.1775923718712755e-05,
+      "loss": 0.0604,
+      "step": 697
+    },
+    {
+      "epoch": 0.4159713945172825,
+      "grad_norm": 1.5762083530426025,
+      "learning_rate": 1.1764004767580454e-05,
+      "loss": 0.0179,
+      "step": 698
+    },
+    {
+      "epoch": 0.4165673420738975,
+      "grad_norm": 3.1598410606384277,
+      "learning_rate": 1.1752085816448154e-05,
+      "loss": 0.0599,
+      "step": 699
+    },
+    {
+      "epoch": 0.4171632896305125,
+      "grad_norm": 2.9568002223968506,
+      "learning_rate": 1.1740166865315854e-05,
+      "loss": 0.0519,
+      "step": 700
+    },
+    {
+      "epoch": 0.41775923718712754,
+      "grad_norm": 3.39498233795166,
+      "learning_rate": 1.1728247914183552e-05,
+      "loss": 0.0388,
+      "step": 701
+    },
+    {
+      "epoch": 0.41835518474374256,
+      "grad_norm": 5.292623996734619,
+      "learning_rate": 1.1716328963051253e-05,
+      "loss": 0.2904,
+      "step": 702
+    },
+    {
+      "epoch": 0.4189511323003576,
+      "grad_norm": 4.077442169189453,
+      "learning_rate": 1.1704410011918951e-05,
+      "loss": 0.1207,
+      "step": 703
+    },
+    {
+      "epoch": 0.4195470798569726,
+      "grad_norm": 5.353928565979004,
+      "learning_rate": 1.1692491060786653e-05,
+      "loss": 0.1931,
+      "step": 704
+    },
+    {
+      "epoch": 0.4201430274135876,
+      "grad_norm": 3.357200860977173,
+      "learning_rate": 1.168057210965435e-05,
+      "loss": 0.0526,
+      "step": 705
+    },
+    {
+      "epoch": 0.42073897497020263,
+      "grad_norm": 1.401711106300354,
+      "learning_rate": 1.1668653158522052e-05,
+      "loss": 0.0167,
+      "step": 706
+    },
+    {
+      "epoch": 0.42133492252681765,
+      "grad_norm": 2.340942859649658,
+      "learning_rate": 1.165673420738975e-05,
+      "loss": 0.0685,
+      "step": 707
+    },
+    {
+      "epoch": 0.42193087008343266,
+      "grad_norm": 4.997816562652588,
+      "learning_rate": 1.1644815256257452e-05,
+      "loss": 0.1771,
+      "step": 708
+    },
+    {
+      "epoch": 0.4225268176400477,
+      "grad_norm": 2.727902412414551,
+      "learning_rate": 1.163289630512515e-05,
+      "loss": 0.0629,
+      "step": 709
+    },
+    {
+      "epoch": 0.4231227651966627,
+      "grad_norm": 3.728635311126709,
+      "learning_rate": 1.1620977353992851e-05,
+      "loss": 0.1262,
+      "step": 710
+    },
+    {
+      "epoch": 0.4237187127532777,
+      "grad_norm": 6.210958003997803,
+      "learning_rate": 1.1609058402860549e-05,
+      "loss": 0.3626,
+      "step": 711
+    },
+    {
+      "epoch": 0.42431466030989273,
+      "grad_norm": 2.9489166736602783,
+      "learning_rate": 1.159713945172825e-05,
+      "loss": 0.1245,
+      "step": 712
+    },
+    {
+      "epoch": 0.42491060786650775,
+      "grad_norm": 6.476369857788086,
+      "learning_rate": 1.1585220500595949e-05,
+      "loss": 0.4989,
+      "step": 713
+    },
+    {
+      "epoch": 0.42550655542312277,
+      "grad_norm": 2.18210506439209,
+      "learning_rate": 1.157330154946365e-05,
+      "loss": 0.0254,
+      "step": 714
+    },
+    {
+      "epoch": 0.4261025029797378,
+      "grad_norm": 7.5067949295043945,
+      "learning_rate": 1.1561382598331348e-05,
+      "loss": 0.5297,
+      "step": 715
+    },
+    {
+      "epoch": 0.4266984505363528,
+      "grad_norm": 6.778183460235596,
+      "learning_rate": 1.1549463647199048e-05,
+      "loss": 0.3382,
+      "step": 716
+    },
+    {
+      "epoch": 0.4272943980929678,
+      "grad_norm": 10.103796005249023,
+      "learning_rate": 1.1537544696066747e-05,
+      "loss": 0.743,
+      "step": 717
+    },
+    {
+      "epoch": 0.42789034564958284,
+      "grad_norm": 2.513467788696289,
+      "learning_rate": 1.1525625744934447e-05,
+      "loss": 0.0505,
+      "step": 718
+    },
+    {
+      "epoch": 0.42848629320619785,
+      "grad_norm": 2.779320240020752,
+      "learning_rate": 1.1513706793802147e-05,
+      "loss": 0.0954,
+      "step": 719
+    },
+    {
+      "epoch": 0.42908224076281287,
+      "grad_norm": 3.824164628982544,
+      "learning_rate": 1.1501787842669847e-05,
+      "loss": 0.2222,
+      "step": 720
+    },
+    {
+      "epoch": 0.4296781883194279,
+      "grad_norm": 2.6972196102142334,
+      "learning_rate": 1.1489868891537546e-05,
+      "loss": 0.0474,
+      "step": 721
+    },
+    {
+      "epoch": 0.4302741358760429,
+      "grad_norm": 4.635561466217041,
+      "learning_rate": 1.1477949940405246e-05,
+      "loss": 0.1863,
+      "step": 722
+    },
+    {
+      "epoch": 0.4308700834326579,
+      "grad_norm": 2.550352096557617,
+      "learning_rate": 1.1466030989272944e-05,
+      "loss": 0.0458,
+      "step": 723
+    },
+    {
+      "epoch": 0.43146603098927294,
+      "grad_norm": 6.033018589019775,
+      "learning_rate": 1.1454112038140646e-05,
+      "loss": 0.2404,
+      "step": 724
+    },
+    {
+      "epoch": 0.43206197854588796,
+      "grad_norm": 3.400806188583374,
+      "learning_rate": 1.1442193087008344e-05,
+      "loss": 0.125,
+      "step": 725
+    },
+    {
+      "epoch": 0.432657926102503,
+      "grad_norm": 2.8533029556274414,
+      "learning_rate": 1.1430274135876045e-05,
+      "loss": 0.0617,
+      "step": 726
+    },
+    {
+      "epoch": 0.433253873659118,
+      "grad_norm": 2.520301103591919,
+      "learning_rate": 1.1418355184743743e-05,
+      "loss": 0.064,
+      "step": 727
+    },
+    {
+      "epoch": 0.433849821215733,
+      "grad_norm": 1.712004542350769,
+      "learning_rate": 1.1406436233611443e-05,
+      "loss": 0.0286,
+      "step": 728
+    },
+    {
+      "epoch": 0.43444576877234803,
+      "grad_norm": 4.453799247741699,
+      "learning_rate": 1.1394517282479143e-05,
+      "loss": 0.1471,
+      "step": 729
+    },
+    {
+      "epoch": 0.43504171632896305,
+      "grad_norm": 5.082167148590088,
+      "learning_rate": 1.138259833134684e-05,
+      "loss": 0.2702,
+      "step": 730
+    },
+    {
+      "epoch": 0.43563766388557806,
+      "grad_norm": 4.99561882019043,
+      "learning_rate": 1.1370679380214542e-05,
+      "loss": 0.19,
+      "step": 731
+    },
+    {
+      "epoch": 0.4362336114421931,
+      "grad_norm": 4.495167255401611,
+      "learning_rate": 1.135876042908224e-05,
+      "loss": 0.261,
+      "step": 732
+    },
+    {
+      "epoch": 0.4368295589988081,
+      "grad_norm": 1.6293431520462036,
+      "learning_rate": 1.1346841477949941e-05,
+      "loss": 0.0246,
+      "step": 733
+    },
+    {
+      "epoch": 0.4374255065554231,
+      "grad_norm": 4.137955188751221,
+      "learning_rate": 1.133492252681764e-05,
+      "loss": 0.0944,
+      "step": 734
+    },
+    {
+      "epoch": 0.43802145411203813,
+      "grad_norm": 1.0005615949630737,
+      "learning_rate": 1.1323003575685341e-05,
+      "loss": 0.0197,
+      "step": 735
+    },
+    {
+      "epoch": 0.43861740166865315,
+      "grad_norm": 2.495532989501953,
+      "learning_rate": 1.1311084624553039e-05,
+      "loss": 0.0483,
+      "step": 736
+    },
+    {
+      "epoch": 0.43921334922526817,
+      "grad_norm": 5.885705471038818,
+      "learning_rate": 1.129916567342074e-05,
+      "loss": 0.503,
+      "step": 737
+    },
+    {
+      "epoch": 0.4398092967818832,
+      "grad_norm": 3.394489049911499,
+      "learning_rate": 1.1287246722288438e-05,
+      "loss": 0.1068,
+      "step": 738
+    },
+    {
+      "epoch": 0.4404052443384982,
+      "grad_norm": 4.112785816192627,
+      "learning_rate": 1.127532777115614e-05,
+      "loss": 0.1092,
+      "step": 739
+    },
+    {
+      "epoch": 0.4410011918951132,
+      "grad_norm": 2.34576678276062,
+      "learning_rate": 1.1263408820023838e-05,
+      "loss": 0.0399,
+      "step": 740
+    },
+    {
+      "epoch": 0.44159713945172824,
+      "grad_norm": 4.662506580352783,
+      "learning_rate": 1.125148986889154e-05,
+      "loss": 0.2551,
+      "step": 741
+    },
+    {
+      "epoch": 0.44219308700834326,
+      "grad_norm": 2.0359137058258057,
+      "learning_rate": 1.1239570917759237e-05,
+      "loss": 0.0441,
+      "step": 742
+    },
+    {
+      "epoch": 0.4427890345649583,
+      "grad_norm": 2.182255506515503,
+      "learning_rate": 1.1227651966626937e-05,
+      "loss": 0.0571,
+      "step": 743
+    },
+    {
+      "epoch": 0.4433849821215733,
+      "grad_norm": 3.577421188354492,
+      "learning_rate": 1.1215733015494637e-05,
+      "loss": 0.1253,
+      "step": 744
+    },
+    {
+      "epoch": 0.4439809296781883,
+      "grad_norm": 3.8679351806640625,
+      "learning_rate": 1.1203814064362337e-05,
+      "loss": 0.1617,
+      "step": 745
+    },
+    {
+      "epoch": 0.4445768772348033,
+      "grad_norm": 1.0977805852890015,
+      "learning_rate": 1.1191895113230036e-05,
+      "loss": 0.0148,
+      "step": 746
+    },
+    {
+      "epoch": 0.44517282479141834,
+      "grad_norm": 7.07128381729126,
+      "learning_rate": 1.1179976162097736e-05,
+      "loss": 0.2979,
+      "step": 747
+    },
+    {
+      "epoch": 0.44576877234803336,
+      "grad_norm": 5.245192050933838,
+      "learning_rate": 1.1168057210965436e-05,
+      "loss": 0.253,
+      "step": 748
+    },
+    {
+      "epoch": 0.4463647199046484,
+      "grad_norm": 2.6536622047424316,
+      "learning_rate": 1.1156138259833135e-05,
+      "loss": 0.0777,
+      "step": 749
+    },
+    {
+      "epoch": 0.4469606674612634,
+      "grad_norm": 4.8840837478637695,
+      "learning_rate": 1.1144219308700835e-05,
+      "loss": 0.2967,
+      "step": 750
+    },
+    {
+      "epoch": 0.4475566150178784,
+      "grad_norm": 7.978186130523682,
+      "learning_rate": 1.1132300357568535e-05,
+      "loss": 0.3119,
+      "step": 751
+    },
+    {
+      "epoch": 0.44815256257449343,
+      "grad_norm": 1.278037190437317,
+      "learning_rate": 1.1120381406436233e-05,
+      "loss": 0.0243,
+      "step": 752
+    },
+    {
+      "epoch": 0.44874851013110845,
+      "grad_norm": 4.259028911590576,
+      "learning_rate": 1.1108462455303934e-05,
+      "loss": 0.1283,
+      "step": 753
+    },
+    {
+      "epoch": 0.44934445768772346,
+      "grad_norm": 4.432330131530762,
+      "learning_rate": 1.1096543504171632e-05,
+      "loss": 0.2504,
+      "step": 754
+    },
+    {
+      "epoch": 0.4499404052443385,
+      "grad_norm": 4.339845180511475,
+      "learning_rate": 1.1084624553039334e-05,
+      "loss": 0.1604,
+      "step": 755
+    },
+    {
+      "epoch": 0.4505363528009535,
+      "grad_norm": 2.067514657974243,
+      "learning_rate": 1.1072705601907032e-05,
+      "loss": 0.0385,
+      "step": 756
+    },
+    {
+      "epoch": 0.4511323003575685,
+      "grad_norm": 3.1854448318481445,
+      "learning_rate": 1.1060786650774733e-05,
+      "loss": 0.1257,
+      "step": 757
+    },
+    {
+      "epoch": 0.45172824791418353,
+      "grad_norm": 2.665107250213623,
+      "learning_rate": 1.1048867699642431e-05,
+      "loss": 0.0633,
+      "step": 758
+    },
+    {
+      "epoch": 0.45232419547079855,
+      "grad_norm": 3.023249864578247,
+      "learning_rate": 1.1036948748510133e-05,
+      "loss": 0.1373,
+      "step": 759
+    },
+    {
+      "epoch": 0.45292014302741357,
+      "grad_norm": 3.842465877532959,
+      "learning_rate": 1.102502979737783e-05,
+      "loss": 0.0733,
+      "step": 760
+    },
+    {
+      "epoch": 0.4535160905840286,
+      "grad_norm": 2.867462635040283,
+      "learning_rate": 1.1013110846245532e-05,
+      "loss": 0.11,
+      "step": 761
+    },
+    {
+      "epoch": 0.4541120381406436,
+      "grad_norm": 3.2108397483825684,
+      "learning_rate": 1.100119189511323e-05,
+      "loss": 0.0965,
+      "step": 762
+    },
+    {
+      "epoch": 0.4547079856972586,
+      "grad_norm": 10.566121101379395,
+      "learning_rate": 1.0989272943980932e-05,
+      "loss": 1.2608,
+      "step": 763
+    },
+    {
+      "epoch": 0.45530393325387364,
+      "grad_norm": 2.1822731494903564,
+      "learning_rate": 1.097735399284863e-05,
+      "loss": 0.059,
+      "step": 764
+    },
+    {
+      "epoch": 0.45589988081048866,
+      "grad_norm": 1.0682462453842163,
+      "learning_rate": 1.096543504171633e-05,
+      "loss": 0.0179,
+      "step": 765
+    },
+    {
+      "epoch": 0.4564958283671037,
+      "grad_norm": 0.4905363917350769,
+      "learning_rate": 1.095351609058403e-05,
+      "loss": 0.0159,
+      "step": 766
+    },
+    {
+      "epoch": 0.4570917759237187,
+      "grad_norm": 3.169326066970825,
+      "learning_rate": 1.0941597139451729e-05,
+      "loss": 0.0825,
+      "step": 767
+    },
+    {
+      "epoch": 0.4576877234803337,
+      "grad_norm": 3.6801364421844482,
+      "learning_rate": 1.0929678188319429e-05,
+      "loss": 0.0318,
+      "step": 768
+    },
+    {
+      "epoch": 0.4582836710369487,
+      "grad_norm": 2.560535192489624,
+      "learning_rate": 1.0917759237187128e-05,
+      "loss": 0.0737,
+      "step": 769
+    },
+    {
+      "epoch": 0.45887961859356374,
+      "grad_norm": 3.1345980167388916,
+      "learning_rate": 1.0905840286054828e-05,
+      "loss": 0.0648,
+      "step": 770
+    },
+    {
+      "epoch": 0.45947556615017876,
+      "grad_norm": 2.198857307434082,
+      "learning_rate": 1.0893921334922528e-05,
+      "loss": 0.0448,
+      "step": 771
+    },
+    {
+      "epoch": 0.4600715137067938,
+      "grad_norm": 3.2347042560577393,
+      "learning_rate": 1.0882002383790226e-05,
+      "loss": 0.0648,
+      "step": 772
+    },
+    {
+      "epoch": 0.4606674612634088,
+      "grad_norm": 5.357144832611084,
+      "learning_rate": 1.0870083432657927e-05,
+      "loss": 0.3072,
+      "step": 773
+    },
+    {
+      "epoch": 0.4612634088200238,
+      "grad_norm": 2.470999240875244,
+      "learning_rate": 1.0858164481525625e-05,
+      "loss": 0.0571,
+      "step": 774
+    },
+    {
+      "epoch": 0.46185935637663883,
+      "grad_norm": 3.8135132789611816,
+      "learning_rate": 1.0846245530393327e-05,
+      "loss": 0.1321,
+      "step": 775
+    },
+    {
+      "epoch": 0.46245530393325385,
+      "grad_norm": 2.938664197921753,
+      "learning_rate": 1.0834326579261025e-05,
+      "loss": 0.0869,
+      "step": 776
+    },
+    {
+      "epoch": 0.46305125148986886,
+      "grad_norm": 20.734119415283203,
+      "learning_rate": 1.0822407628128726e-05,
+      "loss": 0.1628,
+      "step": 777
+    },
+    {
+      "epoch": 0.4636471990464839,
+      "grad_norm": 5.030472755432129,
+      "learning_rate": 1.0810488676996424e-05,
+      "loss": 0.2817,
+      "step": 778
+    },
+    {
+      "epoch": 0.4642431466030989,
+      "grad_norm": 5.104273796081543,
+      "learning_rate": 1.0798569725864126e-05,
+      "loss": 0.3205,
+      "step": 779
+    },
+    {
+      "epoch": 0.464839094159714,
+      "grad_norm": 5.450915336608887,
+      "learning_rate": 1.0786650774731824e-05,
+      "loss": 0.2732,
+      "step": 780
+    },
+    {
+      "epoch": 0.465435041716329,
+      "grad_norm": 3.955986976623535,
+      "learning_rate": 1.0774731823599525e-05,
+      "loss": 0.0764,
+      "step": 781
+    },
+    {
+      "epoch": 0.466030989272944,
+      "grad_norm": 7.30074405670166,
+      "learning_rate": 1.0762812872467223e-05,
+      "loss": 0.4075,
+      "step": 782
+    },
+    {
+      "epoch": 0.466626936829559,
+      "grad_norm": 2.0793118476867676,
+      "learning_rate": 1.0750893921334925e-05,
+      "loss": 0.0489,
+      "step": 783
+    },
+    {
+      "epoch": 0.46722288438617404,
+      "grad_norm": 1.9123713970184326,
+      "learning_rate": 1.0738974970202623e-05,
+      "loss": 0.0277,
+      "step": 784
+    },
+    {
+      "epoch": 0.46781883194278906,
+      "grad_norm": 5.827648639678955,
+      "learning_rate": 1.0727056019070324e-05,
+      "loss": 0.2293,
+      "step": 785
+    },
+    {
+      "epoch": 0.4684147794994041,
+      "grad_norm": 2.1069295406341553,
+      "learning_rate": 1.0715137067938022e-05,
+      "loss": 0.0204,
+      "step": 786
+    },
+    {
+      "epoch": 0.4690107270560191,
+      "grad_norm": 1.8579899072647095,
+      "learning_rate": 1.0703218116805722e-05,
+      "loss": 0.0296,
+      "step": 787
+    },
+    {
+      "epoch": 0.4696066746126341,
+      "grad_norm": 5.065797805786133,
+      "learning_rate": 1.0691299165673422e-05,
+      "loss": 0.2378,
+      "step": 788
+    },
+    {
+      "epoch": 0.47020262216924913,
+      "grad_norm": 2.1179423332214355,
+      "learning_rate": 1.0679380214541121e-05,
+      "loss": 0.0479,
+      "step": 789
+    },
+    {
+      "epoch": 0.47079856972586415,
+      "grad_norm": 3.1752891540527344,
+      "learning_rate": 1.0667461263408821e-05,
+      "loss": 0.0738,
+      "step": 790
+    },
+    {
+      "epoch": 0.47139451728247916,
+      "grad_norm": 1.7890164852142334,
+      "learning_rate": 1.065554231227652e-05,
+      "loss": 0.0383,
+      "step": 791
+    },
+    {
+      "epoch": 0.4719904648390942,
+      "grad_norm": 2.6091272830963135,
+      "learning_rate": 1.064362336114422e-05,
+      "loss": 0.074,
+      "step": 792
+    },
+    {
+      "epoch": 0.4725864123957092,
+      "grad_norm": 2.0145928859710693,
+      "learning_rate": 1.063170441001192e-05,
+      "loss": 0.0258,
+      "step": 793
+    },
+    {
+      "epoch": 0.4731823599523242,
+      "grad_norm": 3.838930368423462,
+      "learning_rate": 1.0619785458879618e-05,
+      "loss": 0.111,
+      "step": 794
+    },
+    {
+      "epoch": 0.47377830750893923,
+      "grad_norm": 2.6002509593963623,
+      "learning_rate": 1.060786650774732e-05,
+      "loss": 0.0404,
+      "step": 795
+    },
+    {
+      "epoch": 0.47437425506555425,
+      "grad_norm": 1.8239020109176636,
+      "learning_rate": 1.0595947556615018e-05,
+      "loss": 0.0198,
+      "step": 796
+    },
+    {
+      "epoch": 0.47497020262216927,
+      "grad_norm": 5.055425643920898,
+      "learning_rate": 1.0584028605482719e-05,
+      "loss": 0.2892,
+      "step": 797
+    },
+    {
+      "epoch": 0.4755661501787843,
+      "grad_norm": 1.8775006532669067,
+      "learning_rate": 1.0572109654350417e-05,
+      "loss": 0.0281,
+      "step": 798
+    },
+    {
+      "epoch": 0.4761620977353993,
+      "grad_norm": 3.077529191970825,
+      "learning_rate": 1.0560190703218119e-05,
+      "loss": 0.0593,
+      "step": 799
+    },
+    {
+      "epoch": 0.4767580452920143,
+      "grad_norm": 7.136275291442871,
+      "learning_rate": 1.0548271752085817e-05,
+      "loss": 0.7126,
+      "step": 800
+    },
+    {
+      "epoch": 0.47735399284862934,
+      "grad_norm": 2.7651515007019043,
+      "learning_rate": 1.0536352800953518e-05,
+      "loss": 0.0598,
+      "step": 801
+    },
+    {
+      "epoch": 0.47794994040524436,
+      "grad_norm": 3.3661513328552246,
+      "learning_rate": 1.0524433849821216e-05,
+      "loss": 0.089,
+      "step": 802
+    },
+    {
+      "epoch": 0.4785458879618594,
+      "grad_norm": 0.8585639595985413,
+      "learning_rate": 1.0512514898688917e-05,
+      "loss": 0.0135,
+      "step": 803
+    },
+    {
+      "epoch": 0.4791418355184744,
+      "grad_norm": 3.092067003250122,
+      "learning_rate": 1.0500595947556616e-05,
+      "loss": 0.0684,
+      "step": 804
+    },
+    {
+      "epoch": 0.4797377830750894,
+      "grad_norm": 4.102750778198242,
+      "learning_rate": 1.0488676996424317e-05,
+      "loss": 0.1201,
+      "step": 805
+    },
+    {
+      "epoch": 0.4803337306317044,
+      "grad_norm": 6.136007308959961,
+      "learning_rate": 1.0476758045292015e-05,
+      "loss": 0.4368,
+      "step": 806
+    },
+    {
+      "epoch": 0.48092967818831944,
+      "grad_norm": 2.0066916942596436,
+      "learning_rate": 1.0464839094159715e-05,
+      "loss": 0.0233,
+      "step": 807
+    },
+    {
+      "epoch": 0.48152562574493446,
+      "grad_norm": 2.638580083847046,
+      "learning_rate": 1.0452920143027414e-05,
+      "loss": 0.0452,
+      "step": 808
+    },
+    {
+      "epoch": 0.4821215733015495,
+      "grad_norm": 3.3853845596313477,
+      "learning_rate": 1.0441001191895114e-05,
+      "loss": 0.1248,
+      "step": 809
+    },
+    {
+      "epoch": 0.4827175208581645,
+      "grad_norm": 3.6625969409942627,
+      "learning_rate": 1.0429082240762814e-05,
+      "loss": 0.0957,
+      "step": 810
+    },
+    {
+      "epoch": 0.4833134684147795,
+      "grad_norm": 2.4347305297851562,
+      "learning_rate": 1.0417163289630514e-05,
+      "loss": 0.053,
+      "step": 811
+    },
+    {
+      "epoch": 0.48390941597139453,
+      "grad_norm": 4.927792549133301,
+      "learning_rate": 1.0405244338498213e-05,
+      "loss": 0.3995,
+      "step": 812
+    },
+    {
+      "epoch": 0.48450536352800955,
+      "grad_norm": 7.650815486907959,
+      "learning_rate": 1.0393325387365913e-05,
+      "loss": 0.449,
+      "step": 813
+    },
+    {
+      "epoch": 0.48510131108462456,
+      "grad_norm": 3.4196665287017822,
+      "learning_rate": 1.0381406436233613e-05,
+      "loss": 0.0894,
+      "step": 814
+    },
+    {
+      "epoch": 0.4856972586412396,
+      "grad_norm": 5.510427951812744,
+      "learning_rate": 1.0369487485101313e-05,
+      "loss": 0.26,
+      "step": 815
+    },
+    {
+      "epoch": 0.4862932061978546,
+      "grad_norm": 1.7290810346603394,
+      "learning_rate": 1.035756853396901e-05,
+      "loss": 0.0263,
+      "step": 816
+    },
+    {
+      "epoch": 0.4868891537544696,
+      "grad_norm": 1.8715295791625977,
+      "learning_rate": 1.0345649582836712e-05,
+      "loss": 0.0269,
+      "step": 817
+    },
+    {
+      "epoch": 0.48748510131108463,
+      "grad_norm": 3.060767650604248,
+      "learning_rate": 1.033373063170441e-05,
+      "loss": 0.0627,
+      "step": 818
+    },
+    {
+      "epoch": 0.48808104886769965,
+      "grad_norm": 2.500493049621582,
+      "learning_rate": 1.0321811680572111e-05,
+      "loss": 0.064,
+      "step": 819
+    },
+    {
+      "epoch": 0.48867699642431467,
