diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,36409 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998737411693973,
+  "eval_steps": 500,
+  "global_step": 5197,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00019239440853750187,
+      "grad_norm": 2.7525458471593183,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 2.0372,
+      "step": 1
+    },
+    {
+      "epoch": 0.00038478881707500374,
+      "grad_norm": 3.3321528814620067,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 2.1372,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005771832256125057,
+      "grad_norm": 3.944387374020291,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 2.1234,
+      "step": 3
+    },
+    {
+      "epoch": 0.0007695776341500075,
+      "grad_norm": 3.229562266204187,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 2.0367,
+      "step": 4
+    },
+    {
+      "epoch": 0.0009619720426875094,
+      "grad_norm": 2.8909143591476454,
+      "learning_rate": 6.41025641025641e-06,
+      "loss": 1.9597,
+      "step": 5
+    },
+    {
+      "epoch": 0.0011543664512250113,
+      "grad_norm": 3.410435117906883,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 1.9149,
+      "step": 6
+    },
+    {
+      "epoch": 0.0013467608597625132,
+      "grad_norm": 2.300554810079252,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 1.8909,
+      "step": 7
+    },
+    {
+      "epoch": 0.001539155268300015,
+      "grad_norm": 3.090665861126571,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 2.0468,
+      "step": 8
+    },
+    {
+      "epoch": 0.001731549676837517,
+      "grad_norm": 2.0948227536865542,
+      "learning_rate": 1.153846153846154e-05,
+      "loss": 1.8387,
+      "step": 9
+    },
+    {
+      "epoch": 0.0019239440853750188,
+      "grad_norm": 2.057968488640536,
+      "learning_rate": 1.282051282051282e-05,
+      "loss": 1.8397,
+      "step": 10
+    },
+    {
+      "epoch": 0.002116338493912521,
+      "grad_norm": 1.5223082417883718,
+      "learning_rate": 1.4102564102564104e-05,
+      "loss": 1.8193,
+      "step": 11
+    },
+    {
+      "epoch": 0.0023087329024500227,
+      "grad_norm": 1.1935463000877207,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 1.8624,
+      "step": 12
+    },
+    {
+      "epoch": 0.0025011273109875245,
+      "grad_norm": 1.1665543351628567,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 1.7492,
+      "step": 13
+    },
+    {
+      "epoch": 0.0026935217195250263,
+      "grad_norm": 1.2934871817264082,
+      "learning_rate": 1.794871794871795e-05,
+      "loss": 1.8863,
+      "step": 14
+    },
+    {
+      "epoch": 0.002885916128062528,
+      "grad_norm": 0.9235693583370461,
+      "learning_rate": 1.923076923076923e-05,
+      "loss": 1.7406,
+      "step": 15
+    },
+    {
+      "epoch": 0.00307831053660003,
+      "grad_norm": 0.971049821770051,
+      "learning_rate": 2.0512820512820512e-05,
+      "loss": 1.7381,
+      "step": 16
+    },
+    {
+      "epoch": 0.0032707049451375318,
+      "grad_norm": 0.8293228609690957,
+      "learning_rate": 2.1794871794871795e-05,
+      "loss": 1.7429,
+      "step": 17
+    },
+    {
+      "epoch": 0.003463099353675034,
+      "grad_norm": 0.9500386578091011,
+      "learning_rate": 2.307692307692308e-05,
+      "loss": 1.6875,
+      "step": 18
+    },
+    {
+      "epoch": 0.003655493762212536,
+      "grad_norm": 0.6974359236292831,
+      "learning_rate": 2.435897435897436e-05,
+      "loss": 1.6986,
+      "step": 19
+    },
+    {
+      "epoch": 0.0038478881707500376,
+      "grad_norm": 0.909510623437065,
+      "learning_rate": 2.564102564102564e-05,
+      "loss": 1.717,
+      "step": 20
+    },
+    {
+      "epoch": 0.00404028257928754,
+      "grad_norm": 0.7378054455865029,
+      "learning_rate": 2.6923076923076923e-05,
+      "loss": 1.6046,
+      "step": 21
+    },
+    {
+      "epoch": 0.004232676987825042,
+      "grad_norm": 0.6573768657794542,
+      "learning_rate": 2.8205128205128207e-05,
+      "loss": 1.5586,
+      "step": 22
+    },
+    {
+      "epoch": 0.0044250713963625435,
+      "grad_norm": 0.591424063753591,
+      "learning_rate": 2.948717948717949e-05,
+      "loss": 1.6279,
+      "step": 23
+    },
+    {
+      "epoch": 0.004617465804900045,
+      "grad_norm": 0.6329184676844749,
+      "learning_rate": 3.0769230769230774e-05,
+      "loss": 1.4297,
+      "step": 24
+    },
+    {
+      "epoch": 0.004809860213437547,
+      "grad_norm": 0.6101505751917261,
+      "learning_rate": 3.205128205128206e-05,
+      "loss": 1.5585,
+      "step": 25
+    },
+    {
+      "epoch": 0.005002254621975049,
+      "grad_norm": 0.5587381379284022,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 1.4957,
+      "step": 26
+    },
+    {
+      "epoch": 0.005194649030512551,
+      "grad_norm": 0.5858283225543386,
+      "learning_rate": 3.461538461538462e-05,
+      "loss": 1.523,
+      "step": 27
+    },
+    {
+      "epoch": 0.005387043439050053,
+      "grad_norm": 0.5305749427400327,
+      "learning_rate": 3.58974358974359e-05,
+      "loss": 1.5648,
+      "step": 28
+    },
+    {
+      "epoch": 0.005579437847587554,
+      "grad_norm": 0.49831555031417485,
+      "learning_rate": 3.717948717948718e-05,
+      "loss": 1.6035,
+      "step": 29
+    },
+    {
+      "epoch": 0.005771832256125056,
+      "grad_norm": 0.49498666201043756,
+      "learning_rate": 3.846153846153846e-05,
+      "loss": 1.4841,
+      "step": 30
+    },
+    {
+      "epoch": 0.005964226664662558,
+      "grad_norm": 0.5306126185222823,
+      "learning_rate": 3.974358974358974e-05,
+      "loss": 1.5928,
+      "step": 31
+    },
+    {
+      "epoch": 0.00615662107320006,
+      "grad_norm": 0.4337095829200314,
+      "learning_rate": 4.1025641025641023e-05,
+      "loss": 1.5603,
+      "step": 32
+    },
+    {
+      "epoch": 0.006349015481737562,
+      "grad_norm": 0.43514674800870196,
+      "learning_rate": 4.230769230769231e-05,
+      "loss": 1.6523,
+      "step": 33
+    },
+    {
+      "epoch": 0.0065414098902750635,
+      "grad_norm": 0.5963298979323705,
+      "learning_rate": 4.358974358974359e-05,
+      "loss": 1.4926,
+      "step": 34
+    },
+    {
+      "epoch": 0.006733804298812566,
+      "grad_norm": 0.44531715285780604,
+      "learning_rate": 4.4871794871794874e-05,
+      "loss": 1.5325,
+      "step": 35
+    },
+    {
+      "epoch": 0.006926198707350068,
+      "grad_norm": 0.3473040471350792,
+      "learning_rate": 4.615384615384616e-05,
+      "loss": 1.4282,
+      "step": 36
+    },
+    {
+      "epoch": 0.00711859311588757,
+      "grad_norm": 0.443491003274745,
+      "learning_rate": 4.7435897435897435e-05,
+      "loss": 1.4698,
+      "step": 37
+    },
+    {
+      "epoch": 0.007310987524425072,
+      "grad_norm": 0.44814406853655325,
+      "learning_rate": 4.871794871794872e-05,
+      "loss": 1.528,
+      "step": 38
+    },
+    {
+      "epoch": 0.0075033819329625735,
+      "grad_norm": 0.5391192848549605,
+      "learning_rate": 5e-05,
+      "loss": 1.5508,
+      "step": 39
+    },
+    {
+      "epoch": 0.007695776341500075,
+      "grad_norm": 0.54926588050491,
+      "learning_rate": 5.128205128205128e-05,
+      "loss": 1.4781,
+      "step": 40
+    },
+    {
+      "epoch": 0.007888170750037577,
+      "grad_norm": 0.46443543207168564,
+      "learning_rate": 5.256410256410257e-05,
+      "loss": 1.4432,
+      "step": 41
+    },
+    {
+      "epoch": 0.00808056515857508,
+      "grad_norm": 0.44321460008702634,
+      "learning_rate": 5.384615384615385e-05,
+      "loss": 1.4928,
+      "step": 42
+    },
+    {
+      "epoch": 0.00827295956711258,
+      "grad_norm": 0.4632820875245947,
+      "learning_rate": 5.512820512820514e-05,
+      "loss": 1.4281,
+      "step": 43
+    },
+    {
+      "epoch": 0.008465353975650083,
+      "grad_norm": 0.42181714931408915,
+      "learning_rate": 5.6410256410256414e-05,
+      "loss": 1.4614,
+      "step": 44
+    },
+    {
+      "epoch": 0.008657748384187584,
+      "grad_norm": 0.45202808445955417,
+      "learning_rate": 5.769230769230769e-05,
+      "loss": 1.512,
+      "step": 45
+    },
+    {
+      "epoch": 0.008850142792725087,
+      "grad_norm": 0.39129615556085906,
+      "learning_rate": 5.897435897435898e-05,
+      "loss": 1.3329,
+      "step": 46
+    },
+    {
+      "epoch": 0.009042537201262588,
+      "grad_norm": 0.44025007115595804,
+      "learning_rate": 6.025641025641026e-05,
+      "loss": 1.4993,
+      "step": 47
+    },
+    {
+      "epoch": 0.00923493160980009,
+      "grad_norm": 0.34779913994346245,
+      "learning_rate": 6.153846153846155e-05,
+      "loss": 1.411,
+      "step": 48
+    },
+    {
+      "epoch": 0.009427326018337592,
+      "grad_norm": 0.3585120048114127,
+      "learning_rate": 6.282051282051282e-05,
+      "loss": 1.4531,
+      "step": 49
+    },
+    {
+      "epoch": 0.009619720426875094,
+      "grad_norm": 0.3955562532849728,
+      "learning_rate": 6.410256410256412e-05,
+      "loss": 1.4165,
+      "step": 50
+    },
+    {
+      "epoch": 0.009812114835412595,
+      "grad_norm": 0.39043156648418725,
+      "learning_rate": 6.538461538461539e-05,
+      "loss": 1.442,
+      "step": 51
+    },
+    {
+      "epoch": 0.010004509243950098,
+      "grad_norm": 0.33706432808884385,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 1.4472,
+      "step": 52
+    },
+    {
+      "epoch": 0.010196903652487599,
+      "grad_norm": 0.3742128066725855,
+      "learning_rate": 6.794871794871795e-05,
+      "loss": 1.3246,
+      "step": 53
+    },
+    {
+      "epoch": 0.010389298061025102,
+      "grad_norm": 0.3929624766913999,
+      "learning_rate": 6.923076923076924e-05,
+      "loss": 1.3529,
+      "step": 54
+    },
+    {
+      "epoch": 0.010581692469562603,
+      "grad_norm": 0.5610802334436438,
+      "learning_rate": 7.051282051282052e-05,
+      "loss": 1.5152,
+      "step": 55
+    },
+    {
+      "epoch": 0.010774086878100105,
+      "grad_norm": 0.5250016533473014,
+      "learning_rate": 7.17948717948718e-05,
+      "loss": 1.3199,
+      "step": 56
+    },
+    {
+      "epoch": 0.010966481286637608,
+      "grad_norm": 0.42322353397161805,
+      "learning_rate": 7.307692307692307e-05,
+      "loss": 1.4614,
+      "step": 57
+    },
+    {
+      "epoch": 0.011158875695175109,
+      "grad_norm": 0.4845674596386445,
+      "learning_rate": 7.435897435897436e-05,
+      "loss": 1.4659,
+      "step": 58
+    },
+    {
+      "epoch": 0.011351270103712612,
+      "grad_norm": 0.34749048932750803,
+      "learning_rate": 7.564102564102564e-05,
+      "loss": 1.3888,
+      "step": 59
+    },
+    {
+      "epoch": 0.011543664512250112,
+      "grad_norm": 0.4017986834980663,
+      "learning_rate": 7.692307692307693e-05,
+      "loss": 1.4199,
+      "step": 60
+    },
+    {
+      "epoch": 0.011736058920787615,
+      "grad_norm": 0.5027208677679172,
+      "learning_rate": 7.820512820512821e-05,
+      "loss": 1.3256,
+      "step": 61
+    },
+    {
+      "epoch": 0.011928453329325116,
+      "grad_norm": 0.5527999152289834,
+      "learning_rate": 7.948717948717948e-05,
+      "loss": 1.3996,
+      "step": 62
+    },
+    {
+      "epoch": 0.012120847737862619,
+      "grad_norm": 0.4462731246263686,
+      "learning_rate": 8.076923076923078e-05,
+      "loss": 1.3843,
+      "step": 63
+    },
+    {
+      "epoch": 0.01231324214640012,
+      "grad_norm": 0.359874528713232,
+      "learning_rate": 8.205128205128205e-05,
+      "loss": 1.2931,
+      "step": 64
+    },
+    {
+      "epoch": 0.012505636554937622,
+      "grad_norm": 0.38360876148067313,
+      "learning_rate": 8.333333333333334e-05,
+      "loss": 1.3122,
+      "step": 65
+    },
+    {
+      "epoch": 0.012698030963475123,
+      "grad_norm": 0.4749001100500843,
+      "learning_rate": 8.461538461538461e-05,
+      "loss": 1.3386,
+      "step": 66
+    },
+    {
+      "epoch": 0.012890425372012626,
+      "grad_norm": 0.568153688337833,
+      "learning_rate": 8.58974358974359e-05,
+      "loss": 1.3523,
+      "step": 67
+    },
+    {
+      "epoch": 0.013082819780550127,
+      "grad_norm": 0.3376201695261382,
+      "learning_rate": 8.717948717948718e-05,
+      "loss": 1.3455,
+      "step": 68
+    },
+    {
+      "epoch": 0.01327521418908763,
+      "grad_norm": 0.41665092992914415,
+      "learning_rate": 8.846153846153847e-05,
+      "loss": 1.3805,
+      "step": 69
+    },
+    {
+      "epoch": 0.013467608597625132,
+      "grad_norm": 0.3775378273997053,
+      "learning_rate": 8.974358974358975e-05,
+      "loss": 1.282,
+      "step": 70
+    },
+    {
+      "epoch": 0.013660003006162633,
+      "grad_norm": 0.4836285793376228,
+      "learning_rate": 9.102564102564103e-05,
+      "loss": 1.3999,
+      "step": 71
+    },
+    {
+      "epoch": 0.013852397414700136,
+      "grad_norm": 0.484212509769501,
+      "learning_rate": 9.230769230769232e-05,
+      "loss": 1.3214,
+      "step": 72
+    },
+    {
+      "epoch": 0.014044791823237637,
+      "grad_norm": 0.5377956432211242,
+      "learning_rate": 9.35897435897436e-05,
+      "loss": 1.3792,
+      "step": 73
+    },
+    {
+      "epoch": 0.01423718623177514,
+      "grad_norm": 0.45721867790444926,
+      "learning_rate": 9.487179487179487e-05,
+      "loss": 1.263,
+      "step": 74
+    },
+    {
+      "epoch": 0.01442958064031264,
+      "grad_norm": 0.4104047180613727,
+      "learning_rate": 9.615384615384617e-05,
+      "loss": 1.3201,
+      "step": 75
+    },
+    {
+      "epoch": 0.014621975048850143,
+      "grad_norm": 0.5907839436204101,
+      "learning_rate": 9.743589743589744e-05,
+      "loss": 1.4332,
+      "step": 76
+    },
+    {
+      "epoch": 0.014814369457387644,
+      "grad_norm": 0.4254853829097721,
+      "learning_rate": 9.871794871794872e-05,
+      "loss": 1.3182,
+      "step": 77
+    },
+    {
+      "epoch": 0.015006763865925147,
+      "grad_norm": 0.5329864075221257,
+      "learning_rate": 0.0001,
+      "loss": 1.3,
+      "step": 78
+    },
+    {
+      "epoch": 0.015199158274462648,
+      "grad_norm": 0.4324036261474306,
+      "learning_rate": 0.00010128205128205129,
+      "loss": 1.3243,
+      "step": 79
+    },
+    {
+      "epoch": 0.01539155268300015,
+      "grad_norm": 0.4290409758966646,
+      "learning_rate": 0.00010256410256410256,
+      "loss": 1.3799,
+      "step": 80
+    },
+    {
+      "epoch": 0.015583947091537651,
+      "grad_norm": 0.3391541691436233,
+      "learning_rate": 0.00010384615384615386,
+      "loss": 1.3717,
+      "step": 81
+    },
+    {
+      "epoch": 0.015776341500075154,
+      "grad_norm": 0.37373787707783,
+      "learning_rate": 0.00010512820512820514,
+      "loss": 1.3731,
+      "step": 82
+    },
+    {
+      "epoch": 0.015968735908612655,
+      "grad_norm": 0.31215329512195983,
+      "learning_rate": 0.00010641025641025641,
+      "loss": 1.3579,
+      "step": 83
+    },
+    {
+      "epoch": 0.01616113031715016,
+      "grad_norm": 0.4288427945434432,
+      "learning_rate": 0.0001076923076923077,
+      "loss": 1.3237,
+      "step": 84
+    },
+    {
+      "epoch": 0.01635352472568766,
+      "grad_norm": 0.5160078251857673,
+      "learning_rate": 0.00010897435897435896,
+      "loss": 1.2972,
+      "step": 85
+    },
+    {
+      "epoch": 0.01654591913422516,
+      "grad_norm": 0.3993394373534561,
+      "learning_rate": 0.00011025641025641027,
+      "loss": 1.3384,
+      "step": 86
+    },
+    {
+      "epoch": 0.016738313542762662,
+      "grad_norm": 0.4366076718871945,
+      "learning_rate": 0.00011153846153846154,
+      "loss": 1.3148,
+      "step": 87
+    },
+    {
+      "epoch": 0.016930707951300167,
+      "grad_norm": 0.4733218205937479,
+      "learning_rate": 0.00011282051282051283,
+      "loss": 1.2973,
+      "step": 88
+    },
+    {
+      "epoch": 0.017123102359837668,
+      "grad_norm": 0.4087246684407547,
+      "learning_rate": 0.0001141025641025641,
+      "loss": 1.3423,
+      "step": 89
+    },
+    {
+      "epoch": 0.01731549676837517,
+      "grad_norm": 0.3992625790555533,
+      "learning_rate": 0.00011538461538461538,
+      "loss": 1.3228,
+      "step": 90
+    },
+    {
+      "epoch": 0.01750789117691267,
+      "grad_norm": 0.5150054008924904,
+      "learning_rate": 0.00011666666666666668,
+      "loss": 1.2881,
+      "step": 91
+    },
+    {
+      "epoch": 0.017700285585450174,
+      "grad_norm": 0.5876754273805255,
+      "learning_rate": 0.00011794871794871796,
+      "loss": 1.3594,
+      "step": 92
+    },
+    {
+      "epoch": 0.017892679993987675,
+      "grad_norm": 0.42631406042289577,
+      "learning_rate": 0.00011923076923076923,
+      "loss": 1.4834,
+      "step": 93
+    },
+    {
+      "epoch": 0.018085074402525176,
+      "grad_norm": 0.3838034281944992,
+      "learning_rate": 0.00012051282051282052,
+      "loss": 1.2706,
+      "step": 94
+    },
+    {
+      "epoch": 0.018277468811062677,
+      "grad_norm": 0.4301521685881343,
+      "learning_rate": 0.00012179487179487179,
+      "loss": 1.2317,
+      "step": 95
+    },
+    {
+      "epoch": 0.01846986321960018,
+      "grad_norm": 0.3865391962403398,
+      "learning_rate": 0.0001230769230769231,
+      "loss": 1.2702,
+      "step": 96
+    },
+    {
+      "epoch": 0.018662257628137682,
+      "grad_norm": 0.46676939700542575,
+      "learning_rate": 0.00012435897435897437,
+      "loss": 1.2659,
+      "step": 97
+    },
+    {
+      "epoch": 0.018854652036675183,
+      "grad_norm": 0.3282694769605565,
+      "learning_rate": 0.00012564102564102564,
+      "loss": 1.3712,
+      "step": 98
+    },
+    {
+      "epoch": 0.019047046445212688,
+      "grad_norm": 0.41323205438808747,
+      "learning_rate": 0.00012692307692307693,
+      "loss": 1.3105,
+      "step": 99
+    },
+    {
+      "epoch": 0.01923944085375019,
+      "grad_norm": 0.39942104467779616,
+      "learning_rate": 0.00012820512820512823,
+      "loss": 1.3306,
+      "step": 100
+    },
+    {
+      "epoch": 0.01943183526228769,
+      "grad_norm": 0.39527401905640397,
+      "learning_rate": 0.0001294871794871795,
+      "loss": 1.3257,
+      "step": 101
+    },
+    {
+      "epoch": 0.01962422967082519,
+      "grad_norm": 0.45162116871270536,
+      "learning_rate": 0.00013076923076923077,
+      "loss": 1.3487,
+      "step": 102
+    },
+    {
+      "epoch": 0.019816624079362695,
+      "grad_norm": 0.45123939487831205,
+      "learning_rate": 0.00013205128205128204,
+      "loss": 1.3389,
+      "step": 103
+    },
+    {
+      "epoch": 0.020009018487900196,
+      "grad_norm": 0.3141520566672659,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 1.2622,
+      "step": 104
+    },
+    {
+      "epoch": 0.020201412896437697,
+      "grad_norm": 0.4351649493507871,
+      "learning_rate": 0.00013461538461538464,
+      "loss": 1.2419,
+      "step": 105
+    },
+    {
+      "epoch": 0.020393807304975198,
+      "grad_norm": 0.39794413282752655,
+      "learning_rate": 0.0001358974358974359,
+      "loss": 1.2614,
+      "step": 106
+    },
+    {
+      "epoch": 0.020586201713512702,
+      "grad_norm": 0.5243932516778702,
+      "learning_rate": 0.00013717948717948718,
+      "loss": 1.2405,
+      "step": 107
+    },
+    {
+      "epoch": 0.020778596122050203,
+      "grad_norm": 0.3886407628185571,
+      "learning_rate": 0.00013846153846153847,
+      "loss": 1.3281,
+      "step": 108
+    },
+    {
+      "epoch": 0.020970990530587704,
+      "grad_norm": 0.42235960401541284,
+      "learning_rate": 0.00013974358974358974,
+      "loss": 1.3742,
+      "step": 109
+    },
+    {
+      "epoch": 0.021163384939125205,
+      "grad_norm": 0.38868927854775726,
+      "learning_rate": 0.00014102564102564104,
+      "loss": 1.185,
+      "step": 110
+    },
+    {
+      "epoch": 0.02135577934766271,
+      "grad_norm": 0.4545771910029529,
+      "learning_rate": 0.0001423076923076923,
+      "loss": 1.2502,
+      "step": 111
+    },
+    {
+      "epoch": 0.02154817375620021,
+      "grad_norm": 0.42840503963264065,
+      "learning_rate": 0.0001435897435897436,
+      "loss": 1.3541,
+      "step": 112
+    },
+    {
+      "epoch": 0.02174056816473771,
+      "grad_norm": 0.5606903363021304,
+      "learning_rate": 0.00014487179487179488,
+      "loss": 1.2254,
+      "step": 113
+    },
+    {
+      "epoch": 0.021932962573275216,
+      "grad_norm": 0.3710188315891035,
+      "learning_rate": 0.00014615384615384615,
+      "loss": 1.213,
+      "step": 114
+    },
+    {
+      "epoch": 0.022125356981812717,
+      "grad_norm": 0.49119061421727545,
+      "learning_rate": 0.00014743589743589745,
+      "loss": 1.2475,
+      "step": 115
+    },
+    {
+      "epoch": 0.022317751390350218,
+      "grad_norm": 0.4781056177708388,
+      "learning_rate": 0.00014871794871794872,
+      "loss": 1.2206,
+      "step": 116
+    },
+    {
+      "epoch": 0.02251014579888772,
+      "grad_norm": 0.42953663682993126,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 1.4172,
+      "step": 117
+    },
+    {
+      "epoch": 0.022702540207425223,
+      "grad_norm": 0.5224096844586757,
+      "learning_rate": 0.00015128205128205128,
+      "loss": 1.3066,
+      "step": 118
+    },
+    {
+      "epoch": 0.022894934615962724,
+      "grad_norm": 0.3840882540830783,
+      "learning_rate": 0.00015256410256410255,
+      "loss": 1.175,
+      "step": 119
+    },
+    {
+      "epoch": 0.023087329024500225,
+      "grad_norm": 0.37664235360882764,
+      "learning_rate": 0.00015384615384615385,
+      "loss": 1.3042,
+      "step": 120
+    },
+    {
+      "epoch": 0.023279723433037726,
+      "grad_norm": 0.44660253558315144,
+      "learning_rate": 0.00015512820512820515,
+      "loss": 1.3376,
+      "step": 121
+    },
+    {
+      "epoch": 0.02347211784157523,
+      "grad_norm": 0.45240697967438576,
+      "learning_rate": 0.00015641025641025642,
+      "loss": 1.3129,
+      "step": 122
+    },
+    {
+      "epoch": 0.02366451225011273,
+      "grad_norm": 0.5488188281205489,
+      "learning_rate": 0.0001576923076923077,
+      "loss": 1.3034,
+      "step": 123
+    },
+    {
+      "epoch": 0.023856906658650232,
+      "grad_norm": 0.4093560819591969,
+      "learning_rate": 0.00015897435897435896,
+      "loss": 1.3536,
+      "step": 124
+    },
+    {
+      "epoch": 0.024049301067187733,
+      "grad_norm": 0.4181518968444485,
+      "learning_rate": 0.00016025641025641028,
+      "loss": 1.2972,
+      "step": 125
+    },
+    {
+      "epoch": 0.024241695475725238,
+      "grad_norm": 0.47822260626271096,
+      "learning_rate": 0.00016153846153846155,
+      "loss": 1.2461,
+      "step": 126
+    },
+    {
+      "epoch": 0.02443408988426274,
+      "grad_norm": 0.3658429455505507,
+      "learning_rate": 0.00016282051282051282,
+      "loss": 1.2409,
+      "step": 127
+    },
+    {
+      "epoch": 0.02462648429280024,
+      "grad_norm": 0.39272623832743203,
+      "learning_rate": 0.0001641025641025641,
+      "loss": 1.2174,
+      "step": 128
+    },
+    {
+      "epoch": 0.024818878701337744,
+      "grad_norm": 0.39762228575064995,
+      "learning_rate": 0.0001653846153846154,
+      "loss": 1.284,
+      "step": 129
+    },
+    {
+      "epoch": 0.025011273109875245,
+      "grad_norm": 0.47852070426660737,
+      "learning_rate": 0.0001666666666666667,
+      "loss": 1.2164,
+      "step": 130
+    },
+    {
+      "epoch": 0.025203667518412746,
+      "grad_norm": 0.37689479811805965,
+      "learning_rate": 0.00016794871794871796,
+      "loss": 1.3237,
+      "step": 131
+    },
+    {
+      "epoch": 0.025396061926950247,
+      "grad_norm": 0.39110709401209226,
+      "learning_rate": 0.00016923076923076923,
+      "loss": 1.2545,
+      "step": 132
+    },
+    {
+      "epoch": 0.02558845633548775,
+      "grad_norm": 0.4310773405987121,
+      "learning_rate": 0.00017051282051282053,
+      "loss": 1.2216,
+      "step": 133
+    },
+    {
+      "epoch": 0.025780850744025252,
+      "grad_norm": 0.3468023433804279,
+      "learning_rate": 0.0001717948717948718,
+      "loss": 1.2678,
+      "step": 134
+    },
+    {
+      "epoch": 0.025973245152562753,
+      "grad_norm": 0.5085016913769101,
+      "learning_rate": 0.0001730769230769231,
+      "loss": 1.2955,
+      "step": 135
+    },
+    {
+      "epoch": 0.026165639561100254,
+      "grad_norm": 0.49287583142064645,
+      "learning_rate": 0.00017435897435897436,
+      "loss": 1.358,
+      "step": 136
+    },
+    {
+      "epoch": 0.02635803396963776,
+      "grad_norm": 0.33228001088489983,
+      "learning_rate": 0.00017564102564102566,
+      "loss": 1.3327,
+      "step": 137
+    },
+    {
+      "epoch": 0.02655042837817526,
+      "grad_norm": 0.3503824504467012,
+      "learning_rate": 0.00017692307692307693,
+      "loss": 1.1518,
+      "step": 138
+    },
+    {
+      "epoch": 0.02674282278671276,
+      "grad_norm": 0.37532956999924594,
+      "learning_rate": 0.00017820512820512823,
+      "loss": 1.2865,
+      "step": 139
+    },
+    {
+      "epoch": 0.026935217195250265,
+      "grad_norm": 0.38406220033381544,
+      "learning_rate": 0.0001794871794871795,
+      "loss": 1.2418,
+      "step": 140
+    },
+    {
+      "epoch": 0.027127611603787766,
+      "grad_norm": 0.39031936695061314,
+      "learning_rate": 0.00018076923076923077,
+      "loss": 1.228,
+      "step": 141
+    },
+    {
+      "epoch": 0.027320006012325267,
+      "grad_norm": 0.34195813274578396,
+      "learning_rate": 0.00018205128205128207,
+      "loss": 1.2056,
+      "step": 142
+    },
+    {
+      "epoch": 0.027512400420862768,
+      "grad_norm": 0.4248888854316749,
+      "learning_rate": 0.00018333333333333334,
+      "loss": 1.3257,
+      "step": 143
+    },
+    {
+      "epoch": 0.027704794829400272,
+      "grad_norm": 0.4210928393092026,
+      "learning_rate": 0.00018461538461538463,
+      "loss": 1.2242,
+      "step": 144
+    },
+    {
+      "epoch": 0.027897189237937773,
+      "grad_norm": 0.4352511878415626,
+      "learning_rate": 0.0001858974358974359,
+      "loss": 1.3441,
+      "step": 145
+    },
+    {
+      "epoch": 0.028089583646475274,
+      "grad_norm": 0.4217866234218403,
+      "learning_rate": 0.0001871794871794872,
+      "loss": 1.2583,
+      "step": 146
+    },
+    {
+      "epoch": 0.028281978055012775,
+      "grad_norm": 0.3974517630815656,
+      "learning_rate": 0.00018846153846153847,
+      "loss": 1.269,
+      "step": 147
+    },
+    {
+      "epoch": 0.02847437246355028,
+      "grad_norm": 0.4216992528085191,
+      "learning_rate": 0.00018974358974358974,
+      "loss": 1.2464,
+      "step": 148
+    },
+    {
+      "epoch": 0.02866676687208778,
+      "grad_norm": 0.33500073758371224,
+      "learning_rate": 0.00019102564102564104,
+      "loss": 1.4107,
+      "step": 149
+    },
+    {
+      "epoch": 0.02885916128062528,
+      "grad_norm": 0.3930034995291226,
+      "learning_rate": 0.00019230769230769233,
+      "loss": 1.2547,
+      "step": 150
+    },
+    {
+      "epoch": 0.029051555689162782,
+      "grad_norm": 0.39243830006177954,
+      "learning_rate": 0.0001935897435897436,
+      "loss": 1.3772,
+      "step": 151
+    },
+    {
+      "epoch": 0.029243950097700287,
+      "grad_norm": 0.43922584612423615,
+      "learning_rate": 0.00019487179487179487,
+      "loss": 1.2391,
+      "step": 152
+    },
+    {
+      "epoch": 0.029436344506237788,
+      "grad_norm": 0.3444854965075257,
+      "learning_rate": 0.00019615384615384615,
+      "loss": 1.3353,
+      "step": 153
+    },
+    {
+      "epoch": 0.02962873891477529,
+      "grad_norm": 0.359517893965557,
+      "learning_rate": 0.00019743589743589744,
+      "loss": 1.2569,
+      "step": 154
+    },
+    {
+      "epoch": 0.029821133323312793,
+      "grad_norm": 0.3968742594316916,
+      "learning_rate": 0.00019871794871794874,
+      "loss": 1.3355,
+      "step": 155
+    },
+    {
+      "epoch": 0.030013527731850294,
+      "grad_norm": 0.34182297551625235,
+      "learning_rate": 0.0002,
+      "loss": 1.2427,
+      "step": 156
+    },
+    {
+      "epoch": 0.030205922140387795,
+      "grad_norm": 0.44423820785007373,
+      "learning_rate": 0.00019999998058057615,
+      "loss": 1.2839,
+      "step": 157
+    },
+    {
+      "epoch": 0.030398316548925296,
+      "grad_norm": 0.42072857308135675,
+      "learning_rate": 0.00019999992232231216,
+      "loss": 1.2678,
+      "step": 158
+    },
+    {
+      "epoch": 0.0305907109574628,
+      "grad_norm": 0.48936094118847634,
+      "learning_rate": 0.0001999998252252306,
+      "loss": 1.2668,
+      "step": 159
+    },
+    {
+      "epoch": 0.0307831053660003,
+      "grad_norm": 0.3442662491489396,
+      "learning_rate": 0.00019999968928936926,
+      "loss": 1.2838,
+      "step": 160
+    },
+    {
+      "epoch": 0.030975499774537802,
+      "grad_norm": 0.3920464401403197,
+      "learning_rate": 0.00019999951451478087,
+      "loss": 1.2205,
+      "step": 161
+    },
+    {
+      "epoch": 0.031167894183075303,
+      "grad_norm": 0.3334924587891562,
+      "learning_rate": 0.00019999930090153334,
+      "loss": 1.3183,
+      "step": 162
+    },
+    {
+      "epoch": 0.03136028859161281,
+      "grad_norm": 0.3582955766263805,
+      "learning_rate": 0.00019999904844970962,
+      "loss": 1.2481,
+      "step": 163
+    },
+    {
+      "epoch": 0.03155268300015031,
+      "grad_norm": 0.34978513788320387,
+      "learning_rate": 0.00019999875715940782,
+      "loss": 1.1672,
+      "step": 164
+    },
+    {
+      "epoch": 0.03174507740868781,
+      "grad_norm": 0.49416109569985456,
+      "learning_rate": 0.000199998427030741,
+      "loss": 1.2332,
+      "step": 165
+    },
+    {
+      "epoch": 0.03193747181722531,
+      "grad_norm": 0.35765291072585936,
+      "learning_rate": 0.00019999805806383738,
+      "loss": 1.2619,
+      "step": 166
+    },
+    {
+      "epoch": 0.03212986622576281,
+      "grad_norm": 0.4847857479606003,
+      "learning_rate": 0.0001999976502588403,
+      "loss": 1.2642,
+      "step": 167
+    },
+    {
+      "epoch": 0.03232226063430032,
+      "grad_norm": 0.48853166022098093,
+      "learning_rate": 0.0001999972036159081,
+      "loss": 1.3542,
+      "step": 168
+    },
+    {
+      "epoch": 0.03251465504283782,
+      "grad_norm": 0.35297868597517823,
+      "learning_rate": 0.00019999671813521435,
+      "loss": 1.2381,
+      "step": 169
+    },
+    {
+      "epoch": 0.03270704945137532,
+      "grad_norm": 0.4769237665603929,
+      "learning_rate": 0.0001999961938169475,
+      "loss": 1.1891,
+      "step": 170
+    },
+    {
+      "epoch": 0.03289944385991282,
+      "grad_norm": 0.36210408771409336,
+      "learning_rate": 0.00019999563066131124,
+      "loss": 1.3756,
+      "step": 171
+    },
+    {
+      "epoch": 0.03309183826845032,
+      "grad_norm": 0.34813424957631334,
+      "learning_rate": 0.00019999502866852425,
+      "loss": 1.293,
+      "step": 172
+    },
+    {
+      "epoch": 0.033284232676987824,
+      "grad_norm": 0.3784075964795684,
+      "learning_rate": 0.0001999943878388204,
+      "loss": 1.3094,
+      "step": 173
+    },
+    {
+      "epoch": 0.033476627085525325,
+      "grad_norm": 0.39718809122566173,
+      "learning_rate": 0.00019999370817244853,
+      "loss": 1.3564,
+      "step": 174
+    },
+    {
+      "epoch": 0.033669021494062826,
+      "grad_norm": 0.45762884116455194,
+      "learning_rate": 0.00019999298966967265,
+      "loss": 1.258,
+      "step": 175
+    },
+    {
+      "epoch": 0.033861415902600334,
+      "grad_norm": 0.381488697387606,
+      "learning_rate": 0.00019999223233077177,
+      "loss": 1.303,
+      "step": 176
+    },
+    {
+      "epoch": 0.034053810311137835,
+      "grad_norm": 0.32182417029429744,
+      "learning_rate": 0.0001999914361560401,
+      "loss": 1.2069,
+      "step": 177
+    },
+    {
+      "epoch": 0.034246204719675336,
+      "grad_norm": 0.33952904696616504,
+      "learning_rate": 0.00019999060114578684,
+      "loss": 1.2738,
+      "step": 178
+    },
+    {
+      "epoch": 0.034438599128212836,
+      "grad_norm": 0.45283663599696256,
+      "learning_rate": 0.00019998972730033622,
+      "loss": 1.2087,
+      "step": 179
+    },
+    {
+      "epoch": 0.03463099353675034,
+      "grad_norm": 0.42777961625466415,
+      "learning_rate": 0.00019998881462002778,
+      "loss": 1.2969,
+      "step": 180
+    },
+    {
+      "epoch": 0.03482338794528784,
+      "grad_norm": 0.4353440055613314,
+      "learning_rate": 0.00019998786310521585,
+      "loss": 1.1811,
+      "step": 181
+    },
+    {
+      "epoch": 0.03501578235382534,
+      "grad_norm": 0.3623049235937626,
+      "learning_rate": 0.00019998687275627006,
+      "loss": 1.2145,
+      "step": 182
+    },
+    {
+      "epoch": 0.03520817676236285,
+      "grad_norm": 0.3633740669851794,
+      "learning_rate": 0.00019998584357357502,
+      "loss": 1.282,
+      "step": 183
+    },
+    {
+      "epoch": 0.03540057117090035,
+      "grad_norm": 0.40672086192782747,
+      "learning_rate": 0.00019998477555753055,
+      "loss": 1.2021,
+      "step": 184
+    },
+    {
+      "epoch": 0.03559296557943785,
+      "grad_norm": 0.3528660946853748,
+      "learning_rate": 0.00019998366870855133,
+      "loss": 1.3171,
+      "step": 185
+    },
+    {
+      "epoch": 0.03578535998797535,
+      "grad_norm": 0.29673419601117756,
+      "learning_rate": 0.0001999825230270673,
+      "loss": 1.2103,
+      "step": 186
+    },
+    {
+      "epoch": 0.03597775439651285,
+      "grad_norm": 0.30415174073622714,
+      "learning_rate": 0.0001999813385135234,
+      "loss": 1.2861,
+      "step": 187
+    },
+    {
+      "epoch": 0.03617014880505035,
+      "grad_norm": 0.3569229070825008,
+      "learning_rate": 0.00019998011516837974,
+      "loss": 1.1805,
+      "step": 188
+    },
+    {
+      "epoch": 0.03636254321358785,
+      "grad_norm": 0.3184620004926606,
+      "learning_rate": 0.0001999788529921114,
+      "loss": 1.3267,
+      "step": 189
+    },
+    {
+      "epoch": 0.036554937622125354,
+      "grad_norm": 0.4710231851905076,
+      "learning_rate": 0.0001999775519852086,
+      "loss": 1.2713,
+      "step": 190
+    },
+    {
+      "epoch": 0.03674733203066286,
+      "grad_norm": 0.31583452059456996,
+      "learning_rate": 0.00019997621214817667,
+      "loss": 1.1783,
+      "step": 191
+    },
+    {
+      "epoch": 0.03693972643920036,
+      "grad_norm": 0.3796671027972158,
+      "learning_rate": 0.000199974833481536,
+      "loss": 1.1852,
+      "step": 192
+    },
+    {
+      "epoch": 0.037132120847737864,
+      "grad_norm": 0.30458775980863195,
+      "learning_rate": 0.00019997341598582195,
+      "loss": 1.3163,
+      "step": 193
+    },
+    {
+      "epoch": 0.037324515256275365,
+      "grad_norm": 0.40876099439003527,
+      "learning_rate": 0.00019997195966158518,
+      "loss": 1.2282,
+      "step": 194
+    },
+    {
+      "epoch": 0.037516909664812866,
+      "grad_norm": 0.2851561930530594,
+      "learning_rate": 0.0001999704645093912,
+      "loss": 1.3051,
+      "step": 195
+    },
+    {
+      "epoch": 0.037709304073350366,
+      "grad_norm": 0.4264152679934513,
+      "learning_rate": 0.0001999689305298208,
+      "loss": 1.3383,
+      "step": 196
+    },
+    {
+      "epoch": 0.03790169848188787,
+      "grad_norm": 0.3968562536557867,
+      "learning_rate": 0.00019996735772346972,
+      "loss": 1.2176,
+      "step": 197
+    },
+    {
+      "epoch": 0.038094092890425375,
+      "grad_norm": 0.3784089645095029,
+      "learning_rate": 0.00019996574609094884,
+      "loss": 1.3513,
+      "step": 198
+    },
+    {
+      "epoch": 0.038286487298962876,
+      "grad_norm": 0.501810530415642,
+      "learning_rate": 0.00019996409563288406,
+      "loss": 1.1565,
+      "step": 199
+    },
+    {
+      "epoch": 0.03847888170750038,
+      "grad_norm": 0.4982537736486703,
+      "learning_rate": 0.00019996240634991642,
+      "loss": 1.1701,
+      "step": 200
+    },
+    {
+      "epoch": 0.03867127611603788,
+      "grad_norm": 0.40509324758121606,
+      "learning_rate": 0.00019996067824270205,
+      "loss": 1.2529,
+      "step": 201
+    },
+    {
+      "epoch": 0.03886367052457538,
+      "grad_norm": 0.424702289607452,
+      "learning_rate": 0.00019995891131191205,
+      "loss": 1.3131,
+      "step": 202
+    },
+    {
+      "epoch": 0.03905606493311288,
+      "grad_norm": 0.41181957289529814,
+      "learning_rate": 0.00019995710555823276,
+      "loss": 1.1718,
+      "step": 203
+    },
+    {
+      "epoch": 0.03924845934165038,
+      "grad_norm": 0.3498527557157088,
+      "learning_rate": 0.00019995526098236547,
+      "loss": 1.2304,
+      "step": 204
+    },
+    {
+      "epoch": 0.03944085375018788,
+      "grad_norm": 0.2965374719080043,
+      "learning_rate": 0.0001999533775850266,
+      "loss": 1.281,
+      "step": 205
+    },
+    {
+      "epoch": 0.03963324815872539,
+      "grad_norm": 0.3680750001537127,
+      "learning_rate": 0.00019995145536694762,
+      "loss": 1.2044,
+      "step": 206
+    },
+    {
+      "epoch": 0.03982564256726289,
+      "grad_norm": 0.42398151925042865,
+      "learning_rate": 0.00019994949432887514,
+      "loss": 1.2261,
+      "step": 207
+    },
+    {
+      "epoch": 0.04001803697580039,
+      "grad_norm": 0.397051080448696,
+      "learning_rate": 0.00019994749447157077,
+      "loss": 1.2953,
+      "step": 208
+    },
+    {
+      "epoch": 0.04021043138433789,
+      "grad_norm": 0.34499463191570895,
+      "learning_rate": 0.00019994545579581123,
+      "loss": 1.2452,
+      "step": 209
+    },
+    {
+      "epoch": 0.040402825792875394,
+      "grad_norm": 0.3798929225789759,
+      "learning_rate": 0.00019994337830238834,
+      "loss": 1.3276,
+      "step": 210
+    },
+    {
+      "epoch": 0.040595220201412895,
+      "grad_norm": 0.3686082097211794,
+      "learning_rate": 0.00019994126199210897,
+      "loss": 1.2147,
+      "step": 211
+    },
+    {
+      "epoch": 0.040787614609950396,
+      "grad_norm": 0.32252571596181884,
+      "learning_rate": 0.00019993910686579507,
+      "loss": 1.2185,
+      "step": 212
+    },
+    {
+      "epoch": 0.0409800090184879,
+      "grad_norm": 0.42545588466081985,
+      "learning_rate": 0.00019993691292428365,
+      "loss": 1.198,
+      "step": 213
+    },
+    {
+      "epoch": 0.041172403427025404,
+      "grad_norm": 0.3269172449298148,
+      "learning_rate": 0.00019993468016842682,
+      "loss": 1.2332,
+      "step": 214
+    },
+    {
+      "epoch": 0.041364797835562905,
+      "grad_norm": 0.33412442290693023,
+      "learning_rate": 0.00019993240859909176,
+      "loss": 1.2392,
+      "step": 215
+    },
+    {
+      "epoch": 0.041557192244100406,
+      "grad_norm": 0.37427266017576016,
+      "learning_rate": 0.00019993009821716074,
+      "loss": 1.1579,
+      "step": 216
+    },
+    {
+      "epoch": 0.04174958665263791,
+      "grad_norm": 0.32304928157084595,
+      "learning_rate": 0.00019992774902353105,
+      "loss": 1.2504,
+      "step": 217
+    },
+    {
+      "epoch": 0.04194198106117541,
+      "grad_norm": 0.4210692012146056,
+      "learning_rate": 0.0001999253610191151,
+      "loss": 1.255,
+      "step": 218
+    },
+    {
+      "epoch": 0.04213437546971291,
+      "grad_norm": 0.39979090531976197,
+      "learning_rate": 0.00019992293420484039,
+      "loss": 1.2154,
+      "step": 219
+    },
+    {
+      "epoch": 0.04232676987825041,
+      "grad_norm": 0.33287705270465007,
+      "learning_rate": 0.00019992046858164944,
+      "loss": 1.3106,
+      "step": 220
+    },
+    {
+      "epoch": 0.04251916428678792,
+      "grad_norm": 0.4267622193847413,
+      "learning_rate": 0.0001999179641504999,
+      "loss": 1.156,
+      "step": 221
+    },
+    {
+      "epoch": 0.04271155869532542,
+      "grad_norm": 0.3489543426400865,
+      "learning_rate": 0.00019991542091236437,
+      "loss": 1.2274,
+      "step": 222
+    },
+    {
+      "epoch": 0.04290395310386292,
+      "grad_norm": 0.3965495981572789,
+      "learning_rate": 0.00019991283886823074,
+      "loss": 1.2429,
+      "step": 223
+    },
+    {
+      "epoch": 0.04309634751240042,
+      "grad_norm": 0.35185812275485573,
+      "learning_rate": 0.00019991021801910177,
+      "loss": 1.1352,
+      "step": 224
+    },
+    {
+      "epoch": 0.04328874192093792,
+      "grad_norm": 0.4269523111931177,
+      "learning_rate": 0.00019990755836599538,
+      "loss": 1.2589,
+      "step": 225
+    },
+    {
+      "epoch": 0.04348113632947542,
+      "grad_norm": 0.3197675379281588,
+      "learning_rate": 0.0001999048599099446,
+      "loss": 1.2607,
+      "step": 226
+    },
+    {
+      "epoch": 0.043673530738012924,
+      "grad_norm": 0.31993439796766976,
+      "learning_rate": 0.00019990212265199738,
+      "loss": 1.3237,
+      "step": 227
+    },
+    {
+      "epoch": 0.04386592514655043,
+      "grad_norm": 0.28977818784032877,
+      "learning_rate": 0.0001998993465932169,
+      "loss": 1.2173,
+      "step": 228
+    },
+    {
+      "epoch": 0.04405831955508793,
+      "grad_norm": 0.3382834816165424,
+      "learning_rate": 0.00019989653173468135,
+      "loss": 1.2356,
+      "step": 229
+    },
+    {
+      "epoch": 0.04425071396362543,
+      "grad_norm": 0.35078039254538645,
+      "learning_rate": 0.000199893678077484,
+      "loss": 1.2551,
+      "step": 230
+    },
+    {
+      "epoch": 0.044443108372162934,
+      "grad_norm": 0.35486481756596094,
+      "learning_rate": 0.00019989078562273314,
+      "loss": 1.2036,
+      "step": 231
+    },
+    {
+      "epoch": 0.044635502780700435,
+      "grad_norm": 0.3642186760167112,
+      "learning_rate": 0.0001998878543715522,
+      "loss": 1.2376,
+      "step": 232
+    },
+    {
+      "epoch": 0.044827897189237936,
+      "grad_norm": 0.3800056324738678,
+      "learning_rate": 0.0001998848843250796,
+      "loss": 1.2665,
+      "step": 233
+    },
+    {
+      "epoch": 0.04502029159777544,
+      "grad_norm": 0.3237256422775167,
+      "learning_rate": 0.00019988187548446894,
+      "loss": 1.133,
+      "step": 234
+    },
+    {
+      "epoch": 0.04521268600631294,
+      "grad_norm": 0.2837551658060299,
+      "learning_rate": 0.0001998788278508888,
+      "loss": 1.1966,
+      "step": 235
+    },
+    {
+      "epoch": 0.045405080414850446,
+      "grad_norm": 0.4134776579615321,
+      "learning_rate": 0.00019987574142552275,
+      "loss": 1.2163,
+      "step": 236
+    },
+    {
+      "epoch": 0.04559747482338795,
+      "grad_norm": 0.3694123337421609,
+      "learning_rate": 0.00019987261620956964,
+      "loss": 1.2597,
+      "step": 237
+    },
+    {
+      "epoch": 0.04578986923192545,
+      "grad_norm": 0.4114395315715834,
+      "learning_rate": 0.00019986945220424324,
+      "loss": 1.3565,
+      "step": 238
+    },
+    {
+      "epoch": 0.04598226364046295,
+      "grad_norm": 0.3084412092455137,
+      "learning_rate": 0.00019986624941077238,
+      "loss": 1.1464,
+      "step": 239
+    },
+    {
+      "epoch": 0.04617465804900045,
+      "grad_norm": 0.349142692471618,
+      "learning_rate": 0.000199863007830401,
+      "loss": 1.2835,
+      "step": 240
+    },
+    {
+      "epoch": 0.04636705245753795,
+      "grad_norm": 0.2732533508503848,
+      "learning_rate": 0.0001998597274643881,
+      "loss": 1.338,
+      "step": 241
+    },
+    {
+      "epoch": 0.04655944686607545,
+      "grad_norm": 0.25748553040026023,
+      "learning_rate": 0.00019985640831400776,
+      "loss": 1.2175,
+      "step": 242
+    },
+    {
+      "epoch": 0.04675184127461296,
+      "grad_norm": 0.32749537921051525,
+      "learning_rate": 0.0001998530503805491,
+      "loss": 1.1979,
+      "step": 243
+    },
+    {
+      "epoch": 0.04694423568315046,
+      "grad_norm": 0.3717967835992418,
+      "learning_rate": 0.00019984965366531623,
+      "loss": 1.2869,
+      "step": 244
+    },
+    {
+      "epoch": 0.04713663009168796,
+      "grad_norm": 0.3410985836741922,
+      "learning_rate": 0.00019984621816962844,
+      "loss": 1.3445,
+      "step": 245
+    },
+    {
+      "epoch": 0.04732902450022546,
+      "grad_norm": 0.3767653735452373,
+      "learning_rate": 0.00019984274389482005,
+      "loss": 1.2968,
+      "step": 246
+    },
+    {
+      "epoch": 0.047521418908762963,
+      "grad_norm": 0.3063748515018053,
+      "learning_rate": 0.00019983923084224045,
+      "loss": 1.3421,
+      "step": 247
+    },
+    {
+      "epoch": 0.047713813317300464,
+      "grad_norm": 0.3953429677010023,
+      "learning_rate": 0.00019983567901325403,
+      "loss": 1.2541,
+      "step": 248
+    },
+    {
+      "epoch": 0.047906207725837965,
+      "grad_norm": 0.35466087149773695,
+      "learning_rate": 0.00019983208840924026,
+      "loss": 1.2191,
+      "step": 249
+    },
+    {
+      "epoch": 0.048098602134375466,
+      "grad_norm": 0.3354957301340148,
+      "learning_rate": 0.0001998284590315937,
+      "loss": 1.1917,
+      "step": 250
+    },
+    {
+      "epoch": 0.048290996542912974,
+      "grad_norm": 0.40041914964435016,
+      "learning_rate": 0.00019982479088172405,
+      "loss": 1.2996,
+      "step": 251
+    },
+    {
+      "epoch": 0.048483390951450475,
+      "grad_norm": 0.3979208234651924,
+      "learning_rate": 0.00019982108396105584,
+      "loss": 1.274,
+      "step": 252
+    },
+    {
+      "epoch": 0.048675785359987976,
+      "grad_norm": 0.4826217626962861,
+      "learning_rate": 0.00019981733827102884,
+      "loss": 1.3042,
+      "step": 253
+    },
+    {
+      "epoch": 0.04886817976852548,
+      "grad_norm": 0.33611727898812377,
+      "learning_rate": 0.00019981355381309789,
+      "loss": 1.2157,
+      "step": 254
+    },
+    {
+      "epoch": 0.04906057417706298,
+      "grad_norm": 0.37058871405251453,
+      "learning_rate": 0.0001998097305887328,
+      "loss": 1.2581,
+      "step": 255
+    },
+    {
+      "epoch": 0.04925296858560048,
+      "grad_norm": 0.3092656056799116,
+      "learning_rate": 0.00019980586859941847,
+      "loss": 1.2189,
+      "step": 256
+    },
+    {
+      "epoch": 0.04944536299413798,
+      "grad_norm": 0.3548142864921303,
+      "learning_rate": 0.00019980196784665478,
+      "loss": 1.1467,
+      "step": 257
+    },
+    {
+      "epoch": 0.04963775740267549,
+      "grad_norm": 0.41277522858899796,
+      "learning_rate": 0.00019979802833195682,
+      "loss": 1.1699,
+      "step": 258
+    },
+    {
+      "epoch": 0.04983015181121299,
+      "grad_norm": 0.34689799716308284,
+      "learning_rate": 0.00019979405005685465,
+      "loss": 1.2887,
+      "step": 259
+    },
+    {
+      "epoch": 0.05002254621975049,
+      "grad_norm": 0.31082236508657474,
+      "learning_rate": 0.00019979003302289335,
+      "loss": 1.2578,
+      "step": 260
+    },
+    {
+      "epoch": 0.05021494062828799,
+      "grad_norm": 0.32128293101778826,
+      "learning_rate": 0.0001997859772316331,
+      "loss": 1.1811,
+      "step": 261
+    },
+    {
+      "epoch": 0.05040733503682549,
+      "grad_norm": 0.39289119700833075,
+      "learning_rate": 0.00019978188268464912,
+      "loss": 1.0934,
+      "step": 262
+    },
+    {
+      "epoch": 0.05059972944536299,
+      "grad_norm": 0.3161307556322266,
+      "learning_rate": 0.0001997777493835317,
+      "loss": 1.2309,
+      "step": 263
+    },
+    {
+      "epoch": 0.050792123853900494,
+      "grad_norm": 0.37715248373313104,
+      "learning_rate": 0.00019977357732988614,
+      "loss": 1.2183,
+      "step": 264
+    },
+    {
+      "epoch": 0.050984518262437994,
+      "grad_norm": 0.3773970132717969,
+      "learning_rate": 0.0001997693665253329,
+      "loss": 1.1416,
+      "step": 265
+    },
+    {
+      "epoch": 0.0511769126709755,
+      "grad_norm": 0.35699919953799203,
+      "learning_rate": 0.0001997651169715073,
+      "loss": 1.2559,
+      "step": 266
+    },
+    {
+      "epoch": 0.051369307079513,
+      "grad_norm": 0.5965240119539538,
+      "learning_rate": 0.00019976082867005984,
+      "loss": 1.1933,
+      "step": 267
+    },
+    {
+      "epoch": 0.051561701488050504,
+      "grad_norm": 0.30273904689782033,
+      "learning_rate": 0.00019975650162265608,
+      "loss": 1.2367,
+      "step": 268
+    },
+    {
+      "epoch": 0.051754095896588005,
+      "grad_norm": 0.38380292113931525,
+      "learning_rate": 0.0001997521358309766,
+      "loss": 1.2624,
+      "step": 269
+    },
+    {
+      "epoch": 0.051946490305125506,
+      "grad_norm": 0.3992247682089908,
+      "learning_rate": 0.00019974773129671701,
+      "loss": 1.2024,
+      "step": 270
+    },
+    {
+      "epoch": 0.05213888471366301,
+      "grad_norm": 0.33516486240992777,
+      "learning_rate": 0.00019974328802158797,
+      "loss": 1.1163,
+      "step": 271
+    },
+    {
+      "epoch": 0.05233127912220051,
+      "grad_norm": 0.36672908086251843,
+      "learning_rate": 0.0001997388060073152,
+      "loss": 1.2088,
+      "step": 272
+    },
+    {
+      "epoch": 0.052523673530738016,
+      "grad_norm": 0.3238539367847135,
+      "learning_rate": 0.00019973428525563947,
+      "loss": 1.262,
+      "step": 273
+    },
+    {
+      "epoch": 0.05271606793927552,
+      "grad_norm": 0.369542699902845,
+      "learning_rate": 0.00019972972576831656,
+      "loss": 1.1273,
+      "step": 274
+    },
+    {
+      "epoch": 0.05290846234781302,
+      "grad_norm": 0.4458707809374286,
+      "learning_rate": 0.0001997251275471174,
+      "loss": 1.1909,
+      "step": 275
+    },
+    {
+      "epoch": 0.05310085675635052,
+      "grad_norm": 0.4577204769176713,
+      "learning_rate": 0.00019972049059382782,
+      "loss": 1.295,
+      "step": 276
+    },
+    {
+      "epoch": 0.05329325116488802,
+      "grad_norm": 0.3116568734534432,
+      "learning_rate": 0.00019971581491024873,
+      "loss": 1.2724,
+      "step": 277
+    },
+    {
+      "epoch": 0.05348564557342552,
+      "grad_norm": 0.3148014098925133,
+      "learning_rate": 0.0001997111004981962,
+      "loss": 1.2434,
+      "step": 278
+    },
+    {
+      "epoch": 0.05367803998196302,
+      "grad_norm": 0.3313017770241379,
+      "learning_rate": 0.00019970634735950115,
+      "loss": 1.2233,
+      "step": 279
+    },
+    {
+      "epoch": 0.05387043439050053,
+      "grad_norm": 0.31638418972864946,
+      "learning_rate": 0.00019970155549600978,
+      "loss": 1.2812,
+      "step": 280
+    },
+    {
+      "epoch": 0.05406282879903803,
+      "grad_norm": 0.3440305445933539,
+      "learning_rate": 0.00019969672490958304,
+      "loss": 1.2002,
+      "step": 281
+    },
+    {
+      "epoch": 0.05425522320757553,
+      "grad_norm": 0.42212263392448246,
+      "learning_rate": 0.0001996918556020972,
+      "loss": 1.2141,
+      "step": 282
+    },
+    {
+      "epoch": 0.05444761761611303,
+      "grad_norm": 0.3913418964374786,
+      "learning_rate": 0.0001996869475754434,
+      "loss": 1.1538,
+      "step": 283
+    },
+    {
+      "epoch": 0.05464001202465053,
+      "grad_norm": 0.255157434086349,
+      "learning_rate": 0.00019968200083152782,
+      "loss": 1.1592,
+      "step": 284
+    },
+    {
+      "epoch": 0.054832406433188034,
+      "grad_norm": 0.3949120063024248,
+      "learning_rate": 0.00019967701537227175,
+      "loss": 1.2124,
+      "step": 285
+    },
+    {
+      "epoch": 0.055024800841725535,
+      "grad_norm": 0.4285206326851563,
+      "learning_rate": 0.00019967199119961152,
+      "loss": 1.2727,
+      "step": 286
+    },
+    {
+      "epoch": 0.055217195250263036,
+      "grad_norm": 0.38899814654007914,
+      "learning_rate": 0.0001996669283154984,
+      "loss": 1.3096,
+      "step": 287
+    },
+    {
+      "epoch": 0.055409589658800544,
+      "grad_norm": 0.36328914022359676,
+      "learning_rate": 0.0001996618267218988,
+      "loss": 1.3222,
+      "step": 288
+    },
+    {
+      "epoch": 0.055601984067338045,
+      "grad_norm": 0.32127260127884855,
+      "learning_rate": 0.00019965668642079408,
+      "loss": 1.2761,
+      "step": 289
+    },
+    {
+      "epoch": 0.055794378475875546,
+      "grad_norm": 0.31629289747544054,
+      "learning_rate": 0.00019965150741418073,
+      "loss": 1.2771,
+      "step": 290
+    },
+    {
+      "epoch": 0.05598677288441305,
+      "grad_norm": 0.3920429739286175,
+      "learning_rate": 0.0001996462897040702,
+      "loss": 1.1744,
+      "step": 291
+    },
+    {
+      "epoch": 0.05617916729295055,
+      "grad_norm": 0.29619481909023426,
+      "learning_rate": 0.0001996410332924889,
+      "loss": 1.262,
+      "step": 292
+    },
+    {
+      "epoch": 0.05637156170148805,
+      "grad_norm": 0.37749915135309287,
+      "learning_rate": 0.0001996357381814785,
+      "loss": 1.1249,
+      "step": 293
+    },
+    {
+      "epoch": 0.05656395611002555,
+      "grad_norm": 0.3146318722157494,
+      "learning_rate": 0.00019963040437309549,
+      "loss": 1.2694,
+      "step": 294
+    },
+    {
+      "epoch": 0.05675635051856306,
+      "grad_norm": 0.3743924130658171,
+      "learning_rate": 0.00019962503186941142,
+      "loss": 1.3085,
+      "step": 295
+    },
+    {
+      "epoch": 0.05694874492710056,
+      "grad_norm": 0.38695803467811296,
+      "learning_rate": 0.00019961962067251298,
+      "loss": 1.1613,
+      "step": 296
+    },
+    {
+      "epoch": 0.05714113933563806,
+      "grad_norm": 0.30996720449326803,
+      "learning_rate": 0.00019961417078450178,
+      "loss": 1.2838,
+      "step": 297
+    },
+    {
+      "epoch": 0.05733353374417556,
+      "grad_norm": 0.3543253614034791,
+      "learning_rate": 0.00019960868220749448,
+      "loss": 1.1588,
+      "step": 298
+    },
+    {
+      "epoch": 0.05752592815271306,
+      "grad_norm": 0.3072955296875589,
+      "learning_rate": 0.00019960315494362284,
+      "loss": 1.2817,
+      "step": 299
+    },
+    {
+      "epoch": 0.05771832256125056,
+      "grad_norm": 0.3433406675719805,
+      "learning_rate": 0.00019959758899503353,
+      "loss": 1.1408,
+      "step": 300
+    },
+    {
+      "epoch": 0.05791071696978806,
+      "grad_norm": 0.3985503674139911,
+      "learning_rate": 0.0001995919843638883,
+      "loss": 1.164,
+      "step": 301
+    },
+    {
+      "epoch": 0.058103111378325564,
+      "grad_norm": 0.3130134403606259,
+      "learning_rate": 0.00019958634105236395,
+      "loss": 1.2136,
+      "step": 302
+    },
+    {
+      "epoch": 0.05829550578686307,
+      "grad_norm": 0.42122450250188936,
+      "learning_rate": 0.00019958065906265228,
+      "loss": 1.1732,
+      "step": 303
+    },
+    {
+      "epoch": 0.05848790019540057,
+      "grad_norm": 0.42812024204470966,
+      "learning_rate": 0.0001995749383969601,
+      "loss": 1.1633,
+      "step": 304
+    },
+    {
+      "epoch": 0.058680294603938074,
+      "grad_norm": 0.2834380855521078,
+      "learning_rate": 0.00019956917905750924,
+      "loss": 1.2696,
+      "step": 305
+    },
+    {
+      "epoch": 0.058872689012475575,
+      "grad_norm": 0.4658231344202302,
+      "learning_rate": 0.00019956338104653657,
+      "loss": 1.26,
+      "step": 306
+    },
+    {
+      "epoch": 0.059065083421013076,
+      "grad_norm": 0.2913686370266581,
+      "learning_rate": 0.00019955754436629399,
+      "loss": 1.2472,
+      "step": 307
+    },
+    {
+      "epoch": 0.05925747782955058,
+      "grad_norm": 0.3246465243061967,
+      "learning_rate": 0.00019955166901904837,
+      "loss": 1.1957,
+      "step": 308
+    },
+    {
+      "epoch": 0.05944987223808808,
+      "grad_norm": 0.3469837804043362,
+      "learning_rate": 0.00019954575500708162,
+      "loss": 1.3085,
+      "step": 309
+    },
+    {
+      "epoch": 0.059642266646625586,
+      "grad_norm": 0.37556550588560683,
+      "learning_rate": 0.0001995398023326907,
+      "loss": 1.2829,
+      "step": 310
+    },
+    {
+      "epoch": 0.05983466105516309,
+      "grad_norm": 0.3499872905769135,
+      "learning_rate": 0.00019953381099818755,
+      "loss": 1.2292,
+      "step": 311
+    },
+    {
+      "epoch": 0.06002705546370059,
+      "grad_norm": 0.3717751806601391,
+      "learning_rate": 0.00019952778100589913,
+      "loss": 1.2553,
+      "step": 312
+    },
+    {
+      "epoch": 0.06021944987223809,
+      "grad_norm": 0.36402968272922626,
+      "learning_rate": 0.00019952171235816747,
+      "loss": 1.1841,
+      "step": 313
+    },
+    {
+      "epoch": 0.06041184428077559,
+      "grad_norm": 0.30485831163497246,
+      "learning_rate": 0.00019951560505734945,
+      "loss": 1.2173,
+      "step": 314
+    },
+    {
+      "epoch": 0.06060423868931309,
+      "grad_norm": 0.32969845452172397,
+      "learning_rate": 0.00019950945910581717,
+      "loss": 1.2912,
+      "step": 315
+    },
+    {
+      "epoch": 0.06079663309785059,
+      "grad_norm": 0.405942459393916,
+      "learning_rate": 0.00019950327450595764,
+      "loss": 1.1841,
+      "step": 316
+    },
+    {
+      "epoch": 0.06098902750638809,
+      "grad_norm": 0.36100896058101106,
+      "learning_rate": 0.00019949705126017287,
+      "loss": 1.1737,
+      "step": 317
+    },
+    {
+      "epoch": 0.0611814219149256,
+      "grad_norm": 0.3654122632134192,
+      "learning_rate": 0.00019949078937087986,
+      "loss": 1.0859,
+      "step": 318
+    },
+    {
+      "epoch": 0.0613738163234631,
+      "grad_norm": 0.3807698199736515,
+      "learning_rate": 0.0001994844888405107,
+      "loss": 1.2496,
+      "step": 319
+    },
+    {
+      "epoch": 0.0615662107320006,
+      "grad_norm": 0.39433143743513216,
+      "learning_rate": 0.00019947814967151244,
+      "loss": 1.2299,
+      "step": 320
+    },
+    {
+      "epoch": 0.0617586051405381,
+      "grad_norm": 0.4138480507334561,
+      "learning_rate": 0.00019947177186634715,
+      "loss": 1.1938,
+      "step": 321
+    },
+    {
+      "epoch": 0.061950999549075604,
+      "grad_norm": 0.3627091646312647,
+      "learning_rate": 0.00019946535542749184,
+      "loss": 1.2097,
+      "step": 322
+    },
+    {
+      "epoch": 0.062143393957613105,
+      "grad_norm": 0.3529420782514917,
+      "learning_rate": 0.00019945890035743866,
+      "loss": 1.2047,
+      "step": 323
+    },
+    {
+      "epoch": 0.062335788366150606,
+      "grad_norm": 0.339502693512079,
+      "learning_rate": 0.00019945240665869465,
+      "loss": 1.2139,
+      "step": 324
+    },
+    {
+      "epoch": 0.0625281827746881,
+      "grad_norm": 0.38730574906823206,
+      "learning_rate": 0.00019944587433378186,
+      "loss": 1.2119,
+      "step": 325
+    },
+    {
+      "epoch": 0.06272057718322561,
+      "grad_norm": 0.3239599185069091,
+      "learning_rate": 0.0001994393033852374,
+      "loss": 1.0793,
+      "step": 326
+    },
+    {
+      "epoch": 0.06291297159176311,
+      "grad_norm": 0.42172393932528346,
+      "learning_rate": 0.00019943269381561334,
+      "loss": 1.2387,
+      "step": 327
+    },
+    {
+      "epoch": 0.06310536600030062,
+      "grad_norm": 0.4649999946630129,
+      "learning_rate": 0.00019942604562747678,
+      "loss": 1.1935,
+      "step": 328
+    },
+    {
+      "epoch": 0.06329776040883812,
+      "grad_norm": 0.3283055534525505,
+      "learning_rate": 0.00019941935882340976,
+      "loss": 1.1858,
+      "step": 329
+    },
+    {
+      "epoch": 0.06349015481737562,
+      "grad_norm": 0.3276058870296969,
+      "learning_rate": 0.00019941263340600939,
+      "loss": 1.2873,
+      "step": 330
+    },
+    {
+      "epoch": 0.06368254922591313,
+      "grad_norm": 0.25504389436073543,
+      "learning_rate": 0.00019940586937788776,
+      "loss": 1.2078,
+      "step": 331
+    },
+    {
+      "epoch": 0.06387494363445062,
+      "grad_norm": 0.3163236598873421,
+      "learning_rate": 0.0001993990667416719,
+      "loss": 1.1835,
+      "step": 332
+    },
+    {
+      "epoch": 0.06406733804298813,
+      "grad_norm": 0.43909878586615,
+      "learning_rate": 0.0001993922255000039,
+      "loss": 1.1776,
+      "step": 333
+    },
+    {
+      "epoch": 0.06425973245152562,
+      "grad_norm": 0.38117519645111986,
+      "learning_rate": 0.0001993853456555408,
+      "loss": 1.1558,
+      "step": 334
+    },
+    {
+      "epoch": 0.06445212686006313,
+      "grad_norm": 0.31423356199533864,
+      "learning_rate": 0.00019937842721095468,
+      "loss": 1.2456,
+      "step": 335
+    },
+    {
+      "epoch": 0.06464452126860064,
+      "grad_norm": 0.27265001833117847,
+      "learning_rate": 0.00019937147016893257,
+      "loss": 1.1418,
+      "step": 336
+    },
+    {
+      "epoch": 0.06483691567713813,
+      "grad_norm": 0.34005108000768947,
+      "learning_rate": 0.00019936447453217646,
+      "loss": 1.2031,
+      "step": 337
+    },
+    {
+      "epoch": 0.06502931008567564,
+      "grad_norm": 0.3491656312670484,
+      "learning_rate": 0.00019935744030340346,
+      "loss": 1.2508,
+      "step": 338
+    },
+    {
+      "epoch": 0.06522170449421313,
+      "grad_norm": 0.3367736624898666,
+      "learning_rate": 0.00019935036748534553,
+      "loss": 1.2166,
+      "step": 339
+    },
+    {
+      "epoch": 0.06541409890275064,
+      "grad_norm": 0.3128517006692993,
+      "learning_rate": 0.0001993432560807497,
+      "loss": 1.2333,
+      "step": 340
+    },
+    {
+      "epoch": 0.06560649331128814,
+      "grad_norm": 0.30348133599339916,
+      "learning_rate": 0.00019933610609237793,
+      "loss": 1.1224,
+      "step": 341
+    },
+    {
+      "epoch": 0.06579888771982564,
+      "grad_norm": 0.29819897968115294,
+      "learning_rate": 0.00019932891752300717,
+      "loss": 1.2299,
+      "step": 342
+    },
+    {
+      "epoch": 0.06599128212836314,
+      "grad_norm": 0.33118202228380633,
+      "learning_rate": 0.00019932169037542946,
+      "loss": 1.1434,
+      "step": 343
+    },
+    {
+      "epoch": 0.06618367653690065,
+      "grad_norm": 0.410488685395728,
+      "learning_rate": 0.00019931442465245165,
+      "loss": 1.2197,
+      "step": 344
+    },
+    {
+      "epoch": 0.06637607094543815,
+      "grad_norm": 0.32967419199508924,
+      "learning_rate": 0.00019930712035689575,
+      "loss": 1.2488,
+      "step": 345
+    },
+    {
+      "epoch": 0.06656846535397565,
+      "grad_norm": 0.303306771079036,
+      "learning_rate": 0.00019929977749159859,
+      "loss": 1.1851,
+      "step": 346
+    },
+    {
+      "epoch": 0.06676085976251316,
+      "grad_norm": 0.3991113102999262,
+      "learning_rate": 0.00019929239605941208,
+      "loss": 1.1829,
+      "step": 347
+    },
+    {
+      "epoch": 0.06695325417105065,
+      "grad_norm": 0.26977548277363106,
+      "learning_rate": 0.0001992849760632031,
+      "loss": 1.2384,
+      "step": 348
+    },
+    {
+      "epoch": 0.06714564857958816,
+      "grad_norm": 0.4217022102813526,
+      "learning_rate": 0.00019927751750585347,
+      "loss": 1.1611,
+      "step": 349
+    },
+    {
+      "epoch": 0.06733804298812565,
+      "grad_norm": 0.27307458080717384,
+      "learning_rate": 0.00019927002039026002,
+      "loss": 1.2025,
+      "step": 350
+    },
+    {
+      "epoch": 0.06753043739666316,
+      "grad_norm": 0.3139832334435657,
+      "learning_rate": 0.00019926248471933454,
+      "loss": 1.228,
+      "step": 351
+    },
+    {
+      "epoch": 0.06772283180520067,
+      "grad_norm": 0.2890948080597137,
+      "learning_rate": 0.0001992549104960038,
+      "loss": 1.2003,
+      "step": 352
+    },
+    {
+      "epoch": 0.06791522621373816,
+      "grad_norm": 0.3364470939670738,
+      "learning_rate": 0.0001992472977232095,
+      "loss": 1.243,
+      "step": 353
+    },
+    {
+      "epoch": 0.06810762062227567,
+      "grad_norm": 0.3106432538005775,
+      "learning_rate": 0.00019923964640390843,
+      "loss": 1.1601,
+      "step": 354
+    },
+    {
+      "epoch": 0.06830001503081316,
+      "grad_norm": 0.2543076746328548,
+      "learning_rate": 0.00019923195654107225,
+      "loss": 1.1535,
+      "step": 355
+    },
+    {
+      "epoch": 0.06849240943935067,
+      "grad_norm": 0.3605226962761155,
+      "learning_rate": 0.00019922422813768758,
+      "loss": 1.2255,
+      "step": 356
+    },
+    {
+      "epoch": 0.06868480384788817,
+      "grad_norm": 0.3069682346498868,
+      "learning_rate": 0.00019921646119675605,
+      "loss": 1.2549,
+      "step": 357
+    },
+    {
+      "epoch": 0.06887719825642567,
+      "grad_norm": 0.3562071033482331,
+      "learning_rate": 0.00019920865572129425,
+      "loss": 1.3144,
+      "step": 358
+    },
+    {
+      "epoch": 0.06906959266496318,
+      "grad_norm": 0.2632783307081333,
+      "learning_rate": 0.00019920081171433379,
+      "loss": 1.2696,
+      "step": 359
+    },
+    {
+      "epoch": 0.06926198707350067,
+      "grad_norm": 0.3930372697927493,
+      "learning_rate": 0.00019919292917892112,
+      "loss": 1.2285,
+      "step": 360
+    },
+    {
+      "epoch": 0.06945438148203818,
+      "grad_norm": 0.29756656546134286,
+      "learning_rate": 0.00019918500811811778,
+      "loss": 1.1767,
+      "step": 361
+    },
+    {
+      "epoch": 0.06964677589057568,
+      "grad_norm": 0.2987196051114864,
+      "learning_rate": 0.00019917704853500016,
+      "loss": 1.1441,
+      "step": 362
+    },
+    {
+      "epoch": 0.06983917029911318,
+      "grad_norm": 0.3273696683415939,
+      "learning_rate": 0.00019916905043265972,
+      "loss": 1.2167,
+      "step": 363
+    },
+    {
+      "epoch": 0.07003156470765068,
+      "grad_norm": 0.32799607870164604,
+      "learning_rate": 0.00019916101381420285,
+      "loss": 1.1706,
+      "step": 364
+    },
+    {
+      "epoch": 0.07022395911618819,
+      "grad_norm": 0.31099871021757747,
+      "learning_rate": 0.00019915293868275083,
+      "loss": 1.1124,
+      "step": 365
+    },
+    {
+      "epoch": 0.0704163535247257,
+      "grad_norm": 0.2596040843750913,
+      "learning_rate": 0.00019914482504143995,
+      "loss": 1.2437,
+      "step": 366
+    },
+    {
+      "epoch": 0.07060874793326319,
+      "grad_norm": 0.34030655508034796,
+      "learning_rate": 0.00019913667289342147,
+      "loss": 1.1917,
+      "step": 367
+    },
+    {
+      "epoch": 0.0708011423418007,
+      "grad_norm": 0.3390501754848551,
+      "learning_rate": 0.0001991284822418616,
+      "loss": 1.1255,
+      "step": 368
+    },
+    {
+      "epoch": 0.07099353675033819,
+      "grad_norm": 0.3128963159659474,
+      "learning_rate": 0.00019912025308994148,
+      "loss": 1.2202,
+      "step": 369
+    },
+    {
+      "epoch": 0.0711859311588757,
+      "grad_norm": 0.3751209174302873,
+      "learning_rate": 0.00019911198544085722,
+      "loss": 1.2319,
+      "step": 370
+    },
+    {
+      "epoch": 0.07137832556741319,
+      "grad_norm": 0.3626056920250688,
+      "learning_rate": 0.00019910367929781988,
+      "loss": 1.2451,
+      "step": 371
+    },
+    {
+      "epoch": 0.0715707199759507,
+      "grad_norm": 0.2651022892019285,
+      "learning_rate": 0.00019909533466405546,
+      "loss": 1.215,
+      "step": 372
+    },
+    {
+      "epoch": 0.0717631143844882,
+      "grad_norm": 0.2695241953067502,
+      "learning_rate": 0.00019908695154280496,
+      "loss": 1.2342,
+      "step": 373
+    },
+    {
+      "epoch": 0.0719555087930257,
+      "grad_norm": 0.396793772597451,
+      "learning_rate": 0.00019907852993732424,
+      "loss": 1.2262,
+      "step": 374
+    },
+    {
+      "epoch": 0.07214790320156321,
+      "grad_norm": 0.3867323324164736,
+      "learning_rate": 0.0001990700698508842,
+      "loss": 1.2584,
+      "step": 375
+    },
+    {
+      "epoch": 0.0723402976101007,
+      "grad_norm": 0.32459859740205627,
+      "learning_rate": 0.0001990615712867706,
+      "loss": 1.2777,
+      "step": 376
+    },
+    {
+      "epoch": 0.07253269201863821,
+      "grad_norm": 0.29589772643405204,
+      "learning_rate": 0.00019905303424828417,
+      "loss": 1.1969,
+      "step": 377
+    },
+    {
+      "epoch": 0.0727250864271757,
+      "grad_norm": 0.2818076727449584,
+      "learning_rate": 0.00019904445873874068,
+      "loss": 1.2332,
+      "step": 378
+    },
+    {
+      "epoch": 0.07291748083571321,
+      "grad_norm": 0.32737308354754885,
+      "learning_rate": 0.00019903584476147065,
+      "loss": 1.1826,
+      "step": 379
+    },
+    {
+      "epoch": 0.07310987524425071,
+      "grad_norm": 0.32575071044631815,
+      "learning_rate": 0.00019902719231981974,
+      "loss": 1.2484,
+      "step": 380
+    },
+    {
+      "epoch": 0.07330226965278822,
+      "grad_norm": 0.3145521605379211,
+      "learning_rate": 0.00019901850141714841,
+      "loss": 1.1781,
+      "step": 381
+    },
+    {
+      "epoch": 0.07349466406132572,
+      "grad_norm": 0.4515392894166184,
+      "learning_rate": 0.0001990097720568321,
+      "loss": 1.2096,
+      "step": 382
+    },
+    {
+      "epoch": 0.07368705846986322,
+      "grad_norm": 0.29729760081447826,
+      "learning_rate": 0.00019900100424226125,
+      "loss": 1.1934,
+      "step": 383
+    },
+    {
+      "epoch": 0.07387945287840073,
+      "grad_norm": 0.3411834562153974,
+      "learning_rate": 0.00019899219797684113,
+      "loss": 1.2245,
+      "step": 384
+    },
+    {
+      "epoch": 0.07407184728693822,
+      "grad_norm": 0.2711191466888693,
+      "learning_rate": 0.000198983353263992,
+      "loss": 1.2488,
+      "step": 385
+    },
+    {
+      "epoch": 0.07426424169547573,
+      "grad_norm": 0.3165981210963245,
+      "learning_rate": 0.00019897447010714905,
+      "loss": 1.1832,
+      "step": 386
+    },
+    {
+      "epoch": 0.07445663610401322,
+      "grad_norm": 0.33524258854840583,
+      "learning_rate": 0.00019896554850976238,
+      "loss": 1.1359,
+      "step": 387
+    },
+    {
+      "epoch": 0.07464903051255073,
+      "grad_norm": 0.26846132333679673,
+      "learning_rate": 0.00019895658847529708,
+      "loss": 1.1609,
+      "step": 388
+    },
+    {
+      "epoch": 0.07484142492108824,
+      "grad_norm": 0.29609002005647955,
+      "learning_rate": 0.00019894759000723306,
+      "loss": 1.2819,
+      "step": 389
+    },
+    {
+      "epoch": 0.07503381932962573,
+      "grad_norm": 0.32203131595754636,
+      "learning_rate": 0.00019893855310906526,
+      "loss": 1.1585,
+      "step": 390
+    },
+    {
+      "epoch": 0.07522621373816324,
+      "grad_norm": 0.4019557977818726,
+      "learning_rate": 0.0001989294777843035,
+      "loss": 1.1633,
+      "step": 391
+    },
+    {
+      "epoch": 0.07541860814670073,
+      "grad_norm": 0.3334773688106313,
+      "learning_rate": 0.00019892036403647254,
+      "loss": 1.2853,
+      "step": 392
+    },
+    {
+      "epoch": 0.07561100255523824,
+      "grad_norm": 0.3209140493686673,
+      "learning_rate": 0.00019891121186911206,
+      "loss": 1.1525,
+      "step": 393
+    },
+    {
+      "epoch": 0.07580339696377573,
+      "grad_norm": 0.3271402047611454,
+      "learning_rate": 0.00019890202128577662,
+      "loss": 1.1953,
+      "step": 394
+    },
+    {
+      "epoch": 0.07599579137231324,
+      "grad_norm": 0.34274408744583185,
+      "learning_rate": 0.00019889279229003576,
+      "loss": 1.1798,
+      "step": 395
+    },
+    {
+      "epoch": 0.07618818578085075,
+      "grad_norm": 0.3867062817728671,
+      "learning_rate": 0.00019888352488547394,
+      "loss": 1.2351,
+      "step": 396
+    },
+    {
+      "epoch": 0.07638058018938824,
+      "grad_norm": 0.2874638157245163,
+      "learning_rate": 0.00019887421907569048,
+      "loss": 1.1834,
+      "step": 397
+    },
+    {
+      "epoch": 0.07657297459792575,
+      "grad_norm": 0.2517062899344747,
+      "learning_rate": 0.00019886487486429964,
+      "loss": 1.2664,
+      "step": 398
+    },
+    {
+      "epoch": 0.07676536900646325,
+      "grad_norm": 0.3447747875977769,
+      "learning_rate": 0.00019885549225493064,
+      "loss": 1.1547,
+      "step": 399
+    },
+    {
+      "epoch": 0.07695776341500075,
+      "grad_norm": 0.28999841111237157,
+      "learning_rate": 0.00019884607125122755,
+      "loss": 1.1282,
+      "step": 400
+    },
+    {
+      "epoch": 0.07715015782353825,
+      "grad_norm": 0.3328577533371222,
+      "learning_rate": 0.0001988366118568494,
+      "loss": 1.1872,
+      "step": 401
+    },
+    {
+      "epoch": 0.07734255223207576,
+      "grad_norm": 0.3472634071603433,
+      "learning_rate": 0.0001988271140754701,
+      "loss": 1.0705,
+      "step": 402
+    },
+    {
+      "epoch": 0.07753494664061325,
+      "grad_norm": 0.28289529408099684,
+      "learning_rate": 0.00019881757791077845,
+      "loss": 1.2342,
+      "step": 403
+    },
+    {
+      "epoch": 0.07772734104915076,
+      "grad_norm": 0.31613099586606025,
+      "learning_rate": 0.00019880800336647824,
+      "loss": 1.2602,
+      "step": 404
+    },
+    {
+      "epoch": 0.07791973545768827,
+      "grad_norm": 0.42006664644376834,
+      "learning_rate": 0.0001987983904462881,
+      "loss": 1.1138,
+      "step": 405
+    },
+    {
+      "epoch": 0.07811212986622576,
+      "grad_norm": 0.35417887060708514,
+      "learning_rate": 0.00019878873915394154,
+      "loss": 1.2261,
+      "step": 406
+    },
+    {
+      "epoch": 0.07830452427476327,
+      "grad_norm": 0.38430248445706694,
+      "learning_rate": 0.00019877904949318703,
+      "loss": 1.2139,
+      "step": 407
+    },
+    {
+      "epoch": 0.07849691868330076,
+      "grad_norm": 0.3548196863630682,
+      "learning_rate": 0.00019876932146778794,
+      "loss": 1.1664,
+      "step": 408
+    },
+    {
+      "epoch": 0.07868931309183827,
+      "grad_norm": 0.3901646019896108,
+      "learning_rate": 0.00019875955508152253,
+      "loss": 1.1328,
+      "step": 409
+    },
+    {
+      "epoch": 0.07888170750037576,
+      "grad_norm": 0.3238762071302361,
+      "learning_rate": 0.0001987497503381839,
+      "loss": 1.1553,
+      "step": 410
+    },
+    {
+      "epoch": 0.07907410190891327,
+      "grad_norm": 0.29814290274840377,
+      "learning_rate": 0.00019873990724158014,
+      "loss": 1.1672,
+      "step": 411
+    },
+    {
+      "epoch": 0.07926649631745078,
+      "grad_norm": 0.35034479493074044,
+      "learning_rate": 0.00019873002579553418,
+      "loss": 1.0997,
+      "step": 412
+    },
+    {
+      "epoch": 0.07945889072598827,
+      "grad_norm": 0.35867542565891897,
+      "learning_rate": 0.00019872010600388392,
+      "loss": 1.1303,
+      "step": 413
+    },
+    {
+      "epoch": 0.07965128513452578,
+      "grad_norm": 0.2909167833413755,
+      "learning_rate": 0.00019871014787048197,
+      "loss": 1.1295,
+      "step": 414
+    },
+    {
+      "epoch": 0.07984367954306328,
+      "grad_norm": 0.3668728514473036,
+      "learning_rate": 0.00019870015139919606,
+      "loss": 1.2464,
+      "step": 415
+    },
+    {
+      "epoch": 0.08003607395160078,
+      "grad_norm": 0.3053516661915177,
+      "learning_rate": 0.00019869011659390866,
+      "loss": 1.1474,
+      "step": 416
+    },
+    {
+      "epoch": 0.08022846836013828,
+      "grad_norm": 0.36341260589354013,
+      "learning_rate": 0.00019868004345851716,
+      "loss": 1.2101,
+      "step": 417
+    },
+    {
+      "epoch": 0.08042086276867579,
+      "grad_norm": 0.33930090088971565,
+      "learning_rate": 0.00019866993199693392,
+      "loss": 1.0921,
+      "step": 418
+    },
+    {
+      "epoch": 0.0806132571772133,
+      "grad_norm": 0.49420476471108643,
+      "learning_rate": 0.000198659782213086,
+      "loss": 1.1695,
+      "step": 419
+    },
+    {
+      "epoch": 0.08080565158575079,
+      "grad_norm": 0.2841104206366389,
+      "learning_rate": 0.00019864959411091556,
+      "loss": 1.3438,
+      "step": 420
+    },
+    {
+      "epoch": 0.0809980459942883,
+      "grad_norm": 0.3438830634633235,
+      "learning_rate": 0.00019863936769437955,
+      "loss": 1.1926,
+      "step": 421
+    },
+    {
+      "epoch": 0.08119044040282579,
+      "grad_norm": 0.32544277162464536,
+      "learning_rate": 0.00019862910296744967,
+      "loss": 1.2137,
+      "step": 422
+    },
+    {
+      "epoch": 0.0813828348113633,
+      "grad_norm": 0.30936189164901584,
+      "learning_rate": 0.00019861879993411275,
+      "loss": 1.2412,
+      "step": 423
+    },
+    {
+      "epoch": 0.08157522921990079,
+      "grad_norm": 0.371471489265311,
+      "learning_rate": 0.00019860845859837032,
+      "loss": 1.0245,
+      "step": 424
+    },
+    {
+      "epoch": 0.0817676236284383,
+      "grad_norm": 0.33541081190268485,
+      "learning_rate": 0.00019859807896423882,
+      "loss": 1.1537,
+      "step": 425
+    },
+    {
+      "epoch": 0.0819600180369758,
+      "grad_norm": 0.3085229952981492,
+      "learning_rate": 0.0001985876610357496,
+      "loss": 1.116,
+      "step": 426
+    },
+    {
+      "epoch": 0.0821524124455133,
+      "grad_norm": 0.28895856037292456,
+      "learning_rate": 0.00019857720481694885,
+      "loss": 1.203,
+      "step": 427
+    },
+    {
+      "epoch": 0.08234480685405081,
+      "grad_norm": 0.45072463668237145,
+      "learning_rate": 0.00019856671031189766,
+      "loss": 1.1595,
+      "step": 428
+    },
+    {
+      "epoch": 0.0825372012625883,
+      "grad_norm": 0.3981656496594179,
+      "learning_rate": 0.000198556177524672,
+      "loss": 1.2791,
+      "step": 429
+    },
+    {
+      "epoch": 0.08272959567112581,
+      "grad_norm": 0.34959095793759987,
+      "learning_rate": 0.0001985456064593626,
+      "loss": 1.1515,
+      "step": 430
+    },
+    {
+      "epoch": 0.0829219900796633,
+      "grad_norm": 0.42780689950352624,
+      "learning_rate": 0.00019853499712007522,
+      "loss": 1.2085,
+      "step": 431
+    },
+    {
+      "epoch": 0.08311438448820081,
+      "grad_norm": 0.3144910521861958,
+      "learning_rate": 0.00019852434951093034,
+      "loss": 1.1487,
+      "step": 432
+    },
+    {
+      "epoch": 0.0833067788967383,
+      "grad_norm": 0.40375939258608123,
+      "learning_rate": 0.00019851366363606346,
+      "loss": 1.2075,
+      "step": 433
+    },
+    {
+      "epoch": 0.08349917330527581,
+      "grad_norm": 0.2530071722324599,
+      "learning_rate": 0.00019850293949962478,
+      "loss": 1.1387,
+      "step": 434
+    },
+    {
+      "epoch": 0.08369156771381332,
+      "grad_norm": 0.28995150953088794,
+      "learning_rate": 0.00019849217710577946,
+      "loss": 1.2065,
+      "step": 435
+    },
+    {
+      "epoch": 0.08388396212235082,
+      "grad_norm": 0.3947416878413457,
+      "learning_rate": 0.00019848137645870747,
+      "loss": 1.1915,
+      "step": 436
+    },
+    {
+      "epoch": 0.08407635653088832,
+      "grad_norm": 0.2822450902075487,
+      "learning_rate": 0.0001984705375626036,
+      "loss": 1.1947,
+      "step": 437
+    },
+    {
+      "epoch": 0.08426875093942582,
+      "grad_norm": 0.35006156835637003,
+      "learning_rate": 0.0001984596604216777,
+      "loss": 1.1896,
+      "step": 438
+    },
+    {
+      "epoch": 0.08446114534796333,
+      "grad_norm": 0.2940332831878667,
+      "learning_rate": 0.0001984487450401542,
+      "loss": 1.2575,
+      "step": 439
+    },
+    {
+      "epoch": 0.08465353975650082,
+      "grad_norm": 0.2750477917770306,
+      "learning_rate": 0.00019843779142227256,
+      "loss": 1.1838,
+      "step": 440
+    },
+    {
+      "epoch": 0.08484593416503833,
+      "grad_norm": 0.44365220556057505,
+      "learning_rate": 0.00019842679957228704,
+      "loss": 1.1236,
+      "step": 441
+    },
+    {
+      "epoch": 0.08503832857357584,
+      "grad_norm": 0.35323008737584244,
+      "learning_rate": 0.00019841576949446675,
+      "loss": 1.2028,
+      "step": 442
+    },
+    {
+      "epoch": 0.08523072298211333,
+      "grad_norm": 0.3065207179608405,
+      "learning_rate": 0.0001984047011930956,
+      "loss": 1.1576,
+      "step": 443
+    },
+    {
+      "epoch": 0.08542311739065084,
+      "grad_norm": 0.337203149463998,
+      "learning_rate": 0.00019839359467247242,
+      "loss": 1.0646,
+      "step": 444
+    },
+    {
+      "epoch": 0.08561551179918833,
+      "grad_norm": 0.3237553476549483,
+      "learning_rate": 0.0001983824499369109,
+      "loss": 1.1432,
+      "step": 445
+    },
+    {
+      "epoch": 0.08580790620772584,
+      "grad_norm": 0.38487437473167235,
+      "learning_rate": 0.00019837126699073947,
+      "loss": 1.2304,
+      "step": 446
+    },
+    {
+      "epoch": 0.08600030061626333,
+      "grad_norm": 0.38638435218236167,
+      "learning_rate": 0.00019836004583830146,
+      "loss": 1.1217,
+      "step": 447
+    },
+    {
+      "epoch": 0.08619269502480084,
+      "grad_norm": 0.4032554737655603,
+      "learning_rate": 0.00019834878648395505,
+      "loss": 1.2968,
+      "step": 448
+    },
+    {
+      "epoch": 0.08638508943333835,
+      "grad_norm": 0.3920953735181232,
+      "learning_rate": 0.00019833748893207325,
+      "loss": 1.1408,
+      "step": 449
+    },
+    {
+      "epoch": 0.08657748384187584,
+      "grad_norm": 0.3308035192207735,
+      "learning_rate": 0.00019832615318704389,
+      "loss": 1.2439,
+      "step": 450
+    },
+    {
+      "epoch": 0.08676987825041335,
+      "grad_norm": 0.3162262954393446,
+      "learning_rate": 0.00019831477925326963,
+      "loss": 1.1697,
+      "step": 451
+    },
+    {
+      "epoch": 0.08696227265895085,
+      "grad_norm": 0.33534237222552854,
+      "learning_rate": 0.00019830336713516799,
+      "loss": 1.1795,
+      "step": 452
+    },
+    {
+      "epoch": 0.08715466706748835,
+      "grad_norm": 0.3577028535782865,
+      "learning_rate": 0.00019829191683717133,
+      "loss": 1.1538,
+      "step": 453
+    },
+    {
+      "epoch": 0.08734706147602585,
+      "grad_norm": 0.39848663571777154,
+      "learning_rate": 0.00019828042836372677,
+      "loss": 1.1412,
+      "step": 454
+    },
+    {
+      "epoch": 0.08753945588456336,
+      "grad_norm": 0.3876119411442906,
+      "learning_rate": 0.00019826890171929632,
+      "loss": 1.1539,
+      "step": 455
+    },
+    {
+      "epoch": 0.08773185029310086,
+      "grad_norm": 0.29051765041640587,
+      "learning_rate": 0.00019825733690835679,
+      "loss": 1.0924,
+      "step": 456
+    },
+    {
+      "epoch": 0.08792424470163836,
+      "grad_norm": 0.27391860563289705,
+      "learning_rate": 0.00019824573393539984,
+      "loss": 1.1334,
+      "step": 457
+    },
+    {
+      "epoch": 0.08811663911017587,
+      "grad_norm": 0.33462249053807647,
+      "learning_rate": 0.0001982340928049319,
+      "loss": 1.125,
+      "step": 458
+    },
+    {
+      "epoch": 0.08830903351871336,
+      "grad_norm": 0.25589837840019536,
+      "learning_rate": 0.00019822241352147427,
+      "loss": 1.1796,
+      "step": 459
+    },
+    {
+      "epoch": 0.08850142792725087,
+      "grad_norm": 0.3066481549327959,
+      "learning_rate": 0.00019821069608956307,
+      "loss": 1.2327,
+      "step": 460
+    },
+    {
+      "epoch": 0.08869382233578836,
+      "grad_norm": 0.36509002454342965,
+      "learning_rate": 0.00019819894051374915,
+      "loss": 1.2575,
+      "step": 461
+    },
+    {
+      "epoch": 0.08888621674432587,
+      "grad_norm": 0.5428321934587285,
+      "learning_rate": 0.0001981871467985983,
+      "loss": 1.1804,
+      "step": 462
+    },
+    {
+      "epoch": 0.08907861115286338,
+      "grad_norm": 0.44837883703400355,
+      "learning_rate": 0.00019817531494869105,
+      "loss": 1.2685,
+      "step": 463
+    },
+    {
+      "epoch": 0.08927100556140087,
+      "grad_norm": 0.42577528440150764,
+      "learning_rate": 0.00019816344496862272,
+      "loss": 1.1823,
+      "step": 464
+    },
+    {
+      "epoch": 0.08946339996993838,
+      "grad_norm": 0.5104164491806513,
+      "learning_rate": 0.00019815153686300352,
+      "loss": 1.1065,
+      "step": 465
+    },
+    {
+      "epoch": 0.08965579437847587,
+      "grad_norm": 0.3073359811722275,
+      "learning_rate": 0.0001981395906364584,
+      "loss": 1.2454,
+      "step": 466
+    },
+    {
+      "epoch": 0.08984818878701338,
+      "grad_norm": 0.3164709627212633,
+      "learning_rate": 0.00019812760629362716,
+      "loss": 1.2492,
+      "step": 467
+    },
+    {
+      "epoch": 0.09004058319555087,
+      "grad_norm": 0.3025378925163634,
+      "learning_rate": 0.0001981155838391643,
+      "loss": 1.2217,
+      "step": 468
+    },
+    {
+      "epoch": 0.09023297760408838,
+      "grad_norm": 0.3857186130510256,
+      "learning_rate": 0.00019810352327773935,
+      "loss": 1.1601,
+      "step": 469
+    },
+    {
+      "epoch": 0.09042537201262588,
+      "grad_norm": 0.35805603980179956,
+      "learning_rate": 0.00019809142461403633,
+      "loss": 1.1157,
+      "step": 470
+    },
+    {
+      "epoch": 0.09061776642116338,
+      "grad_norm": 0.36761443970418983,
+      "learning_rate": 0.00019807928785275434,
+      "loss": 1.2144,
+      "step": 471
+    },
+    {
+      "epoch": 0.09081016082970089,
+      "grad_norm": 0.36531426496163355,
+      "learning_rate": 0.0001980671129986071,
+      "loss": 1.0471,
+      "step": 472
+    },
+    {
+      "epoch": 0.09100255523823839,
+      "grad_norm": 0.341026820881479,
+      "learning_rate": 0.0001980549000563232,
+      "loss": 1.1444,
+      "step": 473
+    },
+    {
+      "epoch": 0.0911949496467759,
+      "grad_norm": 0.2916944170383586,
+      "learning_rate": 0.000198042649030646,
+      "loss": 1.1742,
+      "step": 474
+    },
+    {
+      "epoch": 0.09138734405531339,
+      "grad_norm": 0.32111097544336353,
+      "learning_rate": 0.00019803035992633366,
+      "loss": 1.2338,
+      "step": 475
+    },
+    {
+      "epoch": 0.0915797384638509,
+      "grad_norm": 0.4397000646582326,
+      "learning_rate": 0.00019801803274815917,
+      "loss": 1.13,
+      "step": 476
+    },
+    {
+      "epoch": 0.09177213287238839,
+      "grad_norm": 0.33090615106393406,
+      "learning_rate": 0.00019800566750091016,
+      "loss": 1.164,
+      "step": 477
+    },
+    {
+      "epoch": 0.0919645272809259,
+      "grad_norm": 0.31828284126092543,
+      "learning_rate": 0.00019799326418938924,
+      "loss": 1.1465,
+      "step": 478
+    },
+    {
+      "epoch": 0.0921569216894634,
+      "grad_norm": 0.4274700881284667,
+      "learning_rate": 0.0001979808228184137,
+      "loss": 1.1628,
+      "step": 479
+    },
+    {
+      "epoch": 0.0923493160980009,
+      "grad_norm": 0.3100930905766764,
+      "learning_rate": 0.0001979683433928156,
+      "loss": 1.2112,
+      "step": 480
+    },
+    {
+      "epoch": 0.09254171050653841,
+      "grad_norm": 0.3447305086343119,
+      "learning_rate": 0.0001979558259174418,
+      "loss": 1.1347,
+      "step": 481
+    },
+    {
+      "epoch": 0.0927341049150759,
+      "grad_norm": 0.3096479609034406,
+      "learning_rate": 0.00019794327039715395,
+      "loss": 1.1943,
+      "step": 482
+    },
+    {
+      "epoch": 0.09292649932361341,
+      "grad_norm": 0.3860457764883986,
+      "learning_rate": 0.0001979306768368285,
+      "loss": 1.1265,
+      "step": 483
+    },
+    {
+      "epoch": 0.0931188937321509,
+      "grad_norm": 0.37478427017509097,
+      "learning_rate": 0.0001979180452413566,
+      "loss": 1.1363,
+      "step": 484
+    },
+    {
+      "epoch": 0.09331128814068841,
+      "grad_norm": 0.3416737051733567,
+      "learning_rate": 0.00019790537561564428,
+      "loss": 1.1063,
+      "step": 485
+    },
+    {
+      "epoch": 0.09350368254922592,
+      "grad_norm": 0.33694559127844387,
+      "learning_rate": 0.00019789266796461222,
+      "loss": 1.1177,
+      "step": 486
+    },
+    {
+      "epoch": 0.09369607695776341,
+      "grad_norm": 0.34464276393453,
+      "learning_rate": 0.00019787992229319592,
+      "loss": 1.1136,
+      "step": 487
+    },
+    {
+      "epoch": 0.09388847136630092,
+      "grad_norm": 0.27502450360601083,
+      "learning_rate": 0.00019786713860634567,
+      "loss": 1.1176,
+      "step": 488
+    },
+    {
+      "epoch": 0.09408086577483842,
+      "grad_norm": 0.29014015439601665,
+      "learning_rate": 0.00019785431690902652,
+      "loss": 1.1984,
+      "step": 489
+    },
+    {
+      "epoch": 0.09427326018337592,
+      "grad_norm": 0.27825877470122073,
+      "learning_rate": 0.00019784145720621826,
+      "loss": 1.2074,
+      "step": 490
+    },
+    {
+      "epoch": 0.09446565459191342,
+      "grad_norm": 0.42605752272778963,
+      "learning_rate": 0.00019782855950291542,
+      "loss": 1.2561,
+      "step": 491
+    },
+    {
+      "epoch": 0.09465804900045093,
+      "grad_norm": 0.36320082065467807,
+      "learning_rate": 0.0001978156238041274,
+      "loss": 1.1957,
+      "step": 492
+    },
+    {
+      "epoch": 0.09485044340898843,
+      "grad_norm": 0.3485812315935624,
+      "learning_rate": 0.0001978026501148782,
+      "loss": 1.2829,
+      "step": 493
+    },
+    {
+      "epoch": 0.09504283781752593,
+      "grad_norm": 0.37100459131545566,
+      "learning_rate": 0.00019778963844020665,
+      "loss": 1.2091,
+      "step": 494
+    },
+    {
+      "epoch": 0.09523523222606343,
+      "grad_norm": 0.3254372234617211,
+      "learning_rate": 0.00019777658878516639,
+      "loss": 1.053,
+      "step": 495
+    },
+    {
+      "epoch": 0.09542762663460093,
+      "grad_norm": 0.3382921218252847,
+      "learning_rate": 0.0001977635011548257,
+      "loss": 1.2029,
+      "step": 496
+    },
+    {
+      "epoch": 0.09562002104313844,
+      "grad_norm": 0.33624752210662856,
+      "learning_rate": 0.0001977503755542677,
+      "loss": 1.1296,
+      "step": 497
+    },
+    {
+      "epoch": 0.09581241545167593,
+      "grad_norm": 0.29275302223924876,
+      "learning_rate": 0.00019773721198859022,
+      "loss": 1.1702,
+      "step": 498
+    },
+    {
+      "epoch": 0.09600480986021344,
+      "grad_norm": 0.38733623412226326,
+      "learning_rate": 0.00019772401046290586,
+      "loss": 1.1747,
+      "step": 499
+    },
+    {
+      "epoch": 0.09619720426875093,
+      "grad_norm": 0.32684777724574576,
+      "learning_rate": 0.00019771077098234186,
+      "loss": 1.1797,
+      "step": 500
+    },
+    {
+      "epoch": 0.09638959867728844,
+      "grad_norm": 0.3707423508963498,
+      "learning_rate": 0.00019769749355204032,
+      "loss": 1.2258,
+      "step": 501
+    },
+    {
+      "epoch": 0.09658199308582595,
+      "grad_norm": 0.3035231035025504,
+      "learning_rate": 0.00019768417817715809,
+      "loss": 1.2894,
+      "step": 502
+    },
+    {
+      "epoch": 0.09677438749436344,
+      "grad_norm": 0.3470316433153745,
+      "learning_rate": 0.00019767082486286665,
+      "loss": 1.1925,
+      "step": 503
+    },
+    {
+      "epoch": 0.09696678190290095,
+      "grad_norm": 0.389327656778959,
+      "learning_rate": 0.0001976574336143523,
+      "loss": 1.1069,
+      "step": 504
+    },
+    {
+      "epoch": 0.09715917631143844,
+      "grad_norm": 0.26923590638768,
+      "learning_rate": 0.00019764400443681606,
+      "loss": 1.2889,
+      "step": 505
+    },
+    {
+      "epoch": 0.09735157071997595,
+      "grad_norm": 0.32311857762032764,
+      "learning_rate": 0.00019763053733547366,
+      "loss": 1.2084,
+      "step": 506
+    },
+    {
+      "epoch": 0.09754396512851345,
+      "grad_norm": 0.2721190435899251,
+      "learning_rate": 0.0001976170323155555,
+      "loss": 1.0824,
+      "step": 507
+    },
+    {
+      "epoch": 0.09773635953705095,
+      "grad_norm": 0.2973336404264234,
+      "learning_rate": 0.0001976034893823069,
+      "loss": 1.1494,
+      "step": 508
+    },
+    {
+      "epoch": 0.09792875394558846,
+      "grad_norm": 0.29675934264749443,
+      "learning_rate": 0.0001975899085409876,
+      "loss": 1.1447,
+      "step": 509
+    },
+    {
+      "epoch": 0.09812114835412596,
+      "grad_norm": 0.32625734126550004,
+      "learning_rate": 0.00019757628979687246,
+      "loss": 1.2385,
+      "step": 510
+    },
+    {
+      "epoch": 0.09831354276266346,
+      "grad_norm": 0.33597951594219033,
+      "learning_rate": 0.0001975626331552507,
+      "loss": 1.1841,
+      "step": 511
+    },
+    {
+      "epoch": 0.09850593717120096,
+      "grad_norm": 0.3328618992626306,
+      "learning_rate": 0.00019754893862142643,
+      "loss": 1.1188,
+      "step": 512
+    },
+    {
+      "epoch": 0.09869833157973847,
+      "grad_norm": 0.2718437554018281,
+      "learning_rate": 0.00019753520620071843,
+      "loss": 1.1639,
+      "step": 513
+    },
+    {
+      "epoch": 0.09889072598827596,
+      "grad_norm": 0.3453880777439029,
+      "learning_rate": 0.0001975214358984603,
+      "loss": 1.1144,
+      "step": 514
+    },
+    {
+      "epoch": 0.09908312039681347,
+      "grad_norm": 0.37262831962128673,
+      "learning_rate": 0.00019750762772000014,
+      "loss": 1.2572,
+      "step": 515
+    },
+    {
+      "epoch": 0.09927551480535098,
+      "grad_norm": 0.3214984257975183,
+      "learning_rate": 0.000197493781670701,
+      "loss": 1.1261,
+      "step": 516
+    },
+    {
+      "epoch": 0.09946790921388847,
+      "grad_norm": 0.2928973383182354,
+      "learning_rate": 0.00019747989775594044,
+      "loss": 1.2205,
+      "step": 517
+    },
+    {
+      "epoch": 0.09966030362242598,
+      "grad_norm": 0.284566191322515,
+      "learning_rate": 0.0001974659759811109,
+      "loss": 1.2398,
+      "step": 518
+    },
+    {
+      "epoch": 0.09985269803096347,
+      "grad_norm": 0.4321206416751429,
+      "learning_rate": 0.00019745201635161936,
+      "loss": 1.2418,
+      "step": 519
+    },
+    {
+      "epoch": 0.10004509243950098,
+      "grad_norm": 0.35565386440658175,
+      "learning_rate": 0.00019743801887288763,
+      "loss": 1.1612,
+      "step": 520
+    },
+    {
+      "epoch": 0.10023748684803847,
+      "grad_norm": 0.37881504999540855,
+      "learning_rate": 0.0001974239835503521,
+      "loss": 1.2318,
+      "step": 521
+    },
+    {
+      "epoch": 0.10042988125657598,
+      "grad_norm": 0.2980709180313274,
+      "learning_rate": 0.00019740991038946404,
+      "loss": 1.1698,
+      "step": 522
+    },
+    {
+      "epoch": 0.10062227566511349,
+      "grad_norm": 0.27872265876760344,
+      "learning_rate": 0.0001973957993956892,
+      "loss": 1.2114,
+      "step": 523
+    },
+    {
+      "epoch": 0.10081467007365098,
+      "grad_norm": 0.34523418964020874,
+      "learning_rate": 0.00019738165057450816,
+      "loss": 1.1714,
+      "step": 524
+    },
+    {
+      "epoch": 0.10100706448218849,
+      "grad_norm": 0.2934710661951568,
+      "learning_rate": 0.00019736746393141617,
+      "loss": 1.1414,
+      "step": 525
+    },
+    {
+      "epoch": 0.10119945889072599,
+      "grad_norm": 0.36189734017239755,
+      "learning_rate": 0.00019735323947192316,
+      "loss": 1.1254,
+      "step": 526
+    },
+    {
+      "epoch": 0.10139185329926349,
+      "grad_norm": 0.30398986072937584,
+      "learning_rate": 0.00019733897720155375,
+      "loss": 1.1203,
+      "step": 527
+    },
+    {
+      "epoch": 0.10158424770780099,
+      "grad_norm": 0.31062230964476817,
+      "learning_rate": 0.00019732467712584722,
+      "loss": 1.2721,
+      "step": 528
+    },
+    {
+      "epoch": 0.1017766421163385,
+      "grad_norm": 0.38555634475514416,
+      "learning_rate": 0.0001973103392503576,
+      "loss": 1.1986,
+      "step": 529
+    },
+    {
+      "epoch": 0.10196903652487599,
+      "grad_norm": 0.30085992478038925,
+      "learning_rate": 0.00019729596358065345,
+      "loss": 1.1228,
+      "step": 530
+    },
+    {
+      "epoch": 0.1021614309334135,
+      "grad_norm": 0.36534253873049205,
+      "learning_rate": 0.00019728155012231825,
+      "loss": 1.2129,
+      "step": 531
+    },
+    {
+      "epoch": 0.102353825341951,
+      "grad_norm": 0.33291541665040153,
+      "learning_rate": 0.00019726709888094992,
+      "loss": 1.1892,
+      "step": 532
+    },
+    {
+      "epoch": 0.1025462197504885,
+      "grad_norm": 0.31412881601436915,
+      "learning_rate": 0.0001972526098621612,
+      "loss": 1.1958,
+      "step": 533
+    },
+    {
+      "epoch": 0.102738614159026,
+      "grad_norm": 0.32320825465482145,
+      "learning_rate": 0.00019723808307157948,
+      "loss": 1.2098,
+      "step": 534
+    },
+    {
+      "epoch": 0.1029310085675635,
+      "grad_norm": 0.304629295771607,
+      "learning_rate": 0.00019722351851484676,
+      "loss": 1.203,
+      "step": 535
+    },
+    {
+      "epoch": 0.10312340297610101,
+      "grad_norm": 0.25790056795714794,
+      "learning_rate": 0.00019720891619761974,
+      "loss": 1.1093,
+      "step": 536
+    },
+    {
+      "epoch": 0.1033157973846385,
+      "grad_norm": 0.31643287793830754,
+      "learning_rate": 0.0001971942761255698,
+      "loss": 1.1214,
+      "step": 537
+    },
+    {
+      "epoch": 0.10350819179317601,
+      "grad_norm": 0.28974558037543857,
+      "learning_rate": 0.00019717959830438302,
+      "loss": 1.2127,
+      "step": 538
+    },
+    {
+      "epoch": 0.10370058620171352,
+      "grad_norm": 0.3166167560445395,
+      "learning_rate": 0.00019716488273976003,
+      "loss": 1.1031,
+      "step": 539
+    },
+    {
+      "epoch": 0.10389298061025101,
+      "grad_norm": 0.3114566629739266,
+      "learning_rate": 0.0001971501294374162,
+      "loss": 1.2519,
+      "step": 540
+    },
+    {
+      "epoch": 0.10408537501878852,
+      "grad_norm": 0.36586917893261767,
+      "learning_rate": 0.00019713533840308157,
+      "loss": 1.1902,
+      "step": 541
+    },
+    {
+      "epoch": 0.10427776942732601,
+      "grad_norm": 0.3068846886376433,
+      "learning_rate": 0.00019712050964250082,
+      "loss": 1.1257,
+      "step": 542
+    },
+    {
+      "epoch": 0.10447016383586352,
+      "grad_norm": 0.2996806635344164,
+      "learning_rate": 0.00019710564316143323,
+      "loss": 1.1542,
+      "step": 543
+    },
+    {
+      "epoch": 0.10466255824440102,
+      "grad_norm": 0.3361726600173983,
+      "learning_rate": 0.00019709073896565275,
+      "loss": 1.1917,
+      "step": 544
+    },
+    {
+      "epoch": 0.10485495265293852,
+      "grad_norm": 0.2661540790748322,
+      "learning_rate": 0.00019707579706094807,
+      "loss": 1.1324,
+      "step": 545
+    },
+    {
+      "epoch": 0.10504734706147603,
+      "grad_norm": 0.29974300350194294,
+      "learning_rate": 0.0001970608174531224,
+      "loss": 1.16,
+      "step": 546
+    },
+    {
+      "epoch": 0.10523974147001353,
+      "grad_norm": 0.40286568801141787,
+      "learning_rate": 0.0001970458001479937,
+      "loss": 1.1623,
+      "step": 547
+    },
+    {
+      "epoch": 0.10543213587855103,
+      "grad_norm": 0.3963793410502087,
+      "learning_rate": 0.00019703074515139445,
+      "loss": 1.1597,
+      "step": 548
+    },
+    {
+      "epoch": 0.10562453028708853,
+      "grad_norm": 0.30374122274040766,
+      "learning_rate": 0.00019701565246917183,
+      "loss": 1.1892,
+      "step": 549
+    },
+    {
+      "epoch": 0.10581692469562604,
+      "grad_norm": 0.34744259445162395,
+      "learning_rate": 0.00019700052210718777,
+      "loss": 1.0312,
+      "step": 550
+    },
+    {
+      "epoch": 0.10600931910416353,
+      "grad_norm": 0.237042250882054,
+      "learning_rate": 0.00019698535407131862,
+      "loss": 1.108,
+      "step": 551
+    },
+    {
+      "epoch": 0.10620171351270104,
+      "grad_norm": 0.295802633280917,
+      "learning_rate": 0.00019697014836745553,
+      "loss": 1.068,
+      "step": 552
+    },
+    {
+      "epoch": 0.10639410792123855,
+      "grad_norm": 0.35869197263976094,
+      "learning_rate": 0.00019695490500150418,
+      "loss": 1.125,
+      "step": 553
+    },
+    {
+      "epoch": 0.10658650232977604,
+      "grad_norm": 0.36099269578333354,
+      "learning_rate": 0.00019693962397938496,
+      "loss": 1.1599,
+      "step": 554
+    },
+    {
+      "epoch": 0.10677889673831355,
+      "grad_norm": 0.3698722082113688,
+      "learning_rate": 0.00019692430530703282,
+      "loss": 1.2795,
+      "step": 555
+    },
+    {
+      "epoch": 0.10697129114685104,
+      "grad_norm": 0.33108172990111345,
+      "learning_rate": 0.00019690894899039734,
+      "loss": 1.2043,
+      "step": 556
+    },
+    {
+      "epoch": 0.10716368555538855,
+      "grad_norm": 0.2739477760052005,
+      "learning_rate": 0.00019689355503544275,
+      "loss": 1.2031,
+      "step": 557
+    },
+    {
+      "epoch": 0.10735607996392604,
+      "grad_norm": 0.2573316852826777,
+      "learning_rate": 0.0001968781234481479,
+      "loss": 1.1078,
+      "step": 558
+    },
+    {
+      "epoch": 0.10754847437246355,
+      "grad_norm": 0.396727594032576,
+      "learning_rate": 0.00019686265423450624,
+      "loss": 1.1868,
+      "step": 559
+    },
+    {
+      "epoch": 0.10774086878100106,
+      "grad_norm": 0.31138858864954155,
+      "learning_rate": 0.00019684714740052583,
+      "loss": 1.2124,
+      "step": 560
+    },
+    {
+      "epoch": 0.10793326318953855,
+      "grad_norm": 0.33749954783121383,
+      "learning_rate": 0.00019683160295222934,
+      "loss": 1.0093,
+      "step": 561
+    },
+    {
+      "epoch": 0.10812565759807606,
+      "grad_norm": 0.30779628174143353,
+      "learning_rate": 0.00019681602089565402,
+      "loss": 1.1597,
+      "step": 562
+    },
+    {
+      "epoch": 0.10831805200661355,
+      "grad_norm": 0.3269447594143558,
+      "learning_rate": 0.0001968004012368518,
+      "loss": 1.1063,
+      "step": 563
+    },
+    {
+      "epoch": 0.10851044641515106,
+      "grad_norm": 0.35023284288802403,
+      "learning_rate": 0.0001967847439818892,
+      "loss": 1.2306,
+      "step": 564
+    },
+    {
+      "epoch": 0.10870284082368856,
+      "grad_norm": 0.3425290709701458,
+      "learning_rate": 0.00019676904913684727,
+      "loss": 1.1684,
+      "step": 565
+    },
+    {
+      "epoch": 0.10889523523222606,
+      "grad_norm": 0.36001084910187836,
+      "learning_rate": 0.0001967533167078217,
+      "loss": 1.1282,
+      "step": 566
+    },
+    {
+      "epoch": 0.10908762964076356,
+      "grad_norm": 0.303046847655797,
+      "learning_rate": 0.00019673754670092284,
+      "loss": 1.1031,
+      "step": 567
+    },
+    {
+      "epoch": 0.10928002404930107,
+      "grad_norm": 0.28181341140110044,
+      "learning_rate": 0.00019672173912227553,
+      "loss": 1.2519,
+      "step": 568
+    },
+    {
+      "epoch": 0.10947241845783857,
+      "grad_norm": 0.3041260099407497,
+      "learning_rate": 0.0001967058939780193,
+      "loss": 1.2757,
+      "step": 569
+    },
+    {
+      "epoch": 0.10966481286637607,
+      "grad_norm": 0.2671824279958905,
+      "learning_rate": 0.00019669001127430816,
+      "loss": 1.2552,
+      "step": 570
+    },
+    {
+      "epoch": 0.10985720727491358,
+      "grad_norm": 0.2687800355481347,
+      "learning_rate": 0.00019667409101731083,
+      "loss": 1.1103,
+      "step": 571
+    },
+    {
+      "epoch": 0.11004960168345107,
+      "grad_norm": 0.3036411946803846,
+      "learning_rate": 0.0001966581332132105,
+      "loss": 1.1307,
+      "step": 572
+    },
+    {
+      "epoch": 0.11024199609198858,
+      "grad_norm": 0.29180830681947484,
+      "learning_rate": 0.00019664213786820502,
+      "loss": 1.1453,
+      "step": 573
+    },
+    {
+      "epoch": 0.11043439050052607,
+      "grad_norm": 0.30847019352865407,
+      "learning_rate": 0.00019662610498850683,
+      "loss": 1.2159,
+      "step": 574
+    },
+    {
+      "epoch": 0.11062678490906358,
+      "grad_norm": 0.3073157926850266,
+      "learning_rate": 0.00019661003458034285,
+      "loss": 1.1318,
+      "step": 575
+    },
+    {
+      "epoch": 0.11081917931760109,
+      "grad_norm": 0.2880595028819985,
+      "learning_rate": 0.0001965939266499547,
+      "loss": 1.1767,
+      "step": 576
+    },
+    {
+      "epoch": 0.11101157372613858,
+      "grad_norm": 0.3128324784161832,
+      "learning_rate": 0.00019657778120359847,
+      "loss": 1.1344,
+      "step": 577
+    },
+    {
+      "epoch": 0.11120396813467609,
+      "grad_norm": 0.34450148018842414,
+      "learning_rate": 0.0001965615982475449,
+      "loss": 1.1217,
+      "step": 578
+    },
+    {
+      "epoch": 0.11139636254321358,
+      "grad_norm": 0.37021048106097726,
+      "learning_rate": 0.00019654537778807923,
+      "loss": 1.1722,
+      "step": 579
+    },
+    {
+      "epoch": 0.11158875695175109,
+      "grad_norm": 0.3221029145705702,
+      "learning_rate": 0.00019652911983150136,
+      "loss": 1.2321,
+      "step": 580
+    },
+    {
+      "epoch": 0.11178115136028859,
+      "grad_norm": 0.29322374483179536,
+      "learning_rate": 0.0001965128243841256,
+      "loss": 1.1758,
+      "step": 581
+    },
+    {
+      "epoch": 0.1119735457688261,
+      "grad_norm": 0.3949924836157765,
+      "learning_rate": 0.00019649649145228102,
+      "loss": 1.2365,
+      "step": 582
+    },
+    {
+      "epoch": 0.1121659401773636,
+      "grad_norm": 0.5050059547697262,
+      "learning_rate": 0.00019648012104231106,
+      "loss": 1.1964,
+      "step": 583
+    },
+    {
+      "epoch": 0.1123583345859011,
+      "grad_norm": 0.35969810608901864,
+      "learning_rate": 0.00019646371316057383,
+      "loss": 1.134,
+      "step": 584
+    },
+    {
+      "epoch": 0.1125507289944386,
+      "grad_norm": 0.27437844541592243,
+      "learning_rate": 0.00019644726781344195,
+      "loss": 1.1395,
+      "step": 585
+    },
+    {
+      "epoch": 0.1127431234029761,
+      "grad_norm": 0.3310514422937076,
+      "learning_rate": 0.0001964307850073026,
+      "loss": 1.2128,
+      "step": 586
+    },
+    {
+      "epoch": 0.1129355178115136,
+      "grad_norm": 0.32108661804456523,
+      "learning_rate": 0.00019641426474855758,
+      "loss": 1.1611,
+      "step": 587
+    },
+    {
+      "epoch": 0.1131279122200511,
+      "grad_norm": 0.294995552222423,
+      "learning_rate": 0.00019639770704362307,
+      "loss": 1.2312,
+      "step": 588
+    },
+    {
+      "epoch": 0.11332030662858861,
+      "grad_norm": 0.2956510671437227,
+      "learning_rate": 0.00019638111189892993,
+      "loss": 1.111,
+      "step": 589
+    },
+    {
+      "epoch": 0.11351270103712612,
+      "grad_norm": 0.31541196367129015,
+      "learning_rate": 0.00019636447932092353,
+      "loss": 1.1694,
+      "step": 590
+    },
+    {
+      "epoch": 0.11370509544566361,
+      "grad_norm": 0.3513652794131551,
+      "learning_rate": 0.0001963478093160638,
+      "loss": 1.1923,
+      "step": 591
+    },
+    {
+      "epoch": 0.11389748985420112,
+      "grad_norm": 0.2766388671785139,
+      "learning_rate": 0.00019633110189082512,
+      "loss": 1.2838,
+      "step": 592
+    },
+    {
+      "epoch": 0.11408988426273861,
+      "grad_norm": 0.24850099821971594,
+      "learning_rate": 0.0001963143570516965,
+      "loss": 1.1378,
+      "step": 593
+    },
+    {
+      "epoch": 0.11428227867127612,
+      "grad_norm": 0.2517680673946296,
+      "learning_rate": 0.00019629757480518143,
+      "loss": 1.1493,
+      "step": 594
+    },
+    {
+      "epoch": 0.11447467307981361,
+      "grad_norm": 0.33744656655923166,
+      "learning_rate": 0.00019628075515779796,
+      "loss": 1.1672,
+      "step": 595
+    },
+    {
+      "epoch": 0.11466706748835112,
+      "grad_norm": 0.3949598405631972,
+      "learning_rate": 0.0001962638981160786,
+      "loss": 1.2305,
+      "step": 596
+    },
+    {
+      "epoch": 0.11485946189688861,
+      "grad_norm": 0.33116982611180057,
+      "learning_rate": 0.00019624700368657045,
+      "loss": 1.1266,
+      "step": 597
+    },
+    {
+      "epoch": 0.11505185630542612,
+      "grad_norm": 0.31943827662655716,
+      "learning_rate": 0.00019623007187583515,
+      "loss": 1.1764,
+      "step": 598
+    },
+    {
+      "epoch": 0.11524425071396363,
+      "grad_norm": 0.30767537653897864,
+      "learning_rate": 0.0001962131026904488,
+      "loss": 1.0916,
+      "step": 599
+    },
+    {
+      "epoch": 0.11543664512250112,
+      "grad_norm": 0.2967990341230032,
+      "learning_rate": 0.000196196096137002,
+      "loss": 1.1939,
+      "step": 600
+    },
+    {
+      "epoch": 0.11562903953103863,
+      "grad_norm": 0.3323779023336841,
+      "learning_rate": 0.00019617905222209996,
+      "loss": 1.2729,
+      "step": 601
+    },
+    {
+      "epoch": 0.11582143393957613,
+      "grad_norm": 0.2892387381043252,
+      "learning_rate": 0.00019616197095236227,
+      "loss": 1.1489,
+      "step": 602
+    },
+    {
+      "epoch": 0.11601382834811363,
+      "grad_norm": 0.3309218093655699,
+      "learning_rate": 0.00019614485233442316,
+      "loss": 1.2117,
+      "step": 603
+    },
+    {
+      "epoch": 0.11620622275665113,
+      "grad_norm": 0.3418986857756515,
+      "learning_rate": 0.00019612769637493126,
+      "loss": 1.1968,
+      "step": 604
+    },
+    {
+      "epoch": 0.11639861716518864,
+      "grad_norm": 0.3075216824816062,
+      "learning_rate": 0.0001961105030805498,
+      "loss": 1.0761,
+      "step": 605
+    },
+    {
+      "epoch": 0.11659101157372614,
+      "grad_norm": 0.3786085859243023,
+      "learning_rate": 0.00019609327245795642,
+      "loss": 1.1554,
+      "step": 606
+    },
+    {
+      "epoch": 0.11678340598226364,
+      "grad_norm": 0.4261893933120443,
+      "learning_rate": 0.00019607600451384326,
+      "loss": 1.1284,
+      "step": 607
+    },
+    {
+      "epoch": 0.11697580039080115,
+      "grad_norm": 0.4337950865116119,
+      "learning_rate": 0.00019605869925491706,
+      "loss": 1.1337,
+      "step": 608
+    },
+    {
+      "epoch": 0.11716819479933864,
+      "grad_norm": 0.2589664058294049,
+      "learning_rate": 0.00019604135668789896,
+      "loss": 1.2624,
+      "step": 609
+    },
+    {
+      "epoch": 0.11736058920787615,
+      "grad_norm": 0.3125044485167045,
+      "learning_rate": 0.00019602397681952463,
+      "loss": 1.1433,
+      "step": 610
+    },
+    {
+      "epoch": 0.11755298361641364,
+      "grad_norm": 0.3132992996532646,
+      "learning_rate": 0.00019600655965654412,
+      "loss": 1.1506,
+      "step": 611
+    },
+    {
+      "epoch": 0.11774537802495115,
+      "grad_norm": 0.3475342829246708,
+      "learning_rate": 0.00019598910520572218,
+      "loss": 1.1024,
+      "step": 612
+    },
+    {
+      "epoch": 0.11793777243348866,
+      "grad_norm": 0.3159575506311138,
+      "learning_rate": 0.00019597161347383784,
+      "loss": 1.1873,
+      "step": 613
+    },
+    {
+      "epoch": 0.11813016684202615,
+      "grad_norm": 0.4186559897421048,
+      "learning_rate": 0.00019595408446768472,
+      "loss": 1.1431,
+      "step": 614
+    },
+    {
+      "epoch": 0.11832256125056366,
+      "grad_norm": 0.4317825982999988,
+      "learning_rate": 0.00019593651819407084,
+      "loss": 1.1347,
+      "step": 615
+    },
+    {
+      "epoch": 0.11851495565910115,
+      "grad_norm": 0.2566486445476346,
+      "learning_rate": 0.0001959189146598188,
+      "loss": 1.211,
+      "step": 616
+    },
+    {
+      "epoch": 0.11870735006763866,
+      "grad_norm": 0.3586375479841208,
+      "learning_rate": 0.00019590127387176555,
+      "loss": 1.1637,
+      "step": 617
+    },
+    {
+      "epoch": 0.11889974447617616,
+      "grad_norm": 0.25657719466271134,
+      "learning_rate": 0.00019588359583676263,
+      "loss": 1.1877,
+      "step": 618
+    },
+    {
+      "epoch": 0.11909213888471366,
+      "grad_norm": 0.31393933921644207,
+      "learning_rate": 0.00019586588056167594,
+      "loss": 1.1385,
+      "step": 619
+    },
+    {
+      "epoch": 0.11928453329325117,
+      "grad_norm": 0.3697433194772041,
+      "learning_rate": 0.0001958481280533859,
+      "loss": 1.1338,
+      "step": 620
+    },
+    {
+      "epoch": 0.11947692770178867,
+      "grad_norm": 0.3487066366400267,
+      "learning_rate": 0.0001958303383187874,
+      "loss": 1.2398,
+      "step": 621
+    },
+    {
+      "epoch": 0.11966932211032617,
+      "grad_norm": 0.3608575916161132,
+      "learning_rate": 0.00019581251136478972,
+      "loss": 1.0889,
+      "step": 622
+    },
+    {
+      "epoch": 0.11986171651886367,
+      "grad_norm": 0.3116804686595518,
+      "learning_rate": 0.00019579464719831667,
+      "loss": 1.1536,
+      "step": 623
+    },
+    {
+      "epoch": 0.12005411092740118,
+      "grad_norm": 0.2934512229988681,
+      "learning_rate": 0.00019577674582630652,
+      "loss": 1.1699,
+      "step": 624
+    },
+    {
+      "epoch": 0.12024650533593867,
+      "grad_norm": 0.3360185158130375,
+      "learning_rate": 0.0001957588072557119,
+      "loss": 1.1948,
+      "step": 625
+    },
+    {
+      "epoch": 0.12043889974447618,
+      "grad_norm": 0.2714327580965586,
+      "learning_rate": 0.0001957408314935,
+      "loss": 1.213,
+      "step": 626
+    },
+    {
+      "epoch": 0.12063129415301367,
+      "grad_norm": 0.3423978741298657,
+      "learning_rate": 0.00019572281854665234,
+      "loss": 0.9985,
+      "step": 627
+    },
+    {
+      "epoch": 0.12082368856155118,
+      "grad_norm": 0.31276118810174985,
+      "learning_rate": 0.00019570476842216498,
+      "loss": 1.2471,
+      "step": 628
+    },
+    {
+      "epoch": 0.12101608297008869,
+      "grad_norm": 0.40709474999843404,
+      "learning_rate": 0.00019568668112704838,
+      "loss": 1.2399,
+      "step": 629
+    },
+    {
+      "epoch": 0.12120847737862618,
+      "grad_norm": 0.3063475832949149,
+      "learning_rate": 0.00019566855666832743,
+      "loss": 1.1734,
+      "step": 630
+    },
+    {
+      "epoch": 0.12140087178716369,
+      "grad_norm": 0.45420817847126915,
+      "learning_rate": 0.00019565039505304145,
+      "loss": 1.2326,
+      "step": 631
+    },
+    {
+      "epoch": 0.12159326619570118,
+      "grad_norm": 0.3263169978266844,
+      "learning_rate": 0.0001956321962882442,
+      "loss": 1.1265,
+      "step": 632
+    },
+    {
+      "epoch": 0.12178566060423869,
+      "grad_norm": 0.41784651430485065,
+      "learning_rate": 0.0001956139603810039,
+      "loss": 1.218,
+      "step": 633
+    },
+    {
+      "epoch": 0.12197805501277618,
+      "grad_norm": 0.2917325675032357,
+      "learning_rate": 0.00019559568733840314,
+      "loss": 1.2036,
+      "step": 634
+    },
+    {
+      "epoch": 0.12217044942131369,
+      "grad_norm": 0.31376546729484533,
+      "learning_rate": 0.00019557737716753896,
+      "loss": 1.1802,
+      "step": 635
+    },
+    {
+      "epoch": 0.1223628438298512,
+      "grad_norm": 0.3082219008645867,
+      "learning_rate": 0.00019555902987552283,
+      "loss": 1.2094,
+      "step": 636
+    },
+    {
+      "epoch": 0.1225552382383887,
+      "grad_norm": 0.39551471644425645,
+      "learning_rate": 0.00019554064546948064,
+      "loss": 1.1182,
+      "step": 637
+    },
+    {
+      "epoch": 0.1227476326469262,
+      "grad_norm": 0.31166773309736256,
+      "learning_rate": 0.0001955222239565526,
+      "loss": 1.2583,
+      "step": 638
+    },
+    {
+      "epoch": 0.1229400270554637,
+      "grad_norm": 0.29134632055315,
+      "learning_rate": 0.00019550376534389357,
+      "loss": 1.1864,
+      "step": 639
+    },
+    {
+      "epoch": 0.1231324214640012,
+      "grad_norm": 0.27423356896383116,
+      "learning_rate": 0.00019548526963867252,
+      "loss": 1.0974,
+      "step": 640
+    },
+    {
+      "epoch": 0.1233248158725387,
+      "grad_norm": 0.3271884183989453,
+      "learning_rate": 0.00019546673684807302,
+      "loss": 1.1921,
+      "step": 641
+    },
+    {
+      "epoch": 0.1235172102810762,
+      "grad_norm": 0.3109567121565924,
+      "learning_rate": 0.000195448166979293,
+      "loss": 1.1702,
+      "step": 642
+    },
+    {
+      "epoch": 0.12370960468961371,
+      "grad_norm": 0.2667237796300738,
+      "learning_rate": 0.00019542956003954477,
+      "loss": 1.137,
+      "step": 643
+    },
+    {
+      "epoch": 0.12390199909815121,
+      "grad_norm": 0.44318329785914495,
+      "learning_rate": 0.00019541091603605506,
+      "loss": 1.1498,
+      "step": 644
+    },
+    {
+      "epoch": 0.12409439350668872,
+      "grad_norm": 0.31705928261829636,
+      "learning_rate": 0.000195392234976065,
+      "loss": 1.2049,
+      "step": 645
+    },
+    {
+      "epoch": 0.12428678791522621,
+      "grad_norm": 0.2822955984588036,
+      "learning_rate": 0.00019537351686683003,
+      "loss": 1.2138,
+      "step": 646
+    },
+    {
+      "epoch": 0.12447918232376372,
+      "grad_norm": 0.4290138786224797,
+      "learning_rate": 0.00019535476171562012,
+      "loss": 1.0149,
+      "step": 647
+    },
+    {
+      "epoch": 0.12467157673230121,
+      "grad_norm": 0.3224504244698334,
+      "learning_rate": 0.00019533596952971954,
+      "loss": 1.1165,
+      "step": 648
+    },
+    {
+      "epoch": 0.12486397114083872,
+      "grad_norm": 0.35160525473344234,
+      "learning_rate": 0.00019531714031642696,
+      "loss": 1.1436,
+      "step": 649
+    },
+    {
+      "epoch": 0.1250563655493762,
