diff --git "a/checkpoint-2000/trainer_state.json" "b/checkpoint-2000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2000/trainer_state.json"
@@ -0,0 +1,14034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.5310674455655868,
+  "eval_steps": 500,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0002655337227827934,
+      "grad_norm": 26.09913849342828,
+      "learning_rate": 0.0,
+      "loss": 3.2031,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005310674455655868,
+      "grad_norm": 33.24418612316917,
+      "learning_rate": 2.6525198938992043e-08,
+      "loss": 3.4062,
+      "step": 2
+    },
+    {
+      "epoch": 0.0007966011683483802,
+      "grad_norm": 22.666174254080957,
+      "learning_rate": 5.3050397877984086e-08,
+      "loss": 3.5469,
+      "step": 3
+    },
+    {
+      "epoch": 0.0010621348911311736,
+      "grad_norm": 47.54041711728203,
+      "learning_rate": 7.957559681697613e-08,
+      "loss": 3.3906,
+      "step": 4
+    },
+    {
+      "epoch": 0.0013276686139139671,
+      "grad_norm": 28.74129018702557,
+      "learning_rate": 1.0610079575596817e-07,
+      "loss": 3.7344,
+      "step": 5
+    },
+    {
+      "epoch": 0.0015932023366967605,
+      "grad_norm": 34.823794266535295,
+      "learning_rate": 1.3262599469496022e-07,
+      "loss": 4.875,
+      "step": 6
+    },
+    {
+      "epoch": 0.0018587360594795538,
+      "grad_norm": 33.17874143897226,
+      "learning_rate": 1.5915119363395226e-07,
+      "loss": 4.8125,
+      "step": 7
+    },
+    {
+      "epoch": 0.002124269782262347,
+      "grad_norm": 22.916328083623917,
+      "learning_rate": 1.856763925729443e-07,
+      "loss": 3.5625,
+      "step": 8
+    },
+    {
+      "epoch": 0.002389803505045141,
+      "grad_norm": 22.722952034819503,
+      "learning_rate": 2.1220159151193635e-07,
+      "loss": 2.8281,
+      "step": 9
+    },
+    {
+      "epoch": 0.0026553372278279343,
+      "grad_norm": 71.13233477133232,
+      "learning_rate": 2.387267904509284e-07,
+      "loss": 4.2812,
+      "step": 10
+    },
+    {
+      "epoch": 0.0029208709506107276,
+      "grad_norm": 22.509369620011583,
+      "learning_rate": 2.6525198938992043e-07,
+      "loss": 2.6562,
+      "step": 11
+    },
+    {
+      "epoch": 0.003186404673393521,
+      "grad_norm": 36.02796517304586,
+      "learning_rate": 2.917771883289125e-07,
+      "loss": 3.9375,
+      "step": 12
+    },
+    {
+      "epoch": 0.0034519383961763143,
+      "grad_norm": 35.53736888706161,
+      "learning_rate": 3.183023872679045e-07,
+      "loss": 4.375,
+      "step": 13
+    },
+    {
+      "epoch": 0.0037174721189591076,
+      "grad_norm": 21.726341733924155,
+      "learning_rate": 3.4482758620689656e-07,
+      "loss": 2.8594,
+      "step": 14
+    },
+    {
+      "epoch": 0.003983005841741901,
+      "grad_norm": 26.02926274160067,
+      "learning_rate": 3.713527851458886e-07,
+      "loss": 3.3125,
+      "step": 15
+    },
+    {
+      "epoch": 0.004248539564524694,
+      "grad_norm": 33.549456836516455,
+      "learning_rate": 3.9787798408488065e-07,
+      "loss": 4.5,
+      "step": 16
+    },
+    {
+      "epoch": 0.004514073287307488,
+      "grad_norm": 5704.608112490134,
+      "learning_rate": 4.244031830238727e-07,
+      "loss": 2.6875,
+      "step": 17
+    },
+    {
+      "epoch": 0.004779607010090282,
+      "grad_norm": 46.28223578012096,
+      "learning_rate": 4.5092838196286473e-07,
+      "loss": 3.3906,
+      "step": 18
+    },
+    {
+      "epoch": 0.005045140732873075,
+      "grad_norm": 21.51924983534021,
+      "learning_rate": 4.774535809018568e-07,
+      "loss": 4.0938,
+      "step": 19
+    },
+    {
+      "epoch": 0.0053106744556558685,
+      "grad_norm": 27.91289735068091,
+      "learning_rate": 5.039787798408489e-07,
+      "loss": 3.4062,
+      "step": 20
+    },
+    {
+      "epoch": 0.0055762081784386614,
+      "grad_norm": 38.86707334268686,
+      "learning_rate": 5.305039787798409e-07,
+      "loss": 4.5938,
+      "step": 21
+    },
+    {
+      "epoch": 0.005841741901221455,
+      "grad_norm": 46.02296280217304,
+      "learning_rate": 5.570291777188329e-07,
+      "loss": 4.5312,
+      "step": 22
+    },
+    {
+      "epoch": 0.006107275624004248,
+      "grad_norm": 29.064306674372673,
+      "learning_rate": 5.83554376657825e-07,
+      "loss": 3.7031,
+      "step": 23
+    },
+    {
+      "epoch": 0.006372809346787042,
+      "grad_norm": 22.0053607091662,
+      "learning_rate": 6.10079575596817e-07,
+      "loss": 4.0938,
+      "step": 24
+    },
+    {
+      "epoch": 0.006638343069569836,
+      "grad_norm": 25.74196411791827,
+      "learning_rate": 6.36604774535809e-07,
+      "loss": 3.1875,
+      "step": 25
+    },
+    {
+      "epoch": 0.006903876792352629,
+      "grad_norm": 28.41314685895795,
+      "learning_rate": 6.631299734748012e-07,
+      "loss": 3.5938,
+      "step": 26
+    },
+    {
+      "epoch": 0.007169410515135422,
+      "grad_norm": 24.75234923466472,
+      "learning_rate": 6.896551724137931e-07,
+      "loss": 2.6875,
+      "step": 27
+    },
+    {
+      "epoch": 0.007434944237918215,
+      "grad_norm": 30.66371803716353,
+      "learning_rate": 7.161803713527853e-07,
+      "loss": 4.0312,
+      "step": 28
+    },
+    {
+      "epoch": 0.007700477960701009,
+      "grad_norm": 29.00163943943682,
+      "learning_rate": 7.427055702917772e-07,
+      "loss": 3.3594,
+      "step": 29
+    },
+    {
+      "epoch": 0.007966011683483803,
+      "grad_norm": 25.77157343246292,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 3.1875,
+      "step": 30
+    },
+    {
+      "epoch": 0.008231545406266597,
+      "grad_norm": 22.334502595274675,
+      "learning_rate": 7.957559681697613e-07,
+      "loss": 4.7188,
+      "step": 31
+    },
+    {
+      "epoch": 0.008497079129049389,
+      "grad_norm": 32.603443673328215,
+      "learning_rate": 8.222811671087534e-07,
+      "loss": 3.8906,
+      "step": 32
+    },
+    {
+      "epoch": 0.008762612851832182,
+      "grad_norm": 24.08930675302932,
+      "learning_rate": 8.488063660477454e-07,
+      "loss": 2.8906,
+      "step": 33
+    },
+    {
+      "epoch": 0.009028146574614976,
+      "grad_norm": 32.39366263050705,
+      "learning_rate": 8.753315649867375e-07,
+      "loss": 4.125,
+      "step": 34
+    },
+    {
+      "epoch": 0.00929368029739777,
+      "grad_norm": 29.971764947175103,
+      "learning_rate": 9.018567639257295e-07,
+      "loss": 3.2344,
+      "step": 35
+    },
+    {
+      "epoch": 0.009559214020180564,
+      "grad_norm": 33.19925342837082,
+      "learning_rate": 9.283819628647216e-07,
+      "loss": 4.25,
+      "step": 36
+    },
+    {
+      "epoch": 0.009824747742963356,
+      "grad_norm": 24.097206276546917,
+      "learning_rate": 9.549071618037136e-07,
+      "loss": 2.7031,
+      "step": 37
+    },
+    {
+      "epoch": 0.01009028146574615,
+      "grad_norm": 25.033091440150386,
+      "learning_rate": 9.814323607427057e-07,
+      "loss": 3.5469,
+      "step": 38
+    },
+    {
+      "epoch": 0.010355815188528943,
+      "grad_norm": 19.83584675026491,
+      "learning_rate": 1.0079575596816979e-06,
+      "loss": 3.8125,
+      "step": 39
+    },
+    {
+      "epoch": 0.010621348911311737,
+      "grad_norm": 22.843498610408417,
+      "learning_rate": 1.0344827586206898e-06,
+      "loss": 4.0,
+      "step": 40
+    },
+    {
+      "epoch": 0.01088688263409453,
+      "grad_norm": 27.23163955174518,
+      "learning_rate": 1.0610079575596817e-06,
+      "loss": 4.2188,
+      "step": 41
+    },
+    {
+      "epoch": 0.011152416356877323,
+      "grad_norm": 31.562228731897743,
+      "learning_rate": 1.0875331564986739e-06,
+      "loss": 4.375,
+      "step": 42
+    },
+    {
+      "epoch": 0.011417950079660117,
+      "grad_norm": 22.949386153635178,
+      "learning_rate": 1.1140583554376658e-06,
+      "loss": 3.5156,
+      "step": 43
+    },
+    {
+      "epoch": 0.01168348380244291,
+      "grad_norm": 88.31997625312175,
+      "learning_rate": 1.140583554376658e-06,
+      "loss": 2.7812,
+      "step": 44
+    },
+    {
+      "epoch": 0.011949017525225704,
+      "grad_norm": 14.729164979113321,
+      "learning_rate": 1.16710875331565e-06,
+      "loss": 2.2344,
+      "step": 45
+    },
+    {
+      "epoch": 0.012214551248008496,
+      "grad_norm": 26.318599853460583,
+      "learning_rate": 1.193633952254642e-06,
+      "loss": 3.8125,
+      "step": 46
+    },
+    {
+      "epoch": 0.01248008497079129,
+      "grad_norm": 36.435091476888246,
+      "learning_rate": 1.220159151193634e-06,
+      "loss": 3.5156,
+      "step": 47
+    },
+    {
+      "epoch": 0.012745618693574084,
+      "grad_norm": 21.353387655655027,
+      "learning_rate": 1.2466843501326261e-06,
+      "loss": 2.4219,
+      "step": 48
+    },
+    {
+      "epoch": 0.013011152416356878,
+      "grad_norm": 36.608824688888205,
+      "learning_rate": 1.273209549071618e-06,
+      "loss": 3.6562,
+      "step": 49
+    },
+    {
+      "epoch": 0.013276686139139671,
+      "grad_norm": 32.7469992104952,
+      "learning_rate": 1.29973474801061e-06,
+      "loss": 4.2812,
+      "step": 50
+    },
+    {
+      "epoch": 0.013542219861922463,
+      "grad_norm": 527.1528527718093,
+      "learning_rate": 1.3262599469496024e-06,
+      "loss": 2.4375,
+      "step": 51
+    },
+    {
+      "epoch": 0.013807753584705257,
+      "grad_norm": 27.990609419848134,
+      "learning_rate": 1.3527851458885943e-06,
+      "loss": 2.3906,
+      "step": 52
+    },
+    {
+      "epoch": 0.014073287307488051,
+      "grad_norm": 14.694389603614836,
+      "learning_rate": 1.3793103448275862e-06,
+      "loss": 3.2344,
+      "step": 53
+    },
+    {
+      "epoch": 0.014338821030270845,
+      "grad_norm": 20.735659157221477,
+      "learning_rate": 1.4058355437665782e-06,
+      "loss": 2.8281,
+      "step": 54
+    },
+    {
+      "epoch": 0.014604354753053638,
+      "grad_norm": 31.138348407935453,
+      "learning_rate": 1.4323607427055705e-06,
+      "loss": 3.8906,
+      "step": 55
+    },
+    {
+      "epoch": 0.01486988847583643,
+      "grad_norm": 29.62867209466913,
+      "learning_rate": 1.4588859416445625e-06,
+      "loss": 3.6875,
+      "step": 56
+    },
+    {
+      "epoch": 0.015135422198619224,
+      "grad_norm": 14.03880122581976,
+      "learning_rate": 1.4854111405835544e-06,
+      "loss": 2.1406,
+      "step": 57
+    },
+    {
+      "epoch": 0.015400955921402018,
+      "grad_norm": 33.23697742979803,
+      "learning_rate": 1.5119363395225464e-06,
+      "loss": 3.5,
+      "step": 58
+    },
+    {
+      "epoch": 0.015666489644184812,
+      "grad_norm": 29.453655249544546,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 2.5312,
+      "step": 59
+    },
+    {
+      "epoch": 0.015932023366967606,
+      "grad_norm": 27.29098711329942,
+      "learning_rate": 1.5649867374005307e-06,
+      "loss": 2.6406,
+      "step": 60
+    },
+    {
+      "epoch": 0.0161975570897504,
+      "grad_norm": 28.33472292785293,
+      "learning_rate": 1.5915119363395226e-06,
+      "loss": 2.5,
+      "step": 61
+    },
+    {
+      "epoch": 0.016463090812533193,
+      "grad_norm": 16.521064455634754,
+      "learning_rate": 1.618037135278515e-06,
+      "loss": 2.4219,
+      "step": 62
+    },
+    {
+      "epoch": 0.016728624535315983,
+      "grad_norm": 13.384541738840948,
+      "learning_rate": 1.6445623342175069e-06,
+      "loss": 1.8672,
+      "step": 63
+    },
+    {
+      "epoch": 0.016994158258098777,
+      "grad_norm": 19.701703431707948,
+      "learning_rate": 1.6710875331564988e-06,
+      "loss": 2.4219,
+      "step": 64
+    },
+    {
+      "epoch": 0.01725969198088157,
+      "grad_norm": 32.143288806734,
+      "learning_rate": 1.6976127320954908e-06,
+      "loss": 2.6094,
+      "step": 65
+    },
+    {
+      "epoch": 0.017525225703664365,
+      "grad_norm": 30.63700955458724,
+      "learning_rate": 1.724137931034483e-06,
+      "loss": 2.8125,
+      "step": 66
+    },
+    {
+      "epoch": 0.01779075942644716,
+      "grad_norm": 18.61018137166964,
+      "learning_rate": 1.750663129973475e-06,
+      "loss": 2.125,
+      "step": 67
+    },
+    {
+      "epoch": 0.018056293149229952,
+      "grad_norm": 14.421577541639564,
+      "learning_rate": 1.777188328912467e-06,
+      "loss": 1.7578,
+      "step": 68
+    },
+    {
+      "epoch": 0.018321826872012746,
+      "grad_norm": 29.39406132029808,
+      "learning_rate": 1.803713527851459e-06,
+      "loss": 2.4375,
+      "step": 69
+    },
+    {
+      "epoch": 0.01858736059479554,
+      "grad_norm": 21.159681548926745,
+      "learning_rate": 1.830238726790451e-06,
+      "loss": 2.0938,
+      "step": 70
+    },
+    {
+      "epoch": 0.018852894317578334,
+      "grad_norm": 27.539564746628287,
+      "learning_rate": 1.8567639257294432e-06,
+      "loss": 3.0156,
+      "step": 71
+    },
+    {
+      "epoch": 0.019118428040361127,
+      "grad_norm": 24.190559734623005,
+      "learning_rate": 1.8832891246684352e-06,
+      "loss": 3.0,
+      "step": 72
+    },
+    {
+      "epoch": 0.019383961763143918,
+      "grad_norm": 26.96417513483311,
+      "learning_rate": 1.909814323607427e-06,
+      "loss": 3.0156,
+      "step": 73
+    },
+    {
+      "epoch": 0.01964949548592671,
+      "grad_norm": 24.89585126504584,
+      "learning_rate": 1.9363395225464193e-06,
+      "loss": 2.2344,
+      "step": 74
+    },
+    {
+      "epoch": 0.019915029208709505,
+      "grad_norm": 56.411148213900134,
+      "learning_rate": 1.9628647214854114e-06,
+      "loss": 2.1406,
+      "step": 75
+    },
+    {
+      "epoch": 0.0201805629314923,
+      "grad_norm": 26.867726567521345,
+      "learning_rate": 1.989389920424403e-06,
+      "loss": 1.5469,
+      "step": 76
+    },
+    {
+      "epoch": 0.020446096654275093,
+      "grad_norm": 25.736893839761308,
+      "learning_rate": 2.0159151193633957e-06,
+      "loss": 2.5312,
+      "step": 77
+    },
+    {
+      "epoch": 0.020711630377057887,
+      "grad_norm": 29.10383691197578,
+      "learning_rate": 2.0424403183023874e-06,
+      "loss": 2.0312,
+      "step": 78
+    },
+    {
+      "epoch": 0.02097716409984068,
+      "grad_norm": 16.34789260778447,
+      "learning_rate": 2.0689655172413796e-06,
+      "loss": 2.6719,
+      "step": 79
+    },
+    {
+      "epoch": 0.021242697822623474,
+      "grad_norm": 17.443495153034082,
+      "learning_rate": 2.0954907161803713e-06,
+      "loss": 0.7266,
+      "step": 80
+    },
+    {
+      "epoch": 0.021508231545406268,
+      "grad_norm": 25.598705341280716,
+      "learning_rate": 2.1220159151193635e-06,
+      "loss": 1.2109,
+      "step": 81
+    },
+    {
+      "epoch": 0.02177376526818906,
+      "grad_norm": 143.5401268253218,
+      "learning_rate": 2.1485411140583556e-06,
+      "loss": 1.6016,
+      "step": 82
+    },
+    {
+      "epoch": 0.022039298990971852,
+      "grad_norm": 26.389545474475735,
+      "learning_rate": 2.1750663129973478e-06,
+      "loss": 1.8438,
+      "step": 83
+    },
+    {
+      "epoch": 0.022304832713754646,
+      "grad_norm": 23.96871143922354,
+      "learning_rate": 2.2015915119363395e-06,
+      "loss": 0.9219,
+      "step": 84
+    },
+    {
+      "epoch": 0.02257036643653744,
+      "grad_norm": 38.19763740874266,
+      "learning_rate": 2.2281167108753316e-06,
+      "loss": 0.8359,
+      "step": 85
+    },
+    {
+      "epoch": 0.022835900159320233,
+      "grad_norm": 29.364774056584984,
+      "learning_rate": 2.2546419098143238e-06,
+      "loss": 0.6602,
+      "step": 86
+    },
+    {
+      "epoch": 0.023101433882103027,
+      "grad_norm": 26.845223667722713,
+      "learning_rate": 2.281167108753316e-06,
+      "loss": 0.8086,
+      "step": 87
+    },
+    {
+      "epoch": 0.02336696760488582,
+      "grad_norm": 28.831828873775432,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 1.4062,
+      "step": 88
+    },
+    {
+      "epoch": 0.023632501327668615,
+      "grad_norm": 24.13587932409368,
+      "learning_rate": 2.3342175066313e-06,
+      "loss": 1.3828,
+      "step": 89
+    },
+    {
+      "epoch": 0.02389803505045141,
+      "grad_norm": 24.636469804747414,
+      "learning_rate": 2.360742705570292e-06,
+      "loss": 1.0547,
+      "step": 90
+    },
+    {
+      "epoch": 0.024163568773234202,
+      "grad_norm": 10.515597346677467,
+      "learning_rate": 2.387267904509284e-06,
+      "loss": 0.0923,
+      "step": 91
+    },
+    {
+      "epoch": 0.024429102496016993,
+      "grad_norm": 22.66977819594809,
+      "learning_rate": 2.4137931034482762e-06,
+      "loss": 0.2832,
+      "step": 92
+    },
+    {
+      "epoch": 0.024694636218799786,
+      "grad_norm": 4.5274633525739185,
+      "learning_rate": 2.440318302387268e-06,
+      "loss": 0.0259,
+      "step": 93
+    },
+    {
+      "epoch": 0.02496016994158258,
+      "grad_norm": 9.36517984448498,
+      "learning_rate": 2.46684350132626e-06,
+      "loss": 0.0442,
+      "step": 94
+    },
+    {
+      "epoch": 0.025225703664365374,
+      "grad_norm": 11.897696771384094,
+      "learning_rate": 2.4933687002652523e-06,
+      "loss": 0.1138,
+      "step": 95
+    },
+    {
+      "epoch": 0.025491237387148168,
+      "grad_norm": 32.70173331678797,
+      "learning_rate": 2.5198938992042444e-06,
+      "loss": 6.4688,
+      "step": 96
+    },
+    {
+      "epoch": 0.02575677110993096,
+      "grad_norm": 0.031168975385739393,
+      "learning_rate": 2.546419098143236e-06,
+      "loss": 0.0003,
+      "step": 97
+    },
+    {
+      "epoch": 0.026022304832713755,
+      "grad_norm": 0.7313576810659675,
+      "learning_rate": 2.5729442970822283e-06,
+      "loss": 0.0075,
+      "step": 98
+    },
+    {
+      "epoch": 0.02628783855549655,
+      "grad_norm": 0.5685930769193894,
+      "learning_rate": 2.59946949602122e-06,
+      "loss": 0.0045,
+      "step": 99
+    },
+    {
+      "epoch": 0.026553372278279343,
+      "grad_norm": 4.772903853691814,
+      "learning_rate": 2.625994694960212e-06,
+      "loss": 0.2275,
+      "step": 100
+    },
+    {
+      "epoch": 0.026818906001062136,
+      "grad_norm": 5.102874825538071,
+      "learning_rate": 2.6525198938992047e-06,
+      "loss": 0.5703,
+      "step": 101
+    },
+    {
+      "epoch": 0.027084439723844927,
+      "grad_norm": 0.27402642551717465,
+      "learning_rate": 2.679045092838197e-06,
+      "loss": 0.0028,
+      "step": 102
+    },
+    {
+      "epoch": 0.02734997344662772,
+      "grad_norm": 4.6324372685319695,
+      "learning_rate": 2.7055702917771886e-06,
+      "loss": 0.1367,
+      "step": 103
+    },
+    {
+      "epoch": 0.027615507169410514,
+      "grad_norm": 0.038045626774934026,
+      "learning_rate": 2.7320954907161808e-06,
+      "loss": 0.0007,
+      "step": 104
+    },
+    {
+      "epoch": 0.027881040892193308,
+      "grad_norm": 0.6305735286653126,
+      "learning_rate": 2.7586206896551725e-06,
+      "loss": 0.0097,
+      "step": 105
+    },
+    {
+      "epoch": 0.028146574614976102,
+      "grad_norm": 0.019465351131816563,
+      "learning_rate": 2.7851458885941646e-06,
+      "loss": 0.0003,
+      "step": 106
+    },
+    {
+      "epoch": 0.028412108337758896,
+      "grad_norm": 0.03074317996513282,
+      "learning_rate": 2.8116710875331564e-06,
+      "loss": 0.0004,
+      "step": 107
+    },
+    {
+      "epoch": 0.02867764206054169,
+      "grad_norm": 4.503535246450842,
+      "learning_rate": 2.8381962864721485e-06,
+      "loss": 0.582,
+      "step": 108
+    },
+    {
+      "epoch": 0.028943175783324483,
+      "grad_norm": 4.8728912631764265,
+      "learning_rate": 2.864721485411141e-06,
+      "loss": 0.5469,
+      "step": 109
+    },
+    {
+      "epoch": 0.029208709506107277,
+      "grad_norm": 0.0095940645247496,
+      "learning_rate": 2.891246684350133e-06,
+      "loss": 0.0002,
+      "step": 110
+    },
+    {
+      "epoch": 0.02947424322889007,
+      "grad_norm": 5.105714132066514,
+      "learning_rate": 2.917771883289125e-06,
+      "loss": 0.7383,
+      "step": 111
+    },
+    {
+      "epoch": 0.02973977695167286,
+      "grad_norm": 0.18245285164952776,
+      "learning_rate": 2.944297082228117e-06,
+      "loss": 0.0027,
+      "step": 112
+    },
+    {
+      "epoch": 0.030005310674455655,
+      "grad_norm": 4.137453039540649,
+      "learning_rate": 2.970822281167109e-06,
+      "loss": 0.2158,
+      "step": 113
+    },
+    {
+      "epoch": 0.03027084439723845,
+      "grad_norm": 0.466146965565003,
+      "learning_rate": 2.997347480106101e-06,
+      "loss": 0.0085,
+      "step": 114
+    },
+    {
+      "epoch": 0.030536378120021242,
+      "grad_norm": 4.144598857842222,
+      "learning_rate": 3.0238726790450927e-06,
+      "loss": 0.4355,
+      "step": 115
+    },
+    {
+      "epoch": 0.030801911842804036,
+      "grad_norm": 0.07481190979762373,
+      "learning_rate": 3.0503978779840853e-06,
+      "loss": 0.0012,
+      "step": 116
+    },
+    {
+      "epoch": 0.03106744556558683,
+      "grad_norm": 4.864460397183156,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 0.4883,
+      "step": 117
+    },
+    {
+      "epoch": 0.031332979288369624,
+      "grad_norm": 0.046629399654510524,
+      "learning_rate": 3.103448275862069e-06,
+      "loss": 0.0008,
+      "step": 118
+    },
+    {
+      "epoch": 0.031598513011152414,
+      "grad_norm": 0.9844013414494834,
+      "learning_rate": 3.1299734748010613e-06,
+      "loss": 0.0006,
+      "step": 119
+    },
+    {
+      "epoch": 0.03186404673393521,
+      "grad_norm": 2.8873700648345726,
+      "learning_rate": 3.1564986737400535e-06,
+      "loss": 0.0525,
+      "step": 120
+    },
+    {
+      "epoch": 0.032129580456718,
+      "grad_norm": 13.401562271488295,
+      "learning_rate": 3.183023872679045e-06,
+      "loss": 2.1094,
+      "step": 121
+    },
+    {
+      "epoch": 0.0323951141795008,
+      "grad_norm": 0.9753972945569677,
+      "learning_rate": 3.2095490716180373e-06,
+      "loss": 0.0198,
+      "step": 122
+    },
+    {
+      "epoch": 0.03266064790228359,
+      "grad_norm": 0.052264498037133865,
+      "learning_rate": 3.23607427055703e-06,
+      "loss": 0.0011,
+      "step": 123
+    },
+    {
+      "epoch": 0.032926181625066386,
+      "grad_norm": 0.0275297136646831,
+      "learning_rate": 3.2625994694960216e-06,
+      "loss": 0.0006,
+      "step": 124
+    },
+    {
+      "epoch": 0.03319171534784918,
+      "grad_norm": 0.1637629675751948,
+      "learning_rate": 3.2891246684350138e-06,
+      "loss": 0.0037,
+      "step": 125
+    },
+    {
+      "epoch": 0.03345724907063197,
+      "grad_norm": 4.78491469984592,
+      "learning_rate": 3.3156498673740055e-06,
+      "loss": 0.3535,
+      "step": 126
+    },
+    {
+      "epoch": 0.033722782793414764,
+      "grad_norm": 15.570720009635428,
+      "learning_rate": 3.3421750663129977e-06,
+      "loss": 2.7969,
+      "step": 127
+    },
+    {
+      "epoch": 0.033988316516197555,
+      "grad_norm": 4.505829037318201,
+      "learning_rate": 3.3687002652519894e-06,
+      "loss": 0.3105,
+      "step": 128
+    },
+    {
+      "epoch": 0.03425385023898035,
+      "grad_norm": 0.042401600725809944,
+      "learning_rate": 3.3952254641909815e-06,
+      "loss": 0.0009,
+      "step": 129
+    },
+    {
+      "epoch": 0.03451938396176314,
+      "grad_norm": 0.3272160377202937,
+      "learning_rate": 3.4217506631299737e-06,
+      "loss": 0.0074,
+      "step": 130
+    },
+    {
+      "epoch": 0.03478491768454594,
+      "grad_norm": 0.20468371949697137,
+      "learning_rate": 3.448275862068966e-06,
+      "loss": 0.0036,
+      "step": 131
+    },
+    {
+      "epoch": 0.03505045140732873,
+      "grad_norm": 4.559385816252051,
+      "learning_rate": 3.474801061007958e-06,
+      "loss": 0.2773,
+      "step": 132
+    },
+    {
+      "epoch": 0.03531598513011153,
+      "grad_norm": 0.44947961387998925,
+      "learning_rate": 3.50132625994695e-06,
+      "loss": 0.0093,
+      "step": 133
+    },
+    {
+      "epoch": 0.03558151885289432,
+      "grad_norm": 2.628443790962567,
+      "learning_rate": 3.527851458885942e-06,
+      "loss": 0.0767,
+      "step": 134
+    },
+    {
+      "epoch": 0.035847052575677114,
+      "grad_norm": 0.3344294580416105,
+      "learning_rate": 3.554376657824934e-06,
+      "loss": 0.0079,
+      "step": 135
+    },
+    {
+      "epoch": 0.036112586298459905,
+      "grad_norm": 6.1109142288253375,
+      "learning_rate": 3.5809018567639257e-06,
+      "loss": 0.1455,
+      "step": 136
+    },
+    {
+      "epoch": 0.036378120021242695,
+      "grad_norm": 0.19257798325675396,
+      "learning_rate": 3.607427055702918e-06,
+      "loss": 0.0043,
+      "step": 137
+    },
+    {
+      "epoch": 0.03664365374402549,
+      "grad_norm": 3.878928845456885,
+      "learning_rate": 3.6339522546419104e-06,
+      "loss": 0.1934,
+      "step": 138
+    },
+    {
+      "epoch": 0.03690918746680828,
+      "grad_norm": 1.5262893890482263,
+      "learning_rate": 3.660477453580902e-06,
+      "loss": 0.032,
+      "step": 139
+    },
+    {
+      "epoch": 0.03717472118959108,
+      "grad_norm": 0.19902989840213414,
+      "learning_rate": 3.6870026525198943e-06,
+      "loss": 0.0045,
+      "step": 140
+    },
+    {
+      "epoch": 0.03744025491237387,
+      "grad_norm": 0.5136742696527371,
+      "learning_rate": 3.7135278514588865e-06,
+      "loss": 0.0104,
+      "step": 141
+    },
+    {
+      "epoch": 0.03770578863515667,
+      "grad_norm": 3.86476349919558,
+      "learning_rate": 3.740053050397878e-06,
+      "loss": 0.1367,
+      "step": 142
+    },
+    {
+      "epoch": 0.03797132235793946,
+      "grad_norm": 0.796785610921372,
+      "learning_rate": 3.7665782493368703e-06,
+      "loss": 0.0194,
+      "step": 143
+    },
+    {
+      "epoch": 0.038236856080722255,
+      "grad_norm": 1.5271845218247893,
+      "learning_rate": 3.793103448275862e-06,
+      "loss": 0.0288,
+      "step": 144
+    },
+    {
+      "epoch": 0.038502389803505045,
+      "grad_norm": 2.427314819468491,
+      "learning_rate": 3.819628647214854e-06,
+      "loss": 0.0576,
+      "step": 145
+    },
+    {
+      "epoch": 0.038767923526287836,
+      "grad_norm": 4.117849401975011,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.2441,
+      "step": 146
+    },
+    {
+      "epoch": 0.03903345724907063,
+      "grad_norm": 0.15723067876063995,
+      "learning_rate": 3.8726790450928385e-06,
+      "loss": 0.0039,
+      "step": 147
+    },
+    {
+      "epoch": 0.03929899097185342,
+      "grad_norm": 3.7869596148640228,
+      "learning_rate": 3.89920424403183e-06,
+      "loss": 0.1641,
+      "step": 148
+    },
+    {
+      "epoch": 0.03956452469463622,
+      "grad_norm": 0.37297101582033887,
+      "learning_rate": 3.925729442970823e-06,
+      "loss": 0.0067,
+      "step": 149
+    },
+    {
+      "epoch": 0.03983005841741901,
+      "grad_norm": 3.6378474369832037,
+      "learning_rate": 3.9522546419098145e-06,
+      "loss": 0.0957,
+      "step": 150
+    },
+    {
+      "epoch": 0.04009559214020181,
+      "grad_norm": 0.43825791965553007,
+      "learning_rate": 3.978779840848806e-06,
+      "loss": 0.0105,
+      "step": 151
+    },
+    {
+      "epoch": 0.0403611258629846,
+      "grad_norm": 3.833802266035273,
+      "learning_rate": 4.005305039787799e-06,
+      "loss": 0.1934,
+      "step": 152
+    },
+    {
+      "epoch": 0.040626659585767395,
+      "grad_norm": 0.27648447133818854,
+      "learning_rate": 4.031830238726791e-06,
+      "loss": 0.0059,
+      "step": 153
+    },
+    {
+      "epoch": 0.040892193308550186,
+      "grad_norm": 0.5502618375049242,
+      "learning_rate": 4.058355437665783e-06,
+      "loss": 0.0117,
+      "step": 154
+    },
+    {
+      "epoch": 0.041157727031332976,
+      "grad_norm": 0.3759989484991633,
+      "learning_rate": 4.084880636604775e-06,
+      "loss": 0.0093,
+      "step": 155
+    },
+    {
+      "epoch": 0.04142326075411577,
+      "grad_norm": 0.15725269535355169,
+      "learning_rate": 4.111405835543767e-06,
+      "loss": 0.0033,
+      "step": 156
+    },
+    {
+      "epoch": 0.041688794476898564,
+      "grad_norm": 0.5545627730013464,
+      "learning_rate": 4.137931034482759e-06,
+      "loss": 0.0137,
+      "step": 157
+    },
+    {
+      "epoch": 0.04195432819968136,
+      "grad_norm": 0.3259056057700917,
+      "learning_rate": 4.164456233421751e-06,
+      "loss": 0.0081,
+      "step": 158
+    },
+    {
+      "epoch": 0.04221986192246415,
+      "grad_norm": 0.1865224020504324,
+      "learning_rate": 4.190981432360743e-06,
+      "loss": 0.0041,
+      "step": 159
+    },
+    {
+      "epoch": 0.04248539564524695,
+      "grad_norm": 1.8307778726247308,
+      "learning_rate": 4.217506631299735e-06,
+      "loss": 0.0532,
+      "step": 160
+    },
+    {
+      "epoch": 0.04275092936802974,
+      "grad_norm": 3.8899387815695117,
+      "learning_rate": 4.244031830238727e-06,
+      "loss": 0.1553,
+      "step": 161
+    },
+    {
+      "epoch": 0.043016463090812536,
+      "grad_norm": 0.35213110395365343,
+      "learning_rate": 4.2705570291777195e-06,
+      "loss": 0.0063,
+      "step": 162
+    },
+    {
+      "epoch": 0.043281996813595326,
+      "grad_norm": 0.12739793280337483,
+      "learning_rate": 4.297082228116711e-06,
+      "loss": 0.0029,
+      "step": 163
+    },
+    {
+      "epoch": 0.04354753053637812,
+      "grad_norm": 3.7917485475147767,
+      "learning_rate": 4.323607427055703e-06,
+      "loss": 0.2559,
+      "step": 164
+    },
+    {
+      "epoch": 0.043813064259160914,
+      "grad_norm": 4.222145018176052,
+      "learning_rate": 4.3501326259946955e-06,
+      "loss": 0.3105,
+      "step": 165
+    },
+    {
+      "epoch": 0.044078597981943704,
+      "grad_norm": 0.1510587971806312,
+      "learning_rate": 4.376657824933687e-06,
+      "loss": 0.0036,
+      "step": 166
+    },
+    {
+      "epoch": 0.0443441317047265,
+      "grad_norm": 0.3746365601473909,
+      "learning_rate": 4.403183023872679e-06,
+      "loss": 0.0091,
+      "step": 167
+    },
+    {
+      "epoch": 0.04460966542750929,
+      "grad_norm": 3.6742314366524482,
+      "learning_rate": 4.4297082228116715e-06,
+      "loss": 0.1641,
+      "step": 168
+    },
+    {
+      "epoch": 0.04487519915029209,
+      "grad_norm": 4.169049834925408,
+      "learning_rate": 4.456233421750663e-06,
+      "loss": 0.1367,
+      "step": 169
+    },
+    {
+      "epoch": 0.04514073287307488,
+      "grad_norm": 23.2584513485522,
+      "learning_rate": 4.482758620689656e-06,
+      "loss": 2.375,
+      "step": 170
+    },
+    {
+      "epoch": 0.045406266595857676,
+      "grad_norm": 3.6642006588717737,
+      "learning_rate": 4.5092838196286476e-06,
+      "loss": 0.1465,
+      "step": 171
+    },
+    {
+      "epoch": 0.04567180031864047,
+      "grad_norm": 0.2890573481827513,
+      "learning_rate": 4.535809018567639e-06,
+      "loss": 0.0065,
+      "step": 172
+    },
+    {
+      "epoch": 0.045937334041423264,
+      "grad_norm": 3.68276451135959,
+      "learning_rate": 4.562334217506632e-06,
+      "loss": 0.1836,
+      "step": 173
+    },
+    {
+      "epoch": 0.046202867764206054,
+      "grad_norm": 2.4177768079368986,
+      "learning_rate": 4.588859416445624e-06,
+      "loss": 0.0479,
+      "step": 174
+    },
+    {
+      "epoch": 0.046468401486988845,
+      "grad_norm": 17.645780333783485,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 1.9531,
+      "step": 175
+    },
+    {
+      "epoch": 0.04673393520977164,
+      "grad_norm": 0.9556851330908339,
+      "learning_rate": 4.641909814323608e-06,
+      "loss": 0.0231,
+      "step": 176
+    },
+    {
+      "epoch": 0.04699946893255443,
+      "grad_norm": 0.15825725091826115,
+      "learning_rate": 4.6684350132626e-06,
+      "loss": 0.0035,
+      "step": 177
+    },
+    {
+      "epoch": 0.04726500265533723,
+      "grad_norm": 5.0505715269377065,
+      "learning_rate": 4.694960212201592e-06,
+      "loss": 0.0219,
+      "step": 178
+    },
+    {
+      "epoch": 0.04753053637812002,
+      "grad_norm": 0.19812718466796467,
+      "learning_rate": 4.721485411140584e-06,
+      "loss": 0.0051,
+      "step": 179
+    },
+    {
+      "epoch": 0.04779607010090282,
+      "grad_norm": 3.6161442909421218,
+      "learning_rate": 4.748010610079576e-06,
+      "loss": 0.165,
+      "step": 180
+    },
+    {
+      "epoch": 0.04806160382368561,
+      "grad_norm": 0.322834845757156,
+      "learning_rate": 4.774535809018568e-06,
+      "loss": 0.0065,
+      "step": 181
+    },
+    {
+      "epoch": 0.048327137546468404,
+      "grad_norm": 5.4547556811669935,
+      "learning_rate": 4.80106100795756e-06,
+      "loss": 2.125,
+      "step": 182
+    },
+    {
+      "epoch": 0.048592671269251195,
+      "grad_norm": 3.584757373914313,
+      "learning_rate": 4.8275862068965525e-06,
+      "loss": 0.1553,
+      "step": 183
+    },
+    {
+      "epoch": 0.048858204992033985,
+      "grad_norm": 0.7190093535962121,
+      "learning_rate": 4.854111405835544e-06,
+      "loss": 0.0187,
+      "step": 184
+    },
+    {
+      "epoch": 0.04912373871481678,
+      "grad_norm": 8.285280730208004,
+      "learning_rate": 4.880636604774536e-06,
+      "loss": 2.0156,
+      "step": 185
+    },
+    {
+      "epoch": 0.04938927243759957,
+      "grad_norm": 0.1589100928383723,
+      "learning_rate": 4.9071618037135285e-06,
+      "loss": 0.0035,
+      "step": 186
+    },
+    {
+      "epoch": 0.04965480616038237,
+      "grad_norm": 1.9461569166969948,
+      "learning_rate": 4.93368700265252e-06,
+      "loss": 0.0479,
+      "step": 187
+    },
+    {
+      "epoch": 0.04992033988316516,
+      "grad_norm": 0.28568945003091645,
+      "learning_rate": 4.960212201591512e-06,
+      "loss": 0.0058,
+      "step": 188
+    },
+    {
+      "epoch": 0.05018587360594796,
+      "grad_norm": 0.19319543123307042,
+      "learning_rate": 4.9867374005305045e-06,
+      "loss": 0.0033,
+      "step": 189
+    },
+    {
+      "epoch": 0.05045140732873075,
+      "grad_norm": 4.037258353685831,
+      "learning_rate": 5.013262599469496e-06,
+      "loss": 0.0815,
+      "step": 190
+    },
+    {
+      "epoch": 0.050716941051513545,
+      "grad_norm": 1.3356221226261589,
+      "learning_rate": 5.039787798408489e-06,
+      "loss": 0.0354,
+      "step": 191
+    },
+    {
+      "epoch": 0.050982474774296335,
+      "grad_norm": 0.1480129388225318,
+      "learning_rate": 5.06631299734748e-06,
+      "loss": 0.0026,
+      "step": 192
+    },
+    {
+      "epoch": 0.05124800849707913,
+      "grad_norm": 3.5661629795250045,
+      "learning_rate": 5.092838196286472e-06,
+      "loss": 0.2441,
+      "step": 193
+    },
+    {
+      "epoch": 0.05151354221986192,
+      "grad_norm": 2.1523745771682172,
+      "learning_rate": 5.119363395225465e-06,
+      "loss": 0.0576,
+      "step": 194
+    },
+    {
+      "epoch": 0.05177907594264471,
+      "grad_norm": 0.9192144869318637,
+      "learning_rate": 5.145888594164457e-06,
+      "loss": 0.0208,
+      "step": 195
+    },
+    {
+      "epoch": 0.05204460966542751,
+      "grad_norm": 0.2690072757879695,
+      "learning_rate": 5.172413793103449e-06,
+      "loss": 0.0064,
+      "step": 196
+    },
+    {
+      "epoch": 0.0523101433882103,
+      "grad_norm": 2.7717636534246792,
+      "learning_rate": 5.19893899204244e-06,
+      "loss": 0.0884,
+      "step": 197
+    },
+    {
+      "epoch": 0.0525756771109931,
+      "grad_norm": 7.0745463915342635,
+      "learning_rate": 5.225464190981433e-06,
+      "loss": 2.0156,
+      "step": 198
+    },
+    {
+      "epoch": 0.05284121083377589,
+      "grad_norm": 0.19227286962350826,
+      "learning_rate": 5.251989389920424e-06,
+      "loss": 0.0045,
+      "step": 199
+    },
+    {
+      "epoch": 0.053106744556558685,
+      "grad_norm": 0.8537355801992382,
+      "learning_rate": 5.278514588859417e-06,
+      "loss": 0.0069,
+      "step": 200
+    },
+    {
+      "epoch": 0.053372278279341476,
+      "grad_norm": 0.11929159424700997,
+      "learning_rate": 5.3050397877984095e-06,
+      "loss": 0.0025,
+      "step": 201
+    },
+    {
+      "epoch": 0.05363781200212427,
+      "grad_norm": 0.07790410641989416,
+      "learning_rate": 5.331564986737401e-06,
+      "loss": 0.0013,
+      "step": 202
+    },
+    {
+      "epoch": 0.05390334572490706,
+      "grad_norm": 0.13268954955742412,
+      "learning_rate": 5.358090185676394e-06,
+      "loss": 0.0032,
+      "step": 203
+    },
+    {
+      "epoch": 0.054168879447689854,
+      "grad_norm": 4.439535441327748,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 0.1836,
+      "step": 204
+    },
+    {
+      "epoch": 0.05443441317047265,
+      "grad_norm": 0.2688219627795949,
+      "learning_rate": 5.411140583554377e-06,
+      "loss": 0.0047,
+      "step": 205
+    },
+    {
+      "epoch": 0.05469994689325544,
+      "grad_norm": 0.08483073451016193,
+      "learning_rate": 5.437665782493369e-06,
+      "loss": 0.0019,
+      "step": 206
+    },
+    {
+      "epoch": 0.05496548061603824,
+      "grad_norm": 4.028626588799805,
+      "learning_rate": 5.4641909814323615e-06,
+      "loss": 0.3203,
+      "step": 207
+    },
+    {
+      "epoch": 0.05523101433882103,
+      "grad_norm": 2.4610886737988866,
+      "learning_rate": 5.490716180371354e-06,
+      "loss": 0.0688,
+      "step": 208
+    },
+    {
+      "epoch": 0.055496548061603826,
+      "grad_norm": 1.133315536600772,
+      "learning_rate": 5.517241379310345e-06,
+      "loss": 0.0258,
+      "step": 209
+    },
+    {
+      "epoch": 0.055762081784386616,
+      "grad_norm": 0.7726546713585315,
+      "learning_rate": 5.5437665782493376e-06,
+      "loss": 0.0152,
+      "step": 210
+    },
+    {
+      "epoch": 0.05602761550716941,
+      "grad_norm": 0.38371963709194457,
+      "learning_rate": 5.570291777188329e-06,
+      "loss": 0.0092,
+      "step": 211
+    },
+    {
+      "epoch": 0.056293149229952204,
+      "grad_norm": 0.13225755055257674,
+      "learning_rate": 5.596816976127322e-06,
+      "loss": 0.0023,
+      "step": 212
+    },
+    {
+      "epoch": 0.056558682952734994,
+      "grad_norm": 0.08917614772102375,
+      "learning_rate": 5.623342175066313e-06,
+      "loss": 0.002,
+      "step": 213
+    },
+    {
+      "epoch": 0.05682421667551779,
+      "grad_norm": 4.854086836177229,
+      "learning_rate": 5.649867374005305e-06,
+      "loss": 0.1035,
+      "step": 214
+    },
+    {
+      "epoch": 0.05708975039830058,
+      "grad_norm": 0.04007293390414322,
+      "learning_rate": 5.676392572944297e-06,
+      "loss": 0.0007,
+      "step": 215
+    },
+    {
+      "epoch": 0.05735528412108338,
+      "grad_norm": 0.2848845451163088,
+      "learning_rate": 5.70291777188329e-06,
+      "loss": 0.0064,
+      "step": 216
+    },
+    {
+      "epoch": 0.05762081784386617,
+      "grad_norm": 4.686394489350751,
+      "learning_rate": 5.729442970822282e-06,
+      "loss": 0.2246,
+      "step": 217
+    },
+    {
+      "epoch": 0.057886351566648966,
+      "grad_norm": 0.11120455969442722,
+      "learning_rate": 5.755968169761274e-06,
+      "loss": 0.0029,
+      "step": 218
+    },
+    {
+      "epoch": 0.05815188528943176,
+      "grad_norm": 4.663386143892064,
+      "learning_rate": 5.782493368700266e-06,
+      "loss": 0.1641,
+      "step": 219
+    },
+    {
+      "epoch": 0.058417419012214554,
+      "grad_norm": 0.24512762915181402,
+      "learning_rate": 5.809018567639257e-06,
+      "loss": 0.0036,
+      "step": 220
+    },
+    {
+      "epoch": 0.058682952734997344,
+      "grad_norm": 4.183804090865918,
+      "learning_rate": 5.83554376657825e-06,
+      "loss": 0.2148,
+      "step": 221
+    },
+    {
+      "epoch": 0.05894848645778014,
+      "grad_norm": 0.06396150026208348,
+      "learning_rate": 5.862068965517242e-06,
+      "loss": 0.0012,
+      "step": 222
+    },
+    {
+      "epoch": 0.05921402018056293,
+      "grad_norm": 0.1942578634333766,
+      "learning_rate": 5.888594164456234e-06,
+      "loss": 0.0037,
+      "step": 223
+    },
+    {
+      "epoch": 0.05947955390334572,
+      "grad_norm": 0.07594022266578823,
+      "learning_rate": 5.915119363395227e-06,
+      "loss": 0.002,
+      "step": 224
+    },
+    {
+      "epoch": 0.05974508762612852,
+      "grad_norm": 0.35116426830013603,
+      "learning_rate": 5.941644562334218e-06,
+      "loss": 0.0074,
+      "step": 225
+    },
+    {
+      "epoch": 0.06001062134891131,
+      "grad_norm": 1.578167800096405,
+      "learning_rate": 5.96816976127321e-06,
+      "loss": 0.0356,
+      "step": 226
+    },
+    {
+      "epoch": 0.06027615507169411,
+      "grad_norm": 1.1285595351705588,
+      "learning_rate": 5.994694960212202e-06,
+      "loss": 0.0255,
+      "step": 227
+    },
+    {
+      "epoch": 0.0605416887944769,
+      "grad_norm": 4.241743427696131,
+      "learning_rate": 6.0212201591511945e-06,
+      "loss": 0.4551,
+      "step": 228
+    },
+    {
+      "epoch": 0.060807222517259694,
+      "grad_norm": 7.930922977922742,
+      "learning_rate": 6.0477453580901854e-06,
+      "loss": 1.9453,
+      "step": 229
+    },
+    {
+      "epoch": 0.061072756240042485,
+      "grad_norm": 4.1445314479670925,
+      "learning_rate": 6.074270557029178e-06,
+      "loss": 0.4766,
+      "step": 230
+    },
+    {
+      "epoch": 0.06133828996282528,
+      "grad_norm": 0.022757441023624224,
+      "learning_rate": 6.1007957559681706e-06,
+      "loss": 0.0006,
+      "step": 231
+    },
+    {
+      "epoch": 0.06160382368560807,
+      "grad_norm": 0.04301908301005633,
+      "learning_rate": 6.127320954907162e-06,
+      "loss": 0.0012,
+      "step": 232
+    },
+    {
+      "epoch": 0.06186935740839086,
+      "grad_norm": 3.9993964088934497,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.5352,
+      "step": 233
+    },
+    {
+      "epoch": 0.06213489113117366,
+      "grad_norm": 0.0384831919189867,
+      "learning_rate": 6.180371352785146e-06,
+      "loss": 0.0011,
+      "step": 234
+    },
+    {
+      "epoch": 0.06240042485395645,
+      "grad_norm": 0.11825509016660113,
+      "learning_rate": 6.206896551724138e-06,
+      "loss": 0.0033,
+      "step": 235
+    },
+    {
+      "epoch": 0.06266595857673925,
+      "grad_norm": 0.038756173266160945,
+      "learning_rate": 6.23342175066313e-06,
+      "loss": 0.0011,
+      "step": 236
+    },
+    {
+      "epoch": 0.06293149229952204,
+      "grad_norm": 0.0527621505367311,
+      "learning_rate": 6.259946949602123e-06,
+      "loss": 0.0016,
+      "step": 237
+    },
+    {
+      "epoch": 0.06319702602230483,
+      "grad_norm": 3.4032460542743235,
+      "learning_rate": 6.286472148541115e-06,
+      "loss": 0.4199,
+      "step": 238
+    },
+    {
+      "epoch": 0.06346255974508763,
+      "grad_norm": 1.5026047777466172,
+      "learning_rate": 6.312997347480107e-06,
+      "loss": 0.006,
+      "step": 239
+    },
+    {
+      "epoch": 0.06372809346787042,
+      "grad_norm": 0.09772198768152848,
+      "learning_rate": 6.339522546419099e-06,
+      "loss": 0.0028,
+      "step": 240
+    },
+    {
+      "epoch": 0.06399362719065321,
+      "grad_norm": 3.4974570386194848,
+      "learning_rate": 6.36604774535809e-06,
+      "loss": 0.3203,
+      "step": 241
+    },
+    {
+      "epoch": 0.064259160913436,
+      "grad_norm": 0.05852008470517513,
+      "learning_rate": 6.392572944297083e-06,
+      "loss": 0.0017,
+      "step": 242
+    },
+    {
+      "epoch": 0.0645246946362188,
+      "grad_norm": 0.6206245208833137,
+      "learning_rate": 6.419098143236075e-06,
+      "loss": 0.0187,
+      "step": 243
+    },
+    {
+      "epoch": 0.0647902283590016,
+      "grad_norm": 3.257984380379068,
+      "learning_rate": 6.445623342175067e-06,
+      "loss": 0.2988,
+      "step": 244
+    },
+    {
+      "epoch": 0.06505576208178439,
+      "grad_norm": 3.129808256412029,
+      "learning_rate": 6.47214854111406e-06,
+      "loss": 0.3672,
+      "step": 245
+    },
+    {
+      "epoch": 0.06532129580456718,
+      "grad_norm": 1.0556539855013083,
+      "learning_rate": 6.498673740053051e-06,
+      "loss": 0.0258,
+      "step": 246
+    },
+    {
+      "epoch": 0.06558682952734997,
+      "grad_norm": 2.2148968544822747,
+      "learning_rate": 6.525198938992043e-06,
+      "loss": 0.1367,
+      "step": 247
+    },
+    {
+      "epoch": 0.06585236325013277,
+      "grad_norm": 0.3150693176270268,
+      "learning_rate": 6.551724137931035e-06,
+      "loss": 0.0103,
+      "step": 248
+    },
+    {
+      "epoch": 0.06611789697291556,
+      "grad_norm": 0.9150893306492274,
+      "learning_rate": 6.5782493368700276e-06,
+      "loss": 0.0356,
+      "step": 249
+    },