+      "grad_norm": 2.9807703495025635,
+      "learning_rate": 1.030989272943981e-05,
+      "loss": 0.0812,
+      "step": 820
+    },
+    {
+      "epoch": 0.4892729439809297,
+      "grad_norm": 2.9303271770477295,
+      "learning_rate": 1.0297973778307511e-05,
+      "loss": 0.1412,
+      "step": 821
+    },
+    {
+      "epoch": 0.4898688915375447,
+      "grad_norm": 1.8229219913482666,
+      "learning_rate": 1.0286054827175209e-05,
+      "loss": 0.0586,
+      "step": 822
+    },
+    {
+      "epoch": 0.4904648390941597,
+      "grad_norm": 2.5212156772613525,
+      "learning_rate": 1.027413587604291e-05,
+      "loss": 0.0483,
+      "step": 823
+    },
+    {
+      "epoch": 0.49106078665077474,
+      "grad_norm": 8.439183235168457,
+      "learning_rate": 1.0262216924910608e-05,
+      "loss": 0.5104,
+      "step": 824
+    },
+    {
+      "epoch": 0.49165673420738976,
+      "grad_norm": 2.276721954345703,
+      "learning_rate": 1.025029797377831e-05,
+      "loss": 0.0672,
+      "step": 825
+    },
+    {
+      "epoch": 0.4922526817640048,
+      "grad_norm": 2.139352321624756,
+      "learning_rate": 1.0238379022646008e-05,
+      "loss": 0.0275,
+      "step": 826
+    },
+    {
+      "epoch": 0.4928486293206198,
+      "grad_norm": 3.033491849899292,
+      "learning_rate": 1.022646007151371e-05,
+      "loss": 0.0716,
+      "step": 827
+    },
+    {
+      "epoch": 0.4934445768772348,
+      "grad_norm": 3.7414989471435547,
+      "learning_rate": 1.0214541120381407e-05,
+      "loss": 0.0901,
+      "step": 828
+    },
+    {
+      "epoch": 0.4940405244338498,
+      "grad_norm": 2.414663791656494,
+      "learning_rate": 1.0202622169249107e-05,
+      "loss": 0.0623,
+      "step": 829
+    },
+    {
+      "epoch": 0.49463647199046484,
+      "grad_norm": 3.2372257709503174,
+      "learning_rate": 1.0190703218116807e-05,
+      "loss": 0.1001,
+      "step": 830
+    },
+    {
+      "epoch": 0.49523241954707986,
+      "grad_norm": 3.101979970932007,
+      "learning_rate": 1.0178784266984507e-05,
+      "loss": 0.1026,
+      "step": 831
+    },
+    {
+      "epoch": 0.4958283671036949,
+      "grad_norm": 1.6499109268188477,
+      "learning_rate": 1.0166865315852206e-05,
+      "loss": 0.022,
+      "step": 832
+    },
+    {
+      "epoch": 0.4964243146603099,
+      "grad_norm": 3.2605485916137695,
+      "learning_rate": 1.0154946364719906e-05,
+      "loss": 0.0613,
+      "step": 833
+    },
+    {
+      "epoch": 0.4970202622169249,
+      "grad_norm": 2.228229522705078,
+      "learning_rate": 1.0143027413587606e-05,
+      "loss": 0.069,
+      "step": 834
+    },
+    {
+      "epoch": 0.49761620977353993,
+      "grad_norm": 1.5271751880645752,
+      "learning_rate": 1.0131108462455305e-05,
+      "loss": 0.0237,
+      "step": 835
+    },
+    {
+      "epoch": 0.49821215733015495,
+      "grad_norm": 3.845883846282959,
+      "learning_rate": 1.0119189511323004e-05,
+      "loss": 0.1734,
+      "step": 836
+    },
+    {
+      "epoch": 0.49880810488676997,
+      "grad_norm": 6.340744972229004,
+      "learning_rate": 1.0107270560190705e-05,
+      "loss": 0.5699,
+      "step": 837
+    },
+    {
+      "epoch": 0.499404052443385,
+      "grad_norm": 4.9937825202941895,
+      "learning_rate": 1.0095351609058403e-05,
+      "loss": 0.1785,
+      "step": 838
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.436140060424805,
+      "learning_rate": 1.0083432657926104e-05,
+      "loss": 0.5363,
+      "step": 839
+    },
+    {
+      "epoch": 0.5005959475566151,
+      "grad_norm": 2.3086910247802734,
+      "learning_rate": 1.0071513706793802e-05,
+      "loss": 0.0392,
+      "step": 840
+    },
+    {
+      "epoch": 0.50119189511323,
+      "grad_norm": 8.877634048461914,
+      "learning_rate": 1.0059594755661504e-05,
+      "loss": 0.7652,
+      "step": 841
+    },
+    {
+      "epoch": 0.5017878426698451,
+      "grad_norm": 2.8845701217651367,
+      "learning_rate": 1.0047675804529202e-05,
+      "loss": 0.0502,
+      "step": 842
+    },
+    {
+      "epoch": 0.5023837902264601,
+      "grad_norm": 7.267020225524902,
+      "learning_rate": 1.0035756853396903e-05,
+      "loss": 0.501,
+      "step": 843
+    },
+    {
+      "epoch": 0.5029797377830751,
+      "grad_norm": 3.092458724975586,
+      "learning_rate": 1.0023837902264601e-05,
+      "loss": 0.125,
+      "step": 844
+    },
+    {
+      "epoch": 0.5035756853396901,
+      "grad_norm": 2.8678901195526123,
+      "learning_rate": 1.0011918951132303e-05,
+      "loss": 0.0564,
+      "step": 845
+    },
+    {
+      "epoch": 0.5041716328963052,
+      "grad_norm": 5.802806854248047,
+      "learning_rate": 1e-05,
+      "loss": 0.3051,
+      "step": 846
+    },
+    {
+      "epoch": 0.5047675804529201,
+      "grad_norm": 1.7477421760559082,
+      "learning_rate": 9.9880810488677e-06,
+      "loss": 0.0257,
+      "step": 847
+    },
+    {
+      "epoch": 0.5053635280095352,
+      "grad_norm": 7.440005779266357,
+      "learning_rate": 9.9761620977354e-06,
+      "loss": 0.4641,
+      "step": 848
+    },
+    {
+      "epoch": 0.5059594755661502,
+      "grad_norm": 1.813162922859192,
+      "learning_rate": 9.9642431466031e-06,
+      "loss": 0.0453,
+      "step": 849
+    },
+    {
+      "epoch": 0.5065554231227652,
+      "grad_norm": 4.1152238845825195,
+      "learning_rate": 9.9523241954708e-06,
+      "loss": 0.0814,
+      "step": 850
+    },
+    {
+      "epoch": 0.5071513706793802,
+      "grad_norm": 2.2842793464660645,
+      "learning_rate": 9.9404052443385e-06,
+      "loss": 0.036,
+      "step": 851
+    },
+    {
+      "epoch": 0.5077473182359953,
+      "grad_norm": 4.204278945922852,
+      "learning_rate": 9.9284862932062e-06,
+      "loss": 0.1816,
+      "step": 852
+    },
+    {
+      "epoch": 0.5083432657926102,
+      "grad_norm": 4.989535808563232,
+      "learning_rate": 9.916567342073899e-06,
+      "loss": 0.1615,
+      "step": 853
+    },
+    {
+      "epoch": 0.5089392133492253,
+      "grad_norm": 3.488626003265381,
+      "learning_rate": 9.904648390941599e-06,
+      "loss": 0.1863,
+      "step": 854
+    },
+    {
+      "epoch": 0.5095351609058403,
+      "grad_norm": 4.502569675445557,
+      "learning_rate": 9.892729439809298e-06,
+      "loss": 0.2061,
+      "step": 855
+    },
+    {
+      "epoch": 0.5101311084624554,
+      "grad_norm": 5.368704319000244,
+      "learning_rate": 9.880810488676998e-06,
+      "loss": 0.343,
+      "step": 856
+    },
+    {
+      "epoch": 0.5107270560190703,
+      "grad_norm": 4.670648097991943,
+      "learning_rate": 9.868891537544696e-06,
+      "loss": 0.1705,
+      "step": 857
+    },
+    {
+      "epoch": 0.5113230035756854,
+      "grad_norm": 2.344251871109009,
+      "learning_rate": 9.856972586412396e-06,
+      "loss": 0.0702,
+      "step": 858
+    },
+    {
+      "epoch": 0.5119189511323003,
+      "grad_norm": 2.8786399364471436,
+      "learning_rate": 9.845053635280096e-06,
+      "loss": 0.0688,
+      "step": 859
+    },
+    {
+      "epoch": 0.5125148986889154,
+      "grad_norm": 1.6771944761276245,
+      "learning_rate": 9.833134684147795e-06,
+      "loss": 0.0289,
+      "step": 860
+    },
+    {
+      "epoch": 0.5131108462455304,
+      "grad_norm": 2.164234161376953,
+      "learning_rate": 9.821215733015495e-06,
+      "loss": 0.0546,
+      "step": 861
+    },
+    {
+      "epoch": 0.5137067938021455,
+      "grad_norm": 2.230431079864502,
+      "learning_rate": 9.809296781883195e-06,
+      "loss": 0.0438,
+      "step": 862
+    },
+    {
+      "epoch": 0.5143027413587604,
+      "grad_norm": 5.772002696990967,
+      "learning_rate": 9.797377830750895e-06,
+      "loss": 0.2845,
+      "step": 863
+    },
+    {
+      "epoch": 0.5148986889153755,
+      "grad_norm": 4.240481853485107,
+      "learning_rate": 9.785458879618594e-06,
+      "loss": 0.2245,
+      "step": 864
+    },
+    {
+      "epoch": 0.5154946364719905,
+      "grad_norm": 3.583652973175049,
+      "learning_rate": 9.773539928486294e-06,
+      "loss": 0.0431,
+      "step": 865
+    },
+    {
+      "epoch": 0.5160905840286055,
+      "grad_norm": 1.725081443786621,
+      "learning_rate": 9.761620977353994e-06,
+      "loss": 0.043,
+      "step": 866
+    },
+    {
+      "epoch": 0.5166865315852205,
+      "grad_norm": 2.773484230041504,
+      "learning_rate": 9.749702026221693e-06,
+      "loss": 0.0514,
+      "step": 867
+    },
+    {
+      "epoch": 0.5172824791418356,
+      "grad_norm": 5.514789581298828,
+      "learning_rate": 9.737783075089393e-06,
+      "loss": 0.4276,
+      "step": 868
+    },
+    {
+      "epoch": 0.5178784266984505,
+      "grad_norm": 2.531879186630249,
+      "learning_rate": 9.725864123957093e-06,
+      "loss": 0.0579,
+      "step": 869
+    },
+    {
+      "epoch": 0.5184743742550656,
+      "grad_norm": 5.521443843841553,
+      "learning_rate": 9.713945172824793e-06,
+      "loss": 0.1443,
+      "step": 870
+    },
+    {
+      "epoch": 0.5190703218116806,
+      "grad_norm": 2.0917208194732666,
+      "learning_rate": 9.702026221692492e-06,
+      "loss": 0.0482,
+      "step": 871
+    },
+    {
+      "epoch": 0.5196662693682956,
+      "grad_norm": 2.559394598007202,
+      "learning_rate": 9.690107270560192e-06,
+      "loss": 0.051,
+      "step": 872
+    },
+    {
+      "epoch": 0.5202622169249106,
+      "grad_norm": 6.518402576446533,
+      "learning_rate": 9.678188319427892e-06,
+      "loss": 0.4142,
+      "step": 873
+    },
+    {
+      "epoch": 0.5208581644815257,
+      "grad_norm": 2.548555612564087,
+      "learning_rate": 9.666269368295592e-06,
+      "loss": 0.064,
+      "step": 874
+    },
+    {
+      "epoch": 0.5214541120381406,
+      "grad_norm": 3.3611514568328857,
+      "learning_rate": 9.654350417163291e-06,
+      "loss": 0.1536,
+      "step": 875
+    },
+    {
+      "epoch": 0.5220500595947557,
+      "grad_norm": 1.9860148429870605,
+      "learning_rate": 9.642431466030991e-06,
+      "loss": 0.0476,
+      "step": 876
+    },
+    {
+      "epoch": 0.5226460071513707,
+      "grad_norm": 5.242157459259033,
+      "learning_rate": 9.63051251489869e-06,
+      "loss": 0.3566,
+      "step": 877
+    },
+    {
+      "epoch": 0.5232419547079857,
+      "grad_norm": 2.9342761039733887,
+      "learning_rate": 9.618593563766389e-06,
+      "loss": 0.0656,
+      "step": 878
+    },
+    {
+      "epoch": 0.5238379022646007,
+      "grad_norm": 1.7741440534591675,
+      "learning_rate": 9.606674612634089e-06,
+      "loss": 0.0248,
+      "step": 879
+    },
+    {
+      "epoch": 0.5244338498212158,
+      "grad_norm": 2.2836008071899414,
+      "learning_rate": 9.594755661501788e-06,
+      "loss": 0.0291,
+      "step": 880
+    },
+    {
+      "epoch": 0.5250297973778307,
+      "grad_norm": 6.734463214874268,
+      "learning_rate": 9.582836710369488e-06,
+      "loss": 0.4244,
+      "step": 881
+    },
+    {
+      "epoch": 0.5256257449344458,
+      "grad_norm": 7.955839157104492,
+      "learning_rate": 9.570917759237188e-06,
+      "loss": 0.6723,
+      "step": 882
+    },
+    {
+      "epoch": 0.5262216924910608,
+      "grad_norm": 2.5718812942504883,
+      "learning_rate": 9.558998808104887e-06,
+      "loss": 0.0324,
+      "step": 883
+    },
+    {
+      "epoch": 0.5268176400476758,
+      "grad_norm": 5.10227632522583,
+      "learning_rate": 9.547079856972587e-06,
+      "loss": 0.1951,
+      "step": 884
+    },
+    {
+      "epoch": 0.5274135876042908,
+      "grad_norm": 1.8852691650390625,
+      "learning_rate": 9.535160905840287e-06,
+      "loss": 0.0241,
+      "step": 885
+    },
+    {
+      "epoch": 0.5280095351609059,
+      "grad_norm": 3.4940977096557617,
+      "learning_rate": 9.523241954707987e-06,
+      "loss": 0.1053,
+      "step": 886
+    },
+    {
+      "epoch": 0.5286054827175208,
+      "grad_norm": 10.880402565002441,
+      "learning_rate": 9.511323003575686e-06,
+      "loss": 0.4593,
+      "step": 887
+    },
+    {
+      "epoch": 0.5292014302741359,
+      "grad_norm": 4.85150671005249,
+      "learning_rate": 9.499404052443386e-06,
+      "loss": 0.2323,
+      "step": 888
+    },
+    {
+      "epoch": 0.5297973778307509,
+      "grad_norm": 2.896096706390381,
+      "learning_rate": 9.487485101311086e-06,
+      "loss": 0.0918,
+      "step": 889
+    },
+    {
+      "epoch": 0.5303933253873659,
+      "grad_norm": 2.3204681873321533,
+      "learning_rate": 9.475566150178786e-06,
+      "loss": 0.0486,
+      "step": 890
+    },
+    {
+      "epoch": 0.5309892729439809,
+      "grad_norm": 1.4378235340118408,
+      "learning_rate": 9.463647199046485e-06,
+      "loss": 0.0166,
+      "step": 891
+    },
+    {
+      "epoch": 0.531585220500596,
+      "grad_norm": 2.702502965927124,
+      "learning_rate": 9.451728247914185e-06,
+      "loss": 0.0527,
+      "step": 892
+    },
+    {
+      "epoch": 0.5321811680572109,
+      "grad_norm": 2.885218620300293,
+      "learning_rate": 9.439809296781885e-06,
+      "loss": 0.0819,
+      "step": 893
+    },
+    {
+      "epoch": 0.532777115613826,
+      "grad_norm": 3.9036331176757812,
+      "learning_rate": 9.427890345649584e-06,
+      "loss": 0.0793,
+      "step": 894
+    },
+    {
+      "epoch": 0.533373063170441,
+      "grad_norm": 3.7806947231292725,
+      "learning_rate": 9.415971394517284e-06,
+      "loss": 0.1157,
+      "step": 895
+    },
+    {
+      "epoch": 0.533969010727056,
+      "grad_norm": 5.121149063110352,
+      "learning_rate": 9.404052443384984e-06,
+      "loss": 0.2362,
+      "step": 896
+    },
+    {
+      "epoch": 0.534564958283671,
+      "grad_norm": 3.017591953277588,
+      "learning_rate": 9.392133492252684e-06,
+      "loss": 0.066,
+      "step": 897
+    },
+    {
+      "epoch": 0.5351609058402861,
+      "grad_norm": 2.679819107055664,
+      "learning_rate": 9.380214541120383e-06,
+      "loss": 0.0454,
+      "step": 898
+    },
+    {
+      "epoch": 0.535756853396901,
+      "grad_norm": 3.5048089027404785,
+      "learning_rate": 9.368295589988083e-06,
+      "loss": 0.0583,
+      "step": 899
+    },
+    {
+      "epoch": 0.5363528009535161,
+      "grad_norm": 1.7260154485702515,
+      "learning_rate": 9.356376638855781e-06,
+      "loss": 0.0446,
+      "step": 900
+    },
+    {
+      "epoch": 0.5369487485101311,
+      "grad_norm": 3.1968953609466553,
+      "learning_rate": 9.344457687723481e-06,
+      "loss": 0.0526,
+      "step": 901
+    },
+    {
+      "epoch": 0.5375446960667462,
+      "grad_norm": 3.5607059001922607,
+      "learning_rate": 9.33253873659118e-06,
+      "loss": 0.1363,
+      "step": 902
+    },
+    {
+      "epoch": 0.5381406436233611,
+      "grad_norm": 3.2443525791168213,
+      "learning_rate": 9.32061978545888e-06,
+      "loss": 0.0736,
+      "step": 903
+    },
+    {
+      "epoch": 0.5387365911799762,
+      "grad_norm": 3.2655084133148193,
+      "learning_rate": 9.30870083432658e-06,
+      "loss": 0.078,
+      "step": 904
+    },
+    {
+      "epoch": 0.5393325387365911,
+      "grad_norm": 3.7031915187835693,
+      "learning_rate": 9.29678188319428e-06,
+      "loss": 0.2128,
+      "step": 905
+    },
+    {
+      "epoch": 0.5399284862932062,
+      "grad_norm": 2.430356025695801,
+      "learning_rate": 9.28486293206198e-06,
+      "loss": 0.0793,
+      "step": 906
+    },
+    {
+      "epoch": 0.5405244338498212,
+      "grad_norm": 3.253882884979248,
+      "learning_rate": 9.272943980929678e-06,
+      "loss": 0.1489,
+      "step": 907
+    },
+    {
+      "epoch": 0.5411203814064363,
+      "grad_norm": 2.581735372543335,
+      "learning_rate": 9.261025029797377e-06,
+      "loss": 0.0674,
+      "step": 908
+    },
+    {
+      "epoch": 0.5417163289630512,
+      "grad_norm": 1.8664695024490356,
+      "learning_rate": 9.249106078665077e-06,
+      "loss": 0.0349,
+      "step": 909
+    },
+    {
+      "epoch": 0.5423122765196663,
+      "grad_norm": 8.129162788391113,
+      "learning_rate": 9.237187127532777e-06,
+      "loss": 0.6529,
+      "step": 910
+    },
+    {
+      "epoch": 0.5429082240762813,
+      "grad_norm": 2.8256499767303467,
+      "learning_rate": 9.225268176400477e-06,
+      "loss": 0.0584,
+      "step": 911
+    },
+    {
+      "epoch": 0.5435041716328963,
+      "grad_norm": 6.367729187011719,
+      "learning_rate": 9.213349225268176e-06,
+      "loss": 0.4804,
+      "step": 912
+    },
+    {
+      "epoch": 0.5441001191895113,
+      "grad_norm": 1.6654835939407349,
+      "learning_rate": 9.201430274135876e-06,
+      "loss": 0.0401,
+      "step": 913
+    },
+    {
+      "epoch": 0.5446960667461264,
+      "grad_norm": 0.8940612077713013,
+      "learning_rate": 9.189511323003576e-06,
+      "loss": 0.0171,
+      "step": 914
+    },
+    {
+      "epoch": 0.5452920143027413,
+      "grad_norm": 4.907435417175293,
+      "learning_rate": 9.177592371871275e-06,
+      "loss": 0.2629,
+      "step": 915
+    },
+    {
+      "epoch": 0.5458879618593564,
+      "grad_norm": 3.7550463676452637,
+      "learning_rate": 9.165673420738975e-06,
+      "loss": 0.2344,
+      "step": 916
+    },
+    {
+      "epoch": 0.5464839094159714,
+      "grad_norm": 2.714885950088501,
+      "learning_rate": 9.153754469606675e-06,
+      "loss": 0.0264,
+      "step": 917
+    },
+    {
+      "epoch": 0.5470798569725864,
+      "grad_norm": 2.565814256668091,
+      "learning_rate": 9.141835518474375e-06,
+      "loss": 0.0376,
+      "step": 918
+    },
+    {
+      "epoch": 0.5476758045292014,
+      "grad_norm": 3.824111223220825,
+      "learning_rate": 9.129916567342074e-06,
+      "loss": 0.0977,
+      "step": 919
+    },
+    {
+      "epoch": 0.5482717520858165,
+      "grad_norm": 3.6977102756500244,
+      "learning_rate": 9.117997616209774e-06,
+      "loss": 0.1648,
+      "step": 920
+    },
+    {
+      "epoch": 0.5488676996424314,
+      "grad_norm": 4.884918689727783,
+      "learning_rate": 9.106078665077474e-06,
+      "loss": 0.218,
+      "step": 921
+    },
+    {
+      "epoch": 0.5494636471990465,
+      "grad_norm": 5.428330421447754,
+      "learning_rate": 9.094159713945174e-06,
+      "loss": 0.3814,
+      "step": 922
+    },
+    {
+      "epoch": 0.5500595947556615,
+      "grad_norm": 1.4129438400268555,
+      "learning_rate": 9.082240762812873e-06,
+      "loss": 0.018,
+      "step": 923
+    },
+    {
+      "epoch": 0.5506555423122765,
+      "grad_norm": 1.457028865814209,
+      "learning_rate": 9.070321811680573e-06,
+      "loss": 0.0185,
+      "step": 924
+    },
+    {
+      "epoch": 0.5512514898688915,
+      "grad_norm": 4.178643703460693,
+      "learning_rate": 9.058402860548273e-06,
+      "loss": 0.1838,
+      "step": 925
+    },
+    {
+      "epoch": 0.5518474374255066,
+      "grad_norm": 4.4587836265563965,
+      "learning_rate": 9.046483909415972e-06,
+      "loss": 0.2428,
+      "step": 926
+    },
+    {
+      "epoch": 0.5524433849821215,
+      "grad_norm": 3.8575279712677,
+      "learning_rate": 9.034564958283672e-06,
+      "loss": 0.1159,
+      "step": 927
+    },
+    {
+      "epoch": 0.5530393325387366,
+      "grad_norm": 5.128817081451416,
+      "learning_rate": 9.02264600715137e-06,
+      "loss": 0.35,
+      "step": 928
+    },
+    {
+      "epoch": 0.5536352800953516,
+      "grad_norm": 1.1649953126907349,
+      "learning_rate": 9.01072705601907e-06,
+      "loss": 0.0161,
+      "step": 929
+    },
+    {
+      "epoch": 0.5542312276519666,
+      "grad_norm": 4.548678398132324,
+      "learning_rate": 8.99880810488677e-06,
+      "loss": 0.2536,
+      "step": 930
+    },
+    {
+      "epoch": 0.5548271752085816,
+      "grad_norm": 4.996628761291504,
+      "learning_rate": 8.98688915375447e-06,
+      "loss": 0.3883,
+      "step": 931
+    },
+    {
+      "epoch": 0.5554231227651967,
+      "grad_norm": 2.6920664310455322,
+      "learning_rate": 8.97497020262217e-06,
+      "loss": 0.0523,
+      "step": 932
+    },
+    {
+      "epoch": 0.5560190703218116,
+      "grad_norm": 3.8798282146453857,
+      "learning_rate": 8.963051251489869e-06,
+      "loss": 0.2222,
+      "step": 933
+    },
+    {
+      "epoch": 0.5566150178784267,
+      "grad_norm": 2.5359904766082764,
+      "learning_rate": 8.951132300357569e-06,
+      "loss": 0.0487,
+      "step": 934
+    },
+    {
+      "epoch": 0.5572109654350417,
+      "grad_norm": 5.450087070465088,
+      "learning_rate": 8.939213349225268e-06,
+      "loss": 0.2909,
+      "step": 935
+    },
+    {
+      "epoch": 0.5578069129916567,
+      "grad_norm": 4.621405124664307,
+      "learning_rate": 8.927294398092968e-06,
+      "loss": 0.2136,
+      "step": 936
+    },
+    {
+      "epoch": 0.5584028605482717,
+      "grad_norm": 3.7554264068603516,
+      "learning_rate": 8.915375446960668e-06,
+      "loss": 0.1671,
+      "step": 937
+    },
+    {
+      "epoch": 0.5589988081048868,
+      "grad_norm": 3.642928123474121,
+      "learning_rate": 8.903456495828368e-06,
+      "loss": 0.1316,
+      "step": 938
+    },
+    {
+      "epoch": 0.5595947556615017,
+      "grad_norm": 1.9564480781555176,
+      "learning_rate": 8.891537544696067e-06,
+      "loss": 0.0293,
+      "step": 939
+    },
+    {
+      "epoch": 0.5601907032181168,
+      "grad_norm": 2.611844062805176,
+      "learning_rate": 8.879618593563767e-06,
+      "loss": 0.0655,
+      "step": 940
+    },
+    {
+      "epoch": 0.5607866507747318,
+      "grad_norm": 2.735853672027588,
+      "learning_rate": 8.867699642431467e-06,
+      "loss": 0.0839,
+      "step": 941
+    },
+    {
+      "epoch": 0.5613825983313468,
+      "grad_norm": 4.26708984375,
+      "learning_rate": 8.855780691299166e-06,
+      "loss": 0.1646,
+      "step": 942
+    },
+    {
+      "epoch": 0.5619785458879618,
+      "grad_norm": 3.4753267765045166,
+      "learning_rate": 8.843861740166866e-06,
+      "loss": 0.1222,