+      "grad_norm": 0.38812851323647324,
+      "learning_rate": 0.0001952982740830554,
+      "loss": 1.1181,
+      "step": 650
+    },
+    {
+      "epoch": 0.12524875995791374,
+      "grad_norm": 0.2837084654576034,
+      "learning_rate": 0.00019527937083693231,
+      "loss": 1.2084,
+      "step": 651
+    },
+    {
+      "epoch": 0.12544115436645123,
+      "grad_norm": 0.2984545559635266,
+      "learning_rate": 0.0001952604305853995,
+      "loss": 1.1595,
+      "step": 652
+    },
+    {
+      "epoch": 0.12563354877498872,
+      "grad_norm": 0.27005753072057015,
+      "learning_rate": 0.00019524145333581317,
+      "loss": 1.1116,
+      "step": 653
+    },
+    {
+      "epoch": 0.12582594318352622,
+      "grad_norm": 0.35325056392352405,
+      "learning_rate": 0.00019522243909554377,
+      "loss": 1.0745,
+      "step": 654
+    },
+    {
+      "epoch": 0.12601833759206374,
+      "grad_norm": 0.32629179491927846,
+      "learning_rate": 0.00019520338787197629,
+      "loss": 1.1742,
+      "step": 655
+    },
+    {
+      "epoch": 0.12621073200060123,
+      "grad_norm": 0.33635967732822475,
+      "learning_rate": 0.00019518429967251,
+      "loss": 1.2089,
+      "step": 656
+    },
+    {
+      "epoch": 0.12640312640913873,
+      "grad_norm": 0.2858003747033703,
+      "learning_rate": 0.00019516517450455853,
+      "loss": 1.2316,
+      "step": 657
+    },
+    {
+      "epoch": 0.12659552081767625,
+      "grad_norm": 0.3409332228422794,
+      "learning_rate": 0.00019514601237554988,
+      "loss": 1.2583,
+      "step": 658
+    },
+    {
+      "epoch": 0.12678791522621374,
+      "grad_norm": 0.4402773693540164,
+      "learning_rate": 0.00019512681329292636,
+      "loss": 1.2327,
+      "step": 659
+    },
+    {
+      "epoch": 0.12698030963475124,
+      "grad_norm": 0.3900372680219423,
+      "learning_rate": 0.00019510757726414472,
+      "loss": 1.2713,
+      "step": 660
+    },
+    {
+      "epoch": 0.12717270404328873,
+      "grad_norm": 0.4246181197809418,
+      "learning_rate": 0.000195088304296676,
+      "loss": 1.1656,
+      "step": 661
+    },
+    {
+      "epoch": 0.12736509845182625,
+      "grad_norm": 0.31519949057608326,
+      "learning_rate": 0.00019506899439800557,
+      "loss": 1.1441,
+      "step": 662
+    },
+    {
+      "epoch": 0.12755749286036375,
+      "grad_norm": 0.35574730499498314,
+      "learning_rate": 0.0001950496475756332,
+      "loss": 1.1566,
+      "step": 663
+    },
+    {
+      "epoch": 0.12774988726890124,
+      "grad_norm": 0.4397955821474415,
+      "learning_rate": 0.000195030263837073,
+      "loss": 1.1647,
+      "step": 664
+    },
+    {
+      "epoch": 0.12794228167743876,
+      "grad_norm": 0.32612620971918455,
+      "learning_rate": 0.00019501084318985335,
+      "loss": 1.1546,
+      "step": 665
+    },
+    {
+      "epoch": 0.12813467608597626,
+      "grad_norm": 0.26818914422362755,
+      "learning_rate": 0.000194991385641517,
+      "loss": 1.2254,
+      "step": 666
+    },
+    {
+      "epoch": 0.12832707049451375,
+      "grad_norm": 0.2658810327746161,
+      "learning_rate": 0.00019497189119962105,
+      "loss": 1.191,
+      "step": 667
+    },
+    {
+      "epoch": 0.12851946490305124,
+      "grad_norm": 0.3179415955938622,
+      "learning_rate": 0.00019495235987173693,
+      "loss": 1.1735,
+      "step": 668
+    },
+    {
+      "epoch": 0.12871185931158877,
+      "grad_norm": 0.30240306152889646,
+      "learning_rate": 0.00019493279166545038,
+      "loss": 1.0778,
+      "step": 669
+    },
+    {
+      "epoch": 0.12890425372012626,
+      "grad_norm": 0.5190729683524917,
+      "learning_rate": 0.00019491318658836142,
+      "loss": 1.0943,
+      "step": 670
+    },
+    {
+      "epoch": 0.12909664812866375,
+      "grad_norm": 0.34474065356984984,
+      "learning_rate": 0.0001948935446480845,
+      "loss": 1.1523,
+      "step": 671
+    },
+    {
+      "epoch": 0.12928904253720128,
+      "grad_norm": 0.3722781054637813,
+      "learning_rate": 0.0001948738658522483,
+      "loss": 1.1652,
+      "step": 672
+    },
+    {
+      "epoch": 0.12948143694573877,
+      "grad_norm": 0.36891802402904267,
+      "learning_rate": 0.00019485415020849582,
+      "loss": 1.1259,
+      "step": 673
+    },
+    {
+      "epoch": 0.12967383135427626,
+      "grad_norm": 0.2963423815169123,
+      "learning_rate": 0.0001948343977244844,
+      "loss": 1.2379,
+      "step": 674
+    },
+    {
+      "epoch": 0.12986622576281376,
+      "grad_norm": 0.3373200157829875,
+      "learning_rate": 0.00019481460840788573,
+      "loss": 1.1569,
+      "step": 675
+    },
+    {
+      "epoch": 0.13005862017135128,
+      "grad_norm": 0.2519551023236678,
+      "learning_rate": 0.00019479478226638565,
+      "loss": 1.187,
+      "step": 676
+    },
+    {
+      "epoch": 0.13025101457988877,
+      "grad_norm": 0.3597424367996489,
+      "learning_rate": 0.0001947749193076845,
+      "loss": 1.1901,
+      "step": 677
+    },
+    {
+      "epoch": 0.13044340898842627,
+      "grad_norm": 0.25548776537217965,
+      "learning_rate": 0.00019475501953949672,
+      "loss": 1.2037,
+      "step": 678
+    },
+    {
+      "epoch": 0.13063580339696376,
+      "grad_norm": 0.3216285266643529,
+      "learning_rate": 0.00019473508296955126,
+      "loss": 1.1111,
+      "step": 679
+    },
+    {
+      "epoch": 0.13082819780550128,
+      "grad_norm": 0.28443045254975347,
+      "learning_rate": 0.0001947151096055912,
+      "loss": 1.1015,
+      "step": 680
+    },
+    {
+      "epoch": 0.13102059221403878,
+      "grad_norm": 0.3023199152678026,
+      "learning_rate": 0.00019469509945537397,
+      "loss": 1.1786,
+      "step": 681
+    },
+    {
+      "epoch": 0.13121298662257627,
+      "grad_norm": 0.2601298485674686,
+      "learning_rate": 0.00019467505252667127,
+      "loss": 1.1047,
+      "step": 682
+    },
+    {
+      "epoch": 0.1314053810311138,
+      "grad_norm": 0.2677502741737002,
+      "learning_rate": 0.0001946549688272691,
+      "loss": 1.1468,
+      "step": 683
+    },
+    {
+      "epoch": 0.1315977754396513,
+      "grad_norm": 0.2967509161856476,
+      "learning_rate": 0.0001946348483649678,
+      "loss": 1.1309,
+      "step": 684
+    },
+    {
+      "epoch": 0.13179016984818878,
+      "grad_norm": 0.3527712165405441,
+      "learning_rate": 0.0001946146911475818,
+      "loss": 1.166,
+      "step": 685
+    },
+    {
+      "epoch": 0.13198256425672628,
+      "grad_norm": 0.3550092050671919,
+      "learning_rate": 0.00019459449718294008,
+      "loss": 1.1108,
+      "step": 686
+    },
+    {
+      "epoch": 0.1321749586652638,
+      "grad_norm": 0.2724946168937622,
+      "learning_rate": 0.00019457426647888563,
+      "loss": 1.1882,
+      "step": 687
+    },
+    {
+      "epoch": 0.1323673530738013,
+      "grad_norm": 0.33903673003058277,
+      "learning_rate": 0.00019455399904327585,
+      "loss": 1.2428,
+      "step": 688
+    },
+    {
+      "epoch": 0.13255974748233879,
+      "grad_norm": 0.32867101987082065,
+      "learning_rate": 0.00019453369488398238,
+      "loss": 1.1282,
+      "step": 689
+    },
+    {
+      "epoch": 0.1327521418908763,
+      "grad_norm": 0.233291729397467,
+      "learning_rate": 0.00019451335400889116,
+      "loss": 1.2656,
+      "step": 690
+    },
+    {
+      "epoch": 0.1329445362994138,
+      "grad_norm": 0.2632137189381223,
+      "learning_rate": 0.00019449297642590228,
+      "loss": 1.1921,
+      "step": 691
+    },
+    {
+      "epoch": 0.1331369307079513,
+      "grad_norm": 0.3454510077016807,
+      "learning_rate": 0.00019447256214293025,
+      "loss": 1.2308,
+      "step": 692
+    },
+    {
+      "epoch": 0.1333293251164888,
+      "grad_norm": 0.4226280707358105,
+      "learning_rate": 0.00019445211116790362,
+      "loss": 1.1702,
+      "step": 693
+    },
+    {
+      "epoch": 0.1335217195250263,
+      "grad_norm": 0.29712061464132294,
+      "learning_rate": 0.00019443162350876546,
+      "loss": 1.2282,
+      "step": 694
+    },
+    {
+      "epoch": 0.1337141139335638,
+      "grad_norm": 0.389303579201947,
+      "learning_rate": 0.0001944110991734728,
+      "loss": 1.15,
+      "step": 695
+    },
+    {
+      "epoch": 0.1339065083421013,
+      "grad_norm": 0.26901192469594143,
+      "learning_rate": 0.00019439053816999716,
+      "loss": 1.1921,
+      "step": 696
+    },
+    {
+      "epoch": 0.13409890275063882,
+      "grad_norm": 0.25690282098352424,
+      "learning_rate": 0.00019436994050632414,
+      "loss": 1.1227,
+      "step": 697
+    },
+    {
+      "epoch": 0.13429129715917631,
+      "grad_norm": 0.32254783607969956,
+      "learning_rate": 0.00019434930619045368,
+      "loss": 1.1631,
+      "step": 698
+    },
+    {
+      "epoch": 0.1344836915677138,
+      "grad_norm": 0.25296267034031994,
+      "learning_rate": 0.00019432863523039987,
+      "loss": 1.1099,
+      "step": 699
+    },
+    {
+      "epoch": 0.1346760859762513,
+      "grad_norm": 0.31051863649002237,
+      "learning_rate": 0.00019430792763419107,
+      "loss": 1.1477,
+      "step": 700
+    },
+    {
+      "epoch": 0.13486848038478882,
+      "grad_norm": 0.3935551030369805,
+      "learning_rate": 0.00019428718340986988,
+      "loss": 1.1295,
+      "step": 701
+    },
+    {
+      "epoch": 0.13506087479332632,
+      "grad_norm": 0.3334321571060536,
+      "learning_rate": 0.00019426640256549313,
+      "loss": 1.1976,
+      "step": 702
+    },
+    {
+      "epoch": 0.1352532692018638,
+      "grad_norm": 0.3671116292804533,
+      "learning_rate": 0.00019424558510913186,
+      "loss": 1.2061,
+      "step": 703
+    },
+    {
+      "epoch": 0.13544566361040133,
+      "grad_norm": 0.31295502063785,
+      "learning_rate": 0.00019422473104887134,
+      "loss": 1.1298,
+      "step": 704
+    },
+    {
+      "epoch": 0.13563805801893883,
+      "grad_norm": 0.30832991842179336,
+      "learning_rate": 0.000194203840392811,
+      "loss": 1.2077,
+      "step": 705
+    },
+    {
+      "epoch": 0.13583045242747632,
+      "grad_norm": 0.3441228111516066,
+      "learning_rate": 0.00019418291314906457,
+      "loss": 1.139,
+      "step": 706
+    },
+    {
+      "epoch": 0.13602284683601382,
+      "grad_norm": 0.29980184796539466,
+      "learning_rate": 0.00019416194932576,
+      "loss": 1.1696,
+      "step": 707
+    },
+    {
+      "epoch": 0.13621524124455134,
+      "grad_norm": 0.37352958842959627,
+      "learning_rate": 0.00019414094893103928,
+      "loss": 1.2371,
+      "step": 708
+    },
+    {
+      "epoch": 0.13640763565308883,
+      "grad_norm": 0.2993593241103526,
+      "learning_rate": 0.00019411991197305879,
+      "loss": 1.1402,
+      "step": 709
+    },
+    {
+      "epoch": 0.13660003006162633,
+      "grad_norm": 0.30263817504707197,
+      "learning_rate": 0.00019409883845998904,
+      "loss": 1.2487,
+      "step": 710
+    },
+    {
+      "epoch": 0.13679242447016385,
+      "grad_norm": 0.30046276364974284,
+      "learning_rate": 0.00019407772840001472,
+      "loss": 1.1926,
+      "step": 711
+    },
+    {
+      "epoch": 0.13698481887870134,
+      "grad_norm": 0.3057603250342118,
+      "learning_rate": 0.00019405658180133477,
+      "loss": 1.2414,
+      "step": 712
+    },
+    {
+      "epoch": 0.13717721328723884,
+      "grad_norm": 0.3296249849908774,
+      "learning_rate": 0.00019403539867216224,
+      "loss": 1.0804,
+      "step": 713
+    },
+    {
+      "epoch": 0.13736960769577633,
+      "grad_norm": 0.2865861175683603,
+      "learning_rate": 0.00019401417902072446,
+      "loss": 1.1768,
+      "step": 714
+    },
+    {
+      "epoch": 0.13756200210431385,
+      "grad_norm": 0.33154873450288447,
+      "learning_rate": 0.00019399292285526284,
+      "loss": 0.9976,
+      "step": 715
+    },
+    {
+      "epoch": 0.13775439651285135,
+      "grad_norm": 0.3074844848363132,
+      "learning_rate": 0.00019397163018403308,
+      "loss": 1.2146,
+      "step": 716
+    },
+    {
+      "epoch": 0.13794679092138884,
+      "grad_norm": 0.42828725391083877,
+      "learning_rate": 0.00019395030101530502,
+      "loss": 1.2077,
+      "step": 717
+    },
+    {
+      "epoch": 0.13813918532992636,
+      "grad_norm": 0.31963625158099845,
+      "learning_rate": 0.0001939289353573626,
+      "loss": 1.217,
+      "step": 718
+    },
+    {
+      "epoch": 0.13833157973846386,
+      "grad_norm": 0.27900395869668665,
+      "learning_rate": 0.00019390753321850404,
+      "loss": 1.1168,
+      "step": 719
+    },
+    {
+      "epoch": 0.13852397414700135,
+      "grad_norm": 0.3663090546371539,
+      "learning_rate": 0.00019388609460704168,
+      "loss": 1.1639,
+      "step": 720
+    },
+    {
+      "epoch": 0.13871636855553884,
+      "grad_norm": 0.31428494460147827,
+      "learning_rate": 0.000193864619531302,
+      "loss": 1.1102,
+      "step": 721
+    },
+    {
+      "epoch": 0.13890876296407637,
+      "grad_norm": 0.33792751647033115,
+      "learning_rate": 0.00019384310799962573,
+      "loss": 1.2313,
+      "step": 722
+    },
+    {
+      "epoch": 0.13910115737261386,
+      "grad_norm": 0.28551842776909775,
+      "learning_rate": 0.00019382156002036764,
+      "loss": 1.0712,
+      "step": 723
+    },
+    {
+      "epoch": 0.13929355178115135,
+      "grad_norm": 0.341504059854706,
+      "learning_rate": 0.00019379997560189675,
+      "loss": 1.1658,
+      "step": 724
+    },
+    {
+      "epoch": 0.13948594618968888,
+      "grad_norm": 0.35086663643353283,
+      "learning_rate": 0.0001937783547525962,
+      "loss": 1.1607,
+      "step": 725
+    },
+    {
+      "epoch": 0.13967834059822637,
+      "grad_norm": 0.28325989689285075,
+      "learning_rate": 0.00019375669748086327,
+      "loss": 1.1789,
+      "step": 726
+    },
+    {
+      "epoch": 0.13987073500676386,
+      "grad_norm": 0.33350804230008896,
+      "learning_rate": 0.00019373500379510938,
+      "loss": 1.0801,
+      "step": 727
+    },
+    {
+      "epoch": 0.14006312941530136,
+      "grad_norm": 0.32123906832371024,
+      "learning_rate": 0.00019371327370376016,
+      "loss": 1.0821,
+      "step": 728
+    },
+    {
+      "epoch": 0.14025552382383888,
+      "grad_norm": 0.28028449090200813,
+      "learning_rate": 0.00019369150721525527,
+      "loss": 1.1797,
+      "step": 729
+    },
+    {
+      "epoch": 0.14044791823237637,
+      "grad_norm": 0.3075175308222164,
+      "learning_rate": 0.0001936697043380486,
+      "loss": 1.0829,
+      "step": 730
+    },
+    {
+      "epoch": 0.14064031264091387,
+      "grad_norm": 0.3162620198508448,
+      "learning_rate": 0.00019364786508060808,
+      "loss": 1.158,
+      "step": 731
+    },
+    {
+      "epoch": 0.1408327070494514,
+      "grad_norm": 0.41615403042261184,
+      "learning_rate": 0.0001936259894514159,
+      "loss": 1.2405,
+      "step": 732
+    },
+    {
+      "epoch": 0.14102510145798888,
+      "grad_norm": 0.29028201429154643,
+      "learning_rate": 0.00019360407745896827,
+      "loss": 1.1307,
+      "step": 733
+    },
+    {
+      "epoch": 0.14121749586652638,
+      "grad_norm": 0.3093500865259683,
+      "learning_rate": 0.00019358212911177556,
+      "loss": 1.1043,
+      "step": 734
+    },
+    {
+      "epoch": 0.14140989027506387,
+      "grad_norm": 0.29262454000494764,
+      "learning_rate": 0.0001935601444183622,
+      "loss": 1.2763,
+      "step": 735
+    },
+    {
+      "epoch": 0.1416022846836014,
+      "grad_norm": 0.3386241705717509,
+      "learning_rate": 0.0001935381233872669,
+      "loss": 1.1684,
+      "step": 736
+    },
+    {
+      "epoch": 0.1417946790921389,
+      "grad_norm": 0.30740449007229154,
+      "learning_rate": 0.00019351606602704228,
+      "loss": 1.0948,
+      "step": 737
+    },
+    {
+      "epoch": 0.14198707350067638,
+      "grad_norm": 0.33655278237879394,
+      "learning_rate": 0.0001934939723462552,
+      "loss": 1.1573,
+      "step": 738
+    },
+    {
+      "epoch": 0.14217946790921387,
+      "grad_norm": 0.26829855266766456,
+      "learning_rate": 0.00019347184235348662,
+      "loss": 1.1148,
+      "step": 739
+    },
+    {
+      "epoch": 0.1423718623177514,
+      "grad_norm": 0.40228640762909235,
+      "learning_rate": 0.00019344967605733153,
+      "loss": 1.0593,
+      "step": 740
+    },
+    {
+      "epoch": 0.1425642567262889,
+      "grad_norm": 0.3036716964034863,
+      "learning_rate": 0.0001934274734663991,
+      "loss": 1.2155,
+      "step": 741
+    },
+    {
+      "epoch": 0.14275665113482638,
+      "grad_norm": 0.33668244178189094,
+      "learning_rate": 0.00019340523458931253,
+      "loss": 1.1548,
+      "step": 742
+    },
+    {
+      "epoch": 0.1429490455433639,
+      "grad_norm": 0.3056566736084597,
+      "learning_rate": 0.00019338295943470914,
+      "loss": 0.9979,
+      "step": 743
+    },
+    {
+      "epoch": 0.1431414399519014,
+      "grad_norm": 0.3658077888999858,
+      "learning_rate": 0.00019336064801124035,
+      "loss": 1.2117,
+      "step": 744
+    },
+    {
+      "epoch": 0.1433338343604389,
+      "grad_norm": 0.23031350279498447,
+      "learning_rate": 0.0001933383003275717,
+      "loss": 1.1979,
+      "step": 745
+    },
+    {
+      "epoch": 0.1435262287689764,
+      "grad_norm": 0.28872191286584314,
+      "learning_rate": 0.0001933159163923827,
+      "loss": 1.1149,
+      "step": 746
+    },
+    {
+      "epoch": 0.1437186231775139,
+      "grad_norm": 0.2881579427563588,
+      "learning_rate": 0.00019329349621436708,
+      "loss": 1.1209,
+      "step": 747
+    },
+    {
+      "epoch": 0.1439110175860514,
+      "grad_norm": 0.29406143250864686,
+      "learning_rate": 0.00019327103980223254,
+      "loss": 1.136,
+      "step": 748
+    },
+    {
+      "epoch": 0.1441034119945889,
+      "grad_norm": 0.23922482328152583,
+      "learning_rate": 0.0001932485471647009,
+      "loss": 1.0635,
+      "step": 749
+    },
+    {
+      "epoch": 0.14429580640312642,
+      "grad_norm": 0.4006690705666417,
+      "learning_rate": 0.00019322601831050804,
+      "loss": 1.0788,
+      "step": 750
+    },
+    {
+      "epoch": 0.1444882008116639,
+      "grad_norm": 0.24385411707755578,
+      "learning_rate": 0.00019320345324840395,
+      "loss": 1.2207,
+      "step": 751
+    },
+    {
+      "epoch": 0.1446805952202014,
+      "grad_norm": 0.2973876722404917,
+      "learning_rate": 0.00019318085198715256,
+      "loss": 1.1887,
+      "step": 752
+    },
+    {
+      "epoch": 0.1448729896287389,
+      "grad_norm": 0.3161938501268713,
+      "learning_rate": 0.000193158214535532,
+      "loss": 1.1368,
+      "step": 753
+    },
+    {
+      "epoch": 0.14506538403727642,
+      "grad_norm": 0.32930645037368317,
+      "learning_rate": 0.00019313554090233436,
+      "loss": 1.113,
+      "step": 754
+    },
+    {
+      "epoch": 0.14525777844581392,
+      "grad_norm": 0.33772301031487795,
+      "learning_rate": 0.00019311283109636583,
+      "loss": 1.1531,
+      "step": 755
+    },
+    {
+      "epoch": 0.1454501728543514,
+      "grad_norm": 0.3030071615648059,
+      "learning_rate": 0.00019309008512644667,
+      "loss": 1.2325,
+      "step": 756
+    },
+    {
+      "epoch": 0.14564256726288893,
+      "grad_norm": 0.29485903432025434,
+      "learning_rate": 0.0001930673030014111,
+      "loss": 1.0776,
+      "step": 757
+    },
+    {
+      "epoch": 0.14583496167142643,
+      "grad_norm": 0.3342344262001743,
+      "learning_rate": 0.00019304448473010747,
+      "loss": 1.0979,
+      "step": 758
+    },
+    {
+      "epoch": 0.14602735607996392,
+      "grad_norm": 0.3028861389284532,
+      "learning_rate": 0.00019302163032139814,
+      "loss": 1.0541,
+      "step": 759
+    },
+    {
+      "epoch": 0.14621975048850142,
+      "grad_norm": 0.3698971219635232,
+      "learning_rate": 0.00019299873978415947,
+      "loss": 1.2273,
+      "step": 760
+    },
+    {
+      "epoch": 0.14641214489703894,
+      "grad_norm": 0.38111657029746016,
+      "learning_rate": 0.00019297581312728186,
+      "loss": 1.0848,
+      "step": 761
+    },
+    {
+      "epoch": 0.14660453930557643,
+      "grad_norm": 0.3025925274883586,
+      "learning_rate": 0.0001929528503596698,
+      "loss": 1.0944,
+      "step": 762
+    },
+    {
+      "epoch": 0.14679693371411393,
+      "grad_norm": 0.3663075568500179,
+      "learning_rate": 0.00019292985149024178,
+      "loss": 1.1709,
+      "step": 763
+    },
+    {
+      "epoch": 0.14698932812265145,
+      "grad_norm": 0.2983274522938622,
+      "learning_rate": 0.00019290681652793027,
+      "loss": 1.0795,
+      "step": 764
+    },
+    {
+      "epoch": 0.14718172253118894,
+      "grad_norm": 0.3553278650475889,
+      "learning_rate": 0.0001928837454816818,
+      "loss": 1.1076,
+      "step": 765
+    },
+    {
+      "epoch": 0.14737411693972644,
+      "grad_norm": 0.26420511967690036,
+      "learning_rate": 0.00019286063836045685,
+      "loss": 1.24,
+      "step": 766
+    },
+    {
+      "epoch": 0.14756651134826393,
+      "grad_norm": 0.38557410354088034,
+      "learning_rate": 0.00019283749517322999,
+      "loss": 1.1849,
+      "step": 767
+    },
+    {
+      "epoch": 0.14775890575680145,
+      "grad_norm": 0.3733684238269839,
+      "learning_rate": 0.00019281431592898978,
+      "loss": 1.0788,
+      "step": 768
+    },
+    {
+      "epoch": 0.14795130016533894,
+      "grad_norm": 0.3514037785406932,
+      "learning_rate": 0.0001927911006367388,
+      "loss": 1.1839,
+      "step": 769
+    },
+    {
+      "epoch": 0.14814369457387644,
+      "grad_norm": 0.41333244132631397,
+      "learning_rate": 0.0001927678493054935,
+      "loss": 1.0934,
+      "step": 770
+    },
+    {
+      "epoch": 0.14833608898241396,
+      "grad_norm": 0.28091746417692043,
+      "learning_rate": 0.00019274456194428454,
+      "loss": 1.2135,
+      "step": 771
+    },
+    {
+      "epoch": 0.14852848339095145,
+      "grad_norm": 0.33204022967467484,
+      "learning_rate": 0.0001927212385621564,
+      "loss": 1.2235,
+      "step": 772
+    },
+    {
+      "epoch": 0.14872087779948895,
+      "grad_norm": 0.2821537355021002,
+      "learning_rate": 0.00019269787916816763,
+      "loss": 1.0979,
+      "step": 773
+    },
+    {
+      "epoch": 0.14891327220802644,
+      "grad_norm": 0.324873916493364,
+      "learning_rate": 0.00019267448377139075,
+      "loss": 1.0753,
+      "step": 774
+    },
+    {
+      "epoch": 0.14910566661656396,
+      "grad_norm": 0.33646229080038254,
+      "learning_rate": 0.00019265105238091228,
+      "loss": 1.1231,
+      "step": 775
+    },
+    {
+      "epoch": 0.14929806102510146,
+      "grad_norm": 0.30687158590591035,
+      "learning_rate": 0.00019262758500583263,
+      "loss": 1.1744,
+      "step": 776
+    },
+    {
+      "epoch": 0.14949045543363895,
+      "grad_norm": 0.27220475967262675,
+      "learning_rate": 0.00019260408165526637,
+      "loss": 1.1274,
+      "step": 777
+    },
+    {
+      "epoch": 0.14968284984217647,
+      "grad_norm": 0.4015295724178326,
+      "learning_rate": 0.00019258054233834183,
+      "loss": 1.1402,
+      "step": 778
+    },
+    {
+      "epoch": 0.14987524425071397,
+      "grad_norm": 0.38173730816352625,
+      "learning_rate": 0.00019255696706420148,
+      "loss": 1.1096,
+      "step": 779
+    },
+    {
+      "epoch": 0.15006763865925146,
+      "grad_norm": 0.36746617850090224,
+      "learning_rate": 0.00019253335584200164,
+      "loss": 1.1574,
+      "step": 780
+    },
+    {
+      "epoch": 0.15026003306778896,
+      "grad_norm": 0.39411538039792743,
+      "learning_rate": 0.00019250970868091267,
+      "loss": 1.1671,
+      "step": 781
+    },
+    {
+      "epoch": 0.15045242747632648,
+      "grad_norm": 0.2777342437180254,
+      "learning_rate": 0.00019248602559011882,
+      "loss": 1.1785,
+      "step": 782
+    },
+    {
+      "epoch": 0.15064482188486397,
+      "grad_norm": 0.29772234080795185,
+      "learning_rate": 0.00019246230657881834,
+      "loss": 1.1733,
+      "step": 783
+    },
+    {
+      "epoch": 0.15083721629340147,
+      "grad_norm": 0.3734372712775911,
+      "learning_rate": 0.00019243855165622343,
+      "loss": 1.1536,
+      "step": 784
+    },
+    {
+      "epoch": 0.151029610701939,
+      "grad_norm": 0.3132685824888845,
+      "learning_rate": 0.00019241476083156026,
+      "loss": 1.1949,
+      "step": 785
+    },
+    {
+      "epoch": 0.15122200511047648,
+      "grad_norm": 0.30393801252644465,
+      "learning_rate": 0.00019239093411406886,
+      "loss": 1.2607,
+      "step": 786
+    },
+    {
+      "epoch": 0.15141439951901398,
+      "grad_norm": 0.3348326895793918,
+      "learning_rate": 0.00019236707151300327,
+      "loss": 1.278,
+      "step": 787
+    },
+    {
+      "epoch": 0.15160679392755147,
+      "grad_norm": 0.32052908291461146,
+      "learning_rate": 0.00019234317303763144,
+      "loss": 1.0974,
+      "step": 788
+    },
+    {
+      "epoch": 0.151799188336089,
+      "grad_norm": 0.2877592148113061,
+      "learning_rate": 0.00019231923869723528,
+      "loss": 1.3036,
+      "step": 789
+    },
+    {
+      "epoch": 0.15199158274462649,
+      "grad_norm": 0.2890470422306886,
+      "learning_rate": 0.0001922952685011106,
+      "loss": 1.0717,
+      "step": 790
+    },
+    {
+      "epoch": 0.15218397715316398,
+      "grad_norm": 0.24426070968325814,
+      "learning_rate": 0.00019227126245856716,
+      "loss": 1.105,
+      "step": 791
+    },
+    {
+      "epoch": 0.1523763715617015,
+      "grad_norm": 0.3547678254358933,
+      "learning_rate": 0.00019224722057892862,
+      "loss": 1.2498,
+      "step": 792
+    },
+    {
+      "epoch": 0.152568765970239,
+      "grad_norm": 0.34701905568691055,
+      "learning_rate": 0.00019222314287153255,
+      "loss": 1.1205,
+      "step": 793
+    },
+    {
+      "epoch": 0.1527611603787765,
+      "grad_norm": 0.255481892215242,
+      "learning_rate": 0.00019219902934573048,
+      "loss": 1.084,
+      "step": 794
+    },
+    {
+      "epoch": 0.15295355478731398,
+      "grad_norm": 0.34174367123324506,
+      "learning_rate": 0.00019217488001088784,
+      "loss": 1.1106,
+      "step": 795
+    },
+    {
+      "epoch": 0.1531459491958515,
+      "grad_norm": 0.27720599753661185,
+      "learning_rate": 0.00019215069487638395,
+      "loss": 1.1615,
+      "step": 796
+    },
+    {
+      "epoch": 0.153338343604389,
+      "grad_norm": 0.29089081360146063,
+      "learning_rate": 0.000192126473951612,
+      "loss": 1.1356,
+      "step": 797
+    },
+    {
+      "epoch": 0.1535307380129265,
+      "grad_norm": 0.2809112708989489,
+      "learning_rate": 0.0001921022172459791,
+      "loss": 1.1027,
+      "step": 798
+    },
+    {
+      "epoch": 0.15372313242146401,
+      "grad_norm": 0.2597355632766002,
+      "learning_rate": 0.00019207792476890641,
+      "loss": 1.0677,
+      "step": 799
+    },
+    {
+      "epoch": 0.1539155268300015,
+      "grad_norm": 0.33815897510144577,
+      "learning_rate": 0.00019205359652982868,
+      "loss": 1.1062,
+      "step": 800
+    },
+    {
+      "epoch": 0.154107921238539,
+      "grad_norm": 0.34276618435363154,
+      "learning_rate": 0.0001920292325381948,
+      "loss": 1.1211,
+      "step": 801
+    },
+    {
+      "epoch": 0.1543003156470765,
+      "grad_norm": 0.29298351083394736,
+      "learning_rate": 0.00019200483280346748,
+      "loss": 1.0546,
+      "step": 802
+    },
+    {
+      "epoch": 0.15449271005561402,
+      "grad_norm": 0.26131048122503264,
+      "learning_rate": 0.00019198039733512326,
+      "loss": 1.1759,
+      "step": 803
+    },
+    {
+      "epoch": 0.1546851044641515,
+      "grad_norm": 0.32613803501487826,
+      "learning_rate": 0.00019195592614265261,
+      "loss": 1.1096,
+      "step": 804
+    },
+    {
+      "epoch": 0.154877498872689,
+      "grad_norm": 0.4607077594257645,
+      "learning_rate": 0.00019193141923555984,
+      "loss": 1.0602,
+      "step": 805
+    },
+    {
+      "epoch": 0.1550698932812265,
+      "grad_norm": 0.44246263863816204,
+      "learning_rate": 0.00019190687662336316,
+      "loss": 1.1464,
+      "step": 806
+    },
+    {
+      "epoch": 0.15526228768976402,
+      "grad_norm": 0.299767950010681,
+      "learning_rate": 0.00019188229831559466,
+      "loss": 1.152,
+      "step": 807
+    },
+    {
+      "epoch": 0.15545468209830152,
+      "grad_norm": 0.31157228663338715,
+      "learning_rate": 0.00019185768432180024,
+      "loss": 1.1159,
+      "step": 808
+    },
+    {
+      "epoch": 0.155647076506839,
+      "grad_norm": 0.31860625960099725,
+      "learning_rate": 0.00019183303465153973,
+      "loss": 1.1988,
+      "step": 809
+    },
+    {
+      "epoch": 0.15583947091537653,
+      "grad_norm": 0.37974976781254716,
+      "learning_rate": 0.00019180834931438673,
+      "loss": 1.1397,
+      "step": 810
+    },
+    {
+      "epoch": 0.15603186532391403,
+      "grad_norm": 0.27548286783184456,
+      "learning_rate": 0.00019178362831992878,
+      "loss": 1.1609,
+      "step": 811
+    },
+    {
+      "epoch": 0.15622425973245152,
+      "grad_norm": 0.3175285807908584,
+      "learning_rate": 0.00019175887167776717,
+      "loss": 1.2606,
+      "step": 812
+    },
+    {
+      "epoch": 0.15641665414098901,
+      "grad_norm": 0.29071790808442327,
+      "learning_rate": 0.0001917340793975172,
+      "loss": 1.0629,
+      "step": 813
+    },
+    {
+      "epoch": 0.15660904854952654,
+      "grad_norm": 0.324071592911566,
+      "learning_rate": 0.0001917092514888078,
+      "loss": 1.1034,
+      "step": 814
+    },
+    {
+      "epoch": 0.15680144295806403,
+      "grad_norm": 0.32818180341941944,
+      "learning_rate": 0.00019168438796128193,
+      "loss": 1.0915,
+      "step": 815
+    },
+    {
+      "epoch": 0.15699383736660152,
+      "grad_norm": 0.3068324250807365,
+      "learning_rate": 0.0001916594888245962,
+      "loss": 1.1302,
+      "step": 816
+    },
+    {
+      "epoch": 0.15718623177513905,
+      "grad_norm": 0.31647388090551803,
+      "learning_rate": 0.00019163455408842124,
+      "loss": 1.0526,
+      "step": 817
+    },
+    {
+      "epoch": 0.15737862618367654,
+      "grad_norm": 0.28980654638138603,
+      "learning_rate": 0.00019160958376244136,
+      "loss": 1.0557,
+      "step": 818
+    },
+    {
+      "epoch": 0.15757102059221403,
+      "grad_norm": 0.32202286568568234,
+      "learning_rate": 0.00019158457785635477,
+      "loss": 1.1345,
+      "step": 819
+    },
+    {
+      "epoch": 0.15776341500075153,
+      "grad_norm": 0.31924262071093423,
+      "learning_rate": 0.00019155953637987346,
+      "loss": 1.1974,
+      "step": 820
+    },
+    {
+      "epoch": 0.15795580940928905,
+      "grad_norm": 0.4223233232381367,
+      "learning_rate": 0.0001915344593427233,
+      "loss": 1.1268,
+      "step": 821
+    },
+    {
+      "epoch": 0.15814820381782654,
+      "grad_norm": 0.27051746451790504,
+      "learning_rate": 0.00019150934675464382,
+      "loss": 1.1094,
+      "step": 822
+    },
+    {
+      "epoch": 0.15834059822636404,
+      "grad_norm": 0.45168762181674427,
+      "learning_rate": 0.00019148419862538858,
+      "loss": 1.1691,
+      "step": 823
+    },
+    {
+      "epoch": 0.15853299263490156,
+      "grad_norm": 0.31748888324701396,
+      "learning_rate": 0.00019145901496472472,
+      "loss": 1.0949,
+      "step": 824
+    },
+    {
+      "epoch": 0.15872538704343905,
+      "grad_norm": 0.310329531497297,
+      "learning_rate": 0.00019143379578243335,
+      "loss": 1.1133,
+      "step": 825
+    },
+    {
+      "epoch": 0.15891778145197655,
+      "grad_norm": 0.2846016287860886,
+      "learning_rate": 0.0001914085410883093,
+      "loss": 1.1618,
+      "step": 826
+    },
+    {
+      "epoch": 0.15911017586051404,
+      "grad_norm": 0.3284237977185717,
+      "learning_rate": 0.00019138325089216118,
+      "loss": 1.1741,
+      "step": 827
+    },
+    {
+      "epoch": 0.15930257026905156,
+      "grad_norm": 0.2975616368168792,
+      "learning_rate": 0.00019135792520381142,
+      "loss": 1.1328,
+      "step": 828
+    },
+    {
+      "epoch": 0.15949496467758906,
+      "grad_norm": 0.26313851589609943,
+      "learning_rate": 0.00019133256403309625,
+      "loss": 1.11,
+      "step": 829
+    },
+    {
+      "epoch": 0.15968735908612655,
+      "grad_norm": 0.33184440012914745,
+      "learning_rate": 0.0001913071673898656,
+      "loss": 1.2214,
+      "step": 830
+    },
+    {
+      "epoch": 0.15987975349466407,
+      "grad_norm": 0.34785475366882723,
+      "learning_rate": 0.0001912817352839833,
+      "loss": 1.1308,
+      "step": 831
+    },
+    {
+      "epoch": 0.16007214790320157,
+      "grad_norm": 0.3129360536748428,
+      "learning_rate": 0.00019125626772532683,
+      "loss": 1.1295,
+      "step": 832
+    },
+    {
+      "epoch": 0.16026454231173906,
+      "grad_norm": 0.32037099108764544,
+      "learning_rate": 0.00019123076472378752,
+      "loss": 1.189,
+      "step": 833
+    },
+    {
+      "epoch": 0.16045693672027656,
+      "grad_norm": 0.3289132196265171,
+      "learning_rate": 0.00019120522628927045,
+      "loss": 1.0475,
+      "step": 834
+    },
+    {
+      "epoch": 0.16064933112881408,
+      "grad_norm": 0.27198972535598354,
+      "learning_rate": 0.00019117965243169445,
+      "loss": 1.0564,
+      "step": 835
+    },
+    {
+      "epoch": 0.16084172553735157,
+      "grad_norm": 0.3198607890298292,
+      "learning_rate": 0.00019115404316099213,
+      "loss": 1.1272,
+      "step": 836
+    },
+    {
+      "epoch": 0.16103411994588906,
+      "grad_norm": 0.2450658439115008,
+      "learning_rate": 0.00019112839848710978,
+      "loss": 1.0965,
+      "step": 837
+    },
+    {
+      "epoch": 0.1612265143544266,
+      "grad_norm": 0.36560766257746974,
+      "learning_rate": 0.00019110271842000755,
+      "loss": 1.1042,
+      "step": 838
+    },
+    {
+      "epoch": 0.16141890876296408,
+      "grad_norm": 0.28244008536255183,
+      "learning_rate": 0.00019107700296965927,
+      "loss": 1.216,
+      "step": 839
+    },
+    {
+      "epoch": 0.16161130317150157,
+      "grad_norm": 0.3797374199336646,
+      "learning_rate": 0.0001910512521460525,
+      "loss": 1.1358,
+      "step": 840
+    },
+    {
+      "epoch": 0.16180369758003907,
+      "grad_norm": 0.2863827331712721,
+      "learning_rate": 0.00019102546595918857,
+      "loss": 1.1834,
+      "step": 841
+    },
+    {
+      "epoch": 0.1619960919885766,
+      "grad_norm": 0.44116663877154816,
+      "learning_rate": 0.00019099964441908257,
+      "loss": 1.1378,
+      "step": 842
+    },
+    {
+      "epoch": 0.16218848639711408,
+      "grad_norm": 0.3182949726587924,
+      "learning_rate": 0.00019097378753576324,
+      "loss": 1.1621,
+      "step": 843
+    },
+    {
+      "epoch": 0.16238088080565158,
+      "grad_norm": 0.29900025684267784,
+      "learning_rate": 0.00019094789531927316,
+      "loss": 1.0422,
+      "step": 844
+    },
+    {
+      "epoch": 0.1625732752141891,
+      "grad_norm": 0.3871378818101089,
+      "learning_rate": 0.00019092196777966848,
+      "loss": 1.1545,
+      "step": 845
+    },
+    {
+      "epoch": 0.1627656696227266,
+      "grad_norm": 0.2894517762882134,
+      "learning_rate": 0.00019089600492701924,
+      "loss": 1.1463,
+      "step": 846
+    },
+    {
+      "epoch": 0.1629580640312641,
+      "grad_norm": 0.34456572840821104,
+      "learning_rate": 0.00019087000677140908,
+      "loss": 1.1074,
+      "step": 847
+    },
+    {
+      "epoch": 0.16315045843980158,
+      "grad_norm": 0.2641773341362384,
+      "learning_rate": 0.00019084397332293538,
+      "loss": 1.1674,
+      "step": 848
+    },
+    {
+      "epoch": 0.1633428528483391,
+      "grad_norm": 0.46450522629757435,
+      "learning_rate": 0.00019081790459170926,
+      "loss": 1.0286,
+      "step": 849
+    },
+    {
+      "epoch": 0.1635352472568766,
+      "grad_norm": 0.27296123065945493,
+      "learning_rate": 0.00019079180058785547,
+      "loss": 1.1109,
+      "step": 850
+    },
+    {
+      "epoch": 0.1637276416654141,
+      "grad_norm": 0.3531101160607227,
+      "learning_rate": 0.00019076566132151252,
+      "loss": 1.1079,
+      "step": 851
+    },
+    {
+      "epoch": 0.1639200360739516,
+      "grad_norm": 0.31451057559755285,
+      "learning_rate": 0.0001907394868028326,
+      "loss": 1.1573,
+      "step": 852
+    },
+    {
+      "epoch": 0.1641124304824891,
+      "grad_norm": 0.26160877730349263,
+      "learning_rate": 0.0001907132770419816,
+      "loss": 1.1183,
+      "step": 853
+    },
+    {
+      "epoch": 0.1643048248910266,
+      "grad_norm": 0.2687391961006805,
+      "learning_rate": 0.0001906870320491391,
+      "loss": 1.1199,
+      "step": 854
+    },
+    {
+      "epoch": 0.1644972192995641,
+      "grad_norm": 0.3580615465083372,
+      "learning_rate": 0.00019066075183449835,
+      "loss": 1.0355,
+      "step": 855
+    },
+    {
+      "epoch": 0.16468961370810162,
+      "grad_norm": 0.3155966375753538,
+      "learning_rate": 0.00019063443640826623,
+      "loss": 1.0485,
+      "step": 856
+    },
+    {
+      "epoch": 0.1648820081166391,
+      "grad_norm": 0.3169001927767128,
+      "learning_rate": 0.0001906080857806634,
+      "loss": 1.2239,
+      "step": 857
+    },
+    {
+      "epoch": 0.1650744025251766,
+      "grad_norm": 0.42209780972140887,
+      "learning_rate": 0.00019058169996192417,
+      "loss": 1.1579,
+      "step": 858
+    },
+    {
+      "epoch": 0.16526679693371413,
+      "grad_norm": 0.3491712473462913,
+      "learning_rate": 0.00019055527896229643,
+      "loss": 1.1395,
+      "step": 859
+    },
+    {
+      "epoch": 0.16545919134225162,
+      "grad_norm": 0.26242186175294313,
+      "learning_rate": 0.00019052882279204178,
+      "loss": 1.2278,
+      "step": 860
+    },
+    {
+      "epoch": 0.16565158575078912,
+      "grad_norm": 0.28129617812226465,
+      "learning_rate": 0.00019050233146143556,
+      "loss": 1.127,
+      "step": 861
+    },
+    {
+      "epoch": 0.1658439801593266,
+      "grad_norm": 0.3579369959705776,
+      "learning_rate": 0.00019047580498076662,
+      "loss": 1.1962,
+      "step": 862
+    },
+    {
+      "epoch": 0.16603637456786413,
+      "grad_norm": 0.2963805299448607,
+      "learning_rate": 0.00019044924336033758,
+      "loss": 1.1414,
+      "step": 863
+    },
+    {
+      "epoch": 0.16622876897640163,
+      "grad_norm": 0.3604448636331328,
+      "learning_rate": 0.0001904226466104647,
+      "loss": 1.0947,
+      "step": 864
+    },
+    {
+      "epoch": 0.16642116338493912,
+      "grad_norm": 0.3041402934601379,
+      "learning_rate": 0.0001903960147414778,
+      "loss": 1.0996,
+      "step": 865
+    },
+    {
+      "epoch": 0.1666135577934766,
+      "grad_norm": 0.3301423609008923,
+      "learning_rate": 0.0001903693477637204,
+      "loss": 1.12,
+      "step": 866
+    },
+    {
+      "epoch": 0.16680595220201413,
+      "grad_norm": 0.34247370181151415,
+      "learning_rate": 0.00019034264568754968,
+      "loss": 1.2024,
+      "step": 867
+    },
+    {
+      "epoch": 0.16699834661055163,
+      "grad_norm": 0.30221363752669755,
+      "learning_rate": 0.00019031590852333635,
+      "loss": 1.1315,
+      "step": 868
+    },
+    {
+      "epoch": 0.16719074101908912,
+      "grad_norm": 0.2861518235079073,
+      "learning_rate": 0.00019028913628146488,
+      "loss": 1.1671,
+      "step": 869
+    },
+    {
+      "epoch": 0.16738313542762664,
+      "grad_norm": 0.27077370903842124,
+      "learning_rate": 0.00019026232897233328,
+      "loss": 1.1831,
+      "step": 870
+    },
+    {
+      "epoch": 0.16757552983616414,
+      "grad_norm": 0.2416018295611357,
+      "learning_rate": 0.0001902354866063532,
+      "loss": 1.1248,
+      "step": 871
+    },
+    {
+      "epoch": 0.16776792424470163,
+      "grad_norm": 0.3659243145736444,
+      "learning_rate": 0.0001902086091939499,
+      "loss": 1.1242,
+      "step": 872
+    },
+    {
+      "epoch": 0.16796031865323913,
+      "grad_norm": 0.3676593771940267,
+      "learning_rate": 0.00019018169674556227,
+      "loss": 1.2159,
+      "step": 873
+    },
+    {
+      "epoch": 0.16815271306177665,
+      "grad_norm": 0.38525812623930245,
+      "learning_rate": 0.0001901547492716428,
+      "loss": 1.0718,
+      "step": 874
+    },
+    {
+      "epoch": 0.16834510747031414,
+      "grad_norm": 0.321411107035241,
+      "learning_rate": 0.00019012776678265756,
+      "loss": 1.1952,
+      "step": 875
+    },
+    {
+      "epoch": 0.16853750187885164,
+      "grad_norm": 0.33528943639128383,
+      "learning_rate": 0.00019010074928908622,
+      "loss": 1.1034,
+      "step": 876
+    },
+    {
+      "epoch": 0.16872989628738916,
+      "grad_norm": 0.3329529201785243,
+      "learning_rate": 0.0001900736968014221,
+      "loss": 1.1608,
+      "step": 877
+    },
+    {
+      "epoch": 0.16892229069592665,
+      "grad_norm": 0.4191008002324991,
+      "learning_rate": 0.0001900466093301721,
+      "loss": 1.2329,
+      "step": 878
+    },
+    {
+      "epoch": 0.16911468510446415,
+      "grad_norm": 0.3069985978622253,
+      "learning_rate": 0.0001900194868858566,
+      "loss": 1.112,
+      "step": 879
+    },
+    {
+      "epoch": 0.16930707951300164,
+      "grad_norm": 0.3264232356759452,
+      "learning_rate": 0.0001899923294790097,
+      "loss": 1.176,
+      "step": 880
+    },
+    {
+      "epoch": 0.16949947392153916,
+      "grad_norm": 0.2852300817259893,
+      "learning_rate": 0.00018996513712017898,
+      "loss": 1.1319,
+      "step": 881
+    },
+    {
+      "epoch": 0.16969186833007666,
+      "grad_norm": 0.3794771500255884,
+      "learning_rate": 0.0001899379098199257,
+      "loss": 1.1355,
+      "step": 882
+    },
+    {
+      "epoch": 0.16988426273861415,
+      "grad_norm": 0.30199788724423193,
+      "learning_rate": 0.0001899106475888246,
+      "loss": 1.1564,
+      "step": 883
+    },
+    {
+      "epoch": 0.17007665714715167,
+      "grad_norm": 0.2587179613264272,
+      "learning_rate": 0.00018988335043746403,
+      "loss": 1.1544,
+      "step": 884
+    },
+    {
+      "epoch": 0.17026905155568917,
+      "grad_norm": 0.24636680815062553,
+      "learning_rate": 0.00018985601837644587,
+      "loss": 1.1943,
+      "step": 885
+    },
+    {
+      "epoch": 0.17046144596422666,
+      "grad_norm": 0.28949769006470716,
+      "learning_rate": 0.00018982865141638558,
+      "loss": 1.119,
+      "step": 886
+    },
+    {
+      "epoch": 0.17065384037276415,
+      "grad_norm": 0.2962030770378514,
+      "learning_rate": 0.00018980124956791215,
+      "loss": 1.1325,
+      "step": 887
+    },
+    {
+      "epoch": 0.17084623478130168,
+      "grad_norm": 0.28474813221509665,
+      "learning_rate": 0.00018977381284166817,
+      "loss": 1.0596,
+      "step": 888
+    },
+    {
+      "epoch": 0.17103862918983917,
+      "grad_norm": 0.377399006385986,
+      "learning_rate": 0.00018974634124830977,
+      "loss": 1.0844,
+      "step": 889
+    },
+    {
+      "epoch": 0.17123102359837666,
+      "grad_norm": 0.2566488000812951,
+      "learning_rate": 0.0001897188347985066,
+      "loss": 1.141,
+      "step": 890
+    },
+    {
+      "epoch": 0.17142341800691419,
+      "grad_norm": 0.2895835063247774,
+      "learning_rate": 0.00018969129350294178,
+      "loss": 1.1985,
+      "step": 891
+    },
+    {
+      "epoch": 0.17161581241545168,
+      "grad_norm": 0.3035037609978652,
+      "learning_rate": 0.00018966371737231207,
+      "loss": 1.1131,
+      "step": 892
+    },
+    {
+      "epoch": 0.17180820682398917,
+      "grad_norm": 0.2587965640604509,
+      "learning_rate": 0.00018963610641732777,
+      "loss": 1.1641,
+      "step": 893
+    },
+    {
+      "epoch": 0.17200060123252667,
+      "grad_norm": 0.28287571253611005,
+      "learning_rate": 0.00018960846064871258,
+      "loss": 1.2225,
+      "step": 894
+    },
+    {
+      "epoch": 0.1721929956410642,
+      "grad_norm": 0.2566382349288544,
+      "learning_rate": 0.00018958078007720385,
+      "loss": 1.1806,
+      "step": 895
+    },
+    {
+      "epoch": 0.17238539004960168,
+      "grad_norm": 0.3362457182888903,
+      "learning_rate": 0.00018955306471355235,
+      "loss": 1.1477,
+      "step": 896
+    },
+    {
+      "epoch": 0.17257778445813918,
+      "grad_norm": 0.32165469122272955,
+      "learning_rate": 0.00018952531456852247,
+      "loss": 1.0985,
+      "step": 897
+    },
+    {
+      "epoch": 0.1727701788666767,
+      "grad_norm": 0.3804335438726372,
+      "learning_rate": 0.00018949752965289197,
+      "loss": 1.1691,
+      "step": 898
+    },
+    {
+      "epoch": 0.1729625732752142,
+      "grad_norm": 0.30636012685614056,
+      "learning_rate": 0.0001894697099774523,
+      "loss": 1.0103,
+      "step": 899
+    },
+    {
+      "epoch": 0.1731549676837517,
+      "grad_norm": 0.27844105821472565,
+      "learning_rate": 0.0001894418555530082,
+      "loss": 1.1434,
+      "step": 900
+    },
+    {
+      "epoch": 0.17334736209228918,
+      "grad_norm": 0.37303786723431925,
+      "learning_rate": 0.000189413966390378,
+      "loss": 1.0954,
+      "step": 901
+    },
+    {
+      "epoch": 0.1735397565008267,
+      "grad_norm": 0.2729300729659083,
+      "learning_rate": 0.00018938604250039363,
+      "loss": 1.165,
+      "step": 902
+    },
+    {
+      "epoch": 0.1737321509093642,
+      "grad_norm": 0.3456435945341589,
+      "learning_rate": 0.00018935808389390033,
+      "loss": 1.1683,
+      "step": 903
+    },
+    {
+      "epoch": 0.1739245453179017,
+      "grad_norm": 0.2975897828170863,
+      "learning_rate": 0.0001893300905817569,
+      "loss": 1.0943,
+      "step": 904
+    },
+    {
+      "epoch": 0.1741169397264392,
+      "grad_norm": 0.275977650169121,
+      "learning_rate": 0.00018930206257483564,
+      "loss": 1.1824,
+      "step": 905
+    },
+    {
+      "epoch": 0.1743093341349767,
+      "grad_norm": 0.3222837564004834,
+      "learning_rate": 0.00018927399988402232,
+      "loss": 1.2206,
+      "step": 906
+    },
+    {
+      "epoch": 0.1745017285435142,
+      "grad_norm": 0.32050967591965934,
+      "learning_rate": 0.00018924590252021614,
+      "loss": 1.1409,
+      "step": 907
+    },
+    {
+      "epoch": 0.1746941229520517,
+      "grad_norm": 0.29730979170161537,
+      "learning_rate": 0.00018921777049432984,
+      "loss": 1.1026,
+      "step": 908
+    },
+    {
+      "epoch": 0.17488651736058922,
+      "grad_norm": 0.38965679795992625,
+      "learning_rate": 0.00018918960381728947,
+      "loss": 1.1554,
+      "step": 909
+    },
+    {
+      "epoch": 0.1750789117691267,
+      "grad_norm": 0.2695704834341497,
+      "learning_rate": 0.00018916140250003474,
+      "loss": 1.0725,
+      "step": 910
+    },
+    {
+      "epoch": 0.1752713061776642,
+      "grad_norm": 0.2921095094340929,
+      "learning_rate": 0.0001891331665535187,
+      "loss": 1.1436,
+      "step": 911
+    },
+    {
+      "epoch": 0.17546370058620173,
+      "grad_norm": 0.28035493258201616,
+      "learning_rate": 0.00018910489598870784,
+      "loss": 1.1653,
+      "step": 912
+    },
+    {
+      "epoch": 0.17565609499473922,
+      "grad_norm": 0.31810075713057345,
+      "learning_rate": 0.00018907659081658214,
+      "loss": 1.1106,
+      "step": 913
+    },
+    {
+      "epoch": 0.17584848940327671,
+      "grad_norm": 0.34009373133269605,
+      "learning_rate": 0.00018904825104813498,
+      "loss": 1.2197,
+      "step": 914
+    },
+    {
+      "epoch": 0.1760408838118142,
+      "grad_norm": 0.33559978944965874,
+      "learning_rate": 0.0001890198766943732,
+      "loss": 1.1783,
+      "step": 915
+    },
+    {
+      "epoch": 0.17623327822035173,
+      "grad_norm": 0.3115413830736144,
+      "learning_rate": 0.00018899146776631712,
+      "loss": 1.1791,
+      "step": 916
+    },
+    {
+      "epoch": 0.17642567262888922,
+      "grad_norm": 0.2943923602246328,
+      "learning_rate": 0.0001889630242750004,
+      "loss": 1.0869,
+      "step": 917
+    },
+    {
+      "epoch": 0.17661806703742672,
+      "grad_norm": 0.28803120777551305,
+      "learning_rate": 0.0001889345462314702,
+      "loss": 1.1188,
+      "step": 918
+    },
+    {
+      "epoch": 0.17681046144596424,
+      "grad_norm": 0.25659535236204284,
+      "learning_rate": 0.000188906033646787,
+      "loss": 1.0719,
+      "step": 919
+    },
+    {
+      "epoch": 0.17700285585450173,
+      "grad_norm": 0.331679780896181,
+      "learning_rate": 0.00018887748653202477,
+      "loss": 1.1102,
+      "step": 920
+    },
+    {
+      "epoch": 0.17719525026303923,
+      "grad_norm": 0.25875343699934933,
+      "learning_rate": 0.00018884890489827096,
+      "loss": 1.1138,
+      "step": 921
+    },
+    {
+      "epoch": 0.17738764467157672,
+      "grad_norm": 0.31183044674758037,
+      "learning_rate": 0.00018882028875662627,
+      "loss": 1.1264,
+      "step": 922
+    },
+    {
+      "epoch": 0.17758003908011424,
+      "grad_norm": 0.29341552756913425,
+      "learning_rate": 0.0001887916381182049,
+      "loss": 1.1726,
+      "step": 923
+    },
+    {
+      "epoch": 0.17777243348865174,
+      "grad_norm": 0.3446672146063943,
+      "learning_rate": 0.00018876295299413443,
+      "loss": 1.1337,
+      "step": 924
+    },
+    {
+      "epoch": 0.17796482789718923,
+      "grad_norm": 0.27291731920987883,
+      "learning_rate": 0.00018873423339555584,
+      "loss": 1.0639,
+      "step": 925
+    },
+    {
+      "epoch": 0.17815722230572675,
+      "grad_norm": 0.35088318536829566,
+      "learning_rate": 0.0001887054793336235,
+      "loss": 1.0908,
+      "step": 926
+    },
+    {
+      "epoch": 0.17834961671426425,
+      "grad_norm": 0.3024956335824073,
+      "learning_rate": 0.0001886766908195051,
+      "loss": 1.0772,
+      "step": 927
+    },
+    {
+      "epoch": 0.17854201112280174,
+      "grad_norm": 0.31363029589079705,
+      "learning_rate": 0.00018864786786438185,
+      "loss": 1.1242,
+      "step": 928
+    },
+    {
+      "epoch": 0.17873440553133924,
+      "grad_norm": 0.2978871762659467,
+      "learning_rate": 0.00018861901047944818,
+      "loss": 1.108,
+      "step": 929
+    },
+    {
+      "epoch": 0.17892679993987676,
+      "grad_norm": 0.385312391673112,
+      "learning_rate": 0.00018859011867591202,
+      "loss": 1.2112,
+      "step": 930
+    },
+    {
+      "epoch": 0.17911919434841425,
+      "grad_norm": 0.3779748655240065,
+      "learning_rate": 0.00018856119246499458,
+      "loss": 1.1026,
+      "step": 931
+    },
+    {
+      "epoch": 0.17931158875695175,
+      "grad_norm": 0.275071436808685,
+      "learning_rate": 0.00018853223185793047,
+      "loss": 1.1871,
+      "step": 932
+    },
+    {
+      "epoch": 0.17950398316548924,
+      "grad_norm": 0.3605707147022138,
+      "learning_rate": 0.00018850323686596764,
+      "loss": 1.2237,
+      "step": 933
+    },
+    {
+      "epoch": 0.17969637757402676,
+      "grad_norm": 0.27307640028901337,
+      "learning_rate": 0.00018847420750036747,
+      "loss": 1.1931,
+      "step": 934
+    },
+    {
+      "epoch": 0.17988877198256426,
+      "grad_norm": 0.2774789154423623,
+      "learning_rate": 0.00018844514377240457,
+      "loss": 1.1712,
+      "step": 935
+    },
+    {
+      "epoch": 0.18008116639110175,
+      "grad_norm": 0.2973511135010746,
+      "learning_rate": 0.000188416045693367,
+      "loss": 1.1914,
+      "step": 936
+    },
+    {
+      "epoch": 0.18027356079963927,
+      "grad_norm": 0.3611544428784761,
+      "learning_rate": 0.0001883869132745561,
+      "loss": 1.1679,
+      "step": 937
+    },
+    {
+      "epoch": 0.18046595520817676,
+      "grad_norm": 0.3095639368120166,
+      "learning_rate": 0.0001883577465272866,
+      "loss": 1.0873,
+      "step": 938
+    },
+    {
+      "epoch": 0.18065834961671426,
+      "grad_norm": 0.27455848950725376,
+      "learning_rate": 0.00018832854546288642,
+      "loss": 1.1294,
+      "step": 939
+    },
+    {
+      "epoch": 0.18085074402525175,
+      "grad_norm": 0.3205017163632537,
+      "learning_rate": 0.00018829931009269705,
+      "loss": 1.1613,
+      "step": 940
+    },
+    {
+      "epoch": 0.18104313843378927,
+      "grad_norm": 0.30501737625062797,
+      "learning_rate": 0.0001882700404280731,
+      "loss": 1.0524,
+      "step": 941
+    },
+    {
+      "epoch": 0.18123553284232677,
+      "grad_norm": 0.34182345612171766,
+      "learning_rate": 0.00018824073648038257,
+      "loss": 1.1838,
+      "step": 942
+    },
+    {
+      "epoch": 0.18142792725086426,
+      "grad_norm": 0.31842117371139816,
+      "learning_rate": 0.0001882113982610068,
+      "loss": 0.9934,
+      "step": 943
+    },
+    {
+      "epoch": 0.18162032165940178,
+      "grad_norm": 0.30823825549781764,
+      "learning_rate": 0.00018818202578134042,
+      "loss": 1.1544,
+      "step": 944
+    },
+    {
+      "epoch": 0.18181271606793928,
+      "grad_norm": 0.3999509421889017,
+      "learning_rate": 0.0001881526190527913,
+      "loss": 1.0364,
+      "step": 945
+    },
+    {
+      "epoch": 0.18200511047647677,
+      "grad_norm": 0.2767317165622152,
+      "learning_rate": 0.00018812317808678076,
+      "loss": 1.1617,
+      "step": 946
+    },
+    {
+      "epoch": 0.18219750488501427,
+      "grad_norm": 0.35609542855148246,
+      "learning_rate": 0.0001880937028947433,
+      "loss": 1.106,
+      "step": 947
+    },
+    {
+      "epoch": 0.1823898992935518,
+      "grad_norm": 0.33916490177667435,
+      "learning_rate": 0.00018806419348812672,
+      "loss": 1.1731,
+      "step": 948
+    },
+    {
+      "epoch": 0.18258229370208928,
+      "grad_norm": 0.320424334406229,
+      "learning_rate": 0.00018803464987839216,
+      "loss": 1.0738,
+      "step": 949
+    },
+    {
+      "epoch": 0.18277468811062678,
+      "grad_norm": 0.3057200921750843,
+      "learning_rate": 0.000188005072077014,
+      "loss": 1.1095,
+      "step": 950
+    },
+    {
+      "epoch": 0.1829670825191643,
+      "grad_norm": 0.26604360333033583,
+      "learning_rate": 0.00018797546009547995,
+      "loss": 1.164,
+      "step": 951
+    },
+    {
+      "epoch": 0.1831594769277018,
+      "grad_norm": 0.28116284631909794,
+      "learning_rate": 0.0001879458139452909,
+      "loss": 1.1259,
+      "step": 952
+    },
+    {
+      "epoch": 0.18335187133623929,
+      "grad_norm": 0.30711206838212346,
+      "learning_rate": 0.00018791613363796118,
+      "loss": 1.2019,
+      "step": 953
+    },
+    {
+      "epoch": 0.18354426574477678,
+      "grad_norm": 0.319453796282546,
+      "learning_rate": 0.00018788641918501814,
+      "loss": 1.1093,
+      "step": 954
+    },
+    {
+      "epoch": 0.1837366601533143,
+      "grad_norm": 0.32916626205461036,
+      "learning_rate": 0.00018785667059800264,
+      "loss": 1.208,
+      "step": 955
+    },
+    {
+      "epoch": 0.1839290545618518,
+      "grad_norm": 0.37459457263543294,
+      "learning_rate": 0.00018782688788846865,
+      "loss": 1.1141,
+      "step": 956
+    },
+    {
+      "epoch": 0.1841214489703893,
+      "grad_norm": 0.4593672811701854,
+      "learning_rate": 0.0001877970710679834,
+      "loss": 1.1838,
+      "step": 957
+    },
+    {
+      "epoch": 0.1843138433789268,
+      "grad_norm": 0.29660875397432696,
+      "learning_rate": 0.0001877672201481275,
+      "loss": 1.1834,
+      "step": 958
+    },
+    {
+      "epoch": 0.1845062377874643,
+      "grad_norm": 0.3207005263565337,
+      "learning_rate": 0.0001877373351404946,
+      "loss": 1.1197,
+      "step": 959
+    },
+    {
+      "epoch": 0.1846986321960018,
+      "grad_norm": 0.28124102221285346,
+      "learning_rate": 0.00018770741605669171,
+      "loss": 1.1413,
+      "step": 960
+    },
+    {
+      "epoch": 0.1848910266045393,
+      "grad_norm": 0.33377097507399034,
+      "learning_rate": 0.0001876774629083391,
+      "loss": 1.159,
+      "step": 961
+    },
+    {
+      "epoch": 0.18508342101307682,
+      "grad_norm": 0.33389954881976147,
+      "learning_rate": 0.00018764747570707019,
+      "loss": 1.1281,
+      "step": 962
+    },
+    {
+      "epoch": 0.1852758154216143,
+      "grad_norm": 0.3159594033179389,
+      "learning_rate": 0.00018761745446453168,
+      "loss": 1.1569,
+      "step": 963
+    },
+    {
+      "epoch": 0.1854682098301518,
+      "grad_norm": 0.3180637936569423,
+      "learning_rate": 0.00018758739919238349,
+      "loss": 1.2511,
+      "step": 964
+    },
+    {
+      "epoch": 0.18566060423868933,
+      "grad_norm": 0.2858832191127742,
+      "learning_rate": 0.00018755730990229868,
+      "loss": 1.0574,
+      "step": 965
+    },
+    {
+      "epoch": 0.18585299864722682,
+      "grad_norm": 0.3175702687502775,
+      "learning_rate": 0.00018752718660596367,
+      "loss": 1.159,
+      "step": 966
+    },
+    {
+      "epoch": 0.1860453930557643,
+      "grad_norm": 0.3010254344666017,
+      "learning_rate": 0.00018749702931507796,
+      "loss": 1.1986,
+      "step": 967
+    },
+    {
+      "epoch": 0.1862377874643018,
+      "grad_norm": 0.342427437750701,
+      "learning_rate": 0.00018746683804135429,
+      "loss": 1.1233,
+      "step": 968
+    },
+    {
+      "epoch": 0.18643018187283933,
+      "grad_norm": 0.32207018715516766,
+      "learning_rate": 0.00018743661279651855,
+      "loss": 1.1991,
+      "step": 969
+    },
+    {
+      "epoch": 0.18662257628137682,
+      "grad_norm": 0.3773847455479765,
+      "learning_rate": 0.00018740635359231,
+      "loss": 1.1778,
+      "step": 970
+    },
+    {
+      "epoch": 0.18681497068991432,
+      "grad_norm": 0.2960118645511741,
+      "learning_rate": 0.00018737606044048086,
+      "loss": 1.0833,
+      "step": 971
+    },
+    {
+      "epoch": 0.18700736509845184,
+      "grad_norm": 0.3757855285007589,
+      "learning_rate": 0.00018734573335279668,
+      "loss": 1.1829,
+      "step": 972
+    },
+    {
+      "epoch": 0.18719975950698933,
+      "grad_norm": 0.3996595936719864,
+      "learning_rate": 0.00018731537234103617,
+      "loss": 1.0772,
+      "step": 973
+    },
+    {
+      "epoch": 0.18739215391552683,
+      "grad_norm": 0.2847051068279051,
+      "learning_rate": 0.00018728497741699117,
+      "loss": 1.1134,
+      "step": 974
+    },
+    {
+      "epoch": 0.18758454832406432,
+      "grad_norm": 0.27549970128215284,
+      "learning_rate": 0.0001872545485924667,
+      "loss": 1.1685,
+      "step": 975
+    },
+    {
+      "epoch": 0.18777694273260184,
+      "grad_norm": 0.33484321399459477,
+      "learning_rate": 0.00018722408587928103,
+      "loss": 1.1473,
+      "step": 976
+    },
+    {
+      "epoch": 0.18796933714113934,
+      "grad_norm": 0.35316226832778747,
+      "learning_rate": 0.00018719358928926546,
+      "loss": 1.173,
+      "step": 977
+    },
+    {
+      "epoch": 0.18816173154967683,
+      "grad_norm": 0.3462465232862975,
+      "learning_rate": 0.00018716305883426454,
+      "loss": 1.0195,
+      "step": 978
+    },
+    {
+      "epoch": 0.18835412595821435,
+      "grad_norm": 0.25745398549898946,
+      "learning_rate": 0.00018713249452613598,
+      "loss": 1.1412,
+      "step": 979
+    },
+    {
+      "epoch": 0.18854652036675185,
+      "grad_norm": 0.42279872297244486,
+      "learning_rate": 0.00018710189637675056,
+      "loss": 1.1855,
+      "step": 980
+    },
+    {
+      "epoch": 0.18873891477528934,
+      "grad_norm": 0.3185714083210854,
+      "learning_rate": 0.00018707126439799224,
+      "loss": 1.1447,
+      "step": 981
+    },
+    {
+      "epoch": 0.18893130918382683,
+      "grad_norm": 0.29351698832238055,
+      "learning_rate": 0.00018704059860175817,
+      "loss": 1.1501,
+      "step": 982
+    },
+    {
+      "epoch": 0.18912370359236436,
+      "grad_norm": 0.4140759363951703,
+      "learning_rate": 0.00018700989899995857,
+      "loss": 1.1193,
+      "step": 983
+    },
+    {
+      "epoch": 0.18931609800090185,
+      "grad_norm": 0.3307058322568098,
+      "learning_rate": 0.00018697916560451682,
+      "loss": 1.0535,
+      "step": 984
+    },
+    {
+      "epoch": 0.18950849240943934,
+      "grad_norm": 0.3263000631533168,
+      "learning_rate": 0.0001869483984273694,
+      "loss": 1.1253,
+      "step": 985
+    },
+    {
+      "epoch": 0.18970088681797687,
+      "grad_norm": 0.28214016640278927,
+      "learning_rate": 0.00018691759748046594,
+      "loss": 1.1605,
+      "step": 986
+    },
+    {
+      "epoch": 0.18989328122651436,
+      "grad_norm": 0.27238602443860643,
+      "learning_rate": 0.00018688676277576916,
+      "loss": 1.1358,
+      "step": 987
+    },
+    {
+      "epoch": 0.19008567563505185,
+      "grad_norm": 0.34227702832507034,
+      "learning_rate": 0.00018685589432525492,
+      "loss": 1.2518,
+      "step": 988
+    },
+    {
+      "epoch": 0.19027807004358935,
+      "grad_norm": 0.28004287312705406,
+      "learning_rate": 0.0001868249921409122,
+      "loss": 1.0112,
+      "step": 989
+    },
+    {
+      "epoch": 0.19047046445212687,
+      "grad_norm": 0.2873550611088721,
+      "learning_rate": 0.00018679405623474296,
+      "loss": 1.2342,
+      "step": 990
+    },
+    {
+      "epoch": 0.19066285886066436,
+      "grad_norm": 0.25418784393800076,
+      "learning_rate": 0.00018676308661876243,
+      "loss": 1.0562,
+      "step": 991
+    },
+    {
+      "epoch": 0.19085525326920186,
+      "grad_norm": 0.2835909005007071,
+      "learning_rate": 0.00018673208330499883,
+      "loss": 1.2144,
+      "step": 992
+    },
+    {
+      "epoch": 0.19104764767773938,
+      "grad_norm": 0.3332844782432229,
+      "learning_rate": 0.00018670104630549348,
+      "loss": 1.2369,
+      "step": 993
+    },
+    {
+      "epoch": 0.19124004208627687,
+      "grad_norm": 0.3933352226114899,
+      "learning_rate": 0.0001866699756323008,
+      "loss": 1.1,
+      "step": 994
+    },
+    {
+      "epoch": 0.19143243649481437,
+      "grad_norm": 0.2935380742083632,
+      "learning_rate": 0.0001866388712974883,
+      "loss": 1.1043,
+      "step": 995
+    },
+    {
+      "epoch": 0.19162483090335186,
+      "grad_norm": 0.27358121515773026,
+      "learning_rate": 0.0001866077333131365,
+      "loss": 1.0905,
+      "step": 996
+    },
+    {
+      "epoch": 0.19181722531188938,
+      "grad_norm": 0.306629379220065,
+      "learning_rate": 0.00018657656169133907,
+      "loss": 1.1934,
+      "step": 997
+    },
+    {
+      "epoch": 0.19200961972042688,
+      "grad_norm": 0.28871486180626765,
+      "learning_rate": 0.00018654535644420275,
+      "loss": 1.1733,
+      "step": 998
+    },
+    {
+      "epoch": 0.19220201412896437,
+      "grad_norm": 0.2882640971719597,
+      "learning_rate": 0.00018651411758384718,
+      "loss": 1.1741,
+      "step": 999
+    },
+    {
+      "epoch": 0.19239440853750187,
+      "grad_norm": 0.13947954655889114,
+      "learning_rate": 0.0001864828451224053,
+      "loss": 1.127,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1925868029460394,
+      "grad_norm": 0.3025899712629719,
+      "learning_rate": 0.00018645153907202285,
+      "loss": 1.1595,
+      "step": 1001
+    },
+    {
+      "epoch": 0.19277919735457688,
+      "grad_norm": 0.27681580595475636,
+      "learning_rate": 0.00018642019944485884,
+      "loss": 1.1769,
+      "step": 1002
+    },
+    {
+      "epoch": 0.19297159176311438,
+      "grad_norm": 0.2966134805269456,
+      "learning_rate": 0.00018638882625308522,
+      "loss": 1.0378,
+      "step": 1003
+    },
+    {
+      "epoch": 0.1931639861716519,
+      "grad_norm": 0.35833152967747617,
+      "learning_rate": 0.0001863574195088869,
+      "loss": 1.0837,
+      "step": 1004
+    },
+    {
+      "epoch": 0.1933563805801894,
+      "grad_norm": 0.2958991050800634,
+      "learning_rate": 0.00018632597922446195,
+      "loss": 1.1686,
+      "step": 1005
+    },
+    {
+      "epoch": 0.19354877498872688,
+      "grad_norm": 0.3492786365549858,
+      "learning_rate": 0.0001862945054120214,
+      "loss": 1.2175,
+      "step": 1006
+    },
+    {
+      "epoch": 0.19374116939726438,
+      "grad_norm": 0.31341562808569357,
+      "learning_rate": 0.00018626299808378934,
+      "loss": 1.2116,
+      "step": 1007
+    },
+    {
+      "epoch": 0.1939335638058019,
+      "grad_norm": 0.29824132488910254,
+      "learning_rate": 0.00018623145725200278,
+      "loss": 1.1256,
+      "step": 1008
+    },
+    {
+      "epoch": 0.1941259582143394,
+      "grad_norm": 0.3322377671494098,
+      "learning_rate": 0.0001861998829289119,
+      "loss": 1.1351,
+      "step": 1009
+    },
+    {
+      "epoch": 0.1943183526228769,
+      "grad_norm": 0.343812947236017,
+      "learning_rate": 0.0001861682751267798,
+      "loss": 1.1791,
+      "step": 1010
+    },
+    {
+      "epoch": 0.1945107470314144,
+      "grad_norm": 0.3148311494011386,
+      "learning_rate": 0.00018613663385788252,
+      "loss": 1.1292,
+      "step": 1011
+    },
+    {
+      "epoch": 0.1947031414399519,
+      "grad_norm": 0.24541134382521418,
+      "learning_rate": 0.0001861049591345092,
+      "loss": 1.1716,
+      "step": 1012
+    },
+    {
+      "epoch": 0.1948955358484894,
+      "grad_norm": 0.3182043268234669,
+      "learning_rate": 0.00018607325096896195,
+      "loss": 1.2349,
+      "step": 1013
+    },
+    {
+      "epoch": 0.1950879302570269,
+      "grad_norm": 0.283737706616924,
+      "learning_rate": 0.00018604150937355586,
+      "loss": 1.1599,
+      "step": 1014
+    },
+    {
+      "epoch": 0.19528032466556441,
+      "grad_norm": 0.3329832592951454,
+      "learning_rate": 0.00018600973436061897,
+      "loss": 1.0952,
+      "step": 1015
+    },
+    {
+      "epoch": 0.1954727190741019,
+      "grad_norm": 0.2796509308766962,
+      "learning_rate": 0.00018597792594249234,
+      "loss": 1.033,
+      "step": 1016
+    },
+    {
+      "epoch": 0.1956651134826394,
+      "grad_norm": 0.26260876337805816,
+      "learning_rate": 0.00018594608413153001,
+      "loss": 1.0979,
+      "step": 1017
+    },
+    {
+      "epoch": 0.19585750789117692,
+      "grad_norm": 0.3297883785799931,
+      "learning_rate": 0.00018591420894009897,
+      "loss": 1.2115,
+      "step": 1018
+    },
+    {
+      "epoch": 0.19604990229971442,
+      "grad_norm": 0.29719183536715366,
+      "learning_rate": 0.00018588230038057914,
+      "loss": 1.1183,
+      "step": 1019
+    },
+    {
+      "epoch": 0.1962422967082519,
+      "grad_norm": 0.4358341202415237,
+      "learning_rate": 0.00018585035846536348,
+      "loss": 1.1326,
+      "step": 1020
+    },
+    {
+      "epoch": 0.1964346911167894,
+      "grad_norm": 0.3454508791763001,
+      "learning_rate": 0.00018581838320685782,
+      "loss": 1.1479,
+      "step": 1021
+    },
+    {
+      "epoch": 0.19662708552532693,
+      "grad_norm": 0.30060039291533,
+      "learning_rate": 0.00018578637461748106,
+      "loss": 1.1453,
+      "step": 1022
+    },
+    {
+      "epoch": 0.19681947993386442,
+      "grad_norm": 0.3338543540979239,
+      "learning_rate": 0.00018575433270966486,
+      "loss": 1.1294,
+      "step": 1023
+    },
+    {
+      "epoch": 0.19701187434240192,
+      "grad_norm": 0.27357267072948194,
+      "learning_rate": 0.00018572225749585402,
+      "loss": 1.1806,
+      "step": 1024
+    },
+    {
+      "epoch": 0.19720426875093944,
+      "grad_norm": 0.3120343250408094,
+      "learning_rate": 0.00018569014898850612,
+      "loss": 1.133,
+      "step": 1025
+    },
+    {
+      "epoch": 0.19739666315947693,
+      "grad_norm": 0.3477166900063389,
+      "learning_rate": 0.0001856580072000918,
+      "loss": 1.1008,
+      "step": 1026
+    },
+    {
+      "epoch": 0.19758905756801443,
+      "grad_norm": 0.2234282696276886,
+      "learning_rate": 0.00018562583214309445,
+      "loss": 1.1976,
+      "step": 1027
+    },
+    {
+      "epoch": 0.19778145197655192,
+      "grad_norm": 0.3280499890572628,
+      "learning_rate": 0.0001855936238300106,
+      "loss": 1.1355,
+      "step": 1028
+    },
+    {
+      "epoch": 0.19797384638508944,
+      "grad_norm": 0.2929436441679051,
+      "learning_rate": 0.00018556138227334955,
+      "loss": 1.1707,
+      "step": 1029
+    },
+    {
+      "epoch": 0.19816624079362694,
+      "grad_norm": 0.30070036527879784,
+      "learning_rate": 0.00018552910748563357,
+      "loss": 1.1875,
+      "step": 1030
+    },
+    {
+      "epoch": 0.19835863520216443,
+      "grad_norm": 0.23634706250113938,
+      "learning_rate": 0.00018549679947939778,
+      "loss": 1.149,
+      "step": 1031
+    },
+    {
+      "epoch": 0.19855102961070195,
+      "grad_norm": 0.28545181143680126,
+      "learning_rate": 0.00018546445826719024,
+      "loss": 1.0201,
+      "step": 1032
+    },
+    {
+      "epoch": 0.19874342401923945,
+      "grad_norm": 0.307121954461118,
+      "learning_rate": 0.00018543208386157193,
+      "loss": 1.0307,
+      "step": 1033
+    },
+    {
+      "epoch": 0.19893581842777694,
+      "grad_norm": 0.31654414999389935,
+      "learning_rate": 0.00018539967627511668,
+      "loss": 1.1158,
+      "step": 1034
+    },
+    {
+      "epoch": 0.19912821283631443,
+      "grad_norm": 0.3543921185896822,
+      "learning_rate": 0.00018536723552041123,
+      "loss": 1.1381,
+      "step": 1035
+    },
+    {
+      "epoch": 0.19932060724485196,
+      "grad_norm": 0.2902046620820293,
+      "learning_rate": 0.00018533476161005518,
+      "loss": 1.1096,
+      "step": 1036
+    },
+    {
+      "epoch": 0.19951300165338945,
+      "grad_norm": 0.3746716210376939,
+      "learning_rate": 0.00018530225455666102,
+      "loss": 1.1124,
+      "step": 1037
+    },
+    {
+      "epoch": 0.19970539606192694,
+      "grad_norm": 0.2596079542898459,
+      "learning_rate": 0.00018526971437285417,
+      "loss": 1.1852,
+      "step": 1038
+    },
+    {
+      "epoch": 0.19989779047046446,
+      "grad_norm": 0.25952519079003095,
+      "learning_rate": 0.00018523714107127276,
+      "loss": 1.0874,
+      "step": 1039
+    },
+    {
+      "epoch": 0.20009018487900196,
+      "grad_norm": 0.2974050284024279,
+      "learning_rate": 0.00018520453466456797,
+      "loss": 1.1883,
+      "step": 1040
+    },
+    {
+      "epoch": 0.20028257928753945,
+      "grad_norm": 0.37566020443352655,
+      "learning_rate": 0.00018517189516540378,
+      "loss": 1.0979,
+      "step": 1041
+    },
+    {
+      "epoch": 0.20047497369607695,
+      "grad_norm": 0.3227350978880759,
+      "learning_rate": 0.00018513922258645685,
+      "loss": 1.1349,
+      "step": 1042
+    },
+    {
+      "epoch": 0.20066736810461447,
+      "grad_norm": 0.2707840184031702,
+      "learning_rate": 0.000185106516940417,
+      "loss": 1.1197,
+      "step": 1043
+    },
+    {
+      "epoch": 0.20085976251315196,
+      "grad_norm": 0.24086167817755455,
+      "learning_rate": 0.00018507377823998665,
+      "loss": 1.1774,
+      "step": 1044
+    },
+    {
+      "epoch": 0.20105215692168946,
+      "grad_norm": 0.318161256984643,
+      "learning_rate": 0.0001850410064978811,
+      "loss": 1.0991,
+      "step": 1045
+    },
+    {
+      "epoch": 0.20124455133022698,
+      "grad_norm": 0.3034513893807528,
+      "learning_rate": 0.00018500820172682856,
+      "loss": 1.0625,
+      "step": 1046
+    },
+    {
+      "epoch": 0.20143694573876447,
+      "grad_norm": 0.33655796773506963,
+      "learning_rate": 0.00018497536393957003,
+      "loss": 1.1437,
+      "step": 1047
+    },
+    {
+      "epoch": 0.20162934014730197,
+      "grad_norm": 0.3437961976086482,
+      "learning_rate": 0.0001849424931488593,
+      "loss": 1.1544,
+      "step": 1048
+    },
+    {
+      "epoch": 0.20182173455583946,
+      "grad_norm": 0.2900820367137122,
+      "learning_rate": 0.00018490958936746306,
+      "loss": 1.1205,
+      "step": 1049
+    },
+    {
+      "epoch": 0.20201412896437698,
+      "grad_norm": 0.3036320803489058,
+      "learning_rate": 0.0001848766526081607,
+      "loss": 1.0781,
+      "step": 1050
+    },
+    {
+      "epoch": 0.20220652337291448,
+      "grad_norm": 0.35547222447306837,
+      "learning_rate": 0.0001848436828837445,
+      "loss": 1.2098,
+      "step": 1051
+    },
+    {
+      "epoch": 0.20239891778145197,
+      "grad_norm": 0.40675861481643805,
+      "learning_rate": 0.00018481068020701952,
+      "loss": 0.9895,
+      "step": 1052
+    },
+    {
+      "epoch": 0.2025913121899895,
+      "grad_norm": 0.2858892320887802,
+      "learning_rate": 0.00018477764459080362,
+      "loss": 1.0918,
+      "step": 1053
+    },
+    {
+      "epoch": 0.20278370659852699,
+      "grad_norm": 0.3542974949205585,
+      "learning_rate": 0.00018474457604792746,
+      "loss": 1.1099,
+      "step": 1054
+    },
+    {
+      "epoch": 0.20297610100706448,
+      "grad_norm": 0.2815771555898335,
+      "learning_rate": 0.00018471147459123448,
+      "loss": 1.1638,
+      "step": 1055
+    },
+    {
+      "epoch": 0.20316849541560197,
+      "grad_norm": 0.23213614124799611,
+      "learning_rate": 0.00018467834023358086,
+      "loss": 1.0601,
+      "step": 1056
+    },
+    {
+      "epoch": 0.2033608898241395,
+      "grad_norm": 0.3110000979581457,
+      "learning_rate": 0.0001846451729878357,
+      "loss": 1.1102,
+      "step": 1057
+    },
+    {
+      "epoch": 0.203553284232677,
+      "grad_norm": 0.343155039665011,
+      "learning_rate": 0.00018461197286688067,
+      "loss": 1.0351,
+      "step": 1058
+    },
+    {
+      "epoch": 0.20374567864121448,
+      "grad_norm": 0.31762030521485873,
+      "learning_rate": 0.00018457873988361037,
+      "loss": 1.1313,
+      "step": 1059
+    },
+    {
+      "epoch": 0.20393807304975198,
+      "grad_norm": 0.3034099270607324,
+      "learning_rate": 0.00018454547405093212,
+      "loss": 1.1124,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2041304674582895,
+      "grad_norm": 0.31508523484241846,
+      "learning_rate": 0.00018451217538176598,
+      "loss": 1.0995,
+      "step": 1061
+    },
+    {
+      "epoch": 0.204322861866827,
+      "grad_norm": 0.32737658093192185,
+      "learning_rate": 0.0001844788438890447,
+      "loss": 1.134,
+      "step": 1062
+    },
+    {
+      "epoch": 0.2045152562753645,
+      "grad_norm": 0.30017505137458866,
+      "learning_rate": 0.00018444547958571396,
+      "loss": 1.1007,
+      "step": 1063
+    },
+    {
+      "epoch": 0.204707650683902,
+      "grad_norm": 0.35179798282014624,
+      "learning_rate": 0.00018441208248473197,
+      "loss": 1.1319,
+      "step": 1064
+    },
+    {
+      "epoch": 0.2049000450924395,
+      "grad_norm": 0.39720645495643087,
+      "learning_rate": 0.00018437865259906985,
+      "loss": 1.1498,
+      "step": 1065
+    },
+    {
+      "epoch": 0.205092439500977,
+      "grad_norm": 0.3130211632216441,
+      "learning_rate": 0.00018434518994171135,
+      "loss": 1.0338,
+      "step": 1066
+    },
+    {
+      "epoch": 0.2052848339095145,
+      "grad_norm": 0.31935911128412864,
+      "learning_rate": 0.000184311694525653,
+      "loss": 1.1226,
+      "step": 1067
+    },
+    {
+      "epoch": 0.205477228318052,
+      "grad_norm": 0.2983925784553987,
+      "learning_rate": 0.00018427816636390398,
+      "loss": 1.1327,
+      "step": 1068
+    },
+    {
+      "epoch": 0.2056696227265895,
+      "grad_norm": 0.2986694436938893,
+      "learning_rate": 0.0001842446054694863,
+      "loss": 1.1313,
+      "step": 1069
+    },
+    {
+      "epoch": 0.205862017135127,
+      "grad_norm": 0.27645054405518904,
+      "learning_rate": 0.00018421101185543465,
+      "loss": 1.0862,
+      "step": 1070
+    },
+    {
+      "epoch": 0.20605441154366452,
+      "grad_norm": 0.34654391537794565,
+      "learning_rate": 0.0001841773855347963,
+      "loss": 1.1144,
+      "step": 1071
+    },
+    {
+      "epoch": 0.20624680595220202,
+      "grad_norm": 0.2921548719447003,
+      "learning_rate": 0.0001841437265206314,
+      "loss": 1.093,
+      "step": 1072
+    },
+    {
+      "epoch": 0.2064392003607395,
+      "grad_norm": 0.33910685260576345,
+      "learning_rate": 0.0001841100348260127,
+      "loss": 1.052,
+      "step": 1073
+    },
+    {
+      "epoch": 0.206631594769277,
+      "grad_norm": 0.356207723915354,
+      "learning_rate": 0.0001840763104640257,
+      "loss": 1.1316,
+      "step": 1074
+    },
+    {
+      "epoch": 0.20682398917781453,
+      "grad_norm": 0.3213575450598716,
+      "learning_rate": 0.0001840425534477685,
+      "loss": 1.0683,
+      "step": 1075
+    },
+    {
+      "epoch": 0.20701638358635202,
+      "grad_norm": 0.3611732703231641,
+      "learning_rate": 0.000184008763790352,
+      "loss": 1.1431,
+      "step": 1076
+    },
+    {
+      "epoch": 0.20720877799488951,
+      "grad_norm": 0.2599301226681828,
+      "learning_rate": 0.00018397494150489965,
+      "loss": 1.1671,
+      "step": 1077
+    },
+    {
+      "epoch": 0.20740117240342704,
+      "grad_norm": 0.3573389190698938,
+      "learning_rate": 0.00018394108660454766,
+      "loss": 1.2706,
+      "step": 1078
+    },
+    {
+      "epoch": 0.20759356681196453,
+      "grad_norm": 0.3943059266582131,
+      "learning_rate": 0.00018390719910244487,
+      "loss": 1.1055,
+      "step": 1079
+    },
+    {
+      "epoch": 0.20778596122050202,
+      "grad_norm": 0.2868571124179789,
+      "learning_rate": 0.00018387327901175284,
+      "loss": 1.1002,
+      "step": 1080
+    },
+    {
+      "epoch": 0.20797835562903952,
+      "grad_norm": 0.2559251868316223,
+      "learning_rate": 0.00018383932634564572,
+      "loss": 1.1053,
+      "step": 1081
+    },
+    {
+      "epoch": 0.20817075003757704,
+      "grad_norm": 0.3153264220444252,
+      "learning_rate": 0.00018380534111731028,
+      "loss": 1.2179,
+      "step": 1082
+    },
+    {
+      "epoch": 0.20836314444611453,
+      "grad_norm": 0.34781515789392425,
+      "learning_rate": 0.00018377132333994607,
+      "loss": 1.0962,
+      "step": 1083
+    },
+    {
+      "epoch": 0.20855553885465203,
+      "grad_norm": 0.27326408571733896,
+      "learning_rate": 0.0001837372730267652,
+      "loss": 1.1088,
+      "step": 1084
+    },
+    {
+      "epoch": 0.20874793326318955,
+      "grad_norm": 0.3860607559252603,
+      "learning_rate": 0.00018370319019099234,
+      "loss": 1.0594,
+      "step": 1085
+    },
+    {
+      "epoch": 0.20894032767172704,
+      "grad_norm": 0.30528297689667394,
+      "learning_rate": 0.00018366907484586496,
+      "loss": 1.0382,
+      "step": 1086
+    },
+    {
+      "epoch": 0.20913272208026454,
+      "grad_norm": 0.3847453799085574,
+      "learning_rate": 0.000183634927004633,
+      "loss": 1.0607,
+      "step": 1087
+    },
+    {
+      "epoch": 0.20932511648880203,
+      "grad_norm": 0.38962908628574694,
+      "learning_rate": 0.00018360074668055914,
+      "loss": 1.0069,
+      "step": 1088
+    },
+    {
+      "epoch": 0.20951751089733955,
+      "grad_norm": 0.4797244342763735,
+      "learning_rate": 0.00018356653388691859,
+      "loss": 1.075,
+      "step": 1089
+    },
+    {
+      "epoch": 0.20970990530587705,
+      "grad_norm": 0.38371768004891116,
+      "learning_rate": 0.00018353228863699923,
+      "loss": 1.2317,
+      "step": 1090
+    },
+    {
+      "epoch": 0.20990229971441454,
+      "grad_norm": 0.2648546710256649,
+      "learning_rate": 0.00018349801094410148,
+      "loss": 1.1473,
+      "step": 1091
+    },
+    {
+      "epoch": 0.21009469412295206,
+      "grad_norm": 0.3810434720358997,
+      "learning_rate": 0.00018346370082153844,
+      "loss": 1.0398,
+      "step": 1092
+    },
+    {
+      "epoch": 0.21028708853148956,
+      "grad_norm": 0.29686409378819517,
+      "learning_rate": 0.00018342935828263575,
+      "loss": 1.125,
+      "step": 1093
+    },
+    {
+      "epoch": 0.21047948294002705,
+      "grad_norm": 0.34449014797680627,
+      "learning_rate": 0.00018339498334073168,
+      "loss": 1.1458,
+      "step": 1094
+    },
+    {
+      "epoch": 0.21067187734856455,
+      "grad_norm": 0.25325758368432677,
+      "learning_rate": 0.000183360576009177,
+      "loss": 1.0441,
+      "step": 1095
+    },
+    {
+      "epoch": 0.21086427175710207,
+      "grad_norm": 0.27215896328261197,
+      "learning_rate": 0.00018332613630133517,
+      "loss": 1.1009,
+      "step": 1096
+    },
+    {
+      "epoch": 0.21105666616563956,
+      "grad_norm": 0.38131166614991063,
+      "learning_rate": 0.0001832916642305822,
+      "loss": 1.094,
+      "step": 1097
+    },
+    {
+      "epoch": 0.21124906057417706,
+      "grad_norm": 0.3369507588000508,
+      "learning_rate": 0.00018325715981030657,
+      "loss": 1.2038,
+      "step": 1098
+    },
+    {
+      "epoch": 0.21144145498271458,
+      "grad_norm": 0.35080617220553933,
+      "learning_rate": 0.00018322262305390947,
+      "loss": 1.0997,
+      "step": 1099
+    },
+    {
+      "epoch": 0.21163384939125207,
+      "grad_norm": 0.2889199724957277,
+      "learning_rate": 0.00018318805397480453,
+      "loss": 1.1646,
+      "step": 1100
+    },
+    {
+      "epoch": 0.21182624379978957,
+      "grad_norm": 0.33774441430192076,
+      "learning_rate": 0.00018315345258641802,
+      "loss": 1.1139,
+      "step": 1101
+    },
+    {
+      "epoch": 0.21201863820832706,
+      "grad_norm": 0.28859761991935223,
+      "learning_rate": 0.00018311881890218873,
+      "loss": 1.1036,
+      "step": 1102
+    },
+    {
+      "epoch": 0.21221103261686458,
+      "grad_norm": 0.3200606896820387,
+      "learning_rate": 0.0001830841529355679,
+      "loss": 1.125,
+      "step": 1103
+    },
+    {
+      "epoch": 0.21240342702540208,
+      "grad_norm": 0.5111080731032329,
+      "learning_rate": 0.0001830494547000195,
+      "loss": 1.2038,
+      "step": 1104
+    },
+    {
+      "epoch": 0.21259582143393957,
+      "grad_norm": 0.32682771642658637,
+      "learning_rate": 0.00018301472420901983,
+      "loss": 1.0573,
+      "step": 1105
+    },
+    {
+      "epoch": 0.2127882158424771,
+      "grad_norm": 0.2614531758423323,
+      "learning_rate": 0.00018297996147605787,
+      "loss": 1.0501,
+      "step": 1106
+    },
+    {
+      "epoch": 0.21298061025101458,
+      "grad_norm": 0.27390724391931975,
+      "learning_rate": 0.00018294516651463507,
+      "loss": 1.189,
+      "step": 1107
+    },
+    {
+      "epoch": 0.21317300465955208,
+      "grad_norm": 0.44717349512111515,
+      "learning_rate": 0.00018291033933826533,
+      "loss": 1.1307,
+      "step": 1108
+    },
+    {
+      "epoch": 0.21336539906808957,
+      "grad_norm": 0.2772467096253637,
+      "learning_rate": 0.0001828754799604752,
+      "loss": 1.1356,
+      "step": 1109
+    },
+    {
+      "epoch": 0.2135577934766271,
+      "grad_norm": 0.3173355957614274,
+      "learning_rate": 0.00018284058839480361,
+      "loss": 1.1513,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2137501878851646,
+      "grad_norm": 0.3019262414205688,
+      "learning_rate": 0.00018280566465480206,
+      "loss": 1.1567,
+      "step": 1111
+    },
+    {
+      "epoch": 0.21394258229370208,
+      "grad_norm": 0.3414952526642444,
+      "learning_rate": 0.00018277070875403456,
+      "loss": 1.1014,
+      "step": 1112
+    },
+    {
+      "epoch": 0.2141349767022396,
+      "grad_norm": 0.30910820398515804,
+      "learning_rate": 0.00018273572070607753,
+      "loss": 1.151,
+      "step": 1113
+    },
+    {
+      "epoch": 0.2143273711107771,
+      "grad_norm": 0.3190859453631591,
+      "learning_rate": 0.00018270070052451993,
+      "loss": 1.1189,
+      "step": 1114
+    },
+    {
+      "epoch": 0.2145197655193146,
+      "grad_norm": 0.3587526892073246,
+      "learning_rate": 0.00018266564822296322,
+      "loss": 1.214,
+      "step": 1115
+    },
+    {
+      "epoch": 0.2147121599278521,
+      "grad_norm": 0.3388880499961147,
+      "learning_rate": 0.0001826305638150213,
+      "loss": 1.0513,
+      "step": 1116
+    },
+    {
+      "epoch": 0.2149045543363896,
+      "grad_norm": 0.25139899206117644,
+      "learning_rate": 0.0001825954473143205,
+      "loss": 1.1252,
+      "step": 1117
+    },
+    {
+      "epoch": 0.2150969487449271,
+      "grad_norm": 0.2916843674505857,
+      "learning_rate": 0.00018256029873449974,
+      "loss": 1.1292,
+      "step": 1118
+    },
+    {
+      "epoch": 0.2152893431534646,
+      "grad_norm": 0.29562311422619053,
+      "learning_rate": 0.00018252511808921032,
+      "loss": 1.0445,
+      "step": 1119
+    },
+    {
+      "epoch": 0.21548173756200212,
+      "grad_norm": 0.3544800663138205,
+      "learning_rate": 0.00018248990539211594,
+      "loss": 1.111,
+      "step": 1120
+    },
+    {
+      "epoch": 0.2156741319705396,
+      "grad_norm": 0.27711328478108577,
+      "learning_rate": 0.00018245466065689282,
+      "loss": 1.1431,
+      "step": 1121
+    },
+    {
+      "epoch": 0.2158665263790771,
+      "grad_norm": 0.2840924369934067,
+      "learning_rate": 0.00018241938389722967,
+      "loss": 1.0795,
+      "step": 1122
+    },
+    {
+      "epoch": 0.2160589207876146,
+      "grad_norm": 0.29463204757123623,
+      "learning_rate": 0.0001823840751268275,
+      "loss": 0.9893,
+      "step": 1123
+    },
+    {
+      "epoch": 0.21625131519615212,
+      "grad_norm": 0.2802878329344074,
+      "learning_rate": 0.00018234873435939988,
+      "loss": 1.183,
+      "step": 1124
+    },
+    {
+      "epoch": 0.21644370960468962,
+      "grad_norm": 0.26295221742941355,
+      "learning_rate": 0.00018231336160867275,
+      "loss": 1.0616,
+      "step": 1125
+    },
+    {
+      "epoch": 0.2166361040132271,
+      "grad_norm": 0.27645528237930805,
+      "learning_rate": 0.00018227795688838445,
+      "loss": 1.1208,
+      "step": 1126
+    },
+    {
+      "epoch": 0.2168284984217646,
+      "grad_norm": 0.41457200035887104,
+      "learning_rate": 0.00018224252021228579,
+      "loss": 1.158,
+      "step": 1127
+    },
+    {
+      "epoch": 0.21702089283030213,
+      "grad_norm": 0.2504237141422832,
+      "learning_rate": 0.00018220705159413994,
+      "loss": 1.0543,
+      "step": 1128
+    },
+    {
+      "epoch": 0.21721328723883962,
+      "grad_norm": 0.3564033693238237,
+      "learning_rate": 0.00018217155104772258,
+      "loss": 1.0876,
+      "step": 1129
+    },
+    {
+      "epoch": 0.2174056816473771,
+      "grad_norm": 0.4176021547932737,
+      "learning_rate": 0.0001821360185868216,
+      "loss": 1.1573,
+      "step": 1130
+    },
+    {
+      "epoch": 0.21759807605591464,
+      "grad_norm": 0.27679006739125084,
+      "learning_rate": 0.00018210045422523745,
+      "loss": 1.1081,
+      "step": 1131
+    },
+    {
+      "epoch": 0.21779047046445213,
+      "grad_norm": 0.27540677311560946,
+      "learning_rate": 0.00018206485797678295,
+      "loss": 1.1444,
+      "step": 1132
+    },
+    {
+      "epoch": 0.21798286487298962,
+      "grad_norm": 0.32759307506513685,
+      "learning_rate": 0.00018202922985528322,
+      "loss": 1.2375,
+      "step": 1133
+    },
+    {
+      "epoch": 0.21817525928152712,
+      "grad_norm": 0.38015487387193025,
+      "learning_rate": 0.00018199356987457586,
+      "loss": 1.1722,
+      "step": 1134
+    },
+    {
+      "epoch": 0.21836765369006464,
+      "grad_norm": 0.35890990510196485,
+      "learning_rate": 0.00018195787804851078,
+      "loss": 1.1687,
+      "step": 1135
+    },
+    {
+      "epoch": 0.21856004809860213,
+      "grad_norm": 0.3268318356976673,
+      "learning_rate": 0.00018192215439095024,
+      "loss": 1.0936,
+      "step": 1136
+    },
+    {
+      "epoch": 0.21875244250713963,
+      "grad_norm": 0.2862558292108705,
+      "learning_rate": 0.00018188639891576893,
+      "loss": 1.0997,
+      "step": 1137
+    },
+    {
+      "epoch": 0.21894483691567715,
+      "grad_norm": 0.3225348998051424,
+      "learning_rate": 0.00018185061163685385,
+      "loss": 1.0969,
+      "step": 1138
+    },
+    {
+      "epoch": 0.21913723132421464,
+      "grad_norm": 0.32115413379940383,
+      "learning_rate": 0.0001818147925681044,
+      "loss": 0.995,
+      "step": 1139
+    },
+    {
+      "epoch": 0.21932962573275214,
+      "grad_norm": 0.34817571357384747,
+      "learning_rate": 0.00018177894172343226,
+      "loss": 1.2191,
+      "step": 1140
+    },
+    {
+      "epoch": 0.21952202014128963,
+      "grad_norm": 0.290960976788888,
+      "learning_rate": 0.00018174305911676148,
+      "loss": 1.0787,
+      "step": 1141
+    },
+    {
+      "epoch": 0.21971441454982715,
+      "grad_norm": 0.28993414066683065,
+      "learning_rate": 0.00018170714476202848,
+      "loss": 1.0371,
+      "step": 1142
+    },
+    {
+      "epoch": 0.21990680895836465,
+      "grad_norm": 0.30313896925824735,
+      "learning_rate": 0.00018167119867318197,
+      "loss": 1.069,
+      "step": 1143
+    },
+    {
+      "epoch": 0.22009920336690214,
+      "grad_norm": 0.2418135547481211,
+      "learning_rate": 0.00018163522086418299,
+      "loss": 1.1091,
+      "step": 1144
+    },
+    {
+      "epoch": 0.22029159777543966,
+      "grad_norm": 0.28439664411507437,
+      "learning_rate": 0.00018159921134900487,
+      "loss": 1.1865,
+      "step": 1145
+    },
+    {
+      "epoch": 0.22048399218397716,
+      "grad_norm": 0.3293388321514027,
+      "learning_rate": 0.00018156317014163338,
+      "loss": 1.1025,
+      "step": 1146
+    },
+    {
+      "epoch": 0.22067638659251465,
+      "grad_norm": 0.3454728241335007,
+      "learning_rate": 0.0001815270972560664,
+      "loss": 1.1852,
+      "step": 1147
+    },
+    {
+      "epoch": 0.22086878100105214,
+      "grad_norm": 0.31567062075947233,
+      "learning_rate": 0.00018149099270631434,
+      "loss": 1.0556,
+      "step": 1148
+    },
+    {
+      "epoch": 0.22106117540958967,
+      "grad_norm": 0.3215089924268069,
+      "learning_rate": 0.00018145485650639974,
+      "loss": 1.1194,
+      "step": 1149
+    },
+    {
+      "epoch": 0.22125356981812716,
+      "grad_norm": 0.39734413486853337,
+      "learning_rate": 0.00018141868867035745,
+      "loss": 1.1557,
+      "step": 1150
+    },
+    {
+      "epoch": 0.22144596422666465,
+      "grad_norm": 0.28251570115437874,
+      "learning_rate": 0.00018138248921223467,
+      "loss": 1.1367,
+      "step": 1151
+    },
+    {
+      "epoch": 0.22163835863520218,
+      "grad_norm": 0.3243676259260902,
+      "learning_rate": 0.00018134625814609083,
+      "loss": 1.0262,
+      "step": 1152
+    },
+    {
+      "epoch": 0.22183075304373967,
+      "grad_norm": 0.29467276277484045,
+      "learning_rate": 0.00018130999548599767,
+      "loss": 1.242,
+      "step": 1153
+    },
+    {
+      "epoch": 0.22202314745227716,
+      "grad_norm": 0.2534482334810057,
+      "learning_rate": 0.00018127370124603926,
+      "loss": 1.079,
+      "step": 1154
+    },
+    {
+      "epoch": 0.22221554186081466,
+      "grad_norm": 0.29830626369083707,
+      "learning_rate": 0.00018123737544031176,
+      "loss": 1.0556,
+      "step": 1155
+    },
+    {
+      "epoch": 0.22240793626935218,
+      "grad_norm": 0.28095587277034634,
+      "learning_rate": 0.00018120101808292372,
+      "loss": 1.1014,
+      "step": 1156
+    },
+    {
+      "epoch": 0.22260033067788967,
+      "grad_norm": 0.33055926490969234,
+      "learning_rate": 0.000181164629187996,
+      "loss": 1.1262,
+      "step": 1157
+    },
+    {
+      "epoch": 0.22279272508642717,
+      "grad_norm": 0.2989713967872641,
+      "learning_rate": 0.00018112820876966152,
+      "loss": 1.1338,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2229851194949647,
+      "grad_norm": 0.3102486797503027,
+      "learning_rate": 0.00018109175684206559,
+      "loss": 1.2324,
+      "step": 1159
+    },
+    {
+      "epoch": 0.22317751390350218,
+      "grad_norm": 0.29154857656779337,
+      "learning_rate": 0.00018105527341936574,
+      "loss": 1.1259,
+      "step": 1160
+    },
+    {
+      "epoch": 0.22336990831203968,
+      "grad_norm": 0.28834012144968674,
+      "learning_rate": 0.0001810187585157317,
+      "loss": 1.1156,
+      "step": 1161
+    },
+    {
+      "epoch": 0.22356230272057717,
+      "grad_norm": 0.3206750966434899,
+      "learning_rate": 0.00018098221214534542,
+      "loss": 1.0863,
+      "step": 1162
+    },
+    {
+      "epoch": 0.2237546971291147,
+      "grad_norm": 0.2853576319076801,
+      "learning_rate": 0.00018094563432240107,
+      "loss": 1.0819,
+      "step": 1163
+    },
+    {
+      "epoch": 0.2239470915376522,
+      "grad_norm": 0.3347862601048052,
+      "learning_rate": 0.0001809090250611051,
+      "loss": 1.1518,
+      "step": 1164
+    },
+    {
+      "epoch": 0.22413948594618968,
+      "grad_norm": 0.3679728314356357,
+      "learning_rate": 0.00018087238437567614,
+      "loss": 1.1644,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2243318803547272,
+      "grad_norm": 0.27506822040907014,
+      "learning_rate": 0.00018083571228034496,
+      "loss": 1.1143,
+      "step": 1166
+    },
+    {
+      "epoch": 0.2245242747632647,
+      "grad_norm": 0.3802297920197197,
+      "learning_rate": 0.0001807990087893546,
+      "loss": 1.1691,
+      "step": 1167
+    },
+    {
+      "epoch": 0.2247166691718022,
+      "grad_norm": 0.4106487228143322,
+      "learning_rate": 0.00018076227391696027,
+      "loss": 1.1561,
+      "step": 1168
+    },
+    {
+      "epoch": 0.22490906358033969,
+      "grad_norm": 0.26517102054630975,
+      "learning_rate": 0.0001807255076774294,
+      "loss": 1.05,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2251014579888772,
+      "grad_norm": 0.322212844818469,
+      "learning_rate": 0.00018068871008504152,
+      "loss": 1.0668,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2252938523974147,
+      "grad_norm": 0.30209211301167815,
+      "learning_rate": 0.00018065188115408843,
+      "loss": 1.1396,
+      "step": 1171
+    },
+    {
+      "epoch": 0.2254862468059522,
+      "grad_norm": 0.26949018914889616,
+      "learning_rate": 0.00018061502089887407,
+      "loss": 1.1514,
+      "step": 1172
+    },
+    {
+      "epoch": 0.22567864121448972,
+      "grad_norm": 0.281250937196479,
+      "learning_rate": 0.0001805781293337145,
+      "loss": 1.22,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2258710356230272,
+      "grad_norm": 0.314498538535587,
+      "learning_rate": 0.00018054120647293797,
+      "loss": 1.149,
+      "step": 1174
+    },
+    {
+      "epoch": 0.2260634300315647,
+      "grad_norm": 0.4176947148865461,
+      "learning_rate": 0.00018050425233088495,
+      "loss": 1.174,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2262558244401022,
+      "grad_norm": 0.3435354809332262,
+      "learning_rate": 0.000180467266921908,
+      "loss": 1.2005,
+      "step": 1176
+    },
+    {
+      "epoch": 0.22644821884863972,
+      "grad_norm": 0.31856542143170125,
+      "learning_rate": 0.00018043025026037176,
+      "loss": 1.0522,
+      "step": 1177
+    },
+    {
+      "epoch": 0.22664061325717721,
+      "grad_norm": 0.34809079323389497,
+      "learning_rate": 0.00018039320236065313,
+      "loss": 1.1386,
+      "step": 1178
+    },
+    {
+      "epoch": 0.2268330076657147,
+      "grad_norm": 0.37234319975743585,
+      "learning_rate": 0.0001803561232371411,
+      "loss": 1.2279,
+      "step": 1179
+    },
+    {
+      "epoch": 0.22702540207425223,
+      "grad_norm": 0.2848123093467546,
+      "learning_rate": 0.0001803190129042367,
+      "loss": 1.1423,
+      "step": 1180
+    },
+    {
+      "epoch": 0.22721779648278972,
+      "grad_norm": 0.3219113886369367,
+      "learning_rate": 0.00018028187137635325,
+      "loss": 1.0994,
+      "step": 1181
+    },
+    {
+      "epoch": 0.22741019089132722,
+      "grad_norm": 0.33774998210198537,
+      "learning_rate": 0.000180244698667916,
+      "loss": 1.109,
+      "step": 1182
+    },
+    {
+      "epoch": 0.2276025852998647,
+      "grad_norm": 0.32188700800860826,
+      "learning_rate": 0.00018020749479336247,
+      "loss": 1.157,
+      "step": 1183
+    },
+    {
+      "epoch": 0.22779497970840223,
+      "grad_norm": 0.30941105723702805,
+      "learning_rate": 0.00018017025976714218,
+      "loss": 1.0672,
+      "step": 1184
+    },
+    {
+      "epoch": 0.22798737411693973,
+      "grad_norm": 0.25333574862334907,
+      "learning_rate": 0.00018013299360371686,
+      "loss": 1.0461,
+      "step": 1185
+    },
+    {
+      "epoch": 0.22817976852547722,
+      "grad_norm": 0.32315928483092893,
+      "learning_rate": 0.00018009569631756013,
+      "loss": 1.0754,
+      "step": 1186
+    },
+    {
+      "epoch": 0.22837216293401472,
+      "grad_norm": 0.29511060133275824,
+      "learning_rate": 0.0001800583679231579,
+      "loss": 1.1893,
+      "step": 1187
+    },
+    {
+      "epoch": 0.22856455734255224,
+      "grad_norm": 0.27697451351241503,
+      "learning_rate": 0.0001800210084350081,
+      "loss": 1.1206,
+      "step": 1188
+    },
+    {
+      "epoch": 0.22875695175108973,
+      "grad_norm": 0.3479481888443204,
+      "learning_rate": 0.0001799836178676207,
+      "loss": 1.1708,
+      "step": 1189
+    },
+    {
+      "epoch": 0.22894934615962723,
+      "grad_norm": 0.2795161515312204,
+      "learning_rate": 0.00017994619623551777,
+      "loss": 1.1509,
+      "step": 1190
+    },
+    {
+      "epoch": 0.22914174056816475,
+      "grad_norm": 0.2532610764898941,
+      "learning_rate": 0.00017990874355323344,
+      "loss": 1.1401,
+      "step": 1191
+    },
+    {
+      "epoch": 0.22933413497670224,
+      "grad_norm": 0.319630712163536,
+      "learning_rate": 0.00017987125983531392,
+      "loss": 1.1577,
+      "step": 1192
+    },
+    {
+      "epoch": 0.22952652938523974,
+      "grad_norm": 0.2574571635106825,
+      "learning_rate": 0.00017983374509631742,
+      "loss": 1.0504,
+      "step": 1193
+    },
+    {
+      "epoch": 0.22971892379377723,
+      "grad_norm": 0.3250445193011001,
+      "learning_rate": 0.00017979619935081426,
+      "loss": 1.1761,
+      "step": 1194
+    },
+    {
+      "epoch": 0.22991131820231475,
+      "grad_norm": 0.34924141858396734,
+      "learning_rate": 0.0001797586226133867,
+      "loss": 1.1741,
+      "step": 1195
+    },
+    {
+      "epoch": 0.23010371261085225,
+      "grad_norm": 0.2772679255919962,
+      "learning_rate": 0.00017972101489862924,
+      "loss": 1.0978,
+      "step": 1196
+    },
+    {
+      "epoch": 0.23029610701938974,
+      "grad_norm": 0.28693778256166075,
+      "learning_rate": 0.00017968337622114824,
+      "loss": 1.0746,
+      "step": 1197
+    },
+    {
+      "epoch": 0.23048850142792726,
+      "grad_norm": 0.2754003056772629,
+      "learning_rate": 0.00017964570659556207,
+      "loss": 1.1347,
+      "step": 1198
+    },
+    {
+      "epoch": 0.23068089583646476,
+      "grad_norm": 0.3823588245718991,
+      "learning_rate": 0.0001796080060365012,
+      "loss": 1.2152,
+      "step": 1199
+    },
+    {
+      "epoch": 0.23087329024500225,
+      "grad_norm": 0.26500918067157575,
+      "learning_rate": 0.00017957027455860813,
+      "loss": 1.2071,
+      "step": 1200
+    },
+    {
+      "epoch": 0.23106568465353974,
+      "grad_norm": 0.29457638314716683,
+      "learning_rate": 0.0001795325121765373,
+      "loss": 1.0402,
+      "step": 1201
+    },
+    {
+      "epoch": 0.23125807906207727,
+      "grad_norm": 0.3350062968757272,
+      "learning_rate": 0.00017949471890495522,
+      "loss": 1.1064,
+      "step": 1202
+    },
+    {
+      "epoch": 0.23145047347061476,
+      "grad_norm": 0.29950556650813864,
+      "learning_rate": 0.0001794568947585403,
+      "loss": 1.1487,
+      "step": 1203
+    },
+    {
+      "epoch": 0.23164286787915225,
+      "grad_norm": 0.3487488652341284,
+      "learning_rate": 0.00017941903975198306,
+      "loss": 1.0572,
+      "step": 1204
+    },
+    {
+      "epoch": 0.23183526228768978,
+      "grad_norm": 0.30685417559592054,
+      "learning_rate": 0.00017938115389998593,
+      "loss": 1.0642,
+      "step": 1205
+    },
+    {
+      "epoch": 0.23202765669622727,
+      "grad_norm": 0.34244144646472363,
+      "learning_rate": 0.00017934323721726332,
+      "loss": 1.1652,
+      "step": 1206
+    },
+    {
+      "epoch": 0.23222005110476476,
+      "grad_norm": 0.27706379853450247,
+      "learning_rate": 0.00017930528971854163,
+      "loss": 1.1486,
+      "step": 1207
+    },
+    {
+      "epoch": 0.23241244551330226,
+      "grad_norm": 0.28539288522573164,
+      "learning_rate": 0.00017926731141855928,
+      "loss": 1.1481,
+      "step": 1208
+    },
+    {
+      "epoch": 0.23260483992183978,
+      "grad_norm": 0.35450120298693044,
+      "learning_rate": 0.00017922930233206657,
+      "loss": 1.0455,
+      "step": 1209
+    },
+    {
+      "epoch": 0.23279723433037727,
+      "grad_norm": 0.2667447976577724,
+      "learning_rate": 0.00017919126247382575,
+      "loss": 1.0984,
+      "step": 1210
+    },
+    {
+      "epoch": 0.23298962873891477,
+      "grad_norm": 0.42106334402168993,
+      "learning_rate": 0.00017915319185861118,
+      "loss": 1.0382,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2331820231474523,
+      "grad_norm": 0.28217679770550086,
+      "learning_rate": 0.0001791150905012089,
+      "loss": 1.1259,
+      "step": 1212
+    },
+    {
+      "epoch": 0.23337441755598978,
+      "grad_norm": 0.34894691076199397,
+      "learning_rate": 0.00017907695841641713,
+      "loss": 1.1057,
+      "step": 1213
+    },
+    {
+      "epoch": 0.23356681196452728,
+      "grad_norm": 0.2750436859307013,
+      "learning_rate": 0.00017903879561904598,
+      "loss": 1.1406,
+      "step": 1214
+    },
+    {
+      "epoch": 0.23375920637306477,
+      "grad_norm": 0.26620287407774024,
+      "learning_rate": 0.0001790006021239173,
+      "loss": 1.1932,
+      "step": 1215
+    },
+    {
+      "epoch": 0.2339516007816023,
+      "grad_norm": 0.2938806768287713,
+      "learning_rate": 0.00017896237794586513,
+      "loss": 1.1647,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2341439951901398,
+      "grad_norm": 0.2743820663073566,
+      "learning_rate": 0.0001789241230997352,
+      "loss": 1.1365,
+      "step": 1217
+    },
+    {
+      "epoch": 0.23433638959867728,
+      "grad_norm": 0.3692521698288092,
+      "learning_rate": 0.00017888583760038534,
+      "loss": 1.1886,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2345287840072148,
+      "grad_norm": 0.3158175244074631,
+      "learning_rate": 0.00017884752146268512,
+      "loss": 1.1692,
+      "step": 1219
+    },
+    {
+      "epoch": 0.2347211784157523,
+      "grad_norm": 0.2883654846012361,
+      "learning_rate": 0.00017880917470151615,
+      "loss": 1.1261,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2349135728242898,
+      "grad_norm": 0.28889478488936415,
+      "learning_rate": 0.00017877079733177184,
+      "loss": 0.9975,
+      "step": 1221
+    },
+    {
+      "epoch": 0.23510596723282728,
+      "grad_norm": 0.26912644178255984,
+      "learning_rate": 0.0001787323893683575,
+      "loss": 1.1787,
+      "step": 1222
+    },
+    {
+      "epoch": 0.2352983616413648,
+      "grad_norm": 0.3864415141838015,
+      "learning_rate": 0.0001786939508261904,
+      "loss": 1.1553,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2354907560499023,
+      "grad_norm": 0.32914781984423924,
+      "learning_rate": 0.00017865548172019957,
+      "loss": 1.1807,
+      "step": 1224
+    },
+    {
+      "epoch": 0.2356831504584398,
+      "grad_norm": 0.3063745994020058,
+      "learning_rate": 0.00017861698206532596,
+      "loss": 1.1849,
+      "step": 1225
+    },
+    {
+      "epoch": 0.23587554486697732,
+      "grad_norm": 0.2822649345099394,
+      "learning_rate": 0.00017857845187652246,
+      "loss": 1.188,
+      "step": 1226
+    },
+    {
+      "epoch": 0.2360679392755148,
+      "grad_norm": 0.3428920492026431,
+      "learning_rate": 0.00017853989116875375,
+      "loss": 1.048,
+      "step": 1227
+    },
+    {
+      "epoch": 0.2362603336840523,
+      "grad_norm": 0.29465300250942583,
+      "learning_rate": 0.00017850129995699624,
+      "loss": 1.1281,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2364527280925898,
+      "grad_norm": 0.2749752378776852,
+      "learning_rate": 0.00017846267825623845,
+      "loss": 1.1953,
+      "step": 1229
+    },
+    {
+      "epoch": 0.23664512250112732,
+      "grad_norm": 0.2773158783886057,
+      "learning_rate": 0.00017842402608148053,
+      "loss": 1.1553,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2368375169096648,
+      "grad_norm": 0.31520186939135053,
+      "learning_rate": 0.00017838534344773453,
+      "loss": 1.1738,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2370299113182023,
+      "grad_norm": 0.28568878520482865,
+      "learning_rate": 0.00017834663037002443,
+      "loss": 1.0519,
+      "step": 1232
+    },
+    {
+      "epoch": 0.23722230572673983,
+      "grad_norm": 0.30214167610218123,
+      "learning_rate": 0.00017830788686338584,
+      "loss": 1.1558,
+      "step": 1233
+    },
+    {
+      "epoch": 0.23741470013527732,
+      "grad_norm": 0.27876486396876926,
+      "learning_rate": 0.00017826911294286637,
+      "loss": 1.1407,
+      "step": 1234
+    },
+    {
+      "epoch": 0.23760709454381482,
+      "grad_norm": 0.28188499197703776,
+      "learning_rate": 0.0001782303086235253,
+      "loss": 1.1676,
+      "step": 1235
+    },
+    {
+      "epoch": 0.2377994889523523,
+      "grad_norm": 0.274298371565878,
+      "learning_rate": 0.0001781914739204338,
+      "loss": 1.0646,
+      "step": 1236
+    },
+    {
+      "epoch": 0.23799188336088983,
+      "grad_norm": 0.2638538852788373,
+      "learning_rate": 0.00017815260884867484,
+      "loss": 1.1829,
+      "step": 1237
+    },
+    {
+      "epoch": 0.23818427776942733,
+      "grad_norm": 0.30382234224952787,
+      "learning_rate": 0.00017811371342334317,
+      "loss": 1.1374,
+      "step": 1238
+    },
+    {
+      "epoch": 0.23837667217796482,
+      "grad_norm": 0.2847681047396438,
+      "learning_rate": 0.0001780747876595453,
+      "loss": 1.129,
+      "step": 1239
+    },
+    {
+      "epoch": 0.23856906658650234,
+      "grad_norm": 0.27300639332127213,
+      "learning_rate": 0.00017803583157239956,
+      "loss": 1.1228,
+      "step": 1240
+    },
+    {
+      "epoch": 0.23876146099503984,
+      "grad_norm": 0.2702508456930589,
+      "learning_rate": 0.00017799684517703603,
+      "loss": 1.162,
+      "step": 1241
+    },
+    {
+      "epoch": 0.23895385540357733,
+      "grad_norm": 0.2328771461936807,
+      "learning_rate": 0.0001779578284885966,
+      "loss": 1.1438,
+      "step": 1242
+    },
+    {
+      "epoch": 0.23914624981211483,
+      "grad_norm": 0.3637716867076837,
+      "learning_rate": 0.00017791878152223492,
+      "loss": 1.0683,
+      "step": 1243
+    },
+    {
+      "epoch": 0.23933864422065235,
+      "grad_norm": 0.36480632819887904,
+      "learning_rate": 0.00017787970429311632,
+      "loss": 1.1075,
+      "step": 1244
+    },
+    {
+      "epoch": 0.23953103862918984,
+      "grad_norm": 0.24593441584995046,
+      "learning_rate": 0.00017784059681641797,
+      "loss": 1.1435,
+      "step": 1245
+    },
+    {
+      "epoch": 0.23972343303772733,
+      "grad_norm": 0.2853362577993421,
+      "learning_rate": 0.0001778014591073288,
+      "loss": 1.1555,
+      "step": 1246
+    },
+    {
+      "epoch": 0.23991582744626486,
+      "grad_norm": 0.3349418304761187,
+      "learning_rate": 0.00017776229118104942,
+      "loss": 1.0824,
+      "step": 1247
+    },
+    {
+      "epoch": 0.24010822185480235,
+      "grad_norm": 0.33607502766195857,
+      "learning_rate": 0.0001777230930527922,
+      "loss": 1.0934,
+      "step": 1248
+    },
+    {
+      "epoch": 0.24030061626333984,
+      "grad_norm": 0.253362720817648,
+      "learning_rate": 0.00017768386473778123,
+      "loss": 1.0601,
+      "step": 1249
+    },
+    {
+      "epoch": 0.24049301067187734,
+      "grad_norm": 0.29191360658259746,
+      "learning_rate": 0.00017764460625125235,
+      "loss": 1.1672,
+      "step": 1250
+    },
+    {
+      "epoch": 0.24068540508041486,
+      "grad_norm": 0.286059804578617,
+      "learning_rate": 0.0001776053176084531,
+      "loss": 1.0844,
+      "step": 1251
+    },
+    {
+      "epoch": 0.24087779948895235,
+      "grad_norm": 0.2660256018022285,
+      "learning_rate": 0.00017756599882464273,
+      "loss": 1.0799,
+      "step": 1252
+    },
+    {
+      "epoch": 0.24107019389748985,
+      "grad_norm": 0.290677237969466,
+      "learning_rate": 0.00017752664991509224,
+      "loss": 1.111,
+      "step": 1253
+    },
+    {
+      "epoch": 0.24126258830602734,
+      "grad_norm": 0.30940893460643915,
+      "learning_rate": 0.0001774872708950842,
+      "loss": 1.1163,
+      "step": 1254
+    },
+    {
+      "epoch": 0.24145498271456486,
+      "grad_norm": 0.30167890473086945,
+      "learning_rate": 0.0001774478617799131,
+      "loss": 1.2083,
+      "step": 1255
+    },
+    {
+      "epoch": 0.24164737712310236,
+      "grad_norm": 0.3065808480956745,
+      "learning_rate": 0.00017740842258488488,
+      "loss": 1.0402,
+      "step": 1256
+    },
+    {
+      "epoch": 0.24183977153163985,
+      "grad_norm": 0.2708046298496205,
+      "learning_rate": 0.0001773689533253173,
+      "loss": 1.1752,
+      "step": 1257
+    },
+    {
+      "epoch": 0.24203216594017737,
+      "grad_norm": 0.2902593409578126,
+      "learning_rate": 0.00017732945401653977,
+      "loss": 1.2355,
+      "step": 1258
+    },
+    {
+      "epoch": 0.24222456034871487,
+      "grad_norm": 0.27686437648111334,
+      "learning_rate": 0.0001772899246738934,
+      "loss": 1.1647,
+      "step": 1259
+    },
+    {
+      "epoch": 0.24241695475725236,
+      "grad_norm": 0.31381879978772104,
+      "learning_rate": 0.00017725036531273088,
+      "loss": 0.9973,
+      "step": 1260
+    },
+    {
+      "epoch": 0.24260934916578986,
+      "grad_norm": 0.36802195227643747,
+      "learning_rate": 0.00017721077594841662,
+      "loss": 1.2049,
+      "step": 1261
+    },
+    {
+      "epoch": 0.24280174357432738,
+      "grad_norm": 0.37794334516855255,
+      "learning_rate": 0.00017717115659632667,
+      "loss": 1.2405,
+      "step": 1262
+    },
+    {
+      "epoch": 0.24299413798286487,
+      "grad_norm": 0.29779155499018656,
+      "learning_rate": 0.00017713150727184876,
+      "loss": 1.1347,
+      "step": 1263
+    },
+    {
+      "epoch": 0.24318653239140237,
+      "grad_norm": 0.29407342168966677,
+      "learning_rate": 0.00017709182799038222,
+      "loss": 1.1909,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2433789267999399,
+      "grad_norm": 0.29804169723370155,
+      "learning_rate": 0.000177052118767338,
+      "loss": 1.1199,
+      "step": 1265
+    },
+    {
+      "epoch": 0.24357132120847738,
+      "grad_norm": 0.342090125715215,
+      "learning_rate": 0.00017701237961813875,
+      "loss": 1.1875,
+      "step": 1266
+    },
+    {
+      "epoch": 0.24376371561701488,
+      "grad_norm": 0.35629197989307626,
+      "learning_rate": 0.00017697261055821863,
+      "loss": 1.1291,
+      "step": 1267
+    },
+    {
+      "epoch": 0.24395611002555237,
+      "grad_norm": 0.3234468201376417,
+      "learning_rate": 0.00017693281160302356,
+      "loss": 1.134,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2441485044340899,
+      "grad_norm": 0.29092606485392436,
+      "learning_rate": 0.00017689298276801093,
+      "loss": 1.1479,
+      "step": 1269
+    },
+    {
+      "epoch": 0.24434089884262739,
+      "grad_norm": 0.29993750386802515,
+      "learning_rate": 0.00017685312406864984,
+      "loss": 1.1448,
+      "step": 1270
+    },
+    {
+      "epoch": 0.24453329325116488,
+      "grad_norm": 0.37728993263992666,
+      "learning_rate": 0.00017681323552042096,
+      "loss": 1.0735,
+      "step": 1271
+    },
+    {
+      "epoch": 0.2447256876597024,
+      "grad_norm": 0.34114197040493005,
+      "learning_rate": 0.00017677331713881648,
+      "loss": 1.1709,
+      "step": 1272
+    },
+    {
+      "epoch": 0.2449180820682399,
+      "grad_norm": 0.26461087252963844,
+      "learning_rate": 0.0001767333689393403,
+      "loss": 1.0138,
+      "step": 1273
+    },
+    {
+      "epoch": 0.2451104764767774,
+      "grad_norm": 0.3638451559635149,
+      "learning_rate": 0.00017669339093750785,
+      "loss": 1.0947,
+      "step": 1274
+    },
+    {
+      "epoch": 0.24530287088531488,
+      "grad_norm": 0.35659073194040136,
+      "learning_rate": 0.00017665338314884607,
+      "loss": 1.1172,
+      "step": 1275
+    },
+    {
+      "epoch": 0.2454952652938524,
+      "grad_norm": 0.3402194557182349,
+      "learning_rate": 0.00017661334558889357,
+      "loss": 1.1208,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2456876597023899,
+      "grad_norm": 0.37768772187033833,
+      "learning_rate": 0.00017657327827320045,
+      "loss": 1.1458,
+      "step": 1277
+    },
+    {
+      "epoch": 0.2458800541109274,
+      "grad_norm": 0.30279301402497205,
+      "learning_rate": 0.0001765331812173284,
+      "loss": 1.142,
+      "step": 1278
+    },
+    {
+      "epoch": 0.24607244851946491,
+      "grad_norm": 0.3701728235917981,
+      "learning_rate": 0.00017649305443685068,
+      "loss": 1.1281,
+      "step": 1279
+    },
+    {
+      "epoch": 0.2462648429280024,
+      "grad_norm": 0.3389863139016665,
+      "learning_rate": 0.000176452897947352,
+      "loss": 1.0786,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2464572373365399,
+      "grad_norm": 0.2850160393505762,
+      "learning_rate": 0.00017641271176442876,
+      "loss": 1.0966,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2466496317450774,
+      "grad_norm": 0.3070292820238925,
+      "learning_rate": 0.00017637249590368877,
+      "loss": 1.0818,
+      "step": 1282
+    },
+    {
+      "epoch": 0.24684202615361492,
+      "grad_norm": 0.28431255931479377,
+      "learning_rate": 0.00017633225038075142,
+      "loss": 1.0456,
+      "step": 1283
+    },
+    {
+      "epoch": 0.2470344205621524,
+      "grad_norm": 0.46514505409715884,
+      "learning_rate": 0.00017629197521124755,
+      "loss": 1.1113,
+      "step": 1284
+    },
+    {
+      "epoch": 0.2472268149706899,
+      "grad_norm": 0.3219692065437815,
+      "learning_rate": 0.00017625167041081965,
+      "loss": 1.1805,
+      "step": 1285
+    },
+    {
+      "epoch": 0.24741920937922743,
+      "grad_norm": 0.3857745502111839,
+      "learning_rate": 0.00017621133599512163,
+      "loss": 1.1315,
+      "step": 1286
+    },
+    {
+      "epoch": 0.24761160378776492,
+      "grad_norm": 0.30358705433041244,
+      "learning_rate": 0.00017617097197981889,
+      "loss": 1.1414,
+      "step": 1287
+    },
+    {
+      "epoch": 0.24780399819630242,
+      "grad_norm": 0.3239256918533029,
+      "learning_rate": 0.0001761305783805883,
+      "loss": 1.1976,
+      "step": 1288
+    },
+    {
+      "epoch": 0.2479963926048399,
+      "grad_norm": 0.28888282191266984,
+      "learning_rate": 0.00017609015521311835,
+      "loss": 1.0658,
+      "step": 1289
+    },
+    {
+      "epoch": 0.24818878701337743,
+      "grad_norm": 0.29897663650648887,
+      "learning_rate": 0.0001760497024931089,
+      "loss": 1.1095,
+      "step": 1290
+    },
+    {
+      "epoch": 0.24838118142191493,
+      "grad_norm": 0.36423970435303815,
+      "learning_rate": 0.00017600922023627136,
+      "loss": 1.129,
+      "step": 1291
+    },
+    {
+      "epoch": 0.24857357583045242,
+      "grad_norm": 0.3246330646747353,
+      "learning_rate": 0.0001759687084583285,
+      "loss": 1.1871,
+      "step": 1292
+    },
+    {
+      "epoch": 0.24876597023898994,
+      "grad_norm": 0.27683582059501827,
+      "learning_rate": 0.0001759281671750147,
+      "loss": 1.0987,
+      "step": 1293
+    },
+    {
+      "epoch": 0.24895836464752744,
+      "grad_norm": 0.3391467350342309,
+      "learning_rate": 0.00017588759640207563,
+      "loss": 1.0394,
+      "step": 1294
+    },
+    {
+      "epoch": 0.24915075905606493,
+      "grad_norm": 0.28629417027256443,
+      "learning_rate": 0.00017584699615526858,
+      "loss": 1.1892,
+      "step": 1295
+    },
+    {
+      "epoch": 0.24934315346460242,
+      "grad_norm": 0.29617680549837705,
+      "learning_rate": 0.00017580636645036225,
+      "loss": 1.0634,
+      "step": 1296
+    },
+    {
+      "epoch": 0.24953554787313995,
+      "grad_norm": 0.3241373057473967,
+      "learning_rate": 0.0001757657073031367,
+      "loss": 1.1073,
+      "step": 1297
+    },
+    {
+      "epoch": 0.24972794228167744,
+      "grad_norm": 0.2741305975028379,
+      "learning_rate": 0.00017572501872938342,
+      "loss": 1.1571,
+      "step": 1298
