diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,70033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.2107882485150103,
+  "eval_steps": 500,
+  "global_step": 10000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00032107882485150104,
+      "grad_norm": 15.157858848571777,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 2.4811,
+      "step": 1
+    },
+    {
+      "epoch": 0.0006421576497030021,
+      "grad_norm": 15.802922248840332,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 2.3901,
+      "step": 2
+    },
+    {
+      "epoch": 0.0009632364745545031,
+      "grad_norm": 23.659759521484375,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 2.3375,
+      "step": 3
+    },
+    {
+      "epoch": 0.0012843152994060042,
+      "grad_norm": 27.097450256347656,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 2.3807,
+      "step": 4
+    },
+    {
+      "epoch": 0.0016053941242575051,
+      "grad_norm": 29.78221321105957,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 2.4191,
+      "step": 5
+    },
+    {
+      "epoch": 0.0019264729491090063,
+      "grad_norm": 33.7828254699707,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 2.5859,
+      "step": 6
+    },
+    {
+      "epoch": 0.002247551773960507,
+      "grad_norm": 43.44171142578125,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 2.8257,
+      "step": 7
+    },
+    {
+      "epoch": 0.0025686305988120084,
+      "grad_norm": 51.39571762084961,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 3.0337,
+      "step": 8
+    },
+    {
+      "epoch": 0.0028897094236635095,
+      "grad_norm": 54.60009002685547,
+      "learning_rate": 1.8e-06,
+      "loss": 3.1402,
+      "step": 9
+    },
+    {
+      "epoch": 0.0032107882485150102,
+      "grad_norm": 57.4810791015625,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 3.4821,
+      "step": 10
+    },
+    {
+      "epoch": 0.0035318670733665114,
+      "grad_norm": 61.211952209472656,
+      "learning_rate": 2.2e-06,
+      "loss": 3.4304,
+      "step": 11
+    },
+    {
+      "epoch": 0.0038529458982180125,
+      "grad_norm": 61.09406280517578,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 3.5862,
+      "step": 12
+    },
+    {
+      "epoch": 0.004174024723069513,
+      "grad_norm": 61.53642654418945,
+      "learning_rate": 2.6e-06,
+      "loss": 3.6255,
+      "step": 13
+    },
+    {
+      "epoch": 0.004495103547921014,
+      "grad_norm": 58.33311080932617,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 3.7706,
+      "step": 14
+    },
+    {
+      "epoch": 0.0048161823727725156,
+      "grad_norm": 57.1992301940918,
+      "learning_rate": 3e-06,
+      "loss": 3.7195,
+      "step": 15
+    },
+    {
+      "epoch": 0.005137261197624017,
+      "grad_norm": 55.674320220947266,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 3.6768,
+      "step": 16
+    },
+    {
+      "epoch": 0.005458340022475518,
+      "grad_norm": 54.121376037597656,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 3.9861,
+      "step": 17
+    },
+    {
+      "epoch": 0.005779418847327019,
+      "grad_norm": 52.377349853515625,
+      "learning_rate": 3.6e-06,
+      "loss": 3.8352,
+      "step": 18
+    },
+    {
+      "epoch": 0.00610049767217852,
+      "grad_norm": 50.22018051147461,
+      "learning_rate": 3.8e-06,
+      "loss": 3.9499,
+      "step": 19
+    },
+    {
+      "epoch": 0.0064215764970300205,
+      "grad_norm": 47.37348556518555,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 3.8506,
+      "step": 20
+    },
+    {
+      "epoch": 0.006742655321881522,
+      "grad_norm": 44.72926712036133,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 4.0215,
+      "step": 21
+    },
+    {
+      "epoch": 0.007063734146733023,
+      "grad_norm": 44.81752014160156,
+      "learning_rate": 4.4e-06,
+      "loss": 3.6998,
+      "step": 22
+    },
+    {
+      "epoch": 0.007384812971584524,
+      "grad_norm": 43.61850357055664,
+      "learning_rate": 4.6e-06,
+      "loss": 3.7522,
+      "step": 23
+    },
+    {
+      "epoch": 0.007705891796436025,
+      "grad_norm": 45.95828628540039,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 3.816,
+      "step": 24
+    },
+    {
+      "epoch": 0.008026970621287526,
+      "grad_norm": 49.80207443237305,
+      "learning_rate": 5e-06,
+      "loss": 4.0823,
+      "step": 25
+    },
+    {
+      "epoch": 0.008348049446139027,
+      "grad_norm": 51.24964141845703,
+      "learning_rate": 5.2e-06,
+      "loss": 3.9784,
+      "step": 26
+    },
+    {
+      "epoch": 0.008669128270990529,
+      "grad_norm": 54.43299102783203,
+      "learning_rate": 5.4e-06,
+      "loss": 3.9404,
+      "step": 27
+    },
+    {
+      "epoch": 0.008990207095842029,
+      "grad_norm": 55.02586364746094,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 3.917,
+      "step": 28
+    },
+    {
+      "epoch": 0.00931128592069353,
+      "grad_norm": 55.936119079589844,
+      "learning_rate": 5.8e-06,
+      "loss": 3.852,
+      "step": 29
+    },
+    {
+      "epoch": 0.009632364745545031,
+      "grad_norm": 53.62590789794922,
+      "learning_rate": 6e-06,
+      "loss": 3.8488,
+      "step": 30
+    },
+    {
+      "epoch": 0.009953443570396533,
+      "grad_norm": 50.38666534423828,
+      "learning_rate": 6.2e-06,
+      "loss": 3.3054,
+      "step": 31
+    },
+    {
+      "epoch": 0.010274522395248033,
+      "grad_norm": 49.8444709777832,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 3.6827,
+      "step": 32
+    },
+    {
+      "epoch": 0.010595601220099534,
+      "grad_norm": 46.34492111206055,
+      "learning_rate": 6.6e-06,
+      "loss": 3.7251,
+      "step": 33
+    },
+    {
+      "epoch": 0.010916680044951036,
+      "grad_norm": 45.03758239746094,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 3.6144,
+      "step": 34
+    },
+    {
+      "epoch": 0.011237758869802536,
+      "grad_norm": 40.92512893676758,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 3.3677,
+      "step": 35
+    },
+    {
+      "epoch": 0.011558837694654038,
+      "grad_norm": 36.829925537109375,
+      "learning_rate": 7.2e-06,
+      "loss": 3.1936,
+      "step": 36
+    },
+    {
+      "epoch": 0.011879916519505538,
+      "grad_norm": 40.46910858154297,
+      "learning_rate": 7.4e-06,
+      "loss": 3.4061,
+      "step": 37
+    },
+    {
+      "epoch": 0.01220099534435704,
+      "grad_norm": 54.956382751464844,
+      "learning_rate": 7.6e-06,
+      "loss": 3.5662,
+      "step": 38
+    },
+    {
+      "epoch": 0.01252207416920854,
+      "grad_norm": 57.516902923583984,
+      "learning_rate": 7.8e-06,
+      "loss": 3.0622,
+      "step": 39
+    },
+    {
+      "epoch": 0.012843152994060041,
+      "grad_norm": 61.08616256713867,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 3.5104,
+      "step": 40
+    },
+    {
+      "epoch": 0.013164231818911543,
+      "grad_norm": 58.04061508178711,
+      "learning_rate": 8.200000000000001e-06,
+      "loss": 2.9037,
+      "step": 41
+    },
+    {
+      "epoch": 0.013485310643763043,
+      "grad_norm": 56.112709045410156,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 2.8564,
+      "step": 42
+    },
+    {
+      "epoch": 0.013806389468614545,
+      "grad_norm": 50.86652755737305,
+      "learning_rate": 8.599999999999999e-06,
+      "loss": 3.1863,
+      "step": 43
+    },
+    {
+      "epoch": 0.014127468293466046,
+      "grad_norm": 41.73533630371094,
+      "learning_rate": 8.8e-06,
+      "loss": 3.0879,
+      "step": 44
+    },
+    {
+      "epoch": 0.014448547118317548,
+      "grad_norm": 36.46621322631836,
+      "learning_rate": 9e-06,
+      "loss": 3.1359,
+      "step": 45
+    },
+    {
+      "epoch": 0.014769625943169048,
+      "grad_norm": 28.671815872192383,
+      "learning_rate": 9.2e-06,
+      "loss": 2.9814,
+      "step": 46
+    },
+    {
+      "epoch": 0.01509070476802055,
+      "grad_norm": 24.950931549072266,
+      "learning_rate": 9.4e-06,
+      "loss": 2.8951,
+      "step": 47
+    },
+    {
+      "epoch": 0.01541178359287205,
+      "grad_norm": 27.621028900146484,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 2.6941,
+      "step": 48
+    },
+    {
+      "epoch": 0.01573286241772355,
+      "grad_norm": 27.553085327148438,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 2.5166,
+      "step": 49
+    },
+    {
+      "epoch": 0.016053941242575052,
+      "grad_norm": 26.491987228393555,
+      "learning_rate": 1e-05,
+      "loss": 2.6277,
+      "step": 50
+    },
+    {
+      "epoch": 0.016375020067426554,
+      "grad_norm": 5.4101104736328125,
+      "learning_rate": 1.02e-05,
+      "loss": 1.925,
+      "step": 51
+    },
+    {
+      "epoch": 0.016696098892278053,
+      "grad_norm": 4.891691207885742,
+      "learning_rate": 1.04e-05,
+      "loss": 1.8656,
+      "step": 52
+    },
+    {
+      "epoch": 0.017017177717129555,
+      "grad_norm": 5.655025482177734,
+      "learning_rate": 1.06e-05,
+      "loss": 1.66,
+      "step": 53
+    },
+    {
+      "epoch": 0.017338256541981057,
+      "grad_norm": 6.683502674102783,
+      "learning_rate": 1.08e-05,
+      "loss": 1.6638,
+      "step": 54
+    },
+    {
+      "epoch": 0.01765933536683256,
+      "grad_norm": 6.518195152282715,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 1.6495,
+      "step": 55
+    },
+    {
+      "epoch": 0.017980414191684058,
+      "grad_norm": 6.019643783569336,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 1.7143,
+      "step": 56
+    },
+    {
+      "epoch": 0.01830149301653556,
+      "grad_norm": 5.8600053787231445,
+      "learning_rate": 1.1400000000000001e-05,
+      "loss": 1.3576,
+      "step": 57
+    },
+    {
+      "epoch": 0.01862257184138706,
+      "grad_norm": 6.502221584320068,
+      "learning_rate": 1.16e-05,
+      "loss": 1.4601,
+      "step": 58
+    },
+    {
+      "epoch": 0.01894365066623856,
+      "grad_norm": 5.659090995788574,
+      "learning_rate": 1.18e-05,
+      "loss": 1.4309,
+      "step": 59
+    },
+    {
+      "epoch": 0.019264729491090062,
+      "grad_norm": 5.396006107330322,
+      "learning_rate": 1.2e-05,
+      "loss": 1.6735,
+      "step": 60
+    },
+    {
+      "epoch": 0.019585808315941564,
+      "grad_norm": 5.009734630584717,
+      "learning_rate": 1.22e-05,
+      "loss": 1.5251,
+      "step": 61
+    },
+    {
+      "epoch": 0.019906887140793066,
+      "grad_norm": 5.51364278793335,
+      "learning_rate": 1.24e-05,
+      "loss": 1.8121,
+      "step": 62
+    },
+    {
+      "epoch": 0.020227965965644565,
+      "grad_norm": 4.808631420135498,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 1.5659,
+      "step": 63
+    },
+    {
+      "epoch": 0.020549044790496067,
+      "grad_norm": 3.892801523208618,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 1.6308,
+      "step": 64
+    },
+    {
+      "epoch": 0.02087012361534757,
+      "grad_norm": 3.904027223587036,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 1.5642,
+      "step": 65
+    },
+    {
+      "epoch": 0.021191202440199067,
+      "grad_norm": 4.057150840759277,
+      "learning_rate": 1.32e-05,
+      "loss": 1.4334,
+      "step": 66
+    },
+    {
+      "epoch": 0.02151228126505057,
+      "grad_norm": 4.412062168121338,
+      "learning_rate": 1.3400000000000002e-05,
+      "loss": 1.5825,
+      "step": 67
+    },
+    {
+      "epoch": 0.02183336008990207,
+      "grad_norm": 3.494084358215332,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 1.5724,
+      "step": 68
+    },
+    {
+      "epoch": 0.022154438914753573,
+      "grad_norm": 4.483047008514404,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 1.688,
+      "step": 69
+    },
+    {
+      "epoch": 0.022475517739605072,
+      "grad_norm": 3.417116165161133,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 1.5698,
+      "step": 70
+    },
+    {
+      "epoch": 0.022796596564456574,
+      "grad_norm": 6.7289838790893555,
+      "learning_rate": 1.42e-05,
+      "loss": 1.6379,
+      "step": 71
+    },
+    {
+      "epoch": 0.023117675389308076,
+      "grad_norm": 3.190251111984253,
+      "learning_rate": 1.44e-05,
+      "loss": 1.4405,
+      "step": 72
+    },
+    {
+      "epoch": 0.023438754214159575,
+      "grad_norm": 3.8525373935699463,
+      "learning_rate": 1.4599999999999999e-05,
+      "loss": 1.5232,
+      "step": 73
+    },
+    {
+      "epoch": 0.023759833039011077,
+      "grad_norm": 3.5374369621276855,
+      "learning_rate": 1.48e-05,
+      "loss": 1.7238,
+      "step": 74
+    },
+    {
+      "epoch": 0.02408091186386258,
+      "grad_norm": 5.770689964294434,
+      "learning_rate": 1.5e-05,
+      "loss": 1.5226,
+      "step": 75
+    },
+    {
+      "epoch": 0.02440199068871408,
+      "grad_norm": 4.280745506286621,
+      "learning_rate": 1.52e-05,
+      "loss": 1.6567,
+      "step": 76
+    },
+    {
+      "epoch": 0.02472306951356558,
+      "grad_norm": 3.3552117347717285,
+      "learning_rate": 1.54e-05,
+      "loss": 1.4622,
+      "step": 77
+    },
+    {
+      "epoch": 0.02504414833841708,
+      "grad_norm": 4.613933563232422,
+      "learning_rate": 1.56e-05,
+      "loss": 1.612,
+      "step": 78
+    },
+    {
+      "epoch": 0.025365227163268583,
+      "grad_norm": 4.299830913543701,
+      "learning_rate": 1.58e-05,
+      "loss": 1.7098,
+      "step": 79
+    },
+    {
+      "epoch": 0.025686305988120082,
+      "grad_norm": 3.547947406768799,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.4593,
+      "step": 80
+    },
+    {
+      "epoch": 0.026007384812971584,
+      "grad_norm": 3.5630974769592285,
+      "learning_rate": 1.62e-05,
+      "loss": 1.6422,
+      "step": 81
+    },
+    {
+      "epoch": 0.026328463637823086,
+      "grad_norm": 3.984318256378174,
+      "learning_rate": 1.6400000000000002e-05,
+      "loss": 1.5161,
+      "step": 82
+    },
+    {
+      "epoch": 0.026649542462674588,
+      "grad_norm": 4.502562046051025,
+      "learning_rate": 1.66e-05,
+      "loss": 1.7414,
+      "step": 83
+    },
+    {
+      "epoch": 0.026970621287526086,
+      "grad_norm": 3.2806875705718994,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 1.4668,
+      "step": 84
+    },
+    {
+      "epoch": 0.02729170011237759,
+      "grad_norm": 3.762796401977539,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 1.6335,
+      "step": 85
+    },
+    {
+      "epoch": 0.02761277893722909,
+      "grad_norm": 3.4556496143341064,
+      "learning_rate": 1.7199999999999998e-05,
+      "loss": 1.6127,
+      "step": 86
+    },
+    {
+      "epoch": 0.027933857762080593,
+      "grad_norm": 3.3840537071228027,
+      "learning_rate": 1.74e-05,
+      "loss": 1.4251,
+      "step": 87
+    },
+    {
+      "epoch": 0.02825493658693209,
+      "grad_norm": 3.363680601119995,
+      "learning_rate": 1.76e-05,
+      "loss": 1.6259,
+      "step": 88
+    },
+    {
+      "epoch": 0.028576015411783593,
+      "grad_norm": 3.2910609245300293,
+      "learning_rate": 1.78e-05,
+      "loss": 1.5546,
+      "step": 89
+    },
+    {
+      "epoch": 0.028897094236635095,
+      "grad_norm": 2.9161438941955566,
+      "learning_rate": 1.8e-05,
+      "loss": 1.2032,
+      "step": 90
+    },
+    {
+      "epoch": 0.029218173061486594,
+      "grad_norm": 3.4742963314056396,
+      "learning_rate": 1.8200000000000002e-05,
+      "loss": 1.5664,
+      "step": 91
+    },
+    {
+      "epoch": 0.029539251886338096,
+      "grad_norm": 3.563717842102051,
+      "learning_rate": 1.84e-05,
+      "loss": 1.4364,
+      "step": 92
+    },
+    {
+      "epoch": 0.029860330711189598,
+      "grad_norm": 2.6174569129943848,
+      "learning_rate": 1.86e-05,
+      "loss": 1.4534,
+      "step": 93
+    },
+    {
+      "epoch": 0.0301814095360411,
+      "grad_norm": 5.107251167297363,
+      "learning_rate": 1.88e-05,
+      "loss": 1.4267,
+      "step": 94
+    },
+    {
+      "epoch": 0.0305024883608926,
+      "grad_norm": 3.532844066619873,
+      "learning_rate": 1.9e-05,
+      "loss": 1.5197,
+      "step": 95
+    },
+    {
+      "epoch": 0.0308235671857441,
+      "grad_norm": 3.5595579147338867,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 1.3465,
+      "step": 96
+    },
+    {
+      "epoch": 0.031144646010595602,
+      "grad_norm": 3.7405896186828613,
+      "learning_rate": 1.94e-05,
+      "loss": 1.4418,
+      "step": 97
+    },
+    {
+      "epoch": 0.0314657248354471,
+      "grad_norm": 3.423434257507324,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 1.3589,
+      "step": 98
+    },
+    {
+      "epoch": 0.0317868036602986,
+      "grad_norm": 7.2687602043151855,
+      "learning_rate": 1.9800000000000004e-05,
+      "loss": 1.5237,
+      "step": 99
+    },
+    {
+      "epoch": 0.032107882485150105,
+      "grad_norm": 9.787822723388672,
+      "learning_rate": 2e-05,
+      "loss": 1.3776,
+      "step": 100
+    },
+    {
+      "epoch": 0.03242896131000161,
+      "grad_norm": 3.462364435195923,
+      "learning_rate": 2.0200000000000003e-05,
+      "loss": 1.921,
+      "step": 101
+    },
+    {
+      "epoch": 0.03275004013485311,
+      "grad_norm": 3.6257283687591553,
+      "learning_rate": 2.04e-05,
+      "loss": 1.7884,
+      "step": 102
+    },
+    {
+      "epoch": 0.033071118959704604,
+      "grad_norm": 3.9720280170440674,
+      "learning_rate": 2.06e-05,
+      "loss": 1.5258,
+      "step": 103
+    },
+    {
+      "epoch": 0.033392197784556106,
+      "grad_norm": 3.6847167015075684,
+      "learning_rate": 2.08e-05,
+      "loss": 1.2969,
+      "step": 104
+    },
+    {
+      "epoch": 0.03371327660940761,
+      "grad_norm": 4.045772075653076,
+      "learning_rate": 2.1e-05,
+      "loss": 1.4464,
+      "step": 105
+    },
+    {
+      "epoch": 0.03403435543425911,
+      "grad_norm": 3.264374017715454,
+      "learning_rate": 2.12e-05,
+      "loss": 1.2651,
+      "step": 106
+    },
+    {
+      "epoch": 0.03435543425911061,
+      "grad_norm": 2.960892677307129,
+      "learning_rate": 2.1400000000000002e-05,
+      "loss": 1.1481,
+      "step": 107
+    },
+    {
+      "epoch": 0.034676513083962114,
+      "grad_norm": 2.9885120391845703,
+      "learning_rate": 2.16e-05,
+      "loss": 1.367,
+      "step": 108
+    },
+    {
+      "epoch": 0.034997591908813616,
+      "grad_norm": 3.1829378604888916,
+      "learning_rate": 2.18e-05,
+      "loss": 1.3112,
+      "step": 109
+    },
+    {
+      "epoch": 0.03531867073366512,
+      "grad_norm": 2.956650733947754,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 1.4058,
+      "step": 110
+    },
+    {
+      "epoch": 0.03563974955851661,
+      "grad_norm": 2.93878436088562,
+      "learning_rate": 2.22e-05,
+      "loss": 1.6511,
+      "step": 111
+    },
+    {
+      "epoch": 0.035960828383368115,
+      "grad_norm": 4.111947059631348,
+      "learning_rate": 2.2400000000000002e-05,
+      "loss": 1.2943,
+      "step": 112
+    },
+    {
+      "epoch": 0.03628190720821962,
+      "grad_norm": 3.833181142807007,
+      "learning_rate": 2.26e-05,
+      "loss": 1.3733,
+      "step": 113
+    },
+    {
+      "epoch": 0.03660298603307112,
+      "grad_norm": 2.5650758743286133,
+      "learning_rate": 2.2800000000000002e-05,
+      "loss": 1.4837,
+      "step": 114
+    },
+    {
+      "epoch": 0.03692406485792262,
+      "grad_norm": 2.7790398597717285,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 1.3758,
+      "step": 115
+    },
+    {
+      "epoch": 0.03724514368277412,
+      "grad_norm": 2.0799965858459473,
+      "learning_rate": 2.32e-05,
+      "loss": 1.2325,
+      "step": 116
+    },
+    {
+      "epoch": 0.037566222507625625,
+      "grad_norm": 2.8642420768737793,
+      "learning_rate": 2.3400000000000003e-05,
+      "loss": 1.422,
+      "step": 117
+    },
+    {
+      "epoch": 0.03788730133247712,
+      "grad_norm": 2.840057373046875,
+      "learning_rate": 2.36e-05,
+      "loss": 1.3231,
+      "step": 118
+    },
+    {
+      "epoch": 0.03820838015732862,
+      "grad_norm": 2.1735825538635254,
+      "learning_rate": 2.38e-05,
+      "loss": 1.441,
+      "step": 119
+    },
+    {
+      "epoch": 0.038529458982180124,
+      "grad_norm": 2.365070343017578,
+      "learning_rate": 2.4e-05,
+      "loss": 1.3248,
+      "step": 120
+    },
+    {
+      "epoch": 0.038850537807031627,
+      "grad_norm": 2.179810047149658,
+      "learning_rate": 2.4200000000000002e-05,
+      "loss": 1.4982,
+      "step": 121
+    },
+    {
+      "epoch": 0.03917161663188313,
+      "grad_norm": 2.4042208194732666,
+      "learning_rate": 2.44e-05,
+      "loss": 1.6785,
+      "step": 122
+    },
+    {
+      "epoch": 0.03949269545673463,
+      "grad_norm": 2.441134452819824,
+      "learning_rate": 2.46e-05,
+      "loss": 1.4426,
+      "step": 123
+    },
+    {
+      "epoch": 0.03981377428158613,
+      "grad_norm": 3.4666857719421387,
+      "learning_rate": 2.48e-05,
+      "loss": 1.4485,
+      "step": 124
+    },
+    {
+      "epoch": 0.04013485310643763,
+      "grad_norm": 2.5980589389801025,
+      "learning_rate": 2.5e-05,
+      "loss": 1.3558,
+      "step": 125
+    },
+    {
+      "epoch": 0.04045593193128913,
+      "grad_norm": 3.334627628326416,
+      "learning_rate": 2.5200000000000003e-05,
+      "loss": 1.5553,
+      "step": 126
+    },
+    {
+      "epoch": 0.04077701075614063,
+      "grad_norm": 2.676223039627075,
+      "learning_rate": 2.54e-05,
+      "loss": 1.4527,
+      "step": 127
+    },
+    {
+      "epoch": 0.041098089580992134,
+      "grad_norm": 2.7710494995117188,
+      "learning_rate": 2.5600000000000002e-05,
+      "loss": 1.5945,
+      "step": 128
+    },
+    {
+      "epoch": 0.041419168405843636,
+      "grad_norm": 2.559156894683838,
+      "learning_rate": 2.58e-05,
+      "loss": 1.3244,
+      "step": 129
+    },
+    {
+      "epoch": 0.04174024723069514,
+      "grad_norm": 2.5117619037628174,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 1.5478,
+      "step": 130
+    },
+    {
+      "epoch": 0.04206132605554664,
+      "grad_norm": 2.3930675983428955,
+      "learning_rate": 2.6200000000000003e-05,
+      "loss": 1.3329,
+      "step": 131
+    },
+    {
+      "epoch": 0.042382404880398135,
+      "grad_norm": 2.4628591537475586,
+      "learning_rate": 2.64e-05,
+      "loss": 1.4671,
+      "step": 132
+    },
+    {
+      "epoch": 0.04270348370524964,
+      "grad_norm": 1.9905048608779907,
+      "learning_rate": 2.6600000000000003e-05,
+      "loss": 1.482,
+      "step": 133
+    },
+    {
+      "epoch": 0.04302456253010114,
+      "grad_norm": 2.3793320655822754,
+      "learning_rate": 2.6800000000000004e-05,
+      "loss": 1.4634,
+      "step": 134
+    },
+    {
+      "epoch": 0.04334564135495264,
+      "grad_norm": 2.54632830619812,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 1.2852,
+      "step": 135
+    },
+    {
+      "epoch": 0.04366672017980414,
+      "grad_norm": 3.641115188598633,
+      "learning_rate": 2.7200000000000004e-05,
+      "loss": 1.369,
+      "step": 136
+    },
+    {
+      "epoch": 0.043987799004655645,
+      "grad_norm": 2.443338394165039,
+      "learning_rate": 2.7400000000000002e-05,
+      "loss": 1.3057,
+      "step": 137
+    },
+    {
+      "epoch": 0.04430887782950715,
+      "grad_norm": 3.577340602874756,
+      "learning_rate": 2.7600000000000003e-05,
+      "loss": 1.3757,
+      "step": 138
+    },
+    {
+      "epoch": 0.04462995665435864,
+      "grad_norm": 2.4298670291900635,
+      "learning_rate": 2.7800000000000005e-05,
+      "loss": 1.2701,
+      "step": 139
+    },
+    {
+      "epoch": 0.044951035479210144,
+      "grad_norm": 2.5725412368774414,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 1.3694,
+      "step": 140
+    },
+    {
+      "epoch": 0.045272114304061646,
+      "grad_norm": 2.3307549953460693,
+      "learning_rate": 2.8199999999999998e-05,
+      "loss": 1.3261,
+      "step": 141
+    },
+    {
+      "epoch": 0.04559319312891315,
+      "grad_norm": 3.738875389099121,
+      "learning_rate": 2.84e-05,
+      "loss": 1.3514,
+      "step": 142
+    },
+    {
+      "epoch": 0.04591427195376465,
+      "grad_norm": 2.905665397644043,
+      "learning_rate": 2.86e-05,
+      "loss": 1.0867,
+      "step": 143
+    },
+    {
+      "epoch": 0.04623535077861615,
+      "grad_norm": 3.744802713394165,
+      "learning_rate": 2.88e-05,
+      "loss": 1.1365,
+      "step": 144
+    },
+    {
+      "epoch": 0.046556429603467654,
+      "grad_norm": 3.559023141860962,
+      "learning_rate": 2.9e-05,
+      "loss": 1.2399,
+      "step": 145
+    },
+    {
+      "epoch": 0.04687750842831915,
+      "grad_norm": 3.565185070037842,
+      "learning_rate": 2.9199999999999998e-05,
+      "loss": 1.2912,
+      "step": 146
+    },
+    {
+      "epoch": 0.04719858725317065,
+      "grad_norm": 3.949876308441162,
+      "learning_rate": 2.94e-05,
+      "loss": 1.1523,
+      "step": 147
+    },
+    {
+      "epoch": 0.04751966607802215,
+      "grad_norm": 3.7057857513427734,
+      "learning_rate": 2.96e-05,
+      "loss": 1.1759,
+      "step": 148
+    },
+    {
+      "epoch": 0.047840744902873655,
+      "grad_norm": 2.8545663356781006,
+      "learning_rate": 2.98e-05,
+      "loss": 1.058,
+      "step": 149
+    },
+    {
+      "epoch": 0.04816182372772516,
+      "grad_norm": 5.648036003112793,
+      "learning_rate": 3e-05,
+      "loss": 1.1387,
+      "step": 150
+    },
+    {
+      "epoch": 0.04848290255257666,
+      "grad_norm": 2.1492836475372314,
+      "learning_rate": 3.02e-05,
+      "loss": 1.6236,
+      "step": 151
+    },
+    {
+      "epoch": 0.04880398137742816,
+      "grad_norm": 2.80293607711792,
+      "learning_rate": 3.04e-05,
+      "loss": 1.7931,
+      "step": 152
+    },
+    {
+      "epoch": 0.049125060202279656,
+      "grad_norm": 2.3928093910217285,
+      "learning_rate": 3.06e-05,
+      "loss": 1.2039,
+      "step": 153
+    },
+    {
+      "epoch": 0.04944613902713116,
+      "grad_norm": 2.5905890464782715,
+      "learning_rate": 3.08e-05,
+      "loss": 1.1153,
+      "step": 154
+    },
+    {
+      "epoch": 0.04976721785198266,
+      "grad_norm": 3.102632761001587,
+      "learning_rate": 3.1e-05,
+      "loss": 1.2945,
+      "step": 155
+    },
+    {
+      "epoch": 0.05008829667683416,
+      "grad_norm": 3.5045828819274902,
+      "learning_rate": 3.12e-05,
+      "loss": 1.15,
+      "step": 156
+    },
+    {
+      "epoch": 0.050409375501685665,
+      "grad_norm": 3.210549831390381,
+      "learning_rate": 3.1400000000000004e-05,
+      "loss": 1.2008,
+      "step": 157
+    },
+    {
+      "epoch": 0.05073045432653717,
+      "grad_norm": 2.5296566486358643,
+      "learning_rate": 3.16e-05,
+      "loss": 1.2581,
+      "step": 158
+    },
+    {
+      "epoch": 0.05105153315138867,
+      "grad_norm": 3.071568727493286,
+      "learning_rate": 3.18e-05,
+      "loss": 0.9904,
+      "step": 159
+    },
+    {
+      "epoch": 0.051372611976240164,
+      "grad_norm": 2.763911724090576,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.12,
+      "step": 160
+    },
+    {
+      "epoch": 0.051693690801091666,
+      "grad_norm": 2.2567617893218994,
+      "learning_rate": 3.2200000000000003e-05,
+      "loss": 1.2534,
+      "step": 161
+    },
+    {
+      "epoch": 0.05201476962594317,
+      "grad_norm": 2.49045991897583,
+      "learning_rate": 3.24e-05,
+      "loss": 1.2713,
+      "step": 162
+    },
+    {
+      "epoch": 0.05233584845079467,
+      "grad_norm": 3.3164381980895996,
+      "learning_rate": 3.26e-05,
+      "loss": 1.1137,
+      "step": 163
+    },
+    {
+      "epoch": 0.05265692727564617,
+      "grad_norm": 2.6513657569885254,
+      "learning_rate": 3.2800000000000004e-05,
+      "loss": 1.3505,
+      "step": 164
+    },
+    {
+      "epoch": 0.052978006100497674,
+      "grad_norm": 2.629395008087158,
+      "learning_rate": 3.3e-05,
+      "loss": 1.206,
+      "step": 165
+    },
+    {
+      "epoch": 0.053299084925349176,
+      "grad_norm": 3.333925485610962,
+      "learning_rate": 3.32e-05,
+      "loss": 1.1088,
+      "step": 166
+    },
+    {
+      "epoch": 0.05362016375020067,
+      "grad_norm": 2.59550142288208,
+      "learning_rate": 3.3400000000000005e-05,
+      "loss": 1.2533,
+      "step": 167
+    },
+    {
+      "epoch": 0.05394124257505217,
+      "grad_norm": 2.676664352416992,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 1.2916,
+      "step": 168
+    },
+    {
+      "epoch": 0.054262321399903675,
+      "grad_norm": 2.5266637802124023,
+      "learning_rate": 3.38e-05,
+      "loss": 1.3359,
+      "step": 169
+    },
+    {
+      "epoch": 0.05458340022475518,
+      "grad_norm": 2.394808292388916,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 1.0293,
+      "step": 170
+    },
+    {
+      "epoch": 0.05490447904960668,
+      "grad_norm": 3.7004685401916504,
+      "learning_rate": 3.4200000000000005e-05,
+      "loss": 1.224,
+      "step": 171
+    },
+    {
+      "epoch": 0.05522555787445818,
+      "grad_norm": 2.8753304481506348,
+      "learning_rate": 3.4399999999999996e-05,
+      "loss": 1.2212,
+      "step": 172
+    },
+    {
+      "epoch": 0.05554663669930968,
+      "grad_norm": 2.939927101135254,
+      "learning_rate": 3.46e-05,
+      "loss": 1.4577,
+      "step": 173
+    },
+    {
+      "epoch": 0.055867715524161185,
+      "grad_norm": 2.922861099243164,
+      "learning_rate": 3.48e-05,
+      "loss": 1.2349,
+      "step": 174
+    },
+    {
+      "epoch": 0.05618879434901268,
+      "grad_norm": 2.545013427734375,
+      "learning_rate": 3.5e-05,
+      "loss": 1.0631,
+      "step": 175
+    },
+    {
+      "epoch": 0.05650987317386418,
+      "grad_norm": 3.1497952938079834,
+      "learning_rate": 3.52e-05,
+      "loss": 1.2337,
+      "step": 176
+    },
+    {
+      "epoch": 0.056830951998715684,
+      "grad_norm": 2.5342869758605957,
+      "learning_rate": 3.54e-05,
+      "loss": 1.2889,
+      "step": 177
+    },
+    {
+      "epoch": 0.057152030823567186,
+      "grad_norm": 2.435774087905884,
+      "learning_rate": 3.56e-05,
+      "loss": 1.4616,
+      "step": 178
+    },
+    {
+      "epoch": 0.05747310964841869,
+      "grad_norm": 1.7864798307418823,
+      "learning_rate": 3.58e-05,
+      "loss": 1.0572,
+      "step": 179
+    },
+    {
+      "epoch": 0.05779418847327019,
+      "grad_norm": 1.8739286661148071,
+      "learning_rate": 3.6e-05,
+      "loss": 1.227,
+      "step": 180
+    },
+    {
+      "epoch": 0.05811526729812169,
+      "grad_norm": 3.2398200035095215,
+      "learning_rate": 3.62e-05,
+      "loss": 1.2632,
+      "step": 181
+    },
+    {
+      "epoch": 0.05843634612297319,
+      "grad_norm": 2.340625047683716,
+      "learning_rate": 3.6400000000000004e-05,
+      "loss": 1.2501,
+      "step": 182
+    },
+    {
+      "epoch": 0.05875742494782469,
+      "grad_norm": 2.0247950553894043,
+      "learning_rate": 3.66e-05,
+      "loss": 1.3319,
+      "step": 183
+    },
+    {
+      "epoch": 0.05907850377267619,
+      "grad_norm": 2.264277935028076,
+      "learning_rate": 3.68e-05,
+      "loss": 1.2761,
+      "step": 184
+    },
+    {
+      "epoch": 0.05939958259752769,
+      "grad_norm": 1.971635341644287,
+      "learning_rate": 3.7e-05,
+      "loss": 0.9021,
+      "step": 185
+    },
+    {
+      "epoch": 0.059720661422379195,
+      "grad_norm": 1.713057518005371,
+      "learning_rate": 3.72e-05,
+      "loss": 1.2372,
+      "step": 186
+    },
+    {
+      "epoch": 0.0600417402472307,
+      "grad_norm": 3.2709195613861084,
+      "learning_rate": 3.74e-05,
+      "loss": 1.2062,
+      "step": 187
+    },
+    {
+      "epoch": 0.0603628190720822,
+      "grad_norm": 1.7991886138916016,
+      "learning_rate": 3.76e-05,
+      "loss": 1.0789,
+      "step": 188
+    },
+    {
+      "epoch": 0.060683897896933695,
+      "grad_norm": 3.217481851577759,
+      "learning_rate": 3.7800000000000004e-05,
+      "loss": 1.2653,
+      "step": 189
+    },
+    {
+      "epoch": 0.0610049767217852,
+      "grad_norm": 1.5862120389938354,
+      "learning_rate": 3.8e-05,
+      "loss": 1.2394,
+      "step": 190
+    },
+    {
+      "epoch": 0.0613260555466367,
+      "grad_norm": 1.8591636419296265,
+      "learning_rate": 3.82e-05,
+      "loss": 1.2675,
+      "step": 191
+    },
+    {
+      "epoch": 0.0616471343714882,
+      "grad_norm": 1.6677289009094238,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 1.0502,
+      "step": 192
+    },
+    {
+      "epoch": 0.0619682131963397,
+      "grad_norm": 2.0882251262664795,
+      "learning_rate": 3.86e-05,
+      "loss": 0.9488,
+      "step": 193
+    },
+    {
+      "epoch": 0.062289292021191205,
+      "grad_norm": 1.579268217086792,
+      "learning_rate": 3.88e-05,
+      "loss": 1.1667,
+      "step": 194
+    },
+    {
+      "epoch": 0.0626103708460427,
+      "grad_norm": 1.7781903743743896,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 1.238,
+      "step": 195
+    },
+    {
+      "epoch": 0.0629314496708942,
+      "grad_norm": 2.556015729904175,
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 1.0398,
+      "step": 196
+    },
+    {
+      "epoch": 0.06325252849574571,
+      "grad_norm": 1.9498616456985474,
+      "learning_rate": 3.94e-05,
+      "loss": 1.0116,
+      "step": 197
+    },
+    {
+      "epoch": 0.0635736073205972,
+      "grad_norm": 3.5160417556762695,
+      "learning_rate": 3.960000000000001e-05,
+      "loss": 0.8621,
+      "step": 198
+    },
+    {
+      "epoch": 0.0638946861454487,
+      "grad_norm": 3.2988150119781494,
+      "learning_rate": 3.9800000000000005e-05,
+      "loss": 0.836,
+      "step": 199
+    },
+    {
+      "epoch": 0.06421576497030021,
+      "grad_norm": 3.2342076301574707,
+      "learning_rate": 4e-05,
+      "loss": 0.7853,
+      "step": 200
+    },
+    {
+      "epoch": 0.0645368437951517,
+      "grad_norm": 3.1116089820861816,
+      "learning_rate": 4.02e-05,
+      "loss": 1.5181,
+      "step": 201
+    },
+    {
+      "epoch": 0.06485792262000321,
+      "grad_norm": 4.22519063949585,
+      "learning_rate": 4.0400000000000006e-05,
+      "loss": 1.5198,
+      "step": 202
+    },
+    {
+      "epoch": 0.06517900144485471,
+      "grad_norm": 3.6451752185821533,
+      "learning_rate": 4.0600000000000004e-05,
+      "loss": 1.1846,
+      "step": 203
+    },
+    {
+      "epoch": 0.06550008026970622,
+      "grad_norm": 2.907416582107544,
+      "learning_rate": 4.08e-05,
+      "loss": 1.1008,
+      "step": 204
+    },
+    {
+      "epoch": 0.06582115909455771,
+      "grad_norm": 3.7612717151641846,
+      "learning_rate": 4.1e-05,
+      "loss": 1.0514,
+      "step": 205
+    },
+    {
+      "epoch": 0.06614223791940921,
+      "grad_norm": 2.8270351886749268,
+      "learning_rate": 4.12e-05,
+      "loss": 1.1469,
+      "step": 206
+    },
+    {
+      "epoch": 0.06646331674426072,
+      "grad_norm": 2.9322104454040527,
+      "learning_rate": 4.14e-05,
+      "loss": 1.1255,
+      "step": 207
+    },
+    {
+      "epoch": 0.06678439556911221,
+      "grad_norm": 2.4711625576019287,
+      "learning_rate": 4.16e-05,
+      "loss": 1.0622,
+      "step": 208
+    },
+    {
+      "epoch": 0.06710547439396372,
+      "grad_norm": 1.7461209297180176,
+      "learning_rate": 4.18e-05,
+      "loss": 1.1028,
+      "step": 209
+    },
+    {
+      "epoch": 0.06742655321881522,
+      "grad_norm": 2.2536096572875977,
+      "learning_rate": 4.2e-05,
+      "loss": 1.0743,
+      "step": 210
+    },
+    {
+      "epoch": 0.06774763204366673,
+      "grad_norm": 2.037104606628418,
+      "learning_rate": 4.22e-05,
+      "loss": 1.2257,
+      "step": 211
+    },
+    {
+      "epoch": 0.06806871086851822,
+      "grad_norm": 2.400221586227417,
+      "learning_rate": 4.24e-05,
+      "loss": 1.2784,
+      "step": 212
+    },
+    {
+      "epoch": 0.06838978969336973,
+      "grad_norm": 2.0338759422302246,
+      "learning_rate": 4.26e-05,
+      "loss": 1.1648,
+      "step": 213
+    },
+    {
+      "epoch": 0.06871086851822122,
+      "grad_norm": 3.6578612327575684,
+      "learning_rate": 4.2800000000000004e-05,
+      "loss": 1.1226,
+      "step": 214
+    },
+    {
+      "epoch": 0.06903194734307272,
+      "grad_norm": 2.1461496353149414,
+      "learning_rate": 4.3e-05,
+      "loss": 1.0823,
+      "step": 215
+    },
+    {
+      "epoch": 0.06935302616792423,
+      "grad_norm": 1.87822687625885,
+      "learning_rate": 4.32e-05,
+      "loss": 1.1413,
+      "step": 216
+    },
+    {
+      "epoch": 0.06967410499277572,
+      "grad_norm": 1.7954866886138916,
+      "learning_rate": 4.3400000000000005e-05,
+      "loss": 1.2208,
+      "step": 217
+    },
+    {
+      "epoch": 0.06999518381762723,
+      "grad_norm": 2.6523921489715576,
+      "learning_rate": 4.36e-05,
+      "loss": 1.0858,
+      "step": 218
+    },
+    {
+      "epoch": 0.07031626264247873,
+      "grad_norm": 2.4885449409484863,
+      "learning_rate": 4.38e-05,
+      "loss": 1.0251,
+      "step": 219
+    },
+    {
+      "epoch": 0.07063734146733024,
+      "grad_norm": 2.9225237369537354,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.2586,
+      "step": 220
+    },
+    {
+      "epoch": 0.07095842029218173,
+      "grad_norm": 2.579859495162964,
+      "learning_rate": 4.4200000000000004e-05,
+      "loss": 1.2634,
+      "step": 221
+    },
+    {
+      "epoch": 0.07127949911703323,
+      "grad_norm": 3.152012586593628,
+      "learning_rate": 4.44e-05,
+      "loss": 1.0579,
+      "step": 222
+    },
+    {
+      "epoch": 0.07160057794188474,
+      "grad_norm": 1.843085765838623,
+      "learning_rate": 4.46e-05,
+      "loss": 1.0507,
+      "step": 223
+    },
+    {
+      "epoch": 0.07192165676673623,
+      "grad_norm": 2.062282085418701,
+      "learning_rate": 4.4800000000000005e-05,
+      "loss": 1.3025,
+      "step": 224
+    },
+    {
+      "epoch": 0.07224273559158774,
+      "grad_norm": 1.8540760278701782,
+      "learning_rate": 4.5e-05,
+      "loss": 1.0763,
+      "step": 225
+    },
+    {
+      "epoch": 0.07256381441643923,
+      "grad_norm": 2.1458258628845215,
+      "learning_rate": 4.52e-05,
+      "loss": 1.1163,
+      "step": 226
+    },
+    {
+      "epoch": 0.07288489324129074,
+      "grad_norm": 2.402344226837158,
+      "learning_rate": 4.5400000000000006e-05,
+      "loss": 1.2973,
+      "step": 227
+    },
+    {
+      "epoch": 0.07320597206614224,
+      "grad_norm": 1.6383142471313477,
+      "learning_rate": 4.5600000000000004e-05,
+      "loss": 1.1735,
+      "step": 228
+    },
+    {
+      "epoch": 0.07352705089099373,
+      "grad_norm": 1.9293252229690552,
+      "learning_rate": 4.58e-05,
+      "loss": 1.1113,
+      "step": 229
+    },
+    {
+      "epoch": 0.07384812971584524,
+      "grad_norm": 2.625548839569092,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 1.1503,
+      "step": 230
+    },
+    {
+      "epoch": 0.07416920854069674,
+      "grad_norm": 2.3177950382232666,
+      "learning_rate": 4.6200000000000005e-05,
+      "loss": 0.9863,
+      "step": 231
+    },
+    {
+      "epoch": 0.07449028736554825,
+      "grad_norm": 3.1583642959594727,
+      "learning_rate": 4.64e-05,
+      "loss": 1.3059,
+      "step": 232
+    },
+    {
+      "epoch": 0.07481136619039974,
+      "grad_norm": 1.7886139154434204,
+      "learning_rate": 4.660000000000001e-05,
+      "loss": 1.0193,
+      "step": 233
+    },
+    {
+      "epoch": 0.07513244501525125,
+      "grad_norm": 3.0615925788879395,
+      "learning_rate": 4.6800000000000006e-05,
+      "loss": 1.4904,
+      "step": 234
+    },
+    {
+      "epoch": 0.07545352384010275,
+      "grad_norm": 823.4281005859375,
+      "learning_rate": 4.7e-05,
+      "loss": 1.1034,
+      "step": 235
+    },
+    {
+      "epoch": 0.07577460266495424,
+      "grad_norm": 3.0355687141418457,
+      "learning_rate": 4.72e-05,
+      "loss": 0.9429,
+      "step": 236
+    },
+    {
+      "epoch": 0.07609568148980575,
+      "grad_norm": 556.7845458984375,
+      "learning_rate": 4.74e-05,
+      "loss": 1.1951,
+      "step": 237
+    },
+    {
+      "epoch": 0.07641676031465724,
+      "grad_norm": 3.066845417022705,
+      "learning_rate": 4.76e-05,
+      "loss": 1.1543,
+      "step": 238
+    },
+    {
+      "epoch": 0.07673783913950875,
+      "grad_norm": 2.495962142944336,
+      "learning_rate": 4.78e-05,
+      "loss": 1.1047,
+      "step": 239
+    },
+    {
+      "epoch": 0.07705891796436025,
+      "grad_norm": 12.180785179138184,
+      "learning_rate": 4.8e-05,
+      "loss": 0.9588,
+      "step": 240
+    },
+    {
+      "epoch": 0.07737999678921176,
+      "grad_norm": 3.484467029571533,
+      "learning_rate": 4.82e-05,
+      "loss": 1.1392,
+      "step": 241
+    },
+    {
+      "epoch": 0.07770107561406325,
+      "grad_norm": 2.198939561843872,
+      "learning_rate": 4.8400000000000004e-05,
+      "loss": 1.3359,
+      "step": 242
+    },
+    {
+      "epoch": 0.07802215443891475,
+      "grad_norm": 1.4292407035827637,
+      "learning_rate": 4.86e-05,
+      "loss": 0.9899,
+      "step": 243
+    },
+    {
+      "epoch": 0.07834323326376626,
+      "grad_norm": 1.910438895225525,
+      "learning_rate": 4.88e-05,
+      "loss": 0.928,
+      "step": 244
+    },
+    {
+      "epoch": 0.07866431208861775,
+      "grad_norm": 3.0100486278533936,
+      "learning_rate": 4.9e-05,
+      "loss": 1.4606,
+      "step": 245
+    },
+    {
+      "epoch": 0.07898539091346926,
+      "grad_norm": 2.0493595600128174,
+      "learning_rate": 4.92e-05,
+      "loss": 1.0296,
+      "step": 246
+    },
+    {
+      "epoch": 0.07930646973832076,
+      "grad_norm": 2.848602771759033,
+      "learning_rate": 4.94e-05,
+      "loss": 0.9189,
+      "step": 247
+    },
+    {
+      "epoch": 0.07962754856317227,
+      "grad_norm": 3.749636650085449,
+      "learning_rate": 4.96e-05,
+      "loss": 0.9614,
+      "step": 248
+    },
+    {
+      "epoch": 0.07994862738802376,
+      "grad_norm": 2.882443428039551,
+      "learning_rate": 4.9800000000000004e-05,
+      "loss": 0.872,
+      "step": 249
+    },
+    {
+      "epoch": 0.08026970621287526,
+      "grad_norm": 1.6551190614700317,
+      "learning_rate": 5e-05,
+      "loss": 0.7895,
+      "step": 250
+    },
+    {
+      "epoch": 0.08059078503772676,
+      "grad_norm": 2.3475422859191895,
+      "learning_rate": 5.02e-05,
+      "loss": 1.4449,
+      "step": 251
+    },
+    {
+      "epoch": 0.08091186386257826,
+      "grad_norm": 2.298914670944214,
+      "learning_rate": 5.0400000000000005e-05,
+      "loss": 1.0738,
+      "step": 252
+    },
+    {
+      "epoch": 0.08123294268742977,
+      "grad_norm": 2.8183484077453613,
+      "learning_rate": 5.0600000000000003e-05,
+      "loss": 1.0035,
+      "step": 253
+    },
+    {
+      "epoch": 0.08155402151228126,
+      "grad_norm": 2.7477033138275146,
+      "learning_rate": 5.08e-05,
+      "loss": 1.0145,
+      "step": 254
+    },
+    {
+      "epoch": 0.08187510033713277,
+      "grad_norm": 2.9522218704223633,
+      "learning_rate": 5.1000000000000006e-05,
+      "loss": 0.962,
+      "step": 255
+    },
+    {
+      "epoch": 0.08219617916198427,
+      "grad_norm": 2.496779203414917,
+      "learning_rate": 5.1200000000000004e-05,
+      "loss": 0.899,
+      "step": 256
+    },
+    {
+      "epoch": 0.08251725798683576,
+      "grad_norm": 3.338456392288208,
+      "learning_rate": 5.14e-05,
+      "loss": 1.0442,
+      "step": 257
+    },
+    {
+      "epoch": 0.08283833681168727,
+      "grad_norm": 2.447253465652466,
+      "learning_rate": 5.16e-05,
+      "loss": 1.0634,
+      "step": 258
+    },
+    {
+      "epoch": 0.08315941563653877,
+      "grad_norm": 2.5269699096679688,
+      "learning_rate": 5.1800000000000005e-05,
+      "loss": 1.216,
+      "step": 259
+    },
+    {
+      "epoch": 0.08348049446139028,
+      "grad_norm": 1.807884693145752,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 1.051,
+      "step": 260
+    },
+    {
+      "epoch": 0.08380157328624177,
+      "grad_norm": 3.292743444442749,
+      "learning_rate": 5.22e-05,
+      "loss": 1.0427,
+      "step": 261
+    },
+    {
+      "epoch": 0.08412265211109328,
+      "grad_norm": 1.926658034324646,
+      "learning_rate": 5.2400000000000007e-05,
+      "loss": 1.1237,
+      "step": 262
+    },
+    {
+      "epoch": 0.08444373093594477,
+      "grad_norm": 2.5301294326782227,
+      "learning_rate": 5.2600000000000005e-05,
+      "loss": 1.1137,
+      "step": 263
+    },
+    {
+      "epoch": 0.08476480976079627,
+      "grad_norm": 2.3400182723999023,
+      "learning_rate": 5.28e-05,
+      "loss": 0.9848,
+      "step": 264
+    },
+    {
+      "epoch": 0.08508588858564778,
+      "grad_norm": 1.7750979661941528,
+      "learning_rate": 5.300000000000001e-05,
+      "loss": 1.1295,
+      "step": 265
+    },
+    {
+      "epoch": 0.08540696741049927,
+      "grad_norm": 2.335904598236084,
+      "learning_rate": 5.3200000000000006e-05,
+      "loss": 1.2678,
+      "step": 266
+    },
+    {
+      "epoch": 0.08572804623535078,
+      "grad_norm": 1.9911553859710693,
+      "learning_rate": 5.3400000000000004e-05,
+      "loss": 1.1416,
+      "step": 267
+    },
+    {
+      "epoch": 0.08604912506020228,
+      "grad_norm": 1.7395027875900269,
+      "learning_rate": 5.360000000000001e-05,
+      "loss": 1.1398,
+      "step": 268
+    },
+    {
+      "epoch": 0.08637020388505379,
+      "grad_norm": 1.7891736030578613,
+      "learning_rate": 5.380000000000001e-05,
+      "loss": 1.1321,
+      "step": 269
+    },
+    {
+      "epoch": 0.08669128270990528,
+      "grad_norm": 1.8368785381317139,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 1.0112,
+      "step": 270
+    },
+    {
+      "epoch": 0.08701236153475678,
+      "grad_norm": 1.638079047203064,
+      "learning_rate": 5.420000000000001e-05,
+      "loss": 1.1376,
+      "step": 271
+    },
+    {
+      "epoch": 0.08733344035960829,
+      "grad_norm": 1.9354631900787354,
+      "learning_rate": 5.440000000000001e-05,
+      "loss": 1.1693,
+      "step": 272
+    },
+    {
+      "epoch": 0.08765451918445978,
+      "grad_norm": 2.017259359359741,
+      "learning_rate": 5.4600000000000006e-05,
+      "loss": 1.2163,
+      "step": 273
+    },
+    {
+      "epoch": 0.08797559800931129,
+      "grad_norm": 1.6436361074447632,
+      "learning_rate": 5.4800000000000004e-05,
+      "loss": 0.9737,
+      "step": 274
+    },
+    {
+      "epoch": 0.08829667683416279,
+      "grad_norm": 1.7359622716903687,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 1.2599,
+      "step": 275
+    },
+    {
+      "epoch": 0.0886177556590143,
+      "grad_norm": 1.698771595954895,
+      "learning_rate": 5.520000000000001e-05,
+      "loss": 1.0724,
+      "step": 276
+    },
+    {
+      "epoch": 0.08893883448386579,
+      "grad_norm": 3.8529412746429443,
+      "learning_rate": 5.5400000000000005e-05,
+      "loss": 1.4372,
+      "step": 277
+    },
+    {
+      "epoch": 0.08925991330871728,
+      "grad_norm": 1.826474905014038,
+      "learning_rate": 5.560000000000001e-05,
+      "loss": 1.4767,
+      "step": 278
+    },
+    {
+      "epoch": 0.0895809921335688,
+      "grad_norm": 1.7558326721191406,
+      "learning_rate": 5.580000000000001e-05,
+      "loss": 1.1549,
+      "step": 279
+    },
+    {
+      "epoch": 0.08990207095842029,
+      "grad_norm": 2.5463342666625977,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 0.9819,
+      "step": 280
+    },
+    {
+      "epoch": 0.0902231497832718,
+      "grad_norm": 1.6322652101516724,
+      "learning_rate": 5.620000000000001e-05,
+      "loss": 1.2071,
+      "step": 281
+    },
+    {
+      "epoch": 0.09054422860812329,
+      "grad_norm": 1.6438385248184204,
+      "learning_rate": 5.6399999999999995e-05,
+      "loss": 0.991,
+      "step": 282
+    },
+    {
+      "epoch": 0.0908653074329748,
+      "grad_norm": 1.8053836822509766,
+      "learning_rate": 5.66e-05,
+      "loss": 1.1238,
+      "step": 283
+    },
+    {
+      "epoch": 0.0911863862578263,
+      "grad_norm": 2.149841070175171,
+      "learning_rate": 5.68e-05,
+      "loss": 1.1431,
+      "step": 284
+    },
+    {
+      "epoch": 0.09150746508267779,
+      "grad_norm": 1.8680065870285034,
+      "learning_rate": 5.6999999999999996e-05,
+      "loss": 1.0299,
+      "step": 285
+    },
+    {
+      "epoch": 0.0918285439075293,
+      "grad_norm": 1.402052879333496,
+      "learning_rate": 5.72e-05,
+      "loss": 1.0877,
+      "step": 286
+    },
+    {
+      "epoch": 0.0921496227323808,
+      "grad_norm": 1.4989826679229736,
+      "learning_rate": 5.74e-05,
+      "loss": 0.9015,
+      "step": 287
+    },
+    {
+      "epoch": 0.0924707015572323,
+      "grad_norm": 1.2958741188049316,
+      "learning_rate": 5.76e-05,
+      "loss": 1.0378,
+      "step": 288
+    },
+    {
+      "epoch": 0.0927917803820838,
+      "grad_norm": 2.0742673873901367,
+      "learning_rate": 5.7799999999999995e-05,
+      "loss": 1.0337,
+      "step": 289
+    },
+    {
+      "epoch": 0.09311285920693531,
+      "grad_norm": 1.4535075426101685,
+      "learning_rate": 5.8e-05,
+      "loss": 1.0506,
+      "step": 290
+    },
+    {
+      "epoch": 0.0934339380317868,
+      "grad_norm": 1.7830564975738525,
+      "learning_rate": 5.82e-05,
+      "loss": 1.0116,
+      "step": 291
+    },
+    {
+      "epoch": 0.0937550168566383,
+      "grad_norm": 1.6709142923355103,
+      "learning_rate": 5.8399999999999997e-05,
+      "loss": 0.7538,
+      "step": 292
+    },
+    {
+      "epoch": 0.09407609568148981,
+      "grad_norm": 2.7168803215026855,
+      "learning_rate": 5.86e-05,
+      "loss": 1.1456,
+      "step": 293
+    },
+    {
+      "epoch": 0.0943971745063413,
+      "grad_norm": 1.8454351425170898,
+      "learning_rate": 5.88e-05,
+      "loss": 0.7657,
+      "step": 294
+    },
+    {
+      "epoch": 0.09471825333119281,
+      "grad_norm": 2.2031004428863525,
+      "learning_rate": 5.9e-05,
+      "loss": 0.8091,
+      "step": 295
+    },
+    {
+      "epoch": 0.0950393321560443,
+      "grad_norm": 1.5919681787490845,
+      "learning_rate": 5.92e-05,
+      "loss": 0.9871,
+      "step": 296
+    },
+    {
+      "epoch": 0.09536041098089582,
+      "grad_norm": 2.7647721767425537,
+      "learning_rate": 5.94e-05,
+      "loss": 0.9902,
+      "step": 297
+    },
+    {
+      "epoch": 0.09568148980574731,
+      "grad_norm": 2.6411798000335693,
+      "learning_rate": 5.96e-05,
+      "loss": 0.9511,
+      "step": 298
+    },
+    {
+      "epoch": 0.0960025686305988,
+      "grad_norm": 2.426961898803711,
+      "learning_rate": 5.9800000000000003e-05,
+      "loss": 1.06,
+      "step": 299
+    },
+    {
+      "epoch": 0.09632364745545031,
+      "grad_norm": 2.5059926509857178,
+      "learning_rate": 6e-05,
+      "loss": 0.7106,
+      "step": 300
+    },
+    {
+      "epoch": 0.09664472628030181,
+      "grad_norm": 3.1037328243255615,
+      "learning_rate": 6.02e-05,
+      "loss": 1.5007,
+      "step": 301
+    },
+    {
+      "epoch": 0.09696580510515332,
+      "grad_norm": 2.7043604850769043,
+      "learning_rate": 6.04e-05,
+      "loss": 1.5066,
+      "step": 302
+    },
+    {
+      "epoch": 0.09728688393000481,
+      "grad_norm": 3.1140081882476807,
+      "learning_rate": 6.06e-05,
+      "loss": 1.052,
+      "step": 303
+    },
+    {
+      "epoch": 0.09760796275485632,
+      "grad_norm": 3.623964309692383,
+      "learning_rate": 6.08e-05,
+      "loss": 0.946,
+      "step": 304
+    },
+    {
+      "epoch": 0.09792904157970782,
+      "grad_norm": 3.414036750793457,
+      "learning_rate": 6.1e-05,
+      "loss": 0.9647,
+      "step": 305
+    },
+    {
+      "epoch": 0.09825012040455931,
+      "grad_norm": 3.4817802906036377,
+      "learning_rate": 6.12e-05,
+      "loss": 0.9753,
+      "step": 306
+    },
+    {
+      "epoch": 0.09857119922941082,
+      "grad_norm": 3.067047119140625,
+      "learning_rate": 6.14e-05,
+      "loss": 1.1435,
+      "step": 307
+    },
+    {
+      "epoch": 0.09889227805426232,
+      "grad_norm": 2.2207508087158203,
+      "learning_rate": 6.16e-05,
+      "loss": 0.998,
+      "step": 308
+    },
+    {
+      "epoch": 0.09921335687911383,
+      "grad_norm": 2.4805872440338135,
+      "learning_rate": 6.18e-05,
+      "loss": 0.9144,
+      "step": 309
+    },
+    {
+      "epoch": 0.09953443570396532,
+      "grad_norm": 1.8067208528518677,
+      "learning_rate": 6.2e-05,
+      "loss": 1.1963,
+      "step": 310
+    },
+    {
+      "epoch": 0.09985551452881683,
+      "grad_norm": 1.950782299041748,
+      "learning_rate": 6.220000000000001e-05,
+      "loss": 1.1947,
+      "step": 311
+    },
+    {
+      "epoch": 0.10017659335366833,
+      "grad_norm": 1.88652765750885,
+      "learning_rate": 6.24e-05,
+      "loss": 0.9983,
+      "step": 312
+    },
+    {
+      "epoch": 0.10049767217851982,
+      "grad_norm": 1.9043608903884888,
+      "learning_rate": 6.26e-05,
+      "loss": 0.8638,
+      "step": 313
+    },
+    {
+      "epoch": 0.10081875100337133,
+      "grad_norm": 1.84428071975708,
+      "learning_rate": 6.280000000000001e-05,
+      "loss": 1.0008,
+      "step": 314
+    },
+    {
+      "epoch": 0.10113982982822282,
+      "grad_norm": 1.763142466545105,
+      "learning_rate": 6.3e-05,
+      "loss": 1.3726,
+      "step": 315
+    },
+    {
+      "epoch": 0.10146090865307433,
+      "grad_norm": 1.6071946620941162,
+      "learning_rate": 6.32e-05,
+      "loss": 1.0704,
+      "step": 316
+    },
+    {
+      "epoch": 0.10178198747792583,
+      "grad_norm": 2.2262184619903564,
+      "learning_rate": 6.340000000000001e-05,
+      "loss": 1.0787,
+      "step": 317
+    },
+    {
+      "epoch": 0.10210306630277734,
+      "grad_norm": 2.342939615249634,
+      "learning_rate": 6.36e-05,
+      "loss": 1.096,
+      "step": 318
+    },
+    {
+      "epoch": 0.10242414512762883,
+      "grad_norm": 2.372135639190674,
+      "learning_rate": 6.38e-05,
+      "loss": 1.1364,
+      "step": 319
+    },
+    {
+      "epoch": 0.10274522395248033,
+      "grad_norm": 1.7545585632324219,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 0.9141,
+      "step": 320
+    },
+    {
+      "epoch": 0.10306630277733184,
+      "grad_norm": 2.9137213230133057,
+      "learning_rate": 6.42e-05,
+      "loss": 0.9489,
+      "step": 321
+    },
+    {
+      "epoch": 0.10338738160218333,
+      "grad_norm": 2.827456474304199,
+      "learning_rate": 6.440000000000001e-05,
+      "loss": 1.1006,
+      "step": 322
+    },
+    {
+      "epoch": 0.10370846042703484,
+      "grad_norm": 3.5646371841430664,
+      "learning_rate": 6.460000000000001e-05,
+      "loss": 1.0534,
+      "step": 323
+    },
+    {
+      "epoch": 0.10402953925188634,
+      "grad_norm": 1.827415943145752,
+      "learning_rate": 6.48e-05,
+      "loss": 1.0138,
+      "step": 324
+    },
+    {
+      "epoch": 0.10435061807673784,
+      "grad_norm": 2.491384267807007,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 1.4315,
+      "step": 325
+    },
+    {
+      "epoch": 0.10467169690158934,
+      "grad_norm": 1.7412528991699219,
+      "learning_rate": 6.52e-05,
+      "loss": 0.9961,
+      "step": 326
+    },
+    {
+      "epoch": 0.10499277572644083,
+      "grad_norm": 2.2214128971099854,
+      "learning_rate": 6.54e-05,
+      "loss": 1.2276,
+      "step": 327
+    },
+    {
+      "epoch": 0.10531385455129234,
+      "grad_norm": 1.9566681385040283,
+      "learning_rate": 6.560000000000001e-05,
+      "loss": 1.3574,
+      "step": 328
+    },
+    {
+      "epoch": 0.10563493337614384,
+      "grad_norm": 1.5677928924560547,
+      "learning_rate": 6.58e-05,
+      "loss": 1.1721,
+      "step": 329
+    },
+    {
+      "epoch": 0.10595601220099535,
+      "grad_norm": 2.4960641860961914,
+      "learning_rate": 6.6e-05,
+      "loss": 1.4447,
+      "step": 330
+    },
+    {
+      "epoch": 0.10627709102584684,
+      "grad_norm": 2.4556682109832764,
+      "learning_rate": 6.620000000000001e-05,
+      "loss": 1.0275,
+      "step": 331
+    },
+    {
+      "epoch": 0.10659816985069835,
+      "grad_norm": 2.3573825359344482,
+      "learning_rate": 6.64e-05,
+      "loss": 1.2238,
+      "step": 332
+    },
+    {
+      "epoch": 0.10691924867554985,
+      "grad_norm": 1.6296504735946655,
+      "learning_rate": 6.66e-05,
+      "loss": 1.0889,
+      "step": 333
+    },
+    {
+      "epoch": 0.10724032750040134,
+      "grad_norm": 1.5856961011886597,
+      "learning_rate": 6.680000000000001e-05,
+      "loss": 1.1944,
+      "step": 334
+    },
+    {
+      "epoch": 0.10756140632525285,
+      "grad_norm": 1.3460253477096558,
+      "learning_rate": 6.7e-05,
+      "loss": 1.2147,
+      "step": 335
+    },
+    {
+      "epoch": 0.10788248515010435,
+      "grad_norm": 2.425680637359619,
+      "learning_rate": 6.720000000000001e-05,
+      "loss": 1.0764,
+      "step": 336
+    },
+    {
+      "epoch": 0.10820356397495585,
+      "grad_norm": 1.4525935649871826,
+      "learning_rate": 6.740000000000001e-05,
+      "loss": 1.1013,
+      "step": 337
+    },
+    {
+      "epoch": 0.10852464279980735,
+      "grad_norm": 2.8759169578552246,
+      "learning_rate": 6.76e-05,
+      "loss": 1.1745,
+      "step": 338
+    },
+    {
+      "epoch": 0.10884572162465886,
+      "grad_norm": 2.58520770072937,
+      "learning_rate": 6.780000000000001e-05,
+      "loss": 0.8345,
+      "step": 339
+    },
+    {
+      "epoch": 0.10916680044951035,
+      "grad_norm": 1.522647738456726,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 0.9961,
+      "step": 340
+    },
+    {
+      "epoch": 0.10948787927436185,
+      "grad_norm": 2.1563315391540527,
+      "learning_rate": 6.82e-05,
+      "loss": 1.1935,
+      "step": 341
+    },
+    {
+      "epoch": 0.10980895809921336,
+      "grad_norm": 2.3725993633270264,
+      "learning_rate": 6.840000000000001e-05,
+      "loss": 1.193,
+      "step": 342
+    },
+    {
+      "epoch": 0.11013003692406485,
+      "grad_norm": 1.3734397888183594,
+      "learning_rate": 6.860000000000001e-05,
+      "loss": 1.0738,
+      "step": 343
+    },
+    {
+      "epoch": 0.11045111574891636,
+      "grad_norm": 2.5987794399261475,
+      "learning_rate": 6.879999999999999e-05,
+      "loss": 1.3577,
+      "step": 344
+    },
+    {
+      "epoch": 0.11077219457376786,
+      "grad_norm": 2.429164171218872,
+      "learning_rate": 6.9e-05,
+      "loss": 1.1118,
+      "step": 345
+    },
+    {
+      "epoch": 0.11109327339861937,
+      "grad_norm": 2.4097049236297607,
+      "learning_rate": 6.92e-05,
+      "loss": 1.0425,
+      "step": 346
+    },
+    {
+      "epoch": 0.11141435222347086,
+      "grad_norm": 2.2892768383026123,
+      "learning_rate": 6.939999999999999e-05,
+      "loss": 0.914,
+      "step": 347
+    },
+    {
+      "epoch": 0.11173543104832237,
+      "grad_norm": 2.1729300022125244,
+      "learning_rate": 6.96e-05,
+      "loss": 1.0326,
+      "step": 348
+    },
+    {
+      "epoch": 0.11205650987317387,
+      "grad_norm": 3.1933183670043945,
+      "learning_rate": 6.98e-05,
+      "loss": 0.7324,
+      "step": 349
+    },
+    {
+      "epoch": 0.11237758869802536,
+      "grad_norm": 2.5278120040893555,
+      "learning_rate": 7e-05,
+      "loss": 0.5691,
+      "step": 350
+    },
+    {
+      "epoch": 0.11269866752287687,
+      "grad_norm": 2.725175619125366,
+      "learning_rate": 7.02e-05,
+      "loss": 1.5229,
+      "step": 351
+    },
+    {
+      "epoch": 0.11301974634772836,
+      "grad_norm": 2.0373573303222656,
+      "learning_rate": 7.04e-05,
+      "loss": 1.4782,
+      "step": 352
+    },
+    {
+      "epoch": 0.11334082517257987,
+      "grad_norm": 2.7589669227600098,
+      "learning_rate": 7.06e-05,
+      "loss": 1.0219,
+      "step": 353
+    },
+    {
+      "epoch": 0.11366190399743137,
+      "grad_norm": 3.1588339805603027,
+      "learning_rate": 7.08e-05,
+      "loss": 1.0494,
+      "step": 354
+    },
+    {
+      "epoch": 0.11398298282228288,
+      "grad_norm": 2.983210563659668,
+      "learning_rate": 7.1e-05,
+      "loss": 0.9478,
+      "step": 355
+    },
+    {
+      "epoch": 0.11430406164713437,
+      "grad_norm": 2.4293222427368164,
+      "learning_rate": 7.12e-05,
+      "loss": 0.9519,
+      "step": 356
+    },
+    {
+      "epoch": 0.11462514047198587,
+      "grad_norm": 3.7218642234802246,
+      "learning_rate": 7.14e-05,
+      "loss": 0.8504,
+      "step": 357
+    },
+    {
+      "epoch": 0.11494621929683738,
+      "grad_norm": 2.7284634113311768,
+      "learning_rate": 7.16e-05,
+      "loss": 0.9837,
+      "step": 358
+    },
+    {
+      "epoch": 0.11526729812168887,
+      "grad_norm": 2.264646291732788,
+      "learning_rate": 7.18e-05,
+      "loss": 0.8728,
+      "step": 359
+    },
+    {
+      "epoch": 0.11558837694654038,
+      "grad_norm": 3.814307451248169,
+      "learning_rate": 7.2e-05,
+      "loss": 1.1014,
+      "step": 360
+    },
+    {
+      "epoch": 0.11590945577139188,
+      "grad_norm": 2.3103249073028564,
+      "learning_rate": 7.22e-05,
+      "loss": 1.0027,
+      "step": 361
+    },
+    {
+      "epoch": 0.11623053459624338,
+      "grad_norm": 2.059762716293335,
+      "learning_rate": 7.24e-05,
+      "loss": 1.0825,
+      "step": 362
+    },
+    {
+      "epoch": 0.11655161342109488,
+      "grad_norm": 1.9121147394180298,
+      "learning_rate": 7.26e-05,
+      "loss": 1.0882,
+      "step": 363
+    },
+    {
+      "epoch": 0.11687269224594637,
+      "grad_norm": 1.7068138122558594,
+      "learning_rate": 7.280000000000001e-05,
+      "loss": 1.0976,
+      "step": 364
+    },
+    {
+      "epoch": 0.11719377107079788,
+      "grad_norm": 4.309755802154541,
+      "learning_rate": 7.3e-05,
+      "loss": 1.1305,
+      "step": 365
+    },
+    {
+      "epoch": 0.11751484989564938,
+      "grad_norm": 1.5519227981567383,
+      "learning_rate": 7.32e-05,
+      "loss": 0.9828,
+      "step": 366
+    },
+    {
+      "epoch": 0.11783592872050089,
+      "grad_norm": 2.859467029571533,
+      "learning_rate": 7.340000000000001e-05,
+      "loss": 1.0766,
+      "step": 367
+    },
+    {
+      "epoch": 0.11815700754535238,
+      "grad_norm": 2.033249855041504,
+      "learning_rate": 7.36e-05,
+      "loss": 1.0528,
+      "step": 368
+    },
+    {
+      "epoch": 0.11847808637020389,
+      "grad_norm": 3.5415592193603516,
+      "learning_rate": 7.38e-05,
+      "loss": 1.1399,
+      "step": 369
+    },
+    {
+      "epoch": 0.11879916519505539,
+      "grad_norm": 2.895137071609497,
+      "learning_rate": 7.4e-05,
+      "loss": 1.1532,
+      "step": 370
+    },
+    {
+      "epoch": 0.11912024401990688,
+      "grad_norm": 1.6931350231170654,
+      "learning_rate": 7.42e-05,
+      "loss": 1.1186,
+      "step": 371
+    },
+    {
+      "epoch": 0.11944132284475839,
+      "grad_norm": 1.9246593713760376,
+      "learning_rate": 7.44e-05,
+      "loss": 1.0281,
+      "step": 372
+    },
+    {
+      "epoch": 0.11976240166960989,
+      "grad_norm": 1.789298176765442,
+      "learning_rate": 7.46e-05,
+      "loss": 1.0957,
+      "step": 373
+    },
+    {
+      "epoch": 0.1200834804944614,
+      "grad_norm": 2.8938393592834473,
+      "learning_rate": 7.48e-05,
+      "loss": 0.9665,
+      "step": 374
+    },
+    {
+      "epoch": 0.12040455931931289,
+      "grad_norm": 2.1179943084716797,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 1.1447,
+      "step": 375
+    },
+    {
+      "epoch": 0.1207256381441644,
+      "grad_norm": 1.7801289558410645,
+      "learning_rate": 7.52e-05,
+      "loss": 1.2483,
+      "step": 376
+    },
+    {
+      "epoch": 0.1210467169690159,
+      "grad_norm": 1.6875920295715332,
+      "learning_rate": 7.54e-05,
+      "loss": 1.009,
+      "step": 377
+    },
+    {
+      "epoch": 0.12136779579386739,
+      "grad_norm": 3.599494695663452,
+      "learning_rate": 7.560000000000001e-05,
+      "loss": 1.2534,
+      "step": 378
+    },
+    {
+      "epoch": 0.1216888746187189,
+      "grad_norm": 1.4294577836990356,
+      "learning_rate": 7.58e-05,
+      "loss": 1.1328,
+      "step": 379
+    },
+    {
+      "epoch": 0.1220099534435704,
+      "grad_norm": 1.809747576713562,
+      "learning_rate": 7.6e-05,
+      "loss": 1.0365,
+      "step": 380
+    },
+    {
+      "epoch": 0.1223310322684219,
+      "grad_norm": 1.7651673555374146,
+      "learning_rate": 7.620000000000001e-05,
+      "loss": 1.0842,
+      "step": 381
+    },
+    {
+      "epoch": 0.1226521110932734,
+      "grad_norm": 2.022040843963623,
+      "learning_rate": 7.64e-05,
+      "loss": 1.1518,
+      "step": 382
+    },
+    {
+      "epoch": 0.1229731899181249,
+      "grad_norm": 1.3994489908218384,
+      "learning_rate": 7.66e-05,
+      "loss": 0.9633,
+      "step": 383
+    },
+    {
+      "epoch": 0.1232942687429764,
+      "grad_norm": 1.4892637729644775,
+      "learning_rate": 7.680000000000001e-05,
+      "loss": 1.1681,
+      "step": 384
+    },
+    {
+      "epoch": 0.1236153475678279,
+      "grad_norm": 2.849607467651367,
+      "learning_rate": 7.7e-05,
+      "loss": 1.2844,
+      "step": 385
+    },
+    {
+      "epoch": 0.1239364263926794,
+      "grad_norm": 3.940389394760132,
+      "learning_rate": 7.72e-05,
+      "loss": 1.1795,
+      "step": 386
+    },
+    {
+      "epoch": 0.1242575052175309,
+      "grad_norm": 1.6070865392684937,
+      "learning_rate": 7.740000000000001e-05,
+      "loss": 1.1437,
+      "step": 387
+    },
+    {
+      "epoch": 0.12457858404238241,
+      "grad_norm": 1.3367242813110352,
+      "learning_rate": 7.76e-05,
+      "loss": 0.9778,
+      "step": 388
+    },
+    {
+      "epoch": 0.1248996628672339,
+      "grad_norm": 1.9852445125579834,
+      "learning_rate": 7.780000000000001e-05,
+      "loss": 1.0968,
+      "step": 389
+    },
+    {
+      "epoch": 0.1252207416920854,
+      "grad_norm": 2.6707146167755127,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 1.2846,
+      "step": 390
+    },
+    {
+      "epoch": 0.1255418205169369,
+      "grad_norm": 2.2092552185058594,
+      "learning_rate": 7.82e-05,
+      "loss": 1.0746,
+      "step": 391
+    },
+    {
+      "epoch": 0.1258628993417884,
+      "grad_norm": 2.5755155086517334,
+      "learning_rate": 7.840000000000001e-05,
+      "loss": 0.9431,
+      "step": 392
+    },
+    {
+      "epoch": 0.1261839781666399,
+      "grad_norm": 1.506671667098999,
+      "learning_rate": 7.860000000000001e-05,
+      "loss": 1.0914,
+      "step": 393
+    },
+    {
+      "epoch": 0.12650505699149142,
+      "grad_norm": 1.7302513122558594,
+      "learning_rate": 7.88e-05,
+      "loss": 0.9212,
+      "step": 394
+    },
+    {
+      "epoch": 0.12682613581634292,
+      "grad_norm": 3.166785478591919,
+      "learning_rate": 7.900000000000001e-05,
+      "loss": 1.1882,
+      "step": 395
+    },
+    {
+      "epoch": 0.1271472146411944,
+      "grad_norm": 2.0447499752044678,
+      "learning_rate": 7.920000000000001e-05,
+      "loss": 0.9331,
+      "step": 396
+    },
+    {
+      "epoch": 0.1274682934660459,
+      "grad_norm": 2.1262240409851074,
+      "learning_rate": 7.94e-05,
+      "loss": 0.9565,
+      "step": 397
+    },
+    {
+      "epoch": 0.1277893722908974,
+      "grad_norm": 3.0543501377105713,
+      "learning_rate": 7.960000000000001e-05,
+      "loss": 0.9219,
+      "step": 398
+    },
+    {
+      "epoch": 0.12811045111574892,
+      "grad_norm": 4.1444597244262695,
+      "learning_rate": 7.98e-05,
+      "loss": 0.8325,
+      "step": 399
+    },
+    {
+      "epoch": 0.12843152994060042,
+      "grad_norm": 1.8101000785827637,
+      "learning_rate": 8e-05,
+      "loss": 0.7651,
+      "step": 400
+    },
+    {
+      "epoch": 0.12875260876545191,
+      "grad_norm": 4.419740676879883,
+      "learning_rate": 8.020000000000001e-05,
+      "loss": 1.6615,
+      "step": 401
+    },
+    {
+      "epoch": 0.1290736875903034,
+      "grad_norm": 2.6006617546081543,
+      "learning_rate": 8.04e-05,
+      "loss": 1.6427,
+      "step": 402
+    },
+    {
+      "epoch": 0.12939476641515493,
+      "grad_norm": 3.2348382472991943,
+      "learning_rate": 8.060000000000001e-05,
+      "loss": 1.0184,
+      "step": 403
+    },
+    {
+      "epoch": 0.12971584524000643,
+      "grad_norm": 2.8556830883026123,
+      "learning_rate": 8.080000000000001e-05,
+      "loss": 1.0283,
+      "step": 404
+    },
+    {
+      "epoch": 0.13003692406485792,
+      "grad_norm": 3.4048428535461426,
+      "learning_rate": 8.1e-05,
+      "loss": 1.0695,
+      "step": 405
+    },
+    {
+      "epoch": 0.13035800288970942,
+      "grad_norm": 2.2543652057647705,
+      "learning_rate": 8.120000000000001e-05,
+      "loss": 0.9112,
+      "step": 406
+    },
+    {
+      "epoch": 0.1306790817145609,
+      "grad_norm": 4.042235851287842,
+      "learning_rate": 8.14e-05,
+      "loss": 0.9898,
+      "step": 407
+    },
+    {
+      "epoch": 0.13100016053941244,
+      "grad_norm": 2.8973443508148193,
+      "learning_rate": 8.16e-05,
+      "loss": 0.8473,
+      "step": 408
+    },
+    {
+      "epoch": 0.13132123936426393,
+      "grad_norm": 3.279799461364746,
+      "learning_rate": 8.18e-05,
+      "loss": 1.0893,
+      "step": 409
+    },
+    {
+      "epoch": 0.13164231818911543,
+      "grad_norm": 3.183879852294922,
+      "learning_rate": 8.2e-05,
+      "loss": 1.0263,
+      "step": 410
+    },
+    {
+      "epoch": 0.13196339701396692,
+      "grad_norm": 2.066544532775879,
+      "learning_rate": 8.22e-05,
+      "loss": 1.021,
+      "step": 411
+    },
+    {
+      "epoch": 0.13228447583881842,
+      "grad_norm": 1.7702231407165527,
+      "learning_rate": 8.24e-05,
+      "loss": 1.1344,
+      "step": 412
+    },
+    {
+      "epoch": 0.13260555466366994,
+      "grad_norm": 2.262660026550293,
+      "learning_rate": 8.26e-05,
+      "loss": 1.024,
+      "step": 413
+    },
+    {
+      "epoch": 0.13292663348852143,
+      "grad_norm": 1.833029866218567,
+      "learning_rate": 8.28e-05,
+      "loss": 1.2207,
+      "step": 414
+    },
+    {
+      "epoch": 0.13324771231337293,
+      "grad_norm": 2.144501209259033,
+      "learning_rate": 8.3e-05,
+      "loss": 1.1325,
+      "step": 415
+    },
+    {
+      "epoch": 0.13356879113822442,
+      "grad_norm": 3.092042922973633,
+      "learning_rate": 8.32e-05,
+      "loss": 1.3365,
+      "step": 416
+    },
+    {
+      "epoch": 0.13388986996307595,
+      "grad_norm": 2.136219024658203,
+      "learning_rate": 8.34e-05,
+      "loss": 1.0163,
+      "step": 417
+    },
+    {
+      "epoch": 0.13421094878792744,
+      "grad_norm": 1.7004696130752563,
+      "learning_rate": 8.36e-05,
+      "loss": 0.9517,
+      "step": 418
+    },
+    {
+      "epoch": 0.13453202761277894,
+      "grad_norm": 2.4267795085906982,
+      "learning_rate": 8.38e-05,
+      "loss": 1.0013,
+      "step": 419
+    },
+    {
+      "epoch": 0.13485310643763043,
+      "grad_norm": 2.0210540294647217,
+      "learning_rate": 8.4e-05,
+      "loss": 1.1871,
+      "step": 420
+    },
+    {
+      "epoch": 0.13517418526248193,
+      "grad_norm": 2.205508232116699,
+      "learning_rate": 8.42e-05,
+      "loss": 1.0592,
+      "step": 421
+    },
+    {
+      "epoch": 0.13549526408733345,
+      "grad_norm": 2.3667571544647217,
+      "learning_rate": 8.44e-05,
+      "loss": 0.9743,
+      "step": 422
+    },
+    {
+      "epoch": 0.13581634291218495,
+      "grad_norm": 2.3195512294769287,
+      "learning_rate": 8.46e-05,
+      "loss": 1.1209,
+      "step": 423
+    },
+    {
+      "epoch": 0.13613742173703644,
+      "grad_norm": 2.4299728870391846,
+      "learning_rate": 8.48e-05,
+      "loss": 1.1707,
+      "step": 424
+    },
+    {
+      "epoch": 0.13645850056188794,
+      "grad_norm": 2.341249465942383,
+      "learning_rate": 8.5e-05,
+      "loss": 1.0963,
+      "step": 425
+    },
+    {
+      "epoch": 0.13677957938673946,
+      "grad_norm": 2.3623969554901123,
+      "learning_rate": 8.52e-05,
+      "loss": 1.2913,
+      "step": 426
+    },
+    {
+      "epoch": 0.13710065821159095,
+      "grad_norm": 3.114179849624634,
+      "learning_rate": 8.54e-05,
+      "loss": 1.2193,
+      "step": 427
+    },
+    {
+      "epoch": 0.13742173703644245,
+      "grad_norm": 2.372199058532715,
+      "learning_rate": 8.560000000000001e-05,
+      "loss": 1.1863,
+      "step": 428
+    },
+    {
+      "epoch": 0.13774281586129394,
+      "grad_norm": 2.142880916595459,
+      "learning_rate": 8.58e-05,
+      "loss": 1.0837,
+      "step": 429
+    },
+    {
+      "epoch": 0.13806389468614544,
+      "grad_norm": 2.024472951889038,
+      "learning_rate": 8.6e-05,
+      "loss": 1.036,
+      "step": 430
+    },
+    {
+      "epoch": 0.13838497351099696,
+      "grad_norm": 2.7535400390625,
+      "learning_rate": 8.620000000000001e-05,
+      "loss": 0.9773,
+      "step": 431
+    },
+    {
+      "epoch": 0.13870605233584846,
+      "grad_norm": 2.199688196182251,
+      "learning_rate": 8.64e-05,
+      "loss": 1.0129,
+      "step": 432
+    },
+    {
+      "epoch": 0.13902713116069995,
+      "grad_norm": 2.8626081943511963,
+      "learning_rate": 8.66e-05,
+      "loss": 1.0404,
+      "step": 433
+    },
+    {
+      "epoch": 0.13934820998555145,
+      "grad_norm": 2.4395930767059326,
+      "learning_rate": 8.680000000000001e-05,
+      "loss": 0.9112,
+      "step": 434
+    },
+    {
+      "epoch": 0.13966928881040294,
+      "grad_norm": 2.2529709339141846,
+      "learning_rate": 8.7e-05,
+      "loss": 1.0701,
+      "step": 435
+    },
+    {
+      "epoch": 0.13999036763525446,
+      "grad_norm": 2.0451371669769287,
+      "learning_rate": 8.72e-05,
+      "loss": 1.2465,
+      "step": 436
+    },
+    {
+      "epoch": 0.14031144646010596,
+      "grad_norm": 1.9797896146774292,
+      "learning_rate": 8.740000000000001e-05,
+      "loss": 0.9705,
+      "step": 437
+    },
+    {
+      "epoch": 0.14063252528495745,
+      "grad_norm": 2.4405622482299805,
+      "learning_rate": 8.76e-05,
+      "loss": 1.197,
+      "step": 438
+    },
+    {
+      "epoch": 0.14095360410980895,
+      "grad_norm": 1.761278748512268,
+      "learning_rate": 8.78e-05,
+      "loss": 1.0646,
+      "step": 439
+    },
+    {
+      "epoch": 0.14127468293466047,
+      "grad_norm": 1.8971152305603027,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 0.9596,
+      "step": 440
+    },
+    {
+      "epoch": 0.14159576175951197,
+      "grad_norm": 2.072056531906128,
+      "learning_rate": 8.82e-05,
+      "loss": 0.9762,
+      "step": 441
+    },
+    {
+      "epoch": 0.14191684058436346,
+      "grad_norm": 2.229867935180664,
+      "learning_rate": 8.840000000000001e-05,
+      "loss": 1.1814,
+      "step": 442
+    },
+    {
+      "epoch": 0.14223791940921496,
+      "grad_norm": 1.3738412857055664,
+      "learning_rate": 8.86e-05,
+      "loss": 0.8249,
+      "step": 443
+    },
+    {
+      "epoch": 0.14255899823406645,
+      "grad_norm": 1.5406107902526855,
+      "learning_rate": 8.88e-05,
+      "loss": 0.894,
+      "step": 444
+    },
+    {
+      "epoch": 0.14288007705891798,
+      "grad_norm": 4.295513153076172,
+      "learning_rate": 8.900000000000001e-05,
+      "loss": 0.9362,
+      "step": 445
+    },
+    {
+      "epoch": 0.14320115588376947,
+      "grad_norm": 2.7712790966033936,
+      "learning_rate": 8.92e-05,
+      "loss": 0.8543,
+      "step": 446
+    },
+    {
+      "epoch": 0.14352223470862097,
+      "grad_norm": 2.536055088043213,
+      "learning_rate": 8.94e-05,
+      "loss": 0.9281,
+      "step": 447
+    },
+    {
+      "epoch": 0.14384331353347246,
+      "grad_norm": 3.111720323562622,
+      "learning_rate": 8.960000000000001e-05,
+      "loss": 0.7778,
+      "step": 448
+    },
+    {
+      "epoch": 0.14416439235832396,
+      "grad_norm": 2.919748306274414,
+      "learning_rate": 8.98e-05,
+      "loss": 0.8721,
+      "step": 449
+    },
+    {
+      "epoch": 0.14448547118317548,
+      "grad_norm": 2.425325870513916,
+      "learning_rate": 9e-05,
+      "loss": 0.7305,
+      "step": 450
+    },
+    {
+      "epoch": 0.14480655000802697,
+      "grad_norm": 2.294600009918213,
+      "learning_rate": 9.020000000000001e-05,
+      "loss": 1.6339,
+      "step": 451
+    },
+    {
+      "epoch": 0.14512762883287847,
+      "grad_norm": 2.2173311710357666,
+      "learning_rate": 9.04e-05,
+      "loss": 1.4077,
+      "step": 452
+    },
+    {
+      "epoch": 0.14544870765772996,
+      "grad_norm": 2.2760026454925537,
+      "learning_rate": 9.06e-05,
+      "loss": 1.0451,
+      "step": 453
+    },
+    {
+      "epoch": 0.1457697864825815,
+      "grad_norm": 2.850890636444092,
+      "learning_rate": 9.080000000000001e-05,
+      "loss": 0.9108,
+      "step": 454
+    },
+    {
+      "epoch": 0.14609086530743298,
+      "grad_norm": 2.1252541542053223,
+      "learning_rate": 9.1e-05,
+      "loss": 0.9681,
+      "step": 455
+    },
+    {
+      "epoch": 0.14641194413228448,
+      "grad_norm": 2.2665953636169434,
+      "learning_rate": 9.120000000000001e-05,
+      "loss": 0.9184,
+      "step": 456
+    },
+    {
+      "epoch": 0.14673302295713597,
+      "grad_norm": 3.0463030338287354,
+      "learning_rate": 9.140000000000001e-05,
+      "loss": 0.8551,
+      "step": 457
+    },
+    {
+      "epoch": 0.14705410178198747,
+      "grad_norm": 4.869279384613037,
+      "learning_rate": 9.16e-05,
+      "loss": 1.1517,
+      "step": 458
+    },
+    {
+      "epoch": 0.147375180606839,
+      "grad_norm": 2.495591878890991,
+      "learning_rate": 9.180000000000001e-05,
+      "loss": 1.264,
+      "step": 459
+    },
+    {
+      "epoch": 0.14769625943169049,
+      "grad_norm": 1.5504367351531982,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 1.1832,
+      "step": 460
+    },
+    {
+      "epoch": 0.14801733825654198,
+      "grad_norm": 1.7231016159057617,
+      "learning_rate": 9.22e-05,
+      "loss": 1.0143,
+      "step": 461
+    },
+    {
+      "epoch": 0.14833841708139348,
+      "grad_norm": 2.187086343765259,
+      "learning_rate": 9.240000000000001e-05,
+      "loss": 1.2475,
+      "step": 462
+    },
+    {
+      "epoch": 0.14865949590624497,
+      "grad_norm": 2.143453359603882,
+      "learning_rate": 9.260000000000001e-05,
+      "loss": 1.1353,
+      "step": 463
+    },
+    {
+      "epoch": 0.1489805747310965,
+      "grad_norm": 2.101834774017334,
+      "learning_rate": 9.28e-05,
+      "loss": 1.2927,
+      "step": 464
+    },
+    {
+      "epoch": 0.149301653555948,
+      "grad_norm": 2.4847493171691895,
+      "learning_rate": 9.300000000000001e-05,
+      "loss": 1.2384,
+      "step": 465
+    },
+    {
+      "epoch": 0.14962273238079948,
+      "grad_norm": 2.28727650642395,
+      "learning_rate": 9.320000000000002e-05,
+      "loss": 1.0456,
+      "step": 466
+    },
+    {
+      "epoch": 0.14994381120565098,
+      "grad_norm": 1.9797146320343018,
+      "learning_rate": 9.340000000000001e-05,
+      "loss": 1.1047,
+      "step": 467
+    },
+    {
+      "epoch": 0.1502648900305025,
+      "grad_norm": 2.9211812019348145,
+      "learning_rate": 9.360000000000001e-05,
+      "loss": 1.0695,
+      "step": 468
+    },
+    {
+      "epoch": 0.150585968855354,
+      "grad_norm": 1.7299671173095703,
+      "learning_rate": 9.38e-05,
+      "loss": 0.8681,
+      "step": 469
+    },
+    {
+      "epoch": 0.1509070476802055,
+      "grad_norm": 2.1782050132751465,
+      "learning_rate": 9.4e-05,
+      "loss": 1.1974,
+      "step": 470
+    },
+    {
+      "epoch": 0.151228126505057,
+      "grad_norm": 4.492698669433594,
+      "learning_rate": 9.42e-05,
+      "loss": 1.0819,
+      "step": 471
+    },
+    {
+      "epoch": 0.15154920532990848,
+      "grad_norm": 2.4074699878692627,
+      "learning_rate": 9.44e-05,
+      "loss": 1.1869,
+      "step": 472
+    },
+    {
+      "epoch": 0.15187028415476,
+      "grad_norm": 2.0078368186950684,
+      "learning_rate": 9.46e-05,
+      "loss": 1.0196,
+      "step": 473
+    },
+    {
+      "epoch": 0.1521913629796115,
+      "grad_norm": 1.4234404563903809,
+      "learning_rate": 9.48e-05,
+      "loss": 1.1275,
+      "step": 474
+    },
+    {
+      "epoch": 0.152512441804463,
+      "grad_norm": 1.73353111743927,
+      "learning_rate": 9.5e-05,
+      "loss": 1.103,
+      "step": 475
+    },
+    {
+      "epoch": 0.1528335206293145,
+      "grad_norm": 2.3276383876800537,
+      "learning_rate": 9.52e-05,
+      "loss": 1.0127,
+      "step": 476
+    },
+    {
+      "epoch": 0.15315459945416598,
+      "grad_norm": 2.4894511699676514,
+      "learning_rate": 9.54e-05,
+      "loss": 1.3882,
+      "step": 477
+    },
+    {
+      "epoch": 0.1534756782790175,
+      "grad_norm": 2.496798038482666,
+      "learning_rate": 9.56e-05,
+      "loss": 0.9479,
+      "step": 478
+    },
+    {
+      "epoch": 0.153796757103869,
+      "grad_norm": 1.701995611190796,
+      "learning_rate": 9.58e-05,
+      "loss": 1.0526,
+      "step": 479
+    },
+    {
+      "epoch": 0.1541178359287205,
+      "grad_norm": 1.432371973991394,
+      "learning_rate": 9.6e-05,
+      "loss": 1.1382,
+      "step": 480
+    },
+    {
+      "epoch": 0.154438914753572,
+      "grad_norm": 2.706610918045044,
+      "learning_rate": 9.620000000000001e-05,
+      "loss": 0.9633,
+      "step": 481
+    },
+    {
+      "epoch": 0.15475999357842352,
+      "grad_norm": 2.06569504737854,
+      "learning_rate": 9.64e-05,
+      "loss": 0.9708,
+      "step": 482
+    },
+    {
+      "epoch": 0.155081072403275,
+      "grad_norm": 2.1104109287261963,
+      "learning_rate": 9.66e-05,
+      "loss": 0.9835,
+      "step": 483
+    },
+    {
+      "epoch": 0.1554021512281265,
+      "grad_norm": 1.5272530317306519,
+      "learning_rate": 9.680000000000001e-05,
+      "loss": 1.0398,
+      "step": 484
+    },
+    {
+      "epoch": 0.155723230052978,
+      "grad_norm": 2.510329246520996,
+      "learning_rate": 9.7e-05,
+      "loss": 1.1728,
+      "step": 485
+    },
+    {
+      "epoch": 0.1560443088778295,
+      "grad_norm": 1.6082532405853271,
+      "learning_rate": 9.72e-05,
+      "loss": 1.1114,
+      "step": 486
+    },
+    {
+      "epoch": 0.15636538770268102,
+      "grad_norm": 1.8710626363754272,
+      "learning_rate": 9.74e-05,
+      "loss": 0.8267,
+      "step": 487
+    },
+    {
+      "epoch": 0.15668646652753251,
+      "grad_norm": 1.3604836463928223,
+      "learning_rate": 9.76e-05,
+      "loss": 0.9311,
+      "step": 488
+    },
+    {
+      "epoch": 0.157007545352384,
+      "grad_norm": 1.7233448028564453,
+      "learning_rate": 9.78e-05,
+      "loss": 1.0344,
+      "step": 489
+    },
+    {
+      "epoch": 0.1573286241772355,
+      "grad_norm": 1.827020287513733,
+      "learning_rate": 9.8e-05,
+      "loss": 1.0538,
+      "step": 490
+    },
+    {
+      "epoch": 0.157649703002087,
+      "grad_norm": 2.838731050491333,
+      "learning_rate": 9.82e-05,
+      "loss": 1.0937,
+      "step": 491
+    },
+    {
+      "epoch": 0.15797078182693852,
+      "grad_norm": 2.2758519649505615,
+      "learning_rate": 9.84e-05,
+      "loss": 1.0409,
+      "step": 492
+    },
+    {
+      "epoch": 0.15829186065179002,
+      "grad_norm": 3.4251341819763184,
+      "learning_rate": 9.86e-05,
+      "loss": 1.0816,
+      "step": 493
+    },
+    {
+      "epoch": 0.1586129394766415,
+      "grad_norm": 2.235792398452759,
+      "learning_rate": 9.88e-05,
+      "loss": 0.9386,
+      "step": 494
+    },
+    {
+      "epoch": 0.158934018301493,
+      "grad_norm": 1.5516377687454224,
+      "learning_rate": 9.900000000000001e-05,
+      "loss": 1.0178,
+      "step": 495
+    },
+    {
+      "epoch": 0.15925509712634453,
+      "grad_norm": 2.0881407260894775,
+      "learning_rate": 9.92e-05,
+      "loss": 0.893,
+      "step": 496
+    },
+    {
+      "epoch": 0.15957617595119603,
+      "grad_norm": 2.0493526458740234,
+      "learning_rate": 9.94e-05,
+      "loss": 0.9658,
+      "step": 497
+    },
+    {
+      "epoch": 0.15989725477604752,
+      "grad_norm": 1.7837196588516235,
+      "learning_rate": 9.960000000000001e-05,
+      "loss": 0.7789,
+      "step": 498
+    },
+    {
+      "epoch": 0.16021833360089902,
+      "grad_norm": 2.1364388465881348,
+      "learning_rate": 9.98e-05,
+      "loss": 0.586,
+      "step": 499
+    },
+    {
+      "epoch": 0.1605394124257505,
+      "grad_norm": 2.1484198570251465,
+      "learning_rate": 0.0001,
+      "loss": 0.6172,
+      "step": 500
+    },
+    {
+      "epoch": 0.16086049125060203,
+      "grad_norm": 6.418078899383545,
+      "learning_rate": 0.00010020000000000001,
+      "loss": 1.499,
+      "step": 501
+    },
+    {
+      "epoch": 0.16118157007545353,
+      "grad_norm": 3.4180362224578857,
+      "learning_rate": 0.0001004,
+      "loss": 1.1029,
+      "step": 502
+    },
+    {
+      "epoch": 0.16150264890030502,
+      "grad_norm": 3.427628517150879,
+      "learning_rate": 0.0001006,
+      "loss": 1.1054,
+      "step": 503
+    },
+    {
+      "epoch": 0.16182372772515652,
+      "grad_norm": 3.6065824031829834,
+      "learning_rate": 0.00010080000000000001,
+      "loss": 0.9196,
+      "step": 504
+    },
+    {
+      "epoch": 0.162144806550008,
+      "grad_norm": 2.572456121444702,
+      "learning_rate": 0.000101,
+      "loss": 0.9429,
+      "step": 505
+    },
+    {
+      "epoch": 0.16246588537485954,
+      "grad_norm": 2.5031585693359375,
+      "learning_rate": 0.00010120000000000001,
+      "loss": 0.8701,
+      "step": 506
+    },
+    {
+      "epoch": 0.16278696419971103,
+      "grad_norm": 2.1001455783843994,
+      "learning_rate": 0.00010140000000000001,
+      "loss": 0.992,
+      "step": 507
+    },
+    {
+      "epoch": 0.16310804302456253,
+      "grad_norm": 2.580418109893799,
+      "learning_rate": 0.0001016,
+      "loss": 1.0474,
+      "step": 508
+    },
+    {
+      "epoch": 0.16342912184941402,
+      "grad_norm": 2.187901496887207,
+      "learning_rate": 0.00010180000000000001,
+      "loss": 1.1151,
+      "step": 509
+    },
+    {
+      "epoch": 0.16375020067426554,
+      "grad_norm": 1.5737202167510986,
+      "learning_rate": 0.00010200000000000001,
+      "loss": 1.0749,
+      "step": 510
+    },
+    {
+      "epoch": 0.16407127949911704,
+      "grad_norm": 2.0835020542144775,
+      "learning_rate": 0.0001022,
+      "loss": 0.7957,
+      "step": 511
+    },
+    {
+      "epoch": 0.16439235832396853,
+      "grad_norm": 2.3619582653045654,
+      "learning_rate": 0.00010240000000000001,
+      "loss": 0.8308,
+      "step": 512
+    },
+    {
+      "epoch": 0.16471343714882003,
+      "grad_norm": 1.7399780750274658,
+      "learning_rate": 0.00010260000000000001,
+      "loss": 1.1296,
+      "step": 513
+    },
+    {
+      "epoch": 0.16503451597367153,
+      "grad_norm": 2.110400676727295,
+      "learning_rate": 0.0001028,
+      "loss": 0.8289,
+      "step": 514
+    },
+    {
+      "epoch": 0.16535559479852305,
+      "grad_norm": 1.960696816444397,
+      "learning_rate": 0.00010300000000000001,
+      "loss": 1.1581,
+      "step": 515
+    },
+    {
+      "epoch": 0.16567667362337454,
+      "grad_norm": 1.2459540367126465,
+      "learning_rate": 0.0001032,
+      "loss": 1.0465,
+      "step": 516
+    },
+    {
+      "epoch": 0.16599775244822604,
+      "grad_norm": 1.4018001556396484,
+      "learning_rate": 0.0001034,
+      "loss": 0.9106,
+      "step": 517
+    },
+    {
+      "epoch": 0.16631883127307753,
+      "grad_norm": 1.3802064657211304,
+      "learning_rate": 0.00010360000000000001,
+      "loss": 1.0696,
+      "step": 518
+    },
+    {
+      "epoch": 0.16663991009792903,
+      "grad_norm": 5.539737224578857,
+      "learning_rate": 0.0001038,
+      "loss": 1.0406,
+      "step": 519
+    },
+    {
+      "epoch": 0.16696098892278055,
+      "grad_norm": 2.785435676574707,
+      "learning_rate": 0.00010400000000000001,
+      "loss": 0.9252,
+      "step": 520
+    },
+    {
+      "epoch": 0.16728206774763205,
+      "grad_norm": 1.6449223756790161,
+      "learning_rate": 0.00010420000000000001,
+      "loss": 0.9388,
+      "step": 521
+    },
+    {
+      "epoch": 0.16760314657248354,
+      "grad_norm": 1.6841291189193726,
+      "learning_rate": 0.0001044,
+      "loss": 0.9676,
+      "step": 522
+    },
+    {
+      "epoch": 0.16792422539733504,
+      "grad_norm": 2.328427314758301,
+      "learning_rate": 0.00010460000000000001,
+      "loss": 0.9967,
+      "step": 523
+    },
+    {
+      "epoch": 0.16824530422218656,
+      "grad_norm": 1.5015443563461304,
+      "learning_rate": 0.00010480000000000001,
+      "loss": 0.9851,
+      "step": 524
+    },
+    {
+      "epoch": 0.16856638304703805,
+      "grad_norm": 1.587814450263977,
+      "learning_rate": 0.000105,
+      "loss": 0.9411,
+      "step": 525
+    },
+    {
+      "epoch": 0.16888746187188955,
+      "grad_norm": 2.9774792194366455,
+      "learning_rate": 0.00010520000000000001,
+      "loss": 0.9465,
+      "step": 526
+    },
+    {
+      "epoch": 0.16920854069674104,
+      "grad_norm": 2.033804416656494,
+      "learning_rate": 0.00010540000000000001,
+      "loss": 0.991,
+      "step": 527
+    },
+    {
+      "epoch": 0.16952961952159254,
+      "grad_norm": 2.3683605194091797,
+      "learning_rate": 0.0001056,
+      "loss": 1.0893,
+      "step": 528
+    },
+    {
+      "epoch": 0.16985069834644406,
+      "grad_norm": 2.22643780708313,
+      "learning_rate": 0.00010580000000000001,
+      "loss": 1.1218,
+      "step": 529
+    },
+    {
+      "epoch": 0.17017177717129556,
+      "grad_norm": 2.789557933807373,
+      "learning_rate": 0.00010600000000000002,
+      "loss": 1.0919,
+      "step": 530
+    },
+    {
+      "epoch": 0.17049285599614705,
+      "grad_norm": 1.379966378211975,
+      "learning_rate": 0.0001062,
+      "loss": 1.1469,
+      "step": 531
+    },
+    {
+      "epoch": 0.17081393482099855,
+      "grad_norm": 1.87637197971344,
+      "learning_rate": 0.00010640000000000001,
+      "loss": 1.0153,
+      "step": 532
+    },
+    {
+      "epoch": 0.17113501364585004,
+      "grad_norm": 1.728216528892517,
+      "learning_rate": 0.00010660000000000002,
+      "loss": 1.0824,
+      "step": 533
+    },
+    {
+      "epoch": 0.17145609247070157,
+      "grad_norm": 1.858970046043396,
+      "learning_rate": 0.00010680000000000001,
+      "loss": 1.0331,
+      "step": 534
+    },
+    {
+      "epoch": 0.17177717129555306,
+      "grad_norm": 2.390653133392334,
+      "learning_rate": 0.00010700000000000001,
+      "loss": 1.2246,
+      "step": 535
+    },
+    {
+      "epoch": 0.17209825012040456,
+      "grad_norm": 1.582828402519226,
+      "learning_rate": 0.00010720000000000002,
+      "loss": 1.0276,
+      "step": 536
+    },
+    {
+      "epoch": 0.17241932894525605,
+      "grad_norm": 1.6667520999908447,
+      "learning_rate": 0.00010740000000000001,
+      "loss": 0.899,
+      "step": 537
+    },
+    {
+      "epoch": 0.17274040777010757,
+      "grad_norm": 2.3515865802764893,
+      "learning_rate": 0.00010760000000000001,
+      "loss": 0.9804,
+      "step": 538
+    },
+    {
+      "epoch": 0.17306148659495907,
+      "grad_norm": 2.0802536010742188,
+      "learning_rate": 0.00010780000000000002,
+      "loss": 1.0002,
+      "step": 539
+    },
+    {
+      "epoch": 0.17338256541981056,
+      "grad_norm": 1.2941781282424927,
+      "learning_rate": 0.00010800000000000001,
+      "loss": 1.1026,
+      "step": 540
+    },
+    {
+      "epoch": 0.17370364424466206,
+      "grad_norm": 1.7126529216766357,
+      "learning_rate": 0.00010820000000000001,
+      "loss": 1.1084,
+      "step": 541
+    },
+    {
+      "epoch": 0.17402472306951355,
+      "grad_norm": 1.5774749517440796,
+      "learning_rate": 0.00010840000000000002,
+      "loss": 0.99,
+      "step": 542
+    },
+    {
+      "epoch": 0.17434580189436508,
+      "grad_norm": 1.576080322265625,
+      "learning_rate": 0.00010860000000000001,
+      "loss": 0.9603,
+      "step": 543
+    },
+    {
+      "epoch": 0.17466688071921657,
+      "grad_norm": 1.4444918632507324,
+      "learning_rate": 0.00010880000000000002,
+      "loss": 1.0226,
+      "step": 544
+    },
+    {
+      "epoch": 0.17498795954406807,
+      "grad_norm": 2.149320363998413,
+      "learning_rate": 0.000109,
+      "loss": 0.8809,
+      "step": 545
+    },
+    {
+      "epoch": 0.17530903836891956,
+      "grad_norm": 2.1339616775512695,
+      "learning_rate": 0.00010920000000000001,
+      "loss": 0.9411,
+      "step": 546
+    },
+    {
+      "epoch": 0.17563011719377106,
+      "grad_norm": 3.13826847076416,
+      "learning_rate": 0.00010940000000000002,
+      "loss": 0.9321,
+      "step": 547
+    },
+    {
+      "epoch": 0.17595119601862258,
+      "grad_norm": 2.3245761394500732,
+      "learning_rate": 0.00010960000000000001,
+      "loss": 0.8781,
+      "step": 548
+    },
+    {
+      "epoch": 0.17627227484347407,
+      "grad_norm": 1.5890634059906006,
+      "learning_rate": 0.00010980000000000001,
+      "loss": 0.8076,
+      "step": 549
+    },
+    {
+      "epoch": 0.17659335366832557,
+      "grad_norm": 1.824544906616211,
+      "learning_rate": 0.00011000000000000002,
+      "loss": 0.5724,
+      "step": 550
+    },
+    {
+      "epoch": 0.17691443249317707,
+      "grad_norm": 3.6526408195495605,
+      "learning_rate": 0.00011020000000000001,
+      "loss": 1.5191,
+      "step": 551
+    },
+    {
+      "epoch": 0.1772355113180286,
+      "grad_norm": 3.4367077350616455,
+      "learning_rate": 0.00011040000000000001,
+      "loss": 1.2869,
+      "step": 552
+    },
+    {
+      "epoch": 0.17755659014288008,
+      "grad_norm": 3.5046842098236084,
+      "learning_rate": 0.00011060000000000002,
+      "loss": 1.0015,
+      "step": 553
+    },
+    {
+      "epoch": 0.17787766896773158,
+      "grad_norm": 3.2727510929107666,
+      "learning_rate": 0.00011080000000000001,
+      "loss": 0.9038,
+      "step": 554
+    },
+    {
+      "epoch": 0.17819874779258307,
+      "grad_norm": 2.613293409347534,
+      "learning_rate": 0.00011100000000000001,
+      "loss": 0.9512,
+      "step": 555
+    },
+    {
+      "epoch": 0.17851982661743457,
+      "grad_norm": 4.011002540588379,
+      "learning_rate": 0.00011120000000000002,
+      "loss": 0.8484,
+      "step": 556
+    },
+    {
+      "epoch": 0.1788409054422861,
+      "grad_norm": 3.158597707748413,
+      "learning_rate": 0.00011140000000000001,
+      "loss": 1.0517,
+      "step": 557
+    },
+    {
+      "epoch": 0.1791619842671376,
+      "grad_norm": 2.143503427505493,
+      "learning_rate": 0.00011160000000000002,
+      "loss": 1.0966,
+      "step": 558
+    },
+    {
+      "epoch": 0.17948306309198908,
+      "grad_norm": 2.083723545074463,
+      "learning_rate": 0.00011180000000000002,
+      "loss": 1.0665,
+      "step": 559
+    },
+    {
+      "epoch": 0.17980414191684058,
+      "grad_norm": 1.7016173601150513,
+      "learning_rate": 0.00011200000000000001,
+      "loss": 0.9647,
+      "step": 560
+    },
+    {
+      "epoch": 0.1801252207416921,
+      "grad_norm": 3.817215919494629,
+      "learning_rate": 0.00011220000000000002,
+      "loss": 0.7164,
+      "step": 561
+    },
+    {
+      "epoch": 0.1804462995665436,
+      "grad_norm": 2.013831853866577,
+      "learning_rate": 0.00011240000000000002,
+      "loss": 1.135,
+      "step": 562
+    },
+    {
+      "epoch": 0.1807673783913951,
+      "grad_norm": 2.0431296825408936,
+      "learning_rate": 0.0001126,
+      "loss": 1.1298,
+      "step": 563
+    },
+    {
+      "epoch": 0.18108845721624658,
+      "grad_norm": 2.106182813644409,
+      "learning_rate": 0.00011279999999999999,
+      "loss": 1.1183,
+      "step": 564
+    },
+    {
+      "epoch": 0.18140953604109808,
+      "grad_norm": 2.747995615005493,
+      "learning_rate": 0.000113,
+      "loss": 0.9181,
+      "step": 565
+    },
+    {
+      "epoch": 0.1817306148659496,
+      "grad_norm": 2.186434268951416,
+      "learning_rate": 0.0001132,
+      "loss": 1.3199,
+      "step": 566
+    },
+    {
+      "epoch": 0.1820516936908011,
+      "grad_norm": 2.072404146194458,
+      "learning_rate": 0.00011339999999999999,
+      "loss": 0.9275,
+      "step": 567
+    },
+    {
+      "epoch": 0.1823727725156526,
+      "grad_norm": 2.051323890686035,
+      "learning_rate": 0.0001136,
+      "loss": 0.9817,
+      "step": 568
+    },
+    {
+      "epoch": 0.1826938513405041,
+      "grad_norm": 2.325993061065674,
+      "learning_rate": 0.0001138,
+      "loss": 1.2054,
+      "step": 569
+    },
+    {
+      "epoch": 0.18301493016535558,
+      "grad_norm": 2.8382861614227295,
+      "learning_rate": 0.00011399999999999999,
+      "loss": 1.0196,
+      "step": 570
+    },
+    {
+      "epoch": 0.1833360089902071,
+      "grad_norm": 1.9472187757492065,
+      "learning_rate": 0.0001142,
+      "loss": 1.028,
+      "step": 571
+    },
+    {
+      "epoch": 0.1836570878150586,
+      "grad_norm": 2.2961337566375732,
+      "learning_rate": 0.0001144,
+      "loss": 1.1601,
+      "step": 572
+    },
+    {
+      "epoch": 0.1839781666399101,
+      "grad_norm": 1.6054937839508057,
+      "learning_rate": 0.0001146,
+      "loss": 0.9415,
+      "step": 573
+    },
+    {
+      "epoch": 0.1842992454647616,
+      "grad_norm": 1.5658379793167114,
+      "learning_rate": 0.0001148,
+      "loss": 0.8921,
+      "step": 574
+    },
+    {
+      "epoch": 0.1846203242896131,
+      "grad_norm": 2.6736388206481934,
+      "learning_rate": 0.00011499999999999999,
+      "loss": 1.036,
+      "step": 575
+    },
+    {
+      "epoch": 0.1849414031144646,
+      "grad_norm": 2.179887056350708,
+      "learning_rate": 0.0001152,
+      "loss": 1.3059,
+      "step": 576
+    },
+    {
+      "epoch": 0.1852624819393161,
+      "grad_norm": 1.418059229850769,
+      "learning_rate": 0.0001154,
+      "loss": 1.0964,
+      "step": 577
+    },
+    {
+      "epoch": 0.1855835607641676,
+      "grad_norm": 2.5086400508880615,
+      "learning_rate": 0.00011559999999999999,
+      "loss": 1.1934,
+      "step": 578
+    },
+    {
+      "epoch": 0.1859046395890191,
+      "grad_norm": 2.1473209857940674,
+      "learning_rate": 0.0001158,
+      "loss": 1.0245,
+      "step": 579
+    },
+    {
+      "epoch": 0.18622571841387062,
+      "grad_norm": 4.459160804748535,
+      "learning_rate": 0.000116,
+      "loss": 1.3467,
+      "step": 580
+    },
+    {
+      "epoch": 0.1865467972387221,
+      "grad_norm": 2.073753833770752,
+      "learning_rate": 0.00011619999999999999,
+      "loss": 0.8753,
+      "step": 581
+    },
+    {
+      "epoch": 0.1868678760635736,
+      "grad_norm": 1.9933395385742188,
+      "learning_rate": 0.0001164,
+      "loss": 0.7768,
+      "step": 582
+    },
+    {
+      "epoch": 0.1871889548884251,
+      "grad_norm": 2.649278402328491,
+      "learning_rate": 0.0001166,
+      "loss": 1.2715,
+      "step": 583
+    },
+    {
+      "epoch": 0.1875100337132766,
+      "grad_norm": 3.018387794494629,
+      "learning_rate": 0.00011679999999999999,
+      "loss": 1.139,
+      "step": 584
+    },
+    {
+      "epoch": 0.18783111253812812,
+      "grad_norm": 2.405787467956543,
+      "learning_rate": 0.000117,
+      "loss": 1.1507,
+      "step": 585
+    },
+    {
+      "epoch": 0.18815219136297961,
+      "grad_norm": 2.1480977535247803,
+      "learning_rate": 0.0001172,
+      "loss": 1.1337,
+      "step": 586
+    },
+    {
+      "epoch": 0.1884732701878311,
+      "grad_norm": NaN,
+      "learning_rate": 0.0001172,
+      "loss": 1.233,
+      "step": 587
+    },
+    {
+      "epoch": 0.1887943490126826,
+      "grad_norm": 1.5942261219024658,
+      "learning_rate": 0.0001174,
+      "loss": 0.874,
+      "step": 588
+    },
+    {
+      "epoch": 0.18911542783753413,
+      "grad_norm": 3.029067039489746,
+      "learning_rate": 0.0001176,
+      "loss": 0.9061,
+      "step": 589
+    },
+    {
+      "epoch": 0.18943650666238562,
+      "grad_norm": 1.6258342266082764,
+      "learning_rate": 0.0001178,
+      "loss": 0.8887,
+      "step": 590
+    },
+    {
+      "epoch": 0.18975758548723712,
+      "grad_norm": 1.8686645030975342,
+      "learning_rate": 0.000118,
+      "loss": 0.887,
+      "step": 591
+    },
+    {
+      "epoch": 0.1900786643120886,
+      "grad_norm": 2.315640687942505,
+      "learning_rate": 0.0001182,
+      "loss": 1.2327,
+      "step": 592
+    },
+    {
+      "epoch": 0.1903997431369401,
+      "grad_norm": 2.111231565475464,
+      "learning_rate": 0.0001184,
+      "loss": 1.2629,
+      "step": 593
+    },
+    {
+      "epoch": 0.19072082196179163,
+      "grad_norm": 2.422922372817993,
+      "learning_rate": 0.0001186,
+      "loss": 0.9423,
+      "step": 594
+    },
+    {
+      "epoch": 0.19104190078664313,
+      "grad_norm": 2.0617730617523193,
+      "learning_rate": 0.0001188,
+      "loss": 1.0237,
+      "step": 595
+    },
+    {
+      "epoch": 0.19136297961149462,
+      "grad_norm": 2.295034408569336,
+      "learning_rate": 0.000119,
+      "loss": 0.993,
+      "step": 596
+    },
+    {
+      "epoch": 0.19168405843634612,
+      "grad_norm": 1.7605993747711182,
+      "learning_rate": 0.0001192,
+      "loss": 1.0313,
+      "step": 597
+    },
+    {
+      "epoch": 0.1920051372611976,
+      "grad_norm": 1.9907429218292236,
+      "learning_rate": 0.0001194,
+      "loss": 0.9979,
+      "step": 598
+    },
+    {
+      "epoch": 0.19232621608604913,
+      "grad_norm": 2.0522098541259766,
+      "learning_rate": 0.00011960000000000001,
+      "loss": 0.8497,
+      "step": 599
+    },
+    {
+      "epoch": 0.19264729491090063,
+      "grad_norm": 1.6448557376861572,
+      "learning_rate": 0.0001198,
+      "loss": 0.6443,
+      "step": 600
+    },
+    {
+      "epoch": 0.19296837373575212,
+      "grad_norm": 95.2552719116211,
+      "learning_rate": 0.00012,
+      "loss": 1.6446,
+      "step": 601
+    },
+    {
+      "epoch": 0.19328945256060362,
+      "grad_norm": 50.20346450805664,
+      "learning_rate": 0.00012020000000000001,
+      "loss": 1.4505,
+      "step": 602
+    },
+    {
+      "epoch": 0.19361053138545514,
+      "grad_norm": 3.067426919937134,
+      "learning_rate": 0.0001204,
+      "loss": 1.0486,
+      "step": 603
+    },
+    {
+      "epoch": 0.19393161021030664,
+      "grad_norm": 3.5429553985595703,
+      "learning_rate": 0.0001206,
+      "loss": 0.9873,
+      "step": 604
+    },
+    {
+      "epoch": 0.19425268903515813,
+      "grad_norm": 3.2231061458587646,
+      "learning_rate": 0.0001208,
+      "loss": 0.9301,
+      "step": 605
+    },
+    {
+      "epoch": 0.19457376786000963,
+      "grad_norm": 2.5521445274353027,
+      "learning_rate": 0.000121,
+      "loss": 0.8936,
+      "step": 606
+    },
+    {
+      "epoch": 0.19489484668486112,
+      "grad_norm": 2.4326229095458984,
+      "learning_rate": 0.0001212,
+      "loss": 0.9659,
+      "step": 607
+    },
+    {
+      "epoch": 0.19521592550971265,
+      "grad_norm": 4.246782302856445,
+      "learning_rate": 0.0001214,
+      "loss": 1.0794,
+      "step": 608
+    },
+    {
+      "epoch": 0.19553700433456414,
+      "grad_norm": 3.502556562423706,
+      "learning_rate": 0.0001216,
+      "loss": 1.0255,
+      "step": 609
+    },
+    {
+      "epoch": 0.19585808315941564,
+      "grad_norm": 2.036647319793701,
+      "learning_rate": 0.0001218,
+      "loss": 0.8848,
+      "step": 610
+    },
+    {
+      "epoch": 0.19617916198426713,
+      "grad_norm": 2.284869432449341,
+      "learning_rate": 0.000122,
+      "loss": 0.9834,
+      "step": 611
+    },
+    {
+      "epoch": 0.19650024080911863,
+      "grad_norm": 1.4191306829452515,
+      "learning_rate": 0.00012220000000000002,
+      "loss": 0.945,
+      "step": 612
+    },
+    {
+      "epoch": 0.19682131963397015,
+      "grad_norm": 2.529686450958252,
+      "learning_rate": 0.0001224,
+      "loss": 1.0871,
+      "step": 613
+    },
+    {
+      "epoch": 0.19714239845882164,
+      "grad_norm": 1.5338515043258667,
+      "learning_rate": 0.0001226,
+      "loss": 0.9792,
+      "step": 614
+    },
+    {
+      "epoch": 0.19746347728367314,
+      "grad_norm": 1.4318336248397827,
+      "learning_rate": 0.0001228,
+      "loss": 1.0498,
+      "step": 615
+    },
+    {
+      "epoch": 0.19778455610852463,
+      "grad_norm": 1.7187212705612183,
+      "learning_rate": 0.000123,
+      "loss": 1.0418,
+      "step": 616
+    },
+    {
+      "epoch": 0.19810563493337616,
+      "grad_norm": 1.5892479419708252,
+      "learning_rate": 0.0001232,
+      "loss": 1.1123,
+      "step": 617
+    },
+    {
+      "epoch": 0.19842671375822765,
+      "grad_norm": 2.3069276809692383,
+      "learning_rate": 0.00012340000000000002,
+      "loss": 0.8825,
+      "step": 618
+    },
+    {
+      "epoch": 0.19874779258307915,
+      "grad_norm": 2.5315756797790527,
+      "learning_rate": 0.0001236,
+      "loss": 0.9985,
+      "step": 619
+    },
+    {
+      "epoch": 0.19906887140793064,
+      "grad_norm": 3.1710851192474365,
+      "learning_rate": 0.0001238,
+      "loss": 1.0591,
+      "step": 620
+    },
+    {
+      "epoch": 0.19938995023278214,
+      "grad_norm": 3.0601823329925537,
+      "learning_rate": 0.000124,
+      "loss": 1.1456,
+      "step": 621
+    },
+    {
+      "epoch": 0.19971102905763366,
+      "grad_norm": 4.170618057250977,
+      "learning_rate": 0.0001242,
+      "loss": 1.0819,
+      "step": 622
+    },
+    {
+      "epoch": 0.20003210788248516,
+      "grad_norm": 6.973270416259766,
+      "learning_rate": 0.00012440000000000002,
+      "loss": 1.0412,
+      "step": 623
+    },
+    {
+      "epoch": 0.20035318670733665,
+      "grad_norm": 1.518310546875,
+      "learning_rate": 0.0001246,
+      "loss": 0.9291,
+      "step": 624
+    },
+    {
+      "epoch": 0.20067426553218815,
+      "grad_norm": 1.8494837284088135,
+      "learning_rate": 0.0001248,
+      "loss": 0.8893,
+      "step": 625
+    },
+    {
+      "epoch": 0.20099534435703964,
+      "grad_norm": 1.8757221698760986,
+      "learning_rate": 0.000125,
+      "loss": 0.9862,
+      "step": 626
+    },
+    {
+      "epoch": 0.20131642318189116,
+      "grad_norm": 2.3286845684051514,
+      "learning_rate": 0.0001252,
+      "loss": 1.3974,
+      "step": 627
+    },
+    {
+      "epoch": 0.20163750200674266,
+      "grad_norm": 2.6092865467071533,
+      "learning_rate": 0.0001254,
+      "loss": 1.3019,
+      "step": 628
+    },
+    {
+      "epoch": 0.20195858083159415,
+      "grad_norm": 2.212606906890869,
+      "learning_rate": 0.00012560000000000002,
+      "loss": 1.1727,
+      "step": 629
+    },
+    {
+      "epoch": 0.20227965965644565,
+      "grad_norm": 1.5337756872177124,
+      "learning_rate": 0.0001258,
+      "loss": 1.058,
+      "step": 630
+    },
+    {
+      "epoch": 0.20260073848129717,
+      "grad_norm": 2.063206672668457,
+      "learning_rate": 0.000126,
+      "loss": 1.1522,
+      "step": 631
+    },
+    {
+      "epoch": 0.20292181730614867,
+      "grad_norm": 2.371931552886963,
+      "learning_rate": 0.0001262,
+      "loss": 1.1871,
+      "step": 632
+    },
+    {
+      "epoch": 0.20324289613100016,
+      "grad_norm": 1.9318181276321411,
+      "learning_rate": 0.0001264,
+      "loss": 0.8263,
+      "step": 633
+    },
+    {
+      "epoch": 0.20356397495585166,
+      "grad_norm": 1.6861572265625,
+      "learning_rate": 0.00012660000000000001,
+      "loss": 0.9605,
+      "step": 634
+    },
+    {
+      "epoch": 0.20388505378070315,
+      "grad_norm": 1.6631700992584229,
+      "learning_rate": 0.00012680000000000002,
+      "loss": 1.085,
+      "step": 635
+    },
+    {
+      "epoch": 0.20420613260555467,
+      "grad_norm": 2.279987096786499,
+      "learning_rate": 0.000127,
+      "loss": 1.0187,
+      "step": 636
+    },
+    {
+      "epoch": 0.20452721143040617,
+      "grad_norm": 1.2203277349472046,
+      "learning_rate": 0.0001272,
+      "loss": 0.8815,
+      "step": 637
+    },
+    {
+      "epoch": 0.20484829025525766,
+      "grad_norm": 1.7548545598983765,
+      "learning_rate": 0.0001274,
+      "loss": 1.1048,
+      "step": 638
+    },
+    {
+      "epoch": 0.20516936908010916,
+      "grad_norm": 2.2074527740478516,
+      "learning_rate": 0.0001276,
+      "loss": 0.9279,
+      "step": 639
+    },
+    {
+      "epoch": 0.20549044790496065,
+      "grad_norm": 2.118504762649536,
+      "learning_rate": 0.00012780000000000002,
+      "loss": 1.3603,
+      "step": 640
+    },
+    {
+      "epoch": 0.20581152672981218,
+      "grad_norm": 1.5209214687347412,
+      "learning_rate": 0.00012800000000000002,
+      "loss": 0.9429,
+      "step": 641
+    },
+    {
+      "epoch": 0.20613260555466367,
+      "grad_norm": 1.7410812377929688,
+      "learning_rate": 0.0001282,
+      "loss": 0.7626,
+      "step": 642
+    },
+    {
+      "epoch": 0.20645368437951517,
+      "grad_norm": 2.151503562927246,
+      "learning_rate": 0.0001284,
+      "loss": 1.1012,
+      "step": 643
+    },
+    {
+      "epoch": 0.20677476320436666,
+      "grad_norm": 1.4617127180099487,
+      "learning_rate": 0.0001286,
+      "loss": 1.072,
+      "step": 644
+    },
+    {
+      "epoch": 0.20709584202921819,
+      "grad_norm": 1.6642550230026245,
+      "learning_rate": 0.00012880000000000001,
+      "loss": 1.08,
+      "step": 645
+    },
+    {
+      "epoch": 0.20741692085406968,
+      "grad_norm": 1.8791637420654297,
+      "learning_rate": 0.00012900000000000002,
+      "loss": 0.8933,
+      "step": 646
+    },
+    {
+      "epoch": 0.20773799967892118,
+      "grad_norm": 1.970080018043518,
+      "learning_rate": 0.00012920000000000002,
+      "loss": 0.8507,
+      "step": 647
+    },
+    {
+      "epoch": 0.20805907850377267,
+      "grad_norm": 1.6559851169586182,
+      "learning_rate": 0.0001294,
+      "loss": 0.8517,
+      "step": 648
+    },
+    {
+      "epoch": 0.20838015732862417,
+      "grad_norm": 3.3205556869506836,
+      "learning_rate": 0.0001296,
+      "loss": 0.8304,
+      "step": 649
+    },
+    {
+      "epoch": 0.2087012361534757,
+      "grad_norm": 2.0807406902313232,
+      "learning_rate": 0.0001298,
+      "loss": 0.6431,
+      "step": 650
+    },
+    {
+      "epoch": 0.20902231497832718,
+      "grad_norm": 2.3025104999542236,
+      "learning_rate": 0.00013000000000000002,
+      "loss": 1.3972,
+      "step": 651
+    },
+    {
+      "epoch": 0.20934339380317868,
+      "grad_norm": 2.2260947227478027,
+      "learning_rate": 0.00013020000000000002,
+      "loss": 1.3472,
+      "step": 652
+    },
+    {
+      "epoch": 0.20966447262803017,
+      "grad_norm": 2.1171836853027344,
+      "learning_rate": 0.0001304,
+      "loss": 0.9882,
+      "step": 653
+    },
+    {
+      "epoch": 0.20998555145288167,
+      "grad_norm": 2.8188836574554443,
+      "learning_rate": 0.0001306,
+      "loss": 0.9835,
+      "step": 654
+    },
+    {
+      "epoch": 0.2103066302777332,
+      "grad_norm": 2.3081471920013428,
+      "learning_rate": 0.0001308,
+      "loss": 0.8868,
+      "step": 655
+    },
+    {
+      "epoch": 0.2106277091025847,
+      "grad_norm": 2.6961159706115723,
+      "learning_rate": 0.000131,
+      "loss": 0.8788,
+      "step": 656
+    },
+    {
+      "epoch": 0.21094878792743618,
+      "grad_norm": 2.381976366043091,
+      "learning_rate": 0.00013120000000000002,
+      "loss": 0.8831,
+      "step": 657
+    },
+    {
+      "epoch": 0.21126986675228768,
+      "grad_norm": 1.8818697929382324,
+      "learning_rate": 0.00013140000000000002,
+      "loss": 0.9654,
+      "step": 658
+    },
+    {
+      "epoch": 0.2115909455771392,
+      "grad_norm": 1.7192848920822144,
+      "learning_rate": 0.0001316,
+      "loss": 0.9574,
+      "step": 659
+    },
+    {
+      "epoch": 0.2119120244019907,
+      "grad_norm": 1.9481501579284668,
+      "learning_rate": 0.0001318,
+      "loss": 1.15,
+      "step": 660
+    },
+    {
+      "epoch": 0.2122331032268422,
+      "grad_norm": 1.8775956630706787,
+      "learning_rate": 0.000132,
+      "loss": 0.9595,
+      "step": 661
+    },
+    {
+      "epoch": 0.21255418205169369,
+      "grad_norm": 1.4788551330566406,
+      "learning_rate": 0.00013220000000000001,
+      "loss": 1.0547,
+      "step": 662
+    },
+    {
+      "epoch": 0.21287526087654518,
+      "grad_norm": 1.339841604232788,
+      "learning_rate": 0.00013240000000000002,
+      "loss": 1.0089,
+      "step": 663
+    },
+    {
+      "epoch": 0.2131963397013967,
+      "grad_norm": 2.0733635425567627,
+      "learning_rate": 0.00013260000000000002,
+      "loss": 0.8325,
+      "step": 664
+    },
+    {
+      "epoch": 0.2135174185262482,
+      "grad_norm": 1.9117823839187622,
+      "learning_rate": 0.0001328,
+      "loss": 1.048,
+      "step": 665
+    },
+    {
+      "epoch": 0.2138384973510997,
+      "grad_norm": 1.8078653812408447,
+      "learning_rate": 0.000133,
+      "loss": 0.9458,
+      "step": 666
+    },
+    {
+      "epoch": 0.2141595761759512,
+      "grad_norm": 1.7276756763458252,
+      "learning_rate": 0.0001332,
+      "loss": 1.1302,
+      "step": 667
+    },
+    {
+      "epoch": 0.21448065500080268,
+      "grad_norm": 2.459287166595459,
+      "learning_rate": 0.00013340000000000002,
+      "loss": 1.041,
+      "step": 668
+    },
+    {
+      "epoch": 0.2148017338256542,
+      "grad_norm": 1.7712688446044922,
+      "learning_rate": 0.00013360000000000002,
+      "loss": 1.0259,
+      "step": 669
+    },
+    {
+      "epoch": 0.2151228126505057,
+      "grad_norm": 2.60807728767395,
+      "learning_rate": 0.00013380000000000003,
+      "loss": 1.0492,
+      "step": 670
+    },
+    {
+      "epoch": 0.2154438914753572,
+      "grad_norm": 1.5164037942886353,
+      "learning_rate": 0.000134,
+      "loss": 1.06,
+      "step": 671
+    },
+    {
+      "epoch": 0.2157649703002087,
+      "grad_norm": 1.355437994003296,
+      "learning_rate": 0.0001342,
+      "loss": 1.0805,
+      "step": 672
+    },
+    {
+      "epoch": 0.21608604912506021,
+      "grad_norm": 1.6522624492645264,
+      "learning_rate": 0.00013440000000000001,
+      "loss": 1.1322,
+      "step": 673
+    },
+    {
+      "epoch": 0.2164071279499117,
+      "grad_norm": 2.0744080543518066,
+      "learning_rate": 0.00013460000000000002,
+      "loss": 1.1511,
+      "step": 674
+    },
+    {
+      "epoch": 0.2167282067747632,
+      "grad_norm": 3.5550551414489746,
+      "learning_rate": 0.00013480000000000002,
+      "loss": 1.1194,
+      "step": 675
+    },
+    {
+      "epoch": 0.2170492855996147,
+      "grad_norm": 1.7500160932540894,
+      "learning_rate": 0.00013500000000000003,
+      "loss": 1.0948,
+      "step": 676
+    },
+    {
+      "epoch": 0.2173703644244662,
+      "grad_norm": 2.251464605331421,
+      "learning_rate": 0.0001352,
+      "loss": 1.0829,
+      "step": 677
+    },
+    {
+      "epoch": 0.21769144324931772,
+      "grad_norm": 1.911451816558838,
+      "learning_rate": 0.0001354,
+      "loss": 1.1959,
+      "step": 678
+    },
+    {
+      "epoch": 0.2180125220741692,
+      "grad_norm": 2.199814796447754,
+      "learning_rate": 0.00013560000000000002,
+      "loss": 1.0434,
+      "step": 679
+    },
+    {
+      "epoch": 0.2183336008990207,
+      "grad_norm": 2.2858433723449707,
+      "learning_rate": 0.00013580000000000002,
+      "loss": 0.9276,
+      "step": 680
+    },
+    {
+      "epoch": 0.2186546797238722,
+      "grad_norm": 1.9057203531265259,
+      "learning_rate": 0.00013600000000000003,
+      "loss": 1.1686,
+      "step": 681
+    },
+    {
+      "epoch": 0.2189757585487237,
+      "grad_norm": 1.6700472831726074,
+      "learning_rate": 0.0001362,
+      "loss": 1.0646,
+      "step": 682
+    },
+    {
+      "epoch": 0.21929683737357522,
+      "grad_norm": 1.5589715242385864,
+      "learning_rate": 0.0001364,
+      "loss": 1.0016,
+      "step": 683
+    },
+    {
+      "epoch": 0.21961791619842672,
+      "grad_norm": 1.5741453170776367,
+      "learning_rate": 0.0001366,
+      "loss": 1.0659,
+      "step": 684
+    },
+    {
+      "epoch": 0.2199389950232782,
+      "grad_norm": 2.477905511856079,
+      "learning_rate": 0.00013680000000000002,
+      "loss": 1.0588,
+      "step": 685
+    },
+    {
+      "epoch": 0.2202600738481297,
+      "grad_norm": 1.8854243755340576,
+      "learning_rate": 0.00013700000000000002,
+      "loss": 0.7684,
+      "step": 686
+    },
+    {
+      "epoch": 0.22058115267298123,
+      "grad_norm": 1.5644632577896118,
+      "learning_rate": 0.00013720000000000003,
+      "loss": 1.1827,
+      "step": 687
+    },
+    {
+      "epoch": 0.22090223149783272,
+      "grad_norm": 2.956141948699951,
+      "learning_rate": 0.0001374,
+      "loss": 1.0493,
+      "step": 688
+    },
+    {
+      "epoch": 0.22122331032268422,
+      "grad_norm": 1.7298507690429688,
+      "learning_rate": 0.00013759999999999998,
+      "loss": 1.1918,
+      "step": 689
+    },
+    {
+      "epoch": 0.22154438914753571,
+      "grad_norm": 1.3661199808120728,
+      "learning_rate": 0.0001378,
+      "loss": 0.9168,
+      "step": 690
+    },
+    {
+      "epoch": 0.2218654679723872,
+      "grad_norm": 3.3666417598724365,
+      "learning_rate": 0.000138,
+      "loss": 0.9071,
+      "step": 691
+    },
+    {
+      "epoch": 0.22218654679723873,
+      "grad_norm": 1.8022873401641846,
+      "learning_rate": 0.0001382,
+      "loss": 1.0451,
+      "step": 692
+    },
+    {
+      "epoch": 0.22250762562209023,
+      "grad_norm": 2.2194650173187256,
+      "learning_rate": 0.0001384,
+      "loss": 1.2171,
+      "step": 693
+    },
+    {
+      "epoch": 0.22282870444694172,
+      "grad_norm": 2.7648520469665527,
+      "learning_rate": 0.0001386,
+      "loss": 0.7702,
+      "step": 694
+    },
+    {
+      "epoch": 0.22314978327179322,
+      "grad_norm": 2.6405279636383057,
+      "learning_rate": 0.00013879999999999999,
+      "loss": 0.8529,
+      "step": 695
+    },
+    {
+      "epoch": 0.22347086209664474,
+      "grad_norm": 2.8034744262695312,
+      "learning_rate": 0.000139,
+      "loss": 1.0367,
+      "step": 696
+    },
+    {
+      "epoch": 0.22379194092149624,
+      "grad_norm": 2.957364320755005,
+      "learning_rate": 0.0001392,
+      "loss": 0.9444,
+      "step": 697
+    },
+    {
+      "epoch": 0.22411301974634773,
+      "grad_norm": 1.5870848894119263,
+      "learning_rate": 0.0001394,
+      "loss": 0.9007,
+      "step": 698
+    },
+    {
+      "epoch": 0.22443409857119923,
+      "grad_norm": 1.8650282621383667,
+      "learning_rate": 0.0001396,
+      "loss": 0.7505,
+      "step": 699
+    },
+    {
+      "epoch": 0.22475517739605072,
+      "grad_norm": 3.5914523601531982,
+      "learning_rate": 0.0001398,
+      "loss": 0.8165,
+      "step": 700
+    },
+    {
+      "epoch": 0.22507625622090224,
+      "grad_norm": 6.338365077972412,
+      "learning_rate": 0.00014,
+      "loss": 1.48,
+      "step": 701
+    },
+    {
+      "epoch": 0.22539733504575374,
+      "grad_norm": 2.300706148147583,
+      "learning_rate": 0.0001402,
+      "loss": 1.3409,
+      "step": 702
+    },
+    {
+      "epoch": 0.22571841387060523,
+      "grad_norm": 2.9373714923858643,
+      "learning_rate": 0.0001404,
+      "loss": 1.3802,
+      "step": 703
+    },
+    {
+      "epoch": 0.22603949269545673,
+      "grad_norm": 2.5629308223724365,
+      "learning_rate": 0.0001406,
+      "loss": 0.8298,
+      "step": 704
+    },
+    {
+      "epoch": 0.22636057152030822,
+      "grad_norm": 2.515796184539795,
+      "learning_rate": 0.0001408,
+      "loss": 0.7154,
+      "step": 705
+    },
+    {
+      "epoch": 0.22668165034515975,
+      "grad_norm": 2.616245985031128,
+      "learning_rate": 0.000141,
+      "loss": 0.8415,
+      "step": 706
+    },
+    {
+      "epoch": 0.22700272917001124,
+      "grad_norm": 2.6605446338653564,
+      "learning_rate": 0.0001412,
+      "loss": 0.829,
+      "step": 707
+    },
+    {
+      "epoch": 0.22732380799486274,
+      "grad_norm": 2.9635884761810303,
+      "learning_rate": 0.0001414,
+      "loss": 0.9828,
+      "step": 708
+    },
+    {
+      "epoch": 0.22764488681971423,
+      "grad_norm": 3.6770541667938232,
+      "learning_rate": 0.0001416,
+      "loss": 1.0846,
+      "step": 709
+    },
+    {
+      "epoch": 0.22796596564456575,
+      "grad_norm": 2.220874071121216,
+      "learning_rate": 0.0001418,
+      "loss": 0.9111,
+      "step": 710
+    },
+    {
+      "epoch": 0.22828704446941725,
+      "grad_norm": 2.443786859512329,
+      "learning_rate": 0.000142,
+      "loss": 0.9152,
+      "step": 711
+    },
+    {
+      "epoch": 0.22860812329426874,
+      "grad_norm": 2.4385950565338135,
+      "learning_rate": 0.0001422,
+      "loss": 0.9161,
+      "step": 712
+    },
+    {
+      "epoch": 0.22892920211912024,
+      "grad_norm": 1.6144802570343018,
+      "learning_rate": 0.0001424,
+      "loss": 0.9943,
+      "step": 713
+    },
+    {
+      "epoch": 0.22925028094397173,
+      "grad_norm": 2.6351311206817627,
+      "learning_rate": 0.0001426,
+      "loss": 0.9274,
+      "step": 714
+    },
+    {
+      "epoch": 0.22957135976882326,
+      "grad_norm": 1.7359614372253418,
+      "learning_rate": 0.0001428,
+      "loss": 1.099,
+      "step": 715
+    },
+    {
+      "epoch": 0.22989243859367475,
+      "grad_norm": 1.4439224004745483,
+      "learning_rate": 0.000143,
+      "loss": 1.0443,
+      "step": 716
+    },
+    {
+      "epoch": 0.23021351741852625,
+      "grad_norm": 2.533806085586548,
+      "learning_rate": 0.0001432,
+      "loss": 1.0614,
+      "step": 717
+    },
+    {
+      "epoch": 0.23053459624337774,
+      "grad_norm": 2.2040460109710693,
+      "learning_rate": 0.0001434,
+      "loss": 0.8822,
+      "step": 718
+    },
+    {
+      "epoch": 0.23085567506822924,
+      "grad_norm": 1.4895590543746948,
+      "learning_rate": 0.0001436,
+      "loss": 1.244,
+      "step": 719
+    },
+    {
+      "epoch": 0.23117675389308076,
+      "grad_norm": 2.5897271633148193,
+      "learning_rate": 0.0001438,
+      "loss": 1.2112,
+      "step": 720
+    },
+    {
+      "epoch": 0.23149783271793226,
+      "grad_norm": 1.7311580181121826,
+      "learning_rate": 0.000144,
+      "loss": 0.9964,
+      "step": 721
+    },
+    {
+      "epoch": 0.23181891154278375,
+      "grad_norm": 1.475502610206604,
+      "learning_rate": 0.0001442,
+      "loss": 1.0241,
+      "step": 722
+    },
+    {
+      "epoch": 0.23213999036763525,
+      "grad_norm": 2.242170810699463,
+      "learning_rate": 0.0001444,
+      "loss": 1.2009,
+      "step": 723
+    },
+    {
+      "epoch": 0.23246106919248677,
+      "grad_norm": 1.811926007270813,
+      "learning_rate": 0.0001446,
+      "loss": 0.7842,
+      "step": 724
+    },
+    {
+      "epoch": 0.23278214801733826,
+      "grad_norm": 1.8713488578796387,
+      "learning_rate": 0.0001448,
+      "loss": 1.0161,
+      "step": 725
+    },
+    {
+      "epoch": 0.23310322684218976,
+      "grad_norm": 1.8007389307022095,
+      "learning_rate": 0.000145,
+      "loss": 1.1193,
+      "step": 726
+    },
+    {
+      "epoch": 0.23342430566704125,
+      "grad_norm": 1.432399868965149,
+      "learning_rate": 0.0001452,
+      "loss": 0.9496,
+      "step": 727
+    },
+    {
+      "epoch": 0.23374538449189275,
+      "grad_norm": 1.667504072189331,
+      "learning_rate": 0.0001454,
+      "loss": 1.0116,
+      "step": 728
+    },
+    {
+      "epoch": 0.23406646331674427,
+      "grad_norm": 1.476586103439331,
+      "learning_rate": 0.00014560000000000002,
+      "loss": 1.186,
+      "step": 729
+    },
+    {
+      "epoch": 0.23438754214159577,
+      "grad_norm": 1.715766429901123,
+      "learning_rate": 0.0001458,
+      "loss": 0.9619,
+      "step": 730
+    },
+    {
+      "epoch": 0.23470862096644726,
+      "grad_norm": 1.6196917295455933,
+      "learning_rate": 0.000146,
+      "loss": 0.7596,
+      "step": 731
+    },
+    {
+      "epoch": 0.23502969979129876,
+      "grad_norm": 2.2061188220977783,
+      "learning_rate": 0.0001462,
+      "loss": 0.8178,
+      "step": 732
+    },
+    {
+      "epoch": 0.23535077861615025,
+      "grad_norm": 1.9532783031463623,
+      "learning_rate": 0.0001464,
+      "loss": 1.0697,
+      "step": 733
+    },
+    {
+      "epoch": 0.23567185744100178,
+      "grad_norm": 2.0365941524505615,
+      "learning_rate": 0.0001466,
+      "loss": 0.9422,
+      "step": 734
+    },
+    {
+      "epoch": 0.23599293626585327,
+      "grad_norm": 2.11383318901062,
+      "learning_rate": 0.00014680000000000002,
+      "loss": 1.1384,
+      "step": 735
+    },
+    {
+      "epoch": 0.23631401509070477,
+      "grad_norm": 1.7044955492019653,
+      "learning_rate": 0.000147,
+      "loss": 1.0329,
+      "step": 736
+    },
+    {
+      "epoch": 0.23663509391555626,
+      "grad_norm": 1.1772584915161133,
+      "learning_rate": 0.0001472,
+      "loss": 0.7503,
+      "step": 737
+    },
+    {
+      "epoch": 0.23695617274040778,
+      "grad_norm": 1.8224774599075317,
+      "learning_rate": 0.0001474,
+      "loss": 1.0196,
+      "step": 738
+    },
+    {
+      "epoch": 0.23727725156525928,
+      "grad_norm": 1.4998708963394165,
+      "learning_rate": 0.0001476,
+      "loss": 0.8942,
+      "step": 739
+    },
+    {
+      "epoch": 0.23759833039011077,
+      "grad_norm": 1.2025278806686401,
+      "learning_rate": 0.00014780000000000001,
+      "loss": 0.8372,
+      "step": 740
+    },
+    {
+      "epoch": 0.23791940921496227,
+      "grad_norm": 1.9754714965820312,
+      "learning_rate": 0.000148,
+      "loss": 1.1058,
+      "step": 741
+    },
+    {
+      "epoch": 0.23824048803981376,
+      "grad_norm": 1.4123296737670898,
+      "learning_rate": 0.0001482,
+      "loss": 0.897,
+      "step": 742
+    },
+    {
+      "epoch": 0.2385615668646653,
+      "grad_norm": 1.8686493635177612,
+      "learning_rate": 0.0001484,
+      "loss": 1.1749,
+      "step": 743
+    },
+    {
+      "epoch": 0.23888264568951678,
+      "grad_norm": 2.425736904144287,
+      "learning_rate": 0.0001486,
+      "loss": 0.9771,
+      "step": 744
+    },
+    {
+      "epoch": 0.23920372451436828,
+      "grad_norm": 2.431823492050171,
+      "learning_rate": 0.0001488,
+      "loss": 0.8766,
+      "step": 745
+    },
+    {
+      "epoch": 0.23952480333921977,
+      "grad_norm": 3.2059309482574463,
+      "learning_rate": 0.00014900000000000002,
+      "loss": 0.86,
+      "step": 746
+    },
+    {
+      "epoch": 0.23984588216407127,
+      "grad_norm": 1.9675642251968384,
+      "learning_rate": 0.0001492,
+      "loss": 0.7826,
+      "step": 747
+    },
+    {
+      "epoch": 0.2401669609889228,
+      "grad_norm": 2.1029534339904785,
+      "learning_rate": 0.0001494,
+      "loss": 0.9178,
+      "step": 748
+    },
+    {
+      "epoch": 0.24048803981377428,
+      "grad_norm": 1.725165605545044,
+      "learning_rate": 0.0001496,
+      "loss": 0.7452,
+      "step": 749
+    },
+    {
+      "epoch": 0.24080911863862578,
+      "grad_norm": 2.5603771209716797,
+      "learning_rate": 0.0001498,
+      "loss": 0.7243,
+      "step": 750
+    },
+    {
+      "epoch": 0.24113019746347727,
+      "grad_norm": 3.2743546962738037,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 1.5645,
+      "step": 751
+    },
+    {
+      "epoch": 0.2414512762883288,
+      "grad_norm": 2.220749616622925,
+      "learning_rate": 0.00015020000000000002,
+      "loss": 1.2963,
+      "step": 752
+    },
+    {
+      "epoch": 0.2417723551131803,
+      "grad_norm": 2.2174606323242188,
+      "learning_rate": 0.0001504,
+      "loss": 0.9431,
+      "step": 753
+    },
+    {
+      "epoch": 0.2420934339380318,
+      "grad_norm": 2.5543525218963623,
+      "learning_rate": 0.0001506,
+      "loss": 0.7803,
+      "step": 754
+    },
+    {
+      "epoch": 0.24241451276288328,
+      "grad_norm": 2.6568639278411865,
+      "learning_rate": 0.0001508,
+      "loss": 0.7203,
+      "step": 755
+    },
+    {
+      "epoch": 0.24273559158773478,
+      "grad_norm": 2.6093509197235107,
+      "learning_rate": 0.000151,
+      "loss": 0.7474,
+      "step": 756
+    },
+    {
+      "epoch": 0.2430566704125863,
+      "grad_norm": 3.100468635559082,
+      "learning_rate": 0.00015120000000000002,
+      "loss": 1.1042,
+      "step": 757
+    },
+    {
+      "epoch": 0.2433777492374378,
+      "grad_norm": 2.6489973068237305,
+      "learning_rate": 0.00015140000000000002,
+      "loss": 1.1115,
+      "step": 758
+    },
+    {
+      "epoch": 0.2436988280622893,
+      "grad_norm": 1.9445077180862427,
+      "learning_rate": 0.0001516,
+      "loss": 1.2512,
+      "step": 759
+    },
+    {
+      "epoch": 0.2440199068871408,
+      "grad_norm": 2.7067511081695557,
+      "learning_rate": 0.0001518,
+      "loss": 1.0337,
+      "step": 760
+    },
+    {
+      "epoch": 0.24434098571199228,
+      "grad_norm": 1.793434977531433,
+      "learning_rate": 0.000152,
+      "loss": 1.2837,
+      "step": 761
+    },
+    {
+      "epoch": 0.2446620645368438,
+      "grad_norm": 2.2494122982025146,
+      "learning_rate": 0.0001522,
+      "loss": 0.9891,
+      "step": 762
+    },
+    {
+      "epoch": 0.2449831433616953,
+      "grad_norm": 2.235886812210083,
+      "learning_rate": 0.00015240000000000002,
+      "loss": 1.0547,
+      "step": 763
+    },
+    {
+      "epoch": 0.2453042221865468,
+      "grad_norm": 1.8249210119247437,
+      "learning_rate": 0.00015260000000000002,
+      "loss": 0.9715,
+      "step": 764
+    },
+    {
+      "epoch": 0.2456253010113983,
+      "grad_norm": 1.6558443307876587,
+      "learning_rate": 0.0001528,
+      "loss": 0.9652,
+      "step": 765
+    },
+    {
+      "epoch": 0.2459463798362498,
+      "grad_norm": 1.6599318981170654,
+      "learning_rate": 0.000153,
+      "loss": 1.2118,
+      "step": 766
+    },
+    {
+      "epoch": 0.2462674586611013,
+      "grad_norm": 2.2288310527801514,
+      "learning_rate": 0.0001532,
+      "loss": 1.0511,
+      "step": 767
+    },
+    {
+      "epoch": 0.2465885374859528,
+      "grad_norm": 2.151366949081421,
+      "learning_rate": 0.00015340000000000002,
+      "loss": 1.0277,
+      "step": 768
+    },
+    {
+      "epoch": 0.2469096163108043,
+      "grad_norm": 1.8321627378463745,
+      "learning_rate": 0.00015360000000000002,
+      "loss": 1.0368,
+      "step": 769
+    },
+    {
+      "epoch": 0.2472306951356558,
+      "grad_norm": 1.2890126705169678,
+      "learning_rate": 0.0001538,
+      "loss": 1.0081,
+      "step": 770
+    },
+    {
+      "epoch": 0.24755177396050732,
+      "grad_norm": 2.4398739337921143,
+      "learning_rate": 0.000154,
+      "loss": 0.8832,
+      "step": 771
+    },
+    {
+      "epoch": 0.2478728527853588,
+      "grad_norm": 2.993023157119751,
+      "learning_rate": 0.0001542,
+      "loss": 1.2266,
+      "step": 772
+    },
+    {
+      "epoch": 0.2481939316102103,
+      "grad_norm": 1.544333577156067,
+      "learning_rate": 0.0001544,
+      "loss": 1.0772,
+      "step": 773
+    },
+    {
+      "epoch": 0.2485150104350618,
+      "grad_norm": 1.9021294116973877,
+      "learning_rate": 0.00015460000000000002,
+      "loss": 1.1266,
+      "step": 774
+    },
+    {
+      "epoch": 0.2488360892599133,
+      "grad_norm": 2.129915952682495,
+      "learning_rate": 0.00015480000000000002,
+      "loss": 1.0554,
+      "step": 775
+    },
+    {
+      "epoch": 0.24915716808476482,
+      "grad_norm": 2.156653642654419,
+      "learning_rate": 0.000155,
+      "loss": 0.9094,
+      "step": 776
+    },
+    {
+      "epoch": 0.2494782469096163,
+      "grad_norm": 2.2110507488250732,
+      "learning_rate": 0.0001552,
+      "loss": 1.1234,
+      "step": 777
+    },
+    {
+      "epoch": 0.2497993257344678,
+      "grad_norm": 1.9623451232910156,
+      "learning_rate": 0.0001554,
+      "loss": 1.071,
+      "step": 778
+    },
+    {
+      "epoch": 0.25012040455931933,
+      "grad_norm": 2.1549320220947266,
+      "learning_rate": 0.00015560000000000001,
+      "loss": 1.1625,
+      "step": 779
+    },
+    {
+      "epoch": 0.2504414833841708,
+      "grad_norm": 1.2295719385147095,
+      "learning_rate": 0.00015580000000000002,
+      "loss": 0.8526,
+      "step": 780
+    },
+    {
+      "epoch": 0.2507625622090223,
+      "grad_norm": 1.3640968799591064,
+      "learning_rate": 0.00015600000000000002,
+      "loss": 0.9144,
+      "step": 781
+    },
+    {
+      "epoch": 0.2510836410338738,
+      "grad_norm": 2.0186564922332764,
+      "learning_rate": 0.0001562,
+      "loss": 1.0645,
+      "step": 782
+    },
+    {
+      "epoch": 0.2514047198587253,
+      "grad_norm": 2.253253698348999,
+      "learning_rate": 0.0001564,
+      "loss": 1.087,
+      "step": 783
+    },
+    {
+      "epoch": 0.2517257986835768,
+      "grad_norm": 2.00610089302063,
+      "learning_rate": 0.0001566,
+      "loss": 1.0834,
+      "step": 784
+    },
+    {
+      "epoch": 0.2520468775084283,
+      "grad_norm": 1.227858543395996,
+      "learning_rate": 0.00015680000000000002,
+      "loss": 0.8328,
+      "step": 785
+    },
+    {
+      "epoch": 0.2523679563332798,
+      "grad_norm": 1.6620421409606934,
+      "learning_rate": 0.00015700000000000002,
+      "loss": 0.9007,
+      "step": 786
+    },
+    {
+      "epoch": 0.25268903515813135,
+      "grad_norm": 1.969337821006775,
+      "learning_rate": 0.00015720000000000003,
+      "loss": 0.8842,
+      "step": 787
+    },
+    {
+      "epoch": 0.25301011398298284,
+      "grad_norm": 1.8422551155090332,
+      "learning_rate": 0.0001574,
+      "loss": 1.0329,
+      "step": 788
+    },
+    {
+      "epoch": 0.25333119280783434,
+      "grad_norm": 2.0123472213745117,
+      "learning_rate": 0.0001576,
+      "loss": 1.0015,
+      "step": 789
+    },
+    {
+      "epoch": 0.25365227163268583,
+      "grad_norm": 1.7080883979797363,
+      "learning_rate": 0.00015780000000000001,
+      "loss": 1.0998,
+      "step": 790
+    },
+    {
+      "epoch": 0.25397335045753733,
+      "grad_norm": 3.233981132507324,
+      "learning_rate": 0.00015800000000000002,
+      "loss": 1.132,
+      "step": 791
+    },
+    {
+      "epoch": 0.2542944292823888,
+      "grad_norm": 1.4392627477645874,
+      "learning_rate": 0.00015820000000000002,
+      "loss": 0.8428,
+      "step": 792
+    },
+    {
+      "epoch": 0.2546155081072403,
+      "grad_norm": 1.7838289737701416,
+      "learning_rate": 0.00015840000000000003,
+      "loss": 0.986,
+      "step": 793
+    },
+    {
+      "epoch": 0.2549365869320918,
+      "grad_norm": 2.152930498123169,
+      "learning_rate": 0.0001586,
+      "loss": 0.7264,
+      "step": 794
+    },
+    {
+      "epoch": 0.2552576657569433,
+      "grad_norm": 1.7008942365646362,
+      "learning_rate": 0.0001588,
+      "loss": 1.2359,
+      "step": 795
+    },
+    {
+      "epoch": 0.2555787445817948,
+      "grad_norm": 1.2774722576141357,
+      "learning_rate": 0.00015900000000000002,
+      "loss": 0.8856,
+      "step": 796
+    },
+    {
+      "epoch": 0.25589982340664635,
+      "grad_norm": 1.5711169242858887,
+      "learning_rate": 0.00015920000000000002,
+      "loss": 0.8394,
+      "step": 797
+    },
+    {
+      "epoch": 0.25622090223149785,
+      "grad_norm": 2.9855380058288574,
+      "learning_rate": 0.00015940000000000003,
+      "loss": 0.8879,
+      "step": 798
+    },
+    {
+      "epoch": 0.25654198105634934,
+      "grad_norm": 2.1591596603393555,
+      "learning_rate": 0.0001596,
+      "loss": 0.8859,
+      "step": 799
+    },
+    {
+      "epoch": 0.25686305988120084,
+      "grad_norm": 1.9048887491226196,
+      "learning_rate": 0.0001598,
+      "loss": 0.5577,
+      "step": 800
+    },
+    {
+      "epoch": 0.25718413870605233,
+      "grad_norm": 18.338882446289062,
+      "learning_rate": 0.00016,
+      "loss": 1.7698,
+      "step": 801
+    },
+    {
+      "epoch": 0.25750521753090383,
+      "grad_norm": 2.443437099456787,
+      "learning_rate": 0.00016020000000000002,
+      "loss": 1.5749,
+      "step": 802
+    },
+    {
+      "epoch": 0.2578262963557553,
+      "grad_norm": 2.9275338649749756,
+      "learning_rate": 0.00016040000000000002,
+      "loss": 1.1417,
+      "step": 803
+    },
+    {
+      "epoch": 0.2581473751806068,
+      "grad_norm": 2.1856069564819336,
+      "learning_rate": 0.00016060000000000003,
+      "loss": 0.919,
+      "step": 804
+    },
+    {
+      "epoch": 0.2584684540054583,
+      "grad_norm": 2.806535005569458,
+      "learning_rate": 0.0001608,
+      "loss": 0.8787,
+      "step": 805
+    },
+    {
+      "epoch": 0.25878953283030987,
+      "grad_norm": 2.342740774154663,
+      "learning_rate": 0.000161,
+      "loss": 0.8913,
+      "step": 806
+    },
+    {
+      "epoch": 0.25911061165516136,
+      "grad_norm": 2.3539390563964844,
+      "learning_rate": 0.00016120000000000002,
+      "loss": 0.9415,
+      "step": 807
+    },
+    {
+      "epoch": 0.25943169048001286,
+      "grad_norm": 2.2060728073120117,
+      "learning_rate": 0.00016140000000000002,
+      "loss": 1.1096,
+      "step": 808
+    },
+    {
+      "epoch": 0.25975276930486435,
+      "grad_norm": 1.5557055473327637,
+      "learning_rate": 0.00016160000000000002,
+      "loss": 0.9594,
+      "step": 809
+    },
+    {
+      "epoch": 0.26007384812971585,
+      "grad_norm": 1.8541979789733887,
+      "learning_rate": 0.00016180000000000003,
+      "loss": 0.9541,
+      "step": 810
+    },
+    {
+      "epoch": 0.26039492695456734,
+      "grad_norm": 2.1980228424072266,
+      "learning_rate": 0.000162,
+      "loss": 1.0466,
+      "step": 811
+    },
+    {
+      "epoch": 0.26071600577941884,
+      "grad_norm": 3.6660101413726807,
+      "learning_rate": 0.0001622,
+      "loss": 1.045,
+      "step": 812
+    },
+    {
+      "epoch": 0.26103708460427033,
+      "grad_norm": 1.4750818014144897,
+      "learning_rate": 0.00016240000000000002,
+      "loss": 0.9617,
+      "step": 813
+    },
+    {
+      "epoch": 0.2613581634291218,
+      "grad_norm": 1.7362655401229858,
+      "learning_rate": 0.0001626,
+      "loss": 1.0644,
+      "step": 814
+    },
+    {
+      "epoch": 0.2616792422539734,
+      "grad_norm": 3.413910150527954,
+      "learning_rate": 0.0001628,
+      "loss": 0.9958,
+      "step": 815
+    },
+    {
+      "epoch": 0.26200032107882487,
+      "grad_norm": 1.9933480024337769,
+      "learning_rate": 0.000163,
+      "loss": 1.0984,
+      "step": 816
+    },
+    {
+      "epoch": 0.26232139990367637,
+      "grad_norm": 2.60626220703125,
+      "learning_rate": 0.0001632,
+      "loss": 0.9565,
+      "step": 817
+    },
+    {
+      "epoch": 0.26264247872852786,
+      "grad_norm": 2.5874414443969727,
+      "learning_rate": 0.0001634,
+      "loss": 1.1555,
+      "step": 818
+    },
+    {
+      "epoch": 0.26296355755337936,
+      "grad_norm": 1.991832971572876,
+      "learning_rate": 0.0001636,
+      "loss": 0.879,
+      "step": 819
+    },
+    {
+      "epoch": 0.26328463637823085,
+      "grad_norm": 2.291994333267212,
+      "learning_rate": 0.0001638,
+      "loss": 1.1457,
+      "step": 820
+    },
+    {
+      "epoch": 0.26360571520308235,
+      "grad_norm": 2.4672718048095703,
+      "learning_rate": 0.000164,
+      "loss": 1.0751,
+      "step": 821
+    },
+    {
+      "epoch": 0.26392679402793384,
+      "grad_norm": 1.6756705045700073,
+      "learning_rate": 0.0001642,
+      "loss": 0.8653,
+      "step": 822
+    },
+    {
+      "epoch": 0.26424787285278534,
+      "grad_norm": 1.8445497751235962,
+      "learning_rate": 0.0001644,
+      "loss": 0.9074,
+      "step": 823
+    },
+    {
+      "epoch": 0.26456895167763683,
+      "grad_norm": 1.6544156074523926,
+      "learning_rate": 0.0001646,
+      "loss": 0.9691,
+      "step": 824
+    },
+    {
+      "epoch": 0.2648900305024884,
+      "grad_norm": 1.8549728393554688,
+      "learning_rate": 0.0001648,
+      "loss": 1.1391,
+      "step": 825
+    },
+    {
+      "epoch": 0.2652111093273399,
+      "grad_norm": 2.89859938621521,
+      "learning_rate": 0.000165,
+      "loss": 1.0288,
+      "step": 826
+    },
+    {
+      "epoch": 0.2655321881521914,
+      "grad_norm": 3.0639214515686035,
+      "learning_rate": 0.0001652,
+      "loss": 1.1499,
+      "step": 827
+    },
+    {
+      "epoch": 0.26585326697704287,
+      "grad_norm": 1.4466873407363892,
+      "learning_rate": 0.0001654,
+      "loss": 0.9982,
+      "step": 828
+    },
+    {
+      "epoch": 0.26617434580189436,
+      "grad_norm": 2.002682685852051,
+      "learning_rate": 0.0001656,
+      "loss": 0.9603,
+      "step": 829
+    },
+    {
+      "epoch": 0.26649542462674586,
+      "grad_norm": 2.583599805831909,
+      "learning_rate": 0.0001658,
+      "loss": 1.1498,
+      "step": 830
+    },
+    {
+      "epoch": 0.26681650345159735,
+      "grad_norm": 1.9900764226913452,
+      "learning_rate": 0.000166,
+      "loss": 1.1214,
+      "step": 831
+    },
+    {
+      "epoch": 0.26713758227644885,
+      "grad_norm": 1.384462594985962,
+      "learning_rate": 0.0001662,
+      "loss": 0.7923,
+      "step": 832
+    },
+    {
+      "epoch": 0.26745866110130034,
+      "grad_norm": 7.388343334197998,
+      "learning_rate": 0.0001664,
+      "loss": 1.2836,
+      "step": 833
+    },
+    {
+      "epoch": 0.2677797399261519,
+      "grad_norm": 1.9020540714263916,
+      "learning_rate": 0.0001666,
+      "loss": 1.0163,
+      "step": 834
+    },
+    {
+      "epoch": 0.2681008187510034,
+      "grad_norm": 1.6172999143600464,
+      "learning_rate": 0.0001668,
+      "loss": 1.1151,
+      "step": 835
+    },
+    {
+      "epoch": 0.2684218975758549,
+      "grad_norm": 1.3895915746688843,
+      "learning_rate": 0.000167,
+      "loss": 1.2039,
+      "step": 836
+    },
+    {
+      "epoch": 0.2687429764007064,
+      "grad_norm": 4.345794200897217,
+      "learning_rate": 0.0001672,
+      "loss": 1.4846,
+      "step": 837
+    },
+    {
+      "epoch": 0.2690640552255579,
+      "grad_norm": 1.9014428853988647,
+      "learning_rate": 0.0001674,
+      "loss": 1.164,
+      "step": 838
+    },
+    {
+      "epoch": 0.26938513405040937,
+      "grad_norm": 1.5048143863677979,
+      "learning_rate": 0.0001676,
+      "loss": 0.9134,
+      "step": 839
+    },
+    {
+      "epoch": 0.26970621287526086,
+      "grad_norm": 2.302903890609741,
+      "learning_rate": 0.0001678,
+      "loss": 0.9193,
+      "step": 840
+    },
+    {
+      "epoch": 0.27002729170011236,
+      "grad_norm": 2.238719940185547,
+      "learning_rate": 0.000168,
+      "loss": 0.9618,
+      "step": 841
+    },
+    {
+      "epoch": 0.27034837052496385,
+      "grad_norm": 1.821960210800171,
+      "learning_rate": 0.0001682,
+      "loss": 0.9859,
+      "step": 842
+    },
+    {
+      "epoch": 0.2706694493498154,
+      "grad_norm": 1.5913723707199097,
+      "learning_rate": 0.0001684,
+      "loss": 0.7387,
+      "step": 843
+    },
+    {
+      "epoch": 0.2709905281746669,
+      "grad_norm": 2.4586009979248047,
+      "learning_rate": 0.0001686,
+      "loss": 1.1114,
+      "step": 844
+    },
+    {
+      "epoch": 0.2713116069995184,
+      "grad_norm": 1.5896835327148438,
+      "learning_rate": 0.0001688,
+      "loss": 0.8628,
+      "step": 845
+    },
+    {
+      "epoch": 0.2716326858243699,
+      "grad_norm": 1.7305841445922852,
+      "learning_rate": 0.00016900000000000002,
+      "loss": 0.952,
+      "step": 846
+    },
+    {
+      "epoch": 0.2719537646492214,
+      "grad_norm": 2.2322258949279785,
+      "learning_rate": 0.0001692,
+      "loss": 0.8601,
+      "step": 847
+    },
+    {
+      "epoch": 0.2722748434740729,
+      "grad_norm": 3.5011134147644043,
+      "learning_rate": 0.0001694,
+      "loss": 0.8119,
+      "step": 848
+    },
+    {
+      "epoch": 0.2725959222989244,
+      "grad_norm": 1.6285743713378906,
+      "learning_rate": 0.0001696,
+      "loss": 0.8586,
+      "step": 849
+    },
+    {
+      "epoch": 0.27291700112377587,
+      "grad_norm": 2.2089757919311523,
+      "learning_rate": 0.0001698,
+      "loss": 0.7086,
+      "step": 850
+    },
+    {
+      "epoch": 0.27323807994862737,
+      "grad_norm": 3.487232208251953,
+      "learning_rate": 0.00017,
+      "loss": 1.52,
+      "step": 851
+    },
+    {
+      "epoch": 0.2735591587734789,
+      "grad_norm": 2.836216688156128,
+      "learning_rate": 0.00017020000000000002,
+      "loss": 1.4958,
+      "step": 852
+    },
+    {
+      "epoch": 0.2738802375983304,
+      "grad_norm": 2.933957099914551,
+      "learning_rate": 0.0001704,
+      "loss": 1.1824,
+      "step": 853
+    },
+    {
+      "epoch": 0.2742013164231819,
+      "grad_norm": 2.5374979972839355,
+      "learning_rate": 0.0001706,
+      "loss": 1.0371,
+      "step": 854
+    },
+    {
+      "epoch": 0.2745223952480334,
+      "grad_norm": 2.2211618423461914,
+      "learning_rate": 0.0001708,
+      "loss": 0.8578,
+      "step": 855
+    },
+    {
+      "epoch": 0.2748434740728849,
+      "grad_norm": 2.6845898628234863,
+      "learning_rate": 0.000171,
+      "loss": 0.8998,
+      "step": 856
+    },
+    {
+      "epoch": 0.2751645528977364,
+      "grad_norm": 2.2504897117614746,
+      "learning_rate": 0.00017120000000000001,
+      "loss": 0.8661,
+      "step": 857
+    },
+    {
+      "epoch": 0.2754856317225879,
+      "grad_norm": 2.5308399200439453,
+      "learning_rate": 0.0001714,
+      "loss": 0.8642,
+      "step": 858
+    },
+    {
+      "epoch": 0.2758067105474394,
+      "grad_norm": 2.6598222255706787,
+      "learning_rate": 0.0001716,
+      "loss": 0.8098,
+      "step": 859
+    },
+    {
+      "epoch": 0.2761277893722909,
+      "grad_norm": 1.8385624885559082,
+      "learning_rate": 0.0001718,
+      "loss": 1.0242,
+      "step": 860
+    },
+    {
+      "epoch": 0.2764488681971424,
+      "grad_norm": 2.0007526874542236,
+      "learning_rate": 0.000172,
+      "loss": 1.166,
+      "step": 861
+    },
+    {
+      "epoch": 0.2767699470219939,
+      "grad_norm": 3.164480209350586,
+      "learning_rate": 0.0001722,
+      "loss": 0.9461,
+      "step": 862
+    },
+    {
+      "epoch": 0.2770910258468454,
+      "grad_norm": 1.8709050416946411,
+      "learning_rate": 0.00017240000000000002,
+      "loss": 1.0128,
+      "step": 863
+    },
+    {
+      "epoch": 0.2774121046716969,
+      "grad_norm": 1.6714998483657837,
+      "learning_rate": 0.0001726,
+      "loss": 0.8408,
+      "step": 864
+    },
+    {
+      "epoch": 0.2777331834965484,
+      "grad_norm": 1.7060233354568481,
+      "learning_rate": 0.0001728,
+      "loss": 0.796,
+      "step": 865
+    },
+    {
+      "epoch": 0.2780542623213999,
+      "grad_norm": 1.3818310499191284,
+      "learning_rate": 0.000173,
+      "loss": 1.0166,
+      "step": 866
+    },
+    {
+      "epoch": 0.2783753411462514,
+      "grad_norm": 1.8317921161651611,
+      "learning_rate": 0.0001732,
+      "loss": 1.0847,
+      "step": 867
+    },
+    {
+      "epoch": 0.2786964199711029,
+      "grad_norm": 2.0381312370300293,
+      "learning_rate": 0.0001734,
+      "loss": 0.9889,
+      "step": 868
+    },
+    {
+      "epoch": 0.2790174987959544,
+      "grad_norm": 2.772416353225708,
+      "learning_rate": 0.00017360000000000002,
+      "loss": 1.1122,
+      "step": 869
+    },
+    {
+      "epoch": 0.2793385776208059,
+      "grad_norm": 4.282870292663574,
+      "learning_rate": 0.0001738,
+      "loss": 1.2099,
+      "step": 870
+    },
+    {
+      "epoch": 0.27965965644565743,
+      "grad_norm": 1.2967722415924072,
+      "learning_rate": 0.000174,
+      "loss": 0.9476,
+      "step": 871
+    },
+    {
+      "epoch": 0.27998073527050893,
+      "grad_norm": 1.8298275470733643,
+      "learning_rate": 0.0001742,
+      "loss": 1.0377,
+      "step": 872
+    },
+    {
+      "epoch": 0.2803018140953604,
+      "grad_norm": 2.2432875633239746,
+      "learning_rate": 0.0001744,
+      "loss": 1.0409,
+      "step": 873
+    },
+    {
+      "epoch": 0.2806228929202119,
+      "grad_norm": 1.3698315620422363,
+      "learning_rate": 0.00017460000000000002,
+      "loss": 1.0456,
+      "step": 874
+    },
+    {
+      "epoch": 0.2809439717450634,
+      "grad_norm": 1.9452381134033203,
+      "learning_rate": 0.00017480000000000002,
+      "loss": 0.8272,
+      "step": 875
+    },
+    {
+      "epoch": 0.2812650505699149,
+      "grad_norm": 1.601507306098938,
+      "learning_rate": 0.000175,
+      "loss": 1.0913,
+      "step": 876
+    },
+    {
+      "epoch": 0.2815861293947664,
+      "grad_norm": 3.9005463123321533,
+      "learning_rate": 0.0001752,
+      "loss": 1.4486,
+      "step": 877
+    },
+    {
+      "epoch": 0.2819072082196179,
+      "grad_norm": 2.154240608215332,
+      "learning_rate": 0.0001754,
+      "loss": 0.9732,
+      "step": 878
+    },
+    {
+      "epoch": 0.2822282870444694,
+      "grad_norm": 1.2495157718658447,
+      "learning_rate": 0.0001756,
+      "loss": 1.0217,
+      "step": 879
+    },
+    {
+      "epoch": 0.28254936586932095,
+      "grad_norm": 6.445877552032471,
+      "learning_rate": 0.00017580000000000002,
+      "loss": 1.1539,
+      "step": 880
+    },
+    {
+      "epoch": 0.28287044469417244,
+      "grad_norm": 2.6896142959594727,
+      "learning_rate": 0.00017600000000000002,
+      "loss": 1.0758,
+      "step": 881
+    },
+    {
+      "epoch": 0.28319152351902394,
+      "grad_norm": 1.3675345182418823,
+      "learning_rate": 0.0001762,
+      "loss": 1.0229,
+      "step": 882
+    },
+    {
+      "epoch": 0.28351260234387543,
+      "grad_norm": 9.003292083740234,
+      "learning_rate": 0.0001764,
+      "loss": 1.1875,
+      "step": 883
+    },
+    {
+      "epoch": 0.2838336811687269,
+      "grad_norm": 1.7090665102005005,
+      "learning_rate": 0.0001766,
+      "loss": 1.2299,
+      "step": 884
+    },
+    {
+      "epoch": 0.2841547599935784,
+      "grad_norm": 2.320629119873047,
+      "learning_rate": 0.00017680000000000001,
+      "loss": 1.3493,
+      "step": 885
+    },
+    {
+      "epoch": 0.2844758388184299,
+      "grad_norm": 3.2875213623046875,
+      "learning_rate": 0.00017700000000000002,
+      "loss": 1.0224,
+      "step": 886
+    },
+    {
+      "epoch": 0.2847969176432814,
+      "grad_norm": 3.1284916400909424,
+      "learning_rate": 0.0001772,
+      "loss": 1.0231,
+      "step": 887
+    },
+    {
+      "epoch": 0.2851179964681329,
+      "grad_norm": 1.9187723398208618,
+      "learning_rate": 0.0001774,
+      "loss": 1.0035,
+      "step": 888
+    },
+    {
+      "epoch": 0.2854390752929844,
+      "grad_norm": 1.3869761228561401,
+      "learning_rate": 0.0001776,
+      "loss": 0.8738,
+      "step": 889
+    },
+    {
+      "epoch": 0.28576015411783595,
+      "grad_norm": 1.281718134880066,
+      "learning_rate": 0.0001778,
+      "loss": 0.8464,
+      "step": 890
+    },
+    {
+      "epoch": 0.28608123294268745,
+      "grad_norm": 1.203602910041809,
+      "learning_rate": 0.00017800000000000002,
+      "loss": 0.9197,
+      "step": 891
+    },
+    {
+      "epoch": 0.28640231176753894,
+      "grad_norm": 1.9265743494033813,
+      "learning_rate": 0.00017820000000000002,
+      "loss": 1.0422,
+      "step": 892
+    },
+    {
+      "epoch": 0.28672339059239044,
+      "grad_norm": 1.4553351402282715,
+      "learning_rate": 0.0001784,
+      "loss": 1.1858,
+      "step": 893
+    },
+    {
+      "epoch": 0.28704446941724193,
+      "grad_norm": 1.6324553489685059,
+      "learning_rate": 0.0001786,
+      "loss": 0.9635,
+      "step": 894
+    },
+    {
+      "epoch": 0.2873655482420934,
+      "grad_norm": 4.445520401000977,
+      "learning_rate": 0.0001788,
+      "loss": 1.1842,
+      "step": 895
+    },
+    {
+      "epoch": 0.2876866270669449,
+      "grad_norm": 2.625737428665161,
+      "learning_rate": 0.00017900000000000001,
+      "loss": 1.0714,
+      "step": 896
+    },
+    {
+      "epoch": 0.2880077058917964,
+      "grad_norm": 1.3031972646713257,
+      "learning_rate": 0.00017920000000000002,
+      "loss": 0.9172,
+      "step": 897
+    },
+    {
+      "epoch": 0.2883287847166479,
+      "grad_norm": 1.1816191673278809,
+      "learning_rate": 0.00017940000000000002,
+      "loss": 0.8658,
+      "step": 898
+    },
+    {
+      "epoch": 0.28864986354149946,
+      "grad_norm": 2.7634835243225098,
+      "learning_rate": 0.0001796,
+      "loss": 0.8205,
+      "step": 899
+    },
+    {
+      "epoch": 0.28897094236635096,
+      "grad_norm": 1.715277910232544,
+      "learning_rate": 0.0001798,
+      "loss": 0.6063,
+      "step": 900
+    },
+    {
+      "epoch": 0.28929202119120245,
+      "grad_norm": 5.679165840148926,
+      "learning_rate": 0.00018,
+      "loss": 1.6148,
+      "step": 901
+    },
+    {
+      "epoch": 0.28961310001605395,
+      "grad_norm": 8.520236015319824,
+      "learning_rate": 0.00018020000000000002,
+      "loss": 1.669,
+      "step": 902
+    },
+    {
+      "epoch": 0.28993417884090544,
+      "grad_norm": 5.708471298217773,
+      "learning_rate": 0.00018040000000000002,
+      "loss": 1.0973,
+      "step": 903
+    },
+    {
+      "epoch": 0.29025525766575694,
+      "grad_norm": 4.294043064117432,
+      "learning_rate": 0.00018060000000000003,
+      "loss": 0.9894,
+      "step": 904
+    },
+    {
+      "epoch": 0.29057633649060843,
+      "grad_norm": 4.61492919921875,
+      "learning_rate": 0.0001808,
+      "loss": 0.9369,
+      "step": 905
+    },
+    {
+      "epoch": 0.29089741531545993,
+      "grad_norm": 2.621717691421509,
+      "learning_rate": 0.000181,
+      "loss": 1.0031,
+      "step": 906
+    },
+    {
+      "epoch": 0.2912184941403114,
+      "grad_norm": 2.668614149093628,
+      "learning_rate": 0.0001812,
+      "loss": 0.9567,
+      "step": 907
+    },
+    {
+      "epoch": 0.291539572965163,
+      "grad_norm": 2.6157805919647217,
+      "learning_rate": 0.00018140000000000002,
+      "loss": 1.0925,
+      "step": 908
+    },
+    {
+      "epoch": 0.29186065179001447,
+      "grad_norm": 2.2772955894470215,
+      "learning_rate": 0.00018160000000000002,
+      "loss": 1.0826,
+      "step": 909
+    },
+    {
+      "epoch": 0.29218173061486596,
+      "grad_norm": 2.532787561416626,
+      "learning_rate": 0.00018180000000000003,
+      "loss": 1.0492,
+      "step": 910
+    },
+    {
+      "epoch": 0.29250280943971746,
+      "grad_norm": 2.4008162021636963,
+      "learning_rate": 0.000182,
+      "loss": 0.9815,
+      "step": 911
+    },
+    {
+      "epoch": 0.29282388826456895,
+      "grad_norm": 1.3333711624145508,
+      "learning_rate": 0.0001822,
+      "loss": 1.071,
+      "step": 912
+    },
+    {
+      "epoch": 0.29314496708942045,
+      "grad_norm": 2.2058145999908447,
+      "learning_rate": 0.00018240000000000002,
+      "loss": 1.1503,
+      "step": 913
+    },
+    {
+      "epoch": 0.29346604591427194,
+      "grad_norm": 1.4182907342910767,
+      "learning_rate": 0.00018260000000000002,
+      "loss": 1.1016,
+      "step": 914
+    },
+    {
+      "epoch": 0.29378712473912344,
+      "grad_norm": 2.7944185733795166,
+      "learning_rate": 0.00018280000000000003,
+      "loss": 1.233,
+      "step": 915
+    },
+    {
+      "epoch": 0.29410820356397493,
+      "grad_norm": 1.5195378065109253,
+      "learning_rate": 0.000183,
+      "loss": 1.1977,
+      "step": 916
+    },
+    {
+      "epoch": 0.29442928238882643,
+      "grad_norm": 1.8296455144882202,
+      "learning_rate": 0.0001832,
+      "loss": 0.9808,
+      "step": 917
+    },
+    {
+      "epoch": 0.294750361213678,
+      "grad_norm": 1.8149925470352173,
+      "learning_rate": 0.0001834,
+      "loss": 1.0705,
+      "step": 918
+    },
+    {
+      "epoch": 0.2950714400385295,
+      "grad_norm": 1.741716980934143,
+      "learning_rate": 0.00018360000000000002,
+      "loss": 1.1261,
+      "step": 919
+    },
+    {
+      "epoch": 0.29539251886338097,
+      "grad_norm": 2.0786545276641846,
+      "learning_rate": 0.00018380000000000002,
+      "loss": 1.1481,
+      "step": 920
+    },
+    {
+      "epoch": 0.29571359768823247,
+      "grad_norm": 1.7828667163848877,
+      "learning_rate": 0.00018400000000000003,
+      "loss": 1.034,
+      "step": 921
+    },
+    {
+      "epoch": 0.29603467651308396,
+      "grad_norm": 1.6644797325134277,
+      "learning_rate": 0.0001842,
+      "loss": 1.1322,
+      "step": 922
+    },
+    {
+      "epoch": 0.29635575533793546,
+      "grad_norm": 1.3413432836532593,
+      "learning_rate": 0.0001844,
+      "loss": 1.1023,
+      "step": 923
+    },
+    {
+      "epoch": 0.29667683416278695,
+      "grad_norm": 1.985334873199463,
+      "learning_rate": 0.00018460000000000001,
+      "loss": 0.9872,
+      "step": 924
+    },
+    {
+      "epoch": 0.29699791298763845,
+      "grad_norm": 2.2249557971954346,
+      "learning_rate": 0.00018480000000000002,
+      "loss": 1.0467,
+      "step": 925
+    },
+    {
+      "epoch": 0.29731899181248994,
+      "grad_norm": 1.843141794204712,
+      "learning_rate": 0.00018500000000000002,
+      "loss": 1.0913,
+      "step": 926
+    },
+    {
+      "epoch": 0.2976400706373415,
+      "grad_norm": 1.8132327795028687,
+      "learning_rate": 0.00018520000000000003,
+      "loss": 0.9764,
+      "step": 927
+    },
+    {
+      "epoch": 0.297961149462193,
+      "grad_norm": 2.068922996520996,
+      "learning_rate": 0.0001854,
+      "loss": 0.9522,
+      "step": 928
+    },
+    {
+      "epoch": 0.2982822282870445,
+      "grad_norm": 1.975690245628357,
+      "learning_rate": 0.0001856,
+      "loss": 0.961,
+      "step": 929
+    },
+    {
+      "epoch": 0.298603307111896,
+      "grad_norm": 2.761979103088379,
+      "learning_rate": 0.00018580000000000002,
+      "loss": 1.1271,
+      "step": 930
+    },
+    {
+      "epoch": 0.29892438593674747,
+      "grad_norm": 2.052741527557373,
+      "learning_rate": 0.00018600000000000002,
+      "loss": 1.0824,
+      "step": 931
+    },
+    {
+      "epoch": 0.29924546476159897,
+      "grad_norm": 2.1390645503997803,
+      "learning_rate": 0.00018620000000000003,
+      "loss": 1.2875,
+      "step": 932
+    },
+    {
+      "epoch": 0.29956654358645046,
+      "grad_norm": 2.17155385017395,
+      "learning_rate": 0.00018640000000000003,
+      "loss": 1.0559,
+      "step": 933
+    },
+    {
+      "epoch": 0.29988762241130196,
+      "grad_norm": 1.4446396827697754,
+      "learning_rate": 0.0001866,
+      "loss": 1.052,
+      "step": 934
+    },
+    {
+      "epoch": 0.30020870123615345,
+      "grad_norm": 2.108684778213501,
+      "learning_rate": 0.00018680000000000001,
+      "loss": 0.9973,
+      "step": 935
+    },
+    {
+      "epoch": 0.300529780061005,
+      "grad_norm": 1.7543820142745972,
+      "learning_rate": 0.00018700000000000002,
+      "loss": 1.0001,
+      "step": 936
+    },
+    {
+      "epoch": 0.3008508588858565,
+      "grad_norm": 1.7651071548461914,
+      "learning_rate": 0.00018720000000000002,
+      "loss": 0.8726,
+      "step": 937
+    },
+    {
+      "epoch": 0.301171937710708,
+      "grad_norm": 1.3931697607040405,
+      "learning_rate": 0.00018740000000000003,
+      "loss": 0.9752,
+      "step": 938
+    },
+    {
+      "epoch": 0.3014930165355595,
+      "grad_norm": 1.4437761306762695,
+      "learning_rate": 0.0001876,
+      "loss": 0.8859,
+      "step": 939
+    },
+    {
+      "epoch": 0.301814095360411,
+      "grad_norm": 2.0850377082824707,
+      "learning_rate": 0.0001878,
+      "loss": 1.0133,
+      "step": 940
+    },
+    {
+      "epoch": 0.3021351741852625,
+      "grad_norm": 2.9318671226501465,
+      "learning_rate": 0.000188,
+      "loss": 1.0726,
+      "step": 941
+    },
+    {
+      "epoch": 0.302456253010114,
+      "grad_norm": 1.589202880859375,
+      "learning_rate": 0.0001882,
+      "loss": 0.8896,
+      "step": 942
+    },
+    {
+      "epoch": 0.30277733183496547,
+      "grad_norm": 1.5876160860061646,
+      "learning_rate": 0.0001884,
+      "loss": 0.8636,
+      "step": 943
+    },
+    {
+      "epoch": 0.30309841065981696,
+      "grad_norm": 1.6258230209350586,
+      "learning_rate": 0.0001886,
+      "loss": 1.0084,
+      "step": 944
+    },
+    {
+      "epoch": 0.30341948948466846,
+      "grad_norm": 2.1613311767578125,
+      "learning_rate": 0.0001888,
+      "loss": 0.8062,
+      "step": 945
+    },
+    {
+      "epoch": 0.30374056830952,
+      "grad_norm": 1.6936204433441162,
+      "learning_rate": 0.00018899999999999999,
+      "loss": 1.1147,
+      "step": 946
+    },
+    {
+      "epoch": 0.3040616471343715,
+      "grad_norm": 2.857898473739624,
+      "learning_rate": 0.0001892,
+      "loss": 0.8206,
+      "step": 947
+    },
+    {
+      "epoch": 0.304382725959223,
+      "grad_norm": 1.1513748168945312,
+      "learning_rate": 0.0001894,
+      "loss": 0.7425,
+      "step": 948
+    },
+    {
+      "epoch": 0.3047038047840745,
+      "grad_norm": 1.6653155088424683,
+      "learning_rate": 0.0001896,
+      "loss": 0.9059,
+      "step": 949
+    },
+    {
+      "epoch": 0.305024883608926,
+      "grad_norm": 1.6235933303833008,
+      "learning_rate": 0.0001898,
+      "loss": 0.7386,
+      "step": 950
+    },
+    {
+      "epoch": 0.3053459624337775,
+      "grad_norm": 3.6082000732421875,
+      "learning_rate": 0.00019,
+      "loss": 1.5358,
+      "step": 951
+    },
+    {
+      "epoch": 0.305667041258629,
+      "grad_norm": 2.132099151611328,
+      "learning_rate": 0.0001902,
+      "loss": 1.4807,
+      "step": 952
+    },
+    {
+      "epoch": 0.3059881200834805,
+      "grad_norm": 1.9843887090682983,
+      "learning_rate": 0.0001904,
+      "loss": 0.91,
+      "step": 953
+    },
+    {
+      "epoch": 0.30630919890833197,
+      "grad_norm": 1.8752169609069824,
+      "learning_rate": 0.0001906,
+      "loss": 0.686,
+      "step": 954
+    },
+    {
+      "epoch": 0.3066302777331835,
+      "grad_norm": 3.0422487258911133,
+      "learning_rate": 0.0001908,
+      "loss": 1.0079,
+      "step": 955
+    },
+    {
+      "epoch": 0.306951356558035,
+      "grad_norm": 2.0845787525177,
+      "learning_rate": 0.000191,
+      "loss": 0.8389,
+      "step": 956
+    },
+    {
+      "epoch": 0.3072724353828865,
+      "grad_norm": 2.3233184814453125,
+      "learning_rate": 0.0001912,
+      "loss": 0.9484,
+      "step": 957
+    },
+    {
+      "epoch": 0.307593514207738,
+      "grad_norm": 2.147631883621216,
+      "learning_rate": 0.0001914,
+      "loss": 1.0776,
+      "step": 958
+    },
+    {
+      "epoch": 0.3079145930325895,
+      "grad_norm": 1.7034640312194824,
+      "learning_rate": 0.0001916,
+      "loss": 0.9197,
+      "step": 959
+    },
+    {
+      "epoch": 0.308235671857441,
+      "grad_norm": 1.8278881311416626,
+      "learning_rate": 0.0001918,
+      "loss": 1.1262,
+      "step": 960
+    },
+    {
+      "epoch": 0.3085567506822925,
+      "grad_norm": 2.158137321472168,
+      "learning_rate": 0.000192,
+      "loss": 1.0464,
+      "step": 961
+    },
+    {
+      "epoch": 0.308877829507144,
+      "grad_norm": 1.8429055213928223,
+      "learning_rate": 0.0001922,
+      "loss": 1.1075,
+      "step": 962
+    },
+    {
+      "epoch": 0.3091989083319955,
+      "grad_norm": 1.6389920711517334,
+      "learning_rate": 0.00019240000000000001,
+      "loss": 1.1258,
+      "step": 963
+    },
+    {
+      "epoch": 0.30951998715684703,
+      "grad_norm": 1.337121605873108,
+      "learning_rate": 0.0001926,
+      "loss": 0.789,
+      "step": 964
+    },
+    {
+      "epoch": 0.3098410659816985,
+      "grad_norm": 1.542604684829712,
+      "learning_rate": 0.0001928,
+      "loss": 1.1683,
+      "step": 965
+    },
+    {
+      "epoch": 0.31016214480655,
+      "grad_norm": 1.8985679149627686,
+      "learning_rate": 0.000193,
+      "loss": 0.8494,
+      "step": 966
+    },
+    {
+      "epoch": 0.3104832236314015,
+      "grad_norm": 1.403977870941162,
+      "learning_rate": 0.0001932,
+      "loss": 1.2153,
+      "step": 967
+    },
+    {
+      "epoch": 0.310804302456253,
+      "grad_norm": 1.4651598930358887,
+      "learning_rate": 0.0001934,
+      "loss": 1.0217,
+      "step": 968
+    },
+    {
+      "epoch": 0.3111253812811045,
+      "grad_norm": 1.3526400327682495,
+      "learning_rate": 0.00019360000000000002,
+      "loss": 0.956,
+      "step": 969
+    },
+    {
+      "epoch": 0.311446460105956,
+      "grad_norm": 3.0191409587860107,
+      "learning_rate": 0.0001938,
+      "loss": 1.1278,
+      "step": 970
+    },
+    {
+      "epoch": 0.3117675389308075,
+      "grad_norm": 1.9719985723495483,
+      "learning_rate": 0.000194,
+      "loss": 1.1472,
+      "step": 971
+    },
+    {
+      "epoch": 0.312088617755659,
+      "grad_norm": 2.3695621490478516,
+      "learning_rate": 0.0001942,
+      "loss": 1.1406,
+      "step": 972
+    },
+    {
+      "epoch": 0.31240969658051054,
+      "grad_norm": 1.8187494277954102,
+      "learning_rate": 0.0001944,
+      "loss": 1.0193,
+      "step": 973
+    },
+    {
+      "epoch": 0.31273077540536204,
+      "grad_norm": 1.494602084159851,
+      "learning_rate": 0.00019460000000000001,
+      "loss": 1.0561,
+      "step": 974
+    },
+    {
+      "epoch": 0.31305185423021353,
+      "grad_norm": 2.378328323364258,
+      "learning_rate": 0.0001948,
+      "loss": 0.9976,
+      "step": 975
+    },
+    {
+      "epoch": 0.31337293305506503,
+      "grad_norm": 1.5211881399154663,
+      "learning_rate": 0.000195,
+      "loss": 1.1273,
+      "step": 976
+    },
+    {
+      "epoch": 0.3136940118799165,
+      "grad_norm": 1.2796664237976074,
+      "learning_rate": 0.0001952,
+      "loss": 0.8582,
+      "step": 977
+    },
+    {
+      "epoch": 0.314015090704768,
+      "grad_norm": 1.8555333614349365,
+      "learning_rate": 0.0001954,
+      "loss": 1.0359,
+      "step": 978
+    },
+    {
+      "epoch": 0.3143361695296195,
+      "grad_norm": 1.1893086433410645,
+      "learning_rate": 0.0001956,
+      "loss": 0.8822,
+      "step": 979
+    },
+    {
+      "epoch": 0.314657248354471,
+      "grad_norm": 2.2236552238464355,
+      "learning_rate": 0.00019580000000000002,
+      "loss": 1.0201,
+      "step": 980
+    },
+    {