+    {
+      "epoch": 0.06638343069569835,
+      "grad_norm": 0.28640145943412554,
+      "learning_rate": 6.6047745358090184e-06,
+      "loss": 0.0092,
+      "step": 250
+    },
+    {
+      "epoch": 0.06664896441848114,
+      "grad_norm": 0.463345992401992,
+      "learning_rate": 6.631299734748011e-06,
+      "loss": 0.0165,
+      "step": 251
+    },
+    {
+      "epoch": 0.06691449814126393,
+      "grad_norm": 1.1399589305877227,
+      "learning_rate": 6.657824933687003e-06,
+      "loss": 0.0393,
+      "step": 252
+    },
+    {
+      "epoch": 0.06718003186404674,
+      "grad_norm": 2.377572235801305,
+      "learning_rate": 6.684350132625995e-06,
+      "loss": 0.1641,
+      "step": 253
+    },
+    {
+      "epoch": 0.06744556558682953,
+      "grad_norm": 0.20240560295679277,
+      "learning_rate": 6.710875331564988e-06,
+      "loss": 0.0053,
+      "step": 254
+    },
+    {
+      "epoch": 0.06771109930961232,
+      "grad_norm": 0.5580660153168894,
+      "learning_rate": 6.737400530503979e-06,
+      "loss": 0.015,
+      "step": 255
+    },
+    {
+      "epoch": 0.06797663303239511,
+      "grad_norm": 0.34949012939902974,
+      "learning_rate": 6.763925729442971e-06,
+      "loss": 0.0104,
+      "step": 256
+    },
+    {
+      "epoch": 0.06824216675517791,
+      "grad_norm": 1.5003862236593613,
+      "learning_rate": 6.790450928381963e-06,
+      "loss": 0.0688,
+      "step": 257
+    },
+    {
+      "epoch": 0.0685077004779607,
+      "grad_norm": 3.1203129924267152,
+      "learning_rate": 6.816976127320956e-06,
+      "loss": 0.2246,
+      "step": 258
+    },
+    {
+      "epoch": 0.0687732342007435,
+      "grad_norm": 1.051643999026055,
+      "learning_rate": 6.843501326259947e-06,
+      "loss": 0.0391,
+      "step": 259
+    },
+    {
+      "epoch": 0.06903876792352628,
+      "grad_norm": 0.19030670002007702,
+      "learning_rate": 6.87002652519894e-06,
+      "loss": 0.0058,
+      "step": 260
+    },
+    {
+      "epoch": 0.06930430164630907,
+      "grad_norm": 0.6772208181131325,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 0.0206,
+      "step": 261
+    },
+    {
+      "epoch": 0.06956983536909188,
+      "grad_norm": 2.2887370072088564,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 0.1035,
+      "step": 262
+    },
+    {
+      "epoch": 0.06983536909187467,
+      "grad_norm": 0.6143154696309661,
+      "learning_rate": 6.949602122015916e-06,
+      "loss": 0.0187,
+      "step": 263
+    },
+    {
+      "epoch": 0.07010090281465746,
+      "grad_norm": 2.9212908220956946,
+      "learning_rate": 6.976127320954908e-06,
+      "loss": 0.1641,
+      "step": 264
+    },
+    {
+      "epoch": 0.07036643653744025,
+      "grad_norm": 0.05930635612845523,
+      "learning_rate": 7.0026525198939e-06,
+      "loss": 0.0016,
+      "step": 265
+    },
+    {
+      "epoch": 0.07063197026022305,
+      "grad_norm": 0.21269981419083292,
+      "learning_rate": 7.029177718832891e-06,
+      "loss": 0.0045,
+      "step": 266
+    },
+    {
+      "epoch": 0.07089750398300584,
+      "grad_norm": 3.3491904755168838,
+      "learning_rate": 7.055702917771884e-06,
+      "loss": 0.2461,
+      "step": 267
+    },
+    {
+      "epoch": 0.07116303770578863,
+      "grad_norm": 3.011992184448476,
+      "learning_rate": 7.082228116710876e-06,
+      "loss": 0.2246,
+      "step": 268
+    },
+    {
+      "epoch": 0.07142857142857142,
+      "grad_norm": 3.317414315449119,
+      "learning_rate": 7.108753315649868e-06,
+      "loss": 0.2344,
+      "step": 269
+    },
+    {
+      "epoch": 0.07169410515135423,
+      "grad_norm": 0.3112825686788188,
+      "learning_rate": 7.1352785145888606e-06,
+      "loss": 0.0093,
+      "step": 270
+    },
+    {
+      "epoch": 0.07195963887413702,
+      "grad_norm": 0.05440553438613368,
+      "learning_rate": 7.1618037135278515e-06,
+      "loss": 0.0017,
+      "step": 271
+    },
+    {
+      "epoch": 0.07222517259691981,
+      "grad_norm": 2.1971208118181416,
+      "learning_rate": 7.188328912466844e-06,
+      "loss": 0.1367,
+      "step": 272
+    },
+    {
+      "epoch": 0.0724907063197026,
+      "grad_norm": 12.646903729517486,
+      "learning_rate": 7.214854111405836e-06,
+      "loss": 1.625,
+      "step": 273
+    },
+    {
+      "epoch": 0.07275624004248539,
+      "grad_norm": 0.5134455047996075,
+      "learning_rate": 7.241379310344828e-06,
+      "loss": 0.0166,
+      "step": 274
+    },
+    {
+      "epoch": 0.0730217737652682,
+      "grad_norm": 0.7126267670034715,
+      "learning_rate": 7.267904509283821e-06,
+      "loss": 0.0288,
+      "step": 275
+    },
+    {
+      "epoch": 0.07328730748805098,
+      "grad_norm": 2.247754974205674,
+      "learning_rate": 7.294429708222812e-06,
+      "loss": 0.1367,
+      "step": 276
+    },
+    {
+      "epoch": 0.07355284121083377,
+      "grad_norm": 2.0961429578052546,
+      "learning_rate": 7.320954907161804e-06,
+      "loss": 0.0967,
+      "step": 277
+    },
+    {
+      "epoch": 0.07381837493361657,
+      "grad_norm": 1.5658452401665721,
+      "learning_rate": 7.347480106100796e-06,
+      "loss": 0.0757,
+      "step": 278
+    },
+    {
+      "epoch": 0.07408390865639937,
+      "grad_norm": 1.3474463383395376,
+      "learning_rate": 7.374005305039789e-06,
+      "loss": 0.0635,
+      "step": 279
+    },
+    {
+      "epoch": 0.07434944237918216,
+      "grad_norm": 0.9079733138650365,
+      "learning_rate": 7.40053050397878e-06,
+      "loss": 0.0354,
+      "step": 280
+    },
+    {
+      "epoch": 0.07461497610196495,
+      "grad_norm": 1.423025471379327,
+      "learning_rate": 7.427055702917773e-06,
+      "loss": 0.0579,
+      "step": 281
+    },
+    {
+      "epoch": 0.07488050982474774,
+      "grad_norm": 29.742858003866257,
+      "learning_rate": 7.453580901856765e-06,
+      "loss": 2.2188,
+      "step": 282
+    },
+    {
+      "epoch": 0.07514604354753053,
+      "grad_norm": 1.1822814883714525,
+      "learning_rate": 7.480106100795756e-06,
+      "loss": 0.0579,
+      "step": 283
+    },
+    {
+      "epoch": 0.07541157727031333,
+      "grad_norm": 1.183988881108207,
+      "learning_rate": 7.506631299734749e-06,
+      "loss": 0.053,
+      "step": 284
+    },
+    {
+      "epoch": 0.07567711099309612,
+      "grad_norm": 1.2925771262928738,
+      "learning_rate": 7.533156498673741e-06,
+      "loss": 0.0579,
+      "step": 285
+    },
+    {
+      "epoch": 0.07594264471587892,
+      "grad_norm": 0.6760998544572231,
+      "learning_rate": 7.559681697612733e-06,
+      "loss": 0.0208,
+      "step": 286
+    },
+    {
+      "epoch": 0.0762081784386617,
+      "grad_norm": 1.0282421815983582,
+      "learning_rate": 7.586206896551724e-06,
+      "loss": 0.0396,
+      "step": 287
+    },
+    {
+      "epoch": 0.07647371216144451,
+      "grad_norm": 0.4234422853011549,
+      "learning_rate": 7.612732095490717e-06,
+      "loss": 0.015,
+      "step": 288
+    },
+    {
+      "epoch": 0.0767392458842273,
+      "grad_norm": 0.39859873725327066,
+      "learning_rate": 7.639257294429708e-06,
+      "loss": 0.0135,
+      "step": 289
+    },
+    {
+      "epoch": 0.07700477960701009,
+      "grad_norm": 0.16222305800718434,
+      "learning_rate": 7.6657824933687e-06,
+      "loss": 0.0052,
+      "step": 290
+    },
+    {
+      "epoch": 0.07727031332979288,
+      "grad_norm": 0.11630227304441901,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.0038,
+      "step": 291
+    },
+    {
+      "epoch": 0.07753584705257567,
+      "grad_norm": 0.03372528156970035,
+      "learning_rate": 7.718832891246685e-06,
+      "loss": 0.0008,
+      "step": 292
+    },
+    {
+      "epoch": 0.07780138077535848,
+      "grad_norm": 0.051940974929390066,
+      "learning_rate": 7.745358090185677e-06,
+      "loss": 0.0015,
+      "step": 293
+    },
+    {
+      "epoch": 0.07806691449814127,
+      "grad_norm": 3.0783353845274295,
+      "learning_rate": 7.771883289124669e-06,
+      "loss": 0.2988,
+      "step": 294
+    },
+    {
+      "epoch": 0.07833244822092406,
+      "grad_norm": 3.19288706413644,
+      "learning_rate": 7.79840848806366e-06,
+      "loss": 1.5,
+      "step": 295
+    },
+    {
+      "epoch": 0.07859798194370685,
+      "grad_norm": 0.043571411993144044,
+      "learning_rate": 7.824933687002652e-06,
+      "loss": 0.0012,
+      "step": 296
+    },
+    {
+      "epoch": 0.07886351566648965,
+      "grad_norm": 0.022251985441288053,
+      "learning_rate": 7.851458885941646e-06,
+      "loss": 0.0006,
+      "step": 297
+    },
+    {
+      "epoch": 0.07912904938927244,
+      "grad_norm": 0.08604971258954734,
+      "learning_rate": 7.877984084880637e-06,
+      "loss": 0.002,
+      "step": 298
+    },
+    {
+      "epoch": 0.07939458311205523,
+      "grad_norm": 0.019663492930382675,
+      "learning_rate": 7.904509283819629e-06,
+      "loss": 0.0007,
+      "step": 299
+    },
+    {
+      "epoch": 0.07966011683483802,
+      "grad_norm": 4.060927451869871,
+      "learning_rate": 7.93103448275862e-06,
+      "loss": 0.5469,
+      "step": 300
+    },
+    {
+      "epoch": 0.07992565055762081,
+      "grad_norm": 3.7022259212796285,
+      "learning_rate": 7.957559681697613e-06,
+      "loss": 0.5,
+      "step": 301
+    },
+    {
+      "epoch": 0.08019118428040362,
+      "grad_norm": 0.009537116683502751,
+      "learning_rate": 7.984084880636606e-06,
+      "loss": 0.0002,
+      "step": 302
+    },
+    {
+      "epoch": 0.0804567180031864,
+      "grad_norm": 0.15972134226047616,
+      "learning_rate": 8.010610079575598e-06,
+      "loss": 0.0013,
+      "step": 303
+    },
+    {
+      "epoch": 0.0807222517259692,
+      "grad_norm": 3.667758151535006,
+      "learning_rate": 8.03713527851459e-06,
+      "loss": 0.3203,
+      "step": 304
+    },
+    {
+      "epoch": 0.08098778544875199,
+      "grad_norm": 0.11182970657148031,
+      "learning_rate": 8.063660477453583e-06,
+      "loss": 0.0036,
+      "step": 305
+    },
+    {
+      "epoch": 0.08125331917153479,
+      "grad_norm": 2.9682235846476854,
+      "learning_rate": 8.090185676392573e-06,
+      "loss": 0.3203,
+      "step": 306
+    },
+    {
+      "epoch": 0.08151885289431758,
+      "grad_norm": 2.6544777345452233,
+      "learning_rate": 8.116710875331566e-06,
+      "loss": 0.2148,
+      "step": 307
+    },
+    {
+      "epoch": 0.08178438661710037,
+      "grad_norm": 2.23551129108633,
+      "learning_rate": 8.143236074270558e-06,
+      "loss": 0.2129,
+      "step": 308
+    },
+    {
+      "epoch": 0.08204992033988316,
+      "grad_norm": 12.154843924500748,
+      "learning_rate": 8.16976127320955e-06,
+      "loss": 1.6797,
+      "step": 309
+    },
+    {
+      "epoch": 0.08231545406266595,
+      "grad_norm": 0.15211428151282252,
+      "learning_rate": 8.196286472148541e-06,
+      "loss": 0.006,
+      "step": 310
+    },
+    {
+      "epoch": 0.08258098778544876,
+      "grad_norm": 1.8572071183861802,
+      "learning_rate": 8.222811671087533e-06,
+      "loss": 0.1641,
+      "step": 311
+    },
+    {
+      "epoch": 0.08284652150823155,
+      "grad_norm": 0.04756323667504141,
+      "learning_rate": 8.249336870026527e-06,
+      "loss": 0.0015,
+      "step": 312
+    },
+    {
+      "epoch": 0.08311205523101434,
+      "grad_norm": 1.2740195810197956,
+      "learning_rate": 8.275862068965518e-06,
+      "loss": 0.0752,
+      "step": 313
+    },
+    {
+      "epoch": 0.08337758895379713,
+      "grad_norm": 2.2505964862303536,
+      "learning_rate": 8.30238726790451e-06,
+      "loss": 0.1553,
+      "step": 314
+    },
+    {
+      "epoch": 0.08364312267657993,
+      "grad_norm": 1.2815168496791627,
+      "learning_rate": 8.328912466843502e-06,
+      "loss": 0.0359,
+      "step": 315
+    },
+    {
+      "epoch": 0.08390865639936272,
+      "grad_norm": 0.6640260863661916,
+      "learning_rate": 8.355437665782494e-06,
+      "loss": 0.0356,
+      "step": 316
+    },
+    {
+      "epoch": 0.08417419012214551,
+      "grad_norm": 1.5201761587838876,
+      "learning_rate": 8.381962864721485e-06,
+      "loss": 0.1123,
+      "step": 317
+    },
+    {
+      "epoch": 0.0844397238449283,
+      "grad_norm": 8.490220248388562,
+      "learning_rate": 8.408488063660479e-06,
+      "loss": 1.625,
+      "step": 318
+    },
+    {
+      "epoch": 0.08470525756771109,
+      "grad_norm": 1.220932640130876,
+      "learning_rate": 8.43501326259947e-06,
+      "loss": 0.0762,
+      "step": 319
+    },
+    {
+      "epoch": 0.0849707912904939,
+      "grad_norm": 1.5989309966218423,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 0.1191,
+      "step": 320
+    },
+    {
+      "epoch": 0.08523632501327669,
+      "grad_norm": 0.9427180577076081,
+      "learning_rate": 8.488063660477454e-06,
+      "loss": 0.0583,
+      "step": 321
+    },
+    {
+      "epoch": 0.08550185873605948,
+      "grad_norm": 0.5733118835063594,
+      "learning_rate": 8.514588859416446e-06,
+      "loss": 0.0293,
+      "step": 322
+    },
+    {
+      "epoch": 0.08576739245884227,
+      "grad_norm": 0.14144397139040923,
+      "learning_rate": 8.541114058355439e-06,
+      "loss": 0.0066,
+      "step": 323
+    },
+    {
+      "epoch": 0.08603292618162507,
+      "grad_norm": 2.2159972152363077,
+      "learning_rate": 8.56763925729443e-06,
+      "loss": 0.2676,
+      "step": 324
+    },
+    {
+      "epoch": 0.08629845990440786,
+      "grad_norm": 1.6089684506296742,
+      "learning_rate": 8.594164456233422e-06,
+      "loss": 0.165,
+      "step": 325
+    },
+    {
+      "epoch": 0.08656399362719065,
+      "grad_norm": 1.907931611057086,
+      "learning_rate": 8.620689655172414e-06,
+      "loss": 0.2246,
+      "step": 326
+    },
+    {
+      "epoch": 0.08682952734997344,
+      "grad_norm": 1.3090038245570277,
+      "learning_rate": 8.647214854111406e-06,
+      "loss": 0.0894,
+      "step": 327
+    },
+    {
+      "epoch": 0.08709506107275625,
+      "grad_norm": 0.32857189533914255,
+      "learning_rate": 8.6737400530504e-06,
+      "loss": 0.0167,
+      "step": 328
+    },
+    {
+      "epoch": 0.08736059479553904,
+      "grad_norm": 1.6452882577007435,
+      "learning_rate": 8.700265251989391e-06,
+      "loss": 0.1934,
+      "step": 329
+    },
+    {
+      "epoch": 0.08762612851832183,
+      "grad_norm": 0.9569297269958028,
+      "learning_rate": 8.726790450928383e-06,
+      "loss": 0.0583,
+      "step": 330
+    },
+    {
+      "epoch": 0.08789166224110462,
+      "grad_norm": 1.5135689038875453,
+      "learning_rate": 8.753315649867374e-06,
+      "loss": 0.1641,
+      "step": 331
+    },
+    {
+      "epoch": 0.08815719596388741,
+      "grad_norm": 1.171616362989181,
+      "learning_rate": 8.779840848806366e-06,
+      "loss": 0.0967,
+      "step": 332
+    },
+    {
+      "epoch": 0.08842272968667021,
+      "grad_norm": 1.296129054455011,
+      "learning_rate": 8.806366047745358e-06,
+      "loss": 0.1279,
+      "step": 333
+    },
+    {
+      "epoch": 0.088688263409453,
+      "grad_norm": 1.7346960331999244,
+      "learning_rate": 8.832891246684351e-06,
+      "loss": 0.2031,
+      "step": 334
+    },
+    {
+      "epoch": 0.08895379713223579,
+      "grad_norm": 1.0461737141624319,
+      "learning_rate": 8.859416445623343e-06,
+      "loss": 0.0894,
+      "step": 335
+    },
+    {
+      "epoch": 0.08921933085501858,
+      "grad_norm": 0.6479647062091533,
+      "learning_rate": 8.885941644562335e-06,
+      "loss": 0.0439,
+      "step": 336
+    },
+    {
+      "epoch": 0.08948486457780139,
+      "grad_norm": 0.734572554709422,
+      "learning_rate": 8.912466843501327e-06,
+      "loss": 0.0479,
+      "step": 337
+    },
+    {
+      "epoch": 0.08975039830058418,
+      "grad_norm": 0.37460636172952666,
+      "learning_rate": 8.938992042440318e-06,
+      "loss": 0.0239,
+      "step": 338
+    },
+    {
+      "epoch": 0.09001593202336697,
+      "grad_norm": 0.7525268871585711,
+      "learning_rate": 8.965517241379312e-06,
+      "loss": 0.0579,
+      "step": 339
+    },
+    {
+      "epoch": 0.09028146574614976,
+      "grad_norm": 0.8611894174379208,
+      "learning_rate": 8.992042440318303e-06,
+      "loss": 0.0757,
+      "step": 340
+    },
+    {
+      "epoch": 0.09054699946893255,
+      "grad_norm": 0.7524042144808409,
+      "learning_rate": 9.018567639257295e-06,
+      "loss": 0.0579,
+      "step": 341
+    },
+    {
+      "epoch": 0.09081253319171535,
+      "grad_norm": 0.7389515011548652,
+      "learning_rate": 9.045092838196289e-06,
+      "loss": 0.0579,
+      "step": 342
+    },
+    {
+      "epoch": 0.09107806691449814,
+      "grad_norm": 0.717809617473176,
+      "learning_rate": 9.071618037135279e-06,
+      "loss": 0.0583,
+      "step": 343
+    },
+    {
+      "epoch": 0.09134360063728093,
+      "grad_norm": 3.9747880808948737,
+      "learning_rate": 9.098143236074272e-06,
+      "loss": 1.2812,
+      "step": 344
+    },
+    {
+      "epoch": 0.09160913436006372,
+      "grad_norm": 0.5440174756085783,
+      "learning_rate": 9.124668435013264e-06,
+      "loss": 0.0396,
+      "step": 345
+    },
+    {
+      "epoch": 0.09187466808284653,
+      "grad_norm": 0.5421286985756604,
+      "learning_rate": 9.151193633952255e-06,
+      "loss": 0.04,
+      "step": 346
+    },
+    {
+      "epoch": 0.09214020180562932,
+      "grad_norm": 0.7819761791319009,
+      "learning_rate": 9.177718832891247e-06,
+      "loss": 0.0698,
+      "step": 347
+    },
+    {
+      "epoch": 0.09240573552841211,
+      "grad_norm": 0.6347909735793549,
+      "learning_rate": 9.204244031830239e-06,
+      "loss": 0.0532,
+      "step": 348
+    },
+    {
+      "epoch": 0.0926712692511949,
+      "grad_norm": 6.09104951580698,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 1.3984,
+      "step": 349
+    },
+    {
+      "epoch": 0.09293680297397769,
+      "grad_norm": 1.041426014412066,
+      "learning_rate": 9.257294429708224e-06,
+      "loss": 0.1045,
+      "step": 350
+    },
+    {
+      "epoch": 0.0932023366967605,
+      "grad_norm": 0.47301291400697715,
+      "learning_rate": 9.283819628647216e-06,
+      "loss": 0.027,
+      "step": 351
+    },
+    {
+      "epoch": 0.09346787041954328,
+      "grad_norm": 0.8295019698863545,
+      "learning_rate": 9.310344827586207e-06,
+      "loss": 0.0693,
+      "step": 352
+    },
+    {
+      "epoch": 0.09373340414232607,
+      "grad_norm": 0.5101258233711066,
+      "learning_rate": 9.3368700265252e-06,
+      "loss": 0.0361,
+      "step": 353
+    },
+    {
+      "epoch": 0.09399893786510886,
+      "grad_norm": 0.846504832852156,
+      "learning_rate": 9.363395225464191e-06,
+      "loss": 0.0757,
+      "step": 354
+    },
+    {
+      "epoch": 0.09426447158789167,
+      "grad_norm": 1.228530334987593,
+      "learning_rate": 9.389920424403184e-06,
+      "loss": 0.0825,
+      "step": 355
+    },
+    {
+      "epoch": 0.09453000531067446,
+      "grad_norm": 1.0130610331102055,
+      "learning_rate": 9.416445623342176e-06,
+      "loss": 0.0903,
+      "step": 356
+    },
+    {
+      "epoch": 0.09479553903345725,
+      "grad_norm": 3.315689651222163,
+      "learning_rate": 9.442970822281168e-06,
+      "loss": 0.0698,
+      "step": 357
+    },
+    {
+      "epoch": 0.09506107275624004,
+      "grad_norm": 1.1215958152396903,
+      "learning_rate": 9.46949602122016e-06,
+      "loss": 0.0967,
+      "step": 358
+    },
+    {
+      "epoch": 0.09532660647902283,
+      "grad_norm": 0.9299426765923084,
+      "learning_rate": 9.496021220159151e-06,
+      "loss": 0.0767,
+      "step": 359
+    },
+    {
+      "epoch": 0.09559214020180563,
+      "grad_norm": 0.7265450766257645,
+      "learning_rate": 9.522546419098145e-06,
+      "loss": 0.0588,
+      "step": 360
+    },
+    {
+      "epoch": 0.09585767392458842,
+      "grad_norm": 0.8861616356173306,
+      "learning_rate": 9.549071618037136e-06,
+      "loss": 0.0894,
+      "step": 361
+    },
+    {
+      "epoch": 0.09612320764737121,
+      "grad_norm": 0.5801872302587202,
+      "learning_rate": 9.575596816976128e-06,
+      "loss": 0.0398,
+      "step": 362
+    },
+    {
+      "epoch": 0.096388741370154,
+      "grad_norm": 0.8463320496115545,
+      "learning_rate": 9.60212201591512e-06,
+      "loss": 0.0698,
+      "step": 363
+    },
+    {
+      "epoch": 0.09665427509293681,
+      "grad_norm": 1.0159728812303754,
+      "learning_rate": 9.628647214854112e-06,
+      "loss": 0.0825,
+      "step": 364
+    },
+    {
+      "epoch": 0.0969198088157196,
+      "grad_norm": 0.716278656041375,
+      "learning_rate": 9.655172413793105e-06,
+      "loss": 0.064,
+      "step": 365
+    },
+    {
+      "epoch": 0.09718534253850239,
+      "grad_norm": 0.9734713463730997,
+      "learning_rate": 9.681697612732097e-06,
+      "loss": 0.1035,
+      "step": 366
+    },
+    {
+      "epoch": 0.09745087626128518,
+      "grad_norm": 0.6573062869557463,
+      "learning_rate": 9.708222811671088e-06,
+      "loss": 0.0583,
+      "step": 367
+    },
+    {
+      "epoch": 0.09771640998406797,
+      "grad_norm": 0.4000747396309669,
+      "learning_rate": 9.73474801061008e-06,
+      "loss": 0.0259,
+      "step": 368
+    },
+    {
+      "epoch": 0.09798194370685077,
+      "grad_norm": 0.7259390646225368,
+      "learning_rate": 9.761273209549072e-06,
+      "loss": 0.0698,
+      "step": 369
+    },
+    {
+      "epoch": 0.09824747742963356,
+      "grad_norm": 0.5866962892418395,
+      "learning_rate": 9.787798408488064e-06,
+      "loss": 0.0439,
+      "step": 370
+    },
+    {
+      "epoch": 0.09851301115241635,
+      "grad_norm": 0.5114555343916287,
+      "learning_rate": 9.814323607427057e-06,
+      "loss": 0.0354,
+      "step": 371
+    },
+    {
+      "epoch": 0.09877854487519915,
+      "grad_norm": 0.6409404463637564,
+      "learning_rate": 9.840848806366049e-06,
+      "loss": 0.0525,
+      "step": 372
+    },
+    {
+      "epoch": 0.09904407859798195,
+      "grad_norm": 1.1219844119808053,
+      "learning_rate": 9.86737400530504e-06,
+      "loss": 0.1206,
+      "step": 373
+    },
+    {
+      "epoch": 0.09930961232076474,
+      "grad_norm": 0.553871104594828,
+      "learning_rate": 9.893899204244032e-06,
+      "loss": 0.0396,
+      "step": 374
+    },
+    {
+      "epoch": 0.09957514604354753,
+      "grad_norm": 0.9103969438290072,
+      "learning_rate": 9.920424403183024e-06,
+      "loss": 0.0825,
+      "step": 375
+    },
+    {
+      "epoch": 0.09984067976633032,
+      "grad_norm": 0.6220106264544292,
+      "learning_rate": 9.946949602122017e-06,
+      "loss": 0.053,
+      "step": 376
+    },
+    {
+      "epoch": 0.10010621348911311,
+      "grad_norm": 0.35674878941499344,
+      "learning_rate": 9.973474801061009e-06,
+      "loss": 0.0231,
+      "step": 377
+    },
+    {
+      "epoch": 0.10037174721189591,
+      "grad_norm": 0.829605031932029,
+      "learning_rate": 1e-05,
+      "loss": 0.0698,
+      "step": 378
+    },
+    {
+      "epoch": 0.1006372809346787,
+      "grad_norm": 5.884980972558981,
+      "learning_rate": 9.999999518029487e-06,
+      "loss": 1.0781,
+      "step": 379
+    },
+    {
+      "epoch": 0.1009028146574615,
+      "grad_norm": 0.5316843014168648,
+      "learning_rate": 9.999998072118034e-06,
+      "loss": 0.0396,
+      "step": 380
+    },
+    {
+      "epoch": 0.10116834838024429,
+      "grad_norm": 0.2781221112512736,
+      "learning_rate": 9.99999566226592e-06,
+      "loss": 0.0165,
+      "step": 381
+    },
+    {
+      "epoch": 0.10143388210302709,
+      "grad_norm": 0.48275929349374064,
+      "learning_rate": 9.999992288473615e-06,
+      "loss": 0.0393,
+      "step": 382
+    },
+    {
+      "epoch": 0.10169941582580988,
+      "grad_norm": 0.6278039868358899,
+      "learning_rate": 9.999987950741766e-06,
+      "loss": 0.0354,
+      "step": 383
+    },
+    {
+      "epoch": 0.10196494954859267,
+      "grad_norm": 0.6003466758282822,
+      "learning_rate": 9.99998264907121e-06,
+      "loss": 0.0361,
+      "step": 384
+    },
+    {
+      "epoch": 0.10223048327137546,
+      "grad_norm": 0.8912911423947107,
+      "learning_rate": 9.999976383462966e-06,
+      "loss": 0.0579,
+      "step": 385
+    },
+    {
+      "epoch": 0.10249601699415826,
+      "grad_norm": 0.7329391229020974,
+      "learning_rate": 9.999969153918246e-06,
+      "loss": 0.0435,
+      "step": 386
+    },
+    {
+      "epoch": 0.10276155071694106,
+      "grad_norm": 0.41829397558306736,
+      "learning_rate": 9.999960960438441e-06,
+      "loss": 0.0288,
+      "step": 387
+    },
+    {
+      "epoch": 0.10302708443972385,
+      "grad_norm": 0.9968145270135951,
+      "learning_rate": 9.999951803025134e-06,
+      "loss": 0.0757,
+      "step": 388
+    },
+    {
+      "epoch": 0.10329261816250664,
+      "grad_norm": 0.7676728389082504,
+      "learning_rate": 9.999941681680087e-06,
+      "loss": 0.053,
+      "step": 389
+    },
+    {
+      "epoch": 0.10355815188528943,
+      "grad_norm": 0.31126746105326625,
+      "learning_rate": 9.999930596405254e-06,
+      "loss": 0.0184,
+      "step": 390
+    },
+    {
+      "epoch": 0.10382368560807223,
+      "grad_norm": 0.5108289920792247,
+      "learning_rate": 9.999918547202769e-06,
+      "loss": 0.0354,
+      "step": 391
+    },
+    {
+      "epoch": 0.10408921933085502,
+      "grad_norm": 0.13796497678169306,
+      "learning_rate": 9.999905534074959e-06,
+      "loss": 0.0041,
+      "step": 392
+    },
+    {
+      "epoch": 0.10435475305363781,
+      "grad_norm": 0.7020072318414768,
+      "learning_rate": 9.999891557024329e-06,
+      "loss": 0.0435,
+      "step": 393
+    },
+    {
+      "epoch": 0.1046202867764206,
+      "grad_norm": 0.768122730327616,
+      "learning_rate": 9.999876616053574e-06,
+      "loss": 0.0435,
+      "step": 394
+    },
+    {
+      "epoch": 0.1048858204992034,
+      "grad_norm": 0.3543183385660938,
+      "learning_rate": 9.999860711165577e-06,
+      "loss": 0.0167,
+      "step": 395
+    },
+    {
+      "epoch": 0.1051513542219862,
+      "grad_norm": 0.23030281632455404,
+      "learning_rate": 9.999843842363402e-06,
+      "loss": 0.0115,
+      "step": 396
+    },
+    {
+      "epoch": 0.10541688794476899,
+      "grad_norm": 2.039085244924252,
+      "learning_rate": 9.999826009650302e-06,
+      "loss": 0.1279,
+      "step": 397
+    },
+    {
+      "epoch": 0.10568242166755178,
+      "grad_norm": 0.6863471545307963,
+      "learning_rate": 9.999807213029716e-06,
+      "loss": 0.0352,
+      "step": 398
+    },
+    {
+      "epoch": 0.10594795539033457,
+      "grad_norm": 0.2210748545563893,
+      "learning_rate": 9.999787452505265e-06,
+      "loss": 0.0103,
+      "step": 399
+    },
+    {
+      "epoch": 0.10621348911311737,
+      "grad_norm": 0.5710642294027447,
+      "learning_rate": 9.99976672808076e-06,
+      "loss": 0.0286,
+      "step": 400
+    },
+    {
+      "epoch": 0.10647902283590016,
+      "grad_norm": 1.681840306945433,
+      "learning_rate": 9.999745039760197e-06,
+      "loss": 0.1367,
+      "step": 401
+    },
+    {
+      "epoch": 0.10674455655868295,
+      "grad_norm": 0.5509992577306293,
+      "learning_rate": 9.999722387547759e-06,
+      "loss": 0.0076,
+      "step": 402
+    },
+    {
+      "epoch": 0.10701009028146574,
+      "grad_norm": 0.8731295387203799,
+      "learning_rate": 9.999698771447808e-06,
+      "loss": 0.0258,
+      "step": 403
+    },
+    {
+      "epoch": 0.10727562400424855,
+      "grad_norm": 0.09492425648826142,
+      "learning_rate": 9.9996741914649e-06,
+      "loss": 0.0025,
+      "step": 404
+    },
+    {
+      "epoch": 0.10754115772703134,
+      "grad_norm": 3.5100388797466575,
+      "learning_rate": 9.999648647603774e-06,
+      "loss": 0.332,
+      "step": 405
+    },
+    {
+      "epoch": 0.10780669144981413,
+      "grad_norm": 1.412486941835509,
+      "learning_rate": 9.999622139869356e-06,
+      "loss": 0.0688,
+      "step": 406
+    },
+    {
+      "epoch": 0.10807222517259692,
+      "grad_norm": 0.4019838803661139,
+      "learning_rate": 9.999594668266754e-06,
+      "loss": 0.0184,
+      "step": 407
+    },
+    {
+      "epoch": 0.10833775889537971,
+      "grad_norm": 0.10033143456094161,
+      "learning_rate": 9.999566232801263e-06,
+      "loss": 0.0006,
+      "step": 408
+    },
+    {
+      "epoch": 0.10860329261816251,
+      "grad_norm": 2.671317904601799,
+      "learning_rate": 9.999536833478367e-06,
+      "loss": 0.2031,
+      "step": 409
+    },
+    {
+      "epoch": 0.1088688263409453,
+      "grad_norm": 1.4428641692398059,
+      "learning_rate": 9.999506470303736e-06,
+      "loss": 0.0889,
+      "step": 410
+    },
+    {
+      "epoch": 0.10913436006372809,
+      "grad_norm": 0.8240991675135017,
+      "learning_rate": 9.999475143283219e-06,
+      "loss": 0.032,
+      "step": 411
+    },
+    {
+      "epoch": 0.10939989378651088,
+      "grad_norm": 0.6500544444722121,
+      "learning_rate": 9.999442852422858e-06,
+      "loss": 0.0286,
+      "step": 412
+    },
+    {
+      "epoch": 0.10966542750929369,
+      "grad_norm": 0.43616099112846474,
+      "learning_rate": 9.999409597728878e-06,
+      "loss": 0.0165,
+      "step": 413
+    },
+    {
+      "epoch": 0.10993096123207648,
+      "grad_norm": 0.4976424228594336,
+      "learning_rate": 9.99937537920769e-06,
+      "loss": 0.0229,
+      "step": 414
+    },
+    {
+      "epoch": 0.11019649495485927,
+      "grad_norm": 0.1461188597520504,
+      "learning_rate": 9.99934019686589e-06,
+      "loss": 0.0059,
+      "step": 415
+    },
+    {
+      "epoch": 0.11046202867764206,
+      "grad_norm": 2.5259359528371155,
+      "learning_rate": 9.999304050710264e-06,
+      "loss": 0.2461,
+      "step": 416
+    },
+    {
+      "epoch": 0.11072756240042485,
+      "grad_norm": 0.020285773595493026,
+      "learning_rate": 9.999266940747776e-06,
+      "loss": 0.0006,
+      "step": 417
+    },
+    {
+      "epoch": 0.11099309612320765,
+      "grad_norm": 2.126872840900898,
+      "learning_rate": 9.999228866985585e-06,
+      "loss": 0.1367,
+      "step": 418
+    },
+    {
+      "epoch": 0.11125862984599044,
+      "grad_norm": 0.10900902708813702,
+      "learning_rate": 9.999189829431029e-06,
+      "loss": 0.0035,
+      "step": 419
+    },
+    {
+      "epoch": 0.11152416356877323,
+      "grad_norm": 1.6722307761042388,
+      "learning_rate": 9.999149828091632e-06,
+      "loss": 0.1279,
+      "step": 420
+    },
+    {
+      "epoch": 0.11178969729155602,
+      "grad_norm": 0.3758020820014381,
+      "learning_rate": 9.99910886297511e-06,
+      "loss": 0.0164,
+      "step": 421
+    },
+    {
+      "epoch": 0.11205523101433883,
+      "grad_norm": 0.9309889664471976,
+      "learning_rate": 9.999066934089355e-06,
+      "loss": 0.0476,
+      "step": 422
+    },
+    {
+      "epoch": 0.11232076473712162,
+      "grad_norm": 1.9032139694898287,
+      "learning_rate": 9.999024041442455e-06,
+      "loss": 0.1641,
+      "step": 423
+    },
+    {
+      "epoch": 0.11258629845990441,
+      "grad_norm": 1.9393415053104626,
+      "learning_rate": 9.998980185042678e-06,
+      "loss": 0.1367,
+      "step": 424
+    },
+    {
+      "epoch": 0.1128518321826872,
+      "grad_norm": 1.7335978068394446,
+      "learning_rate": 9.99893536489848e-06,
+      "loss": 0.1553,
+      "step": 425
+    },
+    {
+      "epoch": 0.11311736590546999,
+      "grad_norm": 0.964411541181811,
+      "learning_rate": 9.998889581018498e-06,
+      "loss": 0.0525,
+      "step": 426
+    },
+    {
+      "epoch": 0.11338289962825279,
+      "grad_norm": 0.9732904594255166,
+      "learning_rate": 9.998842833411565e-06,
+      "loss": 0.0479,
+      "step": 427
+    },
+    {
+      "epoch": 0.11364843335103558,
+      "grad_norm": 0.3840885691796018,
+      "learning_rate": 9.998795122086687e-06,
+      "loss": 0.0184,
+      "step": 428
+    },
+    {
+      "epoch": 0.11391396707381837,
+      "grad_norm": 0.47672162986825717,
+      "learning_rate": 9.998746447053063e-06,
+      "loss": 0.0259,
+      "step": 429
+    },
+    {
+      "epoch": 0.11417950079660116,
+      "grad_norm": 0.06479878514865638,
+      "learning_rate": 9.998696808320083e-06,
+      "loss": 0.0018,
+      "step": 430
+    },
+    {
+      "epoch": 0.11444503451938397,
+      "grad_norm": 0.8617385798684972,
+      "learning_rate": 9.99864620589731e-06,
+      "loss": 0.0576,
+      "step": 431
+    },
+    {
+      "epoch": 0.11471056824216676,
+      "grad_norm": 0.25886899017260007,
+      "learning_rate": 9.998594639794502e-06,
+      "loss": 0.0095,
+      "step": 432
+    },
+    {
+      "epoch": 0.11497610196494955,
+      "grad_norm": 0.6784186872332681,
+      "learning_rate": 9.9985421100216e-06,
+      "loss": 0.0393,
+      "step": 433
+    },
+    {
+      "epoch": 0.11524163568773234,
+      "grad_norm": 1.7431553785848855,
+      "learning_rate": 9.998488616588733e-06,
+      "loss": 0.1455,
+      "step": 434
+    },
+    {
+      "epoch": 0.11550716941051513,
+      "grad_norm": 1.9060354513298985,
+      "learning_rate": 9.998434159506211e-06,
+      "loss": 0.1934,
+      "step": 435
+    },
+    {
+      "epoch": 0.11577270313329793,
+      "grad_norm": 0.19317549952322358,
+      "learning_rate": 9.998378738784536e-06,
+      "loss": 0.0092,
+      "step": 436
+    },
+    {
+      "epoch": 0.11603823685608072,
+      "grad_norm": 0.7688675078104924,
+      "learning_rate": 9.998322354434388e-06,
+      "loss": 0.0432,
+      "step": 437
+    },
+    {
+      "epoch": 0.11630377057886351,
+      "grad_norm": 0.7369207362136836,
+      "learning_rate": 9.998265006466642e-06,
+      "loss": 0.0432,
+      "step": 438
+    },
+    {
+      "epoch": 0.1165693043016463,
+      "grad_norm": 0.29436457021835755,
+      "learning_rate": 9.998206694892352e-06,
+      "loss": 0.0146,
+      "step": 439
+    },
+    {
+      "epoch": 0.11683483802442911,
+      "grad_norm": 0.9124118545569474,
+      "learning_rate": 9.998147419722759e-06,
+      "loss": 0.0525,
+      "step": 440
+    },
+    {
+      "epoch": 0.1171003717472119,
+      "grad_norm": 1.1218128409435468,
+      "learning_rate": 9.99808718096929e-06,
+      "loss": 0.0635,
+      "step": 441
+    },
+    {
+      "epoch": 0.11736590546999469,
+      "grad_norm": 1.149876905466913,
+      "learning_rate": 9.99802597864356e-06,
+      "loss": 0.0752,
+      "step": 442
+    },
+    {
+      "epoch": 0.11763143919277748,
+      "grad_norm": 1.0569853424545377,
+      "learning_rate": 9.997963812757368e-06,
+      "loss": 0.0635,
+      "step": 443
+    },
+    {
+      "epoch": 0.11789697291556028,
+      "grad_norm": 0.8351903364180481,
+      "learning_rate": 9.997900683322698e-06,
+      "loss": 0.0476,
+      "step": 444
+    },
+    {
+      "epoch": 0.11816250663834307,
+      "grad_norm": 1.7702924922542935,
+      "learning_rate": 9.997836590351721e-06,
+      "loss": 0.2246,
+      "step": 445
+    },
+    {
+      "epoch": 0.11842804036112586,
+      "grad_norm": 0.14841631560098084,
+      "learning_rate": 9.997771533856794e-06,
+      "loss": 0.0072,
+      "step": 446
+    },
+    {
+      "epoch": 0.11869357408390865,
+      "grad_norm": 0.2930240108885635,
+      "learning_rate": 9.997705513850458e-06,
+      "loss": 0.0165,
+      "step": 447
+    },
+    {
+      "epoch": 0.11895910780669144,
+      "grad_norm": 0.027324664896434907,
+      "learning_rate": 9.997638530345442e-06,
+      "loss": 0.001,
+      "step": 448
+    },
+    {
+      "epoch": 0.11922464152947425,
+      "grad_norm": 0.8050072457659047,
+      "learning_rate": 9.997570583354659e-06,
+      "loss": 0.0435,
+      "step": 449
+    },
+    {
+      "epoch": 0.11949017525225704,
+      "grad_norm": 0.8260959431214968,
+      "learning_rate": 9.997501672891208e-06,
+      "loss": 0.0579,
+      "step": 450
+    },
+    {
+      "epoch": 0.11975570897503983,
+      "grad_norm": 0.050172710801976175,
+      "learning_rate": 9.997431798968376e-06,
+      "loss": 0.0021,
+      "step": 451
+    },
+    {
+      "epoch": 0.12002124269782262,
+      "grad_norm": 0.3333884739733433,
+      "learning_rate": 9.99736096159963e-06,
+      "loss": 0.0147,
+      "step": 452
+    },
+    {
+      "epoch": 0.12028677642060542,
+      "grad_norm": 0.6853848215343389,
+      "learning_rate": 9.997289160798629e-06,
+      "loss": 0.0231,
+      "step": 453
+    },
+    {
+      "epoch": 0.12055231014338821,
+      "grad_norm": 6.935657752668328,
+      "learning_rate": 9.997216396579217e-06,
+      "loss": 1.2891,
+      "step": 454
+    },
+    {
+      "epoch": 0.120817843866171,
+      "grad_norm": 1.2388612554211353,
+      "learning_rate": 9.99714266895542e-06,
+      "loss": 0.0579,
+      "step": 455
+    },
+    {
+      "epoch": 0.1210833775889538,
+      "grad_norm": 0.8108343213894442,
+      "learning_rate": 9.997067977941453e-06,
+      "loss": 0.0354,
+      "step": 456
+    },
+    {
+      "epoch": 0.12134891131173658,
+      "grad_norm": 0.12285291127516768,
+      "learning_rate": 9.996992323551713e-06,
+      "loss": 0.0057,
+      "step": 457
+    },
+    {
+      "epoch": 0.12161444503451939,
+      "grad_norm": 1.8644514519390076,
+      "learning_rate": 9.996915705800786e-06,
+      "loss": 0.2031,
+      "step": 458
+    },
+    {
+      "epoch": 0.12187997875730218,
+      "grad_norm": 0.4192870410785925,
+      "learning_rate": 9.996838124703448e-06,
+      "loss": 0.0118,
+      "step": 459
+    },
+    {
+      "epoch": 0.12214551248008497,
+      "grad_norm": 13.085772336960922,
+      "learning_rate": 9.996759580274649e-06,
+      "loss": 1.1562,
+      "step": 460
+    },
+    {
+      "epoch": 0.12241104620286776,
+      "grad_norm": 1.9383018555391174,
+      "learning_rate": 9.996680072529536e-06,
+      "loss": 0.1455,
+      "step": 461
+    },
+    {
+      "epoch": 0.12267657992565056,
+      "grad_norm": 0.5114158494450113,
+      "learning_rate": 9.996599601483435e-06,
+      "loss": 0.0166,
+      "step": 462
+    },
+    {
+      "epoch": 0.12294211364843335,
+      "grad_norm": 2.071144791223387,
+      "learning_rate": 9.996518167151858e-06,
+      "loss": 0.3105,
+      "step": 463
+    },
+    {
+      "epoch": 0.12320764737121614,
+      "grad_norm": 0.3826668161577709,
+      "learning_rate": 9.996435769550509e-06,
+      "loss": 0.0184,
+      "step": 464
+    },
+    {
+      "epoch": 0.12347318109399893,
+      "grad_norm": 0.11580831376705496,
+      "learning_rate": 9.99635240869527e-06,
+      "loss": 0.0057,
+      "step": 465
+    },
+    {
+      "epoch": 0.12373871481678173,
+      "grad_norm": 2.0940806054901633,
+      "learning_rate": 9.996268084602213e-06,
+      "loss": 0.1836,
+      "step": 466
+    },
+    {
+      "epoch": 0.12400424853956453,
+      "grad_norm": 1.7889639735230736,
+      "learning_rate": 9.996182797287597e-06,
+      "loss": 0.1934,
+      "step": 467
+    },
+    {
+      "epoch": 0.12426978226234732,
+      "grad_norm": 0.274643361135419,
+      "learning_rate": 9.99609654676786e-06,
+      "loss": 0.0095,
+      "step": 468
+    },
+    {
+      "epoch": 0.12453531598513011,
+      "grad_norm": 0.37891937937045345,
+      "learning_rate": 9.996009333059633e-06,
+      "loss": 0.0206,
+      "step": 469
+    },
+    {
+      "epoch": 0.1248008497079129,
+      "grad_norm": 0.5048830191970928,
+      "learning_rate": 9.995921156179729e-06,
+      "loss": 0.0229,
+      "step": 470
+    },
+    {
+      "epoch": 0.1250663834306957,
+      "grad_norm": 2.793136832206918,
+      "learning_rate": 9.995832016145149e-06,
+      "loss": 0.2988,
+      "step": 471
+    },
+    {
+      "epoch": 0.1253319171534785,
+      "grad_norm": 0.9181078850272798,
+      "learning_rate": 9.995741912973073e-06,
+      "loss": 0.0479,
+      "step": 472
+    },
+    {
+      "epoch": 0.12559745087626129,
+      "grad_norm": 0.17214860190032075,
+      "learning_rate": 9.995650846680878e-06,
+      "loss": 0.0025,
+      "step": 473
+    },
+    {
+      "epoch": 0.12586298459904408,
+      "grad_norm": 0.1930329238705798,
+      "learning_rate": 9.995558817286117e-06,
+      "loss": 0.0092,
+      "step": 474
+    },
+    {
+      "epoch": 0.12612851832182687,
+      "grad_norm": 0.3890866531462164,
+      "learning_rate": 9.995465824806534e-06,
+      "loss": 0.0206,
+      "step": 475
+    },
+    {
+      "epoch": 0.12639405204460966,
+      "grad_norm": 0.294426458994664,
+      "learning_rate": 9.995371869260057e-06,
+      "loss": 0.0146,
+      "step": 476
+    },
+    {
+      "epoch": 0.12665958576739245,
+      "grad_norm": 0.49377766768440445,
+      "learning_rate": 9.995276950664796e-06,
+      "loss": 0.0234,
+      "step": 477
+    },
+    {
+      "epoch": 0.12692511949017526,
+      "grad_norm": 1.6403902226241507,
+      "learning_rate": 9.995181069039055e-06,
+      "loss": 0.1035,
+      "step": 478
+    },
+    {
+      "epoch": 0.12719065321295805,
+      "grad_norm": 2.3189750279643517,
+      "learning_rate": 9.995084224401316e-06,
+      "loss": 0.1934,
+      "step": 479
+    },
+    {
+      "epoch": 0.12745618693574085,
+      "grad_norm": 4.2411804627535385,
+      "learning_rate": 9.99498641677025e-06,
+      "loss": 0.7969,
+      "step": 480
+    },
+    {
+      "epoch": 0.12772172065852364,
+      "grad_norm": 0.6775891633342407,
+      "learning_rate": 9.994887646164712e-06,
+      "loss": 0.0354,
+      "step": 481
+    },
+    {
+      "epoch": 0.12798725438130643,
+      "grad_norm": 0.3451145426368071,
+      "learning_rate": 9.994787912603747e-06,
+      "loss": 0.0167,
+      "step": 482
+    },
+    {
+      "epoch": 0.12825278810408922,
+      "grad_norm": 0.7928434044000325,
+      "learning_rate": 9.99468721610658e-06,
+      "loss": 0.0396,
+      "step": 483
+    },
+    {
+      "epoch": 0.128518321826872,
+      "grad_norm": 0.35058719889346857,
+      "learning_rate": 9.994585556692624e-06,
+      "loss": 0.0188,
+      "step": 484
+    },
+    {
+      "epoch": 0.1287838555496548,
+      "grad_norm": 0.10110600298106223,
+      "learning_rate": 9.994482934381478e-06,
+      "loss": 0.0037,
+      "step": 485
+    },
+    {
+      "epoch": 0.1290493892724376,
+      "grad_norm": 0.8111935229536589,
+      "learning_rate": 9.994379349192927e-06,
+      "loss": 0.0393,
+      "step": 486
+    },
+    {
+      "epoch": 0.1293149229952204,
+      "grad_norm": 0.36191339865591954,
+      "learning_rate": 9.994274801146942e-06,
+      "loss": 0.0165,
+      "step": 487
+    },
+    {
+      "epoch": 0.1295804567180032,
+      "grad_norm": 1.4295506331809191,
+      "learning_rate": 9.994169290263675e-06,
+      "loss": 0.0688,
+      "step": 488
+    },
+    {
+      "epoch": 0.12984599044078599,
+      "grad_norm": 1.4477435816269884,
+      "learning_rate": 9.99406281656347e-06,
+      "loss": 0.1191,
+      "step": 489
+    },
+    {
+      "epoch": 0.13011152416356878,
+      "grad_norm": 0.43333547219941215,
+      "learning_rate": 9.993955380066856e-06,
+      "loss": 0.0168,
+      "step": 490
+    },
+    {
+      "epoch": 0.13037705788635157,
+      "grad_norm": 1.135520056624293,
+      "learning_rate": 9.993846980794542e-06,
+      "loss": 0.063,
+      "step": 491
+    },
+    {
+      "epoch": 0.13064259160913436,
+      "grad_norm": 0.043875501106062434,
+      "learning_rate": 9.993737618767426e-06,
+      "loss": 0.0015,
+      "step": 492
+    },
+    {
+      "epoch": 0.13090812533191715,
+      "grad_norm": 0.6090554091571819,
+      "learning_rate": 9.993627294006592e-06,
+      "loss": 0.0286,
+      "step": 493
+    },
+    {
+      "epoch": 0.13117365905469994,
+      "grad_norm": 1.2750886594774946,
+      "learning_rate": 9.993516006533311e-06,
+      "loss": 0.0576,
+      "step": 494
+    },
+    {
+      "epoch": 0.13143919277748273,
+      "grad_norm": 0.09866268692375627,
+      "learning_rate": 9.993403756369037e-06,
+      "loss": 0.0036,
+      "step": 495
+    },
+    {
+      "epoch": 0.13170472650026555,
+      "grad_norm": 2.5466564151997866,
+      "learning_rate": 9.993290543535412e-06,
+      "loss": 0.2148,
+      "step": 496
+    },
+    {
+      "epoch": 0.13197026022304834,
+      "grad_norm": 1.0356354392876912,
+      "learning_rate": 9.993176368054258e-06,
+      "loss": 0.0476,
+      "step": 497
+    },
+    {
+      "epoch": 0.13223579394583113,
+      "grad_norm": 0.1524377921395385,
+      "learning_rate": 9.993061229947591e-06,
+      "loss": 0.0072,
+      "step": 498
+    },
+    {
+      "epoch": 0.13250132766861392,
+      "grad_norm": 5.508400351087429,
+      "learning_rate": 9.992945129237607e-06,
+      "loss": 1.0156,
+      "step": 499
+    },
+    {
+      "epoch": 0.1327668613913967,
+      "grad_norm": 1.8391859755052025,
+      "learning_rate": 9.992828065946687e-06,
+      "loss": 0.2344,
+      "step": 500