+    },
+    {
+      "epoch": 0.24992033669021493,
+      "grad_norm": 0.3541371188862645,
+      "learning_rate": 0.0001756843007449055,
+      "loss": 1.1162,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2501127310987524,
+      "grad_norm": 0.2936939680371704,
+      "learning_rate": 0.00017564355336551727,
+      "loss": 1.0505,
+      "step": 1300
+    },
+    {
+      "epoch": 0.25030512550728995,
+      "grad_norm": 0.2949885721381335,
+      "learning_rate": 0.00017560277660704453,
+      "loss": 1.112,
+      "step": 1301
+    },
+    {
+      "epoch": 0.25049751991582747,
+      "grad_norm": 0.2750557636829198,
+      "learning_rate": 0.00017556197048532456,
+      "loss": 1.1058,
+      "step": 1302
+    },
+    {
+      "epoch": 0.25068991432436494,
+      "grad_norm": 0.264142176765815,
+      "learning_rate": 0.00017552113501620594,
+      "loss": 1.1379,
+      "step": 1303
+    },
+    {
+      "epoch": 0.25088230873290246,
+      "grad_norm": 0.42570119279794666,
+      "learning_rate": 0.00017548027021554874,
+      "loss": 1.1599,
+      "step": 1304
+    },
+    {
+      "epoch": 0.2510747031414399,
+      "grad_norm": 0.2707421360257628,
+      "learning_rate": 0.0001754393760992243,
+      "loss": 1.1376,
+      "step": 1305
+    },
+    {
+      "epoch": 0.25126709754997745,
+      "grad_norm": 0.34846977453024036,
+      "learning_rate": 0.00017539845268311547,
+      "loss": 1.0918,
+      "step": 1306
+    },
+    {
+      "epoch": 0.25145949195851497,
+      "grad_norm": 0.3149632413586983,
+      "learning_rate": 0.00017535749998311647,
+      "loss": 1.0166,
+      "step": 1307
+    },
+    {
+      "epoch": 0.25165188636705244,
+      "grad_norm": 0.35554953079039586,
+      "learning_rate": 0.0001753165180151328,
+      "loss": 1.0824,
+      "step": 1308
+    },
+    {
+      "epoch": 0.25184428077558996,
+      "grad_norm": 0.2896831942210788,
+      "learning_rate": 0.0001752755067950814,
+      "loss": 1.1177,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2520366751841275,
+      "grad_norm": 0.26323499493106395,
+      "learning_rate": 0.0001752344663388906,
+      "loss": 1.1737,
+      "step": 1310
+    },
+    {
+      "epoch": 0.25222906959266495,
+      "grad_norm": 0.3186714366459116,
+      "learning_rate": 0.00017519339666249997,
+      "loss": 1.0434,
+      "step": 1311
+    },
+    {
+      "epoch": 0.25242146400120247,
+      "grad_norm": 0.3465213490943624,
+      "learning_rate": 0.00017515229778186054,
+      "loss": 1.0519,
+      "step": 1312
+    },
+    {
+      "epoch": 0.25261385840974,
+      "grad_norm": 0.334036168381196,
+      "learning_rate": 0.0001751111697129346,
+      "loss": 1.1315,
+      "step": 1313
+    },
+    {
+      "epoch": 0.25280625281827745,
+      "grad_norm": 0.3621019937457814,
+      "learning_rate": 0.00017507001247169587,
+      "loss": 1.1823,
+      "step": 1314
+    },
+    {
+      "epoch": 0.252998647226815,
+      "grad_norm": 0.26110746967417664,
+      "learning_rate": 0.00017502882607412933,
+      "loss": 1.0819,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2531910416353525,
+      "grad_norm": 0.33215603669557836,
+      "learning_rate": 0.00017498761053623128,
+      "loss": 1.0696,
+      "step": 1316
+    },
+    {
+      "epoch": 0.25338343604388996,
+      "grad_norm": 0.3089901410957134,
+      "learning_rate": 0.0001749463658740094,
+      "loss": 1.1566,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2535758304524275,
+      "grad_norm": 0.2676814671588705,
+      "learning_rate": 0.00017490509210348263,
+      "loss": 1.1228,
+      "step": 1318
+    },
+    {
+      "epoch": 0.25376822486096495,
+      "grad_norm": 0.315852099385859,
+      "learning_rate": 0.0001748637892406812,
+      "loss": 1.1903,
+      "step": 1319
+    },
+    {
+      "epoch": 0.2539606192695025,
+      "grad_norm": 0.31970930854664487,
+      "learning_rate": 0.0001748224573016467,
+      "loss": 1.0864,
+      "step": 1320
+    },
+    {
+      "epoch": 0.25415301367804,
+      "grad_norm": 0.27790099021632264,
+      "learning_rate": 0.00017478109630243195,
+      "loss": 1.0916,
+      "step": 1321
+    },
+    {
+      "epoch": 0.25434540808657746,
+      "grad_norm": 0.30558296220478814,
+      "learning_rate": 0.0001747397062591011,
+      "loss": 1.1287,
+      "step": 1322
+    },
+    {
+      "epoch": 0.254537802495115,
+      "grad_norm": 0.2702814127579277,
+      "learning_rate": 0.00017469828718772958,
+      "loss": 1.1779,
+      "step": 1323
+    },
+    {
+      "epoch": 0.2547301969036525,
+      "grad_norm": 0.297505295627331,
+      "learning_rate": 0.00017465683910440405,
+      "loss": 1.1328,
+      "step": 1324
+    },
+    {
+      "epoch": 0.25492259131219,
+      "grad_norm": 0.26769827486566683,
+      "learning_rate": 0.00017461536202522247,
+      "loss": 1.1379,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2551149857207275,
+      "grad_norm": 0.355058893755735,
+      "learning_rate": 0.0001745738559662941,
+      "loss": 1.1411,
+      "step": 1326
+    },
+    {
+      "epoch": 0.255307380129265,
+      "grad_norm": 0.309347595436928,
+      "learning_rate": 0.00017453232094373936,
+      "loss": 1.1696,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2554997745378025,
+      "grad_norm": 0.4476075045732932,
+      "learning_rate": 0.00017449075697369005,
+      "loss": 1.1806,
+      "step": 1328
+    },
+    {
+      "epoch": 0.25569216894634,
+      "grad_norm": 0.3406162517718127,
+      "learning_rate": 0.00017444916407228904,
+      "loss": 1.186,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2558845633548775,
+      "grad_norm": 0.24096468980054248,
+      "learning_rate": 0.0001744075422556906,
+      "loss": 1.1641,
+      "step": 1330
+    },
+    {
+      "epoch": 0.256076957763415,
+      "grad_norm": 0.30086294001288766,
+      "learning_rate": 0.00017436589154006012,
+      "loss": 1.0826,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2562693521719525,
+      "grad_norm": 0.28676636236735553,
+      "learning_rate": 0.00017432421194157432,
+      "loss": 1.1188,
+      "step": 1332
+    },
+    {
+      "epoch": 0.25646174658049,
+      "grad_norm": 0.34417627273508167,
+      "learning_rate": 0.000174282503476421,
+      "loss": 1.0832,
+      "step": 1333
+    },
+    {
+      "epoch": 0.2566541409890275,
+      "grad_norm": 0.35048610341297504,
+      "learning_rate": 0.00017424076616079934,
+      "loss": 1.1534,
+      "step": 1334
+    },
+    {
+      "epoch": 0.256846535397565,
+      "grad_norm": 0.3979590056134015,
+      "learning_rate": 0.00017419900001091953,
+      "loss": 1.1599,
+      "step": 1335
+    },
+    {
+      "epoch": 0.2570389298061025,
+      "grad_norm": 0.33763286483264704,
+      "learning_rate": 0.00017415720504300314,
+      "loss": 1.0098,
+      "step": 1336
+    },
+    {
+      "epoch": 0.25723132421464,
+      "grad_norm": 0.3576401448908834,
+      "learning_rate": 0.00017411538127328283,
+      "loss": 1.1871,
+      "step": 1337
+    },
+    {
+      "epoch": 0.25742371862317753,
+      "grad_norm": 0.35276211463307466,
+      "learning_rate": 0.00017407352871800244,
+      "loss": 1.0371,
+      "step": 1338
+    },
+    {
+      "epoch": 0.257616113031715,
+      "grad_norm": 0.29221167111412355,
+      "learning_rate": 0.00017403164739341706,
+      "loss": 1.1255,
+      "step": 1339
+    },
+    {
+      "epoch": 0.2578085074402525,
+      "grad_norm": 0.36273868784627933,
+      "learning_rate": 0.0001739897373157929,
+      "loss": 1.1714,
+      "step": 1340
+    },
+    {
+      "epoch": 0.25800090184879004,
+      "grad_norm": 0.2755318015574914,
+      "learning_rate": 0.00017394779850140734,
+      "loss": 1.1595,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2581932962573275,
+      "grad_norm": 0.28706153346378777,
+      "learning_rate": 0.00017390583096654896,
+      "loss": 1.0942,
+      "step": 1342
+    },
+    {
+      "epoch": 0.25838569066586503,
+      "grad_norm": 0.3775394537185136,
+      "learning_rate": 0.00017386383472751743,
+      "loss": 1.1813,
+      "step": 1343
+    },
+    {
+      "epoch": 0.25857808507440255,
+      "grad_norm": 0.3323319561680617,
+      "learning_rate": 0.00017382180980062364,
+      "loss": 1.0759,
+      "step": 1344
+    },
+    {
+      "epoch": 0.25877047948294,
+      "grad_norm": 0.255644721439666,
+      "learning_rate": 0.00017377975620218953,
+      "loss": 1.0823,
+      "step": 1345
+    },
+    {
+      "epoch": 0.25896287389147754,
+      "grad_norm": 0.2841415414848638,
+      "learning_rate": 0.00017373767394854834,
+      "loss": 1.1767,
+      "step": 1346
+    },
+    {
+      "epoch": 0.259155268300015,
+      "grad_norm": 0.3228337200242913,
+      "learning_rate": 0.00017369556305604422,
+      "loss": 1.0592,
+      "step": 1347
+    },
+    {
+      "epoch": 0.25934766270855253,
+      "grad_norm": 0.26735867803353425,
+      "learning_rate": 0.00017365342354103264,
+      "loss": 1.1682,
+      "step": 1348
+    },
+    {
+      "epoch": 0.25954005711709005,
+      "grad_norm": 0.29059786890977607,
+      "learning_rate": 0.00017361125541988002,
+      "loss": 1.1384,
+      "step": 1349
+    },
+    {
+      "epoch": 0.2597324515256275,
+      "grad_norm": 0.2771992032868704,
+      "learning_rate": 0.00017356905870896406,
+      "loss": 1.1535,
+      "step": 1350
+    },
+    {
+      "epoch": 0.25992484593416504,
+      "grad_norm": 0.24527673864737073,
+      "learning_rate": 0.0001735268334246734,
+      "loss": 1.1916,
+      "step": 1351
+    },
+    {
+      "epoch": 0.26011724034270256,
+      "grad_norm": 0.28637376823554156,
+      "learning_rate": 0.0001734845795834079,
+      "loss": 1.1167,
+      "step": 1352
+    },
+    {
+      "epoch": 0.26030963475124,
+      "grad_norm": 0.29057489302019385,
+      "learning_rate": 0.00017344229720157845,
+      "loss": 1.043,
+      "step": 1353
+    },
+    {
+      "epoch": 0.26050202915977755,
+      "grad_norm": 0.3455368633759316,
+      "learning_rate": 0.00017339998629560706,
+      "loss": 1.1397,
+      "step": 1354
+    },
+    {
+      "epoch": 0.26069442356831507,
+      "grad_norm": 0.32360951955328404,
+      "learning_rate": 0.00017335764688192678,
+      "loss": 1.0493,
+      "step": 1355
+    },
+    {
+      "epoch": 0.26088681797685254,
+      "grad_norm": 0.34124170919272556,
+      "learning_rate": 0.0001733152789769817,
+      "loss": 1.0926,
+      "step": 1356
+    },
+    {
+      "epoch": 0.26107921238539006,
+      "grad_norm": 0.353349847992069,
+      "learning_rate": 0.00017327288259722713,
+      "loss": 1.0655,
+      "step": 1357
+    },
+    {
+      "epoch": 0.2612716067939275,
+      "grad_norm": 0.2622086953275031,
+      "learning_rate": 0.00017323045775912926,
+      "loss": 1.193,
+      "step": 1358
+    },
+    {
+      "epoch": 0.26146400120246505,
+      "grad_norm": 0.39105781315914395,
+      "learning_rate": 0.00017318800447916542,
+      "loss": 1.1149,
+      "step": 1359
+    },
+    {
+      "epoch": 0.26165639561100257,
+      "grad_norm": 0.30976968640013536,
+      "learning_rate": 0.00017314552277382402,
+      "loss": 1.0476,
+      "step": 1360
+    },
+    {
+      "epoch": 0.26184879001954003,
+      "grad_norm": 0.26619682359698926,
+      "learning_rate": 0.00017310301265960446,
+      "loss": 1.1216,
+      "step": 1361
+    },
+    {
+      "epoch": 0.26204118442807756,
+      "grad_norm": 0.3209847711273689,
+      "learning_rate": 0.00017306047415301706,
+      "loss": 1.0611,
+      "step": 1362
+    },
+    {
+      "epoch": 0.2622335788366151,
+      "grad_norm": 0.2725241445577398,
+      "learning_rate": 0.00017301790727058345,
+      "loss": 1.0797,
+      "step": 1363
+    },
+    {
+      "epoch": 0.26242597324515254,
+      "grad_norm": 0.44269612569447203,
+      "learning_rate": 0.000172975312028836,
+      "loss": 1.0151,
+      "step": 1364
+    },
+    {
+      "epoch": 0.26261836765369007,
+      "grad_norm": 0.3214987078674056,
+      "learning_rate": 0.00017293268844431828,
+      "loss": 1.1846,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2628107620622276,
+      "grad_norm": 0.2755490965984195,
+      "learning_rate": 0.00017289003653358473,
+      "loss": 1.1428,
+      "step": 1366
+    },
+    {
+      "epoch": 0.26300315647076505,
+      "grad_norm": 0.3236620776652681,
+      "learning_rate": 0.00017284735631320093,
+      "loss": 1.0628,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2631955508793026,
+      "grad_norm": 0.43103719426319725,
+      "learning_rate": 0.00017280464779974335,
+      "loss": 1.1225,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2633879452878401,
+      "grad_norm": 0.29182915995559056,
+      "learning_rate": 0.0001727619110097995,
+      "loss": 1.1525,
+      "step": 1369
+    },
+    {
+      "epoch": 0.26358033969637756,
+      "grad_norm": 0.3429150854124409,
+      "learning_rate": 0.00017271914595996782,
+      "loss": 1.1199,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2637727341049151,
+      "grad_norm": 0.2637758803009464,
+      "learning_rate": 0.00017267635266685782,
+      "loss": 1.0769,
+      "step": 1371
+    },
+    {
+      "epoch": 0.26396512851345255,
+      "grad_norm": 0.2935984788774275,
+      "learning_rate": 0.0001726335311470899,
+      "loss": 1.0605,
+      "step": 1372
+    },
+    {
+      "epoch": 0.2641575229219901,
+      "grad_norm": 0.3023912442325294,
+      "learning_rate": 0.00017259068141729542,
+      "loss": 1.1168,
+      "step": 1373
+    },
+    {
+      "epoch": 0.2643499173305276,
+      "grad_norm": 0.3315555055672322,
+      "learning_rate": 0.00017254780349411675,
+      "loss": 1.1239,
+      "step": 1374
+    },
+    {
+      "epoch": 0.26454231173906506,
+      "grad_norm": 0.3352022510656215,
+      "learning_rate": 0.00017250489739420718,
+      "loss": 1.0892,
+      "step": 1375
+    },
+    {
+      "epoch": 0.2647347061476026,
+      "grad_norm": 0.37205721319154095,
+      "learning_rate": 0.00017246196313423093,
+      "loss": 1.0428,
+      "step": 1376
+    },
+    {
+      "epoch": 0.2649271005561401,
+      "grad_norm": 0.3437422788419956,
+      "learning_rate": 0.00017241900073086317,
+      "loss": 1.0495,
+      "step": 1377
+    },
+    {
+      "epoch": 0.26511949496467757,
+      "grad_norm": 0.2800951366405879,
+      "learning_rate": 0.00017237601020079002,
+      "loss": 1.1672,
+      "step": 1378
+    },
+    {
+      "epoch": 0.2653118893732151,
+      "grad_norm": 0.33395863707201673,
+      "learning_rate": 0.0001723329915607085,
+      "loss": 1.1359,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2655042837817526,
+      "grad_norm": 0.27940732767555193,
+      "learning_rate": 0.00017228994482732651,
+      "loss": 1.2608,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2656966781902901,
+      "grad_norm": 0.4140561343379944,
+      "learning_rate": 0.000172246870017363,
+      "loss": 1.0091,
+      "step": 1381
+    },
+    {
+      "epoch": 0.2658890725988276,
+      "grad_norm": 0.31906365134682535,
+      "learning_rate": 0.00017220376714754764,
+      "loss": 1.031,
+      "step": 1382
+    },
+    {
+      "epoch": 0.2660814670073651,
+      "grad_norm": 0.2629626181918077,
+      "learning_rate": 0.00017216063623462112,
+      "loss": 1.1009,
+      "step": 1383
+    },
+    {
+      "epoch": 0.2662738614159026,
+      "grad_norm": 0.3459677194379836,
+      "learning_rate": 0.00017211747729533502,
+      "loss": 1.1798,
+      "step": 1384
+    },
+    {
+      "epoch": 0.2664662558244401,
+      "grad_norm": 0.26964009886145485,
+      "learning_rate": 0.00017207429034645175,
+      "loss": 1.1205,
+      "step": 1385
+    },
+    {
+      "epoch": 0.2666586502329776,
+      "grad_norm": 0.28465305019450465,
+      "learning_rate": 0.0001720310754047446,
+      "loss": 1.1774,
+      "step": 1386
+    },
+    {
+      "epoch": 0.2668510446415151,
+      "grad_norm": 0.3119888541964351,
+      "learning_rate": 0.00017198783248699779,
+      "loss": 1.1608,
+      "step": 1387
+    },
+    {
+      "epoch": 0.2670434390500526,
+      "grad_norm": 0.3944044220649418,
+      "learning_rate": 0.00017194456161000635,
+      "loss": 1.1544,
+      "step": 1388
+    },
+    {
+      "epoch": 0.2672358334585901,
+      "grad_norm": 0.3697790963912528,
+      "learning_rate": 0.0001719012627905762,
+      "loss": 1.0404,
+      "step": 1389
+    },
+    {
+      "epoch": 0.2674282278671276,
+      "grad_norm": 0.28279454414641186,
+      "learning_rate": 0.00017185793604552408,
+      "loss": 1.012,
+      "step": 1390
+    },
+    {
+      "epoch": 0.26762062227566513,
+      "grad_norm": 0.29364721233012775,
+      "learning_rate": 0.00017181458139167764,
+      "loss": 1.0722,
+      "step": 1391
+    },
+    {
+      "epoch": 0.2678130166842026,
+      "grad_norm": 0.30657863225904525,
+      "learning_rate": 0.00017177119884587535,
+      "loss": 1.1353,
+      "step": 1392
+    },
+    {
+      "epoch": 0.2680054110927401,
+      "grad_norm": 0.3001832172375664,
+      "learning_rate": 0.0001717277884249664,
+      "loss": 1.1595,
+      "step": 1393
+    },
+    {
+      "epoch": 0.26819780550127764,
+      "grad_norm": 0.2867900108041073,
+      "learning_rate": 0.00017168435014581094,
+      "loss": 1.1331,
+      "step": 1394
+    },
+    {
+      "epoch": 0.2683901999098151,
+      "grad_norm": 0.3194360701346741,
+      "learning_rate": 0.0001716408840252799,
+      "loss": 1.1501,
+      "step": 1395
+    },
+    {
+      "epoch": 0.26858259431835263,
+      "grad_norm": 0.3434192596775916,
+      "learning_rate": 0.00017159739008025505,
+      "loss": 1.0014,
+      "step": 1396
+    },
+    {
+      "epoch": 0.26877498872689015,
+      "grad_norm": 0.25958055228442156,
+      "learning_rate": 0.0001715538683276289,
+      "loss": 1.1398,
+      "step": 1397
+    },
+    {
+      "epoch": 0.2689673831354276,
+      "grad_norm": 0.2800696009487632,
+      "learning_rate": 0.0001715103187843048,
+      "loss": 1.1148,
+      "step": 1398
+    },
+    {
+      "epoch": 0.26915977754396514,
+      "grad_norm": 0.2666934809970182,
+      "learning_rate": 0.0001714667414671969,
+      "loss": 1.1001,
+      "step": 1399
+    },
+    {
+      "epoch": 0.2693521719525026,
+      "grad_norm": 0.3300122963342882,
+      "learning_rate": 0.0001714231363932301,
+      "loss": 1.1367,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2695445663610401,
+      "grad_norm": 0.288944250780435,
+      "learning_rate": 0.00017137950357934016,
+      "loss": 1.1034,
+      "step": 1401
+    },
+    {
+      "epoch": 0.26973696076957765,
+      "grad_norm": 0.3945039390178446,
+      "learning_rate": 0.00017133584304247353,
+      "loss": 1.2764,
+      "step": 1402
+    },
+    {
+      "epoch": 0.2699293551781151,
+      "grad_norm": 0.34469031247437065,
+      "learning_rate": 0.00017129215479958745,
+      "loss": 1.1197,
+      "step": 1403
+    },
+    {
+      "epoch": 0.27012174958665264,
+      "grad_norm": 0.3067648446670519,
+      "learning_rate": 0.00017124843886765,
+      "loss": 1.0654,
+      "step": 1404
+    },
+    {
+      "epoch": 0.27031414399519016,
+      "grad_norm": 0.27869117175923364,
+      "learning_rate": 0.00017120469526363982,
+      "loss": 1.1095,
+      "step": 1405
+    },
+    {
+      "epoch": 0.2705065384037276,
+      "grad_norm": 0.33802316017498774,
+      "learning_rate": 0.00017116092400454652,
+      "loss": 1.0426,
+      "step": 1406
+    },
+    {
+      "epoch": 0.27069893281226515,
+      "grad_norm": 0.3155777695709668,
+      "learning_rate": 0.00017111712510737035,
+      "loss": 1.1854,
+      "step": 1407
+    },
+    {
+      "epoch": 0.27089132722080267,
+      "grad_norm": 0.30751416378092516,
+      "learning_rate": 0.00017107329858912225,
+      "loss": 1.0578,
+      "step": 1408
+    },
+    {
+      "epoch": 0.27108372162934014,
+      "grad_norm": 0.3518317013774291,
+      "learning_rate": 0.00017102944446682395,
+      "loss": 1.0176,
+      "step": 1409
+    },
+    {
+      "epoch": 0.27127611603787766,
+      "grad_norm": 0.2340350896227295,
+      "learning_rate": 0.00017098556275750787,
+      "loss": 1.0734,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2714685104464152,
+      "grad_norm": 0.2890537821398866,
+      "learning_rate": 0.00017094165347821723,
+      "loss": 1.0906,
+      "step": 1411
+    },
+    {
+      "epoch": 0.27166090485495265,
+      "grad_norm": 0.3395019582990811,
+      "learning_rate": 0.00017089771664600582,
+      "loss": 1.1152,
+      "step": 1412
+    },
+    {
+      "epoch": 0.27185329926349017,
+      "grad_norm": 0.3373935525833467,
+      "learning_rate": 0.00017085375227793818,
+      "loss": 1.0581,
+      "step": 1413
+    },
+    {
+      "epoch": 0.27204569367202763,
+      "grad_norm": 0.29701743997662994,
+      "learning_rate": 0.00017080976039108965,
+      "loss": 1.0521,
+      "step": 1414
+    },
+    {
+      "epoch": 0.27223808808056515,
+      "grad_norm": 0.2540218390388427,
+      "learning_rate": 0.00017076574100254613,
+      "loss": 1.1097,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2724304824891027,
+      "grad_norm": 0.3182432027755049,
+      "learning_rate": 0.00017072169412940422,
+      "loss": 1.0806,
+      "step": 1416
+    },
+    {
+      "epoch": 0.27262287689764014,
+      "grad_norm": 0.3914213197454,
+      "learning_rate": 0.00017067761978877121,
+      "loss": 1.0403,
+      "step": 1417
+    },
+    {
+      "epoch": 0.27281527130617766,
+      "grad_norm": 0.3269191412985446,
+      "learning_rate": 0.00017063351799776513,
+      "loss": 1.119,
+      "step": 1418
+    },
+    {
+      "epoch": 0.2730076657147152,
+      "grad_norm": 0.3803367678562637,
+      "learning_rate": 0.00017058938877351456,
+      "loss": 1.1058,
+      "step": 1419
+    },
+    {
+      "epoch": 0.27320006012325265,
+      "grad_norm": 0.30849186310063365,
+      "learning_rate": 0.00017054523213315878,
+      "loss": 1.1424,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2733924545317902,
+      "grad_norm": 0.2703485857002882,
+      "learning_rate": 0.00017050104809384774,
+      "loss": 1.1049,
+      "step": 1421
+    },
+    {
+      "epoch": 0.2735848489403277,
+      "grad_norm": 0.2921528574845525,
+      "learning_rate": 0.000170456836672742,
+      "loss": 1.0599,
+      "step": 1422
+    },
+    {
+      "epoch": 0.27377724334886516,
+      "grad_norm": 0.3076258718874411,
+      "learning_rate": 0.00017041259788701281,
+      "loss": 1.1367,
+      "step": 1423
+    },
+    {
+      "epoch": 0.2739696377574027,
+      "grad_norm": 0.30124214318469533,
+      "learning_rate": 0.0001703683317538419,
+      "loss": 1.1517,
+      "step": 1424
+    },
+    {
+      "epoch": 0.27416203216594015,
+      "grad_norm": 0.31177408413252455,
+      "learning_rate": 0.00017032403829042182,
+      "loss": 1.1415,
+      "step": 1425
+    },
+    {
+      "epoch": 0.27435442657447767,
+      "grad_norm": 0.30004432972090933,
+      "learning_rate": 0.00017027971751395563,
+      "loss": 1.1988,
+      "step": 1426
+    },
+    {
+      "epoch": 0.2745468209830152,
+      "grad_norm": 0.3106744326241377,
+      "learning_rate": 0.00017023536944165698,
+      "loss": 1.1546,
+      "step": 1427
+    },
+    {
+      "epoch": 0.27473921539155266,
+      "grad_norm": 0.2851602238507599,
+      "learning_rate": 0.00017019099409075014,
+      "loss": 1.086,
+      "step": 1428
+    },
+    {
+      "epoch": 0.2749316098000902,
+      "grad_norm": 0.2634909899667367,
+      "learning_rate": 0.00017014659147847002,
+      "loss": 1.1056,
+      "step": 1429
+    },
+    {
+      "epoch": 0.2751240042086277,
+      "grad_norm": 0.3188866896136964,
+      "learning_rate": 0.00017010216162206208,
+      "loss": 1.1633,
+      "step": 1430
+    },
+    {
+      "epoch": 0.27531639861716517,
+      "grad_norm": 0.28836837788970254,
+      "learning_rate": 0.00017005770453878234,
+      "loss": 1.1182,
+      "step": 1431
+    },
+    {
+      "epoch": 0.2755087930257027,
+      "grad_norm": 0.2975450328661655,
+      "learning_rate": 0.00017001322024589742,
+      "loss": 1.0936,
+      "step": 1432
+    },
+    {
+      "epoch": 0.2757011874342402,
+      "grad_norm": 0.3382751692849956,
+      "learning_rate": 0.00016996870876068452,
+      "loss": 1.1345,
+      "step": 1433
+    },
+    {
+      "epoch": 0.2758935818427777,
+      "grad_norm": 0.34107771338697157,
+      "learning_rate": 0.00016992417010043142,
+      "loss": 1.1076,
+      "step": 1434
+    },
+    {
+      "epoch": 0.2760859762513152,
+      "grad_norm": 0.27712803587893164,
+      "learning_rate": 0.00016987960428243638,
+      "loss": 1.0613,
+      "step": 1435
+    },
+    {
+      "epoch": 0.2762783706598527,
+      "grad_norm": 0.34252903826358966,
+      "learning_rate": 0.00016983501132400824,
+      "loss": 1.1542,
+      "step": 1436
+    },
+    {
+      "epoch": 0.2764707650683902,
+      "grad_norm": 0.3017579731220398,
+      "learning_rate": 0.00016979039124246642,
+      "loss": 1.0011,
+      "step": 1437
+    },
+    {
+      "epoch": 0.2766631594769277,
+      "grad_norm": 0.3650463571416283,
+      "learning_rate": 0.00016974574405514083,
+      "loss": 1.0774,
+      "step": 1438
+    },
+    {
+      "epoch": 0.2768555538854652,
+      "grad_norm": 0.34567173216330377,
+      "learning_rate": 0.00016970106977937192,
+      "loss": 1.1283,
+      "step": 1439
+    },
+    {
+      "epoch": 0.2770479482940027,
+      "grad_norm": 0.3008617277171387,
+      "learning_rate": 0.0001696563684325107,
+      "loss": 1.091,
+      "step": 1440
+    },
+    {
+      "epoch": 0.2772403427025402,
+      "grad_norm": 0.3210892263127466,
+      "learning_rate": 0.00016961164003191862,
+      "loss": 1.0207,
+      "step": 1441
+    },
+    {
+      "epoch": 0.2774327371110777,
+      "grad_norm": 0.4125838840831622,
+      "learning_rate": 0.00016956688459496765,
+      "loss": 1.1289,
+      "step": 1442
+    },
+    {
+      "epoch": 0.2776251315196152,
+      "grad_norm": 0.28647034974871216,
+      "learning_rate": 0.00016952210213904038,
+      "loss": 1.1797,
+      "step": 1443
+    },
+    {
+      "epoch": 0.27781752592815273,
+      "grad_norm": 0.28494862305196333,
+      "learning_rate": 0.0001694772926815297,
+      "loss": 1.0654,
+      "step": 1444
+    },
+    {
+      "epoch": 0.2780099203366902,
+      "grad_norm": 0.2839765423330028,
+      "learning_rate": 0.00016943245623983916,
+      "loss": 1.0816,
+      "step": 1445
+    },
+    {
+      "epoch": 0.2782023147452277,
+      "grad_norm": 0.31956143988872515,
+      "learning_rate": 0.00016938759283138268,
+      "loss": 1.172,
+      "step": 1446
+    },
+    {
+      "epoch": 0.27839470915376524,
+      "grad_norm": 0.34606961562792615,
+      "learning_rate": 0.00016934270247358468,
+      "loss": 1.1393,
+      "step": 1447
+    },
+    {
+      "epoch": 0.2785871035623027,
+      "grad_norm": 0.2799991187980808,
+      "learning_rate": 0.00016929778518388007,
+      "loss": 1.073,
+      "step": 1448
+    },
+    {
+      "epoch": 0.27877949797084023,
+      "grad_norm": 0.2910570538009604,
+      "learning_rate": 0.00016925284097971425,
+      "loss": 1.0928,
+      "step": 1449
+    },
+    {
+      "epoch": 0.27897189237937775,
+      "grad_norm": 0.2681833976207475,
+      "learning_rate": 0.00016920786987854294,
+      "loss": 1.0833,
+      "step": 1450
+    },
+    {
+      "epoch": 0.2791642867879152,
+      "grad_norm": 0.3004191624630394,
+      "learning_rate": 0.00016916287189783252,
+      "loss": 1.0586,
+      "step": 1451
+    },
+    {
+      "epoch": 0.27935668119645274,
+      "grad_norm": 0.2871753365160926,
+      "learning_rate": 0.00016911784705505957,
+      "loss": 1.0896,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2795490756049902,
+      "grad_norm": 0.3046531488285525,
+      "learning_rate": 0.0001690727953677113,
+      "loss": 1.0628,
+      "step": 1453
+    },
+    {
+      "epoch": 0.2797414700135277,
+      "grad_norm": 0.2907399602928292,
+      "learning_rate": 0.00016902771685328523,
+      "loss": 1.1626,
+      "step": 1454
+    },
+    {
+      "epoch": 0.27993386442206525,
+      "grad_norm": 0.3062179049396391,
+      "learning_rate": 0.00016898261152928931,
+      "loss": 1.1066,
+      "step": 1455
+    },
+    {
+      "epoch": 0.2801262588306027,
+      "grad_norm": 0.3351688303157574,
+      "learning_rate": 0.00016893747941324196,
+      "loss": 1.0735,
+      "step": 1456
+    },
+    {
+      "epoch": 0.28031865323914024,
+      "grad_norm": 0.3098324602502187,
+      "learning_rate": 0.000168892320522672,
+      "loss": 1.0838,
+      "step": 1457
+    },
+    {
+      "epoch": 0.28051104764767776,
+      "grad_norm": 0.3733944528578597,
+      "learning_rate": 0.00016884713487511858,
+      "loss": 1.19,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2807034420562152,
+      "grad_norm": 0.2856005012087916,
+      "learning_rate": 0.0001688019224881313,
+      "loss": 1.0703,
+      "step": 1459
+    },
+    {
+      "epoch": 0.28089583646475275,
+      "grad_norm": 0.36357730146748884,
+      "learning_rate": 0.00016875668337927013,
+      "loss": 1.2372,
+      "step": 1460
+    },
+    {
+      "epoch": 0.28108823087329027,
+      "grad_norm": 0.31883751969703866,
+      "learning_rate": 0.00016871141756610544,
+      "loss": 1.1341,
+      "step": 1461
+    },
+    {
+      "epoch": 0.28128062528182773,
+      "grad_norm": 0.24909921113461664,
+      "learning_rate": 0.0001686661250662179,
+      "loss": 1.0997,
+      "step": 1462
+    },
+    {
+      "epoch": 0.28147301969036526,
+      "grad_norm": 0.28181778666966495,
+      "learning_rate": 0.0001686208058971986,
+      "loss": 1.0444,
+      "step": 1463
+    },
+    {
+      "epoch": 0.2816654140989028,
+      "grad_norm": 0.28495260405584993,
+      "learning_rate": 0.00016857546007664905,
+      "loss": 1.1328,
+      "step": 1464
+    },
+    {
+      "epoch": 0.28185780850744024,
+      "grad_norm": 0.27798452060783974,
+      "learning_rate": 0.000168530087622181,
+      "loss": 1.213,
+      "step": 1465
+    },
+    {
+      "epoch": 0.28205020291597777,
+      "grad_norm": 0.49235965559192735,
+      "learning_rate": 0.0001684846885514166,
+      "loss": 1.175,
+      "step": 1466
+    },
+    {
+      "epoch": 0.28224259732451523,
+      "grad_norm": 0.3576291386200579,
+      "learning_rate": 0.00016843926288198828,
+      "loss": 1.0901,
+      "step": 1467
+    },
+    {
+      "epoch": 0.28243499173305275,
+      "grad_norm": 0.29159212153274483,
+      "learning_rate": 0.0001683938106315389,
+      "loss": 1.1877,
+      "step": 1468
+    },
+    {
+      "epoch": 0.2826273861415903,
+      "grad_norm": 0.2723514191399268,
+      "learning_rate": 0.0001683483318177216,
+      "loss": 1.0679,
+      "step": 1469
+    },
+    {
+      "epoch": 0.28281978055012774,
+      "grad_norm": 0.30860933842493043,
+      "learning_rate": 0.00016830282645819976,
+      "loss": 1.1087,
+      "step": 1470
+    },
+    {
+      "epoch": 0.28301217495866526,
+      "grad_norm": 0.33556153547966977,
+      "learning_rate": 0.0001682572945706472,
+      "loss": 1.1703,
+      "step": 1471
+    },
+    {
+      "epoch": 0.2832045693672028,
+      "grad_norm": 0.3015032914858447,
+      "learning_rate": 0.00016821173617274793,
+      "loss": 1.0502,
+      "step": 1472
+    },
+    {
+      "epoch": 0.28339696377574025,
+      "grad_norm": 0.2953371253542578,
+      "learning_rate": 0.00016816615128219635,
+      "loss": 1.1347,
+      "step": 1473
+    },
+    {
+      "epoch": 0.2835893581842778,
+      "grad_norm": 0.3176017935032959,
+      "learning_rate": 0.0001681205399166971,
+      "loss": 1.1705,
+      "step": 1474
+    },
+    {
+      "epoch": 0.2837817525928153,
+      "grad_norm": 0.29309108886207547,
+      "learning_rate": 0.00016807490209396506,
+      "loss": 1.1541,
+      "step": 1475
+    },
+    {
+      "epoch": 0.28397414700135276,
+      "grad_norm": 0.39354989586670625,
+      "learning_rate": 0.00016802923783172552,
+      "loss": 1.0775,
+      "step": 1476
+    },
+    {
+      "epoch": 0.2841665414098903,
+      "grad_norm": 0.29954011805284947,
+      "learning_rate": 0.0001679835471477139,
+      "loss": 1.0235,
+      "step": 1477
+    },
+    {
+      "epoch": 0.28435893581842775,
+      "grad_norm": 0.3035630327059971,
+      "learning_rate": 0.00016793783005967592,
+      "loss": 1.0487,
+      "step": 1478
+    },
+    {
+      "epoch": 0.28455133022696527,
+      "grad_norm": 0.31938507727732696,
+      "learning_rate": 0.0001678920865853676,
+      "loss": 1.0572,
+      "step": 1479
+    },
+    {
+      "epoch": 0.2847437246355028,
+      "grad_norm": 0.29441542933487236,
+      "learning_rate": 0.00016784631674255518,
+      "loss": 1.1836,
+      "step": 1480
+    },
+    {
+      "epoch": 0.28493611904404026,
+      "grad_norm": 0.3107622872292956,
+      "learning_rate": 0.00016780052054901512,
+      "loss": 1.1321,
+      "step": 1481
+    },
+    {
+      "epoch": 0.2851285134525778,
+      "grad_norm": 0.35573137406520317,
+      "learning_rate": 0.00016775469802253414,
+      "loss": 1.1253,
+      "step": 1482
+    },
+    {
+      "epoch": 0.2853209078611153,
+      "grad_norm": 0.2765597068152107,
+      "learning_rate": 0.0001677088491809092,
+      "loss": 1.0699,
+      "step": 1483
+    },
+    {
+      "epoch": 0.28551330226965277,
+      "grad_norm": 0.3639595288307644,
+      "learning_rate": 0.00016766297404194745,
+      "loss": 1.1564,
+      "step": 1484
+    },
+    {
+      "epoch": 0.2857056966781903,
+      "grad_norm": 0.2972230833788457,
+      "learning_rate": 0.00016761707262346625,
+      "loss": 1.0802,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2858980910867278,
+      "grad_norm": 0.2585359595866178,
+      "learning_rate": 0.0001675711449432932,
+      "loss": 1.066,
+      "step": 1486
+    },
+    {
+      "epoch": 0.2860904854952653,
+      "grad_norm": 0.3416958588595029,
+      "learning_rate": 0.00016752519101926607,
+      "loss": 1.1818,
+      "step": 1487
+    },
+    {
+      "epoch": 0.2862828799038028,
+      "grad_norm": 0.24818602492656874,
+      "learning_rate": 0.00016747921086923283,
+      "loss": 1.1695,
+      "step": 1488
+    },
+    {
+      "epoch": 0.2864752743123403,
+      "grad_norm": 0.26713942087382575,
+      "learning_rate": 0.00016743320451105168,
+      "loss": 1.0488,
+      "step": 1489
+    },
+    {
+      "epoch": 0.2866676687208778,
+      "grad_norm": 0.3527701470941162,
+      "learning_rate": 0.00016738717196259094,
+      "loss": 1.144,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2868600631294153,
+      "grad_norm": 0.3293371939936096,
+      "learning_rate": 0.0001673411132417291,
+      "loss": 1.0061,
+      "step": 1491
+    },
+    {
+      "epoch": 0.2870524575379528,
+      "grad_norm": 0.3556116428133735,
+      "learning_rate": 0.0001672950283663548,
+      "loss": 1.1294,
+      "step": 1492
+    },
+    {
+      "epoch": 0.2872448519464903,
+      "grad_norm": 0.2748832749209894,
+      "learning_rate": 0.00016724891735436697,
+      "loss": 1.1753,
+      "step": 1493
+    },
+    {
+      "epoch": 0.2874372463550278,
+      "grad_norm": 0.31851666387671945,
+      "learning_rate": 0.00016720278022367452,
+      "loss": 1.1382,
+      "step": 1494
+    },
+    {
+      "epoch": 0.2876296407635653,
+      "grad_norm": 0.5268915186796314,
+      "learning_rate": 0.00016715661699219663,
+      "loss": 1.0654,
+      "step": 1495
+    },
+    {
+      "epoch": 0.2878220351721028,
+      "grad_norm": 0.34015706564092485,
+      "learning_rate": 0.00016711042767786257,
+      "loss": 1.119,
+      "step": 1496
+    },
+    {
+      "epoch": 0.28801442958064033,
+      "grad_norm": 0.3380177753706035,
+      "learning_rate": 0.00016706421229861166,
+      "loss": 1.0756,
+      "step": 1497
+    },
+    {
+      "epoch": 0.2882068239891778,
+      "grad_norm": 0.251871619832251,
+      "learning_rate": 0.00016701797087239354,
+      "loss": 1.1766,
+      "step": 1498
+    },
+    {
+      "epoch": 0.2883992183977153,
+      "grad_norm": 0.2497903719085726,
+      "learning_rate": 0.0001669717034171677,
+      "loss": 1.0662,
+      "step": 1499
+    },
+    {
+      "epoch": 0.28859161280625284,
+      "grad_norm": 0.271168615319396,
+      "learning_rate": 0.000166925409950904,
+      "loss": 1.1289,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2887840072147903,
+      "grad_norm": 0.28216074211396697,
+      "learning_rate": 0.00016687909049158228,
+      "loss": 1.1391,
+      "step": 1501
+    },
+    {
+      "epoch": 0.2889764016233278,
+      "grad_norm": 0.34081625162184925,
+      "learning_rate": 0.00016683274505719246,
+      "loss": 1.1392,
+      "step": 1502
+    },
+    {
+      "epoch": 0.28916879603186535,
+      "grad_norm": 0.3632388174196172,
+      "learning_rate": 0.00016678637366573454,
+      "loss": 1.0236,
+      "step": 1503
+    },
+    {
+      "epoch": 0.2893611904404028,
+      "grad_norm": 0.3221984456151925,
+      "learning_rate": 0.0001667399763352187,
+      "loss": 1.105,
+      "step": 1504
+    },
+    {
+      "epoch": 0.28955358484894034,
+      "grad_norm": 0.31769912921531707,
+      "learning_rate": 0.0001666935530836651,
+      "loss": 1.186,
+      "step": 1505
+    },
+    {
+      "epoch": 0.2897459792574778,
+      "grad_norm": 0.34271585635347096,
+      "learning_rate": 0.00016664710392910395,
+      "loss": 0.9809,
+      "step": 1506
+    },
+    {
+      "epoch": 0.2899383736660153,
+      "grad_norm": 0.3357288852011628,
+      "learning_rate": 0.00016660062888957563,
+      "loss": 1.1471,
+      "step": 1507
+    },
+    {
+      "epoch": 0.29013076807455285,
+      "grad_norm": 0.31661361899122564,
+      "learning_rate": 0.00016655412798313051,
+      "loss": 1.0597,
+      "step": 1508
+    },
+    {
+      "epoch": 0.2903231624830903,
+      "grad_norm": 0.30237753675077755,
+      "learning_rate": 0.00016650760122782895,
+      "loss": 0.9679,
+      "step": 1509
+    },
+    {
+      "epoch": 0.29051555689162784,
+      "grad_norm": 0.31153421406024145,
+      "learning_rate": 0.00016646104864174145,
+      "loss": 1.1629,
+      "step": 1510
+    },
+    {
+      "epoch": 0.29070795130016536,
+      "grad_norm": 0.32212144065978987,
+      "learning_rate": 0.0001664144702429485,
+      "loss": 1.0928,
+      "step": 1511
+    },
+    {
+      "epoch": 0.2909003457087028,
+      "grad_norm": 0.47514370505666315,
+      "learning_rate": 0.0001663678660495406,
+      "loss": 1.0917,
+      "step": 1512
+    },
+    {
+      "epoch": 0.29109274011724035,
+      "grad_norm": 0.327981366932905,
+      "learning_rate": 0.0001663212360796183,
+      "loss": 1.1701,
+      "step": 1513
+    },
+    {
+      "epoch": 0.29128513452577787,
+      "grad_norm": 0.2578675540893025,
+      "learning_rate": 0.0001662745803512921,
+      "loss": 1.1149,
+      "step": 1514
+    },
+    {
+      "epoch": 0.29147752893431533,
+      "grad_norm": 0.33134955839770963,
+      "learning_rate": 0.00016622789888268258,
+      "loss": 1.1508,
+      "step": 1515
+    },
+    {
+      "epoch": 0.29166992334285285,
+      "grad_norm": 0.29587883752941313,
+      "learning_rate": 0.00016618119169192026,
+      "loss": 1.1707,
+      "step": 1516
+    },
+    {
+      "epoch": 0.2918623177513904,
+      "grad_norm": 0.3259080926822911,
+      "learning_rate": 0.00016613445879714572,
+      "loss": 1.1833,
+      "step": 1517
+    },
+    {
+      "epoch": 0.29205471215992784,
+      "grad_norm": 0.2759774004048962,
+      "learning_rate": 0.00016608770021650943,
+      "loss": 1.0774,
+      "step": 1518
+    },
+    {
+      "epoch": 0.29224710656846536,
+      "grad_norm": 0.34534654655783553,
+      "learning_rate": 0.00016604091596817192,
+      "loss": 1.0763,
+      "step": 1519
+    },
+    {
+      "epoch": 0.29243950097700283,
+      "grad_norm": 0.32048012748563287,
+      "learning_rate": 0.00016599410607030365,
+      "loss": 1.1655,
+      "step": 1520
+    },
+    {
+      "epoch": 0.29263189538554035,
+      "grad_norm": 0.27284069611106254,
+      "learning_rate": 0.00016594727054108498,
+      "loss": 1.1008,
+      "step": 1521
+    },
+    {
+      "epoch": 0.2928242897940779,
+      "grad_norm": 0.2842901508160404,
+      "learning_rate": 0.0001659004093987064,
+      "loss": 1.0221,
+      "step": 1522
+    },
+    {
+      "epoch": 0.29301668420261534,
+      "grad_norm": 0.3000877091087916,
+      "learning_rate": 0.00016585352266136814,
+      "loss": 1.1417,
+      "step": 1523
+    },
+    {
+      "epoch": 0.29320907861115286,
+      "grad_norm": 0.34940099331165325,
+      "learning_rate": 0.00016580661034728053,
+      "loss": 1.0477,
+      "step": 1524
+    },
+    {
+      "epoch": 0.2934014730196904,
+      "grad_norm": 0.28169491807370783,
+      "learning_rate": 0.00016575967247466376,
+      "loss": 1.0167,
+      "step": 1525
+    },
+    {
+      "epoch": 0.29359386742822785,
+      "grad_norm": 0.28410687744689694,
+      "learning_rate": 0.0001657127090617479,
+      "loss": 1.0746,
+      "step": 1526
+    },
+    {
+      "epoch": 0.29378626183676537,
+      "grad_norm": 0.33459806989878466,
+      "learning_rate": 0.0001656657201267731,
+      "loss": 1.083,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2939786562453029,
+      "grad_norm": 0.3048899533795747,
+      "learning_rate": 0.00016561870568798924,
+      "loss": 0.9888,
+      "step": 1528
+    },
+    {
+      "epoch": 0.29417105065384036,
+      "grad_norm": 0.3321949905668031,
+      "learning_rate": 0.00016557166576365622,
+      "loss": 1.163,
+      "step": 1529
+    },
+    {
+      "epoch": 0.2943634450623779,
+      "grad_norm": 0.26981622413587947,
+      "learning_rate": 0.00016552460037204384,
+      "loss": 1.0693,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2945558394709154,
+      "grad_norm": 0.31459955220857305,
+      "learning_rate": 0.00016547750953143167,
+      "loss": 1.0609,
+      "step": 1531
+    },
+    {
+      "epoch": 0.29474823387945287,
+      "grad_norm": 0.3003298648748653,
+      "learning_rate": 0.00016543039326010928,
+      "loss": 1.1169,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2949406282879904,
+      "grad_norm": 0.2875135056200567,
+      "learning_rate": 0.00016538325157637614,
+      "loss": 1.0714,
+      "step": 1533
+    },
+    {
+      "epoch": 0.29513302269652786,
+      "grad_norm": 0.3318638867723934,
+      "learning_rate": 0.00016533608449854147,
+      "loss": 1.1392,
+      "step": 1534
+    },
+    {
+      "epoch": 0.2953254171050654,
+      "grad_norm": 0.3259864783176585,
+      "learning_rate": 0.00016528889204492448,
+      "loss": 1.0244,
+      "step": 1535
+    },
+    {
+      "epoch": 0.2955178115136029,
+      "grad_norm": 0.3374964952180507,
+      "learning_rate": 0.00016524167423385413,
+      "loss": 1.1173,
+      "step": 1536
+    },
+    {
+      "epoch": 0.29571020592214037,
+      "grad_norm": 0.28719876444284703,
+      "learning_rate": 0.00016519443108366927,
+      "loss": 1.234,
+      "step": 1537
+    },
+    {
+      "epoch": 0.2959026003306779,
+      "grad_norm": 0.2739730710761506,
+      "learning_rate": 0.00016514716261271866,
+      "loss": 1.0707,
+      "step": 1538
+    },
+    {
+      "epoch": 0.2960949947392154,
+      "grad_norm": 0.30130410795227114,
+      "learning_rate": 0.00016509986883936074,
+      "loss": 1.101,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2962873891477529,
+      "grad_norm": 0.35028803854119056,
+      "learning_rate": 0.00016505254978196388,
+      "loss": 1.059,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2964797835562904,
+      "grad_norm": 0.32535297219762993,
+      "learning_rate": 0.00016500520545890634,
+      "loss": 1.1654,
+      "step": 1541
+    },
+    {
+      "epoch": 0.2966721779648279,
+      "grad_norm": 0.30817740055072923,
+      "learning_rate": 0.00016495783588857605,
+      "loss": 1.1192,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2968645723733654,
+      "grad_norm": 0.478210489089417,
+      "learning_rate": 0.0001649104410893708,
+      "loss": 1.0413,
+      "step": 1543
+    },
+    {
+      "epoch": 0.2970569667819029,
+      "grad_norm": 0.3068352514438494,
+      "learning_rate": 0.0001648630210796982,
+      "loss": 1.1003,
+      "step": 1544
+    },
+    {
+      "epoch": 0.2972493611904404,
+      "grad_norm": 0.3406848942790489,
+      "learning_rate": 0.0001648155758779756,
+      "loss": 1.0016,
+      "step": 1545
+    },
+    {
+      "epoch": 0.2974417555989779,
+      "grad_norm": 0.2808769192019626,
+      "learning_rate": 0.0001647681055026302,
+      "loss": 1.0166,
+      "step": 1546
+    },
+    {
+      "epoch": 0.2976341500075154,
+      "grad_norm": 0.329848316331849,
+      "learning_rate": 0.000164720609972099,
+      "loss": 1.0867,
+      "step": 1547
+    },
+    {
+      "epoch": 0.2978265444160529,
+      "grad_norm": 0.2809503536357732,
+      "learning_rate": 0.00016467308930482864,
+      "loss": 1.1006,
+      "step": 1548
+    },
+    {
+      "epoch": 0.2980189388245904,
+      "grad_norm": 0.3328033422354526,
+      "learning_rate": 0.00016462554351927557,
+      "loss": 0.9857,
+      "step": 1549
+    },
+    {
+      "epoch": 0.29821133323312793,
+      "grad_norm": 0.35623490764987126,
+      "learning_rate": 0.00016457797263390612,
+      "loss": 1.2333,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2984037276416654,
+      "grad_norm": 0.31260021843495356,
+      "learning_rate": 0.00016453037666719624,
+      "loss": 1.131,
+      "step": 1551
+    },
+    {
+      "epoch": 0.2985961220502029,
+      "grad_norm": 0.30209859915913373,
+      "learning_rate": 0.00016448275563763162,
+      "loss": 1.1302,
+      "step": 1552
+    },
+    {
+      "epoch": 0.29878851645874044,
+      "grad_norm": 0.34068880877156577,
+      "learning_rate": 0.0001644351095637078,
+      "loss": 1.0522,
+      "step": 1553
+    },
+    {
+      "epoch": 0.2989809108672779,
+      "grad_norm": 0.2717034027673641,
+      "learning_rate": 0.00016438743846392985,
+      "loss": 1.1712,
+      "step": 1554
+    },
+    {
+      "epoch": 0.2991733052758154,
+      "grad_norm": 0.3103989627495267,
+      "learning_rate": 0.00016433974235681274,
+      "loss": 1.1326,
+      "step": 1555
+    },
+    {
+      "epoch": 0.29936569968435295,
+      "grad_norm": 0.2402098857090824,
+      "learning_rate": 0.0001642920212608811,
+      "loss": 1.1839,
+      "step": 1556
+    },
+    {
+      "epoch": 0.2995580940928904,
+      "grad_norm": 0.34889125013744243,
+      "learning_rate": 0.00016424427519466924,
+      "loss": 1.1347,
+      "step": 1557
+    },
+    {
+      "epoch": 0.29975048850142794,
+      "grad_norm": 0.363983432710772,
+      "learning_rate": 0.00016419650417672118,
+      "loss": 1.0968,
+      "step": 1558
+    },
+    {
+      "epoch": 0.2999428829099654,
+      "grad_norm": 0.31039175436138483,
+      "learning_rate": 0.00016414870822559064,
+      "loss": 1.2207,
+      "step": 1559
+    },
+    {
+      "epoch": 0.3001352773185029,
+      "grad_norm": 0.3055765865425427,
+      "learning_rate": 0.00016410088735984102,
+      "loss": 1.0434,
+      "step": 1560
+    },
+    {
+      "epoch": 0.30032767172704045,
+      "grad_norm": 0.2667455286187899,
+      "learning_rate": 0.00016405304159804534,
+      "loss": 1.0608,
+      "step": 1561
+    },
+    {
+      "epoch": 0.3005200661355779,
+      "grad_norm": 0.29871428015617846,
+      "learning_rate": 0.00016400517095878643,
+      "loss": 1.0646,
+      "step": 1562
+    },
+    {
+      "epoch": 0.30071246054411543,
+      "grad_norm": 0.42130675647760185,
+      "learning_rate": 0.00016395727546065665,
+      "loss": 0.9163,
+      "step": 1563
+    },
+    {
+      "epoch": 0.30090485495265296,
+      "grad_norm": 0.3271331110157056,
+      "learning_rate": 0.00016390935512225805,
+      "loss": 1.1656,
+      "step": 1564
+    },
+    {
+      "epoch": 0.3010972493611904,
+      "grad_norm": 0.3434655383691047,
+      "learning_rate": 0.00016386140996220232,
+      "loss": 1.0943,
+      "step": 1565
+    },
+    {
+      "epoch": 0.30128964376972794,
+      "grad_norm": 0.3017120632775669,
+      "learning_rate": 0.00016381343999911086,
+      "loss": 1.2189,
+      "step": 1566
+    },
+    {
+      "epoch": 0.30148203817826547,
+      "grad_norm": 0.2531338468220288,
+      "learning_rate": 0.00016376544525161465,
+      "loss": 1.1832,
+      "step": 1567
+    },
+    {
+      "epoch": 0.30167443258680293,
+      "grad_norm": 0.30054357809784965,
+      "learning_rate": 0.00016371742573835426,
+      "loss": 1.0188,
+      "step": 1568
+    },
+    {
+      "epoch": 0.30186682699534045,
+      "grad_norm": 0.2859045038413787,
+      "learning_rate": 0.0001636693814779799,
+      "loss": 1.0915,
+      "step": 1569
+    },
+    {
+      "epoch": 0.302059221403878,
+      "grad_norm": 0.3255328867349781,
+      "learning_rate": 0.00016362131248915144,
+      "loss": 1.0922,
+      "step": 1570
+    },
+    {
+      "epoch": 0.30225161581241544,
+      "grad_norm": 0.31721420954452095,
+      "learning_rate": 0.00016357321879053834,
+      "loss": 1.0463,
+      "step": 1571
+    },
+    {
+      "epoch": 0.30244401022095296,
+      "grad_norm": 0.308636085373308,
+      "learning_rate": 0.00016352510040081963,
+      "loss": 1.1799,
+      "step": 1572
+    },
+    {
+      "epoch": 0.30263640462949043,
+      "grad_norm": 0.3396430145652912,
+      "learning_rate": 0.00016347695733868388,
+      "loss": 1.087,
+      "step": 1573
+    },
+    {
+      "epoch": 0.30282879903802795,
+      "grad_norm": 0.3150167575930741,
+      "learning_rate": 0.00016342878962282938,
+      "loss": 1.0466,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3030211934465655,
+      "grad_norm": 0.23547476718079235,
+      "learning_rate": 0.00016338059727196387,
+      "loss": 1.1615,
+      "step": 1575
+    },
+    {
+      "epoch": 0.30321358785510294,
+      "grad_norm": 0.2981953981187636,
+      "learning_rate": 0.0001633323803048047,
+      "loss": 1.0787,
+      "step": 1576
+    },
+    {
+      "epoch": 0.30340598226364046,
+      "grad_norm": 0.33129995102859633,
+      "learning_rate": 0.0001632841387400788,
+      "loss": 1.1471,
+      "step": 1577
+    },
+    {
+      "epoch": 0.303598376672178,
+      "grad_norm": 0.30979639418539157,
+      "learning_rate": 0.00016323587259652267,
+      "loss": 1.1005,
+      "step": 1578
+    },
+    {
+      "epoch": 0.30379077108071545,
+      "grad_norm": 0.29595478584885476,
+      "learning_rate": 0.00016318758189288226,
+      "loss": 1.0032,
+      "step": 1579
+    },
+    {
+      "epoch": 0.30398316548925297,
+      "grad_norm": 0.36623837203171844,
+      "learning_rate": 0.00016313926664791314,
+      "loss": 1.1667,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3041755598977905,
+      "grad_norm": 0.2983930115362985,
+      "learning_rate": 0.00016309092688038046,
+      "loss": 1.0953,
+      "step": 1581
+    },
+    {
+      "epoch": 0.30436795430632796,
+      "grad_norm": 0.3146073937564928,
+      "learning_rate": 0.00016304256260905872,
+      "loss": 0.9952,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3045603487148655,
+      "grad_norm": 0.29550557175646847,
+      "learning_rate": 0.00016299417385273214,
+      "loss": 1.1821,
+      "step": 1583
+    },
+    {
+      "epoch": 0.304752743123403,
+      "grad_norm": 0.3038707634019499,
+      "learning_rate": 0.00016294576063019428,
+      "loss": 1.1121,
+      "step": 1584
+    },
+    {
+      "epoch": 0.30494513753194047,
+      "grad_norm": 0.3293523071770354,
+      "learning_rate": 0.00016289732296024834,
+      "loss": 1.106,
+      "step": 1585
+    },
+    {
+      "epoch": 0.305137531940478,
+      "grad_norm": 0.29149109941744755,
+      "learning_rate": 0.00016284886086170698,
+      "loss": 1.1194,
+      "step": 1586
+    },
+    {
+      "epoch": 0.30532992634901546,
+      "grad_norm": 0.4592677812140502,
+      "learning_rate": 0.00016280037435339217,
+      "loss": 1.1474,
+      "step": 1587
+    },
+    {
+      "epoch": 0.305522320757553,
+      "grad_norm": 0.3203653400906737,
+      "learning_rate": 0.00016275186345413568,
+      "loss": 1.1605,
+      "step": 1588
+    },
+    {
+      "epoch": 0.3057147151660905,
+      "grad_norm": 0.36307461152220283,
+      "learning_rate": 0.00016270332818277847,
+      "loss": 1.0303,
+      "step": 1589
+    },
+    {
+      "epoch": 0.30590710957462797,
+      "grad_norm": 0.44069341364426956,
+      "learning_rate": 0.00016265476855817116,
+      "loss": 1.0347,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3060995039831655,
+      "grad_norm": 0.29995668120186636,
+      "learning_rate": 0.00016260618459917368,
+      "loss": 1.0528,
+      "step": 1591
+    },
+    {
+      "epoch": 0.306291898391703,
+      "grad_norm": 0.3451138109635149,
+      "learning_rate": 0.00016255757632465553,
+      "loss": 1.1118,
+      "step": 1592
+    },
+    {
+      "epoch": 0.3064842928002405,
+      "grad_norm": 0.2711304810937573,
+      "learning_rate": 0.00016250894375349558,
+      "loss": 1.101,
+      "step": 1593
+    },
+    {
+      "epoch": 0.306676687208778,
+      "grad_norm": 0.38704399296599323,
+      "learning_rate": 0.00016246028690458216,
+      "loss": 1.1577,
+      "step": 1594
+    },
+    {
+      "epoch": 0.3068690816173155,
+      "grad_norm": 0.4133066487034265,
+      "learning_rate": 0.00016241160579681308,
+      "loss": 1.1894,
+      "step": 1595
+    },
+    {
+      "epoch": 0.307061476025853,
+      "grad_norm": 0.28992392284211643,
+      "learning_rate": 0.0001623629004490954,
+      "loss": 1.1236,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3072538704343905,
+      "grad_norm": 0.33565142715411495,
+      "learning_rate": 0.00016231417088034586,
+      "loss": 1.1304,
+      "step": 1597
+    },
+    {
+      "epoch": 0.30744626484292803,
+      "grad_norm": 0.26501111413113954,
+      "learning_rate": 0.0001622654171094904,
+      "loss": 1.0423,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3076386592514655,
+      "grad_norm": 0.36617546738271656,
+      "learning_rate": 0.00016221663915546436,
+      "loss": 1.0918,
+      "step": 1599
+    },
+    {
+      "epoch": 0.307831053660003,
+      "grad_norm": 0.389302797482364,
+      "learning_rate": 0.00016216783703721266,
+      "loss": 1.0707,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3080234480685405,
+      "grad_norm": 0.2773217172383021,
+      "learning_rate": 0.00016211901077368935,
+      "loss": 1.0718,
+      "step": 1601
+    },
+    {
+      "epoch": 0.308215842477078,
+      "grad_norm": 0.28523774796025136,
+      "learning_rate": 0.0001620701603838581,
+      "loss": 1.1037,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3084082368856155,
+      "grad_norm": 0.22574790952974697,
+      "learning_rate": 0.00016202128588669177,
+      "loss": 1.1081,
+      "step": 1603
+    },
+    {
+      "epoch": 0.308600631294153,
+      "grad_norm": 0.2764104018829215,
+      "learning_rate": 0.00016197238730117271,
+      "loss": 1.1793,
+      "step": 1604
+    },
+    {
+      "epoch": 0.3087930257026905,
+      "grad_norm": 0.34309387709313244,
+      "learning_rate": 0.00016192346464629246,
+      "loss": 1.1452,
+      "step": 1605
+    },
+    {
+      "epoch": 0.30898542011122804,
+      "grad_norm": 0.31290656503701925,
+      "learning_rate": 0.00016187451794105214,
+      "loss": 1.1104,
+      "step": 1606
+    },
+    {
+      "epoch": 0.3091778145197655,
+      "grad_norm": 0.288157637887267,
+      "learning_rate": 0.00016182554720446202,
+      "loss": 1.0873,
+      "step": 1607
+    },
+    {
+      "epoch": 0.309370208928303,
+      "grad_norm": 0.2934921643373651,
+      "learning_rate": 0.00016177655245554177,
+      "loss": 1.1279,
+      "step": 1608
+    },
+    {
+      "epoch": 0.30956260333684055,
+      "grad_norm": 0.2972520581216308,
+      "learning_rate": 0.00016172753371332038,
+      "loss": 1.1243,
+      "step": 1609
+    },
+    {
+      "epoch": 0.309754997745378,
+      "grad_norm": 0.29628968296112,
+      "learning_rate": 0.00016167849099683624,
+      "loss": 1.0601,
+      "step": 1610
+    },
+    {
+      "epoch": 0.30994739215391554,
+      "grad_norm": 0.3353959738162529,
+      "learning_rate": 0.00016162942432513687,
+      "loss": 1.1675,
+      "step": 1611
+    },
+    {
+      "epoch": 0.310139786562453,
+      "grad_norm": 0.2632999349649381,
+      "learning_rate": 0.00016158033371727924,
+      "loss": 1.1154,
+      "step": 1612
+    },
+    {
+      "epoch": 0.3103321809709905,
+      "grad_norm": 0.2552675225240679,
+      "learning_rate": 0.00016153121919232962,
+      "loss": 1.0847,
+      "step": 1613
+    },
+    {
+      "epoch": 0.31052457537952805,
+      "grad_norm": 0.35422271810160016,
+      "learning_rate": 0.00016148208076936348,
+      "loss": 1.079,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3107169697880655,
+      "grad_norm": 0.27509491454965834,
+      "learning_rate": 0.0001614329184674656,
+      "loss": 1.1359,
+      "step": 1615
+    },
+    {
+      "epoch": 0.31090936419660303,
+      "grad_norm": 0.30725065984078526,
+      "learning_rate": 0.00016138373230573013,
+      "loss": 1.0937,
+      "step": 1616
+    },
+    {
+      "epoch": 0.31110175860514055,
+      "grad_norm": 0.26424568384200414,
+      "learning_rate": 0.00016133452230326033,
+      "loss": 1.0895,
+      "step": 1617
+    },
+    {
+      "epoch": 0.311294153013678,
+      "grad_norm": 0.33726879817609995,
+      "learning_rate": 0.00016128528847916883,
+      "loss": 1.0272,
+      "step": 1618
+    },
+    {
+      "epoch": 0.31148654742221554,
+      "grad_norm": 0.2984932843827208,
+      "learning_rate": 0.00016123603085257745,
+      "loss": 1.1446,
+      "step": 1619
+    },
+    {
+      "epoch": 0.31167894183075306,
+      "grad_norm": 0.2503779323015421,
+      "learning_rate": 0.00016118674944261732,
+      "loss": 1.1713,
+      "step": 1620
+    },
+    {
+      "epoch": 0.31187133623929053,
+      "grad_norm": 0.2882919565900987,
+      "learning_rate": 0.0001611374442684288,
+      "loss": 1.1407,
+      "step": 1621
+    },
+    {
+      "epoch": 0.31206373064782805,
+      "grad_norm": 0.37607656355574237,
+      "learning_rate": 0.00016108811534916136,
+      "loss": 1.0204,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3122561250563656,
+      "grad_norm": 0.30051842653071015,
+      "learning_rate": 0.00016103876270397386,
+      "loss": 1.1044,
+      "step": 1623
+    },
+    {
+      "epoch": 0.31244851946490304,
+      "grad_norm": 0.3351953421145429,
+      "learning_rate": 0.0001609893863520343,
+      "loss": 1.1175,
+      "step": 1624
+    },
+    {
+      "epoch": 0.31264091387344056,
+      "grad_norm": 0.3709485804872838,
+      "learning_rate": 0.0001609399863125198,
+      "loss": 1.1519,
+      "step": 1625
+    },
+    {
+      "epoch": 0.31283330828197803,
+      "grad_norm": 0.27105795360778984,
+      "learning_rate": 0.00016089056260461688,
+      "loss": 1.1809,
+      "step": 1626
+    },
+    {
+      "epoch": 0.31302570269051555,
+      "grad_norm": 0.3249737759395512,
+      "learning_rate": 0.00016084111524752105,
+      "loss": 1.1303,
+      "step": 1627
+    },
+    {
+      "epoch": 0.31321809709905307,
+      "grad_norm": 0.30231451672539145,
+      "learning_rate": 0.00016079164426043718,
+      "loss": 1.1199,
+      "step": 1628
+    },
+    {
+      "epoch": 0.31341049150759054,
+      "grad_norm": 0.3404165721034413,
+      "learning_rate": 0.0001607421496625791,
+      "loss": 1.0667,
+      "step": 1629
+    },
+    {
+      "epoch": 0.31360288591612806,
+      "grad_norm": 0.3473400020118858,
+      "learning_rate": 0.00016069263147317013,
+      "loss": 1.2436,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3137952803246656,
+      "grad_norm": 0.2635881180537359,
+      "learning_rate": 0.00016064308971144238,
+      "loss": 1.0311,
+      "step": 1631
+    },
+    {
+      "epoch": 0.31398767473320305,
+      "grad_norm": 0.3097943727334572,
+      "learning_rate": 0.00016059352439663739,
+      "loss": 1.1602,
+      "step": 1632
+    },
+    {
+      "epoch": 0.31418006914174057,
+      "grad_norm": 0.24570331745693133,
+      "learning_rate": 0.00016054393554800576,
+      "loss": 1.1309,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3143724635502781,
+      "grad_norm": 0.38554881372537647,
+      "learning_rate": 0.0001604943231848072,
+      "loss": 1.1888,
+      "step": 1634
+    },
+    {
+      "epoch": 0.31456485795881556,
+      "grad_norm": 0.2820389162483143,
+      "learning_rate": 0.00016044468732631057,
+      "loss": 1.1445,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3147572523673531,
+      "grad_norm": 0.27371254530078315,
+      "learning_rate": 0.00016039502799179394,
+      "loss": 0.9879,
+      "step": 1636
+    },
+    {
+      "epoch": 0.3149496467758906,
+      "grad_norm": 0.30841342432884855,
+      "learning_rate": 0.00016034534520054433,
+      "loss": 1.1096,
+      "step": 1637
+    },
+    {
+      "epoch": 0.31514204118442807,
+      "grad_norm": 0.31856710401081456,
+      "learning_rate": 0.000160295638971858,
+      "loss": 1.0031,
+      "step": 1638
+    },
+    {
+      "epoch": 0.3153344355929656,
+      "grad_norm": 0.3274701548655172,
+      "learning_rate": 0.0001602459093250403,
+      "loss": 1.0453,
+      "step": 1639
+    },
+    {
+      "epoch": 0.31552683000150306,
+      "grad_norm": 0.33298410955415725,
+      "learning_rate": 0.0001601961562794056,
+      "loss": 1.1194,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3157192244100406,
+      "grad_norm": 0.33392940564080026,
+      "learning_rate": 0.0001601463798542775,
+      "loss": 1.0302,
+      "step": 1641
+    },
+    {
+      "epoch": 0.3159116188185781,
+      "grad_norm": 0.3143650994119334,
+      "learning_rate": 0.00016009658006898848,
+      "loss": 1.1557,
+      "step": 1642
+    },
+    {
+      "epoch": 0.31610401322711557,
+      "grad_norm": 0.3514850475319936,
+      "learning_rate": 0.00016004675694288026,
+      "loss": 0.9993,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3162964076356531,
+      "grad_norm": 0.30575928311211603,
+      "learning_rate": 0.0001599969104953036,
+      "loss": 1.1048,
+      "step": 1644
+    },
+    {
+      "epoch": 0.3164888020441906,
+      "grad_norm": 0.28060796344719513,
+      "learning_rate": 0.0001599470407456182,
+      "loss": 1.0582,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3166811964527281,
+      "grad_norm": 0.26270948000019373,
+      "learning_rate": 0.00015989714771319299,
+      "loss": 1.1035,
+      "step": 1646
+    },
+    {
+      "epoch": 0.3168735908612656,
+      "grad_norm": 0.3512780011667153,
+      "learning_rate": 0.00015984723141740576,
+      "loss": 1.0329,
+      "step": 1647
+    },
+    {
+      "epoch": 0.3170659852698031,
+      "grad_norm": 0.34227776507782676,
+      "learning_rate": 0.0001597972918776435,
+      "loss": 1.1087,
+      "step": 1648
+    },
+    {
+      "epoch": 0.3172583796783406,
+      "grad_norm": 0.2899848389035279,
+      "learning_rate": 0.00015974732911330208,
+      "loss": 1.0792,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3174507740868781,
+      "grad_norm": 0.2894716805556298,
+      "learning_rate": 0.00015969734314378654,
+      "loss": 1.1692,
+      "step": 1650
+    },
+    {
+      "epoch": 0.31764316849541563,
+      "grad_norm": 0.33311694375190615,
+      "learning_rate": 0.00015964733398851077,
+      "loss": 1.0278,
+      "step": 1651
+    },
+    {
+      "epoch": 0.3178355629039531,
+      "grad_norm": 0.2288311725280411,
+      "learning_rate": 0.00015959730166689783,
+      "loss": 1.1078,
+      "step": 1652
+    },
+    {
+      "epoch": 0.3180279573124906,
+      "grad_norm": 0.3153840508885981,
+      "learning_rate": 0.00015954724619837967,
+      "loss": 1.1157,
+      "step": 1653
+    },
+    {
+      "epoch": 0.3182203517210281,
+      "grad_norm": 0.3111843945999688,
+      "learning_rate": 0.00015949716760239722,
+      "loss": 1.1882,
+      "step": 1654
+    },
+    {
+      "epoch": 0.3184127461295656,
+      "grad_norm": 0.32217455834873554,
+      "learning_rate": 0.00015944706589840046,
+      "loss": 1.1693,
+      "step": 1655
+    },
+    {
+      "epoch": 0.3186051405381031,
+      "grad_norm": 0.34036596071969843,
+      "learning_rate": 0.00015939694110584832,
+      "loss": 1.1236,
+      "step": 1656
+    },
+    {
+      "epoch": 0.3187975349466406,
+      "grad_norm": 0.3013483252959257,
+      "learning_rate": 0.00015934679324420872,
+      "loss": 1.1389,
+      "step": 1657
+    },
+    {
+      "epoch": 0.3189899293551781,
+      "grad_norm": 0.2923871347311417,
+      "learning_rate": 0.00015929662233295843,
+      "loss": 1.0541,
+      "step": 1658
+    },
+    {
+      "epoch": 0.31918232376371564,
+      "grad_norm": 0.31867701210043714,
+      "learning_rate": 0.00015924642839158332,
+      "loss": 1.0832,
+      "step": 1659
+    },
+    {
+      "epoch": 0.3193747181722531,
+      "grad_norm": 0.39856036220128727,
+      "learning_rate": 0.0001591962114395781,
+      "loss": 1.0644,
+      "step": 1660
+    },
+    {
+      "epoch": 0.3195671125807906,
+      "grad_norm": 0.34993384178261716,
+      "learning_rate": 0.00015914597149644652,
+      "loss": 1.1121,
+      "step": 1661
+    },
+    {
+      "epoch": 0.31975950698932815,
+      "grad_norm": 0.2818685497718032,
+      "learning_rate": 0.00015909570858170112,
+      "loss": 1.0815,
+      "step": 1662
+    },
+    {
+      "epoch": 0.3199519013978656,
+      "grad_norm": 0.3257483342553277,
+      "learning_rate": 0.00015904542271486346,
+      "loss": 1.1076,
+      "step": 1663
+    },
+    {
+      "epoch": 0.32014429580640313,
+      "grad_norm": 0.4620397042101499,
+      "learning_rate": 0.00015899511391546402,
+      "loss": 1.0697,
+      "step": 1664
+    },
+    {
+      "epoch": 0.32033669021494066,
+      "grad_norm": 0.28726484021356385,
+      "learning_rate": 0.00015894478220304214,
+      "loss": 1.0706,
+      "step": 1665
+    },
+    {
+      "epoch": 0.3205290846234781,
+      "grad_norm": 0.38799554637874106,
+      "learning_rate": 0.00015889442759714603,
+      "loss": 1.1153,
+      "step": 1666
+    },
+    {
+      "epoch": 0.32072147903201564,
+      "grad_norm": 0.28135786722388934,
+      "learning_rate": 0.00015884405011733292,
+      "loss": 1.078,
+      "step": 1667
+    },
+    {
+      "epoch": 0.3209138734405531,
+      "grad_norm": 0.31900256597094473,
+      "learning_rate": 0.0001587936497831688,
+      "loss": 1.1211,
+      "step": 1668
+    },
+    {
+      "epoch": 0.32110626784909063,
+      "grad_norm": 0.24625409818982552,
+      "learning_rate": 0.00015874322661422856,
+      "loss": 1.1542,
+      "step": 1669
+    },
+    {
+      "epoch": 0.32129866225762815,
+      "grad_norm": 0.3620634302619436,
+      "learning_rate": 0.000158692780630096,
+      "loss": 1.0907,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3214910566661656,
+      "grad_norm": 0.27307794748105374,
+      "learning_rate": 0.0001586423118503638,
+      "loss": 1.0856,
+      "step": 1671
+    },
+    {
+      "epoch": 0.32168345107470314,
+      "grad_norm": 0.3385126516913844,
+      "learning_rate": 0.0001585918202946334,
+      "loss": 1.1228,
+      "step": 1672
+    },
+    {
+      "epoch": 0.32187584548324066,
+      "grad_norm": 0.24589945573280625,
+      "learning_rate": 0.00015854130598251512,
+      "loss": 1.0938,
+      "step": 1673
+    },
+    {
+      "epoch": 0.32206823989177813,
+      "grad_norm": 0.29583788903140473,
+      "learning_rate": 0.0001584907689336282,
+      "loss": 1.1006,
+      "step": 1674
+    },
+    {
+      "epoch": 0.32226063430031565,
+      "grad_norm": 0.2552304059083672,
+      "learning_rate": 0.0001584402091676006,
+      "loss": 1.1358,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3224530287088532,
+      "grad_norm": 0.29829569302398823,
+      "learning_rate": 0.00015838962670406916,
+      "loss": 1.1205,
+      "step": 1676
+    },
+    {
+      "epoch": 0.32264542311739064,
+      "grad_norm": 0.31524272566734557,
+      "learning_rate": 0.00015833902156267956,
+      "loss": 1.0217,
+      "step": 1677
+    },
+    {
+      "epoch": 0.32283781752592816,
+      "grad_norm": 0.30083031982222824,
+      "learning_rate": 0.00015828839376308618,
+      "loss": 1.1027,
+      "step": 1678
+    },
+    {
+      "epoch": 0.3230302119344656,
+      "grad_norm": 0.35217746987846116,
+      "learning_rate": 0.00015823774332495235,
+      "loss": 1.0706,
+      "step": 1679
+    },
+    {
+      "epoch": 0.32322260634300315,
+      "grad_norm": 0.26813025465998713,
+      "learning_rate": 0.0001581870702679501,
+      "loss": 1.1434,
+      "step": 1680
+    },
+    {
+      "epoch": 0.32341500075154067,
+      "grad_norm": 0.2921191933598302,
+      "learning_rate": 0.0001581363746117602,
+      "loss": 1.0904,
+      "step": 1681
+    },
+    {
+      "epoch": 0.32360739516007814,
+      "grad_norm": 0.2670587227394991,
+      "learning_rate": 0.00015808565637607237,
+      "loss": 1.0912,
+      "step": 1682
+    },
+    {
+      "epoch": 0.32379978956861566,
+      "grad_norm": 0.28722073590857866,
+      "learning_rate": 0.00015803491558058488,
+      "loss": 1.0297,
+      "step": 1683
+    },
+    {
+      "epoch": 0.3239921839771532,
+      "grad_norm": 0.27207316853786107,
+      "learning_rate": 0.00015798415224500492,
+      "loss": 1.071,
+      "step": 1684
+    },
+    {
+      "epoch": 0.32418457838569065,
+      "grad_norm": 0.360211668310971,
+      "learning_rate": 0.00015793336638904838,
+      "loss": 1.1801,
+      "step": 1685
+    },
+    {
+      "epoch": 0.32437697279422817,
+      "grad_norm": 0.28057419264824257,
+      "learning_rate": 0.0001578825580324399,
+      "loss": 1.0872,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3245693672027657,
+      "grad_norm": 0.3115687098893564,
+      "learning_rate": 0.0001578317271949129,
+      "loss": 1.0582,
+      "step": 1687
+    },
+    {
+      "epoch": 0.32476176161130316,
+      "grad_norm": 0.3239013387661759,
+      "learning_rate": 0.00015778087389620938,
+      "loss": 1.1821,
+      "step": 1688
+    },
+    {
+      "epoch": 0.3249541560198407,
+      "grad_norm": 0.30370621431040346,
+      "learning_rate": 0.00015772999815608028,
+      "loss": 1.1513,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3251465504283782,
+      "grad_norm": 0.3388270445755482,
+      "learning_rate": 0.00015767909999428513,
+      "loss": 1.048,
+      "step": 1690
+    },
+    {
+      "epoch": 0.32533894483691567,
+      "grad_norm": 0.5014467130105696,
+      "learning_rate": 0.00015762817943059217,
+      "loss": 1.109,
+      "step": 1691
+    },
+    {
+      "epoch": 0.3255313392454532,
+      "grad_norm": 0.3101457911798254,
+      "learning_rate": 0.00015757723648477837,
+      "loss": 1.1951,
+      "step": 1692
+    },
+    {
+      "epoch": 0.32572373365399065,
+      "grad_norm": 0.31983341580099617,
+      "learning_rate": 0.0001575262711766294,
+      "loss": 1.0513,
+      "step": 1693
+    },
+    {
+      "epoch": 0.3259161280625282,
+      "grad_norm": 0.32115860558200393,
+      "learning_rate": 0.00015747528352593956,
+      "loss": 1.1377,
+      "step": 1694
+    },
+    {
+      "epoch": 0.3261085224710657,
+      "grad_norm": 0.26693164410380726,
+      "learning_rate": 0.0001574242735525119,
+      "loss": 1.0741,
+      "step": 1695
+    },
+    {
+      "epoch": 0.32630091687960316,
+      "grad_norm": 0.29261104426639334,
+      "learning_rate": 0.00015737324127615806,
+      "loss": 1.1571,
+      "step": 1696
+    },
+    {
+      "epoch": 0.3264933112881407,
+      "grad_norm": 0.32067235043976067,
+      "learning_rate": 0.00015732218671669844,
+      "loss": 1.1029,
+      "step": 1697
+    },
+    {
+      "epoch": 0.3266857056966782,
+      "grad_norm": 0.2906248562989904,
+      "learning_rate": 0.00015727110989396202,
+      "loss": 1.1475,
+      "step": 1698
+    },
+    {
+      "epoch": 0.3268781001052157,
+      "grad_norm": 0.2867376906048281,
+      "learning_rate": 0.00015722001082778646,
+      "loss": 1.1471,
+      "step": 1699
+    },
+    {
+      "epoch": 0.3270704945137532,
+      "grad_norm": 0.27682117457405814,
+      "learning_rate": 0.00015716888953801804,
+      "loss": 1.0052,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3272628889222907,
+      "grad_norm": 0.34681205850819985,
+      "learning_rate": 0.00015711774604451167,
+      "loss": 1.1095,
+      "step": 1701
+    },
+    {
+      "epoch": 0.3274552833308282,
+      "grad_norm": 0.2947816197773247,
+      "learning_rate": 0.0001570665803671309,
+      "loss": 1.1064,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3276476777393657,
+      "grad_norm": 0.32184530624521973,
+      "learning_rate": 0.00015701539252574792,
+      "loss": 1.0734,
+      "step": 1703
+    },
+    {
+      "epoch": 0.3278400721479032,
+      "grad_norm": 0.2540830853998571,
+      "learning_rate": 0.00015696418254024344,
+      "loss": 1.1088,
+      "step": 1704
+    },
+    {
+      "epoch": 0.3280324665564407,
+      "grad_norm": 0.3484982565630624,
+      "learning_rate": 0.00015691295043050688,
+      "loss": 1.068,
+      "step": 1705
+    },
+    {
+      "epoch": 0.3282248609649782,
+      "grad_norm": 0.2719189864386847,
+      "learning_rate": 0.0001568616962164362,
+      "loss": 1.1032,
+      "step": 1706
+    },
+    {
+      "epoch": 0.3284172553735157,
+      "grad_norm": 0.2705620834447423,
+      "learning_rate": 0.0001568104199179379,
+      "loss": 1.0381,
+      "step": 1707
+    },
+    {
+      "epoch": 0.3286096497820532,
+      "grad_norm": 0.45316165264254743,
+      "learning_rate": 0.00015675912155492712,
+      "loss": 1.0574,
+      "step": 1708
+    },
+    {
+      "epoch": 0.3288020441905907,
+      "grad_norm": 0.2893803363115719,
+      "learning_rate": 0.00015670780114732756,
+      "loss": 1.0125,
+      "step": 1709
+    },
+    {
+      "epoch": 0.3289944385991282,
+      "grad_norm": 0.2836083015471957,
+      "learning_rate": 0.00015665645871507151,
+      "loss": 1.1448,
+      "step": 1710
+    },
+    {
+      "epoch": 0.3291868330076657,
+      "grad_norm": 0.33308071873020717,
+      "learning_rate": 0.00015660509427809974,
+      "loss": 1.1653,
+      "step": 1711
+    },
+    {
+      "epoch": 0.32937922741620324,
+      "grad_norm": 0.32566723301114353,
+      "learning_rate": 0.0001565537078563616,
+      "loss": 1.1916,
+      "step": 1712
+    },
+    {
+      "epoch": 0.3295716218247407,
+      "grad_norm": 0.32567450087158395,
+      "learning_rate": 0.000156502299469815,
+      "loss": 1.12,
+      "step": 1713
+    },
+    {
+      "epoch": 0.3297640162332782,
+      "grad_norm": 0.2915856448588911,
+      "learning_rate": 0.00015645086913842636,
+      "loss": 1.0872,
+      "step": 1714
+    },
+    {
+      "epoch": 0.32995641064181574,
+      "grad_norm": 0.2921176902493695,
+      "learning_rate": 0.00015639941688217065,
+      "loss": 1.0837,
+      "step": 1715
+    },
+    {
+      "epoch": 0.3301488050503532,
+      "grad_norm": 0.2832387107531186,
+      "learning_rate": 0.00015634794272103127,
+      "loss": 1.0409,
+      "step": 1716
+    },
+    {
+      "epoch": 0.33034119945889073,
+      "grad_norm": 0.3914146956229172,
+      "learning_rate": 0.0001562964466750003,
+      "loss": 1.0827,
+      "step": 1717
+    },
+    {
+      "epoch": 0.33053359386742825,
+      "grad_norm": 0.40159459338146336,
+      "learning_rate": 0.0001562449287640781,
+      "loss": 1.1236,
+      "step": 1718
+    },
+    {
+      "epoch": 0.3307259882759657,
+      "grad_norm": 0.29390354481910025,
+      "learning_rate": 0.00015619338900827367,
+      "loss": 1.1759,
+      "step": 1719
+    },
+    {
+      "epoch": 0.33091838268450324,
+      "grad_norm": 0.32539541465338706,
+      "learning_rate": 0.00015614182742760448,
+      "loss": 1.1608,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3311107770930407,
+      "grad_norm": 0.36937310139837315,
+      "learning_rate": 0.00015609024404209643,
+      "loss": 1.1444,
+      "step": 1721
+    },
+    {
+      "epoch": 0.33130317150157823,
+      "grad_norm": 0.285692421423644,
+      "learning_rate": 0.00015603863887178393,
+      "loss": 1.1304,
+      "step": 1722
+    },
+    {
+      "epoch": 0.33149556591011575,
+      "grad_norm": 0.33567645622433384,
+      "learning_rate": 0.00015598701193670982,
+      "loss": 1.1621,
+      "step": 1723
+    },
+    {
+      "epoch": 0.3316879603186532,
+      "grad_norm": 0.2936346138536046,
+      "learning_rate": 0.0001559353632569254,
+      "loss": 1.0891,
+      "step": 1724
+    },
+    {
+      "epoch": 0.33188035472719074,
+      "grad_norm": 0.3329249851761191,
+      "learning_rate": 0.00015588369285249047,
+      "loss": 1.1677,
+      "step": 1725
+    },
+    {
+      "epoch": 0.33207274913572826,
+      "grad_norm": 0.25552820385507224,
+      "learning_rate": 0.00015583200074347316,
+      "loss": 1.2004,
+      "step": 1726
+    },
+    {
+      "epoch": 0.33226514354426573,
+      "grad_norm": 0.33284428871176047,
+      "learning_rate": 0.0001557802869499501,
+      "loss": 1.1082,
+      "step": 1727
+    },
+    {
+      "epoch": 0.33245753795280325,
+      "grad_norm": 0.2794737019628236,
+      "learning_rate": 0.00015572855149200638,
+      "loss": 1.1239,
+      "step": 1728
+    },
+    {
+      "epoch": 0.33264993236134077,
+      "grad_norm": 0.2536638971841869,
+      "learning_rate": 0.0001556767943897354,
+      "loss": 1.0684,
+      "step": 1729
+    },
+    {
+      "epoch": 0.33284232676987824,
+      "grad_norm": 0.2849290901393974,
+      "learning_rate": 0.00015562501566323907,
+      "loss": 1.1189,
+      "step": 1730
+    },
+    {
+      "epoch": 0.33303472117841576,
+      "grad_norm": 0.41521589835162076,
+      "learning_rate": 0.0001555732153326276,
+      "loss": 1.1012,
+      "step": 1731
+    },
+    {
+      "epoch": 0.3332271155869532,
+      "grad_norm": 0.37598631027050283,
+      "learning_rate": 0.00015552139341801967,
+      "loss": 1.0954,
+      "step": 1732
+    },
+    {
+      "epoch": 0.33341950999549075,
+      "grad_norm": 0.3044128262013889,
+      "learning_rate": 0.00015546954993954227,
+      "loss": 1.1882,
+      "step": 1733
+    },
+    {
+      "epoch": 0.33361190440402827,
+      "grad_norm": 0.3680301285876295,
+      "learning_rate": 0.00015541768491733092,
+      "loss": 1.0749,
+      "step": 1734
+    },
+    {
+      "epoch": 0.33380429881256574,
+      "grad_norm": 0.3144030411264564,
+      "learning_rate": 0.00015536579837152926,
+      "loss": 1.059,
+      "step": 1735
+    },
+    {
+      "epoch": 0.33399669322110326,
+      "grad_norm": 0.2706787135809171,
+      "learning_rate": 0.00015531389032228955,
+      "loss": 1.0583,
+      "step": 1736
+    },
+    {
+      "epoch": 0.3341890876296408,
+      "grad_norm": 0.30436458206492695,
+      "learning_rate": 0.00015526196078977217,
+      "loss": 1.155,
+      "step": 1737
+    },
+    {
+      "epoch": 0.33438148203817825,
+      "grad_norm": 0.2969114220531906,
+      "learning_rate": 0.00015521000979414602,
+      "loss": 1.0697,
+      "step": 1738
+    },
+    {
+      "epoch": 0.33457387644671577,
+      "grad_norm": 0.392924646839991,
+      "learning_rate": 0.00015515803735558826,
+      "loss": 1.1072,
+      "step": 1739
+    },
+    {
+      "epoch": 0.3347662708552533,
+      "grad_norm": 0.30896485465806317,
+      "learning_rate": 0.00015510604349428436,
+      "loss": 1.1143,
+      "step": 1740
+    },
+    {
+      "epoch": 0.33495866526379076,
+      "grad_norm": 0.3167503249554492,
+      "learning_rate": 0.00015505402823042818,
+      "loss": 1.0169,
+      "step": 1741
+    },
+    {
+      "epoch": 0.3351510596723283,
+      "grad_norm": 0.2964547561570499,
+      "learning_rate": 0.00015500199158422178,
+      "loss": 1.0048,
+      "step": 1742
+    },
+    {
+      "epoch": 0.3353434540808658,
+      "grad_norm": 0.2723709433325831,
+      "learning_rate": 0.0001549499335758757,
+      "loss": 1.077,
+      "step": 1743
+    },
+    {
+      "epoch": 0.33553584848940327,
+      "grad_norm": 0.30477874378702324,
+      "learning_rate": 0.00015489785422560857,
+      "loss": 1.2128,
+      "step": 1744
+    },
+    {
+      "epoch": 0.3357282428979408,
+      "grad_norm": 0.3686132621298599,
+      "learning_rate": 0.00015484575355364743,
+      "loss": 1.1092,
+      "step": 1745
+    },
+    {
+      "epoch": 0.33592063730647825,
+      "grad_norm": 0.29269390666926565,
+      "learning_rate": 0.00015479363158022764,
+      "loss": 1.1855,
+      "step": 1746
+    },
+    {
+      "epoch": 0.3361130317150158,
+      "grad_norm": 0.2747907752684848,
+      "learning_rate": 0.0001547414883255927,
+      "loss": 1.073,
+      "step": 1747
+    },
+    {
+      "epoch": 0.3363054261235533,
+      "grad_norm": 0.4225370662095908,
+      "learning_rate": 0.0001546893238099945,
+      "loss": 1.1993,
+      "step": 1748
+    },
+    {
+      "epoch": 0.33649782053209076,
+      "grad_norm": 0.30229365873123076,
+      "learning_rate": 0.0001546371380536931,
+      "loss": 1.0499,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3366902149406283,
+      "grad_norm": 0.3224313881088821,
+      "learning_rate": 0.00015458493107695686,
+      "loss": 1.0478,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3368826093491658,
+      "grad_norm": 0.3033038838014268,
+      "learning_rate": 0.00015453270290006238,
+      "loss": 1.0132,
+      "step": 1751
+    },
+    {
+      "epoch": 0.3370750037577033,
+      "grad_norm": 0.27948211864946315,
+      "learning_rate": 0.00015448045354329447,
+      "loss": 1.1255,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3372673981662408,
+      "grad_norm": 0.27273257626454467,
+      "learning_rate": 0.00015442818302694618,
+      "loss": 1.1071,
+      "step": 1753
+    },
+    {
+      "epoch": 0.3374597925747783,
+      "grad_norm": 0.34740510402338404,
+      "learning_rate": 0.0001543758913713188,
+      "loss": 1.1502,
+      "step": 1754
+    },
+    {
+      "epoch": 0.3376521869833158,
+      "grad_norm": 0.38611403755365,
+      "learning_rate": 0.00015432357859672177,
+      "loss": 1.1175,
+      "step": 1755
+    },
+    {
+      "epoch": 0.3378445813918533,
+      "grad_norm": 0.28403726016976716,
+      "learning_rate": 0.00015427124472347278,
+      "loss": 1.0419,
+      "step": 1756
+    },
+    {
+      "epoch": 0.3380369758003908,
+      "grad_norm": 0.26934138562037263,
+      "learning_rate": 0.0001542188897718977,
+      "loss": 1.0526,
+      "step": 1757
+    },
+    {
+      "epoch": 0.3382293702089283,
+      "grad_norm": 0.38889725227347716,
+      "learning_rate": 0.00015416651376233063,
+      "loss": 1.0532,
+      "step": 1758
+    },
+    {
+      "epoch": 0.3384217646174658,
+      "grad_norm": 0.2974947687548203,
+      "learning_rate": 0.00015411411671511377,
+      "loss": 1.1613,
+      "step": 1759
+    },
+    {
+      "epoch": 0.3386141590260033,
+      "grad_norm": 0.2875969969750952,
+      "learning_rate": 0.00015406169865059749,
+      "loss": 1.0862,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3388065534345408,
+      "grad_norm": 0.3539143594883471,
+      "learning_rate": 0.00015400925958914042,
+      "loss": 1.0877,
+      "step": 1761
+    },
+    {
+      "epoch": 0.3389989478430783,
+      "grad_norm": 0.2854611950246478,
+      "learning_rate": 0.00015395679955110925,
+      "loss": 1.0127,
+      "step": 1762
+    },
+    {
+      "epoch": 0.3391913422516158,
+      "grad_norm": 0.30701325381728817,
+      "learning_rate": 0.00015390431855687896,
+      "loss": 1.1032,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3393837366601533,
+      "grad_norm": 0.34027250974421197,
+      "learning_rate": 0.00015385181662683244,
+      "loss": 1.0645,
+      "step": 1764
+    },
+    {
+      "epoch": 0.33957613106869083,
+      "grad_norm": 0.2755512030754894,
+      "learning_rate": 0.00015379929378136087,
+      "loss": 1.1812,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3397685254772283,
+      "grad_norm": 0.2721827475655045,
+      "learning_rate": 0.00015374675004086355,
+      "loss": 1.1167,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3399609198857658,
+      "grad_norm": 0.3153367350088956,
+      "learning_rate": 0.00015369418542574782,
+      "loss": 1.0111,
+      "step": 1767
+    },
+    {
+      "epoch": 0.34015331429430334,
+      "grad_norm": 0.3169366995461158,
+      "learning_rate": 0.00015364159995642917,
+      "loss": 1.0217,
+      "step": 1768
+    },
+    {
+      "epoch": 0.3403457087028408,
+      "grad_norm": 0.32611725661062746,
+      "learning_rate": 0.00015358899365333124,
+      "loss": 1.1728,
+      "step": 1769
+    },
+    {
+      "epoch": 0.34053810311137833,
+      "grad_norm": 0.3338928906087186,
+      "learning_rate": 0.00015353636653688563,
+      "loss": 1.1639,
+      "step": 1770
+    },
+    {
+      "epoch": 0.34073049751991585,
+      "grad_norm": 0.32998448488016646,
+      "learning_rate": 0.0001534837186275322,
+      "loss": 1.0466,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3409228919284533,
+      "grad_norm": 0.28215468049046394,
+      "learning_rate": 0.00015343104994571875,
+      "loss": 1.0381,
+      "step": 1772
+    },
+    {
+      "epoch": 0.34111528633699084,
+      "grad_norm": 0.27450755758158707,
+      "learning_rate": 0.0001533783605119012,
+      "loss": 1.1175,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3413076807455283,
+      "grad_norm": 0.30931338530116964,
+      "learning_rate": 0.00015332565034654344,
+      "loss": 1.1249,
+      "step": 1774
+    },
+    {
+      "epoch": 0.34150007515406583,
+      "grad_norm": 0.38731023149783406,
+      "learning_rate": 0.00015327291947011762,
+      "loss": 1.0035,
+      "step": 1775
+    },
+    {
+      "epoch": 0.34169246956260335,
+      "grad_norm": 0.24864118699835444,
+      "learning_rate": 0.00015322016790310372,
+      "loss": 1.0852,
+      "step": 1776
+    },
+    {
+      "epoch": 0.3418848639711408,
+      "grad_norm": 0.405763130348696,
+      "learning_rate": 0.00015316739566598986,
+      "loss": 1.0656,
+      "step": 1777
+    },
+    {
+      "epoch": 0.34207725837967834,
+      "grad_norm": 0.31255940910451313,
+      "learning_rate": 0.00015311460277927217,
+      "loss": 1.1618,
+      "step": 1778
+    },
+    {
+      "epoch": 0.34226965278821586,
+      "grad_norm": 0.25263431760291793,
+      "learning_rate": 0.0001530617892634548,
+      "loss": 1.0342,
+      "step": 1779
+    },
+    {
+      "epoch": 0.3424620471967533,
+      "grad_norm": 0.30834828150676286,
+      "learning_rate": 0.0001530089551390499,
+      "loss": 1.0734,
+      "step": 1780
+    },
+    {
+      "epoch": 0.34265444160529085,
+      "grad_norm": 0.3297408612407537,
+      "learning_rate": 0.0001529561004265777,
+      "loss": 1.0902,
+      "step": 1781
+    },
+    {
+      "epoch": 0.34284683601382837,
+      "grad_norm": 0.26787666509708885,
+      "learning_rate": 0.00015290322514656626,
+      "loss": 1.1457,
+      "step": 1782
+    },
+    {
+      "epoch": 0.34303923042236584,
+      "grad_norm": 0.2627916678647937,
+      "learning_rate": 0.00015285032931955177,
+      "loss": 1.0799,
+      "step": 1783
+    },
+    {
+      "epoch": 0.34323162483090336,
+      "grad_norm": 0.36148219950596344,
+      "learning_rate": 0.0001527974129660784,
+      "loss": 1.1444,
+      "step": 1784
+    },
+    {
+      "epoch": 0.3434240192394409,
+      "grad_norm": 0.330209573855822,
+      "learning_rate": 0.0001527444761066982,
+      "loss": 1.0294,
+      "step": 1785
+    },
+    {
+      "epoch": 0.34361641364797835,
+      "grad_norm": 0.3197845304826271,
+      "learning_rate": 0.00015269151876197125,
+      "loss": 1.148,
+      "step": 1786
+    },
+    {
+      "epoch": 0.34380880805651587,
+      "grad_norm": 0.2790777484992655,
+      "learning_rate": 0.00015263854095246557,
+      "loss": 1.0771,
+      "step": 1787
+    },
+    {
+      "epoch": 0.34400120246505334,
+      "grad_norm": 0.23711577021571012,
+      "learning_rate": 0.00015258554269875717,
+      "loss": 1.0897,
+      "step": 1788
+    },
+    {
+      "epoch": 0.34419359687359086,
+      "grad_norm": 0.25859153045757366,
+      "learning_rate": 0.00015253252402142988,
+      "loss": 1.0724,
+      "step": 1789
+    },
+    {
+      "epoch": 0.3443859912821284,
+      "grad_norm": 0.34452803862415393,
+      "learning_rate": 0.00015247948494107565,
+      "loss": 1.093,
+      "step": 1790
+    },
+    {
+      "epoch": 0.34457838569066584,
+      "grad_norm": 0.3112044740418616,
+      "learning_rate": 0.00015242642547829417,
+      "loss": 1.1033,
+      "step": 1791
+    },
+    {
+      "epoch": 0.34477078009920337,
+      "grad_norm": 0.3302994700873457,
+      "learning_rate": 0.0001523733456536931,
+      "loss": 1.1365,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3449631745077409,
+      "grad_norm": 0.31761706764435255,
+      "learning_rate": 0.00015232024548788813,
+      "loss": 1.1523,
+      "step": 1793
+    },
+    {
+      "epoch": 0.34515556891627835,
+      "grad_norm": 0.3238261409758097,
+      "learning_rate": 0.00015226712500150268,
+      "loss": 1.0589,
+      "step": 1794
+    },
+    {
+      "epoch": 0.3453479633248159,
+      "grad_norm": 0.32046292439174884,
+      "learning_rate": 0.00015221398421516816,
+      "loss": 1.0966,
+      "step": 1795
+    },
+    {
+      "epoch": 0.3455403577333534,
+      "grad_norm": 0.351197956693725,
+      "learning_rate": 0.0001521608231495238,
+      "loss": 1.1068,
+      "step": 1796
+    },
+    {
+      "epoch": 0.34573275214189086,
+      "grad_norm": 0.3723271805530417,
+      "learning_rate": 0.0001521076418252168,
+      "loss": 1.0115,
+      "step": 1797
+    },
+    {
+      "epoch": 0.3459251465504284,
+      "grad_norm": 0.2827945169050502,
+      "learning_rate": 0.00015205444026290216,
+      "loss": 1.0237,
+      "step": 1798
+    },
+    {
+      "epoch": 0.34611754095896585,
+      "grad_norm": 0.32931992868338894,
+      "learning_rate": 0.00015200121848324275,
+      "loss": 1.1826,
+      "step": 1799
+    },
+    {
+      "epoch": 0.3463099353675034,
+      "grad_norm": 0.2623493266489867,
+      "learning_rate": 0.00015194797650690926,
+      "loss": 0.9022,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3465023297760409,
+      "grad_norm": 0.3204710753398161,
+      "learning_rate": 0.0001518947143545803,
+      "loss": 1.1384,
+      "step": 1801
+    },
+    {
+      "epoch": 0.34669472418457836,
+      "grad_norm": 0.28588826461643324,
+      "learning_rate": 0.0001518414320469423,
+      "loss": 1.0765,
+      "step": 1802
+    },
+    {
+      "epoch": 0.3468871185931159,
+      "grad_norm": 0.297112028634499,
+      "learning_rate": 0.00015178812960468945,
+      "loss": 1.0037,
+      "step": 1803
+    },
+    {
+      "epoch": 0.3470795130016534,
+      "grad_norm": 0.2640300049904744,
+      "learning_rate": 0.0001517348070485238,
+      "loss": 1.0881,
+      "step": 1804
+    },
+    {
+      "epoch": 0.34727190741019087,
+      "grad_norm": 0.3509622595181104,
+      "learning_rate": 0.00015168146439915525,
+      "loss": 1.182,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3474643018187284,
+      "grad_norm": 0.31591870890387047,
+      "learning_rate": 0.00015162810167730143,
+      "loss": 1.1381,
+      "step": 1806
+    },
+    {
+      "epoch": 0.3476566962272659,
+      "grad_norm": 0.30513087254168086,
+      "learning_rate": 0.00015157471890368785,
+      "loss": 1.0488,
+      "step": 1807
+    },
+    {
+      "epoch": 0.3478490906358034,
+      "grad_norm": 0.28825378971561394,
+      "learning_rate": 0.0001515213160990477,
+      "loss": 0.9969,
+      "step": 1808
+    },
+    {
+      "epoch": 0.3480414850443409,
+      "grad_norm": 0.32706413159062864,
+      "learning_rate": 0.00015146789328412212,
+      "loss": 1.0874,
+      "step": 1809
+    },
+    {
+      "epoch": 0.3482338794528784,
+      "grad_norm": 0.32015900627141747,
+      "learning_rate": 0.00015141445047965984,
+      "loss": 1.0852,
+      "step": 1810
+    },
+    {
+      "epoch": 0.3484262738614159,
+      "grad_norm": 0.27149440893241417,
+      "learning_rate": 0.0001513609877064174,
+      "loss": 1.1216,
+      "step": 1811
+    },
+    {
+      "epoch": 0.3486186682699534,
+      "grad_norm": 0.289906861337127,
+      "learning_rate": 0.0001513075049851592,
+      "loss": 1.0222,
+      "step": 1812
+    },
+    {
+      "epoch": 0.3488110626784909,
+      "grad_norm": 0.26056187957917737,
+      "learning_rate": 0.0001512540023366573,
+      "loss": 1.0619,
+      "step": 1813
+    },
+    {
+      "epoch": 0.3490034570870284,
+      "grad_norm": 0.31682954136910635,
+      "learning_rate": 0.00015120047978169144,
+      "loss": 1.1224,
+      "step": 1814
+    },
+    {
+      "epoch": 0.3491958514955659,
+      "grad_norm": 0.3374126923949515,
+      "learning_rate": 0.00015114693734104927,
+      "loss": 1.0508,
+      "step": 1815
+    },
+    {
+      "epoch": 0.3493882459041034,
+      "grad_norm": 0.26288242574283976,
+      "learning_rate": 0.00015109337503552595,
+      "loss": 1.11,
+      "step": 1816
+    },
+    {
+      "epoch": 0.3495806403126409,
+      "grad_norm": 0.34899136002242825,
+      "learning_rate": 0.00015103979288592453,
+      "loss": 1.1022,
+      "step": 1817
+    },
+    {
+      "epoch": 0.34977303472117843,
+      "grad_norm": 0.25210360826426065,
+      "learning_rate": 0.0001509861909130557,
+      "loss": 1.0325,
+      "step": 1818
+    },
+    {
+      "epoch": 0.3499654291297159,
+      "grad_norm": 0.31008390775494704,
+      "learning_rate": 0.00015093256913773786,
+      "loss": 1.0619,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3501578235382534,
+      "grad_norm": 0.32120459374616644,
+      "learning_rate": 0.000150878927580797,
+      "loss": 1.1243,
+      "step": 1820
+    },
+    {
+      "epoch": 0.35035021794679094,
+      "grad_norm": 0.312263341500782,
+      "learning_rate": 0.00015082526626306695,
+      "loss": 1.1324,
+      "step": 1821
+    },
+    {
+      "epoch": 0.3505426123553284,
+      "grad_norm": 0.299100340479796,
+      "learning_rate": 0.0001507715852053892,
+      "loss": 1.0759,
+      "step": 1822
+    },
+    {
+      "epoch": 0.35073500676386593,
+      "grad_norm": 0.24478580044375112,
+      "learning_rate": 0.00015071788442861276,
+      "loss": 1.0094,
+      "step": 1823
+    },
+    {
+      "epoch": 0.35092740117240345,
+      "grad_norm": 0.3220735804964353,
+      "learning_rate": 0.00015066416395359444,
+      "loss": 1.0594,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3511197955809409,
+      "grad_norm": 0.32609772204265813,
+      "learning_rate": 0.00015061042380119864,
+      "loss": 1.1503,
+      "step": 1825
+    },
+    {
+      "epoch": 0.35131218998947844,
+      "grad_norm": 0.3235793516980581,
+      "learning_rate": 0.0001505566639922974,
+      "loss": 1.0348,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3515045843980159,
+      "grad_norm": 0.2738300161583872,
+      "learning_rate": 0.00015050288454777046,
+      "loss": 1.0835,
+      "step": 1827
+    },
+    {
+      "epoch": 0.35169697880655343,
+      "grad_norm": 0.3092118576913086,
+      "learning_rate": 0.0001504490854885051,
+      "loss": 1.0572,
+      "step": 1828
+    },
+    {
+      "epoch": 0.35188937321509095,
+      "grad_norm": 0.2927776653833124,
+      "learning_rate": 0.00015039526683539625,
+      "loss": 1.1122,
+      "step": 1829
+    },
+    {
+      "epoch": 0.3520817676236284,
+      "grad_norm": 0.33700553328467114,
+      "learning_rate": 0.0001503414286093465,
+      "loss": 1.0637,
+      "step": 1830
+    },
+    {
+      "epoch": 0.35227416203216594,
+      "grad_norm": 0.2629123021983572,
+      "learning_rate": 0.00015028757083126592,
+      "loss": 0.9553,
+      "step": 1831
+    },
+    {
+      "epoch": 0.35246655644070346,
+      "grad_norm": 0.3587618540457818,
+      "learning_rate": 0.00015023369352207229,
+      "loss": 1.1639,
+      "step": 1832
+    },
+    {
+      "epoch": 0.3526589508492409,
+      "grad_norm": 0.2597243625037536,
+      "learning_rate": 0.00015017979670269095,
+      "loss": 1.1174,
+      "step": 1833
+    },
+    {
+      "epoch": 0.35285134525777845,
+      "grad_norm": 0.2657473337966056,
+      "learning_rate": 0.0001501258803940548,
+      "loss": 1.0376,
+      "step": 1834
+    },
+    {
+      "epoch": 0.35304373966631597,
+      "grad_norm": 0.2800192506752259,
+      "learning_rate": 0.0001500719446171043,
+      "loss": 1.0763,
+      "step": 1835
+    },
+    {
+      "epoch": 0.35323613407485344,
+      "grad_norm": 0.35641513034473116,
+      "learning_rate": 0.00015001798939278752,
+      "loss": 1.1407,
+      "step": 1836
+    },
+    {
+      "epoch": 0.35342852848339096,
+      "grad_norm": 0.29347724709111284,
+      "learning_rate": 0.00014996401474205997,
+      "loss": 1.1292,
+      "step": 1837
+    },
+    {
+      "epoch": 0.3536209228919285,
+      "grad_norm": 0.39734301227070035,
+      "learning_rate": 0.00014991002068588484,
+      "loss": 1.0724,
+      "step": 1838
+    },
+    {
+      "epoch": 0.35381331730046595,
+      "grad_norm": 0.27775395129926944,
+      "learning_rate": 0.0001498560072452328,
+      "loss": 1.1082,
+      "step": 1839
+    },
+    {
+      "epoch": 0.35400571170900347,
+      "grad_norm": 0.3014670784927019,
+      "learning_rate": 0.00014980197444108205,
+      "loss": 1.0971,
+      "step": 1840
+    },
+    {
+      "epoch": 0.35419810611754093,
+      "grad_norm": 0.3378754581872636,
+      "learning_rate": 0.00014974792229441826,
+      "loss": 1.1237,
+      "step": 1841
+    },
+    {
+      "epoch": 0.35439050052607846,
+      "grad_norm": 0.27128828355443557,
+      "learning_rate": 0.00014969385082623472,
+      "loss": 1.0729,
+      "step": 1842
+    },
+    {
+      "epoch": 0.354582894934616,
+      "grad_norm": 0.3298426117789897,
+      "learning_rate": 0.00014963976005753215,
+      "loss": 1.0834,
+      "step": 1843
+    },
+    {
+      "epoch": 0.35477528934315344,
+      "grad_norm": 0.343659239962405,
+      "learning_rate": 0.00014958565000931876,
+      "loss": 0.9839,
+      "step": 1844
+    },
+    {
+      "epoch": 0.35496768375169097,
+      "grad_norm": 0.278930524450225,
+      "learning_rate": 0.00014953152070261026,
+      "loss": 1.0054,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3551600781602285,
+      "grad_norm": 0.29318306702917285,
+      "learning_rate": 0.0001494773721584299,
+      "loss": 1.1171,
+      "step": 1846
+    },
+    {
+      "epoch": 0.35535247256876595,
+      "grad_norm": 0.30681604941453455,
+      "learning_rate": 0.0001494232043978083,
+      "loss": 1.1272,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3555448669773035,
+      "grad_norm": 0.3147653987995692,
+      "learning_rate": 0.00014936901744178367,
+      "loss": 1.1555,
+      "step": 1848
+    },
+    {
+      "epoch": 0.355737261385841,
+      "grad_norm": 0.303426632944889,
+      "learning_rate": 0.00014931481131140147,
+      "loss": 1.0929,
+      "step": 1849
+    },
+    {
+      "epoch": 0.35592965579437846,
+      "grad_norm": 0.27165012128240573,
+      "learning_rate": 0.00014926058602771484,
+      "loss": 1.137,
+      "step": 1850
+    },
+    {
+      "epoch": 0.356122050202916,
+      "grad_norm": 0.3075775481675345,
+      "learning_rate": 0.00014920634161178425,
+      "loss": 1.1136,
+      "step": 1851
+    },
+    {
+      "epoch": 0.3563144446114535,
+      "grad_norm": 0.34846587137004936,
+      "learning_rate": 0.00014915207808467756,
+      "loss": 1.0719,
+      "step": 1852
+    },
+    {
+      "epoch": 0.356506839019991,
+      "grad_norm": 0.3766285481356455,
+      "learning_rate": 0.0001490977954674701,
+      "loss": 1.0861,
+      "step": 1853
+    },
+    {
+      "epoch": 0.3566992334285285,
+      "grad_norm": 0.2686058052969888,
+      "learning_rate": 0.00014904349378124467,
+      "loss": 1.1716,
+      "step": 1854
+    },
+    {
+      "epoch": 0.35689162783706596,
+      "grad_norm": 0.27299597606529447,
+      "learning_rate": 0.0001489891730470914,
+      "loss": 1.0009,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3570840222456035,
+      "grad_norm": 0.37283271345094865,
+      "learning_rate": 0.00014893483328610777,
+      "loss": 1.0428,
+      "step": 1856
+    },
+    {
+      "epoch": 0.357276416654141,
+      "grad_norm": 0.4309161089859467,
+      "learning_rate": 0.0001488804745193988,
+      "loss": 1.0318,
+      "step": 1857
+    },
+    {
+      "epoch": 0.35746881106267847,
+      "grad_norm": 0.35053829244215123,
+      "learning_rate": 0.00014882609676807675,
+      "loss": 1.0359,
+      "step": 1858
+    },
+    {
+      "epoch": 0.357661205471216,
+      "grad_norm": 0.31613241719091206,
+      "learning_rate": 0.00014877170005326136,
+      "loss": 1.0896,
+      "step": 1859
+    },
+    {
+      "epoch": 0.3578535998797535,
+      "grad_norm": 0.432863418358156,
+      "learning_rate": 0.00014871728439607966,
+      "loss": 1.0949,
+      "step": 1860
+    },
+    {
+      "epoch": 0.358045994288291,
+      "grad_norm": 0.4319979447427512,
+      "learning_rate": 0.00014866284981766606,
+      "loss": 1.0643,
+      "step": 1861
+    },
+    {
+      "epoch": 0.3582383886968285,
+      "grad_norm": 0.3139741896531891,
+      "learning_rate": 0.00014860839633916236,
+      "loss": 1.1716,
+      "step": 1862
+    },
+    {
+      "epoch": 0.358430783105366,
+      "grad_norm": 0.31078885130802425,
+      "learning_rate": 0.0001485539239817176,
+      "loss": 1.0672,
+      "step": 1863
+    },
+    {
+      "epoch": 0.3586231775139035,
+      "grad_norm": 0.3291140952595944,
+      "learning_rate": 0.0001484994327664883,
+      "loss": 1.1039,
+      "step": 1864
+    },
+    {
+      "epoch": 0.358815571922441,
+      "grad_norm": 0.3377318125162411,
+      "learning_rate": 0.0001484449227146381,
+      "loss": 1.1793,
+      "step": 1865
+    },
+    {
+      "epoch": 0.3590079663309785,
+      "grad_norm": 0.29210983351750236,
+      "learning_rate": 0.0001483903938473382,
+      "loss": 1.1435,
+      "step": 1866
+    },
+    {
+      "epoch": 0.359200360739516,
+      "grad_norm": 0.29840851672761753,
+      "learning_rate": 0.00014833584618576695,
+      "loss": 1.1924,
+      "step": 1867
+    },
+    {
+      "epoch": 0.3593927551480535,
+      "grad_norm": 0.33487483726424955,
+      "learning_rate": 0.00014828127975111,
+      "loss": 1.0845,
+      "step": 1868
+    },
+    {
+      "epoch": 0.359585149556591,
+      "grad_norm": 0.286279541824635,
+      "learning_rate": 0.0001482266945645603,
+      "loss": 1.0551,
+      "step": 1869
+    },
+    {
+      "epoch": 0.3597775439651285,
+      "grad_norm": 0.2991117765472499,
+      "learning_rate": 0.00014817209064731817,
+      "loss": 1.1261,
+      "step": 1870
+    },
+    {
+      "epoch": 0.35996993837366603,
+      "grad_norm": 0.33380266102838413,
+      "learning_rate": 0.00014811746802059113,
+      "loss": 1.0044,
+      "step": 1871
+    },
+    {
+      "epoch": 0.3601623327822035,
+      "grad_norm": 0.344713321414533,
+      "learning_rate": 0.000148062826705594,
+      "loss": 1.0734,
+      "step": 1872
+    },
+    {
+      "epoch": 0.360354727190741,
+      "grad_norm": 0.2724585130564941,
+      "learning_rate": 0.00014800816672354877,
+      "loss": 1.0704,
+      "step": 1873
+    },
+    {
+      "epoch": 0.36054712159927854,
+      "grad_norm": 0.2642015365539861,
+      "learning_rate": 0.00014795348809568476,
+      "loss": 1.1128,
+      "step": 1874
+    },
+    {
+      "epoch": 0.360739516007816,
+      "grad_norm": 0.3534706986703852,
+      "learning_rate": 0.00014789879084323857,
+      "loss": 1.1588,
+      "step": 1875
+    },
+    {
+      "epoch": 0.36093191041635353,
+      "grad_norm": 0.25537674601980054,
+      "learning_rate": 0.00014784407498745394,
+      "loss": 1.0076,
+      "step": 1876
+    },
+    {
+      "epoch": 0.36112430482489105,
+      "grad_norm": 0.27894783927574934,
+      "learning_rate": 0.0001477893405495819,
+      "loss": 1.0879,
+      "step": 1877
+    },
+    {
+      "epoch": 0.3613166992334285,
+      "grad_norm": 0.37599560464547327,
+      "learning_rate": 0.00014773458755088068,
+      "loss": 1.1295,
+      "step": 1878
+    },
+    {
+      "epoch": 0.36150909364196604,
+      "grad_norm": 0.36789010878467643,
+      "learning_rate": 0.00014767981601261567,
+      "loss": 1.0879,
+      "step": 1879
+    },
+    {
+      "epoch": 0.3617014880505035,
+      "grad_norm": 0.30677080578676985,
+      "learning_rate": 0.00014762502595605956,
+      "loss": 1.0759,
+      "step": 1880
+    },
+    {
+      "epoch": 0.361893882459041,
+      "grad_norm": 0.27532868669205685,
+      "learning_rate": 0.00014757021740249214,
+      "loss": 1.1114,
+      "step": 1881
+    },
+    {
+      "epoch": 0.36208627686757855,
+      "grad_norm": 0.28652231343501494,
+      "learning_rate": 0.00014751539037320043,
+      "loss": 0.9999,
+      "step": 1882
+    },
+    {
+      "epoch": 0.362278671276116,
+      "grad_norm": 0.2796722496990979,
+      "learning_rate": 0.00014746054488947862,
+      "loss": 1.0701,
+      "step": 1883
+    },
+    {
+      "epoch": 0.36247106568465354,
+      "grad_norm": 0.3076328280684436,
+      "learning_rate": 0.00014740568097262812,
+      "loss": 1.1187,
+      "step": 1884
+    },
+    {
+      "epoch": 0.36266346009319106,
+      "grad_norm": 0.280827738363737,
+      "learning_rate": 0.00014735079864395729,
+      "loss": 1.0844,
+      "step": 1885
+    },
+    {
+      "epoch": 0.3628558545017285,
+      "grad_norm": 0.26475922736819607,
+      "learning_rate": 0.00014729589792478192,
+      "loss": 1.048,
+      "step": 1886
+    },
+    {
+      "epoch": 0.36304824891026605,
+      "grad_norm": 0.3512053032736165,
+      "learning_rate": 0.00014724097883642482,
+      "loss": 1.1478,
+      "step": 1887
+    },
+    {
+      "epoch": 0.36324064331880357,
+      "grad_norm": 0.26768796331119143,
+      "learning_rate": 0.00014718604140021588,
+      "loss": 1.0983,
+      "step": 1888
+    },
+    {
+      "epoch": 0.36343303772734104,
+      "grad_norm": 0.3144776219272371,
+      "learning_rate": 0.0001471310856374922,
+      "loss": 1.0266,
+      "step": 1889
+    },
+    {
+      "epoch": 0.36362543213587856,
+      "grad_norm": 0.302157941849477,
+      "learning_rate": 0.0001470761115695979,
+      "loss": 1.015,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3638178265444161,
+      "grad_norm": 0.22398903107315207,
+      "learning_rate": 0.00014702111921788437,
+      "loss": 1.1601,
+      "step": 1891
+    },
+    {
+      "epoch": 0.36401022095295354,
+      "grad_norm": 0.350950329568758,
+      "learning_rate": 0.00014696610860370996,
+      "loss": 1.0775,
+      "step": 1892
+    },
+    {
+      "epoch": 0.36420261536149107,
+      "grad_norm": 0.38003015568613585,
+      "learning_rate": 0.00014691107974844014,
+      "loss": 1.0482,
+      "step": 1893
+    },
+    {
+      "epoch": 0.36439500977002853,
+      "grad_norm": 0.30681054747715425,
+      "learning_rate": 0.0001468560326734475,
+      "loss": 1.0904,
+      "step": 1894
+    },
+    {
+      "epoch": 0.36458740417856605,
+      "grad_norm": 0.28091694354639735,
+      "learning_rate": 0.00014680096740011172,
+      "loss": 1.1858,
+      "step": 1895
+    },
+    {
+      "epoch": 0.3647797985871036,
+      "grad_norm": 0.291599722468472,
+      "learning_rate": 0.00014674588394981947,
+      "loss": 1.1751,
+      "step": 1896
+    },
+    {
+      "epoch": 0.36497219299564104,
+      "grad_norm": 0.2614752988961378,
+      "learning_rate": 0.00014669078234396453,
+      "loss": 1.0868,
+      "step": 1897
+    },
+    {
+      "epoch": 0.36516458740417856,
+      "grad_norm": 0.2989790066668176,
+      "learning_rate": 0.00014663566260394775,
+      "loss": 1.1482,
+      "step": 1898
+    },
+    {
+      "epoch": 0.3653569818127161,
+      "grad_norm": 0.3141103454325244,
+      "learning_rate": 0.00014658052475117702,
+      "loss": 1.1347,
+      "step": 1899
+    },
+    {
+      "epoch": 0.36554937622125355,
+      "grad_norm": 0.29283849849484106,
+      "learning_rate": 0.0001465253688070672,
+      "loss": 1.1036,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3657417706297911,
+      "grad_norm": 0.26689509763726227,
+      "learning_rate": 0.00014647019479304028,
+      "loss": 1.0175,
+      "step": 1901
+    },
+    {
+      "epoch": 0.3659341650383286,
+      "grad_norm": 0.4247693096984541,
+      "learning_rate": 0.00014641500273052514,
+      "loss": 1.1299,
+      "step": 1902
+    },
+    {
+      "epoch": 0.36612655944686606,
+      "grad_norm": 0.33323301557293616,
+      "learning_rate": 0.0001463597926409578,
+      "loss": 1.0762,
+      "step": 1903
+    },
+    {
+      "epoch": 0.3663189538554036,
+      "grad_norm": 0.32793474311197524,
+      "learning_rate": 0.0001463045645457812,
+      "loss": 1.0773,
+      "step": 1904
+    },
+    {
+      "epoch": 0.3665113482639411,
+      "grad_norm": 0.34854491719274083,
+      "learning_rate": 0.0001462493184664453,
+      "loss": 1.0637,
+      "step": 1905
+    },
+    {
+      "epoch": 0.36670374267247857,
+      "grad_norm": 0.39320003066251424,
+      "learning_rate": 0.000146194054424407,
+      "loss": 1.18,
+      "step": 1906
+    },
+    {
+      "epoch": 0.3668961370810161,
+      "grad_norm": 0.3027938909935693,
+      "learning_rate": 0.00014613877244113032,
+      "loss": 1.0773,
+      "step": 1907
+    },
+    {
+      "epoch": 0.36708853148955356,
+      "grad_norm": 0.3029114678045782,
+      "learning_rate": 0.00014608347253808606,
+      "loss": 1.0866,
+      "step": 1908
+    },
+    {
+      "epoch": 0.3672809258980911,
+      "grad_norm": 0.284655055825206,
+      "learning_rate": 0.0001460281547367521,
+      "loss": 1.1031,
+      "step": 1909
+    },
+    {
+      "epoch": 0.3674733203066286,
+      "grad_norm": 0.36640965978321494,
+      "learning_rate": 0.0001459728190586132,
+      "loss": 1.0476,
+      "step": 1910
+    },
+    {
+      "epoch": 0.36766571471516607,
+      "grad_norm": 0.29089276480726395,
+      "learning_rate": 0.00014591746552516108,
+      "loss": 1.1077,
+      "step": 1911
+    },
+    {
+      "epoch": 0.3678581091237036,
+      "grad_norm": 0.290937824411561,
+      "learning_rate": 0.0001458620941578945,
+      "loss": 1.1023,
+      "step": 1912
+    },
+    {
+      "epoch": 0.3680505035322411,
+      "grad_norm": 0.2891360019158722,
+      "learning_rate": 0.00014580670497831903,
+      "loss": 1.1717,
+      "step": 1913
+    },
+    {
+      "epoch": 0.3682428979407786,
+      "grad_norm": 0.28263787625557674,
+      "learning_rate": 0.00014575129800794718,
+      "loss": 1.0419,
+      "step": 1914
+    },
+    {
+      "epoch": 0.3684352923493161,
+      "grad_norm": 0.30818924629002703,
+      "learning_rate": 0.00014569587326829835,
+      "loss": 1.089,
+      "step": 1915
+    },
+    {
+      "epoch": 0.3686276867578536,
+      "grad_norm": 0.29294698405037756,
+      "learning_rate": 0.0001456404307808989,
+      "loss": 1.1865,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3688200811663911,
+      "grad_norm": 0.29451444432322943,
+      "learning_rate": 0.00014558497056728204,
+      "loss": 1.1114,
+      "step": 1917
+    },
+    {
+      "epoch": 0.3690124755749286,
+      "grad_norm": 0.315494682217328,
+      "learning_rate": 0.00014552949264898794,
+      "loss": 1.1967,
+      "step": 1918
+    },
+    {
+      "epoch": 0.36920486998346613,
+      "grad_norm": 0.2959856348141984,
+      "learning_rate": 0.00014547399704756348,
+      "loss": 1.0801,
+      "step": 1919
+    },
+    {
+      "epoch": 0.3693972643920036,
+      "grad_norm": 0.3163604706547772,
+      "learning_rate": 0.00014541848378456256,
+      "loss": 1.0565,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3695896588005411,
+      "grad_norm": 0.3176996527827041,
+      "learning_rate": 0.00014536295288154592,
+      "loss": 1.0979,
+      "step": 1921
+    },
+    {
+      "epoch": 0.3697820532090786,
+      "grad_norm": 0.28799658236242526,
+      "learning_rate": 0.0001453074043600811,
+      "loss": 1.1257,
+      "step": 1922
+    },
+    {
+      "epoch": 0.3699744476176161,
+      "grad_norm": 0.31411010622190016,
+      "learning_rate": 0.00014525183824174248,
+      "loss": 0.9577,
+      "step": 1923
+    },
+    {
+      "epoch": 0.37016684202615363,
+      "grad_norm": 0.25421927880015555,
+      "learning_rate": 0.00014519625454811135,
+      "loss": 1.1084,
+      "step": 1924
+    },
+    {
+      "epoch": 0.3703592364346911,
+      "grad_norm": 0.32865048509904704,
+      "learning_rate": 0.00014514065330077576,
+      "loss": 1.1339,
+      "step": 1925
+    },
+    {
+      "epoch": 0.3705516308432286,
+      "grad_norm": 0.3242171630569951,
+      "learning_rate": 0.00014508503452133052,
+      "loss": 1.1099,
+      "step": 1926
+    },
+    {
+      "epoch": 0.37074402525176614,
+      "grad_norm": 0.31137280682818663,
+      "learning_rate": 0.00014502939823137745,
+      "loss": 1.1093,
+      "step": 1927
+    },
+    {
+      "epoch": 0.3709364196603036,
+      "grad_norm": 0.27958562705832646,
+      "learning_rate": 0.00014497374445252497,
+      "loss": 1.0681,
+      "step": 1928
+    },
+    {
+      "epoch": 0.37112881406884113,
+      "grad_norm": 0.3657785587537535,
+      "learning_rate": 0.00014491807320638834,
+      "loss": 1.1187,
+      "step": 1929
+    },
+    {
+      "epoch": 0.37132120847737865,
+      "grad_norm": 0.2527408477986618,
+      "learning_rate": 0.0001448623845145897,
+      "loss": 1.0952,
+      "step": 1930
+    },
+    {
+      "epoch": 0.3715136028859161,
+      "grad_norm": 0.2958644187223314,
+      "learning_rate": 0.00014480667839875786,
+      "loss": 1.1191,
+      "step": 1931
+    },
+    {
+      "epoch": 0.37170599729445364,
+      "grad_norm": 0.367129519008423,
+      "learning_rate": 0.00014475095488052843,
+      "loss": 1.0967,
+      "step": 1932
+    },
+    {
+      "epoch": 0.3718983917029911,
+      "grad_norm": 0.3093776216584789,
+      "learning_rate": 0.0001446952139815438,
+      "loss": 1.1141,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3720907861115286,
+      "grad_norm": 0.3012883217871457,
+      "learning_rate": 0.00014463945572345307,
+      "loss": 1.0621,
+      "step": 1934
+    },
+    {
+      "epoch": 0.37228318052006615,
+      "grad_norm": 0.30749918181400143,
+      "learning_rate": 0.00014458368012791213,
+      "loss": 1.1066,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3724755749286036,
+      "grad_norm": 0.3389656998187917,
+      "learning_rate": 0.00014452788721658355,
+      "loss": 1.1911,
+      "step": 1936
+    },
+    {
+      "epoch": 0.37266796933714114,
+      "grad_norm": 0.2576045603691585,
+      "learning_rate": 0.00014447207701113668,
+      "loss": 1.1183,
+      "step": 1937
+    },
+    {
+      "epoch": 0.37286036374567866,
+      "grad_norm": 0.3221752712214336,
+      "learning_rate": 0.00014441624953324755,
+      "loss": 1.1001,
+      "step": 1938
+    },
+    {
+      "epoch": 0.3730527581542161,
+      "grad_norm": 0.3028885729414909,
+      "learning_rate": 0.00014436040480459892,
+      "loss": 1.1018,
+      "step": 1939
+    },
+    {
+      "epoch": 0.37324515256275365,
+      "grad_norm": 0.26494023180236825,
+      "learning_rate": 0.0001443045428468802,
+      "loss": 1.0562,
+      "step": 1940
+    },
+    {
+      "epoch": 0.37343754697129117,
+      "grad_norm": 0.2864666905192389,
+      "learning_rate": 0.0001442486636817876,
+      "loss": 1.1579,
+      "step": 1941
+    },
+    {
+      "epoch": 0.37362994137982863,
+      "grad_norm": 0.27285037768947795,
+      "learning_rate": 0.00014419276733102388,
+      "loss": 1.1256,
+      "step": 1942
+    },
+    {
+      "epoch": 0.37382233578836616,
+      "grad_norm": 0.28600232582889934,
+      "learning_rate": 0.00014413685381629856,
+      "loss": 1.0829,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3740147301969037,
+      "grad_norm": 0.2980899104623963,
+      "learning_rate": 0.0001440809231593278,
+      "loss": 1.0274,
+      "step": 1944
+    },
+    {
+      "epoch": 0.37420712460544114,
+      "grad_norm": 0.3226759823491168,
+      "learning_rate": 0.00014402497538183444,
+      "loss": 1.0545,
+      "step": 1945
+    },
+    {
+      "epoch": 0.37439951901397867,
+      "grad_norm": 0.2695213506096957,
+      "learning_rate": 0.00014396901050554793,
+      "loss": 1.0513,
+      "step": 1946
+    },
+    {
+      "epoch": 0.37459191342251613,