+      "epoch": 0.3149783271793225,
+      "grad_norm": 2.314167022705078,
+      "learning_rate": 0.000196,
+      "loss": 0.9865,
+      "step": 981
+    },
+    {
+      "epoch": 0.315299406004174,
+      "grad_norm": 2.2062315940856934,
+      "learning_rate": 0.0001962,
+      "loss": 1.117,
+      "step": 982
+    },
+    {
+      "epoch": 0.31562048482902555,
+      "grad_norm": 2.284191846847534,
+      "learning_rate": 0.0001964,
+      "loss": 1.1285,
+      "step": 983
+    },
+    {
+      "epoch": 0.31594156365387704,
+      "grad_norm": 2.3977997303009033,
+      "learning_rate": 0.0001966,
+      "loss": 1.2463,
+      "step": 984
+    },
+    {
+      "epoch": 0.31626264247872854,
+      "grad_norm": 2.204314708709717,
+      "learning_rate": 0.0001968,
+      "loss": 0.8849,
+      "step": 985
+    },
+    {
+      "epoch": 0.31658372130358003,
+      "grad_norm": 3.1310677528381348,
+      "learning_rate": 0.00019700000000000002,
+      "loss": 1.2409,
+      "step": 986
+    },
+    {
+      "epoch": 0.31690480012843153,
+      "grad_norm": 1.4078696966171265,
+      "learning_rate": 0.0001972,
+      "loss": 0.8572,
+      "step": 987
+    },
+    {
+      "epoch": 0.317225878953283,
+      "grad_norm": 1.9044156074523926,
+      "learning_rate": 0.0001974,
+      "loss": 0.9414,
+      "step": 988
+    },
+    {
+      "epoch": 0.3175469577781345,
+      "grad_norm": 1.3988277912139893,
+      "learning_rate": 0.0001976,
+      "loss": 1.1101,
+      "step": 989
+    },
+    {
+      "epoch": 0.317868036602986,
+      "grad_norm": 1.4820644855499268,
+      "learning_rate": 0.0001978,
+      "loss": 1.0266,
+      "step": 990
+    },
+    {
+      "epoch": 0.3181891154278375,
+      "grad_norm": 1.5065770149230957,
+      "learning_rate": 0.00019800000000000002,
+      "loss": 0.8049,
+      "step": 991
+    },
+    {
+      "epoch": 0.31851019425268906,
+      "grad_norm": 1.5369571447372437,
+      "learning_rate": 0.00019820000000000002,
+      "loss": 0.9029,
+      "step": 992
+    },
+    {
+      "epoch": 0.31883127307754056,
+      "grad_norm": 4.79142951965332,
+      "learning_rate": 0.0001984,
+      "loss": 0.9041,
+      "step": 993
+    },
+    {
+      "epoch": 0.31915235190239205,
+      "grad_norm": 2.016233205795288,
+      "learning_rate": 0.0001986,
+      "loss": 1.0123,
+      "step": 994
+    },
+    {
+      "epoch": 0.31947343072724355,
+      "grad_norm": 1.5374374389648438,
+      "learning_rate": 0.0001988,
+      "loss": 0.8762,
+      "step": 995
+    },
+    {
+      "epoch": 0.31979450955209504,
+      "grad_norm": 1.3565267324447632,
+      "learning_rate": 0.000199,
+      "loss": 0.7286,
+      "step": 996
+    },
+    {
+      "epoch": 0.32011558837694654,
+      "grad_norm": 2.1161487102508545,
+      "learning_rate": 0.00019920000000000002,
+      "loss": 1.025,
+      "step": 997
+    },
+    {
+      "epoch": 0.32043666720179803,
+      "grad_norm": 1.741025686264038,
+      "learning_rate": 0.00019940000000000002,
+      "loss": 0.7437,
+      "step": 998
+    },
+    {
+      "epoch": 0.3207577460266495,
+      "grad_norm": 2.049365997314453,
+      "learning_rate": 0.0001996,
+      "loss": 0.7756,
+      "step": 999
+    },
+    {
+      "epoch": 0.321078824851501,
+      "grad_norm": 1.8877449035644531,
+      "learning_rate": 0.0001998,
+      "loss": 0.695,
+      "step": 1000
+    },
+    {
+      "epoch": 0.32139990367635257,
+      "grad_norm": 2.1886143684387207,
+      "learning_rate": 0.0002,
+      "loss": 1.4415,
+      "step": 1001
+    },
+    {
+      "epoch": 0.32172098250120407,
+      "grad_norm": 2.847805976867676,
+      "learning_rate": 0.00019999999390765168,
+      "loss": 1.2191,
+      "step": 1002
+    },
+    {
+      "epoch": 0.32204206132605556,
+      "grad_norm": 2.2231152057647705,
+      "learning_rate": 0.0001999999756306074,
+      "loss": 1.0706,
+      "step": 1003
+    },
+    {
+      "epoch": 0.32236314015090706,
+      "grad_norm": 2.219747304916382,
+      "learning_rate": 0.00019999994516886946,
+      "loss": 1.0274,
+      "step": 1004
+    },
+    {
+      "epoch": 0.32268421897575855,
+      "grad_norm": 2.3217408657073975,
+      "learning_rate": 0.00019999990252244151,
+      "loss": 0.9952,
+      "step": 1005
+    },
+    {
+      "epoch": 0.32300529780061005,
+      "grad_norm": 2.1023569107055664,
+      "learning_rate": 0.00019999984769132877,
+      "loss": 0.9075,
+      "step": 1006
+    },
+    {
+      "epoch": 0.32332637662546154,
+      "grad_norm": 2.414299249649048,
+      "learning_rate": 0.00019999978067553796,
+      "loss": 0.916,
+      "step": 1007
+    },
+    {
+      "epoch": 0.32364745545031304,
+      "grad_norm": 1.7825126647949219,
+      "learning_rate": 0.00019999970147507713,
+      "loss": 1.0684,
+      "step": 1008
+    },
+    {
+      "epoch": 0.32396853427516453,
+      "grad_norm": 1.9296294450759888,
+      "learning_rate": 0.00019999961008995606,
+      "loss": 1.1201,
+      "step": 1009
+    },
+    {
+      "epoch": 0.324289613100016,
+      "grad_norm": 1.7151492834091187,
+      "learning_rate": 0.00019999950652018584,
+      "loss": 0.947,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3246106919248676,
+      "grad_norm": 1.7226728200912476,
+      "learning_rate": 0.00019999939076577905,
+      "loss": 1.0448,
+      "step": 1011
+    },
+    {
+      "epoch": 0.3249317707497191,
+      "grad_norm": 1.8700830936431885,
+      "learning_rate": 0.00019999926282674983,
+      "loss": 0.9668,
+      "step": 1012
+    },
+    {
+      "epoch": 0.32525284957457057,
+      "grad_norm": 1.468245267868042,
+      "learning_rate": 0.00019999912270311375,
+      "loss": 0.8688,
+      "step": 1013
+    },
+    {
+      "epoch": 0.32557392839942206,
+      "grad_norm": 2.5088555812835693,
+      "learning_rate": 0.00019999897039488796,
+      "loss": 1.0542,
+      "step": 1014
+    },
+    {
+      "epoch": 0.32589500722427356,
+      "grad_norm": 1.5317341089248657,
+      "learning_rate": 0.0001999988059020909,
+      "loss": 1.1052,
+      "step": 1015
+    },
+    {
+      "epoch": 0.32621608604912505,
+      "grad_norm": 1.1775538921356201,
+      "learning_rate": 0.00019999862922474268,
+      "loss": 1.0154,
+      "step": 1016
+    },
+    {
+      "epoch": 0.32653716487397655,
+      "grad_norm": 1.4249135255813599,
+      "learning_rate": 0.00019999844036286483,
+      "loss": 0.9882,
+      "step": 1017
+    },
+    {
+      "epoch": 0.32685824369882804,
+      "grad_norm": 1.9009687900543213,
+      "learning_rate": 0.00019999823931648035,
+      "loss": 1.173,
+      "step": 1018
+    },
+    {
+      "epoch": 0.32717932252367954,
+      "grad_norm": 2.4772071838378906,
+      "learning_rate": 0.0001999980260856137,
+      "loss": 1.2154,
+      "step": 1019
+    },
+    {
+      "epoch": 0.3275004013485311,
+      "grad_norm": 2.5375864505767822,
+      "learning_rate": 0.00019999780067029094,
+      "loss": 1.1588,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3278214801733826,
+      "grad_norm": 2.0548741817474365,
+      "learning_rate": 0.00019999756307053948,
+      "loss": 1.17,
+      "step": 1021
+    },
+    {
+      "epoch": 0.3281425589982341,
+      "grad_norm": 2.063422441482544,
+      "learning_rate": 0.00019999731328638828,
+      "loss": 0.9767,
+      "step": 1022
+    },
+    {
+      "epoch": 0.3284636378230856,
+      "grad_norm": 1.4420465230941772,
+      "learning_rate": 0.0001999970513178678,
+      "loss": 0.9672,
+      "step": 1023
+    },
+    {
+      "epoch": 0.32878471664793707,
+      "grad_norm": 8.236002922058105,
+      "learning_rate": 0.00019999677716500992,
+      "loss": 1.4308,
+      "step": 1024
+    },
+    {
+      "epoch": 0.32910579547278856,
+      "grad_norm": 1.764096736907959,
+      "learning_rate": 0.0001999964908278481,
+      "loss": 0.889,
+      "step": 1025
+    },
+    {
+      "epoch": 0.32942687429764006,
+      "grad_norm": 1.7221022844314575,
+      "learning_rate": 0.00019999619230641713,
+      "loss": 1.0662,
+      "step": 1026
+    },
+    {
+      "epoch": 0.32974795312249155,
+      "grad_norm": 1.4062892198562622,
+      "learning_rate": 0.00019999588160075348,
+      "loss": 1.034,
+      "step": 1027
+    },
+    {
+      "epoch": 0.33006903194734305,
+      "grad_norm": 1.5284225940704346,
+      "learning_rate": 0.000199995558710895,
+      "loss": 1.0265,
+      "step": 1028
+    },
+    {
+      "epoch": 0.3303901107721946,
+      "grad_norm": 2.0174477100372314,
+      "learning_rate": 0.000199995223636881,
+      "loss": 1.0751,
+      "step": 1029
+    },
+    {
+      "epoch": 0.3307111895970461,
+      "grad_norm": 1.3949508666992188,
+      "learning_rate": 0.0001999948763787523,
+      "loss": 1.0296,
+      "step": 1030
+    },
+    {
+      "epoch": 0.3310322684218976,
+      "grad_norm": 1.5616528987884521,
+      "learning_rate": 0.00019999451693655123,
+      "loss": 1.0017,
+      "step": 1031
+    },
+    {
+      "epoch": 0.3313533472467491,
+      "grad_norm": 1.674304723739624,
+      "learning_rate": 0.00019999414531032158,
+      "loss": 1.1445,
+      "step": 1032
+    },
+    {
+      "epoch": 0.3316744260716006,
+      "grad_norm": 2.715172290802002,
+      "learning_rate": 0.00019999376150010866,
+      "loss": 1.1429,
+      "step": 1033
+    },
+    {
+      "epoch": 0.3319955048964521,
+      "grad_norm": 2.873642683029175,
+      "learning_rate": 0.0001999933655059592,
+      "loss": 0.6858,
+      "step": 1034
+    },
+    {
+      "epoch": 0.33231658372130357,
+      "grad_norm": 1.6501632928848267,
+      "learning_rate": 0.00019999295732792146,
+      "loss": 1.1858,
+      "step": 1035
+    },
+    {
+      "epoch": 0.33263766254615507,
+      "grad_norm": 1.3060718774795532,
+      "learning_rate": 0.00019999253696604522,
+      "loss": 1.0933,
+      "step": 1036
+    },
+    {
+      "epoch": 0.33295874137100656,
+      "grad_norm": 0.9465512633323669,
+      "learning_rate": 0.00019999210442038162,
+      "loss": 0.8341,
+      "step": 1037
+    },
+    {
+      "epoch": 0.33327982019585806,
+      "grad_norm": 1.2716128826141357,
+      "learning_rate": 0.00019999165969098343,
+      "loss": 0.8147,
+      "step": 1038
+    },
+    {
+      "epoch": 0.3336008990207096,
+      "grad_norm": 1.762526035308838,
+      "learning_rate": 0.00019999120277790477,
+      "loss": 1.0783,
+      "step": 1039
+    },
+    {
+      "epoch": 0.3339219778455611,
+      "grad_norm": 2.2095861434936523,
+      "learning_rate": 0.0001999907336812014,
+      "loss": 1.1156,
+      "step": 1040
+    },
+    {
+      "epoch": 0.3342430566704126,
+      "grad_norm": 1.668186068534851,
+      "learning_rate": 0.00019999025240093044,
+      "loss": 0.8137,
+      "step": 1041
+    },
+    {
+      "epoch": 0.3345641354952641,
+      "grad_norm": 1.4128243923187256,
+      "learning_rate": 0.0001999897589371505,
+      "loss": 1.0737,
+      "step": 1042
+    },
+    {
+      "epoch": 0.3348852143201156,
+      "grad_norm": 2.0269153118133545,
+      "learning_rate": 0.00019998925328992175,
+      "loss": 1.0204,
+      "step": 1043
+    },
+    {
+      "epoch": 0.3352062931449671,
+      "grad_norm": 1.2762573957443237,
+      "learning_rate": 0.0001999887354593058,
+      "loss": 1.1014,
+      "step": 1044
+    },
+    {
+      "epoch": 0.3355273719698186,
+      "grad_norm": 1.0873875617980957,
+      "learning_rate": 0.0001999882054453657,
+      "loss": 0.7875,
+      "step": 1045
+    },
+    {
+      "epoch": 0.3358484507946701,
+      "grad_norm": 1.1676608324050903,
+      "learning_rate": 0.00019998766324816607,
+      "loss": 0.9086,
+      "step": 1046
+    },
+    {
+      "epoch": 0.33616952961952157,
+      "grad_norm": 1.2658352851867676,
+      "learning_rate": 0.000199987108867773,
+      "loss": 0.7835,
+      "step": 1047
+    },
+    {
+      "epoch": 0.3364906084443731,
+      "grad_norm": 3.1894469261169434,
+      "learning_rate": 0.00019998654230425395,
+      "loss": 1.0001,
+      "step": 1048
+    },
+    {
+      "epoch": 0.3368116872692246,
+      "grad_norm": 2.7069225311279297,
+      "learning_rate": 0.00019998596355767805,
+      "loss": 0.6518,
+      "step": 1049
+    },
+    {
+      "epoch": 0.3371327660940761,
+      "grad_norm": 1.401463508605957,
+      "learning_rate": 0.00019998537262811577,
+      "loss": 0.6609,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3374538449189276,
+      "grad_norm": 2.923952102661133,
+      "learning_rate": 0.00019998476951563915,
+      "loss": 1.4718,
+      "step": 1051
+    },
+    {
+      "epoch": 0.3377749237437791,
+      "grad_norm": 4.130305767059326,
+      "learning_rate": 0.00019998415422032163,
+      "loss": 1.109,
+      "step": 1052
+    },
+    {
+      "epoch": 0.3380960025686306,
+      "grad_norm": 2.697674036026001,
+      "learning_rate": 0.00019998352674223816,
+      "loss": 0.9685,
+      "step": 1053
+    },
+    {
+      "epoch": 0.3384170813934821,
+      "grad_norm": 13.59544849395752,
+      "learning_rate": 0.00019998288708146527,
+      "loss": 1.4453,
+      "step": 1054
+    },
+    {
+      "epoch": 0.3387381602183336,
+      "grad_norm": 2.960008144378662,
+      "learning_rate": 0.0001999822352380809,
+      "loss": 1.0084,
+      "step": 1055
+    },
+    {
+      "epoch": 0.3390592390431851,
+      "grad_norm": 1.8568581342697144,
+      "learning_rate": 0.00019998157121216444,
+      "loss": 0.8141,
+      "step": 1056
+    },
+    {
+      "epoch": 0.33938031786803663,
+      "grad_norm": 2.4931318759918213,
+      "learning_rate": 0.00019998089500379676,
+      "loss": 0.9077,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3397013966928881,
+      "grad_norm": 2.592142105102539,
+      "learning_rate": 0.00019998020661306037,
+      "loss": 0.967,
+      "step": 1058
+    },
+    {
+      "epoch": 0.3400224755177396,
+      "grad_norm": 2.2816684246063232,
+      "learning_rate": 0.000199979506040039,
+      "loss": 1.0623,
+      "step": 1059
+    },
+    {
+      "epoch": 0.3403435543425911,
+      "grad_norm": 1.5964871644973755,
+      "learning_rate": 0.00019997879328481814,
+      "loss": 0.9404,
+      "step": 1060
+    },
+    {
+      "epoch": 0.3406646331674426,
+      "grad_norm": 1.8557053804397583,
+      "learning_rate": 0.00019997806834748456,
+      "loss": 0.9474,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3409857119922941,
+      "grad_norm": 1.6931266784667969,
+      "learning_rate": 0.0001999773312281266,
+      "loss": 1.216,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3413067908171456,
+      "grad_norm": 1.908123254776001,
+      "learning_rate": 0.00019997658192683412,
+      "loss": 1.0184,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3416278696419971,
+      "grad_norm": 1.5717891454696655,
+      "learning_rate": 0.00019997582044369843,
+      "loss": 1.022,
+      "step": 1064
+    },
+    {
+      "epoch": 0.3419489484668486,
+      "grad_norm": 1.7411060333251953,
+      "learning_rate": 0.00019997504677881224,
+      "loss": 0.9467,
+      "step": 1065
+    },
+    {
+      "epoch": 0.3422700272917001,
+      "grad_norm": 3.3722310066223145,
+      "learning_rate": 0.00019997426093226986,
+      "loss": 0.9437,
+      "step": 1066
+    },
+    {
+      "epoch": 0.34259110611655164,
+      "grad_norm": 1.1751868724822998,
+      "learning_rate": 0.000199973462904167,
+      "loss": 0.9127,
+      "step": 1067
+    },
+    {
+      "epoch": 0.34291218494140313,
+      "grad_norm": 1.7528998851776123,
+      "learning_rate": 0.000199972652694601,
+      "loss": 0.9023,
+      "step": 1068
+    },
+    {
+      "epoch": 0.3432332637662546,
+      "grad_norm": 1.4955263137817383,
+      "learning_rate": 0.00019997183030367048,
+      "loss": 1.3134,
+      "step": 1069
+    },
+    {
+      "epoch": 0.3435543425911061,
+      "grad_norm": 2.127793550491333,
+      "learning_rate": 0.0001999709957314757,
+      "loss": 1.0524,
+      "step": 1070
+    },
+    {
+      "epoch": 0.3438754214159576,
+      "grad_norm": 1.1916999816894531,
+      "learning_rate": 0.00019997014897811833,
+      "loss": 0.8918,
+      "step": 1071
+    },
+    {
+      "epoch": 0.3441965002408091,
+      "grad_norm": 2.142920970916748,
+      "learning_rate": 0.00019996929004370152,
+      "loss": 0.9653,
+      "step": 1072
+    },
+    {
+      "epoch": 0.3445175790656606,
+      "grad_norm": 1.9465276002883911,
+      "learning_rate": 0.00019996841892833,
+      "loss": 1.1276,
+      "step": 1073
+    },
+    {
+      "epoch": 0.3448386578905121,
+      "grad_norm": 4.488142490386963,
+      "learning_rate": 0.00019996753563210985,
+      "loss": 1.2271,
+      "step": 1074
+    },
+    {
+      "epoch": 0.3451597367153636,
+      "grad_norm": 9.381378173828125,
+      "learning_rate": 0.00019996664015514871,
+      "loss": 0.894,
+      "step": 1075
+    },
+    {
+      "epoch": 0.34548081554021515,
+      "grad_norm": 1.1667776107788086,
+      "learning_rate": 0.00019996573249755572,
+      "loss": 0.8693,
+      "step": 1076
+    },
+    {
+      "epoch": 0.34580189436506664,
+      "grad_norm": 1.7447643280029297,
+      "learning_rate": 0.00019996481265944146,
+      "loss": 1.0464,
+      "step": 1077
+    },
+    {
+      "epoch": 0.34612297318991814,
+      "grad_norm": 1.9767264127731323,
+      "learning_rate": 0.000199963880640918,
+      "loss": 1.1109,
+      "step": 1078
+    },
+    {
+      "epoch": 0.34644405201476963,
+      "grad_norm": 1.110445499420166,
+      "learning_rate": 0.00019996293644209887,
+      "loss": 1.0815,
+      "step": 1079
+    },
+    {
+      "epoch": 0.3467651308396211,
+      "grad_norm": 2.0246994495391846,
+      "learning_rate": 0.0001999619800630992,
+      "loss": 1.1254,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3470862096644726,
+      "grad_norm": 1.2259923219680786,
+      "learning_rate": 0.00019996101150403543,
+      "loss": 0.7593,
+      "step": 1081
+    },
+    {
+      "epoch": 0.3474072884893241,
+      "grad_norm": 1.174506425857544,
+      "learning_rate": 0.00019996003076502565,
+      "loss": 1.0263,
+      "step": 1082
+    },
+    {
+      "epoch": 0.3477283673141756,
+      "grad_norm": 1.8576600551605225,
+      "learning_rate": 0.00019995903784618936,
+      "loss": 1.0134,
+      "step": 1083
+    },
+    {
+      "epoch": 0.3480494461390271,
+      "grad_norm": 1.9805476665496826,
+      "learning_rate": 0.00019995803274764747,
+      "loss": 1.0945,
+      "step": 1084
+    },
+    {
+      "epoch": 0.34837052496387866,
+      "grad_norm": 1.4714640378952026,
+      "learning_rate": 0.0001999570154695225,
+      "loss": 1.0238,
+      "step": 1085
+    },
+    {
+      "epoch": 0.34869160378873015,
+      "grad_norm": 2.3097331523895264,
+      "learning_rate": 0.00019995598601193842,
+      "loss": 0.972,
+      "step": 1086
+    },
+    {
+      "epoch": 0.34901268261358165,
+      "grad_norm": 1.2940665483474731,
+      "learning_rate": 0.00019995494437502064,
+      "loss": 0.8954,
+      "step": 1087
+    },
+    {
+      "epoch": 0.34933376143843314,
+      "grad_norm": 1.3209530115127563,
+      "learning_rate": 0.00019995389055889605,
+      "loss": 0.9124,
+      "step": 1088
+    },
+    {
+      "epoch": 0.34965484026328464,
+      "grad_norm": 1.4666227102279663,
+      "learning_rate": 0.0001999528245636931,
+      "loss": 1.0218,
+      "step": 1089
+    },
+    {
+      "epoch": 0.34997591908813613,
+      "grad_norm": 2.331610918045044,
+      "learning_rate": 0.0001999517463895417,
+      "loss": 0.9959,
+      "step": 1090
+    },
+    {
+      "epoch": 0.35029699791298763,
+      "grad_norm": 1.3941978216171265,
+      "learning_rate": 0.00019995065603657316,
+      "loss": 0.9071,
+      "step": 1091
+    },
+    {
+      "epoch": 0.3506180767378391,
+      "grad_norm": 1.3337680101394653,
+      "learning_rate": 0.00019994955350492038,
+      "loss": 1.08,
+      "step": 1092
+    },
+    {
+      "epoch": 0.3509391555626906,
+      "grad_norm": 1.6207735538482666,
+      "learning_rate": 0.0001999484387947177,
+      "loss": 0.972,
+      "step": 1093
+    },
+    {
+      "epoch": 0.3512602343875421,
+      "grad_norm": 1.4743515253067017,
+      "learning_rate": 0.00019994731190610087,
+      "loss": 1.1447,
+      "step": 1094
+    },
+    {
+      "epoch": 0.35158131321239366,
+      "grad_norm": 1.3109478950500488,
+      "learning_rate": 0.0001999461728392073,
+      "loss": 0.9514,
+      "step": 1095
+    },
+    {
+      "epoch": 0.35190239203724516,
+      "grad_norm": 1.4319483041763306,
+      "learning_rate": 0.00019994502159417573,
+      "loss": 0.9688,
+      "step": 1096
+    },
+    {
+      "epoch": 0.35222347086209665,
+      "grad_norm": 1.4653232097625732,
+      "learning_rate": 0.00019994385817114646,
+      "loss": 0.9073,
+      "step": 1097
+    },
+    {
+      "epoch": 0.35254454968694815,
+      "grad_norm": 1.3482818603515625,
+      "learning_rate": 0.00019994268257026118,
+      "loss": 0.7425,
+      "step": 1098
+    },
+    {
+      "epoch": 0.35286562851179964,
+      "grad_norm": 2.7118563652038574,
+      "learning_rate": 0.00019994149479166324,
+      "loss": 0.8003,
+      "step": 1099
+    },
+    {
+      "epoch": 0.35318670733665114,
+      "grad_norm": 1.2395905256271362,
+      "learning_rate": 0.0001999402948354973,
+      "loss": 0.6656,
+      "step": 1100
+    },
+    {
+      "epoch": 0.35350778616150264,
+      "grad_norm": 2.8349075317382812,
+      "learning_rate": 0.0001999390827019096,
+      "loss": 1.369,
+      "step": 1101
+    },
+    {
+      "epoch": 0.35382886498635413,
+      "grad_norm": 1.6994882822036743,
+      "learning_rate": 0.0001999378583910478,
+      "loss": 1.4933,
+      "step": 1102
+    },
+    {
+      "epoch": 0.3541499438112056,
+      "grad_norm": 2.044842481613159,
+      "learning_rate": 0.0001999366219030611,
+      "loss": 1.0908,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3544710226360572,
+      "grad_norm": 2.008758306503296,
+      "learning_rate": 0.00019993537323810014,
+      "loss": 0.963,
+      "step": 1104
+    },
+    {
+      "epoch": 0.35479210146090867,
+      "grad_norm": 2.8811473846435547,
+      "learning_rate": 0.0001999341123963171,
+      "loss": 0.8424,
+      "step": 1105
+    },
+    {
+      "epoch": 0.35511318028576017,
+      "grad_norm": 1.9336885213851929,
+      "learning_rate": 0.00019993283937786563,
+      "loss": 0.9013,
+      "step": 1106
+    },
+    {
+      "epoch": 0.35543425911061166,
+      "grad_norm": 1.6170105934143066,
+      "learning_rate": 0.0001999315541829008,
+      "loss": 0.7265,
+      "step": 1107
+    },
+    {
+      "epoch": 0.35575533793546316,
+      "grad_norm": 2.101649284362793,
+      "learning_rate": 0.0001999302568115792,
+      "loss": 1.1258,
+      "step": 1108
+    },
+    {
+      "epoch": 0.35607641676031465,
+      "grad_norm": 2.104947328567505,
+      "learning_rate": 0.00019992894726405893,
+      "loss": 0.9501,
+      "step": 1109
+    },
+    {
+      "epoch": 0.35639749558516615,
+      "grad_norm": 1.6322426795959473,
+      "learning_rate": 0.00019992762554049955,
+      "loss": 1.0367,
+      "step": 1110
+    },
+    {
+      "epoch": 0.35671857441001764,
+      "grad_norm": 1.6271668672561646,
+      "learning_rate": 0.0001999262916410621,
+      "loss": 0.893,
+      "step": 1111
+    },
+    {
+      "epoch": 0.35703965323486914,
+      "grad_norm": 1.3786333799362183,
+      "learning_rate": 0.00019992494556590916,
+      "loss": 0.9951,
+      "step": 1112
+    },
+    {
+      "epoch": 0.3573607320597207,
+      "grad_norm": 1.2673497200012207,
+      "learning_rate": 0.00019992358731520468,
+      "loss": 0.9283,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3576818108845722,
+      "grad_norm": 1.338150143623352,
+      "learning_rate": 0.0001999222168891142,
+      "loss": 0.9409,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3580028897094237,
+      "grad_norm": 1.344650149345398,
+      "learning_rate": 0.0001999208342878047,
+      "loss": 1.0276,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3583239685342752,
+      "grad_norm": 1.805612325668335,
+      "learning_rate": 0.0001999194395114446,
+      "loss": 1.1324,
+      "step": 1116
+    },
+    {
+      "epoch": 0.35864504735912667,
+      "grad_norm": 21.904685974121094,
+      "learning_rate": 0.00019991803256020393,
+      "loss": 1.1808,
+      "step": 1117
+    },
+    {
+      "epoch": 0.35896612618397816,
+      "grad_norm": 1.6144940853118896,
+      "learning_rate": 0.000199916613434254,
+      "loss": 1.1416,
+      "step": 1118
+    },
+    {
+      "epoch": 0.35928720500882966,
+      "grad_norm": 2.7180378437042236,
+      "learning_rate": 0.00019991518213376787,
+      "loss": 1.0546,
+      "step": 1119
+    },
+    {
+      "epoch": 0.35960828383368115,
+      "grad_norm": 1.8319542407989502,
+      "learning_rate": 0.00019991373865891986,
+      "loss": 0.991,
+      "step": 1120
+    },
+    {
+      "epoch": 0.35992936265853265,
+      "grad_norm": 2.3465638160705566,
+      "learning_rate": 0.00019991228300988585,
+      "loss": 1.0318,
+      "step": 1121
+    },
+    {
+      "epoch": 0.3602504414833842,
+      "grad_norm": 2.1191277503967285,
+      "learning_rate": 0.00019991081518684321,
+      "loss": 1.1896,
+      "step": 1122
+    },
+    {
+      "epoch": 0.3605715203082357,
+      "grad_norm": 7.694565773010254,
+      "learning_rate": 0.00019990933518997084,
+      "loss": 1.3582,
+      "step": 1123
+    },
+    {
+      "epoch": 0.3608925991330872,
+      "grad_norm": 1.5988242626190186,
+      "learning_rate": 0.00019990784301944902,
+      "loss": 1.225,
+      "step": 1124
+    },
+    {
+      "epoch": 0.3612136779579387,
+      "grad_norm": 1.4618570804595947,
+      "learning_rate": 0.00019990633867545955,
+      "loss": 1.152,
+      "step": 1125
+    },
+    {
+      "epoch": 0.3615347567827902,
+      "grad_norm": 1.4518593549728394,
+      "learning_rate": 0.0001999048221581858,
+      "loss": 0.7903,
+      "step": 1126
+    },
+    {
+      "epoch": 0.3618558356076417,
+      "grad_norm": 2.108856439590454,
+      "learning_rate": 0.0001999032934678125,
+      "loss": 0.9205,
+      "step": 1127
+    },
+    {
+      "epoch": 0.36217691443249317,
+      "grad_norm": 1.7413921356201172,
+      "learning_rate": 0.0001999017526045259,
+      "loss": 0.9283,
+      "step": 1128
+    },
+    {
+      "epoch": 0.36249799325734466,
+      "grad_norm": 1.6088074445724487,
+      "learning_rate": 0.00019990019956851382,
+      "loss": 1.0622,
+      "step": 1129
+    },
+    {
+      "epoch": 0.36281907208219616,
+      "grad_norm": 2.035529375076294,
+      "learning_rate": 0.00019989863435996544,
+      "loss": 1.0937,
+      "step": 1130
+    },
+    {
+      "epoch": 0.36314015090704765,
+      "grad_norm": 1.6727184057235718,
+      "learning_rate": 0.00019989705697907149,
+      "loss": 1.133,
+      "step": 1131
+    },
+    {
+      "epoch": 0.3634612297318992,
+      "grad_norm": 4.085322856903076,
+      "learning_rate": 0.00019989546742602414,
+      "loss": 0.8277,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3637823085567507,
+      "grad_norm": 1.4142351150512695,
+      "learning_rate": 0.00019989386570101714,
+      "loss": 1.0965,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3641033873816022,
+      "grad_norm": 2.0783348083496094,
+      "learning_rate": 0.0001998922518042456,
+      "loss": 1.4158,
+      "step": 1134
+    },
+    {
+      "epoch": 0.3644244662064537,
+      "grad_norm": 1.3447226285934448,
+      "learning_rate": 0.00019989062573590616,
+      "loss": 0.8552,
+      "step": 1135
+    },
+    {
+      "epoch": 0.3647455450313052,
+      "grad_norm": 1.5264992713928223,
+      "learning_rate": 0.00019988898749619702,
+      "loss": 0.9846,
+      "step": 1136
+    },
+    {
+      "epoch": 0.3650666238561567,
+      "grad_norm": 1.8115450143814087,
+      "learning_rate": 0.0001998873370853177,
+      "loss": 1.0604,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3653877026810082,
+      "grad_norm": 2.269794225692749,
+      "learning_rate": 0.00019988567450346937,
+      "loss": 0.8845,
+      "step": 1138
+    },
+    {
+      "epoch": 0.36570878150585967,
+      "grad_norm": 1.198961853981018,
+      "learning_rate": 0.0001998839997508546,
+      "loss": 0.9815,
+      "step": 1139
+    },
+    {
+      "epoch": 0.36602986033071117,
+      "grad_norm": 2.3479223251342773,
+      "learning_rate": 0.00019988231282767744,
+      "loss": 0.9144,
+      "step": 1140
+    },
+    {
+      "epoch": 0.3663509391555627,
+      "grad_norm": 1.9824615716934204,
+      "learning_rate": 0.0001998806137341434,
+      "loss": 1.0468,
+      "step": 1141
+    },
+    {
+      "epoch": 0.3666720179804142,
+      "grad_norm": 1.780985713005066,
+      "learning_rate": 0.00019987890247045957,
+      "loss": 1.0552,
+      "step": 1142
+    },
+    {
+      "epoch": 0.3669930968052657,
+      "grad_norm": 1.192861557006836,
+      "learning_rate": 0.00019987717903683448,
+      "loss": 0.8069,
+      "step": 1143
+    },
+    {
+      "epoch": 0.3673141756301172,
+      "grad_norm": 2.568082094192505,
+      "learning_rate": 0.00019987544343347803,
+      "loss": 0.9455,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3676352544549687,
+      "grad_norm": 1.151548981666565,
+      "learning_rate": 0.00019987369566060176,
+      "loss": 0.6518,
+      "step": 1145
+    },
+    {
+      "epoch": 0.3679563332798202,
+      "grad_norm": 2.208691358566284,
+      "learning_rate": 0.0001998719357184186,
+      "loss": 0.7998,
+      "step": 1146
+    },
+    {
+      "epoch": 0.3682774121046717,
+      "grad_norm": 2.1074390411376953,
+      "learning_rate": 0.00019987016360714307,
+      "loss": 1.0609,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3685984909295232,
+      "grad_norm": 1.2993351221084595,
+      "learning_rate": 0.00019986837932699103,
+      "loss": 0.7665,
+      "step": 1148
+    },
+    {
+      "epoch": 0.3689195697543747,
+      "grad_norm": 0.9701393246650696,
+      "learning_rate": 0.00019986658287817987,
+      "loss": 0.5932,
+      "step": 1149
+    },
+    {
+      "epoch": 0.3692406485792262,
+      "grad_norm": 6.661820411682129,
+      "learning_rate": 0.00019986477426092855,
+      "loss": 0.6992,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3695617274040777,
+      "grad_norm": 3.1357054710388184,
+      "learning_rate": 0.0001998629534754574,
+      "loss": 1.5801,
+      "step": 1151
+    },
+    {
+      "epoch": 0.3698828062289292,
+      "grad_norm": 3.9719338417053223,
+      "learning_rate": 0.0001998611205219883,
+      "loss": 1.5882,
+      "step": 1152
+    },
+    {
+      "epoch": 0.3702038850537807,
+      "grad_norm": 3.1035382747650146,
+      "learning_rate": 0.00019985927540074454,
+      "loss": 1.0743,
+      "step": 1153
+    },
+    {
+      "epoch": 0.3705249638786322,
+      "grad_norm": 1.9447718858718872,
+      "learning_rate": 0.00019985741811195097,
+      "loss": 0.8629,
+      "step": 1154
+    },
+    {
+      "epoch": 0.3708460427034837,
+      "grad_norm": 1.993913173675537,
+      "learning_rate": 0.00019985554865583393,
+      "loss": 0.8233,
+      "step": 1155
+    },
+    {
+      "epoch": 0.3711671215283352,
+      "grad_norm": 1.8630907535552979,
+      "learning_rate": 0.0001998536670326212,
+      "loss": 0.8642,
+      "step": 1156
+    },
+    {
+      "epoch": 0.3714882003531867,
+      "grad_norm": 1.8467717170715332,
+      "learning_rate": 0.000199851773242542,
+      "loss": 0.8528,
+      "step": 1157
+    },
+    {
+      "epoch": 0.3718092791780382,
+      "grad_norm": 2.4014041423797607,
+      "learning_rate": 0.0001998498672858271,
+      "loss": 1.0696,
+      "step": 1158
+    },
+    {
+      "epoch": 0.3721303580028897,
+      "grad_norm": 1.915739893913269,
+      "learning_rate": 0.00019984794916270876,
+      "loss": 1.0595,
+      "step": 1159
+    },
+    {
+      "epoch": 0.37245143682774123,
+      "grad_norm": 1.4160325527191162,
+      "learning_rate": 0.00019984601887342073,
+      "loss": 1.1056,
+      "step": 1160
+    },
+    {
+      "epoch": 0.37277251565259273,
+      "grad_norm": 2.503619909286499,
+      "learning_rate": 0.00019984407641819812,
+      "loss": 0.6714,
+      "step": 1161
+    },
+    {
+      "epoch": 0.3730935944774442,
+      "grad_norm": 1.4751476049423218,
+      "learning_rate": 0.00019984212179727766,
+      "loss": 0.9467,
+      "step": 1162
+    },
+    {
+      "epoch": 0.3734146733022957,
+      "grad_norm": 1.4288772344589233,
+      "learning_rate": 0.00019984015501089752,
+      "loss": 1.0041,
+      "step": 1163
+    },
+    {
+      "epoch": 0.3737357521271472,
+      "grad_norm": 1.389568567276001,
+      "learning_rate": 0.00019983817605929733,
+      "loss": 0.9213,
+      "step": 1164
+    },
+    {
+      "epoch": 0.3740568309519987,
+      "grad_norm": 1.2409948110580444,
+      "learning_rate": 0.00019983618494271826,
+      "loss": 0.9687,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3743779097768502,
+      "grad_norm": 1.3233766555786133,
+      "learning_rate": 0.00019983418166140285,
+      "loss": 0.9057,
+      "step": 1166
+    },
+    {
+      "epoch": 0.3746989886017017,
+      "grad_norm": 1.7424918413162231,
+      "learning_rate": 0.00019983216621559524,
+      "loss": 1.3071,
+      "step": 1167
+    },
+    {
+      "epoch": 0.3750200674265532,
+      "grad_norm": 2.5900015830993652,
+      "learning_rate": 0.00019983013860554101,
+      "loss": 0.9437,
+      "step": 1168
+    },
+    {
+      "epoch": 0.37534114625140474,
+      "grad_norm": 1.2561944723129272,
+      "learning_rate": 0.00019982809883148722,
+      "loss": 1.0481,
+      "step": 1169
+    },
+    {
+      "epoch": 0.37566222507625624,
+      "grad_norm": 2.1015701293945312,
+      "learning_rate": 0.00019982604689368239,
+      "loss": 1.1775,
+      "step": 1170
+    },
+    {
+      "epoch": 0.37598330390110773,
+      "grad_norm": 2.039940357208252,
+      "learning_rate": 0.00019982398279237655,
+      "loss": 1.2701,
+      "step": 1171
+    },
+    {
+      "epoch": 0.37630438272595923,
+      "grad_norm": 1.330772042274475,
+      "learning_rate": 0.0001998219065278212,
+      "loss": 0.9725,
+      "step": 1172
+    },
+    {
+      "epoch": 0.3766254615508107,
+      "grad_norm": 2.106564998626709,
+      "learning_rate": 0.00019981981810026934,
+      "loss": 1.0919,
+      "step": 1173
+    },
+    {
+      "epoch": 0.3769465403756622,
+      "grad_norm": 1.5898199081420898,
+      "learning_rate": 0.0001998177175099754,
+      "loss": 0.9706,
+      "step": 1174
+    },
+    {
+      "epoch": 0.3772676192005137,
+      "grad_norm": 1.8556996583938599,
+      "learning_rate": 0.00019981560475719538,
+      "loss": 1.025,
+      "step": 1175
+    },
+    {
+      "epoch": 0.3775886980253652,
+      "grad_norm": 1.008709192276001,
+      "learning_rate": 0.0001998134798421867,
+      "loss": 0.8611,
+      "step": 1176
+    },
+    {
+      "epoch": 0.3779097768502167,
+      "grad_norm": 2.131082057952881,
+      "learning_rate": 0.00019981134276520827,
+      "loss": 1.2977,
+      "step": 1177
+    },
+    {
+      "epoch": 0.37823085567506826,
+      "grad_norm": 1.7559832334518433,
+      "learning_rate": 0.00019980919352652048,
+      "loss": 0.9628,
+      "step": 1178
+    },
+    {
+      "epoch": 0.37855193449991975,
+      "grad_norm": 1.290353536605835,
+      "learning_rate": 0.00019980703212638522,
+      "loss": 1.0724,
+      "step": 1179
+    },
+    {
+      "epoch": 0.37887301332477125,
+      "grad_norm": 1.12660813331604,
+      "learning_rate": 0.00019980485856506582,
+      "loss": 0.9982,
+      "step": 1180
+    },
+    {
+      "epoch": 0.37919409214962274,
+      "grad_norm": 1.5011529922485352,
+      "learning_rate": 0.00019980267284282717,
+      "loss": 0.9109,
+      "step": 1181
+    },
+    {
+      "epoch": 0.37951517097447424,
+      "grad_norm": 1.762739896774292,
+      "learning_rate": 0.00019980047495993554,
+      "loss": 0.9277,
+      "step": 1182
+    },
+    {
+      "epoch": 0.37983624979932573,
+      "grad_norm": 1.3952839374542236,
+      "learning_rate": 0.00019979826491665881,
+      "loss": 0.7895,
+      "step": 1183
+    },
+    {
+      "epoch": 0.3801573286241772,
+      "grad_norm": 1.6314538717269897,
+      "learning_rate": 0.00019979604271326616,
+      "loss": 0.9361,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3804784074490287,
+      "grad_norm": 2.3080692291259766,
+      "learning_rate": 0.00019979380835002846,
+      "loss": 1.1475,
+      "step": 1185
+    },
+    {
+      "epoch": 0.3807994862738802,
+      "grad_norm": 2.15315842628479,
+      "learning_rate": 0.0001997915618272179,
+      "loss": 1.1916,
+      "step": 1186
+    },
+    {
+      "epoch": 0.3811205650987317,
+      "grad_norm": 1.3747003078460693,
+      "learning_rate": 0.00019978930314510824,
+      "loss": 0.8076,
+      "step": 1187
+    },
+    {
+      "epoch": 0.38144164392358326,
+      "grad_norm": 1.4667130708694458,
+      "learning_rate": 0.0001997870323039747,
+      "loss": 0.95,
+      "step": 1188
+    },
+    {
+      "epoch": 0.38176272274843476,
+      "grad_norm": 1.726277470588684,
+      "learning_rate": 0.00019978474930409395,
+      "loss": 0.9735,
+      "step": 1189
+    },
+    {
+      "epoch": 0.38208380157328625,
+      "grad_norm": 2.355656385421753,
+      "learning_rate": 0.00019978245414574417,
+      "loss": 1.0268,
+      "step": 1190
+    },
+    {
+      "epoch": 0.38240488039813775,
+      "grad_norm": 1.840656042098999,
+      "learning_rate": 0.000199780146829205,
+      "loss": 0.9747,
+      "step": 1191
+    },
+    {
+      "epoch": 0.38272595922298924,
+      "grad_norm": 1.4547884464263916,
+      "learning_rate": 0.00019977782735475764,
+      "loss": 1.0729,
+      "step": 1192
+    },
+    {
+      "epoch": 0.38304703804784074,
+      "grad_norm": 1.2916489839553833,
+      "learning_rate": 0.00019977549572268468,
+      "loss": 0.875,
+      "step": 1193
+    },
+    {
+      "epoch": 0.38336811687269223,
+      "grad_norm": 1.5631312131881714,
+      "learning_rate": 0.00019977315193327018,
+      "loss": 0.8071,
+      "step": 1194
+    },
+    {
+      "epoch": 0.38368919569754373,
+      "grad_norm": 1.6755515336990356,
+      "learning_rate": 0.00019977079598679977,
+      "loss": 0.9683,
+      "step": 1195
+    },
+    {
+      "epoch": 0.3840102745223952,
+      "grad_norm": 1.1934847831726074,
+      "learning_rate": 0.00019976842788356055,
+      "loss": 0.666,
+      "step": 1196
+    },
+    {
+      "epoch": 0.3843313533472468,
+      "grad_norm": 1.6635997295379639,
+      "learning_rate": 0.000199766047623841,
+      "loss": 0.8521,
+      "step": 1197
+    },
+    {
+      "epoch": 0.38465243217209827,
+      "grad_norm": 1.6572028398513794,
+      "learning_rate": 0.00019976365520793114,
+      "loss": 0.7134,
+      "step": 1198
+    },
+    {
+      "epoch": 0.38497351099694976,
+      "grad_norm": 1.2580087184906006,
+      "learning_rate": 0.00019976125063612252,
+      "loss": 0.6517,
+      "step": 1199
+    },
+    {
+      "epoch": 0.38529458982180126,
+      "grad_norm": 1.3600796461105347,
+      "learning_rate": 0.00019975883390870817,
+      "loss": 0.5476,
+      "step": 1200
+    },
+    {
+      "epoch": 0.38561566864665275,
+      "grad_norm": 2.1080329418182373,
+      "learning_rate": 0.00019975640502598244,
+      "loss": 1.4574,
+      "step": 1201
+    },
+    {
+      "epoch": 0.38593674747150425,
+      "grad_norm": 1.5729361772537231,
+      "learning_rate": 0.0001997539639882414,
+      "loss": 1.3152,
+      "step": 1202
+    },
+    {
+      "epoch": 0.38625782629635574,
+      "grad_norm": 1.7094898223876953,
+      "learning_rate": 0.00019975151079578237,
+      "loss": 1.1812,
+      "step": 1203
+    },
+    {
+      "epoch": 0.38657890512120724,
+      "grad_norm": 1.5421949625015259,
+      "learning_rate": 0.0001997490454489044,
+      "loss": 1.023,
+      "step": 1204
+    },
+    {
+      "epoch": 0.38689998394605873,
+      "grad_norm": 2.4401865005493164,
+      "learning_rate": 0.00019974656794790775,
+      "loss": 0.866,
+      "step": 1205
+    },
+    {
+      "epoch": 0.3872210627709103,
+      "grad_norm": 1.9785239696502686,
+      "learning_rate": 0.0001997440782930944,
+      "loss": 0.8572,
+      "step": 1206
+    },
+    {
+      "epoch": 0.3875421415957618,
+      "grad_norm": 1.7715507745742798,
+      "learning_rate": 0.00019974157648476766,
+      "loss": 0.7572,
+      "step": 1207
+    },
+    {
+      "epoch": 0.3878632204206133,
+      "grad_norm": 1.8642398118972778,
+      "learning_rate": 0.00019973906252323238,
+      "loss": 0.8058,
+      "step": 1208
+    },
+    {
+      "epoch": 0.38818429924546477,
+      "grad_norm": 1.5486159324645996,
+      "learning_rate": 0.00019973653640879485,
+      "loss": 1.0713,
+      "step": 1209
+    },
+    {
+      "epoch": 0.38850537807031627,
+      "grad_norm": 2.690429449081421,
+      "learning_rate": 0.00019973399814176293,
+      "loss": 1.024,
+      "step": 1210
+    },
+    {
+      "epoch": 0.38882645689516776,
+      "grad_norm": 1.7301387786865234,
+      "learning_rate": 0.00019973144772244582,
+      "loss": 0.9245,
+      "step": 1211
+    },
+    {
+      "epoch": 0.38914753572001926,
+      "grad_norm": 1.1060234308242798,
+      "learning_rate": 0.00019972888515115434,
+      "loss": 0.9201,
+      "step": 1212
+    },
+    {
+      "epoch": 0.38946861454487075,
+      "grad_norm": 2.06071138381958,
+      "learning_rate": 0.0001997263104282007,
+      "loss": 0.8777,
+      "step": 1213
+    },
+    {
+      "epoch": 0.38978969336972225,
+      "grad_norm": 1.2055797576904297,
+      "learning_rate": 0.00019972372355389867,
+      "loss": 0.9946,
+      "step": 1214
+    },
+    {
+      "epoch": 0.39011077219457374,
+      "grad_norm": 1.2430628538131714,
+      "learning_rate": 0.00019972112452856339,
+      "loss": 0.6915,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3904318510194253,
+      "grad_norm": 1.3444316387176514,
+      "learning_rate": 0.00019971851335251158,
+      "loss": 0.9794,
+      "step": 1216
+    },
+    {
+      "epoch": 0.3907529298442768,
+      "grad_norm": 1.4662611484527588,
+      "learning_rate": 0.0001997158900260614,
+      "loss": 1.0779,
+      "step": 1217
+    },
+    {
+      "epoch": 0.3910740086691283,
+      "grad_norm": 1.2781003713607788,
+      "learning_rate": 0.00019971325454953248,
+      "loss": 0.9051,
+      "step": 1218
+    },
+    {
+      "epoch": 0.3913950874939798,
+      "grad_norm": 1.5118160247802734,
+      "learning_rate": 0.00019971060692324598,
+      "loss": 1.3483,
+      "step": 1219
+    },
+    {
+      "epoch": 0.39171616631883127,
+      "grad_norm": 3.3005893230438232,
+      "learning_rate": 0.00019970794714752445,
+      "loss": 1.0498,
+      "step": 1220
+    },
+    {
+      "epoch": 0.39203724514368277,
+      "grad_norm": 2.5156784057617188,
+      "learning_rate": 0.00019970527522269205,
+      "loss": 1.0823,
+      "step": 1221
+    },
+    {
+      "epoch": 0.39235832396853426,
+      "grad_norm": 2.2197351455688477,
+      "learning_rate": 0.00019970259114907425,
+      "loss": 0.9576,
+      "step": 1222
+    },
+    {
+      "epoch": 0.39267940279338576,
+      "grad_norm": 1.4013001918792725,
+      "learning_rate": 0.0001996998949269982,
+      "loss": 1.0194,
+      "step": 1223
+    },
+    {
+      "epoch": 0.39300048161823725,
+      "grad_norm": 1.1013851165771484,
+      "learning_rate": 0.00019969718655679232,
+      "loss": 0.9049,
+      "step": 1224
+    },
+    {
+      "epoch": 0.3933215604430888,
+      "grad_norm": 1.4185377359390259,
+      "learning_rate": 0.00019969446603878673,
+      "loss": 0.8871,
+      "step": 1225
+    },
+    {
+      "epoch": 0.3936426392679403,
+      "grad_norm": 1.4201257228851318,
+      "learning_rate": 0.0001996917333733128,
+      "loss": 0.8107,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3939637180927918,
+      "grad_norm": 1.788902997970581,
+      "learning_rate": 0.0001996889885607036,
+      "loss": 1.1504,
+      "step": 1227
+    },
+    {
+      "epoch": 0.3942847969176433,
+      "grad_norm": 2.3521134853363037,
+      "learning_rate": 0.0001996862316012935,
+      "loss": 1.2274,
+      "step": 1228
+    },
+    {
+      "epoch": 0.3946058757424948,
+      "grad_norm": 2.3193109035491943,
+      "learning_rate": 0.00019968346249541846,
+      "loss": 1.2382,
+      "step": 1229
+    },
+    {
+      "epoch": 0.3949269545673463,
+      "grad_norm": 1.1716614961624146,
+      "learning_rate": 0.0001996806812434159,
+      "loss": 0.8529,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3952480333921978,
+      "grad_norm": 1.5683951377868652,
+      "learning_rate": 0.00019967788784562473,
+      "loss": 1.1172,
+      "step": 1231
+    },
+    {
+      "epoch": 0.39556911221704927,
+      "grad_norm": 1.3435286283493042,
+      "learning_rate": 0.00019967508230238522,
+      "loss": 1.1469,
+      "step": 1232
+    },
+    {
+      "epoch": 0.39589019104190076,
+      "grad_norm": 2.070988893508911,
+      "learning_rate": 0.00019967226461403933,
+      "loss": 1.1132,
+      "step": 1233
+    },
+    {
+      "epoch": 0.3962112698667523,
+      "grad_norm": 1.4258763790130615,
+      "learning_rate": 0.0001996694347809303,
+      "loss": 1.0245,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3965323486916038,
+      "grad_norm": 1.191821813583374,
+      "learning_rate": 0.00019966659280340297,
+      "loss": 1.0535,
+      "step": 1235
+    },
+    {
+      "epoch": 0.3968534275164553,
+      "grad_norm": 1.3982123136520386,
+      "learning_rate": 0.00019966373868180365,
+      "loss": 0.8893,
+      "step": 1236
+    },
+    {
+      "epoch": 0.3971745063413068,
+      "grad_norm": 1.734355092048645,
+      "learning_rate": 0.0001996608724164801,
+      "loss": 1.1121,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3974955851661583,
+      "grad_norm": 1.516452670097351,
+      "learning_rate": 0.00019965799400778152,
+      "loss": 1.0574,
+      "step": 1238
+    },
+    {
+      "epoch": 0.3978166639910098,
+      "grad_norm": 2.239095449447632,
+      "learning_rate": 0.00019965510345605866,
+      "loss": 1.0554,
+      "step": 1239
+    },
+    {
+      "epoch": 0.3981377428158613,
+      "grad_norm": 1.5699093341827393,
+      "learning_rate": 0.00019965220076166376,
+      "loss": 1.0202,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3984588216407128,
+      "grad_norm": 1.7507424354553223,
+      "learning_rate": 0.00019964928592495045,
+      "loss": 0.9151,
+      "step": 1241
+    },
+    {
+      "epoch": 0.3987799004655643,
+      "grad_norm": 2.988863229751587,
+      "learning_rate": 0.0001996463589462739,
+      "loss": 0.8941,
+      "step": 1242
+    },
+    {
+      "epoch": 0.39910097929041577,
+      "grad_norm": 2.8474647998809814,
+      "learning_rate": 0.00019964341982599078,
+      "loss": 0.9185,
+      "step": 1243
+    },
+    {
+      "epoch": 0.3994220581152673,
+      "grad_norm": 1.7581831216812134,
+      "learning_rate": 0.00019964046856445924,
+      "loss": 0.9486,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3997431369401188,
+      "grad_norm": 1.2665711641311646,
+      "learning_rate": 0.00019963750516203884,
+      "loss": 0.816,
+      "step": 1245
+    },
+    {
+      "epoch": 0.4000642157649703,
+      "grad_norm": 1.6899489164352417,
+      "learning_rate": 0.00019963452961909063,
+      "loss": 0.692,
+      "step": 1246
+    },
+    {
+      "epoch": 0.4003852945898218,
+      "grad_norm": 1.0340851545333862,
+      "learning_rate": 0.00019963154193597727,
+      "loss": 0.7222,
+      "step": 1247
+    },
+    {
+      "epoch": 0.4007063734146733,
+      "grad_norm": 2.3070790767669678,
+      "learning_rate": 0.00019962854211306267,
+      "loss": 0.9091,
+      "step": 1248
+    },
+    {
+      "epoch": 0.4010274522395248,
+      "grad_norm": 1.4314565658569336,
+      "learning_rate": 0.0001996255301507125,
+      "loss": 0.6755,
+      "step": 1249
+    },
+    {
+      "epoch": 0.4013485310643763,
+      "grad_norm": 0.9915478825569153,
+      "learning_rate": 0.0001996225060492936,
+      "loss": 0.6343,
+      "step": 1250
+    },
+    {
+      "epoch": 0.4016696098892278,
+      "grad_norm": 1.879836916923523,
+      "learning_rate": 0.00019961946980917456,
+      "loss": 1.3669,
+      "step": 1251
+    },
+    {
+      "epoch": 0.4019906887140793,
+      "grad_norm": 2.172273874282837,
+      "learning_rate": 0.00019961642143072529,
+      "loss": 1.2117,
+      "step": 1252
+    },
+    {
+      "epoch": 0.40231176753893083,
+      "grad_norm": 2.1239922046661377,
+      "learning_rate": 0.00019961336091431727,
+      "loss": 1.0067,
+      "step": 1253
+    },
+    {
+      "epoch": 0.4026328463637823,
+      "grad_norm": 1.5531266927719116,
+      "learning_rate": 0.00019961028826032332,
+      "loss": 0.7723,
+      "step": 1254
+    },
+    {
+      "epoch": 0.4029539251886338,
+      "grad_norm": 1.8615666627883911,
+      "learning_rate": 0.00019960720346911797,
+      "loss": 0.8307,
+      "step": 1255
+    },
+    {
+      "epoch": 0.4032750040134853,
+      "grad_norm": 1.9920127391815186,
+      "learning_rate": 0.00019960410654107697,
+      "loss": 0.7862,
+      "step": 1256
+    },
+    {
+      "epoch": 0.4035960828383368,
+      "grad_norm": 1.7379018068313599,
+      "learning_rate": 0.00019960099747657774,
+      "loss": 0.8312,
+      "step": 1257
+    },
+    {
+      "epoch": 0.4039171616631883,
+      "grad_norm": 2.0180602073669434,
+      "learning_rate": 0.00019959787627599906,
+      "loss": 0.8846,
+      "step": 1258
+    },
+    {
+      "epoch": 0.4042382404880398,
+      "grad_norm": 1.567842960357666,
+      "learning_rate": 0.00019959474293972129,
+      "loss": 1.0824,
+      "step": 1259
+    },
+    {
+      "epoch": 0.4045593193128913,
+      "grad_norm": 2.6913554668426514,
+      "learning_rate": 0.0001995915974681262,
+      "loss": 1.0713,
+      "step": 1260
+    },
+    {
+      "epoch": 0.4048803981377428,
+      "grad_norm": 2.3127224445343018,
+      "learning_rate": 0.00019958843986159704,
+      "loss": 0.8672,
+      "step": 1261
+    },
+    {
+      "epoch": 0.40520147696259434,
+      "grad_norm": 2.383711338043213,
+      "learning_rate": 0.00019958527012051857,
+      "loss": 0.9304,
+      "step": 1262
+    },
+    {
+      "epoch": 0.40552255578744584,
+      "grad_norm": 1.3288320302963257,
+      "learning_rate": 0.000199582088245277,
+      "loss": 1.0369,
+      "step": 1263
+    },
+    {
+      "epoch": 0.40584363461229733,
+      "grad_norm": 1.3985912799835205,
+      "learning_rate": 0.00019957889423626005,
+      "loss": 1.0793,
+      "step": 1264
+    },
+    {
+      "epoch": 0.4061647134371488,
+      "grad_norm": 1.3505330085754395,
+      "learning_rate": 0.00019957568809385694,
+      "loss": 0.8989,
+      "step": 1265
+    },
+    {
+      "epoch": 0.4064857922620003,
+      "grad_norm": 1.663252353668213,
+      "learning_rate": 0.00019957246981845822,
+      "loss": 1.0797,
+      "step": 1266
+    },
+    {
+      "epoch": 0.4068068710868518,
+      "grad_norm": 1.4397826194763184,
+      "learning_rate": 0.0001995692394104561,
+      "loss": 0.8712,
+      "step": 1267
+    },
+    {
+      "epoch": 0.4071279499117033,
+      "grad_norm": 1.1795347929000854,
+      "learning_rate": 0.0001995659968702442,
+      "loss": 1.1528,
+      "step": 1268
+    },
+    {
+      "epoch": 0.4074490287365548,
+      "grad_norm": 2.6851541996002197,
+      "learning_rate": 0.00019956274219821757,
+      "loss": 0.9325,
+      "step": 1269
+    },
+    {
+      "epoch": 0.4077701075614063,
+      "grad_norm": 1.8018854856491089,
+      "learning_rate": 0.00019955947539477284,
+      "loss": 1.1969,
+      "step": 1270
+    },
+    {
+      "epoch": 0.40809118638625785,
+      "grad_norm": 1.442039966583252,
+      "learning_rate": 0.00019955619646030802,
+      "loss": 0.9039,
+      "step": 1271
+    },
+    {
+      "epoch": 0.40841226521110935,
+      "grad_norm": 1.8041740655899048,
+      "learning_rate": 0.0001995529053952226,
+      "loss": 1.0132,
+      "step": 1272
+    },
+    {
+      "epoch": 0.40873334403596084,
+      "grad_norm": 1.6970267295837402,
+      "learning_rate": 0.0001995496021999177,
+      "loss": 1.2557,
+      "step": 1273
+    },
+    {
+      "epoch": 0.40905442286081234,
+      "grad_norm": 2.2318642139434814,
+      "learning_rate": 0.00019954628687479572,
+      "loss": 0.9386,
+      "step": 1274
+    },
+    {
+      "epoch": 0.40937550168566383,
+      "grad_norm": 1.3459020853042603,
+      "learning_rate": 0.00019954295942026064,
+      "loss": 0.979,
+      "step": 1275
+    },
+    {
+      "epoch": 0.40969658051051533,
+      "grad_norm": 1.2683597803115845,
+      "learning_rate": 0.00019953961983671788,
+      "loss": 0.8684,
+      "step": 1276
+    },
+    {
+      "epoch": 0.4100176593353668,
+      "grad_norm": 2.0490283966064453,
+      "learning_rate": 0.0001995362681245744,
+      "loss": 1.1168,
+      "step": 1277
+    },
+    {
+      "epoch": 0.4103387381602183,
+      "grad_norm": 1.9572994709014893,
+      "learning_rate": 0.00019953290428423857,
+      "loss": 0.8459,
+      "step": 1278
+    },
+    {
+      "epoch": 0.4106598169850698,
+      "grad_norm": 1.6033639907836914,
+      "learning_rate": 0.00019952952831612026,
+      "loss": 1.1142,
+      "step": 1279
+    },
+    {
+      "epoch": 0.4109808958099213,
+      "grad_norm": 1.5231640338897705,
+      "learning_rate": 0.00019952614022063084,
+      "loss": 1.1958,
+      "step": 1280
+    },
+    {
+      "epoch": 0.41130197463477286,
+      "grad_norm": 1.5248790979385376,
+      "learning_rate": 0.0001995227399981831,
+      "loss": 1.2117,
+      "step": 1281
+    },
+    {
+      "epoch": 0.41162305345962436,
+      "grad_norm": 1.1884125471115112,
+      "learning_rate": 0.00019951932764919144,
+      "loss": 0.8868,
+      "step": 1282
+    },
+    {
+      "epoch": 0.41194413228447585,
+      "grad_norm": 1.509526014328003,
+      "learning_rate": 0.00019951590317407152,
+      "loss": 1.0724,
+      "step": 1283
+    },
+    {
+      "epoch": 0.41226521110932735,
+      "grad_norm": 1.9606530666351318,
+      "learning_rate": 0.0001995124665732407,
+      "loss": 1.1933,
+      "step": 1284
+    },
+    {
+      "epoch": 0.41258628993417884,
+      "grad_norm": 1.8705965280532837,
+      "learning_rate": 0.00019950901784711764,
+      "loss": 1.1115,
+      "step": 1285
+    },
+    {
+      "epoch": 0.41290736875903034,
+      "grad_norm": 2.0867602825164795,
+      "learning_rate": 0.00019950555699612267,
+      "loss": 1.2534,
+      "step": 1286
+    },
+    {
+      "epoch": 0.41322844758388183,
+      "grad_norm": 1.6749221086502075,
+      "learning_rate": 0.00019950208402067733,
+      "loss": 1.1313,
+      "step": 1287
+    },
+    {
+      "epoch": 0.4135495264087333,
+      "grad_norm": 1.0996752977371216,
+      "learning_rate": 0.00019949859892120491,
+      "loss": 0.9663,
+      "step": 1288
+    },
+    {
+      "epoch": 0.4138706052335848,
+      "grad_norm": 1.0107436180114746,
+      "learning_rate": 0.00019949510169813003,
+      "loss": 0.8468,
+      "step": 1289
+    },
+    {
+      "epoch": 0.41419168405843637,
+      "grad_norm": 1.7090445756912231,
+      "learning_rate": 0.0001994915923518788,
+      "loss": 1.1206,
+      "step": 1290
+    },
+    {
+      "epoch": 0.41451276288328787,
+      "grad_norm": 1.2257174253463745,
+      "learning_rate": 0.00019948807088287883,
+      "loss": 0.9886,
+      "step": 1291
+    },
+    {
+      "epoch": 0.41483384170813936,
+      "grad_norm": 1.3304343223571777,
+      "learning_rate": 0.00019948453729155922,
+      "loss": 0.9679,
+      "step": 1292
+    },
+    {
+      "epoch": 0.41515492053299086,
+      "grad_norm": 3.28393816947937,
+      "learning_rate": 0.00019948099157835047,
+      "loss": 0.9752,
+      "step": 1293
+    },
+    {
+      "epoch": 0.41547599935784235,
+      "grad_norm": 1.4946495294570923,
+      "learning_rate": 0.00019947743374368467,
+      "loss": 0.9856,
+      "step": 1294
+    },
+    {
+      "epoch": 0.41579707818269385,
+      "grad_norm": 1.3553723096847534,
+      "learning_rate": 0.00019947386378799532,
+      "loss": 0.8419,
+      "step": 1295
+    },
+    {
+      "epoch": 0.41611815700754534,
+      "grad_norm": 2.194031238555908,
+      "learning_rate": 0.00019947028171171742,
+      "loss": 0.8151,
+      "step": 1296
+    },
+    {
+      "epoch": 0.41643923583239684,
+      "grad_norm": 1.5983561277389526,
+      "learning_rate": 0.00019946668751528744,
+      "loss": 0.8907,
+      "step": 1297
+    },
+    {
+      "epoch": 0.41676031465724833,
+      "grad_norm": 1.9155287742614746,
+      "learning_rate": 0.00019946308119914323,
+      "loss": 0.9095,
+      "step": 1298
+    },
+    {
+      "epoch": 0.4170813934820999,
+      "grad_norm": 1.078909993171692,
+      "learning_rate": 0.00019945946276372434,
+      "loss": 0.6413,
+      "step": 1299
+    },
+    {
+      "epoch": 0.4174024723069514,
+      "grad_norm": 1.317728042602539,
+      "learning_rate": 0.00019945583220947158,
+      "loss": 0.7406,
+      "step": 1300
+    },
+    {
+      "epoch": 0.4177235511318029,
+      "grad_norm": 1.7083852291107178,
+      "learning_rate": 0.00019945218953682734,
+      "loss": 1.3869,
+      "step": 1301
+    },
+    {
+      "epoch": 0.41804462995665437,
+      "grad_norm": 2.715884208679199,
+      "learning_rate": 0.00019944853474623548,
+      "loss": 1.4385,
+      "step": 1302
+    },
+    {
+      "epoch": 0.41836570878150586,
+      "grad_norm": 1.6352131366729736,
+      "learning_rate": 0.00019944486783814134,
+      "loss": 1.0713,
+      "step": 1303
+    },
+    {
+      "epoch": 0.41868678760635736,
+      "grad_norm": 1.9549245834350586,
+      "learning_rate": 0.00019944118881299168,
+      "loss": 0.876,
+      "step": 1304
+    },
+    {
+      "epoch": 0.41900786643120885,
+      "grad_norm": 3.7763001918792725,
+      "learning_rate": 0.0001994374976712348,
+      "loss": 0.7917,
+      "step": 1305
+    },
+    {
+      "epoch": 0.41932894525606035,
+      "grad_norm": 1.8948286771774292,
+      "learning_rate": 0.00019943379441332047,
+      "loss": 0.8379,
+      "step": 1306
+    },
+    {
+      "epoch": 0.41965002408091184,
+      "grad_norm": 5.399472713470459,
+      "learning_rate": 0.0001994300790396999,
+      "loss": 0.8858,
+      "step": 1307
+    },
+    {
+      "epoch": 0.41997110290576334,
+      "grad_norm": 2.114410400390625,
+      "learning_rate": 0.0001994263515508258,
+      "loss": 0.8726,
+      "step": 1308
+    },
+    {
+      "epoch": 0.4202921817306149,
+      "grad_norm": 2.5239293575286865,
+      "learning_rate": 0.00019942261194715236,
+      "loss": 1.225,
+      "step": 1309
+    },
+    {
+      "epoch": 0.4206132605554664,
+      "grad_norm": 1.2384289503097534,
+      "learning_rate": 0.00019941886022913522,
+      "loss": 1.1042,
+      "step": 1310
+    },
+    {
+      "epoch": 0.4209343393803179,
+      "grad_norm": 1.1283193826675415,
+      "learning_rate": 0.00019941509639723155,
+      "loss": 1.073,
+      "step": 1311
+    },
+    {
+      "epoch": 0.4212554182051694,
+      "grad_norm": 1.4066673517227173,
+      "learning_rate": 0.0001994113204518999,
+      "loss": 1.0025,
+      "step": 1312
+    },
+    {
+      "epoch": 0.42157649703002087,
+      "grad_norm": 1.927209734916687,
+      "learning_rate": 0.00019940753239360047,
+      "loss": 0.8864,
+      "step": 1313
+    },
+    {
+      "epoch": 0.42189757585487236,
+      "grad_norm": 1.1925103664398193,
+      "learning_rate": 0.00019940373222279473,
+      "loss": 0.7443,
+      "step": 1314
+    },
+    {
+      "epoch": 0.42221865467972386,
+      "grad_norm": 0.9499497413635254,
+      "learning_rate": 0.0001993999199399457,
+      "loss": 0.7643,
+      "step": 1315
+    },
+    {
+      "epoch": 0.42253973350457535,
+      "grad_norm": 1.9734052419662476,
+      "learning_rate": 0.000199396095545518,
+      "loss": 0.9895,
+      "step": 1316
+    },
+    {
+      "epoch": 0.42286081232942685,
+      "grad_norm": 1.0893967151641846,
+      "learning_rate": 0.0001993922590399775,
+      "loss": 0.9146,
+      "step": 1317
+    },
+    {
+      "epoch": 0.4231818911542784,
+      "grad_norm": 2.039592742919922,
+      "learning_rate": 0.00019938841042379174,
+      "loss": 1.0294,
+      "step": 1318
+    },
+    {
+      "epoch": 0.4235029699791299,
+      "grad_norm": 1.233694314956665,
+      "learning_rate": 0.00019938454969742968,
+      "loss": 0.8611,
+      "step": 1319
+    },
+    {
+      "epoch": 0.4238240488039814,
+      "grad_norm": 2.1485254764556885,
+      "learning_rate": 0.00019938067686136167,
+      "loss": 1.0674,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4241451276288329,
+      "grad_norm": 1.7402235269546509,
+      "learning_rate": 0.00019937679191605963,
+      "loss": 1.1541,
+      "step": 1321
+    },
+    {
+      "epoch": 0.4244662064536844,
+      "grad_norm": 1.3716280460357666,
+      "learning_rate": 0.00019937289486199696,
+      "loss": 1.1266,
+      "step": 1322
+    },
+    {
+      "epoch": 0.4247872852785359,
+      "grad_norm": 1.6395798921585083,
+      "learning_rate": 0.00019936898569964848,
+      "loss": 0.9904,
+      "step": 1323
+    },
+    {
+      "epoch": 0.42510836410338737,
+      "grad_norm": 1.3475819826126099,
+      "learning_rate": 0.0001993650644294905,
+      "loss": 1.0421,
+      "step": 1324
+    },
+    {
+      "epoch": 0.42542944292823887,
+      "grad_norm": 1.9199656248092651,
+      "learning_rate": 0.00019936113105200085,
+      "loss": 1.1184,
+      "step": 1325
+    },
+    {
+      "epoch": 0.42575052175309036,
+      "grad_norm": 1.6179684400558472,
+      "learning_rate": 0.00019935718556765876,
+      "loss": 1.2786,
+      "step": 1326
+    },
+    {
+      "epoch": 0.4260716005779419,
+      "grad_norm": 1.5954012870788574,
+      "learning_rate": 0.000199353227976945,
+      "loss": 0.9325,
+      "step": 1327
+    },
+    {
+      "epoch": 0.4263926794027934,
+      "grad_norm": 1.832868218421936,
+      "learning_rate": 0.00019934925828034175,
+      "loss": 1.114,
+      "step": 1328
+    },
+    {
+      "epoch": 0.4267137582276449,
+      "grad_norm": 1.5652598142623901,
+      "learning_rate": 0.00019934527647833276,
+      "loss": 1.2399,
+      "step": 1329
+    },
+    {
+      "epoch": 0.4270348370524964,
+      "grad_norm": 2.4210355281829834,
+      "learning_rate": 0.0001993412825714032,
+      "loss": 0.9815,
+      "step": 1330
+    },
+    {
+      "epoch": 0.4273559158773479,
+      "grad_norm": 1.2339181900024414,
+      "learning_rate": 0.00019933727656003963,
+      "loss": 0.9054,
+      "step": 1331
+    },
+    {
+      "epoch": 0.4276769947021994,
+      "grad_norm": 2.5136241912841797,
+      "learning_rate": 0.0001993332584447303,
+      "loss": 1.1086,
+      "step": 1332
+    },
+    {
+      "epoch": 0.4279980735270509,
+      "grad_norm": 1.339241623878479,
+      "learning_rate": 0.00019932922822596473,
+      "loss": 1.2639,
+      "step": 1333
+    },
+    {
+      "epoch": 0.4283191523519024,
+      "grad_norm": 1.0698463916778564,
+      "learning_rate": 0.00019932518590423394,
+      "loss": 0.8401,
+      "step": 1334
+    },
+    {
+      "epoch": 0.42864023117675387,
+      "grad_norm": 2.2309930324554443,
+      "learning_rate": 0.00019932113148003058,
+      "loss": 0.8794,
+      "step": 1335
+    },
+    {
+      "epoch": 0.42896131000160537,
+      "grad_norm": 2.3192408084869385,
+      "learning_rate": 0.00019931706495384863,
+      "loss": 0.9037,
+      "step": 1336
+    },
+    {
+      "epoch": 0.4292823888264569,
+      "grad_norm": 1.2625378370285034,
+      "learning_rate": 0.00019931298632618356,
+      "loss": 0.9458,
+      "step": 1337
+    },
+    {
+      "epoch": 0.4296034676513084,
+      "grad_norm": 2.8485538959503174,
+      "learning_rate": 0.00019930889559753234,
+      "loss": 0.9646,
+      "step": 1338
+    },
+    {
+      "epoch": 0.4299245464761599,
+      "grad_norm": 1.6909540891647339,
+      "learning_rate": 0.00019930479276839344,
+      "loss": 1.0197,
+      "step": 1339
+    },
+    {
+      "epoch": 0.4302456253010114,
+      "grad_norm": 1.6980613470077515,
+      "learning_rate": 0.00019930067783926675,
+      "loss": 0.9211,
+      "step": 1340
+    },
+    {
+      "epoch": 0.4305667041258629,
+      "grad_norm": 0.8947736024856567,
+      "learning_rate": 0.0001992965508106537,
+      "loss": 0.7911,
+      "step": 1341
+    },
+    {
+      "epoch": 0.4308877829507144,
+      "grad_norm": 1.4755980968475342,
+      "learning_rate": 0.00019929241168305714,
+      "loss": 1.0548,
+      "step": 1342
+    },
+    {
+      "epoch": 0.4312088617755659,
+      "grad_norm": 2.1290855407714844,
+      "learning_rate": 0.00019928826045698136,
+      "loss": 0.9762,
+      "step": 1343
+    },
+    {
+      "epoch": 0.4315299406004174,
+      "grad_norm": 1.3842682838439941,
+      "learning_rate": 0.00019928409713293227,
+      "loss": 1.0557,
+      "step": 1344
+    },
+    {
+      "epoch": 0.4318510194252689,
+      "grad_norm": 2.651461601257324,
+      "learning_rate": 0.00019927992171141708,
+      "loss": 0.8827,
+      "step": 1345
+    },
+    {
+      "epoch": 0.43217209825012043,
+      "grad_norm": 3.1933929920196533,
+      "learning_rate": 0.00019927573419294456,
+      "loss": 0.9752,
+      "step": 1346
+    },
+    {
+      "epoch": 0.4324931770749719,
+      "grad_norm": 2.6903903484344482,
+      "learning_rate": 0.000199271534578025,
+      "loss": 1.0742,
+      "step": 1347
+    },
+    {
+      "epoch": 0.4328142558998234,
+      "grad_norm": 0.9604675769805908,
+      "learning_rate": 0.00019926732286717003,
+      "loss": 0.7134,
+      "step": 1348
+    },
+    {
+      "epoch": 0.4331353347246749,
+      "grad_norm": 1.6569525003433228,
+      "learning_rate": 0.0001992630990608929,
+      "loss": 0.6801,
+      "step": 1349
+    },
+    {
+      "epoch": 0.4334564135495264,
+      "grad_norm": 1.8039077520370483,
+      "learning_rate": 0.00019925886315970824,
+      "loss": 0.6657,
+      "step": 1350
+    },
+    {
+      "epoch": 0.4337774923743779,
+      "grad_norm": 2.7572011947631836,
+      "learning_rate": 0.00019925461516413223,
+      "loss": 1.5341,
+      "step": 1351
+    },
+    {
+      "epoch": 0.4340985711992294,
+      "grad_norm": 1.5353821516036987,
+      "learning_rate": 0.0001992503550746824,
+      "loss": 1.2077,
+      "step": 1352
+    },
+    {
+      "epoch": 0.4344196500240809,
+      "grad_norm": 1.8560153245925903,
+      "learning_rate": 0.00019924608289187786,
+      "loss": 1.043,
+      "step": 1353
+    },
+    {
+      "epoch": 0.4347407288489324,
+      "grad_norm": 1.4403096437454224,
+      "learning_rate": 0.00019924179861623915,
+      "loss": 0.7871,
+      "step": 1354
+    },
+    {
+      "epoch": 0.43506180767378394,
+      "grad_norm": 1.403806447982788,
+      "learning_rate": 0.00019923750224828832,
+      "loss": 0.7443,
+      "step": 1355
+    },
+    {
+      "epoch": 0.43538288649863544,
+      "grad_norm": 1.7703312635421753,
+      "learning_rate": 0.00019923319378854887,
+      "loss": 0.9553,
+      "step": 1356
+    },
+    {
+      "epoch": 0.43570396532348693,
+      "grad_norm": 1.7927194833755493,
+      "learning_rate": 0.00019922887323754577,
+      "loss": 0.6829,
+      "step": 1357
+    },
+    {
+      "epoch": 0.4360250441483384,
+      "grad_norm": 3.560272693634033,
+      "learning_rate": 0.00019922454059580544,
+      "loss": 1.0068,
+      "step": 1358
+    },
+    {
+      "epoch": 0.4363461229731899,
+      "grad_norm": 1.5825556516647339,
+      "learning_rate": 0.00019922019586385585,
+      "loss": 1.0626,
+      "step": 1359
+    },
+    {
+      "epoch": 0.4366672017980414,
+      "grad_norm": 1.7079756259918213,
+      "learning_rate": 0.00019921583904222633,
+      "loss": 0.9531,
+      "step": 1360
+    },
+    {
+      "epoch": 0.4369882806228929,
+      "grad_norm": 1.5126279592514038,
+      "learning_rate": 0.0001992114701314478,
+      "loss": 0.9648,
+      "step": 1361
+    },
+    {
+      "epoch": 0.4373093594477444,
+      "grad_norm": 1.6111955642700195,
+      "learning_rate": 0.00019920708913205256,
+      "loss": 1.111,
+      "step": 1362
+    },
+    {
+      "epoch": 0.4376304382725959,
+      "grad_norm": 1.1664777994155884,
+      "learning_rate": 0.00019920269604457446,
+      "loss": 1.0386,
+      "step": 1363
+    },
+    {
+      "epoch": 0.4379515170974474,
+      "grad_norm": 1.0824946165084839,
+      "learning_rate": 0.0001991982908695487,
+      "loss": 0.9912,
+      "step": 1364
+    },
+    {
+      "epoch": 0.43827259592229895,
+      "grad_norm": 1.131016492843628,
+      "learning_rate": 0.00019919387360751217,
+      "loss": 0.8238,
+      "step": 1365
+    },
+    {
+      "epoch": 0.43859367474715044,
+      "grad_norm": 1.307161808013916,
+      "learning_rate": 0.000199189444259003,
+      "loss": 0.8337,
+      "step": 1366
+    },
+    {
+      "epoch": 0.43891475357200194,
+      "grad_norm": 1.598332166671753,
+      "learning_rate": 0.0001991850028245609,
+      "loss": 1.1024,
+      "step": 1367
+    },
+    {
+      "epoch": 0.43923583239685343,
+      "grad_norm": 1.131783366203308,
+      "learning_rate": 0.00019918054930472706,
+      "loss": 0.9586,
+      "step": 1368
+    },
+    {
+      "epoch": 0.4395569112217049,
+      "grad_norm": 1.2683618068695068,
+      "learning_rate": 0.00019917608370004417,
+      "loss": 1.02,
+      "step": 1369
+    },
+    {
+      "epoch": 0.4398779900465564,
+      "grad_norm": 2.1559362411499023,
+      "learning_rate": 0.0001991716060110563,
+      "loss": 0.9757,
+      "step": 1370
+    },
+    {
+      "epoch": 0.4401990688714079,
+      "grad_norm": 1.6462043523788452,
+      "learning_rate": 0.00019916711623830903,
+      "loss": 1.0838,
+      "step": 1371
+    },
+    {
+      "epoch": 0.4405201476962594,
+      "grad_norm": 1.468853235244751,
+      "learning_rate": 0.0001991626143823495,
+      "loss": 1.1971,
+      "step": 1372
+    },
+    {
+      "epoch": 0.4408412265211109,
+      "grad_norm": 1.2045490741729736,
+      "learning_rate": 0.00019915810044372618,
+      "loss": 1.2012,
+      "step": 1373
+    },
+    {
+      "epoch": 0.44116230534596246,
+      "grad_norm": 1.699227213859558,
+      "learning_rate": 0.0001991535744229891,
+      "loss": 0.9933,
+      "step": 1374
+    },
+    {
+      "epoch": 0.44148338417081395,
+      "grad_norm": 1.5638002157211304,
+      "learning_rate": 0.00019914903632068973,
+      "loss": 1.2959,
+      "step": 1375
+    },
+    {
+      "epoch": 0.44180446299566545,
+      "grad_norm": 2.418407440185547,
+      "learning_rate": 0.00019914448613738106,
+      "loss": 1.1246,
+      "step": 1376
+    },
+    {
+      "epoch": 0.44212554182051694,
+      "grad_norm": 2.6279420852661133,
+      "learning_rate": 0.00019913992387361745,
+      "loss": 1.069,
+      "step": 1377
+    },
+    {
+      "epoch": 0.44244662064536844,
+      "grad_norm": 1.4538555145263672,
+      "learning_rate": 0.00019913534952995486,
+      "loss": 0.9135,
+      "step": 1378
+    },
+    {
+      "epoch": 0.44276769947021993,
+      "grad_norm": 1.7238460779190063,
+      "learning_rate": 0.00019913076310695068,
+      "loss": 1.3682,
+      "step": 1379
+    },
+    {
+      "epoch": 0.44308877829507143,
+      "grad_norm": 2.1236801147460938,
+      "learning_rate": 0.00019912616460516364,
+      "loss": 1.0323,
+      "step": 1380
+    },
+    {
+      "epoch": 0.4434098571199229,
+      "grad_norm": 1.527573585510254,
+      "learning_rate": 0.00019912155402515417,
+      "loss": 1.2224,
+      "step": 1381
+    },
+    {
+      "epoch": 0.4437309359447744,
+      "grad_norm": 1.1867547035217285,
+      "learning_rate": 0.00019911693136748403,
+      "loss": 0.9621,
+      "step": 1382
+    },
+    {
+      "epoch": 0.44405201476962597,
+      "grad_norm": 1.631277322769165,
+      "learning_rate": 0.0001991122966327164,
+      "loss": 1.1633,
+      "step": 1383
+    },
+    {
+      "epoch": 0.44437309359447746,
+      "grad_norm": 2.110377550125122,
+      "learning_rate": 0.0001991076498214161,
+      "loss": 1.081,
+      "step": 1384
+    },
+    {
+      "epoch": 0.44469417241932896,
+      "grad_norm": 1.2821978330612183,
+      "learning_rate": 0.0001991029909341493,
+      "loss": 0.8347,
+      "step": 1385
+    },
+    {
+      "epoch": 0.44501525124418045,
+      "grad_norm": 1.3833321332931519,
+      "learning_rate": 0.00019909831997148362,
+      "loss": 1.0652,
+      "step": 1386
+    },
+    {
+      "epoch": 0.44533633006903195,
+      "grad_norm": 1.2607406377792358,
+      "learning_rate": 0.00019909363693398828,
+      "loss": 0.9888,
+      "step": 1387
+    },
+    {
+      "epoch": 0.44565740889388344,
+      "grad_norm": 1.2479708194732666,
+      "learning_rate": 0.00019908894182223388,
+      "loss": 0.9652,
+      "step": 1388
+    },
+    {
+      "epoch": 0.44597848771873494,
+      "grad_norm": 1.9284359216690063,
+      "learning_rate": 0.00019908423463679248,
+      "loss": 1.0546,
+      "step": 1389
+    },
+    {
+      "epoch": 0.44629956654358643,
+      "grad_norm": 1.1474945545196533,
+      "learning_rate": 0.0001990795153782376,
+      "loss": 0.8867,
+      "step": 1390
+    },
+    {
+      "epoch": 0.44662064536843793,
+      "grad_norm": 1.8049343824386597,
+      "learning_rate": 0.00019907478404714436,
+      "loss": 0.921,
+      "step": 1391
+    },
+    {
+      "epoch": 0.4469417241932895,
+      "grad_norm": 2.072420835494995,
+      "learning_rate": 0.0001990700406440892,
+      "loss": 1.0613,
+      "step": 1392
+    },
+    {
+      "epoch": 0.447262803018141,
+      "grad_norm": 1.3974262475967407,
+      "learning_rate": 0.00019906528516965008,
+      "loss": 0.9483,
+      "step": 1393
+    },
+    {
+      "epoch": 0.44758388184299247,
+      "grad_norm": 1.0429906845092773,
+      "learning_rate": 0.0001990605176244065,
+      "loss": 0.8286,
+      "step": 1394
+    },
+    {
+      "epoch": 0.44790496066784397,
+      "grad_norm": 1.2362806797027588,
+      "learning_rate": 0.0001990557380089393,
+      "loss": 1.0296,
+      "step": 1395
+    },
+    {
+      "epoch": 0.44822603949269546,
+      "grad_norm": 1.1999589204788208,
+      "learning_rate": 0.0001990509463238309,
+      "loss": 0.7282,
+      "step": 1396
+    },
+    {
+      "epoch": 0.44854711831754696,
+      "grad_norm": 1.3461198806762695,
+      "learning_rate": 0.00019904614256966512,
+      "loss": 0.9301,
+      "step": 1397
+    },
+    {
+      "epoch": 0.44886819714239845,
+      "grad_norm": 2.5032989978790283,
+      "learning_rate": 0.00019904132674702734,
+      "loss": 0.5983,
+      "step": 1398
+    },
+    {
+      "epoch": 0.44918927596724995,
+      "grad_norm": 1.029445767402649,
+      "learning_rate": 0.0001990364988565043,
+      "loss": 0.7837,
+      "step": 1399
+    },
+    {
+      "epoch": 0.44951035479210144,
+      "grad_norm": 0.8912380337715149,
+      "learning_rate": 0.0001990316588986843,
+      "loss": 0.6475,
+      "step": 1400
+    },
+    {
+      "epoch": 0.44983143361695294,
+      "grad_norm": 4.18993616104126,
+      "learning_rate": 0.00019902680687415705,
+      "loss": 1.7317,
+      "step": 1401
+    },
+    {
+      "epoch": 0.4501525124418045,
+      "grad_norm": 6.351428031921387,
+      "learning_rate": 0.00019902194278351374,
+      "loss": 1.6514,
+      "step": 1402
+    },
+    {
+      "epoch": 0.450473591266656,
+      "grad_norm": 2.5410821437835693,
+      "learning_rate": 0.00019901706662734712,
+      "loss": 1.1215,
+      "step": 1403
+    },
+    {
+      "epoch": 0.4507946700915075,
+      "grad_norm": 1.9113538265228271,
+      "learning_rate": 0.0001990121784062512,
+      "loss": 0.9147,
+      "step": 1404
+    },
+    {
+      "epoch": 0.45111574891635897,
+      "grad_norm": 3.196178913116455,
+      "learning_rate": 0.00019900727812082177,
+      "loss": 0.9308,
+      "step": 1405
+    },
+    {
+      "epoch": 0.45143682774121047,
+      "grad_norm": 1.5448819398880005,
+      "learning_rate": 0.00019900236577165576,
+      "loss": 0.8038,
+      "step": 1406
+    },
+    {
+      "epoch": 0.45175790656606196,
+      "grad_norm": 1.4571887254714966,
+      "learning_rate": 0.0001989974413593518,
+      "loss": 0.7484,
+      "step": 1407
+    },
+    {
+      "epoch": 0.45207898539091346,
+      "grad_norm": 1.5297858715057373,
+      "learning_rate": 0.0001989925048845099,
+      "loss": 0.854,
+      "step": 1408
+    },
+    {
+      "epoch": 0.45240006421576495,
+      "grad_norm": 1.8126001358032227,
+      "learning_rate": 0.00019898755634773158,
+      "loss": 0.9785,
+      "step": 1409
+    },
+    {
+      "epoch": 0.45272114304061645,
+      "grad_norm": 1.3122637271881104,
+      "learning_rate": 0.00019898259574961978,
+      "loss": 1.1906,
+      "step": 1410
+    },
+    {
+      "epoch": 0.453042221865468,
+      "grad_norm": 1.2841436862945557,
+      "learning_rate": 0.0001989776230907789,
+      "loss": 0.7769,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4533633006903195,
+      "grad_norm": 1.6594353914260864,
+      "learning_rate": 0.00019897263837181491,
+      "loss": 0.8995,
+      "step": 1412
+    },
+    {
+      "epoch": 0.453684379515171,
+      "grad_norm": 1.2056258916854858,
+      "learning_rate": 0.0001989676415933351,
+      "loss": 0.9341,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4540054583400225,
+      "grad_norm": 1.2220386266708374,
+      "learning_rate": 0.00019896263275594842,
+      "loss": 0.9202,
+      "step": 1414
+    },
+    {
+      "epoch": 0.454326537164874,
+      "grad_norm": 1.083976149559021,
+      "learning_rate": 0.0001989576118602651,
+      "loss": 0.8357,
+      "step": 1415
+    },
+    {
+      "epoch": 0.4546476159897255,
+      "grad_norm": 1.134037733078003,
+      "learning_rate": 0.00019895257890689696,
+      "loss": 1.0837,
+      "step": 1416
+    },
+    {
+      "epoch": 0.45496869481457697,
+      "grad_norm": 1.6373850107192993,
+      "learning_rate": 0.00019894753389645723,
+      "loss": 1.0875,
+      "step": 1417
+    },
+    {
+      "epoch": 0.45528977363942846,
+      "grad_norm": 1.8361527919769287,
+      "learning_rate": 0.0001989424768295606,
+      "loss": 0.9749,
+      "step": 1418
+    },
+    {
+      "epoch": 0.45561085246427996,
+      "grad_norm": 2.0697522163391113,
+      "learning_rate": 0.00019893740770682335,
+      "loss": 0.9262,
+      "step": 1419
+    },
+    {
+      "epoch": 0.4559319312891315,
+      "grad_norm": 1.6787400245666504,
+      "learning_rate": 0.00019893232652886306,
+      "loss": 1.028,
+      "step": 1420
+    },
+    {
+      "epoch": 0.456253010113983,
+      "grad_norm": 1.6536740064620972,
+      "learning_rate": 0.00019892723329629887,
+      "loss": 1.138,
+      "step": 1421
+    },
+    {
+      "epoch": 0.4565740889388345,
+      "grad_norm": 2.046743631362915,
+      "learning_rate": 0.00019892212800975135,
+      "loss": 0.9917,
+      "step": 1422
+    },
+    {
+      "epoch": 0.456895167763686,
+      "grad_norm": 1.4609099626541138,
+      "learning_rate": 0.00019891701066984262,
+      "loss": 0.9998,
+      "step": 1423
+    },
+    {
+      "epoch": 0.4572162465885375,
+      "grad_norm": 1.2826584577560425,
+      "learning_rate": 0.00019891188127719618,
+      "loss": 1.0006,
+      "step": 1424
+    },
+    {
+      "epoch": 0.457537325413389,
+      "grad_norm": 1.3913053274154663,
+      "learning_rate": 0.00019890673983243706,
+      "loss": 1.098,
+      "step": 1425
+    },
+    {
+      "epoch": 0.4578584042382405,
+      "grad_norm": 2.0255448818206787,
+      "learning_rate": 0.0001989015863361917,
+      "loss": 1.0888,
+      "step": 1426
+    },
+    {
+      "epoch": 0.458179483063092,
+      "grad_norm": 2.3133769035339355,
+      "learning_rate": 0.00019889642078908804,
+      "loss": 1.2081,
+      "step": 1427
+    },
+    {
+      "epoch": 0.45850056188794347,
+      "grad_norm": 1.3124561309814453,
+      "learning_rate": 0.00019889124319175547,
+      "loss": 1.125,
+      "step": 1428
+    },
+    {
+      "epoch": 0.45882164071279496,
+      "grad_norm": 1.5743858814239502,
+      "learning_rate": 0.0001988860535448249,
+      "loss": 1.0983,
+      "step": 1429
+    },
+    {
+      "epoch": 0.4591427195376465,
+      "grad_norm": 1.6619899272918701,
+      "learning_rate": 0.00019888085184892868,
+      "loss": 1.1139,
+      "step": 1430
+    },
+    {
+      "epoch": 0.459463798362498,
+      "grad_norm": 1.2634574174880981,
+      "learning_rate": 0.0001988756381047006,
+      "loss": 1.1412,
+      "step": 1431
+    },
+    {
+      "epoch": 0.4597848771873495,
+      "grad_norm": 1.1055835485458374,
+      "learning_rate": 0.00019887041231277593,
+      "loss": 0.7893,
+      "step": 1432
+    },
+    {
+      "epoch": 0.460105956012201,
+      "grad_norm": 1.8112157583236694,
+      "learning_rate": 0.0001988651744737914,
+      "loss": 0.8947,
+      "step": 1433
+    },
+    {
+      "epoch": 0.4604270348370525,
+      "grad_norm": 0.9498884081840515,
+      "learning_rate": 0.00019885992458838528,
+      "loss": 0.9129,
+      "step": 1434
+    },
+    {
+      "epoch": 0.460748113661904,
+      "grad_norm": 1.6810704469680786,
+      "learning_rate": 0.0001988546626571972,
+      "loss": 1.0324,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4610691924867555,
+      "grad_norm": 1.1200361251831055,
+      "learning_rate": 0.00019884938868086835,
+      "loss": 1.1815,
+      "step": 1436
+    },
+    {
+      "epoch": 0.461390271311607,
+      "grad_norm": 1.532668113708496,
+      "learning_rate": 0.00019884410266004135,
+      "loss": 0.9087,
+      "step": 1437
+    },
+    {
+      "epoch": 0.4617113501364585,
+      "grad_norm": 3.874738931655884,
+      "learning_rate": 0.00019883880459536024,
+      "loss": 0.9593,
+      "step": 1438
+    },
+    {
+      "epoch": 0.46203242896131,
+      "grad_norm": 2.262633800506592,
+      "learning_rate": 0.00019883349448747062,
+      "loss": 0.9921,
+      "step": 1439
+    },
+    {
+      "epoch": 0.4623535077861615,
+      "grad_norm": 1.0738508701324463,
+      "learning_rate": 0.00019882817233701948,
+      "loss": 0.7347,
+      "step": 1440
+    },
+    {
+      "epoch": 0.462674586611013,
+      "grad_norm": 1.7245491743087769,
+      "learning_rate": 0.0001988228381446553,
+      "loss": 0.9579,
+      "step": 1441
+    },
+    {
+      "epoch": 0.4629956654358645,
+      "grad_norm": 1.3573378324508667,
+      "learning_rate": 0.00019881749191102808,
+      "loss": 0.9096,
+      "step": 1442
+    },
+    {
+      "epoch": 0.463316744260716,
+      "grad_norm": 1.2465156316757202,
+      "learning_rate": 0.0001988121336367892,
+      "loss": 0.9165,
+      "step": 1443
+    },
+    {
+      "epoch": 0.4636378230855675,
+      "grad_norm": 1.3146942853927612,
+      "learning_rate": 0.00019880676332259154,
+      "loss": 0.7766,
+      "step": 1444
+    },
+    {
+      "epoch": 0.463958901910419,
+      "grad_norm": 1.2299082279205322,
+      "learning_rate": 0.00019880138096908952,
+      "loss": 0.7833,
+      "step": 1445
+    },
+    {
+      "epoch": 0.4642799807352705,
+      "grad_norm": 1.622185230255127,
+      "learning_rate": 0.00019879598657693891,
+      "loss": 0.7363,
+      "step": 1446
+    },
+    {
+      "epoch": 0.464601059560122,
+      "grad_norm": 1.7409615516662598,
+      "learning_rate": 0.00019879058014679704,
+      "loss": 1.0281,
+      "step": 1447
+    },
+    {
+      "epoch": 0.46492213838497354,
+      "grad_norm": 1.8660942316055298,
+      "learning_rate": 0.00019878516167932261,
+      "loss": 0.8218,
+      "step": 1448
+    },
+    {
+      "epoch": 0.46524321720982503,
+      "grad_norm": 1.445410132408142,
+      "learning_rate": 0.0001987797311751759,
+      "loss": 0.7792,
+      "step": 1449
+    },
+    {
+      "epoch": 0.46556429603467653,
+      "grad_norm": 1.0770496129989624,
+      "learning_rate": 0.00019877428863501856,
+      "loss": 0.6364,
+      "step": 1450
+    },
+    {
+      "epoch": 0.465885374859528,
+      "grad_norm": 1.7615505456924438,
+      "learning_rate": 0.00019876883405951377,
+      "loss": 1.367,
+      "step": 1451
+    },
+    {
+      "epoch": 0.4662064536843795,
+      "grad_norm": 1.661313772201538,
+      "learning_rate": 0.00019876336744932614,
+      "loss": 1.1688,
+      "step": 1452
+    },
+    {
+      "epoch": 0.466527532509231,
+      "grad_norm": 1.480622410774231,
+      "learning_rate": 0.0001987578888051218,
+      "loss": 1.0435,
+      "step": 1453
+    },
+    {
+      "epoch": 0.4668486113340825,
+      "grad_norm": 1.6343483924865723,
+      "learning_rate": 0.00019875239812756825,
+      "loss": 0.9328,
+      "step": 1454
+    },
+    {
+      "epoch": 0.467169690158934,
+      "grad_norm": 2.5799641609191895,
+      "learning_rate": 0.00019874689541733457,
+      "loss": 0.8994,
+      "step": 1455
+    },
+    {
+      "epoch": 0.4674907689837855,
+      "grad_norm": 1.4654110670089722,
+      "learning_rate": 0.00019874138067509117,
+      "loss": 0.7419,
+      "step": 1456
+    },
+    {
+      "epoch": 0.467811847808637,
+      "grad_norm": 1.7656830549240112,
+      "learning_rate": 0.00019873585390151003,
+      "loss": 1.0638,
+      "step": 1457
+    },
+    {
+      "epoch": 0.46813292663348854,
+      "grad_norm": 1.733784794807434,
+      "learning_rate": 0.00019873031509726462,
+      "loss": 1.1878,
+      "step": 1458
+    },
+    {
+      "epoch": 0.46845400545834004,
+      "grad_norm": 2.345583915710449,
+      "learning_rate": 0.00019872476426302982,
+      "loss": 0.98,
+      "step": 1459
+    },
+    {
+      "epoch": 0.46877508428319153,
+      "grad_norm": 1.3415275812149048,
+      "learning_rate": 0.00019871920139948192,
+      "loss": 0.9307,
+      "step": 1460
+    },
+    {
+      "epoch": 0.46909616310804303,
+      "grad_norm": 1.6265960931777954,
+      "learning_rate": 0.0001987136265072988,
+      "loss": 1.0417,
+      "step": 1461
+    },
+    {
+      "epoch": 0.4694172419328945,
+      "grad_norm": 1.1737104654312134,
+      "learning_rate": 0.00019870803958715972,
+      "loss": 0.8448,
+      "step": 1462
+    },
+    {
+      "epoch": 0.469738320757746,
+      "grad_norm": 1.189310908317566,
+      "learning_rate": 0.0001987024406397454,
+      "loss": 1.006,
+      "step": 1463
+    },
+    {
+      "epoch": 0.4700593995825975,
+      "grad_norm": 1.5854594707489014,
+      "learning_rate": 0.00019869682966573813,
+      "loss": 0.9803,
+      "step": 1464
+    },
+    {
+      "epoch": 0.470380478407449,
+      "grad_norm": 1.1072514057159424,
+      "learning_rate": 0.00019869120666582153,
+      "loss": 1.0124,
+      "step": 1465
+    },
+    {
+      "epoch": 0.4707015572323005,
+      "grad_norm": 1.613070011138916,
+      "learning_rate": 0.00019868557164068074,
+      "loss": 1.2908,
+      "step": 1466
+    },
+    {
+      "epoch": 0.47102263605715206,
+      "grad_norm": 1.2394685745239258,
+      "learning_rate": 0.0001986799245910024,
+      "loss": 1.1003,
+      "step": 1467
+    },
+    {
+      "epoch": 0.47134371488200355,
+      "grad_norm": 1.047046422958374,
+      "learning_rate": 0.00019867426551747457,
+      "loss": 1.0153,
+      "step": 1468
+    },
+    {
+      "epoch": 0.47166479370685505,
+      "grad_norm": 1.3292659521102905,
+      "learning_rate": 0.0001986685944207868,
+      "loss": 0.9651,
+      "step": 1469
+    },
+    {
+      "epoch": 0.47198587253170654,
+      "grad_norm": 1.4986034631729126,
+      "learning_rate": 0.0001986629113016301,
+      "loss": 0.9747,
+      "step": 1470
+    },
+    {
+      "epoch": 0.47230695135655804,
+      "grad_norm": 2.4843838214874268,
+      "learning_rate": 0.00019865721616069696,
+      "loss": 1.0655,
+      "step": 1471
+    },
+    {
+      "epoch": 0.47262803018140953,
+      "grad_norm": 1.3794037103652954,
+      "learning_rate": 0.00019865150899868125,
+      "loss": 1.1716,
+      "step": 1472
+    },
+    {
+      "epoch": 0.472949109006261,
+      "grad_norm": 1.1681032180786133,
+      "learning_rate": 0.00019864578981627844,
+      "loss": 1.1593,
+      "step": 1473
+    },
+    {
+      "epoch": 0.4732701878311125,
+      "grad_norm": 1.135563850402832,
+      "learning_rate": 0.00019864005861418535,
+      "loss": 0.9522,
+      "step": 1474
+    },
+    {
+      "epoch": 0.473591266655964,
+      "grad_norm": 1.3052153587341309,
+      "learning_rate": 0.0001986343153931003,
+      "loss": 1.0636,
+      "step": 1475
+    },
+    {
+      "epoch": 0.47391234548081557,
+      "grad_norm": 2.6790149211883545,
+      "learning_rate": 0.00019862856015372317,
+      "loss": 1.0276,
+      "step": 1476
+    },
+    {
+      "epoch": 0.47423342430566706,
+      "grad_norm": 1.5842307806015015,
+      "learning_rate": 0.00019862279289675509,
+      "loss": 0.9003,
+      "step": 1477
+    },
+    {
+      "epoch": 0.47455450313051856,
+      "grad_norm": 1.1842899322509766,
+      "learning_rate": 0.0001986170136228989,
+      "loss": 0.8584,
+      "step": 1478
+    },
+    {
+      "epoch": 0.47487558195537005,
+      "grad_norm": 1.4418596029281616,
+      "learning_rate": 0.0001986112223328587,
+      "loss": 0.9339,
+      "step": 1479
+    },
+    {
+      "epoch": 0.47519666078022155,
+      "grad_norm": 1.6683169603347778,
+      "learning_rate": 0.00019860541902734022,
+      "loss": 0.9629,
+      "step": 1480
+    },
+    {
+      "epoch": 0.47551773960507304,
+      "grad_norm": 1.3546855449676514,
+      "learning_rate": 0.0001985996037070505,
+      "loss": 1.0198,
+      "step": 1481
+    },
+    {
+      "epoch": 0.47583881842992454,
+      "grad_norm": 1.2630248069763184,
+      "learning_rate": 0.00019859377637269815,
+      "loss": 0.9135,
+      "step": 1482
+    },
+    {
+      "epoch": 0.47615989725477603,
+      "grad_norm": 1.2437412738800049,
+      "learning_rate": 0.00019858793702499323,
+      "loss": 0.8743,
+      "step": 1483
+    },
+    {
+      "epoch": 0.4764809760796275,
+      "grad_norm": 2.26094388961792,
+      "learning_rate": 0.00019858208566464724,
+      "loss": 1.038,
+      "step": 1484
+    },
+    {
+      "epoch": 0.476802054904479,
+      "grad_norm": 1.2740654945373535,
+      "learning_rate": 0.00019857622229237313,
+      "loss": 1.0045,
+      "step": 1485
+    },
+    {
+      "epoch": 0.4771231337293306,
+      "grad_norm": 1.8857311010360718,
+      "learning_rate": 0.00019857034690888537,
+      "loss": 0.9932,
+      "step": 1486
+    },
+    {
+      "epoch": 0.47744421255418207,
+      "grad_norm": 1.1165155172348022,
+      "learning_rate": 0.00019856445951489982,
+      "loss": 0.8839,
+      "step": 1487
+    },
+    {
+      "epoch": 0.47776529137903356,
+      "grad_norm": 1.5677917003631592,
+      "learning_rate": 0.00019855856011113384,
+      "loss": 0.8075,
+      "step": 1488
+    },
+    {
+      "epoch": 0.47808637020388506,
+      "grad_norm": 1.1825371980667114,
+      "learning_rate": 0.00019855264869830629,
+      "loss": 0.875,
+      "step": 1489
+    },
+    {
+      "epoch": 0.47840744902873655,
+      "grad_norm": 1.572262167930603,
+      "learning_rate": 0.00019854672527713744,
+      "loss": 1.1585,
+      "step": 1490
+    },
+    {
+      "epoch": 0.47872852785358805,
+      "grad_norm": 2.2230706214904785,
+      "learning_rate": 0.00019854078984834903,
+      "loss": 0.9478,
+      "step": 1491
+    },
+    {
+      "epoch": 0.47904960667843954,
+      "grad_norm": 1.5785419940948486,
+      "learning_rate": 0.00019853484241266428,
+      "loss": 0.8647,
+      "step": 1492
+    },
+    {
+      "epoch": 0.47937068550329104,
+      "grad_norm": 0.8901344537734985,
+      "learning_rate": 0.00019852888297080786,
+      "loss": 0.7951,
+      "step": 1493
+    },
+    {
+      "epoch": 0.47969176432814253,
+      "grad_norm": 1.1752713918685913,
+      "learning_rate": 0.00019852291152350592,
+      "loss": 0.9361,
+      "step": 1494
+    },
+    {
+      "epoch": 0.4800128431529941,
+      "grad_norm": 1.196556568145752,
+      "learning_rate": 0.0001985169280714861,
+      "loss": 0.9662,
+      "step": 1495
+    },
+    {
+      "epoch": 0.4803339219778456,
+      "grad_norm": 1.6458885669708252,
+      "learning_rate": 0.0001985109326154774,
+      "loss": 0.932,
+      "step": 1496
+    },
+    {
+      "epoch": 0.4806550008026971,
+      "grad_norm": 1.7147083282470703,
+      "learning_rate": 0.00019850492515621038,
+      "loss": 0.8323,
+      "step": 1497
+    },
+    {
+      "epoch": 0.48097607962754857,
+      "grad_norm": 1.5965648889541626,
+      "learning_rate": 0.00019849890569441703,
+      "loss": 0.6222,
+      "step": 1498
+    },
+    {
+      "epoch": 0.48129715845240006,
+      "grad_norm": 0.8603938221931458,
+      "learning_rate": 0.00019849287423083078,
+      "loss": 0.4993,
+      "step": 1499
+    },
+    {
+      "epoch": 0.48161823727725156,
+      "grad_norm": 1.216637134552002,
+      "learning_rate": 0.00019848683076618658,
+      "loss": 0.606,
+      "step": 1500
+    },
+    {
+      "epoch": 0.48193931610210305,
+      "grad_norm": 1.512508511543274,
+      "learning_rate": 0.00019848077530122083,
+      "loss": 1.5987,
+      "step": 1501
+    },
+    {
+      "epoch": 0.48226039492695455,
+      "grad_norm": 1.3921582698822021,
+      "learning_rate": 0.00019847470783667127,
+      "loss": 1.3467,
+      "step": 1502
+    },
+    {
+      "epoch": 0.48258147375180604,
+      "grad_norm": 1.5864951610565186,
+      "learning_rate": 0.0001984686283732773,
+      "loss": 1.3512,
+      "step": 1503
+    },
+    {
+      "epoch": 0.4829025525766576,
+      "grad_norm": 1.5356676578521729,
+      "learning_rate": 0.00019846253691177966,
+      "loss": 0.8669,
+      "step": 1504
+    },
+    {
+      "epoch": 0.4832236314015091,
+      "grad_norm": 1.3852877616882324,
+      "learning_rate": 0.00019845643345292054,
+      "loss": 0.677,
+      "step": 1505
+    },
+    {
+      "epoch": 0.4835447102263606,
+      "grad_norm": 1.4612033367156982,
+      "learning_rate": 0.00019845031799744367,
+      "loss": 0.8081,
+      "step": 1506
+    },
+    {
+      "epoch": 0.4838657890512121,
+      "grad_norm": 1.5875556468963623,
+      "learning_rate": 0.0001984441905460942,
+      "loss": 0.783,
+      "step": 1507
+    },
+    {
+      "epoch": 0.4841868678760636,
+      "grad_norm": 2.032498359680176,
+      "learning_rate": 0.00019843805109961868,
+      "loss": 1.2169,
+      "step": 1508
+    },
+    {
+      "epoch": 0.48450794670091507,
+      "grad_norm": 1.468684196472168,
+      "learning_rate": 0.00019843189965876526,
+      "loss": 1.1707,
+      "step": 1509
+    },
+    {
+      "epoch": 0.48482902552576657,
+      "grad_norm": 1.416133165359497,
+      "learning_rate": 0.00019842573622428345,
+      "loss": 1.0083,
+      "step": 1510
+    },
+    {
+      "epoch": 0.48515010435061806,
+      "grad_norm": 1.3447717428207397,
+      "learning_rate": 0.0001984195607969242,
+      "loss": 1.158,
+      "step": 1511
+    },
+    {
+      "epoch": 0.48547118317546956,
+      "grad_norm": 0.9209278225898743,
+      "learning_rate": 0.00019841337337744004,
+      "loss": 0.8482,
+      "step": 1512
+    },
+    {
+      "epoch": 0.48579226200032105,
+      "grad_norm": 1.1930347681045532,
+      "learning_rate": 0.00019840717396658484,
+      "loss": 0.9346,
+      "step": 1513
+    },
+    {
+      "epoch": 0.4861133408251726,
+      "grad_norm": 2.087160587310791,
+      "learning_rate": 0.00019840096256511398,
+      "loss": 1.034,
+      "step": 1514
+    },
+    {
+      "epoch": 0.4864344196500241,
+      "grad_norm": 2.3745105266571045,
+      "learning_rate": 0.00019839473917378434,
+      "loss": 0.9559,
+      "step": 1515
+    },
+    {
+      "epoch": 0.4867554984748756,
+      "grad_norm": 1.2410439252853394,
+      "learning_rate": 0.00019838850379335417,
+      "loss": 1.1425,
+      "step": 1516
+    },
+    {
+      "epoch": 0.4870765772997271,
+      "grad_norm": 1.2628313302993774,
+      "learning_rate": 0.00019838225642458327,
+      "loss": 0.8775,
+      "step": 1517
+    },
+    {
+      "epoch": 0.4873976561245786,
+      "grad_norm": 1.2423824071884155,
+      "learning_rate": 0.00019837599706823284,
+      "loss": 1.1197,
+      "step": 1518
+    },
+    {
+      "epoch": 0.4877187349494301,
+      "grad_norm": 1.6348954439163208,
+      "learning_rate": 0.00019836972572506557,
+      "loss": 1.1318,
+      "step": 1519
+    },
+    {
+      "epoch": 0.4880398137742816,
+      "grad_norm": 1.3113266229629517,
+      "learning_rate": 0.00019836344239584564,
+      "loss": 0.8346,
+      "step": 1520
+    },
+    {
+      "epoch": 0.48836089259913307,
+      "grad_norm": 2.537024974822998,
+      "learning_rate": 0.00019835714708133862,
+      "loss": 0.9601,
+      "step": 1521
+    },
+    {
+      "epoch": 0.48868197142398456,
+      "grad_norm": 1.3241581916809082,
+      "learning_rate": 0.00019835083978231156,
+      "loss": 1.0883,
+      "step": 1522
+    },
+    {
+      "epoch": 0.4890030502488361,
+      "grad_norm": 1.2066631317138672,
+      "learning_rate": 0.00019834452049953297,
+      "loss": 0.8249,
+      "step": 1523
+    },
+    {
+      "epoch": 0.4893241290736876,
+      "grad_norm": 2.023279905319214,
+      "learning_rate": 0.0001983381892337729,
+      "loss": 1.1508,
+      "step": 1524
+    },
+    {
+      "epoch": 0.4896452078985391,
+      "grad_norm": 1.0849757194519043,
+      "learning_rate": 0.00019833184598580276,
+      "loss": 0.8601,
+      "step": 1525
+    },
+    {
+      "epoch": 0.4899662867233906,
+      "grad_norm": 1.6499979496002197,
+      "learning_rate": 0.0001983254907563955,
+      "loss": 1.0404,
+      "step": 1526
+    },
+    {
+      "epoch": 0.4902873655482421,
+      "grad_norm": 2.0686190128326416,
+      "learning_rate": 0.00019831912354632535,
+      "loss": 1.0121,
+      "step": 1527
+    },
+    {
+      "epoch": 0.4906084443730936,
+      "grad_norm": 1.2880290746688843,
+      "learning_rate": 0.0001983127443563683,
+      "loss": 0.9684,
+      "step": 1528
+    },
+    {
+      "epoch": 0.4909295231979451,
+      "grad_norm": 1.8285752534866333,
+      "learning_rate": 0.00019830635318730154,
+      "loss": 1.1573,
+      "step": 1529
+    },
+    {
+      "epoch": 0.4912506020227966,
+      "grad_norm": 1.0872652530670166,
+      "learning_rate": 0.00019829995003990388,
+      "loss": 0.9672,
+      "step": 1530
+    },
+    {
+      "epoch": 0.4915716808476481,
+      "grad_norm": 3.2847740650177,
+      "learning_rate": 0.00019829353491495545,
+      "loss": 1.0548,
+      "step": 1531
+    },
+    {
+      "epoch": 0.4918927596724996,
+      "grad_norm": 1.274139165878296,
+      "learning_rate": 0.00019828710781323792,
+      "loss": 1.1498,
+      "step": 1532
+    },
+    {
+      "epoch": 0.4922138384973511,
+      "grad_norm": 1.1564581394195557,
+      "learning_rate": 0.00019828066873553448,
+      "loss": 1.0376,
+      "step": 1533
+    },
+    {
+      "epoch": 0.4925349173222026,
+      "grad_norm": 2.3864920139312744,
+      "learning_rate": 0.00019827421768262967,
+      "loss": 1.2347,
+      "step": 1534
+    },
+    {
+      "epoch": 0.4928559961470541,
+      "grad_norm": 1.5429226160049438,
+      "learning_rate": 0.0001982677546553095,
+      "loss": 0.7993,
+      "step": 1535
+    },
+    {
+      "epoch": 0.4931770749719056,
+      "grad_norm": 1.7663393020629883,
+      "learning_rate": 0.00019826127965436152,
+      "loss": 1.1139,
+      "step": 1536
+    },
+    {
+      "epoch": 0.4934981537967571,
+      "grad_norm": 1.8667831420898438,
+      "learning_rate": 0.00019825479268057467,
+      "loss": 1.1442,
+      "step": 1537
+    },
+    {
+      "epoch": 0.4938192326216086,
+      "grad_norm": 1.3998355865478516,
+      "learning_rate": 0.0001982482937347394,
+      "loss": 0.9002,
+      "step": 1538
+    },
+    {
+      "epoch": 0.4941403114464601,
+      "grad_norm": 1.7938013076782227,
+      "learning_rate": 0.00019824178281764753,
+      "loss": 1.039,
+      "step": 1539
+    },
+    {
+      "epoch": 0.4944613902713116,
+      "grad_norm": 2.098247528076172,
+      "learning_rate": 0.00019823525993009243,
+      "loss": 0.729,
+      "step": 1540
+    },
+    {
+      "epoch": 0.49478246909616314,
+      "grad_norm": 1.4042150974273682,
+      "learning_rate": 0.0001982287250728689,
+      "loss": 1.1048,
+      "step": 1541
+    },
+    {
+      "epoch": 0.49510354792101463,
+      "grad_norm": 1.135709285736084,
+      "learning_rate": 0.00019822217824677315,
+      "loss": 0.8585,
+      "step": 1542
+    },
+    {
+      "epoch": 0.4954246267458661,
+      "grad_norm": 2.2327728271484375,
+      "learning_rate": 0.0001982156194526029,
+      "loss": 1.0654,
+      "step": 1543
+    },
+    {
+      "epoch": 0.4957457055707176,
+      "grad_norm": 1.6728448867797852,
+      "learning_rate": 0.0001982090486911574,
+      "loss": 0.8941,
+      "step": 1544
+    },
+    {
+      "epoch": 0.4960667843955691,
+      "grad_norm": 1.843428373336792,
+      "learning_rate": 0.0001982024659632372,
+      "loss": 1.0319,
+      "step": 1545
+    },
+    {
+      "epoch": 0.4963878632204206,
+      "grad_norm": 1.780278205871582,
+      "learning_rate": 0.00019819587126964437,
+      "loss": 0.9306,
+      "step": 1546
+    },
+    {
+      "epoch": 0.4967089420452721,
+      "grad_norm": 1.5725250244140625,
+      "learning_rate": 0.00019818926461118253,
+      "loss": 0.8823,
+      "step": 1547
+    },
+    {
+      "epoch": 0.4970300208701236,
+      "grad_norm": 2.787895679473877,
+      "learning_rate": 0.0001981826459886566,
+      "loss": 1.0951,
+      "step": 1548
+    },
+    {
+      "epoch": 0.4973510996949751,
+      "grad_norm": 0.9631800651550293,
+      "learning_rate": 0.00019817601540287306,
+      "loss": 0.637,
+      "step": 1549
+    },
+    {
+      "epoch": 0.4976721785198266,
+      "grad_norm": 1.3815464973449707,
+      "learning_rate": 0.0001981693728546399,
+      "loss": 0.7441,
+      "step": 1550
+    },
+    {
+      "epoch": 0.49799325734467814,
+      "grad_norm": 1.8321810960769653,
+      "learning_rate": 0.00019816271834476642,
+      "loss": 1.4642,
+      "step": 1551
+    },
+    {
+      "epoch": 0.49831433616952964,
+      "grad_norm": 1.4489206075668335,
+      "learning_rate": 0.00019815605187406345,
+      "loss": 1.2029,
+      "step": 1552
+    },
+    {
+      "epoch": 0.49863541499438113,
+      "grad_norm": 1.4845582246780396,
+      "learning_rate": 0.0001981493734433433,
+      "loss": 0.823,
+      "step": 1553
+    },
+    {
+      "epoch": 0.4989564938192326,
+      "grad_norm": 1.642909288406372,
+      "learning_rate": 0.0001981426830534197,
+      "loss": 0.8766,
+      "step": 1554
+    },
+    {
+      "epoch": 0.4992775726440841,
+      "grad_norm": 1.8939411640167236,
+      "learning_rate": 0.00019813598070510792,
+      "loss": 0.9463,
+      "step": 1555
+    },
+    {
+      "epoch": 0.4995986514689356,
+      "grad_norm": 1.7453148365020752,
+      "learning_rate": 0.0001981292663992245,
+      "loss": 0.8896,
+      "step": 1556
+    },
+    {
+      "epoch": 0.4999197302937871,
+      "grad_norm": 1.5144426822662354,
+      "learning_rate": 0.00019812254013658768,
+      "loss": 0.8909,
+      "step": 1557
+    },
+    {
+      "epoch": 0.5002408091186387,
+      "grad_norm": 1.3950881958007812,
+      "learning_rate": 0.00019811580191801697,
+      "loss": 0.9749,
+      "step": 1558
+    },
+    {
+      "epoch": 0.5005618879434901,
+      "grad_norm": 1.4966586828231812,
+      "learning_rate": 0.0001981090517443334,
+      "loss": 0.9999,
+      "step": 1559
+    },
+    {
+      "epoch": 0.5008829667683417,
+      "grad_norm": 1.356341004371643,
+      "learning_rate": 0.0001981022896163595,
+      "loss": 1.192,
+      "step": 1560
+    },
+    {
+      "epoch": 0.5012040455931931,
+      "grad_norm": 3.053656816482544,
+      "learning_rate": 0.00019809551553491916,
+      "loss": 1.1255,
+      "step": 1561
+    },
+    {
+      "epoch": 0.5015251244180446,
+      "grad_norm": 1.3365064859390259,
+      "learning_rate": 0.00019808872950083782,
+      "loss": 0.9588,
+      "step": 1562
+    },
+    {
+      "epoch": 0.5018462032428961,
+      "grad_norm": 0.789279580116272,
+      "learning_rate": 0.00019808193151494232,
+      "loss": 0.747,
+      "step": 1563
+    },
+    {
+      "epoch": 0.5021672820677476,
+      "grad_norm": 1.067087173461914,
+      "learning_rate": 0.000198075121578061,
+      "loss": 0.9783,
+      "step": 1564
+    },
+    {
+      "epoch": 0.5024883608925992,
+      "grad_norm": 1.0413140058517456,
+      "learning_rate": 0.00019806829969102357,
+      "loss": 0.8484,
+      "step": 1565
+    },
+    {
+      "epoch": 0.5028094397174506,
+      "grad_norm": 1.8259743452072144,
+      "learning_rate": 0.0001980614658546613,
+      "loss": 1.0746,
+      "step": 1566
+    },
+    {
+      "epoch": 0.5031305185423022,
+      "grad_norm": 1.282426118850708,
+      "learning_rate": 0.00019805462006980689,
+      "loss": 1.1119,
+      "step": 1567
+    },
+    {
+      "epoch": 0.5034515973671536,
+      "grad_norm": 1.075327754020691,
+      "learning_rate": 0.00019804776233729444,
+      "loss": 0.8866,
+      "step": 1568
+    },
+    {
+      "epoch": 0.5037726761920052,
+      "grad_norm": 1.3174536228179932,
+      "learning_rate": 0.0001980408926579596,
+      "loss": 0.9901,
+      "step": 1569
+    },
+    {
+      "epoch": 0.5040937550168566,
+      "grad_norm": 1.03889000415802,
+      "learning_rate": 0.00019803401103263933,
+      "loss": 0.9264,
+      "step": 1570
+    },
+    {
+      "epoch": 0.5044148338417082,
+      "grad_norm": 1.4712145328521729,
+      "learning_rate": 0.00019802711746217218,
+      "loss": 1.113,
+      "step": 1571
+    },
+    {
+      "epoch": 0.5047359126665596,
+      "grad_norm": 3.3879761695861816,
+      "learning_rate": 0.00019802021194739814,
+      "loss": 0.9667,
+      "step": 1572
+    },
+    {
+      "epoch": 0.5050569914914111,
+      "grad_norm": 1.116847276687622,
+      "learning_rate": 0.00019801329448915862,
+      "loss": 1.0384,
+      "step": 1573
+    },
+    {
+      "epoch": 0.5053780703162627,
+      "grad_norm": 1.0987874269485474,
+      "learning_rate": 0.00019800636508829643,
+      "loss": 1.0302,
+      "step": 1574
+    },
+    {
+      "epoch": 0.5056991491411141,
+      "grad_norm": 1.6565288305282593,
+      "learning_rate": 0.00019799942374565597,
+      "loss": 1.0317,
+      "step": 1575
+    },
+    {
+      "epoch": 0.5060202279659657,
+      "grad_norm": 1.2395991086959839,
+      "learning_rate": 0.00019799247046208297,
+      "loss": 1.1209,
+      "step": 1576
+    },
+    {
+      "epoch": 0.5063413067908171,
+      "grad_norm": 1.0488249063491821,
+      "learning_rate": 0.0001979855052384247,
+      "loss": 0.9093,
+      "step": 1577
+    },
+    {
+      "epoch": 0.5066623856156687,
+      "grad_norm": 1.3731625080108643,
+      "learning_rate": 0.00019797852807552983,
+      "loss": 0.9188,
+      "step": 1578
+    },
+    {
+      "epoch": 0.5069834644405201,
+      "grad_norm": 2.111989974975586,
+      "learning_rate": 0.00019797153897424852,
+      "loss": 1.098,
+      "step": 1579
+    },
+    {
+      "epoch": 0.5073045432653717,
+      "grad_norm": 1.2938238382339478,
+      "learning_rate": 0.00019796453793543238,
+      "loss": 0.8608,
+      "step": 1580
+    },
+    {
+      "epoch": 0.5076256220902231,
+      "grad_norm": 1.4208569526672363,
+      "learning_rate": 0.0001979575249599344,
+      "loss": 1.0857,
+      "step": 1581
+    },
+    {
+      "epoch": 0.5079467009150747,
+      "grad_norm": 1.5202668905258179,
+      "learning_rate": 0.00019795050004860917,
+      "loss": 1.2191,
+      "step": 1582
+    },
+    {
+      "epoch": 0.5082677797399262,
+      "grad_norm": 0.9358740448951721,
+      "learning_rate": 0.00019794346320231265,
+      "loss": 0.9756,
+      "step": 1583
+    },
+    {
+      "epoch": 0.5085888585647776,
+      "grad_norm": 1.4582595825195312,
+      "learning_rate": 0.00019793641442190221,
+      "loss": 1.0251,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5089099373896292,
+      "grad_norm": 1.3542490005493164,
+      "learning_rate": 0.00019792935370823675,
+      "loss": 1.0413,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5092310162144806,
+      "grad_norm": 0.8536126017570496,
+      "learning_rate": 0.00019792228106217658,
+      "loss": 0.7827,
+      "step": 1586
+    },
+    {
+      "epoch": 0.5095520950393322,
+      "grad_norm": 1.294661521911621,
+      "learning_rate": 0.00019791519648458352,
+      "loss": 0.8979,
+      "step": 1587
+    },
+    {
+      "epoch": 0.5098731738641836,
+      "grad_norm": 1.0872153043746948,
+      "learning_rate": 0.00019790809997632076,
+      "loss": 0.8247,
+      "step": 1588
+    },
+    {
+      "epoch": 0.5101942526890352,
+      "grad_norm": 1.6541205644607544,
+      "learning_rate": 0.000197900991538253,
+      "loss": 1.0276,
+      "step": 1589
+    },
+    {
+      "epoch": 0.5105153315138866,
+      "grad_norm": 1.2755903005599976,
+      "learning_rate": 0.00019789387117124637,
+      "loss": 0.9243,
+      "step": 1590
+    },
+    {
+      "epoch": 0.5108364103387382,
+      "grad_norm": 1.2465382814407349,
+      "learning_rate": 0.0001978867388761685,
+      "loss": 0.9222,
+      "step": 1591
+    },
+    {
+      "epoch": 0.5111574891635896,
+      "grad_norm": 0.9154629707336426,
+      "learning_rate": 0.00019787959465388842,
+      "loss": 0.8289,
+      "step": 1592
+    },
+    {
+      "epoch": 0.5114785679884412,
+      "grad_norm": 1.7428443431854248,
+      "learning_rate": 0.00019787243850527664,
+      "loss": 1.0268,
+      "step": 1593
+    },
+    {
+      "epoch": 0.5117996468132927,
+      "grad_norm": 1.0472004413604736,
+      "learning_rate": 0.0001978652704312051,
+      "loss": 0.8157,
+      "step": 1594
+    },
+    {
+      "epoch": 0.5121207256381441,
+      "grad_norm": 1.6456936597824097,
+      "learning_rate": 0.00019785809043254722,
+      "loss": 0.9551,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5124418044629957,
+      "grad_norm": 1.1858004331588745,
+      "learning_rate": 0.00019785089851017787,
+      "loss": 0.9122,
+      "step": 1596
+    },
+    {
+      "epoch": 0.5127628832878471,
+      "grad_norm": 1.2532726526260376,
+      "learning_rate": 0.0001978436946649733,
+      "loss": 0.8158,
+      "step": 1597
+    },
+    {
+      "epoch": 0.5130839621126987,
+      "grad_norm": 0.9633054733276367,
+      "learning_rate": 0.00019783647889781136,
+      "loss": 0.7278,
+      "step": 1598
+    },
+    {
+      "epoch": 0.5134050409375501,
+      "grad_norm": 1.5852664709091187,
+      "learning_rate": 0.00019782925120957124,
+      "loss": 0.8357,
+      "step": 1599
+    },
+    {
+      "epoch": 0.5137261197624017,
+      "grad_norm": 0.9574192762374878,
+      "learning_rate": 0.0001978220116011336,
+      "loss": 0.5688,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5140471985872531,
+      "grad_norm": 1.5146691799163818,
+      "learning_rate": 0.00019781476007338058,
+      "loss": 1.3743,
+      "step": 1601
+    },
+    {
+      "epoch": 0.5143682774121047,
+      "grad_norm": 1.5694231986999512,
+      "learning_rate": 0.00019780749662719573,
+      "loss": 1.6016,
+      "step": 1602
+    },
+    {
+      "epoch": 0.5146893562369562,
+      "grad_norm": 1.2235115766525269,
+      "learning_rate": 0.0001978002212634641,
+      "loss": 1.0506,
+      "step": 1603
+    },
+    {
+      "epoch": 0.5150104350618077,
+      "grad_norm": 1.2383902072906494,
+      "learning_rate": 0.0001977929339830722,
+      "loss": 0.7889,
+      "step": 1604
+    },
+    {
+      "epoch": 0.5153315138866592,
+      "grad_norm": 1.5607134103775024,
+      "learning_rate": 0.0001977856347869079,
+      "loss": 0.9399,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5156525927115106,
+      "grad_norm": 1.2659744024276733,
+      "learning_rate": 0.00019777832367586063,
+      "loss": 0.7502,
+      "step": 1606
+    },
+    {
+      "epoch": 0.5159736715363622,
+      "grad_norm": 1.645847201347351,
+      "learning_rate": 0.00019777100065082118,
+      "loss": 1.0366,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5162947503612136,
+      "grad_norm": 1.4110909700393677,
+      "learning_rate": 0.00019776366571268192,
+      "loss": 1.1166,
+      "step": 1608
+    },
+    {
+      "epoch": 0.5166158291860652,
+      "grad_norm": 1.3182811737060547,
+      "learning_rate": 0.00019775631886233654,
+      "loss": 1.0907,
+      "step": 1609
+    },
+    {
+      "epoch": 0.5169369080109166,
+      "grad_norm": 1.0148051977157593,
+      "learning_rate": 0.0001977489601006802,
+      "loss": 0.7937,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5172579868357682,
+      "grad_norm": 1.499527096748352,
+      "learning_rate": 0.0001977415894286096,
+      "loss": 0.8182,
+      "step": 1611
+    },
+    {
+      "epoch": 0.5175790656606197,
+      "grad_norm": 1.6975308656692505,
+      "learning_rate": 0.0001977342068470228,
+      "loss": 0.9804,
+      "step": 1612
+    },
+    {
+      "epoch": 0.5179001444854712,
+      "grad_norm": 1.4909874200820923,
+      "learning_rate": 0.00019772681235681936,
+      "loss": 1.088,
+      "step": 1613
+    },
+    {
+      "epoch": 0.5182212233103227,
+      "grad_norm": 1.8054780960083008,
+      "learning_rate": 0.00019771940595890027,
+      "loss": 0.8292,
+      "step": 1614
+    },
+    {
+      "epoch": 0.5185423021351742,
+      "grad_norm": 1.4584299325942993,
+      "learning_rate": 0.000197711987654168,
+      "loss": 1.0235,
+      "step": 1615
+    },
+    {
+      "epoch": 0.5188633809600257,
+      "grad_norm": 1.4883002042770386,
+      "learning_rate": 0.0001977045574435264,
+      "loss": 0.9593,
+      "step": 1616
+    },
+    {
+      "epoch": 0.5191844597848772,
+      "grad_norm": 1.0827285051345825,
+      "learning_rate": 0.00019769711532788083,
+      "loss": 0.9392,
+      "step": 1617
+    },
+    {
+      "epoch": 0.5195055386097287,
+      "grad_norm": 1.358249306678772,
+      "learning_rate": 0.0001976896613081381,
+      "loss": 1.1373,
+      "step": 1618
+    },
+    {
+      "epoch": 0.5198266174345801,
+      "grad_norm": 1.4473522901535034,
+      "learning_rate": 0.0001976821953852065,
+      "loss": 0.9981,
+      "step": 1619
+    },
+    {
+      "epoch": 0.5201476962594317,
+      "grad_norm": 1.2284332513809204,
+      "learning_rate": 0.0001976747175599957,
+      "loss": 0.9899,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5204687750842832,
+      "grad_norm": 1.0846779346466064,
+      "learning_rate": 0.0001976672278334168,
+      "loss": 1.0516,
+      "step": 1621
+    },
+    {
+      "epoch": 0.5207898539091347,
+      "grad_norm": 1.5218236446380615,
+      "learning_rate": 0.00019765972620638248,
+      "loss": 0.7433,
+      "step": 1622
+    },
+    {
+      "epoch": 0.5211109327339862,
+      "grad_norm": 1.1767224073410034,
+      "learning_rate": 0.00019765221267980675,
+      "loss": 0.9726,
+      "step": 1623
+    },
+    {
+      "epoch": 0.5214320115588377,
+      "grad_norm": 1.486958384513855,
+      "learning_rate": 0.00019764468725460508,
+      "loss": 0.8919,
+      "step": 1624
+    },
+    {
+      "epoch": 0.5217530903836892,
+      "grad_norm": 1.7597538232803345,
+      "learning_rate": 0.00019763714993169452,
+      "loss": 0.8817,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5220741692085407,
+      "grad_norm": 1.96730637550354,
+      "learning_rate": 0.00019762960071199333,
+      "loss": 1.0347,
+      "step": 1626
+    },
+    {
+      "epoch": 0.5223952480333922,
+      "grad_norm": 1.458665370941162,
+      "learning_rate": 0.0001976220395964215,
+      "loss": 0.9896,
+      "step": 1627
+    },
+    {
+      "epoch": 0.5227163268582437,
+      "grad_norm": 1.2586889266967773,
+      "learning_rate": 0.00019761446658590024,
+      "loss": 0.9721,
+      "step": 1628
+    },
+    {
+      "epoch": 0.5230374056830952,
+      "grad_norm": 1.374340295791626,
+      "learning_rate": 0.00019760688168135232,
+      "loss": 1.1315,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5233584845079468,
+      "grad_norm": 1.393900990486145,
+      "learning_rate": 0.00019759928488370193,
+      "loss": 1.0165,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5236795633327982,
+      "grad_norm": 1.0543795824050903,
+      "learning_rate": 0.00019759167619387476,
+      "loss": 0.8592,
+      "step": 1631
+    },
+    {
+      "epoch": 0.5240006421576497,
+      "grad_norm": 0.9907365441322327,
+      "learning_rate": 0.00019758405561279784,
+      "loss": 1.095,
+      "step": 1632
+    },
+    {
+      "epoch": 0.5243217209825012,
+      "grad_norm": 0.9849667549133301,
+      "learning_rate": 0.00019757642314139977,
+      "loss": 0.8316,
+      "step": 1633
+    },
+    {
+      "epoch": 0.5246427998073527,
+      "grad_norm": 1.4609546661376953,
+      "learning_rate": 0.00019756877878061052,
+      "loss": 0.8168,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5249638786322042,
+      "grad_norm": 1.2530486583709717,
+      "learning_rate": 0.0001975611225313615,
+      "loss": 0.922,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5252849574570557,
+      "grad_norm": 1.1039314270019531,
+      "learning_rate": 0.00019755345439458565,
+      "loss": 0.9009,
+      "step": 1636
+    },
+    {
+      "epoch": 0.5256060362819072,
+      "grad_norm": 1.1519142389297485,
+      "learning_rate": 0.00019754577437121733,
+      "loss": 1.0119,
+      "step": 1637
+    },
+    {
+      "epoch": 0.5259271151067587,
+      "grad_norm": 1.89316725730896,
+      "learning_rate": 0.00019753808246219224,
+      "loss": 0.9892,
+      "step": 1638
+    },
+    {
+      "epoch": 0.5262481939316103,
+      "grad_norm": 1.0566115379333496,
+      "learning_rate": 0.00019753037866844771,
+      "loss": 0.8139,
+      "step": 1639
+    },
+    {
+      "epoch": 0.5265692727564617,
+      "grad_norm": 1.0403966903686523,
+      "learning_rate": 0.00019752266299092236,
+      "loss": 0.9116,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5268903515813133,
+      "grad_norm": 1.4061709642410278,
+      "learning_rate": 0.00019751493543055632,
+      "loss": 1.0763,
+      "step": 1641
+    },
+    {
+      "epoch": 0.5272114304061647,
+      "grad_norm": 1.2321988344192505,
+      "learning_rate": 0.00019750719598829122,
+      "loss": 0.8309,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5275325092310162,
+      "grad_norm": 1.538996696472168,
+      "learning_rate": 0.00019749944466507008,
+      "loss": 1.1677,
+      "step": 1643
+    },
+    {
+      "epoch": 0.5278535880558677,
+      "grad_norm": 1.1386488676071167,
+      "learning_rate": 0.00019749168146183731,
+      "loss": 0.9062,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5281746668807192,
+      "grad_norm": 1.7029637098312378,
+      "learning_rate": 0.0001974839063795389,
+      "loss": 0.898,
+      "step": 1645
+    },
+    {
+      "epoch": 0.5284957457055707,
+      "grad_norm": 1.1630817651748657,
+      "learning_rate": 0.0001974761194191222,
+      "loss": 0.9107,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5288168245304222,
+      "grad_norm": 0.9628080725669861,
+      "learning_rate": 0.00019746832058153602,
+      "loss": 0.6659,
+      "step": 1647
+    },
+    {
+      "epoch": 0.5291379033552737,
+      "grad_norm": 0.8993632197380066,
+      "learning_rate": 0.0001974605098677306,
+      "loss": 0.8283,
+      "step": 1648
+    },
+    {
+      "epoch": 0.5294589821801252,
+      "grad_norm": 3.56905198097229,
+      "learning_rate": 0.00019745268727865774,
+      "loss": 0.6945,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5297800610049768,
+      "grad_norm": 1.1273267269134521,
+      "learning_rate": 0.00019744485281527049,
+      "loss": 0.5489,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5301011398298282,
+      "grad_norm": 1.4737544059753418,
+      "learning_rate": 0.00019743700647852354,
+      "loss": 1.4845,
+      "step": 1651
+    },
+    {
+      "epoch": 0.5304222186546798,
+      "grad_norm": 1.4742531776428223,
+      "learning_rate": 0.00019742914826937288,
+      "loss": 1.327,
+      "step": 1652
+    },
+    {
+      "epoch": 0.5307432974795312,
+      "grad_norm": 1.2791469097137451,
+      "learning_rate": 0.00019742127818877606,
+      "loss": 1.2235,
+      "step": 1653
+    },
+    {
+      "epoch": 0.5310643763043827,
+      "grad_norm": 1.511093258857727,
+      "learning_rate": 0.000197413396237692,
+      "loss": 0.8866,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5313854551292342,
+      "grad_norm": 2.150474786758423,
+      "learning_rate": 0.00019740550241708108,
+      "loss": 0.9557,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5317065339540857,
+      "grad_norm": 1.3363398313522339,
+      "learning_rate": 0.0001973975967279052,
+      "loss": 0.6729,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5320276127789372,
+      "grad_norm": 1.4328546524047852,
+      "learning_rate": 0.0001973896791711275,
+      "loss": 0.6497,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5323486916037887,
+      "grad_norm": 1.1948676109313965,
+      "learning_rate": 0.0001973817497477129,
+      "loss": 0.6963,
+      "step": 1658
+    },
+    {
+      "epoch": 0.5326697704286403,
+      "grad_norm": 1.3898074626922607,
+      "learning_rate": 0.00019737380845862745,
+      "loss": 1.0177,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5329908492534917,
+      "grad_norm": 1.250869631767273,
+      "learning_rate": 0.0001973658553048388,
+      "loss": 0.9914,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5333119280783433,
+      "grad_norm": 1.9206534624099731,
+      "learning_rate": 0.00019735789028731604,
+      "loss": 0.9322,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5336330069031947,
+      "grad_norm": 1.8212529420852661,
+      "learning_rate": 0.00019734991340702966,
+      "loss": 1.1671,
+      "step": 1662
+    },
+    {
+      "epoch": 0.5339540857280463,
+      "grad_norm": 1.4000660181045532,
+      "learning_rate": 0.00019734192466495162,
+      "loss": 1.0022,
+      "step": 1663
+    },
+    {
+      "epoch": 0.5342751645528977,
+      "grad_norm": 0.9797595143318176,
+      "learning_rate": 0.0001973339240620553,
+      "loss": 0.8335,
+      "step": 1664
+    },
+    {
+      "epoch": 0.5345962433777492,
+      "grad_norm": 1.2020220756530762,
+      "learning_rate": 0.0001973259115993156,
+      "loss": 0.9649,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5349173222026007,
+      "grad_norm": 1.1489737033843994,
+      "learning_rate": 0.00019731788727770885,
+      "loss": 0.9856,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5352384010274522,
+      "grad_norm": 1.017920970916748,
+      "learning_rate": 0.00019730985109821266,
+      "loss": 0.899,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5355594798523038,
+      "grad_norm": 0.8471775054931641,
+      "learning_rate": 0.0001973018030618063,
+      "loss": 0.88,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5358805586771552,
+      "grad_norm": 1.4469013214111328,
+      "learning_rate": 0.0001972937431694704,
+      "loss": 1.0509,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5362016375020068,
+      "grad_norm": 1.3785357475280762,
+      "learning_rate": 0.00019728567142218703,
+      "loss": 1.0469,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5365227163268582,
+      "grad_norm": 1.0268610715866089,
+      "learning_rate": 0.00019727758782093967,
+      "loss": 1.0102,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5368437951517098,
+      "grad_norm": 1.2795389890670776,
+      "learning_rate": 0.00019726949236671332,
+      "loss": 1.0537,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5371648739765612,
+      "grad_norm": 3.1554412841796875,
+      "learning_rate": 0.00019726138506049438,
+      "loss": 0.9041,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5374859528014128,
+      "grad_norm": 1.8737775087356567,
+      "learning_rate": 0.00019725326590327066,
+      "loss": 1.0927,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5378070316262642,
+      "grad_norm": 1.8332324028015137,
+      "learning_rate": 0.00019724513489603155,
+      "loss": 1.0715,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5381281104511157,
+      "grad_norm": 1.0385593175888062,
+      "learning_rate": 0.00019723699203976766,
+      "loss": 0.8541,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5384491892759673,
+      "grad_norm": 1.0988675355911255,
+      "learning_rate": 0.00019722883733547128,
+      "loss": 0.7933,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5387702681008187,
+      "grad_norm": 1.472822666168213,
+      "learning_rate": 0.00019722067078413599,
+      "loss": 0.9275,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5390913469256703,
+      "grad_norm": 1.4651199579238892,
+      "learning_rate": 0.00019721249238675688,
+      "loss": 1.2088,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5394124257505217,
+      "grad_norm": 1.5309169292449951,
+      "learning_rate": 0.00019720430214433042,
+      "loss": 1.0605,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5397335045753733,
+      "grad_norm": 1.445311427116394,
+      "learning_rate": 0.00019719610005785465,
+      "loss": 1.0896,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5400545834002247,
+      "grad_norm": 1.6236193180084229,
+      "learning_rate": 0.00019718788612832887,
+      "loss": 1.0223,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5403756622250763,
+      "grad_norm": 1.451526165008545,
+      "learning_rate": 0.00019717966035675397,
+      "loss": 0.8866,
+      "step": 1683
+    },
+    {
+      "epoch": 0.5406967410499277,
+      "grad_norm": 1.2759616374969482,
+      "learning_rate": 0.00019717142274413223,
+      "loss": 1.0889,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5410178198747793,
+      "grad_norm": 1.0535154342651367,
+      "learning_rate": 0.0001971631732914674,
+      "loss": 0.9059,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5413388986996308,
+      "grad_norm": 1.0445899963378906,
+      "learning_rate": 0.0001971549119997646,
+      "loss": 0.9921,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5416599775244823,
+      "grad_norm": 1.2191705703735352,
+      "learning_rate": 0.00019714663887003054,
+      "loss": 1.0312,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5419810563493338,
+      "grad_norm": 1.2853853702545166,
+      "learning_rate": 0.00019713835390327316,
+      "loss": 1.0743,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5423021351741852,
+      "grad_norm": 1.3771874904632568,
+      "learning_rate": 0.000197130057100502,
+      "loss": 1.033,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5426232139990368,
+      "grad_norm": 1.2575607299804688,
+      "learning_rate": 0.00019712174846272805,
+      "loss": 0.94,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5429442928238882,
+      "grad_norm": 1.1774635314941406,
+      "learning_rate": 0.00019711342799096361,
+      "loss": 0.9817,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5432653716487398,
+      "grad_norm": 1.3301347494125366,
+      "learning_rate": 0.00019710509568622258,
+      "loss": 1.0584,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5435864504735912,
+      "grad_norm": 1.1408923864364624,
+      "learning_rate": 0.00019709675154952017,
+      "loss": 0.9083,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5439075292984428,
+      "grad_norm": 1.0724045038223267,
+      "learning_rate": 0.0001970883955818731,
+      "loss": 0.7918,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5442286081232943,
+      "grad_norm": 0.9542874693870544,
+      "learning_rate": 0.00019708002778429955,
+      "loss": 0.9377,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5445496869481458,
+      "grad_norm": 1.1520448923110962,
+      "learning_rate": 0.00019707164815781908,
+      "loss": 0.6503,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5448707657729973,
+      "grad_norm": 1.0709972381591797,
+      "learning_rate": 0.00019706325670345275,
+      "loss": 0.905,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5451918445978488,
+      "grad_norm": 0.7395738959312439,
+      "learning_rate": 0.000197054853422223,
+      "loss": 0.6726,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5455129234227003,
+      "grad_norm": 1.2998228073120117,
+      "learning_rate": 0.00019704643831515374,
+      "loss": 0.8171,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5458340022475517,
+      "grad_norm": 1.2186784744262695,
+      "learning_rate": 0.00019703801138327038,
+      "loss": 0.6297,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5461550810724033,
+      "grad_norm": 1.2121597528457642,
+      "learning_rate": 0.00019702957262759965,
+      "loss": 1.4739,
+      "step": 1701
+    },
+    {
+      "epoch": 0.5464761598972547,
+      "grad_norm": 1.483771800994873,
+      "learning_rate": 0.00019702112204916984,
+      "loss": 1.3041,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5467972387221063,
+      "grad_norm": 1.4804165363311768,
+      "learning_rate": 0.0001970126596490106,
+      "loss": 0.8875,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5471183175469578,
+      "grad_norm": 1.211633324623108,
+      "learning_rate": 0.00019700418542815306,
+      "loss": 0.8732,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5474393963718093,
+      "grad_norm": 1.4509685039520264,
+      "learning_rate": 0.00019699569938762973,
+      "loss": 0.7624,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5477604751966608,
+      "grad_norm": 1.4379287958145142,
+      "learning_rate": 0.00019698720152847468,
+      "loss": 0.6996,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5480815540215123,
+      "grad_norm": 1.7644935846328735,
+      "learning_rate": 0.00019697869185172331,
+      "loss": 1.1261,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5484026328463638,
+      "grad_norm": 1.4107887744903564,
+      "learning_rate": 0.00019697017035841252,
+      "loss": 1.1591,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5487237116712153,
+      "grad_norm": 1.2489043474197388,
+      "learning_rate": 0.0001969616370495806,
+      "loss": 0.8075,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5490447904960668,
+      "grad_norm": 1.4318604469299316,
+      "learning_rate": 0.00019695309192626734,
+      "loss": 0.9552,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5493658693209182,
+      "grad_norm": 1.043524146080017,
+      "learning_rate": 0.0001969445349895139,
+      "loss": 0.7759,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5496869481457698,
+      "grad_norm": 1.1414694786071777,
+      "learning_rate": 0.00019693596624036292,
+      "loss": 1.0484,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5500080269706212,
+      "grad_norm": 1.1262508630752563,
+      "learning_rate": 0.00019692738567985853,
+      "loss": 0.924,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5503291057954728,
+      "grad_norm": 1.4722039699554443,
+      "learning_rate": 0.0001969187933090462,
+      "loss": 0.7755,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5506501846203243,
+      "grad_norm": 1.0937949419021606,
+      "learning_rate": 0.00019691018912897286,
+      "loss": 0.8581,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5509712634451758,
+      "grad_norm": 1.4399571418762207,
+      "learning_rate": 0.00019690157314068696,
+      "loss": 0.9408,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5512923422700273,
+      "grad_norm": 1.3758063316345215,
+      "learning_rate": 0.0001968929453452383,
+      "loss": 1.0267,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5516134210948788,
+      "grad_norm": 1.2351738214492798,
+      "learning_rate": 0.00019688430574367819,
+      "loss": 1.1119,
+      "step": 1718
+    },
+    {
+      "epoch": 0.5519344999197303,
+      "grad_norm": 0.9020370244979858,
+      "learning_rate": 0.00019687565433705926,
+      "loss": 0.8939,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5522555787445818,
+      "grad_norm": 1.680139183998108,
+      "learning_rate": 0.00019686699112643572,
+      "loss": 1.1864,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5525766575694333,
+      "grad_norm": 1.5821527242660522,
+      "learning_rate": 0.0001968583161128631,
+      "loss": 1.125,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5528977363942847,
+      "grad_norm": 1.2951277494430542,
+      "learning_rate": 0.00019684962929739853,
+      "loss": 1.0259,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5532188152191363,
+      "grad_norm": 1.182592749595642,
+      "learning_rate": 0.00019684093068110038,
+      "loss": 1.1474,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5535398940439878,
+      "grad_norm": 1.3489532470703125,
+      "learning_rate": 0.00019683222026502858,
+      "loss": 0.774,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5538609728688393,
+      "grad_norm": 1.1987816095352173,
+      "learning_rate": 0.00019682349805024446,
+      "loss": 1.1013,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5541820516936908,
+      "grad_norm": 1.347397804260254,
+      "learning_rate": 0.0001968147640378108,
+      "loss": 1.1095,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5545031305185423,
+      "grad_norm": 1.4341198205947876,
+      "learning_rate": 0.00019680601822879182,
+      "loss": 1.2611,
+      "step": 1727
+    },
+    {
+      "epoch": 0.5548242093433938,
+      "grad_norm": 1.5567268133163452,
+      "learning_rate": 0.00019679726062425316,
+      "loss": 1.0499,
+      "step": 1728
+    },
+    {
+      "epoch": 0.5551452881682453,
+      "grad_norm": 0.9341502785682678,
+      "learning_rate": 0.00019678849122526187,
+      "loss": 0.9607,
+      "step": 1729
+    },
+    {
+      "epoch": 0.5554663669930968,
+      "grad_norm": 1.2237082719802856,
+      "learning_rate": 0.00019677971003288655,
+      "loss": 1.0517,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5557874458179483,
+      "grad_norm": 0.9689716696739197,
+      "learning_rate": 0.00019677091704819715,
+      "loss": 1.0166,
+      "step": 1731
+    },
+    {
+      "epoch": 0.5561085246427998,
+      "grad_norm": 1.32294762134552,
+      "learning_rate": 0.000196762112272265,
+      "loss": 0.9483,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5564296034676514,
+      "grad_norm": 1.5692658424377441,
+      "learning_rate": 0.00019675329570616298,
+      "loss": 0.8919,
+      "step": 1733
+    },
+    {
+      "epoch": 0.5567506822925028,
+      "grad_norm": 1.022190809249878,
+      "learning_rate": 0.0001967444673509654,
+      "loss": 0.7758,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5570717611173543,
+      "grad_norm": 0.9672703742980957,
+      "learning_rate": 0.00019673562720774792,
+      "loss": 0.9181,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5573928399422058,
+      "grad_norm": 1.6587204933166504,
+      "learning_rate": 0.0001967267752775877,
+      "loss": 0.8566,
+      "step": 1736
+    },
+    {
+      "epoch": 0.5577139187670573,
+      "grad_norm": 1.575669765472412,
+      "learning_rate": 0.0001967179115615633,
+      "loss": 0.9348,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5580349975919088,
+      "grad_norm": 1.2742935419082642,
+      "learning_rate": 0.00019670903606075474,
+      "loss": 0.6976,
+      "step": 1738
+    },
+    {
+      "epoch": 0.5583560764167603,
+      "grad_norm": 2.927837610244751,
+      "learning_rate": 0.00019670014877624353,
+      "loss": 0.9914,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5586771552416118,
+      "grad_norm": 1.1376343965530396,
+      "learning_rate": 0.00019669124970911247,
+      "loss": 0.6877,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5589982340664633,
+      "grad_norm": 1.5874276161193848,
+      "learning_rate": 0.00019668233886044597,
+      "loss": 1.0347,
+      "step": 1741
+    },
+    {
+      "epoch": 0.5593193128913149,
+      "grad_norm": 1.3938146829605103,
+      "learning_rate": 0.0001966734162313297,
+      "loss": 0.8869,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5596403917161663,
+      "grad_norm": 2.021955728530884,
+      "learning_rate": 0.00019666448182285094,
+      "loss": 0.9766,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5599614705410179,
+      "grad_norm": 1.3332651853561401,
+      "learning_rate": 0.00019665553563609825,
+      "loss": 0.9548,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5602825493658693,
+      "grad_norm": 1.0142797231674194,
+      "learning_rate": 0.00019664657767216176,
+      "loss": 0.8287,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5606036281907208,
+      "grad_norm": 1.369992733001709,
+      "learning_rate": 0.00019663760793213296,
+      "loss": 1.0265,
+      "step": 1746
+    },
+    {
+      "epoch": 0.5609247070155723,
+      "grad_norm": 1.5672804117202759,
+      "learning_rate": 0.0001966286264171047,
+      "loss": 0.7512,
+      "step": 1747
+    },
+    {
+      "epoch": 0.5612457858404238,
+      "grad_norm": 1.7840815782546997,
+      "learning_rate": 0.00019661963312817148,
+      "loss": 0.985,
+      "step": 1748
+    },
+    {
+      "epoch": 0.5615668646652753,
+      "grad_norm": 0.9597679972648621,
+      "learning_rate": 0.00019661062806642903,
+      "loss": 0.5063,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5618879434901268,
+      "grad_norm": 1.5060789585113525,
+      "learning_rate": 0.00019660161123297458,
+      "loss": 0.5703,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5622090223149784,
+      "grad_norm": 1.2468721866607666,
+      "learning_rate": 0.00019659258262890683,
+      "loss": 1.4691,
+      "step": 1751
+    },
+    {
+      "epoch": 0.5625301011398298,
+      "grad_norm": 1.2596136331558228,
+      "learning_rate": 0.00019658354225532589,
+      "loss": 0.9999,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5628511799646814,
+      "grad_norm": 1.5096850395202637,
+      "learning_rate": 0.00019657449011333328,
+      "loss": 1.04,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5631722587895328,
+      "grad_norm": 1.6973052024841309,
+      "learning_rate": 0.00019656542620403203,
+      "loss": 0.8056,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5634933376143844,
+      "grad_norm": 1.484620213508606,
+      "learning_rate": 0.00019655635052852647,
+      "loss": 0.7742,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5638144164392358,
+      "grad_norm": 1.2450261116027832,
+      "learning_rate": 0.0001965472630879225,
+      "loss": 0.9274,
+      "step": 1756
+    },
+    {
+      "epoch": 0.5641354952640873,
+      "grad_norm": 1.8217682838439941,
+      "learning_rate": 0.0001965381638833274,
+      "loss": 0.8686,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5644565740889388,
+      "grad_norm": 1.1573996543884277,
+      "learning_rate": 0.00019652905291584984,
+      "loss": 0.9089,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5647776529137903,
+      "grad_norm": 1.1910955905914307,
+      "learning_rate": 0.0001965199301866,
+      "loss": 1.1458,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5650987317386419,
+      "grad_norm": 1.2282229661941528,
+      "learning_rate": 0.00019651079569668945,
+      "loss": 0.9382,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5654198105634933,
+      "grad_norm": 1.6961151361465454,
+      "learning_rate": 0.00019650164944723115,
+      "loss": 1.1206,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5657408893883449,
+      "grad_norm": 1.1346982717514038,
+      "learning_rate": 0.00019649249143933962,
+      "loss": 0.9793,
+      "step": 1762
+    },
+    {
+      "epoch": 0.5660619682131963,
+      "grad_norm": 1.3887813091278076,
+      "learning_rate": 0.00019648332167413067,
+      "loss": 0.9796,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5663830470380479,
+      "grad_norm": 0.7095258831977844,
+      "learning_rate": 0.0001964741401527217,
+      "loss": 0.7365,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5667041258628993,
+      "grad_norm": 1.4679083824157715,
+      "learning_rate": 0.00019646494687623135,
+      "loss": 0.8576,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5670252046877509,
+      "grad_norm": 1.2272289991378784,
+      "learning_rate": 0.00019645574184577982,
+      "loss": 0.9592,
+      "step": 1766
+    },
+    {
+      "epoch": 0.5673462835126023,
+      "grad_norm": 0.9056873917579651,
+      "learning_rate": 0.00019644652506248874,
+      "loss": 0.9373,
+      "step": 1767
+    },
+    {
+      "epoch": 0.5676673623374539,
+      "grad_norm": 1.2756919860839844,
+      "learning_rate": 0.00019643729652748113,
+      "loss": 0.9717,
+      "step": 1768
+    },
+    {
+      "epoch": 0.5679884411623053,
+      "grad_norm": 1.2208895683288574,
+      "learning_rate": 0.00019642805624188147,
+      "loss": 1.0495,
+      "step": 1769
+    },
+    {
+      "epoch": 0.5683095199871568,
+      "grad_norm": 1.9911493062973022,
+      "learning_rate": 0.00019641880420681566,
+      "loss": 1.0649,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5686305988120084,
+      "grad_norm": 1.385181188583374,
+      "learning_rate": 0.00019640954042341103,
+      "loss": 1.0059,
+      "step": 1771
+    },
+    {
+      "epoch": 0.5689516776368598,
+      "grad_norm": 1.642540693283081,
+      "learning_rate": 0.0001964002648927963,
+      "loss": 0.924,
+      "step": 1772
+    },
+    {
+      "epoch": 0.5692727564617114,
+      "grad_norm": 1.2365972995758057,
+      "learning_rate": 0.00019639097761610174,
+      "loss": 0.7404,
+      "step": 1773
+    },
+    {
+      "epoch": 0.5695938352865628,
+      "grad_norm": 1.8943426609039307,
+      "learning_rate": 0.00019638167859445895,
+      "loss": 0.9449,
+      "step": 1774
+    },
+    {
+      "epoch": 0.5699149141114144,
+      "grad_norm": 1.3321115970611572,
+      "learning_rate": 0.000196372367829001,
+      "loss": 0.9601,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5702359929362658,
+      "grad_norm": 1.592492938041687,
+      "learning_rate": 0.0001963630453208623,
+      "loss": 0.9763,
+      "step": 1776
+    },
+    {
+      "epoch": 0.5705570717611174,
+      "grad_norm": 1.3213098049163818,
+      "learning_rate": 0.00019635371107117888,
+      "loss": 1.3418,
+      "step": 1777
+    },
+    {
+      "epoch": 0.5708781505859688,
+      "grad_norm": 1.128093957901001,
+      "learning_rate": 0.000196344365081088,
+      "loss": 0.8828,
+      "step": 1778
+    },
+    {
+      "epoch": 0.5711992294108204,
+      "grad_norm": 1.2909680604934692,
+      "learning_rate": 0.0001963350073517285,
+      "loss": 1.1179,
+      "step": 1779
+    },
+    {
+      "epoch": 0.5715203082356719,
+      "grad_norm": 1.7713595628738403,
+      "learning_rate": 0.00019632563788424053,
+      "loss": 1.0826,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5718413870605233,
+      "grad_norm": 1.069727897644043,
+      "learning_rate": 0.00019631625667976583,
+      "loss": 0.659,
+      "step": 1781
+    },
+    {
+      "epoch": 0.5721624658853749,
+      "grad_norm": 1.3080742359161377,
+      "learning_rate": 0.00019630686373944738,
+      "loss": 0.8356,
+      "step": 1782
+    },
+    {
+      "epoch": 0.5724835447102263,
+      "grad_norm": 1.5017932653427124,
+      "learning_rate": 0.0001962974590644297,
+      "loss": 0.9553,
+      "step": 1783
+    },
+    {
+      "epoch": 0.5728046235350779,
+      "grad_norm": 2.8345601558685303,
+      "learning_rate": 0.00019628804265585877,
+      "loss": 0.984,
+      "step": 1784
+    },
+    {
+      "epoch": 0.5731257023599293,
+      "grad_norm": 0.9238687753677368,
+      "learning_rate": 0.00019627861451488189,
+      "loss": 0.8865,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5734467811847809,
+      "grad_norm": 1.3167964220046997,
+      "learning_rate": 0.0001962691746426479,
+      "loss": 0.8951,
+      "step": 1786
+    },
+    {
+      "epoch": 0.5737678600096323,
+      "grad_norm": 0.8535014986991882,
+      "learning_rate": 0.00019625972304030697,
+      "loss": 0.7118,
+      "step": 1787
+    },
+    {
+      "epoch": 0.5740889388344839,
+      "grad_norm": 1.6030869483947754,
+      "learning_rate": 0.00019625025970901078,
+      "loss": 0.9116,
+      "step": 1788
+    },
+    {
+      "epoch": 0.5744100176593354,
+      "grad_norm": 1.7325750589370728,
+      "learning_rate": 0.0001962407846499124,
+      "loss": 0.9291,
+      "step": 1789
+    },
+    {
+      "epoch": 0.5747310964841869,
+      "grad_norm": 1.3394780158996582,
+      "learning_rate": 0.00019623129786416635,
+      "loss": 0.9792,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5750521753090384,
+      "grad_norm": 1.3643033504486084,
+      "learning_rate": 0.00019622179935292855,
+      "loss": 0.8923,
+      "step": 1791
+    },
+    {
+      "epoch": 0.5753732541338898,
+      "grad_norm": 1.2005494832992554,
+      "learning_rate": 0.00019621228911735636,
+      "loss": 1.0839,
+      "step": 1792
+    },
+    {
+      "epoch": 0.5756943329587414,
+      "grad_norm": 1.9481810331344604,
+      "learning_rate": 0.0001962027671586086,
+      "loss": 1.0381,
+      "step": 1793
+    },
+    {
+      "epoch": 0.5760154117835928,
+      "grad_norm": 1.1878541707992554,
+      "learning_rate": 0.00019619323347784548,
+      "loss": 1.005,
+      "step": 1794
+    },
+    {
+      "epoch": 0.5763364906084444,
+      "grad_norm": 1.2770967483520508,
+      "learning_rate": 0.00019618368807622862,
+      "loss": 1.008,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5766575694332958,
+      "grad_norm": 0.9496734738349915,
+      "learning_rate": 0.00019617413095492114,
+      "loss": 0.7313,
+      "step": 1796
+    },
+    {
+      "epoch": 0.5769786482581474,
+      "grad_norm": 1.6152045726776123,
+      "learning_rate": 0.00019616456211508752,
+      "loss": 0.9678,
+      "step": 1797
+    },
+    {
+      "epoch": 0.5772997270829989,
+      "grad_norm": 1.1719512939453125,
+      "learning_rate": 0.0001961549815578937,
+      "loss": 0.735,
+      "step": 1798
+    },
+    {
+      "epoch": 0.5776208059078504,
+      "grad_norm": 0.9691303372383118,
+      "learning_rate": 0.0001961453892845071,
+      "loss": 0.6029,
+      "step": 1799
+    },
+    {
+      "epoch": 0.5779418847327019,
+      "grad_norm": 1.4163908958435059,
+      "learning_rate": 0.0001961357852960964,
+      "loss": 0.6984,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5782629635575534,
+      "grad_norm": 2.1804752349853516,
+      "learning_rate": 0.0001961261695938319,
+      "loss": 1.5497,
+      "step": 1801
+    },
+    {
+      "epoch": 0.5785840423824049,
+      "grad_norm": 1.3419125080108643,
+      "learning_rate": 0.0001961165421788852,
+      "loss": 1.3835,
+      "step": 1802
+    },
+    {
+      "epoch": 0.5789051212072563,
+      "grad_norm": 1.8257540464401245,
+      "learning_rate": 0.0001961069030524294,
+      "loss": 0.9603,
+      "step": 1803
+    },
+    {
+      "epoch": 0.5792262000321079,
+      "grad_norm": 1.763144612312317,
+      "learning_rate": 0.00019609725221563897,
+      "loss": 0.9591,
+      "step": 1804
+    },
+    {
+      "epoch": 0.5795472788569593,
+      "grad_norm": 1.1090623140335083,
+      "learning_rate": 0.00019608758966968988,
+      "loss": 0.8613,
+      "step": 1805
+    },
+    {
+      "epoch": 0.5798683576818109,
+      "grad_norm": 1.6234127283096313,
+      "learning_rate": 0.00019607791541575943,
+      "loss": 0.8793,
+      "step": 1806
+    },
+    {
+      "epoch": 0.5801894365066624,
+      "grad_norm": 1.1503936052322388,
+      "learning_rate": 0.0001960682294550264,
+      "loss": 0.6553,
+      "step": 1807
+    },
+    {
+      "epoch": 0.5805105153315139,
+      "grad_norm": 1.3944035768508911,
+      "learning_rate": 0.00019605853178867105,
+      "loss": 0.682,
+      "step": 1808
+    },
+    {
+      "epoch": 0.5808315941563654,
+      "grad_norm": 1.1830666065216064,
+      "learning_rate": 0.00019604882241787498,
+      "loss": 0.9909,
+      "step": 1809
+    },
+    {
+      "epoch": 0.5811526729812169,
+      "grad_norm": 1.4636564254760742,
+      "learning_rate": 0.00019603910134382123,
+      "loss": 1.0726,
+      "step": 1810
+    },
+    {
+      "epoch": 0.5814737518060684,
+      "grad_norm": 1.203645944595337,
+      "learning_rate": 0.0001960293685676943,
+      "loss": 0.9043,
+      "step": 1811
+    },
+    {
+      "epoch": 0.5817948306309199,
+      "grad_norm": 1.30366051197052,
+      "learning_rate": 0.00019601962409068012,
+      "loss": 0.9162,
+      "step": 1812
+    },
+    {
+      "epoch": 0.5821159094557714,
+      "grad_norm": 2.1224048137664795,
+      "learning_rate": 0.000196009867913966,
+      "loss": 1.1189,
+      "step": 1813
+    },
+    {
+      "epoch": 0.5824369882806228,
+      "grad_norm": 0.9786831140518188,
+      "learning_rate": 0.00019600010003874069,
+      "loss": 0.8792,
+      "step": 1814
+    },
+    {
+      "epoch": 0.5827580671054744,
+      "grad_norm": 1.0438469648361206,
+      "learning_rate": 0.00019599032046619438,
+      "loss": 1.1121,
+      "step": 1815
+    },
+    {
+      "epoch": 0.583079145930326,
+      "grad_norm": 1.4787214994430542,
+      "learning_rate": 0.0001959805291975187,
+      "loss": 1.1414,
+      "step": 1816
+    },
+    {
+      "epoch": 0.5834002247551774,
+      "grad_norm": 1.3728911876678467,
+      "learning_rate": 0.00019597072623390668,
+      "loss": 0.8714,
+      "step": 1817
+    },
+    {
+      "epoch": 0.5837213035800289,
+      "grad_norm": 1.0536571741104126,
+      "learning_rate": 0.00019596091157655278,
+      "loss": 0.9525,
+      "step": 1818
+    },
+    {
+      "epoch": 0.5840423824048804,
+      "grad_norm": 1.2520835399627686,
+      "learning_rate": 0.0001959510852266529,
+      "loss": 0.7777,
+      "step": 1819
+    },
+    {
+      "epoch": 0.5843634612297319,
+      "grad_norm": 0.8757473826408386,
+      "learning_rate": 0.0001959412471854043,
+      "loss": 1.0104,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5846845400545834,
+      "grad_norm": 1.4845868349075317,
+      "learning_rate": 0.00019593139745400576,
+      "loss": 0.9476,
+      "step": 1821
+    },
+    {
+      "epoch": 0.5850056188794349,
+      "grad_norm": 2.418626546859741,
+      "learning_rate": 0.00019592153603365743,
+      "loss": 1.0727,
+      "step": 1822
+    },
+    {
+      "epoch": 0.5853266977042864,
+      "grad_norm": 1.0435357093811035,
+      "learning_rate": 0.0001959116629255609,
+      "loss": 1.0303,
+      "step": 1823
+    },
+    {
+      "epoch": 0.5856477765291379,
+      "grad_norm": 1.3769514560699463,
+      "learning_rate": 0.0001959017781309192,
+      "loss": 1.0692,
+      "step": 1824
+    },
+    {
+      "epoch": 0.5859688553539895,
+      "grad_norm": 1.0408692359924316,
+      "learning_rate": 0.0001958918816509367,
+      "loss": 0.903,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5862899341788409,
+      "grad_norm": 1.2975987195968628,
+      "learning_rate": 0.0001958819734868193,
+      "loss": 0.9121,
+      "step": 1826
+    },
+    {
+      "epoch": 0.5866110130036924,
+      "grad_norm": 1.479382872581482,
+      "learning_rate": 0.00019587205363977427,
+      "loss": 0.9408,
+      "step": 1827
+    },
+    {
+      "epoch": 0.5869320918285439,
+      "grad_norm": 1.200589656829834,
+      "learning_rate": 0.00019586212211101037,
+      "loss": 1.0283,
+      "step": 1828
+    },
+    {
+      "epoch": 0.5872531706533954,
+      "grad_norm": 1.039357304573059,
+      "learning_rate": 0.0001958521789017376,
+      "loss": 0.8903,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5875742494782469,
+      "grad_norm": 1.5455893278121948,
+      "learning_rate": 0.0001958422240131676,
+      "loss": 1.0474,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5878953283030984,
+      "grad_norm": 1.092637538909912,
+      "learning_rate": 0.00019583225744651333,
+      "loss": 0.9873,
+      "step": 1831
+    },
+    {
+      "epoch": 0.5882164071279499,
+      "grad_norm": 2.400148868560791,
+      "learning_rate": 0.00019582227920298916,
+      "loss": 1.2347,
+      "step": 1832
+    },
+    {
+      "epoch": 0.5885374859528014,
+      "grad_norm": 1.1216742992401123,
+      "learning_rate": 0.00019581228928381098,
+      "loss": 1.0158,
+      "step": 1833
+    },
+    {
+      "epoch": 0.5888585647776529,
+      "grad_norm": 1.167275309562683,
+      "learning_rate": 0.00019580228769019593,
+      "loss": 0.9061,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5891796436025044,
+      "grad_norm": 1.382562518119812,
+      "learning_rate": 0.00019579227442336278,
+      "loss": 1.216,
+      "step": 1835
+    },
+    {
+      "epoch": 0.589500722427356,
+      "grad_norm": 1.2429423332214355,
+      "learning_rate": 0.0001957822494845315,
+      "loss": 0.9084,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5898218012522074,
+      "grad_norm": 1.1769795417785645,
+      "learning_rate": 0.00019577221287492367,
+      "loss": 0.9287,
+      "step": 1837
+    },
+    {
+      "epoch": 0.590142880077059,
+      "grad_norm": 1.2153514623641968,
+      "learning_rate": 0.00019576216459576222,
+      "loss": 1.0391,
+      "step": 1838
+    },
+    {
+      "epoch": 0.5904639589019104,
+      "grad_norm": 1.096490740776062,
+      "learning_rate": 0.00019575210464827149,
+      "loss": 0.9332,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5907850377267619,
+      "grad_norm": 1.9070253372192383,
+      "learning_rate": 0.00019574203303367727,
+      "loss": 0.859,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5911061165516134,
+      "grad_norm": 1.2111899852752686,
+      "learning_rate": 0.00019573194975320673,
+      "loss": 1.0296,
+      "step": 1841
+    },
+    {
+      "epoch": 0.5914271953764649,
+      "grad_norm": 1.3868889808654785,
+      "learning_rate": 0.00019572185480808848,
+      "loss": 1.0413,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5917482742013164,
+      "grad_norm": 1.646368384361267,
+      "learning_rate": 0.00019571174819955263,
+      "loss": 0.9224,
+      "step": 1843
+    },
+    {
+      "epoch": 0.5920693530261679,
+      "grad_norm": 1.1916215419769287,
+      "learning_rate": 0.00019570162992883054,
+      "loss": 0.9388,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5923904318510195,
+      "grad_norm": 1.1193400621414185,
+      "learning_rate": 0.00019569149999715515,
+      "loss": 0.8399,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5927115106758709,
+      "grad_norm": 1.161150574684143,
+      "learning_rate": 0.00019568135840576076,
+      "loss": 0.8907,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5930325895007225,
+      "grad_norm": 1.9916490316390991,
+      "learning_rate": 0.00019567120515588308,
+      "loss": 0.9989,
+      "step": 1847
+    },
+    {
+      "epoch": 0.5933536683255739,
+      "grad_norm": 1.4825630187988281,
+      "learning_rate": 0.0001956610402487592,
+      "loss": 0.6702,
+      "step": 1848
+    },
+    {
+      "epoch": 0.5936747471504255,
+      "grad_norm": 1.547219157218933,
+      "learning_rate": 0.0001956508636856278,
+      "loss": 0.6974,
+      "step": 1849
+    },
+    {
+      "epoch": 0.5939958259752769,
+      "grad_norm": 0.9746522307395935,
+      "learning_rate": 0.00019564067546772878,
+      "loss": 0.499,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5943169048001284,
+      "grad_norm": 1.3846455812454224,
+      "learning_rate": 0.00019563047559630357,
+      "loss": 1.4075,
+      "step": 1851
+    },
+    {
+      "epoch": 0.5946379836249799,
+      "grad_norm": 1.398285984992981,
+      "learning_rate": 0.00019562026407259495,
+      "loss": 0.8492,
+      "step": 1852
+    },
+    {
+      "epoch": 0.5949590624498314,
+      "grad_norm": 3.702108860015869,
+      "learning_rate": 0.00019561004089784723,
+      "loss": 1.2278,
+      "step": 1853
+    },
+    {
+      "epoch": 0.595280141274683,
+      "grad_norm": 1.2834221124649048,
+      "learning_rate": 0.00019559980607330605,
+      "loss": 0.7047,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5956012200995344,
+      "grad_norm": 1.3995305299758911,
+      "learning_rate": 0.00019558955960021849,
+      "loss": 0.7136,
+      "step": 1855
+    },
+    {
+      "epoch": 0.595922298924386,
+      "grad_norm": 1.1465208530426025,
+      "learning_rate": 0.00019557930147983302,
+      "loss": 0.651,
+      "step": 1856
+    },
+    {
+      "epoch": 0.5962433777492374,
+      "grad_norm": 1.3999861478805542,
+      "learning_rate": 0.00019556903171339963,
+      "loss": 0.7879,
+      "step": 1857
+    },
+    {
+      "epoch": 0.596564456574089,
+      "grad_norm": 1.0851597785949707,
+      "learning_rate": 0.00019555875030216954,
+      "loss": 0.8221,
+      "step": 1858
+    },
+    {
+      "epoch": 0.5968855353989404,
+      "grad_norm": 1.2108211517333984,
+      "learning_rate": 0.00019554845724739566,
+      "loss": 0.7746,
+      "step": 1859
+    },
+    {
+      "epoch": 0.597206614223792,
+      "grad_norm": 1.5989631414413452,
+      "learning_rate": 0.00019553815255033205,
+      "loss": 1.028,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5975276930486434,
+      "grad_norm": 1.233514666557312,
+      "learning_rate": 0.00019552783621223436,
+      "loss": 0.6234,
+      "step": 1861
+    },
+    {
+      "epoch": 0.5978487718734949,
+      "grad_norm": 0.9018953442573547,
+      "learning_rate": 0.0001955175082343596,
+      "loss": 0.824,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5981698506983465,
+      "grad_norm": 1.028957486152649,
+      "learning_rate": 0.00019550716861796623,
+      "loss": 0.9231,
+      "step": 1863
+    },
+    {
+      "epoch": 0.5984909295231979,
+      "grad_norm": 1.2545114755630493,
+      "learning_rate": 0.00019549681736431404,
+      "loss": 0.8434,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5988120083480495,
+      "grad_norm": 1.197217583656311,
+      "learning_rate": 0.00019548645447466431,
+      "loss": 0.9775,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5991330871729009,
+      "grad_norm": 1.3030047416687012,
+      "learning_rate": 0.00019547607995027978,
+      "loss": 0.8244,
+      "step": 1866
+    },
+    {
+      "epoch": 0.5994541659977525,
+      "grad_norm": 1.4685194492340088,
+      "learning_rate": 0.00019546569379242444,
+      "loss": 0.9339,
+      "step": 1867
+    },
+    {
+      "epoch": 0.5997752448226039,
+      "grad_norm": 1.9783920049667358,
+      "learning_rate": 0.00019545529600236398,
+      "loss": 0.9437,
+      "step": 1868
+    },
+    {
+      "epoch": 0.6000963236474555,
+      "grad_norm": 1.3374611139297485,
+      "learning_rate": 0.00019544488658136523,
+      "loss": 1.0672,
+      "step": 1869
+    },
+    {
+      "epoch": 0.6004174024723069,
+      "grad_norm": 1.4614479541778564,
+      "learning_rate": 0.0001954344655306965,
+      "loss": 0.9309,
+      "step": 1870
+    },
+    {
+      "epoch": 0.6007384812971585,
+      "grad_norm": 1.359053373336792,
+      "learning_rate": 0.0001954240328516277,
+      "loss": 1.0011,
+      "step": 1871
+    },
+    {
+      "epoch": 0.60105956012201,
+      "grad_norm": 1.4997376203536987,
+      "learning_rate": 0.00019541358854542991,
+      "loss": 1.0507,
+      "step": 1872
+    },
+    {
+      "epoch": 0.6013806389468614,
+      "grad_norm": 1.5839393138885498,
+      "learning_rate": 0.0001954031326133758,
+      "loss": 1.1981,
+      "step": 1873
+    },
+    {
+      "epoch": 0.601701717771713,
+      "grad_norm": 1.4018974304199219,
+      "learning_rate": 0.00019539266505673938,
+      "loss": 1.0872,
+      "step": 1874
+    },
+    {
+      "epoch": 0.6020227965965644,
+      "grad_norm": 1.35053551197052,
+      "learning_rate": 0.00019538218587679605,
+      "loss": 0.8378,
+      "step": 1875
+    },
+    {
+      "epoch": 0.602343875421416,
+      "grad_norm": 0.7986770272254944,
+      "learning_rate": 0.0001953716950748227,
+      "loss": 0.6296,
+      "step": 1876
+    },
+    {
+      "epoch": 0.6026649542462674,
+      "grad_norm": 1.5917302370071411,
+      "learning_rate": 0.0001953611926520976,
+      "loss": 1.0359,
+      "step": 1877
+    },
+    {
+      "epoch": 0.602986033071119,
+      "grad_norm": 1.197108507156372,
+      "learning_rate": 0.00019535067860990046,
+      "loss": 1.1222,
+      "step": 1878
+    },
+    {
+      "epoch": 0.6033071118959704,
+      "grad_norm": 1.6555898189544678,
+      "learning_rate": 0.00019534015294951233,
+      "loss": 1.0129,
+      "step": 1879
+    },
+    {
+      "epoch": 0.603628190720822,
+      "grad_norm": 0.971400797367096,
+      "learning_rate": 0.00019532961567221576,
+      "loss": 0.9535,
+      "step": 1880
+    },
+    {
+      "epoch": 0.6039492695456735,
+      "grad_norm": 1.5631428956985474,
+      "learning_rate": 0.0001953190667792947,
+      "loss": 1.2075,
+      "step": 1881
+    },
+    {
+      "epoch": 0.604270348370525,
+      "grad_norm": 1.193997859954834,
+      "learning_rate": 0.0001953085062720345,
+      "loss": 0.9112,
+      "step": 1882
+    },
+    {
+      "epoch": 0.6045914271953765,
+      "grad_norm": 1.317486047744751,
+      "learning_rate": 0.00019529793415172192,
+      "loss": 1.02,
+      "step": 1883
+    },
+    {
+      "epoch": 0.604912506020228,
+      "grad_norm": 0.9995319247245789,
+      "learning_rate": 0.00019528735041964509,
+      "loss": 0.8154,
+      "step": 1884
+    },
+    {
+      "epoch": 0.6052335848450795,
+      "grad_norm": 1.0479660034179688,
+      "learning_rate": 0.00019527675507709366,
+      "loss": 0.8572,
+      "step": 1885
+    },
+    {
+      "epoch": 0.6055546636699309,
+      "grad_norm": 1.4373043775558472,
+      "learning_rate": 0.00019526614812535864,
+      "loss": 0.9219,
+      "step": 1886
+    },
+    {
+      "epoch": 0.6058757424947825,
+      "grad_norm": 1.3879246711730957,
+      "learning_rate": 0.00019525552956573244,
+      "loss": 0.9512,
+      "step": 1887
+    },
+    {
+      "epoch": 0.6061968213196339,
+      "grad_norm": 1.2492396831512451,
+      "learning_rate": 0.0001952448993995089,
+      "loss": 1.0225,
+      "step": 1888
+    },
+    {
+      "epoch": 0.6065179001444855,
+      "grad_norm": 1.0773907899856567,
+      "learning_rate": 0.00019523425762798329,
+      "loss": 0.7924,
+      "step": 1889
+    },
+    {
+      "epoch": 0.6068389789693369,
+      "grad_norm": 1.1691362857818604,
+      "learning_rate": 0.00019522360425245226,
+      "loss": 0.8436,
+      "step": 1890
+    },
+    {
+      "epoch": 0.6071600577941885,
+      "grad_norm": 1.218471646308899,
+      "learning_rate": 0.00019521293927421388,
+      "loss": 1.0175,
+      "step": 1891
+    },
+    {
+      "epoch": 0.60748113661904,
+      "grad_norm": 0.9042808413505554,
+      "learning_rate": 0.00019520226269456768,
+      "loss": 0.9039,
+      "step": 1892
+    },
+    {
+      "epoch": 0.6078022154438915,
+      "grad_norm": 1.3849420547485352,
+      "learning_rate": 0.00019519157451481454,
+      "loss": 1.1497,
+      "step": 1893
+    },
+    {
+      "epoch": 0.608123294268743,
+      "grad_norm": 1.3135268688201904,
+      "learning_rate": 0.0001951808747362568,
+      "loss": 0.8948,
+      "step": 1894
+    },
+    {
+      "epoch": 0.6084443730935944,
+      "grad_norm": 0.94450443983078,
+      "learning_rate": 0.0001951701633601982,
+      "loss": 0.9394,
+      "step": 1895
+    },
+    {
+      "epoch": 0.608765451918446,
+      "grad_norm": 1.5512750148773193,
+      "learning_rate": 0.00019515944038794384,
+      "loss": 0.9455,
+      "step": 1896
+    },
+    {
+      "epoch": 0.6090865307432974,
+      "grad_norm": 1.3148894309997559,
+      "learning_rate": 0.00019514870582080032,
+      "loss": 0.8146,
+      "step": 1897
+    },
+    {
+      "epoch": 0.609407609568149,
+      "grad_norm": 1.836193323135376,
+      "learning_rate": 0.00019513795966007562,
+      "loss": 0.5957,
+      "step": 1898
+    },
+    {
+      "epoch": 0.6097286883930004,
+      "grad_norm": 1.1311513185501099,
+      "learning_rate": 0.00019512720190707913,
+      "loss": 0.9068,
+      "step": 1899
+    },
+    {
+      "epoch": 0.610049767217852,
+      "grad_norm": 0.6564053893089294,
+      "learning_rate": 0.00019511643256312164,
+      "loss": 0.4583,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6103708460427035,
+      "grad_norm": 1.3620964288711548,
+      "learning_rate": 0.00019510565162951537,
+      "loss": 1.3662,
+      "step": 1901
+    },
+    {
+      "epoch": 0.610691924867555,
+      "grad_norm": 1.2974669933319092,
+      "learning_rate": 0.0001950948591075739,
+      "loss": 1.2901,
+      "step": 1902
+    },
+    {
+      "epoch": 0.6110130036924065,
+      "grad_norm": 1.4451632499694824,
+      "learning_rate": 0.00019508405499861232,
+      "loss": 0.9189,
+      "step": 1903
+    },
+    {
+      "epoch": 0.611334082517258,
+      "grad_norm": 1.583996057510376,
+      "learning_rate": 0.00019507323930394708,
+      "loss": 0.9118,
+      "step": 1904
+    },
+    {
+      "epoch": 0.6116551613421095,
+      "grad_norm": 1.4750548601150513,
+      "learning_rate": 0.00019506241202489602,
+      "loss": 0.7096,
+      "step": 1905
+    },
+    {
+      "epoch": 0.611976240166961,
+      "grad_norm": 1.2632248401641846,
+      "learning_rate": 0.00019505157316277837,
+      "loss": 0.9424,
+      "step": 1906
+    },
+    {
+      "epoch": 0.6122973189918125,
+      "grad_norm": 1.392075538635254,
+      "learning_rate": 0.00019504072271891488,
+      "loss": 0.9846,
+      "step": 1907
+    },
+    {
+      "epoch": 0.6126183978166639,
+      "grad_norm": 1.0871883630752563,
+      "learning_rate": 0.00019502986069462762,
+      "loss": 0.9791,
+      "step": 1908
+    },
+    {
+      "epoch": 0.6129394766415155,
+      "grad_norm": 1.2287681102752686,
+      "learning_rate": 0.00019501898709124008,
+      "loss": 1.0328,
+      "step": 1909
+    },
+    {
+      "epoch": 0.613260555466367,
+      "grad_norm": 1.0907347202301025,
+      "learning_rate": 0.00019500810191007718,
+      "loss": 0.9497,
+      "step": 1910
+    },
+    {
+      "epoch": 0.6135816342912185,