+    },
+    {
+      "epoch": 0.1330323951141795,
+      "grad_norm": 0.2511312439828377,
+      "learning_rate": 9.992710040097402e-06,
+      "loss": 0.01,
+      "step": 501
+    },
+    {
+      "epoch": 0.1332979288369623,
+      "grad_norm": 1.6141514471649456,
+      "learning_rate": 9.992591051712503e-06,
+      "loss": 0.1738,
+      "step": 502
+    },
+    {
+      "epoch": 0.13356346255974508,
+      "grad_norm": 4.287684776700203,
+      "learning_rate": 9.992471100814934e-06,
+      "loss": 0.6641,
+      "step": 503
+    },
+    {
+      "epoch": 0.13382899628252787,
+      "grad_norm": 0.7498793642360017,
+      "learning_rate": 9.992350187427814e-06,
+      "loss": 0.0476,
+      "step": 504
+    },
+    {
+      "epoch": 0.13409453000531069,
+      "grad_norm": 0.7170932959418976,
+      "learning_rate": 9.992228311574462e-06,
+      "loss": 0.0393,
+      "step": 505
+    },
+    {
+      "epoch": 0.13436006372809348,
+      "grad_norm": 1.0630972502146814,
+      "learning_rate": 9.992105473278366e-06,
+      "loss": 0.0688,
+      "step": 506
+    },
+    {
+      "epoch": 0.13462559745087627,
+      "grad_norm": 1.605822286298563,
+      "learning_rate": 9.991981672563211e-06,
+      "loss": 0.0815,
+      "step": 507
+    },
+    {
+      "epoch": 0.13489113117365906,
+      "grad_norm": 0.606270866795599,
+      "learning_rate": 9.991856909452866e-06,
+      "loss": 0.032,
+      "step": 508
+    },
+    {
+      "epoch": 0.13515666489644185,
+      "grad_norm": 1.3621617997303943,
+      "learning_rate": 9.99173118397138e-06,
+      "loss": 0.1035,
+      "step": 509
+    },
+    {
+      "epoch": 0.13542219861922464,
+      "grad_norm": 0.12342567577614295,
+      "learning_rate": 9.991604496142997e-06,
+      "loss": 0.0064,
+      "step": 510
+    },
+    {
+      "epoch": 0.13568773234200743,
+      "grad_norm": 1.044690642843683,
+      "learning_rate": 9.991476845992137e-06,
+      "loss": 0.053,
+      "step": 511
+    },
+    {
+      "epoch": 0.13595326606479022,
+      "grad_norm": 0.35855152173267857,
+      "learning_rate": 9.991348233543408e-06,
+      "loss": 0.0206,
+      "step": 512
+    },
+    {
+      "epoch": 0.136218799787573,
+      "grad_norm": 1.57497690926948,
+      "learning_rate": 9.991218658821609e-06,
+      "loss": 0.2246,
+      "step": 513
+    },
+    {
+      "epoch": 0.13648433351035583,
+      "grad_norm": 0.1113998089521789,
+      "learning_rate": 9.991088121851716e-06,
+      "loss": 0.0057,
+      "step": 514
+    },
+    {
+      "epoch": 0.13674986723313862,
+      "grad_norm": 0.22281969208169422,
+      "learning_rate": 9.990956622658901e-06,
+      "loss": 0.0117,
+      "step": 515
+    },
+    {
+      "epoch": 0.1370154009559214,
+      "grad_norm": 0.2932759441130008,
+      "learning_rate": 9.990824161268509e-06,
+      "loss": 0.0148,
+      "step": 516
+    },
+    {
+      "epoch": 0.1372809346787042,
+      "grad_norm": 0.10622372878984855,
+      "learning_rate": 9.990690737706081e-06,
+      "loss": 0.0057,
+      "step": 517
+    },
+    {
+      "epoch": 0.137546468401487,
+      "grad_norm": 1.5597393946203308,
+      "learning_rate": 9.990556351997338e-06,
+      "loss": 0.1113,
+      "step": 518
+    },
+    {
+      "epoch": 0.13781200212426978,
+      "grad_norm": 1.9903611428670367,
+      "learning_rate": 9.99042100416819e-06,
+      "loss": 0.0703,
+      "step": 519
+    },
+    {
+      "epoch": 0.13807753584705257,
+      "grad_norm": 0.35832314011240834,
+      "learning_rate": 9.990284694244729e-06,
+      "loss": 0.0206,
+      "step": 520
+    },
+    {
+      "epoch": 0.13834306956983536,
+      "grad_norm": 5.2897686383169455,
+      "learning_rate": 9.990147422253234e-06,
+      "loss": 0.8125,
+      "step": 521
+    },
+    {
+      "epoch": 0.13860860329261815,
+      "grad_norm": 0.11713567466813415,
+      "learning_rate": 9.990009188220166e-06,
+      "loss": 0.0039,
+      "step": 522
+    },
+    {
+      "epoch": 0.13887413701540097,
+      "grad_norm": 0.5823075698755563,
+      "learning_rate": 9.989869992172183e-06,
+      "loss": 0.0242,
+      "step": 523
+    },
+    {
+      "epoch": 0.13913967073818376,
+      "grad_norm": 0.5765768333479371,
+      "learning_rate": 9.989729834136113e-06,
+      "loss": 0.0193,
+      "step": 524
+    },
+    {
+      "epoch": 0.13940520446096655,
+      "grad_norm": 0.14344215523221354,
+      "learning_rate": 9.989588714138977e-06,
+      "loss": 0.0039,
+      "step": 525
+    },
+    {
+      "epoch": 0.13967073818374934,
+      "grad_norm": 1.7019424437997808,
+      "learning_rate": 9.989446632207989e-06,
+      "loss": 0.0752,
+      "step": 526
+    },
+    {
+      "epoch": 0.13993627190653213,
+      "grad_norm": 0.2075958214948594,
+      "learning_rate": 9.989303588370531e-06,
+      "loss": 0.0117,
+      "step": 527
+    },
+    {
+      "epoch": 0.14020180562931492,
+      "grad_norm": 0.15118506962449466,
+      "learning_rate": 9.989159582654187e-06,
+      "loss": 0.0082,
+      "step": 528
+    },
+    {
+      "epoch": 0.1404673393520977,
+      "grad_norm": 0.07268392646364422,
+      "learning_rate": 9.989014615086716e-06,
+      "loss": 0.0035,
+      "step": 529
+    },
+    {
+      "epoch": 0.1407328730748805,
+      "grad_norm": 1.4906910982100858,
+      "learning_rate": 9.988868685696067e-06,
+      "loss": 0.2129,
+      "step": 530
+    },
+    {
+      "epoch": 0.14099840679766332,
+      "grad_norm": 1.4326095317287415,
+      "learning_rate": 9.988721794510374e-06,
+      "loss": 0.1836,
+      "step": 531
+    },
+    {
+      "epoch": 0.1412639405204461,
+      "grad_norm": 0.5130258898072058,
+      "learning_rate": 9.988573941557956e-06,
+      "loss": 0.0177,
+      "step": 532
+    },
+    {
+      "epoch": 0.1415294742432289,
+      "grad_norm": 1.3951302100850165,
+      "learning_rate": 9.988425126867317e-06,
+      "loss": 0.1641,
+      "step": 533
+    },
+    {
+      "epoch": 0.1417950079660117,
+      "grad_norm": 1.5620173297534108,
+      "learning_rate": 9.988275350467145e-06,
+      "loss": 0.082,
+      "step": 534
+    },
+    {
+      "epoch": 0.14206054168879448,
+      "grad_norm": 1.5140269849638728,
+      "learning_rate": 9.988124612386319e-06,
+      "loss": 0.1367,
+      "step": 535
+    },
+    {
+      "epoch": 0.14232607541157727,
+      "grad_norm": 0.6063242319852478,
+      "learning_rate": 9.987972912653894e-06,
+      "loss": 0.0393,
+      "step": 536
+    },
+    {
+      "epoch": 0.14259160913436006,
+      "grad_norm": 0.7374682676713821,
+      "learning_rate": 9.987820251299121e-06,
+      "loss": 0.0435,
+      "step": 537
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 1.39588526651874,
+      "learning_rate": 9.987666628351432e-06,
+      "loss": 0.1553,
+      "step": 538
+    },
+    {
+      "epoch": 0.14312267657992564,
+      "grad_norm": 2.2114571113292394,
+      "learning_rate": 9.987512043840437e-06,
+      "loss": 0.1113,
+      "step": 539
+    },
+    {
+      "epoch": 0.14338821030270846,
+      "grad_norm": 0.42061182715165285,
+      "learning_rate": 9.987356497795944e-06,
+      "loss": 0.0286,
+      "step": 540
+    },
+    {
+      "epoch": 0.14365374402549125,
+      "grad_norm": 0.9900996921832054,
+      "learning_rate": 9.987199990247938e-06,
+      "loss": 0.0752,
+      "step": 541
+    },
+    {
+      "epoch": 0.14391927774827404,
+      "grad_norm": 0.7148831432764048,
+      "learning_rate": 9.987042521226594e-06,
+      "loss": 0.0476,
+      "step": 542
+    },
+    {
+      "epoch": 0.14418481147105683,
+      "grad_norm": 1.5856605293259478,
+      "learning_rate": 9.986884090762266e-06,
+      "loss": 0.1123,
+      "step": 543
+    },
+    {
+      "epoch": 0.14445034519383962,
+      "grad_norm": 0.17251194671640718,
+      "learning_rate": 9.986724698885502e-06,
+      "loss": 0.0058,
+      "step": 544
+    },
+    {
+      "epoch": 0.1447158789166224,
+      "grad_norm": 0.7025948256764871,
+      "learning_rate": 9.986564345627027e-06,
+      "loss": 0.0322,
+      "step": 545
+    },
+    {
+      "epoch": 0.1449814126394052,
+      "grad_norm": 0.039152169713499795,
+      "learning_rate": 9.98640303101776e-06,
+      "loss": 0.0011,
+      "step": 546
+    },
+    {
+      "epoch": 0.145246946362188,
+      "grad_norm": 0.7346842884881487,
+      "learning_rate": 9.986240755088797e-06,
+      "loss": 0.033,
+      "step": 547
+    },
+    {
+      "epoch": 0.14551248008497078,
+      "grad_norm": 0.8712665193969032,
+      "learning_rate": 9.986077517871424e-06,
+      "loss": 0.0476,
+      "step": 548
+    },
+    {
+      "epoch": 0.1457780138077536,
+      "grad_norm": 0.4876598992779223,
+      "learning_rate": 9.98591331939711e-06,
+      "loss": 0.0234,
+      "step": 549
+    },
+    {
+      "epoch": 0.1460435475305364,
+      "grad_norm": 0.5723125307338557,
+      "learning_rate": 9.985748159697511e-06,
+      "loss": 0.0432,
+      "step": 550
+    },
+    {
+      "epoch": 0.14630908125331918,
+      "grad_norm": 5.026172323034893,
+      "learning_rate": 9.985582038804469e-06,
+      "loss": 0.7734,
+      "step": 551
+    },
+    {
+      "epoch": 0.14657461497610197,
+      "grad_norm": 0.691831344385777,
+      "learning_rate": 9.985414956750008e-06,
+      "loss": 0.0289,
+      "step": 552
+    },
+    {
+      "epoch": 0.14684014869888476,
+      "grad_norm": 0.5313434024030962,
+      "learning_rate": 9.985246913566343e-06,
+      "loss": 0.0121,
+      "step": 553
+    },
+    {
+      "epoch": 0.14710568242166755,
+      "grad_norm": 0.5022681233131463,
+      "learning_rate": 9.985077909285867e-06,
+      "loss": 0.0391,
+      "step": 554
+    },
+    {
+      "epoch": 0.14737121614445034,
+      "grad_norm": 2.8350298800583738,
+      "learning_rate": 9.984907943941164e-06,
+      "loss": 0.2988,
+      "step": 555
+    },
+    {
+      "epoch": 0.14763674986723313,
+      "grad_norm": 0.5845567847490889,
+      "learning_rate": 9.984737017565003e-06,
+      "loss": 0.0435,
+      "step": 556
+    },
+    {
+      "epoch": 0.14790228359001592,
+      "grad_norm": 5.065900056975599,
+      "learning_rate": 9.984565130190331e-06,
+      "loss": 0.3242,
+      "step": 557
+    },
+    {
+      "epoch": 0.14816781731279874,
+      "grad_norm": 0.9255981265488107,
+      "learning_rate": 9.984392281850293e-06,
+      "loss": 0.0889,
+      "step": 558
+    },
+    {
+      "epoch": 0.14843335103558153,
+      "grad_norm": 0.7464338462336892,
+      "learning_rate": 9.984218472578205e-06,
+      "loss": 0.0576,
+      "step": 559
+    },
+    {
+      "epoch": 0.14869888475836432,
+      "grad_norm": 0.9740861593787775,
+      "learning_rate": 9.984043702407581e-06,
+      "loss": 0.0957,
+      "step": 560
+    },
+    {
+      "epoch": 0.1489644184811471,
+      "grad_norm": 0.4188005366867957,
+      "learning_rate": 9.983867971372113e-06,
+      "loss": 0.0286,
+      "step": 561
+    },
+    {
+      "epoch": 0.1492299522039299,
+      "grad_norm": 3.331842808759764,
+      "learning_rate": 9.983691279505678e-06,
+      "loss": 0.3398,
+      "step": 562
+    },
+    {
+      "epoch": 0.1494954859267127,
+      "grad_norm": 4.35094959797938,
+      "learning_rate": 9.983513626842342e-06,
+      "loss": 0.6523,
+      "step": 563
+    },
+    {
+      "epoch": 0.14976101964949548,
+      "grad_norm": 0.5123642715371224,
+      "learning_rate": 9.983335013416354e-06,
+      "loss": 0.0391,
+      "step": 564
+    },
+    {
+      "epoch": 0.15002655337227827,
+      "grad_norm": 0.796553563912584,
+      "learning_rate": 9.983155439262148e-06,
+      "loss": 0.0579,
+      "step": 565
+    },
+    {
+      "epoch": 0.15029208709506106,
+      "grad_norm": 0.3439490402717059,
+      "learning_rate": 9.982974904414344e-06,
+      "loss": 0.0229,
+      "step": 566
+    },
+    {
+      "epoch": 0.15055762081784388,
+      "grad_norm": 0.4106869676123719,
+      "learning_rate": 9.982793408907747e-06,
+      "loss": 0.0186,
+      "step": 567
+    },
+    {
+      "epoch": 0.15082315454062667,
+      "grad_norm": 0.7143183560111053,
+      "learning_rate": 9.982610952777347e-06,
+      "loss": 0.0393,
+      "step": 568
+    },
+    {
+      "epoch": 0.15108868826340946,
+      "grad_norm": 0.6128505770192598,
+      "learning_rate": 9.982427536058321e-06,
+      "loss": 0.0476,
+      "step": 569
+    },
+    {
+      "epoch": 0.15135422198619225,
+      "grad_norm": 0.2585949344039441,
+      "learning_rate": 9.98224315878603e-06,
+      "loss": 0.0165,
+      "step": 570
+    },
+    {
+      "epoch": 0.15161975570897504,
+      "grad_norm": 0.11924071710544426,
+      "learning_rate": 9.982057820996015e-06,
+      "loss": 0.0039,
+      "step": 571
+    },
+    {
+      "epoch": 0.15188528943175783,
+      "grad_norm": 0.6551332258487033,
+      "learning_rate": 9.98187152272401e-06,
+      "loss": 0.0391,
+      "step": 572
+    },
+    {
+      "epoch": 0.15215082315454062,
+      "grad_norm": 2.691563544451101,
+      "learning_rate": 9.981684264005934e-06,
+      "loss": 0.2031,
+      "step": 573
+    },
+    {
+      "epoch": 0.1524163568773234,
+      "grad_norm": 1.360641707061635,
+      "learning_rate": 9.981496044877882e-06,
+      "loss": 0.1934,
+      "step": 574
+    },
+    {
+      "epoch": 0.1526818906001062,
+      "grad_norm": 0.43346474177116345,
+      "learning_rate": 9.981306865376146e-06,
+      "loss": 0.0139,
+      "step": 575
+    },
+    {
+      "epoch": 0.15294742432288902,
+      "grad_norm": 0.24328200960959998,
+      "learning_rate": 9.981116725537195e-06,
+      "loss": 0.0165,
+      "step": 576
+    },
+    {
+      "epoch": 0.1532129580456718,
+      "grad_norm": 1.0426671122461846,
+      "learning_rate": 9.980925625397686e-06,
+      "loss": 0.0576,
+      "step": 577
+    },
+    {
+      "epoch": 0.1534784917684546,
+      "grad_norm": 0.1999935342186825,
+      "learning_rate": 9.98073356499446e-06,
+      "loss": 0.0131,
+      "step": 578
+    },
+    {
+      "epoch": 0.1537440254912374,
+      "grad_norm": 0.7190524116140479,
+      "learning_rate": 9.980540544364546e-06,
+      "loss": 0.0525,
+      "step": 579
+    },
+    {
+      "epoch": 0.15400955921402018,
+      "grad_norm": 2.4567461753510256,
+      "learning_rate": 9.980346563545156e-06,
+      "loss": 0.1035,
+      "step": 580
+    },
+    {
+      "epoch": 0.15427509293680297,
+      "grad_norm": 1.3907008477488672,
+      "learning_rate": 9.980151622573684e-06,
+      "loss": 0.1279,
+      "step": 581
+    },
+    {
+      "epoch": 0.15454062665958576,
+      "grad_norm": 1.6662654026394559,
+      "learning_rate": 9.979955721487716e-06,
+      "loss": 0.0752,
+      "step": 582
+    },
+    {
+      "epoch": 0.15480616038236855,
+      "grad_norm": 0.2905768001459995,
+      "learning_rate": 9.97975886032502e-06,
+      "loss": 0.0104,
+      "step": 583
+    },
+    {
+      "epoch": 0.15507169410515134,
+      "grad_norm": 0.5722935674089386,
+      "learning_rate": 9.979561039123544e-06,
+      "loss": 0.0391,
+      "step": 584
+    },
+    {
+      "epoch": 0.15533722782793416,
+      "grad_norm": 3.9956006342859354,
+      "learning_rate": 9.979362257921428e-06,
+      "loss": 0.1699,
+      "step": 585
+    },
+    {
+      "epoch": 0.15560276155071695,
+      "grad_norm": 1.4427656836880338,
+      "learning_rate": 9.979162516756995e-06,
+      "loss": 0.0889,
+      "step": 586
+    },
+    {
+      "epoch": 0.15586829527349974,
+      "grad_norm": 1.105527580649903,
+      "learning_rate": 9.978961815668753e-06,
+      "loss": 0.0752,
+      "step": 587
+    },
+    {
+      "epoch": 0.15613382899628253,
+      "grad_norm": 0.1918918255795427,
+      "learning_rate": 9.978760154695393e-06,
+      "loss": 0.0059,
+      "step": 588
+    },
+    {
+      "epoch": 0.15639936271906532,
+      "grad_norm": 0.04571838620835277,
+      "learning_rate": 9.978557533875797e-06,
+      "loss": 0.0011,
+      "step": 589
+    },
+    {
+      "epoch": 0.1566648964418481,
+      "grad_norm": 1.9273246134899191,
+      "learning_rate": 9.978353953249023e-06,
+      "loss": 0.1641,
+      "step": 590
+    },
+    {
+      "epoch": 0.1569304301646309,
+      "grad_norm": 1.3054086665266786,
+      "learning_rate": 9.97814941285432e-06,
+      "loss": 0.0693,
+      "step": 591
+    },
+    {
+      "epoch": 0.1571959638874137,
+      "grad_norm": 0.030533978901150845,
+      "learning_rate": 9.977943912731126e-06,
+      "loss": 0.0007,
+      "step": 592
+    },
+    {
+      "epoch": 0.15746149761019648,
+      "grad_norm": 0.9348392424847526,
+      "learning_rate": 9.977737452919051e-06,
+      "loss": 0.0525,
+      "step": 593
+    },
+    {
+      "epoch": 0.1577270313329793,
+      "grad_norm": 0.6672615655023215,
+      "learning_rate": 9.977530033457906e-06,
+      "loss": 0.0317,
+      "step": 594
+    },
+    {
+      "epoch": 0.1579925650557621,
+      "grad_norm": 0.9167557805831286,
+      "learning_rate": 9.977321654387671e-06,
+      "loss": 0.0525,
+      "step": 595
+    },
+    {
+      "epoch": 0.15825809877854488,
+      "grad_norm": 0.5548577985701605,
+      "learning_rate": 9.977112315748525e-06,
+      "loss": 0.0391,
+      "step": 596
+    },
+    {
+      "epoch": 0.15852363250132767,
+      "grad_norm": 1.0741347246530948,
+      "learning_rate": 9.976902017580824e-06,
+      "loss": 0.0815,
+      "step": 597
+    },
+    {
+      "epoch": 0.15878916622411046,
+      "grad_norm": 0.1973076257512853,
+      "learning_rate": 9.976690759925112e-06,
+      "loss": 0.0058,
+      "step": 598
+    },
+    {
+      "epoch": 0.15905469994689325,
+      "grad_norm": 0.37899730498629314,
+      "learning_rate": 9.976478542822114e-06,
+      "loss": 0.0187,
+      "step": 599
+    },
+    {
+      "epoch": 0.15932023366967604,
+      "grad_norm": 0.030062306649064036,
+      "learning_rate": 9.976265366312746e-06,
+      "loss": 0.0008,
+      "step": 600
+    },
+    {
+      "epoch": 0.15958576739245883,
+      "grad_norm": 0.8706157736283716,
+      "learning_rate": 9.976051230438107e-06,
+      "loss": 0.063,
+      "step": 601
+    },
+    {
+      "epoch": 0.15985130111524162,
+      "grad_norm": 0.06252640600648844,
+      "learning_rate": 9.975836135239475e-06,
+      "loss": 0.0019,
+      "step": 602
+    },
+    {
+      "epoch": 0.16011683483802444,
+      "grad_norm": 1.020235806615748,
+      "learning_rate": 9.975620080758321e-06,
+      "loss": 0.0576,
+      "step": 603
+    },
+    {
+      "epoch": 0.16038236856080723,
+      "grad_norm": 0.6968794165000528,
+      "learning_rate": 9.975403067036298e-06,
+      "loss": 0.0432,
+      "step": 604
+    },
+    {
+      "epoch": 0.16064790228359002,
+      "grad_norm": 2.0675935959691527,
+      "learning_rate": 9.975185094115243e-06,
+      "loss": 0.1455,
+      "step": 605
+    },
+    {
+      "epoch": 0.1609134360063728,
+      "grad_norm": 0.07887764560075496,
+      "learning_rate": 9.97496616203718e-06,
+      "loss": 0.0014,
+      "step": 606
+    },
+    {
+      "epoch": 0.1611789697291556,
+      "grad_norm": 1.0038423253490598,
+      "learning_rate": 9.974746270844313e-06,
+      "loss": 0.0476,
+      "step": 607
+    },
+    {
+      "epoch": 0.1614445034519384,
+      "grad_norm": 0.015512665639925233,
+      "learning_rate": 9.974525420579039e-06,
+      "loss": 0.0003,
+      "step": 608
+    },
+    {
+      "epoch": 0.16171003717472118,
+      "grad_norm": 0.1729310979679677,
+      "learning_rate": 9.97430361128393e-06,
+      "loss": 0.0065,
+      "step": 609
+    },
+    {
+      "epoch": 0.16197557089750397,
+      "grad_norm": 0.9007630781441461,
+      "learning_rate": 9.974080843001755e-06,
+      "loss": 0.0576,
+      "step": 610
+    },
+    {
+      "epoch": 0.16224110462028676,
+      "grad_norm": 0.8783601833486903,
+      "learning_rate": 9.973857115775455e-06,
+      "loss": 0.0342,
+      "step": 611
+    },
+    {
+      "epoch": 0.16250663834306958,
+      "grad_norm": 1.7341324095089548,
+      "learning_rate": 9.973632429648165e-06,
+      "loss": 0.0752,
+      "step": 612
+    },
+    {
+      "epoch": 0.16277217206585237,
+      "grad_norm": 1.4666956044492996,
+      "learning_rate": 9.9734067846632e-06,
+      "loss": 0.0815,
+      "step": 613
+    },
+    {
+      "epoch": 0.16303770578863516,
+      "grad_norm": 0.8113376617255486,
+      "learning_rate": 9.973180180864064e-06,
+      "loss": 0.0317,
+      "step": 614
+    },
+    {
+      "epoch": 0.16330323951141795,
+      "grad_norm": 2.5241610625807303,
+      "learning_rate": 9.972952618294442e-06,
+      "loss": 0.1738,
+      "step": 615
+    },
+    {
+      "epoch": 0.16356877323420074,
+      "grad_norm": 0.008095688510017955,
+      "learning_rate": 9.972724096998206e-06,
+      "loss": 0.0001,
+      "step": 616
+    },
+    {
+      "epoch": 0.16383430695698353,
+      "grad_norm": 0.9981390757792195,
+      "learning_rate": 9.97249461701941e-06,
+      "loss": 0.0576,
+      "step": 617
+    },
+    {
+      "epoch": 0.16409984067976632,
+      "grad_norm": 1.3058310882090556,
+      "learning_rate": 9.9722641784023e-06,
+      "loss": 0.0688,
+      "step": 618
+    },
+    {
+      "epoch": 0.1643653744025491,
+      "grad_norm": 1.0447117296584805,
+      "learning_rate": 9.972032781191297e-06,
+      "loss": 0.0576,
+      "step": 619
+    },
+    {
+      "epoch": 0.1646309081253319,
+      "grad_norm": 0.0992909067061662,
+      "learning_rate": 9.971800425431014e-06,
+      "loss": 0.0025,
+      "step": 620
+    },
+    {
+      "epoch": 0.16489644184811472,
+      "grad_norm": 2.5678651091451097,
+      "learning_rate": 9.971567111166246e-06,
+      "loss": 0.2461,
+      "step": 621
+    },
+    {
+      "epoch": 0.1651619755708975,
+      "grad_norm": 1.3926560176517229,
+      "learning_rate": 9.971332838441973e-06,
+      "loss": 0.0525,
+      "step": 622
+    },
+    {
+      "epoch": 0.1654275092936803,
+      "grad_norm": 0.1312253383616017,
+      "learning_rate": 9.97109760730336e-06,
+      "loss": 0.004,
+      "step": 623
+    },
+    {
+      "epoch": 0.1656930430164631,
+      "grad_norm": 2.1426039577089244,
+      "learning_rate": 9.970861417795757e-06,
+      "loss": 0.1836,
+      "step": 624
+    },
+    {
+      "epoch": 0.16595857673924588,
+      "grad_norm": 1.4296002668699905,
+      "learning_rate": 9.970624269964697e-06,
+      "loss": 0.0889,
+      "step": 625
+    },
+    {
+      "epoch": 0.16622411046202867,
+      "grad_norm": 0.40810968379724855,
+      "learning_rate": 9.970386163855903e-06,
+      "loss": 0.0229,
+      "step": 626
+    },
+    {
+      "epoch": 0.16648964418481146,
+      "grad_norm": 0.026519673065559527,
+      "learning_rate": 9.970147099515277e-06,
+      "loss": 0.0006,
+      "step": 627
+    },
+    {
+      "epoch": 0.16675517790759425,
+      "grad_norm": 0.9111033300992362,
+      "learning_rate": 9.969907076988907e-06,
+      "loss": 0.0576,
+      "step": 628
+    },
+    {
+      "epoch": 0.16702071163037704,
+      "grad_norm": 1.84251401314855,
+      "learning_rate": 9.969666096323068e-06,
+      "loss": 0.1641,
+      "step": 629
+    },
+    {
+      "epoch": 0.16728624535315986,
+      "grad_norm": 1.1977915299686255,
+      "learning_rate": 9.969424157564215e-06,
+      "loss": 0.0752,
+      "step": 630
+    },
+    {
+      "epoch": 0.16755177907594265,
+      "grad_norm": 1.1596796848128883,
+      "learning_rate": 9.969181260758995e-06,
+      "loss": 0.0476,
+      "step": 631
+    },
+    {
+      "epoch": 0.16781731279872544,
+      "grad_norm": 0.7311624184389818,
+      "learning_rate": 9.968937405954235e-06,
+      "loss": 0.0576,
+      "step": 632
+    },
+    {
+      "epoch": 0.16808284652150823,
+      "grad_norm": 0.43066270261344336,
+      "learning_rate": 9.968692593196944e-06,
+      "loss": 0.0317,
+      "step": 633
+    },
+    {
+      "epoch": 0.16834838024429102,
+      "grad_norm": 1.0570009180451556,
+      "learning_rate": 9.968446822534324e-06,
+      "loss": 0.1191,
+      "step": 634
+    },
+    {
+      "epoch": 0.16861391396707381,
+      "grad_norm": 1.7875303182362978,
+      "learning_rate": 9.968200094013751e-06,
+      "loss": 0.1279,
+      "step": 635
+    },
+    {
+      "epoch": 0.1688794476898566,
+      "grad_norm": 0.8109838110303486,
+      "learning_rate": 9.967952407682796e-06,
+      "loss": 0.0525,
+      "step": 636
+    },
+    {
+      "epoch": 0.1691449814126394,
+      "grad_norm": 0.995056083014455,
+      "learning_rate": 9.967703763589208e-06,
+      "loss": 0.0889,
+      "step": 637
+    },
+    {
+      "epoch": 0.16941051513542218,
+      "grad_norm": 0.7317688073360701,
+      "learning_rate": 9.967454161780923e-06,
+      "loss": 0.0576,
+      "step": 638
+    },
+    {
+      "epoch": 0.169676048858205,
+      "grad_norm": 0.27467746292088824,
+      "learning_rate": 9.967203602306062e-06,
+      "loss": 0.0206,
+      "step": 639
+    },
+    {
+      "epoch": 0.1699415825809878,
+      "grad_norm": 1.0717423043804712,
+      "learning_rate": 9.966952085212929e-06,
+      "loss": 0.1455,
+      "step": 640
+    },
+    {
+      "epoch": 0.17020711630377058,
+      "grad_norm": 0.21472091966860316,
+      "learning_rate": 9.966699610550013e-06,
+      "loss": 0.0146,
+      "step": 641
+    },
+    {
+      "epoch": 0.17047265002655337,
+      "grad_norm": 0.5320762490185771,
+      "learning_rate": 9.96644617836599e-06,
+      "loss": 0.0216,
+      "step": 642
+    },
+    {
+      "epoch": 0.17073818374933616,
+      "grad_norm": 0.9624120570928735,
+      "learning_rate": 9.966191788709716e-06,
+      "loss": 0.1035,
+      "step": 643
+    },
+    {
+      "epoch": 0.17100371747211895,
+      "grad_norm": 0.6950471061549814,
+      "learning_rate": 9.965936441630237e-06,
+      "loss": 0.0693,
+      "step": 644
+    },
+    {
+      "epoch": 0.17126925119490174,
+      "grad_norm": 1.060844145170164,
+      "learning_rate": 9.965680137176778e-06,
+      "loss": 0.1553,
+      "step": 645
+    },
+    {
+      "epoch": 0.17153478491768454,
+      "grad_norm": 0.6493031275434652,
+      "learning_rate": 9.965422875398756e-06,
+      "loss": 0.0576,
+      "step": 646
+    },
+    {
+      "epoch": 0.17180031864046733,
+      "grad_norm": 0.2802358255478571,
+      "learning_rate": 9.965164656345765e-06,
+      "loss": 0.0206,
+      "step": 647
+    },
+    {
+      "epoch": 0.17206585236325014,
+      "grad_norm": 1.1889779284390927,
+      "learning_rate": 9.964905480067585e-06,
+      "loss": 0.1113,
+      "step": 648
+    },
+    {
+      "epoch": 0.17233138608603293,
+      "grad_norm": 0.48129568513721377,
+      "learning_rate": 9.964645346614186e-06,
+      "loss": 0.0432,
+      "step": 649
+    },
+    {
+      "epoch": 0.17259691980881572,
+      "grad_norm": 0.5955170121717723,
+      "learning_rate": 9.964384256035715e-06,
+      "loss": 0.0579,
+      "step": 650
+    },
+    {
+      "epoch": 0.17286245353159851,
+      "grad_norm": 0.4053980801589404,
+      "learning_rate": 9.964122208382511e-06,
+      "loss": 0.0354,
+      "step": 651
+    },
+    {
+      "epoch": 0.1731279872543813,
+      "grad_norm": 0.43450493299788406,
+      "learning_rate": 9.96385920370509e-06,
+      "loss": 0.0391,
+      "step": 652
+    },
+    {
+      "epoch": 0.1733935209771641,
+      "grad_norm": 0.4490492804675279,
+      "learning_rate": 9.963595242054158e-06,
+      "loss": 0.0354,
+      "step": 653
+    },
+    {
+      "epoch": 0.17365905469994689,
+      "grad_norm": 0.5527458067392451,
+      "learning_rate": 9.963330323480604e-06,
+      "loss": 0.0525,
+      "step": 654
+    },
+    {
+      "epoch": 0.17392458842272968,
+      "grad_norm": 0.5425679761751954,
+      "learning_rate": 9.9630644480355e-06,
+      "loss": 0.0525,
+      "step": 655
+    },
+    {
+      "epoch": 0.1741901221455125,
+      "grad_norm": 0.35110925559228934,
+      "learning_rate": 9.962797615770105e-06,
+      "loss": 0.0286,
+      "step": 656
+    },
+    {
+      "epoch": 0.17445565586829528,
+      "grad_norm": 0.40723902213459534,
+      "learning_rate": 9.96252982673586e-06,
+      "loss": 0.0354,
+      "step": 657
+    },
+    {
+      "epoch": 0.17472118959107807,
+      "grad_norm": 0.29867369752067024,
+      "learning_rate": 9.962261080984394e-06,
+      "loss": 0.0229,
+      "step": 658
+    },
+    {
+      "epoch": 0.17498672331386086,
+      "grad_norm": 0.30130627923135,
+      "learning_rate": 9.961991378567514e-06,
+      "loss": 0.0229,
+      "step": 659
+    },
+    {
+      "epoch": 0.17525225703664365,
+      "grad_norm": 0.4704399733543959,
+      "learning_rate": 9.961720719537217e-06,
+      "loss": 0.0432,
+      "step": 660
+    },
+    {
+      "epoch": 0.17551779075942645,
+      "grad_norm": 0.8375442861352078,
+      "learning_rate": 9.961449103945684e-06,
+      "loss": 0.0889,
+      "step": 661
+    },
+    {
+      "epoch": 0.17578332448220924,
+      "grad_norm": 0.13114805545544933,
+      "learning_rate": 9.96117653184528e-06,
+      "loss": 0.0045,
+      "step": 662
+    },
+    {
+      "epoch": 0.17604885820499203,
+      "grad_norm": 0.31507559879440783,
+      "learning_rate": 9.960903003288551e-06,
+      "loss": 0.0104,
+      "step": 663
+    },
+    {
+      "epoch": 0.17631439192777482,
+      "grad_norm": 0.38090689585552884,
+      "learning_rate": 9.960628518328231e-06,
+      "loss": 0.0231,
+      "step": 664
+    },
+    {
+      "epoch": 0.17657992565055763,
+      "grad_norm": 0.36367373348514975,
+      "learning_rate": 9.960353077017238e-06,
+      "loss": 0.0286,
+      "step": 665
+    },
+    {
+      "epoch": 0.17684545937334042,
+      "grad_norm": 0.14099382022025622,
+      "learning_rate": 9.960076679408675e-06,
+      "loss": 0.0092,
+      "step": 666
+    },
+    {
+      "epoch": 0.17711099309612321,
+      "grad_norm": 0.3729088508977205,
+      "learning_rate": 9.959799325555825e-06,
+      "loss": 0.0317,
+      "step": 667
+    },
+    {
+      "epoch": 0.177376526818906,
+      "grad_norm": 0.20249754798930547,
+      "learning_rate": 9.95952101551216e-06,
+      "loss": 0.0035,
+      "step": 668
+    },
+    {
+      "epoch": 0.1776420605416888,
+      "grad_norm": 0.728096668821041,
+      "learning_rate": 9.959241749331339e-06,
+      "loss": 0.0391,
+      "step": 669
+    },
+    {
+      "epoch": 0.17790759426447159,
+      "grad_norm": 0.38161937312578453,
+      "learning_rate": 9.958961527067194e-06,
+      "loss": 0.0165,
+      "step": 670
+    },
+    {
+      "epoch": 0.17817312798725438,
+      "grad_norm": 1.1222249017413026,
+      "learning_rate": 9.958680348773753e-06,
+      "loss": 0.1455,
+      "step": 671
+    },
+    {
+      "epoch": 0.17843866171003717,
+      "grad_norm": 1.069255220495418,
+      "learning_rate": 9.958398214505225e-06,
+      "loss": 0.1279,
+      "step": 672
+    },
+    {
+      "epoch": 0.17870419543281996,
+      "grad_norm": 0.6232120562558745,
+      "learning_rate": 9.958115124315996e-06,
+      "loss": 0.0432,
+      "step": 673
+    },
+    {
+      "epoch": 0.17896972915560277,
+      "grad_norm": 0.5800413383472506,
+      "learning_rate": 9.95783107826065e-06,
+      "loss": 0.0125,
+      "step": 674
+    },
+    {
+      "epoch": 0.17923526287838557,
+      "grad_norm": 0.1579792821012369,
+      "learning_rate": 9.957546076393944e-06,
+      "loss": 0.0103,
+      "step": 675
+    },
+    {
+      "epoch": 0.17950079660116836,
+      "grad_norm": 0.8480282976053962,
+      "learning_rate": 9.957260118770822e-06,
+      "loss": 0.0815,
+      "step": 676
+    },
+    {
+      "epoch": 0.17976633032395115,
+      "grad_norm": 0.9382679997758676,
+      "learning_rate": 9.956973205446416e-06,
+      "loss": 0.0957,
+      "step": 677
+    },
+    {
+      "epoch": 0.18003186404673394,
+      "grad_norm": 0.6567431263540028,
+      "learning_rate": 9.956685336476037e-06,
+      "loss": 0.0525,
+      "step": 678
+    },
+    {
+      "epoch": 0.18029739776951673,
+      "grad_norm": 0.9345447093648367,
+      "learning_rate": 9.956396511915182e-06,
+      "loss": 0.0688,
+      "step": 679
+    },
+    {
+      "epoch": 0.18056293149229952,
+      "grad_norm": 1.028643546453173,
+      "learning_rate": 9.956106731819538e-06,
+      "loss": 0.0752,
+      "step": 680
+    },
+    {
+      "epoch": 0.1808284652150823,
+      "grad_norm": 1.1200381609657348,
+      "learning_rate": 9.955815996244967e-06,
+      "loss": 0.0889,
+      "step": 681
+    },
+    {
+      "epoch": 0.1810939989378651,
+      "grad_norm": 0.49251673343344965,
+      "learning_rate": 9.955524305247518e-06,
+      "loss": 0.0391,
+      "step": 682
+    },
+    {
+      "epoch": 0.18135953266064792,
+      "grad_norm": 0.5170146835568467,
+      "learning_rate": 9.955231658883432e-06,
+      "loss": 0.0391,
+      "step": 683
+    },
+    {
+      "epoch": 0.1816250663834307,
+      "grad_norm": 0.7825780874484762,
+      "learning_rate": 9.954938057209121e-06,
+      "loss": 0.0525,
+      "step": 684
+    },
+    {
+      "epoch": 0.1818906001062135,
+      "grad_norm": 1.3290725760750859,
+      "learning_rate": 9.954643500281191e-06,
+      "loss": 0.1934,
+      "step": 685
+    },
+    {
+      "epoch": 0.1821561338289963,
+      "grad_norm": 1.2730693589730842,
+      "learning_rate": 9.95434798815643e-06,
+      "loss": 0.03,
+      "step": 686
+    },
+    {
+      "epoch": 0.18242166755177908,
+      "grad_norm": 0.29840206739057834,
+      "learning_rate": 9.954051520891808e-06,
+      "loss": 0.0229,
+      "step": 687
+    },
+    {
+      "epoch": 0.18268720127456187,
+      "grad_norm": 1.283727575547433,
+      "learning_rate": 9.95375409854448e-06,
+      "loss": 0.1836,
+      "step": 688
+    },
+    {
+      "epoch": 0.18295273499734466,
+      "grad_norm": 0.36148423987976935,
+      "learning_rate": 9.953455721171784e-06,
+      "loss": 0.0286,
+      "step": 689
+    },
+    {
+      "epoch": 0.18321826872012745,
+      "grad_norm": 0.26464506058473897,
+      "learning_rate": 9.953156388831246e-06,
+      "loss": 0.0184,
+      "step": 690
+    },
+    {
+      "epoch": 0.18348380244291024,
+      "grad_norm": 0.26815836434680423,
+      "learning_rate": 9.952856101580574e-06,
+      "loss": 0.0206,
+      "step": 691
+    },
+    {
+      "epoch": 0.18374933616569306,
+      "grad_norm": 0.9433969771302444,
+      "learning_rate": 9.952554859477661e-06,
+      "loss": 0.0815,
+      "step": 692
+    },
+    {
+      "epoch": 0.18401486988847585,
+      "grad_norm": 0.8371591636277368,
+      "learning_rate": 9.95225266258058e-06,
+      "loss": 0.0688,
+      "step": 693
+    },
+    {
+      "epoch": 0.18428040361125864,
+      "grad_norm": 0.31710782740299065,
+      "learning_rate": 9.95194951094759e-06,
+      "loss": 0.0255,
+      "step": 694
+    },
+    {
+      "epoch": 0.18454593733404143,
+      "grad_norm": 0.7792504139919215,
+      "learning_rate": 9.951645404637139e-06,
+      "loss": 0.063,
+      "step": 695
+    },
+    {
+      "epoch": 0.18481147105682422,
+      "grad_norm": 0.24864366834341872,
+      "learning_rate": 9.951340343707852e-06,
+      "loss": 0.0184,
+      "step": 696
+    },
+    {
+      "epoch": 0.185077004779607,
+      "grad_norm": 1.3198814057549204,
+      "learning_rate": 9.951034328218543e-06,
+      "loss": 0.0815,
+      "step": 697
+    },
+    {
+      "epoch": 0.1853425385023898,
+      "grad_norm": 0.1349697962445958,
+      "learning_rate": 9.950727358228207e-06,
+      "loss": 0.004,
+      "step": 698
+    },
+    {
+      "epoch": 0.1856080722251726,
+      "grad_norm": 0.7940800117641265,
+      "learning_rate": 9.950419433796026e-06,
+      "loss": 0.0525,
+      "step": 699
+    },
+    {
+      "epoch": 0.18587360594795538,
+      "grad_norm": 0.17975503032083975,
+      "learning_rate": 9.950110554981363e-06,
+      "loss": 0.0045,
+      "step": 700
+    },
+    {
+      "epoch": 0.1861391396707382,
+      "grad_norm": 0.48368320799159703,
+      "learning_rate": 9.949800721843765e-06,
+      "loss": 0.0391,
+      "step": 701
+    },
+    {
+      "epoch": 0.186404673393521,
+      "grad_norm": 0.31589438590765767,
+      "learning_rate": 9.949489934442966e-06,
+      "loss": 0.0255,
+      "step": 702
+    },
+    {
+      "epoch": 0.18667020711630378,
+      "grad_norm": 0.31533609255408135,
+      "learning_rate": 9.949178192838881e-06,
+      "loss": 0.0258,
+      "step": 703
+    },
+    {
+      "epoch": 0.18693574083908657,
+      "grad_norm": 1.3466582105948832,
+      "learning_rate": 9.94886549709161e-06,
+      "loss": 0.1279,
+      "step": 704
+    },
+    {
+      "epoch": 0.18720127456186936,
+      "grad_norm": 0.910289445658781,
+      "learning_rate": 9.948551847261439e-06,
+      "loss": 0.0815,
+      "step": 705
+    },
+    {
+      "epoch": 0.18746680828465215,
+      "grad_norm": 0.21688834388185138,
+      "learning_rate": 9.948237243408834e-06,
+      "loss": 0.0165,
+      "step": 706
+    },
+    {
+      "epoch": 0.18773234200743494,
+      "grad_norm": 0.021656248810182913,
+      "learning_rate": 9.947921685594448e-06,
+      "loss": 0.0005,
+      "step": 707
+    },
+    {
+      "epoch": 0.18799787573021773,
+      "grad_norm": 0.9878506170969464,
+      "learning_rate": 9.947605173879116e-06,
+      "loss": 0.1279,
+      "step": 708
+    },
+    {
+      "epoch": 0.18826340945300052,
+      "grad_norm": 1.1026344698990722,
+      "learning_rate": 9.947287708323858e-06,
+      "loss": 0.1553,
+      "step": 709
+    },
+    {
+      "epoch": 0.18852894317578334,
+      "grad_norm": 0.6327076367042769,
+      "learning_rate": 9.946969288989878e-06,
+      "loss": 0.0579,
+      "step": 710
+    },
+    {
+      "epoch": 0.18879447689856613,
+      "grad_norm": 0.972281811713341,
+      "learning_rate": 9.946649915938562e-06,
+      "loss": 0.1367,
+      "step": 711
+    },
+    {
+      "epoch": 0.18906001062134892,
+      "grad_norm": 0.9045454418723899,
+      "learning_rate": 9.946329589231484e-06,
+      "loss": 0.1113,
+      "step": 712
+    },
+    {
+      "epoch": 0.1893255443441317,
+      "grad_norm": 1.1652383753491669,
+      "learning_rate": 9.946008308930396e-06,
+      "loss": 0.1035,
+      "step": 713
+    },
+    {
+      "epoch": 0.1895910780669145,
+      "grad_norm": 0.5983398386283125,
+      "learning_rate": 9.94568607509724e-06,
+      "loss": 0.0525,
+      "step": 714
+    },
+    {
+      "epoch": 0.1898566117896973,
+      "grad_norm": 0.9406342065879315,
+      "learning_rate": 9.945362887794139e-06,
+      "loss": 0.1279,
+      "step": 715
+    },
+    {
+      "epoch": 0.19012214551248008,
+      "grad_norm": 1.0678169277527272,
+      "learning_rate": 9.945038747083396e-06,
+      "loss": 0.1455,
+      "step": 716
+    },
+    {
+      "epoch": 0.19038767923526287,
+      "grad_norm": 0.6152395770617992,
+      "learning_rate": 9.944713653027505e-06,
+      "loss": 0.063,
+      "step": 717
+    },
+    {
+      "epoch": 0.19065321295804566,
+      "grad_norm": 0.7306184341026286,
+      "learning_rate": 9.944387605689139e-06,
+      "loss": 0.0693,
+      "step": 718
+    },
+    {
+      "epoch": 0.19091874668082848,
+      "grad_norm": 0.32716061364977655,
+      "learning_rate": 9.944060605131157e-06,
+      "loss": 0.0165,
+      "step": 719
+    },
+    {
+      "epoch": 0.19118428040361127,
+      "grad_norm": 0.9594346998876806,
+      "learning_rate": 9.943732651416597e-06,
+      "loss": 0.0957,
+      "step": 720
+    },
+    {
+      "epoch": 0.19144981412639406,
+      "grad_norm": 0.8678737283379834,
+      "learning_rate": 9.94340374460869e-06,
+      "loss": 0.1035,
+      "step": 721
+    },
+    {
+      "epoch": 0.19171534784917685,
+      "grad_norm": 0.5586020989155928,
+      "learning_rate": 9.943073884770844e-06,
+      "loss": 0.0579,
+      "step": 722
+    },
+    {
+      "epoch": 0.19198088157195964,
+      "grad_norm": 0.6710124613449318,
+      "learning_rate": 9.94274307196665e-06,
+      "loss": 0.0752,
+      "step": 723
+    },
+    {
+      "epoch": 0.19224641529474243,
+      "grad_norm": 0.11750671730822675,
+      "learning_rate": 9.942411306259887e-06,
+      "loss": 0.0041,
+      "step": 724
+    },
+    {
+      "epoch": 0.19251194901752522,
+      "grad_norm": 0.9492120597858461,
+      "learning_rate": 9.942078587714514e-06,
+      "loss": 0.1113,
+      "step": 725
+    },
+    {
+      "epoch": 0.192777482740308,
+      "grad_norm": 0.7578093561638131,
+      "learning_rate": 9.941744916394675e-06,
+      "loss": 0.0889,
+      "step": 726
+    },
+    {
+      "epoch": 0.1930430164630908,
+      "grad_norm": 0.0703405751086318,
+      "learning_rate": 9.9414102923647e-06,
+      "loss": 0.0017,
+      "step": 727
+    },
+    {
+      "epoch": 0.19330855018587362,
+      "grad_norm": 0.2720572743378476,
+      "learning_rate": 9.941074715689097e-06,
+      "loss": 0.0132,
+      "step": 728
+    },
+    {
+      "epoch": 0.1935740839086564,
+      "grad_norm": 0.47926399744773224,
+      "learning_rate": 9.940738186432565e-06,
+      "loss": 0.0479,
+      "step": 729
+    },
+    {
+      "epoch": 0.1938396176314392,
+      "grad_norm": 0.4960442155304019,
+      "learning_rate": 9.94040070465998e-06,
+      "loss": 0.053,
+      "step": 730
+    },
+    {
+      "epoch": 0.194105151354222,
+      "grad_norm": 0.6099810000661499,
+      "learning_rate": 9.940062270436406e-06,
+      "loss": 0.0752,
+      "step": 731
+    },
+    {
+      "epoch": 0.19437068507700478,
+      "grad_norm": 0.5317125466849125,
+      "learning_rate": 9.93972288382709e-06,
+      "loss": 0.0579,
+      "step": 732
+    },
+    {
+      "epoch": 0.19463621879978757,
+      "grad_norm": 0.405350264279859,
+      "learning_rate": 9.93938254489746e-06,
+      "loss": 0.0359,
+      "step": 733
+    },
+    {
+      "epoch": 0.19490175252257036,
+      "grad_norm": 0.5550223239319415,
+      "learning_rate": 9.93904125371313e-06,
+      "loss": 0.063,
+      "step": 734
+    },
+    {
+      "epoch": 0.19516728624535315,
+      "grad_norm": 0.4548926771748096,
+      "learning_rate": 9.938699010339898e-06,
+      "loss": 0.0479,
+      "step": 735
+    },
+    {
+      "epoch": 0.19543281996813594,
+      "grad_norm": 0.4301648170034228,
+      "learning_rate": 9.93835581484374e-06,
+      "loss": 0.0432,
+      "step": 736
+    },
+    {
+      "epoch": 0.19569835369091876,
+      "grad_norm": 0.5759338926245092,
+      "learning_rate": 9.938011667290828e-06,
+      "loss": 0.0476,
+      "step": 737
+    },
+    {
+      "epoch": 0.19596388741370155,
+      "grad_norm": 0.47327836419648434,
+      "learning_rate": 9.9376665677475e-06,
+      "loss": 0.0525,
+      "step": 738
+    },
+    {
+      "epoch": 0.19622942113648434,
+      "grad_norm": 0.1587521869962895,
+      "learning_rate": 9.937320516280297e-06,
+      "loss": 0.0051,
+      "step": 739
+    },
+    {
+      "epoch": 0.19649495485926713,
+      "grad_norm": 0.29964993842166077,
+      "learning_rate": 9.936973512955927e-06,
+      "loss": 0.0132,
+      "step": 740
+    },
+    {
+      "epoch": 0.19676048858204992,
+      "grad_norm": 0.5192574350505638,
+      "learning_rate": 9.936625557841289e-06,
+      "loss": 0.0525,
+      "step": 741
+    },
+    {
+      "epoch": 0.1970260223048327,
+      "grad_norm": 0.6174706812099604,
+      "learning_rate": 9.936276651003465e-06,
+      "loss": 0.0693,
+      "step": 742
+    },
+    {
+      "epoch": 0.1972915560276155,
+      "grad_norm": 0.4060064868344745,
+      "learning_rate": 9.935926792509723e-06,
+      "loss": 0.0435,
+      "step": 743
+    },
+    {
+      "epoch": 0.1975570897503983,
+      "grad_norm": 0.6447420229580227,
+      "learning_rate": 9.935575982427507e-06,
+      "loss": 0.082,
+      "step": 744
+    },
+    {
+      "epoch": 0.19782262347318108,
+      "grad_norm": 0.8009583314059334,
+      "learning_rate": 9.935224220824452e-06,
+      "loss": 0.1191,
+      "step": 745
+    },
+    {
+      "epoch": 0.1980881571959639,
+      "grad_norm": 0.7682777474476555,
+      "learning_rate": 9.934871507768374e-06,
+      "loss": 0.0688,
+      "step": 746
+    },
+    {
+      "epoch": 0.1983536909187467,
+      "grad_norm": 0.27537017109972733,
+      "learning_rate": 9.934517843327269e-06,
+      "loss": 0.0104,
+      "step": 747
+    },
+    {
+      "epoch": 0.19861922464152948,
+      "grad_norm": 0.4708650125151908,
+      "learning_rate": 9.93416322756932e-06,
+      "loss": 0.0479,
+      "step": 748
+    },
+    {
+      "epoch": 0.19888475836431227,
+      "grad_norm": 0.706657435045738,
+      "learning_rate": 9.933807660562898e-06,
+      "loss": 0.082,
+      "step": 749
+    },
+    {
+      "epoch": 0.19915029208709506,
+      "grad_norm": 0.36488237415402874,
+      "learning_rate": 9.933451142376545e-06,
+      "loss": 0.0354,
+      "step": 750
+    },
+    {
+      "epoch": 0.19941582580987785,
+      "grad_norm": 0.4861196249654044,
+      "learning_rate": 9.933093673078994e-06,