+      "grad_norm": 0.38507862268510473,
+      "learning_rate": 0.0001439130285522044,
+      "loss": 1.0645,
+      "step": 1947
+    },
+    {
+      "epoch": 0.37478430783105365,
+      "grad_norm": 0.29164899375142106,
+      "learning_rate": 0.0001438570295435466,
+      "loss": 1.0832,
+      "step": 1948
+    },
+    {
+      "epoch": 0.3749767022395912,
+      "grad_norm": 0.31056433981839915,
+      "learning_rate": 0.00014380101350132388,
+      "loss": 1.1543,
+      "step": 1949
+    },
+    {
+      "epoch": 0.37516909664812864,
+      "grad_norm": 0.2680192797807722,
+      "learning_rate": 0.00014374498044729227,
+      "loss": 1.1139,
+      "step": 1950
+    },
+    {
+      "epoch": 0.37536149105666616,
+      "grad_norm": 0.2821569933474121,
+      "learning_rate": 0.00014368893040321427,
+      "loss": 1.0733,
+      "step": 1951
+    },
+    {
+      "epoch": 0.3755538854652037,
+      "grad_norm": 0.3201459023484589,
+      "learning_rate": 0.00014363286339085914,
+      "loss": 1.1127,
+      "step": 1952
+    },
+    {
+      "epoch": 0.37574627987374115,
+      "grad_norm": 0.29164255169627756,
+      "learning_rate": 0.0001435767794320027,
+      "loss": 1.0496,
+      "step": 1953
+    },
+    {
+      "epoch": 0.3759386742822787,
+      "grad_norm": 0.2906523835438739,
+      "learning_rate": 0.00014352067854842723,
+      "loss": 1.1041,
+      "step": 1954
+    },
+    {
+      "epoch": 0.3761310686908162,
+      "grad_norm": 0.2680536260721687,
+      "learning_rate": 0.0001434645607619217,
+      "loss": 1.1774,
+      "step": 1955
+    },
+    {
+      "epoch": 0.37632346309935366,
+      "grad_norm": 0.25914136042921976,
+      "learning_rate": 0.0001434084260942816,
+      "loss": 1.0613,
+      "step": 1956
+    },
+    {
+      "epoch": 0.3765158575078912,
+      "grad_norm": 0.28862573938948977,
+      "learning_rate": 0.000143352274567309,
+      "loss": 1.0751,
+      "step": 1957
+    },
+    {
+      "epoch": 0.3767082519164287,
+      "grad_norm": 0.3023902117097119,
+      "learning_rate": 0.00014329610620281253,
+      "loss": 1.0843,
+      "step": 1958
+    },
+    {
+      "epoch": 0.37690064632496617,
+      "grad_norm": 0.35296921790886787,
+      "learning_rate": 0.00014323992102260733,
+      "loss": 1.0453,
+      "step": 1959
+    },
+    {
+      "epoch": 0.3770930407335037,
+      "grad_norm": 0.2870603460152669,
+      "learning_rate": 0.00014318371904851503,
+      "loss": 1.0963,
+      "step": 1960
+    },
+    {
+      "epoch": 0.37728543514204116,
+      "grad_norm": 0.26613303087991497,
+      "learning_rate": 0.00014312750030236383,
+      "loss": 1.1107,
+      "step": 1961
+    },
+    {
+      "epoch": 0.3774778295505787,
+      "grad_norm": 0.3064795217516217,
+      "learning_rate": 0.0001430712648059885,
+      "loss": 1.0534,
+      "step": 1962
+    },
+    {
+      "epoch": 0.3776702239591162,
+      "grad_norm": 0.34496826674265324,
+      "learning_rate": 0.00014301501258123025,
+      "loss": 1.0758,
+      "step": 1963
+    },
+    {
+      "epoch": 0.37786261836765367,
+      "grad_norm": 0.34103971079027146,
+      "learning_rate": 0.0001429587436499367,
+      "loss": 1.0877,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3780550127761912,
+      "grad_norm": 0.3143255351612261,
+      "learning_rate": 0.0001429024580339622,
+      "loss": 1.0223,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3782474071847287,
+      "grad_norm": 0.3099969175356247,
+      "learning_rate": 0.00014284615575516738,
+      "loss": 1.1331,
+      "step": 1966
+    },
+    {
+      "epoch": 0.3784398015932662,
+      "grad_norm": 0.2995963604391879,
+      "learning_rate": 0.00014278983683541934,
+      "loss": 1.1212,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3786321960018037,
+      "grad_norm": 0.26594334986357204,
+      "learning_rate": 0.00014273350129659172,
+      "loss": 1.0299,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3788245904103412,
+      "grad_norm": 0.3663847970446549,
+      "learning_rate": 0.00014267714916056464,
+      "loss": 1.1724,
+      "step": 1969
+    },
+    {
+      "epoch": 0.3790169848188787,
+      "grad_norm": 0.2670572257110239,
+      "learning_rate": 0.0001426207804492246,
+      "loss": 1.1415,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3792093792274162,
+      "grad_norm": 0.28044351715522375,
+      "learning_rate": 0.00014256439518446455,
+      "loss": 1.1028,
+      "step": 1971
+    },
+    {
+      "epoch": 0.37940177363595373,
+      "grad_norm": 0.3372370384754339,
+      "learning_rate": 0.00014250799338818388,
+      "loss": 1.1881,
+      "step": 1972
+    },
+    {
+      "epoch": 0.3795941680444912,
+      "grad_norm": 0.28017107256950113,
+      "learning_rate": 0.00014245157508228839,
+      "loss": 1.0563,
+      "step": 1973
+    },
+    {
+      "epoch": 0.3797865624530287,
+      "grad_norm": 0.298544765390121,
+      "learning_rate": 0.00014239514028869033,
+      "loss": 1.1141,
+      "step": 1974
+    },
+    {
+      "epoch": 0.3799789568615662,
+      "grad_norm": 0.30277588433873626,
+      "learning_rate": 0.00014233868902930826,
+      "loss": 1.1369,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3801713512701037,
+      "grad_norm": 0.2943144115185054,
+      "learning_rate": 0.0001422822213260673,
+      "loss": 1.1295,
+      "step": 1976
+    },
+    {
+      "epoch": 0.38036374567864123,
+      "grad_norm": 0.28713604502386886,
+      "learning_rate": 0.00014222573720089874,
+      "loss": 1.0605,
+      "step": 1977
+    },
+    {
+      "epoch": 0.3805561400871787,
+      "grad_norm": 0.33591508231816775,
+      "learning_rate": 0.00014216923667574043,
+      "loss": 1.1793,
+      "step": 1978
+    },
+    {
+      "epoch": 0.3807485344957162,
+      "grad_norm": 0.34764835548957024,
+      "learning_rate": 0.00014211271977253653,
+      "loss": 1.0707,
+      "step": 1979
+    },
+    {
+      "epoch": 0.38094092890425374,
+      "grad_norm": 0.2954517975798876,
+      "learning_rate": 0.00014205618651323752,
+      "loss": 1.1271,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3811333233127912,
+      "grad_norm": 0.3377212307135022,
+      "learning_rate": 0.00014199963691980026,
+      "loss": 1.1592,
+      "step": 1981
+    },
+    {
+      "epoch": 0.3813257177213287,
+      "grad_norm": 0.25759934385625344,
+      "learning_rate": 0.00014194307101418805,
+      "loss": 1.0455,
+      "step": 1982
+    },
+    {
+      "epoch": 0.38151811212986625,
+      "grad_norm": 0.26686972307314477,
+      "learning_rate": 0.00014188648881837033,
+      "loss": 1.1715,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3817105065384037,
+      "grad_norm": 0.29984112996475654,
+      "learning_rate": 0.00014182989035432298,
+      "loss": 1.1334,
+      "step": 1984
+    },
+    {
+      "epoch": 0.38190290094694124,
+      "grad_norm": 0.35255728654768675,
+      "learning_rate": 0.00014177327564402826,
+      "loss": 1.0833,
+      "step": 1985
+    },
+    {
+      "epoch": 0.38209529535547876,
+      "grad_norm": 0.39304632772877357,
+      "learning_rate": 0.00014171664470947465,
+      "loss": 1.0845,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3822876897640162,
+      "grad_norm": 0.3839754251032507,
+      "learning_rate": 0.0001416599975726569,
+      "loss": 1.1347,
+      "step": 1987
+    },
+    {
+      "epoch": 0.38248008417255375,
+      "grad_norm": 0.25505784855196545,
+      "learning_rate": 0.00014160333425557615,
+      "loss": 1.0058,
+      "step": 1988
+    },
+    {
+      "epoch": 0.3826724785810912,
+      "grad_norm": 0.3000442649687535,
+      "learning_rate": 0.00014154665478023976,
+      "loss": 1.1342,
+      "step": 1989
+    },
+    {
+      "epoch": 0.38286487298962874,
+      "grad_norm": 0.2576329171643482,
+      "learning_rate": 0.0001414899591686614,
+      "loss": 1.103,
+      "step": 1990
+    },
+    {
+      "epoch": 0.38305726739816626,
+      "grad_norm": 0.3656325396461047,
+      "learning_rate": 0.000141433247442861,
+      "loss": 0.9734,
+      "step": 1991
+    },
+    {
+      "epoch": 0.3832496618067037,
+      "grad_norm": 0.29380429146326276,
+      "learning_rate": 0.0001413765196248647,
+      "loss": 1.0384,
+      "step": 1992
+    },
+    {
+      "epoch": 0.38344205621524124,
+      "grad_norm": 0.30529503815508935,
+      "learning_rate": 0.00014131977573670498,
+      "loss": 1.1147,
+      "step": 1993
+    },
+    {
+      "epoch": 0.38363445062377877,
+      "grad_norm": 0.32160975147219234,
+      "learning_rate": 0.0001412630158004205,
+      "loss": 1.0596,
+      "step": 1994
+    },
+    {
+      "epoch": 0.38382684503231623,
+      "grad_norm": 0.3663337398837326,
+      "learning_rate": 0.00014120623983805616,
+      "loss": 1.0174,
+      "step": 1995
+    },
+    {
+      "epoch": 0.38401923944085375,
+      "grad_norm": 0.27165311841715334,
+      "learning_rate": 0.00014114944787166307,
+      "loss": 1.1311,
+      "step": 1996
+    },
+    {
+      "epoch": 0.3842116338493913,
+      "grad_norm": 0.28057625346601744,
+      "learning_rate": 0.00014109263992329857,
+      "loss": 1.1139,
+      "step": 1997
+    },
+    {
+      "epoch": 0.38440402825792874,
+      "grad_norm": 0.3152161600145858,
+      "learning_rate": 0.0001410358160150263,
+      "loss": 1.1688,
+      "step": 1998
+    },
+    {
+      "epoch": 0.38459642266646626,
+      "grad_norm": 0.2699550529752809,
+      "learning_rate": 0.0001409789761689159,
+      "loss": 1.1206,
+      "step": 1999
+    },
+    {
+      "epoch": 0.38478881707500373,
+      "grad_norm": 0.18316410270251204,
+      "learning_rate": 0.00014092212040704336,
+      "loss": 1.0787,
+      "step": 2000
+    },
+    {
+      "epoch": 0.38498121148354125,
+      "grad_norm": 0.35663042410706985,
+      "learning_rate": 0.0001408652487514908,
+      "loss": 1.1214,
+      "step": 2001
+    },
+    {
+      "epoch": 0.3851736058920788,
+      "grad_norm": 0.32414486296588063,
+      "learning_rate": 0.0001408083612243465,
+      "loss": 1.0169,
+      "step": 2002
+    },
+    {
+      "epoch": 0.38536600030061624,
+      "grad_norm": 0.32898446387455277,
+      "learning_rate": 0.00014075145784770495,
+      "loss": 1.0604,
+      "step": 2003
+    },
+    {
+      "epoch": 0.38555839470915376,
+      "grad_norm": 0.32755318138444894,
+      "learning_rate": 0.00014069453864366677,
+      "loss": 1.0501,
+      "step": 2004
+    },
+    {
+      "epoch": 0.3857507891176913,
+      "grad_norm": 0.29762364756747844,
+      "learning_rate": 0.00014063760363433866,
+      "loss": 1.1317,
+      "step": 2005
+    },
+    {
+      "epoch": 0.38594318352622875,
+      "grad_norm": 0.2886755174274723,
+      "learning_rate": 0.00014058065284183358,
+      "loss": 1.0229,
+      "step": 2006
+    },
+    {
+      "epoch": 0.38613557793476627,
+      "grad_norm": 0.33066756772643474,
+      "learning_rate": 0.00014052368628827058,
+      "loss": 1.1267,
+      "step": 2007
+    },
+    {
+      "epoch": 0.3863279723433038,
+      "grad_norm": 0.5108689803097722,
+      "learning_rate": 0.00014046670399577477,
+      "loss": 1.1445,
+      "step": 2008
+    },
+    {
+      "epoch": 0.38652036675184126,
+      "grad_norm": 0.367712119078728,
+      "learning_rate": 0.0001404097059864774,
+      "loss": 1.0834,
+      "step": 2009
+    },
+    {
+      "epoch": 0.3867127611603788,
+      "grad_norm": 0.29666278316430783,
+      "learning_rate": 0.00014035269228251587,
+      "loss": 1.0715,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3869051555689163,
+      "grad_norm": 0.3410635961186643,
+      "learning_rate": 0.00014029566290603367,
+      "loss": 1.1426,
+      "step": 2011
+    },
+    {
+      "epoch": 0.38709754997745377,
+      "grad_norm": 0.3064763240114719,
+      "learning_rate": 0.00014023861787918032,
+      "loss": 1.0599,
+      "step": 2012
+    },
+    {
+      "epoch": 0.3872899443859913,
+      "grad_norm": 0.3203840538653993,
+      "learning_rate": 0.00014018155722411144,
+      "loss": 1.1161,
+      "step": 2013
+    },
+    {
+      "epoch": 0.38748233879452876,
+      "grad_norm": 0.3009858416073814,
+      "learning_rate": 0.00014012448096298874,
+      "loss": 1.0631,
+      "step": 2014
+    },
+    {
+      "epoch": 0.3876747332030663,
+      "grad_norm": 0.2715882261239675,
+      "learning_rate": 0.00014006738911798,
+      "loss": 1.1196,
+      "step": 2015
+    },
+    {
+      "epoch": 0.3878671276116038,
+      "grad_norm": 0.2632765294624663,
+      "learning_rate": 0.000140010281711259,
+      "loss": 1.0435,
+      "step": 2016
+    },
+    {
+      "epoch": 0.38805952202014127,
+      "grad_norm": 0.30174660263090725,
+      "learning_rate": 0.00013995315876500565,
+      "loss": 1.0679,
+      "step": 2017
+    },
+    {
+      "epoch": 0.3882519164286788,
+      "grad_norm": 0.2976900181919206,
+      "learning_rate": 0.0001398960203014058,
+      "loss": 1.1669,
+      "step": 2018
+    },
+    {
+      "epoch": 0.3884443108372163,
+      "grad_norm": 0.3002693812398982,
+      "learning_rate": 0.0001398388663426514,
+      "loss": 1.0552,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3886367052457538,
+      "grad_norm": 0.3357480216306192,
+      "learning_rate": 0.00013978169691094037,
+      "loss": 1.0776,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3888290996542913,
+      "grad_norm": 0.29927868197483065,
+      "learning_rate": 0.00013972451202847664,
+      "loss": 1.0676,
+      "step": 2021
+    },
+    {
+      "epoch": 0.3890214940628288,
+      "grad_norm": 0.3286864164580379,
+      "learning_rate": 0.00013966731171747023,
+      "loss": 0.992,
+      "step": 2022
+    },
+    {
+      "epoch": 0.3892138884713663,
+      "grad_norm": 0.3038832491213965,
+      "learning_rate": 0.000139610096000137,
+      "loss": 1.1172,
+      "step": 2023
+    },
+    {
+      "epoch": 0.3894062828799038,
+      "grad_norm": 0.2627361070246382,
+      "learning_rate": 0.00013955286489869895,
+      "loss": 1.132,
+      "step": 2024
+    },
+    {
+      "epoch": 0.38959867728844133,
+      "grad_norm": 0.3438216465287393,
+      "learning_rate": 0.0001394956184353839,
+      "loss": 1.1049,
+      "step": 2025
+    },
+    {
+      "epoch": 0.3897910716969788,
+      "grad_norm": 0.343644278698167,
+      "learning_rate": 0.00013943835663242577,
+      "loss": 1.0887,
+      "step": 2026
+    },
+    {
+      "epoch": 0.3899834661055163,
+      "grad_norm": 0.2617840041858717,
+      "learning_rate": 0.00013938107951206439,
+      "loss": 1.0097,
+      "step": 2027
+    },
+    {
+      "epoch": 0.3901758605140538,
+      "grad_norm": 0.2723666708594852,
+      "learning_rate": 0.00013932378709654547,
+      "loss": 1.0521,
+      "step": 2028
+    },
+    {
+      "epoch": 0.3903682549225913,
+      "grad_norm": 0.32178469621356853,
+      "learning_rate": 0.00013926647940812082,
+      "loss": 1.1396,
+      "step": 2029
+    },
+    {
+      "epoch": 0.39056064933112883,
+      "grad_norm": 0.3511065750811782,
+      "learning_rate": 0.000139209156469048,
+      "loss": 1.2104,
+      "step": 2030
+    },
+    {
+      "epoch": 0.3907530437396663,
+      "grad_norm": 0.3060315973089918,
+      "learning_rate": 0.00013915181830159061,
+      "loss": 1.1405,
+      "step": 2031
+    },
+    {
+      "epoch": 0.3909454381482038,
+      "grad_norm": 0.29461087808642006,
+      "learning_rate": 0.00013909446492801818,
+      "loss": 0.9589,
+      "step": 2032
+    },
+    {
+      "epoch": 0.39113783255674134,
+      "grad_norm": 0.28331014386052233,
+      "learning_rate": 0.00013903709637060605,
+      "loss": 1.1054,
+      "step": 2033
+    },
+    {
+      "epoch": 0.3913302269652788,
+      "grad_norm": 0.317799359866807,
+      "learning_rate": 0.00013897971265163544,
+      "loss": 1.0282,
+      "step": 2034
+    },
+    {
+      "epoch": 0.3915226213738163,
+      "grad_norm": 0.3500943680259453,
+      "learning_rate": 0.00013892231379339368,
+      "loss": 1.0858,
+      "step": 2035
+    },
+    {
+      "epoch": 0.39171501578235385,
+      "grad_norm": 0.3042963158307611,
+      "learning_rate": 0.00013886489981817374,
+      "loss": 1.0279,
+      "step": 2036
+    },
+    {
+      "epoch": 0.3919074101908913,
+      "grad_norm": 0.39582282683952036,
+      "learning_rate": 0.00013880747074827455,
+      "loss": 1.1546,
+      "step": 2037
+    },
+    {
+      "epoch": 0.39209980459942884,
+      "grad_norm": 0.29684025299805616,
+      "learning_rate": 0.00013875002660600086,
+      "loss": 1.1034,
+      "step": 2038
+    },
+    {
+      "epoch": 0.39229219900796636,
+      "grad_norm": 0.3360600449603767,
+      "learning_rate": 0.00013869256741366338,
+      "loss": 1.1364,
+      "step": 2039
+    },
+    {
+      "epoch": 0.3924845934165038,
+      "grad_norm": 0.2939770535631643,
+      "learning_rate": 0.00013863509319357858,
+      "loss": 1.0044,
+      "step": 2040
+    },
+    {
+      "epoch": 0.39267698782504135,
+      "grad_norm": 0.307410315423353,
+      "learning_rate": 0.00013857760396806875,
+      "loss": 1.0466,
+      "step": 2041
+    },
+    {
+      "epoch": 0.3928693822335788,
+      "grad_norm": 0.25265901953556863,
+      "learning_rate": 0.00013852009975946208,
+      "loss": 1.0157,
+      "step": 2042
+    },
+    {
+      "epoch": 0.39306177664211633,
+      "grad_norm": 0.24404275983052462,
+      "learning_rate": 0.00013846258059009253,
+      "loss": 1.066,
+      "step": 2043
+    },
+    {
+      "epoch": 0.39325417105065386,
+      "grad_norm": 0.3280647300136039,
+      "learning_rate": 0.0001384050464822999,
+      "loss": 1.101,
+      "step": 2044
+    },
+    {
+      "epoch": 0.3934465654591913,
+      "grad_norm": 0.3428250585144029,
+      "learning_rate": 0.0001383474974584297,
+      "loss": 1.0671,
+      "step": 2045
+    },
+    {
+      "epoch": 0.39363895986772884,
+      "grad_norm": 0.26512424373300364,
+      "learning_rate": 0.0001382899335408334,
+      "loss": 1.0308,
+      "step": 2046
+    },
+    {
+      "epoch": 0.39383135427626637,
+      "grad_norm": 0.4813952914573968,
+      "learning_rate": 0.00013823235475186814,
+      "loss": 1.0766,
+      "step": 2047
+    },
+    {
+      "epoch": 0.39402374868480383,
+      "grad_norm": 0.2894001090687884,
+      "learning_rate": 0.00013817476111389684,
+      "loss": 1.0691,
+      "step": 2048
+    },
+    {
+      "epoch": 0.39421614309334135,
+      "grad_norm": 0.3104660549042743,
+      "learning_rate": 0.00013811715264928825,
+      "loss": 1.1377,
+      "step": 2049
+    },
+    {
+      "epoch": 0.3944085375018789,
+      "grad_norm": 0.4111115117637331,
+      "learning_rate": 0.00013805952938041673,
+      "loss": 1.0415,
+      "step": 2050
+    },
+    {
+      "epoch": 0.39460093191041634,
+      "grad_norm": 0.3632355462202737,
+      "learning_rate": 0.00013800189132966257,
+      "loss": 1.1263,
+      "step": 2051
+    },
+    {
+      "epoch": 0.39479332631895386,
+      "grad_norm": 0.31626302678159074,
+      "learning_rate": 0.00013794423851941175,
+      "loss": 1.0975,
+      "step": 2052
+    },
+    {
+      "epoch": 0.39498572072749133,
+      "grad_norm": 0.2706800886709484,
+      "learning_rate": 0.00013788657097205592,
+      "loss": 1.1029,
+      "step": 2053
+    },
+    {
+      "epoch": 0.39517811513602885,
+      "grad_norm": 0.25725046537940016,
+      "learning_rate": 0.00013782888870999245,
+      "loss": 1.035,
+      "step": 2054
+    },
+    {
+      "epoch": 0.3953705095445664,
+      "grad_norm": 0.27834633660170705,
+      "learning_rate": 0.0001377711917556245,
+      "loss": 1.1839,
+      "step": 2055
+    },
+    {
+      "epoch": 0.39556290395310384,
+      "grad_norm": 0.42690148735716416,
+      "learning_rate": 0.00013771348013136097,
+      "loss": 1.1049,
+      "step": 2056
+    },
+    {
+      "epoch": 0.39575529836164136,
+      "grad_norm": 0.29549524956082585,
+      "learning_rate": 0.00013765575385961626,
+      "loss": 1.1342,
+      "step": 2057
+    },
+    {
+      "epoch": 0.3959476927701789,
+      "grad_norm": 0.24628504779021126,
+      "learning_rate": 0.0001375980129628107,
+      "loss": 1.1864,
+      "step": 2058
+    },
+    {
+      "epoch": 0.39614008717871635,
+      "grad_norm": 0.27842487613582945,
+      "learning_rate": 0.00013754025746337014,
+      "loss": 1.1022,
+      "step": 2059
+    },
+    {
+      "epoch": 0.39633248158725387,
+      "grad_norm": 0.28060256854335885,
+      "learning_rate": 0.00013748248738372615,
+      "loss": 1.0962,
+      "step": 2060
+    },
+    {
+      "epoch": 0.3965248759957914,
+      "grad_norm": 0.3190526824239674,
+      "learning_rate": 0.00013742470274631598,
+      "loss": 1.099,
+      "step": 2061
+    },
+    {
+      "epoch": 0.39671727040432886,
+      "grad_norm": 0.33726110717618146,
+      "learning_rate": 0.00013736690357358253,
+      "loss": 1.0868,
+      "step": 2062
+    },
+    {
+      "epoch": 0.3969096648128664,
+      "grad_norm": 0.3113877915706845,
+      "learning_rate": 0.00013730908988797427,
+      "loss": 1.0781,
+      "step": 2063
+    },
+    {
+      "epoch": 0.3971020592214039,
+      "grad_norm": 0.2738932258230844,
+      "learning_rate": 0.00013725126171194543,
+      "loss": 1.0563,
+      "step": 2064
+    },
+    {
+      "epoch": 0.39729445362994137,
+      "grad_norm": 0.3452939684872177,
+      "learning_rate": 0.0001371934190679558,
+      "loss": 1.1556,
+      "step": 2065
+    },
+    {
+      "epoch": 0.3974868480384789,
+      "grad_norm": 0.22981383382739284,
+      "learning_rate": 0.00013713556197847074,
+      "loss": 1.1705,
+      "step": 2066
+    },
+    {
+      "epoch": 0.39767924244701636,
+      "grad_norm": 0.26028591749442753,
+      "learning_rate": 0.00013707769046596136,
+      "loss": 0.9801,
+      "step": 2067
+    },
+    {
+      "epoch": 0.3978716368555539,
+      "grad_norm": 0.2918806737392978,
+      "learning_rate": 0.00013701980455290425,
+      "loss": 1.0513,
+      "step": 2068
+    },
+    {
+      "epoch": 0.3980640312640914,
+      "grad_norm": 0.3202449645307156,
+      "learning_rate": 0.00013696190426178162,
+      "loss": 1.2073,
+      "step": 2069
+    },
+    {
+      "epoch": 0.39825642567262887,
+      "grad_norm": 0.3810834203542579,
+      "learning_rate": 0.00013690398961508127,
+      "loss": 1.0588,
+      "step": 2070
+    },
+    {
+      "epoch": 0.3984488200811664,
+      "grad_norm": 0.27711890794106275,
+      "learning_rate": 0.00013684606063529663,
+      "loss": 1.1198,
+      "step": 2071
+    },
+    {
+      "epoch": 0.3986412144897039,
+      "grad_norm": 0.3212684362809215,
+      "learning_rate": 0.00013678811734492657,
+      "loss": 0.9831,
+      "step": 2072
+    },
+    {
+      "epoch": 0.3988336088982414,
+      "grad_norm": 0.30844321733987484,
+      "learning_rate": 0.00013673015976647568,
+      "loss": 1.1314,
+      "step": 2073
+    },
+    {
+      "epoch": 0.3990260033067789,
+      "grad_norm": 0.3166019414123627,
+      "learning_rate": 0.000136672187922454,
+      "loss": 1.0714,
+      "step": 2074
+    },
+    {
+      "epoch": 0.3992183977153164,
+      "grad_norm": 0.38948413266360854,
+      "learning_rate": 0.00013661420183537704,
+      "loss": 1.0322,
+      "step": 2075
+    },
+    {
+      "epoch": 0.3994107921238539,
+      "grad_norm": 0.3153684778434652,
+      "learning_rate": 0.00013655620152776604,
+      "loss": 1.1494,
+      "step": 2076
+    },
+    {
+      "epoch": 0.3996031865323914,
+      "grad_norm": 0.26359494773756625,
+      "learning_rate": 0.0001364981870221476,
+      "loss": 1.1031,
+      "step": 2077
+    },
+    {
+      "epoch": 0.39979558094092893,
+      "grad_norm": 0.2615007390625749,
+      "learning_rate": 0.00013644015834105388,
+      "loss": 1.1718,
+      "step": 2078
+    },
+    {
+      "epoch": 0.3999879753494664,
+      "grad_norm": 0.2770276163713552,
+      "learning_rate": 0.00013638211550702256,
+      "loss": 1.08,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4001803697580039,
+      "grad_norm": 0.2965165584442199,
+      "learning_rate": 0.0001363240585425968,
+      "loss": 1.0811,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4003727641665414,
+      "grad_norm": 0.3613295323674013,
+      "learning_rate": 0.0001362659874703253,
+      "loss": 1.0728,
+      "step": 2081
+    },
+    {
+      "epoch": 0.4005651585750789,
+      "grad_norm": 0.2935570637582771,
+      "learning_rate": 0.00013620790231276214,
+      "loss": 0.9881,
+      "step": 2082
+    },
+    {
+      "epoch": 0.4007575529836164,
+      "grad_norm": 0.32688553000468895,
+      "learning_rate": 0.00013614980309246692,
+      "loss": 1.0613,
+      "step": 2083
+    },
+    {
+      "epoch": 0.4009499473921539,
+      "grad_norm": 0.36265883972935936,
+      "learning_rate": 0.00013609168983200472,
+      "loss": 1.1742,
+      "step": 2084
+    },
+    {
+      "epoch": 0.4011423418006914,
+      "grad_norm": 0.3485944680988614,
+      "learning_rate": 0.00013603356255394614,
+      "loss": 1.1153,
+      "step": 2085
+    },
+    {
+      "epoch": 0.40133473620922894,
+      "grad_norm": 0.28433976686267926,
+      "learning_rate": 0.00013597542128086703,
+      "loss": 1.1619,
+      "step": 2086
+    },
+    {
+      "epoch": 0.4015271306177664,
+      "grad_norm": 0.2900059030464672,
+      "learning_rate": 0.00013591726603534885,
+      "loss": 1.0274,
+      "step": 2087
+    },
+    {
+      "epoch": 0.4017195250263039,
+      "grad_norm": 0.2886868131417094,
+      "learning_rate": 0.00013585909683997842,
+      "loss": 0.9886,
+      "step": 2088
+    },
+    {
+      "epoch": 0.40191191943484145,
+      "grad_norm": 0.3145366087777109,
+      "learning_rate": 0.00013580091371734798,
+      "loss": 1.18,
+      "step": 2089
+    },
+    {
+      "epoch": 0.4021043138433789,
+      "grad_norm": 0.29350833172164986,
+      "learning_rate": 0.00013574271669005518,
+      "loss": 1.0232,
+      "step": 2090
+    },
+    {
+      "epoch": 0.40229670825191644,
+      "grad_norm": 0.31273848002720694,
+      "learning_rate": 0.00013568450578070308,
+      "loss": 1.0609,
+      "step": 2091
+    },
+    {
+      "epoch": 0.40248910266045396,
+      "grad_norm": 0.3180627820275433,
+      "learning_rate": 0.00013562628101190013,
+      "loss": 1.1313,
+      "step": 2092
+    },
+    {
+      "epoch": 0.4026814970689914,
+      "grad_norm": 0.3302072039902513,
+      "learning_rate": 0.00013556804240626017,
+      "loss": 1.0451,
+      "step": 2093
+    },
+    {
+      "epoch": 0.40287389147752894,
+      "grad_norm": 0.30975872280331496,
+      "learning_rate": 0.0001355097899864024,
+      "loss": 1.1584,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4030662858860664,
+      "grad_norm": 0.3376188769221276,
+      "learning_rate": 0.00013545152377495136,
+      "loss": 1.0282,
+      "step": 2095
+    },
+    {
+      "epoch": 0.40325868029460393,
+      "grad_norm": 0.35392391874138956,
+      "learning_rate": 0.00013539324379453698,
+      "loss": 1.1646,
+      "step": 2096
+    },
+    {
+      "epoch": 0.40345107470314145,
+      "grad_norm": 0.34748236877836686,
+      "learning_rate": 0.00013533495006779454,
+      "loss": 1.1262,
+      "step": 2097
+    },
+    {
+      "epoch": 0.4036434691116789,
+      "grad_norm": 0.3661899492574811,
+      "learning_rate": 0.00013527664261736471,
+      "loss": 1.102,
+      "step": 2098
+    },
+    {
+      "epoch": 0.40383586352021644,
+      "grad_norm": 0.32911861384737695,
+      "learning_rate": 0.00013521832146589334,
+      "loss": 1.1295,
+      "step": 2099
+    },
+    {
+      "epoch": 0.40402825792875396,
+      "grad_norm": 0.3158394153870188,
+      "learning_rate": 0.00013515998663603173,
+      "loss": 1.0142,
+      "step": 2100
+    },
+    {
+      "epoch": 0.40422065233729143,
+      "grad_norm": 0.3357591318688121,
+      "learning_rate": 0.00013510163815043646,
+      "loss": 1.0144,
+      "step": 2101
+    },
+    {
+      "epoch": 0.40441304674582895,
+      "grad_norm": 0.30856652318178,
+      "learning_rate": 0.00013504327603176944,
+      "loss": 1.0329,
+      "step": 2102
+    },
+    {
+      "epoch": 0.4046054411543665,
+      "grad_norm": 0.323932369333457,
+      "learning_rate": 0.0001349849003026978,
+      "loss": 1.0971,
+      "step": 2103
+    },
+    {
+      "epoch": 0.40479783556290394,
+      "grad_norm": 0.28865582851014965,
+      "learning_rate": 0.00013492651098589398,
+      "loss": 1.0984,
+      "step": 2104
+    },
+    {
+      "epoch": 0.40499022997144146,
+      "grad_norm": 0.25477100770261774,
+      "learning_rate": 0.0001348681081040358,
+      "loss": 1.1441,
+      "step": 2105
+    },
+    {
+      "epoch": 0.405182624379979,
+      "grad_norm": 0.32261191611519663,
+      "learning_rate": 0.0001348096916798062,
+      "loss": 1.0235,
+      "step": 2106
+    },
+    {
+      "epoch": 0.40537501878851645,
+      "grad_norm": 0.3260341592407033,
+      "learning_rate": 0.00013475126173589344,
+      "loss": 1.0906,
+      "step": 2107
+    },
+    {
+      "epoch": 0.40556741319705397,
+      "grad_norm": 0.4192487630739309,
+      "learning_rate": 0.00013469281829499107,
+      "loss": 1.1488,
+      "step": 2108
+    },
+    {
+      "epoch": 0.40575980760559144,
+      "grad_norm": 0.30952295609392133,
+      "learning_rate": 0.00013463436137979787,
+      "loss": 0.9873,
+      "step": 2109
+    },
+    {
+      "epoch": 0.40595220201412896,
+      "grad_norm": 0.3506024579995169,
+      "learning_rate": 0.00013457589101301774,
+      "loss": 1.1082,
+      "step": 2110
+    },
+    {
+      "epoch": 0.4061445964226665,
+      "grad_norm": 0.3095130709917231,
+      "learning_rate": 0.00013451740721736002,
+      "loss": 1.0673,
+      "step": 2111
+    },
+    {
+      "epoch": 0.40633699083120395,
+      "grad_norm": 0.3069659735134276,
+      "learning_rate": 0.00013445891001553906,
+      "loss": 1.1357,
+      "step": 2112
+    },
+    {
+      "epoch": 0.40652938523974147,
+      "grad_norm": 0.35186699354292944,
+      "learning_rate": 0.00013440039943027451,
+      "loss": 1.0993,
+      "step": 2113
+    },
+    {
+      "epoch": 0.406721779648279,
+      "grad_norm": 0.28273978795415033,
+      "learning_rate": 0.00013434187548429124,
+      "loss": 1.0602,
+      "step": 2114
+    },
+    {
+      "epoch": 0.40691417405681646,
+      "grad_norm": 0.33752178931361904,
+      "learning_rate": 0.00013428333820031922,
+      "loss": 1.0236,
+      "step": 2115
+    },
+    {
+      "epoch": 0.407106568465354,
+      "grad_norm": 0.32474596242932924,
+      "learning_rate": 0.00013422478760109373,
+      "loss": 1.0567,
+      "step": 2116
+    },
+    {
+      "epoch": 0.4072989628738915,
+      "grad_norm": 0.2654352102177828,
+      "learning_rate": 0.00013416622370935507,
+      "loss": 1.0653,
+      "step": 2117
+    },
+    {
+      "epoch": 0.40749135728242897,
+      "grad_norm": 0.3202465990290772,
+      "learning_rate": 0.00013410764654784885,
+      "loss": 1.0901,
+      "step": 2118
+    },
+    {
+      "epoch": 0.4076837516909665,
+      "grad_norm": 0.26980647741216307,
+      "learning_rate": 0.00013404905613932574,
+      "loss": 1.0907,
+      "step": 2119
+    },
+    {
+      "epoch": 0.40787614609950396,
+      "grad_norm": 0.2576526288710288,
+      "learning_rate": 0.00013399045250654152,
+      "loss": 1.056,
+      "step": 2120
+    },
+    {
+      "epoch": 0.4080685405080415,
+      "grad_norm": 0.3072191706241032,
+      "learning_rate": 0.00013393183567225725,
+      "loss": 1.1635,
+      "step": 2121
+    },
+    {
+      "epoch": 0.408260934916579,
+      "grad_norm": 0.2840373706199005,
+      "learning_rate": 0.000133873205659239,
+      "loss": 1.0244,
+      "step": 2122
+    },
+    {
+      "epoch": 0.40845332932511647,
+      "grad_norm": 0.3330726525517525,
+      "learning_rate": 0.000133814562490258,
+      "loss": 1.0733,
+      "step": 2123
+    },
+    {
+      "epoch": 0.408645723733654,
+      "grad_norm": 0.27620918895514857,
+      "learning_rate": 0.00013375590618809054,
+      "loss": 0.9979,
+      "step": 2124
+    },
+    {
+      "epoch": 0.4088381181421915,
+      "grad_norm": 0.30359819970538376,
+      "learning_rate": 0.00013369723677551814,
+      "loss": 1.1688,
+      "step": 2125
+    },
+    {
+      "epoch": 0.409030512550729,
+      "grad_norm": 0.3656900264728042,
+      "learning_rate": 0.00013363855427532724,
+      "loss": 1.0821,
+      "step": 2126
+    },
+    {
+      "epoch": 0.4092229069592665,
+      "grad_norm": 0.25745020351525233,
+      "learning_rate": 0.00013357985871030947,
+      "loss": 1.1077,
+      "step": 2127
+    },
+    {
+      "epoch": 0.409415301367804,
+      "grad_norm": 0.3120952803249164,
+      "learning_rate": 0.00013352115010326155,
+      "loss": 0.9936,
+      "step": 2128
+    },
+    {
+      "epoch": 0.4096076957763415,
+      "grad_norm": 0.245152501357541,
+      "learning_rate": 0.00013346242847698517,
+      "loss": 0.9881,
+      "step": 2129
+    },
+    {
+      "epoch": 0.409800090184879,
+      "grad_norm": 0.38143067763378435,
+      "learning_rate": 0.00013340369385428713,
+      "loss": 1.0871,
+      "step": 2130
+    },
+    {
+      "epoch": 0.40999248459341653,
+      "grad_norm": 0.2904231850863169,
+      "learning_rate": 0.00013334494625797936,
+      "loss": 1.0056,
+      "step": 2131
+    },
+    {
+      "epoch": 0.410184879001954,
+      "grad_norm": 0.31836978508375513,
+      "learning_rate": 0.00013328618571087868,
+      "loss": 1.088,
+      "step": 2132
+    },
+    {
+      "epoch": 0.4103772734104915,
+      "grad_norm": 0.29390994405276855,
+      "learning_rate": 0.000133227412235807,
+      "loss": 1.0907,
+      "step": 2133
+    },
+    {
+      "epoch": 0.410569667819029,
+      "grad_norm": 0.31321200852450837,
+      "learning_rate": 0.0001331686258555913,
+      "loss": 1.1064,
+      "step": 2134
+    },
+    {
+      "epoch": 0.4107620622275665,
+      "grad_norm": 0.3157078732535461,
+      "learning_rate": 0.00013310982659306352,
+      "loss": 1.1283,
+      "step": 2135
+    },
+    {
+      "epoch": 0.410954456636104,
+      "grad_norm": 0.3155262313137496,
+      "learning_rate": 0.00013305101447106063,
+      "loss": 1.152,
+      "step": 2136
+    },
+    {
+      "epoch": 0.4111468510446415,
+      "grad_norm": 0.2691636910706811,
+      "learning_rate": 0.00013299218951242457,
+      "loss": 1.0996,
+      "step": 2137
+    },
+    {
+      "epoch": 0.411339245453179,
+      "grad_norm": 0.2629553413783702,
+      "learning_rate": 0.00013293335174000226,
+      "loss": 1.0581,
+      "step": 2138
+    },
+    {
+      "epoch": 0.41153163986171654,
+      "grad_norm": 0.30474586519998226,
+      "learning_rate": 0.0001328745011766456,
+      "loss": 1.1373,
+      "step": 2139
+    },
+    {
+      "epoch": 0.411724034270254,
+      "grad_norm": 0.2756118623251005,
+      "learning_rate": 0.00013281563784521154,
+      "loss": 1.0728,
+      "step": 2140
+    },
+    {
+      "epoch": 0.4119164286787915,
+      "grad_norm": 0.29252663895150366,
+      "learning_rate": 0.00013275676176856185,
+      "loss": 1.0617,
+      "step": 2141
+    },
+    {
+      "epoch": 0.41210882308732905,
+      "grad_norm": 0.3100632327702502,
+      "learning_rate": 0.00013269787296956334,
+      "loss": 1.0852,
+      "step": 2142
+    },
+    {
+      "epoch": 0.4123012174958665,
+      "grad_norm": 0.3108105143767478,
+      "learning_rate": 0.00013263897147108776,
+      "loss": 1.0525,
+      "step": 2143
+    },
+    {
+      "epoch": 0.41249361190440403,
+      "grad_norm": 0.30276024043743577,
+      "learning_rate": 0.00013258005729601177,
+      "loss": 1.0607,
+      "step": 2144
+    },
+    {
+      "epoch": 0.41268600631294156,
+      "grad_norm": 0.2579126470279425,
+      "learning_rate": 0.00013252113046721692,
+      "loss": 1.0512,
+      "step": 2145
+    },
+    {
+      "epoch": 0.412878400721479,
+      "grad_norm": 0.30291625598100125,
+      "learning_rate": 0.00013246219100758973,
+      "loss": 1.1413,
+      "step": 2146
+    },
+    {
+      "epoch": 0.41307079513001654,
+      "grad_norm": 0.2533645600740764,
+      "learning_rate": 0.00013240323894002167,
+      "loss": 1.0573,
+      "step": 2147
+    },
+    {
+      "epoch": 0.413263189538554,
+      "grad_norm": 0.27270643604806544,
+      "learning_rate": 0.00013234427428740895,
+      "loss": 1.1247,
+      "step": 2148
+    },
+    {
+      "epoch": 0.41345558394709153,
+      "grad_norm": 0.34434617161862013,
+      "learning_rate": 0.00013228529707265277,
+      "loss": 1.1597,
+      "step": 2149
+    },
+    {
+      "epoch": 0.41364797835562905,
+      "grad_norm": 0.3370993543549416,
+      "learning_rate": 0.00013222630731865928,
+      "loss": 1.0698,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4138403727641665,
+      "grad_norm": 0.2915587633582487,
+      "learning_rate": 0.00013216730504833935,
+      "loss": 1.1213,
+      "step": 2151
+    },
+    {
+      "epoch": 0.41403276717270404,
+      "grad_norm": 0.2888318846028617,
+      "learning_rate": 0.00013210829028460883,
+      "loss": 0.9937,
+      "step": 2152
+    },
+    {
+      "epoch": 0.41422516158124156,
+      "grad_norm": 0.2933541912930708,
+      "learning_rate": 0.00013204926305038832,
+      "loss": 1.0192,
+      "step": 2153
+    },
+    {
+      "epoch": 0.41441755598977903,
+      "grad_norm": 0.26981190053274784,
+      "learning_rate": 0.00013199022336860333,
+      "loss": 1.0394,
+      "step": 2154
+    },
+    {
+      "epoch": 0.41460995039831655,
+      "grad_norm": 0.30895432972898323,
+      "learning_rate": 0.00013193117126218424,
+      "loss": 1.0329,
+      "step": 2155
+    },
+    {
+      "epoch": 0.4148023448068541,
+      "grad_norm": 0.39618929841086603,
+      "learning_rate": 0.00013187210675406617,
+      "loss": 1.0169,
+      "step": 2156
+    },
+    {
+      "epoch": 0.41499473921539154,
+      "grad_norm": 0.33622540211306556,
+      "learning_rate": 0.0001318130298671891,
+      "loss": 1.0878,
+      "step": 2157
+    },
+    {
+      "epoch": 0.41518713362392906,
+      "grad_norm": 0.3489650635266243,
+      "learning_rate": 0.00013175394062449778,
+      "loss": 1.0964,
+      "step": 2158
+    },
+    {
+      "epoch": 0.4153795280324666,
+      "grad_norm": 0.3579289855885925,
+      "learning_rate": 0.00013169483904894183,
+      "loss": 1.1028,
+      "step": 2159
+    },
+    {
+      "epoch": 0.41557192244100405,
+      "grad_norm": 0.2952716836493573,
+      "learning_rate": 0.00013163572516347564,
+      "loss": 1.0469,
+      "step": 2160
+    },
+    {
+      "epoch": 0.41576431684954157,
+      "grad_norm": 0.2959757813840895,
+      "learning_rate": 0.00013157659899105833,
+      "loss": 1.0566,
+      "step": 2161
+    },
+    {
+      "epoch": 0.41595671125807904,
+      "grad_norm": 0.2896443710761163,
+      "learning_rate": 0.0001315174605546538,
+      "loss": 1.0327,
+      "step": 2162
+    },
+    {
+      "epoch": 0.41614910566661656,
+      "grad_norm": 0.3863883469690799,
+      "learning_rate": 0.0001314583098772308,
+      "loss": 1.0188,
+      "step": 2163
+    },
+    {
+      "epoch": 0.4163415000751541,
+      "grad_norm": 0.36560905326018345,
+      "learning_rate": 0.00013139914698176273,
+      "loss": 1.1049,
+      "step": 2164
+    },
+    {
+      "epoch": 0.41653389448369155,
+      "grad_norm": 0.30308140765113695,
+      "learning_rate": 0.00013133997189122777,
+      "loss": 1.1425,
+      "step": 2165
+    },
+    {
+      "epoch": 0.41672628889222907,
+      "grad_norm": 0.294607906034493,
+      "learning_rate": 0.00013128078462860886,
+      "loss": 1.0192,
+      "step": 2166
+    },
+    {
+      "epoch": 0.4169186833007666,
+      "grad_norm": 0.32690085651940637,
+      "learning_rate": 0.00013122158521689369,
+      "loss": 1.0381,
+      "step": 2167
+    },
+    {
+      "epoch": 0.41711107770930406,
+      "grad_norm": 0.3327760192353365,
+      "learning_rate": 0.00013116237367907454,
+      "loss": 1.1548,
+      "step": 2168
+    },
+    {
+      "epoch": 0.4173034721178416,
+      "grad_norm": 0.25986497352220245,
+      "learning_rate": 0.00013110315003814855,
+      "loss": 0.9944,
+      "step": 2169
+    },
+    {
+      "epoch": 0.4174958665263791,
+      "grad_norm": 0.27472485431242855,
+      "learning_rate": 0.00013104391431711747,
+      "loss": 1.0126,
+      "step": 2170
+    },
+    {
+      "epoch": 0.41768826093491657,
+      "grad_norm": 0.33484886386237683,
+      "learning_rate": 0.0001309846665389878,
+      "loss": 1.1238,
+      "step": 2171
+    },
+    {
+      "epoch": 0.4178806553434541,
+      "grad_norm": 0.34784220257763554,
+      "learning_rate": 0.0001309254067267707,
+      "loss": 1.0242,
+      "step": 2172
+    },
+    {
+      "epoch": 0.4180730497519916,
+      "grad_norm": 0.39300942378334563,
+      "learning_rate": 0.00013086613490348198,
+      "loss": 1.05,
+      "step": 2173
+    },
+    {
+      "epoch": 0.4182654441605291,
+      "grad_norm": 0.28142923000143394,
+      "learning_rate": 0.00013080685109214207,
+      "loss": 1.0786,
+      "step": 2174
+    },
+    {
+      "epoch": 0.4184578385690666,
+      "grad_norm": 0.2798105651501539,
+      "learning_rate": 0.00013074755531577626,
+      "loss": 1.0239,
+      "step": 2175
+    },
+    {
+      "epoch": 0.41865023297760406,
+      "grad_norm": 0.3211164673409022,
+      "learning_rate": 0.00013068824759741426,
+      "loss": 1.0228,
+      "step": 2176
+    },
+    {
+      "epoch": 0.4188426273861416,
+      "grad_norm": 0.3178939655693238,
+      "learning_rate": 0.0001306289279600905,
+      "loss": 1.0643,
+      "step": 2177
+    },
+    {
+      "epoch": 0.4190350217946791,
+      "grad_norm": 0.28732793536172013,
+      "learning_rate": 0.00013056959642684403,
+      "loss": 1.072,
+      "step": 2178
+    },
+    {
+      "epoch": 0.4192274162032166,
+      "grad_norm": 0.27632839963144906,
+      "learning_rate": 0.0001305102530207186,
+      "loss": 1.0273,
+      "step": 2179
+    },
+    {
+      "epoch": 0.4194198106117541,
+      "grad_norm": 0.27648238248740514,
+      "learning_rate": 0.00013045089776476246,
+      "loss": 1.038,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4196122050202916,
+      "grad_norm": 0.32175572376340095,
+      "learning_rate": 0.0001303915306820285,
+      "loss": 1.1307,
+      "step": 2181
+    },
+    {
+      "epoch": 0.4198045994288291,
+      "grad_norm": 0.2710366545829388,
+      "learning_rate": 0.00013033215179557422,
+      "loss": 1.0766,
+      "step": 2182
+    },
+    {
+      "epoch": 0.4199969938373666,
+      "grad_norm": 0.3124927983215582,
+      "learning_rate": 0.0001302727611284617,
+      "loss": 1.0717,
+      "step": 2183
+    },
+    {
+      "epoch": 0.4201893882459041,
+      "grad_norm": 0.31881710467528623,
+      "learning_rate": 0.00013021335870375765,
+      "loss": 1.036,
+      "step": 2184
+    },
+    {
+      "epoch": 0.4203817826544416,
+      "grad_norm": 0.26895695690636184,
+      "learning_rate": 0.00013015394454453315,
+      "loss": 1.0754,
+      "step": 2185
+    },
+    {
+      "epoch": 0.4205741770629791,
+      "grad_norm": 0.3077667451973139,
+      "learning_rate": 0.00013009451867386412,
+      "loss": 1.0987,
+      "step": 2186
+    },
+    {
+      "epoch": 0.4207665714715166,
+      "grad_norm": 0.3970956054123128,
+      "learning_rate": 0.00013003508111483075,
+      "loss": 1.0918,
+      "step": 2187
+    },
+    {
+      "epoch": 0.4209589658800541,
+      "grad_norm": 0.2570532522133091,
+      "learning_rate": 0.000129975631890518,
+      "loss": 1.1186,
+      "step": 2188
+    },
+    {
+      "epoch": 0.4211513602885916,
+      "grad_norm": 0.2918391916324925,
+      "learning_rate": 0.00012991617102401523,
+      "loss": 1.1065,
+      "step": 2189
+    },
+    {
+      "epoch": 0.4213437546971291,
+      "grad_norm": 0.2923584323407831,
+      "learning_rate": 0.00012985669853841634,
+      "loss": 1.1537,
+      "step": 2190
+    },
+    {
+      "epoch": 0.4215361491056666,
+      "grad_norm": 0.34178013657291517,
+      "learning_rate": 0.0001297972144568198,
+      "loss": 1.0527,
+      "step": 2191
+    },
+    {
+      "epoch": 0.42172854351420413,
+      "grad_norm": 0.3430974974207724,
+      "learning_rate": 0.00012973771880232853,
+      "loss": 1.0659,
+      "step": 2192
+    },
+    {
+      "epoch": 0.4219209379227416,
+      "grad_norm": 0.29206749455891645,
+      "learning_rate": 0.00012967821159804994,
+      "loss": 1.1654,
+      "step": 2193
+    },
+    {
+      "epoch": 0.4221133323312791,
+      "grad_norm": 0.37926292406973056,
+      "learning_rate": 0.00012961869286709594,
+      "loss": 1.0792,
+      "step": 2194
+    },
+    {
+      "epoch": 0.42230572673981664,
+      "grad_norm": 0.3137026053850834,
+      "learning_rate": 0.0001295591626325829,
+      "loss": 1.0235,
+      "step": 2195
+    },
+    {
+      "epoch": 0.4224981211483541,
+      "grad_norm": 0.3351652435566971,
+      "learning_rate": 0.00012949962091763175,
+      "loss": 1.045,
+      "step": 2196
+    },
+    {
+      "epoch": 0.42269051555689163,
+      "grad_norm": 0.31202787933585263,
+      "learning_rate": 0.00012944006774536773,
+      "loss": 1.0712,
+      "step": 2197
+    },
+    {
+      "epoch": 0.42288290996542915,
+      "grad_norm": 0.30090767954904274,
+      "learning_rate": 0.00012938050313892062,
+      "loss": 1.0942,
+      "step": 2198
+    },
+    {
+      "epoch": 0.4230753043739666,
+      "grad_norm": 0.2947107674452307,
+      "learning_rate": 0.00012932092712142468,
+      "loss": 1.0435,
+      "step": 2199
+    },
+    {
+      "epoch": 0.42326769878250414,
+      "grad_norm": 0.3211457650056529,
+      "learning_rate": 0.00012926133971601848,
+      "loss": 1.0717,
+      "step": 2200
+    },
+    {
+      "epoch": 0.4234600931910416,
+      "grad_norm": 0.3864264924279052,
+      "learning_rate": 0.00012920174094584514,
+      "loss": 1.0121,
+      "step": 2201
+    },
+    {
+      "epoch": 0.42365248759957913,
+      "grad_norm": 0.2881779388433144,
+      "learning_rate": 0.00012914213083405212,
+      "loss": 1.0607,
+      "step": 2202
+    },
+    {
+      "epoch": 0.42384488200811665,
+      "grad_norm": 0.3191534191694544,
+      "learning_rate": 0.00012908250940379124,
+      "loss": 1.0203,
+      "step": 2203
+    },
+    {
+      "epoch": 0.4240372764166541,
+      "grad_norm": 0.2507204338262169,
+      "learning_rate": 0.00012902287667821883,
+      "loss": 1.0637,
+      "step": 2204
+    },
+    {
+      "epoch": 0.42422967082519164,
+      "grad_norm": 0.3312245656267057,
+      "learning_rate": 0.0001289632326804956,
+      "loss": 1.0996,
+      "step": 2205
+    },
+    {
+      "epoch": 0.42442206523372916,
+      "grad_norm": 0.47611959534223114,
+      "learning_rate": 0.0001289035774337865,
+      "loss": 1.1233,
+      "step": 2206
+    },
+    {
+      "epoch": 0.42461445964226663,
+      "grad_norm": 0.3415957553888742,
+      "learning_rate": 0.00012884391096126097,
+      "loss": 1.1295,
+      "step": 2207
+    },
+    {
+      "epoch": 0.42480685405080415,
+      "grad_norm": 0.2644265852993455,
+      "learning_rate": 0.00012878423328609281,
+      "loss": 1.1342,
+      "step": 2208
+    },
+    {
+      "epoch": 0.42499924845934167,
+      "grad_norm": 0.25000667995646575,
+      "learning_rate": 0.00012872454443146013,
+      "loss": 1.0312,
+      "step": 2209
+    },
+    {
+      "epoch": 0.42519164286787914,
+      "grad_norm": 0.2782220891682969,
+      "learning_rate": 0.00012866484442054538,
+      "loss": 1.2075,
+      "step": 2210
+    },
+    {
+      "epoch": 0.42538403727641666,
+      "grad_norm": 0.29977025592498524,
+      "learning_rate": 0.00012860513327653535,
+      "loss": 1.1097,
+      "step": 2211
+    },
+    {
+      "epoch": 0.4255764316849542,
+      "grad_norm": 0.2785145450124637,
+      "learning_rate": 0.0001285454110226212,
+      "loss": 1.0123,
+      "step": 2212
+    },
+    {
+      "epoch": 0.42576882609349165,
+      "grad_norm": 0.2928195632179793,
+      "learning_rate": 0.00012848567768199831,
+      "loss": 1.099,
+      "step": 2213
+    },
+    {
+      "epoch": 0.42596122050202917,
+      "grad_norm": 0.2753655286622883,
+      "learning_rate": 0.00012842593327786646,
+      "loss": 1.0469,
+      "step": 2214
+    },
+    {
+      "epoch": 0.42615361491056664,
+      "grad_norm": 0.29230091117996665,
+      "learning_rate": 0.0001283661778334297,
+      "loss": 1.0303,
+      "step": 2215
+    },
+    {
+      "epoch": 0.42634600931910416,
+      "grad_norm": 0.34605414720153976,
+      "learning_rate": 0.00012830641137189628,
+      "loss": 1.0997,
+      "step": 2216
+    },
+    {
+      "epoch": 0.4265384037276417,
+      "grad_norm": 0.33858539393511694,
+      "learning_rate": 0.0001282466339164789,
+      "loss": 1.0509,
+      "step": 2217
+    },
+    {
+      "epoch": 0.42673079813617915,
+      "grad_norm": 0.28442241589571476,
+      "learning_rate": 0.00012818684549039438,
+      "loss": 0.9711,
+      "step": 2218
+    },
+    {
+      "epoch": 0.42692319254471667,
+      "grad_norm": 0.36171167729167664,
+      "learning_rate": 0.00012812704611686387,
+      "loss": 1.092,
+      "step": 2219
+    },
+    {
+      "epoch": 0.4271155869532542,
+      "grad_norm": 0.27227430750327375,
+      "learning_rate": 0.00012806723581911274,
+      "loss": 1.0676,
+      "step": 2220
+    },
+    {
+      "epoch": 0.42730798136179166,
+      "grad_norm": 0.2895761499410537,
+      "learning_rate": 0.00012800741462037064,
+      "loss": 1.0051,
+      "step": 2221
+    },
+    {
+      "epoch": 0.4275003757703292,
+      "grad_norm": 0.3211312693193202,
+      "learning_rate": 0.00012794758254387146,
+      "loss": 1.1371,
+      "step": 2222
+    },
+    {
+      "epoch": 0.4276927701788667,
+      "grad_norm": 0.36817870262920066,
+      "learning_rate": 0.00012788773961285323,
+      "loss": 1.1009,
+      "step": 2223
+    },
+    {
+      "epoch": 0.42788516458740417,
+      "grad_norm": 0.3125801670143624,
+      "learning_rate": 0.00012782788585055828,
+      "loss": 1.1223,
+      "step": 2224
+    },
+    {
+      "epoch": 0.4280775589959417,
+      "grad_norm": 0.40226080161062644,
+      "learning_rate": 0.00012776802128023316,
+      "loss": 1.0634,
+      "step": 2225
+    },
+    {
+      "epoch": 0.4282699534044792,
+      "grad_norm": 0.31390023413359386,
+      "learning_rate": 0.0001277081459251285,
+      "loss": 1.0165,
+      "step": 2226
+    },
+    {
+      "epoch": 0.4284623478130167,
+      "grad_norm": 0.28459209722056183,
+      "learning_rate": 0.0001276482598084993,
+      "loss": 1.0337,
+      "step": 2227
+    },
+    {
+      "epoch": 0.4286547422215542,
+      "grad_norm": 0.27392770914379844,
+      "learning_rate": 0.00012758836295360455,
+      "loss": 1.0162,
+      "step": 2228
+    },
+    {
+      "epoch": 0.42884713663009166,
+      "grad_norm": 0.25426950612596344,
+      "learning_rate": 0.0001275284553837075,
+      "loss": 1.0915,
+      "step": 2229
+    },
+    {
+      "epoch": 0.4290395310386292,
+      "grad_norm": 0.3277360810777147,
+      "learning_rate": 0.00012746853712207568,
+      "loss": 1.0708,
+      "step": 2230
+    },
+    {
+      "epoch": 0.4292319254471667,
+      "grad_norm": 0.28727196793326465,
+      "learning_rate": 0.00012740860819198047,
+      "loss": 1.077,
+      "step": 2231
+    },
+    {
+      "epoch": 0.4294243198557042,
+      "grad_norm": 0.35653124835448086,
+      "learning_rate": 0.0001273486686166977,
+      "loss": 1.0594,
+      "step": 2232
+    },
+    {
+      "epoch": 0.4296167142642417,
+      "grad_norm": 0.323352696347872,
+      "learning_rate": 0.00012728871841950718,
+      "loss": 1.1271,
+      "step": 2233
+    },
+    {
+      "epoch": 0.4298091086727792,
+      "grad_norm": 0.28579166378157655,
+      "learning_rate": 0.00012722875762369288,
+      "loss": 1.0781,
+      "step": 2234
+    },
+    {
+      "epoch": 0.4300015030813167,
+      "grad_norm": 0.3066729640014907,
+      "learning_rate": 0.00012716878625254288,
+      "loss": 1.0445,
+      "step": 2235
+    },
+    {
+      "epoch": 0.4301938974898542,
+      "grad_norm": 0.27755961078088953,
+      "learning_rate": 0.00012710880432934932,
+      "loss": 1.1186,
+      "step": 2236
+    },
+    {
+      "epoch": 0.4303862918983917,
+      "grad_norm": 0.3035864349115272,
+      "learning_rate": 0.0001270488118774086,
+      "loss": 1.1121,
+      "step": 2237
+    },
+    {
+      "epoch": 0.4305786863069292,
+      "grad_norm": 0.3069508843058301,
+      "learning_rate": 0.000126988808920021,
+      "loss": 1.0509,
+      "step": 2238
+    },
+    {
+      "epoch": 0.4307710807154667,
+      "grad_norm": 0.32117458606384475,
+      "learning_rate": 0.000126928795480491,
+      "loss": 1.0658,
+      "step": 2239
+    },
+    {
+      "epoch": 0.43096347512400424,
+      "grad_norm": 0.27043276401306465,
+      "learning_rate": 0.00012686877158212713,
+      "loss": 1.1762,
+      "step": 2240
+    },
+    {
+      "epoch": 0.4311558695325417,
+      "grad_norm": 0.2657502421700463,
+      "learning_rate": 0.00012680873724824198,
+      "loss": 1.1385,
+      "step": 2241
+    },
+    {
+      "epoch": 0.4313482639410792,
+      "grad_norm": 0.3114313159268892,
+      "learning_rate": 0.00012674869250215223,
+      "loss": 1.0139,
+      "step": 2242
+    },
+    {
+      "epoch": 0.4315406583496167,
+      "grad_norm": 0.2555846734850437,
+      "learning_rate": 0.00012668863736717856,
+      "loss": 1.1475,
+      "step": 2243
+    },
+    {
+      "epoch": 0.4317330527581542,
+      "grad_norm": 0.46484680610905427,
+      "learning_rate": 0.00012662857186664559,
+      "loss": 1.1227,
+      "step": 2244
+    },
+    {
+      "epoch": 0.43192544716669173,
+      "grad_norm": 0.252117588429946,
+      "learning_rate": 0.0001265684960238822,
+      "loss": 1.0652,
+      "step": 2245
+    },
+    {
+      "epoch": 0.4321178415752292,
+      "grad_norm": 0.3076639643450356,
+      "learning_rate": 0.0001265084098622211,
+      "loss": 1.0734,
+      "step": 2246
+    },
+    {
+      "epoch": 0.4323102359837667,
+      "grad_norm": 0.3018468849588766,
+      "learning_rate": 0.00012644831340499906,
+      "loss": 0.9336,
+      "step": 2247
+    },
+    {
+      "epoch": 0.43250263039230424,
+      "grad_norm": 0.3231050162300759,
+      "learning_rate": 0.00012638820667555683,
+      "loss": 0.9482,
+      "step": 2248
+    },
+    {
+      "epoch": 0.4326950248008417,
+      "grad_norm": 0.3196641442863389,
+      "learning_rate": 0.00012632808969723926,
+      "loss": 1.0116,
+      "step": 2249
+    },
+    {
+      "epoch": 0.43288741920937923,
+      "grad_norm": 0.3421544817356094,
+      "learning_rate": 0.000126267962493395,
+      "loss": 1.0357,
+      "step": 2250
+    },
+    {
+      "epoch": 0.43307981361791675,
+      "grad_norm": 0.24498736165941662,
+      "learning_rate": 0.0001262078250873768,
+      "loss": 1.0457,
+      "step": 2251
+    },
+    {
+      "epoch": 0.4332722080264542,
+      "grad_norm": 0.32777409315260503,
+      "learning_rate": 0.00012614767750254128,
+      "loss": 1.03,
+      "step": 2252
+    },
+    {
+      "epoch": 0.43346460243499174,
+      "grad_norm": 0.3166004917987926,
+      "learning_rate": 0.00012608751976224915,
+      "loss": 1.1015,
+      "step": 2253
+    },
+    {
+      "epoch": 0.4336569968435292,
+      "grad_norm": 0.34767394458823764,
+      "learning_rate": 0.00012602735188986498,
+      "loss": 1.1033,
+      "step": 2254
+    },
+    {
+      "epoch": 0.43384939125206673,
+      "grad_norm": 0.3214150449244712,
+      "learning_rate": 0.0001259671739087572,
+      "loss": 1.0629,
+      "step": 2255
+    },
+    {
+      "epoch": 0.43404178566060425,
+      "grad_norm": 0.37944127412277756,
+      "learning_rate": 0.00012590698584229835,
+      "loss": 1.0673,
+      "step": 2256
+    },
+    {
+      "epoch": 0.4342341800691417,
+      "grad_norm": 0.3065780828408335,
+      "learning_rate": 0.00012584678771386466,
+      "loss": 1.0475,
+      "step": 2257
+    },
+    {
+      "epoch": 0.43442657447767924,
+      "grad_norm": 0.32283017123688185,
+      "learning_rate": 0.0001257865795468365,
+      "loss": 1.1555,
+      "step": 2258
+    },
+    {
+      "epoch": 0.43461896888621676,
+      "grad_norm": 0.24658988097817017,
+      "learning_rate": 0.000125726361364598,
+      "loss": 1.0975,
+      "step": 2259
+    },
+    {
+      "epoch": 0.4348113632947542,
+      "grad_norm": 0.31873477475517975,
+      "learning_rate": 0.00012566613319053713,
+      "loss": 1.0741,
+      "step": 2260
+    },
+    {
+      "epoch": 0.43500375770329175,
+      "grad_norm": 0.30204281269961986,
+      "learning_rate": 0.00012560589504804592,
+      "loss": 1.0209,
+      "step": 2261
+    },
+    {
+      "epoch": 0.43519615211182927,
+      "grad_norm": 0.3447200061536308,
+      "learning_rate": 0.00012554564696052011,
+      "loss": 1.1049,
+      "step": 2262
+    },
+    {
+      "epoch": 0.43538854652036674,
+      "grad_norm": 0.2833123678050803,
+      "learning_rate": 0.00012548538895135942,
+      "loss": 1.1173,
+      "step": 2263
+    },
+    {
+      "epoch": 0.43558094092890426,
+      "grad_norm": 0.31331049698206237,
+      "learning_rate": 0.00012542512104396728,
+      "loss": 0.992,
+      "step": 2264
+    },
+    {
+      "epoch": 0.4357733353374418,
+      "grad_norm": 0.2669592456599427,
+      "learning_rate": 0.00012536484326175113,
+      "loss": 1.0341,
+      "step": 2265
+    },
+    {
+      "epoch": 0.43596572974597925,
+      "grad_norm": 0.331217984626546,
+      "learning_rate": 0.00012530455562812214,
+      "loss": 0.9767,
+      "step": 2266
+    },
+    {
+      "epoch": 0.43615812415451677,
+      "grad_norm": 0.35354982838335797,
+      "learning_rate": 0.0001252442581664953,
+      "loss": 1.1244,
+      "step": 2267
+    },
+    {
+      "epoch": 0.43635051856305423,
+      "grad_norm": 0.2606668891950679,
+      "learning_rate": 0.0001251839509002895,
+      "loss": 1.1665,
+      "step": 2268
+    },
+    {
+      "epoch": 0.43654291297159176,
+      "grad_norm": 0.2870406769574763,
+      "learning_rate": 0.00012512363385292738,
+      "loss": 1.0324,
+      "step": 2269
+    },
+    {
+      "epoch": 0.4367353073801293,
+      "grad_norm": 0.3217397222066845,
+      "learning_rate": 0.00012506330704783534,
+      "loss": 1.1431,
+      "step": 2270
+    },
+    {
+      "epoch": 0.43692770178866674,
+      "grad_norm": 0.2657327436360576,
+      "learning_rate": 0.00012500297050844366,
+      "loss": 1.0666,
+      "step": 2271
+    },
+    {
+      "epoch": 0.43712009619720427,
+      "grad_norm": 0.36698107717755923,
+      "learning_rate": 0.00012494262425818636,
+      "loss": 0.9727,
+      "step": 2272
+    },
+    {
+      "epoch": 0.4373124906057418,
+      "grad_norm": 0.3092521592366532,
+      "learning_rate": 0.00012488226832050117,
+      "loss": 1.1505,
+      "step": 2273
+    },
+    {
+      "epoch": 0.43750488501427925,
+      "grad_norm": 0.30366464728963216,
+      "learning_rate": 0.00012482190271882973,
+      "loss": 1.1118,
+      "step": 2274
+    },
+    {
+      "epoch": 0.4376972794228168,
+      "grad_norm": 0.35101809678043827,
+      "learning_rate": 0.00012476152747661727,
+      "loss": 1.1599,
+      "step": 2275
+    },
+    {
+      "epoch": 0.4378896738313543,
+      "grad_norm": 0.3078381708471384,
+      "learning_rate": 0.00012470114261731288,
+      "loss": 1.0856,
+      "step": 2276
+    },
+    {
+      "epoch": 0.43808206823989176,
+      "grad_norm": 0.28536122652740276,
+      "learning_rate": 0.0001246407481643693,
+      "loss": 1.0407,
+      "step": 2277
+    },
+    {
+      "epoch": 0.4382744626484293,
+      "grad_norm": 0.31125943751135476,
+      "learning_rate": 0.0001245803441412431,
+      "loss": 1.0388,
+      "step": 2278
+    },
+    {
+      "epoch": 0.4384668570569668,
+      "grad_norm": 0.3637167595517368,
+      "learning_rate": 0.00012451993057139443,
+      "loss": 1.0565,
+      "step": 2279
+    },
+    {
+      "epoch": 0.4386592514655043,
+      "grad_norm": 0.3042062479369799,
+      "learning_rate": 0.0001244595074782873,
+      "loss": 0.9961,
+      "step": 2280
+    },
+    {
+      "epoch": 0.4388516458740418,
+      "grad_norm": 0.32445297057755873,
+      "learning_rate": 0.00012439907488538934,
+      "loss": 1.0426,
+      "step": 2281
+    },
+    {
+      "epoch": 0.43904404028257926,
+      "grad_norm": 0.26772674511517314,
+      "learning_rate": 0.0001243386328161718,
+      "loss": 1.1199,
+      "step": 2282
+    },
+    {
+      "epoch": 0.4392364346911168,
+      "grad_norm": 0.2925346654302083,
+      "learning_rate": 0.00012427818129410976,
+      "loss": 1.1124,
+      "step": 2283
+    },
+    {
+      "epoch": 0.4394288290996543,
+      "grad_norm": 0.33678423980102185,
+      "learning_rate": 0.00012421772034268186,
+      "loss": 1.0548,
+      "step": 2284
+    },
+    {
+      "epoch": 0.43962122350819177,
+      "grad_norm": 0.27495217329774596,
+      "learning_rate": 0.00012415724998537041,
+      "loss": 0.9635,
+      "step": 2285
+    },
+    {
+      "epoch": 0.4398136179167293,
+      "grad_norm": 0.33566196919935115,
+      "learning_rate": 0.00012409677024566144,
+      "loss": 1.0872,
+      "step": 2286
+    },
+    {
+      "epoch": 0.4400060123252668,
+      "grad_norm": 0.3159059612233199,
+      "learning_rate": 0.00012403628114704459,
+      "loss": 1.0789,
+      "step": 2287
+    },
+    {
+      "epoch": 0.4401984067338043,
+      "grad_norm": 0.37142460638117897,
+      "learning_rate": 0.0001239757827130131,
+      "loss": 1.0503,
+      "step": 2288
+    },
+    {
+      "epoch": 0.4403908011423418,
+      "grad_norm": 0.283863797533487,
+      "learning_rate": 0.00012391527496706387,
+      "loss": 1.0236,
+      "step": 2289
+    },
+    {
+      "epoch": 0.4405831955508793,
+      "grad_norm": 0.28568554351328374,
+      "learning_rate": 0.00012385475793269744,
+      "loss": 1.0133,
+      "step": 2290
+    },
+    {
+      "epoch": 0.4407755899594168,
+      "grad_norm": 0.24887157270948593,
+      "learning_rate": 0.0001237942316334179,
+      "loss": 0.9984,
+      "step": 2291
+    },
+    {
+      "epoch": 0.4409679843679543,
+      "grad_norm": 0.3216530668952049,
+      "learning_rate": 0.00012373369609273296,
+      "loss": 1.1074,
+      "step": 2292
+    },
+    {
+      "epoch": 0.44116037877649183,
+      "grad_norm": 0.2760162632127163,
+      "learning_rate": 0.00012367315133415396,
+      "loss": 1.1407,
+      "step": 2293
+    },
+    {
+      "epoch": 0.4413527731850293,
+      "grad_norm": 0.2853925600872062,
+      "learning_rate": 0.00012361259738119575,
+      "loss": 1.0916,
+      "step": 2294
+    },
+    {
+      "epoch": 0.4415451675935668,
+      "grad_norm": 0.26758157504914154,
+      "learning_rate": 0.00012355203425737683,
+      "loss": 1.0632,
+      "step": 2295
+    },
+    {
+      "epoch": 0.4417375620021043,
+      "grad_norm": 0.2867192509295806,
+      "learning_rate": 0.00012349146198621918,
+      "loss": 1.0587,
+      "step": 2296
+    },
+    {
+      "epoch": 0.4419299564106418,
+      "grad_norm": 0.2985482590447692,
+      "learning_rate": 0.0001234308805912484,
+      "loss": 1.0867,
+      "step": 2297
+    },
+    {
+      "epoch": 0.44212235081917933,
+      "grad_norm": 0.30150695817957207,
+      "learning_rate": 0.00012337029009599357,
+      "loss": 1.0863,
+      "step": 2298
+    },
+    {
+      "epoch": 0.4423147452277168,
+      "grad_norm": 0.25723595041437636,
+      "learning_rate": 0.00012330969052398735,
+      "loss": 1.1078,
+      "step": 2299
+    },
+    {
+      "epoch": 0.4425071396362543,
+      "grad_norm": 0.32335571975394967,
+      "learning_rate": 0.00012324908189876595,
+      "loss": 1.0748,
+      "step": 2300
+    },
+    {
+      "epoch": 0.44269953404479184,
+      "grad_norm": 0.2736508636651387,
+      "learning_rate": 0.00012318846424386908,
+      "loss": 1.0912,
+      "step": 2301
+    },
+    {
+      "epoch": 0.4428919284533293,
+      "grad_norm": 0.4178067887573691,
+      "learning_rate": 0.0001231278375828398,
+      "loss": 1.0361,
+      "step": 2302
+    },
+    {
+      "epoch": 0.44308432286186683,
+      "grad_norm": 0.32715779755370955,
+      "learning_rate": 0.00012306720193922498,
+      "loss": 1.1229,
+      "step": 2303
+    },
+    {
+      "epoch": 0.44327671727040435,
+      "grad_norm": 0.2592005503252114,
+      "learning_rate": 0.00012300655733657473,
+      "loss": 1.069,
+      "step": 2304
+    },
+    {
+      "epoch": 0.4434691116789418,
+      "grad_norm": 0.29054898565665616,
+      "learning_rate": 0.0001229459037984427,
+      "loss": 0.9621,
+      "step": 2305
+    },
+    {
+      "epoch": 0.44366150608747934,
+      "grad_norm": 0.30149562517088924,
+      "learning_rate": 0.000122885241348386,
+      "loss": 1.0445,
+      "step": 2306
+    },
+    {
+      "epoch": 0.44385390049601686,
+      "grad_norm": 0.2891102344273007,
+      "learning_rate": 0.0001228245700099653,
+      "loss": 1.0208,
+      "step": 2307
+    },
+    {
+      "epoch": 0.44404629490455433,
+      "grad_norm": 0.35548198313909596,
+      "learning_rate": 0.00012276388980674466,
+      "loss": 1.0507,
+      "step": 2308
+    },
+    {
+      "epoch": 0.44423868931309185,
+      "grad_norm": 0.31786970443848317,
+      "learning_rate": 0.0001227032007622915,
+      "loss": 1.1075,
+      "step": 2309
+    },
+    {
+      "epoch": 0.4444310837216293,
+      "grad_norm": 0.2716463124923323,
+      "learning_rate": 0.00012264250290017673,
+      "loss": 1.0947,
+      "step": 2310
+    },
+    {
+      "epoch": 0.44462347813016684,
+      "grad_norm": 0.310421586742775,
+      "learning_rate": 0.00012258179624397478,
+      "loss": 1.0247,
+      "step": 2311
+    },
+    {
+      "epoch": 0.44481587253870436,
+      "grad_norm": 0.3407677014945833,
+      "learning_rate": 0.00012252108081726336,
+      "loss": 1.1464,
+      "step": 2312
+    },
+    {
+      "epoch": 0.4450082669472418,
+      "grad_norm": 0.34577595553051954,
+      "learning_rate": 0.0001224603566436237,
+      "loss": 1.13,
+      "step": 2313
+    },
+    {
+      "epoch": 0.44520066135577935,
+      "grad_norm": 0.30951874977039784,
+      "learning_rate": 0.00012239962374664028,
+      "loss": 1.1493,
+      "step": 2314
+    },
+    {
+      "epoch": 0.44539305576431687,
+      "grad_norm": 0.3144828648010736,
+      "learning_rate": 0.00012233888214990113,
+      "loss": 1.0235,
+      "step": 2315
+    },
+    {
+      "epoch": 0.44558545017285434,
+      "grad_norm": 0.3111912066543634,
+      "learning_rate": 0.00012227813187699757,
+      "loss": 1.0199,
+      "step": 2316
+    },
+    {
+      "epoch": 0.44577784458139186,
+      "grad_norm": 0.41224042941558886,
+      "learning_rate": 0.00012221737295152428,
+      "loss": 1.0213,
+      "step": 2317
+    },
+    {
+      "epoch": 0.4459702389899294,
+      "grad_norm": 0.29260845775090655,
+      "learning_rate": 0.00012215660539707936,
+      "loss": 1.0794,
+      "step": 2318
+    },
+    {
+      "epoch": 0.44616263339846685,
+      "grad_norm": 0.24297002760316577,
+      "learning_rate": 0.00012209582923726424,
+      "loss": 1.1566,
+      "step": 2319
+    },
+    {
+      "epoch": 0.44635502780700437,
+      "grad_norm": 0.30903627654247995,
+      "learning_rate": 0.00012203504449568362,
+      "loss": 1.1673,
+      "step": 2320
+    },
+    {
+      "epoch": 0.44654742221554183,
+      "grad_norm": 0.28261616514108767,
+      "learning_rate": 0.00012197425119594563,
+      "loss": 1.1039,
+      "step": 2321
+    },
+    {
+      "epoch": 0.44673981662407936,
+      "grad_norm": 0.2736733990665549,
+      "learning_rate": 0.00012191344936166169,
+      "loss": 1.2263,
+      "step": 2322
+    },
+    {
+      "epoch": 0.4469322110326169,
+      "grad_norm": 0.33792510219654087,
+      "learning_rate": 0.00012185263901644653,
+      "loss": 1.0791,
+      "step": 2323
+    },
+    {
+      "epoch": 0.44712460544115434,
+      "grad_norm": 0.3178971140769497,
+      "learning_rate": 0.00012179182018391819,
+      "loss": 0.9916,
+      "step": 2324
+    },
+    {
+      "epoch": 0.44731699984969187,
+      "grad_norm": 0.3007494206127297,
+      "learning_rate": 0.00012173099288769798,
+      "loss": 1.0695,
+      "step": 2325
+    },
+    {
+      "epoch": 0.4475093942582294,
+      "grad_norm": 0.26162239528255915,
+      "learning_rate": 0.00012167015715141057,
+      "loss": 1.1611,
+      "step": 2326
+    },
+    {
+      "epoch": 0.44770178866676685,
+      "grad_norm": 0.28619737465668155,
+      "learning_rate": 0.00012160931299868378,
+      "loss": 1.1205,
+      "step": 2327
+    },
+    {
+      "epoch": 0.4478941830753044,
+      "grad_norm": 0.29636098922637666,
+      "learning_rate": 0.00012154846045314884,
+      "loss": 1.055,
+      "step": 2328
+    },
+    {
+      "epoch": 0.4480865774838419,
+      "grad_norm": 0.2884154527928569,
+      "learning_rate": 0.0001214875995384402,
+      "loss": 0.9509,
+      "step": 2329
+    },
+    {
+      "epoch": 0.44827897189237936,
+      "grad_norm": 0.33756811442876733,
+      "learning_rate": 0.00012142673027819549,
+      "loss": 1.0143,
+      "step": 2330
+    },
+    {
+      "epoch": 0.4484713663009169,
+      "grad_norm": 0.37983895454116856,
+      "learning_rate": 0.00012136585269605558,
+      "loss": 1.0938,
+      "step": 2331
+    },
+    {
+      "epoch": 0.4486637607094544,
+      "grad_norm": 0.2792182801510256,
+      "learning_rate": 0.00012130496681566475,
+      "loss": 1.056,
+      "step": 2332
+    },
+    {
+      "epoch": 0.4488561551179919,
+      "grad_norm": 0.2754923642654477,
+      "learning_rate": 0.00012124407266067028,
+      "loss": 0.998,
+      "step": 2333
+    },
+    {
+      "epoch": 0.4490485495265294,
+      "grad_norm": 0.28017988863404325,
+      "learning_rate": 0.00012118317025472279,
+      "loss": 1.0686,
+      "step": 2334
+    },
+    {
+      "epoch": 0.44924094393506686,
+      "grad_norm": 0.32288983285963274,
+      "learning_rate": 0.00012112225962147605,
+      "loss": 1.1215,
+      "step": 2335
+    },
+    {
+      "epoch": 0.4494333383436044,
+      "grad_norm": 0.26517265302977716,
+      "learning_rate": 0.00012106134078458709,
+      "loss": 1.1099,
+      "step": 2336
+    },
+    {
+      "epoch": 0.4496257327521419,
+      "grad_norm": 0.2984857294534992,
+      "learning_rate": 0.00012100041376771605,
+      "loss": 1.1385,
+      "step": 2337
+    },
+    {
+      "epoch": 0.44981812716067937,
+      "grad_norm": 0.26019049655232596,
+      "learning_rate": 0.0001209394785945263,
+      "loss": 1.0233,
+      "step": 2338
+    },
+    {
+      "epoch": 0.4500105215692169,
+      "grad_norm": 0.3154873324289275,
+      "learning_rate": 0.00012087853528868431,
+      "loss": 1.0183,
+      "step": 2339
+    },
+    {
+      "epoch": 0.4502029159777544,
+      "grad_norm": 0.28033044241149513,
+      "learning_rate": 0.00012081758387385982,
+      "loss": 1.1335,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4503953103862919,
+      "grad_norm": 0.26562508763124576,
+      "learning_rate": 0.00012075662437372566,
+      "loss": 1.0227,
+      "step": 2341
+    },
+    {
+      "epoch": 0.4505877047948294,
+      "grad_norm": 0.2853151766333661,
+      "learning_rate": 0.00012069565681195776,
+      "loss": 1.097,
+      "step": 2342
+    },
+    {
+      "epoch": 0.4507800992033669,
+      "grad_norm": 0.2656617061596788,
+      "learning_rate": 0.0001206346812122352,
+      "loss": 1.0976,
+      "step": 2343
+    },
+    {
+      "epoch": 0.4509724936119044,
+      "grad_norm": 0.31942213283619686,
+      "learning_rate": 0.00012057369759824024,
+      "loss": 1.0778,
+      "step": 2344
+    },
+    {
+      "epoch": 0.4511648880204419,
+      "grad_norm": 0.28674403721840247,
+      "learning_rate": 0.00012051270599365824,
+      "loss": 1.0341,
+      "step": 2345
+    },
+    {
+      "epoch": 0.45135728242897943,
+      "grad_norm": 0.27215019991413036,
+      "learning_rate": 0.00012045170642217755,
+      "loss": 1.0346,
+      "step": 2346
+    },
+    {
+      "epoch": 0.4515496768375169,
+      "grad_norm": 0.3690805921299747,
+      "learning_rate": 0.00012039069890748978,
+      "loss": 1.1343,
+      "step": 2347
+    },
+    {
+      "epoch": 0.4517420712460544,
+      "grad_norm": 0.26925325933876115,
+      "learning_rate": 0.00012032968347328953,
+      "loss": 1.0873,
+      "step": 2348
+    },
+    {
+      "epoch": 0.4519344656545919,
+      "grad_norm": 0.32482317575765907,
+      "learning_rate": 0.00012026866014327444,
+      "loss": 1.1025,
+      "step": 2349
+    },
+    {
+      "epoch": 0.4521268600631294,
+      "grad_norm": 0.287112199346636,
+      "learning_rate": 0.00012020762894114535,
+      "loss": 1.0765,
+      "step": 2350
+    },
+    {
+      "epoch": 0.45231925447166693,
+      "grad_norm": 0.28727869303811654,
+      "learning_rate": 0.00012014658989060599,
+      "loss": 1.1223,
+      "step": 2351
+    },
+    {
+      "epoch": 0.4525116488802044,
+      "grad_norm": 0.35428310771798416,
+      "learning_rate": 0.00012008554301536328,
+      "loss": 1.1729,
+      "step": 2352
+    },
+    {
+      "epoch": 0.4527040432887419,
+      "grad_norm": 0.34666994090451514,
+      "learning_rate": 0.00012002448833912711,
+      "loss": 1.063,
+      "step": 2353
+    },
+    {
+      "epoch": 0.45289643769727944,
+      "grad_norm": 0.2694700834559087,
+      "learning_rate": 0.00011996342588561043,
+      "loss": 1.07,
+      "step": 2354
+    },
+    {
+      "epoch": 0.4530888321058169,
+      "grad_norm": 0.2547057156353204,
+      "learning_rate": 0.00011990235567852917,
+      "loss": 1.0682,
+      "step": 2355
+    },
+    {
+      "epoch": 0.45328122651435443,
+      "grad_norm": 0.31560348872341953,
+      "learning_rate": 0.00011984127774160225,
+      "loss": 1.1713,
+      "step": 2356
+    },
+    {
+      "epoch": 0.45347362092289195,
+      "grad_norm": 0.33723859691101654,
+      "learning_rate": 0.00011978019209855174,
+      "loss": 1.1122,
+      "step": 2357
+    },
+    {
+      "epoch": 0.4536660153314294,
+      "grad_norm": 0.3465027052553532,
+      "learning_rate": 0.00011971909877310253,
+      "loss": 1.1081,
+      "step": 2358
+    },
+    {
+      "epoch": 0.45385840973996694,
+      "grad_norm": 0.34054655641772036,
+      "learning_rate": 0.00011965799778898256,
+      "loss": 1.062,
+      "step": 2359
+    },
+    {
+      "epoch": 0.45405080414850446,
+      "grad_norm": 0.36220274477650694,
+      "learning_rate": 0.00011959688916992277,
+      "loss": 0.9796,
+      "step": 2360
+    },
+    {
+      "epoch": 0.4542431985570419,
+      "grad_norm": 0.3261600758421371,
+      "learning_rate": 0.00011953577293965708,
+      "loss": 1.1029,
+      "step": 2361
+    },
+    {
+      "epoch": 0.45443559296557945,
+      "grad_norm": 0.3322167143751733,
+      "learning_rate": 0.00011947464912192227,
+      "loss": 1.0256,
+      "step": 2362
+    },
+    {
+      "epoch": 0.4546279873741169,
+      "grad_norm": 0.2933363106680554,
+      "learning_rate": 0.00011941351774045815,
+      "loss": 1.0768,
+      "step": 2363
+    },
+    {
+      "epoch": 0.45482038178265444,
+      "grad_norm": 0.35459185899438167,
+      "learning_rate": 0.00011935237881900743,
+      "loss": 1.0544,
+      "step": 2364
+    },
+    {
+      "epoch": 0.45501277619119196,
+      "grad_norm": 0.2766898676397248,
+      "learning_rate": 0.00011929123238131579,
+      "loss": 1.0767,
+      "step": 2365
+    },
+    {
+      "epoch": 0.4552051705997294,
+      "grad_norm": 0.3094980044974744,
+      "learning_rate": 0.00011923007845113176,
+      "loss": 1.085,
+      "step": 2366
+    },
+    {
+      "epoch": 0.45539756500826695,
+      "grad_norm": 0.40095158661505026,
+      "learning_rate": 0.00011916891705220687,
+      "loss": 1.1046,
+      "step": 2367
+    },
+    {
+      "epoch": 0.45558995941680447,
+      "grad_norm": 0.318977039667831,
+      "learning_rate": 0.00011910774820829549,
+      "loss": 1.0942,
+      "step": 2368
+    },
+    {
+      "epoch": 0.45578235382534193,
+      "grad_norm": 0.3072838892086311,
+      "learning_rate": 0.00011904657194315485,
+      "loss": 1.0379,
+      "step": 2369
+    },
+    {
+      "epoch": 0.45597474823387946,
+      "grad_norm": 0.36936268066809186,
+      "learning_rate": 0.00011898538828054517,
+      "loss": 1.1097,
+      "step": 2370
+    },
+    {
+      "epoch": 0.456167142642417,
+      "grad_norm": 0.33227338222367825,
+      "learning_rate": 0.00011892419724422946,
+      "loss": 1.0905,
+      "step": 2371
+    },
+    {
+      "epoch": 0.45635953705095444,
+      "grad_norm": 0.3276191470664485,
+      "learning_rate": 0.00011886299885797356,
+      "loss": 1.1359,
+      "step": 2372
+    },
+    {
+      "epoch": 0.45655193145949197,
+      "grad_norm": 0.36487282188100045,
+      "learning_rate": 0.00011880179314554629,
+      "loss": 1.0813,
+      "step": 2373
+    },
+    {
+      "epoch": 0.45674432586802943,
+      "grad_norm": 0.2645058478472509,
+      "learning_rate": 0.00011874058013071923,
+      "loss": 1.138,
+      "step": 2374
+    },
+    {
+      "epoch": 0.45693672027656695,
+      "grad_norm": 0.34007458867204715,
+      "learning_rate": 0.00011867935983726677,
+      "loss": 1.0535,
+      "step": 2375
+    },
+    {
+      "epoch": 0.4571291146851045,
+      "grad_norm": 0.2850488740324468,
+      "learning_rate": 0.0001186181322889662,
+      "loss": 1.1171,
+      "step": 2376
+    },
+    {
+      "epoch": 0.45732150909364194,
+      "grad_norm": 0.2432625894589558,
+      "learning_rate": 0.00011855689750959757,
+      "loss": 1.0781,
+      "step": 2377
+    },
+    {
+      "epoch": 0.45751390350217946,
+      "grad_norm": 0.29386832480620884,
+      "learning_rate": 0.0001184956555229438,
+      "loss": 1.15,
+      "step": 2378
+    },
+    {
+      "epoch": 0.457706297910717,
+      "grad_norm": 0.26862782200848306,
+      "learning_rate": 0.00011843440635279055,
+      "loss": 1.1044,
+      "step": 2379
+    },
+    {
+      "epoch": 0.45789869231925445,
+      "grad_norm": 0.42155578944834304,
+      "learning_rate": 0.00011837315002292628,
+      "loss": 1.0668,
+      "step": 2380
+    },
+    {
+      "epoch": 0.458091086727792,
+      "grad_norm": 0.3021973754570017,
+      "learning_rate": 0.00011831188655714225,
+      "loss": 1.0575,
+      "step": 2381
+    },
+    {
+      "epoch": 0.4582834811363295,
+      "grad_norm": 0.2629087916374506,
+      "learning_rate": 0.0001182506159792325,
+      "loss": 1.1012,
+      "step": 2382
+    },
+    {
+      "epoch": 0.45847587554486696,
+      "grad_norm": 0.31147553186013294,
+      "learning_rate": 0.00011818933831299382,
+      "loss": 1.0495,
+      "step": 2383
+    },
+    {
+      "epoch": 0.4586682699534045,
+      "grad_norm": 0.28110777583345414,
+      "learning_rate": 0.00011812805358222571,
+      "loss": 1.0646,
+      "step": 2384
+    },
+    {
+      "epoch": 0.458860664361942,
+      "grad_norm": 0.3158535523354326,
+      "learning_rate": 0.00011806676181073048,
+      "loss": 1.1176,
+      "step": 2385
+    },
+    {
+      "epoch": 0.45905305877047947,
+      "grad_norm": 0.30322657882936893,
+      "learning_rate": 0.00011800546302231316,
+      "loss": 1.0483,
+      "step": 2386
+    },
+    {
+      "epoch": 0.459245453179017,
+      "grad_norm": 0.31842168511961255,
+      "learning_rate": 0.00011794415724078147,
+      "loss": 1.1516,
+      "step": 2387
+    },
+    {
+      "epoch": 0.45943784758755446,
+      "grad_norm": 0.31712597940783244,
+      "learning_rate": 0.00011788284448994586,
+      "loss": 1.0873,
+      "step": 2388
+    },
+    {
+      "epoch": 0.459630241996092,
+      "grad_norm": 0.30256333423176934,
+      "learning_rate": 0.00011782152479361957,
+      "loss": 1.0781,
+      "step": 2389
+    },
+    {
+      "epoch": 0.4598226364046295,
+      "grad_norm": 0.29955739856902935,
+      "learning_rate": 0.00011776019817561834,
+      "loss": 0.9694,
+      "step": 2390
+    },
+    {
+      "epoch": 0.46001503081316697,
+      "grad_norm": 0.281558904867827,
+      "learning_rate": 0.00011769886465976086,
+      "loss": 1.0423,
+      "step": 2391
+    },
+    {
+      "epoch": 0.4602074252217045,
+      "grad_norm": 0.3431549084874546,
+      "learning_rate": 0.00011763752426986823,
+      "loss": 1.0413,
+      "step": 2392
+    },
+    {
+      "epoch": 0.460399819630242,
+      "grad_norm": 0.2779997009445957,
+      "learning_rate": 0.00011757617702976442,
+      "loss": 1.1059,
+      "step": 2393
+    },
+    {
+      "epoch": 0.4605922140387795,
+      "grad_norm": 0.2818652570094477,
+      "learning_rate": 0.000117514822963276,
+      "loss": 1.1298,
+      "step": 2394
+    },
+    {
+      "epoch": 0.460784608447317,
+      "grad_norm": 0.26916626681457934,
+      "learning_rate": 0.00011745346209423216,
+      "loss": 1.0699,
+      "step": 2395
+    },
+    {
+      "epoch": 0.4609770028558545,
+      "grad_norm": 0.2742812769718791,
+      "learning_rate": 0.00011739209444646477,
+      "loss": 1.0416,
+      "step": 2396
+    },
+    {
+      "epoch": 0.461169397264392,
+      "grad_norm": 0.33114910849519996,
+      "learning_rate": 0.00011733072004380827,
+      "loss": 1.1922,
+      "step": 2397
+    },
+    {
+      "epoch": 0.4613617916729295,
+      "grad_norm": 0.3338207337392457,
+      "learning_rate": 0.00011726933891009983,
+      "loss": 1.123,
+      "step": 2398
+    },
+    {
+      "epoch": 0.46155418608146703,
+      "grad_norm": 0.3715629482370197,
+      "learning_rate": 0.00011720795106917917,
+      "loss": 1.104,
+      "step": 2399
+    },
+    {
+      "epoch": 0.4617465804900045,
+      "grad_norm": 0.26564206366393767,
+      "learning_rate": 0.00011714655654488859,
+      "loss": 1.1455,
+      "step": 2400
+    },
+    {
+      "epoch": 0.461938974898542,
+      "grad_norm": 0.25886338899516165,
+      "learning_rate": 0.00011708515536107299,
+      "loss": 1.0442,
+      "step": 2401
+    },
+    {
+      "epoch": 0.4621313693070795,
+      "grad_norm": 0.33739270473479455,
+      "learning_rate": 0.00011702374754157997,
+      "loss": 0.9982,
+      "step": 2402
+    },
+    {
+      "epoch": 0.462323763715617,
+      "grad_norm": 0.3516541452146191,
+      "learning_rate": 0.00011696233311025957,
+      "loss": 1.1059,
+      "step": 2403
+    },
+    {
+      "epoch": 0.46251615812415453,
+      "grad_norm": 0.28091670825325904,
+      "learning_rate": 0.00011690091209096441,
+      "loss": 1.0874,
+      "step": 2404
+    },
+    {
+      "epoch": 0.462708552532692,
+      "grad_norm": 0.44757795831758546,
+      "learning_rate": 0.00011683948450754976,
+      "loss": 1.034,
+      "step": 2405
+    },
+    {
+      "epoch": 0.4629009469412295,
+      "grad_norm": 0.28681423657596056,
+      "learning_rate": 0.00011677805038387337,
+      "loss": 1.0373,
+      "step": 2406
+    },
+    {
+      "epoch": 0.46309334134976704,
+      "grad_norm": 0.2703859948163023,
+      "learning_rate": 0.00011671660974379555,
+      "loss": 1.0819,
+      "step": 2407
+    },
+    {
+      "epoch": 0.4632857357583045,
+      "grad_norm": 0.3163916228926271,
+      "learning_rate": 0.00011665516261117912,
+      "loss": 1.1011,
+      "step": 2408
+    },
+    {
+      "epoch": 0.46347813016684203,
+      "grad_norm": 0.27350316299605615,
+      "learning_rate": 0.00011659370900988947,
+      "loss": 0.9953,
+      "step": 2409
+    },
+    {
+      "epoch": 0.46367052457537955,
+      "grad_norm": 0.29232531072109774,
+      "learning_rate": 0.00011653224896379439,
+      "loss": 1.0311,
+      "step": 2410
+    },
+    {
+      "epoch": 0.463862918983917,
+      "grad_norm": 0.36498789740764503,
+      "learning_rate": 0.0001164707824967644,
+      "loss": 0.9932,
+      "step": 2411
+    },
+    {
+      "epoch": 0.46405531339245454,
+      "grad_norm": 0.3140372558749361,
+      "learning_rate": 0.00011640930963267225,
+      "loss": 1.1021,
+      "step": 2412
+    },
+    {
+      "epoch": 0.46424770780099206,
+      "grad_norm": 0.2826486998073901,
+      "learning_rate": 0.00011634783039539328,
+      "loss": 1.0499,
+      "step": 2413
+    },
+    {
+      "epoch": 0.4644401022095295,
+      "grad_norm": 0.28672878644276156,
+      "learning_rate": 0.0001162863448088054,
+      "loss": 1.0147,
+      "step": 2414
+    },
+    {
+      "epoch": 0.46463249661806705,
+      "grad_norm": 0.3462903380271429,
+      "learning_rate": 0.00011622485289678886,
+      "loss": 1.1023,
+      "step": 2415
+    },
+    {
+      "epoch": 0.4648248910266045,
+      "grad_norm": 0.3308961791445071,
+      "learning_rate": 0.00011616335468322641,
+      "loss": 1.0974,
+      "step": 2416
+    },
+    {
+      "epoch": 0.46501728543514204,
+      "grad_norm": 0.32146135483703064,
+      "learning_rate": 0.00011610185019200323,
+      "loss": 0.9963,
+      "step": 2417
+    },
+    {
+      "epoch": 0.46520967984367956,
+      "grad_norm": 0.30784864548896584,
+      "learning_rate": 0.00011604033944700701,
+      "loss": 1.0494,
+      "step": 2418
+    },
+    {
+      "epoch": 0.465402074252217,
+      "grad_norm": 0.3481653940663827,
+      "learning_rate": 0.00011597882247212776,
+      "loss": 0.9466,
+      "step": 2419
+    },
+    {
+      "epoch": 0.46559446866075455,
+      "grad_norm": 0.2670947755386483,
+      "learning_rate": 0.000115917299291258,
+      "loss": 1.0268,
+      "step": 2420
+    },
+    {
+      "epoch": 0.46578686306929207,
+      "grad_norm": 0.3241922940225946,
+      "learning_rate": 0.00011585576992829261,
+      "loss": 1.1111,
+      "step": 2421
+    },
+    {
+      "epoch": 0.46597925747782953,
+      "grad_norm": 0.37070086289730036,
+      "learning_rate": 0.00011579423440712886,
+      "loss": 1.0777,
+      "step": 2422
+    },
+    {
+      "epoch": 0.46617165188636706,
+      "grad_norm": 0.29925540305956333,
+      "learning_rate": 0.00011573269275166652,
+      "loss": 1.1428,
+      "step": 2423
+    },
+    {
+      "epoch": 0.4663640462949046,
+      "grad_norm": 0.32103609892224627,
+      "learning_rate": 0.00011567114498580758,
+      "loss": 1.0564,
+      "step": 2424
+    },
+    {
+      "epoch": 0.46655644070344204,
+      "grad_norm": 0.31649134667432066,
+      "learning_rate": 0.00011560959113345649,
+      "loss": 1.1181,
+      "step": 2425
+    },
+    {
+      "epoch": 0.46674883511197957,
+      "grad_norm": 0.3495530455616566,
+      "learning_rate": 0.00011554803121852005,
+      "loss": 1.0189,
+      "step": 2426
+    },
+    {
+      "epoch": 0.4669412295205171,
+      "grad_norm": 0.2918508303469182,
+      "learning_rate": 0.00011548646526490749,
+      "loss": 1.0675,
+      "step": 2427
+    },
+    {
+      "epoch": 0.46713362392905455,
+      "grad_norm": 0.2988520983363615,
+      "learning_rate": 0.00011542489329653024,
+      "loss": 1.0899,
+      "step": 2428
+    },
+    {
+      "epoch": 0.4673260183375921,
+      "grad_norm": 0.2576049541547143,
+      "learning_rate": 0.00011536331533730221,
+      "loss": 1.0882,
+      "step": 2429
+    },
+    {
+      "epoch": 0.46751841274612954,
+      "grad_norm": 0.26004376842000726,
+      "learning_rate": 0.00011530173141113947,
+      "loss": 1.0869,
+      "step": 2430
+    },
+    {
+      "epoch": 0.46771080715466706,
+      "grad_norm": 0.27800349692250365,
+      "learning_rate": 0.00011524014154196063,
+      "loss": 1.0189,
+      "step": 2431
+    },
+    {
+      "epoch": 0.4679032015632046,
+      "grad_norm": 0.36082387811606265,
+      "learning_rate": 0.00011517854575368642,
+      "loss": 1.0089,
+      "step": 2432
+    },
+    {
+      "epoch": 0.46809559597174205,
+      "grad_norm": 0.2948309937878554,
+      "learning_rate": 0.00011511694407023994,
+      "loss": 1.0338,
+      "step": 2433
+    },
+    {
+      "epoch": 0.4682879903802796,
+      "grad_norm": 0.3286568445270373,
+      "learning_rate": 0.00011505533651554654,
+      "loss": 1.012,
+      "step": 2434
+    },
+    {
+      "epoch": 0.4684803847888171,
+      "grad_norm": 0.3167569352769985,
+      "learning_rate": 0.00011499372311353398,
+      "loss": 1.0574,
+      "step": 2435
+    },
+    {
+      "epoch": 0.46867277919735456,
+      "grad_norm": 0.2639196979482772,
+      "learning_rate": 0.0001149321038881321,
+      "loss": 1.0875,
+      "step": 2436
+    },
+    {
+      "epoch": 0.4688651736058921,
+      "grad_norm": 0.2661156202677314,
+      "learning_rate": 0.00011487047886327313,
+      "loss": 1.1757,
+      "step": 2437
+    },
+    {
+      "epoch": 0.4690575680144296,
+      "grad_norm": 0.3009152148444373,
+      "learning_rate": 0.00011480884806289152,
+      "loss": 1.0668,
+      "step": 2438
+    },
+    {
+      "epoch": 0.46924996242296707,
+      "grad_norm": 0.37544920914394553,
+      "learning_rate": 0.00011474721151092396,
+      "loss": 1.0562,
+      "step": 2439
+    },
+    {
+      "epoch": 0.4694423568315046,
+      "grad_norm": 0.3013747633863066,
+      "learning_rate": 0.00011468556923130942,
+      "loss": 1.1155,
+      "step": 2440
+    },
+    {
+      "epoch": 0.46963475124004206,
+      "grad_norm": 0.2916423561005298,
+      "learning_rate": 0.000114623921247989,
+      "loss": 1.0454,
+      "step": 2441
+    },
+    {
+      "epoch": 0.4698271456485796,
+      "grad_norm": 0.30209189554148286,
+      "learning_rate": 0.00011456226758490602,
+      "loss": 1.1009,
+      "step": 2442
+    },
+    {
+      "epoch": 0.4700195400571171,
+      "grad_norm": 0.4444882427931337,
+      "learning_rate": 0.00011450060826600617,
+      "loss": 1.0391,
+      "step": 2443
+    },
+    {
+      "epoch": 0.47021193446565457,
+      "grad_norm": 0.2856392935289551,
+      "learning_rate": 0.00011443894331523717,
+      "loss": 1.069,
+      "step": 2444
+    },
+    {
+      "epoch": 0.4704043288741921,
+      "grad_norm": 0.2929797319254082,
+      "learning_rate": 0.00011437727275654892,
+      "loss": 1.1212,
+      "step": 2445
+    },
+    {
+      "epoch": 0.4705967232827296,
+      "grad_norm": 0.3388439031139151,
+      "learning_rate": 0.00011431559661389362,
+      "loss": 1.0363,
+      "step": 2446
+    },
+    {
+      "epoch": 0.4707891176912671,
+      "grad_norm": 0.42997177745217463,
+      "learning_rate": 0.00011425391491122556,
+      "loss": 1.0631,
+      "step": 2447
+    },
+    {
+      "epoch": 0.4709815120998046,
+      "grad_norm": 0.26047216024754893,
+      "learning_rate": 0.00011419222767250119,
+      "loss": 1.1293,
+      "step": 2448
+    },
+    {
+      "epoch": 0.4711739065083421,
+      "grad_norm": 0.3067675308083437,
+      "learning_rate": 0.00011413053492167916,
+      "loss": 0.9995,
+      "step": 2449
+    },
+    {
+      "epoch": 0.4713663009168796,
+      "grad_norm": 0.34392169665096517,
+      "learning_rate": 0.00011406883668272014,
+      "loss": 1.1126,
+      "step": 2450
+    },
+    {
+      "epoch": 0.4715586953254171,
+      "grad_norm": 0.28771277526109607,
+      "learning_rate": 0.0001140071329795871,
+      "loss": 1.0986,
+      "step": 2451
+    },
+    {
+      "epoch": 0.47175108973395463,
+      "grad_norm": 0.2966084911052895,
+      "learning_rate": 0.000113945423836245,
+      "loss": 1.1005,
+      "step": 2452
+    },
+    {
+      "epoch": 0.4719434841424921,
+      "grad_norm": 0.32011544620504834,
+      "learning_rate": 0.00011388370927666101,
+      "loss": 0.9713,
+      "step": 2453
+    },
+    {
+      "epoch": 0.4721358785510296,
+      "grad_norm": 0.2906261196653228,
+      "learning_rate": 0.00011382198932480428,
+      "loss": 1.0719,
+      "step": 2454
+    },
+    {
+      "epoch": 0.4723282729595671,
+      "grad_norm": 0.30714586501668545,
+      "learning_rate": 0.00011376026400464617,
+      "loss": 1.0281,
+      "step": 2455
+    },
+    {
+      "epoch": 0.4725206673681046,
+      "grad_norm": 0.3078207126511925,
+      "learning_rate": 0.00011369853334016009,
+      "loss": 1.096,
+      "step": 2456
+    },
+    {
+      "epoch": 0.47271306177664213,
+      "grad_norm": 0.32017475890493713,
+      "learning_rate": 0.0001136367973553215,
+      "loss": 1.0625,
+      "step": 2457
+    },
+    {
+      "epoch": 0.4729054561851796,
+      "grad_norm": 0.32482432949187245,
+      "learning_rate": 0.00011357505607410797,
+      "loss": 1.1041,
+      "step": 2458
+    },
+    {
+      "epoch": 0.4730978505937171,
+      "grad_norm": 0.26549248121833174,
+      "learning_rate": 0.00011351330952049908,
+      "loss": 1.1516,
+      "step": 2459
+    },
+    {
+      "epoch": 0.47329024500225464,
+      "grad_norm": 0.2624649604038464,
+      "learning_rate": 0.00011345155771847646,
+      "loss": 1.0213,
+      "step": 2460
+    },
+    {
+      "epoch": 0.4734826394107921,
+      "grad_norm": 0.30952729581419347,
+      "learning_rate": 0.00011338980069202387,
+      "loss": 1.0981,
+      "step": 2461
+    },
+    {
+      "epoch": 0.4736750338193296,
+      "grad_norm": 0.37943549674014154,
+      "learning_rate": 0.00011332803846512697,
+      "loss": 1.0728,
+      "step": 2462
+    },
+    {
+      "epoch": 0.47386742822786715,
+      "grad_norm": 0.28536607623304083,
+      "learning_rate": 0.00011326627106177347,
+      "loss": 1.0961,
+      "step": 2463
+    },
+    {
+      "epoch": 0.4740598226364046,
+      "grad_norm": 0.3067005267882926,
+      "learning_rate": 0.0001132044985059532,
+      "loss": 1.0373,
+      "step": 2464
+    },
+    {
+      "epoch": 0.47425221704494214,
+      "grad_norm": 0.3204502995861548,
+      "learning_rate": 0.00011314272082165784,
+      "loss": 0.9995,
+      "step": 2465
+    },
+    {
+      "epoch": 0.47444461145347966,
+      "grad_norm": 0.3291924589533158,
+      "learning_rate": 0.00011308093803288118,
+      "loss": 1.0895,
+      "step": 2466
+    },
+    {
+      "epoch": 0.4746370058620171,
+      "grad_norm": 0.3417270692438031,
+      "learning_rate": 0.00011301915016361891,
+      "loss": 1.1515,
+      "step": 2467
+    },
+    {
+      "epoch": 0.47482940027055465,
+      "grad_norm": 0.2985479459112958,
+      "learning_rate": 0.0001129573572378687,
+      "loss": 1.1156,
+      "step": 2468
+    },
+    {
+      "epoch": 0.4750217946790921,
+      "grad_norm": 0.2672237204246719,
+      "learning_rate": 0.00011289555927963031,
+      "loss": 1.0562,
+      "step": 2469
+    },
+    {
+      "epoch": 0.47521418908762963,
+      "grad_norm": 0.2782865846539132,
+      "learning_rate": 0.00011283375631290527,
+      "loss": 1.0884,
+      "step": 2470
+    },
+    {
+      "epoch": 0.47540658349616716,
+      "grad_norm": 0.29891313505133355,
+      "learning_rate": 0.00011277194836169714,
+      "loss": 1.1497,
+      "step": 2471
+    },
+    {
+      "epoch": 0.4755989779047046,
+      "grad_norm": 0.2939161275307998,
+      "learning_rate": 0.00011271013545001143,
+      "loss": 1.1317,
+      "step": 2472
+    },
+    {
+      "epoch": 0.47579137231324214,
+      "grad_norm": 0.2935173681906,
+      "learning_rate": 0.00011264831760185562,
+      "loss": 1.1136,
+      "step": 2473
+    },
+    {
+      "epoch": 0.47598376672177967,
+      "grad_norm": 0.30430278052563287,
+      "learning_rate": 0.00011258649484123896,
+      "loss": 1.0256,
+      "step": 2474
+    },
+    {
+      "epoch": 0.47617616113031713,
+      "grad_norm": 0.327969229597276,
+      "learning_rate": 0.00011252466719217273,
+      "loss": 1.0752,
+      "step": 2475
+    },
+    {
+      "epoch": 0.47636855553885465,
+      "grad_norm": 0.3070924360453279,
+      "learning_rate": 0.00011246283467867012,
+      "loss": 1.0928,
+      "step": 2476
+    },
+    {
+      "epoch": 0.4765609499473922,
+      "grad_norm": 0.26050029622275817,
+      "learning_rate": 0.00011240099732474611,
+      "loss": 1.0053,
+      "step": 2477
+    },
+    {
+      "epoch": 0.47675334435592964,
+      "grad_norm": 0.31549357223842334,
+      "learning_rate": 0.00011233915515441764,
+      "loss": 1.0787,
+      "step": 2478
+    },
+    {
+      "epoch": 0.47694573876446716,
+      "grad_norm": 0.31213070358897843,
+      "learning_rate": 0.00011227730819170348,
+      "loss": 1.1011,
+      "step": 2479
+    },
+    {
+      "epoch": 0.4771381331730047,
+      "grad_norm": 0.30214759097239374,
+      "learning_rate": 0.0001122154564606243,
+      "loss": 1.0055,
+      "step": 2480
+    },
+    {
+      "epoch": 0.47733052758154215,
+      "grad_norm": 0.3315365730666589,
+      "learning_rate": 0.0001121535999852026,
+      "loss": 1.0846,
+      "step": 2481
+    },
+    {
+      "epoch": 0.4775229219900797,
+      "grad_norm": 0.3027198839201215,
+      "learning_rate": 0.0001120917387894627,
+      "loss": 1.1211,
+      "step": 2482
+    },
+    {
+      "epoch": 0.47771531639861714,
+      "grad_norm": 0.29139373354457415,
+      "learning_rate": 0.00011202987289743079,
+      "loss": 0.9735,
+      "step": 2483
+    },
+    {
+      "epoch": 0.47790771080715466,
+      "grad_norm": 0.28622281575804387,
+      "learning_rate": 0.00011196800233313487,
+      "loss": 1.0428,
+      "step": 2484
+    },
+    {
+      "epoch": 0.4781001052156922,
+      "grad_norm": 0.3083370931275903,
+      "learning_rate": 0.00011190612712060476,
+      "loss": 1.0225,
+      "step": 2485
+    },
+    {
+      "epoch": 0.47829249962422965,
+      "grad_norm": 0.3453898740103891,
+      "learning_rate": 0.00011184424728387204,
+      "loss": 1.1167,
+      "step": 2486
+    },
+    {
+      "epoch": 0.47848489403276717,
+      "grad_norm": 0.2809754425406349,
+      "learning_rate": 0.00011178236284697017,
+      "loss": 1.0966,
+      "step": 2487
+    },
+    {
+      "epoch": 0.4786772884413047,
+      "grad_norm": 0.33603653127564614,
+      "learning_rate": 0.00011172047383393434,
+      "loss": 0.9848,
+      "step": 2488
+    },
+    {
+      "epoch": 0.47886968284984216,
+      "grad_norm": 0.32530594572132543,
+      "learning_rate": 0.00011165858026880152,
+      "loss": 1.1285,
+      "step": 2489
+    },
+    {
+      "epoch": 0.4790620772583797,
+      "grad_norm": 0.28477375710938924,
+      "learning_rate": 0.00011159668217561048,
+      "loss": 0.9873,
+      "step": 2490
+    },
+    {
+      "epoch": 0.4792544716669172,
+      "grad_norm": 0.26850697287763875,
+      "learning_rate": 0.0001115347795784017,
+      "loss": 1.1048,
+      "step": 2491
+    },
+    {
+      "epoch": 0.47944686607545467,
+      "grad_norm": 0.32263744208401285,
+      "learning_rate": 0.00011147287250121745,
+      "loss": 1.0445,
+      "step": 2492
+    },
+    {
+      "epoch": 0.4796392604839922,
+      "grad_norm": 0.3527327693555526,
+      "learning_rate": 0.00011141096096810173,
+      "loss": 1.085,
+      "step": 2493
+    },
+    {
+      "epoch": 0.4798316548925297,
+      "grad_norm": 0.304616016772181,
+      "learning_rate": 0.00011134904500310028,
+      "loss": 1.0607,
+      "step": 2494
+    },
+    {
+      "epoch": 0.4800240493010672,
+      "grad_norm": 0.30320494796454556,
+      "learning_rate": 0.00011128712463026048,
+      "loss": 1.1149,
+      "step": 2495
+    },
+    {
+      "epoch": 0.4802164437096047,
+      "grad_norm": 0.2887639606992245,
+      "learning_rate": 0.00011122519987363155,
+      "loss": 1.0105,
+      "step": 2496
+    },
+    {
+      "epoch": 0.48040883811814217,
+      "grad_norm": 0.34694566403040783,
+      "learning_rate": 0.00011116327075726435,
+      "loss": 1.1541,
+      "step": 2497
+    },
+    {
+      "epoch": 0.4806012325266797,
+      "grad_norm": 0.249702910513426,
+      "learning_rate": 0.00011110133730521142,
+      "loss": 1.0795,
+      "step": 2498
+    },
+    {
+      "epoch": 0.4807936269352172,
+      "grad_norm": 0.2852476586916126,
+      "learning_rate": 0.000111039399541527,
+      "loss": 1.0455,
+      "step": 2499
+    },
+    {
+      "epoch": 0.4809860213437547,
+      "grad_norm": 0.3477661351396813,
+      "learning_rate": 0.000110977457490267,
+      "loss": 1.0762,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4811784157522922,
+      "grad_norm": 0.29912510631231753,
+      "learning_rate": 0.000110915511175489,
+      "loss": 1.0767,
+      "step": 2501
+    },
+    {
+      "epoch": 0.4813708101608297,
+      "grad_norm": 0.31702866915248035,
+      "learning_rate": 0.00011085356062125225,
+      "loss": 1.0021,
+      "step": 2502
+    },
+    {
+      "epoch": 0.4815632045693672,
+      "grad_norm": 0.28837950745195445,
+      "learning_rate": 0.00011079160585161759,
+      "loss": 1.1517,
+      "step": 2503
+    },
+    {
+      "epoch": 0.4817555989779047,
+      "grad_norm": 0.271426384992999,
+      "learning_rate": 0.00011072964689064759,
+      "loss": 0.999,
+      "step": 2504
+    },
+    {
+      "epoch": 0.48194799338644223,
+      "grad_norm": 0.32204520349360083,
+      "learning_rate": 0.0001106676837624064,
+      "loss": 1.0527,
+      "step": 2505
+    },
+    {
+      "epoch": 0.4821403877949797,
+      "grad_norm": 0.31410915148149016,
+      "learning_rate": 0.00011060571649095972,
+      "loss": 1.031,
+      "step": 2506
+    },
+    {
+      "epoch": 0.4823327822035172,
+      "grad_norm": 0.27931797715272966,
+      "learning_rate": 0.000110543745100375,
+      "loss": 1.1246,
+      "step": 2507
+    },
+    {
+      "epoch": 0.4825251766120547,
+      "grad_norm": 0.2752501976050457,
+      "learning_rate": 0.00011048176961472114,
+      "loss": 1.1451,
+      "step": 2508
+    },
+    {
+      "epoch": 0.4827175710205922,
+      "grad_norm": 0.2747532456619337,
+      "learning_rate": 0.00011041979005806875,
+      "loss": 1.0734,
+      "step": 2509
+    },
+    {
+      "epoch": 0.48290996542912973,
+      "grad_norm": 0.2342734006645817,
+      "learning_rate": 0.00011035780645449001,
+      "loss": 1.0454,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4831023598376672,
+      "grad_norm": 0.32742955385460815,
+      "learning_rate": 0.00011029581882805857,
+      "loss": 0.9748,
+      "step": 2511
+    },
+    {
+      "epoch": 0.4832947542462047,
+      "grad_norm": 0.37946687327215983,
+      "learning_rate": 0.00011023382720284972,
+      "loss": 1.0628,
+      "step": 2512
+    },
+    {
+      "epoch": 0.48348714865474224,
+      "grad_norm": 0.2732650447123761,
+      "learning_rate": 0.00011017183160294033,
+      "loss": 1.0259,
+      "step": 2513
+    },
+    {
+      "epoch": 0.4836795430632797,
+      "grad_norm": 0.2729585455966021,
+      "learning_rate": 0.00011010983205240877,
+      "loss": 1.0414,
+      "step": 2514
+    },
+    {
+      "epoch": 0.4838719374718172,
+      "grad_norm": 0.32655314274408553,
+      "learning_rate": 0.00011004782857533489,
+      "loss": 1.1244,
+      "step": 2515
+    },
+    {
+      "epoch": 0.48406433188035475,
+      "grad_norm": 0.23700820491114358,
+      "learning_rate": 0.00010998582119580019,
+      "loss": 1.1172,
+      "step": 2516
+    },
+    {
+      "epoch": 0.4842567262888922,
+      "grad_norm": 0.3928723737777664,
+      "learning_rate": 0.00010992380993788762,
+      "loss": 1.0799,
+      "step": 2517
+    },
+    {
+      "epoch": 0.48444912069742974,
+      "grad_norm": 0.2912192721303968,
+      "learning_rate": 0.00010986179482568161,
+      "loss": 1.1404,
+      "step": 2518
+    },
+    {
+      "epoch": 0.48464151510596726,
+      "grad_norm": 0.2978549220616921,
+      "learning_rate": 0.00010979977588326815,
+      "loss": 1.1675,
+      "step": 2519
+    },
+    {
+      "epoch": 0.4848339095145047,
+      "grad_norm": 0.2926954785297245,
+      "learning_rate": 0.00010973775313473464,
+      "loss": 1.1438,
+      "step": 2520
+    },
+    {
+      "epoch": 0.48502630392304225,
+      "grad_norm": 0.36000964866389756,
+      "learning_rate": 0.00010967572660417,
+      "loss": 1.038,
+      "step": 2521
+    },
+    {
+      "epoch": 0.4852186983315797,
+      "grad_norm": 0.37910792556434847,
+      "learning_rate": 0.00010961369631566467,
+      "loss": 1.0837,
+      "step": 2522
+    },
+    {
+      "epoch": 0.48541109274011723,
+      "grad_norm": 0.32581771747503674,
+      "learning_rate": 0.00010955166229331048,
+      "loss": 1.0686,
+      "step": 2523
+    },
+    {
+      "epoch": 0.48560348714865476,
+      "grad_norm": 0.40931583909959296,
+      "learning_rate": 0.00010948962456120068,
+      "loss": 1.1053,
+      "step": 2524
+    },
+    {
+      "epoch": 0.4857958815571922,
+      "grad_norm": 0.2777818201392416,
+      "learning_rate": 0.00010942758314343006,
+      "loss": 1.1047,
+      "step": 2525
+    },
+    {
+      "epoch": 0.48598827596572974,
+      "grad_norm": 0.2723539189762568,
+      "learning_rate": 0.00010936553806409482,
+      "loss": 0.9886,
+      "step": 2526
+    },
+    {
+      "epoch": 0.48618067037426727,
+      "grad_norm": 0.29465289512987186,
+      "learning_rate": 0.00010930348934729248,
+      "loss": 1.1171,
+      "step": 2527
+    },
+    {
+      "epoch": 0.48637306478280473,
+      "grad_norm": 0.34597539212264616,
+      "learning_rate": 0.00010924143701712211,
+      "loss": 1.0029,
+      "step": 2528
+    },
+    {
+      "epoch": 0.48656545919134225,
+      "grad_norm": 0.2786773823360309,
+      "learning_rate": 0.00010917938109768404,
+      "loss": 1.0654,
+      "step": 2529
+    },
+    {
+      "epoch": 0.4867578535998798,
+      "grad_norm": 0.35869410693668063,
+      "learning_rate": 0.00010911732161308015,
+      "loss": 1.0327,
+      "step": 2530
+    },
+    {
+      "epoch": 0.48695024800841724,
+      "grad_norm": 0.30607208498100946,
+      "learning_rate": 0.00010905525858741364,
+      "loss": 1.173,
+      "step": 2531
+    },
+    {
+      "epoch": 0.48714264241695476,
+      "grad_norm": 0.2887048883746236,
+      "learning_rate": 0.000108993192044789,
+      "loss": 1.1132,
+      "step": 2532
+    },
+    {
+      "epoch": 0.4873350368254923,
+      "grad_norm": 0.2722475202006041,
+      "learning_rate": 0.00010893112200931219,
+      "loss": 1.0333,
+      "step": 2533
+    },
+    {
+      "epoch": 0.48752743123402975,
+      "grad_norm": 0.3061443115824018,
+      "learning_rate": 0.00010886904850509052,
+      "loss": 1.0755,
+      "step": 2534
+    },
+    {
+      "epoch": 0.4877198256425673,
+      "grad_norm": 0.28704696881583397,
+      "learning_rate": 0.00010880697155623263,
+      "loss": 1.0567,
+      "step": 2535
+    },
+    {
+      "epoch": 0.48791222005110474,
+      "grad_norm": 0.26753181810345267,
+      "learning_rate": 0.00010874489118684845,
+      "loss": 1.1283,
+      "step": 2536
+    },
+    {
+      "epoch": 0.48810461445964226,
+      "grad_norm": 0.3972482521246636,
+      "learning_rate": 0.00010868280742104928,
+      "loss": 1.1804,
+      "step": 2537
+    },
+    {
+      "epoch": 0.4882970088681798,
+      "grad_norm": 0.4312688126899029,
+      "learning_rate": 0.00010862072028294776,
+      "loss": 1.1485,
+      "step": 2538
+    },
+    {
+      "epoch": 0.48848940327671725,
+      "grad_norm": 0.2605607369109513,
+      "learning_rate": 0.00010855862979665787,
+      "loss": 1.0787,
+      "step": 2539
+    },
+    {
+      "epoch": 0.48868179768525477,
+      "grad_norm": 0.33775397735827983,
+      "learning_rate": 0.00010849653598629476,
+      "loss": 1.0298,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4888741920937923,
+      "grad_norm": 0.3875969639199608,
+      "learning_rate": 0.00010843443887597495,
+      "loss": 1.0571,
+      "step": 2541
+    },
+    {
+      "epoch": 0.48906658650232976,
+      "grad_norm": 0.2666988960590085,
+      "learning_rate": 0.00010837233848981632,
+      "loss": 1.097,
+      "step": 2542
+    },
+    {
+      "epoch": 0.4892589809108673,
+      "grad_norm": 0.31221334810809465,
+      "learning_rate": 0.00010831023485193788,
+      "loss": 1.0807,
+      "step": 2543
+    },
+    {
+      "epoch": 0.4894513753194048,
+      "grad_norm": 0.29251910199626263,
+      "learning_rate": 0.00010824812798645996,
+      "loss": 1.1063,
+      "step": 2544
+    },
+    {
+      "epoch": 0.48964376972794227,
+      "grad_norm": 0.3491128473642241,
+      "learning_rate": 0.00010818601791750417,
+      "loss": 1.0837,
+      "step": 2545
+    },
+    {
+      "epoch": 0.4898361641364798,
+      "grad_norm": 0.24704517407416768,
+      "learning_rate": 0.00010812390466919338,
+      "loss": 1.0647,
+      "step": 2546
+    },
+    {
+      "epoch": 0.4900285585450173,
+      "grad_norm": 0.2613052283228881,
+      "learning_rate": 0.00010806178826565161,
+      "loss": 1.1837,
+      "step": 2547
+    },
+    {
+      "epoch": 0.4902209529535548,
+      "grad_norm": 0.3441270718443512,
+      "learning_rate": 0.00010799966873100417,
+      "loss": 1.1432,
+      "step": 2548
+    },
+    {
+      "epoch": 0.4904133473620923,
+      "grad_norm": 0.2968938160248746,
+      "learning_rate": 0.00010793754608937758,
+      "loss": 1.173,
+      "step": 2549
+    },
+    {
+      "epoch": 0.49060574177062977,
+      "grad_norm": 0.2709027638329477,
+      "learning_rate": 0.00010787542036489955,
+      "loss": 1.1116,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4907981361791673,
+      "grad_norm": 0.2350569375522259,
+      "learning_rate": 0.00010781329158169902,
+      "loss": 1.0758,
+      "step": 2551
+    },
+    {
+      "epoch": 0.4909905305877048,
+      "grad_norm": 0.30759414944729424,
+      "learning_rate": 0.00010775115976390606,
+      "loss": 0.9265,
+      "step": 2552
+    },
+    {
+      "epoch": 0.4911829249962423,
+      "grad_norm": 0.3420019187451639,
+      "learning_rate": 0.00010768902493565196,
+      "loss": 1.0837,
+      "step": 2553
+    },
+    {
+      "epoch": 0.4913753194047798,
+      "grad_norm": 0.26775812639526614,
+      "learning_rate": 0.00010762688712106917,
+      "loss": 1.0271,
+      "step": 2554
+    },
+    {
+      "epoch": 0.4915677138133173,
+      "grad_norm": 0.28723952286715476,
+      "learning_rate": 0.00010756474634429132,
+      "loss": 1.0798,
+      "step": 2555
+    },
+    {
+      "epoch": 0.4917601082218548,
+      "grad_norm": 0.3115782159029881,
+      "learning_rate": 0.00010750260262945314,
+      "loss": 1.131,
+      "step": 2556
+    },
+    {
+      "epoch": 0.4919525026303923,
+      "grad_norm": 0.2638660440405208,
+      "learning_rate": 0.00010744045600069055,
+      "loss": 1.0695,
+      "step": 2557
+    },
+    {
+      "epoch": 0.49214489703892983,
+      "grad_norm": 0.31320999280179135,
+      "learning_rate": 0.00010737830648214062,
+      "loss": 1.0713,
+      "step": 2558
+    },
+    {
+      "epoch": 0.4923372914474673,
+      "grad_norm": 0.2625801367677137,
+      "learning_rate": 0.00010731615409794143,
+      "loss": 1.0681,
+      "step": 2559
+    },
+    {
+      "epoch": 0.4925296858560048,
+      "grad_norm": 0.3148140684694239,
+      "learning_rate": 0.00010725399887223233,
+      "loss": 1.0057,
+      "step": 2560
+    },
+    {
+      "epoch": 0.49272208026454234,
+      "grad_norm": 0.3986659373568962,
+      "learning_rate": 0.00010719184082915363,
+      "loss": 1.1615,
+      "step": 2561
+    },
+    {
+      "epoch": 0.4929144746730798,
+      "grad_norm": 0.3212605066343875,
+      "learning_rate": 0.00010712967999284682,
+      "loss": 1.0646,
+      "step": 2562
+    },
+    {
+      "epoch": 0.4931068690816173,
+      "grad_norm": 0.26949647752321304,
+      "learning_rate": 0.00010706751638745447,
+      "loss": 1.0891,
+      "step": 2563
+    },
+    {
+      "epoch": 0.4932992634901548,
+      "grad_norm": 0.3147194011096115,
+      "learning_rate": 0.00010700535003712022,
+      "loss": 0.9794,
+      "step": 2564
+    },
+    {
+      "epoch": 0.4934916578986923,
+      "grad_norm": 0.27266970059366,
+      "learning_rate": 0.0001069431809659887,
+      "loss": 1.1,
+      "step": 2565
+    },
+    {
+      "epoch": 0.49368405230722984,
+      "grad_norm": 0.3216890395388259,
+      "learning_rate": 0.0001068810091982057,
+      "loss": 1.1474,
+      "step": 2566
+    },
+    {
+      "epoch": 0.4938764467157673,
+      "grad_norm": 0.26261033443954596,
+      "learning_rate": 0.00010681883475791803,
+      "loss": 1.0907,
+      "step": 2567
+    },
+    {
+      "epoch": 0.4940688411243048,
+      "grad_norm": 0.32713879777200705,
+      "learning_rate": 0.0001067566576692735,
+      "loss": 1.1572,
+      "step": 2568
+    },
+    {
+      "epoch": 0.49426123553284235,
+      "grad_norm": 0.2720177629628606,
+      "learning_rate": 0.00010669447795642103,
+      "loss": 1.0116,
+      "step": 2569
+    },
+    {
+      "epoch": 0.4944536299413798,
+      "grad_norm": 0.31697399187316516,
+      "learning_rate": 0.00010663229564351041,
+      "loss": 1.1468,
+      "step": 2570
+    },
+    {
+      "epoch": 0.49464602434991733,
+      "grad_norm": 0.31191288445545967,
+      "learning_rate": 0.00010657011075469259,
+      "loss": 1.0153,
+      "step": 2571
+    },
+    {
+      "epoch": 0.49483841875845486,
+      "grad_norm": 0.2697020764144713,
+      "learning_rate": 0.0001065079233141195,
+      "loss": 1.0992,
+      "step": 2572
+    },
+    {
+      "epoch": 0.4950308131669923,
+      "grad_norm": 0.2833597359763036,
+      "learning_rate": 0.00010644573334594395,
+      "loss": 0.9997,
+      "step": 2573
+    },
+    {
+      "epoch": 0.49522320757552984,
+      "grad_norm": 0.2832566604790744,
+      "learning_rate": 0.00010638354087431985,
+      "loss": 1.0163,
+      "step": 2574
+    },
+    {
+      "epoch": 0.4954156019840673,
+      "grad_norm": 0.33171593960929263,
+      "learning_rate": 0.00010632134592340204,
+      "loss": 1.0766,
+      "step": 2575
+    },
+    {
+      "epoch": 0.49560799639260483,
+      "grad_norm": 0.2822850139561916,
+      "learning_rate": 0.0001062591485173463,
+      "loss": 1.1371,
+      "step": 2576
+    },
+    {
+      "epoch": 0.49580039080114235,
+      "grad_norm": 0.40692033349733203,
+      "learning_rate": 0.00010619694868030942,
+      "loss": 1.0977,
+      "step": 2577
+    },
+    {
+      "epoch": 0.4959927852096798,
+      "grad_norm": 0.29906416732614016,
+      "learning_rate": 0.00010613474643644908,
+      "loss": 1.0658,
+      "step": 2578
+    },
+    {
+      "epoch": 0.49618517961821734,
+      "grad_norm": 0.29619045834352786,
+      "learning_rate": 0.0001060725418099239,
+      "loss": 0.9958,
+      "step": 2579
+    },
+    {
+      "epoch": 0.49637757402675486,
+      "grad_norm": 0.33353541430091754,
+      "learning_rate": 0.00010601033482489346,
+      "loss": 1.0225,
+      "step": 2580
+    },
+    {
+      "epoch": 0.49656996843529233,
+      "grad_norm": 0.2923531560869263,
+      "learning_rate": 0.00010594812550551825,
+      "loss": 1.0753,
+      "step": 2581
+    },
+    {
+      "epoch": 0.49676236284382985,
+      "grad_norm": 0.2730992878560735,
+      "learning_rate": 0.00010588591387595961,
+      "loss": 1.0468,
+      "step": 2582
+    },
+    {
+      "epoch": 0.4969547572523674,
+      "grad_norm": 0.32951452182989416,
+      "learning_rate": 0.00010582369996037984,
+      "loss": 1.075,
+      "step": 2583
+    },
+    {
+      "epoch": 0.49714715166090484,
+      "grad_norm": 0.33759576811525555,
+      "learning_rate": 0.00010576148378294213,
+      "loss": 1.1513,
+      "step": 2584
+    },
+    {
+      "epoch": 0.49733954606944236,
+      "grad_norm": 0.3197998033215524,
+      "learning_rate": 0.0001056992653678105,
+      "loss": 1.0203,
+      "step": 2585
+    },
+    {
+      "epoch": 0.4975319404779799,
+      "grad_norm": 0.2869669207047471,
+      "learning_rate": 0.00010563704473914986,
+      "loss": 1.0632,
+      "step": 2586
+    },
+    {
+      "epoch": 0.49772433488651735,
+      "grad_norm": 0.27943692117962127,
+      "learning_rate": 0.00010557482192112602,
+      "loss": 1.0381,
+      "step": 2587
+    },
+    {
+      "epoch": 0.49791672929505487,
+      "grad_norm": 0.30820481904437813,
+      "learning_rate": 0.00010551259693790556,
+      "loss": 0.9657,
+      "step": 2588
+    },
+    {
+      "epoch": 0.49810912370359234,
+      "grad_norm": 0.2980366953914205,
+      "learning_rate": 0.00010545036981365602,
+      "loss": 1.1031,
+      "step": 2589
+    },
+    {
+      "epoch": 0.49830151811212986,
+      "grad_norm": 0.28084455929546465,
+      "learning_rate": 0.0001053881405725456,
+      "loss": 1.0672,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4984939125206674,
+      "grad_norm": 0.28771732213575407,
+      "learning_rate": 0.00010532590923874349,
+      "loss": 0.9641,
+      "step": 2591
+    },
+    {
+      "epoch": 0.49868630692920485,
+      "grad_norm": 0.34389843173403906,
+      "learning_rate": 0.00010526367583641958,
+      "loss": 1.0802,
+      "step": 2592
+    },
+    {
+      "epoch": 0.49887870133774237,
+      "grad_norm": 0.3732839017386677,
+      "learning_rate": 0.00010520144038974466,
+      "loss": 0.9591,
+      "step": 2593
+    },
+    {
+      "epoch": 0.4990710957462799,
+      "grad_norm": 0.34674764173762207,
+      "learning_rate": 0.00010513920292289021,