+      "grad_norm": 1.0309808254241943,
+      "learning_rate": 0.00019499720515246525,
+      "loss": 0.7238,
+      "step": 1911
+    },
+    {
+      "epoch": 0.61390271311607,
+      "grad_norm": 1.0022937059402466,
+      "learning_rate": 0.00019498629681973206,
+      "loss": 0.8948,
+      "step": 1912
+    },
+    {
+      "epoch": 0.6142237919409215,
+      "grad_norm": 1.440856695175171,
+      "learning_rate": 0.00019497537691320668,
+      "loss": 1.0733,
+      "step": 1913
+    },
+    {
+      "epoch": 0.614544870765773,
+      "grad_norm": 1.5727375745773315,
+      "learning_rate": 0.00019496444543421975,
+      "loss": 1.0399,
+      "step": 1914
+    },
+    {
+      "epoch": 0.6148659495906245,
+      "grad_norm": 1.1591044664382935,
+      "learning_rate": 0.0001949535023841032,
+      "loss": 0.9664,
+      "step": 1915
+    },
+    {
+      "epoch": 0.615187028415476,
+      "grad_norm": 1.1390159130096436,
+      "learning_rate": 0.0001949425477641904,
+      "loss": 0.9239,
+      "step": 1916
+    },
+    {
+      "epoch": 0.6155081072403275,
+      "grad_norm": 1.00704026222229,
+      "learning_rate": 0.00019493158157581615,
+      "loss": 0.9421,
+      "step": 1917
+    },
+    {
+      "epoch": 0.615829186065179,
+      "grad_norm": 2.458767890930176,
+      "learning_rate": 0.00019492060382031662,
+      "loss": 1.1242,
+      "step": 1918
+    },
+    {
+      "epoch": 0.6161502648900306,
+      "grad_norm": 1.4069280624389648,
+      "learning_rate": 0.00019490961449902946,
+      "loss": 1.0579,
+      "step": 1919
+    },
+    {
+      "epoch": 0.616471343714882,
+      "grad_norm": 0.9519532322883606,
+      "learning_rate": 0.00019489861361329366,
+      "loss": 0.7231,
+      "step": 1920
+    },
+    {
+      "epoch": 0.6167924225397335,
+      "grad_norm": 1.2751188278198242,
+      "learning_rate": 0.00019488760116444966,
+      "loss": 1.1183,
+      "step": 1921
+    },
+    {
+      "epoch": 0.617113501364585,
+      "grad_norm": 1.6662013530731201,
+      "learning_rate": 0.00019487657715383926,
+      "loss": 1.1126,
+      "step": 1922
+    },
+    {
+      "epoch": 0.6174345801894365,
+      "grad_norm": 1.7743055820465088,
+      "learning_rate": 0.00019486554158280574,
+      "loss": 0.945,
+      "step": 1923
+    },
+    {
+      "epoch": 0.617755659014288,
+      "grad_norm": 1.3609472513198853,
+      "learning_rate": 0.00019485449445269377,
+      "loss": 0.9343,
+      "step": 1924
+    },
+    {
+      "epoch": 0.6180767378391395,
+      "grad_norm": 2.255979061126709,
+      "learning_rate": 0.00019484343576484933,
+      "loss": 0.9,
+      "step": 1925
+    },
+    {
+      "epoch": 0.618397816663991,
+      "grad_norm": 1.3037607669830322,
+      "learning_rate": 0.00019483236552061994,
+      "loss": 1.059,
+      "step": 1926
+    },
+    {
+      "epoch": 0.6187188954888425,
+      "grad_norm": 1.691819190979004,
+      "learning_rate": 0.00019482128372135446,
+      "loss": 1.0162,
+      "step": 1927
+    },
+    {
+      "epoch": 0.6190399743136941,
+      "grad_norm": 1.8937206268310547,
+      "learning_rate": 0.0001948101903684032,
+      "loss": 0.9981,
+      "step": 1928
+    },
+    {
+      "epoch": 0.6193610531385455,
+      "grad_norm": 1.4838215112686157,
+      "learning_rate": 0.00019479908546311781,
+      "loss": 1.0377,
+      "step": 1929
+    },
+    {
+      "epoch": 0.619682131963397,
+      "grad_norm": 1.2339526414871216,
+      "learning_rate": 0.00019478796900685146,
+      "loss": 0.911,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6200032107882485,
+      "grad_norm": 1.4416377544403076,
+      "learning_rate": 0.0001947768410009586,
+      "loss": 0.9814,
+      "step": 1931
+    },
+    {
+      "epoch": 0.6203242896131,
+      "grad_norm": 1.8542779684066772,
+      "learning_rate": 0.0001947657014467951,
+      "loss": 1.0792,
+      "step": 1932
+    },
+    {
+      "epoch": 0.6206453684379515,
+      "grad_norm": 1.3977333307266235,
+      "learning_rate": 0.00019475455034571838,
+      "loss": 0.9371,
+      "step": 1933
+    },
+    {
+      "epoch": 0.620966447262803,
+      "grad_norm": 1.1695326566696167,
+      "learning_rate": 0.0001947433876990871,
+      "loss": 0.9352,
+      "step": 1934
+    },
+    {
+      "epoch": 0.6212875260876545,
+      "grad_norm": 1.1949793100357056,
+      "learning_rate": 0.00019473221350826142,
+      "loss": 0.8381,
+      "step": 1935
+    },
+    {
+      "epoch": 0.621608604912506,
+      "grad_norm": 1.197929859161377,
+      "learning_rate": 0.0001947210277746029,
+      "loss": 1.0588,
+      "step": 1936
+    },
+    {
+      "epoch": 0.6219296837373576,
+      "grad_norm": 0.9528883695602417,
+      "learning_rate": 0.00019470983049947444,
+      "loss": 0.7636,
+      "step": 1937
+    },
+    {
+      "epoch": 0.622250762562209,
+      "grad_norm": 0.9562763571739197,
+      "learning_rate": 0.00019469862168424042,
+      "loss": 0.9649,
+      "step": 1938
+    },
+    {
+      "epoch": 0.6225718413870606,
+      "grad_norm": 1.5887722969055176,
+      "learning_rate": 0.0001946874013302666,
+      "loss": 0.9054,
+      "step": 1939
+    },
+    {
+      "epoch": 0.622892920211912,
+      "grad_norm": 1.0492193698883057,
+      "learning_rate": 0.0001946761694389202,
+      "loss": 0.9068,
+      "step": 1940
+    },
+    {
+      "epoch": 0.6232139990367636,
+      "grad_norm": 0.9907318353652954,
+      "learning_rate": 0.00019466492601156966,
+      "loss": 0.8565,
+      "step": 1941
+    },
+    {
+      "epoch": 0.623535077861615,
+      "grad_norm": 1.7006665468215942,
+      "learning_rate": 0.00019465367104958505,
+      "loss": 1.0636,
+      "step": 1942
+    },
+    {
+      "epoch": 0.6238561566864665,
+      "grad_norm": 1.3816338777542114,
+      "learning_rate": 0.00019464240455433775,
+      "loss": 0.9676,
+      "step": 1943
+    },
+    {
+      "epoch": 0.624177235511318,
+      "grad_norm": 1.1604748964309692,
+      "learning_rate": 0.00019463112652720054,
+      "loss": 0.8379,
+      "step": 1944
+    },
+    {
+      "epoch": 0.6244983143361695,
+      "grad_norm": 5.007503509521484,
+      "learning_rate": 0.00019461983696954758,
+      "loss": 1.1482,
+      "step": 1945
+    },
+    {
+      "epoch": 0.6248193931610211,
+      "grad_norm": 1.0724717378616333,
+      "learning_rate": 0.00019460853588275454,
+      "loss": 0.7847,
+      "step": 1946
+    },
+    {
+      "epoch": 0.6251404719858725,
+      "grad_norm": 1.2051833868026733,
+      "learning_rate": 0.00019459722326819838,
+      "loss": 0.9865,
+      "step": 1947
+    },
+    {
+      "epoch": 0.6254615508107241,
+      "grad_norm": 1.1510910987854004,
+      "learning_rate": 0.00019458589912725748,
+      "loss": 0.689,
+      "step": 1948
+    },
+    {
+      "epoch": 0.6257826296355755,
+      "grad_norm": 0.7677122354507446,
+      "learning_rate": 0.0001945745634613117,
+      "loss": 0.7075,
+      "step": 1949
+    },
+    {
+      "epoch": 0.6261037084604271,
+      "grad_norm": 1.6661738157272339,
+      "learning_rate": 0.00019456321627174221,
+      "loss": 0.6606,
+      "step": 1950
+    },
+    {
+      "epoch": 0.6264247872852785,
+      "grad_norm": 1.3030189275741577,
+      "learning_rate": 0.0001945518575599317,
+      "loss": 1.3519,
+      "step": 1951
+    },
+    {
+      "epoch": 0.6267458661101301,
+      "grad_norm": 1.27639901638031,
+      "learning_rate": 0.00019454048732726412,
+      "loss": 1.4878,
+      "step": 1952
+    },
+    {
+      "epoch": 0.6270669449349815,
+      "grad_norm": 1.197721004486084,
+      "learning_rate": 0.00019452910557512496,
+      "loss": 1.0872,
+      "step": 1953
+    },
+    {
+      "epoch": 0.627388023759833,
+      "grad_norm": 1.3855386972427368,
+      "learning_rate": 0.000194517712304901,
+      "loss": 0.8866,
+      "step": 1954
+    },
+    {
+      "epoch": 0.6277091025846845,
+      "grad_norm": 1.5599814653396606,
+      "learning_rate": 0.00019450630751798048,
+      "loss": 0.8322,
+      "step": 1955
+    },
+    {
+      "epoch": 0.628030181409536,
+      "grad_norm": 1.247623324394226,
+      "learning_rate": 0.0001944948912157531,
+      "loss": 0.4998,
+      "step": 1956
+    },
+    {
+      "epoch": 0.6283512602343876,
+      "grad_norm": 1.6704400777816772,
+      "learning_rate": 0.00019448346339960982,
+      "loss": 0.739,
+      "step": 1957
+    },
+    {
+      "epoch": 0.628672339059239,
+      "grad_norm": 1.241674542427063,
+      "learning_rate": 0.00019447202407094316,
+      "loss": 0.7236,
+      "step": 1958
+    },
+    {
+      "epoch": 0.6289934178840906,
+      "grad_norm": 1.0023155212402344,
+      "learning_rate": 0.0001944605732311469,
+      "loss": 0.8875,
+      "step": 1959
+    },
+    {
+      "epoch": 0.629314496708942,
+      "grad_norm": 1.710057020187378,
+      "learning_rate": 0.00019444911088161636,
+      "loss": 1.1843,
+      "step": 1960
+    },
+    {
+      "epoch": 0.6296355755337936,
+      "grad_norm": 1.0903862714767456,
+      "learning_rate": 0.00019443763702374812,
+      "loss": 0.9194,
+      "step": 1961
+    },
+    {
+      "epoch": 0.629956654358645,
+      "grad_norm": 0.8305079340934753,
+      "learning_rate": 0.00019442615165894027,
+      "loss": 0.9536,
+      "step": 1962
+    },
+    {
+      "epoch": 0.6302777331834966,
+      "grad_norm": 13.161450386047363,
+      "learning_rate": 0.00019441465478859228,
+      "loss": 0.9715,
+      "step": 1963
+    },
+    {
+      "epoch": 0.630598812008348,
+      "grad_norm": 1.4573770761489868,
+      "learning_rate": 0.000194403146414105,
+      "loss": 1.056,
+      "step": 1964
+    },
+    {
+      "epoch": 0.6309198908331995,
+      "grad_norm": 2.9237208366394043,
+      "learning_rate": 0.00019439162653688065,
+      "loss": 0.8055,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6312409696580511,
+      "grad_norm": 1.0383076667785645,
+      "learning_rate": 0.00019438009515832297,
+      "loss": 0.938,
+      "step": 1966
+    },
+    {
+      "epoch": 0.6315620484829025,
+      "grad_norm": 1.4683743715286255,
+      "learning_rate": 0.00019436855227983695,
+      "loss": 1.0683,
+      "step": 1967
+    },
+    {
+      "epoch": 0.6318831273077541,
+      "grad_norm": 1.3468014001846313,
+      "learning_rate": 0.00019435699790282908,
+      "loss": 1.1329,
+      "step": 1968
+    },
+    {
+      "epoch": 0.6322042061326055,
+      "grad_norm": 1.5651507377624512,
+      "learning_rate": 0.00019434543202870725,
+      "loss": 0.8679,
+      "step": 1969
+    },
+    {
+      "epoch": 0.6325252849574571,
+      "grad_norm": 1.5266989469528198,
+      "learning_rate": 0.0001943338546588807,
+      "loss": 1.2249,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6328463637823085,
+      "grad_norm": 3.1390273571014404,
+      "learning_rate": 0.0001943222657947601,
+      "loss": 1.2727,
+      "step": 1971
+    },
+    {
+      "epoch": 0.6331674426071601,
+      "grad_norm": 1.983047366142273,
+      "learning_rate": 0.00019431066543775752,
+      "loss": 0.8392,
+      "step": 1972
+    },
+    {
+      "epoch": 0.6334885214320115,
+      "grad_norm": 1.197851538658142,
+      "learning_rate": 0.00019429905358928646,
+      "loss": 1.0931,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6338096002568631,
+      "grad_norm": 1.3977620601654053,
+      "learning_rate": 0.00019428743025076174,
+      "loss": 1.1318,
+      "step": 1974
+    },
+    {
+      "epoch": 0.6341306790817146,
+      "grad_norm": 1.217929720878601,
+      "learning_rate": 0.00019427579542359965,
+      "loss": 1.1208,
+      "step": 1975
+    },
+    {
+      "epoch": 0.634451757906566,
+      "grad_norm": 1.4628900289535522,
+      "learning_rate": 0.00019426414910921787,
+      "loss": 1.0285,
+      "step": 1976
+    },
+    {
+      "epoch": 0.6347728367314176,
+      "grad_norm": 1.353716254234314,
+      "learning_rate": 0.00019425249130903543,
+      "loss": 0.9842,
+      "step": 1977
+    },
+    {
+      "epoch": 0.635093915556269,
+      "grad_norm": 1.2540056705474854,
+      "learning_rate": 0.00019424082202447283,
+      "loss": 0.9581,
+      "step": 1978
+    },
+    {
+      "epoch": 0.6354149943811206,
+      "grad_norm": 1.8885056972503662,
+      "learning_rate": 0.0001942291412569519,
+      "loss": 0.9603,
+      "step": 1979
+    },
+    {
+      "epoch": 0.635736073205972,
+      "grad_norm": 1.4361056089401245,
+      "learning_rate": 0.00019421744900789597,
+      "loss": 0.9598,
+      "step": 1980
+    },
+    {
+      "epoch": 0.6360571520308236,
+      "grad_norm": 0.9510266780853271,
+      "learning_rate": 0.00019420574527872968,
+      "loss": 0.8859,
+      "step": 1981
+    },
+    {
+      "epoch": 0.636378230855675,
+      "grad_norm": 1.5164238214492798,
+      "learning_rate": 0.00019419403007087907,
+      "loss": 1.045,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6366993096805266,
+      "grad_norm": 1.3843255043029785,
+      "learning_rate": 0.0001941823033857716,
+      "loss": 0.9721,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6370203885053781,
+      "grad_norm": 1.692832112312317,
+      "learning_rate": 0.0001941705652248362,
+      "loss": 0.8765,
+      "step": 1984
+    },
+    {
+      "epoch": 0.6373414673302296,
+      "grad_norm": 1.3866074085235596,
+      "learning_rate": 0.00019415881558950302,
+      "loss": 1.0165,
+      "step": 1985
+    },
+    {
+      "epoch": 0.6376625461550811,
+      "grad_norm": 0.9165050983428955,
+      "learning_rate": 0.0001941470544812038,
+      "loss": 1.0291,
+      "step": 1986
+    },
+    {
+      "epoch": 0.6379836249799326,
+      "grad_norm": 0.8897827863693237,
+      "learning_rate": 0.0001941352819013716,
+      "loss": 0.7903,
+      "step": 1987
+    },
+    {
+      "epoch": 0.6383047038047841,
+      "grad_norm": 1.3819725513458252,
+      "learning_rate": 0.00019412349785144078,
+      "loss": 1.0681,
+      "step": 1988
+    },
+    {
+      "epoch": 0.6386257826296355,
+      "grad_norm": 1.1194961071014404,
+      "learning_rate": 0.00019411170233284727,
+      "loss": 0.8758,
+      "step": 1989
+    },
+    {
+      "epoch": 0.6389468614544871,
+      "grad_norm": 0.9944826364517212,
+      "learning_rate": 0.00019409989534702833,
+      "loss": 0.8176,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6392679402793385,
+      "grad_norm": 1.1443473100662231,
+      "learning_rate": 0.00019408807689542257,
+      "loss": 1.0307,
+      "step": 1991
+    },
+    {
+      "epoch": 0.6395890191041901,
+      "grad_norm": 0.8666215538978577,
+      "learning_rate": 0.00019407624697947003,
+      "loss": 0.8433,
+      "step": 1992
+    },
+    {
+      "epoch": 0.6399100979290416,
+      "grad_norm": 1.3199023008346558,
+      "learning_rate": 0.00019406440560061216,
+      "loss": 1.0904,
+      "step": 1993
+    },
+    {
+      "epoch": 0.6402311767538931,
+      "grad_norm": 0.9692159295082092,
+      "learning_rate": 0.0001940525527602918,
+      "loss": 0.8195,
+      "step": 1994
+    },
+    {
+      "epoch": 0.6405522555787446,
+      "grad_norm": 1.1354212760925293,
+      "learning_rate": 0.00019404068845995317,
+      "loss": 0.8017,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6408733344035961,
+      "grad_norm": 1.7543773651123047,
+      "learning_rate": 0.0001940288127010419,
+      "loss": 0.8184,
+      "step": 1996
+    },
+    {
+      "epoch": 0.6411944132284476,
+      "grad_norm": 1.025387167930603,
+      "learning_rate": 0.00019401692548500502,
+      "loss": 0.7647,
+      "step": 1997
+    },
+    {
+      "epoch": 0.641515492053299,
+      "grad_norm": 0.7941446304321289,
+      "learning_rate": 0.00019400502681329098,
+      "loss": 0.7321,
+      "step": 1998
+    },
+    {
+      "epoch": 0.6418365708781506,
+      "grad_norm": 1.0787677764892578,
+      "learning_rate": 0.00019399311668734956,
+      "loss": 0.6183,
+      "step": 1999
+    },
+    {
+      "epoch": 0.642157649703002,
+      "grad_norm": 0.9851073026657104,
+      "learning_rate": 0.00019398119510863197,
+      "loss": 0.6669,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6424787285278536,
+      "grad_norm": 1.7059520483016968,
+      "learning_rate": 0.00019396926207859084,
+      "loss": 1.4499,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6427998073527051,
+      "grad_norm": 1.7769484519958496,
+      "learning_rate": 0.00019395731759868018,
+      "loss": 1.2898,
+      "step": 2002
+    },
+    {
+      "epoch": 0.6431208861775566,
+      "grad_norm": 1.4100226163864136,
+      "learning_rate": 0.00019394536167035534,
+      "loss": 0.8129,
+      "step": 2003
+    },
+    {
+      "epoch": 0.6434419650024081,
+      "grad_norm": 1.3856101036071777,
+      "learning_rate": 0.00019393339429507318,
+      "loss": 0.6282,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6437630438272596,
+      "grad_norm": 1.2996238470077515,
+      "learning_rate": 0.00019392141547429183,
+      "loss": 0.7415,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6440841226521111,
+      "grad_norm": 1.237810730934143,
+      "learning_rate": 0.0001939094252094709,
+      "loss": 0.5929,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6444052014769626,
+      "grad_norm": 1.1693778038024902,
+      "learning_rate": 0.00019389742350207141,
+      "loss": 0.8008,
+      "step": 2007
+    },
+    {
+      "epoch": 0.6447262803018141,
+      "grad_norm": 1.4258911609649658,
+      "learning_rate": 0.00019388541035355564,
+      "loss": 1.2064,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6450473591266656,
+      "grad_norm": 1.4066739082336426,
+      "learning_rate": 0.00019387338576538744,
+      "loss": 1.0647,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6453684379515171,
+      "grad_norm": 1.1812559366226196,
+      "learning_rate": 0.0001938613497390319,
+      "loss": 0.9243,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6456895167763685,
+      "grad_norm": 5.421741962432861,
+      "learning_rate": 0.0001938493022759556,
+      "loss": 0.9313,
+      "step": 2011
+    },
+    {
+      "epoch": 0.6460105956012201,
+      "grad_norm": 1.2386350631713867,
+      "learning_rate": 0.0001938372433776265,
+      "loss": 0.6525,
+      "step": 2012
+    },
+    {
+      "epoch": 0.6463316744260716,
+      "grad_norm": 1.1228787899017334,
+      "learning_rate": 0.00019382517304551396,
+      "loss": 0.9062,
+      "step": 2013
+    },
+    {
+      "epoch": 0.6466527532509231,
+      "grad_norm": 0.9291990399360657,
+      "learning_rate": 0.00019381309128108865,
+      "loss": 1.1112,
+      "step": 2014
+    },
+    {
+      "epoch": 0.6469738320757746,
+      "grad_norm": 0.8409721851348877,
+      "learning_rate": 0.00019380099808582278,
+      "loss": 1.0773,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6472949109006261,
+      "grad_norm": 1.0756123065948486,
+      "learning_rate": 0.0001937888934611898,
+      "loss": 0.9227,
+      "step": 2016
+    },
+    {
+      "epoch": 0.6476159897254776,
+      "grad_norm": 1.1562964916229248,
+      "learning_rate": 0.0001937767774086646,
+      "loss": 0.9476,
+      "step": 2017
+    },
+    {
+      "epoch": 0.6479370685503291,
+      "grad_norm": 1.2570315599441528,
+      "learning_rate": 0.00019376464992972356,
+      "loss": 0.8856,
+      "step": 2018
+    },
+    {
+      "epoch": 0.6482581473751806,
+      "grad_norm": 1.1748182773590088,
+      "learning_rate": 0.0001937525110258444,
+      "loss": 0.8097,
+      "step": 2019
+    },
+    {
+      "epoch": 0.648579226200032,
+      "grad_norm": 0.8975158333778381,
+      "learning_rate": 0.00019374036069850608,
+      "loss": 0.9826,
+      "step": 2020
+    },
+    {
+      "epoch": 0.6489003050248836,
+      "grad_norm": 1.5920442342758179,
+      "learning_rate": 0.00019372819894918915,
+      "loss": 1.1392,
+      "step": 2021
+    },
+    {
+      "epoch": 0.6492213838497352,
+      "grad_norm": 0.9668254256248474,
+      "learning_rate": 0.00019371602577937554,
+      "loss": 0.8112,
+      "step": 2022
+    },
+    {
+      "epoch": 0.6495424626745866,
+      "grad_norm": 1.265324592590332,
+      "learning_rate": 0.0001937038411905484,
+      "loss": 0.9334,
+      "step": 2023
+    },
+    {
+      "epoch": 0.6498635414994381,
+      "grad_norm": 1.3954192399978638,
+      "learning_rate": 0.0001936916451841925,
+      "loss": 0.9908,
+      "step": 2024
+    },
+    {
+      "epoch": 0.6501846203242896,
+      "grad_norm": 1.23148775100708,
+      "learning_rate": 0.0001936794377617938,
+      "loss": 1.0773,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6505056991491411,
+      "grad_norm": 1.0222960710525513,
+      "learning_rate": 0.00019366721892483978,
+      "loss": 0.8985,
+      "step": 2026
+    },
+    {
+      "epoch": 0.6508267779739926,
+      "grad_norm": 0.9243746995925903,
+      "learning_rate": 0.00019365498867481923,
+      "loss": 0.6689,
+      "step": 2027
+    },
+    {
+      "epoch": 0.6511478567988441,
+      "grad_norm": 0.9078730344772339,
+      "learning_rate": 0.00019364274701322244,
+      "loss": 0.7267,
+      "step": 2028
+    },
+    {
+      "epoch": 0.6514689356236956,
+      "grad_norm": 1.5736430883407593,
+      "learning_rate": 0.00019363049394154094,
+      "loss": 0.994,
+      "step": 2029
+    },
+    {
+      "epoch": 0.6517900144485471,
+      "grad_norm": 0.8982104659080505,
+      "learning_rate": 0.0001936182294612678,
+      "loss": 0.838,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6521110932733987,
+      "grad_norm": 1.2022709846496582,
+      "learning_rate": 0.00019360595357389735,
+      "loss": 1.2414,
+      "step": 2031
+    },
+    {
+      "epoch": 0.6524321720982501,
+      "grad_norm": 0.9198827147483826,
+      "learning_rate": 0.00019359366628092539,
+      "loss": 0.7564,
+      "step": 2032
+    },
+    {
+      "epoch": 0.6527532509231017,
+      "grad_norm": 1.3141475915908813,
+      "learning_rate": 0.00019358136758384912,
+      "loss": 1.3289,
+      "step": 2033
+    },
+    {
+      "epoch": 0.6530743297479531,
+      "grad_norm": 1.4782156944274902,
+      "learning_rate": 0.00019356905748416702,
+      "loss": 0.9758,
+      "step": 2034
+    },
+    {
+      "epoch": 0.6533954085728046,
+      "grad_norm": 1.2020496129989624,
+      "learning_rate": 0.00019355673598337914,
+      "loss": 0.8308,
+      "step": 2035
+    },
+    {
+      "epoch": 0.6537164873976561,
+      "grad_norm": 1.2560274600982666,
+      "learning_rate": 0.00019354440308298675,
+      "loss": 0.7407,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6540375662225076,
+      "grad_norm": 0.8934011459350586,
+      "learning_rate": 0.00019353205878449258,
+      "loss": 0.865,
+      "step": 2037
+    },
+    {
+      "epoch": 0.6543586450473591,
+      "grad_norm": 0.9289383888244629,
+      "learning_rate": 0.0001935197030894008,
+      "loss": 0.7145,
+      "step": 2038
+    },
+    {
+      "epoch": 0.6546797238722106,
+      "grad_norm": 1.2418136596679688,
+      "learning_rate": 0.00019350733599921683,
+      "loss": 0.8111,
+      "step": 2039
+    },
+    {
+      "epoch": 0.6550008026970622,
+      "grad_norm": 0.9828119277954102,
+      "learning_rate": 0.00019349495751544763,
+      "loss": 0.9877,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6553218815219136,
+      "grad_norm": 0.8384512662887573,
+      "learning_rate": 0.00019348256763960145,
+      "loss": 0.7265,
+      "step": 2041
+    },
+    {
+      "epoch": 0.6556429603467652,
+      "grad_norm": 1.1183538436889648,
+      "learning_rate": 0.000193470166373188,
+      "loss": 0.8691,
+      "step": 2042
+    },
+    {
+      "epoch": 0.6559640391716166,
+      "grad_norm": 1.0585153102874756,
+      "learning_rate": 0.00019345775371771824,
+      "loss": 0.9087,
+      "step": 2043
+    },
+    {
+      "epoch": 0.6562851179964682,
+      "grad_norm": 1.3051196336746216,
+      "learning_rate": 0.0001934453296747047,
+      "loss": 0.8324,
+      "step": 2044
+    },
+    {
+      "epoch": 0.6566061968213196,
+      "grad_norm": 1.0037448406219482,
+      "learning_rate": 0.00019343289424566122,
+      "loss": 0.7462,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6569272756461711,
+      "grad_norm": 1.1593952178955078,
+      "learning_rate": 0.00019342044743210295,
+      "loss": 0.8161,
+      "step": 2046
+    },
+    {
+      "epoch": 0.6572483544710226,
+      "grad_norm": 0.9699090123176575,
+      "learning_rate": 0.00019340798923554657,
+      "loss": 0.8817,
+      "step": 2047
+    },
+    {
+      "epoch": 0.6575694332958741,
+      "grad_norm": 2.443246841430664,
+      "learning_rate": 0.00019339551965751002,
+      "loss": 0.888,
+      "step": 2048
+    },
+    {
+      "epoch": 0.6578905121207257,
+      "grad_norm": 0.9894551634788513,
+      "learning_rate": 0.00019338303869951269,
+      "loss": 0.6925,
+      "step": 2049
+    },
+    {
+      "epoch": 0.6582115909455771,
+      "grad_norm": 1.5795120000839233,
+      "learning_rate": 0.00019337054636307536,
+      "loss": 0.6042,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6585326697704287,
+      "grad_norm": 1.1168063879013062,
+      "learning_rate": 0.00019335804264972018,
+      "loss": 1.449,
+      "step": 2051
+    },
+    {
+      "epoch": 0.6588537485952801,
+      "grad_norm": 1.1965980529785156,
+      "learning_rate": 0.0001933455275609707,
+      "loss": 1.1259,
+      "step": 2052
+    },
+    {
+      "epoch": 0.6591748274201317,
+      "grad_norm": 1.1799712181091309,
+      "learning_rate": 0.0001933330010983518,
+      "loss": 1.0329,
+      "step": 2053
+    },
+    {
+      "epoch": 0.6594959062449831,
+      "grad_norm": 1.4881231784820557,
+      "learning_rate": 0.00019332046326338986,
+      "loss": 0.8752,
+      "step": 2054
+    },
+    {
+      "epoch": 0.6598169850698347,
+      "grad_norm": 1.203697919845581,
+      "learning_rate": 0.00019330791405761252,
+      "loss": 0.7875,
+      "step": 2055
+    },
+    {
+      "epoch": 0.6601380638946861,
+      "grad_norm": 1.563508152961731,
+      "learning_rate": 0.00019329535348254893,
+      "loss": 0.907,
+      "step": 2056
+    },
+    {
+      "epoch": 0.6604591427195377,
+      "grad_norm": 1.3101892471313477,
+      "learning_rate": 0.00019328278153972947,
+      "loss": 0.9096,
+      "step": 2057
+    },
+    {
+      "epoch": 0.6607802215443892,
+      "grad_norm": 1.4064226150512695,
+      "learning_rate": 0.00019327019823068604,
+      "loss": 1.0788,
+      "step": 2058
+    },
+    {
+      "epoch": 0.6611013003692406,
+      "grad_norm": 1.2184844017028809,
+      "learning_rate": 0.00019325760355695188,
+      "loss": 0.9344,
+      "step": 2059
+    },
+    {
+      "epoch": 0.6614223791940922,
+      "grad_norm": 1.2525115013122559,
+      "learning_rate": 0.0001932449975200616,
+      "loss": 0.9809,
+      "step": 2060
+    },
+    {
+      "epoch": 0.6617434580189436,
+      "grad_norm": 1.2751420736312866,
+      "learning_rate": 0.00019323238012155123,
+      "loss": 0.8502,
+      "step": 2061
+    },
+    {
+      "epoch": 0.6620645368437952,
+      "grad_norm": 1.271164059638977,
+      "learning_rate": 0.00019321975136295813,
+      "loss": 0.9252,
+      "step": 2062
+    },
+    {
+      "epoch": 0.6623856156686466,
+      "grad_norm": 0.9167636632919312,
+      "learning_rate": 0.0001932071112458211,
+      "loss": 0.9382,
+      "step": 2063
+    },
+    {
+      "epoch": 0.6627066944934982,
+      "grad_norm": 1.027632474899292,
+      "learning_rate": 0.0001931944597716803,
+      "loss": 0.941,
+      "step": 2064
+    },
+    {
+      "epoch": 0.6630277733183496,
+      "grad_norm": 0.8215566873550415,
+      "learning_rate": 0.00019318179694207725,
+      "loss": 0.8984,
+      "step": 2065
+    },
+    {
+      "epoch": 0.6633488521432012,
+      "grad_norm": 1.2999035120010376,
+      "learning_rate": 0.0001931691227585549,
+      "loss": 0.8938,
+      "step": 2066
+    },
+    {
+      "epoch": 0.6636699309680526,
+      "grad_norm": 0.8892626762390137,
+      "learning_rate": 0.00019315643722265757,
+      "loss": 0.711,
+      "step": 2067
+    },
+    {
+      "epoch": 0.6639910097929042,
+      "grad_norm": 1.4477237462997437,
+      "learning_rate": 0.0001931437403359309,
+      "loss": 1.0521,
+      "step": 2068
+    },
+    {
+      "epoch": 0.6643120886177557,
+      "grad_norm": 1.3851264715194702,
+      "learning_rate": 0.00019313103209992204,
+      "loss": 1.0395,
+      "step": 2069
+    },
+    {
+      "epoch": 0.6646331674426071,
+      "grad_norm": 1.1687705516815186,
+      "learning_rate": 0.0001931183125161794,
+      "loss": 0.8819,
+      "step": 2070
+    },
+    {
+      "epoch": 0.6649542462674587,
+      "grad_norm": 1.4307971000671387,
+      "learning_rate": 0.00019310558158625285,
+      "loss": 1.012,
+      "step": 2071
+    },
+    {
+      "epoch": 0.6652753250923101,
+      "grad_norm": 1.1840988397598267,
+      "learning_rate": 0.00019309283931169356,
+      "loss": 1.034,
+      "step": 2072
+    },
+    {
+      "epoch": 0.6655964039171617,
+      "grad_norm": 1.1295652389526367,
+      "learning_rate": 0.00019308008569405422,
+      "loss": 0.968,
+      "step": 2073
+    },
+    {
+      "epoch": 0.6659174827420131,
+      "grad_norm": 1.238215684890747,
+      "learning_rate": 0.0001930673207348888,
+      "loss": 0.9026,
+      "step": 2074
+    },
+    {
+      "epoch": 0.6662385615668647,
+      "grad_norm": 1.0732163190841675,
+      "learning_rate": 0.00019305454443575262,
+      "loss": 1.0993,
+      "step": 2075
+    },
+    {
+      "epoch": 0.6665596403917161,
+      "grad_norm": 1.2432470321655273,
+      "learning_rate": 0.00019304175679820247,
+      "loss": 1.2107,
+      "step": 2076
+    },
+    {
+      "epoch": 0.6668807192165677,
+      "grad_norm": 1.0395501852035522,
+      "learning_rate": 0.0001930289578237965,
+      "loss": 0.8751,
+      "step": 2077
+    },
+    {
+      "epoch": 0.6672017980414192,
+      "grad_norm": 1.7850005626678467,
+      "learning_rate": 0.00019301614751409416,
+      "loss": 0.9662,
+      "step": 2078
+    },
+    {
+      "epoch": 0.6675228768662707,
+      "grad_norm": 1.5475821495056152,
+      "learning_rate": 0.0001930033258706564,
+      "loss": 1.056,
+      "step": 2079
+    },
+    {
+      "epoch": 0.6678439556911222,
+      "grad_norm": 0.9385414123535156,
+      "learning_rate": 0.00019299049289504553,
+      "loss": 0.8348,
+      "step": 2080
+    },
+    {
+      "epoch": 0.6681650345159736,
+      "grad_norm": 1.2614946365356445,
+      "learning_rate": 0.00019297764858882514,
+      "loss": 0.8467,
+      "step": 2081
+    },
+    {
+      "epoch": 0.6684861133408252,
+      "grad_norm": 1.07332444190979,
+      "learning_rate": 0.00019296479295356035,
+      "loss": 0.9718,
+      "step": 2082
+    },
+    {
+      "epoch": 0.6688071921656766,
+      "grad_norm": 2.2811384201049805,
+      "learning_rate": 0.00019295192599081746,
+      "loss": 1.1172,
+      "step": 2083
+    },
+    {
+      "epoch": 0.6691282709905282,
+      "grad_norm": 0.9308247566223145,
+      "learning_rate": 0.00019293904770216437,
+      "loss": 0.878,
+      "step": 2084
+    },
+    {
+      "epoch": 0.6694493498153796,
+      "grad_norm": 1.3048980236053467,
+      "learning_rate": 0.00019292615808917026,
+      "loss": 0.9988,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6697704286402312,
+      "grad_norm": 2.0807573795318604,
+      "learning_rate": 0.00019291325715340563,
+      "loss": 0.9486,
+      "step": 2086
+    },
+    {
+      "epoch": 0.6700915074650827,
+      "grad_norm": 1.4993385076522827,
+      "learning_rate": 0.00019290034489644246,
+      "loss": 0.9338,
+      "step": 2087
+    },
+    {
+      "epoch": 0.6704125862899342,
+      "grad_norm": 1.1639915704727173,
+      "learning_rate": 0.00019288742131985407,
+      "loss": 0.9569,
+      "step": 2088
+    },
+    {
+      "epoch": 0.6707336651147857,
+      "grad_norm": 1.270796775817871,
+      "learning_rate": 0.00019287448642521513,
+      "loss": 0.904,
+      "step": 2089
+    },
+    {
+      "epoch": 0.6710547439396372,
+      "grad_norm": 1.7701547145843506,
+      "learning_rate": 0.00019286154021410173,
+      "loss": 1.0951,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6713758227644887,
+      "grad_norm": 1.2145819664001465,
+      "learning_rate": 0.00019284858268809137,
+      "loss": 0.7358,
+      "step": 2091
+    },
+    {
+      "epoch": 0.6716969015893401,
+      "grad_norm": 1.4250010251998901,
+      "learning_rate": 0.00019283561384876284,
+      "loss": 1.0274,
+      "step": 2092
+    },
+    {
+      "epoch": 0.6720179804141917,
+      "grad_norm": 1.232686996459961,
+      "learning_rate": 0.00019282263369769633,
+      "loss": 1.0309,
+      "step": 2093
+    },
+    {
+      "epoch": 0.6723390592390431,
+      "grad_norm": 1.1875118017196655,
+      "learning_rate": 0.00019280964223647348,
+      "loss": 0.789,
+      "step": 2094
+    },
+    {
+      "epoch": 0.6726601380638947,
+      "grad_norm": 1.035929560661316,
+      "learning_rate": 0.00019279663946667727,
+      "loss": 0.8858,
+      "step": 2095
+    },
+    {
+      "epoch": 0.6729812168887462,
+      "grad_norm": 1.323173999786377,
+      "learning_rate": 0.000192783625389892,
+      "loss": 1.0185,
+      "step": 2096
+    },
+    {
+      "epoch": 0.6733022957135977,
+      "grad_norm": 1.4377621412277222,
+      "learning_rate": 0.00019277060000770342,
+      "loss": 0.6274,
+      "step": 2097
+    },
+    {
+      "epoch": 0.6736233745384492,
+      "grad_norm": 0.6943148374557495,
+      "learning_rate": 0.00019275756332169867,
+      "loss": 0.5938,
+      "step": 2098
+    },
+    {
+      "epoch": 0.6739444533633007,
+      "grad_norm": 1.0975568294525146,
+      "learning_rate": 0.00019274451533346615,
+      "loss": 0.5224,
+      "step": 2099
+    },
+    {
+      "epoch": 0.6742655321881522,
+      "grad_norm": 0.7381674647331238,
+      "learning_rate": 0.00019273145604459577,
+      "loss": 0.5018,
+      "step": 2100
+    },
+    {
+      "epoch": 0.6745866110130037,
+      "grad_norm": 1.5068893432617188,
+      "learning_rate": 0.00019271838545667876,
+      "loss": 1.3901,
+      "step": 2101
+    },
+    {
+      "epoch": 0.6749076898378552,
+      "grad_norm": 4.2163472175598145,
+      "learning_rate": 0.0001927053035713077,
+      "loss": 0.9407,
+      "step": 2102
+    },
+    {
+      "epoch": 0.6752287686627066,
+      "grad_norm": 1.432925820350647,
+      "learning_rate": 0.00019269221039007665,
+      "loss": 0.9272,
+      "step": 2103
+    },
+    {
+      "epoch": 0.6755498474875582,
+      "grad_norm": 1.374459981918335,
+      "learning_rate": 0.0001926791059145809,
+      "loss": 0.7463,
+      "step": 2104
+    },
+    {
+      "epoch": 0.6758709263124097,
+      "grad_norm": 1.4353713989257812,
+      "learning_rate": 0.0001926659901464172,
+      "loss": 0.8202,
+      "step": 2105
+    },
+    {
+      "epoch": 0.6761920051372612,
+      "grad_norm": 1.1444305181503296,
+      "learning_rate": 0.00019265286308718372,
+      "loss": 0.7957,
+      "step": 2106
+    },
+    {
+      "epoch": 0.6765130839621127,
+      "grad_norm": 1.4592642784118652,
+      "learning_rate": 0.00019263972473847993,
+      "loss": 0.6475,
+      "step": 2107
+    },
+    {
+      "epoch": 0.6768341627869642,
+      "grad_norm": 1.0431547164916992,
+      "learning_rate": 0.00019262657510190666,
+      "loss": 0.8323,
+      "step": 2108
+    },
+    {
+      "epoch": 0.6771552416118157,
+      "grad_norm": 1.0479865074157715,
+      "learning_rate": 0.00019261341417906621,
+      "loss": 0.9267,
+      "step": 2109
+    },
+    {
+      "epoch": 0.6774763204366672,
+      "grad_norm": 1.3619253635406494,
+      "learning_rate": 0.00019260024197156214,
+      "loss": 1.0108,
+      "step": 2110
+    },
+    {
+      "epoch": 0.6777973992615187,
+      "grad_norm": 1.088303565979004,
+      "learning_rate": 0.0001925870584809995,
+      "loss": 0.9458,
+      "step": 2111
+    },
+    {
+      "epoch": 0.6781184780863702,
+      "grad_norm": 1.064591646194458,
+      "learning_rate": 0.00019257386370898457,
+      "loss": 0.8121,
+      "step": 2112
+    },
+    {
+      "epoch": 0.6784395569112217,
+      "grad_norm": 0.9795963764190674,
+      "learning_rate": 0.00019256065765712522,
+      "loss": 1.006,
+      "step": 2113
+    },
+    {
+      "epoch": 0.6787606357360733,
+      "grad_norm": 1.218254804611206,
+      "learning_rate": 0.00019254744032703048,
+      "loss": 0.866,
+      "step": 2114
+    },
+    {
+      "epoch": 0.6790817145609247,
+      "grad_norm": 0.9186314940452576,
+      "learning_rate": 0.00019253421172031086,
+      "loss": 0.922,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6794027933857762,
+      "grad_norm": 1.187469244003296,
+      "learning_rate": 0.00019252097183857823,
+      "loss": 0.7671,
+      "step": 2116
+    },
+    {
+      "epoch": 0.6797238722106277,
+      "grad_norm": 1.1949058771133423,
+      "learning_rate": 0.0001925077206834458,
+      "loss": 0.9525,
+      "step": 2117
+    },
+    {
+      "epoch": 0.6800449510354792,
+      "grad_norm": 1.0139573812484741,
+      "learning_rate": 0.00019249445825652824,
+      "loss": 0.8317,
+      "step": 2118
+    },
+    {
+      "epoch": 0.6803660298603307,
+      "grad_norm": 1.2550780773162842,
+      "learning_rate": 0.0001924811845594415,
+      "loss": 1.0096,
+      "step": 2119
+    },
+    {
+      "epoch": 0.6806871086851822,
+      "grad_norm": 1.0419883728027344,
+      "learning_rate": 0.00019246789959380295,
+      "loss": 0.9382,
+      "step": 2120
+    },
+    {
+      "epoch": 0.6810081875100337,
+      "grad_norm": 0.8842172622680664,
+      "learning_rate": 0.00019245460336123134,
+      "loss": 1.0352,
+      "step": 2121
+    },
+    {
+      "epoch": 0.6813292663348852,
+      "grad_norm": 1.3533904552459717,
+      "learning_rate": 0.00019244129586334672,
+      "loss": 1.1023,
+      "step": 2122
+    },
+    {
+      "epoch": 0.6816503451597368,
+      "grad_norm": 1.0904252529144287,
+      "learning_rate": 0.0001924279771017706,
+      "loss": 1.0112,
+      "step": 2123
+    },
+    {
+      "epoch": 0.6819714239845882,
+      "grad_norm": 1.2503504753112793,
+      "learning_rate": 0.00019241464707812585,
+      "loss": 0.8125,
+      "step": 2124
+    },
+    {
+      "epoch": 0.6822925028094398,
+      "grad_norm": 1.4106786251068115,
+      "learning_rate": 0.0001924013057940367,
+      "loss": 0.9185,
+      "step": 2125
+    },
+    {
+      "epoch": 0.6826135816342912,
+      "grad_norm": 1.1263327598571777,
+      "learning_rate": 0.0001923879532511287,
+      "loss": 0.9063,
+      "step": 2126
+    },
+    {
+      "epoch": 0.6829346604591428,
+      "grad_norm": 2.7052721977233887,
+      "learning_rate": 0.00019237458945102882,
+      "loss": 0.9509,
+      "step": 2127
+    },
+    {
+      "epoch": 0.6832557392839942,
+      "grad_norm": 1.1038817167282104,
+      "learning_rate": 0.00019236121439536542,
+      "loss": 0.9669,
+      "step": 2128
+    },
+    {
+      "epoch": 0.6835768181088457,
+      "grad_norm": 1.2589917182922363,
+      "learning_rate": 0.00019234782808576824,
+      "loss": 0.9313,
+      "step": 2129
+    },
+    {
+      "epoch": 0.6838978969336972,
+      "grad_norm": 1.6602294445037842,
+      "learning_rate": 0.0001923344305238683,
+      "loss": 1.0642,
+      "step": 2130
+    },
+    {
+      "epoch": 0.6842189757585487,
+      "grad_norm": 0.957186758518219,
+      "learning_rate": 0.00019232102171129811,
+      "loss": 0.7523,
+      "step": 2131
+    },
+    {
+      "epoch": 0.6845400545834002,
+      "grad_norm": 0.8682358264923096,
+      "learning_rate": 0.00019230760164969143,
+      "loss": 0.8184,
+      "step": 2132
+    },
+    {
+      "epoch": 0.6848611334082517,
+      "grad_norm": 1.2894452810287476,
+      "learning_rate": 0.0001922941703406835,
+      "loss": 0.7959,
+      "step": 2133
+    },
+    {
+      "epoch": 0.6851822122331033,
+      "grad_norm": 0.9780199527740479,
+      "learning_rate": 0.00019228072778591088,
+      "loss": 0.865,
+      "step": 2134
+    },
+    {
+      "epoch": 0.6855032910579547,
+      "grad_norm": 1.7117539644241333,
+      "learning_rate": 0.0001922672739870115,
+      "loss": 0.9489,
+      "step": 2135
+    },
+    {
+      "epoch": 0.6858243698828063,
+      "grad_norm": 1.3718947172164917,
+      "learning_rate": 0.00019225380894562463,
+      "loss": 1.0358,
+      "step": 2136
+    },
+    {
+      "epoch": 0.6861454487076577,
+      "grad_norm": 1.2108441591262817,
+      "learning_rate": 0.00019224033266339102,
+      "loss": 1.023,
+      "step": 2137
+    },
+    {
+      "epoch": 0.6864665275325093,
+      "grad_norm": 1.9689425230026245,
+      "learning_rate": 0.00019222684514195264,
+      "loss": 0.8403,
+      "step": 2138
+    },
+    {
+      "epoch": 0.6867876063573607,
+      "grad_norm": 1.626488208770752,
+      "learning_rate": 0.00019221334638295294,
+      "loss": 0.889,
+      "step": 2139
+    },
+    {
+      "epoch": 0.6871086851822122,
+      "grad_norm": 1.0145803689956665,
+      "learning_rate": 0.0001921998363880367,
+      "loss": 0.9659,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6874297640070637,
+      "grad_norm": 2.1822826862335205,
+      "learning_rate": 0.00019218631515885006,
+      "loss": 0.8542,
+      "step": 2141
+    },
+    {
+      "epoch": 0.6877508428319152,
+      "grad_norm": 0.9943452477455139,
+      "learning_rate": 0.00019217278269704056,
+      "loss": 0.8499,
+      "step": 2142
+    },
+    {
+      "epoch": 0.6880719216567668,
+      "grad_norm": 0.675469696521759,
+      "learning_rate": 0.00019215923900425707,
+      "loss": 0.6101,
+      "step": 2143
+    },
+    {
+      "epoch": 0.6883930004816182,
+      "grad_norm": 1.052608609199524,
+      "learning_rate": 0.00019214568408214985,
+      "loss": 0.8342,
+      "step": 2144
+    },
+    {
+      "epoch": 0.6887140793064698,
+      "grad_norm": 0.9476191997528076,
+      "learning_rate": 0.00019213211793237057,
+      "loss": 0.8063,
+      "step": 2145
+    },
+    {
+      "epoch": 0.6890351581313212,
+      "grad_norm": 1.0476568937301636,
+      "learning_rate": 0.00019211854055657215,
+      "loss": 0.6778,
+      "step": 2146
+    },
+    {
+      "epoch": 0.6893562369561728,
+      "grad_norm": 1.4577487707138062,
+      "learning_rate": 0.00019210495195640895,
+      "loss": 0.7882,
+      "step": 2147
+    },
+    {
+      "epoch": 0.6896773157810242,
+      "grad_norm": 1.045127034187317,
+      "learning_rate": 0.0001920913521335368,
+      "loss": 0.6774,
+      "step": 2148
+    },
+    {
+      "epoch": 0.6899983946058758,
+      "grad_norm": 0.8924066424369812,
+      "learning_rate": 0.00019207774108961272,
+      "loss": 0.764,
+      "step": 2149
+    },
+    {
+      "epoch": 0.6903194734307272,
+      "grad_norm": 1.1722183227539062,
+      "learning_rate": 0.00019206411882629517,
+      "loss": 0.7897,
+      "step": 2150
+    },
+    {
+      "epoch": 0.6906405522555787,
+      "grad_norm": 1.1900309324264526,
+      "learning_rate": 0.00019205048534524406,
+      "loss": 1.2467,
+      "step": 2151
+    },
+    {
+      "epoch": 0.6909616310804303,
+      "grad_norm": 1.3887277841567993,
+      "learning_rate": 0.00019203684064812045,
+      "loss": 1.2293,
+      "step": 2152
+    },
+    {
+      "epoch": 0.6912827099052817,
+      "grad_norm": 1.1636091470718384,
+      "learning_rate": 0.00019202318473658705,
+      "loss": 0.8659,
+      "step": 2153
+    },
+    {
+      "epoch": 0.6916037887301333,
+      "grad_norm": 1.2683323621749878,
+      "learning_rate": 0.0001920095176123077,
+      "loss": 0.7881,
+      "step": 2154
+    },
+    {
+      "epoch": 0.6919248675549847,
+      "grad_norm": 1.227537989616394,
+      "learning_rate": 0.00019199583927694772,
+      "loss": 0.841,
+      "step": 2155
+    },
+    {
+      "epoch": 0.6922459463798363,
+      "grad_norm": 1.9995452165603638,
+      "learning_rate": 0.00019198214973217378,
+      "loss": 0.8185,
+      "step": 2156
+    },
+    {
+      "epoch": 0.6925670252046877,
+      "grad_norm": 1.2101178169250488,
+      "learning_rate": 0.00019196844897965393,
+      "loss": 0.5357,
+      "step": 2157
+    },
+    {
+      "epoch": 0.6928881040295393,
+      "grad_norm": 0.990922749042511,
+      "learning_rate": 0.00019195473702105748,
+      "loss": 0.8727,
+      "step": 2158
+    },
+    {
+      "epoch": 0.6932091828543907,
+      "grad_norm": 1.5191935300827026,
+      "learning_rate": 0.0001919410138580553,
+      "loss": 1.1395,
+      "step": 2159
+    },
+    {
+      "epoch": 0.6935302616792423,
+      "grad_norm": 1.074205994606018,
+      "learning_rate": 0.00019192727949231945,
+      "loss": 1.1821,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6938513405040938,
+      "grad_norm": 1.0748862028121948,
+      "learning_rate": 0.00019191353392552344,
+      "loss": 0.9226,
+      "step": 2161
+    },
+    {
+      "epoch": 0.6941724193289452,
+      "grad_norm": 1.0452769994735718,
+      "learning_rate": 0.00019189977715934213,
+      "loss": 0.8429,
+      "step": 2162
+    },
+    {
+      "epoch": 0.6944934981537968,
+      "grad_norm": 1.0717366933822632,
+      "learning_rate": 0.00019188600919545174,
+      "loss": 0.8875,
+      "step": 2163
+    },
+    {
+      "epoch": 0.6948145769786482,
+      "grad_norm": 0.7801916599273682,
+      "learning_rate": 0.00019187223003552985,
+      "loss": 0.8721,
+      "step": 2164
+    },
+    {
+      "epoch": 0.6951356558034998,
+      "grad_norm": 1.0750560760498047,
+      "learning_rate": 0.0001918584396812554,
+      "loss": 1.103,
+      "step": 2165
+    },
+    {
+      "epoch": 0.6954567346283512,
+      "grad_norm": 1.0144506692886353,
+      "learning_rate": 0.00019184463813430873,
+      "loss": 1.0441,
+      "step": 2166
+    },
+    {
+      "epoch": 0.6957778134532028,
+      "grad_norm": 1.0422799587249756,
+      "learning_rate": 0.00019183082539637146,
+      "loss": 0.8684,
+      "step": 2167
+    },
+    {
+      "epoch": 0.6960988922780542,
+      "grad_norm": 1.1190441846847534,
+      "learning_rate": 0.0001918170014691267,
+      "loss": 0.8669,
+      "step": 2168
+    },
+    {
+      "epoch": 0.6964199711029058,
+      "grad_norm": 1.2634443044662476,
+      "learning_rate": 0.0001918031663542588,
+      "loss": 0.8319,
+      "step": 2169
+    },
+    {
+      "epoch": 0.6967410499277573,
+      "grad_norm": 1.256981611251831,
+      "learning_rate": 0.0001917893200534536,
+      "loss": 0.9238,
+      "step": 2170
+    },
+    {
+      "epoch": 0.6970621287526088,
+      "grad_norm": 1.5736911296844482,
+      "learning_rate": 0.00019177546256839812,
+      "loss": 1.1809,
+      "step": 2171
+    },
+    {
+      "epoch": 0.6973832075774603,
+      "grad_norm": 0.890072762966156,
+      "learning_rate": 0.00019176159390078094,
+      "loss": 0.8873,
+      "step": 2172
+    },
+    {
+      "epoch": 0.6977042864023117,
+      "grad_norm": 0.9746190309524536,
+      "learning_rate": 0.00019174771405229186,
+      "loss": 0.8685,
+      "step": 2173
+    },
+    {
+      "epoch": 0.6980253652271633,
+      "grad_norm": 1.4396299123764038,
+      "learning_rate": 0.00019173382302462214,
+      "loss": 0.9388,
+      "step": 2174
+    },
+    {
+      "epoch": 0.6983464440520147,
+      "grad_norm": 1.3588879108428955,
+      "learning_rate": 0.00019171992081946435,
+      "loss": 1.0749,
+      "step": 2175
+    },
+    {
+      "epoch": 0.6986675228768663,
+      "grad_norm": 2.2067346572875977,
+      "learning_rate": 0.0001917060074385124,
+      "loss": 1.0014,
+      "step": 2176
+    },
+    {
+      "epoch": 0.6989886017017177,
+      "grad_norm": 1.2330677509307861,
+      "learning_rate": 0.00019169208288346166,
+      "loss": 1.1379,
+      "step": 2177
+    },
+    {
+      "epoch": 0.6993096805265693,
+      "grad_norm": 1.0044240951538086,
+      "learning_rate": 0.0001916781471560087,
+      "loss": 0.9738,
+      "step": 2178
+    },
+    {
+      "epoch": 0.6996307593514208,
+      "grad_norm": 1.3933671712875366,
+      "learning_rate": 0.00019166420025785164,
+      "loss": 1.0522,
+      "step": 2179
+    },
+    {
+      "epoch": 0.6999518381762723,
+      "grad_norm": 0.9893519282341003,
+      "learning_rate": 0.0001916502421906898,
+      "loss": 0.8907,
+      "step": 2180
+    },
+    {
+      "epoch": 0.7002729170011238,
+      "grad_norm": 1.0318580865859985,
+      "learning_rate": 0.00019163627295622397,
+      "loss": 0.9829,
+      "step": 2181
+    },
+    {
+      "epoch": 0.7005939958259753,
+      "grad_norm": 1.4720815420150757,
+      "learning_rate": 0.00019162229255615624,
+      "loss": 0.9374,
+      "step": 2182
+    },
+    {
+      "epoch": 0.7009150746508268,
+      "grad_norm": 1.006734013557434,
+      "learning_rate": 0.00019160830099219006,
+      "loss": 1.04,
+      "step": 2183
+    },
+    {
+      "epoch": 0.7012361534756782,
+      "grad_norm": 2.0803046226501465,
+      "learning_rate": 0.00019159429826603032,
+      "loss": 1.1813,
+      "step": 2184
+    },
+    {
+      "epoch": 0.7015572323005298,
+      "grad_norm": 1.770952820777893,
+      "learning_rate": 0.00019158028437938317,
+      "loss": 1.0018,
+      "step": 2185
+    },
+    {
+      "epoch": 0.7018783111253812,
+      "grad_norm": 1.0838485956192017,
+      "learning_rate": 0.00019156625933395614,
+      "loss": 1.0851,
+      "step": 2186
+    },
+    {
+      "epoch": 0.7021993899502328,
+      "grad_norm": 1.384687066078186,
+      "learning_rate": 0.00019155222313145816,
+      "loss": 0.9151,
+      "step": 2187
+    },
+    {
+      "epoch": 0.7025204687750842,
+      "grad_norm": 0.8908778429031372,
+      "learning_rate": 0.0001915381757735995,
+      "loss": 0.9169,
+      "step": 2188
+    },
+    {
+      "epoch": 0.7028415475999358,
+      "grad_norm": 0.814580500125885,
+      "learning_rate": 0.00019152411726209176,
+      "loss": 0.5335,
+      "step": 2189
+    },
+    {
+      "epoch": 0.7031626264247873,
+      "grad_norm": 1.491808533668518,
+      "learning_rate": 0.000191510047598648,
+      "loss": 0.9649,
+      "step": 2190
+    },
+    {
+      "epoch": 0.7034837052496388,
+      "grad_norm": 1.308821678161621,
+      "learning_rate": 0.0001914959667849825,
+      "loss": 0.9609,
+      "step": 2191
+    },
+    {
+      "epoch": 0.7038047840744903,
+      "grad_norm": 1.6235363483428955,
+      "learning_rate": 0.00019148187482281097,
+      "loss": 0.8551,
+      "step": 2192
+    },
+    {
+      "epoch": 0.7041258628993418,
+      "grad_norm": 1.9721722602844238,
+      "learning_rate": 0.0001914677717138505,
+      "loss": 0.8588,
+      "step": 2193
+    },
+    {
+      "epoch": 0.7044469417241933,
+      "grad_norm": 1.485297679901123,
+      "learning_rate": 0.00019145365745981948,
+      "loss": 0.7772,
+      "step": 2194
+    },
+    {
+      "epoch": 0.7047680205490447,
+      "grad_norm": 1.1313917636871338,
+      "learning_rate": 0.00019143953206243776,
+      "loss": 0.8873,
+      "step": 2195
+    },
+    {
+      "epoch": 0.7050890993738963,
+      "grad_norm": 1.1025351285934448,
+      "learning_rate": 0.00019142539552342638,
+      "loss": 0.8683,
+      "step": 2196
+    },
+    {
+      "epoch": 0.7054101781987477,
+      "grad_norm": 1.7644602060317993,
+      "learning_rate": 0.0001914112478445079,
+      "loss": 0.6767,
+      "step": 2197
+    },
+    {
+      "epoch": 0.7057312570235993,
+      "grad_norm": 0.9576209783554077,
+      "learning_rate": 0.00019139708902740613,
+      "loss": 0.6193,
+      "step": 2198
+    },
+    {
+      "epoch": 0.7060523358484508,
+      "grad_norm": 1.4080398082733154,
+      "learning_rate": 0.0001913829190738463,
+      "loss": 0.7909,
+      "step": 2199
+    },
+    {
+      "epoch": 0.7063734146733023,
+      "grad_norm": 1.04661226272583,
+      "learning_rate": 0.000191368737985555,
+      "loss": 0.6024,
+      "step": 2200
+    },
+    {
+      "epoch": 0.7066944934981538,
+      "grad_norm": 1.448823094367981,
+      "learning_rate": 0.0001913545457642601,
+      "loss": 1.4655,
+      "step": 2201
+    },
+    {
+      "epoch": 0.7070155723230053,
+      "grad_norm": 1.4213786125183105,
+      "learning_rate": 0.0001913403424116909,
+      "loss": 1.2999,
+      "step": 2202
+    },
+    {
+      "epoch": 0.7073366511478568,
+      "grad_norm": 1.3486974239349365,
+      "learning_rate": 0.00019132612792957808,
+      "loss": 0.946,
+      "step": 2203
+    },
+    {
+      "epoch": 0.7076577299727083,
+      "grad_norm": 1.5810649394989014,
+      "learning_rate": 0.00019131190231965356,
+      "loss": 0.8893,
+      "step": 2204
+    },
+    {
+      "epoch": 0.7079788087975598,
+      "grad_norm": 1.3420324325561523,
+      "learning_rate": 0.00019129766558365074,
+      "loss": 0.7434,
+      "step": 2205
+    },
+    {
+      "epoch": 0.7082998876224113,
+      "grad_norm": 1.3447771072387695,
+      "learning_rate": 0.0001912834177233043,
+      "loss": 0.8466,
+      "step": 2206
+    },
+    {
+      "epoch": 0.7086209664472628,
+      "grad_norm": 1.1809029579162598,
+      "learning_rate": 0.0001912691587403503,
+      "loss": 0.6462,
+      "step": 2207
+    },
+    {
+      "epoch": 0.7089420452721144,
+      "grad_norm": 1.2314188480377197,
+      "learning_rate": 0.00019125488863652615,
+      "loss": 0.5693,
+      "step": 2208
+    },
+    {
+      "epoch": 0.7092631240969658,
+      "grad_norm": 1.04121994972229,
+      "learning_rate": 0.00019124060741357063,
+      "loss": 0.8847,
+      "step": 2209
+    },
+    {
+      "epoch": 0.7095842029218173,
+      "grad_norm": 1.0673465728759766,
+      "learning_rate": 0.00019122631507322387,
+      "loss": 1.0044,
+      "step": 2210
+    },
+    {
+      "epoch": 0.7099052817466688,
+      "grad_norm": 0.6472902894020081,
+      "learning_rate": 0.0001912120116172273,
+      "loss": 0.7363,
+      "step": 2211
+    },
+    {
+      "epoch": 0.7102263605715203,
+      "grad_norm": 0.9594241380691528,
+      "learning_rate": 0.00019119769704732382,
+      "loss": 0.7698,
+      "step": 2212
+    },
+    {
+      "epoch": 0.7105474393963718,
+      "grad_norm": 1.1029056310653687,
+      "learning_rate": 0.0001911833713652576,
+      "loss": 1.025,
+      "step": 2213
+    },
+    {
+      "epoch": 0.7108685182212233,
+      "grad_norm": 1.1992616653442383,
+      "learning_rate": 0.00019116903457277413,
+      "loss": 1.0412,
+      "step": 2214
+    },
+    {
+      "epoch": 0.7111895970460748,
+      "grad_norm": 1.4310470819473267,
+      "learning_rate": 0.00019115468667162038,
+      "loss": 0.8216,
+      "step": 2215
+    },
+    {
+      "epoch": 0.7115106758709263,
+      "grad_norm": 1.5409244298934937,
+      "learning_rate": 0.00019114032766354453,
+      "loss": 0.5844,
+      "step": 2216
+    },
+    {
+      "epoch": 0.7118317546957779,
+      "grad_norm": 1.4199776649475098,
+      "learning_rate": 0.00019112595755029624,
+      "loss": 1.0082,
+      "step": 2217
+    },
+    {
+      "epoch": 0.7121528335206293,
+      "grad_norm": 1.1927727460861206,
+      "learning_rate": 0.0001911115763336264,
+      "loss": 0.8184,
+      "step": 2218
+    },
+    {
+      "epoch": 0.7124739123454809,
+      "grad_norm": 0.9337337017059326,
+      "learning_rate": 0.0001910971840152874,
+      "loss": 0.8842,
+      "step": 2219
+    },
+    {
+      "epoch": 0.7127949911703323,
+      "grad_norm": 0.9035426378250122,
+      "learning_rate": 0.0001910827805970328,
+      "loss": 0.8288,
+      "step": 2220
+    },
+    {
+      "epoch": 0.7131160699951838,
+      "grad_norm": 1.3903597593307495,
+      "learning_rate": 0.00019106836608061772,
+      "loss": 0.8678,
+      "step": 2221
+    },
+    {
+      "epoch": 0.7134371488200353,
+      "grad_norm": 1.2221095561981201,
+      "learning_rate": 0.00019105394046779845,
+      "loss": 0.7565,
+      "step": 2222
+    },
+    {
+      "epoch": 0.7137582276448868,
+      "grad_norm": 1.1277964115142822,
+      "learning_rate": 0.00019103950376033276,
+      "loss": 0.9622,
+      "step": 2223
+    },
+    {
+      "epoch": 0.7140793064697383,
+      "grad_norm": 1.453009843826294,
+      "learning_rate": 0.00019102505595997965,
+      "loss": 0.9864,
+      "step": 2224
+    },
+    {
+      "epoch": 0.7144003852945898,
+      "grad_norm": 1.2029697895050049,
+      "learning_rate": 0.00019101059706849957,
+      "loss": 1.1135,
+      "step": 2225
+    },
+    {
+      "epoch": 0.7147214641194414,
+      "grad_norm": 1.6069939136505127,
+      "learning_rate": 0.00019099612708765434,
+      "loss": 1.078,
+      "step": 2226
+    },
+    {
+      "epoch": 0.7150425429442928,
+      "grad_norm": 1.1098419427871704,
+      "learning_rate": 0.000190981646019207,
+      "loss": 0.8924,
+      "step": 2227
+    },
+    {
+      "epoch": 0.7153636217691444,
+      "grad_norm": 1.2886768579483032,
+      "learning_rate": 0.0001909671538649221,
+      "loss": 1.105,
+      "step": 2228
+    },
+    {
+      "epoch": 0.7156847005939958,
+      "grad_norm": 0.9400178790092468,
+      "learning_rate": 0.00019095265062656544,
+      "loss": 1.1208,
+      "step": 2229
+    },
+    {
+      "epoch": 0.7160057794188474,
+      "grad_norm": 1.2319140434265137,
+      "learning_rate": 0.00019093813630590418,
+      "loss": 0.8575,
+      "step": 2230
+    },
+    {
+      "epoch": 0.7163268582436988,
+      "grad_norm": 1.095289945602417,
+      "learning_rate": 0.00019092361090470688,
+      "loss": 0.842,
+      "step": 2231
+    },
+    {
+      "epoch": 0.7166479370685503,
+      "grad_norm": 1.421386957168579,
+      "learning_rate": 0.00019090907442474334,
+      "loss": 1.0803,
+      "step": 2232
+    },
+    {
+      "epoch": 0.7169690158934018,
+      "grad_norm": 1.5871707201004028,
+      "learning_rate": 0.00019089452686778488,
+      "loss": 1.0677,
+      "step": 2233
+    },
+    {
+      "epoch": 0.7172900947182533,
+      "grad_norm": 1.591778039932251,
+      "learning_rate": 0.00019087996823560402,
+      "loss": 1.0679,
+      "step": 2234
+    },
+    {
+      "epoch": 0.7176111735431049,
+      "grad_norm": 2.42740797996521,
+      "learning_rate": 0.0001908653985299747,
+      "loss": 1.0548,
+      "step": 2235
+    },
+    {
+      "epoch": 0.7179322523679563,
+      "grad_norm": 1.0320085287094116,
+      "learning_rate": 0.0001908508177526722,
+      "loss": 0.7861,
+      "step": 2236
+    },
+    {
+      "epoch": 0.7182533311928079,
+      "grad_norm": 1.18021821975708,
+      "learning_rate": 0.00019083622590547312,
+      "loss": 0.8066,
+      "step": 2237
+    },
+    {
+      "epoch": 0.7185744100176593,
+      "grad_norm": 1.0343506336212158,
+      "learning_rate": 0.00019082162299015546,
+      "loss": 0.8142,
+      "step": 2238
+    },
+    {
+      "epoch": 0.7188954888425109,
+      "grad_norm": 1.4337462186813354,
+      "learning_rate": 0.00019080700900849851,
+      "loss": 0.9664,
+      "step": 2239
+    },
+    {
+      "epoch": 0.7192165676673623,
+      "grad_norm": 1.5604829788208008,
+      "learning_rate": 0.000190792383962283,
+      "loss": 1.024,
+      "step": 2240
+    },
+    {
+      "epoch": 0.7195376464922139,
+      "grad_norm": 1.152066707611084,
+      "learning_rate": 0.00019077774785329087,
+      "loss": 1.0221,
+      "step": 2241
+    },
+    {
+      "epoch": 0.7198587253170653,
+      "grad_norm": 1.9510457515716553,
+      "learning_rate": 0.00019076310068330554,
+      "loss": 1.0129,
+      "step": 2242
+    },
+    {
+      "epoch": 0.7201798041419168,
+      "grad_norm": 1.1222991943359375,
+      "learning_rate": 0.0001907484424541117,
+      "loss": 0.8,
+      "step": 2243
+    },
+    {
+      "epoch": 0.7205008829667684,
+      "grad_norm": 1.8768889904022217,
+      "learning_rate": 0.00019073377316749542,
+      "loss": 0.8479,
+      "step": 2244
+    },
+    {
+      "epoch": 0.7208219617916198,
+      "grad_norm": 1.251123309135437,
+      "learning_rate": 0.00019071909282524413,
+      "loss": 1.0588,
+      "step": 2245
+    },
+    {
+      "epoch": 0.7211430406164714,
+      "grad_norm": 1.2855134010314941,
+      "learning_rate": 0.0001907044014291465,
+      "loss": 0.9369,
+      "step": 2246
+    },
+    {
+      "epoch": 0.7214641194413228,
+      "grad_norm": 0.9821279644966125,
+      "learning_rate": 0.0001906896989809927,
+      "loss": 0.6617,
+      "step": 2247
+    },
+    {
+      "epoch": 0.7217851982661744,
+      "grad_norm": 1.674775242805481,
+      "learning_rate": 0.00019067498548257423,
+      "loss": 0.7747,
+      "step": 2248
+    },
+    {
+      "epoch": 0.7221062770910258,
+      "grad_norm": 1.1248822212219238,
+      "learning_rate": 0.00019066026093568378,
+      "loss": 0.739,
+      "step": 2249
+    },
+    {
+      "epoch": 0.7224273559158774,
+      "grad_norm": 0.8824935555458069,
+      "learning_rate": 0.00019064552534211554,
+      "loss": 0.5349,
+      "step": 2250
+    },
+    {
+      "epoch": 0.7227484347407288,
+      "grad_norm": 1.197925329208374,
+      "learning_rate": 0.000190630778703665,
+      "loss": 1.3444,
+      "step": 2251
+    },
+    {
+      "epoch": 0.7230695135655804,
+      "grad_norm": 0.9842731356620789,
+      "learning_rate": 0.00019061602102212898,
+      "loss": 1.3156,
+      "step": 2252
+    },
+    {
+      "epoch": 0.7233905923904318,
+      "grad_norm": 1.0988041162490845,
+      "learning_rate": 0.0001906012522993057,
+      "loss": 1.1763,
+      "step": 2253
+    },
+    {
+      "epoch": 0.7237116712152833,
+      "grad_norm": 1.067678451538086,
+      "learning_rate": 0.0001905864725369946,
+      "loss": 0.8682,
+      "step": 2254
+    },
+    {
+      "epoch": 0.7240327500401349,
+      "grad_norm": 1.3331475257873535,
+      "learning_rate": 0.00019057168173699664,
+      "loss": 0.7538,
+      "step": 2255
+    },
+    {
+      "epoch": 0.7243538288649863,
+      "grad_norm": 1.1442729234695435,
+      "learning_rate": 0.00019055687990111398,
+      "loss": 0.7896,
+      "step": 2256
+    },
+    {
+      "epoch": 0.7246749076898379,
+      "grad_norm": 1.0878592729568481,
+      "learning_rate": 0.0001905420670311502,
+      "loss": 0.5776,
+      "step": 2257
+    },
+    {
+      "epoch": 0.7249959865146893,
+      "grad_norm": 1.124906301498413,
+      "learning_rate": 0.00019052724312891014,
+      "loss": 0.8746,
+      "step": 2258
+    },
+    {
+      "epoch": 0.7253170653395409,
+      "grad_norm": 1.1294090747833252,
+      "learning_rate": 0.00019051240819620014,
+      "loss": 1.0163,
+      "step": 2259
+    },
+    {
+      "epoch": 0.7256381441643923,
+      "grad_norm": 1.1215919256210327,
+      "learning_rate": 0.0001904975622348278,
+      "loss": 1.042,
+      "step": 2260
+    },
+    {
+      "epoch": 0.7259592229892439,
+      "grad_norm": 0.8851989507675171,
+      "learning_rate": 0.00019048270524660196,
+      "loss": 0.6864,
+      "step": 2261
+    },
+    {
+      "epoch": 0.7262803018140953,
+      "grad_norm": 0.9922202229499817,
+      "learning_rate": 0.00019046783723333297,
+      "loss": 0.7434,
+      "step": 2262
+    },
+    {
+      "epoch": 0.7266013806389469,
+      "grad_norm": 1.0879603624343872,
+      "learning_rate": 0.00019045295819683242,
+      "loss": 0.7117,
+      "step": 2263
+    },
+    {
+      "epoch": 0.7269224594637984,
+      "grad_norm": 1.3650341033935547,
+      "learning_rate": 0.0001904380681389133,
+      "loss": 0.9001,
+      "step": 2264
+    },
+    {
+      "epoch": 0.7272435382886498,
+      "grad_norm": 1.5964879989624023,
+      "learning_rate": 0.00019042316706138987,
+      "loss": 1.0102,
+      "step": 2265
+    },
+    {
+      "epoch": 0.7275646171135014,
+      "grad_norm": 0.8338374495506287,
+      "learning_rate": 0.00019040825496607786,
+      "loss": 0.8816,
+      "step": 2266
+    },
+    {
+      "epoch": 0.7278856959383528,
+      "grad_norm": 1.0702760219573975,
+      "learning_rate": 0.00019039333185479418,
+      "loss": 0.8459,
+      "step": 2267
+    },
+    {
+      "epoch": 0.7282067747632044,
+      "grad_norm": 0.976639449596405,
+      "learning_rate": 0.0001903783977293572,
+      "loss": 0.8101,
+      "step": 2268
+    },
+    {
+      "epoch": 0.7285278535880558,
+      "grad_norm": 1.2155683040618896,
+      "learning_rate": 0.00019036345259158667,
+      "loss": 1.0607,
+      "step": 2269
+    },
+    {
+      "epoch": 0.7288489324129074,
+      "grad_norm": 1.2647230625152588,
+      "learning_rate": 0.0001903484964433035,
+      "loss": 1.1008,
+      "step": 2270
+    },
+    {
+      "epoch": 0.7291700112377588,
+      "grad_norm": 0.9674025177955627,
+      "learning_rate": 0.0001903335292863301,
+      "loss": 0.7245,
+      "step": 2271
+    },
+    {
+      "epoch": 0.7294910900626104,
+      "grad_norm": 1.2419253587722778,
+      "learning_rate": 0.00019031855112249015,
+      "loss": 0.8682,
+      "step": 2272
+    },
+    {
+      "epoch": 0.7298121688874619,
+      "grad_norm": 0.9692038297653198,
+      "learning_rate": 0.00019030356195360874,
+      "loss": 0.9075,
+      "step": 2273
+    },
+    {
+      "epoch": 0.7301332477123134,
+      "grad_norm": 1.1425732374191284,
+      "learning_rate": 0.0001902885617815122,
+      "loss": 0.7757,
+      "step": 2274
+    },
+    {
+      "epoch": 0.7304543265371649,
+      "grad_norm": 0.9863192439079285,
+      "learning_rate": 0.0001902735506080283,
+      "loss": 0.8154,
+      "step": 2275
+    },
+    {
+      "epoch": 0.7307754053620164,
+      "grad_norm": 0.9807154536247253,
+      "learning_rate": 0.00019025852843498607,
+      "loss": 0.9582,
+      "step": 2276
+    },
+    {
+      "epoch": 0.7310964841868679,
+      "grad_norm": 1.2582796812057495,
+      "learning_rate": 0.00019024349526421594,
+      "loss": 1.009,
+      "step": 2277
+    },
+    {
+      "epoch": 0.7314175630117193,
+      "grad_norm": 1.0475612878799438,
+      "learning_rate": 0.00019022845109754966,
+      "loss": 0.997,
+      "step": 2278
+    },
+    {
+      "epoch": 0.7317386418365709,
+      "grad_norm": 1.6267993450164795,
+      "learning_rate": 0.00019021339593682028,
+      "loss": 0.9519,
+      "step": 2279
+    },
+    {
+      "epoch": 0.7320597206614223,
+      "grad_norm": 1.2485898733139038,
+      "learning_rate": 0.00019019832978386228,
+      "loss": 0.8931,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7323807994862739,
+      "grad_norm": 0.9349825382232666,
+      "learning_rate": 0.0001901832526405114,
+      "loss": 0.8948,
+      "step": 2281
+    },
+    {
+      "epoch": 0.7327018783111254,
+      "grad_norm": 1.5776710510253906,
+      "learning_rate": 0.00019016816450860474,
+      "loss": 0.9504,
+      "step": 2282
+    },
+    {
+      "epoch": 0.7330229571359769,
+      "grad_norm": 0.9056895971298218,
+      "learning_rate": 0.0001901530653899807,
+      "loss": 0.8712,
+      "step": 2283
+    },
+    {
+      "epoch": 0.7333440359608284,
+      "grad_norm": 1.3263038396835327,
+      "learning_rate": 0.00019013795528647912,
+      "loss": 0.8974,
+      "step": 2284
+    },
+    {
+      "epoch": 0.7336651147856799,
+      "grad_norm": 1.0501210689544678,
+      "learning_rate": 0.00019012283419994115,
+      "loss": 0.888,
+      "step": 2285
+    },
+    {
+      "epoch": 0.7339861936105314,
+      "grad_norm": 1.9327582120895386,
+      "learning_rate": 0.00019010770213220916,
+      "loss": 1.0191,
+      "step": 2286
+    },
+    {
+      "epoch": 0.7343072724353829,
+      "grad_norm": 0.9051433205604553,
+      "learning_rate": 0.000190092559085127,
+      "loss": 0.736,
+      "step": 2287
+    },
+    {
+      "epoch": 0.7346283512602344,
+      "grad_norm": 2.308243989944458,
+      "learning_rate": 0.00019007740506053983,
+      "loss": 0.9449,
+      "step": 2288
+    },
+    {
+      "epoch": 0.7349494300850858,
+      "grad_norm": 0.9226866960525513,
+      "learning_rate": 0.00019006224006029406,
+      "loss": 0.9529,
+      "step": 2289
+    },
+    {
+      "epoch": 0.7352705089099374,
+      "grad_norm": 1.1376844644546509,
+      "learning_rate": 0.0001900470640862375,
+      "loss": 0.8272,
+      "step": 2290
+    },
+    {
+      "epoch": 0.7355915877347889,
+      "grad_norm": 1.475791335105896,
+      "learning_rate": 0.00019003187714021938,
+      "loss": 1.1438,
+      "step": 2291
+    },
+    {
+      "epoch": 0.7359126665596404,
+      "grad_norm": 0.9544022679328918,
+      "learning_rate": 0.00019001667922409008,
+      "loss": 0.8274,
+      "step": 2292
+    },
+    {
+      "epoch": 0.7362337453844919,
+      "grad_norm": 1.031445860862732,
+      "learning_rate": 0.00019000147033970148,
+      "loss": 0.7776,
+      "step": 2293
+    },
+    {
+      "epoch": 0.7365548242093434,
+      "grad_norm": 0.8652722835540771,
+      "learning_rate": 0.00018998625048890672,
+      "loss": 0.7621,
+      "step": 2294
+    },
+    {
+      "epoch": 0.7368759030341949,
+      "grad_norm": 1.1719671487808228,
+      "learning_rate": 0.0001899710196735603,
+      "loss": 0.9207,
+      "step": 2295
+    },
+    {
+      "epoch": 0.7371969818590464,
+      "grad_norm": 1.2041524648666382,
+      "learning_rate": 0.00018995577789551803,
+      "loss": 0.8044,
+      "step": 2296
+    },
+    {
+      "epoch": 0.7375180606838979,
+      "grad_norm": 1.1736308336257935,
+      "learning_rate": 0.0001899405251566371,
+      "loss": 0.8288,
+      "step": 2297
+    },
+    {
+      "epoch": 0.7378391395087494,
+      "grad_norm": 0.8900883793830872,
+      "learning_rate": 0.000189925261458776,
+      "loss": 0.7487,
+      "step": 2298
+    },
+    {
+      "epoch": 0.7381602183336009,
+      "grad_norm": 0.7392171621322632,
+      "learning_rate": 0.00018990998680379456,
+      "loss": 0.6899,
+      "step": 2299
+    },
+    {
+      "epoch": 0.7384812971584525,
+      "grad_norm": 1.4253010749816895,
+      "learning_rate": 0.00018989470119355398,
+      "loss": 0.6706,
+      "step": 2300
+    },
+    {
+      "epoch": 0.7388023759833039,
+      "grad_norm": 0.8190407752990723,
+      "learning_rate": 0.0001898794046299167,
+      "loss": 1.2835,
+      "step": 2301
+    },
+    {
+      "epoch": 0.7391234548081554,
+      "grad_norm": 1.0139261484146118,
+      "learning_rate": 0.00018986409711474665,
+      "loss": 0.9747,
+      "step": 2302
+    },
+    {
+      "epoch": 0.7394445336330069,
+      "grad_norm": 1.0037243366241455,
+      "learning_rate": 0.00018984877864990888,
+      "loss": 0.7472,
+      "step": 2303
+    },
+    {
+      "epoch": 0.7397656124578584,
+      "grad_norm": 1.3445838689804077,
+      "learning_rate": 0.00018983344923727003,
+      "loss": 0.8011,
+      "step": 2304
+    },
+    {
+      "epoch": 0.7400866912827099,
+      "grad_norm": 1.1861538887023926,
+      "learning_rate": 0.00018981810887869785,
+      "loss": 0.5261,
+      "step": 2305
+    },
+    {
+      "epoch": 0.7404077701075614,
+      "grad_norm": 1.1027861833572388,
+      "learning_rate": 0.00018980275757606157,
+      "loss": 0.6695,
+      "step": 2306
+    },
+    {
+      "epoch": 0.7407288489324129,
+      "grad_norm": 1.280778169631958,
+      "learning_rate": 0.0001897873953312317,
+      "loss": 0.7354,
+      "step": 2307
+    },
+    {
+      "epoch": 0.7410499277572644,
+      "grad_norm": 1.1856905221939087,
+      "learning_rate": 0.00018977202214608,
+      "loss": 1.0707,
+      "step": 2308
+    },
+    {
+      "epoch": 0.7413710065821159,
+      "grad_norm": 0.8002036809921265,
+      "learning_rate": 0.00018975663802247976,
+      "loss": 0.763,
+      "step": 2309
+    },
+    {
+      "epoch": 0.7416920854069674,
+      "grad_norm": 1.3091520071029663,
+      "learning_rate": 0.0001897412429623054,
+      "loss": 1.1745,
+      "step": 2310
+    },
+    {
+      "epoch": 0.742013164231819,
+      "grad_norm": 1.0086097717285156,
+      "learning_rate": 0.00018972583696743285,
+      "loss": 1.039,
+      "step": 2311
+    },
+    {
+      "epoch": 0.7423342430566704,
+      "grad_norm": 1.0016741752624512,
+      "learning_rate": 0.00018971042003973924,
+      "loss": 0.9677,
+      "step": 2312
+    },
+    {
+      "epoch": 0.742655321881522,
+      "grad_norm": 1.2029709815979004,
+      "learning_rate": 0.000189694992181103,
+      "loss": 1.062,
+      "step": 2313
+    },
+    {
+      "epoch": 0.7429764007063734,
+      "grad_norm": 0.854852020740509,
+      "learning_rate": 0.00018967955339340407,
+      "loss": 0.8209,
+      "step": 2314
+    },
+    {
+      "epoch": 0.7432974795312249,
+      "grad_norm": 0.892218828201294,
+      "learning_rate": 0.00018966410367852362,
+      "loss": 1.0757,
+      "step": 2315
+    },
+    {
+      "epoch": 0.7436185583560764,
+      "grad_norm": 0.8785187005996704,
+      "learning_rate": 0.00018964864303834406,
+      "loss": 0.9317,
+      "step": 2316
+    },
+    {
+      "epoch": 0.7439396371809279,
+      "grad_norm": 0.999880850315094,
+      "learning_rate": 0.0001896331714747493,
+      "loss": 0.8786,
+      "step": 2317
+    },
+    {
+      "epoch": 0.7442607160057794,
+      "grad_norm": 1.373669147491455,
+      "learning_rate": 0.0001896176889896245,
+      "loss": 1.0008,
+      "step": 2318
+    },
+    {
+      "epoch": 0.7445817948306309,
+      "grad_norm": 1.0708602666854858,
+      "learning_rate": 0.0001896021955848561,
+      "loss": 1.0163,
+      "step": 2319
+    },
+    {
+      "epoch": 0.7449028736554825,
+      "grad_norm": 1.0174452066421509,
+      "learning_rate": 0.00018958669126233199,
+      "loss": 0.9382,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7452239524803339,
+      "grad_norm": 1.2005442380905151,
+      "learning_rate": 0.0001895711760239413,
+      "loss": 0.8355,
+      "step": 2321
+    },
+    {
+      "epoch": 0.7455450313051855,
+      "grad_norm": 1.3386541604995728,
+      "learning_rate": 0.0001895556498715745,
+      "loss": 0.8852,
+      "step": 2322
+    },
+    {
+      "epoch": 0.7458661101300369,
+      "grad_norm": 1.0639350414276123,
+      "learning_rate": 0.0001895401128071234,
+      "loss": 0.9128,
+      "step": 2323
+    },
+    {
+      "epoch": 0.7461871889548884,
+      "grad_norm": 0.9678146243095398,
+      "learning_rate": 0.00018952456483248119,
+      "loss": 0.9178,
+      "step": 2324
+    },
+    {
+      "epoch": 0.7465082677797399,
+      "grad_norm": 1.260991096496582,
+      "learning_rate": 0.00018950900594954227,
+      "loss": 0.9372,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7468293466045914,
+      "grad_norm": 1.2211194038391113,
+      "learning_rate": 0.00018949343616020252,
+      "loss": 0.8576,
+      "step": 2326
+    },
+    {
+      "epoch": 0.7471504254294429,
+      "grad_norm": 1.1809611320495605,
+      "learning_rate": 0.00018947785546635904,
+      "loss": 0.9619,
+      "step": 2327
+    },
+    {
+      "epoch": 0.7474715042542944,
+      "grad_norm": 2.355696678161621,
+      "learning_rate": 0.00018946226386991027,
+      "loss": 1.1663,
+      "step": 2328
+    },
+    {
+      "epoch": 0.747792583079146,
+      "grad_norm": 1.0631635189056396,
+      "learning_rate": 0.000189446661372756,
+      "loss": 0.8316,
+      "step": 2329
+    },
+    {
+      "epoch": 0.7481136619039974,
+      "grad_norm": 1.7369928359985352,
+      "learning_rate": 0.0001894310479767974,
+      "loss": 0.8755,
+      "step": 2330
+    },
+    {
+      "epoch": 0.748434740728849,
+      "grad_norm": 2.2223386764526367,
+      "learning_rate": 0.0001894154236839368,
+      "loss": 1.1898,
+      "step": 2331
+    },
+    {
+      "epoch": 0.7487558195537004,
+      "grad_norm": 1.2930457592010498,
+      "learning_rate": 0.00018939978849607814,
+      "loss": 0.8362,
+      "step": 2332
+    },
+    {
+      "epoch": 0.749076898378552,
+      "grad_norm": 0.9054352045059204,
+      "learning_rate": 0.0001893841424151264,
+      "loss": 0.8712,
+      "step": 2333
+    },
+    {
+      "epoch": 0.7493979772034034,
+      "grad_norm": 1.637449026107788,
+      "learning_rate": 0.000189368485442988,
+      "loss": 1.0388,
+      "step": 2334
+    },
+    {
+      "epoch": 0.749719056028255,
+      "grad_norm": 1.306647777557373,
+      "learning_rate": 0.00018935281758157078,
+      "loss": 0.8911,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7500401348531064,
+      "grad_norm": 1.2520787715911865,
+      "learning_rate": 0.00018933713883278376,
+      "loss": 0.8615,
+      "step": 2336
+    },
+    {
+      "epoch": 0.7503612136779579,
+      "grad_norm": 1.5709820985794067,
+      "learning_rate": 0.0001893214491985374,
+      "loss": 0.8583,
+      "step": 2337
+    },
+    {
+      "epoch": 0.7506822925028095,
+      "grad_norm": 1.1838668584823608,
+      "learning_rate": 0.00018930574868074334,
+      "loss": 0.9393,
+      "step": 2338
+    },
+    {
+      "epoch": 0.7510033713276609,
+      "grad_norm": 1.7159479856491089,
+      "learning_rate": 0.0001892900372813147,
+      "loss": 0.7978,
+      "step": 2339
+    },
+    {
+      "epoch": 0.7513244501525125,
+      "grad_norm": 0.8659077882766724,
+      "learning_rate": 0.00018927431500216586,
+      "loss": 0.8576,
+      "step": 2340
+    },
+    {
+      "epoch": 0.7516455289773639,
+      "grad_norm": 1.2940974235534668,
+      "learning_rate": 0.00018925858184521256,
+      "loss": 0.7817,
+      "step": 2341
+    },
+    {
+      "epoch": 0.7519666078022155,
+      "grad_norm": 1.3620824813842773,
+      "learning_rate": 0.0001892428378123718,
+      "loss": 1.0294,
+      "step": 2342
+    },
+    {
+      "epoch": 0.7522876866270669,
+      "grad_norm": 1.067583680152893,
+      "learning_rate": 0.00018922708290556198,
+      "loss": 0.6811,
+      "step": 2343
+    },
+    {
+      "epoch": 0.7526087654519185,
+      "grad_norm": 1.2058768272399902,
+      "learning_rate": 0.0001892113171267027,
+      "loss": 0.8882,
+      "step": 2344
+    },
+    {
+      "epoch": 0.7529298442767699,
+      "grad_norm": 1.3467442989349365,
+      "learning_rate": 0.0001891955404777151,
+      "loss": 0.6371,
+      "step": 2345
+    },
+    {
+      "epoch": 0.7532509231016215,
+      "grad_norm": 0.8499478697776794,
+      "learning_rate": 0.00018917975296052142,
+      "loss": 0.7051,
+      "step": 2346
+    },
+    {
+      "epoch": 0.753572001926473,
+      "grad_norm": 1.8910988569259644,
+      "learning_rate": 0.00018916395457704534,
+      "loss": 0.7791,
+      "step": 2347
+    },
+    {
+      "epoch": 0.7538930807513244,
+      "grad_norm": 1.0912384986877441,
+      "learning_rate": 0.00018914814532921187,
+      "loss": 0.6608,
+      "step": 2348
+    },
+    {
+      "epoch": 0.754214159576176,
+      "grad_norm": 0.564804196357727,
+      "learning_rate": 0.00018913232521894732,
+      "loss": 0.4543,
+      "step": 2349
+    },
+    {
+      "epoch": 0.7545352384010274,
+      "grad_norm": 0.8918449878692627,
+      "learning_rate": 0.00018911649424817933,
+      "loss": 0.4786,
+      "step": 2350
+    },
+    {
+      "epoch": 0.754856317225879,
+      "grad_norm": 1.0368508100509644,
+      "learning_rate": 0.0001891006524188368,
+      "loss": 1.2446,
+      "step": 2351
+    },
+    {
+      "epoch": 0.7551773960507304,
+      "grad_norm": 1.0669269561767578,
+      "learning_rate": 0.00018908479973285005,
+      "loss": 1.1619,
+      "step": 2352
+    },
+    {
+      "epoch": 0.755498474875582,
+      "grad_norm": 1.34662663936615,
+      "learning_rate": 0.00018906893619215066,
+      "loss": 0.9378,
+      "step": 2353
+    },
+    {
+      "epoch": 0.7558195537004334,
+      "grad_norm": 1.1142168045043945,
+      "learning_rate": 0.0001890530617986716,
+      "loss": 0.6657,
+      "step": 2354
+    },
+    {
+      "epoch": 0.756140632525285,
+      "grad_norm": 1.00096595287323,
+      "learning_rate": 0.00018903717655434707,
+      "loss": 0.6454,
+      "step": 2355
+    },
+    {
+      "epoch": 0.7564617113501365,
+      "grad_norm": 1.4639570713043213,
+      "learning_rate": 0.00018902128046111266,
+      "loss": 0.6425,
+      "step": 2356
+    },
+    {
+      "epoch": 0.756782790174988,
+      "grad_norm": 1.1114808320999146,
+      "learning_rate": 0.00018900537352090524,
+      "loss": 0.7002,
+      "step": 2357
+    },
+    {
+      "epoch": 0.7571038689998395,
+      "grad_norm": 1.290031909942627,
+      "learning_rate": 0.00018898945573566308,
+      "loss": 0.8956,
+      "step": 2358
+    },
+    {
+      "epoch": 0.7574249478246909,
+      "grad_norm": 1.2316958904266357,
+      "learning_rate": 0.00018897352710732564,
+      "loss": 0.8771,
+      "step": 2359
+    },
+    {
+      "epoch": 0.7577460266495425,
+      "grad_norm": 1.4026559591293335,
+      "learning_rate": 0.00018895758763783383,
+      "loss": 1.0492,
+      "step": 2360
+    },
+    {
+      "epoch": 0.7580671054743939,
+      "grad_norm": 0.8219192624092102,
+      "learning_rate": 0.00018894163732912977,
+      "loss": 0.8172,
+      "step": 2361
+    },
+    {
+      "epoch": 0.7583881842992455,
+      "grad_norm": 0.9561658501625061,
+      "learning_rate": 0.000188925676183157,
+      "loss": 0.8669,
+      "step": 2362
+    },
+    {
+      "epoch": 0.7587092631240969,
+      "grad_norm": 0.9221644997596741,
+      "learning_rate": 0.00018890970420186033,
+      "loss": 0.701,
+      "step": 2363
+    },
+    {
+      "epoch": 0.7590303419489485,
+      "grad_norm": 1.0757882595062256,
+      "learning_rate": 0.0001888937213871859,
+      "loss": 0.7451,
+      "step": 2364
+    },
+    {
+      "epoch": 0.7593514207738,
+      "grad_norm": 1.2633696794509888,
+      "learning_rate": 0.00018887772774108116,
+      "loss": 0.9686,
+      "step": 2365
+    },
+    {
+      "epoch": 0.7596724995986515,
+      "grad_norm": 1.1361249685287476,
+      "learning_rate": 0.0001888617232654949,
+      "loss": 1.032,
+      "step": 2366
+    },
+    {
+      "epoch": 0.759993578423503,
+      "grad_norm": 1.2562459707260132,
+      "learning_rate": 0.00018884570796237718,
+      "loss": 0.9559,
+      "step": 2367
+    },
+    {
+      "epoch": 0.7603146572483545,
+      "grad_norm": 0.9838060140609741,
+      "learning_rate": 0.00018882968183367947,
+      "loss": 0.9249,
+      "step": 2368
+    },
+    {
+      "epoch": 0.760635736073206,
+      "grad_norm": 0.8353525996208191,
+      "learning_rate": 0.00018881364488135448,
+      "loss": 0.8962,
+      "step": 2369
+    },
+    {
+      "epoch": 0.7609568148980574,
+      "grad_norm": 1.24613618850708,
+      "learning_rate": 0.00018879759710735622,
+      "loss": 0.927,
+      "step": 2370
+    },
+    {
+      "epoch": 0.761277893722909,
+      "grad_norm": 1.1252332925796509,
+      "learning_rate": 0.00018878153851364013,
+      "loss": 0.9016,
+      "step": 2371
+    },
+    {
+      "epoch": 0.7615989725477604,
+      "grad_norm": 0.9512665271759033,
+      "learning_rate": 0.00018876546910216288,
+      "loss": 0.9949,
+      "step": 2372
+    },
+    {
+      "epoch": 0.761920051372612,
+      "grad_norm": 1.5389331579208374,
+      "learning_rate": 0.00018874938887488248,
+      "loss": 1.0535,
+      "step": 2373
+    },
+    {
+      "epoch": 0.7622411301974634,
+      "grad_norm": 0.9745280146598816,
+      "learning_rate": 0.00018873329783375824,
+      "loss": 0.9594,
+      "step": 2374
+    },
+    {
+      "epoch": 0.762562209022315,
+      "grad_norm": 1.1034287214279175,
+      "learning_rate": 0.0001887171959807508,
+      "loss": 0.9506,
+      "step": 2375
+    },
+    {
+      "epoch": 0.7628832878471665,
+      "grad_norm": 1.1382795572280884,
+      "learning_rate": 0.00018870108331782217,
+      "loss": 0.771,
+      "step": 2376
+    },
+    {
+      "epoch": 0.763204366672018,
+      "grad_norm": 1.042029619216919,
+      "learning_rate": 0.0001886849598469356,
+      "loss": 0.8624,
+      "step": 2377
+    },
+    {
+      "epoch": 0.7635254454968695,
+      "grad_norm": 1.0116221904754639,
+      "learning_rate": 0.00018866882557005567,
+      "loss": 0.9996,
+      "step": 2378
+    },
+    {
+      "epoch": 0.763846524321721,
+      "grad_norm": 0.6933526992797852,
+      "learning_rate": 0.00018865268048914828,
+      "loss": 0.657,
+      "step": 2379
+    },
+    {
+      "epoch": 0.7641676031465725,
+      "grad_norm": 1.2314203977584839,
+      "learning_rate": 0.0001886365246061807,
+      "loss": 1.1146,
+      "step": 2380
+    },
+    {
+      "epoch": 0.764488681971424,
+      "grad_norm": 1.3230317831039429,
+      "learning_rate": 0.00018862035792312147,
+      "loss": 0.8496,
+      "step": 2381
+    },
+    {
+      "epoch": 0.7648097607962755,
+      "grad_norm": 0.9279382824897766,
+      "learning_rate": 0.00018860418044194045,
+      "loss": 0.8664,
+      "step": 2382
+    },
+    {
+      "epoch": 0.7651308396211269,
+      "grad_norm": 1.3693788051605225,
+      "learning_rate": 0.00018858799216460881,
+      "loss": 0.935,
+      "step": 2383
+    },
+    {
+      "epoch": 0.7654519184459785,
+      "grad_norm": 1.4639575481414795,
+      "learning_rate": 0.00018857179309309901,
+      "loss": 0.9289,
+      "step": 2384
+    },
+    {
+      "epoch": 0.76577299727083,
+      "grad_norm": 1.192893147468567,
+      "learning_rate": 0.00018855558322938493,
+      "loss": 0.841,
+      "step": 2385
+    },
+    {
+      "epoch": 0.7660940760956815,
+      "grad_norm": 2.4525952339172363,
+      "learning_rate": 0.0001885393625754416,
+      "loss": 0.9004,
+      "step": 2386
+    },
+    {
+      "epoch": 0.766415154920533,
+      "grad_norm": 0.7626288533210754,
+      "learning_rate": 0.00018852313113324552,
+      "loss": 0.5966,
+      "step": 2387
+    },
+    {
+      "epoch": 0.7667362337453845,
+      "grad_norm": 1.5377343893051147,
+      "learning_rate": 0.00018850688890477445,
+      "loss": 0.9392,
+      "step": 2388
+    },
+    {
+      "epoch": 0.767057312570236,
+      "grad_norm": 1.555207371711731,
+      "learning_rate": 0.00018849063589200743,
+      "loss": 1.0675,
+      "step": 2389
+    },
+    {
+      "epoch": 0.7673783913950875,
+      "grad_norm": 1.0154179334640503,
+      "learning_rate": 0.00018847437209692486,
+      "loss": 0.6848,
+      "step": 2390
+    },
+    {
+      "epoch": 0.767699470219939,
+      "grad_norm": 1.9315353631973267,
+      "learning_rate": 0.0001884580975215084,
+      "loss": 0.8662,
+      "step": 2391
+    },
+    {
+      "epoch": 0.7680205490447904,
+      "grad_norm": 1.4244552850723267,
+      "learning_rate": 0.0001884418121677411,
+      "loss": 1.0924,
+      "step": 2392
+    },
+    {
+      "epoch": 0.768341627869642,
+      "grad_norm": 1.190861701965332,
+      "learning_rate": 0.00018842551603760724,
+      "loss": 0.8864,
+      "step": 2393
+    },
+    {
+      "epoch": 0.7686627066944935,
+      "grad_norm": 0.8999507427215576,
+      "learning_rate": 0.0001884092091330925,
+      "loss": 0.6656,
+      "step": 2394
+    },
+    {
+      "epoch": 0.768983785519345,
+      "grad_norm": 1.2089276313781738,
+      "learning_rate": 0.00018839289145618378,
+      "loss": 0.7695,
+      "step": 2395
+    },
+    {
+      "epoch": 0.7693048643441965,
+      "grad_norm": 0.7257319092750549,
+      "learning_rate": 0.00018837656300886937,
+      "loss": 0.5922,
+      "step": 2396
+    },
+    {
+      "epoch": 0.769625943169048,
+      "grad_norm": 0.9687317609786987,
+      "learning_rate": 0.00018836022379313883,
+      "loss": 0.6401,
+      "step": 2397
+    },
+    {
+      "epoch": 0.7699470219938995,
+      "grad_norm": 1.0825469493865967,
+      "learning_rate": 0.000188343873810983,
+      "loss": 0.8336,
+      "step": 2398
+    },
+    {
+      "epoch": 0.770268100818751,
+      "grad_norm": 1.286629319190979,
+      "learning_rate": 0.00018832751306439418,
+      "loss": 0.9844,
+      "step": 2399
+    },
+    {
+      "epoch": 0.7705891796436025,
+      "grad_norm": 0.8030791878700256,
+      "learning_rate": 0.0001883111415553658,
+      "loss": 0.5217,
+      "step": 2400
+    },
+    {
+      "epoch": 0.770910258468454,
+      "grad_norm": 1.3023558855056763,
+      "learning_rate": 0.00018829475928589271,
+      "loss": 1.1508,
+      "step": 2401
+    },
+    {
+      "epoch": 0.7712313372933055,
+      "grad_norm": 1.1620362997055054,
+      "learning_rate": 0.00018827836625797103,
+      "loss": 1.1917,
+      "step": 2402
+    },
+    {
+      "epoch": 0.7715524161181571,
+      "grad_norm": 1.134852647781372,
+      "learning_rate": 0.00018826196247359817,
+      "loss": 0.8599,
+      "step": 2403
+    },
+    {
+      "epoch": 0.7718734949430085,
+      "grad_norm": 0.962922990322113,
+      "learning_rate": 0.00018824554793477294,
+      "loss": 0.708,
+      "step": 2404
+    },
+    {
+      "epoch": 0.77219457376786,
+      "grad_norm": 1.1392818689346313,
+      "learning_rate": 0.00018822912264349534,
+      "loss": 0.6393,
+      "step": 2405
+    },
+    {
+      "epoch": 0.7725156525927115,
+      "grad_norm": 1.5677722692489624,
+      "learning_rate": 0.00018821268660176678,
+      "loss": 0.7293,
+      "step": 2406
+    },
+    {
+      "epoch": 0.772836731417563,
+      "grad_norm": 1.1908127069473267,
+      "learning_rate": 0.00018819623981158995,
+      "loss": 0.8353,
+      "step": 2407
+    },
+    {
+      "epoch": 0.7731578102424145,
+      "grad_norm": 0.9969918727874756,
+      "learning_rate": 0.00018817978227496883,
+      "loss": 0.9109,
+      "step": 2408
+    },
+    {
+      "epoch": 0.773478889067266,
+      "grad_norm": 0.8602403402328491,
+      "learning_rate": 0.0001881633139939087,
+      "loss": 0.9064,
+      "step": 2409
+    },
+    {
+      "epoch": 0.7737999678921175,
+      "grad_norm": 0.8332744836807251,
+      "learning_rate": 0.0001881468349704162,
+      "loss": 0.8529,
+      "step": 2410
+    },
+    {
+      "epoch": 0.774121046716969,
+      "grad_norm": 1.0467489957809448,
+      "learning_rate": 0.0001881303452064992,
+      "loss": 0.8339,
+      "step": 2411
+    },
+    {
+      "epoch": 0.7744421255418206,
+      "grad_norm": 1.618408441543579,
+      "learning_rate": 0.00018811384470416705,
+      "loss": 0.7408,
+      "step": 2412
+    },
+    {
+      "epoch": 0.774763204366672,
+      "grad_norm": 1.06851327419281,
+      "learning_rate": 0.00018809733346543013,
+      "loss": 0.857,
+      "step": 2413
+    },
+    {
+      "epoch": 0.7750842831915236,
+      "grad_norm": 1.3492529392242432,
+      "learning_rate": 0.00018808081149230036,
+      "loss": 0.8873,
+      "step": 2414
+    },
+    {
+      "epoch": 0.775405362016375,
+      "grad_norm": 1.1693357229232788,
+      "learning_rate": 0.00018806427878679093,
+      "loss": 1.1389,
+      "step": 2415
+    },
+    {
+      "epoch": 0.7757264408412265,
+      "grad_norm": 1.2862006425857544,
+      "learning_rate": 0.0001880477353509162,
+      "loss": 0.9218,
+      "step": 2416
+    },
+    {
+      "epoch": 0.776047519666078,
+      "grad_norm": 0.9942718744277954,
+      "learning_rate": 0.00018803118118669202,
+      "loss": 0.7227,
+      "step": 2417
+    },
+    {
+      "epoch": 0.7763685984909295,
+      "grad_norm": 1.2188372611999512,
+      "learning_rate": 0.00018801461629613546,
+      "loss": 0.9224,
+      "step": 2418
+    },
+    {
+      "epoch": 0.776689677315781,
+      "grad_norm": 1.206196904182434,
+      "learning_rate": 0.00018799804068126485,
+      "loss": 0.9985,
+      "step": 2419
+    },
+    {
+      "epoch": 0.7770107561406325,
+      "grad_norm": 1.327648639678955,
+      "learning_rate": 0.0001879814543440999,
+      "loss": 0.932,
+      "step": 2420
+    },
+    {
+      "epoch": 0.7773318349654841,
+      "grad_norm": 1.5388193130493164,
+      "learning_rate": 0.00018796485728666165,
+      "loss": 0.9052,
+      "step": 2421
+    },
+    {
+      "epoch": 0.7776529137903355,
+      "grad_norm": 1.7653731107711792,
+      "learning_rate": 0.00018794824951097236,
+      "loss": 1.032,
+      "step": 2422
+    },
+    {
+      "epoch": 0.7779739926151871,
+      "grad_norm": 1.1087538003921509,
+      "learning_rate": 0.00018793163101905563,
+      "loss": 1.283,
+      "step": 2423
+    },
+    {
+      "epoch": 0.7782950714400385,
+      "grad_norm": 1.3659240007400513,
+      "learning_rate": 0.0001879150018129364,
+      "loss": 1.0172,
+      "step": 2424
+    },
+    {
+      "epoch": 0.7786161502648901,
+      "grad_norm": 0.9709173440933228,
+      "learning_rate": 0.00018789836189464086,
+      "loss": 0.9285,
+      "step": 2425
+    },
+    {
+      "epoch": 0.7789372290897415,
+      "grad_norm": 1.2485631704330444,
+      "learning_rate": 0.00018788171126619653,
+      "loss": 0.9621,
+      "step": 2426
+    },
+    {
+      "epoch": 0.779258307914593,
+      "grad_norm": 1.27434504032135,
+      "learning_rate": 0.0001878650499296323,
+      "loss": 0.984,
+      "step": 2427
+    },
+    {
+      "epoch": 0.7795793867394445,
+      "grad_norm": 1.2106627225875854,
+      "learning_rate": 0.00018784837788697823,
+      "loss": 0.9109,
+      "step": 2428
+    },
+    {
+      "epoch": 0.779900465564296,
+      "grad_norm": 0.9945119023323059,
+      "learning_rate": 0.00018783169514026578,
+      "loss": 1.0478,
+      "step": 2429
+    },
+    {
+      "epoch": 0.7802215443891475,
+      "grad_norm": 2.2942955493927,
+      "learning_rate": 0.00018781500169152773,
+      "loss": 1.0132,
+      "step": 2430
+    },
+    {
+      "epoch": 0.780542623213999,
+      "grad_norm": 1.0083764791488647,
+      "learning_rate": 0.00018779829754279805,
+      "loss": 1.0231,
+      "step": 2431
+    },
+    {
+      "epoch": 0.7808637020388506,
+      "grad_norm": 1.2520384788513184,
+      "learning_rate": 0.00018778158269611218,
+      "loss": 0.8075,
+      "step": 2432
+    },
+    {
+      "epoch": 0.781184780863702,
+      "grad_norm": 1.1380735635757446,
+      "learning_rate": 0.00018776485715350671,
+      "loss": 0.8545,
+      "step": 2433
+    },
+    {
+      "epoch": 0.7815058596885536,
+      "grad_norm": 1.0106878280639648,
+      "learning_rate": 0.00018774812091701962,
+      "loss": 0.8459,
+      "step": 2434
+    },
+    {
+      "epoch": 0.781826938513405,
+      "grad_norm": 1.134502649307251,
+      "learning_rate": 0.00018773137398869015,
+      "loss": 0.9232,
+      "step": 2435
+    },
+    {
+      "epoch": 0.7821480173382566,
+      "grad_norm": 1.356572151184082,
+      "learning_rate": 0.00018771461637055888,
+      "loss": 0.9854,
+      "step": 2436
+    },
+    {
+      "epoch": 0.782469096163108,
+      "grad_norm": 1.156056523323059,
+      "learning_rate": 0.0001876978480646677,
+      "loss": 0.9877,
+      "step": 2437
+    },
+    {
+      "epoch": 0.7827901749879596,
+      "grad_norm": 1.1275618076324463,
+      "learning_rate": 0.00018768106907305973,
+      "loss": 0.7909,
+      "step": 2438
+    },
+    {
+      "epoch": 0.783111253812811,
+      "grad_norm": 0.8512035012245178,
+      "learning_rate": 0.00018766427939777945,
+      "loss": 0.9474,
+      "step": 2439
+    },
+    {
+      "epoch": 0.7834323326376625,
+      "grad_norm": 1.328540563583374,
+      "learning_rate": 0.00018764747904087263,
+      "loss": 0.9969,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7837534114625141,
+      "grad_norm": 0.8241159319877625,
+      "learning_rate": 0.00018763066800438636,
+      "loss": 0.5979,
+      "step": 2441
+    },
+    {
+      "epoch": 0.7840744902873655,
+      "grad_norm": 1.58833646774292,
+      "learning_rate": 0.00018761384629036902,
+      "loss": 0.9713,
+      "step": 2442
+    },
+    {
+      "epoch": 0.7843955691122171,
+      "grad_norm": 1.2584527730941772,
+      "learning_rate": 0.00018759701390087027,
+      "loss": 1.0281,
+      "step": 2443
+    },
+    {
+      "epoch": 0.7847166479370685,
+      "grad_norm": 0.8919162154197693,
+      "learning_rate": 0.0001875801708379411,
+      "loss": 0.9197,
+      "step": 2444
+    },
+    {
+      "epoch": 0.7850377267619201,
+      "grad_norm": 2.073561906814575,
+      "learning_rate": 0.00018756331710363374,
+      "loss": 0.9238,
+      "step": 2445
+    },
+    {
+      "epoch": 0.7853588055867715,
+      "grad_norm": 0.9129263162612915,
+      "learning_rate": 0.0001875464527000018,
+      "loss": 0.8524,
+      "step": 2446
+    },
+    {
+      "epoch": 0.7856798844116231,
+      "grad_norm": 1.1756608486175537,
+      "learning_rate": 0.00018752957762910018,
+      "loss": 0.9372,
+      "step": 2447
+    },
+    {
+      "epoch": 0.7860009632364745,
+      "grad_norm": 1.1006115674972534,
+      "learning_rate": 0.000187512691892985,
+      "loss": 0.8847,
+      "step": 2448
+    },
+    {
+      "epoch": 0.786322042061326,
+      "grad_norm": 1.1314774751663208,
+      "learning_rate": 0.0001874957954937138,
+      "loss": 0.724,
+      "step": 2449
+    },
+    {
+      "epoch": 0.7866431208861776,
+      "grad_norm": 0.5091024041175842,
+      "learning_rate": 0.0001874788884333453,
+      "loss": 0.4491,
+      "step": 2450
+    },
+    {
+      "epoch": 0.786964199711029,
+      "grad_norm": 0.8892128467559814,
+      "learning_rate": 0.00018746197071393958,
+      "loss": 1.5905,
+      "step": 2451
+    },
+    {
+      "epoch": 0.7872852785358806,
+      "grad_norm": 0.9708239436149597,
+      "learning_rate": 0.00018744504233755805,
+      "loss": 1.2197,
+      "step": 2452
+    },
+    {
+      "epoch": 0.787606357360732,
+      "grad_norm": 1.2645493745803833,
+      "learning_rate": 0.00018742810330626337,
+      "loss": 0.8199,
+      "step": 2453
+    },
+    {
+      "epoch": 0.7879274361855836,
+      "grad_norm": 0.9297351837158203,
+      "learning_rate": 0.00018741115362211949,
+      "loss": 0.6392,
+      "step": 2454
+    },
+    {
+      "epoch": 0.788248515010435,
+      "grad_norm": 1.3193695545196533,
+      "learning_rate": 0.0001873941932871917,
+      "loss": 0.7132,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7885695938352866,
+      "grad_norm": 1.7561746835708618,
+      "learning_rate": 0.00018737722230354655,
+      "loss": 0.607,
+      "step": 2456
+    },
+    {
+      "epoch": 0.788890672660138,
+      "grad_norm": 1.2228854894638062,
+      "learning_rate": 0.00018736024067325188,
+      "loss": 1.0434,
+      "step": 2457
+    },
+    {
+      "epoch": 0.7892117514849896,
+      "grad_norm": 1.381565809249878,
+      "learning_rate": 0.0001873432483983769,
+      "loss": 1.0155,
+      "step": 2458
+    },
+    {
+      "epoch": 0.7895328303098411,
+      "grad_norm": 1.2549446821212769,
+      "learning_rate": 0.00018732624548099204,
+      "loss": 0.6954,
+      "step": 2459
+    },
+    {
+      "epoch": 0.7898539091346926,
+      "grad_norm": 1.154805064201355,
+      "learning_rate": 0.00018730923192316902,
+      "loss": 0.6555,
+      "step": 2460
+    },
+    {
+      "epoch": 0.7901749879595441,
+      "grad_norm": 0.7833784818649292,
+      "learning_rate": 0.00018729220772698097,
+      "loss": 0.8146,
+      "step": 2461
+    },
+    {
+      "epoch": 0.7904960667843955,
+      "grad_norm": 0.9661710858345032,
+      "learning_rate": 0.0001872751728945022,
+      "loss": 0.9448,
+      "step": 2462
+    },
+    {
+      "epoch": 0.7908171456092471,
+      "grad_norm": 0.9815911054611206,
+      "learning_rate": 0.00018725812742780834,
+      "loss": 0.9413,
+      "step": 2463
+    },
+    {
+      "epoch": 0.7911382244340985,
+      "grad_norm": 0.823572039604187,
+      "learning_rate": 0.0001872410713289763,
+      "loss": 0.8122,
+      "step": 2464
+    },
+    {
+      "epoch": 0.7914593032589501,
+      "grad_norm": 0.8957955837249756,
+      "learning_rate": 0.0001872240046000844,
+      "loss": 0.8169,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7917803820838015,
+      "grad_norm": 1.1782604455947876,
+      "learning_rate": 0.00018720692724321207,
+      "loss": 0.8323,
+      "step": 2466
+    },
+    {
+      "epoch": 0.7921014609086531,
+      "grad_norm": 0.887837827205658,
+      "learning_rate": 0.0001871898392604402,
+      "loss": 0.8662,
+      "step": 2467
+    },
+    {
+      "epoch": 0.7924225397335046,
+      "grad_norm": 0.903867781162262,
+      "learning_rate": 0.0001871727406538509,
+      "loss": 1.0044,
+      "step": 2468
+    },
+    {
+      "epoch": 0.7927436185583561,
+      "grad_norm": 0.8444812297821045,
+      "learning_rate": 0.00018715563142552758,
+      "loss": 0.9237,
+      "step": 2469
+    },
+    {
+      "epoch": 0.7930646973832076,
+      "grad_norm": 1.5705596208572388,
+      "learning_rate": 0.00018713851157755492,
+      "loss": 0.9272,
+      "step": 2470
+    },
+    {
+      "epoch": 0.7933857762080591,
+      "grad_norm": 0.9843676090240479,
+      "learning_rate": 0.00018712138111201895,
+      "loss": 0.9544,
+      "step": 2471
+    },
+    {
+      "epoch": 0.7937068550329106,
+      "grad_norm": 1.2598280906677246,
+      "learning_rate": 0.00018710424003100698,
+      "loss": 0.8223,
+      "step": 2472
+    },
+    {
+      "epoch": 0.794027933857762,
+      "grad_norm": 0.903620183467865,
+      "learning_rate": 0.00018708708833660754,
+      "loss": 0.6692,
+      "step": 2473
+    },
+    {
+      "epoch": 0.7943490126826136,
+      "grad_norm": 1.244788408279419,
+      "learning_rate": 0.00018706992603091058,
+      "loss": 1.0025,
+      "step": 2474
+    },
+    {
+      "epoch": 0.794670091507465,
+      "grad_norm": 1.1981257200241089,
+      "learning_rate": 0.00018705275311600722,
+      "loss": 0.8593,
+      "step": 2475
+    },
+    {
+      "epoch": 0.7949911703323166,
+      "grad_norm": 0.8418868184089661,
+      "learning_rate": 0.00018703556959398998,
+      "loss": 0.7955,
+      "step": 2476
+    },
+    {
+      "epoch": 0.7953122491571681,
+      "grad_norm": 0.8239766955375671,
+      "learning_rate": 0.0001870183754669526,
+      "loss": 0.796,
+      "step": 2477
+    },
+    {
+      "epoch": 0.7956333279820196,
+      "grad_norm": 1.1803929805755615,
+      "learning_rate": 0.0001870011707369901,
+      "loss": 1.0684,
+      "step": 2478
+    },
+    {
+      "epoch": 0.7959544068068711,
+      "grad_norm": 1.2590785026550293,
+      "learning_rate": 0.0001869839554061988,
+      "loss": 0.8635,
+      "step": 2479
+    },
+    {
+      "epoch": 0.7962754856317226,
+      "grad_norm": 1.4843522310256958,
+      "learning_rate": 0.00018696672947667646,
+      "loss": 0.7823,
+      "step": 2480
+    },
+    {
+      "epoch": 0.7965965644565741,
+      "grad_norm": 0.9807465076446533,
+      "learning_rate": 0.0001869494929505219,
+      "loss": 0.8197,
+      "step": 2481
+    },
+    {
+      "epoch": 0.7969176432814256,
+      "grad_norm": 1.2640116214752197,
+      "learning_rate": 0.0001869322458298354,
+      "loss": 0.9056,
+      "step": 2482
+    },
+    {
+      "epoch": 0.7972387221062771,
+      "grad_norm": 1.0154175758361816,
+      "learning_rate": 0.0001869149881167184,
+      "loss": 0.8573,
+      "step": 2483
+    },
+    {
+      "epoch": 0.7975598009311285,
+      "grad_norm": 1.578016996383667,
+      "learning_rate": 0.00018689771981327376,
+      "loss": 0.9812,
+      "step": 2484
+    },
+    {
+      "epoch": 0.7978808797559801,
+      "grad_norm": 0.9746622443199158,
+      "learning_rate": 0.00018688044092160551,
+      "loss": 0.8187,
+      "step": 2485
+    },
+    {
+      "epoch": 0.7982019585808315,
+      "grad_norm": 1.5574668645858765,
+      "learning_rate": 0.00018686315144381913,
+      "loss": 0.7508,
+      "step": 2486
+    },
+    {
+      "epoch": 0.7985230374056831,
+      "grad_norm": 0.9885238409042358,
+      "learning_rate": 0.00018684585138202122,
+      "loss": 0.8462,
+      "step": 2487
+    },
+    {
+      "epoch": 0.7988441162305346,
+      "grad_norm": 1.220053791999817,
+      "learning_rate": 0.00018682854073831973,
+      "loss": 0.9497,
+      "step": 2488
+    },
+    {
+      "epoch": 0.7991651950553861,
+      "grad_norm": 0.8417456746101379,
+      "learning_rate": 0.00018681121951482393,
+      "loss": 0.6397,
+      "step": 2489
+    },
+    {
+      "epoch": 0.7994862738802376,
+      "grad_norm": 1.1695717573165894,
+      "learning_rate": 0.00018679388771364436,
+      "loss": 0.6754,
+      "step": 2490
+    },
+    {
+      "epoch": 0.7998073527050891,
+      "grad_norm": 1.0518317222595215,
+      "learning_rate": 0.00018677654533689287,
+      "loss": 0.8917,
+      "step": 2491
+    },
+    {
+      "epoch": 0.8001284315299406,
+      "grad_norm": 1.123487114906311,
+      "learning_rate": 0.0001867591923866825,
+      "loss": 0.977,
+      "step": 2492
+    },
+    {
+      "epoch": 0.8004495103547921,
+      "grad_norm": 1.3612045049667358,
+      "learning_rate": 0.00018674182886512774,
+      "loss": 0.8039,
+      "step": 2493
+    },
+    {
+      "epoch": 0.8007705891796436,
+      "grad_norm": 1.34379243850708,
+      "learning_rate": 0.00018672445477434425,
+      "loss": 0.8616,
+      "step": 2494
+    },
+    {
+      "epoch": 0.801091668004495,
+      "grad_norm": 0.8489518761634827,
+      "learning_rate": 0.000186707070116449,
+      "loss": 0.7731,
+      "step": 2495
+    },
+    {
+      "epoch": 0.8014127468293466,
+      "grad_norm": 1.2689493894577026,
+      "learning_rate": 0.00018668967489356028,
+      "loss": 0.7308,
+      "step": 2496
+    },
+    {
+      "epoch": 0.8017338256541982,
+      "grad_norm": 1.7729278802871704,
+      "learning_rate": 0.00018667226910779765,
+      "loss": 0.6356,
+      "step": 2497
+    },
+    {
+      "epoch": 0.8020549044790496,
+      "grad_norm": 0.9951494932174683,
+      "learning_rate": 0.00018665485276128188,
+      "loss": 0.9106,
+      "step": 2498
+    },
+    {
+      "epoch": 0.8023759833039011,
+      "grad_norm": 1.5119075775146484,
+      "learning_rate": 0.00018663742585613518,
+      "loss": 0.7492,
+      "step": 2499
+    },
+    {
+      "epoch": 0.8026970621287526,
+      "grad_norm": 1.1886110305786133,
+      "learning_rate": 0.00018661998839448094,
+      "loss": 0.5986,
+      "step": 2500
+    },
+    {
+      "epoch": 0.8030181409536041,
+      "grad_norm": 1.4970862865447998,
+      "learning_rate": 0.00018660254037844388,
+      "loss": 1.4061,
+      "step": 2501
+    },
+    {
+      "epoch": 0.8033392197784556,
+      "grad_norm": 1.2722395658493042,
+      "learning_rate": 0.00018658508181014995,
+      "loss": 1.2115,
+      "step": 2502
+    },
+    {
+      "epoch": 0.8036602986033071,
+      "grad_norm": 1.3451300859451294,
+      "learning_rate": 0.00018656761269172643,
+      "loss": 0.9685,
+      "step": 2503
+    },
+    {
+      "epoch": 0.8039813774281586,
+      "grad_norm": 1.2562284469604492,
+      "learning_rate": 0.0001865501330253019,
+      "loss": 0.7432,
+      "step": 2504
+    },
+    {
+      "epoch": 0.8043024562530101,
+      "grad_norm": 1.0537221431732178,
+      "learning_rate": 0.00018653264281300622,
+      "loss": 0.7564,
+      "step": 2505
+    },
+    {
+      "epoch": 0.8046235350778617,
+      "grad_norm": 1.2182271480560303,
+      "learning_rate": 0.00018651514205697046,
+      "loss": 0.7227,
+      "step": 2506
+    },
+    {
+      "epoch": 0.8049446139027131,
+      "grad_norm": 0.9667792320251465,
+      "learning_rate": 0.00018649763075932708,
+      "loss": 0.6023,
+      "step": 2507
+    },
+    {
+      "epoch": 0.8052656927275647,
+      "grad_norm": 1.413209319114685,
+      "learning_rate": 0.00018648010892220978,
+      "loss": 0.9355,
+      "step": 2508
+    },
+    {
+      "epoch": 0.8055867715524161,
+      "grad_norm": 1.3829725980758667,
+      "learning_rate": 0.0001864625765477535,
+      "loss": 0.7942,
+      "step": 2509
+    },
+    {
+      "epoch": 0.8059078503772676,
+      "grad_norm": 0.8415265083312988,
+      "learning_rate": 0.00018644503363809457,
+      "loss": 0.7862,
+      "step": 2510
+    },
+    {
+      "epoch": 0.8062289292021191,
+      "grad_norm": 1.514174222946167,
+      "learning_rate": 0.0001864274801953705,
+      "loss": 1.0991,
+      "step": 2511
+    },
+    {
+      "epoch": 0.8065500080269706,
+      "grad_norm": 1.3786250352859497,
+      "learning_rate": 0.0001864099162217201,
+      "loss": 0.8292,
+      "step": 2512
+    },
+    {
+      "epoch": 0.8068710868518221,
+      "grad_norm": 1.0749139785766602,
+      "learning_rate": 0.00018639234171928353,
+      "loss": 0.9804,
+      "step": 2513
+    },
+    {
+      "epoch": 0.8071921656766736,
+      "grad_norm": 0.8056579828262329,
+      "learning_rate": 0.0001863747566902022,
+      "loss": 0.8092,
+      "step": 2514
+    },
+    {
+      "epoch": 0.8075132445015252,
+      "grad_norm": 1.0548845529556274,
+      "learning_rate": 0.00018635716113661873,
+      "loss": 0.7606,
+      "step": 2515
+    },
+    {
+      "epoch": 0.8078343233263766,
+      "grad_norm": 0.9919301271438599,
+      "learning_rate": 0.00018633955506067718,
+      "loss": 1.0425,
+      "step": 2516
+    },
+    {
+      "epoch": 0.8081554021512282,
+      "grad_norm": 0.918125569820404,
+      "learning_rate": 0.0001863219384645227,
+      "loss": 0.8707,
+      "step": 2517
+    },
+    {
+      "epoch": 0.8084764809760796,
+      "grad_norm": 1.2536126375198364,
+      "learning_rate": 0.0001863043113503019,
+      "loss": 1.1534,
+      "step": 2518
+    },
+    {
+      "epoch": 0.8087975598009312,
+      "grad_norm": 1.1653470993041992,
+      "learning_rate": 0.0001862866737201625,
+      "loss": 0.8979,
+      "step": 2519
+    },
+    {
+      "epoch": 0.8091186386257826,
+      "grad_norm": 1.016139030456543,
+      "learning_rate": 0.00018626902557625368,
+      "loss": 0.7874,
+      "step": 2520
+    },
+    {
+      "epoch": 0.8094397174506341,
+      "grad_norm": 1.2816565036773682,
+      "learning_rate": 0.00018625136692072575,
+      "loss": 0.8338,
+      "step": 2521
+    },
+    {
+      "epoch": 0.8097607962754856,
+      "grad_norm": 1.5095349550247192,
+      "learning_rate": 0.0001862336977557304,
+      "loss": 1.0349,
+      "step": 2522
+    },
+    {
+      "epoch": 0.8100818751003371,
+      "grad_norm": 1.753609299659729,
+      "learning_rate": 0.00018621601808342056,
+      "loss": 0.932,
+      "step": 2523
+    },
+    {
+      "epoch": 0.8104029539251887,
+      "grad_norm": 1.1441911458969116,
+      "learning_rate": 0.00018619832790595043,
+      "loss": 1.0346,
+      "step": 2524
+    },
+    {
+      "epoch": 0.8107240327500401,
+      "grad_norm": 1.1393861770629883,
+      "learning_rate": 0.0001861806272254755,
+      "loss": 0.8634,
+      "step": 2525
+    },
+    {
+      "epoch": 0.8110451115748917,
+      "grad_norm": 0.9023036956787109,
+      "learning_rate": 0.00018616291604415258,
+      "loss": 0.7442,
+      "step": 2526
+    },
+    {
+      "epoch": 0.8113661903997431,
+      "grad_norm": 0.9922143220901489,
+      "learning_rate": 0.0001861451943641397,
+      "loss": 0.893,
+      "step": 2527
+    },
+    {
+      "epoch": 0.8116872692245947,
+      "grad_norm": 1.2461748123168945,
+      "learning_rate": 0.00018612746218759618,
+      "loss": 1.0521,
+      "step": 2528
+    },
+    {
+      "epoch": 0.8120083480494461,
+      "grad_norm": 1.3164596557617188,
+      "learning_rate": 0.00018610971951668265,
+      "loss": 0.9787,
+      "step": 2529
+    },
+    {
+      "epoch": 0.8123294268742977,
+      "grad_norm": 1.3357511758804321,
+      "learning_rate": 0.000186091966353561,
+      "loss": 0.8539,
+      "step": 2530
+    },
+    {
+      "epoch": 0.8126505056991491,
+      "grad_norm": 0.9411560297012329,
+      "learning_rate": 0.0001860742027003944,
+      "loss": 1.0298,
+      "step": 2531
+    },
+    {
+      "epoch": 0.8129715845240006,
+      "grad_norm": 0.8316774368286133,
+      "learning_rate": 0.00018605642855934725,
+      "loss": 0.8219,
+      "step": 2532
+    },
+    {
+      "epoch": 0.8132926633488522,
+      "grad_norm": 1.3453702926635742,
+      "learning_rate": 0.00018603864393258534,
+      "loss": 0.9055,
+      "step": 2533
+    },
+    {
+      "epoch": 0.8136137421737036,
+      "grad_norm": 0.9413175582885742,
+      "learning_rate": 0.00018602084882227566,
+      "loss": 0.9423,
+      "step": 2534
+    },
+    {
+      "epoch": 0.8139348209985552,
+      "grad_norm": 0.9945281744003296,
+      "learning_rate": 0.00018600304323058647,
+      "loss": 0.9676,
+      "step": 2535
+    },
+    {
+      "epoch": 0.8142558998234066,
+      "grad_norm": 1.2609598636627197,
+      "learning_rate": 0.00018598522715968736,
+      "loss": 0.877,
+      "step": 2536
+    },
+    {
+      "epoch": 0.8145769786482582,
+      "grad_norm": 1.2044451236724854,
+      "learning_rate": 0.0001859674006117491,
+      "loss": 1.0152,
+      "step": 2537
+    },
+    {
+      "epoch": 0.8148980574731096,
+      "grad_norm": 1.1789966821670532,
+      "learning_rate": 0.00018594956358894388,
+      "loss": 0.7836,
+      "step": 2538
+    },
+    {
+      "epoch": 0.8152191362979612,
+      "grad_norm": 2.6686882972717285,
+      "learning_rate": 0.00018593171609344503,
+      "loss": 0.7848,
+      "step": 2539
+    },
+    {
+      "epoch": 0.8155402151228126,
+      "grad_norm": 0.900510311126709,
+      "learning_rate": 0.00018591385812742725,
+      "loss": 0.8703,
+      "step": 2540
+    },
+    {
+      "epoch": 0.8158612939476642,
+      "grad_norm": 0.9134002923965454,
+      "learning_rate": 0.00018589598969306645,
+      "loss": 0.7962,
+      "step": 2541
+    },
+    {
+      "epoch": 0.8161823727725157,
+      "grad_norm": 1.1984633207321167,
+      "learning_rate": 0.00018587811079253985,
+      "loss": 0.984,
+      "step": 2542
+    },
+    {
+      "epoch": 0.8165034515973671,
+      "grad_norm": 1.6189078092575073,
+      "learning_rate": 0.00018586022142802597,
+      "loss": 1.013,
+      "step": 2543
+    },
+    {
+      "epoch": 0.8168245304222187,
+      "grad_norm": 1.4700372219085693,
+      "learning_rate": 0.00018584232160170452,
+      "loss": 1.1043,
+      "step": 2544
+    },
+    {
+      "epoch": 0.8171456092470701,
+      "grad_norm": 1.035675287246704,
+      "learning_rate": 0.0001858244113157566,
+      "loss": 0.8351,
+      "step": 2545
+    },
+    {
+      "epoch": 0.8174666880719217,
+      "grad_norm": 1.505395531654358,
+      "learning_rate": 0.00018580649057236447,
+      "loss": 0.8241,
+      "step": 2546
+    },
+    {
+      "epoch": 0.8177877668967731,
+      "grad_norm": 1.6375924348831177,
+      "learning_rate": 0.00018578855937371173,
+      "loss": 0.6124,
+      "step": 2547
+    },
+    {
+      "epoch": 0.8181088457216247,
+      "grad_norm": 0.5977188944816589,
+      "learning_rate": 0.0001857706177219833,
+      "loss": 0.5429,
+      "step": 2548
+    },
+    {
+      "epoch": 0.8184299245464761,
+      "grad_norm": 0.6796979308128357,
+      "learning_rate": 0.00018575266561936523,
+      "loss": 0.5959,
+      "step": 2549
+    },
+    {
+      "epoch": 0.8187510033713277,
+      "grad_norm": 0.6622598171234131,
+      "learning_rate": 0.00018573470306804498,
+      "loss": 0.5182,
+      "step": 2550
+    },
+    {
+      "epoch": 0.8190720821961791,
+      "grad_norm": 0.9923622012138367,
+      "learning_rate": 0.00018571673007021123,
+      "loss": 1.2836,
+      "step": 2551
+    },
+    {
+      "epoch": 0.8193931610210307,
+      "grad_norm": 1.1485919952392578,
+      "learning_rate": 0.00018569874662805393,
+      "loss": 1.3207,
+      "step": 2552
+    },
+    {
+      "epoch": 0.8197142398458822,
+      "grad_norm": 0.9416031837463379,
+      "learning_rate": 0.0001856807527437643,
+      "loss": 0.9675,
+      "step": 2553
+    },
+    {
+      "epoch": 0.8200353186707336,
+      "grad_norm": 1.0910120010375977,
+      "learning_rate": 0.00018566274841953483,
+      "loss": 0.8907,
+      "step": 2554
+    },
+    {
+      "epoch": 0.8203563974955852,
+      "grad_norm": 1.061464786529541,
+      "learning_rate": 0.00018564473365755935,
+      "loss": 0.6902,
+      "step": 2555
+    },
+    {
+      "epoch": 0.8206774763204366,
+      "grad_norm": 1.0407187938690186,
+      "learning_rate": 0.00018562670846003284,
+      "loss": 0.6008,
+      "step": 2556
+    },
+    {
+      "epoch": 0.8209985551452882,
+      "grad_norm": 1.0993876457214355,
+      "learning_rate": 0.0001856086728291516,
+      "loss": 0.7588,
+      "step": 2557
+    },
+    {
+      "epoch": 0.8213196339701396,
+      "grad_norm": 1.2485312223434448,
+      "learning_rate": 0.00018559062676711332,
+      "loss": 0.9553,
+      "step": 2558
+    },
+    {
+      "epoch": 0.8216407127949912,
+      "grad_norm": 1.184083342552185,
+      "learning_rate": 0.00018557257027611675,
+      "loss": 1.0036,
+      "step": 2559
+    },
+    {
+      "epoch": 0.8219617916198426,
+      "grad_norm": 1.219896674156189,
+      "learning_rate": 0.00018555450335836206,
+      "loss": 0.9131,
+      "step": 2560
+    },
+    {
+      "epoch": 0.8222828704446942,
+      "grad_norm": 0.8921108841896057,
+      "learning_rate": 0.00018553642601605068,
+      "loss": 0.7521,
+      "step": 2561
+    },
+    {
+      "epoch": 0.8226039492695457,
+      "grad_norm": 1.3684065341949463,
+      "learning_rate": 0.0001855183382513852,
+      "loss": 0.8931,
+      "step": 2562
+    },
+    {
+      "epoch": 0.8229250280943972,
+      "grad_norm": 0.8778723478317261,
+      "learning_rate": 0.00018550024006656966,
+      "loss": 0.8721,
+      "step": 2563
+    },
+    {
+      "epoch": 0.8232461069192487,
+      "grad_norm": 1.4915300607681274,
+      "learning_rate": 0.00018548213146380918,
+      "loss": 0.9267,
+      "step": 2564
+    },
+    {
+      "epoch": 0.8235671857441001,
+      "grad_norm": 1.0653166770935059,
+      "learning_rate": 0.0001854640124453103,
+      "loss": 0.9316,
+      "step": 2565
+    },
+    {
+      "epoch": 0.8238882645689517,
+      "grad_norm": 1.8792741298675537,
+      "learning_rate": 0.00018544588301328075,
+      "loss": 0.9906,
+      "step": 2566
+    },
+    {
+      "epoch": 0.8242093433938031,
+      "grad_norm": 1.2481502294540405,
+      "learning_rate": 0.0001854277431699295,
+      "loss": 0.9631,
+      "step": 2567
+    },
+    {
+      "epoch": 0.8245304222186547,
+      "grad_norm": 1.4939336776733398,
+      "learning_rate": 0.00018540959291746693,
+      "loss": 0.9405,
+      "step": 2568
+    },
+    {
+      "epoch": 0.8248515010435061,
+      "grad_norm": 1.0083293914794922,
+      "learning_rate": 0.0001853914322581045,
+      "loss": 0.9924,
+      "step": 2569
+    },
+    {
+      "epoch": 0.8251725798683577,
+      "grad_norm": 1.3089494705200195,
+      "learning_rate": 0.00018537326119405506,
+      "loss": 0.899,
+      "step": 2570
+    },
+    {
+      "epoch": 0.8254936586932092,
+      "grad_norm": 1.0954056978225708,
+      "learning_rate": 0.00018535507972753274,
+      "loss": 0.977,
+      "step": 2571
+    },
+    {
+      "epoch": 0.8258147375180607,
+      "grad_norm": 0.9227792620658875,
+      "learning_rate": 0.00018533688786075288,
+      "loss": 0.7727,
+      "step": 2572
+    },
+    {
+      "epoch": 0.8261358163429122,
+      "grad_norm": 1.0734128952026367,
+      "learning_rate": 0.00018531868559593204,
+      "loss": 1.0625,
+      "step": 2573
+    },
+    {
+      "epoch": 0.8264568951677637,
+      "grad_norm": 0.9797103404998779,
+      "learning_rate": 0.00018530047293528819,
+      "loss": 1.0345,
+      "step": 2574
+    },
+    {
+      "epoch": 0.8267779739926152,
+      "grad_norm": 1.2002710103988647,
+      "learning_rate": 0.00018528224988104044,
+      "loss": 0.9635,
+      "step": 2575
+    },
+    {
+      "epoch": 0.8270990528174667,
+      "grad_norm": 1.1027469635009766,
+      "learning_rate": 0.00018526401643540922,
+      "loss": 0.8992,
+      "step": 2576
+    },
+    {
+      "epoch": 0.8274201316423182,
+      "grad_norm": 1.8506336212158203,
+      "learning_rate": 0.00018524577260061627,
+      "loss": 0.9184,
+      "step": 2577
+    },
+    {
+      "epoch": 0.8277412104671696,
+      "grad_norm": 1.1233233213424683,
+      "learning_rate": 0.0001852275183788845,
+      "loss": 0.9805,
+      "step": 2578
+    },
+    {
+      "epoch": 0.8280622892920212,
+      "grad_norm": 1.0749657154083252,
+      "learning_rate": 0.0001852092537724381,
+      "loss": 0.999,
+      "step": 2579
+    },
+    {
+      "epoch": 0.8283833681168727,
+      "grad_norm": 1.4672601222991943,
+      "learning_rate": 0.00018519097878350263,
+      "loss": 1.0324,
+      "step": 2580
+    },
+    {
+      "epoch": 0.8287044469417242,
+      "grad_norm": 1.0896575450897217,
+      "learning_rate": 0.00018517269341430476,
+      "loss": 0.9614,
+      "step": 2581
+    },
+    {
+      "epoch": 0.8290255257665757,
+      "grad_norm": 1.345746397972107,
+      "learning_rate": 0.00018515439766707262,
+      "loss": 0.9667,
+      "step": 2582
+    },
+    {
+      "epoch": 0.8293466045914272,
+      "grad_norm": 0.8869307041168213,
+      "learning_rate": 0.00018513609154403534,
+      "loss": 0.8017,
+      "step": 2583
+    },
+    {
+      "epoch": 0.8296676834162787,
+      "grad_norm": 1.2558448314666748,
+      "learning_rate": 0.00018511777504742362,
+      "loss": 0.9269,
+      "step": 2584
+    },
+    {
+      "epoch": 0.8299887622411302,
+      "grad_norm": 1.5162900686264038,
+      "learning_rate": 0.00018509944817946922,
+      "loss": 0.9928,
+      "step": 2585
+    },
+    {
+      "epoch": 0.8303098410659817,
+      "grad_norm": 1.0953887701034546,
+      "learning_rate": 0.00018508111094240514,
+      "loss": 0.7442,
+      "step": 2586
+    },
+    {
+      "epoch": 0.8306309198908332,
+      "grad_norm": 0.9997647404670715,
+      "learning_rate": 0.00018506276333846579,
+      "loss": 0.8279,
+      "step": 2587
+    },
+    {
+      "epoch": 0.8309519987156847,
+      "grad_norm": 0.8126912117004395,
+      "learning_rate": 0.00018504440536988673,
+      "loss": 0.6527,
+      "step": 2588
+    },
+    {
+      "epoch": 0.8312730775405363,
+      "grad_norm": 0.7459484338760376,
+      "learning_rate": 0.00018502603703890488,
+      "loss": 0.7202,
+      "step": 2589
+    },
+    {
+      "epoch": 0.8315941563653877,
+      "grad_norm": 1.0116833448410034,
+      "learning_rate": 0.00018500765834775828,
+      "loss": 0.8612,
+      "step": 2590
+    },
+    {
+      "epoch": 0.8319152351902392,
+      "grad_norm": 1.7899558544158936,
+      "learning_rate": 0.00018498926929868642,
+      "loss": 1.0347,
+      "step": 2591
+    },
+    {
+      "epoch": 0.8322363140150907,
+      "grad_norm": 1.1006872653961182,
+      "learning_rate": 0.00018497086989392988,
+      "loss": 0.9109,
+      "step": 2592
+    },
+    {
+      "epoch": 0.8325573928399422,
+      "grad_norm": 2.0811855792999268,
+      "learning_rate": 0.00018495246013573054,
+      "loss": 0.9743,
+      "step": 2593
+    },
+    {
+      "epoch": 0.8328784716647937,
+      "grad_norm": 1.6364414691925049,
+      "learning_rate": 0.00018493404002633166,
+      "loss": 1.1582,
+      "step": 2594
+    },
+    {
+      "epoch": 0.8331995504896452,
+      "grad_norm": 1.5922627449035645,
+      "learning_rate": 0.00018491560956797765,
+      "loss": 0.7103,
+      "step": 2595
+    },
+    {
+      "epoch": 0.8335206293144967,
+      "grad_norm": 1.3164536952972412,
+      "learning_rate": 0.00018489716876291415,
+      "loss": 1.0022,
+      "step": 2596
+    },
+    {
+      "epoch": 0.8338417081393482,
+      "grad_norm": 0.8407567739486694,
+      "learning_rate": 0.0001848787176133882,
+      "loss": 0.7196,
+      "step": 2597
+    },
+    {
+      "epoch": 0.8341627869641998,
+      "grad_norm": 1.0273271799087524,
+      "learning_rate": 0.00018486025612164794,
+      "loss": 0.7724,
+      "step": 2598
+    },
+    {
+      "epoch": 0.8344838657890512,
+      "grad_norm": 0.7496247291564941,
+      "learning_rate": 0.0001848417842899429,
+      "loss": 0.5834,
+      "step": 2599
+    },
+    {
+      "epoch": 0.8348049446139028,
+      "grad_norm": 0.6890442371368408,
+      "learning_rate": 0.00018482330212052378,
+      "loss": 0.6603,
+      "step": 2600
+    },
+    {
+      "epoch": 0.8351260234387542,
+      "grad_norm": 1.2235509157180786,
+      "learning_rate": 0.0001848048096156426,
+      "loss": 1.3789,
+      "step": 2601
+    },
+    {
+      "epoch": 0.8354471022636057,
+      "grad_norm": 0.9681376218795776,
+      "learning_rate": 0.00018478630677755262,
+      "loss": 0.9582,
+      "step": 2602
+    },
+    {
+      "epoch": 0.8357681810884572,
+      "grad_norm": 1.0608433485031128,
+      "learning_rate": 0.00018476779360850832,
+      "loss": 0.7719,
+      "step": 2603
+    },
+    {
+      "epoch": 0.8360892599133087,
+      "grad_norm": 1.0130113363265991,
+      "learning_rate": 0.00018474927011076552,
+      "loss": 0.822,
+      "step": 2604
+    },
+    {
+      "epoch": 0.8364103387381602,
+      "grad_norm": 1.0975375175476074,
+      "learning_rate": 0.0001847307362865812,
+      "loss": 0.7262,
+      "step": 2605
+    },
+    {
+      "epoch": 0.8367314175630117,
+      "grad_norm": 0.8877381086349487,
+      "learning_rate": 0.00018471219213821375,
+      "loss": 0.5131,
+      "step": 2606
+    },
+    {
+      "epoch": 0.8370524963878632,
+      "grad_norm": 1.0735324621200562,
+      "learning_rate": 0.00018469363766792255,
+      "loss": 0.6324,
+      "step": 2607
+    },
+    {
+      "epoch": 0.8373735752127147,
+      "grad_norm": 0.9613997936248779,
+      "learning_rate": 0.00018467507287796856,
+      "loss": 0.8337,
+      "step": 2608
+    },
+    {
+      "epoch": 0.8376946540375663,
+      "grad_norm": 1.0283273458480835,
+      "learning_rate": 0.0001846564977706138,
+      "loss": 0.8794,
+      "step": 2609
+    },
+    {
+      "epoch": 0.8380157328624177,
+      "grad_norm": 1.1222115755081177,
+      "learning_rate": 0.00018463791234812153,
+      "loss": 0.9956,
+      "step": 2610
+    },
+    {
+      "epoch": 0.8383368116872693,
+      "grad_norm": 0.8465763330459595,
+      "learning_rate": 0.00018461931661275643,
+      "loss": 0.8214,
+      "step": 2611
+    },
+    {
+      "epoch": 0.8386578905121207,
+      "grad_norm": 1.5119593143463135,
+      "learning_rate": 0.00018460071056678422,
+      "loss": 1.0962,
+      "step": 2612
+    },
+    {
+      "epoch": 0.8389789693369722,
+      "grad_norm": 0.7871680855751038,
+      "learning_rate": 0.00018458209421247208,
+      "loss": 0.6603,
+      "step": 2613
+    },
+    {
+      "epoch": 0.8393000481618237,
+      "grad_norm": 0.7733088731765747,
+      "learning_rate": 0.00018456346755208833,
+      "loss": 0.8264,
+      "step": 2614
+    },
+    {
+      "epoch": 0.8396211269866752,
+      "grad_norm": 0.8866434693336487,
+      "learning_rate": 0.00018454483058790255,
+      "loss": 0.7796,
+      "step": 2615
+    },
+    {
+      "epoch": 0.8399422058115267,
+      "grad_norm": 0.9938840270042419,
+      "learning_rate": 0.00018452618332218563,
+      "loss": 1.0233,
+      "step": 2616
+    },
+    {
+      "epoch": 0.8402632846363782,
+      "grad_norm": 1.1391087770462036,
+      "learning_rate": 0.00018450752575720967,
+      "loss": 0.8702,
+      "step": 2617
+    },
+    {
+      "epoch": 0.8405843634612298,
+      "grad_norm": 1.0326181650161743,
+      "learning_rate": 0.00018448885789524802,
+      "loss": 1.1119,
+      "step": 2618
+    },
+    {
+      "epoch": 0.8409054422860812,
+      "grad_norm": 1.0216528177261353,
+      "learning_rate": 0.00018447017973857532,
+      "loss": 0.9276,
+      "step": 2619
+    },
+    {
+      "epoch": 0.8412265211109328,
+      "grad_norm": 1.6243035793304443,
+      "learning_rate": 0.00018445149128946744,
+      "loss": 0.7381,
+      "step": 2620
+    },
+    {
+      "epoch": 0.8415475999357842,
+      "grad_norm": 1.0764888525009155,
+      "learning_rate": 0.00018443279255020152,
+      "loss": 1.0441,
+      "step": 2621
+    },
+    {
+      "epoch": 0.8418686787606358,
+      "grad_norm": 0.9923635721206665,
+      "learning_rate": 0.00018441408352305594,
+      "loss": 0.9418,
+      "step": 2622
+    },
+    {
+      "epoch": 0.8421897575854872,
+      "grad_norm": 1.145235538482666,
+      "learning_rate": 0.00018439536421031033,
+      "loss": 0.8303,
+      "step": 2623
+    },
+    {
+      "epoch": 0.8425108364103387,
+      "grad_norm": 1.0477502346038818,
+      "learning_rate": 0.0001843766346142456,
+      "loss": 0.9138,
+      "step": 2624
+    },
+    {
+      "epoch": 0.8428319152351902,
+      "grad_norm": 0.9633687734603882,
+      "learning_rate": 0.0001843578947371439,
+      "loss": 0.7834,
+      "step": 2625
+    },
+    {
+      "epoch": 0.8431529940600417,
+      "grad_norm": 0.8642808198928833,
+      "learning_rate": 0.0001843391445812886,
+      "loss": 0.9069,
+      "step": 2626
+    },
+    {
+      "epoch": 0.8434740728848933,
+      "grad_norm": 1.472249984741211,
+      "learning_rate": 0.00018432038414896434,
+      "loss": 0.8867,
+      "step": 2627
+    },
+    {
+      "epoch": 0.8437951517097447,
+      "grad_norm": 1.1954059600830078,
+      "learning_rate": 0.00018430161344245707,
+      "loss": 0.9881,
+      "step": 2628
+    },
+    {
+      "epoch": 0.8441162305345963,
+      "grad_norm": 0.8830155730247498,
+      "learning_rate": 0.0001842828324640539,
+      "loss": 0.6729,
+      "step": 2629
+    },
+    {
+      "epoch": 0.8444373093594477,
+      "grad_norm": 1.3963950872421265,
+      "learning_rate": 0.00018426404121604323,
+      "loss": 1.0219,
+      "step": 2630
+    },
+    {
+      "epoch": 0.8447583881842993,
+      "grad_norm": 1.185807704925537,
+      "learning_rate": 0.00018424523970071477,
+      "loss": 0.9033,
+      "step": 2631
+    },
+    {
+      "epoch": 0.8450794670091507,
+      "grad_norm": 0.9597305059432983,
+      "learning_rate": 0.0001842264279203594,
+      "loss": 1.0832,
+      "step": 2632
+    },
+    {
+      "epoch": 0.8454005458340023,
+      "grad_norm": 0.8251401782035828,
+      "learning_rate": 0.00018420760587726923,
+      "loss": 0.7083,
+      "step": 2633
+    },
+    {
+      "epoch": 0.8457216246588537,
+      "grad_norm": 1.246910572052002,
+      "learning_rate": 0.00018418877357373776,
+      "loss": 1.1274,
+      "step": 2634
+    },
+    {
+      "epoch": 0.8460427034837052,
+      "grad_norm": 1.0847307443618774,
+      "learning_rate": 0.00018416993101205958,
+      "loss": 0.8948,
+      "step": 2635
+    },
+    {
+      "epoch": 0.8463637823085568,
+      "grad_norm": 1.0887922048568726,
+      "learning_rate": 0.00018415107819453062,
+      "loss": 1.111,
+      "step": 2636
+    },
+    {
+      "epoch": 0.8466848611334082,
+      "grad_norm": 0.9815394878387451,
+      "learning_rate": 0.00018413221512344805,
+      "loss": 0.8554,
+      "step": 2637
+    },
+    {
+      "epoch": 0.8470059399582598,
+      "grad_norm": 1.085922122001648,
+      "learning_rate": 0.00018411334180111027,
+      "loss": 0.7271,
+      "step": 2638
+    },
+    {
+      "epoch": 0.8473270187831112,
+      "grad_norm": 1.687097430229187,
+      "learning_rate": 0.00018409445822981693,
+      "loss": 0.9256,
+      "step": 2639
+    },
+    {
+      "epoch": 0.8476480976079628,
+      "grad_norm": 1.0354297161102295,
+      "learning_rate": 0.00018407556441186893,
+      "loss": 0.7517,
+      "step": 2640
+    },
+    {
+      "epoch": 0.8479691764328142,
+      "grad_norm": 1.799309492111206,
+      "learning_rate": 0.00018405666034956844,
+      "loss": 0.888,
+      "step": 2641
+    },
+    {
+      "epoch": 0.8482902552576658,
+      "grad_norm": 1.172786831855774,
+      "learning_rate": 0.00018403774604521886,
+      "loss": 0.9754,
+      "step": 2642
+    },
+    {
+      "epoch": 0.8486113340825172,
+      "grad_norm": 1.1656599044799805,
+      "learning_rate": 0.00018401882150112484,
+      "loss": 0.8011,
+      "step": 2643
+    },
+    {
+      "epoch": 0.8489324129073688,
+      "grad_norm": 1.0657223463058472,
+      "learning_rate": 0.00018399988671959227,
+      "loss": 0.8496,
+      "step": 2644
+    },
+    {
+      "epoch": 0.8492534917322203,
+      "grad_norm": 1.51372230052948,
+      "learning_rate": 0.0001839809417029283,
+      "loss": 0.9055,
+      "step": 2645
+    },
+    {
+      "epoch": 0.8495745705570718,
+      "grad_norm": 1.423654556274414,
+      "learning_rate": 0.00018396198645344135,
+      "loss": 0.8467,
+      "step": 2646
+    },
+    {
+      "epoch": 0.8498956493819233,
+      "grad_norm": 2.3389534950256348,
+      "learning_rate": 0.000183943020973441,
+      "loss": 0.8266,
+      "step": 2647
+    },
+    {
+      "epoch": 0.8502167282067747,
+      "grad_norm": 0.9923876523971558,
+      "learning_rate": 0.00018392404526523817,
+      "loss": 0.7428,
+      "step": 2648
+    },
+    {
+      "epoch": 0.8505378070316263,
+      "grad_norm": 0.8069606423377991,
+      "learning_rate": 0.000183905059331145,
+      "loss": 0.5436,
+      "step": 2649
+    },
+    {
+      "epoch": 0.8508588858564777,
+      "grad_norm": 0.787869930267334,
+      "learning_rate": 0.0001838860631734749,
+      "loss": 0.4813,
+      "step": 2650
+    },
+    {
+      "epoch": 0.8511799646813293,
+      "grad_norm": 1.2456567287445068,
+      "learning_rate": 0.00018386705679454242,
+      "loss": 1.2521,
+      "step": 2651
+    },
+    {
+      "epoch": 0.8515010435061807,
+      "grad_norm": 1.0350762605667114,
+      "learning_rate": 0.00018384804019666345,
+      "loss": 1.4132,
+      "step": 2652
+    },
+    {
+      "epoch": 0.8518221223310323,
+      "grad_norm": 1.5614581108093262,
+      "learning_rate": 0.00018382901338215516,
+      "loss": 0.9286,
+      "step": 2653
+    },
+    {
+      "epoch": 0.8521432011558838,
+      "grad_norm": 1.5894172191619873,
+      "learning_rate": 0.00018380997635333585,
+      "loss": 0.7567,
+      "step": 2654
+    },
+    {
+      "epoch": 0.8524642799807353,
+      "grad_norm": 1.0441161394119263,
+      "learning_rate": 0.00018379092911252514,
+      "loss": 0.5728,
+      "step": 2655
+    },
+    {
+      "epoch": 0.8527853588055868,
+      "grad_norm": 1.1320891380310059,
+      "learning_rate": 0.0001837718716620439,
+      "loss": 0.5784,
+      "step": 2656
+    },
+    {
+      "epoch": 0.8531064376304383,
+      "grad_norm": 1.0418906211853027,
+      "learning_rate": 0.0001837528040042142,
+      "loss": 0.5328,
+      "step": 2657
+    },
+    {
+      "epoch": 0.8534275164552898,
+      "grad_norm": 1.1594635248184204,
+      "learning_rate": 0.00018373372614135936,
+      "loss": 0.7899,
+      "step": 2658
+    },
+    {
+      "epoch": 0.8537485952801412,
+      "grad_norm": 1.4470210075378418,
+      "learning_rate": 0.000183714638075804,
+      "loss": 1.0683,
+      "step": 2659
+    },
+    {
+      "epoch": 0.8540696741049928,
+      "grad_norm": 0.8838327527046204,
+      "learning_rate": 0.0001836955398098739,
+      "loss": 0.8985,
+      "step": 2660
+    },
+    {
+      "epoch": 0.8543907529298442,
+      "grad_norm": 0.9166834354400635,
+      "learning_rate": 0.00018367643134589617,
+      "loss": 0.9936,
+      "step": 2661
+    },
+    {
+      "epoch": 0.8547118317546958,
+      "grad_norm": 0.998890221118927,
+      "learning_rate": 0.0001836573126861991,
+      "loss": 0.9257,
+      "step": 2662
+    },
+    {
+      "epoch": 0.8550329105795473,
+      "grad_norm": 1.1718299388885498,
+      "learning_rate": 0.00018363818383311225,
+      "loss": 1.0218,
+      "step": 2663
+    },
+    {
+      "epoch": 0.8553539894043988,
+      "grad_norm": 1.2179023027420044,
+      "learning_rate": 0.0001836190447889664,
+      "loss": 1.0098,
+      "step": 2664
+    },
+    {
+      "epoch": 0.8556750682292503,
+      "grad_norm": 1.2525956630706787,
+      "learning_rate": 0.00018359989555609353,
+      "loss": 0.8362,
+      "step": 2665
+    },
+    {
+      "epoch": 0.8559961470541018,
+      "grad_norm": 1.1447235345840454,
+      "learning_rate": 0.00018358073613682706,
+      "loss": 0.825,
+      "step": 2666
+    },
+    {
+      "epoch": 0.8563172258789533,
+      "grad_norm": 1.0088034868240356,
+      "learning_rate": 0.00018356156653350137,
+      "loss": 0.8964,
+      "step": 2667
+    },
+    {
+      "epoch": 0.8566383047038048,
+      "grad_norm": 1.095577597618103,
+      "learning_rate": 0.00018354238674845225,
+      "loss": 0.8193,
+      "step": 2668
+    },
+    {
+      "epoch": 0.8569593835286563,
+      "grad_norm": 1.3043547868728638,
+      "learning_rate": 0.00018352319678401676,
+      "loss": 1.0768,
+      "step": 2669
+    },
+    {
+      "epoch": 0.8572804623535077,
+      "grad_norm": 1.7319622039794922,
+      "learning_rate": 0.00018350399664253305,
+      "loss": 0.9552,
+      "step": 2670
+    },
+    {
+      "epoch": 0.8576015411783593,
+      "grad_norm": 1.5099067687988281,
+      "learning_rate": 0.00018348478632634066,
+      "loss": 1.0515,
+      "step": 2671
+    },
+    {
+      "epoch": 0.8579226200032107,
+      "grad_norm": 1.069471001625061,
+      "learning_rate": 0.0001834655658377803,
+      "loss": 0.8308,
+      "step": 2672
+    },
+    {
+      "epoch": 0.8582436988280623,
+      "grad_norm": 0.9602334499359131,
+      "learning_rate": 0.00018344633517919392,
+      "loss": 0.8164,
+      "step": 2673
+    },
+    {
+      "epoch": 0.8585647776529138,
+      "grad_norm": 1.416114091873169,
+      "learning_rate": 0.00018342709435292473,
+      "loss": 0.9465,
+      "step": 2674
+    },
+    {
+      "epoch": 0.8588858564777653,
+      "grad_norm": 1.6077988147735596,
+      "learning_rate": 0.00018340784336131713,
+      "loss": 0.907,
+      "step": 2675
+    },
+    {
+      "epoch": 0.8592069353026168,
+      "grad_norm": 1.6769777536392212,
+      "learning_rate": 0.00018338858220671682,
+      "loss": 0.8495,
+      "step": 2676
+    },
+    {
+      "epoch": 0.8595280141274683,
+      "grad_norm": 0.7927365899085999,
+      "learning_rate": 0.00018336931089147073,
+      "loss": 0.7901,
+      "step": 2677
+    },
+    {
+      "epoch": 0.8598490929523198,
+      "grad_norm": 0.780259907245636,
+      "learning_rate": 0.00018335002941792698,
+      "loss": 0.714,
+      "step": 2678
+    },
+    {
+      "epoch": 0.8601701717771713,
+      "grad_norm": 1.1689207553863525,
+      "learning_rate": 0.000183330737788435,
+      "loss": 0.9876,
+      "step": 2679
+    },
+    {
+      "epoch": 0.8604912506020228,
+      "grad_norm": 1.2812306880950928,
+      "learning_rate": 0.00018331143600534535,
+      "loss": 1.1554,
+      "step": 2680
+    },
+    {
+      "epoch": 0.8608123294268742,
+      "grad_norm": 1.0278218984603882,
+      "learning_rate": 0.00018329212407100994,
+      "loss": 0.75,
+      "step": 2681
+    },
+    {
+      "epoch": 0.8611334082517258,
+      "grad_norm": 1.0668963193893433,
+      "learning_rate": 0.0001832728019877819,
+      "loss": 0.8591,
+      "step": 2682
+    },
+    {
+      "epoch": 0.8614544870765773,
+      "grad_norm": 1.1060330867767334,
+      "learning_rate": 0.0001832534697580155,
+      "loss": 1.0402,
+      "step": 2683
+    },
+    {
+      "epoch": 0.8617755659014288,
+      "grad_norm": 1.13728928565979,
+      "learning_rate": 0.00018323412738406635,
+      "loss": 0.8727,
+      "step": 2684
+    },
+    {
+      "epoch": 0.8620966447262803,
+      "grad_norm": 1.4041433334350586,
+      "learning_rate": 0.00018321477486829126,
+      "loss": 0.949,
+      "step": 2685
+    },
+    {
+      "epoch": 0.8624177235511318,
+      "grad_norm": 1.0087988376617432,
+      "learning_rate": 0.00018319541221304827,
+      "loss": 0.8199,
+      "step": 2686
+    },
+    {
+      "epoch": 0.8627388023759833,
+      "grad_norm": 0.8733333945274353,
+      "learning_rate": 0.00018317603942069664,
+      "loss": 0.8297,
+      "step": 2687
+    },
+    {
+      "epoch": 0.8630598812008348,
+      "grad_norm": 1.4244256019592285,
+      "learning_rate": 0.00018315665649359692,
+      "loss": 0.9084,
+      "step": 2688
+    },
+    {
+      "epoch": 0.8633809600256863,
+      "grad_norm": 0.8898962736129761,
+      "learning_rate": 0.00018313726343411086,
+      "loss": 0.7508,
+      "step": 2689
+    },
+    {
+      "epoch": 0.8637020388505378,
+      "grad_norm": 1.4125561714172363,
+      "learning_rate": 0.0001831178602446014,
+      "loss": 0.8825,
+      "step": 2690
+    },
+    {
+      "epoch": 0.8640231176753893,
+      "grad_norm": 0.8810192346572876,
+      "learning_rate": 0.00018309844692743283,
+      "loss": 0.8294,
+      "step": 2691
+    },
+    {
+      "epoch": 0.8643441965002409,
+      "grad_norm": 2.355106830596924,
+      "learning_rate": 0.00018307902348497056,
+      "loss": 0.6482,
+      "step": 2692
+    },
+    {
+      "epoch": 0.8646652753250923,
+      "grad_norm": 0.9504086375236511,
+      "learning_rate": 0.00018305958991958127,
+      "loss": 0.7463,
+      "step": 2693
+    },
+    {
+      "epoch": 0.8649863541499438,
+      "grad_norm": 1.0893597602844238,
+      "learning_rate": 0.0001830401462336329,
+      "loss": 0.8072,
+      "step": 2694
+    },
+    {
+      "epoch": 0.8653074329747953,
+      "grad_norm": 0.7756816148757935,
+      "learning_rate": 0.0001830206924294946,
+      "loss": 0.6733,
+      "step": 2695
+    },
+    {
+      "epoch": 0.8656285117996468,
+      "grad_norm": 0.8419047594070435,
+      "learning_rate": 0.00018300122850953675,
+      "loss": 0.6931,
+      "step": 2696
+    },
+    {
+      "epoch": 0.8659495906244983,
+      "grad_norm": 0.8006436228752136,
+      "learning_rate": 0.00018298175447613096,
+      "loss": 0.6323,
+      "step": 2697
+    },
+    {
+      "epoch": 0.8662706694493498,
+      "grad_norm": 0.9535905122756958,
+      "learning_rate": 0.00018296227033165013,
+      "loss": 0.6195,
+      "step": 2698
+    },
+    {
+      "epoch": 0.8665917482742013,
+      "grad_norm": 0.7773939371109009,
+      "learning_rate": 0.00018294277607846832,
+      "loss": 0.5081,
+      "step": 2699
+    },
+    {
+      "epoch": 0.8669128270990528,
+      "grad_norm": 2.4419236183166504,
+      "learning_rate": 0.0001829232717189608,
+      "loss": 0.7581,
+      "step": 2700
+    },
+    {
+      "epoch": 0.8672339059239044,
+      "grad_norm": 0.919539213180542,
+      "learning_rate": 0.00018290375725550417,
+      "loss": 1.1346,
+      "step": 2701
+    },
+    {
+      "epoch": 0.8675549847487558,
+      "grad_norm": 0.9994780421257019,
+      "learning_rate": 0.0001828842326904762,
+      "loss": 0.9724,
+      "step": 2702
+    },
+    {
+      "epoch": 0.8678760635736074,
+      "grad_norm": 1.1944332122802734,
+      "learning_rate": 0.00018286469802625589,
+      "loss": 0.8469,
+      "step": 2703
+    },
+    {
+      "epoch": 0.8681971423984588,
+      "grad_norm": 1.1087571382522583,
+      "learning_rate": 0.00018284515326522346,
+      "loss": 0.6316,
+      "step": 2704
+    },
+    {
+      "epoch": 0.8685182212233103,
+      "grad_norm": 1.079188346862793,
+      "learning_rate": 0.00018282559840976042,
+      "loss": 0.5803,
+      "step": 2705
+    },
+    {
+      "epoch": 0.8688393000481618,
+      "grad_norm": 0.9261389374732971,
+      "learning_rate": 0.00018280603346224945,
+      "loss": 0.5796,
+      "step": 2706
+    },
+    {
+      "epoch": 0.8691603788730133,
+      "grad_norm": 1.1797443628311157,
+      "learning_rate": 0.00018278645842507448,
+      "loss": 0.7453,
+      "step": 2707
+    },
+    {
+      "epoch": 0.8694814576978648,
+      "grad_norm": 1.0713974237442017,
+      "learning_rate": 0.00018276687330062065,
+      "loss": 0.994,
+      "step": 2708
+    },
+    {
+      "epoch": 0.8698025365227163,
+      "grad_norm": 1.1797808408737183,
+      "learning_rate": 0.00018274727809127438,
+      "loss": 0.7688,
+      "step": 2709
+    },
+    {
+      "epoch": 0.8701236153475679,
+      "grad_norm": 0.9432153105735779,
+      "learning_rate": 0.00018272767279942328,
+      "loss": 1.0629,
+      "step": 2710
+    },
+    {
+      "epoch": 0.8704446941724193,
+      "grad_norm": 0.9762241840362549,
+      "learning_rate": 0.00018270805742745617,
+      "loss": 1.0634,
+      "step": 2711
+    },
+    {
+      "epoch": 0.8707657729972709,
+      "grad_norm": 0.8841230273246765,
+      "learning_rate": 0.00018268843197776318,
+      "loss": 0.9472,
+      "step": 2712
+    },
+    {
+      "epoch": 0.8710868518221223,
+      "grad_norm": 1.085910439491272,
+      "learning_rate": 0.00018266879645273556,
+      "loss": 0.9591,
+      "step": 2713
+    },
+    {
+      "epoch": 0.8714079306469739,
+      "grad_norm": 0.8603227138519287,
+      "learning_rate": 0.00018264915085476583,
+      "loss": 0.8901,
+      "step": 2714
+    },
+    {
+      "epoch": 0.8717290094718253,
+      "grad_norm": 1.1412354707717896,
+      "learning_rate": 0.0001826294951862478,
+      "loss": 0.8354,
+      "step": 2715
+    },
+    {
+      "epoch": 0.8720500882966769,
+      "grad_norm": 0.9545773267745972,
+      "learning_rate": 0.00018260982944957638,
+      "loss": 1.0581,
+      "step": 2716
+    },
+    {
+      "epoch": 0.8723711671215283,
+      "grad_norm": 1.4084380865097046,
+      "learning_rate": 0.00018259015364714787,
+      "loss": 0.9433,
+      "step": 2717
+    },
+    {
+      "epoch": 0.8726922459463798,
+      "grad_norm": 1.183406949043274,
+      "learning_rate": 0.00018257046778135964,
+      "loss": 0.9878,
+      "step": 2718
+    },
+    {
+      "epoch": 0.8730133247712314,
+      "grad_norm": 0.791094958782196,
+      "learning_rate": 0.00018255077185461038,
+      "loss": 0.8771,
+      "step": 2719
+    },
+    {
+      "epoch": 0.8733344035960828,
+      "grad_norm": 1.0063689947128296,
+      "learning_rate": 0.00018253106586929997,
+      "loss": 0.8799,
+      "step": 2720
+    },
+    {
+      "epoch": 0.8736554824209344,
+      "grad_norm": 1.1546752452850342,
+      "learning_rate": 0.00018251134982782952,
+      "loss": 1.0092,
+      "step": 2721
+    },
+    {
+      "epoch": 0.8739765612457858,
+      "grad_norm": 1.525952696800232,
+      "learning_rate": 0.00018249162373260141,
+      "loss": 0.9339,
+      "step": 2722
+    },
+    {
+      "epoch": 0.8742976400706374,
+      "grad_norm": 1.4072719812393188,
+      "learning_rate": 0.0001824718875860191,
+      "loss": 0.8635,
+      "step": 2723
+    },
+    {
+      "epoch": 0.8746187188954888,
+      "grad_norm": 1.070801854133606,
+      "learning_rate": 0.00018245214139048753,
+      "loss": 1.107,
+      "step": 2724
+    },
+    {
+      "epoch": 0.8749397977203404,
+      "grad_norm": 0.9687029123306274,
+      "learning_rate": 0.0001824323851484126,
+      "loss": 0.6512,
+      "step": 2725
+    },
+    {
+      "epoch": 0.8752608765451918,
+      "grad_norm": 1.0900896787643433,
+      "learning_rate": 0.00018241261886220154,
+      "loss": 1.0686,
+      "step": 2726
+    },
+    {
+      "epoch": 0.8755819553700434,
+      "grad_norm": 1.1022065877914429,
+      "learning_rate": 0.00018239284253426295,
+      "loss": 0.9941,
+      "step": 2727
+    },
+    {
+      "epoch": 0.8759030341948948,
+      "grad_norm": 0.7542319297790527,
+      "learning_rate": 0.00018237305616700637,
+      "loss": 0.783,
+      "step": 2728
+    },
+    {
+      "epoch": 0.8762241130197463,
+      "grad_norm": 1.090649127960205,
+      "learning_rate": 0.00018235325976284275,
+      "loss": 0.8426,
+      "step": 2729
+    },
+    {
+      "epoch": 0.8765451918445979,
+      "grad_norm": 0.8522968888282776,
+      "learning_rate": 0.00018233345332418423,
+      "loss": 0.689,
+      "step": 2730
+    },
+    {
+      "epoch": 0.8768662706694493,
+      "grad_norm": 0.9584747552871704,
+      "learning_rate": 0.0001823136368534442,
+      "loss": 0.6896,
+      "step": 2731
+    },
+    {
+      "epoch": 0.8771873494943009,
+      "grad_norm": 1.115478754043579,
+      "learning_rate": 0.00018229381035303718,
+      "loss": 0.72,
+      "step": 2732
+    },
+    {
+      "epoch": 0.8775084283191523,
+      "grad_norm": 0.9085679650306702,
+      "learning_rate": 0.000182273973825379,
+      "loss": 0.6926,
+      "step": 2733
+    },
+    {
+      "epoch": 0.8778295071440039,
+      "grad_norm": 1.5002387762069702,
+      "learning_rate": 0.00018225412727288667,
+      "loss": 1.0068,
+      "step": 2734
+    },
+    {
+      "epoch": 0.8781505859688553,
+      "grad_norm": 1.0805754661560059,
+      "learning_rate": 0.00018223427069797844,
+      "loss": 0.8847,
+      "step": 2735
+    },
+    {
+      "epoch": 0.8784716647937069,
+      "grad_norm": 1.1780469417572021,
+      "learning_rate": 0.00018221440410307374,
+      "loss": 0.798,
+      "step": 2736
+    },
+    {
+      "epoch": 0.8787927436185583,
+      "grad_norm": 1.034372091293335,
+      "learning_rate": 0.0001821945274905933,
+      "loss": 1.0309,
+      "step": 2737
+    },
+    {
+      "epoch": 0.8791138224434099,
+      "grad_norm": 1.680379867553711,
+      "learning_rate": 0.00018217464086295904,
+      "loss": 1.0186,
+      "step": 2738
+    },
+    {
+      "epoch": 0.8794349012682614,
+      "grad_norm": 1.7660083770751953,
+      "learning_rate": 0.00018215474422259402,
+      "loss": 0.8119,
+      "step": 2739
+    },
+    {
+      "epoch": 0.8797559800931128,
+      "grad_norm": 1.747609257698059,
+      "learning_rate": 0.00018213483757192263,
+      "loss": 0.5625,
+      "step": 2740
+    },
+    {
+      "epoch": 0.8800770589179644,
+      "grad_norm": 1.156343936920166,
+      "learning_rate": 0.00018211492091337042,
+      "loss": 1.0188,
+      "step": 2741
+    },
+    {
+      "epoch": 0.8803981377428158,
+      "grad_norm": 1.104516863822937,
+      "learning_rate": 0.00018209499424936415,
+      "loss": 0.6375,
+      "step": 2742
+    },
+    {
+      "epoch": 0.8807192165676674,
+      "grad_norm": 3.744687080383301,
+      "learning_rate": 0.0001820750575823319,
+      "loss": 0.9079,
+      "step": 2743
+    },
+    {
+      "epoch": 0.8810402953925188,
+      "grad_norm": 1.1239784955978394,
+      "learning_rate": 0.00018205511091470283,
+      "loss": 0.7289,
+      "step": 2744
+    },
+    {
+      "epoch": 0.8813613742173704,
+      "grad_norm": 0.7483261823654175,
+      "learning_rate": 0.0001820351542489074,
+      "loss": 0.5682,
+      "step": 2745
+    },
+    {
+      "epoch": 0.8816824530422218,
+      "grad_norm": 1.107564926147461,
+      "learning_rate": 0.00018201518758737724,
+      "loss": 0.6602,
+      "step": 2746
+    },
+    {
+      "epoch": 0.8820035318670734,
+      "grad_norm": 4.356315612792969,
+      "learning_rate": 0.00018199521093254523,
+      "loss": 0.7124,
+      "step": 2747
+    },
+    {
+      "epoch": 0.8823246106919249,
+      "grad_norm": 1.0629299879074097,
+      "learning_rate": 0.00018197522428684552,
+      "loss": 0.4317,
+      "step": 2748
+    },
+    {
+      "epoch": 0.8826456895167764,
+      "grad_norm": 0.7678312063217163,
+      "learning_rate": 0.0001819552276527134,
+      "loss": 0.447,
+      "step": 2749
+    },
+    {
+      "epoch": 0.8829667683416279,
+      "grad_norm": 1.091256856918335,
+      "learning_rate": 0.00018193522103258537,
+      "loss": 0.5988,
+      "step": 2750
+    },
+    {
+      "epoch": 0.8832878471664793,
+      "grad_norm": 1.013898253440857,
+      "learning_rate": 0.0001819152044288992,
+      "loss": 1.3746,
+      "step": 2751
+    },
+    {
+      "epoch": 0.8836089259913309,
+      "grad_norm": 0.9919261336326599,
+      "learning_rate": 0.00018189517784409381,
+      "loss": 1.3712,
+      "step": 2752
+    },
+    {
+      "epoch": 0.8839300048161823,
+      "grad_norm": 1.0880647897720337,
+      "learning_rate": 0.00018187514128060946,
+      "loss": 0.9107,
+      "step": 2753
+    },
+    {
+      "epoch": 0.8842510836410339,
+      "grad_norm": 1.6532173156738281,
+      "learning_rate": 0.0001818550947408875,
+      "loss": 0.8713,
+      "step": 2754
+    },
+    {
+      "epoch": 0.8845721624658853,
+      "grad_norm": 0.9895550608634949,
+      "learning_rate": 0.0001818350382273705,
+      "loss": 0.6474,
+      "step": 2755
+    },
+    {
+      "epoch": 0.8848932412907369,
+      "grad_norm": 1.1515357494354248,
+      "learning_rate": 0.00018181497174250236,
+      "loss": 0.5816,
+      "step": 2756
+    },
+    {
+      "epoch": 0.8852143201155884,
+      "grad_norm": 1.0654630661010742,
+      "learning_rate": 0.00018179489528872807,
+      "loss": 0.5709,
+      "step": 2757
+    },
+    {
+      "epoch": 0.8855353989404399,
+      "grad_norm": 1.0934091806411743,
+      "learning_rate": 0.00018177480886849388,
+      "loss": 0.9673,
+      "step": 2758
+    },
+    {
+      "epoch": 0.8858564777652914,
+      "grad_norm": 2.52614426612854,
+      "learning_rate": 0.0001817547124842473,
+      "loss": 0.9605,
+      "step": 2759
+    },
+    {
+      "epoch": 0.8861775565901429,
+      "grad_norm": 0.8149087429046631,
+      "learning_rate": 0.00018173460613843701,
+      "loss": 0.981,
+      "step": 2760
+    },
+    {
+      "epoch": 0.8864986354149944,
+      "grad_norm": 0.758175253868103,
+      "learning_rate": 0.00018171448983351284,
+      "loss": 0.7961,
+      "step": 2761
+    },
+    {
+      "epoch": 0.8868197142398458,
+      "grad_norm": 0.8692430257797241,
+      "learning_rate": 0.00018169436357192602,
+      "loss": 0.782,
+      "step": 2762
+    },
+    {
+      "epoch": 0.8871407930646974,
+      "grad_norm": 1.1529359817504883,
+      "learning_rate": 0.00018167422735612877,
+      "loss": 0.9836,
+      "step": 2763
+    },
+    {
+      "epoch": 0.8874618718895488,
+      "grad_norm": 0.9792761206626892,
+      "learning_rate": 0.00018165408118857464,
+      "loss": 0.9161,
+      "step": 2764
+    },
+    {
+      "epoch": 0.8877829507144004,
+      "grad_norm": 1.1091268062591553,
+      "learning_rate": 0.00018163392507171842,
+      "loss": 0.9312,
+      "step": 2765
+    },
+    {
+      "epoch": 0.8881040295392519,
+      "grad_norm": 0.8531142473220825,
+      "learning_rate": 0.00018161375900801604,
+      "loss": 0.6937,
+      "step": 2766
+    },
+    {
+      "epoch": 0.8884251083641034,
+      "grad_norm": 0.9302113652229309,
+      "learning_rate": 0.00018159358299992467,
+      "loss": 0.9284,
+      "step": 2767
+    },
+    {
+      "epoch": 0.8887461871889549,
+      "grad_norm": 1.1668531894683838,
+      "learning_rate": 0.00018157339704990275,
+      "loss": 1.0318,
+      "step": 2768
+    },
+    {
+      "epoch": 0.8890672660138064,
+      "grad_norm": 0.8978874683380127,
+      "learning_rate": 0.00018155320116040982,
+      "loss": 0.9906,
+      "step": 2769
+    },
+    {
+      "epoch": 0.8893883448386579,
+      "grad_norm": 1.0586960315704346,
+      "learning_rate": 0.00018153299533390672,
+      "loss": 0.8478,
+      "step": 2770
+    },
+    {
+      "epoch": 0.8897094236635094,
+      "grad_norm": 1.5482786893844604,
+      "learning_rate": 0.00018151277957285543,
+      "loss": 0.8452,
+      "step": 2771
+    },
+    {
+      "epoch": 0.8900305024883609,
+      "grad_norm": 1.1627875566482544,
+      "learning_rate": 0.00018149255387971922,
+      "loss": 1.1749,
+      "step": 2772
+    },
+    {
+      "epoch": 0.8903515813132123,
+      "grad_norm": 1.5141947269439697,
+      "learning_rate": 0.00018147231825696252,
+      "loss": 0.9226,
+      "step": 2773
+    },
+    {
+      "epoch": 0.8906726601380639,
+      "grad_norm": 0.9088945984840393,
+      "learning_rate": 0.00018145207270705096,
+      "loss": 0.8038,
+      "step": 2774
+    },
+    {
+      "epoch": 0.8909937389629154,
+      "grad_norm": 1.2037333250045776,
+      "learning_rate": 0.0001814318172324514,
+      "loss": 0.926,
+      "step": 2775
+    },
+    {
+      "epoch": 0.8913148177877669,
+      "grad_norm": 1.1055127382278442,
+      "learning_rate": 0.00018141155183563193,
+      "loss": 0.8784,
+      "step": 2776
+    },
+    {
+      "epoch": 0.8916358966126184,
+      "grad_norm": 0.8244720101356506,
+      "learning_rate": 0.00018139127651906184,
+      "loss": 0.8521,
+      "step": 2777
+    },
+    {
+      "epoch": 0.8919569754374699,
+      "grad_norm": 1.0560851097106934,
+      "learning_rate": 0.00018137099128521156,
+      "loss": 0.7876,
+      "step": 2778
+    },
+    {
+      "epoch": 0.8922780542623214,
+      "grad_norm": 0.7370025515556335,
+      "learning_rate": 0.0001813506961365528,
+      "loss": 0.7918,
+      "step": 2779
+    },
+    {
+      "epoch": 0.8925991330871729,
+      "grad_norm": 0.9496878385543823,
+      "learning_rate": 0.00018133039107555852,
+      "loss": 0.9127,
+      "step": 2780
+    },
+    {
+      "epoch": 0.8929202119120244,
+      "grad_norm": 0.897685706615448,
+      "learning_rate": 0.00018131007610470276,
+      "loss": 0.8017,
+      "step": 2781
+    },
+    {
+      "epoch": 0.8932412907368759,
+      "grad_norm": 1.8280360698699951,
+      "learning_rate": 0.0001812897512264609,
+      "loss": 1.2183,
+      "step": 2782
+    },
+    {
+      "epoch": 0.8935623695617274,
+      "grad_norm": 1.164359450340271,
+      "learning_rate": 0.0001812694164433094,
+      "loss": 0.8879,
+      "step": 2783
+    },
+    {
+      "epoch": 0.893883448386579,
+      "grad_norm": 0.8923983573913574,
+      "learning_rate": 0.00018124907175772604,
+      "loss": 0.8414,
+      "step": 2784
+    },
+    {
+      "epoch": 0.8942045272114304,
+      "grad_norm": 1.2509326934814453,
+      "learning_rate": 0.0001812287171721897,
+      "loss": 1.0194,
+      "step": 2785
+    },
+    {
+      "epoch": 0.894525606036282,
+      "grad_norm": 0.9996488094329834,
+      "learning_rate": 0.00018120835268918063,
+      "loss": 0.9708,
+      "step": 2786
+    },
+    {
+      "epoch": 0.8948466848611334,
+      "grad_norm": 0.8612101078033447,
+      "learning_rate": 0.0001811879783111801,
+      "loss": 0.7101,
+      "step": 2787
+    },
+    {
+      "epoch": 0.8951677636859849,
+      "grad_norm": 1.1495945453643799,
+      "learning_rate": 0.00018116759404067064,
+      "loss": 0.7831,
+      "step": 2788
+    },
+    {
+      "epoch": 0.8954888425108364,
+      "grad_norm": 0.9605495929718018,
+      "learning_rate": 0.00018114719988013612,
+      "loss": 0.6721,
+      "step": 2789
+    },
+    {
+      "epoch": 0.8958099213356879,
+      "grad_norm": 1.1470704078674316,
+      "learning_rate": 0.00018112679583206137,
+      "loss": 0.6784,
+      "step": 2790
+    },
+    {
+      "epoch": 0.8961310001605394,
+      "grad_norm": 1.1517149209976196,
+      "learning_rate": 0.00018110638189893267,
+      "loss": 0.9704,
+      "step": 2791
+    },
+    {
+      "epoch": 0.8964520789853909,
+      "grad_norm": 1.7360931634902954,
+      "learning_rate": 0.00018108595808323736,
+      "loss": 0.7464,
+      "step": 2792
+    },
+    {
+      "epoch": 0.8967731578102424,
+      "grad_norm": 0.8036109805107117,
+      "learning_rate": 0.000181065524387464,
+      "loss": 0.7334,
+      "step": 2793
+    },
+    {
+      "epoch": 0.8970942366350939,
+      "grad_norm": 0.5651756525039673,
+      "learning_rate": 0.0001810450808141024,
+      "loss": 0.5383,
+      "step": 2794
+    },
+    {
+      "epoch": 0.8974153154599455,
+      "grad_norm": 1.0314606428146362,
+      "learning_rate": 0.00018102462736564355,
+      "loss": 0.6901,
+      "step": 2795
+    },
+    {
+      "epoch": 0.8977363942847969,
+      "grad_norm": 1.6273174285888672,
+      "learning_rate": 0.00018100416404457961,
+      "loss": 0.7406,
+      "step": 2796
+    },
+    {
+      "epoch": 0.8980574731096485,
+      "grad_norm": 1.643372893333435,
+      "learning_rate": 0.00018098369085340398,
+      "loss": 0.6796,
+      "step": 2797
+    },
+    {
+      "epoch": 0.8983785519344999,
+      "grad_norm": 1.2636449337005615,
+      "learning_rate": 0.00018096320779461132,
+      "loss": 0.6756,
+      "step": 2798
+    },
+    {
+      "epoch": 0.8986996307593514,
+      "grad_norm": 1.1232649087905884,
+      "learning_rate": 0.00018094271487069735,
+      "loss": 0.5207,
+      "step": 2799
+    },
+    {
+      "epoch": 0.8990207095842029,
+      "grad_norm": 0.5236806869506836,
+      "learning_rate": 0.00018092221208415907,
+      "loss": 0.4062,
+      "step": 2800
+    },
+    {
+      "epoch": 0.8993417884090544,
+      "grad_norm": 0.7692060470581055,
+      "learning_rate": 0.00018090169943749476,
+      "loss": 0.9762,
+      "step": 2801
+    },
+    {
+      "epoch": 0.8996628672339059,
+      "grad_norm": 0.8355540633201599,
+      "learning_rate": 0.00018088117693320374,
+      "loss": 1.2168,
+      "step": 2802
+    },
+    {
+      "epoch": 0.8999839460587574,
+      "grad_norm": 0.8926533460617065,
+      "learning_rate": 0.00018086064457378665,
+      "loss": 0.7676,
+      "step": 2803
+    },
+    {
+      "epoch": 0.900305024883609,
+      "grad_norm": 1.108173131942749,
+      "learning_rate": 0.00018084010236174534,
+      "loss": 0.552,
+      "step": 2804
+    },
+    {
+      "epoch": 0.9006261037084604,
+      "grad_norm": 1.1268749237060547,
+      "learning_rate": 0.00018081955029958274,
+      "loss": 0.5069,
+      "step": 2805
+    },
+    {
+      "epoch": 0.900947182533312,
+      "grad_norm": 1.0203336477279663,
+      "learning_rate": 0.00018079898838980305,
+      "loss": 0.6792,
+      "step": 2806
+    },
+    {
+      "epoch": 0.9012682613581634,
+      "grad_norm": 1.2625064849853516,
+      "learning_rate": 0.00018077841663491175,
+      "loss": 0.6916,
+      "step": 2807
+    },
+    {
+      "epoch": 0.901589340183015,
+      "grad_norm": 1.1094204187393188,
+      "learning_rate": 0.0001807578350374154,
+      "loss": 1.146,
+      "step": 2808
+    },
+    {
+      "epoch": 0.9019104190078664,
+      "grad_norm": 1.1322005987167358,
+      "learning_rate": 0.00018073724359982186,
+      "loss": 0.8353,
+      "step": 2809
+    },
+    {
+      "epoch": 0.9022314978327179,
+      "grad_norm": 0.9779659509658813,
+      "learning_rate": 0.00018071664232464002,
+      "loss": 0.7259,
+      "step": 2810
+    },
+    {
+      "epoch": 0.9025525766575694,
+      "grad_norm": 1.0345367193222046,
+      "learning_rate": 0.00018069603121438022,
+      "loss": 0.7205,
+      "step": 2811
+    },
+    {
+      "epoch": 0.9028736554824209,
+      "grad_norm": 1.5466442108154297,
+      "learning_rate": 0.00018067541027155375,
+      "loss": 1.3155,
+      "step": 2812
+    },
+    {
+      "epoch": 0.9031947343072725,
+      "grad_norm": 1.0183089971542358,
+      "learning_rate": 0.00018065477949867327,
+      "loss": 0.7655,
+      "step": 2813
+    },
+    {
+      "epoch": 0.9035158131321239,
+      "grad_norm": 0.7953174114227295,
+      "learning_rate": 0.00018063413889825254,
+      "loss": 0.8727,
+      "step": 2814
+    },
+    {
+      "epoch": 0.9038368919569755,
+      "grad_norm": 1.6211440563201904,
+      "learning_rate": 0.0001806134884728066,
+      "loss": 0.9523,
+      "step": 2815
+    },
+    {
+      "epoch": 0.9041579707818269,
+      "grad_norm": 1.3157118558883667,
+      "learning_rate": 0.00018059282822485158,
+      "loss": 1.0292,
+      "step": 2816
+    },
+    {
+      "epoch": 0.9044790496066785,
+      "grad_norm": 1.2418993711471558,
+      "learning_rate": 0.00018057215815690494,
+      "loss": 0.8633,
+      "step": 2817
+    },
+    {
+      "epoch": 0.9048001284315299,
+      "grad_norm": 1.4473196268081665,
+      "learning_rate": 0.00018055147827148523,
+      "loss": 0.8806,
+      "step": 2818
+    },
+    {
+      "epoch": 0.9051212072563815,
+      "grad_norm": 1.058814525604248,
+      "learning_rate": 0.0001805307885711122,
+      "loss": 0.8762,
+      "step": 2819
+    },
+    {
+      "epoch": 0.9054422860812329,
+      "grad_norm": 0.8622485995292664,
+      "learning_rate": 0.0001805100890583069,
+      "loss": 0.8043,
+      "step": 2820
+    },
+    {
+      "epoch": 0.9057633649060844,
+      "grad_norm": 1.5182201862335205,
+      "learning_rate": 0.0001804893797355914,
+      "loss": 1.0228,
+      "step": 2821
+    },
+    {
+      "epoch": 0.906084443730936,
+      "grad_norm": 0.8368403315544128,
+      "learning_rate": 0.00018046866060548918,
+      "loss": 0.7839,
+      "step": 2822
+    },
+    {
+      "epoch": 0.9064055225557874,
+      "grad_norm": 1.0724780559539795,
+      "learning_rate": 0.00018044793167052477,
+      "loss": 0.9187,
+      "step": 2823
+    },
+    {
+      "epoch": 0.906726601380639,
+      "grad_norm": 1.226404070854187,
+      "learning_rate": 0.00018042719293322388,
+      "loss": 0.9023,
+      "step": 2824
+    },
+    {
+      "epoch": 0.9070476802054904,
+      "grad_norm": 0.865028977394104,
+      "learning_rate": 0.00018040644439611348,
+      "loss": 0.6866,
+      "step": 2825
+    },
+    {
+      "epoch": 0.907368759030342,
+      "grad_norm": 1.3817094564437866,
+      "learning_rate": 0.00018038568606172173,
+      "loss": 1.2747,
+      "step": 2826
+    },
+    {
+      "epoch": 0.9076898378551934,
+      "grad_norm": 1.4625390768051147,
+      "learning_rate": 0.00018036491793257798,
+      "loss": 0.9156,
+      "step": 2827
+    },
+    {
+      "epoch": 0.908010916680045,
+      "grad_norm": 1.0158188343048096,
+      "learning_rate": 0.00018034414001121278,
+      "loss": 0.9246,
+      "step": 2828
+    },
+    {
+      "epoch": 0.9083319955048964,
+      "grad_norm": 0.98722243309021,
+      "learning_rate": 0.0001803233523001578,
+      "loss": 1.1086,
+      "step": 2829
+    },
+    {
+      "epoch": 0.908653074329748,
+      "grad_norm": 0.8799543976783752,
+      "learning_rate": 0.000180302554801946,
+      "loss": 0.724,
+      "step": 2830
+    },
+    {
+      "epoch": 0.9089741531545995,
+      "grad_norm": 1.1787347793579102,
+      "learning_rate": 0.00018028174751911146,
+      "loss": 1.0376,
+      "step": 2831
+    },
+    {
+      "epoch": 0.909295231979451,
+      "grad_norm": 1.1940038204193115,
+      "learning_rate": 0.00018026093045418954,
+      "loss": 0.7911,
+      "step": 2832
+    },
+    {
+      "epoch": 0.9096163108043025,
+      "grad_norm": 0.9036272168159485,
+      "learning_rate": 0.0001802401036097167,
+      "loss": 0.7765,
+      "step": 2833
+    },
+    {
+      "epoch": 0.9099373896291539,
+      "grad_norm": 1.148139238357544,
+      "learning_rate": 0.00018021926698823059,
+      "loss": 0.8591,
+      "step": 2834
+    },
+    {
+      "epoch": 0.9102584684540055,
+      "grad_norm": 0.8454437851905823,
+      "learning_rate": 0.00018019842059227012,
+      "loss": 0.7655,
+      "step": 2835
+    },
+    {
+      "epoch": 0.9105795472788569,
+      "grad_norm": 1.5326919555664062,
+      "learning_rate": 0.0001801775644243754,
+      "loss": 0.9971,
+      "step": 2836
+    },
+    {
+      "epoch": 0.9109006261037085,
+      "grad_norm": 1.6905158758163452,
+      "learning_rate": 0.00018015669848708767,
+      "loss": 1.1873,
+      "step": 2837
+    },
+    {
+      "epoch": 0.9112217049285599,
+      "grad_norm": 0.8212069869041443,
+      "learning_rate": 0.00018013582278294935,
+      "loss": 0.7388,
+      "step": 2838
+    },
+    {
+      "epoch": 0.9115427837534115,
+      "grad_norm": 1.243600606918335,
+      "learning_rate": 0.00018011493731450413,
+      "loss": 0.8089,
+      "step": 2839
+    },
+    {
+      "epoch": 0.911863862578263,
+      "grad_norm": 1.1280215978622437,
+      "learning_rate": 0.0001800940420842968,
+      "loss": 0.8168,
+      "step": 2840