+      "step": 943
+    },
+    {
+      "epoch": 0.5625744934445769,
+      "grad_norm": 5.538196563720703,
+      "learning_rate": 8.831942789034566e-06,
+      "loss": 0.2744,
+      "step": 944
+    },
+    {
+      "epoch": 0.5631704410011918,
+      "grad_norm": 4.448379039764404,
+      "learning_rate": 8.820023837902266e-06,
+      "loss": 0.2101,
+      "step": 945
+    },
+    {
+      "epoch": 0.5637663885578069,
+      "grad_norm": 2.9264047145843506,
+      "learning_rate": 8.808104886769965e-06,
+      "loss": 0.0733,
+      "step": 946
+    },
+    {
+      "epoch": 0.5643623361144219,
+      "grad_norm": 2.4338572025299072,
+      "learning_rate": 8.796185935637665e-06,
+      "loss": 0.0377,
+      "step": 947
+    },
+    {
+      "epoch": 0.564958283671037,
+      "grad_norm": 3.4269251823425293,
+      "learning_rate": 8.784266984505365e-06,
+      "loss": 0.1038,
+      "step": 948
+    },
+    {
+      "epoch": 0.5655542312276519,
+      "grad_norm": 2.5800302028656006,
+      "learning_rate": 8.772348033373065e-06,
+      "loss": 0.0551,
+      "step": 949
+    },
+    {
+      "epoch": 0.566150178784267,
+      "grad_norm": 3.4629178047180176,
+      "learning_rate": 8.760429082240763e-06,
+      "loss": 0.1321,
+      "step": 950
+    },
+    {
+      "epoch": 0.566746126340882,
+      "grad_norm": 3.4136455059051514,
+      "learning_rate": 8.748510131108462e-06,
+      "loss": 0.1572,
+      "step": 951
+    },
+    {
+      "epoch": 0.567342073897497,
+      "grad_norm": 1.0053603649139404,
+      "learning_rate": 8.736591179976162e-06,
+      "loss": 0.0121,
+      "step": 952
+    },
+    {
+      "epoch": 0.567938021454112,
+      "grad_norm": 1.4704949855804443,
+      "learning_rate": 8.724672228843862e-06,
+      "loss": 0.0236,
+      "step": 953
+    },
+    {
+      "epoch": 0.5685339690107271,
+      "grad_norm": 3.032944440841675,
+      "learning_rate": 8.712753277711562e-06,
+      "loss": 0.0339,
+      "step": 954
+    },
+    {
+      "epoch": 0.569129916567342,
+      "grad_norm": 2.986560344696045,
+      "learning_rate": 8.700834326579261e-06,
+      "loss": 0.0681,
+      "step": 955
+    },
+    {
+      "epoch": 0.5697258641239571,
+      "grad_norm": 4.405220985412598,
+      "learning_rate": 8.688915375446961e-06,
+      "loss": 0.2573,
+      "step": 956
+    },
+    {
+      "epoch": 0.570321811680572,
+      "grad_norm": 7.3769001960754395,
+      "learning_rate": 8.67699642431466e-06,
+      "loss": 0.4854,
+      "step": 957
+    },
+    {
+      "epoch": 0.5709177592371871,
+      "grad_norm": 7.263244152069092,
+      "learning_rate": 8.66507747318236e-06,
+      "loss": 0.5094,
+      "step": 958
+    },
+    {
+      "epoch": 0.5715137067938022,
+      "grad_norm": 3.1772360801696777,
+      "learning_rate": 8.65315852205006e-06,
+      "loss": 0.0756,
+      "step": 959
+    },
+    {
+      "epoch": 0.5721096543504172,
+      "grad_norm": 3.401954174041748,
+      "learning_rate": 8.64123957091776e-06,
+      "loss": 0.1251,
+      "step": 960
+    },
+    {
+      "epoch": 0.5727056019070322,
+      "grad_norm": 3.326629638671875,
+      "learning_rate": 8.62932061978546e-06,
+      "loss": 0.0864,
+      "step": 961
+    },
+    {
+      "epoch": 0.5733015494636472,
+      "grad_norm": 4.005965232849121,
+      "learning_rate": 8.61740166865316e-06,
+      "loss": 0.1729,
+      "step": 962
+    },
+    {
+      "epoch": 0.5738974970202623,
+      "grad_norm": 6.6543402671813965,
+      "learning_rate": 8.605482717520859e-06,
+      "loss": 0.3675,
+      "step": 963
+    },
+    {
+      "epoch": 0.5744934445768772,
+      "grad_norm": 2.849501132965088,
+      "learning_rate": 8.593563766388559e-06,
+      "loss": 0.0553,
+      "step": 964
+    },
+    {
+      "epoch": 0.5750893921334923,
+      "grad_norm": 3.096726894378662,
+      "learning_rate": 8.581644815256259e-06,
+      "loss": 0.0653,
+      "step": 965
+    },
+    {
+      "epoch": 0.5756853396901073,
+      "grad_norm": 5.49264669418335,
+      "learning_rate": 8.569725864123958e-06,
+      "loss": 0.3785,
+      "step": 966
+    },
+    {
+      "epoch": 0.5762812872467223,
+      "grad_norm": 2.8589844703674316,
+      "learning_rate": 8.557806912991658e-06,
+      "loss": 0.0395,
+      "step": 967
+    },
+    {
+      "epoch": 0.5768772348033373,
+      "grad_norm": 1.381852626800537,
+      "learning_rate": 8.545887961859358e-06,
+      "loss": 0.0117,
+      "step": 968
+    },
+    {
+      "epoch": 0.5774731823599524,
+      "grad_norm": 2.2843375205993652,
+      "learning_rate": 8.533969010727057e-06,
+      "loss": 0.0802,
+      "step": 969
+    },
+    {
+      "epoch": 0.5780691299165673,
+      "grad_norm": 2.7274208068847656,
+      "learning_rate": 8.522050059594757e-06,
+      "loss": 0.0703,
+      "step": 970
+    },
+    {
+      "epoch": 0.5786650774731824,
+      "grad_norm": 2.8992300033569336,
+      "learning_rate": 8.510131108462455e-06,
+      "loss": 0.074,
+      "step": 971
+    },
+    {
+      "epoch": 0.5792610250297974,
+      "grad_norm": 2.2076029777526855,
+      "learning_rate": 8.498212157330155e-06,
+      "loss": 0.0389,
+      "step": 972
+    },
+    {
+      "epoch": 0.5798569725864124,
+      "grad_norm": 2.3622515201568604,
+      "learning_rate": 8.486293206197855e-06,
+      "loss": 0.0795,
+      "step": 973
+    },
+    {
+      "epoch": 0.5804529201430274,
+      "grad_norm": 2.5536694526672363,
+      "learning_rate": 8.474374255065554e-06,
+      "loss": 0.0597,
+      "step": 974
+    },
+    {
+      "epoch": 0.5810488676996425,
+      "grad_norm": 2.97017502784729,
+      "learning_rate": 8.462455303933254e-06,
+      "loss": 0.0863,
+      "step": 975
+    },
+    {
+      "epoch": 0.5816448152562574,
+      "grad_norm": 2.2850182056427,
+      "learning_rate": 8.450536352800954e-06,
+      "loss": 0.0688,
+      "step": 976
+    },
+    {
+      "epoch": 0.5822407628128725,
+      "grad_norm": 5.303691864013672,
+      "learning_rate": 8.438617401668654e-06,
+      "loss": 0.3093,
+      "step": 977
+    },
+    {
+      "epoch": 0.5828367103694875,
+      "grad_norm": 2.183743953704834,
+      "learning_rate": 8.426698450536353e-06,
+      "loss": 0.1004,
+      "step": 978
+    },
+    {
+      "epoch": 0.5834326579261025,
+      "grad_norm": 2.4150192737579346,
+      "learning_rate": 8.414779499404053e-06,
+      "loss": 0.0905,
+      "step": 979
+    },
+    {
+      "epoch": 0.5840286054827175,
+      "grad_norm": 2.5931613445281982,
+      "learning_rate": 8.402860548271753e-06,
+      "loss": 0.0445,
+      "step": 980
+    },
+    {
+      "epoch": 0.5846245530393326,
+      "grad_norm": 3.2930822372436523,
+      "learning_rate": 8.390941597139453e-06,
+      "loss": 0.0768,
+      "step": 981
+    },
+    {
+      "epoch": 0.5852205005959475,
+      "grad_norm": 2.0512821674346924,
+      "learning_rate": 8.379022646007152e-06,
+      "loss": 0.039,
+      "step": 982
+    },
+    {
+      "epoch": 0.5858164481525626,
+      "grad_norm": 3.983140230178833,
+      "learning_rate": 8.367103694874852e-06,
+      "loss": 0.1713,
+      "step": 983
+    },
+    {
+      "epoch": 0.5864123957091776,
+      "grad_norm": 1.427815556526184,
+      "learning_rate": 8.355184743742552e-06,
+      "loss": 0.0241,
+      "step": 984
+    },
+    {
+      "epoch": 0.5870083432657927,
+      "grad_norm": 3.5479249954223633,
+      "learning_rate": 8.343265792610251e-06,
+      "loss": 0.1009,
+      "step": 985
+    },
+    {
+      "epoch": 0.5876042908224076,
+      "grad_norm": 3.082832098007202,
+      "learning_rate": 8.331346841477951e-06,
+      "loss": 0.0755,
+      "step": 986
+    },
+    {
+      "epoch": 0.5882002383790227,
+      "grad_norm": 2.122084856033325,
+      "learning_rate": 8.319427890345651e-06,
+      "loss": 0.064,
+      "step": 987
+    },
+    {
+      "epoch": 0.5887961859356377,
+      "grad_norm": 3.058931589126587,
+      "learning_rate": 8.30750893921335e-06,
+      "loss": 0.1336,
+      "step": 988
+    },
+    {
+      "epoch": 0.5893921334922527,
+      "grad_norm": 1.4611798524856567,
+      "learning_rate": 8.29558998808105e-06,
+      "loss": 0.0214,
+      "step": 989
+    },
+    {
+      "epoch": 0.5899880810488677,
+      "grad_norm": 2.0266125202178955,
+      "learning_rate": 8.28367103694875e-06,
+      "loss": 0.0478,
+      "step": 990
+    },
+    {
+      "epoch": 0.5905840286054828,
+      "grad_norm": 2.9834725856781006,
+      "learning_rate": 8.27175208581645e-06,
+      "loss": 0.1261,
+      "step": 991
+    },
+    {
+      "epoch": 0.5911799761620977,
+      "grad_norm": 3.0291907787323,
+      "learning_rate": 8.259833134684148e-06,
+      "loss": 0.0682,
+      "step": 992
+    },
+    {
+      "epoch": 0.5917759237187128,
+      "grad_norm": 1.4318104982376099,
+      "learning_rate": 8.247914183551848e-06,
+      "loss": 0.011,
+      "step": 993
+    },
+    {
+      "epoch": 0.5923718712753278,
+      "grad_norm": 5.025737285614014,
+      "learning_rate": 8.235995232419547e-06,
+      "loss": 0.1002,
+      "step": 994
+    },
+    {
+      "epoch": 0.5929678188319428,
+      "grad_norm": 0.9305858612060547,
+      "learning_rate": 8.224076281287247e-06,
+      "loss": 0.0114,
+      "step": 995
+    },
+    {
+      "epoch": 0.5935637663885578,
+      "grad_norm": 3.4280264377593994,
+      "learning_rate": 8.212157330154947e-06,
+      "loss": 0.1473,
+      "step": 996
+    },
+    {
+      "epoch": 0.5941597139451729,
+      "grad_norm": 5.026422023773193,
+      "learning_rate": 8.200238379022647e-06,
+      "loss": 0.259,
+      "step": 997
+    },
+    {
+      "epoch": 0.5947556615017878,
+      "grad_norm": 3.4882593154907227,
+      "learning_rate": 8.188319427890346e-06,
+      "loss": 0.0881,
+      "step": 998
+    },
+    {
+      "epoch": 0.5953516090584029,
+      "grad_norm": 2.6533875465393066,
+      "learning_rate": 8.176400476758046e-06,
+      "loss": 0.0529,
+      "step": 999
+    },
+    {
+      "epoch": 0.5959475566150179,
+      "grad_norm": 2.4922659397125244,
+      "learning_rate": 8.164481525625746e-06,
+      "loss": 0.0646,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5965435041716329,
+      "grad_norm": 2.8225927352905273,
+      "learning_rate": 8.152562574493445e-06,
+      "loss": 0.0359,
+      "step": 1001
+    },
+    {
+      "epoch": 0.5971394517282479,
+      "grad_norm": 2.3172740936279297,
+      "learning_rate": 8.140643623361145e-06,
+      "loss": 0.0564,
+      "step": 1002
+    },
+    {
+      "epoch": 0.597735399284863,
+      "grad_norm": 3.772636890411377,
+      "learning_rate": 8.128724672228845e-06,
+      "loss": 0.1983,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5983313468414779,
+      "grad_norm": 2.7776310443878174,
+      "learning_rate": 8.116805721096545e-06,
+      "loss": 0.0838,
+      "step": 1004
+    },
+    {
+      "epoch": 0.598927294398093,
+      "grad_norm": 1.278112769126892,
+      "learning_rate": 8.104886769964244e-06,
+      "loss": 0.0107,
+      "step": 1005
+    },
+    {
+      "epoch": 0.599523241954708,
+      "grad_norm": 4.212291717529297,
+      "learning_rate": 8.092967818831944e-06,
+      "loss": 0.1457,
+      "step": 1006
+    },
+    {
+      "epoch": 0.600119189511323,
+      "grad_norm": 3.647627115249634,
+      "learning_rate": 8.081048867699644e-06,
+      "loss": 0.076,
+      "step": 1007
+    },
+    {
+      "epoch": 0.600715137067938,
+      "grad_norm": 3.2856578826904297,
+      "learning_rate": 8.069129916567344e-06,
+      "loss": 0.0668,
+      "step": 1008
+    },
+    {
+      "epoch": 0.6013110846245531,
+      "grad_norm": 4.387614727020264,
+      "learning_rate": 8.057210965435043e-06,
+      "loss": 0.2219,
+      "step": 1009
+    },
+    {
+      "epoch": 0.601907032181168,
+      "grad_norm": 4.314347743988037,
+      "learning_rate": 8.045292014302743e-06,
+      "loss": 0.1428,
+      "step": 1010
+    },
+    {
+      "epoch": 0.6025029797377831,
+      "grad_norm": 2.286726713180542,
+      "learning_rate": 8.033373063170443e-06,
+      "loss": 0.067,
+      "step": 1011
+    },
+    {
+      "epoch": 0.6030989272943981,
+      "grad_norm": 1.9418418407440186,
+      "learning_rate": 8.021454112038142e-06,
+      "loss": 0.0361,
+      "step": 1012
+    },
+    {
+      "epoch": 0.6036948748510131,
+      "grad_norm": 3.5930075645446777,
+      "learning_rate": 8.00953516090584e-06,
+      "loss": 0.0838,
+      "step": 1013
+    },
+    {
+      "epoch": 0.6042908224076281,
+      "grad_norm": 4.444027900695801,
+      "learning_rate": 7.99761620977354e-06,
+      "loss": 0.2832,
+      "step": 1014
+    },
+    {
+      "epoch": 0.6048867699642432,
+      "grad_norm": 3.3959922790527344,
+      "learning_rate": 7.98569725864124e-06,
+      "loss": 0.1136,
+      "step": 1015
+    },
+    {
+      "epoch": 0.6054827175208581,
+      "grad_norm": 3.846806526184082,
+      "learning_rate": 7.97377830750894e-06,
+      "loss": 0.1172,
+      "step": 1016
+    },
+    {
+      "epoch": 0.6060786650774732,
+      "grad_norm": 5.049923419952393,
+      "learning_rate": 7.96185935637664e-06,
+      "loss": 0.411,
+      "step": 1017
+    },
+    {
+      "epoch": 0.6066746126340882,
+      "grad_norm": 5.467438220977783,
+      "learning_rate": 7.94994040524434e-06,
+      "loss": 0.3307,
+      "step": 1018
+    },
+    {
+      "epoch": 0.6072705601907032,
+      "grad_norm": 1.621523141860962,
+      "learning_rate": 7.938021454112039e-06,
+      "loss": 0.0245,
+      "step": 1019
+    },
+    {
+      "epoch": 0.6078665077473182,
+      "grad_norm": 5.420456409454346,
+      "learning_rate": 7.926102502979739e-06,
+      "loss": 0.3369,
+      "step": 1020
+    },
+    {
+      "epoch": 0.6084624553039333,
+      "grad_norm": 3.2818987369537354,
+      "learning_rate": 7.914183551847438e-06,
+      "loss": 0.1501,
+      "step": 1021
+    },
+    {
+      "epoch": 0.6090584028605482,
+      "grad_norm": 1.9436466693878174,
+      "learning_rate": 7.902264600715138e-06,
+      "loss": 0.0194,
+      "step": 1022
+    },
+    {
+      "epoch": 0.6096543504171633,
+      "grad_norm": 3.301459550857544,
+      "learning_rate": 7.890345649582838e-06,
+      "loss": 0.0929,
+      "step": 1023
+    },
+    {
+      "epoch": 0.6102502979737783,
+      "grad_norm": 5.303291320800781,
+      "learning_rate": 7.878426698450538e-06,
+      "loss": 0.1891,
+      "step": 1024
+    },
+    {
+      "epoch": 0.6108462455303934,
+      "grad_norm": 4.092094898223877,
+      "learning_rate": 7.866507747318237e-06,
+      "loss": 0.1446,
+      "step": 1025
+    },
+    {
+      "epoch": 0.6114421930870083,
+      "grad_norm": 2.92805814743042,
+      "learning_rate": 7.854588796185935e-06,
+      "loss": 0.0885,
+      "step": 1026
+    },
+    {
+      "epoch": 0.6120381406436234,
+      "grad_norm": 2.568646192550659,
+      "learning_rate": 7.842669845053635e-06,
+      "loss": 0.0496,
+      "step": 1027
+    },
+    {
+      "epoch": 0.6126340882002383,
+      "grad_norm": 1.2362128496170044,
+      "learning_rate": 7.830750893921335e-06,
+      "loss": 0.0133,
+      "step": 1028
+    },
+    {
+      "epoch": 0.6132300357568534,
+      "grad_norm": 5.4793572425842285,
+      "learning_rate": 7.818831942789035e-06,
+      "loss": 0.2841,
+      "step": 1029
+    },
+    {
+      "epoch": 0.6138259833134684,
+      "grad_norm": 5.73588752746582,
+      "learning_rate": 7.806912991656734e-06,
+      "loss": 0.3436,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6144219308700835,
+      "grad_norm": 2.729397773742676,
+      "learning_rate": 7.794994040524434e-06,
+      "loss": 0.0396,
+      "step": 1031
+    },
+    {
+      "epoch": 0.6150178784266984,
+      "grad_norm": 3.307756185531616,
+      "learning_rate": 7.783075089392134e-06,
+      "loss": 0.097,
+      "step": 1032
+    },
+    {
+      "epoch": 0.6156138259833135,
+      "grad_norm": 2.5629005432128906,
+      "learning_rate": 7.771156138259833e-06,
+      "loss": 0.0616,
+      "step": 1033
+    },
+    {
+      "epoch": 0.6162097735399285,
+      "grad_norm": 2.6664211750030518,
+      "learning_rate": 7.759237187127533e-06,
+      "loss": 0.0691,
+      "step": 1034
+    },
+    {
+      "epoch": 0.6168057210965435,
+      "grad_norm": 2.790714979171753,
+      "learning_rate": 7.747318235995233e-06,
+      "loss": 0.0419,
+      "step": 1035
+    },
+    {
+      "epoch": 0.6174016686531585,
+      "grad_norm": 2.87893009185791,
+      "learning_rate": 7.735399284862933e-06,
+      "loss": 0.0644,
+      "step": 1036
+    },
+    {
+      "epoch": 0.6179976162097736,
+      "grad_norm": 3.411217451095581,
+      "learning_rate": 7.723480333730632e-06,
+      "loss": 0.1082,
+      "step": 1037
+    },
+    {
+      "epoch": 0.6185935637663885,
+      "grad_norm": 3.5509350299835205,
+      "learning_rate": 7.711561382598332e-06,
+      "loss": 0.1144,
+      "step": 1038
+    },
+    {
+      "epoch": 0.6191895113230036,
+      "grad_norm": 3.757441997528076,
+      "learning_rate": 7.699642431466032e-06,
+      "loss": 0.1904,
+      "step": 1039
+    },
+    {
+      "epoch": 0.6197854588796186,
+      "grad_norm": 3.2949378490448,
+      "learning_rate": 7.687723480333732e-06,
+      "loss": 0.1345,
+      "step": 1040
+    },
+    {
+      "epoch": 0.6203814064362336,
+      "grad_norm": 1.231120228767395,
+      "learning_rate": 7.675804529201431e-06,
+      "loss": 0.0213,
+      "step": 1041
+    },
+    {
+      "epoch": 0.6209773539928486,
+      "grad_norm": 5.333119869232178,
+      "learning_rate": 7.66388557806913e-06,
+      "loss": 0.3259,
+      "step": 1042
+    },
+    {
+      "epoch": 0.6215733015494637,
+      "grad_norm": 4.492021560668945,
+      "learning_rate": 7.651966626936829e-06,
+      "loss": 0.2067,
+      "step": 1043
+    },
+    {
+      "epoch": 0.6221692491060786,
+      "grad_norm": 2.070573091506958,
+      "learning_rate": 7.640047675804529e-06,
+      "loss": 0.055,
+      "step": 1044
+    },
+    {
+      "epoch": 0.6227651966626937,
+      "grad_norm": 4.437075138092041,
+      "learning_rate": 7.628128724672229e-06,
+      "loss": 0.1695,
+      "step": 1045
+    },
+    {
+      "epoch": 0.6233611442193087,
+      "grad_norm": 2.7593233585357666,
+      "learning_rate": 7.616209773539929e-06,
+      "loss": 0.0671,
+      "step": 1046
+    },
+    {
+      "epoch": 0.6239570917759237,
+      "grad_norm": 4.070964813232422,
+      "learning_rate": 7.604290822407628e-06,
+      "loss": 0.1926,
+      "step": 1047
+    },
+    {
+      "epoch": 0.6245530393325387,
+      "grad_norm": 3.0797886848449707,
+      "learning_rate": 7.592371871275328e-06,
+      "loss": 0.0743,
+      "step": 1048
+    },
+    {
+      "epoch": 0.6251489868891538,
+      "grad_norm": 3.8007187843322754,
+      "learning_rate": 7.5804529201430274e-06,
+      "loss": 0.1122,
+      "step": 1049
+    },
+    {
+      "epoch": 0.6257449344457687,
+      "grad_norm": 3.06937313079834,
+      "learning_rate": 7.568533969010727e-06,
+      "loss": 0.0462,
+      "step": 1050
+    },
+    {
+      "epoch": 0.6263408820023838,
+      "grad_norm": 1.294691801071167,
+      "learning_rate": 7.556615017878427e-06,
+      "loss": 0.0146,
+      "step": 1051
+    },
+    {
+      "epoch": 0.6269368295589988,
+      "grad_norm": 3.303410053253174,
+      "learning_rate": 7.544696066746127e-06,
+      "loss": 0.0543,
+      "step": 1052
+    },
+    {
+      "epoch": 0.6275327771156138,
+      "grad_norm": 4.026877403259277,
+      "learning_rate": 7.532777115613826e-06,
+      "loss": 0.229,
+      "step": 1053
+    },
+    {
+      "epoch": 0.6281287246722288,
+      "grad_norm": 2.8060317039489746,
+      "learning_rate": 7.520858164481526e-06,
+      "loss": 0.0748,
+      "step": 1054
+    },
+    {
+      "epoch": 0.6287246722288439,
+      "grad_norm": 2.307131052017212,
+      "learning_rate": 7.508939213349226e-06,
+      "loss": 0.0734,
+      "step": 1055
+    },
+    {
+      "epoch": 0.6293206197854588,
+      "grad_norm": 4.481800556182861,
+      "learning_rate": 7.4970202622169255e-06,
+      "loss": 0.0756,
+      "step": 1056
+    },
+    {
+      "epoch": 0.6299165673420739,
+      "grad_norm": 4.39906644821167,
+      "learning_rate": 7.485101311084625e-06,
+      "loss": 0.1561,
+      "step": 1057
+    },
+    {
+      "epoch": 0.6305125148986889,
+      "grad_norm": 2.539372205734253,
+      "learning_rate": 7.473182359952324e-06,
+      "loss": 0.0735,
+      "step": 1058
+    },
+    {
+      "epoch": 0.6311084624553039,
+      "grad_norm": 4.0483551025390625,
+      "learning_rate": 7.461263408820024e-06,
+      "loss": 0.1165,
+      "step": 1059
+    },
+    {
+      "epoch": 0.6317044100119189,
+      "grad_norm": 2.972843647003174,
+      "learning_rate": 7.449344457687724e-06,
+      "loss": 0.0559,
+      "step": 1060
+    },
+    {
+      "epoch": 0.632300357568534,
+      "grad_norm": 4.183932304382324,
+      "learning_rate": 7.437425506555423e-06,
+      "loss": 0.2673,
+      "step": 1061
+    },
+    {
+      "epoch": 0.6328963051251489,
+      "grad_norm": 5.052440643310547,
+      "learning_rate": 7.425506555423123e-06,
+      "loss": 0.3387,
+      "step": 1062
+    },
+    {
+      "epoch": 0.633492252681764,
+      "grad_norm": 4.317245006561279,
+      "learning_rate": 7.413587604290823e-06,
+      "loss": 0.1971,
+      "step": 1063
+    },
+    {
+      "epoch": 0.634088200238379,
+      "grad_norm": 2.183413028717041,
+      "learning_rate": 7.4016686531585225e-06,
+      "loss": 0.0396,
+      "step": 1064
+    },
+    {
+      "epoch": 0.634684147794994,
+      "grad_norm": 1.8442902565002441,
+      "learning_rate": 7.389749702026222e-06,
+      "loss": 0.0333,
+      "step": 1065
+    },
+    {
+      "epoch": 0.635280095351609,
+      "grad_norm": 3.147921323776245,
+      "learning_rate": 7.377830750893922e-06,
+      "loss": 0.0631,
+      "step": 1066
+    },
+    {