+      "loss": 1.0662,
+      "step": 2594
+    },
+    {
+      "epoch": 0.49926349015481736,
+      "grad_norm": 0.29948221452749685,
+      "learning_rate": 0.00010507696346002858,
+      "loss": 1.0667,
+      "step": 2595
+    },
+    {
+      "epoch": 0.4994558845633549,
+      "grad_norm": 0.3475302187613384,
+      "learning_rate": 0.00010501472202533284,
+      "loss": 1.0526,
+      "step": 2596
+    },
+    {
+      "epoch": 0.4996482789718924,
+      "grad_norm": 0.2997366532667675,
+      "learning_rate": 0.00010495247864297684,
+      "loss": 1.0594,
+      "step": 2597
+    },
+    {
+      "epoch": 0.49984067338042987,
+      "grad_norm": 0.30994459615061065,
+      "learning_rate": 0.00010489023333713521,
+      "loss": 0.9625,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5000330677889674,
+      "grad_norm": 0.3686286763345276,
+      "learning_rate": 0.00010482798613198329,
+      "loss": 0.998,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5002254621975049,
+      "grad_norm": 0.3163621335454191,
+      "learning_rate": 0.00010476573705169719,
+      "loss": 1.0923,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5004178566060424,
+      "grad_norm": 0.30144952669238584,
+      "learning_rate": 0.00010470348612045375,
+      "loss": 1.0036,
+      "step": 2601
+    },
+    {
+      "epoch": 0.5006102510145799,
+      "grad_norm": 0.37770094153244205,
+      "learning_rate": 0.00010464123336243048,
+      "loss": 1.0699,
+      "step": 2602
+    },
+    {
+      "epoch": 0.5008026454231174,
+      "grad_norm": 0.35502403376775843,
+      "learning_rate": 0.00010457897880180565,
+      "loss": 1.0344,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5009950398316549,
+      "grad_norm": 0.25372534454374396,
+      "learning_rate": 0.00010451672246275825,
+      "loss": 0.9606,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5011874342401924,
+      "grad_norm": 0.2835736745333487,
+      "learning_rate": 0.00010445446436946789,
+      "loss": 1.0621,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5013798286487299,
+      "grad_norm": 0.31098446920641104,
+      "learning_rate": 0.00010439220454611487,
+      "loss": 1.0722,
+      "step": 2606
+    },
+    {
+      "epoch": 0.5015722230572675,
+      "grad_norm": 0.31501677875249645,
+      "learning_rate": 0.00010432994301688021,
+      "loss": 1.0762,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5017646174658049,
+      "grad_norm": 0.2656806264910574,
+      "learning_rate": 0.00010426767980594558,
+      "loss": 1.092,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5019570118743424,
+      "grad_norm": 0.29719386992571994,
+      "learning_rate": 0.00010420541493749332,
+      "loss": 0.975,
+      "step": 2609
+    },
+    {
+      "epoch": 0.5021494062828799,
+      "grad_norm": 0.2930741864291432,
+      "learning_rate": 0.00010414314843570634,
+      "loss": 1.0567,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5023418006914174,
+      "grad_norm": 0.27144962176639403,
+      "learning_rate": 0.00010408088032476822,
+      "loss": 1.0395,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5025341950999549,
+      "grad_norm": 0.30055243602535087,
+      "learning_rate": 0.00010401861062886325,
+      "loss": 1.1904,
+      "step": 2612
+    },
+    {
+      "epoch": 0.5027265895084924,
+      "grad_norm": 0.3549779485460265,
+      "learning_rate": 0.00010395633937217621,
+      "loss": 1.0692,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5029189839170299,
+      "grad_norm": 0.26843258273395965,
+      "learning_rate": 0.00010389406657889254,
+      "loss": 0.9925,
+      "step": 2614
+    },
+    {
+      "epoch": 0.5031113783255674,
+      "grad_norm": 0.2574939342570903,
+      "learning_rate": 0.00010383179227319826,
+      "loss": 1.0854,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5033037727341049,
+      "grad_norm": 0.27246358490382644,
+      "learning_rate": 0.00010376951647928005,
+      "loss": 1.1691,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5034961671426424,
+      "grad_norm": 0.274051757221271,
+      "learning_rate": 0.00010370723922132506,
+      "loss": 1.1617,
+      "step": 2617
+    },
+    {
+      "epoch": 0.5036885615511799,
+      "grad_norm": 0.3553250776188068,
+      "learning_rate": 0.00010364496052352108,
+      "loss": 0.9463,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5038809559597174,
+      "grad_norm": 0.2898745997092644,
+      "learning_rate": 0.00010358268041005643,
+      "loss": 1.0145,
+      "step": 2619
+    },
+    {
+      "epoch": 0.504073350368255,
+      "grad_norm": 0.24409249658862053,
+      "learning_rate": 0.00010352039890511998,
+      "loss": 1.0222,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5042657447767924,
+      "grad_norm": 0.2889115040400644,
+      "learning_rate": 0.00010345811603290118,
+      "loss": 1.1149,
+      "step": 2621
+    },
+    {
+      "epoch": 0.5044581391853299,
+      "grad_norm": 0.31833577913532923,
+      "learning_rate": 0.00010339583181758996,
+      "loss": 1.0173,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5046505335938675,
+      "grad_norm": 0.26892391856935105,
+      "learning_rate": 0.00010333354628337679,
+      "loss": 0.9919,
+      "step": 2623
+    },
+    {
+      "epoch": 0.5048429280024049,
+      "grad_norm": 0.29372557612701433,
+      "learning_rate": 0.00010327125945445264,
+      "loss": 1.1192,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5050353224109424,
+      "grad_norm": 0.30514287790701333,
+      "learning_rate": 0.00010320897135500905,
+      "loss": 1.0816,
+      "step": 2625
+    },
+    {
+      "epoch": 0.50522771681948,
+      "grad_norm": 0.2663340382110766,
+      "learning_rate": 0.00010314668200923792,
+      "loss": 0.9706,
+      "step": 2626
+    },
+    {
+      "epoch": 0.5054201112280174,
+      "grad_norm": 0.27497538646337466,
+      "learning_rate": 0.00010308439144133177,
+      "loss": 0.9269,
+      "step": 2627
+    },
+    {
+      "epoch": 0.5056125056365549,
+      "grad_norm": 0.3747878051056086,
+      "learning_rate": 0.00010302209967548353,
+      "loss": 1.0395,
+      "step": 2628
+    },
+    {
+      "epoch": 0.5058049000450925,
+      "grad_norm": 0.32699886016865964,
+      "learning_rate": 0.00010295980673588659,
+      "loss": 1.0573,
+      "step": 2629
+    },
+    {
+      "epoch": 0.50599729445363,
+      "grad_norm": 0.33481225110411705,
+      "learning_rate": 0.00010289751264673484,
+      "loss": 1.0797,
+      "step": 2630
+    },
+    {
+      "epoch": 0.5061896888621674,
+      "grad_norm": 0.3733809136960564,
+      "learning_rate": 0.00010283521743222255,
+      "loss": 1.0912,
+      "step": 2631
+    },
+    {
+      "epoch": 0.506382083270705,
+      "grad_norm": 0.29952830016660215,
+      "learning_rate": 0.00010277292111654446,
+      "loss": 1.0953,
+      "step": 2632
+    },
+    {
+      "epoch": 0.5065744776792425,
+      "grad_norm": 0.28969819219624027,
+      "learning_rate": 0.00010271062372389581,
+      "loss": 1.0718,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5067668720877799,
+      "grad_norm": 0.2855058383205973,
+      "learning_rate": 0.00010264832527847212,
+      "loss": 1.106,
+      "step": 2634
+    },
+    {
+      "epoch": 0.5069592664963174,
+      "grad_norm": 0.2866604186099265,
+      "learning_rate": 0.00010258602580446941,
+      "loss": 1.0022,
+      "step": 2635
+    },
+    {
+      "epoch": 0.507151660904855,
+      "grad_norm": 0.3028260283123913,
+      "learning_rate": 0.00010252372532608405,
+      "loss": 1.1272,
+      "step": 2636
+    },
+    {
+      "epoch": 0.5073440553133924,
+      "grad_norm": 0.35341763494903256,
+      "learning_rate": 0.0001024614238675129,
+      "loss": 1.0565,
+      "step": 2637
+    },
+    {
+      "epoch": 0.5075364497219299,
+      "grad_norm": 0.2813090653405737,
+      "learning_rate": 0.00010239912145295303,
+      "loss": 1.077,
+      "step": 2638
+    },
+    {
+      "epoch": 0.5077288441304675,
+      "grad_norm": 0.28657049227187126,
+      "learning_rate": 0.00010233681810660207,
+      "loss": 1.0751,
+      "step": 2639
+    },
+    {
+      "epoch": 0.507921238539005,
+      "grad_norm": 0.42781320547043034,
+      "learning_rate": 0.00010227451385265787,
+      "loss": 0.9878,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5081136329475424,
+      "grad_norm": 0.39459958921270927,
+      "learning_rate": 0.00010221220871531869,
+      "loss": 0.9328,
+      "step": 2641
+    },
+    {
+      "epoch": 0.50830602735608,
+      "grad_norm": 0.4018612321690641,
+      "learning_rate": 0.00010214990271878318,
+      "loss": 1.0537,
+      "step": 2642
+    },
+    {
+      "epoch": 0.5084984217646175,
+      "grad_norm": 0.29053851124061514,
+      "learning_rate": 0.00010208759588725016,
+      "loss": 1.0487,
+      "step": 2643
+    },
+    {
+      "epoch": 0.5086908161731549,
+      "grad_norm": 0.29738007242546566,
+      "learning_rate": 0.000102025288244919,
+      "loss": 1.0101,
+      "step": 2644
+    },
+    {
+      "epoch": 0.5088832105816925,
+      "grad_norm": 0.3250214277297258,
+      "learning_rate": 0.00010196297981598922,
+      "loss": 1.1369,
+      "step": 2645
+    },
+    {
+      "epoch": 0.50907560499023,
+      "grad_norm": 0.34190619789289134,
+      "learning_rate": 0.00010190067062466069,
+      "loss": 1.0531,
+      "step": 2646
+    },
+    {
+      "epoch": 0.5092679993987674,
+      "grad_norm": 0.2947098929850498,
+      "learning_rate": 0.00010183836069513359,
+      "loss": 1.1089,
+      "step": 2647
+    },
+    {
+      "epoch": 0.509460393807305,
+      "grad_norm": 0.3536138621507215,
+      "learning_rate": 0.00010177605005160838,
+      "loss": 1.1267,
+      "step": 2648
+    },
+    {
+      "epoch": 0.5096527882158425,
+      "grad_norm": 0.26788902121616626,
+      "learning_rate": 0.00010171373871828578,
+      "loss": 1.1427,
+      "step": 2649
+    },
+    {
+      "epoch": 0.50984518262438,
+      "grad_norm": 0.30130205048109276,
+      "learning_rate": 0.00010165142671936685,
+      "loss": 1.0723,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5100375770329175,
+      "grad_norm": 0.2770160095983166,
+      "learning_rate": 0.00010158911407905278,
+      "loss": 1.1712,
+      "step": 2651
+    },
+    {
+      "epoch": 0.510229971441455,
+      "grad_norm": 0.26671936004694313,
+      "learning_rate": 0.00010152680082154513,
+      "loss": 1.0766,
+      "step": 2652
+    },
+    {
+      "epoch": 0.5104223658499925,
+      "grad_norm": 0.40072532513761644,
+      "learning_rate": 0.0001014644869710456,
+      "loss": 0.9855,
+      "step": 2653
+    },
+    {
+      "epoch": 0.51061476025853,
+      "grad_norm": 0.26219903596687183,
+      "learning_rate": 0.00010140217255175625,
+      "loss": 1.0791,
+      "step": 2654
+    },
+    {
+      "epoch": 0.5108071546670675,
+      "grad_norm": 0.3228343803495144,
+      "learning_rate": 0.00010133985758787921,
+      "loss": 0.99,
+      "step": 2655
+    },
+    {
+      "epoch": 0.510999549075605,
+      "grad_norm": 0.3244082266353016,
+      "learning_rate": 0.00010127754210361693,
+      "loss": 1.1227,
+      "step": 2656
+    },
+    {
+      "epoch": 0.5111919434841425,
+      "grad_norm": 0.36065187450740854,
+      "learning_rate": 0.00010121522612317203,
+      "loss": 1.0787,
+      "step": 2657
+    },
+    {
+      "epoch": 0.51138433789268,
+      "grad_norm": 0.34035338975275603,
+      "learning_rate": 0.00010115290967074728,
+      "loss": 1.1208,
+      "step": 2658
+    },
+    {
+      "epoch": 0.5115767323012175,
+      "grad_norm": 0.30493698593909646,
+      "learning_rate": 0.00010109059277054574,
+      "loss": 1.1078,
+      "step": 2659
+    },
+    {
+      "epoch": 0.511769126709755,
+      "grad_norm": 0.2840535902614037,
+      "learning_rate": 0.00010102827544677049,
+      "loss": 1.0772,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5119615211182925,
+      "grad_norm": 0.2988125811168846,
+      "learning_rate": 0.00010096595772362491,
+      "loss": 1.1629,
+      "step": 2661
+    },
+    {
+      "epoch": 0.51215391552683,
+      "grad_norm": 0.32361070861201896,
+      "learning_rate": 0.0001009036396253125,
+      "loss": 1.0723,
+      "step": 2662
+    },
+    {
+      "epoch": 0.5123463099353674,
+      "grad_norm": 0.27752404145549714,
+      "learning_rate": 0.00010084132117603689,
+      "loss": 1.0606,
+      "step": 2663
+    },
+    {
+      "epoch": 0.512538704343905,
+      "grad_norm": 0.25688663722710564,
+      "learning_rate": 0.00010077900240000179,
+      "loss": 1.0527,
+      "step": 2664
+    },
+    {
+      "epoch": 0.5127310987524425,
+      "grad_norm": 0.28328637659502315,
+      "learning_rate": 0.00010071668332141115,
+      "loss": 1.0329,
+      "step": 2665
+    },
+    {
+      "epoch": 0.51292349316098,
+      "grad_norm": 0.3342898186566314,
+      "learning_rate": 0.00010065436396446899,
+      "loss": 1.0472,
+      "step": 2666
+    },
+    {
+      "epoch": 0.5131158875695175,
+      "grad_norm": 0.33043726905356363,
+      "learning_rate": 0.00010059204435337937,
+      "loss": 1.0561,
+      "step": 2667
+    },
+    {
+      "epoch": 0.513308281978055,
+      "grad_norm": 0.34945810910931496,
+      "learning_rate": 0.00010052972451234657,
+      "loss": 1.0958,
+      "step": 2668
+    },
+    {
+      "epoch": 0.5135006763865925,
+      "grad_norm": 0.3279197009321726,
+      "learning_rate": 0.00010046740446557486,
+      "loss": 1.0391,
+      "step": 2669
+    },
+    {
+      "epoch": 0.51369307079513,
+      "grad_norm": 0.3317184582762969,
+      "learning_rate": 0.00010040508423726865,
+      "loss": 1.0203,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5138854652036675,
+      "grad_norm": 0.24702965120855871,
+      "learning_rate": 0.00010034276385163237,
+      "loss": 1.0539,
+      "step": 2671
+    },
+    {
+      "epoch": 0.514077859612205,
+      "grad_norm": 0.29554405798530037,
+      "learning_rate": 0.00010028044333287057,
+      "loss": 1.0724,
+      "step": 2672
+    },
+    {
+      "epoch": 0.5142702540207426,
+      "grad_norm": 0.3113869697297527,
+      "learning_rate": 0.0001002181227051878,
+      "loss": 0.9427,
+      "step": 2673
+    },
+    {
+      "epoch": 0.51446264842928,
+      "grad_norm": 0.2982163121180696,
+      "learning_rate": 0.00010015580199278873,
+      "loss": 0.9998,
+      "step": 2674
+    },
+    {
+      "epoch": 0.5146550428378175,
+      "grad_norm": 0.4630578795642548,
+      "learning_rate": 0.00010009348121987794,
+      "loss": 1.0666,
+      "step": 2675
+    },
+    {
+      "epoch": 0.5148474372463551,
+      "grad_norm": 0.3424864477705635,
+      "learning_rate": 0.0001000311604106601,
+      "loss": 1.0087,
+      "step": 2676
+    },
+    {
+      "epoch": 0.5150398316548925,
+      "grad_norm": 0.2749273603356742,
+      "learning_rate": 9.996883958933992e-05,
+      "loss": 1.0186,
+      "step": 2677
+    },
+    {
+      "epoch": 0.51523222606343,
+      "grad_norm": 0.2751288045530157,
+      "learning_rate": 9.990651878012211e-05,
+      "loss": 1.1156,
+      "step": 2678
+    },
+    {
+      "epoch": 0.5154246204719676,
+      "grad_norm": 0.3852014664002532,
+      "learning_rate": 9.984419800721132e-05,
+      "loss": 1.0128,
+      "step": 2679
+    },
+    {
+      "epoch": 0.515617014880505,
+      "grad_norm": 0.3370617303217846,
+      "learning_rate": 9.97818772948122e-05,
+      "loss": 0.9892,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5158094092890425,
+      "grad_norm": 0.36322859436231897,
+      "learning_rate": 9.971955666712944e-05,
+      "loss": 1.0977,
+      "step": 2681
+    },
+    {
+      "epoch": 0.5160018036975801,
+      "grad_norm": 0.2849679576817036,
+      "learning_rate": 9.965723614836764e-05,
+      "loss": 0.9967,
+      "step": 2682
+    },
+    {
+      "epoch": 0.5161941981061176,
+      "grad_norm": 0.32749669764721134,
+      "learning_rate": 9.959491576273139e-05,
+      "loss": 1.0315,
+      "step": 2683
+    },
+    {
+      "epoch": 0.516386592514655,
+      "grad_norm": 0.2980660635842453,
+      "learning_rate": 9.953259553442518e-05,
+      "loss": 1.0892,
+      "step": 2684
+    },
+    {
+      "epoch": 0.5165789869231926,
+      "grad_norm": 0.3130228369711336,
+      "learning_rate": 9.947027548765348e-05,
+      "loss": 1.1218,
+      "step": 2685
+    },
+    {
+      "epoch": 0.5167713813317301,
+      "grad_norm": 0.31813700200054223,
+      "learning_rate": 9.940795564662063e-05,
+      "loss": 1.1224,
+      "step": 2686
+    },
+    {
+      "epoch": 0.5169637757402675,
+      "grad_norm": 0.29391047872832265,
+      "learning_rate": 9.934563603553102e-05,
+      "loss": 1.0267,
+      "step": 2687
+    },
+    {
+      "epoch": 0.5171561701488051,
+      "grad_norm": 0.3142164764833463,
+      "learning_rate": 9.928331667858886e-05,
+      "loss": 1.0478,
+      "step": 2688
+    },
+    {
+      "epoch": 0.5173485645573426,
+      "grad_norm": 0.3191948303423121,
+      "learning_rate": 9.922099759999822e-05,
+      "loss": 1.0669,
+      "step": 2689
+    },
+    {
+      "epoch": 0.51754095896588,
+      "grad_norm": 0.3188475774024312,
+      "learning_rate": 9.915867882396315e-05,
+      "loss": 1.0429,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5177333533744175,
+      "grad_norm": 0.2909051017286928,
+      "learning_rate": 9.909636037468753e-05,
+      "loss": 1.0452,
+      "step": 2691
+    },
+    {
+      "epoch": 0.5179257477829551,
+      "grad_norm": 0.30835000947059094,
+      "learning_rate": 9.903404227637508e-05,
+      "loss": 0.972,
+      "step": 2692
+    },
+    {
+      "epoch": 0.5181181421914925,
+      "grad_norm": 0.41145153340626306,
+      "learning_rate": 9.897172455322953e-05,
+      "loss": 1.0764,
+      "step": 2693
+    },
+    {
+      "epoch": 0.51831053660003,
+      "grad_norm": 0.39122752702358476,
+      "learning_rate": 9.890940722945428e-05,
+      "loss": 1.0617,
+      "step": 2694
+    },
+    {
+      "epoch": 0.5185029310085676,
+      "grad_norm": 0.39610199342283875,
+      "learning_rate": 9.884709032925273e-05,
+      "loss": 1.0784,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5186953254171051,
+      "grad_norm": 0.3397739407149826,
+      "learning_rate": 9.878477387682803e-05,
+      "loss": 0.9895,
+      "step": 2696
+    },
+    {
+      "epoch": 0.5188877198256425,
+      "grad_norm": 0.3009485489557237,
+      "learning_rate": 9.872245789638308e-05,
+      "loss": 1.0658,
+      "step": 2697
+    },
+    {
+      "epoch": 0.5190801142341801,
+      "grad_norm": 0.2802731105031589,
+      "learning_rate": 9.866014241212079e-05,
+      "loss": 1.0106,
+      "step": 2698
+    },
+    {
+      "epoch": 0.5192725086427176,
+      "grad_norm": 0.2824960458523637,
+      "learning_rate": 9.859782744824376e-05,
+      "loss": 1.0747,
+      "step": 2699
+    },
+    {
+      "epoch": 0.519464903051255,
+      "grad_norm": 0.2927205430549778,
+      "learning_rate": 9.85355130289544e-05,
+      "loss": 1.0738,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5196572974597926,
+      "grad_norm": 0.32586985138104474,
+      "learning_rate": 9.84731991784549e-05,
+      "loss": 1.0894,
+      "step": 2701
+    },
+    {
+      "epoch": 0.5198496918683301,
+      "grad_norm": 0.2859626837060444,
+      "learning_rate": 9.841088592094725e-05,
+      "loss": 1.0044,
+      "step": 2702
+    },
+    {
+      "epoch": 0.5200420862768675,
+      "grad_norm": 0.3431843579012332,
+      "learning_rate": 9.834857328063316e-05,
+      "loss": 1.0482,
+      "step": 2703
+    },
+    {
+      "epoch": 0.5202344806854051,
+      "grad_norm": 0.32562736513819074,
+      "learning_rate": 9.828626128171421e-05,
+      "loss": 1.0656,
+      "step": 2704
+    },
+    {
+      "epoch": 0.5204268750939426,
+      "grad_norm": 0.2618859474564516,
+      "learning_rate": 9.822394994839165e-05,
+      "loss": 1.087,
+      "step": 2705
+    },
+    {
+      "epoch": 0.52061926950248,
+      "grad_norm": 0.27248647838384255,
+      "learning_rate": 9.816163930486642e-05,
+      "loss": 1.0671,
+      "step": 2706
+    },
+    {
+      "epoch": 0.5208116639110176,
+      "grad_norm": 0.3484565938222194,
+      "learning_rate": 9.809932937533934e-05,
+      "loss": 1.1142,
+      "step": 2707
+    },
+    {
+      "epoch": 0.5210040583195551,
+      "grad_norm": 0.2768075250068996,
+      "learning_rate": 9.803702018401083e-05,
+      "loss": 1.0704,
+      "step": 2708
+    },
+    {
+      "epoch": 0.5211964527280926,
+      "grad_norm": 0.3144587533028435,
+      "learning_rate": 9.7974711755081e-05,
+      "loss": 1.0815,
+      "step": 2709
+    },
+    {
+      "epoch": 0.5213888471366301,
+      "grad_norm": 0.27773867743775527,
+      "learning_rate": 9.791240411274982e-05,
+      "loss": 1.0781,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5215812415451676,
+      "grad_norm": 0.3125349683222927,
+      "learning_rate": 9.785009728121686e-05,
+      "loss": 1.0468,
+      "step": 2711
+    },
+    {
+      "epoch": 0.5217736359537051,
+      "grad_norm": 0.3185643889770487,
+      "learning_rate": 9.778779128468132e-05,
+      "loss": 1.0691,
+      "step": 2712
+    },
+    {
+      "epoch": 0.5219660303622427,
+      "grad_norm": 0.2479368912483558,
+      "learning_rate": 9.772548614734217e-05,
+      "loss": 1.0711,
+      "step": 2713
+    },
+    {
+      "epoch": 0.5221584247707801,
+      "grad_norm": 0.36081129278836305,
+      "learning_rate": 9.766318189339797e-05,
+      "loss": 1.0686,
+      "step": 2714
+    },
+    {
+      "epoch": 0.5223508191793176,
+      "grad_norm": 0.28520848572758933,
+      "learning_rate": 9.760087854704698e-05,
+      "loss": 1.0164,
+      "step": 2715
+    },
+    {
+      "epoch": 0.522543213587855,
+      "grad_norm": 0.33016854865971146,
+      "learning_rate": 9.753857613248714e-05,
+      "loss": 1.1244,
+      "step": 2716
+    },
+    {
+      "epoch": 0.5227356079963926,
+      "grad_norm": 0.3292921902789938,
+      "learning_rate": 9.747627467391596e-05,
+      "loss": 1.0906,
+      "step": 2717
+    },
+    {
+      "epoch": 0.5229280024049301,
+      "grad_norm": 0.29538709272466723,
+      "learning_rate": 9.741397419553063e-05,
+      "loss": 1.05,
+      "step": 2718
+    },
+    {
+      "epoch": 0.5231203968134676,
+      "grad_norm": 0.3110439058725263,
+      "learning_rate": 9.735167472152791e-05,
+      "loss": 0.9908,
+      "step": 2719
+    },
+    {
+      "epoch": 0.5233127912220051,
+      "grad_norm": 0.3158559888079872,
+      "learning_rate": 9.728937627610424e-05,
+      "loss": 1.0779,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5235051856305426,
+      "grad_norm": 0.286200450977235,
+      "learning_rate": 9.722707888345552e-05,
+      "loss": 1.0948,
+      "step": 2721
+    },
+    {
+      "epoch": 0.5236975800390801,
+      "grad_norm": 0.28452082130062206,
+      "learning_rate": 9.716478256777747e-05,
+      "loss": 1.0167,
+      "step": 2722
+    },
+    {
+      "epoch": 0.5238899744476176,
+      "grad_norm": 0.2979739181344111,
+      "learning_rate": 9.710248735326519e-05,
+      "loss": 1.0223,
+      "step": 2723
+    },
+    {
+      "epoch": 0.5240823688561551,
+      "grad_norm": 0.33157498620222703,
+      "learning_rate": 9.704019326411343e-05,
+      "loss": 1.0609,
+      "step": 2724
+    },
+    {
+      "epoch": 0.5242747632646926,
+      "grad_norm": 0.3655860162714879,
+      "learning_rate": 9.697790032451651e-05,
+      "loss": 1.0378,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5244671576732302,
+      "grad_norm": 0.3881593466629942,
+      "learning_rate": 9.691560855866826e-05,
+      "loss": 1.1249,
+      "step": 2726
+    },
+    {
+      "epoch": 0.5246595520817676,
+      "grad_norm": 0.301132490633845,
+      "learning_rate": 9.685331799076209e-05,
+      "loss": 1.0453,
+      "step": 2727
+    },
+    {
+      "epoch": 0.5248519464903051,
+      "grad_norm": 0.26668304761279166,
+      "learning_rate": 9.679102864499099e-05,
+      "loss": 1.1148,
+      "step": 2728
+    },
+    {
+      "epoch": 0.5250443408988427,
+      "grad_norm": 0.3376845559654377,
+      "learning_rate": 9.672874054554738e-05,
+      "loss": 1.1388,
+      "step": 2729
+    },
+    {
+      "epoch": 0.5252367353073801,
+      "grad_norm": 0.33459083057765177,
+      "learning_rate": 9.666645371662324e-05,
+      "loss": 1.1106,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5254291297159176,
+      "grad_norm": 0.26500515784677214,
+      "learning_rate": 9.660416818241007e-05,
+      "loss": 1.035,
+      "step": 2731
+    },
+    {
+      "epoch": 0.5256215241244552,
+      "grad_norm": 0.33717677608962726,
+      "learning_rate": 9.654188396709882e-05,
+      "loss": 1.1347,
+      "step": 2732
+    },
+    {
+      "epoch": 0.5258139185329926,
+      "grad_norm": 0.3369284977260401,
+      "learning_rate": 9.647960109488003e-05,
+      "loss": 1.1271,
+      "step": 2733
+    },
+    {
+      "epoch": 0.5260063129415301,
+      "grad_norm": 0.311167262364201,
+      "learning_rate": 9.641731958994359e-05,
+      "loss": 1.0501,
+      "step": 2734
+    },
+    {
+      "epoch": 0.5261987073500677,
+      "grad_norm": 0.27073114029613154,
+      "learning_rate": 9.635503947647894e-05,
+      "loss": 1.1564,
+      "step": 2735
+    },
+    {
+      "epoch": 0.5263911017586052,
+      "grad_norm": 0.3227954134919288,
+      "learning_rate": 9.629276077867497e-05,
+      "loss": 1.0419,
+      "step": 2736
+    },
+    {
+      "epoch": 0.5265834961671426,
+      "grad_norm": 0.29174345704097504,
+      "learning_rate": 9.623048352071998e-05,
+      "loss": 1.0503,
+      "step": 2737
+    },
+    {
+      "epoch": 0.5267758905756802,
+      "grad_norm": 0.30989115194163197,
+      "learning_rate": 9.616820772680173e-05,
+      "loss": 1.031,
+      "step": 2738
+    },
+    {
+      "epoch": 0.5269682849842177,
+      "grad_norm": 0.30412929310668163,
+      "learning_rate": 9.610593342110746e-05,
+      "loss": 1.0234,
+      "step": 2739
+    },
+    {
+      "epoch": 0.5271606793927551,
+      "grad_norm": 0.2926047636961451,
+      "learning_rate": 9.604366062782381e-05,
+      "loss": 1.0525,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5273530738012927,
+      "grad_norm": 0.4168850642209728,
+      "learning_rate": 9.598138937113676e-05,
+      "loss": 0.9568,
+      "step": 2741
+    },
+    {
+      "epoch": 0.5275454682098302,
+      "grad_norm": 0.3078878204238304,
+      "learning_rate": 9.591911967523179e-05,
+      "loss": 1.0086,
+      "step": 2742
+    },
+    {
+      "epoch": 0.5277378626183676,
+      "grad_norm": 0.24207485728750425,
+      "learning_rate": 9.585685156429369e-05,
+      "loss": 1.0984,
+      "step": 2743
+    },
+    {
+      "epoch": 0.5279302570269051,
+      "grad_norm": 0.3322540073081415,
+      "learning_rate": 9.579458506250669e-05,
+      "loss": 1.0875,
+      "step": 2744
+    },
+    {
+      "epoch": 0.5281226514354427,
+      "grad_norm": 0.26985254397003916,
+      "learning_rate": 9.573232019405441e-05,
+      "loss": 1.0612,
+      "step": 2745
+    },
+    {
+      "epoch": 0.5283150458439801,
+      "grad_norm": 0.27146483728943643,
+      "learning_rate": 9.567005698311981e-05,
+      "loss": 1.1319,
+      "step": 2746
+    },
+    {
+      "epoch": 0.5285074402525176,
+      "grad_norm": 0.31292999571775915,
+      "learning_rate": 9.560779545388517e-05,
+      "loss": 1.0504,
+      "step": 2747
+    },
+    {
+      "epoch": 0.5286998346610552,
+      "grad_norm": 0.382001753843954,
+      "learning_rate": 9.554553563053215e-05,
+      "loss": 1.044,
+      "step": 2748
+    },
+    {
+      "epoch": 0.5288922290695927,
+      "grad_norm": 0.315149746667852,
+      "learning_rate": 9.54832775372418e-05,
+      "loss": 1.0604,
+      "step": 2749
+    },
+    {
+      "epoch": 0.5290846234781301,
+      "grad_norm": 0.3229981866016716,
+      "learning_rate": 9.542102119819434e-05,
+      "loss": 1.1197,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5292770178866677,
+      "grad_norm": 0.3187424846693923,
+      "learning_rate": 9.535876663756954e-05,
+      "loss": 1.078,
+      "step": 2751
+    },
+    {
+      "epoch": 0.5294694122952052,
+      "grad_norm": 0.3615588509829996,
+      "learning_rate": 9.529651387954627e-05,
+      "loss": 1.0651,
+      "step": 2752
+    },
+    {
+      "epoch": 0.5296618067037426,
+      "grad_norm": 0.31526782514947366,
+      "learning_rate": 9.523426294830284e-05,
+      "loss": 1.0682,
+      "step": 2753
+    },
+    {
+      "epoch": 0.5298542011122802,
+      "grad_norm": 0.28015305728954015,
+      "learning_rate": 9.517201386801675e-05,
+      "loss": 1.1625,
+      "step": 2754
+    },
+    {
+      "epoch": 0.5300465955208177,
+      "grad_norm": 0.27640149620410465,
+      "learning_rate": 9.510976666286484e-05,
+      "loss": 1.1234,
+      "step": 2755
+    },
+    {
+      "epoch": 0.5302389899293551,
+      "grad_norm": 0.2843590225786937,
+      "learning_rate": 9.504752135702317e-05,
+      "loss": 1.0188,
+      "step": 2756
+    },
+    {
+      "epoch": 0.5304313843378927,
+      "grad_norm": 0.3227943181420178,
+      "learning_rate": 9.498527797466718e-05,
+      "loss": 1.0053,
+      "step": 2757
+    },
+    {
+      "epoch": 0.5306237787464302,
+      "grad_norm": 0.3720956893322388,
+      "learning_rate": 9.492303653997146e-05,
+      "loss": 1.0461,
+      "step": 2758
+    },
+    {
+      "epoch": 0.5308161731549677,
+      "grad_norm": 0.34037755114619667,
+      "learning_rate": 9.48607970771098e-05,
+      "loss": 1.0244,
+      "step": 2759
+    },
+    {
+      "epoch": 0.5310085675635052,
+      "grad_norm": 0.3133560430418856,
+      "learning_rate": 9.479855961025536e-05,
+      "loss": 1.0927,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5312009619720427,
+      "grad_norm": 0.36269324634286576,
+      "learning_rate": 9.473632416358044e-05,
+      "loss": 1.0558,
+      "step": 2761
+    },
+    {
+      "epoch": 0.5313933563805802,
+      "grad_norm": 0.2903963671912293,
+      "learning_rate": 9.467409076125652e-05,
+      "loss": 1.1751,
+      "step": 2762
+    },
+    {
+      "epoch": 0.5315857507891177,
+      "grad_norm": 0.3160080584023893,
+      "learning_rate": 9.461185942745443e-05,
+      "loss": 1.0913,
+      "step": 2763
+    },
+    {
+      "epoch": 0.5317781451976552,
+      "grad_norm": 0.33624050187679827,
+      "learning_rate": 9.454963018634402e-05,
+      "loss": 0.8959,
+      "step": 2764
+    },
+    {
+      "epoch": 0.5319705396061927,
+      "grad_norm": 0.30220024132171014,
+      "learning_rate": 9.448740306209446e-05,
+      "loss": 1.0349,
+      "step": 2765
+    },
+    {
+      "epoch": 0.5321629340147302,
+      "grad_norm": 0.27063084653526037,
+      "learning_rate": 9.442517807887403e-05,
+      "loss": 1.0956,
+      "step": 2766
+    },
+    {
+      "epoch": 0.5323553284232677,
+      "grad_norm": 0.30000972601986187,
+      "learning_rate": 9.436295526085015e-05,
+      "loss": 1.1024,
+      "step": 2767
+    },
+    {
+      "epoch": 0.5325477228318052,
+      "grad_norm": 0.30672501398732777,
+      "learning_rate": 9.430073463218951e-05,
+      "loss": 0.9967,
+      "step": 2768
+    },
+    {
+      "epoch": 0.5327401172403426,
+      "grad_norm": 0.29709113816444055,
+      "learning_rate": 9.42385162170579e-05,
+      "loss": 0.9671,
+      "step": 2769
+    },
+    {
+      "epoch": 0.5329325116488802,
+      "grad_norm": 0.297654727456846,
+      "learning_rate": 9.417630003962018e-05,
+      "loss": 1.0928,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5331249060574177,
+      "grad_norm": 0.3028868168277822,
+      "learning_rate": 9.411408612404042e-05,
+      "loss": 1.0585,
+      "step": 2771
+    },
+    {
+      "epoch": 0.5333173004659552,
+      "grad_norm": 0.3616804117522196,
+      "learning_rate": 9.405187449448179e-05,
+      "loss": 1.0256,
+      "step": 2772
+    },
+    {
+      "epoch": 0.5335096948744927,
+      "grad_norm": 0.3291472649954265,
+      "learning_rate": 9.398966517510653e-05,
+      "loss": 1.1663,
+      "step": 2773
+    },
+    {
+      "epoch": 0.5337020892830302,
+      "grad_norm": 0.29427033422444704,
+      "learning_rate": 9.392745819007611e-05,
+      "loss": 1.0456,
+      "step": 2774
+    },
+    {
+      "epoch": 0.5338944836915677,
+      "grad_norm": 0.29669086798730604,
+      "learning_rate": 9.386525356355094e-05,
+      "loss": 1.0459,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5340868781001052,
+      "grad_norm": 0.31207769708090366,
+      "learning_rate": 9.38030513196906e-05,
+      "loss": 1.0466,
+      "step": 2776
+    },
+    {
+      "epoch": 0.5342792725086427,
+      "grad_norm": 0.37966080780743977,
+      "learning_rate": 9.37408514826537e-05,
+      "loss": 1.0014,
+      "step": 2777
+    },
+    {
+      "epoch": 0.5344716669171802,
+      "grad_norm": 0.26385790450294977,
+      "learning_rate": 9.3678654076598e-05,
+      "loss": 1.0082,
+      "step": 2778
+    },
+    {
+      "epoch": 0.5346640613257178,
+      "grad_norm": 0.3218037031651044,
+      "learning_rate": 9.361645912568016e-05,
+      "loss": 1.0125,
+      "step": 2779
+    },
+    {
+      "epoch": 0.5348564557342552,
+      "grad_norm": 0.30518090575277684,
+      "learning_rate": 9.355426665405606e-05,
+      "loss": 1.0402,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5350488501427927,
+      "grad_norm": 0.3041275829512792,
+      "learning_rate": 9.349207668588052e-05,
+      "loss": 1.0473,
+      "step": 2781
+    },
+    {
+      "epoch": 0.5352412445513303,
+      "grad_norm": 0.2707767455349892,
+      "learning_rate": 9.342988924530742e-05,
+      "loss": 1.1194,
+      "step": 2782
+    },
+    {
+      "epoch": 0.5354336389598677,
+      "grad_norm": 0.2970468228020865,
+      "learning_rate": 9.336770435648964e-05,
+      "loss": 1.0255,
+      "step": 2783
+    },
+    {
+      "epoch": 0.5356260333684052,
+      "grad_norm": 0.31351423831171826,
+      "learning_rate": 9.330552204357904e-05,
+      "loss": 1.0481,
+      "step": 2784
+    },
+    {
+      "epoch": 0.5358184277769428,
+      "grad_norm": 0.3845474184550142,
+      "learning_rate": 9.324334233072648e-05,
+      "loss": 0.9651,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5360108221854802,
+      "grad_norm": 0.281176962524045,
+      "learning_rate": 9.318116524208198e-05,
+      "loss": 1.0951,
+      "step": 2786
+    },
+    {
+      "epoch": 0.5362032165940177,
+      "grad_norm": 0.2914111318352684,
+      "learning_rate": 9.311899080179433e-05,
+      "loss": 1.0827,
+      "step": 2787
+    },
+    {
+      "epoch": 0.5363956110025553,
+      "grad_norm": 0.3564120679051981,
+      "learning_rate": 9.305681903401132e-05,
+      "loss": 1.0194,
+      "step": 2788
+    },
+    {
+      "epoch": 0.5365880054110927,
+      "grad_norm": 0.24276411649445256,
+      "learning_rate": 9.299464996287983e-05,
+      "loss": 1.0857,
+      "step": 2789
+    },
+    {
+      "epoch": 0.5367803998196302,
+      "grad_norm": 0.30739416618655285,
+      "learning_rate": 9.293248361254556e-05,
+      "loss": 0.9993,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5369727942281678,
+      "grad_norm": 0.4960691729042597,
+      "learning_rate": 9.287032000715318e-05,
+      "loss": 1.0502,
+      "step": 2791
+    },
+    {
+      "epoch": 0.5371651886367053,
+      "grad_norm": 0.31124029201715964,
+      "learning_rate": 9.28081591708464e-05,
+      "loss": 1.0489,
+      "step": 2792
+    },
+    {
+      "epoch": 0.5373575830452427,
+      "grad_norm": 0.2740784549147511,
+      "learning_rate": 9.274600112776769e-05,
+      "loss": 1.0375,
+      "step": 2793
+    },
+    {
+      "epoch": 0.5375499774537803,
+      "grad_norm": 0.27141699117857976,
+      "learning_rate": 9.268384590205858e-05,
+      "loss": 1.0799,
+      "step": 2794
+    },
+    {
+      "epoch": 0.5377423718623178,
+      "grad_norm": 0.32863385363595055,
+      "learning_rate": 9.262169351785943e-05,
+      "loss": 1.042,
+      "step": 2795
+    },
+    {
+      "epoch": 0.5379347662708552,
+      "grad_norm": 0.3299091517342768,
+      "learning_rate": 9.255954399930947e-05,
+      "loss": 0.9722,
+      "step": 2796
+    },
+    {
+      "epoch": 0.5381271606793927,
+      "grad_norm": 0.32779796428934976,
+      "learning_rate": 9.249739737054686e-05,
+      "loss": 1.135,
+      "step": 2797
+    },
+    {
+      "epoch": 0.5383195550879303,
+      "grad_norm": 0.31454421970021534,
+      "learning_rate": 9.24352536557087e-05,
+      "loss": 1.057,
+      "step": 2798
+    },
+    {
+      "epoch": 0.5385119494964677,
+      "grad_norm": 0.26750739049993183,
+      "learning_rate": 9.237311287893086e-05,
+      "loss": 1.0418,
+      "step": 2799
+    },
+    {
+      "epoch": 0.5387043439050052,
+      "grad_norm": 0.2731401553352244,
+      "learning_rate": 9.231097506434807e-05,
+      "loss": 1.0911,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5388967383135428,
+      "grad_norm": 0.29354989195287307,
+      "learning_rate": 9.224884023609397e-05,
+      "loss": 1.0442,
+      "step": 2801
+    },
+    {
+      "epoch": 0.5390891327220803,
+      "grad_norm": 0.35260721010315077,
+      "learning_rate": 9.218670841830098e-05,
+      "loss": 0.9677,
+      "step": 2802
+    },
+    {
+      "epoch": 0.5392815271306177,
+      "grad_norm": 0.2802246546142186,
+      "learning_rate": 9.212457963510044e-05,
+      "loss": 1.1106,
+      "step": 2803
+    },
+    {
+      "epoch": 0.5394739215391553,
+      "grad_norm": 0.3358029617604148,
+      "learning_rate": 9.206245391062244e-05,
+      "loss": 1.1165,
+      "step": 2804
+    },
+    {
+      "epoch": 0.5396663159476928,
+      "grad_norm": 0.3166541082824362,
+      "learning_rate": 9.200033126899584e-05,
+      "loss": 1.043,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5398587103562302,
+      "grad_norm": 0.30795967955353276,
+      "learning_rate": 9.193821173434842e-05,
+      "loss": 1.0437,
+      "step": 2806
+    },
+    {
+      "epoch": 0.5400511047647678,
+      "grad_norm": 0.3327238498922548,
+      "learning_rate": 9.187609533080667e-05,
+      "loss": 1.0617,
+      "step": 2807
+    },
+    {
+      "epoch": 0.5402434991733053,
+      "grad_norm": 0.36117726524740024,
+      "learning_rate": 9.181398208249583e-05,
+      "loss": 1.0449,
+      "step": 2808
+    },
+    {
+      "epoch": 0.5404358935818427,
+      "grad_norm": 0.2822965590599794,
+      "learning_rate": 9.175187201354004e-05,
+      "loss": 1.0556,
+      "step": 2809
+    },
+    {
+      "epoch": 0.5406282879903803,
+      "grad_norm": 0.34551883558973434,
+      "learning_rate": 9.168976514806216e-05,
+      "loss": 1.0436,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5408206823989178,
+      "grad_norm": 0.3276728211785904,
+      "learning_rate": 9.162766151018372e-05,
+      "loss": 1.0218,
+      "step": 2811
+    },
+    {
+      "epoch": 0.5410130768074553,
+      "grad_norm": 0.2809564264393135,
+      "learning_rate": 9.156556112402507e-05,
+      "loss": 1.1238,
+      "step": 2812
+    },
+    {
+      "epoch": 0.5412054712159928,
+      "grad_norm": 0.32923288760595176,
+      "learning_rate": 9.150346401370527e-05,
+      "loss": 1.0828,
+      "step": 2813
+    },
+    {
+      "epoch": 0.5413978656245303,
+      "grad_norm": 0.30982590545276434,
+      "learning_rate": 9.144137020334214e-05,
+      "loss": 1.1131,
+      "step": 2814
+    },
+    {
+      "epoch": 0.5415902600330678,
+      "grad_norm": 0.2968755265338655,
+      "learning_rate": 9.137927971705222e-05,
+      "loss": 1.008,
+      "step": 2815
+    },
+    {
+      "epoch": 0.5417826544416053,
+      "grad_norm": 0.2661602971152169,
+      "learning_rate": 9.131719257895074e-05,
+      "loss": 1.1025,
+      "step": 2816
+    },
+    {
+      "epoch": 0.5419750488501428,
+      "grad_norm": 0.31899816939093295,
+      "learning_rate": 9.125510881315158e-05,
+      "loss": 1.0911,
+      "step": 2817
+    },
+    {
+      "epoch": 0.5421674432586803,
+      "grad_norm": 0.31194303994512873,
+      "learning_rate": 9.119302844376741e-05,
+      "loss": 1.0762,
+      "step": 2818
+    },
+    {
+      "epoch": 0.5423598376672178,
+      "grad_norm": 0.30479570213560897,
+      "learning_rate": 9.113095149490952e-05,
+      "loss": 1.0359,
+      "step": 2819
+    },
+    {
+      "epoch": 0.5425522320757553,
+      "grad_norm": 0.3249136726390395,
+      "learning_rate": 9.106887799068782e-05,
+      "loss": 1.0743,
+      "step": 2820
+    },
+    {
+      "epoch": 0.5427446264842928,
+      "grad_norm": 0.3636120345725698,
+      "learning_rate": 9.100680795521104e-05,
+      "loss": 1.0495,
+      "step": 2821
+    },
+    {
+      "epoch": 0.5429370208928304,
+      "grad_norm": 0.2848339273714078,
+      "learning_rate": 9.09447414125864e-05,
+      "loss": 1.0268,
+      "step": 2822
+    },
+    {
+      "epoch": 0.5431294153013678,
+      "grad_norm": 0.2857247053117168,
+      "learning_rate": 9.088267838691986e-05,
+      "loss": 1.0178,
+      "step": 2823
+    },
+    {
+      "epoch": 0.5433218097099053,
+      "grad_norm": 0.3487514775644445,
+      "learning_rate": 9.0820618902316e-05,
+      "loss": 1.0549,
+      "step": 2824
+    },
+    {
+      "epoch": 0.5435142041184428,
+      "grad_norm": 0.3300207026944498,
+      "learning_rate": 9.075856298287795e-05,
+      "loss": 0.943,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5437065985269803,
+      "grad_norm": 0.311900787783297,
+      "learning_rate": 9.069651065270752e-05,
+      "loss": 1.0415,
+      "step": 2826
+    },
+    {
+      "epoch": 0.5438989929355178,
+      "grad_norm": 0.2603950382145914,
+      "learning_rate": 9.06344619359052e-05,
+      "loss": 1.0866,
+      "step": 2827
+    },
+    {
+      "epoch": 0.5440913873440553,
+      "grad_norm": 0.33786791580426334,
+      "learning_rate": 9.057241685656995e-05,
+      "loss": 1.0252,
+      "step": 2828
+    },
+    {
+      "epoch": 0.5442837817525928,
+      "grad_norm": 0.37359857562233495,
+      "learning_rate": 9.051037543879932e-05,
+      "loss": 1.003,
+      "step": 2829
+    },
+    {
+      "epoch": 0.5444761761611303,
+      "grad_norm": 0.2969417967748203,
+      "learning_rate": 9.044833770668956e-05,
+      "loss": 1.1049,
+      "step": 2830
+    },
+    {
+      "epoch": 0.5446685705696678,
+      "grad_norm": 0.31430576504033775,
+      "learning_rate": 9.038630368433535e-05,
+      "loss": 1.0444,
+      "step": 2831
+    },
+    {
+      "epoch": 0.5448609649782054,
+      "grad_norm": 0.31783943377008056,
+      "learning_rate": 9.032427339583e-05,
+      "loss": 1.1372,
+      "step": 2832
+    },
+    {
+      "epoch": 0.5450533593867428,
+      "grad_norm": 0.2957587310922381,
+      "learning_rate": 9.026224686526538e-05,
+      "loss": 1.0809,
+      "step": 2833
+    },
+    {
+      "epoch": 0.5452457537952803,
+      "grad_norm": 0.27822396836129787,
+      "learning_rate": 9.020022411673187e-05,
+      "loss": 1.031,
+      "step": 2834
+    },
+    {
+      "epoch": 0.5454381482038179,
+      "grad_norm": 0.300938603064847,
+      "learning_rate": 9.01382051743184e-05,
+      "loss": 1.1484,
+      "step": 2835
+    },
+    {
+      "epoch": 0.5456305426123553,
+      "grad_norm": 0.31966311928483815,
+      "learning_rate": 9.007619006211241e-05,
+      "loss": 1.079,
+      "step": 2836
+    },
+    {
+      "epoch": 0.5458229370208928,
+      "grad_norm": 0.2749902856625382,
+      "learning_rate": 9.00141788041998e-05,
+      "loss": 1.0762,
+      "step": 2837
+    },
+    {
+      "epoch": 0.5460153314294304,
+      "grad_norm": 0.3276725006229355,
+      "learning_rate": 8.99521714246651e-05,
+      "loss": 1.012,
+      "step": 2838
+    },
+    {
+      "epoch": 0.5462077258379678,
+      "grad_norm": 0.35701775296019184,
+      "learning_rate": 8.989016794759127e-05,
+      "loss": 1.1177,
+      "step": 2839
+    },
+    {
+      "epoch": 0.5464001202465053,
+      "grad_norm": 0.34521768665052943,
+      "learning_rate": 8.982816839705969e-05,
+      "loss": 1.0087,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5465925146550429,
+      "grad_norm": 0.31296761727600764,
+      "learning_rate": 8.976617279715031e-05,
+      "loss": 1.0251,
+      "step": 2841
+    },
+    {
+      "epoch": 0.5467849090635803,
+      "grad_norm": 0.2877284160127235,
+      "learning_rate": 8.970418117194146e-05,
+      "loss": 1.1764,
+      "step": 2842
+    },
+    {
+      "epoch": 0.5469773034721178,
+      "grad_norm": 0.3491685011984224,
+      "learning_rate": 8.964219354550999e-05,
+      "loss": 1.0731,
+      "step": 2843
+    },
+    {
+      "epoch": 0.5471696978806554,
+      "grad_norm": 0.33580397212952245,
+      "learning_rate": 8.958020994193124e-05,
+      "loss": 1.0638,
+      "step": 2844
+    },
+    {
+      "epoch": 0.5473620922891929,
+      "grad_norm": 0.3692386492957108,
+      "learning_rate": 8.951823038527888e-05,
+      "loss": 1.0258,
+      "step": 2845
+    },
+    {
+      "epoch": 0.5475544866977303,
+      "grad_norm": 0.2894896857350827,
+      "learning_rate": 8.945625489962503e-05,
+      "loss": 1.0558,
+      "step": 2846
+    },
+    {
+      "epoch": 0.5477468811062679,
+      "grad_norm": 0.2858192761845107,
+      "learning_rate": 8.93942835090403e-05,
+      "loss": 1.008,
+      "step": 2847
+    },
+    {
+      "epoch": 0.5479392755148054,
+      "grad_norm": 0.25297580804698744,
+      "learning_rate": 8.933231623759365e-05,
+      "loss": 1.0477,
+      "step": 2848
+    },
+    {
+      "epoch": 0.5481316699233428,
+      "grad_norm": 0.2966275998196543,
+      "learning_rate": 8.927035310935242e-05,
+      "loss": 1.0029,
+      "step": 2849
+    },
+    {
+      "epoch": 0.5483240643318803,
+      "grad_norm": 0.4199080177543365,
+      "learning_rate": 8.920839414838242e-05,
+      "loss": 0.9477,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5485164587404179,
+      "grad_norm": 0.3665132119139109,
+      "learning_rate": 8.914643937874778e-05,
+      "loss": 1.0608,
+      "step": 2851
+    },
+    {
+      "epoch": 0.5487088531489553,
+      "grad_norm": 0.2966821211781953,
+      "learning_rate": 8.908448882451104e-05,
+      "loss": 0.9727,
+      "step": 2852
+    },
+    {
+      "epoch": 0.5489012475574928,
+      "grad_norm": 0.3102070609033994,
+      "learning_rate": 8.902254250973305e-05,
+      "loss": 1.0988,
+      "step": 2853
+    },
+    {
+      "epoch": 0.5490936419660304,
+      "grad_norm": 0.294385063859879,
+      "learning_rate": 8.896060045847304e-05,
+      "loss": 1.1261,
+      "step": 2854
+    },
+    {
+      "epoch": 0.5492860363745679,
+      "grad_norm": 0.30667231559898744,
+      "learning_rate": 8.889866269478859e-05,
+      "loss": 1.0785,
+      "step": 2855
+    },
+    {
+      "epoch": 0.5494784307831053,
+      "grad_norm": 0.3517437039855562,
+      "learning_rate": 8.883672924273566e-05,
+      "loss": 1.0161,
+      "step": 2856
+    },
+    {
+      "epoch": 0.5496708251916429,
+      "grad_norm": 0.37631058254100214,
+      "learning_rate": 8.877480012636847e-05,
+      "loss": 1.0885,
+      "step": 2857
+    },
+    {
+      "epoch": 0.5498632196001804,
+      "grad_norm": 0.33444593087880226,
+      "learning_rate": 8.871287536973953e-05,
+      "loss": 1.0089,
+      "step": 2858
+    },
+    {
+      "epoch": 0.5500556140087178,
+      "grad_norm": 0.29141826135560583,
+      "learning_rate": 8.865095499689977e-05,
+      "loss": 1.0822,
+      "step": 2859
+    },
+    {
+      "epoch": 0.5502480084172554,
+      "grad_norm": 0.37836689058252226,
+      "learning_rate": 8.858903903189831e-05,
+      "loss": 1.035,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5504404028257929,
+      "grad_norm": 0.2808594228189031,
+      "learning_rate": 8.852712749878254e-05,
+      "loss": 1.0318,
+      "step": 2861
+    },
+    {
+      "epoch": 0.5506327972343303,
+      "grad_norm": 0.31959706265304494,
+      "learning_rate": 8.846522042159832e-05,
+      "loss": 1.0232,
+      "step": 2862
+    },
+    {
+      "epoch": 0.5508251916428679,
+      "grad_norm": 0.2804531424986975,
+      "learning_rate": 8.840331782438953e-05,
+      "loss": 0.9697,
+      "step": 2863
+    },
+    {
+      "epoch": 0.5510175860514054,
+      "grad_norm": 0.2676175782294614,
+      "learning_rate": 8.834141973119849e-05,
+      "loss": 1.0522,
+      "step": 2864
+    },
+    {
+      "epoch": 0.5512099804599428,
+      "grad_norm": 0.2596985852011845,
+      "learning_rate": 8.82795261660657e-05,
+      "loss": 1.0758,
+      "step": 2865
+    },
+    {
+      "epoch": 0.5514023748684804,
+      "grad_norm": 0.41373696903188156,
+      "learning_rate": 8.821763715302987e-05,
+      "loss": 1.0527,
+      "step": 2866
+    },
+    {
+      "epoch": 0.5515947692770179,
+      "grad_norm": 0.320969070662106,
+      "learning_rate": 8.815575271612797e-05,
+      "loss": 1.0917,
+      "step": 2867
+    },
+    {
+      "epoch": 0.5517871636855554,
+      "grad_norm": 0.3547632523405083,
+      "learning_rate": 8.809387287939528e-05,
+      "loss": 1.0736,
+      "step": 2868
+    },
+    {
+      "epoch": 0.5519795580940929,
+      "grad_norm": 0.44054579481182676,
+      "learning_rate": 8.803199766686516e-05,
+      "loss": 1.1252,
+      "step": 2869
+    },
+    {
+      "epoch": 0.5521719525026304,
+      "grad_norm": 0.41091653661454736,
+      "learning_rate": 8.797012710256923e-05,
+      "loss": 1.0356,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5523643469111679,
+      "grad_norm": 0.2967478708848712,
+      "learning_rate": 8.790826121053733e-05,
+      "loss": 1.1003,
+      "step": 2871
+    },
+    {
+      "epoch": 0.5525567413197054,
+      "grad_norm": 0.29904638106311077,
+      "learning_rate": 8.784640001479741e-05,
+      "loss": 1.0463,
+      "step": 2872
+    },
+    {
+      "epoch": 0.5527491357282429,
+      "grad_norm": 0.3125859749474423,
+      "learning_rate": 8.77845435393757e-05,
+      "loss": 1.0916,
+      "step": 2873
+    },
+    {
+      "epoch": 0.5529415301367804,
+      "grad_norm": 0.37931912983022686,
+      "learning_rate": 8.772269180829653e-05,
+      "loss": 1.1105,
+      "step": 2874
+    },
+    {
+      "epoch": 0.553133924545318,
+      "grad_norm": 0.34877374229130115,
+      "learning_rate": 8.766084484558237e-05,
+      "loss": 1.077,
+      "step": 2875
+    },
+    {
+      "epoch": 0.5533263189538554,
+      "grad_norm": 0.26242106157551015,
+      "learning_rate": 8.759900267525391e-05,
+      "loss": 1.1204,
+      "step": 2876
+    },
+    {
+      "epoch": 0.5535187133623929,
+      "grad_norm": 0.2804202584959949,
+      "learning_rate": 8.753716532132992e-05,
+      "loss": 1.0686,
+      "step": 2877
+    },
+    {
+      "epoch": 0.5537111077709304,
+      "grad_norm": 0.3944097406678049,
+      "learning_rate": 8.747533280782726e-05,
+      "loss": 1.0112,
+      "step": 2878
+    },
+    {
+      "epoch": 0.5539035021794679,
+      "grad_norm": 0.2898151481611907,
+      "learning_rate": 8.741350515876104e-05,
+      "loss": 1.0187,
+      "step": 2879
+    },
+    {
+      "epoch": 0.5540958965880054,
+      "grad_norm": 0.2912261695515056,
+      "learning_rate": 8.735168239814439e-05,
+      "loss": 0.9711,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5542882909965429,
+      "grad_norm": 0.3013843622351798,
+      "learning_rate": 8.728986454998858e-05,
+      "loss": 0.9723,
+      "step": 2881
+    },
+    {
+      "epoch": 0.5544806854050804,
+      "grad_norm": 0.3256206939588518,
+      "learning_rate": 8.72280516383029e-05,
+      "loss": 0.9352,
+      "step": 2882
+    },
+    {
+      "epoch": 0.5546730798136179,
+      "grad_norm": 0.29571597280881323,
+      "learning_rate": 8.716624368709477e-05,
+      "loss": 1.0715,
+      "step": 2883
+    },
+    {
+      "epoch": 0.5548654742221554,
+      "grad_norm": 0.3016978852087846,
+      "learning_rate": 8.71044407203697e-05,
+      "loss": 1.0182,
+      "step": 2884
+    },
+    {
+      "epoch": 0.555057868630693,
+      "grad_norm": 0.2964638439991812,
+      "learning_rate": 8.704264276213129e-05,
+      "loss": 1.0537,
+      "step": 2885
+    },
+    {
+      "epoch": 0.5552502630392304,
+      "grad_norm": 0.2798235691934867,
+      "learning_rate": 8.698084983638111e-05,
+      "loss": 0.9875,
+      "step": 2886
+    },
+    {
+      "epoch": 0.5554426574477679,
+      "grad_norm": 0.3035192055389043,
+      "learning_rate": 8.691906196711884e-05,
+      "loss": 1.1402,
+      "step": 2887
+    },
+    {
+      "epoch": 0.5556350518563055,
+      "grad_norm": 0.330054918020496,
+      "learning_rate": 8.685727917834217e-05,
+      "loss": 1.0881,
+      "step": 2888
+    },
+    {
+      "epoch": 0.5558274462648429,
+      "grad_norm": 0.3268851139586036,
+      "learning_rate": 8.679550149404685e-05,
+      "loss": 1.0162,
+      "step": 2889
+    },
+    {
+      "epoch": 0.5560198406733804,
+      "grad_norm": 0.3106436762216314,
+      "learning_rate": 8.673372893822654e-05,
+      "loss": 1.0518,
+      "step": 2890
+    },
+    {
+      "epoch": 0.556212235081918,
+      "grad_norm": 0.28321711918575027,
+      "learning_rate": 8.667196153487307e-05,
+      "loss": 1.0468,
+      "step": 2891
+    },
+    {
+      "epoch": 0.5564046294904554,
+      "grad_norm": 0.3593566557919029,
+      "learning_rate": 8.661019930797614e-05,
+      "loss": 1.0154,
+      "step": 2892
+    },
+    {
+      "epoch": 0.5565970238989929,
+      "grad_norm": 0.25402597082912576,
+      "learning_rate": 8.654844228152355e-05,
+      "loss": 1.0784,
+      "step": 2893
+    },
+    {
+      "epoch": 0.5567894183075305,
+      "grad_norm": 0.39642091437982707,
+      "learning_rate": 8.648669047950097e-05,
+      "loss": 1.0931,
+      "step": 2894
+    },
+    {
+      "epoch": 0.556981812716068,
+      "grad_norm": 0.3282317294342435,
+      "learning_rate": 8.642494392589206e-05,
+      "loss": 1.0096,
+      "step": 2895
+    },
+    {
+      "epoch": 0.5571742071246054,
+      "grad_norm": 0.31148082446589526,
+      "learning_rate": 8.63632026446785e-05,
+      "loss": 1.0613,
+      "step": 2896
+    },
+    {
+      "epoch": 0.557366601533143,
+      "grad_norm": 0.35293748890427656,
+      "learning_rate": 8.630146665983992e-05,
+      "loss": 1.1124,
+      "step": 2897
+    },
+    {
+      "epoch": 0.5575589959416805,
+      "grad_norm": 0.2922389511492973,
+      "learning_rate": 8.623973599535385e-05,
+      "loss": 1.0646,
+      "step": 2898
+    },
+    {
+      "epoch": 0.5577513903502179,
+      "grad_norm": 0.25579634072324514,
+      "learning_rate": 8.617801067519574e-05,
+      "loss": 1.1038,
+      "step": 2899
+    },
+    {
+      "epoch": 0.5579437847587555,
+      "grad_norm": 0.34931617057055997,
+      "learning_rate": 8.611629072333904e-05,
+      "loss": 1.0979,
+      "step": 2900
+    },
+    {
+      "epoch": 0.558136179167293,
+      "grad_norm": 0.26884340317373434,
+      "learning_rate": 8.605457616375503e-05,
+      "loss": 1.0341,
+      "step": 2901
+    },
+    {
+      "epoch": 0.5583285735758304,
+      "grad_norm": 0.3298652036955405,
+      "learning_rate": 8.599286702041292e-05,
+      "loss": 0.9975,
+      "step": 2902
+    },
+    {
+      "epoch": 0.5585209679843679,
+      "grad_norm": 0.31408532841608927,
+      "learning_rate": 8.593116331727987e-05,
+      "loss": 1.0318,
+      "step": 2903
+    },
+    {
+      "epoch": 0.5587133623929055,
+      "grad_norm": 0.2922561135077166,
+      "learning_rate": 8.586946507832088e-05,
+      "loss": 1.0685,
+      "step": 2904
+    },
+    {
+      "epoch": 0.5589057568014429,
+      "grad_norm": 0.2765400696849115,
+      "learning_rate": 8.580777232749883e-05,
+      "loss": 1.0929,
+      "step": 2905
+    },
+    {
+      "epoch": 0.5590981512099804,
+      "grad_norm": 0.3356421625174065,
+      "learning_rate": 8.574608508877448e-05,
+      "loss": 1.0141,
+      "step": 2906
+    },
+    {
+      "epoch": 0.559290545618518,
+      "grad_norm": 0.49676670681708085,
+      "learning_rate": 8.568440338610638e-05,
+      "loss": 1.1265,
+      "step": 2907
+    },
+    {
+      "epoch": 0.5594829400270555,
+      "grad_norm": 0.31312977063260056,
+      "learning_rate": 8.562272724345108e-05,
+      "loss": 1.0335,
+      "step": 2908
+    },
+    {
+      "epoch": 0.5596753344355929,
+      "grad_norm": 0.38195926565004296,
+      "learning_rate": 8.556105668476286e-05,
+      "loss": 0.9858,
+      "step": 2909
+    },
+    {
+      "epoch": 0.5598677288441305,
+      "grad_norm": 0.3150881476193059,
+      "learning_rate": 8.549939173399385e-05,
+      "loss": 1.0554,
+      "step": 2910
+    },
+    {
+      "epoch": 0.560060123252668,
+      "grad_norm": 0.40396081695784297,
+      "learning_rate": 8.5437732415094e-05,
+      "loss": 1.0308,
+      "step": 2911
+    },
+    {
+      "epoch": 0.5602525176612054,
+      "grad_norm": 0.3041510627946518,
+      "learning_rate": 8.537607875201106e-05,
+      "loss": 1.1022,
+      "step": 2912
+    },
+    {
+      "epoch": 0.560444912069743,
+      "grad_norm": 0.29310168589718943,
+      "learning_rate": 8.531443076869059e-05,
+      "loss": 1.0168,
+      "step": 2913
+    },
+    {
+      "epoch": 0.5606373064782805,
+      "grad_norm": 0.29803076151224067,
+      "learning_rate": 8.525278848907604e-05,
+      "loss": 1.0347,
+      "step": 2914
+    },
+    {
+      "epoch": 0.5608297008868179,
+      "grad_norm": 0.3522448099447787,
+      "learning_rate": 8.519115193710849e-05,
+      "loss": 1.0471,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5610220952953555,
+      "grad_norm": 0.35125451333791413,
+      "learning_rate": 8.512952113672689e-05,
+      "loss": 1.1041,
+      "step": 2916
+    },
+    {
+      "epoch": 0.561214489703893,
+      "grad_norm": 0.33007617837125985,
+      "learning_rate": 8.506789611186794e-05,
+      "loss": 1.0859,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5614068841124304,
+      "grad_norm": 0.3282077574145357,
+      "learning_rate": 8.500627688646607e-05,
+      "loss": 1.0369,
+      "step": 2918
+    },
+    {
+      "epoch": 0.561599278520968,
+      "grad_norm": 0.3574013581671308,
+      "learning_rate": 8.494466348445345e-05,
+      "loss": 1.0112,
+      "step": 2919
+    },
+    {
+      "epoch": 0.5617916729295055,
+      "grad_norm": 0.32669518195491326,
+      "learning_rate": 8.48830559297601e-05,
+      "loss": 1.0256,
+      "step": 2920
+    },
+    {
+      "epoch": 0.561984067338043,
+      "grad_norm": 0.34915107762167114,
+      "learning_rate": 8.48214542463136e-05,
+      "loss": 1.0796,
+      "step": 2921
+    },
+    {
+      "epoch": 0.5621764617465805,
+      "grad_norm": 0.34200888221421405,
+      "learning_rate": 8.475985845803939e-05,
+      "loss": 1.1363,
+      "step": 2922
+    },
+    {
+      "epoch": 0.562368856155118,
+      "grad_norm": 0.3233520630722986,
+      "learning_rate": 8.469826858886054e-05,
+      "loss": 1.0679,
+      "step": 2923
+    },
+    {
+      "epoch": 0.5625612505636555,
+      "grad_norm": 0.3220418829222748,
+      "learning_rate": 8.463668466269784e-05,
+      "loss": 1.1137,
+      "step": 2924
+    },
+    {
+      "epoch": 0.562753644972193,
+      "grad_norm": 0.3493961359395474,
+      "learning_rate": 8.457510670346976e-05,
+      "loss": 1.0761,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5629460393807305,
+      "grad_norm": 0.2522171366067136,
+      "learning_rate": 8.451353473509253e-05,
+      "loss": 1.114,
+      "step": 2926
+    },
+    {
+      "epoch": 0.563138433789268,
+      "grad_norm": 0.42238325404920435,
+      "learning_rate": 8.445196878147996e-05,
+      "loss": 1.0114,
+      "step": 2927
+    },
+    {
+      "epoch": 0.5633308281978056,
+      "grad_norm": 0.30275201960687115,
+      "learning_rate": 8.439040886654355e-05,
+      "loss": 1.1037,
+      "step": 2928
+    },
+    {
+      "epoch": 0.563523222606343,
+      "grad_norm": 0.33375987212409763,
+      "learning_rate": 8.432885501419247e-05,
+      "loss": 1.0279,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5637156170148805,
+      "grad_norm": 0.3221588554133989,
+      "learning_rate": 8.426730724833354e-05,
+      "loss": 1.1009,
+      "step": 2930
+    },
+    {
+      "epoch": 0.563908011423418,
+      "grad_norm": 0.37092319935498647,
+      "learning_rate": 8.420576559287112e-05,
+      "loss": 1.1381,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5641004058319555,
+      "grad_norm": 0.2897348933751357,
+      "learning_rate": 8.414423007170741e-05,
+      "loss": 1.0954,
+      "step": 2932
+    },
+    {
+      "epoch": 0.564292800240493,
+      "grad_norm": 0.3659774461234699,
+      "learning_rate": 8.4082700708742e-05,
+      "loss": 1.0984,
+      "step": 2933
+    },
+    {
+      "epoch": 0.5644851946490305,
+      "grad_norm": 0.46693933336326643,
+      "learning_rate": 8.402117752787226e-05,
+      "loss": 1.0542,
+      "step": 2934
+    },
+    {
+      "epoch": 0.564677589057568,
+      "grad_norm": 0.32568646658369843,
+      "learning_rate": 8.395966055299303e-05,
+      "loss": 1.0483,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5648699834661055,
+      "grad_norm": 0.3442078687631252,
+      "learning_rate": 8.389814980799678e-05,
+      "loss": 1.0723,
+      "step": 2936
+    },
+    {
+      "epoch": 0.565062377874643,
+      "grad_norm": 0.3472095691264837,
+      "learning_rate": 8.38366453167736e-05,
+      "loss": 0.9716,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5652547722831806,
+      "grad_norm": 0.27964568766938036,
+      "learning_rate": 8.377514710321117e-05,
+      "loss": 1.1544,
+      "step": 2938
+    },
+    {
+      "epoch": 0.565447166691718,
+      "grad_norm": 0.3408981584310334,
+      "learning_rate": 8.371365519119462e-05,
+      "loss": 1.131,
+      "step": 2939
+    },
+    {
+      "epoch": 0.5656395611002555,
+      "grad_norm": 0.2911958386346275,
+      "learning_rate": 8.365216960460676e-05,
+      "loss": 1.0964,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5658319555087931,
+      "grad_norm": 0.3073056917380063,
+      "learning_rate": 8.35906903673278e-05,
+      "loss": 1.0181,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5660243499173305,
+      "grad_norm": 0.2922271104890582,
+      "learning_rate": 8.352921750323562e-05,
+      "loss": 1.0217,
+      "step": 2942
+    },
+    {
+      "epoch": 0.566216744325868,
+      "grad_norm": 0.3290922456340551,
+      "learning_rate": 8.346775103620559e-05,
+      "loss": 1.0823,
+      "step": 2943
+    },
+    {
+      "epoch": 0.5664091387344056,
+      "grad_norm": 0.2818427572569486,
+      "learning_rate": 8.340629099011057e-05,
+      "loss": 1.0648,
+      "step": 2944
+    },
+    {
+      "epoch": 0.566601533142943,
+      "grad_norm": 0.36915445556975685,
+      "learning_rate": 8.334483738882089e-05,
+      "loss": 1.0104,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5667939275514805,
+      "grad_norm": 0.27973847369457555,
+      "learning_rate": 8.328339025620449e-05,
+      "loss": 1.1257,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5669863219600181,
+      "grad_norm": 0.39155358428980874,
+      "learning_rate": 8.322194961612667e-05,
+      "loss": 1.1273,
+      "step": 2947
+    },
+    {
+      "epoch": 0.5671787163685555,
+      "grad_norm": 0.325140818309315,
+      "learning_rate": 8.316051549245025e-05,
+      "loss": 1.0835,
+      "step": 2948
+    },
+    {
+      "epoch": 0.567371110777093,
+      "grad_norm": 0.3459094343374345,
+      "learning_rate": 8.309908790903561e-05,
+      "loss": 1.0567,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5675635051856306,
+      "grad_norm": 0.2705266112923663,
+      "learning_rate": 8.303766688974046e-05,
+      "loss": 1.0434,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5677558995941681,
+      "grad_norm": 0.2727472811050495,
+      "learning_rate": 8.297625245842005e-05,
+      "loss": 1.2022,
+      "step": 2951
+    },
+    {
+      "epoch": 0.5679482940027055,
+      "grad_norm": 0.319155395199419,
+      "learning_rate": 8.291484463892702e-05,
+      "loss": 0.9956,
+      "step": 2952
+    },
+    {
+      "epoch": 0.5681406884112431,
+      "grad_norm": 0.28603855577669535,
+      "learning_rate": 8.285344345511146e-05,
+      "loss": 1.1022,
+      "step": 2953
+    },
+    {
+      "epoch": 0.5683330828197806,
+      "grad_norm": 0.37313080220443556,
+      "learning_rate": 8.279204893082084e-05,
+      "loss": 1.1066,
+      "step": 2954
+    },
+    {
+      "epoch": 0.568525477228318,
+      "grad_norm": 0.30771450982283066,
+      "learning_rate": 8.273066108990018e-05,
+      "loss": 1.0124,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5687178716368555,
+      "grad_norm": 0.2856987042573155,
+      "learning_rate": 8.266927995619174e-05,
+      "loss": 1.0554,
+      "step": 2956
+    },
+    {
+      "epoch": 0.5689102660453931,
+      "grad_norm": 0.3043271138540745,
+      "learning_rate": 8.260790555353526e-05,
+      "loss": 0.9545,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5691026604539305,
+      "grad_norm": 0.3432701605487397,
+      "learning_rate": 8.254653790576787e-05,
+      "loss": 1.1622,
+      "step": 2958
+    },
+    {
+      "epoch": 0.569295054862468,
+      "grad_norm": 0.28219474642821746,
+      "learning_rate": 8.248517703672403e-05,
+      "loss": 1.0287,
+      "step": 2959
+    },
+    {
+      "epoch": 0.5694874492710056,
+      "grad_norm": 0.2936952098314389,
+      "learning_rate": 8.242382297023559e-05,
+      "loss": 1.209,
+      "step": 2960
+    },
+    {
+      "epoch": 0.569679843679543,
+      "grad_norm": 0.3374468229543128,
+      "learning_rate": 8.23624757301318e-05,
+      "loss": 0.9702,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5698722380880805,
+      "grad_norm": 0.27543060302655936,
+      "learning_rate": 8.230113534023918e-05,
+      "loss": 0.9777,
+      "step": 2962
+    },
+    {
+      "epoch": 0.5700646324966181,
+      "grad_norm": 0.33936999941642565,
+      "learning_rate": 8.223980182438167e-05,
+      "loss": 1.0399,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5702570269051556,
+      "grad_norm": 0.29275017099012624,
+      "learning_rate": 8.217847520638048e-05,
+      "loss": 1.0333,
+      "step": 2964
+    },
+    {
+      "epoch": 0.570449421313693,
+      "grad_norm": 0.31328004913119656,
+      "learning_rate": 8.211715551005415e-05,
+      "loss": 0.9984,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5706418157222306,
+      "grad_norm": 0.256653750000591,
+      "learning_rate": 8.205584275921854e-05,
+      "loss": 1.0715,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5708342101307681,
+      "grad_norm": 0.329455369514639,
+      "learning_rate": 8.199453697768686e-05,
+      "loss": 1.0999,
+      "step": 2967
+    },
+    {
+      "epoch": 0.5710266045393055,
+      "grad_norm": 0.28984053603135707,
+      "learning_rate": 8.193323818926954e-05,
+      "loss": 1.0483,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5712189989478431,
+      "grad_norm": 0.3095215667263857,
+      "learning_rate": 8.187194641777431e-05,
+      "loss": 1.0846,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5714113933563806,
+      "grad_norm": 0.30774485147540814,
+      "learning_rate": 8.181066168700622e-05,
+      "loss": 1.0476,
+      "step": 2970
+    },
+    {
+      "epoch": 0.571603787764918,
+      "grad_norm": 0.3387347512175656,
+      "learning_rate": 8.174938402076754e-05,
+      "loss": 1.0258,
+      "step": 2971
+    },
+    {
+      "epoch": 0.5717961821734556,
+      "grad_norm": 0.35602066290546375,
+      "learning_rate": 8.168811344285776e-05,
+      "loss": 1.0413,
+      "step": 2972
+    },
+    {
+      "epoch": 0.5719885765819931,
+      "grad_norm": 0.3015380111079108,
+      "learning_rate": 8.162684997707374e-05,
+      "loss": 1.0803,
+      "step": 2973
+    },
+    {
+      "epoch": 0.5721809709905306,
+      "grad_norm": 0.2963561122363027,
+      "learning_rate": 8.156559364720947e-05,
+      "loss": 1.1005,
+      "step": 2974
+    },
+    {
+      "epoch": 0.5723733653990681,
+      "grad_norm": 0.329341006860998,
+      "learning_rate": 8.150434447705623e-05,
+      "loss": 1.0479,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5725657598076056,
+      "grad_norm": 0.30062702471568525,
+      "learning_rate": 8.144310249040246e-05,
+      "loss": 1.0658,
+      "step": 2976
+    },
+    {
+      "epoch": 0.5727581542161431,
+      "grad_norm": 0.2585440426067988,
+      "learning_rate": 8.138186771103382e-05,
+      "loss": 1.0614,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5729505486246806,
+      "grad_norm": 0.3532383685540636,
+      "learning_rate": 8.132064016273324e-05,
+      "loss": 1.0534,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5731429430332181,
+      "grad_norm": 0.35197392508075986,
+      "learning_rate": 8.12594198692808e-05,
+      "loss": 1.1429,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5733353374417556,
+      "grad_norm": 0.3552647550885328,
+      "learning_rate": 8.119820685445372e-05,
+      "loss": 1.0337,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5735277318502932,
+      "grad_norm": 0.2798084376744557,
+      "learning_rate": 8.113700114202648e-05,
+      "loss": 1.0314,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5737201262588306,
+      "grad_norm": 0.3122306524448876,
+      "learning_rate": 8.107580275577058e-05,
+      "loss": 1.0146,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5739125206673681,
+      "grad_norm": 0.31672008977998933,
+      "learning_rate": 8.101461171945483e-05,
+      "loss": 1.1221,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5741049150759056,
+      "grad_norm": 0.3142657351882375,
+      "learning_rate": 8.095342805684515e-05,
+      "loss": 1.1336,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5742973094844431,
+      "grad_norm": 0.289786399207669,
+      "learning_rate": 8.089225179170455e-05,
+      "loss": 1.1975,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5744897038929806,
+      "grad_norm": 0.29182138906010224,
+      "learning_rate": 8.083108294779314e-05,
+      "loss": 1.1137,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5746820983015181,
+      "grad_norm": 0.3214063840165335,
+      "learning_rate": 8.076992154886826e-05,
+      "loss": 1.1256,
+      "step": 2987
+    },
+    {
+      "epoch": 0.5748744927100556,
+      "grad_norm": 0.31043283097334473,
+      "learning_rate": 8.070876761868426e-05,
+      "loss": 1.0392,
+      "step": 2988
+    },
+    {
+      "epoch": 0.5750668871185931,
+      "grad_norm": 0.3678254248700128,
+      "learning_rate": 8.064762118099258e-05,
+      "loss": 0.9888,
+      "step": 2989
+    },
+    {
+      "epoch": 0.5752592815271306,
+      "grad_norm": 0.32502399308706836,
+      "learning_rate": 8.058648225954188e-05,
+      "loss": 1.0283,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5754516759356681,
+      "grad_norm": 0.46103901803142605,
+      "learning_rate": 8.052535087807774e-05,
+      "loss": 1.0449,
+      "step": 2991
+    },
+    {
+      "epoch": 0.5756440703442056,
+      "grad_norm": 0.2848126321191939,
+      "learning_rate": 8.046422706034293e-05,
+      "loss": 1.1115,
+      "step": 2992
+    },
+    {
+      "epoch": 0.5758364647527431,
+      "grad_norm": 0.3880332796301805,
+      "learning_rate": 8.040311083007725e-05,
+      "loss": 1.0332,
+      "step": 2993
+    },
+    {
+      "epoch": 0.5760288591612807,
+      "grad_norm": 0.27972481698711205,
+      "learning_rate": 8.034200221101746e-05,
+      "loss": 1.0622,
+      "step": 2994
+    },
+    {
+      "epoch": 0.5762212535698181,
+      "grad_norm": 0.29036261568266625,
+      "learning_rate": 8.028090122689748e-05,
+      "loss": 1.067,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5764136479783556,
+      "grad_norm": 0.31195394162882,
+      "learning_rate": 8.021980790144827e-05,
+      "loss": 1.007,
+      "step": 2996
+    },
+    {
+      "epoch": 0.5766060423868932,
+      "grad_norm": 0.29190211397801924,
+      "learning_rate": 8.015872225839776e-05,
+      "loss": 1.0594,
+      "step": 2997
+    },
+    {
+      "epoch": 0.5767984367954306,
+      "grad_norm": 0.3460122315939625,
+      "learning_rate": 8.009764432147087e-05,
+      "loss": 1.0253,
+      "step": 2998
+    },
+    {
+      "epoch": 0.5769908312039681,
+      "grad_norm": 0.3134818257765632,
+      "learning_rate": 8.00365741143896e-05,
+      "loss": 1.083,
+      "step": 2999
+    },
+    {
+      "epoch": 0.5771832256125057,
+      "grad_norm": 0.19852239844831535,
+      "learning_rate": 7.997551166087292e-05,
+      "loss": 1.0636,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5773756200210431,
+      "grad_norm": 0.32157172158787256,
+      "learning_rate": 7.991445698463672e-05,
+      "loss": 1.0905,
+      "step": 3001
+    },
+    {
+      "epoch": 0.5775680144295806,
+      "grad_norm": 0.2595311745107518,
+      "learning_rate": 7.985341010939402e-05,
+      "loss": 1.0962,
+      "step": 3002
+    },
+    {
+      "epoch": 0.5777604088381182,
+      "grad_norm": 0.2592225998516401,
+      "learning_rate": 7.979237105885467e-05,
+      "loss": 1.0211,
+      "step": 3003
+    },
+    {
+      "epoch": 0.5779528032466557,
+      "grad_norm": 0.2841325486468255,
+      "learning_rate": 7.973133985672558e-05,
+      "loss": 1.0899,
+      "step": 3004
+    },
+    {
+      "epoch": 0.5781451976551931,
+      "grad_norm": 0.3920423417893378,
+      "learning_rate": 7.967031652671051e-05,
+      "loss": 0.988,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5783375920637307,
+      "grad_norm": 0.3315538864014733,
+      "learning_rate": 7.960930109251023e-05,
+      "loss": 1.1081,
+      "step": 3006
+    },
+    {
+      "epoch": 0.5785299864722682,
+      "grad_norm": 0.38049696106104786,
+      "learning_rate": 7.954829357782243e-05,
+      "loss": 1.0139,
+      "step": 3007
+    },
+    {
+      "epoch": 0.5787223808808056,
+      "grad_norm": 0.3031004736351881,
+      "learning_rate": 7.948729400634178e-05,
+      "loss": 0.9951,
+      "step": 3008
+    },
+    {
+      "epoch": 0.5789147752893432,
+      "grad_norm": 0.3076935334332335,
+      "learning_rate": 7.942630240175977e-05,
+      "loss": 1.0051,
+      "step": 3009
+    },
+    {
+      "epoch": 0.5791071696978807,
+      "grad_norm": 0.254168632252162,
+      "learning_rate": 7.936531878776483e-05,
+      "loss": 1.0902,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5792995641064181,
+      "grad_norm": 0.35922989990736226,
+      "learning_rate": 7.930434318804229e-05,
+      "loss": 1.121,
+      "step": 3011
+    },
+    {
+      "epoch": 0.5794919585149556,
+      "grad_norm": 0.38525056963375937,
+      "learning_rate": 7.924337562627435e-05,
+      "loss": 0.9958,
+      "step": 3012
+    },
+    {
+      "epoch": 0.5796843529234932,
+      "grad_norm": 0.3169114236494042,
+      "learning_rate": 7.918241612614017e-05,
+      "loss": 1.0714,
+      "step": 3013
+    },
+    {
+      "epoch": 0.5798767473320307,
+      "grad_norm": 0.35633178952283684,
+      "learning_rate": 7.91214647113157e-05,
+      "loss": 1.1072,
+      "step": 3014
+    },
+    {
+      "epoch": 0.5800691417405681,
+      "grad_norm": 0.33817475160978044,
+      "learning_rate": 7.906052140547373e-05,
+      "loss": 1.0128,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5802615361491057,
+      "grad_norm": 0.4629633589105797,
+      "learning_rate": 7.899958623228397e-05,
+      "loss": 0.9721,
+      "step": 3016
+    },
+    {
+      "epoch": 0.5804539305576432,
+      "grad_norm": 0.401536867567892,
+      "learning_rate": 7.893865921541295e-05,
+      "loss": 1.0179,
+      "step": 3017
+    },
+    {
+      "epoch": 0.5806463249661806,
+      "grad_norm": 0.34724558848245557,
+      "learning_rate": 7.887774037852395e-05,
+      "loss": 1.0602,
+      "step": 3018
+    },
+    {
+      "epoch": 0.5808387193747182,
+      "grad_norm": 0.3169874278396651,
+      "learning_rate": 7.881682974527723e-05,
+      "loss": 1.0927,
+      "step": 3019
+    },
+    {
+      "epoch": 0.5810311137832557,
+      "grad_norm": 0.2858684245012883,
+      "learning_rate": 7.875592733932973e-05,
+      "loss": 1.0323,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5812235081917931,
+      "grad_norm": 0.2731989063280718,
+      "learning_rate": 7.869503318433528e-05,
+      "loss": 1.0445,
+      "step": 3021
+    },
+    {
+      "epoch": 0.5814159026003307,
+      "grad_norm": 0.3103015954746588,
+      "learning_rate": 7.863414730394443e-05,
+      "loss": 1.09,
+      "step": 3022
+    },
+    {
+      "epoch": 0.5816082970088682,
+      "grad_norm": 0.30203232913115585,
+      "learning_rate": 7.857326972180455e-05,
+      "loss": 1.1297,
+      "step": 3023
+    },
+    {
+      "epoch": 0.5818006914174056,
+      "grad_norm": 0.26525352154377424,
+      "learning_rate": 7.85124004615598e-05,
+      "loss": 1.1041,
+      "step": 3024
+    },
+    {
+      "epoch": 0.5819930858259432,
+      "grad_norm": 0.31640303526051006,
+      "learning_rate": 7.845153954685115e-05,
+      "loss": 1.0337,
+      "step": 3025
+    },
+    {
+      "epoch": 0.5821854802344807,
+      "grad_norm": 0.27913055950021226,
+      "learning_rate": 7.839068700131623e-05,
+      "loss": 1.1239,
+      "step": 3026
+    },
+    {
+      "epoch": 0.5823778746430182,
+      "grad_norm": 0.28708002362777646,
+      "learning_rate": 7.832984284858947e-05,
+      "loss": 1.1184,
+      "step": 3027
+    },
+    {
+      "epoch": 0.5825702690515557,
+      "grad_norm": 0.3422422390565635,
+      "learning_rate": 7.826900711230205e-05,
+      "loss": 1.0974,
+      "step": 3028
+    },
+    {
+      "epoch": 0.5827626634600932,
+      "grad_norm": 0.2948247056925184,
+      "learning_rate": 7.820817981608185e-05,
+      "loss": 1.128,
+      "step": 3029
+    },
+    {
+      "epoch": 0.5829550578686307,
+      "grad_norm": 0.2803711512964012,
+      "learning_rate": 7.814736098355347e-05,
+      "loss": 1.0595,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5831474522771682,
+      "grad_norm": 0.31119626206646295,
+      "learning_rate": 7.808655063833832e-05,
+      "loss": 1.0101,
+      "step": 3031
+    },
+    {
+      "epoch": 0.5833398466857057,
+      "grad_norm": 0.3009823259329833,
+      "learning_rate": 7.802574880405438e-05,
+      "loss": 1.0398,
+      "step": 3032
+    },
+    {
+      "epoch": 0.5835322410942432,
+      "grad_norm": 0.32715195524593876,
+      "learning_rate": 7.79649555043164e-05,
+      "loss": 1.1906,
+      "step": 3033
+    },
+    {
+      "epoch": 0.5837246355027808,
+      "grad_norm": 0.33079442156366184,
+      "learning_rate": 7.790417076273581e-05,
+      "loss": 1.0983,
+      "step": 3034
+    },
+    {
+      "epoch": 0.5839170299113182,
+      "grad_norm": 0.26756437043326453,
+      "learning_rate": 7.784339460292064e-05,
+      "loss": 1.1034,
+      "step": 3035
+    },
+    {
+      "epoch": 0.5841094243198557,
+      "grad_norm": 0.3008002530743136,
+      "learning_rate": 7.778262704847569e-05,
+      "loss": 1.0354,
+      "step": 3036
+    },
+    {
+      "epoch": 0.5843018187283932,
+      "grad_norm": 0.30001681408558467,
+      "learning_rate": 7.772186812300244e-05,
+      "loss": 1.0414,
+      "step": 3037
+    },
+    {
+      "epoch": 0.5844942131369307,
+      "grad_norm": 0.27750164021739476,
+      "learning_rate": 7.766111785009889e-05,
+      "loss": 0.9629,
+      "step": 3038
+    },
+    {
+      "epoch": 0.5846866075454682,
+      "grad_norm": 0.3760468994500705,
+      "learning_rate": 7.760037625335973e-05,
+      "loss": 1.0661,
+      "step": 3039
+    },
+    {
+      "epoch": 0.5848790019540057,
+      "grad_norm": 0.3308287060489697,
+      "learning_rate": 7.753964335637634e-05,
+      "loss": 1.0127,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5850713963625432,
+      "grad_norm": 0.2836083274299015,
+      "learning_rate": 7.747891918273666e-05,
+      "loss": 0.9892,
+      "step": 3041
+    },
+    {
+      "epoch": 0.5852637907710807,
+      "grad_norm": 0.2986784878300844,
+      "learning_rate": 7.741820375602523e-05,
+      "loss": 1.0234,
+      "step": 3042
+    },
+    {
+      "epoch": 0.5854561851796182,
+      "grad_norm": 0.3013287247190927,
+      "learning_rate": 7.735749709982328e-05,
+      "loss": 1.075,
+      "step": 3043
+    },
+    {
+      "epoch": 0.5856485795881557,
+      "grad_norm": 0.3070801058232072,
+      "learning_rate": 7.729679923770854e-05,
+      "loss": 1.0188,
+      "step": 3044
+    },
+    {
+      "epoch": 0.5858409739966932,
+      "grad_norm": 0.3052756652102331,
+      "learning_rate": 7.723611019325538e-05,
+      "loss": 0.9934,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5860333684052307,
+      "grad_norm": 0.27134755611052147,
+      "learning_rate": 7.717542999003471e-05,
+      "loss": 1.0411,
+      "step": 3046
+    },
+    {
+      "epoch": 0.5862257628137683,
+      "grad_norm": 0.3270743296681542,
+      "learning_rate": 7.711475865161398e-05,
+      "loss": 1.0282,
+      "step": 3047
+    },
+    {
+      "epoch": 0.5864181572223057,
+      "grad_norm": 0.3186476049801471,
+      "learning_rate": 7.705409620155734e-05,
+      "loss": 1.0951,
+      "step": 3048
+    },
+    {
+      "epoch": 0.5866105516308432,
+      "grad_norm": 0.31250314585223327,
+      "learning_rate": 7.699344266342528e-05,
+      "loss": 1.0227,
+      "step": 3049
+    },
+    {
+      "epoch": 0.5868029460393808,
+      "grad_norm": 0.25580630747247385,
+      "learning_rate": 7.693279806077503e-05,
+      "loss": 1.0338,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5869953404479182,
+      "grad_norm": 0.31017788484111014,
+      "learning_rate": 7.687216241716021e-05,
+      "loss": 1.0808,
+      "step": 3051
+    },
+    {
+      "epoch": 0.5871877348564557,
+      "grad_norm": 0.29349995299992276,
+      "learning_rate": 7.681153575613098e-05,
+      "loss": 1.0587,
+      "step": 3052
+    },
+    {
+      "epoch": 0.5873801292649933,
+      "grad_norm": 0.3139967126275711,
+      "learning_rate": 7.675091810123404e-05,
+      "loss": 0.9752,
+      "step": 3053
+    },
+    {
+      "epoch": 0.5875725236735307,
+      "grad_norm": 0.2680331480372597,
+      "learning_rate": 7.669030947601265e-05,
+      "loss": 1.0127,
+      "step": 3054
+    },
+    {
+      "epoch": 0.5877649180820682,
+      "grad_norm": 0.26250042079949676,
+      "learning_rate": 7.662970990400646e-05,
+      "loss": 1.0904,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5879573124906058,
+      "grad_norm": 0.36141049467350617,
+      "learning_rate": 7.656911940875162e-05,
+      "loss": 1.0701,
+      "step": 3056
+    },
+    {
+      "epoch": 0.5881497068991433,
+      "grad_norm": 0.3029203904354708,
+      "learning_rate": 7.650853801378084e-05,
+      "loss": 1.0959,
+      "step": 3057
+    },
+    {
+      "epoch": 0.5883421013076807,
+      "grad_norm": 0.28161542771351783,
+      "learning_rate": 7.644796574262321e-05,
+      "loss": 1.0199,
+      "step": 3058
+    },
+    {
+      "epoch": 0.5885344957162183,
+      "grad_norm": 0.27539699107600873,
+      "learning_rate": 7.638740261880423e-05,
+      "loss": 1.0165,
+      "step": 3059
+    },
+    {
+      "epoch": 0.5887268901247558,
+      "grad_norm": 0.3074529165873919,
+      "learning_rate": 7.632684866584605e-05,
+      "loss": 1.0613,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5889192845332932,
+      "grad_norm": 0.2468202849305989,
+      "learning_rate": 7.626630390726703e-05,
+      "loss": 1.0617,
+      "step": 3061
+    },
+    {
+      "epoch": 0.5891116789418308,
+      "grad_norm": 0.3238953587284659,
+      "learning_rate": 7.620576836658212e-05,
+      "loss": 1.0019,
+      "step": 3062
+    },
+    {
+      "epoch": 0.5893040733503683,
+      "grad_norm": 0.32862481835404034,
+      "learning_rate": 7.614524206730259e-05,
+      "loss": 1.1729,
+      "step": 3063
+    },
+    {
+      "epoch": 0.5894964677589057,
+      "grad_norm": 0.3223158404509194,
+      "learning_rate": 7.608472503293614e-05,
+      "loss": 1.0441,
+      "step": 3064
+    },
+    {
+      "epoch": 0.5896888621674432,
+      "grad_norm": 0.27374881913205357,
+      "learning_rate": 7.60242172869869e-05,
+      "loss": 1.0838,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5898812565759808,
+      "grad_norm": 0.3405303338348711,
+      "learning_rate": 7.596371885295541e-05,
+      "loss": 1.0497,
+      "step": 3066
+    },
+    {
+      "epoch": 0.5900736509845182,
+      "grad_norm": 0.2720651033528268,
+      "learning_rate": 7.590322975433857e-05,
+      "loss": 1.0814,
+      "step": 3067
+    },
+    {
+      "epoch": 0.5902660453930557,
+      "grad_norm": 0.2740581800085928,
+      "learning_rate": 7.584275001462961e-05,
+      "loss": 1.0741,
+      "step": 3068
+    },
+    {
+      "epoch": 0.5904584398015933,
+      "grad_norm": 0.2919123866647432,
+      "learning_rate": 7.578227965731819e-05,
+      "loss": 1.0061,
+      "step": 3069
+    },
+    {
+      "epoch": 0.5906508342101308,
+      "grad_norm": 0.26393511203481274,
+      "learning_rate": 7.572181870589028e-05,
+      "loss": 1.0526,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5908432286186682,
+      "grad_norm": 0.32518174277827583,
+      "learning_rate": 7.56613671838282e-05,
+      "loss": 1.1785,
+      "step": 3071
+    },
+    {
+      "epoch": 0.5910356230272058,
+      "grad_norm": 0.2630773409765149,
+      "learning_rate": 7.560092511461068e-05,
+      "loss": 1.0677,
+      "step": 3072
+    },
+    {
+      "epoch": 0.5912280174357433,
+      "grad_norm": 0.27534005804629985,
+      "learning_rate": 7.554049252171269e-05,
+      "loss": 1.1679,
+      "step": 3073
+    },
+    {
+      "epoch": 0.5914204118442807,
+      "grad_norm": 0.3148975311461333,
+      "learning_rate": 7.548006942860557e-05,
+      "loss": 1.0578,
+      "step": 3074
+    },
+    {
+      "epoch": 0.5916128062528183,
+      "grad_norm": 0.37808263234443806,
+      "learning_rate": 7.541965585875694e-05,
+      "loss": 0.9947,
+      "step": 3075
+    },
+    {
+      "epoch": 0.5918052006613558,
+      "grad_norm": 0.2886138057187265,
+      "learning_rate": 7.535925183563073e-05,
+      "loss": 1.0711,
+      "step": 3076
+    },
+    {
+      "epoch": 0.5919975950698932,
+      "grad_norm": 0.35547266160823177,
+      "learning_rate": 7.529885738268714e-05,
+      "loss": 1.0901,
+      "step": 3077
+    },
+    {
+      "epoch": 0.5921899894784308,
+      "grad_norm": 0.37904667679195964,
+      "learning_rate": 7.523847252338274e-05,
+      "loss": 1.1462,
+      "step": 3078
+    },
+    {
+      "epoch": 0.5923823838869683,
+      "grad_norm": 0.37811002078595307,
+      "learning_rate": 7.51780972811703e-05,
+      "loss": 1.0059,
+      "step": 3079
+    },
+    {
+      "epoch": 0.5925747782955058,
+      "grad_norm": 0.346049962005794,
+      "learning_rate": 7.511773167949884e-05,
+      "loss": 1.0912,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5927671727040433,
+      "grad_norm": 0.3892857884151602,
+      "learning_rate": 7.505737574181369e-05,
+      "loss": 1.0876,
+      "step": 3081
+    },
+    {
+      "epoch": 0.5929595671125808,
+      "grad_norm": 0.3093045855396968,
+      "learning_rate": 7.499702949155633e-05,
+      "loss": 1.0942,
+      "step": 3082
+    },
+    {
+      "epoch": 0.5931519615211183,
+      "grad_norm": 0.33446027352832136,
+      "learning_rate": 7.493669295216467e-05,
+      "loss": 1.0442,
+      "step": 3083
+    },
+    {
+      "epoch": 0.5933443559296558,
+      "grad_norm": 0.3364260486611449,
+      "learning_rate": 7.487636614707266e-05,
+      "loss": 1.1104,
+      "step": 3084
+    },
+    {
+      "epoch": 0.5935367503381933,
+      "grad_norm": 0.3575690478645272,
+      "learning_rate": 7.48160490997105e-05,
+      "loss": 1.0642,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5937291447467308,
+      "grad_norm": 0.28266921046206767,
+      "learning_rate": 7.475574183350471e-05,
+      "loss": 1.0196,
+      "step": 3086
+    },
+    {
+      "epoch": 0.5939215391552684,
+      "grad_norm": 0.35033837672162804,
+      "learning_rate": 7.469544437187789e-05,
+      "loss": 1.0062,
+      "step": 3087
+    },
+    {
+      "epoch": 0.5941139335638058,
+      "grad_norm": 0.298906654512516,
+      "learning_rate": 7.463515673824888e-05,
+      "loss": 1.0135,
+      "step": 3088
+    },
+    {
+      "epoch": 0.5943063279723433,
+      "grad_norm": 0.31273259577774803,
+      "learning_rate": 7.457487895603272e-05,
+      "loss": 1.0093,
+      "step": 3089
+    },
+    {
+      "epoch": 0.5944987223808808,
+      "grad_norm": 0.3084641365259164,
+      "learning_rate": 7.45146110486406e-05,
+      "loss": 1.0624,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5946911167894183,
+      "grad_norm": 0.27133471846089224,
+      "learning_rate": 7.44543530394799e-05,
+      "loss": 1.0353,
+      "step": 3091
+    },
+    {
+      "epoch": 0.5948835111979558,
+      "grad_norm": 0.3536885418529265,
+      "learning_rate": 7.439410495195412e-05,
+      "loss": 1.1073,
+      "step": 3092
+    },
+    {
+      "epoch": 0.5950759056064933,
+      "grad_norm": 0.2689452442593363,
+      "learning_rate": 7.433386680946288e-05,
+      "loss": 1.077,
+      "step": 3093
+    },
+    {
+      "epoch": 0.5952683000150308,
+      "grad_norm": 0.32634513396455317,
+      "learning_rate": 7.427363863540202e-05,
+      "loss": 1.1045,
+      "step": 3094
+    },
+    {
+      "epoch": 0.5954606944235683,
+      "grad_norm": 0.2526804935657253,
+      "learning_rate": 7.421342045316351e-05,
+      "loss": 1.0189,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5956530888321058,
+      "grad_norm": 0.32919939522606007,
+      "learning_rate": 7.415321228613535e-05,
+      "loss": 1.0347,
+      "step": 3096
+    },
+    {
+      "epoch": 0.5958454832406433,
+      "grad_norm": 0.3915455066319134,
+      "learning_rate": 7.409301415770167e-05,
+      "loss": 1.0579,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5960378776491808,
+      "grad_norm": 0.27195460632896507,
+      "learning_rate": 7.40328260912428e-05,
+      "loss": 0.9482,
+      "step": 3098
+    },
+    {
+      "epoch": 0.5962302720577183,
+      "grad_norm": 0.2936165252582911,
+      "learning_rate": 7.397264811013506e-05,
+      "loss": 1.0041,
+      "step": 3099
+    },
+    {
+      "epoch": 0.5964226664662559,
+      "grad_norm": 0.3811574263997021,
+      "learning_rate": 7.391248023775083e-05,
+      "loss": 1.209,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5966150608747933,
+      "grad_norm": 0.28079577432916897,
+      "learning_rate": 7.385232249745872e-05,
+      "loss": 1.0937,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5968074552833308,
+      "grad_norm": 0.30109437393785105,
+      "learning_rate": 7.379217491262325e-05,
+      "loss": 0.9977,
+      "step": 3102
+    },
+    {
+      "epoch": 0.5969998496918684,
+      "grad_norm": 0.32664253263075066,
+      "learning_rate": 7.373203750660505e-05,
+      "loss": 1.048,
+      "step": 3103
+    },
+    {
+      "epoch": 0.5971922441004058,
+      "grad_norm": 0.2778939205123568,
+      "learning_rate": 7.367191030276079e-05,
+      "loss": 0.9911,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5973846385089433,
+      "grad_norm": 0.4213010065390305,
+      "learning_rate": 7.361179332444318e-05,
+      "loss": 1.0832,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5975770329174809,
+      "grad_norm": 0.30629462463239404,
+      "learning_rate": 7.355168659500095e-05,
+      "loss": 1.0928,
+      "step": 3106
+    },
+    {
+      "epoch": 0.5977694273260183,
+      "grad_norm": 0.2759873905420565,
+      "learning_rate": 7.349159013777891e-05,
+      "loss": 1.0361,
+      "step": 3107
+    },
+    {
+      "epoch": 0.5979618217345558,
+      "grad_norm": 0.3345478739575179,
+      "learning_rate": 7.343150397611782e-05,
+      "loss": 1.0821,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5981542161430934,
+      "grad_norm": 0.3085659634945026,
+      "learning_rate": 7.337142813335444e-05,
+      "loss": 1.0104,
+      "step": 3109
+    },
+    {
+      "epoch": 0.5983466105516309,
+      "grad_norm": 0.3366321058620241,
+      "learning_rate": 7.33113626328215e-05,
+      "loss": 1.0143,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5985390049601683,
+      "grad_norm": 0.3297433585855268,
+      "learning_rate": 7.325130749784782e-05,
+      "loss": 0.9981,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5987313993687059,
+      "grad_norm": 0.4111893929532169,
+      "learning_rate": 7.319126275175801e-05,
+      "loss": 0.9975,
+      "step": 3112
+    },
+    {
+      "epoch": 0.5989237937772434,
+      "grad_norm": 0.2689472354130831,
+      "learning_rate": 7.31312284178729e-05,
+      "loss": 1.0536,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5991161881857808,
+      "grad_norm": 0.2905316066106125,
+      "learning_rate": 7.307120451950901e-05,
+      "loss": 1.0809,
+      "step": 3114
+    },
+    {
+      "epoch": 0.5993085825943184,
+      "grad_norm": 0.29788026421655234,
+      "learning_rate": 7.301119107997905e-05,
+      "loss": 1.0956,
+      "step": 3115
+    },
+    {
+      "epoch": 0.5995009770028559,
+      "grad_norm": 0.29102152972921824,
+      "learning_rate": 7.295118812259145e-05,
+      "loss": 1.1763,
+      "step": 3116
+    },
+    {
+      "epoch": 0.5996933714113933,
+      "grad_norm": 0.3107038025831909,
+      "learning_rate": 7.289119567065068e-05,
+      "loss": 1.0535,
+      "step": 3117
+    },
+    {
+      "epoch": 0.5998857658199308,
+      "grad_norm": 0.24707902033391899,
+      "learning_rate": 7.283121374745715e-05,
+      "loss": 1.0461,
+      "step": 3118
+    },
+    {
+      "epoch": 0.6000781602284684,
+      "grad_norm": 0.3486109517776961,
+      "learning_rate": 7.277124237630712e-05,
+      "loss": 1.0346,
+      "step": 3119
+    },
+    {
+      "epoch": 0.6002705546370058,
+      "grad_norm": 0.3034906473475664,
+      "learning_rate": 7.271128158049283e-05,
+      "loss": 1.1119,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6004629490455433,
+      "grad_norm": 0.33242807184737416,
+      "learning_rate": 7.265133138330233e-05,
+      "loss": 1.0505,
+      "step": 3121
+    },
+    {
+      "epoch": 0.6006553434540809,
+      "grad_norm": 0.3396780950237764,
+      "learning_rate": 7.259139180801955e-05,
+      "loss": 0.9709,
+      "step": 3122
+    },
+    {
+      "epoch": 0.6008477378626184,
+      "grad_norm": 0.3992152627169244,
+      "learning_rate": 7.253146287792434e-05,
+      "loss": 0.9487,
+      "step": 3123
+    },
+    {
+      "epoch": 0.6010401322711558,
+      "grad_norm": 0.2662520459623596,
+      "learning_rate": 7.247154461629247e-05,
+      "loss": 1.1125,
+      "step": 3124
+    },
+    {
+      "epoch": 0.6012325266796934,
+      "grad_norm": 0.40625749861933913,
+      "learning_rate": 7.241163704639546e-05,
+      "loss": 0.9254,
+      "step": 3125
+    },
+    {
+      "epoch": 0.6014249210882309,
+      "grad_norm": 0.32228443916868366,
+      "learning_rate": 7.23517401915007e-05,
+      "loss": 1.1109,
+      "step": 3126
+    },
+    {
+      "epoch": 0.6016173154967683,
+      "grad_norm": 0.3376790795045926,
+      "learning_rate": 7.229185407487148e-05,
+      "loss": 1.0386,
+      "step": 3127
+    },
+    {
+      "epoch": 0.6018097099053059,
+      "grad_norm": 0.2844271214485223,
+      "learning_rate": 7.223197871976689e-05,
+      "loss": 1.0249,
+      "step": 3128
+    },
+    {
+      "epoch": 0.6020021043138434,
+      "grad_norm": 0.3107249107787389,
+      "learning_rate": 7.21721141494417e-05,
+      "loss": 1.0029,
+      "step": 3129
+    },
+    {
+      "epoch": 0.6021944987223808,
+      "grad_norm": 0.3715500084873061,
+      "learning_rate": 7.211226038714678e-05,
+      "loss": 1.075,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6023868931309184,
+      "grad_norm": 0.293414959678292,
+      "learning_rate": 7.205241745612856e-05,
+      "loss": 1.1004,
+      "step": 3131
+    },
+    {
+      "epoch": 0.6025792875394559,
+      "grad_norm": 0.3369710821044232,
+      "learning_rate": 7.199258537962936e-05,
+      "loss": 1.0494,
+      "step": 3132
+    },
+    {
+      "epoch": 0.6027716819479934,
+      "grad_norm": 0.2796816523812784,
+      "learning_rate": 7.193276418088729e-05,
+      "loss": 1.1183,
+      "step": 3133
+    },
+    {
+      "epoch": 0.6029640763565309,
+      "grad_norm": 0.31454650834902614,
+      "learning_rate": 7.187295388313617e-05,
+      "loss": 1.0692,
+      "step": 3134
+    },
+    {
+      "epoch": 0.6031564707650684,
+      "grad_norm": 0.29282374430420566,
+      "learning_rate": 7.181315450960562e-05,
+      "loss": 1.0385,
+      "step": 3135
+    },
+    {
+      "epoch": 0.6033488651736059,
+      "grad_norm": 0.29889728769600493,
+      "learning_rate": 7.175336608352112e-05,
+      "loss": 1.0322,
+      "step": 3136
+    },
+    {
+      "epoch": 0.6035412595821434,
+      "grad_norm": 0.4074927991287567,
+      "learning_rate": 7.169358862810373e-05,
+      "loss": 1.132,
+      "step": 3137
+    },
+    {
+      "epoch": 0.6037336539906809,
+      "grad_norm": 0.2950774894541228,
+      "learning_rate": 7.163382216657034e-05,
+      "loss": 1.031,
+      "step": 3138
+    },
+    {
+      "epoch": 0.6039260483992184,
+      "grad_norm": 0.2611881653496529,
+      "learning_rate": 7.157406672213356e-05,
+      "loss": 1.032,
+      "step": 3139
+    },
+    {
+      "epoch": 0.604118442807756,
+      "grad_norm": 0.34845492103692877,
+      "learning_rate": 7.151432231800172e-05,
+      "loss": 0.9868,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6043108372162934,
+      "grad_norm": 0.3038089797497765,
+      "learning_rate": 7.145458897737881e-05,
+      "loss": 1.0732,
+      "step": 3141
+    },
+    {
+      "epoch": 0.6045032316248309,
+      "grad_norm": 0.3137752390729656,
+      "learning_rate": 7.139486672346466e-05,
+      "loss": 1.108,
+      "step": 3142
+    },
+    {
+      "epoch": 0.6046956260333685,
+      "grad_norm": 0.3694910468844507,
+      "learning_rate": 7.133515557945463e-05,
+      "loss": 1.0062,
+      "step": 3143
+    },
+    {
+      "epoch": 0.6048880204419059,
+      "grad_norm": 0.3294255713228448,
+      "learning_rate": 7.12754555685399e-05,
+      "loss": 1.04,
+      "step": 3144
+    },
+    {
+      "epoch": 0.6050804148504434,
+      "grad_norm": 0.33653669660879837,
+      "learning_rate": 7.121576671390722e-05,
+      "loss": 1.0484,
+      "step": 3145
+    },
+    {
+      "epoch": 0.6052728092589809,
+      "grad_norm": 0.3020546419585607,
+      "learning_rate": 7.115608903873905e-05,
+      "loss": 1.1138,
+      "step": 3146
+    },
+    {
+      "epoch": 0.6054652036675184,
+      "grad_norm": 0.3310334704837057,
+      "learning_rate": 7.109642256621353e-05,
+      "loss": 0.9903,
+      "step": 3147
+    },
+    {
+      "epoch": 0.6056575980760559,
+      "grad_norm": 0.299325820097342,
+      "learning_rate": 7.103676731950443e-05,
+      "loss": 1.1136,
+      "step": 3148
+    },
+    {
+      "epoch": 0.6058499924845934,
+      "grad_norm": 0.3068790577592849,
+      "learning_rate": 7.097712332178117e-05,
+      "loss": 1.16,
+      "step": 3149
+    },
+    {
+      "epoch": 0.606042386893131,
+      "grad_norm": 0.36793760486500554,
+      "learning_rate": 7.09174905962088e-05,
+      "loss": 1.0094,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6062347813016684,
+      "grad_norm": 0.35934732007828213,
+      "learning_rate": 7.085786916594794e-05,
+      "loss": 1.1305,
+      "step": 3151
+    },
+    {
+      "epoch": 0.6064271757102059,
+      "grad_norm": 0.2848795920834444,
+      "learning_rate": 7.07982590541549e-05,
+      "loss": 1.0565,
+      "step": 3152
+    },
+    {
+      "epoch": 0.6066195701187435,
+      "grad_norm": 0.352791679113492,
+      "learning_rate": 7.073866028398153e-05,
+      "loss": 1.0512,
+      "step": 3153
+    },
+    {
+      "epoch": 0.6068119645272809,
+      "grad_norm": 0.3810594999166601,
+      "learning_rate": 7.067907287857535e-05,
+      "loss": 0.9948,
+      "step": 3154
+    },
+    {
+      "epoch": 0.6070043589358184,
+      "grad_norm": 0.320146452926806,
+      "learning_rate": 7.061949686107938e-05,
+      "loss": 1.1565,
+      "step": 3155
+    },
+    {
+      "epoch": 0.607196753344356,
+      "grad_norm": 0.2911436133839381,
+      "learning_rate": 7.055993225463231e-05,
+      "loss": 1.1068,
+      "step": 3156
+    },
+    {
+      "epoch": 0.6073891477528934,
+      "grad_norm": 0.3633095823610109,
+      "learning_rate": 7.050037908236831e-05,
+      "loss": 1.0505,
+      "step": 3157
+    },
+    {
+      "epoch": 0.6075815421614309,
+      "grad_norm": 0.33426229956446274,
+      "learning_rate": 7.04408373674171e-05,
+      "loss": 1.0152,
+      "step": 3158
+    },
+    {
+      "epoch": 0.6077739365699685,
+      "grad_norm": 0.32742563091903526,
+      "learning_rate": 7.03813071329041e-05,
+      "loss": 1.147,
+      "step": 3159
+    },
+    {
+      "epoch": 0.6079663309785059,
+      "grad_norm": 0.2790811793523364,
+      "learning_rate": 7.032178840195008e-05,
+      "loss": 1.0405,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6081587253870434,
+      "grad_norm": 0.2952337042047172,
+      "learning_rate": 7.026228119767148e-05,
+      "loss": 1.0971,
+      "step": 3161
+    },
+    {
+      "epoch": 0.608351119795581,
+      "grad_norm": 0.27164504353733254,
+      "learning_rate": 7.020278554318023e-05,
+      "loss": 1.087,
+      "step": 3162
+    },
+    {
+      "epoch": 0.6085435142041185,
+      "grad_norm": 0.3558037065668216,
+      "learning_rate": 7.014330146158366e-05,
+      "loss": 1.0625,
+      "step": 3163
+    },
+    {
+      "epoch": 0.6087359086126559,
+      "grad_norm": 0.31779233447839716,
+      "learning_rate": 7.008382897598477e-05,
+      "loss": 1.0981,
+      "step": 3164
+    },
+    {
+      "epoch": 0.6089283030211935,
+      "grad_norm": 0.2657203520374658,
+      "learning_rate": 7.0024368109482e-05,
+      "loss": 0.9672,
+      "step": 3165
+    },
+    {
+      "epoch": 0.609120697429731,
+      "grad_norm": 0.2971543985820181,
+      "learning_rate": 6.996491888516927e-05,
+      "loss": 0.9752,
+      "step": 3166
+    },
+    {
+      "epoch": 0.6093130918382684,
+      "grad_norm": 0.3199890294894419,
+      "learning_rate": 6.990548132613592e-05,
+      "loss": 1.0581,
+      "step": 3167
+    },
+    {
+      "epoch": 0.609505486246806,
+      "grad_norm": 0.30185881267616743,
+      "learning_rate": 6.984605545546686e-05,
+      "loss": 1.1385,
+      "step": 3168
+    },
+    {
+      "epoch": 0.6096978806553435,
+      "grad_norm": 0.31439506448656324,
+      "learning_rate": 6.97866412962424e-05,
+      "loss": 1.0981,
+      "step": 3169
+    },
+    {
+      "epoch": 0.6098902750638809,
+      "grad_norm": 0.2975749646567161,
+      "learning_rate": 6.972723887153828e-05,
+      "loss": 1.1096,
+      "step": 3170
+    },
+    {
+      "epoch": 0.6100826694724184,
+      "grad_norm": 0.32543734781439976,
+      "learning_rate": 6.966784820442577e-05,
+      "loss": 1.0293,
+      "step": 3171
+    },
+    {
+      "epoch": 0.610275063880956,
+      "grad_norm": 0.301005268570769,
+      "learning_rate": 6.960846931797152e-05,
+      "loss": 1.0759,
+      "step": 3172
+    },
+    {
+      "epoch": 0.6104674582894934,
+      "grad_norm": 0.40349220413301146,
+      "learning_rate": 6.954910223523756e-05,
+      "loss": 1.1402,
+      "step": 3173
+    },
+    {
+      "epoch": 0.6106598526980309,
+      "grad_norm": 0.28264332051518454,
+      "learning_rate": 6.948974697928143e-05,
+      "loss": 1.1521,
+      "step": 3174
+    },
+    {
+      "epoch": 0.6108522471065685,
+      "grad_norm": 0.35597297323871074,
+      "learning_rate": 6.943040357315598e-05,
+      "loss": 0.9359,
+      "step": 3175
+    },
+    {
+      "epoch": 0.611044641515106,
+      "grad_norm": 0.29310468134660095,
+      "learning_rate": 6.937107203990952e-05,
+      "loss": 1.0784,
+      "step": 3176
+    },
+    {
+      "epoch": 0.6112370359236434,
+      "grad_norm": 0.31643840882070373,
+      "learning_rate": 6.931175240258577e-05,
+      "loss": 1.1196,
+      "step": 3177
+    },
+    {
+      "epoch": 0.611429430332181,
+      "grad_norm": 0.348644595596235,
+      "learning_rate": 6.925244468422376e-05,
+      "loss": 1.0125,
+      "step": 3178
+    },
+    {
+      "epoch": 0.6116218247407185,
+      "grad_norm": 0.35654473063218517,
+      "learning_rate": 6.919314890785793e-05,
+      "loss": 1.1178,
+      "step": 3179
+    },
+    {
+      "epoch": 0.6118142191492559,
+      "grad_norm": 0.3131207296767454,
+      "learning_rate": 6.913386509651807e-05,
+      "loss": 1.0278,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6120066135577935,
+      "grad_norm": 0.2890381392340612,
+      "learning_rate": 6.907459327322934e-05,
+      "loss": 1.1188,
+      "step": 3181
+    },
+    {
+      "epoch": 0.612199007966331,
+      "grad_norm": 0.3468639764201469,
+      "learning_rate": 6.90153334610122e-05,
+      "loss": 1.0762,
+      "step": 3182
+    },
+    {
+      "epoch": 0.6123914023748684,
+      "grad_norm": 0.3118330279435034,
+      "learning_rate": 6.895608568288254e-05,
+      "loss": 1.077,
+      "step": 3183
+    },
+    {
+      "epoch": 0.612583796783406,
+      "grad_norm": 0.35913788631242155,
+      "learning_rate": 6.889684996185148e-05,
+      "loss": 1.0556,
+      "step": 3184
+    },
+    {
+      "epoch": 0.6127761911919435,
+      "grad_norm": 0.3073499316941991,
+      "learning_rate": 6.88376263209255e-05,
+      "loss": 1.0525,
+      "step": 3185
+    },
+    {
+      "epoch": 0.612968585600481,
+      "grad_norm": 0.27311256791698535,
+      "learning_rate": 6.877841478310638e-05,
+      "loss": 1.0137,
+      "step": 3186
+    },
+    {
+      "epoch": 0.6131609800090185,
+      "grad_norm": 0.36096549011145274,
+      "learning_rate": 6.871921537139116e-05,
+      "loss": 1.0076,
+      "step": 3187
+    },
+    {
+      "epoch": 0.613353374417556,
+      "grad_norm": 0.30737596598278466,
+      "learning_rate": 6.866002810877225e-05,
+      "loss": 1.0725,
+      "step": 3188
+    },
+    {
+      "epoch": 0.6135457688260935,
+      "grad_norm": 0.38236401188781555,
+      "learning_rate": 6.860085301823729e-05,
+      "loss": 1.1162,
+      "step": 3189
+    },
+    {
+      "epoch": 0.613738163234631,
+      "grad_norm": 0.32156577713072626,
+      "learning_rate": 6.854169012276923e-05,
+      "loss": 1.0488,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6139305576431685,
+      "grad_norm": 0.30994306625391804,
+      "learning_rate": 6.848253944534622e-05,
+      "loss": 1.054,
+      "step": 3191
+    },
+    {
+      "epoch": 0.614122952051706,
+      "grad_norm": 0.349438353805856,
+      "learning_rate": 6.84234010089417e-05,
+      "loss": 1.1381,
+      "step": 3192
+    },
+    {
+      "epoch": 0.6143153464602435,
+      "grad_norm": 0.3443864850029389,
+      "learning_rate": 6.836427483652436e-05,
+      "loss": 1.0583,
+      "step": 3193
+    },
+    {
+      "epoch": 0.614507740868781,
+      "grad_norm": 0.30440869719274616,
+      "learning_rate": 6.830516095105816e-05,
+      "loss": 0.9697,
+      "step": 3194
+    },
+    {
+      "epoch": 0.6147001352773185,
+      "grad_norm": 0.3196110425371395,
+      "learning_rate": 6.824605937550223e-05,
+      "loss": 0.9898,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6148925296858561,
+      "grad_norm": 0.2729958040406237,
+      "learning_rate": 6.818697013281092e-05,
+      "loss": 1.0265,
+      "step": 3196
+    },
+    {
+      "epoch": 0.6150849240943935,
+      "grad_norm": 0.35942339061496686,
+      "learning_rate": 6.812789324593386e-05,
+      "loss": 1.0153,
+      "step": 3197
+    },
+    {
+      "epoch": 0.615277318502931,
+      "grad_norm": 0.487622474816348,
+      "learning_rate": 6.80688287378158e-05,
+      "loss": 1.0876,
+      "step": 3198
+    },
+    {
+      "epoch": 0.6154697129114685,
+      "grad_norm": 0.3069970110857509,
+      "learning_rate": 6.800977663139666e-05,
+      "loss": 1.0567,
+      "step": 3199
+    },
+    {
+      "epoch": 0.615662107320006,
+      "grad_norm": 0.32958271293881347,
+      "learning_rate": 6.79507369496117e-05,
+      "loss": 1.0528,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6158545017285435,
+      "grad_norm": 0.25059074857258157,
+      "learning_rate": 6.789170971539118e-05,
+      "loss": 1.1515,
+      "step": 3201
+    },
+    {
+      "epoch": 0.616046896137081,
+      "grad_norm": 0.29879244255037085,
+      "learning_rate": 6.783269495166065e-05,
+      "loss": 1.1002,
+      "step": 3202
+    },
+    {
+      "epoch": 0.6162392905456185,
+      "grad_norm": 0.2688747575922325,
+      "learning_rate": 6.777369268134076e-05,
+      "loss": 1.0205,
+      "step": 3203
+    },
+    {
+      "epoch": 0.616431684954156,
+      "grad_norm": 0.38096085837503346,
+      "learning_rate": 6.771470292734723e-05,
+      "loss": 0.9968,
+      "step": 3204
+    },
+    {
+      "epoch": 0.6166240793626935,
+      "grad_norm": 0.332673446138928,
+      "learning_rate": 6.765572571259106e-05,
+      "loss": 1.0151,
+      "step": 3205
+    },
+    {
+      "epoch": 0.616816473771231,
+      "grad_norm": 0.32996935799721544,
+      "learning_rate": 6.759676105997834e-05,
+      "loss": 1.0917,
+      "step": 3206
+    },
+    {
+      "epoch": 0.6170088681797685,
+      "grad_norm": 0.26740183024326003,
+      "learning_rate": 6.753780899241027e-05,
+      "loss": 1.0526,
+      "step": 3207
+    },
+    {
+      "epoch": 0.617201262588306,
+      "grad_norm": 0.3648189581681417,
+      "learning_rate": 6.74788695327831e-05,
+      "loss": 1.0182,
+      "step": 3208
+    },
+    {
+      "epoch": 0.6173936569968436,
+      "grad_norm": 0.3163587900480176,
+      "learning_rate": 6.741994270398826e-05,
+      "loss": 1.0459,
+      "step": 3209
+    },
+    {
+      "epoch": 0.617586051405381,
+      "grad_norm": 0.327200191085277,
+      "learning_rate": 6.736102852891227e-05,
+      "loss": 1.0633,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6177784458139185,
+      "grad_norm": 0.336514780512139,
+      "learning_rate": 6.730212703043665e-05,
+      "loss": 1.1718,
+      "step": 3211
+    },
+    {
+      "epoch": 0.6179708402224561,
+      "grad_norm": 0.30205308924607555,
+      "learning_rate": 6.724323823143818e-05,
+      "loss": 1.0267,
+      "step": 3212
+    },
+    {
+      "epoch": 0.6181632346309935,
+      "grad_norm": 0.2748087610486462,
+      "learning_rate": 6.718436215478848e-05,
+      "loss": 1.0177,
+      "step": 3213
+    },
+    {
+      "epoch": 0.618355629039531,
+      "grad_norm": 0.3258845059063351,
+      "learning_rate": 6.712549882335441e-05,
+      "loss": 1.0164,
+      "step": 3214
+    },
+    {
+      "epoch": 0.6185480234480686,
+      "grad_norm": 0.3405240358466416,
+      "learning_rate": 6.70666482599978e-05,
+      "loss": 1.0605,
+      "step": 3215
+    },
+    {
+      "epoch": 0.618740417856606,
+      "grad_norm": 0.3316771301275599,
+      "learning_rate": 6.700781048757547e-05,
+      "loss": 1.0885,
+      "step": 3216
+    },
+    {
+      "epoch": 0.6189328122651435,
+      "grad_norm": 0.3479830177651135,
+      "learning_rate": 6.69489855289394e-05,
+      "loss": 1.0174,
+      "step": 3217
+    },
+    {
+      "epoch": 0.6191252066736811,
+      "grad_norm": 0.33697682656409855,
+      "learning_rate": 6.689017340693648e-05,
+      "loss": 1.0733,
+      "step": 3218
+    },
+    {
+      "epoch": 0.6193176010822186,
+      "grad_norm": 0.3202779683593554,
+      "learning_rate": 6.683137414440872e-05,
+      "loss": 1.0256,
+      "step": 3219
+    },
+    {
+      "epoch": 0.619509995490756,
+      "grad_norm": 0.3330827951961716,
+      "learning_rate": 6.677258776419305e-05,
+      "loss": 0.9833,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6197023898992936,
+      "grad_norm": 0.3830597198297923,
+      "learning_rate": 6.671381428912138e-05,
+      "loss": 1.0219,
+      "step": 3221
+    },
+    {
+      "epoch": 0.6198947843078311,
+      "grad_norm": 0.3510247204133472,
+      "learning_rate": 6.66550537420207e-05,
+      "loss": 1.0097,
+      "step": 3222
+    },
+    {
+      "epoch": 0.6200871787163685,
+      "grad_norm": 0.3423656046146432,
+      "learning_rate": 6.659630614571288e-05,
+      "loss": 0.9948,
+      "step": 3223
+    },
+    {
+      "epoch": 0.620279573124906,
+      "grad_norm": 0.2948880138448875,
+      "learning_rate": 6.653757152301487e-05,
+      "loss": 1.069,
+      "step": 3224
+    },
+    {
+      "epoch": 0.6204719675334436,
+      "grad_norm": 0.3815063867172276,
+      "learning_rate": 6.647884989673849e-05,
+      "loss": 1.0371,
+      "step": 3225
+    },
+    {
+      "epoch": 0.620664361941981,
+      "grad_norm": 0.3078550601774885,
+      "learning_rate": 6.642014128969055e-05,
+      "loss": 1.0692,
+      "step": 3226
+    },
+    {
+      "epoch": 0.6208567563505185,
+      "grad_norm": 0.34606514694043655,
+      "learning_rate": 6.63614457246728e-05,
+      "loss": 1.0321,
+      "step": 3227
+    },
+    {
+      "epoch": 0.6210491507590561,
+      "grad_norm": 0.41116911806315476,
+      "learning_rate": 6.630276322448188e-05,
+      "loss": 1.0745,
+      "step": 3228
+    },
+    {
+      "epoch": 0.6212415451675936,
+      "grad_norm": 0.24472103250606803,
+      "learning_rate": 6.624409381190945e-05,
+      "loss": 1.0985,
+      "step": 3229
+    },
+    {
+      "epoch": 0.621433939576131,
+      "grad_norm": 0.31352932435117903,
+      "learning_rate": 6.618543750974202e-05,
+      "loss": 1.0179,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6216263339846686,
+      "grad_norm": 0.3151752206717778,
+      "learning_rate": 6.612679434076102e-05,
+      "loss": 1.0628,
+      "step": 3231
+    },
+    {
+      "epoch": 0.6218187283932061,
+      "grad_norm": 0.3065423050285731,
+      "learning_rate": 6.606816432774278e-05,
+      "loss": 1.0622,
+      "step": 3232
+    },
+    {
+      "epoch": 0.6220111228017435,
+      "grad_norm": 0.29479034068650284,
+      "learning_rate": 6.600954749345851e-05,
+      "loss": 1.107,
+      "step": 3233
+    },
+    {
+      "epoch": 0.6222035172102811,
+      "grad_norm": 0.4010322666023779,
+      "learning_rate": 6.595094386067429e-05,
+      "loss": 1.0238,
+      "step": 3234
+    },
+    {
+      "epoch": 0.6223959116188186,
+      "grad_norm": 0.4423963767043969,
+      "learning_rate": 6.589235345215117e-05,
+      "loss": 1.0631,
+      "step": 3235
+    },
+    {
+      "epoch": 0.622588306027356,
+      "grad_norm": 0.3586024441649116,
+      "learning_rate": 6.583377629064494e-05,
+      "loss": 1.0822,
+      "step": 3236
+    },
+    {
+      "epoch": 0.6227807004358936,
+      "grad_norm": 0.28908757771648613,
+      "learning_rate": 6.57752123989063e-05,
+      "loss": 1.0478,
+      "step": 3237
+    },
+    {
+      "epoch": 0.6229730948444311,
+      "grad_norm": 0.364994836191207,
+      "learning_rate": 6.571666179968079e-05,
+      "loss": 1.0865,
+      "step": 3238
+    },
+    {
+      "epoch": 0.6231654892529686,
+      "grad_norm": 0.2709653216374744,
+      "learning_rate": 6.56581245157088e-05,
+      "loss": 1.0485,
+      "step": 3239
+    },
+    {
+      "epoch": 0.6233578836615061,
+      "grad_norm": 0.3209669544971997,
+      "learning_rate": 6.55996005697255e-05,
+      "loss": 1.0074,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6235502780700436,
+      "grad_norm": 0.35443750198506224,
+      "learning_rate": 6.554108998446095e-05,
+      "loss": 1.0011,
+      "step": 3241
+    },
+    {
+      "epoch": 0.6237426724785811,
+      "grad_norm": 0.3361987303876914,
+      "learning_rate": 6.548259278264e-05,
+      "loss": 1.0608,
+      "step": 3242
+    },
+    {
+      "epoch": 0.6239350668871186,
+      "grad_norm": 0.3776409290138578,
+      "learning_rate": 6.542410898698226e-05,
+      "loss": 1.1058,
+      "step": 3243
+    },
+    {
+      "epoch": 0.6241274612956561,
+      "grad_norm": 0.3164591304376817,
+      "learning_rate": 6.536563862020218e-05,
+      "loss": 1.1111,
+      "step": 3244
+    },
+    {
+      "epoch": 0.6243198557041936,
+      "grad_norm": 0.33195484277223375,
+      "learning_rate": 6.530718170500895e-05,
+      "loss": 1.0232,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6245122501127311,
+      "grad_norm": 0.32228404606207534,
+      "learning_rate": 6.524873826410658e-05,
+      "loss": 1.0066,
+      "step": 3246
+    },
+    {
+      "epoch": 0.6247046445212686,
+      "grad_norm": 0.3515019637636345,
+      "learning_rate": 6.519030832019383e-05,