+    },
+    {
+      "epoch": 0.9121849414031145,
+      "grad_norm": 1.2804031372070312,
+      "learning_rate": 0.00018007313709487334,
+      "loss": 1.0121,
+      "step": 2841
+    },
+    {
+      "epoch": 0.912506020227966,
+      "grad_norm": 1.406071662902832,
+      "learning_rate": 0.0001800522223487811,
+      "loss": 0.8801,
+      "step": 2842
+    },
+    {
+      "epoch": 0.9128270990528174,
+      "grad_norm": 1.1531569957733154,
+      "learning_rate": 0.0001800312978485683,
+      "loss": 0.9636,
+      "step": 2843
+    },
+    {
+      "epoch": 0.913148177877669,
+      "grad_norm": 2.095679521560669,
+      "learning_rate": 0.00018001036359678469,
+      "loss": 0.9504,
+      "step": 2844
+    },
+    {
+      "epoch": 0.9134692567025204,
+      "grad_norm": 0.9581215381622314,
+      "learning_rate": 0.00017998941959598095,
+      "loss": 0.798,
+      "step": 2845
+    },
+    {
+      "epoch": 0.913790335527372,
+      "grad_norm": 0.78875333070755,
+      "learning_rate": 0.00017996846584870908,
+      "loss": 0.7823,
+      "step": 2846
+    },
+    {
+      "epoch": 0.9141114143522234,
+      "grad_norm": 1.2259639501571655,
+      "learning_rate": 0.0001799475023575222,
+      "loss": 0.8185,
+      "step": 2847
+    },
+    {
+      "epoch": 0.914432493177075,
+      "grad_norm": 1.0644735097885132,
+      "learning_rate": 0.00017992652912497464,
+      "loss": 0.6927,
+      "step": 2848
+    },
+    {
+      "epoch": 0.9147535720019264,
+      "grad_norm": 1.1323128938674927,
+      "learning_rate": 0.00017990554615362198,
+      "loss": 0.7886,
+      "step": 2849
+    },
+    {
+      "epoch": 0.915074650826778,
+      "grad_norm": 0.6815614700317383,
+      "learning_rate": 0.00017988455344602092,
+      "loss": 0.5591,
+      "step": 2850
+    },
+    {
+      "epoch": 0.9153957296516295,
+      "grad_norm": 0.9420286417007446,
+      "learning_rate": 0.00017986355100472928,
+      "loss": 1.0964,
+      "step": 2851
+    },
+    {
+      "epoch": 0.915716808476481,
+      "grad_norm": 0.9961267113685608,
+      "learning_rate": 0.00017984253883230627,
+      "loss": 1.2273,
+      "step": 2852
+    },
+    {
+      "epoch": 0.9160378873013325,
+      "grad_norm": 0.790002167224884,
+      "learning_rate": 0.00017982151693131203,
+      "loss": 0.8297,
+      "step": 2853
+    },
+    {
+      "epoch": 0.916358966126184,
+      "grad_norm": 1.0446445941925049,
+      "learning_rate": 0.0001798004853043081,
+      "loss": 0.7856,
+      "step": 2854
+    },
+    {
+      "epoch": 0.9166800449510355,
+      "grad_norm": 0.9418595433235168,
+      "learning_rate": 0.0001797794439538571,
+      "loss": 0.5708,
+      "step": 2855
+    },
+    {
+      "epoch": 0.9170011237758869,
+      "grad_norm": 1.0176951885223389,
+      "learning_rate": 0.00017975839288252287,
+      "loss": 0.6077,
+      "step": 2856
+    },
+    {
+      "epoch": 0.9173222026007385,
+      "grad_norm": 1.1099709272384644,
+      "learning_rate": 0.00017973733209287036,
+      "loss": 0.7867,
+      "step": 2857
+    },
+    {
+      "epoch": 0.9176432814255899,
+      "grad_norm": 1.4851493835449219,
+      "learning_rate": 0.00017971626158746584,
+      "loss": 1.0428,
+      "step": 2858
+    },
+    {
+      "epoch": 0.9179643602504415,
+      "grad_norm": 1.387786865234375,
+      "learning_rate": 0.00017969518136887663,
+      "loss": 0.9532,
+      "step": 2859
+    },
+    {
+      "epoch": 0.918285439075293,
+      "grad_norm": 0.9932065010070801,
+      "learning_rate": 0.00017967409143967132,
+      "loss": 0.8475,
+      "step": 2860
+    },
+    {
+      "epoch": 0.9186065179001445,
+      "grad_norm": 0.8978595733642578,
+      "learning_rate": 0.00017965299180241963,
+      "loss": 1.0245,
+      "step": 2861
+    },
+    {
+      "epoch": 0.918927596724996,
+      "grad_norm": 1.0536555051803589,
+      "learning_rate": 0.00017963188245969253,
+      "loss": 0.814,
+      "step": 2862
+    },
+    {
+      "epoch": 0.9192486755498475,
+      "grad_norm": 1.004948616027832,
+      "learning_rate": 0.00017961076341406208,
+      "loss": 0.7993,
+      "step": 2863
+    },
+    {
+      "epoch": 0.919569754374699,
+      "grad_norm": 0.9587133526802063,
+      "learning_rate": 0.0001795896346681016,
+      "loss": 0.7497,
+      "step": 2864
+    },
+    {
+      "epoch": 0.9198908331995505,
+      "grad_norm": 0.9380829930305481,
+      "learning_rate": 0.00017956849622438554,
+      "loss": 0.7871,
+      "step": 2865
+    },
+    {
+      "epoch": 0.920211912024402,
+      "grad_norm": 0.9536609053611755,
+      "learning_rate": 0.00017954734808548958,
+      "loss": 0.8406,
+      "step": 2866
+    },
+    {
+      "epoch": 0.9205329908492534,
+      "grad_norm": 0.9843786954879761,
+      "learning_rate": 0.00017952619025399057,
+      "loss": 0.9366,
+      "step": 2867
+    },
+    {
+      "epoch": 0.920854069674105,
+      "grad_norm": 0.8176027536392212,
+      "learning_rate": 0.00017950502273246649,
+      "loss": 0.8689,
+      "step": 2868
+    },
+    {
+      "epoch": 0.9211751484989565,
+      "grad_norm": 0.9969504475593567,
+      "learning_rate": 0.00017948384552349657,
+      "loss": 1.0414,
+      "step": 2869
+    },
+    {
+      "epoch": 0.921496227323808,
+      "grad_norm": 1.0049349069595337,
+      "learning_rate": 0.00017946265862966114,
+      "loss": 0.8943,
+      "step": 2870
+    },
+    {
+      "epoch": 0.9218173061486595,
+      "grad_norm": 1.8080761432647705,
+      "learning_rate": 0.00017944146205354182,
+      "loss": 1.0632,
+      "step": 2871
+    },
+    {
+      "epoch": 0.922138384973511,
+      "grad_norm": 0.8137750625610352,
+      "learning_rate": 0.00017942025579772132,
+      "loss": 0.7279,
+      "step": 2872
+    },
+    {
+      "epoch": 0.9224594637983625,
+      "grad_norm": 0.9603304862976074,
+      "learning_rate": 0.00017939903986478355,
+      "loss": 0.9531,
+      "step": 2873
+    },
+    {
+      "epoch": 0.922780542623214,
+      "grad_norm": 1.556550145149231,
+      "learning_rate": 0.0001793778142573136,
+      "loss": 0.8103,
+      "step": 2874
+    },
+    {
+      "epoch": 0.9231016214480655,
+      "grad_norm": 1.2223728895187378,
+      "learning_rate": 0.0001793565789778978,
+      "loss": 0.9159,
+      "step": 2875
+    },
+    {
+      "epoch": 0.923422700272917,
+      "grad_norm": 1.2416236400604248,
+      "learning_rate": 0.00017933533402912354,
+      "loss": 0.8169,
+      "step": 2876
+    },
+    {
+      "epoch": 0.9237437790977685,
+      "grad_norm": 1.7050341367721558,
+      "learning_rate": 0.00017931407941357947,
+      "loss": 0.9111,
+      "step": 2877
+    },
+    {
+      "epoch": 0.92406485792262,
+      "grad_norm": 1.7200475931167603,
+      "learning_rate": 0.0001792928151338554,
+      "loss": 1.0214,
+      "step": 2878
+    },
+    {
+      "epoch": 0.9243859367474715,
+      "grad_norm": 1.3585445880889893,
+      "learning_rate": 0.00017927154119254236,
+      "loss": 1.0152,
+      "step": 2879
+    },
+    {
+      "epoch": 0.924707015572323,
+      "grad_norm": 1.661501407623291,
+      "learning_rate": 0.00017925025759223245,
+      "loss": 0.7003,
+      "step": 2880
+    },
+    {
+      "epoch": 0.9250280943971745,
+      "grad_norm": 0.8686938285827637,
+      "learning_rate": 0.00017922896433551907,
+      "loss": 0.8456,
+      "step": 2881
+    },
+    {
+      "epoch": 0.925349173222026,
+      "grad_norm": 1.1402549743652344,
+      "learning_rate": 0.00017920766142499672,
+      "loss": 0.91,
+      "step": 2882
+    },
+    {
+      "epoch": 0.9256702520468775,
+      "grad_norm": 1.2883620262145996,
+      "learning_rate": 0.00017918634886326108,
+      "loss": 0.834,
+      "step": 2883
+    },
+    {
+      "epoch": 0.925991330871729,
+      "grad_norm": 0.8248382210731506,
+      "learning_rate": 0.00017916502665290903,
+      "loss": 0.7228,
+      "step": 2884
+    },
+    {
+      "epoch": 0.9263124096965805,
+      "grad_norm": 1.0151740312576294,
+      "learning_rate": 0.0001791436947965386,
+      "loss": 0.8951,
+      "step": 2885
+    },
+    {
+      "epoch": 0.926633488521432,
+      "grad_norm": 1.5327341556549072,
+      "learning_rate": 0.00017912235329674902,
+      "loss": 0.912,
+      "step": 2886
+    },
+    {
+      "epoch": 0.9269545673462836,
+      "grad_norm": 1.1290961503982544,
+      "learning_rate": 0.0001791010021561407,
+      "loss": 0.8542,
+      "step": 2887
+    },
+    {
+      "epoch": 0.927275646171135,
+      "grad_norm": 1.0894137620925903,
+      "learning_rate": 0.0001790796413773152,
+      "loss": 0.9589,
+      "step": 2888
+    },
+    {
+      "epoch": 0.9275967249959866,
+      "grad_norm": 1.0757745504379272,
+      "learning_rate": 0.0001790582709628753,
+      "loss": 0.9288,
+      "step": 2889
+    },
+    {
+      "epoch": 0.927917803820838,
+      "grad_norm": 0.8483561277389526,
+      "learning_rate": 0.0001790368909154249,
+      "loss": 0.852,
+      "step": 2890
+    },
+    {
+      "epoch": 0.9282388826456895,
+      "grad_norm": 1.5198390483856201,
+      "learning_rate": 0.00017901550123756906,
+      "loss": 0.7621,
+      "step": 2891
+    },
+    {
+      "epoch": 0.928559961470541,
+      "grad_norm": 1.2075666189193726,
+      "learning_rate": 0.00017899410193191406,
+      "loss": 0.9783,
+      "step": 2892
+    },
+    {
+      "epoch": 0.9288810402953925,
+      "grad_norm": 1.1517733335494995,
+      "learning_rate": 0.00017897269300106737,
+      "loss": 0.9489,
+      "step": 2893
+    },
+    {
+      "epoch": 0.929202119120244,
+      "grad_norm": 1.0166940689086914,
+      "learning_rate": 0.0001789512744476376,
+      "loss": 0.8871,
+      "step": 2894
+    },
+    {
+      "epoch": 0.9295231979450955,
+      "grad_norm": 0.7956641912460327,
+      "learning_rate": 0.0001789298462742345,
+      "loss": 0.6786,
+      "step": 2895
+    },
+    {
+      "epoch": 0.9298442767699471,
+      "grad_norm": 1.2172540426254272,
+      "learning_rate": 0.00017890840848346908,
+      "loss": 0.7532,
+      "step": 2896
+    },
+    {
+      "epoch": 0.9301653555947985,
+      "grad_norm": 1.0997369289398193,
+      "learning_rate": 0.00017888696107795342,
+      "loss": 0.8759,
+      "step": 2897
+    },
+    {
+      "epoch": 0.9304864344196501,
+      "grad_norm": 0.9977442026138306,
+      "learning_rate": 0.00017886550406030085,
+      "loss": 0.7112,
+      "step": 2898
+    },
+    {
+      "epoch": 0.9308075132445015,
+      "grad_norm": 1.1181156635284424,
+      "learning_rate": 0.00017884403743312582,
+      "loss": 0.6298,
+      "step": 2899
+    },
+    {
+      "epoch": 0.9311285920693531,
+      "grad_norm": 0.6041698455810547,
+      "learning_rate": 0.00017882256119904403,
+      "loss": 0.4929,
+      "step": 2900
+    },
+    {
+      "epoch": 0.9314496708942045,
+      "grad_norm": 0.8137570023536682,
+      "learning_rate": 0.00017880107536067218,
+      "loss": 0.9051,
+      "step": 2901
+    },
+    {
+      "epoch": 0.931770749719056,
+      "grad_norm": 0.838915228843689,
+      "learning_rate": 0.0001787795799206284,
+      "loss": 1.2534,
+      "step": 2902
+    },
+    {
+      "epoch": 0.9320918285439075,
+      "grad_norm": 0.7378472685813904,
+      "learning_rate": 0.00017875807488153175,
+      "loss": 0.6947,
+      "step": 2903
+    },
+    {
+      "epoch": 0.932412907368759,
+      "grad_norm": 0.8197500705718994,
+      "learning_rate": 0.00017873656024600254,
+      "loss": 0.7199,
+      "step": 2904
+    },
+    {
+      "epoch": 0.9327339861936105,
+      "grad_norm": 1.34345281124115,
+      "learning_rate": 0.00017871503601666233,
+      "loss": 0.737,
+      "step": 2905
+    },
+    {
+      "epoch": 0.933055065018462,
+      "grad_norm": 1.0809118747711182,
+      "learning_rate": 0.00017869350219613375,
+      "loss": 0.5367,
+      "step": 2906
+    },
+    {
+      "epoch": 0.9333761438433136,
+      "grad_norm": 0.9395135045051575,
+      "learning_rate": 0.0001786719587870406,
+      "loss": 0.4426,
+      "step": 2907
+    },
+    {
+      "epoch": 0.933697222668165,
+      "grad_norm": 0.8438148498535156,
+      "learning_rate": 0.00017865040579200794,
+      "loss": 0.5694,
+      "step": 2908
+    },
+    {
+      "epoch": 0.9340183014930166,
+      "grad_norm": 0.836493194103241,
+      "learning_rate": 0.00017862884321366188,
+      "loss": 0.8198,
+      "step": 2909
+    },
+    {
+      "epoch": 0.934339380317868,
+      "grad_norm": 1.9980072975158691,
+      "learning_rate": 0.0001786072710546298,
+      "loss": 0.9351,
+      "step": 2910
+    },
+    {
+      "epoch": 0.9346604591427196,
+      "grad_norm": 1.1404621601104736,
+      "learning_rate": 0.0001785856893175402,
+      "loss": 0.93,
+      "step": 2911
+    },
+    {
+      "epoch": 0.934981537967571,
+      "grad_norm": 1.2216063737869263,
+      "learning_rate": 0.00017856409800502272,
+      "loss": 0.811,
+      "step": 2912
+    },
+    {
+      "epoch": 0.9353026167924225,
+      "grad_norm": 0.7324502468109131,
+      "learning_rate": 0.00017854249711970818,
+      "loss": 0.6464,
+      "step": 2913
+    },
+    {
+      "epoch": 0.935623695617274,
+      "grad_norm": 0.8018614053726196,
+      "learning_rate": 0.00017852088666422863,
+      "loss": 0.8989,
+      "step": 2914
+    },
+    {
+      "epoch": 0.9359447744421255,
+      "grad_norm": 0.8852442502975464,
+      "learning_rate": 0.00017849926664121726,
+      "loss": 0.7535,
+      "step": 2915
+    },
+    {
+      "epoch": 0.9362658532669771,
+      "grad_norm": 1.2402286529541016,
+      "learning_rate": 0.0001784776370533083,
+      "loss": 0.9203,
+      "step": 2916
+    },
+    {
+      "epoch": 0.9365869320918285,
+      "grad_norm": 0.9982452988624573,
+      "learning_rate": 0.00017845599790313735,
+      "loss": 0.8699,
+      "step": 2917
+    },
+    {
+      "epoch": 0.9369080109166801,
+      "grad_norm": 1.3398115634918213,
+      "learning_rate": 0.000178434349193341,
+      "loss": 0.9541,
+      "step": 2918
+    },
+    {
+      "epoch": 0.9372290897415315,
+      "grad_norm": 1.2586543560028076,
+      "learning_rate": 0.00017841269092655715,
+      "loss": 0.8738,
+      "step": 2919
+    },
+    {
+      "epoch": 0.9375501685663831,
+      "grad_norm": 0.7111466526985168,
+      "learning_rate": 0.00017839102310542477,
+      "loss": 0.7364,
+      "step": 2920
+    },
+    {
+      "epoch": 0.9378712473912345,
+      "grad_norm": 1.377344012260437,
+      "learning_rate": 0.000178369345732584,
+      "loss": 0.9916,
+      "step": 2921
+    },
+    {
+      "epoch": 0.9381923262160861,
+      "grad_norm": 0.881392240524292,
+      "learning_rate": 0.00017834765881067616,
+      "loss": 0.8958,
+      "step": 2922
+    },
+    {
+      "epoch": 0.9385134050409375,
+      "grad_norm": 0.8309992551803589,
+      "learning_rate": 0.00017832596234234376,
+      "loss": 0.6358,
+      "step": 2923
+    },
+    {
+      "epoch": 0.938834483865789,
+      "grad_norm": 0.9630143046379089,
+      "learning_rate": 0.00017830425633023043,
+      "loss": 0.7877,
+      "step": 2924
+    },
+    {
+      "epoch": 0.9391555626906406,
+      "grad_norm": 0.92414790391922,
+      "learning_rate": 0.000178282540776981,
+      "loss": 0.9282,
+      "step": 2925
+    },
+    {
+      "epoch": 0.939476641515492,
+      "grad_norm": 1.1467437744140625,
+      "learning_rate": 0.0001782608156852414,
+      "loss": 0.8531,
+      "step": 2926
+    },
+    {
+      "epoch": 0.9397977203403436,
+      "grad_norm": 1.1309285163879395,
+      "learning_rate": 0.0001782390810576588,
+      "loss": 0.969,
+      "step": 2927
+    },
+    {
+      "epoch": 0.940118799165195,
+      "grad_norm": 1.0550163984298706,
+      "learning_rate": 0.00017821733689688153,
+      "loss": 0.8446,
+      "step": 2928
+    },
+    {
+      "epoch": 0.9404398779900466,
+      "grad_norm": 0.9224303364753723,
+      "learning_rate": 0.000178195583205559,
+      "loss": 0.9563,
+      "step": 2929
+    },
+    {
+      "epoch": 0.940760956814898,
+      "grad_norm": 1.0767320394515991,
+      "learning_rate": 0.00017817381998634185,
+      "loss": 0.8241,
+      "step": 2930
+    },
+    {
+      "epoch": 0.9410820356397496,
+      "grad_norm": 0.7739982604980469,
+      "learning_rate": 0.00017815204724188187,
+      "loss": 0.5874,
+      "step": 2931
+    },
+    {
+      "epoch": 0.941403114464601,
+      "grad_norm": 1.8265200853347778,
+      "learning_rate": 0.000178130264974832,
+      "loss": 1.0106,
+      "step": 2932
+    },
+    {
+      "epoch": 0.9417241932894526,
+      "grad_norm": 0.7935619950294495,
+      "learning_rate": 0.0001781084731878463,
+      "loss": 0.7087,
+      "step": 2933
+    },
+    {
+      "epoch": 0.9420452721143041,
+      "grad_norm": 1.314613938331604,
+      "learning_rate": 0.00017808667188358012,
+      "loss": 1.1337,
+      "step": 2934
+    },
+    {
+      "epoch": 0.9423663509391556,
+      "grad_norm": 1.5377808809280396,
+      "learning_rate": 0.00017806486106468981,
+      "loss": 1.074,
+      "step": 2935
+    },
+    {
+      "epoch": 0.9426874297640071,
+      "grad_norm": 0.8148975968360901,
+      "learning_rate": 0.000178043040733833,
+      "loss": 0.6838,
+      "step": 2936
+    },
+    {
+      "epoch": 0.9430085085888585,
+      "grad_norm": 1.7382673025131226,
+      "learning_rate": 0.00017802121089366836,
+      "loss": 0.731,
+      "step": 2937
+    },
+    {
+      "epoch": 0.9433295874137101,
+      "grad_norm": 1.164584994316101,
+      "learning_rate": 0.00017799937154685586,
+      "loss": 1.0691,
+      "step": 2938
+    },
+    {
+      "epoch": 0.9436506662385615,
+      "grad_norm": 1.2241766452789307,
+      "learning_rate": 0.00017797752269605653,
+      "loss": 0.9289,
+      "step": 2939
+    },
+    {
+      "epoch": 0.9439717450634131,
+      "grad_norm": 1.3293641805648804,
+      "learning_rate": 0.00017795566434393258,
+      "loss": 0.7479,
+      "step": 2940
+    },
+    {
+      "epoch": 0.9442928238882645,
+      "grad_norm": 1.1661382913589478,
+      "learning_rate": 0.00017793379649314744,
+      "loss": 0.9162,
+      "step": 2941
+    },
+    {
+      "epoch": 0.9446139027131161,
+      "grad_norm": 1.2080588340759277,
+      "learning_rate": 0.00017791191914636554,
+      "loss": 0.7161,
+      "step": 2942
+    },
+    {
+      "epoch": 0.9449349815379676,
+      "grad_norm": 1.0982539653778076,
+      "learning_rate": 0.00017789003230625266,
+      "loss": 0.8409,
+      "step": 2943
+    },
+    {
+      "epoch": 0.9452560603628191,
+      "grad_norm": 1.0143113136291504,
+      "learning_rate": 0.00017786813597547562,
+      "loss": 0.7521,
+      "step": 2944
+    },
+    {
+      "epoch": 0.9455771391876706,
+      "grad_norm": 1.105790615081787,
+      "learning_rate": 0.00017784623015670238,
+      "loss": 0.9074,
+      "step": 2945
+    },
+    {
+      "epoch": 0.945898218012522,
+      "grad_norm": 0.9993615746498108,
+      "learning_rate": 0.00017782431485260212,
+      "loss": 0.8263,
+      "step": 2946
+    },
+    {
+      "epoch": 0.9462192968373736,
+      "grad_norm": 0.969996452331543,
+      "learning_rate": 0.00017780239006584515,
+      "loss": 0.6984,
+      "step": 2947
+    },
+    {
+      "epoch": 0.946540375662225,
+      "grad_norm": 0.934131383895874,
+      "learning_rate": 0.00017778045579910302,
+      "loss": 0.7838,
+      "step": 2948
+    },
+    {
+      "epoch": 0.9468614544870766,
+      "grad_norm": 0.7031441330909729,
+      "learning_rate": 0.0001777585120550482,
+      "loss": 0.4672,
+      "step": 2949
+    },
+    {
+      "epoch": 0.947182533311928,
+      "grad_norm": 1.3736737966537476,
+      "learning_rate": 0.0001777365588363546,
+      "loss": 0.6343,
+      "step": 2950
+    },
+    {
+      "epoch": 0.9475036121367796,
+      "grad_norm": 1.2793693542480469,
+      "learning_rate": 0.0001777145961456971,
+      "loss": 1.0671,
+      "step": 2951
+    },
+    {
+      "epoch": 0.9478246909616311,
+      "grad_norm": 1.0697038173675537,
+      "learning_rate": 0.0001776926239857518,
+      "loss": 0.9461,
+      "step": 2952
+    },
+    {
+      "epoch": 0.9481457697864826,
+      "grad_norm": 1.163883090019226,
+      "learning_rate": 0.00017767064235919592,
+      "loss": 0.8222,
+      "step": 2953
+    },
+    {
+      "epoch": 0.9484668486113341,
+      "grad_norm": 1.1357632875442505,
+      "learning_rate": 0.00017764865126870786,
+      "loss": 0.6185,
+      "step": 2954
+    },
+    {
+      "epoch": 0.9487879274361856,
+      "grad_norm": 1.044090986251831,
+      "learning_rate": 0.0001776266507169672,
+      "loss": 0.5034,
+      "step": 2955
+    },
+    {
+      "epoch": 0.9491090062610371,
+      "grad_norm": 1.0913854837417603,
+      "learning_rate": 0.0001776046407066546,
+      "loss": 0.4259,
+      "step": 2956
+    },
+    {
+      "epoch": 0.9494300850858886,
+      "grad_norm": 0.9846095442771912,
+      "learning_rate": 0.00017758262124045195,
+      "loss": 0.464,
+      "step": 2957
+    },
+    {
+      "epoch": 0.9497511639107401,
+      "grad_norm": 1.2047371864318848,
+      "learning_rate": 0.0001775605923210422,
+      "loss": 0.9376,
+      "step": 2958
+    },
+    {
+      "epoch": 0.9500722427355915,
+      "grad_norm": 2.1599411964416504,
+      "learning_rate": 0.0001775385539511096,
+      "loss": 1.0841,
+      "step": 2959
+    },
+    {
+      "epoch": 0.9503933215604431,
+      "grad_norm": 1.1163110733032227,
+      "learning_rate": 0.00017751650613333935,
+      "loss": 0.8837,
+      "step": 2960
+    },
+    {
+      "epoch": 0.9507144003852946,
+      "grad_norm": 1.0459928512573242,
+      "learning_rate": 0.00017749444887041799,
+      "loss": 0.9614,
+      "step": 2961
+    },
+    {
+      "epoch": 0.9510354792101461,
+      "grad_norm": 0.897926390171051,
+      "learning_rate": 0.00017747238216503307,
+      "loss": 0.8908,
+      "step": 2962
+    },
+    {
+      "epoch": 0.9513565580349976,
+      "grad_norm": 0.9355558753013611,
+      "learning_rate": 0.00017745030601987337,
+      "loss": 0.8321,
+      "step": 2963
+    },
+    {
+      "epoch": 0.9516776368598491,
+      "grad_norm": 0.7442376613616943,
+      "learning_rate": 0.00017742822043762888,
+      "loss": 0.627,
+      "step": 2964
+    },
+    {
+      "epoch": 0.9519987156847006,
+      "grad_norm": 0.7983213067054749,
+      "learning_rate": 0.00017740612542099053,
+      "loss": 0.8255,
+      "step": 2965
+    },
+    {
+      "epoch": 0.9523197945095521,
+      "grad_norm": 1.141676425933838,
+      "learning_rate": 0.00017738402097265064,
+      "loss": 0.9653,
+      "step": 2966
+    },
+    {
+      "epoch": 0.9526408733344036,
+      "grad_norm": 0.9923720955848694,
+      "learning_rate": 0.0001773619070953025,
+      "loss": 0.7411,
+      "step": 2967
+    },
+    {
+      "epoch": 0.952961952159255,
+      "grad_norm": 1.1766963005065918,
+      "learning_rate": 0.00017733978379164066,
+      "loss": 1.008,
+      "step": 2968
+    },
+    {
+      "epoch": 0.9532830309841066,
+      "grad_norm": 1.0631762742996216,
+      "learning_rate": 0.00017731765106436073,
+      "loss": 0.8621,
+      "step": 2969
+    },
+    {
+      "epoch": 0.953604109808958,
+      "grad_norm": 0.9990087747573853,
+      "learning_rate": 0.00017729550891615957,
+      "loss": 0.8143,
+      "step": 2970
+    },
+    {
+      "epoch": 0.9539251886338096,
+      "grad_norm": 0.9863772988319397,
+      "learning_rate": 0.00017727335734973512,
+      "loss": 0.8346,
+      "step": 2971
+    },
+    {
+      "epoch": 0.9542462674586611,
+      "grad_norm": 1.3948047161102295,
+      "learning_rate": 0.00017725119636778644,
+      "loss": 0.7758,
+      "step": 2972
+    },
+    {
+      "epoch": 0.9545673462835126,
+      "grad_norm": 1.3944754600524902,
+      "learning_rate": 0.00017722902597301383,
+      "loss": 0.9348,
+      "step": 2973
+    },
+    {
+      "epoch": 0.9548884251083641,
+      "grad_norm": 1.196573257446289,
+      "learning_rate": 0.00017720684616811866,
+      "loss": 0.9216,
+      "step": 2974
+    },
+    {
+      "epoch": 0.9552095039332156,
+      "grad_norm": 1.4773893356323242,
+      "learning_rate": 0.0001771846569558035,
+      "loss": 1.0486,
+      "step": 2975
+    },
+    {
+      "epoch": 0.9555305827580671,
+      "grad_norm": 1.0879054069519043,
+      "learning_rate": 0.00017716245833877201,
+      "loss": 0.7509,
+      "step": 2976
+    },
+    {
+      "epoch": 0.9558516615829186,
+      "grad_norm": 0.9991624355316162,
+      "learning_rate": 0.00017714025031972903,
+      "loss": 0.7418,
+      "step": 2977
+    },
+    {
+      "epoch": 0.9561727404077701,
+      "grad_norm": 1.0229377746582031,
+      "learning_rate": 0.00017711803290138052,
+      "loss": 0.8292,
+      "step": 2978
+    },
+    {
+      "epoch": 0.9564938192326216,
+      "grad_norm": 0.9216927289962769,
+      "learning_rate": 0.00017709580608643363,
+      "loss": 0.7256,
+      "step": 2979
+    },
+    {
+      "epoch": 0.9568148980574731,
+      "grad_norm": 1.1415516138076782,
+      "learning_rate": 0.0001770735698775966,
+      "loss": 0.9657,
+      "step": 2980
+    },
+    {
+      "epoch": 0.9571359768823247,
+      "grad_norm": 1.4062246084213257,
+      "learning_rate": 0.00017705132427757895,
+      "loss": 1.0777,
+      "step": 2981
+    },
+    {
+      "epoch": 0.9574570557071761,
+      "grad_norm": 1.2299398183822632,
+      "learning_rate": 0.00017702906928909108,
+      "loss": 1.1786,
+      "step": 2982
+    },
+    {
+      "epoch": 0.9577781345320276,
+      "grad_norm": 1.3134931325912476,
+      "learning_rate": 0.0001770068049148448,
+      "loss": 1.0092,
+      "step": 2983
+    },
+    {
+      "epoch": 0.9580992133568791,
+      "grad_norm": 1.0825444459915161,
+      "learning_rate": 0.00017698453115755293,
+      "loss": 0.6504,
+      "step": 2984
+    },
+    {
+      "epoch": 0.9584202921817306,
+      "grad_norm": 1.2085427045822144,
+      "learning_rate": 0.00017696224801992945,
+      "loss": 0.9662,
+      "step": 2985
+    },
+    {
+      "epoch": 0.9587413710065821,
+      "grad_norm": 1.1950358152389526,
+      "learning_rate": 0.0001769399555046895,
+      "loss": 0.8202,
+      "step": 2986
+    },
+    {
+      "epoch": 0.9590624498314336,
+      "grad_norm": 1.1132547855377197,
+      "learning_rate": 0.00017691765361454938,
+      "loss": 0.9621,
+      "step": 2987
+    },
+    {
+      "epoch": 0.9593835286562851,
+      "grad_norm": 0.9960354566574097,
+      "learning_rate": 0.00017689534235222648,
+      "loss": 0.9149,
+      "step": 2988
+    },
+    {
+      "epoch": 0.9597046074811366,
+      "grad_norm": 1.1196013689041138,
+      "learning_rate": 0.00017687302172043933,
+      "loss": 0.8249,
+      "step": 2989
+    },
+    {
+      "epoch": 0.9600256863059882,
+      "grad_norm": 1.1874531507492065,
+      "learning_rate": 0.00017685069172190766,
+      "loss": 0.8245,
+      "step": 2990
+    },
+    {
+      "epoch": 0.9603467651308396,
+      "grad_norm": 1.0298069715499878,
+      "learning_rate": 0.00017682835235935236,
+      "loss": 0.7936,
+      "step": 2991
+    },
+    {
+      "epoch": 0.9606678439556912,
+      "grad_norm": 1.3679097890853882,
+      "learning_rate": 0.00017680600363549533,
+      "loss": 0.8939,
+      "step": 2992
+    },
+    {
+      "epoch": 0.9609889227805426,
+      "grad_norm": 0.9797439575195312,
+      "learning_rate": 0.00017678364555305978,
+      "loss": 0.6787,
+      "step": 2993
+    },
+    {
+      "epoch": 0.9613100016053941,
+      "grad_norm": 1.0256136655807495,
+      "learning_rate": 0.00017676127811476987,
+      "loss": 0.6678,
+      "step": 2994
+    },
+    {
+      "epoch": 0.9616310804302456,
+      "grad_norm": 1.0464802980422974,
+      "learning_rate": 0.0001767389013233511,
+      "loss": 0.6863,
+      "step": 2995
+    },
+    {
+      "epoch": 0.9619521592550971,
+      "grad_norm": 1.0149427652359009,
+      "learning_rate": 0.00017671651518153,
+      "loss": 0.6783,
+      "step": 2996
+    },
+    {
+      "epoch": 0.9622732380799486,
+      "grad_norm": 0.8050268292427063,
+      "learning_rate": 0.00017669411969203417,
+      "loss": 0.7203,
+      "step": 2997
+    },
+    {
+      "epoch": 0.9625943169048001,
+      "grad_norm": 0.896780252456665,
+      "learning_rate": 0.00017667171485759252,
+      "loss": 0.575,
+      "step": 2998
+    },
+    {
+      "epoch": 0.9629153957296517,
+      "grad_norm": 0.813828706741333,
+      "learning_rate": 0.00017664930068093498,
+      "loss": 0.7544,
+      "step": 2999
+    },
+    {
+      "epoch": 0.9632364745545031,
+      "grad_norm": 0.922319769859314,
+      "learning_rate": 0.00017662687716479266,
+      "loss": 0.5257,
+      "step": 3000
+    },
+    {
+      "epoch": 0.9635575533793547,
+      "grad_norm": 1.0877915620803833,
+      "learning_rate": 0.0001766044443118978,
+      "loss": 1.0061,
+      "step": 3001
+    },
+    {
+      "epoch": 0.9638786322042061,
+      "grad_norm": 1.045781135559082,
+      "learning_rate": 0.00017658200212498378,
+      "loss": 0.9707,
+      "step": 3002
+    },
+    {
+      "epoch": 0.9641997110290577,
+      "grad_norm": 1.023611307144165,
+      "learning_rate": 0.00017655955060678506,
+      "loss": 0.7057,
+      "step": 3003
+    },
+    {
+      "epoch": 0.9645207898539091,
+      "grad_norm": 1.043655514717102,
+      "learning_rate": 0.00017653708976003737,
+      "loss": 0.5738,
+      "step": 3004
+    },
+    {
+      "epoch": 0.9648418686787607,
+      "grad_norm": 0.8784222602844238,
+      "learning_rate": 0.00017651461958747745,
+      "loss": 0.5103,
+      "step": 3005
+    },
+    {
+      "epoch": 0.9651629475036121,
+      "grad_norm": 1.0596736669540405,
+      "learning_rate": 0.0001764921400918432,
+      "loss": 0.7845,
+      "step": 3006
+    },
+    {
+      "epoch": 0.9654840263284636,
+      "grad_norm": 1.0180422067642212,
+      "learning_rate": 0.0001764696512758737,
+      "loss": 0.8355,
+      "step": 3007
+    },
+    {
+      "epoch": 0.9658051051533152,
+      "grad_norm": 0.8986467719078064,
+      "learning_rate": 0.00017644715314230918,
+      "loss": 0.8457,
+      "step": 3008
+    },
+    {
+      "epoch": 0.9661261839781666,
+      "grad_norm": 0.9878740906715393,
+      "learning_rate": 0.0001764246456938909,
+      "loss": 0.7791,
+      "step": 3009
+    },
+    {
+      "epoch": 0.9664472628030182,
+      "grad_norm": 0.8736189603805542,
+      "learning_rate": 0.00017640212893336142,
+      "loss": 0.8261,
+      "step": 3010
+    },
+    {
+      "epoch": 0.9667683416278696,
+      "grad_norm": 0.9508904218673706,
+      "learning_rate": 0.00017637960286346425,
+      "loss": 0.8671,
+      "step": 3011
+    },
+    {
+      "epoch": 0.9670894204527212,
+      "grad_norm": 1.029610514640808,
+      "learning_rate": 0.00017635706748694413,
+      "loss": 0.8247,
+      "step": 3012
+    },
+    {
+      "epoch": 0.9674104992775726,
+      "grad_norm": 1.0193253755569458,
+      "learning_rate": 0.000176334522806547,
+      "loss": 0.8908,
+      "step": 3013
+    },
+    {
+      "epoch": 0.9677315781024242,
+      "grad_norm": 1.048719882965088,
+      "learning_rate": 0.00017631196882501973,
+      "loss": 0.985,
+      "step": 3014
+    },
+    {
+      "epoch": 0.9680526569272756,
+      "grad_norm": 0.8610148429870605,
+      "learning_rate": 0.00017628940554511061,
+      "loss": 0.7015,
+      "step": 3015
+    },
+    {
+      "epoch": 0.9683737357521272,
+      "grad_norm": 0.8026429414749146,
+      "learning_rate": 0.00017626683296956882,
+      "loss": 0.7906,
+      "step": 3016
+    },
+    {
+      "epoch": 0.9686948145769787,
+      "grad_norm": 1.1094107627868652,
+      "learning_rate": 0.0001762442511011448,
+      "loss": 1.057,
+      "step": 3017
+    },
+    {
+      "epoch": 0.9690158934018301,
+      "grad_norm": 1.203913688659668,
+      "learning_rate": 0.00017622165994259,
+      "loss": 0.7759,
+      "step": 3018
+    },
+    {
+      "epoch": 0.9693369722266817,
+      "grad_norm": 0.8827437162399292,
+      "learning_rate": 0.0001761990594966572,
+      "loss": 0.948,
+      "step": 3019
+    },
+    {
+      "epoch": 0.9696580510515331,
+      "grad_norm": 1.1221593618392944,
+      "learning_rate": 0.0001761764497661001,
+      "loss": 0.9794,
+      "step": 3020
+    },
+    {
+      "epoch": 0.9699791298763847,
+      "grad_norm": 0.9867079854011536,
+      "learning_rate": 0.0001761538307536737,
+      "loss": 1.082,
+      "step": 3021
+    },
+    {
+      "epoch": 0.9703002087012361,
+      "grad_norm": 5.697073936462402,
+      "learning_rate": 0.000176131202462134,
+      "loss": 1.1723,
+      "step": 3022
+    },
+    {
+      "epoch": 0.9706212875260877,
+      "grad_norm": 0.8379837870597839,
+      "learning_rate": 0.0001761085648942382,
+      "loss": 0.6992,
+      "step": 3023
+    },
+    {
+      "epoch": 0.9709423663509391,
+      "grad_norm": 1.10336172580719,
+      "learning_rate": 0.00017608591805274464,
+      "loss": 0.9453,
+      "step": 3024
+    },
+    {
+      "epoch": 0.9712634451757907,
+      "grad_norm": 1.2008827924728394,
+      "learning_rate": 0.00017606326194041273,
+      "loss": 0.8227,
+      "step": 3025
+    },
+    {
+      "epoch": 0.9715845240006421,
+      "grad_norm": 0.9085438251495361,
+      "learning_rate": 0.0001760405965600031,
+      "loss": 0.8305,
+      "step": 3026
+    },
+    {
+      "epoch": 0.9719056028254937,
+      "grad_norm": 1.2480196952819824,
+      "learning_rate": 0.00017601792191427741,
+      "loss": 1.0148,
+      "step": 3027
+    },
+    {
+      "epoch": 0.9722266816503452,
+      "grad_norm": 0.7692457437515259,
+      "learning_rate": 0.0001759952380059986,
+      "loss": 0.7948,
+      "step": 3028
+    },
+    {
+      "epoch": 0.9725477604751966,
+      "grad_norm": 1.0976073741912842,
+      "learning_rate": 0.00017597254483793048,
+      "loss": 0.9424,
+      "step": 3029
+    },
+    {
+      "epoch": 0.9728688393000482,
+      "grad_norm": 0.9526044130325317,
+      "learning_rate": 0.00017594984241283825,
+      "loss": 0.8496,
+      "step": 3030
+    },
+    {
+      "epoch": 0.9731899181248996,
+      "grad_norm": 0.7856842279434204,
+      "learning_rate": 0.00017592713073348807,
+      "loss": 0.6682,
+      "step": 3031
+    },
+    {
+      "epoch": 0.9735109969497512,
+      "grad_norm": 1.1540440320968628,
+      "learning_rate": 0.00017590440980264738,
+      "loss": 0.8165,
+      "step": 3032
+    },
+    {
+      "epoch": 0.9738320757746026,
+      "grad_norm": 1.6081081628799438,
+      "learning_rate": 0.00017588167962308458,
+      "loss": 1.0609,
+      "step": 3033
+    },
+    {
+      "epoch": 0.9741531545994542,
+      "grad_norm": 0.7435811758041382,
+      "learning_rate": 0.00017585894019756925,
+      "loss": 0.6044,
+      "step": 3034
+    },
+    {
+      "epoch": 0.9744742334243056,
+      "grad_norm": 0.9327528476715088,
+      "learning_rate": 0.0001758361915288722,
+      "loss": 0.8865,
+      "step": 3035
+    },
+    {
+      "epoch": 0.9747953122491572,
+      "grad_norm": 0.8505873680114746,
+      "learning_rate": 0.00017581343361976524,
+      "loss": 0.7442,
+      "step": 3036
+    },
+    {
+      "epoch": 0.9751163910740087,
+      "grad_norm": 0.8487008213996887,
+      "learning_rate": 0.00017579066647302133,
+      "loss": 0.7235,
+      "step": 3037
+    },
+    {
+      "epoch": 0.9754374698988602,
+      "grad_norm": 1.0709935426712036,
+      "learning_rate": 0.00017576789009141465,
+      "loss": 0.9574,
+      "step": 3038
+    },
+    {
+      "epoch": 0.9757585487237117,
+      "grad_norm": 0.9425415992736816,
+      "learning_rate": 0.00017574510447772039,
+      "loss": 0.7343,
+      "step": 3039
+    },
+    {
+      "epoch": 0.9760796275485631,
+      "grad_norm": 1.0272222757339478,
+      "learning_rate": 0.00017572230963471488,
+      "loss": 0.9407,
+      "step": 3040
+    },
+    {
+      "epoch": 0.9764007063734147,
+      "grad_norm": 1.2393730878829956,
+      "learning_rate": 0.00017569950556517566,
+      "loss": 0.93,
+      "step": 3041
+    },
+    {
+      "epoch": 0.9767217851982661,
+      "grad_norm": 1.0345895290374756,
+      "learning_rate": 0.00017567669227188128,
+      "loss": 0.6844,
+      "step": 3042
+    },
+    {
+      "epoch": 0.9770428640231177,
+      "grad_norm": 1.0678569078445435,
+      "learning_rate": 0.0001756538697576115,
+      "loss": 0.5873,
+      "step": 3043
+    },
+    {
+      "epoch": 0.9773639428479691,
+      "grad_norm": 1.1300995349884033,
+      "learning_rate": 0.0001756310380251472,
+      "loss": 0.8044,
+      "step": 3044
+    },
+    {
+      "epoch": 0.9776850216728207,
+      "grad_norm": 1.4100874662399292,
+      "learning_rate": 0.00017560819707727033,
+      "loss": 0.9445,
+      "step": 3045
+    },
+    {
+      "epoch": 0.9780061004976722,
+      "grad_norm": 0.9837530255317688,
+      "learning_rate": 0.00017558534691676397,
+      "loss": 0.6677,
+      "step": 3046
+    },
+    {
+      "epoch": 0.9783271793225237,
+      "grad_norm": 1.2359012365341187,
+      "learning_rate": 0.00017556248754641235,
+      "loss": 0.8114,
+      "step": 3047
+    },
+    {
+      "epoch": 0.9786482581473752,
+      "grad_norm": 0.6732703447341919,
+      "learning_rate": 0.00017553961896900087,
+      "loss": 0.5295,
+      "step": 3048
+    },
+    {
+      "epoch": 0.9789693369722267,
+      "grad_norm": 1.0407007932662964,
+      "learning_rate": 0.00017551674118731591,
+      "loss": 0.714,
+      "step": 3049
+    },
+    {
+      "epoch": 0.9792904157970782,
+      "grad_norm": 0.6249756217002869,
+      "learning_rate": 0.00017549385420414514,
+      "loss": 0.4261,
+      "step": 3050
+    },
+    {
+      "epoch": 0.9796114946219296,
+      "grad_norm": 0.8719810247421265,
+      "learning_rate": 0.00017547095802227723,
+      "loss": 0.9618,
+      "step": 3051
+    },
+    {
+      "epoch": 0.9799325734467812,
+      "grad_norm": 0.9490910768508911,
+      "learning_rate": 0.00017544805264450196,
+      "loss": 0.925,
+      "step": 3052
+    },
+    {
+      "epoch": 0.9802536522716326,
+      "grad_norm": 1.0451079607009888,
+      "learning_rate": 0.00017542513807361037,
+      "loss": 0.8166,
+      "step": 3053
+    },
+    {
+      "epoch": 0.9805747310964842,
+      "grad_norm": 1.0744620561599731,
+      "learning_rate": 0.00017540221431239453,
+      "loss": 0.5847,
+      "step": 3054
+    },
+    {
+      "epoch": 0.9808958099213357,
+      "grad_norm": 1.3825839757919312,
+      "learning_rate": 0.00017537928136364755,
+      "loss": 0.4744,
+      "step": 3055
+    },
+    {
+      "epoch": 0.9812168887461872,
+      "grad_norm": 1.0428171157836914,
+      "learning_rate": 0.0001753563392301638,
+      "loss": 0.7581,
+      "step": 3056
+    },
+    {
+      "epoch": 0.9815379675710387,
+      "grad_norm": 1.382125973701477,
+      "learning_rate": 0.0001753333879147387,
+      "loss": 0.9721,
+      "step": 3057
+    },
+    {
+      "epoch": 0.9818590463958902,
+      "grad_norm": 0.9676405787467957,
+      "learning_rate": 0.00017531042742016876,
+      "loss": 0.9361,
+      "step": 3058
+    },
+    {
+      "epoch": 0.9821801252207417,
+      "grad_norm": 0.9411641955375671,
+      "learning_rate": 0.00017528745774925172,
+      "loss": 0.9277,
+      "step": 3059
+    },
+    {
+      "epoch": 0.9825012040455932,
+      "grad_norm": 1.0995005369186401,
+      "learning_rate": 0.00017526447890478633,
+      "loss": 0.9208,
+      "step": 3060
+    },
+    {
+      "epoch": 0.9828222828704447,
+      "grad_norm": 0.9930598735809326,
+      "learning_rate": 0.00017524149088957245,
+      "loss": 0.798,
+      "step": 3061
+    },
+    {
+      "epoch": 0.9831433616952961,
+      "grad_norm": 0.6785845756530762,
+      "learning_rate": 0.00017521849370641114,
+      "loss": 0.8495,
+      "step": 3062
+    },
+    {
+      "epoch": 0.9834644405201477,
+      "grad_norm": 0.7817926406860352,
+      "learning_rate": 0.00017519548735810456,
+      "loss": 0.787,
+      "step": 3063
+    },
+    {
+      "epoch": 0.9837855193449992,
+      "grad_norm": 0.8893129825592041,
+      "learning_rate": 0.00017517247184745593,
+      "loss": 0.9347,
+      "step": 3064
+    },
+    {
+      "epoch": 0.9841065981698507,
+      "grad_norm": 0.8218941688537598,
+      "learning_rate": 0.00017514944717726962,
+      "loss": 0.7465,
+      "step": 3065
+    },
+    {
+      "epoch": 0.9844276769947022,
+      "grad_norm": 1.1597498655319214,
+      "learning_rate": 0.00017512641335035113,
+      "loss": 0.9377,
+      "step": 3066
+    },
+    {
+      "epoch": 0.9847487558195537,
+      "grad_norm": 1.1089355945587158,
+      "learning_rate": 0.00017510337036950703,
+      "loss": 1.0179,
+      "step": 3067
+    },
+    {
+      "epoch": 0.9850698346444052,
+      "grad_norm": 0.9803544878959656,
+      "learning_rate": 0.0001750803182375451,
+      "loss": 1.0501,
+      "step": 3068
+    },
+    {
+      "epoch": 0.9853909134692567,
+      "grad_norm": 1.3686444759368896,
+      "learning_rate": 0.00017505725695727412,
+      "loss": 1.0379,
+      "step": 3069
+    },
+    {
+      "epoch": 0.9857119922941082,
+      "grad_norm": 1.3570899963378906,
+      "learning_rate": 0.00017503418653150405,
+      "loss": 0.8903,
+      "step": 3070
+    },
+    {
+      "epoch": 0.9860330711189597,
+      "grad_norm": 1.031009554862976,
+      "learning_rate": 0.00017501110696304596,
+      "loss": 0.8283,
+      "step": 3071
+    },
+    {
+      "epoch": 0.9863541499438112,
+      "grad_norm": 1.3445180654525757,
+      "learning_rate": 0.00017498801825471203,
+      "loss": 0.9019,
+      "step": 3072
+    },
+    {
+      "epoch": 0.9866752287686628,
+      "grad_norm": 0.8621981143951416,
+      "learning_rate": 0.00017496492040931552,
+      "loss": 0.7137,
+      "step": 3073
+    },
+    {
+      "epoch": 0.9869963075935142,
+      "grad_norm": 1.229825735092163,
+      "learning_rate": 0.00017494181342967083,
+      "loss": 1.0208,
+      "step": 3074
+    },
+    {
+      "epoch": 0.9873173864183657,
+      "grad_norm": 0.9832011461257935,
+      "learning_rate": 0.00017491869731859353,
+      "loss": 0.7463,
+      "step": 3075
+    },
+    {
+      "epoch": 0.9876384652432172,
+      "grad_norm": 0.7911801934242249,
+      "learning_rate": 0.00017489557207890023,
+      "loss": 0.8144,
+      "step": 3076
+    },
+    {
+      "epoch": 0.9879595440680687,
+      "grad_norm": 0.849236011505127,
+      "learning_rate": 0.0001748724377134086,
+      "loss": 0.7658,
+      "step": 3077
+    },
+    {
+      "epoch": 0.9882806228929202,
+      "grad_norm": 0.7961871027946472,
+      "learning_rate": 0.0001748492942249376,
+      "loss": 0.8704,
+      "step": 3078
+    },
+    {
+      "epoch": 0.9886017017177717,
+      "grad_norm": 1.293888807296753,
+      "learning_rate": 0.00017482614161630714,
+      "loss": 0.9067,
+      "step": 3079
+    },
+    {
+      "epoch": 0.9889227805426232,
+      "grad_norm": 1.098433256149292,
+      "learning_rate": 0.00017480297989033825,
+      "loss": 0.9446,
+      "step": 3080
+    },
+    {
+      "epoch": 0.9892438593674747,
+      "grad_norm": 1.4394235610961914,
+      "learning_rate": 0.0001747798090498532,
+      "loss": 0.7206,
+      "step": 3081
+    },
+    {
+      "epoch": 0.9895649381923263,
+      "grad_norm": 1.047020435333252,
+      "learning_rate": 0.00017475662909767522,
+      "loss": 0.9727,
+      "step": 3082
+    },
+    {
+      "epoch": 0.9898860170171777,
+      "grad_norm": 0.9796097278594971,
+      "learning_rate": 0.00017473344003662877,
+      "loss": 0.8787,
+      "step": 3083
+    },
+    {
+      "epoch": 0.9902070958420293,
+      "grad_norm": 0.9924256801605225,
+      "learning_rate": 0.00017471024186953936,
+      "loss": 0.7212,
+      "step": 3084
+    },
+    {
+      "epoch": 0.9905281746668807,
+      "grad_norm": 1.5609381198883057,
+      "learning_rate": 0.00017468703459923356,
+      "loss": 0.9614,
+      "step": 3085
+    },
+    {
+      "epoch": 0.9908492534917323,
+      "grad_norm": 0.9042816162109375,
+      "learning_rate": 0.00017466381822853915,
+      "loss": 0.8655,
+      "step": 3086
+    },
+    {
+      "epoch": 0.9911703323165837,
+      "grad_norm": 1.2502557039260864,
+      "learning_rate": 0.00017464059276028497,
+      "loss": 0.8242,
+      "step": 3087
+    },
+    {
+      "epoch": 0.9914914111414352,
+      "grad_norm": 1.919909119606018,
+      "learning_rate": 0.00017461735819730096,
+      "loss": 0.8001,
+      "step": 3088
+    },
+    {
+      "epoch": 0.9918124899662867,
+      "grad_norm": 1.0362643003463745,
+      "learning_rate": 0.00017459411454241822,
+      "loss": 0.6697,
+      "step": 3089
+    },
+    {
+      "epoch": 0.9921335687911382,
+      "grad_norm": 1.6059205532073975,
+      "learning_rate": 0.00017457086179846888,
+      "loss": 0.898,
+      "step": 3090
+    },
+    {
+      "epoch": 0.9924546476159897,
+      "grad_norm": 1.2551214694976807,
+      "learning_rate": 0.00017454759996828623,
+      "loss": 0.8499,
+      "step": 3091
+    },
+    {
+      "epoch": 0.9927757264408412,
+      "grad_norm": 0.9736458659172058,
+      "learning_rate": 0.00017452432905470464,
+      "loss": 0.8884,
+      "step": 3092
+    },
+    {
+      "epoch": 0.9930968052656928,
+      "grad_norm": 1.4679756164550781,
+      "learning_rate": 0.00017450104906055963,
+      "loss": 0.9162,
+      "step": 3093
+    },
+    {
+      "epoch": 0.9934178840905442,
+      "grad_norm": 0.9897777438163757,
+      "learning_rate": 0.00017447775998868776,
+      "loss": 0.7051,
+      "step": 3094
+    },
+    {
+      "epoch": 0.9937389629153958,
+      "grad_norm": 0.9767149686813354,
+      "learning_rate": 0.00017445446184192676,
+      "loss": 0.7716,
+      "step": 3095
+    },
+    {
+      "epoch": 0.9940600417402472,
+      "grad_norm": 1.2053884267807007,
+      "learning_rate": 0.0001744311546231154,
+      "loss": 0.8858,
+      "step": 3096
+    },
+    {
+      "epoch": 0.9943811205650988,
+      "grad_norm": 1.160662293434143,
+      "learning_rate": 0.00017440783833509366,
+      "loss": 0.5346,
+      "step": 3097
+    },
+    {
+      "epoch": 0.9947021993899502,
+      "grad_norm": 0.992720365524292,
+      "learning_rate": 0.00017438451298070252,
+      "loss": 0.6864,
+      "step": 3098
+    },
+    {
+      "epoch": 0.9950232782148017,
+      "grad_norm": 0.705855667591095,
+      "learning_rate": 0.0001743611785627841,
+      "loss": 0.6184,
+      "step": 3099
+    },
+    {
+      "epoch": 0.9953443570396532,
+      "grad_norm": 1.4650824069976807,
+      "learning_rate": 0.00017433783508418162,
+      "loss": 0.4631,
+      "step": 3100
+    },
+    {
+      "epoch": 0.9956654358645047,
+      "grad_norm": 1.113871693611145,
+      "learning_rate": 0.00017431448254773944,
+      "loss": 1.1905,
+      "step": 3101
+    },
+    {
+      "epoch": 0.9959865146893563,
+      "grad_norm": 1.3432302474975586,
+      "learning_rate": 0.00017429112095630295,
+      "loss": 0.6531,
+      "step": 3102
+    },
+    {
+      "epoch": 0.9963075935142077,
+      "grad_norm": 1.2041120529174805,
+      "learning_rate": 0.00017426775031271875,
+      "loss": 0.8234,
+      "step": 3103
+    },
+    {
+      "epoch": 0.9966286723390593,
+      "grad_norm": 0.980356752872467,
+      "learning_rate": 0.00017424437061983446,
+      "loss": 1.0048,
+      "step": 3104
+    },
+    {
+      "epoch": 0.9969497511639107,
+      "grad_norm": 0.9739211201667786,
+      "learning_rate": 0.00017422098188049883,
+      "loss": 0.648,
+      "step": 3105
+    },
+    {
+      "epoch": 0.9972708299887623,
+      "grad_norm": 1.0046653747558594,
+      "learning_rate": 0.00017419758409756164,
+      "loss": 0.8891,
+      "step": 3106
+    },
+    {
+      "epoch": 0.9975919088136137,
+      "grad_norm": 0.8817589282989502,
+      "learning_rate": 0.00017417417727387394,
+      "loss": 0.8698,
+      "step": 3107
+    },
+    {
+      "epoch": 0.9979129876384653,
+      "grad_norm": 0.8024824857711792,
+      "learning_rate": 0.0001741507614122877,
+      "loss": 0.7637,
+      "step": 3108
+    },
+    {
+      "epoch": 0.9982340664633167,
+      "grad_norm": 1.0565115213394165,
+      "learning_rate": 0.0001741273365156561,
+      "loss": 0.9549,
+      "step": 3109
+    },
+    {
+      "epoch": 0.9985551452881682,
+      "grad_norm": 1.4340211153030396,
+      "learning_rate": 0.00017410390258683345,
+      "loss": 0.9085,
+      "step": 3110
+    },
+    {
+      "epoch": 0.9988762241130198,
+      "grad_norm": 0.8469492197036743,
+      "learning_rate": 0.000174080459628675,
+      "loss": 0.7797,
+      "step": 3111
+    },
+    {
+      "epoch": 0.9991973029378712,
+      "grad_norm": 2.2130372524261475,
+      "learning_rate": 0.00017405700764403726,
+      "loss": 0.7562,
+      "step": 3112
+    },
+    {
+      "epoch": 0.9995183817627228,
+      "grad_norm": 0.984209418296814,
+      "learning_rate": 0.00017403354663577783,
+      "loss": 0.9092,
+      "step": 3113
+    },
+    {
+      "epoch": 0.9998394605875742,
+      "grad_norm": 1.7540405988693237,
+      "learning_rate": 0.00017401007660675525,
+      "loss": 0.6898,
+      "step": 3114
+    },
+    {
+      "epoch": 1.0001605394124258,
+      "grad_norm": 0.4434870779514313,
+      "learning_rate": 0.00017398659755982936,
+      "loss": 0.5678,
+      "step": 3115
+    },
+    {
+      "epoch": 1.0004816182372773,
+      "grad_norm": 0.6553979516029358,
+      "learning_rate": 0.000173963109497861,
+      "loss": 1.0352,
+      "step": 3116
+    },
+    {
+      "epoch": 1.0008026970621287,
+      "grad_norm": 0.8394825458526611,
+      "learning_rate": 0.00017393961242371205,
+      "loss": 0.4912,
+      "step": 3117
+    },
+    {
+      "epoch": 1.0011237758869802,
+      "grad_norm": 1.0343304872512817,
+      "learning_rate": 0.00017391610634024564,
+      "loss": 0.6326,
+      "step": 3118
+    },
+    {
+      "epoch": 1.0014448547118318,
+      "grad_norm": 0.8101858496665955,
+      "learning_rate": 0.0001738925912503259,
+      "loss": 0.5709,
+      "step": 3119
+    },
+    {
+      "epoch": 1.0017659335366833,
+      "grad_norm": 0.8755748271942139,
+      "learning_rate": 0.000173869067156818,
+      "loss": 0.377,
+      "step": 3120
+    },
+    {
+      "epoch": 1.0020870123615349,
+      "grad_norm": 0.9467462301254272,
+      "learning_rate": 0.00017384553406258842,
+      "loss": 0.6883,
+      "step": 3121
+    },
+    {
+      "epoch": 1.0024080911863862,
+      "grad_norm": 0.8870365619659424,
+      "learning_rate": 0.0001738219919705044,
+      "loss": 0.7174,
+      "step": 3122
+    },
+    {
+      "epoch": 1.0027291700112377,
+      "grad_norm": 1.0948724746704102,
+      "learning_rate": 0.00017379844088343468,
+      "loss": 0.9455,
+      "step": 3123
+    },
+    {
+      "epoch": 1.0030502488360893,
+      "grad_norm": 0.8313069939613342,
+      "learning_rate": 0.00017377488080424876,
+      "loss": 0.6838,
+      "step": 3124
+    },
+    {
+      "epoch": 1.0033713276609408,
+      "grad_norm": 0.9020499587059021,
+      "learning_rate": 0.0001737513117358174,
+      "loss": 0.7241,
+      "step": 3125
+    },
+    {
+      "epoch": 1.0036924064857922,
+      "grad_norm": 1.03434419631958,
+      "learning_rate": 0.0001737277336810124,
+      "loss": 0.5053,
+      "step": 3126
+    },
+    {
+      "epoch": 1.0040134853106437,
+      "grad_norm": 0.6919084191322327,
+      "learning_rate": 0.00017370414664270674,
+      "loss": 0.5537,
+      "step": 3127
+    },
+    {
+      "epoch": 1.0043345641354953,
+      "grad_norm": 0.7697117924690247,
+      "learning_rate": 0.00017368055062377434,
+      "loss": 0.5852,
+      "step": 3128
+    },
+    {
+      "epoch": 1.0046556429603468,
+      "grad_norm": 0.8752188086509705,
+      "learning_rate": 0.00017365694562709034,
+      "loss": 0.6626,
+      "step": 3129
+    },
+    {
+      "epoch": 1.0049767217851984,
+      "grad_norm": 1.3550692796707153,
+      "learning_rate": 0.00017363333165553092,
+      "loss": 0.5762,
+      "step": 3130
+    },
+    {
+      "epoch": 1.0052978006100497,
+      "grad_norm": 1.0785512924194336,
+      "learning_rate": 0.00017360970871197346,
+      "loss": 0.6023,
+      "step": 3131
+    },
+    {
+      "epoch": 1.0056188794349012,
+      "grad_norm": 0.6877568960189819,
+      "learning_rate": 0.0001735860767992962,
+      "loss": 0.5769,
+      "step": 3132
+    },
+    {
+      "epoch": 1.0059399582597528,
+      "grad_norm": 1.3127740621566772,
+      "learning_rate": 0.00017356243592037875,
+      "loss": 0.5348,
+      "step": 3133
+    },
+    {
+      "epoch": 1.0062610370846043,
+      "grad_norm": 1.041380763053894,
+      "learning_rate": 0.0001735387860781016,
+      "loss": 0.8355,
+      "step": 3134
+    },
+    {
+      "epoch": 1.0065821159094557,
+      "grad_norm": 1.9624297618865967,
+      "learning_rate": 0.00017351512727534644,
+      "loss": 0.7688,
+      "step": 3135
+    },
+    {
+      "epoch": 1.0069031947343072,
+      "grad_norm": 1.034769892692566,
+      "learning_rate": 0.000173491459514996,
+      "loss": 0.6289,
+      "step": 3136
+    },
+    {
+      "epoch": 1.0072242735591588,
+      "grad_norm": 0.9164260625839233,
+      "learning_rate": 0.00017346778279993415,
+      "loss": 0.6022,
+      "step": 3137
+    },
+    {
+      "epoch": 1.0075453523840103,
+      "grad_norm": 0.9046801328659058,
+      "learning_rate": 0.0001734440971330458,
+      "loss": 0.7448,
+      "step": 3138
+    },
+    {
+      "epoch": 1.0078664312088619,
+      "grad_norm": 1.0306909084320068,
+      "learning_rate": 0.00017342040251721702,
+      "loss": 0.738,
+      "step": 3139
+    },
+    {
+      "epoch": 1.0081875100337132,
+      "grad_norm": 1.137430191040039,
+      "learning_rate": 0.0001733966989553349,
+      "loss": 0.7579,
+      "step": 3140
+    },
+    {
+      "epoch": 1.0085085888585648,
+      "grad_norm": 0.7871647477149963,
+      "learning_rate": 0.00017337298645028764,
+      "loss": 0.5093,
+      "step": 3141
+    },
+    {
+      "epoch": 1.0088296676834163,
+      "grad_norm": 0.9174445867538452,
+      "learning_rate": 0.00017334926500496456,
+      "loss": 0.6771,
+      "step": 3142
+    },
+    {
+      "epoch": 1.0091507465082679,
+      "grad_norm": 0.8707671761512756,
+      "learning_rate": 0.00017332553462225602,
+      "loss": 0.698,
+      "step": 3143
+    },
+    {
+      "epoch": 1.0094718253331192,
+      "grad_norm": 1.3214155435562134,
+      "learning_rate": 0.00017330179530505348,
+      "loss": 0.7583,
+      "step": 3144
+    },
+    {
+      "epoch": 1.0097929041579707,
+      "grad_norm": 0.8207698464393616,
+      "learning_rate": 0.00017327804705624957,
+      "loss": 0.6534,
+      "step": 3145
+    },
+    {
+      "epoch": 1.0101139829828223,
+      "grad_norm": 0.9219654202461243,
+      "learning_rate": 0.0001732542898787379,
+      "loss": 0.5564,
+      "step": 3146
+    },
+    {
+      "epoch": 1.0104350618076738,
+      "grad_norm": 0.9626970887184143,
+      "learning_rate": 0.00017323052377541317,
+      "loss": 0.758,
+      "step": 3147
+    },
+    {
+      "epoch": 1.0107561406325254,
+      "grad_norm": 0.8818027377128601,
+      "learning_rate": 0.0001732067487491713,
+      "loss": 0.7285,
+      "step": 3148
+    },
+    {
+      "epoch": 1.0110772194573767,
+      "grad_norm": 1.1400665044784546,
+      "learning_rate": 0.0001731829648029091,
+      "loss": 0.7988,
+      "step": 3149
+    },
+    {
+      "epoch": 1.0113982982822283,
+      "grad_norm": 1.296278476715088,
+      "learning_rate": 0.0001731591719395247,
+      "loss": 0.5509,
+      "step": 3150
+    },
+    {
+      "epoch": 1.0117193771070798,
+      "grad_norm": 0.886254608631134,
+      "learning_rate": 0.00017313537016191706,
+      "loss": 0.5616,
+      "step": 3151
+    },
+    {
+      "epoch": 1.0120404559319314,
+      "grad_norm": 0.9640228748321533,
+      "learning_rate": 0.00017311155947298643,
+      "loss": 0.6448,
+      "step": 3152
+    },
+    {
+      "epoch": 1.0123615347567827,
+      "grad_norm": 1.8710490465164185,
+      "learning_rate": 0.00017308773987563406,
+      "loss": 0.5566,
+      "step": 3153
+    },
+    {
+      "epoch": 1.0126826135816343,
+      "grad_norm": 1.0365729331970215,
+      "learning_rate": 0.00017306391137276224,
+      "loss": 0.7364,
+      "step": 3154
+    },
+    {
+      "epoch": 1.0130036924064858,
+      "grad_norm": 1.0165393352508545,
+      "learning_rate": 0.00017304007396727448,
+      "loss": 0.6858,
+      "step": 3155
+    },
+    {
+      "epoch": 1.0133247712313374,
+      "grad_norm": 1.3366596698760986,
+      "learning_rate": 0.00017301622766207527,
+      "loss": 0.6922,
+      "step": 3156
+    },
+    {
+      "epoch": 1.013645850056189,
+      "grad_norm": 1.253536581993103,
+      "learning_rate": 0.00017299237246007015,
+      "loss": 0.7062,
+      "step": 3157
+    },
+    {
+      "epoch": 1.0139669288810402,
+      "grad_norm": 0.9793636202812195,
+      "learning_rate": 0.00017296850836416588,
+      "loss": 0.6105,
+      "step": 3158
+    },
+    {
+      "epoch": 1.0142880077058918,
+      "grad_norm": 1.0851402282714844,
+      "learning_rate": 0.00017294463537727024,
+      "loss": 0.6383,
+      "step": 3159
+    },
+    {
+      "epoch": 1.0146090865307433,
+      "grad_norm": 0.8400055766105652,
+      "learning_rate": 0.000172920753502292,
+      "loss": 0.6843,
+      "step": 3160
+    },
+    {
+      "epoch": 1.0149301653555949,
+      "grad_norm": 1.1264251470565796,
+      "learning_rate": 0.00017289686274214118,
+      "loss": 0.5577,
+      "step": 3161
+    },
+    {
+      "epoch": 1.0152512441804462,
+      "grad_norm": 1.5338242053985596,
+      "learning_rate": 0.0001728729630997287,
+      "loss": 0.4795,
+      "step": 3162
+    },
+    {
+      "epoch": 1.0155723230052978,
+      "grad_norm": 0.7459766864776611,
+      "learning_rate": 0.00017284905457796675,
+      "loss": 0.4932,
+      "step": 3163
+    },
+    {
+      "epoch": 1.0158934018301493,
+      "grad_norm": 0.6073582172393799,
+      "learning_rate": 0.00017282513717976848,
+      "loss": 0.3484,
+      "step": 3164
+    },
+    {
+      "epoch": 1.0162144806550009,
+      "grad_norm": 0.8098936080932617,
+      "learning_rate": 0.00017280121090804812,
+      "loss": 0.4744,
+      "step": 3165
+    },
+    {
+      "epoch": 1.0165355594798524,
+      "grad_norm": 1.1062911748886108,
+      "learning_rate": 0.00017277727576572107,
+      "loss": 1.023,
+      "step": 3166
+    },
+    {
+      "epoch": 1.0168566383047037,
+      "grad_norm": 1.3004356622695923,
+      "learning_rate": 0.00017275333175570368,
+      "loss": 0.7562,
+      "step": 3167
+    },
+    {
+      "epoch": 1.0171777171295553,
+      "grad_norm": 1.5146620273590088,
+      "learning_rate": 0.00017272937888091353,
+      "loss": 0.6177,
+      "step": 3168
+    },
+    {
+      "epoch": 1.0174987959544068,
+      "grad_norm": 1.1251643896102905,
+      "learning_rate": 0.0001727054171442692,
+      "loss": 0.4857,
+      "step": 3169
+    },
+    {
+      "epoch": 1.0178198747792584,
+      "grad_norm": 0.8808386921882629,
+      "learning_rate": 0.0001726814465486903,
+      "loss": 0.2652,
+      "step": 3170
+    },
+    {
+      "epoch": 1.0181409536041097,
+      "grad_norm": 1.0438605546951294,
+      "learning_rate": 0.0001726574670970976,
+      "loss": 0.5482,
+      "step": 3171
+    },
+    {
+      "epoch": 1.0184620324289613,
+      "grad_norm": 3.0426223278045654,
+      "learning_rate": 0.00017263347879241291,
+      "loss": 0.7843,
+      "step": 3172
+    },
+    {
+      "epoch": 1.0187831112538128,
+      "grad_norm": 0.8068621754646301,
+      "learning_rate": 0.00017260948163755918,
+      "loss": 0.7201,
+      "step": 3173
+    },
+    {
+      "epoch": 1.0191041900786644,
+      "grad_norm": 0.9241558313369751,
+      "learning_rate": 0.00017258547563546038,
+      "loss": 0.719,
+      "step": 3174
+    },
+    {
+      "epoch": 1.019425268903516,
+      "grad_norm": 0.9264689087867737,
+      "learning_rate": 0.00017256146078904153,
+      "loss": 0.8149,
+      "step": 3175
+    },
+    {
+      "epoch": 1.0197463477283673,
+      "grad_norm": 0.7384179830551147,
+      "learning_rate": 0.00017253743710122875,
+      "loss": 0.5689,
+      "step": 3176
+    },
+    {
+      "epoch": 1.0200674265532188,
+      "grad_norm": 1.2575445175170898,
+      "learning_rate": 0.00017251340457494934,
+      "loss": 0.6949,
+      "step": 3177
+    },
+    {
+      "epoch": 1.0203885053780704,
+      "grad_norm": 0.8369306921958923,
+      "learning_rate": 0.0001724893632131315,
+      "loss": 0.6599,
+      "step": 3178
+    },
+    {
+      "epoch": 1.020709584202922,
+      "grad_norm": 1.2006936073303223,
+      "learning_rate": 0.0001724653130187047,
+      "loss": 0.6771,
+      "step": 3179
+    },
+    {
+      "epoch": 1.0210306630277732,
+      "grad_norm": 0.7915051579475403,
+      "learning_rate": 0.00017244125399459926,
+      "loss": 0.7149,
+      "step": 3180
+    },
+    {
+      "epoch": 1.0213517418526248,
+      "grad_norm": 1.273621916770935,
+      "learning_rate": 0.00017241718614374678,
+      "loss": 0.7439,
+      "step": 3181
+    },
+    {
+      "epoch": 1.0216728206774763,
+      "grad_norm": 0.8545572757720947,
+      "learning_rate": 0.0001723931094690798,
+      "loss": 0.72,
+      "step": 3182
+    },
+    {
+      "epoch": 1.0219938995023279,
+      "grad_norm": 0.8133922815322876,
+      "learning_rate": 0.00017236902397353205,
+      "loss": 0.5812,
+      "step": 3183
+    },
+    {
+      "epoch": 1.0223149783271792,
+      "grad_norm": 1.2312626838684082,
+      "learning_rate": 0.00017234492966003824,
+      "loss": 0.6452,
+      "step": 3184
+    },
+    {
+      "epoch": 1.0226360571520308,
+      "grad_norm": 1.0506107807159424,
+      "learning_rate": 0.00017232082653153422,
+      "loss": 0.697,
+      "step": 3185
+    },
+    {
+      "epoch": 1.0229571359768823,
+      "grad_norm": 0.9486799240112305,
+      "learning_rate": 0.00017229671459095683,
+      "loss": 0.8227,
+      "step": 3186
+    },
+    {
+      "epoch": 1.0232782148017339,
+      "grad_norm": 0.8300420641899109,
+      "learning_rate": 0.0001722725938412441,
+      "loss": 0.7616,
+      "step": 3187
+    },
+    {
+      "epoch": 1.0235992936265854,
+      "grad_norm": 1.1895195245742798,
+      "learning_rate": 0.00017224846428533499,
+      "loss": 0.8887,
+      "step": 3188
+    },
+    {
+      "epoch": 1.0239203724514367,
+      "grad_norm": 0.9756418466567993,
+      "learning_rate": 0.0001722243259261697,
+      "loss": 0.5706,
+      "step": 3189
+    },
+    {
+      "epoch": 1.0242414512762883,
+      "grad_norm": 0.9104772210121155,
+      "learning_rate": 0.00017220017876668934,
+      "loss": 0.6564,
+      "step": 3190
+    },
+    {
+      "epoch": 1.0245625301011398,
+      "grad_norm": 0.8529773950576782,
+      "learning_rate": 0.00017217602280983623,
+      "loss": 0.664,
+      "step": 3191
+    },
+    {
+      "epoch": 1.0248836089259914,
+      "grad_norm": 1.0890889167785645,
+      "learning_rate": 0.00017215185805855368,
+      "loss": 0.5597,
+      "step": 3192
+    },
+    {
+      "epoch": 1.0252046877508427,
+      "grad_norm": 0.9177720546722412,
+      "learning_rate": 0.00017212768451578607,
+      "loss": 0.8042,
+      "step": 3193
+    },
+    {
+      "epoch": 1.0255257665756943,
+      "grad_norm": 0.8690738677978516,
+      "learning_rate": 0.00017210350218447887,
+      "loss": 0.6965,
+      "step": 3194
+    },
+    {
+      "epoch": 1.0258468454005458,
+      "grad_norm": 0.9098244309425354,
+      "learning_rate": 0.00017207931106757868,
+      "loss": 0.5547,
+      "step": 3195
+    },
+    {
+      "epoch": 1.0261679242253974,
+      "grad_norm": 0.8864629864692688,
+      "learning_rate": 0.00017205511116803306,
+      "loss": 0.6666,
+      "step": 3196
+    },
+    {
+      "epoch": 1.026489003050249,
+      "grad_norm": 0.9058293700218201,
+      "learning_rate": 0.0001720309024887907,
+      "loss": 0.6675,
+      "step": 3197
+    },
+    {
+      "epoch": 1.0268100818751003,
+      "grad_norm": 1.372377634048462,
+      "learning_rate": 0.00017200668503280136,
+      "loss": 0.7582,
+      "step": 3198
+    },
+    {
+      "epoch": 1.0271311606999518,
+      "grad_norm": 0.7774912118911743,
+      "learning_rate": 0.0001719824588030159,
+      "loss": 0.6099,
+      "step": 3199
+    },
+    {
+      "epoch": 1.0274522395248034,
+      "grad_norm": 1.0164647102355957,
+      "learning_rate": 0.00017195822380238615,
+      "loss": 0.7993,
+      "step": 3200
+    },
+    {
+      "epoch": 1.027773318349655,
+      "grad_norm": 0.9077091813087463,
+      "learning_rate": 0.0001719339800338651,
+      "loss": 0.7304,
+      "step": 3201
+    },
+    {
+      "epoch": 1.0280943971745062,
+      "grad_norm": 1.8159345388412476,
+      "learning_rate": 0.00017190972750040682,
+      "loss": 0.6752,
+      "step": 3202
+    },
+    {
+      "epoch": 1.0284154759993578,
+      "grad_norm": 1.1675589084625244,
+      "learning_rate": 0.00017188546620496635,
+      "loss": 0.8066,
+      "step": 3203
+    },
+    {
+      "epoch": 1.0287365548242093,
+      "grad_norm": 1.103464126586914,
+      "learning_rate": 0.00017186119615049988,
+      "loss": 0.6484,
+      "step": 3204
+    },
+    {
+      "epoch": 1.0290576336490609,
+      "grad_norm": 1.758217692375183,
+      "learning_rate": 0.00017183691733996462,
+      "loss": 0.6094,
+      "step": 3205
+    },
+    {
+      "epoch": 1.0293787124739124,
+      "grad_norm": 1.3336718082427979,
+      "learning_rate": 0.00017181262977631888,
+      "loss": 0.6994,
+      "step": 3206
+    },
+    {
+      "epoch": 1.0296997912987638,
+      "grad_norm": 1.3006914854049683,
+      "learning_rate": 0.00017178833346252206,
+      "loss": 0.5225,
+      "step": 3207
+    },
+    {
+      "epoch": 1.0300208701236153,
+      "grad_norm": 1.329729437828064,
+      "learning_rate": 0.00017176402840153455,
+      "loss": 0.6402,
+      "step": 3208
+    },
+    {
+      "epoch": 1.0303419489484669,
+      "grad_norm": 1.2530442476272583,
+      "learning_rate": 0.00017173971459631787,
+      "loss": 0.6543,
+      "step": 3209
+    },
+    {
+      "epoch": 1.0306630277733184,
+      "grad_norm": 1.1140762567520142,
+      "learning_rate": 0.0001717153920498346,
+      "loss": 0.5683,
+      "step": 3210
+    },
+    {
+      "epoch": 1.0309841065981697,
+      "grad_norm": 1.3410987854003906,
+      "learning_rate": 0.0001716910607650483,
+      "loss": 0.5983,
+      "step": 3211
+    },
+    {
+      "epoch": 1.0313051854230213,
+      "grad_norm": 1.056822419166565,
+      "learning_rate": 0.0001716667207449237,
+      "loss": 0.5871,
+      "step": 3212
+    },
+    {
+      "epoch": 1.0316262642478728,
+      "grad_norm": 1.11257004737854,
+      "learning_rate": 0.0001716423719924266,
+      "loss": 0.4527,
+      "step": 3213
+    },
+    {
+      "epoch": 1.0319473430727244,
+      "grad_norm": 1.4045213460922241,
+      "learning_rate": 0.00017161801451052376,
+      "loss": 0.7051,
+      "step": 3214
+    },
+    {
+      "epoch": 1.032268421897576,
+      "grad_norm": 0.6990142464637756,
+      "learning_rate": 0.00017159364830218312,
+      "loss": 0.69,
+      "step": 3215
+    },
+    {
+      "epoch": 1.0325895007224273,
+      "grad_norm": 0.6937003135681152,
+      "learning_rate": 0.0001715692733703736,
+      "loss": 0.9796,
+      "step": 3216
+    },
+    {
+      "epoch": 1.0329105795472788,
+      "grad_norm": 0.7836512923240662,
+      "learning_rate": 0.00017154488971806518,
+      "loss": 0.5325,
+      "step": 3217
+    },
+    {
+      "epoch": 1.0332316583721304,
+      "grad_norm": 0.7977432012557983,
+      "learning_rate": 0.00017152049734822902,
+      "loss": 0.5356,
+      "step": 3218
+    },
+    {
+      "epoch": 1.033552737196982,
+      "grad_norm": 0.8344630599021912,
+      "learning_rate": 0.00017149609626383717,
+      "loss": 0.529,
+      "step": 3219
+    },
+    {
+      "epoch": 1.0338738160218333,
+      "grad_norm": 0.9146289229393005,
+      "learning_rate": 0.00017147168646786286,
+      "loss": 0.4378,
+      "step": 3220
+    },
+    {
+      "epoch": 1.0341948948466848,
+      "grad_norm": 0.8820130825042725,
+      "learning_rate": 0.00017144726796328034,
+      "loss": 0.4912,
+      "step": 3221
+    },
+    {
+      "epoch": 1.0345159736715364,
+      "grad_norm": 0.8433278203010559,
+      "learning_rate": 0.00017142284075306497,
+      "loss": 0.5559,
+      "step": 3222
+    },
+    {
+      "epoch": 1.034837052496388,
+      "grad_norm": 1.3478479385375977,
+      "learning_rate": 0.0001713984048401931,
+      "loss": 0.8736,
+      "step": 3223
+    },
+    {
+      "epoch": 1.0351581313212395,
+      "grad_norm": 1.062957763671875,
+      "learning_rate": 0.00017137396022764214,
+      "loss": 0.7858,
+      "step": 3224
+    },
+    {
+      "epoch": 1.0354792101460908,
+      "grad_norm": 1.3038783073425293,
+      "learning_rate": 0.00017134950691839065,
+      "loss": 0.8089,
+      "step": 3225
+    },
+    {
+      "epoch": 1.0358002889709423,
+      "grad_norm": 1.0557838678359985,
+      "learning_rate": 0.00017132504491541818,
+      "loss": 0.5426,
+      "step": 3226
+    },
+    {
+      "epoch": 1.036121367795794,
+      "grad_norm": 0.7496936321258545,
+      "learning_rate": 0.0001713005742217053,
+      "loss": 0.5222,
+      "step": 3227
+    },
+    {
+      "epoch": 1.0364424466206454,
+      "grad_norm": 0.9153050184249878,
+      "learning_rate": 0.00017127609484023377,
+      "loss": 0.6843,
+      "step": 3228
+    },
+    {
+      "epoch": 1.0367635254454968,
+      "grad_norm": 1.0914545059204102,
+      "learning_rate": 0.00017125160677398626,
+      "loss": 0.5231,
+      "step": 3229
+    },
+    {
+      "epoch": 1.0370846042703483,
+      "grad_norm": 0.7990017533302307,
+      "learning_rate": 0.0001712271100259466,
+      "loss": 0.6105,
+      "step": 3230
+    },
+    {
+      "epoch": 1.0374056830951999,
+      "grad_norm": 0.8752117156982422,
+      "learning_rate": 0.00017120260459909967,
+      "loss": 0.6104,
+      "step": 3231
+    },
+    {
+      "epoch": 1.0377267619200514,
+      "grad_norm": 1.1408075094223022,
+      "learning_rate": 0.0001711780904964313,
+      "loss": 0.8112,
+      "step": 3232
+    },
+    {
+      "epoch": 1.038047840744903,
+      "grad_norm": 0.7575783133506775,
+      "learning_rate": 0.00017115356772092857,
+      "loss": 0.4848,
+      "step": 3233
+    },
+    {
+      "epoch": 1.0383689195697543,
+      "grad_norm": 1.2850323915481567,
+      "learning_rate": 0.0001711290362755794,
+      "loss": 0.7803,
+      "step": 3234
+    },
+    {
+      "epoch": 1.0386899983946059,
+      "grad_norm": 0.9859423637390137,
+      "learning_rate": 0.00017110449616337289,
+      "loss": 0.6785,
+      "step": 3235
+    },
+    {
+      "epoch": 1.0390110772194574,
+      "grad_norm": 1.224280834197998,
+      "learning_rate": 0.00017107994738729926,
+      "loss": 0.806,
+      "step": 3236
+    },
+    {
+      "epoch": 1.039332156044309,
+      "grad_norm": 1.1364513635635376,
+      "learning_rate": 0.00017105538995034963,
+      "loss": 0.8497,
+      "step": 3237
+    },
+    {
+      "epoch": 1.0396532348691603,
+      "grad_norm": 0.712236225605011,
+      "learning_rate": 0.00017103082385551627,
+      "loss": 0.4612,
+      "step": 3238
+    },
+    {
+      "epoch": 1.0399743136940118,
+      "grad_norm": 0.8277900218963623,
+      "learning_rate": 0.0001710062491057925,
+      "loss": 0.6745,
+      "step": 3239
+    },
+    {
+      "epoch": 1.0402953925188634,
+      "grad_norm": 1.1629518270492554,
+      "learning_rate": 0.00017098166570417262,
+      "loss": 0.7966,
+      "step": 3240
+    },
+    {
+      "epoch": 1.040616471343715,
+      "grad_norm": 1.143298864364624,
+      "learning_rate": 0.0001709570736536521,
+      "loss": 0.5971,
+      "step": 3241
+    },
+    {
+      "epoch": 1.0409375501685665,
+      "grad_norm": 1.7243444919586182,
+      "learning_rate": 0.0001709324729572274,
+      "loss": 0.521,
+      "step": 3242
+    },
+    {
+      "epoch": 1.0412586289934178,
+      "grad_norm": 0.7155652046203613,
+      "learning_rate": 0.000170907863617896,
+      "loss": 0.5353,
+      "step": 3243
+    },
+    {
+      "epoch": 1.0415797078182694,
+      "grad_norm": 0.8861052393913269,
+      "learning_rate": 0.00017088324563865656,
+      "loss": 0.7525,
+      "step": 3244
+    },
+    {
+      "epoch": 1.041900786643121,
+      "grad_norm": 0.786766529083252,
+      "learning_rate": 0.00017085861902250865,
+      "loss": 0.6192,
+      "step": 3245
+    },
+    {
+      "epoch": 1.0422218654679725,
+      "grad_norm": 1.0128397941589355,
+      "learning_rate": 0.0001708339837724529,
+      "loss": 0.6272,
+      "step": 3246
+    },
+    {
+      "epoch": 1.0425429442928238,
+      "grad_norm": 0.9892904162406921,
+      "learning_rate": 0.0001708093398914911,
+      "loss": 0.5604,
+      "step": 3247
+    },
+    {
+      "epoch": 1.0428640231176753,
+      "grad_norm": 1.106776237487793,
+      "learning_rate": 0.00017078468738262602,
+      "loss": 0.7397,
+      "step": 3248
+    },
+    {
+      "epoch": 1.043185101942527,
+      "grad_norm": 1.0783487558364868,
+      "learning_rate": 0.00017076002624886154,
+      "loss": 0.6305,
+      "step": 3249
+    },
+    {
+      "epoch": 1.0435061807673784,
+      "grad_norm": 1.2297683954238892,
+      "learning_rate": 0.00017073535649320248,
+      "loss": 0.6252,
+      "step": 3250
+    },
+    {
+      "epoch": 1.04382725959223,
+      "grad_norm": 0.9655210375785828,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 0.6704,
+      "step": 3251
+    },
+    {
+      "epoch": 1.0441483384170813,
+      "grad_norm": 1.2158546447753906,
+      "learning_rate": 0.00017068599112822543,
+      "loss": 0.6869,
+      "step": 3252
+    },
+    {
+      "epoch": 1.0444694172419329,
+      "grad_norm": 1.0447654724121094,
+      "learning_rate": 0.0001706612955249225,
+      "loss": 0.4427,
+      "step": 3253
+    },
+    {
+      "epoch": 1.0447904960667844,
+      "grad_norm": 0.8590002059936523,
+      "learning_rate": 0.000170636591311755,
+      "loss": 0.5222,
+      "step": 3254
+    },
+    {
+      "epoch": 1.045111574891636,
+      "grad_norm": 1.114013433456421,
+      "learning_rate": 0.00017061187849173317,
+      "loss": 0.5776,
+      "step": 3255
+    },
+    {
+      "epoch": 1.0454326537164873,
+      "grad_norm": 0.8874313235282898,
+      "learning_rate": 0.0001705871570678681,
+      "loss": 0.6153,
+      "step": 3256
+    },
+    {
+      "epoch": 1.0457537325413389,
+      "grad_norm": 1.2078914642333984,
+      "learning_rate": 0.0001705624270431721,
+      "loss": 0.7155,
+      "step": 3257
+    },
+    {
+      "epoch": 1.0460748113661904,
+      "grad_norm": 0.9867888689041138,
+      "learning_rate": 0.00017053768842065833,
+      "loss": 0.5113,
+      "step": 3258
+    },
+    {
+      "epoch": 1.046395890191042,
+      "grad_norm": 0.6994266510009766,
+      "learning_rate": 0.00017051294120334125,
+      "loss": 0.5073,
+      "step": 3259
+    },
+    {
+      "epoch": 1.0467169690158935,
+      "grad_norm": 1.1606996059417725,
+      "learning_rate": 0.00017048818539423615,
+      "loss": 0.5518,
+      "step": 3260
+    },
+    {
+      "epoch": 1.0470380478407448,
+      "grad_norm": 0.9645611643791199,
+      "learning_rate": 0.00017046342099635948,
+      "loss": 0.5303,
+      "step": 3261
+    },
+    {
+      "epoch": 1.0473591266655964,
+      "grad_norm": 0.6704840660095215,
+      "learning_rate": 0.00017043864801272868,
+      "loss": 0.4778,
+      "step": 3262
+    },
+    {
+      "epoch": 1.047680205490448,
+      "grad_norm": 0.7522701025009155,
+      "learning_rate": 0.0001704138664463623,
+      "loss": 0.5107,
+      "step": 3263
+    },
+    {
+      "epoch": 1.0480012843152995,
+      "grad_norm": 1.1835343837738037,
+      "learning_rate": 0.00017038907630027988,
+      "loss": 0.516,
+      "step": 3264
+    },
+    {
+      "epoch": 1.0483223631401508,
+      "grad_norm": 0.9107168316841125,
+      "learning_rate": 0.00017036427757750205,
+      "loss": 0.7626,
+      "step": 3265
+    },
+    {
+      "epoch": 1.0486434419650024,
+      "grad_norm": 0.9731441736221313,
+      "learning_rate": 0.00017033947028105039,
+      "loss": 0.848,
+      "step": 3266
+    },
+    {
+      "epoch": 1.048964520789854,
+      "grad_norm": 1.073034644126892,
+      "learning_rate": 0.00017031465441394767,
+      "loss": 0.875,
+      "step": 3267
+    },
+    {
+      "epoch": 1.0492855996147055,
+      "grad_norm": 1.114713430404663,
+      "learning_rate": 0.00017028982997921758,
+      "loss": 0.5253,
+      "step": 3268
+    },
+    {
+      "epoch": 1.049606678439557,
+      "grad_norm": 1.0419694185256958,
+      "learning_rate": 0.00017026499697988493,
+      "loss": 0.6738,
+      "step": 3269
+    },
+    {
+      "epoch": 1.0499277572644083,
+      "grad_norm": 1.0044914484024048,
+      "learning_rate": 0.0001702401554189755,
+      "loss": 0.4802,
+      "step": 3270
+    },
+    {
+      "epoch": 1.05024883608926,
+      "grad_norm": 0.8162968158721924,
+      "learning_rate": 0.00017021530529951625,
+      "loss": 0.3764,
+      "step": 3271
+    },
+    {
+      "epoch": 1.0505699149141114,
+      "grad_norm": 0.8983839154243469,
+      "learning_rate": 0.000170190446624535,
+      "loss": 0.5214,
+      "step": 3272
+    },
+    {
+      "epoch": 1.050890993738963,
+      "grad_norm": 1.0330827236175537,
+      "learning_rate": 0.00017016557939706075,
+      "loss": 0.7508,
+      "step": 3273
+    },
+    {
+      "epoch": 1.0512120725638143,
+      "grad_norm": 0.8711283802986145,
+      "learning_rate": 0.00017014070362012348,
+      "loss": 0.6736,
+      "step": 3274
+    },
+    {
+      "epoch": 1.0515331513886659,
+      "grad_norm": 0.8156806230545044,
+      "learning_rate": 0.00017011581929675425,
+      "loss": 0.633,
+      "step": 3275
+    },
+    {
+      "epoch": 1.0518542302135174,
+      "grad_norm": 0.9561333060264587,
+      "learning_rate": 0.0001700909264299851,
+      "loss": 0.7371,
+      "step": 3276
+    },
+    {
+      "epoch": 1.052175309038369,
+      "grad_norm": 0.7813916802406311,
+      "learning_rate": 0.00017006602502284918,
+      "loss": 0.5109,
+      "step": 3277
+    },
+    {
+      "epoch": 1.0524963878632203,
+      "grad_norm": 0.8601284027099609,
+      "learning_rate": 0.00017004111507838064,
+      "loss": 0.5188,
+      "step": 3278
+    },
+    {
+      "epoch": 1.0528174666880719,
+      "grad_norm": 1.1225234270095825,
+      "learning_rate": 0.00017001619659961467,
+      "loss": 0.6945,
+      "step": 3279
+    },
+    {
+      "epoch": 1.0531385455129234,
+      "grad_norm": 1.0475960969924927,
+      "learning_rate": 0.00016999126958958756,
+      "loss": 0.6678,
+      "step": 3280
+    },
+    {
+      "epoch": 1.053459624337775,
+      "grad_norm": 1.015073537826538,
+      "learning_rate": 0.00016996633405133655,
+      "loss": 0.8029,
+      "step": 3281
+    },
+    {
+      "epoch": 1.0537807031626265,
+      "grad_norm": 1.1307599544525146,
+      "learning_rate": 0.00016994138998789997,
+      "loss": 0.6215,
+      "step": 3282
+    },
+    {
+      "epoch": 1.0541017819874778,
+      "grad_norm": 1.1362276077270508,
+      "learning_rate": 0.00016991643740231716,
+      "loss": 0.7888,
+      "step": 3283
+    },
+    {
+      "epoch": 1.0544228608123294,
+      "grad_norm": 1.2660646438598633,
+      "learning_rate": 0.0001698914762976285,
+      "loss": 0.7849,
+      "step": 3284
+    },
+    {
+      "epoch": 1.054743939637181,
+      "grad_norm": 0.9536210298538208,
+      "learning_rate": 0.00016986650667687552,
+      "loss": 0.6164,
+      "step": 3285
+    },
+    {
+      "epoch": 1.0550650184620325,
+      "grad_norm": 0.9557814598083496,
+      "learning_rate": 0.0001698415285431006,
+      "loss": 0.6736,
+      "step": 3286
+    },
+    {
+      "epoch": 1.0553860972868838,
+      "grad_norm": 1.0140166282653809,
+      "learning_rate": 0.00016981654189934727,
+      "loss": 0.6405,
+      "step": 3287
+    },
+    {
+      "epoch": 1.0557071761117354,
+      "grad_norm": 0.9973203539848328,
+      "learning_rate": 0.0001697915467486601,
+      "loss": 0.7855,
+      "step": 3288
+    },
+    {
+      "epoch": 1.056028254936587,
+      "grad_norm": 1.143325924873352,
+      "learning_rate": 0.00016976654309408464,
+      "loss": 0.8278,
+      "step": 3289
+    },
+    {
+      "epoch": 1.0563493337614385,
+      "grad_norm": 1.1182677745819092,
+      "learning_rate": 0.00016974153093866757,
+      "loss": 0.6487,
+      "step": 3290
+    },
+    {
+      "epoch": 1.05667041258629,
+      "grad_norm": 0.8752659559249878,
+      "learning_rate": 0.00016971651028545648,
+      "loss": 0.7054,
+      "step": 3291
+    },
+    {
+      "epoch": 1.0569914914111413,
+      "grad_norm": 1.3598425388336182,
+      "learning_rate": 0.00016969148113750007,
+      "loss": 0.8206,
+      "step": 3292
+    },
+    {
+      "epoch": 1.057312570235993,
+      "grad_norm": 0.8707671165466309,
+      "learning_rate": 0.00016966644349784808,
+      "loss": 0.6105,
+      "step": 3293
+    },
+    {
+      "epoch": 1.0576336490608444,
+      "grad_norm": 0.9644749760627747,
+      "learning_rate": 0.0001696413973695513,
+      "loss": 0.5752,
+      "step": 3294
+    },
+    {
+      "epoch": 1.057954727885696,
+      "grad_norm": 0.894107460975647,
+      "learning_rate": 0.00016961634275566146,
+      "loss": 0.6371,
+      "step": 3295
+    },
+    {
+      "epoch": 1.0582758067105473,
+      "grad_norm": 1.123781442642212,
+      "learning_rate": 0.00016959127965923142,
+      "loss": 0.6209,
+      "step": 3296
+    },
+    {
+      "epoch": 1.0585968855353989,
+      "grad_norm": 1.2167694568634033,
+      "learning_rate": 0.0001695662080833151,
+      "loss": 0.576,
+      "step": 3297
+    },
+    {
+      "epoch": 1.0589179643602504,
+      "grad_norm": 1.0875262022018433,
+      "learning_rate": 0.00016954112803096728,
+      "loss": 0.6831,
+      "step": 3298
+    },
+    {
+      "epoch": 1.059239043185102,
+      "grad_norm": 0.9846530556678772,
+      "learning_rate": 0.000169516039505244,
+      "loss": 0.6525,
+      "step": 3299
+    },
+    {
+      "epoch": 1.0595601220099535,
+      "grad_norm": 0.9327017664909363,
+      "learning_rate": 0.00016949094250920217,
+      "loss": 0.5837,
+      "step": 3300
+    },
+    {
+      "epoch": 1.0598812008348049,
+      "grad_norm": 0.8693040013313293,
+      "learning_rate": 0.00016946583704589973,
+      "loss": 0.7224,
+      "step": 3301
+    },
+    {
+      "epoch": 1.0602022796596564,
+      "grad_norm": 1.8250339031219482,
+      "learning_rate": 0.00016944072311839581,
+      "loss": 0.6624,
+      "step": 3302
+    },
+    {
+      "epoch": 1.060523358484508,
+      "grad_norm": 1.6133804321289062,
+      "learning_rate": 0.0001694156007297504,
+      "loss": 0.6358,
+      "step": 3303
+    },
+    {
+      "epoch": 1.0608444373093595,
+      "grad_norm": 1.2496116161346436,
+      "learning_rate": 0.00016939046988302458,
+      "loss": 0.788,
+      "step": 3304
+    },
+    {
+      "epoch": 1.0611655161342108,
+      "grad_norm": 0.8245418071746826,
+      "learning_rate": 0.0001693653305812805,
+      "loss": 0.6231,
+      "step": 3305
+    },
+    {
+      "epoch": 1.0614865949590624,
+      "grad_norm": 0.9475066661834717,
+      "learning_rate": 0.0001693401828275813,
+      "loss": 0.6326,
+      "step": 3306
+    },
+    {
+      "epoch": 1.061807673783914,
+      "grad_norm": 0.8447438478469849,
+      "learning_rate": 0.00016931502662499118,
+      "loss": 0.4383,
+      "step": 3307
+    },
+    {
+      "epoch": 1.0621287526087655,
+      "grad_norm": 0.9479712247848511,
+      "learning_rate": 0.00016928986197657525,
+      "loss": 0.4397,
+      "step": 3308
+    },
+    {
+      "epoch": 1.062449831433617,
+      "grad_norm": 1.2100830078125,
+      "learning_rate": 0.0001692646888853999,
+      "loss": 0.6239,
+      "step": 3309
+    },
+    {
+      "epoch": 1.0627709102584684,
+      "grad_norm": 1.2680039405822754,
+      "learning_rate": 0.00016923950735453226,
+      "loss": 0.5405,
+      "step": 3310
+    },
+    {
+      "epoch": 1.06309198908332,
+      "grad_norm": 0.9997744560241699,
+      "learning_rate": 0.0001692143173870407,
+      "loss": 0.5493,
+      "step": 3311
+    },
+    {
+      "epoch": 1.0634130679081715,
+      "grad_norm": 0.9026280641555786,
+      "learning_rate": 0.0001691891189859945,
+      "loss": 0.5545,
+      "step": 3312
+    },
+    {
+      "epoch": 1.063734146733023,
+      "grad_norm": 1.112392544746399,
+      "learning_rate": 0.000169163912154464,
+      "loss": 0.6452,
+      "step": 3313
+    },
+    {
+      "epoch": 1.0640552255578744,
+      "grad_norm": 0.7569609880447388,
+      "learning_rate": 0.00016913869689552064,
+      "loss": 0.4284,
+      "step": 3314
+    },
+    {
+      "epoch": 1.064376304382726,
+      "grad_norm": 0.8292409181594849,
+      "learning_rate": 0.0001691134732122368,
+      "loss": 0.5371,
+      "step": 3315
+    },
+    {
+      "epoch": 1.0646973832075775,
+      "grad_norm": 0.8496055006980896,
+      "learning_rate": 0.00016908824110768584,
+      "loss": 1.1306,
+      "step": 3316
+    },
+    {
+      "epoch": 1.065018462032429,
+      "grad_norm": 0.9334166049957275,
+      "learning_rate": 0.00016906300058494228,
+      "loss": 0.619,
+      "step": 3317
+    },
+    {
+      "epoch": 1.0653395408572806,
+      "grad_norm": 0.9362530708312988,
+      "learning_rate": 0.00016903775164708163,
+      "loss": 0.6312,
+      "step": 3318
+    },
+    {
+      "epoch": 1.0656606196821319,
+      "grad_norm": 0.7941862344741821,
+      "learning_rate": 0.00016901249429718032,
+      "loss": 0.5325,
+      "step": 3319
+    },
+    {
+      "epoch": 1.0659816985069834,
+      "grad_norm": 0.8351782560348511,
+      "learning_rate": 0.00016898722853831593,
+      "loss": 0.4099,
+      "step": 3320
+    },
+    {
+      "epoch": 1.066302777331835,
+      "grad_norm": 2.7632505893707275,
+      "learning_rate": 0.000168961954373567,
+      "loss": 0.6414,
+      "step": 3321
+    },
+    {
+      "epoch": 1.0666238561566865,
+      "grad_norm": 1.1203904151916504,
+      "learning_rate": 0.00016893667180601312,
+      "loss": 0.7519,
+      "step": 3322
+    },
+    {
+      "epoch": 1.0669449349815379,
+      "grad_norm": 1.0219404697418213,
+      "learning_rate": 0.00016891138083873487,
+      "loss": 0.6174,
+      "step": 3323
+    },
+    {
+      "epoch": 1.0672660138063894,
+      "grad_norm": 0.9879265427589417,
+      "learning_rate": 0.00016888608147481388,
+      "loss": 0.636,
+      "step": 3324
+    },
+    {
+      "epoch": 1.067587092631241,
+      "grad_norm": 0.7659774422645569,
+      "learning_rate": 0.00016886077371733283,
+      "loss": 0.5238,
+      "step": 3325
+    },
+    {
+      "epoch": 1.0679081714560925,
+      "grad_norm": 0.9733467698097229,
+      "learning_rate": 0.0001688354575693754,
+      "loss": 0.7139,
+      "step": 3326
+    },
+    {
+      "epoch": 1.068229250280944,
+      "grad_norm": 0.7653968334197998,
+      "learning_rate": 0.0001688101330340263,
+      "loss": 0.4363,
+      "step": 3327
+    },
+    {
+      "epoch": 1.0685503291057954,
+      "grad_norm": 1.0636018514633179,
+      "learning_rate": 0.0001687848001143711,
+      "loss": 0.7243,
+      "step": 3328
+    },
+    {
+      "epoch": 1.068871407930647,
+      "grad_norm": 1.1847553253173828,
+      "learning_rate": 0.00016875945881349676,
+      "loss": 0.7352,
+      "step": 3329
+    },
+    {
+      "epoch": 1.0691924867554985,
+      "grad_norm": 1.2986525297164917,
+      "learning_rate": 0.00016873410913449091,
+      "loss": 0.7651,
+      "step": 3330
+    },
+    {
+      "epoch": 1.06951356558035,
+      "grad_norm": 0.921399712562561,
+      "learning_rate": 0.0001687087510804423,
+      "loss": 0.6639,
+      "step": 3331
+    },
+    {
+      "epoch": 1.0698346444052014,
+      "grad_norm": 0.900202751159668,
+      "learning_rate": 0.00016868338465444085,
+      "loss": 0.6414,
+      "step": 3332
+    },
+    {
+      "epoch": 1.070155723230053,
+      "grad_norm": 0.7372778654098511,
+      "learning_rate": 0.00016865800985957726,
+      "loss": 0.518,
+      "step": 3333
+    },
+    {
+      "epoch": 1.0704768020549045,
+      "grad_norm": 1.0439753532409668,
+      "learning_rate": 0.0001686326266989435,
+      "loss": 0.7912,
+      "step": 3334
+    },
+    {
+      "epoch": 1.070797880879756,
+      "grad_norm": 0.9654948711395264,
+      "learning_rate": 0.0001686072351756323,
+      "loss": 0.7365,
+      "step": 3335
+    },
+    {
+      "epoch": 1.0711189597046076,
+      "grad_norm": 0.9803206324577332,
+      "learning_rate": 0.00016858183529273765,
+      "loss": 0.5448,
+      "step": 3336
+    },
+    {
+      "epoch": 1.071440038529459,
+      "grad_norm": 0.9181073307991028,
+      "learning_rate": 0.00016855642705335437,
+      "loss": 0.6209,
+      "step": 3337
+    },
+    {
+      "epoch": 1.0717611173543105,
+      "grad_norm": 1.7982044219970703,
+      "learning_rate": 0.0001685310104605784,
+      "loss": 0.6342,
+      "step": 3338
+    },
+    {
+      "epoch": 1.072082196179162,
+      "grad_norm": 1.0803323984146118,
+      "learning_rate": 0.0001685055855175067,
+      "loss": 0.8042,
+      "step": 3339
+    },
+    {
+      "epoch": 1.0724032750040136,
+      "grad_norm": 0.936269223690033,
+      "learning_rate": 0.0001684801522272372,
+      "loss": 0.589,
+      "step": 3340
+    },
+    {
+      "epoch": 1.0727243538288649,
+      "grad_norm": 0.9068983197212219,
+      "learning_rate": 0.00016845471059286887,
+      "loss": 0.5997,
+      "step": 3341
+    },
+    {
+      "epoch": 1.0730454326537164,
+      "grad_norm": 1.2561771869659424,
+      "learning_rate": 0.0001684292606175017,
+      "loss": 0.6738,
+      "step": 3342
+    },
+    {
+      "epoch": 1.073366511478568,
+      "grad_norm": 2.2465908527374268,
+      "learning_rate": 0.00016840380230423668,
+      "loss": 0.7389,
+      "step": 3343
+    },
+    {
+      "epoch": 1.0736875903034195,
+      "grad_norm": 1.6561983823776245,
+      "learning_rate": 0.0001683783356561759,
+      "loss": 0.8363,
+      "step": 3344
+    },
+    {
+      "epoch": 1.074008669128271,
+      "grad_norm": 0.843169629573822,
+      "learning_rate": 0.00016835286067642228,
+      "loss": 0.6381,
+      "step": 3345
+    },
+    {
+      "epoch": 1.0743297479531224,
+      "grad_norm": 0.7259941697120667,
+      "learning_rate": 0.00016832737736807994,
+      "loss": 0.5703,
+      "step": 3346
+    },
+    {
+      "epoch": 1.074650826777974,
+      "grad_norm": 1.018009901046753,
+      "learning_rate": 0.00016830188573425389,
+      "loss": 0.7034,
+      "step": 3347
+    },
+    {
+      "epoch": 1.0749719056028255,
+      "grad_norm": 0.8844952583312988,
+      "learning_rate": 0.00016827638577805026,
+      "loss": 0.6634,
+      "step": 3348
+    },
+    {
+      "epoch": 1.075292984427677,
+      "grad_norm": 0.9090389609336853,
+      "learning_rate": 0.0001682508775025762,
+      "loss": 0.5809,
+      "step": 3349
+    },
+    {
+      "epoch": 1.0756140632525284,
+      "grad_norm": 0.8599204421043396,
+      "learning_rate": 0.00016822536091093965,
+      "loss": 0.6162,
+      "step": 3350
+    },
+    {
+      "epoch": 1.07593514207738,
+      "grad_norm": 1.0319000482559204,
+      "learning_rate": 0.00016819983600624986,
+      "loss": 0.6512,
+      "step": 3351
+    },
+    {
+      "epoch": 1.0762562209022315,
+      "grad_norm": 0.8953613042831421,
+      "learning_rate": 0.0001681743027916169,
+      "loss": 0.4827,
+      "step": 3352
+    },
+    {
+      "epoch": 1.076577299727083,
+      "grad_norm": 0.7418705821037292,
+      "learning_rate": 0.000168148761270152,
+      "loss": 0.4673,
+      "step": 3353
+    },
+    {
+      "epoch": 1.0768983785519346,
+      "grad_norm": 1.4244052171707153,
+      "learning_rate": 0.0001681232114449672,
+      "loss": 0.728,
+      "step": 3354
+    },
+    {
+      "epoch": 1.077219457376786,
+      "grad_norm": 0.9894974827766418,
+      "learning_rate": 0.00016809765331917575,
+      "loss": 0.6679,
+      "step": 3355
+    },
+    {
+      "epoch": 1.0775405362016375,
+      "grad_norm": 0.8581247925758362,
+      "learning_rate": 0.0001680720868958918,
+      "loss": 0.5517,
+      "step": 3356
+    },
+    {
+      "epoch": 1.077861615026489,
+      "grad_norm": 0.8083166480064392,
+      "learning_rate": 0.00016804651217823053,
+      "loss": 0.5014,
+      "step": 3357
+    },
+    {
+      "epoch": 1.0781826938513406,
+      "grad_norm": 1.293942928314209,
+      "learning_rate": 0.00016802092916930818,
+      "loss": 0.724,
+      "step": 3358
+    },
+    {
+      "epoch": 1.078503772676192,
+      "grad_norm": 0.8754002451896667,
+      "learning_rate": 0.00016799533787224192,
+      "loss": 0.5268,
+      "step": 3359
+    },
+    {
+      "epoch": 1.0788248515010435,
+      "grad_norm": 0.9752049446105957,
+      "learning_rate": 0.00016796973829015,
+      "loss": 0.5621,
+      "step": 3360
+    },
+    {
+      "epoch": 1.079145930325895,
+      "grad_norm": 0.9150457978248596,
+      "learning_rate": 0.00016794413042615168,
+      "loss": 0.4815,
+      "step": 3361
+    },
+    {
+      "epoch": 1.0794670091507466,
+      "grad_norm": 0.992889940738678,
+      "learning_rate": 0.00016791851428336711,
+      "loss": 0.5026,
+      "step": 3362
+    },
+    {
+      "epoch": 1.079788087975598,
+      "grad_norm": 0.6349382996559143,
+      "learning_rate": 0.00016789288986491762,
+      "loss": 0.3832,
+      "step": 3363
+    },
+    {
+      "epoch": 1.0801091668004494,
+      "grad_norm": 0.9240784645080566,
+      "learning_rate": 0.00016786725717392545,
+      "loss": 0.4269,
+      "step": 3364
+    },
+    {
+      "epoch": 1.080430245625301,
+      "grad_norm": 0.9478742480278015,
+      "learning_rate": 0.00016784161621351382,
+      "loss": 0.9252,
+      "step": 3365
+    },
+    {
+      "epoch": 1.0807513244501525,
+      "grad_norm": 0.8908936381340027,
+      "learning_rate": 0.0001678159669868071,
+      "loss": 0.9994,
+      "step": 3366
+    },
+    {
+      "epoch": 1.081072403275004,
+      "grad_norm": 0.7920323610305786,
+      "learning_rate": 0.00016779030949693044,
+      "loss": 0.5493,
+      "step": 3367
+    },
+    {
+      "epoch": 1.0813934820998554,
+      "grad_norm": 1.6557127237319946,
+      "learning_rate": 0.00016776464374701025,
+      "loss": 0.638,
+      "step": 3368
+    },
+    {
+      "epoch": 1.081714560924707,
+      "grad_norm": 0.9681761860847473,
+      "learning_rate": 0.00016773896974017373,
+      "loss": 0.5296,
+      "step": 3369
+    },
+    {
+      "epoch": 1.0820356397495585,
+      "grad_norm": 0.812778651714325,
+      "learning_rate": 0.00016771328747954925,
+      "loss": 0.3148,
+      "step": 3370
+    },
+    {
+      "epoch": 1.08235671857441,
+      "grad_norm": 0.8901128768920898,
+      "learning_rate": 0.00016768759696826608,
+      "loss": 0.6074,
+      "step": 3371
+    },
+    {
+      "epoch": 1.0826777973992616,
+      "grad_norm": 1.0377326011657715,
+      "learning_rate": 0.00016766189820945456,
+      "loss": 0.5944,
+      "step": 3372
+    },
+    {
+      "epoch": 1.082998876224113,
+      "grad_norm": 0.9749375581741333,
+      "learning_rate": 0.00016763619120624594,
+      "loss": 0.6133,
+      "step": 3373
+    },
+    {
+      "epoch": 1.0833199550489645,
+      "grad_norm": 1.1712234020233154,
+      "learning_rate": 0.00016761047596177263,
+      "loss": 0.735,
+      "step": 3374
+    },
+    {
+      "epoch": 1.083641033873816,
+      "grad_norm": 0.9546951651573181,
+      "learning_rate": 0.00016758475247916787,
+      "loss": 0.6357,
+      "step": 3375
+    },
+    {
+      "epoch": 1.0839621126986676,
+      "grad_norm": 1.6145797967910767,
+      "learning_rate": 0.00016755902076156604,
+      "loss": 0.8267,
+      "step": 3376
+    },
+    {
+      "epoch": 1.084283191523519,
+      "grad_norm": 0.9280446767807007,
+      "learning_rate": 0.00016753328081210245,
+      "loss": 0.7201,
+      "step": 3377
+    },
+    {
+      "epoch": 1.0846042703483705,
+      "grad_norm": 0.7923296689987183,
+      "learning_rate": 0.00016750753263391346,
+      "loss": 0.6021,
+      "step": 3378
+    },
+    {
+      "epoch": 1.084925349173222,
+      "grad_norm": 0.9804905652999878,
+      "learning_rate": 0.00016748177623013638,
+      "loss": 0.544,
+      "step": 3379
+    },
+    {
+      "epoch": 1.0852464279980736,
+      "grad_norm": 0.815140962600708,
+      "learning_rate": 0.00016745601160390958,
+      "loss": 0.6108,
+      "step": 3380
+    },
+    {
+      "epoch": 1.0855675068229251,
+      "grad_norm": 0.842607855796814,
+      "learning_rate": 0.00016743023875837233,
+      "loss": 0.6777,
+      "step": 3381
+    },
+    {
+      "epoch": 1.0858885856477765,
+      "grad_norm": 0.7987306118011475,
+      "learning_rate": 0.00016740445769666509,
+      "loss": 0.672,
+      "step": 3382
+    },
+    {
+      "epoch": 1.086209664472628,
+      "grad_norm": 0.8180364966392517,
+      "learning_rate": 0.0001673786684219291,
+      "loss": 0.5932,
+      "step": 3383
+    },
+    {
+      "epoch": 1.0865307432974796,
+      "grad_norm": 1.0452377796173096,
+      "learning_rate": 0.00016735287093730676,
+      "loss": 0.5845,
+      "step": 3384
+    },
+    {
+      "epoch": 1.0868518221223311,
+      "grad_norm": 1.3078027963638306,
+      "learning_rate": 0.00016732706524594137,
+      "loss": 0.632,
+      "step": 3385
+    },
+    {
+      "epoch": 1.0871729009471824,
+      "grad_norm": 1.3132222890853882,
+      "learning_rate": 0.00016730125135097735,
+      "loss": 0.5767,
+      "step": 3386
+    },
+    {
+      "epoch": 1.087493979772034,
+      "grad_norm": 1.1473840475082397,
+      "learning_rate": 0.00016727542925555996,
+      "loss": 0.7189,
+      "step": 3387
+    },
+    {
+      "epoch": 1.0878150585968855,
+      "grad_norm": 1.0360926389694214,
+      "learning_rate": 0.00016724959896283559,
+      "loss": 0.6505,
+      "step": 3388
+    },
+    {
+      "epoch": 1.088136137421737,
+      "grad_norm": 0.9349396228790283,
+      "learning_rate": 0.00016722376047595164,
+      "loss": 0.7034,
+      "step": 3389
+    },
+    {
+      "epoch": 1.0884572162465886,
+      "grad_norm": 1.2072162628173828,
+      "learning_rate": 0.0001671979137980563,
+      "loss": 0.6266,
+      "step": 3390
+    },
+    {
+      "epoch": 1.08877829507144,
+      "grad_norm": 1.2379400730133057,
+      "learning_rate": 0.00016717205893229903,
+      "loss": 0.6909,
+      "step": 3391
+    },
+    {
+      "epoch": 1.0890993738962915,
+      "grad_norm": 0.8755012154579163,
+      "learning_rate": 0.00016714619588183014,
+      "loss": 0.7037,
+      "step": 3392
+    },
+    {
+      "epoch": 1.089420452721143,
+      "grad_norm": 1.0528205633163452,
+      "learning_rate": 0.00016712032464980095,
+      "loss": 0.6186,
+      "step": 3393
+    },
+    {
+      "epoch": 1.0897415315459946,
+      "grad_norm": 0.9427685141563416,
+      "learning_rate": 0.0001670944452393638,
+      "loss": 0.7993,
+      "step": 3394
+    },
+    {
+      "epoch": 1.090062610370846,
+      "grad_norm": 0.9150214791297913,
+      "learning_rate": 0.000167068557653672,
+      "loss": 0.739,
+      "step": 3395
+    },
+    {
+      "epoch": 1.0903836891956975,
+      "grad_norm": 2.4165000915527344,
+      "learning_rate": 0.0001670426618958799,
+      "loss": 0.6721,
+      "step": 3396
+    },
+    {
+      "epoch": 1.090704768020549,
+      "grad_norm": 0.821657121181488,
+      "learning_rate": 0.00016701675796914286,
+      "loss": 0.6333,
+      "step": 3397
+    },
+    {
+      "epoch": 1.0910258468454006,
+      "grad_norm": 0.8281720876693726,
+      "learning_rate": 0.0001669908458766171,
+      "loss": 0.6536,
+      "step": 3398
+    },
+    {
+      "epoch": 1.0913469256702522,
+      "grad_norm": 0.7724263668060303,
+      "learning_rate": 0.00016696492562145996,
+      "loss": 0.5401,
+      "step": 3399
+    },
+    {
+      "epoch": 1.0916680044951035,
+      "grad_norm": 1.0343469381332397,
+      "learning_rate": 0.00016693899720682977,
+      "loss": 0.6455,
+      "step": 3400
+    },
+    {
+      "epoch": 1.091989083319955,
+      "grad_norm": 0.9886199831962585,
+      "learning_rate": 0.00016691306063588583,
+      "loss": 0.4821,
+      "step": 3401
+    },
+    {
+      "epoch": 1.0923101621448066,
+      "grad_norm": 0.8337252736091614,
+      "learning_rate": 0.00016688711591178842,
+      "loss": 0.5141,
+      "step": 3402
+    },
+    {
+      "epoch": 1.0926312409696581,
+      "grad_norm": 0.8266648054122925,
+      "learning_rate": 0.00016686116303769882,
+      "loss": 0.5598,
+      "step": 3403
+    },
+    {
+      "epoch": 1.0929523197945095,
+      "grad_norm": 0.9328741431236267,
+      "learning_rate": 0.0001668352020167793,
+      "loss": 0.7067,
+      "step": 3404
+    },
+    {
+      "epoch": 1.093273398619361,
+      "grad_norm": 1.163390040397644,
+      "learning_rate": 0.00016680923285219317,
+      "loss": 0.6493,
+      "step": 3405
+    },
+    {
+      "epoch": 1.0935944774442126,
+      "grad_norm": 1.0952308177947998,
+      "learning_rate": 0.00016678325554710468,
+      "loss": 0.6423,
+      "step": 3406
+    },
+    {
+      "epoch": 1.0939155562690641,
+      "grad_norm": 1.341279149055481,
+      "learning_rate": 0.00016675727010467906,
+      "loss": 0.751,
+      "step": 3407
+    },
+    {
+      "epoch": 1.0942366350939157,
+      "grad_norm": 0.7064843773841858,
+      "learning_rate": 0.0001667312765280826,
+      "loss": 0.4315,
+      "step": 3408
+    },
+    {
+      "epoch": 1.094557713918767,
+      "grad_norm": 1.5312671661376953,
+      "learning_rate": 0.00016670527482048246,
+      "loss": 0.5555,
+      "step": 3409
+    },
+    {
+      "epoch": 1.0948787927436185,
+      "grad_norm": 0.8741896152496338,
+      "learning_rate": 0.00016667926498504696,
+      "loss": 0.573,
+      "step": 3410
+    },
+    {
+      "epoch": 1.09519987156847,
+      "grad_norm": 0.6594826579093933,
+      "learning_rate": 0.00016665324702494524,
+      "loss": 0.4327,
+      "step": 3411
+    },
+    {
+      "epoch": 1.0955209503933216,
+      "grad_norm": 0.7997058033943176,
+      "learning_rate": 0.0001666272209433476,
+      "loss": 0.4166,
+      "step": 3412
+    },
+    {
+      "epoch": 1.095842029218173,
+      "grad_norm": 1.526983380317688,
+      "learning_rate": 0.00016660118674342517,
+      "loss": 0.6392,
+      "step": 3413
+    },
+    {
+      "epoch": 1.0961631080430245,
+      "grad_norm": 0.7481285929679871,
+      "learning_rate": 0.00016657514442835014,
+      "loss": 0.4221,
+      "step": 3414
+    },
+    {
+      "epoch": 1.096484186867876,
+      "grad_norm": 0.7330780625343323,
+      "learning_rate": 0.00016654909400129575,
+      "loss": 0.8458,
+      "step": 3415
+    },
+    {
+      "epoch": 1.0968052656927276,
+      "grad_norm": 0.798095703125,
+      "learning_rate": 0.00016652303546543608,
+      "loss": 1.141,
+      "step": 3416
+    },
+    {
+      "epoch": 1.0971263445175792,
+      "grad_norm": 1.091076374053955,
+      "learning_rate": 0.00016649696882394633,
+      "loss": 0.6236,
+      "step": 3417
+    },
+    {
+      "epoch": 1.0974474233424305,
+      "grad_norm": 1.0143516063690186,
+      "learning_rate": 0.00016647089408000266,
+      "loss": 0.5477,
+      "step": 3418
+    },
+    {
+      "epoch": 1.097768502167282,
+      "grad_norm": 0.8810601234436035,
+      "learning_rate": 0.00016644481123678217,
+      "loss": 0.3834,
+      "step": 3419
+    },
+    {
+      "epoch": 1.0980895809921336,
+      "grad_norm": 1.0156632661819458,
+      "learning_rate": 0.00016641872029746297,
+      "loss": 0.3836,
+      "step": 3420
+    },
+    {
+      "epoch": 1.0984106598169852,
+      "grad_norm": 1.0656507015228271,
+      "learning_rate": 0.00016639262126522418,
+      "loss": 0.5248,
+      "step": 3421
+    },
+    {
+      "epoch": 1.0987317386418365,
+      "grad_norm": 0.9195857048034668,
+      "learning_rate": 0.00016636651414324587,
+      "loss": 0.3556,
+      "step": 3422
+    },
+    {
+      "epoch": 1.099052817466688,
+      "grad_norm": 1.0913866758346558,
+      "learning_rate": 0.00016634039893470912,
+      "loss": 0.8998,
+      "step": 3423
+    },
+    {
+      "epoch": 1.0993738962915396,
+      "grad_norm": 1.01368248462677,
+      "learning_rate": 0.000166314275642796,
+      "loss": 0.8026,
+      "step": 3424
+    },
+    {
+      "epoch": 1.0996949751163911,
+      "grad_norm": 0.8792101740837097,
+      "learning_rate": 0.00016628814427068953,
+      "loss": 0.7112,
+      "step": 3425
+    },
+    {
+      "epoch": 1.1000160539412427,
+      "grad_norm": 1.1131765842437744,
+      "learning_rate": 0.00016626200482157378,
+      "loss": 0.7527,
+      "step": 3426
+    },
+    {
+      "epoch": 1.100337132766094,
+      "grad_norm": 0.7965254783630371,
+      "learning_rate": 0.00016623585729863368,
+      "loss": 0.5228,
+      "step": 3427
+    },
+    {
+      "epoch": 1.1006582115909456,
+      "grad_norm": 0.7073742747306824,
+      "learning_rate": 0.00016620970170505534,
+      "loss": 0.5549,
+      "step": 3428
+    },
+    {
+      "epoch": 1.1009792904157971,
+      "grad_norm": 1.1967341899871826,
+      "learning_rate": 0.00016618353804402568,
+      "loss": 0.8142,
+      "step": 3429
+    },
+    {
+      "epoch": 1.1013003692406487,
+      "grad_norm": 1.0835646390914917,
+      "learning_rate": 0.00016615736631873262,
+      "loss": 0.7728,
+      "step": 3430
+    },
+    {
+      "epoch": 1.1016214480655,
+      "grad_norm": 0.8416442275047302,
+      "learning_rate": 0.00016613118653236518,
+      "loss": 0.5987,
+      "step": 3431
+    },
+    {
+      "epoch": 1.1019425268903515,
+      "grad_norm": 3.7327215671539307,
+      "learning_rate": 0.00016610499868811328,
+      "loss": 0.8418,
+      "step": 3432
+    },
+    {
+      "epoch": 1.102263605715203,
+      "grad_norm": 0.9980570077896118,
+      "learning_rate": 0.00016607880278916777,
+      "loss": 0.5245,
+      "step": 3433
+    },
+    {
+      "epoch": 1.1025846845400546,
+      "grad_norm": 0.9758525490760803,
+      "learning_rate": 0.0001660525988387206,
+      "loss": 0.6669,
+      "step": 3434
+    },
+    {
+      "epoch": 1.1029057633649062,
+      "grad_norm": 1.14530611038208,
+      "learning_rate": 0.00016602638683996463,
+      "loss": 0.7346,
+      "step": 3435
+    },
+    {
+      "epoch": 1.1032268421897575,
+      "grad_norm": 1.0736594200134277,
+      "learning_rate": 0.0001660001667960937,
+      "loss": 0.8106,
+      "step": 3436
+    },
+    {
+      "epoch": 1.103547921014609,
+      "grad_norm": 1.2492066621780396,
+      "learning_rate": 0.00016597393871030264,
+      "loss": 0.6462,
+      "step": 3437
+    },
+    {
+      "epoch": 1.1038689998394606,
+      "grad_norm": 1.3556599617004395,
+      "learning_rate": 0.0001659477025857872,
+      "loss": 0.8159,
+      "step": 3438
+    },
+    {
+      "epoch": 1.1041900786643122,
+      "grad_norm": 0.9458972811698914,
+      "learning_rate": 0.00016592145842574433,
+      "loss": 0.6553,
+      "step": 3439
+    },
+    {
+      "epoch": 1.1045111574891635,
+      "grad_norm": 1.0294822454452515,
+      "learning_rate": 0.0001658952062333717,
+      "loss": 0.6278,
+      "step": 3440
+    },
+    {
+      "epoch": 1.104832236314015,
+      "grad_norm": 1.8948813676834106,
+      "learning_rate": 0.00016586894601186805,
+      "loss": 0.7584,
+      "step": 3441
+    },
+    {
+      "epoch": 1.1051533151388666,
+      "grad_norm": 1.0939890146255493,
+      "learning_rate": 0.00016584267776443317,
+      "loss": 0.7916,
+      "step": 3442
+    },
+    {
+      "epoch": 1.1054743939637182,
+      "grad_norm": 1.3582258224487305,
+      "learning_rate": 0.00016581640149426768,
+      "loss": 0.6503,
+      "step": 3443
+    },
+    {
+      "epoch": 1.1057954727885697,
+      "grad_norm": 1.082184910774231,
+      "learning_rate": 0.00016579011720457333,
+      "loss": 0.6415,
+      "step": 3444
+    },
+    {
+      "epoch": 1.106116551613421,
+      "grad_norm": 0.8474713563919067,
+      "learning_rate": 0.00016576382489855274,
+      "loss": 0.5673,
+      "step": 3445
+    },
+    {
+      "epoch": 1.1064376304382726,
+      "grad_norm": 0.9523148536682129,
+      "learning_rate": 0.0001657375245794096,
+      "loss": 0.6361,
+      "step": 3446
+    },
+    {
+      "epoch": 1.1067587092631241,
+      "grad_norm": 1.021826982498169,
+      "learning_rate": 0.00016571121625034847,
+      "loss": 0.7183,
+      "step": 3447
+    },
+    {
+      "epoch": 1.1070797880879757,
+      "grad_norm": 1.3967097997665405,
+      "learning_rate": 0.00016568489991457497,
+      "loss": 0.7846,
+      "step": 3448
+    },
+    {
+      "epoch": 1.107400866912827,
+      "grad_norm": 0.7168581485748291,
+      "learning_rate": 0.00016565857557529566,
+      "loss": 0.4492,
+      "step": 3449
+    },
+    {
+      "epoch": 1.1077219457376786,
+      "grad_norm": 0.7148603200912476,
+      "learning_rate": 0.00016563224323571806,
+      "loss": 0.4595,
+      "step": 3450
+    },
+    {
+      "epoch": 1.1080430245625301,
+      "grad_norm": 1.2017083168029785,
+      "learning_rate": 0.00016560590289905073,
+      "loss": 0.5782,
+      "step": 3451
+    },
+    {
+      "epoch": 1.1083641033873817,
+      "grad_norm": 0.8505964279174805,
+      "learning_rate": 0.00016557955456850313,
+      "loss": 0.5918,
+      "step": 3452
+    },
+    {
+      "epoch": 1.108685182212233,
+      "grad_norm": 1.3165613412857056,
+      "learning_rate": 0.00016555319824728575,
+      "loss": 0.6645,
+      "step": 3453
+    },
+    {
+      "epoch": 1.1090062610370846,
+      "grad_norm": 0.9442883729934692,
+      "learning_rate": 0.00016552683393860997,
+      "loss": 0.6143,
+      "step": 3454
+    },
+    {
+      "epoch": 1.109327339861936,
+      "grad_norm": 1.0663771629333496,
+      "learning_rate": 0.00016550046164568827,
+      "loss": 0.5655,
+      "step": 3455
+    },
+    {
+      "epoch": 1.1096484186867877,
+      "grad_norm": 0.8024634122848511,
+      "learning_rate": 0.00016547408137173396,
+      "loss": 0.4356,
+      "step": 3456
+    },
+    {
+      "epoch": 1.1099694975116392,
+      "grad_norm": 1.0509380102157593,
+      "learning_rate": 0.00016544769311996148,
+      "loss": 0.5454,
+      "step": 3457
+    },
+    {
+      "epoch": 1.1102905763364905,
+      "grad_norm": 1.2511743307113647,
+      "learning_rate": 0.00016542129689358612,
+      "loss": 0.7829,
+      "step": 3458
+    },
+    {
+      "epoch": 1.110611655161342,
+      "grad_norm": 0.8298420906066895,
+      "learning_rate": 0.00016539489269582416,
+      "loss": 0.5684,
+      "step": 3459
+    },
+    {
+      "epoch": 1.1109327339861936,
+      "grad_norm": 0.8872331380844116,
+      "learning_rate": 0.00016536848052989291,
+      "loss": 0.6568,
+      "step": 3460
+    },
+    {
+      "epoch": 1.1112538128110452,
+      "grad_norm": 1.0351537466049194,
+      "learning_rate": 0.00016534206039901057,
+      "loss": 0.5094,
+      "step": 3461
+    },
+    {
+      "epoch": 1.1115748916358965,
+      "grad_norm": 0.9417381286621094,
+      "learning_rate": 0.00016531563230639637,
+      "loss": 0.4403,
+      "step": 3462
+    },
+    {
+      "epoch": 1.111895970460748,
+      "grad_norm": 1.471907615661621,
+      "learning_rate": 0.0001652891962552705,
+      "loss": 0.5832,
+      "step": 3463
+    },
+    {
+      "epoch": 1.1122170492855996,
+      "grad_norm": 0.8464019894599915,
+      "learning_rate": 0.00016526275224885411,
+      "loss": 0.4628,
+      "step": 3464
+    },
+    {
+      "epoch": 1.1125381281104512,
+      "grad_norm": 0.7151861190795898,
+      "learning_rate": 0.00016523630029036931,
+      "loss": 0.6354,
+      "step": 3465
+    },
+    {
+      "epoch": 1.1128592069353027,
+      "grad_norm": 0.8638876080513,
+      "learning_rate": 0.00016520984038303924,
+      "loss": 0.9545,
+      "step": 3466
+    },
+    {
+      "epoch": 1.113180285760154,
+      "grad_norm": 0.7509087920188904,
+      "learning_rate": 0.0001651833725300879,
+      "loss": 0.8626,
+      "step": 3467
+    },
+    {
+      "epoch": 1.1135013645850056,
+      "grad_norm": 0.7917301058769226,
+      "learning_rate": 0.00016515689673474033,
+      "loss": 0.5097,
+      "step": 3468
+    },
+    {
+      "epoch": 1.1138224434098571,
+      "grad_norm": 1.0008916854858398,
+      "learning_rate": 0.00016513041300022255,
+      "loss": 0.4359,
+      "step": 3469
+    },
+    {
+      "epoch": 1.1141435222347087,
+      "grad_norm": 1.335162878036499,
+      "learning_rate": 0.0001651039213297615,
+      "loss": 0.3395,
+      "step": 3470
+    },
+    {
+      "epoch": 1.11446460105956,
+      "grad_norm": 0.8309035897254944,
+      "learning_rate": 0.0001650774217265851,
+      "loss": 0.4172,
+      "step": 3471
+    },
+    {
+      "epoch": 1.1147856798844116,
+      "grad_norm": 1.0509357452392578,
+      "learning_rate": 0.00016505091419392228,
+      "loss": 0.9272,
+      "step": 3472
+    },
+    {
+      "epoch": 1.1151067587092631,
+      "grad_norm": 1.047833800315857,
+      "learning_rate": 0.00016502439873500289,
+      "loss": 0.8088,
+      "step": 3473
+    },
+    {
+      "epoch": 1.1154278375341147,
+      "grad_norm": 0.9323020577430725,
+      "learning_rate": 0.00016499787535305776,
+      "loss": 0.713,
+      "step": 3474
+    },
+    {
+      "epoch": 1.1157489163589662,
+      "grad_norm": 0.902013897895813,
+      "learning_rate": 0.00016497134405131866,
+      "loss": 0.644,
+      "step": 3475
+    },
+    {
+      "epoch": 1.1160699951838176,
+      "grad_norm": 0.8789449334144592,
+      "learning_rate": 0.00016494480483301836,
+      "loss": 0.6663,
+      "step": 3476
+    },
+    {
+      "epoch": 1.116391074008669,
+      "grad_norm": 0.7771266102790833,
+      "learning_rate": 0.00016491825770139062,
+      "loss": 0.7559,
+      "step": 3477
+    },
+    {
+      "epoch": 1.1167121528335207,
+      "grad_norm": 0.9561981558799744,
+      "learning_rate": 0.00016489170265967008,
+      "loss": 0.7239,
+      "step": 3478
+    },
+    {
+      "epoch": 1.1170332316583722,
+      "grad_norm": 0.9018192887306213,
+      "learning_rate": 0.00016486513971109243,
+      "loss": 0.8175,
+      "step": 3479
+    },
+    {
+      "epoch": 1.1173543104832235,
+      "grad_norm": 1.1967219114303589,
+      "learning_rate": 0.0001648385688588942,
+      "loss": 0.7613,
+      "step": 3480
+    },
+    {
+      "epoch": 1.117675389308075,
+      "grad_norm": 0.9554547667503357,
+      "learning_rate": 0.0001648119901063131,
+      "loss": 0.4639,
+      "step": 3481
+    },
+    {
+      "epoch": 1.1179964681329266,
+      "grad_norm": 1.2637308835983276,
+      "learning_rate": 0.00016478540345658759,
+      "loss": 0.7384,
+      "step": 3482
+    },
+    {
+      "epoch": 1.1183175469577782,
+      "grad_norm": 0.9255183935165405,
+      "learning_rate": 0.00016475880891295716,
+      "loss": 0.7321,
+      "step": 3483
+    },
+    {
+      "epoch": 1.1186386257826297,
+      "grad_norm": 1.0068658590316772,
+      "learning_rate": 0.0001647322064786623,
+      "loss": 0.6394,
+      "step": 3484
+    },
+    {
+      "epoch": 1.118959704607481,
+      "grad_norm": 1.1133503913879395,
+      "learning_rate": 0.00016470559615694446,
+      "loss": 0.713,
+      "step": 3485
+    },
+    {
+      "epoch": 1.1192807834323326,
+      "grad_norm": 0.8424758911132812,
+      "learning_rate": 0.000164678977951046,
+      "loss": 0.692,
+      "step": 3486
+    },
+    {
+      "epoch": 1.1196018622571842,
+      "grad_norm": 1.4674954414367676,
+      "learning_rate": 0.0001646523518642102,
+      "loss": 0.7578,
+      "step": 3487
+    },
+    {
+      "epoch": 1.1199229410820357,
+      "grad_norm": 1.0364352464675903,
+      "learning_rate": 0.00016462571789968152,
+      "loss": 0.6651,
+      "step": 3488
+    },
+    {
+      "epoch": 1.120244019906887,
+      "grad_norm": 0.8905995488166809,
+      "learning_rate": 0.0001645990760607051,
+      "loss": 0.5398,
+      "step": 3489
+    },
+    {
+      "epoch": 1.1205650987317386,
+      "grad_norm": 0.9402886629104614,
+      "learning_rate": 0.00016457242635052724,
+      "loss": 0.6423,
+      "step": 3490
+    },
+    {
+      "epoch": 1.1208861775565901,
+      "grad_norm": 1.093013882637024,
+      "learning_rate": 0.00016454576877239507,
+      "loss": 0.629,
+      "step": 3491
+    },
+    {
+      "epoch": 1.1212072563814417,
+      "grad_norm": 1.6441494226455688,
+      "learning_rate": 0.00016451910332955679,
+      "loss": 0.8024,
+      "step": 3492
+    },
+    {
+      "epoch": 1.1215283352062932,
+      "grad_norm": 1.1009982824325562,
+      "learning_rate": 0.00016449243002526144,
+      "loss": 0.6887,
+      "step": 3493
+    },
+    {
+      "epoch": 1.1218494140311446,
+      "grad_norm": 0.9255393743515015,
+      "learning_rate": 0.00016446574886275913,
+      "loss": 0.6147,
+      "step": 3494
+    },
+    {
+      "epoch": 1.1221704928559961,
+      "grad_norm": 1.203102469444275,
+      "learning_rate": 0.0001644390598453009,
+      "loss": 0.7035,
+      "step": 3495
+    },
+    {
+      "epoch": 1.1224915716808477,
+      "grad_norm": 1.5952414274215698,
+      "learning_rate": 0.00016441236297613866,
+      "loss": 0.753,
+      "step": 3496
+    },
+    {
+      "epoch": 1.1228126505056992,
+      "grad_norm": 1.0285307168960571,
+      "learning_rate": 0.0001643856582585254,
+      "loss": 0.6581,
+      "step": 3497
+    },
+    {
+      "epoch": 1.1231337293305506,
+      "grad_norm": 1.1580029726028442,
+      "learning_rate": 0.00016435894569571496,
+      "loss": 0.75,
+      "step": 3498
+    },
+    {
+      "epoch": 1.123454808155402,
+      "grad_norm": 2.223442792892456,
+      "learning_rate": 0.0001643322252909622,
+      "loss": 0.751,
+      "step": 3499
+    },
+    {
+      "epoch": 1.1237758869802537,
+      "grad_norm": 1.224037528038025,
+      "learning_rate": 0.00016430549704752294,
+      "loss": 0.6832,
+      "step": 3500
+    },
+    {
+      "epoch": 1.1240969658051052,
+      "grad_norm": 0.9943028092384338,
+      "learning_rate": 0.00016427876096865394,
+      "loss": 0.639,
+      "step": 3501
+    },
+    {
+      "epoch": 1.1244180446299565,
+      "grad_norm": 1.723598599433899,
+      "learning_rate": 0.00016425201705761288,
+      "loss": 0.7863,
+      "step": 3502
+    },
+    {
+      "epoch": 1.124739123454808,
+      "grad_norm": 0.7261873483657837,
+      "learning_rate": 0.00016422526531765846,
+      "loss": 0.53,
+      "step": 3503
+    },
+    {
+      "epoch": 1.1250602022796596,
+      "grad_norm": 0.9602160453796387,
+      "learning_rate": 0.00016419850575205024,
+      "loss": 0.4901,
+      "step": 3504
+    },
+    {
+      "epoch": 1.1253812811045112,
+      "grad_norm": 0.8210904002189636,
+      "learning_rate": 0.00016417173836404887,
+      "loss": 0.4663,
+      "step": 3505
+    },
+    {
+      "epoch": 1.1257023599293627,
+      "grad_norm": 1.026781439781189,
+      "learning_rate": 0.00016414496315691581,
+      "loss": 0.6553,
+      "step": 3506
+    },
+    {
+      "epoch": 1.126023438754214,
+      "grad_norm": 0.9842190146446228,
+      "learning_rate": 0.00016411818013391355,
+      "loss": 0.4576,
+      "step": 3507
+    },
+    {
+      "epoch": 1.1263445175790656,
+      "grad_norm": 0.8347097635269165,
+      "learning_rate": 0.00016409138929830553,
+      "loss": 0.478,
+      "step": 3508
+    },
+    {
+      "epoch": 1.1266655964039172,
+      "grad_norm": 1.0178107023239136,
+      "learning_rate": 0.00016406459065335615,
+      "loss": 0.621,
+      "step": 3509
+    },
+    {
+      "epoch": 1.1269866752287687,
+      "grad_norm": 0.8920689225196838,
+      "learning_rate": 0.00016403778420233075,
+      "loss": 0.4913,
+      "step": 3510
+    },
+    {
+      "epoch": 1.12730775405362,
+      "grad_norm": 0.9431245923042297,
+      "learning_rate": 0.00016401096994849557,
+      "loss": 0.5671,
+      "step": 3511
+    },
+    {
+      "epoch": 1.1276288328784716,
+      "grad_norm": 1.142845869064331,
+      "learning_rate": 0.00016398414789511786,
+      "loss": 0.6351,
+      "step": 3512
+    },
+    {
+      "epoch": 1.1279499117033231,
+      "grad_norm": 0.4318004250526428,
+      "learning_rate": 0.0001639573180454658,
+      "loss": 0.3718,
+      "step": 3513
+    },
+    {
+      "epoch": 1.1282709905281747,
+      "grad_norm": 0.7192524671554565,
+      "learning_rate": 0.00016393048040280855,
+      "loss": 0.411,
+      "step": 3514
+    },
+    {
+      "epoch": 1.1285920693530263,
+      "grad_norm": 0.7344934344291687,
+      "learning_rate": 0.00016390363497041622,
+      "loss": 0.7704,
+      "step": 3515
+    },
+    {
+      "epoch": 1.1289131481778776,
+      "grad_norm": 0.746320366859436,
+      "learning_rate": 0.00016387678175155978,
+      "loss": 1.1028,
+      "step": 3516
+    },
+    {
+      "epoch": 1.1292342270027291,
+      "grad_norm": 0.9115157127380371,
+      "learning_rate": 0.00016384992074951123,
+      "loss": 1.1808,
+      "step": 3517
+    },
+    {
+      "epoch": 1.1295553058275807,
+      "grad_norm": 1.0850014686584473,
+      "learning_rate": 0.00016382305196754356,
+      "loss": 0.6813,
+      "step": 3518
+    },
+    {
+      "epoch": 1.1298763846524322,
+      "grad_norm": 1.0016835927963257,
+      "learning_rate": 0.00016379617540893056,
+      "loss": 0.5984,
+      "step": 3519
+    },
+    {
+      "epoch": 1.1301974634772836,
+      "grad_norm": 0.8297609090805054,
+      "learning_rate": 0.0001637692910769471,
+      "loss": 0.3999,
+      "step": 3520
+    },
+    {
+      "epoch": 1.130518542302135,
+      "grad_norm": 0.8285101056098938,
+      "learning_rate": 0.000163742398974869,
+      "loss": 0.4791,
+      "step": 3521
+    },
+    {
+      "epoch": 1.1308396211269867,
+      "grad_norm": 0.9352186918258667,
+      "learning_rate": 0.00016371549910597287,
+      "loss": 0.5488,
+      "step": 3522
+    },
+    {
+      "epoch": 1.1311606999518382,
+      "grad_norm": 1.118675947189331,
+      "learning_rate": 0.0001636885914735365,
+      "loss": 0.7107,
+      "step": 3523
+    },
+    {
+      "epoch": 1.1314817787766898,
+      "grad_norm": 0.8197215795516968,
+      "learning_rate": 0.00016366167608083843,
+      "loss": 0.6317,
+      "step": 3524
+    },
+    {
+      "epoch": 1.131802857601541,
+      "grad_norm": 0.8849331736564636,
+      "learning_rate": 0.00016363475293115824,
+      "loss": 0.7617,
+      "step": 3525
+    },
+    {
+      "epoch": 1.1321239364263926,
+      "grad_norm": 1.0144376754760742,
+      "learning_rate": 0.0001636078220277764,
+      "loss": 0.7364,
+      "step": 3526
+    },
+    {
+      "epoch": 1.1324450152512442,
+      "grad_norm": 0.8314408659934998,
+      "learning_rate": 0.00016358088337397442,
+      "loss": 0.759,
+      "step": 3527
+    },
+    {
+      "epoch": 1.1327660940760957,
+      "grad_norm": 0.8000319004058838,
+      "learning_rate": 0.00016355393697303465,
+      "loss": 0.6728,
+      "step": 3528
+    },
+    {
+      "epoch": 1.133087172900947,
+      "grad_norm": 0.9634177684783936,
+      "learning_rate": 0.00016352698282824044,
+      "loss": 0.7623,
+      "step": 3529
+    },
+    {
+      "epoch": 1.1334082517257986,
+      "grad_norm": 0.7649347186088562,
+      "learning_rate": 0.00016350002094287609,
+      "loss": 0.5196,
+      "step": 3530
+    },
+    {
+      "epoch": 1.1337293305506502,
+      "grad_norm": 0.8428080081939697,
+      "learning_rate": 0.00016347305132022677,
+      "loss": 0.6598,
+      "step": 3531
+    },
+    {
+      "epoch": 1.1340504093755017,
+      "grad_norm": 0.9351022243499756,
+      "learning_rate": 0.0001634460739635787,
+      "loss": 0.667,
+      "step": 3532
+    },
+    {
+      "epoch": 1.1343714882003533,
+      "grad_norm": 0.854694128036499,
+      "learning_rate": 0.00016341908887621895,
+      "loss": 0.7199,
+      "step": 3533
+    },
+    {
+      "epoch": 1.1346925670252046,
+      "grad_norm": 0.8774638175964355,
+      "learning_rate": 0.00016339209606143563,
+      "loss": 0.4986,
+      "step": 3534
+    },
+    {
+      "epoch": 1.1350136458500562,
+      "grad_norm": 0.8318768739700317,
+      "learning_rate": 0.00016336509552251766,
+      "loss": 0.7203,
+      "step": 3535
+    },
+    {
+      "epoch": 1.1353347246749077,
+      "grad_norm": 1.0216394662857056,
+      "learning_rate": 0.000163338087262755,
+      "loss": 0.7023,
+      "step": 3536
+    },
+    {
+      "epoch": 1.1356558034997593,
+      "grad_norm": 1.4008989334106445,
+      "learning_rate": 0.00016331107128543857,
+      "loss": 0.8295,
+      "step": 3537
+    },
+    {
+      "epoch": 1.1359768823246106,
+      "grad_norm": 0.8423093557357788,
+      "learning_rate": 0.00016328404759386014,
+      "loss": 0.6274,
+      "step": 3538
+    },
+    {
+      "epoch": 1.1362979611494621,
+      "grad_norm": 1.1864322423934937,
+      "learning_rate": 0.00016325701619131246,
+      "loss": 0.6257,
+      "step": 3539
+    },
+    {
+      "epoch": 1.1366190399743137,
+      "grad_norm": 1.03271484375,
+      "learning_rate": 0.00016322997708108922,
+      "loss": 0.712,
+      "step": 3540
+    },
+    {
+      "epoch": 1.1369401187991652,
+      "grad_norm": 0.9949981570243835,
+      "learning_rate": 0.0001632029302664851,
+      "loss": 0.7167,
+      "step": 3541
+    },
+    {
+      "epoch": 1.1372611976240168,
+      "grad_norm": 0.9992634057998657,
+      "learning_rate": 0.00016317587575079563,
+      "loss": 0.6267,
+      "step": 3542
+    },
+    {
+      "epoch": 1.1375822764488681,
+      "grad_norm": 0.9867181181907654,
+      "learning_rate": 0.00016314881353731732,
+      "loss": 0.6953,
+      "step": 3543
+    },
+    {
+      "epoch": 1.1379033552737197,
+      "grad_norm": 1.025295615196228,
+      "learning_rate": 0.00016312174362934765,
+      "loss": 0.8544,
+      "step": 3544
+    },
+    {
+      "epoch": 1.1382244340985712,
+      "grad_norm": 0.9153071641921997,
+      "learning_rate": 0.00016309466603018496,
+      "loss": 0.6237,
+      "step": 3545
+    },
+    {
+      "epoch": 1.1385455129234228,
+      "grad_norm": 0.8474112749099731,
+      "learning_rate": 0.00016306758074312864,
+      "loss": 0.6815,
+      "step": 3546
+    },
+    {
+      "epoch": 1.138866591748274,
+      "grad_norm": 0.9534016847610474,
+      "learning_rate": 0.0001630404877714789,
+      "loss": 0.7636,
+      "step": 3547
+    },
+    {
+      "epoch": 1.1391876705731256,
+      "grad_norm": 1.3331577777862549,
+      "learning_rate": 0.00016301338711853693,
+      "loss": 0.8063,
+      "step": 3548
+    },
+    {
+      "epoch": 1.1395087493979772,
+      "grad_norm": 1.318224549293518,
+      "learning_rate": 0.00016298627878760487,
+      "loss": 0.7616,
+      "step": 3549
+    },
+    {
+      "epoch": 1.1398298282228287,
+      "grad_norm": 1.084550142288208,
+      "learning_rate": 0.00016295916278198584,
+      "loss": 0.7024,
+      "step": 3550
+    },
+    {
+      "epoch": 1.1401509070476803,
+      "grad_norm": 1.946920394897461,
+      "learning_rate": 0.00016293203910498376,
+      "loss": 0.8216,
+      "step": 3551
+    },
+    {
+      "epoch": 1.1404719858725316,
+      "grad_norm": 0.9112229943275452,
+      "learning_rate": 0.0001629049077599036,
+      "loss": 0.59,
+      "step": 3552
+    },
+    {
+      "epoch": 1.1407930646973832,
+      "grad_norm": 0.9161878824234009,
+      "learning_rate": 0.0001628777687500513,
+      "loss": 0.4761,
+      "step": 3553
+    },
+    {
+      "epoch": 1.1411141435222347,
+      "grad_norm": 0.8824045658111572,
+      "learning_rate": 0.00016285062207873355,
+      "loss": 0.6317,
+      "step": 3554
+    },
+    {
+      "epoch": 1.1414352223470863,
+      "grad_norm": 0.897005021572113,
+      "learning_rate": 0.00016282346774925817,
+      "loss": 0.5112,
+      "step": 3555
+    },
+    {
+      "epoch": 1.1417563011719376,
+      "grad_norm": 0.7413773536682129,
+      "learning_rate": 0.00016279630576493382,
+      "loss": 0.5975,
+      "step": 3556
+    },
+    {
+      "epoch": 1.1420773799967892,
+      "grad_norm": 0.8787861466407776,
+      "learning_rate": 0.00016276913612907007,
+      "loss": 0.508,
+      "step": 3557
+    },
+    {
+      "epoch": 1.1423984588216407,
+      "grad_norm": 1.0276175737380981,
+      "learning_rate": 0.0001627419588449775,
+      "loss": 0.6004,
+      "step": 3558
+    },
+    {
+      "epoch": 1.1427195376464923,
+      "grad_norm": 1.1329381465911865,
+      "learning_rate": 0.00016271477391596753,
+      "loss": 0.5258,
+      "step": 3559
+    },
+    {
+      "epoch": 1.1430406164713438,
+      "grad_norm": 0.9113026857376099,
+      "learning_rate": 0.0001626875813453526,
+      "loss": 0.5028,
+      "step": 3560
+    },
+    {
+      "epoch": 1.1433616952961951,
+      "grad_norm": 1.3290276527404785,
+      "learning_rate": 0.00016266038113644607,
+      "loss": 0.6505,
+      "step": 3561
+    },
+    {
+      "epoch": 1.1436827741210467,
+      "grad_norm": 0.9289746284484863,
+      "learning_rate": 0.00016263317329256213,
+      "loss": 0.449,
+      "step": 3562
+    },
+    {
+      "epoch": 1.1440038529458982,
+      "grad_norm": 1.1701115369796753,
+      "learning_rate": 0.00016260595781701604,
+      "loss": 0.6017,
+      "step": 3563
+    },
+    {
+      "epoch": 1.1443249317707498,
+      "grad_norm": 1.0579955577850342,
+      "learning_rate": 0.0001625787347131239,
+      "loss": 0.3807,
+      "step": 3564
+    },
+    {
+      "epoch": 1.1446460105956011,
+      "grad_norm": 0.7256002426147461,
+      "learning_rate": 0.00016255150398420271,
+      "loss": 0.6831,
+      "step": 3565
+    },
+    {
+      "epoch": 1.1449670894204527,
+      "grad_norm": 0.8824403882026672,
+      "learning_rate": 0.00016252426563357055,
+      "loss": 0.8835,
+      "step": 3566
+    },
+    {
+      "epoch": 1.1452881682453042,
+      "grad_norm": 1.1135727167129517,
+      "learning_rate": 0.00016249701966454625,
+      "loss": 0.4474,
+      "step": 3567
+    },
+    {
+      "epoch": 1.1456092470701558,
+      "grad_norm": 1.1596729755401611,
+      "learning_rate": 0.0001624697660804497,
+      "loss": 0.4549,
+      "step": 3568
+    },
+    {
+      "epoch": 1.1459303258950073,
+      "grad_norm": 1.1575850248336792,
+      "learning_rate": 0.00016244250488460158,
+      "loss": 0.3866,
+      "step": 3569
+    },
+    {
+      "epoch": 1.1462514047198586,
+      "grad_norm": 1.0404236316680908,
+      "learning_rate": 0.00016241523608032373,
+      "loss": 0.7547,
+      "step": 3570
+    },
+    {
+      "epoch": 1.1465724835447102,
+      "grad_norm": 1.020094633102417,
+      "learning_rate": 0.00016238795967093864,
+      "loss": 0.752,
+      "step": 3571
+    },
+    {
+      "epoch": 1.1468935623695617,
+      "grad_norm": 1.119242787361145,
+      "learning_rate": 0.00016236067565976992,
+      "loss": 0.755,
+      "step": 3572
+    },
+    {
+      "epoch": 1.1472146411944133,
+      "grad_norm": 0.830894947052002,
+      "learning_rate": 0.00016233338405014202,
+      "loss": 0.639,
+      "step": 3573
+    },
+    {
+      "epoch": 1.1475357200192646,
+      "grad_norm": 0.7884930968284607,
+      "learning_rate": 0.00016230608484538034,
+      "loss": 0.546,
+      "step": 3574
+    },
+    {
+      "epoch": 1.1478567988441162,
+      "grad_norm": 1.13563072681427,
+      "learning_rate": 0.00016227877804881127,
+      "loss": 0.7434,
+      "step": 3575
+    },
+    {
+      "epoch": 1.1481778776689677,
+      "grad_norm": 0.8976322412490845,
+      "learning_rate": 0.00016225146366376198,
+      "loss": 0.5961,
+      "step": 3576
+    },
+    {
+      "epoch": 1.1484989564938193,
+      "grad_norm": 0.8841371536254883,
+      "learning_rate": 0.00016222414169356065,
+      "loss": 0.5675,
+      "step": 3577
+    },
+    {
+      "epoch": 1.1488200353186708,
+      "grad_norm": 1.0855551958084106,
+      "learning_rate": 0.00016219681214153643,
+      "loss": 0.8424,
+      "step": 3578
+    },
+    {
+      "epoch": 1.1491411141435222,
+      "grad_norm": 1.0877726078033447,
+      "learning_rate": 0.0001621694750110193,
+      "loss": 0.8544,
+      "step": 3579
+    },
+    {
+      "epoch": 1.1494621929683737,
+      "grad_norm": 0.8400183320045471,
+      "learning_rate": 0.0001621421303053402,
+      "loss": 0.6311,
+      "step": 3580
+    },
+    {
+      "epoch": 1.1497832717932253,
+      "grad_norm": 0.8696182370185852,
+      "learning_rate": 0.00016211477802783103,
+      "loss": 0.5355,
+      "step": 3581
+    },
+    {
+      "epoch": 1.1501043506180768,
+      "grad_norm": 1.010580062866211,
+      "learning_rate": 0.0001620874181818246,
+      "loss": 0.7799,
+      "step": 3582
+    },
+    {
+      "epoch": 1.1504254294429281,
+      "grad_norm": 0.9656292200088501,
+      "learning_rate": 0.00016206005077065458,
+      "loss": 0.7286,
+      "step": 3583
+    },
+    {
+      "epoch": 1.1507465082677797,
+      "grad_norm": 1.5450822114944458,
+      "learning_rate": 0.00016203267579765563,
+      "loss": 0.791,
+      "step": 3584
+    },
+    {
+      "epoch": 1.1510675870926312,
+      "grad_norm": 0.836439847946167,
+      "learning_rate": 0.00016200529326616328,
+      "loss": 0.6037,
+      "step": 3585
+    },
+    {
+      "epoch": 1.1513886659174828,
+      "grad_norm": 1.2512813806533813,
+      "learning_rate": 0.00016197790317951403,
+      "loss": 0.5621,
+      "step": 3586
+    },
+    {
+      "epoch": 1.1517097447423343,
+      "grad_norm": 1.2249475717544556,
+      "learning_rate": 0.00016195050554104528,
+      "loss": 0.7086,
+      "step": 3587
+    },
+    {
+      "epoch": 1.1520308235671857,
+      "grad_norm": 1.0801141262054443,
+      "learning_rate": 0.00016192310035409536,
+      "loss": 0.7006,
+      "step": 3588
+    },
+    {
+      "epoch": 1.1523519023920372,
+      "grad_norm": 1.20231294631958,
+      "learning_rate": 0.00016189568762200348,
+      "loss": 0.6001,
+      "step": 3589
+    },
+    {
+      "epoch": 1.1526729812168888,
+      "grad_norm": 1.2221317291259766,
+      "learning_rate": 0.00016186826734810979,
+      "loss": 0.8367,
+      "step": 3590
+    },
+    {
+      "epoch": 1.1529940600417403,
+      "grad_norm": 0.8664124011993408,
+      "learning_rate": 0.0001618408395357554,
+      "loss": 0.646,
+      "step": 3591
+    },
+    {
+      "epoch": 1.1533151388665916,
+      "grad_norm": 1.5447945594787598,
+      "learning_rate": 0.00016181340418828233,
+      "loss": 0.7546,
+      "step": 3592
+    },
+    {
+      "epoch": 1.1536362176914432,
+      "grad_norm": 0.9267856478691101,
+      "learning_rate": 0.00016178596130903344,
+      "loss": 0.7593,
+      "step": 3593
+    },
+    {
+      "epoch": 1.1539572965162948,
+      "grad_norm": 1.0048587322235107,
+      "learning_rate": 0.0001617585109013526,
+      "loss": 0.6133,
+      "step": 3594
+    },
+    {
+      "epoch": 1.1542783753411463,
+      "grad_norm": 1.1687082052230835,
+      "learning_rate": 0.00016173105296858452,
+      "loss": 0.5192,
+      "step": 3595
+    },
+    {
+      "epoch": 1.1545994541659979,
+      "grad_norm": 0.8441829085350037,
+      "learning_rate": 0.00016170358751407487,
+      "loss": 0.6674,
+      "step": 3596
+    },
+    {
+      "epoch": 1.1549205329908492,
+      "grad_norm": 0.8441125154495239,
+      "learning_rate": 0.00016167611454117025,
+      "loss": 0.5651,
+      "step": 3597
+    },
+    {
+      "epoch": 1.1552416118157007,
+      "grad_norm": 0.8785204291343689,
+      "learning_rate": 0.0001616486340532182,
+      "loss": 0.6461,
+      "step": 3598
+    },
+    {
+      "epoch": 1.1555626906405523,
+      "grad_norm": 0.8196426033973694,
+      "learning_rate": 0.00016162114605356703,
+      "loss": 0.6182,
+      "step": 3599
+    },
+    {
+      "epoch": 1.1558837694654038,
+      "grad_norm": 1.4259721040725708,
+      "learning_rate": 0.0001615936505455662,
+      "loss": 0.7115,
+      "step": 3600
+    },
+    {
+      "epoch": 1.1562048482902552,
+      "grad_norm": 1.1216281652450562,
+      "learning_rate": 0.0001615661475325658,
+      "loss": 0.5116,
+      "step": 3601
+    },
+    {
+      "epoch": 1.1565259271151067,
+      "grad_norm": 0.9022599458694458,
+      "learning_rate": 0.00016153863701791717,
+      "loss": 0.5675,
+      "step": 3602
+    },
+    {
+      "epoch": 1.1568470059399583,
+      "grad_norm": 1.1778627634048462,
+      "learning_rate": 0.00016151111900497225,
+      "loss": 0.6868,
+      "step": 3603
+    },
+    {
+      "epoch": 1.1571680847648098,
+      "grad_norm": 1.2097725868225098,
+      "learning_rate": 0.00016148359349708402,
+      "loss": 0.6071,
+      "step": 3604
+    },
+    {
+      "epoch": 1.1574891635896614,
+      "grad_norm": 1.0817495584487915,
+      "learning_rate": 0.00016145606049760644,
+      "loss": 0.871,
+      "step": 3605
+    },
+    {
+      "epoch": 1.1578102424145127,
+      "grad_norm": 1.1181918382644653,
+      "learning_rate": 0.00016142852000989433,
+      "loss": 0.6375,
+      "step": 3606
+    },
+    {
+      "epoch": 1.1581313212393642,
+      "grad_norm": 1.0407376289367676,
+      "learning_rate": 0.00016140097203730337,
+      "loss": 0.6119,
+      "step": 3607
+    },
+    {
+      "epoch": 1.1584524000642158,
+      "grad_norm": 0.9020042419433594,
+      "learning_rate": 0.00016137341658319023,
+      "loss": 0.5714,
+      "step": 3608
+    },
+    {
+      "epoch": 1.1587734788890673,
+      "grad_norm": 1.4991035461425781,
+      "learning_rate": 0.00016134585365091243,
+      "loss": 0.8278,
+      "step": 3609
+    },
+    {
+      "epoch": 1.1590945577139187,
+      "grad_norm": 1.3282997608184814,
+      "learning_rate": 0.00016131828324382846,
+      "loss": 0.5376,
+      "step": 3610
+    },
+    {
+      "epoch": 1.1594156365387702,
+      "grad_norm": 3.6863720417022705,
+      "learning_rate": 0.00016129070536529766,
+      "loss": 0.5749,
+      "step": 3611
+    },
+    {
+      "epoch": 1.1597367153636218,
+      "grad_norm": 0.7200700640678406,
+      "learning_rate": 0.00016126312001868033,
+      "loss": 0.4114,
+      "step": 3612
+    },
+    {
+      "epoch": 1.1600577941884733,
+      "grad_norm": 0.7452567219734192,
+      "learning_rate": 0.00016123552720733765,
+      "loss": 0.5016,
+      "step": 3613
+    },
+    {
+      "epoch": 1.1603788730133249,
+      "grad_norm": 0.543973982334137,
+      "learning_rate": 0.00016120792693463174,
+      "loss": 0.4231,
+      "step": 3614
+    },
+    {
+      "epoch": 1.1606999518381762,
+      "grad_norm": 0.718130886554718,
+      "learning_rate": 0.00016118031920392558,
+      "loss": 0.6174,
+      "step": 3615
+    },
+    {
+      "epoch": 1.1610210306630278,
+      "grad_norm": 0.9145506024360657,
+      "learning_rate": 0.00016115270401858314,
+      "loss": 0.9531,
+      "step": 3616
+    },
+    {
+      "epoch": 1.1613421094878793,
+      "grad_norm": 0.9525867700576782,
+      "learning_rate": 0.00016112508138196917,
+      "loss": 0.6047,
+      "step": 3617
+    },
+    {
+      "epoch": 1.1616631883127309,
+      "grad_norm": 0.9359449744224548,
+      "learning_rate": 0.00016109745129744946,
+      "loss": 0.4499,
+      "step": 3618
+    },
+    {
+      "epoch": 1.1619842671375822,
+      "grad_norm": 0.8167291283607483,
+      "learning_rate": 0.00016106981376839066,
+      "loss": 0.4337,
+      "step": 3619
+    },
+    {
+      "epoch": 1.1623053459624337,
+      "grad_norm": 1.0462777614593506,
+      "learning_rate": 0.00016104216879816026,
+      "loss": 0.5782,
+      "step": 3620
+    },
+    {
+      "epoch": 1.1626264247872853,
+      "grad_norm": 0.8575799465179443,
+      "learning_rate": 0.0001610145163901268,
+      "loss": 0.7717,
+      "step": 3621
+    },
+    {
+      "epoch": 1.1629475036121368,
+      "grad_norm": 0.911267876625061,
+      "learning_rate": 0.00016098685654765955,
+      "loss": 0.786,
+      "step": 3622
+    },
+    {
+      "epoch": 1.1632685824369884,
+      "grad_norm": 0.5476121306419373,
+      "learning_rate": 0.0001609591892741288,
+      "loss": 0.4105,
+      "step": 3623
+    },
+    {
+      "epoch": 1.1635896612618397,
+      "grad_norm": 1.0486372709274292,
+      "learning_rate": 0.0001609315145729058,
+      "loss": 0.7131,
+      "step": 3624
+    },
+    {
+      "epoch": 1.1639107400866913,
+      "grad_norm": 0.9732680320739746,
+      "learning_rate": 0.00016090383244736256,
+      "loss": 0.6468,
+      "step": 3625
+    },
+    {
+      "epoch": 1.1642318189115428,
+      "grad_norm": 0.9579371213912964,
+      "learning_rate": 0.00016087614290087208,
+      "loss": 0.651,
+      "step": 3626
+    },
+    {
+      "epoch": 1.1645528977363944,
+      "grad_norm": 0.8852554559707642,
+      "learning_rate": 0.0001608484459368082,
+      "loss": 0.734,
+      "step": 3627
+    },
+    {
+      "epoch": 1.1648739765612457,
+      "grad_norm": 1.0034711360931396,
+      "learning_rate": 0.00016082074155854582,
+      "loss": 0.6897,
+      "step": 3628
+    },
+    {
+      "epoch": 1.1651950553860972,
+      "grad_norm": 1.2633333206176758,
+      "learning_rate": 0.00016079302976946055,
+      "loss": 0.5851,
+      "step": 3629
+    },
+    {
+      "epoch": 1.1655161342109488,
+      "grad_norm": 0.9554821848869324,
+      "learning_rate": 0.000160765310572929,
+      "loss": 0.6042,
+      "step": 3630
+    },
+    {
+      "epoch": 1.1658372130358003,
+      "grad_norm": 1.3010461330413818,
+      "learning_rate": 0.00016073758397232868,
+      "loss": 0.7993,
+      "step": 3631
+    },
+    {
+      "epoch": 1.166158291860652,
+      "grad_norm": 2.045257091522217,
+      "learning_rate": 0.000160709849971038,
+      "loss": 0.8856,
+      "step": 3632
+    },
+    {
+      "epoch": 1.1664793706855032,
+      "grad_norm": 1.0188322067260742,
+      "learning_rate": 0.00016068210857243624,
+      "loss": 0.7296,
+      "step": 3633
+    },
+    {
+      "epoch": 1.1668004495103548,
+      "grad_norm": 0.7480521202087402,
+      "learning_rate": 0.0001606543597799036,
+      "loss": 0.4887,
+      "step": 3634
+    },
+    {
+      "epoch": 1.1671215283352063,
+      "grad_norm": 1.0140467882156372,
+      "learning_rate": 0.00016062660359682124,
+      "loss": 0.8066,
+      "step": 3635
+    },
+    {
+      "epoch": 1.1674426071600579,
+      "grad_norm": 1.1531801223754883,
+      "learning_rate": 0.0001605988400265711,
+      "loss": 0.7233,
+      "step": 3636
+    },
+    {
+      "epoch": 1.1677636859849092,
+      "grad_norm": 0.8829801678657532,
+      "learning_rate": 0.00016057106907253616,
+      "loss": 0.6,
+      "step": 3637
+    },
+    {
+      "epoch": 1.1680847648097608,
+      "grad_norm": 1.2912678718566895,
+      "learning_rate": 0.00016054329073810015,
+      "loss": 0.6598,
+      "step": 3638
+    },
+    {
+      "epoch": 1.1684058436346123,
+      "grad_norm": 2.6909449100494385,
+      "learning_rate": 0.0001605155050266478,
+      "loss": 0.8908,
+      "step": 3639
+    },
+    {
+      "epoch": 1.1687269224594639,
+      "grad_norm": 1.1661655902862549,
+      "learning_rate": 0.00016048771194156477,
+      "loss": 0.8248,
+      "step": 3640
+    },
+    {
+      "epoch": 1.1690480012843154,
+      "grad_norm": 0.8896608948707581,
+      "learning_rate": 0.0001604599114862375,
+      "loss": 0.6519,
+      "step": 3641
+    },
+    {
+      "epoch": 1.1693690801091667,
+      "grad_norm": 1.0213063955307007,
+      "learning_rate": 0.0001604321036640534,
+      "loss": 0.598,
+      "step": 3642
+    },
+    {
+      "epoch": 1.1696901589340183,
+      "grad_norm": 1.0079537630081177,
+      "learning_rate": 0.00016040428847840079,
+      "loss": 0.5429,
+      "step": 3643
+    },
+    {
+      "epoch": 1.1700112377588698,
+      "grad_norm": 1.1655068397521973,
+      "learning_rate": 0.00016037646593266883,
+      "loss": 0.6184,
+      "step": 3644
+    },
+    {
+      "epoch": 1.1703323165837214,
+      "grad_norm": 1.1363699436187744,
+      "learning_rate": 0.00016034863603024767,
+      "loss": 0.6852,
+      "step": 3645
+    },
+    {
+      "epoch": 1.1706533954085727,
+      "grad_norm": 2.0418121814727783,
+      "learning_rate": 0.00016032079877452825,
+      "loss": 0.6902,
+      "step": 3646
+    },
+    {
+      "epoch": 1.1709744742334243,
+      "grad_norm": 0.8260238170623779,
+      "learning_rate": 0.00016029295416890248,
+      "loss": 0.5518,
+      "step": 3647
+    },
+    {
+      "epoch": 1.1712955530582758,
+      "grad_norm": 0.9947139620780945,
+      "learning_rate": 0.00016026510221676311,
+      "loss": 0.6918,
+      "step": 3648
+    },
+    {
+      "epoch": 1.1716166318831274,
+      "grad_norm": 1.0895559787750244,
+      "learning_rate": 0.00016023724292150385,
+      "loss": 0.8454,
+      "step": 3649
+    },
+    {
+      "epoch": 1.171937710707979,
+      "grad_norm": 1.3528664112091064,
+      "learning_rate": 0.00016020937628651927,
+      "loss": 0.6834,
+      "step": 3650
+    },
+    {
+      "epoch": 1.1722587895328302,
+      "grad_norm": 1.0099211931228638,
+      "learning_rate": 0.00016018150231520486,
+      "loss": 0.5097,
+      "step": 3651
+    },
+    {
+      "epoch": 1.1725798683576818,
+      "grad_norm": 1.2328240871429443,
+      "learning_rate": 0.0001601536210109569,
+      "loss": 0.6222,
+      "step": 3652
+    },
+    {
+      "epoch": 1.1729009471825333,
+      "grad_norm": 0.7499989867210388,
+      "learning_rate": 0.0001601257323771727,
+      "loss": 0.4985,
+      "step": 3653
+    },
+    {
+      "epoch": 1.173222026007385,
+      "grad_norm": 1.5503602027893066,
+      "learning_rate": 0.00016009783641725034,
+      "loss": 0.7534,
+      "step": 3654
+    },
+    {
+      "epoch": 1.1735431048322362,
+      "grad_norm": 1.1560394763946533,
+      "learning_rate": 0.00016006993313458896,
+      "loss": 0.5315,
+      "step": 3655
+    },
+    {
+      "epoch": 1.1738641836570878,
+      "grad_norm": 1.0473573207855225,
+      "learning_rate": 0.00016004202253258842,
+      "loss": 0.7029,
+      "step": 3656
+    },
+    {
+      "epoch": 1.1741852624819393,
+      "grad_norm": 0.7268290519714355,
+      "learning_rate": 0.00016001410461464956,
+      "loss": 0.3893,
+      "step": 3657
+    },
+    {
+      "epoch": 1.1745063413067909,
+      "grad_norm": 0.8399918079376221,
+      "learning_rate": 0.0001599861793841741,
+      "loss": 0.6149,
+      "step": 3658
+    },
+    {
+      "epoch": 1.1748274201316424,
+      "grad_norm": 0.9458096027374268,
+      "learning_rate": 0.00015995824684456465,
+      "loss": 0.6191,
+      "step": 3659
+    },
+    {
+      "epoch": 1.1751484989564938,
+      "grad_norm": 0.9515113234519958,
+      "learning_rate": 0.00015993030699922468,
+      "loss": 0.6762,
+      "step": 3660
+    },
+    {
+      "epoch": 1.1754695777813453,
+      "grad_norm": 0.9861279726028442,
+      "learning_rate": 0.0001599023598515586,
+      "loss": 0.4427,
+      "step": 3661
+    },
+    {
+      "epoch": 1.1757906566061969,
+      "grad_norm": 0.9347848892211914,
+      "learning_rate": 0.00015987440540497167,
+      "loss": 0.4939,
+      "step": 3662
+    },
+    {
+      "epoch": 1.1761117354310484,
+      "grad_norm": 0.8265044093132019,
+      "learning_rate": 0.00015984644366287006,
+      "loss": 0.525,
+      "step": 3663
+    },
+    {
+      "epoch": 1.1764328142558997,
+      "grad_norm": 1.6481767892837524,
+      "learning_rate": 0.00015981847462866083,
+      "loss": 0.4591,
+      "step": 3664
+    },
+    {
+      "epoch": 1.1767538930807513,
+      "grad_norm": 0.738932728767395,
+      "learning_rate": 0.0001597904983057519,
+      "loss": 0.5946,
+      "step": 3665
+    },
+    {
+      "epoch": 1.1770749719056028,
+      "grad_norm": 0.8453285694122314,
+      "learning_rate": 0.00015976251469755214,
+      "loss": 1.0439,
+      "step": 3666
+    },
+    {
+      "epoch": 1.1773960507304544,
+      "grad_norm": 1.1691863536834717,
+      "learning_rate": 0.00015973452380747122,
+      "loss": 0.7068,
+      "step": 3667
+    },
+    {
+      "epoch": 1.177717129555306,
+      "grad_norm": 0.9889507293701172,
+      "learning_rate": 0.00015970652563891978,
+      "loss": 0.6272,
+      "step": 3668
+    },
+    {
+      "epoch": 1.1780382083801573,
+      "grad_norm": 0.9118053317070007,
+      "learning_rate": 0.00015967852019530926,
+      "loss": 0.5458,
+      "step": 3669
+    },
+    {
+      "epoch": 1.1783592872050088,
+      "grad_norm": 0.9585251808166504,
+      "learning_rate": 0.00015965050748005215,
+      "loss": 0.3483,
+      "step": 3670
+    },
+    {
+      "epoch": 1.1786803660298604,
+      "grad_norm": 0.7852297425270081,
+      "learning_rate": 0.0001596224874965616,
+      "loss": 0.3958,
+      "step": 3671
+    },
+    {
+      "epoch": 1.179001444854712,
+      "grad_norm": 1.148110270500183,
+      "learning_rate": 0.0001595944602482518,
+      "loss": 0.7532,
+      "step": 3672
+    },
+    {
+      "epoch": 1.1793225236795633,
+      "grad_norm": 1.0729384422302246,
+      "learning_rate": 0.0001595664257385378,
+      "loss": 0.7822,
+      "step": 3673
+    },
+    {
+      "epoch": 1.1796436025044148,
+      "grad_norm": 1.0979061126708984,
+      "learning_rate": 0.00015953838397083552,
+      "loss": 0.8229,
+      "step": 3674
+    },
+    {
+      "epoch": 1.1799646813292664,
+      "grad_norm": 0.9034024477005005,
+      "learning_rate": 0.00015951033494856175,
+      "loss": 0.6591,
+      "step": 3675
+    },
+    {
+      "epoch": 1.180285760154118,
+      "grad_norm": 0.832550048828125,
+      "learning_rate": 0.00015948227867513415,
+      "loss": 0.6285,
+      "step": 3676
+    },
+    {
+      "epoch": 1.1806068389789695,
+      "grad_norm": 0.8613409399986267,
+      "learning_rate": 0.00015945421515397133,
+      "loss": 0.5975,
+      "step": 3677
+    },
+    {
+      "epoch": 1.1809279178038208,
+      "grad_norm": 0.762203574180603,
+      "learning_rate": 0.00015942614438849275,
+      "loss": 0.5549,
+      "step": 3678
+    },
+    {
+      "epoch": 1.1812489966286723,
+      "grad_norm": 0.8697448968887329,
+      "learning_rate": 0.0001593980663821187,
+      "loss": 0.7776,
+      "step": 3679
+    },
+    {
+      "epoch": 1.1815700754535239,
+      "grad_norm": 0.8074071407318115,
+      "learning_rate": 0.00015936998113827048,
+      "loss": 0.6768,
+      "step": 3680
+    },
+    {
+      "epoch": 1.1818911542783754,
+      "grad_norm": 0.7951146364212036,
+      "learning_rate": 0.00015934188866037016,
+      "loss": 0.6422,
+      "step": 3681
+    },
+    {
+      "epoch": 1.1822122331032268,
+      "grad_norm": 0.7514532804489136,
+      "learning_rate": 0.0001593137889518407,
+      "loss": 0.5805,
+      "step": 3682
+    },
+    {
+      "epoch": 1.1825333119280783,
+      "grad_norm": 0.7481774687767029,
+      "learning_rate": 0.00015928568201610595,
+      "loss": 0.5634,
+      "step": 3683
+    },
+    {
+      "epoch": 1.1828543907529299,
+      "grad_norm": 1.1757893562316895,
+      "learning_rate": 0.00015925756785659069,
+      "loss": 0.8346,
+      "step": 3684
+    },
+    {
+      "epoch": 1.1831754695777814,
+      "grad_norm": 1.0502216815948486,
+      "learning_rate": 0.00015922944647672052,
+      "loss": 0.6632,
+      "step": 3685
+    },
+    {
+      "epoch": 1.183496548402633,
+      "grad_norm": 0.7584315538406372,
+      "learning_rate": 0.00015920131787992197,
+      "loss": 0.5628,
+      "step": 3686
+    },
+    {
+      "epoch": 1.1838176272274843,
+      "grad_norm": 0.8802182674407959,
+      "learning_rate": 0.0001591731820696224,
+      "loss": 0.6358,
+      "step": 3687
+    },
+    {
+      "epoch": 1.1841387060523358,
+      "grad_norm": 0.9516767263412476,
+      "learning_rate": 0.0001591450390492501,
+      "loss": 0.6534,
+      "step": 3688
+    },
+    {
+      "epoch": 1.1844597848771874,
+      "grad_norm": 0.8391969799995422,
+      "learning_rate": 0.0001591168888222342,
+      "loss": 0.6248,
+      "step": 3689
+    },
+    {
+      "epoch": 1.184780863702039,
+      "grad_norm": 1.1127198934555054,
+      "learning_rate": 0.00015908873139200473,
+      "loss": 0.6813,
+      "step": 3690
+    },
+    {
+      "epoch": 1.1851019425268903,
+      "grad_norm": 1.2551233768463135,
+      "learning_rate": 0.00015906056676199255,
+      "loss": 0.577,
+      "step": 3691
+    },
+    {
+      "epoch": 1.1854230213517418,
+      "grad_norm": 1.1656824350357056,
+      "learning_rate": 0.00015903239493562948,
+      "loss": 0.8192,
+      "step": 3692
+    },
+    {
+      "epoch": 1.1857441001765934,
+      "grad_norm": 0.7559661269187927,
+      "learning_rate": 0.00015900421591634814,
+      "loss": 0.5647,
+      "step": 3693
+    },
+    {
+      "epoch": 1.186065179001445,
+      "grad_norm": 0.8689810037612915,
+      "learning_rate": 0.00015897602970758206,
+      "loss": 0.6108,
+      "step": 3694
+    },
+    {
+      "epoch": 1.1863862578262965,
+      "grad_norm": 1.167493462562561,
+      "learning_rate": 0.00015894783631276567,
+      "loss": 0.6969,
+      "step": 3695
+    },
+    {
+      "epoch": 1.1867073366511478,
+      "grad_norm": 1.027427077293396,
+      "learning_rate": 0.0001589196357353342,
+      "loss": 0.6776,
+      "step": 3696
+    },
+    {
+      "epoch": 1.1870284154759994,
+      "grad_norm": 1.0420182943344116,
+      "learning_rate": 0.00015889142797872387,
+      "loss": 0.606,
+      "step": 3697
+    },
+    {
+      "epoch": 1.187349494300851,
+      "grad_norm": 0.8472828269004822,
+      "learning_rate": 0.0001588632130463717,
+      "loss": 0.5962,
+      "step": 3698
+    },
+    {
+      "epoch": 1.1876705731257025,
+      "grad_norm": 0.9795169830322266,
+      "learning_rate": 0.00015883499094171554,
+      "loss": 0.59,
+      "step": 3699
+    },
+    {
+      "epoch": 1.1879916519505538,
+      "grad_norm": 1.1912764310836792,
+      "learning_rate": 0.0001588067616681942,
+      "loss": 0.7445,
+      "step": 3700
+    },
+    {
+      "epoch": 1.1883127307754053,
+      "grad_norm": 1.2817906141281128,
+      "learning_rate": 0.00015877852522924732,
+      "loss": 0.447,
+      "step": 3701
+    },
+    {
+      "epoch": 1.1886338096002569,
+      "grad_norm": 1.1116948127746582,
+      "learning_rate": 0.00015875028162831546,
+      "loss": 0.7283,
+      "step": 3702
+    },
+    {
+      "epoch": 1.1889548884251084,
+      "grad_norm": 1.298180341720581,
+      "learning_rate": 0.00015872203086883996,
+      "loss": 0.6075,
+      "step": 3703
+    },
+    {
+      "epoch": 1.18927596724996,
+      "grad_norm": 1.0509965419769287,
+      "learning_rate": 0.00015869377295426316,
+      "loss": 0.5656,
+      "step": 3704
+    },
+    {
+      "epoch": 1.1895970460748113,
+      "grad_norm": 1.1119554042816162,
+      "learning_rate": 0.00015866550788802813,
+      "loss": 0.5898,
+      "step": 3705
+    },
+    {
+      "epoch": 1.1899181248996629,
+      "grad_norm": 1.1349931955337524,
+      "learning_rate": 0.00015863723567357892,
+      "loss": 0.7535,
+      "step": 3706
+    },
+    {
+      "epoch": 1.1902392037245144,
+      "grad_norm": 1.1311873197555542,
+      "learning_rate": 0.00015860895631436043,
+      "loss": 0.6378,
+      "step": 3707
+    },
+    {
+      "epoch": 1.1905602825493657,
+      "grad_norm": 1.291495680809021,
+      "learning_rate": 0.0001585806698138184,
+      "loss": 0.6692,
+      "step": 3708
+    },
+    {
+      "epoch": 1.1908813613742173,
+      "grad_norm": 0.741792619228363,
+      "learning_rate": 0.0001585523761753994,
+      "loss": 0.3739,
+      "step": 3709
+    },
+    {
+      "epoch": 1.1912024401990688,
+      "grad_norm": 1.0837442874908447,
+      "learning_rate": 0.00015852407540255104,
+      "loss": 0.5723,
+      "step": 3710
+    },
+    {
+      "epoch": 1.1915235190239204,
+      "grad_norm": 0.8327803015708923,
+      "learning_rate": 0.00015849576749872157,
+      "loss": 0.565,
+      "step": 3711
+    },
+    {
+      "epoch": 1.191844597848772,
+      "grad_norm": 0.62799072265625,
+      "learning_rate": 0.00015846745246736026,
+      "loss": 0.3977,
+      "step": 3712
+    },
+    {
+      "epoch": 1.1921656766736235,
+      "grad_norm": 0.8305913209915161,
+      "learning_rate": 0.00015843913031191723,
+      "loss": 0.4562,
+      "step": 3713
+    },
+    {
+      "epoch": 1.1924867554984748,
+      "grad_norm": 0.7770129442214966,
+      "learning_rate": 0.00015841080103584342,
+      "loss": 0.4003,
+      "step": 3714
+    },
+    {
+      "epoch": 1.1928078343233264,
+      "grad_norm": 0.9295758605003357,
+      "learning_rate": 0.0001583824646425907,
+      "loss": 0.803,
+      "step": 3715
+    },
+    {
+      "epoch": 1.193128913148178,
+      "grad_norm": 0.8279706239700317,
+      "learning_rate": 0.00015835412113561175,
+      "loss": 0.3903,
+      "step": 3716
+    },
+    {
+      "epoch": 1.1934499919730293,
+      "grad_norm": 0.8809469938278198,
+      "learning_rate": 0.00015832577051836015,
+      "loss": 0.771,
+      "step": 3717
+    },
+    {
+      "epoch": 1.1937710707978808,
+      "grad_norm": 1.3099849224090576,
+      "learning_rate": 0.00015829741279429035,
+      "loss": 0.7353,
+      "step": 3718
+    },
+    {
+      "epoch": 1.1940921496227324,
+      "grad_norm": 0.8764384388923645,
+      "learning_rate": 0.00015826904796685762,
+      "loss": 0.5436,
+      "step": 3719
+    },
+    {
+      "epoch": 1.194413228447584,
+      "grad_norm": 0.8666921257972717,
+      "learning_rate": 0.00015824067603951812,
+      "loss": 0.323,
+      "step": 3720
+    },
+    {
+      "epoch": 1.1947343072724355,
+      "grad_norm": 0.9187259674072266,
+      "learning_rate": 0.00015821229701572896,
+      "loss": 0.5078,
+      "step": 3721
+    },
+    {
+      "epoch": 1.1950553860972868,
+      "grad_norm": 0.9329090118408203,
+      "learning_rate": 0.00015818391089894796,
+      "loss": 0.4697,
+      "step": 3722
+    },
+    {
+      "epoch": 1.1953764649221383,
+      "grad_norm": 0.9224849939346313,
+      "learning_rate": 0.00015815551769263387,
+      "loss": 0.6722,
+      "step": 3723
+    },
+    {
+      "epoch": 1.1956975437469899,
+      "grad_norm": 0.9811240434646606,
+      "learning_rate": 0.0001581271174002464,
+      "loss": 0.7558,
+      "step": 3724
+    },
+    {
+      "epoch": 1.1960186225718414,
+      "grad_norm": 0.8810230493545532,
+      "learning_rate": 0.000158098710025246,
+      "loss": 0.6703,
+      "step": 3725
+    },
+    {
+      "epoch": 1.1963397013966928,
+      "grad_norm": 0.852672815322876,
+      "learning_rate": 0.00015807029557109398,
+      "loss": 0.6907,
+      "step": 3726
+    },
+    {
+      "epoch": 1.1966607802215443,
+      "grad_norm": 0.9756046533584595,
+      "learning_rate": 0.0001580418740412526,
+      "loss": 0.7458,
+      "step": 3727
+    },
+    {
+      "epoch": 1.1969818590463959,
+      "grad_norm": 0.7994744181632996,
+      "learning_rate": 0.00015801344543918495,
+      "loss": 0.572,
+      "step": 3728
+    },
+    {
+      "epoch": 1.1973029378712474,
+      "grad_norm": 0.8432713747024536,
+      "learning_rate": 0.00015798500976835493,
+      "loss": 0.567,
+      "step": 3729
+    },
+    {
+      "epoch": 1.197624016696099,
+      "grad_norm": 0.8624154925346375,
+      "learning_rate": 0.00015795656703222736,
+      "loss": 0.5077,
+      "step": 3730
+    },
+    {
+      "epoch": 1.1979450955209503,
+      "grad_norm": 1.059381127357483,
+      "learning_rate": 0.0001579281172342679,
+      "loss": 0.6805,
+      "step": 3731
+    },
+    {
+      "epoch": 1.1982661743458018,
+      "grad_norm": 1.291218638420105,
+      "learning_rate": 0.00015789966037794306,
+      "loss": 0.816,
+      "step": 3732
+    },
+    {
+      "epoch": 1.1985872531706534,
+      "grad_norm": 1.0364362001419067,
+      "learning_rate": 0.00015787119646672025,
+      "loss": 0.7035,
+      "step": 3733
+    },
+    {
+      "epoch": 1.198908331995505,
+      "grad_norm": 0.8102630972862244,
+      "learning_rate": 0.00015784272550406765,
+      "loss": 0.4892,
+      "step": 3734
+    },
+    {
+      "epoch": 1.1992294108203563,
+      "grad_norm": 0.8558127284049988,
+      "learning_rate": 0.00015781424749345446,
+      "loss": 0.6971,
+      "step": 3735
+    },
+    {
+      "epoch": 1.1995504896452078,
+      "grad_norm": 0.9906354546546936,
+      "learning_rate": 0.00015778576243835054,
+      "loss": 0.5872,
+      "step": 3736
+    },
+    {
+      "epoch": 1.1998715684700594,
+      "grad_norm": 1.178027629852295,
+      "learning_rate": 0.00015775727034222675,
+      "loss": 0.7676,
+      "step": 3737
+    },
+    {
+      "epoch": 1.200192647294911,
+      "grad_norm": 0.6666367053985596,
+      "learning_rate": 0.0001577287712085548,
+      "loss": 0.4359,
+      "step": 3738
+    },
+    {
+      "epoch": 1.2005137261197625,
+      "grad_norm": 0.7421877980232239,
+      "learning_rate": 0.00015770026504080718,
+      "loss": 0.582,
+      "step": 3739
+    },
+    {
+      "epoch": 1.2008348049446138,
+      "grad_norm": 1.3498107194900513,
+      "learning_rate": 0.00015767175184245726,
+      "loss": 0.656,
+      "step": 3740
+    },
+    {
+      "epoch": 1.2011558837694654,
+      "grad_norm": 0.896021842956543,
+      "learning_rate": 0.00015764323161697935,
+      "loss": 0.4453,
+      "step": 3741
+    },
+    {
+      "epoch": 1.201476962594317,
+      "grad_norm": 0.9143558740615845,
+      "learning_rate": 0.00015761470436784846,
+      "loss": 0.4917,
+      "step": 3742
+    },
+    {
+      "epoch": 1.2017980414191685,
+      "grad_norm": 1.4460296630859375,
+      "learning_rate": 0.0001575861700985407,
+      "loss": 0.8317,
+      "step": 3743
+    },
+    {
+      "epoch": 1.2021191202440198,
+      "grad_norm": 1.7859466075897217,
+      "learning_rate": 0.0001575576288125327,
+      "loss": 0.6328,
+      "step": 3744
+    },
+    {
+      "epoch": 1.2024401990688713,
+      "grad_norm": 0.773352324962616,
+      "learning_rate": 0.00015752908051330227,
+      "loss": 0.4811,
+      "step": 3745
+    },
+    {
+      "epoch": 1.202761277893723,
+      "grad_norm": 0.9597437381744385,
+      "learning_rate": 0.00015750052520432787,
+      "loss": 0.6322,
+      "step": 3746
+    },
+    {
+      "epoch": 1.2030823567185744,
+      "grad_norm": 1.0898431539535522,
+      "learning_rate": 0.00015747196288908887,
+      "loss": 0.6503,
+      "step": 3747
+    },
+    {
+      "epoch": 1.203403435543426,
+      "grad_norm": 1.4335774183273315,
+      "learning_rate": 0.00015744339357106558,
+      "loss": 0.502,
+      "step": 3748
+    },
+    {
+      "epoch": 1.2037245143682773,
+      "grad_norm": 1.5681631565093994,
+      "learning_rate": 0.000157414817253739,
+      "loss": 0.7586,
+      "step": 3749
+    },
+    {
+      "epoch": 1.2040455931931289,
+      "grad_norm": 1.2081352472305298,
+      "learning_rate": 0.00015738623394059107,
+      "loss": 0.743,
+      "step": 3750
+    },
+    {
+      "epoch": 1.2043666720179804,
+      "grad_norm": 0.9064257740974426,
+      "learning_rate": 0.0001573576436351046,
+      "loss": 0.6028,
+      "step": 3751
+    },
+    {
+      "epoch": 1.204687750842832,
+      "grad_norm": 0.9491926431655884,
+      "learning_rate": 0.00015732904634076328,
+      "loss": 0.5614,
+      "step": 3752
+    },
+    {
+      "epoch": 1.2050088296676833,
+      "grad_norm": 1.0897586345672607,
+      "learning_rate": 0.00015730044206105156,
+      "loss": 0.5848,
+      "step": 3753
+    },
+    {
+      "epoch": 1.2053299084925349,
+      "grad_norm": 0.8406479358673096,
+      "learning_rate": 0.00015727183079945476,
+      "loss": 0.4931,
+      "step": 3754
+    },
+    {
+      "epoch": 1.2056509873173864,
+      "grad_norm": 0.9714513421058655,
+      "learning_rate": 0.0001572432125594591,
+      "loss": 0.4901,
+      "step": 3755
+    },
+    {
+      "epoch": 1.205972066142238,
+      "grad_norm": 1.135326623916626,
+      "learning_rate": 0.00015721458734455163,
+      "loss": 0.6236,
+      "step": 3756
+    },
+    {
+      "epoch": 1.2062931449670895,
+      "grad_norm": 2.5608773231506348,
+      "learning_rate": 0.00015718595515822027,
+      "loss": 0.5221,
+      "step": 3757
+    },
+    {
+      "epoch": 1.2066142237919408,
+      "grad_norm": 0.9237856864929199,
+      "learning_rate": 0.0001571573160039537,
+      "loss": 0.5451,
+      "step": 3758
+    },
+    {
+      "epoch": 1.2069353026167924,
+      "grad_norm": 1.0552841424942017,
+      "learning_rate": 0.00015712866988524155,
+      "loss": 0.5602,
+      "step": 3759
+    },
+    {
+      "epoch": 1.207256381441644,
+      "grad_norm": 1.123316764831543,
+      "learning_rate": 0.0001571000168055743,
+      "loss": 0.6103,
+      "step": 3760
+    },
+    {
+      "epoch": 1.2075774602664955,
+      "grad_norm": 0.7353083491325378,
+      "learning_rate": 0.0001570713567684432,
+      "loss": 0.4778,
+      "step": 3761
+    },
+    {
+      "epoch": 1.2078985390913468,
+      "grad_norm": 1.4208050966262817,
+      "learning_rate": 0.00015704268977734037,
+      "loss": 0.529,
+      "step": 3762
+    },
+    {
+      "epoch": 1.2082196179161984,
+      "grad_norm": 1.0792194604873657,
+      "learning_rate": 0.00015701401583575884,
+      "loss": 0.5751,
+      "step": 3763
+    },
+    {
+      "epoch": 1.20854069674105,
+      "grad_norm": 0.7297836542129517,
+      "learning_rate": 0.00015698533494719238,
+      "loss": 0.4981,
+      "step": 3764
+    },
+    {
+      "epoch": 1.2088617755659015,
+      "grad_norm": 0.9594284892082214,
+      "learning_rate": 0.00015695664711513576,
+      "loss": 0.9049,
+      "step": 3765
+    },
+    {
+      "epoch": 1.209182854390753,
+      "grad_norm": 0.8230898976325989,
+      "learning_rate": 0.00015692795234308445,
+      "loss": 0.964,
+      "step": 3766
+    },
+    {
+      "epoch": 1.2095039332156043,
+      "grad_norm": 0.7228332757949829,
+      "learning_rate": 0.0001568992506345348,
+      "loss": 1.1967,
+      "step": 3767
+    },
+    {
+      "epoch": 1.209825012040456,
+      "grad_norm": 0.8351823687553406,
+      "learning_rate": 0.00015687054199298408,
+      "loss": 0.5122,
+      "step": 3768
+    },
+    {
+      "epoch": 1.2101460908653074,
+      "grad_norm": 1.0156910419464111,
+      "learning_rate": 0.0001568418264219303,
+      "loss": 0.4224,
+      "step": 3769
+    },
+    {
+      "epoch": 1.210467169690159,
+      "grad_norm": 0.9522204995155334,
+      "learning_rate": 0.0001568131039248724,
+      "loss": 0.4744,