+      "loss": 0.0525,
+      "step": 751
+    },
+    {
+      "epoch": 0.19968135953266064,
+      "grad_norm": 0.5291778145248267,
+      "learning_rate": 9.932735252739168e-06,
+      "loss": 0.0393,
+      "step": 752
+    },
+    {
+      "epoch": 0.19994689325544343,
+      "grad_norm": 0.35001230301241637,
+      "learning_rate": 9.93237588142616e-06,
+      "loss": 0.0354,
+      "step": 753
+    },
+    {
+      "epoch": 0.20021242697822622,
+      "grad_norm": 2.417388946512036,
+      "learning_rate": 9.932015559209255e-06,
+      "loss": 0.0277,
+      "step": 754
+    },
+    {
+      "epoch": 0.20047796070100904,
+      "grad_norm": 0.1251433570133762,
+      "learning_rate": 9.931654286157915e-06,
+      "loss": 0.0032,
+      "step": 755
+    },
+    {
+      "epoch": 0.20074349442379183,
+      "grad_norm": 0.209414281849796,
+      "learning_rate": 9.931292062341793e-06,
+      "loss": 0.0047,
+      "step": 756
+    },
+    {
+      "epoch": 0.20100902814657462,
+      "grad_norm": 0.008292667150507884,
+      "learning_rate": 9.930928887830721e-06,
+      "loss": 0.0002,
+      "step": 757
+    },
+    {
+      "epoch": 0.2012745618693574,
+      "grad_norm": 0.38647994443034445,
+      "learning_rate": 9.930564762694715e-06,
+      "loss": 0.0393,
+      "step": 758
+    },
+    {
+      "epoch": 0.2015400955921402,
+      "grad_norm": 0.6378261596904626,
+      "learning_rate": 9.930199687003972e-06,
+      "loss": 0.0579,
+      "step": 759
+    },
+    {
+      "epoch": 0.201805629314923,
+      "grad_norm": 2.3592667920352035,
+      "learning_rate": 9.929833660828878e-06,
+      "loss": 0.1455,
+      "step": 760
+    },
+    {
+      "epoch": 0.20207116303770578,
+      "grad_norm": 0.6334539242279383,
+      "learning_rate": 9.929466684239993e-06,
+      "loss": 0.0525,
+      "step": 761
+    },
+    {
+      "epoch": 0.20233669676048857,
+      "grad_norm": 0.30334797132909946,
+      "learning_rate": 9.929098757308071e-06,
+      "loss": 0.0286,
+      "step": 762
+    },
+    {
+      "epoch": 0.20260223048327136,
+      "grad_norm": 0.30034960277983547,
+      "learning_rate": 9.928729880104042e-06,
+      "loss": 0.0258,
+      "step": 763
+    },
+    {
+      "epoch": 0.20286776420605418,
+      "grad_norm": 0.7158437599898999,
+      "learning_rate": 9.928360052699019e-06,
+      "loss": 0.0579,
+      "step": 764
+    },
+    {
+      "epoch": 0.20313329792883697,
+      "grad_norm": 0.4427519026261093,
+      "learning_rate": 9.927989275164305e-06,
+      "loss": 0.0432,
+      "step": 765
+    },
+    {
+      "epoch": 0.20339883165161976,
+      "grad_norm": 1.223371374907059,
+      "learning_rate": 9.927617547571377e-06,
+      "loss": 0.0957,
+      "step": 766
+    },
+    {
+      "epoch": 0.20366436537440255,
+      "grad_norm": 0.8525259837165133,
+      "learning_rate": 9.927244869991905e-06,
+      "loss": 0.1191,
+      "step": 767
+    },
+    {
+      "epoch": 0.20392989909718534,
+      "grad_norm": 0.7529517092699473,
+      "learning_rate": 9.92687124249773e-06,
+      "loss": 0.0889,
+      "step": 768
+    },
+    {
+      "epoch": 0.20419543281996813,
+      "grad_norm": 0.4231343502487387,
+      "learning_rate": 9.926496665160887e-06,
+      "loss": 0.0393,
+      "step": 769
+    },
+    {
+      "epoch": 0.20446096654275092,
+      "grad_norm": 0.02781249909477673,
+      "learning_rate": 9.926121138053591e-06,
+      "loss": 0.0005,
+      "step": 770
+    },
+    {
+      "epoch": 0.2047265002655337,
+      "grad_norm": 0.32019858843907156,
+      "learning_rate": 9.925744661248238e-06,
+      "loss": 0.0286,
+      "step": 771
+    },
+    {
+      "epoch": 0.20499203398831653,
+      "grad_norm": 0.34210995299202096,
+      "learning_rate": 9.925367234817407e-06,
+      "loss": 0.0317,
+      "step": 772
+    },
+    {
+      "epoch": 0.20525756771109932,
+      "grad_norm": 0.29667952505177037,
+      "learning_rate": 9.924988858833862e-06,
+      "loss": 0.0258,
+      "step": 773
+    },
+    {
+      "epoch": 0.2055231014338821,
+      "grad_norm": 0.775170587796152,
+      "learning_rate": 9.924609533370551e-06,
+      "loss": 0.082,
+      "step": 774
+    },
+    {
+      "epoch": 0.2057886351566649,
+      "grad_norm": 0.8776207151240932,
+      "learning_rate": 9.924229258500601e-06,
+      "loss": 0.0688,
+      "step": 775
+    },
+    {
+      "epoch": 0.2060541688794477,
+      "grad_norm": 0.34921604699592496,
+      "learning_rate": 9.923848034297327e-06,
+      "loss": 0.0286,
+      "step": 776
+    },
+    {
+      "epoch": 0.20631970260223048,
+      "grad_norm": 0.7094770354343616,
+      "learning_rate": 9.923465860834224e-06,
+      "loss": 0.063,
+      "step": 777
+    },
+    {
+      "epoch": 0.20658523632501327,
+      "grad_norm": 0.8576553061781719,
+      "learning_rate": 9.923082738184969e-06,
+      "loss": 0.0688,
+      "step": 778
+    },
+    {
+      "epoch": 0.20685077004779606,
+      "grad_norm": 0.4617621570202573,
+      "learning_rate": 9.922698666423424e-06,
+      "loss": 0.0126,
+      "step": 779
+    },
+    {
+      "epoch": 0.20711630377057885,
+      "grad_norm": 0.2701329483492822,
+      "learning_rate": 9.922313645623634e-06,
+      "loss": 0.0205,
+      "step": 780
+    },
+    {
+      "epoch": 0.20738183749336167,
+      "grad_norm": 1.2246770118493904,
+      "learning_rate": 9.921927675859828e-06,
+      "loss": 0.0576,
+      "step": 781
+    },
+    {
+      "epoch": 0.20764737121614446,
+      "grad_norm": 0.29636608608865783,
+      "learning_rate": 9.921540757206414e-06,
+      "loss": 0.0258,
+      "step": 782
+    },
+    {
+      "epoch": 0.20791290493892725,
+      "grad_norm": 1.8125953283598475,
+      "learning_rate": 9.921152889737985e-06,
+      "loss": 0.1367,
+      "step": 783
+    },
+    {
+      "epoch": 0.20817843866171004,
+      "grad_norm": 0.03047349596838622,
+      "learning_rate": 9.920764073529319e-06,
+      "loss": 0.0006,
+      "step": 784
+    },
+    {
+      "epoch": 0.20844397238449283,
+      "grad_norm": 0.16957814720325154,
+      "learning_rate": 9.920374308655375e-06,
+      "loss": 0.0041,
+      "step": 785
+    },
+    {
+      "epoch": 0.20870950610727562,
+      "grad_norm": 0.7111174333404663,
+      "learning_rate": 9.919983595191295e-06,
+      "loss": 0.0165,
+      "step": 786
+    },
+    {
+      "epoch": 0.2089750398300584,
+      "grad_norm": 1.244139227465599,
+      "learning_rate": 9.919591933212403e-06,
+      "loss": 0.0815,
+      "step": 787
+    },
+    {
+      "epoch": 0.2092405735528412,
+      "grad_norm": 0.34737267129055766,
+      "learning_rate": 9.919199322794209e-06,
+      "loss": 0.0286,
+      "step": 788
+    },
+    {
+      "epoch": 0.209506107275624,
+      "grad_norm": 0.9063191327124598,
+      "learning_rate": 9.9188057640124e-06,
+      "loss": 0.0479,
+      "step": 789
+    },
+    {
+      "epoch": 0.2097716409984068,
+      "grad_norm": 1.1996030088378673,
+      "learning_rate": 9.918411256942852e-06,
+      "loss": 0.1279,
+      "step": 790
+    },
+    {
+      "epoch": 0.2100371747211896,
+      "grad_norm": 0.3272617386081918,
+      "learning_rate": 9.918015801661622e-06,
+      "loss": 0.0231,
+      "step": 791
+    },
+    {
+      "epoch": 0.2103027084439724,
+      "grad_norm": 0.20150941486275553,
+      "learning_rate": 9.91761939824495e-06,
+      "loss": 0.0146,
+      "step": 792
+    },
+    {
+      "epoch": 0.21056824216675518,
+      "grad_norm": 0.9406775259417158,
+      "learning_rate": 9.917222046769252e-06,
+      "loss": 0.1035,
+      "step": 793
+    },
+    {
+      "epoch": 0.21083377588953797,
+      "grad_norm": 0.33726088549905536,
+      "learning_rate": 9.916823747311138e-06,
+      "loss": 0.0255,
+      "step": 794
+    },
+    {
+      "epoch": 0.21109930961232076,
+      "grad_norm": 0.24491234727145536,
+      "learning_rate": 9.916424499947395e-06,
+      "loss": 0.0184,
+      "step": 795
+    },
+    {
+      "epoch": 0.21136484333510355,
+      "grad_norm": 0.1878113074555544,
+      "learning_rate": 9.91602430475499e-06,
+      "loss": 0.0115,
+      "step": 796
+    },
+    {
+      "epoch": 0.21163037705788634,
+      "grad_norm": 0.030677918403306813,
+      "learning_rate": 9.91562316181108e-06,
+      "loss": 0.0008,
+      "step": 797
+    },
+    {
+      "epoch": 0.21189591078066913,
+      "grad_norm": 0.8660378755286174,
+      "learning_rate": 9.915221071193e-06,
+      "loss": 0.0476,
+      "step": 798
+    },
+    {
+      "epoch": 0.21216144450345195,
+      "grad_norm": 0.8549555937928719,
+      "learning_rate": 9.914818032978263e-06,
+      "loss": 0.0525,
+      "step": 799
+    },
+    {
+      "epoch": 0.21242697822623474,
+      "grad_norm": 0.6372317187162225,
+      "learning_rate": 9.914414047244578e-06,
+      "loss": 0.0317,
+      "step": 800
+    },
+    {
+      "epoch": 0.21269251194901753,
+      "grad_norm": 0.9907320369841311,
+      "learning_rate": 9.914009114069824e-06,
+      "loss": 0.0525,
+      "step": 801
+    },
+    {
+      "epoch": 0.21295804567180032,
+      "grad_norm": 1.3053615357778507,
+      "learning_rate": 9.913603233532067e-06,
+      "loss": 0.0391,
+      "step": 802
+    },
+    {
+      "epoch": 0.2132235793945831,
+      "grad_norm": 0.5689635893757348,
+      "learning_rate": 9.913196405709557e-06,
+      "loss": 0.0286,
+      "step": 803
+    },
+    {
+      "epoch": 0.2134891131173659,
+      "grad_norm": 1.538445081246666,
+      "learning_rate": 9.912788630680727e-06,
+      "loss": 0.0525,
+      "step": 804
+    },
+    {
+      "epoch": 0.2137546468401487,
+      "grad_norm": 0.4053859965192937,
+      "learning_rate": 9.91237990852419e-06,
+      "loss": 0.0165,
+      "step": 805
+    },
+    {
+      "epoch": 0.21402018056293148,
+      "grad_norm": 0.12115693287671359,
+      "learning_rate": 9.911970239318743e-06,
+      "loss": 0.0072,
+      "step": 806
+    },
+    {
+      "epoch": 0.21428571428571427,
+      "grad_norm": 0.02612612601513698,
+      "learning_rate": 9.911559623143364e-06,
+      "loss": 0.0004,
+      "step": 807
+    },
+    {
+      "epoch": 0.2145512480084971,
+      "grad_norm": 0.155521917709447,
+      "learning_rate": 9.911148060077216e-06,
+      "loss": 0.0092,
+      "step": 808
+    },
+    {
+      "epoch": 0.21481678173127988,
+      "grad_norm": 0.800502691704114,
+      "learning_rate": 9.910735550199647e-06,
+      "loss": 0.0317,
+      "step": 809
+    },
+    {
+      "epoch": 0.21508231545406267,
+      "grad_norm": 1.7498059645077653,
+      "learning_rate": 9.910322093590177e-06,
+      "loss": 0.1035,
+      "step": 810
+    },
+    {
+      "epoch": 0.21534784917684546,
+      "grad_norm": 0.5768858949644357,
+      "learning_rate": 9.90990769032852e-06,
+      "loss": 0.0255,
+      "step": 811
+    },
+    {
+      "epoch": 0.21561338289962825,
+      "grad_norm": 0.17436803776687265,
+      "learning_rate": 9.909492340494567e-06,
+      "loss": 0.0064,
+      "step": 812
+    },
+    {
+      "epoch": 0.21587891662241104,
+      "grad_norm": 1.2847308699648239,
+      "learning_rate": 9.909076044168393e-06,
+      "loss": 0.0476,
+      "step": 813
+    },
+    {
+      "epoch": 0.21614445034519383,
+      "grad_norm": 0.20641696265719725,
+      "learning_rate": 9.908658801430256e-06,
+      "loss": 0.0115,
+      "step": 814
+    },
+    {
+      "epoch": 0.21640998406797662,
+      "grad_norm": 0.4594930508016625,
+      "learning_rate": 9.908240612360593e-06,
+      "loss": 0.0205,
+      "step": 815
+    },
+    {
+      "epoch": 0.21667551779075941,
+      "grad_norm": 1.6460520494176336,
+      "learning_rate": 9.907821477040028e-06,
+      "loss": 0.0391,
+      "step": 816
+    },
+    {
+      "epoch": 0.21694105151354223,
+      "grad_norm": 0.006952372103528881,
+      "learning_rate": 9.907401395549364e-06,
+      "loss": 0.0,
+      "step": 817
+    },
+    {
+      "epoch": 0.21720658523632502,
+      "grad_norm": 1.625155011661868,
+      "learning_rate": 9.90698036796959e-06,
+      "loss": 0.2344,
+      "step": 818
+    },
+    {
+      "epoch": 0.2174721189591078,
+      "grad_norm": 0.0352752424800895,
+      "learning_rate": 9.906558394381872e-06,
+      "loss": 0.0017,
+      "step": 819
+    },
+    {
+      "epoch": 0.2177376526818906,
+      "grad_norm": 0.20879590616740903,
+      "learning_rate": 9.906135474867562e-06,
+      "loss": 0.0103,
+      "step": 820
+    },
+    {
+      "epoch": 0.2180031864046734,
+      "grad_norm": 1.6599267611839752,
+      "learning_rate": 9.905711609508198e-06,
+      "loss": 0.2676,
+      "step": 821
+    },
+    {
+      "epoch": 0.21826872012745618,
+      "grad_norm": 3.9796641629384975,
+      "learning_rate": 9.90528679838549e-06,
+      "loss": 0.1553,
+      "step": 822
+    },
+    {
+      "epoch": 0.21853425385023897,
+      "grad_norm": 0.0648341014830216,
+      "learning_rate": 9.904861041581341e-06,
+      "loss": 0.0035,
+      "step": 823
+    },
+    {
+      "epoch": 0.21879978757302176,
+      "grad_norm": 11.724369911160045,
+      "learning_rate": 9.904434339177831e-06,
+      "loss": 0.4199,
+      "step": 824
+    },
+    {
+      "epoch": 0.21906532129580455,
+      "grad_norm": 1.7629568442469772,
+      "learning_rate": 9.904006691257224e-06,
+      "loss": 0.2559,
+      "step": 825
+    },
+    {
+      "epoch": 0.21933085501858737,
+      "grad_norm": 0.10426293175973483,
+      "learning_rate": 9.903578097901963e-06,
+      "loss": 0.0064,
+      "step": 826
+    },
+    {
+      "epoch": 0.21959638874137016,
+      "grad_norm": 1.4924171036944354,
+      "learning_rate": 9.903148559194677e-06,
+      "loss": 0.1641,
+      "step": 827
+    },
+    {
+      "epoch": 0.21986192246415295,
+      "grad_norm": 1.3050509840328823,
+      "learning_rate": 9.902718075218176e-06,
+      "loss": 0.1934,
+      "step": 828
+    },
+    {
+      "epoch": 0.22012745618693574,
+      "grad_norm": 1.235993047459142,
+      "learning_rate": 9.902286646055453e-06,
+      "loss": 0.0576,
+      "step": 829
+    },
+    {
+      "epoch": 0.22039298990971853,
+      "grad_norm": 0.24248121114396165,
+      "learning_rate": 9.90185427178968e-06,
+      "loss": 0.0165,
+      "step": 830
+    },
+    {
+      "epoch": 0.22065852363250132,
+      "grad_norm": 4.175312997922109,
+      "learning_rate": 9.901420952504217e-06,
+      "loss": 0.0576,
+      "step": 831
+    },
+    {
+      "epoch": 0.22092405735528411,
+      "grad_norm": 3.2184596570988093,
+      "learning_rate": 9.9009866882826e-06,
+      "loss": 0.1279,
+      "step": 832
+    },
+    {
+      "epoch": 0.2211895910780669,
+      "grad_norm": 1.0077714273038476,
+      "learning_rate": 9.900551479208553e-06,
+      "loss": 0.0576,
+      "step": 833
+    },
+    {
+      "epoch": 0.2214551248008497,
+      "grad_norm": 0.5299304251634451,
+      "learning_rate": 9.900115325365975e-06,
+      "loss": 0.0432,
+      "step": 834
+    },
+    {
+      "epoch": 0.2217206585236325,
+      "grad_norm": 0.25144714721412126,
+      "learning_rate": 9.899678226838956e-06,
+      "loss": 0.0206,
+      "step": 835
+    },
+    {
+      "epoch": 0.2219861922464153,
+      "grad_norm": 1.4383846378372553,
+      "learning_rate": 9.89924018371176e-06,
+      "loss": 0.1191,
+      "step": 836
+    },
+    {
+      "epoch": 0.2222517259691981,
+      "grad_norm": 0.6128666274034679,
+      "learning_rate": 9.898801196068839e-06,
+      "loss": 0.0391,
+      "step": 837
+    },
+    {
+      "epoch": 0.22251725969198088,
+      "grad_norm": 1.6083587605392458,
+      "learning_rate": 9.898361263994823e-06,
+      "loss": 0.1035,
+      "step": 838
+    },
+    {
+      "epoch": 0.22278279341476367,
+      "grad_norm": 0.33105381078233614,
+      "learning_rate": 9.897920387574525e-06,
+      "loss": 0.0286,
+      "step": 839
+    },
+    {
+      "epoch": 0.22304832713754646,
+      "grad_norm": 0.581750205311232,
+      "learning_rate": 9.897478566892942e-06,
+      "loss": 0.0317,
+      "step": 840
+    },
+    {
+      "epoch": 0.22331386086032926,
+      "grad_norm": 1.103725743553766,
+      "learning_rate": 9.897035802035253e-06,
+      "loss": 0.0354,
+      "step": 841
+    },
+    {
+      "epoch": 0.22357939458311205,
+      "grad_norm": 0.41623629278157465,
+      "learning_rate": 9.896592093086816e-06,
+      "loss": 0.0354,
+      "step": 842
+    },
+    {
+      "epoch": 0.22384492830589484,
+      "grad_norm": 0.22870378430920638,
+      "learning_rate": 9.896147440133173e-06,
+      "loss": 0.002,
+      "step": 843
+    },
+    {
+      "epoch": 0.22411046202867765,
+      "grad_norm": 0.026554099075059814,
+      "learning_rate": 9.89570184326005e-06,
+      "loss": 0.0005,
+      "step": 844
+    },
+    {
+      "epoch": 0.22437599575146044,
+      "grad_norm": 1.0737697123349998,
+      "learning_rate": 9.895255302553349e-06,
+      "loss": 0.1641,
+      "step": 845
+    },
+    {
+      "epoch": 0.22464152947424323,
+      "grad_norm": 1.05755659361386,
+      "learning_rate": 9.89480781809916e-06,
+      "loss": 0.0688,
+      "step": 846
+    },
+    {
+      "epoch": 0.22490706319702602,
+      "grad_norm": 0.8318386749828283,
+      "learning_rate": 9.894359389983753e-06,
+      "loss": 0.1191,
+      "step": 847
+    },
+    {
+      "epoch": 0.22517259691980882,
+      "grad_norm": 0.38706813856524885,
+      "learning_rate": 9.89391001829358e-06,
+      "loss": 0.0354,
+      "step": 848
+    },
+    {
+      "epoch": 0.2254381306425916,
+      "grad_norm": 0.06799335395718269,
+      "learning_rate": 9.893459703115272e-06,
+      "loss": 0.0017,
+      "step": 849
+    },
+    {
+      "epoch": 0.2257036643653744,
+      "grad_norm": 0.36794444497590895,
+      "learning_rate": 9.893008444535647e-06,
+      "loss": 0.0093,
+      "step": 850
+    },
+    {
+      "epoch": 0.22596919808815719,
+      "grad_norm": 0.5439922261966197,
+      "learning_rate": 9.892556242641702e-06,
+      "loss": 0.0117,
+      "step": 851
+    },
+    {
+      "epoch": 0.22623473181093998,
+      "grad_norm": 0.8380785010711241,
+      "learning_rate": 9.892103097520615e-06,
+      "loss": 0.063,
+      "step": 852
+    },
+    {
+      "epoch": 0.2265002655337228,
+      "grad_norm": 0.5431138700099414,
+      "learning_rate": 9.89164900925975e-06,
+      "loss": 0.0288,
+      "step": 853
+    },
+    {
+      "epoch": 0.22676579925650558,
+      "grad_norm": 0.8146895584075347,
+      "learning_rate": 9.891193977946645e-06,
+      "loss": 0.1113,
+      "step": 854
+    },
+    {
+      "epoch": 0.22703133297928837,
+      "grad_norm": 0.6831359882459529,
+      "learning_rate": 9.890738003669029e-06,
+      "loss": 0.0688,
+      "step": 855
+    },
+    {
+      "epoch": 0.22729686670207117,
+      "grad_norm": 0.4121057526788246,
+      "learning_rate": 9.890281086514806e-06,
+      "loss": 0.0391,
+      "step": 856
+    },
+    {
+      "epoch": 0.22756240042485396,
+      "grad_norm": 0.7258123968019558,
+      "learning_rate": 9.889823226572066e-06,
+      "loss": 0.0752,
+      "step": 857
+    },
+    {
+      "epoch": 0.22782793414763675,
+      "grad_norm": 0.8294544573466394,
+      "learning_rate": 9.889364423929076e-06,
+      "loss": 0.1035,
+      "step": 858
+    },
+    {
+      "epoch": 0.22809346787041954,
+      "grad_norm": 1.051757101289739,
+      "learning_rate": 9.888904678674292e-06,
+      "loss": 0.1113,
+      "step": 859
+    },
+    {
+      "epoch": 0.22835900159320233,
+      "grad_norm": 1.3061340829563401,
+      "learning_rate": 9.888443990896344e-06,
+      "loss": 0.1113,
+      "step": 860
+    },
+    {
+      "epoch": 0.22862453531598512,
+      "grad_norm": 0.8993459080630722,
+      "learning_rate": 9.887982360684049e-06,
+      "loss": 0.0889,
+      "step": 861
+    },
+    {
+      "epoch": 0.22889006903876793,
+      "grad_norm": 0.7347323587084027,
+      "learning_rate": 9.887519788126404e-06,
+      "loss": 0.0752,
+      "step": 862
+    },
+    {
+      "epoch": 0.22915560276155073,
+      "grad_norm": 0.5477598732831455,
+      "learning_rate": 9.887056273312584e-06,
+      "loss": 0.0354,
+      "step": 863
+    },
+    {
+      "epoch": 0.22942113648433352,
+      "grad_norm": 0.35585991259844624,
+      "learning_rate": 9.886591816331953e-06,
+      "loss": 0.0233,
+      "step": 864
+    },
+    {
+      "epoch": 0.2296866702071163,
+      "grad_norm": 1.3171584886462089,
+      "learning_rate": 9.886126417274054e-06,
+      "loss": 0.1191,
+      "step": 865
+    },
+    {
+      "epoch": 0.2299522039298991,
+      "grad_norm": 0.5959273314376112,
+      "learning_rate": 9.885660076228606e-06,
+      "loss": 0.0354,
+      "step": 866
+    },
+    {
+      "epoch": 0.2302177376526819,
+      "grad_norm": 0.5887513008564544,
+      "learning_rate": 9.885192793285516e-06,
+      "loss": 0.063,
+      "step": 867
+    },
+    {
+      "epoch": 0.23048327137546468,
+      "grad_norm": 1.2261572998536514,
+      "learning_rate": 9.884724568534873e-06,
+      "loss": 0.1191,
+      "step": 868
+    },
+    {
+      "epoch": 0.23074880509824747,
+      "grad_norm": 3.6206803166086385,
+      "learning_rate": 9.884255402066943e-06,
+      "loss": 0.2891,
+      "step": 869
+    },
+    {
+      "epoch": 0.23101433882103026,
+      "grad_norm": 0.8824142077915467,
+      "learning_rate": 9.883785293972175e-06,
+      "loss": 0.0957,
+      "step": 870
+    },
+    {
+      "epoch": 0.23127987254381308,
+      "grad_norm": 0.46423245513429523,
+      "learning_rate": 9.8833142443412e-06,
+      "loss": 0.0393,
+      "step": 871
+    },
+    {
+      "epoch": 0.23154540626659587,
+      "grad_norm": 0.5138918912335925,
+      "learning_rate": 9.882842253264835e-06,
+      "loss": 0.0525,
+      "step": 872
+    },
+    {
+      "epoch": 0.23181093998937866,
+      "grad_norm": 0.5843357342302304,
+      "learning_rate": 9.882369320834068e-06,
+      "loss": 0.0693,
+      "step": 873
+    },
+    {
+      "epoch": 0.23207647371216145,
+      "grad_norm": 0.2520466827681795,
+      "learning_rate": 9.88189544714008e-06,
+      "loss": 0.0074,
+      "step": 874
+    },
+    {
+      "epoch": 0.23234200743494424,
+      "grad_norm": 0.5794422421534507,
+      "learning_rate": 9.881420632274226e-06,
+      "loss": 0.0693,
+      "step": 875
+    },
+    {
+      "epoch": 0.23260754115772703,
+      "grad_norm": 0.44568917664701574,
+      "learning_rate": 9.880944876328045e-06,
+      "loss": 0.0354,
+      "step": 876
+    },
+    {
+      "epoch": 0.23287307488050982,
+      "grad_norm": 0.5250220184195629,
+      "learning_rate": 9.88046817939326e-06,
+      "loss": 0.0579,
+      "step": 877
+    },
+    {
+      "epoch": 0.2331386086032926,
+      "grad_norm": 0.5550066871107131,
+      "learning_rate": 9.879990541561766e-06,
+      "loss": 0.063,
+      "step": 878
+    },
+    {
+      "epoch": 0.2334041423260754,
+      "grad_norm": 0.4307639707627933,
+      "learning_rate": 9.879511962925651e-06,
+      "loss": 0.0476,
+      "step": 879
+    },
+    {
+      "epoch": 0.23366967604885822,
+      "grad_norm": 2.7902947643287472,
+      "learning_rate": 9.879032443577179e-06,
+      "loss": 0.2344,
+      "step": 880
+    },
+    {
+      "epoch": 0.233935209771641,
+      "grad_norm": 0.14563410670518806,
+      "learning_rate": 9.878551983608794e-06,
+      "loss": 0.0036,
+      "step": 881
+    },
+    {
+      "epoch": 0.2342007434944238,
+      "grad_norm": 0.8774501103628786,
+      "learning_rate": 9.878070583113123e-06,
+      "loss": 0.0957,
+      "step": 882
+    },
+    {
+      "epoch": 0.2344662772172066,
+      "grad_norm": 0.5560581059426071,
+      "learning_rate": 9.877588242182976e-06,
+      "loss": 0.063,
+      "step": 883
+    },
+    {
+      "epoch": 0.23473181093998938,
+      "grad_norm": 3.7081276425034457,
+      "learning_rate": 9.877104960911341e-06,
+      "loss": 0.3301,
+      "step": 884
+    },
+    {
+      "epoch": 0.23499734466277217,
+      "grad_norm": 0.22236979379444108,
+      "learning_rate": 9.87662073939139e-06,
+      "loss": 0.0074,
+      "step": 885
+    },
+    {
+      "epoch": 0.23526287838555496,
+      "grad_norm": 0.4092243618958261,
+      "learning_rate": 9.876135577716475e-06,
+      "loss": 0.0259,
+      "step": 886
+    },
+    {
+      "epoch": 0.23552841210833775,
+      "grad_norm": 0.37933414096989954,
+      "learning_rate": 9.875649475980129e-06,
+      "loss": 0.0432,
+      "step": 887
+    },
+    {
+      "epoch": 0.23579394583112057,
+      "grad_norm": 0.5651151674620623,
+      "learning_rate": 9.875162434276066e-06,
+      "loss": 0.0579,
+      "step": 888
+    },
+    {
+      "epoch": 0.23605947955390336,
+      "grad_norm": 0.5162559290907474,
+      "learning_rate": 9.874674452698184e-06,
+      "loss": 0.0479,
+      "step": 889
+    },
+    {
+      "epoch": 0.23632501327668615,
+      "grad_norm": 0.637236975984389,
+      "learning_rate": 9.874185531340561e-06,
+      "loss": 0.0889,
+      "step": 890
+    },
+    {
+      "epoch": 0.23659054699946894,
+      "grad_norm": 0.37992930731240715,
+      "learning_rate": 9.87369567029745e-06,
+      "loss": 0.0432,
+      "step": 891
+    },
+    {
+      "epoch": 0.23685608072225173,
+      "grad_norm": 0.4304890878920463,
+      "learning_rate": 9.873204869663296e-06,
+      "loss": 0.0479,
+      "step": 892
+    },
+    {
+      "epoch": 0.23712161444503452,
+      "grad_norm": 0.15235504129186367,
+      "learning_rate": 9.872713129532717e-06,
+      "loss": 0.0052,
+      "step": 893
+    },
+    {
+      "epoch": 0.2373871481678173,
+      "grad_norm": 0.6165775215284957,
+      "learning_rate": 9.872220450000514e-06,
+      "loss": 0.0393,
+      "step": 894
+    },
+    {
+      "epoch": 0.2376526818906001,
+      "grad_norm": 0.11125726497429037,
+      "learning_rate": 9.87172683116167e-06,
+      "loss": 0.0018,
+      "step": 895
+    },
+    {
+      "epoch": 0.2379182156133829,
+      "grad_norm": 0.3081309691181636,
+      "learning_rate": 9.871232273111351e-06,
+      "loss": 0.0317,
+      "step": 896
+    },
+    {
+      "epoch": 0.2381837493361657,
+      "grad_norm": 0.428579523646676,
+      "learning_rate": 9.8707367759449e-06,
+      "loss": 0.0479,
+      "step": 897
+    },
+    {
+      "epoch": 0.2384492830589485,
+      "grad_norm": 1.2174720808088506,
+      "learning_rate": 9.870240339757844e-06,
+      "loss": 0.0752,
+      "step": 898
+    },
+    {
+      "epoch": 0.2387148167817313,
+      "grad_norm": 0.3515404509638987,
+      "learning_rate": 9.869742964645889e-06,
+      "loss": 0.0393,
+      "step": 899
+    },
+    {
+      "epoch": 0.23898035050451408,
+      "grad_norm": 0.36157621557162617,
+      "learning_rate": 9.869244650704924e-06,
+      "loss": 0.0393,
+      "step": 900
+    },
+    {
+      "epoch": 0.23924588422729687,
+      "grad_norm": 0.4552592244911392,
+      "learning_rate": 9.868745398031018e-06,
+      "loss": 0.0479,
+      "step": 901
+    },
+    {
+      "epoch": 0.23951141795007966,
+      "grad_norm": 0.7564817410807659,
+      "learning_rate": 9.86824520672042e-06,
+      "loss": 0.1191,
+      "step": 902
+    },
+    {
+      "epoch": 0.23977695167286245,
+      "grad_norm": 0.3519836929818125,
+      "learning_rate": 9.867744076869564e-06,
+      "loss": 0.0354,
+      "step": 903
+    },
+    {
+      "epoch": 0.24004248539564524,
+      "grad_norm": 0.5911859305106711,
+      "learning_rate": 9.867242008575056e-06,
+      "loss": 0.0579,
+      "step": 904
+    },
+    {
+      "epoch": 0.24030801911842803,
+      "grad_norm": 0.7584559784767734,
+      "learning_rate": 9.866739001933695e-06,
+      "loss": 0.1113,
+      "step": 905
+    },
+    {
+      "epoch": 0.24057355284121085,
+      "grad_norm": 0.7691613201005841,
+      "learning_rate": 9.866235057042453e-06,
+      "loss": 0.1035,
+      "step": 906
+    },
+    {
+      "epoch": 0.24083908656399364,
+      "grad_norm": 0.7497833122031072,
+      "learning_rate": 9.865730173998482e-06,
+      "loss": 0.1191,
+      "step": 907
+    },
+    {
+      "epoch": 0.24110462028677643,
+      "grad_norm": 0.754572113206245,
+      "learning_rate": 9.86522435289912e-06,
+      "loss": 0.1279,
+      "step": 908
+    },
+    {
+      "epoch": 0.24137015400955922,
+      "grad_norm": 0.8672227298778474,
+      "learning_rate": 9.864717593841884e-06,
+      "loss": 0.0752,
+      "step": 909
+    },
+    {
+      "epoch": 0.241635687732342,
+      "grad_norm": 0.4009604044104567,
+      "learning_rate": 9.864209896924468e-06,
+      "loss": 0.0393,
+      "step": 910
+    },
+    {
+      "epoch": 0.2419012214551248,
+      "grad_norm": 0.2966195431778511,
+      "learning_rate": 9.863701262244753e-06,
+      "loss": 0.032,
+      "step": 911
+    },
+    {
+      "epoch": 0.2421667551779076,
+      "grad_norm": 0.5999702384639466,
+      "learning_rate": 9.863191689900796e-06,
+      "loss": 0.063,
+      "step": 912
+    },
+    {
+      "epoch": 0.24243228890069038,
+      "grad_norm": 0.5132953676012183,
+      "learning_rate": 9.862681179990838e-06,
+      "loss": 0.0231,
+      "step": 913
+    },
+    {
+      "epoch": 0.24269782262347317,
+      "grad_norm": 0.6668918417364565,
+      "learning_rate": 9.862169732613298e-06,
+      "loss": 0.0525,
+      "step": 914
+    },
+    {
+      "epoch": 0.242963356346256,
+      "grad_norm": 0.6465213387617673,
+      "learning_rate": 9.861657347866778e-06,
+      "loss": 0.0957,
+      "step": 915
+    },
+    {
+      "epoch": 0.24322889006903878,
+      "grad_norm": 0.2980696299642875,
+      "learning_rate": 9.861144025850058e-06,
+      "loss": 0.0317,
+      "step": 916
+    },
+    {
+      "epoch": 0.24349442379182157,
+      "grad_norm": 0.5561195146487935,
+      "learning_rate": 9.860629766662103e-06,
+      "loss": 0.0354,
+      "step": 917
+    },
+    {
+      "epoch": 0.24375995751460436,
+      "grad_norm": 0.8511473475660397,
+      "learning_rate": 9.860114570402055e-06,
+      "loss": 0.0354,
+      "step": 918
+    },
+    {
+      "epoch": 0.24402549123738715,
+      "grad_norm": 1.0193129893064767,
+      "learning_rate": 9.859598437169236e-06,
+      "loss": 0.082,
+      "step": 919
+    },
+    {
+      "epoch": 0.24429102496016994,
+      "grad_norm": 0.38753475869451304,
+      "learning_rate": 9.859081367063154e-06,
+      "loss": 0.0432,
+      "step": 920
+    },
+    {
+      "epoch": 0.24455655868295273,
+      "grad_norm": 1.2376062406421997,
+      "learning_rate": 9.858563360183491e-06,
+      "loss": 0.0815,
+      "step": 921
+    },
+    {
+      "epoch": 0.24482209240573552,
+      "grad_norm": 0.2614059783802485,
+      "learning_rate": 9.858044416630113e-06,
+      "loss": 0.0288,
+      "step": 922
+    },
+    {
+      "epoch": 0.2450876261285183,
+      "grad_norm": 0.7298755606796804,
+      "learning_rate": 9.857524536503067e-06,
+      "loss": 0.1191,
+      "step": 923
+    },
+    {
+      "epoch": 0.24535315985130113,
+      "grad_norm": 0.2655675767167442,
+      "learning_rate": 9.85700371990258e-06,
+      "loss": 0.0286,
+      "step": 924
+    },
+    {
+      "epoch": 0.24561869357408392,
+      "grad_norm": 0.6867605169602318,
+      "learning_rate": 9.85648196692906e-06,
+      "loss": 0.063,
+      "step": 925
+    },
+    {
+      "epoch": 0.2458842272968667,
+      "grad_norm": 0.7154487582123846,
+      "learning_rate": 9.855959277683092e-06,
+      "loss": 0.1113,
+      "step": 926
+    },
+    {
+      "epoch": 0.2461497610196495,
+      "grad_norm": 0.7662936924895251,
+      "learning_rate": 9.855435652265446e-06,
+      "loss": 0.0889,
+      "step": 927
+    },
+    {
+      "epoch": 0.2464152947424323,
+      "grad_norm": 0.9128535852883102,
+      "learning_rate": 9.854911090777071e-06,
+      "loss": 0.0815,
+      "step": 928
+    },
+    {
+      "epoch": 0.24668082846521508,
+      "grad_norm": 0.6423625754170631,
+      "learning_rate": 9.854385593319096e-06,
+      "loss": 0.082,
+      "step": 929
+    },
+    {
+      "epoch": 0.24694636218799787,
+      "grad_norm": 0.3506830449134971,
+      "learning_rate": 9.853859159992831e-06,
+      "loss": 0.0393,
+      "step": 930
+    },
+    {
+      "epoch": 0.24721189591078066,
+      "grad_norm": 0.6686174900920513,
+      "learning_rate": 9.853331790899765e-06,
+      "loss": 0.0957,
+      "step": 931
+    },
+    {
+      "epoch": 0.24747742963356345,
+      "grad_norm": 0.5803696610157401,
+      "learning_rate": 9.852803486141571e-06,
+      "loss": 0.0479,
+      "step": 932
+    },
+    {
+      "epoch": 0.24774296335634627,
+      "grad_norm": 0.6367144973315528,
+      "learning_rate": 9.852274245820097e-06,
+      "loss": 0.0889,
+      "step": 933
+    },
+    {
+      "epoch": 0.24800849707912906,
+      "grad_norm": 0.6019312344886419,
+      "learning_rate": 9.851744070037376e-06,
+      "loss": 0.082,
+      "step": 934
+    },
+    {
+      "epoch": 0.24827403080191185,
+      "grad_norm": 0.40743643859287837,
+      "learning_rate": 9.851212958895621e-06,
+      "loss": 0.032,
+      "step": 935
+    },
+    {
+      "epoch": 0.24853956452469464,
+      "grad_norm": 0.5172725409243906,
+      "learning_rate": 9.85068091249722e-06,
+      "loss": 0.063,
+      "step": 936
+    },
+    {
+      "epoch": 0.24880509824747743,
+      "grad_norm": 0.4756469334653875,
+      "learning_rate": 9.850147930944748e-06,
+      "loss": 0.0525,
+      "step": 937
+    },
+    {
+      "epoch": 0.24907063197026022,
+      "grad_norm": 0.4365306093001885,
+      "learning_rate": 9.849614014340959e-06,
+      "loss": 0.0432,
+      "step": 938
+    },
+    {
+      "epoch": 0.249336165693043,
+      "grad_norm": 0.6047425723994772,
+      "learning_rate": 9.849079162788782e-06,
+      "loss": 0.0752,
+      "step": 939
+    },
+    {
+      "epoch": 0.2496016994158258,
+      "grad_norm": 0.5680420397578201,
+      "learning_rate": 9.848543376391333e-06,
+      "loss": 0.0693,
+      "step": 940
+    },
+    {
+      "epoch": 0.2498672331386086,
+      "grad_norm": 0.8494400098616096,
+      "learning_rate": 9.848006655251904e-06,
+      "loss": 0.0393,
+      "step": 941
+    },
+    {
+      "epoch": 0.2501327668613914,
+      "grad_norm": 0.20707760476213186,
+      "learning_rate": 9.847468999473967e-06,
+      "loss": 0.0074,
+      "step": 942
+    },
+    {
+      "epoch": 0.25039830058417417,
+      "grad_norm": 0.5599934007032675,
+      "learning_rate": 9.846930409161179e-06,
+      "loss": 0.063,
+      "step": 943
+    },
+    {
+      "epoch": 0.250663834306957,
+      "grad_norm": 0.5738129646742892,
+      "learning_rate": 9.846390884417374e-06,
+      "loss": 0.0693,
+      "step": 944
+    },
+    {
+      "epoch": 0.25092936802973975,
+      "grad_norm": 0.3121394791058085,
+      "learning_rate": 9.845850425346563e-06,
+      "loss": 0.0093,
+      "step": 945
+    },
+    {
+      "epoch": 0.25119490175252257,
+      "grad_norm": 0.9012275305666327,
+      "learning_rate": 9.845309032052941e-06,
+      "loss": 0.1035,
+      "step": 946
+    },
+    {
+      "epoch": 0.2514604354753054,
+      "grad_norm": 0.8132551945926066,
+      "learning_rate": 9.844766704640883e-06,
+      "loss": 0.1035,
+      "step": 947
+    },
+    {
+      "epoch": 0.25172596919808815,
+      "grad_norm": 0.5059511213268292,
+      "learning_rate": 9.844223443214942e-06,
+      "loss": 0.0576,
+      "step": 948
+    },
+    {
+      "epoch": 0.25199150292087097,
+      "grad_norm": 0.5004469864929803,
+      "learning_rate": 9.843679247879856e-06,
+      "loss": 0.0579,
+      "step": 949
+    },
+    {
+      "epoch": 0.25225703664365373,
+      "grad_norm": 0.5325050038377581,
+      "learning_rate": 9.843134118740535e-06,
+      "loss": 0.063,
+      "step": 950
+    },
+    {
+      "epoch": 0.25252257036643655,
+      "grad_norm": 0.0018393961535569554,
+      "learning_rate": 9.842588055902076e-06,
+      "loss": 0.0,
+      "step": 951
+    },
+    {
+      "epoch": 0.2527881040892193,
+      "grad_norm": 0.29281234077102586,
+      "learning_rate": 9.842041059469752e-06,
+      "loss": 0.0117,
+      "step": 952
+    },
+    {
+      "epoch": 0.25305363781200213,
+      "grad_norm": 4.655974030022523,
+      "learning_rate": 9.841493129549018e-06,
+      "loss": 0.4082,
+      "step": 953
+    },
+    {
+      "epoch": 0.2533191715347849,
+      "grad_norm": 0.345197837145536,
+      "learning_rate": 9.840944266245511e-06,
+      "loss": 0.0354,
+      "step": 954
+    },
+    {
+      "epoch": 0.2535847052575677,
+      "grad_norm": 0.5150522007650085,
+      "learning_rate": 9.840394469665043e-06,
+      "loss": 0.032,
+      "step": 955
+    },
+    {
+      "epoch": 0.25385023898035053,
+      "grad_norm": 4.355092056512744,
+      "learning_rate": 9.839843739913606e-06,
+      "loss": 0.3438,
+      "step": 956
+    },
+    {
+      "epoch": 0.2541157727031333,
+      "grad_norm": 0.5533036225395148,
+      "learning_rate": 9.839292077097378e-06,
+      "loss": 0.063,
+      "step": 957
+    },
+    {
+      "epoch": 0.2543813064259161,
+      "grad_norm": 0.47245181913716905,
+      "learning_rate": 9.838739481322711e-06,
+      "loss": 0.032,
+      "step": 958
+    },
+    {
+      "epoch": 0.25464684014869887,
+      "grad_norm": 0.015777567489265145,
+      "learning_rate": 9.838185952696139e-06,
+      "loss": 0.0004,
+      "step": 959
+    },
+    {
+      "epoch": 0.2549123738714817,
+      "grad_norm": 0.3070632948499519,
+      "learning_rate": 9.837631491324379e-06,
+      "loss": 0.0317,
+      "step": 960
+    },
+    {
+      "epoch": 0.25517790759426445,
+      "grad_norm": 0.6417501675077917,
+      "learning_rate": 9.83707609731432e-06,
+      "loss": 0.0525,
+      "step": 961
+    },
+    {
+      "epoch": 0.25544344131704727,
+      "grad_norm": 1.103140251013234,
+      "learning_rate": 9.836519770773038e-06,
+      "loss": 0.0889,
+      "step": 962
+    },
+    {
+      "epoch": 0.25570897503983003,
+      "grad_norm": 0.36859782996784346,
+      "learning_rate": 9.835962511807786e-06,
+      "loss": 0.0391,
+      "step": 963
+    },
+    {
+      "epoch": 0.25597450876261285,
+      "grad_norm": 0.3995475012233262,
+      "learning_rate": 9.835404320525997e-06,
+      "loss": 0.0432,
+      "step": 964
+    },
+    {
+      "epoch": 0.25624004248539567,
+      "grad_norm": 0.6669653652071353,
+      "learning_rate": 9.834845197035282e-06,
+      "loss": 0.0752,
+      "step": 965
+    },
+    {
+      "epoch": 0.25650557620817843,
+      "grad_norm": 0.9463153272300644,
+      "learning_rate": 9.834285141443436e-06,
+      "loss": 0.082,
+      "step": 966
+    },
+    {
+      "epoch": 0.25677110993096125,
+      "grad_norm": 0.7472061020733514,
+      "learning_rate": 9.833724153858428e-06,
+      "loss": 0.0479,
+      "step": 967
+    },
+    {
+      "epoch": 0.257036643653744,
+      "grad_norm": 0.5964140946194706,
+      "learning_rate": 9.833162234388413e-06,
+      "loss": 0.0525,
+      "step": 968
+    },
+    {
+      "epoch": 0.25730217737652683,
+      "grad_norm": 0.69941976438153,
+      "learning_rate": 9.832599383141722e-06,
+      "loss": 0.0957,
+      "step": 969
+    },
+    {
+      "epoch": 0.2575677110993096,
+      "grad_norm": 0.6938979357435278,
+      "learning_rate": 9.832035600226863e-06,
+      "loss": 0.0957,
+      "step": 970
+    },
+    {
+      "epoch": 0.2578332448220924,
+      "grad_norm": 1.6505906761311266,
+      "learning_rate": 9.83147088575253e-06,
+      "loss": 0.0957,
+      "step": 971
+    },
+    {
+      "epoch": 0.2580987785448752,
+      "grad_norm": 0.2571270449075492,
+      "learning_rate": 9.830905239827592e-06,
+      "loss": 0.0258,
+      "step": 972
+    },
+    {
+      "epoch": 0.258364312267658,
+      "grad_norm": 0.17362195702763106,
+      "learning_rate": 9.830338662561098e-06,
+      "loss": 0.0059,
+      "step": 973
+    },
+    {
+      "epoch": 0.2586298459904408,
+      "grad_norm": 0.3252416871542705,
+      "learning_rate": 9.82977115406228e-06,
+      "loss": 0.0354,
+      "step": 974
+    },
+    {
+      "epoch": 0.2588953797132236,
+      "grad_norm": 0.7560895172752308,
+      "learning_rate": 9.829202714440544e-06,
+      "loss": 0.0752,
+      "step": 975
+    },
+    {
+      "epoch": 0.2591609134360064,
+      "grad_norm": 0.8515843549515478,
+      "learning_rate": 9.828633343805482e-06,
+      "loss": 0.032,
+      "step": 976
+    },
+    {
+      "epoch": 0.25942644715878915,
+      "grad_norm": 0.06839867591689461,
+      "learning_rate": 9.828063042266858e-06,
+      "loss": 0.0018,
+      "step": 977
+    },
+    {
+      "epoch": 0.25969198088157197,
+      "grad_norm": 0.2973877835283162,
+      "learning_rate": 9.827491809934621e-06,
+      "loss": 0.0317,
+      "step": 978
+    },
+    {
+      "epoch": 0.25995751460435473,
+      "grad_norm": 0.050237992088855266,
+      "learning_rate": 9.826919646918899e-06,
+      "loss": 0.0014,
+      "step": 979
+    },
+    {
+      "epoch": 0.26022304832713755,
+      "grad_norm": 0.6244940971022646,
+      "learning_rate": 9.826346553329997e-06,
+      "loss": 0.063,
+      "step": 980
+    },
+    {
+      "epoch": 0.2604885820499203,
+      "grad_norm": 0.3832845757482919,
+      "learning_rate": 9.825772529278402e-06,
+      "loss": 0.0393,
+      "step": 981
+    },
+    {
+      "epoch": 0.26075411577270313,
+      "grad_norm": 0.5607756019076436,
+      "learning_rate": 9.825197574874775e-06,
+      "loss": 0.0288,
+      "step": 982
+    },
+    {
+      "epoch": 0.26101964949548595,
+      "grad_norm": 0.3614253694825283,
+      "learning_rate": 9.824621690229966e-06,
+      "loss": 0.0354,
+      "step": 983
+    },
+    {
+      "epoch": 0.2612851832182687,
+      "grad_norm": 1.7457333721108772,
+      "learning_rate": 9.824044875454995e-06,
+      "loss": 0.0815,
+      "step": 984
+    },
+    {
+      "epoch": 0.26155071694105153,
+      "grad_norm": 0.39656420534134673,
+      "learning_rate": 9.823467130661068e-06,
+      "loss": 0.0354,
+      "step": 985
+    },
+    {
+      "epoch": 0.2618162506638343,
+      "grad_norm": 0.4441491434154354,
+      "learning_rate": 9.822888455959564e-06,
+      "loss": 0.0391,
+      "step": 986
+    },
+    {
+      "epoch": 0.2620817843866171,
+      "grad_norm": 0.9091935280621182,
+      "learning_rate": 9.822308851462046e-06,
+      "loss": 0.063,
+      "step": 987
+    },
+    {
+      "epoch": 0.2623473181093999,
+      "grad_norm": 0.2809293161710395,
+      "learning_rate": 9.821728317280257e-06,
+      "loss": 0.0258,
+      "step": 988
+    },
+    {
+      "epoch": 0.2626128518321827,
+      "grad_norm": 0.25690503502749795,
+      "learning_rate": 9.821146853526116e-06,
+      "loss": 0.0231,
+      "step": 989
+    },
+    {
+      "epoch": 0.26287838555496545,
+      "grad_norm": 0.9504240651916706,
+      "learning_rate": 9.820564460311719e-06,
+      "loss": 0.1035,
+      "step": 990
+    },
+    {
+      "epoch": 0.2631439192777483,
+      "grad_norm": 1.062595886433007,
+      "learning_rate": 9.81998113774935e-06,
+      "loss": 0.1553,
+      "step": 991
+    },
+    {
+      "epoch": 0.2634094530005311,
+      "grad_norm": 0.08306616186641885,
+      "learning_rate": 9.819396885951462e-06,
+      "loss": 0.0022,
+      "step": 992
+    },
+    {
+      "epoch": 0.26367498672331385,
+      "grad_norm": 0.00883581144917326,
+      "learning_rate": 9.818811705030696e-06,
+      "loss": 0.0002,
+      "step": 993
+    },
+    {
+      "epoch": 0.26394052044609667,
+      "grad_norm": 0.1998801649088699,
+      "learning_rate": 9.818225595099864e-06,
+      "loss": 0.0165,
+      "step": 994
+    },
+    {
+      "epoch": 0.26420605416887943,
+      "grad_norm": 0.8818331545317252,
+      "learning_rate": 9.817638556271964e-06,
+      "loss": 0.1367,
+      "step": 995
+    },
+    {
+      "epoch": 0.26447158789166225,
+      "grad_norm": 1.6642224007144948,
+      "learning_rate": 9.81705058866017e-06,
+      "loss": 0.0752,
+      "step": 996
+    },
+    {
+      "epoch": 0.264737121614445,
+      "grad_norm": 0.719955269890806,
+      "learning_rate": 9.816461692377833e-06,
+      "loss": 0.0317,
+      "step": 997
+    },
+    {
+      "epoch": 0.26500265533722783,
+      "grad_norm": 0.7209406456988616,
+      "learning_rate": 9.815871867538486e-06,
+      "loss": 0.0476,
+      "step": 998
+    },
+    {
+      "epoch": 0.2652681890600106,
+      "grad_norm": 0.814061894388583,
+      "learning_rate": 9.815281114255841e-06,
+      "loss": 0.063,
+      "step": 999
+    },
+    {
+      "epoch": 0.2655337227827934,
+      "grad_norm": 0.9966166060448677,
+      "learning_rate": 9.81468943264379e-06,
+      "loss": 0.1035,
+      "step": 1000
+    },
+    {
+      "epoch": 0.26579925650557623,
+      "grad_norm": 0.22419755071513356,
+      "learning_rate": 9.814096822816398e-06,
+      "loss": 0.0061,
+      "step": 1001
+    },
+    {
+      "epoch": 0.266064790228359,
+      "grad_norm": 0.2754600582043281,