+      "loss": 1.086,
+      "step": 3247
+    },
+    {
+      "epoch": 0.6248970389298061,
+      "grad_norm": 0.28337674561169207,
+      "learning_rate": 6.513189189596423e-05,
+      "loss": 1.0629,
+      "step": 3248
+    },
+    {
+      "epoch": 0.6250894333383437,
+      "grad_norm": 0.2759755152994146,
+      "learning_rate": 6.507348901410604e-05,
+      "loss": 0.9868,
+      "step": 3249
+    },
+    {
+      "epoch": 0.6252818277468811,
+      "grad_norm": 0.30357710842318314,
+      "learning_rate": 6.501509969730224e-05,
+      "loss": 1.0086,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6254742221554186,
+      "grad_norm": 0.29845308238407653,
+      "learning_rate": 6.49567239682306e-05,
+      "loss": 1.1055,
+      "step": 3251
+    },
+    {
+      "epoch": 0.6256666165639561,
+      "grad_norm": 0.2740601001913743,
+      "learning_rate": 6.489836184956352e-05,
+      "loss": 1.076,
+      "step": 3252
+    },
+    {
+      "epoch": 0.6258590109724936,
+      "grad_norm": 0.28177946664996745,
+      "learning_rate": 6.484001336396828e-05,
+      "loss": 1.1389,
+      "step": 3253
+    },
+    {
+      "epoch": 0.6260514053810311,
+      "grad_norm": 0.3243522137813481,
+      "learning_rate": 6.478167853410668e-05,
+      "loss": 1.0381,
+      "step": 3254
+    },
+    {
+      "epoch": 0.6262437997895686,
+      "grad_norm": 0.24804116839338353,
+      "learning_rate": 6.472335738263534e-05,
+      "loss": 1.0058,
+      "step": 3255
+    },
+    {
+      "epoch": 0.6264361941981061,
+      "grad_norm": 0.2771303461685542,
+      "learning_rate": 6.466504993220548e-05,
+      "loss": 1.0447,
+      "step": 3256
+    },
+    {
+      "epoch": 0.6266285886066436,
+      "grad_norm": 0.36070024374836523,
+      "learning_rate": 6.460675620546305e-05,
+      "loss": 1.0533,
+      "step": 3257
+    },
+    {
+      "epoch": 0.6268209830151811,
+      "grad_norm": 0.3293182096919963,
+      "learning_rate": 6.454847622504867e-05,
+      "loss": 1.0899,
+      "step": 3258
+    },
+    {
+      "epoch": 0.6270133774237187,
+      "grad_norm": 0.31036914850569086,
+      "learning_rate": 6.449021001359763e-05,
+      "loss": 1.155,
+      "step": 3259
+    },
+    {
+      "epoch": 0.6272057718322561,
+      "grad_norm": 0.3653809423664422,
+      "learning_rate": 6.443195759373984e-05,
+      "loss": 1.1249,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6273981662407936,
+      "grad_norm": 0.2988202569348402,
+      "learning_rate": 6.43737189880999e-05,
+      "loss": 1.109,
+      "step": 3261
+    },
+    {
+      "epoch": 0.6275905606493312,
+      "grad_norm": 0.2737358679990404,
+      "learning_rate": 6.431549421929694e-05,
+      "loss": 1.0509,
+      "step": 3262
+    },
+    {
+      "epoch": 0.6277829550578686,
+      "grad_norm": 0.3102302212792746,
+      "learning_rate": 6.42572833099448e-05,
+      "loss": 0.9636,
+      "step": 3263
+    },
+    {
+      "epoch": 0.6279753494664061,
+      "grad_norm": 0.29140370713712194,
+      "learning_rate": 6.419908628265203e-05,
+      "loss": 1.0477,
+      "step": 3264
+    },
+    {
+      "epoch": 0.6281677438749437,
+      "grad_norm": 0.29649158880804677,
+      "learning_rate": 6.41409031600216e-05,
+      "loss": 1.1414,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6283601382834811,
+      "grad_norm": 0.3382830964293109,
+      "learning_rate": 6.408273396465117e-05,
+      "loss": 1.0583,
+      "step": 3266
+    },
+    {
+      "epoch": 0.6285525326920186,
+      "grad_norm": 0.3261742714060955,
+      "learning_rate": 6.4024578719133e-05,
+      "loss": 1.0536,
+      "step": 3267
+    },
+    {
+      "epoch": 0.6287449271005562,
+      "grad_norm": 0.3211658155217184,
+      "learning_rate": 6.396643744605391e-05,
+      "loss": 1.0728,
+      "step": 3268
+    },
+    {
+      "epoch": 0.6289373215090936,
+      "grad_norm": 0.2976463044815811,
+      "learning_rate": 6.390831016799526e-05,
+      "loss": 1.0521,
+      "step": 3269
+    },
+    {
+      "epoch": 0.6291297159176311,
+      "grad_norm": 0.33262577793029213,
+      "learning_rate": 6.38501969075331e-05,
+      "loss": 1.0809,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6293221103261687,
+      "grad_norm": 0.277220432009812,
+      "learning_rate": 6.379209768723791e-05,
+      "loss": 1.0484,
+      "step": 3271
+    },
+    {
+      "epoch": 0.6295145047347062,
+      "grad_norm": 0.33964435513218144,
+      "learning_rate": 6.373401252967475e-05,
+      "loss": 1.0739,
+      "step": 3272
+    },
+    {
+      "epoch": 0.6297068991432436,
+      "grad_norm": 0.28450375692843183,
+      "learning_rate": 6.367594145740324e-05,
+      "loss": 1.0327,
+      "step": 3273
+    },
+    {
+      "epoch": 0.6298992935517812,
+      "grad_norm": 0.32661271925115803,
+      "learning_rate": 6.361788449297747e-05,
+      "loss": 1.0674,
+      "step": 3274
+    },
+    {
+      "epoch": 0.6300916879603187,
+      "grad_norm": 0.315459592093191,
+      "learning_rate": 6.355984165894613e-05,
+      "loss": 1.085,
+      "step": 3275
+    },
+    {
+      "epoch": 0.6302840823688561,
+      "grad_norm": 0.253299074954734,
+      "learning_rate": 6.350181297785241e-05,
+      "loss": 1.0749,
+      "step": 3276
+    },
+    {
+      "epoch": 0.6304764767773936,
+      "grad_norm": 0.3352430348465554,
+      "learning_rate": 6.344379847223398e-05,
+      "loss": 0.9969,
+      "step": 3277
+    },
+    {
+      "epoch": 0.6306688711859312,
+      "grad_norm": 0.44857188716230184,
+      "learning_rate": 6.338579816462298e-05,
+      "loss": 1.0771,
+      "step": 3278
+    },
+    {
+      "epoch": 0.6308612655944686,
+      "grad_norm": 0.31878567263193996,
+      "learning_rate": 6.332781207754605e-05,
+      "loss": 1.0634,
+      "step": 3279
+    },
+    {
+      "epoch": 0.6310536600030061,
+      "grad_norm": 0.4220603210907803,
+      "learning_rate": 6.326984023352435e-05,
+      "loss": 1.0782,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6312460544115437,
+      "grad_norm": 0.3500112673391026,
+      "learning_rate": 6.321188265507342e-05,
+      "loss": 1.0784,
+      "step": 3281
+    },
+    {
+      "epoch": 0.6314384488200812,
+      "grad_norm": 0.3317206323233369,
+      "learning_rate": 6.31539393647034e-05,
+      "loss": 1.0612,
+      "step": 3282
+    },
+    {
+      "epoch": 0.6316308432286186,
+      "grad_norm": 0.2834538104859952,
+      "learning_rate": 6.309601038491874e-05,
+      "loss": 1.0232,
+      "step": 3283
+    },
+    {
+      "epoch": 0.6318232376371562,
+      "grad_norm": 0.2876697113533709,
+      "learning_rate": 6.303809573821842e-05,
+      "loss": 1.1626,
+      "step": 3284
+    },
+    {
+      "epoch": 0.6320156320456937,
+      "grad_norm": 0.32032901784811857,
+      "learning_rate": 6.298019544709579e-05,
+      "loss": 1.0028,
+      "step": 3285
+    },
+    {
+      "epoch": 0.6322080264542311,
+      "grad_norm": 0.31393383093016775,
+      "learning_rate": 6.292230953403866e-05,
+      "loss": 1.0758,
+      "step": 3286
+    },
+    {
+      "epoch": 0.6324004208627687,
+      "grad_norm": 0.2866606285030481,
+      "learning_rate": 6.286443802152926e-05,
+      "loss": 1.0882,
+      "step": 3287
+    },
+    {
+      "epoch": 0.6325928152713062,
+      "grad_norm": 0.3566030827974186,
+      "learning_rate": 6.280658093204422e-05,
+      "loss": 1.126,
+      "step": 3288
+    },
+    {
+      "epoch": 0.6327852096798436,
+      "grad_norm": 0.2882110299183065,
+      "learning_rate": 6.274873828805458e-05,
+      "loss": 0.9535,
+      "step": 3289
+    },
+    {
+      "epoch": 0.6329776040883812,
+      "grad_norm": 0.28770803337268513,
+      "learning_rate": 6.269091011202575e-05,
+      "loss": 1.036,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6331699984969187,
+      "grad_norm": 0.3597252464793629,
+      "learning_rate": 6.26330964264175e-05,
+      "loss": 1.0505,
+      "step": 3291
+    },
+    {
+      "epoch": 0.6333623929054562,
+      "grad_norm": 0.31478618153593463,
+      "learning_rate": 6.257529725368405e-05,
+      "loss": 1.0855,
+      "step": 3292
+    },
+    {
+      "epoch": 0.6335547873139937,
+      "grad_norm": 0.3279002575403472,
+      "learning_rate": 6.251751261627385e-05,
+      "loss": 1.0276,
+      "step": 3293
+    },
+    {
+      "epoch": 0.6337471817225312,
+      "grad_norm": 0.27649089506198865,
+      "learning_rate": 6.245974253662987e-05,
+      "loss": 1.0958,
+      "step": 3294
+    },
+    {
+      "epoch": 0.6339395761310687,
+      "grad_norm": 0.3285416682656933,
+      "learning_rate": 6.240198703718931e-05,
+      "loss": 0.9991,
+      "step": 3295
+    },
+    {
+      "epoch": 0.6341319705396062,
+      "grad_norm": 0.2884009081311535,
+      "learning_rate": 6.234424614038375e-05,
+      "loss": 1.0619,
+      "step": 3296
+    },
+    {
+      "epoch": 0.6343243649481437,
+      "grad_norm": 0.3255285238518406,
+      "learning_rate": 6.22865198686391e-05,
+      "loss": 1.0076,
+      "step": 3297
+    },
+    {
+      "epoch": 0.6345167593566812,
+      "grad_norm": 0.31144663446899296,
+      "learning_rate": 6.22288082443755e-05,
+      "loss": 1.0091,
+      "step": 3298
+    },
+    {
+      "epoch": 0.6347091537652187,
+      "grad_norm": 0.3434456634130591,
+      "learning_rate": 6.217111129000759e-05,
+      "loss": 1.0673,
+      "step": 3299
+    },
+    {
+      "epoch": 0.6349015481737562,
+      "grad_norm": 0.2909305213438964,
+      "learning_rate": 6.211342902794413e-05,
+      "loss": 0.9468,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6350939425822937,
+      "grad_norm": 0.2805083613239678,
+      "learning_rate": 6.205576148058828e-05,
+      "loss": 1.0027,
+      "step": 3301
+    },
+    {
+      "epoch": 0.6352863369908313,
+      "grad_norm": 0.28429525215888946,
+      "learning_rate": 6.199810867033746e-05,
+      "loss": 0.9934,
+      "step": 3302
+    },
+    {
+      "epoch": 0.6354787313993687,
+      "grad_norm": 0.3054596757460676,
+      "learning_rate": 6.19404706195833e-05,
+      "loss": 0.9942,
+      "step": 3303
+    },
+    {
+      "epoch": 0.6356711258079062,
+      "grad_norm": 0.48592238735603327,
+      "learning_rate": 6.188284735071177e-05,
+      "loss": 1.1198,
+      "step": 3304
+    },
+    {
+      "epoch": 0.6358635202164437,
+      "grad_norm": 0.31043074682586275,
+      "learning_rate": 6.182523888610316e-05,
+      "loss": 1.1537,
+      "step": 3305
+    },
+    {
+      "epoch": 0.6360559146249812,
+      "grad_norm": 0.3056883480277573,
+      "learning_rate": 6.176764524813187e-05,
+      "loss": 1.0194,
+      "step": 3306
+    },
+    {
+      "epoch": 0.6362483090335187,
+      "grad_norm": 0.3853795908283225,
+      "learning_rate": 6.171006645916661e-05,
+      "loss": 1.011,
+      "step": 3307
+    },
+    {
+      "epoch": 0.6364407034420562,
+      "grad_norm": 0.33361669580306874,
+      "learning_rate": 6.165250254157031e-05,
+      "loss": 0.9554,
+      "step": 3308
+    },
+    {
+      "epoch": 0.6366330978505937,
+      "grad_norm": 0.43820426918359007,
+      "learning_rate": 6.159495351770017e-05,
+      "loss": 1.0356,
+      "step": 3309
+    },
+    {
+      "epoch": 0.6368254922591312,
+      "grad_norm": 0.38145749560677095,
+      "learning_rate": 6.15374194099075e-05,
+      "loss": 0.989,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6370178866676687,
+      "grad_norm": 0.2591346268038298,
+      "learning_rate": 6.147990024053796e-05,
+      "loss": 1.0879,
+      "step": 3311
+    },
+    {
+      "epoch": 0.6372102810762063,
+      "grad_norm": 0.3173659968597233,
+      "learning_rate": 6.142239603193127e-05,
+      "loss": 1.0605,
+      "step": 3312
+    },
+    {
+      "epoch": 0.6374026754847437,
+      "grad_norm": 0.3097825813450799,
+      "learning_rate": 6.136490680642146e-05,
+      "loss": 1.0771,
+      "step": 3313
+    },
+    {
+      "epoch": 0.6375950698932812,
+      "grad_norm": 0.31390018539583237,
+      "learning_rate": 6.130743258633667e-05,
+      "loss": 1.0579,
+      "step": 3314
+    },
+    {
+      "epoch": 0.6377874643018188,
+      "grad_norm": 0.33347954798568236,
+      "learning_rate": 6.124997339399916e-05,
+      "loss": 1.0346,
+      "step": 3315
+    },
+    {
+      "epoch": 0.6379798587103562,
+      "grad_norm": 0.27345445949181163,
+      "learning_rate": 6.11925292517255e-05,
+      "loss": 1.1647,
+      "step": 3316
+    },
+    {
+      "epoch": 0.6381722531188937,
+      "grad_norm": 0.2737259722260716,
+      "learning_rate": 6.113510018182627e-05,
+      "loss": 1.0115,
+      "step": 3317
+    },
+    {
+      "epoch": 0.6383646475274313,
+      "grad_norm": 0.45835645437873973,
+      "learning_rate": 6.107768620660632e-05,
+      "loss": 1.1887,
+      "step": 3318
+    },
+    {
+      "epoch": 0.6385570419359687,
+      "grad_norm": 0.38738529902933116,
+      "learning_rate": 6.102028734836456e-05,
+      "loss": 1.0255,
+      "step": 3319
+    },
+    {
+      "epoch": 0.6387494363445062,
+      "grad_norm": 0.2910298616448152,
+      "learning_rate": 6.0962903629394e-05,
+      "loss": 1.0668,
+      "step": 3320
+    },
+    {
+      "epoch": 0.6389418307530438,
+      "grad_norm": 0.341935564357545,
+      "learning_rate": 6.090553507198187e-05,
+      "loss": 1.0782,
+      "step": 3321
+    },
+    {
+      "epoch": 0.6391342251615812,
+      "grad_norm": 0.3237868686841295,
+      "learning_rate": 6.084818169840938e-05,
+      "loss": 1.037,
+      "step": 3322
+    },
+    {
+      "epoch": 0.6393266195701187,
+      "grad_norm": 0.2971280606010649,
+      "learning_rate": 6.079084353095201e-05,
+      "loss": 1.0298,
+      "step": 3323
+    },
+    {
+      "epoch": 0.6395190139786563,
+      "grad_norm": 0.38124794607764034,
+      "learning_rate": 6.07335205918792e-05,
+      "loss": 1.0519,
+      "step": 3324
+    },
+    {
+      "epoch": 0.6397114083871938,
+      "grad_norm": 0.2874092020563654,
+      "learning_rate": 6.0676212903454543e-05,
+      "loss": 1.0473,
+      "step": 3325
+    },
+    {
+      "epoch": 0.6399038027957312,
+      "grad_norm": 0.33483377221512484,
+      "learning_rate": 6.061892048793567e-05,
+      "loss": 1.1512,
+      "step": 3326
+    },
+    {
+      "epoch": 0.6400961972042688,
+      "grad_norm": 0.3253224483316621,
+      "learning_rate": 6.056164336757426e-05,
+      "loss": 1.0252,
+      "step": 3327
+    },
+    {
+      "epoch": 0.6402885916128063,
+      "grad_norm": 0.37329254897434155,
+      "learning_rate": 6.050438156461613e-05,
+      "loss": 1.0117,
+      "step": 3328
+    },
+    {
+      "epoch": 0.6404809860213437,
+      "grad_norm": 0.3605613571270172,
+      "learning_rate": 6.0447135101301077e-05,
+      "loss": 0.9967,
+      "step": 3329
+    },
+    {
+      "epoch": 0.6406733804298813,
+      "grad_norm": 0.2952491863544519,
+      "learning_rate": 6.038990399986302e-05,
+      "loss": 1.1094,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6408657748384188,
+      "grad_norm": 0.3031487556295676,
+      "learning_rate": 6.0332688282529804e-05,
+      "loss": 0.9857,
+      "step": 3331
+    },
+    {
+      "epoch": 0.6410581692469562,
+      "grad_norm": 0.3723881085078127,
+      "learning_rate": 6.027548797152336e-05,
+      "loss": 1.0544,
+      "step": 3332
+    },
+    {
+      "epoch": 0.6412505636554937,
+      "grad_norm": 0.3062297014261907,
+      "learning_rate": 6.0218303089059626e-05,
+      "loss": 1.0451,
+      "step": 3333
+    },
+    {
+      "epoch": 0.6414429580640313,
+      "grad_norm": 0.3113549757993061,
+      "learning_rate": 6.016113365734861e-05,
+      "loss": 1.0768,
+      "step": 3334
+    },
+    {
+      "epoch": 0.6416353524725688,
+      "grad_norm": 0.25129588602287267,
+      "learning_rate": 6.010397969859421e-05,
+      "loss": 1.1132,
+      "step": 3335
+    },
+    {
+      "epoch": 0.6418277468811062,
+      "grad_norm": 0.3210328629829216,
+      "learning_rate": 6.004684123499436e-05,
+      "loss": 0.9612,
+      "step": 3336
+    },
+    {
+      "epoch": 0.6420201412896438,
+      "grad_norm": 0.3427142782343593,
+      "learning_rate": 5.998971828874102e-05,
+      "loss": 1.0178,
+      "step": 3337
+    },
+    {
+      "epoch": 0.6422125356981813,
+      "grad_norm": 0.33394889645808373,
+      "learning_rate": 5.9932610882020046e-05,
+      "loss": 1.1197,
+      "step": 3338
+    },
+    {
+      "epoch": 0.6424049301067187,
+      "grad_norm": 0.3112312897238979,
+      "learning_rate": 5.987551903701127e-05,
+      "loss": 1.0553,
+      "step": 3339
+    },
+    {
+      "epoch": 0.6425973245152563,
+      "grad_norm": 0.40182320808874666,
+      "learning_rate": 5.9818442775888595e-05,
+      "loss": 0.9917,
+      "step": 3340
+    },
+    {
+      "epoch": 0.6427897189237938,
+      "grad_norm": 0.3205377821789851,
+      "learning_rate": 5.97613821208197e-05,
+      "loss": 0.9541,
+      "step": 3341
+    },
+    {
+      "epoch": 0.6429821133323312,
+      "grad_norm": 0.32416902476703935,
+      "learning_rate": 5.9704337093966344e-05,
+      "loss": 0.9948,
+      "step": 3342
+    },
+    {
+      "epoch": 0.6431745077408688,
+      "grad_norm": 0.28765774184505194,
+      "learning_rate": 5.9647307717484143e-05,
+      "loss": 1.0529,
+      "step": 3343
+    },
+    {
+      "epoch": 0.6433669021494063,
+      "grad_norm": 0.36221757232170504,
+      "learning_rate": 5.959029401352262e-05,
+      "loss": 1.106,
+      "step": 3344
+    },
+    {
+      "epoch": 0.6435592965579437,
+      "grad_norm": 0.322293932757898,
+      "learning_rate": 5.9533296004225235e-05,
+      "loss": 1.0424,
+      "step": 3345
+    },
+    {
+      "epoch": 0.6437516909664813,
+      "grad_norm": 0.3342649613279946,
+      "learning_rate": 5.947631371172942e-05,
+      "loss": 1.0927,
+      "step": 3346
+    },
+    {
+      "epoch": 0.6439440853750188,
+      "grad_norm": 0.2914043254352665,
+      "learning_rate": 5.941934715816642e-05,
+      "loss": 1.0099,
+      "step": 3347
+    },
+    {
+      "epoch": 0.6441364797835563,
+      "grad_norm": 0.32566465982223153,
+      "learning_rate": 5.936239636566137e-05,
+      "loss": 1.0321,
+      "step": 3348
+    },
+    {
+      "epoch": 0.6443288741920938,
+      "grad_norm": 0.31998849983665684,
+      "learning_rate": 5.930546135633327e-05,
+      "loss": 1.0827,
+      "step": 3349
+    },
+    {
+      "epoch": 0.6445212686006313,
+      "grad_norm": 0.32359460296217146,
+      "learning_rate": 5.924854215229508e-05,
+      "loss": 1.0949,
+      "step": 3350
+    },
+    {
+      "epoch": 0.6447136630091688,
+      "grad_norm": 0.28767243369843587,
+      "learning_rate": 5.91916387756535e-05,
+      "loss": 1.0691,
+      "step": 3351
+    },
+    {
+      "epoch": 0.6449060574177063,
+      "grad_norm": 0.371534559297127,
+      "learning_rate": 5.9134751248509236e-05,
+      "loss": 1.1254,
+      "step": 3352
+    },
+    {
+      "epoch": 0.6450984518262438,
+      "grad_norm": 0.4585563484531112,
+      "learning_rate": 5.9077879592956675e-05,
+      "loss": 1.1102,
+      "step": 3353
+    },
+    {
+      "epoch": 0.6452908462347813,
+      "grad_norm": 0.3170741341399339,
+      "learning_rate": 5.902102383108414e-05,
+      "loss": 1.1067,
+      "step": 3354
+    },
+    {
+      "epoch": 0.6454832406433189,
+      "grad_norm": 0.3122303871654554,
+      "learning_rate": 5.896418398497377e-05,
+      "loss": 1.0845,
+      "step": 3355
+    },
+    {
+      "epoch": 0.6456756350518563,
+      "grad_norm": 0.4088156798252746,
+      "learning_rate": 5.890736007670144e-05,
+      "loss": 1.1419,
+      "step": 3356
+    },
+    {
+      "epoch": 0.6458680294603938,
+      "grad_norm": 0.34816947482301286,
+      "learning_rate": 5.8850552128336954e-05,
+      "loss": 1.0364,
+      "step": 3357
+    },
+    {
+      "epoch": 0.6460604238689313,
+      "grad_norm": 0.3482682499654772,
+      "learning_rate": 5.879376016194387e-05,
+      "loss": 1.0198,
+      "step": 3358
+    },
+    {
+      "epoch": 0.6462528182774688,
+      "grad_norm": 0.28158505635506154,
+      "learning_rate": 5.873698419957952e-05,
+      "loss": 0.9989,
+      "step": 3359
+    },
+    {
+      "epoch": 0.6464452126860063,
+      "grad_norm": 0.2686635351483379,
+      "learning_rate": 5.868022426329505e-05,
+      "loss": 0.9743,
+      "step": 3360
+    },
+    {
+      "epoch": 0.6466376070945438,
+      "grad_norm": 0.305189263070437,
+      "learning_rate": 5.862348037513532e-05,
+      "loss": 1.1192,
+      "step": 3361
+    },
+    {
+      "epoch": 0.6468300015030813,
+      "grad_norm": 0.27525600757420376,
+      "learning_rate": 5.856675255713905e-05,
+      "loss": 1.0303,
+      "step": 3362
+    },
+    {
+      "epoch": 0.6470223959116188,
+      "grad_norm": 0.34737589184864753,
+      "learning_rate": 5.851004083133862e-05,
+      "loss": 1.0955,
+      "step": 3363
+    },
+    {
+      "epoch": 0.6472147903201563,
+      "grad_norm": 0.319264546088472,
+      "learning_rate": 5.845334521976027e-05,
+      "loss": 1.0478,
+      "step": 3364
+    },
+    {
+      "epoch": 0.6474071847286939,
+      "grad_norm": 0.30090464359742247,
+      "learning_rate": 5.8396665744423885e-05,
+      "loss": 1.0532,
+      "step": 3365
+    },
+    {
+      "epoch": 0.6475995791372313,
+      "grad_norm": 0.3937634631338097,
+      "learning_rate": 5.8340002427343164e-05,
+      "loss": 1.1079,
+      "step": 3366
+    },
+    {
+      "epoch": 0.6477919735457688,
+      "grad_norm": 0.325883695652783,
+      "learning_rate": 5.828335529052541e-05,
+      "loss": 1.0908,
+      "step": 3367
+    },
+    {
+      "epoch": 0.6479843679543064,
+      "grad_norm": 0.38670909174801216,
+      "learning_rate": 5.822672435597172e-05,
+      "loss": 1.0535,
+      "step": 3368
+    },
+    {
+      "epoch": 0.6481767623628438,
+      "grad_norm": 0.29345715629570895,
+      "learning_rate": 5.817010964567702e-05,
+      "loss": 1.0534,
+      "step": 3369
+    },
+    {
+      "epoch": 0.6483691567713813,
+      "grad_norm": 0.29964248551292266,
+      "learning_rate": 5.811351118162969e-05,
+      "loss": 1.0095,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6485615511799189,
+      "grad_norm": 0.31196840409199006,
+      "learning_rate": 5.8056928985811963e-05,
+      "loss": 1.0901,
+      "step": 3371
+    },
+    {
+      "epoch": 0.6487539455884563,
+      "grad_norm": 0.45743176378003714,
+      "learning_rate": 5.8000363080199736e-05,
+      "loss": 1.1343,
+      "step": 3372
+    },
+    {
+      "epoch": 0.6489463399969938,
+      "grad_norm": 0.2975454931728827,
+      "learning_rate": 5.79438134867625e-05,
+      "loss": 1.0949,
+      "step": 3373
+    },
+    {
+      "epoch": 0.6491387344055314,
+      "grad_norm": 0.30430171352238705,
+      "learning_rate": 5.7887280227463484e-05,
+      "loss": 1.1414,
+      "step": 3374
+    },
+    {
+      "epoch": 0.6493311288140688,
+      "grad_norm": 0.2816355779911321,
+      "learning_rate": 5.783076332425956e-05,
+      "loss": 1.0915,
+      "step": 3375
+    },
+    {
+      "epoch": 0.6495235232226063,
+      "grad_norm": 0.33298574715222334,
+      "learning_rate": 5.777426279910125e-05,
+      "loss": 1.0832,
+      "step": 3376
+    },
+    {
+      "epoch": 0.6497159176311439,
+      "grad_norm": 0.31926373221196686,
+      "learning_rate": 5.771777867393274e-05,
+      "loss": 1.1044,
+      "step": 3377
+    },
+    {
+      "epoch": 0.6499083120396814,
+      "grad_norm": 0.33555709307971854,
+      "learning_rate": 5.766131097069174e-05,
+      "loss": 1.1369,
+      "step": 3378
+    },
+    {
+      "epoch": 0.6501007064482188,
+      "grad_norm": 0.38930249397283606,
+      "learning_rate": 5.760485971130969e-05,
+      "loss": 1.16,
+      "step": 3379
+    },
+    {
+      "epoch": 0.6502931008567564,
+      "grad_norm": 0.2709197697942827,
+      "learning_rate": 5.75484249177116e-05,
+      "loss": 1.1145,
+      "step": 3380
+    },
+    {
+      "epoch": 0.6504854952652939,
+      "grad_norm": 0.2839442611120986,
+      "learning_rate": 5.749200661181611e-05,
+      "loss": 1.0224,
+      "step": 3381
+    },
+    {
+      "epoch": 0.6506778896738313,
+      "grad_norm": 0.32321847117929114,
+      "learning_rate": 5.743560481553547e-05,
+      "loss": 0.9826,
+      "step": 3382
+    },
+    {
+      "epoch": 0.6508702840823689,
+      "grad_norm": 0.3210531190288898,
+      "learning_rate": 5.737921955077541e-05,
+      "loss": 1.0741,
+      "step": 3383
+    },
+    {
+      "epoch": 0.6510626784909064,
+      "grad_norm": 0.3433622705707647,
+      "learning_rate": 5.732285083943536e-05,
+      "loss": 1.0504,
+      "step": 3384
+    },
+    {
+      "epoch": 0.6512550728994438,
+      "grad_norm": 0.3621977358194807,
+      "learning_rate": 5.7266498703408325e-05,
+      "loss": 0.962,
+      "step": 3385
+    },
+    {
+      "epoch": 0.6514474673079813,
+      "grad_norm": 0.31938858392337316,
+      "learning_rate": 5.721016316458068e-05,
+      "loss": 1.0408,
+      "step": 3386
+    },
+    {
+      "epoch": 0.6516398617165189,
+      "grad_norm": 0.3011130459759042,
+      "learning_rate": 5.715384424483268e-05,
+      "loss": 1.062,
+      "step": 3387
+    },
+    {
+      "epoch": 0.6518322561250564,
+      "grad_norm": 0.2710309344278669,
+      "learning_rate": 5.7097541966037816e-05,
+      "loss": 1.0639,
+      "step": 3388
+    },
+    {
+      "epoch": 0.6520246505335938,
+      "grad_norm": 0.3672832366698096,
+      "learning_rate": 5.7041256350063286e-05,
+      "loss": 1.1049,
+      "step": 3389
+    },
+    {
+      "epoch": 0.6522170449421314,
+      "grad_norm": 0.3618610991051004,
+      "learning_rate": 5.698498741876982e-05,
+      "loss": 1.0615,
+      "step": 3390
+    },
+    {
+      "epoch": 0.6524094393506689,
+      "grad_norm": 0.317243270274209,
+      "learning_rate": 5.692873519401154e-05,
+      "loss": 1.0592,
+      "step": 3391
+    },
+    {
+      "epoch": 0.6526018337592063,
+      "grad_norm": 0.3164909340877395,
+      "learning_rate": 5.6872499697636195e-05,
+      "loss": 1.0266,
+      "step": 3392
+    },
+    {
+      "epoch": 0.6527942281677439,
+      "grad_norm": 0.30335243238724957,
+      "learning_rate": 5.681628095148501e-05,
+      "loss": 1.0562,
+      "step": 3393
+    },
+    {
+      "epoch": 0.6529866225762814,
+      "grad_norm": 0.30089979251518684,
+      "learning_rate": 5.6760078977392706e-05,
+      "loss": 0.9966,
+      "step": 3394
+    },
+    {
+      "epoch": 0.6531790169848188,
+      "grad_norm": 0.340657217900467,
+      "learning_rate": 5.67038937971875e-05,
+      "loss": 0.9891,
+      "step": 3395
+    },
+    {
+      "epoch": 0.6533714113933564,
+      "grad_norm": 0.3019685620085431,
+      "learning_rate": 5.664772543269101e-05,
+      "loss": 1.0971,
+      "step": 3396
+    },
+    {
+      "epoch": 0.6535638058018939,
+      "grad_norm": 0.3114039455358227,
+      "learning_rate": 5.659157390571842e-05,
+      "loss": 1.0454,
+      "step": 3397
+    },
+    {
+      "epoch": 0.6537562002104313,
+      "grad_norm": 0.339025833106066,
+      "learning_rate": 5.653543923807832e-05,
+      "loss": 1.0296,
+      "step": 3398
+    },
+    {
+      "epoch": 0.6539485946189689,
+      "grad_norm": 0.3494007175989049,
+      "learning_rate": 5.6479321451572784e-05,
+      "loss": 1.0859,
+      "step": 3399
+    },
+    {
+      "epoch": 0.6541409890275064,
+      "grad_norm": 0.26910039043189565,
+      "learning_rate": 5.6423220567997316e-05,
+      "loss": 1.0297,
+      "step": 3400
+    },
+    {
+      "epoch": 0.6543333834360439,
+      "grad_norm": 0.633069797895338,
+      "learning_rate": 5.636713660914087e-05,
+      "loss": 1.0332,
+      "step": 3401
+    },
+    {
+      "epoch": 0.6545257778445814,
+      "grad_norm": 0.3077163804044054,
+      "learning_rate": 5.631106959678575e-05,
+      "loss": 0.9649,
+      "step": 3402
+    },
+    {
+      "epoch": 0.6547181722531189,
+      "grad_norm": 0.28593647186853255,
+      "learning_rate": 5.625501955270776e-05,
+      "loss": 0.9928,
+      "step": 3403
+    },
+    {
+      "epoch": 0.6549105666616564,
+      "grad_norm": 0.28051843679710936,
+      "learning_rate": 5.619898649867612e-05,
+      "loss": 1.0593,
+      "step": 3404
+    },
+    {
+      "epoch": 0.655102961070194,
+      "grad_norm": 0.32209970322772813,
+      "learning_rate": 5.614297045645339e-05,
+      "loss": 1.012,
+      "step": 3405
+    },
+    {
+      "epoch": 0.6552953554787314,
+      "grad_norm": 0.2749127172078766,
+      "learning_rate": 5.6086971447795624e-05,
+      "loss": 1.0729,
+      "step": 3406
+    },
+    {
+      "epoch": 0.6554877498872689,
+      "grad_norm": 0.2986434152466038,
+      "learning_rate": 5.603098949445209e-05,
+      "loss": 1.0294,
+      "step": 3407
+    },
+    {
+      "epoch": 0.6556801442958065,
+      "grad_norm": 0.33526433260027283,
+      "learning_rate": 5.597502461816557e-05,
+      "loss": 1.0707,
+      "step": 3408
+    },
+    {
+      "epoch": 0.6558725387043439,
+      "grad_norm": 0.35657004141077403,
+      "learning_rate": 5.591907684067221e-05,
+      "loss": 1.1018,
+      "step": 3409
+    },
+    {
+      "epoch": 0.6560649331128814,
+      "grad_norm": 0.49097401316341893,
+      "learning_rate": 5.586314618370145e-05,
+      "loss": 1.1616,
+      "step": 3410
+    },
+    {
+      "epoch": 0.6562573275214189,
+      "grad_norm": 0.3289508251553119,
+      "learning_rate": 5.580723266897616e-05,
+      "loss": 1.0972,
+      "step": 3411
+    },
+    {
+      "epoch": 0.6564497219299564,
+      "grad_norm": 0.2982384439715662,
+      "learning_rate": 5.575133631821243e-05,
+      "loss": 1.0481,
+      "step": 3412
+    },
+    {
+      "epoch": 0.6566421163384939,
+      "grad_norm": 0.3354765393620317,
+      "learning_rate": 5.5695457153119804e-05,
+      "loss": 1.0972,
+      "step": 3413
+    },
+    {
+      "epoch": 0.6568345107470314,
+      "grad_norm": 0.3580272923048353,
+      "learning_rate": 5.563959519540114e-05,
+      "loss": 1.0572,
+      "step": 3414
+    },
+    {
+      "epoch": 0.6570269051555689,
+      "grad_norm": 0.27018878384234063,
+      "learning_rate": 5.5583750466752435e-05,
+      "loss": 1.0714,
+      "step": 3415
+    },
+    {
+      "epoch": 0.6572192995641064,
+      "grad_norm": 0.36090736243022997,
+      "learning_rate": 5.552792298886334e-05,
+      "loss": 1.0173,
+      "step": 3416
+    },
+    {
+      "epoch": 0.6574116939726439,
+      "grad_norm": 0.27242805653495944,
+      "learning_rate": 5.547211278341646e-05,
+      "loss": 1.0593,
+      "step": 3417
+    },
+    {
+      "epoch": 0.6576040883811815,
+      "grad_norm": 0.2957701975950919,
+      "learning_rate": 5.541631987208788e-05,
+      "loss": 1.0237,
+      "step": 3418
+    },
+    {
+      "epoch": 0.6577964827897189,
+      "grad_norm": 0.3007347148460491,
+      "learning_rate": 5.536054427654698e-05,
+      "loss": 1.0863,
+      "step": 3419
+    },
+    {
+      "epoch": 0.6579888771982564,
+      "grad_norm": 0.5546170878502891,
+      "learning_rate": 5.530478601845623e-05,
+      "loss": 0.9821,
+      "step": 3420
+    },
+    {
+      "epoch": 0.658181271606794,
+      "grad_norm": 0.2968469323713588,
+      "learning_rate": 5.52490451194716e-05,
+      "loss": 1.1035,
+      "step": 3421
+    },
+    {
+      "epoch": 0.6583736660153314,
+      "grad_norm": 0.32686786577907967,
+      "learning_rate": 5.5193321601242156e-05,
+      "loss": 1.078,
+      "step": 3422
+    },
+    {
+      "epoch": 0.6585660604238689,
+      "grad_norm": 0.32951964267826644,
+      "learning_rate": 5.513761548541031e-05,
+      "loss": 1.1353,
+      "step": 3423
+    },
+    {
+      "epoch": 0.6587584548324065,
+      "grad_norm": 0.2980474553296739,
+      "learning_rate": 5.5081926793611694e-05,
+      "loss": 1.0147,
+      "step": 3424
+    },
+    {
+      "epoch": 0.6589508492409439,
+      "grad_norm": 0.28767968988064563,
+      "learning_rate": 5.502625554747508e-05,
+      "loss": 1.1066,
+      "step": 3425
+    },
+    {
+      "epoch": 0.6591432436494814,
+      "grad_norm": 0.35069418962184845,
+      "learning_rate": 5.497060176862259e-05,
+      "loss": 1.0916,
+      "step": 3426
+    },
+    {
+      "epoch": 0.659335638058019,
+      "grad_norm": 0.30115289278766205,
+      "learning_rate": 5.4914965478669475e-05,
+      "loss": 1.1414,
+      "step": 3427
+    },
+    {
+      "epoch": 0.6595280324665564,
+      "grad_norm": 0.31274459691709966,
+      "learning_rate": 5.485934669922428e-05,
+      "loss": 1.0208,
+      "step": 3428
+    },
+    {
+      "epoch": 0.6597204268750939,
+      "grad_norm": 0.2979652159762174,
+      "learning_rate": 5.480374545188866e-05,
+      "loss": 1.067,
+      "step": 3429
+    },
+    {
+      "epoch": 0.6599128212836315,
+      "grad_norm": 0.2778377680927135,
+      "learning_rate": 5.4748161758257544e-05,
+      "loss": 1.071,
+      "step": 3430
+    },
+    {
+      "epoch": 0.660105215692169,
+      "grad_norm": 0.3373862880581623,
+      "learning_rate": 5.469259563991893e-05,
+      "loss": 1.0109,
+      "step": 3431
+    },
+    {
+      "epoch": 0.6602976101007064,
+      "grad_norm": 0.3287814288371032,
+      "learning_rate": 5.4637047118454096e-05,
+      "loss": 1.0454,
+      "step": 3432
+    },
+    {
+      "epoch": 0.660490004509244,
+      "grad_norm": 0.3289065610296748,
+      "learning_rate": 5.458151621543743e-05,
+      "loss": 0.989,
+      "step": 3433
+    },
+    {
+      "epoch": 0.6606823989177815,
+      "grad_norm": 0.3021531338177011,
+      "learning_rate": 5.4526002952436526e-05,
+      "loss": 1.0527,
+      "step": 3434
+    },
+    {
+      "epoch": 0.6608747933263189,
+      "grad_norm": 0.4688877912363289,
+      "learning_rate": 5.447050735101211e-05,
+      "loss": 1.1106,
+      "step": 3435
+    },
+    {
+      "epoch": 0.6610671877348565,
+      "grad_norm": 0.3150331400644717,
+      "learning_rate": 5.441502943271797e-05,
+      "loss": 1.1096,
+      "step": 3436
+    },
+    {
+      "epoch": 0.661259582143394,
+      "grad_norm": 0.3228519058353967,
+      "learning_rate": 5.4359569219101114e-05,
+      "loss": 1.0497,
+      "step": 3437
+    },
+    {
+      "epoch": 0.6614519765519314,
+      "grad_norm": 0.35287891576341984,
+      "learning_rate": 5.4304126731701665e-05,
+      "loss": 1.0826,
+      "step": 3438
+    },
+    {
+      "epoch": 0.6616443709604689,
+      "grad_norm": 0.28556050017271517,
+      "learning_rate": 5.424870199205283e-05,
+      "loss": 1.0472,
+      "step": 3439
+    },
+    {
+      "epoch": 0.6618367653690065,
+      "grad_norm": 0.3707419496719028,
+      "learning_rate": 5.4193295021681e-05,
+      "loss": 1.0794,
+      "step": 3440
+    },
+    {
+      "epoch": 0.662029159777544,
+      "grad_norm": 0.3303223526555379,
+      "learning_rate": 5.413790584210551e-05,
+      "loss": 1.0782,
+      "step": 3441
+    },
+    {
+      "epoch": 0.6622215541860814,
+      "grad_norm": 0.3031476381153801,
+      "learning_rate": 5.408253447483892e-05,
+      "loss": 1.0633,
+      "step": 3442
+    },
+    {
+      "epoch": 0.662413948594619,
+      "grad_norm": 0.2584004580094845,
+      "learning_rate": 5.4027180941386877e-05,
+      "loss": 1.0665,
+      "step": 3443
+    },
+    {
+      "epoch": 0.6626063430031565,
+      "grad_norm": 0.3619300719303694,
+      "learning_rate": 5.397184526324792e-05,
+      "loss": 1.0445,
+      "step": 3444
+    },
+    {
+      "epoch": 0.6627987374116939,
+      "grad_norm": 0.37435043793646805,
+      "learning_rate": 5.391652746191398e-05,
+      "loss": 1.0261,
+      "step": 3445
+    },
+    {
+      "epoch": 0.6629911318202315,
+      "grad_norm": 0.33021817770856465,
+      "learning_rate": 5.3861227558869695e-05,
+      "loss": 1.1074,
+      "step": 3446
+    },
+    {
+      "epoch": 0.663183526228769,
+      "grad_norm": 0.28730902128760144,
+      "learning_rate": 5.3805945575592975e-05,
+      "loss": 1.0366,
+      "step": 3447
+    },
+    {
+      "epoch": 0.6633759206373064,
+      "grad_norm": 0.2989894849763508,
+      "learning_rate": 5.3750681533554735e-05,
+      "loss": 1.0828,
+      "step": 3448
+    },
+    {
+      "epoch": 0.663568315045844,
+      "grad_norm": 0.3450384306105209,
+      "learning_rate": 5.369543545421883e-05,
+      "loss": 0.9548,
+      "step": 3449
+    },
+    {
+      "epoch": 0.6637607094543815,
+      "grad_norm": 0.28340658318717854,
+      "learning_rate": 5.3640207359042224e-05,
+      "loss": 0.9798,
+      "step": 3450
+    },
+    {
+      "epoch": 0.663953103862919,
+      "grad_norm": 0.29969066035073144,
+      "learning_rate": 5.358499726947488e-05,
+      "loss": 1.0045,
+      "step": 3451
+    },
+    {
+      "epoch": 0.6641454982714565,
+      "grad_norm": 0.2877993635804976,
+      "learning_rate": 5.352980520695974e-05,
+      "loss": 1.0319,
+      "step": 3452
+    },
+    {
+      "epoch": 0.664337892679994,
+      "grad_norm": 0.3597380567808401,
+      "learning_rate": 5.347463119293283e-05,
+      "loss": 0.9405,
+      "step": 3453
+    },
+    {
+      "epoch": 0.6645302870885315,
+      "grad_norm": 0.32679688968707826,
+      "learning_rate": 5.3419475248823014e-05,
+      "loss": 1.147,
+      "step": 3454
+    },
+    {
+      "epoch": 0.664722681497069,
+      "grad_norm": 0.44192355010270074,
+      "learning_rate": 5.3364337396052265e-05,
+      "loss": 1.0927,
+      "step": 3455
+    },
+    {
+      "epoch": 0.6649150759056065,
+      "grad_norm": 0.30312076736198956,
+      "learning_rate": 5.3309217656035496e-05,
+      "loss": 1.0591,
+      "step": 3456
+    },
+    {
+      "epoch": 0.665107470314144,
+      "grad_norm": 0.3016535358737794,
+      "learning_rate": 5.3254116050180555e-05,
+      "loss": 1.012,
+      "step": 3457
+    },
+    {
+      "epoch": 0.6652998647226815,
+      "grad_norm": 0.3318437075038821,
+      "learning_rate": 5.31990325998883e-05,
+      "loss": 1.0944,
+      "step": 3458
+    },
+    {
+      "epoch": 0.665492259131219,
+      "grad_norm": 0.30450078277603604,
+      "learning_rate": 5.314396732655253e-05,
+      "loss": 1.0385,
+      "step": 3459
+    },
+    {
+      "epoch": 0.6656846535397565,
+      "grad_norm": 0.2715940306666047,
+      "learning_rate": 5.3088920251559895e-05,
+      "loss": 1.1106,
+      "step": 3460
+    },
+    {
+      "epoch": 0.665877047948294,
+      "grad_norm": 0.31196214500018166,
+      "learning_rate": 5.303389139629007e-05,
+      "loss": 1.1233,
+      "step": 3461
+    },
+    {
+      "epoch": 0.6660694423568315,
+      "grad_norm": 0.416855790692469,
+      "learning_rate": 5.297888078211564e-05,
+      "loss": 1.0223,
+      "step": 3462
+    },
+    {
+      "epoch": 0.666261836765369,
+      "grad_norm": 0.3102542867239069,
+      "learning_rate": 5.292388843040208e-05,
+      "loss": 1.052,
+      "step": 3463
+    },
+    {
+      "epoch": 0.6664542311739065,
+      "grad_norm": 0.28655538199164865,
+      "learning_rate": 5.286891436250785e-05,
+      "loss": 1.0021,
+      "step": 3464
+    },
+    {
+      "epoch": 0.666646625582444,
+      "grad_norm": 0.2959347403424418,
+      "learning_rate": 5.281395859978414e-05,
+      "loss": 1.107,
+      "step": 3465
+    },
+    {
+      "epoch": 0.6668390199909815,
+      "grad_norm": 0.3403928199945639,
+      "learning_rate": 5.2759021163575186e-05,
+      "loss": 1.0195,
+      "step": 3466
+    },
+    {
+      "epoch": 0.667031414399519,
+      "grad_norm": 0.29092788411017356,
+      "learning_rate": 5.27041020752181e-05,
+      "loss": 1.1889,
+      "step": 3467
+    },
+    {
+      "epoch": 0.6672238088080565,
+      "grad_norm": 0.3396006773293743,
+      "learning_rate": 5.2649201356042696e-05,
+      "loss": 1.1107,
+      "step": 3468
+    },
+    {
+      "epoch": 0.667416203216594,
+      "grad_norm": 0.3535984837643357,
+      "learning_rate": 5.259431902737194e-05,
+      "loss": 1.0237,
+      "step": 3469
+    },
+    {
+      "epoch": 0.6676085976251315,
+      "grad_norm": 0.27173861392264975,
+      "learning_rate": 5.2539455110521385e-05,
+      "loss": 1.1442,
+      "step": 3470
+    },
+    {
+      "epoch": 0.667800992033669,
+      "grad_norm": 0.33403624635639617,
+      "learning_rate": 5.248460962679958e-05,
+      "loss": 1.0057,
+      "step": 3471
+    },
+    {
+      "epoch": 0.6679933864422065,
+      "grad_norm": 0.39720732637496525,
+      "learning_rate": 5.2429782597507905e-05,
+      "loss": 1.0161,
+      "step": 3472
+    },
+    {
+      "epoch": 0.668185780850744,
+      "grad_norm": 0.31011479046520346,
+      "learning_rate": 5.237497404394044e-05,
+      "loss": 1.0047,
+      "step": 3473
+    },
+    {
+      "epoch": 0.6683781752592816,
+      "grad_norm": 0.2985203849976459,
+      "learning_rate": 5.232018398738435e-05,
+      "loss": 1.0197,
+      "step": 3474
+    },
+    {
+      "epoch": 0.668570569667819,
+      "grad_norm": 0.31537123268295464,
+      "learning_rate": 5.2265412449119355e-05,
+      "loss": 1.0304,
+      "step": 3475
+    },
+    {
+      "epoch": 0.6687629640763565,
+      "grad_norm": 0.33981738740965567,
+      "learning_rate": 5.221065945041811e-05,
+      "loss": 1.0955,
+      "step": 3476
+    },
+    {
+      "epoch": 0.6689553584848941,
+      "grad_norm": 0.35608856289475144,
+      "learning_rate": 5.215592501254609e-05,
+      "loss": 1.0608,
+      "step": 3477
+    },
+    {
+      "epoch": 0.6691477528934315,
+      "grad_norm": 0.2864630861142581,
+      "learning_rate": 5.2101209156761465e-05,
+      "loss": 1.1058,
+      "step": 3478
+    },
+    {
+      "epoch": 0.669340147301969,
+      "grad_norm": 0.3163995990800733,
+      "learning_rate": 5.2046511904315265e-05,
+      "loss": 1.0911,
+      "step": 3479
+    },
+    {
+      "epoch": 0.6695325417105066,
+      "grad_norm": 0.3721352096299449,
+      "learning_rate": 5.1991833276451275e-05,
+      "loss": 1.0843,
+      "step": 3480
+    },
+    {
+      "epoch": 0.669724936119044,
+      "grad_norm": 0.3721352096299449,
+      "learning_rate": 5.1991833276451275e-05,
+      "loss": 1.0888,
+      "step": 3481
+    },
+    {
+      "epoch": 0.6699173305275815,
+      "grad_norm": 0.2591002969131627,
+      "learning_rate": 5.193717329440604e-05,
+      "loss": 1.0547,
+      "step": 3482
+    },
+    {
+      "epoch": 0.6701097249361191,
+      "grad_norm": 0.34860001072649993,
+      "learning_rate": 5.188253197940889e-05,
+      "loss": 1.0656,
+      "step": 3483
+    },
+    {
+      "epoch": 0.6703021193446566,
+      "grad_norm": 0.37402840864169773,
+      "learning_rate": 5.182790935268185e-05,
+      "loss": 1.0289,
+      "step": 3484
+    },
+    {
+      "epoch": 0.670494513753194,
+      "grad_norm": 0.32250155083277543,
+      "learning_rate": 5.177330543543971e-05,
+      "loss": 0.9671,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6706869081617316,
+      "grad_norm": 0.34620336151375863,
+      "learning_rate": 5.171872024889004e-05,
+      "loss": 1.049,
+      "step": 3486
+    },
+    {
+      "epoch": 0.6708793025702691,
+      "grad_norm": 0.26336357452895937,
+      "learning_rate": 5.1664153814233064e-05,
+      "loss": 0.9925,
+      "step": 3487
+    },
+    {
+      "epoch": 0.6710716969788065,
+      "grad_norm": 0.317969382613369,
+      "learning_rate": 5.160960615266178e-05,
+      "loss": 0.9775,
+      "step": 3488
+    },
+    {
+      "epoch": 0.6712640913873441,
+      "grad_norm": 0.3359508723889803,
+      "learning_rate": 5.155507728536191e-05,
+      "loss": 1.0986,
+      "step": 3489
+    },
+    {
+      "epoch": 0.6714564857958816,
+      "grad_norm": 0.32933401454688754,
+      "learning_rate": 5.150056723351173e-05,
+      "loss": 0.9926,
+      "step": 3490
+    },
+    {
+      "epoch": 0.671648880204419,
+      "grad_norm": 0.2614247531279279,
+      "learning_rate": 5.1446076018282396e-05,
+      "loss": 1.0509,
+      "step": 3491
+    },
+    {
+      "epoch": 0.6718412746129565,
+      "grad_norm": 0.38743634026232615,
+      "learning_rate": 5.139160366083765e-05,
+      "loss": 1.0221,
+      "step": 3492
+    },
+    {
+      "epoch": 0.6720336690214941,
+      "grad_norm": 0.3017251897735763,
+      "learning_rate": 5.133715018233393e-05,
+      "loss": 0.9652,
+      "step": 3493
+    },
+    {
+      "epoch": 0.6722260634300316,
+      "grad_norm": 0.38869117469922626,
+      "learning_rate": 5.1282715603920374e-05,
+      "loss": 1.1011,
+      "step": 3494
+    },
+    {
+      "epoch": 0.672418457838569,
+      "grad_norm": 0.3224318578621361,
+      "learning_rate": 5.1228299946738655e-05,
+      "loss": 1.1207,
+      "step": 3495
+    },
+    {
+      "epoch": 0.6726108522471066,
+      "grad_norm": 0.3540349437109095,
+      "learning_rate": 5.117390323192326e-05,
+      "loss": 1.0257,
+      "step": 3496
+    },
+    {
+      "epoch": 0.6728032466556441,
+      "grad_norm": 0.3352667382460881,
+      "learning_rate": 5.111952548060126e-05,
+      "loss": 1.1285,
+      "step": 3497
+    },
+    {
+      "epoch": 0.6729956410641815,
+      "grad_norm": 0.4835129632139507,
+      "learning_rate": 5.106516671389223e-05,
+      "loss": 0.9994,
+      "step": 3498
+    },
+    {
+      "epoch": 0.6731880354727191,
+      "grad_norm": 0.5230661074481276,
+      "learning_rate": 5.101082695290865e-05,
+      "loss": 1.0477,
+      "step": 3499
+    },
+    {
+      "epoch": 0.6733804298812566,
+      "grad_norm": 0.3298560235714514,
+      "learning_rate": 5.095650621875534e-05,
+      "loss": 1.0949,
+      "step": 3500
+    },
+    {
+      "epoch": 0.673572824289794,
+      "grad_norm": 0.34964875895227787,
+      "learning_rate": 5.09022045325299e-05,
+      "loss": 1.0916,
+      "step": 3501
+    },
+    {
+      "epoch": 0.6737652186983316,
+      "grad_norm": 0.31618872130600445,
+      "learning_rate": 5.084792191532248e-05,
+      "loss": 1.0596,
+      "step": 3502
+    },
+    {
+      "epoch": 0.6739576131068691,
+      "grad_norm": 0.32154843125943255,
+      "learning_rate": 5.079365838821579e-05,
+      "loss": 0.9787,
+      "step": 3503
+    },
+    {
+      "epoch": 0.6741500075154065,
+      "grad_norm": 0.27095671650963016,
+      "learning_rate": 5.073941397228518e-05,
+      "loss": 1.0794,
+      "step": 3504
+    },
+    {
+      "epoch": 0.6743424019239441,
+      "grad_norm": 0.3240090981442025,
+      "learning_rate": 5.068518868859854e-05,
+      "loss": 1.0712,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6745347963324816,
+      "grad_norm": 0.3172136719068298,
+      "learning_rate": 5.0630982558216363e-05,
+      "loss": 1.0907,
+      "step": 3506
+    },
+    {
+      "epoch": 0.6747271907410191,
+      "grad_norm": 0.2982821252582512,
+      "learning_rate": 5.057679560219173e-05,
+      "loss": 1.0676,
+      "step": 3507
+    },
+    {
+      "epoch": 0.6749195851495566,
+      "grad_norm": 0.30414175589834497,
+      "learning_rate": 5.052262784157014e-05,
+      "loss": 0.9762,
+      "step": 3508
+    },
+    {
+      "epoch": 0.6751119795580941,
+      "grad_norm": 0.28099959197297614,
+      "learning_rate": 5.046847929738971e-05,
+      "loss": 1.0931,
+      "step": 3509
+    },
+    {
+      "epoch": 0.6753043739666316,
+      "grad_norm": 0.2699023791158635,
+      "learning_rate": 5.041434999068126e-05,
+      "loss": 1.0864,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6754967683751691,
+      "grad_norm": 0.29990433839149544,
+      "learning_rate": 5.036023994246787e-05,
+      "loss": 1.1284,
+      "step": 3511
+    },
+    {
+      "epoch": 0.6756891627837066,
+      "grad_norm": 0.3503481004861861,
+      "learning_rate": 5.0306149173765316e-05,
+      "loss": 1.0889,
+      "step": 3512
+    },
+    {
+      "epoch": 0.6758815571922441,
+      "grad_norm": 0.3040101789968439,
+      "learning_rate": 5.0252077705581755e-05,
+      "loss": 1.0654,
+      "step": 3513
+    },
+    {
+      "epoch": 0.6760739516007817,
+      "grad_norm": 0.37375892634670016,
+      "learning_rate": 5.019802555891798e-05,
+      "loss": 0.9018,
+      "step": 3514
+    },
+    {
+      "epoch": 0.6762663460093191,
+      "grad_norm": 0.31041405079337997,
+      "learning_rate": 5.014399275476721e-05,
+      "loss": 1.0327,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6764587404178566,
+      "grad_norm": 0.3816054179403178,
+      "learning_rate": 5.008997931411517e-05,
+      "loss": 1.0226,
+      "step": 3516
+    },
+    {
+      "epoch": 0.6766511348263942,
+      "grad_norm": 0.4172256314260427,
+      "learning_rate": 5.0035985257940024e-05,
+      "loss": 1.0311,
+      "step": 3517
+    },
+    {
+      "epoch": 0.6768435292349316,
+      "grad_norm": 0.2819875109896288,
+      "learning_rate": 4.998201060721253e-05,
+      "loss": 1.0974,
+      "step": 3518
+    },
+    {
+      "epoch": 0.6770359236434691,
+      "grad_norm": 0.3468176654535024,
+      "learning_rate": 4.992805538289571e-05,
+      "loss": 1.097,
+      "step": 3519
+    },
+    {
+      "epoch": 0.6772283180520066,
+      "grad_norm": 0.3037974777451656,
+      "learning_rate": 4.98741196059452e-05,
+      "loss": 0.94,
+      "step": 3520
+    },
+    {
+      "epoch": 0.6774207124605441,
+      "grad_norm": 0.26826748104198406,
+      "learning_rate": 4.982020329730904e-05,
+      "loss": 1.0274,
+      "step": 3521
+    },
+    {
+      "epoch": 0.6776131068690816,
+      "grad_norm": 0.3040532996708861,
+      "learning_rate": 4.97663064779277e-05,
+      "loss": 1.0814,
+      "step": 3522
+    },
+    {
+      "epoch": 0.6778055012776191,
+      "grad_norm": 0.3539878059616163,
+      "learning_rate": 4.971242916873412e-05,
+      "loss": 0.9866,
+      "step": 3523
+    },
+    {
+      "epoch": 0.6779978956861566,
+      "grad_norm": 0.2927001674477416,
+      "learning_rate": 4.965857139065354e-05,
+      "loss": 1.0578,
+      "step": 3524
+    },
+    {
+      "epoch": 0.6781902900946941,
+      "grad_norm": 0.4008429496385881,
+      "learning_rate": 4.960473316460376e-05,
+      "loss": 1.0021,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6783826845032316,
+      "grad_norm": 0.34462676651482826,
+      "learning_rate": 4.955091451149495e-05,
+      "loss": 1.0711,
+      "step": 3526
+    },
+    {
+      "epoch": 0.6785750789117692,
+      "grad_norm": 0.40334991637959616,
+      "learning_rate": 4.9497115452229534e-05,
+      "loss": 1.1514,
+      "step": 3527
+    },
+    {
+      "epoch": 0.6787674733203066,
+      "grad_norm": 0.3433344075847281,
+      "learning_rate": 4.944333600770261e-05,
+      "loss": 0.9917,
+      "step": 3528
+    },
+    {
+      "epoch": 0.6789598677288441,
+      "grad_norm": 0.30932715422008966,
+      "learning_rate": 4.938957619880138e-05,
+      "loss": 1.0267,
+      "step": 3529
+    },
+    {
+      "epoch": 0.6791522621373817,
+      "grad_norm": 0.3747694416111552,
+      "learning_rate": 4.933583604640557e-05,
+      "loss": 1.0406,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6793446565459191,
+      "grad_norm": 0.39469199866750787,
+      "learning_rate": 4.928211557138728e-05,
+      "loss": 0.9877,
+      "step": 3531
+    },
+    {
+      "epoch": 0.6795370509544566,
+      "grad_norm": 0.35919781003192447,
+      "learning_rate": 4.922841479461083e-05,
+      "loss": 0.98,
+      "step": 3532
+    },
+    {
+      "epoch": 0.6797294453629942,
+      "grad_norm": 0.3120814571239843,
+      "learning_rate": 4.917473373693304e-05,
+      "loss": 1.014,
+      "step": 3533
+    },
+    {
+      "epoch": 0.6799218397715316,
+      "grad_norm": 0.39006451101516376,
+      "learning_rate": 4.912107241920302e-05,
+      "loss": 1.0582,
+      "step": 3534
+    },
+    {
+      "epoch": 0.6801142341800691,
+      "grad_norm": 0.35400627449530375,
+      "learning_rate": 4.906743086226218e-05,
+      "loss": 1.0813,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6803066285886067,
+      "grad_norm": 0.3158028565257898,
+      "learning_rate": 4.901380908694434e-05,
+      "loss": 1.0287,
+      "step": 3536
+    },
+    {
+      "epoch": 0.6804990229971442,
+      "grad_norm": 0.4489557526750681,
+      "learning_rate": 4.8960207114075495e-05,
+      "loss": 1.0546,
+      "step": 3537
+    },
+    {
+      "epoch": 0.6806914174056816,
+      "grad_norm": 0.33445995832322456,
+      "learning_rate": 4.890662496447407e-05,
+      "loss": 0.9672,
+      "step": 3538
+    },
+    {
+      "epoch": 0.6808838118142192,
+      "grad_norm": 0.36170982280497127,
+      "learning_rate": 4.885306265895076e-05,
+      "loss": 0.9894,
+      "step": 3539
+    },
+    {
+      "epoch": 0.6810762062227567,
+      "grad_norm": 0.33625501551835196,
+      "learning_rate": 4.879952021830856e-05,
+      "loss": 1.1454,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6812686006312941,
+      "grad_norm": 0.328461528299424,
+      "learning_rate": 4.8745997663342755e-05,
+      "loss": 1.0563,
+      "step": 3541
+    },
+    {
+      "epoch": 0.6814609950398317,
+      "grad_norm": 0.3550336805649042,
+      "learning_rate": 4.8692495014840825e-05,
+      "loss": 1.0457,
+      "step": 3542
+    },
+    {
+      "epoch": 0.6816533894483692,
+      "grad_norm": 0.33109470753925446,
+      "learning_rate": 4.863901229358261e-05,
+      "loss": 1.0013,
+      "step": 3543
+    },
+    {
+      "epoch": 0.6818457838569066,
+      "grad_norm": 0.289906702426852,
+      "learning_rate": 4.858554952034019e-05,
+      "loss": 1.0227,
+      "step": 3544
+    },
+    {
+      "epoch": 0.6820381782654441,
+      "grad_norm": 0.2893291582895103,
+      "learning_rate": 4.853210671587789e-05,
+      "loss": 1.0507,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6822305726739817,
+      "grad_norm": 0.39603861748236213,
+      "learning_rate": 4.8478683900952274e-05,
+      "loss": 1.0317,
+      "step": 3546
+    },
+    {
+      "epoch": 0.6824229670825191,
+      "grad_norm": 0.29220139134084816,
+      "learning_rate": 4.842528109631218e-05,
+      "loss": 0.9988,
+      "step": 3547
+    },
+    {
+      "epoch": 0.6826153614910566,
+      "grad_norm": 0.46930101398512997,
+      "learning_rate": 4.8371898322698585e-05,
+      "loss": 1.0519,
+      "step": 3548
+    },
+    {
+      "epoch": 0.6828077558995942,
+      "grad_norm": 0.3386570994592851,
+      "learning_rate": 4.831853560084477e-05,
+      "loss": 0.9849,
+      "step": 3549
+    },
+    {
+      "epoch": 0.6830001503081317,
+      "grad_norm": 0.3711144063340892,
+      "learning_rate": 4.82651929514762e-05,
+      "loss": 0.9882,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6831925447166691,
+      "grad_norm": 0.2900557521153995,
+      "learning_rate": 4.821187039531056e-05,
+      "loss": 1.0449,
+      "step": 3551
+    },
+    {
+      "epoch": 0.6833849391252067,
+      "grad_norm": 0.3049776128344479,
+      "learning_rate": 4.815856795305772e-05,
+      "loss": 1.1149,
+      "step": 3552
+    },
+    {
+      "epoch": 0.6835773335337442,
+      "grad_norm": 0.2783425655822952,
+      "learning_rate": 4.810528564541969e-05,
+      "loss": 1.0782,
+      "step": 3553
+    },
+    {
+      "epoch": 0.6837697279422816,
+      "grad_norm": 0.36016407989698923,
+      "learning_rate": 4.805202349309074e-05,
+      "loss": 0.9969,
+      "step": 3554
+    },
+    {
+      "epoch": 0.6839621223508192,
+      "grad_norm": 0.3084802855475764,
+      "learning_rate": 4.79987815167573e-05,
+      "loss": 1.062,
+      "step": 3555
+    },
+    {
+      "epoch": 0.6841545167593567,
+      "grad_norm": 0.27659298163284235,
+      "learning_rate": 4.7945559737097834e-05,
+      "loss": 1.0389,
+      "step": 3556
+    },
+    {
+      "epoch": 0.6843469111678941,
+      "grad_norm": 0.36830292708209716,
+      "learning_rate": 4.789235817478322e-05,
+      "loss": 1.051,
+      "step": 3557
+    },
+    {
+      "epoch": 0.6845393055764317,
+      "grad_norm": 0.3247416035929323,
+      "learning_rate": 4.7839176850476206e-05,
+      "loss": 1.0173,
+      "step": 3558
+    },
+    {
+      "epoch": 0.6847316999849692,
+      "grad_norm": 0.25811498348476397,
+      "learning_rate": 4.778601578483187e-05,
+      "loss": 1.0816,
+      "step": 3559
+    },
+    {
+      "epoch": 0.6849240943935067,
+      "grad_norm": 0.2975820688406356,
+      "learning_rate": 4.773287499849737e-05,
+      "loss": 0.9899,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6851164888020442,
+      "grad_norm": 0.3687199888790893,
+      "learning_rate": 4.767975451211191e-05,
+      "loss": 1.1033,
+      "step": 3561
+    },
+    {
+      "epoch": 0.6853088832105817,
+      "grad_norm": 0.308506714084704,
+      "learning_rate": 4.7626654346306923e-05,
+      "loss": 1.0434,
+      "step": 3562
+    },
+    {
+      "epoch": 0.6855012776191192,
+      "grad_norm": 0.36557095390192673,
+      "learning_rate": 4.757357452170588e-05,
+      "loss": 1.0686,
+      "step": 3563
+    },
+    {
+      "epoch": 0.6856936720276567,
+      "grad_norm": 0.2873479699893437,
+      "learning_rate": 4.752051505892438e-05,
+      "loss": 1.0846,
+      "step": 3564
+    },
+    {
+      "epoch": 0.6858860664361942,
+      "grad_norm": 0.36870745999412574,
+      "learning_rate": 4.746747597857014e-05,
+      "loss": 1.0681,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6860784608447317,
+      "grad_norm": 0.34317991543935006,
+      "learning_rate": 4.741445730124288e-05,
+      "loss": 1.0514,
+      "step": 3566
+    },
+    {
+      "epoch": 0.6862708552532693,
+      "grad_norm": 0.34686465917834397,
+      "learning_rate": 4.736145904753445e-05,
+      "loss": 0.9692,
+      "step": 3567
+    },
+    {
+      "epoch": 0.6864632496618067,
+      "grad_norm": 0.37449660756103303,
+      "learning_rate": 4.7308481238028765e-05,
+      "loss": 1.0592,
+      "step": 3568
+    },
+    {
+      "epoch": 0.6866556440703442,
+      "grad_norm": 0.2600848956522634,
+      "learning_rate": 4.725552389330182e-05,
+      "loss": 1.1187,
+      "step": 3569
+    },
+    {
+      "epoch": 0.6868480384788818,
+      "grad_norm": 0.29988467331887764,
+      "learning_rate": 4.720258703392161e-05,
+      "loss": 1.0976,
+      "step": 3570
+    },
+    {
+      "epoch": 0.6870404328874192,
+      "grad_norm": 0.3414643017885808,
+      "learning_rate": 4.7149670680448256e-05,
+      "loss": 0.989,
+      "step": 3571
+    },
+    {
+      "epoch": 0.6872328272959567,
+      "grad_norm": 0.28334956104705594,
+      "learning_rate": 4.709677485343377e-05,
+      "loss": 1.0395,
+      "step": 3572
+    },
+    {
+      "epoch": 0.6874252217044942,
+      "grad_norm": 0.32554643996706106,
+      "learning_rate": 4.704389957342237e-05,
+      "loss": 1.0765,
+      "step": 3573
+    },
+    {
+      "epoch": 0.6876176161130317,
+      "grad_norm": 0.3839251163213743,
+      "learning_rate": 4.699104486095008e-05,
+      "loss": 1.0684,
+      "step": 3574
+    },
+    {
+      "epoch": 0.6878100105215692,
+      "grad_norm": 0.2912143580861565,
+      "learning_rate": 4.6938210736545195e-05,
+      "loss": 1.0125,
+      "step": 3575
+    },
+    {
+      "epoch": 0.6880024049301067,
+      "grad_norm": 0.39430444951430527,
+      "learning_rate": 4.688539722072786e-05,
+      "loss": 1.06,
+      "step": 3576
+    },
+    {
+      "epoch": 0.6881947993386442,
+      "grad_norm": 0.5465363646831795,
+      "learning_rate": 4.6832604334010164e-05,
+      "loss": 1.0244,
+      "step": 3577
+    },
+    {
+      "epoch": 0.6883871937471817,
+      "grad_norm": 0.28907495375669995,
+      "learning_rate": 4.67798320968963e-05,
+      "loss": 1.056,
+      "step": 3578
+    },
+    {
+      "epoch": 0.6885795881557192,
+      "grad_norm": 0.31659417428817904,
+      "learning_rate": 4.672708052988239e-05,
+      "loss": 1.0531,
+      "step": 3579
+    },
+    {
+      "epoch": 0.6887719825642568,
+      "grad_norm": 0.3604903424708685,
+      "learning_rate": 4.667434965345654e-05,
+      "loss": 0.9765,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6889643769727942,
+      "grad_norm": 0.31353467027059173,
+      "learning_rate": 4.6621639488098854e-05,
+      "loss": 1.0254,
+      "step": 3581
+    },
+    {
+      "epoch": 0.6891567713813317,
+      "grad_norm": 0.3131835511266756,
+      "learning_rate": 4.656895005428127e-05,
+      "loss": 1.1014,
+      "step": 3582
+    },
+    {
+      "epoch": 0.6893491657898693,
+      "grad_norm": 0.2960827827209148,
+      "learning_rate": 4.6516281372467806e-05,
+      "loss": 1.1676,
+      "step": 3583
+    },
+    {
+      "epoch": 0.6895415601984067,
+      "grad_norm": 0.2783333735238189,
+      "learning_rate": 4.64636334631144e-05,
+      "loss": 0.9882,
+      "step": 3584
+    },
+    {
+      "epoch": 0.6897339546069442,
+      "grad_norm": 0.34259578523253326,
+      "learning_rate": 4.6411006346668776e-05,
+      "loss": 1.0231,
+      "step": 3585
+    },
+    {
+      "epoch": 0.6899263490154818,
+      "grad_norm": 0.4171859908601556,
+      "learning_rate": 4.635840004357086e-05,
+      "loss": 1.0481,
+      "step": 3586
+    },
+    {
+      "epoch": 0.6901187434240192,
+      "grad_norm": 0.3401331914954342,
+      "learning_rate": 4.630581457425222e-05,
+      "loss": 1.069,
+      "step": 3587
+    },
+    {
+      "epoch": 0.6903111378325567,
+      "grad_norm": 0.3149676949564201,
+      "learning_rate": 4.625324995913648e-05,
+      "loss": 1.141,
+      "step": 3588
+    },
+    {
+      "epoch": 0.6905035322410943,
+      "grad_norm": 0.3252719993142569,
+      "learning_rate": 4.620070621863917e-05,
+      "loss": 1.0408,
+      "step": 3589
+    },
+    {
+      "epoch": 0.6906959266496318,
+      "grad_norm": 0.29594764723436207,
+      "learning_rate": 4.614818337316759e-05,
+      "loss": 1.0119,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6908883210581692,
+      "grad_norm": 0.2833495514491127,
+      "learning_rate": 4.6095681443121066e-05,
+      "loss": 1.0631,
+      "step": 3591
+    },
+    {
+      "epoch": 0.6910807154667068,
+      "grad_norm": 0.3137384228123955,
+      "learning_rate": 4.604320044889072e-05,
+      "loss": 1.0145,
+      "step": 3592
+    },
+    {
+      "epoch": 0.6912731098752443,
+      "grad_norm": 0.3435444837394111,
+      "learning_rate": 4.599074041085958e-05,
+      "loss": 1.0179,
+      "step": 3593
+    },
+    {
+      "epoch": 0.6914655042837817,
+      "grad_norm": 0.39489853221544463,
+      "learning_rate": 4.593830134940256e-05,
+      "loss": 1.1021,
+      "step": 3594
+    },
+    {
+      "epoch": 0.6916578986923193,
+      "grad_norm": 0.27555621352680026,
+      "learning_rate": 4.588588328488629e-05,
+      "loss": 1.0369,
+      "step": 3595
+    },
+    {
+      "epoch": 0.6918502931008568,
+      "grad_norm": 0.3053879678905768,
+      "learning_rate": 4.583348623766941e-05,
+      "loss": 1.114,
+      "step": 3596
+    },
+    {
+      "epoch": 0.6920426875093942,
+      "grad_norm": 0.31209583353961534,
+      "learning_rate": 4.578111022810231e-05,
+      "loss": 0.9983,
+      "step": 3597
+    },
+    {
+      "epoch": 0.6922350819179317,
+      "grad_norm": 0.3036329270805974,
+      "learning_rate": 4.5728755276527226e-05,
+      "loss": 1.0477,
+      "step": 3598
+    },
+    {
+      "epoch": 0.6924274763264693,
+      "grad_norm": 0.2977631134446947,
+      "learning_rate": 4.567642140327823e-05,
+      "loss": 1.118,
+      "step": 3599
+    },
+    {
+      "epoch": 0.6926198707350067,
+      "grad_norm": 0.323468790462587,
+      "learning_rate": 4.562410862868123e-05,
+      "loss": 1.1009,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6928122651435442,
+      "grad_norm": 0.31469573480339924,
+      "learning_rate": 4.5571816973053825e-05,
+      "loss": 1.0514,
+      "step": 3601
+    },
+    {
+      "epoch": 0.6930046595520818,
+      "grad_norm": 0.3581819043883575,
+      "learning_rate": 4.551954645670557e-05,
+      "loss": 1.0984,
+      "step": 3602
+    },
+    {
+      "epoch": 0.6931970539606193,
+      "grad_norm": 0.283659053333895,
+      "learning_rate": 4.5467297099937625e-05,
+      "loss": 1.0603,
+      "step": 3603
+    },
+    {
+      "epoch": 0.6933894483691567,
+      "grad_norm": 0.262658386806912,
+      "learning_rate": 4.5415068923043133e-05,
+      "loss": 1.072,
+      "step": 3604
+    },
+    {
+      "epoch": 0.6935818427776943,
+      "grad_norm": 0.2886571006290498,
+      "learning_rate": 4.5362861946306936e-05,
+      "loss": 1.038,
+      "step": 3605
+    },
+    {
+      "epoch": 0.6937742371862318,
+      "grad_norm": 0.2837306873457255,
+      "learning_rate": 4.531067619000553e-05,
+      "loss": 1.0021,
+      "step": 3606
+    },
+    {
+      "epoch": 0.6939666315947692,
+      "grad_norm": 0.40396626050076123,
+      "learning_rate": 4.525851167440731e-05,
+      "loss": 1.0649,
+      "step": 3607
+    },
+    {
+      "epoch": 0.6941590260033068,
+      "grad_norm": 0.27674427213065805,
+      "learning_rate": 4.5206368419772406e-05,
+      "loss": 1.0525,
+      "step": 3608
+    },
+    {
+      "epoch": 0.6943514204118443,
+      "grad_norm": 0.34574891123886986,
+      "learning_rate": 4.5154246446352544e-05,
+      "loss": 0.9968,
+      "step": 3609
+    },
+    {
+      "epoch": 0.6945438148203817,
+      "grad_norm": 0.2811762999622638,
+      "learning_rate": 4.5102145774391455e-05,
+      "loss": 0.9743,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6947362092289193,
+      "grad_norm": 0.31103172756328284,
+      "learning_rate": 4.505006642412433e-05,
+      "loss": 0.9706,
+      "step": 3611
+    },
+    {
+      "epoch": 0.6949286036374568,
+      "grad_norm": 0.3319691141468076,
+      "learning_rate": 4.4998008415778205e-05,
+      "loss": 0.9846,
+      "step": 3612
+    },
+    {
+      "epoch": 0.6951209980459943,
+      "grad_norm": 0.3542622944509499,
+      "learning_rate": 4.494597176957186e-05,
+      "loss": 1.0153,
+      "step": 3613
+    },
+    {
+      "epoch": 0.6953133924545318,
+      "grad_norm": 0.27408377469551215,
+      "learning_rate": 4.4893956505715614e-05,
+      "loss": 1.0739,
+      "step": 3614
+    },
+    {
+      "epoch": 0.6955057868630693,
+      "grad_norm": 0.2900926564026216,
+      "learning_rate": 4.484196264441176e-05,
+      "loss": 1.0093,
+      "step": 3615
+    },
+    {
+      "epoch": 0.6956981812716068,
+      "grad_norm": 0.27153232008243816,
+      "learning_rate": 4.478999020585399e-05,
+      "loss": 0.9417,
+      "step": 3616
+    },
+    {
+      "epoch": 0.6958905756801443,
+      "grad_norm": 0.3506619741613685,
+      "learning_rate": 4.4738039210227836e-05,
+      "loss": 1.0928,
+      "step": 3617
+    },
+    {
+      "epoch": 0.6960829700886818,
+      "grad_norm": 0.318977313180713,
+      "learning_rate": 4.468610967771051e-05,
+      "loss": 1.086,
+      "step": 3618
+    },
+    {
+      "epoch": 0.6962753644972193,
+      "grad_norm": 0.3011208266798856,
+      "learning_rate": 4.463420162847076e-05,
+      "loss": 0.9993,
+      "step": 3619
+    },
+    {
+      "epoch": 0.6964677589057569,
+      "grad_norm": 0.3382065454700353,
+      "learning_rate": 4.4582315082669124e-05,
+      "loss": 1.0384,
+      "step": 3620
+    },
+    {
+      "epoch": 0.6966601533142943,
+      "grad_norm": 0.31788701459963614,
+      "learning_rate": 4.453045006045773e-05,
+      "loss": 1.0346,
+      "step": 3621
+    },
+    {
+      "epoch": 0.6968525477228318,
+      "grad_norm": 0.3610214002020846,
+      "learning_rate": 4.4478606581980354e-05,
+      "loss": 1.0278,
+      "step": 3622
+    },
+    {
+      "epoch": 0.6970449421313694,
+      "grad_norm": 0.31798060911254394,
+      "learning_rate": 4.4426784667372444e-05,
+      "loss": 1.0491,
+      "step": 3623
+    },
+    {
+      "epoch": 0.6972373365399068,
+      "grad_norm": 0.3165145664033429,
+      "learning_rate": 4.437498433676097e-05,
+      "loss": 1.1319,
+      "step": 3624
+    },
+    {
+      "epoch": 0.6974297309484443,
+      "grad_norm": 0.3490203065216595,
+      "learning_rate": 4.432320561026461e-05,
+      "loss": 1.0663,
+      "step": 3625
+    },
+    {
+      "epoch": 0.6976221253569818,
+      "grad_norm": 0.321369796440476,
+      "learning_rate": 4.4271448507993626e-05,
+      "loss": 1.0723,
+      "step": 3626
+    },
+    {
+      "epoch": 0.6978145197655193,
+      "grad_norm": 0.3229977161547563,
+      "learning_rate": 4.4219713050049884e-05,
+      "loss": 0.9756,
+      "step": 3627
+    },
+    {
+      "epoch": 0.6980069141740568,
+      "grad_norm": 0.3661307168782783,
+      "learning_rate": 4.416799925652684e-05,
+      "loss": 0.9889,
+      "step": 3628
+    },
+    {
+      "epoch": 0.6981993085825943,
+      "grad_norm": 0.28575720349807143,
+      "learning_rate": 4.4116307147509564e-05,
+      "loss": 1.074,
+      "step": 3629
+    },
+    {
+      "epoch": 0.6983917029911318,
+      "grad_norm": 0.3255365371927097,
+      "learning_rate": 4.4064636743074605e-05,
+      "loss": 1.1065,
+      "step": 3630
+    },
+    {
+      "epoch": 0.6985840973996693,
+      "grad_norm": 0.28518210359942675,
+      "learning_rate": 4.4012988063290194e-05,
+      "loss": 0.9744,
+      "step": 3631
+    },
+    {
+      "epoch": 0.6987764918082068,
+      "grad_norm": 0.30779999292049,
+      "learning_rate": 4.3961361128216084e-05,
+      "loss": 1.0211,
+      "step": 3632
+    },
+    {
+      "epoch": 0.6989688862167444,
+      "grad_norm": 0.2905723137000605,
+      "learning_rate": 4.390975595790358e-05,
+      "loss": 1.0688,
+      "step": 3633
+    },
+    {
+      "epoch": 0.6991612806252818,
+      "grad_norm": 0.3212770448905719,
+      "learning_rate": 4.385817257239556e-05,
+      "loss": 1.024,
+      "step": 3634
+    },
+    {
+      "epoch": 0.6993536750338193,
+      "grad_norm": 0.3174745426718206,
+      "learning_rate": 4.380661099172636e-05,
+      "loss": 1.034,
+      "step": 3635
+    },
+    {
+      "epoch": 0.6995460694423569,
+      "grad_norm": 0.3260547521158565,
+      "learning_rate": 4.375507123592194e-05,
+      "loss": 1.0072,
+      "step": 3636
+    },
+    {
+      "epoch": 0.6997384638508943,
+      "grad_norm": 0.3348747934083643,
+      "learning_rate": 4.370355332499977e-05,
+      "loss": 1.098,
+      "step": 3637
+    },
+    {
+      "epoch": 0.6999308582594318,
+      "grad_norm": 0.3083344727748014,
+      "learning_rate": 4.365205727896872e-05,
+      "loss": 1.0648,
+      "step": 3638
+    },
+    {
+      "epoch": 0.7001232526679694,
+      "grad_norm": 0.301352824560482,
+      "learning_rate": 4.3600583117829395e-05,
+      "loss": 1.0803,
+      "step": 3639
+    },
+    {
+      "epoch": 0.7003156470765068,
+      "grad_norm": 0.30552325020482884,
+      "learning_rate": 4.354913086157366e-05,
+      "loss": 1.007,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7005080414850443,
+      "grad_norm": 0.3234255184277871,
+      "learning_rate": 4.3497700530185015e-05,
+      "loss": 1.0255,
+      "step": 3641
+    },
+    {
+      "epoch": 0.7007004358935819,
+      "grad_norm": 0.36707484512526145,
+      "learning_rate": 4.344629214363845e-05,
+      "loss": 1.1103,
+      "step": 3642
+    },
+    {
+      "epoch": 0.7008928303021194,
+      "grad_norm": 0.35193929458091555,
+      "learning_rate": 4.33949057219003e-05,
+      "loss": 0.9848,
+      "step": 3643
+    },
+    {
+      "epoch": 0.7010852247106568,
+      "grad_norm": 0.3691778899647307,
+      "learning_rate": 4.3343541284928515e-05,
+      "loss": 1.0949,
+      "step": 3644
+    },
+    {
+      "epoch": 0.7012776191191944,
+      "grad_norm": 0.4275166515800625,
+      "learning_rate": 4.329219885267244e-05,
+      "loss": 1.0213,
+      "step": 3645
+    },
+    {
+      "epoch": 0.7014700135277319,
+      "grad_norm": 0.3176436192150303,
+      "learning_rate": 4.324087844507289e-05,
+      "loss": 1.0829,
+      "step": 3646
+    },
+    {
+      "epoch": 0.7016624079362693,
+      "grad_norm": 0.3690684161969959,
+      "learning_rate": 4.318958008206214e-05,
+      "loss": 1.095,
+      "step": 3647
+    },
+    {
+      "epoch": 0.7018548023448069,
+      "grad_norm": 0.31173441345064723,
+      "learning_rate": 4.3138303783563835e-05,
+      "loss": 1.0174,
+      "step": 3648
+    },
+    {
+      "epoch": 0.7020471967533444,
+      "grad_norm": 0.35401815098570905,
+      "learning_rate": 4.308704956949313e-05,
+      "loss": 0.9801,
+      "step": 3649
+    },
+    {
+      "epoch": 0.7022395911618818,
+      "grad_norm": 0.2810103337747946,
+      "learning_rate": 4.303581745975655e-05,
+      "loss": 1.0704,
+      "step": 3650
+    },
+    {
+      "epoch": 0.7024319855704194,
+      "grad_norm": 0.3321944505964623,
+      "learning_rate": 4.298460747425208e-05,
+      "loss": 1.0488,
+      "step": 3651
+    },
+    {
+      "epoch": 0.7026243799789569,
+      "grad_norm": 0.35588553122779976,
+      "learning_rate": 4.293341963286912e-05,
+      "loss": 1.002,
+      "step": 3652
+    },
+    {
+      "epoch": 0.7028167743874943,
+      "grad_norm": 0.5000776742160565,
+      "learning_rate": 4.288225395548835e-05,
+      "loss": 1.0413,
+      "step": 3653
+    },
+    {
+      "epoch": 0.7030091687960318,
+      "grad_norm": 0.28922850719296755,
+      "learning_rate": 4.283111046198198e-05,
+      "loss": 1.0107,
+      "step": 3654
+    },
+    {
+      "epoch": 0.7032015632045694,
+      "grad_norm": 0.3814514918786738,
+      "learning_rate": 4.277998917221354e-05,
+      "loss": 1.1259,
+      "step": 3655
+    },
+    {
+      "epoch": 0.7033939576131069,
+      "grad_norm": 0.3195508583181781,
+      "learning_rate": 4.272889010603798e-05,
+      "loss": 1.0463,
+      "step": 3656
+    },
+    {
+      "epoch": 0.7035863520216443,
+      "grad_norm": 0.2943539622186965,
+      "learning_rate": 4.267781328330155e-05,
+      "loss": 1.0733,
+      "step": 3657
+    },
+    {
+      "epoch": 0.7037787464301819,
+      "grad_norm": 0.27271477986595266,
+      "learning_rate": 4.262675872384196e-05,
+      "loss": 1.067,
+      "step": 3658
+    },
+    {
+      "epoch": 0.7039711408387194,
+      "grad_norm": 0.3224113081689263,
+      "learning_rate": 4.257572644748813e-05,
+      "loss": 1.0631,
+      "step": 3659
+    },
+    {
+      "epoch": 0.7041635352472568,
+      "grad_norm": 0.34867810832027757,
+      "learning_rate": 4.2524716474060454e-05,
+      "loss": 1.0439,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7043559296557944,
+      "grad_norm": 0.3055074411484674,
+      "learning_rate": 4.2473728823370604e-05,
+      "loss": 1.0913,
+      "step": 3661
+    },
+    {
+      "epoch": 0.7045483240643319,
+      "grad_norm": 0.2693879156783923,
+      "learning_rate": 4.242276351522161e-05,
+      "loss": 1.0455,
+      "step": 3662
+    },
+    {
+      "epoch": 0.7047407184728693,
+      "grad_norm": 0.3377767490572996,
+      "learning_rate": 4.237182056940784e-05,
+      "loss": 1.0492,
+      "step": 3663
+    },
+    {
+      "epoch": 0.7049331128814069,
+      "grad_norm": 0.3453557523605591,
+      "learning_rate": 4.232090000571488e-05,
+      "loss": 1.0132,
+      "step": 3664
+    },
+    {
+      "epoch": 0.7051255072899444,
+      "grad_norm": 0.3931272846701584,
+      "learning_rate": 4.227000184391972e-05,
+      "loss": 1.0836,
+      "step": 3665
+    },
+    {
+      "epoch": 0.7053179016984819,
+      "grad_norm": 0.2821680074923524,
+      "learning_rate": 4.221912610379065e-05,
+      "loss": 1.0373,
+      "step": 3666
+    },
+    {
+      "epoch": 0.7055102961070194,
+      "grad_norm": 0.4156339281689855,
+      "learning_rate": 4.2168272805087114e-05,
+      "loss": 0.9768,
+      "step": 3667
+    },
+    {
+      "epoch": 0.7057026905155569,
+      "grad_norm": 0.331543348965038,
+      "learning_rate": 4.2117441967560115e-05,
+      "loss": 0.9686,
+      "step": 3668
+    },
+    {
+      "epoch": 0.7058950849240944,
+      "grad_norm": 0.358836274801957,
+      "learning_rate": 4.2066633610951636e-05,
+      "loss": 1.0072,
+      "step": 3669
+    },
+    {
+      "epoch": 0.7060874793326319,
+      "grad_norm": 0.2611148059855862,
+      "learning_rate": 4.2015847754995086e-05,
+      "loss": 1.0759,
+      "step": 3670
+    },
+    {
+      "epoch": 0.7062798737411694,
+      "grad_norm": 0.3091379333248745,
+      "learning_rate": 4.196508441941516e-05,
+      "loss": 0.908,
+      "step": 3671
+    },
+    {
+      "epoch": 0.7064722681497069,
+      "grad_norm": 0.35632687792019546,
+      "learning_rate": 4.191434362392768e-05,
+      "loss": 0.9345,
+      "step": 3672
+    },
+    {
+      "epoch": 0.7066646625582444,
+      "grad_norm": 0.3338972341642524,
+      "learning_rate": 4.18636253882398e-05,
+      "loss": 1.011,
+      "step": 3673
+    },
+    {
+      "epoch": 0.7068570569667819,
+      "grad_norm": 0.27346720179809914,
+      "learning_rate": 4.181292973204992e-05,
+      "loss": 1.0944,
+      "step": 3674
+    },
+    {
+      "epoch": 0.7070494513753194,
+      "grad_norm": 0.30661424884024435,
+      "learning_rate": 4.176225667504765e-05,
+      "loss": 1.1006,
+      "step": 3675
+    },
+    {
+      "epoch": 0.707241845783857,
+      "grad_norm": 0.3247441684160884,
+      "learning_rate": 4.171160623691383e-05,
+      "loss": 1.0559,
+      "step": 3676
+    },
+    {
+      "epoch": 0.7074342401923944,
+      "grad_norm": 0.2849133293486218,
+      "learning_rate": 4.1660978437320474e-05,
+      "loss": 1.0741,
+      "step": 3677
+    },
+    {
+      "epoch": 0.7076266346009319,
+      "grad_norm": 0.3425235450710564,
+      "learning_rate": 4.161037329593085e-05,
+      "loss": 1.0535,
+      "step": 3678
+    },
+    {
+      "epoch": 0.7078190290094694,
+      "grad_norm": 0.34727074422066867,
+      "learning_rate": 4.155979083239942e-05,
+      "loss": 1.023,
+      "step": 3679
+    },
+    {
+      "epoch": 0.7080114234180069,
+      "grad_norm": 0.3755806831889047,
+      "learning_rate": 4.150923106637181e-05,
+      "loss": 1.0091,
+      "step": 3680
+    },
+    {
+      "epoch": 0.7082038178265444,
+      "grad_norm": 0.2915675120716691,
+      "learning_rate": 4.145869401748491e-05,
+      "loss": 1.0241,
+      "step": 3681
+    },
+    {
+      "epoch": 0.7083962122350819,
+      "grad_norm": 0.2646728702913228,
+      "learning_rate": 4.1408179705366636e-05,
+      "loss": 1.0533,
+      "step": 3682
+    },
+    {
+      "epoch": 0.7085886066436194,
+      "grad_norm": 0.3001666301177163,
+      "learning_rate": 4.135768814963622e-05,
+      "loss": 1.0303,
+      "step": 3683
+    },
+    {
+      "epoch": 0.7087810010521569,
+      "grad_norm": 0.360544215986514,
+      "learning_rate": 4.130721936990399e-05,
+      "loss": 1.0656,
+      "step": 3684
+    },
+    {
+      "epoch": 0.7089733954606944,
+      "grad_norm": 0.2977311383710479,
+      "learning_rate": 4.125677338577144e-05,
+      "loss": 0.9727,
+      "step": 3685
+    },
+    {
+      "epoch": 0.709165789869232,
+      "grad_norm": 0.30164742924262233,
+      "learning_rate": 4.1206350216831214e-05,
+      "loss": 0.9284,
+      "step": 3686
+    },
+    {
+      "epoch": 0.7093581842777694,
+      "grad_norm": 0.28890038409130697,
+      "learning_rate": 4.115594988266711e-05,
+      "loss": 1.0666,
+      "step": 3687
+    },
+    {
+      "epoch": 0.7095505786863069,
+      "grad_norm": 0.31157720589399535,
+      "learning_rate": 4.110557240285398e-05,
+      "loss": 1.095,
+      "step": 3688
+    },
+    {
+      "epoch": 0.7097429730948445,
+      "grad_norm": 0.3148937274621826,
+      "learning_rate": 4.105521779695789e-05,
+      "loss": 1.031,
+      "step": 3689
+    },
+    {
+      "epoch": 0.7099353675033819,
+      "grad_norm": 0.3438890727145004,
+      "learning_rate": 4.100488608453599e-05,
+      "loss": 1.1107,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7101277619119194,
+      "grad_norm": 0.3141463560938008,
+      "learning_rate": 4.095457728513652e-05,
+      "loss": 1.0204,
+      "step": 3691
+    },
+    {
+      "epoch": 0.710320156320457,
+      "grad_norm": 0.30817618613909464,
+      "learning_rate": 4.09042914182989e-05,
+      "loss": 0.9865,
+      "step": 3692
+    },
+    {
+      "epoch": 0.7105125507289944,
+      "grad_norm": 0.371041857371169,
+      "learning_rate": 4.08540285035535e-05,
+      "loss": 1.0582,
+      "step": 3693
+    },
+    {
+      "epoch": 0.7107049451375319,
+      "grad_norm": 0.3289110495655149,
+      "learning_rate": 4.0803788560421885e-05,
+      "loss": 1.1455,
+      "step": 3694
+    },
+    {
+      "epoch": 0.7108973395460695,
+      "grad_norm": 0.292030176544287,
+      "learning_rate": 4.075357160841671e-05,
+      "loss": 0.9832,
+      "step": 3695
+    },
+    {
+      "epoch": 0.711089733954607,
+      "grad_norm": 0.27344448611216143,
+      "learning_rate": 4.0703377667041554e-05,
+      "loss": 0.9701,
+      "step": 3696
+    },
+    {
+      "epoch": 0.7112821283631444,
+      "grad_norm": 0.30585706691093373,
+      "learning_rate": 4.065320675579132e-05,
+      "loss": 0.9758,
+      "step": 3697
+    },
+    {
+      "epoch": 0.711474522771682,
+      "grad_norm": 0.34736995015019284,
+      "learning_rate": 4.060305889415168e-05,
+      "loss": 1.0565,
+      "step": 3698
+    },
+    {
+      "epoch": 0.7116669171802195,
+      "grad_norm": 0.3438355222148382,
+      "learning_rate": 4.055293410159954e-05,
+      "loss": 1.0976,
+      "step": 3699
+    },
+    {
+      "epoch": 0.7118593115887569,
+      "grad_norm": 0.3368300244498483,
+      "learning_rate": 4.050283239760282e-05,
+      "loss": 1.084,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7120517059972945,
+      "grad_norm": 0.44270531166203037,
+      "learning_rate": 4.045275380162038e-05,
+      "loss": 1.1315,
+      "step": 3701
+    },
+    {
+      "epoch": 0.712244100405832,
+      "grad_norm": 0.3009240606340105,
+      "learning_rate": 4.04026983331022e-05,
+      "loss": 1.0968,
+      "step": 3702
+    },
+    {
+      "epoch": 0.7124364948143694,
+      "grad_norm": 0.30989802760591656,
+      "learning_rate": 4.035266601148924e-05,
+      "loss": 1.0408,
+      "step": 3703
+    },
+    {
+      "epoch": 0.712628889222907,
+      "grad_norm": 0.28894025278079455,
+      "learning_rate": 4.030265685621349e-05,
+      "loss": 1.0601,
+      "step": 3704
+    },
+    {
+      "epoch": 0.7128212836314445,
+      "grad_norm": 0.3529244750906975,
+      "learning_rate": 4.025267088669796e-05,
+      "loss": 0.944,
+      "step": 3705
+    },
+    {
+      "epoch": 0.713013678039982,
+      "grad_norm": 0.2755090473065579,
+      "learning_rate": 4.020270812235656e-05,
+      "loss": 1.0735,
+      "step": 3706
+    },
+    {
+      "epoch": 0.7132060724485194,
+      "grad_norm": 0.2756465759392184,
+      "learning_rate": 4.015276858259427e-05,
+      "loss": 0.9892,
+      "step": 3707
+    },
+    {
+      "epoch": 0.713398466857057,
+      "grad_norm": 0.27570867421132506,
+      "learning_rate": 4.0102852286807044e-05,
+      "loss": 0.9672,
+      "step": 3708
+    },
+    {
+      "epoch": 0.7135908612655945,
+      "grad_norm": 0.2769737548017259,
+      "learning_rate": 4.005295925438181e-05,
+      "loss": 1.0487,
+      "step": 3709
+    },
+    {
+      "epoch": 0.7137832556741319,
+      "grad_norm": 0.3824239219285524,
+      "learning_rate": 4.000308950469646e-05,
+      "loss": 1.0881,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7139756500826695,
+      "grad_norm": 0.300542991742788,
+      "learning_rate": 3.995324305711976e-05,
+      "loss": 1.0832,
+      "step": 3711
+    },
+    {
+      "epoch": 0.714168044491207,
+      "grad_norm": 0.38485709232922743,
+      "learning_rate": 3.990341993101154e-05,
+      "loss": 0.9982,
+      "step": 3712
+    },
+    {
+      "epoch": 0.7143604388997444,
+      "grad_norm": 0.36948846441757627,
+      "learning_rate": 3.985362014572256e-05,
+      "loss": 1.0375,
+      "step": 3713
+    },
+    {
+      "epoch": 0.714552833308282,
+      "grad_norm": 0.33271159089885494,
+      "learning_rate": 3.980384372059438e-05,
+      "loss": 1.087,
+      "step": 3714
+    },
+    {
+      "epoch": 0.7147452277168195,
+      "grad_norm": 0.3070608102249598,
+      "learning_rate": 3.97540906749597e-05,
+      "loss": 1.0418,
+      "step": 3715
+    },
+    {
+      "epoch": 0.7149376221253569,
+      "grad_norm": 0.278222487182402,
+      "learning_rate": 3.9704361028142025e-05,
+      "loss": 1.0597,
+      "step": 3716
+    },
+    {
+      "epoch": 0.7151300165338945,
+      "grad_norm": 0.33907652157565016,
+      "learning_rate": 3.9654654799455695e-05,
+      "loss": 1.0245,
+      "step": 3717
+    },
+    {
+      "epoch": 0.715322410942432,
+      "grad_norm": 0.3346865490073867,
+      "learning_rate": 3.960497200820609e-05,
+      "loss": 0.9931,
+      "step": 3718
+    },
+    {
+      "epoch": 0.7155148053509695,
+      "grad_norm": 0.27766037000656824,
+      "learning_rate": 3.955531267368941e-05,
+      "loss": 1.0313,
+      "step": 3719
+    },
+    {
+      "epoch": 0.715707199759507,
+      "grad_norm": 0.32186423231949907,
+      "learning_rate": 3.9505676815192794e-05,
+      "loss": 1.1072,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7158995941680445,
+      "grad_norm": 0.29875133578593793,
+      "learning_rate": 3.945606445199427e-05,
+      "loss": 1.0587,
+      "step": 3721
+    },
+    {
+      "epoch": 0.716091988576582,
+      "grad_norm": 0.44261304846040084,
+      "learning_rate": 3.9406475603362617e-05,
+      "loss": 1.0142,
+      "step": 3722
+    },
+    {
+      "epoch": 0.7162843829851195,
+      "grad_norm": 0.3336360338894305,
+      "learning_rate": 3.9356910288557626e-05,
+      "loss": 1.0217,
+      "step": 3723
+    },
+    {
+      "epoch": 0.716476777393657,
+      "grad_norm": 0.36172411772231067,
+      "learning_rate": 3.9307368526829924e-05,
+      "loss": 1.0965,
+      "step": 3724
+    },
+    {
+      "epoch": 0.7166691718021945,
+      "grad_norm": 0.32432511609127174,
+      "learning_rate": 3.925785033742085e-05,
+      "loss": 0.9878,
+      "step": 3725
+    },
+    {
+      "epoch": 0.716861566210732,
+      "grad_norm": 0.27051167653474395,
+      "learning_rate": 3.920835573956285e-05,
+      "loss": 1.0523,
+      "step": 3726
+    },
+    {
+      "epoch": 0.7170539606192695,
+      "grad_norm": 0.26130992964006505,
+      "learning_rate": 3.915888475247894e-05,
+      "loss": 1.0498,
+      "step": 3727
+    },
+    {
+      "epoch": 0.717246355027807,
+      "grad_norm": 0.34151546602238086,
+      "learning_rate": 3.910943739538313e-05,
+      "loss": 1.0417,
+      "step": 3728
+    },
+    {
+      "epoch": 0.7174387494363446,
+      "grad_norm": 0.2965158721769046,
+      "learning_rate": 3.906001368748022e-05,
+      "loss": 0.9544,
+      "step": 3729
+    },
+    {
+      "epoch": 0.717631143844882,
+      "grad_norm": 0.37161904876305424,
+      "learning_rate": 3.901061364796574e-05,
+      "loss": 0.9319,
+      "step": 3730
+    },
+    {
+      "epoch": 0.7178235382534195,
+      "grad_norm": 0.3515770974258548,
+      "learning_rate": 3.896123729602615e-05,
+      "loss": 0.9907,
+      "step": 3731
+    },
+    {
+      "epoch": 0.718015932661957,
+      "grad_norm": 0.30045487767797413,
+      "learning_rate": 3.891188465083865e-05,
+      "loss": 1.0438,
+      "step": 3732
+    },
+    {
+      "epoch": 0.7182083270704945,
+      "grad_norm": 0.28262652086020346,
+      "learning_rate": 3.8862555731571216e-05,
+      "loss": 1.0546,
+      "step": 3733
+    },
+    {
+      "epoch": 0.718400721479032,
+      "grad_norm": 0.3462546537820959,
+      "learning_rate": 3.88132505573827e-05,
+      "loss": 1.0559,
+      "step": 3734
+    },
+    {
+      "epoch": 0.7185931158875695,
+      "grad_norm": 0.27948409223515214,
+      "learning_rate": 3.876396914742257e-05,
+      "loss": 0.9437,
+      "step": 3735
+    },
+    {
+      "epoch": 0.718785510296107,
+      "grad_norm": 0.30889913743720837,
+      "learning_rate": 3.87147115208312e-05,
+      "loss": 1.0806,
+      "step": 3736
+    },
+    {
+      "epoch": 0.7189779047046445,
+      "grad_norm": 0.3472597324022128,
+      "learning_rate": 3.8665477696739685e-05,
+      "loss": 1.0979,
+      "step": 3737
+    },
+    {
+      "epoch": 0.719170299113182,
+      "grad_norm": 0.32232113001805635,
+      "learning_rate": 3.861626769426988e-05,
+      "loss": 1.0403,
+      "step": 3738
+    },
+    {
+      "epoch": 0.7193626935217196,
+      "grad_norm": 0.3208620241975116,
+      "learning_rate": 3.856708153253438e-05,
+      "loss": 1.0075,
+      "step": 3739
+    },
+    {
+      "epoch": 0.719555087930257,
+      "grad_norm": 0.3617577173368042,
+      "learning_rate": 3.8517919230636546e-05,
+      "loss": 1.0665,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7197474823387945,
+      "grad_norm": 0.3985323588661623,
+      "learning_rate": 3.846878080767039e-05,
+      "loss": 0.9795,
+      "step": 3741
+    },
+    {
+      "epoch": 0.7199398767473321,
+      "grad_norm": 0.32962981865633395,
+      "learning_rate": 3.841966628272079e-05,
+      "loss": 0.966,
+      "step": 3742
+    },
+    {
+      "epoch": 0.7201322711558695,
+      "grad_norm": 0.3720995664827796,
+      "learning_rate": 3.837057567486314e-05,
+      "loss": 1.0475,
+      "step": 3743
+    },
+    {
+      "epoch": 0.720324665564407,
+      "grad_norm": 0.3428358332024321,
+      "learning_rate": 3.832150900316377e-05,
+      "loss": 1.0912,
+      "step": 3744
+    },
+    {
+      "epoch": 0.7205170599729446,
+      "grad_norm": 0.2905157740356335,
+      "learning_rate": 3.827246628667962e-05,
+      "loss": 1.0004,
+      "step": 3745
+    },
+    {
+      "epoch": 0.720709454381482,
+      "grad_norm": 0.300929230146176,
+      "learning_rate": 3.8223447544458256e-05,
+      "loss": 1.0903,
+      "step": 3746
+    },
+    {
+      "epoch": 0.7209018487900195,
+      "grad_norm": 0.301120293317064,
+      "learning_rate": 3.817445279553801e-05,
+      "loss": 1.032,
+      "step": 3747
+    },
+    {
+      "epoch": 0.7210942431985571,
+      "grad_norm": 0.33428298687700736,
+      "learning_rate": 3.8125482058947906e-05,
+      "loss": 0.9867,
+      "step": 3748
+    },
+    {
+      "epoch": 0.7212866376070945,
+      "grad_norm": 0.4316125347231823,
+      "learning_rate": 3.807653535370752e-05,
+      "loss": 1.0085,
+      "step": 3749
+    },
+    {
+      "epoch": 0.721479032015632,
+      "grad_norm": 0.2887936628347958,
+      "learning_rate": 3.802761269882734e-05,
+      "loss": 0.9862,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7216714264241696,
+      "grad_norm": 0.2708283673844958,
+      "learning_rate": 3.797871411330824e-05,
+      "loss": 0.9947,
+      "step": 3751
+    },
+    {
+      "epoch": 0.7218638208327071,
+      "grad_norm": 0.31822615890375244,
+      "learning_rate": 3.7929839616141914e-05,
+      "loss": 1.0154,
+      "step": 3752
+    },
+    {
+      "epoch": 0.7220562152412445,
+      "grad_norm": 0.38637026262251845,
+      "learning_rate": 3.788098922631067e-05,
+      "loss": 1.1266,
+      "step": 3753
+    },
+    {
+      "epoch": 0.7222486096497821,
+      "grad_norm": 0.338785243238196,
+      "learning_rate": 3.7832162962787355e-05,
+      "loss": 0.9855,
+      "step": 3754
+    },
+    {
+      "epoch": 0.7224410040583196,
+      "grad_norm": 0.31961264096618797,
+      "learning_rate": 3.7783360844535656e-05,
+      "loss": 1.0101,
+      "step": 3755
+    },
+    {
+      "epoch": 0.722633398466857,
+      "grad_norm": 0.31727969585760973,
+      "learning_rate": 3.773458289050963e-05,
+      "loss": 1.0261,
+      "step": 3756
+    },
+    {
+      "epoch": 0.7228257928753946,
+      "grad_norm": 0.35221786714551934,
+      "learning_rate": 3.7685829119654134e-05,
+      "loss": 1.1314,
+      "step": 3757
+    },
+    {
+      "epoch": 0.7230181872839321,
+      "grad_norm": 0.3157238242748574,
+      "learning_rate": 3.763709955090461e-05,
+      "loss": 1.1029,
+      "step": 3758
+    },
+    {
+      "epoch": 0.7232105816924695,
+      "grad_norm": 0.2831064025789112,
+      "learning_rate": 3.758839420318696e-05,
+      "loss": 1.043,
+      "step": 3759
+    },
+    {
+      "epoch": 0.723402976101007,
+      "grad_norm": 0.43844636356805766,
+      "learning_rate": 3.753971309541784e-05,
+      "loss": 1.0919,
+      "step": 3760
+    },
+    {
+      "epoch": 0.7235953705095446,
+      "grad_norm": 0.3869646349147955,
+      "learning_rate": 3.7491056246504433e-05,
+      "loss": 1.1238,
+      "step": 3761
+    },
+    {
+      "epoch": 0.723787764918082,
+      "grad_norm": 0.321153005189178,
+      "learning_rate": 3.7442423675344474e-05,
+      "loss": 1.1164,
+      "step": 3762
+    },
+    {
+      "epoch": 0.7239801593266195,
+      "grad_norm": 0.43787386905458286,
+      "learning_rate": 3.739381540082635e-05,
+      "loss": 1.0211,
+      "step": 3763
+    },
+    {
+      "epoch": 0.7241725537351571,
+      "grad_norm": 0.3400332352368309,
+      "learning_rate": 3.734523144182887e-05,
+      "loss": 1.0387,
+      "step": 3764
+    },
+    {
+      "epoch": 0.7243649481436946,
+      "grad_norm": 0.3013033084876963,
+      "learning_rate": 3.729667181722154e-05,
+      "loss": 1.0887,
+      "step": 3765
+    },
+    {
+      "epoch": 0.724557342552232,
+      "grad_norm": 0.2726036749211284,
+      "learning_rate": 3.7248136545864344e-05,
+      "loss": 1.123,
+      "step": 3766
+    },
+    {
+      "epoch": 0.7247497369607696,
+      "grad_norm": 0.39547155414047175,
+      "learning_rate": 3.7199625646607825e-05,
+      "loss": 1.072,
+      "step": 3767
+    },
+    {
+      "epoch": 0.7249421313693071,
+      "grad_norm": 0.2665373450133298,
+      "learning_rate": 3.7151139138293054e-05,
+      "loss": 1.0858,
+      "step": 3768
+    },
+    {
+      "epoch": 0.7251345257778445,
+      "grad_norm": 0.31535191192121326,
+      "learning_rate": 3.7102677039751666e-05,
+      "loss": 1.0084,
+      "step": 3769
+    },
+    {
+      "epoch": 0.7253269201863821,
+      "grad_norm": 0.42017617585244926,
+      "learning_rate": 3.705423936980572e-05,
+      "loss": 1.0712,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7255193145949196,
+      "grad_norm": 0.27582195092043776,
+      "learning_rate": 3.700582614726791e-05,
+      "loss": 1.0112,
+      "step": 3771
+    },
+    {
+      "epoch": 0.725711709003457,
+      "grad_norm": 0.28966556909723074,
+      "learning_rate": 3.6957437390941275e-05,
+      "loss": 1.0827,
+      "step": 3772
+    },
+    {
+      "epoch": 0.7259041034119946,
+      "grad_norm": 0.279856243846349,
+      "learning_rate": 3.690907311961955e-05,
+      "loss": 1.045,
+      "step": 3773
+    },
+    {
+      "epoch": 0.7260964978205321,
+      "grad_norm": 0.29136893719638884,
+      "learning_rate": 3.6860733352086864e-05,
+      "loss": 1.1533,
+      "step": 3774
+    },
+    {
+      "epoch": 0.7262888922290696,
+      "grad_norm": 0.33973439977357034,
+      "learning_rate": 3.681241810711776e-05,
+      "loss": 1.0134,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7264812866376071,
+      "grad_norm": 0.31523364878187615,
+      "learning_rate": 3.6764127403477344e-05,
+      "loss": 0.9708,
+      "step": 3776
+    },
+    {
+      "epoch": 0.7266736810461446,
+      "grad_norm": 0.337605590526865,
+      "learning_rate": 3.671586125992123e-05,
+      "loss": 1.0967,
+      "step": 3777
+    },
+    {
+      "epoch": 0.7268660754546821,
+      "grad_norm": 0.32663225825593967,
+      "learning_rate": 3.6667619695195285e-05,
+      "loss": 1.0049,
+      "step": 3778
+    },
+    {
+      "epoch": 0.7270584698632196,
+      "grad_norm": 0.3487857365650158,
+      "learning_rate": 3.661940272803616e-05,
+      "loss": 1.0463,
+      "step": 3779
+    },
+    {
+      "epoch": 0.7272508642717571,
+      "grad_norm": 0.41716348293929606,
+      "learning_rate": 3.657121037717064e-05,
+      "loss": 1.0468,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7274432586802946,
+      "grad_norm": 0.3096807012547874,
+      "learning_rate": 3.652304266131612e-05,
+      "loss": 1.1228,
+      "step": 3781
+    },
+    {
+      "epoch": 0.7276356530888322,
+      "grad_norm": 0.28589831071669314,
+      "learning_rate": 3.6474899599180423e-05,
+      "loss": 1.0348,
+      "step": 3782
+    },
+    {
+      "epoch": 0.7278280474973696,
+      "grad_norm": 0.44004740383958785,
+      "learning_rate": 3.642678120946168e-05,
+      "loss": 0.9969,
+      "step": 3783
+    },
+    {
+      "epoch": 0.7280204419059071,
+      "grad_norm": 0.2890710011956558,
+      "learning_rate": 3.6378687510848575e-05,
+      "loss": 1.0167,
+      "step": 3784
+    },
+    {
+      "epoch": 0.7282128363144446,
+      "grad_norm": 0.3483685844095603,
+      "learning_rate": 3.633061852202012e-05,
+      "loss": 1.0003,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7284052307229821,
+      "grad_norm": 0.39234394570675146,
+      "learning_rate": 3.628257426164577e-05,
+      "loss": 1.0001,
+      "step": 3786
+    },
+    {
+      "epoch": 0.7285976251315196,
+      "grad_norm": 0.2650185580329702,
+      "learning_rate": 3.62345547483854e-05,
+      "loss": 1.0101,
+      "step": 3787
+    },
+    {
+      "epoch": 0.7287900195400571,
+      "grad_norm": 0.3376377883217269,
+      "learning_rate": 3.6186560000889155e-05,
+      "loss": 1.0657,
+      "step": 3788
+    },
+    {
+      "epoch": 0.7289824139485946,
+      "grad_norm": 0.41767604358984806,
+      "learning_rate": 3.613859003779769e-05,
+      "loss": 1.0208,
+      "step": 3789
+    },
+    {
+      "epoch": 0.7291748083571321,
+      "grad_norm": 0.29866232461856085,
+      "learning_rate": 3.609064487774199e-05,
+      "loss": 1.0896,
+      "step": 3790
+    },
+    {
+      "epoch": 0.7293672027656696,
+      "grad_norm": 0.3550225661979497,
+      "learning_rate": 3.6042724539343376e-05,
+      "loss": 1.1159,
+      "step": 3791
+    },
+    {
+      "epoch": 0.7295595971742072,
+      "grad_norm": 0.31662749067494306,
+      "learning_rate": 3.599482904121361e-05,
+      "loss": 1.0126,
+      "step": 3792
+    },
+    {
+      "epoch": 0.7297519915827446,
+      "grad_norm": 0.3046494714183812,
+      "learning_rate": 3.594695840195468e-05,
+      "loss": 1.0667,
+      "step": 3793
+    },
+    {
+      "epoch": 0.7299443859912821,
+      "grad_norm": 0.28274908213991673,
+      "learning_rate": 3.589911264015902e-05,
+      "loss": 1.0634,
+      "step": 3794
+    },
+    {
+      "epoch": 0.7301367803998197,
+      "grad_norm": 0.34078390297765465,
+      "learning_rate": 3.585129177440938e-05,
+      "loss": 1.0763,
+      "step": 3795
+    },
+    {
+      "epoch": 0.7303291748083571,
+      "grad_norm": 0.5116109315261538,
+      "learning_rate": 3.580349582327882e-05,
+      "loss": 1.0075,
+      "step": 3796
+    },
+    {
+      "epoch": 0.7305215692168946,
+      "grad_norm": 0.3205760990103904,
+      "learning_rate": 3.575572480533076e-05,
+      "loss": 1.033,
+      "step": 3797
+    },
+    {
+      "epoch": 0.7307139636254322,
+      "grad_norm": 0.33090927368722733,
+      "learning_rate": 3.570797873911892e-05,
+      "loss": 0.9595,
+      "step": 3798
+    },
+    {
+      "epoch": 0.7309063580339696,
+      "grad_norm": 0.31168680661249126,
+      "learning_rate": 3.5660257643187276e-05,
+      "loss": 0.9647,
+      "step": 3799
+    },
+    {
+      "epoch": 0.7310987524425071,
+      "grad_norm": 0.33753469324900864,
+      "learning_rate": 3.561256153607021e-05,
+      "loss": 1.0939,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7312911468510447,
+      "grad_norm": 0.3055531827749562,
+      "learning_rate": 3.556489043629224e-05,
+      "loss": 1.0224,
+      "step": 3801
+    },
+    {
+      "epoch": 0.7314835412595821,
+      "grad_norm": 0.4349685632480251,
+      "learning_rate": 3.5517244362368365e-05,
+      "loss": 0.987,
+      "step": 3802
+    },
+    {
+      "epoch": 0.7316759356681196,
+      "grad_norm": 0.29842656083531044,
+      "learning_rate": 3.546962333280379e-05,
+      "loss": 0.9958,
+      "step": 3803
+    },
+    {
+      "epoch": 0.7318683300766572,
+      "grad_norm": 0.3617704818980536,
+      "learning_rate": 3.542202736609389e-05,
+      "loss": 1.04,
+      "step": 3804
+    },
+    {
+      "epoch": 0.7320607244851947,
+      "grad_norm": 0.28628580888877486,
+      "learning_rate": 3.5374456480724426e-05,
+      "loss": 0.9949,
+      "step": 3805
+    },
+    {
+      "epoch": 0.7322531188937321,
+      "grad_norm": 0.31285519070919615,
+      "learning_rate": 3.532691069517142e-05,
+      "loss": 0.9728,
+      "step": 3806
+    },
+    {
+      "epoch": 0.7324455133022697,
+      "grad_norm": 0.40701208946703527,
+      "learning_rate": 3.527939002790101e-05,
+      "loss": 1.0585,
+      "step": 3807
+    },
+    {
+      "epoch": 0.7326379077108072,
+      "grad_norm": 0.31365602556048144,
+      "learning_rate": 3.52318944973698e-05,
+      "loss": 0.8284,
+      "step": 3808
+    },
+    {
+      "epoch": 0.7328303021193446,
+      "grad_norm": 0.3465266680547385,
+      "learning_rate": 3.518442412202441e-05,
+      "loss": 0.9903,
+      "step": 3809
+    },
+    {
+      "epoch": 0.7330226965278822,
+      "grad_norm": 0.30421390179652025,
+      "learning_rate": 3.5136978920301825e-05,
+      "loss": 0.9775,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7332150909364197,
+      "grad_norm": 0.37369570297910537,
+      "learning_rate": 3.508955891062924e-05,
+      "loss": 1.1015,
+      "step": 3811
+    },
+    {
+      "epoch": 0.7334074853449571,
+      "grad_norm": 0.3694988509279056,
+      "learning_rate": 3.504216411142398e-05,
+      "loss": 0.9695,
+      "step": 3812
+    },
+    {
+      "epoch": 0.7335998797534946,
+      "grad_norm": 0.2809510869138268,
+      "learning_rate": 3.499479454109367e-05,
+      "loss": 1.0136,
+      "step": 3813
+    },
+    {
+      "epoch": 0.7337922741620322,
+      "grad_norm": 0.32389434632168307,
+      "learning_rate": 3.4947450218036106e-05,
+      "loss": 1.1118,
+      "step": 3814
+    },
+    {
+      "epoch": 0.7339846685705697,
+      "grad_norm": 0.2935391619293556,
+      "learning_rate": 3.490013116063928e-05,
+      "loss": 1.0271,
+      "step": 3815
+    },
+    {
+      "epoch": 0.7341770629791071,
+      "grad_norm": 0.3841594257340414,
+      "learning_rate": 3.485283738728139e-05,
+      "loss": 1.0188,
+      "step": 3816
+    },
+    {
+      "epoch": 0.7343694573876447,
+      "grad_norm": 0.3392109881109757,
+      "learning_rate": 3.480556891633074e-05,
+      "loss": 1.0101,
+      "step": 3817
+    },
+    {
+      "epoch": 0.7345618517961822,
+      "grad_norm": 0.33709491767906086,
+      "learning_rate": 3.475832576614589e-05,
+      "loss": 1.0193,
+      "step": 3818
+    },
+    {
+      "epoch": 0.7347542462047196,
+      "grad_norm": 0.42010514938698673,
+      "learning_rate": 3.4711107955075536e-05,
+      "loss": 1.0292,
+      "step": 3819
+    },
+    {
+      "epoch": 0.7349466406132572,
+      "grad_norm": 0.3482491943345016,
+      "learning_rate": 3.466391550145852e-05,
+      "loss": 0.9591,
+      "step": 3820
+    },
+    {
+      "epoch": 0.7351390350217947,
+      "grad_norm": 0.2903810344626169,
+      "learning_rate": 3.461674842362389e-05,
+      "loss": 0.9785,
+      "step": 3821
+    },
+    {
+      "epoch": 0.7353314294303321,
+      "grad_norm": 0.3051341660022261,
+      "learning_rate": 3.456960673989074e-05,
+      "loss": 1.0277,
+      "step": 3822
+    },
+    {
+      "epoch": 0.7355238238388697,
+      "grad_norm": 0.32840846531318874,
+      "learning_rate": 3.452249046856836e-05,
+      "loss": 1.0349,
+      "step": 3823
+    },
+    {
+      "epoch": 0.7357162182474072,
+      "grad_norm": 0.3086958749095765,
+      "learning_rate": 3.447539962795619e-05,
+      "loss": 0.9733,
+      "step": 3824
+    },
+    {
+      "epoch": 0.7359086126559446,
+      "grad_norm": 0.29347635630651686,
+      "learning_rate": 3.442833423634377e-05,
+      "loss": 1.0347,
+      "step": 3825
+    },
+    {
+      "epoch": 0.7361010070644822,
+      "grad_norm": 0.2727939117876327,
+      "learning_rate": 3.4381294312010745e-05,
+      "loss": 0.9761,
+      "step": 3826
+    },
+    {
+      "epoch": 0.7362934014730197,
+      "grad_norm": 0.34392564930358976,
+      "learning_rate": 3.433427987322693e-05,
+      "loss": 1.0714,
+      "step": 3827
+    },
+    {
+      "epoch": 0.7364857958815572,
+      "grad_norm": 0.29948809123823283,
+      "learning_rate": 3.4287290938252106e-05,
+      "loss": 0.9936,
+      "step": 3828
+    },
+    {
+      "epoch": 0.7366781902900947,
+      "grad_norm": 0.3825672834034742,
+      "learning_rate": 3.424032752533627e-05,
+      "loss": 1.1535,
+      "step": 3829
+    },
+    {
+      "epoch": 0.7368705846986322,
+      "grad_norm": 0.270002640652531,
+      "learning_rate": 3.4193389652719476e-05,
+      "loss": 1.0852,
+      "step": 3830
+    },
+    {
+      "epoch": 0.7370629791071697,
+      "grad_norm": 0.26936672858848226,
+      "learning_rate": 3.414647733863185e-05,
+      "loss": 1.0606,
+      "step": 3831
+    },
+    {
+      "epoch": 0.7372553735157072,
+      "grad_norm": 0.3856095111550712,
+      "learning_rate": 3.4099590601293626e-05,
+      "loss": 0.9965,
+      "step": 3832
+    },
+    {
+      "epoch": 0.7374477679242447,
+      "grad_norm": 0.28977195742475814,
+      "learning_rate": 3.4052729458915023e-05,
+      "loss": 0.991,
+      "step": 3833
+    },
+    {
+      "epoch": 0.7376401623327822,
+      "grad_norm": 0.29524355198726887,
+      "learning_rate": 3.4005893929696377e-05,
+      "loss": 1.1655,
+      "step": 3834
+    },
+    {
+      "epoch": 0.7378325567413198,
+      "grad_norm": 0.26160246240202195,
+      "learning_rate": 3.395908403182811e-05,
+      "loss": 1.0802,
+      "step": 3835
+    },
+    {
+      "epoch": 0.7380249511498572,
+      "grad_norm": 0.36378109459526253,
+      "learning_rate": 3.391229978349056e-05,
+      "loss": 1.0913,
+      "step": 3836
+    },
+    {
+      "epoch": 0.7382173455583947,
+      "grad_norm": 0.3885348797468423,
+      "learning_rate": 3.386554120285431e-05,
+      "loss": 0.9901,
+      "step": 3837
+    },
+    {
+      "epoch": 0.7384097399669323,
+      "grad_norm": 0.3113625643214789,
+      "learning_rate": 3.381880830807975e-05,
+      "loss": 1.0248,
+      "step": 3838
+    },
+    {
+      "epoch": 0.7386021343754697,
+      "grad_norm": 0.3314536204084258,
+      "learning_rate": 3.3772101117317435e-05,
+      "loss": 0.9322,
+      "step": 3839
+    },
+    {
+      "epoch": 0.7387945287840072,
+      "grad_norm": 0.29580523255470403,
+      "learning_rate": 3.372541964870795e-05,
+      "loss": 1.0826,
+      "step": 3840
+    },
+    {
+      "epoch": 0.7389869231925447,
+      "grad_norm": 0.33361814715852317,
+      "learning_rate": 3.367876392038174e-05,
+      "loss": 1.0389,
+      "step": 3841
+    },
+    {
+      "epoch": 0.7391793176010822,
+      "grad_norm": 0.3084181829564584,
+      "learning_rate": 3.363213395045941e-05,
+      "loss": 1.0238,
+      "step": 3842
+    },
+    {
+      "epoch": 0.7393717120096197,
+      "grad_norm": 0.3401781008304183,
+      "learning_rate": 3.3585529757051504e-05,
+      "loss": 1.003,
+      "step": 3843
+    },
+    {
+      "epoch": 0.7395641064181572,
+      "grad_norm": 0.3145841059695358,
+      "learning_rate": 3.353895135825854e-05,
+      "loss": 1.0826,
+      "step": 3844
+    },
+    {
+      "epoch": 0.7397565008266948,
+      "grad_norm": 0.33358532570006877,
+      "learning_rate": 3.349239877217108e-05,
+      "loss": 1.1339,
+      "step": 3845
+    },
+    {
+      "epoch": 0.7399488952352322,
+      "grad_norm": 0.282023660392982,
+      "learning_rate": 3.3445872016869516e-05,
+      "loss": 1.0499,
+      "step": 3846
+    },
+    {
+      "epoch": 0.7401412896437697,
+      "grad_norm": 0.2601702279941731,
+      "learning_rate": 3.339937111042437e-05,
+      "loss": 1.0775,
+      "step": 3847
+    },
+    {
+      "epoch": 0.7403336840523073,
+      "grad_norm": 0.34465223879362933,
+      "learning_rate": 3.3352896070896054e-05,
+      "loss": 0.9581,
+      "step": 3848
+    },
+    {
+      "epoch": 0.7405260784608447,
+      "grad_norm": 0.31119601244903744,
+      "learning_rate": 3.330644691633492e-05,
+      "loss": 1.0303,
+      "step": 3849
+    },
+    {
+      "epoch": 0.7407184728693822,
+      "grad_norm": 0.29162235865954,
+      "learning_rate": 3.3260023664781326e-05,
+      "loss": 0.9509,
+      "step": 3850
+    },
+    {
+      "epoch": 0.7409108672779198,
+      "grad_norm": 0.27088071763972155,
+      "learning_rate": 3.321362633426547e-05,
+      "loss": 1.0597,
+      "step": 3851
+    },
+    {
+      "epoch": 0.7411032616864572,
+      "grad_norm": 0.2968302711086647,
+      "learning_rate": 3.316725494280757e-05,
+      "loss": 1.0536,
+      "step": 3852
+    },
+    {
+      "epoch": 0.7412956560949947,
+      "grad_norm": 0.2881510912997479,
+      "learning_rate": 3.312090950841775e-05,
+      "loss": 1.0461,
+      "step": 3853
+    },
+    {
+      "epoch": 0.7414880505035323,
+      "grad_norm": 0.41564915405098773,
+      "learning_rate": 3.307459004909599e-05,
+      "loss": 0.9426,
+      "step": 3854
+    },
+    {
+      "epoch": 0.7416804449120697,
+      "grad_norm": 0.34597871012365344,
+      "learning_rate": 3.3028296582832284e-05,
+      "loss": 1.0022,
+      "step": 3855
+    },
+    {
+      "epoch": 0.7418728393206072,
+      "grad_norm": 0.385068863284406,
+      "learning_rate": 3.298202912760652e-05,
+      "loss": 1.0388,
+      "step": 3856
+    },
+    {
+      "epoch": 0.7420652337291448,
+      "grad_norm": 0.34505895322585384,
+      "learning_rate": 3.2935787701388344e-05,
+      "loss": 0.9887,
+      "step": 3857
+    },
+    {
+      "epoch": 0.7422576281376823,
+      "grad_norm": 0.4414506370320824,
+      "learning_rate": 3.288957232213745e-05,
+      "loss": 0.9956,
+      "step": 3858
+    },
+    {
+      "epoch": 0.7424500225462197,
+      "grad_norm": 0.3406219814756699,
+      "learning_rate": 3.284338300780336e-05,
+      "loss": 1.0236,
+      "step": 3859
+    },
+    {
+      "epoch": 0.7426424169547573,
+      "grad_norm": 0.31442420535558924,
+      "learning_rate": 3.279721977632546e-05,
+      "loss": 1.1384,
+      "step": 3860
+    },
+    {
+      "epoch": 0.7428348113632948,
+      "grad_norm": 0.31375976906486563,
+      "learning_rate": 3.275108264563306e-05,
+      "loss": 1.03,
+      "step": 3861
+    },
+    {
+      "epoch": 0.7430272057718322,
+      "grad_norm": 0.31300874958598546,
+      "learning_rate": 3.270497163364521e-05,
+      "loss": 0.9889,
+      "step": 3862
+    },
+    {
+      "epoch": 0.7432196001803698,
+      "grad_norm": 0.2705761982144218,
+      "learning_rate": 3.2658886758270943e-05,
+      "loss": 0.9692,
+      "step": 3863
+    },
+    {
+      "epoch": 0.7434119945889073,
+      "grad_norm": 0.30297177469103076,
+      "learning_rate": 3.261282803740911e-05,
+      "loss": 0.9891,
+      "step": 3864
+    },
+    {
+      "epoch": 0.7436043889974447,
+      "grad_norm": 0.38406072684084736,
+      "learning_rate": 3.256679548894831e-05,
+      "loss": 1.0813,
+      "step": 3865
+    },
+    {
+      "epoch": 0.7437967834059822,
+      "grad_norm": 0.29591747237840993,
+      "learning_rate": 3.252078913076718e-05,
+      "loss": 1.0739,
+      "step": 3866
+    },
+    {
+      "epoch": 0.7439891778145198,
+      "grad_norm": 0.28621020136855685,
+      "learning_rate": 3.247480898073395e-05,
+      "loss": 0.9873,
+      "step": 3867
+    },
+    {
+      "epoch": 0.7441815722230573,
+      "grad_norm": 0.3644907579246381,
+      "learning_rate": 3.242885505670681e-05,
+      "loss": 1.0777,
+      "step": 3868
+    },
+    {
+      "epoch": 0.7443739666315947,
+      "grad_norm": 0.28254687455682775,
+      "learning_rate": 3.238292737653379e-05,
+      "loss": 0.9716,
+      "step": 3869
+    },
+    {
+      "epoch": 0.7445663610401323,
+      "grad_norm": 0.38046555828035444,
+      "learning_rate": 3.233702595805258e-05,
+      "loss": 1.0137,
+      "step": 3870
+    },
+    {
+      "epoch": 0.7447587554486698,
+      "grad_norm": 0.3342343396210659,
+      "learning_rate": 3.229115081909082e-05,
+      "loss": 1.0624,
+      "step": 3871
+    },
+    {
+      "epoch": 0.7449511498572072,
+      "grad_norm": 0.3530750681785038,
+      "learning_rate": 3.224530197746587e-05,
+      "loss": 1.0749,
+      "step": 3872
+    },
+    {
+      "epoch": 0.7451435442657448,
+      "grad_norm": 0.2819775581917406,
+      "learning_rate": 3.219947945098489e-05,
+      "loss": 1.0432,
+      "step": 3873
+    },
+    {
+      "epoch": 0.7453359386742823,
+      "grad_norm": 0.28390548768316626,
+      "learning_rate": 3.215368325744485e-05,
+      "loss": 1.091,
+      "step": 3874
+    },
+    {
+      "epoch": 0.7455283330828197,
+      "grad_norm": 0.35488816437454646,
+      "learning_rate": 3.2107913414632426e-05,
+      "loss": 1.0809,
+      "step": 3875
+    },
+    {
+      "epoch": 0.7457207274913573,
+      "grad_norm": 0.3100437466636983,
+      "learning_rate": 3.206216994032411e-05,
+      "loss": 1.0866,
+      "step": 3876
+    },
+    {
+      "epoch": 0.7459131218998948,
+      "grad_norm": 0.33137820929544354,
+      "learning_rate": 3.2016452852286127e-05,
+      "loss": 1.0101,
+      "step": 3877
+    },
+    {
+      "epoch": 0.7461055163084322,
+      "grad_norm": 0.4217550504657977,
+      "learning_rate": 3.197076216827449e-05,
+      "loss": 1.116,
+      "step": 3878
+    },
+    {
+      "epoch": 0.7462979107169698,
+      "grad_norm": 0.316701280473312,
+      "learning_rate": 3.192509790603496e-05,
+      "loss": 1.1168,
+      "step": 3879
+    },
+    {
+      "epoch": 0.7464903051255073,
+      "grad_norm": 0.2863085336509443,
+      "learning_rate": 3.187946008330295e-05,
+      "loss": 1.1201,
+      "step": 3880
+    },
+    {
+      "epoch": 0.7466826995340448,
+      "grad_norm": 0.3388751009514164,
+      "learning_rate": 3.183384871780367e-05,
+      "loss": 0.9836,
+      "step": 3881
+    },
+    {
+      "epoch": 0.7468750939425823,
+      "grad_norm": 0.27931731976612073,
+      "learning_rate": 3.178826382725212e-05,
+      "loss": 1.0166,
+      "step": 3882
+    },
+    {
+      "epoch": 0.7470674883511198,
+      "grad_norm": 0.40005571496549147,
+      "learning_rate": 3.1742705429352826e-05,
+      "loss": 1.0421,
+      "step": 3883
+    },
+    {
+      "epoch": 0.7472598827596573,
+      "grad_norm": 0.33287206561056276,
+      "learning_rate": 3.169717354180025e-05,
+      "loss": 1.0173,
+      "step": 3884
+    },
+    {
+      "epoch": 0.7474522771681948,
+      "grad_norm": 0.27804798886789184,
+      "learning_rate": 3.165166818227845e-05,
+      "loss": 1.0846,
+      "step": 3885
+    },
+    {
+      "epoch": 0.7476446715767323,
+      "grad_norm": 0.28062620713526976,
+      "learning_rate": 3.160618936846111e-05,
+      "loss": 1.0661,
+      "step": 3886
+    },
+    {
+      "epoch": 0.7478370659852698,
+      "grad_norm": 0.33588095161644405,
+      "learning_rate": 3.156073711801172e-05,
+      "loss": 1.0366,
+      "step": 3887
+    },
+    {
+      "epoch": 0.7480294603938074,
+      "grad_norm": 0.3675736406382842,
+      "learning_rate": 3.151531144858344e-05,
+      "loss": 1.0801,
+      "step": 3888
+    },
+    {
+      "epoch": 0.7482218548023448,
+      "grad_norm": 0.3139361084371443,
+      "learning_rate": 3.1469912377818986e-05,
+      "loss": 1.0132,
+      "step": 3889
+    },
+    {
+      "epoch": 0.7484142492108823,
+      "grad_norm": 0.2738547897392836,
+      "learning_rate": 3.142453992335096e-05,
+      "loss": 1.0006,
+      "step": 3890
+    },
+    {
+      "epoch": 0.7486066436194199,
+      "grad_norm": 0.31015001363801614,
+      "learning_rate": 3.137919410280139e-05,
+      "loss": 0.9684,
+      "step": 3891
+    },
+    {
+      "epoch": 0.7487990380279573,
+      "grad_norm": 0.33435989965073243,
+      "learning_rate": 3.1333874933782115e-05,
+      "loss": 0.9925,
+      "step": 3892
+    },
+    {
+      "epoch": 0.7489914324364948,
+      "grad_norm": 0.27193849841389395,
+      "learning_rate": 3.128858243389461e-05,
+      "loss": 1.0905,
+      "step": 3893
+    },
+    {
+      "epoch": 0.7491838268450323,
+      "grad_norm": 0.3005729293682963,
+      "learning_rate": 3.124331662072987e-05,
+      "loss": 1.0217,
+      "step": 3894
+    },
+    {
+      "epoch": 0.7493762212535698,
+      "grad_norm": 0.28884142465361506,
+      "learning_rate": 3.119807751186872e-05,
+      "loss": 1.0116,
+      "step": 3895
+    },
+    {
+      "epoch": 0.7495686156621073,
+      "grad_norm": 0.3888879568439692,
+      "learning_rate": 3.115286512488144e-05,
+      "loss": 1.0656,
+      "step": 3896
+    },
+    {
+      "epoch": 0.7497610100706448,
+      "grad_norm": 0.3283868279599659,
+      "learning_rate": 3.110767947732801e-05,
+      "loss": 1.0659,
+      "step": 3897
+    },
+    {
+      "epoch": 0.7499534044791824,
+      "grad_norm": 0.29821654741921305,
+      "learning_rate": 3.106252058675806e-05,
+      "loss": 1.0394,
+      "step": 3898
+    },
+    {
+      "epoch": 0.7501457988877198,