+      "step": 3770
+    },
+    {
+      "epoch": 1.2107882485150103,
+      "grad_norm": 0.835364818572998,
+      "learning_rate": 0.00015678437450531013,
+      "loss": 0.3386,
+      "step": 3771
+    },
+    {
+      "epoch": 1.2111093273398619,
+      "grad_norm": 1.0940889120101929,
+      "learning_rate": 0.00015675563816674407,
+      "loss": 0.4991,
+      "step": 3772
+    },
+    {
+      "epoch": 1.2114304061647134,
+      "grad_norm": 0.7849900722503662,
+      "learning_rate": 0.00015672689491267567,
+      "loss": 0.3609,
+      "step": 3773
+    },
+    {
+      "epoch": 1.211751484989565,
+      "grad_norm": 1.0691943168640137,
+      "learning_rate": 0.00015669814474660718,
+      "loss": 0.8524,
+      "step": 3774
+    },
+    {
+      "epoch": 1.2120725638144165,
+      "grad_norm": 2.4347715377807617,
+      "learning_rate": 0.0001566693876720417,
+      "loss": 0.6352,
+      "step": 3775
+    },
+    {
+      "epoch": 1.2123936426392679,
+      "grad_norm": 0.8152544498443604,
+      "learning_rate": 0.00015664062369248328,
+      "loss": 0.5616,
+      "step": 3776
+    },
+    {
+      "epoch": 1.2127147214641194,
+      "grad_norm": 0.7303499579429626,
+      "learning_rate": 0.00015661185281143667,
+      "loss": 0.546,
+      "step": 3777
+    },
+    {
+      "epoch": 1.213035800288971,
+      "grad_norm": 0.9726660251617432,
+      "learning_rate": 0.0001565830750324075,
+      "loss": 0.5368,
+      "step": 3778
+    },
+    {
+      "epoch": 1.2133568791138225,
+      "grad_norm": 0.8433571457862854,
+      "learning_rate": 0.0001565542903589023,
+      "loss": 0.5876,
+      "step": 3779
+    },
+    {
+      "epoch": 1.2136779579386738,
+      "grad_norm": 0.7719855308532715,
+      "learning_rate": 0.00015652549879442834,
+      "loss": 0.6917,
+      "step": 3780
+    },
+    {
+      "epoch": 1.2139990367635254,
+      "grad_norm": 0.931829571723938,
+      "learning_rate": 0.0001564967003424938,
+      "loss": 0.7677,
+      "step": 3781
+    },
+    {
+      "epoch": 1.214320115588377,
+      "grad_norm": 1.2775667905807495,
+      "learning_rate": 0.00015646789500660773,
+      "loss": 0.8883,
+      "step": 3782
+    },
+    {
+      "epoch": 1.2146411944132285,
+      "grad_norm": 0.9605757594108582,
+      "learning_rate": 0.00015643908279027992,
+      "loss": 0.6806,
+      "step": 3783
+    },
+    {
+      "epoch": 1.21496227323808,
+      "grad_norm": 0.9084275960922241,
+      "learning_rate": 0.00015641026369702106,
+      "loss": 0.6906,
+      "step": 3784
+    },
+    {
+      "epoch": 1.2152833520629314,
+      "grad_norm": 0.9735769629478455,
+      "learning_rate": 0.00015638143773034267,
+      "loss": 0.6756,
+      "step": 3785
+    },
+    {
+      "epoch": 1.215604430887783,
+      "grad_norm": 1.068384051322937,
+      "learning_rate": 0.00015635260489375714,
+      "loss": 0.7305,
+      "step": 3786
+    },
+    {
+      "epoch": 1.2159255097126345,
+      "grad_norm": 0.9031590223312378,
+      "learning_rate": 0.00015632376519077767,
+      "loss": 0.6657,
+      "step": 3787
+    },
+    {
+      "epoch": 1.216246588537486,
+      "grad_norm": 1.1685518026351929,
+      "learning_rate": 0.0001562949186249182,
+      "loss": 0.8324,
+      "step": 3788
+    },
+    {
+      "epoch": 1.2165676673623373,
+      "grad_norm": 0.9389743804931641,
+      "learning_rate": 0.0001562660651996937,
+      "loss": 0.6689,
+      "step": 3789
+    },
+    {
+      "epoch": 1.216888746187189,
+      "grad_norm": 0.8513121008872986,
+      "learning_rate": 0.0001562372049186198,
+      "loss": 0.5915,
+      "step": 3790
+    },
+    {
+      "epoch": 1.2172098250120404,
+      "grad_norm": 1.3301939964294434,
+      "learning_rate": 0.00015620833778521307,
+      "loss": 0.5882,
+      "step": 3791
+    },
+    {
+      "epoch": 1.217530903836892,
+      "grad_norm": 1.1245044469833374,
+      "learning_rate": 0.00015617946380299088,
+      "loss": 0.6076,
+      "step": 3792
+    },
+    {
+      "epoch": 1.2178519826617435,
+      "grad_norm": 0.8256964087486267,
+      "learning_rate": 0.00015615058297547145,
+      "loss": 0.5895,
+      "step": 3793
+    },
+    {
+      "epoch": 1.2181730614865949,
+      "grad_norm": 1.246702790260315,
+      "learning_rate": 0.0001561216953061738,
+      "loss": 0.6992,
+      "step": 3794
+    },
+    {
+      "epoch": 1.2184941403114464,
+      "grad_norm": 0.8872213959693909,
+      "learning_rate": 0.0001560928007986178,
+      "loss": 0.547,
+      "step": 3795
+    },
+    {
+      "epoch": 1.218815219136298,
+      "grad_norm": 1.465101718902588,
+      "learning_rate": 0.0001560638994563242,
+      "loss": 0.879,
+      "step": 3796
+    },
+    {
+      "epoch": 1.2191362979611495,
+      "grad_norm": 2.137561321258545,
+      "learning_rate": 0.00015603499128281448,
+      "loss": 0.5223,
+      "step": 3797
+    },
+    {
+      "epoch": 1.2194573767860009,
+      "grad_norm": 1.957757830619812,
+      "learning_rate": 0.00015600607628161103,
+      "loss": 0.5569,
+      "step": 3798
+    },
+    {
+      "epoch": 1.2197784556108524,
+      "grad_norm": 0.9652926921844482,
+      "learning_rate": 0.00015597715445623712,
+      "loss": 0.6782,
+      "step": 3799
+    },
+    {
+      "epoch": 1.220099534435704,
+      "grad_norm": 1.3139662742614746,
+      "learning_rate": 0.0001559482258102167,
+      "loss": 0.5627,
+      "step": 3800
+    },
+    {
+      "epoch": 1.2204206132605555,
+      "grad_norm": 1.0370103120803833,
+      "learning_rate": 0.0001559192903470747,
+      "loss": 0.5761,
+      "step": 3801
+    },
+    {
+      "epoch": 1.220741692085407,
+      "grad_norm": 1.4218648672103882,
+      "learning_rate": 0.00015589034807033677,
+      "loss": 0.6787,
+      "step": 3802
+    },
+    {
+      "epoch": 1.2210627709102584,
+      "grad_norm": 1.046571969985962,
+      "learning_rate": 0.00015586139898352946,
+      "loss": 0.5636,
+      "step": 3803
+    },
+    {
+      "epoch": 1.22138384973511,
+      "grad_norm": 0.8225265145301819,
+      "learning_rate": 0.00015583244309018014,
+      "loss": 0.5093,
+      "step": 3804
+    },
+    {
+      "epoch": 1.2217049285599615,
+      "grad_norm": 0.8345786929130554,
+      "learning_rate": 0.000155803480393817,
+      "loss": 0.6215,
+      "step": 3805
+    },
+    {
+      "epoch": 1.222026007384813,
+      "grad_norm": 1.0300137996673584,
+      "learning_rate": 0.00015577451089796905,
+      "loss": 0.5036,
+      "step": 3806
+    },
+    {
+      "epoch": 1.2223470862096644,
+      "grad_norm": 1.6953532695770264,
+      "learning_rate": 0.00015574553460616608,
+      "loss": 0.5542,
+      "step": 3807
+    },
+    {
+      "epoch": 1.222668165034516,
+      "grad_norm": 1.2669715881347656,
+      "learning_rate": 0.00015571655152193885,
+      "loss": 0.5519,
+      "step": 3808
+    },
+    {
+      "epoch": 1.2229892438593675,
+      "grad_norm": 0.9114670753479004,
+      "learning_rate": 0.00015568756164881882,
+      "loss": 0.5563,
+      "step": 3809
+    },
+    {
+      "epoch": 1.223310322684219,
+      "grad_norm": 1.259053349494934,
+      "learning_rate": 0.00015565856499033832,
+      "loss": 0.5622,
+      "step": 3810
+    },
+    {
+      "epoch": 1.2236314015090706,
+      "grad_norm": 1.13980233669281,
+      "learning_rate": 0.0001556295615500305,
+      "loss": 0.528,
+      "step": 3811
+    },
+    {
+      "epoch": 1.223952480333922,
+      "grad_norm": 1.1629769802093506,
+      "learning_rate": 0.0001556005513314293,
+      "loss": 0.389,
+      "step": 3812
+    },
+    {
+      "epoch": 1.2242735591587735,
+      "grad_norm": 1.1004542112350464,
+      "learning_rate": 0.00015557153433806966,
+      "loss": 0.43,
+      "step": 3813
+    },
+    {
+      "epoch": 1.224594637983625,
+      "grad_norm": 1.0557912588119507,
+      "learning_rate": 0.00015554251057348713,
+      "loss": 0.5082,
+      "step": 3814
+    },
+    {
+      "epoch": 1.2249157168084766,
+      "grad_norm": 0.8889212012290955,
+      "learning_rate": 0.0001555134800412181,
+      "loss": 0.9576,
+      "step": 3815
+    },
+    {
+      "epoch": 1.2252367956333279,
+      "grad_norm": 0.7801089882850647,
+      "learning_rate": 0.00015548444274479995,
+      "loss": 1.078,
+      "step": 3816
+    },
+    {
+      "epoch": 1.2255578744581794,
+      "grad_norm": 0.8572964668273926,
+      "learning_rate": 0.00015545539868777074,
+      "loss": 0.5412,
+      "step": 3817
+    },
+    {
+      "epoch": 1.225878953283031,
+      "grad_norm": 1.0506659746170044,
+      "learning_rate": 0.00015542634787366942,
+      "loss": 0.5478,
+      "step": 3818
+    },
+    {
+      "epoch": 1.2262000321078825,
+      "grad_norm": 0.9632019996643066,
+      "learning_rate": 0.00015539729030603574,
+      "loss": 0.4899,
+      "step": 3819
+    },
+    {
+      "epoch": 1.226521110932734,
+      "grad_norm": 0.9268298149108887,
+      "learning_rate": 0.00015536822598841024,
+      "loss": 0.2964,
+      "step": 3820
+    },
+    {
+      "epoch": 1.2268421897575854,
+      "grad_norm": 0.7900720834732056,
+      "learning_rate": 0.00015533915492433443,
+      "loss": 0.4777,
+      "step": 3821
+    },
+    {
+      "epoch": 1.227163268582437,
+      "grad_norm": 0.9437267780303955,
+      "learning_rate": 0.0001553100771173504,
+      "loss": 0.3384,
+      "step": 3822
+    },
+    {
+      "epoch": 1.2274843474072885,
+      "grad_norm": 0.995335578918457,
+      "learning_rate": 0.00015528099257100127,
+      "loss": 0.5097,
+      "step": 3823
+    },
+    {
+      "epoch": 1.22780542623214,
+      "grad_norm": 1.3662070035934448,
+      "learning_rate": 0.00015525190128883083,
+      "loss": 0.7474,
+      "step": 3824
+    },
+    {
+      "epoch": 1.2281265050569914,
+      "grad_norm": 1.1693106889724731,
+      "learning_rate": 0.00015522280327438388,
+      "loss": 0.6907,
+      "step": 3825
+    },
+    {
+      "epoch": 1.228447583881843,
+      "grad_norm": 0.9598641395568848,
+      "learning_rate": 0.0001551936985312058,
+      "loss": 0.651,
+      "step": 3826
+    },
+    {
+      "epoch": 1.2287686627066945,
+      "grad_norm": 0.9100416898727417,
+      "learning_rate": 0.00015516458706284303,
+      "loss": 0.6725,
+      "step": 3827
+    },
+    {
+      "epoch": 1.229089741531546,
+      "grad_norm": 0.8291442394256592,
+      "learning_rate": 0.00015513546887284264,
+      "loss": 0.621,
+      "step": 3828
+    },
+    {
+      "epoch": 1.2294108203563976,
+      "grad_norm": 0.8551065325737,
+      "learning_rate": 0.0001551063439647526,
+      "loss": 0.593,
+      "step": 3829
+    },
+    {
+      "epoch": 1.229731899181249,
+      "grad_norm": 0.6962476968765259,
+      "learning_rate": 0.00015507721234212172,
+      "loss": 0.6304,
+      "step": 3830
+    },
+    {
+      "epoch": 1.2300529780061005,
+      "grad_norm": 0.8959445357322693,
+      "learning_rate": 0.00015504807400849958,
+      "loss": 0.7921,
+      "step": 3831
+    },
+    {
+      "epoch": 1.230374056830952,
+      "grad_norm": 0.88429856300354,
+      "learning_rate": 0.0001550189289674366,
+      "loss": 0.6881,
+      "step": 3832
+    },
+    {
+      "epoch": 1.2306951356558036,
+      "grad_norm": 1.0589004755020142,
+      "learning_rate": 0.000154989777222484,
+      "loss": 0.6597,
+      "step": 3833
+    },
+    {
+      "epoch": 1.231016214480655,
+      "grad_norm": 0.932532548904419,
+      "learning_rate": 0.00015496061877719384,
+      "loss": 0.6299,
+      "step": 3834
+    },
+    {
+      "epoch": 1.2313372933055065,
+      "grad_norm": 1.763346552848816,
+      "learning_rate": 0.000154931453635119,
+      "loss": 0.6717,
+      "step": 3835
+    },
+    {
+      "epoch": 1.231658372130358,
+      "grad_norm": 2.0481505393981934,
+      "learning_rate": 0.0001549022817998132,
+      "loss": 0.6489,
+      "step": 3836
+    },
+    {
+      "epoch": 1.2319794509552096,
+      "grad_norm": 0.8745352029800415,
+      "learning_rate": 0.00015487310327483086,
+      "loss": 0.6171,
+      "step": 3837
+    },
+    {
+      "epoch": 1.232300529780061,
+      "grad_norm": 0.8354154825210571,
+      "learning_rate": 0.00015484391806372733,
+      "loss": 0.669,
+      "step": 3838
+    },
+    {
+      "epoch": 1.2326216086049124,
+      "grad_norm": 0.9111486673355103,
+      "learning_rate": 0.00015481472617005876,
+      "loss": 0.7414,
+      "step": 3839
+    },
+    {
+      "epoch": 1.232942687429764,
+      "grad_norm": 1.0252684354782104,
+      "learning_rate": 0.00015478552759738207,
+      "loss": 0.724,
+      "step": 3840
+    },
+    {
+      "epoch": 1.2332637662546155,
+      "grad_norm": 1.018760323524475,
+      "learning_rate": 0.00015475632234925504,
+      "loss": 0.5888,
+      "step": 3841
+    },
+    {
+      "epoch": 1.233584845079467,
+      "grad_norm": 1.5923256874084473,
+      "learning_rate": 0.00015472711042923621,
+      "loss": 0.6708,
+      "step": 3842
+    },
+    {
+      "epoch": 1.2339059239043184,
+      "grad_norm": 1.3504046201705933,
+      "learning_rate": 0.00015469789184088497,
+      "loss": 0.6927,
+      "step": 3843
+    },
+    {
+      "epoch": 1.23422700272917,
+      "grad_norm": 0.9088150262832642,
+      "learning_rate": 0.00015466866658776155,
+      "loss": 0.6003,
+      "step": 3844
+    },
+    {
+      "epoch": 1.2345480815540215,
+      "grad_norm": 0.9223833680152893,
+      "learning_rate": 0.00015463943467342693,
+      "loss": 0.7096,
+      "step": 3845
+    },
+    {
+      "epoch": 1.234869160378873,
+      "grad_norm": 0.9848785400390625,
+      "learning_rate": 0.0001546101961014429,
+      "loss": 0.6357,
+      "step": 3846
+    },
+    {
+      "epoch": 1.2351902392037246,
+      "grad_norm": 0.9770264029502869,
+      "learning_rate": 0.00015458095087537218,
+      "loss": 0.5513,
+      "step": 3847
+    },
+    {
+      "epoch": 1.235511318028576,
+      "grad_norm": 0.8450601696968079,
+      "learning_rate": 0.00015455169899877813,
+      "loss": 0.571,
+      "step": 3848
+    },
+    {
+      "epoch": 1.2358323968534275,
+      "grad_norm": 0.8427115678787231,
+      "learning_rate": 0.00015452244047522502,
+      "loss": 0.637,
+      "step": 3849
+    },
+    {
+      "epoch": 1.236153475678279,
+      "grad_norm": 0.7641344666481018,
+      "learning_rate": 0.00015449317530827794,
+      "loss": 0.564,
+      "step": 3850
+    },
+    {
+      "epoch": 1.2364745545031306,
+      "grad_norm": 1.1168718338012695,
+      "learning_rate": 0.00015446390350150273,
+      "loss": 0.8165,
+      "step": 3851
+    },
+    {
+      "epoch": 1.236795633327982,
+      "grad_norm": 1.0445823669433594,
+      "learning_rate": 0.0001544346250584661,
+      "loss": 0.5067,
+      "step": 3852
+    },
+    {
+      "epoch": 1.2371167121528335,
+      "grad_norm": 0.9036487340927124,
+      "learning_rate": 0.00015440533998273547,
+      "loss": 0.5537,
+      "step": 3853
+    },
+    {
+      "epoch": 1.237437790977685,
+      "grad_norm": 0.7309536933898926,
+      "learning_rate": 0.00015437604827787927,
+      "loss": 0.4998,
+      "step": 3854
+    },
+    {
+      "epoch": 1.2377588698025366,
+      "grad_norm": 0.7521160840988159,
+      "learning_rate": 0.0001543467499474665,
+      "loss": 0.4568,
+      "step": 3855
+    },
+    {
+      "epoch": 1.2380799486273881,
+      "grad_norm": 1.4212712049484253,
+      "learning_rate": 0.00015431744499506706,
+      "loss": 0.6532,
+      "step": 3856
+    },
+    {
+      "epoch": 1.2384010274522395,
+      "grad_norm": 0.758643627166748,
+      "learning_rate": 0.00015428813342425177,
+      "loss": 0.436,
+      "step": 3857
+    },
+    {
+      "epoch": 1.238722106277091,
+      "grad_norm": 0.8695588707923889,
+      "learning_rate": 0.00015425881523859207,
+      "loss": 0.4969,
+      "step": 3858
+    },
+    {
+      "epoch": 1.2390431851019426,
+      "grad_norm": 1.0201971530914307,
+      "learning_rate": 0.0001542294904416603,
+      "loss": 0.5688,
+      "step": 3859
+    },
+    {
+      "epoch": 1.239364263926794,
+      "grad_norm": 0.7259702682495117,
+      "learning_rate": 0.00015420015903702962,
+      "loss": 0.4379,
+      "step": 3860
+    },
+    {
+      "epoch": 1.2396853427516454,
+      "grad_norm": 1.1150838136672974,
+      "learning_rate": 0.000154170821028274,
+      "loss": 0.6326,
+      "step": 3861
+    },
+    {
+      "epoch": 1.240006421576497,
+      "grad_norm": 0.9569064974784851,
+      "learning_rate": 0.00015414147641896813,
+      "loss": 0.4673,
+      "step": 3862
+    },
+    {
+      "epoch": 1.2403275004013485,
+      "grad_norm": 0.6631338000297546,
+      "learning_rate": 0.00015411212521268758,
+      "loss": 0.4138,
+      "step": 3863
+    },
+    {
+      "epoch": 1.2406485792262,
+      "grad_norm": 1.1545629501342773,
+      "learning_rate": 0.00015408276741300873,
+      "loss": 0.4498,
+      "step": 3864
+    },
+    {
+      "epoch": 1.2409696580510516,
+      "grad_norm": 3.294052839279175,
+      "learning_rate": 0.00015405340302350871,
+      "loss": 0.7942,
+      "step": 3865
+    },
+    {
+      "epoch": 1.241290736875903,
+      "grad_norm": 0.7048175930976868,
+      "learning_rate": 0.0001540240320477655,
+      "loss": 0.727,
+      "step": 3866
+    },
+    {
+      "epoch": 1.2416118157007545,
+      "grad_norm": 0.6912996172904968,
+      "learning_rate": 0.00015399465448935788,
+      "loss": 0.7886,
+      "step": 3867
+    },
+    {
+      "epoch": 1.241932894525606,
+      "grad_norm": 0.8170081377029419,
+      "learning_rate": 0.00015396527035186537,
+      "loss": 0.4395,
+      "step": 3868
+    },
+    {
+      "epoch": 1.2422539733504576,
+      "grad_norm": 0.9392108917236328,
+      "learning_rate": 0.00015393587963886835,
+      "loss": 0.3322,
+      "step": 3869
+    },
+    {
+      "epoch": 1.242575052175309,
+      "grad_norm": 1.0908229351043701,
+      "learning_rate": 0.00015390648235394803,
+      "loss": 0.3907,
+      "step": 3870
+    },
+    {
+      "epoch": 1.2428961310001605,
+      "grad_norm": 1.2291988134384155,
+      "learning_rate": 0.0001538770785006863,
+      "loss": 0.5189,
+      "step": 3871
+    },
+    {
+      "epoch": 1.243217209825012,
+      "grad_norm": 0.9118694067001343,
+      "learning_rate": 0.00015384766808266602,
+      "loss": 0.702,
+      "step": 3872
+    },
+    {
+      "epoch": 1.2435382886498636,
+      "grad_norm": 1.1265658140182495,
+      "learning_rate": 0.00015381825110347074,
+      "loss": 0.6291,
+      "step": 3873
+    },
+    {
+      "epoch": 1.2438593674747151,
+      "grad_norm": 0.9491468667984009,
+      "learning_rate": 0.00015378882756668478,
+      "loss": 0.7169,
+      "step": 3874
+    },
+    {
+      "epoch": 1.2441804462995665,
+      "grad_norm": 0.8333759307861328,
+      "learning_rate": 0.00015375939747589335,
+      "loss": 0.5432,
+      "step": 3875
+    },
+    {
+      "epoch": 1.244501525124418,
+      "grad_norm": 0.7121776342391968,
+      "learning_rate": 0.0001537299608346824,
+      "loss": 0.4806,
+      "step": 3876
+    },
+    {
+      "epoch": 1.2448226039492696,
+      "grad_norm": 0.9570683836936951,
+      "learning_rate": 0.0001537005176466387,
+      "loss": 0.6031,
+      "step": 3877
+    },
+    {
+      "epoch": 1.2451436827741211,
+      "grad_norm": 0.8433529734611511,
+      "learning_rate": 0.00015367106791534983,
+      "loss": 0.7327,
+      "step": 3878
+    },
+    {
+      "epoch": 1.2454647615989725,
+      "grad_norm": 0.8592220544815063,
+      "learning_rate": 0.0001536416116444041,
+      "loss": 0.5449,
+      "step": 3879
+    },
+    {
+      "epoch": 1.245785840423824,
+      "grad_norm": 0.9075251221656799,
+      "learning_rate": 0.00015361214883739076,
+      "loss": 0.6982,
+      "step": 3880
+    },
+    {
+      "epoch": 1.2461069192486756,
+      "grad_norm": 1.0835230350494385,
+      "learning_rate": 0.00015358267949789966,
+      "loss": 0.6304,
+      "step": 3881
+    },
+    {
+      "epoch": 1.246427998073527,
+      "grad_norm": 0.9230034947395325,
+      "learning_rate": 0.00015355320362952162,
+      "loss": 0.631,
+      "step": 3882
+    },
+    {
+      "epoch": 1.2467490768983787,
+      "grad_norm": 1.1106213331222534,
+      "learning_rate": 0.00015352372123584814,
+      "loss": 0.7395,
+      "step": 3883
+    },
+    {
+      "epoch": 1.24707015572323,
+      "grad_norm": 1.3075162172317505,
+      "learning_rate": 0.00015349423232047162,
+      "loss": 0.6999,
+      "step": 3884
+    },
+    {
+      "epoch": 1.2473912345480815,
+      "grad_norm": 1.0641080141067505,
+      "learning_rate": 0.00015346473688698513,
+      "loss": 0.6968,
+      "step": 3885
+    },
+    {
+      "epoch": 1.247712313372933,
+      "grad_norm": 0.8962423801422119,
+      "learning_rate": 0.00015343523493898265,
+      "loss": 0.5366,
+      "step": 3886
+    },
+    {
+      "epoch": 1.2480333921977846,
+      "grad_norm": 1.0747464895248413,
+      "learning_rate": 0.00015340572648005888,
+      "loss": 0.5723,
+      "step": 3887
+    },
+    {
+      "epoch": 1.248354471022636,
+      "grad_norm": 1.0071834325790405,
+      "learning_rate": 0.0001533762115138093,
+      "loss": 0.6652,
+      "step": 3888
+    },
+    {
+      "epoch": 1.2486755498474875,
+      "grad_norm": 1.272632360458374,
+      "learning_rate": 0.0001533466900438303,
+      "loss": 0.7782,
+      "step": 3889
+    },
+    {
+      "epoch": 1.248996628672339,
+      "grad_norm": 1.1216950416564941,
+      "learning_rate": 0.00015331716207371888,
+      "loss": 0.6813,
+      "step": 3890
+    },
+    {
+      "epoch": 1.2493177074971906,
+      "grad_norm": 0.9694088697433472,
+      "learning_rate": 0.000153287627607073,
+      "loss": 0.5781,
+      "step": 3891
+    },
+    {
+      "epoch": 1.2496387863220422,
+      "grad_norm": 1.916099190711975,
+      "learning_rate": 0.00015325808664749135,
+      "loss": 0.8175,
+      "step": 3892
+    },
+    {
+      "epoch": 1.2499598651468935,
+      "grad_norm": 0.9841503500938416,
+      "learning_rate": 0.0001532285391985734,
+      "loss": 0.6193,
+      "step": 3893
+    },
+    {
+      "epoch": 1.250280943971745,
+      "grad_norm": 1.1670957803726196,
+      "learning_rate": 0.0001531989852639194,
+      "loss": 0.7807,
+      "step": 3894
+    },
+    {
+      "epoch": 1.2506020227965966,
+      "grad_norm": 1.1303762197494507,
+      "learning_rate": 0.0001531694248471304,
+      "loss": 0.7768,
+      "step": 3895
+    },
+    {
+      "epoch": 1.250923101621448,
+      "grad_norm": 0.9170710444450378,
+      "learning_rate": 0.00015313985795180828,
+      "loss": 0.6192,
+      "step": 3896
+    },
+    {
+      "epoch": 1.2512441804462995,
+      "grad_norm": 1.1706234216690063,
+      "learning_rate": 0.00015311028458155567,
+      "loss": 0.6049,
+      "step": 3897
+    },
+    {
+      "epoch": 1.251565259271151,
+      "grad_norm": 1.188920021057129,
+      "learning_rate": 0.00015308070473997598,
+      "loss": 0.6907,
+      "step": 3898
+    },
+    {
+      "epoch": 1.2518863380960026,
+      "grad_norm": 0.8956981301307678,
+      "learning_rate": 0.0001530511184306734,
+      "loss": 0.5518,
+      "step": 3899
+    },
+    {
+      "epoch": 1.2522074169208541,
+      "grad_norm": 0.978665828704834,
+      "learning_rate": 0.00015302152565725298,
+      "loss": 0.5264,
+      "step": 3900
+    },
+    {
+      "epoch": 1.2525284957457057,
+      "grad_norm": 0.822583794593811,
+      "learning_rate": 0.0001529919264233205,
+      "loss": 0.5521,
+      "step": 3901
+    },
+    {
+      "epoch": 1.252849574570557,
+      "grad_norm": 0.7783530354499817,
+      "learning_rate": 0.00015296232073248251,
+      "loss": 0.4333,
+      "step": 3902
+    },
+    {
+      "epoch": 1.2531706533954086,
+      "grad_norm": 0.7691320776939392,
+      "learning_rate": 0.00015293270858834644,
+      "loss": 0.4828,
+      "step": 3903
+    },
+    {
+      "epoch": 1.2534917322202601,
+      "grad_norm": 1.0390291213989258,
+      "learning_rate": 0.00015290308999452032,
+      "loss": 0.7419,
+      "step": 3904
+    },
+    {
+      "epoch": 1.2538128110451114,
+      "grad_norm": 0.8240483999252319,
+      "learning_rate": 0.00015287346495461315,
+      "loss": 0.5811,
+      "step": 3905
+    },
+    {
+      "epoch": 1.254133889869963,
+      "grad_norm": 1.1478774547576904,
+      "learning_rate": 0.00015284383347223472,
+      "loss": 0.7649,
+      "step": 3906
+    },
+    {
+      "epoch": 1.2544549686948145,
+      "grad_norm": 1.2211809158325195,
+      "learning_rate": 0.00015281419555099546,
+      "loss": 0.6535,
+      "step": 3907
+    },
+    {
+      "epoch": 1.254776047519666,
+      "grad_norm": 1.166593074798584,
+      "learning_rate": 0.00015278455119450664,
+      "loss": 0.6659,
+      "step": 3908
+    },
+    {
+      "epoch": 1.2550971263445176,
+      "grad_norm": 0.8463481068611145,
+      "learning_rate": 0.00015275490040638038,
+      "loss": 0.508,
+      "step": 3909
+    },
+    {
+      "epoch": 1.2554182051693692,
+      "grad_norm": 1.5510456562042236,
+      "learning_rate": 0.00015272524319022955,
+      "loss": 0.3536,
+      "step": 3910
+    },
+    {
+      "epoch": 1.2557392839942205,
+      "grad_norm": 0.9114112854003906,
+      "learning_rate": 0.00015269557954966778,
+      "loss": 0.5686,
+      "step": 3911
+    },
+    {
+      "epoch": 1.256060362819072,
+      "grad_norm": 0.8683449029922485,
+      "learning_rate": 0.00015266590948830947,
+      "loss": 0.4499,
+      "step": 3912
+    },
+    {
+      "epoch": 1.2563814416439236,
+      "grad_norm": 0.9325652122497559,
+      "learning_rate": 0.00015263623300976978,
+      "loss": 0.4684,
+      "step": 3913
+    },
+    {
+      "epoch": 1.256702520468775,
+      "grad_norm": 0.4808101952075958,
+      "learning_rate": 0.00015260655011766484,
+      "loss": 0.3881,
+      "step": 3914
+    },
+    {
+      "epoch": 1.2570235992936265,
+      "grad_norm": 0.9658359289169312,
+      "learning_rate": 0.00015257686081561134,
+      "loss": 0.9497,
+      "step": 3915
+    },
+    {
+      "epoch": 1.257344678118478,
+      "grad_norm": 0.7216698527336121,
+      "learning_rate": 0.00015254716510722677,
+      "loss": 0.7033,
+      "step": 3916
+    },
+    {
+      "epoch": 1.2576657569433296,
+      "grad_norm": 0.9119216799736023,
+      "learning_rate": 0.0001525174629961296,
+      "loss": 0.6078,
+      "step": 3917
+    },
+    {
+      "epoch": 1.2579868357681812,
+      "grad_norm": 0.9573872685432434,
+      "learning_rate": 0.00015248775448593882,
+      "loss": 0.5487,
+      "step": 3918
+    },
+    {
+      "epoch": 1.2583079145930327,
+      "grad_norm": 0.8556592464447021,
+      "learning_rate": 0.00015245803958027434,
+      "loss": 0.3048,
+      "step": 3919
+    },
+    {
+      "epoch": 1.258628993417884,
+      "grad_norm": 0.7970437407493591,
+      "learning_rate": 0.00015242831828275692,
+      "loss": 0.5525,
+      "step": 3920
+    },
+    {
+      "epoch": 1.2589500722427356,
+      "grad_norm": 1.1400879621505737,
+      "learning_rate": 0.00015239859059700794,
+      "loss": 0.7523,
+      "step": 3921
+    },
+    {
+      "epoch": 1.2592711510675871,
+      "grad_norm": 1.1350511312484741,
+      "learning_rate": 0.00015236885652664963,
+      "loss": 0.6163,
+      "step": 3922
+    },
+    {
+      "epoch": 1.2595922298924385,
+      "grad_norm": 1.0047470331192017,
+      "learning_rate": 0.00015233911607530497,
+      "loss": 0.5577,
+      "step": 3923
+    },
+    {
+      "epoch": 1.25991330871729,
+      "grad_norm": 1.0919958353042603,
+      "learning_rate": 0.0001523093692465978,
+      "loss": 0.5031,
+      "step": 3924
+    },
+    {
+      "epoch": 1.2602343875421416,
+      "grad_norm": 0.7224966883659363,
+      "learning_rate": 0.00015227961604415266,
+      "loss": 0.5719,
+      "step": 3925
+    },
+    {
+      "epoch": 1.2605554663669931,
+      "grad_norm": 0.9928240776062012,
+      "learning_rate": 0.0001522498564715949,
+      "loss": 0.6879,
+      "step": 3926
+    },
+    {
+      "epoch": 1.2608765451918447,
+      "grad_norm": 0.8768775463104248,
+      "learning_rate": 0.0001522200905325506,
+      "loss": 0.6634,
+      "step": 3927
+    },
+    {
+      "epoch": 1.2611976240166962,
+      "grad_norm": 0.8131898641586304,
+      "learning_rate": 0.00015219031823064666,
+      "loss": 0.5647,
+      "step": 3928
+    },
+    {
+      "epoch": 1.2615187028415475,
+      "grad_norm": 1.0961014032363892,
+      "learning_rate": 0.0001521605395695108,
+      "loss": 0.6579,
+      "step": 3929
+    },
+    {
+      "epoch": 1.261839781666399,
+      "grad_norm": 0.9365900754928589,
+      "learning_rate": 0.0001521307545527714,
+      "loss": 0.6051,
+      "step": 3930
+    },
+    {
+      "epoch": 1.2621608604912506,
+      "grad_norm": 1.2082931995391846,
+      "learning_rate": 0.00015210096318405767,
+      "loss": 0.7062,
+      "step": 3931
+    },
+    {
+      "epoch": 1.262481939316102,
+      "grad_norm": 0.8848322033882141,
+      "learning_rate": 0.0001520711654669996,
+      "loss": 0.6358,
+      "step": 3932
+    },
+    {
+      "epoch": 1.2628030181409535,
+      "grad_norm": 0.8963106274604797,
+      "learning_rate": 0.00015204136140522799,
+      "loss": 0.5787,
+      "step": 3933
+    },
+    {
+      "epoch": 1.263124096965805,
+      "grad_norm": 1.1440925598144531,
+      "learning_rate": 0.0001520115510023743,
+      "loss": 0.6726,
+      "step": 3934
+    },
+    {
+      "epoch": 1.2634451757906566,
+      "grad_norm": 1.4616795778274536,
+      "learning_rate": 0.00015198173426207094,
+      "loss": 0.8507,
+      "step": 3935
+    },
+    {
+      "epoch": 1.2637662546155082,
+      "grad_norm": 1.1262779235839844,
+      "learning_rate": 0.00015195191118795096,
+      "loss": 0.6682,
+      "step": 3936
+    },
+    {
+      "epoch": 1.2640873334403597,
+      "grad_norm": 0.878693699836731,
+      "learning_rate": 0.00015192208178364816,
+      "loss": 0.6938,
+      "step": 3937
+    },
+    {
+      "epoch": 1.264408412265211,
+      "grad_norm": 1.1921688318252563,
+      "learning_rate": 0.00015189224605279718,
+      "loss": 0.6734,
+      "step": 3938
+    },
+    {
+      "epoch": 1.2647294910900626,
+      "grad_norm": 1.2111103534698486,
+      "learning_rate": 0.00015186240399903342,
+      "loss": 0.6716,
+      "step": 3939
+    },
+    {
+      "epoch": 1.2650505699149142,
+      "grad_norm": 1.1788506507873535,
+      "learning_rate": 0.00015183255562599307,
+      "loss": 0.6973,
+      "step": 3940
+    },
+    {
+      "epoch": 1.2653716487397655,
+      "grad_norm": 1.2426005601882935,
+      "learning_rate": 0.00015180270093731303,
+      "loss": 0.5777,
+      "step": 3941
+    },
+    {
+      "epoch": 1.265692727564617,
+      "grad_norm": 1.2214686870574951,
+      "learning_rate": 0.000151772839936631,
+      "loss": 0.8486,
+      "step": 3942
+    },
+    {
+      "epoch": 1.2660138063894686,
+      "grad_norm": 0.9162314534187317,
+      "learning_rate": 0.0001517429726275855,
+      "loss": 0.4804,
+      "step": 3943
+    },
+    {
+      "epoch": 1.2663348852143201,
+      "grad_norm": 1.108119249343872,
+      "learning_rate": 0.00015171309901381572,
+      "loss": 0.7479,
+      "step": 3944
+    },
+    {
+      "epoch": 1.2666559640391717,
+      "grad_norm": 0.9174817204475403,
+      "learning_rate": 0.00015168321909896172,
+      "loss": 0.6093,
+      "step": 3945
+    },
+    {
+      "epoch": 1.2669770428640232,
+      "grad_norm": 1.142285704612732,
+      "learning_rate": 0.0001516533328866642,
+      "loss": 0.6486,
+      "step": 3946
+    },
+    {
+      "epoch": 1.2672981216888746,
+      "grad_norm": 1.0875240564346313,
+      "learning_rate": 0.00015162344038056476,
+      "loss": 0.7344,
+      "step": 3947
+    },
+    {
+      "epoch": 1.2676192005137261,
+      "grad_norm": 0.7915586233139038,
+      "learning_rate": 0.00015159354158430572,
+      "loss": 0.5285,
+      "step": 3948
+    },
+    {
+      "epoch": 1.2679402793385777,
+      "grad_norm": 1.0847499370574951,
+      "learning_rate": 0.00015156363650153012,
+      "loss": 0.5974,
+      "step": 3949
+    },
+    {
+      "epoch": 1.268261358163429,
+      "grad_norm": 0.7492148876190186,
+      "learning_rate": 0.00015153372513588182,
+      "loss": 0.5093,
+      "step": 3950
+    },
+    {
+      "epoch": 1.2685824369882805,
+      "grad_norm": 1.150233507156372,
+      "learning_rate": 0.00015150380749100545,
+      "loss": 0.6276,
+      "step": 3951
+    },
+    {
+      "epoch": 1.268903515813132,
+      "grad_norm": 0.8965975642204285,
+      "learning_rate": 0.00015147388357054632,
+      "loss": 0.6621,
+      "step": 3952
+    },
+    {
+      "epoch": 1.2692245946379836,
+      "grad_norm": 1.1043106317520142,
+      "learning_rate": 0.00015144395337815064,
+      "loss": 0.6958,
+      "step": 3953
+    },
+    {
+      "epoch": 1.2695456734628352,
+      "grad_norm": 1.1139799356460571,
+      "learning_rate": 0.0001514140169174653,
+      "loss": 0.613,
+      "step": 3954
+    },
+    {
+      "epoch": 1.2698667522876868,
+      "grad_norm": 1.0345196723937988,
+      "learning_rate": 0.00015138407419213796,
+      "loss": 0.5999,
+      "step": 3955
+    },
+    {
+      "epoch": 1.270187831112538,
+      "grad_norm": 1.0387682914733887,
+      "learning_rate": 0.00015135412520581702,
+      "loss": 0.6301,
+      "step": 3956
+    },
+    {
+      "epoch": 1.2705089099373896,
+      "grad_norm": 0.9722599387168884,
+      "learning_rate": 0.0001513241699621517,
+      "loss": 0.6064,
+      "step": 3957
+    },
+    {
+      "epoch": 1.2708299887622412,
+      "grad_norm": 1.886247158050537,
+      "learning_rate": 0.00015129420846479196,
+      "loss": 0.7651,
+      "step": 3958
+    },
+    {
+      "epoch": 1.2711510675870925,
+      "grad_norm": 1.071523666381836,
+      "learning_rate": 0.0001512642407173885,
+      "loss": 0.5445,
+      "step": 3959
+    },
+    {
+      "epoch": 1.271472146411944,
+      "grad_norm": 0.7643309235572815,
+      "learning_rate": 0.00015123426672359285,
+      "loss": 0.3537,
+      "step": 3960
+    },
+    {
+      "epoch": 1.2717932252367956,
+      "grad_norm": 1.0264235734939575,
+      "learning_rate": 0.00015120428648705717,
+      "loss": 0.6581,
+      "step": 3961
+    },
+    {
+      "epoch": 1.2721143040616472,
+      "grad_norm": 1.0826537609100342,
+      "learning_rate": 0.00015117430001143452,
+      "loss": 0.6621,
+      "step": 3962
+    },
+    {
+      "epoch": 1.2724353828864987,
+      "grad_norm": 0.5251657366752625,
+      "learning_rate": 0.0001511443073003786,
+      "loss": 0.3996,
+      "step": 3963
+    },
+    {
+      "epoch": 1.2727564617113503,
+      "grad_norm": 0.6575037240982056,
+      "learning_rate": 0.000151114308357544,
+      "loss": 0.3963,
+      "step": 3964
+    },
+    {
+      "epoch": 1.2730775405362016,
+      "grad_norm": 0.7141722440719604,
+      "learning_rate": 0.000151084303186586,
+      "loss": 0.6904,
+      "step": 3965
+    },
+    {
+      "epoch": 1.2733986193610531,
+      "grad_norm": 0.8512976169586182,
+      "learning_rate": 0.0001510542917911606,
+      "loss": 0.999,
+      "step": 3966
+    },
+    {
+      "epoch": 1.2737196981859047,
+      "grad_norm": 0.7313768863677979,
+      "learning_rate": 0.0001510242741749246,
+      "loss": 1.1832,
+      "step": 3967
+    },
+    {
+      "epoch": 1.274040777010756,
+      "grad_norm": 0.7449411153793335,
+      "learning_rate": 0.00015099425034153553,
+      "loss": 0.6666,
+      "step": 3968
+    },
+    {
+      "epoch": 1.2743618558356076,
+      "grad_norm": 0.7838384509086609,
+      "learning_rate": 0.00015096422029465178,
+      "loss": 0.4504,
+      "step": 3969
+    },
+    {
+      "epoch": 1.2746829346604591,
+      "grad_norm": 0.6540324091911316,
+      "learning_rate": 0.00015093418403793238,
+      "loss": 0.3508,
+      "step": 3970
+    },
+    {
+      "epoch": 1.2750040134853107,
+      "grad_norm": 0.8694151043891907,
+      "learning_rate": 0.00015090414157503714,
+      "loss": 0.3151,
+      "step": 3971
+    },
+    {
+      "epoch": 1.2753250923101622,
+      "grad_norm": 1.0265659093856812,
+      "learning_rate": 0.00015087409290962667,
+      "loss": 0.3798,
+      "step": 3972
+    },
+    {
+      "epoch": 1.2756461711350138,
+      "grad_norm": 0.7983294129371643,
+      "learning_rate": 0.0001508440380453623,
+      "loss": 0.7452,
+      "step": 3973
+    },
+    {
+      "epoch": 1.275967249959865,
+      "grad_norm": 1.3019341230392456,
+      "learning_rate": 0.0001508139769859061,
+      "loss": 0.6089,
+      "step": 3974
+    },
+    {
+      "epoch": 1.2762883287847167,
+      "grad_norm": 0.756151556968689,
+      "learning_rate": 0.00015078390973492092,
+      "loss": 0.6198,
+      "step": 3975
+    },
+    {
+      "epoch": 1.2766094076095682,
+      "grad_norm": 0.797984778881073,
+      "learning_rate": 0.00015075383629607042,
+      "loss": 0.5604,
+      "step": 3976
+    },
+    {
+      "epoch": 1.2769304864344195,
+      "grad_norm": 0.7493342161178589,
+      "learning_rate": 0.00015072375667301893,
+      "loss": 0.6128,
+      "step": 3977
+    },
+    {
+      "epoch": 1.277251565259271,
+      "grad_norm": 0.8932711482048035,
+      "learning_rate": 0.00015069367086943154,
+      "loss": 0.5593,
+      "step": 3978
+    },
+    {
+      "epoch": 1.2775726440841226,
+      "grad_norm": 0.9008491039276123,
+      "learning_rate": 0.0001506635788889741,
+      "loss": 0.4842,
+      "step": 3979
+    },
+    {
+      "epoch": 1.2778937229089742,
+      "grad_norm": 0.8058199882507324,
+      "learning_rate": 0.00015063348073531324,
+      "loss": 0.6554,
+      "step": 3980
+    },
+    {
+      "epoch": 1.2782148017338257,
+      "grad_norm": 1.0209277868270874,
+      "learning_rate": 0.00015060337641211637,
+      "loss": 0.5969,
+      "step": 3981
+    },
+    {
+      "epoch": 1.2785358805586773,
+      "grad_norm": 1.4035515785217285,
+      "learning_rate": 0.0001505732659230516,
+      "loss": 0.7034,
+      "step": 3982
+    },
+    {
+      "epoch": 1.2788569593835286,
+      "grad_norm": 0.963071882724762,
+      "learning_rate": 0.0001505431492717878,
+      "loss": 0.5722,
+      "step": 3983
+    },
+    {
+      "epoch": 1.2791780382083802,
+      "grad_norm": 1.2736893892288208,
+      "learning_rate": 0.0001505130264619945,
+      "loss": 0.6295,
+      "step": 3984
+    },
+    {
+      "epoch": 1.2794991170332317,
+      "grad_norm": 0.9955981373786926,
+      "learning_rate": 0.0001504828974973422,
+      "loss": 0.7137,
+      "step": 3985
+    },
+    {
+      "epoch": 1.279820195858083,
+      "grad_norm": 1.3226053714752197,
+      "learning_rate": 0.00015045276238150192,
+      "loss": 0.7224,
+      "step": 3986
+    },
+    {
+      "epoch": 1.2801412746829346,
+      "grad_norm": 1.3493560552597046,
+      "learning_rate": 0.00015042262111814565,
+      "loss": 0.932,
+      "step": 3987
+    },
+    {
+      "epoch": 1.2804623535077861,
+      "grad_norm": 0.8732056617736816,
+      "learning_rate": 0.00015039247371094588,
+      "loss": 0.5749,
+      "step": 3988
+    },
+    {
+      "epoch": 1.2807834323326377,
+      "grad_norm": 0.9701531529426575,
+      "learning_rate": 0.0001503623201635761,
+      "loss": 0.6856,
+      "step": 3989
+    },
+    {
+      "epoch": 1.2811045111574892,
+      "grad_norm": 1.385581135749817,
+      "learning_rate": 0.00015033216047971031,
+      "loss": 0.7327,
+      "step": 3990
+    },
+    {
+      "epoch": 1.2814255899823408,
+      "grad_norm": 0.7699181437492371,
+      "learning_rate": 0.00015030199466302353,
+      "loss": 0.5799,
+      "step": 3991
+    },
+    {
+      "epoch": 1.2817466688071921,
+      "grad_norm": 1.5015044212341309,
+      "learning_rate": 0.00015027182271719122,
+      "loss": 0.8638,
+      "step": 3992
+    },
+    {
+      "epoch": 1.2820677476320437,
+      "grad_norm": 0.9158933758735657,
+      "learning_rate": 0.00015024164464588982,
+      "loss": 0.7239,
+      "step": 3993
+    },
+    {
+      "epoch": 1.2823888264568952,
+      "grad_norm": 1.1398961544036865,
+      "learning_rate": 0.0001502114604527964,
+      "loss": 0.7259,
+      "step": 3994
+    },
+    {
+      "epoch": 1.2827099052817466,
+      "grad_norm": 0.9500359892845154,
+      "learning_rate": 0.00015018127014158886,
+      "loss": 0.5439,
+      "step": 3995
+    },
+    {
+      "epoch": 1.283030984106598,
+      "grad_norm": 0.9452093243598938,
+      "learning_rate": 0.00015015107371594573,
+      "loss": 0.617,
+      "step": 3996
+    },
+    {
+      "epoch": 1.2833520629314497,
+      "grad_norm": 0.881150484085083,
+      "learning_rate": 0.00015012087117954642,
+      "loss": 0.4604,
+      "step": 3997
+    },
+    {
+      "epoch": 1.2836731417563012,
+      "grad_norm": 1.126312494277954,
+      "learning_rate": 0.000150090662536071,
+      "loss": 0.8014,
+      "step": 3998
+    },
+    {
+      "epoch": 1.2839942205811528,
+      "grad_norm": 1.0232257843017578,
+      "learning_rate": 0.0001500604477892003,
+      "loss": 0.7495,
+      "step": 3999
+    },
+    {
+      "epoch": 1.2843152994060043,
+      "grad_norm": 0.9484618902206421,
+      "learning_rate": 0.00015003022694261585,
+      "loss": 0.6574,
+      "step": 4000
+    },
+    {
+      "epoch": 1.2846363782308556,
+      "grad_norm": 0.8673555850982666,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.6955,
+      "step": 4001
+    },
+    {
+      "epoch": 1.2849574570557072,
+      "grad_norm": 0.842471718788147,
+      "learning_rate": 0.00014996976696503587,
+      "loss": 0.5209,
+      "step": 4002
+    },
+    {
+      "epoch": 1.2852785358805587,
+      "grad_norm": 0.6719858050346375,
+      "learning_rate": 0.00014993952784140717,
+      "loss": 0.4186,
+      "step": 4003
+    },
+    {
+      "epoch": 1.28559961470541,
+      "grad_norm": 1.7661453485488892,
+      "learning_rate": 0.00014990928263279848,
+      "loss": 0.5979,
+      "step": 4004
+    },
+    {
+      "epoch": 1.2859206935302616,
+      "grad_norm": 1.1179763078689575,
+      "learning_rate": 0.00014987903134289508,
+      "loss": 0.5886,
+      "step": 4005
+    },
+    {
+      "epoch": 1.2862417723551132,
+      "grad_norm": 0.9625594615936279,
+      "learning_rate": 0.00014984877397538303,
+      "loss": 0.5743,
+      "step": 4006
+    },
+    {
+      "epoch": 1.2865628511799647,
+      "grad_norm": 1.0250779390335083,
+      "learning_rate": 0.0001498185105339491,
+      "loss": 0.5355,
+      "step": 4007
+    },
+    {
+      "epoch": 1.2868839300048163,
+      "grad_norm": 1.0908957719802856,
+      "learning_rate": 0.00014978824102228076,
+      "loss": 0.5967,
+      "step": 4008
+    },
+    {
+      "epoch": 1.2872050088296678,
+      "grad_norm": 0.8703522086143494,
+      "learning_rate": 0.00014975796544406625,
+      "loss": 0.5883,
+      "step": 4009
+    },
+    {
+      "epoch": 1.2875260876545191,
+      "grad_norm": 0.976186990737915,
+      "learning_rate": 0.0001497276838029946,
+      "loss": 0.5918,
+      "step": 4010
+    },
+    {
+      "epoch": 1.2878471664793707,
+      "grad_norm": 1.2151916027069092,
+      "learning_rate": 0.00014969739610275556,
+      "loss": 0.4542,
+      "step": 4011
+    },
+    {
+      "epoch": 1.2881682453042222,
+      "grad_norm": 0.8926727175712585,
+      "learning_rate": 0.0001496671023470395,
+      "loss": 0.5784,
+      "step": 4012
+    },
+    {
+      "epoch": 1.2884893241290736,
+      "grad_norm": 1.0263932943344116,
+      "learning_rate": 0.0001496368025395377,
+      "loss": 0.5809,
+      "step": 4013
+    },
+    {
+      "epoch": 1.2888104029539251,
+      "grad_norm": 1.3325320482254028,
+      "learning_rate": 0.00014960649668394207,
+      "loss": 0.4552,
+      "step": 4014
+    },
+    {
+      "epoch": 1.2891314817787767,
+      "grad_norm": 1.0666171312332153,
+      "learning_rate": 0.00014957618478394529,
+      "loss": 1.0576,
+      "step": 4015
+    },
+    {
+      "epoch": 1.2894525606036282,
+      "grad_norm": 0.7908462285995483,
+      "learning_rate": 0.00014954586684324078,
+      "loss": 1.0942,
+      "step": 4016
+    },
+    {
+      "epoch": 1.2897736394284798,
+      "grad_norm": 0.7149842381477356,
+      "learning_rate": 0.00014951554286552266,
+      "loss": 0.8738,
+      "step": 4017
+    },
+    {
+      "epoch": 1.2900947182533313,
+      "grad_norm": 0.8624144196510315,
+      "learning_rate": 0.00014948521285448586,
+      "loss": 0.9059,
+      "step": 4018
+    },
+    {
+      "epoch": 1.2904157970781827,
+      "grad_norm": 0.8445031642913818,
+      "learning_rate": 0.00014945487681382598,
+      "loss": 0.6081,
+      "step": 4019
+    },
+    {
+      "epoch": 1.2907368759030342,
+      "grad_norm": 0.8286342620849609,
+      "learning_rate": 0.00014942453474723935,
+      "loss": 0.4396,
+      "step": 4020
+    },
+    {
+      "epoch": 1.2910579547278858,
+      "grad_norm": 0.7982895374298096,
+      "learning_rate": 0.0001493941866584231,
+      "loss": 0.3624,
+      "step": 4021
+    },
+    {
+      "epoch": 1.291379033552737,
+      "grad_norm": 0.7270027995109558,
+      "learning_rate": 0.00014936383255107505,
+      "loss": 0.3002,
+      "step": 4022
+    },
+    {
+      "epoch": 1.2917001123775886,
+      "grad_norm": 0.838603138923645,
+      "learning_rate": 0.0001493334724288937,
+      "loss": 0.7561,
+      "step": 4023
+    },
+    {
+      "epoch": 1.2920211912024402,
+      "grad_norm": 1.2956212759017944,
+      "learning_rate": 0.0001493031062955784,
+      "loss": 0.6119,
+      "step": 4024
+    },
+    {
+      "epoch": 1.2923422700272917,
+      "grad_norm": 0.8435970544815063,
+      "learning_rate": 0.00014927273415482915,
+      "loss": 0.5659,
+      "step": 4025
+    },
+    {
+      "epoch": 1.2926633488521433,
+      "grad_norm": 0.8424569368362427,
+      "learning_rate": 0.00014924235601034672,
+      "loss": 0.5801,
+      "step": 4026
+    },
+    {
+      "epoch": 1.2929844276769948,
+      "grad_norm": 1.0110530853271484,
+      "learning_rate": 0.00014921197186583255,
+      "loss": 0.6738,
+      "step": 4027
+    },
+    {
+      "epoch": 1.2933055065018462,
+      "grad_norm": 0.7541561126708984,
+      "learning_rate": 0.0001491815817249889,
+      "loss": 0.5595,
+      "step": 4028
+    },
+    {
+      "epoch": 1.2936265853266977,
+      "grad_norm": 0.9687687158584595,
+      "learning_rate": 0.0001491511855915187,
+      "loss": 0.6474,
+      "step": 4029
+    },
+    {
+      "epoch": 1.2939476641515493,
+      "grad_norm": 0.9450034499168396,
+      "learning_rate": 0.00014912078346912563,
+      "loss": 0.5455,
+      "step": 4030
+    },
+    {
+      "epoch": 1.2942687429764006,
+      "grad_norm": 0.8591203689575195,
+      "learning_rate": 0.00014909037536151409,
+      "loss": 0.5695,
+      "step": 4031
+    },
+    {
+      "epoch": 1.2945898218012521,
+      "grad_norm": 1.0376840829849243,
+      "learning_rate": 0.0001490599612723892,
+      "loss": 0.7787,
+      "step": 4032
+    },
+    {
+      "epoch": 1.2949109006261037,
+      "grad_norm": 0.9652031064033508,
+      "learning_rate": 0.00014902954120545687,
+      "loss": 0.6709,
+      "step": 4033
+    },
+    {
+      "epoch": 1.2952319794509553,
+      "grad_norm": 0.8163144588470459,
+      "learning_rate": 0.00014899911516442365,
+      "loss": 0.6559,
+      "step": 4034
+    },
+    {
+      "epoch": 1.2955530582758068,
+      "grad_norm": 1.0613020658493042,
+      "learning_rate": 0.00014896868315299693,
+      "loss": 0.5966,
+      "step": 4035
+    },
+    {
+      "epoch": 1.2958741371006581,
+      "grad_norm": 0.9142425656318665,
+      "learning_rate": 0.00014893824517488464,
+      "loss": 0.6798,
+      "step": 4036
+    },
+    {
+      "epoch": 1.2961952159255097,
+      "grad_norm": 0.8986964225769043,
+      "learning_rate": 0.00014890780123379564,
+      "loss": 0.7447,
+      "step": 4037
+    },
+    {
+      "epoch": 1.2965162947503612,
+      "grad_norm": 1.133459448814392,
+      "learning_rate": 0.0001488773513334394,
+      "loss": 0.567,
+      "step": 4038
+    },
+    {
+      "epoch": 1.2968373735752128,
+      "grad_norm": 1.0823416709899902,
+      "learning_rate": 0.0001488468954775262,
+      "loss": 0.7199,
+      "step": 4039
+    },
+    {
+      "epoch": 1.297158452400064,
+      "grad_norm": 0.8796799182891846,
+      "learning_rate": 0.00014881643366976692,
+      "loss": 0.5962,
+      "step": 4040
+    },
+    {
+      "epoch": 1.2974795312249157,
+      "grad_norm": 1.0232266187667847,
+      "learning_rate": 0.0001487859659138733,
+      "loss": 0.7782,
+      "step": 4041
+    },
+    {
+      "epoch": 1.2978006100497672,
+      "grad_norm": 1.0571743249893188,
+      "learning_rate": 0.00014875549221355768,
+      "loss": 0.6525,
+      "step": 4042
+    },
+    {
+      "epoch": 1.2981216888746188,
+      "grad_norm": 1.7555642127990723,
+      "learning_rate": 0.00014872501257253323,
+      "loss": 0.8595,
+      "step": 4043
+    },
+    {
+      "epoch": 1.2984427676994703,
+      "grad_norm": 1.0583701133728027,
+      "learning_rate": 0.00014869452699451383,
+      "loss": 0.6666,
+      "step": 4044
+    },
+    {
+      "epoch": 1.2987638465243216,
+      "grad_norm": 1.1438575983047485,
+      "learning_rate": 0.000148664035483214,
+      "loss": 0.8125,
+      "step": 4045
+    },
+    {
+      "epoch": 1.2990849253491732,
+      "grad_norm": 1.0455180406570435,
+      "learning_rate": 0.00014863353804234905,
+      "loss": 0.5961,
+      "step": 4046
+    },
+    {
+      "epoch": 1.2994060041740247,
+      "grad_norm": 1.0730552673339844,
+      "learning_rate": 0.00014860303467563503,
+      "loss": 0.4922,
+      "step": 4047
+    },
+    {
+      "epoch": 1.2997270829988763,
+      "grad_norm": 1.0210821628570557,
+      "learning_rate": 0.00014857252538678865,
+      "loss": 0.6711,
+      "step": 4048
+    },
+    {
+      "epoch": 1.3000481618237276,
+      "grad_norm": 1.0794029235839844,
+      "learning_rate": 0.0001485420101795274,
+      "loss": 0.6992,
+      "step": 4049
+    },
+    {
+      "epoch": 1.3003692406485792,
+      "grad_norm": 1.1778931617736816,
+      "learning_rate": 0.00014851148905756947,
+      "loss": 0.5997,
+      "step": 4050
+    },
+    {
+      "epoch": 1.3006903194734307,
+      "grad_norm": 1.5340932607650757,
+      "learning_rate": 0.00014848096202463372,
+      "loss": 0.4667,
+      "step": 4051
+    },
+    {
+      "epoch": 1.3010113982982823,
+      "grad_norm": 1.405383825302124,
+      "learning_rate": 0.0001484504290844398,
+      "loss": 0.7285,
+      "step": 4052
+    },
+    {
+      "epoch": 1.3013324771231338,
+      "grad_norm": 0.9958581328392029,
+      "learning_rate": 0.00014841989024070809,
+      "loss": 0.651,
+      "step": 4053
+    },
+    {
+      "epoch": 1.3016535559479852,
+      "grad_norm": 1.1035085916519165,
+      "learning_rate": 0.00014838934549715963,
+      "loss": 0.6727,
+      "step": 4054
+    },
+    {
+      "epoch": 1.3019746347728367,
+      "grad_norm": 0.9837401509284973,
+      "learning_rate": 0.00014835879485751617,
+      "loss": 0.6048,
+      "step": 4055
+    },
+    {
+      "epoch": 1.3022957135976883,
+      "grad_norm": 1.033929467201233,
+      "learning_rate": 0.00014832823832550024,
+      "loss": 0.709,
+      "step": 4056
+    },
+    {
+      "epoch": 1.3026167924225398,
+      "grad_norm": 1.9119707345962524,
+      "learning_rate": 0.00014829767590483506,
+      "loss": 0.7885,
+      "step": 4057
+    },
+    {
+      "epoch": 1.3029378712473911,
+      "grad_norm": 1.1973304748535156,
+      "learning_rate": 0.0001482671075992446,
+      "loss": 0.6681,
+      "step": 4058
+    },
+    {
+      "epoch": 1.3032589500722427,
+      "grad_norm": 0.8669224977493286,
+      "learning_rate": 0.00014823653341245353,
+      "loss": 0.5783,
+      "step": 4059
+    },
+    {
+      "epoch": 1.3035800288970942,
+      "grad_norm": 1.0361067056655884,
+      "learning_rate": 0.00014820595334818712,
+      "loss": 0.7104,
+      "step": 4060
+    },
+    {
+      "epoch": 1.3039011077219458,
+      "grad_norm": 0.9638859033584595,
+      "learning_rate": 0.00014817536741017152,
+      "loss": 0.609,
+      "step": 4061
+    },
+    {
+      "epoch": 1.3042221865467973,
+      "grad_norm": 0.9824711084365845,
+      "learning_rate": 0.00014814477560213358,
+      "loss": 0.526,
+      "step": 4062
+    },
+    {
+      "epoch": 1.3045432653716487,
+      "grad_norm": 0.6022173762321472,
+      "learning_rate": 0.00014811417792780075,
+      "loss": 0.4001,
+      "step": 4063
+    },
+    {
+      "epoch": 1.3048643441965002,
+      "grad_norm": 0.7990144491195679,
+      "learning_rate": 0.00014808357439090127,
+      "loss": 0.4216,
+      "step": 4064
+    },
+    {
+      "epoch": 1.3051854230213518,
+      "grad_norm": 0.676643431186676,
+      "learning_rate": 0.00014805296499516407,
+      "loss": 0.7422,
+      "step": 4065
+    },
+    {
+      "epoch": 1.3055065018462033,
+      "grad_norm": 0.8250537514686584,
+      "learning_rate": 0.0001480223497443189,
+      "loss": 1.0526,
+      "step": 4066
+    },
+    {
+      "epoch": 1.3058275806710546,
+      "grad_norm": 1.0050231218338013,
+      "learning_rate": 0.00014799172864209608,
+      "loss": 0.6552,
+      "step": 4067
+    },
+    {
+      "epoch": 1.3061486594959062,
+      "grad_norm": 0.8684215545654297,
+      "learning_rate": 0.00014796110169222666,
+      "loss": 0.4027,
+      "step": 4068
+    },
+    {
+      "epoch": 1.3064697383207577,
+      "grad_norm": 0.89919513463974,
+      "learning_rate": 0.0001479304688984425,
+      "loss": 0.4916,
+      "step": 4069
+    },
+    {
+      "epoch": 1.3067908171456093,
+      "grad_norm": 0.8932663798332214,
+      "learning_rate": 0.00014789983026447612,
+      "loss": 0.2584,
+      "step": 4070
+    },
+    {
+      "epoch": 1.3071118959704608,
+      "grad_norm": 0.8465712666511536,
+      "learning_rate": 0.0001478691857940607,
+      "loss": 0.3093,
+      "step": 4071
+    },
+    {
+      "epoch": 1.3074329747953122,
+      "grad_norm": 0.7077059149742126,
+      "learning_rate": 0.00014783853549093018,
+      "loss": 0.4673,
+      "step": 4072
+    },
+    {
+      "epoch": 1.3077540536201637,
+      "grad_norm": 1.0644139051437378,
+      "learning_rate": 0.00014780787935881923,
+      "loss": 0.7199,
+      "step": 4073
+    },
+    {
+      "epoch": 1.3080751324450153,
+      "grad_norm": 0.7375410795211792,
+      "learning_rate": 0.0001477772174014632,
+      "loss": 0.5636,
+      "step": 4074
+    },
+    {
+      "epoch": 1.3083962112698668,
+      "grad_norm": 0.9191805720329285,
+      "learning_rate": 0.00014774654962259812,
+      "loss": 0.663,
+      "step": 4075
+    },
+    {
+      "epoch": 1.3087172900947182,
+      "grad_norm": 0.9933133721351624,
+      "learning_rate": 0.00014771587602596084,
+      "loss": 0.6267,
+      "step": 4076
+    },
+    {
+      "epoch": 1.3090383689195697,
+      "grad_norm": 0.7604119777679443,
+      "learning_rate": 0.0001476851966152888,
+      "loss": 0.5479,
+      "step": 4077
+    },
+    {
+      "epoch": 1.3093594477444213,
+      "grad_norm": 0.8366871476173401,
+      "learning_rate": 0.0001476545113943202,
+      "loss": 0.643,
+      "step": 4078
+    },
+    {
+      "epoch": 1.3096805265692728,
+      "grad_norm": 0.8511562943458557,
+      "learning_rate": 0.0001476238203667939,
+      "loss": 0.5903,
+      "step": 4079
+    },
+    {
+      "epoch": 1.3100016053941244,
+      "grad_norm": 0.9097726941108704,
+      "learning_rate": 0.0001475931235364496,
+      "loss": 0.6136,
+      "step": 4080
+    },
+    {
+      "epoch": 1.3103226842189757,
+      "grad_norm": 1.0176063776016235,
+      "learning_rate": 0.00014756242090702756,
+      "loss": 0.6633,
+      "step": 4081
+    },
+    {
+      "epoch": 1.3106437630438272,
+      "grad_norm": 1.1305146217346191,
+      "learning_rate": 0.00014753171248226875,
+      "loss": 0.7559,
+      "step": 4082
+    },
+    {
+      "epoch": 1.3109648418686788,
+      "grad_norm": 0.7558575868606567,
+      "learning_rate": 0.00014750099826591498,
+      "loss": 0.5253,
+      "step": 4083
+    },
+    {
+      "epoch": 1.3112859206935303,
+      "grad_norm": 0.850307047367096,
+      "learning_rate": 0.00014747027826170867,
+      "loss": 0.6287,
+      "step": 4084
+    },
+    {
+      "epoch": 1.3116069995183817,
+      "grad_norm": 1.0060899257659912,
+      "learning_rate": 0.00014743955247339293,
+      "loss": 0.7239,
+      "step": 4085
+    },
+    {
+      "epoch": 1.3119280783432332,
+      "grad_norm": 0.9706814885139465,
+      "learning_rate": 0.0001474088209047116,
+      "loss": 0.6527,
+      "step": 4086
+    },
+    {
+      "epoch": 1.3122491571680848,
+      "grad_norm": 0.847439706325531,
+      "learning_rate": 0.00014737808355940932,
+      "loss": 0.6468,
+      "step": 4087
+    },
+    {
+      "epoch": 1.3125702359929363,
+      "grad_norm": 0.9411894679069519,
+      "learning_rate": 0.0001473473404412312,
+      "loss": 0.6667,
+      "step": 4088
+    },
+    {
+      "epoch": 1.3128913148177879,
+      "grad_norm": 1.1132397651672363,
+      "learning_rate": 0.00014731659155392332,
+      "loss": 0.5735,
+      "step": 4089
+    },
+    {
+      "epoch": 1.3132123936426392,
+      "grad_norm": 1.0457525253295898,
+      "learning_rate": 0.00014728583690123224,
+      "loss": 0.6922,
+      "step": 4090
+    },
+    {
+      "epoch": 1.3135334724674907,
+      "grad_norm": 0.8525384068489075,
+      "learning_rate": 0.00014725507648690543,
+      "loss": 0.634,
+      "step": 4091
+    },
+    {
+      "epoch": 1.3138545512923423,
+      "grad_norm": 0.7392247319221497,
+      "learning_rate": 0.00014722431031469083,
+      "loss": 0.491,
+      "step": 4092
+    },
+    {
+      "epoch": 1.3141756301171938,
+      "grad_norm": 1.0142507553100586,
+      "learning_rate": 0.0001471935383883373,
+      "loss": 0.7623,
+      "step": 4093
+    },
+    {
+      "epoch": 1.3144967089420452,
+      "grad_norm": 1.3004919290542603,
+      "learning_rate": 0.00014716276071159422,
+      "loss": 0.7285,
+      "step": 4094
+    },
+    {
+      "epoch": 1.3148177877668967,
+      "grad_norm": 1.129610538482666,
+      "learning_rate": 0.00014713197728821183,
+      "loss": 0.7317,
+      "step": 4095
+    },
+    {
+      "epoch": 1.3151388665917483,
+      "grad_norm": 0.9293155670166016,
+      "learning_rate": 0.000147101188121941,
+      "loss": 0.6089,
+      "step": 4096
+    },
+    {
+      "epoch": 1.3154599454165998,
+      "grad_norm": 0.9089670777320862,
+      "learning_rate": 0.0001470703932165333,
+      "loss": 0.6274,
+      "step": 4097
+    },
+    {
+      "epoch": 1.3157810242414514,
+      "grad_norm": 1.2431045770645142,
+      "learning_rate": 0.0001470395925757409,
+      "loss": 0.6852,
+      "step": 4098
+    },
+    {
+      "epoch": 1.3161021030663027,
+      "grad_norm": 0.9798834919929504,
+      "learning_rate": 0.00014700878620331684,
+      "loss": 0.7175,
+      "step": 4099
+    },
+    {
+      "epoch": 1.3164231818911543,
+      "grad_norm": 0.9327998757362366,
+      "learning_rate": 0.0001469779741030148,
+      "loss": 0.6003,
+      "step": 4100
+    },
+    {
+      "epoch": 1.3167442607160058,
+      "grad_norm": 1.2201324701309204,
+      "learning_rate": 0.00014694715627858908,
+      "loss": 0.5154,
+      "step": 4101
+    },
+    {
+      "epoch": 1.3170653395408574,
+      "grad_norm": 0.912707507610321,
+      "learning_rate": 0.0001469163327337948,
+      "loss": 0.4929,
+      "step": 4102
+    },
+    {
+      "epoch": 1.3173864183657087,
+      "grad_norm": 1.2306265830993652,
+      "learning_rate": 0.0001468855034723877,
+      "loss": 0.6194,
+      "step": 4103
+    },
+    {
+      "epoch": 1.3177074971905602,
+      "grad_norm": 1.114705204963684,
+      "learning_rate": 0.00014685466849812418,
+      "loss": 0.6703,
+      "step": 4104
+    },
+    {
+      "epoch": 1.3180285760154118,
+      "grad_norm": 0.8249489068984985,
+      "learning_rate": 0.00014682382781476146,
+      "loss": 0.533,
+      "step": 4105
+    },
+    {
+      "epoch": 1.3183496548402633,
+      "grad_norm": 1.0677366256713867,
+      "learning_rate": 0.00014679298142605734,
+      "loss": 0.6168,
+      "step": 4106
+    },
+    {
+      "epoch": 1.318670733665115,
+      "grad_norm": 0.7703533172607422,
+      "learning_rate": 0.0001467621293357704,
+      "loss": 0.4699,
+      "step": 4107
+    },
+    {
+      "epoch": 1.3189918124899662,
+      "grad_norm": 1.1350317001342773,
+      "learning_rate": 0.0001467312715476598,
+      "loss": 0.5877,
+      "step": 4108
+    },
+    {
+      "epoch": 1.3193128913148178,
+      "grad_norm": 0.9653196930885315,
+      "learning_rate": 0.00014670040806548555,
+      "loss": 0.5712,
+      "step": 4109
+    },
+    {
+      "epoch": 1.3196339701396693,
+      "grad_norm": 1.1715449094772339,
+      "learning_rate": 0.0001466695388930082,
+      "loss": 0.5182,
+      "step": 4110
+    },
+    {
+      "epoch": 1.3199550489645209,
+      "grad_norm": 0.7922108769416809,
+      "learning_rate": 0.00014663866403398913,
+      "loss": 0.5427,
+      "step": 4111
+    },
+    {
+      "epoch": 1.3202761277893722,
+      "grad_norm": 0.9558143615722656,
+      "learning_rate": 0.0001466077834921903,
+      "loss": 0.5072,
+      "step": 4112
+    },
+    {
+      "epoch": 1.3205972066142238,
+      "grad_norm": 0.6734508872032166,
+      "learning_rate": 0.00014657689727137443,
+      "loss": 0.4252,
+      "step": 4113
+    },
+    {
+      "epoch": 1.3209182854390753,
+      "grad_norm": 0.826819658279419,
+      "learning_rate": 0.0001465460053753049,
+      "loss": 0.4521,
+      "step": 4114
+    },
+    {
+      "epoch": 1.3212393642639269,
+      "grad_norm": 0.8778970241546631,
+      "learning_rate": 0.00014651510780774583,
+      "loss": 0.5193,
+      "step": 4115
+    },
+    {
+      "epoch": 1.3215604430887784,
+      "grad_norm": 0.7205748558044434,
+      "learning_rate": 0.000146484204572462,
+      "loss": 0.8583,
+      "step": 4116
+    },
+    {
+      "epoch": 1.3218815219136297,
+      "grad_norm": 0.5651400685310364,
+      "learning_rate": 0.0001464532956732188,
+      "loss": 0.4929,
+      "step": 4117
+    },
+    {
+      "epoch": 1.3222026007384813,
+      "grad_norm": 0.8473796248435974,
+      "learning_rate": 0.0001464223811137824,
+      "loss": 0.5065,
+      "step": 4118
+    },
+    {
+      "epoch": 1.3225236795633328,
+      "grad_norm": 0.8395222425460815,
+      "learning_rate": 0.0001463914608979197,
+      "loss": 0.63,
+      "step": 4119
+    },
+    {
+      "epoch": 1.3228447583881842,
+      "grad_norm": 0.8078770041465759,
+      "learning_rate": 0.00014636053502939823,
+      "loss": 0.3171,
+      "step": 4120
+    },
+    {
+      "epoch": 1.3231658372130357,
+      "grad_norm": 0.7320200204849243,
+      "learning_rate": 0.00014632960351198618,
+      "loss": 0.4024,
+      "step": 4121
+    },
+    {
+      "epoch": 1.3234869160378873,
+      "grad_norm": 0.9218945503234863,
+      "learning_rate": 0.00014629866634945248,
+      "loss": 0.3652,
+      "step": 4122
+    },
+    {
+      "epoch": 1.3238079948627388,
+      "grad_norm": 0.6042816042900085,
+      "learning_rate": 0.0001462677235455667,
+      "loss": 0.2944,
+      "step": 4123
+    },
+    {
+      "epoch": 1.3241290736875904,
+      "grad_norm": 1.0619382858276367,
+      "learning_rate": 0.00014623677510409918,
+      "loss": 0.7914,
+      "step": 4124
+    },
+    {
+      "epoch": 1.324450152512442,
+      "grad_norm": 1.0162999629974365,
+      "learning_rate": 0.00014620582102882089,
+      "loss": 0.7339,
+      "step": 4125
+    },
+    {
+      "epoch": 1.3247712313372932,
+      "grad_norm": 1.0029937028884888,
+      "learning_rate": 0.00014617486132350343,
+      "loss": 0.63,
+      "step": 4126
+    },
+    {
+      "epoch": 1.3250923101621448,
+      "grad_norm": 0.9867237210273743,
+      "learning_rate": 0.00014614389599191917,
+      "loss": 0.7598,
+      "step": 4127
+    },
+    {
+      "epoch": 1.3254133889869963,
+      "grad_norm": 0.8086850643157959,
+      "learning_rate": 0.00014611292503784117,
+      "loss": 0.5079,
+      "step": 4128
+    },
+    {
+      "epoch": 1.3257344678118477,
+      "grad_norm": 0.8222152590751648,
+      "learning_rate": 0.0001460819484650431,
+      "loss": 0.551,
+      "step": 4129
+    },
+    {
+      "epoch": 1.3260555466366992,
+      "grad_norm": 0.84171462059021,
+      "learning_rate": 0.0001460509662772994,
+      "loss": 0.6227,
+      "step": 4130
+    },
+    {
+      "epoch": 1.3263766254615508,
+      "grad_norm": 0.8350204825401306,
+      "learning_rate": 0.00014601997847838518,
+      "loss": 0.626,
+      "step": 4131
+    },
+    {
+      "epoch": 1.3266977042864023,
+      "grad_norm": 1.4779672622680664,
+      "learning_rate": 0.00014598898507207615,
+      "loss": 0.6658,
+      "step": 4132
+    },
+    {
+      "epoch": 1.3270187831112539,
+      "grad_norm": 0.7730790376663208,
+      "learning_rate": 0.00014595798606214882,
+      "loss": 0.5676,
+      "step": 4133
+    },
+    {
+      "epoch": 1.3273398619361054,
+      "grad_norm": 1.5839117765426636,
+      "learning_rate": 0.00014592698145238028,
+      "loss": 0.7933,
+      "step": 4134
+    },
+    {
+      "epoch": 1.3276609407609568,
+      "grad_norm": 0.9410939812660217,
+      "learning_rate": 0.00014589597124654833,
+      "loss": 0.6305,
+      "step": 4135
+    },
+    {
+      "epoch": 1.3279820195858083,
+      "grad_norm": 1.0541006326675415,
+      "learning_rate": 0.00014586495544843152,
+      "loss": 0.7623,
+      "step": 4136
+    },
+    {
+      "epoch": 1.3283030984106599,
+      "grad_norm": 0.8780885934829712,
+      "learning_rate": 0.000145833934061809,
+      "loss": 0.6705,
+      "step": 4137
+    },
+    {
+      "epoch": 1.3286241772355112,
+      "grad_norm": 0.790573239326477,
+      "learning_rate": 0.00014580290709046066,
+      "loss": 0.6865,
+      "step": 4138
+    },
+    {
+      "epoch": 1.3289452560603627,
+      "grad_norm": 0.7406685948371887,
+      "learning_rate": 0.000145771874538167,
+      "loss": 0.4454,
+      "step": 4139
+    },
+    {
+      "epoch": 1.3292663348852143,
+      "grad_norm": 1.0333536863327026,
+      "learning_rate": 0.0001457408364087093,
+      "loss": 0.7394,
+      "step": 4140
+    },
+    {
+      "epoch": 1.3295874137100658,
+      "grad_norm": 0.9966391324996948,
+      "learning_rate": 0.00014570979270586945,
+      "loss": 0.5846,
+      "step": 4141
+    },
+    {
+      "epoch": 1.3299084925349174,
+      "grad_norm": 1.0538800954818726,
+      "learning_rate": 0.00014567874343342997,
+      "loss": 0.7812,
+      "step": 4142
+    },
+    {
+      "epoch": 1.330229571359769,
+      "grad_norm": 0.90020352602005,
+      "learning_rate": 0.00014564768859517418,
+      "loss": 0.62,
+      "step": 4143
+    },
+    {
+      "epoch": 1.3305506501846203,
+      "grad_norm": 1.0670877695083618,
+      "learning_rate": 0.00014561662819488597,
+      "loss": 0.6715,
+      "step": 4144
+    },
+    {
+      "epoch": 1.3308717290094718,
+      "grad_norm": 1.155572772026062,
+      "learning_rate": 0.00014558556223635003,
+      "loss": 0.6894,
+      "step": 4145
+    },
+    {
+      "epoch": 1.3311928078343234,
+      "grad_norm": 1.7336785793304443,
+      "learning_rate": 0.00014555449072335157,
+      "loss": 0.8215,
+      "step": 4146
+    },
+    {
+      "epoch": 1.3315138866591747,
+      "grad_norm": 1.0733205080032349,
+      "learning_rate": 0.00014552341365967658,
+      "loss": 0.7403,
+      "step": 4147
+    },
+    {
+      "epoch": 1.3318349654840262,
+      "grad_norm": 0.9724444150924683,
+      "learning_rate": 0.00014549233104911178,
+      "loss": 0.6041,
+      "step": 4148
+    },
+    {
+      "epoch": 1.3321560443088778,
+      "grad_norm": 1.006319522857666,
+      "learning_rate": 0.0001454612428954444,
+      "loss": 0.6566,
+      "step": 4149
+    },
+    {
+      "epoch": 1.3324771231337293,
+      "grad_norm": 0.736380398273468,
+      "learning_rate": 0.00014543014920246247,
+      "loss": 0.5568,
+      "step": 4150
+    },
+    {
+      "epoch": 1.332798201958581,
+      "grad_norm": 1.084572196006775,
+      "learning_rate": 0.00014539904997395468,
+      "loss": 0.5559,
+      "step": 4151
+    },
+    {
+      "epoch": 1.3331192807834324,
+      "grad_norm": 0.8436475396156311,
+      "learning_rate": 0.00014536794521371037,
+      "loss": 0.5434,
+      "step": 4152
+    },
+    {
+      "epoch": 1.3334403596082838,
+      "grad_norm": 1.179206371307373,
+      "learning_rate": 0.00014533683492551952,
+      "loss": 0.7242,
+      "step": 4153
+    },
+    {
+      "epoch": 1.3337614384331353,
+      "grad_norm": 1.3009215593338013,
+      "learning_rate": 0.0001453057191131729,
+      "loss": 0.6785,
+      "step": 4154
+    },
+    {
+      "epoch": 1.3340825172579869,
+      "grad_norm": 1.1896605491638184,
+      "learning_rate": 0.0001452745977804618,
+      "loss": 0.6612,
+      "step": 4155
+    },
+    {
+      "epoch": 1.3344035960828382,
+      "grad_norm": 1.1616253852844238,
+      "learning_rate": 0.00014524347093117828,
+      "loss": 0.6739,
+      "step": 4156
+    },
+    {
+      "epoch": 1.3347246749076898,
+      "grad_norm": 0.7783265113830566,
+      "learning_rate": 0.00014521233856911508,
+      "loss": 0.5076,
+      "step": 4157
+    },
+    {
+      "epoch": 1.3350457537325413,
+      "grad_norm": 1.7753092050552368,
+      "learning_rate": 0.00014518120069806557,
+      "loss": 0.5744,
+      "step": 4158
+    },
+    {
+      "epoch": 1.3353668325573929,
+      "grad_norm": 1.8857307434082031,
+      "learning_rate": 0.00014515005732182383,
+      "loss": 0.8165,
+      "step": 4159
+    },
+    {
+      "epoch": 1.3356879113822444,
+      "grad_norm": 0.8890649676322937,
+      "learning_rate": 0.00014511890844418453,
+      "loss": 0.3913,
+      "step": 4160
+    },
+    {
+      "epoch": 1.336008990207096,
+      "grad_norm": 0.691440999507904,
+      "learning_rate": 0.00014508775406894307,
+      "loss": 0.495,
+      "step": 4161
+    },
+    {
+      "epoch": 1.3363300690319473,
+      "grad_norm": 1.066584587097168,
+      "learning_rate": 0.0001450565941998956,
+      "loss": 0.6014,
+      "step": 4162
+    },
+    {
+      "epoch": 1.3366511478567988,
+      "grad_norm": 1.0590673685073853,
+      "learning_rate": 0.00014502542884083875,
+      "loss": 0.4447,
+      "step": 4163
+    },
+    {
+      "epoch": 1.3369722266816504,
+      "grad_norm": 0.7496399283409119,
+      "learning_rate": 0.00014499425799557,
+      "loss": 0.3962,
+      "step": 4164
+    },
+    {
+      "epoch": 1.3372933055065017,
+      "grad_norm": 0.5220407843589783,
+      "learning_rate": 0.0001449630816678874,
+      "loss": 0.4822,
+      "step": 4165
+    },
+    {
+      "epoch": 1.3376143843313533,
+      "grad_norm": 0.6849836111068726,
+      "learning_rate": 0.00014493189986158965,
+      "loss": 0.875,
+      "step": 4166
+    },
+    {
+      "epoch": 1.3379354631562048,
+      "grad_norm": 0.9034988284111023,
+      "learning_rate": 0.00014490071258047623,
+      "loss": 0.7436,
+      "step": 4167
+    },
+    {
+      "epoch": 1.3382565419810564,
+      "grad_norm": 0.7391397356987,
+      "learning_rate": 0.0001448695198283472,
+      "loss": 0.5641,
+      "step": 4168
+    },
+    {
+      "epoch": 1.338577620805908,
+      "grad_norm": 0.5824310779571533,
+      "learning_rate": 0.00014483832160900326,
+      "loss": 0.3517,
+      "step": 4169
+    },
+    {
+      "epoch": 1.3388986996307595,
+      "grad_norm": 0.7779441475868225,
+      "learning_rate": 0.0001448071179262458,
+      "loss": 0.4648,
+      "step": 4170
+    },
+    {
+      "epoch": 1.3392197784556108,
+      "grad_norm": 0.7949094176292419,
+      "learning_rate": 0.00014477590878387696,
+      "loss": 0.3886,
+      "step": 4171
+    },
+    {
+      "epoch": 1.3395408572804623,
+      "grad_norm": 0.7168573141098022,
+      "learning_rate": 0.0001447446941856995,
+      "loss": 0.5573,
+      "step": 4172
+    },
+    {
+      "epoch": 1.339861936105314,
+      "grad_norm": 0.9068509340286255,
+      "learning_rate": 0.00014471347413551672,
+      "loss": 0.7493,
+      "step": 4173
+    },
+    {
+      "epoch": 1.3401830149301652,
+      "grad_norm": 0.9333740472793579,
+      "learning_rate": 0.00014468224863713278,
+      "loss": 0.8158,
+      "step": 4174
+    },
+    {
+      "epoch": 1.3405040937550168,
+      "grad_norm": 0.8705101609230042,
+      "learning_rate": 0.00014465101769435234,
+      "loss": 0.7667,
+      "step": 4175
+    },
+    {
+      "epoch": 1.3408251725798683,
+      "grad_norm": 0.9098716974258423,
+      "learning_rate": 0.00014461978131098088,
+      "loss": 0.6398,
+      "step": 4176
+    },
+    {
+      "epoch": 1.3411462514047199,
+      "grad_norm": 0.7872979044914246,
+      "learning_rate": 0.00014458853949082443,
+      "loss": 0.5239,
+      "step": 4177
+    },
+    {
+      "epoch": 1.3414673302295714,
+      "grad_norm": 0.6730207800865173,
+      "learning_rate": 0.00014455729223768966,
+      "loss": 0.4844,
+      "step": 4178
+    },
+    {
+      "epoch": 1.341788409054423,
+      "grad_norm": 0.7260086536407471,
+      "learning_rate": 0.00014452603955538397,
+      "loss": 0.6197,
+      "step": 4179
+    },
+    {
+      "epoch": 1.3421094878792743,
+      "grad_norm": 0.8892760276794434,
+      "learning_rate": 0.00014449478144771543,
+      "loss": 0.6995,
+      "step": 4180
+    },
+    {
+      "epoch": 1.3424305667041259,
+      "grad_norm": 1.4739090204238892,
+      "learning_rate": 0.00014446351791849276,
+      "loss": 0.7409,
+      "step": 4181
+    },
+    {
+      "epoch": 1.3427516455289774,
+      "grad_norm": 0.7913824915885925,
+      "learning_rate": 0.0001444322489715253,
+      "loss": 0.6096,
+      "step": 4182
+    },
+    {
+      "epoch": 1.3430727243538287,
+      "grad_norm": 0.9372351169586182,
+      "learning_rate": 0.00014440097461062307,
+      "loss": 0.5211,
+      "step": 4183
+    },
+    {
+      "epoch": 1.3433938031786803,
+      "grad_norm": 1.0964716672897339,
+      "learning_rate": 0.00014436969483959676,
+      "loss": 0.7583,
+      "step": 4184
+    },
+    {
+      "epoch": 1.3437148820035318,
+      "grad_norm": 1.6157090663909912,
+      "learning_rate": 0.00014433840966225772,
+      "loss": 0.6351,
+      "step": 4185
+    },
+    {
+      "epoch": 1.3440359608283834,
+      "grad_norm": 0.8620708584785461,
+      "learning_rate": 0.00014430711908241798,
+      "loss": 0.5351,
+      "step": 4186
+    },
+    {
+      "epoch": 1.344357039653235,
+      "grad_norm": 0.814059317111969,
+      "learning_rate": 0.0001442758231038902,
+      "loss": 0.6808,
+      "step": 4187
+    },
+    {
+      "epoch": 1.3446781184780865,
+      "grad_norm": 1.1919108629226685,
+      "learning_rate": 0.0001442445217304876,
+      "loss": 0.8476,
+      "step": 4188
+    },
+    {
+      "epoch": 1.3449991973029378,
+      "grad_norm": 1.3358224630355835,
+      "learning_rate": 0.00014421321496602428,
+      "loss": 0.8407,
+      "step": 4189
+    },
+    {
+      "epoch": 1.3453202761277894,
+      "grad_norm": 1.180011510848999,
+      "learning_rate": 0.00014418190281431482,
+      "loss": 0.5575,
+      "step": 4190
+    },
+    {
+      "epoch": 1.345641354952641,
+      "grad_norm": 1.2531673908233643,
+      "learning_rate": 0.00014415058527917452,
+      "loss": 0.661,
+      "step": 4191
+    },
+    {
+      "epoch": 1.3459624337774923,
+      "grad_norm": 0.8697596192359924,
+      "learning_rate": 0.00014411926236441934,
+      "loss": 0.7263,
+      "step": 4192
+    },
+    {
+      "epoch": 1.3462835126023438,
+      "grad_norm": 1.6186376810073853,
+      "learning_rate": 0.00014408793407386588,
+      "loss": 0.6973,
+      "step": 4193
+    },
+    {
+      "epoch": 1.3466045914271954,
+      "grad_norm": 0.9928228259086609,
+      "learning_rate": 0.00014405660041133132,
+      "loss": 0.5972,
+      "step": 4194
+    },
+    {
+      "epoch": 1.346925670252047,
+      "grad_norm": 1.2978379726409912,
+      "learning_rate": 0.00014402526138063373,
+      "loss": 0.5707,
+      "step": 4195
+    },
+    {
+      "epoch": 1.3472467490768985,
+      "grad_norm": 1.1947479248046875,
+      "learning_rate": 0.00014399391698559152,
+      "loss": 0.7005,
+      "step": 4196
+    },
+    {
+      "epoch": 1.34756782790175,
+      "grad_norm": 1.2066709995269775,
+      "learning_rate": 0.000143962567230024,
+      "loss": 0.6688,
+      "step": 4197
+    },
+    {
+      "epoch": 1.3478889067266013,
+      "grad_norm": 0.8478728532791138,
+      "learning_rate": 0.000143931212117751,
+      "loss": 0.5476,
+      "step": 4198
+    },
+    {
+      "epoch": 1.3482099855514529,
+      "grad_norm": 1.081978678703308,
+      "learning_rate": 0.00014389985165259308,
+      "loss": 0.684,
+      "step": 4199
+    },
+    {
+      "epoch": 1.3485310643763044,
+      "grad_norm": 1.2415368556976318,
+      "learning_rate": 0.0001438684858383714,
+      "loss": 0.7176,
+      "step": 4200
+    },
+    {
+      "epoch": 1.3488521432011558,
+      "grad_norm": 1.0181922912597656,
+      "learning_rate": 0.00014383711467890774,
+      "loss": 0.6968,
+      "step": 4201
+    },
+    {
+      "epoch": 1.3491732220260073,
+      "grad_norm": 1.017606258392334,
+      "learning_rate": 0.00014380573817802467,
+      "loss": 0.474,
+      "step": 4202
+    },
+    {
+      "epoch": 1.3494943008508589,
+      "grad_norm": 1.2191901206970215,
+      "learning_rate": 0.00014377435633954527,
+      "loss": 0.5747,
+      "step": 4203
+    },
+    {
+      "epoch": 1.3498153796757104,
+      "grad_norm": 0.7296000719070435,
+      "learning_rate": 0.00014374296916729336,
+      "loss": 0.4687,
+      "step": 4204
+    },
+    {
+      "epoch": 1.350136458500562,
+      "grad_norm": 1.0375962257385254,
+      "learning_rate": 0.0001437115766650933,
+      "loss": 0.5935,
+      "step": 4205
+    },
+    {
+      "epoch": 1.3504575373254135,
+      "grad_norm": 0.8629053831100464,
+      "learning_rate": 0.00014368017883677024,
+      "loss": 0.5089,
+      "step": 4206
+    },
+    {
+      "epoch": 1.3507786161502648,
+      "grad_norm": 1.310308575630188,
+      "learning_rate": 0.0001436487756861499,
+      "loss": 0.7386,
+      "step": 4207
+    },
+    {
+      "epoch": 1.3510996949751164,
+      "grad_norm": 0.7340025901794434,
+      "learning_rate": 0.0001436173672170586,
+      "loss": 0.492,
+      "step": 4208
+    },
+    {
+      "epoch": 1.351420773799968,
+      "grad_norm": 0.8930183053016663,
+      "learning_rate": 0.00014358595343332342,
+      "loss": 0.4347,
+      "step": 4209
+    },
+    {
+      "epoch": 1.3517418526248193,
+      "grad_norm": 1.247145175933838,
+      "learning_rate": 0.00014355453433877204,
+      "loss": 0.5153,
+      "step": 4210
+    },
+    {
+      "epoch": 1.3520629314496708,
+      "grad_norm": 1.237648844718933,
+      "learning_rate": 0.00014352310993723277,
+      "loss": 0.5992,
+      "step": 4211
+    },
+    {
+      "epoch": 1.3523840102745224,
+      "grad_norm": 0.776263952255249,
+      "learning_rate": 0.00014349168023253456,
+      "loss": 0.5426,
+      "step": 4212
+    },
+    {
+      "epoch": 1.352705089099374,
+      "grad_norm": 1.0824544429779053,
+      "learning_rate": 0.00014346024522850703,
+      "loss": 0.5301,
+      "step": 4213
+    },
+    {
+      "epoch": 1.3530261679242255,
+      "grad_norm": 0.7915077209472656,
+      "learning_rate": 0.00014342880492898048,
+      "loss": 0.456,
+      "step": 4214
+    },
+    {
+      "epoch": 1.353347246749077,
+      "grad_norm": 0.845167338848114,
+      "learning_rate": 0.00014339735933778576,
+      "loss": 0.8329,
+      "step": 4215
+    },
+    {
+      "epoch": 1.3536683255739284,
+      "grad_norm": 0.7723715901374817,
+      "learning_rate": 0.00014336590845875446,
+      "loss": 0.914,
+      "step": 4216
+    },
+    {
+      "epoch": 1.35398940439878,
+      "grad_norm": 0.7588359713554382,
+      "learning_rate": 0.00014333445229571873,
+      "loss": 0.7267,
+      "step": 4217
+    },
+    {
+      "epoch": 1.3543104832236315,
+      "grad_norm": 0.733163058757782,
+      "learning_rate": 0.00014330299085251144,
+      "loss": 0.486,
+      "step": 4218
+    },
+    {
+      "epoch": 1.3546315620484828,
+      "grad_norm": 0.6390459537506104,
+      "learning_rate": 0.00014327152413296608,
+      "loss": 0.3084,
+      "step": 4219
+    },
+    {
+      "epoch": 1.3549526408733343,
+      "grad_norm": 0.8938314318656921,
+      "learning_rate": 0.00014324005214091676,
+      "loss": 0.4455,
+      "step": 4220
+    },
+    {
+      "epoch": 1.3552737196981859,
+      "grad_norm": 1.0440847873687744,
+      "learning_rate": 0.00014320857488019824,
+      "loss": 0.4297,
+      "step": 4221
+    },
+    {
+      "epoch": 1.3555947985230374,
+      "grad_norm": 0.8476107716560364,
+      "learning_rate": 0.00014317709235464593,
+      "loss": 0.5257,
+      "step": 4222
+    },
+    {
+      "epoch": 1.355915877347889,
+      "grad_norm": 0.9037560224533081,
+      "learning_rate": 0.0001431456045680959,
+      "loss": 0.6943,
+      "step": 4223
+    },
+    {
+      "epoch": 1.3562369561727405,
+      "grad_norm": 0.875007152557373,
+      "learning_rate": 0.00014311411152438482,
+      "loss": 0.6148,
+      "step": 4224
+    },
+    {
+      "epoch": 1.3565580349975919,
+      "grad_norm": 0.829646646976471,
+      "learning_rate": 0.00014308261322735005,
+      "loss": 0.7564,
+      "step": 4225
+    },
+    {
+      "epoch": 1.3568791138224434,
+      "grad_norm": 0.7148510217666626,
+      "learning_rate": 0.00014305110968082952,
+      "loss": 0.5596,
+      "step": 4226
+    },
+    {
+      "epoch": 1.357200192647295,
+      "grad_norm": 0.7443501949310303,
+      "learning_rate": 0.00014301960088866186,
+      "loss": 0.6532,
+      "step": 4227
+    },
+    {
+      "epoch": 1.3575212714721463,
+      "grad_norm": 0.9788792729377747,
+      "learning_rate": 0.00014298808685468635,
+      "loss": 0.5607,
+      "step": 4228
+    },
+    {
+      "epoch": 1.3578423502969978,
+      "grad_norm": 0.7856297492980957,
+      "learning_rate": 0.00014295656758274284,
+      "loss": 0.5779,
+      "step": 4229
+    },
+    {
+      "epoch": 1.3581634291218494,
+      "grad_norm": 0.7464038133621216,
+      "learning_rate": 0.00014292504307667186,
+      "loss": 0.4625,
+      "step": 4230
+    },
+    {
+      "epoch": 1.358484507946701,
+      "grad_norm": 0.958886981010437,
+      "learning_rate": 0.0001428935133403146,
+      "loss": 0.6014,
+      "step": 4231
+    },
+    {
+      "epoch": 1.3588055867715525,
+      "grad_norm": 1.1012500524520874,
+      "learning_rate": 0.00014286197837751286,
+      "loss": 0.9194,
+      "step": 4232
+    },
+    {
+      "epoch": 1.359126665596404,
+      "grad_norm": 0.8240073323249817,
+      "learning_rate": 0.00014283043819210905,
+      "loss": 0.5618,
+      "step": 4233
+    },
+    {
+      "epoch": 1.3594477444212554,
+      "grad_norm": 0.8680970072746277,
+      "learning_rate": 0.00014279889278794627,
+      "loss": 0.6105,
+      "step": 4234
+    },
+    {
+      "epoch": 1.359768823246107,
+      "grad_norm": 1.032063364982605,
+      "learning_rate": 0.00014276734216886821,
+      "loss": 0.7325,
+      "step": 4235
+    },
+    {
+      "epoch": 1.3600899020709585,
+      "grad_norm": 0.9061574935913086,
+      "learning_rate": 0.00014273578633871927,
+      "loss": 0.6026,
+      "step": 4236
+    },
+    {
+      "epoch": 1.3604109808958098,
+      "grad_norm": 0.7440208196640015,
+      "learning_rate": 0.00014270422530134432,
+      "loss": 0.552,
+      "step": 4237
+    },
+    {
+      "epoch": 1.3607320597206614,
+      "grad_norm": 1.1810888051986694,
+      "learning_rate": 0.00014267265906058914,
+      "loss": 0.7499,
+      "step": 4238
+    },
+    {
+      "epoch": 1.361053138545513,
+      "grad_norm": 0.903121292591095,
+      "learning_rate": 0.0001426410876202999,
+      "loss": 0.6064,
+      "step": 4239
+    },
+    {
+      "epoch": 1.3613742173703645,
+      "grad_norm": 0.9991443157196045,
+      "learning_rate": 0.00014260951098432343,
+      "loss": 0.661,
+      "step": 4240
+    },
+    {
+      "epoch": 1.361695296195216,
+      "grad_norm": 0.9029455184936523,
+      "learning_rate": 0.00014257792915650728,
+      "loss": 0.5712,
+      "step": 4241
+    },
+    {
+      "epoch": 1.3620163750200676,
+      "grad_norm": 0.899472713470459,
+      "learning_rate": 0.00014254634214069963,
+      "loss": 0.6596,
+      "step": 4242
+    },
+    {
+      "epoch": 1.3623374538449189,
+      "grad_norm": 0.7812873125076294,
+      "learning_rate": 0.00014251474994074928,
+      "loss": 0.5881,
+      "step": 4243
+    },
+    {
+      "epoch": 1.3626585326697704,
+      "grad_norm": 1.202508807182312,
+      "learning_rate": 0.00014248315256050557,
+      "loss": 0.6451,
+      "step": 4244
+    },
+    {
+      "epoch": 1.362979611494622,
+      "grad_norm": 0.7961817979812622,
+      "learning_rate": 0.0001424515500038186,
+      "loss": 0.4156,
+      "step": 4245
+    },
+    {
+      "epoch": 1.3633006903194733,
+      "grad_norm": 0.9045254588127136,
+      "learning_rate": 0.00014241994227453901,
+      "loss": 0.5355,
+      "step": 4246
+    },
+    {
+      "epoch": 1.3636217691443249,
+      "grad_norm": 1.0353204011917114,
+      "learning_rate": 0.00014238832937651816,
+      "loss": 0.6559,
+      "step": 4247
+    },
+    {
+      "epoch": 1.3639428479691764,
+      "grad_norm": 0.866054117679596,
+      "learning_rate": 0.00014235671131360798,
+      "loss": 0.562,
+      "step": 4248
+    },
+    {
+      "epoch": 1.364263926794028,
+      "grad_norm": 1.5949488878250122,
+      "learning_rate": 0.00014232508808966098,
+      "loss": 0.7606,
+      "step": 4249
+    },
+    {
+      "epoch": 1.3645850056188795,
+      "grad_norm": 1.053368091583252,
+      "learning_rate": 0.00014229345970853032,
+      "loss": 0.6,
+      "step": 4250
+    },
+    {
+      "epoch": 1.364906084443731,
+      "grad_norm": 0.8603706359863281,
+      "learning_rate": 0.00014226182617406996,
+      "loss": 0.5477,
+      "step": 4251
+    },
+    {
+      "epoch": 1.3652271632685824,
+      "grad_norm": 2.3181567192077637,
+      "learning_rate": 0.00014223018749013423,
+      "loss": 0.6106,
+      "step": 4252
+    },
+    {
+      "epoch": 1.365548242093434,
+      "grad_norm": 0.8432567715644836,
+      "learning_rate": 0.0001421985436605783,
+      "loss": 0.4574,
+      "step": 4253
+    },
+    {
+      "epoch": 1.3658693209182855,
+      "grad_norm": 1.323652982711792,
+      "learning_rate": 0.0001421668946892578,
+      "loss": 0.7721,
+      "step": 4254
+    },
+    {
+      "epoch": 1.3661903997431368,
+      "grad_norm": 0.9810237288475037,
+      "learning_rate": 0.0001421352405800291,
+      "loss": 0.5758,
+      "step": 4255
+    },
+    {
+      "epoch": 1.3665114785679884,
+      "grad_norm": 1.0755305290222168,
+      "learning_rate": 0.00014210358133674912,
+      "loss": 0.5253,
+      "step": 4256
+    },
+    {
+      "epoch": 1.36683255739284,
+      "grad_norm": 1.2368508577346802,
+      "learning_rate": 0.00014207191696327548,
+      "loss": 0.5937,
+      "step": 4257
+    },
+    {
+      "epoch": 1.3671536362176915,
+      "grad_norm": 0.9989250302314758,
+      "learning_rate": 0.00014204024746346637,
+      "loss": 0.7571,
+      "step": 4258
+    },
+    {
+      "epoch": 1.367474715042543,
+      "grad_norm": 0.9901091456413269,
+      "learning_rate": 0.00014200857284118066,
+      "loss": 0.6444,
+      "step": 4259
+    },
+    {
+      "epoch": 1.3677957938673946,
+      "grad_norm": 0.8036117553710938,
+      "learning_rate": 0.00014197689310027772,
+      "loss": 0.4396,
+      "step": 4260
+    },
+    {
+      "epoch": 1.368116872692246,
+      "grad_norm": 1.049436330795288,
+      "learning_rate": 0.00014194520824461771,
+      "loss": 0.5589,
+      "step": 4261
+    },
+    {
+      "epoch": 1.3684379515170975,
+      "grad_norm": 0.8437821865081787,
+      "learning_rate": 0.00014191351827806133,
+      "loss": 0.5044,
+      "step": 4262
+    },
+    {
+      "epoch": 1.368759030341949,
+      "grad_norm": 1.1721175909042358,
+      "learning_rate": 0.00014188182320446985,
+      "loss": 0.4725,
+      "step": 4263
+    },
+    {
+      "epoch": 1.3690801091668003,
+      "grad_norm": 0.7921808958053589,
+      "learning_rate": 0.00014185012302770527,
+      "loss": 0.4147,
+      "step": 4264
+    },
+    {
+      "epoch": 1.369401187991652,
+      "grad_norm": 1.2697014808654785,
+      "learning_rate": 0.00014181841775163013,
+      "loss": 0.9084,
+      "step": 4265
+    },
+    {
+      "epoch": 1.3697222668165034,
+      "grad_norm": 0.8648690581321716,
+      "learning_rate": 0.0001417867073801077,
+      "loss": 0.9195,
+      "step": 4266
+    },
+    {
+      "epoch": 1.370043345641355,
+      "grad_norm": 0.7991724014282227,
+      "learning_rate": 0.00014175499191700167,
+      "loss": 0.7083,
+      "step": 4267
+    },
+    {
+      "epoch": 1.3703644244662065,
+      "grad_norm": 0.8012338876724243,
+      "learning_rate": 0.00014172327136617656,
+      "loss": 0.4684,
+      "step": 4268
+    },
+    {
+      "epoch": 1.3706855032910579,
+      "grad_norm": 0.9364102482795715,
+      "learning_rate": 0.00014169154573149737,
+      "loss": 0.383,
+      "step": 4269
+    },
+    {
+      "epoch": 1.3710065821159094,
+      "grad_norm": 1.1031969785690308,
+      "learning_rate": 0.0001416598150168298,
+      "loss": 0.5984,
+      "step": 4270
+    },
+    {
+      "epoch": 1.371327660940761,
+      "grad_norm": 1.1069051027297974,
+      "learning_rate": 0.00014162807922604012,
+      "loss": 0.7745,
+      "step": 4271
+    },
+    {
+      "epoch": 1.3716487397656125,
+      "grad_norm": 1.0749973058700562,
+      "learning_rate": 0.00014159633836299527,
+      "loss": 0.7569,
+      "step": 4272
+    },
+    {
+      "epoch": 1.3719698185904639,
+      "grad_norm": 0.8237097263336182,
+      "learning_rate": 0.0001415645924315628,
+      "loss": 0.7982,
+      "step": 4273
+    },
+    {
+      "epoch": 1.3722908974153154,
+      "grad_norm": 0.9842055439949036,
+      "learning_rate": 0.0001415328414356108,
+      "loss": 0.6246,
+      "step": 4274
+    },
+    {
+      "epoch": 1.372611976240167,
+      "grad_norm": 0.9688705205917358,
+      "learning_rate": 0.00014150108537900805,
+      "loss": 0.5753,
+      "step": 4275
+    },
+    {
+      "epoch": 1.3729330550650185,
+      "grad_norm": 0.6846550703048706,
+      "learning_rate": 0.00014146932426562392,
+      "loss": 0.5259,
+      "step": 4276
+    },
+    {
+      "epoch": 1.37325413388987,
+      "grad_norm": 1.078979730606079,
+      "learning_rate": 0.00014143755809932845,
+      "loss": 0.6775,
+      "step": 4277
+    },
+    {
+      "epoch": 1.3735752127147214,
+      "grad_norm": 0.8213169574737549,
+      "learning_rate": 0.00014140578688399218,
+      "loss": 0.4362,
+      "step": 4278
+    },
+    {
+      "epoch": 1.373896291539573,
+      "grad_norm": 1.00385320186615,
+      "learning_rate": 0.00014137401062348638,
+      "loss": 0.6659,
+      "step": 4279
+    },
+    {
+      "epoch": 1.3742173703644245,
+      "grad_norm": 0.8310264945030212,
+      "learning_rate": 0.0001413422293216829,
+      "loss": 0.8067,
+      "step": 4280
+    },
+    {
+      "epoch": 1.374538449189276,
+      "grad_norm": 0.888613224029541,
+      "learning_rate": 0.0001413104429824542,
+      "loss": 0.688,
+      "step": 4281
+    },
+    {
+      "epoch": 1.3748595280141274,
+      "grad_norm": 0.9263864755630493,
+      "learning_rate": 0.0001412786516096733,
+      "loss": 0.5974,
+      "step": 4282
+    },
+    {
+      "epoch": 1.375180606838979,
+      "grad_norm": 1.2327766418457031,
+      "learning_rate": 0.00014124685520721392,
+      "loss": 0.8498,
+      "step": 4283
+    },
+    {
+      "epoch": 1.3755016856638305,
+      "grad_norm": 0.7652939558029175,
+      "learning_rate": 0.00014121505377895038,
+      "loss": 0.4643,
+      "step": 4284
+    },
+    {
+      "epoch": 1.375822764488682,
+      "grad_norm": 0.8574872016906738,
+      "learning_rate": 0.00014118324732875748,
+      "loss": 0.5679,
+      "step": 4285
+    },
+    {
+      "epoch": 1.3761438433135336,
+      "grad_norm": 1.3594255447387695,
+      "learning_rate": 0.00014115143586051088,
+      "loss": 0.7465,
+      "step": 4286
+    },
+    {
+      "epoch": 1.376464922138385,
+      "grad_norm": 0.8461220264434814,
+      "learning_rate": 0.00014111961937808665,
+      "loss": 0.4636,
+      "step": 4287
+    },
+    {
+      "epoch": 1.3767860009632364,
+      "grad_norm": 0.9702261686325073,
+      "learning_rate": 0.0001410877978853615,
+      "loss": 0.6617,
+      "step": 4288
+    },
+    {
+      "epoch": 1.377107079788088,
+      "grad_norm": 0.8296035528182983,
+      "learning_rate": 0.0001410559713862128,
+      "loss": 0.5687,
+      "step": 4289
+    },
+    {
+      "epoch": 1.3774281586129395,
+      "grad_norm": 1.2875922918319702,
+      "learning_rate": 0.00014102413988451856,
+      "loss": 0.8027,
+      "step": 4290
+    },
+    {
+      "epoch": 1.3777492374377909,
+      "grad_norm": 0.8944180607795715,
+      "learning_rate": 0.00014099230338415728,
+      "loss": 0.6401,
+      "step": 4291
+    },
+    {
+      "epoch": 1.3780703162626424,
+      "grad_norm": 0.9776332378387451,
+      "learning_rate": 0.00014096046188900822,
+      "loss": 0.7287,
+      "step": 4292
+    },
+    {
+      "epoch": 1.378391395087494,
+      "grad_norm": 0.7893264293670654,
+      "learning_rate": 0.00014092861540295108,
+      "loss": 0.5128,
+      "step": 4293
+    },
+    {
+      "epoch": 1.3787124739123455,
+      "grad_norm": 1.7183669805526733,
+      "learning_rate": 0.0001408967639298663,
+      "loss": 0.5958,
+      "step": 4294
+    },
+    {
+      "epoch": 1.379033552737197,
+      "grad_norm": 1.1158559322357178,
+      "learning_rate": 0.00014086490747363493,
+      "loss": 0.8365,
+      "step": 4295
+    },
+    {
+      "epoch": 1.3793546315620484,
+      "grad_norm": 1.0233869552612305,
+      "learning_rate": 0.00014083304603813848,
+      "loss": 0.7533,
+      "step": 4296
+    },
+    {
+      "epoch": 1.3796757103869,
+      "grad_norm": 1.1017775535583496,
+      "learning_rate": 0.0001408011796272593,
+      "loss": 0.6726,
+      "step": 4297
+    },
+    {
+      "epoch": 1.3799967892117515,
+      "grad_norm": 0.8513362407684326,
+      "learning_rate": 0.0001407693082448801,
+      "loss": 0.5752,
+      "step": 4298
+    },
+    {
+      "epoch": 1.380317868036603,
+      "grad_norm": 1.0489625930786133,
+      "learning_rate": 0.00014073743189488435,
+      "loss": 0.5832,
+      "step": 4299
+    },
+    {
+      "epoch": 1.3806389468614544,
+      "grad_norm": 0.8657923340797424,
+      "learning_rate": 0.00014070555058115614,
+      "loss": 0.6029,
+      "step": 4300
+    },
+    {
+      "epoch": 1.380960025686306,
+      "grad_norm": 1.106916904449463,
+      "learning_rate": 0.00014067366430758004,
+      "loss": 0.655,
+      "step": 4301
+    },
+    {
+      "epoch": 1.3812811045111575,
+      "grad_norm": 0.9051868915557861,
+      "learning_rate": 0.00014064177307804133,
+      "loss": 0.6185,
+      "step": 4302
+    },
+    {
+      "epoch": 1.381602183336009,
+      "grad_norm": 0.950446605682373,
+      "learning_rate": 0.00014060987689642581,
+      "loss": 0.6738,
+      "step": 4303
+    },
+    {
+      "epoch": 1.3819232621608606,
+      "grad_norm": 1.1080812215805054,
+      "learning_rate": 0.00014057797576662,
+      "loss": 0.6398,
+      "step": 4304
+    },
+    {
+      "epoch": 1.382244340985712,
+      "grad_norm": 0.6714284420013428,
+      "learning_rate": 0.00014054606969251095,
+      "loss": 0.4397,
+      "step": 4305
+    },
+    {
+      "epoch": 1.3825654198105635,
+      "grad_norm": 0.9164280891418457,
+      "learning_rate": 0.00014051415867798628,
+      "loss": 0.5544,
+      "step": 4306
+    },
+    {
+      "epoch": 1.382886498635415,
+      "grad_norm": 0.951714277267456,
+      "learning_rate": 0.00014048224272693424,
+      "loss": 0.6772,
+      "step": 4307
+    },
+    {
+      "epoch": 1.3832075774602666,
+      "grad_norm": 0.9855425357818604,
+      "learning_rate": 0.00014045032184324373,
+      "loss": 0.6,
+      "step": 4308
+    },
+    {
+      "epoch": 1.383528656285118,
+      "grad_norm": 0.9370409250259399,
+      "learning_rate": 0.00014041839603080422,
+      "loss": 0.5603,
+      "step": 4309
+    },
+    {
+      "epoch": 1.3838497351099694,
+      "grad_norm": 0.9768514633178711,
+      "learning_rate": 0.00014038646529350579,
+      "loss": 0.5308,
+      "step": 4310
+    },
+    {
+      "epoch": 1.384170813934821,
+      "grad_norm": 1.384358525276184,
+      "learning_rate": 0.00014035452963523902,
+      "loss": 0.6596,
+      "step": 4311
+    },
+    {
+      "epoch": 1.3844918927596725,
+      "grad_norm": 1.2001529932022095,
+      "learning_rate": 0.00014032258905989522,
+      "loss": 0.4874,
+      "step": 4312
+    },
+    {
+      "epoch": 1.384812971584524,
+      "grad_norm": 0.5518187284469604,
+      "learning_rate": 0.00014029064357136628,
+      "loss": 0.3581,
+      "step": 4313
+    },
+    {
+      "epoch": 1.3851340504093754,
+      "grad_norm": 0.7382297515869141,
+      "learning_rate": 0.00014025869317354462,
+      "loss": 0.4,
+      "step": 4314
+    },
+    {
+      "epoch": 1.385455129234227,
+      "grad_norm": 0.8723937273025513,
+      "learning_rate": 0.00014022673787032332,
+      "loss": 0.6065,
+      "step": 4315
+    },
+    {
+      "epoch": 1.3857762080590785,
+      "grad_norm": 0.8456537127494812,
+      "learning_rate": 0.00014019477766559604,
+      "loss": 0.5173,
+      "step": 4316
+    },
+    {
+      "epoch": 1.38609728688393,
+      "grad_norm": 0.8314757347106934,
+      "learning_rate": 0.000140162812563257,
+      "loss": 0.5405,
+      "step": 4317
+    },
+    {
+      "epoch": 1.3864183657087814,
+      "grad_norm": 0.9915850162506104,
+      "learning_rate": 0.00014013084256720107,
+      "loss": 0.4016,
+      "step": 4318
+    },
+    {
+      "epoch": 1.386739444533633,
+      "grad_norm": 0.6893324851989746,
+      "learning_rate": 0.00014009886768132375,
+      "loss": 0.2416,
+      "step": 4319
+    },
+    {
+      "epoch": 1.3870605233584845,
+      "grad_norm": 0.7434089183807373,
+      "learning_rate": 0.000140066887909521,
+      "loss": 0.2607,
+      "step": 4320
+    },
+    {
+      "epoch": 1.387381602183336,
+      "grad_norm": 0.7426462769508362,
+      "learning_rate": 0.00014003490325568954,
+      "loss": 0.6526,
+      "step": 4321
+    },
+    {
+      "epoch": 1.3877026810081876,
+      "grad_norm": 1.0482592582702637,
+      "learning_rate": 0.00014000291372372647,
+      "loss": 0.6866,
+      "step": 4322
+    },
+    {
+      "epoch": 1.388023759833039,
+      "grad_norm": 0.7865654826164246,
+      "learning_rate": 0.00013997091931752977,
+      "loss": 0.7286,
+      "step": 4323
+    },
+    {
+      "epoch": 1.3883448386578905,
+      "grad_norm": 0.7435041666030884,
+      "learning_rate": 0.00013993892004099777,
+      "loss": 0.6065,
+      "step": 4324
+    },
+    {
+      "epoch": 1.388665917482742,
+      "grad_norm": 1.0697741508483887,
+      "learning_rate": 0.00013990691589802954,
+      "loss": 0.7525,
+      "step": 4325
+    },
+    {
+      "epoch": 1.3889869963075936,
+      "grad_norm": 0.6864974498748779,
+      "learning_rate": 0.00013987490689252463,
+      "loss": 0.5124,
+      "step": 4326
+    },
+    {
+      "epoch": 1.389308075132445,
+      "grad_norm": 1.0872437953948975,
+      "learning_rate": 0.00013984289302838328,
+      "loss": 0.7208,
+      "step": 4327
+    },
+    {
+      "epoch": 1.3896291539572965,
+      "grad_norm": 0.8134937286376953,
+      "learning_rate": 0.00013981087430950628,
+      "loss": 0.5676,
+      "step": 4328
+    },
+    {
+      "epoch": 1.389950232782148,
+      "grad_norm": 0.8556815981864929,
+      "learning_rate": 0.000139778850739795,
+      "loss": 0.6281,
+      "step": 4329
+    },
+    {
+      "epoch": 1.3902713116069996,
+      "grad_norm": 0.8196367025375366,
+      "learning_rate": 0.0001397468223231514,
+      "loss": 0.5989,
+      "step": 4330
+    },
+    {
+      "epoch": 1.3905923904318511,
+      "grad_norm": 0.7392389178276062,
+      "learning_rate": 0.00013971478906347806,
+      "loss": 0.5696,
+      "step": 4331
+    },
+    {
+      "epoch": 1.3909134692567025,
+      "grad_norm": 0.8934568166732788,
+      "learning_rate": 0.0001396827509646782,
+      "loss": 0.649,
+      "step": 4332
+    },
+    {
+      "epoch": 1.391234548081554,
+      "grad_norm": 0.861387312412262,
+      "learning_rate": 0.00013965070803065543,
+      "loss": 0.5132,
+      "step": 4333
+    },
+    {
+      "epoch": 1.3915556269064056,
+      "grad_norm": 0.8012207746505737,
+      "learning_rate": 0.00013961866026531417,
+      "loss": 0.4985,
+      "step": 4334
+    },
+    {
+      "epoch": 1.391876705731257,
+      "grad_norm": 1.5408804416656494,
+      "learning_rate": 0.00013958660767255938,
+      "loss": 0.8035,
+      "step": 4335
+    },
+    {
+      "epoch": 1.3921977845561084,
+      "grad_norm": 1.0151411294937134,
+      "learning_rate": 0.00013955455025629651,
+      "loss": 0.6018,
+      "step": 4336
+    },
+    {
+      "epoch": 1.39251886338096,
+      "grad_norm": 0.9835482835769653,
+      "learning_rate": 0.00013952248802043165,
+      "loss": 0.625,
+      "step": 4337
+    },
+    {
+      "epoch": 1.3928399422058115,
+      "grad_norm": 1.0157198905944824,
+      "learning_rate": 0.00013949042096887153,
+      "loss": 0.6432,
+      "step": 4338
+    },
+    {
+      "epoch": 1.393161021030663,
+      "grad_norm": 1.2168697118759155,
+      "learning_rate": 0.0001394583491055234,
+      "loss": 0.4917,
+      "step": 4339
+    },
+    {
+      "epoch": 1.3934820998555146,
+      "grad_norm": 1.4631038904190063,
+      "learning_rate": 0.00013942627243429512,
+      "loss": 0.6929,
+      "step": 4340
+    },
+    {
+      "epoch": 1.393803178680366,
+      "grad_norm": 1.6047730445861816,
+      "learning_rate": 0.00013939419095909512,
+      "loss": 0.6978,
+      "step": 4341
+    },
+    {
+      "epoch": 1.3941242575052175,
+      "grad_norm": 1.1495589017868042,
+      "learning_rate": 0.00013936210468383246,
+      "loss": 0.8062,
+      "step": 4342
+    },
+    {
+      "epoch": 1.394445336330069,
+      "grad_norm": 1.0219823122024536,
+      "learning_rate": 0.00013933001361241673,
+      "loss": 0.6822,
+      "step": 4343
+    },
+    {
+      "epoch": 1.3947664151549206,
+      "grad_norm": 0.9300022125244141,
+      "learning_rate": 0.00013929791774875815,
+      "loss": 0.6795,
+      "step": 4344
+    },
+    {
+      "epoch": 1.395087493979772,
+      "grad_norm": 0.8773917555809021,
+      "learning_rate": 0.00013926581709676751,
+      "loss": 0.4946,
+      "step": 4345
+    },
+    {
+      "epoch": 1.3954085728046235,
+      "grad_norm": 0.8553141951560974,
+      "learning_rate": 0.00013923371166035616,
+      "loss": 0.6083,
+      "step": 4346
+    },
+    {
+      "epoch": 1.395729651629475,
+      "grad_norm": 0.9813185930252075,
+      "learning_rate": 0.00013920160144343603,
+      "loss": 0.8011,
+      "step": 4347
+    },
+    {
+      "epoch": 1.3960507304543266,
+      "grad_norm": 1.4478272199630737,
+      "learning_rate": 0.0001391694864499197,
+      "loss": 0.5279,
+      "step": 4348
+    },
+    {
+      "epoch": 1.3963718092791781,
+      "grad_norm": 0.9650620818138123,
+      "learning_rate": 0.00013913736668372026,
+      "loss": 0.6811,
+      "step": 4349
+    },
+    {
+      "epoch": 1.3966928881040295,
+      "grad_norm": 1.1189510822296143,
+      "learning_rate": 0.00013910524214875137,
+      "loss": 0.7923,
+      "step": 4350
+    },
+    {
+      "epoch": 1.397013966928881,
+      "grad_norm": 1.1408734321594238,
+      "learning_rate": 0.00013907311284892736,
+      "loss": 0.6301,
+      "step": 4351
+    },
+    {
+      "epoch": 1.3973350457537326,
+      "grad_norm": 1.3011447191238403,
+      "learning_rate": 0.00013904097878816312,
+      "loss": 0.4263,
+      "step": 4352
+    },
+    {
+      "epoch": 1.3976561245785841,
+      "grad_norm": 0.7979041337966919,
+      "learning_rate": 0.00013900883997037397,
+      "loss": 0.534,
+      "step": 4353
+    },
+    {
+      "epoch": 1.3979772034034355,
+      "grad_norm": 0.9429032206535339,
+      "learning_rate": 0.00013897669639947606,
+      "loss": 0.7349,
+      "step": 4354
+    },
+    {
+      "epoch": 1.398298282228287,
+      "grad_norm": 1.0224515199661255,
+      "learning_rate": 0.00013894454807938586,
+      "loss": 0.5381,
+      "step": 4355
+    },
+    {
+      "epoch": 1.3986193610531386,
+      "grad_norm": 0.8756483197212219,
+      "learning_rate": 0.0001389123950140206,
+      "loss": 0.5863,
+      "step": 4356
+    },
+    {
+      "epoch": 1.39894043987799,
+      "grad_norm": 1.352267861366272,
+      "learning_rate": 0.0001388802372072981,
+      "loss": 0.6053,
+      "step": 4357
+    },
+    {
+      "epoch": 1.3992615187028417,
+      "grad_norm": 0.9977678060531616,
+      "learning_rate": 0.00013884807466313663,
+      "loss": 0.6707,
+      "step": 4358
+    },
+    {
+      "epoch": 1.399582597527693,
+      "grad_norm": 0.9317201972007751,
+      "learning_rate": 0.00013881590738545508,
+      "loss": 0.6246,
+      "step": 4359
+    },
+    {
+      "epoch": 1.3999036763525445,
+      "grad_norm": 0.8313597440719604,
+      "learning_rate": 0.00013878373537817292,
+      "loss": 0.4458,
+      "step": 4360
+    },
+    {
+      "epoch": 1.400224755177396,
+      "grad_norm": 0.9241980314254761,
+      "learning_rate": 0.0001387515586452103,
+      "loss": 0.566,
+      "step": 4361
+    },
+    {
+      "epoch": 1.4005458340022474,
+      "grad_norm": 0.5518043041229248,
+      "learning_rate": 0.00013871937719048779,
+      "loss": 0.3723,
+      "step": 4362
+    },
+    {
+      "epoch": 1.400866912827099,
+      "grad_norm": 0.598588764667511,
+      "learning_rate": 0.00013868719101792665,
+      "loss": 0.354,
+      "step": 4363
+    },
+    {
+      "epoch": 1.4011879916519505,
+      "grad_norm": 1.0155926942825317,
+      "learning_rate": 0.00013865500013144857,
+      "loss": 0.614,
+      "step": 4364
+    },
+    {
+      "epoch": 1.401509070476802,
+      "grad_norm": 0.5616649389266968,
+      "learning_rate": 0.000138622804534976,
+      "loss": 0.4872,
+      "step": 4365
+    },
+    {
+      "epoch": 1.4018301493016536,
+      "grad_norm": 0.6463456749916077,
+      "learning_rate": 0.00013859060423243187,
+      "loss": 1.0319,
+      "step": 4366
+    },
+    {
+      "epoch": 1.4021512281265052,
+      "grad_norm": 0.708520770072937,
+      "learning_rate": 0.00013855839922773968,
+      "loss": 1.1378,
+      "step": 4367
+    },
+    {
+      "epoch": 1.4024723069513565,
+      "grad_norm": 0.6274012327194214,
+      "learning_rate": 0.00013852618952482347,
+      "loss": 0.4126,
+      "step": 4368
+    },
+    {
+      "epoch": 1.402793385776208,
+      "grad_norm": 0.7933291792869568,
+      "learning_rate": 0.00013849397512760795,
+      "loss": 0.4245,
+      "step": 4369
+    },
+    {
+      "epoch": 1.4031144646010596,
+      "grad_norm": 0.7001267671585083,
+      "learning_rate": 0.0001384617560400183,
+      "loss": 0.3049,
+      "step": 4370
+    },
+    {
+      "epoch": 1.403435543425911,
+      "grad_norm": 0.7495763897895813,
+      "learning_rate": 0.00013842953226598037,
+      "loss": 0.3643,
+      "step": 4371
+    },
+    {
+      "epoch": 1.4037566222507625,
+      "grad_norm": 0.7259708046913147,
+      "learning_rate": 0.00013839730380942053,
+      "loss": 0.3979,
+      "step": 4372
+    },
+    {
+      "epoch": 1.404077701075614,
+      "grad_norm": 1.0623929500579834,
+      "learning_rate": 0.00013836507067426564,
+      "loss": 0.9446,
+      "step": 4373
+    },
+    {
+      "epoch": 1.4043987799004656,
+      "grad_norm": 0.9666998386383057,
+      "learning_rate": 0.00013833283286444328,
+      "loss": 0.7658,
+      "step": 4374
+    },
+    {
+      "epoch": 1.4047198587253171,
+      "grad_norm": 0.7665534019470215,
+      "learning_rate": 0.0001383005903838815,
+      "loss": 0.5662,
+      "step": 4375
+    },
+    {
+      "epoch": 1.4050409375501687,
+      "grad_norm": 1.0555462837219238,
+      "learning_rate": 0.000138268343236509,
+      "loss": 0.7391,
+      "step": 4376
+    },
+    {
+      "epoch": 1.40536201637502,
+      "grad_norm": 0.9639325141906738,
+      "learning_rate": 0.0001382360914262549,
+      "loss": 0.6964,
+      "step": 4377
+    },
+    {
+      "epoch": 1.4056830951998716,
+      "grad_norm": 0.8347415924072266,
+      "learning_rate": 0.0001382038349570491,
+      "loss": 0.5785,
+      "step": 4378
+    },
+    {
+      "epoch": 1.406004174024723,
+      "grad_norm": 0.8627128005027771,
+      "learning_rate": 0.00013817157383282184,
+      "loss": 0.6525,
+      "step": 4379
+    },
+    {
+      "epoch": 1.4063252528495744,
+      "grad_norm": 1.1546865701675415,
+      "learning_rate": 0.00013813930805750413,
+      "loss": 0.7064,
+      "step": 4380
+    },
+    {
+      "epoch": 1.406646331674426,
+      "grad_norm": 0.9595782160758972,
+      "learning_rate": 0.00013810703763502744,
+      "loss": 0.593,
+      "step": 4381
+    },
+    {
+      "epoch": 1.4069674104992775,
+      "grad_norm": 0.7686512470245361,
+      "learning_rate": 0.00013807476256932376,
+      "loss": 0.4359,
+      "step": 4382
+    },
+    {
+      "epoch": 1.407288489324129,
+      "grad_norm": 1.0802874565124512,
+      "learning_rate": 0.00013804248286432578,
+      "loss": 0.8544,
+      "step": 4383
+    },
+    {
+      "epoch": 1.4076095681489806,
+      "grad_norm": 0.8571449518203735,
+      "learning_rate": 0.00013801019852396665,
+      "loss": 0.6631,
+      "step": 4384
+    },
+    {
+      "epoch": 1.4079306469738322,
+      "grad_norm": 0.9189439415931702,
+      "learning_rate": 0.00013797790955218014,
+      "loss": 0.6817,
+      "step": 4385
+    },
+    {
+      "epoch": 1.4082517257986835,
+      "grad_norm": 0.814200222492218,
+      "learning_rate": 0.00013794561595290052,
+      "loss": 0.5863,
+      "step": 4386
+    },
+    {
+      "epoch": 1.408572804623535,
+      "grad_norm": 1.049676537513733,
+      "learning_rate": 0.0001379133177300627,
+      "loss": 0.6476,
+      "step": 4387
+    },
+    {
+      "epoch": 1.4088938834483866,
+      "grad_norm": 0.9915246367454529,
+      "learning_rate": 0.00013788101488760215,
+      "loss": 0.6418,
+      "step": 4388
+    },
+    {
+      "epoch": 1.409214962273238,
+      "grad_norm": 1.644821047782898,
+      "learning_rate": 0.00013784870742945482,
+      "loss": 0.776,
+      "step": 4389
+    },
+    {
+      "epoch": 1.4095360410980895,
+      "grad_norm": 0.856365978717804,
+      "learning_rate": 0.00013781639535955732,
+      "loss": 0.5186,
+      "step": 4390
+    },
+    {
+      "epoch": 1.409857119922941,
+      "grad_norm": 0.9367002844810486,
+      "learning_rate": 0.00013778407868184672,
+      "loss": 0.7263,
+      "step": 4391
+    },
+    {
+      "epoch": 1.4101781987477926,
+      "grad_norm": 0.8069283962249756,
+      "learning_rate": 0.00013775175740026078,
+      "loss": 0.6471,
+      "step": 4392
+    },
+    {
+      "epoch": 1.4104992775726442,
+      "grad_norm": 0.9524908065795898,
+      "learning_rate": 0.00013771943151873767,
+      "loss": 0.762,
+      "step": 4393
+    },
+    {
+      "epoch": 1.4108203563974957,
+      "grad_norm": 1.1511192321777344,
+      "learning_rate": 0.00013768710104121627,
+      "loss": 0.5543,
+      "step": 4394
+    },
+    {
+      "epoch": 1.411141435222347,
+      "grad_norm": 0.8177773952484131,
+      "learning_rate": 0.00013765476597163594,
+      "loss": 0.6148,
+      "step": 4395
+    },
+    {
+      "epoch": 1.4114625140471986,
+      "grad_norm": 1.0919883251190186,
+      "learning_rate": 0.00013762242631393655,
+      "loss": 0.7347,
+      "step": 4396
+    },
+    {
+      "epoch": 1.4117835928720501,
+      "grad_norm": 0.9575279951095581,
+      "learning_rate": 0.0001375900820720587,
+      "loss": 0.6493,
+      "step": 4397
+    },
+    {
+      "epoch": 1.4121046716969015,
+      "grad_norm": 0.8994300365447998,
+      "learning_rate": 0.0001375577332499433,
+      "loss": 0.6778,
+      "step": 4398
+    },
+    {
+      "epoch": 1.412425750521753,
+      "grad_norm": 1.2716469764709473,
+      "learning_rate": 0.0001375253798515321,
+      "loss": 0.794,
+      "step": 4399
+    },
+    {
+      "epoch": 1.4127468293466046,
+      "grad_norm": 0.9935747385025024,
+      "learning_rate": 0.00013749302188076717,
+      "loss": 0.688,
+      "step": 4400
+    },
+    {
+      "epoch": 1.413067908171456,
+      "grad_norm": 1.3132479190826416,
+      "learning_rate": 0.00013746065934159123,
+      "loss": 0.6833,
+      "step": 4401
+    },
+    {
+      "epoch": 1.4133889869963077,
+      "grad_norm": 0.9004419445991516,
+      "learning_rate": 0.00013742829223794759,
+      "loss": 0.5584,
+      "step": 4402
+    },
+    {
+      "epoch": 1.4137100658211592,
+      "grad_norm": 0.8987523317337036,
+      "learning_rate": 0.00013739592057378003,
+      "loss": 0.5162,
+      "step": 4403
+    },
+    {
+      "epoch": 1.4140311446460105,
+      "grad_norm": 1.0713285207748413,
+      "learning_rate": 0.00013736354435303305,
+      "loss": 0.6563,
+      "step": 4404
+    },
+    {
+      "epoch": 1.414352223470862,
+      "grad_norm": 0.9994504451751709,
+      "learning_rate": 0.0001373311635796515,
+      "loss": 0.6936,
+      "step": 4405
+    },
+    {
+      "epoch": 1.4146733022957136,
+      "grad_norm": 1.2554911375045776,
+      "learning_rate": 0.0001372987782575809,
+      "loss": 0.6968,
+      "step": 4406
+    },
+    {
+      "epoch": 1.414994381120565,
+      "grad_norm": 0.9798457622528076,
+      "learning_rate": 0.0001372663883907673,
+      "loss": 0.648,
+      "step": 4407
+    },
+    {
+      "epoch": 1.4153154599454165,
+      "grad_norm": 1.1374400854110718,
+      "learning_rate": 0.00013723399398315734,
+      "loss": 0.6739,
+      "step": 4408
+    },
+    {
+      "epoch": 1.415636538770268,
+      "grad_norm": 0.9649138450622559,
+      "learning_rate": 0.00013720159503869815,
+      "loss": 0.5802,
+      "step": 4409
+    },
+    {
+      "epoch": 1.4159576175951196,
+      "grad_norm": 1.5144495964050293,
+      "learning_rate": 0.00013716919156133746,
+      "loss": 0.6543,
+      "step": 4410
+    },
+    {
+      "epoch": 1.4162786964199712,
+      "grad_norm": 1.2621392011642456,
+      "learning_rate": 0.00013713678355502351,
+      "loss": 0.6814,
+      "step": 4411
+    },
+    {
+      "epoch": 1.4165997752448227,
+      "grad_norm": 0.90981125831604,
+      "learning_rate": 0.0001371043710237051,
+      "loss": 0.5224,
+      "step": 4412
+    },
+    {
+      "epoch": 1.416920854069674,
+      "grad_norm": 0.5423502922058105,
+      "learning_rate": 0.00013707195397133165,
+      "loss": 0.369,
+      "step": 4413
+    },
+    {
+      "epoch": 1.4172419328945256,
+      "grad_norm": 0.6597849130630493,
+      "learning_rate": 0.0001370395324018531,
+      "loss": 0.4365,
+      "step": 4414
+    },
+    {
+      "epoch": 1.4175630117193772,
+      "grad_norm": 0.6346552968025208,
+      "learning_rate": 0.00013700710631921984,
+      "loss": 0.5433,
+      "step": 4415
+    },
+    {
+      "epoch": 1.4178840905442285,
+      "grad_norm": 0.8271416425704956,
+      "learning_rate": 0.00013697467572738295,
+      "loss": 0.9643,
+      "step": 4416
+    },
+    {
+      "epoch": 1.41820516936908,
+      "grad_norm": 0.77828449010849,
+      "learning_rate": 0.00013694224063029396,
+      "loss": 0.8363,
+      "step": 4417
+    },
+    {
+      "epoch": 1.4185262481939316,
+      "grad_norm": 0.9805882573127747,
+      "learning_rate": 0.00013690980103190503,
+      "loss": 0.4089,
+      "step": 4418
+    },
+    {
+      "epoch": 1.4188473270187831,
+      "grad_norm": 0.9755864143371582,
+      "learning_rate": 0.00013687735693616876,
+      "loss": 0.3488,
+      "step": 4419
+    },
+    {
+      "epoch": 1.4191684058436347,
+      "grad_norm": 0.9460292458534241,
+      "learning_rate": 0.00013684490834703843,
+      "loss": 0.3581,
+      "step": 4420
+    },
+    {
+      "epoch": 1.4194894846684862,
+      "grad_norm": 0.8120269179344177,
+      "learning_rate": 0.00013681245526846783,
+      "loss": 0.4829,
+      "step": 4421
+    },
+    {
+      "epoch": 1.4198105634933376,
+      "grad_norm": 0.925631046295166,
+      "learning_rate": 0.00013677999770441115,
+      "loss": 0.7187,
+      "step": 4422
+    },
+    {
+      "epoch": 1.4201316423181891,
+      "grad_norm": 0.8358470797538757,
+      "learning_rate": 0.00013674753565882334,
+      "loss": 0.6196,
+      "step": 4423
+    },
+    {
+      "epoch": 1.4204527211430407,
+      "grad_norm": 1.1134681701660156,
+      "learning_rate": 0.0001367150691356598,
+      "loss": 0.7756,
+      "step": 4424
+    },
+    {
+      "epoch": 1.420773799967892,
+      "grad_norm": 0.7973681688308716,
+      "learning_rate": 0.00013668259813887643,
+      "loss": 0.5704,
+      "step": 4425
+    },
+    {
+      "epoch": 1.4210948787927435,
+      "grad_norm": 0.8833571076393127,
+      "learning_rate": 0.00013665012267242974,
+      "loss": 0.7326,
+      "step": 4426
+    },
+    {
+      "epoch": 1.421415957617595,
+      "grad_norm": 0.85068678855896,
+      "learning_rate": 0.00013661764274027678,
+      "loss": 0.5295,
+      "step": 4427
+    },
+    {
+      "epoch": 1.4217370364424466,
+      "grad_norm": 0.5901229977607727,
+      "learning_rate": 0.00013658515834637512,
+      "loss": 0.4465,
+      "step": 4428
+    },
+    {
+      "epoch": 1.4220581152672982,
+      "grad_norm": 0.7249242067337036,
+      "learning_rate": 0.0001365526694946829,
+      "loss": 0.6021,
+      "step": 4429
+    },
+    {
+      "epoch": 1.4223791940921497,
+      "grad_norm": 0.8554542660713196,
+      "learning_rate": 0.0001365201761891588,
+      "loss": 0.5952,
+      "step": 4430
+    },
+    {
+      "epoch": 1.422700272917001,
+      "grad_norm": 1.1563721895217896,
+      "learning_rate": 0.00013648767843376196,
+      "loss": 0.6391,
+      "step": 4431
+    },
+    {
+      "epoch": 1.4230213517418526,
+      "grad_norm": 0.7757640480995178,
+      "learning_rate": 0.0001364551762324522,
+      "loss": 0.5914,
+      "step": 4432
+    },
+    {
+      "epoch": 1.4233424305667042,
+      "grad_norm": 0.9488760828971863,
+      "learning_rate": 0.00013642266958918984,
+      "loss": 0.678,
+      "step": 4433
+    },
+    {
+      "epoch": 1.4236635093915555,
+      "grad_norm": 0.8043296337127686,
+      "learning_rate": 0.00013639015850793563,
+      "loss": 0.6684,
+      "step": 4434
+    },
+    {
+      "epoch": 1.423984588216407,
+      "grad_norm": 1.0018153190612793,
+      "learning_rate": 0.000136357642992651,
+      "loss": 0.8602,
+      "step": 4435
+    },
+    {
+      "epoch": 1.4243056670412586,
+      "grad_norm": 0.8006235361099243,
+      "learning_rate": 0.00013632512304729785,
+      "loss": 0.6058,
+      "step": 4436
+    },
+    {
+      "epoch": 1.4246267458661102,
+      "grad_norm": 0.9268879890441895,
+      "learning_rate": 0.00013629259867583863,
+      "loss": 0.6974,
+      "step": 4437
+    },
+    {
+      "epoch": 1.4249478246909617,
+      "grad_norm": 0.8567279577255249,
+      "learning_rate": 0.00013626006988223636,
+      "loss": 0.5062,
+      "step": 4438
+    },
+    {
+      "epoch": 1.4252689035158133,
+      "grad_norm": 1.0375492572784424,
+      "learning_rate": 0.00013622753667045457,
+      "loss": 0.7928,
+      "step": 4439
+    },
+    {
+      "epoch": 1.4255899823406646,
+      "grad_norm": 0.6748252511024475,
+      "learning_rate": 0.00013619499904445734,
+      "loss": 0.4775,
+      "step": 4440
+    },
+    {
+      "epoch": 1.4259110611655161,
+      "grad_norm": 0.8363870978355408,
+      "learning_rate": 0.00013616245700820922,
+      "loss": 0.6386,
+      "step": 4441
+    },
+    {
+      "epoch": 1.4262321399903677,
+      "grad_norm": 1.0793986320495605,
+      "learning_rate": 0.00013612991056567543,
+      "loss": 0.5909,
+      "step": 4442
+    },
+    {
+      "epoch": 1.426553218815219,
+      "grad_norm": 0.9321547150611877,
+      "learning_rate": 0.00013609735972082166,
+      "loss": 0.7883,
+      "step": 4443
+    },
+    {
+      "epoch": 1.4268742976400706,
+      "grad_norm": 1.0707225799560547,
+      "learning_rate": 0.00013606480447761409,
+      "loss": 0.6685,
+      "step": 4444
+    },
+    {
+      "epoch": 1.4271953764649221,
+      "grad_norm": 0.9186208248138428,
+      "learning_rate": 0.00013603224484001948,
+      "loss": 0.6659,
+      "step": 4445
+    },
+    {
+      "epoch": 1.4275164552897737,
+      "grad_norm": 1.3622729778289795,
+      "learning_rate": 0.00013599968081200514,
+      "loss": 0.6381,
+      "step": 4446
+    },
+    {
+      "epoch": 1.4278375341146252,
+      "grad_norm": 0.9809760451316833,
+      "learning_rate": 0.0001359671123975389,
+      "loss": 0.5467,
+      "step": 4447
+    },
+    {
+      "epoch": 1.4281586129394768,
+      "grad_norm": 1.0845415592193604,
+      "learning_rate": 0.00013593453960058908,
+      "loss": 0.6937,
+      "step": 4448
+    },
+    {
+      "epoch": 1.428479691764328,
+      "grad_norm": 1.316145896911621,
+      "learning_rate": 0.00013590196242512463,
+      "loss": 0.7624,
+      "step": 4449
+    },
+    {
+      "epoch": 1.4288007705891796,
+      "grad_norm": 0.8910375833511353,
+      "learning_rate": 0.00013586938087511494,
+      "loss": 0.532,
+      "step": 4450
+    },
+    {
+      "epoch": 1.4291218494140312,
+      "grad_norm": 1.1741594076156616,
+      "learning_rate": 0.00013583679495453,
+      "loss": 0.5843,
+      "step": 4451
+    },
+    {
+      "epoch": 1.4294429282388825,
+      "grad_norm": 0.8071622252464294,
+      "learning_rate": 0.00013580420466734037,
+      "loss": 0.5575,
+      "step": 4452
+    },
+    {
+      "epoch": 1.429764007063734,
+      "grad_norm": 1.1299529075622559,
+      "learning_rate": 0.00013577161001751694,
+      "loss": 0.4591,
+      "step": 4453
+    },
+    {
+      "epoch": 1.4300850858885856,
+      "grad_norm": 1.1855186223983765,
+      "learning_rate": 0.00013573901100903134,
+      "loss": 0.6654,
+      "step": 4454
+    },
+    {
+      "epoch": 1.4304061647134372,
+      "grad_norm": 1.0821056365966797,
+      "learning_rate": 0.00013570640764585566,
+      "loss": 0.6234,
+      "step": 4455
+    },
+    {
+      "epoch": 1.4307272435382887,
+      "grad_norm": 1.0219788551330566,
+      "learning_rate": 0.00013567379993196252,
+      "loss": 0.5468,
+      "step": 4456
+    },
+    {
+      "epoch": 1.4310483223631403,
+      "grad_norm": 1.1291626691818237,
+      "learning_rate": 0.00013564118787132506,
+      "loss": 0.782,
+      "step": 4457
+    },
+    {
+      "epoch": 1.4313694011879916,
+      "grad_norm": 0.9877511262893677,
+      "learning_rate": 0.000135608571467917,
+      "loss": 0.5307,
+      "step": 4458
+    },
+    {
+      "epoch": 1.4316904800128432,
+      "grad_norm": 0.9503813982009888,
+      "learning_rate": 0.0001355759507257125,
+      "loss": 0.5951,
+      "step": 4459
+    },
+    {
+      "epoch": 1.4320115588376947,
+      "grad_norm": 0.9377137422561646,
+      "learning_rate": 0.0001355433256486863,
+      "loss": 0.4201,
+      "step": 4460
+    },
+    {
+      "epoch": 1.432332637662546,
+      "grad_norm": 0.9886158108711243,
+      "learning_rate": 0.0001355106962408137,
+      "loss": 0.5017,
+      "step": 4461
+    },
+    {
+      "epoch": 1.4326537164873976,
+      "grad_norm": 1.0197458267211914,
+      "learning_rate": 0.0001354780625060705,
+      "loss": 0.4774,
+      "step": 4462
+    },
+    {
+      "epoch": 1.4329747953122491,
+      "grad_norm": 0.9758638143539429,
+      "learning_rate": 0.00013544542444843299,
+      "loss": 0.4103,
+      "step": 4463
+    },
+    {
+      "epoch": 1.4332958741371007,
+      "grad_norm": 0.7870468497276306,
+      "learning_rate": 0.000135412782071878,
+      "loss": 0.4326,
+      "step": 4464
+    },
+    {
+      "epoch": 1.4336169529619522,
+      "grad_norm": 0.738239049911499,
+      "learning_rate": 0.00013538013538038295,
+      "loss": 0.8734,
+      "step": 4465
+    },
+    {
+      "epoch": 1.4339380317868038,
+      "grad_norm": 0.6055427193641663,
+      "learning_rate": 0.00013534748437792573,
+      "loss": 0.7507,
+      "step": 4466
+    },
+    {
+      "epoch": 1.4342591106116551,
+      "grad_norm": 0.7751237154006958,
+      "learning_rate": 0.00013531482906848475,
+      "loss": 0.3607,
+      "step": 4467
+    },
+    {
+      "epoch": 1.4345801894365067,
+      "grad_norm": 0.7698683738708496,
+      "learning_rate": 0.000135282169456039,
+      "loss": 0.3953,
+      "step": 4468
+    },
+    {
+      "epoch": 1.4349012682613582,
+      "grad_norm": 0.8559454679489136,
+      "learning_rate": 0.00013524950554456784,
+      "loss": 0.4011,
+      "step": 4469
+    },
+    {
+      "epoch": 1.4352223470862095,
+      "grad_norm": 1.0439728498458862,
+      "learning_rate": 0.00013521683733805145,
+      "loss": 0.775,
+      "step": 4470
+    },
+    {
+      "epoch": 1.435543425911061,
+      "grad_norm": 0.8648383021354675,
+      "learning_rate": 0.00013518416484047018,
+      "loss": 0.6064,
+      "step": 4471
+    },
+    {
+      "epoch": 1.4358645047359127,
+      "grad_norm": 1.1003457307815552,
+      "learning_rate": 0.0001351514880558052,
+      "loss": 0.5959,
+      "step": 4472
+    },
+    {
+      "epoch": 1.4361855835607642,
+      "grad_norm": 0.8835905194282532,
+      "learning_rate": 0.00013511880698803801,
+      "loss": 0.579,
+      "step": 4473
+    },
+    {
+      "epoch": 1.4365066623856158,
+      "grad_norm": 0.654793381690979,
+      "learning_rate": 0.00013508612164115068,
+      "loss": 0.4533,
+      "step": 4474
+    },
+    {
+      "epoch": 1.4368277412104673,
+      "grad_norm": 0.832150399684906,
+      "learning_rate": 0.0001350534320191259,
+      "loss": 0.59,
+      "step": 4475
+    },
+    {
+      "epoch": 1.4371488200353186,
+      "grad_norm": 0.805191695690155,
+      "learning_rate": 0.00013502073812594675,
+      "loss": 0.5468,
+      "step": 4476
+    },
+    {
+      "epoch": 1.4374698988601702,
+      "grad_norm": 1.0974748134613037,
+      "learning_rate": 0.00013498803996559692,
+      "loss": 0.8118,
+      "step": 4477
+    },
+    {
+      "epoch": 1.4377909776850217,
+      "grad_norm": 1.0176666975021362,
+      "learning_rate": 0.0001349553375420605,
+      "loss": 0.6775,
+      "step": 4478
+    },
+    {
+      "epoch": 1.438112056509873,
+      "grad_norm": 1.417904019355774,
+      "learning_rate": 0.00013492263085932224,
+      "loss": 0.7306,
+      "step": 4479
+    },
+    {
+      "epoch": 1.4384331353347246,
+      "grad_norm": 1.5575790405273438,
+      "learning_rate": 0.00013488991992136734,
+      "loss": 0.7895,
+      "step": 4480
+    },
+    {
+      "epoch": 1.4387542141595762,
+      "grad_norm": 0.9798082113265991,
+      "learning_rate": 0.00013485720473218154,
+      "loss": 0.6748,
+      "step": 4481
+    },
+    {
+      "epoch": 1.4390752929844277,
+      "grad_norm": 1.1362134218215942,
+      "learning_rate": 0.00013482448529575106,
+      "loss": 0.7912,
+      "step": 4482
+    },
+    {
+      "epoch": 1.4393963718092793,
+      "grad_norm": 0.7490417957305908,
+      "learning_rate": 0.0001347917616160627,
+      "loss": 0.549,
+      "step": 4483
+    },
+    {
+      "epoch": 1.4397174506341308,
+      "grad_norm": 1.027808427810669,
+      "learning_rate": 0.0001347590336971037,
+      "loss": 0.7783,
+      "step": 4484
+    },
+    {
+      "epoch": 1.4400385294589821,
+      "grad_norm": 0.8055115938186646,
+      "learning_rate": 0.0001347263015428619,
+      "loss": 0.5538,
+      "step": 4485
+    },
+    {
+      "epoch": 1.4403596082838337,
+      "grad_norm": 1.0068027973175049,
+      "learning_rate": 0.00013469356515732558,
+      "loss": 0.8646,
+      "step": 4486
+    },
+    {
+      "epoch": 1.4406806871086852,
+      "grad_norm": 1.301796555519104,
+      "learning_rate": 0.00013466082454448362,
+      "loss": 0.7354,
+      "step": 4487
+    },
+    {
+      "epoch": 1.4410017659335366,
+      "grad_norm": 0.9036690592765808,
+      "learning_rate": 0.0001346280797083253,
+      "loss": 0.5607,
+      "step": 4488
+    },
+    {
+      "epoch": 1.4413228447583881,
+      "grad_norm": 0.928037703037262,
+      "learning_rate": 0.0001345953306528405,
+      "loss": 0.7001,
+      "step": 4489
+    },
+    {
+      "epoch": 1.4416439235832397,
+      "grad_norm": 0.9121316075325012,
+      "learning_rate": 0.00013456257738201957,
+      "loss": 0.561,
+      "step": 4490
+    },
+    {
+      "epoch": 1.4419650024080912,
+      "grad_norm": 1.3250795602798462,
+      "learning_rate": 0.00013452981989985348,
+      "loss": 0.81,
+      "step": 4491
+    },
+    {
+      "epoch": 1.4422860812329428,
+      "grad_norm": 0.8794850707054138,
+      "learning_rate": 0.00013449705821033355,
+      "loss": 0.7304,
+      "step": 4492
+    },
+    {
+      "epoch": 1.4426071600577943,
+      "grad_norm": 1.1400325298309326,
+      "learning_rate": 0.0001344642923174517,
+      "loss": 0.6693,
+      "step": 4493
+    },
+    {
+      "epoch": 1.4429282388826457,
+      "grad_norm": 1.1574450731277466,
+      "learning_rate": 0.00013443152222520038,
+      "loss": 0.5701,
+      "step": 4494
+    },
+    {
+      "epoch": 1.4432493177074972,
+      "grad_norm": 1.0430783033370972,
+      "learning_rate": 0.00013439874793757254,
+      "loss": 0.7403,
+      "step": 4495
+    },
+    {
+      "epoch": 1.4435703965323488,
+      "grad_norm": 0.968147337436676,
+      "learning_rate": 0.00013436596945856164,
+      "loss": 0.7173,
+      "step": 4496
+    },
+    {
+      "epoch": 1.4438914753572,
+      "grad_norm": 0.7893198132514954,
+      "learning_rate": 0.00013433318679216153,
+      "loss": 0.5111,
+      "step": 4497
+    },
+    {
+      "epoch": 1.4442125541820516,
+      "grad_norm": 0.9785538911819458,
+      "learning_rate": 0.00013430039994236678,
+      "loss": 0.5591,
+      "step": 4498
+    },
+    {
+      "epoch": 1.4445336330069032,
+      "grad_norm": 1.5484408140182495,
+      "learning_rate": 0.00013426760891317236,
+      "loss": 0.7088,
+      "step": 4499
+    },
+    {
+      "epoch": 1.4448547118317547,
+      "grad_norm": 0.9762724041938782,
+      "learning_rate": 0.00013423481370857375,
+      "loss": 0.5886,
+      "step": 4500
+    },
+    {
+      "epoch": 1.4451757906566063,
+      "grad_norm": 1.3003888130187988,
+      "learning_rate": 0.00013420201433256689,
+      "loss": 0.7262,
+      "step": 4501
+    },
+    {
+      "epoch": 1.4454968694814578,
+      "grad_norm": 0.7951440811157227,
+      "learning_rate": 0.00013416921078914835,
+      "loss": 0.4705,
+      "step": 4502
+    },
+    {
+      "epoch": 1.4458179483063092,
+      "grad_norm": 0.8316115140914917,
+      "learning_rate": 0.0001341364030823151,
+      "loss": 0.5546,
+      "step": 4503
+    },
+    {
+      "epoch": 1.4461390271311607,
+      "grad_norm": 1.0351200103759766,
+      "learning_rate": 0.0001341035912160647,
+      "loss": 0.5174,
+      "step": 4504
+    },
+    {
+      "epoch": 1.4464601059560123,
+      "grad_norm": 1.0522757768630981,
+      "learning_rate": 0.0001340707751943952,
+      "loss": 0.6695,
+      "step": 4505
+    },
+    {
+      "epoch": 1.4467811847808636,
+      "grad_norm": 0.8904996514320374,
+      "learning_rate": 0.00013403795502130503,
+      "loss": 0.5893,
+      "step": 4506
+    },
+    {
+      "epoch": 1.4471022636057151,
+      "grad_norm": 0.866426408290863,
+      "learning_rate": 0.0001340051307007933,
+      "loss": 0.6581,
+      "step": 4507
+    },
+    {
+      "epoch": 1.4474233424305667,
+      "grad_norm": 0.922524094581604,
+      "learning_rate": 0.00013397230223685956,
+      "loss": 0.5626,
+      "step": 4508
+    },
+    {
+      "epoch": 1.4477444212554182,
+      "grad_norm": 1.0396236181259155,
+      "learning_rate": 0.00013393946963350382,
+      "loss": 0.524,
+      "step": 4509
+    },
+    {
+      "epoch": 1.4480655000802698,
+      "grad_norm": 1.3418883085250854,
+      "learning_rate": 0.00013390663289472666,
+      "loss": 0.6468,
+      "step": 4510
+    },
+    {
+      "epoch": 1.4483865789051211,
+      "grad_norm": 0.7169436812400818,
+      "learning_rate": 0.00013387379202452917,
+      "loss": 0.4527,
+      "step": 4511
+    },
+    {
+      "epoch": 1.4487076577299727,
+      "grad_norm": 1.3441863059997559,
+      "learning_rate": 0.00013384094702691281,
+      "loss": 0.5522,
+      "step": 4512
+    },
+    {
+      "epoch": 1.4490287365548242,
+      "grad_norm": 0.7752569913864136,
+      "learning_rate": 0.00013380809790587974,
+      "loss": 0.441,
+      "step": 4513
+    },
+    {
+      "epoch": 1.4493498153796758,
+      "grad_norm": 0.8399336934089661,
+      "learning_rate": 0.00013377524466543248,
+      "loss": 0.4828,
+      "step": 4514
+    },
+    {
+      "epoch": 1.449670894204527,
+      "grad_norm": 0.9468197822570801,
+      "learning_rate": 0.00013374238730957412,
+      "loss": 0.6818,
+      "step": 4515
+    },
+    {
+      "epoch": 1.4499919730293787,
+      "grad_norm": 0.5419335961341858,
+      "learning_rate": 0.00013370952584230823,
+      "loss": 0.7663,
+      "step": 4516
+    },
+    {
+      "epoch": 1.4503130518542302,
+      "grad_norm": 0.7655202150344849,
+      "learning_rate": 0.00013367666026763882,
+      "loss": 0.6944,
+      "step": 4517
+    },
+    {
+      "epoch": 1.4506341306790818,
+      "grad_norm": 0.7268679141998291,
+      "learning_rate": 0.00013364379058957056,
+      "loss": 0.3399,
+      "step": 4518
+    },
+    {
+      "epoch": 1.4509552095039333,
+      "grad_norm": 0.787480354309082,
+      "learning_rate": 0.00013361091681210845,
+      "loss": 0.4652,
+      "step": 4519
+    },
+    {
+      "epoch": 1.4512762883287846,
+      "grad_norm": 0.9196494221687317,
+      "learning_rate": 0.00013357803893925807,
+      "loss": 0.5813,
+      "step": 4520
+    },
+    {
+      "epoch": 1.4515973671536362,
+      "grad_norm": 0.7516376376152039,
+      "learning_rate": 0.00013354515697502553,
+      "loss": 0.3046,
+      "step": 4521
+    },
+    {
+      "epoch": 1.4519184459784877,
+      "grad_norm": 0.7327350974082947,
+      "learning_rate": 0.00013351227092341732,
+      "loss": 0.6606,
+      "step": 4522
+    },
+    {
+      "epoch": 1.4522395248033393,
+      "grad_norm": 0.9102757573127747,
+      "learning_rate": 0.0001334793807884406,
+      "loss": 0.7293,
+      "step": 4523
+    },
+    {
+      "epoch": 1.4525606036281906,
+      "grad_norm": 0.8285297155380249,
+      "learning_rate": 0.00013344648657410282,
+      "loss": 0.5445,
+      "step": 4524
+    },
+    {
+      "epoch": 1.4528816824530422,
+      "grad_norm": 1.0268747806549072,
+      "learning_rate": 0.00013341358828441218,
+      "loss": 0.7724,
+      "step": 4525
+    },
+    {
+      "epoch": 1.4532027612778937,
+      "grad_norm": 0.7675826549530029,
+      "learning_rate": 0.0001333806859233771,
+      "loss": 0.5759,
+      "step": 4526
+    },
+    {
+      "epoch": 1.4535238401027453,
+      "grad_norm": 0.8815962672233582,
+      "learning_rate": 0.00013334777949500673,
+      "loss": 0.7797,
+      "step": 4527
+    },
+    {
+      "epoch": 1.4538449189275968,
+      "grad_norm": 0.7846258282661438,
+      "learning_rate": 0.00013331486900331057,
+      "loss": 0.5603,
+      "step": 4528
+    },
+    {
+      "epoch": 1.4541659977524481,
+      "grad_norm": 1.0539195537567139,
+      "learning_rate": 0.00013328195445229868,
+      "loss": 0.7525,
+      "step": 4529
+    },
+    {
+      "epoch": 1.4544870765772997,
+      "grad_norm": 0.9988424777984619,
+      "learning_rate": 0.0001332490358459816,
+      "loss": 0.7323,
+      "step": 4530
+    },
+    {
+      "epoch": 1.4548081554021512,
+      "grad_norm": 0.7499678730964661,
+      "learning_rate": 0.00013321611318837032,
+      "loss": 0.6056,
+      "step": 4531
+    },
+    {
+      "epoch": 1.4551292342270028,
+      "grad_norm": 0.7443246841430664,
+      "learning_rate": 0.00013318318648347646,
+      "loss": 0.4723,
+      "step": 4532
+    },
+    {
+      "epoch": 1.4554503130518541,
+      "grad_norm": 0.7882360219955444,
+      "learning_rate": 0.00013315025573531198,
+      "loss": 0.5804,
+      "step": 4533
+    },
+    {
+      "epoch": 1.4557713918767057,
+      "grad_norm": 0.9871779680252075,
+      "learning_rate": 0.00013311732094788937,
+      "loss": 0.715,
+      "step": 4534
+    },
+    {
+      "epoch": 1.4560924707015572,
+      "grad_norm": 0.8856052756309509,
+      "learning_rate": 0.00013308438212522163,
+      "loss": 0.6614,
+      "step": 4535
+    },
+    {
+      "epoch": 1.4564135495264088,
+      "grad_norm": 0.9039063453674316,
+      "learning_rate": 0.0001330514392713223,
+      "loss": 0.7465,
+      "step": 4536
+    },
+    {
+      "epoch": 1.4567346283512603,
+      "grad_norm": 0.7212230563163757,
+      "learning_rate": 0.00013301849239020537,
+      "loss": 0.6114,
+      "step": 4537
+    },
+    {
+      "epoch": 1.4570557071761117,
+      "grad_norm": 0.8474130630493164,
+      "learning_rate": 0.00013298554148588528,
+      "loss": 0.5634,
+      "step": 4538
+    },
+    {
+      "epoch": 1.4573767860009632,
+      "grad_norm": 1.1321617364883423,
+      "learning_rate": 0.00013295258656237702,
+      "loss": 0.6322,
+      "step": 4539
+    },
+    {
+      "epoch": 1.4576978648258148,
+      "grad_norm": 1.0911418199539185,
+      "learning_rate": 0.00013291962762369608,
+      "loss": 0.6877,
+      "step": 4540
+    },
+    {
+      "epoch": 1.4580189436506663,
+      "grad_norm": 0.9263336658477783,
+      "learning_rate": 0.00013288666467385833,
+      "loss": 0.5769,
+      "step": 4541
+    },
+    {
+      "epoch": 1.4583400224755176,
+      "grad_norm": 0.677189826965332,
+      "learning_rate": 0.00013285369771688025,
+      "loss": 0.5574,
+      "step": 4542
+    },
+    {
+      "epoch": 1.4586611013003692,
+      "grad_norm": 0.9409946203231812,
+      "learning_rate": 0.0001328207267567788,
+      "loss": 0.6771,
+      "step": 4543
+    },
+    {
+      "epoch": 1.4589821801252207,
+      "grad_norm": 0.848575234413147,
+      "learning_rate": 0.00013278775179757131,
+      "loss": 0.5214,
+      "step": 4544
+    },
+    {
+      "epoch": 1.4593032589500723,
+      "grad_norm": 0.8765393495559692,
+      "learning_rate": 0.0001327547728432757,
+      "loss": 0.5847,
+      "step": 4545
+    },
+    {
+      "epoch": 1.4596243377749238,
+      "grad_norm": 0.8433587551116943,
+      "learning_rate": 0.0001327217898979104,
+      "loss": 0.4884,
+      "step": 4546
+    },
+    {
+      "epoch": 1.4599454165997752,
+      "grad_norm": 1.2517951726913452,
+      "learning_rate": 0.00013268880296549425,
+      "loss": 0.5431,
+      "step": 4547
+    },
+    {
+      "epoch": 1.4602664954246267,
+      "grad_norm": 0.8801559209823608,
+      "learning_rate": 0.00013265581205004661,
+      "loss": 0.5063,
+      "step": 4548
+    },
+    {
+      "epoch": 1.4605875742494783,
+      "grad_norm": 1.0915852785110474,
+      "learning_rate": 0.00013262281715558736,
+      "loss": 0.6934,
+      "step": 4549
+    },
+    {
+      "epoch": 1.4609086530743298,
+      "grad_norm": 1.139147400856018,
+      "learning_rate": 0.00013258981828613678,
+      "loss": 0.6021,
+      "step": 4550
+    },
+    {
+      "epoch": 1.4612297318991812,
+      "grad_norm": 1.0436866283416748,
+      "learning_rate": 0.00013255681544571568,
+      "loss": 0.8493,
+      "step": 4551
+    },
+    {
+      "epoch": 1.4615508107240327,
+      "grad_norm": 0.8446107506752014,
+      "learning_rate": 0.0001325238086383454,
+      "loss": 0.4369,
+      "step": 4552
+    },
+    {
+      "epoch": 1.4618718895488843,
+      "grad_norm": 2.1884477138519287,
+      "learning_rate": 0.00013249079786804765,
+      "loss": 0.6191,
+      "step": 4553
+    },
+    {
+      "epoch": 1.4621929683737358,
+      "grad_norm": 0.8910212516784668,
+      "learning_rate": 0.00013245778313884478,
+      "loss": 0.4856,
+      "step": 4554
+    },
+    {
+      "epoch": 1.4625140471985874,
+      "grad_norm": 0.6236334443092346,
+      "learning_rate": 0.00013242476445475944,
+      "loss": 0.4026,
+      "step": 4555
+    },
+    {
+      "epoch": 1.4628351260234387,
+      "grad_norm": 1.155137538909912,
+      "learning_rate": 0.00013239174181981495,
+      "loss": 0.7141,
+      "step": 4556
+    },
+    {
+      "epoch": 1.4631562048482902,
+      "grad_norm": 1.5037484169006348,
+      "learning_rate": 0.00013235871523803496,
+      "loss": 0.7128,
+      "step": 4557
+    },
+    {
+      "epoch": 1.4634772836731418,
+      "grad_norm": 0.9598552584648132,
+      "learning_rate": 0.0001323256847134437,
+      "loss": 0.6375,
+      "step": 4558
+    },
+    {
+      "epoch": 1.4637983624979933,
+      "grad_norm": 0.9860318303108215,
+      "learning_rate": 0.00013229265025006576,
+      "loss": 0.572,
+      "step": 4559
+    },
+    {
+      "epoch": 1.4641194413228447,
+      "grad_norm": 0.9279815554618835,
+      "learning_rate": 0.00013225961185192638,
+      "loss": 0.4483,
+      "step": 4560
+    },
+    {
+      "epoch": 1.4644405201476962,
+      "grad_norm": 0.8800573945045471,
+      "learning_rate": 0.00013222656952305113,
+      "loss": 0.5297,
+      "step": 4561
+    },
+    {
+      "epoch": 1.4647615989725478,
+      "grad_norm": 1.1138170957565308,
+      "learning_rate": 0.00013219352326746613,
+      "loss": 0.5575,
+      "step": 4562
+    },
+    {
+      "epoch": 1.4650826777973993,
+      "grad_norm": 0.760184645652771,
+      "learning_rate": 0.00013216047308919798,
+      "loss": 0.4005,
+      "step": 4563
+    },
+    {
+      "epoch": 1.4654037566222509,
+      "grad_norm": 0.6601788997650146,
+      "learning_rate": 0.00013212741899227374,
+      "loss": 0.3951,
+      "step": 4564
+    },
+    {
+      "epoch": 1.4657248354471022,
+      "grad_norm": 0.825016438961029,
+      "learning_rate": 0.00013209436098072095,
+      "loss": 0.7009,
+      "step": 4565
+    },
+    {
+      "epoch": 1.4660459142719537,
+      "grad_norm": 0.652347981929779,
+      "learning_rate": 0.00013206129905856765,
+      "loss": 0.6756,
+      "step": 4566
+    },
+    {
+      "epoch": 1.4663669930968053,
+      "grad_norm": 0.5992451906204224,
+      "learning_rate": 0.00013202823322984228,
+      "loss": 0.6338,
+      "step": 4567
+    },
+    {
+      "epoch": 1.4666880719216568,
+      "grad_norm": 0.8534404039382935,
+      "learning_rate": 0.00013199516349857385,
+      "loss": 0.5359,
+      "step": 4568
+    },
+    {
+      "epoch": 1.4670091507465082,
+      "grad_norm": 0.7051871418952942,
+      "learning_rate": 0.0001319620898687918,
+      "loss": 0.3355,
+      "step": 4569
+    },
+    {
+      "epoch": 1.4673302295713597,
+      "grad_norm": 0.9312176704406738,
+      "learning_rate": 0.00013192901234452607,
+      "loss": 0.4381,
+      "step": 4570
+    },
+    {
+      "epoch": 1.4676513083962113,
+      "grad_norm": 1.0084874629974365,
+      "learning_rate": 0.00013189593092980702,
+      "loss": 0.3896,
+      "step": 4571
+    },
+    {
+      "epoch": 1.4679723872210628,
+      "grad_norm": 0.9059427380561829,
+      "learning_rate": 0.00013186284562866554,
+      "loss": 0.3917,
+      "step": 4572
+    },
+    {
+      "epoch": 1.4682934660459144,
+      "grad_norm": 0.9761437773704529,
+      "learning_rate": 0.00013182975644513296,
+      "loss": 0.6641,
+      "step": 4573
+    },
+    {
+      "epoch": 1.4686145448707657,
+      "grad_norm": 0.8613228797912598,
+      "learning_rate": 0.00013179666338324108,
+      "loss": 0.5554,
+      "step": 4574
+    },
+    {
+      "epoch": 1.4689356236956173,
+      "grad_norm": 0.9337467551231384,
+      "learning_rate": 0.00013176356644702225,
+      "loss": 0.82,
+      "step": 4575
+    },
+    {
+      "epoch": 1.4692567025204688,
+      "grad_norm": 0.6009548306465149,
+      "learning_rate": 0.00013173046564050924,
+      "loss": 0.4591,
+      "step": 4576
+    },
+    {
+      "epoch": 1.4695777813453204,
+      "grad_norm": 0.8322634100914001,
+      "learning_rate": 0.0001316973609677352,
+      "loss": 0.6572,
+      "step": 4577
+    },
+    {
+      "epoch": 1.4698988601701717,
+      "grad_norm": 0.9558845162391663,
+      "learning_rate": 0.00013166425243273385,
+      "loss": 0.7128,
+      "step": 4578
+    },
+    {
+      "epoch": 1.4702199389950232,
+      "grad_norm": 0.9433383345603943,
+      "learning_rate": 0.0001316311400395394,
+      "loss": 0.6818,
+      "step": 4579
+    },
+    {
+      "epoch": 1.4705410178198748,
+      "grad_norm": 0.7871822118759155,
+      "learning_rate": 0.0001315980237921865,
+      "loss": 0.6238,
+      "step": 4580
+    },
+    {
+      "epoch": 1.4708620966447263,
+      "grad_norm": 0.8970574140548706,
+      "learning_rate": 0.00013156490369471027,
+      "loss": 0.6878,
+      "step": 4581
+    },
+    {
+      "epoch": 1.4711831754695779,
+      "grad_norm": 1.0396785736083984,
+      "learning_rate": 0.00013153177975114625,
+      "loss": 0.7369,
+      "step": 4582
+    },
+    {
+      "epoch": 1.4715042542944292,
+      "grad_norm": 0.8116987943649292,
+      "learning_rate": 0.0001314986519655305,
+      "loss": 0.6017,
+      "step": 4583
+    },
+    {
+      "epoch": 1.4718253331192808,
+      "grad_norm": 1.061795711517334,
+      "learning_rate": 0.00013146552034189955,
+      "loss": 0.5786,
+      "step": 4584
+    },
+    {
+      "epoch": 1.4721464119441323,
+      "grad_norm": 0.9677996039390564,
+      "learning_rate": 0.0001314323848842904,
+      "loss": 0.4793,
+      "step": 4585
+    },
+    {
+      "epoch": 1.4724674907689839,
+      "grad_norm": 0.8710011839866638,
+      "learning_rate": 0.00013139924559674052,
+      "loss": 0.6054,
+      "step": 4586
+    },
+    {
+      "epoch": 1.4727885695938352,
+      "grad_norm": 0.9556978344917297,
+      "learning_rate": 0.00013136610248328778,
+      "loss": 0.6016,
+      "step": 4587
+    },
+    {
+      "epoch": 1.4731096484186867,
+      "grad_norm": 1.0169155597686768,
+      "learning_rate": 0.0001313329555479706,
+      "loss": 0.5464,
+      "step": 4588
+    },
+    {
+      "epoch": 1.4734307272435383,
+      "grad_norm": 0.9183655977249146,
+      "learning_rate": 0.00013129980479482782,
+      "loss": 0.7172,
+      "step": 4589
+    },
+    {
+      "epoch": 1.4737518060683898,
+      "grad_norm": 0.7591384053230286,
+      "learning_rate": 0.00013126665022789879,
+      "loss": 0.6191,
+      "step": 4590
+    },
+    {
+      "epoch": 1.4740728848932414,
+      "grad_norm": 0.8953664898872375,
+      "learning_rate": 0.00013123349185122327,
+      "loss": 0.532,
+      "step": 4591
+    },
+    {
+      "epoch": 1.4743939637180927,
+      "grad_norm": 1.7143136262893677,
+      "learning_rate": 0.0001312003296688415,
+      "loss": 0.581,
+      "step": 4592
+    },
+    {
+      "epoch": 1.4747150425429443,
+      "grad_norm": 0.8613825440406799,
+      "learning_rate": 0.0001311671636847942,
+      "loss": 0.5599,
+      "step": 4593
+    },
+    {
+      "epoch": 1.4750361213677958,
+      "grad_norm": 0.901800274848938,
+      "learning_rate": 0.00013113399390312256,
+      "loss": 0.5636,
+      "step": 4594
+    },
+    {
+      "epoch": 1.4753572001926474,
+      "grad_norm": 0.9919823408126831,
+      "learning_rate": 0.0001311008203278682,
+      "loss": 0.6591,
+      "step": 4595
+    },
+    {
+      "epoch": 1.4756782790174987,
+      "grad_norm": 1.3764430284500122,
+      "learning_rate": 0.0001310676429630732,
+      "loss": 0.6994,
+      "step": 4596
+    },
+    {
+      "epoch": 1.4759993578423503,
+      "grad_norm": 1.025925874710083,
+      "learning_rate": 0.00013103446181278013,
+      "loss": 0.571,
+      "step": 4597
+    },
+    {
+      "epoch": 1.4763204366672018,
+      "grad_norm": 0.9428420662879944,
+      "learning_rate": 0.00013100127688103205,
+      "loss": 0.6081,
+      "step": 4598
+    },
+    {
+      "epoch": 1.4766415154920534,
+      "grad_norm": 0.8529512882232666,
+      "learning_rate": 0.00013096808817187242,
+      "loss": 0.4775,
+      "step": 4599
+    },
+    {
+      "epoch": 1.476962594316905,
+      "grad_norm": 1.0675090551376343,
+      "learning_rate": 0.0001309348956893452,
+      "loss": 0.7771,
+      "step": 4600
+    },
+    {
+      "epoch": 1.4772836731417562,
+      "grad_norm": 1.3736308813095093,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 0.6906,
+      "step": 4601
+    },
+    {
+      "epoch": 1.4776047519666078,
+      "grad_norm": 1.8220711946487427,
+      "learning_rate": 0.00013086849942036595,
+      "loss": 0.704,
+      "step": 4602
+    },
+    {
+      "epoch": 1.4779258307914593,
+      "grad_norm": 0.9620269536972046,
+      "learning_rate": 0.00013083529564200417,
+      "loss": 0.526,
+      "step": 4603
+    },
+    {
+      "epoch": 1.4782469096163107,
+      "grad_norm": 0.6493226289749146,
+      "learning_rate": 0.00013080208810645514,
+      "loss": 0.3707,
+      "step": 4604
+    },
+    {
+      "epoch": 1.4785679884411622,
+      "grad_norm": 1.0979892015457153,
+      "learning_rate": 0.0001307688768177651,
+      "loss": 0.6589,
+      "step": 4605
+    },
+    {
+      "epoch": 1.4788890672660138,
+      "grad_norm": 0.8274226784706116,
+      "learning_rate": 0.00013073566177998074,
+      "loss": 0.5467,
+      "step": 4606
+    },
+    {
+      "epoch": 1.4792101460908653,
+      "grad_norm": 1.2158207893371582,
+      "learning_rate": 0.0001307024429971492,
+      "loss": 0.5996,
+      "step": 4607
+    },
+    {
+      "epoch": 1.4795312249157169,
+      "grad_norm": 1.105692982673645,
+      "learning_rate": 0.00013066922047331813,
+      "loss": 0.5918,
+      "step": 4608
+    },
+    {
+      "epoch": 1.4798523037405684,
+      "grad_norm": 0.8877955675125122,
+      "learning_rate": 0.00013063599421253558,
+      "loss": 0.5614,
+      "step": 4609
+    },
+    {
+      "epoch": 1.4801733825654197,
+      "grad_norm": 0.8180930018424988,
+      "learning_rate": 0.0001306027642188501,
+      "loss": 0.5291,
+      "step": 4610
+    },
+    {
+      "epoch": 1.4804944613902713,
+      "grad_norm": 1.009455680847168,
+      "learning_rate": 0.00013056953049631057,
+      "loss": 0.4543,
+      "step": 4611
+    },
+    {
+      "epoch": 1.4808155402151228,
+      "grad_norm": 0.779489278793335,
+      "learning_rate": 0.0001305362930489665,
+      "loss": 0.5272,
+      "step": 4612
+    },
+    {
+      "epoch": 1.4811366190399742,
+      "grad_norm": 0.9974098801612854,
+      "learning_rate": 0.0001305030518808678,
+      "loss": 0.4729,
+      "step": 4613
+    },
+    {
+      "epoch": 1.4814576978648257,
+      "grad_norm": 0.7673447132110596,
+      "learning_rate": 0.00013046980699606467,
+      "loss": 0.345,
+      "step": 4614
+    },
+    {
+      "epoch": 1.4817787766896773,
+      "grad_norm": 0.6975226402282715,
+      "learning_rate": 0.00013043655839860803,
+      "loss": 0.6528,
+      "step": 4615
+    },
+    {
+      "epoch": 1.4820998555145288,
+      "grad_norm": 0.652308464050293,
+      "learning_rate": 0.00013040330609254903,
+      "loss": 0.5996,
+      "step": 4616
+    },
+    {
+      "epoch": 1.4824209343393804,
+      "grad_norm": 0.7014802694320679,
+      "learning_rate": 0.00013037005008193942,
+      "loss": 0.6009,
+      "step": 4617
+    },
+    {
+      "epoch": 1.482742013164232,
+      "grad_norm": 0.8486183881759644,
+      "learning_rate": 0.00013033679037083138,
+      "loss": 0.5923,
+      "step": 4618
+    },
+    {
+      "epoch": 1.4830630919890833,
+      "grad_norm": 0.7879297137260437,
+      "learning_rate": 0.00013030352696327742,
+      "loss": 0.3314,
+      "step": 4619
+    },
+    {
+      "epoch": 1.4833841708139348,
+      "grad_norm": 0.6520053148269653,
+      "learning_rate": 0.0001302702598633306,
+      "loss": 0.3041,
+      "step": 4620
+    },
+    {
+      "epoch": 1.4837052496387864,
+      "grad_norm": 0.8596171736717224,
+      "learning_rate": 0.00013023698907504446,
+      "loss": 0.3313,
+      "step": 4621
+    },
+    {
+      "epoch": 1.4840263284636377,
+      "grad_norm": 0.5723639130592346,
+      "learning_rate": 0.0001302037146024729,
+      "loss": 0.188,
+      "step": 4622
+    },
+    {
+      "epoch": 1.4843474072884892,
+      "grad_norm": 0.7225738167762756,
+      "learning_rate": 0.00013017043644967036,
+      "loss": 0.3561,
+      "step": 4623
+    },
+    {
+      "epoch": 1.4846684861133408,
+      "grad_norm": 0.9386597871780396,
+      "learning_rate": 0.00013013715462069166,
+      "loss": 0.7271,
+      "step": 4624
+    },
+    {
+      "epoch": 1.4849895649381923,
+      "grad_norm": 1.0117478370666504,
+      "learning_rate": 0.00013010386911959206,
+      "loss": 0.806,
+      "step": 4625
+    },
+    {
+      "epoch": 1.485310643763044,
+      "grad_norm": 0.9228258728981018,
+      "learning_rate": 0.00013007057995042732,
+      "loss": 0.5942,
+      "step": 4626
+    },
+    {
+      "epoch": 1.4856317225878954,
+      "grad_norm": 0.9322078227996826,
+      "learning_rate": 0.0001300372871172536,
+      "loss": 0.6389,
+      "step": 4627
+    },
+    {
+      "epoch": 1.4859528014127468,
+      "grad_norm": 1.0570144653320312,
+      "learning_rate": 0.00013000399062412763,
+      "loss": 0.6589,
+      "step": 4628
+    },
+    {
+      "epoch": 1.4862738802375983,
+      "grad_norm": 0.8913359642028809,
+      "learning_rate": 0.0001299706904751064,
+      "loss": 0.6874,
+      "step": 4629
+    },
+    {
+      "epoch": 1.4865949590624499,
+      "grad_norm": 0.8339388966560364,
+      "learning_rate": 0.0001299373866742474,
+      "loss": 0.6398,
+      "step": 4630
+    },
+    {
+      "epoch": 1.4869160378873012,
+      "grad_norm": 0.6980022192001343,
+      "learning_rate": 0.00012990407922560868,
+      "loss": 0.4593,
+      "step": 4631
+    },
+    {
+      "epoch": 1.4872371167121528,
+      "grad_norm": 0.9570392370223999,
+      "learning_rate": 0.00012987076813324858,
+      "loss": 0.5951,
+      "step": 4632
+    },
+    {
+      "epoch": 1.4875581955370043,
+      "grad_norm": 0.8993356227874756,
+      "learning_rate": 0.00012983745340122604,
+      "loss": 0.7096,
+      "step": 4633
+    },
+    {
+      "epoch": 1.4878792743618559,
+      "grad_norm": 0.9646190404891968,
+      "learning_rate": 0.0001298041350336003,
+      "loss": 0.662,
+      "step": 4634
+    },
+    {
+      "epoch": 1.4882003531867074,
+      "grad_norm": 0.8036277294158936,
+      "learning_rate": 0.00012977081303443107,
+      "loss": 0.5762,
+      "step": 4635
+    },
+    {
+      "epoch": 1.488521432011559,
+      "grad_norm": 1.3247047662734985,
+      "learning_rate": 0.0001297374874077786,
+      "loss": 0.6607,
+      "step": 4636
+    },
+    {
+      "epoch": 1.4888425108364103,
+      "grad_norm": 0.9229692816734314,
+      "learning_rate": 0.0001297041581577035,
+      "loss": 0.6021,
+      "step": 4637
+    },
+    {
+      "epoch": 1.4891635896612618,
+      "grad_norm": 0.8703567981719971,
+      "learning_rate": 0.00012967082528826684,
+      "loss": 0.5731,
+      "step": 4638
+    },
+    {
+      "epoch": 1.4894846684861134,
+      "grad_norm": 1.0840383768081665,
+      "learning_rate": 0.00012963748880353011,
+      "loss": 0.6943,
+      "step": 4639
+    },
+    {
+      "epoch": 1.4898057473109647,
+      "grad_norm": 0.9261624217033386,
+      "learning_rate": 0.00012960414870755524,
+      "loss": 0.5193,
+      "step": 4640
+    },
+    {
+      "epoch": 1.4901268261358163,
+      "grad_norm": 1.1405514478683472,
+      "learning_rate": 0.00012957080500440468,
+      "loss": 0.709,
+      "step": 4641
+    },
+    {
+      "epoch": 1.4904479049606678,
+      "grad_norm": 1.0057867765426636,
+      "learning_rate": 0.00012953745769814123,
+      "loss": 0.6075,
+      "step": 4642
+    },
+    {
+      "epoch": 1.4907689837855194,
+      "grad_norm": 0.86097252368927,
+      "learning_rate": 0.00012950410679282815,
+      "loss": 0.4511,
+      "step": 4643
+    },
+    {
+      "epoch": 1.491090062610371,
+      "grad_norm": 1.2975316047668457,
+      "learning_rate": 0.00012947075229252915,
+      "loss": 0.7537,
+      "step": 4644
+    },
+    {
+      "epoch": 1.4914111414352225,
+      "grad_norm": 1.0226067304611206,
+      "learning_rate": 0.00012943739420130836,
+      "loss": 0.497,
+      "step": 4645
+    },
+    {
+      "epoch": 1.4917322202600738,
+      "grad_norm": 1.0293515920639038,
+      "learning_rate": 0.0001294040325232304,
+      "loss": 0.5741,
+      "step": 4646
+    },
+    {
+      "epoch": 1.4920532990849253,
+      "grad_norm": 0.7942186594009399,
+      "learning_rate": 0.0001293706672623603,
+      "loss": 0.5992,
+      "step": 4647
+    },
+    {
+      "epoch": 1.492374377909777,
+      "grad_norm": 0.8032208681106567,
+      "learning_rate": 0.00012933729842276343,
+      "loss": 0.4878,
+      "step": 4648
+    },
+    {
+      "epoch": 1.4926954567346282,
+      "grad_norm": 0.9675914645195007,
+      "learning_rate": 0.00012930392600850573,
+      "loss": 0.5973,
+      "step": 4649
+    },
+    {
+      "epoch": 1.4930165355594798,
+      "grad_norm": 1.0577799081802368,
+      "learning_rate": 0.0001292705500236536,
+      "loss": 0.7173,
+      "step": 4650
+    },
+    {
+      "epoch": 1.4933376143843313,
+      "grad_norm": 0.9797313213348389,
+      "learning_rate": 0.00012923717047227368,
+      "loss": 0.5796,
+      "step": 4651
+    },
+    {
+      "epoch": 1.4936586932091829,
+      "grad_norm": 1.041603446006775,
+      "learning_rate": 0.00012920378735843327,
+      "loss": 0.4257,
+      "step": 4652
+    },
+    {
+      "epoch": 1.4939797720340344,
+      "grad_norm": 0.7669867873191833,
+      "learning_rate": 0.0001291704006861999,
+      "loss": 0.5239,
+      "step": 4653
+    },
+    {
+      "epoch": 1.494300850858886,
+      "grad_norm": 0.8458186388015747,
+      "learning_rate": 0.00012913701045964173,
+      "loss": 0.4882,
+      "step": 4654
+    },
+    {
+      "epoch": 1.4946219296837373,
+      "grad_norm": 0.9558455944061279,
+      "learning_rate": 0.00012910361668282719,
+      "loss": 0.53,
+      "step": 4655
+    },
+    {
+      "epoch": 1.4949430085085889,
+      "grad_norm": 0.9773117899894714,
+      "learning_rate": 0.00012907021935982526,
+      "loss": 0.5155,
+      "step": 4656
+    },
+    {
+      "epoch": 1.4952640873334404,
+      "grad_norm": 1.0766710042953491,
+      "learning_rate": 0.00012903681849470527,
+      "loss": 0.5791,
+      "step": 4657
+    },
+    {
+      "epoch": 1.4955851661582917,
+      "grad_norm": 0.9035495519638062,
+      "learning_rate": 0.00012900341409153703,
+      "loss": 0.5649,
+      "step": 4658
+    },
+    {
+      "epoch": 1.4959062449831433,
+      "grad_norm": 1.3426756858825684,
+      "learning_rate": 0.00012897000615439075,
+      "loss": 0.4312,
+      "step": 4659
+    },
+    {
+      "epoch": 1.4962273238079948,
+      "grad_norm": 1.0395222902297974,
+      "learning_rate": 0.0001289365946873371,
+      "loss": 0.5155,
+      "step": 4660
+    },
+    {
+      "epoch": 1.4965484026328464,
+      "grad_norm": 0.9974014163017273,
+      "learning_rate": 0.00012890317969444716,
+      "loss": 0.6004,
+      "step": 4661
+    },
+    {
+      "epoch": 1.496869481457698,
+      "grad_norm": 0.9123116135597229,
+      "learning_rate": 0.00012886976117979247,
+      "loss": 0.6294,
+      "step": 4662
+    },
+    {
+      "epoch": 1.4971905602825495,
+      "grad_norm": 0.7099243402481079,
+      "learning_rate": 0.00012883633914744492,
+      "loss": 0.3992,
+      "step": 4663
+    },
+    {
+      "epoch": 1.4975116391074008,
+      "grad_norm": 0.47554677724838257,
+      "learning_rate": 0.00012880291360147693,
+      "loss": 0.3285,
+      "step": 4664
+    },
+    {
+      "epoch": 1.4978327179322524,
+      "grad_norm": 0.7324989438056946,
+      "learning_rate": 0.00012876948454596128,
+      "loss": 0.6776,
+      "step": 4665
+    },
+    {
+      "epoch": 1.498153796757104,
+      "grad_norm": 0.677893877029419,
+      "learning_rate": 0.00012873605198497124,
+      "loss": 0.5352,
+      "step": 4666
+    },
+    {
+      "epoch": 1.4984748755819552,
+      "grad_norm": 0.9297817945480347,
+      "learning_rate": 0.00012870261592258037,
+      "loss": 0.451,
+      "step": 4667
+    },
+    {
+      "epoch": 1.4987959544068068,
+      "grad_norm": 0.8155097961425781,
+      "learning_rate": 0.00012866917636286286,
+      "loss": 0.5778,
+      "step": 4668
+    },
+    {
+      "epoch": 1.4991170332316583,
+      "grad_norm": 0.7590045928955078,
+      "learning_rate": 0.00012863573330989313,
+      "loss": 0.3355,
+      "step": 4669
+    },
+    {
+      "epoch": 1.49943811205651,
+      "grad_norm": 0.9300277233123779,
+      "learning_rate": 0.0001286022867677462,
+      "loss": 0.3444,
+      "step": 4670
+    },
+    {
+      "epoch": 1.4997591908813614,
+      "grad_norm": 0.7806175947189331,
+      "learning_rate": 0.00012856883674049736,
+      "loss": 0.4089,
+      "step": 4671
+    },
+    {
+      "epoch": 1.500080269706213,
+      "grad_norm": 0.7708941102027893,
+      "learning_rate": 0.00012853538323222242,
+      "loss": 0.5435,
+      "step": 4672
+    },
+    {
+      "epoch": 1.5004013485310643,
+      "grad_norm": 1.0800435543060303,
+      "learning_rate": 0.0001285019262469976,
+      "loss": 0.6181,
+      "step": 4673
+    },
+    {
+      "epoch": 1.5007224273559159,
+      "grad_norm": 1.2180079221725464,
+      "learning_rate": 0.00012846846578889955,
+      "loss": 0.705,
+      "step": 4674
+    },
+    {
+      "epoch": 1.5010435061807674,
+      "grad_norm": 1.081411600112915,
+      "learning_rate": 0.00012843500186200527,
+      "loss": 0.6284,
+      "step": 4675
+    },
+    {
+      "epoch": 1.5013645850056188,
+      "grad_norm": 0.7021897435188293,
+      "learning_rate": 0.00012840153447039228,
+      "loss": 0.5129,
+      "step": 4676
+    },
+    {
+      "epoch": 1.5016856638304703,
+      "grad_norm": 0.9453624486923218,
+      "learning_rate": 0.00012836806361813844,
+      "loss": 0.5492,
+      "step": 4677
+    },
+    {
+      "epoch": 1.5020067426553219,
+      "grad_norm": 0.9042197465896606,
+      "learning_rate": 0.00012833458930932212,
+      "loss": 0.7457,
+      "step": 4678
+    },
+    {
+      "epoch": 1.5023278214801734,
+      "grad_norm": 1.1903046369552612,
+      "learning_rate": 0.00012830111154802203,
+      "loss": 0.5891,
+      "step": 4679
+    },
+    {
+      "epoch": 1.502648900305025,
+      "grad_norm": 0.779662013053894,
+      "learning_rate": 0.00012826763033831735,
+      "loss": 0.665,
+      "step": 4680
+    },
+    {
+      "epoch": 1.5029699791298765,
+      "grad_norm": 0.9202207326889038,
+      "learning_rate": 0.00012823414568428768,
+      "loss": 0.7502,
+      "step": 4681
+    },
+    {
+      "epoch": 1.5032910579547278,
+      "grad_norm": 0.9345004558563232,
+      "learning_rate": 0.00012820065759001293,
+      "loss": 0.6421,
+      "step": 4682
+    },
+    {
+      "epoch": 1.5036121367795794,
+      "grad_norm": 0.7800152897834778,
+      "learning_rate": 0.00012816716605957367,
+      "loss": 0.5856,
+      "step": 4683
+    },
+    {
+      "epoch": 1.503933215604431,
+      "grad_norm": 0.8760406374931335,
+      "learning_rate": 0.00012813367109705063,
+      "loss": 0.6065,
+      "step": 4684
+    },
+    {
+      "epoch": 1.5042542944292823,
+      "grad_norm": 0.9938950538635254,
+      "learning_rate": 0.00012810017270652513,
+      "loss": 0.7377,
+      "step": 4685
+    },
+    {
+      "epoch": 1.5045753732541338,
+      "grad_norm": 0.9110274910926819,
+      "learning_rate": 0.0001280666708920788,
+      "loss": 0.6981,
+      "step": 4686
+    },
+    {
+      "epoch": 1.5048964520789854,
+      "grad_norm": 0.902837336063385,
+      "learning_rate": 0.00012803316565779377,
+      "loss": 0.6673,
+      "step": 4687
+    },
+    {
+      "epoch": 1.505217530903837,
+      "grad_norm": 0.8513844013214111,
+      "learning_rate": 0.0001279996570077525,
+      "loss": 0.623,
+      "step": 4688
+    },
+    {
+      "epoch": 1.5055386097286885,
+      "grad_norm": 0.8973796963691711,
+      "learning_rate": 0.000127966144946038,
+      "loss": 0.5205,
+      "step": 4689
+    },
+    {
+      "epoch": 1.50585968855354,
+      "grad_norm": 0.9134697318077087,
+      "learning_rate": 0.00012793262947673355,
+      "loss": 0.6839,
+      "step": 4690
+    },
+    {
+      "epoch": 1.5061807673783913,
+      "grad_norm": 1.0323461294174194,
+      "learning_rate": 0.00012789911060392294,
+      "loss": 0.8083,
+      "step": 4691
+    },
+    {
+      "epoch": 1.506501846203243,
+      "grad_norm": 0.8523952960968018,
+      "learning_rate": 0.00012786558833169031,
+      "loss": 0.4945,
+      "step": 4692
+    },
+    {
+      "epoch": 1.5068229250280945,
+      "grad_norm": 1.019515037536621,
+      "learning_rate": 0.0001278320626641203,
+      "loss": 0.5435,
+      "step": 4693
+    },
+    {
+      "epoch": 1.5071440038529458,
+      "grad_norm": 0.8791556358337402,
+      "learning_rate": 0.00012779853360529785,
+      "loss": 0.6037,
+      "step": 4694
+    },
+    {
+      "epoch": 1.5074650826777973,
+      "grad_norm": 0.9916840195655823,
+      "learning_rate": 0.00012776500115930842,
+      "loss": 0.581,
+      "step": 4695
+    },
+    {
+      "epoch": 1.5077861615026489,
+      "grad_norm": 0.7988699674606323,
+      "learning_rate": 0.00012773146533023782,
+      "loss": 0.5452,
+      "step": 4696
+    },
+    {
+      "epoch": 1.5081072403275004,
+      "grad_norm": 0.7620382308959961,
+      "learning_rate": 0.00012769792612217224,
+      "loss": 0.5199,
+      "step": 4697
+    },
+    {
+      "epoch": 1.508428319152352,
+      "grad_norm": 1.141939401626587,
+      "learning_rate": 0.0001276643835391984,
+      "loss": 0.7435,
+      "step": 4698
+    },
+    {
+      "epoch": 1.5087493979772035,
+      "grad_norm": 0.9198951125144958,
+      "learning_rate": 0.00012763083758540337,
+      "loss": 0.6203,
+      "step": 4699
+    },
+    {
+      "epoch": 1.5090704768020549,
+      "grad_norm": 0.7245876789093018,
+      "learning_rate": 0.0001275972882648746,
+      "loss": 0.5134,