+      "epoch": 0.6358760429082241,
+      "grad_norm": 2.4488799571990967,
+      "learning_rate": 7.365911799761622e-06,
+      "loss": 0.0876,
+      "step": 1067
+    },
+    {
+      "epoch": 0.636471990464839,
+      "grad_norm": 2.3631789684295654,
+      "learning_rate": 7.3539928486293215e-06,
+      "loss": 0.0437,
+      "step": 1068
+    },
+    {
+      "epoch": 0.6370679380214541,
+      "grad_norm": 2.2388243675231934,
+      "learning_rate": 7.34207389749702e-06,
+      "loss": 0.0485,
+      "step": 1069
+    },
+    {
+      "epoch": 0.6376638855780691,
+      "grad_norm": 3.376847267150879,
+      "learning_rate": 7.33015494636472e-06,
+      "loss": 0.092,
+      "step": 1070
+    },
+    {
+      "epoch": 0.6382598331346842,
+      "grad_norm": 1.2180335521697998,
+      "learning_rate": 7.31823599523242e-06,
+      "loss": 0.0237,
+      "step": 1071
+    },
+    {
+      "epoch": 0.6388557806912991,
+      "grad_norm": 2.344777822494507,
+      "learning_rate": 7.3063170441001195e-06,
+      "loss": 0.0451,
+      "step": 1072
+    },
+    {
+      "epoch": 0.6394517282479142,
+      "grad_norm": 4.088088512420654,
+      "learning_rate": 7.294398092967819e-06,
+      "loss": 0.1247,
+      "step": 1073
+    },
+    {
+      "epoch": 0.6400476758045291,
+      "grad_norm": 2.0176923274993896,
+      "learning_rate": 7.282479141835519e-06,
+      "loss": 0.0327,
+      "step": 1074
+    },
+    {
+      "epoch": 0.6406436233611442,
+      "grad_norm": 2.024414300918579,
+      "learning_rate": 7.270560190703219e-06,
+      "loss": 0.0488,
+      "step": 1075
+    },
+    {
+      "epoch": 0.6412395709177592,
+      "grad_norm": 1.8544275760650635,
+      "learning_rate": 7.2586412395709185e-06,
+      "loss": 0.045,
+      "step": 1076
+    },
+    {
+      "epoch": 0.6418355184743743,
+      "grad_norm": 7.20472526550293,
+      "learning_rate": 7.246722288438618e-06,
+      "loss": 0.7238,
+      "step": 1077
+    },
+    {
+      "epoch": 0.6424314660309892,
+      "grad_norm": 2.909888982772827,
+      "learning_rate": 7.234803337306318e-06,
+      "loss": 0.1086,
+      "step": 1078
+    },
+    {
+      "epoch": 0.6430274135876043,
+      "grad_norm": 1.6136006116867065,
+      "learning_rate": 7.222884386174017e-06,
+      "loss": 0.016,
+      "step": 1079
+    },
+    {
+      "epoch": 0.6436233611442194,
+      "grad_norm": 5.893424034118652,
+      "learning_rate": 7.2109654350417165e-06,
+      "loss": 0.3131,
+      "step": 1080
+    },
+    {
+      "epoch": 0.6442193087008343,
+      "grad_norm": 2.5028076171875,
+      "learning_rate": 7.199046483909416e-06,
+      "loss": 0.0402,
+      "step": 1081
+    },
+    {
+      "epoch": 0.6448152562574494,
+      "grad_norm": 4.470302581787109,
+      "learning_rate": 7.187127532777116e-06,
+      "loss": 0.1435,
+      "step": 1082
+    },
+    {
+      "epoch": 0.6454112038140644,
+      "grad_norm": 2.2946817874908447,
+      "learning_rate": 7.175208581644816e-06,
+      "loss": 0.0581,
+      "step": 1083
+    },
+    {
+      "epoch": 0.6460071513706794,
+      "grad_norm": 2.486341714859009,
+      "learning_rate": 7.1632896305125155e-06,
+      "loss": 0.0445,
+      "step": 1084
+    },
+    {
+      "epoch": 0.6466030989272944,
+      "grad_norm": 3.089010238647461,
+      "learning_rate": 7.151370679380215e-06,
+      "loss": 0.1217,
+      "step": 1085
+    },
+    {
+      "epoch": 0.6471990464839095,
+      "grad_norm": 2.1066575050354004,
+      "learning_rate": 7.139451728247915e-06,
+      "loss": 0.0433,
+      "step": 1086
+    },
+    {
+      "epoch": 0.6477949940405244,
+      "grad_norm": 6.139880180358887,
+      "learning_rate": 7.127532777115615e-06,
+      "loss": 0.287,
+      "step": 1087
+    },
+    {
+      "epoch": 0.6483909415971395,
+      "grad_norm": 3.1939146518707275,
+      "learning_rate": 7.115613825983314e-06,
+      "loss": 0.0744,
+      "step": 1088
+    },
+    {
+      "epoch": 0.6489868891537545,
+      "grad_norm": 2.6073572635650635,
+      "learning_rate": 7.103694874851014e-06,
+      "loss": 0.0496,
+      "step": 1089
+    },
+    {
+      "epoch": 0.6495828367103695,
+      "grad_norm": 1.9987437725067139,
+      "learning_rate": 7.091775923718713e-06,
+      "loss": 0.0299,
+      "step": 1090
+    },
+    {
+      "epoch": 0.6501787842669845,
+      "grad_norm": 1.4185389280319214,
+      "learning_rate": 7.079856972586413e-06,
+      "loss": 0.0226,
+      "step": 1091
+    },
+    {
+      "epoch": 0.6507747318235996,
+      "grad_norm": 3.0944974422454834,
+      "learning_rate": 7.0679380214541124e-06,
+      "loss": 0.1,
+      "step": 1092
+    },
+    {
+      "epoch": 0.6513706793802145,
+      "grad_norm": 2.222912549972534,
+      "learning_rate": 7.056019070321812e-06,
+      "loss": 0.0622,
+      "step": 1093
+    },
+    {
+      "epoch": 0.6519666269368296,
+      "grad_norm": 3.5529370307922363,
+      "learning_rate": 7.044100119189512e-06,
+      "loss": 0.0803,
+      "step": 1094
+    },
+    {
+      "epoch": 0.6525625744934446,
+      "grad_norm": 3.151442289352417,
+      "learning_rate": 7.032181168057212e-06,
+      "loss": 0.0744,
+      "step": 1095
+    },
+    {
+      "epoch": 0.6531585220500596,
+      "grad_norm": 3.739518165588379,
+      "learning_rate": 7.020262216924911e-06,
+      "loss": 0.1443,
+      "step": 1096
+    },
+    {
+      "epoch": 0.6537544696066746,
+      "grad_norm": 2.3120760917663574,
+      "learning_rate": 7.008343265792611e-06,
+      "loss": 0.05,
+      "step": 1097
+    },
+    {
+      "epoch": 0.6543504171632897,
+      "grad_norm": 2.5826151371002197,
+      "learning_rate": 6.996424314660311e-06,
+      "loss": 0.0748,
+      "step": 1098
+    },
+    {
+      "epoch": 0.6549463647199046,
+      "grad_norm": 3.638197660446167,
+      "learning_rate": 6.9845053635280106e-06,
+      "loss": 0.1448,
+      "step": 1099
+    },
+    {
+      "epoch": 0.6555423122765197,
+      "grad_norm": 2.879312515258789,
+      "learning_rate": 6.97258641239571e-06,
+      "loss": 0.0607,
+      "step": 1100
+    },
+    {
+      "epoch": 0.6561382598331347,
+      "grad_norm": 2.5166313648223877,
+      "learning_rate": 6.960667461263409e-06,
+      "loss": 0.0553,
+      "step": 1101
+    },
+    {
+      "epoch": 0.6567342073897497,
+      "grad_norm": 3.6005442142486572,
+      "learning_rate": 6.948748510131109e-06,
+      "loss": 0.1566,
+      "step": 1102
+    },
+    {
+      "epoch": 0.6573301549463647,
+      "grad_norm": 3.298604726791382,
+      "learning_rate": 6.936829558998809e-06,
+      "loss": 0.1164,
+      "step": 1103
+    },
+    {
+      "epoch": 0.6579261025029798,
+      "grad_norm": 3.7708985805511475,
+      "learning_rate": 6.924910607866508e-06,
+      "loss": 0.1973,
+      "step": 1104
+    },
+    {
+      "epoch": 0.6585220500595947,
+      "grad_norm": 2.5936386585235596,
+      "learning_rate": 6.912991656734208e-06,
+      "loss": 0.0488,
+      "step": 1105
+    },
+    {
+      "epoch": 0.6591179976162098,
+      "grad_norm": 3.5538878440856934,
+      "learning_rate": 6.901072705601908e-06,
+      "loss": 0.1146,
+      "step": 1106
+    },
+    {
+      "epoch": 0.6597139451728248,
+      "grad_norm": 2.4304397106170654,
+      "learning_rate": 6.8891537544696076e-06,
+      "loss": 0.057,
+      "step": 1107
+    },
+    {
+      "epoch": 0.6603098927294399,
+      "grad_norm": 3.828990936279297,
+      "learning_rate": 6.877234803337307e-06,
+      "loss": 0.0736,
+      "step": 1108
+    },
+    {
+      "epoch": 0.6609058402860548,
+      "grad_norm": 3.4269118309020996,
+      "learning_rate": 6.865315852205007e-06,
+      "loss": 0.1351,
+      "step": 1109
+    },
+    {
+      "epoch": 0.6615017878426699,
+      "grad_norm": 4.157618999481201,
+      "learning_rate": 6.853396901072707e-06,
+      "loss": 0.1825,
+      "step": 1110
+    },
+    {
+      "epoch": 0.6620977353992848,
+      "grad_norm": 2.10422420501709,
+      "learning_rate": 6.841477949940406e-06,
+      "loss": 0.0404,
+      "step": 1111
+    },
+    {
+      "epoch": 0.6626936829558999,
+      "grad_norm": 2.9045560359954834,
+      "learning_rate": 6.829558998808105e-06,
+      "loss": 0.075,
+      "step": 1112
+    },
+    {
+      "epoch": 0.6632896305125149,
+      "grad_norm": 1.960497498512268,
+      "learning_rate": 6.817640047675805e-06,
+      "loss": 0.021,
+      "step": 1113
+    },
+    {
+      "epoch": 0.66388557806913,
+      "grad_norm": 10.49171257019043,
+      "learning_rate": 6.805721096543505e-06,
+      "loss": 1.0764,
+      "step": 1114
+    },
+    {
+      "epoch": 0.6644815256257449,
+      "grad_norm": 5.002450942993164,
+      "learning_rate": 6.7938021454112046e-06,
+      "loss": 0.22,
+      "step": 1115
+    },
+    {
+      "epoch": 0.66507747318236,
+      "grad_norm": 4.224698066711426,
+      "learning_rate": 6.781883194278904e-06,
+      "loss": 0.2691,
+      "step": 1116
+    },
+    {
+      "epoch": 0.665673420738975,
+      "grad_norm": 1.7933399677276611,
+      "learning_rate": 6.769964243146604e-06,
+      "loss": 0.0261,
+      "step": 1117
+    },
+    {
+      "epoch": 0.66626936829559,
+      "grad_norm": 3.6782877445220947,
+      "learning_rate": 6.758045292014304e-06,
+      "loss": 0.1116,
+      "step": 1118
+    },
+    {
+      "epoch": 0.666865315852205,
+      "grad_norm": 2.1333539485931396,
+      "learning_rate": 6.7461263408820035e-06,
+      "loss": 0.0353,
+      "step": 1119
+    },
+    {
+      "epoch": 0.6674612634088201,
+      "grad_norm": 6.434367656707764,
+      "learning_rate": 6.734207389749703e-06,
+      "loss": 0.4895,
+      "step": 1120
+    },
+    {
+      "epoch": 0.668057210965435,
+      "grad_norm": 2.36881685256958,
+      "learning_rate": 6.722288438617403e-06,
+      "loss": 0.074,
+      "step": 1121
+    },
+    {
+      "epoch": 0.6686531585220501,
+      "grad_norm": 3.0191922187805176,
+      "learning_rate": 6.710369487485102e-06,
+      "loss": 0.0438,
+      "step": 1122
+    },
+    {
+      "epoch": 0.6692491060786651,
+      "grad_norm": 3.490842819213867,
+      "learning_rate": 6.6984505363528015e-06,
+      "loss": 0.123,
+      "step": 1123
+    },
+    {
+      "epoch": 0.6698450536352801,
+      "grad_norm": 2.7544801235198975,
+      "learning_rate": 6.686531585220501e-06,
+      "loss": 0.0349,
+      "step": 1124
+    },
+    {
+      "epoch": 0.6704410011918951,
+      "grad_norm": 3.816267251968384,
+      "learning_rate": 6.674612634088201e-06,
+      "loss": 0.1585,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6710369487485102,
+      "grad_norm": 2.1254501342773438,
+      "learning_rate": 6.662693682955901e-06,
+      "loss": 0.0397,
+      "step": 1126
+    },
+    {
+      "epoch": 0.6716328963051251,
+      "grad_norm": 3.215665102005005,
+      "learning_rate": 6.6507747318236005e-06,
+      "loss": 0.1017,
+      "step": 1127
+    },
+    {
+      "epoch": 0.6722288438617402,
+      "grad_norm": 2.086947202682495,
+      "learning_rate": 6.6388557806913e-06,
+      "loss": 0.046,
+      "step": 1128
+    },
+    {
+      "epoch": 0.6728247914183552,
+      "grad_norm": 2.104255437850952,
+      "learning_rate": 6.626936829559e-06,
+      "loss": 0.063,
+      "step": 1129
+    },
+    {
+      "epoch": 0.6734207389749702,
+      "grad_norm": 5.049071788787842,
+      "learning_rate": 6.6150178784267e-06,
+      "loss": 0.2939,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6740166865315852,
+      "grad_norm": 3.7174763679504395,
+      "learning_rate": 6.603098927294399e-06,
+      "loss": 0.1097,
+      "step": 1131
+    },
+    {
+      "epoch": 0.6746126340882003,
+      "grad_norm": 5.160399913787842,
+      "learning_rate": 6.591179976162099e-06,
+      "loss": 0.2053,
+      "step": 1132
+    },
+    {
+      "epoch": 0.6752085816448152,
+      "grad_norm": 1.5703755617141724,
+      "learning_rate": 6.579261025029798e-06,
+      "loss": 0.028,
+      "step": 1133
+    },
+    {
+      "epoch": 0.6758045292014303,
+      "grad_norm": 4.515401363372803,
+      "learning_rate": 6.567342073897498e-06,
+      "loss": 0.3906,
+      "step": 1134
+    },
+    {
+      "epoch": 0.6764004767580453,
+      "grad_norm": 5.331261157989502,
+      "learning_rate": 6.5554231227651975e-06,
+      "loss": 0.1932,
+      "step": 1135
+    },
+    {
+      "epoch": 0.6769964243146603,
+      "grad_norm": 3.3486618995666504,
+      "learning_rate": 6.543504171632897e-06,
+      "loss": 0.0891,
+      "step": 1136
+    },
+    {
+      "epoch": 0.6775923718712753,
+      "grad_norm": 2.135518789291382,
+      "learning_rate": 6.531585220500597e-06,
+      "loss": 0.0305,
+      "step": 1137
+    },
+    {
+      "epoch": 0.6781883194278904,
+      "grad_norm": 1.9690325260162354,
+      "learning_rate": 6.519666269368297e-06,
+      "loss": 0.0278,
+      "step": 1138
+    },
+    {
+      "epoch": 0.6787842669845053,
+      "grad_norm": 2.266768217086792,
+      "learning_rate": 6.507747318235996e-06,
+      "loss": 0.0271,
+      "step": 1139
+    },
+    {
+      "epoch": 0.6793802145411204,
+      "grad_norm": 3.0908663272857666,
+      "learning_rate": 6.495828367103696e-06,
+      "loss": 0.061,
+      "step": 1140
+    },
+    {
+      "epoch": 0.6799761620977354,
+      "grad_norm": 2.4925196170806885,
+      "learning_rate": 6.483909415971396e-06,
+      "loss": 0.0903,
+      "step": 1141
+    },
+    {
+      "epoch": 0.6805721096543504,
+      "grad_norm": 3.298534631729126,
+      "learning_rate": 6.4719904648390956e-06,
+      "loss": 0.0868,
+      "step": 1142
+    },
+    {
+      "epoch": 0.6811680572109654,
+      "grad_norm": 5.865818977355957,
+      "learning_rate": 6.4600715137067945e-06,
+      "loss": 0.4865,
+      "step": 1143
+    },
+    {
+      "epoch": 0.6817640047675805,
+      "grad_norm": 4.295893669128418,
+      "learning_rate": 6.448152562574494e-06,
+      "loss": 0.0509,
+      "step": 1144
+    },
+    {
+      "epoch": 0.6823599523241954,
+      "grad_norm": 2.810593843460083,
+      "learning_rate": 6.436233611442194e-06,
+      "loss": 0.058,
+      "step": 1145
+    },
+    {
+      "epoch": 0.6829558998808105,
+      "grad_norm": 3.281193256378174,
+      "learning_rate": 6.424314660309893e-06,
+      "loss": 0.0936,
+      "step": 1146
+    },
+    {
+      "epoch": 0.6835518474374255,
+      "grad_norm": 3.481616735458374,
+      "learning_rate": 6.4123957091775925e-06,
+      "loss": 0.1252,
+      "step": 1147
+    },
+    {
+      "epoch": 0.6841477949940405,
+      "grad_norm": 5.1532063484191895,
+      "learning_rate": 6.400476758045292e-06,
+      "loss": 0.1758,
+      "step": 1148
+    },
+    {
+      "epoch": 0.6847437425506555,
+      "grad_norm": 2.2513818740844727,
+      "learning_rate": 6.388557806912992e-06,
+      "loss": 0.0465,
+      "step": 1149
+    },
+    {
+      "epoch": 0.6853396901072706,
+      "grad_norm": 2.7745208740234375,
+      "learning_rate": 6.376638855780692e-06,
+      "loss": 0.0643,
+      "step": 1150
+    },
+    {
+      "epoch": 0.6859356376638855,
+      "grad_norm": 3.495253562927246,
+      "learning_rate": 6.364719904648391e-06,
+      "loss": 0.062,
+      "step": 1151
+    },
+    {
+      "epoch": 0.6865315852205006,
+      "grad_norm": 2.093095302581787,
+      "learning_rate": 6.35280095351609e-06,
+      "loss": 0.0283,
+      "step": 1152
+    },
+    {
+      "epoch": 0.6871275327771156,
+      "grad_norm": 3.031529426574707,
+      "learning_rate": 6.34088200238379e-06,
+      "loss": 0.1104,
+      "step": 1153
+    },
+    {
+      "epoch": 0.6877234803337307,
+      "grad_norm": 6.3146185874938965,
+      "learning_rate": 6.32896305125149e-06,
+      "loss": 0.3832,
+      "step": 1154
+    },
+    {
+      "epoch": 0.6883194278903456,
+      "grad_norm": 2.0355019569396973,
+      "learning_rate": 6.3170441001191895e-06,
+      "loss": 0.0335,
+      "step": 1155
+    },
+    {
+      "epoch": 0.6889153754469607,
+      "grad_norm": 2.2712013721466064,
+      "learning_rate": 6.305125148986889e-06,
+      "loss": 0.049,
+      "step": 1156
+    },
+    {
+      "epoch": 0.6895113230035756,
+      "grad_norm": 2.6364786624908447,
+      "learning_rate": 6.293206197854589e-06,
+      "loss": 0.0659,
+      "step": 1157
+    },
+    {
+      "epoch": 0.6901072705601907,
+      "grad_norm": 4.340022087097168,
+      "learning_rate": 6.281287246722289e-06,
+      "loss": 0.1584,
+      "step": 1158
+    },
+    {
+      "epoch": 0.6907032181168057,
+      "grad_norm": 6.797107219696045,
+      "learning_rate": 6.2693682955899884e-06,
+      "loss": 0.4719,
+      "step": 1159
+    },
+    {
+      "epoch": 0.6912991656734208,
+      "grad_norm": 1.3068870306015015,
+      "learning_rate": 6.257449344457688e-06,
+      "loss": 0.0245,
+      "step": 1160
+    },
+    {
+      "epoch": 0.6918951132300357,
+      "grad_norm": 2.322540521621704,
+      "learning_rate": 6.245530393325387e-06,
+      "loss": 0.0243,
+      "step": 1161
+    },
+    {
+      "epoch": 0.6924910607866508,
+      "grad_norm": 2.8117356300354004,
+      "learning_rate": 6.233611442193087e-06,
+      "loss": 0.0738,
+      "step": 1162
+    },
+    {
+      "epoch": 0.6930870083432658,
+      "grad_norm": 3.0581936836242676,
+      "learning_rate": 6.2216924910607865e-06,
+      "loss": 0.0407,
+      "step": 1163
+    },
+    {
+      "epoch": 0.6936829558998808,
+      "grad_norm": 3.5459225177764893,
+      "learning_rate": 6.209773539928486e-06,
+      "loss": 0.0898,
+      "step": 1164
+    },
+    {
+      "epoch": 0.6942789034564958,
+      "grad_norm": 0.9564624428749084,
+      "learning_rate": 6.197854588796186e-06,
+      "loss": 0.0135,
+      "step": 1165
+    },
+    {
+      "epoch": 0.6948748510131109,
+      "grad_norm": 2.2051024436950684,
+      "learning_rate": 6.185935637663886e-06,
+      "loss": 0.0408,
+      "step": 1166
+    },
+    {
+      "epoch": 0.6954707985697258,
+      "grad_norm": 2.494473695755005,
+      "learning_rate": 6.1740166865315854e-06,
+      "loss": 0.0457,
+      "step": 1167
+    },
+    {
+      "epoch": 0.6960667461263409,
+      "grad_norm": 7.206812858581543,
+      "learning_rate": 6.162097735399285e-06,
+      "loss": 0.5905,
+      "step": 1168
+    },
+    {
+      "epoch": 0.6966626936829559,
+      "grad_norm": 2.628601312637329,
+      "learning_rate": 6.150178784266985e-06,
+      "loss": 0.0289,
+      "step": 1169
+    },
+    {
+      "epoch": 0.6972586412395709,
+      "grad_norm": 4.6316142082214355,
+      "learning_rate": 6.138259833134685e-06,
+      "loss": 0.2242,
+      "step": 1170
+    },
+    {
+      "epoch": 0.6978545887961859,
+      "grad_norm": 2.3409907817840576,
+      "learning_rate": 6.126340882002384e-06,
+      "loss": 0.1002,
+      "step": 1171
+    },
+    {
+      "epoch": 0.698450536352801,
+      "grad_norm": 2.0637707710266113,
+      "learning_rate": 6.114421930870083e-06,
+      "loss": 0.0371,
+      "step": 1172
+    },
+    {
+      "epoch": 0.6990464839094159,
+      "grad_norm": 1.5731134414672852,
+      "learning_rate": 6.102502979737783e-06,
+      "loss": 0.0191,
+      "step": 1173
+    },
+    {
+      "epoch": 0.699642431466031,
+      "grad_norm": 4.803267478942871,
+      "learning_rate": 6.090584028605483e-06,
+      "loss": 0.0915,
+      "step": 1174
+    },
+    {
+      "epoch": 0.700238379022646,
+      "grad_norm": 2.069693088531494,
+      "learning_rate": 6.0786650774731824e-06,
+      "loss": 0.0518,
+      "step": 1175
+    },
+    {
+      "epoch": 0.700834326579261,
+      "grad_norm": 3.0883407592773438,
+      "learning_rate": 6.066746126340882e-06,
+      "loss": 0.1514,
+      "step": 1176
+    },
+    {
+      "epoch": 0.701430274135876,
+      "grad_norm": 4.128971099853516,
+      "learning_rate": 6.054827175208582e-06,
+      "loss": 0.1026,
+      "step": 1177
+    },
+    {
+      "epoch": 0.7020262216924911,
+      "grad_norm": 2.3611960411071777,
+      "learning_rate": 6.042908224076282e-06,
+      "loss": 0.0397,
+      "step": 1178
+    },
+    {
+      "epoch": 0.702622169249106,
+      "grad_norm": 4.395667552947998,
+      "learning_rate": 6.030989272943981e-06,
+      "loss": 0.1427,
+      "step": 1179
+    },
+    {
+      "epoch": 0.7032181168057211,
+      "grad_norm": 2.4245283603668213,
+      "learning_rate": 6.019070321811681e-06,
+      "loss": 0.0293,
+      "step": 1180
+    },
+    {
+      "epoch": 0.7038140643623361,
+      "grad_norm": 2.081998109817505,
+      "learning_rate": 6.007151370679381e-06,
+      "loss": 0.0487,
+      "step": 1181
+    },
+    {
+      "epoch": 0.7044100119189511,
+      "grad_norm": 2.83132266998291,
+      "learning_rate": 5.9952324195470806e-06,
+      "loss": 0.0612,
+      "step": 1182
+    },
+    {
+      "epoch": 0.7050059594755661,
+      "grad_norm": 1.9806642532348633,
+      "learning_rate": 5.9833134684147794e-06,
+      "loss": 0.0776,
+      "step": 1183
+    },
+    {
+      "epoch": 0.7056019070321812,
+      "grad_norm": 1.929106593132019,
+      "learning_rate": 5.971394517282479e-06,
+      "loss": 0.016,
+      "step": 1184
+    },
+    {
+      "epoch": 0.7061978545887961,
+      "grad_norm": 1.2401010990142822,
+      "learning_rate": 5.959475566150179e-06,
+      "loss": 0.0193,
+      "step": 1185
+    },
+    {
+      "epoch": 0.7067938021454112,
+      "grad_norm": 2.925126552581787,
+      "learning_rate": 5.947556615017879e-06,
+      "loss": 0.0712,
+      "step": 1186
+    },
+    {
+      "epoch": 0.7073897497020262,
+      "grad_norm": 4.228752136230469,
+      "learning_rate": 5.935637663885578e-06,
+      "loss": 0.1487,
+      "step": 1187
+    },
+    {
+      "epoch": 0.7079856972586412,
+      "grad_norm": 2.2906155586242676,
+      "learning_rate": 5.923718712753278e-06,
+      "loss": 0.0201,
+      "step": 1188
+    },
+    {
+      "epoch": 0.7085816448152562,
+      "grad_norm": 2.392747640609741,
+      "learning_rate": 5.911799761620978e-06,
+      "loss": 0.0596,
+      "step": 1189