+      "grad_norm": 0.39433267274864786,
+      "learning_rate": 3.1017388470710716e-05,
+      "loss": 0.9995,
+      "step": 3899
+    },
+    {
+      "epoch": 0.7503381932962573,
+      "grad_norm": 0.3221394384989637,
+      "learning_rate": 3.097228314671481e-05,
+      "loss": 1.0771,
+      "step": 3900
+    },
+    {
+      "epoch": 0.7505305877047949,
+      "grad_norm": 0.28913411844908465,
+      "learning_rate": 3.092720463228872e-05,
+      "loss": 1.1104,
+      "step": 3901
+    },
+    {
+      "epoch": 0.7507229821133323,
+      "grad_norm": 0.32193391627664836,
+      "learning_rate": 3.0882152944940424e-05,
+      "loss": 1.057,
+      "step": 3902
+    },
+    {
+      "epoch": 0.7509153765218698,
+      "grad_norm": 0.28515267815000667,
+      "learning_rate": 3.0837128102167515e-05,
+      "loss": 1.0559,
+      "step": 3903
+    },
+    {
+      "epoch": 0.7511077709304074,
+      "grad_norm": 0.3467078382476179,
+      "learning_rate": 3.079213012145705e-05,
+      "loss": 1.0364,
+      "step": 3904
+    },
+    {
+      "epoch": 0.7513001653389448,
+      "grad_norm": 0.31431388025175716,
+      "learning_rate": 3.0747159020285765e-05,
+      "loss": 1.0523,
+      "step": 3905
+    },
+    {
+      "epoch": 0.7514925597474823,
+      "grad_norm": 0.3110355117013477,
+      "learning_rate": 3.070221481611992e-05,
+      "loss": 1.0449,
+      "step": 3906
+    },
+    {
+      "epoch": 0.7516849541560199,
+      "grad_norm": 0.33757083018869277,
+      "learning_rate": 3.065729752641532e-05,
+      "loss": 1.0544,
+      "step": 3907
+    },
+    {
+      "epoch": 0.7518773485645573,
+      "grad_norm": 0.30862324850587997,
+      "learning_rate": 3.061240716861735e-05,
+      "loss": 1.0093,
+      "step": 3908
+    },
+    {
+      "epoch": 0.7520697429730948,
+      "grad_norm": 0.3579949480363705,
+      "learning_rate": 3.056754376016087e-05,
+      "loss": 1.1937,
+      "step": 3909
+    },
+    {
+      "epoch": 0.7522621373816324,
+      "grad_norm": 0.2808916414164792,
+      "learning_rate": 3.05227073184703e-05,
+      "loss": 1.0566,
+      "step": 3910
+    },
+    {
+      "epoch": 0.7524545317901699,
+      "grad_norm": 0.31708005491534186,
+      "learning_rate": 3.047789786095967e-05,
+      "loss": 0.9274,
+      "step": 3911
+    },
+    {
+      "epoch": 0.7526469261987073,
+      "grad_norm": 0.2577452537286519,
+      "learning_rate": 3.043311540503233e-05,
+      "loss": 1.0037,
+      "step": 3912
+    },
+    {
+      "epoch": 0.7528393206072449,
+      "grad_norm": 0.36581308032506854,
+      "learning_rate": 3.0388359968081392e-05,
+      "loss": 1.0002,
+      "step": 3913
+    },
+    {
+      "epoch": 0.7530317150157824,
+      "grad_norm": 0.3511404345581437,
+      "learning_rate": 3.0343631567489327e-05,
+      "loss": 1.088,
+      "step": 3914
+    },
+    {
+      "epoch": 0.7532241094243198,
+      "grad_norm": 0.4218062449676212,
+      "learning_rate": 3.0298930220628086e-05,
+      "loss": 0.998,
+      "step": 3915
+    },
+    {
+      "epoch": 0.7534165038328574,
+      "grad_norm": 0.34087832583861805,
+      "learning_rate": 3.0254255944859187e-05,
+      "loss": 1.1089,
+      "step": 3916
+    },
+    {
+      "epoch": 0.7536088982413949,
+      "grad_norm": 0.2992297630211009,
+      "learning_rate": 3.0209608757533625e-05,
+      "loss": 1.0122,
+      "step": 3917
+    },
+    {
+      "epoch": 0.7538012926499323,
+      "grad_norm": 0.3383432519795849,
+      "learning_rate": 3.0164988675991768e-05,
+      "loss": 1.1073,
+      "step": 3918
+    },
+    {
+      "epoch": 0.7539936870584698,
+      "grad_norm": 0.2708614192043716,
+      "learning_rate": 3.0120395717563653e-05,
+      "loss": 1.04,
+      "step": 3919
+    },
+    {
+      "epoch": 0.7541860814670074,
+      "grad_norm": 0.31613015846150794,
+      "learning_rate": 3.0075829899568597e-05,
+      "loss": 1.0496,
+      "step": 3920
+    },
+    {
+      "epoch": 0.7543784758755449,
+      "grad_norm": 0.278483091937346,
+      "learning_rate": 3.0031291239315475e-05,
+      "loss": 1.0537,
+      "step": 3921
+    },
+    {
+      "epoch": 0.7545708702840823,
+      "grad_norm": 0.33118449322408994,
+      "learning_rate": 2.9986779754102613e-05,
+      "loss": 1.0932,
+      "step": 3922
+    },
+    {
+      "epoch": 0.7547632646926199,
+      "grad_norm": 0.27147766967211845,
+      "learning_rate": 2.9942295461217696e-05,
+      "loss": 1.0697,
+      "step": 3923
+    },
+    {
+      "epoch": 0.7549556591011574,
+      "grad_norm": 0.2959924821180381,
+      "learning_rate": 2.9897838377937947e-05,
+      "loss": 1.0275,
+      "step": 3924
+    },
+    {
+      "epoch": 0.7551480535096948,
+      "grad_norm": 0.29312862332293554,
+      "learning_rate": 2.985340852152999e-05,
+      "loss": 0.9969,
+      "step": 3925
+    },
+    {
+      "epoch": 0.7553404479182324,
+      "grad_norm": 0.29010737144046506,
+      "learning_rate": 2.9809005909249864e-05,
+      "loss": 0.9731,
+      "step": 3926
+    },
+    {
+      "epoch": 0.7555328423267699,
+      "grad_norm": 0.2893645304415515,
+      "learning_rate": 2.9764630558343067e-05,
+      "loss": 0.9537,
+      "step": 3927
+    },
+    {
+      "epoch": 0.7557252367353073,
+      "grad_norm": 0.3111618087743285,
+      "learning_rate": 2.9720282486044403e-05,
+      "loss": 1.0596,
+      "step": 3928
+    },
+    {
+      "epoch": 0.7559176311438449,
+      "grad_norm": 0.34666246722633387,
+      "learning_rate": 2.9675961709578194e-05,
+      "loss": 1.0351,
+      "step": 3929
+    },
+    {
+      "epoch": 0.7561100255523824,
+      "grad_norm": 0.40738874844638784,
+      "learning_rate": 2.9631668246158107e-05,
+      "loss": 1.168,
+      "step": 3930
+    },
+    {
+      "epoch": 0.7563024199609198,
+      "grad_norm": 0.44619290769415537,
+      "learning_rate": 2.958740211298722e-05,
+      "loss": 0.9292,
+      "step": 3931
+    },
+    {
+      "epoch": 0.7564948143694574,
+      "grad_norm": 0.3505806871130929,
+      "learning_rate": 2.9543163327258017e-05,
+      "loss": 1.0345,
+      "step": 3932
+    },
+    {
+      "epoch": 0.7566872087779949,
+      "grad_norm": 0.3199085047775709,
+      "learning_rate": 2.9498951906152272e-05,
+      "loss": 1.0892,
+      "step": 3933
+    },
+    {
+      "epoch": 0.7568796031865324,
+      "grad_norm": 0.3700027640340649,
+      "learning_rate": 2.9454767866841227e-05,
+      "loss": 1.0796,
+      "step": 3934
+    },
+    {
+      "epoch": 0.7570719975950699,
+      "grad_norm": 0.3488826836303916,
+      "learning_rate": 2.9410611226485452e-05,
+      "loss": 1.0706,
+      "step": 3935
+    },
+    {
+      "epoch": 0.7572643920036074,
+      "grad_norm": 0.2997754622357488,
+      "learning_rate": 2.9366482002234874e-05,
+      "loss": 1.0116,
+      "step": 3936
+    },
+    {
+      "epoch": 0.7574567864121449,
+      "grad_norm": 0.35128922618999814,
+      "learning_rate": 2.932238021122877e-05,
+      "loss": 0.9734,
+      "step": 3937
+    },
+    {
+      "epoch": 0.7576491808206824,
+      "grad_norm": 0.28937176822519944,
+      "learning_rate": 2.9278305870595812e-05,
+      "loss": 1.0275,
+      "step": 3938
+    },
+    {
+      "epoch": 0.7578415752292199,
+      "grad_norm": 0.3409422796756947,
+      "learning_rate": 2.92342589974539e-05,
+      "loss": 1.0885,
+      "step": 3939
+    },
+    {
+      "epoch": 0.7580339696377574,
+      "grad_norm": 0.284463571268749,
+      "learning_rate": 2.9190239608910387e-05,
+      "loss": 0.9913,
+      "step": 3940
+    },
+    {
+      "epoch": 0.758226364046295,
+      "grad_norm": 0.2989666518432513,
+      "learning_rate": 2.9146247722061802e-05,
+      "loss": 1.0357,
+      "step": 3941
+    },
+    {
+      "epoch": 0.7584187584548324,
+      "grad_norm": 0.3633222837419775,
+      "learning_rate": 2.9102283353994186e-05,
+      "loss": 1.0235,
+      "step": 3942
+    },
+    {
+      "epoch": 0.7586111528633699,
+      "grad_norm": 0.33979945860743493,
+      "learning_rate": 2.9058346521782797e-05,
+      "loss": 0.9876,
+      "step": 3943
+    },
+    {
+      "epoch": 0.7588035472719075,
+      "grad_norm": 0.3062559852702967,
+      "learning_rate": 2.9014437242492132e-05,
+      "loss": 1.0306,
+      "step": 3944
+    },
+    {
+      "epoch": 0.7589959416804449,
+      "grad_norm": 0.37490172386932863,
+      "learning_rate": 2.897055553317607e-05,
+      "loss": 0.9268,
+      "step": 3945
+    },
+    {
+      "epoch": 0.7591883360889824,
+      "grad_norm": 0.3175710898752762,
+      "learning_rate": 2.8926701410877798e-05,
+      "loss": 1.0257,
+      "step": 3946
+    },
+    {
+      "epoch": 0.7593807304975199,
+      "grad_norm": 0.3396488431854097,
+      "learning_rate": 2.8882874892629652e-05,
+      "loss": 1.0612,
+      "step": 3947
+    },
+    {
+      "epoch": 0.7595731249060574,
+      "grad_norm": 0.31035183217118995,
+      "learning_rate": 2.883907599545348e-05,
+      "loss": 1.0694,
+      "step": 3948
+    },
+    {
+      "epoch": 0.7597655193145949,
+      "grad_norm": 0.2860007827521177,
+      "learning_rate": 2.8795304736360184e-05,
+      "loss": 1.1063,
+      "step": 3949
+    },
+    {
+      "epoch": 0.7599579137231324,
+      "grad_norm": 0.3330913290487703,
+      "learning_rate": 2.8751561132350024e-05,
+      "loss": 1.1057,
+      "step": 3950
+    },
+    {
+      "epoch": 0.76015030813167,
+      "grad_norm": 0.34261009963535755,
+      "learning_rate": 2.8707845200412565e-05,
+      "loss": 1.0867,
+      "step": 3951
+    },
+    {
+      "epoch": 0.7603427025402074,
+      "grad_norm": 0.37266434923093356,
+      "learning_rate": 2.866415695752649e-05,
+      "loss": 1.0558,
+      "step": 3952
+    },
+    {
+      "epoch": 0.7605350969487449,
+      "grad_norm": 0.31175911979080434,
+      "learning_rate": 2.862049642065986e-05,
+      "loss": 1.031,
+      "step": 3953
+    },
+    {
+      "epoch": 0.7607274913572825,
+      "grad_norm": 0.370050910254662,
+      "learning_rate": 2.857686360676991e-05,
+      "loss": 1.1007,
+      "step": 3954
+    },
+    {
+      "epoch": 0.7609198857658199,
+      "grad_norm": 0.3421057723878658,
+      "learning_rate": 2.853325853280312e-05,
+      "loss": 1.0797,
+      "step": 3955
+    },
+    {
+      "epoch": 0.7611122801743574,
+      "grad_norm": 0.3832631987897684,
+      "learning_rate": 2.848968121569524e-05,
+      "loss": 1.054,
+      "step": 3956
+    },
+    {
+      "epoch": 0.761304674582895,
+      "grad_norm": 0.37275664296860284,
+      "learning_rate": 2.8446131672371135e-05,
+      "loss": 1.0912,
+      "step": 3957
+    },
+    {
+      "epoch": 0.7614970689914324,
+      "grad_norm": 0.3304387465465502,
+      "learning_rate": 2.8402609919744972e-05,
+      "loss": 0.983,
+      "step": 3958
+    },
+    {
+      "epoch": 0.7616894633999699,
+      "grad_norm": 0.30194453818165623,
+      "learning_rate": 2.8359115974720096e-05,
+      "loss": 1.0302,
+      "step": 3959
+    },
+    {
+      "epoch": 0.7618818578085075,
+      "grad_norm": 0.28967776736045325,
+      "learning_rate": 2.8315649854189062e-05,
+      "loss": 1.129,
+      "step": 3960
+    },
+    {
+      "epoch": 0.762074252217045,
+      "grad_norm": 0.30055465406949444,
+      "learning_rate": 2.8272211575033636e-05,
+      "loss": 1.069,
+      "step": 3961
+    },
+    {
+      "epoch": 0.7622666466255824,
+      "grad_norm": 0.3180402868348469,
+      "learning_rate": 2.8228801154124685e-05,
+      "loss": 1.0631,
+      "step": 3962
+    },
+    {
+      "epoch": 0.76245904103412,
+      "grad_norm": 0.3229325701714879,
+      "learning_rate": 2.8185418608322344e-05,
+      "loss": 1.0492,
+      "step": 3963
+    },
+    {
+      "epoch": 0.7626514354426575,
+      "grad_norm": 0.3128977147863264,
+      "learning_rate": 2.8142063954475927e-05,
+      "loss": 1.0094,
+      "step": 3964
+    },
+    {
+      "epoch": 0.7628438298511949,
+      "grad_norm": 0.3463474623140369,
+      "learning_rate": 2.8098737209423797e-05,
+      "loss": 1.0431,
+      "step": 3965
+    },
+    {
+      "epoch": 0.7630362242597325,
+      "grad_norm": 0.35924171537174326,
+      "learning_rate": 2.8055438389993637e-05,
+      "loss": 1.0167,
+      "step": 3966
+    },
+    {
+      "epoch": 0.76322861866827,
+      "grad_norm": 0.4179875926715584,
+      "learning_rate": 2.801216751300223e-05,
+      "loss": 1.1283,
+      "step": 3967
+    },
+    {
+      "epoch": 0.7634210130768074,
+      "grad_norm": 0.3081530657285735,
+      "learning_rate": 2.7968924595255407e-05,
+      "loss": 1.0301,
+      "step": 3968
+    },
+    {
+      "epoch": 0.763613407485345,
+      "grad_norm": 0.31872859290972033,
+      "learning_rate": 2.792570965354829e-05,
+      "loss": 1.1135,
+      "step": 3969
+    },
+    {
+      "epoch": 0.7638058018938825,
+      "grad_norm": 0.3244520426746989,
+      "learning_rate": 2.788252270466497e-05,
+      "loss": 1.1205,
+      "step": 3970
+    },
+    {
+      "epoch": 0.7639981963024199,
+      "grad_norm": 0.3705981929460267,
+      "learning_rate": 2.7839363765378858e-05,
+      "loss": 0.9965,
+      "step": 3971
+    },
+    {
+      "epoch": 0.7641905907109575,
+      "grad_norm": 0.31379350959289637,
+      "learning_rate": 2.7796232852452377e-05,
+      "loss": 0.9563,
+      "step": 3972
+    },
+    {
+      "epoch": 0.764382985119495,
+      "grad_norm": 0.31576895213515366,
+      "learning_rate": 2.7753129982637026e-05,
+      "loss": 1.0151,
+      "step": 3973
+    },
+    {
+      "epoch": 0.7645753795280325,
+      "grad_norm": 0.2757703721846459,
+      "learning_rate": 2.7710055172673488e-05,
+      "loss": 1.0403,
+      "step": 3974
+    },
+    {
+      "epoch": 0.7647677739365699,
+      "grad_norm": 0.3968575039251857,
+      "learning_rate": 2.7667008439291554e-05,
+      "loss": 0.9872,
+      "step": 3975
+    },
+    {
+      "epoch": 0.7649601683451075,
+      "grad_norm": 0.3630510456845327,
+      "learning_rate": 2.762398979920998e-05,
+      "loss": 1.1546,
+      "step": 3976
+    },
+    {
+      "epoch": 0.765152562753645,
+      "grad_norm": 0.26759335699516074,
+      "learning_rate": 2.7580999269136855e-05,
+      "loss": 1.0407,
+      "step": 3977
+    },
+    {
+      "epoch": 0.7653449571621824,
+      "grad_norm": 0.2884702975494463,
+      "learning_rate": 2.753803686576909e-05,
+      "loss": 1.0551,
+      "step": 3978
+    },
+    {
+      "epoch": 0.76553735157072,
+      "grad_norm": 0.3156002348613051,
+      "learning_rate": 2.7495102605792823e-05,
+      "loss": 1.0669,
+      "step": 3979
+    },
+    {
+      "epoch": 0.7657297459792575,
+      "grad_norm": 0.2885612326710155,
+      "learning_rate": 2.7452196505883265e-05,
+      "loss": 1.0386,
+      "step": 3980
+    },
+    {
+      "epoch": 0.7659221403877949,
+      "grad_norm": 0.3047927811751598,
+      "learning_rate": 2.740931858270459e-05,
+      "loss": 1.087,
+      "step": 3981
+    },
+    {
+      "epoch": 0.7661145347963325,
+      "grad_norm": 0.30871845918423163,
+      "learning_rate": 2.736646885291011e-05,
+      "loss": 1.0866,
+      "step": 3982
+    },
+    {
+      "epoch": 0.76630692920487,
+      "grad_norm": 0.3388384889666327,
+      "learning_rate": 2.7323647333142177e-05,
+      "loss": 1.0817,
+      "step": 3983
+    },
+    {
+      "epoch": 0.7664993236134074,
+      "grad_norm": 0.46564906665760697,
+      "learning_rate": 2.7280854040032165e-05,
+      "loss": 1.0577,
+      "step": 3984
+    },
+    {
+      "epoch": 0.766691718021945,
+      "grad_norm": 0.38134204213507994,
+      "learning_rate": 2.723808899020054e-05,
+      "loss": 0.9901,
+      "step": 3985
+    },
+    {
+      "epoch": 0.7668841124304825,
+      "grad_norm": 0.2965798398979628,
+      "learning_rate": 2.7195352200256674e-05,
+      "loss": 1.0966,
+      "step": 3986
+    },
+    {
+      "epoch": 0.76707650683902,
+      "grad_norm": 0.32761091388500674,
+      "learning_rate": 2.7152643686799094e-05,
+      "loss": 1.0445,
+      "step": 3987
+    },
+    {
+      "epoch": 0.7672689012475575,
+      "grad_norm": 0.2864715208624762,
+      "learning_rate": 2.710996346641528e-05,
+      "loss": 1.0355,
+      "step": 3988
+    },
+    {
+      "epoch": 0.767461295656095,
+      "grad_norm": 0.29936767644953005,
+      "learning_rate": 2.7067311555681753e-05,
+      "loss": 1.0557,
+      "step": 3989
+    },
+    {
+      "epoch": 0.7676536900646325,
+      "grad_norm": 0.3538820552764843,
+      "learning_rate": 2.702468797116403e-05,
+      "loss": 1.0068,
+      "step": 3990
+    },
+    {
+      "epoch": 0.76784608447317,
+      "grad_norm": 0.2966248862206346,
+      "learning_rate": 2.6982092729416587e-05,
+      "loss": 1.0498,
+      "step": 3991
+    },
+    {
+      "epoch": 0.7680384788817075,
+      "grad_norm": 0.29188308974065,
+      "learning_rate": 2.693952584698294e-05,
+      "loss": 1.0997,
+      "step": 3992
+    },
+    {
+      "epoch": 0.768230873290245,
+      "grad_norm": 0.3067091517693415,
+      "learning_rate": 2.6896987340395607e-05,
+      "loss": 1.0556,
+      "step": 3993
+    },
+    {
+      "epoch": 0.7684232676987826,
+      "grad_norm": 0.355518303275076,
+      "learning_rate": 2.6854477226175966e-05,
+      "loss": 1.0518,
+      "step": 3994
+    },
+    {
+      "epoch": 0.76861566210732,
+      "grad_norm": 0.39579897449486906,
+      "learning_rate": 2.6811995520834542e-05,
+      "loss": 1.1298,
+      "step": 3995
+    },
+    {
+      "epoch": 0.7688080565158575,
+      "grad_norm": 0.30281406871292027,
+      "learning_rate": 2.676954224087075e-05,
+      "loss": 1.021,
+      "step": 3996
+    },
+    {
+      "epoch": 0.7690004509243951,
+      "grad_norm": 0.3343113865114453,
+      "learning_rate": 2.6727117402772884e-05,
+      "loss": 1.0501,
+      "step": 3997
+    },
+    {
+      "epoch": 0.7691928453329325,
+      "grad_norm": 0.33618485022413086,
+      "learning_rate": 2.668472102301829e-05,
+      "loss": 0.9161,
+      "step": 3998
+    },
+    {
+      "epoch": 0.76938523974147,
+      "grad_norm": 0.3314610417572652,
+      "learning_rate": 2.664235311807327e-05,
+      "loss": 0.9741,
+      "step": 3999
+    },
+    {
+      "epoch": 0.7695776341500075,
+      "grad_norm": 0.3814914781027189,
+      "learning_rate": 2.6600013704392946e-05,
+      "loss": 1.1172,
+      "step": 4000
+    },
+    {
+      "epoch": 0.769770028558545,
+      "grad_norm": 0.37555382252086283,
+      "learning_rate": 2.6557702798421568e-05,
+      "loss": 0.9957,
+      "step": 4001
+    },
+    {
+      "epoch": 0.7699624229670825,
+      "grad_norm": 0.3285949684646368,
+      "learning_rate": 2.6515420416592106e-05,
+      "loss": 0.9902,
+      "step": 4002
+    },
+    {
+      "epoch": 0.77015481737562,
+      "grad_norm": 0.3272101517196965,
+      "learning_rate": 2.6473166575326604e-05,
+      "loss": 1.0029,
+      "step": 4003
+    },
+    {
+      "epoch": 0.7703472117841575,
+      "grad_norm": 0.3428322977780469,
+      "learning_rate": 2.643094129103598e-05,
+      "loss": 1.0134,
+      "step": 4004
+    },
+    {
+      "epoch": 0.770539606192695,
+      "grad_norm": 0.38987663832369335,
+      "learning_rate": 2.6388744580119974e-05,
+      "loss": 1.0692,
+      "step": 4005
+    },
+    {
+      "epoch": 0.7707320006012325,
+      "grad_norm": 0.30997584813617124,
+      "learning_rate": 2.6346576458967398e-05,
+      "loss": 1.077,
+      "step": 4006
+    },
+    {
+      "epoch": 0.7709243950097701,
+      "grad_norm": 0.32291783511857325,
+      "learning_rate": 2.630443694395579e-05,
+      "loss": 1.026,
+      "step": 4007
+    },
+    {
+      "epoch": 0.7711167894183075,
+      "grad_norm": 0.3869910695556213,
+      "learning_rate": 2.626232605145168e-05,
+      "loss": 1.0728,
+      "step": 4008
+    },
+    {
+      "epoch": 0.771309183826845,
+      "grad_norm": 0.2626535001714557,
+      "learning_rate": 2.6220243797810485e-05,
+      "loss": 1.0662,
+      "step": 4009
+    },
+    {
+      "epoch": 0.7715015782353826,
+      "grad_norm": 0.28559239083046223,
+      "learning_rate": 2.6178190199376396e-05,
+      "loss": 1.0688,
+      "step": 4010
+    },
+    {
+      "epoch": 0.77169397264392,
+      "grad_norm": 0.2975260127829443,
+      "learning_rate": 2.6136165272482594e-05,
+      "loss": 1.0007,
+      "step": 4011
+    },
+    {
+      "epoch": 0.7718863670524575,
+      "grad_norm": 0.2799224205086977,
+      "learning_rate": 2.6094169033451066e-05,
+      "loss": 1.1006,
+      "step": 4012
+    },
+    {
+      "epoch": 0.7720787614609951,
+      "grad_norm": 0.3384434653268322,
+      "learning_rate": 2.6052201498592667e-05,
+      "loss": 1.0642,
+      "step": 4013
+    },
+    {
+      "epoch": 0.7722711558695325,
+      "grad_norm": 0.38372593099633906,
+      "learning_rate": 2.6010262684207133e-05,
+      "loss": 1.1053,
+      "step": 4014
+    },
+    {
+      "epoch": 0.77246355027807,
+      "grad_norm": 0.3479139414621463,
+      "learning_rate": 2.596835260658297e-05,
+      "loss": 1.062,
+      "step": 4015
+    },
+    {
+      "epoch": 0.7726559446866076,
+      "grad_norm": 0.35415062773453876,
+      "learning_rate": 2.5926471281997577e-05,
+      "loss": 1.1097,
+      "step": 4016
+    },
+    {
+      "epoch": 0.772848339095145,
+      "grad_norm": 0.3110110335495718,
+      "learning_rate": 2.588461872671719e-05,
+      "loss": 1.1365,
+      "step": 4017
+    },
+    {
+      "epoch": 0.7730407335036825,
+      "grad_norm": 0.36186703670247633,
+      "learning_rate": 2.5842794956996865e-05,
+      "loss": 1.1057,
+      "step": 4018
+    },
+    {
+      "epoch": 0.7732331279122201,
+      "grad_norm": 0.33453047473056935,
+      "learning_rate": 2.5800999989080487e-05,
+      "loss": 1.0175,
+      "step": 4019
+    },
+    {
+      "epoch": 0.7734255223207576,
+      "grad_norm": 0.308594799556121,
+      "learning_rate": 2.575923383920069e-05,
+      "loss": 1.0826,
+      "step": 4020
+    },
+    {
+      "epoch": 0.773617916729295,
+      "grad_norm": 0.3469832755249014,
+      "learning_rate": 2.5717496523578998e-05,
+      "loss": 1.198,
+      "step": 4021
+    },
+    {
+      "epoch": 0.7738103111378326,
+      "grad_norm": 0.3534675435459272,
+      "learning_rate": 2.567578805842572e-05,
+      "loss": 1.1401,
+      "step": 4022
+    },
+    {
+      "epoch": 0.7740027055463701,
+      "grad_norm": 0.33708336044036163,
+      "learning_rate": 2.5634108459939877e-05,
+      "loss": 0.9757,
+      "step": 4023
+    },
+    {
+      "epoch": 0.7741950999549075,
+      "grad_norm": 0.3608360995548934,
+      "learning_rate": 2.5592457744309404e-05,
+      "loss": 1.0324,
+      "step": 4024
+    },
+    {
+      "epoch": 0.7743874943634451,
+      "grad_norm": 0.4737744576119984,
+      "learning_rate": 2.555083592771098e-05,
+      "loss": 1.0569,
+      "step": 4025
+    },
+    {
+      "epoch": 0.7745798887719826,
+      "grad_norm": 0.3067064860873511,
+      "learning_rate": 2.5509243026309982e-05,
+      "loss": 1.0173,
+      "step": 4026
+    },
+    {
+      "epoch": 0.77477228318052,
+      "grad_norm": 0.3430453853679413,
+      "learning_rate": 2.546767905626063e-05,
+      "loss": 1.0577,
+      "step": 4027
+    },
+    {
+      "epoch": 0.7749646775890575,
+      "grad_norm": 0.3397715439844128,
+      "learning_rate": 2.5426144033705935e-05,
+      "loss": 0.9836,
+      "step": 4028
+    },
+    {
+      "epoch": 0.7751570719975951,
+      "grad_norm": 0.417246128979903,
+      "learning_rate": 2.5384637974777515e-05,
+      "loss": 0.9902,
+      "step": 4029
+    },
+    {
+      "epoch": 0.7753494664061326,
+      "grad_norm": 0.27043056158554457,
+      "learning_rate": 2.5343160895595974e-05,
+      "loss": 1.0361,
+      "step": 4030
+    },
+    {
+      "epoch": 0.77554186081467,
+      "grad_norm": 0.27470490079667864,
+      "learning_rate": 2.530171281227044e-05,
+      "loss": 0.9939,
+      "step": 4031
+    },
+    {
+      "epoch": 0.7757342552232076,
+      "grad_norm": 0.2983011796892629,
+      "learning_rate": 2.52602937408989e-05,
+      "loss": 0.9877,
+      "step": 4032
+    },
+    {
+      "epoch": 0.7759266496317451,
+      "grad_norm": 0.3408758864071013,
+      "learning_rate": 2.5218903697568076e-05,
+      "loss": 1.0961,
+      "step": 4033
+    },
+    {
+      "epoch": 0.7761190440402825,
+      "grad_norm": 0.4308523035375894,
+      "learning_rate": 2.5177542698353317e-05,
+      "loss": 1.0146,
+      "step": 4034
+    },
+    {
+      "epoch": 0.7763114384488201,
+      "grad_norm": 0.2615637980872723,
+      "learning_rate": 2.5136210759318812e-05,
+      "loss": 1.035,
+      "step": 4035
+    },
+    {
+      "epoch": 0.7765038328573576,
+      "grad_norm": 0.30360119620260373,
+      "learning_rate": 2.5094907896517383e-05,
+      "loss": 1.0835,
+      "step": 4036
+    },
+    {
+      "epoch": 0.776696227265895,
+      "grad_norm": 0.348795839967339,
+      "learning_rate": 2.5053634125990587e-05,
+      "loss": 1.0246,
+      "step": 4037
+    },
+    {
+      "epoch": 0.7768886216744326,
+      "grad_norm": 0.33124235723714646,
+      "learning_rate": 2.5012389463768737e-05,
+      "loss": 1.0569,
+      "step": 4038
+    },
+    {
+      "epoch": 0.7770810160829701,
+      "grad_norm": 0.3895946928595779,
+      "learning_rate": 2.4971173925870693e-05,
+      "loss": 1.0102,
+      "step": 4039
+    },
+    {
+      "epoch": 0.7772734104915076,
+      "grad_norm": 0.27675749777566394,
+      "learning_rate": 2.4929987528304144e-05,
+      "loss": 1.0166,
+      "step": 4040
+    },
+    {
+      "epoch": 0.7774658049000451,
+      "grad_norm": 0.29956076371373347,
+      "learning_rate": 2.4888830287065412e-05,
+      "loss": 1.1142,
+      "step": 4041
+    },
+    {
+      "epoch": 0.7776581993085826,
+      "grad_norm": 0.3301533640561284,
+      "learning_rate": 2.484770221813949e-05,
+      "loss": 1.0976,
+      "step": 4042
+    },
+    {
+      "epoch": 0.7778505937171201,
+      "grad_norm": 0.34265420369693467,
+      "learning_rate": 2.4806603337500067e-05,
+      "loss": 0.9647,
+      "step": 4043
+    },
+    {
+      "epoch": 0.7780429881256576,
+      "grad_norm": 0.2734141583041449,
+      "learning_rate": 2.476553366110944e-05,
+      "loss": 1.0359,
+      "step": 4044
+    },
+    {
+      "epoch": 0.7782353825341951,
+      "grad_norm": 0.3015593986202266,
+      "learning_rate": 2.4724493204918596e-05,
+      "loss": 1.1342,
+      "step": 4045
+    },
+    {
+      "epoch": 0.7784277769427326,
+      "grad_norm": 0.2836461622925106,
+      "learning_rate": 2.4683481984867207e-05,
+      "loss": 1.074,
+      "step": 4046
+    },
+    {
+      "epoch": 0.7786201713512702,
+      "grad_norm": 0.43075273223704935,
+      "learning_rate": 2.4642500016883528e-05,
+      "loss": 1.0518,
+      "step": 4047
+    },
+    {
+      "epoch": 0.7788125657598076,
+      "grad_norm": 0.2819826605565669,
+      "learning_rate": 2.4601547316884543e-05,
+      "loss": 0.9599,
+      "step": 4048
+    },
+    {
+      "epoch": 0.7790049601683451,
+      "grad_norm": 0.3592609629587011,
+      "learning_rate": 2.4560623900775727e-05,
+      "loss": 1.0204,
+      "step": 4049
+    },
+    {
+      "epoch": 0.7791973545768827,
+      "grad_norm": 0.31598771590263175,
+      "learning_rate": 2.4519729784451295e-05,
+      "loss": 1.0321,
+      "step": 4050
+    },
+    {
+      "epoch": 0.7793897489854201,
+      "grad_norm": 0.3115038335678325,
+      "learning_rate": 2.4478864983794093e-05,
+      "loss": 1.0711,
+      "step": 4051
+    },
+    {
+      "epoch": 0.7795821433939576,
+      "grad_norm": 0.3357033311514998,
+      "learning_rate": 2.443802951467544e-05,
+      "loss": 1.0346,
+      "step": 4052
+    },
+    {
+      "epoch": 0.7797745378024951,
+      "grad_norm": 0.3419375048866406,
+      "learning_rate": 2.4397223392955447e-05,
+      "loss": 1.0076,
+      "step": 4053
+    },
+    {
+      "epoch": 0.7799669322110326,
+      "grad_norm": 0.39253540462547415,
+      "learning_rate": 2.4356446634482754e-05,
+      "loss": 1.0256,
+      "step": 4054
+    },
+    {
+      "epoch": 0.7801593266195701,
+      "grad_norm": 0.3394767314481405,
+      "learning_rate": 2.4315699255094515e-05,
+      "loss": 1.1117,
+      "step": 4055
+    },
+    {
+      "epoch": 0.7803517210281076,
+      "grad_norm": 0.3528472401618156,
+      "learning_rate": 2.4274981270616583e-05,
+      "loss": 0.9493,
+      "step": 4056
+    },
+    {
+      "epoch": 0.7805441154366451,
+      "grad_norm": 0.2999978646617924,
+      "learning_rate": 2.4234292696863358e-05,
+      "loss": 1.0493,
+      "step": 4057
+    },
+    {
+      "epoch": 0.7807365098451826,
+      "grad_norm": 0.3305828017367092,
+      "learning_rate": 2.419363354963776e-05,
+      "loss": 0.934,
+      "step": 4058
+    },
+    {
+      "epoch": 0.7809289042537201,
+      "grad_norm": 0.3131905305565756,
+      "learning_rate": 2.4153003844731425e-05,
+      "loss": 1.0465,
+      "step": 4059
+    },
+    {
+      "epoch": 0.7811212986622577,
+      "grad_norm": 0.311768628899431,
+      "learning_rate": 2.4112403597924384e-05,
+      "loss": 1.0466,
+      "step": 4060
+    },
+    {
+      "epoch": 0.7813136930707951,
+      "grad_norm": 0.31447781093658556,
+      "learning_rate": 2.407183282498534e-05,
+      "loss": 1.0875,
+      "step": 4061
+    },
+    {
+      "epoch": 0.7815060874793326,
+      "grad_norm": 0.31905288290593364,
+      "learning_rate": 2.403129154167153e-05,
+      "loss": 1.0108,
+      "step": 4062
+    },
+    {
+      "epoch": 0.7816984818878702,
+      "grad_norm": 0.2985873251330728,
+      "learning_rate": 2.3990779763728666e-05,
+      "loss": 1.0744,
+      "step": 4063
+    },
+    {
+      "epoch": 0.7818908762964076,
+      "grad_norm": 0.2806404781368688,
+      "learning_rate": 2.3950297506891084e-05,
+      "loss": 1.0115,
+      "step": 4064
+    },
+    {
+      "epoch": 0.7820832707049451,
+      "grad_norm": 0.30439516972919456,
+      "learning_rate": 2.390984478688164e-05,
+      "loss": 1.0799,
+      "step": 4065
+    },
+    {
+      "epoch": 0.7822756651134827,
+      "grad_norm": 0.28342634146353857,
+      "learning_rate": 2.386942161941169e-05,
+      "loss": 1.0225,
+      "step": 4066
+    },
+    {
+      "epoch": 0.7824680595220201,
+      "grad_norm": 0.38117154873694775,
+      "learning_rate": 2.3829028020181155e-05,
+      "loss": 1.1151,
+      "step": 4067
+    },
+    {
+      "epoch": 0.7826604539305576,
+      "grad_norm": 0.36278378623171886,
+      "learning_rate": 2.37886640048784e-05,
+      "loss": 0.9536,
+      "step": 4068
+    },
+    {
+      "epoch": 0.7828528483390952,
+      "grad_norm": 0.33845556870268984,
+      "learning_rate": 2.374832958918035e-05,
+      "loss": 1.0774,
+      "step": 4069
+    },
+    {
+      "epoch": 0.7830452427476327,
+      "grad_norm": 0.2963930998501609,
+      "learning_rate": 2.370802478875245e-05,
+      "loss": 1.1321,
+      "step": 4070
+    },
+    {
+      "epoch": 0.7832376371561701,
+      "grad_norm": 0.2681621272910231,
+      "learning_rate": 2.3667749619248614e-05,
+      "loss": 1.1041,
+      "step": 4071
+    },
+    {
+      "epoch": 0.7834300315647077,
+      "grad_norm": 0.27603200007243145,
+      "learning_rate": 2.362750409631127e-05,
+      "loss": 1.1226,
+      "step": 4072
+    },
+    {
+      "epoch": 0.7836224259732452,
+      "grad_norm": 0.28788382478046276,
+      "learning_rate": 2.3587288235571258e-05,
+      "loss": 1.0804,
+      "step": 4073
+    },
+    {
+      "epoch": 0.7838148203817826,
+      "grad_norm": 0.3370535206192947,
+      "learning_rate": 2.3547102052648006e-05,
+      "loss": 0.9467,
+      "step": 4074
+    },
+    {
+      "epoch": 0.7840072147903202,
+      "grad_norm": 0.34986406364633094,
+      "learning_rate": 2.350694556314934e-05,
+      "loss": 1.0536,
+      "step": 4075
+    },
+    {
+      "epoch": 0.7841996091988577,
+      "grad_norm": 0.27240378127577797,
+      "learning_rate": 2.3466818782671596e-05,
+      "loss": 1.0921,
+      "step": 4076
+    },
+    {
+      "epoch": 0.7843920036073951,
+      "grad_norm": 0.2812033355321229,
+      "learning_rate": 2.3426721726799573e-05,
+      "loss": 1.0325,
+      "step": 4077
+    },
+    {
+      "epoch": 0.7845843980159327,
+      "grad_norm": 0.34031903232887356,
+      "learning_rate": 2.3386654411106447e-05,
+      "loss": 1.0514,
+      "step": 4078
+    },
+    {
+      "epoch": 0.7847767924244702,
+      "grad_norm": 0.3046854755934307,
+      "learning_rate": 2.3346616851153933e-05,
+      "loss": 1.0023,
+      "step": 4079
+    },
+    {
+      "epoch": 0.7849691868330076,
+      "grad_norm": 0.32665786393550217,
+      "learning_rate": 2.330660906249218e-05,
+      "loss": 1.0162,
+      "step": 4080
+    },
+    {
+      "epoch": 0.7851615812415451,
+      "grad_norm": 0.3214540785873001,
+      "learning_rate": 2.3266631060659682e-05,
+      "loss": 1.0427,
+      "step": 4081
+    },
+    {
+      "epoch": 0.7853539756500827,
+      "grad_norm": 0.35325189378993976,
+      "learning_rate": 2.3226682861183503e-05,
+      "loss": 0.9747,
+      "step": 4082
+    },
+    {
+      "epoch": 0.7855463700586202,
+      "grad_norm": 0.3210495994588188,
+      "learning_rate": 2.318676447957907e-05,
+      "loss": 1.0729,
+      "step": 4083
+    },
+    {
+      "epoch": 0.7857387644671576,
+      "grad_norm": 0.2880568003418139,
+      "learning_rate": 2.3146875931350166e-05,
+      "loss": 0.9654,
+      "step": 4084
+    },
+    {
+      "epoch": 0.7859311588756952,
+      "grad_norm": 0.28606074795437947,
+      "learning_rate": 2.310701723198908e-05,
+      "loss": 1.0605,
+      "step": 4085
+    },
+    {
+      "epoch": 0.7861235532842327,
+      "grad_norm": 0.28690140399214414,
+      "learning_rate": 2.306718839697648e-05,
+      "loss": 1.0703,
+      "step": 4086
+    },
+    {
+      "epoch": 0.7863159476927701,
+      "grad_norm": 0.28690140399214414,
+      "learning_rate": 2.306718839697648e-05,
+      "loss": 1.102,
+      "step": 4087
+    },
+    {
+      "epoch": 0.7865083421013077,
+      "grad_norm": 0.2901468566759315,
+      "learning_rate": 2.3027389441781366e-05,
+      "loss": 1.0635,
+      "step": 4088
+    },
+    {
+      "epoch": 0.7867007365098452,
+      "grad_norm": 0.4363590695428236,
+      "learning_rate": 2.2987620381861285e-05,
+      "loss": 1.0649,
+      "step": 4089
+    },
+    {
+      "epoch": 0.7868931309183826,
+      "grad_norm": 0.30619329026243897,
+      "learning_rate": 2.2947881232662006e-05,
+      "loss": 1.0209,
+      "step": 4090
+    },
+    {
+      "epoch": 0.7870855253269202,
+      "grad_norm": 0.3140748196324708,
+      "learning_rate": 2.290817200961779e-05,
+      "loss": 0.9419,
+      "step": 4091
+    },
+    {
+      "epoch": 0.7872779197354577,
+      "grad_norm": 0.276800857021627,
+      "learning_rate": 2.2868492728151258e-05,
+      "loss": 1.0503,
+      "step": 4092
+    },
+    {
+      "epoch": 0.7874703141439952,
+      "grad_norm": 0.2720990516313734,
+      "learning_rate": 2.282884340367334e-05,
+      "loss": 1.0954,
+      "step": 4093
+    },
+    {
+      "epoch": 0.7876627085525327,
+      "grad_norm": 0.3307703329262707,
+      "learning_rate": 2.2789224051583403e-05,
+      "loss": 0.9635,
+      "step": 4094
+    },
+    {
+      "epoch": 0.7878551029610702,
+      "grad_norm": 0.33231629737588114,
+      "learning_rate": 2.274963468726914e-05,
+      "loss": 1.0863,
+      "step": 4095
+    },
+    {
+      "epoch": 0.7880474973696077,
+      "grad_norm": 0.3252652585237453,
+      "learning_rate": 2.2710075326106617e-05,
+      "loss": 1.0173,
+      "step": 4096
+    },
+    {
+      "epoch": 0.7882398917781452,
+      "grad_norm": 0.3220119322351853,
+      "learning_rate": 2.2670545983460243e-05,
+      "loss": 1.0327,
+      "step": 4097
+    },
+    {
+      "epoch": 0.7884322861866827,
+      "grad_norm": 0.3501618840649719,
+      "learning_rate": 2.263104667468272e-05,
+      "loss": 1.0293,
+      "step": 4098
+    },
+    {
+      "epoch": 0.7886246805952202,
+      "grad_norm": 0.31475961614477027,
+      "learning_rate": 2.259157741511515e-05,
+      "loss": 1.0332,
+      "step": 4099
+    },
+    {
+      "epoch": 0.7888170750037578,
+      "grad_norm": 0.37200819679911085,
+      "learning_rate": 2.2552138220086927e-05,
+      "loss": 1.0637,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7890094694122952,
+      "grad_norm": 0.3278409661272568,
+      "learning_rate": 2.2512729104915786e-05,
+      "loss": 0.9842,
+      "step": 4101
+    },
+    {
+      "epoch": 0.7892018638208327,
+      "grad_norm": 0.2960526646228673,
+      "learning_rate": 2.2473350084907806e-05,
+      "loss": 1.0734,
+      "step": 4102
+    },
+    {
+      "epoch": 0.7893942582293703,
+      "grad_norm": 0.392294481516786,
+      "learning_rate": 2.243400117535729e-05,
+      "loss": 1.0329,
+      "step": 4103
+    },
+    {
+      "epoch": 0.7895866526379077,
+      "grad_norm": 0.4708160441664302,
+      "learning_rate": 2.2394682391546928e-05,
+      "loss": 0.9351,
+      "step": 4104
+    },
+    {
+      "epoch": 0.7897790470464452,
+      "grad_norm": 0.34827263568985245,
+      "learning_rate": 2.23553937487477e-05,
+      "loss": 1.0339,
+      "step": 4105
+    },
+    {
+      "epoch": 0.7899714414549827,
+      "grad_norm": 0.3523262710672966,
+      "learning_rate": 2.2316135262218785e-05,
+      "loss": 0.9533,
+      "step": 4106
+    },
+    {
+      "epoch": 0.7901638358635202,
+      "grad_norm": 0.3467557651511863,
+      "learning_rate": 2.2276906947207844e-05,
+      "loss": 1.0628,
+      "step": 4107
+    },
+    {
+      "epoch": 0.7903562302720577,
+      "grad_norm": 0.2813250676517662,
+      "learning_rate": 2.2237708818950607e-05,
+      "loss": 0.9986,
+      "step": 4108
+    },
+    {
+      "epoch": 0.7905486246805952,
+      "grad_norm": 0.30823412400173916,
+      "learning_rate": 2.2198540892671215e-05,
+      "loss": 1.0536,
+      "step": 4109
+    },
+    {
+      "epoch": 0.7907410190891327,
+      "grad_norm": 0.29080862082233805,
+      "learning_rate": 2.215940318358206e-05,
+      "loss": 1.0727,
+      "step": 4110
+    },
+    {
+      "epoch": 0.7909334134976702,
+      "grad_norm": 0.3680696143843968,
+      "learning_rate": 2.2120295706883698e-05,
+      "loss": 1.0495,
+      "step": 4111
+    },
+    {
+      "epoch": 0.7911258079062077,
+      "grad_norm": 0.31380449212150596,
+      "learning_rate": 2.20812184777651e-05,
+      "loss": 1.0288,
+      "step": 4112
+    },
+    {
+      "epoch": 0.7913182023147453,
+      "grad_norm": 0.2743720210322785,
+      "learning_rate": 2.204217151140342e-05,
+      "loss": 1.0336,
+      "step": 4113
+    },
+    {
+      "epoch": 0.7915105967232827,
+      "grad_norm": 0.34140327841539325,
+      "learning_rate": 2.2003154822963978e-05,
+      "loss": 1.0649,
+      "step": 4114
+    },
+    {
+      "epoch": 0.7917029911318202,
+      "grad_norm": 0.3045718665457332,
+      "learning_rate": 2.196416842760046e-05,
+      "loss": 0.9731,
+      "step": 4115
+    },
+    {
+      "epoch": 0.7918953855403578,
+      "grad_norm": 0.33949965571767204,
+      "learning_rate": 2.1925212340454736e-05,
+      "loss": 0.9842,
+      "step": 4116
+    },
+    {
+      "epoch": 0.7920877799488952,
+      "grad_norm": 0.3102855794383734,
+      "learning_rate": 2.1886286576656835e-05,
+      "loss": 1.0188,
+      "step": 4117
+    },
+    {
+      "epoch": 0.7922801743574327,
+      "grad_norm": 0.3383555322216625,
+      "learning_rate": 2.184739115132517e-05,
+      "loss": 1.0952,
+      "step": 4118
+    },
+    {
+      "epoch": 0.7924725687659703,
+      "grad_norm": 0.302114197381981,
+      "learning_rate": 2.1808526079566217e-05,
+      "loss": 1.0001,
+      "step": 4119
+    },
+    {
+      "epoch": 0.7926649631745077,
+      "grad_norm": 0.30712967246705486,
+      "learning_rate": 2.176969137647472e-05,
+      "loss": 1.0103,
+      "step": 4120
+    },
+    {
+      "epoch": 0.7928573575830452,
+      "grad_norm": 0.33138108721276416,
+      "learning_rate": 2.1730887057133677e-05,
+      "loss": 1.0247,
+      "step": 4121
+    },
+    {
+      "epoch": 0.7930497519915828,
+      "grad_norm": 0.34536020755544034,
+      "learning_rate": 2.1692113136614177e-05,
+      "loss": 0.9655,
+      "step": 4122
+    },
+    {
+      "epoch": 0.7932421464001203,
+      "grad_norm": 0.2878807027628804,
+      "learning_rate": 2.1653369629975595e-05,
+      "loss": 1.0556,
+      "step": 4123
+    },
+    {
+      "epoch": 0.7934345408086577,
+      "grad_norm": 0.302434305619991,
+      "learning_rate": 2.1614656552265456e-05,
+      "loss": 1.0239,
+      "step": 4124
+    },
+    {
+      "epoch": 0.7936269352171953,
+      "grad_norm": 0.2625604788862428,
+      "learning_rate": 2.1575973918519486e-05,
+      "loss": 1.0518,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7938193296257328,
+      "grad_norm": 0.3250231557601072,
+      "learning_rate": 2.1537321743761584e-05,
+      "loss": 0.9883,
+      "step": 4126
+    },
+    {
+      "epoch": 0.7940117240342702,
+      "grad_norm": 0.3161924556206067,
+      "learning_rate": 2.1498700043003773e-05,
+      "loss": 1.1181,
+      "step": 4127
+    },
+    {
+      "epoch": 0.7942041184428078,
+      "grad_norm": 0.36467580323113497,
+      "learning_rate": 2.1460108831246296e-05,
+      "loss": 1.0927,
+      "step": 4128
+    },
+    {
+      "epoch": 0.7943965128513453,
+      "grad_norm": 0.30205927433122226,
+      "learning_rate": 2.1421548123477532e-05,
+      "loss": 1.1026,
+      "step": 4129
+    },
+    {
+      "epoch": 0.7945889072598827,
+      "grad_norm": 0.3389016658532511,
+      "learning_rate": 2.138301793467401e-05,
+      "loss": 1.0773,
+      "step": 4130
+    },
+    {
+      "epoch": 0.7947813016684203,
+      "grad_norm": 0.2936165561136769,
+      "learning_rate": 2.1344518279800452e-05,
+      "loss": 1.0376,
+      "step": 4131
+    },
+    {
+      "epoch": 0.7949736960769578,
+      "grad_norm": 0.3244203334235697,
+      "learning_rate": 2.1306049173809615e-05,
+      "loss": 1.0074,
+      "step": 4132
+    },
+    {
+      "epoch": 0.7951660904854952,
+      "grad_norm": 0.3179106524364337,
+      "learning_rate": 2.1267610631642498e-05,
+      "loss": 1.0484,
+      "step": 4133
+    },
+    {
+      "epoch": 0.7953584848940327,
+      "grad_norm": 0.2872263205387523,
+      "learning_rate": 2.1229202668228197e-05,
+      "loss": 1.086,
+      "step": 4134
+    },
+    {
+      "epoch": 0.7955508793025703,
+      "grad_norm": 0.31792386893201335,
+      "learning_rate": 2.1190825298483852e-05,
+      "loss": 1.0655,
+      "step": 4135
+    },
+    {
+      "epoch": 0.7957432737111078,
+      "grad_norm": 0.3239416101435148,
+      "learning_rate": 2.115247853731488e-05,
+      "loss": 1.0877,
+      "step": 4136
+    },
+    {
+      "epoch": 0.7959356681196452,
+      "grad_norm": 0.28517445114722245,
+      "learning_rate": 2.11141623996147e-05,
+      "loss": 1.102,
+      "step": 4137
+    },
+    {
+      "epoch": 0.7961280625281828,
+      "grad_norm": 0.32037924197412265,
+      "learning_rate": 2.107587690026481e-05,
+      "loss": 0.9985,
+      "step": 4138
+    },
+    {
+      "epoch": 0.7963204569367203,
+      "grad_norm": 0.2873254753145235,
+      "learning_rate": 2.103762205413493e-05,
+      "loss": 1.0813,
+      "step": 4139
+    },
+    {
+      "epoch": 0.7965128513452577,
+      "grad_norm": 0.3233192744702412,
+      "learning_rate": 2.0999397876082728e-05,
+      "loss": 1.035,
+      "step": 4140
+    },
+    {
+      "epoch": 0.7967052457537953,
+      "grad_norm": 0.2961188079461089,
+      "learning_rate": 2.0961204380954036e-05,
+      "loss": 0.9644,
+      "step": 4141
+    },
+    {
+      "epoch": 0.7968976401623328,
+      "grad_norm": 0.33631847165289963,
+      "learning_rate": 2.092304158358286e-05,
+      "loss": 1.102,
+      "step": 4142
+    },
+    {
+      "epoch": 0.7970900345708702,
+      "grad_norm": 0.33500412017339415,
+      "learning_rate": 2.0884909498791104e-05,
+      "loss": 1.004,
+      "step": 4143
+    },
+    {
+      "epoch": 0.7972824289794078,
+      "grad_norm": 0.30883720658660496,
+      "learning_rate": 2.084680814138885e-05,
+      "loss": 0.9892,
+      "step": 4144
+    },
+    {
+      "epoch": 0.7974748233879453,
+      "grad_norm": 0.2818898328146198,
+      "learning_rate": 2.080873752617426e-05,
+      "loss": 0.9751,
+      "step": 4145
+    },
+    {
+      "epoch": 0.7976672177964828,
+      "grad_norm": 0.3585731462953241,
+      "learning_rate": 2.0770697667933437e-05,
+      "loss": 1.1006,
+      "step": 4146
+    },
+    {
+      "epoch": 0.7978596122050203,
+      "grad_norm": 0.2926286408042117,
+      "learning_rate": 2.073268858144074e-05,
+      "loss": 0.9301,
+      "step": 4147
+    },
+    {
+      "epoch": 0.7980520066135578,
+      "grad_norm": 0.29597234117131,
+      "learning_rate": 2.0694710281458373e-05,
+      "loss": 1.0619,
+      "step": 4148
+    },
+    {
+      "epoch": 0.7982444010220953,
+      "grad_norm": 0.3089719433180061,
+      "learning_rate": 2.0656762782736693e-05,
+      "loss": 1.1035,
+      "step": 4149
+    },
+    {
+      "epoch": 0.7984367954306328,
+      "grad_norm": 0.3454499371124246,
+      "learning_rate": 2.0618846100014112e-05,
+      "loss": 0.986,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7986291898391703,
+      "grad_norm": 0.2938694679943425,
+      "learning_rate": 2.058096024801697e-05,
+      "loss": 1.2005,
+      "step": 4151
+    },
+    {
+      "epoch": 0.7988215842477078,
+      "grad_norm": 0.3510400028153644,
+      "learning_rate": 2.0543105241459715e-05,
+      "loss": 1.0692,
+      "step": 4152
+    },
+    {
+      "epoch": 0.7990139786562453,
+      "grad_norm": 0.3770395642067201,
+      "learning_rate": 2.0505281095044804e-05,
+      "loss": 1.0484,
+      "step": 4153
+    },
+    {
+      "epoch": 0.7992063730647828,
+      "grad_norm": 0.361746577092918,
+      "learning_rate": 2.0467487823462695e-05,
+      "loss": 1.0363,
+      "step": 4154
+    },
+    {
+      "epoch": 0.7993987674733203,
+      "grad_norm": 0.31233626713597756,
+      "learning_rate": 2.0429725441391888e-05,
+      "loss": 0.9825,
+      "step": 4155
+    },
+    {
+      "epoch": 0.7995911618818579,
+      "grad_norm": 0.3245864247833704,
+      "learning_rate": 2.0391993963498813e-05,
+      "loss": 0.9743,
+      "step": 4156
+    },
+    {
+      "epoch": 0.7997835562903953,
+      "grad_norm": 0.3105149021097417,
+      "learning_rate": 2.0354293404437965e-05,
+      "loss": 0.9958,
+      "step": 4157
+    },
+    {
+      "epoch": 0.7999759506989328,
+      "grad_norm": 0.3379959066219544,
+      "learning_rate": 2.0316623778851783e-05,
+      "loss": 0.969,
+      "step": 4158
+    },
+    {
+      "epoch": 0.8001683451074704,
+      "grad_norm": 0.2859122086892328,
+      "learning_rate": 2.027898510137075e-05,
+      "loss": 1.0724,
+      "step": 4159
+    },
+    {
+      "epoch": 0.8003607395160078,
+      "grad_norm": 0.34975418711012796,
+      "learning_rate": 2.024137738661329e-05,
+      "loss": 0.9916,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8005531339245453,
+      "grad_norm": 0.283130966063517,
+      "learning_rate": 2.0203800649185788e-05,
+      "loss": 1.0596,
+      "step": 4161
+    },
+    {
+      "epoch": 0.8007455283330828,
+      "grad_norm": 0.4038353386203545,
+      "learning_rate": 2.0166254903682603e-05,
+      "loss": 1.1049,
+      "step": 4162
+    },
+    {
+      "epoch": 0.8009379227416203,
+      "grad_norm": 0.3347735083989096,
+      "learning_rate": 2.0128740164686134e-05,
+      "loss": 1.0415,
+      "step": 4163
+    },
+    {
+      "epoch": 0.8011303171501578,
+      "grad_norm": 0.33174471154216084,
+      "learning_rate": 2.009125644676656e-05,
+      "loss": 1.0897,
+      "step": 4164
+    },
+    {
+      "epoch": 0.8013227115586953,
+      "grad_norm": 0.2889547781948744,
+      "learning_rate": 2.0053803764482227e-05,
+      "loss": 1.0109,
+      "step": 4165
+    },
+    {
+      "epoch": 0.8015151059672329,
+      "grad_norm": 0.3208742641100502,
+      "learning_rate": 2.0016382132379318e-05,
+      "loss": 1.0773,
+      "step": 4166
+    },
+    {
+      "epoch": 0.8017075003757703,
+      "grad_norm": 0.3036094637011469,
+      "learning_rate": 1.9978991564991913e-05,
+      "loss": 1.0771,
+      "step": 4167
+    },
+    {
+      "epoch": 0.8018998947843078,
+      "grad_norm": 0.4080670188996166,
+      "learning_rate": 1.994163207684212e-05,
+      "loss": 1.0422,
+      "step": 4168
+    },
+    {
+      "epoch": 0.8020922891928454,
+      "grad_norm": 0.3492900620252333,
+      "learning_rate": 1.9904303682439897e-05,
+      "loss": 1.063,
+      "step": 4169
+    },
+    {
+      "epoch": 0.8022846836013828,
+      "grad_norm": 0.2751590063723222,
+      "learning_rate": 1.986700639628316e-05,
+      "loss": 0.9833,
+      "step": 4170
+    },
+    {
+      "epoch": 0.8024770780099203,
+      "grad_norm": 0.34622534863558896,
+      "learning_rate": 1.9829740232857808e-05,
+      "loss": 0.9761,
+      "step": 4171
+    },
+    {
+      "epoch": 0.8026694724184579,
+      "grad_norm": 0.3159037009911428,
+      "learning_rate": 1.9792505206637524e-05,
+      "loss": 0.944,
+      "step": 4172
+    },
+    {
+      "epoch": 0.8028618668269953,
+      "grad_norm": 0.2983641727337406,
+      "learning_rate": 1.9755301332083997e-05,
+      "loss": 0.945,
+      "step": 4173
+    },
+    {
+      "epoch": 0.8030542612355328,
+      "grad_norm": 0.3546053658805946,
+      "learning_rate": 1.971812862364679e-05,
+      "loss": 1.0628,
+      "step": 4174
+    },
+    {
+      "epoch": 0.8032466556440704,
+      "grad_norm": 0.44468887222521297,
+      "learning_rate": 1.9680987095763313e-05,
+      "loss": 1.1259,
+      "step": 4175
+    },
+    {
+      "epoch": 0.8034390500526079,
+      "grad_norm": 0.30665208772499136,
+      "learning_rate": 1.9643876762858937e-05,
+      "loss": 1.0578,
+      "step": 4176
+    },
+    {
+      "epoch": 0.8036314444611453,
+      "grad_norm": 0.2890170031227219,
+      "learning_rate": 1.9606797639346874e-05,
+      "loss": 1.0149,
+      "step": 4177
+    },
+    {
+      "epoch": 0.8038238388696829,
+      "grad_norm": 0.30784286792350757,
+      "learning_rate": 1.9569749739628242e-05,
+      "loss": 1.0046,
+      "step": 4178
+    },
+    {
+      "epoch": 0.8040162332782204,
+      "grad_norm": 0.2784803979754186,
+      "learning_rate": 1.9532733078092037e-05,
+      "loss": 1.0657,
+      "step": 4179
+    },
+    {
+      "epoch": 0.8042086276867578,
+      "grad_norm": 0.3399332301498935,
+      "learning_rate": 1.949574766911506e-05,
+      "loss": 1.0492,
+      "step": 4180
+    },
+    {
+      "epoch": 0.8044010220952954,
+      "grad_norm": 0.32112326353734333,
+      "learning_rate": 1.9458793527062035e-05,
+      "loss": 1.108,
+      "step": 4181
+    },
+    {
+      "epoch": 0.8045934165038329,
+      "grad_norm": 0.30822347698308206,
+      "learning_rate": 1.9421870666285524e-05,
+      "loss": 1.0282,
+      "step": 4182
+    },
+    {
+      "epoch": 0.8047858109123703,
+      "grad_norm": 0.3359397490488531,
+      "learning_rate": 1.9384979101125943e-05,
+      "loss": 1.0438,
+      "step": 4183
+    },
+    {
+      "epoch": 0.8049782053209079,
+      "grad_norm": 0.2557747986998689,
+      "learning_rate": 1.934811884591159e-05,
+      "loss": 1.1201,
+      "step": 4184
+    },
+    {
+      "epoch": 0.8051705997294454,
+      "grad_norm": 0.32207592881392766,
+      "learning_rate": 1.9311289914958497e-05,
+      "loss": 0.961,
+      "step": 4185
+    },
+    {
+      "epoch": 0.8053629941379828,
+      "grad_norm": 0.36412405685639604,
+      "learning_rate": 1.9274492322570615e-05,
+      "loss": 1.0364,
+      "step": 4186
+    },
+    {
+      "epoch": 0.8055553885465203,
+      "grad_norm": 0.2712936491840502,
+      "learning_rate": 1.923772608303972e-05,
+      "loss": 1.0981,
+      "step": 4187
+    },
+    {
+      "epoch": 0.8057477829550579,
+      "grad_norm": 0.31081075169339534,
+      "learning_rate": 1.920099121064539e-05,
+      "loss": 1.0806,
+      "step": 4188
+    },
+    {
+      "epoch": 0.8059401773635954,
+      "grad_norm": 0.3567761356839382,
+      "learning_rate": 1.9164287719655062e-05,
+      "loss": 1.0246,
+      "step": 4189
+    },
+    {
+      "epoch": 0.8061325717721328,
+      "grad_norm": 0.3404313493082168,
+      "learning_rate": 1.912761562432388e-05,
+      "loss": 1.0629,
+      "step": 4190
+    },
+    {
+      "epoch": 0.8063249661806704,
+      "grad_norm": 0.31730127105323097,
+      "learning_rate": 1.90909749388949e-05,
+      "loss": 1.1147,
+      "step": 4191
+    },
+    {
+      "epoch": 0.8065173605892079,
+      "grad_norm": 0.3250360068257254,
+      "learning_rate": 1.905436567759896e-05,
+      "loss": 1.007,
+      "step": 4192
+    },
+    {
+      "epoch": 0.8067097549977453,
+      "grad_norm": 0.3563080686918488,
+      "learning_rate": 1.901778785465461e-05,
+      "loss": 0.9851,
+      "step": 4193
+    },
+    {
+      "epoch": 0.8069021494062829,
+      "grad_norm": 0.3655226410932478,
+      "learning_rate": 1.898124148426832e-05,
+      "loss": 0.9737,
+      "step": 4194
+    },
+    {
+      "epoch": 0.8070945438148204,
+      "grad_norm": 0.31423297386602744,
+      "learning_rate": 1.8944726580634288e-05,
+      "loss": 1.1074,
+      "step": 4195
+    },
+    {
+      "epoch": 0.8072869382233578,
+      "grad_norm": 0.31576186727571104,
+      "learning_rate": 1.8908243157934423e-05,
+      "loss": 1.0248,
+      "step": 4196
+    },
+    {
+      "epoch": 0.8074793326318954,
+      "grad_norm": 0.3002823181128077,
+      "learning_rate": 1.8871791230338497e-05,
+      "loss": 1.1222,
+      "step": 4197
+    },
+    {
+      "epoch": 0.8076717270404329,
+      "grad_norm": 0.4105801184366847,
+      "learning_rate": 1.883537081200404e-05,
+      "loss": 0.9563,
+      "step": 4198
+    },
+    {
+      "epoch": 0.8078641214489704,
+      "grad_norm": 0.392641757221708,
+      "learning_rate": 1.8798981917076252e-05,
+      "loss": 1.0261,
+      "step": 4199
+    },
+    {
+      "epoch": 0.8080565158575079,
+      "grad_norm": 0.39782525902805865,
+      "learning_rate": 1.8762624559688256e-05,
+      "loss": 1.097,
+      "step": 4200
+    },
+    {
+      "epoch": 0.8082489102660454,
+      "grad_norm": 0.34534136262599896,
+      "learning_rate": 1.872629875396076e-05,
+      "loss": 1.0925,
+      "step": 4201
+    },
+    {
+      "epoch": 0.8084413046745829,
+      "grad_norm": 0.3509857390532199,
+      "learning_rate": 1.8690004514002313e-05,
+      "loss": 0.9611,
+      "step": 4202
+    },
+    {
+      "epoch": 0.8086336990831204,
+      "grad_norm": 0.3678830763617161,
+      "learning_rate": 1.86537418539092e-05,
+      "loss": 1.059,
+      "step": 4203
+    },
+    {
+      "epoch": 0.8088260934916579,
+      "grad_norm": 0.3260728921005822,
+      "learning_rate": 1.861751078776538e-05,
+      "loss": 0.9822,
+      "step": 4204
+    },
+    {
+      "epoch": 0.8090184879001954,
+      "grad_norm": 0.2638828516373688,
+      "learning_rate": 1.858131132964259e-05,
+      "loss": 1.0576,
+      "step": 4205
+    },
+    {
+      "epoch": 0.809210882308733,
+      "grad_norm": 0.2785485216506586,
+      "learning_rate": 1.8545143493600292e-05,
+      "loss": 0.9734,
+      "step": 4206
+    },
+    {
+      "epoch": 0.8094032767172704,
+      "grad_norm": 0.3809734279293067,
+      "learning_rate": 1.8509007293685667e-05,
+      "loss": 1.0106,
+      "step": 4207
+    },
+    {
+      "epoch": 0.8095956711258079,
+      "grad_norm": 0.36715539055737184,
+      "learning_rate": 1.8472902743933607e-05,
+      "loss": 1.0611,
+      "step": 4208
+    },
+    {
+      "epoch": 0.8097880655343455,
+      "grad_norm": 0.35580377380689066,
+      "learning_rate": 1.8436829858366657e-05,
+      "loss": 1.0038,
+      "step": 4209
+    },
+    {
+      "epoch": 0.8099804599428829,
+      "grad_norm": 0.39956965072590633,
+      "learning_rate": 1.840078865099514e-05,
+      "loss": 1.0844,
+      "step": 4210
+    },
+    {
+      "epoch": 0.8101728543514204,
+      "grad_norm": 0.31385576585429487,
+      "learning_rate": 1.8364779135817044e-05,
+      "loss": 1.0548,
+      "step": 4211
+    },
+    {
+      "epoch": 0.810365248759958,
+      "grad_norm": 0.35592901429668644,
+      "learning_rate": 1.8328801326818046e-05,
+      "loss": 1.0441,
+      "step": 4212
+    },
+    {
+      "epoch": 0.8105576431684954,
+      "grad_norm": 0.3017258135852835,
+      "learning_rate": 1.829285523797155e-05,
+      "loss": 1.0783,
+      "step": 4213
+    },
+    {
+      "epoch": 0.8107500375770329,
+      "grad_norm": 0.35817488309160284,
+      "learning_rate": 1.8256940883238538e-05,
+      "loss": 0.988,
+      "step": 4214
+    },
+    {
+      "epoch": 0.8109424319855704,
+      "grad_norm": 0.3878324351241288,
+      "learning_rate": 1.822105827656776e-05,
+      "loss": 1.159,
+      "step": 4215
+    },
+    {
+      "epoch": 0.8111348263941079,
+      "grad_norm": 0.3971895888591802,
+      "learning_rate": 1.818520743189561e-05,
+      "loss": 1.0366,
+      "step": 4216
+    },
+    {
+      "epoch": 0.8113272208026454,
+      "grad_norm": 0.36828337174297815,
+      "learning_rate": 1.814938836314615e-05,
+      "loss": 1.0499,
+      "step": 4217
+    },
+    {
+      "epoch": 0.8115196152111829,
+      "grad_norm": 0.31104518170028983,
+      "learning_rate": 1.8113601084231092e-05,
+      "loss": 1.0339,
+      "step": 4218
+    },
+    {
+      "epoch": 0.8117120096197205,
+      "grad_norm": 0.3135971832808061,
+      "learning_rate": 1.8077845609049782e-05,
+      "loss": 0.9906,
+      "step": 4219
+    },
+    {
+      "epoch": 0.8119044040282579,
+      "grad_norm": 0.35736961857313027,
+      "learning_rate": 1.8042121951489256e-05,
+      "loss": 1.0318,
+      "step": 4220
+    },
+    {
+      "epoch": 0.8120967984367954,
+      "grad_norm": 0.34467707918278906,
+      "learning_rate": 1.800643012542418e-05,
+      "loss": 1.0819,
+      "step": 4221
+    },
+    {
+      "epoch": 0.812289192845333,
+      "grad_norm": 0.3352034971155479,
+      "learning_rate": 1.7970770144716774e-05,
+      "loss": 1.0092,
+      "step": 4222
+    },
+    {
+      "epoch": 0.8124815872538704,
+      "grad_norm": 0.31872179140101714,
+      "learning_rate": 1.7935142023217057e-05,
+      "loss": 0.8971,
+      "step": 4223
+    },
+    {
+      "epoch": 0.8126739816624079,
+      "grad_norm": 0.34400960317966045,
+      "learning_rate": 1.789954577476257e-05,
+      "loss": 1.0387,
+      "step": 4224
+    },
+    {
+      "epoch": 0.8128663760709455,
+      "grad_norm": 0.3223066208945814,
+      "learning_rate": 1.786398141317843e-05,
+      "loss": 1.0419,
+      "step": 4225
+    },
+    {
+      "epoch": 0.8130587704794829,
+      "grad_norm": 0.3153892303556073,
+      "learning_rate": 1.7828448952277453e-05,
+      "loss": 1.1225,
+      "step": 4226
+    },
+    {
+      "epoch": 0.8132511648880204,
+      "grad_norm": 0.2751318071489437,
+      "learning_rate": 1.7792948405860077e-05,
+      "loss": 1.0039,
+      "step": 4227
+    },
+    {
+      "epoch": 0.813443559296558,
+      "grad_norm": 0.30815973679032055,
+      "learning_rate": 1.7757479787714217e-05,
+      "loss": 1.0134,
+      "step": 4228
+    },
+    {
+      "epoch": 0.8136359537050954,
+      "grad_norm": 0.34134710946116986,
+      "learning_rate": 1.7722043111615573e-05,
+      "loss": 1.0606,
+      "step": 4229
+    },
+    {
+      "epoch": 0.8138283481136329,
+      "grad_norm": 0.3277704416233078,
+      "learning_rate": 1.7686638391327272e-05,
+      "loss": 1.0352,
+      "step": 4230
+    },
+    {
+      "epoch": 0.8140207425221705,
+      "grad_norm": 0.3642671718728077,
+      "learning_rate": 1.765126564060011e-05,
+      "loss": 1.05,
+      "step": 4231
+    },
+    {
+      "epoch": 0.814213136930708,
+      "grad_norm": 0.322402312179894,
+      "learning_rate": 1.7615924873172507e-05,
+      "loss": 1.0651,
+      "step": 4232
+    },
+    {
+      "epoch": 0.8144055313392454,
+      "grad_norm": 0.2989501305355325,
+      "learning_rate": 1.7580616102770354e-05,
+      "loss": 1.0064,
+      "step": 4233
+    },
+    {
+      "epoch": 0.814597925747783,
+      "grad_norm": 0.30865533082422814,
+      "learning_rate": 1.754533934310717e-05,
+      "loss": 1.0269,
+      "step": 4234
+    },
+    {
+      "epoch": 0.8147903201563205,
+      "grad_norm": 0.2813133994460399,
+      "learning_rate": 1.7510094607884074e-05,
+      "loss": 1.0484,
+      "step": 4235
+    },
+    {
+      "epoch": 0.8149827145648579,
+      "grad_norm": 0.29309108104577763,
+      "learning_rate": 1.74748819107897e-05,
+      "loss": 1.0361,
+      "step": 4236
+    },
+    {
+      "epoch": 0.8151751089733955,
+      "grad_norm": 0.31017364253929985,
+      "learning_rate": 1.7439701265500273e-05,
+      "loss": 1.0103,
+      "step": 4237
+    },
+    {
+      "epoch": 0.815367503381933,
+      "grad_norm": 0.33639039321781367,
+      "learning_rate": 1.7404552685679508e-05,
+      "loss": 1.0795,
+      "step": 4238
+    },
+    {
+      "epoch": 0.8155598977904704,
+      "grad_norm": 0.31604691099661286,
+      "learning_rate": 1.7369436184978736e-05,
+      "loss": 1.0319,
+      "step": 4239
+    },
+    {
+      "epoch": 0.8157522921990079,
+      "grad_norm": 0.33841243657213893,
+      "learning_rate": 1.7334351777036805e-05,
+      "loss": 1.0887,
+      "step": 4240
+    },
+    {
+      "epoch": 0.8159446866075455,
+      "grad_norm": 0.30609860047683907,
+      "learning_rate": 1.729929947548008e-05,
+      "loss": 1.0226,
+      "step": 4241
+    },
+    {
+      "epoch": 0.816137081016083,
+      "grad_norm": 0.293788790133934,
+      "learning_rate": 1.7264279293922502e-05,
+      "loss": 1.0551,
+      "step": 4242
+    },
+    {
+      "epoch": 0.8163294754246204,
+      "grad_norm": 0.33061289289998036,
+      "learning_rate": 1.7229291245965462e-05,
+      "loss": 1.0399,
+      "step": 4243
+    },
+    {
+      "epoch": 0.816521869833158,
+      "grad_norm": 0.35072322030946385,
+      "learning_rate": 1.7194335345197932e-05,
+      "loss": 0.9791,
+      "step": 4244
+    },
+    {
+      "epoch": 0.8167142642416955,
+      "grad_norm": 0.32148579613729894,
+      "learning_rate": 1.715941160519641e-05,
+      "loss": 0.959,
+      "step": 4245
+    },
+    {
+      "epoch": 0.8169066586502329,
+      "grad_norm": 0.29383959213907057,
+      "learning_rate": 1.7124520039524803e-05,
+      "loss": 1.0015,
+      "step": 4246
+    },
+    {
+      "epoch": 0.8170990530587705,
+      "grad_norm": 0.2820336703221806,
+      "learning_rate": 1.7089660661734685e-05,
+      "loss": 1.0566,
+      "step": 4247
+    },
+    {
+      "epoch": 0.817291447467308,
+      "grad_norm": 0.2942055912019547,
+      "learning_rate": 1.705483348536496e-05,
+      "loss": 1.0596,
+      "step": 4248
+    },
+    {
+      "epoch": 0.8174838418758454,
+      "grad_norm": 0.3349384521380425,
+      "learning_rate": 1.702003852394214e-05,
+      "loss": 1.0737,
+      "step": 4249
+    },
+    {
+      "epoch": 0.817676236284383,
+      "grad_norm": 0.29960967337845307,
+      "learning_rate": 1.6985275790980203e-05,
+      "loss": 1.1079,
+      "step": 4250
+    },
+    {
+      "epoch": 0.8178686306929205,
+      "grad_norm": 0.3228212536301577,
+      "learning_rate": 1.6950545299980526e-05,
+      "loss": 1.041,
+      "step": 4251
+    },
+    {
+      "epoch": 0.818061025101458,
+      "grad_norm": 0.3152113373953018,
+      "learning_rate": 1.691584706443209e-05,
+      "loss": 1.0687,
+      "step": 4252
+    },
+    {
+      "epoch": 0.8182534195099955,
+      "grad_norm": 0.390827064483229,
+      "learning_rate": 1.6881181097811304e-05,
+      "loss": 1.0104,
+      "step": 4253
+    },
+    {
+      "epoch": 0.818445813918533,
+      "grad_norm": 0.3317988171243961,
+      "learning_rate": 1.684654741358198e-05,
+      "loss": 1.1326,
+      "step": 4254
+    },
+    {
+      "epoch": 0.8186382083270705,
+      "grad_norm": 0.3001968785209655,
+      "learning_rate": 1.681194602519546e-05,
+      "loss": 1.1085,
+      "step": 4255
+    },
+    {
+      "epoch": 0.818830602735608,
+      "grad_norm": 0.40915101868692316,
+      "learning_rate": 1.677737694609055e-05,
+      "loss": 1.0702,
+      "step": 4256
+    },
+    {
+      "epoch": 0.8190229971441455,
+      "grad_norm": 0.36230968687389703,
+      "learning_rate": 1.674284018969342e-05,
+      "loss": 1.0588,
+      "step": 4257
+    },
+    {
+      "epoch": 0.819215391552683,
+      "grad_norm": 0.3530382872668718,
+      "learning_rate": 1.6708335769417827e-05,
+      "loss": 0.9842,
+      "step": 4258
+    },
+    {
+      "epoch": 0.8194077859612205,
+      "grad_norm": 0.3191849370261009,
+      "learning_rate": 1.667386369866484e-05,
+      "loss": 1.128,
+      "step": 4259
+    },
+    {
+      "epoch": 0.819600180369758,
+      "grad_norm": 0.34008937395332556,
+      "learning_rate": 1.6639423990823012e-05,
+      "loss": 0.9227,
+      "step": 4260
+    },
+    {
+      "epoch": 0.8197925747782955,
+      "grad_norm": 0.2841751541218453,
+      "learning_rate": 1.660501665926838e-05,
+      "loss": 1.035,
+      "step": 4261
+    },
+    {
+      "epoch": 0.8199849691868331,
+      "grad_norm": 0.34085841556893776,
+      "learning_rate": 1.6570641717364277e-05,
+      "loss": 1.0777,
+      "step": 4262
+    },
+    {
+      "epoch": 0.8201773635953705,
+      "grad_norm": 0.4086163305010936,
+      "learning_rate": 1.653629917846159e-05,
+      "loss": 1.086,
+      "step": 4263
+    },
+    {
+      "epoch": 0.820369758003908,
+      "grad_norm": 0.3371792459661734,
+      "learning_rate": 1.6501989055898535e-05,
+      "loss": 1.1721,
+      "step": 4264
+    },
+    {
+      "epoch": 0.8205621524124456,
+      "grad_norm": 0.2815788276431441,
+      "learning_rate": 1.6467711363000793e-05,
+      "loss": 1.0934,
+      "step": 4265
+    },
+    {
+      "epoch": 0.820754546820983,
+      "grad_norm": 0.3628916186036442,
+      "learning_rate": 1.643346611308144e-05,
+      "loss": 1.0818,
+      "step": 4266
+    },
+    {
+      "epoch": 0.8209469412295205,
+      "grad_norm": 0.29654224487139946,
+      "learning_rate": 1.6399253319440887e-05,
+      "loss": 1.0465,
+      "step": 4267
+    },
+    {
+      "epoch": 0.821139335638058,
+      "grad_norm": 0.2937533177962177,
+      "learning_rate": 1.6365072995367004e-05,
+      "loss": 0.9724,
+      "step": 4268
+    },
+    {
+      "epoch": 0.8213317300465955,
+      "grad_norm": 0.3265450810091649,
+      "learning_rate": 1.6330925154135058e-05,
+      "loss": 1.0522,
+      "step": 4269
+    },
+    {
+      "epoch": 0.821524124455133,
+      "grad_norm": 0.4547358240015819,
+      "learning_rate": 1.629680980900765e-05,
+      "loss": 1.0914,
+      "step": 4270
+    },
+    {
+      "epoch": 0.8217165188636705,
+      "grad_norm": 0.2686243073655931,
+      "learning_rate": 1.6262726973234843e-05,
+      "loss": 1.0879,
+      "step": 4271
+    },
+    {
+      "epoch": 0.821908913272208,
+      "grad_norm": 0.3189454370898761,
+      "learning_rate": 1.622867666005393e-05,
+      "loss": 1.0132,
+      "step": 4272
+    },
+    {
+      "epoch": 0.8221013076807455,
+      "grad_norm": 0.3747357956313708,
+      "learning_rate": 1.619465888268972e-05,
+      "loss": 0.9731,
+      "step": 4273
+    },
+    {
+      "epoch": 0.822293702089283,
+      "grad_norm": 0.32076015561625526,
+      "learning_rate": 1.616067365435433e-05,
+      "loss": 1.1012,
+      "step": 4274
+    },
+    {
+      "epoch": 0.8224860964978206,
+      "grad_norm": 0.3542059668112701,
+      "learning_rate": 1.6126720988247167e-05,
+      "loss": 0.9566,
+      "step": 4275
+    },
+    {
+      "epoch": 0.822678490906358,
+      "grad_norm": 0.30753471248970976,
+      "learning_rate": 1.609280089755515e-05,
+      "loss": 0.9947,
+      "step": 4276
+    },
+    {
+      "epoch": 0.8228708853148955,
+      "grad_norm": 0.3010294221868473,
+      "learning_rate": 1.605891339545237e-05,
+      "loss": 1.1124,
+      "step": 4277
+    },
+    {
+      "epoch": 0.8230632797234331,
+      "grad_norm": 0.3522633893705517,
+      "learning_rate": 1.6025058495100385e-05,
+      "loss": 1.0783,
+      "step": 4278
+    },
+    {
+      "epoch": 0.8232556741319705,
+      "grad_norm": 0.3547404714593905,
+      "learning_rate": 1.5991236209648053e-05,
+      "loss": 1.0576,
+      "step": 4279
+    },
+    {
+      "epoch": 0.823448068540508,
+      "grad_norm": 0.2914039621274619,
+      "learning_rate": 1.5957446552231526e-05,
+      "loss": 1.0655,
+      "step": 4280
+    },
+    {
+      "epoch": 0.8236404629490456,
+      "grad_norm": 0.2785104060807799,
+      "learning_rate": 1.5923689535974305e-05,
+      "loss": 1.0728,
+      "step": 4281
+    },
+    {
+      "epoch": 0.823832857357583,
+      "grad_norm": 0.4817029438779027,
+      "learning_rate": 1.588996517398731e-05,
+      "loss": 1.0681,
+      "step": 4282
+    },
+    {
+      "epoch": 0.8240252517661205,
+      "grad_norm": 0.3406957508689732,
+      "learning_rate": 1.5856273479368612e-05,
+      "loss": 1.0252,
+      "step": 4283
+    },
+    {
+      "epoch": 0.8242176461746581,
+      "grad_norm": 0.3832770419294321,
+      "learning_rate": 1.582261446520371e-05,
+      "loss": 1.0021,
+      "step": 4284
+    },
+    {
+      "epoch": 0.8244100405831956,
+      "grad_norm": 0.35433853652434877,
+      "learning_rate": 1.5788988144565398e-05,
+      "loss": 1.0356,
+      "step": 4285
+    },
+    {
+      "epoch": 0.824602434991733,
+      "grad_norm": 0.313297882112553,
+      "learning_rate": 1.575539453051369e-05,
+      "loss": 0.9684,
+      "step": 4286
+    },
+    {
+      "epoch": 0.8247948294002706,
+      "grad_norm": 0.34122480873000194,
+      "learning_rate": 1.5721833636096027e-05,
+      "loss": 1.0628,
+      "step": 4287
+    },
+    {
+      "epoch": 0.8249872238088081,
+      "grad_norm": 0.30057871198672154,
+      "learning_rate": 1.568830547434703e-05,
+      "loss": 1.0253,
+      "step": 4288
+    },
+    {
+      "epoch": 0.8251796182173455,
+      "grad_norm": 0.31207426829107815,
+      "learning_rate": 1.565481005828866e-05,
+      "loss": 1.0598,
+      "step": 4289
+    },
+    {
+      "epoch": 0.8253720126258831,
+      "grad_norm": 0.3398292973627893,
+      "learning_rate": 1.5621347400930175e-05,
+      "loss": 1.0325,
+      "step": 4290
+    },
+    {
+      "epoch": 0.8255644070344206,
+      "grad_norm": 0.4328416035109433,
+      "learning_rate": 1.5587917515268048e-05,
+      "loss": 1.0173,
+      "step": 4291
+    },
+    {
+      "epoch": 0.825756801442958,
+      "grad_norm": 0.29339438472408647,
+      "learning_rate": 1.5554520414286067e-05,
+      "loss": 1.014,
+      "step": 4292
+    },
+    {
+      "epoch": 0.8259491958514956,
+      "grad_norm": 0.347747159471914,
+      "learning_rate": 1.552115611095529e-05,
+      "loss": 1.0482,
+      "step": 4293
+    },
+    {
+      "epoch": 0.8261415902600331,
+      "grad_norm": 0.34020470481192366,
+      "learning_rate": 1.5487824618234047e-05,
+      "loss": 1.0218,
+      "step": 4294
+    },
+    {
+      "epoch": 0.8263339846685706,
+      "grad_norm": 0.27092672233754667,
+      "learning_rate": 1.54545259490679e-05,
+      "loss": 1.0925,
+      "step": 4295
+    },
+    {
+      "epoch": 0.826526379077108,
+      "grad_norm": 0.28364747899207743,
+      "learning_rate": 1.5421260116389637e-05,
+      "loss": 1.1335,
+      "step": 4296
+    },
+    {
+      "epoch": 0.8267187734856456,
+      "grad_norm": 0.2954520583421916,
+      "learning_rate": 1.5388027133119342e-05,
+      "loss": 1.0517,
+      "step": 4297
+    },
+    {
+      "epoch": 0.8269111678941831,
+      "grad_norm": 0.32338285222766205,
+      "learning_rate": 1.535482701216433e-05,
+      "loss": 1.1162,
+      "step": 4298
+    },
+    {
+      "epoch": 0.8271035623027205,
+      "grad_norm": 0.3834289541311469,
+      "learning_rate": 1.5321659766419128e-05,
+      "loss": 1.0339,
+      "step": 4299
+    },
+    {
+      "epoch": 0.8272959567112581,
+      "grad_norm": 0.32922442128799073,
+      "learning_rate": 1.5288525408765564e-05,
+      "loss": 1.0048,
+      "step": 4300
+    },
+    {
+      "epoch": 0.8274883511197956,
+      "grad_norm": 0.3496787458178123,
+      "learning_rate": 1.5255423952072567e-05,
+      "loss": 0.963,
+      "step": 4301
+    },
+    {
+      "epoch": 0.827680745528333,
+      "grad_norm": 0.28275325227655657,
+      "learning_rate": 1.5222355409196398e-05,
+      "loss": 0.9594,
+      "step": 4302
+    },
+    {
+      "epoch": 0.8278731399368706,
+      "grad_norm": 0.36837139051448053,
+      "learning_rate": 1.5189319792980516e-05,
+      "loss": 1.0367,
+      "step": 4303
+    },
+    {
+      "epoch": 0.8280655343454081,
+      "grad_norm": 0.2954420350130657,
+      "learning_rate": 1.5156317116255513e-05,
+      "loss": 1.0374,
+      "step": 4304
+    },
+    {
+      "epoch": 0.8282579287539455,
+      "grad_norm": 0.38732593372836654,
+      "learning_rate": 1.5123347391839305e-05,
+      "loss": 1.0311,
+      "step": 4305
+    },
+    {
+      "epoch": 0.8284503231624831,
+      "grad_norm": 0.3092292055949576,
+      "learning_rate": 1.5090410632536967e-05,
+      "loss": 0.9978,
+      "step": 4306
+    },
+    {
+      "epoch": 0.8286427175710206,
+      "grad_norm": 0.3002129319227162,
+      "learning_rate": 1.5057506851140702e-05,
+      "loss": 1.088,
+      "step": 4307
+    },
+    {
+      "epoch": 0.8288351119795581,
+      "grad_norm": 0.310423549819509,
+      "learning_rate": 1.5024636060429997e-05,
+      "loss": 1.0912,
+      "step": 4308
+    },
+    {
+      "epoch": 0.8290275063880956,
+      "grad_norm": 0.30640879170819296,
+      "learning_rate": 1.4991798273171465e-05,
+      "loss": 1.0805,
+      "step": 4309
+    },
+    {
+      "epoch": 0.8292199007966331,
+      "grad_norm": 0.3112398536626859,
+      "learning_rate": 1.4958993502118901e-05,
+      "loss": 1.0377,
+      "step": 4310
+    },
+    {
+      "epoch": 0.8294122952051706,
+      "grad_norm": 0.31469996971858366,
+      "learning_rate": 1.4926221760013392e-05,
+      "loss": 1.0506,
+      "step": 4311
+    },
+    {
+      "epoch": 0.8296046896137081,
+      "grad_norm": 0.4022606241327596,
+      "learning_rate": 1.4893483059583014e-05,
+      "loss": 1.067,
+      "step": 4312
+    },
+    {
+      "epoch": 0.8297970840222456,
+      "grad_norm": 0.3062598702182294,
+      "learning_rate": 1.4860777413543137e-05,
+      "loss": 1.1611,
+      "step": 4313
+    },
+    {
+      "epoch": 0.8299894784307831,
+      "grad_norm": 0.3889458385107906,
+      "learning_rate": 1.4828104834596268e-05,
+      "loss": 0.9347,
+      "step": 4314
+    },
+    {
+      "epoch": 0.8301818728393207,
+      "grad_norm": 0.33928997942988326,
+      "learning_rate": 1.4795465335432035e-05,
+      "loss": 0.945,
+      "step": 4315
+    },
+    {
+      "epoch": 0.8303742672478581,
+      "grad_norm": 0.37590035648628306,
+      "learning_rate": 1.4762858928727241e-05,
+      "loss": 1.0919,
+      "step": 4316
+    },
+    {
+      "epoch": 0.8305666616563956,
+      "grad_norm": 0.28229467482843107,
+      "learning_rate": 1.4730285627145856e-05,
+      "loss": 1.0615,
+      "step": 4317
+    },
+    {
+      "epoch": 0.8307590560649332,
+      "grad_norm": 0.3403883019779579,
+      "learning_rate": 1.4697745443338984e-05,
+      "loss": 1.0085,
+      "step": 4318
+    },
+    {
+      "epoch": 0.8309514504734706,
+      "grad_norm": 0.3108788465793099,
+      "learning_rate": 1.4665238389944857e-05,
+      "loss": 1.0103,
+      "step": 4319
+    },
+    {
+      "epoch": 0.8311438448820081,
+      "grad_norm": 0.3495461101060005,
+      "learning_rate": 1.4632764479588801e-05,
+      "loss": 1.0028,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8313362392905456,
+      "grad_norm": 0.3453501409524611,
+      "learning_rate": 1.4600323724883335e-05,
+      "loss": 0.9827,
+      "step": 4321
+    },
+    {
+      "epoch": 0.8315286336990831,
+      "grad_norm": 0.295247115020464,
+      "learning_rate": 1.4567916138428072e-05,
+      "loss": 1.0298,
+      "step": 4322
+    },
+    {
+      "epoch": 0.8317210281076206,
+      "grad_norm": 0.3712462561953952,
+      "learning_rate": 1.4535541732809754e-05,
+      "loss": 1.0807,
+      "step": 4323
+    },
+    {
+      "epoch": 0.8319134225161581,
+      "grad_norm": 0.34911513546897444,
+      "learning_rate": 1.4503200520602245e-05,
+      "loss": 0.9743,
+      "step": 4324
+    },
+    {
+      "epoch": 0.8321058169246957,
+      "grad_norm": 0.34512041064861837,
+      "learning_rate": 1.4470892514366441e-05,
+      "loss": 1.0319,
+      "step": 4325
+    },
+    {
+      "epoch": 0.8322982113332331,
+      "grad_norm": 0.30514220474916104,
+      "learning_rate": 1.4438617726650439e-05,
+      "loss": 1.0587,
+      "step": 4326
+    },
+    {
+      "epoch": 0.8324906057417706,
+      "grad_norm": 0.2873211724007842,
+      "learning_rate": 1.4406376169989388e-05,
+      "loss": 1.0059,
+      "step": 4327
+    },
+    {
+      "epoch": 0.8326830001503082,
+      "grad_norm": 0.29469615819157635,
+      "learning_rate": 1.437416785690554e-05,
+      "loss": 1.0403,
+      "step": 4328
+    },
+    {
+      "epoch": 0.8328753945588456,
+      "grad_norm": 0.3553241588287279,
+      "learning_rate": 1.4341992799908255e-05,
+      "loss": 0.9453,
+      "step": 4329
+    },
+    {
+      "epoch": 0.8330677889673831,
+      "grad_norm": 0.30591594191161253,
+      "learning_rate": 1.4309851011493903e-05,
+      "loss": 1.0181,
+      "step": 4330
+    },
+    {
+      "epoch": 0.8332601833759207,
+      "grad_norm": 0.2885401639784365,
+      "learning_rate": 1.4277742504146008e-05,
+      "loss": 1.0681,
+      "step": 4331
+    },
+    {
+      "epoch": 0.8334525777844581,
+      "grad_norm": 0.2939893159795581,
+      "learning_rate": 1.4245667290335174e-05,
+      "loss": 0.9951,
+      "step": 4332
+    },
+    {
+      "epoch": 0.8336449721929956,
+      "grad_norm": 0.28758006529882046,
+      "learning_rate": 1.4213625382518968e-05,
+      "loss": 1.1012,
+      "step": 4333
+    },
+    {
+      "epoch": 0.8338373666015332,
+      "grad_norm": 0.28569834353715723,
+      "learning_rate": 1.4181616793142172e-05,
+      "loss": 1.0933,
+      "step": 4334
+    },
+    {
+      "epoch": 0.8340297610100706,
+      "grad_norm": 0.35958759492633907,
+      "learning_rate": 1.4149641534636549e-05,
+      "loss": 1.0671,
+      "step": 4335
+    },
+    {
+      "epoch": 0.8342221554186081,
+      "grad_norm": 0.2852292793421516,
+      "learning_rate": 1.4117699619420877e-05,
+      "loss": 1.0427,
+      "step": 4336
+    },
+    {
+      "epoch": 0.8344145498271457,
+      "grad_norm": 0.31080925371030704,
+      "learning_rate": 1.4085791059901076e-05,
+      "loss": 1.0799,
+      "step": 4337
+    },
+    {
+      "epoch": 0.8346069442356832,
+      "grad_norm": 0.31890650968048345,
+      "learning_rate": 1.4053915868470013e-05,
+      "loss": 1.0675,
+      "step": 4338
+    },
+    {
+      "epoch": 0.8347993386442206,
+      "grad_norm": 0.3143199905768238,
+      "learning_rate": 1.402207405750765e-05,
+      "loss": 1.0101,
+      "step": 4339
+    },
+    {
+      "epoch": 0.8349917330527582,
+      "grad_norm": 0.31384670029948,
+      "learning_rate": 1.399026563938105e-05,
+      "loss": 0.9138,
+      "step": 4340
+    },
+    {
+      "epoch": 0.8351841274612957,
+      "grad_norm": 0.2978722899973814,
+      "learning_rate": 1.3958490626444154e-05,
+      "loss": 1.021,
+      "step": 4341
+    },
+    {
+      "epoch": 0.8353765218698331,
+      "grad_norm": 0.286398059175232,
+      "learning_rate": 1.3926749031038056e-05,
+      "loss": 0.9644,
+      "step": 4342
+    },
+    {
+      "epoch": 0.8355689162783707,
+      "grad_norm": 0.3273241169556011,
+      "learning_rate": 1.3895040865490816e-05,
+      "loss": 1.0414,
+      "step": 4343
+    },
+    {
+      "epoch": 0.8357613106869082,
+      "grad_norm": 0.5008963752393932,
+      "learning_rate": 1.3863366142117506e-05,
+      "loss": 1.0676,
+      "step": 4344
+    },
+    {
+      "epoch": 0.8359537050954456,
+      "grad_norm": 0.40182739734768624,
+      "learning_rate": 1.383172487322023e-05,
+      "loss": 1.0994,
+      "step": 4345
+    },
+    {
+      "epoch": 0.8361460995039832,
+      "grad_norm": 0.32240367638137035,
+      "learning_rate": 1.3800117071088104e-05,
+      "loss": 1.0284,
+      "step": 4346
+    },
+    {
+      "epoch": 0.8363384939125207,
+      "grad_norm": 0.30346940699395386,
+      "learning_rate": 1.3768542747997215e-05,
+      "loss": 1.0364,
+      "step": 4347
+    },
+    {
+      "epoch": 0.8365308883210582,
+      "grad_norm": 0.35952692288069216,
+      "learning_rate": 1.3737001916210713e-05,
+      "loss": 1.0285,
+      "step": 4348
+    },
+    {
+      "epoch": 0.8367232827295956,
+      "grad_norm": 0.3280341536673114,
+      "learning_rate": 1.3705494587978628e-05,
+      "loss": 0.9689,
+      "step": 4349
+    },
+    {
+      "epoch": 0.8369156771381332,
+      "grad_norm": 0.29841163661727804,
+      "learning_rate": 1.3674020775538077e-05,
+      "loss": 1.0264,
+      "step": 4350
+    },
+    {
+      "epoch": 0.8371080715466707,
+      "grad_norm": 0.3300173606199424,
+      "learning_rate": 1.3642580491113122e-05,
+      "loss": 1.0486,
+      "step": 4351
+    },
+    {
+      "epoch": 0.8373004659552081,
+      "grad_norm": 0.29583793172814193,
+      "learning_rate": 1.3611173746914795e-05,
+      "loss": 1.1074,
+      "step": 4352
+    },
+    {
+      "epoch": 0.8374928603637457,
+      "grad_norm": 0.3259445866010344,
+      "learning_rate": 1.3579800555141165e-05,
+      "loss": 1.0406,
+      "step": 4353
+    },
+    {
+      "epoch": 0.8376852547722832,
+      "grad_norm": 0.3081893561443836,
+      "learning_rate": 1.3548460927977159e-05,
+      "loss": 1.0318,
+      "step": 4354
+    },
+    {
+      "epoch": 0.8378776491808206,
+      "grad_norm": 0.39999903393994896,
+      "learning_rate": 1.351715487759474e-05,
+      "loss": 1.0302,
+      "step": 4355
+    },
+    {
+      "epoch": 0.8380700435893582,
+      "grad_norm": 0.3317252535529246,
+      "learning_rate": 1.3485882416152818e-05,
+      "loss": 1.0327,
+      "step": 4356
+    },
+    {
+      "epoch": 0.8382624379978957,
+      "grad_norm": 0.34708946151619857,
+      "learning_rate": 1.3454643555797274e-05,
+      "loss": 1.1173,
+      "step": 4357
+    },
+    {
+      "epoch": 0.8384548324064331,
+      "grad_norm": 0.40143747456039425,
+      "learning_rate": 1.342343830866093e-05,
+      "loss": 1.0181,
+      "step": 4358
+    },
+    {
+      "epoch": 0.8386472268149707,
+      "grad_norm": 0.28548195570072665,
+      "learning_rate": 1.3392266686863509e-05,
+      "loss": 1.0654,
+      "step": 4359
+    },
+    {
+      "epoch": 0.8388396212235082,
+      "grad_norm": 0.31076940655849183,
+      "learning_rate": 1.3361128702511716e-05,
+      "loss": 0.9509,
+      "step": 4360
+    },
+    {
+      "epoch": 0.8390320156320457,
+      "grad_norm": 0.2757276326788152,
+      "learning_rate": 1.3330024367699223e-05,
+      "loss": 1.1054,
+      "step": 4361
+    },
+    {
+      "epoch": 0.8392244100405832,
+      "grad_norm": 0.3736669343343043,
+      "learning_rate": 1.3298953694506522e-05,
+      "loss": 1.1259,
+      "step": 4362
+    },
+    {
+      "epoch": 0.8394168044491207,
+      "grad_norm": 0.3804990149363424,
+      "learning_rate": 1.326791669500117e-05,
+      "loss": 1.0513,
+      "step": 4363
+    },
+    {
+      "epoch": 0.8396091988576582,
+      "grad_norm": 0.3154312484184179,
+      "learning_rate": 1.3236913381237592e-05,
+      "loss": 0.9786,
+      "step": 4364
+    },
+    {
+      "epoch": 0.8398015932661957,
+      "grad_norm": 0.3857616024777847,
+      "learning_rate": 1.3205943765257055e-05,
+      "loss": 1.0056,
+      "step": 4365
+    },
+    {
+      "epoch": 0.8399939876747332,
+      "grad_norm": 0.3399896988079937,
+      "learning_rate": 1.317500785908783e-05,
+      "loss": 1.0466,
+      "step": 4366
+    },
+    {
+      "epoch": 0.8401863820832707,
+      "grad_norm": 0.2961339211810674,
+      "learning_rate": 1.31441056747451e-05,
+      "loss": 1.0089,
+      "step": 4367
+    },
+    {
+      "epoch": 0.8403787764918083,
+      "grad_norm": 0.27330043522945013,
+      "learning_rate": 1.3113237224230834e-05,
+      "loss": 0.9903,
+      "step": 4368
+    },
+    {
+      "epoch": 0.8405711709003457,
+      "grad_norm": 0.302922675005817,
+      "learning_rate": 1.3082402519534076e-05,
+      "loss": 0.9964,
+      "step": 4369
+    },
+    {
+      "epoch": 0.8407635653088832,
+      "grad_norm": 0.3113824707386979,
+      "learning_rate": 1.3051601572630612e-05,
+      "loss": 1.0192,
+      "step": 4370
+    },
+    {
+      "epoch": 0.8409559597174208,
+      "grad_norm": 0.39076202495571843,
+      "learning_rate": 1.3020834395483195e-05,
+      "loss": 1.0051,
+      "step": 4371
+    },
+    {
+      "epoch": 0.8411483541259582,
+      "grad_norm": 0.3388408607420535,
+      "learning_rate": 1.2990101000041444e-05,
+      "loss": 0.9489,
+      "step": 4372
+    },
+    {
+      "epoch": 0.8413407485344957,
+      "grad_norm": 0.3241563187854764,
+      "learning_rate": 1.2959401398241843e-05,
+      "loss": 1.0729,
+      "step": 4373
+    },
+    {
+      "epoch": 0.8415331429430332,
+      "grad_norm": 0.2942550788767924,
+      "learning_rate": 1.2928735602007769e-05,
+      "loss": 1.0623,
+      "step": 4374
+    },
+    {
+      "epoch": 0.8417255373515707,
+      "grad_norm": 0.2972173317913256,
+      "learning_rate": 1.2898103623249457e-05,
+      "loss": 1.0782,
+      "step": 4375
+    },
+    {
+      "epoch": 0.8419179317601082,
+      "grad_norm": 0.31317053641199444,
+      "learning_rate": 1.2867505473864028e-05,
+      "loss": 1.0302,
+      "step": 4376
+    },
+    {
+      "epoch": 0.8421103261686457,
+      "grad_norm": 0.3347588087456929,
+      "learning_rate": 1.2836941165735462e-05,
+      "loss": 1.0332,
+      "step": 4377
+    },
+    {
+      "epoch": 0.8423027205771833,
+      "grad_norm": 0.34002030210517803,
+      "learning_rate": 1.280641071073455e-05,
+      "loss": 1.0942,
+      "step": 4378
+    },
+    {
+      "epoch": 0.8424951149857207,
+      "grad_norm": 0.31416757185788263,
+      "learning_rate": 1.277591412071899e-05,
+      "loss": 1.181,
+      "step": 4379
+    },
+    {
+      "epoch": 0.8426875093942582,
+      "grad_norm": 0.2513049540304737,
+      "learning_rate": 1.2745451407533293e-05,
+      "loss": 0.9742,
+      "step": 4380
+    },
+    {
+      "epoch": 0.8428799038027958,
+      "grad_norm": 0.33396854133990356,
+      "learning_rate": 1.271502258300885e-05,
+      "loss": 1.0648,
+      "step": 4381
+    },
+    {
+      "epoch": 0.8430722982113332,
+      "grad_norm": 0.28811172966431414,
+      "learning_rate": 1.2684627658963867e-05,
+      "loss": 0.9838,
+      "step": 4382
+    },
+    {
+      "epoch": 0.8432646926198707,
+      "grad_norm": 0.3344094463958978,
+      "learning_rate": 1.2654266647203339e-05,
+      "loss": 1.0583,
+      "step": 4383
+    },
+    {
+      "epoch": 0.8434570870284083,
+      "grad_norm": 0.29059186098623796,
+      "learning_rate": 1.262393955951916e-05,
+      "loss": 0.9516,
+      "step": 4384
+    },
+    {
+      "epoch": 0.8436494814369457,
+      "grad_norm": 0.3154616925093502,
+      "learning_rate": 1.2593646407690051e-05,
+      "loss": 1.0943,
+      "step": 4385
+    },
+    {
+      "epoch": 0.8438418758454832,
+      "grad_norm": 0.4595138985021085,
+      "learning_rate": 1.2563387203481448e-05,
+      "loss": 1.078,
+      "step": 4386
+    },
+    {
+      "epoch": 0.8440342702540208,
+      "grad_norm": 0.34485173350393894,
+      "learning_rate": 1.2533161958645756e-05,
+      "loss": 0.9703,
+      "step": 4387
+    },
+    {
+      "epoch": 0.8442266646625582,
+      "grad_norm": 0.29043708174197574,
+      "learning_rate": 1.2502970684922066e-05,
+      "loss": 1.0914,
+      "step": 4388
+    },
+    {
+      "epoch": 0.8444190590710957,
+      "grad_norm": 0.33807419121247195,
+      "learning_rate": 1.2472813394036343e-05,
+      "loss": 1.1189,
+      "step": 4389
+    },
+    {
+      "epoch": 0.8446114534796333,
+      "grad_norm": 0.3627737516281703,
+      "learning_rate": 1.2442690097701327e-05,
+      "loss": 1.1022,
+      "step": 4390
+    },
+    {
+      "epoch": 0.8448038478881708,
+      "grad_norm": 0.300897351451403,
+      "learning_rate": 1.2412600807616525e-05,
+      "loss": 1.061,
+      "step": 4391
+    },
+    {
+      "epoch": 0.8449962422967082,
+      "grad_norm": 0.30350554626579285,
+      "learning_rate": 1.2382545535468315e-05,
+      "loss": 1.0349,
+      "step": 4392
+    },
+    {
+      "epoch": 0.8451886367052458,
+      "grad_norm": 0.3102167279099889,
+      "learning_rate": 1.2352524292929824e-05,
+      "loss": 1.0592,
+      "step": 4393
+    },
+    {
+      "epoch": 0.8453810311137833,
+      "grad_norm": 0.2902124229608211,
+      "learning_rate": 1.2322537091660912e-05,
+      "loss": 0.9995,
+      "step": 4394
+    },
+    {
+      "epoch": 0.8455734255223207,
+      "grad_norm": 0.35997306625094294,
+      "learning_rate": 1.22925839433083e-05,
+      "loss": 1.0911,
+      "step": 4395
+    },
+    {
+      "epoch": 0.8457658199308583,
+      "grad_norm": 0.2790983922193345,
+      "learning_rate": 1.2262664859505435e-05,
+      "loss": 1.0048,
+      "step": 4396
+    },
+    {
+      "epoch": 0.8459582143393958,
+      "grad_norm": 0.31723225473807776,
+      "learning_rate": 1.223277985187251e-05,
+      "loss": 1.029,
+      "step": 4397
+    },
+    {
+      "epoch": 0.8461506087479332,
+      "grad_norm": 0.34931807081170924,
+      "learning_rate": 1.2202928932016589e-05,
+      "loss": 1.0319,
+      "step": 4398
+    },
+    {
+      "epoch": 0.8463430031564708,
+      "grad_norm": 0.25205369282964313,
+      "learning_rate": 1.217311211153137e-05,
+      "loss": 1.0531,
+      "step": 4399
+    },
+    {
+      "epoch": 0.8465353975650083,
+      "grad_norm": 0.28878951052224605,
+      "learning_rate": 1.2143329401997372e-05,
+      "loss": 1.0261,
+      "step": 4400
+    },
+    {
+      "epoch": 0.8467277919735458,
+      "grad_norm": 0.30475683176157253,
+      "learning_rate": 1.2113580814981884e-05,
+      "loss": 1.0791,
+      "step": 4401
+    },
+    {
+      "epoch": 0.8469201863820832,
+      "grad_norm": 0.3590265373440576,
+      "learning_rate": 1.2083866362038865e-05,
+      "loss": 1.0454,
+      "step": 4402
+    },
+    {
+      "epoch": 0.8471125807906208,
+      "grad_norm": 0.3147493282361996,
+      "learning_rate": 1.2054186054709105e-05,
+      "loss": 1.0376,
+      "step": 4403
+    },
+    {
+      "epoch": 0.8473049751991583,
+      "grad_norm": 0.35656484326336557,
+      "learning_rate": 1.2024539904520072e-05,
+      "loss": 1.1018,
+      "step": 4404
+    },
+    {
+      "epoch": 0.8474973696076957,
+      "grad_norm": 0.31804220118379734,
+      "learning_rate": 1.1994927922985998e-05,
+      "loss": 0.9712,
+      "step": 4405
+    },
+    {
+      "epoch": 0.8476897640162333,
+      "grad_norm": 0.2815920389719055,
+      "learning_rate": 1.1965350121607865e-05,
+      "loss": 1.0473,
+      "step": 4406
+    },
+    {
+      "epoch": 0.8478821584247708,
+      "grad_norm": 0.3697235808703823,
+      "learning_rate": 1.1935806511873304e-05,
+      "loss": 1.0633,
+      "step": 4407
+    },
+    {
+      "epoch": 0.8480745528333082,
+      "grad_norm": 0.33882787882348747,
+      "learning_rate": 1.1906297105256725e-05,
+      "loss": 0.9101,
+      "step": 4408
+    },
+    {
+      "epoch": 0.8482669472418458,
+      "grad_norm": 0.38126204769137373,
+      "learning_rate": 1.187682191321925e-05,
+      "loss": 0.9732,
+      "step": 4409
+    },
+    {
+      "epoch": 0.8484593416503833,
+      "grad_norm": 0.32230133575299447,
+      "learning_rate": 1.1847380947208698e-05,
+      "loss": 1.1817,
+      "step": 4410
+    },
+    {
+      "epoch": 0.8486517360589207,
+      "grad_norm": 0.2687968482367284,
+      "learning_rate": 1.1817974218659622e-05,
+      "loss": 1.0705,
+      "step": 4411
+    },
+    {
+      "epoch": 0.8488441304674583,
+      "grad_norm": 0.29467017204237417,
+      "learning_rate": 1.178860173899321e-05,
+      "loss": 1.0357,
+      "step": 4412
+    },
+    {
+      "epoch": 0.8490365248759958,
+      "grad_norm": 0.3513843597260413,
+      "learning_rate": 1.1759263519617437e-05,
+      "loss": 0.9443,
+      "step": 4413
+    },
+    {
+      "epoch": 0.8492289192845333,
+      "grad_norm": 0.33476851316917355,
+      "learning_rate": 1.172995957192693e-05,
+      "loss": 1.0129,
+      "step": 4414
+    },
+    {
+      "epoch": 0.8494213136930708,
+      "grad_norm": 0.31263259081617994,
+      "learning_rate": 1.1700689907302953e-05,
+      "loss": 1.0095,
+      "step": 4415
+    },
+    {
+      "epoch": 0.8496137081016083,
+      "grad_norm": 0.3430615587125829,
+      "learning_rate": 1.167145453711358e-05,
+      "loss": 1.0359,
+      "step": 4416
+    },
+    {
+      "epoch": 0.8498061025101458,
+      "grad_norm": 0.3062578995978995,
+      "learning_rate": 1.1642253472713426e-05,
+      "loss": 1.003,
+      "step": 4417
+    },
+    {
+      "epoch": 0.8499984969186833,
+      "grad_norm": 0.33071943347271965,
+      "learning_rate": 1.161308672544389e-05,
+      "loss": 0.9642,
+      "step": 4418
+    },
+    {
+      "epoch": 0.8501908913272208,
+      "grad_norm": 0.3224894187239732,
+      "learning_rate": 1.1583954306633005e-05,
+      "loss": 1.0617,
+      "step": 4419
+    },
+    {
+      "epoch": 0.8503832857357583,
+      "grad_norm": 0.331184047299763,
+      "learning_rate": 1.1554856227595433e-05,
+      "loss": 1.197,
+      "step": 4420
+    },
+    {
+      "epoch": 0.8505756801442959,
+      "grad_norm": 0.48725607724834413,
+      "learning_rate": 1.1525792499632527e-05,
+      "loss": 1.0276,
+      "step": 4421
+    },
+    {
+      "epoch": 0.8507680745528333,
+      "grad_norm": 0.396870438466712,
+      "learning_rate": 1.1496763134032363e-05,
+      "loss": 0.9782,
+      "step": 4422
+    },
+    {
+      "epoch": 0.8509604689613708,
+      "grad_norm": 0.30087970612157156,
+      "learning_rate": 1.1467768142069546e-05,
+      "loss": 0.9694,
+      "step": 4423
+    },
+    {
+      "epoch": 0.8511528633699084,
+      "grad_norm": 0.3325945603099869,
+      "learning_rate": 1.1438807535005436e-05,
+      "loss": 0.9962,
+      "step": 4424
+    },
+    {
+      "epoch": 0.8513452577784458,
+      "grad_norm": 0.33261011360842196,
+      "learning_rate": 1.1409881324088011e-05,
+      "loss": 1.0458,
+      "step": 4425
+    },
+    {
+      "epoch": 0.8515376521869833,
+      "grad_norm": 0.3049445624066469,
+      "learning_rate": 1.138098952055181e-05,
+      "loss": 1.0112,
+      "step": 4426
+    },
+    {
+      "epoch": 0.8517300465955208,
+      "grad_norm": 0.38766045486114176,
+      "learning_rate": 1.1352132135618166e-05,
+      "loss": 1.0492,
+      "step": 4427
+    },
+    {
+      "epoch": 0.8519224410040583,
+      "grad_norm": 0.31099305666207355,
+      "learning_rate": 1.1323309180494901e-05,
+      "loss": 0.9956,
+      "step": 4428
+    },
+    {
+      "epoch": 0.8521148354125958,
+      "grad_norm": 0.42968829303413725,
+      "learning_rate": 1.1294520666376518e-05,
+      "loss": 1.1178,
+      "step": 4429
+    },
+    {
+      "epoch": 0.8523072298211333,
+      "grad_norm": 0.36525606478135414,
+      "learning_rate": 1.1265766604444172e-05,
+      "loss": 0.9432,
+      "step": 4430
+    },
+    {
+      "epoch": 0.8524996242296708,
+      "grad_norm": 0.33189102350562616,
+      "learning_rate": 1.1237047005865576e-05,
+      "loss": 1.1322,
+      "step": 4431
+    },
+    {
+      "epoch": 0.8526920186382083,
+      "grad_norm": 0.3414862663003093,
+      "learning_rate": 1.1208361881795116e-05,
+      "loss": 0.9851,
+      "step": 4432
+    },
+    {
+      "epoch": 0.8528844130467458,
+      "grad_norm": 0.28830398683799496,
+      "learning_rate": 1.1179711243373736e-05,
+      "loss": 0.9595,
+      "step": 4433
+    },
+    {
+      "epoch": 0.8530768074552834,
+      "grad_norm": 0.3474831865506262,
+      "learning_rate": 1.1151095101729047e-05,
+      "loss": 1.0068,
+      "step": 4434
+    },
+    {
+      "epoch": 0.8532692018638208,
+      "grad_norm": 0.3107078594784587,
+      "learning_rate": 1.1122513467975237e-05,
+      "loss": 0.9611,
+      "step": 4435
+    },
+    {
+      "epoch": 0.8534615962723583,
+      "grad_norm": 0.33472913486005623,
+      "learning_rate": 1.1093966353213037e-05,
+      "loss": 1.0413,
+      "step": 4436
+    },
+    {
+      "epoch": 0.8536539906808959,
+      "grad_norm": 0.3484026202151843,
+      "learning_rate": 1.1065453768529843e-05,
+      "loss": 1.0699,
+      "step": 4437
+    },
+    {
+      "epoch": 0.8538463850894333,
+      "grad_norm": 0.32119362521413797,
+      "learning_rate": 1.1036975724999609e-05,
+      "loss": 0.9327,
+      "step": 4438
+    },
+    {
+      "epoch": 0.8540387794979708,
+      "grad_norm": 0.28734280143011365,
+      "learning_rate": 1.1008532233682877e-05,
+      "loss": 1.101,
+      "step": 4439
+    },
+    {
+      "epoch": 0.8542311739065084,
+      "grad_norm": 0.3014044660527849,
+      "learning_rate": 1.098012330562681e-05,
+      "loss": 1.0343,
+      "step": 4440
+    },
+    {
+      "epoch": 0.8544235683150458,
+      "grad_norm": 0.29342925649486995,
+      "learning_rate": 1.0951748951865048e-05,
+      "loss": 1.0573,
+      "step": 4441
+    },
+    {
+      "epoch": 0.8546159627235833,
+      "grad_norm": 0.32888223382561116,
+      "learning_rate": 1.0923409183417887e-05,
+      "loss": 1.1137,
+      "step": 4442
+    },
+    {
+      "epoch": 0.8548083571321209,
+      "grad_norm": 0.3158956758836071,
+      "learning_rate": 1.08951040112922e-05,
+      "loss": 1.0781,
+      "step": 4443
+    },
+    {
+      "epoch": 0.8550007515406584,
+      "grad_norm": 0.36033958326935306,
+      "learning_rate": 1.0866833446481317e-05,
+      "loss": 1.0316,
+      "step": 4444
+    },
+    {
+      "epoch": 0.8551931459491958,
+      "grad_norm": 0.3338713449986007,
+      "learning_rate": 1.0838597499965275e-05,
+      "loss": 1.0277,
+      "step": 4445
+    },
+    {
+      "epoch": 0.8553855403577334,
+      "grad_norm": 0.33614920307354396,
+      "learning_rate": 1.0810396182710536e-05,
+      "loss": 0.9098,
+      "step": 4446
+    },
+    {
+      "epoch": 0.8555779347662709,
+      "grad_norm": 0.33525744216441417,
+      "learning_rate": 1.0782229505670193e-05,
+      "loss": 1.0694,
+      "step": 4447
+    },
+    {
+      "epoch": 0.8557703291748083,
+      "grad_norm": 0.3394241324598925,
+      "learning_rate": 1.0754097479783876e-05,
+      "loss": 1.0566,
+      "step": 4448
+    },
+    {
+      "epoch": 0.8559627235833459,
+      "grad_norm": 0.3018279365084254,
+      "learning_rate": 1.0726000115977696e-05,
+      "loss": 0.9997,
+      "step": 4449
+    },
+    {
+      "epoch": 0.8561551179918834,
+      "grad_norm": 0.40704980939486435,
+      "learning_rate": 1.069793742516435e-05,
+      "loss": 1.123,
+      "step": 4450
+    },
+    {
+      "epoch": 0.8563475124004208,
+      "grad_norm": 0.2919924860493612,
+      "learning_rate": 1.0669909418243118e-05,
+      "loss": 0.9917,
+      "step": 4451
+    },
+    {
+      "epoch": 0.8565399068089584,
+      "grad_norm": 0.2790355169328536,
+      "learning_rate": 1.0641916106099692e-05,
+      "loss": 1.0248,
+      "step": 4452
+    },
+    {
+      "epoch": 0.8567323012174959,
+      "grad_norm": 0.37835937772774725,
+      "learning_rate": 1.0613957499606386e-05,
+      "loss": 1.0291,
+      "step": 4453
+    },
+    {
+      "epoch": 0.8569246956260334,
+      "grad_norm": 0.42342398844313034,
+      "learning_rate": 1.0586033609622004e-05,
+      "loss": 1.0717,
+      "step": 4454
+    },
+    {
+      "epoch": 0.8571170900345708,
+      "grad_norm": 0.317702221107483,
+      "learning_rate": 1.0558144446991836e-05,
+      "loss": 0.974,
+      "step": 4455
+    },
+    {
+      "epoch": 0.8573094844431084,
+      "grad_norm": 0.373112834970984,
+      "learning_rate": 1.0530290022547728e-05,
+      "loss": 0.9997,
+      "step": 4456
+    },
+    {
+      "epoch": 0.8575018788516459,
+      "grad_norm": 0.3300060489963877,
+      "learning_rate": 1.0502470347108018e-05,
+      "loss": 1.0404,
+      "step": 4457
+    },
+    {
+      "epoch": 0.8576942732601833,
+      "grad_norm": 0.33122842325647844,
+      "learning_rate": 1.0474685431477539e-05,
+      "loss": 1.0045,
+      "step": 4458
+    },
+    {
+      "epoch": 0.8578866676687209,
+      "grad_norm": 0.3178185048306649,
+      "learning_rate": 1.0446935286447656e-05,
+      "loss": 0.992,
+      "step": 4459
+    },
+    {
+      "epoch": 0.8580790620772584,
+      "grad_norm": 0.31261366375069094,
+      "learning_rate": 1.0419219922796176e-05,
+      "loss": 1.0118,
+      "step": 4460
+    },
+    {
+      "epoch": 0.8582714564857958,
+      "grad_norm": 0.30546684428292226,
+      "learning_rate": 1.0391539351287439e-05,
+      "loss": 1.084,
+      "step": 4461
+    },
+    {
+      "epoch": 0.8584638508943334,
+      "grad_norm": 0.2832597379638999,
+      "learning_rate": 1.0363893582672247e-05,
+      "loss": 1.0793,
+      "step": 4462
+    },
+    {
+      "epoch": 0.8586562453028709,
+      "grad_norm": 0.3228207422481123,
+      "learning_rate": 1.0336282627687921e-05,
+      "loss": 1.0698,
+      "step": 4463
+    },
+    {
+      "epoch": 0.8588486397114083,
+      "grad_norm": 0.3454503461978924,
+      "learning_rate": 1.030870649705825e-05,
+      "loss": 1.0485,
+      "step": 4464
+    },
+    {
+      "epoch": 0.8590410341199459,
+      "grad_norm": 0.30329830668170643,
+      "learning_rate": 1.0281165201493438e-05,
+      "loss": 0.962,
+      "step": 4465
+    },
+    {
+      "epoch": 0.8592334285284834,
+      "grad_norm": 0.3168778069988205,
+      "learning_rate": 1.025365875169023e-05,
+      "loss": 1.0196,
+      "step": 4466
+    },
+    {
+      "epoch": 0.8594258229370209,
+      "grad_norm": 0.3360349260654396,
+      "learning_rate": 1.0226187158331823e-05,
+      "loss": 1.0133,
+      "step": 4467
+    },
+    {
+      "epoch": 0.8596182173455584,
+      "grad_norm": 0.2829251034812311,
+      "learning_rate": 1.0198750432087855e-05,
+      "loss": 1.098,
+      "step": 4468
+    },
+    {
+      "epoch": 0.8598106117540959,
+      "grad_norm": 0.33258295644506225,
+      "learning_rate": 1.0171348583614459e-05,
+      "loss": 0.9965,
+      "step": 4469
+    },
+    {
+      "epoch": 0.8600030061626334,
+      "grad_norm": 0.3873773548625855,
+      "learning_rate": 1.0143981623554156e-05,
+      "loss": 1.0556,
+      "step": 4470
+    },
+    {
+      "epoch": 0.8601954005711709,
+      "grad_norm": 0.3466738624280973,
+      "learning_rate": 1.0116649562535985e-05,
+      "loss": 0.998,
+      "step": 4471
+    },
+    {
+      "epoch": 0.8603877949797084,
+      "grad_norm": 0.3646238098517531,
+      "learning_rate": 1.0089352411175423e-05,
+      "loss": 1.0928,
+      "step": 4472
+    },
+    {
+      "epoch": 0.8605801893882459,
+      "grad_norm": 0.27685121866733553,
+      "learning_rate": 1.0062090180074291e-05,
+      "loss": 1.037,
+      "step": 4473
+    },
+    {
+      "epoch": 0.8607725837967835,
+      "grad_norm": 0.34181690377305196,
+      "learning_rate": 1.0034862879821027e-05,
+      "loss": 1.0747,
+      "step": 4474
+    },
+    {
+      "epoch": 0.8609649782053209,
+      "grad_norm": 0.3083060809958894,
+      "learning_rate": 1.000767052099033e-05,
+      "loss": 1.088,
+      "step": 4475
+    },
+    {
+      "epoch": 0.8611573726138584,
+      "grad_norm": 0.3537574821480526,
+      "learning_rate": 9.98051311414342e-06,
+      "loss": 1.0503,
+      "step": 4476
+    },
+    {
+      "epoch": 0.861349767022396,
+      "grad_norm": 0.34305103563259565,
+      "learning_rate": 9.953390669827945e-06,
+      "loss": 0.9898,
+      "step": 4477
+    },
+    {
+      "epoch": 0.8615421614309334,
+      "grad_norm": 0.3032818888889282,
+      "learning_rate": 9.926303198577913e-06,
+      "loss": 1.0187,
+      "step": 4478
+    },
+    {
+      "epoch": 0.8617345558394709,
+      "grad_norm": 0.29898397552424566,
+      "learning_rate": 9.899250710913765e-06,
+      "loss": 1.0202,
+      "step": 4479
+    },
+    {
+      "epoch": 0.8619269502480085,
+      "grad_norm": 0.3177122180422633,
+      "learning_rate": 9.872233217342463e-06,
+      "loss": 1.0814,
+      "step": 4480
+    },
+    {
+      "epoch": 0.8621193446565459,
+      "grad_norm": 0.2985889291179135,
+      "learning_rate": 9.845250728357215e-06,
+      "loss": 1.0034,
+      "step": 4481
+    },
+    {
+      "epoch": 0.8623117390650834,
+      "grad_norm": 0.36401806520978847,
+      "learning_rate": 9.818303254437722e-06,
+      "loss": 0.9344,
+      "step": 4482
+    },
+    {
+      "epoch": 0.8625041334736209,
+      "grad_norm": 0.2835813223927997,
+      "learning_rate": 9.791390806050115e-06,
+      "loss": 1.0655,
+      "step": 4483
+    },
+    {
+      "epoch": 0.8626965278821584,
+      "grad_norm": 0.30858071706827567,
+      "learning_rate": 9.764513393646812e-06,
+      "loss": 1.0191,
+      "step": 4484
+    },
+    {
+      "epoch": 0.8628889222906959,
+      "grad_norm": 0.3089258645673938,
+      "learning_rate": 9.737671027666728e-06,
+      "loss": 0.9848,
+      "step": 4485
+    },
+    {
+      "epoch": 0.8630813166992334,
+      "grad_norm": 0.36369330480420137,
+      "learning_rate": 9.710863718535135e-06,
+      "loss": 0.9762,
+      "step": 4486
+    },
+    {
+      "epoch": 0.863273711107771,
+      "grad_norm": 0.3027704579909306,
+      "learning_rate": 9.684091476663659e-06,
+      "loss": 0.9363,
+      "step": 4487
+    },
+    {
+      "epoch": 0.8634661055163084,
+      "grad_norm": 0.33084617800809374,
+      "learning_rate": 9.657354312450362e-06,
+      "loss": 1.007,
+      "step": 4488
+    },
+    {
+      "epoch": 0.8636584999248459,
+      "grad_norm": 0.3315757145446283,
+      "learning_rate": 9.630652236279625e-06,
+      "loss": 1.071,
+      "step": 4489
+    },
+    {
+      "epoch": 0.8638508943333835,
+      "grad_norm": 0.3172335488087789,
+      "learning_rate": 9.603985258522218e-06,
+      "loss": 1.0596,
+      "step": 4490
+    },
+    {
+      "epoch": 0.8640432887419209,
+      "grad_norm": 0.3501319764249724,
+      "learning_rate": 9.577353389535315e-06,
+      "loss": 1.0877,
+      "step": 4491
+    },
+    {
+      "epoch": 0.8642356831504584,
+      "grad_norm": 0.37297691990062776,
+      "learning_rate": 9.550756639662417e-06,
+      "loss": 1.0372,
+      "step": 4492
+    },
+    {
+      "epoch": 0.864428077558996,
+      "grad_norm": 0.38715522066816577,
+      "learning_rate": 9.524195019233407e-06,
+      "loss": 1.1032,
+      "step": 4493
+    },
+    {
+      "epoch": 0.8646204719675334,
+      "grad_norm": 0.349409455189122,
+      "learning_rate": 9.497668538564474e-06,
+      "loss": 1.0524,
+      "step": 4494
+    },
+    {
+      "epoch": 0.8648128663760709,
+      "grad_norm": 0.3200069302607972,
+      "learning_rate": 9.471177207958237e-06,
+      "loss": 0.9815,
+      "step": 4495
+    },
+    {
+      "epoch": 0.8650052607846085,
+      "grad_norm": 0.2894757497068607,
+      "learning_rate": 9.444721037703597e-06,
+      "loss": 1.0473,
+      "step": 4496
+    },
+    {
+      "epoch": 0.865197655193146,
+      "grad_norm": 0.25782538962289675,
+      "learning_rate": 9.418300038075845e-06,
+      "loss": 1.0448,
+      "step": 4497
+    },
+    {
+      "epoch": 0.8653900496016834,
+      "grad_norm": 0.25219221351477167,
+      "learning_rate": 9.391914219336606e-06,
+      "loss": 1.0396,
+      "step": 4498
+    },
+    {
+      "epoch": 0.865582444010221,
+      "grad_norm": 0.3577759140001364,
+      "learning_rate": 9.365563591733784e-06,
+      "loss": 1.0004,
+      "step": 4499
+    },
+    {
+      "epoch": 0.8657748384187585,
+      "grad_norm": 0.3580821477834573,
+      "learning_rate": 9.33924816550168e-06,
+      "loss": 1.0903,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8659672328272959,
+      "grad_norm": 0.42143140171027066,
+      "learning_rate": 9.31296795086093e-06,
+      "loss": 1.0454,
+      "step": 4501
+    },
+    {
+      "epoch": 0.8661596272358335,
+      "grad_norm": 0.33710189601644214,
+      "learning_rate": 9.28672295801839e-06,
+      "loss": 0.9482,
+      "step": 4502
+    },
+    {
+      "epoch": 0.866352021644371,
+      "grad_norm": 0.26576983296396556,
+      "learning_rate": 9.260513197167397e-06,
+      "loss": 1.0863,
+      "step": 4503
+    },
+    {
+      "epoch": 0.8665444160529084,
+      "grad_norm": 0.3011347453154511,
+      "learning_rate": 9.234338678487509e-06,
+      "loss": 1.0969,
+      "step": 4504
+    },
+    {
+      "epoch": 0.866736810461446,
+      "grad_norm": 0.29311712535303436,
+      "learning_rate": 9.208199412144558e-06,
+      "loss": 1.0271,
+      "step": 4505
+    },
+    {
+      "epoch": 0.8669292048699835,
+      "grad_norm": 0.3459572268262598,
+      "learning_rate": 9.182095408290781e-06,
+      "loss": 1.1073,
+      "step": 4506
+    },
+    {
+      "epoch": 0.867121599278521,
+      "grad_norm": 0.2678475190803086,
+      "learning_rate": 9.156026677064632e-06,
+      "loss": 0.9845,
+      "step": 4507
+    },
+    {
+      "epoch": 0.8673139936870584,
+      "grad_norm": 0.35735510750568455,
+      "learning_rate": 9.129993228590916e-06,
+      "loss": 1.1411,
+      "step": 4508
+    },
+    {
+      "epoch": 0.867506388095596,
+      "grad_norm": 0.4133557860551575,
+      "learning_rate": 9.103995072980765e-06,
+      "loss": 1.0795,
+      "step": 4509
+    },
+    {
+      "epoch": 0.8676987825041335,
+      "grad_norm": 0.2867012334098863,
+      "learning_rate": 9.078032220331522e-06,
+      "loss": 1.0124,
+      "step": 4510
+    },
+    {
+      "epoch": 0.8678911769126709,
+      "grad_norm": 0.2612459563249222,
+      "learning_rate": 9.052104680726858e-06,
+      "loss": 1.0226,
+      "step": 4511
+    },
+    {
+      "epoch": 0.8680835713212085,
+      "grad_norm": 0.2875637050035914,
+      "learning_rate": 9.026212464236772e-06,
+      "loss": 0.9402,
+      "step": 4512
+    },
+    {
+      "epoch": 0.868275965729746,
+      "grad_norm": 0.3676081864459413,
+      "learning_rate": 9.000355580917463e-06,
+      "loss": 0.9936,
+      "step": 4513
+    },
+    {
+      "epoch": 0.8684683601382834,
+      "grad_norm": 0.33022578201744995,
+      "learning_rate": 8.974534040811445e-06,
+      "loss": 0.9945,
+      "step": 4514
+    },
+    {
+      "epoch": 0.868660754546821,
+      "grad_norm": 0.2643701704320734,
+      "learning_rate": 8.948747853947526e-06,
+      "loss": 0.9493,
+      "step": 4515
+    },
+    {
+      "epoch": 0.8688531489553585,
+      "grad_norm": 0.3177045285295756,
+      "learning_rate": 8.922997030340752e-06,
+      "loss": 1.0616,
+      "step": 4516
+    },
+    {
+      "epoch": 0.869045543363896,
+      "grad_norm": 0.2909658218247964,
+      "learning_rate": 8.897281579992467e-06,
+      "loss": 1.0885,
+      "step": 4517
+    },
+    {
+      "epoch": 0.8692379377724335,
+      "grad_norm": 0.3174905933862179,
+      "learning_rate": 8.871601512890238e-06,
+      "loss": 1.0604,
+      "step": 4518
+    },
+    {
+      "epoch": 0.869430332180971,
+      "grad_norm": 0.3716458273941534,
+      "learning_rate": 8.845956839007896e-06,
+      "loss": 0.9767,
+      "step": 4519
+    },
+    {
+      "epoch": 0.8696227265895085,
+      "grad_norm": 0.3361585594569247,
+      "learning_rate": 8.820347568305542e-06,
+      "loss": 1.043,
+      "step": 4520
+    },
+    {
+      "epoch": 0.869815120998046,
+      "grad_norm": 0.3424274173679851,
+      "learning_rate": 8.794773710729542e-06,
+      "loss": 0.9721,
+      "step": 4521
+    },
+    {
+      "epoch": 0.8700075154065835,
+      "grad_norm": 0.2796008606602198,
+      "learning_rate": 8.769235276212495e-06,
+      "loss": 0.9811,
+      "step": 4522
+    },
+    {
+      "epoch": 0.870199909815121,
+      "grad_norm": 0.36144247503136384,
+      "learning_rate": 8.74373227467319e-06,
+      "loss": 1.0633,
+      "step": 4523
+    },
+    {
+      "epoch": 0.8703923042236585,
+      "grad_norm": 0.3732452968777644,
+      "learning_rate": 8.718264716016722e-06,
+      "loss": 1.0302,
+      "step": 4524
+    },
+    {
+      "epoch": 0.870584698632196,
+      "grad_norm": 0.2953599648420802,
+      "learning_rate": 8.692832610134427e-06,
+      "loss": 1.105,
+      "step": 4525
+    },
+    {
+      "epoch": 0.8707770930407335,
+      "grad_norm": 0.26832474254996325,
+      "learning_rate": 8.66743596690377e-06,
+      "loss": 1.0644,
+      "step": 4526
+    },
+    {
+      "epoch": 0.870969487449271,
+      "grad_norm": 0.3027354183839997,
+      "learning_rate": 8.642074796188592e-06,
+      "loss": 1.0099,
+      "step": 4527
+    },
+    {
+      "epoch": 0.8711618818578085,
+      "grad_norm": 0.27651364483762547,
+      "learning_rate": 8.61674910783884e-06,
+      "loss": 1.0008,
+      "step": 4528
+    },
+    {
+      "epoch": 0.871354276266346,
+      "grad_norm": 0.3557336831020569,
+      "learning_rate": 8.59145891169072e-06,
+      "loss": 0.9959,
+      "step": 4529
+    },
+    {
+      "epoch": 0.8715466706748836,
+      "grad_norm": 0.34495167014227857,
+      "learning_rate": 8.566204217566664e-06,
+      "loss": 1.0266,
+      "step": 4530
+    },
+    {
+      "epoch": 0.871739065083421,
+      "grad_norm": 0.29654934009419176,
+      "learning_rate": 8.540985035275272e-06,
+      "loss": 1.085,
+      "step": 4531
+    },
+    {
+      "epoch": 0.8719314594919585,
+      "grad_norm": 0.3617659512440195,
+      "learning_rate": 8.515801374611432e-06,
+      "loss": 1.1141,
+      "step": 4532
+    },
+    {
+      "epoch": 0.8721238539004961,
+      "grad_norm": 0.30700110714809614,
+      "learning_rate": 8.490653245356185e-06,
+      "loss": 0.9967,
+      "step": 4533
+    },
+    {
+      "epoch": 0.8723162483090335,
+      "grad_norm": 0.29819207582559937,
+      "learning_rate": 8.465540657276727e-06,
+      "loss": 1.0536,
+      "step": 4534
+    },
+    {
+      "epoch": 0.872508642717571,
+      "grad_norm": 0.3569705737680106,
+      "learning_rate": 8.44046362012656e-06,
+      "loss": 0.8938,
+      "step": 4535
+    },
+    {
+      "epoch": 0.8727010371261085,
+      "grad_norm": 0.3579119333163085,
+      "learning_rate": 8.415422143645245e-06,
+      "loss": 1.0691,
+      "step": 4536
+    },
+    {
+      "epoch": 0.872893431534646,
+      "grad_norm": 0.30332627861015554,
+      "learning_rate": 8.39041623755864e-06,
+      "loss": 0.9906,
+      "step": 4537
+    },
+    {
+      "epoch": 0.8730858259431835,
+      "grad_norm": 0.3289154092150236,
+      "learning_rate": 8.365445911578785e-06,
+      "loss": 0.9788,
+      "step": 4538
+    },
+    {
+      "epoch": 0.873278220351721,
+      "grad_norm": 0.31544088970501094,
+      "learning_rate": 8.340511175403809e-06,
+      "loss": 1.0725,
+      "step": 4539
+    },
+    {
+      "epoch": 0.8734706147602586,
+      "grad_norm": 0.3000155065610411,
+      "learning_rate": 8.315612038718101e-06,
+      "loss": 0.9786,
+      "step": 4540
+    },
+    {
+      "epoch": 0.873663009168796,
+      "grad_norm": 0.2997212359600715,
+      "learning_rate": 8.290748511192215e-06,
+      "loss": 1.0573,
+      "step": 4541
+    },
+    {
+      "epoch": 0.8738554035773335,
+      "grad_norm": 0.38680784507249677,
+      "learning_rate": 8.265920602482824e-06,
+      "loss": 1.06,
+      "step": 4542
+    },
+    {
+      "epoch": 0.8740477979858711,
+      "grad_norm": 0.3134152454131542,
+      "learning_rate": 8.241128322232817e-06,
+      "loss": 1.0345,
+      "step": 4543
+    },
+    {
+      "epoch": 0.8742401923944085,
+      "grad_norm": 0.30801958987495803,
+      "learning_rate": 8.216371680071244e-06,
+      "loss": 1.0268,
+      "step": 4544
+    },
+    {
+      "epoch": 0.874432586802946,
+      "grad_norm": 0.4017090241313671,
+      "learning_rate": 8.191650685613272e-06,
+      "loss": 1.1076,
+      "step": 4545
+    },
+    {
+      "epoch": 0.8746249812114836,
+      "grad_norm": 0.28283552919225213,
+      "learning_rate": 8.166965348460299e-06,
+      "loss": 0.992,
+      "step": 4546
+    },
+    {
+      "epoch": 0.874817375620021,
+      "grad_norm": 0.30768950704447573,
+      "learning_rate": 8.142315678199763e-06,
+      "loss": 1.1308,
+      "step": 4547
+    },
+    {
+      "epoch": 0.8750097700285585,
+      "grad_norm": 0.29105323878644324,
+      "learning_rate": 8.117701684405343e-06,
+      "loss": 1.1231,
+      "step": 4548
+    },
+    {
+      "epoch": 0.8752021644370961,
+      "grad_norm": 0.3444143380118533,
+      "learning_rate": 8.093123376636836e-06,