+      "step": 4700
+    },
+    {
+      "epoch": 1.5093915556269064,
+      "grad_norm": 0.7252978086471558,
+      "learning_rate": 0.0001275637355816999,
+      "loss": 0.4891,
+      "step": 4701
+    },
+    {
+      "epoch": 1.5097126344517577,
+      "grad_norm": 0.7837473750114441,
+      "learning_rate": 0.0001275301795399677,
+      "loss": 0.4528,
+      "step": 4702
+    },
+    {
+      "epoch": 1.5100337132766093,
+      "grad_norm": 0.8873403668403625,
+      "learning_rate": 0.00012749662014376663,
+      "loss": 0.5449,
+      "step": 4703
+    },
+    {
+      "epoch": 1.5103547921014608,
+      "grad_norm": 0.9590548872947693,
+      "learning_rate": 0.00012746305739718577,
+      "loss": 0.5991,
+      "step": 4704
+    },
+    {
+      "epoch": 1.5106758709263124,
+      "grad_norm": 1.017732858657837,
+      "learning_rate": 0.00012742949130431467,
+      "loss": 0.5145,
+      "step": 4705
+    },
+    {
+      "epoch": 1.510996949751164,
+      "grad_norm": 0.720311164855957,
+      "learning_rate": 0.00012739592186924328,
+      "loss": 0.4315,
+      "step": 4706
+    },
+    {
+      "epoch": 1.5113180285760155,
+      "grad_norm": 0.7741522192955017,
+      "learning_rate": 0.00012736234909606185,
+      "loss": 0.529,
+      "step": 4707
+    },
+    {
+      "epoch": 1.511639107400867,
+      "grad_norm": 1.115466594696045,
+      "learning_rate": 0.00012732877298886124,
+      "loss": 0.6952,
+      "step": 4708
+    },
+    {
+      "epoch": 1.5119601862257184,
+      "grad_norm": 0.8059918880462646,
+      "learning_rate": 0.00012729519355173254,
+      "loss": 0.5197,
+      "step": 4709
+    },
+    {
+      "epoch": 1.51228126505057,
+      "grad_norm": 1.1203827857971191,
+      "learning_rate": 0.0001272616107887673,
+      "loss": 0.5233,
+      "step": 4710
+    },
+    {
+      "epoch": 1.5126023438754213,
+      "grad_norm": 1.13717520236969,
+      "learning_rate": 0.00012722802470405744,
+      "loss": 0.5846,
+      "step": 4711
+    },
+    {
+      "epoch": 1.5129234227002728,
+      "grad_norm": 0.5460472106933594,
+      "learning_rate": 0.0001271944353016954,
+      "loss": 0.3628,
+      "step": 4712
+    },
+    {
+      "epoch": 1.5132445015251244,
+      "grad_norm": 0.9577854871749878,
+      "learning_rate": 0.00012716084258577388,
+      "loss": 0.5651,
+      "step": 4713
+    },
+    {
+      "epoch": 1.513565580349976,
+      "grad_norm": 0.5291711091995239,
+      "learning_rate": 0.0001271272465603861,
+      "loss": 0.3436,
+      "step": 4714
+    },
+    {
+      "epoch": 1.5138866591748275,
+      "grad_norm": 0.6409555673599243,
+      "learning_rate": 0.0001270936472296256,
+      "loss": 0.7887,
+      "step": 4715
+    },
+    {
+      "epoch": 1.514207737999679,
+      "grad_norm": 0.6114294528961182,
+      "learning_rate": 0.00012706004459758636,
+      "loss": 0.7388,
+      "step": 4716
+    },
+    {
+      "epoch": 1.5145288168245306,
+      "grad_norm": 0.6325629353523254,
+      "learning_rate": 0.00012702643866836278,
+      "loss": 0.416,
+      "step": 4717
+    },
+    {
+      "epoch": 1.5148498956493819,
+      "grad_norm": 0.9128963351249695,
+      "learning_rate": 0.00012699282944604967,
+      "loss": 0.4181,
+      "step": 4718
+    },
+    {
+      "epoch": 1.5151709744742334,
+      "grad_norm": 0.7952489256858826,
+      "learning_rate": 0.00012695921693474212,
+      "loss": 0.2609,
+      "step": 4719
+    },
+    {
+      "epoch": 1.5154920532990848,
+      "grad_norm": 0.715871274471283,
+      "learning_rate": 0.00012692560113853574,
+      "loss": 0.269,
+      "step": 4720
+    },
+    {
+      "epoch": 1.5158131321239363,
+      "grad_norm": 0.9184644222259521,
+      "learning_rate": 0.00012689198206152657,
+      "loss": 0.5186,
+      "step": 4721
+    },
+    {
+      "epoch": 1.5161342109487879,
+      "grad_norm": 1.1602627038955688,
+      "learning_rate": 0.00012685835970781097,
+      "loss": 0.736,
+      "step": 4722
+    },
+    {
+      "epoch": 1.5164552897736394,
+      "grad_norm": 1.0593267679214478,
+      "learning_rate": 0.0001268247340814857,
+      "loss": 0.6185,
+      "step": 4723
+    },
+    {
+      "epoch": 1.516776368598491,
+      "grad_norm": 0.8425109386444092,
+      "learning_rate": 0.00012679110518664795,
+      "loss": 0.555,
+      "step": 4724
+    },
+    {
+      "epoch": 1.5170974474233425,
+      "grad_norm": 1.03610360622406,
+      "learning_rate": 0.00012675747302739527,
+      "loss": 0.8006,
+      "step": 4725
+    },
+    {
+      "epoch": 1.517418526248194,
+      "grad_norm": 0.6738163828849792,
+      "learning_rate": 0.00012672383760782568,
+      "loss": 0.431,
+      "step": 4726
+    },
+    {
+      "epoch": 1.5177396050730454,
+      "grad_norm": 1.9449659585952759,
+      "learning_rate": 0.00012669019893203759,
+      "loss": 0.658,
+      "step": 4727
+    },
+    {
+      "epoch": 1.518060683897897,
+      "grad_norm": 0.9231836795806885,
+      "learning_rate": 0.00012665655700412967,
+      "loss": 0.6246,
+      "step": 4728
+    },
+    {
+      "epoch": 1.5183817627227483,
+      "grad_norm": 0.884365975856781,
+      "learning_rate": 0.00012662291182820114,
+      "loss": 0.6044,
+      "step": 4729
+    },
+    {
+      "epoch": 1.5187028415475998,
+      "grad_norm": 0.9164744019508362,
+      "learning_rate": 0.00012658926340835156,
+      "loss": 0.6805,
+      "step": 4730
+    },
+    {
+      "epoch": 1.5190239203724514,
+      "grad_norm": 1.071161150932312,
+      "learning_rate": 0.00012655561174868088,
+      "loss": 0.6354,
+      "step": 4731
+    },
+    {
+      "epoch": 1.519344999197303,
+      "grad_norm": 1.1153299808502197,
+      "learning_rate": 0.00012652195685328946,
+      "loss": 0.8903,
+      "step": 4732
+    },
+    {
+      "epoch": 1.5196660780221545,
+      "grad_norm": 1.0313321352005005,
+      "learning_rate": 0.00012648829872627807,
+      "loss": 0.5328,
+      "step": 4733
+    },
+    {
+      "epoch": 1.519987156847006,
+      "grad_norm": 1.025863528251648,
+      "learning_rate": 0.00012645463737174782,
+      "loss": 0.6995,
+      "step": 4734
+    },
+    {
+      "epoch": 1.5203082356718576,
+      "grad_norm": 0.8044242858886719,
+      "learning_rate": 0.00012642097279380027,
+      "loss": 0.6276,
+      "step": 4735
+    },
+    {
+      "epoch": 1.520629314496709,
+      "grad_norm": 1.842185139656067,
+      "learning_rate": 0.0001263873049965373,
+      "loss": 0.6605,
+      "step": 4736
+    },
+    {
+      "epoch": 1.5209503933215605,
+      "grad_norm": 1.1269233226776123,
+      "learning_rate": 0.00012635363398406127,
+      "loss": 0.6219,
+      "step": 4737
+    },
+    {
+      "epoch": 1.5212714721464118,
+      "grad_norm": 1.2168982028961182,
+      "learning_rate": 0.00012631995976047488,
+      "loss": 0.9384,
+      "step": 4738
+    },
+    {
+      "epoch": 1.5215925509712633,
+      "grad_norm": 0.9263548254966736,
+      "learning_rate": 0.0001262862823298812,
+      "loss": 0.6786,
+      "step": 4739
+    },
+    {
+      "epoch": 1.5219136297961149,
+      "grad_norm": 0.9486871361732483,
+      "learning_rate": 0.00012625260169638378,
+      "loss": 0.7042,
+      "step": 4740
+    },
+    {
+      "epoch": 1.5222347086209664,
+      "grad_norm": 1.2256191968917847,
+      "learning_rate": 0.00012621891786408648,
+      "loss": 0.8557,
+      "step": 4741
+    },
+    {
+      "epoch": 1.522555787445818,
+      "grad_norm": 0.9984161853790283,
+      "learning_rate": 0.00012618523083709357,
+      "loss": 0.6621,
+      "step": 4742
+    },
+    {
+      "epoch": 1.5228768662706695,
+      "grad_norm": 1.0978655815124512,
+      "learning_rate": 0.00012615154061950968,
+      "loss": 0.6981,
+      "step": 4743
+    },
+    {
+      "epoch": 1.523197945095521,
+      "grad_norm": 0.8551576137542725,
+      "learning_rate": 0.00012611784721543995,
+      "loss": 0.6271,
+      "step": 4744
+    },
+    {
+      "epoch": 1.5235190239203724,
+      "grad_norm": 0.907645583152771,
+      "learning_rate": 0.00012608415062898972,
+      "loss": 0.6353,
+      "step": 4745
+    },
+    {
+      "epoch": 1.523840102745224,
+      "grad_norm": 1.3650954961776733,
+      "learning_rate": 0.00012605045086426487,
+      "loss": 0.5611,
+      "step": 4746
+    },
+    {
+      "epoch": 1.5241611815700753,
+      "grad_norm": 0.9021798372268677,
+      "learning_rate": 0.00012601674792537157,
+      "loss": 0.6619,
+      "step": 4747
+    },
+    {
+      "epoch": 1.5244822603949268,
+      "grad_norm": 1.0378648042678833,
+      "learning_rate": 0.00012598304181641646,
+      "loss": 0.6465,
+      "step": 4748
+    },
+    {
+      "epoch": 1.5248033392197784,
+      "grad_norm": 1.0701501369476318,
+      "learning_rate": 0.00012594933254150655,
+      "loss": 0.6018,
+      "step": 4749
+    },
+    {
+      "epoch": 1.52512441804463,
+      "grad_norm": 0.9513067603111267,
+      "learning_rate": 0.00012591562010474914,
+      "loss": 0.5637,
+      "step": 4750
+    },
+    {
+      "epoch": 1.5254454968694815,
+      "grad_norm": 0.8903157114982605,
+      "learning_rate": 0.00012588190451025207,
+      "loss": 0.6278,
+      "step": 4751
+    },
+    {
+      "epoch": 1.525766575694333,
+      "grad_norm": 1.711920142173767,
+      "learning_rate": 0.00012584818576212345,
+      "loss": 0.6439,
+      "step": 4752
+    },
+    {
+      "epoch": 1.5260876545191846,
+      "grad_norm": 1.3381930589675903,
+      "learning_rate": 0.0001258144638644718,
+      "loss": 0.6182,
+      "step": 4753
+    },
+    {
+      "epoch": 1.526408733344036,
+      "grad_norm": 0.8519673347473145,
+      "learning_rate": 0.000125780738821406,
+      "loss": 0.5031,
+      "step": 4754
+    },
+    {
+      "epoch": 1.5267298121688875,
+      "grad_norm": 1.123969554901123,
+      "learning_rate": 0.0001257470106370354,
+      "loss": 0.702,
+      "step": 4755
+    },
+    {
+      "epoch": 1.5270508909937388,
+      "grad_norm": 0.7065092921257019,
+      "learning_rate": 0.00012571327931546963,
+      "loss": 0.438,
+      "step": 4756
+    },
+    {
+      "epoch": 1.5273719698185904,
+      "grad_norm": 1.5530750751495361,
+      "learning_rate": 0.00012567954486081878,
+      "loss": 0.5497,
+      "step": 4757
+    },
+    {
+      "epoch": 1.527693048643442,
+      "grad_norm": 1.869837760925293,
+      "learning_rate": 0.0001256458072771933,
+      "loss": 0.5909,
+      "step": 4758
+    },
+    {
+      "epoch": 1.5280141274682935,
+      "grad_norm": 0.8413248658180237,
+      "learning_rate": 0.00012561206656870396,
+      "loss": 0.5934,
+      "step": 4759
+    },
+    {
+      "epoch": 1.528335206293145,
+      "grad_norm": 0.7256144285202026,
+      "learning_rate": 0.00012557832273946205,
+      "loss": 0.4381,
+      "step": 4760
+    },
+    {
+      "epoch": 1.5286562851179966,
+      "grad_norm": 1.1708571910858154,
+      "learning_rate": 0.00012554457579357905,
+      "loss": 0.5908,
+      "step": 4761
+    },
+    {
+      "epoch": 1.5289773639428481,
+      "grad_norm": 0.7625494003295898,
+      "learning_rate": 0.00012551082573516705,
+      "loss": 0.4367,
+      "step": 4762
+    },
+    {
+      "epoch": 1.5292984427676994,
+      "grad_norm": 0.81993567943573,
+      "learning_rate": 0.00012547707256833823,
+      "loss": 0.5166,
+      "step": 4763
+    },
+    {
+      "epoch": 1.529619521592551,
+      "grad_norm": 0.6082926392555237,
+      "learning_rate": 0.00012544331629720543,
+      "loss": 0.4116,
+      "step": 4764
+    },
+    {
+      "epoch": 1.5299406004174023,
+      "grad_norm": 0.818605899810791,
+      "learning_rate": 0.00012540955692588173,
+      "loss": 0.6703,
+      "step": 4765
+    },
+    {
+      "epoch": 1.5302616792422539,
+      "grad_norm": 0.6201203465461731,
+      "learning_rate": 0.00012537579445848058,
+      "loss": 1.0005,
+      "step": 4766
+    },
+    {
+      "epoch": 1.5305827580671054,
+      "grad_norm": 0.7237659692764282,
+      "learning_rate": 0.00012534202889911584,
+      "loss": 0.6653,
+      "step": 4767
+    },
+    {
+      "epoch": 1.530903836891957,
+      "grad_norm": 0.7227284908294678,
+      "learning_rate": 0.00012530826025190174,
+      "loss": 0.5275,
+      "step": 4768
+    },
+    {
+      "epoch": 1.5312249157168085,
+      "grad_norm": 0.7782474160194397,
+      "learning_rate": 0.00012527448852095295,
+      "loss": 0.4055,
+      "step": 4769
+    },
+    {
+      "epoch": 1.53154599454166,
+      "grad_norm": 0.8507752418518066,
+      "learning_rate": 0.00012524071371038434,
+      "loss": 0.2621,
+      "step": 4770
+    },
+    {
+      "epoch": 1.5318670733665116,
+      "grad_norm": 0.8687464594841003,
+      "learning_rate": 0.0001252069358243114,
+      "loss": 0.3743,
+      "step": 4771
+    },
+    {
+      "epoch": 1.532188152191363,
+      "grad_norm": 0.7505998015403748,
+      "learning_rate": 0.00012517315486684972,
+      "loss": 0.4383,
+      "step": 4772
+    },
+    {
+      "epoch": 1.5325092310162145,
+      "grad_norm": 0.9363805055618286,
+      "learning_rate": 0.0001251393708421155,
+      "loss": 0.7721,
+      "step": 4773
+    },
+    {
+      "epoch": 1.5328303098410658,
+      "grad_norm": 0.9391258955001831,
+      "learning_rate": 0.00012510558375422522,
+      "loss": 0.7531,
+      "step": 4774
+    },
+    {
+      "epoch": 1.5331513886659174,
+      "grad_norm": 0.9769285917282104,
+      "learning_rate": 0.0001250717936072957,
+      "loss": 0.8322,
+      "step": 4775
+    },
+    {
+      "epoch": 1.533472467490769,
+      "grad_norm": 0.670604944229126,
+      "learning_rate": 0.00012503800040544416,
+      "loss": 0.4766,
+      "step": 4776
+    },
+    {
+      "epoch": 1.5337935463156205,
+      "grad_norm": 0.7332351803779602,
+      "learning_rate": 0.00012500420415278822,
+      "loss": 0.6418,
+      "step": 4777
+    },
+    {
+      "epoch": 1.534114625140472,
+      "grad_norm": 0.7202760577201843,
+      "learning_rate": 0.00012497040485344584,
+      "loss": 0.5964,
+      "step": 4778
+    },
+    {
+      "epoch": 1.5344357039653236,
+      "grad_norm": 0.9445421695709229,
+      "learning_rate": 0.0001249366025115354,
+      "loss": 0.6184,
+      "step": 4779
+    },
+    {
+      "epoch": 1.5347567827901751,
+      "grad_norm": 0.7218775749206543,
+      "learning_rate": 0.0001249027971311756,
+      "loss": 0.5217,
+      "step": 4780
+    },
+    {
+      "epoch": 1.5350778616150265,
+      "grad_norm": 0.7306594252586365,
+      "learning_rate": 0.0001248689887164855,
+      "loss": 0.5849,
+      "step": 4781
+    },
+    {
+      "epoch": 1.535398940439878,
+      "grad_norm": 0.942954421043396,
+      "learning_rate": 0.00012483517727158454,
+      "loss": 0.6694,
+      "step": 4782
+    },
+    {
+      "epoch": 1.5357200192647293,
+      "grad_norm": 0.7879648208618164,
+      "learning_rate": 0.00012480136280059256,
+      "loss": 0.5698,
+      "step": 4783
+    },
+    {
+      "epoch": 1.536041098089581,
+      "grad_norm": 0.835502028465271,
+      "learning_rate": 0.00012476754530762977,
+      "loss": 0.63,
+      "step": 4784
+    },
+    {
+      "epoch": 1.5363621769144324,
+      "grad_norm": 0.9831227660179138,
+      "learning_rate": 0.00012473372479681672,
+      "loss": 0.7113,
+      "step": 4785
+    },
+    {
+      "epoch": 1.536683255739284,
+      "grad_norm": 0.9741390347480774,
+      "learning_rate": 0.00012469990127227432,
+      "loss": 0.6879,
+      "step": 4786
+    },
+    {
+      "epoch": 1.5370043345641355,
+      "grad_norm": 0.7839673757553101,
+      "learning_rate": 0.00012466607473812387,
+      "loss": 0.5085,
+      "step": 4787
+    },
+    {
+      "epoch": 1.537325413388987,
+      "grad_norm": 0.9261371493339539,
+      "learning_rate": 0.000124632245198487,
+      "loss": 0.6465,
+      "step": 4788
+    },
+    {
+      "epoch": 1.5376464922138386,
+      "grad_norm": 1.3706035614013672,
+      "learning_rate": 0.0001245984126574858,
+      "loss": 0.8686,
+      "step": 4789
+    },
+    {
+      "epoch": 1.53796757103869,
+      "grad_norm": 0.6611718535423279,
+      "learning_rate": 0.00012456457711924266,
+      "loss": 0.3897,
+      "step": 4790
+    },
+    {
+      "epoch": 1.5382886498635415,
+      "grad_norm": 0.9132871031761169,
+      "learning_rate": 0.00012453073858788026,
+      "loss": 0.6996,
+      "step": 4791
+    },
+    {
+      "epoch": 1.5386097286883929,
+      "grad_norm": 0.9300896525382996,
+      "learning_rate": 0.00012449689706752178,
+      "loss": 0.7055,
+      "step": 4792
+    },
+    {
+      "epoch": 1.5389308075132444,
+      "grad_norm": 0.7551592588424683,
+      "learning_rate": 0.00012446305256229073,
+      "loss": 0.6199,
+      "step": 4793
+    },
+    {
+      "epoch": 1.539251886338096,
+      "grad_norm": 0.9444193243980408,
+      "learning_rate": 0.00012442920507631092,
+      "loss": 0.6397,
+      "step": 4794
+    },
+    {
+      "epoch": 1.5395729651629475,
+      "grad_norm": 0.8001585006713867,
+      "learning_rate": 0.00012439535461370658,
+      "loss": 0.4635,
+      "step": 4795
+    },
+    {
+      "epoch": 1.539894043987799,
+      "grad_norm": 0.9087579250335693,
+      "learning_rate": 0.00012436150117860225,
+      "loss": 0.5755,
+      "step": 4796
+    },
+    {
+      "epoch": 1.5402151228126506,
+      "grad_norm": 0.9415332674980164,
+      "learning_rate": 0.00012432764477512292,
+      "loss": 0.6303,
+      "step": 4797
+    },
+    {
+      "epoch": 1.5405362016375022,
+      "grad_norm": 0.8173394203186035,
+      "learning_rate": 0.00012429378540739386,
+      "loss": 0.5832,
+      "step": 4798
+    },
+    {
+      "epoch": 1.5408572804623535,
+      "grad_norm": 1.0621390342712402,
+      "learning_rate": 0.00012425992307954075,
+      "loss": 0.6614,
+      "step": 4799
+    },
+    {
+      "epoch": 1.541178359287205,
+      "grad_norm": 1.3637477159500122,
+      "learning_rate": 0.0001242260577956896,
+      "loss": 0.6073,
+      "step": 4800
+    },
+    {
+      "epoch": 1.5414994381120564,
+      "grad_norm": 1.0318750143051147,
+      "learning_rate": 0.00012419218955996676,
+      "loss": 0.6623,
+      "step": 4801
+    },
+    {
+      "epoch": 1.541820516936908,
+      "grad_norm": 1.1521292924880981,
+      "learning_rate": 0.00012415831837649905,
+      "loss": 0.5285,
+      "step": 4802
+    },
+    {
+      "epoch": 1.5421415957617595,
+      "grad_norm": 1.148503065109253,
+      "learning_rate": 0.0001241244442494135,
+      "loss": 0.5444,
+      "step": 4803
+    },
+    {
+      "epoch": 1.542462674586611,
+      "grad_norm": 0.9957001805305481,
+      "learning_rate": 0.0001240905671828376,
+      "loss": 0.5206,
+      "step": 4804
+    },
+    {
+      "epoch": 1.5427837534114626,
+      "grad_norm": 0.6883019804954529,
+      "learning_rate": 0.00012405668718089917,
+      "loss": 0.4488,
+      "step": 4805
+    },
+    {
+      "epoch": 1.5431048322363141,
+      "grad_norm": 1.0081177949905396,
+      "learning_rate": 0.00012402280424772639,
+      "loss": 0.6462,
+      "step": 4806
+    },
+    {
+      "epoch": 1.5434259110611657,
+      "grad_norm": 0.8516654372215271,
+      "learning_rate": 0.00012398891838744778,
+      "loss": 0.5383,
+      "step": 4807
+    },
+    {
+      "epoch": 1.543746989886017,
+      "grad_norm": 0.8348969221115112,
+      "learning_rate": 0.0001239550296041922,
+      "loss": 0.484,
+      "step": 4808
+    },
+    {
+      "epoch": 1.5440680687108685,
+      "grad_norm": 1.3261665105819702,
+      "learning_rate": 0.00012392113790208895,
+      "loss": 0.5407,
+      "step": 4809
+    },
+    {
+      "epoch": 1.5443891475357199,
+      "grad_norm": 1.114880919456482,
+      "learning_rate": 0.0001238872432852676,
+      "loss": 0.7105,
+      "step": 4810
+    },
+    {
+      "epoch": 1.5447102263605714,
+      "grad_norm": 0.8749669790267944,
+      "learning_rate": 0.0001238533457578581,
+      "loss": 0.4329,
+      "step": 4811
+    },
+    {
+      "epoch": 1.545031305185423,
+      "grad_norm": 0.7413171529769897,
+      "learning_rate": 0.0001238194453239908,
+      "loss": 0.3563,
+      "step": 4812
+    },
+    {
+      "epoch": 1.5453523840102745,
+      "grad_norm": 0.7285036444664001,
+      "learning_rate": 0.0001237855419877963,
+      "loss": 0.4244,
+      "step": 4813
+    },
+    {
+      "epoch": 1.545673462835126,
+      "grad_norm": 0.6822255253791809,
+      "learning_rate": 0.0001237516357534057,
+      "loss": 0.3974,
+      "step": 4814
+    },
+    {
+      "epoch": 1.5459945416599776,
+      "grad_norm": 0.44284263253211975,
+      "learning_rate": 0.0001237177266249503,
+      "loss": 0.6239,
+      "step": 4815
+    },
+    {
+      "epoch": 1.5463156204848292,
+      "grad_norm": 0.6735876798629761,
+      "learning_rate": 0.0001236838146065619,
+      "loss": 1.0601,
+      "step": 4816
+    },
+    {
+      "epoch": 1.5466366993096805,
+      "grad_norm": 0.6792389154434204,
+      "learning_rate": 0.00012364989970237248,
+      "loss": 0.6631,
+      "step": 4817
+    },
+    {
+      "epoch": 1.546957778134532,
+      "grad_norm": 0.8029338717460632,
+      "learning_rate": 0.00012361598191651453,
+      "loss": 0.446,
+      "step": 4818
+    },
+    {
+      "epoch": 1.5472788569593834,
+      "grad_norm": 0.5691446661949158,
+      "learning_rate": 0.00012358206125312083,
+      "loss": 0.2494,
+      "step": 4819
+    },
+    {
+      "epoch": 1.547599935784235,
+      "grad_norm": 0.7446752786636353,
+      "learning_rate": 0.00012354813771632447,
+      "loss": 0.3416,
+      "step": 4820
+    },
+    {
+      "epoch": 1.5479210146090865,
+      "grad_norm": 0.5998992323875427,
+      "learning_rate": 0.000123514211310259,
+      "loss": 0.2877,
+      "step": 4821
+    },
+    {
+      "epoch": 1.548242093433938,
+      "grad_norm": 0.9401177763938904,
+      "learning_rate": 0.00012348028203905818,
+      "loss": 0.7846,
+      "step": 4822
+    },
+    {
+      "epoch": 1.5485631722587896,
+      "grad_norm": 0.9916895627975464,
+      "learning_rate": 0.00012344634990685624,
+      "loss": 0.8004,
+      "step": 4823
+    },
+    {
+      "epoch": 1.5488842510836411,
+      "grad_norm": 0.8018080592155457,
+      "learning_rate": 0.0001234124149177877,
+      "loss": 0.6483,
+      "step": 4824
+    },
+    {
+      "epoch": 1.5492053299084927,
+      "grad_norm": 0.9374871850013733,
+      "learning_rate": 0.00012337847707598738,
+      "loss": 0.7748,
+      "step": 4825
+    },
+    {
+      "epoch": 1.549526408733344,
+      "grad_norm": 0.975395917892456,
+      "learning_rate": 0.00012334453638559057,
+      "loss": 0.6303,
+      "step": 4826
+    },
+    {
+      "epoch": 1.5498474875581956,
+      "grad_norm": 1.0059309005737305,
+      "learning_rate": 0.0001233105928507328,
+      "loss": 0.4923,
+      "step": 4827
+    },
+    {
+      "epoch": 1.550168566383047,
+      "grad_norm": 0.9571434259414673,
+      "learning_rate": 0.00012327664647554998,
+      "loss": 0.7088,
+      "step": 4828
+    },
+    {
+      "epoch": 1.5504896452078984,
+      "grad_norm": 0.9377606511116028,
+      "learning_rate": 0.0001232426972641784,
+      "loss": 0.7066,
+      "step": 4829
+    },
+    {
+      "epoch": 1.55081072403275,
+      "grad_norm": 0.748374342918396,
+      "learning_rate": 0.00012320874522075468,
+      "loss": 0.4916,
+      "step": 4830
+    },
+    {
+      "epoch": 1.5511318028576015,
+      "grad_norm": 0.8802591562271118,
+      "learning_rate": 0.00012317479034941573,
+      "loss": 0.5684,
+      "step": 4831
+    },
+    {
+      "epoch": 1.551452881682453,
+      "grad_norm": 0.7750247120857239,
+      "learning_rate": 0.0001231408326542989,
+      "loss": 0.5093,
+      "step": 4832
+    },
+    {
+      "epoch": 1.5517739605073047,
+      "grad_norm": 0.7303853034973145,
+      "learning_rate": 0.00012310687213954182,
+      "loss": 0.582,
+      "step": 4833
+    },
+    {
+      "epoch": 1.5520950393321562,
+      "grad_norm": 0.9113678932189941,
+      "learning_rate": 0.0001230729088092824,
+      "loss": 0.5847,
+      "step": 4834
+    },
+    {
+      "epoch": 1.5524161181570075,
+      "grad_norm": 0.9570289254188538,
+      "learning_rate": 0.00012303894266765908,
+      "loss": 0.622,
+      "step": 4835
+    },
+    {
+      "epoch": 1.552737196981859,
+      "grad_norm": 0.9025993347167969,
+      "learning_rate": 0.00012300497371881046,
+      "loss": 0.5098,
+      "step": 4836
+    },
+    {
+      "epoch": 1.5530582758067104,
+      "grad_norm": 1.2178800106048584,
+      "learning_rate": 0.00012297100196687557,
+      "loss": 0.6402,
+      "step": 4837
+    },
+    {
+      "epoch": 1.553379354631562,
+      "grad_norm": 0.9967452883720398,
+      "learning_rate": 0.00012293702741599378,
+      "loss": 0.6723,
+      "step": 4838
+    },
+    {
+      "epoch": 1.5537004334564135,
+      "grad_norm": 0.7148991823196411,
+      "learning_rate": 0.00012290305007030478,
+      "loss": 0.5875,
+      "step": 4839
+    },
+    {
+      "epoch": 1.554021512281265,
+      "grad_norm": 1.1900482177734375,
+      "learning_rate": 0.00012286906993394856,
+      "loss": 0.6827,
+      "step": 4840
+    },
+    {
+      "epoch": 1.5543425911061166,
+      "grad_norm": 1.2599159479141235,
+      "learning_rate": 0.00012283508701106557,
+      "loss": 0.8436,
+      "step": 4841
+    },
+    {
+      "epoch": 1.5546636699309682,
+      "grad_norm": 1.0731333494186401,
+      "learning_rate": 0.0001228011013057965,
+      "loss": 0.7042,
+      "step": 4842
+    },
+    {
+      "epoch": 1.5549847487558197,
+      "grad_norm": 0.7968559861183167,
+      "learning_rate": 0.0001227671128222824,
+      "loss": 0.5391,
+      "step": 4843
+    },
+    {
+      "epoch": 1.555305827580671,
+      "grad_norm": 1.2837064266204834,
+      "learning_rate": 0.00012273312156466464,
+      "loss": 0.8585,
+      "step": 4844
+    },
+    {
+      "epoch": 1.5556269064055226,
+      "grad_norm": 0.7288888692855835,
+      "learning_rate": 0.00012269912753708502,
+      "loss": 0.5193,
+      "step": 4845
+    },
+    {
+      "epoch": 1.555947985230374,
+      "grad_norm": 0.8178963661193848,
+      "learning_rate": 0.0001226651307436855,
+      "loss": 0.5001,
+      "step": 4846
+    },
+    {
+      "epoch": 1.5562690640552255,
+      "grad_norm": 1.0722037553787231,
+      "learning_rate": 0.0001226311311886086,
+      "loss": 0.5924,
+      "step": 4847
+    },
+    {
+      "epoch": 1.556590142880077,
+      "grad_norm": 0.9165698885917664,
+      "learning_rate": 0.00012259712887599698,
+      "loss": 0.5656,
+      "step": 4848
+    },
+    {
+      "epoch": 1.5569112217049286,
+      "grad_norm": 1.0873520374298096,
+      "learning_rate": 0.00012256312380999376,
+      "loss": 0.7788,
+      "step": 4849
+    },
+    {
+      "epoch": 1.5572323005297801,
+      "grad_norm": 1.1277052164077759,
+      "learning_rate": 0.00012252911599474237,
+      "loss": 0.6183,
+      "step": 4850
+    },
+    {
+      "epoch": 1.5575533793546317,
+      "grad_norm": 0.8544272780418396,
+      "learning_rate": 0.0001224951054343865,
+      "loss": 0.5859,
+      "step": 4851
+    },
+    {
+      "epoch": 1.5578744581794832,
+      "grad_norm": 1.0256620645523071,
+      "learning_rate": 0.0001224610921330703,
+      "loss": 0.663,
+      "step": 4852
+    },
+    {
+      "epoch": 1.5581955370043346,
+      "grad_norm": 0.9969653487205505,
+      "learning_rate": 0.00012242707609493814,
+      "loss": 0.6084,
+      "step": 4853
+    },
+    {
+      "epoch": 1.558516615829186,
+      "grad_norm": 0.8547719717025757,
+      "learning_rate": 0.00012239305732413477,
+      "loss": 0.5886,
+      "step": 4854
+    },
+    {
+      "epoch": 1.5588376946540374,
+      "grad_norm": 1.0089030265808105,
+      "learning_rate": 0.0001223590358248053,
+      "loss": 0.6111,
+      "step": 4855
+    },
+    {
+      "epoch": 1.559158773478889,
+      "grad_norm": 1.0773385763168335,
+      "learning_rate": 0.00012232501160109514,
+      "loss": 0.5771,
+      "step": 4856
+    },
+    {
+      "epoch": 1.5594798523037405,
+      "grad_norm": 0.8894745707511902,
+      "learning_rate": 0.00012229098465715006,
+      "loss": 0.5668,
+      "step": 4857
+    },
+    {
+      "epoch": 1.559800931128592,
+      "grad_norm": 0.9482760429382324,
+      "learning_rate": 0.00012225695499711606,
+      "loss": 0.4387,
+      "step": 4858
+    },
+    {
+      "epoch": 1.5601220099534436,
+      "grad_norm": 1.192726969718933,
+      "learning_rate": 0.00012222292262513965,
+      "loss": 0.5511,
+      "step": 4859
+    },
+    {
+      "epoch": 1.5604430887782952,
+      "grad_norm": 0.9910078644752502,
+      "learning_rate": 0.0001221888875453675,
+      "loss": 0.6603,
+      "step": 4860
+    },
+    {
+      "epoch": 1.5607641676031467,
+      "grad_norm": 1.0606755018234253,
+      "learning_rate": 0.00012215484976194676,
+      "loss": 0.534,
+      "step": 4861
+    },
+    {
+      "epoch": 1.561085246427998,
+      "grad_norm": 1.2194321155548096,
+      "learning_rate": 0.00012212080927902474,
+      "loss": 0.4939,
+      "step": 4862
+    },
+    {
+      "epoch": 1.5614063252528496,
+      "grad_norm": 0.4965793490409851,
+      "learning_rate": 0.00012208676610074917,
+      "loss": 0.3877,
+      "step": 4863
+    },
+    {
+      "epoch": 1.561727404077701,
+      "grad_norm": 0.7706876993179321,
+      "learning_rate": 0.00012205272023126821,
+      "loss": 0.4616,
+      "step": 4864
+    },
+    {
+      "epoch": 1.5620484829025525,
+      "grad_norm": 0.7944642305374146,
+      "learning_rate": 0.00012201867167473015,
+      "loss": 0.9236,
+      "step": 4865
+    },
+    {
+      "epoch": 1.562369561727404,
+      "grad_norm": 0.7954860329627991,
+      "learning_rate": 0.00012198462043528376,
+      "loss": 0.9766,
+      "step": 4866
+    },
+    {
+      "epoch": 1.5626906405522556,
+      "grad_norm": 0.6089110970497131,
+      "learning_rate": 0.00012195056651707806,
+      "loss": 0.7265,
+      "step": 4867
+    },
+    {
+      "epoch": 1.5630117193771071,
+      "grad_norm": 0.6708793640136719,
+      "learning_rate": 0.00012191650992426238,
+      "loss": 0.3664,
+      "step": 4868
+    },
+    {
+      "epoch": 1.5633327982019587,
+      "grad_norm": 1.029671549797058,
+      "learning_rate": 0.00012188245066098647,
+      "loss": 0.4111,
+      "step": 4869
+    },
+    {
+      "epoch": 1.5636538770268102,
+      "grad_norm": 0.8126581311225891,
+      "learning_rate": 0.00012184838873140032,
+      "loss": 0.3168,
+      "step": 4870
+    },
+    {
+      "epoch": 1.5639749558516616,
+      "grad_norm": 0.7029819488525391,
+      "learning_rate": 0.00012181432413965428,
+      "loss": 0.4029,
+      "step": 4871
+    },
+    {
+      "epoch": 1.5642960346765131,
+      "grad_norm": 0.7372133135795593,
+      "learning_rate": 0.000121780256889899,
+      "loss": 0.7746,
+      "step": 4872
+    },
+    {
+      "epoch": 1.5646171135013645,
+      "grad_norm": 0.8719139099121094,
+      "learning_rate": 0.00012174618698628549,
+      "loss": 0.6874,
+      "step": 4873
+    },
+    {
+      "epoch": 1.564938192326216,
+      "grad_norm": 0.8879600167274475,
+      "learning_rate": 0.00012171211443296505,
+      "loss": 0.6931,
+      "step": 4874
+    },
+    {
+      "epoch": 1.5652592711510676,
+      "grad_norm": 0.8598337769508362,
+      "learning_rate": 0.00012167803923408934,
+      "loss": 0.5866,
+      "step": 4875
+    },
+    {
+      "epoch": 1.565580349975919,
+      "grad_norm": 0.6961860656738281,
+      "learning_rate": 0.00012164396139381029,
+      "loss": 0.4646,
+      "step": 4876
+    },
+    {
+      "epoch": 1.5659014288007707,
+      "grad_norm": 0.7586543560028076,
+      "learning_rate": 0.00012160988091628022,
+      "loss": 0.5582,
+      "step": 4877
+    },
+    {
+      "epoch": 1.5662225076256222,
+      "grad_norm": 0.82035231590271,
+      "learning_rate": 0.00012157579780565169,
+      "loss": 0.5765,
+      "step": 4878
+    },
+    {
+      "epoch": 1.5665435864504738,
+      "grad_norm": 0.6588397026062012,
+      "learning_rate": 0.00012154171206607764,
+      "loss": 0.4265,
+      "step": 4879
+    },
+    {
+      "epoch": 1.566864665275325,
+      "grad_norm": 0.7135642766952515,
+      "learning_rate": 0.00012150762370171136,
+      "loss": 0.4743,
+      "step": 4880
+    },
+    {
+      "epoch": 1.5671857441001766,
+      "grad_norm": 0.7233707308769226,
+      "learning_rate": 0.00012147353271670634,
+      "loss": 0.474,
+      "step": 4881
+    },
+    {
+      "epoch": 1.567506822925028,
+      "grad_norm": 0.773969829082489,
+      "learning_rate": 0.00012143943911521647,
+      "loss": 0.5916,
+      "step": 4882
+    },
+    {
+      "epoch": 1.5678279017498795,
+      "grad_norm": 0.9881671667098999,
+      "learning_rate": 0.000121405342901396,
+      "loss": 0.6945,
+      "step": 4883
+    },
+    {
+      "epoch": 1.568148980574731,
+      "grad_norm": 1.0888925790786743,
+      "learning_rate": 0.00012137124407939943,
+      "loss": 0.5791,
+      "step": 4884
+    },
+    {
+      "epoch": 1.5684700593995826,
+      "grad_norm": 1.0212979316711426,
+      "learning_rate": 0.00012133714265338161,
+      "loss": 0.7161,
+      "step": 4885
+    },
+    {
+      "epoch": 1.5687911382244342,
+      "grad_norm": 0.8693848848342896,
+      "learning_rate": 0.00012130303862749767,
+      "loss": 0.6095,
+      "step": 4886
+    },
+    {
+      "epoch": 1.5691122170492857,
+      "grad_norm": 1.1114261150360107,
+      "learning_rate": 0.00012126893200590308,
+      "loss": 0.6328,
+      "step": 4887
+    },
+    {
+      "epoch": 1.569433295874137,
+      "grad_norm": 0.9835912585258484,
+      "learning_rate": 0.00012123482279275365,
+      "loss": 0.6541,
+      "step": 4888
+    },
+    {
+      "epoch": 1.5697543746989886,
+      "grad_norm": 0.8264690041542053,
+      "learning_rate": 0.00012120071099220549,
+      "loss": 0.4623,
+      "step": 4889
+    },
+    {
+      "epoch": 1.5700754535238401,
+      "grad_norm": 0.8061067461967468,
+      "learning_rate": 0.00012116659660841499,
+      "loss": 0.524,
+      "step": 4890
+    },
+    {
+      "epoch": 1.5703965323486915,
+      "grad_norm": 0.846877932548523,
+      "learning_rate": 0.00012113247964553888,
+      "loss": 0.6132,
+      "step": 4891
+    },
+    {
+      "epoch": 1.570717611173543,
+      "grad_norm": 1.043820858001709,
+      "learning_rate": 0.00012109836010773424,
+      "loss": 0.6175,
+      "step": 4892
+    },
+    {
+      "epoch": 1.5710386899983946,
+      "grad_norm": 1.267200231552124,
+      "learning_rate": 0.0001210642379991584,
+      "loss": 0.6943,
+      "step": 4893
+    },
+    {
+      "epoch": 1.5713597688232461,
+      "grad_norm": 1.0244359970092773,
+      "learning_rate": 0.00012103011332396908,
+      "loss": 0.6472,
+      "step": 4894
+    },
+    {
+      "epoch": 1.5716808476480977,
+      "grad_norm": 1.727343201637268,
+      "learning_rate": 0.00012099598608632428,
+      "loss": 0.5938,
+      "step": 4895
+    },
+    {
+      "epoch": 1.5720019264729492,
+      "grad_norm": 0.8495163917541504,
+      "learning_rate": 0.0001209618562903822,
+      "loss": 0.5524,
+      "step": 4896
+    },
+    {
+      "epoch": 1.5723230052978006,
+      "grad_norm": 1.0026789903640747,
+      "learning_rate": 0.00012092772394030152,
+      "loss": 0.5923,
+      "step": 4897
+    },
+    {
+      "epoch": 1.572644084122652,
+      "grad_norm": 1.362273097038269,
+      "learning_rate": 0.00012089358904024117,
+      "loss": 0.7854,
+      "step": 4898
+    },
+    {
+      "epoch": 1.5729651629475037,
+      "grad_norm": 0.79168701171875,
+      "learning_rate": 0.00012085945159436038,
+      "loss": 0.5297,
+      "step": 4899
+    },
+    {
+      "epoch": 1.573286241772355,
+      "grad_norm": 1.0517579317092896,
+      "learning_rate": 0.00012082531160681869,
+      "loss": 0.5276,
+      "step": 4900
+    },
+    {
+      "epoch": 1.5736073205972065,
+      "grad_norm": 1.1384689807891846,
+      "learning_rate": 0.00012079116908177593,
+      "loss": 0.4365,
+      "step": 4901
+    },
+    {
+      "epoch": 1.573928399422058,
+      "grad_norm": 0.916374683380127,
+      "learning_rate": 0.00012075702402339231,
+      "loss": 0.7095,
+      "step": 4902
+    },
+    {
+      "epoch": 1.5742494782469096,
+      "grad_norm": 0.8095363974571228,
+      "learning_rate": 0.00012072287643582825,
+      "loss": 0.5287,
+      "step": 4903
+    },
+    {
+      "epoch": 1.5745705570717612,
+      "grad_norm": 0.7615143060684204,
+      "learning_rate": 0.0001206887263232446,
+      "loss": 0.5034,
+      "step": 4904
+    },
+    {
+      "epoch": 1.5748916358966127,
+      "grad_norm": 0.9699037671089172,
+      "learning_rate": 0.00012065457368980236,
+      "loss": 0.4912,
+      "step": 4905
+    },
+    {
+      "epoch": 1.575212714721464,
+      "grad_norm": 1.3205819129943848,
+      "learning_rate": 0.00012062041853966298,
+      "loss": 0.7739,
+      "step": 4906
+    },
+    {
+      "epoch": 1.5755337935463156,
+      "grad_norm": 1.0966612100601196,
+      "learning_rate": 0.00012058626087698814,
+      "loss": 0.5963,
+      "step": 4907
+    },
+    {
+      "epoch": 1.5758548723711672,
+      "grad_norm": 1.8393914699554443,
+      "learning_rate": 0.00012055210070593988,
+      "loss": 0.5436,
+      "step": 4908
+    },
+    {
+      "epoch": 1.5761759511960185,
+      "grad_norm": 1.010202407836914,
+      "learning_rate": 0.00012051793803068046,
+      "loss": 0.5538,
+      "step": 4909
+    },
+    {
+      "epoch": 1.57649703002087,
+      "grad_norm": 0.9576059579849243,
+      "learning_rate": 0.00012048377285537256,
+      "loss": 0.4986,
+      "step": 4910
+    },
+    {
+      "epoch": 1.5768181088457216,
+      "grad_norm": 1.2450673580169678,
+      "learning_rate": 0.00012044960518417903,
+      "loss": 0.6183,
+      "step": 4911
+    },
+    {
+      "epoch": 1.5771391876705732,
+      "grad_norm": 0.9183635115623474,
+      "learning_rate": 0.00012041543502126318,
+      "loss": 0.5388,
+      "step": 4912
+    },
+    {
+      "epoch": 1.5774602664954247,
+      "grad_norm": 0.63742595911026,
+      "learning_rate": 0.0001203812623707885,
+      "loss": 0.3583,
+      "step": 4913
+    },
+    {
+      "epoch": 1.5777813453202763,
+      "grad_norm": 0.7811853885650635,
+      "learning_rate": 0.00012034708723691881,
+      "loss": 0.4776,
+      "step": 4914
+    },
+    {
+      "epoch": 1.5781024241451276,
+      "grad_norm": 0.6186936497688293,
+      "learning_rate": 0.00012031290962381823,
+      "loss": 0.5344,
+      "step": 4915
+    },
+    {
+      "epoch": 1.5784235029699791,
+      "grad_norm": 0.5814709663391113,
+      "learning_rate": 0.00012027872953565125,
+      "loss": 0.766,
+      "step": 4916
+    },
+    {
+      "epoch": 1.5787445817948307,
+      "grad_norm": 0.767975926399231,
+      "learning_rate": 0.00012024454697658261,
+      "loss": 0.4243,
+      "step": 4917
+    },
+    {
+      "epoch": 1.579065660619682,
+      "grad_norm": 0.8333182334899902,
+      "learning_rate": 0.00012021036195077731,
+      "loss": 0.4496,
+      "step": 4918
+    },
+    {
+      "epoch": 1.5793867394445336,
+      "grad_norm": 0.9080848693847656,
+      "learning_rate": 0.0001201761744624007,
+      "loss": 0.2366,
+      "step": 4919
+    },
+    {
+      "epoch": 1.5797078182693851,
+      "grad_norm": 0.7227024435997009,
+      "learning_rate": 0.00012014198451561842,
+      "loss": 0.2954,
+      "step": 4920
+    },
+    {
+      "epoch": 1.5800288970942367,
+      "grad_norm": 0.7397527694702148,
+      "learning_rate": 0.00012010779211459648,
+      "loss": 0.2761,
+      "step": 4921
+    },
+    {
+      "epoch": 1.5803499759190882,
+      "grad_norm": 0.753508448600769,
+      "learning_rate": 0.00012007359726350105,
+      "loss": 0.4536,
+      "step": 4922
+    },
+    {
+      "epoch": 1.5806710547439398,
+      "grad_norm": 1.0282949209213257,
+      "learning_rate": 0.00012003939996649865,
+      "loss": 0.7643,
+      "step": 4923
+    },
+    {
+      "epoch": 1.580992133568791,
+      "grad_norm": 0.9250288605690002,
+      "learning_rate": 0.00012000520022775617,
+      "loss": 0.7326,
+      "step": 4924
+    },
+    {
+      "epoch": 1.5813132123936426,
+      "grad_norm": 0.898415207862854,
+      "learning_rate": 0.00011997099805144069,
+      "loss": 0.6032,
+      "step": 4925
+    },
+    {
+      "epoch": 1.5816342912184942,
+      "grad_norm": 0.9478784799575806,
+      "learning_rate": 0.00011993679344171973,
+      "loss": 0.6384,
+      "step": 4926
+    },
+    {
+      "epoch": 1.5819553700433455,
+      "grad_norm": 1.0748721361160278,
+      "learning_rate": 0.00011990258640276094,
+      "loss": 0.3729,
+      "step": 4927
+    },
+    {
+      "epoch": 1.582276448868197,
+      "grad_norm": 0.7065726518630981,
+      "learning_rate": 0.00011986837693873237,
+      "loss": 0.5082,
+      "step": 4928
+    },
+    {
+      "epoch": 1.5825975276930486,
+      "grad_norm": 0.789629340171814,
+      "learning_rate": 0.00011983416505380234,
+      "loss": 0.5867,
+      "step": 4929
+    },
+    {
+      "epoch": 1.5829186065179002,
+      "grad_norm": 0.8611066341400146,
+      "learning_rate": 0.00011979995075213946,
+      "loss": 0.6781,
+      "step": 4930
+    },
+    {
+      "epoch": 1.5832396853427517,
+      "grad_norm": 0.8396785259246826,
+      "learning_rate": 0.00011976573403791262,
+      "loss": 0.6383,
+      "step": 4931
+    },
+    {
+      "epoch": 1.5835607641676033,
+      "grad_norm": 0.779766321182251,
+      "learning_rate": 0.00011973151491529107,
+      "loss": 0.6317,
+      "step": 4932
+    },
+    {
+      "epoch": 1.5838818429924546,
+      "grad_norm": 0.8786318898200989,
+      "learning_rate": 0.00011969729338844429,
+      "loss": 0.683,
+      "step": 4933
+    },
+    {
+      "epoch": 1.5842029218173062,
+      "grad_norm": 1.0004550218582153,
+      "learning_rate": 0.000119663069461542,
+      "loss": 0.6376,
+      "step": 4934
+    },
+    {
+      "epoch": 1.5845240006421577,
+      "grad_norm": 0.8501437306404114,
+      "learning_rate": 0.0001196288431387544,
+      "loss": 0.6203,
+      "step": 4935
+    },
+    {
+      "epoch": 1.584845079467009,
+      "grad_norm": 1.1006617546081543,
+      "learning_rate": 0.00011959461442425177,
+      "loss": 0.8655,
+      "step": 4936
+    },
+    {
+      "epoch": 1.5851661582918606,
+      "grad_norm": 1.571334719657898,
+      "learning_rate": 0.00011956038332220483,
+      "loss": 0.5745,
+      "step": 4937
+    },
+    {
+      "epoch": 1.5854872371167121,
+      "grad_norm": 0.8604090213775635,
+      "learning_rate": 0.00011952614983678452,
+      "loss": 0.5597,
+      "step": 4938
+    },
+    {
+      "epoch": 1.5858083159415637,
+      "grad_norm": 0.9236046671867371,
+      "learning_rate": 0.00011949191397216206,
+      "loss": 0.5838,
+      "step": 4939
+    },
+    {
+      "epoch": 1.5861293947664152,
+      "grad_norm": 0.8020102381706238,
+      "learning_rate": 0.00011945767573250903,
+      "loss": 0.5327,
+      "step": 4940
+    },
+    {
+      "epoch": 1.5864504735912668,
+      "grad_norm": 1.1067618131637573,
+      "learning_rate": 0.0001194234351219972,
+      "loss": 0.5908,
+      "step": 4941
+    },
+    {
+      "epoch": 1.5867715524161181,
+      "grad_norm": 1.1690189838409424,
+      "learning_rate": 0.00011938919214479876,
+      "loss": 0.6194,
+      "step": 4942
+    },
+    {
+      "epoch": 1.5870926312409697,
+      "grad_norm": 0.883562445640564,
+      "learning_rate": 0.00011935494680508606,
+      "loss": 0.487,
+      "step": 4943
+    },
+    {
+      "epoch": 1.587413710065821,
+      "grad_norm": 0.8721025586128235,
+      "learning_rate": 0.00011932069910703176,
+      "loss": 0.4942,
+      "step": 4944
+    },
+    {
+      "epoch": 1.5877347888906725,
+      "grad_norm": 1.2519181966781616,
+      "learning_rate": 0.0001192864490548089,
+      "loss": 0.6544,
+      "step": 4945
+    },
+    {
+      "epoch": 1.588055867715524,
+      "grad_norm": 0.8696319460868835,
+      "learning_rate": 0.00011925219665259075,
+      "loss": 0.5955,
+      "step": 4946
+    },
+    {
+      "epoch": 1.5883769465403756,
+      "grad_norm": 0.7267146706581116,
+      "learning_rate": 0.00011921794190455082,
+      "loss": 0.5205,
+      "step": 4947
+    },
+    {
+      "epoch": 1.5886980253652272,
+      "grad_norm": 0.9043694138526917,
+      "learning_rate": 0.00011918368481486297,
+      "loss": 0.6211,
+      "step": 4948
+    },
+    {
+      "epoch": 1.5890191041900787,
+      "grad_norm": 0.7768150568008423,
+      "learning_rate": 0.00011914942538770131,
+      "loss": 0.5542,
+      "step": 4949
+    },
+    {
+      "epoch": 1.5893401830149303,
+      "grad_norm": 1.070797324180603,
+      "learning_rate": 0.00011911516362724024,
+      "loss": 0.5892,
+      "step": 4950
+    },
+    {
+      "epoch": 1.5896612618397816,
+      "grad_norm": 0.8269063234329224,
+      "learning_rate": 0.00011908089953765449,
+      "loss": 0.5293,
+      "step": 4951
+    },
+    {
+      "epoch": 1.5899823406646332,
+      "grad_norm": 0.8718098402023315,
+      "learning_rate": 0.00011904663312311901,
+      "loss": 0.5791,
+      "step": 4952
+    },
+    {
+      "epoch": 1.5903034194894845,
+      "grad_norm": 0.6878900527954102,
+      "learning_rate": 0.00011901236438780902,
+      "loss": 0.3673,
+      "step": 4953
+    },
+    {
+      "epoch": 1.590624498314336,
+      "grad_norm": 0.8858333230018616,
+      "learning_rate": 0.00011897809333590014,
+      "loss": 0.5325,
+      "step": 4954
+    },
+    {
+      "epoch": 1.5909455771391876,
+      "grad_norm": 0.894816517829895,
+      "learning_rate": 0.00011894381997156813,
+      "loss": 0.4611,
+      "step": 4955
+    },
+    {
+      "epoch": 1.5912666559640392,
+      "grad_norm": 1.1014503240585327,
+      "learning_rate": 0.00011890954429898912,
+      "loss": 0.6538,
+      "step": 4956
+    },
+    {
+      "epoch": 1.5915877347888907,
+      "grad_norm": 0.9481959342956543,
+      "learning_rate": 0.00011887526632233954,
+      "loss": 0.5716,
+      "step": 4957
+    },
+    {
+      "epoch": 1.5919088136137423,
+      "grad_norm": 1.010277271270752,
+      "learning_rate": 0.00011884098604579597,
+      "loss": 0.7025,
+      "step": 4958
+    },
+    {
+      "epoch": 1.5922298924385938,
+      "grad_norm": 0.8406404256820679,
+      "learning_rate": 0.00011880670347353539,
+      "loss": 0.4381,
+      "step": 4959
+    },
+    {
+      "epoch": 1.5925509712634451,
+      "grad_norm": 0.948979914188385,
+      "learning_rate": 0.00011877241860973507,
+      "loss": 0.5589,
+      "step": 4960
+    },
+    {
+      "epoch": 1.5928720500882967,
+      "grad_norm": 0.9533628821372986,
+      "learning_rate": 0.00011873813145857249,
+      "loss": 0.5339,
+      "step": 4961
+    },
+    {
+      "epoch": 1.593193128913148,
+      "grad_norm": 1.116711974143982,
+      "learning_rate": 0.0001187038420242254,
+      "loss": 0.4653,
+      "step": 4962
+    },
+    {
+      "epoch": 1.5935142077379996,
+      "grad_norm": 0.6900465488433838,
+      "learning_rate": 0.0001186695503108719,
+      "loss": 0.4061,
+      "step": 4963
+    },
+    {
+      "epoch": 1.5938352865628511,
+      "grad_norm": 0.548774003982544,
+      "learning_rate": 0.00011863525632269032,
+      "loss": 0.3729,
+      "step": 4964
+    },
+    {
+      "epoch": 1.5941563653877027,
+      "grad_norm": 0.519191324710846,
+      "learning_rate": 0.0001186009600638593,
+      "loss": 0.5593,
+      "step": 4965
+    },
+    {
+      "epoch": 1.5944774442125542,
+      "grad_norm": 0.5707334280014038,
+      "learning_rate": 0.00011856666153855776,
+      "loss": 0.4427,
+      "step": 4966
+    },
+    {
+      "epoch": 1.5947985230374058,
+      "grad_norm": 0.5691962242126465,
+      "learning_rate": 0.00011853236075096474,
+      "loss": 0.264,
+      "step": 4967
+    },
+    {
+      "epoch": 1.5951196018622573,
+      "grad_norm": 0.8893996477127075,
+      "learning_rate": 0.00011849805770525983,
+      "loss": 0.4208,
+      "step": 4968
+    },
+    {
+      "epoch": 1.5954406806871086,
+      "grad_norm": 1.1631325483322144,
+      "learning_rate": 0.0001184637524056227,
+      "loss": 0.2257,
+      "step": 4969
+    },
+    {
+      "epoch": 1.5957617595119602,
+      "grad_norm": 0.7988505363464355,
+      "learning_rate": 0.00011842944485623335,
+      "loss": 0.2805,
+      "step": 4970
+    },
+    {
+      "epoch": 1.5960828383368115,
+      "grad_norm": 0.6999688148498535,
+      "learning_rate": 0.00011839513506127203,
+      "loss": 0.3814,
+      "step": 4971
+    },
+    {
+      "epoch": 1.596403917161663,
+      "grad_norm": 0.7029821276664734,
+      "learning_rate": 0.0001183608230249193,
+      "loss": 0.5627,
+      "step": 4972
+    },
+    {
+      "epoch": 1.5967249959865146,
+      "grad_norm": 0.8678010702133179,
+      "learning_rate": 0.00011832650875135598,
+      "loss": 0.7244,
+      "step": 4973
+    },
+    {
+      "epoch": 1.5970460748113662,
+      "grad_norm": 0.8419655561447144,
+      "learning_rate": 0.00011829219224476318,
+      "loss": 0.6425,
+      "step": 4974
+    },
+    {
+      "epoch": 1.5973671536362177,
+      "grad_norm": 0.9832744598388672,
+      "learning_rate": 0.00011825787350932222,
+      "loss": 0.7158,
+      "step": 4975
+    },
+    {
+      "epoch": 1.5976882324610693,
+      "grad_norm": 0.9112509489059448,
+      "learning_rate": 0.00011822355254921478,
+      "loss": 0.5684,
+      "step": 4976
+    },
+    {
+      "epoch": 1.5980093112859208,
+      "grad_norm": 1.4586632251739502,
+      "learning_rate": 0.00011818922936862269,
+      "loss": 0.4982,
+      "step": 4977
+    },
+    {
+      "epoch": 1.5983303901107722,
+      "grad_norm": 0.8100280165672302,
+      "learning_rate": 0.00011815490397172821,
+      "loss": 0.6925,
+      "step": 4978
+    },
+    {
+      "epoch": 1.5986514689356237,
+      "grad_norm": 0.8247403502464294,
+      "learning_rate": 0.00011812057636271374,
+      "loss": 0.6636,
+      "step": 4979
+    },
+    {
+      "epoch": 1.598972547760475,
+      "grad_norm": 0.8497698903083801,
+      "learning_rate": 0.00011808624654576202,
+      "loss": 0.5672,
+      "step": 4980
+    },
+    {
+      "epoch": 1.5992936265853266,
+      "grad_norm": 0.9448472261428833,
+      "learning_rate": 0.00011805191452505602,
+      "loss": 0.6503,
+      "step": 4981
+    },
+    {
+      "epoch": 1.5996147054101781,
+      "grad_norm": 1.0496573448181152,
+      "learning_rate": 0.00011801758030477897,
+      "loss": 0.8055,
+      "step": 4982
+    },
+    {
+      "epoch": 1.5999357842350297,
+      "grad_norm": 0.7179785966873169,
+      "learning_rate": 0.00011798324388911444,
+      "loss": 0.4881,
+      "step": 4983
+    },
+    {
+      "epoch": 1.6002568630598812,
+      "grad_norm": 0.8595174551010132,
+      "learning_rate": 0.00011794890528224618,
+      "loss": 0.615,
+      "step": 4984
+    },
+    {
+      "epoch": 1.6005779418847328,
+      "grad_norm": 0.8961060047149658,
+      "learning_rate": 0.00011791456448835825,
+      "loss": 0.6427,
+      "step": 4985
+    },
+    {
+      "epoch": 1.6008990207095843,
+      "grad_norm": 0.963125467300415,
+      "learning_rate": 0.00011788022151163495,
+      "loss": 0.7784,
+      "step": 4986
+    },
+    {
+      "epoch": 1.6012200995344357,
+      "grad_norm": 0.7629128694534302,
+      "learning_rate": 0.00011784587635626094,
+      "loss": 0.509,
+      "step": 4987
+    },
+    {
+      "epoch": 1.6015411783592872,
+      "grad_norm": 1.0571444034576416,
+      "learning_rate": 0.000117811529026421,
+      "loss": 0.7576,
+      "step": 4988
+    },
+    {
+      "epoch": 1.6018622571841385,
+      "grad_norm": 1.0005722045898438,
+      "learning_rate": 0.00011777717952630031,
+      "loss": 0.6151,
+      "step": 4989
+    },
+    {
+      "epoch": 1.60218333600899,
+      "grad_norm": 0.7066619396209717,
+      "learning_rate": 0.00011774282786008422,
+      "loss": 0.5324,
+      "step": 4990
+    },
+    {
+      "epoch": 1.6025044148338417,
+      "grad_norm": 0.6396430730819702,
+      "learning_rate": 0.00011770847403195834,
+      "loss": 0.5243,
+      "step": 4991
+    },
+    {
+      "epoch": 1.6028254936586932,
+      "grad_norm": 1.4114651679992676,
+      "learning_rate": 0.00011767411804610864,
+      "loss": 0.4948,
+      "step": 4992
+    },
+    {
+      "epoch": 1.6031465724835448,
+      "grad_norm": 0.9310147166252136,
+      "learning_rate": 0.00011763975990672125,
+      "loss": 0.7609,
+      "step": 4993
+    },
+    {
+      "epoch": 1.6034676513083963,
+      "grad_norm": 0.8665596842765808,
+      "learning_rate": 0.00011760539961798262,
+      "loss": 0.6365,
+      "step": 4994
+    },
+    {
+      "epoch": 1.6037887301332479,
+      "grad_norm": 0.831099808216095,
+      "learning_rate": 0.00011757103718407947,
+      "loss": 0.5575,
+      "step": 4995
+    },
+    {
+      "epoch": 1.6041098089580992,
+      "grad_norm": 0.9257915616035461,
+      "learning_rate": 0.00011753667260919872,
+      "loss": 0.5272,
+      "step": 4996
+    },
+    {
+      "epoch": 1.6044308877829507,
+      "grad_norm": 1.372697353363037,
+      "learning_rate": 0.00011750230589752762,
+      "loss": 0.5669,
+      "step": 4997
+    },
+    {
+      "epoch": 1.604751966607802,
+      "grad_norm": 0.978588342666626,
+      "learning_rate": 0.00011746793705325363,
+      "loss": 0.532,
+      "step": 4998
+    },
+    {
+      "epoch": 1.6050730454326536,
+      "grad_norm": 1.0235109329223633,
+      "learning_rate": 0.00011743356608056449,
+      "loss": 0.7292,
+      "step": 4999
+    },
+    {
+      "epoch": 1.6053941242575052,
+      "grad_norm": 1.004314661026001,
+      "learning_rate": 0.0001173991929836482,
+      "loss": 0.5686,
+      "step": 5000
+    },
+    {
+      "epoch": 1.6057152030823567,
+      "grad_norm": 1.8537969589233398,
+      "learning_rate": 0.00011736481776669306,
+      "loss": 0.4302,
+      "step": 5001
+    },
+    {
+      "epoch": 1.6060362819072083,
+      "grad_norm": 0.9163577556610107,
+      "learning_rate": 0.00011733044043388752,
+      "loss": 0.5876,
+      "step": 5002
+    },
+    {
+      "epoch": 1.6063573607320598,
+      "grad_norm": 1.2079983949661255,
+      "learning_rate": 0.00011729606098942039,
+      "loss": 0.6811,
+      "step": 5003
+    },
+    {
+      "epoch": 1.6066784395569114,
+      "grad_norm": 1.0708050727844238,
+      "learning_rate": 0.00011726167943748067,
+      "loss": 0.5863,
+      "step": 5004
+    },
+    {
+      "epoch": 1.6069995183817627,
+      "grad_norm": 0.7910842895507812,
+      "learning_rate": 0.00011722729578225769,
+      "loss": 0.4729,
+      "step": 5005
+    },
+    {
+      "epoch": 1.6073205972066142,
+      "grad_norm": 0.8176678419113159,
+      "learning_rate": 0.00011719291002794096,
+      "loss": 0.5986,
+      "step": 5006
+    },
+    {
+      "epoch": 1.6076416760314656,
+      "grad_norm": 1.354212760925293,
+      "learning_rate": 0.0001171585221787203,
+      "loss": 0.5428,
+      "step": 5007
+    },
+    {
+      "epoch": 1.6079627548563171,
+      "grad_norm": 1.120445966720581,
+      "learning_rate": 0.00011712413223878578,
+      "loss": 0.5019,
+      "step": 5008
+    },
+    {
+      "epoch": 1.6082838336811687,
+      "grad_norm": 0.7986057996749878,
+      "learning_rate": 0.00011708974021232769,
+      "loss": 0.4202,
+      "step": 5009
+    },
+    {
+      "epoch": 1.6086049125060202,
+      "grad_norm": 0.9796577095985413,
+      "learning_rate": 0.00011705534610353657,
+      "loss": 0.5712,
+      "step": 5010
+    },
+    {
+      "epoch": 1.6089259913308718,
+      "grad_norm": 1.225272297859192,
+      "learning_rate": 0.00011702094991660326,
+      "loss": 0.615,
+      "step": 5011
+    },
+    {
+      "epoch": 1.6092470701557233,
+      "grad_norm": 0.8488008975982666,
+      "learning_rate": 0.00011698655165571886,
+      "loss": 0.4476,
+      "step": 5012
+    },
+    {
+      "epoch": 1.6095681489805749,
+      "grad_norm": 0.8429365158081055,
+      "learning_rate": 0.00011695215132507464,
+      "loss": 0.5331,
+      "step": 5013
+    },
+    {
+      "epoch": 1.6098892278054262,
+      "grad_norm": 0.8646475672721863,
+      "learning_rate": 0.00011691774892886222,
+      "loss": 0.3656,
+      "step": 5014
+    },
+    {
+      "epoch": 1.6102103066302778,
+      "grad_norm": 0.603814423084259,
+      "learning_rate": 0.00011688334447127338,
+      "loss": 0.3372,
+      "step": 5015
+    },
+    {
+      "epoch": 1.610531385455129,
+      "grad_norm": 0.5882185697555542,
+      "learning_rate": 0.00011684893795650027,
+      "loss": 0.9782,
+      "step": 5016
+    },
+    {
+      "epoch": 1.6108524642799806,
+      "grad_norm": 0.5701581239700317,
+      "learning_rate": 0.00011681452938873516,
+      "loss": 0.3712,
+      "step": 5017
+    },
+    {
+      "epoch": 1.6111735431048322,
+      "grad_norm": 0.9942889213562012,
+      "learning_rate": 0.00011678011877217065,
+      "loss": 0.4755,
+      "step": 5018
+    },
+    {
+      "epoch": 1.6114946219296837,
+      "grad_norm": 0.8010011911392212,
+      "learning_rate": 0.00011674570611099955,
+      "loss": 0.3993,
+      "step": 5019
+    },
+    {
+      "epoch": 1.6118157007545353,
+      "grad_norm": 0.7642001509666443,
+      "learning_rate": 0.00011671129140941499,
+      "loss": 0.4508,
+      "step": 5020
+    },
+    {
+      "epoch": 1.6121367795793868,
+      "grad_norm": 0.7442882061004639,
+      "learning_rate": 0.00011667687467161024,
+      "loss": 0.5295,
+      "step": 5021
+    },
+    {
+      "epoch": 1.6124578584042384,
+      "grad_norm": 0.8074431419372559,
+      "learning_rate": 0.00011664245590177892,
+      "loss": 0.6415,
+      "step": 5022
+    },
+    {
+      "epoch": 1.6127789372290897,
+      "grad_norm": 0.8954799175262451,
+      "learning_rate": 0.0001166080351041148,
+      "loss": 0.6567,
+      "step": 5023
+    },
+    {
+      "epoch": 1.6131000160539413,
+      "grad_norm": 0.8098952174186707,
+      "learning_rate": 0.00011657361228281199,
+      "loss": 0.6164,
+      "step": 5024
+    },
+    {
+      "epoch": 1.6134210948787926,
+      "grad_norm": 1.059260368347168,
+      "learning_rate": 0.00011653918744206478,
+      "loss": 0.4956,
+      "step": 5025
+    },
+    {
+      "epoch": 1.6137421737036441,
+      "grad_norm": 0.6929076910018921,
+      "learning_rate": 0.00011650476058606777,
+      "loss": 0.5855,
+      "step": 5026
+    },
+    {
+      "epoch": 1.6140632525284957,
+      "grad_norm": 0.9725719094276428,
+      "learning_rate": 0.00011647033171901573,
+      "loss": 0.6258,
+      "step": 5027
+    },
+    {
+      "epoch": 1.6143843313533472,
+      "grad_norm": 0.8998417258262634,
+      "learning_rate": 0.00011643590084510379,
+      "loss": 0.6263,
+      "step": 5028
+    },
+    {
+      "epoch": 1.6147054101781988,
+      "grad_norm": 1.7190628051757812,
+      "learning_rate": 0.00011640146796852711,
+      "loss": 0.6147,
+      "step": 5029
+    },
+    {
+      "epoch": 1.6150264890030503,
+      "grad_norm": 0.8948490023612976,
+      "learning_rate": 0.00011636703309348133,
+      "loss": 0.6867,
+      "step": 5030
+    },
+    {
+      "epoch": 1.615347567827902,
+      "grad_norm": 0.9756568670272827,
+      "learning_rate": 0.00011633259622416224,
+      "loss": 0.7751,
+      "step": 5031
+    },
+    {
+      "epoch": 1.6156686466527532,
+      "grad_norm": 0.9897345304489136,
+      "learning_rate": 0.00011629815736476581,
+      "loss": 0.6944,
+      "step": 5032
+    },
+    {
+      "epoch": 1.6159897254776048,
+      "grad_norm": 1.4569859504699707,
+      "learning_rate": 0.00011626371651948838,
+      "loss": 0.5682,
+      "step": 5033
+    },
+    {
+      "epoch": 1.616310804302456,
+      "grad_norm": 1.191692590713501,
+      "learning_rate": 0.00011622927369252638,
+      "loss": 0.7242,
+      "step": 5034
+    },
+    {
+      "epoch": 1.6166318831273077,
+      "grad_norm": 0.7807984948158264,
+      "learning_rate": 0.00011619482888807662,
+      "loss": 0.5721,
+      "step": 5035
+    },
+    {
+      "epoch": 1.6169529619521592,
+      "grad_norm": 0.7991989254951477,
+      "learning_rate": 0.00011616038211033613,
+      "loss": 0.5288,
+      "step": 5036
+    },
+    {
+      "epoch": 1.6172740407770108,
+      "grad_norm": 0.8879060745239258,
+      "learning_rate": 0.00011612593336350208,
+      "loss": 0.4782,
+      "step": 5037
+    },
+    {
+      "epoch": 1.6175951196018623,
+      "grad_norm": 0.8001835942268372,
+      "learning_rate": 0.00011609148265177193,
+      "loss": 0.504,
+      "step": 5038
+    },
+    {
+      "epoch": 1.6179161984267139,
+      "grad_norm": 0.9068782925605774,
+      "learning_rate": 0.00011605702997934345,
+      "loss": 0.5692,
+      "step": 5039
+    },
+    {
+      "epoch": 1.6182372772515654,
+      "grad_norm": 0.9409818649291992,
+      "learning_rate": 0.00011602257535041459,
+      "loss": 0.5669,
+      "step": 5040
+    },
+    {
+      "epoch": 1.6185583560764167,
+      "grad_norm": 0.9030987620353699,
+      "learning_rate": 0.0001159881187691835,
+      "loss": 0.4359,
+      "step": 5041
+    },
+    {
+      "epoch": 1.6188794349012683,
+      "grad_norm": 1.0598105192184448,
+      "learning_rate": 0.00011595366023984864,
+      "loss": 0.693,
+      "step": 5042
+    },
+    {
+      "epoch": 1.6192005137261196,
+      "grad_norm": 1.0515077114105225,
+      "learning_rate": 0.00011591919976660868,
+      "loss": 0.6296,
+      "step": 5043
+    },
+    {
+      "epoch": 1.6195215925509712,
+      "grad_norm": 0.8490827083587646,
+      "learning_rate": 0.00011588473735366249,
+      "loss": 0.5571,
+      "step": 5044
+    },
+    {
+      "epoch": 1.6198426713758227,
+      "grad_norm": 0.8692120909690857,
+      "learning_rate": 0.0001158502730052093,
+      "loss": 0.5099,
+      "step": 5045
+    },
+    {
+      "epoch": 1.6201637502006743,
+      "grad_norm": 1.1289715766906738,
+      "learning_rate": 0.00011581580672544838,
+      "loss": 0.7012,
+      "step": 5046
+    },
+    {
+      "epoch": 1.6204848290255258,
+      "grad_norm": 1.281795859336853,
+      "learning_rate": 0.0001157813385185794,
+      "loss": 0.7737,
+      "step": 5047
+    },
+    {
+      "epoch": 1.6208059078503774,
+      "grad_norm": 1.1194076538085938,
+      "learning_rate": 0.00011574686838880215,
+      "loss": 0.633,
+      "step": 5048
+    },
+    {
+      "epoch": 1.621126986675229,
+      "grad_norm": 0.8962092399597168,
+      "learning_rate": 0.00011571239634031679,
+      "loss": 0.6357,
+      "step": 5049
+    },
+    {
+      "epoch": 1.6214480655000802,
+      "grad_norm": 0.9650201797485352,
+      "learning_rate": 0.00011567792237732358,
+      "loss": 0.5478,
+      "step": 5050
+    },
+    {
+      "epoch": 1.6217691443249318,
+      "grad_norm": 0.8813601732254028,
+      "learning_rate": 0.0001156434465040231,
+      "loss": 0.4708,
+      "step": 5051
+    },
+    {
+      "epoch": 1.6220902231497831,
+      "grad_norm": 0.8140305876731873,
+      "learning_rate": 0.0001156089687246161,
+      "loss": 0.485,
+      "step": 5052
+    },
+    {
+      "epoch": 1.6224113019746347,
+      "grad_norm": 1.1410284042358398,
+      "learning_rate": 0.00011557448904330362,
+      "loss": 0.6938,
+      "step": 5053
+    },
+    {
+      "epoch": 1.6227323807994862,
+      "grad_norm": 1.0146127939224243,
+      "learning_rate": 0.0001155400074642869,
+      "loss": 0.6013,
+      "step": 5054
+    },
+    {
+      "epoch": 1.6230534596243378,
+      "grad_norm": 1.0160109996795654,
+      "learning_rate": 0.00011550552399176739,
+      "loss": 0.4768,
+      "step": 5055
+    },
+    {
+      "epoch": 1.6233745384491893,
+      "grad_norm": 1.1787000894546509,
+      "learning_rate": 0.00011547103862994684,
+      "loss": 0.5916,
+      "step": 5056
+    },
+    {
+      "epoch": 1.6236956172740409,
+      "grad_norm": 0.8274866342544556,
+      "learning_rate": 0.00011543655138302714,
+      "loss": 0.5369,
+      "step": 5057
+    },
+    {
+      "epoch": 1.6240166960988924,
+      "grad_norm": 0.7210972905158997,
+      "learning_rate": 0.00011540206225521046,
+      "loss": 0.3732,
+      "step": 5058
+    },
+    {
+      "epoch": 1.6243377749237438,
+      "grad_norm": 1.1187121868133545,
+      "learning_rate": 0.00011536757125069923,
+      "loss": 0.658,
+      "step": 5059
+    },
+    {
+      "epoch": 1.6246588537485953,
+      "grad_norm": 0.6988892555236816,
+      "learning_rate": 0.00011533307837369607,
+      "loss": 0.3963,
+      "step": 5060
+    },
+    {
+      "epoch": 1.6249799325734466,
+      "grad_norm": 0.8323413133621216,
+      "learning_rate": 0.00011529858362840382,
+      "loss": 0.3751,
+      "step": 5061
+    },
+    {
+      "epoch": 1.6253010113982982,
+      "grad_norm": 1.572111964225769,
+      "learning_rate": 0.00011526408701902556,
+      "loss": 0.5879,
+      "step": 5062
+    },
+    {
+      "epoch": 1.6256220902231497,
+      "grad_norm": 0.6193574666976929,
+      "learning_rate": 0.00011522958854976458,
+      "loss": 0.3986,
+      "step": 5063
+    },
+    {
+      "epoch": 1.6259431690480013,
+      "grad_norm": 0.6282438635826111,
+      "learning_rate": 0.00011519508822482446,
+      "loss": 0.389,
+      "step": 5064
+    },
+    {
+      "epoch": 1.6262642478728528,
+      "grad_norm": 0.6472731828689575,
+      "learning_rate": 0.00011516058604840891,
+      "loss": 0.7557,
+      "step": 5065
+    },
+    {
+      "epoch": 1.6265853266977044,
+      "grad_norm": 0.5297544598579407,
+      "learning_rate": 0.00011512608202472194,
+      "loss": 0.7176,
+      "step": 5066
+    },
+    {
+      "epoch": 1.626906405522556,
+      "grad_norm": 0.6865007281303406,
+      "learning_rate": 0.00011509157615796776,
+      "loss": 0.3564,
+      "step": 5067
+    },
+    {
+      "epoch": 1.6272274843474073,
+      "grad_norm": 0.6907351613044739,
+      "learning_rate": 0.00011505706845235078,
+      "loss": 0.2377,
+      "step": 5068
+    },
+    {
+      "epoch": 1.6275485631722588,
+      "grad_norm": 0.8242547512054443,
+      "learning_rate": 0.00011502255891207572,
+      "loss": 0.2439,
+      "step": 5069
+    },
+    {
+      "epoch": 1.6278696419971102,
+      "grad_norm": 0.5880115032196045,
+      "learning_rate": 0.0001149880475413474,
+      "loss": 0.2065,
+      "step": 5070
+    },
+    {
+      "epoch": 1.6281907208219617,
+      "grad_norm": 0.6865981817245483,
+      "learning_rate": 0.00011495353434437098,
+      "loss": 0.1763,
+      "step": 5071
+    },
+    {
+      "epoch": 1.6285117996468133,
+      "grad_norm": 0.7891075015068054,
+      "learning_rate": 0.00011491901932535172,
+      "loss": 0.5724,
+      "step": 5072
+    },
+    {
+      "epoch": 1.6288328784716648,
+      "grad_norm": 1.0175186395645142,
+      "learning_rate": 0.00011488450248849522,
+      "loss": 0.884,
+      "step": 5073
+    },
+    {
+      "epoch": 1.6291539572965164,
+      "grad_norm": 0.808527946472168,
+      "learning_rate": 0.00011484998383800726,
+      "loss": 0.5626,
+      "step": 5074
+    },
+    {
+      "epoch": 1.629475036121368,
+      "grad_norm": 0.8275448679924011,
+      "learning_rate": 0.00011481546337809381,
+      "loss": 0.5928,
+      "step": 5075
+    },
+    {
+      "epoch": 1.6297961149462195,
+      "grad_norm": 0.7020288109779358,
+      "learning_rate": 0.00011478094111296109,
+      "loss": 0.4619,
+      "step": 5076
+    },
+    {
+      "epoch": 1.6301171937710708,
+      "grad_norm": 0.7639602422714233,
+      "learning_rate": 0.0001147464170468155,
+      "loss": 0.5046,
+      "step": 5077
+    },
+    {
+      "epoch": 1.6304382725959223,
+      "grad_norm": 0.7162854671478271,
+      "learning_rate": 0.00011471189118386375,
+      "loss": 0.5293,
+      "step": 5078
+    },
+    {
+      "epoch": 1.6307593514207737,
+      "grad_norm": 0.826367199420929,
+      "learning_rate": 0.00011467736352831266,
+      "loss": 0.4409,
+      "step": 5079
+    },
+    {
+      "epoch": 1.6310804302456252,
+      "grad_norm": 0.9901473522186279,
+      "learning_rate": 0.00011464283408436938,
+      "loss": 0.7239,
+      "step": 5080
+    },
+    {
+      "epoch": 1.6314015090704768,
+      "grad_norm": 1.07185697555542,
+      "learning_rate": 0.00011460830285624118,
+      "loss": 0.6114,
+      "step": 5081
+    },
+    {
+      "epoch": 1.6317225878953283,
+      "grad_norm": 1.0970301628112793,
+      "learning_rate": 0.00011457376984813557,
+      "loss": 0.6966,
+      "step": 5082
+    },
+    {
+      "epoch": 1.6320436667201799,
+      "grad_norm": 0.672584056854248,
+      "learning_rate": 0.00011453923506426032,
+      "loss": 0.4641,
+      "step": 5083
+    },
+    {
+      "epoch": 1.6323647455450314,
+      "grad_norm": 0.9183497428894043,
+      "learning_rate": 0.00011450469850882337,
+      "loss": 0.7209,
+      "step": 5084
+    },
+    {
+      "epoch": 1.632685824369883,
+      "grad_norm": 1.120105266571045,
+      "learning_rate": 0.00011447016018603292,
+      "loss": 0.7857,
+      "step": 5085
+    },
+    {
+      "epoch": 1.6330069031947343,
+      "grad_norm": 1.3121768236160278,
+      "learning_rate": 0.00011443562010009731,
+      "loss": 0.6645,
+      "step": 5086
+    },
+    {
+      "epoch": 1.6333279820195858,
+      "grad_norm": 0.8911232352256775,
+      "learning_rate": 0.00011440107825522521,
+      "loss": 0.5584,
+      "step": 5087
+    },
+    {
+      "epoch": 1.6336490608444372,
+      "grad_norm": 0.9025933146476746,
+      "learning_rate": 0.00011436653465562542,
+      "loss": 0.5403,
+      "step": 5088
+    },
+    {
+      "epoch": 1.6339701396692887,
+      "grad_norm": 0.9944098591804504,
+      "learning_rate": 0.00011433198930550695,
+      "loss": 0.5797,
+      "step": 5089
+    },
+    {
+      "epoch": 1.6342912184941403,
+      "grad_norm": 0.8134103417396545,
+      "learning_rate": 0.00011429744220907903,
+      "loss": 0.6454,
+      "step": 5090
+    },
+    {
+      "epoch": 1.6346122973189918,
+      "grad_norm": 0.9299569129943848,
+      "learning_rate": 0.00011426289337055119,
+      "loss": 0.6705,
+      "step": 5091
+    },
+    {
+      "epoch": 1.6349333761438434,
+      "grad_norm": 1.1119047403335571,
+      "learning_rate": 0.00011422834279413301,
+      "loss": 0.5778,
+      "step": 5092
+    },
+    {
+      "epoch": 1.635254454968695,
+      "grad_norm": 0.860736072063446,
+      "learning_rate": 0.00011419379048403444,
+      "loss": 0.687,
+      "step": 5093
+    },
+    {
+      "epoch": 1.6355755337935465,
+      "grad_norm": 1.328277349472046,
+      "learning_rate": 0.00011415923644446557,
+      "loss": 0.8227,
+      "step": 5094
+    },
+    {
+      "epoch": 1.6358966126183978,
+      "grad_norm": 1.0155832767486572,
+      "learning_rate": 0.00011412468067963669,
+      "loss": 0.61,
+      "step": 5095
+    },
+    {
+      "epoch": 1.6362176914432494,
+      "grad_norm": 0.8249773383140564,
+      "learning_rate": 0.00011409012319375827,
+      "loss": 0.454,
+      "step": 5096
+    },
+    {
+      "epoch": 1.6365387702681007,
+      "grad_norm": 0.8771111369132996,
+      "learning_rate": 0.00011405556399104109,
+      "loss": 0.6191,
+      "step": 5097
+    },
+    {
+      "epoch": 1.6368598490929522,
+      "grad_norm": 1.0984089374542236,
+      "learning_rate": 0.00011402100307569612,
+      "loss": 0.8062,
+      "step": 5098
+    },
+    {
+      "epoch": 1.6371809279178038,
+      "grad_norm": 0.7686770558357239,
+      "learning_rate": 0.00011398644045193444,
+      "loss": 0.4661,
+      "step": 5099
+    },
+    {
+      "epoch": 1.6375020067426553,
+      "grad_norm": 0.8252708911895752,
+      "learning_rate": 0.00011395187612396738,
+      "loss": 0.5061,
+      "step": 5100
+    },
+    {
+      "epoch": 1.6378230855675069,
+      "grad_norm": 1.2015011310577393,
+      "learning_rate": 0.00011391731009600654,
+      "loss": 0.6285,
+      "step": 5101
+    },
+    {
+      "epoch": 1.6381441643923584,
+      "grad_norm": 1.1227675676345825,
+      "learning_rate": 0.00011388274237226371,
+      "loss": 0.8034,
+      "step": 5102
+    },
+    {
+      "epoch": 1.63846524321721,
+      "grad_norm": 0.8477095365524292,
+      "learning_rate": 0.00011384817295695083,
+      "loss": 0.5733,
+      "step": 5103
+    },
+    {
+      "epoch": 1.6387863220420613,
+      "grad_norm": 0.7261363863945007,
+      "learning_rate": 0.00011381360185428007,
+      "loss": 0.3891,
+      "step": 5104
+    },
+    {
+      "epoch": 1.6391074008669129,
+      "grad_norm": 0.9272900223731995,
+      "learning_rate": 0.0001137790290684638,
+      "loss": 0.3385,
+      "step": 5105
+    },
+    {
+      "epoch": 1.6394284796917642,
+      "grad_norm": 1.001617431640625,
+      "learning_rate": 0.00011374445460371466,
+      "loss": 0.5374,
+      "step": 5106
+    },
+    {
+      "epoch": 1.6397495585166157,
+      "grad_norm": 0.8543923497200012,
+      "learning_rate": 0.00011370987846424546,
+      "loss": 0.5697,
+      "step": 5107
+    },
+    {
+      "epoch": 1.6400706373414673,
+      "grad_norm": 0.9451769590377808,
+      "learning_rate": 0.0001136753006542691,
+      "loss": 0.4807,
+      "step": 5108
+    },
+    {
+      "epoch": 1.6403917161663188,
+      "grad_norm": 0.6929908394813538,
+      "learning_rate": 0.00011364072117799885,
+      "loss": 0.4472,
+      "step": 5109
+    },
+    {
+      "epoch": 1.6407127949911704,
+      "grad_norm": 1.1182005405426025,
+      "learning_rate": 0.00011360614003964809,
+      "loss": 0.5891,
+      "step": 5110
+    },
+    {
+      "epoch": 1.641033873816022,
+      "grad_norm": 1.0447207689285278,
+      "learning_rate": 0.00011357155724343045,
+      "loss": 0.5113,
+      "step": 5111
+    },
+    {
+      "epoch": 1.6413549526408735,
+      "grad_norm": 0.7325683832168579,
+      "learning_rate": 0.00011353697279355973,
+      "loss": 0.3998,
+      "step": 5112
+    },
+    {
+      "epoch": 1.6416760314657248,
+      "grad_norm": 0.6994525790214539,
+      "learning_rate": 0.00011350238669424993,
+      "loss": 0.441,
+      "step": 5113
+    },
+    {
+      "epoch": 1.6419971102905764,
+      "grad_norm": 1.6808966398239136,
+      "learning_rate": 0.00011346779894971527,
+      "loss": 0.5446,
+      "step": 5114
+    },
+    {
+      "epoch": 1.6423181891154277,
+      "grad_norm": 0.5908676981925964,
+      "learning_rate": 0.00011343320956417014,
+      "loss": 0.8066,
+      "step": 5115
+    },
+    {
+      "epoch": 1.6426392679402793,
+      "grad_norm": 0.6339381337165833,
+      "learning_rate": 0.00011339861854182922,
+      "loss": 0.7703,
+      "step": 5116
+    },
+    {
+      "epoch": 1.6429603467651308,
+      "grad_norm": 0.7568504810333252,
+      "learning_rate": 0.00011336402588690726,
+      "loss": 0.5826,
+      "step": 5117
+    },
+    {
+      "epoch": 1.6432814255899824,
+      "grad_norm": 0.6824870705604553,
+      "learning_rate": 0.00011332943160361926,
+      "loss": 0.2832,
+      "step": 5118
+    },
+    {
+      "epoch": 1.643602504414834,
+      "grad_norm": 0.7612968683242798,
+      "learning_rate": 0.00011329483569618045,
+      "loss": 0.291,
+      "step": 5119
+    },
+    {
+      "epoch": 1.6439235832396855,
+      "grad_norm": 0.7312465906143188,
+      "learning_rate": 0.00011326023816880625,
+      "loss": 0.2184,
+      "step": 5120
+    },
+    {
+      "epoch": 1.6442446620645368,
+      "grad_norm": 0.7438070178031921,
+      "learning_rate": 0.00011322563902571226,
+      "loss": 0.3823,
+      "step": 5121
+    },
+    {
+      "epoch": 1.6445657408893883,
+      "grad_norm": 1.085114598274231,
+      "learning_rate": 0.00011319103827111426,
+      "loss": 0.6073,
+      "step": 5122
+    },
+    {
+      "epoch": 1.64488681971424,
+      "grad_norm": 0.8549251556396484,
+      "learning_rate": 0.00011315643590922827,
+      "loss": 0.6365,
+      "step": 5123
+    },
+    {
+      "epoch": 1.6452078985390912,
+      "grad_norm": 0.8248443603515625,
+      "learning_rate": 0.00011312183194427046,
+      "loss": 0.5795,
+      "step": 5124
+    },
+    {
+      "epoch": 1.6455289773639428,
+      "grad_norm": 0.9894840121269226,
+      "learning_rate": 0.00011308722638045724,
+      "loss": 0.5455,
+      "step": 5125
+    },
+    {
+      "epoch": 1.6458500561887943,
+      "grad_norm": 0.7669897079467773,
+      "learning_rate": 0.00011305261922200519,
+      "loss": 0.537,
+      "step": 5126
+    },
+    {
+      "epoch": 1.6461711350136459,
+      "grad_norm": 0.9028288722038269,
+      "learning_rate": 0.00011301801047313105,
+      "loss": 0.6561,
+      "step": 5127
+    },
+    {
+      "epoch": 1.6464922138384974,
+      "grad_norm": 0.8793859481811523,
+      "learning_rate": 0.00011298340013805184,
+      "loss": 0.6243,
+      "step": 5128
+    },
+    {
+      "epoch": 1.646813292663349,
+      "grad_norm": 0.7349303364753723,
+      "learning_rate": 0.00011294878822098469,
+      "loss": 0.5293,
+      "step": 5129
+    },
+    {
+      "epoch": 1.6471343714882003,
+      "grad_norm": 0.8124568462371826,
+      "learning_rate": 0.000112914174726147,
+      "loss": 0.5847,
+      "step": 5130
+    },
+    {
+      "epoch": 1.6474554503130519,
+      "grad_norm": 1.0718876123428345,
+      "learning_rate": 0.0001128795596577563,
+      "loss": 0.6405,
+      "step": 5131
+    },
+    {
+      "epoch": 1.6477765291379034,
+      "grad_norm": 0.7873112559318542,
+      "learning_rate": 0.0001128449430200303,
+      "loss": 0.5677,
+      "step": 5132
+    },
+    {
+      "epoch": 1.6480976079627547,
+      "grad_norm": 0.8473823666572571,
+      "learning_rate": 0.00011281032481718697,
+      "loss": 0.6249,
+      "step": 5133
+    },
+    {
+      "epoch": 1.6484186867876063,
+      "grad_norm": 0.8866380453109741,
+      "learning_rate": 0.0001127757050534444,
+      "loss": 0.7474,
+      "step": 5134
+    },
+    {
+      "epoch": 1.6487397656124578,
+      "grad_norm": 0.9700244069099426,
+      "learning_rate": 0.00011274108373302095,
+      "loss": 0.7294,
+      "step": 5135
+    },
+    {
+      "epoch": 1.6490608444373094,
+      "grad_norm": 0.9701023697853088,
+      "learning_rate": 0.00011270646086013505,
+      "loss": 0.8279,
+      "step": 5136
+    },
+    {
+      "epoch": 1.649381923262161,
+      "grad_norm": 0.9571521282196045,
+      "learning_rate": 0.00011267183643900548,
+      "loss": 0.4458,
+      "step": 5137
+    },
+    {
+      "epoch": 1.6497030020870125,
+      "grad_norm": 0.8921281099319458,
+      "learning_rate": 0.00011263721047385105,
+      "loss": 0.588,
+      "step": 5138
+    },
+    {
+      "epoch": 1.6500240809118638,
+      "grad_norm": 1.0599466562271118,
+      "learning_rate": 0.00011260258296889086,
+      "loss": 0.7079,
+      "step": 5139
+    },
+    {
+      "epoch": 1.6503451597367154,
+      "grad_norm": 0.9576237201690674,
+      "learning_rate": 0.00011256795392834419,
+      "loss": 0.6869,
+      "step": 5140
+    },
+    {
+      "epoch": 1.650666238561567,
+      "grad_norm": 1.035122036933899,
+      "learning_rate": 0.00011253332335643043,
+      "loss": 0.6376,
+      "step": 5141
+    },
+    {
+      "epoch": 1.6509873173864182,
+      "grad_norm": 1.0403311252593994,
+      "learning_rate": 0.00011249869125736925,
+      "loss": 0.5694,
+      "step": 5142
+    },
+    {
+      "epoch": 1.6513083962112698,
+      "grad_norm": 0.9405300617218018,
+      "learning_rate": 0.00011246405763538046,
+      "loss": 0.5892,
+      "step": 5143
+    },
+    {
+      "epoch": 1.6516294750361213,
+      "grad_norm": 1.0725064277648926,
+      "learning_rate": 0.00011242942249468402,
+      "loss": 0.7886,
+      "step": 5144
+    },
+    {
+      "epoch": 1.651950553860973,
+      "grad_norm": 1.1589728593826294,
+      "learning_rate": 0.00011239478583950018,
+      "loss": 0.6928,
+      "step": 5145
+    },
+    {
+      "epoch": 1.6522716326858244,
+      "grad_norm": 0.9433485865592957,
+      "learning_rate": 0.00011236014767404927,
+      "loss": 0.673,
+      "step": 5146
+    },
+    {
+      "epoch": 1.652592711510676,
+      "grad_norm": 0.7959662079811096,
+      "learning_rate": 0.00011232550800255188,
+      "loss": 0.555,
+      "step": 5147
+    },
+    {
+      "epoch": 1.6529137903355273,
+      "grad_norm": 0.8811464905738831,
+      "learning_rate": 0.00011229086682922869,
+      "loss": 0.5339,
+      "step": 5148
+    },
+    {
+      "epoch": 1.6532348691603789,
+      "grad_norm": 1.0386700630187988,
+      "learning_rate": 0.00011225622415830068,
+      "loss": 0.5246,
+      "step": 5149
+    },
+    {
+      "epoch": 1.6535559479852304,
+      "grad_norm": 0.907658040523529,
+      "learning_rate": 0.00011222157999398895,
+      "loss": 0.6589,
+      "step": 5150
+    },
+    {
+      "epoch": 1.6538770268100818,
+      "grad_norm": 0.6662940979003906,
+      "learning_rate": 0.00011218693434051475,
+      "loss": 0.5202,
+      "step": 5151
+    },
+    {
+      "epoch": 1.6541981056349333,
+      "grad_norm": 0.8499528765678406,
+      "learning_rate": 0.00011215228720209958,
+      "loss": 0.5316,
+      "step": 5152
+    },
+    {
+      "epoch": 1.6545191844597849,
+      "grad_norm": 0.8833222389221191,
+      "learning_rate": 0.00011211763858296507,
+      "loss": 0.534,
+      "step": 5153
+    },
+    {
+      "epoch": 1.6548402632846364,
+      "grad_norm": 0.8427216410636902,
+      "learning_rate": 0.00011208298848733305,
+      "loss": 0.4688,
+      "step": 5154
+    },
+    {
+      "epoch": 1.655161342109488,
+      "grad_norm": 0.808330774307251,
+      "learning_rate": 0.00011204833691942553,
+      "loss": 0.5204,
+      "step": 5155
+    },
+    {
+      "epoch": 1.6554824209343395,
+      "grad_norm": 0.6491779685020447,
+      "learning_rate": 0.00011201368388346471,
+      "loss": 0.383,
+      "step": 5156
+    },
+    {
+      "epoch": 1.6558034997591908,
+      "grad_norm": 1.3244563341140747,
+      "learning_rate": 0.00011197902938367298,
+      "loss": 0.5214,
+      "step": 5157
+    },
+    {
+      "epoch": 1.6561245785840424,
+      "grad_norm": 1.1561415195465088,
+      "learning_rate": 0.0001119443734242728,
+      "loss": 0.5834,
+      "step": 5158
+    },
+    {
+      "epoch": 1.656445657408894,
+      "grad_norm": 1.0299034118652344,
+      "learning_rate": 0.00011190971600948699,
+      "loss": 0.5443,
+      "step": 5159
+    },
+    {
+      "epoch": 1.6567667362337453,
+      "grad_norm": 0.8660802841186523,
+      "learning_rate": 0.0001118750571435384,
+      "loss": 0.4689,
+      "step": 5160
+    },
+    {
+      "epoch": 1.6570878150585968,
+      "grad_norm": 0.6859344840049744,
+      "learning_rate": 0.00011184039683065013,
+      "loss": 0.4929,
+      "step": 5161
+    },
+    {
+      "epoch": 1.6574088938834484,
+      "grad_norm": 1.2435836791992188,
+      "learning_rate": 0.00011180573507504537,
+      "loss": 0.5144,
+      "step": 5162
+    },
+    {
+      "epoch": 1.6577299727083,
+      "grad_norm": 0.9810152649879456,
+      "learning_rate": 0.00011177107188094764,
+      "loss": 0.4428,
+      "step": 5163
+    },
+    {
+      "epoch": 1.6580510515331515,
+      "grad_norm": 2.1121833324432373,
+      "learning_rate": 0.00011173640725258052,
+      "loss": 0.5309,
+      "step": 5164
+    },
+    {
+      "epoch": 1.658372130358003,
+      "grad_norm": 0.8293464183807373,
+      "learning_rate": 0.00011170174119416776,
+      "loss": 0.7781,
+      "step": 5165
+    },
+    {
+      "epoch": 1.6586932091828543,
+      "grad_norm": 0.5578963160514832,
+      "learning_rate": 0.00011166707370993333,
+      "loss": 0.5859,
+      "step": 5166
+    },
+    {
+      "epoch": 1.659014288007706,
+      "grad_norm": 0.5009366869926453,
+      "learning_rate": 0.00011163240480410135,
+      "loss": 0.6605,
+      "step": 5167
+    },
+    {
+      "epoch": 1.6593353668325574,
+      "grad_norm": 0.7301324605941772,
+      "learning_rate": 0.00011159773448089614,
+      "loss": 0.4758,
+      "step": 5168
+    },
+    {
+      "epoch": 1.6596564456574088,
+      "grad_norm": 0.6743203401565552,
+      "learning_rate": 0.00011156306274454218,
+      "loss": 0.3494,
+      "step": 5169
+    },
+    {
+      "epoch": 1.6599775244822603,
+      "grad_norm": 0.7202017903327942,
+      "learning_rate": 0.00011152838959926408,
+      "loss": 0.3711,
+      "step": 5170
+    },
+    {
+      "epoch": 1.6602986033071119,
+      "grad_norm": 0.6269795298576355,
+      "learning_rate": 0.00011149371504928668,
+      "loss": 0.2464,
+      "step": 5171
+    },
+    {
+      "epoch": 1.6606196821319634,
+      "grad_norm": 0.5528831481933594,
+      "learning_rate": 0.00011145903909883495,
+      "loss": 0.2153,
+      "step": 5172
+    },
+    {
+      "epoch": 1.660940760956815,
+      "grad_norm": 0.695706307888031,
+      "learning_rate": 0.00011142436175213409,
+      "loss": 0.5348,
+      "step": 5173
+    },
+    {
+      "epoch": 1.6612618397816665,
+      "grad_norm": 0.8045232892036438,
+      "learning_rate": 0.0001113896830134094,
+      "loss": 0.6501,
+      "step": 5174
+    },
+    {
+      "epoch": 1.6615829186065179,
+      "grad_norm": 1.1666730642318726,
+      "learning_rate": 0.00011135500288688636,
+      "loss": 0.6264,
+      "step": 5175
+    },
+    {
+      "epoch": 1.6619039974313694,
+      "grad_norm": 0.9477275013923645,
+      "learning_rate": 0.0001113203213767907,
+      "loss": 0.5997,
+      "step": 5176
+    },
+    {
+      "epoch": 1.662225076256221,
+      "grad_norm": 0.7591928839683533,
+      "learning_rate": 0.00011128563848734816,
+      "loss": 0.4988,
+      "step": 5177
+    },
+    {
+      "epoch": 1.6625461550810723,
+      "grad_norm": 1.0086168050765991,
+      "learning_rate": 0.00011125095422278486,
+      "loss": 0.5753,
+      "step": 5178
+    },
+    {
+      "epoch": 1.6628672339059238,
+      "grad_norm": 0.78431636095047,
+      "learning_rate": 0.0001112162685873269,
+      "loss": 0.6154,
+      "step": 5179
+    },
+    {
+      "epoch": 1.6631883127307754,
+      "grad_norm": 0.7329674959182739,
+      "learning_rate": 0.00011118158158520064,
+      "loss": 0.4773,
+      "step": 5180
+    },
+    {
+      "epoch": 1.663509391555627,
+      "grad_norm": 0.6751540303230286,
+      "learning_rate": 0.00011114689322063255,
+      "loss": 0.4195,
+      "step": 5181
+    },
+    {
+      "epoch": 1.6638304703804785,
+      "grad_norm": 0.8054059147834778,
+      "learning_rate": 0.00011111220349784937,
+      "loss": 0.5682,
+      "step": 5182
+    },
+    {
+      "epoch": 1.66415154920533,
+      "grad_norm": 1.0262163877487183,
+      "learning_rate": 0.00011107751242107787,
+      "loss": 0.8966,
+      "step": 5183
+    },
+    {
+      "epoch": 1.6644726280301814,
+      "grad_norm": 1.2697820663452148,
+      "learning_rate": 0.00011104281999454511,
+      "loss": 0.6868,
+      "step": 5184
+    },
+    {
+      "epoch": 1.664793706855033,
+      "grad_norm": 1.0350483655929565,
+      "learning_rate": 0.00011100812622247822,
+      "loss": 0.5431,
+      "step": 5185
+    },
+    {
+      "epoch": 1.6651147856798842,
+      "grad_norm": 0.7967930436134338,
+      "learning_rate": 0.00011097343110910452,
+      "loss": 0.3997,
+      "step": 5186
+    },
+    {
+      "epoch": 1.6654358645047358,
+      "grad_norm": 1.0638127326965332,
+      "learning_rate": 0.00011093873465865157,
+      "loss": 0.8347,
+      "step": 5187
+    },
+    {
+      "epoch": 1.6657569433295873,
+      "grad_norm": 0.8634060621261597,
+      "learning_rate": 0.00011090403687534697,
+      "loss": 0.5585,
+      "step": 5188
+    },
+    {
+      "epoch": 1.666078022154439,
+      "grad_norm": 1.0660980939865112,
+      "learning_rate": 0.00011086933776341852,
+      "loss": 0.722,
+      "step": 5189
+    },
+    {
+      "epoch": 1.6663991009792904,
+      "grad_norm": 1.0140385627746582,
+      "learning_rate": 0.00011083463732709425,
+      "loss": 0.5254,
+      "step": 5190
+    },
+    {
+      "epoch": 1.666720179804142,
+      "grad_norm": 1.2487151622772217,
+      "learning_rate": 0.0001107999355706023,
+      "loss": 0.6787,
+      "step": 5191
+    },
+    {
+      "epoch": 1.6670412586289935,
+      "grad_norm": 1.3040145635604858,
+      "learning_rate": 0.00011076523249817094,
+      "loss": 0.8144,
+      "step": 5192
+    },
+    {
+      "epoch": 1.6673623374538449,
+      "grad_norm": 0.9779611229896545,
+      "learning_rate": 0.00011073052811402867,
+      "loss": 0.6663,
+      "step": 5193
+    },
+    {
+      "epoch": 1.6676834162786964,
+      "grad_norm": 0.7410480976104736,
+      "learning_rate": 0.0001106958224224041,
+      "loss": 0.49,
+      "step": 5194
+    },
+    {
+      "epoch": 1.6680044951035478,
+      "grad_norm": 1.0676357746124268,
+      "learning_rate": 0.000110661115427526,
+      "loss": 0.6552,
+      "step": 5195
+    },
+    {
+      "epoch": 1.6683255739283993,
+      "grad_norm": 0.8795901536941528,
+      "learning_rate": 0.00011062640713362333,
+      "loss": 0.6453,
+      "step": 5196
+    },
+    {
+      "epoch": 1.6686466527532509,
+      "grad_norm": 0.8016363382339478,
+      "learning_rate": 0.0001105916975449252,
+      "loss": 0.5491,
+      "step": 5197
+    },
+    {
+      "epoch": 1.6689677315781024,
+      "grad_norm": 0.7814456820487976,
+      "learning_rate": 0.00011055698666566084,
+      "loss": 0.5246,
+      "step": 5198
+    },
+    {
+      "epoch": 1.669288810402954,
+      "grad_norm": 1.1860097646713257,
+      "learning_rate": 0.00011052227450005967,
+      "loss": 0.5291,
+      "step": 5199
+    },
+    {
+      "epoch": 1.6696098892278055,
+      "grad_norm": 0.7847801446914673,
+      "learning_rate": 0.00011048756105235125,
+      "loss": 0.5715,
+      "step": 5200
+    },
+    {
+      "epoch": 1.669930968052657,
+      "grad_norm": 0.6763100028038025,
+      "learning_rate": 0.00011045284632676536,
+      "loss": 0.4152,
+      "step": 5201
+    },
+    {
+      "epoch": 1.6702520468775084,
+      "grad_norm": 1.2399777173995972,
+      "learning_rate": 0.00011041813032753183,
+      "loss": 0.504,
+      "step": 5202
+    },
+    {
+      "epoch": 1.67057312570236,
+      "grad_norm": 0.9793505668640137,
+      "learning_rate": 0.00011038341305888074,
+      "loss": 0.7147,
+      "step": 5203
+    },
+    {
+      "epoch": 1.6708942045272113,
+      "grad_norm": 0.6837410926818848,
+      "learning_rate": 0.00011034869452504226,
+      "loss": 0.381,
+      "step": 5204
+    },
+    {
+      "epoch": 1.6712152833520628,
+      "grad_norm": 1.278186321258545,
+      "learning_rate": 0.00011031397473024674,
+      "loss": 0.5128,
+      "step": 5205
+    },
+    {
+      "epoch": 1.6715363621769144,
+      "grad_norm": 0.8159211277961731,
+      "learning_rate": 0.00011027925367872469,
+      "loss": 0.5699,
+      "step": 5206
+    },
+    {
+      "epoch": 1.671857441001766,
+      "grad_norm": 0.7557732462882996,
+      "learning_rate": 0.00011024453137470677,
+      "loss": 0.5079,
+      "step": 5207
+    },
+    {
+      "epoch": 1.6721785198266175,
+      "grad_norm": 0.9018636345863342,
+      "learning_rate": 0.00011020980782242376,
+      "loss": 0.533,
+      "step": 5208
+    },
+    {
+      "epoch": 1.672499598651469,
+      "grad_norm": 0.8341492414474487,
+      "learning_rate": 0.00011017508302610664,
+      "loss": 0.5423,
+      "step": 5209
+    },
+    {
+      "epoch": 1.6728206774763206,
+      "grad_norm": 1.1660737991333008,
+      "learning_rate": 0.00011014035698998651,
+      "loss": 0.4989,
+      "step": 5210
+    },
+    {
+      "epoch": 1.673141756301172,
+      "grad_norm": 0.9293453097343445,
+      "learning_rate": 0.00011010562971829463,
+      "loss": 0.4339,
+      "step": 5211
+    },
+    {
+      "epoch": 1.6734628351260235,
+      "grad_norm": 0.6236379146575928,
+      "learning_rate": 0.00011007090121526245,
+      "loss": 0.4249,
+      "step": 5212
+    },
+    {
+      "epoch": 1.6737839139508748,
+      "grad_norm": 0.8735912442207336,
+      "learning_rate": 0.00011003617148512149,
+      "loss": 0.5046,
+      "step": 5213
+    },
+    {
+      "epoch": 1.6741049927757263,
+      "grad_norm": 1.3508676290512085,
+      "learning_rate": 0.00011000144053210348,
+      "loss": 0.3542,
+      "step": 5214
+    },
+    {
+      "epoch": 1.6744260716005779,
+      "grad_norm": 1.2363876104354858,
+      "learning_rate": 0.0001099667083604403,
+      "loss": 0.4835,
+      "step": 5215
+    },
+    {
+      "epoch": 1.6747471504254294,
+      "grad_norm": 0.650037944316864,
+      "learning_rate": 0.00010993197497436391,
+      "loss": 0.7668,
+      "step": 5216
+    },
+    {
+      "epoch": 1.675068229250281,
+      "grad_norm": 0.6067812442779541,
+      "learning_rate": 0.00010989724037810652,
+      "loss": 0.5965,
+      "step": 5217
+    },
+    {
+      "epoch": 1.6753893080751325,
+      "grad_norm": 0.8295800685882568,
+      "learning_rate": 0.00010986250457590039,
+      "loss": 0.4706,
+      "step": 5218
+    },
+    {
+      "epoch": 1.675710386899984,
+      "grad_norm": 0.9519297480583191,
+      "learning_rate": 0.00010982776757197799,
+      "loss": 0.3641,
+      "step": 5219
+    },
+    {
+      "epoch": 1.6760314657248354,
+      "grad_norm": 0.8043132424354553,
+      "learning_rate": 0.00010979302937057192,
+      "loss": 0.3151,
+      "step": 5220
+    },
+    {
+      "epoch": 1.676352544549687,
+      "grad_norm": 0.8412678837776184,
+      "learning_rate": 0.00010975828997591495,
+      "loss": 0.4017,
+      "step": 5221
+    },
+    {
+      "epoch": 1.6766736233745383,
+      "grad_norm": 0.5943979620933533,
+      "learning_rate": 0.00010972354939223996,
+      "loss": 0.1791,
+      "step": 5222
+    },
+    {
+      "epoch": 1.6769947021993898,
+      "grad_norm": 0.7364015579223633,
+      "learning_rate": 0.00010968880762377993,
+      "loss": 0.2705,
+      "step": 5223
+    },
+    {
+      "epoch": 1.6773157810242414,
+      "grad_norm": 0.9227198362350464,
+      "learning_rate": 0.00010965406467476808,
+      "loss": 0.6049,
+      "step": 5224
+    },
+    {
+      "epoch": 1.677636859849093,
+      "grad_norm": 0.7940647602081299,
+      "learning_rate": 0.00010961932054943778,
+      "loss": 0.5055,
+      "step": 5225
+    },
+    {
+      "epoch": 1.6779579386739445,
+      "grad_norm": 0.8924132585525513,
+      "learning_rate": 0.00010958457525202241,
+      "loss": 0.6137,
+      "step": 5226
+    },
+    {
+      "epoch": 1.678279017498796,
+      "grad_norm": 0.7673733830451965,
+      "learning_rate": 0.00010954982878675563,
+      "loss": 0.5312,
+      "step": 5227
+    },
+    {
+      "epoch": 1.6786000963236476,
+      "grad_norm": 0.7759941816329956,
+      "learning_rate": 0.00010951508115787119,
+      "loss": 0.4362,
+      "step": 5228
+    },
+    {
+      "epoch": 1.678921175148499,
+      "grad_norm": 0.9239017963409424,
+      "learning_rate": 0.00010948033236960294,
+      "loss": 0.6432,
+      "step": 5229
+    },
+    {
+      "epoch": 1.6792422539733505,
+      "grad_norm": 0.9103022813796997,
+      "learning_rate": 0.00010944558242618496,
+      "loss": 0.6079,
+      "step": 5230
+    },
+    {
+      "epoch": 1.6795633327982018,
+      "grad_norm": 1.2982368469238281,
+      "learning_rate": 0.00010941083133185146,
+      "loss": 0.7558,
+      "step": 5231
+    },
+    {
+      "epoch": 1.6798844116230534,
+      "grad_norm": 0.9765358567237854,
+      "learning_rate": 0.00010937607909083667,
+      "loss": 0.5457,
+      "step": 5232
+    },
+    {
+      "epoch": 1.680205490447905,
+      "grad_norm": 0.7929934859275818,
+      "learning_rate": 0.00010934132570737507,
+      "loss": 0.5489,
+      "step": 5233
+    },
+    {
+      "epoch": 1.6805265692727565,
+      "grad_norm": 0.8707826733589172,
+      "learning_rate": 0.00010930657118570126,
+      "loss": 0.5285,
+      "step": 5234
+    },
+    {
+      "epoch": 1.680847648097608,
+      "grad_norm": 0.8279105424880981,
+      "learning_rate": 0.00010927181553005002,
+      "loss": 0.6373,
+      "step": 5235
+    },
+    {
+      "epoch": 1.6811687269224596,
+      "grad_norm": 1.1090543270111084,
+      "learning_rate": 0.00010923705874465618,
+      "loss": 0.5366,
+      "step": 5236
+    },
+    {
+      "epoch": 1.681489805747311,
+      "grad_norm": 0.6475349068641663,
+      "learning_rate": 0.00010920230083375473,
+      "loss": 0.4386,
+      "step": 5237
+    },
+    {
+      "epoch": 1.6818108845721624,
+      "grad_norm": 1.031524419784546,
+      "learning_rate": 0.00010916754180158082,
+      "loss": 0.6921,
+      "step": 5238
+    },
+    {
+      "epoch": 1.682131963397014,
+      "grad_norm": 1.1546660661697388,
+      "learning_rate": 0.00010913278165236978,
+      "loss": 0.6612,
+      "step": 5239
+    },
+    {
+      "epoch": 1.6824530422218653,
+      "grad_norm": 1.4143015146255493,
+      "learning_rate": 0.00010909802039035701,
+      "loss": 0.7917,
+      "step": 5240
+    },
+    {
+      "epoch": 1.6827741210467169,
+      "grad_norm": 1.350466012954712,
+      "learning_rate": 0.00010906325801977804,
+      "loss": 0.7994,
+      "step": 5241
+    },
+    {
+      "epoch": 1.6830951998715684,
+      "grad_norm": 0.7425410747528076,
+      "learning_rate": 0.00010902849454486856,
+      "loss": 0.5108,
+      "step": 5242
+    },
+    {
+      "epoch": 1.68341627869642,
+      "grad_norm": 1.0292056798934937,
+      "learning_rate": 0.00010899372996986439,
+      "loss": 0.5466,
+      "step": 5243
+    },
+    {
+      "epoch": 1.6837373575212715,
+      "grad_norm": 0.9850762486457825,
+      "learning_rate": 0.00010895896429900154,
+      "loss": 0.5561,
+      "step": 5244
+    },
+    {
+      "epoch": 1.684058436346123,
+      "grad_norm": 1.2769955396652222,
+      "learning_rate": 0.00010892419753651606,
+      "loss": 0.614,
+      "step": 5245
+    },
+    {
+      "epoch": 1.6843795151709746,
+      "grad_norm": 0.877348005771637,
+      "learning_rate": 0.00010888942968664417,
+      "loss": 0.5857,
+      "step": 5246
+    },
+    {
+      "epoch": 1.684700593995826,
+      "grad_norm": 1.039420485496521,
+      "learning_rate": 0.00010885466075362223,
+      "loss": 0.5347,
+      "step": 5247
+    },
+    {
+      "epoch": 1.6850216728206775,
+      "grad_norm": 1.2962031364440918,
+      "learning_rate": 0.00010881989074168673,
+      "loss": 0.86,
+      "step": 5248
+    },
+    {
+      "epoch": 1.6853427516455288,
+      "grad_norm": 1.009063720703125,
+      "learning_rate": 0.00010878511965507434,
+      "loss": 0.6275,
+      "step": 5249
+    },
+    {
+      "epoch": 1.6856638304703804,
+      "grad_norm": 0.9977805614471436,
+      "learning_rate": 0.00010875034749802173,
+      "loss": 0.6225,
+      "step": 5250
+    },
+    {
+      "epoch": 1.685984909295232,
+      "grad_norm": 0.8478687405586243,
+      "learning_rate": 0.00010871557427476583,
+      "loss": 0.5791,
+      "step": 5251
+    },
+    {
+      "epoch": 1.6863059881200835,
+      "grad_norm": 1.0061284303665161,
+      "learning_rate": 0.00010868079998954364,
+      "loss": 0.5899,
+      "step": 5252
+    },
+    {
+      "epoch": 1.686627066944935,
+      "grad_norm": 0.7541671395301819,
+      "learning_rate": 0.0001086460246465923,
+      "loss": 0.5005,
+      "step": 5253
+    },
+    {
+      "epoch": 1.6869481457697866,
+      "grad_norm": 0.8530805706977844,
+      "learning_rate": 0.00010861124825014908,
+      "loss": 0.5079,
+      "step": 5254
+    },
+    {
+      "epoch": 1.6872692245946381,
+      "grad_norm": 0.8451884984970093,
+      "learning_rate": 0.00010857647080445139,
+      "loss": 0.5617,
+      "step": 5255
+    },
+    {
+      "epoch": 1.6875903034194895,
+      "grad_norm": 1.1053026914596558,
+      "learning_rate": 0.00010854169231373676,
+      "loss": 0.6804,
+      "step": 5256
+    },
+    {
+      "epoch": 1.687911382244341,
+      "grad_norm": 0.7350571751594543,
+      "learning_rate": 0.00010850691278224281,
+      "loss": 0.5118,
+      "step": 5257
+    },
+    {
+      "epoch": 1.6882324610691923,
+      "grad_norm": 0.7682427167892456,
+      "learning_rate": 0.00010847213221420736,
+      "loss": 0.4746,
+      "step": 5258
+    },
+    {
+      "epoch": 1.6885535398940439,
+      "grad_norm": 1.612838625907898,
+      "learning_rate": 0.00010843735061386828,
+      "loss": 0.6223,
+      "step": 5259
+    },
+    {
+      "epoch": 1.6888746187188954,
+      "grad_norm": 0.8637731671333313,
+      "learning_rate": 0.00010840256798546364,
+      "loss": 0.4417,
+      "step": 5260
+    },
+    {
+      "epoch": 1.689195697543747,
+      "grad_norm": 1.2856574058532715,
+      "learning_rate": 0.00010836778433323158,
+      "loss": 0.6902,
+      "step": 5261
+    },
+    {
+      "epoch": 1.6895167763685985,
+      "grad_norm": 1.089842438697815,
+      "learning_rate": 0.00010833299966141035,
+      "loss": 0.5839,
+      "step": 5262
+    },
+    {
+      "epoch": 1.68983785519345,
+      "grad_norm": 0.7094756364822388,
+      "learning_rate": 0.0001082982139742384,
+      "loss": 0.418,
+      "step": 5263
+    },
+    {
+      "epoch": 1.6901589340183016,
+      "grad_norm": 0.7684889435768127,
+      "learning_rate": 0.00010826342727595426,
+      "loss": 0.3724,
+      "step": 5264
+    },
+    {
+      "epoch": 1.690480012843153,
+      "grad_norm": 1.0270161628723145,
+      "learning_rate": 0.00010822863957079656,
+      "loss": 1.0122,
+      "step": 5265
+    },
+    {
+      "epoch": 1.6908010916680045,
+      "grad_norm": 0.5250154137611389,
+      "learning_rate": 0.0001081938508630041,
+      "loss": 0.8635,
+      "step": 5266
+    },
+    {
+      "epoch": 1.6911221704928558,
+      "grad_norm": 0.612006664276123,
+      "learning_rate": 0.00010815906115681578,
+      "loss": 0.8014,
+      "step": 5267
+    },
+    {
+      "epoch": 1.6914432493177074,
+      "grad_norm": 0.5928871631622314,
+      "learning_rate": 0.00010812427045647058,
+      "loss": 0.4363,
+      "step": 5268
+    },
+    {
+      "epoch": 1.691764328142559,
+      "grad_norm": 0.7500852346420288,
+      "learning_rate": 0.00010808947876620767,
+      "loss": 0.4729,
+      "step": 5269
+    },
+    {
+      "epoch": 1.6920854069674105,
+      "grad_norm": 0.7043389081954956,
+      "learning_rate": 0.00010805468609026632,
+      "loss": 0.225,
+      "step": 5270
+    },
+    {
+      "epoch": 1.692406485792262,
+      "grad_norm": 1.1725765466690063,
+      "learning_rate": 0.00010801989243288589,
+      "loss": 0.2793,
+      "step": 5271
+    },
+    {
+      "epoch": 1.6927275646171136,
+      "grad_norm": 0.7991273403167725,
+      "learning_rate": 0.0001079850977983059,
+      "loss": 0.4612,
+      "step": 5272
+    },
+    {
+      "epoch": 1.6930486434419652,
+      "grad_norm": 0.8243421316146851,
+      "learning_rate": 0.00010795030219076599,
+      "loss": 0.5367,
+      "step": 5273
+    },
+    {
+      "epoch": 1.6933697222668165,
+      "grad_norm": 0.7552120685577393,
+      "learning_rate": 0.00010791550561450586,
+      "loss": 0.5766,
+      "step": 5274
+    },
+    {
+      "epoch": 1.693690801091668,
+      "grad_norm": 0.7614461779594421,
+      "learning_rate": 0.00010788070807376536,
+      "loss": 0.5462,
+      "step": 5275
+    },
+    {
+      "epoch": 1.6940118799165194,
+      "grad_norm": 1.0038031339645386,
+      "learning_rate": 0.0001078459095727845,
+      "loss": 0.6166,
+      "step": 5276
+    },
+    {
+      "epoch": 1.694332958741371,
+      "grad_norm": 0.7951825857162476,
+      "learning_rate": 0.00010781111011580336,
+      "loss": 0.5819,
+      "step": 5277
+    },
+    {
+      "epoch": 1.6946540375662225,
+      "grad_norm": 0.9689728021621704,
+      "learning_rate": 0.00010777630970706217,
+      "loss": 0.5629,
+      "step": 5278
+    },
+    {
+      "epoch": 1.694975116391074,
+      "grad_norm": 1.189405918121338,
+      "learning_rate": 0.00010774150835080119,
+      "loss": 0.4832,
+      "step": 5279
+    },
+    {
+      "epoch": 1.6952961952159256,
+      "grad_norm": 0.8709222674369812,
+      "learning_rate": 0.00010770670605126092,
+      "loss": 0.5692,
+      "step": 5280
+    },
+    {
+      "epoch": 1.6956172740407771,
+      "grad_norm": 0.9132195115089417,
+      "learning_rate": 0.00010767190281268187,
+      "loss": 0.4311,
+      "step": 5281
+    },
+    {
+      "epoch": 1.6959383528656287,
+      "grad_norm": 0.7161700129508972,
+      "learning_rate": 0.00010763709863930476,
+      "loss": 0.4397,
+      "step": 5282
+    },
+    {
+      "epoch": 1.69625943169048,
+      "grad_norm": 0.9493613839149475,
+      "learning_rate": 0.00010760229353537033,
+      "loss": 0.6239,
+      "step": 5283
+    },
+    {
+      "epoch": 1.6965805105153315,
+      "grad_norm": 1.3455016613006592,
+      "learning_rate": 0.00010756748750511953,
+      "loss": 0.7217,
+      "step": 5284
+    },
+    {
+      "epoch": 1.6969015893401829,
+      "grad_norm": 0.9821035265922546,
+      "learning_rate": 0.00010753268055279329,
+      "loss": 0.5342,
+      "step": 5285
+    },
+    {
+      "epoch": 1.6972226681650344,
+      "grad_norm": 0.8994725346565247,
+      "learning_rate": 0.00010749787268263279,
+      "loss": 0.6187,
+      "step": 5286
+    },
+    {
+      "epoch": 1.697543746989886,
+      "grad_norm": 1.6420204639434814,
+      "learning_rate": 0.00010746306389887924,
+      "loss": 0.7116,
+      "step": 5287
+    },
+    {
+      "epoch": 1.6978648258147375,
+      "grad_norm": 1.0258111953735352,
+      "learning_rate": 0.000107428254205774,
+      "loss": 0.6124,
+      "step": 5288
+    },
+    {
+      "epoch": 1.698185904639589,
+      "grad_norm": 1.5440269708633423,
+      "learning_rate": 0.00010739344360755852,
+      "loss": 0.6184,
+      "step": 5289
+    },
+    {
+      "epoch": 1.6985069834644406,
+      "grad_norm": 0.8540660738945007,
+      "learning_rate": 0.00010735863210847433,
+      "loss": 0.496,
+      "step": 5290
+    },
+    {
+      "epoch": 1.6988280622892922,
+      "grad_norm": 0.9297248721122742,
+      "learning_rate": 0.00010732381971276318,
+      "loss": 0.5762,
+      "step": 5291
+    },
+    {
+      "epoch": 1.6991491411141435,
+      "grad_norm": 0.9052714705467224,
+      "learning_rate": 0.0001072890064246668,
+      "loss": 0.5208,
+      "step": 5292
+    },
+    {
+      "epoch": 1.699470219938995,
+      "grad_norm": 0.836807370185852,
+      "learning_rate": 0.0001072541922484271,
+      "loss": 0.4136,
+      "step": 5293
+    },
+    {
+      "epoch": 1.6997912987638464,
+      "grad_norm": 1.4985709190368652,
+      "learning_rate": 0.0001072193771882861,
+      "loss": 0.9237,
+      "step": 5294
+    },
+    {
+      "epoch": 1.700112377588698,
+      "grad_norm": 1.0867552757263184,
+      "learning_rate": 0.00010718456124848583,
+      "loss": 0.5195,
+      "step": 5295
+    },
+    {
+      "epoch": 1.7004334564135495,
+      "grad_norm": 1.109985589981079,
+      "learning_rate": 0.0001071497444332686,
+      "loss": 0.7572,
+      "step": 5296
+    },
+    {
+      "epoch": 1.700754535238401,
+      "grad_norm": 1.2597237825393677,
+      "learning_rate": 0.00010711492674687671,
+      "loss": 0.603,
+      "step": 5297
+    },
+    {
+      "epoch": 1.7010756140632526,
+      "grad_norm": 1.0145076513290405,
+      "learning_rate": 0.00010708010819355256,
+      "loss": 0.5953,
+      "step": 5298
+    },
+    {
+      "epoch": 1.7013966928881041,
+      "grad_norm": 1.2275587320327759,
+      "learning_rate": 0.0001070452887775387,
+      "loss": 0.6326,
+      "step": 5299
+    },
+    {
+      "epoch": 1.7017177717129557,
+      "grad_norm": 0.6137974858283997,
+      "learning_rate": 0.00010701046850307777,
+      "loss": 0.3397,
+      "step": 5300
+    },
+    {
+      "epoch": 1.702038850537807,
+      "grad_norm": 0.7091952562332153,
+      "learning_rate": 0.00010697564737441252,
+      "loss": 0.3673,
+      "step": 5301
+    },
+    {
+      "epoch": 1.7023599293626586,
+      "grad_norm": 0.9978876113891602,
+      "learning_rate": 0.00010694082539578585,
+      "loss": 0.5951,
+      "step": 5302
+    },
+    {
+      "epoch": 1.70268100818751,
+      "grad_norm": 1.3031730651855469,
+      "learning_rate": 0.00010690600257144061,
+      "loss": 0.72,
+      "step": 5303
+    },
+    {
+      "epoch": 1.7030020870123614,
+      "grad_norm": 1.1480228900909424,
+      "learning_rate": 0.00010687117890561988,
+      "loss": 0.6365,
+      "step": 5304
+    },
+    {
+      "epoch": 1.703323165837213,
+      "grad_norm": 0.8998190760612488,
+      "learning_rate": 0.00010683635440256687,
+      "loss": 0.4303,
+      "step": 5305
+    },
+    {
+      "epoch": 1.7036442446620645,
+      "grad_norm": 0.8919349312782288,
+      "learning_rate": 0.00010680152906652483,
+      "loss": 0.5207,
+      "step": 5306
+    },
+    {
+      "epoch": 1.703965323486916,
+      "grad_norm": 1.5924605131149292,
+      "learning_rate": 0.00010676670290173709,
+      "loss": 0.96,
+      "step": 5307
+    },
+    {
+      "epoch": 1.7042864023117676,
+      "grad_norm": 1.0313186645507812,
+      "learning_rate": 0.00010673187591244714,
+      "loss": 0.5008,
+      "step": 5308
+    },
+    {
+      "epoch": 1.7046074811366192,
+      "grad_norm": 0.8978110551834106,
+      "learning_rate": 0.00010669704810289851,
+      "loss": 0.5782,
+      "step": 5309
+    },
+    {
+      "epoch": 1.7049285599614705,
+      "grad_norm": 0.9183505177497864,
+      "learning_rate": 0.00010666221947733486,
+      "loss": 0.4536,
+      "step": 5310
+    },
+    {
+      "epoch": 1.705249638786322,
+      "grad_norm": 1.1860748529434204,
+      "learning_rate": 0.00010662739004000005,
+      "loss": 0.5646,
+      "step": 5311
+    },
+    {
+      "epoch": 1.7055707176111734,
+      "grad_norm": 0.6028177738189697,
+      "learning_rate": 0.0001065925597951378,
+      "loss": 0.3459,
+      "step": 5312
+    },
+    {
+      "epoch": 1.705891796436025,
+      "grad_norm": 0.9374269843101501,
+      "learning_rate": 0.00010655772874699217,
+      "loss": 0.4143,
+      "step": 5313
+    },
+    {
+      "epoch": 1.7062128752608765,
+      "grad_norm": 0.8540414571762085,
+      "learning_rate": 0.00010652289689980714,
+      "loss": 0.4658,
+      "step": 5314
+    },
+    {
+      "epoch": 1.706533954085728,
+      "grad_norm": 0.7660170197486877,
+      "learning_rate": 0.00010648806425782695,
+      "loss": 0.687,
+      "step": 5315
+    },
+    {
+      "epoch": 1.7068550329105796,
+      "grad_norm": 0.5973314642906189,
+      "learning_rate": 0.00010645323082529581,
+      "loss": 0.7844,
+      "step": 5316
+    },
+    {
+      "epoch": 1.7071761117354312,
+      "grad_norm": 0.6430346965789795,
+      "learning_rate": 0.00010641839660645805,
+      "loss": 0.6883,
+      "step": 5317
+    },
+    {
+      "epoch": 1.7074971905602827,
+      "grad_norm": 0.7452960014343262,
+      "learning_rate": 0.00010638356160555816,
+      "loss": 0.5505,
+      "step": 5318
+    },
+    {
+      "epoch": 1.707818269385134,
+      "grad_norm": 0.7918518781661987,
+      "learning_rate": 0.00010634872582684061,
+      "loss": 0.5305,
+      "step": 5319
+    },
+    {
+      "epoch": 1.7081393482099856,
+      "grad_norm": 0.7591331005096436,
+      "learning_rate": 0.00010631388927455013,
+      "loss": 0.2967,
+      "step": 5320
+    },
+    {
+      "epoch": 1.708460427034837,
+      "grad_norm": 0.6316062211990356,
+      "learning_rate": 0.00010627905195293135,
+      "loss": 0.3833,
+      "step": 5321
+    },
+    {
+      "epoch": 1.7087815058596885,
+      "grad_norm": 0.77117919921875,
+      "learning_rate": 0.00010624421386622916,
+      "loss": 0.5453,
+      "step": 5322
+    },
+    {
+      "epoch": 1.70910258468454,
+      "grad_norm": 0.8424499034881592,
+      "learning_rate": 0.00010620937501868841,
+      "loss": 0.684,
+      "step": 5323
+    },
+    {
+      "epoch": 1.7094236635093916,
+      "grad_norm": 0.6682990789413452,
+      "learning_rate": 0.0001061745354145542,
+      "loss": 0.4461,
+      "step": 5324
+    },
+    {
+      "epoch": 1.7097447423342431,
+      "grad_norm": 0.8375151753425598,
+      "learning_rate": 0.00010613969505807156,
+      "loss": 0.773,
+      "step": 5325
+    },
+    {
+      "epoch": 1.7100658211590947,
+      "grad_norm": 0.800901472568512,
+      "learning_rate": 0.00010610485395348571,
+      "loss": 0.6591,
+      "step": 5326
+    },
+    {
+      "epoch": 1.7103868999839462,
+      "grad_norm": 0.9238450527191162,
+      "learning_rate": 0.00010607001210504191,
+      "loss": 0.5129,
+      "step": 5327
+    },
+    {
+      "epoch": 1.7107079788087975,
+      "grad_norm": 0.8702148199081421,
+      "learning_rate": 0.00010603516951698556,
+      "loss": 0.5476,
+      "step": 5328
+    },
+    {
+      "epoch": 1.711029057633649,
+      "grad_norm": 0.7507149577140808,
+      "learning_rate": 0.00010600032619356209,
+      "loss": 0.6374,
+      "step": 5329
+    },
+    {
+      "epoch": 1.7113501364585004,
+      "grad_norm": 0.9397144317626953,
+      "learning_rate": 0.00010596548213901708,
+      "loss": 0.579,
+      "step": 5330
+    },
+    {
+      "epoch": 1.711671215283352,
+      "grad_norm": 0.8215104937553406,
+      "learning_rate": 0.00010593063735759618,
+      "loss": 0.726,
+      "step": 5331
+    },
+    {
+      "epoch": 1.7119922941082035,
+      "grad_norm": 1.175023078918457,
+      "learning_rate": 0.0001058957918535451,
+      "loss": 0.6882,
+      "step": 5332
+    },
+    {
+      "epoch": 1.712313372933055,
+      "grad_norm": 0.8529446125030518,
+      "learning_rate": 0.00010586094563110964,
+      "loss": 0.6363,
+      "step": 5333
+    },
+    {
+      "epoch": 1.7126344517579066,
+      "grad_norm": 1.1046276092529297,
+      "learning_rate": 0.00010582609869453577,
+      "loss": 0.6866,
+      "step": 5334
+    },
+    {
+      "epoch": 1.7129555305827582,
+      "grad_norm": 0.8943831920623779,
+      "learning_rate": 0.00010579125104806944,
+      "loss": 0.6993,
+      "step": 5335
+    },
+    {
+      "epoch": 1.7132766094076097,
+      "grad_norm": 0.9862803220748901,
+      "learning_rate": 0.00010575640269595675,
+      "loss": 0.6898,
+      "step": 5336
+    },
+    {
+      "epoch": 1.713597688232461,
+      "grad_norm": 0.8223783373832703,
+      "learning_rate": 0.00010572155364244382,
+      "loss": 0.6043,
+      "step": 5337
+    },
+    {
+      "epoch": 1.7139187670573126,
+      "grad_norm": 0.6628785729408264,
+      "learning_rate": 0.00010568670389177696,
+      "loss": 0.3829,
+      "step": 5338
+    },
+    {
+      "epoch": 1.714239845882164,
+      "grad_norm": 1.1625560522079468,
+      "learning_rate": 0.00010565185344820247,
+      "loss": 0.6914,
+      "step": 5339
+    },
+    {
+      "epoch": 1.7145609247070155,
+      "grad_norm": 1.0547271966934204,
+      "learning_rate": 0.00010561700231596678,
+      "loss": 0.7001,
+      "step": 5340
+    },
+    {
+      "epoch": 1.714882003531867,
+      "grad_norm": 0.8884708881378174,
+      "learning_rate": 0.00010558215049931638,
+      "loss": 0.6464,
+      "step": 5341
+    },
+    {
+      "epoch": 1.7152030823567186,
+      "grad_norm": 0.7549629807472229,
+      "learning_rate": 0.00010554729800249792,
+      "loss": 0.4945,
+      "step": 5342
+    },
+    {
+      "epoch": 1.7155241611815701,
+      "grad_norm": 0.7881496548652649,
+      "learning_rate": 0.00010551244482975798,
+      "loss": 0.6036,
+      "step": 5343
+    },
+    {
+      "epoch": 1.7158452400064217,
+      "grad_norm": 0.919924259185791,
+      "learning_rate": 0.00010547759098534335,
+      "loss": 0.5523,
+      "step": 5344
+    },
+    {
+      "epoch": 1.7161663188312732,
+      "grad_norm": 1.0688406229019165,
+      "learning_rate": 0.00010544273647350092,
+      "loss": 0.4713,
+      "step": 5345
+    },
+    {
+      "epoch": 1.7164873976561246,
+      "grad_norm": 1.295059323310852,
+      "learning_rate": 0.00010540788129847756,
+      "loss": 0.6204,
+      "step": 5346
+    },
+    {
+      "epoch": 1.7168084764809761,
+      "grad_norm": 1.1029874086380005,
+      "learning_rate": 0.00010537302546452022,
+      "loss": 0.7223,
+      "step": 5347
+    },
+    {
+      "epoch": 1.7171295553058274,
+      "grad_norm": 0.7325824499130249,
+      "learning_rate": 0.00010533816897587606,
+      "loss": 0.3234,
+      "step": 5348
+    },
+    {
+      "epoch": 1.717450634130679,
+      "grad_norm": 0.9354049563407898,
+      "learning_rate": 0.00010530331183679218,
+      "loss": 0.5902,
+      "step": 5349
+    },
+    {
+      "epoch": 1.7177717129555305,
+      "grad_norm": 0.8697565793991089,
+      "learning_rate": 0.00010526845405151586,
+      "loss": 0.4462,
+      "step": 5350
+    },
+    {
+      "epoch": 1.718092791780382,
+      "grad_norm": 0.930233895778656,
+      "learning_rate": 0.0001052335956242944,
+      "loss": 0.6314,
+      "step": 5351
+    },
+    {
+      "epoch": 1.7184138706052337,
+      "grad_norm": 0.7648993730545044,
+      "learning_rate": 0.00010519873655937516,
+      "loss": 0.471,
+      "step": 5352
+    },
+    {
+      "epoch": 1.7187349494300852,
+      "grad_norm": 0.9538902640342712,
+      "learning_rate": 0.00010516387686100566,
+      "loss": 0.5847,
+      "step": 5353
+    },
+    {
+      "epoch": 1.7190560282549368,
+      "grad_norm": 0.749451220035553,
+      "learning_rate": 0.00010512901653343344,
+      "loss": 0.4562,
+      "step": 5354
+    },
+    {
+      "epoch": 1.719377107079788,
+      "grad_norm": 0.8436141610145569,
+      "learning_rate": 0.00010509415558090609,
+      "loss": 0.5758,
+      "step": 5355
+    },
+    {
+      "epoch": 1.7196981859046396,
+      "grad_norm": 1.4007991552352905,
+      "learning_rate": 0.00010505929400767134,
+      "loss": 0.6263,
+      "step": 5356
+    },
+    {
+      "epoch": 1.720019264729491,
+      "grad_norm": 0.9559081196784973,
+      "learning_rate": 0.00010502443181797697,
+      "loss": 0.5488,
+      "step": 5357
+    },
+    {
+      "epoch": 1.7203403435543425,
+      "grad_norm": 0.9755504131317139,
+      "learning_rate": 0.00010498956901607083,
+      "loss": 0.6127,
+      "step": 5358
+    },
+    {
+      "epoch": 1.720661422379194,
+      "grad_norm": 0.9046206474304199,
+      "learning_rate": 0.00010495470560620083,
+      "loss": 0.4139,
+      "step": 5359
+    },
+    {
+      "epoch": 1.7209825012040456,
+      "grad_norm": 0.9097113609313965,
+      "learning_rate": 0.00010491984159261496,
+      "loss": 0.5977,
+      "step": 5360
+    },
+    {
+      "epoch": 1.7213035800288972,
+      "grad_norm": 0.5500661730766296,
+      "learning_rate": 0.00010488497697956135,
+      "loss": 0.3321,
+      "step": 5361
+    },
+    {
+      "epoch": 1.7216246588537487,
+      "grad_norm": 0.9392353296279907,
+      "learning_rate": 0.00010485011177128807,
+      "loss": 0.4554,
+      "step": 5362
+    },
+    {
+      "epoch": 1.7219457376786,
+      "grad_norm": 0.7440015077590942,
+      "learning_rate": 0.00010481524597204342,
+      "loss": 0.419,
+      "step": 5363
+    },
+    {
+      "epoch": 1.7222668165034516,
+      "grad_norm": 0.6105871796607971,
+      "learning_rate": 0.00010478037958607568,
+      "loss": 0.4671,
+      "step": 5364
+    },
+    {
+      "epoch": 1.7225878953283031,
+      "grad_norm": 0.8352953195571899,
+      "learning_rate": 0.00010474551261763314,
+      "loss": 0.6749,
+      "step": 5365
+    },
+    {
+      "epoch": 1.7229089741531545,
+      "grad_norm": 0.5296282768249512,
+      "learning_rate": 0.00010471064507096426,
+      "loss": 0.7636,
+      "step": 5366
+    },
+    {
+      "epoch": 1.723230052978006,
+      "grad_norm": 0.717930793762207,
+      "learning_rate": 0.00010467577695031762,
+      "loss": 0.3502,
+      "step": 5367
+    },
+    {
+      "epoch": 1.7235511318028576,
+      "grad_norm": 0.6417704820632935,
+      "learning_rate": 0.00010464090825994173,
+      "loss": 0.2754,
+      "step": 5368
+    },
+    {
+      "epoch": 1.7238722106277091,
+      "grad_norm": 0.6707653999328613,
+      "learning_rate": 0.00010460603900408523,
+      "loss": 0.1641,
+      "step": 5369
+    },
+    {
+      "epoch": 1.7241932894525607,
+      "grad_norm": 2.9203500747680664,
+      "learning_rate": 0.00010457116918699688,
+      "loss": 0.4299,
+      "step": 5370
+    },
+    {
+      "epoch": 1.7245143682774122,
+      "grad_norm": 0.8975229859352112,
+      "learning_rate": 0.00010453629881292538,
+      "loss": 0.689,
+      "step": 5371
+    },
+    {
+      "epoch": 1.7248354471022636,
+      "grad_norm": 0.8682465553283691,
+      "learning_rate": 0.00010450142788611965,
+      "loss": 0.6525,
+      "step": 5372
+    },
+    {
+      "epoch": 1.725156525927115,
+      "grad_norm": 1.0204635858535767,
+      "learning_rate": 0.00010446655641082862,
+      "loss": 0.6479,
+      "step": 5373
+    },
+    {
+      "epoch": 1.7254776047519667,
+      "grad_norm": 0.8298991918563843,
+      "learning_rate": 0.00010443168439130122,
+      "loss": 0.4763,
+      "step": 5374
+    },
+    {
+      "epoch": 1.725798683576818,
+      "grad_norm": 0.9152095913887024,
+      "learning_rate": 0.0001043968118317865,
+      "loss": 0.6022,
+      "step": 5375
+    },
+    {
+      "epoch": 1.7261197624016695,
+      "grad_norm": 0.5959125757217407,
+      "learning_rate": 0.00010436193873653361,
+      "loss": 0.414,
+      "step": 5376
+    },
+    {
+      "epoch": 1.726440841226521,
+      "grad_norm": 1.1426783800125122,
+      "learning_rate": 0.00010432706510979171,
+      "loss": 0.6795,
+      "step": 5377
+    },
+    {
+      "epoch": 1.7267619200513726,
+      "grad_norm": 0.8990039825439453,
+      "learning_rate": 0.00010429219095581007,
+      "loss": 0.6957,
+      "step": 5378
+    },
+    {
+      "epoch": 1.7270829988762242,
+      "grad_norm": 0.9482988119125366,
+      "learning_rate": 0.00010425731627883797,
+      "loss": 0.575,
+      "step": 5379
+    },
+    {
+      "epoch": 1.7274040777010757,
+      "grad_norm": 0.8645418286323547,
+      "learning_rate": 0.0001042224410831248,
+      "loss": 0.5164,
+      "step": 5380
+    },
+    {
+      "epoch": 1.727725156525927,
+      "grad_norm": 0.8730982542037964,
+      "learning_rate": 0.00010418756537291996,
+      "loss": 0.5756,
+      "step": 5381
+    },
+    {
+      "epoch": 1.7280462353507786,
+      "grad_norm": 0.8636887669563293,
+      "learning_rate": 0.00010415268915247303,
+      "loss": 0.6095,
+      "step": 5382
+    },
+    {
+      "epoch": 1.7283673141756302,
+      "grad_norm": 1.0389565229415894,
+      "learning_rate": 0.00010411781242603352,
+      "loss": 0.6065,
+      "step": 5383
+    },
+    {
+      "epoch": 1.7286883930004815,
+      "grad_norm": 1.2318190336227417,
+      "learning_rate": 0.00010408293519785101,
+      "loss": 0.6049,
+      "step": 5384
+    },
+    {
+      "epoch": 1.729009471825333,
+      "grad_norm": 1.0228158235549927,
+      "learning_rate": 0.00010404805747217526,
+      "loss": 0.7854,
+      "step": 5385
+    },
+    {
+      "epoch": 1.7293305506501846,
+      "grad_norm": 0.8208790421485901,
+      "learning_rate": 0.00010401317925325598,
+      "loss": 0.5619,
+      "step": 5386
+    },
+    {
+      "epoch": 1.7296516294750361,
+      "grad_norm": 0.8615589737892151,
+      "learning_rate": 0.000103978300545343,
+      "loss": 0.5876,
+      "step": 5387
+    },
+    {
+      "epoch": 1.7299727082998877,
+      "grad_norm": 0.7015345096588135,
+      "learning_rate": 0.00010394342135268613,
+      "loss": 0.4422,
+      "step": 5388
+    },
+    {
+      "epoch": 1.7302937871247392,
+      "grad_norm": 0.7104688286781311,
+      "learning_rate": 0.00010390854167953537,
+      "loss": 0.5317,
+      "step": 5389
+    },
+    {
+      "epoch": 1.7306148659495906,
+      "grad_norm": 1.4996254444122314,
+      "learning_rate": 0.00010387366153014062,
+      "loss": 0.6955,
+      "step": 5390
+    },
+    {
+      "epoch": 1.7309359447744421,
+      "grad_norm": 1.07838773727417,
+      "learning_rate": 0.00010383878090875201,
+      "loss": 0.4943,
+      "step": 5391
+    },
+    {
+      "epoch": 1.7312570235992937,
+      "grad_norm": 1.10397207736969,
+      "learning_rate": 0.00010380389981961958,
+      "loss": 0.8043,
+      "step": 5392
+    },
+    {
+      "epoch": 1.731578102424145,
+      "grad_norm": 1.2021815776824951,
+      "learning_rate": 0.00010376901826699348,
+      "loss": 0.5963,
+      "step": 5393
+    },
+    {
+      "epoch": 1.7318991812489966,
+      "grad_norm": 0.8675143122673035,
+      "learning_rate": 0.00010373413625512394,
+      "loss": 0.5774,
+      "step": 5394
+    },
+    {
+      "epoch": 1.732220260073848,
+      "grad_norm": 1.3191437721252441,
+      "learning_rate": 0.0001036992537882612,
+      "loss": 0.7901,
+      "step": 5395
+    },
+    {
+      "epoch": 1.7325413388986997,
+      "grad_norm": 0.9630816578865051,
+      "learning_rate": 0.00010366437087065564,
+      "loss": 0.6245,
+      "step": 5396
+    },
+    {
+      "epoch": 1.7328624177235512,
+      "grad_norm": 0.8633376359939575,
+      "learning_rate": 0.00010362948750655759,
+      "loss": 0.6033,
+      "step": 5397
+    },
+    {
+      "epoch": 1.7331834965484028,
+      "grad_norm": 1.0866016149520874,
+      "learning_rate": 0.0001035946037002175,
+      "loss": 0.6458,
+      "step": 5398
+    },
+    {
+      "epoch": 1.733504575373254,
+      "grad_norm": 1.0186740159988403,
+      "learning_rate": 0.00010355971945588585,
+      "loss": 0.6645,
+      "step": 5399
+    },
+    {
+      "epoch": 1.7338256541981056,
+      "grad_norm": 1.092057228088379,
+      "learning_rate": 0.0001035248347778132,
+      "loss": 0.5861,
+      "step": 5400
+    },
+    {
+      "epoch": 1.7341467330229572,
+      "grad_norm": 1.0183131694793701,
+      "learning_rate": 0.00010348994967025012,
+      "loss": 0.518,
+      "step": 5401
+    },
+    {
+      "epoch": 1.7344678118478085,
+      "grad_norm": 0.9326373338699341,
+      "learning_rate": 0.00010345506413744726,
+      "loss": 0.5674,
+      "step": 5402
+    },
+    {
+      "epoch": 1.73478889067266,
+      "grad_norm": 0.947853684425354,
+      "learning_rate": 0.0001034201781836553,
+      "loss": 0.5808,
+      "step": 5403
+    },
+    {
+      "epoch": 1.7351099694975116,
+      "grad_norm": 0.948047935962677,
+      "learning_rate": 0.00010338529181312497,
+      "loss": 0.4687,
+      "step": 5404
+    },
+    {
+      "epoch": 1.7354310483223632,
+      "grad_norm": 1.0557464361190796,
+      "learning_rate": 0.00010335040503010716,
+      "loss": 0.5766,
+      "step": 5405
+    },
+    {
+      "epoch": 1.7357521271472147,
+      "grad_norm": 1.128117322921753,
+      "learning_rate": 0.00010331551783885263,
+      "loss": 0.5139,
+      "step": 5406
+    },
+    {
+      "epoch": 1.7360732059720663,
+      "grad_norm": 0.9319320917129517,
+      "learning_rate": 0.00010328063024361232,
+      "loss": 0.5615,
+      "step": 5407
+    },
+    {
+      "epoch": 1.7363942847969176,
+      "grad_norm": 1.243923544883728,
+      "learning_rate": 0.00010324574224863717,
+      "loss": 0.6024,
+      "step": 5408
+    },
+    {
+      "epoch": 1.7367153636217691,
+      "grad_norm": 1.1786973476409912,
+      "learning_rate": 0.00010321085385817817,
+      "loss": 0.5292,
+      "step": 5409
+    },
+    {
+      "epoch": 1.7370364424466207,
+      "grad_norm": 1.0594767332077026,
+      "learning_rate": 0.00010317596507648636,
+      "loss": 0.5848,
+      "step": 5410
+    },
+    {
+      "epoch": 1.737357521271472,
+      "grad_norm": 0.8402320742607117,
+      "learning_rate": 0.00010314107590781284,
+      "loss": 0.4564,
+      "step": 5411
+    },
+    {
+      "epoch": 1.7376786000963236,
+      "grad_norm": 1.0813262462615967,
+      "learning_rate": 0.00010310618635640876,
+      "loss": 0.465,
+      "step": 5412
+    },
+    {
+      "epoch": 1.7379996789211751,
+      "grad_norm": 0.6268415451049805,
+      "learning_rate": 0.00010307129642652528,
+      "loss": 0.4441,
+      "step": 5413
+    },
+    {
+      "epoch": 1.7383207577460267,
+      "grad_norm": 0.6250267028808594,
+      "learning_rate": 0.00010303640612241363,
+      "loss": 0.3773,
+      "step": 5414
+    },
+    {
+      "epoch": 1.7386418365708782,
+      "grad_norm": 0.6857768893241882,
+      "learning_rate": 0.00010300151544832512,
+      "loss": 0.6693,
+      "step": 5415
+    },
+    {
+      "epoch": 1.7389629153957298,
+      "grad_norm": 0.5000283122062683,
+      "learning_rate": 0.00010296662440851108,
+      "loss": 0.6734,
+      "step": 5416
+    },
+    {
+      "epoch": 1.739283994220581,
+      "grad_norm": 0.5204218626022339,
+      "learning_rate": 0.00010293173300722285,
+      "loss": 0.6129,
+      "step": 5417
+    },
+    {
+      "epoch": 1.7396050730454327,
+      "grad_norm": 0.6385371685028076,
+      "learning_rate": 0.00010289684124871182,
+      "loss": 0.6554,
+      "step": 5418
+    },
+    {
+      "epoch": 1.7399261518702842,
+      "grad_norm": 0.5886783599853516,
+      "learning_rate": 0.00010286194913722948,
+      "loss": 0.1828,
+      "step": 5419
+    },
+    {
+      "epoch": 1.7402472306951355,
+      "grad_norm": 0.9899042844772339,
+      "learning_rate": 0.00010282705667702734,
+      "loss": 0.3291,
+      "step": 5420
+    },
+    {
+      "epoch": 1.740568309519987,
+      "grad_norm": 0.702520489692688,
+      "learning_rate": 0.0001027921638723569,
+      "loss": 0.3839,
+      "step": 5421
+    },
+    {
+      "epoch": 1.7408893883448386,
+      "grad_norm": 0.5995388627052307,
+      "learning_rate": 0.00010275727072746977,
+      "loss": 0.2942,
+      "step": 5422
+    },
+    {
+      "epoch": 1.7412104671696902,
+      "grad_norm": 1.077636957168579,
+      "learning_rate": 0.00010272237724661753,
+      "loss": 0.7508,
+      "step": 5423
+    },
+    {
+      "epoch": 1.7415315459945417,
+      "grad_norm": 0.8297303915023804,
+      "learning_rate": 0.00010268748343405192,
+      "loss": 0.6525,
+      "step": 5424
+    },
+    {
+      "epoch": 1.7418526248193933,
+      "grad_norm": 0.8193151950836182,
+      "learning_rate": 0.0001026525892940246,
+      "loss": 0.6898,
+      "step": 5425
+    },
+    {
+      "epoch": 1.7421737036442446,
+      "grad_norm": 1.1228710412979126,
+      "learning_rate": 0.00010261769483078733,
+      "loss": 0.7282,
+      "step": 5426
+    },
+    {
+      "epoch": 1.7424947824690962,
+      "grad_norm": 0.6626234650611877,
+      "learning_rate": 0.00010258280004859188,
+      "loss": 0.4688,
+      "step": 5427
+    },
+    {
+      "epoch": 1.7428158612939475,
+      "grad_norm": 1.0781296491622925,
+      "learning_rate": 0.00010254790495169006,
+      "loss": 0.5331,
+      "step": 5428
+    },
+    {
+      "epoch": 1.743136940118799,
+      "grad_norm": 0.7310062646865845,
+      "learning_rate": 0.00010251300954433376,
+      "loss": 0.5715,
+      "step": 5429
+    },
+    {
+      "epoch": 1.7434580189436506,
+      "grad_norm": 0.6906360387802124,
+      "learning_rate": 0.00010247811383077487,
+      "loss": 0.474,
+      "step": 5430
+    },
+    {
+      "epoch": 1.7437790977685022,
+      "grad_norm": 0.8053506016731262,
+      "learning_rate": 0.00010244321781526533,
+      "loss": 0.5084,
+      "step": 5431
+    },
+    {
+      "epoch": 1.7441001765933537,
+      "grad_norm": 0.930061399936676,
+      "learning_rate": 0.0001024083215020571,
+      "loss": 0.6584,
+      "step": 5432
+    },
+    {
+      "epoch": 1.7444212554182053,
+      "grad_norm": 1.089030146598816,
+      "learning_rate": 0.0001023734248954022,
+      "loss": 0.6163,
+      "step": 5433
+    },
+    {
+      "epoch": 1.7447423342430568,
+      "grad_norm": 0.9254440665245056,
+      "learning_rate": 0.00010233852799955268,
+      "loss": 0.6622,
+      "step": 5434
+    },
+    {
+      "epoch": 1.7450634130679081,
+      "grad_norm": 0.9075636267662048,
+      "learning_rate": 0.00010230363081876064,
+      "loss": 0.6482,
+      "step": 5435
+    },
+    {
+      "epoch": 1.7453844918927597,
+      "grad_norm": 1.084458589553833,
+      "learning_rate": 0.00010226873335727816,
+      "loss": 0.6901,
+      "step": 5436
+    },
+    {
+      "epoch": 1.745705570717611,
+      "grad_norm": 1.6565884351730347,
+      "learning_rate": 0.00010223383561935738,
+      "loss": 0.597,
+      "step": 5437
+    },
+    {
+      "epoch": 1.7460266495424626,
+      "grad_norm": 0.997448742389679,
+      "learning_rate": 0.00010219893760925052,
+      "loss": 0.6115,
+      "step": 5438
+    },
+    {
+      "epoch": 1.7463477283673141,
+      "grad_norm": 0.9751862287521362,
+      "learning_rate": 0.00010216403933120979,
+      "loss": 0.5718,
+      "step": 5439
+    },
+    {
+      "epoch": 1.7466688071921657,
+      "grad_norm": 0.9030267596244812,
+      "learning_rate": 0.0001021291407894874,
+      "loss": 0.6273,
+      "step": 5440
+    },
+    {
+      "epoch": 1.7469898860170172,
+      "grad_norm": 1.0032966136932373,
+      "learning_rate": 0.0001020942419883357,
+      "loss": 0.6734,
+      "step": 5441
+    },
+    {
+      "epoch": 1.7473109648418688,
+      "grad_norm": 0.7883462309837341,
+      "learning_rate": 0.00010205934293200696,
+      "loss": 0.4985,
+      "step": 5442
+    },
+    {
+      "epoch": 1.7476320436667203,
+      "grad_norm": 1.4333808422088623,
+      "learning_rate": 0.00010202444362475352,
+      "loss": 0.5427,
+      "step": 5443
+    },
+    {
+      "epoch": 1.7479531224915716,
+      "grad_norm": 0.9737802147865295,
+      "learning_rate": 0.0001019895440708278,
+      "loss": 0.6445,
+      "step": 5444
+    },
+    {
+      "epoch": 1.7482742013164232,
+      "grad_norm": 0.8467430472373962,
+      "learning_rate": 0.00010195464427448213,
+      "loss": 0.5475,
+      "step": 5445
+    },
+    {
+      "epoch": 1.7485952801412745,
+      "grad_norm": 0.855104923248291,
+      "learning_rate": 0.000101919744239969,
+      "loss": 0.5532,
+      "step": 5446
+    },
+    {
+      "epoch": 1.748916358966126,
+      "grad_norm": 1.96385657787323,
+      "learning_rate": 0.00010188484397154084,
+      "loss": 0.6679,
+      "step": 5447
+    },
+    {
+      "epoch": 1.7492374377909776,
+      "grad_norm": 0.9854167699813843,
+      "learning_rate": 0.00010184994347345016,
+      "loss": 0.5125,
+      "step": 5448
+    },
+    {
+      "epoch": 1.7495585166158292,
+      "grad_norm": 1.1345828771591187,
+      "learning_rate": 0.00010181504274994948,
+      "loss": 0.577,
+      "step": 5449
+    },
+    {
+      "epoch": 1.7498795954406807,
+      "grad_norm": 0.941738486289978,
+      "learning_rate": 0.00010178014180529136,
+      "loss": 0.4594,
+      "step": 5450
+    },
+    {
+      "epoch": 1.7502006742655323,
+      "grad_norm": 0.9602155685424805,
+      "learning_rate": 0.00010174524064372837,
+      "loss": 0.7691,
+      "step": 5451
+    },
+    {
+      "epoch": 1.7505217530903838,
+      "grad_norm": 1.012502908706665,
+      "learning_rate": 0.00010171033926951305,
+      "loss": 0.6743,
+      "step": 5452
+    },
+    {
+      "epoch": 1.7508428319152352,
+      "grad_norm": 1.0123480558395386,
+      "learning_rate": 0.00010167543768689815,
+      "loss": 0.5418,
+      "step": 5453
+    },
+    {
+      "epoch": 1.7511639107400867,
+      "grad_norm": 0.9171655774116516,
+      "learning_rate": 0.00010164053590013622,
+      "loss": 0.4985,
+      "step": 5454
+    },
+    {
+      "epoch": 1.751484989564938,
+      "grad_norm": 0.9921781420707703,
+      "learning_rate": 0.00010160563391347996,
+      "loss": 0.6584,
+      "step": 5455
+    },
+    {
+      "epoch": 1.7518060683897896,
+      "grad_norm": 1.2568939924240112,
+      "learning_rate": 0.00010157073173118208,
+      "loss": 0.7265,
+      "step": 5456
+    },
+    {
+      "epoch": 1.7521271472146411,
+      "grad_norm": 1.0861806869506836,
+      "learning_rate": 0.0001015358293574953,
+      "loss": 0.6727,
+      "step": 5457
+    },
+    {
+      "epoch": 1.7524482260394927,
+      "grad_norm": 0.9609785676002502,
+      "learning_rate": 0.00010150092679667238,
+      "loss": 0.52,
+      "step": 5458
+    },
+    {
+      "epoch": 1.7527693048643442,
+      "grad_norm": 1.046840786933899,
+      "learning_rate": 0.00010146602405296607,
+      "loss": 0.6536,
+      "step": 5459
+    },
+    {
+      "epoch": 1.7530903836891958,
+      "grad_norm": 1.095132827758789,
+      "learning_rate": 0.00010143112113062918,
+      "loss": 0.467,
+      "step": 5460
+    },
+    {
+      "epoch": 1.7534114625140473,
+      "grad_norm": 0.7272664904594421,
+      "learning_rate": 0.00010139621803391455,
+      "loss": 0.4159,
+      "step": 5461
+    },
+    {
+      "epoch": 1.7537325413388987,
+      "grad_norm": 0.7782832980155945,
+      "learning_rate": 0.00010136131476707496,
+      "loss": 0.5137,
+      "step": 5462
+    },
+    {
+      "epoch": 1.7540536201637502,
+      "grad_norm": 1.0231236219406128,
+      "learning_rate": 0.0001013264113343633,
+      "loss": 0.5287,
+      "step": 5463
+    },
+    {
+      "epoch": 1.7543746989886015,
+      "grad_norm": 0.7744680643081665,
+      "learning_rate": 0.00010129150774003245,
+      "loss": 0.4127,
+      "step": 5464
+    },
+    {
+      "epoch": 1.754695777813453,
+      "grad_norm": 0.7057910561561584,
+      "learning_rate": 0.00010125660398833528,
+      "loss": 0.6325,
+      "step": 5465
+    },
+    {
+      "epoch": 1.7550168566383046,
+      "grad_norm": 0.5882181525230408,
+      "learning_rate": 0.00010122170008352472,
+      "loss": 1.0208,
+      "step": 5466
+    },
+    {
+      "epoch": 1.7553379354631562,
+      "grad_norm": 0.5369687080383301,
+      "learning_rate": 0.00010118679602985373,
+      "loss": 0.6156,
+      "step": 5467
+    },
+    {
+      "epoch": 1.7556590142880077,
+      "grad_norm": 0.8291602730751038,
+      "learning_rate": 0.00010115189183157523,
+      "loss": 0.4136,
+      "step": 5468
+    },
+    {
+      "epoch": 1.7559800931128593,
+      "grad_norm": 0.6269962787628174,
+      "learning_rate": 0.00010111698749294223,
+      "loss": 0.2929,
+      "step": 5469
+    },
+    {
+      "epoch": 1.7563011719377108,
+      "grad_norm": 0.68831467628479,
+      "learning_rate": 0.00010108208301820767,
+      "loss": 0.2129,
+      "step": 5470
+    },
+    {
+      "epoch": 1.7566222507625622,
+      "grad_norm": 0.7378066778182983,
+      "learning_rate": 0.00010104717841162458,
+      "loss": 0.1843,
+      "step": 5471
+    },
+    {
+      "epoch": 1.7569433295874137,
+      "grad_norm": 0.9951292276382446,
+      "learning_rate": 0.000101012273677446,
+      "loss": 0.5378,
+      "step": 5472
+    },
+    {
+      "epoch": 1.757264408412265,
+      "grad_norm": 0.9426655173301697,
+      "learning_rate": 0.00010097736881992492,
+      "loss": 0.6999,
+      "step": 5473
+    },
+    {
+      "epoch": 1.7575854872371166,
+      "grad_norm": 0.7657575607299805,
+      "learning_rate": 0.00010094246384331442,
+      "loss": 0.5443,
+      "step": 5474
+    },
+    {
+      "epoch": 1.7579065660619682,
+      "grad_norm": 1.0222439765930176,
+      "learning_rate": 0.00010090755875186753,
+      "loss": 0.8449,
+      "step": 5475
+    },
+    {
+      "epoch": 1.7582276448868197,
+      "grad_norm": 0.7700451016426086,
+      "learning_rate": 0.0001008726535498374,
+      "loss": 0.729,
+      "step": 5476
+    },
+    {
+      "epoch": 1.7585487237116713,
+      "grad_norm": 0.9244915246963501,
+      "learning_rate": 0.00010083774824147708,
+      "loss": 0.595,
+      "step": 5477
+    },
+    {
+      "epoch": 1.7588698025365228,
+      "grad_norm": 0.7973812222480774,
+      "learning_rate": 0.00010080284283103965,
+      "loss": 0.6733,
+      "step": 5478
+    },
+    {
+      "epoch": 1.7591908813613744,
+      "grad_norm": 0.7147784233093262,
+      "learning_rate": 0.00010076793732277829,
+      "loss": 0.5151,
+      "step": 5479
+    },
+    {
+      "epoch": 1.7595119601862257,
+      "grad_norm": 0.7087793946266174,
+      "learning_rate": 0.00010073303172094606,
+      "loss": 0.4801,
+      "step": 5480
+    },
+    {
+      "epoch": 1.7598330390110772,
+      "grad_norm": 0.9641126394271851,
+      "learning_rate": 0.00010069812602979615,
+      "loss": 0.5686,
+      "step": 5481
+    },
+    {
+      "epoch": 1.7601541178359286,
+      "grad_norm": 0.755591630935669,
+      "learning_rate": 0.00010066322025358173,
+      "loss": 0.4766,
+      "step": 5482
+    },
+    {
+      "epoch": 1.7604751966607801,
+      "grad_norm": 0.9040724635124207,
+      "learning_rate": 0.00010062831439655591,
+      "loss": 0.578,
+      "step": 5483
+    },
+    {
+      "epoch": 1.7607962754856317,
+      "grad_norm": 0.9179210066795349,
+      "learning_rate": 0.00010059340846297189,
+      "loss": 0.7127,
+      "step": 5484
+    },
+    {
+      "epoch": 1.7611173543104832,
+      "grad_norm": 0.9386594891548157,
+      "learning_rate": 0.00010055850245708283,
+      "loss": 0.6266,
+      "step": 5485
+    },
+    {
+      "epoch": 1.7614384331353348,
+      "grad_norm": 0.9094860553741455,
+      "learning_rate": 0.00010052359638314195,
+      "loss": 0.5722,
+      "step": 5486
+    },
+    {
+      "epoch": 1.7617595119601863,
+      "grad_norm": 1.1115118265151978,
+      "learning_rate": 0.00010048869024540247,
+      "loss": 0.6238,
+      "step": 5487
+    },
+    {
+      "epoch": 1.7620805907850379,
+      "grad_norm": 0.83982914686203,
+      "learning_rate": 0.00010045378404811756,
+      "loss": 0.6233,
+      "step": 5488
+    },
+    {
+      "epoch": 1.7624016696098892,
+      "grad_norm": 1.0576977729797363,
+      "learning_rate": 0.0001004188777955404,
+      "loss": 0.7304,
+      "step": 5489
+    },
+    {
+      "epoch": 1.7627227484347407,
+      "grad_norm": 0.9903758764266968,
+      "learning_rate": 0.00010038397149192426,
+      "loss": 0.5852,
+      "step": 5490
+    },
+    {
+      "epoch": 1.763043827259592,
+      "grad_norm": 1.2004951238632202,
+      "learning_rate": 0.00010034906514152238,
+      "loss": 0.6996,
+      "step": 5491
+    },
+    {
+      "epoch": 1.7633649060844436,
+      "grad_norm": 1.251308798789978,
+      "learning_rate": 0.00010031415874858797,
+      "loss": 0.7741,
+      "step": 5492
+    },
+    {
+      "epoch": 1.7636859849092952,
+      "grad_norm": 1.0517253875732422,
+      "learning_rate": 0.00010027925231737428,
+      "loss": 0.4877,
+      "step": 5493
+    },
+    {
+      "epoch": 1.7640070637341467,
+      "grad_norm": 1.2016655206680298,
+      "learning_rate": 0.00010024434585213451,
+      "loss": 0.6398,
+      "step": 5494
+    },
+    {
+      "epoch": 1.7643281425589983,
+      "grad_norm": 1.0655850172042847,
+      "learning_rate": 0.00010020943935712192,
+      "loss": 0.5722,
+      "step": 5495
+    },
+    {
+      "epoch": 1.7646492213838498,
+      "grad_norm": 1.114829182624817,
+      "learning_rate": 0.00010017453283658984,
+      "loss": 0.5166,
+      "step": 5496
+    },
+    {
+      "epoch": 1.7649703002087014,
+      "grad_norm": 0.8498656153678894,
+      "learning_rate": 0.00010013962629479146,
+      "loss": 0.5585,
+      "step": 5497
+    },
+    {
+      "epoch": 1.7652913790335527,
+      "grad_norm": 0.9619441628456116,
+      "learning_rate": 0.00010010471973598002,
+      "loss": 0.6372,
+      "step": 5498
+    },
+    {
+      "epoch": 1.7656124578584043,
+      "grad_norm": 0.7192568778991699,
+      "learning_rate": 0.00010006981316440875,
+      "loss": 0.4638,
+      "step": 5499
+    },
+    {
+      "epoch": 1.7659335366832556,
+      "grad_norm": 0.8864771127700806,
+      "learning_rate": 0.00010003490658433101,
+      "loss": 0.6032,
+      "step": 5500
+    },
+    {
+      "epoch": 1.7662546155081071,
+      "grad_norm": 1.0561635494232178,
+      "learning_rate": 0.0001,
+      "loss": 0.5758,
+      "step": 5501
+    },
+    {
+      "epoch": 1.7665756943329587,
+      "grad_norm": 0.9847275018692017,
+      "learning_rate": 9.996509341566904e-05,
+      "loss": 0.5311,
+      "step": 5502
+    },
+    {
+      "epoch": 1.7668967731578102,
+      "grad_norm": 0.7542006373405457,
+      "learning_rate": 9.993018683559126e-05,
+      "loss": 0.4759,
+      "step": 5503
+    },
+    {
+      "epoch": 1.7672178519826618,
+      "grad_norm": 1.13614821434021,
+      "learning_rate": 9.989528026402003e-05,
+      "loss": 0.4959,
+      "step": 5504
+    },
+    {
+      "epoch": 1.7675389308075133,
+      "grad_norm": 0.8526762127876282,
+      "learning_rate": 9.986037370520857e-05,
+      "loss": 0.5481,
+      "step": 5505
+    },
+    {
+      "epoch": 1.767860009632365,
+      "grad_norm": 0.771710216999054,
+      "learning_rate": 9.98254671634102e-05,
+      "loss": 0.5153,
+      "step": 5506
+    },
+    {
+      "epoch": 1.7681810884572162,
+      "grad_norm": 0.8657694458961487,
+      "learning_rate": 9.979056064287806e-05,
+      "loss": 0.5242,
+      "step": 5507
+    },
+    {
+      "epoch": 1.7685021672820678,
+      "grad_norm": 0.9876973628997803,
+      "learning_rate": 9.975565414786551e-05,
+      "loss": 0.5711,
+      "step": 5508
+    },
+    {
+      "epoch": 1.768823246106919,
+      "grad_norm": 0.7581015229225159,
+      "learning_rate": 9.972074768262576e-05,
+      "loss": 0.5151,
+      "step": 5509
+    },
+    {
+      "epoch": 1.7691443249317707,
+      "grad_norm": 0.7137047052383423,
+      "learning_rate": 9.968584125141204e-05,
+      "loss": 0.3444,
+      "step": 5510
+    },
+    {
+      "epoch": 1.7694654037566222,
+      "grad_norm": 0.4818606376647949,
+      "learning_rate": 9.965093485847767e-05,
+      "loss": 0.3025,
+      "step": 5511
+    },
+    {
+      "epoch": 1.7697864825814738,
+      "grad_norm": 0.567258894443512,
+      "learning_rate": 9.961602850807576e-05,
+      "loss": 0.3447,
+      "step": 5512
+    },
+    {
+      "epoch": 1.7701075614063253,
+      "grad_norm": 0.9824522137641907,
+      "learning_rate": 9.958112220445963e-05,
+      "loss": 0.3899,
+      "step": 5513
+    },
+    {
+      "epoch": 1.7704286402311769,
+      "grad_norm": 0.9963672161102295,
+      "learning_rate": 9.954621595188247e-05,
+      "loss": 0.4292,
+      "step": 5514
+    },
+    {
+      "epoch": 1.7707497190560284,
+      "grad_norm": 0.6725934147834778,
+      "learning_rate": 9.951130975459757e-05,
+      "loss": 0.8659,
+      "step": 5515
+    },
+    {
+      "epoch": 1.7710707978808797,
+      "grad_norm": 0.6370208263397217,
+      "learning_rate": 9.947640361685804e-05,
+      "loss": 1.0013,
+      "step": 5516
+    },
+    {
+      "epoch": 1.7713918767057313,
+      "grad_norm": 0.7263522744178772,
+      "learning_rate": 9.94414975429172e-05,
+      "loss": 0.1642,
+      "step": 5517
+    },
+    {
+      "epoch": 1.7717129555305826,
+      "grad_norm": 0.8269543647766113,
+      "learning_rate": 9.940659153702813e-05,
+      "loss": 0.3893,
+      "step": 5518
+    },
+    {
+      "epoch": 1.7720340343554342,
+      "grad_norm": 0.6357668042182922,
+      "learning_rate": 9.937168560344412e-05,
+      "loss": 0.1586,
+      "step": 5519
+    },
+    {
+      "epoch": 1.7723551131802857,
+      "grad_norm": 0.5424615144729614,
+      "learning_rate": 9.933677974641831e-05,
+      "loss": 0.2211,
+      "step": 5520
+    },
+    {
+      "epoch": 1.7726761920051373,
+      "grad_norm": 0.8307551741600037,
+      "learning_rate": 9.930187397020386e-05,
+      "loss": 0.5533,
+      "step": 5521
+    },
+    {
+      "epoch": 1.7729972708299888,
+      "grad_norm": 0.8896917104721069,
+      "learning_rate": 9.926696827905395e-05,
+      "loss": 0.5725,
+      "step": 5522
+    },
+    {
+      "epoch": 1.7733183496548404,
+      "grad_norm": 1.096503496170044,
+      "learning_rate": 9.923206267722173e-05,
+      "loss": 0.8626,
+      "step": 5523
+    },
+    {
+      "epoch": 1.773639428479692,
+      "grad_norm": 0.8159449696540833,
+      "learning_rate": 9.919715716896036e-05,
+      "loss": 0.6556,
+      "step": 5524
+    },
+    {
+      "epoch": 1.7739605073045432,
+      "grad_norm": 0.8842390775680542,
+      "learning_rate": 9.916225175852293e-05,
+      "loss": 0.6443,
+      "step": 5525
+    },
+    {
+      "epoch": 1.7742815861293948,
+      "grad_norm": 0.7416340708732605,
+      "learning_rate": 9.912734645016263e-05,
+      "loss": 0.4824,
+      "step": 5526
+    },
+    {
+      "epoch": 1.7746026649542461,
+      "grad_norm": 0.804015040397644,
+      "learning_rate": 9.909244124813247e-05,
+      "loss": 0.4857,
+      "step": 5527
+    },
+    {
+      "epoch": 1.7749237437790977,
+      "grad_norm": 0.7028852105140686,
+      "learning_rate": 9.90575361566856e-05,
+      "loss": 0.4962,
+      "step": 5528
+    },
+    {
+      "epoch": 1.7752448226039492,
+      "grad_norm": 0.8853591084480286,
+      "learning_rate": 9.902263118007513e-05,
+      "loss": 0.6105,
+      "step": 5529
+    },
+    {
+      "epoch": 1.7755659014288008,
+      "grad_norm": 0.9714551568031311,
+      "learning_rate": 9.898772632255403e-05,
+      "loss": 0.6712,
+      "step": 5530
+    },
+    {
+      "epoch": 1.7758869802536523,
+      "grad_norm": 0.8881363272666931,
+      "learning_rate": 9.895282158837545e-05,
+      "loss": 0.5934,
+      "step": 5531
+    },
+    {
+      "epoch": 1.7762080590785039,
+      "grad_norm": 1.1298437118530273,
+      "learning_rate": 9.891791698179235e-05,
+      "loss": 0.6977,
+      "step": 5532
+    },
+    {
+      "epoch": 1.7765291379033554,
+      "grad_norm": 0.7341294288635254,
+      "learning_rate": 9.888301250705779e-05,
+      "loss": 0.522,
+      "step": 5533
+    },
+    {
+      "epoch": 1.7768502167282068,
+      "grad_norm": 0.9081159234046936,
+      "learning_rate": 9.884810816842475e-05,
+      "loss": 0.6074,
+      "step": 5534
+    },
+    {
+      "epoch": 1.7771712955530583,
+      "grad_norm": 1.22751784324646,
+      "learning_rate": 9.881320397014629e-05,
+      "loss": 0.7584,
+      "step": 5535
+    },
+    {
+      "epoch": 1.7774923743779096,
+      "grad_norm": 0.7882629632949829,
+      "learning_rate": 9.877829991647528e-05,
+      "loss": 0.4237,
+      "step": 5536
+    },
+    {
+      "epoch": 1.7778134532027612,
+      "grad_norm": 0.7696181535720825,
+      "learning_rate": 9.874339601166473e-05,
+      "loss": 0.3827,
+      "step": 5537
+    },
+    {
+      "epoch": 1.7781345320276127,
+      "grad_norm": 1.0486514568328857,
+      "learning_rate": 9.87084922599676e-05,
+      "loss": 0.4434,
+      "step": 5538
+    },
+    {
+      "epoch": 1.7784556108524643,
+      "grad_norm": 0.6344658732414246,
+      "learning_rate": 9.867358866563673e-05,
+      "loss": 0.3858,
+      "step": 5539
+    },
+    {
+      "epoch": 1.7787766896773158,
+      "grad_norm": 0.9014244079589844,
+      "learning_rate": 9.863868523292509e-05,
+      "loss": 0.6374,
+      "step": 5540
+    },
+    {
+      "epoch": 1.7790977685021674,
+      "grad_norm": 0.9677129983901978,
+      "learning_rate": 9.860378196608549e-05,
+      "loss": 0.652,
+      "step": 5541
+    },
+    {
+      "epoch": 1.779418847327019,
+      "grad_norm": 1.439113736152649,
+      "learning_rate": 9.856887886937083e-05,
+      "loss": 0.7972,
+      "step": 5542
+    },
+    {
+      "epoch": 1.7797399261518703,
+      "grad_norm": 1.0146809816360474,
+      "learning_rate": 9.853397594703394e-05,
+      "loss": 0.5427,
+      "step": 5543
+    },
+    {
+      "epoch": 1.7800610049767218,
+      "grad_norm": 1.0371732711791992,
+      "learning_rate": 9.849907320332766e-05,
+      "loss": 0.526,
+      "step": 5544
+    },
+    {
+      "epoch": 1.7803820838015731,
+      "grad_norm": 0.9912870526313782,
+      "learning_rate": 9.846417064250471e-05,
+      "loss": 0.6929,
+      "step": 5545
+    },
+    {
+      "epoch": 1.7807031626264247,
+      "grad_norm": 0.9019957780838013,
+      "learning_rate": 9.842926826881796e-05,
+      "loss": 0.5505,
+      "step": 5546
+    },
+    {
+      "epoch": 1.7810242414512762,
+      "grad_norm": 0.9842864274978638,
+      "learning_rate": 9.839436608652007e-05,
+      "loss": 0.6031,
+      "step": 5547
+    },
+    {
+      "epoch": 1.7813453202761278,
+      "grad_norm": 1.212700605392456,
+      "learning_rate": 9.83594640998638e-05,
+      "loss": 0.5913,
+      "step": 5548
+    },
+    {
+      "epoch": 1.7816663991009793,
+      "grad_norm": 1.029821753501892,
+      "learning_rate": 9.832456231310189e-05,
+      "loss": 0.546,
+      "step": 5549
+    },
+    {
+      "epoch": 1.781987477925831,
+      "grad_norm": 0.77622389793396,
+      "learning_rate": 9.828966073048693e-05,
+      "loss": 0.4759,
+      "step": 5550
+    },
+    {
+      "epoch": 1.7823085567506824,
+      "grad_norm": 0.9894078373908997,
+      "learning_rate": 9.825475935627165e-05,
+      "loss": 0.6129,
+      "step": 5551
+    },
+    {
+      "epoch": 1.7826296355755338,
+      "grad_norm": 0.7435919642448425,
+      "learning_rate": 9.821985819470863e-05,
+      "loss": 0.5363,
+      "step": 5552
+    },
+    {
+      "epoch": 1.7829507144003853,
+      "grad_norm": 0.7949634790420532,
+      "learning_rate": 9.818495725005054e-05,
+      "loss": 0.492,
+      "step": 5553
+    },
+    {
+      "epoch": 1.7832717932252367,
+      "grad_norm": 0.7472631931304932,
+      "learning_rate": 9.815005652654985e-05,
+      "loss": 0.4438,
+      "step": 5554
+    },
+    {
+      "epoch": 1.7835928720500882,
+      "grad_norm": 1.0186270475387573,
+      "learning_rate": 9.81151560284592e-05,
+      "loss": 0.602,
+      "step": 5555
+    },
+    {
+      "epoch": 1.7839139508749398,
+      "grad_norm": 0.807303786277771,
+      "learning_rate": 9.808025576003104e-05,
+      "loss": 0.5087,
+      "step": 5556
+    },
+    {
+      "epoch": 1.7842350296997913,
+      "grad_norm": 0.872272253036499,
+      "learning_rate": 9.804535572551789e-05,
+      "loss": 0.5801,
+      "step": 5557
+    },
+    {
+      "epoch": 1.7845561085246429,
+      "grad_norm": 0.9321494102478027,
+      "learning_rate": 9.801045592917226e-05,
+      "loss": 0.4829,
+      "step": 5558
+    },
+    {
+      "epoch": 1.7848771873494944,
+      "grad_norm": 1.4904452562332153,
+      "learning_rate": 9.797555637524649e-05,
+      "loss": 0.5219,
+      "step": 5559
+    },
+    {
+      "epoch": 1.785198266174346,
+      "grad_norm": 1.060113787651062,
+      "learning_rate": 9.794065706799306e-05,
+      "loss": 0.641,
+      "step": 5560
+    },
+    {
+      "epoch": 1.7855193449991973,
+      "grad_norm": 0.7635378241539001,
+      "learning_rate": 9.790575801166432e-05,
+      "loss": 0.4252,
+      "step": 5561
+    },
+    {
+      "epoch": 1.7858404238240488,
+      "grad_norm": 0.670326292514801,
+      "learning_rate": 9.78708592105126e-05,
+      "loss": 0.3945,
+      "step": 5562
+    },
+    {
+      "epoch": 1.7861615026489002,
+      "grad_norm": 0.8592783212661743,
+      "learning_rate": 9.783596066879022e-05,
+      "loss": 0.3788,
+      "step": 5563
+    },
+    {
+      "epoch": 1.7864825814737517,
+      "grad_norm": 0.7472242712974548,
+      "learning_rate": 9.78010623907495e-05,
+      "loss": 0.353,
+      "step": 5564
+    },
+    {
+      "epoch": 1.7868036602986033,
+      "grad_norm": 0.5249154567718506,
+      "learning_rate": 9.776616438064264e-05,
+      "loss": 0.6485,
+      "step": 5565
+    },
+    {
+      "epoch": 1.7871247391234548,
+      "grad_norm": 0.6192828416824341,
+      "learning_rate": 9.773126664272186e-05,
+      "loss": 0.9866,
+      "step": 5566
+    },
+    {
+      "epoch": 1.7874458179483064,
+      "grad_norm": 0.5821790099143982,
+      "learning_rate": 9.76963691812394e-05,
+      "loss": 0.3801,
+      "step": 5567
+    },
+    {
+      "epoch": 1.787766896773158,
+      "grad_norm": 0.7041050791740417,
+      "learning_rate": 9.766147200044732e-05,
+      "loss": 0.2845,
+      "step": 5568
+    },
+    {
+      "epoch": 1.7880879755980095,
+      "grad_norm": 0.7531072497367859,
+      "learning_rate": 9.762657510459783e-05,
+      "loss": 0.2994,
+      "step": 5569
+    },
+    {
+      "epoch": 1.7884090544228608,
+      "grad_norm": 0.6308495998382568,
+      "learning_rate": 9.759167849794292e-05,
+      "loss": 0.1686,
+      "step": 5570
+    },
+    {
+      "epoch": 1.7887301332477124,
+      "grad_norm": 0.5816466808319092,
+      "learning_rate": 9.755678218473469e-05,
+      "loss": 0.2938,
+      "step": 5571
+    },
+    {
+      "epoch": 1.7890512120725637,
+      "grad_norm": 0.8111523389816284,
+      "learning_rate": 9.752188616922518e-05,
+      "loss": 0.7485,
+      "step": 5572
+    },
+    {
+      "epoch": 1.7893722908974152,
+      "grad_norm": 0.9488875865936279,
+      "learning_rate": 9.748699045566626e-05,
+      "loss": 0.6345,
+      "step": 5573
+    },
+    {
+      "epoch": 1.7896933697222668,
+      "grad_norm": 0.9654393196105957,
+      "learning_rate": 9.745209504830996e-05,
+      "loss": 0.669,
+      "step": 5574
+    },
+    {
+      "epoch": 1.7900144485471183,
+      "grad_norm": 0.9864271879196167,
+      "learning_rate": 9.741719995140814e-05,
+      "loss": 0.6693,
+      "step": 5575
+    },
+    {
+      "epoch": 1.7903355273719699,
+      "grad_norm": 0.7702199220657349,
+      "learning_rate": 9.73823051692127e-05,
+      "loss": 0.6806,
+      "step": 5576
+    },
+    {
+      "epoch": 1.7906566061968214,
+      "grad_norm": 0.7644592523574829,
+      "learning_rate": 9.734741070597539e-05,
+      "loss": 0.6401,
+      "step": 5577
+    },
+    {
+      "epoch": 1.790977685021673,
+      "grad_norm": 0.8069849610328674,
+      "learning_rate": 9.73125165659481e-05,
+      "loss": 0.5884,
+      "step": 5578
+    },
+    {
+      "epoch": 1.7912987638465243,
+      "grad_norm": 1.0152876377105713,
+      "learning_rate": 9.727762275338246e-05,
+      "loss": 0.5491,
+      "step": 5579
+    },
+    {
+      "epoch": 1.7916198426713759,
+      "grad_norm": 0.7692520618438721,
+      "learning_rate": 9.724272927253025e-05,
+      "loss": 0.5617,
+      "step": 5580
+    },
+    {
+      "epoch": 1.7919409214962272,
+      "grad_norm": 0.7524055242538452,
+      "learning_rate": 9.720783612764314e-05,
+      "loss": 0.535,
+      "step": 5581
+    },
+    {
+      "epoch": 1.7922620003210787,
+      "grad_norm": 0.7909640669822693,
+      "learning_rate": 9.717294332297268e-05,
+      "loss": 0.4679,
+      "step": 5582
+    },
+    {
+      "epoch": 1.7925830791459303,
+      "grad_norm": 1.540922999382019,
+      "learning_rate": 9.713805086277054e-05,
+      "loss": 0.6385,
+      "step": 5583
+    },
+    {
+      "epoch": 1.7929041579707818,
+      "grad_norm": 1.7500687837600708,
+      "learning_rate": 9.710315875128819e-05,
+      "loss": 0.6036,
+      "step": 5584
+    },
+    {
+      "epoch": 1.7932252367956334,
+      "grad_norm": 0.8995754718780518,
+      "learning_rate": 9.706826699277718e-05,
+      "loss": 0.7047,
+      "step": 5585
+    },
+    {
+      "epoch": 1.793546315620485,
+      "grad_norm": 0.9600510001182556,
+      "learning_rate": 9.703337559148892e-05,
+      "loss": 0.4654,
+      "step": 5586
+    },
+    {
+      "epoch": 1.7938673944453365,
+      "grad_norm": 3.3275082111358643,
+      "learning_rate": 9.699848455167489e-05,
+      "loss": 0.7396,
+      "step": 5587
+    },
+    {
+      "epoch": 1.7941884732701878,
+      "grad_norm": 1.0256354808807373,
+      "learning_rate": 9.696359387758636e-05,
+      "loss": 0.7023,
+      "step": 5588
+    },
+    {
+      "epoch": 1.7945095520950394,
+      "grad_norm": 0.8502017855644226,
+      "learning_rate": 9.692870357347473e-05,
+      "loss": 0.4102,
+      "step": 5589
+    },
+    {
+      "epoch": 1.7948306309198907,
+      "grad_norm": 0.9669037461280823,
+      "learning_rate": 9.689381364359129e-05,
+      "loss": 0.668,
+      "step": 5590
+    },
+    {
+      "epoch": 1.7951517097447423,
+      "grad_norm": 1.028659462928772,
+      "learning_rate": 9.685892409218717e-05,
+      "loss": 0.493,
+      "step": 5591
+    },
+    {
+      "epoch": 1.7954727885695938,
+      "grad_norm": 0.9129486680030823,
+      "learning_rate": 9.682403492351369e-05,
+      "loss": 0.6085,
+      "step": 5592
+    },
+    {
+      "epoch": 1.7957938673944454,
+      "grad_norm": 1.1581437587738037,
+      "learning_rate": 9.678914614182185e-05,
+      "loss": 0.6057,
+      "step": 5593
+    },
+    {
+      "epoch": 1.796114946219297,
+      "grad_norm": 1.030382513999939,
+      "learning_rate": 9.675425775136286e-05,
+      "loss": 0.6529,
+      "step": 5594
+    },
+    {
+      "epoch": 1.7964360250441485,
+      "grad_norm": 0.7557877898216248,
+      "learning_rate": 9.671936975638768e-05,
+      "loss": 0.417,
+      "step": 5595
+    },
+    {
+      "epoch": 1.796757103869,
+      "grad_norm": 0.8263580799102783,
+      "learning_rate": 9.668448216114739e-05,
+      "loss": 0.4406,
+      "step": 5596
+    },
+    {
+      "epoch": 1.7970781826938513,
+      "grad_norm": 1.1736435890197754,
+      "learning_rate": 9.664959496989285e-05,
+      "loss": 0.7772,
+      "step": 5597
+    },
+    {
+      "epoch": 1.7973992615187029,
+      "grad_norm": 0.8506044149398804,
+      "learning_rate": 9.661470818687503e-05,
+      "loss": 0.541,
+      "step": 5598
+    },
+    {
+      "epoch": 1.7977203403435542,
+      "grad_norm": 1.1008727550506592,
+      "learning_rate": 9.657982181634475e-05,
+      "loss": 0.6642,
+      "step": 5599
+    },
+    {
+      "epoch": 1.7980414191684058,
+      "grad_norm": 1.1743552684783936,
+      "learning_rate": 9.654493586255278e-05,
+      "loss": 0.6096,
+      "step": 5600
+    },
+    {
+      "epoch": 1.7983624979932573,
+      "grad_norm": 0.5854097604751587,
+      "learning_rate": 9.651005032974994e-05,
+      "loss": 0.3086,
+      "step": 5601
+    },
+    {
+      "epoch": 1.7986835768181089,
+      "grad_norm": 1.1236282587051392,
+      "learning_rate": 9.647516522218683e-05,
+      "loss": 0.3826,
+      "step": 5602
+    },
+    {
+      "epoch": 1.7990046556429604,
+      "grad_norm": 1.1320160627365112,
+      "learning_rate": 9.644028054411416e-05,
+      "loss": 0.6647,
+      "step": 5603
+    },
+    {
+      "epoch": 1.799325734467812,
+      "grad_norm": 0.9424121975898743,
+      "learning_rate": 9.64053962997825e-05,
+      "loss": 0.4697,
+      "step": 5604
+    },
+    {
+      "epoch": 1.7996468132926633,
+      "grad_norm": 0.765177845954895,
+      "learning_rate": 9.637051249344243e-05,
+      "loss": 0.401,
+      "step": 5605
+    },
+    {
+      "epoch": 1.7999678921175148,
+      "grad_norm": 0.8307339549064636,
+      "learning_rate": 9.633562912934436e-05,
+      "loss": 0.3938,
+      "step": 5606
+    },
+    {
+      "epoch": 1.8002889709423664,
+      "grad_norm": 1.0224002599716187,
+      "learning_rate": 9.630074621173883e-05,
+      "loss": 0.6176,
+      "step": 5607
+    },
+    {
+      "epoch": 1.8006100497672177,
+      "grad_norm": 1.1550755500793457,
+      "learning_rate": 9.62658637448761e-05,
+      "loss": 0.4616,
+      "step": 5608
+    },
+    {
+      "epoch": 1.8009311285920693,
+      "grad_norm": 0.8593536615371704,
+      "learning_rate": 9.623098173300654e-05,
+      "loss": 0.443,
+      "step": 5609
+    },
+    {
+      "epoch": 1.8012522074169208,
+      "grad_norm": 1.4475479125976562,
+      "learning_rate": 9.619610018038048e-05,
+      "loss": 0.6054,
+      "step": 5610
+    },
+    {
+      "epoch": 1.8015732862417724,
+      "grad_norm": 1.1326322555541992,
+      "learning_rate": 9.616121909124801e-05,
+      "loss": 0.3655,
+      "step": 5611
+    },
+    {
+      "epoch": 1.801894365066624,
+      "grad_norm": 0.7792235016822815,
+      "learning_rate": 9.612633846985941e-05,
+      "loss": 0.4951,
+      "step": 5612
+    },
+    {
+      "epoch": 1.8022154438914755,
+      "grad_norm": 0.576023280620575,
+      "learning_rate": 9.609145832046465e-05,
+      "loss": 0.3493,
+      "step": 5613
+    },
+    {
+      "epoch": 1.8025365227163268,
+      "grad_norm": 0.7932665944099426,
+      "learning_rate": 9.605657864731388e-05,
+      "loss": 0.3994,
+      "step": 5614
+    },
+    {
+      "epoch": 1.8028576015411784,
+      "grad_norm": 0.6708626747131348,
+      "learning_rate": 9.602169945465702e-05,
+      "loss": 0.5825,
+      "step": 5615
+    },
+    {
+      "epoch": 1.80317868036603,
+      "grad_norm": 0.534807026386261,
+      "learning_rate": 9.598682074674405e-05,
+      "loss": 0.8931,
+      "step": 5616
+    },
+    {
+      "epoch": 1.8034997591908812,
+      "grad_norm": 0.5789874196052551,
+      "learning_rate": 9.595194252782477e-05,
+      "loss": 0.3229,
+      "step": 5617
+    },
+    {
+      "epoch": 1.8038208380157328,
+      "grad_norm": 0.7923842668533325,
+      "learning_rate": 9.591706480214901e-05,
+      "loss": 0.4642,
+      "step": 5618
+    },
+    {
+      "epoch": 1.8041419168405843,
+      "grad_norm": 0.6035703420639038,
+      "learning_rate": 9.588218757396655e-05,
+      "loss": 0.36,
+      "step": 5619
+    },
+    {
+      "epoch": 1.8044629956654359,
+      "grad_norm": 0.6047983765602112,
+      "learning_rate": 9.584731084752699e-05,
+      "loss": 0.2533,
+      "step": 5620
+    },
+    {
+      "epoch": 1.8047840744902874,
+      "grad_norm": 0.8523005843162537,
+      "learning_rate": 9.581243462708006e-05,
+      "loss": 0.2647,
+      "step": 5621
+    },
+    {
+      "epoch": 1.805105153315139,
+      "grad_norm": 0.5535145998001099,
+      "learning_rate": 9.577755891687523e-05,
+      "loss": 0.137,
+      "step": 5622
+    },
+    {
+      "epoch": 1.8054262321399903,
+      "grad_norm": 0.7448154091835022,
+      "learning_rate": 9.574268372116205e-05,
+      "loss": 0.4489,
+      "step": 5623
+    },
+    {
+      "epoch": 1.8057473109648419,
+      "grad_norm": 1.0387903451919556,
+      "learning_rate": 9.570780904418993e-05,
+      "loss": 0.7846,
+      "step": 5624
+    },
+    {
+      "epoch": 1.8060683897896934,
+      "grad_norm": 0.9373582005500793,
+      "learning_rate": 9.567293489020831e-05,
+      "loss": 0.7087,
+      "step": 5625
+    },
+    {
+      "epoch": 1.8063894686145447,
+      "grad_norm": 0.9567878842353821,
+      "learning_rate": 9.563806126346642e-05,
+      "loss": 0.5429,
+      "step": 5626
+    },
+    {
+      "epoch": 1.8067105474393963,
+      "grad_norm": 1.1631771326065063,
+      "learning_rate": 9.560318816821353e-05,
+      "loss": 0.778,
+      "step": 5627
+    },
+    {
+      "epoch": 1.8070316262642478,
+      "grad_norm": 0.7982397675514221,
+      "learning_rate": 9.556831560869882e-05,
+      "loss": 0.6363,
+      "step": 5628
+    },
+    {
+      "epoch": 1.8073527050890994,
+      "grad_norm": 0.7375602126121521,
+      "learning_rate": 9.55334435891714e-05,
+      "loss": 0.5203,
+      "step": 5629
+    },
+    {
+      "epoch": 1.807673783913951,
+      "grad_norm": 0.8686415553092957,
+      "learning_rate": 9.549857211388037e-05,
+      "loss": 0.6616,