+      "learning_rate": 9.813503284887916e-06,
+      "loss": 0.0258,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2663303239511418,
+      "grad_norm": 1.1239681806680273,
+      "learning_rate": 9.812908818972771e-06,
+      "loss": 0.0889,
+      "step": 1003
+    },
+    {
+      "epoch": 0.2665958576739246,
+      "grad_norm": 0.5803488339467218,
+      "learning_rate": 9.81231342518557e-06,
+      "loss": 0.0476,
+      "step": 1004
+    },
+    {
+      "epoch": 0.2668613913967074,
+      "grad_norm": 0.35519796165827866,
+      "learning_rate": 9.811717103641096e-06,
+      "loss": 0.0286,
+      "step": 1005
+    },
+    {
+      "epoch": 0.26712692511949016,
+      "grad_norm": 2.462660544438473,
+      "learning_rate": 9.811119854454313e-06,
+      "loss": 0.0957,
+      "step": 1006
+    },
+    {
+      "epoch": 0.267392458842273,
+      "grad_norm": 0.0859060653931223,
+      "learning_rate": 9.810521677740363e-06,
+      "loss": 0.0022,
+      "step": 1007
+    },
+    {
+      "epoch": 0.26765799256505574,
+      "grad_norm": 0.21988261467431752,
+      "learning_rate": 9.80992257361457e-06,
+      "loss": 0.0206,
+      "step": 1008
+    },
+    {
+      "epoch": 0.26792352628783855,
+      "grad_norm": 0.37935970444692796,
+      "learning_rate": 9.809322542192432e-06,
+      "loss": 0.0317,
+      "step": 1009
+    },
+    {
+      "epoch": 0.26818906001062137,
+      "grad_norm": 0.3156766850538133,
+      "learning_rate": 9.808721583589628e-06,
+      "loss": 0.0286,
+      "step": 1010
+    },
+    {
+      "epoch": 0.26845459373340413,
+      "grad_norm": 0.17021294848589957,
+      "learning_rate": 9.808119697922016e-06,
+      "loss": 0.0146,
+      "step": 1011
+    },
+    {
+      "epoch": 0.26872012745618695,
+      "grad_norm": 1.3132137393142103,
+      "learning_rate": 9.807516885305634e-06,
+      "loss": 0.1641,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2689856611789697,
+      "grad_norm": 0.14707242483307284,
+      "learning_rate": 9.806913145856697e-06,
+      "loss": 0.0115,
+      "step": 1013
+    },
+    {
+      "epoch": 0.26925119490175253,
+      "grad_norm": 3.511760120759758,
+      "learning_rate": 9.806308479691595e-06,
+      "loss": 0.2061,
+      "step": 1014
+    },
+    {
+      "epoch": 0.2695167286245353,
+      "grad_norm": 0.8061004291701489,
+      "learning_rate": 9.805702886926904e-06,
+      "loss": 0.0261,
+      "step": 1015
+    },
+    {
+      "epoch": 0.2697822623473181,
+      "grad_norm": 0.2234690978519963,
+      "learning_rate": 9.805096367679375e-06,
+      "loss": 0.0184,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2700477960701009,
+      "grad_norm": 0.44975353193446077,
+      "learning_rate": 9.804488922065937e-06,
+      "loss": 0.0258,
+      "step": 1017
+    },
+    {
+      "epoch": 0.2703133297928837,
+      "grad_norm": 0.6109060230396508,
+      "learning_rate": 9.8038805502037e-06,
+      "loss": 0.0354,
+      "step": 1018
+    },
+    {
+      "epoch": 0.2705788635156665,
+      "grad_norm": 0.19828294757841264,
+      "learning_rate": 9.803271252209947e-06,
+      "loss": 0.0184,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2708443972384493,
+      "grad_norm": 0.25772979651426625,
+      "learning_rate": 9.802661028202147e-06,
+      "loss": 0.0165,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2711099309612321,
+      "grad_norm": 1.1839479936613346,
+      "learning_rate": 9.802049878297942e-06,
+      "loss": 0.0525,
+      "step": 1021
+    },
+    {
+      "epoch": 0.27137546468401486,
+      "grad_norm": 0.31430038721241105,
+      "learning_rate": 9.801437802615156e-06,
+      "loss": 0.0082,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2716409984067977,
+      "grad_norm": 1.1023674078125263,
+      "learning_rate": 9.80082480127179e-06,
+      "loss": 0.1191,
+      "step": 1023
+    },
+    {
+      "epoch": 0.27190653212958044,
+      "grad_norm": 0.22162338231371304,
+      "learning_rate": 9.800210874386022e-06,
+      "loss": 0.0117,
+      "step": 1024
+    },
+    {
+      "epoch": 0.27217206585236325,
+      "grad_norm": 0.020666537436477367,
+      "learning_rate": 9.79959602207621e-06,
+      "loss": 0.0004,
+      "step": 1025
+    },
+    {
+      "epoch": 0.272437599575146,
+      "grad_norm": 0.1640864385317412,
+      "learning_rate": 9.798980244460892e-06,
+      "loss": 0.0131,
+      "step": 1026
+    },
+    {
+      "epoch": 0.27270313329792883,
+      "grad_norm": 1.33096980689375,
+      "learning_rate": 9.798363541658782e-06,
+      "loss": 0.1455,
+      "step": 1027
+    },
+    {
+      "epoch": 0.27296866702071165,
+      "grad_norm": 0.0030561000010345305,
+      "learning_rate": 9.797745913788772e-06,
+      "loss": 0.0001,
+      "step": 1028
+    },
+    {
+      "epoch": 0.2732342007434944,
+      "grad_norm": 0.3353151513139151,
+      "learning_rate": 9.797127360969933e-06,
+      "loss": 0.0093,
+      "step": 1029
+    },
+    {
+      "epoch": 0.27349973446627723,
+      "grad_norm": 0.24279565510695517,
+      "learning_rate": 9.796507883321519e-06,
+      "loss": 0.0165,
+      "step": 1030
+    },
+    {
+      "epoch": 0.27376526818906,
+      "grad_norm": 0.6409923619811838,
+      "learning_rate": 9.795887480962951e-06,
+      "loss": 0.0206,
+      "step": 1031
+    },
+    {
+      "epoch": 0.2740308019118428,
+      "grad_norm": 0.1865018425314411,
+      "learning_rate": 9.79526615401384e-06,
+      "loss": 0.0146,
+      "step": 1032
+    },
+    {
+      "epoch": 0.2742963356346256,
+      "grad_norm": 0.5691428064000503,
+      "learning_rate": 9.79464390259397e-06,
+      "loss": 0.0354,
+      "step": 1033
+    },
+    {
+      "epoch": 0.2745618693574084,
+      "grad_norm": 1.0718553071053665,
+      "learning_rate": 9.794020726823303e-06,
+      "loss": 0.1934,
+      "step": 1034
+    },
+    {
+      "epoch": 0.27482740308019116,
+      "grad_norm": 1.338016801862817,
+      "learning_rate": 9.79339662682198e-06,
+      "loss": 0.1279,
+      "step": 1035
+    },
+    {
+      "epoch": 0.275092936802974,
+      "grad_norm": 0.3434764844856958,
+      "learning_rate": 9.79277160271032e-06,
+      "loss": 0.0229,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2753584705257568,
+      "grad_norm": 0.9536560542927063,
+      "learning_rate": 9.79214565460882e-06,
+      "loss": 0.1367,
+      "step": 1037
+    },
+    {
+      "epoch": 0.27562400424853956,
+      "grad_norm": 0.2777647682007733,
+      "learning_rate": 9.791518782638157e-06,
+      "loss": 0.0229,
+      "step": 1038
+    },
+    {
+      "epoch": 0.2758895379713224,
+      "grad_norm": 2.8846935746827778,
+      "learning_rate": 9.790890986919184e-06,
+      "loss": 0.1035,
+      "step": 1039
+    },
+    {
+      "epoch": 0.27615507169410514,
+      "grad_norm": 0.9622843736990357,
+      "learning_rate": 9.79026226757293e-06,
+      "loss": 0.1455,
+      "step": 1040
+    },
+    {
+      "epoch": 0.27642060541688795,
+      "grad_norm": 1.3517743387903762,
+      "learning_rate": 9.789632624720608e-06,
+      "loss": 0.1191,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2766861391396707,
+      "grad_norm": 0.3315039166067069,
+      "learning_rate": 9.789002058483604e-06,
+      "loss": 0.0286,
+      "step": 1042
+    },
+    {
+      "epoch": 0.27695167286245354,
+      "grad_norm": 0.0228123965731073,
+      "learning_rate": 9.788370568983483e-06,
+      "loss": 0.0005,
+      "step": 1043
+    },
+    {
+      "epoch": 0.2772172065852363,
+      "grad_norm": 1.4788814320601762,
+      "learning_rate": 9.787738156341992e-06,
+      "loss": 0.0393,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2774827403080191,
+      "grad_norm": 1.146212270628032,
+      "learning_rate": 9.78710482068105e-06,
+      "loss": 0.0649,
+      "step": 1045
+    },
+    {
+      "epoch": 0.27774827403080193,
+      "grad_norm": 0.5981452393985338,
+      "learning_rate": 9.786470562122758e-06,
+      "loss": 0.0391,
+      "step": 1046
+    },
+    {
+      "epoch": 0.2780138077535847,
+      "grad_norm": 0.10195539381598936,
+      "learning_rate": 9.78583538078939e-06,
+      "loss": 0.0026,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2782793414763675,
+      "grad_norm": 0.3185752679243096,
+      "learning_rate": 9.785199276803408e-06,
+      "loss": 0.0317,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2785448751991503,
+      "grad_norm": 0.3967640673554989,
+      "learning_rate": 9.78456225028744e-06,
+      "loss": 0.0432,
+      "step": 1049
+    },
+    {
+      "epoch": 0.2788104089219331,
+      "grad_norm": 0.7810124894163271,
+      "learning_rate": 9.783924301364297e-06,
+      "loss": 0.0184,
+      "step": 1050
+    },
+    {
+      "epoch": 0.27907594264471586,
+      "grad_norm": 0.08506478322426231,
+      "learning_rate": 9.783285430156972e-06,
+      "loss": 0.0022,
+      "step": 1051
+    },
+    {
+      "epoch": 0.2793414763674987,
+      "grad_norm": 0.7008776117228477,
+      "learning_rate": 9.782645636788628e-06,
+      "loss": 0.0476,
+      "step": 1052
+    },
+    {
+      "epoch": 0.27960701009028144,
+      "grad_norm": 0.3314725833848813,
+      "learning_rate": 9.782004921382612e-06,
+      "loss": 0.0354,
+      "step": 1053
+    },
+    {
+      "epoch": 0.27987254381306426,
+      "grad_norm": 0.03864368830561217,
+      "learning_rate": 9.781363284062446e-06,
+      "loss": 0.0008,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2801380775358471,
+      "grad_norm": 0.5162549754187832,
+      "learning_rate": 9.78072072495183e-06,
+      "loss": 0.0479,
+      "step": 1055
+    },
+    {
+      "epoch": 0.28040361125862984,
+      "grad_norm": 0.5385016039999986,
+      "learning_rate": 9.780077244174642e-06,
+      "loss": 0.0184,
+      "step": 1056
+    },
+    {
+      "epoch": 0.28066914498141265,
+      "grad_norm": 0.39671754689268834,
+      "learning_rate": 9.779432841854937e-06,
+      "loss": 0.0391,
+      "step": 1057
+    },
+    {
+      "epoch": 0.2809346787041954,
+      "grad_norm": 0.3533088509471452,
+      "learning_rate": 9.778787518116948e-06,
+      "loss": 0.0354,
+      "step": 1058
+    },
+    {
+      "epoch": 0.28120021242697824,
+      "grad_norm": 0.8001393279845374,
+      "learning_rate": 9.778141273085085e-06,
+      "loss": 0.0752,
+      "step": 1059
+    },
+    {
+      "epoch": 0.281465746149761,
+      "grad_norm": 0.9390525477977365,
+      "learning_rate": 9.777494106883938e-06,
+      "loss": 0.063,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2817312798725438,
+      "grad_norm": 0.09034297328274829,
+      "learning_rate": 9.776846019638273e-06,
+      "loss": 0.0015,
+      "step": 1061
+    },
+    {
+      "epoch": 0.28199681359532663,
+      "grad_norm": 0.3397054585962954,
+      "learning_rate": 9.776197011473034e-06,
+      "loss": 0.0073,
+      "step": 1062
+    },
+    {
+      "epoch": 0.2822623473181094,
+      "grad_norm": 0.8269003857531309,
+      "learning_rate": 9.775547082513339e-06,
+      "loss": 0.0391,
+      "step": 1063
+    },
+    {
+      "epoch": 0.2825278810408922,
+      "grad_norm": 11.321097522244969,
+      "learning_rate": 9.77489623288449e-06,
+      "loss": 0.3203,
+      "step": 1064
+    },
+    {
+      "epoch": 0.282793414763675,
+      "grad_norm": 0.3961581646443348,
+      "learning_rate": 9.774244462711962e-06,
+      "loss": 0.0354,
+      "step": 1065
+    },
+    {
+      "epoch": 0.2830589484864578,
+      "grad_norm": 0.9979585458391305,
+      "learning_rate": 9.773591772121409e-06,
+      "loss": 0.0815,
+      "step": 1066
+    },
+    {
+      "epoch": 0.28332448220924056,
+      "grad_norm": 0.24274949520810776,
+      "learning_rate": 9.77293816123866e-06,
+      "loss": 0.0051,
+      "step": 1067
+    },
+    {
+      "epoch": 0.2835900159320234,
+      "grad_norm": 5.344035497507486,
+      "learning_rate": 9.772283630189727e-06,
+      "loss": 0.2773,
+      "step": 1068
+    },
+    {
+      "epoch": 0.28385554965480614,
+      "grad_norm": 0.45994201443551747,
+      "learning_rate": 9.771628179100793e-06,
+      "loss": 0.0391,
+      "step": 1069
+    },
+    {
+      "epoch": 0.28412108337758896,
+      "grad_norm": 1.055834204920163,
+      "learning_rate": 9.770971808098221e-06,
+      "loss": 0.0525,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2843866171003718,
+      "grad_norm": 0.5618926736078067,
+      "learning_rate": 9.770314517308554e-06,
+      "loss": 0.0317,
+      "step": 1071
+    },
+    {
+      "epoch": 0.28465215082315454,
+      "grad_norm": 0.24186740542409346,
+      "learning_rate": 9.76965630685851e-06,
+      "loss": 0.0206,
+      "step": 1072
+    },
+    {
+      "epoch": 0.28491768454593736,
+      "grad_norm": 0.9072361060971639,
+      "learning_rate": 9.768997176874982e-06,
+      "loss": 0.1191,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2851832182687201,
+      "grad_norm": 1.1218511212461664,
+      "learning_rate": 9.768337127485044e-06,
+      "loss": 0.0576,
+      "step": 1074
+    },
+    {
+      "epoch": 0.28544875199150294,
+      "grad_norm": 0.2496818724866189,
+      "learning_rate": 9.767676158815944e-06,
+      "loss": 0.0073,
+      "step": 1075
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 1.5823043624366155,
+      "learning_rate": 9.76701427099511e-06,
+      "loss": 0.0688,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2859798194370685,
+      "grad_norm": 2.327890956163168,
+      "learning_rate": 9.766351464150146e-06,
+      "loss": 0.1738,
+      "step": 1077
+    },
+    {
+      "epoch": 0.2862453531598513,
+      "grad_norm": 1.8266555516390197,
+      "learning_rate": 9.765687738408834e-06,
+      "loss": 0.0957,
+      "step": 1078
+    },
+    {
+      "epoch": 0.2865108868826341,
+      "grad_norm": 0.21974129050224078,
+      "learning_rate": 9.765023093899133e-06,
+      "loss": 0.0184,
+      "step": 1079
+    },
+    {
+      "epoch": 0.2867764206054169,
+      "grad_norm": 1.213711313563219,
+      "learning_rate": 9.764357530749178e-06,
+      "loss": 0.0576,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2870419543281997,
+      "grad_norm": 2.075553937576423,
+      "learning_rate": 9.76369104908728e-06,
+      "loss": 0.0957,
+      "step": 1081
+    },
+    {
+      "epoch": 0.2873074880509825,
+      "grad_norm": 0.927208051761271,
+      "learning_rate": 9.76302364904193e-06,
+      "loss": 0.1455,
+      "step": 1082
+    },
+    {
+      "epoch": 0.28757302177376526,
+      "grad_norm": 0.7356955614292838,
+      "learning_rate": 9.762355330741797e-06,
+      "loss": 0.0231,
+      "step": 1083
+    },
+    {
+      "epoch": 0.2878385554965481,
+      "grad_norm": 0.7816526361217149,
+      "learning_rate": 9.761686094315721e-06,
+      "loss": 0.0432,
+      "step": 1084
+    },
+    {
+      "epoch": 0.28810408921933084,
+      "grad_norm": 1.7402446527684008,
+      "learning_rate": 9.761015939892727e-06,
+      "loss": 0.0957,
+      "step": 1085
+    },
+    {
+      "epoch": 0.28836962294211366,
+      "grad_norm": 0.05762993692685398,
+      "learning_rate": 9.760344867602008e-06,
+      "loss": 0.0003,
+      "step": 1086
+    },
+    {
+      "epoch": 0.2886351566648964,
+      "grad_norm": 0.024188141344319784,
+      "learning_rate": 9.759672877572943e-06,
+      "loss": 0.0005,
+      "step": 1087
+    },
+    {
+      "epoch": 0.28890069038767924,
+      "grad_norm": 0.13702148519441235,
+      "learning_rate": 9.758999969935083e-06,
+      "loss": 0.0017,
+      "step": 1088
+    },
+    {
+      "epoch": 0.28916622411046206,
+      "grad_norm": 0.23185659329270913,
+      "learning_rate": 9.758326144818155e-06,
+      "loss": 0.0206,
+      "step": 1089
+    },
+    {
+      "epoch": 0.2894317578332448,
+      "grad_norm": 2.020274635548338,
+      "learning_rate": 9.757651402352067e-06,
+      "loss": 0.1191,
+      "step": 1090
+    },
+    {
+      "epoch": 0.28969729155602764,
+      "grad_norm": 0.23791843579850122,
+      "learning_rate": 9.756975742666899e-06,
+      "loss": 0.0231,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2899628252788104,
+      "grad_norm": 1.5926155722260418,
+      "learning_rate": 9.756299165892911e-06,
+      "loss": 0.0576,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2902283590015932,
+      "grad_norm": 0.29759291019807677,
+      "learning_rate": 9.755621672160538e-06,
+      "loss": 0.0258,
+      "step": 1093
+    },
+    {
+      "epoch": 0.290493892724376,
+      "grad_norm": 0.3257843081640135,
+      "learning_rate": 9.754943261600396e-06,
+      "loss": 0.0206,
+      "step": 1094
+    },
+    {
+      "epoch": 0.2907594264471588,
+      "grad_norm": 0.24127331860074858,
+      "learning_rate": 9.754263934343272e-06,
+      "loss": 0.0206,
+      "step": 1095
+    },
+    {
+      "epoch": 0.29102496016994156,
+      "grad_norm": 0.05613263082076969,
+      "learning_rate": 9.753583690520132e-06,
+      "loss": 0.0007,
+      "step": 1096
+    },
+    {
+      "epoch": 0.2912904938927244,
+      "grad_norm": 0.15571605076216163,
+      "learning_rate": 9.75290253026212e-06,
+      "loss": 0.0131,
+      "step": 1097
+    },
+    {
+      "epoch": 0.2915560276155072,
+      "grad_norm": 0.9548153892509748,
+      "learning_rate": 9.752220453700556e-06,
+      "loss": 0.1455,
+      "step": 1098
+    },
+    {
+      "epoch": 0.29182156133828996,
+      "grad_norm": 0.03386757581041895,
+      "learning_rate": 9.751537460966935e-06,
+      "loss": 0.0008,
+      "step": 1099
+    },
+    {
+      "epoch": 0.2920870950610728,
+      "grad_norm": 1.0372337414427726,
+      "learning_rate": 9.750853552192931e-06,
+      "loss": 0.0889,
+      "step": 1100
+    },
+    {
+      "epoch": 0.29235262878385554,
+      "grad_norm": 0.7281567499028863,
+      "learning_rate": 9.750168727510394e-06,
+      "loss": 0.0286,
+      "step": 1101
+    },
+    {
+      "epoch": 0.29261816250663836,
+      "grad_norm": 0.6380207840024593,
+      "learning_rate": 9.749482987051349e-06,
+      "loss": 0.0391,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2928836962294211,
+      "grad_norm": 0.9729212672078682,
+      "learning_rate": 9.748796330947998e-06,
+      "loss": 0.0889,
+      "step": 1103
+    },
+    {
+      "epoch": 0.29314922995220394,
+      "grad_norm": 1.0018172347011092,
+      "learning_rate": 9.748108759332722e-06,
+      "loss": 0.0815,
+      "step": 1104
+    },
+    {
+      "epoch": 0.2934147636749867,
+      "grad_norm": 0.004923265696314747,
+      "learning_rate": 9.747420272338078e-06,
+      "loss": 0.0001,
+      "step": 1105
+    },
+    {
+      "epoch": 0.2936802973977695,
+      "grad_norm": 0.022021697905026522,
+      "learning_rate": 9.746730870096794e-06,
+      "loss": 0.0005,
+      "step": 1106
+    },
+    {
+      "epoch": 0.29394583112055234,
+      "grad_norm": 1.6710028936610164,
+      "learning_rate": 9.74604055274178e-06,
+      "loss": 0.0815,
+      "step": 1107
+    },
+    {
+      "epoch": 0.2942113648433351,
+      "grad_norm": 1.3634452754585142,
+      "learning_rate": 9.745349320406123e-06,
+      "loss": 0.0432,
+      "step": 1108
+    },
+    {
+      "epoch": 0.2944768985661179,
+      "grad_norm": 0.8146412207177869,
+      "learning_rate": 9.744657173223082e-06,
+      "loss": 0.1113,
+      "step": 1109
+    },
+    {
+      "epoch": 0.2947424322889007,
+      "grad_norm": 1.2206960698761407,
+      "learning_rate": 9.743964111326098e-06,
+      "loss": 0.0688,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2950079660116835,
+      "grad_norm": 0.042821624435204764,
+      "learning_rate": 9.743270134848782e-06,
+      "loss": 0.0004,
+      "step": 1111
+    },
+    {
+      "epoch": 0.29527349973446626,
+      "grad_norm": 0.6103965509662032,
+      "learning_rate": 9.742575243924925e-06,
+      "loss": 0.0206,
+      "step": 1112
+    },
+    {
+      "epoch": 0.2955390334572491,
+      "grad_norm": 0.31275659015640817,
+      "learning_rate": 9.741879438688495e-06,
+      "loss": 0.0317,
+      "step": 1113
+    },
+    {
+      "epoch": 0.29580456718003184,
+      "grad_norm": 0.7391124276417567,
+      "learning_rate": 9.741182719273636e-06,
+      "loss": 0.0815,
+      "step": 1114
+    },
+    {
+      "epoch": 0.29607010090281466,
+      "grad_norm": 1.8465376752624503,
+      "learning_rate": 9.740485085814665e-06,
+      "loss": 0.1113,
+      "step": 1115
+    },
+    {
+      "epoch": 0.2963356346255975,
+      "grad_norm": 0.8365414433585058,
+      "learning_rate": 9.739786538446076e-06,
+      "loss": 0.0576,
+      "step": 1116
+    },
+    {
+      "epoch": 0.29660116834838024,
+      "grad_norm": 0.7185562716195945,
+      "learning_rate": 9.739087077302547e-06,
+      "loss": 0.063,
+      "step": 1117
+    },
+    {
+      "epoch": 0.29686670207116306,
+      "grad_norm": 0.24881622042316806,
+      "learning_rate": 9.738386702518918e-06,
+      "loss": 0.0082,
+      "step": 1118
+    },
+    {
+      "epoch": 0.2971322357939458,
+      "grad_norm": 0.82204521512429,
+      "learning_rate": 9.737685414230218e-06,
+      "loss": 0.0476,
+      "step": 1119
+    },
+    {
+      "epoch": 0.29739776951672864,
+      "grad_norm": 0.4442702937110363,
+      "learning_rate": 9.736983212571646e-06,
+      "loss": 0.0476,
+      "step": 1120
+    },
+    {
+      "epoch": 0.2976633032395114,
+      "grad_norm": 0.017044609666170246,
+      "learning_rate": 9.73628009767858e-06,
+      "loss": 0.0004,
+      "step": 1121
+    },
+    {
+      "epoch": 0.2979288369622942,
+      "grad_norm": 1.2421240055318807,
+      "learning_rate": 9.73557606968657e-06,
+      "loss": 0.1035,
+      "step": 1122
+    },
+    {
+      "epoch": 0.298194370685077,
+      "grad_norm": 0.048110722657898286,
+      "learning_rate": 9.734871128731342e-06,
+      "loss": 0.0006,
+      "step": 1123
+    },
+    {
+      "epoch": 0.2984599044078598,
+      "grad_norm": 4.6600239173016185,
+      "learning_rate": 9.734165274948806e-06,
+      "loss": 0.5469,
+      "step": 1124
+    },
+    {
+      "epoch": 0.2987254381306426,
+      "grad_norm": 0.7244351120756174,
+      "learning_rate": 9.733458508475038e-06,
+      "loss": 0.0576,
+      "step": 1125
+    },
+    {
+      "epoch": 0.2989909718534254,
+      "grad_norm": 0.004283320379219228,
+      "learning_rate": 9.732750829446295e-06,
+      "loss": 0.0001,
+      "step": 1126
+    },
+    {
+      "epoch": 0.2992565055762082,
+      "grad_norm": 0.7629355667225856,
+      "learning_rate": 9.73204223799901e-06,
+      "loss": 0.1191,
+      "step": 1127
+    },
+    {
+      "epoch": 0.29952203929899096,
+      "grad_norm": 0.6208152245082953,
+      "learning_rate": 9.731332734269791e-06,
+      "loss": 0.0479,
+      "step": 1128
+    },
+    {
+      "epoch": 0.2997875730217738,
+      "grad_norm": 0.6038756411851064,
+      "learning_rate": 9.730622318395422e-06,
+      "loss": 0.0576,
+      "step": 1129
+    },
+    {
+      "epoch": 0.30005310674455654,
+      "grad_norm": 0.6098024815769895,
+      "learning_rate": 9.729910990512862e-06,
+      "loss": 0.0752,
+      "step": 1130
+    },
+    {
+      "epoch": 0.30031864046733936,
+      "grad_norm": 0.6672028744949045,
+      "learning_rate": 9.729198750759248e-06,
+      "loss": 0.0688,
+      "step": 1131
+    },
+    {
+      "epoch": 0.3005841741901221,
+      "grad_norm": 0.42173926069039175,
+      "learning_rate": 9.72848559927189e-06,
+      "loss": 0.0432,
+      "step": 1132
+    },
+    {
+      "epoch": 0.30084970791290494,
+      "grad_norm": 0.4224982111516302,
+      "learning_rate": 9.727771536188275e-06,
+      "loss": 0.0476,
+      "step": 1133
+    },
+    {
+      "epoch": 0.30111524163568776,
+      "grad_norm": 0.818705810028351,
+      "learning_rate": 9.727056561646067e-06,
+      "loss": 0.0815,
+      "step": 1134
+    },
+    {
+      "epoch": 0.3013807753584705,
+      "grad_norm": 0.4673281241576781,
+      "learning_rate": 9.726340675783105e-06,
+      "loss": 0.0391,
+      "step": 1135
+    },
+    {
+      "epoch": 0.30164630908125334,
+      "grad_norm": 0.6576605686901854,
+      "learning_rate": 9.725623878737402e-06,
+      "loss": 0.0752,
+      "step": 1136
+    },
+    {
+      "epoch": 0.3019118428040361,
+      "grad_norm": 0.5547502104537158,
+      "learning_rate": 9.724906170647151e-06,
+      "loss": 0.0479,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3021773765268189,
+      "grad_norm": 0.35579660396826157,
+      "learning_rate": 9.724187551650712e-06,
+      "loss": 0.0132,
+      "step": 1138
+    },
+    {
+      "epoch": 0.3024429102496017,
+      "grad_norm": 0.4736433372724878,
+      "learning_rate": 9.723468021886632e-06,
+      "loss": 0.0525,
+      "step": 1139
+    },
+    {
+      "epoch": 0.3027084439723845,
+      "grad_norm": 0.02789314147339348,
+      "learning_rate": 9.722747581493625e-06,
+      "loss": 0.0007,
+      "step": 1140
+    },
+    {
+      "epoch": 0.30297397769516726,
+      "grad_norm": 0.21925197694495466,
+      "learning_rate": 9.722026230610584e-06,
+      "loss": 0.0206,
+      "step": 1141
+    },
+    {
+      "epoch": 0.3032395114179501,
+      "grad_norm": 0.010099703111864153,
+      "learning_rate": 9.721303969376576e-06,
+      "loss": 0.0003,
+      "step": 1142
+    },
+    {
+      "epoch": 0.3035050451407329,
+      "grad_norm": 0.31933908172577696,
+      "learning_rate": 9.720580797930845e-06,
+      "loss": 0.0354,
+      "step": 1143
+    },
+    {
+      "epoch": 0.30377057886351566,
+      "grad_norm": 1.1192764070316166,
+      "learning_rate": 9.719856716412813e-06,
+      "loss": 0.0889,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3040361125862985,
+      "grad_norm": 0.6684300524852204,
+      "learning_rate": 9.71913172496207e-06,
+      "loss": 0.1113,
+      "step": 1145
+    },
+    {
+      "epoch": 0.30430164630908124,
+      "grad_norm": 0.29373443475884653,
+      "learning_rate": 9.718405823718386e-06,
+      "loss": 0.0317,
+      "step": 1146
+    },
+    {
+      "epoch": 0.30456718003186406,
+      "grad_norm": 0.3658446950988331,
+      "learning_rate": 9.71767901282171e-06,
+      "loss": 0.0432,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3048327137546468,
+      "grad_norm": 0.567931719494275,
+      "learning_rate": 9.716951292412158e-06,
+      "loss": 0.0432,
+      "step": 1148
+    },
+    {
+      "epoch": 0.30509824747742964,
+      "grad_norm": 0.06610736970007457,
+      "learning_rate": 9.71622266263003e-06,
+      "loss": 0.0016,
+      "step": 1149
+    },
+    {
+      "epoch": 0.3053637812002124,
+      "grad_norm": 0.24553686436218464,
+      "learning_rate": 9.715493123615793e-06,
+      "loss": 0.0258,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3056293149229952,
+      "grad_norm": 0.3986171599233714,
+      "learning_rate": 9.714762675510098e-06,
+      "loss": 0.0432,
+      "step": 1151
+    },
+    {
+      "epoch": 0.30589484864577804,
+      "grad_norm": 0.03893798946143774,
+      "learning_rate": 9.714031318453763e-06,
+      "loss": 0.001,
+      "step": 1152
+    },
+    {
+      "epoch": 0.3061603823685608,
+      "grad_norm": 0.30020768346470134,
+      "learning_rate": 9.713299052587788e-06,
+      "loss": 0.0066,
+      "step": 1153
+    },
+    {
+      "epoch": 0.3064259160913436,
+      "grad_norm": 0.7987344718538759,
+      "learning_rate": 9.712565878053344e-06,
+      "loss": 0.1191,
+      "step": 1154
+    },
+    {
+      "epoch": 0.3066914498141264,
+      "grad_norm": 0.2734737664568203,
+      "learning_rate": 9.711831794991777e-06,
+      "loss": 0.0286,
+      "step": 1155
+    },
+    {
+      "epoch": 0.3069569835369092,
+      "grad_norm": 0.9407886341613316,
+      "learning_rate": 9.711096803544612e-06,
+      "loss": 0.0206,
+      "step": 1156
+    },
+    {
+      "epoch": 0.30722251725969196,
+      "grad_norm": 0.2662920174573857,
+      "learning_rate": 9.710360903853545e-06,
+      "loss": 0.0286,
+      "step": 1157
+    },
+    {
+      "epoch": 0.3074880509824748,
+      "grad_norm": 0.40813353628480753,
+      "learning_rate": 9.70962409606045e-06,
+      "loss": 0.0432,
+      "step": 1158
+    },
+    {
+      "epoch": 0.30775358470525754,
+      "grad_norm": 0.33115218272803004,
+      "learning_rate": 9.708886380307373e-06,
+      "loss": 0.0286,
+      "step": 1159
+    },
+    {
+      "epoch": 0.30801911842804036,
+      "grad_norm": 2.752464726652913,
+      "learning_rate": 9.70814775673654e-06,
+      "loss": 0.1113,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3082846521508232,
+      "grad_norm": 0.21100244178718114,
+      "learning_rate": 9.707408225490343e-06,
+      "loss": 0.0206,
+      "step": 1161
+    },
+    {
+      "epoch": 0.30855018587360594,
+      "grad_norm": 0.273597747914562,
+      "learning_rate": 9.706667786711362e-06,
+      "loss": 0.0286,
+      "step": 1162
+    },
+    {
+      "epoch": 0.30881571959638876,
+      "grad_norm": 0.5455974875060108,
+      "learning_rate": 9.705926440542342e-06,
+      "loss": 0.0476,
+      "step": 1163
+    },
+    {
+      "epoch": 0.3090812533191715,
+      "grad_norm": 0.04996096845556894,
+      "learning_rate": 9.705184187126205e-06,
+      "loss": 0.001,
+      "step": 1164
+    },
+    {
+      "epoch": 0.30934678704195434,
+      "grad_norm": 0.7661826955072533,
+      "learning_rate": 9.704441026606047e-06,
+      "loss": 0.0957,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3096123207647371,
+      "grad_norm": 0.22898596466557175,
+      "learning_rate": 9.703696959125144e-06,
+      "loss": 0.0038,
+      "step": 1166
+    },
+    {
+      "epoch": 0.3098778544875199,
+      "grad_norm": 1.0310557558318094,
+      "learning_rate": 9.702951984826943e-06,
+      "loss": 0.1113,
+      "step": 1167
+    },
+    {
+      "epoch": 0.3101433882103027,
+      "grad_norm": 0.9506415865468816,
+      "learning_rate": 9.702206103855065e-06,
+      "loss": 0.1367,
+      "step": 1168
+    },
+    {
+      "epoch": 0.3104089219330855,
+      "grad_norm": 0.9519356526243661,
+      "learning_rate": 9.701459316353309e-06,
+      "loss": 0.0752,
+      "step": 1169
+    },
+    {
+      "epoch": 0.3106744556558683,
+      "grad_norm": 0.4403895621363803,
+      "learning_rate": 9.700711622465645e-06,
+      "loss": 0.0391,
+      "step": 1170
+    },
+    {
+      "epoch": 0.3109399893786511,
+      "grad_norm": 0.8734691695278628,
+      "learning_rate": 9.699963022336219e-06,
+      "loss": 0.0815,
+      "step": 1171
+    },
+    {
+      "epoch": 0.3112055231014339,
+      "grad_norm": 0.03692544587443828,
+      "learning_rate": 9.69921351610935e-06,
+      "loss": 0.001,
+      "step": 1172
+    },
+    {
+      "epoch": 0.31147105682421666,
+      "grad_norm": 0.38902223564103433,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.0354,
+      "step": 1173
+    },
+    {
+      "epoch": 0.3117365905469995,
+      "grad_norm": 0.4322873123614092,
+      "learning_rate": 9.697711785941459e-06,
+      "loss": 0.0391,
+      "step": 1174
+    },
+    {
+      "epoch": 0.31200212426978224,
+      "grad_norm": 1.04700194984681,
+      "learning_rate": 9.696959562289949e-06,
+      "loss": 0.1035,
+      "step": 1175
+    },
+    {
+      "epoch": 0.31226765799256506,
+      "grad_norm": 0.6455068902710267,
+      "learning_rate": 9.69620643312003e-06,
+      "loss": 0.0525,
+      "step": 1176
+    },
+    {
+      "epoch": 0.3125331917153478,
+      "grad_norm": 0.19469026077885143,
+      "learning_rate": 9.695452398576897e-06,
+      "loss": 0.0206,
+      "step": 1177
+    },
+    {
+      "epoch": 0.31279872543813064,
+      "grad_norm": 0.552300258388761,
+      "learning_rate": 9.69469745880592e-06,
+      "loss": 0.0576,
+      "step": 1178
+    },
+    {
+      "epoch": 0.31306425916091346,
+      "grad_norm": 0.15378094703192938,
+      "learning_rate": 9.693941613952642e-06,
+      "loss": 0.0146,
+      "step": 1179
+    },
+    {
+      "epoch": 0.3133297928836962,
+      "grad_norm": 0.7149303778962087,
+      "learning_rate": 9.693184864162781e-06,
+      "loss": 0.0525,
+      "step": 1180
+    },
+    {
+      "epoch": 0.31359532660647904,
+      "grad_norm": 12.606742773671819,
+      "learning_rate": 9.692427209582227e-06,
+      "loss": 0.3516,
+      "step": 1181
+    },
+    {
+      "epoch": 0.3138608603292618,
+      "grad_norm": 0.02382454479686282,
+      "learning_rate": 9.691668650357052e-06,
+      "loss": 0.0005,
+      "step": 1182
+    },
+    {
+      "epoch": 0.3141263940520446,
+      "grad_norm": 0.6509559366250702,
+      "learning_rate": 9.690909186633493e-06,
+      "loss": 0.063,
+      "step": 1183
+    },
+    {
+      "epoch": 0.3143919277748274,
+      "grad_norm": 0.22407883750706376,
+      "learning_rate": 9.690148818557966e-06,
+      "loss": 0.0231,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3146574614976102,
+      "grad_norm": 1.194851766940795,
+      "learning_rate": 9.689387546277062e-06,
+      "loss": 0.0752,
+      "step": 1185
+    },
+    {
+      "epoch": 0.31492299522039296,
+      "grad_norm": 1.5889517299989575,
+      "learning_rate": 9.688625369937548e-06,
+      "loss": 0.1113,
+      "step": 1186
+    },
+    {
+      "epoch": 0.3151885289431758,
+      "grad_norm": 0.7528605787401368,
+      "learning_rate": 9.687862289686356e-06,
+      "loss": 0.1035,
+      "step": 1187
+    },
+    {
+      "epoch": 0.3154540626659586,
+      "grad_norm": 0.10046386842043101,
+      "learning_rate": 9.687098305670606e-06,
+      "loss": 0.0025,
+      "step": 1188
+    },
+    {
+      "epoch": 0.31571959638874136,
+      "grad_norm": 0.744431616154801,
+      "learning_rate": 9.68633341803758e-06,
+      "loss": 0.1113,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3159851301115242,
+      "grad_norm": 0.3171470340086223,
+      "learning_rate": 9.685567626934742e-06,
+      "loss": 0.0354,
+      "step": 1190
+    },
+    {
+      "epoch": 0.31625066383430694,
+      "grad_norm": 0.24491006741762908,
+      "learning_rate": 9.684800932509726e-06,
+      "loss": 0.0258,
+      "step": 1191
+    },
+    {
+      "epoch": 0.31651619755708976,
+      "grad_norm": 0.057338171719011885,
+      "learning_rate": 9.684033334910342e-06,
+      "loss": 0.0014,
+      "step": 1192
+    },
+    {
+      "epoch": 0.3167817312798725,
+      "grad_norm": 0.2935840183769047,
+      "learning_rate": 9.683264834284575e-06,
+      "loss": 0.0258,
+      "step": 1193
+    },
+    {
+      "epoch": 0.31704726500265534,
+      "grad_norm": 0.7590487718984419,
+      "learning_rate": 9.682495430780581e-06,
+      "loss": 0.1455,
+      "step": 1194
+    },
+    {
+      "epoch": 0.3173127987254381,
+      "grad_norm": 1.1684076403623338,
+      "learning_rate": 9.681725124546695e-06,
+      "loss": 0.0693,
+      "step": 1195
+    },
+    {
+      "epoch": 0.3175783324482209,
+      "grad_norm": 1.6438208494368356,
+      "learning_rate": 9.680953915731418e-06,
+      "loss": 0.082,
+      "step": 1196
+    },
+    {
+      "epoch": 0.31784386617100374,
+      "grad_norm": 0.8695124690281735,
+      "learning_rate": 9.680181804483435e-06,
+      "loss": 0.0479,
+      "step": 1197
+    },
+    {
+      "epoch": 0.3181093998937865,
+      "grad_norm": 0.20326281659079895,
+      "learning_rate": 9.679408790951597e-06,
+      "loss": 0.0231,
+      "step": 1198
+    },
+    {
+      "epoch": 0.3183749336165693,
+      "grad_norm": 0.6573685494777957,
+      "learning_rate": 9.678634875284932e-06,
+      "loss": 0.0187,
+      "step": 1199
+    },
+    {
+      "epoch": 0.3186404673393521,
+      "grad_norm": 0.7687488235704806,
+      "learning_rate": 9.677860057632642e-06,
+      "loss": 0.1455,
+      "step": 1200
+    },
+    {
+      "epoch": 0.3189060010621349,
+      "grad_norm": 0.12038614362749626,
+      "learning_rate": 9.677084338144105e-06,
+      "loss": 0.0117,
+      "step": 1201
+    },
+    {
+      "epoch": 0.31917153478491767,
+      "grad_norm": 0.22290803237319895,
+      "learning_rate": 9.676307716968866e-06,
+      "loss": 0.0258,
+      "step": 1202
+    },
+    {
+      "epoch": 0.3194370685077005,
+      "grad_norm": 0.08514537747421477,
+      "learning_rate": 9.675530194256652e-06,
+      "loss": 0.0018,
+      "step": 1203
+    },
+    {
+      "epoch": 0.31970260223048325,
+      "grad_norm": 0.5579141610152237,
+      "learning_rate": 9.674751770157362e-06,
+      "loss": 0.0233,
+      "step": 1204
+    },
+    {
+      "epoch": 0.31996813595326606,
+      "grad_norm": 0.312483949561754,
+      "learning_rate": 9.673972444821063e-06,
+      "loss": 0.0354,
+      "step": 1205
+    },
+    {
+      "epoch": 0.3202336696760489,
+      "grad_norm": 0.22701164555191825,
+      "learning_rate": 9.673192218398e-06,
+      "loss": 0.0258,
+      "step": 1206
+    },
+    {
+      "epoch": 0.32049920339883164,
+      "grad_norm": 0.1487192071650038,
+      "learning_rate": 9.672411091038592e-06,
+      "loss": 0.0146,
+      "step": 1207
+    },
+    {
+      "epoch": 0.32076473712161446,
+      "grad_norm": 1.3761300654988238,
+      "learning_rate": 9.671629062893433e-06,
+      "loss": 0.1279,
+      "step": 1208
+    },
+    {
+      "epoch": 0.3210302708443972,
+      "grad_norm": 0.1704844264079489,
+      "learning_rate": 9.670846134113286e-06,
+      "loss": 0.0184,
+      "step": 1209
+    },
+    {
+      "epoch": 0.32129580456718004,
+      "grad_norm": 0.19819201079023002,
+      "learning_rate": 9.670062304849093e-06,
+      "loss": 0.0047,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3215613382899628,
+      "grad_norm": 0.1838454976972731,
+      "learning_rate": 9.669277575251965e-06,
+      "loss": 0.0052,
+      "step": 1211
+    },
+    {
+      "epoch": 0.3218268720127456,
+      "grad_norm": 0.6558129007222472,
+      "learning_rate": 9.668491945473191e-06,
+      "loss": 0.0148,
+      "step": 1212
+    },
+    {
+      "epoch": 0.3220924057355284,
+      "grad_norm": 0.34799526026312055,
+      "learning_rate": 9.667705415664228e-06,
+      "loss": 0.0393,
+      "step": 1213
+    },
+    {
+      "epoch": 0.3223579394583112,
+      "grad_norm": 0.5171512089272212,
+      "learning_rate": 9.666917985976712e-06,
+      "loss": 0.0148,
+      "step": 1214
+    },
+    {
+      "epoch": 0.322623473181094,
+      "grad_norm": 0.2062886511175874,
+      "learning_rate": 9.66612965656245e-06,
+      "loss": 0.0231,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3228890069038768,
+      "grad_norm": 0.7738586973681455,
+      "learning_rate": 9.665340427573422e-06,
+      "loss": 0.1641,
+      "step": 1216
+    },
+    {
+      "epoch": 0.3231545406266596,
+      "grad_norm": 0.03671678604671888,
+      "learning_rate": 9.664550299161782e-06,
+      "loss": 0.0008,
+      "step": 1217
+    },
+    {
+      "epoch": 0.32342007434944237,
+      "grad_norm": 0.19085387586302657,
+      "learning_rate": 9.663759271479858e-06,
+      "loss": 0.0206,
+      "step": 1218
+    },
+    {
+      "epoch": 0.3236856080722252,
+      "grad_norm": 3.034553850378646,
+      "learning_rate": 9.66296734468015e-06,
+      "loss": 0.1553,
+      "step": 1219
+    },
+    {
+      "epoch": 0.32395114179500795,
+      "grad_norm": 0.026411465752670175,
+      "learning_rate": 9.662174518915332e-06,
+      "loss": 0.0006,
+      "step": 1220
+    },
+    {
+      "epoch": 0.32421667551779076,
+      "grad_norm": 0.08856791563455327,
+      "learning_rate": 9.661380794338252e-06,
+      "loss": 0.002,
+      "step": 1221
+    },
+    {
+      "epoch": 0.3244822092405735,
+      "grad_norm": 0.04449604335631291,
+      "learning_rate": 9.660586171101931e-06,
+      "loss": 0.0008,
+      "step": 1222
+    },
+    {
+      "epoch": 0.32474774296335635,
+      "grad_norm": 0.8025875942279423,
+      "learning_rate": 9.659790649359563e-06,
+      "loss": 0.1738,
+      "step": 1223
+    },
+    {
+      "epoch": 0.32501327668613916,
+      "grad_norm": 0.13222478587189693,
+      "learning_rate": 9.658994229264514e-06,
+      "loss": 0.0131,
+      "step": 1224
+    },
+    {
+      "epoch": 0.3252788104089219,
+      "grad_norm": 0.6260229647329109,
+      "learning_rate": 9.658196910970327e-06,
+      "loss": 0.0889,
+      "step": 1225
+    },
+    {
+      "epoch": 0.32554434413170474,
+      "grad_norm": 0.7658967348933747,
+      "learning_rate": 9.657398694630713e-06,
+      "loss": 0.1553,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3258098778544875,
+      "grad_norm": 0.48947788594768415,
+      "learning_rate": 9.65659958039956e-06,
+      "loss": 0.0288,
+      "step": 1227
+    },
+    {
+      "epoch": 0.3260754115772703,
+      "grad_norm": 0.8585792462640025,
+      "learning_rate": 9.655799568430926e-06,
+      "loss": 0.0688,
+      "step": 1228
+    },
+    {
+      "epoch": 0.3263409453000531,
+      "grad_norm": 0.6527231816101399,
+      "learning_rate": 9.654998658879046e-06,
+      "loss": 0.1113,
+      "step": 1229
+    },
+    {
+      "epoch": 0.3266064790228359,
+      "grad_norm": 0.6582417207082769,
+      "learning_rate": 9.654196851898325e-06,
+      "loss": 0.1191,
+      "step": 1230
+    },
+    {
+      "epoch": 0.32687201274561867,
+      "grad_norm": 0.6645575253832666,
+      "learning_rate": 9.653394147643343e-06,
+      "loss": 0.1113,
+      "step": 1231
+    },
+    {
+      "epoch": 0.3271375464684015,
+      "grad_norm": 0.27026297493457185,
+      "learning_rate": 9.65259054626885e-06,
+      "loss": 0.0317,
+      "step": 1232
+    },
+    {
+      "epoch": 0.3274030801911843,
+      "grad_norm": 0.6383288218728779,
+      "learning_rate": 9.651786047929772e-06,
+      "loss": 0.1035,
+      "step": 1233
+    },
+    {
+      "epoch": 0.32766861391396707,
+      "grad_norm": 0.011355216588088562,
+      "learning_rate": 9.650980652781208e-06,
+      "loss": 0.0003,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3279341476367499,
+      "grad_norm": 0.5036691216178738,
+      "learning_rate": 9.650174360978427e-06,
+      "loss": 0.0525,
+      "step": 1235
+    },
+    {
+      "epoch": 0.32819968135953265,
+      "grad_norm": 0.48956110736136377,
+      "learning_rate": 9.649367172676873e-06,
+      "loss": 0.0317,
+      "step": 1236
+    },
+    {
+      "epoch": 0.32846521508231546,
+      "grad_norm": 0.6193502355457858,
+      "learning_rate": 9.648559088032162e-06,
+      "loss": 0.0957,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3287307488050982,
+      "grad_norm": 0.7256539061385159,
+      "learning_rate": 9.647750107200082e-06,
+      "loss": 0.0889,
+      "step": 1238
+    },
+    {
+      "epoch": 0.32899628252788105,
+      "grad_norm": 0.5224109219073716,
+      "learning_rate": 9.646940230336599e-06,
+      "loss": 0.0045,
+      "step": 1239
+    },
+    {
+      "epoch": 0.3292618162506638,
+      "grad_norm": 0.2037818062231921,
+      "learning_rate": 9.646129457597844e-06,
+      "loss": 0.0023,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3295273499734466,
+      "grad_norm": 0.4607652592533831,
+      "learning_rate": 9.645317789140125e-06,
+      "loss": 0.0579,
+      "step": 1241
+    },
+    {
+      "epoch": 0.32979288369622944,
+      "grad_norm": 0.9063414384859513,
+      "learning_rate": 9.644505225119922e-06,
+      "loss": 0.1113,
+      "step": 1242
+    },
+    {
+      "epoch": 0.3300584174190122,
+      "grad_norm": 0.2770486401552903,
+      "learning_rate": 9.64369176569389e-06,
+      "loss": 0.0187,
+      "step": 1243
+    },
+    {
+      "epoch": 0.330323951141795,
+      "grad_norm": 0.4514415661962752,
+      "learning_rate": 9.642877411018852e-06,
+      "loss": 0.0525,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3305894848645778,
+      "grad_norm": 0.4388178911161381,
+      "learning_rate": 9.642062161251807e-06,
+      "loss": 0.0525,
+      "step": 1245
+    },
+    {
+      "epoch": 0.3308550185873606,
+      "grad_norm": 0.5887566574525436,
+      "learning_rate": 9.641246016549925e-06,
+      "loss": 0.0579,
+      "step": 1246
+    },
+    {
+      "epoch": 0.33112055231014337,
+      "grad_norm": 2.5823916037440813,
+      "learning_rate": 9.64042897707055e-06,
+      "loss": 0.2344,
+      "step": 1247
+    },
+    {
+      "epoch": 0.3313860860329262,
+      "grad_norm": 0.3156225772793237,
+      "learning_rate": 9.639611042971198e-06,
+      "loss": 0.0258,
+      "step": 1248
+    },
+    {
+      "epoch": 0.33165161975570895,
+      "grad_norm": 0.4480395609995155,
+      "learning_rate": 9.638792214409556e-06,
+      "loss": 0.0479,
+      "step": 1249
+    },
+    {
+      "epoch": 0.33191715347849177,
+      "grad_norm": 0.7209237690266596,
+      "learning_rate": 9.637972491543482e-06,
+      "loss": 0.0957,
+      "step": 1250
+    },
+    {
+      "epoch": 0.3321826872012746,
+      "grad_norm": 0.790962238735122,
+      "learning_rate": 9.637151874531014e-06,
+      "loss": 0.1035,
+      "step": 1251
+    },
+    {
+      "epoch": 0.33244822092405735,
+      "grad_norm": 0.26847764212042974,