+      "loss": 1.0038,
+      "step": 4549
+    },
+    {
+      "epoch": 0.8753945588456336,
+      "grad_norm": 0.37826490436062077,
+      "learning_rate": 8.06858076444017e-06,
+      "loss": 1.0436,
+      "step": 4550
+    },
+    {
+      "epoch": 0.875586953254171,
+      "grad_norm": 0.31686602377310846,
+      "learning_rate": 8.044073857347423e-06,
+      "loss": 1.0862,
+      "step": 4551
+    },
+    {
+      "epoch": 0.8757793476627086,
+      "grad_norm": 0.3167485328632148,
+      "learning_rate": 8.019602664876758e-06,
+      "loss": 0.9895,
+      "step": 4552
+    },
+    {
+      "epoch": 0.8759717420712461,
+      "grad_norm": 0.3726995070754788,
+      "learning_rate": 7.995167196532527e-06,
+      "loss": 0.9494,
+      "step": 4553
+    },
+    {
+      "epoch": 0.8761641364797835,
+      "grad_norm": 0.2971354953411799,
+      "learning_rate": 7.970767461805217e-06,
+      "loss": 0.9925,
+      "step": 4554
+    },
+    {
+      "epoch": 0.8763565308883211,
+      "grad_norm": 0.2672533224321114,
+      "learning_rate": 7.946403470171327e-06,
+      "loss": 1.0142,
+      "step": 4555
+    },
+    {
+      "epoch": 0.8765489252968586,
+      "grad_norm": 0.30759314843219776,
+      "learning_rate": 7.922075231093629e-06,
+      "loss": 1.1331,
+      "step": 4556
+    },
+    {
+      "epoch": 0.876741319705396,
+      "grad_norm": 0.26725015363847765,
+      "learning_rate": 7.897782754020888e-06,
+      "loss": 1.0258,
+      "step": 4557
+    },
+    {
+      "epoch": 0.8769337141139336,
+      "grad_norm": 0.3614669287403115,
+      "learning_rate": 7.873526048388026e-06,
+      "loss": 1.0257,
+      "step": 4558
+    },
+    {
+      "epoch": 0.8771261085224711,
+      "grad_norm": 0.2792742565674579,
+      "learning_rate": 7.849305123616092e-06,
+      "loss": 0.9808,
+      "step": 4559
+    },
+    {
+      "epoch": 0.8773185029310085,
+      "grad_norm": 0.34917237286106484,
+      "learning_rate": 7.825119989112173e-06,
+      "loss": 1.1074,
+      "step": 4560
+    },
+    {
+      "epoch": 0.877510897339546,
+      "grad_norm": 0.36783188142796946,
+      "learning_rate": 7.800970654269512e-06,
+      "loss": 1.0171,
+      "step": 4561
+    },
+    {
+      "epoch": 0.8777032917480836,
+      "grad_norm": 0.3004651117437256,
+      "learning_rate": 7.776857128467463e-06,
+      "loss": 1.0281,
+      "step": 4562
+    },
+    {
+      "epoch": 0.8778956861566211,
+      "grad_norm": 0.3177625927176387,
+      "learning_rate": 7.75277942107141e-06,
+      "loss": 1.0023,
+      "step": 4563
+    },
+    {
+      "epoch": 0.8780880805651585,
+      "grad_norm": 0.32300275833253295,
+      "learning_rate": 7.728737541432863e-06,
+      "loss": 1.0401,
+      "step": 4564
+    },
+    {
+      "epoch": 0.8782804749736961,
+      "grad_norm": 0.3425531359765448,
+      "learning_rate": 7.704731498889428e-06,
+      "loss": 1.0772,
+      "step": 4565
+    },
+    {
+      "epoch": 0.8784728693822336,
+      "grad_norm": 0.29130708313697373,
+      "learning_rate": 7.680761302764727e-06,
+      "loss": 0.9022,
+      "step": 4566
+    },
+    {
+      "epoch": 0.878665263790771,
+      "grad_norm": 0.2933954423000946,
+      "learning_rate": 7.656826962368579e-06,
+      "loss": 1.0647,
+      "step": 4567
+    },
+    {
+      "epoch": 0.8788576581993086,
+      "grad_norm": 0.3998976598753328,
+      "learning_rate": 7.632928486996749e-06,
+      "loss": 1.0291,
+      "step": 4568
+    },
+    {
+      "epoch": 0.8790500526078461,
+      "grad_norm": 0.29213728572412817,
+      "learning_rate": 7.609065885931155e-06,
+      "loss": 1.0231,
+      "step": 4569
+    },
+    {
+      "epoch": 0.8792424470163835,
+      "grad_norm": 0.298579230197274,
+      "learning_rate": 7.585239168439761e-06,
+      "loss": 1.0777,
+      "step": 4570
+    },
+    {
+      "epoch": 0.8794348414249211,
+      "grad_norm": 0.3758451651902908,
+      "learning_rate": 7.561448343776567e-06,
+      "loss": 1.1126,
+      "step": 4571
+    },
+    {
+      "epoch": 0.8796272358334586,
+      "grad_norm": 0.3424006288082153,
+      "learning_rate": 7.537693421181658e-06,
+      "loss": 1.0157,
+      "step": 4572
+    },
+    {
+      "epoch": 0.879819630241996,
+      "grad_norm": 0.33944984462898,
+      "learning_rate": 7.513974409881186e-06,
+      "loss": 1.0515,
+      "step": 4573
+    },
+    {
+      "epoch": 0.8800120246505336,
+      "grad_norm": 0.2985959864275465,
+      "learning_rate": 7.490291319087339e-06,
+      "loss": 1.0288,
+      "step": 4574
+    },
+    {
+      "epoch": 0.8802044190590711,
+      "grad_norm": 0.41024714339703505,
+      "learning_rate": 7.46664415799837e-06,
+      "loss": 1.0506,
+      "step": 4575
+    },
+    {
+      "epoch": 0.8803968134676086,
+      "grad_norm": 0.36109715019961514,
+      "learning_rate": 7.443032935798533e-06,
+      "loss": 1.016,
+      "step": 4576
+    },
+    {
+      "epoch": 0.8805892078761461,
+      "grad_norm": 0.374191845666455,
+      "learning_rate": 7.419457661658169e-06,
+      "loss": 0.9572,
+      "step": 4577
+    },
+    {
+      "epoch": 0.8807816022846836,
+      "grad_norm": 0.32999386144732745,
+      "learning_rate": 7.3959183447336434e-06,
+      "loss": 1.0133,
+      "step": 4578
+    },
+    {
+      "epoch": 0.8809739966932211,
+      "grad_norm": 0.32854606115096013,
+      "learning_rate": 7.372414994167354e-06,
+      "loss": 1.0821,
+      "step": 4579
+    },
+    {
+      "epoch": 0.8811663911017587,
+      "grad_norm": 0.2847110915474937,
+      "learning_rate": 7.3489476190877536e-06,
+      "loss": 1.0574,
+      "step": 4580
+    },
+    {
+      "epoch": 0.8813587855102961,
+      "grad_norm": 0.27303589687485225,
+      "learning_rate": 7.325516228609264e-06,
+      "loss": 1.0641,
+      "step": 4581
+    },
+    {
+      "epoch": 0.8815511799188336,
+      "grad_norm": 0.33401077357174613,
+      "learning_rate": 7.302120831832382e-06,
+      "loss": 1.0438,
+      "step": 4582
+    },
+    {
+      "epoch": 0.8817435743273712,
+      "grad_norm": 0.3630441409035198,
+      "learning_rate": 7.278761437843629e-06,
+      "loss": 1.0187,
+      "step": 4583
+    },
+    {
+      "epoch": 0.8819359687359086,
+      "grad_norm": 0.30384362015195976,
+      "learning_rate": 7.255438055715469e-06,
+      "loss": 1.0756,
+      "step": 4584
+    },
+    {
+      "epoch": 0.8821283631444461,
+      "grad_norm": 0.31486811664333225,
+      "learning_rate": 7.232150694506512e-06,
+      "loss": 0.9881,
+      "step": 4585
+    },
+    {
+      "epoch": 0.8823207575529837,
+      "grad_norm": 0.36614796103834135,
+      "learning_rate": 7.208899363261234e-06,
+      "loss": 0.9636,
+      "step": 4586
+    },
+    {
+      "epoch": 0.8825131519615211,
+      "grad_norm": 0.29384081111177496,
+      "learning_rate": 7.185684071010224e-06,
+      "loss": 1.0296,
+      "step": 4587
+    },
+    {
+      "epoch": 0.8827055463700586,
+      "grad_norm": 0.3427269356235219,
+      "learning_rate": 7.162504826770033e-06,
+      "loss": 1.0921,
+      "step": 4588
+    },
+    {
+      "epoch": 0.8828979407785961,
+      "grad_norm": 0.29503300125278814,
+      "learning_rate": 7.139361639543185e-06,
+      "loss": 1.0112,
+      "step": 4589
+    },
+    {
+      "epoch": 0.8830903351871336,
+      "grad_norm": 0.3988848634130886,
+      "learning_rate": 7.116254518318221e-06,
+      "loss": 0.8527,
+      "step": 4590
+    },
+    {
+      "epoch": 0.8832827295956711,
+      "grad_norm": 0.3362618641133164,
+      "learning_rate": 7.0931834720697535e-06,
+      "loss": 0.9526,
+      "step": 4591
+    },
+    {
+      "epoch": 0.8834751240042086,
+      "grad_norm": 0.3171940337183918,
+      "learning_rate": 7.070148509758223e-06,
+      "loss": 1.0747,
+      "step": 4592
+    },
+    {
+      "epoch": 0.8836675184127462,
+      "grad_norm": 0.31391739649335915,
+      "learning_rate": 7.0471496403301975e-06,
+      "loss": 1.0683,
+      "step": 4593
+    },
+    {
+      "epoch": 0.8838599128212836,
+      "grad_norm": 0.32567208559468985,
+      "learning_rate": 7.024186872718164e-06,
+      "loss": 1.1047,
+      "step": 4594
+    },
+    {
+      "epoch": 0.8840523072298211,
+      "grad_norm": 0.288861639589572,
+      "learning_rate": 7.001260215840566e-06,
+      "loss": 1.098,
+      "step": 4595
+    },
+    {
+      "epoch": 0.8842447016383587,
+      "grad_norm": 0.3427373607584587,
+      "learning_rate": 6.978369678601893e-06,
+      "loss": 1.0368,
+      "step": 4596
+    },
+    {
+      "epoch": 0.8844370960468961,
+      "grad_norm": 0.34831122126737557,
+      "learning_rate": 6.955515269892532e-06,
+      "loss": 1.0635,
+      "step": 4597
+    },
+    {
+      "epoch": 0.8846294904554336,
+      "grad_norm": 0.30863953906433567,
+      "learning_rate": 6.932696998588895e-06,
+      "loss": 1.0007,
+      "step": 4598
+    },
+    {
+      "epoch": 0.8848218848639712,
+      "grad_norm": 0.382562898683625,
+      "learning_rate": 6.909914873553347e-06,
+      "loss": 1.0074,
+      "step": 4599
+    },
+    {
+      "epoch": 0.8850142792725086,
+      "grad_norm": 0.3511536472059468,
+      "learning_rate": 6.887168903634178e-06,
+      "loss": 1.1182,
+      "step": 4600
+    },
+    {
+      "epoch": 0.8852066736810461,
+      "grad_norm": 0.3796847833354549,
+      "learning_rate": 6.864459097665654e-06,
+      "loss": 1.059,
+      "step": 4601
+    },
+    {
+      "epoch": 0.8853990680895837,
+      "grad_norm": 0.3153127901217798,
+      "learning_rate": 6.8417854644680205e-06,
+      "loss": 1.0374,
+      "step": 4602
+    },
+    {
+      "epoch": 0.8855914624981212,
+      "grad_norm": 0.2685441588513206,
+      "learning_rate": 6.819148012847454e-06,
+      "loss": 1.0042,
+      "step": 4603
+    },
+    {
+      "epoch": 0.8857838569066586,
+      "grad_norm": 0.3201018442721621,
+      "learning_rate": 6.796546751596089e-06,
+      "loss": 1.1054,
+      "step": 4604
+    },
+    {
+      "epoch": 0.8859762513151962,
+      "grad_norm": 0.3488306027320683,
+      "learning_rate": 6.773981689491959e-06,
+      "loss": 1.0534,
+      "step": 4605
+    },
+    {
+      "epoch": 0.8861686457237337,
+      "grad_norm": 0.3036267090649368,
+      "learning_rate": 6.75145283529911e-06,
+      "loss": 0.9497,
+      "step": 4606
+    },
+    {
+      "epoch": 0.8863610401322711,
+      "grad_norm": 0.44302319637642057,
+      "learning_rate": 6.728960197767475e-06,
+      "loss": 1.0181,
+      "step": 4607
+    },
+    {
+      "epoch": 0.8865534345408087,
+      "grad_norm": 0.4058703377346576,
+      "learning_rate": 6.706503785632934e-06,
+      "loss": 1.0813,
+      "step": 4608
+    },
+    {
+      "epoch": 0.8867458289493462,
+      "grad_norm": 0.33406822452214446,
+      "learning_rate": 6.68408360761732e-06,
+      "loss": 1.0352,
+      "step": 4609
+    },
+    {
+      "epoch": 0.8869382233578836,
+      "grad_norm": 0.27904225537842964,
+      "learning_rate": 6.661699672428334e-06,
+      "loss": 1.0031,
+      "step": 4610
+    },
+    {
+      "epoch": 0.8871306177664212,
+      "grad_norm": 0.41040572374800766,
+      "learning_rate": 6.639351988759657e-06,
+      "loss": 1.0692,
+      "step": 4611
+    },
+    {
+      "epoch": 0.8873230121749587,
+      "grad_norm": 0.32900727877422314,
+      "learning_rate": 6.6170405652909e-06,
+      "loss": 0.9534,
+      "step": 4612
+    },
+    {
+      "epoch": 0.8875154065834961,
+      "grad_norm": 0.3737738495513355,
+      "learning_rate": 6.594765410687487e-06,
+      "loss": 1.0797,
+      "step": 4613
+    },
+    {
+      "epoch": 0.8877078009920337,
+      "grad_norm": 0.30707814793928245,
+      "learning_rate": 6.57252653360092e-06,
+      "loss": 0.9994,
+      "step": 4614
+    },
+    {
+      "epoch": 0.8879001954005712,
+      "grad_norm": 0.33435425358324994,
+      "learning_rate": 6.550323942668468e-06,
+      "loss": 0.9721,
+      "step": 4615
+    },
+    {
+      "epoch": 0.8880925898091087,
+      "grad_norm": 0.3251041998289954,
+      "learning_rate": 6.528157646513378e-06,
+      "loss": 1.1378,
+      "step": 4616
+    },
+    {
+      "epoch": 0.8882849842176461,
+      "grad_norm": 0.3913190590863455,
+      "learning_rate": 6.506027653744795e-06,
+      "loss": 1.0791,
+      "step": 4617
+    },
+    {
+      "epoch": 0.8884773786261837,
+      "grad_norm": 0.32990941547926167,
+      "learning_rate": 6.483933972957734e-06,
+      "loss": 1.1513,
+      "step": 4618
+    },
+    {
+      "epoch": 0.8886697730347212,
+      "grad_norm": 0.4322966205047125,
+      "learning_rate": 6.461876612733109e-06,
+      "loss": 1.0151,
+      "step": 4619
+    },
+    {
+      "epoch": 0.8888621674432586,
+      "grad_norm": 0.2921533053054054,
+      "learning_rate": 6.439855581637799e-06,
+      "loss": 1.0156,
+      "step": 4620
+    },
+    {
+      "epoch": 0.8890545618517962,
+      "grad_norm": 0.33251370626902865,
+      "learning_rate": 6.417870888224475e-06,
+      "loss": 1.0399,
+      "step": 4621
+    },
+    {
+      "epoch": 0.8892469562603337,
+      "grad_norm": 0.2952245531857302,
+      "learning_rate": 6.39592254103174e-06,
+      "loss": 0.9662,
+      "step": 4622
+    },
+    {
+      "epoch": 0.8894393506688711,
+      "grad_norm": 0.32120802741526594,
+      "learning_rate": 6.37401054858412e-06,
+      "loss": 1.0353,
+      "step": 4623
+    },
+    {
+      "epoch": 0.8896317450774087,
+      "grad_norm": 0.3218940622511122,
+      "learning_rate": 6.352134919391928e-06,
+      "loss": 1.0695,
+      "step": 4624
+    },
+    {
+      "epoch": 0.8898241394859462,
+      "grad_norm": 0.31885607353444745,
+      "learning_rate": 6.330295661951435e-06,
+      "loss": 1.0587,
+      "step": 4625
+    },
+    {
+      "epoch": 0.8900165338944837,
+      "grad_norm": 0.34241441708118187,
+      "learning_rate": 6.308492784744746e-06,
+      "loss": 0.9965,
+      "step": 4626
+    },
+    {
+      "epoch": 0.8902089283030212,
+      "grad_norm": 0.31364800543367083,
+      "learning_rate": 6.2867262962398535e-06,
+      "loss": 1.0373,
+      "step": 4627
+    },
+    {
+      "epoch": 0.8904013227115587,
+      "grad_norm": 0.34227924847973723,
+      "learning_rate": 6.2649962048906276e-06,
+      "loss": 0.9645,
+      "step": 4628
+    },
+    {
+      "epoch": 0.8905937171200962,
+      "grad_norm": 0.3110099557536748,
+      "learning_rate": 6.24330251913674e-06,
+      "loss": 0.9479,
+      "step": 4629
+    },
+    {
+      "epoch": 0.8907861115286337,
+      "grad_norm": 0.36315800752798055,
+      "learning_rate": 6.221645247403807e-06,
+      "loss": 0.9701,
+      "step": 4630
+    },
+    {
+      "epoch": 0.8909785059371712,
+      "grad_norm": 0.36357466131651023,
+      "learning_rate": 6.200024398103255e-06,
+      "loss": 1.0991,
+      "step": 4631
+    },
+    {
+      "epoch": 0.8911709003457087,
+      "grad_norm": 0.3419904365021918,
+      "learning_rate": 6.178439979632366e-06,
+      "loss": 0.9602,
+      "step": 4632
+    },
+    {
+      "epoch": 0.8913632947542462,
+      "grad_norm": 0.29102130463619397,
+      "learning_rate": 6.156892000374292e-06,
+      "loss": 1.0854,
+      "step": 4633
+    },
+    {
+      "epoch": 0.8915556891627837,
+      "grad_norm": 0.3310144371550895,
+      "learning_rate": 6.135380468698005e-06,
+      "loss": 1.0653,
+      "step": 4634
+    },
+    {
+      "epoch": 0.8917480835713212,
+      "grad_norm": 0.2821726495862142,
+      "learning_rate": 6.113905392958341e-06,
+      "loss": 1.0389,
+      "step": 4635
+    },
+    {
+      "epoch": 0.8919404779798588,
+      "grad_norm": 0.38318683412149257,
+      "learning_rate": 6.092466781495976e-06,
+      "loss": 1.059,
+      "step": 4636
+    },
+    {
+      "epoch": 0.8921328723883962,
+      "grad_norm": 0.32372680121821257,
+      "learning_rate": 6.071064642637403e-06,
+      "loss": 0.9398,
+      "step": 4637
+    },
+    {
+      "epoch": 0.8923252667969337,
+      "grad_norm": 0.31883471511113454,
+      "learning_rate": 6.049698984695007e-06,
+      "loss": 1.0942,
+      "step": 4638
+    },
+    {
+      "epoch": 0.8925176612054713,
+      "grad_norm": 0.31037935158238317,
+      "learning_rate": 6.028369815966917e-06,
+      "loss": 1.0672,
+      "step": 4639
+    },
+    {
+      "epoch": 0.8927100556140087,
+      "grad_norm": 0.4310545184398492,
+      "learning_rate": 6.007077144737161e-06,
+      "loss": 1.0144,
+      "step": 4640
+    },
+    {
+      "epoch": 0.8929024500225462,
+      "grad_norm": 0.3275514347235563,
+      "learning_rate": 5.985820979275569e-06,
+      "loss": 1.0361,
+      "step": 4641
+    },
+    {
+      "epoch": 0.8930948444310837,
+      "grad_norm": 0.3120789409608759,
+      "learning_rate": 5.96460132783776e-06,
+      "loss": 1.0356,
+      "step": 4642
+    },
+    {
+      "epoch": 0.8932872388396212,
+      "grad_norm": 0.2993952674923048,
+      "learning_rate": 5.943418198665252e-06,
+      "loss": 1.0275,
+      "step": 4643
+    },
+    {
+      "epoch": 0.8934796332481587,
+      "grad_norm": 0.35402351847110725,
+      "learning_rate": 5.922271599985285e-06,
+      "loss": 1.0626,
+      "step": 4644
+    },
+    {
+      "epoch": 0.8936720276566962,
+      "grad_norm": 0.38675466255553903,
+      "learning_rate": 5.901161540010968e-06,
+      "loss": 0.9366,
+      "step": 4645
+    },
+    {
+      "epoch": 0.8938644220652338,
+      "grad_norm": 0.35253814137892703,
+      "learning_rate": 5.8800880269412326e-06,
+      "loss": 1.0693,
+      "step": 4646
+    },
+    {
+      "epoch": 0.8940568164737712,
+      "grad_norm": 0.34140731566655314,
+      "learning_rate": 5.8590510689607415e-06,
+      "loss": 0.9287,
+      "step": 4647
+    },
+    {
+      "epoch": 0.8942492108823087,
+      "grad_norm": 0.3205469940104918,
+      "learning_rate": 5.838050674240025e-06,
+      "loss": 1.053,
+      "step": 4648
+    },
+    {
+      "epoch": 0.8944416052908463,
+      "grad_norm": 0.32893328930589,
+      "learning_rate": 5.817086850935416e-06,
+      "loss": 0.9709,
+      "step": 4649
+    },
+    {
+      "epoch": 0.8946339996993837,
+      "grad_norm": 0.2938137267581275,
+      "learning_rate": 5.796159607189e-06,
+      "loss": 1.0779,
+      "step": 4650
+    },
+    {
+      "epoch": 0.8948263941079212,
+      "grad_norm": 0.3436050832639803,
+      "learning_rate": 5.775268951128676e-06,
+      "loss": 1.0848,
+      "step": 4651
+    },
+    {
+      "epoch": 0.8950187885164588,
+      "grad_norm": 0.328391758232989,
+      "learning_rate": 5.754414890868154e-06,
+      "loss": 1.0354,
+      "step": 4652
+    },
+    {
+      "epoch": 0.8952111829249962,
+      "grad_norm": 0.28888688147049585,
+      "learning_rate": 5.733597434506877e-06,
+      "loss": 1.0025,
+      "step": 4653
+    },
+    {
+      "epoch": 0.8954035773335337,
+      "grad_norm": 0.28007441025842733,
+      "learning_rate": 5.712816590130132e-06,
+      "loss": 1.0741,
+      "step": 4654
+    },
+    {
+      "epoch": 0.8955959717420713,
+      "grad_norm": 0.26421831273018,
+      "learning_rate": 5.6920723658089534e-06,
+      "loss": 0.9876,
+      "step": 4655
+    },
+    {
+      "epoch": 0.8957883661506088,
+      "grad_norm": 0.27173567424146877,
+      "learning_rate": 5.671364769600163e-06,
+      "loss": 1.0714,
+      "step": 4656
+    },
+    {
+      "epoch": 0.8959807605591462,
+      "grad_norm": 0.26447474562109996,
+      "learning_rate": 5.650693809546348e-06,
+      "loss": 1.0254,
+      "step": 4657
+    },
+    {
+      "epoch": 0.8961731549676838,
+      "grad_norm": 0.2975042806371661,
+      "learning_rate": 5.630059493675865e-06,
+      "loss": 1.0022,
+      "step": 4658
+    },
+    {
+      "epoch": 0.8963655493762213,
+      "grad_norm": 0.48556404937171793,
+      "learning_rate": 5.609461830002849e-06,
+      "loss": 1.0284,
+      "step": 4659
+    },
+    {
+      "epoch": 0.8965579437847587,
+      "grad_norm": 0.286139606108514,
+      "learning_rate": 5.5889008265271995e-06,
+      "loss": 1.0405,
+      "step": 4660
+    },
+    {
+      "epoch": 0.8967503381932963,
+      "grad_norm": 0.3164533873164714,
+      "learning_rate": 5.5683764912345615e-06,
+      "loss": 1.1313,
+      "step": 4661
+    },
+    {
+      "epoch": 0.8969427326018338,
+      "grad_norm": 0.282476107216376,
+      "learning_rate": 5.547888832096382e-06,
+      "loss": 1.042,
+      "step": 4662
+    },
+    {
+      "epoch": 0.8971351270103712,
+      "grad_norm": 0.307869337456247,
+      "learning_rate": 5.527437857069784e-06,
+      "loss": 1.085,
+      "step": 4663
+    },
+    {
+      "epoch": 0.8973275214189088,
+      "grad_norm": 0.3145844460023375,
+      "learning_rate": 5.507023574097725e-06,
+      "loss": 1.1131,
+      "step": 4664
+    },
+    {
+      "epoch": 0.8975199158274463,
+      "grad_norm": 0.4021742817787599,
+      "learning_rate": 5.486645991108874e-06,
+      "loss": 1.0051,
+      "step": 4665
+    },
+    {
+      "epoch": 0.8977123102359837,
+      "grad_norm": 0.3137743874729658,
+      "learning_rate": 5.4663051160176225e-06,
+      "loss": 1.035,
+      "step": 4666
+    },
+    {
+      "epoch": 0.8979047046445213,
+      "grad_norm": 0.28991803622695916,
+      "learning_rate": 5.446000956724173e-06,
+      "loss": 0.9653,
+      "step": 4667
+    },
+    {
+      "epoch": 0.8980970990530588,
+      "grad_norm": 0.28534420867723764,
+      "learning_rate": 5.4257335211143955e-06,
+      "loss": 0.9961,
+      "step": 4668
+    },
+    {
+      "epoch": 0.8982894934615963,
+      "grad_norm": 0.32819704564998353,
+      "learning_rate": 5.4055028170599374e-06,
+      "loss": 1.0489,
+      "step": 4669
+    },
+    {
+      "epoch": 0.8984818878701337,
+      "grad_norm": 0.32864091963419684,
+      "learning_rate": 5.3853088524181914e-06,
+      "loss": 1.0522,
+      "step": 4670
+    },
+    {
+      "epoch": 0.8986742822786713,
+      "grad_norm": 0.3361236809262748,
+      "learning_rate": 5.365151635032218e-06,
+      "loss": 1.0878,
+      "step": 4671
+    },
+    {
+      "epoch": 0.8988666766872088,
+      "grad_norm": 0.32311371417787205,
+      "learning_rate": 5.345031172730875e-06,
+      "loss": 1.028,
+      "step": 4672
+    },
+    {
+      "epoch": 0.8990590710957462,
+      "grad_norm": 0.29418311661565816,
+      "learning_rate": 5.3249474733287345e-06,
+      "loss": 0.9015,
+      "step": 4673
+    },
+    {
+      "epoch": 0.8992514655042838,
+      "grad_norm": 0.344500548694273,
+      "learning_rate": 5.3049005446260455e-06,
+      "loss": 1.0007,
+      "step": 4674
+    },
+    {
+      "epoch": 0.8994438599128213,
+      "grad_norm": 0.36044099566013865,
+      "learning_rate": 5.284890394408826e-06,
+      "loss": 0.9919,
+      "step": 4675
+    },
+    {
+      "epoch": 0.8996362543213587,
+      "grad_norm": 0.31737076944270715,
+      "learning_rate": 5.2649170304487576e-06,
+      "loss": 1.0594,
+      "step": 4676
+    },
+    {
+      "epoch": 0.8998286487298963,
+      "grad_norm": 0.299635053084897,
+      "learning_rate": 5.2449804605032685e-06,
+      "loss": 1.0815,
+      "step": 4677
+    },
+    {
+      "epoch": 0.9000210431384338,
+      "grad_norm": 0.28145817307932297,
+      "learning_rate": 5.225080692315532e-06,
+      "loss": 1.0752,
+      "step": 4678
+    },
+    {
+      "epoch": 0.9002134375469713,
+      "grad_norm": 0.4469966026197235,
+      "learning_rate": 5.205217733614353e-06,
+      "loss": 1.042,
+      "step": 4679
+    },
+    {
+      "epoch": 0.9004058319555088,
+      "grad_norm": 0.30239234010479293,
+      "learning_rate": 5.185391592114286e-06,
+      "loss": 1.0656,
+      "step": 4680
+    },
+    {
+      "epoch": 0.9005982263640463,
+      "grad_norm": 0.35043673045070767,
+      "learning_rate": 5.165602275515591e-06,
+      "loss": 1.0589,
+      "step": 4681
+    },
+    {
+      "epoch": 0.9007906207725838,
+      "grad_norm": 0.36906667940377264,
+      "learning_rate": 5.145849791504187e-06,
+      "loss": 1.0616,
+      "step": 4682
+    },
+    {
+      "epoch": 0.9009830151811213,
+      "grad_norm": 0.31964155258238713,
+      "learning_rate": 5.126134147751716e-06,
+      "loss": 1.0624,
+      "step": 4683
+    },
+    {
+      "epoch": 0.9011754095896588,
+      "grad_norm": 0.30297233254917755,
+      "learning_rate": 5.106455351915507e-06,
+      "loss": 1.0991,
+      "step": 4684
+    },
+    {
+      "epoch": 0.9013678039981963,
+      "grad_norm": 0.25680122431409214,
+      "learning_rate": 5.0868134116385805e-06,
+      "loss": 1.0324,
+      "step": 4685
+    },
+    {
+      "epoch": 0.9015601984067338,
+      "grad_norm": 0.3376287318935145,
+      "learning_rate": 5.067208334549656e-06,
+      "loss": 1.1278,
+      "step": 4686
+    },
+    {
+      "epoch": 0.9017525928152713,
+      "grad_norm": 0.3834491733753379,
+      "learning_rate": 5.047640128263087e-06,
+      "loss": 1.0904,
+      "step": 4687
+    },
+    {
+      "epoch": 0.9019449872238088,
+      "grad_norm": 0.33098987845620953,
+      "learning_rate": 5.028108800378961e-06,
+      "loss": 1.0127,
+      "step": 4688
+    },
+    {
+      "epoch": 0.9021373816323464,
+      "grad_norm": 0.3162574053758955,
+      "learning_rate": 5.008614358483022e-06,
+      "loss": 1.1005,
+      "step": 4689
+    },
+    {
+      "epoch": 0.9023297760408838,
+      "grad_norm": 0.3370414185244734,
+      "learning_rate": 4.9891568101466666e-06,
+      "loss": 0.9485,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9025221704494213,
+      "grad_norm": 0.2859260081361284,
+      "learning_rate": 4.969736162927019e-06,
+      "loss": 0.9994,
+      "step": 4691
+    },
+    {
+      "epoch": 0.9027145648579589,
+      "grad_norm": 0.31877757567776244,
+      "learning_rate": 4.9503524243667994e-06,
+      "loss": 1.0455,
+      "step": 4692
+    },
+    {
+      "epoch": 0.9029069592664963,
+      "grad_norm": 0.37209508640453964,
+      "learning_rate": 4.931005601994432e-06,
+      "loss": 0.9753,
+      "step": 4693
+    },
+    {
+      "epoch": 0.9030993536750338,
+      "grad_norm": 0.344925861162728,
+      "learning_rate": 4.911695703324038e-06,
+      "loss": 1.1069,
+      "step": 4694
+    },
+    {
+      "epoch": 0.9032917480835713,
+      "grad_norm": 0.3407600719310235,
+      "learning_rate": 4.892422735855285e-06,
+      "loss": 1.0365,
+      "step": 4695
+    },
+    {
+      "epoch": 0.9034841424921088,
+      "grad_norm": 0.3288644993736094,
+      "learning_rate": 4.873186707073663e-06,
+      "loss": 1.0268,
+      "step": 4696
+    },
+    {
+      "epoch": 0.9036765369006463,
+      "grad_norm": 0.4030587255941124,
+      "learning_rate": 4.853987624450151e-06,
+      "loss": 0.9906,
+      "step": 4697
+    },
+    {
+      "epoch": 0.9038689313091838,
+      "grad_norm": 0.3603505485429001,
+      "learning_rate": 4.834825495441475e-06,
+      "loss": 1.0784,
+      "step": 4698
+    },
+    {
+      "epoch": 0.9040613257177214,
+      "grad_norm": 0.3685601711605079,
+      "learning_rate": 4.815700327490014e-06,
+      "loss": 0.9907,
+      "step": 4699
+    },
+    {
+      "epoch": 0.9042537201262588,
+      "grad_norm": 0.29755955115289273,
+      "learning_rate": 4.796612128023725e-06,
+      "loss": 1.0372,
+      "step": 4700
+    },
+    {
+      "epoch": 0.9044461145347963,
+      "grad_norm": 0.3702446507454852,
+      "learning_rate": 4.777560904456235e-06,
+      "loss": 0.9751,
+      "step": 4701
+    },
+    {
+      "epoch": 0.9046385089433339,
+      "grad_norm": 0.33105706567839577,
+      "learning_rate": 4.758546664186869e-06,
+      "loss": 1.1285,
+      "step": 4702
+    },
+    {
+      "epoch": 0.9048309033518713,
+      "grad_norm": 0.3012697441491406,
+      "learning_rate": 4.739569414600498e-06,
+      "loss": 1.1422,
+      "step": 4703
+    },
+    {
+      "epoch": 0.9050232977604088,
+      "grad_norm": 0.48643189319635904,
+      "learning_rate": 4.720629163067703e-06,
+      "loss": 0.9624,
+      "step": 4704
+    },
+    {
+      "epoch": 0.9052156921689464,
+      "grad_norm": 0.27923402852355445,
+      "learning_rate": 4.70172591694461e-06,
+      "loss": 1.0395,
+      "step": 4705
+    },
+    {
+      "epoch": 0.9054080865774838,
+      "grad_norm": 0.37346233164118836,
+      "learning_rate": 4.682859683573048e-06,
+      "loss": 1.0817,
+      "step": 4706
+    },
+    {
+      "epoch": 0.9056004809860213,
+      "grad_norm": 0.294511129772034,
+      "learning_rate": 4.664030470280467e-06,
+      "loss": 0.9809,
+      "step": 4707
+    },
+    {
+      "epoch": 0.9057928753945589,
+      "grad_norm": 0.33917260682397626,
+      "learning_rate": 4.645238284379883e-06,
+      "loss": 1.0332,
+      "step": 4708
+    },
+    {
+      "epoch": 0.9059852698030963,
+      "grad_norm": 0.3418696516815014,
+      "learning_rate": 4.6264831331699675e-06,
+      "loss": 1.0228,
+      "step": 4709
+    },
+    {
+      "epoch": 0.9061776642116338,
+      "grad_norm": 0.2785906410099841,
+      "learning_rate": 4.60776502393504e-06,
+      "loss": 1.0427,
+      "step": 4710
+    },
+    {
+      "epoch": 0.9063700586201714,
+      "grad_norm": 0.2751496166182031,
+      "learning_rate": 4.589083963944951e-06,
+      "loss": 1.1048,
+      "step": 4711
+    },
+    {
+      "epoch": 0.9065624530287089,
+      "grad_norm": 0.3469497046864914,
+      "learning_rate": 4.570439960455242e-06,
+      "loss": 1.1205,
+      "step": 4712
+    },
+    {
+      "epoch": 0.9067548474372463,
+      "grad_norm": 0.32243734446344646,
+      "learning_rate": 4.551833020707008e-06,
+      "loss": 1.0166,
+      "step": 4713
+    },
+    {
+      "epoch": 0.9069472418457839,
+      "grad_norm": 0.2835201392756422,
+      "learning_rate": 4.533263151926981e-06,
+      "loss": 0.9664,
+      "step": 4714
+    },
+    {
+      "epoch": 0.9071396362543214,
+      "grad_norm": 0.3344748355979359,
+      "learning_rate": 4.5147303613275016e-06,
+      "loss": 1.045,
+      "step": 4715
+    },
+    {
+      "epoch": 0.9073320306628588,
+      "grad_norm": 0.3730153351141198,
+      "learning_rate": 4.496234656106457e-06,
+      "loss": 1.0579,
+      "step": 4716
+    },
+    {
+      "epoch": 0.9075244250713964,
+      "grad_norm": 0.390287045584428,
+      "learning_rate": 4.477776043447379e-06,
+      "loss": 1.0027,
+      "step": 4717
+    },
+    {
+      "epoch": 0.9077168194799339,
+      "grad_norm": 0.3310936579781912,
+      "learning_rate": 4.459354530519377e-06,
+      "loss": 1.0485,
+      "step": 4718
+    },
+    {
+      "epoch": 0.9079092138884713,
+      "grad_norm": 0.3354224436322591,
+      "learning_rate": 4.440970124477173e-06,
+      "loss": 0.9852,
+      "step": 4719
+    },
+    {
+      "epoch": 0.9081016082970089,
+      "grad_norm": 0.2848795605285452,
+      "learning_rate": 4.422622832461054e-06,
+      "loss": 1.0174,
+      "step": 4720
+    },
+    {
+      "epoch": 0.9082940027055464,
+      "grad_norm": 0.3156990972632789,
+      "learning_rate": 4.404312661596877e-06,
+      "loss": 1.0204,
+      "step": 4721
+    },
+    {
+      "epoch": 0.9084863971140839,
+      "grad_norm": 0.29160329378948585,
+      "learning_rate": 4.386039618996119e-06,
+      "loss": 1.1066,
+      "step": 4722
+    },
+    {
+      "epoch": 0.9086787915226213,
+      "grad_norm": 0.27796622696943685,
+      "learning_rate": 4.367803711755814e-06,
+      "loss": 1.0305,
+      "step": 4723
+    },
+    {
+      "epoch": 0.9088711859311589,
+      "grad_norm": 0.317004259576644,
+      "learning_rate": 4.349604946958563e-06,
+      "loss": 1.1259,
+      "step": 4724
+    },
+    {
+      "epoch": 0.9090635803396964,
+      "grad_norm": 0.29366726007516236,
+      "learning_rate": 4.331443331672591e-06,
+      "loss": 1.0927,
+      "step": 4725
+    },
+    {
+      "epoch": 0.9092559747482338,
+      "grad_norm": 0.3504554452641467,
+      "learning_rate": 4.313318872951633e-06,
+      "loss": 1.0233,
+      "step": 4726
+    },
+    {
+      "epoch": 0.9094483691567714,
+      "grad_norm": 0.39513607869209655,
+      "learning_rate": 4.295231577835024e-06,
+      "loss": 1.0777,
+      "step": 4727
+    },
+    {
+      "epoch": 0.9096407635653089,
+      "grad_norm": 0.36810292242258597,
+      "learning_rate": 4.277181453347678e-06,
+      "loss": 1.0282,
+      "step": 4728
+    },
+    {
+      "epoch": 0.9098331579738463,
+      "grad_norm": 0.36434477303207796,
+      "learning_rate": 4.259168506500022e-06,
+      "loss": 1.0809,
+      "step": 4729
+    },
+    {
+      "epoch": 0.9100255523823839,
+      "grad_norm": 0.35647505279118624,
+      "learning_rate": 4.241192744288091e-06,
+      "loss": 1.0299,
+      "step": 4730
+    },
+    {
+      "epoch": 0.9102179467909214,
+      "grad_norm": 0.401373897766685,
+      "learning_rate": 4.223254173693492e-06,
+      "loss": 1.0311,
+      "step": 4731
+    },
+    {
+      "epoch": 0.9104103411994589,
+      "grad_norm": 0.3423815390074625,
+      "learning_rate": 4.205352801683327e-06,
+      "loss": 0.9968,
+      "step": 4732
+    },
+    {
+      "epoch": 0.9106027356079964,
+      "grad_norm": 0.34650929862977964,
+      "learning_rate": 4.187488635210302e-06,
+      "loss": 1.0492,
+      "step": 4733
+    },
+    {
+      "epoch": 0.9107951300165339,
+      "grad_norm": 0.3320047163071482,
+      "learning_rate": 4.1696616812126335e-06,
+      "loss": 1.0063,
+      "step": 4734
+    },
+    {
+      "epoch": 0.9109875244250714,
+      "grad_norm": 0.2984260035123869,
+      "learning_rate": 4.151871946614116e-06,
+      "loss": 1.0452,
+      "step": 4735
+    },
+    {
+      "epoch": 0.9111799188336089,
+      "grad_norm": 0.3287456793464544,
+      "learning_rate": 4.13411943832408e-06,
+      "loss": 1.0614,
+      "step": 4736
+    },
+    {
+      "epoch": 0.9113723132421464,
+      "grad_norm": 0.3538509811310011,
+      "learning_rate": 4.116404163237386e-06,
+      "loss": 0.9903,
+      "step": 4737
+    },
+    {
+      "epoch": 0.9115647076506839,
+      "grad_norm": 0.311754307606841,
+      "learning_rate": 4.098726128234443e-06,
+      "loss": 1.0264,
+      "step": 4738
+    },
+    {
+      "epoch": 0.9117571020592214,
+      "grad_norm": 0.31461694776761406,
+      "learning_rate": 4.081085340181223e-06,
+      "loss": 1.0102,
+      "step": 4739
+    },
+    {
+      "epoch": 0.9119494964677589,
+      "grad_norm": 0.3428107142670629,
+      "learning_rate": 4.06348180592917e-06,
+      "loss": 1.0326,
+      "step": 4740
+    },
+    {
+      "epoch": 0.9121418908762964,
+      "grad_norm": 0.3130639226250299,
+      "learning_rate": 4.0459155323153035e-06,
+      "loss": 1.0298,
+      "step": 4741
+    },
+    {
+      "epoch": 0.912334285284834,
+      "grad_norm": 0.31660763786422674,
+      "learning_rate": 4.028386526162175e-06,
+      "loss": 1.1499,
+      "step": 4742
+    },
+    {
+      "epoch": 0.9125266796933714,
+      "grad_norm": 0.35070366408568776,
+      "learning_rate": 4.010894794277831e-06,
+      "loss": 1.0468,
+      "step": 4743
+    },
+    {
+      "epoch": 0.9127190741019089,
+      "grad_norm": 0.32266840253917106,
+      "learning_rate": 3.993440343455878e-06,
+      "loss": 0.9768,
+      "step": 4744
+    },
+    {
+      "epoch": 0.9129114685104465,
+      "grad_norm": 0.30485752247821407,
+      "learning_rate": 3.976023180475397e-06,
+      "loss": 0.9941,
+      "step": 4745
+    },
+    {
+      "epoch": 0.9131038629189839,
+      "grad_norm": 0.349700848622084,
+      "learning_rate": 3.95864331210104e-06,
+      "loss": 1.1024,
+      "step": 4746
+    },
+    {
+      "epoch": 0.9132962573275214,
+      "grad_norm": 0.35768207410177166,
+      "learning_rate": 3.941300745082932e-06,
+      "loss": 1.0066,
+      "step": 4747
+    },
+    {
+      "epoch": 0.9134886517360589,
+      "grad_norm": 0.3486707573125071,
+      "learning_rate": 3.923995486156717e-06,
+      "loss": 0.9458,
+      "step": 4748
+    },
+    {
+      "epoch": 0.9136810461445964,
+      "grad_norm": 0.29821996715749854,
+      "learning_rate": 3.906727542043598e-06,
+      "loss": 1.0813,
+      "step": 4749
+    },
+    {
+      "epoch": 0.9138734405531339,
+      "grad_norm": 0.34515012431476105,
+      "learning_rate": 3.889496919450208e-06,
+      "loss": 1.0102,
+      "step": 4750
+    },
+    {
+      "epoch": 0.9140658349616714,
+      "grad_norm": 0.25496242531786867,
+      "learning_rate": 3.872303625068729e-06,
+      "loss": 1.0312,
+      "step": 4751
+    },
+    {
+      "epoch": 0.914258229370209,
+      "grad_norm": 0.34750426348871405,
+      "learning_rate": 3.8551476655768525e-06,
+      "loss": 0.9874,
+      "step": 4752
+    },
+    {
+      "epoch": 0.9144506237787464,
+      "grad_norm": 0.2744330157187026,
+      "learning_rate": 3.8380290476377255e-06,
+      "loss": 0.9901,
+      "step": 4753
+    },
+    {
+      "epoch": 0.9146430181872839,
+      "grad_norm": 0.3876490329659078,
+      "learning_rate": 3.820947777900063e-06,
+      "loss": 0.9687,
+      "step": 4754
+    },
+    {
+      "epoch": 0.9148354125958215,
+      "grad_norm": 0.39139861433816614,
+      "learning_rate": 3.803903862998004e-06,
+      "loss": 1.0299,
+      "step": 4755
+    },
+    {
+      "epoch": 0.9150278070043589,
+      "grad_norm": 0.3063743765811644,
+      "learning_rate": 3.7868973095512185e-06,
+      "loss": 1.0577,
+      "step": 4756
+    },
+    {
+      "epoch": 0.9152202014128964,
+      "grad_norm": 0.36828417153808,
+      "learning_rate": 3.7699281241648564e-06,
+      "loss": 0.9977,
+      "step": 4757
+    },
+    {
+      "epoch": 0.915412595821434,
+      "grad_norm": 0.28780300258749514,
+      "learning_rate": 3.7529963134295465e-06,
+      "loss": 1.0596,
+      "step": 4758
+    },
+    {
+      "epoch": 0.9156049902299714,
+      "grad_norm": 0.3461580433326025,
+      "learning_rate": 3.7361018839214057e-06,
+      "loss": 1.0413,
+      "step": 4759
+    },
+    {
+      "epoch": 0.9157973846385089,
+      "grad_norm": 0.2850723602444769,
+      "learning_rate": 3.7192448422020743e-06,
+      "loss": 1.0108,
+      "step": 4760
+    },
+    {
+      "epoch": 0.9159897790470465,
+      "grad_norm": 0.2810026001645126,
+      "learning_rate": 3.7024251948185816e-06,
+      "loss": 1.0281,
+      "step": 4761
+    },
+    {
+      "epoch": 0.916182173455584,
+      "grad_norm": 0.3203215907471018,
+      "learning_rate": 3.6856429483035028e-06,
+      "loss": 0.9606,
+      "step": 4762
+    },
+    {
+      "epoch": 0.9163745678641214,
+      "grad_norm": 0.3064091568721602,
+      "learning_rate": 3.66889810917489e-06,
+      "loss": 0.95,
+      "step": 4763
+    },
+    {
+      "epoch": 0.916566962272659,
+      "grad_norm": 0.31491530367666204,
+      "learning_rate": 3.6521906839362185e-06,
+      "loss": 1.1035,
+      "step": 4764
+    },
+    {
+      "epoch": 0.9167593566811965,
+      "grad_norm": 0.32663413046198564,
+      "learning_rate": 3.635520679076465e-06,
+      "loss": 1.0336,
+      "step": 4765
+    },
+    {
+      "epoch": 0.9169517510897339,
+      "grad_norm": 0.335316225236997,
+      "learning_rate": 3.6188881010700726e-06,
+      "loss": 0.9827,
+      "step": 4766
+    },
+    {
+      "epoch": 0.9171441454982715,
+      "grad_norm": 0.2799735952248801,
+      "learning_rate": 3.6022929563769513e-06,
+      "loss": 1.0344,
+      "step": 4767
+    },
+    {
+      "epoch": 0.917336539906809,
+      "grad_norm": 0.30110546928367454,
+      "learning_rate": 3.585735251442457e-06,
+      "loss": 1.0451,
+      "step": 4768
+    },
+    {
+      "epoch": 0.9175289343153464,
+      "grad_norm": 0.30754811420297273,
+      "learning_rate": 3.5692149926974006e-06,
+      "loss": 1.0564,
+      "step": 4769
+    },
+    {
+      "epoch": 0.917721328723884,
+      "grad_norm": 0.4117794890640593,
+      "learning_rate": 3.552732186558072e-06,
+      "loss": 1.0439,
+      "step": 4770
+    },
+    {
+      "epoch": 0.9179137231324215,
+      "grad_norm": 0.2867302225609774,
+      "learning_rate": 3.536286839426195e-06,
+      "loss": 1.0288,
+      "step": 4771
+    },
+    {
+      "epoch": 0.9181061175409589,
+      "grad_norm": 0.3223331881370476,
+      "learning_rate": 3.5198789576889603e-06,
+      "loss": 1.0712,
+      "step": 4772
+    },
+    {
+      "epoch": 0.9182985119494965,
+      "grad_norm": 0.299777090994648,
+      "learning_rate": 3.5035085477190143e-06,
+      "loss": 0.9606,
+      "step": 4773
+    },
+    {
+      "epoch": 0.918490906358034,
+      "grad_norm": 0.2979044891398794,
+      "learning_rate": 3.4871756158744052e-06,
+      "loss": 1.1323,
+      "step": 4774
+    },
+    {
+      "epoch": 0.9186833007665715,
+      "grad_norm": 0.3504658917550687,
+      "learning_rate": 3.470880168498669e-06,
+      "loss": 0.9905,
+      "step": 4775
+    },
+    {
+      "epoch": 0.9188756951751089,
+      "grad_norm": 0.28519204077072113,
+      "learning_rate": 3.454622211920766e-06,
+      "loss": 1.0676,
+      "step": 4776
+    },
+    {
+      "epoch": 0.9190680895836465,
+      "grad_norm": 0.3749364264858349,
+      "learning_rate": 3.4384017524551116e-06,
+      "loss": 1.0138,
+      "step": 4777
+    },
+    {
+      "epoch": 0.919260483992184,
+      "grad_norm": 0.29620547010939036,
+      "learning_rate": 3.422218796401544e-06,
+      "loss": 1.0629,
+      "step": 4778
+    },
+    {
+      "epoch": 0.9194528784007214,
+      "grad_norm": 0.3448874604740178,
+      "learning_rate": 3.4060733500453247e-06,
+      "loss": 1.0431,
+      "step": 4779
+    },
+    {
+      "epoch": 0.919645272809259,
+      "grad_norm": 0.4293918180477315,
+      "learning_rate": 3.3899654196571597e-06,
+      "loss": 0.9424,
+      "step": 4780
+    },
+    {
+      "epoch": 0.9198376672177965,
+      "grad_norm": 0.2685364751795157,
+      "learning_rate": 3.3738950114932e-06,
+      "loss": 1.1123,
+      "step": 4781
+    },
+    {
+      "epoch": 0.9200300616263339,
+      "grad_norm": 0.3094663767757603,
+      "learning_rate": 3.3578621317949756e-06,
+      "loss": 0.9291,
+      "step": 4782
+    },
+    {
+      "epoch": 0.9202224560348715,
+      "grad_norm": 0.31426169410071264,
+      "learning_rate": 3.3418667867895047e-06,
+      "loss": 1.0479,
+      "step": 4783
+    },
+    {
+      "epoch": 0.920414850443409,
+      "grad_norm": 0.3005711694545352,
+      "learning_rate": 3.3259089826891854e-06,
+      "loss": 1.0308,
+      "step": 4784
+    },
+    {
+      "epoch": 0.9206072448519464,
+      "grad_norm": 0.32478489128523025,
+      "learning_rate": 3.309988725691837e-06,
+      "loss": 1.0518,
+      "step": 4785
+    },
+    {
+      "epoch": 0.920799639260484,
+      "grad_norm": 0.378754173468823,
+      "learning_rate": 3.294106021980714e-06,
+      "loss": 0.9836,
+      "step": 4786
+    },
+    {
+      "epoch": 0.9209920336690215,
+      "grad_norm": 0.2864945578879739,
+      "learning_rate": 3.2782608777244705e-06,
+      "loss": 1.097,
+      "step": 4787
+    },
+    {
+      "epoch": 0.921184428077559,
+      "grad_norm": 0.3110923092227332,
+      "learning_rate": 3.2624532990771505e-06,
+      "loss": 1.1217,
+      "step": 4788
+    },
+    {
+      "epoch": 0.9213768224860965,
+      "grad_norm": 0.3725423853538233,
+      "learning_rate": 3.2466832921782985e-06,
+      "loss": 1.0299,
+      "step": 4789
+    },
+    {
+      "epoch": 0.921569216894634,
+      "grad_norm": 0.3205879519436053,
+      "learning_rate": 3.2309508631527486e-06,
+      "loss": 1.0659,
+      "step": 4790
+    },
+    {
+      "epoch": 0.9217616113031715,
+      "grad_norm": 0.33535870039183635,
+      "learning_rate": 3.215256018110824e-06,
+      "loss": 1.0597,
+      "step": 4791
+    },
+    {
+      "epoch": 0.921954005711709,
+      "grad_norm": 0.31841686704841865,
+      "learning_rate": 3.1995987631482148e-06,
+      "loss": 1.0244,
+      "step": 4792
+    },
+    {
+      "epoch": 0.9221464001202465,
+      "grad_norm": 0.2891392043589409,
+      "learning_rate": 3.1839791043460023e-06,
+      "loss": 1.0667,
+      "step": 4793
+    },
+    {
+      "epoch": 0.922338794528784,
+      "grad_norm": 0.350886995142571,
+      "learning_rate": 3.168397047770699e-06,
+      "loss": 1.0267,
+      "step": 4794
+    },
+    {
+      "epoch": 0.9225311889373216,
+      "grad_norm": 0.33852892217693087,
+      "learning_rate": 3.1528525994741877e-06,
+      "loss": 0.9916,
+      "step": 4795
+    },
+    {
+      "epoch": 0.922723583345859,
+      "grad_norm": 0.44013632400182945,
+      "learning_rate": 3.13734576549376e-06,
+      "loss": 1.0408,
+      "step": 4796
+    },
+    {
+      "epoch": 0.9229159777543965,
+      "grad_norm": 0.32845052873146846,
+      "learning_rate": 3.1218765518520985e-06,
+      "loss": 0.9718,
+      "step": 4797
+    },
+    {
+      "epoch": 0.9231083721629341,
+      "grad_norm": 0.32663070994418064,
+      "learning_rate": 3.1064449645572537e-06,
+      "loss": 1.0918,
+      "step": 4798
+    },
+    {
+      "epoch": 0.9233007665714715,
+      "grad_norm": 0.26077891738687076,
+      "learning_rate": 3.0910510096026745e-06,
+      "loss": 0.991,
+      "step": 4799
+    },
+    {
+      "epoch": 0.923493160980009,
+      "grad_norm": 0.34353592786377535,
+      "learning_rate": 3.0756946929672014e-06,
+      "loss": 0.9836,
+      "step": 4800
+    },
+    {
+      "epoch": 0.9236855553885466,
+      "grad_norm": 0.32664461616404566,
+      "learning_rate": 3.0603760206150524e-06,
+      "loss": 0.9964,
+      "step": 4801
+    },
+    {
+      "epoch": 0.923877949797084,
+      "grad_norm": 0.3206455614642741,
+      "learning_rate": 3.045094998495834e-06,
+      "loss": 1.001,
+      "step": 4802
+    },
+    {
+      "epoch": 0.9240703442056215,
+      "grad_norm": 0.2951839621408488,
+      "learning_rate": 3.029851632544489e-06,
+      "loss": 0.9973,
+      "step": 4803
+    },
+    {
+      "epoch": 0.924262738614159,
+      "grad_norm": 0.37290732523734266,
+      "learning_rate": 3.0146459286813923e-06,
+      "loss": 1.1221,
+      "step": 4804
+    },
+    {
+      "epoch": 0.9244551330226966,
+      "grad_norm": 0.29236760879957674,
+      "learning_rate": 2.9994778928122636e-06,
+      "loss": 1.0356,
+      "step": 4805
+    },
+    {
+      "epoch": 0.924647527431234,
+      "grad_norm": 0.2931870899031517,
+      "learning_rate": 2.9843475308281575e-06,
+      "loss": 0.9466,
+      "step": 4806
+    },
+    {
+      "epoch": 0.9248399218397715,
+      "grad_norm": 0.31973296672323614,
+      "learning_rate": 2.9692548486055847e-06,
+      "loss": 1.0945,
+      "step": 4807
+    },
+    {
+      "epoch": 0.9250323162483091,
+      "grad_norm": 0.3661219127547001,
+      "learning_rate": 2.954199852006334e-06,
+      "loss": 1.0427,
+      "step": 4808
+    },
+    {
+      "epoch": 0.9252247106568465,
+      "grad_norm": 0.4175494631158559,
+      "learning_rate": 2.9391825468775945e-06,
+      "loss": 1.0891,
+      "step": 4809
+    },
+    {
+      "epoch": 0.925417105065384,
+      "grad_norm": 0.2788125545134578,
+      "learning_rate": 2.9242029390519456e-06,
+      "loss": 1.0199,
+      "step": 4810
+    },
+    {
+      "epoch": 0.9256094994739216,
+      "grad_norm": 0.35030552760024536,
+      "learning_rate": 2.909261034347255e-06,
+      "loss": 0.9827,
+      "step": 4811
+    },
+    {
+      "epoch": 0.925801893882459,
+      "grad_norm": 0.27581407804012614,
+      "learning_rate": 2.8943568385667918e-06,
+      "loss": 0.9738,
+      "step": 4812
+    },
+    {
+      "epoch": 0.9259942882909965,
+      "grad_norm": 0.44156556297116145,
+      "learning_rate": 2.8794903574992037e-06,
+      "loss": 1.0266,
+      "step": 4813
+    },
+    {
+      "epoch": 0.9261866826995341,
+      "grad_norm": 0.29859310101493625,
+      "learning_rate": 2.8646615969184276e-06,
+      "loss": 1.0063,
+      "step": 4814
+    },
+    {
+      "epoch": 0.9263790771080715,
+      "grad_norm": 0.36351924132348623,
+      "learning_rate": 2.8498705625838117e-06,
+      "loss": 0.9289,
+      "step": 4815
+    },
+    {
+      "epoch": 0.926571471516609,
+      "grad_norm": 0.30891981525308526,
+      "learning_rate": 2.8351172602399946e-06,
+      "loss": 1.032,
+      "step": 4816
+    },
+    {
+      "epoch": 0.9267638659251466,
+      "grad_norm": 0.31948945035477055,
+      "learning_rate": 2.820401695616992e-06,
+      "loss": 1.069,
+      "step": 4817
+    },
+    {
+      "epoch": 0.9269562603336841,
+      "grad_norm": 0.2909736294090731,
+      "learning_rate": 2.8057238744301996e-06,
+      "loss": 1.0696,
+      "step": 4818
+    },
+    {
+      "epoch": 0.9271486547422215,
+      "grad_norm": 0.2860334733992794,
+      "learning_rate": 2.7910838023802676e-06,
+      "loss": 1.0031,
+      "step": 4819
+    },
+    {
+      "epoch": 0.9273410491507591,
+      "grad_norm": 0.3069103045701495,
+      "learning_rate": 2.7764814851532482e-06,
+      "loss": 1.1075,
+      "step": 4820
+    },
+    {
+      "epoch": 0.9275334435592966,
+      "grad_norm": 0.27667505843985585,
+      "learning_rate": 2.761916928420527e-06,
+      "loss": 1.0639,
+      "step": 4821
+    },
+    {
+      "epoch": 0.927725837967834,
+      "grad_norm": 0.3386318817559325,
+      "learning_rate": 2.74739013783879e-06,
+      "loss": 0.8886,
+      "step": 4822
+    },
+    {
+      "epoch": 0.9279182323763716,
+      "grad_norm": 0.3547979920898592,
+      "learning_rate": 2.7329011190500796e-06,
+      "loss": 1.028,
+      "step": 4823
+    },
+    {
+      "epoch": 0.9281106267849091,
+      "grad_norm": 0.35650567572881114,
+      "learning_rate": 2.7184498776817612e-06,
+      "loss": 1.0684,
+      "step": 4824
+    },
+    {
+      "epoch": 0.9283030211934465,
+      "grad_norm": 0.32744332231790724,
+      "learning_rate": 2.704036419346534e-06,
+      "loss": 1.1264,
+      "step": 4825
+    },
+    {
+      "epoch": 0.9284954156019841,
+      "grad_norm": 0.3674114391060293,
+      "learning_rate": 2.689660749642442e-06,
+      "loss": 1.0133,
+      "step": 4826
+    },
+    {
+      "epoch": 0.9286878100105216,
+      "grad_norm": 0.37083950050765935,
+      "learning_rate": 2.675322874152786e-06,
+      "loss": 1.0182,
+      "step": 4827
+    },
+    {
+      "epoch": 0.928880204419059,
+      "grad_norm": 0.3209775086730098,
+      "learning_rate": 2.6610227984462554e-06,
+      "loss": 1.0543,
+      "step": 4828
+    },
+    {
+      "epoch": 0.9290725988275965,
+      "grad_norm": 0.2825439751894937,
+      "learning_rate": 2.6467605280768413e-06,
+      "loss": 0.9806,
+      "step": 4829
+    },
+    {
+      "epoch": 0.9292649932361341,
+      "grad_norm": 0.2878430630772789,
+      "learning_rate": 2.632536068583824e-06,
+      "loss": 1.0246,
+      "step": 4830
+    },
+    {
+      "epoch": 0.9294573876446716,
+      "grad_norm": 0.2675421152066387,
+      "learning_rate": 2.6183494254918505e-06,
+      "loss": 1.1268,
+      "step": 4831
+    },
+    {
+      "epoch": 0.929649782053209,
+      "grad_norm": 0.3494602575053354,
+      "learning_rate": 2.6042006043108247e-06,
+      "loss": 0.9946,
+      "step": 4832
+    },
+    {
+      "epoch": 0.9298421764617466,
+      "grad_norm": 0.2922246085136682,
+      "learning_rate": 2.590089610535984e-06,
+      "loss": 1.0399,
+      "step": 4833
+    },
+    {
+      "epoch": 0.9300345708702841,
+      "grad_norm": 0.3528907431025616,
+      "learning_rate": 2.5760164496479e-06,
+      "loss": 1.0514,
+      "step": 4834
+    },
+    {
+      "epoch": 0.9302269652788215,
+      "grad_norm": 0.3598896569766778,
+      "learning_rate": 2.5619811271123894e-06,
+      "loss": 1.0533,
+      "step": 4835
+    },
+    {
+      "epoch": 0.9304193596873591,
+      "grad_norm": 0.3560654831586113,
+      "learning_rate": 2.5479836483806587e-06,
+      "loss": 0.9881,
+      "step": 4836
+    },
+    {
+      "epoch": 0.9306117540958966,
+      "grad_norm": 0.3913011561841819,
+      "learning_rate": 2.534024018889114e-06,
+      "loss": 1.0933,
+      "step": 4837
+    },
+    {
+      "epoch": 0.930804148504434,
+      "grad_norm": 0.339144569976722,
+      "learning_rate": 2.5201022440595522e-06,
+      "loss": 1.0482,
+      "step": 4838
+    },
+    {
+      "epoch": 0.9309965429129716,
+      "grad_norm": 0.32288385554851634,
+      "learning_rate": 2.506218329299026e-06,
+      "loss": 1.0367,
+      "step": 4839
+    },
+    {
+      "epoch": 0.9311889373215091,
+      "grad_norm": 0.32793579252765287,
+      "learning_rate": 2.492372279999866e-06,
+      "loss": 1.0376,
+      "step": 4840
+    },
+    {
+      "epoch": 0.9313813317300466,
+      "grad_norm": 0.3498260259802165,
+      "learning_rate": 2.4785641015397374e-06,
+      "loss": 1.031,
+      "step": 4841
+    },
+    {
+      "epoch": 0.9315737261385841,
+      "grad_norm": 0.41037160210456186,
+      "learning_rate": 2.464793799281573e-06,
+      "loss": 0.9956,
+      "step": 4842
+    },
+    {
+      "epoch": 0.9317661205471216,
+      "grad_norm": 0.3285375397793912,
+      "learning_rate": 2.4510613785735938e-06,
+      "loss": 1.025,
+      "step": 4843
+    },
+    {
+      "epoch": 0.9319585149556591,
+      "grad_norm": 0.35561665183650015,
+      "learning_rate": 2.4373668447493224e-06,
+      "loss": 1.0575,
+      "step": 4844
+    },
+    {
+      "epoch": 0.9321509093641966,
+      "grad_norm": 0.25749452794270217,
+      "learning_rate": 2.4237102031275607e-06,
+      "loss": 1.0743,
+      "step": 4845
+    },
+    {
+      "epoch": 0.9323433037727341,
+      "grad_norm": 0.33505290025284906,
+      "learning_rate": 2.4100914590123756e-06,
+      "loss": 1.0315,
+      "step": 4846
+    },
+    {
+      "epoch": 0.9325356981812716,
+      "grad_norm": 0.28728856362608546,
+      "learning_rate": 2.3965106176931374e-06,
+      "loss": 0.9858,
+      "step": 4847
+    },
+    {
+      "epoch": 0.9327280925898092,
+      "grad_norm": 0.4902951527240888,
+      "learning_rate": 2.382967684444493e-06,
+      "loss": 1.0443,
+      "step": 4848
+    },
+    {
+      "epoch": 0.9329204869983466,
+      "grad_norm": 0.2917632548051919,
+      "learning_rate": 2.3694626645263675e-06,
+      "loss": 1.0427,
+      "step": 4849
+    },
+    {
+      "epoch": 0.9331128814068841,
+      "grad_norm": 0.33975483972911147,
+      "learning_rate": 2.3559955631839434e-06,
+      "loss": 1.0502,
+      "step": 4850
+    },
+    {
+      "epoch": 0.9333052758154217,
+      "grad_norm": 0.40146935445843523,
+      "learning_rate": 2.3425663856476932e-06,
+      "loss": 0.9184,
+      "step": 4851
+    },
+    {
+      "epoch": 0.9334976702239591,
+      "grad_norm": 0.31735660398853427,
+      "learning_rate": 2.3291751371333437e-06,
+      "loss": 0.9794,
+      "step": 4852
+    },
+    {
+      "epoch": 0.9336900646324966,
+      "grad_norm": 0.310236063303638,
+      "learning_rate": 2.3158218228419127e-06,
+      "loss": 1.0753,
+      "step": 4853
+    },
+    {
+      "epoch": 0.9338824590410342,
+      "grad_norm": 0.3425510296524538,
+      "learning_rate": 2.3025064479596627e-06,
+      "loss": 1.0535,
+      "step": 4854
+    },
+    {
+      "epoch": 0.9340748534495716,
+      "grad_norm": 0.3584980665652243,
+      "learning_rate": 2.2892290176581676e-06,
+      "loss": 1.0197,
+      "step": 4855
+    },
+    {
+      "epoch": 0.9342672478581091,
+      "grad_norm": 0.28912489809723535,
+      "learning_rate": 2.2759895370941807e-06,
+      "loss": 1.0575,
+      "step": 4856
+    },
+    {
+      "epoch": 0.9344596422666466,
+      "grad_norm": 0.3486576476561234,
+      "learning_rate": 2.2627880114097776e-06,
+      "loss": 0.9741,
+      "step": 4857
+    },
+    {
+      "epoch": 0.9346520366751842,
+      "grad_norm": 0.29947452971608934,
+      "learning_rate": 2.249624445732301e-06,
+      "loss": 1.033,
+      "step": 4858
+    },
+    {
+      "epoch": 0.9348444310837216,
+      "grad_norm": 0.33413340824935367,
+      "learning_rate": 2.2364988451742953e-06,
+      "loss": 0.9967,
+      "step": 4859
+    },
+    {
+      "epoch": 0.9350368254922591,
+      "grad_norm": 0.33405221722967166,
+      "learning_rate": 2.223411214833637e-06,
+      "loss": 1.0954,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9352292199007967,
+      "grad_norm": 0.4607464547692085,
+      "learning_rate": 2.210361559793361e-06,
+      "loss": 0.9939,
+      "step": 4861
+    },
+    {
+      "epoch": 0.9354216143093341,
+      "grad_norm": 0.25654193414478754,
+      "learning_rate": 2.1973498851218242e-06,
+      "loss": 1.048,
+      "step": 4862
+    },
+    {
+      "epoch": 0.9356140087178716,
+      "grad_norm": 0.3289380619374026,
+      "learning_rate": 2.1843761958726285e-06,
+      "loss": 1.122,
+      "step": 4863
+    },
+    {
+      "epoch": 0.9358064031264092,
+      "grad_norm": 0.3376021402986719,
+      "learning_rate": 2.1714404970845647e-06,
+      "loss": 1.0825,
+      "step": 4864
+    },
+    {
+      "epoch": 0.9359987975349466,
+      "grad_norm": 0.33197297073054194,
+      "learning_rate": 2.158542793781759e-06,
+      "loss": 1.1005,
+      "step": 4865
+    },
+    {
+      "epoch": 0.9361911919434841,
+      "grad_norm": 0.3715262749437805,
+      "learning_rate": 2.1456830909734936e-06,
+      "loss": 1.0331,
+      "step": 4866
+    },
+    {
+      "epoch": 0.9363835863520217,
+      "grad_norm": 0.31017615300071755,
+      "learning_rate": 2.1328613936543397e-06,
+      "loss": 0.986,
+      "step": 4867
+    },
+    {
+      "epoch": 0.9365759807605591,
+      "grad_norm": 0.4285549502927926,
+      "learning_rate": 2.1200777068041134e-06,
+      "loss": 1.0437,
+      "step": 4868
+    },
+    {
+      "epoch": 0.9367683751690966,
+      "grad_norm": 0.28864745103907863,
+      "learning_rate": 2.10733203538781e-06,
+      "loss": 1.0907,
+      "step": 4869
+    },
+    {
+      "epoch": 0.9369607695776342,
+      "grad_norm": 0.36904236344609426,
+      "learning_rate": 2.0946243843557367e-06,
+      "loss": 1.0278,
+      "step": 4870
+    },
+    {
+      "epoch": 0.9371531639861717,
+      "grad_norm": 0.3102728966580874,
+      "learning_rate": 2.081954758643401e-06,
+      "loss": 1.029,
+      "step": 4871
+    },
+    {
+      "epoch": 0.9373455583947091,
+      "grad_norm": 0.3123550210871251,
+      "learning_rate": 2.0693231631715103e-06,
+      "loss": 0.9711,
+      "step": 4872
+    },
+    {
+      "epoch": 0.9375379528032467,
+      "grad_norm": 0.26796339723537277,
+      "learning_rate": 2.0567296028460637e-06,
+      "loss": 1.0617,
+      "step": 4873
+    },
+    {
+      "epoch": 0.9377303472117842,
+      "grad_norm": 0.3169742849441562,
+      "learning_rate": 2.0441740825582257e-06,
+      "loss": 1.0811,
+      "step": 4874
+    },
+    {
+      "epoch": 0.9379227416203216,
+      "grad_norm": 0.3962702167905515,
+      "learning_rate": 2.03165660718444e-06,
+      "loss": 1.1389,
+      "step": 4875
+    },
+    {
+      "epoch": 0.9381151360288592,
+      "grad_norm": 0.38869223617079507,
+      "learning_rate": 2.019177181586329e-06,
+      "loss": 1.0143,
+      "step": 4876
+    },
+    {
+      "epoch": 0.9383075304373967,
+      "grad_norm": 0.43113120042528336,
+      "learning_rate": 2.0067358106107716e-06,
+      "loss": 1.0278,
+      "step": 4877
+    },
+    {
+      "epoch": 0.9384999248459341,
+      "grad_norm": 0.4273010434622452,
+      "learning_rate": 1.994332499089846e-06,
+      "loss": 1.0516,
+      "step": 4878
+    },
+    {
+      "epoch": 0.9386923192544717,
+      "grad_norm": 0.33401338414813536,
+      "learning_rate": 1.9819672518408663e-06,
+      "loss": 0.9897,
+      "step": 4879
+    },
+    {
+      "epoch": 0.9388847136630092,
+      "grad_norm": 0.2903052212541772,
+      "learning_rate": 1.9696400736663457e-06,
+      "loss": 1.0415,
+      "step": 4880
+    },
+    {
+      "epoch": 0.9390771080715467,
+      "grad_norm": 0.2885973544109154,
+      "learning_rate": 1.9573509693540102e-06,
+      "loss": 1.0472,
+      "step": 4881
+    },
+    {
+      "epoch": 0.9392695024800841,
+      "grad_norm": 0.4055233808292139,
+      "learning_rate": 1.945099943676809e-06,
+      "loss": 1.0238,
+      "step": 4882
+    },
+    {
+      "epoch": 0.9394618968886217,
+      "grad_norm": 0.3043600360792589,
+      "learning_rate": 1.9328870013929135e-06,
+      "loss": 0.9858,
+      "step": 4883
+    },
+    {
+      "epoch": 0.9396542912971592,
+      "grad_norm": 0.3288107196283699,
+      "learning_rate": 1.9207121472456847e-06,
+      "loss": 1.0178,
+      "step": 4884
+    },
+    {
+      "epoch": 0.9398466857056966,
+      "grad_norm": 0.34658482573316357,
+      "learning_rate": 1.9085753859636734e-06,
+      "loss": 1.0898,
+      "step": 4885
+    },
+    {
+      "epoch": 0.9400390801142342,
+      "grad_norm": 0.3172365309727629,
+      "learning_rate": 1.8964767222606871e-06,
+      "loss": 1.025,
+      "step": 4886
+    },
+    {
+      "epoch": 0.9402314745227717,
+      "grad_norm": 0.34031796787447277,
+      "learning_rate": 1.8844161608356781e-06,
+      "loss": 1.0427,
+      "step": 4887
+    },
+    {
+      "epoch": 0.9404238689313091,
+      "grad_norm": 0.3291207418764745,
+      "learning_rate": 1.872393706372866e-06,
+      "loss": 1.0009,
+      "step": 4888
+    },
+    {
+      "epoch": 0.9406162633398467,
+      "grad_norm": 0.29478539067629855,
+      "learning_rate": 1.8604093635416153e-06,
+      "loss": 1.0464,
+      "step": 4889
+    },
+    {
+      "epoch": 0.9408086577483842,
+      "grad_norm": 0.3623751718748766,
+      "learning_rate": 1.8484631369964922e-06,
+      "loss": 1.0602,
+      "step": 4890
+    },
+    {
+      "epoch": 0.9410010521569216,
+      "grad_norm": 0.30845842469050583,
+      "learning_rate": 1.836555031377285e-06,
+      "loss": 1.0668,
+      "step": 4891
+    },
+    {
+      "epoch": 0.9411934465654592,
+      "grad_norm": 0.32249605054941965,
+      "learning_rate": 1.824685051308983e-06,
+      "loss": 1.091,
+      "step": 4892
+    },
+    {
+      "epoch": 0.9413858409739967,
+      "grad_norm": 0.3138256728496758,
+      "learning_rate": 1.8128532014017096e-06,
+      "loss": 1.0716,
+      "step": 4893
+    },
+    {
+      "epoch": 0.9415782353825342,
+      "grad_norm": 0.3119478462358224,
+      "learning_rate": 1.8010594862508668e-06,
+      "loss": 0.942,
+      "step": 4894
+    },
+    {
+      "epoch": 0.9417706297910717,
+      "grad_norm": 0.3488029324246239,
+      "learning_rate": 1.789303910436968e-06,
+      "loss": 1.0603,
+      "step": 4895
+    },
+    {
+      "epoch": 0.9419630241996092,
+      "grad_norm": 0.3787800306507009,
+      "learning_rate": 1.777586478525739e-06,
+      "loss": 1.046,
+      "step": 4896
+    },
+    {
+      "epoch": 0.9421554186081467,
+      "grad_norm": 0.3135274644887523,
+      "learning_rate": 1.7659071950681172e-06,
+      "loss": 0.9332,
+      "step": 4897
+    },
+    {
+      "epoch": 0.9423478130166842,
+      "grad_norm": 0.33137046188757624,
+      "learning_rate": 1.754266064600174e-06,
+      "loss": 1.0091,
+      "step": 4898
+    },
+    {
+      "epoch": 0.9425402074252217,
+      "grad_norm": 0.3065716222440659,
+      "learning_rate": 1.7426630916432041e-06,
+      "loss": 1.0374,
+      "step": 4899
+    },
+    {
+      "epoch": 0.9427326018337592,
+      "grad_norm": 0.31380857607841134,
+      "learning_rate": 1.7310982807036912e-06,
+      "loss": 1.0093,
+      "step": 4900
+    },
+    {
+      "epoch": 0.9429249962422968,
+      "grad_norm": 0.3762180087028263,
+      "learning_rate": 1.7195716362732427e-06,
+      "loss": 1.0377,
+      "step": 4901
+    },
+    {
+      "epoch": 0.9431173906508342,
+      "grad_norm": 0.3295357694390801,
+      "learning_rate": 1.7080831628286887e-06,
+      "loss": 1.0466,
+      "step": 4902
+    },
+    {
+      "epoch": 0.9433097850593717,
+      "grad_norm": 0.38756594399970046,
+      "learning_rate": 1.696632864832015e-06,
+      "loss": 1.03,
+      "step": 4903
+    },
+    {
+      "epoch": 0.9435021794679093,
+      "grad_norm": 0.29547189408230545,
+      "learning_rate": 1.685220746730387e-06,
+      "loss": 1.0756,
+      "step": 4904
+    },
+    {
+      "epoch": 0.9436945738764467,
+      "grad_norm": 0.36152742957456924,
+      "learning_rate": 1.6738468129561369e-06,
+      "loss": 0.9748,
+      "step": 4905
+    },
+    {
+      "epoch": 0.9438869682849842,
+      "grad_norm": 0.3394094491262785,
+      "learning_rate": 1.6625110679267641e-06,
+      "loss": 0.9952,
+      "step": 4906
+    },
+    {
+      "epoch": 0.9440793626935218,
+      "grad_norm": 0.32356067482846507,
+      "learning_rate": 1.6512135160449583e-06,
+      "loss": 1.0275,
+      "step": 4907
+    },
+    {
+      "epoch": 0.9442717571020592,
+      "grad_norm": 0.2934165285307212,
+      "learning_rate": 1.6399541616985648e-06,
+      "loss": 1.0749,
+      "step": 4908
+    },
+    {
+      "epoch": 0.9444641515105967,
+      "grad_norm": 0.3739368505174029,
+      "learning_rate": 1.6287330092605524e-06,
+      "loss": 1.0416,
+      "step": 4909
+    },
+    {
+      "epoch": 0.9446565459191342,
+      "grad_norm": 0.31729146999040525,
+      "learning_rate": 1.6175500630891127e-06,
+      "loss": 1.1309,
+      "step": 4910
+    },
+    {
+      "epoch": 0.9448489403276717,
+      "grad_norm": 0.27925966909469846,
+      "learning_rate": 1.6064053275275714e-06,
+      "loss": 1.0902,
+      "step": 4911
+    },
+    {
+      "epoch": 0.9450413347362092,
+      "grad_norm": 0.3534517232238773,
+      "learning_rate": 1.5952988069044106e-06,
+      "loss": 1.0863,
+      "step": 4912
+    },
+    {
+      "epoch": 0.9452337291447467,
+      "grad_norm": 0.2966449356315687,
+      "learning_rate": 1.5842305055332795e-06,
+      "loss": 0.9766,
+      "step": 4913
+    },
+    {
+      "epoch": 0.9454261235532843,
+      "grad_norm": 0.2932166784805485,
+      "learning_rate": 1.573200427712973e-06,
+      "loss": 1.0655,
+      "step": 4914
+    },
+    {
+      "epoch": 0.9456185179618217,
+      "grad_norm": 0.35760784565349485,
+      "learning_rate": 1.562208577727442e-06,
+      "loss": 1.0548,
+      "step": 4915
+    },
+    {
+      "epoch": 0.9458109123703592,
+      "grad_norm": 0.3592867540870145,
+      "learning_rate": 1.5512549598458048e-06,
+      "loss": 1.0426,
+      "step": 4916
+    },
+    {
+      "epoch": 0.9460033067788968,
+      "grad_norm": 0.276766046699416,
+      "learning_rate": 1.540339578322314e-06,
+      "loss": 1.0428,
+      "step": 4917
+    },
+    {
+      "epoch": 0.9461957011874342,
+      "grad_norm": 0.39779999778007813,
+      "learning_rate": 1.5294624373963895e-06,
+      "loss": 0.9154,
+      "step": 4918
+    },
+    {
+      "epoch": 0.9463880955959717,
+      "grad_norm": 0.31654156787619875,
+      "learning_rate": 1.5186235412925742e-06,
+      "loss": 1.0365,
+      "step": 4919
+    },
+    {
+      "epoch": 0.9465804900045093,
+      "grad_norm": 0.3234701154471859,
+      "learning_rate": 1.5078228942205674e-06,
+      "loss": 0.9417,
+      "step": 4920
+    },
+    {
+      "epoch": 0.9467728844130467,
+      "grad_norm": 0.2823261131733285,
+      "learning_rate": 1.497060500375236e-06,
+      "loss": 1.0225,
+      "step": 4921
+    },
+    {
+      "epoch": 0.9469652788215842,
+      "grad_norm": 0.325333468865082,
+      "learning_rate": 1.4863363639365357e-06,
+      "loss": 1.0717,
+      "step": 4922
+    },
+    {
+      "epoch": 0.9471576732301218,
+      "grad_norm": 0.2910093377836334,
+      "learning_rate": 1.4756504890696466e-06,
+      "loss": 1.0304,
+      "step": 4923
+    },
+    {
+      "epoch": 0.9473500676386593,
+      "grad_norm": 0.3535415888828865,
+      "learning_rate": 1.4650028799247928e-06,
+      "loss": 1.0075,
+      "step": 4924
+    },
+    {
+      "epoch": 0.9475424620471967,
+      "grad_norm": 0.36128552754942406,
+      "learning_rate": 1.454393540637411e-06,
+      "loss": 1.0177,
+      "step": 4925
+    },
+    {
+      "epoch": 0.9477348564557343,
+      "grad_norm": 0.336278605392095,
+      "learning_rate": 1.4438224753280382e-06,
+      "loss": 0.9692,
+      "step": 4926
+    },
+    {
+      "epoch": 0.9479272508642718,
+      "grad_norm": 0.36851277106489816,
+      "learning_rate": 1.4332896881023461e-06,
+      "loss": 1.0513,
+      "step": 4927
+    },
+    {
+      "epoch": 0.9481196452728092,
+      "grad_norm": 0.34292967990781215,
+      "learning_rate": 1.422795183051151e-06,
+      "loss": 1.1092,
+      "step": 4928
+    },
+    {
+      "epoch": 0.9483120396813468,
+      "grad_norm": 0.3954331229329331,
+      "learning_rate": 1.4123389642504148e-06,
+      "loss": 1.0335,
+      "step": 4929
+    },
+    {
+      "epoch": 0.9485044340898843,
+      "grad_norm": 0.34196522169793797,
+      "learning_rate": 1.401921035761189e-06,
+      "loss": 1.054,
+      "step": 4930
+    },
+    {
+      "epoch": 0.9486968284984217,
+      "grad_norm": 0.3558813609551621,
+      "learning_rate": 1.3915414016296924e-06,
+      "loss": 0.9845,
+      "step": 4931
+    },
+    {
+      "epoch": 0.9488892229069593,
+      "grad_norm": 0.3317569179637117,
+      "learning_rate": 1.381200065887256e-06,
+      "loss": 1.0452,
+      "step": 4932
+    },
+    {
+      "epoch": 0.9490816173154968,
+      "grad_norm": 0.3213494368941408,
+      "learning_rate": 1.3708970325503223e-06,
+      "loss": 1.0054,
+      "step": 4933
+    },
+    {
+      "epoch": 0.9492740117240343,
+      "grad_norm": 0.2872465885520479,
+      "learning_rate": 1.3606323056204795e-06,
+      "loss": 1.0079,
+      "step": 4934
+    },
+    {
+      "epoch": 0.9494664061325718,
+      "grad_norm": 0.29249274691955834,
+      "learning_rate": 1.3504058890844273e-06,
+      "loss": 1.0959,
+      "step": 4935
+    },
+    {
+      "epoch": 0.9496588005411093,
+      "grad_norm": 0.28275085751362944,
+      "learning_rate": 1.3402177869139886e-06,
+      "loss": 1.0508,
+      "step": 4936
+    },
+    {
+      "epoch": 0.9498511949496468,
+      "grad_norm": 0.3054640897212883,
+      "learning_rate": 1.3300680030661095e-06,
+      "loss": 1.0043,
+      "step": 4937
+    },
+    {
+      "epoch": 0.9500435893581842,
+      "grad_norm": 0.30570198243457886,
+      "learning_rate": 1.3199565414828364e-06,
+      "loss": 1.0938,
+      "step": 4938
+    },
+    {
+      "epoch": 0.9502359837667218,
+      "grad_norm": 0.37211481598858215,
+      "learning_rate": 1.309883406091361e-06,
+      "loss": 0.9454,
+      "step": 4939
+    },
+    {
+      "epoch": 0.9504283781752593,
+      "grad_norm": 0.3242052176726218,
+      "learning_rate": 1.2998486008039545e-06,
+      "loss": 1.0366,
+      "step": 4940
+    },
+    {
+      "epoch": 0.9506207725837967,
+      "grad_norm": 0.2985978591366246,
+      "learning_rate": 1.2898521295180322e-06,
+      "loss": 1.0992,
+      "step": 4941
+    },
+    {
+      "epoch": 0.9508131669923343,
+      "grad_norm": 0.29500611801476373,
+      "learning_rate": 1.2798939961161215e-06,
+      "loss": 1.0179,
+      "step": 4942
+    },
+    {
+      "epoch": 0.9510055614008718,
+      "grad_norm": 0.2729088391853405,
+      "learning_rate": 1.269974204465818e-06,
+      "loss": 0.9948,
+      "step": 4943
+    },
+    {
+      "epoch": 0.9511979558094092,
+      "grad_norm": 0.3050495645978466,
+      "learning_rate": 1.260092758419862e-06,
+      "loss": 1.0503,
+      "step": 4944
+    },
+    {
+      "epoch": 0.9513903502179468,
+      "grad_norm": 0.3945562672310988,
+      "learning_rate": 1.2502496618161164e-06,
+      "loss": 0.9892,
+      "step": 4945
+    },
+    {
+      "epoch": 0.9515827446264843,
+      "grad_norm": 0.34853183191628206,
+      "learning_rate": 1.2404449184774902e-06,
+      "loss": 1.1257,
+      "step": 4946
+    },
+    {
+      "epoch": 0.9517751390350218,
+      "grad_norm": 0.3467851955053795,
+      "learning_rate": 1.2306785322120595e-06,
+      "loss": 1.0521,
+      "step": 4947
+    },
+    {
+      "epoch": 0.9519675334435593,
+      "grad_norm": 0.3058773332816911,
+      "learning_rate": 1.220950506812968e-06,
+      "loss": 1.0327,
+      "step": 4948
+    },
+    {
+      "epoch": 0.9521599278520968,
+      "grad_norm": 0.30765152112549815,
+      "learning_rate": 1.2112608460584707e-06,
+      "loss": 1.0455,
+      "step": 4949
+    },
+    {
+      "epoch": 0.9523523222606343,
+      "grad_norm": 0.5386394603532704,
+      "learning_rate": 1.2016095537119243e-06,
+      "loss": 1.0611,
+      "step": 4950
+    },
+    {
+      "epoch": 0.9525447166691718,
+      "grad_norm": 0.366068053972542,
+      "learning_rate": 1.1919966335217636e-06,
+      "loss": 0.8987,
+      "step": 4951
+    },
+    {
+      "epoch": 0.9527371110777093,
+      "grad_norm": 0.3007287526569373,
+      "learning_rate": 1.1824220892215464e-06,
+      "loss": 1.061,
+      "step": 4952
+    },
+    {
+      "epoch": 0.9529295054862468,
+      "grad_norm": 0.38681754054034423,
+      "learning_rate": 1.1728859245299205e-06,
+      "loss": 0.9447,
+      "step": 4953
+    },
+    {
+      "epoch": 0.9531218998947844,
+      "grad_norm": 0.295780045648606,
+      "learning_rate": 1.1633881431506122e-06,
+      "loss": 0.9852,
+      "step": 4954
+    },
+    {
+      "epoch": 0.9533142943033218,
+      "grad_norm": 0.3220803327742914,
+      "learning_rate": 1.1539287487724593e-06,
+      "loss": 1.0308,
+      "step": 4955
+    },
+    {
+      "epoch": 0.9535066887118593,
+      "grad_norm": 0.3312860254546172,
+      "learning_rate": 1.1445077450693786e-06,
+      "loss": 1.0285,
+      "step": 4956
+    },
+    {
+      "epoch": 0.9536990831203969,
+      "grad_norm": 0.42106776361564996,
+      "learning_rate": 1.1351251357003655e-06,
+      "loss": 0.9361,
+      "step": 4957
+    },
+    {
+      "epoch": 0.9538914775289343,
+      "grad_norm": 0.32889789776165806,
+      "learning_rate": 1.1257809243095386e-06,
+      "loss": 0.9823,
+      "step": 4958
+    },
+    {
+      "epoch": 0.9540838719374718,
+      "grad_norm": 0.39465940212204376,
+      "learning_rate": 1.1164751145260722e-06,
+      "loss": 1.0534,
+      "step": 4959
+    },
+    {
+      "epoch": 0.9542762663460094,
+      "grad_norm": 0.3986728447282168,
+      "learning_rate": 1.1072077099642418e-06,
+      "loss": 1.05,
+      "step": 4960
+    },
+    {
+      "epoch": 0.9544686607545468,
+      "grad_norm": 0.32610161967454426,
+      "learning_rate": 1.097978714223391e-06,
+      "loss": 0.9867,
+      "step": 4961
+    },
+    {
+      "epoch": 0.9546610551630843,
+      "grad_norm": 0.3264805381065678,
+      "learning_rate": 1.0887881308879633e-06,
+      "loss": 1.0906,
+      "step": 4962
+    },
+    {
+      "epoch": 0.9548534495716218,
+      "grad_norm": 0.31076733448210864,
+      "learning_rate": 1.07963596352747e-06,
+      "loss": 1.075,
+      "step": 4963
+    },
+    {
+      "epoch": 0.9550458439801593,
+      "grad_norm": 0.3190122657108563,
+      "learning_rate": 1.0705222156965012e-06,
+      "loss": 1.0424,
+      "step": 4964
+    },
+    {
+      "epoch": 0.9552382383886968,
+      "grad_norm": 0.30517956610751173,
+      "learning_rate": 1.0614468909347474e-06,
+      "loss": 0.9354,
+      "step": 4965
+    },
+    {
+      "epoch": 0.9554306327972343,
+      "grad_norm": 0.3692653648118493,
+      "learning_rate": 1.0524099927669563e-06,
+      "loss": 1.0236,
+      "step": 4966
+    },
+    {
+      "epoch": 0.9556230272057719,
+      "grad_norm": 0.296500174680699,
+      "learning_rate": 1.043411524702942e-06,
+      "loss": 1.0736,
+      "step": 4967
+    },
+    {
+      "epoch": 0.9558154216143093,
+      "grad_norm": 0.32263885549526844,
+      "learning_rate": 1.03445149023762e-06,
+      "loss": 1.0013,
+      "step": 4968
+    },
+    {
+      "epoch": 0.9560078160228468,
+      "grad_norm": 0.3705868130695738,
+      "learning_rate": 1.0255298928509627e-06,
+      "loss": 1.109,
+      "step": 4969
+    },
+    {
+      "epoch": 0.9562002104313844,
+      "grad_norm": 0.3052829339008226,
+      "learning_rate": 1.016646736007998e-06,
+      "loss": 1.0115,
+      "step": 4970
+    },
+    {
+      "epoch": 0.9563926048399218,
+      "grad_norm": 0.3362393744463195,
+      "learning_rate": 1.0078020231588768e-06,
+      "loss": 0.9704,
+      "step": 4971
+    },
+    {
+      "epoch": 0.9565849992484593,
+      "grad_norm": 0.31742447278834984,
+      "learning_rate": 9.989957577387521e-07,
+      "loss": 1.0067,
+      "step": 4972
+    },
+    {
+      "epoch": 0.9567773936569969,
+      "grad_norm": 0.3048006284582907,
+      "learning_rate": 9.902279431678873e-07,
+      "loss": 1.0663,
+      "step": 4973
+    },
+    {
+      "epoch": 0.9569697880655343,
+      "grad_norm": 0.40837297051339433,
+      "learning_rate": 9.814985828516033e-07,
+      "loss": 1.0326,
+      "step": 4974
+    },
+    {
+      "epoch": 0.9571621824740718,
+      "grad_norm": 0.28414315019383085,
+      "learning_rate": 9.728076801802655e-07,
+      "loss": 1.0578,
+      "step": 4975
+    },
+    {
+      "epoch": 0.9573545768826094,
+      "grad_norm": 0.3394737525573108,
+      "learning_rate": 9.641552385293518e-07,
+      "loss": 0.9949,
+      "step": 4976
+    },
+    {
+      "epoch": 0.9575469712911469,
+      "grad_norm": 0.37620313313250575,
+      "learning_rate": 9.555412612593518e-07,
+      "loss": 1.0158,
+      "step": 4977
+    },
+    {
+      "epoch": 0.9577393656996843,
+      "grad_norm": 0.3284786956023174,
+      "learning_rate": 9.469657517158226e-07,
+      "loss": 1.028,
+      "step": 4978
+    },
+    {
+      "epoch": 0.9579317601082219,
+      "grad_norm": 0.32029102021052164,
+      "learning_rate": 9.384287132294223e-07,
+      "loss": 0.9752,
+      "step": 4979
+    },
+    {
+      "epoch": 0.9581241545167594,
+      "grad_norm": 0.2817345331905995,
+      "learning_rate": 9.299301491158207e-07,
+      "loss": 1.0367,
+      "step": 4980
+    },
+    {
+      "epoch": 0.9583165489252968,
+      "grad_norm": 0.36074091813494247,
+      "learning_rate": 9.214700626757666e-07,
+      "loss": 1.034,
+      "step": 4981
+    },
+    {
+      "epoch": 0.9585089433338344,
+      "grad_norm": 0.3045402123270288,
+      "learning_rate": 9.130484571950537e-07,
+      "loss": 1.0372,
+      "step": 4982
+    },
+    {
+      "epoch": 0.9587013377423719,
+      "grad_norm": 0.2874965714266688,
+      "learning_rate": 9.046653359445323e-07,
+      "loss": 1.0093,
+      "step": 4983
+    },
+    {
+      "epoch": 0.9588937321509093,
+      "grad_norm": 0.28719417590798496,
+      "learning_rate": 8.963207021801423e-07,
+      "loss": 1.0915,
+      "step": 4984
+    },
+    {
+      "epoch": 0.9590861265594469,
+      "grad_norm": 0.3894590181618534,
+      "learning_rate": 8.880145591428024e-07,
+      "loss": 0.9573,
+      "step": 4985
+    },
+    {
+      "epoch": 0.9592785209679844,
+      "grad_norm": 0.36760468451037237,
+      "learning_rate": 8.797469100585431e-07,
+      "loss": 0.9966,
+      "step": 4986
+    },
+    {
+      "epoch": 0.9594709153765218,
+      "grad_norm": 0.34228324657728326,
+      "learning_rate": 8.715177581384182e-07,
+      "loss": 1.0008,
+      "step": 4987
+    },
+    {
+      "epoch": 0.9596633097850594,
+      "grad_norm": 0.5079965020003488,
+      "learning_rate": 8.633271065785486e-07,
+      "loss": 1.0156,
+      "step": 4988
+    },
+    {
+      "epoch": 0.9598557041935969,
+      "grad_norm": 0.32560821376502563,
+      "learning_rate": 8.551749585600677e-07,
+      "loss": 1.0717,
+      "step": 4989
+    },
+    {
+      "epoch": 0.9600480986021344,
+      "grad_norm": 0.36258177524586555,
+      "learning_rate": 8.470613172491981e-07,
+      "loss": 1.1304,
+      "step": 4990
+    },
+    {
+      "epoch": 0.9602404930106718,
+      "grad_norm": 0.34090086472935294,
+      "learning_rate": 8.389861857971748e-07,
+      "loss": 1.1374,
+      "step": 4991
+    },
+    {
+      "epoch": 0.9604328874192094,
+      "grad_norm": 0.41406228945182855,
+      "learning_rate": 8.309495673402779e-07,
+      "loss": 1.13,
+      "step": 4992
+    },
+    {
+      "epoch": 0.9606252818277469,
+      "grad_norm": 0.31467940982811504,
+      "learning_rate": 8.229514649998437e-07,
+      "loss": 1.0519,
+      "step": 4993
+    },
+    {
+      "epoch": 0.9608176762362843,
+      "grad_norm": 0.3936063919661145,
+      "learning_rate": 8.149918818822433e-07,
+      "loss": 0.9858,
+      "step": 4994
+    },
+    {
+      "epoch": 0.9610100706448219,
+      "grad_norm": 0.30032133813124073,
+      "learning_rate": 8.070708210788924e-07,
+      "loss": 1.0487,
+      "step": 4995
+    },
+    {
+      "epoch": 0.9612024650533594,
+      "grad_norm": 0.2545971837000099,
+      "learning_rate": 7.991882856662302e-07,
+      "loss": 1.0984,
+      "step": 4996
+    },
+    {
+      "epoch": 0.9613948594618968,
+      "grad_norm": 0.3253419585783581,
+      "learning_rate": 7.913442787057523e-07,
+      "loss": 1.0599,
+      "step": 4997
+    },
+    {
+      "epoch": 0.9615872538704344,
+      "grad_norm": 0.37692848928140704,
+      "learning_rate": 7.835388032439661e-07,
+      "loss": 0.9417,
+      "step": 4998
+    },
+    {
+      "epoch": 0.9617796482789719,
+      "grad_norm": 0.3228267566390857,
+      "learning_rate": 7.757718623124466e-07,
+      "loss": 0.9511,
+      "step": 4999
+    },
+    {
+      "epoch": 0.9619720426875094,
+      "grad_norm": 0.3952488429491064,
+      "learning_rate": 7.680434589277697e-07,
+      "loss": 1.0789,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9621644370960469,
+      "grad_norm": 0.30622710527485747,
+      "learning_rate": 7.603535960915675e-07,
+      "loss": 1.0395,
+      "step": 5001
+    },
+    {
+      "epoch": 0.9623568315045844,
+      "grad_norm": 0.38342706353059464,
+      "learning_rate": 7.527022767904956e-07,
+      "loss": 1.0077,
+      "step": 5002
+    },
+    {
+      "epoch": 0.9625492259131219,
+      "grad_norm": 0.3395337716599965,
+      "learning_rate": 7.450895039962214e-07,
+      "loss": 1.0927,
+      "step": 5003
+    },
+    {
+      "epoch": 0.9627416203216594,
+      "grad_norm": 0.36232484857569147,
+      "learning_rate": 7.375152806654683e-07,
+      "loss": 1.0387,
+      "step": 5004
+    },
+    {
+      "epoch": 0.9629340147301969,
+      "grad_norm": 0.36656435099597173,
+      "learning_rate": 7.299796097399947e-07,
+      "loss": 1.067,
+      "step": 5005
+    },
+    {
+      "epoch": 0.9631264091387344,
+      "grad_norm": 0.29315110436382213,
+      "learning_rate": 7.224824941465369e-07,
+      "loss": 1.0228,
+      "step": 5006
+    },
+    {
+      "epoch": 0.963318803547272,
+      "grad_norm": 0.43515984259373935,
+      "learning_rate": 7.150239367969102e-07,
+      "loss": 0.9789,
+      "step": 5007
+    },
+    {
+      "epoch": 0.9635111979558094,
+      "grad_norm": 0.3110854369609877,
+      "learning_rate": 7.076039405879309e-07,
+      "loss": 1.0924,
+      "step": 5008
+    },
+    {
+      "epoch": 0.9637035923643469,
+      "grad_norm": 0.28387957421165444,
+      "learning_rate": 7.002225084014269e-07,
+      "loss": 1.0357,
+      "step": 5009
+    },
+    {
+      "epoch": 0.9638959867728845,
+      "grad_norm": 0.29887345785884456,
+      "learning_rate": 6.928796431042717e-07,
+      "loss": 1.0634,
+      "step": 5010
+    },
+    {
+      "epoch": 0.9640883811814219,
+      "grad_norm": 0.3141623923742202,
+      "learning_rate": 6.855753475483506e-07,
+      "loss": 0.9347,
+      "step": 5011
+    },
+    {
+      "epoch": 0.9642807755899594,
+      "grad_norm": 0.33119300666646084,
+      "learning_rate": 6.783096245705611e-07,
+      "loss": 1.0173,
+      "step": 5012
+    },
+    {
+      "epoch": 0.964473169998497,
+      "grad_norm": 0.4166242696396946,
+      "learning_rate": 6.710824769928348e-07,
+      "loss": 1.0347,
+      "step": 5013
+    },
+    {
+      "epoch": 0.9646655644070344,
+      "grad_norm": 0.33297074336934146,
+      "learning_rate": 6.63893907622104e-07,
+      "loss": 1.1196,
+      "step": 5014
+    },
+    {
+      "epoch": 0.9648579588155719,
+      "grad_norm": 0.29337790425361476,
+      "learning_rate": 6.567439192503245e-07,
+      "loss": 0.9234,
+      "step": 5015
+    },
+    {
+      "epoch": 0.9650503532241094,
+      "grad_norm": 0.42082504983687435,
+      "learning_rate": 6.496325146544746e-07,
+      "loss": 1.0825,
+      "step": 5016
+    },
+    {
+      "epoch": 0.965242747632647,
+      "grad_norm": 0.3632439017386509,
+      "learning_rate": 6.425596965965452e-07,
+      "loss": 1.0366,
+      "step": 5017
+    },
+    {
+      "epoch": 0.9654351420411844,
+      "grad_norm": 0.3082693157834947,
+      "learning_rate": 6.35525467823539e-07,
+      "loss": 1.1006,
+      "step": 5018
+    },
+    {
+      "epoch": 0.9656275364497219,
+      "grad_norm": 0.31182322007077123,
+      "learning_rate": 6.285298310674703e-07,
+      "loss": 1.0385,
+      "step": 5019
+    },
+    {
+      "epoch": 0.9658199308582595,
+      "grad_norm": 0.3423418253615423,
+      "learning_rate": 6.215727890453438e-07,
+      "loss": 1.0634,
+      "step": 5020
+    },
+    {
+      "epoch": 0.9660123252667969,
+      "grad_norm": 0.31689270309642814,
+      "learning_rate": 6.14654344459209e-07,
+      "loss": 1.0528,
+      "step": 5021
+    },
+    {
+      "epoch": 0.9662047196753344,
+      "grad_norm": 0.40301768880378613,
+      "learning_rate": 6.077744999961165e-07,
+      "loss": 1.0881,
+      "step": 5022
+    },
+    {
+      "epoch": 0.966397114083872,
+      "grad_norm": 0.31492110205436796,
+      "learning_rate": 6.009332583281069e-07,
+      "loss": 1.0787,
+      "step": 5023
+    },
+    {
+      "epoch": 0.9665895084924094,
+      "grad_norm": 0.4023107389473161,
+      "learning_rate": 5.941306221122545e-07,
+      "loss": 1.075,
+      "step": 5024
+    },
+    {
+      "epoch": 0.9667819029009469,
+      "grad_norm": 0.30721715303343167,
+      "learning_rate": 5.873665939906015e-07,
+      "loss": 1.0737,
+      "step": 5025
+    },
+    {
+      "epoch": 0.9669742973094845,
+      "grad_norm": 0.29014115403983065,
+      "learning_rate": 5.806411765902353e-07,
+      "loss": 0.9845,
+      "step": 5026
+    },
+    {
+      "epoch": 0.9671666917180219,
+      "grad_norm": 0.26137856671050164,
+      "learning_rate": 5.739543725232332e-07,
+      "loss": 1.0461,
+      "step": 5027
+    },
+    {
+      "epoch": 0.9673590861265594,
+      "grad_norm": 0.2759912417552023,
+      "learning_rate": 5.673061843866623e-07,
+      "loss": 1.0338,
+      "step": 5028
+    },
+    {
+      "epoch": 0.967551480535097,
+      "grad_norm": 0.2845290218412168,
+      "learning_rate": 5.606966147626124e-07,
+      "loss": 1.1266,
+      "step": 5029
+    },
+    {
+      "epoch": 0.9677438749436345,
+      "grad_norm": 0.3496918140083736,
+      "learning_rate": 5.541256662181526e-07,
+      "loss": 1.0113,
+      "step": 5030
+    },
+    {
+      "epoch": 0.9679362693521719,
+      "grad_norm": 0.32659747761414937,
+      "learning_rate": 5.475933413053636e-07,
+      "loss": 1.0333,
+      "step": 5031
+    },
+    {
+      "epoch": 0.9681286637607095,
+      "grad_norm": 0.3675553389359608,
+      "learning_rate": 5.410996425613379e-07,
+      "loss": 1.0833,
+      "step": 5032
+    },
+    {
+      "epoch": 0.968321058169247,
+      "grad_norm": 0.4808990978514393,
+      "learning_rate": 5.346445725081473e-07,
+      "loss": 0.914,
+      "step": 5033
+    },
+    {
+      "epoch": 0.9685134525777844,
+      "grad_norm": 0.35795587440588744,
+      "learning_rate": 5.282281336528638e-07,
+      "loss": 0.9903,
+      "step": 5034
+    },
+    {
+      "epoch": 0.968705846986322,
+      "grad_norm": 0.35282334183642,
+      "learning_rate": 5.218503284875609e-07,
+      "loss": 1.0208,
+      "step": 5035
+    },
+    {
+      "epoch": 0.9688982413948595,
+      "grad_norm": 0.2902959575167372,
+      "learning_rate": 5.155111594893014e-07,
+      "loss": 0.9751,
+      "step": 5036
+    },
+    {
+      "epoch": 0.9690906358033969,
+      "grad_norm": 0.3599339035019339,
+      "learning_rate": 5.092106291201604e-07,
+      "loss": 1.099,
+      "step": 5037
+    },
+    {
+      "epoch": 0.9692830302119345,
+      "grad_norm": 0.3438592152853123,
+      "learning_rate": 5.029487398271692e-07,
+      "loss": 1.0155,
+      "step": 5038
+    },
+    {
+      "epoch": 0.969475424620472,
+      "grad_norm": 0.2727622230106225,
+      "learning_rate": 4.967254940423715e-07,
+      "loss": 1.0899,
+      "step": 5039
+    },
+    {
+      "epoch": 0.9696678190290094,
+      "grad_norm": 0.3187102551863971,
+      "learning_rate": 4.905408941828338e-07,
+      "loss": 1.0093,
+      "step": 5040
+    },
+    {
+      "epoch": 0.969860213437547,
+      "grad_norm": 0.27321792094726255,
+      "learning_rate": 4.843949426505567e-07,
+      "loss": 0.9545,
+      "step": 5041
+    },
+    {
+      "epoch": 0.9700526078460845,
+      "grad_norm": 0.45757931008923125,
+      "learning_rate": 4.782876418325755e-07,
+      "loss": 1.0078,
+      "step": 5042
+    },
+    {
+      "epoch": 0.970245002254622,
+      "grad_norm": 0.33484521601404077,
+      "learning_rate": 4.7221899410087033e-07,
+      "loss": 1.0748,
+      "step": 5043
+    },
+    {
+      "epoch": 0.9704373966631594,
+      "grad_norm": 0.2762389527650955,
+      "learning_rate": 4.6618900181245595e-07,
+      "loss": 1.1174,
+      "step": 5044
+    },
+    {
+      "epoch": 0.970629791071697,
+      "grad_norm": 0.3223238593624723,
+      "learning_rate": 4.6019766730930336e-07,
+      "loss": 1.032,
+      "step": 5045
+    },
+    {
+      "epoch": 0.9708221854802345,
+      "grad_norm": 0.33345687239029953,
+      "learning_rate": 4.5424499291838454e-07,
+      "loss": 1.0953,
+      "step": 5046
+    },
+    {
+      "epoch": 0.9710145798887719,
+      "grad_norm": 0.30811885961325575,
+      "learning_rate": 4.4833098095165004e-07,
+      "loss": 1.0191,
+      "step": 5047
+    },
+    {
+      "epoch": 0.9712069742973095,
+      "grad_norm": 0.3169431005266869,
+      "learning_rate": 4.4245563370601815e-07,
+      "loss": 1.0009,
+      "step": 5048
+    },
+    {
+      "epoch": 0.971399368705847,
+      "grad_norm": 0.31354157329219307,
+      "learning_rate": 4.366189534634191e-07,
+      "loss": 0.9344,
+      "step": 5049
+    },
+    {
+      "epoch": 0.9715917631143844,
+      "grad_norm": 0.31487998402505285,
+      "learning_rate": 4.308209424907506e-07,
+      "loss": 1.1079,
+      "step": 5050
+    },
+    {
+      "epoch": 0.971784157522922,
+      "grad_norm": 0.28691385148451876,
+      "learning_rate": 4.250616030399002e-07,
+      "loss": 1.083,
+      "step": 5051
+    },
+    {
+      "epoch": 0.9719765519314595,
+      "grad_norm": 0.3350178958167011,
+      "learning_rate": 4.1934093734771195e-07,
+      "loss": 0.9912,
+      "step": 5052
+    },
+    {
+      "epoch": 0.972168946339997,
+      "grad_norm": 0.37849520683837934,
+      "learning_rate": 4.1365894763604197e-07,
+      "loss": 0.9879,
+      "step": 5053
+    },
+    {
+      "epoch": 0.9723613407485345,
+      "grad_norm": 0.37069367516727025,
+      "learning_rate": 4.080156361117027e-07,
+      "loss": 1.0408,
+      "step": 5054
+    },
+    {
+      "epoch": 0.972553735157072,
+      "grad_norm": 0.3137751182780175,
+      "learning_rate": 4.024110049664853e-07,
+      "loss": 1.0364,
+      "step": 5055
+    },
+    {
+      "epoch": 0.9727461295656095,
+      "grad_norm": 0.29747626355891726,
+      "learning_rate": 3.9684505637718194e-07,
+      "loss": 1.0502,
+      "step": 5056
+    },
+    {
+      "epoch": 0.972938523974147,
+      "grad_norm": 0.3152025452518086,
+      "learning_rate": 3.913177925055189e-07,
+      "loss": 1.07,
+      "step": 5057
+    },
+    {
+      "epoch": 0.9731309183826845,
+      "grad_norm": 0.33432316652220423,
+      "learning_rate": 3.858292154982457e-07,
+      "loss": 0.9501,
+      "step": 5058
+    },
+    {
+      "epoch": 0.973323312791222,
+      "grad_norm": 0.310274214150185,
+      "learning_rate": 3.8037932748704597e-07,
+      "loss": 1.0468,
+      "step": 5059
+    },
+    {
+      "epoch": 0.9735157071997596,
+      "grad_norm": 0.27870253609330775,
+      "learning_rate": 3.7496813058859326e-07,
+      "loss": 1.0257,
+      "step": 5060
+    },
+    {
+      "epoch": 0.973708101608297,
+      "grad_norm": 0.41910230664417786,
+      "learning_rate": 3.6959562690455087e-07,
+      "loss": 1.1828,
+      "step": 5061
+    },
+    {
+      "epoch": 0.9739004960168345,
+      "grad_norm": 0.303718678540597,
+      "learning_rate": 3.642618185215163e-07,
+      "loss": 1.0768,
+      "step": 5062
+    },
+    {
+      "epoch": 0.9740928904253721,
+      "grad_norm": 0.3351723168140292,
+      "learning_rate": 3.5896670751109916e-07,
+      "loss": 1.1378,
+      "step": 5063
+    },
+    {
+      "epoch": 0.9742852848339095,
+      "grad_norm": 0.28292247744285975,
+      "learning_rate": 3.537102959298322e-07,
+      "loss": 0.9996,
+      "step": 5064
+    },
+    {
+      "epoch": 0.974477679242447,
+      "grad_norm": 0.29316251503126217,
+      "learning_rate": 3.484925858192822e-07,
+      "loss": 0.9714,
+      "step": 5065
+    },
+    {
+      "epoch": 0.9746700736509846,
+      "grad_norm": 0.3365946720449322,
+      "learning_rate": 3.4331357920591724e-07,
+      "loss": 1.0325,
+      "step": 5066
+    },
+    {
+      "epoch": 0.974862468059522,
+      "grad_norm": 0.26037443573801905,
+      "learning_rate": 3.38173278101217e-07,
+      "loss": 0.9809,
+      "step": 5067
+    },
+    {
+      "epoch": 0.9750548624680595,
+      "grad_norm": 0.3373501754356196,
+      "learning_rate": 3.3307168450160685e-07,
+      "loss": 1.0276,
+      "step": 5068
+    },
+    {
+      "epoch": 0.975247256876597,
+      "grad_norm": 0.34352897012496336,
+      "learning_rate": 3.280088003885018e-07,
+      "loss": 1.0656,
+      "step": 5069
+    },
+    {
+      "epoch": 0.9754396512851345,
+      "grad_norm": 0.42857668201851623,
+      "learning_rate": 3.2298462772825113e-07,
+      "loss": 0.9532,
+      "step": 5070
+    },
+    {
+      "epoch": 0.975632045693672,
+      "grad_norm": 0.42496313488765447,
+      "learning_rate": 3.17999168472205e-07,
+      "loss": 0.9518,
+      "step": 5071
+    },
+    {
+      "epoch": 0.9758244401022095,
+      "grad_norm": 0.2899132290117504,
+      "learning_rate": 3.130524245566369e-07,
+      "loss": 0.972,
+      "step": 5072
+    },
+    {
+      "epoch": 0.9760168345107471,
+      "grad_norm": 0.3431746855107751,
+      "learning_rate": 3.081443979028098e-07,
+      "loss": 1.0186,
+      "step": 5073
+    },
+    {
+      "epoch": 0.9762092289192845,
+      "grad_norm": 0.2804886758532411,
+      "learning_rate": 3.032750904169546e-07,
+      "loss": 1.0124,
+      "step": 5074
+    },
+    {
+      "epoch": 0.976401623327822,
+      "grad_norm": 0.3399834976319933,
+      "learning_rate": 2.9844450399024726e-07,
+      "loss": 1.1384,
+      "step": 5075
+    },
+    {
+      "epoch": 0.9765940177363596,
+      "grad_norm": 0.2830378097148043,
+      "learning_rate": 2.9365264049884267e-07,
+      "loss": 1.0306,
+      "step": 5076
+    },
+    {
+      "epoch": 0.976786412144897,
+      "grad_norm": 0.3123175290278857,
+      "learning_rate": 2.8889950180382986e-07,
+      "loss": 0.9695,
+      "step": 5077
+    },
+    {
+      "epoch": 0.9769788065534345,
+      "grad_norm": 0.33577519483208257,
+      "learning_rate": 2.8418508975127654e-07,
+      "loss": 1.0596,
+      "step": 5078
+    },
+    {
+      "epoch": 0.9771712009619721,
+      "grad_norm": 0.307192702623424,
+      "learning_rate": 2.795094061722181e-07,
+      "loss": 1.0036,
+      "step": 5079
+    },
+    {
+      "epoch": 0.9773635953705095,
+      "grad_norm": 0.4440475783331159,
+      "learning_rate": 2.7487245288261296e-07,
+      "loss": 1.1049,
+      "step": 5080
+    },
+    {
+      "epoch": 0.977555989779047,
+      "grad_norm": 0.3156220079680518,
+      "learning_rate": 2.702742316834206e-07,
+      "loss": 1.0217,
+      "step": 5081
+    },
+    {
+      "epoch": 0.9777483841875846,
+      "grad_norm": 0.42724938870948,
+      "learning_rate": 2.657147443605457e-07,
+      "loss": 1.095,
+      "step": 5082
+    },
+    {
+      "epoch": 0.977940778596122,
+      "grad_norm": 0.2819177191675086,
+      "learning_rate": 2.6119399268480507e-07,
+      "loss": 0.9718,
+      "step": 5083
+    },
+    {
+      "epoch": 0.9781331730046595,
+      "grad_norm": 0.2867413296030326,
+      "learning_rate": 2.5671197841203865e-07,
+      "loss": 1.017,
+      "step": 5084
+    },
+    {
+      "epoch": 0.9783255674131971,
+      "grad_norm": 0.3131370832798454,
+      "learning_rate": 2.5226870328299845e-07,
+      "loss": 0.9642,
+      "step": 5085
+    },
+    {
+      "epoch": 0.9785179618217346,
+      "grad_norm": 0.284876663165116,
+      "learning_rate": 2.478641690233929e-07,
+      "loss": 0.9948,
+      "step": 5086
+    },
+    {
+      "epoch": 0.978710356230272,
+      "grad_norm": 0.16414187797845742,
+      "learning_rate": 2.434983773439092e-07,
+      "loss": 1.0587,
+      "step": 5087
+    },
+    {
+      "epoch": 0.9789027506388096,
+      "grad_norm": 0.3792642984444722,
+      "learning_rate": 2.3917132994016877e-07,
+      "loss": 1.0413,
+      "step": 5088
+    },
+    {
+      "epoch": 0.9790951450473471,
+      "grad_norm": 0.3166014315663511,
+      "learning_rate": 2.3488302849272724e-07,
+      "loss": 1.0379,
+      "step": 5089
+    },
+    {
+      "epoch": 0.9792875394558845,
+      "grad_norm": 0.2807327494227699,
+      "learning_rate": 2.3063347466713013e-07,
+      "loss": 0.982,
+      "step": 5090
+    },
+    {
+      "epoch": 0.9794799338644221,
+      "grad_norm": 0.29118100069509606,
+      "learning_rate": 2.264226701138461e-07,
+      "loss": 0.95,
+      "step": 5091
+    },
+    {
+      "epoch": 0.9796723282729596,
+      "grad_norm": 0.34729023109741464,
+      "learning_rate": 2.2225061646830025e-07,
+      "loss": 1.0836,
+      "step": 5092
+    },
+    {
+      "epoch": 0.979864722681497,
+      "grad_norm": 0.29257850627314713,
+      "learning_rate": 2.181173153508853e-07,
+      "loss": 1.0233,
+      "step": 5093
+    },
+    {
+      "epoch": 0.9800571170900346,
+      "grad_norm": 0.3959917265023297,
+      "learning_rate": 2.1402276836691714e-07,
+      "loss": 0.992,
+      "step": 5094
+    },
+    {
+      "epoch": 0.9802495114985721,
+      "grad_norm": 0.2590900735118314,
+      "learning_rate": 2.0996697710666812e-07,
+      "loss": 1.097,
+      "step": 5095
+    },
+    {
+      "epoch": 0.9804419059071096,
+      "grad_norm": 0.3654215070810998,
+      "learning_rate": 2.0594994314536707e-07,
+      "loss": 0.9906,
+      "step": 5096
+    },
+    {
+      "epoch": 0.980634300315647,
+      "grad_norm": 0.3604543846432549,
+      "learning_rate": 2.0197166804317712e-07,
+      "loss": 1.0263,
+      "step": 5097
+    },
+    {
+      "epoch": 0.9808266947241846,
+      "grad_norm": 0.3052356194126838,
+      "learning_rate": 1.9803215334522896e-07,
+      "loss": 0.9838,
+      "step": 5098
+    },
+    {
+      "epoch": 0.9810190891327221,
+      "grad_norm": 0.331316123888874,
+      "learning_rate": 1.941314005815653e-07,
+      "loss": 1.1548,
+      "step": 5099
+    },
+    {
+      "epoch": 0.9812114835412595,
+      "grad_norm": 0.29670341958253454,
+      "learning_rate": 1.9026941126721875e-07,
+      "loss": 0.9693,
+      "step": 5100
+    },
+    {
+      "epoch": 0.9814038779497971,
+      "grad_norm": 0.3153939739119349,
+      "learning_rate": 1.8644618690211168e-07,
+      "loss": 0.9661,
+      "step": 5101
+    },
+    {
+      "epoch": 0.9815962723583346,
+      "grad_norm": 0.3161423649297131,
+      "learning_rate": 1.826617289711563e-07,
+      "loss": 0.9762,
+      "step": 5102
+    },
+    {
+      "epoch": 0.981788666766872,
+      "grad_norm": 0.2851548919511647,
+      "learning_rate": 1.7891603894418796e-07,
+      "loss": 1.0781,
+      "step": 5103
+    },
+    {
+      "epoch": 0.9819810611754096,
+      "grad_norm": 0.3695966861691063,
+      "learning_rate": 1.7520911827598742e-07,
+      "loss": 1.0483,
+      "step": 5104
+    },
+    {
+      "epoch": 0.9821734555839471,
+      "grad_norm": 0.3364653546491064,
+      "learning_rate": 1.7154096840629185e-07,
+      "loss": 1.0092,
+      "step": 5105
+    },
+    {
+      "epoch": 0.9823658499924846,
+      "grad_norm": 0.2859723773642625,
+      "learning_rate": 1.679115907597617e-07,
+      "loss": 1.0642,
+      "step": 5106
+    },
+    {
+      "epoch": 0.9825582444010221,
+      "grad_norm": 0.3627135038541458,
+      "learning_rate": 1.643209867460027e-07,
+      "loss": 1.0595,
+      "step": 5107
+    },
+    {
+      "epoch": 0.9827506388095596,
+      "grad_norm": 0.34671507806669843,
+      "learning_rate": 1.6076915775956604e-07,
+      "loss": 1.0372,
+      "step": 5108
+    },
+    {
+      "epoch": 0.9829430332180971,
+      "grad_norm": 0.37906382876933126,
+      "learning_rate": 1.5725610517994815e-07,
+      "loss": 1.0274,
+      "step": 5109
+    },
+    {
+      "epoch": 0.9831354276266346,
+      "grad_norm": 0.3896087181310295,
+      "learning_rate": 1.537818303715688e-07,
+      "loss": 1.0223,
+      "step": 5110
+    },
+    {
+      "epoch": 0.9833278220351721,
+      "grad_norm": 0.31701559870235946,
+      "learning_rate": 1.50346334683793e-07,
+      "loss": 1.0776,
+      "step": 5111
+    },
+    {
+      "epoch": 0.9835202164437096,
+      "grad_norm": 0.33106207329083365,
+      "learning_rate": 1.4694961945093122e-07,
+      "loss": 1.0536,
+      "step": 5112
+    },
+    {
+      "epoch": 0.9837126108522471,
+      "grad_norm": 0.30840765701709627,
+      "learning_rate": 1.4359168599223928e-07,
+      "loss": 0.9884,
+      "step": 5113
+    },
+    {
+      "epoch": 0.9839050052607846,
+      "grad_norm": 0.40430042383446857,
+      "learning_rate": 1.4027253561188502e-07,
+      "loss": 1.0541,
+      "step": 5114
+    },
+    {
+      "epoch": 0.9840973996693221,
+      "grad_norm": 0.3676395713742101,
+      "learning_rate": 1.3699216959899287e-07,
+      "loss": 1.0052,
+      "step": 5115
+    },
+    {
+      "epoch": 0.9842897940778597,
+      "grad_norm": 0.3718158973689816,
+      "learning_rate": 1.3375058922763252e-07,
+      "loss": 1.0406,
+      "step": 5116
+    },
+    {
+      "epoch": 0.9844821884863971,
+      "grad_norm": 0.30102185062051356,
+      "learning_rate": 1.3054779575677466e-07,
+      "loss": 1.0683,
+      "step": 5117
+    },
+    {
+      "epoch": 0.9846745828949346,
+      "grad_norm": 0.30723822686900915,
+      "learning_rate": 1.2738379043035764e-07,
+      "loss": 0.9981,
+      "step": 5118
+    },
+    {
+      "epoch": 0.9848669773034722,
+      "grad_norm": 0.37907539466338597,
+      "learning_rate": 1.2425857447725398e-07,
+      "loss": 1.0811,
+      "step": 5119
+    },
+    {
+      "epoch": 0.9850593717120096,
+      "grad_norm": 0.41688853701192846,
+      "learning_rate": 1.211721491112372e-07,
+      "loss": 1.0408,
+      "step": 5120
+    },
+    {
+      "epoch": 0.9852517661205471,
+      "grad_norm": 0.30608020355653576,
+      "learning_rate": 1.1812451553107062e-07,
+      "loss": 1.0056,
+      "step": 5121
+    },
+    {
+      "epoch": 0.9854441605290847,
+      "grad_norm": 0.3520025563703071,
+      "learning_rate": 1.1511567492038522e-07,
+      "loss": 1.0606,
+      "step": 5122
+    },
+    {
+      "epoch": 0.9856365549376221,
+      "grad_norm": 0.3068441943985922,
+      "learning_rate": 1.1214562844781285e-07,
+      "loss": 1.0366,
+      "step": 5123
+    },
+    {
+      "epoch": 0.9858289493461596,
+      "grad_norm": 0.3249513870562328,
+      "learning_rate": 1.0921437726686411e-07,
+      "loss": 1.0075,
+      "step": 5124
+    },
+    {
+      "epoch": 0.9860213437546971,
+      "grad_norm": 0.3626847659538625,
+      "learning_rate": 1.0632192251601725e-07,
+      "loss": 1.0602,
+      "step": 5125
+    },
+    {
+      "epoch": 0.9862137381632347,
+      "grad_norm": 0.30729084202484547,
+      "learning_rate": 1.0346826531865139e-07,
+      "loss": 0.966,
+      "step": 5126
+    },
+    {
+      "epoch": 0.9864061325717721,
+      "grad_norm": 0.3507531533210151,
+      "learning_rate": 1.006534067831022e-07,
+      "loss": 1.0929,
+      "step": 5127
+    },
+    {
+      "epoch": 0.9865985269803096,
+      "grad_norm": 0.40352441135005895,
+      "learning_rate": 9.78773480026396e-08,
+      "loss": 1.0611,
+      "step": 5128
+    },
+    {
+      "epoch": 0.9867909213888472,
+      "grad_norm": 0.31468960125532663,
+      "learning_rate": 9.514009005543445e-08,
+      "loss": 0.9624,
+      "step": 5129
+    },
+    {
+      "epoch": 0.9869833157973846,
+      "grad_norm": 0.3270197500831794,
+      "learning_rate": 9.244163400462525e-08,
+      "loss": 1.0326,
+      "step": 5130
+    },
+    {
+      "epoch": 0.9871757102059221,
+      "grad_norm": 0.29821584705824833,
+      "learning_rate": 8.978198089824031e-08,
+      "loss": 1.0625,
+      "step": 5131
+    },
+    {
+      "epoch": 0.9873681046144597,
+      "grad_norm": 0.36897523605551197,
+      "learning_rate": 8.716113176927554e-08,
+      "loss": 1.1163,
+      "step": 5132
+    },
+    {
+      "epoch": 0.9875604990229971,
+      "grad_norm": 0.32400594888563705,
+      "learning_rate": 8.457908763562783e-08,
+      "loss": 1.1224,
+      "step": 5133
+    },
+    {
+      "epoch": 0.9877528934315346,
+      "grad_norm": 0.3092625063397082,
+      "learning_rate": 8.203584950013943e-08,
+      "loss": 1.0871,
+      "step": 5134
+    },
+    {
+      "epoch": 0.9879452878400722,
+      "grad_norm": 0.2826061850679332,
+      "learning_rate": 7.953141835057576e-08,
+      "loss": 1.0583,
+      "step": 5135
+    },
+    {
+      "epoch": 0.9881376822486097,
+      "grad_norm": 0.30800872364108206,
+      "learning_rate": 7.706579515962542e-08,
+      "loss": 1.0506,
+      "step": 5136
+    },
+    {
+      "epoch": 0.9883300766571471,
+      "grad_norm": 0.27389570162999394,
+      "learning_rate": 7.463898088490018e-08,
+      "loss": 1.0334,
+      "step": 5137
+    },
+    {
+      "epoch": 0.9885224710656847,
+      "grad_norm": 0.294756176525652,
+      "learning_rate": 7.225097646895718e-08,
+      "loss": 1.022,
+      "step": 5138
+    },
+    {
+      "epoch": 0.9887148654742222,
+      "grad_norm": 0.34231691884217796,
+      "learning_rate": 6.990178283927674e-08,
+      "loss": 0.9789,
+      "step": 5139
+    },
+    {
+      "epoch": 0.9889072598827596,
+      "grad_norm": 0.26817788047365737,
+      "learning_rate": 6.759140090824013e-08,
+      "loss": 1.0641,
+      "step": 5140
+    },
+    {
+      "epoch": 0.9890996542912972,
+      "grad_norm": 0.25716053792488497,
+      "learning_rate": 6.53198315731851e-08,
+      "loss": 1.0427,
+      "step": 5141
+    },
+    {
+      "epoch": 0.9892920486998347,
+      "grad_norm": 0.32935953149064995,
+      "learning_rate": 6.308707571636153e-08,
+      "loss": 1.05,
+      "step": 5142
+    },
+    {
+      "epoch": 0.9894844431083721,
+      "grad_norm": 0.328380784771582,
+      "learning_rate": 6.089313420494236e-08,
+      "loss": 0.9967,
+      "step": 5143
+    },
+    {
+      "epoch": 0.9896768375169097,
+      "grad_norm": 0.37057095491843356,
+      "learning_rate": 5.87380078910349e-08,
+      "loss": 1.0385,
+      "step": 5144
+    },
+    {
+      "epoch": 0.9898692319254472,
+      "grad_norm": 0.3532213545283488,
+      "learning_rate": 5.6621697611658474e-08,
+      "loss": 1.0159,
+      "step": 5145
+    },
+    {
+      "epoch": 0.9900616263339846,
+      "grad_norm": 0.3264479235789096,
+      "learning_rate": 5.454420418877782e-08,
+      "loss": 1.0119,
+      "step": 5146
+    },
+    {
+      "epoch": 0.9902540207425222,
+      "grad_norm": 0.32391000736233166,
+      "learning_rate": 5.2505528429247494e-08,
+      "loss": 1.1292,
+      "step": 5147
+    },
+    {
+      "epoch": 0.9904464151510597,
+      "grad_norm": 0.40009508655518117,
+      "learning_rate": 5.0505671124878584e-08,
+      "loss": 1.1856,
+      "step": 5148
+    },
+    {
+      "epoch": 0.9906388095595972,
+      "grad_norm": 0.27838687609714285,
+      "learning_rate": 4.8544633052383106e-08,
+      "loss": 1.0858,
+      "step": 5149
+    },
+    {
+      "epoch": 0.9908312039681346,
+      "grad_norm": 0.3505331434814559,
+      "learning_rate": 4.662241497341846e-08,
+      "loss": 1.0012,
+      "step": 5150
+    },
+    {
+      "epoch": 0.9910235983766722,
+      "grad_norm": 0.3735866147332912,
+      "learning_rate": 4.473901763454302e-08,
+      "loss": 1.0039,
+      "step": 5151
+    },
+    {
+      "epoch": 0.9912159927852097,
+      "grad_norm": 0.35050639185913424,
+      "learning_rate": 4.289444176724944e-08,
+      "loss": 0.9755,
+      "step": 5152
+    },
+    {
+      "epoch": 0.9914083871937471,
+      "grad_norm": 0.4015214322041215,
+      "learning_rate": 4.108868808794242e-08,
+      "loss": 1.099,
+      "step": 5153
+    },
+    {
+      "epoch": 0.9916007816022847,
+      "grad_norm": 0.32490188732318753,
+      "learning_rate": 3.932175729797205e-08,
+      "loss": 1.0825,
+      "step": 5154
+    },
+    {
+      "epoch": 0.9917931760108222,
+      "grad_norm": 0.25621099687365334,
+      "learning_rate": 3.759365008357829e-08,
+      "loss": 0.9776,
+      "step": 5155
+    },
+    {
+      "epoch": 0.9919855704193596,
+      "grad_norm": 0.30612496529530275,
+      "learning_rate": 3.590436711594647e-08,
+      "loss": 1.0419,
+      "step": 5156
+    },
+    {
+      "epoch": 0.9921779648278972,
+      "grad_norm": 0.27549390219302733,
+      "learning_rate": 3.425390905117398e-08,
+      "loss": 0.9952,
+      "step": 5157
+    },
+    {
+      "epoch": 0.9923703592364347,
+      "grad_norm": 0.25044499402764225,
+      "learning_rate": 3.26422765302814e-08,
+      "loss": 0.9608,
+      "step": 5158
+    },
+    {
+      "epoch": 0.9925627536449722,
+      "grad_norm": 0.3473895293144665,
+      "learning_rate": 3.1069470179201364e-08,
+      "loss": 1.0073,
+      "step": 5159
+    },
+    {
+      "epoch": 0.9927551480535097,
+      "grad_norm": 0.2801863283558563,
+      "learning_rate": 2.9535490608789684e-08,
+      "loss": 0.992,
+      "step": 5160
+    },
+    {
+      "epoch": 0.9929475424620472,
+      "grad_norm": 0.28783072718150426,
+      "learning_rate": 2.8040338414847545e-08,
+      "loss": 1.0381,
+      "step": 5161
+    },
+    {
+      "epoch": 0.9931399368705847,
+      "grad_norm": 0.35476132654503884,
+      "learning_rate": 2.6584014178054895e-08,
+      "loss": 1.0589,
+      "step": 5162
+    },
+    {
+      "epoch": 0.9933323312791222,
+      "grad_norm": 0.36852576995944003,
+      "learning_rate": 2.516651846403706e-08,
+      "loss": 1.0105,
+      "step": 5163
+    },
+    {
+      "epoch": 0.9935247256876597,
+      "grad_norm": 0.32577321750325783,
+      "learning_rate": 2.378785182333143e-08,
+      "loss": 0.9502,
+      "step": 5164
+    },
+    {
+      "epoch": 0.9937171200961972,
+      "grad_norm": 0.28902393769938606,
+      "learning_rate": 2.2448014791398574e-08,
+      "loss": 1.0236,
+      "step": 5165
+    },
+    {
+      "epoch": 0.9939095145047347,
+      "grad_norm": 0.2782511558091309,
+      "learning_rate": 2.1147007888622228e-08,
+      "loss": 1.0375,
+      "step": 5166
+    },
+    {
+      "epoch": 0.9941019089132722,
+      "grad_norm": 0.33963338490072537,
+      "learning_rate": 1.9884831620287093e-08,
+      "loss": 1.0853,
+      "step": 5167
+    },
+    {
+      "epoch": 0.9942943033218097,
+      "grad_norm": 0.26228132825875633,
+      "learning_rate": 1.8661486476612145e-08,
+      "loss": 1.0217,
+      "step": 5168
+    },
+    {
+      "epoch": 0.9944866977303473,
+      "grad_norm": 0.3150664604179331,
+      "learning_rate": 1.747697293272843e-08,
+      "loss": 0.9769,
+      "step": 5169
+    },
+    {
+      "epoch": 0.9946790921388847,
+      "grad_norm": 0.27546539074040105,
+      "learning_rate": 1.633129144870127e-08,
+      "loss": 1.0444,
+      "step": 5170
+    },
+    {
+      "epoch": 0.9948714865474222,
+      "grad_norm": 0.31547501045666676,
+      "learning_rate": 1.522444246947474e-08,
+      "loss": 1.042,
+      "step": 5171
+    },
+    {
+      "epoch": 0.9950638809559598,
+      "grad_norm": 0.2797112311490469,
+      "learning_rate": 1.4156426424960512e-08,
+      "loss": 1.0165,
+      "step": 5172
+    },
+    {
+      "epoch": 0.9952562753644972,
+      "grad_norm": 0.42737115683387134,
+      "learning_rate": 1.3127243729949001e-08,
+      "loss": 0.9595,
+      "step": 5173
+    },
+    {
+      "epoch": 0.9954486697730347,
+      "grad_norm": 0.3362270391859536,
+      "learning_rate": 1.2136894784176011e-08,
+      "loss": 0.9807,
+      "step": 5174
+    },
+    {
+      "epoch": 0.9956410641815723,
+      "grad_norm": 0.29380800797610535,
+      "learning_rate": 1.1185379972256105e-08,
+      "loss": 0.9628,
+      "step": 5175
+    },
+    {
+      "epoch": 0.9958334585901097,
+      "grad_norm": 0.27825084019984586,
+      "learning_rate": 1.0272699663782526e-08,
+      "loss": 1.0753,
+      "step": 5176
+    },
+    {
+      "epoch": 0.9960258529986472,
+      "grad_norm": 0.3120808822918355,
+      "learning_rate": 9.398854213193974e-09,
+      "loss": 1.0719,
+      "step": 5177
+    },
+    {
+      "epoch": 0.9962182474071847,
+      "grad_norm": 0.303026423702011,
+      "learning_rate": 8.563843959907836e-09,
+      "loss": 1.0695,
+      "step": 5178
+    },
+    {
+      "epoch": 0.9964106418157223,
+      "grad_norm": 0.3408252513090465,
+      "learning_rate": 7.767669228231356e-09,
+      "loss": 1.043,
+      "step": 5179
+    },
+    {
+      "epoch": 0.9966030362242597,
+      "grad_norm": 0.2870483299653104,
+      "learning_rate": 7.01033032737275e-09,
+      "loss": 1.0582,
+      "step": 5180
+    },
+    {
+      "epoch": 0.9967954306327972,
+      "grad_norm": 0.32622364042112134,
+      "learning_rate": 6.291827551474505e-09,
+      "loss": 1.0088,
+      "step": 5181
+    },
+    {
+      "epoch": 0.9969878250413348,
+      "grad_norm": 0.4938456626228281,
+      "learning_rate": 5.6121611796133845e-09,
+      "loss": 0.9896,
+      "step": 5182
+    },
+    {
+      "epoch": 0.9971802194498722,
+      "grad_norm": 0.341240759710936,
+      "learning_rate": 4.971331475756014e-09,
+      "loss": 0.9955,
+      "step": 5183
+    },
+    {
+      "epoch": 0.9973726138584097,
+      "grad_norm": 0.3593343391171791,
+      "learning_rate": 4.369338688781088e-09,
+      "loss": 1.1007,
+      "step": 5184
+    },
+    {
+      "epoch": 0.9975650082669473,
+      "grad_norm": 0.2646884522334642,
+      "learning_rate": 3.806183052512679e-09,
+      "loss": 1.0633,
+      "step": 5185
+    },
+    {
+      "epoch": 0.9977574026754847,
+      "grad_norm": 0.38411537205633367,
+      "learning_rate": 3.2818647856647234e-09,
+      "loss": 1.1244,
+      "step": 5186
+    },
+    {
+      "epoch": 0.9979497970840222,
+      "grad_norm": 0.3403012202566673,
+      "learning_rate": 2.7963840918854288e-09,
+      "loss": 1.0468,
+      "step": 5187
+    },
+    {
+      "epoch": 0.9981421914925598,
+      "grad_norm": 0.30320713046567127,
+      "learning_rate": 2.3497411597128703e-09,
+      "loss": 1.0754,
+      "step": 5188
+    },
+    {
+      "epoch": 0.9983345859010972,
+      "grad_norm": 0.363354538467187,
+      "learning_rate": 1.9419361626416e-09,
+      "loss": 1.0282,
+      "step": 5189
+    },
+    {
+      "epoch": 0.9985269803096347,
+      "grad_norm": 0.2894238073581119,
+      "learning_rate": 1.5729692590338297e-09,
+      "loss": 1.022,
+      "step": 5190
+    },
+    {
+      "epoch": 0.9987193747181723,
+      "grad_norm": 0.3261138013320079,
+      "learning_rate": 1.2428405921971476e-09,
+      "loss": 1.0666,
+      "step": 5191
+    },
+    {
+      "epoch": 0.9989117691267098,
+      "grad_norm": 0.34898553867330984,
+      "learning_rate": 9.515502903734153e-10,
+      "loss": 1.0095,
+      "step": 5192
+    },
+    {
+      "epoch": 0.9991041635352472,
+      "grad_norm": 0.3063028761263027,
+      "learning_rate": 6.990984666610523e-10,
+      "loss": 1.0384,
+      "step": 5193
+    },
+    {
+      "epoch": 0.9992965579437848,
+      "grad_norm": 0.29267951087106403,
+      "learning_rate": 4.854852191371606e-10,
+      "loss": 1.0509,
+      "step": 5194
+    },
+    {
+      "epoch": 0.9994889523523223,
+      "grad_norm": 0.29660899633363924,
+      "learning_rate": 3.1071063075760466e-10,
+      "loss": 1.0022,
+      "step": 5195
+    },
+    {
+      "epoch": 0.9996813467608597,
+      "grad_norm": 0.28400778819313277,
+      "learning_rate": 1.7477476940142012e-10,
+      "loss": 0.979,
+      "step": 5196
+    },
+    {
+      "epoch": 0.9998737411693973,
+      "grad_norm": 0.44440155061503905,
+      "learning_rate": 7.767768785971185e-11,
+      "loss": 1.0221,
+      "step": 5197
+    },
+    {
+      "epoch": 0.9998737411693973,
+      "step": 5197,
+      "total_flos": 2841295497035776.0,
+      "train_loss": 0.08884279989888857,
+      "train_runtime": 6271.7197,
+      "train_samples_per_second": 106.079,
+      "train_steps_per_second": 0.829
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5197,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 250,
+  "total_flos": 2841295497035776.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}