+      "step": 5630
+    },
+    {
+      "epoch": 1.8079948627388025,
+      "grad_norm": 0.9008349776268005,
+      "learning_rate": 9.546370118707463e-05,
+      "loss": 0.7094,
+      "step": 5631
+    },
+    {
+      "epoch": 1.8083159415636538,
+      "grad_norm": 0.7942934036254883,
+      "learning_rate": 9.542883081300316e-05,
+      "loss": 0.5465,
+      "step": 5632
+    },
+    {
+      "epoch": 1.8086370203885054,
+      "grad_norm": 0.939593493938446,
+      "learning_rate": 9.539396099591476e-05,
+      "loss": 0.6448,
+      "step": 5633
+    },
+    {
+      "epoch": 1.808958099213357,
+      "grad_norm": 0.8011131286621094,
+      "learning_rate": 9.53590917400583e-05,
+      "loss": 0.5164,
+      "step": 5634
+    },
+    {
+      "epoch": 1.8092791780382083,
+      "grad_norm": 0.973583459854126,
+      "learning_rate": 9.532422304968243e-05,
+      "loss": 0.6263,
+      "step": 5635
+    },
+    {
+      "epoch": 1.8096002568630598,
+      "grad_norm": 0.788236677646637,
+      "learning_rate": 9.528935492903575e-05,
+      "loss": 0.4855,
+      "step": 5636
+    },
+    {
+      "epoch": 1.8099213356879114,
+      "grad_norm": 0.9553484916687012,
+      "learning_rate": 9.525448738236691e-05,
+      "loss": 0.7004,
+      "step": 5637
+    },
+    {
+      "epoch": 1.810242414512763,
+      "grad_norm": 0.9981603622436523,
+      "learning_rate": 9.521962041392436e-05,
+      "loss": 0.64,
+      "step": 5638
+    },
+    {
+      "epoch": 1.8105634933376145,
+      "grad_norm": 0.7514563798904419,
+      "learning_rate": 9.518475402795661e-05,
+      "loss": 0.5227,
+      "step": 5639
+    },
+    {
+      "epoch": 1.810884572162466,
+      "grad_norm": 0.8019845485687256,
+      "learning_rate": 9.514988822871193e-05,
+      "loss": 0.4965,
+      "step": 5640
+    },
+    {
+      "epoch": 1.8112056509873173,
+      "grad_norm": 0.9751089811325073,
+      "learning_rate": 9.511502302043868e-05,
+      "loss": 0.4515,
+      "step": 5641
+    },
+    {
+      "epoch": 1.811526729812169,
+      "grad_norm": 1.0139288902282715,
+      "learning_rate": 9.508015840738503e-05,
+      "loss": 0.6349,
+      "step": 5642
+    },
+    {
+      "epoch": 1.8118478086370204,
+      "grad_norm": 0.7157825231552124,
+      "learning_rate": 9.504529439379921e-05,
+      "loss": 0.4664,
+      "step": 5643
+    },
+    {
+      "epoch": 1.8121688874618718,
+      "grad_norm": 0.942258358001709,
+      "learning_rate": 9.501043098392924e-05,
+      "loss": 0.5903,
+      "step": 5644
+    },
+    {
+      "epoch": 1.8124899662867233,
+      "grad_norm": 0.9379916787147522,
+      "learning_rate": 9.497556818202306e-05,
+      "loss": 0.432,
+      "step": 5645
+    },
+    {
+      "epoch": 1.8128110451115749,
+      "grad_norm": 0.9485198259353638,
+      "learning_rate": 9.494070599232868e-05,
+      "loss": 0.5559,
+      "step": 5646
+    },
+    {
+      "epoch": 1.8131321239364264,
+      "grad_norm": 1.1162331104278564,
+      "learning_rate": 9.490584441909392e-05,
+      "loss": 0.613,
+      "step": 5647
+    },
+    {
+      "epoch": 1.813453202761278,
+      "grad_norm": 1.1678434610366821,
+      "learning_rate": 9.48709834665666e-05,
+      "loss": 0.5808,
+      "step": 5648
+    },
+    {
+      "epoch": 1.8137742815861295,
+      "grad_norm": 0.8702452182769775,
+      "learning_rate": 9.483612313899435e-05,
+      "loss": 0.583,
+      "step": 5649
+    },
+    {
+      "epoch": 1.8140953604109809,
+      "grad_norm": 1.071449637413025,
+      "learning_rate": 9.480126344062487e-05,
+      "loss": 0.5968,
+      "step": 5650
+    },
+    {
+      "epoch": 1.8144164392358324,
+      "grad_norm": 1.1288082599639893,
+      "learning_rate": 9.476640437570562e-05,
+      "loss": 0.5764,
+      "step": 5651
+    },
+    {
+      "epoch": 1.814737518060684,
+      "grad_norm": 1.0096368789672852,
+      "learning_rate": 9.473154594848415e-05,
+      "loss": 0.5777,
+      "step": 5652
+    },
+    {
+      "epoch": 1.8150585968855353,
+      "grad_norm": 0.9093414545059204,
+      "learning_rate": 9.469668816320784e-05,
+      "loss": 0.5467,
+      "step": 5653
+    },
+    {
+      "epoch": 1.8153796757103868,
+      "grad_norm": 0.9198897480964661,
+      "learning_rate": 9.466183102412395e-05,
+      "loss": 0.5423,
+      "step": 5654
+    },
+    {
+      "epoch": 1.8157007545352384,
+      "grad_norm": 0.9140397310256958,
+      "learning_rate": 9.462697453547979e-05,
+      "loss": 0.5262,
+      "step": 5655
+    },
+    {
+      "epoch": 1.81602183336009,
+      "grad_norm": 1.1737873554229736,
+      "learning_rate": 9.459211870152245e-05,
+      "loss": 0.572,
+      "step": 5656
+    },
+    {
+      "epoch": 1.8163429121849415,
+      "grad_norm": 0.6453090906143188,
+      "learning_rate": 9.455726352649911e-05,
+      "loss": 0.3562,
+      "step": 5657
+    },
+    {
+      "epoch": 1.816663991009793,
+      "grad_norm": 0.9809330105781555,
+      "learning_rate": 9.452240901465663e-05,
+      "loss": 0.4382,
+      "step": 5658
+    },
+    {
+      "epoch": 1.8169850698346444,
+      "grad_norm": 1.092984914779663,
+      "learning_rate": 9.448755517024206e-05,
+      "loss": 0.5466,
+      "step": 5659
+    },
+    {
+      "epoch": 1.817306148659496,
+      "grad_norm": 0.7490816712379456,
+      "learning_rate": 9.445270199750212e-05,
+      "loss": 0.4304,
+      "step": 5660
+    },
+    {
+      "epoch": 1.8176272274843472,
+      "grad_norm": 0.6709809303283691,
+      "learning_rate": 9.441784950068362e-05,
+      "loss": 0.3968,
+      "step": 5661
+    },
+    {
+      "epoch": 1.8179483063091988,
+      "grad_norm": 0.8209671378135681,
+      "learning_rate": 9.438299768403327e-05,
+      "loss": 0.4619,
+      "step": 5662
+    },
+    {
+      "epoch": 1.8182693851340503,
+      "grad_norm": 0.5862700939178467,
+      "learning_rate": 9.434814655179755e-05,
+      "loss": 0.3755,
+      "step": 5663
+    },
+    {
+      "epoch": 1.818590463958902,
+      "grad_norm": 0.5392091870307922,
+      "learning_rate": 9.43132961082231e-05,
+      "loss": 0.3564,
+      "step": 5664
+    },
+    {
+      "epoch": 1.8189115427837534,
+      "grad_norm": 0.6820465326309204,
+      "learning_rate": 9.427844635755619e-05,
+      "loss": 0.6728,
+      "step": 5665
+    },
+    {
+      "epoch": 1.819232621608605,
+      "grad_norm": 0.4257480800151825,
+      "learning_rate": 9.424359730404329e-05,
+      "loss": 0.4742,
+      "step": 5666
+    },
+    {
+      "epoch": 1.8195537004334565,
+      "grad_norm": 0.7783001065254211,
+      "learning_rate": 9.420874895193056e-05,
+      "loss": 0.3928,
+      "step": 5667
+    },
+    {
+      "epoch": 1.8198747792583079,
+      "grad_norm": 0.7277984619140625,
+      "learning_rate": 9.417390130546426e-05,
+      "loss": 0.4529,
+      "step": 5668
+    },
+    {
+      "epoch": 1.8201958580831594,
+      "grad_norm": 0.4899481236934662,
+      "learning_rate": 9.413905436889035e-05,
+      "loss": 0.2177,
+      "step": 5669
+    },
+    {
+      "epoch": 1.8205169369080108,
+      "grad_norm": 0.7720094323158264,
+      "learning_rate": 9.410420814645493e-05,
+      "loss": 0.2418,
+      "step": 5670
+    },
+    {
+      "epoch": 1.8208380157328623,
+      "grad_norm": 0.7921152114868164,
+      "learning_rate": 9.406936264240386e-05,
+      "loss": 0.3818,
+      "step": 5671
+    },
+    {
+      "epoch": 1.8211590945577139,
+      "grad_norm": 0.8348581194877625,
+      "learning_rate": 9.403451786098294e-05,
+      "loss": 0.5024,
+      "step": 5672
+    },
+    {
+      "epoch": 1.8214801733825654,
+      "grad_norm": 0.9092313051223755,
+      "learning_rate": 9.399967380643796e-05,
+      "loss": 0.6206,
+      "step": 5673
+    },
+    {
+      "epoch": 1.821801252207417,
+      "grad_norm": 0.9595973491668701,
+      "learning_rate": 9.396483048301448e-05,
+      "loss": 0.5598,
+      "step": 5674
+    },
+    {
+      "epoch": 1.8221223310322685,
+      "grad_norm": 0.7221679091453552,
+      "learning_rate": 9.392998789495811e-05,
+      "loss": 0.463,
+      "step": 5675
+    },
+    {
+      "epoch": 1.82244340985712,
+      "grad_norm": 0.6640863418579102,
+      "learning_rate": 9.38951460465143e-05,
+      "loss": 0.4028,
+      "step": 5676
+    },
+    {
+      "epoch": 1.8227644886819714,
+      "grad_norm": 0.8685281872749329,
+      "learning_rate": 9.386030494192846e-05,
+      "loss": 0.4908,
+      "step": 5677
+    },
+    {
+      "epoch": 1.823085567506823,
+      "grad_norm": 0.8399313688278198,
+      "learning_rate": 9.382546458544582e-05,
+      "loss": 0.6339,
+      "step": 5678
+    },
+    {
+      "epoch": 1.8234066463316743,
+      "grad_norm": 0.8233359456062317,
+      "learning_rate": 9.37906249813116e-05,
+      "loss": 0.5707,
+      "step": 5679
+    },
+    {
+      "epoch": 1.8237277251565258,
+      "grad_norm": 0.9246627688407898,
+      "learning_rate": 9.375578613377089e-05,
+      "loss": 0.5633,
+      "step": 5680
+    },
+    {
+      "epoch": 1.8240488039813774,
+      "grad_norm": 0.9575353860855103,
+      "learning_rate": 9.372094804706867e-05,
+      "loss": 0.622,
+      "step": 5681
+    },
+    {
+      "epoch": 1.824369882806229,
+      "grad_norm": 0.6907000541687012,
+      "learning_rate": 9.368611072544992e-05,
+      "loss": 0.4413,
+      "step": 5682
+    },
+    {
+      "epoch": 1.8246909616310805,
+      "grad_norm": 0.8728267550468445,
+      "learning_rate": 9.36512741731594e-05,
+      "loss": 0.4828,
+      "step": 5683
+    },
+    {
+      "epoch": 1.825012040455932,
+      "grad_norm": 0.8942257165908813,
+      "learning_rate": 9.361643839444188e-05,
+      "loss": 0.6054,
+      "step": 5684
+    },
+    {
+      "epoch": 1.8253331192807836,
+      "grad_norm": 1.4165688753128052,
+      "learning_rate": 9.358160339354194e-05,
+      "loss": 0.6781,
+      "step": 5685
+    },
+    {
+      "epoch": 1.825654198105635,
+      "grad_norm": 1.0946552753448486,
+      "learning_rate": 9.354676917470422e-05,
+      "loss": 0.6444,
+      "step": 5686
+    },
+    {
+      "epoch": 1.8259752769304864,
+      "grad_norm": 0.9636126756668091,
+      "learning_rate": 9.351193574217306e-05,
+      "loss": 0.5896,
+      "step": 5687
+    },
+    {
+      "epoch": 1.8262963557553378,
+      "grad_norm": 0.9439002871513367,
+      "learning_rate": 9.347710310019288e-05,
+      "loss": 0.5716,
+      "step": 5688
+    },
+    {
+      "epoch": 1.8266174345801893,
+      "grad_norm": 1.3570852279663086,
+      "learning_rate": 9.344227125300788e-05,
+      "loss": 0.5469,
+      "step": 5689
+    },
+    {
+      "epoch": 1.8269385134050409,
+      "grad_norm": 1.0195239782333374,
+      "learning_rate": 9.340744020486222e-05,
+      "loss": 0.6181,
+      "step": 5690
+    },
+    {
+      "epoch": 1.8272595922298924,
+      "grad_norm": 0.9529807567596436,
+      "learning_rate": 9.337260996000002e-05,
+      "loss": 0.5538,
+      "step": 5691
+    },
+    {
+      "epoch": 1.827580671054744,
+      "grad_norm": 1.627726674079895,
+      "learning_rate": 9.333778052266513e-05,
+      "loss": 0.6368,
+      "step": 5692
+    },
+    {
+      "epoch": 1.8279017498795955,
+      "grad_norm": 1.326042652130127,
+      "learning_rate": 9.330295189710152e-05,
+      "loss": 0.6902,
+      "step": 5693
+    },
+    {
+      "epoch": 1.828222828704447,
+      "grad_norm": 0.8660145998001099,
+      "learning_rate": 9.32681240875529e-05,
+      "loss": 0.5452,
+      "step": 5694
+    },
+    {
+      "epoch": 1.8285439075292984,
+      "grad_norm": 0.9196065664291382,
+      "learning_rate": 9.323329709826294e-05,
+      "loss": 0.5653,
+      "step": 5695
+    },
+    {
+      "epoch": 1.82886498635415,
+      "grad_norm": 0.7746536135673523,
+      "learning_rate": 9.319847093347522e-05,
+      "loss": 0.4199,
+      "step": 5696
+    },
+    {
+      "epoch": 1.8291860651790013,
+      "grad_norm": 0.7608351707458496,
+      "learning_rate": 9.316364559743314e-05,
+      "loss": 0.4165,
+      "step": 5697
+    },
+    {
+      "epoch": 1.8295071440038528,
+      "grad_norm": 0.7243719696998596,
+      "learning_rate": 9.312882109438013e-05,
+      "loss": 0.4451,
+      "step": 5698
+    },
+    {
+      "epoch": 1.8298282228287044,
+      "grad_norm": 1.2507250308990479,
+      "learning_rate": 9.309399742855942e-05,
+      "loss": 0.7069,
+      "step": 5699
+    },
+    {
+      "epoch": 1.830149301653556,
+      "grad_norm": 0.8393792510032654,
+      "learning_rate": 9.30591746042142e-05,
+      "loss": 0.4762,
+      "step": 5700
+    },
+    {
+      "epoch": 1.8304703804784075,
+      "grad_norm": 0.8377349376678467,
+      "learning_rate": 9.302435262558747e-05,
+      "loss": 0.4423,
+      "step": 5701
+    },
+    {
+      "epoch": 1.830791459303259,
+      "grad_norm": 0.9344905614852905,
+      "learning_rate": 9.298953149692225e-05,
+      "loss": 0.5336,
+      "step": 5702
+    },
+    {
+      "epoch": 1.8311125381281106,
+      "grad_norm": 0.8581468462944031,
+      "learning_rate": 9.295471122246131e-05,
+      "loss": 0.5264,
+      "step": 5703
+    },
+    {
+      "epoch": 1.831433616952962,
+      "grad_norm": 0.9583359956741333,
+      "learning_rate": 9.291989180644747e-05,
+      "loss": 0.5861,
+      "step": 5704
+    },
+    {
+      "epoch": 1.8317546957778135,
+      "grad_norm": 1.2595868110656738,
+      "learning_rate": 9.288507325312335e-05,
+      "loss": 0.5907,
+      "step": 5705
+    },
+    {
+      "epoch": 1.8320757746026648,
+      "grad_norm": 0.8175548315048218,
+      "learning_rate": 9.285025556673141e-05,
+      "loss": 0.5518,
+      "step": 5706
+    },
+    {
+      "epoch": 1.8323968534275163,
+      "grad_norm": 1.2646697759628296,
+      "learning_rate": 9.281543875151419e-05,
+      "loss": 0.5245,
+      "step": 5707
+    },
+    {
+      "epoch": 1.832717932252368,
+      "grad_norm": 0.8525310754776001,
+      "learning_rate": 9.278062281171393e-05,
+      "loss": 0.5834,
+      "step": 5708
+    },
+    {
+      "epoch": 1.8330390110772194,
+      "grad_norm": 1.1546502113342285,
+      "learning_rate": 9.274580775157294e-05,
+      "loss": 0.6688,
+      "step": 5709
+    },
+    {
+      "epoch": 1.833360089902071,
+      "grad_norm": 0.5779723525047302,
+      "learning_rate": 9.271099357533321e-05,
+      "loss": 0.3218,
+      "step": 5710
+    },
+    {
+      "epoch": 1.8336811687269226,
+      "grad_norm": 0.8706488013267517,
+      "learning_rate": 9.267618028723686e-05,
+      "loss": 0.4382,
+      "step": 5711
+    },
+    {
+      "epoch": 1.834002247551774,
+      "grad_norm": 0.9869682788848877,
+      "learning_rate": 9.264136789152567e-05,
+      "loss": 0.3801,
+      "step": 5712
+    },
+    {
+      "epoch": 1.8343233263766254,
+      "grad_norm": 0.890150785446167,
+      "learning_rate": 9.26065563924415e-05,
+      "loss": 0.5917,
+      "step": 5713
+    },
+    {
+      "epoch": 1.834644405201477,
+      "grad_norm": 0.7428485155105591,
+      "learning_rate": 9.257174579422605e-05,
+      "loss": 0.4025,
+      "step": 5714
+    },
+    {
+      "epoch": 1.8349654840263283,
+      "grad_norm": 0.6964389085769653,
+      "learning_rate": 9.253693610112078e-05,
+      "loss": 0.5247,
+      "step": 5715
+    },
+    {
+      "epoch": 1.8352865628511799,
+      "grad_norm": 0.4811036288738251,
+      "learning_rate": 9.250212731736726e-05,
+      "loss": 0.7248,
+      "step": 5716
+    },
+    {
+      "epoch": 1.8356076416760314,
+      "grad_norm": 0.4891749322414398,
+      "learning_rate": 9.246731944720675e-05,
+      "loss": 0.5256,
+      "step": 5717
+    },
+    {
+      "epoch": 1.835928720500883,
+      "grad_norm": 0.6794456839561462,
+      "learning_rate": 9.243251249488052e-05,
+      "loss": 0.3973,
+      "step": 5718
+    },
+    {
+      "epoch": 1.8362497993257345,
+      "grad_norm": 0.5766389966011047,
+      "learning_rate": 9.239770646462968e-05,
+      "loss": 0.2345,
+      "step": 5719
+    },
+    {
+      "epoch": 1.836570878150586,
+      "grad_norm": 0.8062403202056885,
+      "learning_rate": 9.236290136069528e-05,
+      "loss": 0.5047,
+      "step": 5720
+    },
+    {
+      "epoch": 1.8368919569754376,
+      "grad_norm": 0.6303856372833252,
+      "learning_rate": 9.232809718731814e-05,
+      "loss": 0.2518,
+      "step": 5721
+    },
+    {
+      "epoch": 1.837213035800289,
+      "grad_norm": 1.0692548751831055,
+      "learning_rate": 9.229329394873911e-05,
+      "loss": 0.8745,
+      "step": 5722
+    },
+    {
+      "epoch": 1.8375341146251405,
+      "grad_norm": 0.8411319255828857,
+      "learning_rate": 9.225849164919885e-05,
+      "loss": 0.5388,
+      "step": 5723
+    },
+    {
+      "epoch": 1.8378551934499918,
+      "grad_norm": 0.8031153678894043,
+      "learning_rate": 9.222369029293787e-05,
+      "loss": 0.6806,
+      "step": 5724
+    },
+    {
+      "epoch": 1.8381762722748434,
+      "grad_norm": 0.6173913478851318,
+      "learning_rate": 9.218888988419668e-05,
+      "loss": 0.4643,
+      "step": 5725
+    },
+    {
+      "epoch": 1.838497351099695,
+      "grad_norm": 0.7925007939338684,
+      "learning_rate": 9.215409042721552e-05,
+      "loss": 0.5676,
+      "step": 5726
+    },
+    {
+      "epoch": 1.8388184299245465,
+      "grad_norm": 0.9955210089683533,
+      "learning_rate": 9.211929192623467e-05,
+      "loss": 0.7278,
+      "step": 5727
+    },
+    {
+      "epoch": 1.839139508749398,
+      "grad_norm": 0.7641330361366272,
+      "learning_rate": 9.208449438549415e-05,
+      "loss": 0.4462,
+      "step": 5728
+    },
+    {
+      "epoch": 1.8394605875742496,
+      "grad_norm": 0.7429258227348328,
+      "learning_rate": 9.204969780923403e-05,
+      "loss": 0.6244,
+      "step": 5729
+    },
+    {
+      "epoch": 1.8397816663991011,
+      "grad_norm": 0.8412753939628601,
+      "learning_rate": 9.201490220169408e-05,
+      "loss": 0.7808,
+      "step": 5730
+    },
+    {
+      "epoch": 1.8401027452239525,
+      "grad_norm": 0.8708707094192505,
+      "learning_rate": 9.198010756711412e-05,
+      "loss": 0.6767,
+      "step": 5731
+    },
+    {
+      "epoch": 1.840423824048804,
+      "grad_norm": 0.7798470258712769,
+      "learning_rate": 9.194531390973371e-05,
+      "loss": 0.5248,
+      "step": 5732
+    },
+    {
+      "epoch": 1.8407449028736553,
+      "grad_norm": 0.8160192966461182,
+      "learning_rate": 9.191052123379234e-05,
+      "loss": 0.5404,
+      "step": 5733
+    },
+    {
+      "epoch": 1.8410659816985069,
+      "grad_norm": 1.1252083778381348,
+      "learning_rate": 9.187572954352947e-05,
+      "loss": 0.7795,
+      "step": 5734
+    },
+    {
+      "epoch": 1.8413870605233584,
+      "grad_norm": 0.8956120610237122,
+      "learning_rate": 9.184093884318425e-05,
+      "loss": 0.6233,
+      "step": 5735
+    },
+    {
+      "epoch": 1.84170813934821,
+      "grad_norm": 0.9571072459220886,
+      "learning_rate": 9.180614913699592e-05,
+      "loss": 0.6217,
+      "step": 5736
+    },
+    {
+      "epoch": 1.8420292181730615,
+      "grad_norm": 0.7272530794143677,
+      "learning_rate": 9.177136042920344e-05,
+      "loss": 0.4461,
+      "step": 5737
+    },
+    {
+      "epoch": 1.842350296997913,
+      "grad_norm": 1.0026044845581055,
+      "learning_rate": 9.173657272404576e-05,
+      "loss": 0.7131,
+      "step": 5738
+    },
+    {
+      "epoch": 1.8426713758227646,
+      "grad_norm": 0.8981397747993469,
+      "learning_rate": 9.17017860257616e-05,
+      "loss": 0.5629,
+      "step": 5739
+    },
+    {
+      "epoch": 1.842992454647616,
+      "grad_norm": 1.2423765659332275,
+      "learning_rate": 9.166700033858969e-05,
+      "loss": 0.7977,
+      "step": 5740
+    },
+    {
+      "epoch": 1.8433135334724675,
+      "grad_norm": 0.7401012778282166,
+      "learning_rate": 9.163221566676847e-05,
+      "loss": 0.4719,
+      "step": 5741
+    },
+    {
+      "epoch": 1.8436346122973188,
+      "grad_norm": 0.9083794951438904,
+      "learning_rate": 9.159743201453638e-05,
+      "loss": 0.5664,
+      "step": 5742
+    },
+    {
+      "epoch": 1.8439556911221704,
+      "grad_norm": 0.7999395728111267,
+      "learning_rate": 9.156264938613174e-05,
+      "loss": 0.5195,
+      "step": 5743
+    },
+    {
+      "epoch": 1.844276769947022,
+      "grad_norm": 0.8803184628486633,
+      "learning_rate": 9.152786778579267e-05,
+      "loss": 0.4888,
+      "step": 5744
+    },
+    {
+      "epoch": 1.8445978487718735,
+      "grad_norm": 0.923516571521759,
+      "learning_rate": 9.14930872177572e-05,
+      "loss": 0.4521,
+      "step": 5745
+    },
+    {
+      "epoch": 1.844918927596725,
+      "grad_norm": 0.7799448370933533,
+      "learning_rate": 9.145830768626327e-05,
+      "loss": 0.5848,
+      "step": 5746
+    },
+    {
+      "epoch": 1.8452400064215766,
+      "grad_norm": 0.9645082354545593,
+      "learning_rate": 9.142352919554862e-05,
+      "loss": 0.5881,
+      "step": 5747
+    },
+    {
+      "epoch": 1.8455610852464281,
+      "grad_norm": 0.8418881297111511,
+      "learning_rate": 9.138875174985091e-05,
+      "loss": 0.4551,
+      "step": 5748
+    },
+    {
+      "epoch": 1.8458821640712795,
+      "grad_norm": 0.8868610858917236,
+      "learning_rate": 9.135397535340773e-05,
+      "loss": 0.3873,
+      "step": 5749
+    },
+    {
+      "epoch": 1.846203242896131,
+      "grad_norm": 1.052437424659729,
+      "learning_rate": 9.131920001045638e-05,
+      "loss": 0.5356,
+      "step": 5750
+    },
+    {
+      "epoch": 1.8465243217209824,
+      "grad_norm": 0.9833017587661743,
+      "learning_rate": 9.128442572523417e-05,
+      "loss": 0.5061,
+      "step": 5751
+    },
+    {
+      "epoch": 1.846845400545834,
+      "grad_norm": 1.3260668516159058,
+      "learning_rate": 9.12496525019783e-05,
+      "loss": 0.5013,
+      "step": 5752
+    },
+    {
+      "epoch": 1.8471664793706855,
+      "grad_norm": 1.056301474571228,
+      "learning_rate": 9.121488034492569e-05,
+      "loss": 0.6949,
+      "step": 5753
+    },
+    {
+      "epoch": 1.847487558195537,
+      "grad_norm": 0.8964768648147583,
+      "learning_rate": 9.11801092583133e-05,
+      "loss": 0.5426,
+      "step": 5754
+    },
+    {
+      "epoch": 1.8478086370203886,
+      "grad_norm": 0.8216196894645691,
+      "learning_rate": 9.114533924637778e-05,
+      "loss": 0.4087,
+      "step": 5755
+    },
+    {
+      "epoch": 1.84812971584524,
+      "grad_norm": 0.7172057032585144,
+      "learning_rate": 9.111057031335585e-05,
+      "loss": 0.417,
+      "step": 5756
+    },
+    {
+      "epoch": 1.8484507946700917,
+      "grad_norm": 1.1012330055236816,
+      "learning_rate": 9.107580246348395e-05,
+      "loss": 0.3882,
+      "step": 5757
+    },
+    {
+      "epoch": 1.848771873494943,
+      "grad_norm": 0.856521725654602,
+      "learning_rate": 9.104103570099848e-05,
+      "loss": 0.4772,
+      "step": 5758
+    },
+    {
+      "epoch": 1.8490929523197945,
+      "grad_norm": 1.0229532718658447,
+      "learning_rate": 9.100627003013562e-05,
+      "loss": 0.4683,
+      "step": 5759
+    },
+    {
+      "epoch": 1.8494140311446459,
+      "grad_norm": 1.44857919216156,
+      "learning_rate": 9.097150545513145e-05,
+      "loss": 0.7295,
+      "step": 5760
+    },
+    {
+      "epoch": 1.8497351099694974,
+      "grad_norm": 0.8216368556022644,
+      "learning_rate": 9.093674198022201e-05,
+      "loss": 0.4156,
+      "step": 5761
+    },
+    {
+      "epoch": 1.850056188794349,
+      "grad_norm": 0.5136415362358093,
+      "learning_rate": 9.090197960964301e-05,
+      "loss": 0.3143,
+      "step": 5762
+    },
+    {
+      "epoch": 1.8503772676192005,
+      "grad_norm": 0.7286933660507202,
+      "learning_rate": 9.086721834763024e-05,
+      "loss": 0.3688,
+      "step": 5763
+    },
+    {
+      "epoch": 1.850698346444052,
+      "grad_norm": 0.9348329305648804,
+      "learning_rate": 9.083245819841918e-05,
+      "loss": 0.5614,
+      "step": 5764
+    },
+    {
+      "epoch": 1.8510194252689036,
+      "grad_norm": 1.0801033973693848,
+      "learning_rate": 9.07976991662453e-05,
+      "loss": 0.7294,
+      "step": 5765
+    },
+    {
+      "epoch": 1.8513405040937552,
+      "grad_norm": 0.467505544424057,
+      "learning_rate": 9.076294125534383e-05,
+      "loss": 0.9349,
+      "step": 5766
+    },
+    {
+      "epoch": 1.8516615829186065,
+      "grad_norm": 0.6000930070877075,
+      "learning_rate": 9.072818446994999e-05,
+      "loss": 0.5034,
+      "step": 5767
+    },
+    {
+      "epoch": 1.851982661743458,
+      "grad_norm": 0.7102051973342896,
+      "learning_rate": 9.069342881429876e-05,
+      "loss": 0.2786,
+      "step": 5768
+    },
+    {
+      "epoch": 1.8523037405683094,
+      "grad_norm": 0.6671398878097534,
+      "learning_rate": 9.065867429262496e-05,
+      "loss": 0.1777,
+      "step": 5769
+    },
+    {
+      "epoch": 1.852624819393161,
+      "grad_norm": 0.6379199624061584,
+      "learning_rate": 9.062392090916337e-05,
+      "loss": 0.3455,
+      "step": 5770
+    },
+    {
+      "epoch": 1.8529458982180125,
+      "grad_norm": 0.7193529009819031,
+      "learning_rate": 9.058916866814858e-05,
+      "loss": 0.2457,
+      "step": 5771
+    },
+    {
+      "epoch": 1.853266977042864,
+      "grad_norm": 1.0792187452316284,
+      "learning_rate": 9.055441757381506e-05,
+      "loss": 0.7677,
+      "step": 5772
+    },
+    {
+      "epoch": 1.8535880558677156,
+      "grad_norm": 0.8163480758666992,
+      "learning_rate": 9.051966763039707e-05,
+      "loss": 0.5803,
+      "step": 5773
+    },
+    {
+      "epoch": 1.8539091346925671,
+      "grad_norm": 0.7993916869163513,
+      "learning_rate": 9.048491884212884e-05,
+      "loss": 0.5471,
+      "step": 5774
+    },
+    {
+      "epoch": 1.8542302135174187,
+      "grad_norm": 0.9378834962844849,
+      "learning_rate": 9.045017121324438e-05,
+      "loss": 0.5756,
+      "step": 5775
+    },
+    {
+      "epoch": 1.85455129234227,
+      "grad_norm": 0.7326672077178955,
+      "learning_rate": 9.04154247479776e-05,
+      "loss": 0.3347,
+      "step": 5776
+    },
+    {
+      "epoch": 1.8548723711671216,
+      "grad_norm": 0.8225162625312805,
+      "learning_rate": 9.038067945056227e-05,
+      "loss": 0.6454,
+      "step": 5777
+    },
+    {
+      "epoch": 1.8551934499919729,
+      "grad_norm": 1.0159610509872437,
+      "learning_rate": 9.034593532523193e-05,
+      "loss": 0.5996,
+      "step": 5778
+    },
+    {
+      "epoch": 1.8555145288168244,
+      "grad_norm": 0.8151166439056396,
+      "learning_rate": 9.03111923762201e-05,
+      "loss": 0.6339,
+      "step": 5779
+    },
+    {
+      "epoch": 1.855835607641676,
+      "grad_norm": 0.956063449382782,
+      "learning_rate": 9.027645060776006e-05,
+      "loss": 0.5437,
+      "step": 5780
+    },
+    {
+      "epoch": 1.8561566864665275,
+      "grad_norm": 0.8743849396705627,
+      "learning_rate": 9.024171002408506e-05,
+      "loss": 0.5587,
+      "step": 5781
+    },
+    {
+      "epoch": 1.856477765291379,
+      "grad_norm": 0.8803306818008423,
+      "learning_rate": 9.020697062942807e-05,
+      "loss": 0.5749,
+      "step": 5782
+    },
+    {
+      "epoch": 1.8567988441162306,
+      "grad_norm": 1.1139497756958008,
+      "learning_rate": 9.017223242802204e-05,
+      "loss": 0.5244,
+      "step": 5783
+    },
+    {
+      "epoch": 1.8571199229410822,
+      "grad_norm": 0.8290317058563232,
+      "learning_rate": 9.013749542409963e-05,
+      "loss": 0.5412,
+      "step": 5784
+    },
+    {
+      "epoch": 1.8574410017659335,
+      "grad_norm": 0.9109243154525757,
+      "learning_rate": 9.01027596218935e-05,
+      "loss": 0.7158,
+      "step": 5785
+    },
+    {
+      "epoch": 1.857762080590785,
+      "grad_norm": 0.8047156929969788,
+      "learning_rate": 9.006802502563612e-05,
+      "loss": 0.4727,
+      "step": 5786
+    },
+    {
+      "epoch": 1.8580831594156364,
+      "grad_norm": 0.9587835073471069,
+      "learning_rate": 9.003329163955972e-05,
+      "loss": 0.6326,
+      "step": 5787
+    },
+    {
+      "epoch": 1.858404238240488,
+      "grad_norm": 0.8253732919692993,
+      "learning_rate": 8.999855946789653e-05,
+      "loss": 0.5527,
+      "step": 5788
+    },
+    {
+      "epoch": 1.8587253170653395,
+      "grad_norm": 1.1531040668487549,
+      "learning_rate": 8.99638285148785e-05,
+      "loss": 0.4517,
+      "step": 5789
+    },
+    {
+      "epoch": 1.859046395890191,
+      "grad_norm": 0.7936963438987732,
+      "learning_rate": 8.992909878473758e-05,
+      "loss": 0.492,
+      "step": 5790
+    },
+    {
+      "epoch": 1.8593674747150426,
+      "grad_norm": 1.0635939836502075,
+      "learning_rate": 8.989437028170537e-05,
+      "loss": 0.5887,
+      "step": 5791
+    },
+    {
+      "epoch": 1.8596885535398942,
+      "grad_norm": 4.376264572143555,
+      "learning_rate": 8.985964301001353e-05,
+      "loss": 0.5299,
+      "step": 5792
+    },
+    {
+      "epoch": 1.8600096323647457,
+      "grad_norm": 0.9289653897285461,
+      "learning_rate": 8.982491697389338e-05,
+      "loss": 0.6141,
+      "step": 5793
+    },
+    {
+      "epoch": 1.860330711189597,
+      "grad_norm": 1.19081449508667,
+      "learning_rate": 8.979019217757625e-05,
+      "loss": 0.9054,
+      "step": 5794
+    },
+    {
+      "epoch": 1.8606517900144486,
+      "grad_norm": 0.8353030681610107,
+      "learning_rate": 8.975546862529328e-05,
+      "loss": 0.5393,
+      "step": 5795
+    },
+    {
+      "epoch": 1.8609728688393,
+      "grad_norm": 0.9625294804573059,
+      "learning_rate": 8.972074632127533e-05,
+      "loss": 0.5229,
+      "step": 5796
+    },
+    {
+      "epoch": 1.8612939476641515,
+      "grad_norm": 0.9256024956703186,
+      "learning_rate": 8.96860252697533e-05,
+      "loss": 0.3256,
+      "step": 5797
+    },
+    {
+      "epoch": 1.861615026489003,
+      "grad_norm": 0.9383046627044678,
+      "learning_rate": 8.965130547495776e-05,
+      "loss": 0.7085,
+      "step": 5798
+    },
+    {
+      "epoch": 1.8619361053138546,
+      "grad_norm": 0.8655479550361633,
+      "learning_rate": 8.961658694111929e-05,
+      "loss": 0.5807,
+      "step": 5799
+    },
+    {
+      "epoch": 1.8622571841387061,
+      "grad_norm": 1.0678274631500244,
+      "learning_rate": 8.958186967246816e-05,
+      "loss": 0.6035,
+      "step": 5800
+    },
+    {
+      "epoch": 1.8625782629635577,
+      "grad_norm": 0.6425728797912598,
+      "learning_rate": 8.954715367323468e-05,
+      "loss": 0.3561,
+      "step": 5801
+    },
+    {
+      "epoch": 1.8628993417884092,
+      "grad_norm": 0.835169792175293,
+      "learning_rate": 8.951243894764876e-05,
+      "loss": 0.5272,
+      "step": 5802
+    },
+    {
+      "epoch": 1.8632204206132605,
+      "grad_norm": 0.9119851589202881,
+      "learning_rate": 8.947772549994035e-05,
+      "loss": 0.5669,
+      "step": 5803
+    },
+    {
+      "epoch": 1.863541499438112,
+      "grad_norm": 0.8746310472488403,
+      "learning_rate": 8.944301333433922e-05,
+      "loss": 0.5494,
+      "step": 5804
+    },
+    {
+      "epoch": 1.8638625782629634,
+      "grad_norm": 1.2345632314682007,
+      "learning_rate": 8.940830245507483e-05,
+      "loss": 0.751,
+      "step": 5805
+    },
+    {
+      "epoch": 1.864183657087815,
+      "grad_norm": 1.1940757036209106,
+      "learning_rate": 8.93735928663767e-05,
+      "loss": 0.6379,
+      "step": 5806
+    },
+    {
+      "epoch": 1.8645047359126665,
+      "grad_norm": 1.1962798833847046,
+      "learning_rate": 8.933888457247402e-05,
+      "loss": 0.5358,
+      "step": 5807
+    },
+    {
+      "epoch": 1.864825814737518,
+      "grad_norm": 1.165390968322754,
+      "learning_rate": 8.930417757759592e-05,
+      "loss": 0.5781,
+      "step": 5808
+    },
+    {
+      "epoch": 1.8651468935623696,
+      "grad_norm": 0.9763917922973633,
+      "learning_rate": 8.926947188597134e-05,
+      "loss": 0.4743,
+      "step": 5809
+    },
+    {
+      "epoch": 1.8654679723872212,
+      "grad_norm": 1.0017492771148682,
+      "learning_rate": 8.923476750182908e-05,
+      "loss": 0.6824,
+      "step": 5810
+    },
+    {
+      "epoch": 1.8657890512120727,
+      "grad_norm": 1.386108160018921,
+      "learning_rate": 8.920006442939772e-05,
+      "loss": 0.595,
+      "step": 5811
+    },
+    {
+      "epoch": 1.866110130036924,
+      "grad_norm": 0.5562759041786194,
+      "learning_rate": 8.916536267290578e-05,
+      "loss": 0.3123,
+      "step": 5812
+    },
+    {
+      "epoch": 1.8664312088617756,
+      "grad_norm": 0.5364285111427307,
+      "learning_rate": 8.913066223658151e-05,
+      "loss": 0.3485,
+      "step": 5813
+    },
+    {
+      "epoch": 1.866752287686627,
+      "grad_norm": 0.5442401170730591,
+      "learning_rate": 8.909596312465306e-05,
+      "loss": 0.3492,
+      "step": 5814
+    },
+    {
+      "epoch": 1.8670733665114785,
+      "grad_norm": 0.5919818878173828,
+      "learning_rate": 8.906126534134848e-05,
+      "loss": 0.8261,
+      "step": 5815
+    },
+    {
+      "epoch": 1.86739444533633,
+      "grad_norm": 0.5136668682098389,
+      "learning_rate": 8.902656889089548e-05,
+      "loss": 0.7693,
+      "step": 5816
+    },
+    {
+      "epoch": 1.8677155241611816,
+      "grad_norm": 0.7814684510231018,
+      "learning_rate": 8.89918737775218e-05,
+      "loss": 0.456,
+      "step": 5817
+    },
+    {
+      "epoch": 1.8680366029860331,
+      "grad_norm": 0.5206146240234375,
+      "learning_rate": 8.895718000545489e-05,
+      "loss": 0.2985,
+      "step": 5818
+    },
+    {
+      "epoch": 1.8683576818108847,
+      "grad_norm": 0.5971502065658569,
+      "learning_rate": 8.892248757892214e-05,
+      "loss": 0.2403,
+      "step": 5819
+    },
+    {
+      "epoch": 1.8686787606357362,
+      "grad_norm": 0.8546911478042603,
+      "learning_rate": 8.888779650215068e-05,
+      "loss": 0.2383,
+      "step": 5820
+    },
+    {
+      "epoch": 1.8689998394605876,
+      "grad_norm": 0.9832369089126587,
+      "learning_rate": 8.885310677936746e-05,
+      "loss": 0.7438,
+      "step": 5821
+    },
+    {
+      "epoch": 1.8693209182854391,
+      "grad_norm": 1.0020122528076172,
+      "learning_rate": 8.88184184147994e-05,
+      "loss": 0.8816,
+      "step": 5822
+    },
+    {
+      "epoch": 1.8696419971102904,
+      "grad_norm": 0.7644429206848145,
+      "learning_rate": 8.878373141267311e-05,
+      "loss": 0.6192,
+      "step": 5823
+    },
+    {
+      "epoch": 1.869963075935142,
+      "grad_norm": 0.6598174571990967,
+      "learning_rate": 8.874904577721518e-05,
+      "loss": 0.4108,
+      "step": 5824
+    },
+    {
+      "epoch": 1.8702841547599935,
+      "grad_norm": 0.7471329569816589,
+      "learning_rate": 8.871436151265184e-05,
+      "loss": 0.5137,
+      "step": 5825
+    },
+    {
+      "epoch": 1.870605233584845,
+      "grad_norm": 0.910190999507904,
+      "learning_rate": 8.867967862320934e-05,
+      "loss": 0.5835,
+      "step": 5826
+    },
+    {
+      "epoch": 1.8709263124096966,
+      "grad_norm": 0.7468095421791077,
+      "learning_rate": 8.864499711311362e-05,
+      "loss": 0.417,
+      "step": 5827
+    },
+    {
+      "epoch": 1.8712473912345482,
+      "grad_norm": 0.8930068016052246,
+      "learning_rate": 8.861031698659063e-05,
+      "loss": 0.6094,
+      "step": 5828
+    },
+    {
+      "epoch": 1.8715684700593997,
+      "grad_norm": 0.793305516242981,
+      "learning_rate": 8.857563824786596e-05,
+      "loss": 0.5365,
+      "step": 5829
+    },
+    {
+      "epoch": 1.871889548884251,
+      "grad_norm": 0.6843454837799072,
+      "learning_rate": 8.854096090116508e-05,
+      "loss": 0.4672,
+      "step": 5830
+    },
+    {
+      "epoch": 1.8722106277091026,
+      "grad_norm": 0.9117445349693298,
+      "learning_rate": 8.850628495071336e-05,
+      "loss": 0.5671,
+      "step": 5831
+    },
+    {
+      "epoch": 1.872531706533954,
+      "grad_norm": 0.6631526350975037,
+      "learning_rate": 8.847161040073594e-05,
+      "loss": 0.3177,
+      "step": 5832
+    },
+    {
+      "epoch": 1.8728527853588055,
+      "grad_norm": 0.716529130935669,
+      "learning_rate": 8.843693725545786e-05,
+      "loss": 0.4629,
+      "step": 5833
+    },
+    {
+      "epoch": 1.873173864183657,
+      "grad_norm": 1.5731027126312256,
+      "learning_rate": 8.840226551910387e-05,
+      "loss": 0.8211,
+      "step": 5834
+    },
+    {
+      "epoch": 1.8734949430085086,
+      "grad_norm": 1.7772483825683594,
+      "learning_rate": 8.836759519589867e-05,
+      "loss": 0.7299,
+      "step": 5835
+    },
+    {
+      "epoch": 1.8738160218333602,
+      "grad_norm": 1.1696412563323975,
+      "learning_rate": 8.833292629006668e-05,
+      "loss": 0.5431,
+      "step": 5836
+    },
+    {
+      "epoch": 1.8741371006582117,
+      "grad_norm": 1.2018239498138428,
+      "learning_rate": 8.829825880583226e-05,
+      "loss": 0.5802,
+      "step": 5837
+    },
+    {
+      "epoch": 1.8744581794830633,
+      "grad_norm": 1.0494779348373413,
+      "learning_rate": 8.826359274741953e-05,
+      "loss": 0.7327,
+      "step": 5838
+    },
+    {
+      "epoch": 1.8747792583079146,
+      "grad_norm": 1.106572151184082,
+      "learning_rate": 8.822892811905237e-05,
+      "loss": 0.6034,
+      "step": 5839
+    },
+    {
+      "epoch": 1.8751003371327661,
+      "grad_norm": 0.8521847724914551,
+      "learning_rate": 8.819426492495464e-05,
+      "loss": 0.504,
+      "step": 5840
+    },
+    {
+      "epoch": 1.8754214159576175,
+      "grad_norm": 0.9584136605262756,
+      "learning_rate": 8.81596031693499e-05,
+      "loss": 0.546,
+      "step": 5841
+    },
+    {
+      "epoch": 1.875742494782469,
+      "grad_norm": 1.2079963684082031,
+      "learning_rate": 8.812494285646163e-05,
+      "loss": 0.491,
+      "step": 5842
+    },
+    {
+      "epoch": 1.8760635736073206,
+      "grad_norm": 0.8575901389122009,
+      "learning_rate": 8.809028399051302e-05,
+      "loss": 0.6095,
+      "step": 5843
+    },
+    {
+      "epoch": 1.8763846524321721,
+      "grad_norm": 0.8896117806434631,
+      "learning_rate": 8.805562657572723e-05,
+      "loss": 0.5603,
+      "step": 5844
+    },
+    {
+      "epoch": 1.8767057312570237,
+      "grad_norm": 0.928081750869751,
+      "learning_rate": 8.802097061632705e-05,
+      "loss": 0.5653,
+      "step": 5845
+    },
+    {
+      "epoch": 1.8770268100818752,
+      "grad_norm": 1.1048555374145508,
+      "learning_rate": 8.79863161165353e-05,
+      "loss": 0.5764,
+      "step": 5846
+    },
+    {
+      "epoch": 1.8773478889067265,
+      "grad_norm": 0.8389851450920105,
+      "learning_rate": 8.79516630805745e-05,
+      "loss": 0.6391,
+      "step": 5847
+    },
+    {
+      "epoch": 1.877668967731578,
+      "grad_norm": 0.9550890922546387,
+      "learning_rate": 8.791701151266696e-05,
+      "loss": 0.6019,
+      "step": 5848
+    },
+    {
+      "epoch": 1.8779900465564296,
+      "grad_norm": 0.8240407109260559,
+      "learning_rate": 8.788236141703498e-05,
+      "loss": 0.4589,
+      "step": 5849
+    },
+    {
+      "epoch": 1.878311125381281,
+      "grad_norm": 0.9926547408103943,
+      "learning_rate": 8.784771279790044e-05,
+      "loss": 0.5949,
+      "step": 5850
+    },
+    {
+      "epoch": 1.8786322042061325,
+      "grad_norm": 1.015940546989441,
+      "learning_rate": 8.781306565948528e-05,
+      "loss": 0.611,
+      "step": 5851
+    },
+    {
+      "epoch": 1.878953283030984,
+      "grad_norm": 0.912274181842804,
+      "learning_rate": 8.777842000601105e-05,
+      "loss": 0.488,
+      "step": 5852
+    },
+    {
+      "epoch": 1.8792743618558356,
+      "grad_norm": 1.0201482772827148,
+      "learning_rate": 8.774377584169933e-05,
+      "loss": 0.6077,
+      "step": 5853
+    },
+    {
+      "epoch": 1.8795954406806872,
+      "grad_norm": 1.027444839477539,
+      "learning_rate": 8.77091331707713e-05,
+      "loss": 0.6085,
+      "step": 5854
+    },
+    {
+      "epoch": 1.8799165195055387,
+      "grad_norm": 0.821168065071106,
+      "learning_rate": 8.767449199744814e-05,
+      "loss": 0.5152,
+      "step": 5855
+    },
+    {
+      "epoch": 1.88023759833039,
+      "grad_norm": 1.338505506515503,
+      "learning_rate": 8.763985232595075e-05,
+      "loss": 0.5792,
+      "step": 5856
+    },
+    {
+      "epoch": 1.8805586771552416,
+      "grad_norm": 0.9771504402160645,
+      "learning_rate": 8.760521416049983e-05,
+      "loss": 0.6102,
+      "step": 5857
+    },
+    {
+      "epoch": 1.8808797559800932,
+      "grad_norm": 0.9123767018318176,
+      "learning_rate": 8.7570577505316e-05,
+      "loss": 0.5544,
+      "step": 5858
+    },
+    {
+      "epoch": 1.8812008348049445,
+      "grad_norm": 0.6781209707260132,
+      "learning_rate": 8.753594236461957e-05,
+      "loss": 0.4612,
+      "step": 5859
+    },
+    {
+      "epoch": 1.881521913629796,
+      "grad_norm": 0.8213598132133484,
+      "learning_rate": 8.750130874263077e-05,
+      "loss": 0.338,
+      "step": 5860
+    },
+    {
+      "epoch": 1.8818429924546476,
+      "grad_norm": 1.2444860935211182,
+      "learning_rate": 8.746667664356956e-05,
+      "loss": 0.6963,
+      "step": 5861
+    },
+    {
+      "epoch": 1.8821640712794991,
+      "grad_norm": 0.9669928550720215,
+      "learning_rate": 8.743204607165583e-05,
+      "loss": 0.3941,
+      "step": 5862
+    },
+    {
+      "epoch": 1.8824851501043507,
+      "grad_norm": 0.7179502248764038,
+      "learning_rate": 8.739741703110913e-05,
+      "loss": 0.3945,
+      "step": 5863
+    },
+    {
+      "epoch": 1.8828062289292022,
+      "grad_norm": 0.9342414736747742,
+      "learning_rate": 8.736278952614898e-05,
+      "loss": 0.4656,
+      "step": 5864
+    },
+    {
+      "epoch": 1.8831273077540536,
+      "grad_norm": 0.5712354779243469,
+      "learning_rate": 8.732816356099456e-05,
+      "loss": 0.8414,
+      "step": 5865
+    },
+    {
+      "epoch": 1.8834483865789051,
+      "grad_norm": 0.5650877952575684,
+      "learning_rate": 8.729353913986496e-05,
+      "loss": 0.77,
+      "step": 5866
+    },
+    {
+      "epoch": 1.8837694654037567,
+      "grad_norm": 0.6072726845741272,
+      "learning_rate": 8.72589162669791e-05,
+      "loss": 0.3555,
+      "step": 5867
+    },
+    {
+      "epoch": 1.884090544228608,
+      "grad_norm": 0.5450513362884521,
+      "learning_rate": 8.722429494655561e-05,
+      "loss": 0.2534,
+      "step": 5868
+    },
+    {
+      "epoch": 1.8844116230534596,
+      "grad_norm": 0.7012900114059448,
+      "learning_rate": 8.718967518281307e-05,
+      "loss": 0.4425,
+      "step": 5869
+    },
+    {
+      "epoch": 1.884732701878311,
+      "grad_norm": 0.7500693798065186,
+      "learning_rate": 8.715505697996971e-05,
+      "loss": 0.221,
+      "step": 5870
+    },
+    {
+      "epoch": 1.8850537807031627,
+      "grad_norm": 0.9719879627227783,
+      "learning_rate": 8.712044034224374e-05,
+      "loss": 0.8356,
+      "step": 5871
+    },
+    {
+      "epoch": 1.8853748595280142,
+      "grad_norm": 0.8122969269752502,
+      "learning_rate": 8.708582527385301e-05,
+      "loss": 0.623,
+      "step": 5872
+    },
+    {
+      "epoch": 1.8856959383528658,
+      "grad_norm": 1.0290989875793457,
+      "learning_rate": 8.705121177901532e-05,
+      "loss": 0.7414,
+      "step": 5873
+    },
+    {
+      "epoch": 1.886017017177717,
+      "grad_norm": 1.0142052173614502,
+      "learning_rate": 8.70165998619482e-05,
+      "loss": 0.6093,
+      "step": 5874
+    },
+    {
+      "epoch": 1.8863380960025686,
+      "grad_norm": 0.9030125141143799,
+      "learning_rate": 8.698198952686896e-05,
+      "loss": 0.6493,
+      "step": 5875
+    },
+    {
+      "epoch": 1.8866591748274202,
+      "grad_norm": 0.8314472436904907,
+      "learning_rate": 8.694738077799488e-05,
+      "loss": 0.6468,
+      "step": 5876
+    },
+    {
+      "epoch": 1.8869802536522715,
+      "grad_norm": 0.7704334855079651,
+      "learning_rate": 8.691277361954279e-05,
+      "loss": 0.4876,
+      "step": 5877
+    },
+    {
+      "epoch": 1.887301332477123,
+      "grad_norm": 0.8338825702667236,
+      "learning_rate": 8.687816805572956e-05,
+      "loss": 0.6363,
+      "step": 5878
+    },
+    {
+      "epoch": 1.8876224113019746,
+      "grad_norm": 0.8600116968154907,
+      "learning_rate": 8.684356409077176e-05,
+      "loss": 0.5554,
+      "step": 5879
+    },
+    {
+      "epoch": 1.8879434901268262,
+      "grad_norm": 0.8333651423454285,
+      "learning_rate": 8.680896172888576e-05,
+      "loss": 0.6546,
+      "step": 5880
+    },
+    {
+      "epoch": 1.8882645689516777,
+      "grad_norm": 0.9315165877342224,
+      "learning_rate": 8.677436097428775e-05,
+      "loss": 0.5263,
+      "step": 5881
+    },
+    {
+      "epoch": 1.8885856477765293,
+      "grad_norm": 0.9922060370445251,
+      "learning_rate": 8.673976183119376e-05,
+      "loss": 0.6409,
+      "step": 5882
+    },
+    {
+      "epoch": 1.8889067266013806,
+      "grad_norm": 0.9471839070320129,
+      "learning_rate": 8.670516430381958e-05,
+      "loss": 0.7208,
+      "step": 5883
+    },
+    {
+      "epoch": 1.8892278054262321,
+      "grad_norm": 1.207506537437439,
+      "learning_rate": 8.667056839638075e-05,
+      "loss": 0.8187,
+      "step": 5884
+    },
+    {
+      "epoch": 1.8895488842510837,
+      "grad_norm": 1.2479336261749268,
+      "learning_rate": 8.663597411309279e-05,
+      "loss": 0.4916,
+      "step": 5885
+    },
+    {
+      "epoch": 1.889869963075935,
+      "grad_norm": 0.7998619675636292,
+      "learning_rate": 8.660138145817079e-05,
+      "loss": 0.5774,
+      "step": 5886
+    },
+    {
+      "epoch": 1.8901910419007866,
+      "grad_norm": 0.6762514710426331,
+      "learning_rate": 8.656679043582986e-05,
+      "loss": 0.4435,
+      "step": 5887
+    },
+    {
+      "epoch": 1.8905121207256381,
+      "grad_norm": 2.044832944869995,
+      "learning_rate": 8.653220105028474e-05,
+      "loss": 0.6974,
+      "step": 5888
+    },
+    {
+      "epoch": 1.8908331995504897,
+      "grad_norm": 0.8140920400619507,
+      "learning_rate": 8.64976133057501e-05,
+      "loss": 0.6755,
+      "step": 5889
+    },
+    {
+      "epoch": 1.8911542783753412,
+      "grad_norm": 0.837049663066864,
+      "learning_rate": 8.646302720644027e-05,
+      "loss": 0.4787,
+      "step": 5890
+    },
+    {
+      "epoch": 1.8914753572001928,
+      "grad_norm": 0.9118859767913818,
+      "learning_rate": 8.642844275656957e-05,
+      "loss": 0.6674,
+      "step": 5891
+    },
+    {
+      "epoch": 1.891796436025044,
+      "grad_norm": 0.8164528012275696,
+      "learning_rate": 8.639385996035194e-05,
+      "loss": 0.5175,
+      "step": 5892
+    },
+    {
+      "epoch": 1.8921175148498957,
+      "grad_norm": 1.1621068716049194,
+      "learning_rate": 8.635927882200116e-05,
+      "loss": 0.5832,
+      "step": 5893
+    },
+    {
+      "epoch": 1.8924385936747472,
+      "grad_norm": 0.7055744528770447,
+      "learning_rate": 8.632469934573094e-05,
+      "loss": 0.5454,
+      "step": 5894
+    },
+    {
+      "epoch": 1.8927596724995985,
+      "grad_norm": 1.0053250789642334,
+      "learning_rate": 8.629012153575458e-05,
+      "loss": 0.7157,
+      "step": 5895
+    },
+    {
+      "epoch": 1.89308075132445,
+      "grad_norm": 0.8789286613464355,
+      "learning_rate": 8.625554539628535e-05,
+      "loss": 0.5651,
+      "step": 5896
+    },
+    {
+      "epoch": 1.8934018301493016,
+      "grad_norm": 0.8847569227218628,
+      "learning_rate": 8.62209709315362e-05,
+      "loss": 0.648,
+      "step": 5897
+    },
+    {
+      "epoch": 1.8937229089741532,
+      "grad_norm": 0.8232566714286804,
+      "learning_rate": 8.618639814571996e-05,
+      "loss": 0.473,
+      "step": 5898
+    },
+    {
+      "epoch": 1.8940439877990047,
+      "grad_norm": 0.8590362668037415,
+      "learning_rate": 8.615182704304918e-05,
+      "loss": 0.6142,
+      "step": 5899
+    },
+    {
+      "epoch": 1.8943650666238563,
+      "grad_norm": 1.0845537185668945,
+      "learning_rate": 8.611725762773631e-05,
+      "loss": 0.5017,
+      "step": 5900
+    },
+    {
+      "epoch": 1.8946861454487076,
+      "grad_norm": 0.7190443277359009,
+      "learning_rate": 8.608268990399349e-05,
+      "loss": 0.4462,
+      "step": 5901
+    },
+    {
+      "epoch": 1.8950072242735592,
+      "grad_norm": 0.7313873171806335,
+      "learning_rate": 8.604812387603265e-05,
+      "loss": 0.4465,
+      "step": 5902
+    },
+    {
+      "epoch": 1.8953283030984105,
+      "grad_norm": 0.816182553768158,
+      "learning_rate": 8.601355954806561e-05,
+      "loss": 0.4299,
+      "step": 5903
+    },
+    {
+      "epoch": 1.895649381923262,
+      "grad_norm": 1.0076704025268555,
+      "learning_rate": 8.597899692430389e-05,
+      "loss": 0.5754,
+      "step": 5904
+    },
+    {
+      "epoch": 1.8959704607481136,
+      "grad_norm": 0.9564604759216309,
+      "learning_rate": 8.594443600895892e-05,
+      "loss": 0.4443,
+      "step": 5905
+    },
+    {
+      "epoch": 1.8962915395729651,
+      "grad_norm": 0.6225574612617493,
+      "learning_rate": 8.590987680624174e-05,
+      "loss": 0.3169,
+      "step": 5906
+    },
+    {
+      "epoch": 1.8966126183978167,
+      "grad_norm": 1.0120649337768555,
+      "learning_rate": 8.587531932036335e-05,
+      "loss": 0.6126,
+      "step": 5907
+    },
+    {
+      "epoch": 1.8969336972226682,
+      "grad_norm": 1.029392957687378,
+      "learning_rate": 8.584076355553444e-05,
+      "loss": 0.5688,
+      "step": 5908
+    },
+    {
+      "epoch": 1.8972547760475198,
+      "grad_norm": 1.1245523691177368,
+      "learning_rate": 8.580620951596557e-05,
+      "loss": 0.5589,
+      "step": 5909
+    },
+    {
+      "epoch": 1.8975758548723711,
+      "grad_norm": 0.8369296789169312,
+      "learning_rate": 8.577165720586703e-05,
+      "loss": 0.3727,
+      "step": 5910
+    },
+    {
+      "epoch": 1.8978969336972227,
+      "grad_norm": 0.8176161646842957,
+      "learning_rate": 8.573710662944885e-05,
+      "loss": 0.3624,
+      "step": 5911
+    },
+    {
+      "epoch": 1.898218012522074,
+      "grad_norm": 0.9345711469650269,
+      "learning_rate": 8.570255779092098e-05,
+      "loss": 0.4462,
+      "step": 5912
+    },
+    {
+      "epoch": 1.8985390913469256,
+      "grad_norm": 1.294392466545105,
+      "learning_rate": 8.566801069449305e-05,
+      "loss": 0.3714,
+      "step": 5913
+    },
+    {
+      "epoch": 1.898860170171777,
+      "grad_norm": 0.5765014886856079,
+      "learning_rate": 8.56334653443746e-05,
+      "loss": 0.3126,
+      "step": 5914
+    },
+    {
+      "epoch": 1.8991812489966287,
+      "grad_norm": 0.6456901431083679,
+      "learning_rate": 8.559892174477479e-05,
+      "loss": 0.3416,
+      "step": 5915
+    },
+    {
+      "epoch": 1.8995023278214802,
+      "grad_norm": 0.4262344539165497,
+      "learning_rate": 8.55643798999027e-05,
+      "loss": 0.4155,
+      "step": 5916
+    },
+    {
+      "epoch": 1.8998234066463318,
+      "grad_norm": 0.4280303120613098,
+      "learning_rate": 8.55298398139671e-05,
+      "loss": 0.4822,
+      "step": 5917
+    },
+    {
+      "epoch": 1.9001444854711833,
+      "grad_norm": 0.6191977858543396,
+      "learning_rate": 8.549530149117664e-05,
+      "loss": 0.5129,
+      "step": 5918
+    },
+    {
+      "epoch": 1.9004655642960346,
+      "grad_norm": 0.5976957678794861,
+      "learning_rate": 8.546076493573972e-05,
+      "loss": 0.2694,
+      "step": 5919
+    },
+    {
+      "epoch": 1.9007866431208862,
+      "grad_norm": 0.6138757467269897,
+      "learning_rate": 8.542623015186445e-05,
+      "loss": 0.1946,
+      "step": 5920
+    },
+    {
+      "epoch": 1.9011077219457375,
+      "grad_norm": 0.624046802520752,
+      "learning_rate": 8.539169714375885e-05,
+      "loss": 0.131,
+      "step": 5921
+    },
+    {
+      "epoch": 1.901428800770589,
+      "grad_norm": 0.7090097069740295,
+      "learning_rate": 8.535716591563062e-05,
+      "loss": 0.3913,
+      "step": 5922
+    },
+    {
+      "epoch": 1.9017498795954406,
+      "grad_norm": 1.0457631349563599,
+      "learning_rate": 8.532263647168735e-05,
+      "loss": 0.6625,
+      "step": 5923
+    },
+    {
+      "epoch": 1.9020709584202922,
+      "grad_norm": 0.8407219648361206,
+      "learning_rate": 8.528810881613626e-05,
+      "loss": 0.6366,
+      "step": 5924
+    },
+    {
+      "epoch": 1.9023920372451437,
+      "grad_norm": 0.9645373225212097,
+      "learning_rate": 8.525358295318454e-05,
+      "loss": 0.524,
+      "step": 5925
+    },
+    {
+      "epoch": 1.9027131160699953,
+      "grad_norm": 0.8919804096221924,
+      "learning_rate": 8.521905888703893e-05,
+      "loss": 0.5851,
+      "step": 5926
+    },
+    {
+      "epoch": 1.9030341948948468,
+      "grad_norm": 0.740583598613739,
+      "learning_rate": 8.51845366219062e-05,
+      "loss": 0.5141,
+      "step": 5927
+    },
+    {
+      "epoch": 1.9033552737196981,
+      "grad_norm": 0.7482326030731201,
+      "learning_rate": 8.515001616199279e-05,
+      "loss": 0.4804,
+      "step": 5928
+    },
+    {
+      "epoch": 1.9036763525445497,
+      "grad_norm": 0.9648141860961914,
+      "learning_rate": 8.511549751150479e-05,
+      "loss": 0.6938,
+      "step": 5929
+    },
+    {
+      "epoch": 1.903997431369401,
+      "grad_norm": 0.8778332471847534,
+      "learning_rate": 8.508098067464832e-05,
+      "loss": 0.6264,
+      "step": 5930
+    },
+    {
+      "epoch": 1.9043185101942526,
+      "grad_norm": 0.9466199278831482,
+      "learning_rate": 8.504646565562906e-05,
+      "loss": 0.6328,
+      "step": 5931
+    },
+    {
+      "epoch": 1.9046395890191041,
+      "grad_norm": 1.0974069833755493,
+      "learning_rate": 8.501195245865263e-05,
+      "loss": 0.8376,
+      "step": 5932
+    },
+    {
+      "epoch": 1.9049606678439557,
+      "grad_norm": 1.1989012956619263,
+      "learning_rate": 8.497744108792429e-05,
+      "loss": 0.7932,
+      "step": 5933
+    },
+    {
+      "epoch": 1.9052817466688072,
+      "grad_norm": 1.0475575923919678,
+      "learning_rate": 8.494293154764924e-05,
+      "loss": 0.6127,
+      "step": 5934
+    },
+    {
+      "epoch": 1.9056028254936588,
+      "grad_norm": 0.8935070633888245,
+      "learning_rate": 8.490842384203225e-05,
+      "loss": 0.4574,
+      "step": 5935
+    },
+    {
+      "epoch": 1.9059239043185103,
+      "grad_norm": 0.9734428524971008,
+      "learning_rate": 8.487391797527808e-05,
+      "loss": 0.6899,
+      "step": 5936
+    },
+    {
+      "epoch": 1.9062449831433617,
+      "grad_norm": 1.1841446161270142,
+      "learning_rate": 8.483941395159114e-05,
+      "loss": 0.5652,
+      "step": 5937
+    },
+    {
+      "epoch": 1.9065660619682132,
+      "grad_norm": 0.7886440753936768,
+      "learning_rate": 8.480491177517557e-05,
+      "loss": 0.5065,
+      "step": 5938
+    },
+    {
+      "epoch": 1.9068871407930645,
+      "grad_norm": 1.349183440208435,
+      "learning_rate": 8.477041145023546e-05,
+      "loss": 0.6979,
+      "step": 5939
+    },
+    {
+      "epoch": 1.907208219617916,
+      "grad_norm": 0.7376853227615356,
+      "learning_rate": 8.473591298097448e-05,
+      "loss": 0.485,
+      "step": 5940
+    },
+    {
+      "epoch": 1.9075292984427676,
+      "grad_norm": 0.7252330183982849,
+      "learning_rate": 8.47014163715962e-05,
+      "loss": 0.4728,
+      "step": 5941
+    },
+    {
+      "epoch": 1.9078503772676192,
+      "grad_norm": 1.054677128791809,
+      "learning_rate": 8.466692162630392e-05,
+      "loss": 0.7022,
+      "step": 5942
+    },
+    {
+      "epoch": 1.9081714560924707,
+      "grad_norm": 0.9820946455001831,
+      "learning_rate": 8.463242874930079e-05,
+      "loss": 0.4489,
+      "step": 5943
+    },
+    {
+      "epoch": 1.9084925349173223,
+      "grad_norm": 0.9880334734916687,
+      "learning_rate": 8.459793774478957e-05,
+      "loss": 0.5668,
+      "step": 5944
+    },
+    {
+      "epoch": 1.9088136137421738,
+      "grad_norm": 0.8617029190063477,
+      "learning_rate": 8.456344861697289e-05,
+      "loss": 0.5731,
+      "step": 5945
+    },
+    {
+      "epoch": 1.9091346925670252,
+      "grad_norm": 0.9127218127250671,
+      "learning_rate": 8.452896137005321e-05,
+      "loss": 0.5747,
+      "step": 5946
+    },
+    {
+      "epoch": 1.9094557713918767,
+      "grad_norm": 0.8752137422561646,
+      "learning_rate": 8.449447600823262e-05,
+      "loss": 0.533,
+      "step": 5947
+    },
+    {
+      "epoch": 1.909776850216728,
+      "grad_norm": 0.9676142334938049,
+      "learning_rate": 8.445999253571315e-05,
+      "loss": 0.6118,
+      "step": 5948
+    },
+    {
+      "epoch": 1.9100979290415796,
+      "grad_norm": 1.485176920890808,
+      "learning_rate": 8.442551095669639e-05,
+      "loss": 0.491,
+      "step": 5949
+    },
+    {
+      "epoch": 1.9104190078664312,
+      "grad_norm": 0.8751528263092041,
+      "learning_rate": 8.439103127538392e-05,
+      "loss": 0.5723,
+      "step": 5950
+    },
+    {
+      "epoch": 1.9107400866912827,
+      "grad_norm": 0.8707427382469177,
+      "learning_rate": 8.435655349597689e-05,
+      "loss": 0.5169,
+      "step": 5951
+    },
+    {
+      "epoch": 1.9110611655161343,
+      "grad_norm": 1.3300323486328125,
+      "learning_rate": 8.432207762267644e-05,
+      "loss": 0.7869,
+      "step": 5952
+    },
+    {
+      "epoch": 1.9113822443409858,
+      "grad_norm": 0.8493139147758484,
+      "learning_rate": 8.428760365968326e-05,
+      "loss": 0.4855,
+      "step": 5953
+    },
+    {
+      "epoch": 1.9117033231658374,
+      "grad_norm": 0.7276837825775146,
+      "learning_rate": 8.425313161119787e-05,
+      "loss": 0.449,
+      "step": 5954
+    },
+    {
+      "epoch": 1.9120244019906887,
+      "grad_norm": 0.7487722635269165,
+      "learning_rate": 8.421866148142066e-05,
+      "loss": 0.5161,
+      "step": 5955
+    },
+    {
+      "epoch": 1.9123454808155402,
+      "grad_norm": 0.7644703984260559,
+      "learning_rate": 8.418419327455164e-05,
+      "loss": 0.4887,
+      "step": 5956
+    },
+    {
+      "epoch": 1.9126665596403916,
+      "grad_norm": 1.1047660112380981,
+      "learning_rate": 8.414972699479075e-05,
+      "loss": 0.6635,
+      "step": 5957
+    },
+    {
+      "epoch": 1.9129876384652431,
+      "grad_norm": 1.2923800945281982,
+      "learning_rate": 8.41152626463375e-05,
+      "loss": 0.8645,
+      "step": 5958
+    },
+    {
+      "epoch": 1.9133087172900947,
+      "grad_norm": 0.7669944167137146,
+      "learning_rate": 8.408080023339133e-05,
+      "loss": 0.5723,
+      "step": 5959
+    },
+    {
+      "epoch": 1.9136297961149462,
+      "grad_norm": 0.7324369549751282,
+      "learning_rate": 8.404633976015134e-05,
+      "loss": 0.4862,
+      "step": 5960
+    },
+    {
+      "epoch": 1.9139508749397978,
+      "grad_norm": 0.8924105763435364,
+      "learning_rate": 8.401188123081653e-05,
+      "loss": 0.4537,
+      "step": 5961
+    },
+    {
+      "epoch": 1.9142719537646493,
+      "grad_norm": 0.8175060153007507,
+      "learning_rate": 8.397742464958547e-05,
+      "loss": 0.5257,
+      "step": 5962
+    },
+    {
+      "epoch": 1.9145930325895009,
+      "grad_norm": 0.9657145738601685,
+      "learning_rate": 8.394297002065658e-05,
+      "loss": 0.4518,
+      "step": 5963
+    },
+    {
+      "epoch": 1.9149141114143522,
+      "grad_norm": 0.5297245383262634,
+      "learning_rate": 8.390851734822808e-05,
+      "loss": 0.366,
+      "step": 5964
+    },
+    {
+      "epoch": 1.9152351902392037,
+      "grad_norm": 0.5978273153305054,
+      "learning_rate": 8.387406663649795e-05,
+      "loss": 0.5728,
+      "step": 5965
+    },
+    {
+      "epoch": 1.915556269064055,
+      "grad_norm": 0.3594565987586975,
+      "learning_rate": 8.383961788966391e-05,
+      "loss": 0.4309,
+      "step": 5966
+    },
+    {
+      "epoch": 1.9158773478889066,
+      "grad_norm": 0.5129295587539673,
+      "learning_rate": 8.380517111192337e-05,
+      "loss": 0.7305,
+      "step": 5967
+    },
+    {
+      "epoch": 1.9161984267137582,
+      "grad_norm": 0.6634910702705383,
+      "learning_rate": 8.377072630747364e-05,
+      "loss": 0.5793,
+      "step": 5968
+    },
+    {
+      "epoch": 1.9165195055386097,
+      "grad_norm": 0.5111119747161865,
+      "learning_rate": 8.373628348051165e-05,
+      "loss": 0.2366,
+      "step": 5969
+    },
+    {
+      "epoch": 1.9168405843634613,
+      "grad_norm": 0.6472846269607544,
+      "learning_rate": 8.37018426352342e-05,
+      "loss": 0.1341,
+      "step": 5970
+    },
+    {
+      "epoch": 1.9171616631883128,
+      "grad_norm": 0.5201217532157898,
+      "learning_rate": 8.366740377583781e-05,
+      "loss": 0.2317,
+      "step": 5971
+    },
+    {
+      "epoch": 1.9174827420131644,
+      "grad_norm": 0.4701801836490631,
+      "learning_rate": 8.363296690651868e-05,
+      "loss": 0.3096,
+      "step": 5972
+    },
+    {
+      "epoch": 1.9178038208380157,
+      "grad_norm": 0.9514907598495483,
+      "learning_rate": 8.359853203147291e-05,
+      "loss": 0.7662,
+      "step": 5973
+    },
+    {
+      "epoch": 1.9181248996628673,
+      "grad_norm": 0.7471129298210144,
+      "learning_rate": 8.356409915489625e-05,
+      "loss": 0.4633,
+      "step": 5974
+    },
+    {
+      "epoch": 1.9184459784877186,
+      "grad_norm": 0.7893948554992676,
+      "learning_rate": 8.352966828098428e-05,
+      "loss": 0.5721,
+      "step": 5975
+    },
+    {
+      "epoch": 1.9187670573125701,
+      "grad_norm": 0.6973623633384705,
+      "learning_rate": 8.349523941393224e-05,
+      "loss": 0.5921,
+      "step": 5976
+    },
+    {
+      "epoch": 1.9190881361374217,
+      "grad_norm": 0.6330721378326416,
+      "learning_rate": 8.346081255793525e-05,
+      "loss": 0.4046,
+      "step": 5977
+    },
+    {
+      "epoch": 1.9194092149622732,
+      "grad_norm": 0.844613254070282,
+      "learning_rate": 8.342638771718802e-05,
+      "loss": 0.6565,
+      "step": 5978
+    },
+    {
+      "epoch": 1.9197302937871248,
+      "grad_norm": 0.8681631088256836,
+      "learning_rate": 8.339196489588523e-05,
+      "loss": 0.5315,
+      "step": 5979
+    },
+    {
+      "epoch": 1.9200513726119763,
+      "grad_norm": 0.8076524138450623,
+      "learning_rate": 8.335754409822114e-05,
+      "loss": 0.621,
+      "step": 5980
+    },
+    {
+      "epoch": 1.9203724514368279,
+      "grad_norm": 0.8629617094993591,
+      "learning_rate": 8.332312532838978e-05,
+      "loss": 0.5126,
+      "step": 5981
+    },
+    {
+      "epoch": 1.9206935302616792,
+      "grad_norm": 0.8805716633796692,
+      "learning_rate": 8.328870859058506e-05,
+      "loss": 0.6354,
+      "step": 5982
+    },
+    {
+      "epoch": 1.9210146090865308,
+      "grad_norm": 0.9122462272644043,
+      "learning_rate": 8.325429388900046e-05,
+      "loss": 0.6502,
+      "step": 5983
+    },
+    {
+      "epoch": 1.921335687911382,
+      "grad_norm": 0.7924925684928894,
+      "learning_rate": 8.321988122782937e-05,
+      "loss": 0.6105,
+      "step": 5984
+    },
+    {
+      "epoch": 1.9216567667362336,
+      "grad_norm": 0.9744467735290527,
+      "learning_rate": 8.318547061126485e-05,
+      "loss": 0.6674,
+      "step": 5985
+    },
+    {
+      "epoch": 1.9219778455610852,
+      "grad_norm": 0.8236846327781677,
+      "learning_rate": 8.315106204349976e-05,
+      "loss": 0.5076,
+      "step": 5986
+    },
+    {
+      "epoch": 1.9222989243859367,
+      "grad_norm": 0.8699111342430115,
+      "learning_rate": 8.311665552872662e-05,
+      "loss": 0.6371,
+      "step": 5987
+    },
+    {
+      "epoch": 1.9226200032107883,
+      "grad_norm": 0.9178052544593811,
+      "learning_rate": 8.30822510711378e-05,
+      "loss": 0.7132,
+      "step": 5988
+    },
+    {
+      "epoch": 1.9229410820356398,
+      "grad_norm": 1.2815251350402832,
+      "learning_rate": 8.30478486749254e-05,
+      "loss": 0.6409,
+      "step": 5989
+    },
+    {
+      "epoch": 1.9232621608604914,
+      "grad_norm": 0.7497451901435852,
+      "learning_rate": 8.301344834428116e-05,
+      "loss": 0.4567,
+      "step": 5990
+    },
+    {
+      "epoch": 1.9235832396853427,
+      "grad_norm": 0.8337712287902832,
+      "learning_rate": 8.297905008339677e-05,
+      "loss": 0.603,
+      "step": 5991
+    },
+    {
+      "epoch": 1.9239043185101943,
+      "grad_norm": 1.0814270973205566,
+      "learning_rate": 8.294465389646345e-05,
+      "loss": 0.7705,
+      "step": 5992
+    },
+    {
+      "epoch": 1.9242253973350456,
+      "grad_norm": 0.9741120934486389,
+      "learning_rate": 8.291025978767235e-05,
+      "loss": 0.7132,
+      "step": 5993
+    },
+    {
+      "epoch": 1.9245464761598972,
+      "grad_norm": 0.6710167527198792,
+      "learning_rate": 8.287586776121423e-05,
+      "loss": 0.416,
+      "step": 5994
+    },
+    {
+      "epoch": 1.9248675549847487,
+      "grad_norm": 0.6828485131263733,
+      "learning_rate": 8.284147782127971e-05,
+      "loss": 0.4111,
+      "step": 5995
+    },
+    {
+      "epoch": 1.9251886338096003,
+      "grad_norm": 1.174616813659668,
+      "learning_rate": 8.280708997205904e-05,
+      "loss": 0.6958,
+      "step": 5996
+    },
+    {
+      "epoch": 1.9255097126344518,
+      "grad_norm": 1.070745825767517,
+      "learning_rate": 8.277270421774235e-05,
+      "loss": 0.7379,
+      "step": 5997
+    },
+    {
+      "epoch": 1.9258307914593034,
+      "grad_norm": 0.9961045384407043,
+      "learning_rate": 8.273832056251937e-05,
+      "loss": 0.467,
+      "step": 5998
+    },
+    {
+      "epoch": 1.926151870284155,
+      "grad_norm": 0.8946678638458252,
+      "learning_rate": 8.270393901057964e-05,
+      "loss": 0.5288,
+      "step": 5999
+    },
+    {
+      "epoch": 1.9264729491090062,
+      "grad_norm": 0.7546229362487793,
+      "learning_rate": 8.266955956611253e-05,
+      "loss": 0.4404,
+      "step": 6000
+    },
+    {
+      "epoch": 1.9267940279338578,
+      "grad_norm": 0.8234146237373352,
+      "learning_rate": 8.263518223330697e-05,
+      "loss": 0.5327,
+      "step": 6001
+    },
+    {
+      "epoch": 1.9271151067587091,
+      "grad_norm": 0.760515570640564,
+      "learning_rate": 8.26008070163518e-05,
+      "loss": 0.4604,
+      "step": 6002
+    },
+    {
+      "epoch": 1.9274361855835607,
+      "grad_norm": 1.1221221685409546,
+      "learning_rate": 8.256643391943551e-05,
+      "loss": 0.5529,
+      "step": 6003
+    },
+    {
+      "epoch": 1.9277572644084122,
+      "grad_norm": 0.8846840858459473,
+      "learning_rate": 8.25320629467464e-05,
+      "loss": 0.4192,
+      "step": 6004
+    },
+    {
+      "epoch": 1.9280783432332638,
+      "grad_norm": 0.8447346091270447,
+      "learning_rate": 8.249769410247239e-05,
+      "loss": 0.4966,
+      "step": 6005
+    },
+    {
+      "epoch": 1.9283994220581153,
+      "grad_norm": 1.1361325979232788,
+      "learning_rate": 8.24633273908013e-05,
+      "loss": 0.555,
+      "step": 6006
+    },
+    {
+      "epoch": 1.9287205008829669,
+      "grad_norm": 1.2147115468978882,
+      "learning_rate": 8.242896281592057e-05,
+      "loss": 0.699,
+      "step": 6007
+    },
+    {
+      "epoch": 1.9290415797078184,
+      "grad_norm": 0.7966561317443848,
+      "learning_rate": 8.239460038201739e-05,
+      "loss": 0.4552,
+      "step": 6008
+    },
+    {
+      "epoch": 1.9293626585326697,
+      "grad_norm": 0.9599836468696594,
+      "learning_rate": 8.236024009327879e-05,
+      "loss": 0.6289,
+      "step": 6009
+    },
+    {
+      "epoch": 1.9296837373575213,
+      "grad_norm": 1.475778341293335,
+      "learning_rate": 8.23258819538914e-05,
+      "loss": 0.4814,
+      "step": 6010
+    },
+    {
+      "epoch": 1.9300048161823726,
+      "grad_norm": 1.2509891986846924,
+      "learning_rate": 8.229152596804168e-05,
+      "loss": 0.4725,
+      "step": 6011
+    },
+    {
+      "epoch": 1.9303258950072242,
+      "grad_norm": 0.7716514468193054,
+      "learning_rate": 8.225717213991579e-05,
+      "loss": 0.4775,
+      "step": 6012
+    },
+    {
+      "epoch": 1.9306469738320757,
+      "grad_norm": 0.9235790967941284,
+      "learning_rate": 8.222282047369971e-05,
+      "loss": 0.6091,
+      "step": 6013
+    },
+    {
+      "epoch": 1.9309680526569273,
+      "grad_norm": 0.5926783084869385,
+      "learning_rate": 8.218847097357898e-05,
+      "loss": 0.3537,
+      "step": 6014
+    },
+    {
+      "epoch": 1.9312891314817788,
+      "grad_norm": 0.5469971299171448,
+      "learning_rate": 8.215412364373907e-05,
+      "loss": 0.5765,
+      "step": 6015
+    },
+    {
+      "epoch": 1.9316102103066304,
+      "grad_norm": 0.5453217029571533,
+      "learning_rate": 8.211977848836506e-05,
+      "loss": 0.8021,
+      "step": 6016
+    },
+    {
+      "epoch": 1.931931289131482,
+      "grad_norm": 0.5584077835083008,
+      "learning_rate": 8.208543551164178e-05,
+      "loss": 0.2779,
+      "step": 6017
+    },
+    {
+      "epoch": 1.9322523679563333,
+      "grad_norm": 1.3637664318084717,
+      "learning_rate": 8.205109471775387e-05,
+      "loss": 0.5096,
+      "step": 6018
+    },
+    {
+      "epoch": 1.9325734467811848,
+      "grad_norm": 0.6770474910736084,
+      "learning_rate": 8.201675611088558e-05,
+      "loss": 0.3166,
+      "step": 6019
+    },
+    {
+      "epoch": 1.9328945256060361,
+      "grad_norm": 0.7160546183586121,
+      "learning_rate": 8.198241969522107e-05,
+      "loss": 0.1541,
+      "step": 6020
+    },
+    {
+      "epoch": 1.9332156044308877,
+      "grad_norm": 0.6660668253898621,
+      "learning_rate": 8.194808547494401e-05,
+      "loss": 0.5054,
+      "step": 6021
+    },
+    {
+      "epoch": 1.9335366832557392,
+      "grad_norm": 0.7240803241729736,
+      "learning_rate": 8.191375345423799e-05,
+      "loss": 0.5679,
+      "step": 6022
+    },
+    {
+      "epoch": 1.9338577620805908,
+      "grad_norm": 0.9651069045066833,
+      "learning_rate": 8.187942363728625e-05,
+      "loss": 0.5552,
+      "step": 6023
+    },
+    {
+      "epoch": 1.9341788409054423,
+      "grad_norm": 0.6748064160346985,
+      "learning_rate": 8.184509602827181e-05,
+      "loss": 0.5365,
+      "step": 6024
+    },
+    {
+      "epoch": 1.934499919730294,
+      "grad_norm": 0.7690969705581665,
+      "learning_rate": 8.181077063137733e-05,
+      "loss": 0.5248,
+      "step": 6025
+    },
+    {
+      "epoch": 1.9348209985551454,
+      "grad_norm": 0.7835337519645691,
+      "learning_rate": 8.177644745078526e-05,
+      "loss": 0.5647,
+      "step": 6026
+    },
+    {
+      "epoch": 1.9351420773799968,
+      "grad_norm": 0.7683063745498657,
+      "learning_rate": 8.174212649067781e-05,
+      "loss": 0.485,
+      "step": 6027
+    },
+    {
+      "epoch": 1.9354631562048483,
+      "grad_norm": 0.9228124618530273,
+      "learning_rate": 8.170780775523684e-05,
+      "loss": 0.6106,
+      "step": 6028
+    },
+    {
+      "epoch": 1.9357842350296997,
+      "grad_norm": 0.8772586584091187,
+      "learning_rate": 8.167349124864405e-05,
+      "loss": 0.5138,
+      "step": 6029
+    },
+    {
+      "epoch": 1.9361053138545512,
+      "grad_norm": 0.8027740716934204,
+      "learning_rate": 8.163917697508072e-05,
+      "loss": 0.583,
+      "step": 6030
+    },
+    {
+      "epoch": 1.9364263926794028,
+      "grad_norm": 0.9190700054168701,
+      "learning_rate": 8.160486493872798e-05,
+      "loss": 0.7091,
+      "step": 6031
+    },
+    {
+      "epoch": 1.9367474715042543,
+      "grad_norm": 0.9426003098487854,
+      "learning_rate": 8.157055514376666e-05,
+      "loss": 0.6263,
+      "step": 6032
+    },
+    {
+      "epoch": 1.9370685503291059,
+      "grad_norm": 0.918766975402832,
+      "learning_rate": 8.153624759437732e-05,
+      "loss": 0.669,
+      "step": 6033
+    },
+    {
+      "epoch": 1.9373896291539574,
+      "grad_norm": 0.9351176023483276,
+      "learning_rate": 8.15019422947402e-05,
+      "loss": 0.6859,
+      "step": 6034
+    },
+    {
+      "epoch": 1.937710707978809,
+      "grad_norm": 0.8434003591537476,
+      "learning_rate": 8.146763924903527e-05,
+      "loss": 0.5469,
+      "step": 6035
+    },
+    {
+      "epoch": 1.9380317868036603,
+      "grad_norm": 0.8188439607620239,
+      "learning_rate": 8.14333384614423e-05,
+      "loss": 0.5615,
+      "step": 6036
+    },
+    {
+      "epoch": 1.9383528656285118,
+      "grad_norm": 0.782618522644043,
+      "learning_rate": 8.139903993614068e-05,
+      "loss": 0.5117,
+      "step": 6037
+    },
+    {
+      "epoch": 1.9386739444533632,
+      "grad_norm": 0.882308840751648,
+      "learning_rate": 8.136474367730969e-05,
+      "loss": 0.5433,
+      "step": 6038
+    },
+    {
+      "epoch": 1.9389950232782147,
+      "grad_norm": 1.1491224765777588,
+      "learning_rate": 8.133044968912811e-05,
+      "loss": 0.5036,
+      "step": 6039
+    },
+    {
+      "epoch": 1.9393161021030663,
+      "grad_norm": 1.1491897106170654,
+      "learning_rate": 8.129615797577461e-05,
+      "loss": 0.63,
+      "step": 6040
+    },
+    {
+      "epoch": 1.9396371809279178,
+      "grad_norm": 0.8906580805778503,
+      "learning_rate": 8.126186854142752e-05,
+      "loss": 0.4457,
+      "step": 6041
+    },
+    {
+      "epoch": 1.9399582597527694,
+      "grad_norm": 0.7406003475189209,
+      "learning_rate": 8.122758139026495e-05,
+      "loss": 0.4659,
+      "step": 6042
+    },
+    {
+      "epoch": 1.940279338577621,
+      "grad_norm": 0.8870121240615845,
+      "learning_rate": 8.119329652646463e-05,
+      "loss": 0.6007,
+      "step": 6043
+    },
+    {
+      "epoch": 1.9406004174024725,
+      "grad_norm": 1.128095269203186,
+      "learning_rate": 8.115901395420407e-05,
+      "loss": 0.6736,
+      "step": 6044
+    },
+    {
+      "epoch": 1.9409214962273238,
+      "grad_norm": 1.0389516353607178,
+      "learning_rate": 8.11247336776605e-05,
+      "loss": 0.6387,
+      "step": 6045
+    },
+    {
+      "epoch": 1.9412425750521753,
+      "grad_norm": 1.045128583908081,
+      "learning_rate": 8.109045570101086e-05,
+      "loss": 0.6013,
+      "step": 6046
+    },
+    {
+      "epoch": 1.9415636538770267,
+      "grad_norm": 0.6581218838691711,
+      "learning_rate": 8.105618002843189e-05,
+      "loss": 0.4185,
+      "step": 6047
+    },
+    {
+      "epoch": 1.9418847327018782,
+      "grad_norm": 0.7015986442565918,
+      "learning_rate": 8.102190666409987e-05,
+      "loss": 0.3757,
+      "step": 6048
+    },
+    {
+      "epoch": 1.9422058115267298,
+      "grad_norm": 1.0713157653808594,
+      "learning_rate": 8.0987635612191e-05,
+      "loss": 0.5619,
+      "step": 6049
+    },
+    {
+      "epoch": 1.9425268903515813,
+      "grad_norm": 0.5126787424087524,
+      "learning_rate": 8.095336687688102e-05,
+      "loss": 0.2985,
+      "step": 6050
+    },
+    {
+      "epoch": 1.9428479691764329,
+      "grad_norm": 0.9810903668403625,
+      "learning_rate": 8.091910046234552e-05,
+      "loss": 0.5227,
+      "step": 6051
+    },
+    {
+      "epoch": 1.9431690480012844,
+      "grad_norm": 1.0264313220977783,
+      "learning_rate": 8.088483637275979e-05,
+      "loss": 0.591,
+      "step": 6052
+    },
+    {
+      "epoch": 1.943490126826136,
+      "grad_norm": 0.7621950507164001,
+      "learning_rate": 8.085057461229872e-05,
+      "loss": 0.4689,
+      "step": 6053
+    },
+    {
+      "epoch": 1.9438112056509873,
+      "grad_norm": 0.6205912828445435,
+      "learning_rate": 8.081631518513704e-05,
+      "loss": 0.339,
+      "step": 6054
+    },
+    {
+      "epoch": 1.9441322844758389,
+      "grad_norm": 0.9565242528915405,
+      "learning_rate": 8.078205809544917e-05,
+      "loss": 0.5811,
+      "step": 6055
+    },
+    {
+      "epoch": 1.9444533633006902,
+      "grad_norm": 1.0997697114944458,
+      "learning_rate": 8.074780334740928e-05,
+      "loss": 0.6262,
+      "step": 6056
+    },
+    {
+      "epoch": 1.9447744421255417,
+      "grad_norm": 0.8352198004722595,
+      "learning_rate": 8.071355094519109e-05,
+      "loss": 0.455,
+      "step": 6057
+    },
+    {
+      "epoch": 1.9450955209503933,
+      "grad_norm": 1.3071821928024292,
+      "learning_rate": 8.067930089296827e-05,
+      "loss": 0.6426,
+      "step": 6058
+    },
+    {
+      "epoch": 1.9454165997752448,
+      "grad_norm": 0.8504396677017212,
+      "learning_rate": 8.064505319491398e-05,
+      "loss": 0.4168,
+      "step": 6059
+    },
+    {
+      "epoch": 1.9457376786000964,
+      "grad_norm": 1.1026358604431152,
+      "learning_rate": 8.061080785520126e-05,
+      "loss": 0.5795,
+      "step": 6060
+    },
+    {
+      "epoch": 1.946058757424948,
+      "grad_norm": 0.9876211285591125,
+      "learning_rate": 8.057656487800282e-05,
+      "loss": 0.4293,
+      "step": 6061
+    },
+    {
+      "epoch": 1.9463798362497995,
+      "grad_norm": 0.8381431102752686,
+      "learning_rate": 8.0542324267491e-05,
+      "loss": 0.4613,
+      "step": 6062
+    },
+    {
+      "epoch": 1.9467009150746508,
+      "grad_norm": 0.9041491150856018,
+      "learning_rate": 8.050808602783795e-05,
+      "loss": 0.4021,
+      "step": 6063
+    },
+    {
+      "epoch": 1.9470219938995024,
+      "grad_norm": 0.8426972031593323,
+      "learning_rate": 8.047385016321552e-05,
+      "loss": 0.5205,
+      "step": 6064
+    },
+    {
+      "epoch": 1.9473430727243537,
+      "grad_norm": 0.5573900938034058,
+      "learning_rate": 8.04396166777952e-05,
+      "loss": 0.6376,
+      "step": 6065
+    },
+    {
+      "epoch": 1.9476641515492052,
+      "grad_norm": 0.5124462246894836,
+      "learning_rate": 8.040538557574822e-05,
+      "loss": 0.7874,
+      "step": 6066
+    },
+    {
+      "epoch": 1.9479852303740568,
+      "grad_norm": 0.7582365870475769,
+      "learning_rate": 8.037115686124564e-05,
+      "loss": 0.3713,
+      "step": 6067
+    },
+    {
+      "epoch": 1.9483063091989083,
+      "grad_norm": 1.0878230333328247,
+      "learning_rate": 8.033693053845801e-05,
+      "loss": 0.5994,
+      "step": 6068
+    },
+    {
+      "epoch": 1.94862738802376,
+      "grad_norm": 0.8553785681724548,
+      "learning_rate": 8.030270661155574e-05,
+      "loss": 0.2631,
+      "step": 6069
+    },
+    {
+      "epoch": 1.9489484668486114,
+      "grad_norm": 0.5939182043075562,
+      "learning_rate": 8.026848508470897e-05,
+      "loss": 0.2166,
+      "step": 6070
+    },
+    {
+      "epoch": 1.949269545673463,
+      "grad_norm": 0.5438021421432495,
+      "learning_rate": 8.023426596208739e-05,
+      "loss": 0.1312,
+      "step": 6071
+    },
+    {
+      "epoch": 1.9495906244983143,
+      "grad_norm": 0.6619899272918701,
+      "learning_rate": 8.020004924786059e-05,
+      "loss": 0.2189,
+      "step": 6072
+    },
+    {
+      "epoch": 1.9499117033231659,
+      "grad_norm": 0.8623759150505066,
+      "learning_rate": 8.016583494619769e-05,
+      "loss": 0.7344,
+      "step": 6073
+    },
+    {
+      "epoch": 1.9502327821480172,
+      "grad_norm": 0.8690182566642761,
+      "learning_rate": 8.013162306126765e-05,
+      "loss": 0.6706,
+      "step": 6074
+    },
+    {
+      "epoch": 1.9505538609728688,
+      "grad_norm": 0.8359203338623047,
+      "learning_rate": 8.009741359723906e-05,
+      "loss": 0.7241,
+      "step": 6075
+    },
+    {
+      "epoch": 1.9508749397977203,
+      "grad_norm": 0.9023215770721436,
+      "learning_rate": 8.00632065582803e-05,
+      "loss": 0.662,
+      "step": 6076
+    },
+    {
+      "epoch": 1.9511960186225719,
+      "grad_norm": 0.8437222242355347,
+      "learning_rate": 8.002900194855932e-05,
+      "loss": 0.4732,
+      "step": 6077
+    },
+    {
+      "epoch": 1.9515170974474234,
+      "grad_norm": 1.18110990524292,
+      "learning_rate": 7.999479977224384e-05,
+      "loss": 0.7354,
+      "step": 6078
+    },
+    {
+      "epoch": 1.951838176272275,
+      "grad_norm": 0.7391937971115112,
+      "learning_rate": 7.996060003350139e-05,
+      "loss": 0.5063,
+      "step": 6079
+    },
+    {
+      "epoch": 1.9521592550971263,
+      "grad_norm": 0.937690794467926,
+      "learning_rate": 7.992640273649898e-05,
+      "loss": 0.5532,
+      "step": 6080
+    },
+    {
+      "epoch": 1.9524803339219778,
+      "grad_norm": 0.7518649101257324,
+      "learning_rate": 7.989220788540355e-05,
+      "loss": 0.5798,
+      "step": 6081
+    },
+    {
+      "epoch": 1.9528014127468294,
+      "grad_norm": 1.227014422416687,
+      "learning_rate": 7.985801548438157e-05,
+      "loss": 0.5997,
+      "step": 6082
+    },
+    {
+      "epoch": 1.9531224915716807,
+      "grad_norm": 0.9930256009101868,
+      "learning_rate": 7.982382553759931e-05,
+      "loss": 0.7134,
+      "step": 6083
+    },
+    {
+      "epoch": 1.9534435703965323,
+      "grad_norm": 0.8083244562149048,
+      "learning_rate": 7.97896380492227e-05,
+      "loss": 0.5235,
+      "step": 6084
+    },
+    {
+      "epoch": 1.9537646492213838,
+      "grad_norm": 0.8794909119606018,
+      "learning_rate": 7.975545302341743e-05,
+      "loss": 0.6455,
+      "step": 6085
+    },
+    {
+      "epoch": 1.9540857280462354,
+      "grad_norm": 0.7878838181495667,
+      "learning_rate": 7.972127046434878e-05,
+      "loss": 0.5327,
+      "step": 6086
+    },
+    {
+      "epoch": 1.954406806871087,
+      "grad_norm": 0.861558198928833,
+      "learning_rate": 7.96870903761818e-05,
+      "loss": 0.4608,
+      "step": 6087
+    },
+    {
+      "epoch": 1.9547278856959385,
+      "grad_norm": 0.7832056283950806,
+      "learning_rate": 7.965291276308124e-05,
+      "loss": 0.5346,
+      "step": 6088
+    },
+    {
+      "epoch": 1.9550489645207898,
+      "grad_norm": 0.9764854907989502,
+      "learning_rate": 7.961873762921153e-05,
+      "loss": 0.5555,
+      "step": 6089
+    },
+    {
+      "epoch": 1.9553700433456414,
+      "grad_norm": 1.0838629007339478,
+      "learning_rate": 7.958456497873685e-05,
+      "loss": 0.6255,
+      "step": 6090
+    },
+    {
+      "epoch": 1.955691122170493,
+      "grad_norm": 1.2348774671554565,
+      "learning_rate": 7.955039481582097e-05,
+      "loss": 0.4626,
+      "step": 6091
+    },
+    {
+      "epoch": 1.9560122009953442,
+      "grad_norm": 0.8627070784568787,
+      "learning_rate": 7.951622714462746e-05,
+      "loss": 0.5652,
+      "step": 6092
+    },
+    {
+      "epoch": 1.9563332798201958,
+      "grad_norm": 1.0902684926986694,
+      "learning_rate": 7.948206196931954e-05,
+      "loss": 0.3824,
+      "step": 6093
+    },
+    {
+      "epoch": 1.9566543586450473,
+      "grad_norm": 0.9122470617294312,
+      "learning_rate": 7.944789929406016e-05,
+      "loss": 0.4996,
+      "step": 6094
+    },
+    {
+      "epoch": 1.9569754374698989,
+      "grad_norm": 0.9444713592529297,
+      "learning_rate": 7.941373912301189e-05,
+      "loss": 0.782,
+      "step": 6095
+    },
+    {
+      "epoch": 1.9572965162947504,
+      "grad_norm": 0.7925707697868347,
+      "learning_rate": 7.937958146033705e-05,
+      "loss": 0.5442,
+      "step": 6096
+    },
+    {
+      "epoch": 1.957617595119602,
+      "grad_norm": 0.7286838889122009,
+      "learning_rate": 7.934542631019768e-05,
+      "loss": 0.4491,
+      "step": 6097
+    },
+    {
+      "epoch": 1.9579386739444533,
+      "grad_norm": 0.904128909111023,
+      "learning_rate": 7.931127367675543e-05,
+      "loss": 0.6565,
+      "step": 6098
+    },
+    {
+      "epoch": 1.9582597527693049,
+      "grad_norm": 0.8943635821342468,
+      "learning_rate": 7.927712356417176e-05,
+      "loss": 0.5327,
+      "step": 6099
+    },
+    {
+      "epoch": 1.9585808315941564,
+      "grad_norm": 0.8871841430664062,
+      "learning_rate": 7.92429759766077e-05,
+      "loss": 0.5109,
+      "step": 6100
+    },
+    {
+      "epoch": 1.9589019104190077,
+      "grad_norm": 0.6855352520942688,
+      "learning_rate": 7.920883091822408e-05,
+      "loss": 0.4418,
+      "step": 6101
+    },
+    {
+      "epoch": 1.9592229892438593,
+      "grad_norm": 0.8009575009346008,
+      "learning_rate": 7.917468839318132e-05,
+      "loss": 0.4312,
+      "step": 6102
+    },
+    {
+      "epoch": 1.9595440680687108,
+      "grad_norm": 0.8051626682281494,
+      "learning_rate": 7.914054840563963e-05,
+      "loss": 0.4664,
+      "step": 6103
+    },
+    {
+      "epoch": 1.9598651468935624,
+      "grad_norm": 0.9816034436225891,
+      "learning_rate": 7.910641095975886e-05,
+      "loss": 0.5121,
+      "step": 6104
+    },
+    {
+      "epoch": 1.960186225718414,
+      "grad_norm": 1.6955353021621704,
+      "learning_rate": 7.907227605969849e-05,
+      "loss": 0.6544,
+      "step": 6105
+    },
+    {
+      "epoch": 1.9605073045432655,
+      "grad_norm": 0.8258131146430969,
+      "learning_rate": 7.903814370961784e-05,
+      "loss": 0.4505,
+      "step": 6106
+    },
+    {
+      "epoch": 1.9608283833681168,
+      "grad_norm": 1.0741841793060303,
+      "learning_rate": 7.900401391367576e-05,
+      "loss": 0.636,
+      "step": 6107
+    },
+    {
+      "epoch": 1.9611494621929684,
+      "grad_norm": 0.7029876112937927,
+      "learning_rate": 7.896988667603093e-05,
+      "loss": 0.4387,
+      "step": 6108
+    },
+    {
+      "epoch": 1.96147054101782,
+      "grad_norm": 0.8194742798805237,
+      "learning_rate": 7.893576200084159e-05,
+      "loss": 0.5288,
+      "step": 6109
+    },
+    {
+      "epoch": 1.9617916198426713,
+      "grad_norm": 0.9146851301193237,
+      "learning_rate": 7.89016398922658e-05,
+      "loss": 0.504,
+      "step": 6110
+    },
+    {
+      "epoch": 1.9621126986675228,
+      "grad_norm": 0.760689914226532,
+      "learning_rate": 7.886752035446114e-05,
+      "loss": 0.4793,
+      "step": 6111
+    },
+    {
+      "epoch": 1.9624337774923744,
+      "grad_norm": 0.5260710120201111,
+      "learning_rate": 7.883340339158505e-05,
+      "loss": 0.3192,
+      "step": 6112
+    },
+    {
+      "epoch": 1.962754856317226,
+      "grad_norm": 0.7600740194320679,
+      "learning_rate": 7.879928900779456e-05,
+      "loss": 0.3748,
+      "step": 6113
+    },
+    {
+      "epoch": 1.9630759351420775,
+      "grad_norm": 0.8149203658103943,
+      "learning_rate": 7.876517720724636e-05,
+      "loss": 0.5003,
+      "step": 6114
+    },
+    {
+      "epoch": 1.963397013966929,
+      "grad_norm": 0.630042314529419,
+      "learning_rate": 7.873106799409695e-05,
+      "loss": 0.6224,
+      "step": 6115
+    },
+    {
+      "epoch": 1.9637180927917803,
+      "grad_norm": 0.6109962463378906,
+      "learning_rate": 7.869696137250235e-05,
+      "loss": 0.6135,
+      "step": 6116
+    },
+    {
+      "epoch": 1.9640391716166319,
+      "grad_norm": 0.4882936179637909,
+      "learning_rate": 7.866285734661841e-05,
+      "loss": 0.3282,
+      "step": 6117
+    },
+    {
+      "epoch": 1.9643602504414834,
+      "grad_norm": 0.5604743957519531,
+      "learning_rate": 7.862875592060056e-05,
+      "loss": 0.2431,
+      "step": 6118
+    },
+    {
+      "epoch": 1.9646813292663348,
+      "grad_norm": 0.4672786593437195,
+      "learning_rate": 7.8594657098604e-05,
+      "loss": 0.183,
+      "step": 6119
+    },
+    {
+      "epoch": 1.9650024080911863,
+      "grad_norm": 0.4110962450504303,
+      "learning_rate": 7.856056088478352e-05,
+      "loss": 0.1327,
+      "step": 6120
+    },
+    {
+      "epoch": 1.9653234869160379,
+      "grad_norm": 0.5428380370140076,
+      "learning_rate": 7.852646728329368e-05,
+      "loss": 0.2487,
+      "step": 6121
+    },
+    {
+      "epoch": 1.9656445657408894,
+      "grad_norm": 0.7151157855987549,
+      "learning_rate": 7.849237629828869e-05,
+      "loss": 0.2885,
+      "step": 6122
+    },
+    {
+      "epoch": 1.965965644565741,
+      "grad_norm": 0.8936823606491089,
+      "learning_rate": 7.845828793392236e-05,
+      "loss": 0.7795,
+      "step": 6123
+    },
+    {
+      "epoch": 1.9662867233905925,
+      "grad_norm": 0.8891395330429077,
+      "learning_rate": 7.842420219434833e-05,
+      "loss": 0.4893,
+      "step": 6124
+    },
+    {
+      "epoch": 1.9666078022154438,
+      "grad_norm": 0.8358086943626404,
+      "learning_rate": 7.83901190837198e-05,
+      "loss": 0.7111,
+      "step": 6125
+    },
+    {
+      "epoch": 1.9669288810402954,
+      "grad_norm": 0.7215904593467712,
+      "learning_rate": 7.835603860618972e-05,
+      "loss": 0.402,
+      "step": 6126
+    },
+    {
+      "epoch": 1.967249959865147,
+      "grad_norm": 0.7132273316383362,
+      "learning_rate": 7.832196076591067e-05,
+      "loss": 0.4719,
+      "step": 6127
+    },
+    {
+      "epoch": 1.9675710386899983,
+      "grad_norm": 0.9818544387817383,
+      "learning_rate": 7.828788556703498e-05,
+      "loss": 0.8142,
+      "step": 6128
+    },
+    {
+      "epoch": 1.9678921175148498,
+      "grad_norm": 1.1569818258285522,
+      "learning_rate": 7.825381301371452e-05,
+      "loss": 0.5118,
+      "step": 6129
+    },
+    {
+      "epoch": 1.9682131963397014,
+      "grad_norm": 0.9389197826385498,
+      "learning_rate": 7.821974311010102e-05,
+      "loss": 0.6396,
+      "step": 6130
+    },
+    {
+      "epoch": 1.968534275164553,
+      "grad_norm": 1.039969563484192,
+      "learning_rate": 7.818567586034577e-05,
+      "loss": 0.7603,
+      "step": 6131
+    },
+    {
+      "epoch": 1.9688553539894045,
+      "grad_norm": 1.0185346603393555,
+      "learning_rate": 7.81516112685997e-05,
+      "loss": 0.6472,
+      "step": 6132
+    },
+    {
+      "epoch": 1.969176432814256,
+      "grad_norm": 0.9028285145759583,
+      "learning_rate": 7.811754933901358e-05,
+      "loss": 0.7387,
+      "step": 6133
+    },
+    {
+      "epoch": 1.9694975116391074,
+      "grad_norm": 1.1110644340515137,
+      "learning_rate": 7.808349007573763e-05,
+      "loss": 0.5489,
+      "step": 6134
+    },
+    {
+      "epoch": 1.969818590463959,
+      "grad_norm": 1.0398848056793213,
+      "learning_rate": 7.804943348292197e-05,
+      "loss": 0.7001,
+      "step": 6135
+    },
+    {
+      "epoch": 1.9701396692888105,
+      "grad_norm": 0.7271793484687805,
+      "learning_rate": 7.801537956471624e-05,
+      "loss": 0.5753,
+      "step": 6136
+    },
+    {
+      "epoch": 1.9704607481136618,
+      "grad_norm": 1.480493426322937,
+      "learning_rate": 7.798132832526986e-05,
+      "loss": 0.657,
+      "step": 6137
+    },
+    {
+      "epoch": 1.9707818269385133,
+      "grad_norm": 0.8976219892501831,
+      "learning_rate": 7.79472797687318e-05,
+      "loss": 0.5007,
+      "step": 6138
+    },
+    {
+      "epoch": 1.9711029057633649,
+      "grad_norm": 0.8589239716529846,
+      "learning_rate": 7.791323389925084e-05,
+      "loss": 0.603,
+      "step": 6139
+    },
+    {
+      "epoch": 1.9714239845882164,
+      "grad_norm": 0.8826812505722046,
+      "learning_rate": 7.787919072097531e-05,
+      "loss": 0.6224,
+      "step": 6140
+    },
+    {
+      "epoch": 1.971745063413068,
+      "grad_norm": 0.7189533114433289,
+      "learning_rate": 7.784515023805328e-05,
+      "loss": 0.4416,
+      "step": 6141
+    },
+    {
+      "epoch": 1.9720661422379195,
+      "grad_norm": 1.1884037256240845,
+      "learning_rate": 7.781111245463252e-05,
+      "loss": 0.8795,
+      "step": 6142
+    },
+    {
+      "epoch": 1.9723872210627709,
+      "grad_norm": 1.107159972190857,
+      "learning_rate": 7.777707737486037e-05,
+      "loss": 0.6781,
+      "step": 6143
+    },
+    {
+      "epoch": 1.9727082998876224,
+      "grad_norm": 0.7705017924308777,
+      "learning_rate": 7.774304500288394e-05,
+      "loss": 0.5458,
+      "step": 6144
+    },
+    {
+      "epoch": 1.9730293787124737,
+      "grad_norm": 0.8601257801055908,
+      "learning_rate": 7.770901534284995e-05,
+      "loss": 0.4366,
+      "step": 6145
+    },
+    {
+      "epoch": 1.9733504575373253,
+      "grad_norm": 0.6789402961730957,
+      "learning_rate": 7.767498839890488e-05,
+      "loss": 0.4362,
+      "step": 6146
+    },
+    {
+      "epoch": 1.9736715363621768,
+      "grad_norm": 0.7042021155357361,
+      "learning_rate": 7.76409641751947e-05,
+      "loss": 0.4543,
+      "step": 6147
+    },
+    {
+      "epoch": 1.9739926151870284,
+      "grad_norm": 0.9554926753044128,
+      "learning_rate": 7.760694267586525e-05,
+      "loss": 0.7007,
+      "step": 6148
+    },
+    {
+      "epoch": 1.97431369401188,
+      "grad_norm": 0.9321420192718506,
+      "learning_rate": 7.75729239050619e-05,
+      "loss": 0.6791,
+      "step": 6149
+    },
+    {
+      "epoch": 1.9746347728367315,
+      "grad_norm": 0.8884396553039551,
+      "learning_rate": 7.753890786692972e-05,
+      "loss": 0.4627,
+      "step": 6150
+    },
+    {
+      "epoch": 1.974955851661583,
+      "grad_norm": 0.805990993976593,
+      "learning_rate": 7.750489456561352e-05,
+      "loss": 0.4353,
+      "step": 6151
+    },
+    {
+      "epoch": 1.9752769304864344,
+      "grad_norm": 0.9722745418548584,
+      "learning_rate": 7.747088400525766e-05,
+      "loss": 0.5465,
+      "step": 6152
+    },
+    {
+      "epoch": 1.975598009311286,
+      "grad_norm": 0.662632405757904,
+      "learning_rate": 7.743687619000626e-05,
+      "loss": 0.4702,
+      "step": 6153
+    },
+    {
+      "epoch": 1.9759190881361373,
+      "grad_norm": 0.8062264919281006,
+      "learning_rate": 7.740287112400303e-05,
+      "loss": 0.5142,
+      "step": 6154
+    },
+    {
+      "epoch": 1.9762401669609888,
+      "grad_norm": 0.8860646486282349,
+      "learning_rate": 7.736886881139142e-05,
+      "loss": 0.4718,
+      "step": 6155
+    },
+    {
+      "epoch": 1.9765612457858404,
+      "grad_norm": 0.928923487663269,
+      "learning_rate": 7.733486925631447e-05,
+      "loss": 0.5866,
+      "step": 6156
+    },
+    {
+      "epoch": 1.976882324610692,
+      "grad_norm": 0.9381113648414612,
+      "learning_rate": 7.730087246291502e-05,
+      "loss": 0.4805,
+      "step": 6157
+    },
+    {
+      "epoch": 1.9772034034355435,
+      "grad_norm": 0.8858740329742432,
+      "learning_rate": 7.726687843533538e-05,
+      "loss": 0.5572,
+      "step": 6158
+    },
+    {
+      "epoch": 1.977524482260395,
+      "grad_norm": 0.9326648712158203,
+      "learning_rate": 7.723288717771761e-05,
+      "loss": 0.5227,
+      "step": 6159
+    },
+    {
+      "epoch": 1.9778455610852466,
+      "grad_norm": 0.9480475187301636,
+      "learning_rate": 7.719889869420353e-05,
+      "loss": 0.4524,
+      "step": 6160
+    },
+    {
+      "epoch": 1.978166639910098,
+      "grad_norm": 0.7123903632164001,
+      "learning_rate": 7.716491298893442e-05,
+      "loss": 0.4944,
+      "step": 6161
+    },
+    {
+      "epoch": 1.9784877187349494,
+      "grad_norm": 0.5559878349304199,
+      "learning_rate": 7.713093006605145e-05,
+      "loss": 0.3082,
+      "step": 6162
+    },
+    {
+      "epoch": 1.9788087975598008,
+      "grad_norm": 0.6076563000679016,
+      "learning_rate": 7.709694992969526e-05,
+      "loss": 0.4055,
+      "step": 6163
+    },
+    {
+      "epoch": 1.9791298763846523,
+      "grad_norm": 0.5390298366546631,
+      "learning_rate": 7.706297258400624e-05,
+      "loss": 0.3068,
+      "step": 6164
+    },
+    {
+      "epoch": 1.9794509552095039,
+      "grad_norm": 0.533420205116272,
+      "learning_rate": 7.702899803312443e-05,
+      "loss": 0.5381,
+      "step": 6165
+    },
+    {
+      "epoch": 1.9797720340343554,
+      "grad_norm": 0.5672399997711182,
+      "learning_rate": 7.699502628118958e-05,
+      "loss": 0.9303,
+      "step": 6166
+    },
+    {
+      "epoch": 1.980093112859207,
+      "grad_norm": 0.7802035212516785,
+      "learning_rate": 7.696105733234098e-05,
+      "loss": 0.5104,
+      "step": 6167
+    },
+    {
+      "epoch": 1.9804141916840585,
+      "grad_norm": 0.6914035081863403,
+      "learning_rate": 7.692709119071762e-05,
+      "loss": 0.5653,
+      "step": 6168
+    },
+    {
+      "epoch": 1.98073527050891,
+      "grad_norm": 0.6555715203285217,
+      "learning_rate": 7.689312786045823e-05,
+      "loss": 0.3267,
+      "step": 6169
+    },
+    {
+      "epoch": 1.9810563493337614,
+      "grad_norm": 0.8108289241790771,
+      "learning_rate": 7.685916734570112e-05,
+      "loss": 0.2342,
+      "step": 6170
+    },
+    {
+      "epoch": 1.981377428158613,
+      "grad_norm": 0.6541457772254944,
+      "learning_rate": 7.682520965058428e-05,
+      "loss": 0.4736,
+      "step": 6171
+    },
+    {
+      "epoch": 1.9816985069834643,
+      "grad_norm": 0.9171504378318787,
+      "learning_rate": 7.679125477924534e-05,
+      "loss": 0.6021,
+      "step": 6172
+    },
+    {
+      "epoch": 1.9820195858083158,
+      "grad_norm": 0.798708975315094,
+      "learning_rate": 7.67573027358216e-05,
+      "loss": 0.537,
+      "step": 6173
+    },
+    {
+      "epoch": 1.9823406646331674,
+      "grad_norm": 0.8044041991233826,
+      "learning_rate": 7.672335352445002e-05,
+      "loss": 0.5214,
+      "step": 6174
+    },
+    {
+      "epoch": 1.982661743458019,
+      "grad_norm": 0.9100499153137207,
+      "learning_rate": 7.668940714926725e-05,
+      "loss": 0.6087,
+      "step": 6175
+    },
+    {
+      "epoch": 1.9829828222828705,
+      "grad_norm": 0.7136359810829163,
+      "learning_rate": 7.66554636144095e-05,
+      "loss": 0.4973,
+      "step": 6176
+    },
+    {
+      "epoch": 1.983303901107722,
+      "grad_norm": 0.6627728343009949,
+      "learning_rate": 7.662152292401264e-05,
+      "loss": 0.4051,
+      "step": 6177
+    },
+    {
+      "epoch": 1.9836249799325736,
+      "grad_norm": 0.6324448585510254,
+      "learning_rate": 7.658758508221234e-05,
+      "loss": 0.4427,
+      "step": 6178
+    },
+    {
+      "epoch": 1.983946058757425,
+      "grad_norm": 0.8954194188117981,
+      "learning_rate": 7.655365009314374e-05,
+      "loss": 0.6454,
+      "step": 6179
+    },
+    {
+      "epoch": 1.9842671375822765,
+      "grad_norm": 0.952394425868988,
+      "learning_rate": 7.651971796094183e-05,
+      "loss": 0.5537,
+      "step": 6180
+    },
+    {
+      "epoch": 1.9845882164071278,
+      "grad_norm": 0.9424409866333008,
+      "learning_rate": 7.6485788689741e-05,
+      "loss": 0.5597,
+      "step": 6181
+    },
+    {
+      "epoch": 1.9849092952319793,
+      "grad_norm": 1.0070812702178955,
+      "learning_rate": 7.645186228367554e-05,
+      "loss": 0.7455,
+      "step": 6182
+    },
+    {
+      "epoch": 1.985230374056831,
+      "grad_norm": 0.8608502149581909,
+      "learning_rate": 7.641793874687918e-05,
+      "loss": 0.678,
+      "step": 6183
+    },
+    {
+      "epoch": 1.9855514528816824,
+      "grad_norm": 1.0382908582687378,
+      "learning_rate": 7.638401808348548e-05,
+      "loss": 0.6999,
+      "step": 6184
+    },
+    {
+      "epoch": 1.985872531706534,
+      "grad_norm": 1.0702378749847412,
+      "learning_rate": 7.635010029762756e-05,
+      "loss": 0.8023,
+      "step": 6185
+    },
+    {
+      "epoch": 1.9861936105313855,
+      "grad_norm": 0.9120118618011475,
+      "learning_rate": 7.631618539343814e-05,
+      "loss": 0.6575,
+      "step": 6186
+    },
+    {
+      "epoch": 1.986514689356237,
+      "grad_norm": 0.8130465149879456,
+      "learning_rate": 7.628227337504972e-05,
+      "loss": 0.5327,
+      "step": 6187
+    },
+    {
+      "epoch": 1.9868357681810884,
+      "grad_norm": 1.0626096725463867,
+      "learning_rate": 7.62483642465943e-05,
+      "loss": 0.6589,
+      "step": 6188
+    },
+    {
+      "epoch": 1.98715684700594,
+      "grad_norm": 1.0040377378463745,
+      "learning_rate": 7.621445801220371e-05,
+      "loss": 0.6529,
+      "step": 6189
+    },
+    {
+      "epoch": 1.9874779258307913,
+      "grad_norm": 0.877011775970459,
+      "learning_rate": 7.618055467600922e-05,
+      "loss": 0.5563,
+      "step": 6190
+    },
+    {
+      "epoch": 1.9877990046556429,
+      "grad_norm": 0.8776607513427734,
+      "learning_rate": 7.614665424214193e-05,
+      "loss": 0.6501,
+      "step": 6191
+    },
+    {
+      "epoch": 1.9881200834804944,
+      "grad_norm": 1.0523321628570557,
+      "learning_rate": 7.611275671473245e-05,
+      "loss": 0.6605,
+      "step": 6192
+    },
+    {
+      "epoch": 1.988441162305346,
+      "grad_norm": 0.7821586728096008,
+      "learning_rate": 7.607886209791107e-05,
+      "loss": 0.5578,
+      "step": 6193
+    },
+    {
+      "epoch": 1.9887622411301975,
+      "grad_norm": 0.8066548109054565,
+      "learning_rate": 7.604497039580785e-05,
+      "loss": 0.5154,
+      "step": 6194
+    },
+    {
+      "epoch": 1.989083319955049,
+      "grad_norm": 1.202035665512085,
+      "learning_rate": 7.601108161255226e-05,
+      "loss": 0.5195,
+      "step": 6195
+    },
+    {
+      "epoch": 1.9894043987799006,
+      "grad_norm": 0.838572084903717,
+      "learning_rate": 7.597719575227364e-05,
+      "loss": 0.5041,
+      "step": 6196
+    },
+    {
+      "epoch": 1.989725477604752,
+      "grad_norm": 0.9432242512702942,
+      "learning_rate": 7.594331281910082e-05,
+      "loss": 0.6335,
+      "step": 6197
+    },
+    {
+      "epoch": 1.9900465564296035,
+      "grad_norm": 0.9394798278808594,
+      "learning_rate": 7.590943281716241e-05,
+      "loss": 0.4824,
+      "step": 6198
+    },
+    {
+      "epoch": 1.9903676352544548,
+      "grad_norm": 0.6274194717407227,
+      "learning_rate": 7.587555575058649e-05,
+      "loss": 0.377,
+      "step": 6199
+    },
+    {
+      "epoch": 1.9906887140793064,
+      "grad_norm": 1.0759146213531494,
+      "learning_rate": 7.584168162350098e-05,
+      "loss": 0.5105,
+      "step": 6200
+    },
+    {
+      "epoch": 1.991009792904158,
+      "grad_norm": 0.6160650253295898,
+      "learning_rate": 7.580781044003324e-05,
+      "loss": 0.4023,
+      "step": 6201
+    },
+    {
+      "epoch": 1.9913308717290095,
+      "grad_norm": 0.6983049511909485,
+      "learning_rate": 7.577394220431042e-05,
+      "loss": 0.3685,
+      "step": 6202
+    },
+    {
+      "epoch": 1.991651950553861,
+      "grad_norm": 0.6337112784385681,
+      "learning_rate": 7.574007692045928e-05,
+      "loss": 0.41,
+      "step": 6203
+    },
+    {
+      "epoch": 1.9919730293787126,
+      "grad_norm": 0.9221318960189819,
+      "learning_rate": 7.570621459260615e-05,
+      "loss": 0.5264,
+      "step": 6204
+    },
+    {
+      "epoch": 1.9922941082035641,
+      "grad_norm": 0.8658796548843384,
+      "learning_rate": 7.567235522487712e-05,
+      "loss": 0.5192,
+      "step": 6205
+    },
+    {
+      "epoch": 1.9926151870284154,
+      "grad_norm": 0.8173865079879761,
+      "learning_rate": 7.563849882139776e-05,
+      "loss": 0.464,
+      "step": 6206
+    },
+    {
+      "epoch": 1.992936265853267,
+      "grad_norm": 0.7882019877433777,
+      "learning_rate": 7.560464538629344e-05,
+      "loss": 0.4191,
+      "step": 6207
+    },
+    {
+      "epoch": 1.9932573446781183,
+      "grad_norm": 1.4528001546859741,
+      "learning_rate": 7.557079492368909e-05,
+      "loss": 0.3734,
+      "step": 6208
+    },
+    {
+      "epoch": 1.9935784235029699,
+      "grad_norm": 1.0732624530792236,
+      "learning_rate": 7.553694743770928e-05,
+      "loss": 0.5081,
+      "step": 6209
+    },
+    {
+      "epoch": 1.9938995023278214,
+      "grad_norm": 0.9734475612640381,
+      "learning_rate": 7.550310293247823e-05,
+      "loss": 0.4522,
+      "step": 6210
+    },
+    {
+      "epoch": 1.994220581152673,
+      "grad_norm": 1.118553638458252,
+      "learning_rate": 7.546926141211974e-05,
+      "loss": 0.5817,
+      "step": 6211
+    },
+    {
+      "epoch": 1.9945416599775245,
+      "grad_norm": 0.7407261729240417,
+      "learning_rate": 7.543542288075739e-05,
+      "loss": 0.3776,
+      "step": 6212
+    },
+    {
+      "epoch": 1.994862738802376,
+      "grad_norm": 0.7426897287368774,
+      "learning_rate": 7.54015873425142e-05,
+      "loss": 0.3893,
+      "step": 6213
+    },
+    {
+      "epoch": 1.9951838176272276,
+      "grad_norm": 0.4387853443622589,
+      "learning_rate": 7.536775480151303e-05,
+      "loss": 0.2971,
+      "step": 6214
+    },
+    {
+      "epoch": 1.995504896452079,
+      "grad_norm": 0.4856942594051361,
+      "learning_rate": 7.533392526187617e-05,
+      "loss": 0.659,
+      "step": 6215
+    },
+    {
+      "epoch": 1.9958259752769305,
+      "grad_norm": 0.49128448963165283,
+      "learning_rate": 7.530009872772572e-05,
+      "loss": 0.6117,
+      "step": 6216
+    },
+    {
+      "epoch": 1.9961470541017818,
+      "grad_norm": 0.819917619228363,
+      "learning_rate": 7.526627520318329e-05,
+      "loss": 0.6772,
+      "step": 6217
+    },
+    {
+      "epoch": 1.9964681329266334,
+      "grad_norm": 0.7481715083122253,
+      "learning_rate": 7.523245469237026e-05,
+      "loss": 0.4364,
+      "step": 6218
+    },
+    {
+      "epoch": 1.996789211751485,
+      "grad_norm": 0.9296541810035706,
+      "learning_rate": 7.519863719940748e-05,
+      "loss": 0.5882,
+      "step": 6219
+    },
+    {
+      "epoch": 1.9971102905763365,
+      "grad_norm": 0.8651626706123352,
+      "learning_rate": 7.516482272841549e-05,
+      "loss": 0.6071,
+      "step": 6220
+    },
+    {
+      "epoch": 1.997431369401188,
+      "grad_norm": 0.8119394183158875,
+      "learning_rate": 7.513101128351454e-05,
+      "loss": 0.4912,
+      "step": 6221
+    },
+    {
+      "epoch": 1.9977524482260396,
+      "grad_norm": 1.2624839544296265,
+      "learning_rate": 7.50972028688244e-05,
+      "loss": 0.5303,
+      "step": 6222
+    },
+    {
+      "epoch": 1.9980735270508911,
+      "grad_norm": 1.1252477169036865,
+      "learning_rate": 7.506339748846461e-05,
+      "loss": 0.7267,
+      "step": 6223
+    },
+    {
+      "epoch": 1.9983946058757425,
+      "grad_norm": 0.8501682877540588,
+      "learning_rate": 7.502959514655414e-05,
+      "loss": 0.6461,
+      "step": 6224
+    },
+    {
+      "epoch": 1.998715684700594,
+      "grad_norm": 0.9397971034049988,
+      "learning_rate": 7.499579584721179e-05,
+      "loss": 0.491,
+      "step": 6225
+    },
+    {
+      "epoch": 1.9990367635254453,
+      "grad_norm": 0.8632334470748901,
+      "learning_rate": 7.496199959455584e-05,
+      "loss": 0.5389,
+      "step": 6226
+    },
+    {
+      "epoch": 1.999357842350297,
+      "grad_norm": 1.034378170967102,
+      "learning_rate": 7.492820639270434e-05,
+      "loss": 0.4877,
+      "step": 6227
+    },
+    {
+      "epoch": 1.9996789211751484,
+      "grad_norm": 0.5917540192604065,
+      "learning_rate": 7.489441624577485e-05,
+      "loss": 0.3548,
+      "step": 6228
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.1308932304382324,
+      "learning_rate": 7.486062915788452e-05,
+      "loss": 0.6238,
+      "step": 6229
+    },
+    {
+      "epoch": 2.0003210788248516,
+      "grad_norm": 0.3727356791496277,
+      "learning_rate": 7.48268451331503e-05,
+      "loss": 0.5456,
+      "step": 6230
+    },
+    {
+      "epoch": 2.000642157649703,
+      "grad_norm": 0.40441709756851196,
+      "learning_rate": 7.479306417568864e-05,
+      "loss": 0.1526,
+      "step": 6231
+    },
+    {
+      "epoch": 2.0009632364745547,
+      "grad_norm": 0.27756381034851074,
+      "learning_rate": 7.475928628961566e-05,
+      "loss": 0.0981,
+      "step": 6232
+    },
+    {
+      "epoch": 2.001284315299406,
+      "grad_norm": 0.480465292930603,
+      "learning_rate": 7.472551147904708e-05,
+      "loss": 0.126,
+      "step": 6233
+    },
+    {
+      "epoch": 2.0016053941242573,
+      "grad_norm": 0.26273319125175476,
+      "learning_rate": 7.469173974809826e-05,
+      "loss": 0.0916,
+      "step": 6234
+    },
+    {
+      "epoch": 2.001926472949109,
+      "grad_norm": 0.4176071286201477,
+      "learning_rate": 7.465797110088417e-05,
+      "loss": 0.2166,
+      "step": 6235
+    },
+    {
+      "epoch": 2.0022475517739604,
+      "grad_norm": 0.5277458429336548,
+      "learning_rate": 7.462420554151944e-05,
+      "loss": 0.3709,
+      "step": 6236
+    },
+    {
+      "epoch": 2.002568630598812,
+      "grad_norm": 0.7284386157989502,
+      "learning_rate": 7.459044307411832e-05,
+      "loss": 0.4687,
+      "step": 6237
+    },
+    {
+      "epoch": 2.0028897094236635,
+      "grad_norm": 0.6903724074363708,
+      "learning_rate": 7.45566837027946e-05,
+      "loss": 0.4989,
+      "step": 6238
+    },
+    {
+      "epoch": 2.003210788248515,
+      "grad_norm": 0.7660374641418457,
+      "learning_rate": 7.45229274316618e-05,
+      "loss": 0.4508,
+      "step": 6239
+    },
+    {
+      "epoch": 2.0035318670733666,
+      "grad_norm": 0.5654639005661011,
+      "learning_rate": 7.448917426483299e-05,
+      "loss": 0.39,
+      "step": 6240
+    },
+    {
+      "epoch": 2.003852945898218,
+      "grad_norm": 0.6629282236099243,
+      "learning_rate": 7.445542420642097e-05,
+      "loss": 0.3377,
+      "step": 6241
+    },
+    {
+      "epoch": 2.0041740247230697,
+      "grad_norm": 0.7881885170936584,
+      "learning_rate": 7.442167726053797e-05,
+      "loss": 0.3575,
+      "step": 6242
+    },
+    {
+      "epoch": 2.004495103547921,
+      "grad_norm": 0.6721541881561279,
+      "learning_rate": 7.438793343129605e-05,
+      "loss": 0.3635,
+      "step": 6243
+    },
+    {
+      "epoch": 2.0048161823727724,
+      "grad_norm": 0.8870039582252502,
+      "learning_rate": 7.435419272280672e-05,
+      "loss": 0.4121,
+      "step": 6244
+    },
+    {
+      "epoch": 2.005137261197624,
+      "grad_norm": 0.9872992038726807,
+      "learning_rate": 7.432045513918122e-05,
+      "loss": 0.4399,
+      "step": 6245
+    },
+    {
+      "epoch": 2.0054583400224755,
+      "grad_norm": 0.9258590340614319,
+      "learning_rate": 7.42867206845304e-05,
+      "loss": 0.3429,
+      "step": 6246
+    },
+    {
+      "epoch": 2.005779418847327,
+      "grad_norm": 0.7288257479667664,
+      "learning_rate": 7.425298936296463e-05,
+      "loss": 0.3793,
+      "step": 6247
+    },
+    {
+      "epoch": 2.0061004976721786,
+      "grad_norm": 0.6890774965286255,
+      "learning_rate": 7.421926117859403e-05,
+      "loss": 0.3466,
+      "step": 6248
+    },
+    {
+      "epoch": 2.00642157649703,
+      "grad_norm": 0.7121484875679016,
+      "learning_rate": 7.418553613552825e-05,
+      "loss": 0.3155,
+      "step": 6249
+    },
+    {
+      "epoch": 2.0067426553218817,
+      "grad_norm": 0.5670762658119202,
+      "learning_rate": 7.415181423787659e-05,
+      "loss": 0.2762,
+      "step": 6250
+    },
+    {
+      "epoch": 2.0070637341467332,
+      "grad_norm": 0.684097945690155,
+      "learning_rate": 7.411809548974792e-05,
+      "loss": 0.3319,
+      "step": 6251
+    },
+    {
+      "epoch": 2.0073848129715843,
+      "grad_norm": 0.8124287724494934,
+      "learning_rate": 7.408437989525085e-05,
+      "loss": 0.3208,
+      "step": 6252
+    },
+    {
+      "epoch": 2.007705891796436,
+      "grad_norm": 0.9157565236091614,
+      "learning_rate": 7.405066745849346e-05,
+      "loss": 0.4214,
+      "step": 6253
+    },
+    {
+      "epoch": 2.0080269706212874,
+      "grad_norm": 0.7512825131416321,
+      "learning_rate": 7.401695818358353e-05,
+      "loss": 0.2616,
+      "step": 6254
+    },
+    {
+      "epoch": 2.008348049446139,
+      "grad_norm": 0.834835410118103,
+      "learning_rate": 7.398325207462846e-05,
+      "loss": 0.3358,
+      "step": 6255
+    },
+    {
+      "epoch": 2.0086691282709905,
+      "grad_norm": 1.0159118175506592,
+      "learning_rate": 7.394954913573517e-05,
+      "loss": 0.4667,
+      "step": 6256
+    },
+    {
+      "epoch": 2.008990207095842,
+      "grad_norm": 0.9639686942100525,
+      "learning_rate": 7.391584937101033e-05,
+      "loss": 0.2684,
+      "step": 6257
+    },
+    {
+      "epoch": 2.0093112859206936,
+      "grad_norm": 0.9616678357124329,
+      "learning_rate": 7.38821527845601e-05,
+      "loss": 0.4674,
+      "step": 6258
+    },
+    {
+      "epoch": 2.009632364745545,
+      "grad_norm": 1.6692441701889038,
+      "learning_rate": 7.384845938049031e-05,
+      "loss": 0.4352,
+      "step": 6259
+    },
+    {
+      "epoch": 2.0099534435703967,
+      "grad_norm": 0.8217843770980835,
+      "learning_rate": 7.381476916290644e-05,
+      "loss": 0.3344,
+      "step": 6260
+    },
+    {
+      "epoch": 2.010274522395248,
+      "grad_norm": 1.3268312215805054,
+      "learning_rate": 7.378108213591355e-05,
+      "loss": 0.5378,
+      "step": 6261
+    },
+    {
+      "epoch": 2.0105956012200994,
+      "grad_norm": 0.9014345407485962,
+      "learning_rate": 7.374739830361621e-05,
+      "loss": 0.3766,
+      "step": 6262
+    },
+    {
+      "epoch": 2.010916680044951,
+      "grad_norm": 0.8533119559288025,
+      "learning_rate": 7.37137176701188e-05,
+      "loss": 0.2654,
+      "step": 6263
+    },
+    {
+      "epoch": 2.0112377588698025,
+      "grad_norm": 0.9980586171150208,
+      "learning_rate": 7.368004023952517e-05,
+      "loss": 0.3431,
+      "step": 6264
+    },
+    {
+      "epoch": 2.011558837694654,
+      "grad_norm": 0.7079142332077026,
+      "learning_rate": 7.364636601593875e-05,
+      "loss": 0.2977,
+      "step": 6265
+    },
+    {
+      "epoch": 2.0118799165195056,
+      "grad_norm": 0.8518555164337158,
+      "learning_rate": 7.361269500346274e-05,
+      "loss": 0.3331,
+      "step": 6266
+    },
+    {
+      "epoch": 2.012200995344357,
+      "grad_norm": 0.9751218557357788,
+      "learning_rate": 7.357902720619976e-05,
+      "loss": 0.324,
+      "step": 6267
+    },
+    {
+      "epoch": 2.0125220741692087,
+      "grad_norm": 1.0797030925750732,
+      "learning_rate": 7.354536262825219e-05,
+      "loss": 0.399,
+      "step": 6268
+    },
+    {
+      "epoch": 2.0128431529940602,
+      "grad_norm": 1.0779436826705933,
+      "learning_rate": 7.351170127372191e-05,
+      "loss": 0.3863,
+      "step": 6269
+    },
+    {
+      "epoch": 2.0131642318189114,
+      "grad_norm": 0.9085304737091064,
+      "learning_rate": 7.347804314671055e-05,
+      "loss": 0.3424,
+      "step": 6270
+    },
+    {
+      "epoch": 2.013485310643763,
+      "grad_norm": 0.8026139140129089,
+      "learning_rate": 7.344438825131911e-05,
+      "loss": 0.3144,
+      "step": 6271
+    },
+    {
+      "epoch": 2.0138063894686145,
+      "grad_norm": 0.9833243489265442,
+      "learning_rate": 7.341073659164848e-05,
+      "loss": 0.3074,
+      "step": 6272
+    },
+    {
+      "epoch": 2.014127468293466,
+      "grad_norm": 1.829190969467163,
+      "learning_rate": 7.33770881717989e-05,
+      "loss": 0.3371,
+      "step": 6273
+    },
+    {
+      "epoch": 2.0144485471183176,
+      "grad_norm": 0.8061204552650452,
+      "learning_rate": 7.334344299587035e-05,
+      "loss": 0.3001,
+      "step": 6274
+    },
+    {
+      "epoch": 2.014769625943169,
+      "grad_norm": 0.5113372206687927,
+      "learning_rate": 7.330980106796246e-05,
+      "loss": 0.2632,
+      "step": 6275
+    },
+    {
+      "epoch": 2.0150907047680207,
+      "grad_norm": 0.5458338260650635,
+      "learning_rate": 7.327616239217431e-05,
+      "loss": 0.279,
+      "step": 6276
+    },
+    {
+      "epoch": 2.015411783592872,
+      "grad_norm": 0.7398447394371033,
+      "learning_rate": 7.324252697260474e-05,
+      "loss": 0.2997,
+      "step": 6277
+    },
+    {
+      "epoch": 2.0157328624177238,
+      "grad_norm": 0.7369372248649597,
+      "learning_rate": 7.320889481335207e-05,
+      "loss": 0.2572,
+      "step": 6278
+    },
+    {
+      "epoch": 2.016053941242575,
+      "grad_norm": 0.4629496932029724,
+      "learning_rate": 7.317526591851433e-05,
+      "loss": 0.2779,
+      "step": 6279
+    },
+    {
+      "epoch": 2.0163750200674264,
+      "grad_norm": 0.5169734954833984,
+      "learning_rate": 7.314164029218904e-05,
+      "loss": 0.3046,
+      "step": 6280
+    },
+    {
+      "epoch": 2.016696098892278,
+      "grad_norm": 0.6138988733291626,
+      "learning_rate": 7.310801793847344e-05,
+      "loss": 0.6675,
+      "step": 6281
+    },
+    {
+      "epoch": 2.0170171777171295,
+      "grad_norm": 0.6308935284614563,
+      "learning_rate": 7.307439886146428e-05,
+      "loss": 0.1614,
+      "step": 6282
+    },
+    {
+      "epoch": 2.017338256541981,
+      "grad_norm": 0.7147278785705566,
+      "learning_rate": 7.30407830652579e-05,
+      "loss": 0.2836,
+      "step": 6283
+    },
+    {
+      "epoch": 2.0176593353668326,
+      "grad_norm": 0.5563004612922668,
+      "learning_rate": 7.300717055395039e-05,
+      "loss": 0.1,
+      "step": 6284
+    },
+    {
+      "epoch": 2.017980414191684,
+      "grad_norm": 0.6495770812034607,
+      "learning_rate": 7.297356133163721e-05,
+      "loss": 0.1254,
+      "step": 6285
+    },
+    {
+      "epoch": 2.0183014930165357,
+      "grad_norm": 0.6263350248336792,
+      "learning_rate": 7.293995540241366e-05,
+      "loss": 0.2167,
+      "step": 6286
+    },
+    {
+      "epoch": 2.0186225718413873,
+      "grad_norm": 0.7048402428627014,
+      "learning_rate": 7.290635277037442e-05,
+      "loss": 0.3179,
+      "step": 6287
+    },
+    {
+      "epoch": 2.0189436506662384,
+      "grad_norm": 0.934867799282074,
+      "learning_rate": 7.287275343961392e-05,
+      "loss": 0.4398,
+      "step": 6288
+    },
+    {
+      "epoch": 2.01926472949109,
+      "grad_norm": 0.650744616985321,
+      "learning_rate": 7.283915741422612e-05,
+      "loss": 0.4162,
+      "step": 6289
+    },
+    {
+      "epoch": 2.0195858083159415,
+      "grad_norm": 0.9540171027183533,
+      "learning_rate": 7.280556469830464e-05,
+      "loss": 0.4748,
+      "step": 6290
+    },
+    {
+      "epoch": 2.019906887140793,
+      "grad_norm": 0.6486244201660156,
+      "learning_rate": 7.277197529594257e-05,
+      "loss": 0.3104,
+      "step": 6291
+    },
+    {
+      "epoch": 2.0202279659656446,
+      "grad_norm": 0.7935119867324829,
+      "learning_rate": 7.273838921123272e-05,
+      "loss": 0.3595,
+      "step": 6292
+    },
+    {
+      "epoch": 2.020549044790496,
+      "grad_norm": 0.6961763501167297,
+      "learning_rate": 7.270480644826749e-05,
+      "loss": 0.2754,
+      "step": 6293
+    },
+    {
+      "epoch": 2.0208701236153477,
+      "grad_norm": 0.8212137818336487,
+      "learning_rate": 7.267122701113876e-05,
+      "loss": 0.3855,
+      "step": 6294
+    },
+    {
+      "epoch": 2.0211912024401992,
+      "grad_norm": 0.6237925887107849,
+      "learning_rate": 7.263765090393817e-05,
+      "loss": 0.2879,
+      "step": 6295
+    },
+    {
+      "epoch": 2.021512281265051,
+      "grad_norm": 0.5799258947372437,
+      "learning_rate": 7.260407813075676e-05,
+      "loss": 0.29,
+      "step": 6296
+    },
+    {
+      "epoch": 2.021833360089902,
+      "grad_norm": 0.8160949349403381,
+      "learning_rate": 7.257050869568535e-05,
+      "loss": 0.4371,
+      "step": 6297
+    },
+    {
+      "epoch": 2.0221544389147534,
+      "grad_norm": 0.9621134400367737,
+      "learning_rate": 7.253694260281425e-05,
+      "loss": 0.492,
+      "step": 6298
+    },
+    {
+      "epoch": 2.022475517739605,
+      "grad_norm": 0.7058377861976624,
+      "learning_rate": 7.250337985623342e-05,
+      "loss": 0.299,
+      "step": 6299
+    },
+    {
+      "epoch": 2.0227965965644565,
+      "grad_norm": 0.7675186991691589,
+      "learning_rate": 7.246982046003234e-05,
+      "loss": 0.2944,
+      "step": 6300
+    },
+    {
+      "epoch": 2.023117675389308,
+      "grad_norm": 0.910317063331604,
+      "learning_rate": 7.243626441830009e-05,
+      "loss": 0.3878,
+      "step": 6301
+    },
+    {
+      "epoch": 2.0234387542141596,
+      "grad_norm": 0.9493252038955688,
+      "learning_rate": 7.240271173512546e-05,
+      "loss": 0.3264,
+      "step": 6302
+    },
+    {
+      "epoch": 2.023759833039011,
+      "grad_norm": 1.0064270496368408,
+      "learning_rate": 7.236916241459663e-05,
+      "loss": 0.3125,
+      "step": 6303
+    },
+    {
+      "epoch": 2.0240809118638627,
+      "grad_norm": 0.7101002335548401,
+      "learning_rate": 7.233561646080161e-05,
+      "loss": 0.2849,
+      "step": 6304
+    },
+    {
+      "epoch": 2.0244019906887143,
+      "grad_norm": 0.728187084197998,
+      "learning_rate": 7.230207387782776e-05,
+      "loss": 0.2976,
+      "step": 6305
+    },
+    {
+      "epoch": 2.0247230695135654,
+      "grad_norm": 0.8924553394317627,
+      "learning_rate": 7.226853466976222e-05,
+      "loss": 0.4461,
+      "step": 6306
+    },
+    {
+      "epoch": 2.025044148338417,
+      "grad_norm": 0.7692164182662964,
+      "learning_rate": 7.22349988406916e-05,
+      "loss": 0.3043,
+      "step": 6307
+    },
+    {
+      "epoch": 2.0253652271632685,
+      "grad_norm": 0.7273018956184387,
+      "learning_rate": 7.220146639470218e-05,
+      "loss": 0.321,
+      "step": 6308
+    },
+    {
+      "epoch": 2.02568630598812,
+      "grad_norm": 0.9611377716064453,
+      "learning_rate": 7.216793733587976e-05,
+      "loss": 0.2974,
+      "step": 6309
+    },
+    {
+      "epoch": 2.0260073848129716,
+      "grad_norm": 0.8985442519187927,
+      "learning_rate": 7.21344116683097e-05,
+      "loss": 0.3249,
+      "step": 6310
+    },
+    {
+      "epoch": 2.026328463637823,
+      "grad_norm": 0.7551534175872803,
+      "learning_rate": 7.210088939607708e-05,
+      "loss": 0.3713,
+      "step": 6311
+    },
+    {
+      "epoch": 2.0266495424626747,
+      "grad_norm": 1.1302769184112549,
+      "learning_rate": 7.206737052326645e-05,
+      "loss": 0.3553,
+      "step": 6312
+    },
+    {
+      "epoch": 2.0269706212875263,
+      "grad_norm": 0.7023568153381348,
+      "learning_rate": 7.203385505396203e-05,
+      "loss": 0.345,
+      "step": 6313
+    },
+    {
+      "epoch": 2.027291700112378,
+      "grad_norm": 1.19681715965271,
+      "learning_rate": 7.20003429922475e-05,
+      "loss": 0.4365,
+      "step": 6314
+    },
+    {
+      "epoch": 2.027612778937229,
+      "grad_norm": 1.0040020942687988,
+      "learning_rate": 7.196683434220625e-05,
+      "loss": 0.3879,
+      "step": 6315
+    },
+    {
+      "epoch": 2.0279338577620805,
+      "grad_norm": 0.9291061758995056,
+      "learning_rate": 7.193332910792124e-05,
+      "loss": 0.4166,
+      "step": 6316
+    },
+    {
+      "epoch": 2.028254936586932,
+      "grad_norm": 0.5969425439834595,
+      "learning_rate": 7.18998272934749e-05,
+      "loss": 0.2834,
+      "step": 6317
+    },
+    {
+      "epoch": 2.0285760154117836,
+      "grad_norm": 0.7998864650726318,
+      "learning_rate": 7.186632890294941e-05,
+      "loss": 0.3245,
+      "step": 6318
+    },
+    {
+      "epoch": 2.028897094236635,
+      "grad_norm": 0.6707684993743896,
+      "learning_rate": 7.183283394042634e-05,
+      "loss": 0.2533,
+      "step": 6319
+    },
+    {
+      "epoch": 2.0292181730614867,
+      "grad_norm": 0.6057593822479248,
+      "learning_rate": 7.179934240998706e-05,
+      "loss": 0.2811,
+      "step": 6320
+    },
+    {
+      "epoch": 2.029539251886338,
+      "grad_norm": 0.9882358908653259,
+      "learning_rate": 7.176585431571235e-05,
+      "loss": 0.4067,
+      "step": 6321
+    },
+    {
+      "epoch": 2.0298603307111898,
+      "grad_norm": 0.688306450843811,
+      "learning_rate": 7.173236966168268e-05,
+      "loss": 0.2973,
+      "step": 6322
+    },
+    {
+      "epoch": 2.0301814095360413,
+      "grad_norm": 0.886023223400116,
+      "learning_rate": 7.169888845197798e-05,
+      "loss": 0.3354,
+      "step": 6323
+    },
+    {
+      "epoch": 2.0305024883608924,
+      "grad_norm": 0.6586227416992188,
+      "learning_rate": 7.166541069067792e-05,
+      "loss": 0.2762,
+      "step": 6324
+    },
+    {
+      "epoch": 2.030823567185744,
+      "grad_norm": 1.0189096927642822,
+      "learning_rate": 7.163193638186158e-05,
+      "loss": 0.2873,
+      "step": 6325
+    },
+    {
+      "epoch": 2.0311446460105955,
+      "grad_norm": 0.48503080010414124,
+      "learning_rate": 7.159846552960774e-05,
+      "loss": 0.2642,
+      "step": 6326
+    },
+    {
+      "epoch": 2.031465724835447,
+      "grad_norm": 0.4218836724758148,
+      "learning_rate": 7.156499813799476e-05,
+      "loss": 0.247,
+      "step": 6327
+    },
+    {
+      "epoch": 2.0317868036602986,
+      "grad_norm": 0.5482301712036133,
+      "learning_rate": 7.153153421110048e-05,
+      "loss": 0.268,
+      "step": 6328
+    },
+    {
+      "epoch": 2.03210788248515,
+      "grad_norm": 0.47911396622657776,
+      "learning_rate": 7.149807375300239e-05,
+      "loss": 0.2903,
+      "step": 6329
+    },
+    {
+      "epoch": 2.0324289613100017,
+      "grad_norm": 0.5212134718894958,
+      "learning_rate": 7.146461676777756e-05,
+      "loss": 0.8643,
+      "step": 6330
+    },
+    {
+      "epoch": 2.0327500401348533,
+      "grad_norm": 0.43788644671440125,
+      "learning_rate": 7.143116325950265e-05,
+      "loss": 0.3721,
+      "step": 6331
+    },
+    {
+      "epoch": 2.033071118959705,
+      "grad_norm": 0.5127784609794617,
+      "learning_rate": 7.139771323225381e-05,
+      "loss": 0.2185,
+      "step": 6332
+    },
+    {
+      "epoch": 2.033392197784556,
+      "grad_norm": 0.5072081685066223,
+      "learning_rate": 7.136426669010689e-05,
+      "loss": 0.2602,
+      "step": 6333
+    },
+    {
+      "epoch": 2.0337132766094075,
+      "grad_norm": 0.7099177241325378,
+      "learning_rate": 7.13308236371372e-05,
+      "loss": 0.413,
+      "step": 6334
+    },
+    {
+      "epoch": 2.034034355434259,
+      "grad_norm": 0.9787275791168213,
+      "learning_rate": 7.129738407741964e-05,
+      "loss": 0.5135,
+      "step": 6335
+    },
+    {
+      "epoch": 2.0343554342591106,
+      "grad_norm": 0.8141676783561707,
+      "learning_rate": 7.126394801502882e-05,
+      "loss": 0.3424,
+      "step": 6336
+    },
+    {
+      "epoch": 2.034676513083962,
+      "grad_norm": 0.7322252988815308,
+      "learning_rate": 7.123051545403874e-05,
+      "loss": 0.3168,
+      "step": 6337
+    },
+    {
+      "epoch": 2.0349975919088137,
+      "grad_norm": 0.8565484285354614,
+      "learning_rate": 7.119708639852312e-05,
+      "loss": 0.2875,
+      "step": 6338
+    },
+    {
+      "epoch": 2.0353186707336652,
+      "grad_norm": 0.731372594833374,
+      "learning_rate": 7.11636608525551e-05,
+      "loss": 0.3009,
+      "step": 6339
+    },
+    {
+      "epoch": 2.035639749558517,
+      "grad_norm": 0.9566566348075867,
+      "learning_rate": 7.113023882020757e-05,
+      "loss": 0.3571,
+      "step": 6340
+    },
+    {
+      "epoch": 2.0359608283833683,
+      "grad_norm": 1.051526665687561,
+      "learning_rate": 7.109682030555283e-05,
+      "loss": 0.3774,
+      "step": 6341
+    },
+    {
+      "epoch": 2.0362819072082194,
+      "grad_norm": 0.8493756651878357,
+      "learning_rate": 7.106340531266292e-05,
+      "loss": 0.2954,
+      "step": 6342
+    },
+    {
+      "epoch": 2.036602986033071,
+      "grad_norm": 1.491635799407959,
+      "learning_rate": 7.102999384560927e-05,
+      "loss": 0.4563,
+      "step": 6343
+    },
+    {
+      "epoch": 2.0369240648579225,
+      "grad_norm": 0.7433634400367737,
+      "learning_rate": 7.099658590846299e-05,
+      "loss": 0.3044,
+      "step": 6344
+    },
+    {
+      "epoch": 2.037245143682774,
+      "grad_norm": 0.8622017502784729,
+      "learning_rate": 7.096318150529477e-05,
+      "loss": 0.2676,
+      "step": 6345
+    },
+    {
+      "epoch": 2.0375662225076256,
+      "grad_norm": 0.8949119448661804,
+      "learning_rate": 7.092978064017475e-05,
+      "loss": 0.2809,
+      "step": 6346
+    },
+    {
+      "epoch": 2.037887301332477,
+      "grad_norm": 1.0402297973632812,
+      "learning_rate": 7.089638331717284e-05,
+      "loss": 0.4512,
+      "step": 6347
+    },
+    {
+      "epoch": 2.0382083801573287,
+      "grad_norm": 0.9292569756507874,
+      "learning_rate": 7.08629895403583e-05,
+      "loss": 0.4013,
+      "step": 6348
+    },
+    {
+      "epoch": 2.0385294589821803,
+      "grad_norm": 1.625335454940796,
+      "learning_rate": 7.082959931380011e-05,
+      "loss": 0.3156,
+      "step": 6349
+    },
+    {
+      "epoch": 2.038850537807032,
+      "grad_norm": 0.6953380703926086,
+      "learning_rate": 7.079621264156675e-05,
+      "loss": 0.3309,
+      "step": 6350
+    },
+    {
+      "epoch": 2.039171616631883,
+      "grad_norm": 0.7905119061470032,
+      "learning_rate": 7.076282952772633e-05,
+      "loss": 0.3588,
+      "step": 6351
+    },
+    {
+      "epoch": 2.0394926954567345,
+      "grad_norm": 0.7801741361618042,
+      "learning_rate": 7.072944997634646e-05,
+      "loss": 0.2859,
+      "step": 6352
+    },
+    {
+      "epoch": 2.039813774281586,
+      "grad_norm": 1.3757237195968628,
+      "learning_rate": 7.069607399149428e-05,
+      "loss": 0.4971,
+      "step": 6353
+    },
+    {
+      "epoch": 2.0401348531064376,
+      "grad_norm": 0.847542405128479,
+      "learning_rate": 7.06627015772366e-05,
+      "loss": 0.3273,
+      "step": 6354
+    },
+    {
+      "epoch": 2.040455931931289,
+      "grad_norm": 0.6177198886871338,
+      "learning_rate": 7.062933273763975e-05,
+      "loss": 0.2351,
+      "step": 6355
+    },
+    {
+      "epoch": 2.0407770107561407,
+      "grad_norm": 0.9541710019111633,
+      "learning_rate": 7.059596747676962e-05,
+      "loss": 0.3357,
+      "step": 6356
+    },
+    {
+      "epoch": 2.0410980895809923,
+      "grad_norm": 0.8325964212417603,
+      "learning_rate": 7.056260579869165e-05,
+      "loss": 0.2587,
+      "step": 6357
+    },
+    {
+      "epoch": 2.041419168405844,
+      "grad_norm": 0.8677617907524109,
+      "learning_rate": 7.052924770747087e-05,
+      "loss": 0.2691,
+      "step": 6358
+    },
+    {
+      "epoch": 2.0417402472306954,
+      "grad_norm": 0.9467921257019043,
+      "learning_rate": 7.049589320717186e-05,
+      "loss": 0.3171,
+      "step": 6359
+    },
+    {
+      "epoch": 2.0420613260555465,
+      "grad_norm": 0.9890326261520386,
+      "learning_rate": 7.04625423018588e-05,
+      "loss": 0.3976,
+      "step": 6360
+    },
+    {
+      "epoch": 2.042382404880398,
+      "grad_norm": 0.6989203095436096,
+      "learning_rate": 7.042919499559537e-05,
+      "loss": 0.2859,
+      "step": 6361
+    },
+    {
+      "epoch": 2.0427034837052496,
+      "grad_norm": 0.8812395930290222,
+      "learning_rate": 7.039585129244477e-05,
+      "loss": 0.3466,
+      "step": 6362
+    },
+    {
+      "epoch": 2.043024562530101,
+      "grad_norm": 0.8233697414398193,
+      "learning_rate": 7.036251119646992e-05,
+      "loss": 0.2523,
+      "step": 6363
+    },
+    {
+      "epoch": 2.0433456413549527,
+      "grad_norm": 0.9117799401283264,
+      "learning_rate": 7.032917471173318e-05,
+      "loss": 0.3386,
+      "step": 6364
+    },
+    {
+      "epoch": 2.043666720179804,
+      "grad_norm": 0.9386948347091675,
+      "learning_rate": 7.029584184229653e-05,
+      "loss": 0.2908,
+      "step": 6365
+    },
+    {
+      "epoch": 2.0439877990046558,
+      "grad_norm": 0.7460491061210632,
+      "learning_rate": 7.026251259222141e-05,
+      "loss": 0.2804,
+      "step": 6366
+    },
+    {
+      "epoch": 2.0443088778295073,
+      "grad_norm": 0.7433897256851196,
+      "learning_rate": 7.022918696556896e-05,
+      "loss": 0.2923,
+      "step": 6367
+    },
+    {
+      "epoch": 2.0446299566543584,
+      "grad_norm": 0.6513417959213257,
+      "learning_rate": 7.019586496639974e-05,
+      "loss": 0.2907,
+      "step": 6368
+    },
+    {
+      "epoch": 2.04495103547921,
+      "grad_norm": 0.9250189661979675,
+      "learning_rate": 7.016254659877398e-05,
+      "loss": 0.3394,
+      "step": 6369
+    },
+    {
+      "epoch": 2.0452721143040615,
+      "grad_norm": 0.848927915096283,
+      "learning_rate": 7.012923186675144e-05,
+      "loss": 0.3345,
+      "step": 6370
+    },
+    {
+      "epoch": 2.045593193128913,
+      "grad_norm": 0.718906044960022,
+      "learning_rate": 7.009592077439134e-05,
+      "loss": 0.3041,
+      "step": 6371
+    },
+    {
+      "epoch": 2.0459142719537646,
+      "grad_norm": 0.887864351272583,
+      "learning_rate": 7.00626133257526e-05,
+      "loss": 0.3669,
+      "step": 6372
+    },
+    {
+      "epoch": 2.046235350778616,
+      "grad_norm": 0.7558139562606812,
+      "learning_rate": 7.002930952489362e-05,
+      "loss": 0.2784,
+      "step": 6373
+    },
+    {
+      "epoch": 2.0465564296034677,
+      "grad_norm": 0.6548720002174377,
+      "learning_rate": 6.999600937587239e-05,
+      "loss": 0.2803,
+      "step": 6374
+    },
+    {
+      "epoch": 2.0468775084283193,
+      "grad_norm": 0.7359789609909058,
+      "learning_rate": 6.996271288274636e-05,
+      "loss": 0.2794,
+      "step": 6375
+    },
+    {
+      "epoch": 2.047198587253171,
+      "grad_norm": 0.7246059775352478,
+      "learning_rate": 6.992942004957271e-05,
+      "loss": 0.2569,
+      "step": 6376
+    },
+    {
+      "epoch": 2.047519666078022,
+      "grad_norm": 0.4649535119533539,
+      "learning_rate": 6.989613088040796e-05,
+      "loss": 0.2646,
+      "step": 6377
+    },
+    {
+      "epoch": 2.0478407449028735,
+      "grad_norm": 1.727747917175293,
+      "learning_rate": 6.986284537930838e-05,
+      "loss": 0.2704,
+      "step": 6378
+    },
+    {
+      "epoch": 2.048161823727725,
+      "grad_norm": 0.7226576805114746,
+      "learning_rate": 6.982956355032968e-05,
+      "loss": 0.2848,
+      "step": 6379
+    },
+    {
+      "epoch": 2.0484829025525766,
+      "grad_norm": 0.4427557587623596,
+      "learning_rate": 6.979628539752711e-05,
+      "loss": 0.5042,
+      "step": 6380
+    },
+    {
+      "epoch": 2.048803981377428,
+      "grad_norm": 0.6116136312484741,
+      "learning_rate": 6.976301092495556e-05,
+      "loss": 0.7693,
+      "step": 6381
+    },
+    {
+      "epoch": 2.0491250602022797,
+      "grad_norm": 0.5171028971672058,
+      "learning_rate": 6.972974013666942e-05,
+      "loss": 0.3138,
+      "step": 6382
+    },
+    {
+      "epoch": 2.0494461390271312,
+      "grad_norm": 0.5506100058555603,
+      "learning_rate": 6.969647303672262e-05,
+      "loss": 0.1486,
+      "step": 6383
+    },
+    {
+      "epoch": 2.049767217851983,
+      "grad_norm": 0.48883306980133057,
+      "learning_rate": 6.966320962916864e-05,
+      "loss": 0.1554,
+      "step": 6384
+    },
+    {
+      "epoch": 2.0500882966768343,
+      "grad_norm": 0.720169186592102,
+      "learning_rate": 6.962994991806059e-05,
+      "loss": 0.3594,
+      "step": 6385
+    },
+    {
+      "epoch": 2.0504093755016854,
+      "grad_norm": 0.9285702705383301,
+      "learning_rate": 6.959669390745097e-05,
+      "loss": 0.5368,
+      "step": 6386
+    },
+    {
+      "epoch": 2.050730454326537,
+      "grad_norm": 0.9640303254127502,
+      "learning_rate": 6.956344160139201e-05,
+      "loss": 0.3814,
+      "step": 6387
+    },
+    {
+      "epoch": 2.0510515331513886,
+      "grad_norm": 0.9004988670349121,
+      "learning_rate": 6.953019300393538e-05,
+      "loss": 0.4218,
+      "step": 6388
+    },
+    {
+      "epoch": 2.05137261197624,
+      "grad_norm": 0.4966624975204468,
+      "learning_rate": 6.949694811913225e-05,
+      "loss": 0.206,
+      "step": 6389
+    },
+    {
+      "epoch": 2.0516936908010917,
+      "grad_norm": 0.7329800724983215,
+      "learning_rate": 6.946370695103353e-05,
+      "loss": 0.3075,
+      "step": 6390
+    },
+    {
+      "epoch": 2.052014769625943,
+      "grad_norm": 0.8683103322982788,
+      "learning_rate": 6.943046950368944e-05,
+      "loss": 0.4042,
+      "step": 6391
+    },
+    {
+      "epoch": 2.0523358484507948,
+      "grad_norm": 0.7453745603561401,
+      "learning_rate": 6.939723578114993e-05,
+      "loss": 0.325,
+      "step": 6392
+    },
+    {
+      "epoch": 2.0526569272756463,
+      "grad_norm": 0.8619517683982849,
+      "learning_rate": 6.93640057874644e-05,
+      "loss": 0.3574,
+      "step": 6393
+    },
+    {
+      "epoch": 2.052978006100498,
+      "grad_norm": 0.7723063826560974,
+      "learning_rate": 6.93307795266819e-05,
+      "loss": 0.3337,
+      "step": 6394
+    },
+    {
+      "epoch": 2.053299084925349,
+      "grad_norm": 0.7700170278549194,
+      "learning_rate": 6.929755700285081e-05,
+      "loss": 0.4027,
+      "step": 6395
+    },
+    {
+      "epoch": 2.0536201637502005,
+      "grad_norm": 0.7963249087333679,
+      "learning_rate": 6.92643382200193e-05,
+      "loss": 0.3438,
+      "step": 6396
+    },
+    {
+      "epoch": 2.053941242575052,
+      "grad_norm": 0.8687177896499634,
+      "learning_rate": 6.923112318223496e-05,
+      "loss": 0.4145,
+      "step": 6397
+    },
+    {
+      "epoch": 2.0542623213999036,
+      "grad_norm": 0.988239049911499,
+      "learning_rate": 6.91979118935449e-05,
+      "loss": 0.3573,
+      "step": 6398
+    },
+    {
+      "epoch": 2.054583400224755,
+      "grad_norm": 1.586459755897522,
+      "learning_rate": 6.916470435799587e-05,
+      "loss": 0.3187,
+      "step": 6399
+    },
+    {
+      "epoch": 2.0549044790496067,
+      "grad_norm": 0.8232465386390686,
+      "learning_rate": 6.913150057963404e-05,
+      "loss": 0.4458,
+      "step": 6400
+    },
+    {
+      "epoch": 2.0552255578744583,
+      "grad_norm": 0.6517797708511353,
+      "learning_rate": 6.909830056250527e-05,
+      "loss": 0.2838,
+      "step": 6401
+    },
+    {
+      "epoch": 2.05554663669931,
+      "grad_norm": 0.7707570195198059,
+      "learning_rate": 6.90651043106548e-05,
+      "loss": 0.3037,
+      "step": 6402
+    },
+    {
+      "epoch": 2.0558677155241614,
+      "grad_norm": 1.065109133720398,
+      "learning_rate": 6.90319118281276e-05,
+      "loss": 0.5664,
+      "step": 6403
+    },
+    {
+      "epoch": 2.0561887943490125,
+      "grad_norm": 0.7648012042045593,
+      "learning_rate": 6.899872311896795e-05,
+      "loss": 0.392,
+      "step": 6404
+    },
+    {
+      "epoch": 2.056509873173864,
+      "grad_norm": 0.7669097185134888,
+      "learning_rate": 6.896553818721989e-05,
+      "loss": 0.2889,
+      "step": 6405
+    },
+    {
+      "epoch": 2.0568309519987156,
+      "grad_norm": 1.1542384624481201,
+      "learning_rate": 6.893235703692685e-05,
+      "loss": 0.3426,
+      "step": 6406
+    },
+    {
+      "epoch": 2.057152030823567,
+      "grad_norm": 0.5195868015289307,
+      "learning_rate": 6.889917967213185e-05,
+      "loss": 0.2784,
+      "step": 6407
+    },
+    {
+      "epoch": 2.0574731096484187,
+      "grad_norm": 0.7854254841804504,
+      "learning_rate": 6.88660060968775e-05,
+      "loss": 0.3684,
+      "step": 6408
+    },
+    {
+      "epoch": 2.0577941884732702,
+      "grad_norm": 1.0029950141906738,
+      "learning_rate": 6.883283631520582e-05,
+      "loss": 0.4543,
+      "step": 6409
+    },
+    {
+      "epoch": 2.0581152672981218,
+      "grad_norm": 1.1565372943878174,
+      "learning_rate": 6.879967033115853e-05,
+      "loss": 0.4042,
+      "step": 6410
+    },
+    {
+      "epoch": 2.0584363461229733,
+      "grad_norm": 1.5860395431518555,
+      "learning_rate": 6.876650814877674e-05,
+      "loss": 0.3917,
+      "step": 6411
+    },
+    {
+      "epoch": 2.058757424947825,
+      "grad_norm": 1.0455715656280518,
+      "learning_rate": 6.873334977210122e-05,
+      "loss": 0.3976,
+      "step": 6412
+    },
+    {
+      "epoch": 2.059078503772676,
+      "grad_norm": 0.671190619468689,
+      "learning_rate": 6.870019520517217e-05,
+      "loss": 0.31,
+      "step": 6413
+    },
+    {
+      "epoch": 2.0593995825975275,
+      "grad_norm": 0.5373357534408569,
+      "learning_rate": 6.866704445202943e-05,
+      "loss": 0.2763,
+      "step": 6414
+    },
+    {
+      "epoch": 2.059720661422379,
+      "grad_norm": 1.1658103466033936,
+      "learning_rate": 6.863389751671225e-05,
+      "loss": 0.3591,
+      "step": 6415
+    },
+    {
+      "epoch": 2.0600417402472306,
+      "grad_norm": 0.7469364404678345,
+      "learning_rate": 6.860075440325951e-05,
+      "loss": 0.3128,
+      "step": 6416
+    },
+    {
+      "epoch": 2.060362819072082,
+      "grad_norm": 1.2972486019134521,
+      "learning_rate": 6.856761511570963e-05,
+      "loss": 0.3338,
+      "step": 6417
+    },
+    {
+      "epoch": 2.0606838978969337,
+      "grad_norm": 0.8493379950523376,
+      "learning_rate": 6.853447965810046e-05,
+      "loss": 0.277,
+      "step": 6418
+    },
+    {
+      "epoch": 2.0610049767217853,
+      "grad_norm": 0.7463499307632446,
+      "learning_rate": 6.850134803446954e-05,
+      "loss": 0.3535,
+      "step": 6419
+    },
+    {
+      "epoch": 2.061326055546637,
+      "grad_norm": 0.6443728804588318,
+      "learning_rate": 6.846822024885379e-05,
+      "loss": 0.2776,
+      "step": 6420
+    },
+    {
+      "epoch": 2.0616471343714884,
+      "grad_norm": 1.100448489189148,
+      "learning_rate": 6.843509630528977e-05,
+      "loss": 0.3022,
+      "step": 6421
+    },
+    {
+      "epoch": 2.0619682131963395,
+      "grad_norm": 0.7827779054641724,
+      "learning_rate": 6.840197620781349e-05,
+      "loss": 0.3248,
+      "step": 6422
+    },
+    {
+      "epoch": 2.062289292021191,
+      "grad_norm": 0.8405880331993103,
+      "learning_rate": 6.836885996046061e-05,
+      "loss": 0.3416,
+      "step": 6423
+    },
+    {
+      "epoch": 2.0626103708460426,
+      "grad_norm": 0.832291841506958,
+      "learning_rate": 6.833574756726618e-05,
+      "loss": 0.3128,
+      "step": 6424
+    },
+    {
+      "epoch": 2.062931449670894,
+      "grad_norm": 0.8597456812858582,
+      "learning_rate": 6.830263903226483e-05,
+      "loss": 0.3529,
+      "step": 6425
+    },
+    {
+      "epoch": 2.0632525284957457,
+      "grad_norm": 0.6993900537490845,
+      "learning_rate": 6.826953435949081e-05,
+      "loss": 0.3465,
+      "step": 6426
+    },
+    {
+      "epoch": 2.0635736073205972,
+      "grad_norm": 0.8503331542015076,
+      "learning_rate": 6.823643355297773e-05,
+      "loss": 0.3257,
+      "step": 6427
+    },
+    {
+      "epoch": 2.063894686145449,
+      "grad_norm": 0.460232138633728,
+      "learning_rate": 6.820333661675893e-05,
+      "loss": 0.2558,
+      "step": 6428
+    },
+    {
+      "epoch": 2.0642157649703003,
+      "grad_norm": 0.6968713402748108,
+      "learning_rate": 6.817024355486706e-05,
+      "loss": 0.3211,
+      "step": 6429
+    },
+    {
+      "epoch": 2.064536843795152,
+      "grad_norm": 0.532193660736084,
+      "learning_rate": 6.81371543713345e-05,
+      "loss": 0.8584,
+      "step": 6430
+    },
+    {
+      "epoch": 2.064857922620003,
+      "grad_norm": 0.474894255399704,
+      "learning_rate": 6.8104069070193e-05,
+      "loss": 0.6014,
+      "step": 6431
+    },
+    {
+      "epoch": 2.0651790014448546,
+      "grad_norm": 0.5201006531715393,
+      "learning_rate": 6.807098765547397e-05,
+      "loss": 0.2353,
+      "step": 6432
+    },
+    {
+      "epoch": 2.065500080269706,
+      "grad_norm": 0.5294368267059326,
+      "learning_rate": 6.803791013120822e-05,
+      "loss": 0.1455,
+      "step": 6433
+    },
+    {
+      "epoch": 2.0658211590945577,
+      "grad_norm": 0.3815780580043793,
+      "learning_rate": 6.800483650142617e-05,
+      "loss": 0.0917,
+      "step": 6434
+    },
+    {
+      "epoch": 2.066142237919409,
+      "grad_norm": 0.5428910851478577,
+      "learning_rate": 6.797176677015775e-05,
+      "loss": 0.1991,
+      "step": 6435
+    },
+    {
+      "epoch": 2.0664633167442608,
+      "grad_norm": 0.46137359738349915,
+      "learning_rate": 6.793870094143238e-05,
+      "loss": 0.1742,
+      "step": 6436
+    },
+    {
+      "epoch": 2.0667843955691123,
+      "grad_norm": 0.6830509901046753,
+      "learning_rate": 6.790563901927907e-05,
+      "loss": 0.4138,
+      "step": 6437
+    },
+    {
+      "epoch": 2.067105474393964,
+      "grad_norm": 0.5673910975456238,
+      "learning_rate": 6.787258100772627e-05,
+      "loss": 0.2538,
+      "step": 6438
+    },
+    {
+      "epoch": 2.0674265532188154,
+      "grad_norm": 0.769719123840332,
+      "learning_rate": 6.783952691080203e-05,
+      "loss": 0.257,
+      "step": 6439
+    },
+    {
+      "epoch": 2.0677476320436665,
+      "grad_norm": 0.6150414347648621,
+      "learning_rate": 6.780647673253391e-05,
+      "loss": 0.2459,
+      "step": 6440
+    },
+    {
+      "epoch": 2.068068710868518,
+      "grad_norm": 0.5758988857269287,
+      "learning_rate": 6.77734304769489e-05,
+      "loss": 0.2108,
+      "step": 6441
+    },
+    {
+      "epoch": 2.0683897896933696,
+      "grad_norm": 0.979168713092804,
+      "learning_rate": 6.774038814807369e-05,
+      "loss": 0.4103,
+      "step": 6442
+    },
+    {
+      "epoch": 2.068710868518221,
+      "grad_norm": 0.8934927582740784,
+      "learning_rate": 6.770734974993426e-05,
+      "loss": 0.3478,
+      "step": 6443
+    },
+    {
+      "epoch": 2.0690319473430727,
+      "grad_norm": 1.4555315971374512,
+      "learning_rate": 6.767431528655635e-05,
+      "loss": 0.3517,
+      "step": 6444
+    },
+    {
+      "epoch": 2.0693530261679243,
+      "grad_norm": 0.8350145220756531,
+      "learning_rate": 6.764128476196505e-05,
+      "loss": 0.3284,
+      "step": 6445
+    },
+    {
+      "epoch": 2.069674104992776,
+      "grad_norm": 0.7142661809921265,
+      "learning_rate": 6.760825818018508e-05,
+      "loss": 0.2966,
+      "step": 6446
+    },
+    {
+      "epoch": 2.0699951838176274,
+      "grad_norm": 1.0782002210617065,
+      "learning_rate": 6.757523554524056e-05,
+      "loss": 0.4881,
+      "step": 6447
+    },
+    {
+      "epoch": 2.070316262642479,
+      "grad_norm": 0.9906001687049866,
+      "learning_rate": 6.754221686115525e-05,
+      "loss": 0.4128,
+      "step": 6448
+    },
+    {
+      "epoch": 2.07063734146733,
+      "grad_norm": 0.8098526000976562,
+      "learning_rate": 6.750920213195238e-05,
+      "loss": 0.3199,
+      "step": 6449
+    },
+    {
+      "epoch": 2.0709584202921816,
+      "grad_norm": 1.1390670537948608,
+      "learning_rate": 6.747619136165463e-05,
+      "loss": 0.41,
+      "step": 6450
+    },
+    {
+      "epoch": 2.071279499117033,
+      "grad_norm": 0.8910248875617981,
+      "learning_rate": 6.744318455428436e-05,
+      "loss": 0.3355,
+      "step": 6451
+    },
+    {
+      "epoch": 2.0716005779418847,
+      "grad_norm": 0.9463450312614441,
+      "learning_rate": 6.741018171386326e-05,
+      "loss": 0.4561,
+      "step": 6452
+    },
+    {
+      "epoch": 2.0719216567667362,
+      "grad_norm": 0.7821532487869263,
+      "learning_rate": 6.737718284441267e-05,
+      "loss": 0.3068,
+      "step": 6453
+    },
+    {
+      "epoch": 2.072242735591588,
+      "grad_norm": 1.0844594240188599,
+      "learning_rate": 6.734418794995337e-05,
+      "loss": 0.4195,
+      "step": 6454
+    },
+    {
+      "epoch": 2.0725638144164393,
+      "grad_norm": 0.7623462677001953,
+      "learning_rate": 6.731119703450577e-05,
+      "loss": 0.3326,
+      "step": 6455
+    },
+    {
+      "epoch": 2.072884893241291,
+      "grad_norm": 0.7975969314575195,
+      "learning_rate": 6.727821010208962e-05,
+      "loss": 0.3355,
+      "step": 6456
+    },
+    {
+      "epoch": 2.0732059720661424,
+      "grad_norm": 1.0122053623199463,
+      "learning_rate": 6.724522715672432e-05,
+      "loss": 0.4545,
+      "step": 6457
+    },
+    {
+      "epoch": 2.0735270508909935,
+      "grad_norm": 0.9668064117431641,
+      "learning_rate": 6.721224820242875e-05,
+      "loss": 0.381,
+      "step": 6458
+    },
+    {
+      "epoch": 2.073848129715845,
+      "grad_norm": 0.8006826043128967,
+      "learning_rate": 6.717927324322124e-05,
+      "loss": 0.3322,
+      "step": 6459
+    },
+    {
+      "epoch": 2.0741692085406966,
+      "grad_norm": 0.6765208840370178,
+      "learning_rate": 6.714630228311978e-05,
+      "loss": 0.2891,
+      "step": 6460
+    },
+    {
+      "epoch": 2.074490287365548,
+      "grad_norm": 0.8815526366233826,
+      "learning_rate": 6.711333532614168e-05,
+      "loss": 0.3118,
+      "step": 6461
+    },
+    {
+      "epoch": 2.0748113661903997,
+      "grad_norm": 0.7715243101119995,
+      "learning_rate": 6.708037237630395e-05,
+      "loss": 0.3149,
+      "step": 6462
+    },
+    {
+      "epoch": 2.0751324450152513,
+      "grad_norm": 0.7587166428565979,
+      "learning_rate": 6.704741343762295e-05,
+      "loss": 0.3618,
+      "step": 6463
+    },
+    {
+      "epoch": 2.075453523840103,
+      "grad_norm": 0.6574379205703735,
+      "learning_rate": 6.701445851411472e-05,
+      "loss": 0.3009,
+      "step": 6464
+    },
+    {
+      "epoch": 2.0757746026649544,
+      "grad_norm": 1.0423983335494995,
+      "learning_rate": 6.698150760979463e-05,
+      "loss": 0.3738,
+      "step": 6465
+    },
+    {
+      "epoch": 2.076095681489806,
+      "grad_norm": 0.5287320613861084,
+      "learning_rate": 6.69485607286777e-05,
+      "loss": 0.2526,
+      "step": 6466
+    },
+    {
+      "epoch": 2.076416760314657,
+      "grad_norm": 0.7871254086494446,
+      "learning_rate": 6.69156178747784e-05,
+      "loss": 0.2799,
+      "step": 6467
+    },
+    {
+      "epoch": 2.0767378391395086,
+      "grad_norm": 1.0920846462249756,
+      "learning_rate": 6.688267905211067e-05,
+      "loss": 0.3249,
+      "step": 6468
+    },
+    {
+      "epoch": 2.07705891796436,
+      "grad_norm": 1.0607701539993286,
+      "learning_rate": 6.684974426468808e-05,
+      "loss": 0.363,
+      "step": 6469
+    },
+    {
+      "epoch": 2.0773799967892117,
+      "grad_norm": 0.6999905705451965,
+      "learning_rate": 6.681681351652356e-05,
+      "loss": 0.3041,
+      "step": 6470
+    },
+    {
+      "epoch": 2.0777010756140633,
+      "grad_norm": 1.2206069231033325,
+      "learning_rate": 6.67838868116297e-05,
+      "loss": 0.3103,
+      "step": 6471
+    },
+    {
+      "epoch": 2.078022154438915,
+      "grad_norm": 0.7365018725395203,
+      "learning_rate": 6.675096415401842e-05,
+      "loss": 0.3293,
+      "step": 6472
+    },
+    {
+      "epoch": 2.0783432332637664,
+      "grad_norm": 0.6262559294700623,
+      "learning_rate": 6.671804554770135e-05,
+      "loss": 0.2744,
+      "step": 6473
+    },
+    {
+      "epoch": 2.078664312088618,
+      "grad_norm": 0.5494680404663086,
+      "learning_rate": 6.668513099668944e-05,
+      "loss": 0.2866,
+      "step": 6474
+    },
+    {
+      "epoch": 2.0789853909134695,
+      "grad_norm": 0.6878925561904907,
+      "learning_rate": 6.66522205049933e-05,
+      "loss": 0.3257,
+      "step": 6475
+    },
+    {
+      "epoch": 2.0793064697383206,
+      "grad_norm": 0.7526207566261292,
+      "learning_rate": 6.661931407662292e-05,
+      "loss": 0.3167,
+      "step": 6476
+    },
+    {
+      "epoch": 2.079627548563172,
+      "grad_norm": 1.070246934890747,
+      "learning_rate": 6.658641171558785e-05,
+      "loss": 0.3273,
+      "step": 6477
+    },
+    {
+      "epoch": 2.0799486273880237,
+      "grad_norm": 0.7761482000350952,
+      "learning_rate": 6.65535134258972e-05,
+      "loss": 0.304,
+      "step": 6478
+    },
+    {
+      "epoch": 2.080269706212875,
+      "grad_norm": 0.6462170481681824,
+      "learning_rate": 6.652061921155943e-05,
+      "loss": 0.2834,
+      "step": 6479
+    },
+    {
+      "epoch": 2.0805907850377268,
+      "grad_norm": 0.5869211554527283,
+      "learning_rate": 6.648772907658272e-05,
+      "loss": 0.5811,
+      "step": 6480
+    },
+    {
+      "epoch": 2.0809118638625783,
+      "grad_norm": 0.397668719291687,
+      "learning_rate": 6.64548430249745e-05,
+      "loss": 0.2837,
+      "step": 6481
+    },
+    {
+      "epoch": 2.08123294268743,
+      "grad_norm": 0.5283700227737427,
+      "learning_rate": 6.642196106074194e-05,
+      "loss": 0.2056,
+      "step": 6482
+    },
+    {
+      "epoch": 2.0815540215122814,
+      "grad_norm": 0.5720309615135193,
+      "learning_rate": 6.638908318789156e-05,
+      "loss": 0.1464,
+      "step": 6483
+    },
+    {
+      "epoch": 2.081875100337133,
+      "grad_norm": 0.4584559202194214,
+      "learning_rate": 6.635620941042945e-05,
+      "loss": 0.2292,
+      "step": 6484
+    },
+    {
+      "epoch": 2.082196179161984,
+      "grad_norm": 0.5467604994773865,
+      "learning_rate": 6.63233397323612e-05,
+      "loss": 0.2558,
+      "step": 6485
+    },
+    {
+      "epoch": 2.0825172579868356,
+      "grad_norm": 0.6699813008308411,
+      "learning_rate": 6.62904741576918e-05,
+      "loss": 0.3378,
+      "step": 6486
+    },
+    {
+      "epoch": 2.082838336811687,
+      "grad_norm": 0.9696993231773376,
+      "learning_rate": 6.62576126904259e-05,
+      "loss": 0.3986,
+      "step": 6487
+    },
+    {
+      "epoch": 2.0831594156365387,
+      "grad_norm": 0.788311243057251,
+      "learning_rate": 6.622475533456751e-05,
+      "loss": 0.361,
+      "step": 6488
+    },
+    {
+      "epoch": 2.0834804944613903,
+      "grad_norm": 1.0402991771697998,
+      "learning_rate": 6.61919020941203e-05,
+      "loss": 0.3738,
+      "step": 6489
+    },
+    {
+      "epoch": 2.083801573286242,
+      "grad_norm": 0.8160973191261292,
+      "learning_rate": 6.61590529730872e-05,
+      "loss": 0.3736,
+      "step": 6490
+    },
+    {
+      "epoch": 2.0841226521110934,
+      "grad_norm": 0.6715680360794067,
+      "learning_rate": 6.612620797547087e-05,
+      "loss": 0.3151,
+      "step": 6491
+    },
+    {
+      "epoch": 2.084443730935945,
+      "grad_norm": 0.8539772629737854,
+      "learning_rate": 6.609336710527332e-05,
+      "loss": 0.4157,
+      "step": 6492
+    },
+    {
+      "epoch": 2.0847648097607965,
+      "grad_norm": 0.8843332529067993,
+      "learning_rate": 6.606053036649619e-05,
+      "loss": 0.3158,
+      "step": 6493
+    },
+    {
+      "epoch": 2.0850858885856476,
+      "grad_norm": 0.9574533700942993,
+      "learning_rate": 6.602769776314049e-05,
+      "loss": 0.3908,
+      "step": 6494
+    },
+    {
+      "epoch": 2.085406967410499,
+      "grad_norm": 0.8649710416793823,
+      "learning_rate": 6.599486929920673e-05,
+      "loss": 0.3199,
+      "step": 6495
+    },
+    {
+      "epoch": 2.0857280462353507,
+      "grad_norm": 1.0361381769180298,
+      "learning_rate": 6.5962044978695e-05,
+      "loss": 0.4014,
+      "step": 6496
+    },
+    {
+      "epoch": 2.0860491250602022,
+      "grad_norm": 1.0464524030685425,
+      "learning_rate": 6.592922480560483e-05,
+      "loss": 0.2594,
+      "step": 6497
+    },
+    {
+      "epoch": 2.086370203885054,
+      "grad_norm": 1.0081483125686646,
+      "learning_rate": 6.589640878393531e-05,
+      "loss": 0.3912,
+      "step": 6498
+    },
+    {
+      "epoch": 2.0866912827099053,
+      "grad_norm": 0.9563593864440918,
+      "learning_rate": 6.58635969176849e-05,
+      "loss": 0.453,
+      "step": 6499
+    },
+    {
+      "epoch": 2.087012361534757,
+      "grad_norm": 0.8743892908096313,
+      "learning_rate": 6.583078921085167e-05,
+      "loss": 0.3783,
+      "step": 6500
+    },
+    {
+      "epoch": 2.0873334403596084,
+      "grad_norm": 0.964160680770874,
+      "learning_rate": 6.579798566743314e-05,
+      "loss": 0.3375,
+      "step": 6501
+    },
+    {
+      "epoch": 2.08765451918446,
+      "grad_norm": 1.0127390623092651,
+      "learning_rate": 6.57651862914263e-05,
+      "loss": 0.4236,
+      "step": 6502
+    },
+    {
+      "epoch": 2.087975598009311,
+      "grad_norm": 0.7326671481132507,
+      "learning_rate": 6.573239108682768e-05,
+      "loss": 0.3081,
+      "step": 6503
+    },
+    {
+      "epoch": 2.0882966768341626,
+      "grad_norm": 0.9221829771995544,
+      "learning_rate": 6.569960005763323e-05,
+      "loss": 0.284,
+      "step": 6504
+    },
+    {
+      "epoch": 2.088617755659014,
+      "grad_norm": 0.6714749336242676,
+      "learning_rate": 6.566681320783849e-05,
+      "loss": 0.2871,
+      "step": 6505
+    },
+    {
+      "epoch": 2.0889388344838657,
+      "grad_norm": 0.9831109642982483,
+      "learning_rate": 6.56340305414384e-05,
+      "loss": 0.3815,
+      "step": 6506
+    },
+    {
+      "epoch": 2.0892599133087173,
+      "grad_norm": 0.8641473650932312,
+      "learning_rate": 6.560125206242746e-05,
+      "loss": 0.3888,
+      "step": 6507
+    },
+    {
+      "epoch": 2.089580992133569,
+      "grad_norm": 0.9115124344825745,
+      "learning_rate": 6.55684777747996e-05,
+      "loss": 0.3683,
+      "step": 6508
+    },
+    {
+      "epoch": 2.0899020709584204,
+      "grad_norm": 0.6893644332885742,
+      "learning_rate": 6.55357076825483e-05,
+      "loss": 0.2572,
+      "step": 6509
+    },
+    {
+      "epoch": 2.090223149783272,
+      "grad_norm": 1.011763572692871,
+      "learning_rate": 6.550294178966647e-05,
+      "loss": 0.3746,
+      "step": 6510
+    },
+    {
+      "epoch": 2.0905442286081235,
+      "grad_norm": 0.739936888217926,
+      "learning_rate": 6.547018010014654e-05,
+      "loss": 0.338,
+      "step": 6511
+    },
+    {
+      "epoch": 2.0908653074329746,
+      "grad_norm": 0.9537821412086487,
+      "learning_rate": 6.543742261798045e-05,
+      "loss": 0.2584,
+      "step": 6512
+    },
+    {
+      "epoch": 2.091186386257826,
+      "grad_norm": 0.6458030343055725,
+      "learning_rate": 6.540466934715953e-05,
+      "loss": 0.3046,
+      "step": 6513
+    },
+    {
+      "epoch": 2.0915074650826777,
+      "grad_norm": 0.7992889881134033,
+      "learning_rate": 6.537192029167474e-05,
+      "loss": 0.3166,
+      "step": 6514
+    },
+    {
+      "epoch": 2.0918285439075293,
+      "grad_norm": 0.651301383972168,
+      "learning_rate": 6.53391754555164e-05,
+      "loss": 0.2818,
+      "step": 6515
+    },
+    {
+      "epoch": 2.092149622732381,
+      "grad_norm": 0.6748250722885132,
+      "learning_rate": 6.530643484267443e-05,
+      "loss": 0.3025,
+      "step": 6516
+    },
+    {
+      "epoch": 2.0924707015572324,
+      "grad_norm": 0.7246717810630798,
+      "learning_rate": 6.52736984571381e-05,
+      "loss": 0.258,
+      "step": 6517
+    },
+    {
+      "epoch": 2.092791780382084,
+      "grad_norm": 0.7231757640838623,
+      "learning_rate": 6.52409663028963e-05,
+      "loss": 0.2886,
+      "step": 6518
+    },
+    {
+      "epoch": 2.0931128592069355,
+      "grad_norm": 0.8351329565048218,
+      "learning_rate": 6.520823838393731e-05,
+      "loss": 0.3284,
+      "step": 6519
+    },
+    {
+      "epoch": 2.093433938031787,
+      "grad_norm": 0.9581044316291809,
+      "learning_rate": 6.517551470424895e-05,
+      "loss": 0.3134,
+      "step": 6520
+    },
+    {
+      "epoch": 2.093755016856638,
+      "grad_norm": 0.9175045490264893,
+      "learning_rate": 6.51427952678185e-05,
+      "loss": 0.3739,
+      "step": 6521
+    },
+    {
+      "epoch": 2.0940760956814897,
+      "grad_norm": 0.9837835431098938,
+      "learning_rate": 6.511008007863268e-05,
+      "loss": 0.4271,
+      "step": 6522
+    },
+    {
+      "epoch": 2.094397174506341,
+      "grad_norm": 0.467220664024353,
+      "learning_rate": 6.507736914067781e-05,
+      "loss": 0.2381,
+      "step": 6523
+    },
+    {
+      "epoch": 2.0947182533311928,
+      "grad_norm": 1.0905086994171143,
+      "learning_rate": 6.504466245793955e-05,
+      "loss": 0.4082,
+      "step": 6524
+    },
+    {
+      "epoch": 2.0950393321560443,
+      "grad_norm": 0.6342186331748962,
+      "learning_rate": 6.501196003440314e-05,
+      "loss": 0.268,
+      "step": 6525
+    },
+    {
+      "epoch": 2.095360410980896,
+      "grad_norm": 0.6534708142280579,
+      "learning_rate": 6.497926187405326e-05,
+      "loss": 0.261,
+      "step": 6526
+    },
+    {
+      "epoch": 2.0956814898057474,
+      "grad_norm": 0.5003716349601746,
+      "learning_rate": 6.494656798087412e-05,
+      "loss": 0.2658,
+      "step": 6527
+    },
+    {
+      "epoch": 2.096002568630599,
+      "grad_norm": 0.6496462225914001,
+      "learning_rate": 6.49138783588493e-05,
+      "loss": 0.3374,
+      "step": 6528
+    },
+    {
+      "epoch": 2.0963236474554505,
+      "grad_norm": 0.8161367177963257,
+      "learning_rate": 6.488119301196201e-05,
+      "loss": 0.3874,
+      "step": 6529
+    },
+    {
+      "epoch": 2.0966447262803016,
+      "grad_norm": 0.5122060775756836,
+      "learning_rate": 6.484851194419484e-05,
+      "loss": 0.6031,
+      "step": 6530
+    },
+    {
+      "epoch": 2.096965805105153,
+      "grad_norm": 0.43576663732528687,
+      "learning_rate": 6.481583515952983e-05,
+      "loss": 0.6629,
+      "step": 6531
+    },
+    {
+      "epoch": 2.0972868839300047,
+      "grad_norm": 0.484679639339447,
+      "learning_rate": 6.478316266194861e-05,
+      "loss": 0.2203,
+      "step": 6532
+    },
+    {
+      "epoch": 2.0976079627548563,
+      "grad_norm": 0.5761163830757141,
+      "learning_rate": 6.475049445543215e-05,
+      "loss": 0.2443,
+      "step": 6533
+    },
+    {
+      "epoch": 2.097929041579708,
+      "grad_norm": 0.5513499975204468,
+      "learning_rate": 6.471783054396105e-05,
+      "loss": 0.1622,
+      "step": 6534
+    },
+    {
+      "epoch": 2.0982501204045594,
+      "grad_norm": 0.522097110748291,
+      "learning_rate": 6.468517093151525e-05,
+      "loss": 0.1611,
+      "step": 6535
+    },
+    {
+      "epoch": 2.098571199229411,
+      "grad_norm": 0.7834147214889526,
+      "learning_rate": 6.465251562207431e-05,
+      "loss": 0.3935,
+      "step": 6536
+    },
+    {
+      "epoch": 2.0988922780542625,
+      "grad_norm": 1.0069727897644043,
+      "learning_rate": 6.461986461961706e-05,
+      "loss": 0.4389,
+      "step": 6537
+    },
+    {
+      "epoch": 2.099213356879114,
+      "grad_norm": 1.162589192390442,
+      "learning_rate": 6.458721792812204e-05,
+      "loss": 0.4037,
+      "step": 6538
+    },
+    {
+      "epoch": 2.099534435703965,
+      "grad_norm": 0.7293219566345215,
+      "learning_rate": 6.455457555156705e-05,
+      "loss": 0.341,
+      "step": 6539
+    },
+    {
+      "epoch": 2.0998555145288167,
+      "grad_norm": 0.6971396803855896,
+      "learning_rate": 6.452193749392952e-05,
+      "loss": 0.2258,
+      "step": 6540
+    },
+    {
+      "epoch": 2.1001765933536682,
+      "grad_norm": 0.7499396800994873,
+      "learning_rate": 6.448930375918631e-05,
+      "loss": 0.2789,
+      "step": 6541
+    },
+    {
+      "epoch": 2.10049767217852,
+      "grad_norm": 0.637330174446106,
+      "learning_rate": 6.44566743513137e-05,
+      "loss": 0.301,
+      "step": 6542
+    },
+    {
+      "epoch": 2.1008187510033713,
+      "grad_norm": 0.9882725477218628,
+      "learning_rate": 6.442404927428751e-05,
+      "loss": 0.499,
+      "step": 6543
+    },
+    {
+      "epoch": 2.101139829828223,
+      "grad_norm": 0.7943243980407715,
+      "learning_rate": 6.4391428532083e-05,
+      "loss": 0.324,
+      "step": 6544
+    },
+    {
+      "epoch": 2.1014609086530744,
+      "grad_norm": 0.8991156816482544,
+      "learning_rate": 6.435881212867493e-05,
+      "loss": 0.3999,
+      "step": 6545
+    },
+    {
+      "epoch": 2.101781987477926,
+      "grad_norm": 0.8025935292243958,
+      "learning_rate": 6.432620006803746e-05,
+      "loss": 0.3236,
+      "step": 6546
+    },
+    {
+      "epoch": 2.1021030663027775,
+      "grad_norm": 0.795685350894928,
+      "learning_rate": 6.429359235414436e-05,
+      "loss": 0.3178,
+      "step": 6547
+    },
+    {
+      "epoch": 2.1024241451276287,
+      "grad_norm": 0.8055205345153809,
+      "learning_rate": 6.426098899096868e-05,
+      "loss": 0.3545,
+      "step": 6548
+    },
+    {
+      "epoch": 2.10274522395248,
+      "grad_norm": 0.9374400973320007,
+      "learning_rate": 6.422838998248307e-05,
+      "loss": 0.4094,
+      "step": 6549
+    },
+    {
+      "epoch": 2.1030663027773318,
+      "grad_norm": 1.0976170301437378,
+      "learning_rate": 6.419579533265968e-05,
+      "loss": 0.3392,
+      "step": 6550
+    },
+    {
+      "epoch": 2.1033873816021833,
+      "grad_norm": 0.8357941508293152,
+      "learning_rate": 6.416320504546997e-05,
+      "loss": 0.3858,
+      "step": 6551
+    },
+    {
+      "epoch": 2.103708460427035,
+      "grad_norm": 1.0488425493240356,
+      "learning_rate": 6.413061912488508e-05,
+      "loss": 0.4397,
+      "step": 6552
+    },
+    {
+      "epoch": 2.1040295392518864,
+      "grad_norm": 0.8378785848617554,
+      "learning_rate": 6.409803757487538e-05,
+      "loss": 0.3655,
+      "step": 6553
+    },
+    {
+      "epoch": 2.104350618076738,
+      "grad_norm": 1.1138861179351807,
+      "learning_rate": 6.406546039941094e-05,
+      "loss": 0.3253,
+      "step": 6554
+    },
+    {
+      "epoch": 2.1046716969015895,
+      "grad_norm": 1.0650768280029297,
+      "learning_rate": 6.403288760246112e-05,
+      "loss": 0.3725,
+      "step": 6555
+    },
+    {
+      "epoch": 2.1049927757264406,
+      "grad_norm": 1.272623896598816,
+      "learning_rate": 6.40003191879949e-05,
+      "loss": 0.3764,
+      "step": 6556
+    },
+    {
+      "epoch": 2.105313854551292,
+      "grad_norm": 1.197476863861084,
+      "learning_rate": 6.396775515998055e-05,
+      "loss": 0.3797,
+      "step": 6557
+    },
+    {
+      "epoch": 2.1056349333761437,
+      "grad_norm": 0.9312015771865845,
+      "learning_rate": 6.393519552238591e-05,
+      "loss": 0.3614,
+      "step": 6558
+    },
+    {
+      "epoch": 2.1059560122009953,
+      "grad_norm": 1.1965104341506958,
+      "learning_rate": 6.390264027917836e-05,
+      "loss": 0.4493,
+      "step": 6559
+    },
+    {
+      "epoch": 2.106277091025847,
+      "grad_norm": 0.7930313348770142,
+      "learning_rate": 6.387008943432455e-05,
+      "loss": 0.3172,
+      "step": 6560
+    },
+    {
+      "epoch": 2.1065981698506984,
+      "grad_norm": 0.645929753780365,
+      "learning_rate": 6.383754299179079e-05,
+      "loss": 0.3191,
+      "step": 6561
+    },
+    {
+      "epoch": 2.10691924867555,
+      "grad_norm": 0.7615616917610168,
+      "learning_rate": 6.380500095554268e-05,
+      "loss": 0.2763,
+      "step": 6562
+    },
+    {
+      "epoch": 2.1072403275004015,
+      "grad_norm": 1.1042746305465698,
+      "learning_rate": 6.377246332954544e-05,
+      "loss": 0.3695,
+      "step": 6563
+    },
+    {
+      "epoch": 2.107561406325253,
+      "grad_norm": 0.7434027791023254,
+      "learning_rate": 6.373993011776367e-05,
+      "loss": 0.3348,
+      "step": 6564
+    },
+    {
+      "epoch": 2.107882485150104,
+      "grad_norm": 0.8379579782485962,