+    },
+    {
+      "epoch": 0.7091775923718713,
+      "grad_norm": 1.8827886581420898,
+      "learning_rate": 5.8998808104886775e-06,
+      "loss": 0.028,
+      "step": 1190
+    },
+    {
+      "epoch": 0.7097735399284862,
+      "grad_norm": 5.079347610473633,
+      "learning_rate": 5.887961859356377e-06,
+      "loss": 0.2279,
+      "step": 1191
+    },
+    {
+      "epoch": 0.7103694874851013,
+      "grad_norm": 2.2446863651275635,
+      "learning_rate": 5.876042908224077e-06,
+      "loss": 0.0344,
+      "step": 1192
+    },
+    {
+      "epoch": 0.7109654350417163,
+      "grad_norm": 1.5279927253723145,
+      "learning_rate": 5.864123957091776e-06,
+      "loss": 0.0266,
+      "step": 1193
+    },
+    {
+      "epoch": 0.7115613825983313,
+      "grad_norm": 2.473679542541504,
+      "learning_rate": 5.852205005959476e-06,
+      "loss": 0.0481,
+      "step": 1194
+    },
+    {
+      "epoch": 0.7121573301549463,
+      "grad_norm": 2.3891243934631348,
+      "learning_rate": 5.840286054827175e-06,
+      "loss": 0.0673,
+      "step": 1195
+    },
+    {
+      "epoch": 0.7127532777115614,
+      "grad_norm": 2.470102310180664,
+      "learning_rate": 5.828367103694875e-06,
+      "loss": 0.0377,
+      "step": 1196
+    },
+    {
+      "epoch": 0.7133492252681763,
+      "grad_norm": 2.922252655029297,
+      "learning_rate": 5.816448152562575e-06,
+      "loss": 0.0431,
+      "step": 1197
+    },
+    {
+      "epoch": 0.7139451728247914,
+      "grad_norm": 2.768221616744995,
+      "learning_rate": 5.8045292014302745e-06,
+      "loss": 0.0528,
+      "step": 1198
+    },
+    {
+      "epoch": 0.7145411203814065,
+      "grad_norm": 3.103853940963745,
+      "learning_rate": 5.792610250297974e-06,
+      "loss": 0.1055,
+      "step": 1199
+    },
+    {
+      "epoch": 0.7151370679380215,
+      "grad_norm": 2.8077375888824463,
+      "learning_rate": 5.780691299165674e-06,
+      "loss": 0.0507,
+      "step": 1200
+    },
+    {
+      "epoch": 0.7157330154946365,
+      "grad_norm": 2.822129487991333,
+      "learning_rate": 5.768772348033374e-06,
+      "loss": 0.0967,
+      "step": 1201
+    },
+    {
+      "epoch": 0.7163289630512515,
+      "grad_norm": 7.454535961151123,
+      "learning_rate": 5.7568533969010735e-06,
+      "loss": 0.5716,
+      "step": 1202
+    },
+    {
+      "epoch": 0.7169249106078666,
+      "grad_norm": 4.591785907745361,
+      "learning_rate": 5.744934445768773e-06,
+      "loss": 0.2139,
+      "step": 1203
+    },
+    {
+      "epoch": 0.7175208581644815,
+      "grad_norm": 2.651623010635376,
+      "learning_rate": 5.733015494636472e-06,
+      "loss": 0.0262,
+      "step": 1204
+    },
+    {
+      "epoch": 0.7181168057210966,
+      "grad_norm": 1.9551035165786743,
+      "learning_rate": 5.721096543504172e-06,
+      "loss": 0.0343,
+      "step": 1205
+    },
+    {
+      "epoch": 0.7187127532777116,
+      "grad_norm": 3.3551743030548096,
+      "learning_rate": 5.7091775923718715e-06,
+      "loss": 0.0969,
+      "step": 1206
+    },
+    {
+      "epoch": 0.7193087008343266,
+      "grad_norm": 3.6219136714935303,
+      "learning_rate": 5.697258641239571e-06,
+      "loss": 0.0986,
+      "step": 1207
+    },
+    {
+      "epoch": 0.7199046483909416,
+      "grad_norm": 3.2070531845092773,
+      "learning_rate": 5.685339690107271e-06,
+      "loss": 0.087,
+      "step": 1208
+    },
+    {
+      "epoch": 0.7205005959475567,
+      "grad_norm": 2.1673078536987305,
+      "learning_rate": 5.673420738974971e-06,
+      "loss": 0.0316,
+      "step": 1209
+    },
+    {
+      "epoch": 0.7210965435041716,
+      "grad_norm": 2.0260696411132812,
+      "learning_rate": 5.6615017878426705e-06,
+      "loss": 0.0255,
+      "step": 1210
+    },
+    {
+      "epoch": 0.7216924910607867,
+      "grad_norm": 2.75852632522583,
+      "learning_rate": 5.64958283671037e-06,
+      "loss": 0.0348,
+      "step": 1211
+    },
+    {
+      "epoch": 0.7222884386174017,
+      "grad_norm": 2.978618621826172,
+      "learning_rate": 5.63766388557807e-06,
+      "loss": 0.0666,
+      "step": 1212
+    },
+    {
+      "epoch": 0.7228843861740167,
+      "grad_norm": 5.50506591796875,
+      "learning_rate": 5.62574493444577e-06,
+      "loss": 0.2366,
+      "step": 1213
+    },
+    {
+      "epoch": 0.7234803337306317,
+      "grad_norm": 5.543450832366943,
+      "learning_rate": 5.6138259833134685e-06,
+      "loss": 0.1997,
+      "step": 1214
+    },
+    {
+      "epoch": 0.7240762812872468,
+      "grad_norm": 3.9785401821136475,
+      "learning_rate": 5.601907032181168e-06,
+      "loss": 0.1779,
+      "step": 1215
+    },
+    {
+      "epoch": 0.7246722288438617,
+      "grad_norm": 3.2885096073150635,
+      "learning_rate": 5.589988081048868e-06,
+      "loss": 0.093,
+      "step": 1216
+    },
+    {
+      "epoch": 0.7252681764004768,
+      "grad_norm": 2.642819881439209,
+      "learning_rate": 5.578069129916568e-06,
+      "loss": 0.0635,
+      "step": 1217
+    },
+    {
+      "epoch": 0.7258641239570918,
+      "grad_norm": 3.8590922355651855,
+      "learning_rate": 5.5661501787842675e-06,
+      "loss": 0.1173,
+      "step": 1218
+    },
+    {
+      "epoch": 0.7264600715137068,
+      "grad_norm": 2.8430287837982178,
+      "learning_rate": 5.554231227651967e-06,
+      "loss": 0.0603,
+      "step": 1219
+    },
+    {
+      "epoch": 0.7270560190703218,
+      "grad_norm": 3.5365285873413086,
+      "learning_rate": 5.542312276519667e-06,
+      "loss": 0.1419,
+      "step": 1220
+    },
+    {
+      "epoch": 0.7276519666269369,
+      "grad_norm": 1.5074772834777832,
+      "learning_rate": 5.530393325387367e-06,
+      "loss": 0.0279,
+      "step": 1221
+    },
+    {
+      "epoch": 0.7282479141835518,
+      "grad_norm": 2.2593233585357666,
+      "learning_rate": 5.518474374255066e-06,
+      "loss": 0.053,
+      "step": 1222
+    },
+    {
+      "epoch": 0.7288438617401669,
+      "grad_norm": 2.690866231918335,
+      "learning_rate": 5.506555423122766e-06,
+      "loss": 0.0291,
+      "step": 1223
+    },
+    {
+      "epoch": 0.7294398092967819,
+      "grad_norm": 2.543686866760254,
+      "learning_rate": 5.494636471990466e-06,
+      "loss": 0.0818,
+      "step": 1224
+    },
+    {
+      "epoch": 0.7300357568533969,
+      "grad_norm": 3.4255502223968506,
+      "learning_rate": 5.482717520858165e-06,
+      "loss": 0.1687,
+      "step": 1225
+    },
+    {
+      "epoch": 0.7306317044100119,
+      "grad_norm": 2.494204044342041,
+      "learning_rate": 5.4707985697258645e-06,
+      "loss": 0.0397,
+      "step": 1226
+    },
+    {
+      "epoch": 0.731227651966627,
+      "grad_norm": 4.699764728546143,
+      "learning_rate": 5.458879618593564e-06,
+      "loss": 0.1838,
+      "step": 1227
+    },
+    {
+      "epoch": 0.7318235995232419,
+      "grad_norm": 2.8754186630249023,
+      "learning_rate": 5.446960667461264e-06,
+      "loss": 0.0497,
+      "step": 1228
+    },
+    {
+      "epoch": 0.732419547079857,
+      "grad_norm": 1.708220362663269,
+      "learning_rate": 5.435041716328964e-06,
+      "loss": 0.0308,
+      "step": 1229
+    },
+    {
+      "epoch": 0.733015494636472,
+      "grad_norm": 2.385408639907837,
+      "learning_rate": 5.423122765196663e-06,
+      "loss": 0.0515,
+      "step": 1230
+    },
+    {
+      "epoch": 0.733611442193087,
+      "grad_norm": 2.4741907119750977,
+      "learning_rate": 5.411203814064363e-06,
+      "loss": 0.043,
+      "step": 1231
+    },
+    {
+      "epoch": 0.734207389749702,
+      "grad_norm": 2.2943315505981445,
+      "learning_rate": 5.399284862932063e-06,
+      "loss": 0.0577,
+      "step": 1232
+    },
+    {
+      "epoch": 0.7348033373063171,
+      "grad_norm": 1.793779969215393,
+      "learning_rate": 5.3873659117997626e-06,
+      "loss": 0.0242,
+      "step": 1233
+    },
+    {
+      "epoch": 0.735399284862932,
+      "grad_norm": 2.8842623233795166,
+      "learning_rate": 5.375446960667462e-06,
+      "loss": 0.0805,
+      "step": 1234
+    },
+    {
+      "epoch": 0.7359952324195471,
+      "grad_norm": 2.0644900798797607,
+      "learning_rate": 5.363528009535162e-06,
+      "loss": 0.0369,
+      "step": 1235
+    },
+    {
+      "epoch": 0.7365911799761621,
+      "grad_norm": 1.883152961730957,
+      "learning_rate": 5.351609058402861e-06,
+      "loss": 0.0217,
+      "step": 1236
+    },
+    {
+      "epoch": 0.7371871275327772,
+      "grad_norm": 2.6587488651275635,
+      "learning_rate": 5.339690107270561e-06,
+      "loss": 0.0724,
+      "step": 1237
+    },
+    {
+      "epoch": 0.7377830750893921,
+      "grad_norm": 3.9151549339294434,
+      "learning_rate": 5.32777115613826e-06,
+      "loss": 0.1188,
+      "step": 1238
+    },
+    {
+      "epoch": 0.7383790226460072,
+      "grad_norm": 7.870319366455078,
+      "learning_rate": 5.31585220500596e-06,
+      "loss": 0.8665,
+      "step": 1239
+    },
+    {
+      "epoch": 0.7389749702026222,
+      "grad_norm": 0.6723051071166992,
+      "learning_rate": 5.30393325387366e-06,
+      "loss": 0.0135,
+      "step": 1240
+    },
+    {
+      "epoch": 0.7395709177592372,
+      "grad_norm": 6.917332649230957,
+      "learning_rate": 5.2920143027413596e-06,
+      "loss": 0.4533,
+      "step": 1241
+    },
+    {
+      "epoch": 0.7401668653158522,
+      "grad_norm": 0.8312991261482239,
+      "learning_rate": 5.280095351609059e-06,
+      "loss": 0.0123,
+      "step": 1242
+    },
+    {
+      "epoch": 0.7407628128724673,
+      "grad_norm": 2.468366861343384,
+      "learning_rate": 5.268176400476759e-06,
+      "loss": 0.0228,
+      "step": 1243
+    },
+    {
+      "epoch": 0.7413587604290822,
+      "grad_norm": 1.7204393148422241,
+      "learning_rate": 5.256257449344459e-06,
+      "loss": 0.0214,
+      "step": 1244
+    },
+    {
+      "epoch": 0.7419547079856973,
+      "grad_norm": 3.438488721847534,
+      "learning_rate": 5.2443384982121585e-06,
+      "loss": 0.0758,
+      "step": 1245
+    },
+    {
+      "epoch": 0.7425506555423123,
+      "grad_norm": 3.128673791885376,
+      "learning_rate": 5.232419547079857e-06,
+      "loss": 0.2049,
+      "step": 1246
+    },
+    {
+      "epoch": 0.7431466030989273,
+      "grad_norm": 4.451478481292725,
+      "learning_rate": 5.220500595947557e-06,
+      "loss": 0.0592,
+      "step": 1247
+    },
+    {
+      "epoch": 0.7437425506555423,
+      "grad_norm": 3.058401107788086,
+      "learning_rate": 5.208581644815257e-06,
+      "loss": 0.1025,
+      "step": 1248
+    },
+    {
+      "epoch": 0.7443384982121574,
+      "grad_norm": 3.6950948238372803,
+      "learning_rate": 5.1966626936829566e-06,
+      "loss": 0.1312,
+      "step": 1249
+    },
+    {
+      "epoch": 0.7449344457687723,
+      "grad_norm": 3.580315589904785,
+      "learning_rate": 5.184743742550656e-06,
+      "loss": 0.1679,
+      "step": 1250
+    },
+    {
+      "epoch": 0.7455303933253874,
+      "grad_norm": 1.4871916770935059,
+      "learning_rate": 5.172824791418356e-06,
+      "loss": 0.0199,
+      "step": 1251
+    },
+    {
+      "epoch": 0.7461263408820024,
+      "grad_norm": 3.6483335494995117,
+      "learning_rate": 5.160905840286056e-06,
+      "loss": 0.126,
+      "step": 1252
+    },
+    {
+      "epoch": 0.7467222884386174,
+      "grad_norm": 2.335641622543335,
+      "learning_rate": 5.1489868891537555e-06,
+      "loss": 0.0392,
+      "step": 1253
+    },
+    {
+      "epoch": 0.7473182359952324,
+      "grad_norm": 3.3537986278533936,
+      "learning_rate": 5.137067938021455e-06,
+      "loss": 0.1027,
+      "step": 1254
+    },
+    {
+      "epoch": 0.7479141835518475,
+      "grad_norm": 1.8556506633758545,
+      "learning_rate": 5.125148986889155e-06,
+      "loss": 0.0126,
+      "step": 1255
+    },
+    {
+      "epoch": 0.7485101311084624,
+      "grad_norm": 5.477853298187256,
+      "learning_rate": 5.113230035756855e-06,
+      "loss": 0.188,
+      "step": 1256
+    },
+    {
+      "epoch": 0.7491060786650775,
+      "grad_norm": 5.612376689910889,
+      "learning_rate": 5.1013110846245535e-06,
+      "loss": 0.2914,
+      "step": 1257
+    },
+    {
+      "epoch": 0.7497020262216925,
+      "grad_norm": 3.4262020587921143,
+      "learning_rate": 5.089392133492253e-06,
+      "loss": 0.0639,
+      "step": 1258
+    },
+    {
+      "epoch": 0.7502979737783075,
+      "grad_norm": 2.3721022605895996,
+      "learning_rate": 5.077473182359953e-06,
+      "loss": 0.0663,
+      "step": 1259
+    },
+    {
+      "epoch": 0.7508939213349225,
+      "grad_norm": 2.726559638977051,
+      "learning_rate": 5.065554231227653e-06,
+      "loss": 0.078,
+      "step": 1260
+    },
+    {
+      "epoch": 0.7514898688915376,
+      "grad_norm": 2.3778347969055176,
+      "learning_rate": 5.0536352800953525e-06,
+      "loss": 0.0426,
+      "step": 1261
+    },
+    {
+      "epoch": 0.7520858164481525,
+      "grad_norm": 6.475907802581787,
+      "learning_rate": 5.041716328963052e-06,
+      "loss": 0.4502,
+      "step": 1262
+    },
+    {
+      "epoch": 0.7526817640047676,
+      "grad_norm": 4.859015464782715,
+      "learning_rate": 5.029797377830752e-06,
+      "loss": 0.2577,
+      "step": 1263
+    },
+    {
+      "epoch": 0.7532777115613826,
+      "grad_norm": 4.157678127288818,
+      "learning_rate": 5.017878426698452e-06,
+      "loss": 0.2725,
+      "step": 1264
+    },
+    {
+      "epoch": 0.7538736591179976,
+      "grad_norm": 2.740621328353882,
+      "learning_rate": 5.005959475566151e-06,
+      "loss": 0.0379,
+      "step": 1265
+    },
+    {
+      "epoch": 0.7544696066746126,
+      "grad_norm": 3.7304000854492188,
+      "learning_rate": 4.99404052443385e-06,
+      "loss": 0.1241,
+      "step": 1266
+    },
+    {
+      "epoch": 0.7550655542312277,
+      "grad_norm": 3.14554762840271,
+      "learning_rate": 4.98212157330155e-06,
+      "loss": 0.0406,
+      "step": 1267
+    },
+    {
+      "epoch": 0.7556615017878426,
+      "grad_norm": 2.6353635787963867,
+      "learning_rate": 4.97020262216925e-06,
+      "loss": 0.0597,
+      "step": 1268
+    },
+    {
+      "epoch": 0.7562574493444577,
+      "grad_norm": 2.8757503032684326,
+      "learning_rate": 4.9582836710369495e-06,
+      "loss": 0.0869,
+      "step": 1269
+    },
+    {
+      "epoch": 0.7568533969010727,
+      "grad_norm": 3.174990653991699,
+      "learning_rate": 4.946364719904649e-06,
+      "loss": 0.063,
+      "step": 1270
+    },
+    {
+      "epoch": 0.7574493444576877,
+      "grad_norm": 6.198691368103027,
+      "learning_rate": 4.934445768772348e-06,
+      "loss": 0.2734,
+      "step": 1271
+    },
+    {
+      "epoch": 0.7580452920143027,
+      "grad_norm": 3.2936997413635254,
+      "learning_rate": 4.922526817640048e-06,
+      "loss": 0.1434,
+      "step": 1272
+    },
+    {
+      "epoch": 0.7586412395709178,
+      "grad_norm": 3.24391508102417,
+      "learning_rate": 4.9106078665077475e-06,
+      "loss": 0.0948,
+      "step": 1273
+    },
+    {
+      "epoch": 0.7592371871275327,
+      "grad_norm": 3.3039894104003906,
+      "learning_rate": 4.898688915375447e-06,
+      "loss": 0.1084,
+      "step": 1274
+    },
+    {
+      "epoch": 0.7598331346841478,
+      "grad_norm": 1.8673690557479858,
+      "learning_rate": 4.886769964243147e-06,
+      "loss": 0.0251,
+      "step": 1275
+    },
+    {
+      "epoch": 0.7604290822407628,
+      "grad_norm": 4.618743419647217,
+      "learning_rate": 4.874851013110847e-06,
+      "loss": 0.206,
+      "step": 1276
+    },
+    {
+      "epoch": 0.7610250297973778,
+      "grad_norm": 3.8007113933563232,
+      "learning_rate": 4.8629320619785465e-06,
+      "loss": 0.2136,
+      "step": 1277
+    },
+    {
+      "epoch": 0.7616209773539928,
+      "grad_norm": 1.2952100038528442,
+      "learning_rate": 4.851013110846246e-06,
+      "loss": 0.0209,
+      "step": 1278
+    },
+    {
+      "epoch": 0.7622169249106079,
+      "grad_norm": 2.23616623878479,
+      "learning_rate": 4.839094159713946e-06,
+      "loss": 0.0351,
+      "step": 1279
+    },
+    {
+      "epoch": 0.7628128724672228,
+      "grad_norm": 1.867742896080017,
+      "learning_rate": 4.827175208581646e-06,
+      "loss": 0.0265,
+      "step": 1280
+    },
+    {
+      "epoch": 0.7634088200238379,
+      "grad_norm": 2.373424768447876,
+      "learning_rate": 4.815256257449345e-06,
+      "loss": 0.0509,
+      "step": 1281
+    },
+    {
+      "epoch": 0.7640047675804529,
+      "grad_norm": 2.160205125808716,
+      "learning_rate": 4.803337306317044e-06,
+      "loss": 0.0324,
+      "step": 1282
+    },
+    {
+      "epoch": 0.764600715137068,
+      "grad_norm": 7.456882476806641,
+      "learning_rate": 4.791418355184744e-06,
+      "loss": 0.6713,
+      "step": 1283
+    },
+    {
+      "epoch": 0.7651966626936829,
+      "grad_norm": 3.692147970199585,
+      "learning_rate": 4.779499404052444e-06,
+      "loss": 0.1569,
+      "step": 1284
+    },
+    {
+      "epoch": 0.765792610250298,
+      "grad_norm": 1.6925747394561768,
+      "learning_rate": 4.7675804529201435e-06,
+      "loss": 0.0165,
+      "step": 1285
+    },
+    {
+      "epoch": 0.766388557806913,
+      "grad_norm": 1.4829703569412231,
+      "learning_rate": 4.755661501787843e-06,
+      "loss": 0.0289,
+      "step": 1286
+    },
+    {
+      "epoch": 0.766984505363528,
+      "grad_norm": 6.628187656402588,
+      "learning_rate": 4.743742550655543e-06,
+      "loss": 0.3501,
+      "step": 1287
+    },
+    {
+      "epoch": 0.767580452920143,
+      "grad_norm": 1.9022228717803955,
+      "learning_rate": 4.731823599523243e-06,
+      "loss": 0.0338,
+      "step": 1288
+    },
+    {
+      "epoch": 0.7681764004767581,
+      "grad_norm": 5.0756707191467285,
+      "learning_rate": 4.719904648390942e-06,
+      "loss": 0.1289,
+      "step": 1289
+    },
+    {
+      "epoch": 0.768772348033373,
+      "grad_norm": 3.395425319671631,
+      "learning_rate": 4.707985697258642e-06,
+      "loss": 0.0751,
+      "step": 1290
+    },
+    {
+      "epoch": 0.7693682955899881,
+      "grad_norm": 2.5421316623687744,
+      "learning_rate": 4.696066746126342e-06,
+      "loss": 0.0354,
+      "step": 1291
+    },
+    {
+      "epoch": 0.7699642431466031,
+      "grad_norm": 10.178092956542969,
+      "learning_rate": 4.6841477949940416e-06,
+      "loss": 0.8973,
+      "step": 1292
+    },
+    {
+      "epoch": 0.7705601907032181,
+      "grad_norm": 1.9432107210159302,
+      "learning_rate": 4.6722288438617405e-06,
+      "loss": 0.0278,
+      "step": 1293
+    },
+    {
+      "epoch": 0.7711561382598331,
+      "grad_norm": 2.1611554622650146,
+      "learning_rate": 4.66030989272944e-06,
+      "loss": 0.0371,
+      "step": 1294
+    },
+    {
+      "epoch": 0.7717520858164482,
+      "grad_norm": 2.5997865200042725,
+      "learning_rate": 4.64839094159714e-06,
+      "loss": 0.0673,
+      "step": 1295
+    },
+    {
+      "epoch": 0.7723480333730631,
+      "grad_norm": 2.871330499649048,
+      "learning_rate": 4.636471990464839e-06,
+      "loss": 0.0584,
+      "step": 1296
+    },
+    {
+      "epoch": 0.7729439809296782,
+      "grad_norm": 2.370143175125122,
+      "learning_rate": 4.6245530393325385e-06,
+      "loss": 0.0416,
+      "step": 1297
+    },
+    {
+      "epoch": 0.7735399284862932,
+      "grad_norm": 4.48471736907959,
+      "learning_rate": 4.612634088200238e-06,
+      "loss": 0.2054,
+      "step": 1298
+    },
+    {
+      "epoch": 0.7741358760429082,
+      "grad_norm": 1.8504489660263062,
+      "learning_rate": 4.600715137067938e-06,
+      "loss": 0.0466,
+      "step": 1299
+    },
+    {
+      "epoch": 0.7747318235995232,
+      "grad_norm": 5.12666130065918,
+      "learning_rate": 4.588796185935638e-06,
+      "loss": 0.3279,
+      "step": 1300
+    },
+    {
+      "epoch": 0.7753277711561383,
+      "grad_norm": 2.2526676654815674,
+      "learning_rate": 4.5768772348033374e-06,
+      "loss": 0.0327,
+      "step": 1301
+    },
+    {
+      "epoch": 0.7759237187127532,
+      "grad_norm": 2.2608699798583984,
+      "learning_rate": 4.564958283671037e-06,
+      "loss": 0.0418,
+      "step": 1302
+    },
+    {
+      "epoch": 0.7765196662693683,
+      "grad_norm": 2.5429494380950928,
+      "learning_rate": 4.553039332538737e-06,
+      "loss": 0.0326,
+      "step": 1303
+    },
+    {
+      "epoch": 0.7771156138259833,
+      "grad_norm": 2.0945937633514404,
+      "learning_rate": 4.541120381406437e-06,
+      "loss": 0.0329,
+      "step": 1304
+    },
+    {
+      "epoch": 0.7777115613825983,
+      "grad_norm": 2.5666866302490234,
+      "learning_rate": 4.529201430274136e-06,
+      "loss": 0.0411,
+      "step": 1305
+    },
+    {
+      "epoch": 0.7783075089392133,
+      "grad_norm": 2.3858096599578857,
+      "learning_rate": 4.517282479141836e-06,
+      "loss": 0.0492,
+      "step": 1306
+    },
+    {
+      "epoch": 0.7789034564958284,
+      "grad_norm": 2.043741464614868,
+      "learning_rate": 4.505363528009535e-06,
+      "loss": 0.0235,
+      "step": 1307
+    },
+    {
+      "epoch": 0.7794994040524433,
+      "grad_norm": 3.832977771759033,
+      "learning_rate": 4.493444576877235e-06,
+      "loss": 0.1472,
+      "step": 1308
+    },
+    {
+      "epoch": 0.7800953516090584,
+      "grad_norm": 2.4840798377990723,
+      "learning_rate": 4.4815256257449344e-06,
+      "loss": 0.0441,
+      "step": 1309
+    },
+    {
+      "epoch": 0.7806912991656734,
+      "grad_norm": 3.7343192100524902,
+      "learning_rate": 4.469606674612634e-06,
+      "loss": 0.092,
+      "step": 1310
+    },
+    {
+      "epoch": 0.7812872467222884,
+      "grad_norm": 3.164261817932129,
+      "learning_rate": 4.457687723480334e-06,
+      "loss": 0.1036,
+      "step": 1311
+    },
+    {
+      "epoch": 0.7818831942789034,
+      "grad_norm": 4.32907772064209,