+      "learning_rate": 9.636330363530352e-06,
+      "loss": 0.0206,
+      "step": 1252
+    },
+    {
+      "epoch": 0.33271375464684017,
+      "grad_norm": 0.13095952869121744,
+      "learning_rate": 9.63550795869988e-06,
+      "loss": 0.0042,
+      "step": 1253
+    },
+    {
+      "epoch": 0.33297928836962293,
+      "grad_norm": 0.4485281612188027,
+      "learning_rate": 9.63468466019814e-06,
+      "loss": 0.0479,
+      "step": 1254
+    },
+    {
+      "epoch": 0.33324482209240575,
+      "grad_norm": 6.678236601224771,
+      "learning_rate": 9.633860468183862e-06,
+      "loss": 0.6133,
+      "step": 1255
+    },
+    {
+      "epoch": 0.3335103558151885,
+      "grad_norm": 0.03640018770094486,
+      "learning_rate": 9.633035382815936e-06,
+      "loss": 0.0009,
+      "step": 1256
+    },
+    {
+      "epoch": 0.3337758895379713,
+      "grad_norm": 0.20995200600370745,
+      "learning_rate": 9.63220940425343e-06,
+      "loss": 0.0147,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3340414232607541,
+      "grad_norm": 0.7744165649910679,
+      "learning_rate": 9.631382532655583e-06,
+      "loss": 0.1035,
+      "step": 1258
+    },
+    {
+      "epoch": 0.3343069569835369,
+      "grad_norm": 0.8521274871394714,
+      "learning_rate": 9.630554768181806e-06,
+      "loss": 0.1279,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3345724907063197,
+      "grad_norm": 0.013869617394598415,
+      "learning_rate": 9.62972611099168e-06,
+      "loss": 0.0004,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3348380244291025,
+      "grad_norm": 0.7984701653609472,
+      "learning_rate": 9.628896561244965e-06,
+      "loss": 0.1201,
+      "step": 1261
+    },
+    {
+      "epoch": 0.3351035581518853,
+      "grad_norm": 0.18442932803572457,
+      "learning_rate": 9.628066119101583e-06,
+      "loss": 0.0093,
+      "step": 1262
+    },
+    {
+      "epoch": 0.33536909187466807,
+      "grad_norm": 0.7728685034480455,
+      "learning_rate": 9.627234784721637e-06,
+      "loss": 0.1113,
+      "step": 1263
+    },
+    {
+      "epoch": 0.3356346255974509,
+      "grad_norm": 0.28335521981673617,
+      "learning_rate": 9.6264025582654e-06,
+      "loss": 0.0233,
+      "step": 1264
+    },
+    {
+      "epoch": 0.33590015932023365,
+      "grad_norm": 0.399075752307739,
+      "learning_rate": 9.62556943989331e-06,
+      "loss": 0.0479,
+      "step": 1265
+    },
+    {
+      "epoch": 0.33616569304301647,
+      "grad_norm": 0.12658474094636998,
+      "learning_rate": 9.624735429765988e-06,
+      "loss": 0.0041,
+      "step": 1266
+    },
+    {
+      "epoch": 0.33643122676579923,
+      "grad_norm": 0.5280828189652471,
+      "learning_rate": 9.623900528044216e-06,
+      "loss": 0.0752,
+      "step": 1267
+    },
+    {
+      "epoch": 0.33669676048858205,
+      "grad_norm": 0.24243489448797254,
+      "learning_rate": 9.623064734888959e-06,
+      "loss": 0.0066,
+      "step": 1268
+    },
+    {
+      "epoch": 0.33696229421136487,
+      "grad_norm": 0.4700632898354242,
+      "learning_rate": 9.622228050461345e-06,
+      "loss": 0.0693,
+      "step": 1269
+    },
+    {
+      "epoch": 0.33722782793414763,
+      "grad_norm": 0.41440456235100415,
+      "learning_rate": 9.621390474922675e-06,
+      "loss": 0.0579,
+      "step": 1270
+    },
+    {
+      "epoch": 0.33749336165693045,
+      "grad_norm": 0.7223449992115132,
+      "learning_rate": 9.620552008434426e-06,
+      "loss": 0.0752,
+      "step": 1271
+    },
+    {
+      "epoch": 0.3377588953797132,
+      "grad_norm": 0.4596889344479549,
+      "learning_rate": 9.619712651158243e-06,
+      "loss": 0.0693,
+      "step": 1272
+    },
+    {
+      "epoch": 0.338024429102496,
+      "grad_norm": 0.4441012257660843,
+      "learning_rate": 9.618872403255947e-06,
+      "loss": 0.063,
+      "step": 1273
+    },
+    {
+      "epoch": 0.3382899628252788,
+      "grad_norm": 0.1553935211134242,
+      "learning_rate": 9.618031264889529e-06,
+      "loss": 0.0066,
+      "step": 1274
+    },
+    {
+      "epoch": 0.3385554965480616,
+      "grad_norm": 0.42387440934012344,
+      "learning_rate": 9.617189236221143e-06,
+      "loss": 0.0579,
+      "step": 1275
+    },
+    {
+      "epoch": 0.33882103027084437,
+      "grad_norm": 0.47243312888409494,
+      "learning_rate": 9.61634631741313e-06,
+      "loss": 0.0752,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3390865639936272,
+      "grad_norm": 0.06774040615620845,
+      "learning_rate": 9.61550250862799e-06,
+      "loss": 0.0018,
+      "step": 1277
+    },
+    {
+      "epoch": 0.33935209771641,
+      "grad_norm": 0.38367232223776365,
+      "learning_rate": 9.614657810028402e-06,
+      "loss": 0.0525,
+      "step": 1278
+    },
+    {
+      "epoch": 0.33961763143919277,
+      "grad_norm": 0.3392200068088174,
+      "learning_rate": 9.613812221777212e-06,
+      "loss": 0.0131,
+      "step": 1279
+    },
+    {
+      "epoch": 0.3398831651619756,
+      "grad_norm": 0.42002324638804434,
+      "learning_rate": 9.612965744037441e-06,
+      "loss": 0.0479,
+      "step": 1280
+    },
+    {
+      "epoch": 0.34014869888475835,
+      "grad_norm": 0.5274772223331019,
+      "learning_rate": 9.61211837697228e-06,
+      "loss": 0.0693,
+      "step": 1281
+    },
+    {
+      "epoch": 0.34041423260754117,
+      "grad_norm": 0.35860555791803894,
+      "learning_rate": 9.61127012074509e-06,
+      "loss": 0.0479,
+      "step": 1282
+    },
+    {
+      "epoch": 0.34067976633032393,
+      "grad_norm": 0.5038903692242176,
+      "learning_rate": 9.610420975519407e-06,
+      "loss": 0.0752,
+      "step": 1283
+    },
+    {
+      "epoch": 0.34094530005310675,
+      "grad_norm": 0.4033059759112738,
+      "learning_rate": 9.609570941458935e-06,
+      "loss": 0.0579,
+      "step": 1284
+    },
+    {
+      "epoch": 0.3412108337758895,
+      "grad_norm": 0.5660112789357162,
+      "learning_rate": 9.60872001872755e-06,
+      "loss": 0.0752,
+      "step": 1285
+    },
+    {
+      "epoch": 0.34147636749867233,
+      "grad_norm": 0.3263429089982916,
+      "learning_rate": 9.607868207489297e-06,
+      "loss": 0.0393,
+      "step": 1286
+    },
+    {
+      "epoch": 0.34174190122145515,
+      "grad_norm": 0.37153713665523197,
+      "learning_rate": 9.607015507908401e-06,
+      "loss": 0.0479,
+      "step": 1287
+    },
+    {
+      "epoch": 0.3420074349442379,
+      "grad_norm": 0.4755797521549427,
+      "learning_rate": 9.606161920149251e-06,
+      "loss": 0.0579,
+      "step": 1288
+    },
+    {
+      "epoch": 0.3422729686670207,
+      "grad_norm": 0.5767413183549783,
+      "learning_rate": 9.605307444376407e-06,
+      "loss": 0.063,
+      "step": 1289
+    },
+    {
+      "epoch": 0.3425385023898035,
+      "grad_norm": 0.06779770472600782,
+      "learning_rate": 9.604452080754601e-06,
+      "loss": 0.0017,
+      "step": 1290
+    },
+    {
+      "epoch": 0.3428040361125863,
+      "grad_norm": 0.5248146878342517,
+      "learning_rate": 9.603595829448737e-06,
+      "loss": 0.0815,
+      "step": 1291
+    },
+    {
+      "epoch": 0.34306956983536907,
+      "grad_norm": 3.5626821708062617,
+      "learning_rate": 9.602738690623895e-06,
+      "loss": 0.377,
+      "step": 1292
+    },
+    {
+      "epoch": 0.3433351035581519,
+      "grad_norm": 0.5707292470276522,
+      "learning_rate": 9.601880664445313e-06,
+      "loss": 0.0957,
+      "step": 1293
+    },
+    {
+      "epoch": 0.34360063728093465,
+      "grad_norm": 0.41750308884117965,
+      "learning_rate": 9.601021751078416e-06,
+      "loss": 0.0393,
+      "step": 1294
+    },
+    {
+      "epoch": 0.34386617100371747,
+      "grad_norm": 0.5159271503322392,
+      "learning_rate": 9.60016195068879e-06,
+      "loss": 0.063,
+      "step": 1295
+    },
+    {
+      "epoch": 0.3441317047265003,
+      "grad_norm": 0.5216179998375932,
+      "learning_rate": 9.599301263442194e-06,
+      "loss": 0.0688,
+      "step": 1296
+    },
+    {
+      "epoch": 0.34439723844928305,
+      "grad_norm": 0.30090609879974806,
+      "learning_rate": 9.598439689504556e-06,
+      "loss": 0.0391,
+      "step": 1297
+    },
+    {
+      "epoch": 0.34466277217206587,
+      "grad_norm": 0.3707745206401245,
+      "learning_rate": 9.597577229041981e-06,
+      "loss": 0.0525,
+      "step": 1298
+    },
+    {
+      "epoch": 0.34492830589484863,
+      "grad_norm": 2.103947714882032,
+      "learning_rate": 9.596713882220739e-06,
+      "loss": 0.1035,
+      "step": 1299
+    },
+    {
+      "epoch": 0.34519383961763145,
+      "grad_norm": 0.5141149862836718,
+      "learning_rate": 9.595849649207274e-06,
+      "loss": 0.0693,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3454593733404142,
+      "grad_norm": 0.44396106651225886,
+      "learning_rate": 9.594984530168198e-06,
+      "loss": 0.0525,
+      "step": 1301
+    },
+    {
+      "epoch": 0.34572490706319703,
+      "grad_norm": 0.3455247284911814,
+      "learning_rate": 9.594118525270299e-06,
+      "loss": 0.0479,
+      "step": 1302
+    },
+    {
+      "epoch": 0.34599044078597985,
+      "grad_norm": 0.7585784664106814,
+      "learning_rate": 9.593251634680531e-06,
+      "loss": 0.0889,
+      "step": 1303
+    },
+    {
+      "epoch": 0.3462559745087626,
+      "grad_norm": 0.36123832598406336,
+      "learning_rate": 9.59238385856602e-06,
+      "loss": 0.0479,
+      "step": 1304
+    },
+    {
+      "epoch": 0.3465215082315454,
+      "grad_norm": 0.5097193438244858,
+      "learning_rate": 9.591515197094064e-06,
+      "loss": 0.082,
+      "step": 1305
+    },
+    {
+      "epoch": 0.3467870419543282,
+      "grad_norm": 0.2653818151867503,
+      "learning_rate": 9.59064565043213e-06,
+      "loss": 0.0354,
+      "step": 1306
+    },
+    {
+      "epoch": 0.347052575677111,
+      "grad_norm": 0.5864846744145519,
+      "learning_rate": 9.589775218747854e-06,
+      "loss": 0.0889,
+      "step": 1307
+    },
+    {
+      "epoch": 0.34731810939989377,
+      "grad_norm": 0.3781614336308014,
+      "learning_rate": 9.58890390220905e-06,
+      "loss": 0.0131,
+      "step": 1308
+    },
+    {
+      "epoch": 0.3475836431226766,
+      "grad_norm": 0.30092084572433875,
+      "learning_rate": 9.588031700983693e-06,
+      "loss": 0.0393,
+      "step": 1309
+    },
+    {
+      "epoch": 0.34784917684545935,
+      "grad_norm": 0.35904973427780035,
+      "learning_rate": 9.587158615239935e-06,
+      "loss": 0.0354,
+      "step": 1310
+    },
+    {
+      "epoch": 0.34811471056824217,
+      "grad_norm": 0.5600811056316272,
+      "learning_rate": 9.586284645146096e-06,
+      "loss": 0.0579,
+      "step": 1311
+    },
+    {
+      "epoch": 0.348380244291025,
+      "grad_norm": 0.05426252506203742,
+      "learning_rate": 9.58540979087067e-06,
+      "loss": 0.0015,
+      "step": 1312
+    },
+    {
+      "epoch": 0.34864577801380775,
+      "grad_norm": 0.5870410341108383,
+      "learning_rate": 9.584534052582314e-06,
+      "loss": 0.1035,
+      "step": 1313
+    },
+    {
+      "epoch": 0.34891131173659057,
+      "grad_norm": 0.6294228311893199,
+      "learning_rate": 9.583657430449862e-06,
+      "loss": 0.082,
+      "step": 1314
+    },
+    {
+      "epoch": 0.34917684545937333,
+      "grad_norm": 0.4895276277368698,
+      "learning_rate": 9.582779924642317e-06,
+      "loss": 0.0184,
+      "step": 1315
+    },
+    {
+      "epoch": 0.34944237918215615,
+      "grad_norm": 0.5733825038864722,
+      "learning_rate": 9.581901535328853e-06,
+      "loss": 0.082,
+      "step": 1316
+    },
+    {
+      "epoch": 0.3497079129049389,
+      "grad_norm": 0.32782198798301165,
+      "learning_rate": 9.58102226267881e-06,
+      "loss": 0.0146,
+      "step": 1317
+    },
+    {
+      "epoch": 0.34997344662772173,
+      "grad_norm": 0.590823967005489,
+      "learning_rate": 9.580142106861704e-06,
+      "loss": 0.0693,
+      "step": 1318
+    },
+    {
+      "epoch": 0.3502389803505045,
+      "grad_norm": 0.28663624606998783,
+      "learning_rate": 9.579261068047217e-06,
+      "loss": 0.0147,
+      "step": 1319
+    },
+    {
+      "epoch": 0.3505045140732873,
+      "grad_norm": 0.07002167741569795,
+      "learning_rate": 9.578379146405202e-06,
+      "loss": 0.002,
+      "step": 1320
+    },
+    {
+      "epoch": 0.35077004779607013,
+      "grad_norm": 0.3823299154936616,
+      "learning_rate": 9.577496342105686e-06,
+      "loss": 0.0476,
+      "step": 1321
+    },
+    {
+      "epoch": 0.3510355815188529,
+      "grad_norm": 0.41555476819705756,
+      "learning_rate": 9.576612655318863e-06,
+      "loss": 0.0184,
+      "step": 1322
+    },
+    {
+      "epoch": 0.3513011152416357,
+      "grad_norm": 0.6565947270509365,
+      "learning_rate": 9.575728086215093e-06,
+      "loss": 0.0479,
+      "step": 1323
+    },
+    {
+      "epoch": 0.35156664896441847,
+      "grad_norm": 0.28826966637792456,
+      "learning_rate": 9.574842634964917e-06,
+      "loss": 0.0354,
+      "step": 1324
+    },
+    {
+      "epoch": 0.3518321826872013,
+      "grad_norm": 0.47440584122085944,
+      "learning_rate": 9.573956301739033e-06,
+      "loss": 0.0576,
+      "step": 1325
+    },
+    {
+      "epoch": 0.35209771640998405,
+      "grad_norm": 0.013825970291321273,
+      "learning_rate": 9.573069086708322e-06,
+      "loss": 0.0003,
+      "step": 1326
+    },
+    {
+      "epoch": 0.35236325013276687,
+      "grad_norm": 0.308417343496391,
+      "learning_rate": 9.572180990043822e-06,
+      "loss": 0.0393,
+      "step": 1327
+    },
+    {
+      "epoch": 0.35262878385554963,
+      "grad_norm": 0.023751366893711805,
+      "learning_rate": 9.571292011916753e-06,
+      "loss": 0.0006,
+      "step": 1328
+    },
+    {
+      "epoch": 0.35289431757833245,
+      "grad_norm": 0.6910550885091158,
+      "learning_rate": 9.570402152498498e-06,
+      "loss": 0.0688,
+      "step": 1329
+    },
+    {
+      "epoch": 0.35315985130111527,
+      "grad_norm": 0.5584726753157215,
+      "learning_rate": 9.56951141196061e-06,
+      "loss": 0.063,
+      "step": 1330
+    },
+    {
+      "epoch": 0.35342538502389803,
+      "grad_norm": 0.3771730171037822,
+      "learning_rate": 9.568619790474815e-06,
+      "loss": 0.0432,
+      "step": 1331
+    },
+    {
+      "epoch": 0.35369091874668085,
+      "grad_norm": 0.006598573947351961,
+      "learning_rate": 9.567727288213005e-06,
+      "loss": 0.0001,
+      "step": 1332
+    },
+    {
+      "epoch": 0.3539564524694636,
+      "grad_norm": 0.24569141953841403,
+      "learning_rate": 9.566833905347245e-06,
+      "loss": 0.0286,
+      "step": 1333
+    },
+    {
+      "epoch": 0.35422198619224643,
+      "grad_norm": 0.31994068602155107,
+      "learning_rate": 9.56593964204977e-06,
+      "loss": 0.0354,
+      "step": 1334
+    },
+    {
+      "epoch": 0.3544875199150292,
+      "grad_norm": 0.48426314470272797,
+      "learning_rate": 9.565044498492984e-06,
+      "loss": 0.0476,
+      "step": 1335
+    },
+    {
+      "epoch": 0.354753053637812,
+      "grad_norm": 0.4284293731700466,
+      "learning_rate": 9.564148474849455e-06,
+      "loss": 0.0476,
+      "step": 1336
+    },
+    {
+      "epoch": 0.3550185873605948,
+      "grad_norm": 0.5223639882465799,
+      "learning_rate": 9.563251571291932e-06,
+      "loss": 0.0432,
+      "step": 1337
+    },
+    {
+      "epoch": 0.3552841210833776,
+      "grad_norm": 0.3682885994989024,
+      "learning_rate": 9.562353787993321e-06,
+      "loss": 0.0432,
+      "step": 1338
+    },
+    {
+      "epoch": 0.3555496548061604,
+      "grad_norm": 0.7276856913645094,
+      "learning_rate": 9.56145512512671e-06,
+      "loss": 0.0957,
+      "step": 1339
+    },
+    {
+      "epoch": 0.35581518852894317,
+      "grad_norm": 0.28021489029743807,
+      "learning_rate": 9.56055558286535e-06,
+      "loss": 0.0317,
+      "step": 1340
+    },
+    {
+      "epoch": 0.356080722251726,
+      "grad_norm": 0.01832087245829523,
+      "learning_rate": 9.559655161382658e-06,
+      "loss": 0.0004,
+      "step": 1341
+    },
+    {
+      "epoch": 0.35634625597450875,
+      "grad_norm": 0.914955086500517,
+      "learning_rate": 9.558753860852228e-06,
+      "loss": 0.0752,
+      "step": 1342
+    },
+    {
+      "epoch": 0.35661178969729157,
+      "grad_norm": 0.20459946602158102,
+      "learning_rate": 9.557851681447818e-06,
+      "loss": 0.0206,
+      "step": 1343
+    },
+    {
+      "epoch": 0.35687732342007433,
+      "grad_norm": 0.27576350826669555,
+      "learning_rate": 9.556948623343358e-06,
+      "loss": 0.0286,
+      "step": 1344
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 0.2761915334889806,
+      "learning_rate": 9.55604468671295e-06,
+      "loss": 0.0255,
+      "step": 1345
+    },
+    {
+      "epoch": 0.3574083908656399,
+      "grad_norm": 0.8233226661583315,
+      "learning_rate": 9.555139871730859e-06,
+      "loss": 0.0354,
+      "step": 1346
+    },
+    {
+      "epoch": 0.35767392458842273,
+      "grad_norm": 0.1926615669215923,
+      "learning_rate": 9.554234178571524e-06,
+      "loss": 0.0165,
+      "step": 1347
+    },
+    {
+      "epoch": 0.35793945831120555,
+      "grad_norm": 0.3098753295444041,
+      "learning_rate": 9.55332760740955e-06,
+      "loss": 0.0286,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3582049920339883,
+      "grad_norm": 3.431526479191698,
+      "learning_rate": 9.552420158419715e-06,
+      "loss": 0.1191,
+      "step": 1349
+    },
+    {
+      "epoch": 0.35847052575677113,
+      "grad_norm": 0.0625932742877244,
+      "learning_rate": 9.551511831776966e-06,
+      "loss": 0.001,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3587360594795539,
+      "grad_norm": 1.1478320124700105,
+      "learning_rate": 9.550602627656415e-06,
+      "loss": 0.0889,
+      "step": 1351
+    },
+    {
+      "epoch": 0.3590015932023367,
+      "grad_norm": 1.003238754359967,
+      "learning_rate": 9.549692546233348e-06,
+      "loss": 0.0688,
+      "step": 1352
+    },
+    {
+      "epoch": 0.3592671269251195,
+      "grad_norm": 0.25770729335411,
+      "learning_rate": 9.548781587683214e-06,
+      "loss": 0.0061,
+      "step": 1353
+    },
+    {
+      "epoch": 0.3595326606479023,
+      "grad_norm": 1.55958556462164,
+      "learning_rate": 9.54786975218164e-06,
+      "loss": 0.0688,
+      "step": 1354
+    },
+    {
+      "epoch": 0.35979819437068505,
+      "grad_norm": 1.2409731636678565,
+      "learning_rate": 9.546957039904414e-06,
+      "loss": 0.0688,
+      "step": 1355
+    },
+    {
+      "epoch": 0.36006372809346787,
+      "grad_norm": 0.032263196311481473,
+      "learning_rate": 9.546043451027498e-06,
+      "loss": 0.0007,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3603292618162507,
+      "grad_norm": 0.4590699936620603,
+      "learning_rate": 9.54512898572702e-06,
+      "loss": 0.0205,
+      "step": 1357
+    },
+    {
+      "epoch": 0.36059479553903345,
+      "grad_norm": 6.157784939523085,
+      "learning_rate": 9.544213644179278e-06,
+      "loss": 0.4668,
+      "step": 1358
+    },
+    {
+      "epoch": 0.36086032926181627,
+      "grad_norm": 1.0445567828018743,
+      "learning_rate": 9.54329742656074e-06,
+      "loss": 0.1035,
+      "step": 1359
+    },
+    {
+      "epoch": 0.36112586298459903,
+      "grad_norm": 0.900975787190914,
+      "learning_rate": 9.54238033304804e-06,
+      "loss": 0.0525,
+      "step": 1360
+    },
+    {
+      "epoch": 0.36139139670738185,
+      "grad_norm": 2.1445461499571827,
+      "learning_rate": 9.541462363817986e-06,
+      "loss": 0.1279,
+      "step": 1361
+    },
+    {
+      "epoch": 0.3616569304301646,
+      "grad_norm": 2.6058708901982754,
+      "learning_rate": 9.540543519047547e-06,
+      "loss": 0.0889,
+      "step": 1362
+    },
+    {
+      "epoch": 0.36192246415294743,
+      "grad_norm": 0.30837506165256673,
+      "learning_rate": 9.53962379891387e-06,
+      "loss": 0.0255,
+      "step": 1363
+    },
+    {
+      "epoch": 0.3621879978757302,
+      "grad_norm": 0.6781227526654185,
+      "learning_rate": 9.538703203594266e-06,
+      "loss": 0.0476,
+      "step": 1364
+    },
+    {
+      "epoch": 0.362453531598513,
+      "grad_norm": 0.0014988796322645392,
+      "learning_rate": 9.53778173326621e-06,
+      "loss": 0.0,
+      "step": 1365
+    },
+    {
+      "epoch": 0.36271906532129583,
+      "grad_norm": 0.4848181673027792,
+      "learning_rate": 9.536859388107359e-06,
+      "loss": 0.0354,
+      "step": 1366
+    },
+    {
+      "epoch": 0.3629845990440786,
+      "grad_norm": 0.30798774656542016,
+      "learning_rate": 9.535936168295521e-06,
+      "loss": 0.0255,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3632501327668614,
+      "grad_norm": 1.1103162493296537,
+      "learning_rate": 9.535012074008688e-06,
+      "loss": 0.1191,
+      "step": 1368
+    },
+    {
+      "epoch": 0.3635156664896442,
+      "grad_norm": 1.115968824289187,
+      "learning_rate": 9.53408710542501e-06,
+      "loss": 0.1934,
+      "step": 1369
+    },
+    {
+      "epoch": 0.363781200212427,
+      "grad_norm": 0.1558534145654595,
+      "learning_rate": 9.533161262722815e-06,
+      "loss": 0.0131,
+      "step": 1370
+    },
+    {
+      "epoch": 0.36404673393520975,
+      "grad_norm": 0.042070066792334465,
+      "learning_rate": 9.532234546080592e-06,
+      "loss": 0.0008,
+      "step": 1371
+    },
+    {
+      "epoch": 0.3643122676579926,
+      "grad_norm": 0.05020993746733037,
+      "learning_rate": 9.531306955677003e-06,
+      "loss": 0.0012,
+      "step": 1372
+    },
+    {
+      "epoch": 0.36457780138077533,
+      "grad_norm": 0.3478946473030989,
+      "learning_rate": 9.530378491690872e-06,
+      "loss": 0.0255,
+      "step": 1373
+    },
+    {
+      "epoch": 0.36484333510355815,
+      "grad_norm": 0.532671213666101,
+      "learning_rate": 9.5294491543012e-06,
+      "loss": 0.0286,
+      "step": 1374
+    },
+    {
+      "epoch": 0.36510886882634097,
+      "grad_norm": 0.0104645404151327,
+      "learning_rate": 9.52851894368715e-06,
+      "loss": 0.0002,
+      "step": 1375
+    },
+    {
+      "epoch": 0.36537440254912373,
+      "grad_norm": 0.6407476894876308,
+      "learning_rate": 9.527587860028055e-06,
+      "loss": 0.0165,
+      "step": 1376
+    },
+    {
+      "epoch": 0.36563993627190655,
+      "grad_norm": 1.683239750297336,
+      "learning_rate": 9.526655903503423e-06,
+      "loss": 0.0688,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3659054699946893,
+      "grad_norm": 0.1956603056666467,
+      "learning_rate": 9.525723074292916e-06,
+      "loss": 0.0184,
+      "step": 1378
+    },
+    {
+      "epoch": 0.36617100371747213,
+      "grad_norm": 0.6108332993871155,
+      "learning_rate": 9.524789372576377e-06,
+      "loss": 0.0317,
+      "step": 1379
+    },
+    {
+      "epoch": 0.3664365374402549,
+      "grad_norm": 0.03952181334754172,
+      "learning_rate": 9.523854798533814e-06,
+      "loss": 0.0012,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3667020711630377,
+      "grad_norm": 0.18831659774689863,
+      "learning_rate": 9.522919352345398e-06,
+      "loss": 0.0024,
+      "step": 1381
+    },
+    {
+      "epoch": 0.3669676048858205,
+      "grad_norm": 0.8786083910112925,
+      "learning_rate": 9.521983034191472e-06,
+      "loss": 0.0432,
+      "step": 1382
+    },
+    {
+      "epoch": 0.3672331386086033,
+      "grad_norm": 0.1730242308325762,
+      "learning_rate": 9.521045844252552e-06,
+      "loss": 0.0146,
+      "step": 1383
+    },
+    {
+      "epoch": 0.3674986723313861,
+      "grad_norm": 1.2876116546269887,
+      "learning_rate": 9.520107782709314e-06,
+      "loss": 0.0815,
+      "step": 1384
+    },
+    {
+      "epoch": 0.3677642060541689,
+      "grad_norm": 1.3082256972483228,
+      "learning_rate": 9.519168849742603e-06,
+      "loss": 0.0752,
+      "step": 1385
+    },
+    {
+      "epoch": 0.3680297397769517,
+      "grad_norm": 0.9335061419191978,
+      "learning_rate": 9.518229045533438e-06,
+      "loss": 0.0576,
+      "step": 1386
+    },
+    {
+      "epoch": 0.36829527349973445,
+      "grad_norm": 2.446950303059504,
+      "learning_rate": 9.517288370263e-06,
+      "loss": 0.2246,
+      "step": 1387
+    },
+    {
+      "epoch": 0.3685608072225173,
+      "grad_norm": 0.6333100630168161,
+      "learning_rate": 9.51634682411264e-06,
+      "loss": 0.0123,
+      "step": 1388
+    },
+    {
+      "epoch": 0.36882634094530004,
+      "grad_norm": 1.1503551358840165,
+      "learning_rate": 9.51540440726388e-06,
+      "loss": 0.1035,
+      "step": 1389
+    },
+    {
+      "epoch": 0.36909187466808285,
+      "grad_norm": 0.3449715903261567,
+      "learning_rate": 9.514461119898401e-06,
+      "loss": 0.0092,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3693574083908656,
+      "grad_norm": 0.23906913907246705,
+      "learning_rate": 9.513516962198064e-06,
+      "loss": 0.0229,
+      "step": 1391
+    },
+    {
+      "epoch": 0.36962294211364843,
+      "grad_norm": 0.4848006955734177,
+      "learning_rate": 9.512571934344887e-06,
+      "loss": 0.0117,
+      "step": 1392
+    },
+    {
+      "epoch": 0.36988847583643125,
+      "grad_norm": 0.9490909420873752,
+      "learning_rate": 9.511626036521064e-06,
+      "loss": 0.0889,
+      "step": 1393
+    },
+    {
+      "epoch": 0.370154009559214,
+      "grad_norm": 3.92147853906124,
+      "learning_rate": 9.510679268908949e-06,
+      "loss": 0.1934,
+      "step": 1394
+    },
+    {
+      "epoch": 0.37041954328199683,
+      "grad_norm": 2.8735279348285796,
+      "learning_rate": 9.509731631691071e-06,
+      "loss": 0.2129,
+      "step": 1395
+    },
+    {
+      "epoch": 0.3706850770047796,
+      "grad_norm": 0.2436877641705072,
+      "learning_rate": 9.50878312505012e-06,
+      "loss": 0.0205,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3709506107275624,
+      "grad_norm": 1.3855718841062967,
+      "learning_rate": 9.50783374916896e-06,
+      "loss": 0.063,
+      "step": 1397
+    },
+    {
+      "epoch": 0.3712161444503452,
+      "grad_norm": 0.39133070889265453,
+      "learning_rate": 9.506883504230618e-06,
+      "loss": 0.0354,
+      "step": 1398
+    },
+    {
+      "epoch": 0.371481678173128,
+      "grad_norm": 0.6442273431688003,
+      "learning_rate": 9.50593239041829e-06,
+      "loss": 0.0476,
+      "step": 1399
+    },
+    {
+      "epoch": 0.37174721189591076,
+      "grad_norm": 0.3654458756752601,
+      "learning_rate": 9.504980407915338e-06,
+      "loss": 0.0354,
+      "step": 1400
+    },
+    {
+      "epoch": 0.3720127456186936,
+      "grad_norm": 0.8586712209152337,
+      "learning_rate": 9.504027556905297e-06,
+      "loss": 0.1279,
+      "step": 1401
+    },
+    {
+      "epoch": 0.3722782793414764,
+      "grad_norm": 0.6773527175343205,
+      "learning_rate": 9.503073837571862e-06,
+      "loss": 0.0576,
+      "step": 1402
+    },
+    {
+      "epoch": 0.37254381306425915,
+      "grad_norm": 0.4186954158863423,
+      "learning_rate": 9.5021192500989e-06,
+      "loss": 0.0352,
+      "step": 1403
+    },
+    {
+      "epoch": 0.372809346787042,
+      "grad_norm": 1.9132653529788708,
+      "learning_rate": 9.501163794670445e-06,
+      "loss": 0.1738,
+      "step": 1404
+    },
+    {
+      "epoch": 0.37307488050982474,
+      "grad_norm": 2.2952198425920582,
+      "learning_rate": 9.500207471470695e-06,
+      "loss": 0.0815,
+      "step": 1405
+    },
+    {
+      "epoch": 0.37334041423260755,
+      "grad_norm": 0.40586858759644967,
+      "learning_rate": 9.49925028068402e-06,
+      "loss": 0.0165,
+      "step": 1406
+    },
+    {
+      "epoch": 0.3736059479553903,
+      "grad_norm": 0.18480641707554119,
+      "learning_rate": 9.498292222494955e-06,
+      "loss": 0.0035,
+      "step": 1407
+    },
+    {
+      "epoch": 0.37387148167817313,
+      "grad_norm": 0.7772486998466412,
+      "learning_rate": 9.497333297088202e-06,
+      "loss": 0.1191,
+      "step": 1408
+    },
+    {
+      "epoch": 0.3741370154009559,
+      "grad_norm": 0.2965999353034397,
+      "learning_rate": 9.496373504648632e-06,
+      "loss": 0.0286,
+      "step": 1409
+    },
+    {
+      "epoch": 0.3744025491237387,
+      "grad_norm": 2.294123343793471,
+      "learning_rate": 9.495412845361279e-06,
+      "loss": 0.1934,
+      "step": 1410
+    },
+    {
+      "epoch": 0.37466808284652153,
+      "grad_norm": 0.0026701934360799534,
+      "learning_rate": 9.494451319411347e-06,
+      "loss": 0.0001,
+      "step": 1411
+    },
+    {
+      "epoch": 0.3749336165693043,
+      "grad_norm": 1.4911265196857602,
+      "learning_rate": 9.493488926984208e-06,
+      "loss": 0.063,
+      "step": 1412
+    },
+    {
+      "epoch": 0.3751991502920871,
+      "grad_norm": 0.9596610809296374,
+      "learning_rate": 9.4925256682654e-06,
+      "loss": 0.0752,
+      "step": 1413
+    },
+    {
+      "epoch": 0.3754646840148699,
+      "grad_norm": 0.8806164654939721,
+      "learning_rate": 9.491561543440627e-06,
+      "loss": 0.0391,
+      "step": 1414
+    },
+    {
+      "epoch": 0.3757302177376527,
+      "grad_norm": 0.7690458874265063,
+      "learning_rate": 9.490596552695762e-06,
+      "loss": 0.1191,
+      "step": 1415
+    },
+    {
+      "epoch": 0.37599575146043546,
+      "grad_norm": 0.6872344076762994,
+      "learning_rate": 9.489630696216842e-06,
+      "loss": 0.0576,
+      "step": 1416
+    },
+    {
+      "epoch": 0.3762612851832183,
+      "grad_norm": 0.48727643854025365,
+      "learning_rate": 9.488663974190077e-06,
+      "loss": 0.0354,
+      "step": 1417
+    },
+    {
+      "epoch": 0.37652681890600104,
+      "grad_norm": 0.7530220920460418,
+      "learning_rate": 9.487696386801835e-06,
+      "loss": 0.0752,
+      "step": 1418
+    },
+    {
+      "epoch": 0.37679235262878386,
+      "grad_norm": 0.29673241857480187,
+      "learning_rate": 9.486727934238656e-06,
+      "loss": 0.0079,
+      "step": 1419
+    },
+    {
+      "epoch": 0.3770578863515667,
+      "grad_norm": 0.3529062224624686,
+      "learning_rate": 9.485758616687248e-06,
+      "loss": 0.0093,
+      "step": 1420
+    },
+    {
+      "epoch": 0.37732342007434944,
+      "grad_norm": 0.09031390787732817,
+      "learning_rate": 9.484788434334484e-06,
+      "loss": 0.0028,
+      "step": 1421
+    },
+    {
+      "epoch": 0.37758895379713225,
+      "grad_norm": 0.6189002563525746,
+      "learning_rate": 9.483817387367403e-06,
+      "loss": 0.0889,
+      "step": 1422
+    },
+    {
+      "epoch": 0.377854487519915,
+      "grad_norm": 0.8659301327268721,
+      "learning_rate": 9.482845475973212e-06,
+      "loss": 0.1113,
+      "step": 1423
+    },
+    {
+      "epoch": 0.37812002124269783,
+      "grad_norm": 0.025135957238805916,
+      "learning_rate": 9.481872700339282e-06,
+      "loss": 0.0007,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3783855549654806,
+      "grad_norm": 0.20693260579313302,
+      "learning_rate": 9.480899060653154e-06,
+      "loss": 0.0105,
+      "step": 1425
+    },
+    {
+      "epoch": 0.3786510886882634,
+      "grad_norm": 3.127258712138675,
+      "learning_rate": 9.479924557102536e-06,
+      "loss": 0.2344,
+      "step": 1426
+    },
+    {
+      "epoch": 0.3789166224110462,
+      "grad_norm": 0.20096836647101268,
+      "learning_rate": 9.478949189875299e-06,
+      "loss": 0.004,
+      "step": 1427
+    },
+    {
+      "epoch": 0.379182156133829,
+      "grad_norm": 1.2506042183343944,
+      "learning_rate": 9.47797295915948e-06,
+      "loss": 0.1113,
+      "step": 1428
+    },
+    {
+      "epoch": 0.3794476898566118,
+      "grad_norm": 0.6735848636167625,
+      "learning_rate": 9.47699586514329e-06,
+      "loss": 0.0688,
+      "step": 1429
+    },
+    {
+      "epoch": 0.3797132235793946,
+      "grad_norm": 0.8729493010310873,
+      "learning_rate": 9.476017908015098e-06,
+      "loss": 0.0815,
+      "step": 1430
+    },
+    {
+      "epoch": 0.3799787573021774,
+      "grad_norm": 0.19365690895180718,
+      "learning_rate": 9.475039087963443e-06,
+      "loss": 0.0041,
+      "step": 1431
+    },
+    {
+      "epoch": 0.38024429102496016,
+      "grad_norm": 0.5438261709069115,
+      "learning_rate": 9.47405940517703e-06,
+      "loss": 0.063,
+      "step": 1432
+    },
+    {
+      "epoch": 0.380509824747743,
+      "grad_norm": 0.32752487349022863,
+      "learning_rate": 9.47307885984473e-06,
+      "loss": 0.0432,
+      "step": 1433
+    },
+    {
+      "epoch": 0.38077535847052574,
+      "grad_norm": 0.45526266402944837,
+      "learning_rate": 9.47209745215558e-06,
+      "loss": 0.032,
+      "step": 1434
+    },
+    {
+      "epoch": 0.38104089219330856,
+      "grad_norm": 1.0213144942650498,
+      "learning_rate": 9.471115182298786e-06,
+      "loss": 0.1035,
+      "step": 1435
+    },
+    {
+      "epoch": 0.3813064259160913,
+      "grad_norm": 0.4413563467345107,
+      "learning_rate": 9.470132050463716e-06,
+      "loss": 0.0354,
+      "step": 1436
+    },
+    {
+      "epoch": 0.38157195963887414,
+      "grad_norm": 0.0613716953521828,
+      "learning_rate": 9.469148056839907e-06,
+      "loss": 0.002,
+      "step": 1437
+    },
+    {
+      "epoch": 0.38183749336165695,
+      "grad_norm": 0.8153858333079927,
+      "learning_rate": 9.468163201617063e-06,
+      "loss": 0.082,
+      "step": 1438
+    },
+    {
+      "epoch": 0.3821030270844397,
+      "grad_norm": 0.5563381804784219,
+      "learning_rate": 9.467177484985048e-06,
+      "loss": 0.063,
+      "step": 1439
+    },
+    {
+      "epoch": 0.38236856080722254,
+      "grad_norm": 0.4245368835018735,
+      "learning_rate": 9.466190907133901e-06,
+      "loss": 0.0479,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3826340945300053,
+      "grad_norm": 0.33400027625309914,
+      "learning_rate": 9.46520346825382e-06,
+      "loss": 0.0432,
+      "step": 1441
+    },
+    {
+      "epoch": 0.3828996282527881,
+      "grad_norm": 0.14631192270559776,
+      "learning_rate": 9.464215168535175e-06,
+      "loss": 0.0045,
+      "step": 1442
+    },
+    {
+      "epoch": 0.3831651619755709,
+      "grad_norm": 0.4620703244763112,
+      "learning_rate": 9.463226008168493e-06,
+      "loss": 0.0432,
+      "step": 1443
+    },
+    {
+      "epoch": 0.3834306956983537,
+      "grad_norm": 0.6625505596890234,
+      "learning_rate": 9.462235987344478e-06,
+      "loss": 0.0579,
+      "step": 1444
+    },
+    {
+      "epoch": 0.38369622942113646,
+      "grad_norm": 0.24691441798104835,
+      "learning_rate": 9.461245106253991e-06,
+      "loss": 0.0105,
+      "step": 1445
+    },
+    {
+      "epoch": 0.3839617631439193,
+      "grad_norm": 0.6337463070104612,
+      "learning_rate": 9.460253365088063e-06,
+      "loss": 0.0752,
+      "step": 1446
+    },
+    {
+      "epoch": 0.3842272968667021,
+      "grad_norm": 0.08857326498123558,
+      "learning_rate": 9.45926076403789e-06,
+      "loss": 0.0028,
+      "step": 1447
+    },
+    {
+      "epoch": 0.38449283058948486,
+      "grad_norm": 0.614060872702447,
+      "learning_rate": 9.458267303294833e-06,
+      "loss": 0.0957,
+      "step": 1448
+    },
+    {
+      "epoch": 0.3847583643122677,
+      "grad_norm": 0.34706341100621624,
+      "learning_rate": 9.457272983050421e-06,
+      "loss": 0.0432,
+      "step": 1449
+    },
+    {
+      "epoch": 0.38502389803505044,
+      "grad_norm": 0.5244496312054608,
+      "learning_rate": 9.456277803496345e-06,
+      "loss": 0.082,
+      "step": 1450
+    },
+    {
+      "epoch": 0.38528943175783326,
+      "grad_norm": 0.04965359509992197,
+      "learning_rate": 9.455281764824467e-06,
+      "loss": 0.0015,
+      "step": 1451
+    },
+    {
+      "epoch": 0.385554965480616,
+      "grad_norm": 0.4205660220211263,
+      "learning_rate": 9.45428486722681e-06,
+      "loss": 0.0286,
+      "step": 1452
+    },
+    {
+      "epoch": 0.38582049920339884,
+      "grad_norm": 0.5607004309206753,
+      "learning_rate": 9.453287110895564e-06,
+      "loss": 0.063,
+      "step": 1453
+    },
+    {
+      "epoch": 0.3860860329261816,
+      "grad_norm": 0.47148899218207774,
+      "learning_rate": 9.452288496023085e-06,
+      "loss": 0.0576,
+      "step": 1454
+    },
+    {
+      "epoch": 0.3863515666489644,
+      "grad_norm": 0.46004679938747756,
+      "learning_rate": 9.451289022801894e-06,
+      "loss": 0.0579,
+      "step": 1455
+    },
+    {
+      "epoch": 0.38661710037174724,
+      "grad_norm": 0.4292369995431107,
+      "learning_rate": 9.450288691424676e-06,
+      "loss": 0.0317,
+      "step": 1456
+    },
+    {
+      "epoch": 0.38688263409453,
+      "grad_norm": 0.9283721518542083,
+      "learning_rate": 9.449287502084284e-06,
+      "loss": 0.0957,
+      "step": 1457
+    },
+    {
+      "epoch": 0.3871481678173128,
+      "grad_norm": 1.199281966245374,
+      "learning_rate": 9.448285454973739e-06,
+      "loss": 0.063,
+      "step": 1458
+    },
+    {
+      "epoch": 0.3874137015400956,
+      "grad_norm": 1.5176795239281788,
+      "learning_rate": 9.447282550286217e-06,
+      "loss": 0.1191,
+      "step": 1459
+    },
+    {
+      "epoch": 0.3876792352628784,
+      "grad_norm": 0.574751228105574,
+      "learning_rate": 9.446278788215074e-06,
+      "loss": 0.0889,
+      "step": 1460
+    },
+    {
+      "epoch": 0.38794476898566116,
+      "grad_norm": 0.05157646996776102,
+      "learning_rate": 9.445274168953818e-06,
+      "loss": 0.0012,
+      "step": 1461
+    },
+    {
+      "epoch": 0.388210302708444,
+      "grad_norm": 0.7857880559484768,
+      "learning_rate": 9.444268692696129e-06,
+      "loss": 0.082,
+      "step": 1462
+    },
+    {
+      "epoch": 0.38847583643122674,
+      "grad_norm": 0.3007386860700484,
+      "learning_rate": 9.44326235963585e-06,
+      "loss": 0.0393,
+      "step": 1463
+    },
+    {
+      "epoch": 0.38874137015400956,
+      "grad_norm": 0.3055671787388409,
+      "learning_rate": 9.442255169966995e-06,
+      "loss": 0.0354,
+      "step": 1464
+    },
+    {
+      "epoch": 0.3890069038767924,
+      "grad_norm": 0.5568087984102646,
+      "learning_rate": 9.441247123883733e-06,
+      "loss": 0.0889,
+      "step": 1465
+    },
+    {
+      "epoch": 0.38927243759957514,
+      "grad_norm": 0.5895505413874539,
+      "learning_rate": 9.440238221580404e-06,
+      "loss": 0.1035,
+      "step": 1466
+    },
+    {
+      "epoch": 0.38953797132235796,
+      "grad_norm": 0.5616300389774914,
+      "learning_rate": 9.439228463251515e-06,
+      "loss": 0.0957,
+      "step": 1467
+    },
+    {
+      "epoch": 0.3898035050451407,
+      "grad_norm": 0.33063941694631,
+      "learning_rate": 9.438217849091732e-06,
+      "loss": 0.0393,
+      "step": 1468
+    },
+    {
+      "epoch": 0.39006903876792354,
+      "grad_norm": 0.2674438568163559,
+      "learning_rate": 9.437206379295895e-06,
+      "loss": 0.0354,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3903345724907063,
+      "grad_norm": 0.7808160049243492,
+      "learning_rate": 9.436194054058998e-06,
+      "loss": 0.0752,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3906001062134891,
+      "grad_norm": 0.5509891118998486,
+      "learning_rate": 9.435180873576208e-06,
+      "loss": 0.0957,
+      "step": 1471
+    },
+    {
+      "epoch": 0.3908656399362719,
+      "grad_norm": 0.6425294720673329,
+      "learning_rate": 9.434166838042854e-06,
+      "loss": 0.0889,
+      "step": 1472
+    },
+    {
+      "epoch": 0.3911311736590547,
+      "grad_norm": 1.8808615893246776,
+      "learning_rate": 9.433151947654428e-06,
+      "loss": 0.1279,
+      "step": 1473
+    },
+    {
+      "epoch": 0.3913967073818375,
+      "grad_norm": 0.6663733492038063,
+      "learning_rate": 9.432136202606594e-06,
+      "loss": 0.0184,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3916622411046203,
+      "grad_norm": 0.5208712307214396,
+      "learning_rate": 9.431119603095169e-06,
+      "loss": 0.0889,
+      "step": 1475
+    },
+    {
+      "epoch": 0.3919277748274031,
+      "grad_norm": 0.2713853966362412,
+      "learning_rate": 9.430102149316147e-06,
+      "loss": 0.0354,
+      "step": 1476
+    },
+    {
+      "epoch": 0.39219330855018586,
+      "grad_norm": 0.5349973334569837,
+      "learning_rate": 9.429083841465679e-06,
+      "loss": 0.0957,
+      "step": 1477
+    },
+    {
+      "epoch": 0.3924588422729687,
+      "grad_norm": 0.1267971200504465,
+      "learning_rate": 9.428064679740081e-06,
+      "loss": 0.0025,
+      "step": 1478
+    },
+    {
+      "epoch": 0.39272437599575144,
+      "grad_norm": 0.5108936337689303,
+      "learning_rate": 9.427044664335837e-06,
+      "loss": 0.0815,
+      "step": 1479
+    },
+    {
+      "epoch": 0.39298990971853426,
+      "grad_norm": 0.33305136147342523,
+      "learning_rate": 9.426023795449595e-06,
+      "loss": 0.0479,
+      "step": 1480
+    },
+    {
+      "epoch": 0.393255443441317,
+      "grad_norm": 0.2984668226042693,
+      "learning_rate": 9.425002073278165e-06,
+      "loss": 0.0432,
+      "step": 1481
+    },
+    {
+      "epoch": 0.39352097716409984,
+      "grad_norm": 0.33394165373144796,
+      "learning_rate": 9.423979498018521e-06,
+      "loss": 0.0432,
+      "step": 1482
+    },
+    {
+      "epoch": 0.39378651088688266,
+      "grad_norm": 0.49293937391002435,
+      "learning_rate": 9.422956069867808e-06,
+      "loss": 0.0889,
+      "step": 1483
+    },
+    {
+      "epoch": 0.3940520446096654,
+      "grad_norm": 0.5210948201626876,
+      "learning_rate": 9.421931789023329e-06,
+      "loss": 0.0579,
+      "step": 1484
+    },
+    {
+      "epoch": 0.39431757833244824,
+      "grad_norm": 0.41304821586273616,
+      "learning_rate": 9.420906655682553e-06,
+      "loss": 0.0432,
+      "step": 1485
+    },
+    {
+      "epoch": 0.394583112055231,
+      "grad_norm": 0.46657739059291675,
+      "learning_rate": 9.41988067004311e-06,
+      "loss": 0.0354,
+      "step": 1486
+    },
+    {
+      "epoch": 0.3948486457780138,
+      "grad_norm": 0.026843605916078473,
+      "learning_rate": 9.418853832302806e-06,
+      "loss": 0.0007,
+      "step": 1487
+    },
+    {
+      "epoch": 0.3951141795007966,
+      "grad_norm": 0.2666054121708235,
+      "learning_rate": 9.417826142659596e-06,
+      "loss": 0.0103,
+      "step": 1488
+    },
+    {
+      "epoch": 0.3953797132235794,
+      "grad_norm": 0.3517104605858895,
+      "learning_rate": 9.416797601311611e-06,
+      "loss": 0.032,
+      "step": 1489
+    },
+    {
+      "epoch": 0.39564524694636216,
+      "grad_norm": 0.448265208584516,
+      "learning_rate": 9.415768208457139e-06,
+      "loss": 0.0693,
+      "step": 1490
+    },
+    {
+      "epoch": 0.395910780669145,
+      "grad_norm": 0.4432887319257767,
+      "learning_rate": 9.414737964294636e-06,
+      "loss": 0.063,
+      "step": 1491
+    },
+    {
+      "epoch": 0.3961763143919278,
+      "grad_norm": 0.15615182853960044,
+      "learning_rate": 9.41370686902272e-06,
+      "loss": 0.0057,
+      "step": 1492
+    },
+    {
+      "epoch": 0.39644184811471056,
+      "grad_norm": 0.40174854437847995,
+      "learning_rate": 9.412674922840175e-06,
+      "loss": 0.0432,
+      "step": 1493
+    },
+    {
+      "epoch": 0.3967073818374934,
+      "grad_norm": 0.5334231337886949,
+      "learning_rate": 9.411642125945945e-06,
+      "loss": 0.0752,
+      "step": 1494
+    },
+    {
+      "epoch": 0.39697291556027614,
+      "grad_norm": 0.39350262505624917,
+      "learning_rate": 9.410608478539146e-06,
+      "loss": 0.0525,
+      "step": 1495
+    },
+    {
+      "epoch": 0.39723844928305896,
+      "grad_norm": 0.5104750407099554,
+      "learning_rate": 9.409573980819052e-06,
+      "loss": 0.0432,
+      "step": 1496
+    },
+    {
+      "epoch": 0.3975039830058417,
+      "grad_norm": 0.018301303401147544,
+      "learning_rate": 9.408538632985098e-06,
+      "loss": 0.0005,
+      "step": 1497
+    },
+    {
+      "epoch": 0.39776951672862454,
+      "grad_norm": 2.925231221732229,
+      "learning_rate": 9.407502435236889e-06,
+      "loss": 0.0184,
+      "step": 1498
+    },
+    {
+      "epoch": 0.3980350504514073,
+      "grad_norm": 0.43976667471652925,
+      "learning_rate": 9.406465387774194e-06,
+      "loss": 0.032,
+      "step": 1499
+    },
+    {
+      "epoch": 0.3983005841741901,
+      "grad_norm": 0.004447437986864806,
+      "learning_rate": 9.405427490796941e-06,
+      "loss": 0.0001,
+      "step": 1500
+    },
+    {
+      "epoch": 0.39856611789697294,
+      "grad_norm": 0.22189544659424998,
+      "learning_rate": 9.404388744505225e-06,
+      "loss": 0.0147,
+      "step": 1501
+    },
+    {
+      "epoch": 0.3988316516197557,