+      "learning_rate": 4.445768772348034e-06,
+      "loss": 0.1157,
+      "step": 1312
+    },
+    {
+      "epoch": 0.7824791418355185,
+      "grad_norm": 2.589921474456787,
+      "learning_rate": 4.433849821215733e-06,
+      "loss": 0.0535,
+      "step": 1313
+    },
+    {
+      "epoch": 0.7830750893921334,
+      "grad_norm": 5.096065044403076,
+      "learning_rate": 4.421930870083433e-06,
+      "loss": 0.2161,
+      "step": 1314
+    },
+    {
+      "epoch": 0.7836710369487485,
+      "grad_norm": 2.6442270278930664,
+      "learning_rate": 4.410011918951133e-06,
+      "loss": 0.0819,
+      "step": 1315
+    },
+    {
+      "epoch": 0.7842669845053635,
+      "grad_norm": 2.651536226272583,
+      "learning_rate": 4.3980929678188326e-06,
+      "loss": 0.0856,
+      "step": 1316
+    },
+    {
+      "epoch": 0.7848629320619785,
+      "grad_norm": 2.5488345623016357,
+      "learning_rate": 4.386174016686532e-06,
+      "loss": 0.0525,
+      "step": 1317
+    },
+    {
+      "epoch": 0.7854588796185935,
+      "grad_norm": 3.5000293254852295,
+      "learning_rate": 4.374255065554231e-06,
+      "loss": 0.0376,
+      "step": 1318
+    },
+    {
+      "epoch": 0.7860548271752086,
+      "grad_norm": 4.375251770019531,
+      "learning_rate": 4.362336114421931e-06,
+      "loss": 0.1395,
+      "step": 1319
+    },
+    {
+      "epoch": 0.7866507747318237,
+      "grad_norm": 2.8396964073181152,
+      "learning_rate": 4.350417163289631e-06,
+      "loss": 0.0766,
+      "step": 1320
+    },
+    {
+      "epoch": 0.7872467222884386,
+      "grad_norm": 3.052922487258911,
+      "learning_rate": 4.33849821215733e-06,
+      "loss": 0.0838,
+      "step": 1321
+    },
+    {
+      "epoch": 0.7878426698450537,
+      "grad_norm": 2.043560743331909,
+      "learning_rate": 4.32657926102503e-06,
+      "loss": 0.0301,
+      "step": 1322
+    },
+    {
+      "epoch": 0.7884386174016687,
+      "grad_norm": 1.574734091758728,
+      "learning_rate": 4.31466030989273e-06,
+      "loss": 0.0205,
+      "step": 1323
+    },
+    {
+      "epoch": 0.7890345649582837,
+      "grad_norm": 1.8588253259658813,
+      "learning_rate": 4.3027413587604295e-06,
+      "loss": 0.0361,
+      "step": 1324
+    },
+    {
+      "epoch": 0.7896305125148987,
+      "grad_norm": 2.7645130157470703,
+      "learning_rate": 4.290822407628129e-06,
+      "loss": 0.0473,
+      "step": 1325
+    },
+    {
+      "epoch": 0.7902264600715138,
+      "grad_norm": 2.636229991912842,
+      "learning_rate": 4.278903456495829e-06,
+      "loss": 0.1063,
+      "step": 1326
+    },
+    {
+      "epoch": 0.7908224076281287,
+      "grad_norm": 2.616288423538208,
+      "learning_rate": 4.266984505363529e-06,
+      "loss": 0.0395,
+      "step": 1327
+    },
+    {
+      "epoch": 0.7914183551847438,
+      "grad_norm": 3.121781826019287,
+      "learning_rate": 4.255065554231228e-06,
+      "loss": 0.0597,
+      "step": 1328
+    },
+    {
+      "epoch": 0.7920143027413588,
+      "grad_norm": 4.313778400421143,
+      "learning_rate": 4.243146603098927e-06,
+      "loss": 0.1665,
+      "step": 1329
+    },
+    {
+      "epoch": 0.7926102502979738,
+      "grad_norm": 5.76975154876709,
+      "learning_rate": 4.231227651966627e-06,
+      "loss": 0.185,
+      "step": 1330
+    },
+    {
+      "epoch": 0.7932061978545888,
+      "grad_norm": 2.8637681007385254,
+      "learning_rate": 4.219308700834327e-06,
+      "loss": 0.0525,
+      "step": 1331
+    },
+    {
+      "epoch": 0.7938021454112039,
+      "grad_norm": 2.148899793624878,
+      "learning_rate": 4.2073897497020265e-06,
+      "loss": 0.0526,
+      "step": 1332
+    },
+    {
+      "epoch": 0.7943980929678188,
+      "grad_norm": 3.347210168838501,
+      "learning_rate": 4.195470798569726e-06,
+      "loss": 0.0717,
+      "step": 1333
+    },
+    {
+      "epoch": 0.7949940405244339,
+      "grad_norm": 2.132847547531128,
+      "learning_rate": 4.183551847437426e-06,
+      "loss": 0.0597,
+      "step": 1334
+    },
+    {
+      "epoch": 0.7955899880810489,
+      "grad_norm": 8.103880882263184,
+      "learning_rate": 4.171632896305126e-06,
+      "loss": 0.6,
+      "step": 1335
+    },
+    {
+      "epoch": 0.7961859356376639,
+      "grad_norm": 5.309153079986572,
+      "learning_rate": 4.1597139451728255e-06,
+      "loss": 0.2293,
+      "step": 1336
+    },
+    {
+      "epoch": 0.7967818831942789,
+      "grad_norm": 3.466413974761963,
+      "learning_rate": 4.147794994040525e-06,
+      "loss": 0.078,
+      "step": 1337
+    },
+    {
+      "epoch": 0.797377830750894,
+      "grad_norm": 2.7010772228240967,
+      "learning_rate": 4.135876042908225e-06,
+      "loss": 0.1543,
+      "step": 1338
+    },
+    {
+      "epoch": 0.7979737783075089,
+      "grad_norm": 3.2330362796783447,
+      "learning_rate": 4.123957091775924e-06,
+      "loss": 0.1463,
+      "step": 1339
+    },
+    {
+      "epoch": 0.798569725864124,
+      "grad_norm": 1.607108235359192,
+      "learning_rate": 4.1120381406436235e-06,
+      "loss": 0.0316,
+      "step": 1340
+    },
+    {
+      "epoch": 0.799165673420739,
+      "grad_norm": 3.044491767883301,
+      "learning_rate": 4.100119189511323e-06,
+      "loss": 0.0658,
+      "step": 1341
+    },
+    {
+      "epoch": 0.799761620977354,
+      "grad_norm": 1.8893952369689941,
+      "learning_rate": 4.088200238379023e-06,
+      "loss": 0.0234,
+      "step": 1342
+    },
+    {
+      "epoch": 0.800357568533969,
+      "grad_norm": 2.9888243675231934,
+      "learning_rate": 4.076281287246723e-06,
+      "loss": 0.1029,
+      "step": 1343
+    },
+    {
+      "epoch": 0.8009535160905841,
+      "grad_norm": 3.1113743782043457,
+      "learning_rate": 4.0643623361144225e-06,
+      "loss": 0.0853,
+      "step": 1344
+    },
+    {
+      "epoch": 0.801549463647199,
+      "grad_norm": 4.227092742919922,
+      "learning_rate": 4.052443384982122e-06,
+      "loss": 0.1202,
+      "step": 1345
+    },
+    {
+      "epoch": 0.8021454112038141,
+      "grad_norm": 2.237760066986084,
+      "learning_rate": 4.040524433849822e-06,
+      "loss": 0.0631,
+      "step": 1346
+    },
+    {
+      "epoch": 0.8027413587604291,
+      "grad_norm": 2.2760984897613525,
+      "learning_rate": 4.028605482717522e-06,
+      "loss": 0.0452,
+      "step": 1347
+    },
+    {
+      "epoch": 0.8033373063170441,
+      "grad_norm": 3.9873101711273193,
+      "learning_rate": 4.016686531585221e-06,
+      "loss": 0.2078,
+      "step": 1348
+    },
+    {
+      "epoch": 0.8039332538736591,
+      "grad_norm": 2.6308178901672363,
+      "learning_rate": 4.00476758045292e-06,
+      "loss": 0.0644,
+      "step": 1349
+    },
+    {
+      "epoch": 0.8045292014302742,
+      "grad_norm": 4.436225891113281,
+      "learning_rate": 3.99284862932062e-06,
+      "loss": 0.1182,
+      "step": 1350
+    },
+    {
+      "epoch": 0.8051251489868891,
+      "grad_norm": 5.318544387817383,
+      "learning_rate": 3.98092967818832e-06,
+      "loss": 0.1562,
+      "step": 1351
+    },
+    {
+      "epoch": 0.8057210965435042,
+      "grad_norm": 3.7578964233398438,
+      "learning_rate": 3.9690107270560195e-06,
+      "loss": 0.1002,
+      "step": 1352
+    },
+    {
+      "epoch": 0.8063170441001192,
+      "grad_norm": 2.1111109256744385,
+      "learning_rate": 3.957091775923719e-06,
+      "loss": 0.0325,
+      "step": 1353
+    },
+    {
+      "epoch": 0.8069129916567342,
+      "grad_norm": 4.955950736999512,
+      "learning_rate": 3.945172824791419e-06,
+      "loss": 0.2022,
+      "step": 1354
+    },
+    {
+      "epoch": 0.8075089392133492,
+      "grad_norm": 2.041447162628174,
+      "learning_rate": 3.933253873659119e-06,
+      "loss": 0.0268,
+      "step": 1355
+    },
+    {
+      "epoch": 0.8081048867699643,
+      "grad_norm": 3.834017038345337,
+      "learning_rate": 3.9213349225268175e-06,
+      "loss": 0.1884,
+      "step": 1356
+    },
+    {
+      "epoch": 0.8087008343265792,
+      "grad_norm": 3.3759191036224365,
+      "learning_rate": 3.909415971394517e-06,
+      "loss": 0.0914,
+      "step": 1357
+    },
+    {
+      "epoch": 0.8092967818831943,
+      "grad_norm": 1.9740601778030396,
+      "learning_rate": 3.897497020262217e-06,
+      "loss": 0.0265,
+      "step": 1358
+    },
+    {
+      "epoch": 0.8098927294398093,
+      "grad_norm": 2.5155625343322754,
+      "learning_rate": 3.885578069129917e-06,
+      "loss": 0.0377,
+      "step": 1359
+    },
+    {
+      "epoch": 0.8104886769964244,
+      "grad_norm": 4.44327449798584,
+      "learning_rate": 3.8736591179976165e-06,
+      "loss": 0.1306,
+      "step": 1360
+    },
+    {
+      "epoch": 0.8110846245530393,
+      "grad_norm": 1.1348023414611816,
+      "learning_rate": 3.861740166865316e-06,
+      "loss": 0.0131,
+      "step": 1361
+    },
+    {
+      "epoch": 0.8116805721096544,
+      "grad_norm": 2.43721342086792,
+      "learning_rate": 3.849821215733016e-06,
+      "loss": 0.0387,
+      "step": 1362
+    },
+    {
+      "epoch": 0.8122765196662693,
+      "grad_norm": 4.001234531402588,
+      "learning_rate": 3.837902264600716e-06,
+      "loss": 0.1544,
+      "step": 1363
+    },
+    {
+      "epoch": 0.8128724672228844,
+      "grad_norm": 2.763822555541992,
+      "learning_rate": 3.8259833134684145e-06,
+      "loss": 0.0666,
+      "step": 1364
+    },
+    {
+      "epoch": 0.8134684147794994,
+      "grad_norm": 3.676327705383301,
+      "learning_rate": 3.8140643623361147e-06,
+      "loss": 0.0999,
+      "step": 1365
+    },
+    {
+      "epoch": 0.8140643623361145,
+      "grad_norm": 5.467711448669434,
+      "learning_rate": 3.802145411203814e-06,
+      "loss": 0.3507,
+      "step": 1366
+    },
+    {
+      "epoch": 0.8146603098927294,
+      "grad_norm": 3.18977427482605,
+      "learning_rate": 3.7902264600715137e-06,
+      "loss": 0.1071,
+      "step": 1367
+    },
+    {
+      "epoch": 0.8152562574493445,
+      "grad_norm": 5.722716331481934,
+      "learning_rate": 3.7783075089392134e-06,
+      "loss": 0.1534,
+      "step": 1368
+    },
+    {
+      "epoch": 0.8158522050059595,
+      "grad_norm": 2.572305679321289,
+      "learning_rate": 3.766388557806913e-06,
+      "loss": 0.0489,
+      "step": 1369
+    },
+    {
+      "epoch": 0.8164481525625745,
+      "grad_norm": 2.576174736022949,
+      "learning_rate": 3.754469606674613e-06,
+      "loss": 0.0335,
+      "step": 1370
+    },
+    {
+      "epoch": 0.8170441001191895,
+      "grad_norm": 4.052601337432861,
+      "learning_rate": 3.7425506555423126e-06,
+      "loss": 0.1586,
+      "step": 1371
+    },
+    {
+      "epoch": 0.8176400476758046,
+      "grad_norm": 8.502437591552734,
+      "learning_rate": 3.730631704410012e-06,
+      "loss": 0.7609,
+      "step": 1372
+    },
+    {
+      "epoch": 0.8182359952324195,
+      "grad_norm": 6.36151123046875,
+      "learning_rate": 3.7187127532777117e-06,
+      "loss": 0.541,
+      "step": 1373
+    },
+    {
+      "epoch": 0.8188319427890346,
+      "grad_norm": 2.5999956130981445,
+      "learning_rate": 3.7067938021454114e-06,
+      "loss": 0.0563,
+      "step": 1374
+    },
+    {
+      "epoch": 0.8194278903456496,
+      "grad_norm": 3.626056671142578,
+      "learning_rate": 3.694874851013111e-06,
+      "loss": 0.1097,
+      "step": 1375
+    },
+    {
+      "epoch": 0.8200238379022646,
+      "grad_norm": 2.969913959503174,
+      "learning_rate": 3.682955899880811e-06,
+      "loss": 0.0437,
+      "step": 1376
+    },
+    {
+      "epoch": 0.8206197854588796,
+      "grad_norm": 2.4127302169799805,
+      "learning_rate": 3.67103694874851e-06,
+      "loss": 0.0487,
+      "step": 1377
+    },
+    {
+      "epoch": 0.8212157330154947,
+      "grad_norm": 2.7202298641204834,
+      "learning_rate": 3.65911799761621e-06,
+      "loss": 0.0877,
+      "step": 1378
+    },
+    {
+      "epoch": 0.8218116805721096,
+      "grad_norm": 3.23714017868042,
+      "learning_rate": 3.6471990464839096e-06,
+      "loss": 0.1007,
+      "step": 1379
+    },
+    {
+      "epoch": 0.8224076281287247,
+      "grad_norm": 1.4859226942062378,
+      "learning_rate": 3.6352800953516094e-06,
+      "loss": 0.0262,
+      "step": 1380
+    },
+    {
+      "epoch": 0.8230035756853397,
+      "grad_norm": 2.2871851921081543,
+      "learning_rate": 3.623361144219309e-06,
+      "loss": 0.0499,
+      "step": 1381
+    },
+    {
+      "epoch": 0.8235995232419547,
+      "grad_norm": 3.2374980449676514,
+      "learning_rate": 3.6114421930870084e-06,
+      "loss": 0.0861,
+      "step": 1382
+    },
+    {
+      "epoch": 0.8241954707985697,
+      "grad_norm": 3.2266061305999756,
+      "learning_rate": 3.599523241954708e-06,
+      "loss": 0.1245,
+      "step": 1383
+    },
+    {
+      "epoch": 0.8247914183551848,
+      "grad_norm": 2.2869575023651123,
+      "learning_rate": 3.587604290822408e-06,
+      "loss": 0.0502,
+      "step": 1384
+    },
+    {
+      "epoch": 0.8253873659117997,
+      "grad_norm": 3.517930269241333,
+      "learning_rate": 3.5756853396901076e-06,
+      "loss": 0.0788,
+      "step": 1385
+    },
+    {
+      "epoch": 0.8259833134684148,
+      "grad_norm": 2.613194704055786,
+      "learning_rate": 3.5637663885578073e-06,
+      "loss": 0.0891,
+      "step": 1386
+    },
+    {
+      "epoch": 0.8265792610250298,
+      "grad_norm": 4.878419876098633,
+      "learning_rate": 3.551847437425507e-06,
+      "loss": 0.1571,
+      "step": 1387
+    },
+    {
+      "epoch": 0.8271752085816448,
+      "grad_norm": 3.7913482189178467,
+      "learning_rate": 3.5399284862932064e-06,
+      "loss": 0.1042,
+      "step": 1388
+    },
+    {
+      "epoch": 0.8277711561382598,
+      "grad_norm": 2.6399805545806885,
+      "learning_rate": 3.528009535160906e-06,
+      "loss": 0.0484,
+      "step": 1389
+    },
+    {
+      "epoch": 0.8283671036948749,
+      "grad_norm": 1.8688760995864868,
+      "learning_rate": 3.516090584028606e-06,
+      "loss": 0.0207,
+      "step": 1390
+    },
+    {
+      "epoch": 0.8289630512514898,
+      "grad_norm": 6.469098091125488,
+      "learning_rate": 3.5041716328963056e-06,
+      "loss": 0.4868,
+      "step": 1391
+    },
+    {
+      "epoch": 0.8295589988081049,
+      "grad_norm": 4.183995723724365,
+      "learning_rate": 3.4922526817640053e-06,
+      "loss": 0.1388,
+      "step": 1392
+    },
+    {
+      "epoch": 0.8301549463647199,
+      "grad_norm": 1.8615021705627441,
+      "learning_rate": 3.4803337306317046e-06,
+      "loss": 0.0224,
+      "step": 1393
+    },
+    {
+      "epoch": 0.8307508939213349,
+      "grad_norm": 3.0067672729492188,
+      "learning_rate": 3.4684147794994043e-06,
+      "loss": 0.0593,
+      "step": 1394
+    },
+    {
+      "epoch": 0.8313468414779499,
+      "grad_norm": 2.0249879360198975,
+      "learning_rate": 3.456495828367104e-06,
+      "loss": 0.0385,
+      "step": 1395
+    },
+    {
+      "epoch": 0.831942789034565,
+      "grad_norm": 3.6021742820739746,
+      "learning_rate": 3.4445768772348038e-06,
+      "loss": 0.0805,
+      "step": 1396
+    },
+    {
+      "epoch": 0.8325387365911799,
+      "grad_norm": 4.234012603759766,
+      "learning_rate": 3.4326579261025035e-06,
+      "loss": 0.1668,
+      "step": 1397
+    },
+    {
+      "epoch": 0.833134684147795,
+      "grad_norm": 2.1181154251098633,
+      "learning_rate": 3.420738974970203e-06,
+      "loss": 0.0471,
+      "step": 1398
+    },
+    {
+      "epoch": 0.83373063170441,
+      "grad_norm": 1.4991300106048584,
+      "learning_rate": 3.4088200238379025e-06,
+      "loss": 0.0112,
+      "step": 1399
+    },
+    {
+      "epoch": 0.834326579261025,
+      "grad_norm": 2.884045124053955,
+      "learning_rate": 3.3969010727056023e-06,
+      "loss": 0.0641,
+      "step": 1400
+    },
+    {
+      "epoch": 0.83492252681764,
+      "grad_norm": 0.7166174650192261,
+      "learning_rate": 3.384982121573302e-06,
+      "loss": 0.008,
+      "step": 1401
+    },
+    {
+      "epoch": 0.8355184743742551,
+      "grad_norm": 2.3950607776641846,
+      "learning_rate": 3.3730631704410017e-06,
+      "loss": 0.0563,
+      "step": 1402
+    },
+    {
+      "epoch": 0.83611442193087,
+      "grad_norm": 2.2635176181793213,
+      "learning_rate": 3.3611442193087015e-06,
+      "loss": 0.0536,
+      "step": 1403
+    },
+    {
+      "epoch": 0.8367103694874851,
+      "grad_norm": 4.702054023742676,
+      "learning_rate": 3.3492252681764008e-06,
+      "loss": 0.1241,
+      "step": 1404
+    },
+    {
+      "epoch": 0.8373063170441001,
+      "grad_norm": 2.1027443408966064,
+      "learning_rate": 3.3373063170441005e-06,
+      "loss": 0.0283,
+      "step": 1405
+    },
+    {
+      "epoch": 0.8379022646007152,
+      "grad_norm": 3.670628309249878,
+      "learning_rate": 3.3253873659118002e-06,
+      "loss": 0.0768,
+      "step": 1406
+    },
+    {
+      "epoch": 0.8384982121573301,
+      "grad_norm": 6.088999271392822,
+      "learning_rate": 3.3134684147795e-06,
+      "loss": 0.3618,
+      "step": 1407
+    },
+    {
+      "epoch": 0.8390941597139452,
+      "grad_norm": 2.691434144973755,
+      "learning_rate": 3.3015494636471997e-06,
+      "loss": 0.0447,
+      "step": 1408
+    },
+    {
+      "epoch": 0.8396901072705601,
+      "grad_norm": 0.9001061320304871,
+      "learning_rate": 3.289630512514899e-06,
+      "loss": 0.0113,
+      "step": 1409
+    },
+    {
+      "epoch": 0.8402860548271752,
+      "grad_norm": 2.283468723297119,
+      "learning_rate": 3.2777115613825987e-06,
+      "loss": 0.0292,
+      "step": 1410
+    },
+    {
+      "epoch": 0.8408820023837902,
+      "grad_norm": 4.893710613250732,
+      "learning_rate": 3.2657926102502985e-06,
+      "loss": 0.2056,
+      "step": 1411
+    },
+    {
+      "epoch": 0.8414779499404053,
+      "grad_norm": 2.945300579071045,
+      "learning_rate": 3.253873659117998e-06,
+      "loss": 0.0656,
+      "step": 1412
+    },
+    {
+      "epoch": 0.8420738974970202,
+      "grad_norm": 2.6728291511535645,
+      "learning_rate": 3.241954707985698e-06,
+      "loss": 0.0311,
+      "step": 1413
+    },
+    {
+      "epoch": 0.8426698450536353,
+      "grad_norm": 2.7345974445343018,
+      "learning_rate": 3.2300357568533972e-06,
+      "loss": 0.0414,
+      "step": 1414
+    },
+    {
+      "epoch": 0.8432657926102503,
+      "grad_norm": 3.424111843109131,
+      "learning_rate": 3.218116805721097e-06,
+      "loss": 0.1263,
+      "step": 1415
+    },
+    {
+      "epoch": 0.8438617401668653,
+      "grad_norm": 2.366262197494507,
+      "learning_rate": 3.2061978545887963e-06,
+      "loss": 0.0349,
+      "step": 1416
+    },
+    {
+      "epoch": 0.8444576877234803,
+      "grad_norm": 2.9889485836029053,
+      "learning_rate": 3.194278903456496e-06,
+      "loss": 0.0529,
+      "step": 1417
+    },
+    {
+      "epoch": 0.8450536352800954,
+      "grad_norm": 3.130985736846924,
+      "learning_rate": 3.1823599523241953e-06,
+      "loss": 0.1384,
+      "step": 1418
+    },
+    {
+      "epoch": 0.8456495828367103,
+      "grad_norm": 3.6531012058258057,
+      "learning_rate": 3.170441001191895e-06,
+      "loss": 0.1293,
+      "step": 1419
+    },
+    {
+      "epoch": 0.8462455303933254,
+      "grad_norm": 2.514796257019043,
+      "learning_rate": 3.1585220500595948e-06,
+      "loss": 0.0586,
+      "step": 1420
+    },
+    {
+      "epoch": 0.8468414779499404,
+      "grad_norm": 2.093137741088867,
+      "learning_rate": 3.1466030989272945e-06,
+      "loss": 0.0241,
+      "step": 1421
+    },
+    {
+      "epoch": 0.8474374255065554,
+      "grad_norm": 3.3545241355895996,
+      "learning_rate": 3.1346841477949942e-06,
+      "loss": 0.1476,
+      "step": 1422
+    },
+    {
+      "epoch": 0.8480333730631704,
+      "grad_norm": 4.011752128601074,
+      "learning_rate": 3.1227651966626935e-06,
+      "loss": 0.1126,
+      "step": 1423
+    },
+    {
+      "epoch": 0.8486293206197855,
+      "grad_norm": 2.252500295639038,
+      "learning_rate": 3.1108462455303933e-06,
+      "loss": 0.024,
+      "step": 1424
+    },
+    {
+      "epoch": 0.8492252681764004,
+      "grad_norm": 2.8281123638153076,
+      "learning_rate": 3.098927294398093e-06,
+      "loss": 0.0535,
+      "step": 1425
+    },
+    {
+      "epoch": 0.8498212157330155,
+      "grad_norm": 2.3812878131866455,
+      "learning_rate": 3.0870083432657927e-06,
+      "loss": 0.041,
+      "step": 1426
+    },
+    {
+      "epoch": 0.8504171632896305,
+      "grad_norm": 2.7914116382598877,
+      "learning_rate": 3.0750893921334925e-06,
+      "loss": 0.0524,
+      "step": 1427
+    },
+    {
+      "epoch": 0.8510131108462455,
+      "grad_norm": 2.006333589553833,
+      "learning_rate": 3.063170441001192e-06,
+      "loss": 0.0326,
+      "step": 1428
+    },
+    {
+      "epoch": 0.8516090584028605,
+      "grad_norm": 3.6660690307617188,
+      "learning_rate": 3.0512514898688915e-06,
+      "loss": 0.087,
+      "step": 1429
+    },
+    {
+      "epoch": 0.8522050059594756,
+      "grad_norm": 6.571730613708496,
+      "learning_rate": 3.0393325387365912e-06,
+      "loss": 0.3269,
+      "step": 1430
+    },
+    {
+      "epoch": 0.8528009535160905,
+      "grad_norm": 3.672942638397217,
+      "learning_rate": 3.027413587604291e-06,
+      "loss": 0.1303,
+      "step": 1431
+    },
+    {
+      "epoch": 0.8533969010727056,
+      "grad_norm": 4.3475022315979,
+      "learning_rate": 3.0154946364719907e-06,
+      "loss": 0.1921,
+      "step": 1432
+    },
+    {
+      "epoch": 0.8539928486293206,
+      "grad_norm": 1.4469280242919922,
+      "learning_rate": 3.0035756853396904e-06,
+      "loss": 0.0123,
+      "step": 1433
+    },
+    {
+      "epoch": 0.8545887961859356,
+      "grad_norm": 4.477729797363281,
+      "learning_rate": 2.9916567342073897e-06,
+      "loss": 0.1485,
+      "step": 1434
+    },
+    {
+      "epoch": 0.8551847437425506,
+      "grad_norm": 2.2894351482391357,
+      "learning_rate": 2.9797377830750894e-06,
+      "loss": 0.0514,
+      "step": 1435
+    },
+    {
+      "epoch": 0.8557806912991657,
+      "grad_norm": 1.8487260341644287,
+      "learning_rate": 2.967818831942789e-06,
+      "loss": 0.0313,
+      "step": 1436
+    },
+    {
+      "epoch": 0.8563766388557806,
+      "grad_norm": 0.9441725015640259,
+      "learning_rate": 2.955899880810489e-06,
+      "loss": 0.0156,
+      "step": 1437
+    },
+    {
+      "epoch": 0.8569725864123957,
+      "grad_norm": 4.637929916381836,
+      "learning_rate": 2.9439809296781886e-06,
+      "loss": 0.1507,
+      "step": 1438
+    },
+    {
+      "epoch": 0.8575685339690108,
+      "grad_norm": 2.537447690963745,
+      "learning_rate": 2.932061978545888e-06,
+      "loss": 0.0523,
+      "step": 1439
+    },
+    {
+      "epoch": 0.8581644815256257,
+      "grad_norm": 4.204341411590576,
+      "learning_rate": 2.9201430274135877e-06,
+      "loss": 0.1369,
+      "step": 1440
+    },
+    {
+      "epoch": 0.8587604290822408,
+      "grad_norm": 2.0728156566619873,
+      "learning_rate": 2.9082240762812874e-06,
+      "loss": 0.0285,
+      "step": 1441
+    },
+    {
+      "epoch": 0.8593563766388558,
+      "grad_norm": 6.536391258239746,
+      "learning_rate": 2.896305125148987e-06,
+      "loss": 0.4043,
+      "step": 1442
+    },
+    {
+      "epoch": 0.8599523241954709,
+      "grad_norm": 2.446197748184204,
+      "learning_rate": 2.884386174016687e-06,
+      "loss": 0.0546,
+      "step": 1443
+    },
+    {
+      "epoch": 0.8605482717520858,
+      "grad_norm": 3.9211833477020264,
+      "learning_rate": 2.8724672228843866e-06,
+      "loss": 0.1174,
+      "step": 1444
+    },
+    {
+      "epoch": 0.8611442193087009,
+      "grad_norm": 2.775639533996582,
+      "learning_rate": 2.860548271752086e-06,
+      "loss": 0.0624,
+      "step": 1445
+    },
+    {
+      "epoch": 0.8617401668653158,
+      "grad_norm": 2.2616357803344727,
+      "learning_rate": 2.8486293206197856e-06,
+      "loss": 0.0341,
+      "step": 1446
+    },
+    {
+      "epoch": 0.8623361144219309,
+      "grad_norm": 5.573391437530518,
+      "learning_rate": 2.8367103694874854e-06,
+      "loss": 0.2898,
+      "step": 1447
+    },
+    {
+      "epoch": 0.8629320619785459,
+      "grad_norm": 3.1679506301879883,
+      "learning_rate": 2.824791418355185e-06,
+      "loss": 0.0729,
+      "step": 1448
+    },
+    {
+      "epoch": 0.863528009535161,
+      "grad_norm": 1.4207916259765625,
+      "learning_rate": 2.812872467222885e-06,
+      "loss": 0.016,
+      "step": 1449
+    },
+    {
+      "epoch": 0.8641239570917759,
+      "grad_norm": 2.910435914993286,
+      "learning_rate": 2.800953516090584e-06,
+      "loss": 0.0347,
+      "step": 1450
+    },
+    {
+      "epoch": 0.864719904648391,
+      "grad_norm": 3.005824565887451,
+      "learning_rate": 2.789034564958284e-06,
+      "loss": 0.0486,
+      "step": 1451
+    },
+    {
+      "epoch": 0.865315852205006,
+      "grad_norm": 2.193031072616577,
+      "learning_rate": 2.7771156138259836e-06,
+      "loss": 0.033,
+      "step": 1452
+    },
+    {
+      "epoch": 0.865911799761621,
+      "grad_norm": 2.646733045578003,
+      "learning_rate": 2.7651966626936833e-06,
+      "loss": 0.0322,
+      "step": 1453
+    },
+    {
+      "epoch": 0.866507747318236,
+      "grad_norm": 2.7761952877044678,
+      "learning_rate": 2.753277711561383e-06,
+      "loss": 0.0724,
+      "step": 1454
+    },
+    {
+      "epoch": 0.8671036948748511,
+      "grad_norm": 6.215579509735107,
+      "learning_rate": 2.7413587604290824e-06,
+      "loss": 0.4059,
+      "step": 1455
+    },
+    {
+      "epoch": 0.867699642431466,
+      "grad_norm": 2.5482070446014404,
+      "learning_rate": 2.729439809296782e-06,
+      "loss": 0.0569,
+      "step": 1456
+    },
+    {
+      "epoch": 0.8682955899880811,
+      "grad_norm": 3.668600559234619,
+      "learning_rate": 2.717520858164482e-06,
+      "loss": 0.0524,
+      "step": 1457
+    },
+    {
+      "epoch": 0.8688915375446961,
+      "grad_norm": 4.553552627563477,
+      "learning_rate": 2.7056019070321816e-06,
+      "loss": 0.1382,
+      "step": 1458
+    },
+    {
+      "epoch": 0.8694874851013111,
+      "grad_norm": 4.349089622497559,
+      "learning_rate": 2.6936829558998813e-06,
+      "loss": 0.1644,
+      "step": 1459
+    },
+    {
+      "epoch": 0.8700834326579261,
+      "grad_norm": 2.398308753967285,
+      "learning_rate": 2.681764004767581e-06,
+      "loss": 0.0429,
+      "step": 1460
+    },
+    {
+      "epoch": 0.8706793802145412,
+      "grad_norm": 2.8538005352020264,
+      "learning_rate": 2.6698450536352803e-06,
+      "loss": 0.0914,
+      "step": 1461
+    },
+    {
+      "epoch": 0.8712753277711561,
+      "grad_norm": 8.578140258789062,
+      "learning_rate": 2.65792610250298e-06,
+      "loss": 0.4157,
+      "step": 1462
+    },
+    {
+      "epoch": 0.8718712753277712,
+      "grad_norm": 3.0562233924865723,
+      "learning_rate": 2.6460071513706798e-06,
+      "loss": 0.0817,
+      "step": 1463
+    },
+    {
+      "epoch": 0.8724672228843862,
+      "grad_norm": 3.730272054672241,
+      "learning_rate": 2.6340882002383795e-06,
+      "loss": 0.142,
+      "step": 1464
+    },
+    {
+      "epoch": 0.8730631704410012,
+      "grad_norm": 1.9700920581817627,
+      "learning_rate": 2.6221692491060792e-06,
+      "loss": 0.0364,
+      "step": 1465
+    },
+    {
+      "epoch": 0.8736591179976162,
+      "grad_norm": 2.811162233352661,
+      "learning_rate": 2.6102502979737785e-06,
+      "loss": 0.0471,
+      "step": 1466
+    },
+    {
+      "epoch": 0.8742550655542313,
+      "grad_norm": 2.3930492401123047,
+      "learning_rate": 2.5983313468414783e-06,
+      "loss": 0.0269,
+      "step": 1467
+    },
+    {
+      "epoch": 0.8748510131108462,
+      "grad_norm": 4.60854434967041,
+      "learning_rate": 2.586412395709178e-06,
+      "loss": 0.1627,
+      "step": 1468
+    },
+    {
+      "epoch": 0.8754469606674613,
+      "grad_norm": 2.0338821411132812,
+      "learning_rate": 2.5744934445768777e-06,
+      "loss": 0.0313,
+      "step": 1469
+    },
+    {
+      "epoch": 0.8760429082240763,
+      "grad_norm": 3.286135196685791,
+      "learning_rate": 2.5625744934445775e-06,
+      "loss": 0.074,
+      "step": 1470
+    },
+    {
+      "epoch": 0.8766388557806913,
+      "grad_norm": 1.6500340700149536,
+      "learning_rate": 2.5506555423122768e-06,
+      "loss": 0.014,
+      "step": 1471
+    },
+    {
+      "epoch": 0.8772348033373063,
+      "grad_norm": 3.7283742427825928,
+      "learning_rate": 2.5387365911799765e-06,
+      "loss": 0.1182,
+      "step": 1472
+    },
+    {
+      "epoch": 0.8778307508939214,
+      "grad_norm": 4.866960525512695,
+      "learning_rate": 2.5268176400476762e-06,
+      "loss": 0.2723,
+      "step": 1473
+    },
+    {
+      "epoch": 0.8784266984505363,
+      "grad_norm": 2.5320045948028564,
+      "learning_rate": 2.514898688915376e-06,
+      "loss": 0.0878,
+      "step": 1474
+    },
+    {
+      "epoch": 0.8790226460071514,
+      "grad_norm": 2.8843231201171875,
+      "learning_rate": 2.5029797377830757e-06,
+      "loss": 0.0714,
+      "step": 1475
+    },
+    {
+      "epoch": 0.8796185935637664,
+      "grad_norm": 1.5001121759414673,
+      "learning_rate": 2.491060786650775e-06,
+      "loss": 0.0148,
+      "step": 1476
+    },
+    {
+      "epoch": 0.8802145411203814,
+      "grad_norm": 6.51433801651001,
+      "learning_rate": 2.4791418355184747e-06,
+      "loss": 0.3725,
+      "step": 1477
+    },
+    {
+      "epoch": 0.8808104886769964,
+      "grad_norm": 3.046895980834961,
+      "learning_rate": 2.467222884386174e-06,
+      "loss": 0.0786,
+      "step": 1478
+    },
+    {
+      "epoch": 0.8814064362336115,
+      "grad_norm": 3.9337496757507324,
+      "learning_rate": 2.4553039332538738e-06,
+      "loss": 0.0554,
+      "step": 1479
+    },
+    {
+      "epoch": 0.8820023837902264,
+      "grad_norm": 1.8371058702468872,
+      "learning_rate": 2.4433849821215735e-06,
+      "loss": 0.0327,
+      "step": 1480
+    },
+    {
+      "epoch": 0.8825983313468415,
+      "grad_norm": 2.3079159259796143,
+      "learning_rate": 2.4314660309892732e-06,
+      "loss": 0.0663,
+      "step": 1481
+    },
+    {
+      "epoch": 0.8831942789034565,
+      "grad_norm": 3.908693313598633,
+      "learning_rate": 2.419547079856973e-06,
+      "loss": 0.1249,
+      "step": 1482
+    },
+    {
+      "epoch": 0.8837902264600715,
+      "grad_norm": 4.256025791168213,
+      "learning_rate": 2.4076281287246727e-06,
+      "loss": 0.1705,
+      "step": 1483
+    },
+    {
+      "epoch": 0.8843861740166865,
+      "grad_norm": 2.2597267627716064,
+      "learning_rate": 2.395709177592372e-06,
+      "loss": 0.0824,
+      "step": 1484
+    },
+    {
+      "epoch": 0.8849821215733016,
+      "grad_norm": 1.906375527381897,
+      "learning_rate": 2.3837902264600717e-06,
+      "loss": 0.0319,
+      "step": 1485
+    },
+    {
+      "epoch": 0.8855780691299165,
+      "grad_norm": 2.0713112354278564,
+      "learning_rate": 2.3718712753277715e-06,
+      "loss": 0.0354,
+      "step": 1486
+    },
+    {
+      "epoch": 0.8861740166865316,
+      "grad_norm": 2.6114954948425293,
+      "learning_rate": 2.359952324195471e-06,
+      "loss": 0.0468,
+      "step": 1487
+    },
+    {
+      "epoch": 0.8867699642431466,
+      "grad_norm": 2.957955837249756,
+      "learning_rate": 2.348033373063171e-06,
+      "loss": 0.0784,
+      "step": 1488
+    },
+    {
+      "epoch": 0.8873659117997617,
+      "grad_norm": 2.2929766178131104,
+      "learning_rate": 2.3361144219308702e-06,
+      "loss": 0.0411,
+      "step": 1489
+    },
+    {
+      "epoch": 0.8879618593563766,
+      "grad_norm": 4.406092643737793,
+      "learning_rate": 2.32419547079857e-06,
+      "loss": 0.3203,
+      "step": 1490
+    },
+    {
+      "epoch": 0.8885578069129917,
+      "grad_norm": 2.072228193283081,
+      "learning_rate": 2.3122765196662693e-06,
+      "loss": 0.0233,
+      "step": 1491
+    },
+    {
+      "epoch": 0.8891537544696066,
+      "grad_norm": 1.0417461395263672,
+      "learning_rate": 2.300357568533969e-06,
+      "loss": 0.0147,
+      "step": 1492
+    },
+    {
+      "epoch": 0.8897497020262217,
+      "grad_norm": 1.4013792276382446,
+      "learning_rate": 2.2884386174016687e-06,
+      "loss": 0.0178,
+      "step": 1493
+    },
+    {
+      "epoch": 0.8903456495828367,
+      "grad_norm": 1.801497220993042,
+      "learning_rate": 2.2765196662693685e-06,
+      "loss": 0.0149,
+      "step": 1494
+    },
+    {
+      "epoch": 0.8909415971394518,
+      "grad_norm": 3.8414456844329834,
+      "learning_rate": 2.264600715137068e-06,
+      "loss": 0.1001,
+      "step": 1495
+    },
+    {
+      "epoch": 0.8915375446960667,
+      "grad_norm": 2.6731340885162354,
+      "learning_rate": 2.2526817640047675e-06,
+      "loss": 0.056,
+      "step": 1496
+    },
+    {
+      "epoch": 0.8921334922526818,
+      "grad_norm": 1.7622570991516113,
+      "learning_rate": 2.2407628128724672e-06,
+      "loss": 0.0312,
+      "step": 1497
+    },
+    {
+      "epoch": 0.8927294398092968,
+      "grad_norm": 2.419862985610962,
+      "learning_rate": 2.228843861740167e-06,
+      "loss": 0.0555,
+      "step": 1498
+    },
+    {
+      "epoch": 0.8933253873659118,
+      "grad_norm": 4.080726623535156,
+      "learning_rate": 2.2169249106078667e-06,
+      "loss": 0.1372,
+      "step": 1499
+    },
+    {
+      "epoch": 0.8939213349225268,
+      "grad_norm": 2.2697556018829346,
+      "learning_rate": 2.2050059594755664e-06,
+      "loss": 0.0403,
+      "step": 1500
+    },
+    {
+      "epoch": 0.8945172824791419,
+      "grad_norm": 1.9136961698532104,
+      "learning_rate": 2.193087008343266e-06,
+      "loss": 0.0251,
+      "step": 1501
+    },
+    {
+      "epoch": 0.8951132300357568,
+      "grad_norm": 1.2775863409042358,
+      "learning_rate": 2.1811680572109654e-06,
+      "loss": 0.0222,
+      "step": 1502
+    },
+    {
+      "epoch": 0.8957091775923719,
+      "grad_norm": 2.4846091270446777,
+      "learning_rate": 2.169249106078665e-06,
+      "loss": 0.0576,
+      "step": 1503
+    },
+    {
+      "epoch": 0.8963051251489869,
+      "grad_norm": 3.0741591453552246,
+      "learning_rate": 2.157330154946365e-06,
+      "loss": 0.0603,
+      "step": 1504
+    },
+    {
+      "epoch": 0.8969010727056019,
+      "grad_norm": 4.0735015869140625,
+      "learning_rate": 2.1454112038140646e-06,
+      "loss": 0.0891,
+      "step": 1505
+    },
+    {
+      "epoch": 0.8974970202622169,
+      "grad_norm": 7.8920817375183105,
+      "learning_rate": 2.1334922526817644e-06,
+      "loss": 0.4072,
+      "step": 1506
+    },
+    {
+      "epoch": 0.898092967818832,
+      "grad_norm": 3.011532783508301,
+      "learning_rate": 2.1215733015494637e-06,
+      "loss": 0.0565,
+      "step": 1507
+    },
+    {
+      "epoch": 0.8986889153754469,
+      "grad_norm": 3.2934131622314453,
+      "learning_rate": 2.1096543504171634e-06,
+      "loss": 0.1384,
+      "step": 1508
+    },
+    {
+      "epoch": 0.899284862932062,
+      "grad_norm": 2.269780158996582,
+      "learning_rate": 2.097735399284863e-06,
+      "loss": 0.0366,
+      "step": 1509
+    },
+    {
+      "epoch": 0.899880810488677,
+      "grad_norm": 3.0726797580718994,
+      "learning_rate": 2.085816448152563e-06,
+      "loss": 0.0758,
+      "step": 1510
+    },
+    {
+      "epoch": 0.900476758045292,
+      "grad_norm": 1.3454055786132812,
+      "learning_rate": 2.0738974970202626e-06,
+      "loss": 0.0191,
+      "step": 1511
+    },
+    {
+      "epoch": 0.901072705601907,
+      "grad_norm": 1.6831629276275635,
+      "learning_rate": 2.061978545887962e-06,
+      "loss": 0.026,
+      "step": 1512
+    },
+    {
+      "epoch": 0.9016686531585221,
+      "grad_norm": 5.08050012588501,
+      "learning_rate": 2.0500595947556616e-06,
+      "loss": 0.2799,
+      "step": 1513
+    },
+    {
+      "epoch": 0.902264600715137,
+      "grad_norm": 1.5127214193344116,
+      "learning_rate": 2.0381406436233614e-06,
+      "loss": 0.0241,
+      "step": 1514
+    },
+    {
+      "epoch": 0.9028605482717521,
+      "grad_norm": 4.180816173553467,
+      "learning_rate": 2.026221692491061e-06,
+      "loss": 0.1696,
+      "step": 1515
+    },
+    {
+      "epoch": 0.9034564958283671,
+      "grad_norm": 1.197894811630249,
+      "learning_rate": 2.014302741358761e-06,
+      "loss": 0.013,
+      "step": 1516
+    },
+    {
+      "epoch": 0.9040524433849821,
+      "grad_norm": 4.050142765045166,
+      "learning_rate": 2.00238379022646e-06,
+      "loss": 0.1307,
+      "step": 1517
+    },
+    {
+      "epoch": 0.9046483909415971,
+      "grad_norm": 2.7219483852386475,
+      "learning_rate": 1.99046483909416e-06,
+      "loss": 0.0614,
+      "step": 1518
+    },
+    {
+      "epoch": 0.9052443384982122,
+      "grad_norm": 2.3871822357177734,
+      "learning_rate": 1.9785458879618596e-06,
+      "loss": 0.0511,
+      "step": 1519
+    },
+    {
+      "epoch": 0.9058402860548271,
+      "grad_norm": 2.3907227516174316,
+      "learning_rate": 1.9666269368295593e-06,
+      "loss": 0.0438,
+      "step": 1520
+    },
+    {
+      "epoch": 0.9064362336114422,
+      "grad_norm": 2.9418601989746094,
+      "learning_rate": 1.9547079856972586e-06,
+      "loss": 0.0742,
+      "step": 1521
+    },
+    {
+      "epoch": 0.9070321811680572,
+      "grad_norm": 3.893791913986206,
+      "learning_rate": 1.9427890345649584e-06,
+      "loss": 0.0817,
+      "step": 1522
+    },
+    {
+      "epoch": 0.9076281287246722,
+      "grad_norm": 5.582831382751465,
+      "learning_rate": 1.930870083432658e-06,
+      "loss": 0.1957,
+      "step": 1523
+    },
+    {
+      "epoch": 0.9082240762812872,
+      "grad_norm": 1.5124911069869995,
+      "learning_rate": 1.918951132300358e-06,
+      "loss": 0.0223,
+      "step": 1524
+    },
+    {
+      "epoch": 0.9088200238379023,
+      "grad_norm": 4.09805965423584,
+      "learning_rate": 1.9070321811680573e-06,
+      "loss": 0.1793,
+      "step": 1525
+    },
+    {
+      "epoch": 0.9094159713945172,
+      "grad_norm": 1.7736088037490845,
+      "learning_rate": 1.8951132300357569e-06,
+      "loss": 0.041,
+      "step": 1526
+    },
+    {
+      "epoch": 0.9100119189511323,
+      "grad_norm": 3.9493987560272217,
+      "learning_rate": 1.8831942789034566e-06,
+      "loss": 0.0758,
+      "step": 1527
+    },
+    {
+      "epoch": 0.9106078665077473,
+      "grad_norm": 2.371330976486206,
+      "learning_rate": 1.8712753277711563e-06,
+      "loss": 0.0722,
+      "step": 1528
+    },
+    {
+      "epoch": 0.9112038140643623,
+      "grad_norm": 3.869568109512329,
+      "learning_rate": 1.8593563766388558e-06,
+      "loss": 0.1019,
+      "step": 1529
+    },
+    {
+      "epoch": 0.9117997616209773,
+      "grad_norm": 13.198070526123047,
+      "learning_rate": 1.8474374255065556e-06,
+      "loss": 1.0408,
+      "step": 1530
+    },
+    {
+      "epoch": 0.9123957091775924,
+      "grad_norm": 3.0595428943634033,
+      "learning_rate": 1.835518474374255e-06,
+      "loss": 0.0649,
+      "step": 1531
+    },
+    {
+      "epoch": 0.9129916567342073,
+      "grad_norm": 1.4733202457427979,
+      "learning_rate": 1.8235995232419548e-06,
+      "loss": 0.0131,
+      "step": 1532
+    },
+    {
+      "epoch": 0.9135876042908224,
+      "grad_norm": 2.596548080444336,
+      "learning_rate": 1.8116805721096545e-06,
+      "loss": 0.0428,
+      "step": 1533
+    },
+    {
+      "epoch": 0.9141835518474374,
+      "grad_norm": 3.699479103088379,
+      "learning_rate": 1.799761620977354e-06,
+      "loss": 0.0965,
+      "step": 1534
+    },
+    {
+      "epoch": 0.9147794994040525,
+      "grad_norm": 3.1251769065856934,
+      "learning_rate": 1.7878426698450538e-06,
+      "loss": 0.0442,
+      "step": 1535
+    },
+    {
+      "epoch": 0.9153754469606674,
+      "grad_norm": 2.2160158157348633,
+      "learning_rate": 1.7759237187127535e-06,
+      "loss": 0.0564,
+      "step": 1536
+    },
+    {
+      "epoch": 0.9159713945172825,
+      "grad_norm": 1.5512466430664062,
+      "learning_rate": 1.764004767580453e-06,
+      "loss": 0.0206,
+      "step": 1537
+    },
+    {
+      "epoch": 0.9165673420738975,
+      "grad_norm": 3.299013376235962,
+      "learning_rate": 1.7520858164481528e-06,
+      "loss": 0.1035,
+      "step": 1538
+    },
+    {
+      "epoch": 0.9171632896305125,
+      "grad_norm": 3.2081875801086426,
+      "learning_rate": 1.7401668653158523e-06,
+      "loss": 0.1057,
+      "step": 1539
+    },
+    {
+      "epoch": 0.9177592371871275,
+      "grad_norm": 3.062593698501587,
+      "learning_rate": 1.728247914183552e-06,
+      "loss": 0.1328,
+      "step": 1540
+    },
+    {
+      "epoch": 0.9183551847437426,
+      "grad_norm": 3.771188497543335,
+      "learning_rate": 1.7163289630512518e-06,
+      "loss": 0.099,
+      "step": 1541
+    },
+    {
+      "epoch": 0.9189511323003575,
+      "grad_norm": 2.8904314041137695,
+      "learning_rate": 1.7044100119189513e-06,
+      "loss": 0.0596,
+      "step": 1542
+    },
+    {
+      "epoch": 0.9195470798569726,
+      "grad_norm": 0.8863335251808167,
+      "learning_rate": 1.692491060786651e-06,
+      "loss": 0.0209,
+      "step": 1543
+    },
+    {
+      "epoch": 0.9201430274135876,
+      "grad_norm": 2.9656665325164795,
+      "learning_rate": 1.6805721096543507e-06,
+      "loss": 0.0713,
+      "step": 1544
+    },
+    {
+      "epoch": 0.9207389749702026,
+      "grad_norm": 3.9244027137756348,
+      "learning_rate": 1.6686531585220503e-06,
+      "loss": 0.0937,
+      "step": 1545
+    },
+    {
+      "epoch": 0.9213349225268176,
+      "grad_norm": 3.685025215148926,
+      "learning_rate": 1.65673420738975e-06,
+      "loss": 0.0991,
+      "step": 1546
+    },
+    {
+      "epoch": 0.9219308700834327,
+      "grad_norm": 2.4365339279174805,
+      "learning_rate": 1.6448152562574495e-06,
+      "loss": 0.0531,
+      "step": 1547
+    },
+    {
+      "epoch": 0.9225268176400476,
+      "grad_norm": 3.6056761741638184,
+      "learning_rate": 1.6328963051251492e-06,
+      "loss": 0.0702,
+      "step": 1548
+    },
+    {
+      "epoch": 0.9231227651966627,
+      "grad_norm": 2.510640859603882,
+      "learning_rate": 1.620977353992849e-06,
+      "loss": 0.0494,
+      "step": 1549
+    },
+    {
+      "epoch": 0.9237187127532777,
+      "grad_norm": 6.178497314453125,
+      "learning_rate": 1.6090584028605485e-06,
+      "loss": 0.3186,
+      "step": 1550
+    },
+    {
+      "epoch": 0.9243146603098927,
+      "grad_norm": 2.538041830062866,
+      "learning_rate": 1.597139451728248e-06,
+      "loss": 0.0389,
+      "step": 1551
+    },
+    {
+      "epoch": 0.9249106078665077,
+      "grad_norm": 2.3738415241241455,
+      "learning_rate": 1.5852205005959475e-06,
+      "loss": 0.0369,
+      "step": 1552
+    },
+    {
+      "epoch": 0.9255065554231228,
+      "grad_norm": 2.3635497093200684,
+      "learning_rate": 1.5733015494636472e-06,
+      "loss": 0.0593,
+      "step": 1553
+    },
+    {
+      "epoch": 0.9261025029797377,
+      "grad_norm": 4.1664347648620605,
+      "learning_rate": 1.5613825983313468e-06,
+      "loss": 0.169,
+      "step": 1554
+    },
+    {
+      "epoch": 0.9266984505363528,
+      "grad_norm": 2.4234297275543213,
+      "learning_rate": 1.5494636471990465e-06,
+      "loss": 0.0553,
+      "step": 1555
+    },
+    {
+      "epoch": 0.9272943980929678,
+      "grad_norm": 1.5025144815444946,
+      "learning_rate": 1.5375446960667462e-06,
+      "loss": 0.0164,
+      "step": 1556
+    },
+    {
+      "epoch": 0.9278903456495828,
+      "grad_norm": 2.7447378635406494,