+      "grad_norm": 0.5323862127877603,
+      "learning_rate": 9.403349149099304e-06,
+      "loss": 0.0354,
+      "step": 1502
+    },
+    {
+      "epoch": 0.3990971853425385,
+      "grad_norm": 0.7824556156322893,
+      "learning_rate": 9.4023087047796e-06,
+      "loss": 0.1035,
+      "step": 1503
+    },
+    {
+      "epoch": 0.3993627190653213,
+      "grad_norm": 0.537555019938912,
+      "learning_rate": 9.401267411746697e-06,
+      "loss": 0.0752,
+      "step": 1504
+    },
+    {
+      "epoch": 0.3996282527881041,
+      "grad_norm": 1.2599751022767143,
+      "learning_rate": 9.400225270201347e-06,
+      "loss": 0.1279,
+      "step": 1505
+    },
+    {
+      "epoch": 0.39989378651088686,
+      "grad_norm": 0.39943457727783255,
+      "learning_rate": 9.399182280344458e-06,
+      "loss": 0.0286,
+      "step": 1506
+    },
+    {
+      "epoch": 0.4001593202336697,
+      "grad_norm": 0.3951237763375591,
+      "learning_rate": 9.398138442377112e-06,
+      "loss": 0.0354,
+      "step": 1507
+    },
+    {
+      "epoch": 0.40042485395645244,
+      "grad_norm": 0.3348310768931224,
+      "learning_rate": 9.397093756500544e-06,
+      "loss": 0.0258,
+      "step": 1508
+    },
+    {
+      "epoch": 0.40069038767923526,
+      "grad_norm": 0.6183702558458398,
+      "learning_rate": 9.396048222916159e-06,
+      "loss": 0.0889,
+      "step": 1509
+    },
+    {
+      "epoch": 0.4009559214020181,
+      "grad_norm": 0.7389099287695277,
+      "learning_rate": 9.395001841825523e-06,
+      "loss": 0.0957,
+      "step": 1510
+    },
+    {
+      "epoch": 0.40122145512480084,
+      "grad_norm": 0.5072802900638348,
+      "learning_rate": 9.393954613430365e-06,
+      "loss": 0.0752,
+      "step": 1511
+    },
+    {
+      "epoch": 0.40148698884758366,
+      "grad_norm": 0.6807209999699108,
+      "learning_rate": 9.392906537932582e-06,
+      "loss": 0.0957,
+      "step": 1512
+    },
+    {
+      "epoch": 0.4017525225703664,
+      "grad_norm": 0.4704376229884201,
+      "learning_rate": 9.391857615534226e-06,
+      "loss": 0.0286,
+      "step": 1513
+    },
+    {
+      "epoch": 0.40201805629314924,
+      "grad_norm": 0.028093132307861573,
+      "learning_rate": 9.39080784643752e-06,
+      "loss": 0.0006,
+      "step": 1514
+    },
+    {
+      "epoch": 0.402283590015932,
+      "grad_norm": 0.5084581122752261,
+      "learning_rate": 9.389757230844845e-06,
+      "loss": 0.0688,
+      "step": 1515
+    },
+    {
+      "epoch": 0.4025491237387148,
+      "grad_norm": 0.006193390834703029,
+      "learning_rate": 9.388705768958748e-06,
+      "loss": 0.0001,
+      "step": 1516
+    },
+    {
+      "epoch": 0.4028146574614976,
+      "grad_norm": 0.46353104622161806,
+      "learning_rate": 9.38765346098194e-06,
+      "loss": 0.0525,
+      "step": 1517
+    },
+    {
+      "epoch": 0.4030801911842804,
+      "grad_norm": 0.49616262621142954,
+      "learning_rate": 9.386600307117293e-06,
+      "loss": 0.0393,
+      "step": 1518
+    },
+    {
+      "epoch": 0.4033457249070632,
+      "grad_norm": 0.0815583003071057,
+      "learning_rate": 9.38554630756784e-06,
+      "loss": 0.0022,
+      "step": 1519
+    },
+    {
+      "epoch": 0.403611258629846,
+      "grad_norm": 0.48994190511403873,
+      "learning_rate": 9.384491462536783e-06,
+      "loss": 0.0354,
+      "step": 1520
+    },
+    {
+      "epoch": 0.4038767923526288,
+      "grad_norm": 0.11251221678106954,
+      "learning_rate": 9.383435772227481e-06,
+      "loss": 0.004,
+      "step": 1521
+    },
+    {
+      "epoch": 0.40414232607541156,
+      "grad_norm": 0.5397658046795061,
+      "learning_rate": 9.382379236843459e-06,
+      "loss": 0.0688,
+      "step": 1522
+    },
+    {
+      "epoch": 0.4044078597981944,
+      "grad_norm": 0.09364212325756312,
+      "learning_rate": 9.381321856588407e-06,
+      "loss": 0.0032,
+      "step": 1523
+    },
+    {
+      "epoch": 0.40467339352097714,
+      "grad_norm": 0.9230122195682332,
+      "learning_rate": 9.380263631666175e-06,
+      "loss": 0.0815,
+      "step": 1524
+    },
+    {
+      "epoch": 0.40493892724375996,
+      "grad_norm": 0.9314503563752995,
+      "learning_rate": 9.379204562280775e-06,
+      "loss": 0.0889,
+      "step": 1525
+    },
+    {
+      "epoch": 0.4052044609665427,
+      "grad_norm": 0.60796669683741,
+      "learning_rate": 9.378144648636382e-06,
+      "loss": 0.0688,
+      "step": 1526
+    },
+    {
+      "epoch": 0.40546999468932554,
+      "grad_norm": 0.5554197315940151,
+      "learning_rate": 9.377083890937337e-06,
+      "loss": 0.0576,
+      "step": 1527
+    },
+    {
+      "epoch": 0.40573552841210836,
+      "grad_norm": 0.12196493290704465,
+      "learning_rate": 9.37602228938814e-06,
+      "loss": 0.0028,
+      "step": 1528
+    },
+    {
+      "epoch": 0.4060010621348911,
+      "grad_norm": 0.3126139063656892,
+      "learning_rate": 9.374959844193456e-06,
+      "loss": 0.0354,
+      "step": 1529
+    },
+    {
+      "epoch": 0.40626659585767394,
+      "grad_norm": 0.011854084112567173,
+      "learning_rate": 9.373896555558113e-06,
+      "loss": 0.0003,
+      "step": 1530
+    },
+    {
+      "epoch": 0.4065321295804567,
+      "grad_norm": 3.841192190633747,
+      "learning_rate": 9.372832423687099e-06,
+      "loss": 0.2031,
+      "step": 1531
+    },
+    {
+      "epoch": 0.4067976633032395,
+      "grad_norm": 0.6503926893484147,
+      "learning_rate": 9.371767448785566e-06,
+      "loss": 0.0184,
+      "step": 1532
+    },
+    {
+      "epoch": 0.4070631970260223,
+      "grad_norm": 0.8960971685517782,
+      "learning_rate": 9.37070163105883e-06,
+      "loss": 0.0815,
+      "step": 1533
+    },
+    {
+      "epoch": 0.4073287307488051,
+      "grad_norm": 0.011994467119951549,
+      "learning_rate": 9.369634970712365e-06,
+      "loss": 0.0003,
+      "step": 1534
+    },
+    {
+      "epoch": 0.4075942644715879,
+      "grad_norm": 0.00513544623691197,
+      "learning_rate": 9.368567467951815e-06,
+      "loss": 0.0001,
+      "step": 1535
+    },
+    {
+      "epoch": 0.4078597981943707,
+      "grad_norm": 0.043827008510423704,
+      "learning_rate": 9.367499122982979e-06,
+      "loss": 0.0009,
+      "step": 1536
+    },
+    {
+      "epoch": 0.4081253319171535,
+      "grad_norm": 0.46712106544993076,
+      "learning_rate": 9.366429936011822e-06,
+      "loss": 0.0476,
+      "step": 1537
+    },
+    {
+      "epoch": 0.40839086563993626,
+      "grad_norm": 0.07965123738499298,
+      "learning_rate": 9.365359907244469e-06,
+      "loss": 0.0019,
+      "step": 1538
+    },
+    {
+      "epoch": 0.4086563993627191,
+      "grad_norm": 0.5002938341280985,
+      "learning_rate": 9.364289036887214e-06,
+      "loss": 0.0476,
+      "step": 1539
+    },
+    {
+      "epoch": 0.40892193308550184,
+      "grad_norm": 0.5932213624042697,
+      "learning_rate": 9.3632173251465e-06,
+      "loss": 0.0286,
+      "step": 1540
+    },
+    {
+      "epoch": 0.40918746680828466,
+      "grad_norm": 0.01294158924164034,
+      "learning_rate": 9.362144772228949e-06,
+      "loss": 0.0004,
+      "step": 1541
+    },
+    {
+      "epoch": 0.4094530005310674,
+      "grad_norm": 0.6159693393016078,
+      "learning_rate": 9.361071378341332e-06,
+      "loss": 0.0231,
+      "step": 1542
+    },
+    {
+      "epoch": 0.40971853425385024,
+      "grad_norm": 0.09689935660402307,
+      "learning_rate": 9.359997143690586e-06,
+      "loss": 0.0019,
+      "step": 1543
+    },
+    {
+      "epoch": 0.40998406797663306,
+      "grad_norm": 0.3554406490463387,
+      "learning_rate": 9.358922068483813e-06,
+      "loss": 0.0391,
+      "step": 1544
+    },
+    {
+      "epoch": 0.4102496016994158,
+      "grad_norm": 4.665948195663397,
+      "learning_rate": 9.357846152928275e-06,
+      "loss": 0.3535,
+      "step": 1545
+    },
+    {
+      "epoch": 0.41051513542219864,
+      "grad_norm": 0.6643692186727801,
+      "learning_rate": 9.356769397231391e-06,
+      "loss": 0.0206,
+      "step": 1546
+    },
+    {
+      "epoch": 0.4107806691449814,
+      "grad_norm": 0.3417631602945585,
+      "learning_rate": 9.355691801600754e-06,
+      "loss": 0.0354,
+      "step": 1547
+    },
+    {
+      "epoch": 0.4110462028677642,
+      "grad_norm": 0.04476331486140047,
+      "learning_rate": 9.354613366244108e-06,
+      "loss": 0.0006,
+      "step": 1548
+    },
+    {
+      "epoch": 0.411311736590547,
+      "grad_norm": 0.6284446073342773,
+      "learning_rate": 9.35353409136936e-06,
+      "loss": 0.0815,
+      "step": 1549
+    },
+    {
+      "epoch": 0.4115772703133298,
+      "grad_norm": 0.3822928656858541,
+      "learning_rate": 9.352453977184588e-06,
+      "loss": 0.0391,
+      "step": 1550
+    },
+    {
+      "epoch": 0.41184280403611256,
+      "grad_norm": 0.6046689508647543,
+      "learning_rate": 9.35137302389802e-06,
+      "loss": 0.1035,
+      "step": 1551
+    },
+    {
+      "epoch": 0.4121083377588954,
+      "grad_norm": 0.00728328734429902,
+      "learning_rate": 9.350291231718051e-06,
+      "loss": 0.0001,
+      "step": 1552
+    },
+    {
+      "epoch": 0.4123738714816782,
+      "grad_norm": 0.41563519871352395,
+      "learning_rate": 9.349208600853242e-06,
+      "loss": 0.0131,
+      "step": 1553
+    },
+    {
+      "epoch": 0.41263940520446096,
+      "grad_norm": 0.8926182157973549,
+      "learning_rate": 9.348125131512309e-06,
+      "loss": 0.0688,
+      "step": 1554
+    },
+    {
+      "epoch": 0.4129049389272438,
+      "grad_norm": 0.682749649537968,
+      "learning_rate": 9.34704082390413e-06,
+      "loss": 0.0889,
+      "step": 1555
+    },
+    {
+      "epoch": 0.41317047265002654,
+      "grad_norm": 0.004689924901466316,
+      "learning_rate": 9.34595567823775e-06,
+      "loss": 0.0001,
+      "step": 1556
+    },
+    {
+      "epoch": 0.41343600637280936,
+      "grad_norm": 0.6858612644220904,
+      "learning_rate": 9.344869694722372e-06,
+      "loss": 0.0889,
+      "step": 1557
+    },
+    {
+      "epoch": 0.4137015400955921,
+      "grad_norm": 0.4520811232153325,
+      "learning_rate": 9.343782873567358e-06,
+      "loss": 0.0391,
+      "step": 1558
+    },
+    {
+      "epoch": 0.41396707381837494,
+      "grad_norm": 0.3326484721722159,
+      "learning_rate": 9.342695214982236e-06,
+      "loss": 0.0354,
+      "step": 1559
+    },
+    {
+      "epoch": 0.4142326075411577,
+      "grad_norm": 0.3853243104104061,
+      "learning_rate": 9.341606719176695e-06,
+      "loss": 0.0432,
+      "step": 1560
+    },
+    {
+      "epoch": 0.4144981412639405,
+      "grad_norm": 0.01786226839288024,
+      "learning_rate": 9.340517386360584e-06,
+      "loss": 0.0004,
+      "step": 1561
+    },
+    {
+      "epoch": 0.41476367498672334,
+      "grad_norm": 0.014836540197822963,
+      "learning_rate": 9.33942721674391e-06,
+      "loss": 0.0003,
+      "step": 1562
+    },
+    {
+      "epoch": 0.4150292087095061,
+      "grad_norm": 0.6491829250530613,
+      "learning_rate": 9.33833621053685e-06,
+      "loss": 0.063,
+      "step": 1563
+    },
+    {
+      "epoch": 0.4152947424322889,
+      "grad_norm": 0.0008502111840714258,
+      "learning_rate": 9.337244367949733e-06,
+      "loss": 0.0,
+      "step": 1564
+    },
+    {
+      "epoch": 0.4155602761550717,
+      "grad_norm": 0.4693791006428915,
+      "learning_rate": 9.336151689193055e-06,
+      "loss": 0.0476,
+      "step": 1565
+    },
+    {
+      "epoch": 0.4158258098778545,
+      "grad_norm": 0.638806505853796,
+      "learning_rate": 9.335058174477472e-06,
+      "loss": 0.0525,
+      "step": 1566
+    },
+    {
+      "epoch": 0.41609134360063726,
+      "grad_norm": 3.7713130277807108,
+      "learning_rate": 9.333963824013798e-06,
+      "loss": 0.2246,
+      "step": 1567
+    },
+    {
+      "epoch": 0.4163568773234201,
+      "grad_norm": 0.004494212269452434,
+      "learning_rate": 9.332868638013016e-06,
+      "loss": 0.0001,
+      "step": 1568
+    },
+    {
+      "epoch": 0.41662241104620285,
+      "grad_norm": 0.20743155158952067,
+      "learning_rate": 9.331772616686262e-06,
+      "loss": 0.0051,
+      "step": 1569
+    },
+    {
+      "epoch": 0.41688794476898566,
+      "grad_norm": 0.6985128357597653,
+      "learning_rate": 9.330675760244834e-06,
+      "loss": 0.0432,
+      "step": 1570
+    },
+    {
+      "epoch": 0.4171534784917685,
+      "grad_norm": 0.9089238790009025,
+      "learning_rate": 9.329578068900195e-06,
+      "loss": 0.063,
+      "step": 1571
+    },
+    {
+      "epoch": 0.41741901221455124,
+      "grad_norm": 2.064976993600294,
+      "learning_rate": 9.328479542863968e-06,
+      "loss": 0.1455,
+      "step": 1572
+    },
+    {
+      "epoch": 0.41768454593733406,
+      "grad_norm": 0.4081315774110765,
+      "learning_rate": 9.327380182347937e-06,
+      "loss": 0.0354,
+      "step": 1573
+    },
+    {
+      "epoch": 0.4179500796601168,
+      "grad_norm": 0.34553567048811684,
+      "learning_rate": 9.32627998756404e-06,
+      "loss": 0.0354,
+      "step": 1574
+    },
+    {
+      "epoch": 0.41821561338289964,
+      "grad_norm": 0.74457240073377,
+      "learning_rate": 9.325178958724387e-06,
+      "loss": 0.0476,
+      "step": 1575
+    },
+    {
+      "epoch": 0.4184811471056824,
+      "grad_norm": 0.02077584538656709,
+      "learning_rate": 9.32407709604124e-06,
+      "loss": 0.0005,
+      "step": 1576
+    },
+    {
+      "epoch": 0.4187466808284652,
+      "grad_norm": 1.5760043575275893,
+      "learning_rate": 9.322974399727029e-06,
+      "loss": 0.1191,
+      "step": 1577
+    },
+    {
+      "epoch": 0.419012214551248,
+      "grad_norm": 0.08166506720345244,
+      "learning_rate": 9.321870869994336e-06,
+      "loss": 0.0013,
+      "step": 1578
+    },
+    {
+      "epoch": 0.4192777482740308,
+      "grad_norm": 0.9348396362017879,
+      "learning_rate": 9.320766507055912e-06,
+      "loss": 0.0688,
+      "step": 1579
+    },
+    {
+      "epoch": 0.4195432819968136,
+      "grad_norm": 0.27331732617626087,
+      "learning_rate": 9.319661311124664e-06,
+      "loss": 0.0317,
+      "step": 1580
+    },
+    {
+      "epoch": 0.4198088157195964,
+      "grad_norm": 5.750235838182297,
+      "learning_rate": 9.318555282413662e-06,
+      "loss": 0.3809,
+      "step": 1581
+    },
+    {
+      "epoch": 0.4200743494423792,
+      "grad_norm": 0.6368103985830037,
+      "learning_rate": 9.317448421136134e-06,
+      "loss": 0.0476,
+      "step": 1582
+    },
+    {
+      "epoch": 0.42033988316516196,
+      "grad_norm": 0.2827804927558854,
+      "learning_rate": 9.316340727505469e-06,
+      "loss": 0.0317,
+      "step": 1583
+    },
+    {
+      "epoch": 0.4206054168879448,
+      "grad_norm": 0.30307381764984753,
+      "learning_rate": 9.315232201735217e-06,
+      "loss": 0.0064,
+      "step": 1584
+    },
+    {
+      "epoch": 0.42087095061072755,
+      "grad_norm": 0.21252100392930384,
+      "learning_rate": 9.314122844039094e-06,
+      "loss": 0.0231,
+      "step": 1585
+    },
+    {
+      "epoch": 0.42113648433351036,
+      "grad_norm": 1.2522860606912725,
+      "learning_rate": 9.313012654630963e-06,
+      "loss": 0.1035,
+      "step": 1586
+    },
+    {
+      "epoch": 0.4214020180562931,
+      "grad_norm": 0.022875048396212985,
+      "learning_rate": 9.311901633724861e-06,
+      "loss": 0.0006,
+      "step": 1587
+    },
+    {
+      "epoch": 0.42166755177907594,
+      "grad_norm": 0.9338167878973481,
+      "learning_rate": 9.310789781534978e-06,
+      "loss": 0.0286,
+      "step": 1588
+    },
+    {
+      "epoch": 0.42193308550185876,
+      "grad_norm": 0.6025789069506893,
+      "learning_rate": 9.309677098275667e-06,
+      "loss": 0.0476,
+      "step": 1589
+    },
+    {
+      "epoch": 0.4221986192246415,
+      "grad_norm": 0.15935459440059752,
+      "learning_rate": 9.308563584161439e-06,
+      "loss": 0.0146,
+      "step": 1590
+    },
+    {
+      "epoch": 0.42246415294742434,
+      "grad_norm": 0.09990679440548342,
+      "learning_rate": 9.307449239406967e-06,
+      "loss": 0.0021,
+      "step": 1591
+    },
+    {
+      "epoch": 0.4227296866702071,
+      "grad_norm": 0.4308256760120562,
+      "learning_rate": 9.306334064227082e-06,
+      "loss": 0.0206,
+      "step": 1592
+    },
+    {
+      "epoch": 0.4229952203929899,
+      "grad_norm": 0.17557875823654812,
+      "learning_rate": 9.305218058836778e-06,
+      "loss": 0.0165,
+      "step": 1593
+    },
+    {
+      "epoch": 0.4232607541157727,
+      "grad_norm": 0.8469708090699332,
+      "learning_rate": 9.304101223451209e-06,
+      "loss": 0.0391,
+      "step": 1594
+    },
+    {
+      "epoch": 0.4235262878385555,
+      "grad_norm": 0.9742488100637656,
+      "learning_rate": 9.302983558285686e-06,
+      "loss": 0.1738,
+      "step": 1595
+    },
+    {
+      "epoch": 0.42379182156133827,
+      "grad_norm": 0.3807711068184408,
+      "learning_rate": 9.301865063555681e-06,
+      "loss": 0.0255,
+      "step": 1596
+    },
+    {
+      "epoch": 0.4240573552841211,
+      "grad_norm": 0.1865814006843838,
+      "learning_rate": 9.30074573947683e-06,
+      "loss": 0.0146,
+      "step": 1597
+    },
+    {
+      "epoch": 0.4243228890069039,
+      "grad_norm": 0.9795085871771619,
+      "learning_rate": 9.299625586264918e-06,
+      "loss": 0.1455,
+      "step": 1598
+    },
+    {
+      "epoch": 0.42458842272968667,
+      "grad_norm": 0.8224575743387007,
+      "learning_rate": 9.298504604135906e-06,
+      "loss": 0.1641,
+      "step": 1599
+    },
+    {
+      "epoch": 0.4248539564524695,
+      "grad_norm": 0.08459635905989443,
+      "learning_rate": 9.297382793305902e-06,
+      "loss": 0.0017,
+      "step": 1600
+    },
+    {
+      "epoch": 0.42511949017525225,
+      "grad_norm": 0.9247975416587106,
+      "learning_rate": 9.29626015399118e-06,
+      "loss": 0.0525,
+      "step": 1601
+    },
+    {
+      "epoch": 0.42538502389803506,
+      "grad_norm": 0.9030669338926891,
+      "learning_rate": 9.295136686408165e-06,
+      "loss": 0.0815,
+      "step": 1602
+    },
+    {
+      "epoch": 0.4256505576208178,
+      "grad_norm": 0.176488322331327,
+      "learning_rate": 9.294012390773457e-06,
+      "loss": 0.0184,
+      "step": 1603
+    },
+    {
+      "epoch": 0.42591609134360064,
+      "grad_norm": 0.7575042176122994,
+      "learning_rate": 9.292887267303803e-06,
+      "loss": 0.1113,
+      "step": 1604
+    },
+    {
+      "epoch": 0.4261816250663834,
+      "grad_norm": 2.8198240780302513,
+      "learning_rate": 9.291761316216115e-06,
+      "loss": 0.0889,
+      "step": 1605
+    },
+    {
+      "epoch": 0.4264471587891662,
+      "grad_norm": 0.80505821260424,
+      "learning_rate": 9.290634537727459e-06,
+      "loss": 0.0889,
+      "step": 1606
+    },
+    {
+      "epoch": 0.42671269251194904,
+      "grad_norm": 0.4004451010986326,
+      "learning_rate": 9.28950693205507e-06,
+      "loss": 0.0092,
+      "step": 1607
+    },
+    {
+      "epoch": 0.4269782262347318,
+      "grad_norm": 0.05174174947474149,
+      "learning_rate": 9.288378499416334e-06,
+      "loss": 0.0013,
+      "step": 1608
+    },
+    {
+      "epoch": 0.4272437599575146,
+      "grad_norm": 0.2673693573498579,
+      "learning_rate": 9.287249240028799e-06,
+      "loss": 0.0258,
+      "step": 1609
+    },
+    {
+      "epoch": 0.4275092936802974,
+      "grad_norm": 0.20068759525709715,
+      "learning_rate": 9.286119154110173e-06,
+      "loss": 0.0051,
+      "step": 1610
+    },
+    {
+      "epoch": 0.4277748274030802,
+      "grad_norm": 0.2052076895631866,
+      "learning_rate": 9.284988241878326e-06,
+      "loss": 0.0206,
+      "step": 1611
+    },
+    {
+      "epoch": 0.42804036112586297,
+      "grad_norm": 0.1826457677627445,
+      "learning_rate": 9.283856503551283e-06,
+      "loss": 0.0165,
+      "step": 1612
+    },
+    {
+      "epoch": 0.4283058948486458,
+      "grad_norm": 0.016102516202645443,
+      "learning_rate": 9.282723939347227e-06,
+      "loss": 0.0003,
+      "step": 1613
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.06556414219099314,
+      "learning_rate": 9.281590549484508e-06,
+      "loss": 0.0015,
+      "step": 1614
+    },
+    {
+      "epoch": 0.42883696229421137,
+      "grad_norm": 0.6492833856406688,
+      "learning_rate": 9.280456334181626e-06,
+      "loss": 0.1035,
+      "step": 1615
+    },
+    {
+      "epoch": 0.4291024960169942,
+      "grad_norm": 0.26676720638479,
+      "learning_rate": 9.279321293657248e-06,
+      "loss": 0.0255,
+      "step": 1616
+    },
+    {
+      "epoch": 0.42936802973977695,
+      "grad_norm": 1.0528365853705146,
+      "learning_rate": 9.278185428130194e-06,
+      "loss": 0.0476,
+      "step": 1617
+    },
+    {
+      "epoch": 0.42963356346255976,
+      "grad_norm": 0.300356633549314,
+      "learning_rate": 9.277048737819445e-06,
+      "loss": 0.0255,
+      "step": 1618
+    },
+    {
+      "epoch": 0.4298990971853425,
+      "grad_norm": 0.28681150202389294,
+      "learning_rate": 9.275911222944143e-06,
+      "loss": 0.0317,
+      "step": 1619
+    },
+    {
+      "epoch": 0.43016463090812534,
+      "grad_norm": 0.41351911300181254,
+      "learning_rate": 9.274772883723587e-06,
+      "loss": 0.0317,
+      "step": 1620
+    },
+    {
+      "epoch": 0.4304301646309081,
+      "grad_norm": 0.3435709971804024,
+      "learning_rate": 9.273633720377235e-06,
+      "loss": 0.0354,
+      "step": 1621
+    },
+    {
+      "epoch": 0.4306956983536909,
+      "grad_norm": 0.006280834272583653,
+      "learning_rate": 9.272493733124706e-06,
+      "loss": 0.0002,
+      "step": 1622
+    },
+    {
+      "epoch": 0.4309612320764737,
+      "grad_norm": 0.31013270569482426,
+      "learning_rate": 9.271352922185772e-06,
+      "loss": 0.0286,
+      "step": 1623
+    },
+    {
+      "epoch": 0.4312267657992565,
+      "grad_norm": 0.003193536264216083,
+      "learning_rate": 9.270211287780374e-06,
+      "loss": 0.0001,
+      "step": 1624
+    },
+    {
+      "epoch": 0.4314922995220393,
+      "grad_norm": 1.2572420728217566,
+      "learning_rate": 9.269068830128599e-06,
+      "loss": 0.0815,
+      "step": 1625
+    },
+    {
+      "epoch": 0.4317578332448221,
+      "grad_norm": 0.16424497737231764,
+      "learning_rate": 9.267925549450703e-06,
+      "loss": 0.0092,
+      "step": 1626
+    },
+    {
+      "epoch": 0.4320233669676049,
+      "grad_norm": 0.0051496845292951345,
+      "learning_rate": 9.266781445967098e-06,
+      "loss": 0.0001,
+      "step": 1627
+    },
+    {
+      "epoch": 0.43228890069038767,
+      "grad_norm": 0.4131524580991801,
+      "learning_rate": 9.26563651989835e-06,
+      "loss": 0.0255,
+      "step": 1628
+    },
+    {
+      "epoch": 0.4325544344131705,
+      "grad_norm": 0.005905247669898074,
+      "learning_rate": 9.264490771465191e-06,
+      "loss": 0.0001,
+      "step": 1629
+    },
+    {
+      "epoch": 0.43281996813595325,
+      "grad_norm": 0.11239806675648503,
+      "learning_rate": 9.263344200888507e-06,
+      "loss": 0.005,
+      "step": 1630
+    },
+    {
+      "epoch": 0.43308550185873607,
+      "grad_norm": 1.9861947240668674,
+      "learning_rate": 9.26219680838934e-06,
+      "loss": 0.1279,
+      "step": 1631
+    },
+    {
+      "epoch": 0.43335103558151883,
+      "grad_norm": 0.1940044319795506,
+      "learning_rate": 9.2610485941889e-06,
+      "loss": 0.0131,
+      "step": 1632
+    },
+    {
+      "epoch": 0.43361656930430165,
+      "grad_norm": 0.2272907921942844,
+      "learning_rate": 9.259899558508543e-06,
+      "loss": 0.0205,
+      "step": 1633
+    },
+    {
+      "epoch": 0.43388210302708446,
+      "grad_norm": 0.4199007338504671,
+      "learning_rate": 9.258749701569793e-06,
+      "loss": 0.0092,
+      "step": 1634
+    },
+    {
+      "epoch": 0.4341476367498672,
+      "grad_norm": 0.5230506299864369,
+      "learning_rate": 9.257599023594326e-06,
+      "loss": 0.0352,
+      "step": 1635
+    },
+    {
+      "epoch": 0.43441317047265005,
+      "grad_norm": 1.7928341347867058,
+      "learning_rate": 9.256447524803983e-06,
+      "loss": 0.2129,
+      "step": 1636
+    },
+    {
+      "epoch": 0.4346787041954328,
+      "grad_norm": 1.0771613409897511,
+      "learning_rate": 9.255295205420757e-06,
+      "loss": 0.1035,
+      "step": 1637
+    },
+    {
+      "epoch": 0.4349442379182156,
+      "grad_norm": 0.23823363260187153,
+      "learning_rate": 9.254142065666802e-06,
+      "loss": 0.0229,
+      "step": 1638
+    },
+    {
+      "epoch": 0.4352097716409984,
+      "grad_norm": 1.7503348124699936,
+      "learning_rate": 9.252988105764428e-06,
+      "loss": 0.2129,
+      "step": 1639
+    },
+    {
+      "epoch": 0.4354753053637812,
+      "grad_norm": 0.002717147032072083,
+      "learning_rate": 9.251833325936108e-06,
+      "loss": 0.0001,
+      "step": 1640
+    },
+    {
+      "epoch": 0.43574083908656397,
+      "grad_norm": 0.43583660633350435,
+      "learning_rate": 9.25067772640447e-06,
+      "loss": 0.0229,
+      "step": 1641
+    },
+    {
+      "epoch": 0.4360063728093468,
+      "grad_norm": 0.01550132302821912,
+      "learning_rate": 9.249521307392294e-06,
+      "loss": 0.0003,
+      "step": 1642
+    },
+    {
+      "epoch": 0.4362719065321296,
+      "grad_norm": 0.21256823089610236,
+      "learning_rate": 9.248364069122531e-06,
+      "loss": 0.0146,
+      "step": 1643
+    },
+    {
+      "epoch": 0.43653744025491237,
+      "grad_norm": 0.2227015636760905,
+      "learning_rate": 9.247206011818281e-06,
+      "loss": 0.0229,
+      "step": 1644
+    },
+    {
+      "epoch": 0.4368029739776952,
+      "grad_norm": 0.3421213770839244,
+      "learning_rate": 9.246047135702802e-06,
+      "loss": 0.0286,
+      "step": 1645
+    },
+    {
+      "epoch": 0.43706850770047795,
+      "grad_norm": 0.2175367153596953,
+      "learning_rate": 9.244887440999515e-06,
+      "loss": 0.0056,
+      "step": 1646
+    },
+    {
+      "epoch": 0.43733404142326077,
+      "grad_norm": 0.2561588514828441,
+      "learning_rate": 9.24372692793199e-06,
+      "loss": 0.0255,
+      "step": 1647
+    },
+    {
+      "epoch": 0.43759957514604353,
+      "grad_norm": 0.5116398168106552,
+      "learning_rate": 9.242565596723965e-06,
+      "loss": 0.0103,
+      "step": 1648
+    },
+    {
+      "epoch": 0.43786510886882635,
+      "grad_norm": 0.004941866451653809,
+      "learning_rate": 9.241403447599329e-06,
+      "loss": 0.0001,
+      "step": 1649
+    },
+    {
+      "epoch": 0.4381306425916091,
+      "grad_norm": 0.07087433757924969,
+      "learning_rate": 9.24024048078213e-06,
+      "loss": 0.0015,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4383961763143919,
+      "grad_norm": 0.33106687803013696,
+      "learning_rate": 9.239076696496577e-06,
+      "loss": 0.0354,
+      "step": 1651
+    },
+    {
+      "epoch": 0.43866171003717475,
+      "grad_norm": 0.0641176743240912,
+      "learning_rate": 9.237912094967031e-06,
+      "loss": 0.0012,
+      "step": 1652
+    },
+    {
+      "epoch": 0.4389272437599575,
+      "grad_norm": 0.30760602554788097,
+      "learning_rate": 9.236746676418016e-06,
+      "loss": 0.0255,
+      "step": 1653
+    },
+    {
+      "epoch": 0.4391927774827403,
+      "grad_norm": 0.1564547333112549,
+      "learning_rate": 9.235580441074208e-06,
+      "loss": 0.0129,
+      "step": 1654
+    },
+    {
+      "epoch": 0.4394583112055231,
+      "grad_norm": 1.3722315058379413,
+      "learning_rate": 9.234413389160446e-06,
+      "loss": 0.1191,
+      "step": 1655
+    },
+    {
+      "epoch": 0.4397238449283059,
+      "grad_norm": 1.2711619824610763,
+      "learning_rate": 9.233245520901723e-06,
+      "loss": 0.0815,
+      "step": 1656
+    },
+    {
+      "epoch": 0.43998937865108867,
+      "grad_norm": 0.8074581064566986,
+      "learning_rate": 9.23207683652319e-06,
+      "loss": 0.0317,
+      "step": 1657
+    },
+    {
+      "epoch": 0.4402549123738715,
+      "grad_norm": 0.18847549880372136,
+      "learning_rate": 9.230907336250155e-06,
+      "loss": 0.0165,
+      "step": 1658
+    },
+    {
+      "epoch": 0.44052044609665425,
+      "grad_norm": 0.3653202272308303,
+      "learning_rate": 9.229737020308084e-06,
+      "loss": 0.0092,
+      "step": 1659
+    },
+    {
+      "epoch": 0.44078597981943707,
+      "grad_norm": 5.255587093506221,
+      "learning_rate": 9.228565888922602e-06,
+      "loss": 0.3984,
+      "step": 1660
+    },
+    {
+      "epoch": 0.4410515135422199,
+      "grad_norm": 0.3264226155423392,
+      "learning_rate": 9.227393942319487e-06,
+      "loss": 0.0255,
+      "step": 1661
+    },
+    {
+      "epoch": 0.44131704726500265,
+      "grad_norm": 0.34863206865018187,
+      "learning_rate": 9.226221180724677e-06,
+      "loss": 0.0205,
+      "step": 1662
+    },
+    {
+      "epoch": 0.44158258098778547,
+      "grad_norm": 0.28323093506885416,
+      "learning_rate": 9.225047604364268e-06,
+      "loss": 0.0229,
+      "step": 1663
+    },
+    {
+      "epoch": 0.44184811471056823,
+      "grad_norm": 0.17631014929000818,
+      "learning_rate": 9.223873213464511e-06,
+      "loss": 0.0115,
+      "step": 1664
+    },
+    {
+      "epoch": 0.44211364843335105,
+      "grad_norm": 1.099027808353995,
+      "learning_rate": 9.222698008251814e-06,
+      "loss": 0.0432,
+      "step": 1665
+    },
+    {
+      "epoch": 0.4423791821561338,
+      "grad_norm": 0.10350955443306387,
+      "learning_rate": 9.221521988952742e-06,
+      "loss": 0.0021,
+      "step": 1666
+    },
+    {
+      "epoch": 0.44264471587891663,
+      "grad_norm": 0.5117982387799549,
+      "learning_rate": 9.22034515579402e-06,
+      "loss": 0.0317,
+      "step": 1667
+    },
+    {
+      "epoch": 0.4429102496016994,
+      "grad_norm": 0.17219084419699793,
+      "learning_rate": 9.219167509002526e-06,
+      "loss": 0.0092,
+      "step": 1668
+    },
+    {
+      "epoch": 0.4431757833244822,
+      "grad_norm": 1.6574596891900364,
+      "learning_rate": 9.217989048805297e-06,
+      "loss": 0.0957,
+      "step": 1669
+    },
+    {
+      "epoch": 0.443441317047265,
+      "grad_norm": 1.3273618450366584,
+      "learning_rate": 9.216809775429524e-06,
+      "loss": 0.1836,
+      "step": 1670
+    },
+    {
+      "epoch": 0.4437068507700478,
+      "grad_norm": 1.7850762788827483,
+      "learning_rate": 9.215629689102561e-06,
+      "loss": 0.1113,
+      "step": 1671
+    },
+    {
+      "epoch": 0.4439723844928306,
+      "grad_norm": 0.0017107214111774336,
+      "learning_rate": 9.214448790051913e-06,
+      "loss": 0.0,
+      "step": 1672
+    },
+    {
+      "epoch": 0.44423791821561337,
+      "grad_norm": 0.28213652008092643,
+      "learning_rate": 9.213267078505243e-06,
+      "loss": 0.0072,
+      "step": 1673
+    },
+    {
+      "epoch": 0.4445034519383962,
+      "grad_norm": 0.1888622989632113,
+      "learning_rate": 9.21208455469037e-06,
+      "loss": 0.0164,
+      "step": 1674
+    },
+    {
+      "epoch": 0.44476898566117895,
+      "grad_norm": 2.354678143727365,
+      "learning_rate": 9.210901218835273e-06,
+      "loss": 0.1367,
+      "step": 1675
+    },
+    {
+      "epoch": 0.44503451938396177,
+      "grad_norm": 0.1314357064594671,
+      "learning_rate": 9.209717071168083e-06,
+      "loss": 0.0092,
+      "step": 1676
+    },
+    {
+      "epoch": 0.44530005310674453,
+      "grad_norm": 2.3597528685967464,
+      "learning_rate": 9.208532111917092e-06,
+      "loss": 0.0957,
+      "step": 1677
+    },
+    {
+      "epoch": 0.44556558682952735,
+      "grad_norm": 0.015474073285452805,
+      "learning_rate": 9.207346341310744e-06,
+      "loss": 0.0003,
+      "step": 1678
+    },
+    {
+      "epoch": 0.44583112055231017,
+      "grad_norm": 1.151474305722994,
+      "learning_rate": 9.20615975957764e-06,
+      "loss": 0.0476,
+      "step": 1679
+    },
+    {
+      "epoch": 0.44609665427509293,
+      "grad_norm": 1.5488215701473362,
+      "learning_rate": 9.204972366946546e-06,
+      "loss": 0.2246,
+      "step": 1680
+    },
+    {
+      "epoch": 0.44636218799787575,
+      "grad_norm": 0.0018658514561948273,
+      "learning_rate": 9.203784163646368e-06,
+      "loss": 0.0,
+      "step": 1681
+    },
+    {
+      "epoch": 0.4466277217206585,
+      "grad_norm": 0.006199377876020424,
+      "learning_rate": 9.202595149906185e-06,
+      "loss": 0.0002,
+      "step": 1682
+    },
+    {
+      "epoch": 0.44689325544344133,
+      "grad_norm": 0.0004928732427429256,
+      "learning_rate": 9.201405325955222e-06,
+      "loss": 0.0,
+      "step": 1683
+    },
+    {
+      "epoch": 0.4471587891662241,
+      "grad_norm": 1.3116595608924697,
+      "learning_rate": 9.200214692022862e-06,
+      "loss": 0.1367,
+      "step": 1684
+    },
+    {
+      "epoch": 0.4474243228890069,
+      "grad_norm": 0.27607253424721656,
+      "learning_rate": 9.199023248338648e-06,
+      "loss": 0.0205,
+      "step": 1685
+    },
+    {
+      "epoch": 0.44768985661178967,
+      "grad_norm": 0.8567058953910064,
+      "learning_rate": 9.197830995132275e-06,
+      "loss": 0.0206,
+      "step": 1686
+    },
+    {
+      "epoch": 0.4479553903345725,
+      "grad_norm": 3.4163189556120814,
+      "learning_rate": 9.196637932633593e-06,
+      "loss": 0.1553,
+      "step": 1687
+    },
+    {
+      "epoch": 0.4482209240573553,
+      "grad_norm": 0.5354290085411026,
+      "learning_rate": 9.195444061072612e-06,
+      "loss": 0.0352,
+      "step": 1688
+    },
+    {
+      "epoch": 0.44848645778013807,
+      "grad_norm": 0.24446326873599736,
+      "learning_rate": 9.194249380679498e-06,
+      "loss": 0.0205,
+      "step": 1689
+    },
+    {
+      "epoch": 0.4487519915029209,
+      "grad_norm": 2.035095641007781,
+      "learning_rate": 9.19305389168457e-06,
+      "loss": 0.0688,
+      "step": 1690
+    },
+    {
+      "epoch": 0.44901752522570365,
+      "grad_norm": 0.0008471435364572292,
+      "learning_rate": 9.191857594318304e-06,
+      "loss": 0.0,
+      "step": 1691
+    },
+    {
+      "epoch": 0.44928305894848647,
+      "grad_norm": 3.829454052397207,
+      "learning_rate": 9.190660488811332e-06,
+      "loss": 0.5664,
+      "step": 1692
+    },
+    {
+      "epoch": 0.44954859267126923,
+      "grad_norm": 1.1644189483281735,
+      "learning_rate": 9.189462575394443e-06,
+      "loss": 0.1113,
+      "step": 1693
+    },
+    {
+      "epoch": 0.44981412639405205,
+      "grad_norm": 0.6076990567042442,
+      "learning_rate": 9.188263854298578e-06,
+      "loss": 0.0476,
+      "step": 1694
+    },
+    {
+      "epoch": 0.4500796601168348,
+      "grad_norm": 0.22821643461061178,
+      "learning_rate": 9.187064325754838e-06,
+      "loss": 0.0229,
+      "step": 1695
+    },
+    {
+      "epoch": 0.45034519383961763,
+      "grad_norm": 0.517920311097021,
+      "learning_rate": 9.185863989994479e-06,
+      "loss": 0.0432,
+      "step": 1696
+    },
+    {
+      "epoch": 0.45061072756240045,
+      "grad_norm": 1.6939168341092024,
+      "learning_rate": 9.18466284724891e-06,
+      "loss": 0.1035,
+      "step": 1697
+    },
+    {
+      "epoch": 0.4508762612851832,
+      "grad_norm": 0.2848964703780022,
+      "learning_rate": 9.183460897749697e-06,
+      "loss": 0.0229,
+      "step": 1698
+    },
+    {
+      "epoch": 0.45114179500796603,
+      "grad_norm": 0.7200198808392976,
+      "learning_rate": 9.182258141728563e-06,
+      "loss": 0.1455,
+      "step": 1699
+    },
+    {
+      "epoch": 0.4514073287307488,
+      "grad_norm": 0.6382191231542441,
+      "learning_rate": 9.181054579417384e-06,
+      "loss": 0.0525,
+      "step": 1700
+    },
+    {
+      "epoch": 0.4516728624535316,
+      "grad_norm": 0.7739371440317542,
+      "learning_rate": 9.179850211048193e-06,
+      "loss": 0.1279,
+      "step": 1701
+    },
+    {
+      "epoch": 0.45193839617631437,
+      "grad_norm": 0.05514363697004764,
+      "learning_rate": 9.178645036853177e-06,
+      "loss": 0.0017,
+      "step": 1702
+    },
+    {
+      "epoch": 0.4522039298990972,
+      "grad_norm": 0.7474045741866328,
+      "learning_rate": 9.177439057064684e-06,
+      "loss": 0.1279,
+      "step": 1703
+    },
+    {
+      "epoch": 0.45246946362187995,
+      "grad_norm": 0.25830195125237543,
+      "learning_rate": 9.176232271915206e-06,
+      "loss": 0.0255,
+      "step": 1704
+    },
+    {
+      "epoch": 0.45273499734466277,
+      "grad_norm": 0.08723998970372462,
+      "learning_rate": 9.175024681637401e-06,
+      "loss": 0.0031,
+      "step": 1705
+    },
+    {
+      "epoch": 0.4530005310674456,
+      "grad_norm": 0.7724681106268992,
+      "learning_rate": 9.173816286464077e-06,
+      "loss": 0.0576,
+      "step": 1706
+    },
+    {
+      "epoch": 0.45326606479022835,
+      "grad_norm": 0.2962473622329823,
+      "learning_rate": 9.172607086628197e-06,
+      "loss": 0.0286,
+      "step": 1707
+    },
+    {
+      "epoch": 0.45353159851301117,
+      "grad_norm": 0.5255810866203752,
+      "learning_rate": 9.171397082362885e-06,
+      "loss": 0.0286,
+      "step": 1708
+    },
+    {
+      "epoch": 0.45379713223579393,
+      "grad_norm": 0.527170789292327,
+      "learning_rate": 9.170186273901411e-06,
+      "loss": 0.0286,
+      "step": 1709
+    },
+    {
+      "epoch": 0.45406266595857675,
+      "grad_norm": 0.38467745191339814,
+      "learning_rate": 9.168974661477206e-06,
+      "loss": 0.0206,
+      "step": 1710
+    },
+    {
+      "epoch": 0.4543281996813595,
+      "grad_norm": 0.29047817939552817,
+      "learning_rate": 9.167762245323855e-06,
+      "loss": 0.0354,
+      "step": 1711
+    },
+    {
+      "epoch": 0.45459373340414233,
+      "grad_norm": 0.834962595717288,
+      "learning_rate": 9.166549025675098e-06,
+      "loss": 0.0576,
+      "step": 1712
+    },
+    {
+      "epoch": 0.4548592671269251,
+      "grad_norm": 0.8325104633372797,
+      "learning_rate": 9.165335002764828e-06,
+      "loss": 0.0889,
+      "step": 1713
+    },
+    {
+      "epoch": 0.4551248008497079,
+      "grad_norm": 0.29406829163167375,
+      "learning_rate": 9.164120176827097e-06,
+      "loss": 0.0317,
+      "step": 1714
+    },
+    {
+      "epoch": 0.45539033457249073,
+      "grad_norm": 0.8674015109840324,
+      "learning_rate": 9.162904548096103e-06,
+      "loss": 0.1191,
+      "step": 1715
+    },
+    {
+      "epoch": 0.4556558682952735,
+      "grad_norm": 0.2980629365474458,
+      "learning_rate": 9.16168811680621e-06,
+      "loss": 0.0354,
+      "step": 1716
+    },
+    {
+      "epoch": 0.4559214020180563,
+      "grad_norm": 0.236925754930231,
+      "learning_rate": 9.160470883191934e-06,
+      "loss": 0.0286,
+      "step": 1717
+    },
+    {
+      "epoch": 0.4561869357408391,
+      "grad_norm": 0.21775609808125498,
+      "learning_rate": 9.159252847487939e-06,
+      "loss": 0.0255,
+      "step": 1718
+    },
+    {
+      "epoch": 0.4564524694636219,
+      "grad_norm": 0.05357758947992938,
+      "learning_rate": 9.158034009929046e-06,
+      "loss": 0.0019,
+      "step": 1719
+    },
+    {
+      "epoch": 0.45671800318640465,
+      "grad_norm": 0.7295630149525557,
+      "learning_rate": 9.156814370750238e-06,
+      "loss": 0.0525,
+      "step": 1720
+    },
+    {
+      "epoch": 0.45698353690918747,
+      "grad_norm": 0.06168367786760944,
+      "learning_rate": 9.155593930186644e-06,
+      "loss": 0.0024,
+      "step": 1721
+    },
+    {
+      "epoch": 0.45724907063197023,
+      "grad_norm": 1.351984217373161,
+      "learning_rate": 9.15437268847355e-06,
+      "loss": 0.0815,
+      "step": 1722
+    },
+    {
+      "epoch": 0.45751460435475305,
+      "grad_norm": 0.10497376351174942,
+      "learning_rate": 9.153150645846399e-06,
+      "loss": 0.0036,
+      "step": 1723
+    },
+    {
+      "epoch": 0.45778013807753587,
+      "grad_norm": 0.18119641858710658,
+      "learning_rate": 9.151927802540783e-06,
+      "loss": 0.0045,
+      "step": 1724
+    },
+    {
+      "epoch": 0.45804567180031863,
+      "grad_norm": 0.9837813441272035,
+      "learning_rate": 9.150704158792456e-06,
+      "loss": 0.0688,
+      "step": 1725
+    },
+    {
+      "epoch": 0.45831120552310145,
+      "grad_norm": 0.8532756218779173,
+      "learning_rate": 9.14947971483732e-06,
+      "loss": 0.0576,
+      "step": 1726
+    },
+    {
+      "epoch": 0.4585767392458842,
+      "grad_norm": 0.6798270406461934,
+      "learning_rate": 9.148254470911434e-06,
+      "loss": 0.1113,
+      "step": 1727
+    },
+    {
+      "epoch": 0.45884227296866703,
+      "grad_norm": 0.33188951195468896,
+      "learning_rate": 9.14702842725101e-06,
+      "loss": 0.0317,
+      "step": 1728
+    },
+    {
+      "epoch": 0.4591078066914498,
+      "grad_norm": 0.9547380036001121,
+      "learning_rate": 9.145801584092415e-06,
+      "loss": 0.0688,
+      "step": 1729
+    },
+    {
+      "epoch": 0.4593733404142326,
+      "grad_norm": 1.121208129214251,
+      "learning_rate": 9.144573941672169e-06,
+      "loss": 0.0317,
+      "step": 1730
+    },
+    {
+      "epoch": 0.4596388741370154,
+      "grad_norm": 0.018532306902938646,
+      "learning_rate": 9.143345500226948e-06,
+      "loss": 0.0006,
+      "step": 1731
+    },
+    {
+      "epoch": 0.4599044078597982,
+      "grad_norm": 0.6486549833157388,
+      "learning_rate": 9.14211625999358e-06,
+      "loss": 0.1279,
+      "step": 1732
+    },
+    {
+      "epoch": 0.460169941582581,
+      "grad_norm": 0.2387504469567946,
+      "learning_rate": 9.14088622120905e-06,
+      "loss": 0.0082,
+      "step": 1733
+    },
+    {
+      "epoch": 0.4604354753053638,
+      "grad_norm": 2.1639487181238826,
+      "learning_rate": 9.139655384110494e-06,
+      "loss": 0.1641,
+      "step": 1734
+    },
+    {
+      "epoch": 0.4607010090281466,
+      "grad_norm": 0.6344741659386001,
+      "learning_rate": 9.1384237489352e-06,
+      "loss": 0.1113,
+      "step": 1735
+    },
+    {
+      "epoch": 0.46096654275092935,
+      "grad_norm": 0.6929723214303033,
+      "learning_rate": 9.13719131592062e-06,
+      "loss": 0.0889,
+      "step": 1736
+    },
+    {
+      "epoch": 0.46123207647371217,
+      "grad_norm": 0.38365443204668837,
+      "learning_rate": 9.135958085304345e-06,
+      "loss": 0.0432,
+      "step": 1737
+    },
+    {
+      "epoch": 0.46149761019649493,
+      "grad_norm": 0.4578266373131249,
+      "learning_rate": 9.134724057324132e-06,
+      "loss": 0.0476,
+      "step": 1738
+    },
+    {
+      "epoch": 0.46176314391927775,
+      "grad_norm": 0.3533084463962813,
+      "learning_rate": 9.133489232217882e-06,
+      "loss": 0.0103,
+      "step": 1739
+    },
+    {
+      "epoch": 0.4620286776420605,
+      "grad_norm": 0.3101849952354096,
+      "learning_rate": 9.13225361022366e-06,
+      "loss": 0.0391,
+      "step": 1740
+    },
+    {
+      "epoch": 0.46229421136484333,
+      "grad_norm": 1.3062414453586326,
+      "learning_rate": 9.131017191579679e-06,
+      "loss": 0.1035,
+      "step": 1741
+    },
+    {
+      "epoch": 0.46255974508762615,
+      "grad_norm": 0.28116099076278234,
+      "learning_rate": 9.129779976524303e-06,
+      "loss": 0.0354,
+      "step": 1742
+    },
+    {
+      "epoch": 0.4628252788104089,
+      "grad_norm": 0.31019250951818894,
+      "learning_rate": 9.128541965296051e-06,
+      "loss": 0.0391,
+      "step": 1743
+    },
+    {
+      "epoch": 0.46309081253319173,
+      "grad_norm": 1.032116360292696,
+      "learning_rate": 9.127303158133602e-06,
+      "loss": 0.0752,
+      "step": 1744
+    },
+    {
+      "epoch": 0.4633563462559745,
+      "grad_norm": 0.6052388800690558,
+      "learning_rate": 9.126063555275782e-06,
+      "loss": 0.0576,
+      "step": 1745
+    },
+    {
+      "epoch": 0.4636218799787573,
+      "grad_norm": 0.04310564970851126,
+      "learning_rate": 9.12482315696157e-06,
+      "loss": 0.0015,
+      "step": 1746
+    },
+    {
+      "epoch": 0.4638874137015401,
+      "grad_norm": 0.7380757773397368,
+      "learning_rate": 9.123581963430102e-06,
+      "loss": 0.063,
+      "step": 1747
+    },
+    {
+      "epoch": 0.4641529474243229,
+      "grad_norm": 0.8495163211366225,
+      "learning_rate": 9.122339974920664e-06,
+      "loss": 0.0688,
+      "step": 1748
+    },
+    {
+      "epoch": 0.46441848114710566,
+      "grad_norm": 0.17953121426797672,
+      "learning_rate": 9.121097191672697e-06,
+      "loss": 0.0082,
+      "step": 1749
+    },
+    {
+      "epoch": 0.4646840148698885,
+      "grad_norm": 0.6531547283200894,
+      "learning_rate": 9.119853613925796e-06,
+      "loss": 0.1367,
+      "step": 1750
+    },
+    {
+      "epoch": 0.4649495485926713,
+      "grad_norm": 1.3480427367731387,
+      "learning_rate": 9.118609241919707e-06,
+      "loss": 0.0688,