+      "learning_rate": 1.5256257449344457e-06,
+      "loss": 0.0431,
+      "step": 1557
+    },
+    {
+      "epoch": 0.9284862932061978,
+      "grad_norm": 2.2552292346954346,
+      "learning_rate": 1.5137067938021455e-06,
+      "loss": 0.0299,
+      "step": 1558
+    },
+    {
+      "epoch": 0.9290822407628129,
+      "grad_norm": 3.6033928394317627,
+      "learning_rate": 1.5017878426698452e-06,
+      "loss": 0.1066,
+      "step": 1559
+    },
+    {
+      "epoch": 0.929678188319428,
+      "grad_norm": 2.74593186378479,
+      "learning_rate": 1.4898688915375447e-06,
+      "loss": 0.0762,
+      "step": 1560
+    },
+    {
+      "epoch": 0.9302741358760429,
+      "grad_norm": 3.92916202545166,
+      "learning_rate": 1.4779499404052445e-06,
+      "loss": 0.1554,
+      "step": 1561
+    },
+    {
+      "epoch": 0.930870083432658,
+      "grad_norm": 2.7704644203186035,
+      "learning_rate": 1.466030989272944e-06,
+      "loss": 0.0827,
+      "step": 1562
+    },
+    {
+      "epoch": 0.9314660309892729,
+      "grad_norm": 1.7968323230743408,
+      "learning_rate": 1.4541120381406437e-06,
+      "loss": 0.0133,
+      "step": 1563
+    },
+    {
+      "epoch": 0.932061978545888,
+      "grad_norm": 1.763094425201416,
+      "learning_rate": 1.4421930870083434e-06,
+      "loss": 0.0293,
+      "step": 1564
+    },
+    {
+      "epoch": 0.932657926102503,
+      "grad_norm": 2.50945782661438,
+      "learning_rate": 1.430274135876043e-06,
+      "loss": 0.0305,
+      "step": 1565
+    },
+    {
+      "epoch": 0.933253873659118,
+      "grad_norm": 3.618032693862915,
+      "learning_rate": 1.4183551847437427e-06,
+      "loss": 0.2028,
+      "step": 1566
+    },
+    {
+      "epoch": 0.933849821215733,
+      "grad_norm": 3.760291814804077,
+      "learning_rate": 1.4064362336114424e-06,
+      "loss": 0.07,
+      "step": 1567
+    },
+    {
+      "epoch": 0.9344457687723481,
+      "grad_norm": 3.859217882156372,
+      "learning_rate": 1.394517282479142e-06,
+      "loss": 0.1645,
+      "step": 1568
+    },
+    {
+      "epoch": 0.935041716328963,
+      "grad_norm": 5.459234237670898,
+      "learning_rate": 1.3825983313468417e-06,
+      "loss": 0.2834,
+      "step": 1569
+    },
+    {
+      "epoch": 0.9356376638855781,
+      "grad_norm": 2.329538106918335,
+      "learning_rate": 1.3706793802145412e-06,
+      "loss": 0.036,
+      "step": 1570
+    },
+    {
+      "epoch": 0.9362336114421931,
+      "grad_norm": 2.236847400665283,
+      "learning_rate": 1.358760429082241e-06,
+      "loss": 0.0342,
+      "step": 1571
+    },
+    {
+      "epoch": 0.9368295589988082,
+      "grad_norm": 3.093256950378418,
+      "learning_rate": 1.3468414779499406e-06,
+      "loss": 0.0823,
+      "step": 1572
+    },
+    {
+      "epoch": 0.9374255065554231,
+      "grad_norm": 2.1845829486846924,
+      "learning_rate": 1.3349225268176402e-06,
+      "loss": 0.0321,
+      "step": 1573
+    },
+    {
+      "epoch": 0.9380214541120382,
+      "grad_norm": 2.3721911907196045,
+      "learning_rate": 1.3230035756853399e-06,
+      "loss": 0.0407,
+      "step": 1574
+    },
+    {
+      "epoch": 0.9386174016686532,
+      "grad_norm": 3.254718065261841,
+      "learning_rate": 1.3110846245530396e-06,
+      "loss": 0.0636,
+      "step": 1575
+    },
+    {
+      "epoch": 0.9392133492252682,
+      "grad_norm": 8.438640594482422,
+      "learning_rate": 1.2991656734207391e-06,
+      "loss": 0.4648,
+      "step": 1576
+    },
+    {
+      "epoch": 0.9398092967818832,
+      "grad_norm": 2.1724629402160645,
+      "learning_rate": 1.2872467222884389e-06,
+      "loss": 0.0355,
+      "step": 1577
+    },
+    {
+      "epoch": 0.9404052443384983,
+      "grad_norm": 1.1918535232543945,
+      "learning_rate": 1.2753277711561384e-06,
+      "loss": 0.017,
+      "step": 1578
+    },
+    {
+      "epoch": 0.9410011918951132,
+      "grad_norm": 1.5868169069290161,
+      "learning_rate": 1.2634088200238381e-06,
+      "loss": 0.0323,
+      "step": 1579
+    },
+    {
+      "epoch": 0.9415971394517283,
+      "grad_norm": 1.609222412109375,
+      "learning_rate": 1.2514898688915378e-06,
+      "loss": 0.0211,
+      "step": 1580
+    },
+    {
+      "epoch": 0.9421930870083433,
+      "grad_norm": 2.7724976539611816,
+      "learning_rate": 1.2395709177592374e-06,
+      "loss": 0.0287,
+      "step": 1581
+    },
+    {
+      "epoch": 0.9427890345649583,
+      "grad_norm": 3.7431795597076416,
+      "learning_rate": 1.2276519666269369e-06,
+      "loss": 0.1367,
+      "step": 1582
+    },
+    {
+      "epoch": 0.9433849821215733,
+      "grad_norm": 1.8297927379608154,
+      "learning_rate": 1.2157330154946366e-06,
+      "loss": 0.0342,
+      "step": 1583
+    },
+    {
+      "epoch": 0.9439809296781884,
+      "grad_norm": 4.0742058753967285,
+      "learning_rate": 1.2038140643623363e-06,
+      "loss": 0.1623,
+      "step": 1584
+    },
+    {
+      "epoch": 0.9445768772348033,
+      "grad_norm": 3.9633004665374756,
+      "learning_rate": 1.1918951132300359e-06,
+      "loss": 0.0866,
+      "step": 1585
+    },
+    {
+      "epoch": 0.9451728247914184,
+      "grad_norm": 2.410407543182373,
+      "learning_rate": 1.1799761620977356e-06,
+      "loss": 0.0515,
+      "step": 1586
+    },
+    {
+      "epoch": 0.9457687723480334,
+      "grad_norm": 6.614712238311768,
+      "learning_rate": 1.1680572109654351e-06,
+      "loss": 0.3694,
+      "step": 1587
+    },
+    {
+      "epoch": 0.9463647199046484,
+      "grad_norm": 2.4917004108428955,
+      "learning_rate": 1.1561382598331346e-06,
+      "loss": 0.0427,
+      "step": 1588
+    },
+    {
+      "epoch": 0.9469606674612634,
+      "grad_norm": 2.6107895374298096,
+      "learning_rate": 1.1442193087008344e-06,
+      "loss": 0.0398,
+      "step": 1589
+    },
+    {
+      "epoch": 0.9475566150178785,
+      "grad_norm": 1.960059404373169,
+      "learning_rate": 1.132300357568534e-06,
+      "loss": 0.0182,
+      "step": 1590
+    },
+    {
+      "epoch": 0.9481525625744934,
+      "grad_norm": 1.7278741598129272,
+      "learning_rate": 1.1203814064362336e-06,
+      "loss": 0.0211,
+      "step": 1591
+    },
+    {
+      "epoch": 0.9487485101311085,
+      "grad_norm": 5.575089454650879,
+      "learning_rate": 1.1084624553039333e-06,
+      "loss": 0.426,
+      "step": 1592
+    },
+    {
+      "epoch": 0.9493444576877235,
+      "grad_norm": 1.820632815361023,
+      "learning_rate": 1.096543504171633e-06,
+      "loss": 0.0244,
+      "step": 1593
+    },
+    {
+      "epoch": 0.9499404052443385,
+      "grad_norm": 2.8208107948303223,
+      "learning_rate": 1.0846245530393326e-06,
+      "loss": 0.0682,
+      "step": 1594
+    },
+    {
+      "epoch": 0.9505363528009535,
+      "grad_norm": 10.904874801635742,
+      "learning_rate": 1.0727056019070323e-06,
+      "loss": 0.9416,
+      "step": 1595
+    },
+    {
+      "epoch": 0.9511323003575686,
+      "grad_norm": 2.1957781314849854,
+      "learning_rate": 1.0607866507747318e-06,
+      "loss": 0.024,
+      "step": 1596
+    },
+    {
+      "epoch": 0.9517282479141835,
+      "grad_norm": 3.5120437145233154,
+      "learning_rate": 1.0488676996424316e-06,
+      "loss": 0.1083,
+      "step": 1597
+    },
+    {
+      "epoch": 0.9523241954707986,
+      "grad_norm": 1.7157765626907349,
+      "learning_rate": 1.0369487485101313e-06,
+      "loss": 0.0393,
+      "step": 1598
+    },
+    {
+      "epoch": 0.9529201430274136,
+      "grad_norm": 3.9517483711242676,
+      "learning_rate": 1.0250297973778308e-06,
+      "loss": 0.1693,
+      "step": 1599
+    },
+    {
+      "epoch": 0.9535160905840286,
+      "grad_norm": 5.213629722595215,
+      "learning_rate": 1.0131108462455305e-06,
+      "loss": 0.3,
+      "step": 1600
+    },
+    {
+      "epoch": 0.9541120381406436,
+      "grad_norm": 5.265130043029785,
+      "learning_rate": 1.00119189511323e-06,
+      "loss": 0.3323,
+      "step": 1601
+    },
+    {
+      "epoch": 0.9547079856972587,
+      "grad_norm": 2.6802587509155273,
+      "learning_rate": 9.892729439809298e-07,
+      "loss": 0.0607,
+      "step": 1602
+    },
+    {
+      "epoch": 0.9553039332538736,
+      "grad_norm": 1.5229887962341309,
+      "learning_rate": 9.773539928486293e-07,
+      "loss": 0.0163,
+      "step": 1603
+    },
+    {
+      "epoch": 0.9558998808104887,
+      "grad_norm": 2.0646655559539795,
+      "learning_rate": 9.65435041716329e-07,
+      "loss": 0.0403,
+      "step": 1604
+    },
+    {
+      "epoch": 0.9564958283671037,
+      "grad_norm": 0.8683291673660278,
+      "learning_rate": 9.535160905840287e-07,
+      "loss": 0.0134,
+      "step": 1605
+    },
+    {
+      "epoch": 0.9570917759237187,
+      "grad_norm": 5.255694389343262,
+      "learning_rate": 9.415971394517283e-07,
+      "loss": 0.194,
+      "step": 1606
+    },
+    {
+      "epoch": 0.9576877234803337,
+      "grad_norm": 1.8716148138046265,
+      "learning_rate": 9.296781883194279e-07,
+      "loss": 0.0301,
+      "step": 1607
+    },
+    {
+      "epoch": 0.9582836710369488,
+      "grad_norm": 1.2501001358032227,
+      "learning_rate": 9.177592371871275e-07,
+      "loss": 0.0119,
+      "step": 1608
+    },
+    {
+      "epoch": 0.9588796185935637,
+      "grad_norm": 1.719030499458313,
+      "learning_rate": 9.058402860548273e-07,
+      "loss": 0.0234,
+      "step": 1609
+    },
+    {
+      "epoch": 0.9594755661501788,
+      "grad_norm": 1.1037157773971558,
+      "learning_rate": 8.939213349225269e-07,
+      "loss": 0.0206,
+      "step": 1610
+    },
+    {
+      "epoch": 0.9600715137067938,
+      "grad_norm": 2.3820929527282715,
+      "learning_rate": 8.820023837902265e-07,
+      "loss": 0.0315,
+      "step": 1611
+    },
+    {
+      "epoch": 0.9606674612634089,
+      "grad_norm": 2.370861053466797,
+      "learning_rate": 8.700834326579261e-07,
+      "loss": 0.0408,
+      "step": 1612
+    },
+    {
+      "epoch": 0.9612634088200238,
+      "grad_norm": 2.7671847343444824,
+      "learning_rate": 8.581644815256259e-07,
+      "loss": 0.0502,
+      "step": 1613
+    },
+    {
+      "epoch": 0.9618593563766389,
+      "grad_norm": 2.477128267288208,
+      "learning_rate": 8.462455303933255e-07,
+      "loss": 0.0525,
+      "step": 1614
+    },
+    {
+      "epoch": 0.9624553039332538,
+      "grad_norm": 2.7510149478912354,
+      "learning_rate": 8.343265792610251e-07,
+      "loss": 0.0529,
+      "step": 1615
+    },
+    {
+      "epoch": 0.9630512514898689,
+      "grad_norm": 2.2813541889190674,
+      "learning_rate": 8.224076281287248e-07,
+      "loss": 0.0296,
+      "step": 1616
+    },
+    {
+      "epoch": 0.9636471990464839,
+      "grad_norm": 1.6292246580123901,
+      "learning_rate": 8.104886769964245e-07,
+      "loss": 0.0341,
+      "step": 1617
+    },
+    {
+      "epoch": 0.964243146603099,
+      "grad_norm": 7.052913665771484,
+      "learning_rate": 7.98569725864124e-07,
+      "loss": 0.4176,
+      "step": 1618
+    },
+    {
+      "epoch": 0.9648390941597139,
+      "grad_norm": 2.7143383026123047,
+      "learning_rate": 7.866507747318236e-07,
+      "loss": 0.066,
+      "step": 1619
+    },
+    {
+      "epoch": 0.965435041716329,
+      "grad_norm": 5.9636759757995605,
+      "learning_rate": 7.747318235995232e-07,
+      "loss": 0.1947,
+      "step": 1620
+    },
+    {
+      "epoch": 0.966030989272944,
+      "grad_norm": 2.5450968742370605,
+      "learning_rate": 7.628128724672229e-07,
+      "loss": 0.0292,
+      "step": 1621
+    },
+    {
+      "epoch": 0.966626936829559,
+      "grad_norm": 2.512603759765625,
+      "learning_rate": 7.508939213349226e-07,
+      "loss": 0.0448,
+      "step": 1622
+    },
+    {
+      "epoch": 0.967222884386174,
+      "grad_norm": 3.5826632976531982,
+      "learning_rate": 7.389749702026222e-07,
+      "loss": 0.1174,
+      "step": 1623
+    },
+    {
+      "epoch": 0.9678188319427891,
+      "grad_norm": 2.238163948059082,
+      "learning_rate": 7.270560190703219e-07,
+      "loss": 0.0392,
+      "step": 1624
+    },
+    {
+      "epoch": 0.968414779499404,
+      "grad_norm": 1.9223339557647705,
+      "learning_rate": 7.151370679380215e-07,
+      "loss": 0.0318,
+      "step": 1625
+    },
+    {
+      "epoch": 0.9690107270560191,
+      "grad_norm": 4.5131754875183105,
+      "learning_rate": 7.032181168057212e-07,
+      "loss": 0.0646,
+      "step": 1626
+    },
+    {
+      "epoch": 0.9696066746126341,
+      "grad_norm": 3.0408236980438232,
+      "learning_rate": 6.912991656734208e-07,
+      "loss": 0.0818,
+      "step": 1627
+    },
+    {
+      "epoch": 0.9702026221692491,
+      "grad_norm": 3.072122097015381,
+      "learning_rate": 6.793802145411205e-07,
+      "loss": 0.1289,
+      "step": 1628
+    },
+    {
+      "epoch": 0.9707985697258641,
+      "grad_norm": 2.496825933456421,
+      "learning_rate": 6.674612634088201e-07,
+      "loss": 0.042,
+      "step": 1629
+    },
+    {
+      "epoch": 0.9713945172824792,
+      "grad_norm": 9.735997200012207,
+      "learning_rate": 6.555423122765198e-07,
+      "loss": 0.9161,
+      "step": 1630
+    },
+    {
+      "epoch": 0.9719904648390941,
+      "grad_norm": 2.282886505126953,
+      "learning_rate": 6.436233611442194e-07,
+      "loss": 0.034,
+      "step": 1631
+    },
+    {
+      "epoch": 0.9725864123957092,
+      "grad_norm": 4.7012224197387695,
+      "learning_rate": 6.317044100119191e-07,
+      "loss": 0.1152,
+      "step": 1632
+    },
+    {
+      "epoch": 0.9731823599523242,
+      "grad_norm": 3.7380337715148926,
+      "learning_rate": 6.197854588796187e-07,
+      "loss": 0.1087,
+      "step": 1633
+    },
+    {
+      "epoch": 0.9737783075089392,
+      "grad_norm": 1.624183177947998,
+      "learning_rate": 6.078665077473183e-07,
+      "loss": 0.019,
+      "step": 1634
+    },
+    {
+      "epoch": 0.9743742550655542,
+      "grad_norm": 4.622048377990723,
+      "learning_rate": 5.959475566150179e-07,
+      "loss": 0.2335,
+      "step": 1635
+    },
+    {
+      "epoch": 0.9749702026221693,
+      "grad_norm": 9.421975135803223,
+      "learning_rate": 5.840286054827176e-07,
+      "loss": 0.8621,
+      "step": 1636
+    },
+    {
+      "epoch": 0.9755661501787842,
+      "grad_norm": 5.438080787658691,
+      "learning_rate": 5.721096543504172e-07,
+      "loss": 0.2537,
+      "step": 1637
+    },
+    {
+      "epoch": 0.9761620977353993,
+      "grad_norm": 0.9332717061042786,
+      "learning_rate": 5.601907032181168e-07,
+      "loss": 0.016,
+      "step": 1638
+    },
+    {
+      "epoch": 0.9767580452920143,
+      "grad_norm": 1.7406736612319946,
+      "learning_rate": 5.482717520858165e-07,
+      "loss": 0.0303,
+      "step": 1639
+    },
+    {
+      "epoch": 0.9773539928486293,
+      "grad_norm": 3.5902836322784424,
+      "learning_rate": 5.363528009535162e-07,
+      "loss": 0.1278,
+      "step": 1640
+    },
+    {
+      "epoch": 0.9779499404052443,
+      "grad_norm": 6.565382480621338,
+      "learning_rate": 5.244338498212158e-07,
+      "loss": 0.4488,
+      "step": 1641
+    },
+    {
+      "epoch": 0.9785458879618594,
+      "grad_norm": 2.7775580883026123,
+      "learning_rate": 5.125148986889154e-07,
+      "loss": 0.0626,
+      "step": 1642
+    },
+    {
+      "epoch": 0.9791418355184743,
+      "grad_norm": 1.5376158952713013,
+      "learning_rate": 5.00595947556615e-07,
+      "loss": 0.0313,
+      "step": 1643
+    },
+    {
+      "epoch": 0.9797377830750894,
+      "grad_norm": 7.309384346008301,
+      "learning_rate": 4.886769964243147e-07,
+      "loss": 0.4811,
+      "step": 1644
+    },
+    {
+      "epoch": 0.9803337306317044,
+      "grad_norm": 4.446257591247559,
+      "learning_rate": 4.7675804529201433e-07,
+      "loss": 0.1741,
+      "step": 1645
+    },
+    {
+      "epoch": 0.9809296781883194,
+      "grad_norm": 9.520309448242188,
+      "learning_rate": 4.6483909415971396e-07,
+      "loss": 0.7938,
+      "step": 1646
+    },
+    {
+      "epoch": 0.9815256257449344,
+      "grad_norm": 2.090595245361328,
+      "learning_rate": 4.5292014302741364e-07,
+      "loss": 0.0389,
+      "step": 1647
+    },
+    {
+      "epoch": 0.9821215733015495,
+      "grad_norm": 3.8076796531677246,
+      "learning_rate": 4.4100119189511326e-07,
+      "loss": 0.2072,
+      "step": 1648
+    },
+    {
+      "epoch": 0.9827175208581644,
+      "grad_norm": 3.1650187969207764,
+      "learning_rate": 4.2908224076281294e-07,
+      "loss": 0.0775,
+      "step": 1649
+    },
+    {
+      "epoch": 0.9833134684147795,
+      "grad_norm": 2.303884983062744,
+      "learning_rate": 4.1716328963051256e-07,
+      "loss": 0.0183,
+      "step": 1650
+    },
+    {
+      "epoch": 0.9839094159713945,
+      "grad_norm": 2.7000458240509033,
+      "learning_rate": 4.0524433849821224e-07,
+      "loss": 0.0403,
+      "step": 1651
+    },
+    {
+      "epoch": 0.9845053635280095,
+      "grad_norm": 4.3197855949401855,
+      "learning_rate": 3.933253873659118e-07,
+      "loss": 0.219,
+      "step": 1652
+    },
+    {
+      "epoch": 0.9851013110846245,
+      "grad_norm": 2.888159990310669,
+      "learning_rate": 3.8140643623361144e-07,
+      "loss": 0.0603,
+      "step": 1653
+    },
+    {
+      "epoch": 0.9856972586412396,
+      "grad_norm": 3.3117964267730713,
+      "learning_rate": 3.694874851013111e-07,
+      "loss": 0.1404,
+      "step": 1654
+    },
+    {
+      "epoch": 0.9862932061978545,
+      "grad_norm": 3.0056722164154053,
+      "learning_rate": 3.5756853396901074e-07,
+      "loss": 0.0827,
+      "step": 1655
+    },
+    {
+      "epoch": 0.9868891537544696,
+      "grad_norm": 1.5030121803283691,
+      "learning_rate": 3.456495828367104e-07,
+      "loss": 0.0377,
+      "step": 1656
+    },
+    {
+      "epoch": 0.9874851013110846,
+      "grad_norm": 1.8468106985092163,
+      "learning_rate": 3.3373063170441004e-07,
+      "loss": 0.0224,
+      "step": 1657
+    },
+    {
+      "epoch": 0.9880810488676997,
+      "grad_norm": 4.9262800216674805,
+      "learning_rate": 3.218116805721097e-07,
+      "loss": 0.2281,
+      "step": 1658
+    },
+    {
+      "epoch": 0.9886769964243146,
+      "grad_norm": 2.864992141723633,
+      "learning_rate": 3.0989272943980934e-07,
+      "loss": 0.0293,
+      "step": 1659
+    },
+    {
+      "epoch": 0.9892729439809297,
+      "grad_norm": 4.808154106140137,
+      "learning_rate": 2.9797377830750897e-07,
+      "loss": 0.1918,
+      "step": 1660
+    },
+    {
+      "epoch": 0.9898688915375446,
+      "grad_norm": 0.6254331469535828,
+      "learning_rate": 2.860548271752086e-07,
+      "loss": 0.01,
+      "step": 1661
+    },
+    {
+      "epoch": 0.9904648390941597,
+      "grad_norm": 4.171772480010986,
+      "learning_rate": 2.7413587604290827e-07,
+      "loss": 0.1316,
+      "step": 1662
+    },
+    {
+      "epoch": 0.9910607866507747,
+      "grad_norm": 3.624047040939331,
+      "learning_rate": 2.622169249106079e-07,
+      "loss": 0.0659,
+      "step": 1663
+    },
+    {
+      "epoch": 0.9916567342073898,
+      "grad_norm": 4.160861492156982,
+      "learning_rate": 2.502979737783075e-07,
+      "loss": 0.2409,
+      "step": 1664
+    },
+    {
+      "epoch": 0.9922526817640047,
+      "grad_norm": 2.50158429145813,
+      "learning_rate": 2.3837902264600717e-07,
+      "loss": 0.0379,
+      "step": 1665
+    },
+    {
+      "epoch": 0.9928486293206198,
+      "grad_norm": 1.942334532737732,
+      "learning_rate": 2.2646007151370682e-07,
+      "loss": 0.0273,
+      "step": 1666
+    },
+    {
+      "epoch": 0.9934445768772348,
+      "grad_norm": 3.129732847213745,
+      "learning_rate": 2.1454112038140647e-07,
+      "loss": 0.0481,
+      "step": 1667
+    },
+    {
+      "epoch": 0.9940405244338498,
+      "grad_norm": 2.985731363296509,
+      "learning_rate": 2.0262216924910612e-07,
+      "loss": 0.1185,
+      "step": 1668
+    },
+    {
+      "epoch": 0.9946364719904648,
+      "grad_norm": 1.9783278703689575,
+      "learning_rate": 1.9070321811680572e-07,
+      "loss": 0.027,
+      "step": 1669
+    },
+    {
+      "epoch": 0.9952324195470799,
+      "grad_norm": 3.403944253921509,
+      "learning_rate": 1.7878426698450537e-07,
+      "loss": 0.0492,
+      "step": 1670
+    },
+    {
+      "epoch": 0.9958283671036948,
+      "grad_norm": 3.436485528945923,
+      "learning_rate": 1.6686531585220502e-07,
+      "loss": 0.1356,
+      "step": 1671
+    },
+    {
+      "epoch": 0.9964243146603099,
+      "grad_norm": 3.2049596309661865,
+      "learning_rate": 1.5494636471990467e-07,
+      "loss": 0.0878,
+      "step": 1672
+    },
+    {
+      "epoch": 0.9970202622169249,
+      "grad_norm": 5.231047630310059,
+      "learning_rate": 1.430274135876043e-07,
+      "loss": 0.2155,
+      "step": 1673
+    },
+    {
+      "epoch": 0.9976162097735399,
+      "grad_norm": 4.0553669929504395,
+      "learning_rate": 1.3110846245530395e-07,
+      "loss": 0.2573,
+      "step": 1674
+    },
+    {
+      "epoch": 0.9982121573301549,
+      "grad_norm": 4.210700988769531,
+      "learning_rate": 1.1918951132300358e-07,
+      "loss": 0.1692,
+      "step": 1675
+    },
+    {
+      "epoch": 0.99880810488677,
+      "grad_norm": 4.255580902099609,
+      "learning_rate": 1.0727056019070323e-07,
+      "loss": 0.3026,
+      "step": 1676
+    },
+    {
+      "epoch": 0.9994040524433849,
+      "grad_norm": 2.1463119983673096,
+      "learning_rate": 9.535160905840286e-08,
+      "loss": 0.0319,
+      "step": 1677
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.412320137023926,
+      "learning_rate": 8.343265792610251e-08,
+      "loss": 0.2996,
+      "step": 1678
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1678,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 296127271403520.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}