+      "step": 1751
+    },
+    {
+      "epoch": 0.46521508231545405,
+      "grad_norm": 0.5732194141376209,
+      "learning_rate": 9.11736407589433e-06,
+      "loss": 0.1191,
+      "step": 1752
+    },
+    {
+      "epoch": 0.46548061603823687,
+      "grad_norm": 0.615762764256231,
+      "learning_rate": 9.11611811608972e-06,
+      "loss": 0.1113,
+      "step": 1753
+    },
+    {
+      "epoch": 0.46574614976101963,
+      "grad_norm": 0.7635883205751591,
+      "learning_rate": 9.11487136274608e-06,
+      "loss": 0.0693,
+      "step": 1754
+    },
+    {
+      "epoch": 0.46601168348380245,
+      "grad_norm": 0.5733271972141396,
+      "learning_rate": 9.113623816103775e-06,
+      "loss": 0.0889,
+      "step": 1755
+    },
+    {
+      "epoch": 0.4662772172065852,
+      "grad_norm": 0.23390235066147239,
+      "learning_rate": 9.112375476403313e-06,
+      "loss": 0.0317,
+      "step": 1756
+    },
+    {
+      "epoch": 0.46654275092936803,
+      "grad_norm": 0.627686461965525,
+      "learning_rate": 9.111126343885359e-06,
+      "loss": 0.0752,
+      "step": 1757
+    },
+    {
+      "epoch": 0.4668082846521508,
+      "grad_norm": 1.1290944290448022,
+      "learning_rate": 9.109876418790733e-06,
+      "loss": 0.1113,
+      "step": 1758
+    },
+    {
+      "epoch": 0.4670738183749336,
+      "grad_norm": 0.5093788174002621,
+      "learning_rate": 9.108625701360403e-06,
+      "loss": 0.0957,
+      "step": 1759
+    },
+    {
+      "epoch": 0.46733935209771643,
+      "grad_norm": 0.5396898654005425,
+      "learning_rate": 9.107374191835496e-06,
+      "loss": 0.0752,
+      "step": 1760
+    },
+    {
+      "epoch": 0.4676048858204992,
+      "grad_norm": 0.31226491832256825,
+      "learning_rate": 9.106121890457289e-06,
+      "loss": 0.0432,
+      "step": 1761
+    },
+    {
+      "epoch": 0.467870419543282,
+      "grad_norm": 0.5714355485781667,
+      "learning_rate": 9.104868797467206e-06,
+      "loss": 0.0286,
+      "step": 1762
+    },
+    {
+      "epoch": 0.4681359532660648,
+      "grad_norm": 0.2957638989705724,
+      "learning_rate": 9.103614913106832e-06,
+      "loss": 0.0432,
+      "step": 1763
+    },
+    {
+      "epoch": 0.4684014869888476,
+      "grad_norm": 0.292588618562859,
+      "learning_rate": 9.1023602376179e-06,
+      "loss": 0.0432,
+      "step": 1764
+    },
+    {
+      "epoch": 0.46866702071163036,
+      "grad_norm": 0.33751587589467996,
+      "learning_rate": 9.101104771242297e-06,
+      "loss": 0.0525,
+      "step": 1765
+    },
+    {
+      "epoch": 0.4689325544344132,
+      "grad_norm": 0.394806774742472,
+      "learning_rate": 9.099848514222063e-06,
+      "loss": 0.0579,
+      "step": 1766
+    },
+    {
+      "epoch": 0.469198088157196,
+      "grad_norm": 0.30284921769372897,
+      "learning_rate": 9.098591466799388e-06,
+      "loss": 0.0479,
+      "step": 1767
+    },
+    {
+      "epoch": 0.46946362187997875,
+      "grad_norm": 0.06033720701590172,
+      "learning_rate": 9.097333629216618e-06,
+      "loss": 0.0019,
+      "step": 1768
+    },
+    {
+      "epoch": 0.46972915560276157,
+      "grad_norm": 0.6433340242166607,
+      "learning_rate": 9.096075001716245e-06,
+      "loss": 0.082,
+      "step": 1769
+    },
+    {
+      "epoch": 0.46999468932554433,
+      "grad_norm": 0.47188366036526186,
+      "learning_rate": 9.094815584540922e-06,
+      "loss": 0.063,
+      "step": 1770
+    },
+    {
+      "epoch": 0.47026022304832715,
+      "grad_norm": 0.48271838829532376,
+      "learning_rate": 9.093555377933449e-06,
+      "loss": 0.0957,
+      "step": 1771
+    },
+    {
+      "epoch": 0.4705257567711099,
+      "grad_norm": 0.47768086513551467,
+      "learning_rate": 9.092294382136778e-06,
+      "loss": 0.0889,
+      "step": 1772
+    },
+    {
+      "epoch": 0.47079129049389273,
+      "grad_norm": 0.28143225047541554,
+      "learning_rate": 9.091032597394012e-06,
+      "loss": 0.0432,
+      "step": 1773
+    },
+    {
+      "epoch": 0.4710568242166755,
+      "grad_norm": 0.3048274878976099,
+      "learning_rate": 9.089770023948413e-06,
+      "loss": 0.0132,
+      "step": 1774
+    },
+    {
+      "epoch": 0.4713223579394583,
+      "grad_norm": 0.4934823224302044,
+      "learning_rate": 9.088506662043387e-06,
+      "loss": 0.0688,
+      "step": 1775
+    },
+    {
+      "epoch": 0.47158789166224113,
+      "grad_norm": 0.4824507932270506,
+      "learning_rate": 9.087242511922497e-06,
+      "loss": 0.0889,
+      "step": 1776
+    },
+    {
+      "epoch": 0.4718534253850239,
+      "grad_norm": 2.06956586384691,
+      "learning_rate": 9.085977573829453e-06,
+      "loss": 0.1367,
+      "step": 1777
+    },
+    {
+      "epoch": 0.4721189591078067,
+      "grad_norm": 0.16024574451614348,
+      "learning_rate": 9.084711848008122e-06,
+      "loss": 0.0065,
+      "step": 1778
+    },
+    {
+      "epoch": 0.4723844928305895,
+      "grad_norm": 0.392020268202222,
+      "learning_rate": 9.083445334702525e-06,
+      "loss": 0.0317,
+      "step": 1779
+    },
+    {
+      "epoch": 0.4726500265533723,
+      "grad_norm": 0.2638470201441458,
+      "learning_rate": 9.082178034156823e-06,
+      "loss": 0.0391,
+      "step": 1780
+    },
+    {
+      "epoch": 0.47291556027615506,
+      "grad_norm": 0.35572403768500227,
+      "learning_rate": 9.080909946615344e-06,
+      "loss": 0.0525,
+      "step": 1781
+    },
+    {
+      "epoch": 0.4731810939989379,
+      "grad_norm": 0.1322872189240058,
+      "learning_rate": 9.079641072322555e-06,
+      "loss": 0.0045,
+      "step": 1782
+    },
+    {
+      "epoch": 0.47344662772172064,
+      "grad_norm": 0.4683993996526506,
+      "learning_rate": 9.078371411523085e-06,
+      "loss": 0.0889,
+      "step": 1783
+    },
+    {
+      "epoch": 0.47371216144450345,
+      "grad_norm": 0.05846428329547589,
+      "learning_rate": 9.077100964461705e-06,
+      "loss": 0.0017,
+      "step": 1784
+    },
+    {
+      "epoch": 0.4739776951672863,
+      "grad_norm": 0.4950840613938997,
+      "learning_rate": 9.075829731383342e-06,
+      "loss": 0.082,
+      "step": 1785
+    },
+    {
+      "epoch": 0.47424322889006904,
+      "grad_norm": 0.409989728413392,
+      "learning_rate": 9.07455771253308e-06,
+      "loss": 0.0525,
+      "step": 1786
+    },
+    {
+      "epoch": 0.47450876261285185,
+      "grad_norm": 0.540891443973615,
+      "learning_rate": 9.073284908156145e-06,
+      "loss": 0.0476,
+      "step": 1787
+    },
+    {
+      "epoch": 0.4747742963356346,
+      "grad_norm": 0.013696281162261802,
+      "learning_rate": 9.07201131849792e-06,
+      "loss": 0.0004,
+      "step": 1788
+    },
+    {
+      "epoch": 0.47503983005841743,
+      "grad_norm": 0.23943299662174408,
+      "learning_rate": 9.070736943803937e-06,
+      "loss": 0.0104,
+      "step": 1789
+    },
+    {
+      "epoch": 0.4753053637812002,
+      "grad_norm": 0.4895048851505992,
+      "learning_rate": 9.069461784319882e-06,
+      "loss": 0.0432,
+      "step": 1790
+    },
+    {
+      "epoch": 0.475570897503983,
+      "grad_norm": 0.10004712225856449,
+      "learning_rate": 9.068185840291588e-06,
+      "loss": 0.0025,
+      "step": 1791
+    },
+    {
+      "epoch": 0.4758364312267658,
+      "grad_norm": 0.159183886065426,
+      "learning_rate": 9.066909111965046e-06,
+      "loss": 0.0041,
+      "step": 1792
+    },
+    {
+      "epoch": 0.4761019649495486,
+      "grad_norm": 0.5751229116192297,
+      "learning_rate": 9.065631599586392e-06,
+      "loss": 0.0752,
+      "step": 1793
+    },
+    {
+      "epoch": 0.4763674986723314,
+      "grad_norm": 0.2988995897362944,
+      "learning_rate": 9.064353303401912e-06,
+      "loss": 0.0432,
+      "step": 1794
+    },
+    {
+      "epoch": 0.4766330323951142,
+      "grad_norm": 0.42552366259459906,
+      "learning_rate": 9.063074223658052e-06,
+      "loss": 0.0693,
+      "step": 1795
+    },
+    {
+      "epoch": 0.476898566117897,
+      "grad_norm": 1.4837334028705988,
+      "learning_rate": 9.061794360601401e-06,
+      "loss": 0.1113,
+      "step": 1796
+    },
+    {
+      "epoch": 0.47716409984067976,
+      "grad_norm": 0.6192662336206393,
+      "learning_rate": 9.060513714478701e-06,
+      "loss": 0.0752,
+      "step": 1797
+    },
+    {
+      "epoch": 0.4774296335634626,
+      "grad_norm": 0.321229007267076,
+      "learning_rate": 9.059232285536848e-06,
+      "loss": 0.0476,
+      "step": 1798
+    },
+    {
+      "epoch": 0.47769516728624534,
+      "grad_norm": 0.2890108799839838,
+      "learning_rate": 9.057950074022884e-06,
+      "loss": 0.0432,
+      "step": 1799
+    },
+    {
+      "epoch": 0.47796070100902815,
+      "grad_norm": 0.07149010496675735,
+      "learning_rate": 9.056667080184004e-06,
+      "loss": 0.0025,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4782262347318109,
+      "grad_norm": 0.39389094255878604,
+      "learning_rate": 9.055383304267555e-06,
+      "loss": 0.0258,
+      "step": 1801
+    },
+    {
+      "epoch": 0.47849176845459374,
+      "grad_norm": 0.33439676094633747,
+      "learning_rate": 9.054098746521033e-06,
+      "loss": 0.0479,
+      "step": 1802
+    },
+    {
+      "epoch": 0.47875730217737655,
+      "grad_norm": 0.4026126537880819,
+      "learning_rate": 9.052813407192087e-06,
+      "loss": 0.0525,
+      "step": 1803
+    },
+    {
+      "epoch": 0.4790228359001593,
+      "grad_norm": 0.13383065541461073,
+      "learning_rate": 9.051527286528514e-06,
+      "loss": 0.004,
+      "step": 1804
+    },
+    {
+      "epoch": 0.47928836962294213,
+      "grad_norm": 0.45913557201091415,
+      "learning_rate": 9.050240384778265e-06,
+      "loss": 0.0525,
+      "step": 1805
+    },
+    {
+      "epoch": 0.4795539033457249,
+      "grad_norm": 0.32449255612491684,
+      "learning_rate": 9.048952702189437e-06,
+      "loss": 0.0131,
+      "step": 1806
+    },
+    {
+      "epoch": 0.4798194370685077,
+      "grad_norm": 0.23468227496683847,
+      "learning_rate": 9.047664239010281e-06,
+      "loss": 0.0317,
+      "step": 1807
+    },
+    {
+      "epoch": 0.4800849707912905,
+      "grad_norm": 0.4626085648207346,
+      "learning_rate": 9.046374995489197e-06,
+      "loss": 0.0576,
+      "step": 1808
+    },
+    {
+      "epoch": 0.4803505045140733,
+      "grad_norm": 0.2436457613182265,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.0354,
+      "step": 1809
+    },
+    {
+      "epoch": 0.48061603823685606,
+      "grad_norm": 0.6559558699245994,
+      "learning_rate": 9.043794168415603e-06,
+      "loss": 0.0206,
+      "step": 1810
+    },
+    {
+      "epoch": 0.4808815719596389,
+      "grad_norm": 1.0632918592533553,
+      "learning_rate": 9.042502585360644e-06,
+      "loss": 0.0815,
+      "step": 1811
+    },
+    {
+      "epoch": 0.4811471056824217,
+      "grad_norm": 0.24478371121705456,
+      "learning_rate": 9.041210222958863e-06,
+      "loss": 0.0286,
+      "step": 1812
+    },
+    {
+      "epoch": 0.48141263940520446,
+      "grad_norm": 1.4444380543418034,
+      "learning_rate": 9.039917081459415e-06,
+      "loss": 0.0986,
+      "step": 1813
+    },
+    {
+      "epoch": 0.4816781731279873,
+      "grad_norm": 0.7045925020545539,
+      "learning_rate": 9.038623161111597e-06,
+      "loss": 0.1035,
+      "step": 1814
+    },
+    {
+      "epoch": 0.48194370685077004,
+      "grad_norm": 0.6197386448287133,
+      "learning_rate": 9.037328462164866e-06,
+      "loss": 0.1113,
+      "step": 1815
+    },
+    {
+      "epoch": 0.48220924057355286,
+      "grad_norm": 0.1857729974346812,
+      "learning_rate": 9.036032984868825e-06,
+      "loss": 0.0206,
+      "step": 1816
+    },
+    {
+      "epoch": 0.4824747742963356,
+      "grad_norm": 0.24507656947520706,
+      "learning_rate": 9.034736729473223e-06,
+      "loss": 0.0286,
+      "step": 1817
+    },
+    {
+      "epoch": 0.48274030801911844,
+      "grad_norm": 0.18927032479525302,
+      "learning_rate": 9.033439696227966e-06,
+      "loss": 0.0258,
+      "step": 1818
+    },
+    {
+      "epoch": 0.4830058417419012,
+      "grad_norm": 0.06063697384987017,
+      "learning_rate": 9.032141885383106e-06,
+      "loss": 0.0017,
+      "step": 1819
+    },
+    {
+      "epoch": 0.483271375464684,
+      "grad_norm": 0.22908240400686025,
+      "learning_rate": 9.030843297188842e-06,
+      "loss": 0.0286,
+      "step": 1820
+    },
+    {
+      "epoch": 0.48353690918746683,
+      "grad_norm": 0.029382370073099087,
+      "learning_rate": 9.029543931895533e-06,
+      "loss": 0.0007,
+      "step": 1821
+    },
+    {
+      "epoch": 0.4838024429102496,
+      "grad_norm": 4.222452192211459,
+      "learning_rate": 9.028243789753676e-06,
+      "loss": 0.2129,
+      "step": 1822
+    },
+    {
+      "epoch": 0.4840679766330324,
+      "grad_norm": 0.2060789874561621,
+      "learning_rate": 9.026942871013927e-06,
+      "loss": 0.0057,
+      "step": 1823
+    },
+    {
+      "epoch": 0.4843335103558152,
+      "grad_norm": 0.6395860695177051,
+      "learning_rate": 9.025641175927084e-06,
+      "loss": 0.0525,
+      "step": 1824
+    },
+    {
+      "epoch": 0.484599044078598,
+      "grad_norm": 0.6417176946489609,
+      "learning_rate": 9.024338704744102e-06,
+      "loss": 0.0476,
+      "step": 1825
+    },
+    {
+      "epoch": 0.48486457780138076,
+      "grad_norm": 0.1480376723553451,
+      "learning_rate": 9.02303545771608e-06,
+      "loss": 0.0184,
+      "step": 1826
+    },
+    {
+      "epoch": 0.4851301115241636,
+      "grad_norm": 1.0502927779842128,
+      "learning_rate": 9.021731435094267e-06,
+      "loss": 0.1367,
+      "step": 1827
+    },
+    {
+      "epoch": 0.48539564524694634,
+      "grad_norm": 0.4100889619332376,
+      "learning_rate": 9.020426637130069e-06,
+      "loss": 0.0184,
+      "step": 1828
+    },
+    {
+      "epoch": 0.48566117896972916,
+      "grad_norm": 1.263025388192655,
+      "learning_rate": 9.01912106407503e-06,
+      "loss": 0.1035,
+      "step": 1829
+    },
+    {
+      "epoch": 0.485926712692512,
+      "grad_norm": 0.6821914603573469,
+      "learning_rate": 9.01781471618085e-06,
+      "loss": 0.1641,
+      "step": 1830
+    },
+    {
+      "epoch": 0.48619224641529474,
+      "grad_norm": 0.1830730022933736,
+      "learning_rate": 9.016507593699379e-06,
+      "loss": 0.0231,
+      "step": 1831
+    },
+    {
+      "epoch": 0.48645778013807756,
+      "grad_norm": 0.1500688928562703,
+      "learning_rate": 9.015199696882614e-06,
+      "loss": 0.0184,
+      "step": 1832
+    },
+    {
+      "epoch": 0.4867233138608603,
+      "grad_norm": 0.19372694562652476,
+      "learning_rate": 9.013891025982704e-06,
+      "loss": 0.0231,
+      "step": 1833
+    },
+    {
+      "epoch": 0.48698884758364314,
+      "grad_norm": 0.7571475485333456,
+      "learning_rate": 9.01258158125194e-06,
+      "loss": 0.1279,
+      "step": 1834
+    },
+    {
+      "epoch": 0.4872543813064259,
+      "grad_norm": 0.6523517733514138,
+      "learning_rate": 9.011271362942774e-06,
+      "loss": 0.1279,
+      "step": 1835
+    },
+    {
+      "epoch": 0.4875199150292087,
+      "grad_norm": 0.2681899167683771,
+      "learning_rate": 9.009960371307798e-06,
+      "loss": 0.0286,
+      "step": 1836
+    },
+    {
+      "epoch": 0.4877854487519915,
+      "grad_norm": 0.7058528797659482,
+      "learning_rate": 9.008648606599755e-06,
+      "loss": 0.1455,
+      "step": 1837
+    },
+    {
+      "epoch": 0.4880509824747743,
+      "grad_norm": 0.12442443775061221,
+      "learning_rate": 9.007336069071537e-06,
+      "loss": 0.0039,
+      "step": 1838
+    },
+    {
+      "epoch": 0.4883165161975571,
+      "grad_norm": 0.20140444385699593,
+      "learning_rate": 9.00602275897619e-06,
+      "loss": 0.0286,
+      "step": 1839
+    },
+    {
+      "epoch": 0.4885820499203399,
+      "grad_norm": 0.5373289060422746,
+      "learning_rate": 9.0047086765669e-06,
+      "loss": 0.1113,
+      "step": 1840
+    },
+    {
+      "epoch": 0.4888475836431227,
+      "grad_norm": 0.705010285755384,
+      "learning_rate": 9.003393822097005e-06,
+      "loss": 0.0576,
+      "step": 1841
+    },
+    {
+      "epoch": 0.48911311736590546,
+      "grad_norm": 0.20375285524703615,
+      "learning_rate": 9.002078195820002e-06,
+      "loss": 0.0286,
+      "step": 1842
+    },
+    {
+      "epoch": 0.4893786510886883,
+      "grad_norm": 0.2973123126251027,
+      "learning_rate": 9.000761797989521e-06,
+      "loss": 0.0391,
+      "step": 1843
+    },
+    {
+      "epoch": 0.48964418481147104,
+      "grad_norm": 0.5261967824579288,
+      "learning_rate": 8.99944462885935e-06,
+      "loss": 0.0391,
+      "step": 1844
+    },
+    {
+      "epoch": 0.48990971853425386,
+      "grad_norm": 0.641071008582391,
+      "learning_rate": 8.998126688683423e-06,
+      "loss": 0.0957,
+      "step": 1845
+    },
+    {
+      "epoch": 0.4901752522570366,
+      "grad_norm": 2.8556542247006456,
+      "learning_rate": 8.996807977715824e-06,
+      "loss": 0.1836,
+      "step": 1846
+    },
+    {
+      "epoch": 0.49044078597981944,
+      "grad_norm": 0.21588840632781164,
+      "learning_rate": 8.995488496210785e-06,
+      "loss": 0.0317,
+      "step": 1847
+    },
+    {
+      "epoch": 0.49070631970260226,
+      "grad_norm": 0.5200930577904636,
+      "learning_rate": 8.994168244422687e-06,
+      "loss": 0.0957,
+      "step": 1848
+    },
+    {
+      "epoch": 0.490971853425385,
+      "grad_norm": 0.5946617117384666,
+      "learning_rate": 8.992847222606059e-06,
+      "loss": 0.0815,
+      "step": 1849
+    },
+    {
+      "epoch": 0.49123738714816784,
+      "grad_norm": 0.5329049241163124,
+      "learning_rate": 8.991525431015575e-06,
+      "loss": 0.1035,
+      "step": 1850
+    },
+    {
+      "epoch": 0.4915029208709506,
+      "grad_norm": 0.13421039859666778,
+      "learning_rate": 8.990202869906067e-06,
+      "loss": 0.0038,
+      "step": 1851
+    },
+    {
+      "epoch": 0.4917684545937334,
+      "grad_norm": 0.21205384524729126,
+      "learning_rate": 8.988879539532502e-06,
+      "loss": 0.0035,
+      "step": 1852
+    },
+    {
+      "epoch": 0.4920339883165162,
+      "grad_norm": 0.3961762407408791,
+      "learning_rate": 8.987555440150009e-06,
+      "loss": 0.0576,
+      "step": 1853
+    },
+    {
+      "epoch": 0.492299522039299,
+      "grad_norm": 0.5124722957876611,
+      "learning_rate": 8.986230572013856e-06,
+      "loss": 0.0688,
+      "step": 1854
+    },
+    {
+      "epoch": 0.49256505576208176,
+      "grad_norm": 0.31830392157316406,
+      "learning_rate": 8.98490493537946e-06,
+      "loss": 0.0476,
+      "step": 1855
+    },
+    {
+      "epoch": 0.4928305894848646,
+      "grad_norm": 0.4475338780438215,
+      "learning_rate": 8.983578530502391e-06,
+      "loss": 0.0688,
+      "step": 1856
+    },
+    {
+      "epoch": 0.4930961232076474,
+      "grad_norm": 0.5131627119793176,
+      "learning_rate": 8.982251357638366e-06,
+      "loss": 0.0432,
+      "step": 1857
+    },
+    {
+      "epoch": 0.49336165693043016,
+      "grad_norm": 0.03746990293964691,
+      "learning_rate": 8.980923417043241e-06,
+      "loss": 0.001,
+      "step": 1858
+    },
+    {
+      "epoch": 0.493627190653213,
+      "grad_norm": 0.37701188868972424,
+      "learning_rate": 8.979594708973034e-06,
+      "loss": 0.0576,
+      "step": 1859
+    },
+    {
+      "epoch": 0.49389272437599574,
+      "grad_norm": 0.3246063141561153,
+      "learning_rate": 8.978265233683903e-06,
+      "loss": 0.0476,
+      "step": 1860
+    },
+    {
+      "epoch": 0.49415825809877856,
+      "grad_norm": 0.2529206787160113,
+      "learning_rate": 8.976934991432153e-06,
+      "loss": 0.0117,
+      "step": 1861
+    },
+    {
+      "epoch": 0.4944237918215613,
+      "grad_norm": 0.4707397561581439,
+      "learning_rate": 8.97560398247424e-06,
+      "loss": 0.0479,
+      "step": 1862
+    },
+    {
+      "epoch": 0.49468932554434414,
+      "grad_norm": 0.35709255896991166,
+      "learning_rate": 8.974272207066767e-06,
+      "loss": 0.0525,
+      "step": 1863
+    },
+    {
+      "epoch": 0.4949548592671269,
+      "grad_norm": 0.35165939869522234,
+      "learning_rate": 8.972939665466485e-06,
+      "loss": 0.0525,
+      "step": 1864
+    },
+    {
+      "epoch": 0.4952203929899097,
+      "grad_norm": 0.44447870911726006,
+      "learning_rate": 8.971606357930293e-06,
+      "loss": 0.0576,
+      "step": 1865
+    },
+    {
+      "epoch": 0.49548592671269254,
+      "grad_norm": 0.01898523611296749,
+      "learning_rate": 8.970272284715234e-06,
+      "loss": 0.0005,
+      "step": 1866
+    },
+    {
+      "epoch": 0.4957514604354753,
+      "grad_norm": 0.4586388691281053,
+      "learning_rate": 8.968937446078505e-06,
+      "loss": 0.0688,
+      "step": 1867
+    },
+    {
+      "epoch": 0.4960169941582581,
+      "grad_norm": 0.35720555993159125,
+      "learning_rate": 8.967601842277445e-06,
+      "loss": 0.0525,
+      "step": 1868
+    },
+    {
+      "epoch": 0.4962825278810409,
+      "grad_norm": 0.36644379984601816,
+      "learning_rate": 8.966265473569545e-06,
+      "loss": 0.0525,
+      "step": 1869
+    },
+    {
+      "epoch": 0.4965480616038237,
+      "grad_norm": 0.011010370117200359,
+      "learning_rate": 8.964928340212436e-06,
+      "loss": 0.0003,
+      "step": 1870
+    },
+    {
+      "epoch": 0.49681359532660646,
+      "grad_norm": 0.5993583089811542,
+      "learning_rate": 8.963590442463907e-06,
+      "loss": 0.082,
+      "step": 1871
+    },
+    {
+      "epoch": 0.4970791290493893,
+      "grad_norm": 0.4181666595587938,
+      "learning_rate": 8.962251780581888e-06,
+      "loss": 0.0354,
+      "step": 1872
+    },
+    {
+      "epoch": 0.49734466277217204,
+      "grad_norm": 0.4221668256111686,
+      "learning_rate": 8.960912354824455e-06,
+      "loss": 0.0393,
+      "step": 1873
+    },
+    {
+      "epoch": 0.49761019649495486,
+      "grad_norm": 0.3977116516433932,
+      "learning_rate": 8.959572165449834e-06,
+      "loss": 0.0576,
+      "step": 1874
+    },
+    {
+      "epoch": 0.4978757302177377,
+      "grad_norm": 0.010735836440220152,
+      "learning_rate": 8.9582312127164e-06,
+      "loss": 0.0003,
+      "step": 1875
+    },
+    {
+      "epoch": 0.49814126394052044,
+      "grad_norm": 0.27808107918242736,
+      "learning_rate": 8.95688949688267e-06,
+      "loss": 0.0354,
+      "step": 1876
+    },
+    {
+      "epoch": 0.49840679766330326,
+      "grad_norm": 0.0016704416964712005,
+      "learning_rate": 8.955547018207311e-06,
+      "loss": 0.0,
+      "step": 1877
+    },
+    {
+      "epoch": 0.498672331386086,
+      "grad_norm": 0.5686188793893922,
+      "learning_rate": 8.954203776949141e-06,
+      "loss": 0.0476,
+      "step": 1878
+    },
+    {
+      "epoch": 0.49893786510886884,
+      "grad_norm": 0.39056710727137667,
+      "learning_rate": 8.952859773367117e-06,
+      "loss": 0.0525,
+      "step": 1879
+    },
+    {
+      "epoch": 0.4992033988316516,
+      "grad_norm": 0.512461057922991,
+      "learning_rate": 8.951515007720348e-06,
+      "loss": 0.0525,
+      "step": 1880
+    },
+    {
+      "epoch": 0.4994689325544344,
+      "grad_norm": 0.07275683481894793,
+      "learning_rate": 8.950169480268089e-06,
+      "loss": 0.0022,
+      "step": 1881
+    },
+    {
+      "epoch": 0.4997344662772172,
+      "grad_norm": 0.5906991505102981,
+      "learning_rate": 8.948823191269742e-06,
+      "loss": 0.063,
+      "step": 1882
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6168155161242778,
+      "learning_rate": 8.947476140984856e-06,
+      "loss": 0.0688,
+      "step": 1883
+    },
+    {
+      "epoch": 0.5002655337227828,
+      "grad_norm": 0.5204598871312067,
+      "learning_rate": 8.946128329673126e-06,
+      "loss": 0.0752,
+      "step": 1884
+    },
+    {
+      "epoch": 0.5005310674455656,
+      "grad_norm": 1.8571138477453206,
+      "learning_rate": 8.944779757594397e-06,
+      "loss": 0.2129,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5007966011683483,
+      "grad_norm": 0.49090367613182434,
+      "learning_rate": 8.943430425008653e-06,
+      "loss": 0.063,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5010621348911312,
+      "grad_norm": 0.2501068284298111,
+      "learning_rate": 8.942080332176031e-06,
+      "loss": 0.0286,
+      "step": 1887
+    },
+    {
+      "epoch": 0.501327668613914,
+      "grad_norm": 0.19234786774639814,
+      "learning_rate": 8.940729479356814e-06,
+      "loss": 0.0073,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5015932023366968,
+      "grad_norm": 1.5237924486409677,
+      "learning_rate": 8.93937786681143e-06,
+      "loss": 0.0889,
+      "step": 1889
+    },
+    {
+      "epoch": 0.5018587360594795,
+      "grad_norm": 0.5459929255095108,
+      "learning_rate": 8.938025494800454e-06,
+      "loss": 0.063,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5021242697822623,
+      "grad_norm": 0.25938046401016135,
+      "learning_rate": 8.936672363584607e-06,
+      "loss": 0.0354,
+      "step": 1891
+    },
+    {
+      "epoch": 0.5023898035050451,
+      "grad_norm": 0.32388794224012313,
+      "learning_rate": 8.935318473424756e-06,
+      "loss": 0.0432,
+      "step": 1892
+    },
+    {
+      "epoch": 0.502655337227828,
+      "grad_norm": 0.526908944612307,
+      "learning_rate": 8.933963824581919e-06,
+      "loss": 0.0693,
+      "step": 1893
+    },
+    {
+      "epoch": 0.5029208709506108,
+      "grad_norm": 0.27900639395149873,
+      "learning_rate": 8.932608417317253e-06,
+      "loss": 0.0092,
+      "step": 1894
+    },
+    {
+      "epoch": 0.5031864046733935,
+      "grad_norm": 0.3445447813472986,
+      "learning_rate": 8.931252251892061e-06,
+      "loss": 0.0354,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5034519383961763,
+      "grad_norm": 0.44933449962511285,
+      "learning_rate": 8.929895328567804e-06,
+      "loss": 0.0476,
+      "step": 1896
+    },
+    {
+      "epoch": 0.5037174721189591,
+      "grad_norm": 0.03865608508633661,
+      "learning_rate": 8.928537647606074e-06,
+      "loss": 0.001,
+      "step": 1897
+    },
+    {
+      "epoch": 0.5039830058417419,
+      "grad_norm": 0.20536476890220312,
+      "learning_rate": 8.92717920926862e-06,
+      "loss": 0.0258,
+      "step": 1898
+    },
+    {
+      "epoch": 0.5042485395645246,
+      "grad_norm": 0.2603825837594336,
+      "learning_rate": 8.92582001381733e-06,
+      "loss": 0.0317,
+      "step": 1899
+    },
+    {
+      "epoch": 0.5045140732873075,
+      "grad_norm": 0.664354927635177,
+      "learning_rate": 8.924460061514241e-06,
+      "loss": 0.1553,
+      "step": 1900
+    },
+    {
+      "epoch": 0.5047796070100903,
+      "grad_norm": 0.1417729671988901,
+      "learning_rate": 8.92309935262154e-06,
+      "loss": 0.0165,
+      "step": 1901
+    },
+    {
+      "epoch": 0.5050451407328731,
+      "grad_norm": 0.7573282597398386,
+      "learning_rate": 8.92173788740155e-06,
+      "loss": 0.0391,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5053106744556559,
+      "grad_norm": 0.6651934597634159,
+      "learning_rate": 8.920375666116749e-06,
+      "loss": 0.1367,
+      "step": 1903
+    },
+    {
+      "epoch": 0.5055762081784386,
+      "grad_norm": 0.630368828497892,
+      "learning_rate": 8.919012689029756e-06,
+      "loss": 0.1279,
+      "step": 1904
+    },
+    {
+      "epoch": 0.5058417419012214,
+      "grad_norm": 0.011800849809816667,
+      "learning_rate": 8.917648956403338e-06,
+      "loss": 0.0003,
+      "step": 1905
+    },
+    {
+      "epoch": 0.5061072756240043,
+      "grad_norm": 0.1916335585910718,
+      "learning_rate": 8.916284468500404e-06,
+      "loss": 0.0231,
+      "step": 1906
+    },
+    {
+      "epoch": 0.5063728093467871,
+      "grad_norm": 0.1990089576514881,
+      "learning_rate": 8.914919225584014e-06,
+      "loss": 0.0258,
+      "step": 1907
+    },
+    {
+      "epoch": 0.5066383430695698,
+      "grad_norm": 0.42788748968525386,
+      "learning_rate": 8.913553227917366e-06,
+      "loss": 0.0165,
+      "step": 1908
+    },
+    {
+      "epoch": 0.5069038767923526,
+      "grad_norm": 0.6043691154555516,
+      "learning_rate": 8.912186475763815e-06,
+      "loss": 0.1113,
+      "step": 1909
+    },
+    {
+      "epoch": 0.5071694105151354,
+      "grad_norm": 0.20100736554436516,
+      "learning_rate": 8.910818969386852e-06,
+      "loss": 0.0258,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5074349442379182,
+      "grad_norm": 0.2668002216539039,
+      "learning_rate": 8.909450709050115e-06,
+      "loss": 0.0067,
+      "step": 1911
+    },
+    {
+      "epoch": 0.5077004779607011,
+      "grad_norm": 0.25391486889680465,
+      "learning_rate": 8.908081695017388e-06,
+      "loss": 0.0286,
+      "step": 1912
+    },
+    {
+      "epoch": 0.5079660116834838,
+      "grad_norm": 0.10326740021263013,
+      "learning_rate": 8.906711927552601e-06,
+      "loss": 0.0022,
+      "step": 1913
+    },
+    {
+      "epoch": 0.5082315454062666,
+      "grad_norm": 0.7183992803108435,
+      "learning_rate": 8.905341406919832e-06,
+      "loss": 0.1455,
+      "step": 1914
+    },
+    {
+      "epoch": 0.5084970791290494,
+      "grad_norm": 0.24497092264223638,
+      "learning_rate": 8.903970133383297e-06,
+      "loss": 0.0286,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5087626128518322,
+      "grad_norm": 0.5045822122027199,
+      "learning_rate": 8.902598107207364e-06,
+      "loss": 0.0476,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5090281465746149,
+      "grad_norm": 0.21870695038083438,
+      "learning_rate": 8.901225328656543e-06,
+      "loss": 0.0258,
+      "step": 1917
+    },
+    {
+      "epoch": 0.5092936802973977,
+      "grad_norm": 0.006806934116766017,
+      "learning_rate": 8.89985179799549e-06,
+      "loss": 0.0002,
+      "step": 1918
+    },
+    {
+      "epoch": 0.5095592140201806,
+      "grad_norm": 0.06217815771623346,
+      "learning_rate": 8.898477515489003e-06,
+      "loss": 0.0013,
+      "step": 1919
+    },
+    {
+      "epoch": 0.5098247477429634,
+      "grad_norm": 0.15502279422567414,
+      "learning_rate": 8.897102481402031e-06,
+      "loss": 0.0036,
+      "step": 1920
+    },
+    {
+      "epoch": 0.5100902814657462,
+      "grad_norm": 0.2254494341774315,
+      "learning_rate": 8.895726695999663e-06,
+      "loss": 0.0286,
+      "step": 1921
+    },
+    {
+      "epoch": 0.5103558151885289,
+      "grad_norm": 4.048223290258338,
+      "learning_rate": 8.894350159547132e-06,
+      "loss": 0.2559,
+      "step": 1922
+    },
+    {
+      "epoch": 0.5106213489113117,
+      "grad_norm": 0.2224313022582158,
+      "learning_rate": 8.892972872309821e-06,
+      "loss": 0.0258,
+      "step": 1923
+    },
+    {
+      "epoch": 0.5108868826340945,
+      "grad_norm": 0.2649153994192695,
+      "learning_rate": 8.891594834553252e-06,
+      "loss": 0.0317,
+      "step": 1924
+    },
+    {
+      "epoch": 0.5111524163568774,
+      "grad_norm": 0.32381344159451636,
+      "learning_rate": 8.890216046543099e-06,
+      "loss": 0.0354,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5114179500796601,
+      "grad_norm": 0.18911967304759786,
+      "learning_rate": 8.888836508545172e-06,
+      "loss": 0.0258,
+      "step": 1926
+    },
+    {
+      "epoch": 0.5116834838024429,
+      "grad_norm": 0.15794943866572278,
+      "learning_rate": 8.887456220825429e-06,
+      "loss": 0.0206,
+      "step": 1927
+    },
+    {
+      "epoch": 0.5119490175252257,
+      "grad_norm": 0.8298787508237971,
+      "learning_rate": 8.886075183649976e-06,
+      "loss": 0.0889,
+      "step": 1928
+    },
+    {
+      "epoch": 0.5122145512480085,
+      "grad_norm": 0.14098262352173152,
+      "learning_rate": 8.884693397285058e-06,
+      "loss": 0.0045,
+      "step": 1929
+    },
+    {
+      "epoch": 0.5124800849707913,
+      "grad_norm": 0.5112261546913817,
+      "learning_rate": 8.883310861997073e-06,
+      "loss": 0.0391,
+      "step": 1930
+    },
+    {
+      "epoch": 0.512745618693574,
+      "grad_norm": 0.0710276758168645,
+      "learning_rate": 8.881927578052549e-06,
+      "loss": 0.0014,
+      "step": 1931
+    },
+    {
+      "epoch": 0.5130111524163569,
+      "grad_norm": 0.7156148636827978,
+      "learning_rate": 8.880543545718173e-06,
+      "loss": 0.1191,
+      "step": 1932
+    },
+    {
+      "epoch": 0.5132766861391397,
+      "grad_norm": 0.7498314694324334,
+      "learning_rate": 8.879158765260767e-06,
+      "loss": 0.1279,
+      "step": 1933
+    },
+    {
+      "epoch": 0.5135422198619225,
+      "grad_norm": 0.16462904141113316,
+      "learning_rate": 8.877773236947302e-06,
+      "loss": 0.0206,
+      "step": 1934
+    },
+    {
+      "epoch": 0.5138077535847052,
+      "grad_norm": 0.8601863622602933,
+      "learning_rate": 8.876386961044892e-06,
+      "loss": 0.0752,
+      "step": 1935
+    },
+    {
+      "epoch": 0.514073287307488,
+      "grad_norm": 0.6362870475286894,
+      "learning_rate": 8.874999937820793e-06,
+      "loss": 0.1279,
+      "step": 1936
+    },
+    {
+      "epoch": 0.5143388210302708,
+      "grad_norm": 0.1624202878388607,
+      "learning_rate": 8.873612167542408e-06,
+      "loss": 0.0206,
+      "step": 1937
+    },
+    {
+      "epoch": 0.5146043547530537,
+      "grad_norm": 0.15596786935599946,
+      "learning_rate": 8.87222365047728e-06,
+      "loss": 0.0206,
+      "step": 1938
+    },
+    {
+      "epoch": 0.5148698884758365,
+      "grad_norm": 1.0578529242706238,
+      "learning_rate": 8.870834386893101e-06,
+      "loss": 0.0815,
+      "step": 1939
+    },
+    {
+      "epoch": 0.5151354221986192,
+      "grad_norm": 0.32309548985685693,
+      "learning_rate": 8.869444377057706e-06,
+      "loss": 0.0105,
+      "step": 1940
+    },
+    {
+      "epoch": 0.515400955921402,
+      "grad_norm": 0.6279202187077859,
+      "learning_rate": 8.868053621239072e-06,
+      "loss": 0.1279,
+      "step": 1941
+    },
+    {
+      "epoch": 0.5156664896441848,
+      "grad_norm": 0.29499114303717844,
+      "learning_rate": 8.866662119705317e-06,
+      "loss": 0.0391,
+      "step": 1942
+    },
+    {
+      "epoch": 0.5159320233669676,
+      "grad_norm": 0.22730083243000213,
+      "learning_rate": 8.865269872724708e-06,
+      "loss": 0.0286,
+      "step": 1943
+    },
+    {
+      "epoch": 0.5161975570897503,
+      "grad_norm": 0.8126774785685559,
+      "learning_rate": 8.863876880565656e-06,
+      "loss": 0.0286,
+      "step": 1944
+    },
+    {
+      "epoch": 0.5164630908125332,
+      "grad_norm": 0.5189987862432444,
+      "learning_rate": 8.862483143496713e-06,
+      "loss": 0.0286,
+      "step": 1945
+    },
+    {
+      "epoch": 0.516728624535316,
+      "grad_norm": 0.3079203367541235,
+      "learning_rate": 8.86108866178657e-06,
+      "loss": 0.0354,
+      "step": 1946
+    },
+    {
+      "epoch": 0.5169941582580988,
+      "grad_norm": 0.30146182888816936,
+      "learning_rate": 8.859693435704073e-06,
+      "loss": 0.0391,
+      "step": 1947
+    },
+    {
+      "epoch": 0.5172596919808816,
+      "grad_norm": 0.18766405767921446,
+      "learning_rate": 8.858297465518201e-06,
+      "loss": 0.0258,
+      "step": 1948
+    },
+    {
+      "epoch": 0.5175252257036643,
+      "grad_norm": 0.48345252694955393,
+      "learning_rate": 8.856900751498083e-06,
+      "loss": 0.0476,
+      "step": 1949
+    },
+    {
+      "epoch": 0.5177907594264471,
+      "grad_norm": 0.28999764767448305,
+      "learning_rate": 8.855503293912987e-06,
+      "loss": 0.0354,
+      "step": 1950
+    },
+    {
+      "epoch": 0.51805629314923,
+      "grad_norm": 0.301468091100063,
+      "learning_rate": 8.854105093032328e-06,
+      "loss": 0.0354,
+      "step": 1951
+    },
+    {
+      "epoch": 0.5183218268720128,
+      "grad_norm": 0.22535390470370426,
+      "learning_rate": 8.852706149125662e-06,
+      "loss": 0.0286,
+      "step": 1952
+    },
+    {
+      "epoch": 0.5185873605947955,
+      "grad_norm": 0.2907903784108349,
+      "learning_rate": 8.851306462462689e-06,
+      "loss": 0.0317,
+      "step": 1953
+    },
+    {
+      "epoch": 0.5188528943175783,
+      "grad_norm": 0.6566340180383853,
+      "learning_rate": 8.849906033313251e-06,
+      "loss": 0.0525,
+      "step": 1954
+    },
+    {
+      "epoch": 0.5191184280403611,
+      "grad_norm": 0.3414604558845798,
+      "learning_rate": 8.848504861947335e-06,
+      "loss": 0.0354,
+      "step": 1955
+    },
+    {
+      "epoch": 0.5193839617631439,
+      "grad_norm": 0.15095251988649963,
+      "learning_rate": 8.847102948635074e-06,
+      "loss": 0.0184,
+      "step": 1956
+    },
+    {
+      "epoch": 0.5196494954859268,
+      "grad_norm": 0.8850305174499467,
+      "learning_rate": 8.845700293646733e-06,
+      "loss": 0.063,
+      "step": 1957
+    },
+    {
+      "epoch": 0.5199150292087095,
+      "grad_norm": 0.18094733973006374,
+      "learning_rate": 8.844296897252733e-06,
+      "loss": 0.0039,
+      "step": 1958
+    },
+    {
+      "epoch": 0.5201805629314923,
+      "grad_norm": 0.0024453736726058376,
+      "learning_rate": 8.84289275972363e-06,
+      "loss": 0.0001,
+      "step": 1959
+    },
+    {
+      "epoch": 0.5204460966542751,
+      "grad_norm": 4.799487348660996,
+      "learning_rate": 8.841487881330126e-06,
+      "loss": 0.2891,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5207116303770579,
+      "grad_norm": 0.00898819327183765,
+      "learning_rate": 8.840082262343064e-06,
+      "loss": 0.0002,
+      "step": 1961
+    },
+    {
+      "epoch": 0.5209771640998406,
+      "grad_norm": 0.11036124994527378,
+      "learning_rate": 8.838675903033432e-06,
+      "loss": 0.0117,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5212426978226234,
+      "grad_norm": 0.17194701223020112,
+      "learning_rate": 8.837268803672358e-06,
+      "loss": 0.0184,
+      "step": 1963
+    },
+    {
+      "epoch": 0.5215082315454063,
+      "grad_norm": 0.19802383357405098,
+      "learning_rate": 8.835860964531115e-06,
+      "loss": 0.0206,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5217737652681891,
+      "grad_norm": 0.11041893858226348,
+      "learning_rate": 8.834452385881121e-06,
+      "loss": 0.0117,
+      "step": 1965
+    },
+    {
+      "epoch": 0.5220392989909719,
+      "grad_norm": 0.6156090855902048,
+      "learning_rate": 8.833043067993926e-06,
+      "loss": 0.0233,
+      "step": 1966
+    },
+    {
+      "epoch": 0.5223048327137546,
+      "grad_norm": 0.14201533586884124,
+      "learning_rate": 8.831633011141237e-06,
+      "loss": 0.0146,
+      "step": 1967
+    },
+    {
+      "epoch": 0.5225703664365374,
+      "grad_norm": 0.09288200859356922,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.0092,
+      "step": 1968
+    },
+    {
+      "epoch": 0.5228359001593202,
+      "grad_norm": 0.9615613402120565,
+      "learning_rate": 8.828810681626876e-06,
+      "loss": 0.1455,
+      "step": 1969
+    },
+    {
+      "epoch": 0.5231014338821031,
+      "grad_norm": 1.3056914957728305,
+      "learning_rate": 8.827398409509318e-06,
+      "loss": 0.0476,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5233669676048858,
+      "grad_norm": 0.9123728894775064,
+      "learning_rate": 8.825985399514488e-06,
+      "loss": 0.1738,
+      "step": 1971
+    },
+    {
+      "epoch": 0.5236325013276686,
+      "grad_norm": 0.5116325107103561,
+      "learning_rate": 8.824571651914796e-06,
+      "loss": 0.0354,
+      "step": 1972
+    },
+    {
+      "epoch": 0.5238980350504514,
+      "grad_norm": 1.0931368397378376,
+      "learning_rate": 8.823157166982794e-06,
+      "loss": 0.0889,
+      "step": 1973
+    },
+    {
+      "epoch": 0.5241635687732342,
+      "grad_norm": 1.5038247940641931,
+      "learning_rate": 8.821741944991182e-06,
+      "loss": 0.1191,
+      "step": 1974
+    },
+    {
+      "epoch": 0.524429102496017,
+      "grad_norm": 2.1410167922056185,
+      "learning_rate": 8.820325986212796e-06,
+      "loss": 0.0957,
+      "step": 1975
+    },
+    {
+      "epoch": 0.5246946362187997,
+      "grad_norm": 1.4124271244242217,
+      "learning_rate": 8.818909290920617e-06,
+      "loss": 0.063,
+      "step": 1976
+    },
+    {
+      "epoch": 0.5249601699415826,
+      "grad_norm": 0.044914449959208055,
+      "learning_rate": 8.817491859387765e-06,
+      "loss": 0.001,
+      "step": 1977
+    },
+    {
+      "epoch": 0.5252257036643654,
+      "grad_norm": 0.7192623272990704,
+      "learning_rate": 8.816073691887506e-06,
+      "loss": 0.1455,
+      "step": 1978
+    },
+    {
+      "epoch": 0.5254912373871482,
+      "grad_norm": 0.4995422114884727,
+      "learning_rate": 8.814654788693246e-06,
+      "loss": 0.0391,
+      "step": 1979
+    },
+    {
+      "epoch": 0.5257567711099309,
+      "grad_norm": 0.2947556592020414,
+      "learning_rate": 8.813235150078532e-06,
+      "loss": 0.0286,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5260223048327137,
+      "grad_norm": 0.2865413479084726,
+      "learning_rate": 8.811814776317056e-06,
+      "loss": 0.0255,
+      "step": 1981
+    },
+    {
+      "epoch": 0.5262878385554965,
+      "grad_norm": 2.2341427898489776,
+      "learning_rate": 8.810393667682645e-06,
+      "loss": 0.0815,
+      "step": 1982
+    },
+    {
+      "epoch": 0.5265533722782794,
+      "grad_norm": 0.1433749071821932,
+      "learning_rate": 8.808971824449276e-06,
+      "loss": 0.0165,
+      "step": 1983
+    },
+    {
+      "epoch": 0.5268189060010622,
+      "grad_norm": 1.1251534229014521,
+      "learning_rate": 8.80754924689106e-06,
+      "loss": 0.0815,
+      "step": 1984
+    },
+    {
+      "epoch": 0.5270844397238449,
+      "grad_norm": 0.9046870543343221,
+      "learning_rate": 8.806125935282258e-06,
+      "loss": 0.0231,
+      "step": 1985
+    },
+    {
+      "epoch": 0.5273499734466277,
+      "grad_norm": 2.59520506547441,
+      "learning_rate": 8.804701889897261e-06,
+      "loss": 0.1641,
+      "step": 1986
+    },
+    {
+      "epoch": 0.5276155071694105,
+      "grad_norm": 1.1343653355959855,
+      "learning_rate": 8.803277111010613e-06,
+      "loss": 0.0688,
+      "step": 1987
+    },
+    {
+      "epoch": 0.5278810408921933,
+      "grad_norm": 1.0935958000724277,
+      "learning_rate": 8.801851598896996e-06,
+      "loss": 0.0293,
+      "step": 1988
+    },
+    {
+      "epoch": 0.528146574614976,
+      "grad_norm": 0.13718069103720956,
+      "learning_rate": 8.800425353831227e-06,
+      "loss": 0.0165,
+      "step": 1989
+    },
+    {
+      "epoch": 0.5284121083377589,
+      "grad_norm": 0.1382767894506031,
+      "learning_rate": 8.798998376088273e-06,
+      "loss": 0.0165,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5286776420605417,
+      "grad_norm": 0.6992769860574649,
+      "learning_rate": 8.797570665943236e-06,
+      "loss": 0.1553,
+      "step": 1991
+    },
+    {
+      "epoch": 0.5289431757833245,
+      "grad_norm": 0.7959070902740847,
+      "learning_rate": 8.796142223671362e-06,
+      "loss": 0.1279,
+      "step": 1992
+    },
+    {
+      "epoch": 0.5292087095061073,
+      "grad_norm": 0.003383773068786995,
+      "learning_rate": 8.79471304954804e-06,
+      "loss": 0.0001,
+      "step": 1993
+    },
+    {
+      "epoch": 0.52947424322889,
+      "grad_norm": 1.6911270754687433,
+      "learning_rate": 8.793283143848796e-06,
+      "loss": 0.1035,
+      "step": 1994
+    },
+    {
+      "epoch": 0.5297397769516728,
+      "grad_norm": 0.25896993348139163,
+      "learning_rate": 8.791852506849301e-06,
+      "loss": 0.0258,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5300053106744557,
+      "grad_norm": 0.1990072374637002,
+      "learning_rate": 8.790421138825362e-06,
+      "loss": 0.0229,
+      "step": 1996
+    },
+    {
+      "epoch": 0.5302708443972385,
+      "grad_norm": 0.12393114254257008,
+      "learning_rate": 8.788989040052932e-06,
+      "loss": 0.0146,
+      "step": 1997
+    },
+    {
+      "epoch": 0.5305363781200212,
+      "grad_norm": 0.1350871742151799,
+      "learning_rate": 8.787556210808101e-06,
+      "loss": 0.0165,
+      "step": 1998
+    },
+    {
+      "epoch": 0.530801911842804,
+      "grad_norm": 0.13938411153478383,
+      "learning_rate": 8.786122651367104e-06,
+      "loss": 0.0184,
+      "step": 1999
+    },
+    {
+      "epoch": 0.5310674455655868,
+      "grad_norm": 1.1760763074160063,
+      "learning_rate": 8.784688362006311e-06,
+      "loss": 0.1191,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 7532,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 381199319040000.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}