+      "learning_rate": 6.370740132416138e-05,
+      "loss": 0.3715,
+      "step": 6565
+    },
+    {
+      "epoch": 2.1082035639749557,
+      "grad_norm": 0.7400662899017334,
+      "learning_rate": 6.367487695270217e-05,
+      "loss": 0.3288,
+      "step": 6566
+    },
+    {
+      "epoch": 2.1085246427998072,
+      "grad_norm": 0.8709506392478943,
+      "learning_rate": 6.364235700734903e-05,
+      "loss": 0.3382,
+      "step": 6567
+    },
+    {
+      "epoch": 2.1088457216246588,
+      "grad_norm": 1.0580693483352661,
+      "learning_rate": 6.360984149206439e-05,
+      "loss": 0.3497,
+      "step": 6568
+    },
+    {
+      "epoch": 2.1091668004495103,
+      "grad_norm": 1.0599806308746338,
+      "learning_rate": 6.357733041081018e-05,
+      "loss": 0.3606,
+      "step": 6569
+    },
+    {
+      "epoch": 2.109487879274362,
+      "grad_norm": 0.7007160782814026,
+      "learning_rate": 6.35448237675478e-05,
+      "loss": 0.2877,
+      "step": 6570
+    },
+    {
+      "epoch": 2.1098089580992134,
+      "grad_norm": 0.9604334831237793,
+      "learning_rate": 6.351232156623803e-05,
+      "loss": 0.3173,
+      "step": 6571
+    },
+    {
+      "epoch": 2.110130036924065,
+      "grad_norm": 0.7274121642112732,
+      "learning_rate": 6.347982381084123e-05,
+      "loss": 0.2848,
+      "step": 6572
+    },
+    {
+      "epoch": 2.1104511157489165,
+      "grad_norm": 0.8316487669944763,
+      "learning_rate": 6.344733050531713e-05,
+      "loss": 0.3307,
+      "step": 6573
+    },
+    {
+      "epoch": 2.1107721945737676,
+      "grad_norm": 0.6824556589126587,
+      "learning_rate": 6.341484165362487e-05,
+      "loss": 0.2692,
+      "step": 6574
+    },
+    {
+      "epoch": 2.111093273398619,
+      "grad_norm": 0.9442518949508667,
+      "learning_rate": 6.338235725972325e-05,
+      "loss": 0.326,
+      "step": 6575
+    },
+    {
+      "epoch": 2.1114143522234707,
+      "grad_norm": 0.46426621079444885,
+      "learning_rate": 6.334987732757029e-05,
+      "loss": 0.2598,
+      "step": 6576
+    },
+    {
+      "epoch": 2.1117354310483223,
+      "grad_norm": 0.6108339428901672,
+      "learning_rate": 6.33174018611236e-05,
+      "loss": 0.3211,
+      "step": 6577
+    },
+    {
+      "epoch": 2.112056509873174,
+      "grad_norm": 0.8628884553909302,
+      "learning_rate": 6.328493086434023e-05,
+      "loss": 0.3061,
+      "step": 6578
+    },
+    {
+      "epoch": 2.1123775886980254,
+      "grad_norm": 0.45920509099960327,
+      "learning_rate": 6.325246434117668e-05,
+      "loss": 0.2858,
+      "step": 6579
+    },
+    {
+      "epoch": 2.112698667522877,
+      "grad_norm": 0.36384066939353943,
+      "learning_rate": 6.322000229558887e-05,
+      "loss": 0.3652,
+      "step": 6580
+    },
+    {
+      "epoch": 2.1130197463477285,
+      "grad_norm": 0.6161776781082153,
+      "learning_rate": 6.318754473153221e-05,
+      "loss": 0.4469,
+      "step": 6581
+    },
+    {
+      "epoch": 2.11334082517258,
+      "grad_norm": 0.5256691575050354,
+      "learning_rate": 6.315509165296159e-05,
+      "loss": 0.2341,
+      "step": 6582
+    },
+    {
+      "epoch": 2.113661903997431,
+      "grad_norm": 0.5201000571250916,
+      "learning_rate": 6.312264306383124e-05,
+      "loss": 0.2947,
+      "step": 6583
+    },
+    {
+      "epoch": 2.1139829828222827,
+      "grad_norm": 0.5829021334648132,
+      "learning_rate": 6.309019896809503e-05,
+      "loss": 0.2341,
+      "step": 6584
+    },
+    {
+      "epoch": 2.1143040616471342,
+      "grad_norm": 0.5891327261924744,
+      "learning_rate": 6.305775936970606e-05,
+      "loss": 0.1354,
+      "step": 6585
+    },
+    {
+      "epoch": 2.114625140471986,
+      "grad_norm": 0.4353516697883606,
+      "learning_rate": 6.302532427261709e-05,
+      "loss": 0.1752,
+      "step": 6586
+    },
+    {
+      "epoch": 2.1149462192968373,
+      "grad_norm": 0.6292280554771423,
+      "learning_rate": 6.299289368078016e-05,
+      "loss": 0.327,
+      "step": 6587
+    },
+    {
+      "epoch": 2.115267298121689,
+      "grad_norm": 0.7321916818618774,
+      "learning_rate": 6.296046759814694e-05,
+      "loss": 0.3184,
+      "step": 6588
+    },
+    {
+      "epoch": 2.1155883769465404,
+      "grad_norm": 1.0000382661819458,
+      "learning_rate": 6.292804602866834e-05,
+      "loss": 0.3432,
+      "step": 6589
+    },
+    {
+      "epoch": 2.115909455771392,
+      "grad_norm": 0.9078216552734375,
+      "learning_rate": 6.289562897629492e-05,
+      "loss": 0.4655,
+      "step": 6590
+    },
+    {
+      "epoch": 2.1162305345962436,
+      "grad_norm": 0.827180802822113,
+      "learning_rate": 6.286321644497655e-05,
+      "loss": 0.369,
+      "step": 6591
+    },
+    {
+      "epoch": 2.1165516134210947,
+      "grad_norm": 0.5973712205886841,
+      "learning_rate": 6.283080843866256e-05,
+      "loss": 0.2657,
+      "step": 6592
+    },
+    {
+      "epoch": 2.116872692245946,
+      "grad_norm": 0.7066929936408997,
+      "learning_rate": 6.27984049613019e-05,
+      "loss": 0.3177,
+      "step": 6593
+    },
+    {
+      "epoch": 2.1171937710707978,
+      "grad_norm": 0.8273329734802246,
+      "learning_rate": 6.276600601684267e-05,
+      "loss": 0.3679,
+      "step": 6594
+    },
+    {
+      "epoch": 2.1175148498956493,
+      "grad_norm": 0.9276707768440247,
+      "learning_rate": 6.27336116092327e-05,
+      "loss": 0.4267,
+      "step": 6595
+    },
+    {
+      "epoch": 2.117835928720501,
+      "grad_norm": 0.9455394744873047,
+      "learning_rate": 6.27012217424191e-05,
+      "loss": 0.3272,
+      "step": 6596
+    },
+    {
+      "epoch": 2.1181570075453524,
+      "grad_norm": 1.1003650426864624,
+      "learning_rate": 6.266883642034853e-05,
+      "loss": 0.4239,
+      "step": 6597
+    },
+    {
+      "epoch": 2.118478086370204,
+      "grad_norm": 1.107857346534729,
+      "learning_rate": 6.263645564696696e-05,
+      "loss": 0.436,
+      "step": 6598
+    },
+    {
+      "epoch": 2.1187991651950555,
+      "grad_norm": 0.7510982155799866,
+      "learning_rate": 6.260407942621998e-05,
+      "loss": 0.3712,
+      "step": 6599
+    },
+    {
+      "epoch": 2.119120244019907,
+      "grad_norm": 0.7718844413757324,
+      "learning_rate": 6.257170776205245e-05,
+      "loss": 0.3182,
+      "step": 6600
+    },
+    {
+      "epoch": 2.119441322844758,
+      "grad_norm": 0.8712736964225769,
+      "learning_rate": 6.25393406584088e-05,
+      "loss": 0.3044,
+      "step": 6601
+    },
+    {
+      "epoch": 2.1197624016696097,
+      "grad_norm": 1.5617177486419678,
+      "learning_rate": 6.25069781192329e-05,
+      "loss": 0.42,
+      "step": 6602
+    },
+    {
+      "epoch": 2.1200834804944613,
+      "grad_norm": 0.9841622710227966,
+      "learning_rate": 6.247462014846792e-05,
+      "loss": 0.399,
+      "step": 6603
+    },
+    {
+      "epoch": 2.120404559319313,
+      "grad_norm": 0.9455597400665283,
+      "learning_rate": 6.24422667500567e-05,
+      "loss": 0.4426,
+      "step": 6604
+    },
+    {
+      "epoch": 2.1207256381441644,
+      "grad_norm": 0.6564199924468994,
+      "learning_rate": 6.240991792794133e-05,
+      "loss": 0.2695,
+      "step": 6605
+    },
+    {
+      "epoch": 2.121046716969016,
+      "grad_norm": 0.9546607732772827,
+      "learning_rate": 6.237757368606345e-05,
+      "loss": 0.4005,
+      "step": 6606
+    },
+    {
+      "epoch": 2.1213677957938675,
+      "grad_norm": 0.9721749424934387,
+      "learning_rate": 6.234523402836407e-05,
+      "loss": 0.4454,
+      "step": 6607
+    },
+    {
+      "epoch": 2.121688874618719,
+      "grad_norm": 1.2680310010910034,
+      "learning_rate": 6.231289895878375e-05,
+      "loss": 0.3899,
+      "step": 6608
+    },
+    {
+      "epoch": 2.1220099534435706,
+      "grad_norm": 0.8545310497283936,
+      "learning_rate": 6.228056848126236e-05,
+      "loss": 0.3709,
+      "step": 6609
+    },
+    {
+      "epoch": 2.1223310322684217,
+      "grad_norm": 0.8629288077354431,
+      "learning_rate": 6.224824259973925e-05,
+      "loss": 0.2811,
+      "step": 6610
+    },
+    {
+      "epoch": 2.1226521110932732,
+      "grad_norm": 1.1735135316848755,
+      "learning_rate": 6.22159213181533e-05,
+      "loss": 0.3109,
+      "step": 6611
+    },
+    {
+      "epoch": 2.122973189918125,
+      "grad_norm": 0.7792378067970276,
+      "learning_rate": 6.21836046404427e-05,
+      "loss": 0.3302,
+      "step": 6612
+    },
+    {
+      "epoch": 2.1232942687429763,
+      "grad_norm": 0.7332779169082642,
+      "learning_rate": 6.215129257054522e-05,
+      "loss": 0.3134,
+      "step": 6613
+    },
+    {
+      "epoch": 2.123615347567828,
+      "grad_norm": 0.6187347769737244,
+      "learning_rate": 6.211898511239787e-05,
+      "loss": 0.2562,
+      "step": 6614
+    },
+    {
+      "epoch": 2.1239364263926794,
+      "grad_norm": 0.8484505414962769,
+      "learning_rate": 6.20866822699373e-05,
+      "loss": 0.2516,
+      "step": 6615
+    },
+    {
+      "epoch": 2.124257505217531,
+      "grad_norm": 0.8732132911682129,
+      "learning_rate": 6.205438404709947e-05,
+      "loss": 0.3172,
+      "step": 6616
+    },
+    {
+      "epoch": 2.1245785840423825,
+      "grad_norm": 0.8905081152915955,
+      "learning_rate": 6.20220904478199e-05,
+      "loss": 0.3439,
+      "step": 6617
+    },
+    {
+      "epoch": 2.124899662867234,
+      "grad_norm": 0.7029334306716919,
+      "learning_rate": 6.198980147603339e-05,
+      "loss": 0.2728,
+      "step": 6618
+    },
+    {
+      "epoch": 2.125220741692085,
+      "grad_norm": 0.8085559010505676,
+      "learning_rate": 6.195751713567427e-05,
+      "loss": 0.2663,
+      "step": 6619
+    },
+    {
+      "epoch": 2.1255418205169367,
+      "grad_norm": 0.8118886351585388,
+      "learning_rate": 6.192523743067626e-05,
+      "loss": 0.3628,
+      "step": 6620
+    },
+    {
+      "epoch": 2.1258628993417883,
+      "grad_norm": 0.5868399143218994,
+      "learning_rate": 6.18929623649726e-05,
+      "loss": 0.2687,
+      "step": 6621
+    },
+    {
+      "epoch": 2.12618397816664,
+      "grad_norm": 0.5492334961891174,
+      "learning_rate": 6.18606919424959e-05,
+      "loss": 0.2749,
+      "step": 6622
+    },
+    {
+      "epoch": 2.1265050569914914,
+      "grad_norm": 1.074462652206421,
+      "learning_rate": 6.182842616717817e-05,
+      "loss": 0.3686,
+      "step": 6623
+    },
+    {
+      "epoch": 2.126826135816343,
+      "grad_norm": 0.8293436765670776,
+      "learning_rate": 6.179616504295092e-05,
+      "loss": 0.3272,
+      "step": 6624
+    },
+    {
+      "epoch": 2.1271472146411945,
+      "grad_norm": 1.137435793876648,
+      "learning_rate": 6.176390857374507e-05,
+      "loss": 0.3109,
+      "step": 6625
+    },
+    {
+      "epoch": 2.127468293466046,
+      "grad_norm": 0.7483497858047485,
+      "learning_rate": 6.173165676349103e-05,
+      "loss": 0.343,
+      "step": 6626
+    },
+    {
+      "epoch": 2.1277893722908976,
+      "grad_norm": 0.40780845284461975,
+      "learning_rate": 6.169940961611854e-05,
+      "loss": 0.2489,
+      "step": 6627
+    },
+    {
+      "epoch": 2.1281104511157487,
+      "grad_norm": 0.6803835034370422,
+      "learning_rate": 6.166716713555674e-05,
+      "loss": 0.3458,
+      "step": 6628
+    },
+    {
+      "epoch": 2.1284315299406003,
+      "grad_norm": 0.6361788511276245,
+      "learning_rate": 6.163492932573438e-05,
+      "loss": 0.3,
+      "step": 6629
+    },
+    {
+      "epoch": 2.128752608765452,
+      "grad_norm": 0.5756948590278625,
+      "learning_rate": 6.160269619057951e-05,
+      "loss": 0.6565,
+      "step": 6630
+    },
+    {
+      "epoch": 2.1290736875903034,
+      "grad_norm": 0.5994390845298767,
+      "learning_rate": 6.157046773401964e-05,
+      "loss": 0.4455,
+      "step": 6631
+    },
+    {
+      "epoch": 2.129394766415155,
+      "grad_norm": 0.661205530166626,
+      "learning_rate": 6.153824395998168e-05,
+      "loss": 0.1752,
+      "step": 6632
+    },
+    {
+      "epoch": 2.1297158452400065,
+      "grad_norm": 0.5093392729759216,
+      "learning_rate": 6.150602487239206e-05,
+      "loss": 0.0911,
+      "step": 6633
+    },
+    {
+      "epoch": 2.130036924064858,
+      "grad_norm": 0.4452694356441498,
+      "learning_rate": 6.147381047517655e-05,
+      "loss": 0.0949,
+      "step": 6634
+    },
+    {
+      "epoch": 2.1303580028897096,
+      "grad_norm": 0.4948364496231079,
+      "learning_rate": 6.144160077226036e-05,
+      "loss": 0.1567,
+      "step": 6635
+    },
+    {
+      "epoch": 2.130679081714561,
+      "grad_norm": 0.5426846742630005,
+      "learning_rate": 6.140939576756817e-05,
+      "loss": 0.2952,
+      "step": 6636
+    },
+    {
+      "epoch": 2.131000160539412,
+      "grad_norm": 0.6515873074531555,
+      "learning_rate": 6.1377195465024e-05,
+      "loss": 0.3118,
+      "step": 6637
+    },
+    {
+      "epoch": 2.1313212393642638,
+      "grad_norm": 0.7269318103790283,
+      "learning_rate": 6.134499986855144e-05,
+      "loss": 0.3818,
+      "step": 6638
+    },
+    {
+      "epoch": 2.1316423181891153,
+      "grad_norm": 0.7911025285720825,
+      "learning_rate": 6.131280898207339e-05,
+      "loss": 0.3881,
+      "step": 6639
+    },
+    {
+      "epoch": 2.131963397013967,
+      "grad_norm": 0.8551725149154663,
+      "learning_rate": 6.128062280951224e-05,
+      "loss": 0.3669,
+      "step": 6640
+    },
+    {
+      "epoch": 2.1322844758388184,
+      "grad_norm": 0.8696125745773315,
+      "learning_rate": 6.12484413547897e-05,
+      "loss": 0.3246,
+      "step": 6641
+    },
+    {
+      "epoch": 2.13260555466367,
+      "grad_norm": 0.9644296765327454,
+      "learning_rate": 6.121626462182707e-05,
+      "loss": 0.3639,
+      "step": 6642
+    },
+    {
+      "epoch": 2.1329266334885215,
+      "grad_norm": 0.9500702619552612,
+      "learning_rate": 6.118409261454494e-05,
+      "loss": 0.2879,
+      "step": 6643
+    },
+    {
+      "epoch": 2.133247712313373,
+      "grad_norm": 0.942509114742279,
+      "learning_rate": 6.11519253368634e-05,
+      "loss": 0.3482,
+      "step": 6644
+    },
+    {
+      "epoch": 2.1335687911382246,
+      "grad_norm": 0.7259782552719116,
+      "learning_rate": 6.111976279270192e-05,
+      "loss": 0.3214,
+      "step": 6645
+    },
+    {
+      "epoch": 2.1338898699630757,
+      "grad_norm": 0.7801980972290039,
+      "learning_rate": 6.108760498597938e-05,
+      "loss": 0.3301,
+      "step": 6646
+    },
+    {
+      "epoch": 2.1342109487879273,
+      "grad_norm": 0.8624287247657776,
+      "learning_rate": 6.105545192061416e-05,
+      "loss": 0.3717,
+      "step": 6647
+    },
+    {
+      "epoch": 2.134532027612779,
+      "grad_norm": 0.8877105712890625,
+      "learning_rate": 6.1023303600523975e-05,
+      "loss": 0.4454,
+      "step": 6648
+    },
+    {
+      "epoch": 2.1348531064376304,
+      "grad_norm": 0.7793693542480469,
+      "learning_rate": 6.099116002962604e-05,
+      "loss": 0.3266,
+      "step": 6649
+    },
+    {
+      "epoch": 2.135174185262482,
+      "grad_norm": 1.096024990081787,
+      "learning_rate": 6.09590212118369e-05,
+      "loss": 0.4317,
+      "step": 6650
+    },
+    {
+      "epoch": 2.1354952640873335,
+      "grad_norm": 0.889217734336853,
+      "learning_rate": 6.092688715107264e-05,
+      "loss": 0.4424,
+      "step": 6651
+    },
+    {
+      "epoch": 2.135816342912185,
+      "grad_norm": 0.8628430962562561,
+      "learning_rate": 6.089475785124863e-05,
+      "loss": 0.3435,
+      "step": 6652
+    },
+    {
+      "epoch": 2.1361374217370366,
+      "grad_norm": 0.8696472644805908,
+      "learning_rate": 6.086263331627976e-05,
+      "loss": 0.3443,
+      "step": 6653
+    },
+    {
+      "epoch": 2.136458500561888,
+      "grad_norm": 0.9807567596435547,
+      "learning_rate": 6.0830513550080335e-05,
+      "loss": 0.3259,
+      "step": 6654
+    },
+    {
+      "epoch": 2.1367795793867392,
+      "grad_norm": 0.9957831501960754,
+      "learning_rate": 6.0798398556563976e-05,
+      "loss": 0.3485,
+      "step": 6655
+    },
+    {
+      "epoch": 2.137100658211591,
+      "grad_norm": 1.0580285787582397,
+      "learning_rate": 6.076628833964388e-05,
+      "loss": 0.4838,
+      "step": 6656
+    },
+    {
+      "epoch": 2.1374217370364423,
+      "grad_norm": 0.7610054016113281,
+      "learning_rate": 6.073418290323251e-05,
+      "loss": 0.3736,
+      "step": 6657
+    },
+    {
+      "epoch": 2.137742815861294,
+      "grad_norm": 0.8939658999443054,
+      "learning_rate": 6.070208225124185e-05,
+      "loss": 0.3186,
+      "step": 6658
+    },
+    {
+      "epoch": 2.1380638946861454,
+      "grad_norm": 0.9913893342018127,
+      "learning_rate": 6.066998638758326e-05,
+      "loss": 0.3687,
+      "step": 6659
+    },
+    {
+      "epoch": 2.138384973510997,
+      "grad_norm": 0.6824525594711304,
+      "learning_rate": 6.063789531616757e-05,
+      "loss": 0.277,
+      "step": 6660
+    },
+    {
+      "epoch": 2.1387060523358485,
+      "grad_norm": 0.935042679309845,
+      "learning_rate": 6.0605809040904894e-05,
+      "loss": 0.3631,
+      "step": 6661
+    },
+    {
+      "epoch": 2.1390271311607,
+      "grad_norm": 0.7435742020606995,
+      "learning_rate": 6.05737275657049e-05,
+      "loss": 0.3319,
+      "step": 6662
+    },
+    {
+      "epoch": 2.1393482099855516,
+      "grad_norm": 0.6883519291877747,
+      "learning_rate": 6.054165089447663e-05,
+      "loss": 0.3006,
+      "step": 6663
+    },
+    {
+      "epoch": 2.1396692888104027,
+      "grad_norm": 0.7741847038269043,
+      "learning_rate": 6.0509579031128485e-05,
+      "loss": 0.3047,
+      "step": 6664
+    },
+    {
+      "epoch": 2.1399903676352543,
+      "grad_norm": 0.8180326223373413,
+      "learning_rate": 6.047751197956838e-05,
+      "loss": 0.3593,
+      "step": 6665
+    },
+    {
+      "epoch": 2.140311446460106,
+      "grad_norm": 1.200945496559143,
+      "learning_rate": 6.0445449743703516e-05,
+      "loss": 0.3234,
+      "step": 6666
+    },
+    {
+      "epoch": 2.1406325252849574,
+      "grad_norm": 0.9025703072547913,
+      "learning_rate": 6.0413392327440635e-05,
+      "loss": 0.3599,
+      "step": 6667
+    },
+    {
+      "epoch": 2.140953604109809,
+      "grad_norm": 0.9718348979949951,
+      "learning_rate": 6.03813397346858e-05,
+      "loss": 0.3025,
+      "step": 6668
+    },
+    {
+      "epoch": 2.1412746829346605,
+      "grad_norm": 0.9838557243347168,
+      "learning_rate": 6.034929196934459e-05,
+      "loss": 0.3157,
+      "step": 6669
+    },
+    {
+      "epoch": 2.141595761759512,
+      "grad_norm": 0.940606951713562,
+      "learning_rate": 6.031724903532183e-05,
+      "loss": 0.3326,
+      "step": 6670
+    },
+    {
+      "epoch": 2.1419168405843636,
+      "grad_norm": 0.6957113742828369,
+      "learning_rate": 6.0285210936521955e-05,
+      "loss": 0.2739,
+      "step": 6671
+    },
+    {
+      "epoch": 2.142237919409215,
+      "grad_norm": 0.977173388004303,
+      "learning_rate": 6.025317767684864e-05,
+      "loss": 0.3673,
+      "step": 6672
+    },
+    {
+      "epoch": 2.1425589982340663,
+      "grad_norm": 0.6859129071235657,
+      "learning_rate": 6.022114926020504e-05,
+      "loss": 0.3025,
+      "step": 6673
+    },
+    {
+      "epoch": 2.142880077058918,
+      "grad_norm": 0.8257904052734375,
+      "learning_rate": 6.018912569049376e-05,
+      "loss": 0.3357,
+      "step": 6674
+    },
+    {
+      "epoch": 2.1432011558837694,
+      "grad_norm": 0.5588051080703735,
+      "learning_rate": 6.015710697161674e-05,
+      "loss": 0.2849,
+      "step": 6675
+    },
+    {
+      "epoch": 2.143522234708621,
+      "grad_norm": 0.5934134125709534,
+      "learning_rate": 6.012509310747538e-05,
+      "loss": 0.2849,
+      "step": 6676
+    },
+    {
+      "epoch": 2.1438433135334725,
+      "grad_norm": 0.800035834312439,
+      "learning_rate": 6.009308410197047e-05,
+      "loss": 0.2704,
+      "step": 6677
+    },
+    {
+      "epoch": 2.144164392358324,
+      "grad_norm": 0.42928749322891235,
+      "learning_rate": 6.006107995900224e-05,
+      "loss": 0.2584,
+      "step": 6678
+    },
+    {
+      "epoch": 2.1444854711831756,
+      "grad_norm": 0.4132596552371979,
+      "learning_rate": 6.002908068247024e-05,
+      "loss": 0.2802,
+      "step": 6679
+    },
+    {
+      "epoch": 2.144806550008027,
+      "grad_norm": 0.4437423348426819,
+      "learning_rate": 5.999708627627354e-05,
+      "loss": 0.4253,
+      "step": 6680
+    },
+    {
+      "epoch": 2.1451276288328787,
+      "grad_norm": 0.4113101065158844,
+      "learning_rate": 5.9965096744310526e-05,
+      "loss": 0.2371,
+      "step": 6681
+    },
+    {
+      "epoch": 2.1454487076577298,
+      "grad_norm": 0.513230562210083,
+      "learning_rate": 5.9933112090479006e-05,
+      "loss": 0.1586,
+      "step": 6682
+    },
+    {
+      "epoch": 2.1457697864825813,
+      "grad_norm": 0.5119988918304443,
+      "learning_rate": 5.990113231867629e-05,
+      "loss": 0.1865,
+      "step": 6683
+    },
+    {
+      "epoch": 2.146090865307433,
+      "grad_norm": 0.6047758460044861,
+      "learning_rate": 5.9869157432798926e-05,
+      "loss": 0.2309,
+      "step": 6684
+    },
+    {
+      "epoch": 2.1464119441322844,
+      "grad_norm": 0.5919579267501831,
+      "learning_rate": 5.9837187436743016e-05,
+      "loss": 0.1818,
+      "step": 6685
+    },
+    {
+      "epoch": 2.146733022957136,
+      "grad_norm": 0.44815778732299805,
+      "learning_rate": 5.9805222334404e-05,
+      "loss": 0.1748,
+      "step": 6686
+    },
+    {
+      "epoch": 2.1470541017819875,
+      "grad_norm": 0.4934236705303192,
+      "learning_rate": 5.977326212967671e-05,
+      "loss": 0.1963,
+      "step": 6687
+    },
+    {
+      "epoch": 2.147375180606839,
+      "grad_norm": 0.8252405524253845,
+      "learning_rate": 5.974130682645538e-05,
+      "loss": 0.3972,
+      "step": 6688
+    },
+    {
+      "epoch": 2.1476962594316906,
+      "grad_norm": 1.336822509765625,
+      "learning_rate": 5.9709356428633746e-05,
+      "loss": 0.3795,
+      "step": 6689
+    },
+    {
+      "epoch": 2.148017338256542,
+      "grad_norm": 0.7905248999595642,
+      "learning_rate": 5.967741094010479e-05,
+      "loss": 0.4087,
+      "step": 6690
+    },
+    {
+      "epoch": 2.1483384170813933,
+      "grad_norm": 0.7174348831176758,
+      "learning_rate": 5.964547036476099e-05,
+      "loss": 0.3429,
+      "step": 6691
+    },
+    {
+      "epoch": 2.148659495906245,
+      "grad_norm": 0.7305684685707092,
+      "learning_rate": 5.961353470649426e-05,
+      "loss": 0.2866,
+      "step": 6692
+    },
+    {
+      "epoch": 2.1489805747310964,
+      "grad_norm": 0.6904447078704834,
+      "learning_rate": 5.9581603969195766e-05,
+      "loss": 0.3,
+      "step": 6693
+    },
+    {
+      "epoch": 2.149301653555948,
+      "grad_norm": 0.8633493185043335,
+      "learning_rate": 5.954967815675627e-05,
+      "loss": 0.3736,
+      "step": 6694
+    },
+    {
+      "epoch": 2.1496227323807995,
+      "grad_norm": 0.8246678113937378,
+      "learning_rate": 5.951775727306577e-05,
+      "loss": 0.359,
+      "step": 6695
+    },
+    {
+      "epoch": 2.149943811205651,
+      "grad_norm": 1.400453805923462,
+      "learning_rate": 5.9485841322013755e-05,
+      "loss": 0.454,
+      "step": 6696
+    },
+    {
+      "epoch": 2.1502648900305026,
+      "grad_norm": 2.0813472270965576,
+      "learning_rate": 5.94539303074891e-05,
+      "loss": 0.3139,
+      "step": 6697
+    },
+    {
+      "epoch": 2.150585968855354,
+      "grad_norm": 0.9798773527145386,
+      "learning_rate": 5.9422024233380013e-05,
+      "loss": 0.4055,
+      "step": 6698
+    },
+    {
+      "epoch": 2.1509070476802057,
+      "grad_norm": 0.9480912685394287,
+      "learning_rate": 5.9390123103574215e-05,
+      "loss": 0.3727,
+      "step": 6699
+    },
+    {
+      "epoch": 2.151228126505057,
+      "grad_norm": 0.9623056650161743,
+      "learning_rate": 5.935822692195869e-05,
+      "loss": 0.2921,
+      "step": 6700
+    },
+    {
+      "epoch": 2.1515492053299083,
+      "grad_norm": 0.8746601343154907,
+      "learning_rate": 5.9326335692419995e-05,
+      "loss": 0.3345,
+      "step": 6701
+    },
+    {
+      "epoch": 2.15187028415476,
+      "grad_norm": 0.725217878818512,
+      "learning_rate": 5.929444941884388e-05,
+      "loss": 0.346,
+      "step": 6702
+    },
+    {
+      "epoch": 2.1521913629796114,
+      "grad_norm": 0.8172053098678589,
+      "learning_rate": 5.9262568105115654e-05,
+      "loss": 0.3468,
+      "step": 6703
+    },
+    {
+      "epoch": 2.152512441804463,
+      "grad_norm": 0.648247241973877,
+      "learning_rate": 5.92306917551199e-05,
+      "loss": 0.2908,
+      "step": 6704
+    },
+    {
+      "epoch": 2.1528335206293145,
+      "grad_norm": 0.7141852378845215,
+      "learning_rate": 5.9198820372740726e-05,
+      "loss": 0.2882,
+      "step": 6705
+    },
+    {
+      "epoch": 2.153154599454166,
+      "grad_norm": 0.8479703068733215,
+      "learning_rate": 5.916695396186154e-05,
+      "loss": 0.3407,
+      "step": 6706
+    },
+    {
+      "epoch": 2.1534756782790176,
+      "grad_norm": 0.8146336674690247,
+      "learning_rate": 5.913509252636511e-05,
+      "loss": 0.2869,
+      "step": 6707
+    },
+    {
+      "epoch": 2.153796757103869,
+      "grad_norm": 0.6545037627220154,
+      "learning_rate": 5.910323607013373e-05,
+      "loss": 0.2742,
+      "step": 6708
+    },
+    {
+      "epoch": 2.1541178359287203,
+      "grad_norm": 0.7448673248291016,
+      "learning_rate": 5.907138459704895e-05,
+      "loss": 0.3192,
+      "step": 6709
+    },
+    {
+      "epoch": 2.154438914753572,
+      "grad_norm": 0.7796539664268494,
+      "learning_rate": 5.903953811099183e-05,
+      "loss": 0.2947,
+      "step": 6710
+    },
+    {
+      "epoch": 2.1547599935784234,
+      "grad_norm": 1.0737991333007812,
+      "learning_rate": 5.900769661584272e-05,
+      "loss": 0.4188,
+      "step": 6711
+    },
+    {
+      "epoch": 2.155081072403275,
+      "grad_norm": 0.6433249711990356,
+      "learning_rate": 5.8975860115481487e-05,
+      "loss": 0.2853,
+      "step": 6712
+    },
+    {
+      "epoch": 2.1554021512281265,
+      "grad_norm": 0.8667881488800049,
+      "learning_rate": 5.8944028613787206e-05,
+      "loss": 0.3175,
+      "step": 6713
+    },
+    {
+      "epoch": 2.155723230052978,
+      "grad_norm": 1.061166524887085,
+      "learning_rate": 5.8912202114638524e-05,
+      "loss": 0.4216,
+      "step": 6714
+    },
+    {
+      "epoch": 2.1560443088778296,
+      "grad_norm": 0.9208434224128723,
+      "learning_rate": 5.88803806219134e-05,
+      "loss": 0.3846,
+      "step": 6715
+    },
+    {
+      "epoch": 2.156365387702681,
+      "grad_norm": 0.6994331479072571,
+      "learning_rate": 5.884856413948913e-05,
+      "loss": 0.2805,
+      "step": 6716
+    },
+    {
+      "epoch": 2.1566864665275327,
+      "grad_norm": 0.9039559960365295,
+      "learning_rate": 5.881675267124254e-05,
+      "loss": 0.3893,
+      "step": 6717
+    },
+    {
+      "epoch": 2.157007545352384,
+      "grad_norm": 0.8214314579963684,
+      "learning_rate": 5.8784946221049666e-05,
+      "loss": 0.3455,
+      "step": 6718
+    },
+    {
+      "epoch": 2.1573286241772354,
+      "grad_norm": 0.5890683531761169,
+      "learning_rate": 5.8753144792786096e-05,
+      "loss": 0.2904,
+      "step": 6719
+    },
+    {
+      "epoch": 2.157649703002087,
+      "grad_norm": 0.8071637749671936,
+      "learning_rate": 5.87213483903267e-05,
+      "loss": 0.3424,
+      "step": 6720
+    },
+    {
+      "epoch": 2.1579707818269385,
+      "grad_norm": 0.5049499869346619,
+      "learning_rate": 5.868955701754584e-05,
+      "loss": 0.2401,
+      "step": 6721
+    },
+    {
+      "epoch": 2.15829186065179,
+      "grad_norm": 0.7438219785690308,
+      "learning_rate": 5.86577706783171e-05,
+      "loss": 0.2888,
+      "step": 6722
+    },
+    {
+      "epoch": 2.1586129394766416,
+      "grad_norm": 0.5607714056968689,
+      "learning_rate": 5.862598937651365e-05,
+      "loss": 0.28,
+      "step": 6723
+    },
+    {
+      "epoch": 2.158934018301493,
+      "grad_norm": 0.6200205087661743,
+      "learning_rate": 5.8594213116007855e-05,
+      "loss": 0.3038,
+      "step": 6724
+    },
+    {
+      "epoch": 2.1592550971263447,
+      "grad_norm": 0.9494879841804504,
+      "learning_rate": 5.856244190067159e-05,
+      "loss": 0.3483,
+      "step": 6725
+    },
+    {
+      "epoch": 2.159576175951196,
+      "grad_norm": 0.9687539339065552,
+      "learning_rate": 5.853067573437612e-05,
+      "loss": 0.3126,
+      "step": 6726
+    },
+    {
+      "epoch": 2.1598972547760473,
+      "grad_norm": 0.35860729217529297,
+      "learning_rate": 5.849891462099198e-05,
+      "loss": 0.2379,
+      "step": 6727
+    },
+    {
+      "epoch": 2.160218333600899,
+      "grad_norm": 0.4721701443195343,
+      "learning_rate": 5.8467158564389234e-05,
+      "loss": 0.2996,
+      "step": 6728
+    },
+    {
+      "epoch": 2.1605394124257504,
+      "grad_norm": 2.956984758377075,
+      "learning_rate": 5.8435407568437216e-05,
+      "loss": 0.2978,
+      "step": 6729
+    },
+    {
+      "epoch": 2.160860491250602,
+      "grad_norm": 0.5004953742027283,
+      "learning_rate": 5.8403661637004736e-05,
+      "loss": 0.6088,
+      "step": 6730
+    },
+    {
+      "epoch": 2.1611815700754535,
+      "grad_norm": 0.43807387351989746,
+      "learning_rate": 5.83719207739599e-05,
+      "loss": 0.6434,
+      "step": 6731
+    },
+    {
+      "epoch": 2.161502648900305,
+      "grad_norm": 0.5199769139289856,
+      "learning_rate": 5.834018498317024e-05,
+      "loss": 0.5589,
+      "step": 6732
+    },
+    {
+      "epoch": 2.1618237277251566,
+      "grad_norm": 0.46276336908340454,
+      "learning_rate": 5.8308454268502675e-05,
+      "loss": 0.2407,
+      "step": 6733
+    },
+    {
+      "epoch": 2.162144806550008,
+      "grad_norm": 0.47607436776161194,
+      "learning_rate": 5.82767286338235e-05,
+      "loss": 0.2192,
+      "step": 6734
+    },
+    {
+      "epoch": 2.1624658853748597,
+      "grad_norm": 0.36103343963623047,
+      "learning_rate": 5.8245008082998364e-05,
+      "loss": 0.0913,
+      "step": 6735
+    },
+    {
+      "epoch": 2.162786964199711,
+      "grad_norm": 0.6850721836090088,
+      "learning_rate": 5.82132926198923e-05,
+      "loss": 0.353,
+      "step": 6736
+    },
+    {
+      "epoch": 2.1631080430245624,
+      "grad_norm": 0.7700496315956116,
+      "learning_rate": 5.818158224836987e-05,
+      "loss": 0.4083,
+      "step": 6737
+    },
+    {
+      "epoch": 2.163429121849414,
+      "grad_norm": 0.7721123695373535,
+      "learning_rate": 5.814987697229471e-05,
+      "loss": 0.382,
+      "step": 6738
+    },
+    {
+      "epoch": 2.1637502006742655,
+      "grad_norm": 0.7945840954780579,
+      "learning_rate": 5.8118176795530176e-05,
+      "loss": 0.3534,
+      "step": 6739
+    },
+    {
+      "epoch": 2.164071279499117,
+      "grad_norm": 0.5969064831733704,
+      "learning_rate": 5.808648172193868e-05,
+      "loss": 0.2527,
+      "step": 6740
+    },
+    {
+      "epoch": 2.1643923583239686,
+      "grad_norm": 0.9726594686508179,
+      "learning_rate": 5.805479175538229e-05,
+      "loss": 0.3638,
+      "step": 6741
+    },
+    {
+      "epoch": 2.16471343714882,
+      "grad_norm": 0.7358959913253784,
+      "learning_rate": 5.802310689972233e-05,
+      "loss": 0.3867,
+      "step": 6742
+    },
+    {
+      "epoch": 2.1650345159736717,
+      "grad_norm": 0.7400618195533752,
+      "learning_rate": 5.799142715881938e-05,
+      "loss": 0.2788,
+      "step": 6743
+    },
+    {
+      "epoch": 2.1653555947985232,
+      "grad_norm": 0.8120781779289246,
+      "learning_rate": 5.795975253653364e-05,
+      "loss": 0.3746,
+      "step": 6744
+    },
+    {
+      "epoch": 2.1656766736233743,
+      "grad_norm": 0.7224815487861633,
+      "learning_rate": 5.792808303672454e-05,
+      "loss": 0.3336,
+      "step": 6745
+    },
+    {
+      "epoch": 2.165997752448226,
+      "grad_norm": 0.7844706773757935,
+      "learning_rate": 5.789641866325091e-05,
+      "loss": 0.3321,
+      "step": 6746
+    },
+    {
+      "epoch": 2.1663188312730774,
+      "grad_norm": 1.100580096244812,
+      "learning_rate": 5.786475941997094e-05,
+      "loss": 0.3908,
+      "step": 6747
+    },
+    {
+      "epoch": 2.166639910097929,
+      "grad_norm": 0.7362777590751648,
+      "learning_rate": 5.783310531074223e-05,
+      "loss": 0.3151,
+      "step": 6748
+    },
+    {
+      "epoch": 2.1669609889227806,
+      "grad_norm": 0.9936489462852478,
+      "learning_rate": 5.780145633942173e-05,
+      "loss": 0.401,
+      "step": 6749
+    },
+    {
+      "epoch": 2.167282067747632,
+      "grad_norm": 0.7832114696502686,
+      "learning_rate": 5.7769812509865773e-05,
+      "loss": 0.3097,
+      "step": 6750
+    },
+    {
+      "epoch": 2.1676031465724837,
+      "grad_norm": 0.7143785357475281,
+      "learning_rate": 5.773817382593008e-05,
+      "loss": 0.2509,
+      "step": 6751
+    },
+    {
+      "epoch": 2.167924225397335,
+      "grad_norm": 0.9188450574874878,
+      "learning_rate": 5.770654029146969e-05,
+      "loss": 0.4234,
+      "step": 6752
+    },
+    {
+      "epoch": 2.1682453042221868,
+      "grad_norm": 0.7946859002113342,
+      "learning_rate": 5.7674911910339094e-05,
+      "loss": 0.28,
+      "step": 6753
+    },
+    {
+      "epoch": 2.168566383047038,
+      "grad_norm": 1.0123013257980347,
+      "learning_rate": 5.764328868639208e-05,
+      "loss": 0.3555,
+      "step": 6754
+    },
+    {
+      "epoch": 2.1688874618718894,
+      "grad_norm": 1.0456223487854004,
+      "learning_rate": 5.7611670623481864e-05,
+      "loss": 0.28,
+      "step": 6755
+    },
+    {
+      "epoch": 2.169208540696741,
+      "grad_norm": 0.765620231628418,
+      "learning_rate": 5.758005772546097e-05,
+      "loss": 0.3154,
+      "step": 6756
+    },
+    {
+      "epoch": 2.1695296195215925,
+      "grad_norm": 0.8351200222969055,
+      "learning_rate": 5.754844999618144e-05,
+      "loss": 0.3056,
+      "step": 6757
+    },
+    {
+      "epoch": 2.169850698346444,
+      "grad_norm": 0.7080745697021484,
+      "learning_rate": 5.7516847439494435e-05,
+      "loss": 0.2799,
+      "step": 6758
+    },
+    {
+      "epoch": 2.1701717771712956,
+      "grad_norm": 0.7718917727470398,
+      "learning_rate": 5.748525005925074e-05,
+      "loss": 0.3316,
+      "step": 6759
+    },
+    {
+      "epoch": 2.170492855996147,
+      "grad_norm": 0.6026411652565002,
+      "learning_rate": 5.7453657859300415e-05,
+      "loss": 0.2607,
+      "step": 6760
+    },
+    {
+      "epoch": 2.1708139348209987,
+      "grad_norm": 1.0333638191223145,
+      "learning_rate": 5.7422070843492734e-05,
+      "loss": 0.2856,
+      "step": 6761
+    },
+    {
+      "epoch": 2.1711350136458503,
+      "grad_norm": 0.8613486289978027,
+      "learning_rate": 5.739048901567665e-05,
+      "loss": 0.3556,
+      "step": 6762
+    },
+    {
+      "epoch": 2.1714560924707014,
+      "grad_norm": 0.7801258563995361,
+      "learning_rate": 5.735891237970015e-05,
+      "loss": 0.3481,
+      "step": 6763
+    },
+    {
+      "epoch": 2.171777171295553,
+      "grad_norm": 0.8479055762290955,
+      "learning_rate": 5.732734093941087e-05,
+      "loss": 0.3521,
+      "step": 6764
+    },
+    {
+      "epoch": 2.1720982501204045,
+      "grad_norm": 0.8801478743553162,
+      "learning_rate": 5.7295774698655655e-05,
+      "loss": 0.3246,
+      "step": 6765
+    },
+    {
+      "epoch": 2.172419328945256,
+      "grad_norm": 0.860466718673706,
+      "learning_rate": 5.7264213661280765e-05,
+      "loss": 0.3524,
+      "step": 6766
+    },
+    {
+      "epoch": 2.1727404077701076,
+      "grad_norm": 0.7384558320045471,
+      "learning_rate": 5.723265783113181e-05,
+      "loss": 0.2897,
+      "step": 6767
+    },
+    {
+      "epoch": 2.173061486594959,
+      "grad_norm": 0.9084320068359375,
+      "learning_rate": 5.720110721205376e-05,
+      "loss": 0.3288,
+      "step": 6768
+    },
+    {
+      "epoch": 2.1733825654198107,
+      "grad_norm": 0.7406836152076721,
+      "learning_rate": 5.716956180789098e-05,
+      "loss": 0.3548,
+      "step": 6769
+    },
+    {
+      "epoch": 2.1737036442446622,
+      "grad_norm": 0.8314934372901917,
+      "learning_rate": 5.713802162248718e-05,
+      "loss": 0.3294,
+      "step": 6770
+    },
+    {
+      "epoch": 2.1740247230695138,
+      "grad_norm": 1.2621833086013794,
+      "learning_rate": 5.710648665968543e-05,
+      "loss": 0.4228,
+      "step": 6771
+    },
+    {
+      "epoch": 2.174345801894365,
+      "grad_norm": 0.7521808743476868,
+      "learning_rate": 5.707495692332816e-05,
+      "loss": 0.2837,
+      "step": 6772
+    },
+    {
+      "epoch": 2.1746668807192164,
+      "grad_norm": 1.019662857055664,
+      "learning_rate": 5.704343241725719e-05,
+      "loss": 0.3995,
+      "step": 6773
+    },
+    {
+      "epoch": 2.174987959544068,
+      "grad_norm": 0.661185085773468,
+      "learning_rate": 5.701191314531364e-05,
+      "loss": 0.2683,
+      "step": 6774
+    },
+    {
+      "epoch": 2.1753090383689195,
+      "grad_norm": 0.9004967212677002,
+      "learning_rate": 5.6980399111338156e-05,
+      "loss": 0.3288,
+      "step": 6775
+    },
+    {
+      "epoch": 2.175630117193771,
+      "grad_norm": 0.5190272927284241,
+      "learning_rate": 5.694889031917047e-05,
+      "loss": 0.2803,
+      "step": 6776
+    },
+    {
+      "epoch": 2.1759511960186226,
+      "grad_norm": 0.6035006642341614,
+      "learning_rate": 5.691738677265e-05,
+      "loss": 0.2716,
+      "step": 6777
+    },
+    {
+      "epoch": 2.176272274843474,
+      "grad_norm": 0.888232409954071,
+      "learning_rate": 5.6885888475615204e-05,
+      "loss": 0.3035,
+      "step": 6778
+    },
+    {
+      "epoch": 2.1765933536683257,
+      "grad_norm": 0.47930851578712463,
+      "learning_rate": 5.6854395431904094e-05,
+      "loss": 0.255,
+      "step": 6779
+    },
+    {
+      "epoch": 2.1769144324931773,
+      "grad_norm": 0.3152773678302765,
+      "learning_rate": 5.68229076453541e-05,
+      "loss": 0.3738,
+      "step": 6780
+    },
+    {
+      "epoch": 2.1772355113180284,
+      "grad_norm": 0.5532098412513733,
+      "learning_rate": 5.679142511980175e-05,
+      "loss": 0.7954,
+      "step": 6781
+    },
+    {
+      "epoch": 2.17755659014288,
+      "grad_norm": 0.48920655250549316,
+      "learning_rate": 5.675994785908329e-05,
+      "loss": 0.4281,
+      "step": 6782
+    },
+    {
+      "epoch": 2.1778776689677315,
+      "grad_norm": 0.5712006688117981,
+      "learning_rate": 5.6728475867033925e-05,
+      "loss": 0.255,
+      "step": 6783
+    },
+    {
+      "epoch": 2.178198747792583,
+      "grad_norm": 0.6813241839408875,
+      "learning_rate": 5.669700914748857e-05,
+      "loss": 0.1684,
+      "step": 6784
+    },
+    {
+      "epoch": 2.1785198266174346,
+      "grad_norm": 0.3276176452636719,
+      "learning_rate": 5.666554770428129e-05,
+      "loss": 0.0962,
+      "step": 6785
+    },
+    {
+      "epoch": 2.178840905442286,
+      "grad_norm": 0.8344230651855469,
+      "learning_rate": 5.663409154124557e-05,
+      "loss": 0.3953,
+      "step": 6786
+    },
+    {
+      "epoch": 2.1791619842671377,
+      "grad_norm": 0.9453107714653015,
+      "learning_rate": 5.6602640662214256e-05,
+      "loss": 0.4333,
+      "step": 6787
+    },
+    {
+      "epoch": 2.1794830630919892,
+      "grad_norm": 0.737591564655304,
+      "learning_rate": 5.657119507101954e-05,
+      "loss": 0.3033,
+      "step": 6788
+    },
+    {
+      "epoch": 2.179804141916841,
+      "grad_norm": 0.7458065748214722,
+      "learning_rate": 5.653975477149298e-05,
+      "loss": 0.2871,
+      "step": 6789
+    },
+    {
+      "epoch": 2.180125220741692,
+      "grad_norm": 0.9435691833496094,
+      "learning_rate": 5.6508319767465465e-05,
+      "loss": 0.4635,
+      "step": 6790
+    },
+    {
+      "epoch": 2.1804462995665435,
+      "grad_norm": 0.611143946647644,
+      "learning_rate": 5.647689006276726e-05,
+      "loss": 0.2361,
+      "step": 6791
+    },
+    {
+      "epoch": 2.180767378391395,
+      "grad_norm": 0.6122483611106873,
+      "learning_rate": 5.6445465661227994e-05,
+      "loss": 0.3083,
+      "step": 6792
+    },
+    {
+      "epoch": 2.1810884572162466,
+      "grad_norm": 0.9717581272125244,
+      "learning_rate": 5.641404656667661e-05,
+      "loss": 0.4157,
+      "step": 6793
+    },
+    {
+      "epoch": 2.181409536041098,
+      "grad_norm": 0.8596290349960327,
+      "learning_rate": 5.6382632782941405e-05,
+      "loss": 0.3547,
+      "step": 6794
+    },
+    {
+      "epoch": 2.1817306148659497,
+      "grad_norm": 0.9642032384872437,
+      "learning_rate": 5.635122431385016e-05,
+      "loss": 0.3883,
+      "step": 6795
+    },
+    {
+      "epoch": 2.182051693690801,
+      "grad_norm": 0.8566706776618958,
+      "learning_rate": 5.63198211632298e-05,
+      "loss": 0.395,
+      "step": 6796
+    },
+    {
+      "epoch": 2.1823727725156528,
+      "grad_norm": 0.8098425269126892,
+      "learning_rate": 5.6288423334906735e-05,
+      "loss": 0.3301,
+      "step": 6797
+    },
+    {
+      "epoch": 2.1826938513405043,
+      "grad_norm": 0.8384745121002197,
+      "learning_rate": 5.6257030832706695e-05,
+      "loss": 0.3198,
+      "step": 6798
+    },
+    {
+      "epoch": 2.1830149301653554,
+      "grad_norm": 0.8153161406517029,
+      "learning_rate": 5.622564366045472e-05,
+      "loss": 0.3101,
+      "step": 6799
+    },
+    {
+      "epoch": 2.183336008990207,
+      "grad_norm": 0.8854767084121704,
+      "learning_rate": 5.619426182197536e-05,
+      "loss": 0.3009,
+      "step": 6800
+    },
+    {
+      "epoch": 2.1836570878150585,
+      "grad_norm": 0.9710304737091064,
+      "learning_rate": 5.616288532109225e-05,
+      "loss": 0.3954,
+      "step": 6801
+    },
+    {
+      "epoch": 2.18397816663991,
+      "grad_norm": 0.9379624128341675,
+      "learning_rate": 5.6131514161628626e-05,
+      "loss": 0.3856,
+      "step": 6802
+    },
+    {
+      "epoch": 2.1842992454647616,
+      "grad_norm": 1.2100353240966797,
+      "learning_rate": 5.610014834740693e-05,
+      "loss": 0.5041,
+      "step": 6803
+    },
+    {
+      "epoch": 2.184620324289613,
+      "grad_norm": 0.8026917576789856,
+      "learning_rate": 5.6068787882249005e-05,
+      "loss": 0.28,
+      "step": 6804
+    },
+    {
+      "epoch": 2.1849414031144647,
+      "grad_norm": 0.776182234287262,
+      "learning_rate": 5.6037432769976064e-05,
+      "loss": 0.2885,
+      "step": 6805
+    },
+    {
+      "epoch": 2.1852624819393163,
+      "grad_norm": 0.9188643097877502,
+      "learning_rate": 5.6006083014408484e-05,
+      "loss": 0.4413,
+      "step": 6806
+    },
+    {
+      "epoch": 2.185583560764168,
+      "grad_norm": 0.9212349057197571,
+      "learning_rate": 5.5974738619366295e-05,
+      "loss": 0.4209,
+      "step": 6807
+    },
+    {
+      "epoch": 2.185904639589019,
+      "grad_norm": 1.0901522636413574,
+      "learning_rate": 5.594339958866867e-05,
+      "loss": 0.3755,
+      "step": 6808
+    },
+    {
+      "epoch": 2.1862257184138705,
+      "grad_norm": 1.0959854125976562,
+      "learning_rate": 5.591206592613416e-05,
+      "loss": 0.3334,
+      "step": 6809
+    },
+    {
+      "epoch": 2.186546797238722,
+      "grad_norm": 0.8223209381103516,
+      "learning_rate": 5.588073763558068e-05,
+      "loss": 0.3331,
+      "step": 6810
+    },
+    {
+      "epoch": 2.1868678760635736,
+      "grad_norm": 0.9825518131256104,
+      "learning_rate": 5.584941472082549e-05,
+      "loss": 0.3669,
+      "step": 6811
+    },
+    {
+      "epoch": 2.187188954888425,
+      "grad_norm": 0.7833166122436523,
+      "learning_rate": 5.5818097185685206e-05,
+      "loss": 0.3309,
+      "step": 6812
+    },
+    {
+      "epoch": 2.1875100337132767,
+      "grad_norm": 0.5879184603691101,
+      "learning_rate": 5.578678503397574e-05,
+      "loss": 0.2564,
+      "step": 6813
+    },
+    {
+      "epoch": 2.1878311125381282,
+      "grad_norm": 0.9265559315681458,
+      "learning_rate": 5.575547826951242e-05,
+      "loss": 0.3454,
+      "step": 6814
+    },
+    {
+      "epoch": 2.18815219136298,
+      "grad_norm": 0.9046943187713623,
+      "learning_rate": 5.572417689610987e-05,
+      "loss": 0.3133,
+      "step": 6815
+    },
+    {
+      "epoch": 2.1884732701878313,
+      "grad_norm": 1.2076610326766968,
+      "learning_rate": 5.5692880917582046e-05,
+      "loss": 0.4036,
+      "step": 6816
+    },
+    {
+      "epoch": 2.1887943490126824,
+      "grad_norm": 0.8415603637695312,
+      "learning_rate": 5.566159033774225e-05,
+      "loss": 0.3282,
+      "step": 6817
+    },
+    {
+      "epoch": 2.189115427837534,
+      "grad_norm": 0.725054144859314,
+      "learning_rate": 5.5630305160403275e-05,
+      "loss": 0.2602,
+      "step": 6818
+    },
+    {
+      "epoch": 2.1894365066623855,
+      "grad_norm": 0.7742078900337219,
+      "learning_rate": 5.5599025389376935e-05,
+      "loss": 0.3131,
+      "step": 6819
+    },
+    {
+      "epoch": 2.189757585487237,
+      "grad_norm": 0.7798224091529846,
+      "learning_rate": 5.556775102847475e-05,
+      "loss": 0.3724,
+      "step": 6820
+    },
+    {
+      "epoch": 2.1900786643120886,
+      "grad_norm": 0.964368462562561,
+      "learning_rate": 5.553648208150728e-05,
+      "loss": 0.3715,
+      "step": 6821
+    },
+    {
+      "epoch": 2.19039974313694,
+      "grad_norm": 1.1923866271972656,
+      "learning_rate": 5.5505218552284565e-05,
+      "loss": 0.3996,
+      "step": 6822
+    },
+    {
+      "epoch": 2.1907208219617917,
+      "grad_norm": 0.852208137512207,
+      "learning_rate": 5.547396044461608e-05,
+      "loss": 0.3929,
+      "step": 6823
+    },
+    {
+      "epoch": 2.1910419007866433,
+      "grad_norm": 0.5591673254966736,
+      "learning_rate": 5.544270776231038e-05,
+      "loss": 0.2536,
+      "step": 6824
+    },
+    {
+      "epoch": 2.191362979611495,
+      "grad_norm": 0.6292114853858948,
+      "learning_rate": 5.541146050917561e-05,
+      "loss": 0.3284,
+      "step": 6825
+    },
+    {
+      "epoch": 2.191684058436346,
+      "grad_norm": 0.8230541944503784,
+      "learning_rate": 5.5380218689019125e-05,
+      "loss": 0.3711,
+      "step": 6826
+    },
+    {
+      "epoch": 2.1920051372611975,
+      "grad_norm": 0.500881016254425,
+      "learning_rate": 5.5348982305647643e-05,
+      "loss": 0.2666,
+      "step": 6827
+    },
+    {
+      "epoch": 2.192326216086049,
+      "grad_norm": 0.8684194087982178,
+      "learning_rate": 5.5317751362867234e-05,
+      "loss": 0.3415,
+      "step": 6828
+    },
+    {
+      "epoch": 2.1926472949109006,
+      "grad_norm": 1.012944221496582,
+      "learning_rate": 5.5286525864483285e-05,
+      "loss": 0.3696,
+      "step": 6829
+    },
+    {
+      "epoch": 2.192968373735752,
+      "grad_norm": 0.5349951386451721,
+      "learning_rate": 5.525530581430054e-05,
+      "loss": 0.6524,
+      "step": 6830
+    },
+    {
+      "epoch": 2.1932894525606037,
+      "grad_norm": 0.5182074308395386,
+      "learning_rate": 5.522409121612304e-05,
+      "loss": 0.5625,
+      "step": 6831
+    },
+    {
+      "epoch": 2.1936105313854553,
+      "grad_norm": 0.450907438993454,
+      "learning_rate": 5.519288207375422e-05,
+      "loss": 0.2453,
+      "step": 6832
+    },
+    {
+      "epoch": 2.193931610210307,
+      "grad_norm": 0.4998016655445099,
+      "learning_rate": 5.5161678390996796e-05,
+      "loss": 0.1582,
+      "step": 6833
+    },
+    {
+      "epoch": 2.1942526890351584,
+      "grad_norm": 0.7120688557624817,
+      "learning_rate": 5.513048017165284e-05,
+      "loss": 0.2156,
+      "step": 6834
+    },
+    {
+      "epoch": 2.1945737678600095,
+      "grad_norm": 0.4577905237674713,
+      "learning_rate": 5.509928741952379e-05,
+      "loss": 0.203,
+      "step": 6835
+    },
+    {
+      "epoch": 2.194894846684861,
+      "grad_norm": 0.744748055934906,
+      "learning_rate": 5.506810013841036e-05,
+      "loss": 0.377,
+      "step": 6836
+    },
+    {
+      "epoch": 2.1952159255097126,
+      "grad_norm": 1.0785094499588013,
+      "learning_rate": 5.50369183321126e-05,
+      "loss": 0.4875,
+      "step": 6837
+    },
+    {
+      "epoch": 2.195537004334564,
+      "grad_norm": 0.8972413539886475,
+      "learning_rate": 5.5005742004430025e-05,
+      "loss": 0.398,
+      "step": 6838
+    },
+    {
+      "epoch": 2.1958580831594157,
+      "grad_norm": 0.7619357705116272,
+      "learning_rate": 5.497457115916127e-05,
+      "loss": 0.3321,
+      "step": 6839
+    },
+    {
+      "epoch": 2.196179161984267,
+      "grad_norm": 0.711033046245575,
+      "learning_rate": 5.494340580010441e-05,
+      "loss": 0.3118,
+      "step": 6840
+    },
+    {
+      "epoch": 2.1965002408091188,
+      "grad_norm": 0.715796172618866,
+      "learning_rate": 5.491224593105695e-05,
+      "loss": 0.3553,
+      "step": 6841
+    },
+    {
+      "epoch": 2.1968213196339703,
+      "grad_norm": 0.6636841297149658,
+      "learning_rate": 5.488109155581549e-05,
+      "loss": 0.2737,
+      "step": 6842
+    },
+    {
+      "epoch": 2.197142398458822,
+      "grad_norm": 0.918535590171814,
+      "learning_rate": 5.484994267817624e-05,
+      "loss": 0.284,
+      "step": 6843
+    },
+    {
+      "epoch": 2.197463477283673,
+      "grad_norm": 0.7877380847930908,
+      "learning_rate": 5.481879930193443e-05,
+      "loss": 0.3617,
+      "step": 6844
+    },
+    {
+      "epoch": 2.1977845561085245,
+      "grad_norm": 0.9507259130477905,
+      "learning_rate": 5.478766143088492e-05,
+      "loss": 0.3557,
+      "step": 6845
+    },
+    {
+      "epoch": 2.198105634933376,
+      "grad_norm": 0.8569693565368652,
+      "learning_rate": 5.475652906882173e-05,
+      "loss": 0.3243,
+      "step": 6846
+    },
+    {
+      "epoch": 2.1984267137582276,
+      "grad_norm": 0.8883730173110962,
+      "learning_rate": 5.4725402219538236e-05,
+      "loss": 0.3438,
+      "step": 6847
+    },
+    {
+      "epoch": 2.198747792583079,
+      "grad_norm": 0.8322269320487976,
+      "learning_rate": 5.469428088682717e-05,
+      "loss": 0.3581,
+      "step": 6848
+    },
+    {
+      "epoch": 2.1990688714079307,
+      "grad_norm": 0.8799639344215393,
+      "learning_rate": 5.466316507448048e-05,
+      "loss": 0.3936,
+      "step": 6849
+    },
+    {
+      "epoch": 2.1993899502327823,
+      "grad_norm": 1.0511608123779297,
+      "learning_rate": 5.4632054786289656e-05,
+      "loss": 0.4143,
+      "step": 6850
+    },
+    {
+      "epoch": 2.199711029057634,
+      "grad_norm": 0.8037779927253723,
+      "learning_rate": 5.4600950026045326e-05,
+      "loss": 0.3097,
+      "step": 6851
+    },
+    {
+      "epoch": 2.2000321078824854,
+      "grad_norm": 0.8404053449630737,
+      "learning_rate": 5.4569850797537536e-05,
+      "loss": 0.3181,
+      "step": 6852
+    },
+    {
+      "epoch": 2.2003531867073365,
+      "grad_norm": 1.0227787494659424,
+      "learning_rate": 5.4538757104555615e-05,
+      "loss": 0.3921,
+      "step": 6853
+    },
+    {
+      "epoch": 2.200674265532188,
+      "grad_norm": 0.5781304240226746,
+      "learning_rate": 5.4507668950888245e-05,
+      "loss": 0.2198,
+      "step": 6854
+    },
+    {
+      "epoch": 2.2009953443570396,
+      "grad_norm": 0.811543881893158,
+      "learning_rate": 5.447658634032338e-05,
+      "loss": 0.3287,
+      "step": 6855
+    },
+    {
+      "epoch": 2.201316423181891,
+      "grad_norm": 0.7216514945030212,
+      "learning_rate": 5.4445509276648466e-05,
+      "loss": 0.3109,
+      "step": 6856
+    },
+    {
+      "epoch": 2.2016375020067427,
+      "grad_norm": 0.7720044851303101,
+      "learning_rate": 5.441443776365003e-05,
+      "loss": 0.3122,
+      "step": 6857
+    },
+    {
+      "epoch": 2.2019585808315942,
+      "grad_norm": 0.8112689852714539,
+      "learning_rate": 5.438337180511406e-05,
+      "loss": 0.3681,
+      "step": 6858
+    },
+    {
+      "epoch": 2.202279659656446,
+      "grad_norm": 0.9748257398605347,
+      "learning_rate": 5.435231140482587e-05,
+      "loss": 0.4623,
+      "step": 6859
+    },
+    {
+      "epoch": 2.2026007384812973,
+      "grad_norm": 0.9538961052894592,
+      "learning_rate": 5.4321256566570036e-05,
+      "loss": 0.4044,
+      "step": 6860
+    },
+    {
+      "epoch": 2.202921817306149,
+      "grad_norm": 0.7297887206077576,
+      "learning_rate": 5.4290207294130615e-05,
+      "loss": 0.2882,
+      "step": 6861
+    },
+    {
+      "epoch": 2.203242896131,
+      "grad_norm": 0.8942319750785828,
+      "learning_rate": 5.42591635912907e-05,
+      "loss": 0.3243,
+      "step": 6862
+    },
+    {
+      "epoch": 2.2035639749558515,
+      "grad_norm": 0.6959256529808044,
+      "learning_rate": 5.4228125461833024e-05,
+      "loss": 0.2407,
+      "step": 6863
+    },
+    {
+      "epoch": 2.203885053780703,
+      "grad_norm": 0.8571872115135193,
+      "learning_rate": 5.419709290953936e-05,
+      "loss": 0.3134,
+      "step": 6864
+    },
+    {
+      "epoch": 2.2042061326055546,
+      "grad_norm": 0.8868092894554138,
+      "learning_rate": 5.416606593819101e-05,
+      "loss": 0.3696,
+      "step": 6865
+    },
+    {
+      "epoch": 2.204527211430406,
+      "grad_norm": 1.1420269012451172,
+      "learning_rate": 5.4135044551568546e-05,
+      "loss": 0.338,
+      "step": 6866
+    },
+    {
+      "epoch": 2.2048482902552577,
+      "grad_norm": 0.7926510572433472,
+      "learning_rate": 5.4104028753451696e-05,
+      "loss": 0.3127,
+      "step": 6867
+    },
+    {
+      "epoch": 2.2051693690801093,
+      "grad_norm": 0.7856837511062622,
+      "learning_rate": 5.407301854761977e-05,
+      "loss": 0.2912,
+      "step": 6868
+    },
+    {
+      "epoch": 2.205490447904961,
+      "grad_norm": 0.9291871786117554,
+      "learning_rate": 5.404201393785122e-05,
+      "loss": 0.3568,
+      "step": 6869
+    },
+    {
+      "epoch": 2.2058115267298124,
+      "grad_norm": 0.7121933102607727,
+      "learning_rate": 5.401101492792386e-05,
+      "loss": 0.3451,
+      "step": 6870
+    },
+    {
+      "epoch": 2.2061326055546635,
+      "grad_norm": 0.6835899949073792,
+      "learning_rate": 5.398002152161484e-05,
+      "loss": 0.284,
+      "step": 6871
+    },
+    {
+      "epoch": 2.206453684379515,
+      "grad_norm": 0.9216342568397522,
+      "learning_rate": 5.394903372270062e-05,
+      "loss": 0.357,
+      "step": 6872
+    },
+    {
+      "epoch": 2.2067747632043666,
+      "grad_norm": 0.6000509262084961,
+      "learning_rate": 5.3918051534956926e-05,
+      "loss": 0.2943,
+      "step": 6873
+    },
+    {
+      "epoch": 2.207095842029218,
+      "grad_norm": 1.0322999954223633,
+      "learning_rate": 5.388707496215888e-05,
+      "loss": 0.3784,
+      "step": 6874
+    },
+    {
+      "epoch": 2.2074169208540697,
+      "grad_norm": 0.9377049803733826,
+      "learning_rate": 5.3856104008080876e-05,
+      "loss": 0.3036,
+      "step": 6875
+    },
+    {
+      "epoch": 2.2077379996789213,
+      "grad_norm": 0.6841163039207458,
+      "learning_rate": 5.382513867649663e-05,
+      "loss": 0.2938,
+      "step": 6876
+    },
+    {
+      "epoch": 2.208059078503773,
+      "grad_norm": 0.9732836484909058,
+      "learning_rate": 5.3794178971179165e-05,
+      "loss": 0.377,
+      "step": 6877
+    },
+    {
+      "epoch": 2.2083801573286244,
+      "grad_norm": 0.5027754306793213,
+      "learning_rate": 5.3763224895900846e-05,
+      "loss": 0.2797,
+      "step": 6878
+    },
+    {
+      "epoch": 2.208701236153476,
+      "grad_norm": 0.867052435874939,
+      "learning_rate": 5.373227645443332e-05,
+      "loss": 0.3347,
+      "step": 6879
+    },
+    {
+      "epoch": 2.209022314978327,
+      "grad_norm": 0.4138392210006714,
+      "learning_rate": 5.3701333650547525e-05,
+      "loss": 0.4415,
+      "step": 6880
+    },
+    {
+      "epoch": 2.2093433938031786,
+      "grad_norm": 0.4743240475654602,
+      "learning_rate": 5.3670396488013854e-05,
+      "loss": 0.4268,
+      "step": 6881
+    },
+    {
+      "epoch": 2.20966447262803,
+      "grad_norm": 0.6370947957038879,
+      "learning_rate": 5.3639464970601775e-05,
+      "loss": 0.5011,
+      "step": 6882
+    },
+    {
+      "epoch": 2.2099855514528817,
+      "grad_norm": 0.6120859980583191,
+      "learning_rate": 5.360853910208028e-05,
+      "loss": 0.4556,
+      "step": 6883
+    },
+    {
+      "epoch": 2.210306630277733,
+      "grad_norm": 0.5497114062309265,
+      "learning_rate": 5.357761888621764e-05,
+      "loss": 0.1802,
+      "step": 6884
+    },
+    {
+      "epoch": 2.2106277091025848,
+      "grad_norm": 0.7469709515571594,
+      "learning_rate": 5.3546704326781236e-05,
+      "loss": 0.1502,
+      "step": 6885
+    },
+    {
+      "epoch": 2.2109487879274363,
+      "grad_norm": 0.6369648575782776,
+      "learning_rate": 5.351579542753807e-05,
+      "loss": 0.2719,
+      "step": 6886
+    },
+    {
+      "epoch": 2.211269866752288,
+      "grad_norm": 0.8421298861503601,
+      "learning_rate": 5.348489219225416e-05,
+      "loss": 0.4426,
+      "step": 6887
+    },
+    {
+      "epoch": 2.2115909455771394,
+      "grad_norm": 0.9989409446716309,
+      "learning_rate": 5.345399462469509e-05,
+      "loss": 0.4848,
+      "step": 6888
+    },
+    {
+      "epoch": 2.2119120244019905,
+      "grad_norm": 0.9956933259963989,
+      "learning_rate": 5.3423102728625574e-05,
+      "loss": 0.4123,
+      "step": 6889
+    },
+    {
+      "epoch": 2.212233103226842,
+      "grad_norm": 0.6040937900543213,
+      "learning_rate": 5.3392216507809714e-05,
+      "loss": 0.2475,
+      "step": 6890
+    },
+    {
+      "epoch": 2.2125541820516936,
+      "grad_norm": 0.8480896353721619,
+      "learning_rate": 5.33613359660109e-05,
+      "loss": 0.3257,
+      "step": 6891
+    },
+    {
+      "epoch": 2.212875260876545,
+      "grad_norm": 0.824813961982727,
+      "learning_rate": 5.33304611069918e-05,
+      "loss": 0.3541,
+      "step": 6892
+    },
+    {
+      "epoch": 2.2131963397013967,
+      "grad_norm": 1.2338895797729492,
+      "learning_rate": 5.329959193451448e-05,
+      "loss": 0.347,
+      "step": 6893
+    },
+    {
+      "epoch": 2.2135174185262483,
+      "grad_norm": 0.9513993263244629,
+      "learning_rate": 5.326872845234021e-05,
+      "loss": 0.3895,
+      "step": 6894
+    },
+    {
+      "epoch": 2.2138384973511,
+      "grad_norm": 0.9307944774627686,
+      "learning_rate": 5.3237870664229636e-05,
+      "loss": 0.4291,
+      "step": 6895
+    },
+    {
+      "epoch": 2.2141595761759514,
+      "grad_norm": 0.8363532423973083,
+      "learning_rate": 5.320701857394268e-05,
+      "loss": 0.3269,
+      "step": 6896
+    },
+    {
+      "epoch": 2.2144806550008025,
+      "grad_norm": 0.873292863368988,
+      "learning_rate": 5.317617218523856e-05,
+      "loss": 0.4286,
+      "step": 6897
+    },
+    {
+      "epoch": 2.214801733825654,
+      "grad_norm": 0.7528856992721558,
+      "learning_rate": 5.3145331501875796e-05,
+      "loss": 0.3506,
+      "step": 6898
+    },
+    {
+      "epoch": 2.2151228126505056,
+      "grad_norm": 0.7959845662117004,
+      "learning_rate": 5.311449652761235e-05,
+      "loss": 0.3377,
+      "step": 6899
+    },
+    {
+      "epoch": 2.215443891475357,
+      "grad_norm": 0.954590916633606,
+      "learning_rate": 5.3083667266205194e-05,
+      "loss": 0.3659,
+      "step": 6900
+    },
+    {
+      "epoch": 2.2157649703002087,
+      "grad_norm": 0.9657596945762634,
+      "learning_rate": 5.305284372141095e-05,
+      "loss": 0.437,
+      "step": 6901
+    },
+    {
+      "epoch": 2.2160860491250602,
+      "grad_norm": 0.9250769019126892,
+      "learning_rate": 5.302202589698525e-05,
+      "loss": 0.3796,
+      "step": 6902
+    },
+    {
+      "epoch": 2.216407127949912,
+      "grad_norm": 0.8130130171775818,
+      "learning_rate": 5.299121379668316e-05,
+      "loss": 0.286,
+      "step": 6903
+    },
+    {
+      "epoch": 2.2167282067747633,
+      "grad_norm": 0.8332054018974304,
+      "learning_rate": 5.296040742425916e-05,
+      "loss": 0.3795,
+      "step": 6904
+    },
+    {
+      "epoch": 2.217049285599615,
+      "grad_norm": 0.786918580532074,
+      "learning_rate": 5.292960678346675e-05,
+      "loss": 0.3355,
+      "step": 6905
+    },
+    {
+      "epoch": 2.217370364424466,
+      "grad_norm": 1.060431718826294,
+      "learning_rate": 5.2898811878059e-05,
+      "loss": 0.4117,
+      "step": 6906
+    },
+    {
+      "epoch": 2.2176914432493176,
+      "grad_norm": 0.7190358638763428,
+      "learning_rate": 5.286802271178815e-05,
+      "loss": 0.2718,
+      "step": 6907
+    },
+    {
+      "epoch": 2.218012522074169,
+      "grad_norm": 1.1578835248947144,
+      "learning_rate": 5.2837239288405784e-05,
+      "loss": 0.4461,
+      "step": 6908
+    },
+    {
+      "epoch": 2.2183336008990207,
+      "grad_norm": 0.8902946710586548,
+      "learning_rate": 5.2806461611662735e-05,
+      "loss": 0.3296,
+      "step": 6909
+    },
+    {
+      "epoch": 2.218654679723872,
+      "grad_norm": 0.8117751479148865,
+      "learning_rate": 5.277568968530919e-05,
+      "loss": 0.3455,
+      "step": 6910
+    },
+    {
+      "epoch": 2.2189757585487238,
+      "grad_norm": 0.6965182423591614,
+      "learning_rate": 5.274492351309461e-05,
+      "loss": 0.2972,
+      "step": 6911
+    },
+    {
+      "epoch": 2.2192968373735753,
+      "grad_norm": 0.7272025942802429,
+      "learning_rate": 5.271416309876776e-05,
+      "loss": 0.2932,
+      "step": 6912
+    },
+    {
+      "epoch": 2.219617916198427,
+      "grad_norm": 0.7494564056396484,
+      "learning_rate": 5.26834084460767e-05,
+      "loss": 0.2976,
+      "step": 6913
+    },
+    {
+      "epoch": 2.2199389950232784,
+      "grad_norm": 0.7828695774078369,
+      "learning_rate": 5.265265955876879e-05,
+      "loss": 0.2948,
+      "step": 6914
+    },
+    {
+      "epoch": 2.2202600738481295,
+      "grad_norm": 0.8334441781044006,
+      "learning_rate": 5.2621916440590715e-05,
+      "loss": 0.3339,
+      "step": 6915
+    },
+    {
+      "epoch": 2.220581152672981,
+      "grad_norm": 0.6018253564834595,
+      "learning_rate": 5.259117909528839e-05,
+      "loss": 0.265,
+      "step": 6916
+    },
+    {
+      "epoch": 2.2209022314978326,
+      "grad_norm": 1.2306244373321533,
+      "learning_rate": 5.256044752660709e-05,
+      "loss": 0.2778,
+      "step": 6917
+    },
+    {
+      "epoch": 2.221223310322684,
+      "grad_norm": 1.4875001907348633,
+      "learning_rate": 5.2529721738291315e-05,
+      "loss": 0.3668,
+      "step": 6918
+    },
+    {
+      "epoch": 2.2215443891475357,
+      "grad_norm": 1.1924291849136353,
+      "learning_rate": 5.2499001734085044e-05,
+      "loss": 0.4159,
+      "step": 6919
+    },
+    {
+      "epoch": 2.2218654679723873,
+      "grad_norm": 0.8212143182754517,
+      "learning_rate": 5.2468287517731276e-05,
+      "loss": 0.3293,
+      "step": 6920
+    },
+    {
+      "epoch": 2.222186546797239,
+      "grad_norm": 0.7851700782775879,
+      "learning_rate": 5.243757909297247e-05,
+      "loss": 0.3297,
+      "step": 6921
+    },
+    {
+      "epoch": 2.2225076256220904,
+      "grad_norm": 0.7045104503631592,
+      "learning_rate": 5.2406876463550445e-05,
+      "loss": 0.2812,
+      "step": 6922
+    },
+    {
+      "epoch": 2.222828704446942,
+      "grad_norm": 1.1284433603286743,
+      "learning_rate": 5.237617963320608e-05,
+      "loss": 0.3043,
+      "step": 6923
+    },
+    {
+      "epoch": 2.223149783271793,
+      "grad_norm": 0.7409510016441345,
+      "learning_rate": 5.234548860567985e-05,
+      "loss": 0.3132,
+      "step": 6924
+    },
+    {
+      "epoch": 2.2234708620966446,
+      "grad_norm": 1.1001427173614502,
+      "learning_rate": 5.2314803384711195e-05,
+      "loss": 0.3493,
+      "step": 6925
+    },
+    {
+      "epoch": 2.223791940921496,
+      "grad_norm": 0.7392867803573608,
+      "learning_rate": 5.2284123974039154e-05,
+      "loss": 0.3222,
+      "step": 6926
+    },
+    {
+      "epoch": 2.2241130197463477,
+      "grad_norm": 0.556638777256012,
+      "learning_rate": 5.225345037740186e-05,
+      "loss": 0.2963,
+      "step": 6927
+    },
+    {
+      "epoch": 2.2244340985711992,
+      "grad_norm": 0.4246133863925934,
+      "learning_rate": 5.222278259853681e-05,
+      "loss": 0.2487,
+      "step": 6928
+    },
+    {
+      "epoch": 2.2247551773960508,
+      "grad_norm": 1.3218295574188232,
+      "learning_rate": 5.2192120641180786e-05,
+      "loss": 0.2931,
+      "step": 6929
+    },
+    {
+      "epoch": 2.2250762562209023,
+      "grad_norm": 0.4780152440071106,
+      "learning_rate": 5.216146450906984e-05,
+      "loss": 0.4964,
+      "step": 6930
+    },
+    {
+      "epoch": 2.225397335045754,
+      "grad_norm": 0.41451188921928406,
+      "learning_rate": 5.213081420593933e-05,
+      "loss": 0.4104,
+      "step": 6931
+    },
+    {
+      "epoch": 2.2257184138706054,
+      "grad_norm": 0.5102071762084961,
+      "learning_rate": 5.210016973552391e-05,
+      "loss": 0.2379,
+      "step": 6932
+    },
+    {
+      "epoch": 2.2260394926954565,
+      "grad_norm": 0.32184290885925293,
+      "learning_rate": 5.20695311015575e-05,
+      "loss": 0.0819,
+      "step": 6933
+    },
+    {
+      "epoch": 2.226360571520308,
+      "grad_norm": 0.5118618011474609,
+      "learning_rate": 5.2038898307773354e-05,
+      "loss": 0.1675,
+      "step": 6934
+    },
+    {
+      "epoch": 2.2266816503451596,
+      "grad_norm": 0.561114490032196,
+      "learning_rate": 5.200827135790396e-05,
+      "loss": 0.2358,
+      "step": 6935
+    },
+    {
+      "epoch": 2.227002729170011,
+      "grad_norm": 0.8075849413871765,
+      "learning_rate": 5.197765025568109e-05,
+      "loss": 0.4558,
+      "step": 6936
+    },
+    {
+      "epoch": 2.2273238079948627,
+      "grad_norm": 0.7261122465133667,
+      "learning_rate": 5.194703500483593e-05,
+      "loss": 0.3476,
+      "step": 6937
+    },
+    {
+      "epoch": 2.2276448868197143,
+      "grad_norm": 0.8632923364639282,
+      "learning_rate": 5.1916425609098775e-05,
+      "loss": 0.4246,
+      "step": 6938
+    },
+    {
+      "epoch": 2.227965965644566,
+      "grad_norm": 0.8326036334037781,
+      "learning_rate": 5.188582207219931e-05,
+      "loss": 0.3473,
+      "step": 6939
+    },
+    {
+      "epoch": 2.2282870444694174,
+      "grad_norm": 0.9028269648551941,
+      "learning_rate": 5.1855224397866476e-05,
+      "loss": 0.3969,
+      "step": 6940
+    },
+    {
+      "epoch": 2.228608123294269,
+      "grad_norm": 0.8235176205635071,
+      "learning_rate": 5.182463258982846e-05,
+      "loss": 0.2969,
+      "step": 6941
+    },
+    {
+      "epoch": 2.22892920211912,
+      "grad_norm": 0.8503168821334839,
+      "learning_rate": 5.179404665181291e-05,
+      "loss": 0.3176,
+      "step": 6942
+    },
+    {
+      "epoch": 2.2292502809439716,
+      "grad_norm": 0.8072776794433594,
+      "learning_rate": 5.1763466587546485e-05,
+      "loss": 0.3188,
+      "step": 6943
+    },
+    {
+      "epoch": 2.229571359768823,
+      "grad_norm": 0.9978686571121216,
+      "learning_rate": 5.1732892400755376e-05,
+      "loss": 0.3749,
+      "step": 6944
+    },
+    {
+      "epoch": 2.2298924385936747,
+      "grad_norm": 0.9184989929199219,
+      "learning_rate": 5.170232409516496e-05,
+      "loss": 0.4544,
+      "step": 6945
+    },
+    {
+      "epoch": 2.2302135174185262,
+      "grad_norm": 0.7581484913825989,
+      "learning_rate": 5.1671761674499765e-05,
+      "loss": 0.2865,
+      "step": 6946
+    },
+    {
+      "epoch": 2.230534596243378,
+      "grad_norm": 0.6982542276382446,
+      "learning_rate": 5.1641205142483894e-05,
+      "loss": 0.285,
+      "step": 6947
+    },
+    {
+      "epoch": 2.2308556750682293,
+      "grad_norm": 1.0385242700576782,
+      "learning_rate": 5.16106545028404e-05,
+      "loss": 0.3419,
+      "step": 6948
+    },
+    {
+      "epoch": 2.231176753893081,
+      "grad_norm": 0.9583960175514221,
+      "learning_rate": 5.158010975929193e-05,
+      "loss": 0.3959,
+      "step": 6949
+    },
+    {
+      "epoch": 2.2314978327179325,
+      "grad_norm": 1.119672417640686,
+      "learning_rate": 5.1549570915560206e-05,
+      "loss": 0.4012,
+      "step": 6950
+    },
+    {
+      "epoch": 2.2318189115427836,
+      "grad_norm": 0.9875466823577881,
+      "learning_rate": 5.15190379753663e-05,
+      "loss": 0.4324,
+      "step": 6951
+    },
+    {
+      "epoch": 2.232139990367635,
+      "grad_norm": 0.9298665523529053,
+      "learning_rate": 5.148851094243057e-05,
+      "loss": 0.4168,
+      "step": 6952
+    },
+    {
+      "epoch": 2.2324610691924867,
+      "grad_norm": 0.7571067810058594,
+      "learning_rate": 5.145798982047261e-05,
+      "loss": 0.3063,
+      "step": 6953
+    },
+    {
+      "epoch": 2.232782148017338,
+      "grad_norm": 0.9960950613021851,
+      "learning_rate": 5.1427474613211356e-05,
+      "loss": 0.4125,
+      "step": 6954
+    },
+    {
+      "epoch": 2.2331032268421898,
+      "grad_norm": 0.9957188367843628,
+      "learning_rate": 5.1396965324364986e-05,
+      "loss": 0.3128,
+      "step": 6955
+    },
+    {
+      "epoch": 2.2334243056670413,
+      "grad_norm": 0.9226759672164917,
+      "learning_rate": 5.1366461957650954e-05,
+      "loss": 0.3247,
+      "step": 6956
+    },
+    {
+      "epoch": 2.233745384491893,
+      "grad_norm": 0.7779395580291748,
+      "learning_rate": 5.133596451678603e-05,
+      "loss": 0.297,
+      "step": 6957
+    },
+    {
+      "epoch": 2.2340664633167444,
+      "grad_norm": 1.165313482284546,
+      "learning_rate": 5.13054730054862e-05,
+      "loss": 0.4306,
+      "step": 6958
+    },
+    {
+      "epoch": 2.234387542141596,
+      "grad_norm": 0.7935650944709778,
+      "learning_rate": 5.127498742746675e-05,
+      "loss": 0.3009,
+      "step": 6959
+    },
+    {
+      "epoch": 2.234708620966447,
+      "grad_norm": 1.200869083404541,
+      "learning_rate": 5.1244507786442356e-05,
+      "loss": 0.4304,
+      "step": 6960
+    },
+    {
+      "epoch": 2.2350296997912986,
+      "grad_norm": 1.077374815940857,
+      "learning_rate": 5.121403408612672e-05,
+      "loss": 0.498,
+      "step": 6961
+    },
+    {
+      "epoch": 2.23535077861615,
+      "grad_norm": 0.8770790100097656,
+      "learning_rate": 5.1183566330233124e-05,
+      "loss": 0.3659,
+      "step": 6962
+    },
+    {
+      "epoch": 2.2356718574410017,
+      "grad_norm": 0.8858685493469238,
+      "learning_rate": 5.115310452247386e-05,
+      "loss": 0.3483,
+      "step": 6963
+    },
+    {
+      "epoch": 2.2359929362658533,
+      "grad_norm": 0.9457120895385742,
+      "learning_rate": 5.112264866656059e-05,
+      "loss": 0.4015,
+      "step": 6964
+    },
+    {
+      "epoch": 2.236314015090705,
+      "grad_norm": 0.7275859713554382,
+      "learning_rate": 5.1092198766204415e-05,
+      "loss": 0.3293,
+      "step": 6965
+    },
+    {
+      "epoch": 2.2366350939155564,
+      "grad_norm": 0.8628625869750977,
+      "learning_rate": 5.1061754825115374e-05,
+      "loss": 0.3513,
+      "step": 6966
+    },
+    {
+      "epoch": 2.236956172740408,
+      "grad_norm": 1.0785568952560425,
+      "learning_rate": 5.103131684700314e-05,
+      "loss": 0.3859,
+      "step": 6967
+    },
+    {
+      "epoch": 2.2372772515652595,
+      "grad_norm": 1.0810168981552124,
+      "learning_rate": 5.100088483557634e-05,
+      "loss": 0.3165,
+      "step": 6968
+    },
+    {
+      "epoch": 2.2375983303901106,
+      "grad_norm": 0.8149095177650452,
+      "learning_rate": 5.097045879454313e-05,
+      "loss": 0.3059,
+      "step": 6969
+    },
+    {
+      "epoch": 2.237919409214962,
+      "grad_norm": 0.921907901763916,
+      "learning_rate": 5.0940038727610796e-05,
+      "loss": 0.279,
+      "step": 6970
+    },
+    {
+      "epoch": 2.2382404880398137,
+      "grad_norm": 0.6431214213371277,
+      "learning_rate": 5.090962463848592e-05,
+      "loss": 0.3469,
+      "step": 6971
+    },
+    {
+      "epoch": 2.2385615668646652,
+      "grad_norm": 0.8860642313957214,
+      "learning_rate": 5.087921653087437e-05,
+      "loss": 0.4101,
+      "step": 6972
+    },
+    {
+      "epoch": 2.238882645689517,
+      "grad_norm": 0.930432140827179,
+      "learning_rate": 5.0848814408481305e-05,
+      "loss": 0.3629,
+      "step": 6973
+    },
+    {
+      "epoch": 2.2392037245143683,
+      "grad_norm": 1.0363874435424805,
+      "learning_rate": 5.0818418275011104e-05,
+      "loss": 0.2949,
+      "step": 6974
+    },
+    {
+      "epoch": 2.23952480333922,
+      "grad_norm": 1.0017728805541992,
+      "learning_rate": 5.0788028134167456e-05,
+      "loss": 0.37,
+      "step": 6975
+    },
+    {
+      "epoch": 2.2398458821640714,
+      "grad_norm": 0.7278321385383606,
+      "learning_rate": 5.07576439896533e-05,
+      "loss": 0.2832,
+      "step": 6976
+    },
+    {
+      "epoch": 2.240166960988923,
+      "grad_norm": 1.1001341342926025,
+      "learning_rate": 5.072726584517086e-05,
+      "loss": 0.3137,
+      "step": 6977
+    },
+    {
+      "epoch": 2.240488039813774,
+      "grad_norm": 0.9127258062362671,
+      "learning_rate": 5.069689370442161e-05,
+      "loss": 0.2938,
+      "step": 6978
+    },
+    {
+      "epoch": 2.2408091186386256,
+      "grad_norm": 0.6705423593521118,
+      "learning_rate": 5.066652757110628e-05,
+      "loss": 0.2919,
+      "step": 6979
+    },
+    {
+      "epoch": 2.241130197463477,
+      "grad_norm": 0.6100792288780212,
+      "learning_rate": 5.0636167448924987e-05,
+      "loss": 0.8073,
+      "step": 6980
+    },
+    {
+      "epoch": 2.2414512762883287,
+      "grad_norm": 0.6084210872650146,
+      "learning_rate": 5.0605813341576924e-05,
+      "loss": 0.8022,
+      "step": 6981
+    },
+    {
+      "epoch": 2.2417723551131803,
+      "grad_norm": 0.42793241143226624,
+      "learning_rate": 5.057546525276068e-05,
+      "loss": 0.2486,
+      "step": 6982
+    },
+    {
+      "epoch": 2.242093433938032,
+      "grad_norm": 0.3993290662765503,
+      "learning_rate": 5.054512318617406e-05,
+      "loss": 0.1494,
+      "step": 6983
+    },
+    {
+      "epoch": 2.2424145127628834,
+      "grad_norm": 0.5702289938926697,
+      "learning_rate": 5.051478714551414e-05,
+      "loss": 0.3078,
+      "step": 6984
+    },
+    {
+      "epoch": 2.242735591587735,
+      "grad_norm": 0.3860454559326172,
+      "learning_rate": 5.048445713447738e-05,
+      "loss": 0.0808,
+      "step": 6985
+    },
+    {
+      "epoch": 2.2430566704125865,
+      "grad_norm": 0.7141852974891663,
+      "learning_rate": 5.045413315675924e-05,
+      "loss": 0.3344,
+      "step": 6986
+    },
+    {
+      "epoch": 2.2433777492374376,
+      "grad_norm": 0.7017838954925537,
+      "learning_rate": 5.0423815216054724e-05,
+      "loss": 0.348,
+      "step": 6987
+    },
+    {
+      "epoch": 2.243698828062289,
+      "grad_norm": 1.1939187049865723,
+      "learning_rate": 5.039350331605794e-05,
+      "loss": 0.4238,
+      "step": 6988
+    },
+    {
+      "epoch": 2.2440199068871407,
+      "grad_norm": 0.8673609495162964,
+      "learning_rate": 5.036319746046232e-05,
+      "loss": 0.4174,
+      "step": 6989
+    },
+    {
+      "epoch": 2.2443409857119923,
+      "grad_norm": 0.7690039873123169,
+      "learning_rate": 5.033289765296054e-05,
+      "loss": 0.3594,
+      "step": 6990
+    },
+    {
+      "epoch": 2.244662064536844,
+      "grad_norm": 0.624330997467041,
+      "learning_rate": 5.0302603897244474e-05,
+      "loss": 0.2757,
+      "step": 6991
+    },
+    {
+      "epoch": 2.2449831433616954,
+      "grad_norm": 0.8041682839393616,
+      "learning_rate": 5.0272316197005396e-05,
+      "loss": 0.347,
+      "step": 6992
+    },
+    {
+      "epoch": 2.245304222186547,
+      "grad_norm": 0.6567179560661316,
+      "learning_rate": 5.024203455593375e-05,
+      "loss": 0.2858,
+      "step": 6993
+    },
+    {
+      "epoch": 2.2456253010113985,
+      "grad_norm": 0.9437269568443298,
+      "learning_rate": 5.021175897771927e-05,
+      "loss": 0.3507,
+      "step": 6994
+    },
+    {
+      "epoch": 2.24594637983625,
+      "grad_norm": 1.0030709505081177,
+      "learning_rate": 5.018148946605092e-05,
+      "loss": 0.4534,
+      "step": 6995
+    },
+    {
+      "epoch": 2.246267458661101,
+      "grad_norm": 0.8728824853897095,
+      "learning_rate": 5.015122602461698e-05,
+      "loss": 0.3801,
+      "step": 6996
+    },
+    {
+      "epoch": 2.2465885374859527,
+      "grad_norm": 0.9060454964637756,
+      "learning_rate": 5.012096865710494e-05,
+      "loss": 0.3869,
+      "step": 6997
+    },
+    {
+      "epoch": 2.246909616310804,
+      "grad_norm": 0.8747976422309875,
+      "learning_rate": 5.0090717367201554e-05,
+      "loss": 0.3823,
+      "step": 6998
+    },
+    {
+      "epoch": 2.2472306951356558,
+      "grad_norm": 0.9838657975196838,
+      "learning_rate": 5.006047215859289e-05,
+      "loss": 0.2901,
+      "step": 6999
+    },
+    {
+      "epoch": 2.2475517739605073,
+      "grad_norm": 0.7466594576835632,
+      "learning_rate": 5.003023303496419e-05,
+      "loss": 0.3505,
+      "step": 7000
+    },
+    {
+      "epoch": 2.247872852785359,
+      "grad_norm": 0.8229215145111084,
+      "learning_rate": 5.000000000000002e-05,
+      "loss": 0.3138,
+      "step": 7001
+    },
+    {
+      "epoch": 2.2481939316102104,
+      "grad_norm": 0.7786136865615845,
+      "learning_rate": 4.996977305738415e-05,
+      "loss": 0.3044,
+      "step": 7002
+    },
+    {
+      "epoch": 2.248515010435062,
+      "grad_norm": 0.6571848392486572,
+      "learning_rate": 4.9939552210799755e-05,
+      "loss": 0.2649,
+      "step": 7003
+    },
+    {
+      "epoch": 2.248836089259913,
+      "grad_norm": 0.6646254062652588,
+      "learning_rate": 4.990933746392899e-05,
+      "loss": 0.3347,
+      "step": 7004
+    },
+    {
+      "epoch": 2.2491571680847646,
+      "grad_norm": 0.8542558550834656,
+      "learning_rate": 4.98791288204536e-05,
+      "loss": 0.3518,
+      "step": 7005
+    },
+    {
+      "epoch": 2.249478246909616,
+      "grad_norm": 0.726729154586792,
+      "learning_rate": 4.9848926284054255e-05,
+      "loss": 0.2939,
+      "step": 7006
+    },
+    {
+      "epoch": 2.2497993257344677,
+      "grad_norm": 0.9049625396728516,
+      "learning_rate": 4.981872985841115e-05,
+      "loss": 0.3657,
+      "step": 7007
+    },
+    {
+      "epoch": 2.2501204045593193,
+      "grad_norm": 0.9269988536834717,
+      "learning_rate": 4.978853954720364e-05,
+      "loss": 0.2894,
+      "step": 7008
+    },
+    {
+      "epoch": 2.250441483384171,
+      "grad_norm": 1.2654590606689453,
+      "learning_rate": 4.97583553541102e-05,
+      "loss": 0.4061,
+      "step": 7009
+    },
+    {
+      "epoch": 2.2507625622090224,
+      "grad_norm": 0.8163503408432007,
+      "learning_rate": 4.97281772828088e-05,
+      "loss": 0.336,
+      "step": 7010
+    },
+    {
+      "epoch": 2.251083641033874,
+      "grad_norm": 1.1745022535324097,
+      "learning_rate": 4.969800533697649e-05,
+      "loss": 0.3993,
+      "step": 7011
+    },
+    {
+      "epoch": 2.2514047198587255,
+      "grad_norm": 0.7915329933166504,
+      "learning_rate": 4.966783952028967e-05,
+      "loss": 0.3401,
+      "step": 7012
+    },
+    {
+      "epoch": 2.2517257986835766,
+      "grad_norm": 0.9306576251983643,
+      "learning_rate": 4.9637679836423924e-05,
+      "loss": 0.3605,
+      "step": 7013
+    },
+    {
+      "epoch": 2.252046877508428,
+      "grad_norm": 0.5942294597625732,
+      "learning_rate": 4.960752628905412e-05,
+      "loss": 0.2566,
+      "step": 7014
+    },
+    {
+      "epoch": 2.2523679563332797,
+      "grad_norm": 0.743722140789032,
+      "learning_rate": 4.957737888185439e-05,
+      "loss": 0.2979,
+      "step": 7015
+    },
+    {
+      "epoch": 2.2526890351581312,
+      "grad_norm": 0.7660803198814392,
+      "learning_rate": 4.9547237618498085e-05,
+      "loss": 0.2865,
+      "step": 7016
+    },
+    {
+      "epoch": 2.253010113982983,
+      "grad_norm": 0.5819270014762878,
+      "learning_rate": 4.9517102502657845e-05,
+      "loss": 0.2663,
+      "step": 7017
+    },
+    {
+      "epoch": 2.2533311928078343,
+      "grad_norm": 0.7460042834281921,
+      "learning_rate": 4.9486973538005535e-05,
+      "loss": 0.2898,
+      "step": 7018
+    },
+    {
+      "epoch": 2.253652271632686,
+      "grad_norm": 0.8672288060188293,
+      "learning_rate": 4.945685072821227e-05,
+      "loss": 0.3627,
+      "step": 7019
+    },
+    {
+      "epoch": 2.2539733504575374,
+      "grad_norm": 1.0746335983276367,
+      "learning_rate": 4.9426734076948436e-05,
+      "loss": 0.3425,
+      "step": 7020
+    },
+    {
+      "epoch": 2.254294429282389,
+      "grad_norm": 0.6159694790840149,
+      "learning_rate": 4.939662358788364e-05,
+      "loss": 0.2663,
+      "step": 7021
+    },
+    {
+      "epoch": 2.25461550810724,
+      "grad_norm": 0.8164811730384827,
+      "learning_rate": 4.9366519264686725e-05,
+      "loss": 0.251,
+      "step": 7022
+    },
+    {
+      "epoch": 2.2549365869320916,
+      "grad_norm": 0.7953729033470154,
+      "learning_rate": 4.933642111102594e-05,
+      "loss": 0.3071,
+      "step": 7023
+    },
+    {
+      "epoch": 2.255257665756943,
+      "grad_norm": 0.7565558552742004,
+      "learning_rate": 4.9306329130568474e-05,
+      "loss": 0.3028,
+      "step": 7024
+    },
+    {
+      "epoch": 2.2555787445817947,
+      "grad_norm": 1.2023780345916748,
+      "learning_rate": 4.927624332698109e-05,
+      "loss": 0.4058,
+      "step": 7025
+    },
+    {
+      "epoch": 2.2558998234066463,
+      "grad_norm": 0.4921356439590454,
+      "learning_rate": 4.924616370392961e-05,
+      "loss": 0.2675,
+      "step": 7026
+    },
+    {
+      "epoch": 2.256220902231498,
+      "grad_norm": 0.8888304233551025,
+      "learning_rate": 4.921609026507907e-05,
+      "loss": 0.3284,
+      "step": 7027
+    },
+    {
+      "epoch": 2.2565419810563494,
+      "grad_norm": 0.6142435669898987,
+      "learning_rate": 4.918602301409395e-05,
+      "loss": 0.2956,
+      "step": 7028
+    },
+    {
+      "epoch": 2.256863059881201,
+      "grad_norm": 0.3907153308391571,
+      "learning_rate": 4.915596195463773e-05,
+      "loss": 0.2773,
+      "step": 7029
+    },
+    {
+      "epoch": 2.2571841387060525,
+      "grad_norm": 0.5676328539848328,
+      "learning_rate": 4.912590709037335e-05,
+      "loss": 0.6601,
+      "step": 7030
+    },
+    {
+      "epoch": 2.2575052175309036,
+      "grad_norm": 0.5613508224487305,
+      "learning_rate": 4.909585842496287e-05,
+      "loss": 0.4196,
+      "step": 7031
+    },
+    {
+      "epoch": 2.257826296355755,
+      "grad_norm": 0.6825602054595947,
+      "learning_rate": 4.906581596206764e-05,
+      "loss": 0.4297,
+      "step": 7032
+    },
+    {
+      "epoch": 2.2581473751806067,
+      "grad_norm": 0.6425372362136841,
+      "learning_rate": 4.9035779705348226e-05,
+      "loss": 0.1434,
+      "step": 7033
+    },
+    {
+      "epoch": 2.2584684540054583,
+      "grad_norm": 0.3993998169898987,
+      "learning_rate": 4.900574965846447e-05,
+      "loss": 0.1526,
+      "step": 7034
+    },
+    {
+      "epoch": 2.25878953283031,
+      "grad_norm": 0.9571073651313782,
+      "learning_rate": 4.8975725825075435e-05,
+      "loss": 0.4906,
+      "step": 7035
+    },
+    {
+      "epoch": 2.2591106116551614,
+      "grad_norm": 0.9649335741996765,
+      "learning_rate": 4.894570820883944e-05,
+      "loss": 0.4216,
+      "step": 7036
+    },
+    {
+      "epoch": 2.259431690480013,
+      "grad_norm": 0.8993098735809326,
+      "learning_rate": 4.8915696813414026e-05,
+      "loss": 0.4621,
+      "step": 7037
+    },
+    {
+      "epoch": 2.2597527693048645,
+      "grad_norm": 0.8691626191139221,
+      "learning_rate": 4.888569164245601e-05,
+      "loss": 0.3543,
+      "step": 7038
+    },
+    {
+      "epoch": 2.260073848129716,
+      "grad_norm": 1.0925235748291016,
+      "learning_rate": 4.885569269962142e-05,
+      "loss": 0.3965,
+      "step": 7039
+    },
+    {
+      "epoch": 2.260394926954567,
+      "grad_norm": 0.9121451377868652,
+      "learning_rate": 4.8825699988565485e-05,
+      "loss": 0.3612,
+      "step": 7040
+    },
+    {
+      "epoch": 2.2607160057794187,
+      "grad_norm": 0.8157550096511841,
+      "learning_rate": 4.8795713512942865e-05,
+      "loss": 0.3485,
+      "step": 7041
+    },
+    {
+      "epoch": 2.26103708460427,
+      "grad_norm": 0.691774845123291,
+      "learning_rate": 4.8765733276407156e-05,
+      "loss": 0.285,
+      "step": 7042
+    },
+    {
+      "epoch": 2.2613581634291218,
+      "grad_norm": 0.9489686489105225,
+      "learning_rate": 4.8735759282611516e-05,
+      "loss": 0.3201,
+      "step": 7043
+    },
+    {
+      "epoch": 2.2616792422539733,
+      "grad_norm": 0.9636274576187134,
+      "learning_rate": 4.870579153520807e-05,
+      "loss": 0.3511,
+      "step": 7044
+    },
+    {
+      "epoch": 2.262000321078825,
+      "grad_norm": 0.8838746547698975,
+      "learning_rate": 4.867583003784829e-05,
+      "loss": 0.3283,
+      "step": 7045
+    },
+    {
+      "epoch": 2.2623213999036764,
+      "grad_norm": 0.8975582718849182,
+      "learning_rate": 4.864587479418302e-05,
+      "loss": 0.3509,
+      "step": 7046
+    },
+    {
+      "epoch": 2.262642478728528,
+      "grad_norm": 0.8758841156959534,
+      "learning_rate": 4.861592580786205e-05,
+      "loss": 0.3312,
+      "step": 7047
+    },
+    {
+      "epoch": 2.2629635575533795,
+      "grad_norm": 0.8865678310394287,
+      "learning_rate": 4.858598308253473e-05,
+      "loss": 0.4163,
+      "step": 7048
+    },
+    {
+      "epoch": 2.2632846363782306,
+      "grad_norm": 0.7309367656707764,
+      "learning_rate": 4.8556046621849346e-05,
+      "loss": 0.3007,
+      "step": 7049
+    },
+    {
+      "epoch": 2.263605715203082,
+      "grad_norm": 1.0383028984069824,
+      "learning_rate": 4.852611642945368e-05,
+      "loss": 0.4668,
+      "step": 7050
+    },
+    {
+      "epoch": 2.2639267940279337,
+      "grad_norm": 0.8784850239753723,
+      "learning_rate": 4.8496192508994576e-05,
+      "loss": 0.2978,
+      "step": 7051
+    },
+    {
+      "epoch": 2.2642478728527853,
+      "grad_norm": 0.8731421232223511,
+      "learning_rate": 4.84662748641182e-05,
+      "loss": 0.3559,
+      "step": 7052
+    },
+    {
+      "epoch": 2.264568951677637,
+      "grad_norm": 0.8055264353752136,
+      "learning_rate": 4.8436363498469906e-05,
+      "loss": 0.333,
+      "step": 7053
+    },
+    {
+      "epoch": 2.2648900305024884,
+      "grad_norm": 0.7497748732566833,
+      "learning_rate": 4.840645841569431e-05,
+      "loss": 0.2937,
+      "step": 7054
+    },
+    {
+      "epoch": 2.26521110932734,
+      "grad_norm": 0.9812029600143433,
+      "learning_rate": 4.837655961943526e-05,
+      "loss": 0.4044,
+      "step": 7055
+    },
+    {
+      "epoch": 2.2655321881521915,
+      "grad_norm": 0.9688860177993774,
+      "learning_rate": 4.834666711333582e-05,
+      "loss": 0.3708,
+      "step": 7056
+    },
+    {
+      "epoch": 2.265853266977043,
+      "grad_norm": 0.8438839912414551,
+      "learning_rate": 4.8316780901038314e-05,
+      "loss": 0.2835,
+      "step": 7057
+    },
+    {
+      "epoch": 2.266174345801894,
+      "grad_norm": 1.0307844877243042,
+      "learning_rate": 4.828690098618429e-05,
+      "loss": 0.3232,
+      "step": 7058
+    },
+    {
+      "epoch": 2.2664954246267457,
+      "grad_norm": 0.6895515322685242,
+      "learning_rate": 4.825702737241452e-05,
+      "loss": 0.2816,
+      "step": 7059
+    },
+    {
+      "epoch": 2.2668165034515972,
+      "grad_norm": 0.8338602185249329,
+      "learning_rate": 4.822716006336897e-05,
+      "loss": 0.3282,
+      "step": 7060
+    },
+    {
+      "epoch": 2.267137582276449,
+      "grad_norm": 0.7110063433647156,
+      "learning_rate": 4.8197299062686995e-05,
+      "loss": 0.2868,
+      "step": 7061
+    },
+    {
+      "epoch": 2.2674586611013003,
+      "grad_norm": 0.8078935146331787,
+      "learning_rate": 4.816744437400697e-05,
+      "loss": 0.3183,
+      "step": 7062
+    },
+    {
+      "epoch": 2.267779739926152,
+      "grad_norm": 0.830915093421936,
+      "learning_rate": 4.8137596000966614e-05,
+      "loss": 0.3452,
+      "step": 7063
+    },
+    {
+      "epoch": 2.2681008187510034,
+      "grad_norm": 1.4586313962936401,
+      "learning_rate": 4.810775394720286e-05,
+      "loss": 0.4637,
+      "step": 7064
+    },
+    {
+      "epoch": 2.268421897575855,
+      "grad_norm": 0.9930241703987122,
+      "learning_rate": 4.807791821635186e-05,
+      "loss": 0.3232,
+      "step": 7065
+    },
+    {
+      "epoch": 2.2687429764007065,
+      "grad_norm": 0.7299748063087463,
+      "learning_rate": 4.8048088812049096e-05,
+      "loss": 0.313,
+      "step": 7066
+    },
+    {
+      "epoch": 2.2690640552255577,
+      "grad_norm": 0.6193024516105652,
+      "learning_rate": 4.8018265737929044e-05,
+      "loss": 0.285,
+      "step": 7067
+    },
+    {
+      "epoch": 2.269385134050409,
+      "grad_norm": 0.773335874080658,
+      "learning_rate": 4.798844899762568e-05,
+      "loss": 0.3412,
+      "step": 7068
+    },
+    {
+      "epoch": 2.2697062128752608,
+      "grad_norm": 0.9206930994987488,
+      "learning_rate": 4.7958638594772064e-05,
+      "loss": 0.3609,
+      "step": 7069
+    },
+    {
+      "epoch": 2.2700272917001123,
+      "grad_norm": 1.3238471746444702,
+      "learning_rate": 4.792883453300042e-05,
+      "loss": 0.3777,
+      "step": 7070
+    },
+    {
+      "epoch": 2.270348370524964,
+      "grad_norm": 0.8293914794921875,
+      "learning_rate": 4.78990368159424e-05,
+      "loss": 0.3592,
+      "step": 7071
+    },
+    {
+      "epoch": 2.2706694493498154,
+      "grad_norm": 0.9413627982139587,
+      "learning_rate": 4.786924544722864e-05,
+      "loss": 0.377,
+      "step": 7072
+    },
+    {
+      "epoch": 2.270990528174667,
+      "grad_norm": 0.9093666672706604,
+      "learning_rate": 4.783946043048923e-05,
+      "loss": 0.3278,
+      "step": 7073
+    },
+    {
+      "epoch": 2.2713116069995185,
+      "grad_norm": 0.8311013579368591,
+      "learning_rate": 4.780968176935333e-05,
+      "loss": 0.2915,
+      "step": 7074
+    },
+    {
+      "epoch": 2.27163268582437,
+      "grad_norm": 0.9607753157615662,
+      "learning_rate": 4.7779909467449414e-05,
+      "loss": 0.2904,
+      "step": 7075
+    },
+    {
+      "epoch": 2.271953764649221,
+      "grad_norm": 0.5113980174064636,
+      "learning_rate": 4.7750143528405126e-05,
+      "loss": 0.2676,
+      "step": 7076
+    },
+    {
+      "epoch": 2.2722748434740727,
+      "grad_norm": 0.7223218679428101,
+      "learning_rate": 4.7720383955847345e-05,
+      "loss": 0.291,
+      "step": 7077
+    },
+    {
+      "epoch": 2.2725959222989243,
+      "grad_norm": 0.48104968667030334,
+      "learning_rate": 4.769063075340222e-05,
+      "loss": 0.2631,
+      "step": 7078
+    },
+    {
+      "epoch": 2.272917001123776,
+      "grad_norm": 1.0848753452301025,
+      "learning_rate": 4.766088392469506e-05,
+      "loss": 0.3135,
+      "step": 7079
+    },
+    {
+      "epoch": 2.2732380799486274,
+      "grad_norm": 0.40874695777893066,
+      "learning_rate": 4.763114347335043e-05,
+      "loss": 0.4864,
+      "step": 7080
+    },
+    {
+      "epoch": 2.273559158773479,
+      "grad_norm": 0.5546656250953674,
+      "learning_rate": 4.7601409402992106e-05,
+      "loss": 0.3189,
+      "step": 7081
+    },
+    {
+      "epoch": 2.2738802375983305,
+      "grad_norm": 0.678066611289978,
+      "learning_rate": 4.757168171724311e-05,
+      "loss": 0.3145,
+      "step": 7082
+    },
+    {
+      "epoch": 2.274201316423182,
+      "grad_norm": 0.4192000925540924,
+      "learning_rate": 4.7541960419725626e-05,
+      "loss": 0.1832,
+      "step": 7083
+    },
+    {
+      "epoch": 2.2745223952480336,
+      "grad_norm": 0.24414893984794617,
+      "learning_rate": 4.7512245514061225e-05,
+      "loss": 0.0799,
+      "step": 7084
+    },
+    {
+      "epoch": 2.2748434740728847,
+      "grad_norm": 0.634087085723877,
+      "learning_rate": 4.748253700387042e-05,
+      "loss": 0.139,
+      "step": 7085
+    },
+    {
+      "epoch": 2.2751645528977362,
+      "grad_norm": 0.9760504364967346,
+      "learning_rate": 4.745283489277325e-05,
+      "loss": 0.522,
+      "step": 7086
+    },
+    {
+      "epoch": 2.2754856317225878,
+      "grad_norm": 0.9048032164573669,
+      "learning_rate": 4.742313918438872e-05,
+      "loss": 0.4195,
+      "step": 7087
+    },
+    {
+      "epoch": 2.2758067105474393,
+      "grad_norm": 0.9374763369560242,
+      "learning_rate": 4.739344988233516e-05,
+      "loss": 0.3933,
+      "step": 7088
+    },
+    {
+      "epoch": 2.276127789372291,
+      "grad_norm": 1.0608501434326172,
+      "learning_rate": 4.736376699023023e-05,
+      "loss": 0.4224,
+      "step": 7089
+    },
+    {
+      "epoch": 2.2764488681971424,
+      "grad_norm": 0.7070604562759399,
+      "learning_rate": 4.7334090511690554e-05,
+      "loss": 0.3017,
+      "step": 7090
+    },
+    {
+      "epoch": 2.276769947021994,
+      "grad_norm": 0.7293145060539246,
+      "learning_rate": 4.7304420450332244e-05,
+      "loss": 0.297,
+      "step": 7091
+    },
+    {
+      "epoch": 2.2770910258468455,
+      "grad_norm": 0.7650532722473145,
+      "learning_rate": 4.7274756809770446e-05,
+      "loss": 0.3605,
+      "step": 7092
+    },
+    {
+      "epoch": 2.277412104671697,
+      "grad_norm": 0.9350664019584656,
+      "learning_rate": 4.724509959361961e-05,
+      "loss": 0.3853,
+      "step": 7093
+    },
+    {
+      "epoch": 2.277733183496548,
+      "grad_norm": 0.8287744522094727,
+      "learning_rate": 4.721544880549337e-05,
+      "loss": 0.3868,
+      "step": 7094
+    },
+    {
+      "epoch": 2.2780542623213997,
+      "grad_norm": 0.7390232086181641,
+      "learning_rate": 4.7185804449004565e-05,
+      "loss": 0.3578,
+      "step": 7095
+    },
+    {
+      "epoch": 2.2783753411462513,
+      "grad_norm": 0.8684871196746826,
+      "learning_rate": 4.71561665277653e-05,
+      "loss": 0.3543,
+      "step": 7096
+    },
+    {
+      "epoch": 2.278696419971103,
+      "grad_norm": 0.6932856440544128,
+      "learning_rate": 4.712653504538683e-05,
+      "loss": 0.296,
+      "step": 7097
+    },
+    {
+      "epoch": 2.2790174987959544,
+      "grad_norm": 1.051287055015564,
+      "learning_rate": 4.70969100054797e-05,
+      "loss": 0.5495,
+      "step": 7098
+    },
+    {
+      "epoch": 2.279338577620806,
+      "grad_norm": 0.8018210530281067,
+      "learning_rate": 4.706729141165361e-05,
+      "loss": 0.2714,
+      "step": 7099
+    },
+    {
+      "epoch": 2.2796596564456575,
+      "grad_norm": 1.3825582265853882,
+      "learning_rate": 4.7037679267517495e-05,
+      "loss": 0.3997,
+      "step": 7100
+    },
+    {
+      "epoch": 2.279980735270509,
+      "grad_norm": 0.8683844804763794,
+      "learning_rate": 4.700807357667952e-05,
+      "loss": 0.3403,
+      "step": 7101
+    },
+    {
+      "epoch": 2.2803018140953606,
+      "grad_norm": 0.9737560749053955,
+      "learning_rate": 4.697847434274704e-05,
+      "loss": 0.3464,
+      "step": 7102
+    },
+    {
+      "epoch": 2.2806228929202117,
+      "grad_norm": 0.7259578704833984,
+      "learning_rate": 4.694888156932658e-05,
+      "loss": 0.3272,
+      "step": 7103
+    },
+    {
+      "epoch": 2.2809439717450632,
+      "grad_norm": 0.7659843564033508,
+      "learning_rate": 4.6919295260024054e-05,
+      "loss": 0.32,
+      "step": 7104
+    },
+    {
+      "epoch": 2.281265050569915,
+      "grad_norm": 1.1575708389282227,
+      "learning_rate": 4.688971541844436e-05,
+      "loss": 0.4335,
+      "step": 7105
+    },
+    {
+      "epoch": 2.2815861293947663,
+      "grad_norm": 0.7869051098823547,
+      "learning_rate": 4.68601420481917e-05,
+      "loss": 0.3432,
+      "step": 7106
+    },
+    {
+      "epoch": 2.281907208219618,
+      "grad_norm": 1.0924334526062012,
+      "learning_rate": 4.6830575152869616e-05,
+      "loss": 0.3652,
+      "step": 7107
+    },
+    {
+      "epoch": 2.2822282870444695,
+      "grad_norm": 0.8119152188301086,
+      "learning_rate": 4.6801014736080596e-05,
+      "loss": 0.2975,
+      "step": 7108
+    },
+    {
+      "epoch": 2.282549365869321,
+      "grad_norm": 0.9879711270332336,
+      "learning_rate": 4.6771460801426635e-05,
+      "loss": 0.3559,
+      "step": 7109
+    },
+    {
+      "epoch": 2.2828704446941726,
+      "grad_norm": 0.7146173119544983,
+      "learning_rate": 4.674191335250865e-05,
+      "loss": 0.2943,
+      "step": 7110
+    },
+    {
+      "epoch": 2.283191523519024,
+      "grad_norm": 0.7338638305664062,
+      "learning_rate": 4.6712372392927e-05,
+      "loss": 0.2994,
+      "step": 7111
+    },
+    {
+      "epoch": 2.283512602343875,
+      "grad_norm": 1.170583963394165,
+      "learning_rate": 4.668283792628114e-05,
+      "loss": 0.3887,
+      "step": 7112
+    },
+    {
+      "epoch": 2.2838336811687268,
+      "grad_norm": 0.9542192816734314,
+      "learning_rate": 4.665330995616974e-05,
+      "loss": 0.3925,
+      "step": 7113
+    },
+    {
+      "epoch": 2.2841547599935783,
+      "grad_norm": 0.7801535129547119,
+      "learning_rate": 4.6623788486190725e-05,
+      "loss": 0.2912,
+      "step": 7114
+    },
+    {
+      "epoch": 2.28447583881843,
+      "grad_norm": 1.0590436458587646,
+      "learning_rate": 4.6594273519941154e-05,
+      "loss": 0.3585,
+      "step": 7115
+    },
+    {
+      "epoch": 2.2847969176432814,
+      "grad_norm": 0.7104535102844238,
+      "learning_rate": 4.656476506101737e-05,
+      "loss": 0.2778,
+      "step": 7116
+    },
+    {
+      "epoch": 2.285117996468133,
+      "grad_norm": 0.7049524784088135,
+      "learning_rate": 4.653526311301488e-05,
+      "loss": 0.3042,
+      "step": 7117
+    },
+    {
+      "epoch": 2.2854390752929845,
+      "grad_norm": 1.7429343461990356,
+      "learning_rate": 4.65057676795284e-05,
+      "loss": 0.3351,
+      "step": 7118
+    },
+    {
+      "epoch": 2.285760154117836,
+      "grad_norm": 0.6940548419952393,
+      "learning_rate": 4.647627876415186e-05,
+      "loss": 0.2715,
+      "step": 7119
+    },
+    {
+      "epoch": 2.2860812329426876,
+      "grad_norm": 0.7398602366447449,
+      "learning_rate": 4.6446796370478394e-05,
+      "loss": 0.3182,
+      "step": 7120
+    },
+    {
+      "epoch": 2.2864023117675387,
+      "grad_norm": 1.2548072338104248,
+      "learning_rate": 4.6417320502100316e-05,
+      "loss": 0.3538,
+      "step": 7121
+    },
+    {
+      "epoch": 2.2867233905923903,
+      "grad_norm": 0.6731551289558411,
+      "learning_rate": 4.6387851162609275e-05,
+      "loss": 0.3286,
+      "step": 7122
+    },
+    {
+      "epoch": 2.287044469417242,
+      "grad_norm": 0.6933743357658386,
+      "learning_rate": 4.6358388355595904e-05,
+      "loss": 0.2885,
+      "step": 7123
+    },
+    {
+      "epoch": 2.2873655482420934,
+      "grad_norm": 1.1013816595077515,
+      "learning_rate": 4.632893208465021e-05,
+      "loss": 0.4317,
+      "step": 7124
+    },
+    {
+      "epoch": 2.287686627066945,
+      "grad_norm": 0.7992168664932251,
+      "learning_rate": 4.629948235336133e-05,
+      "loss": 0.317,
+      "step": 7125
+    },
+    {
+      "epoch": 2.2880077058917965,
+      "grad_norm": 0.8445520401000977,
+      "learning_rate": 4.6270039165317605e-05,
+      "loss": 0.3382,
+      "step": 7126
+    },
+    {
+      "epoch": 2.288328784716648,
+      "grad_norm": 0.7868288159370422,
+      "learning_rate": 4.62406025241067e-05,
+      "loss": 0.2714,
+      "step": 7127
+    },
+    {
+      "epoch": 2.2886498635414996,
+      "grad_norm": 0.7132185697555542,
+      "learning_rate": 4.621117243331523e-05,
+      "loss": 0.3171,
+      "step": 7128
+    },
+    {
+      "epoch": 2.288970942366351,
+      "grad_norm": 0.7570092082023621,
+      "learning_rate": 4.6181748896529273e-05,
+      "loss": 0.3292,
+      "step": 7129
+    },
+    {
+      "epoch": 2.2892920211912022,
+      "grad_norm": 0.5305631160736084,
+      "learning_rate": 4.615233191733398e-05,
+      "loss": 0.6934,
+      "step": 7130
+    },
+    {
+      "epoch": 2.289613100016054,
+      "grad_norm": 0.5653072595596313,
+      "learning_rate": 4.612292149931369e-05,
+      "loss": 0.7784,
+      "step": 7131
+    },
+    {
+      "epoch": 2.2899341788409053,
+      "grad_norm": 0.5794984102249146,
+      "learning_rate": 4.6093517646052034e-05,
+      "loss": 0.4634,
+      "step": 7132
+    },
+    {
+      "epoch": 2.290255257665757,
+      "grad_norm": 0.5841457843780518,
+      "learning_rate": 4.6064120361131656e-05,
+      "loss": 0.4593,
+      "step": 7133
+    },
+    {
+      "epoch": 2.2905763364906084,
+      "grad_norm": 0.6066222786903381,
+      "learning_rate": 4.603472964813466e-05,
+      "loss": 0.1791,
+      "step": 7134
+    },
+    {
+      "epoch": 2.29089741531546,
+      "grad_norm": 0.4143172800540924,
+      "learning_rate": 4.600534551064215e-05,
+      "loss": 0.0885,
+      "step": 7135
+    },
+    {
+      "epoch": 2.2912184941403115,
+      "grad_norm": 0.424333393573761,
+      "learning_rate": 4.59759679522345e-05,
+      "loss": 0.1697,
+      "step": 7136
+    },
+    {
+      "epoch": 2.291539572965163,
+      "grad_norm": 0.6681946516036987,
+      "learning_rate": 4.5946596976491295e-05,
+      "loss": 0.2762,
+      "step": 7137
+    },
+    {
+      "epoch": 2.2918606517900146,
+      "grad_norm": 0.9873928427696228,
+      "learning_rate": 4.591723258699127e-05,
+      "loss": 0.3777,
+      "step": 7138
+    },
+    {
+      "epoch": 2.2921817306148657,
+      "grad_norm": 0.8079444766044617,
+      "learning_rate": 4.588787478731242e-05,
+      "loss": 0.4295,
+      "step": 7139
+    },
+    {
+      "epoch": 2.2925028094397173,
+      "grad_norm": 0.8375623822212219,
+      "learning_rate": 4.5858523581031884e-05,
+      "loss": 0.2995,
+      "step": 7140
+    },
+    {
+      "epoch": 2.292823888264569,
+      "grad_norm": 1.0319302082061768,
+      "learning_rate": 4.582917897172603e-05,
+      "loss": 0.4127,
+      "step": 7141
+    },
+    {
+      "epoch": 2.2931449670894204,
+      "grad_norm": 0.7338537573814392,
+      "learning_rate": 4.579984096297038e-05,
+      "loss": 0.3225,
+      "step": 7142
+    },
+    {
+      "epoch": 2.293466045914272,
+      "grad_norm": 0.9782087206840515,
+      "learning_rate": 4.577050955833973e-05,
+      "loss": 0.3765,
+      "step": 7143
+    },
+    {
+      "epoch": 2.2937871247391235,
+      "grad_norm": 0.8286393284797668,
+      "learning_rate": 4.574118476140794e-05,
+      "loss": 0.3714,
+      "step": 7144
+    },
+    {
+      "epoch": 2.294108203563975,
+      "grad_norm": 0.7754277586936951,
+      "learning_rate": 4.5711866575748276e-05,
+      "loss": 0.3201,
+      "step": 7145
+    },
+    {
+      "epoch": 2.2944292823888266,
+      "grad_norm": 0.8050145506858826,
+      "learning_rate": 4.568255500493292e-05,
+      "loss": 0.3388,
+      "step": 7146
+    },
+    {
+      "epoch": 2.294750361213678,
+      "grad_norm": 0.9979050755500793,
+      "learning_rate": 4.565325005253356e-05,
+      "loss": 0.3467,
+      "step": 7147
+    },
+    {
+      "epoch": 2.2950714400385293,
+      "grad_norm": 0.6032088398933411,
+      "learning_rate": 4.5623951722120736e-05,
+      "loss": 0.2315,
+      "step": 7148
+    },
+    {
+      "epoch": 2.295392518863381,
+      "grad_norm": 0.8849102258682251,
+      "learning_rate": 4.559466001726451e-05,
+      "loss": 0.3303,
+      "step": 7149
+    },
+    {
+      "epoch": 2.2957135976882324,
+      "grad_norm": 0.9479863047599792,
+      "learning_rate": 4.5565374941533965e-05,
+      "loss": 0.3925,
+      "step": 7150
+    },
+    {
+      "epoch": 2.296034676513084,
+      "grad_norm": 0.8778762817382812,
+      "learning_rate": 4.5536096498497295e-05,
+      "loss": 0.3579,
+      "step": 7151
+    },
+    {
+      "epoch": 2.2963557553379355,
+      "grad_norm": 1.0045291185379028,
+      "learning_rate": 4.5506824691722126e-05,
+      "loss": 0.3575,
+      "step": 7152
+    },
+    {
+      "epoch": 2.296676834162787,
+      "grad_norm": 1.0932892560958862,
+      "learning_rate": 4.5477559524774994e-05,
+      "loss": 0.4418,
+      "step": 7153
+    },
+    {
+      "epoch": 2.2969979129876386,
+      "grad_norm": 1.10843026638031,
+      "learning_rate": 4.5448301001221895e-05,
+      "loss": 0.5555,
+      "step": 7154
+    },
+    {
+      "epoch": 2.29731899181249,
+      "grad_norm": 0.7526745200157166,
+      "learning_rate": 4.541904912462784e-05,
+      "loss": 0.3106,
+      "step": 7155
+    },
+    {
+      "epoch": 2.2976400706373417,
+      "grad_norm": 0.8937078714370728,
+      "learning_rate": 4.5389803898557106e-05,
+      "loss": 0.3546,
+      "step": 7156
+    },
+    {
+      "epoch": 2.2979611494621928,
+      "grad_norm": 0.660315752029419,
+      "learning_rate": 4.5360565326573104e-05,
+      "loss": 0.2998,
+      "step": 7157
+    },
+    {
+      "epoch": 2.2982822282870443,
+      "grad_norm": 1.0045713186264038,
+      "learning_rate": 4.5331333412238475e-05,
+      "loss": 0.497,
+      "step": 7158
+    },
+    {
+      "epoch": 2.298603307111896,
+      "grad_norm": 1.044492244720459,
+      "learning_rate": 4.530210815911504e-05,
+      "loss": 0.4079,
+      "step": 7159
+    },
+    {
+      "epoch": 2.2989243859367474,
+      "grad_norm": 0.894120991230011,
+      "learning_rate": 4.527288957076382e-05,
+      "loss": 0.364,
+      "step": 7160
+    },
+    {
+      "epoch": 2.299245464761599,
+      "grad_norm": 0.7585509419441223,
+      "learning_rate": 4.524367765074499e-05,
+      "loss": 0.2742,
+      "step": 7161
+    },
+    {
+      "epoch": 2.2995665435864505,
+      "grad_norm": 1.1006792783737183,
+      "learning_rate": 4.5214472402617944e-05,
+      "loss": 0.4268,
+      "step": 7162
+    },
+    {
+      "epoch": 2.299887622411302,
+      "grad_norm": 0.7505506873130798,
+      "learning_rate": 4.518527382994127e-05,
+      "loss": 0.3131,
+      "step": 7163
+    },
+    {
+      "epoch": 2.3002087012361536,
+      "grad_norm": 0.645842432975769,
+      "learning_rate": 4.515608193627265e-05,
+      "loss": 0.2745,
+      "step": 7164
+    },
+    {
+      "epoch": 2.300529780061005,
+      "grad_norm": 0.8791124820709229,
+      "learning_rate": 4.512689672516918e-05,
+      "loss": 0.4013,
+      "step": 7165
+    },
+    {
+      "epoch": 2.3008508588858563,
+      "grad_norm": 0.6512730121612549,
+      "learning_rate": 4.5097718200186814e-05,
+      "loss": 0.2472,
+      "step": 7166
+    },
+    {
+      "epoch": 2.301171937710708,
+      "grad_norm": 0.7656790018081665,
+      "learning_rate": 4.506854636488103e-05,
+      "loss": 0.3403,
+      "step": 7167
+    },
+    {
+      "epoch": 2.3014930165355594,
+      "grad_norm": 0.6794445514678955,
+      "learning_rate": 4.50393812228062e-05,
+      "loss": 0.3161,
+      "step": 7168
+    },
+    {
+      "epoch": 2.301814095360411,
+      "grad_norm": 0.6210803389549255,
+      "learning_rate": 4.501022277751602e-05,
+      "loss": 0.2855,
+      "step": 7169
+    },
+    {
+      "epoch": 2.3021351741852625,
+      "grad_norm": 0.748984694480896,
+      "learning_rate": 4.498107103256346e-05,
+      "loss": 0.2983,
+      "step": 7170
+    },
+    {
+      "epoch": 2.302456253010114,
+      "grad_norm": 0.7318181395530701,
+      "learning_rate": 4.495192599150044e-05,
+      "loss": 0.2941,
+      "step": 7171
+    },
+    {
+      "epoch": 2.3027773318349656,
+      "grad_norm": 0.8461513519287109,
+      "learning_rate": 4.4922787657878294e-05,
+      "loss": 0.3321,
+      "step": 7172
+    },
+    {
+      "epoch": 2.303098410659817,
+      "grad_norm": 0.4881146252155304,
+      "learning_rate": 4.48936560352474e-05,
+      "loss": 0.2739,
+      "step": 7173
+    },
+    {
+      "epoch": 2.3034194894846687,
+      "grad_norm": 1.058592677116394,
+      "learning_rate": 4.4864531127157374e-05,
+      "loss": 0.381,
+      "step": 7174
+    },
+    {
+      "epoch": 2.30374056830952,
+      "grad_norm": 0.6712564826011658,
+      "learning_rate": 4.483541293715698e-05,
+      "loss": 0.3067,
+      "step": 7175
+    },
+    {
+      "epoch": 2.3040616471343713,
+      "grad_norm": 0.9052884578704834,
+      "learning_rate": 4.480630146879419e-05,
+      "loss": 0.2888,
+      "step": 7176
+    },
+    {
+      "epoch": 2.304382725959223,
+      "grad_norm": 0.6388865113258362,
+      "learning_rate": 4.4777196725616146e-05,
+      "loss": 0.3019,
+      "step": 7177
+    },
+    {
+      "epoch": 2.3047038047840744,
+      "grad_norm": 1.7161591053009033,
+      "learning_rate": 4.474809871116916e-05,
+      "loss": 0.3845,
+      "step": 7178
+    },
+    {
+      "epoch": 2.305024883608926,
+      "grad_norm": 0.959423840045929,
+      "learning_rate": 4.471900742899876e-05,
+      "loss": 0.3427,
+      "step": 7179
+    },
+    {
+      "epoch": 2.3053459624337775,
+      "grad_norm": 0.47565457224845886,
+      "learning_rate": 4.4689922882649626e-05,
+      "loss": 0.6239,
+      "step": 7180
+    },
+    {
+      "epoch": 2.305667041258629,
+      "grad_norm": 0.5845574140548706,
+      "learning_rate": 4.46608450756656e-05,
+      "loss": 0.6796,
+      "step": 7181
+    },
+    {
+      "epoch": 2.3059881200834806,
+      "grad_norm": 0.6896982192993164,
+      "learning_rate": 4.463177401158975e-05,
+      "loss": 0.3241,
+      "step": 7182
+    },
+    {
+      "epoch": 2.306309198908332,
+      "grad_norm": 0.45037001371383667,
+      "learning_rate": 4.460270969396429e-05,
+      "loss": 0.1748,
+      "step": 7183
+    },
+    {
+      "epoch": 2.3066302777331833,
+      "grad_norm": 0.4934408366680145,
+      "learning_rate": 4.457365212633058e-05,
+      "loss": 0.2459,
+      "step": 7184
+    },
+    {
+      "epoch": 2.306951356558035,
+      "grad_norm": 0.6035980582237244,
+      "learning_rate": 4.45446013122293e-05,
+      "loss": 0.2141,
+      "step": 7185
+    },
+    {
+      "epoch": 2.3072724353828864,
+      "grad_norm": 0.8812675476074219,
+      "learning_rate": 4.451555725520009e-05,
+      "loss": 0.4984,
+      "step": 7186
+    },
+    {
+      "epoch": 2.307593514207738,
+      "grad_norm": 0.7593944072723389,
+      "learning_rate": 4.44865199587819e-05,
+      "loss": 0.3683,
+      "step": 7187
+    },
+    {
+      "epoch": 2.3079145930325895,
+      "grad_norm": 0.6698897480964661,
+      "learning_rate": 4.4457489426512947e-05,
+      "loss": 0.2978,
+      "step": 7188
+    },
+    {
+      "epoch": 2.308235671857441,
+      "grad_norm": 0.8218667507171631,
+      "learning_rate": 4.4428465661930343e-05,
+      "loss": 0.3963,
+      "step": 7189
+    },
+    {
+      "epoch": 2.3085567506822926,
+      "grad_norm": 0.8323404788970947,
+      "learning_rate": 4.43994486685707e-05,
+      "loss": 0.3453,
+      "step": 7190
+    },
+    {
+      "epoch": 2.308877829507144,
+      "grad_norm": 0.8024272322654724,
+      "learning_rate": 4.437043844996952e-05,
+      "loss": 0.2979,
+      "step": 7191
+    },
+    {
+      "epoch": 2.3091989083319957,
+      "grad_norm": 1.1842060089111328,
+      "learning_rate": 4.43414350096617e-05,
+      "loss": 0.4108,
+      "step": 7192
+    },
+    {
+      "epoch": 2.309519987156847,
+      "grad_norm": 0.9761362671852112,
+      "learning_rate": 4.431243835118124e-05,
+      "loss": 0.4397,
+      "step": 7193
+    },
+    {
+      "epoch": 2.3098410659816984,
+      "grad_norm": 0.8918820023536682,
+      "learning_rate": 4.428344847806116e-05,
+      "loss": 0.3433,
+      "step": 7194
+    },
+    {
+      "epoch": 2.31016214480655,
+      "grad_norm": 0.9746502041816711,
+      "learning_rate": 4.425446539383393e-05,
+      "loss": 0.404,
+      "step": 7195
+    },
+    {
+      "epoch": 2.3104832236314015,
+      "grad_norm": 0.7089119553565979,
+      "learning_rate": 4.4225489102030995e-05,
+      "loss": 0.241,
+      "step": 7196
+    },
+    {
+      "epoch": 2.310804302456253,
+      "grad_norm": 0.9124135971069336,
+      "learning_rate": 4.419651960618302e-05,
+      "loss": 0.3361,
+      "step": 7197
+    },
+    {
+      "epoch": 2.3111253812811046,
+      "grad_norm": 0.95118647813797,
+      "learning_rate": 4.4167556909819874e-05,
+      "loss": 0.3814,
+      "step": 7198
+    },
+    {
+      "epoch": 2.311446460105956,
+      "grad_norm": 1.066606044769287,
+      "learning_rate": 4.413860101647055e-05,
+      "loss": 0.3397,
+      "step": 7199
+    },
+    {
+      "epoch": 2.3117675389308077,
+      "grad_norm": 0.8261109590530396,
+      "learning_rate": 4.4109651929663256e-05,
+      "loss": 0.2883,
+      "step": 7200
+    },
+    {
+      "epoch": 2.312088617755659,
+      "grad_norm": 0.9831299185752869,
+      "learning_rate": 4.4080709652925336e-05,
+      "loss": 0.387,
+      "step": 7201
+    },
+    {
+      "epoch": 2.3124096965805103,
+      "grad_norm": 1.2849937677383423,
+      "learning_rate": 4.4051774189783315e-05,
+      "loss": 0.4081,
+      "step": 7202
+    },
+    {
+      "epoch": 2.312730775405362,
+      "grad_norm": 1.1613420248031616,
+      "learning_rate": 4.4022845543762915e-05,
+      "loss": 0.3917,
+      "step": 7203
+    },
+    {
+      "epoch": 2.3130518542302134,
+      "grad_norm": 0.9455772638320923,
+      "learning_rate": 4.399392371838897e-05,
+      "loss": 0.334,
+      "step": 7204
+    },
+    {
+      "epoch": 2.313372933055065,
+      "grad_norm": 0.8039372563362122,
+      "learning_rate": 4.396500871718555e-05,
+      "loss": 0.338,
+      "step": 7205
+    },
+    {
+      "epoch": 2.3136940118799165,
+      "grad_norm": 0.7344679832458496,
+      "learning_rate": 4.393610054367585e-05,
+      "loss": 0.3172,
+      "step": 7206
+    },
+    {
+      "epoch": 2.314015090704768,
+      "grad_norm": 0.6306678056716919,
+      "learning_rate": 4.39071992013822e-05,
+      "loss": 0.2823,
+      "step": 7207
+    },
+    {
+      "epoch": 2.3143361695296196,
+      "grad_norm": 0.9511836171150208,
+      "learning_rate": 4.387830469382624e-05,
+      "loss": 0.4142,
+      "step": 7208
+    },
+    {
+      "epoch": 2.314657248354471,
+      "grad_norm": 0.9060647487640381,
+      "learning_rate": 4.3849417024528564e-05,
+      "loss": 0.3342,
+      "step": 7209
+    },
+    {
+      "epoch": 2.3149783271793227,
+      "grad_norm": 0.821698784828186,
+      "learning_rate": 4.382053619700912e-05,
+      "loss": 0.3357,
+      "step": 7210
+    },
+    {
+      "epoch": 2.315299406004174,
+      "grad_norm": 0.7825961709022522,
+      "learning_rate": 4.379166221478697e-05,
+      "loss": 0.3329,
+      "step": 7211
+    },
+    {
+      "epoch": 2.3156204848290254,
+      "grad_norm": 0.9795680046081543,
+      "learning_rate": 4.3762795081380215e-05,
+      "loss": 0.3623,
+      "step": 7212
+    },
+    {
+      "epoch": 2.315941563653877,
+      "grad_norm": 0.7996529936790466,
+      "learning_rate": 4.3733934800306366e-05,
+      "loss": 0.2732,
+      "step": 7213
+    },
+    {
+      "epoch": 2.3162626424787285,
+      "grad_norm": 0.6189697980880737,
+      "learning_rate": 4.37050813750818e-05,
+      "loss": 0.2654,
+      "step": 7214
+    },
+    {
+      "epoch": 2.31658372130358,
+      "grad_norm": 0.7410408854484558,
+      "learning_rate": 4.367623480922236e-05,
+      "loss": 0.3059,
+      "step": 7215
+    },
+    {
+      "epoch": 2.3169048001284316,
+      "grad_norm": 0.8010258078575134,
+      "learning_rate": 4.364739510624286e-05,
+      "loss": 0.3012,
+      "step": 7216
+    },
+    {
+      "epoch": 2.317225878953283,
+      "grad_norm": 0.9137861728668213,
+      "learning_rate": 4.361856226965733e-05,
+      "loss": 0.391,
+      "step": 7217
+    },
+    {
+      "epoch": 2.3175469577781347,
+      "grad_norm": 0.7650619745254517,
+      "learning_rate": 4.3589736302978954e-05,
+      "loss": 0.3483,
+      "step": 7218
+    },
+    {
+      "epoch": 2.3178680366029862,
+      "grad_norm": 0.6533841490745544,
+      "learning_rate": 4.356091720972011e-05,
+      "loss": 0.2859,
+      "step": 7219
+    },
+    {
+      "epoch": 2.3181891154278373,
+      "grad_norm": 0.5599982738494873,
+      "learning_rate": 4.3532104993392306e-05,
+      "loss": 0.2805,
+      "step": 7220
+    },
+    {
+      "epoch": 2.318510194252689,
+      "grad_norm": 1.0042898654937744,
+      "learning_rate": 4.350329965750621e-05,
+      "loss": 0.4129,
+      "step": 7221
+    },
+    {
+      "epoch": 2.3188312730775404,
+      "grad_norm": 0.5694493055343628,
+      "learning_rate": 4.347450120557169e-05,
+      "loss": 0.269,
+      "step": 7222
+    },
+    {
+      "epoch": 2.319152351902392,
+      "grad_norm": 0.872643232345581,
+      "learning_rate": 4.3445709641097745e-05,
+      "loss": 0.3538,
+      "step": 7223
+    },
+    {
+      "epoch": 2.3194734307272435,
+      "grad_norm": 0.4571438133716583,
+      "learning_rate": 4.341692496759252e-05,
+      "loss": 0.2497,
+      "step": 7224
+    },
+    {
+      "epoch": 2.319794509552095,
+      "grad_norm": 0.6154420375823975,
+      "learning_rate": 4.3388147188563325e-05,
+      "loss": 0.2781,
+      "step": 7225
+    },
+    {
+      "epoch": 2.3201155883769466,
+      "grad_norm": 0.6062678098678589,
+      "learning_rate": 4.335937630751674e-05,
+      "loss": 0.2748,
+      "step": 7226
+    },
+    {
+      "epoch": 2.320436667201798,
+      "grad_norm": 0.8544726371765137,
+      "learning_rate": 4.333061232795826e-05,
+      "loss": 0.3392,
+      "step": 7227
+    },
+    {
+      "epoch": 2.3207577460266497,
+      "grad_norm": 0.8965178728103638,
+      "learning_rate": 4.3301855253392864e-05,
+      "loss": 0.3342,
+      "step": 7228
+    },
+    {
+      "epoch": 2.321078824851501,
+      "grad_norm": 0.7547941207885742,
+      "learning_rate": 4.327310508732437e-05,
+      "loss": 0.3099,
+      "step": 7229
+    },
+    {
+      "epoch": 2.3213999036763524,
+      "grad_norm": 0.43631595373153687,
+      "learning_rate": 4.324436183325593e-05,
+      "loss": 0.5266,
+      "step": 7230
+    },
+    {
+      "epoch": 2.321720982501204,
+      "grad_norm": 0.44915318489074707,
+      "learning_rate": 4.32156254946899e-05,
+      "loss": 0.586,
+      "step": 7231
+    },
+    {
+      "epoch": 2.3220420613260555,
+      "grad_norm": 0.4304579198360443,
+      "learning_rate": 4.3186896075127595e-05,
+      "loss": 0.3149,
+      "step": 7232
+    },
+    {
+      "epoch": 2.322363140150907,
+      "grad_norm": 0.5537540316581726,
+      "learning_rate": 4.315817357806974e-05,
+      "loss": 0.4514,
+      "step": 7233
+    },
+    {
+      "epoch": 2.3226842189757586,
+      "grad_norm": 0.4857487380504608,
+      "learning_rate": 4.3129458007015946e-05,
+      "loss": 0.3249,
+      "step": 7234
+    },
+    {
+      "epoch": 2.32300529780061,
+      "grad_norm": 0.44234228134155273,
+      "learning_rate": 4.310074936546521e-05,
+      "loss": 0.2186,
+      "step": 7235
+    },
+    {
+      "epoch": 2.3233263766254617,
+      "grad_norm": 0.29130613803863525,
+      "learning_rate": 4.307204765691558e-05,
+      "loss": 0.0781,
+      "step": 7236
+    },
+    {
+      "epoch": 2.3236474554503133,
+      "grad_norm": 0.6856487393379211,
+      "learning_rate": 4.304335288486426e-05,
+      "loss": 0.3749,
+      "step": 7237
+    },
+    {
+      "epoch": 2.3239685342751644,
+      "grad_norm": 0.7507144212722778,
+      "learning_rate": 4.301466505280762e-05,
+      "loss": 0.3243,
+      "step": 7238
+    },
+    {
+      "epoch": 2.324289613100016,
+      "grad_norm": 0.8678911328315735,
+      "learning_rate": 4.29859841642412e-05,
+      "loss": 0.4114,
+      "step": 7239
+    },
+    {
+      "epoch": 2.3246106919248675,
+      "grad_norm": 0.8836848139762878,
+      "learning_rate": 4.295731022265966e-05,
+      "loss": 0.3738,
+      "step": 7240
+    },
+    {
+      "epoch": 2.324931770749719,
+      "grad_norm": 0.7541016936302185,
+      "learning_rate": 4.2928643231556844e-05,
+      "loss": 0.3071,
+      "step": 7241
+    },
+    {
+      "epoch": 2.3252528495745706,
+      "grad_norm": 0.8909046053886414,
+      "learning_rate": 4.289998319442573e-05,
+      "loss": 0.4172,
+      "step": 7242
+    },
+    {
+      "epoch": 2.325573928399422,
+      "grad_norm": 0.733267605304718,
+      "learning_rate": 4.287133011475847e-05,
+      "loss": 0.3147,
+      "step": 7243
+    },
+    {
+      "epoch": 2.3258950072242737,
+      "grad_norm": 0.8695718050003052,
+      "learning_rate": 4.2842683996046327e-05,
+      "loss": 0.3655,
+      "step": 7244
+    },
+    {
+      "epoch": 2.326216086049125,
+      "grad_norm": 0.867275059223175,
+      "learning_rate": 4.2814044841779745e-05,
+      "loss": 0.4448,
+      "step": 7245
+    },
+    {
+      "epoch": 2.3265371648739768,
+      "grad_norm": 0.9108153581619263,
+      "learning_rate": 4.27854126554484e-05,
+      "loss": 0.3652,
+      "step": 7246
+    },
+    {
+      "epoch": 2.326858243698828,
+      "grad_norm": 0.5802766680717468,
+      "learning_rate": 4.2756787440540936e-05,
+      "loss": 0.2382,
+      "step": 7247
+    },
+    {
+      "epoch": 2.3271793225236794,
+      "grad_norm": 0.9157735109329224,
+      "learning_rate": 4.2728169200545286e-05,
+      "loss": 0.4234,
+      "step": 7248
+    },
+    {
+      "epoch": 2.327500401348531,
+      "grad_norm": 0.817433774471283,
+      "learning_rate": 4.26995579389485e-05,
+      "loss": 0.3119,
+      "step": 7249
+    },
+    {
+      "epoch": 2.3278214801733825,
+      "grad_norm": 0.793776273727417,
+      "learning_rate": 4.267095365923672e-05,
+      "loss": 0.3468,
+      "step": 7250
+    },
+    {
+      "epoch": 2.328142558998234,
+      "grad_norm": 0.7414565086364746,
+      "learning_rate": 4.264235636489542e-05,
+      "loss": 0.3007,
+      "step": 7251
+    },
+    {
+      "epoch": 2.3284636378230856,
+      "grad_norm": 0.74152672290802,
+      "learning_rate": 4.261376605940894e-05,
+      "loss": 0.2868,
+      "step": 7252
+    },
+    {
+      "epoch": 2.328784716647937,
+      "grad_norm": 0.9075784683227539,
+      "learning_rate": 4.2585182746261035e-05,
+      "loss": 0.3053,
+      "step": 7253
+    },
+    {
+      "epoch": 2.3291057954727887,
+      "grad_norm": 0.8360973000526428,
+      "learning_rate": 4.2556606428934443e-05,
+      "loss": 0.3616,
+      "step": 7254
+    },
+    {
+      "epoch": 2.3294268742976403,
+      "grad_norm": 1.0343592166900635,
+      "learning_rate": 4.252803711091112e-05,
+      "loss": 0.2661,
+      "step": 7255
+    },
+    {
+      "epoch": 2.3297479531224914,
+      "grad_norm": 0.8651792407035828,
+      "learning_rate": 4.249947479567218e-05,
+      "loss": 0.398,
+      "step": 7256
+    },
+    {
+      "epoch": 2.330069031947343,
+      "grad_norm": 0.8300990462303162,
+      "learning_rate": 4.2470919486697744e-05,
+      "loss": 0.3371,
+      "step": 7257
+    },
+    {
+      "epoch": 2.3303901107721945,
+      "grad_norm": 0.7871789932250977,
+      "learning_rate": 4.244237118746731e-05,
+      "loss": 0.2825,
+      "step": 7258
+    },
+    {
+      "epoch": 2.330711189597046,
+      "grad_norm": 0.9724371433258057,
+      "learning_rate": 4.2413829901459344e-05,
+      "loss": 0.2868,
+      "step": 7259
+    },
+    {
+      "epoch": 2.3310322684218976,
+      "grad_norm": 1.072011113166809,
+      "learning_rate": 4.238529563215153e-05,
+      "loss": 0.3748,
+      "step": 7260
+    },
+    {
+      "epoch": 2.331353347246749,
+      "grad_norm": 1.1171770095825195,
+      "learning_rate": 4.235676838302068e-05,
+      "loss": 0.5082,
+      "step": 7261
+    },
+    {
+      "epoch": 2.3316744260716007,
+      "grad_norm": 0.8948763012886047,
+      "learning_rate": 4.232824815754276e-05,
+      "loss": 0.3269,
+      "step": 7262
+    },
+    {
+      "epoch": 2.3319955048964522,
+      "grad_norm": 0.9017735123634338,
+      "learning_rate": 4.229973495919286e-05,
+      "loss": 0.3732,
+      "step": 7263
+    },
+    {
+      "epoch": 2.332316583721304,
+      "grad_norm": 0.8788909316062927,
+      "learning_rate": 4.227122879144523e-05,
+      "loss": 0.3456,
+      "step": 7264
+    },
+    {
+      "epoch": 2.332637662546155,
+      "grad_norm": 0.7098720669746399,
+      "learning_rate": 4.224272965777326e-05,
+      "loss": 0.2543,
+      "step": 7265
+    },
+    {
+      "epoch": 2.3329587413710065,
+      "grad_norm": 1.32795250415802,
+      "learning_rate": 4.221423756164948e-05,
+      "loss": 0.3675,
+      "step": 7266
+    },
+    {
+      "epoch": 2.333279820195858,
+      "grad_norm": 0.7488431930541992,
+      "learning_rate": 4.2185752506545585e-05,
+      "loss": 0.3516,
+      "step": 7267
+    },
+    {
+      "epoch": 2.3336008990207096,
+      "grad_norm": 0.5389450788497925,
+      "learning_rate": 4.215727449593233e-05,
+      "loss": 0.2551,
+      "step": 7268
+    },
+    {
+      "epoch": 2.333921977845561,
+      "grad_norm": 0.5667807459831238,
+      "learning_rate": 4.212880353327979e-05,
+      "loss": 0.2823,
+      "step": 7269
+    },
+    {
+      "epoch": 2.3342430566704127,
+      "grad_norm": 0.6024496555328369,
+      "learning_rate": 4.210033962205694e-05,
+      "loss": 0.2439,
+      "step": 7270
+    },
+    {
+      "epoch": 2.334564135495264,
+      "grad_norm": 0.6424733400344849,
+      "learning_rate": 4.207188276573214e-05,
+      "loss": 0.2782,
+      "step": 7271
+    },
+    {
+      "epoch": 2.3348852143201158,
+      "grad_norm": 1.2001317739486694,
+      "learning_rate": 4.204343296777265e-05,
+      "loss": 0.3473,
+      "step": 7272
+    },
+    {
+      "epoch": 2.3352062931449673,
+      "grad_norm": 0.7400069236755371,
+      "learning_rate": 4.201499023164508e-05,
+      "loss": 0.3178,
+      "step": 7273
+    },
+    {
+      "epoch": 2.3355273719698184,
+      "grad_norm": 0.6566978693008423,
+      "learning_rate": 4.1986554560815096e-05,
+      "loss": 0.3215,
+      "step": 7274
+    },
+    {
+      "epoch": 2.33584845079467,
+      "grad_norm": 0.6812121272087097,
+      "learning_rate": 4.195812595874739e-05,
+      "loss": 0.2494,
+      "step": 7275
+    },
+    {
+      "epoch": 2.3361695296195215,
+      "grad_norm": 0.8875928521156311,
+      "learning_rate": 4.1929704428906026e-05,
+      "loss": 0.3022,
+      "step": 7276
+    },
+    {
+      "epoch": 2.336490608444373,
+      "grad_norm": 0.6352400779724121,
+      "learning_rate": 4.190128997475402e-05,
+      "loss": 0.2617,
+      "step": 7277
+    },
+    {
+      "epoch": 2.3368116872692246,
+      "grad_norm": 0.5926759243011475,
+      "learning_rate": 4.1872882599753605e-05,
+      "loss": 0.2664,
+      "step": 7278
+    },
+    {
+      "epoch": 2.337132766094076,
+      "grad_norm": 0.41694337129592896,
+      "learning_rate": 4.184448230736613e-05,
+      "loss": 0.2739,
+      "step": 7279
+    },
+    {
+      "epoch": 2.3374538449189277,
+      "grad_norm": 0.5080072283744812,
+      "learning_rate": 4.181608910105207e-05,
+      "loss": 0.6664,
+      "step": 7280
+    },
+    {
+      "epoch": 2.3377749237437793,
+      "grad_norm": 0.468008428812027,
+      "learning_rate": 4.1787702984271074e-05,
+      "loss": 0.4515,
+      "step": 7281
+    },
+    {
+      "epoch": 2.338096002568631,
+      "grad_norm": 0.5170953869819641,
+      "learning_rate": 4.175932396048188e-05,
+      "loss": 0.2252,
+      "step": 7282
+    },
+    {
+      "epoch": 2.338417081393482,
+      "grad_norm": 0.46261003613471985,
+      "learning_rate": 4.173095203314241e-05,
+      "loss": 0.1327,
+      "step": 7283
+    },
+    {
+      "epoch": 2.3387381602183335,
+      "grad_norm": 0.5765483975410461,
+      "learning_rate": 4.170258720570968e-05,
+      "loss": 0.2548,
+      "step": 7284
+    },
+    {
+      "epoch": 2.339059239043185,
+      "grad_norm": 0.6683604121208191,
+      "learning_rate": 4.167422948163986e-05,
+      "loss": 0.2663,
+      "step": 7285
+    },
+    {
+      "epoch": 2.3393803178680366,
+      "grad_norm": 0.32332003116607666,
+      "learning_rate": 4.1645878864388266e-05,
+      "loss": 0.1391,
+      "step": 7286
+    },
+    {
+      "epoch": 2.339701396692888,
+      "grad_norm": 1.4612623453140259,
+      "learning_rate": 4.161753535740932e-05,
+      "loss": 0.3675,
+      "step": 7287
+    },
+    {
+      "epoch": 2.3400224755177397,
+      "grad_norm": 0.8074957728385925,
+      "learning_rate": 4.158919896415656e-05,
+      "loss": 0.3862,
+      "step": 7288
+    },
+    {
+      "epoch": 2.3403435543425912,
+      "grad_norm": 0.7950018048286438,
+      "learning_rate": 4.15608696880828e-05,
+      "loss": 0.3471,
+      "step": 7289
+    },
+    {
+      "epoch": 2.3406646331674428,
+      "grad_norm": 0.8618911504745483,
+      "learning_rate": 4.153254753263974e-05,
+      "loss": 0.3643,
+      "step": 7290
+    },
+    {
+      "epoch": 2.3409857119922943,
+      "grad_norm": 0.9194211959838867,
+      "learning_rate": 4.150423250127845e-05,
+      "loss": 0.4025,
+      "step": 7291
+    },
+    {
+      "epoch": 2.3413067908171454,
+      "grad_norm": 0.7834305167198181,
+      "learning_rate": 4.1475924597449024e-05,
+      "loss": 0.3126,
+      "step": 7292
+    },
+    {
+      "epoch": 2.341627869641997,
+      "grad_norm": 0.6242783665657043,
+      "learning_rate": 4.144762382460059e-05,
+      "loss": 0.2834,
+      "step": 7293
+    },
+    {
+      "epoch": 2.3419489484668485,
+      "grad_norm": 0.8568835854530334,
+      "learning_rate": 4.141933018618165e-05,
+      "loss": 0.3792,
+      "step": 7294
+    },
+    {
+      "epoch": 2.3422700272917,
+      "grad_norm": 1.036650538444519,
+      "learning_rate": 4.1391043685639576e-05,
+      "loss": 0.4148,
+      "step": 7295
+    },
+    {
+      "epoch": 2.3425911061165516,
+      "grad_norm": 0.9023113250732422,
+      "learning_rate": 4.1362764326421064e-05,
+      "loss": 0.3962,
+      "step": 7296
+    },
+    {
+      "epoch": 2.342912184941403,
+      "grad_norm": 0.9023537635803223,
+      "learning_rate": 4.133449211197188e-05,
+      "loss": 0.3818,
+      "step": 7297
+    },
+    {
+      "epoch": 2.3432332637662547,
+      "grad_norm": 1.3327304124832153,
+      "learning_rate": 4.130622704573685e-05,
+      "loss": 0.2894,
+      "step": 7298
+    },
+    {
+      "epoch": 2.3435543425911063,
+      "grad_norm": 0.8340547680854797,
+      "learning_rate": 4.1277969131160045e-05,
+      "loss": 0.3798,
+      "step": 7299
+    },
+    {
+      "epoch": 2.343875421415958,
+      "grad_norm": 0.6969391107559204,
+      "learning_rate": 4.1249718371684564e-05,
+      "loss": 0.3505,
+      "step": 7300
+    },
+    {
+      "epoch": 2.344196500240809,
+      "grad_norm": 0.7478737235069275,
+      "learning_rate": 4.12214747707527e-05,
+      "loss": 0.3221,
+      "step": 7301
+    },
+    {
+      "epoch": 2.3445175790656605,
+      "grad_norm": 1.0204554796218872,
+      "learning_rate": 4.1193238331805826e-05,
+      "loss": 0.4387,
+      "step": 7302
+    },
+    {
+      "epoch": 2.344838657890512,
+      "grad_norm": 1.032119870185852,
+      "learning_rate": 4.11650090582845e-05,
+      "loss": 0.3519,
+      "step": 7303
+    },
+    {
+      "epoch": 2.3451597367153636,
+      "grad_norm": 0.8070648908615112,
+      "learning_rate": 4.1136786953628334e-05,
+      "loss": 0.3181,
+      "step": 7304
+    },
+    {
+      "epoch": 2.345480815540215,
+      "grad_norm": 0.9976387023925781,
+      "learning_rate": 4.110857202127615e-05,
+      "loss": 0.2975,
+      "step": 7305
+    },
+    {
+      "epoch": 2.3458018943650667,
+      "grad_norm": 0.7166529297828674,
+      "learning_rate": 4.1080364264665774e-05,
+      "loss": 0.2918,
+      "step": 7306
+    },
+    {
+      "epoch": 2.3461229731899182,
+      "grad_norm": 1.119057297706604,
+      "learning_rate": 4.1052163687234366e-05,
+      "loss": 0.3691,
+      "step": 7307
+    },
+    {
+      "epoch": 2.34644405201477,
+      "grad_norm": 0.8771776556968689,
+      "learning_rate": 4.1023970292417935e-05,
+      "loss": 0.3273,
+      "step": 7308
+    },
+    {
+      "epoch": 2.3467651308396213,
+      "grad_norm": 1.007441759109497,
+      "learning_rate": 4.099578408365191e-05,
+      "loss": 0.39,
+      "step": 7309
+    },
+    {
+      "epoch": 2.3470862096644725,
+      "grad_norm": 0.7695625424385071,
+      "learning_rate": 4.096760506437057e-05,
+      "loss": 0.3232,
+      "step": 7310
+    },
+    {
+      "epoch": 2.347407288489324,
+      "grad_norm": 0.8058918714523315,
+      "learning_rate": 4.093943323800745e-05,
+      "loss": 0.279,
+      "step": 7311
+    },
+    {
+      "epoch": 2.3477283673141756,
+      "grad_norm": 1.1080282926559448,
+      "learning_rate": 4.0911268607995325e-05,
+      "loss": 0.4143,
+      "step": 7312
+    },
+    {
+      "epoch": 2.348049446139027,
+      "grad_norm": 0.9627980589866638,
+      "learning_rate": 4.08831111777658e-05,
+      "loss": 0.3703,
+      "step": 7313
+    },
+    {
+      "epoch": 2.3483705249638787,
+      "grad_norm": 0.8395978808403015,
+      "learning_rate": 4.08549609507499e-05,
+      "loss": 0.3836,
+      "step": 7314
+    },
+    {
+      "epoch": 2.34869160378873,
+      "grad_norm": 0.8003190755844116,
+      "learning_rate": 4.08268179303776e-05,
+      "loss": 0.3474,
+      "step": 7315
+    },
+    {
+      "epoch": 2.3490126826135818,
+      "grad_norm": 0.987792432308197,
+      "learning_rate": 4.0798682120078044e-05,
+      "loss": 0.3746,
+      "step": 7316
+    },
+    {
+      "epoch": 2.3493337614384333,
+      "grad_norm": 0.8142846822738647,
+      "learning_rate": 4.077055352327953e-05,
+      "loss": 0.3273,
+      "step": 7317
+    },
+    {
+      "epoch": 2.349654840263285,
+      "grad_norm": 0.8976936340332031,
+      "learning_rate": 4.074243214340934e-05,
+      "loss": 0.3165,
+      "step": 7318
+    },
+    {
+      "epoch": 2.349975919088136,
+      "grad_norm": 0.8904353380203247,
+      "learning_rate": 4.071431798389408e-05,
+      "loss": 0.296,
+      "step": 7319
+    },
+    {
+      "epoch": 2.3502969979129875,
+      "grad_norm": 1.0113767385482788,
+      "learning_rate": 4.068621104815934e-05,
+      "loss": 0.3628,
+      "step": 7320
+    },
+    {
+      "epoch": 2.350618076737839,
+      "grad_norm": 0.5627405643463135,
+      "learning_rate": 4.065811133962987e-05,
+      "loss": 0.2582,
+      "step": 7321
+    },
+    {
+      "epoch": 2.3509391555626906,
+      "grad_norm": 0.9106473326683044,
+      "learning_rate": 4.063001886172952e-05,
+      "loss": 0.3047,
+      "step": 7322
+    },
+    {
+      "epoch": 2.351260234387542,
+      "grad_norm": 1.275215744972229,
+      "learning_rate": 4.0601933617881294e-05,
+      "loss": 0.3613,
+      "step": 7323
+    },
+    {
+      "epoch": 2.3515813132123937,
+      "grad_norm": 0.7992632985115051,
+      "learning_rate": 4.057385561150727e-05,
+      "loss": 0.3085,
+      "step": 7324
+    },
+    {
+      "epoch": 2.3519023920372453,
+      "grad_norm": 0.7819589972496033,
+      "learning_rate": 4.05457848460287e-05,
+      "loss": 0.3735,
+      "step": 7325
+    },
+    {
+      "epoch": 2.352223470862097,
+      "grad_norm": 1.0428626537322998,
+      "learning_rate": 4.0517721324865884e-05,
+      "loss": 0.3611,
+      "step": 7326
+    },
+    {
+      "epoch": 2.3525445496869484,
+      "grad_norm": 0.7494415640830994,
+      "learning_rate": 4.048966505143831e-05,
+      "loss": 0.3112,
+      "step": 7327
+    },
+    {
+      "epoch": 2.3528656285117995,
+      "grad_norm": 0.7105958461761475,
+      "learning_rate": 4.0461616029164526e-05,
+      "loss": 0.3408,
+      "step": 7328
+    },
+    {
+      "epoch": 2.353186707336651,
+      "grad_norm": 0.5049455165863037,
+      "learning_rate": 4.0433574261462206e-05,
+      "loss": 0.3125,
+      "step": 7329
+    },
+    {
+      "epoch": 2.3535077861615026,
+      "grad_norm": 0.4792654514312744,
+      "learning_rate": 4.040553975174823e-05,
+      "loss": 0.6885,
+      "step": 7330
+    },
+    {
+      "epoch": 2.353828864986354,
+      "grad_norm": 0.450774222612381,
+      "learning_rate": 4.037751250343841e-05,
+      "loss": 0.4495,
+      "step": 7331
+    },
+    {
+      "epoch": 2.3541499438112057,
+      "grad_norm": 0.5843284130096436,
+      "learning_rate": 4.0349492519947904e-05,
+      "loss": 0.5532,
+      "step": 7332
+    },
+    {
+      "epoch": 2.3544710226360572,
+      "grad_norm": 0.535630464553833,
+      "learning_rate": 4.032147980469072e-05,
+      "loss": 0.316,
+      "step": 7333
+    },
+    {
+      "epoch": 2.354792101460909,
+      "grad_norm": 0.4793131649494171,
+      "learning_rate": 4.0293474361080244e-05,
+      "loss": 0.1362,
+      "step": 7334
+    },
+    {
+      "epoch": 2.3551131802857603,
+      "grad_norm": 0.4620800316333771,
+      "learning_rate": 4.026547619252883e-05,
+      "loss": 0.1559,
+      "step": 7335
+    },
+    {
+      "epoch": 2.355434259110612,
+      "grad_norm": 0.44343894720077515,
+      "learning_rate": 4.023748530244789e-05,
+      "loss": 0.162,
+      "step": 7336
+    },
+    {
+      "epoch": 2.355755337935463,
+      "grad_norm": 0.5997874140739441,
+      "learning_rate": 4.020950169424815e-05,
+      "loss": 0.2213,
+      "step": 7337
+    },
+    {
+      "epoch": 2.3560764167603145,
+      "grad_norm": 0.9611859917640686,
+      "learning_rate": 4.018152537133919e-05,
+      "loss": 0.5503,
+      "step": 7338
+    },
+    {
+      "epoch": 2.356397495585166,
+      "grad_norm": 0.9722205996513367,
+      "learning_rate": 4.015355633712996e-05,
+      "loss": 0.4235,
+      "step": 7339
+    },
+    {
+      "epoch": 2.3567185744100176,
+      "grad_norm": 0.6975258588790894,
+      "learning_rate": 4.012559459502835e-05,
+      "loss": 0.3188,
+      "step": 7340
+    },
+    {
+      "epoch": 2.357039653234869,
+      "grad_norm": 0.7221542000770569,
+      "learning_rate": 4.009764014844143e-05,
+      "loss": 0.3202,
+      "step": 7341
+    },
+    {
+      "epoch": 2.3573607320597207,
+      "grad_norm": 0.7738749980926514,
+      "learning_rate": 4.006969300077534e-05,
+      "loss": 0.2977,
+      "step": 7342
+    },
+    {
+      "epoch": 2.3576818108845723,
+      "grad_norm": 0.7268139123916626,
+      "learning_rate": 4.004175315543538e-05,
+      "loss": 0.2917,
+      "step": 7343
+    },
+    {
+      "epoch": 2.358002889709424,
+      "grad_norm": 0.9446465373039246,
+      "learning_rate": 4.001382061582593e-05,
+      "loss": 0.3598,
+      "step": 7344
+    },
+    {
+      "epoch": 2.3583239685342754,
+      "grad_norm": 1.0309242010116577,
+      "learning_rate": 3.9985895385350456e-05,
+      "loss": 0.349,
+      "step": 7345
+    },
+    {
+      "epoch": 2.3586450473591265,
+      "grad_norm": 0.9522349834442139,
+      "learning_rate": 3.9957977467411615e-05,
+      "loss": 0.3701,
+      "step": 7346
+    },
+    {
+      "epoch": 2.358966126183978,
+      "grad_norm": 0.9686524271965027,
+      "learning_rate": 3.9930066865411075e-05,
+      "loss": 0.4144,
+      "step": 7347
+    },
+    {
+      "epoch": 2.3592872050088296,
+      "grad_norm": 0.9719577431678772,
+      "learning_rate": 3.990216358274969e-05,
+      "loss": 0.3456,
+      "step": 7348
+    },
+    {
+      "epoch": 2.359608283833681,
+      "grad_norm": 0.8357756733894348,
+      "learning_rate": 3.987426762282733e-05,
+      "loss": 0.3546,
+      "step": 7349
+    },
+    {
+      "epoch": 2.3599293626585327,
+      "grad_norm": 0.7465749979019165,
+      "learning_rate": 3.9846378989043156e-05,
+      "loss": 0.3632,
+      "step": 7350
+    },
+    {
+      "epoch": 2.3602504414833843,
+      "grad_norm": 1.1760458946228027,
+      "learning_rate": 3.981849768479517e-05,
+      "loss": 0.3777,
+      "step": 7351
+    },
+    {
+      "epoch": 2.360571520308236,
+      "grad_norm": 0.82599937915802,
+      "learning_rate": 3.979062371348075e-05,
+      "loss": 0.3254,
+      "step": 7352
+    },
+    {
+      "epoch": 2.3608925991330874,
+      "grad_norm": 0.9712439179420471,
+      "learning_rate": 3.976275707849616e-05,
+      "loss": 0.4092,
+      "step": 7353
+    },
+    {
+      "epoch": 2.361213677957939,
+      "grad_norm": 1.3784904479980469,
+      "learning_rate": 3.973489778323688e-05,
+      "loss": 0.433,
+      "step": 7354
+    },
+    {
+      "epoch": 2.36153475678279,
+      "grad_norm": 0.9245650172233582,
+      "learning_rate": 3.9707045831097555e-05,
+      "loss": 0.343,
+      "step": 7355
+    },
+    {
+      "epoch": 2.3618558356076416,
+      "grad_norm": 0.994295060634613,
+      "learning_rate": 3.967920122547175e-05,
+      "loss": 0.3945,
+      "step": 7356
+    },
+    {
+      "epoch": 2.362176914432493,
+      "grad_norm": 0.8298449516296387,
+      "learning_rate": 3.9651363969752344e-05,
+      "loss": 0.3194,
+      "step": 7357
+    },
+    {
+      "epoch": 2.3624979932573447,
+      "grad_norm": 1.0529513359069824,
+      "learning_rate": 3.962353406733117e-05,
+      "loss": 0.3887,
+      "step": 7358
+    },
+    {
+      "epoch": 2.362819072082196,
+      "grad_norm": 1.0410268306732178,
+      "learning_rate": 3.9595711521599224e-05,
+      "loss": 0.4083,
+      "step": 7359
+    },
+    {
+      "epoch": 2.3631401509070478,
+      "grad_norm": 0.7109187245368958,
+      "learning_rate": 3.956789633594661e-05,
+      "loss": 0.2836,
+      "step": 7360
+    },
+    {
+      "epoch": 2.3634612297318993,
+      "grad_norm": 1.154443383216858,
+      "learning_rate": 3.954008851376252e-05,
+      "loss": 0.3007,
+      "step": 7361
+    },
+    {
+      "epoch": 2.363782308556751,
+      "grad_norm": 0.7189384698867798,
+      "learning_rate": 3.9512288058435256e-05,
+      "loss": 0.2745,
+      "step": 7362
+    },
+    {
+      "epoch": 2.3641033873816024,
+      "grad_norm": 0.5539619326591492,
+      "learning_rate": 3.948449497335219e-05,
+      "loss": 0.2625,
+      "step": 7363
+    },
+    {
+      "epoch": 2.3644244662064535,
+      "grad_norm": 0.5823628306388855,
+      "learning_rate": 3.945670926189987e-05,
+      "loss": 0.2553,
+      "step": 7364
+    },
+    {
+      "epoch": 2.364745545031305,
+      "grad_norm": 0.793720006942749,
+      "learning_rate": 3.942893092746387e-05,
+      "loss": 0.315,
+      "step": 7365
+    },
+    {
+      "epoch": 2.3650666238561566,
+      "grad_norm": 0.7939404845237732,
+      "learning_rate": 3.940115997342891e-05,
+      "loss": 0.3487,
+      "step": 7366
+    },
+    {
+      "epoch": 2.365387702681008,
+      "grad_norm": 0.80466628074646,
+      "learning_rate": 3.9373396403178786e-05,
+      "loss": 0.3018,
+      "step": 7367
+    },
+    {
+      "epoch": 2.3657087815058597,
+      "grad_norm": 1.0654207468032837,
+      "learning_rate": 3.9345640220096417e-05,
+      "loss": 0.4041,
+      "step": 7368
+    },
+    {
+      "epoch": 2.3660298603307113,
+      "grad_norm": 1.0553652048110962,
+      "learning_rate": 3.931789142756377e-05,
+      "loss": 0.3632,
+      "step": 7369
+    },
+    {
+      "epoch": 2.366350939155563,
+      "grad_norm": 0.7207555770874023,
+      "learning_rate": 3.9290150028962044e-05,
+      "loss": 0.3082,
+      "step": 7370
+    },
+    {
+      "epoch": 2.3666720179804144,
+      "grad_norm": 1.4064568281173706,
+      "learning_rate": 3.9262416027671356e-05,
+      "loss": 0.4244,
+      "step": 7371
+    },
+    {
+      "epoch": 2.366993096805266,
+      "grad_norm": 0.7677832841873169,
+      "learning_rate": 3.9234689427071006e-05,
+      "loss": 0.2716,
+      "step": 7372
+    },
+    {
+      "epoch": 2.367314175630117,
+      "grad_norm": 0.6761558651924133,
+      "learning_rate": 3.920697023053949e-05,
+      "loss": 0.2968,
+      "step": 7373
+    },
+    {
+      "epoch": 2.3676352544549686,
+      "grad_norm": 0.7641769051551819,
+      "learning_rate": 3.917925844145418e-05,
+      "loss": 0.3176,
+      "step": 7374
+    },
+    {
+      "epoch": 2.36795633327982,
+      "grad_norm": 0.5635340213775635,
+      "learning_rate": 3.915155406319181e-05,
+      "loss": 0.2882,
+      "step": 7375
+    },
+    {
+      "epoch": 2.3682774121046717,
+      "grad_norm": 0.689179539680481,
+      "learning_rate": 3.9123857099127936e-05,
+      "loss": 0.2785,
+      "step": 7376
+    },
+    {
+      "epoch": 2.3685984909295232,
+      "grad_norm": 0.6026139259338379,
+      "learning_rate": 3.9096167552637454e-05,
+      "loss": 0.2739,
+      "step": 7377
+    },
+    {
+      "epoch": 2.368919569754375,
+      "grad_norm": 0.6535028219223022,
+      "learning_rate": 3.9068485427094205e-05,
+      "loss": 0.2989,
+      "step": 7378
+    },
+    {
+      "epoch": 2.3692406485792263,
+      "grad_norm": 1.0952959060668945,
+      "learning_rate": 3.904081072587119e-05,
+      "loss": 0.3505,
+      "step": 7379
+    },
+    {
+      "epoch": 2.369561727404078,
+      "grad_norm": 0.41570642590522766,
+      "learning_rate": 3.9013143452340475e-05,
+      "loss": 0.3985,
+      "step": 7380
+    },
+    {
+      "epoch": 2.3698828062289294,
+      "grad_norm": 0.5073022246360779,
+      "learning_rate": 3.8985483609873244e-05,
+      "loss": 0.4265,
+      "step": 7381
+    },
+    {
+      "epoch": 2.3702038850537805,
+      "grad_norm": 0.5192015171051025,
+      "learning_rate": 3.895783120183976e-05,
+      "loss": 0.1128,
+      "step": 7382
+    },
+    {
+      "epoch": 2.370524963878632,
+      "grad_norm": 0.46728768944740295,
+      "learning_rate": 3.893018623160938e-05,
+      "loss": 0.1821,
+      "step": 7383
+    },
+    {
+      "epoch": 2.3708460427034836,
+      "grad_norm": 0.4381241202354431,
+      "learning_rate": 3.890254870255055e-05,
+      "loss": 0.1916,
+      "step": 7384
+    },
+    {
+      "epoch": 2.371167121528335,
+      "grad_norm": 0.6580232977867126,
+      "learning_rate": 3.887491861803085e-05,
+      "loss": 0.3314,
+      "step": 7385
+    },
+    {
+      "epoch": 2.3714882003531867,
+      "grad_norm": 1.0685169696807861,
+      "learning_rate": 3.8847295981416896e-05,
+      "loss": 0.5294,
+      "step": 7386
+    },
+    {
+      "epoch": 2.3718092791780383,
+      "grad_norm": 1.036335825920105,
+      "learning_rate": 3.88196807960744e-05,
+      "loss": 0.4178,
+      "step": 7387
+    },
+    {
+      "epoch": 2.37213035800289,
+      "grad_norm": 0.7811726331710815,
+      "learning_rate": 3.879207306536829e-05,
+      "loss": 0.3555,
+      "step": 7388
+    },
+    {
+      "epoch": 2.3724514368277414,
+      "grad_norm": 0.7611583471298218,
+      "learning_rate": 3.876447279266238e-05,
+      "loss": 0.2854,
+      "step": 7389
+    },
+    {
+      "epoch": 2.372772515652593,
+      "grad_norm": 0.8770664930343628,
+      "learning_rate": 3.8736879981319695e-05,
+      "loss": 0.398,
+      "step": 7390
+    },
+    {
+      "epoch": 2.373093594477444,
+      "grad_norm": 1.128092885017395,
+      "learning_rate": 3.8709294634702376e-05,
+      "loss": 0.3486,
+      "step": 7391
+    },
+    {
+      "epoch": 2.3734146733022956,
+      "grad_norm": 1.1329503059387207,
+      "learning_rate": 3.868171675617155e-05,
+      "loss": 0.2897,
+      "step": 7392
+    },
+    {
+      "epoch": 2.373735752127147,
+      "grad_norm": 0.9214721322059631,
+      "learning_rate": 3.8654146349087606e-05,
+      "loss": 0.3414,
+      "step": 7393
+    },
+    {
+      "epoch": 2.3740568309519987,
+      "grad_norm": 0.9622142910957336,
+      "learning_rate": 3.862658341680977e-05,
+      "loss": 0.2992,
+      "step": 7394
+    },
+    {
+      "epoch": 2.3743779097768503,
+      "grad_norm": 0.94224613904953,
+      "learning_rate": 3.859902796269663e-05,
+      "loss": 0.2816,
+      "step": 7395
+    },
+    {
+      "epoch": 2.374698988601702,
+      "grad_norm": 0.7693343758583069,
+      "learning_rate": 3.857147999010567e-05,
+      "loss": 0.2936,
+      "step": 7396
+    },
+    {
+      "epoch": 2.3750200674265534,
+      "grad_norm": 0.846794068813324,
+      "learning_rate": 3.854393950239355e-05,
+      "loss": 0.2686,
+      "step": 7397
+    },
+    {
+      "epoch": 2.375341146251405,
+      "grad_norm": 0.8755759000778198,
+      "learning_rate": 3.851640650291603e-05,
+      "loss": 0.3316,
+      "step": 7398
+    },
+    {
+      "epoch": 2.3756622250762565,
+      "grad_norm": 0.867134153842926,
+      "learning_rate": 3.8488880995027786e-05,
+      "loss": 0.2939,
+      "step": 7399
+    },
+    {
+      "epoch": 2.3759833039011076,
+      "grad_norm": 0.9087311029434204,
+      "learning_rate": 3.846136298208285e-05,
+      "loss": 0.4108,
+      "step": 7400
+    },
+    {
+      "epoch": 2.376304382725959,
+      "grad_norm": 1.0739927291870117,
+      "learning_rate": 3.843385246743417e-05,
+      "loss": 0.5237,
+      "step": 7401
+    },
+    {
+      "epoch": 2.3766254615508107,
+      "grad_norm": 0.8434580564498901,
+      "learning_rate": 3.840634945443382e-05,
+      "loss": 0.3213,
+      "step": 7402
+    },
+    {
+      "epoch": 2.376946540375662,
+      "grad_norm": 0.887002170085907,
+      "learning_rate": 3.837885394643296e-05,
+      "loss": 0.3309,
+      "step": 7403
+    },
+    {
+      "epoch": 2.3772676192005138,
+      "grad_norm": 0.9817523956298828,
+      "learning_rate": 3.835136594678183e-05,
+      "loss": 0.3857,
+      "step": 7404
+    },
+    {
+      "epoch": 2.3775886980253653,
+      "grad_norm": 0.9014565944671631,
+      "learning_rate": 3.832388545882976e-05,
+      "loss": 0.3774,
+      "step": 7405
+    },
+    {
+      "epoch": 2.377909776850217,
+      "grad_norm": 0.9643536806106567,
+      "learning_rate": 3.829641248592515e-05,
+      "loss": 0.3548,
+      "step": 7406
+    },
+    {
+      "epoch": 2.3782308556750684,
+      "grad_norm": 0.7037240266799927,
+      "learning_rate": 3.8268947031415514e-05,
+      "loss": 0.2919,
+      "step": 7407
+    },
+    {
+      "epoch": 2.37855193449992,
+      "grad_norm": 0.9688251614570618,
+      "learning_rate": 3.824148909864744e-05,
+      "loss": 0.3215,
+      "step": 7408
+    },
+    {
+      "epoch": 2.378873013324771,
+      "grad_norm": 0.6717649698257446,
+      "learning_rate": 3.821403869096658e-05,
+      "loss": 0.3176,
+      "step": 7409
+    },
+    {
+      "epoch": 2.3791940921496226,
+      "grad_norm": 0.7562946081161499,
+      "learning_rate": 3.818659581171766e-05,
+      "loss": 0.3263,
+      "step": 7410
+    },
+    {
+      "epoch": 2.379515170974474,
+      "grad_norm": 0.9467628598213196,
+      "learning_rate": 3.8159160464244606e-05,
+      "loss": 0.3804,
+      "step": 7411
+    },
+    {
+      "epoch": 2.3798362497993257,
+      "grad_norm": 0.824565052986145,
+      "learning_rate": 3.81317326518902e-05,
+      "loss": 0.3667,
+      "step": 7412
+    },
+    {
+      "epoch": 2.3801573286241773,
+      "grad_norm": 0.9573568105697632,
+      "learning_rate": 3.8104312377996564e-05,
+      "loss": 0.4441,
+      "step": 7413
+    },
+    {
+      "epoch": 2.380478407449029,
+      "grad_norm": 0.6948115229606628,
+      "learning_rate": 3.807689964590466e-05,
+      "loss": 0.2781,
+      "step": 7414
+    },
+    {
+      "epoch": 2.3807994862738804,
+      "grad_norm": 0.6470552682876587,
+      "learning_rate": 3.8049494458954725e-05,
+      "loss": 0.3053,
+      "step": 7415
+    },
+    {
+      "epoch": 2.3811205650987315,
+      "grad_norm": 0.6895164847373962,
+      "learning_rate": 3.802209682048602e-05,
+      "loss": 0.3258,
+      "step": 7416
+    },
+    {
+      "epoch": 2.3814416439235835,
+      "grad_norm": 1.2019386291503906,
+      "learning_rate": 3.799470673383674e-05,
+      "loss": 0.3939,
+      "step": 7417
+    },
+    {
+      "epoch": 2.3817627227484346,
+      "grad_norm": 0.884876012802124,
+      "learning_rate": 3.796732420234443e-05,
+      "loss": 0.3349,
+      "step": 7418
+    },
+    {
+      "epoch": 2.382083801573286,
+      "grad_norm": 0.8389701247215271,
+      "learning_rate": 3.793994922934544e-05,
+      "loss": 0.3409,
+      "step": 7419
+    },
+    {
+      "epoch": 2.3824048803981377,
+      "grad_norm": 0.9509716629981995,
+      "learning_rate": 3.791258181817542e-05,
+      "loss": 0.3532,
+      "step": 7420
+    },
+    {
+      "epoch": 2.3827259592229892,
+      "grad_norm": 0.8582257628440857,
+      "learning_rate": 3.788522197216897e-05,
+      "loss": 0.3035,
+      "step": 7421
+    },
+    {
+      "epoch": 2.383047038047841,
+      "grad_norm": 0.9152477979660034,
+      "learning_rate": 3.785786969465981e-05,
+      "loss": 0.3324,
+      "step": 7422
+    },
+    {
+      "epoch": 2.3833681168726923,
+      "grad_norm": 0.8005543947219849,
+      "learning_rate": 3.783052498898073e-05,
+      "loss": 0.3398,
+      "step": 7423
+    },
+    {
+      "epoch": 2.383689195697544,
+      "grad_norm": 0.5987676382064819,
+      "learning_rate": 3.7803187858463605e-05,
+      "loss": 0.287,
+      "step": 7424
+    },
+    {
+      "epoch": 2.384010274522395,
+      "grad_norm": 0.4976361393928528,
+      "learning_rate": 3.777585830643937e-05,
+      "loss": 0.2578,
+      "step": 7425
+    },
+    {
+      "epoch": 2.384331353347247,
+      "grad_norm": 0.7815603017807007,
+      "learning_rate": 3.774853633623806e-05,
+      "loss": 0.3316,
+      "step": 7426
+    },
+    {
+      "epoch": 2.384652432172098,
+      "grad_norm": 0.36369022727012634,
+      "learning_rate": 3.7721221951188765e-05,
+      "loss": 0.2671,
+      "step": 7427
+    },
+    {
+      "epoch": 2.3849735109969497,
+      "grad_norm": 0.5498353242874146,
+      "learning_rate": 3.769391515461966e-05,
+      "loss": 0.2742,
+      "step": 7428
+    },
+    {
+      "epoch": 2.385294589821801,
+      "grad_norm": 0.7687121629714966,
+      "learning_rate": 3.766661594985801e-05,
+      "loss": 0.2926,
+      "step": 7429
+    },
+    {
+      "epoch": 2.3856156686466528,
+      "grad_norm": 0.5641812086105347,
+      "learning_rate": 3.7639324340230085e-05,
+      "loss": 0.5458,
+      "step": 7430
+    },
+    {
+      "epoch": 2.3859367474715043,
+      "grad_norm": 0.49390679597854614,
+      "learning_rate": 3.7612040329061405e-05,
+      "loss": 0.5351,
+      "step": 7431
+    },
+    {
+      "epoch": 2.386257826296356,
+      "grad_norm": 0.48149552941322327,
+      "learning_rate": 3.7584763919676294e-05,
+      "loss": 0.2797,
+      "step": 7432
+    },
+    {
+      "epoch": 2.3865789051212074,
+      "grad_norm": 0.45125648379325867,
+      "learning_rate": 3.755749511539845e-05,
+      "loss": 0.2214,
+      "step": 7433
+    },
+    {
+      "epoch": 2.3868999839460585,
+      "grad_norm": 0.28079113364219666,
+      "learning_rate": 3.753023391955037e-05,
+      "loss": 0.0733,
+      "step": 7434
+    },
+    {
+      "epoch": 2.3872210627709105,
+      "grad_norm": 0.6530325412750244,
+      "learning_rate": 3.7502980335453774e-05,
+      "loss": 0.3432,
+      "step": 7435
+    },
+    {
+      "epoch": 2.3875421415957616,
+      "grad_norm": 0.9349000453948975,
+      "learning_rate": 3.747573436642951e-05,
+      "loss": 0.4311,
+      "step": 7436
+    },
+    {
+      "epoch": 2.387863220420613,
+      "grad_norm": 0.8858532905578613,
+      "learning_rate": 3.7448496015797295e-05,
+      "loss": 0.4246,
+      "step": 7437
+    },
+    {
+      "epoch": 2.3881842992454647,
+      "grad_norm": 1.0124034881591797,
+      "learning_rate": 3.742126528687614e-05,
+      "loss": 0.3771,
+      "step": 7438
+    },
+    {
+      "epoch": 2.3885053780703163,
+      "grad_norm": 0.7207339406013489,
+      "learning_rate": 3.739404218298398e-05,
+      "loss": 0.2938,
+      "step": 7439
+    },
+    {
+      "epoch": 2.388826456895168,
+      "grad_norm": 0.8159089684486389,
+      "learning_rate": 3.736682670743787e-05,
+      "loss": 0.3231,
+      "step": 7440
+    },
+    {
+      "epoch": 2.3891475357200194,
+      "grad_norm": 0.8773292303085327,
+      "learning_rate": 3.733961886355398e-05,
+      "loss": 0.3656,
+      "step": 7441
+    },
+    {
+      "epoch": 2.389468614544871,
+      "grad_norm": 0.7147423624992371,
+      "learning_rate": 3.7312418654647406e-05,
+      "loss": 0.311,
+      "step": 7442
+    },
+    {
+      "epoch": 2.389789693369722,
+      "grad_norm": 0.8349779844284058,
+      "learning_rate": 3.728522608403249e-05,
+      "loss": 0.3847,
+      "step": 7443
+    },
+    {
+      "epoch": 2.3901107721945736,
+      "grad_norm": 0.7083280682563782,
+      "learning_rate": 3.725804115502254e-05,
+      "loss": 0.265,
+      "step": 7444
+    },
+    {
+      "epoch": 2.390431851019425,
+      "grad_norm": 0.8586387634277344,
+      "learning_rate": 3.7230863870929964e-05,
+      "loss": 0.3667,
+      "step": 7445
+    },
+    {
+      "epoch": 2.3907529298442767,
+      "grad_norm": 0.7868812680244446,
+      "learning_rate": 3.720369423506622e-05,
+      "loss": 0.335,
+      "step": 7446
+    },
+    {
+      "epoch": 2.3910740086691282,
+      "grad_norm": 0.919206976890564,
+      "learning_rate": 3.717653225074186e-05,
+      "loss": 0.4075,
+      "step": 7447
+    },
+    {
+      "epoch": 2.3913950874939798,
+      "grad_norm": 0.9069111347198486,
+      "learning_rate": 3.714937792126647e-05,
+      "loss": 0.3928,
+      "step": 7448
+    },
+    {
+      "epoch": 2.3917161663188313,
+      "grad_norm": 0.7450005412101746,
+      "learning_rate": 3.712223124994875e-05,
+      "loss": 0.3179,
+      "step": 7449
+    },
+    {
+      "epoch": 2.392037245143683,
+      "grad_norm": 0.90874844789505,
+      "learning_rate": 3.709509224009641e-05,
+      "loss": 0.3934,
+      "step": 7450
+    },
+    {
+      "epoch": 2.3923583239685344,
+      "grad_norm": 0.797484815120697,
+      "learning_rate": 3.7067960895016275e-05,
+      "loss": 0.3171,
+      "step": 7451
+    },
+    {
+      "epoch": 2.3926794027933855,
+      "grad_norm": 1.188254475593567,
+      "learning_rate": 3.704083721801421e-05,
+      "loss": 0.4681,
+      "step": 7452
+    },
+    {
+      "epoch": 2.393000481618237,
+      "grad_norm": 0.8540223836898804,
+      "learning_rate": 3.701372121239512e-05,
+      "loss": 0.2897,
+      "step": 7453
+    },
+    {
+      "epoch": 2.3933215604430886,
+      "grad_norm": 0.6741155385971069,
+      "learning_rate": 3.698661288146311e-05,
+      "loss": 0.265,
+      "step": 7454
+    },
+    {
+      "epoch": 2.39364263926794,
+      "grad_norm": 0.8153589367866516,
+      "learning_rate": 3.695951222852112e-05,
+      "loss": 0.3477,
+      "step": 7455
+    },
+    {
+      "epoch": 2.3939637180927917,
+      "grad_norm": 0.7138598561286926,
+      "learning_rate": 3.6932419256871406e-05,
+      "loss": 0.3001,
+      "step": 7456
+    },
+    {
+      "epoch": 2.3942847969176433,
+      "grad_norm": 1.1851181983947754,
+      "learning_rate": 3.690533396981504e-05,
+      "loss": 0.4506,
+      "step": 7457
+    },
+    {
+      "epoch": 2.394605875742495,
+      "grad_norm": 0.9050045013427734,
+      "learning_rate": 3.687825637065236e-05,
+      "loss": 0.3637,
+      "step": 7458
+    },
+    {
+      "epoch": 2.3949269545673464,
+      "grad_norm": 0.7797844409942627,
+      "learning_rate": 3.685118646268272e-05,
+      "loss": 0.336,
+      "step": 7459
+    },
+    {
+      "epoch": 2.395248033392198,
+      "grad_norm": 1.0224764347076416,
+      "learning_rate": 3.682412424920438e-05,
+      "loss": 0.3581,
+      "step": 7460
+    },
+    {
+      "epoch": 2.395569112217049,
+      "grad_norm": 0.8182712197303772,
+      "learning_rate": 3.679706973351491e-05,
+      "loss": 0.307,
+      "step": 7461
+    },
+    {
+      "epoch": 2.3958901910419006,
+      "grad_norm": 0.8494866490364075,
+      "learning_rate": 3.677002291891078e-05,
+      "loss": 0.3239,
+      "step": 7462
+    },
+    {
+      "epoch": 2.396211269866752,
+      "grad_norm": 0.9537984728813171,
+      "learning_rate": 3.674298380868756e-05,
+      "loss": 0.3285,
+      "step": 7463
+    },
+    {
+      "epoch": 2.3965323486916037,
+      "grad_norm": 0.9284290075302124,
+      "learning_rate": 3.6715952406139885e-05,
+      "loss": 0.3925,
+      "step": 7464
+    },
+    {
+      "epoch": 2.3968534275164552,
+      "grad_norm": 1.1413391828536987,
+      "learning_rate": 3.668892871456144e-05,
+      "loss": 0.3875,
+      "step": 7465
+    },
+    {
+      "epoch": 2.397174506341307,
+      "grad_norm": 0.987825870513916,
+      "learning_rate": 3.6661912737245e-05,
+      "loss": 0.3362,
+      "step": 7466
+    },
+    {
+      "epoch": 2.3974955851661583,
+      "grad_norm": 0.8978787660598755,
+      "learning_rate": 3.6634904477482354e-05,
+      "loss": 0.3855,
+      "step": 7467
+    },
+    {
+      "epoch": 2.39781666399101,
+      "grad_norm": 0.7444339990615845,
+      "learning_rate": 3.6607903938564405e-05,
+      "loss": 0.2704,
+      "step": 7468
+    },
+    {
+      "epoch": 2.3981377428158615,
+      "grad_norm": 1.1076825857162476,
+      "learning_rate": 3.6580911123781056e-05,
+      "loss": 0.4605,
+      "step": 7469
+    },
+    {
+      "epoch": 2.3984588216407126,
+      "grad_norm": 0.8118841052055359,
+      "learning_rate": 3.6553926036421326e-05,
+      "loss": 0.2976,
+      "step": 7470
+    },
+    {
+      "epoch": 2.398779900465564,
+      "grad_norm": 0.8877241611480713,
+      "learning_rate": 3.6526948679773257e-05,
+      "loss": 0.3332,
+      "step": 7471
+    },
+    {
+      "epoch": 2.3991009792904157,
+      "grad_norm": 0.7601701617240906,
+      "learning_rate": 3.649997905712396e-05,
+      "loss": 0.3221,
+      "step": 7472
+    },
+    {
+      "epoch": 2.399422058115267,
+      "grad_norm": 1.4962290525436401,
+      "learning_rate": 3.647301717175956e-05,
+      "loss": 0.4457,
+      "step": 7473
+    },
+    {
+      "epoch": 2.3997431369401188,
+      "grad_norm": 0.8516422510147095,
+      "learning_rate": 3.6446063026965384e-05,
+      "loss": 0.3285,
+      "step": 7474
+    },
+    {
+      "epoch": 2.4000642157649703,
+      "grad_norm": 0.5712471008300781,
+      "learning_rate": 3.641911662602559e-05,
+      "loss": 0.2645,
+      "step": 7475
+    },
+    {
+      "epoch": 2.400385294589822,
+      "grad_norm": 0.7448149919509888,
+      "learning_rate": 3.6392177972223594e-05,
+      "loss": 0.3071,
+      "step": 7476
+    },
+    {
+      "epoch": 2.4007063734146734,
+      "grad_norm": 0.5864403247833252,
+      "learning_rate": 3.6365247068841814e-05,
+      "loss": 0.2586,
+      "step": 7477
+    },
+    {
+      "epoch": 2.401027452239525,
+      "grad_norm": 0.6279107928276062,
+      "learning_rate": 3.633832391916159e-05,
+      "loss": 0.2896,
+      "step": 7478
+    },
+    {
+      "epoch": 2.401348531064376,
+      "grad_norm": 0.7097012996673584,
+      "learning_rate": 3.631140852646355e-05,
+      "loss": 0.3143,
+      "step": 7479
+    },
+    {
+      "epoch": 2.4016696098892276,
+      "grad_norm": 0.42605191469192505,
+      "learning_rate": 3.628450089402713e-05,
+      "loss": 0.4756,
+      "step": 7480
+    },
+    {
+      "epoch": 2.401990688714079,
+      "grad_norm": 0.4604340195655823,
+      "learning_rate": 3.6257601025131026e-05,
+      "loss": 0.7031,
+      "step": 7481
+    },
+    {
+      "epoch": 2.4023117675389307,
+      "grad_norm": 0.3688209652900696,
+      "learning_rate": 3.62307089230529e-05,
+      "loss": 0.247,
+      "step": 7482
+    },
+    {
+      "epoch": 2.4026328463637823,
+      "grad_norm": 0.48397693037986755,
+      "learning_rate": 3.6203824591069456e-05,
+      "loss": 0.2488,
+      "step": 7483
+    },
+    {
+      "epoch": 2.402953925188634,
+      "grad_norm": 0.6211444735527039,
+      "learning_rate": 3.6176948032456473e-05,
+      "loss": 0.3253,
+      "step": 7484
+    },
+    {
+      "epoch": 2.4032750040134854,
+      "grad_norm": 0.5152127146720886,
+      "learning_rate": 3.615007925048878e-05,
+      "loss": 0.209,
+      "step": 7485
+    },
+    {
+      "epoch": 2.403596082838337,
+      "grad_norm": 0.44418853521347046,
+      "learning_rate": 3.612321824844024e-05,
+      "loss": 0.0993,
+      "step": 7486
+    },
+    {
+      "epoch": 2.4039171616631885,
+      "grad_norm": 0.5577342510223389,
+      "learning_rate": 3.60963650295838e-05,
+      "loss": 0.2749,
+      "step": 7487
+    },
+    {
+      "epoch": 2.4042382404880396,
+      "grad_norm": 0.7818149328231812,
+      "learning_rate": 3.606951959719145e-05,
+      "loss": 0.4772,
+      "step": 7488
+    },
+    {
+      "epoch": 2.404559319312891,
+      "grad_norm": 0.7911422252655029,
+      "learning_rate": 3.604268195453421e-05,
+      "loss": 0.3716,
+      "step": 7489
+    },
+    {
+      "epoch": 2.4048803981377427,
+      "grad_norm": 0.8051799535751343,
+      "learning_rate": 3.601585210488218e-05,
+      "loss": 0.4383,
+      "step": 7490
+    },
+    {
+      "epoch": 2.4052014769625942,
+      "grad_norm": 0.8039785623550415,
+      "learning_rate": 3.5989030051504434e-05,
+      "loss": 0.3691,
+      "step": 7491
+    },
+    {
+      "epoch": 2.405522555787446,
+      "grad_norm": 0.6632509827613831,
+      "learning_rate": 3.59622157976693e-05,
+      "loss": 0.2926,
+      "step": 7492
+    },
+    {
+      "epoch": 2.4058436346122973,
+      "grad_norm": 0.8523510694503784,
+      "learning_rate": 3.5935409346643835e-05,
+      "loss": 0.3532,
+      "step": 7493
+    },
+    {
+      "epoch": 2.406164713437149,
+      "grad_norm": 0.9791433811187744,
+      "learning_rate": 3.590861070169449e-05,
+      "loss": 0.3427,
+      "step": 7494
+    },
+    {
+      "epoch": 2.4064857922620004,
+      "grad_norm": 0.6575258374214172,
+      "learning_rate": 3.5881819866086484e-05,
+      "loss": 0.3172,
+      "step": 7495
+    },
+    {
+      "epoch": 2.406806871086852,
+      "grad_norm": 0.833772599697113,
+      "learning_rate": 3.585503684308421e-05,
+      "loss": 0.414,
+      "step": 7496
+    },
+    {
+      "epoch": 2.407127949911703,
+      "grad_norm": 0.9059905409812927,
+      "learning_rate": 3.582826163595119e-05,
+      "loss": 0.3809,
+      "step": 7497
+    },
+    {
+      "epoch": 2.4074490287365546,
+      "grad_norm": 1.0018295049667358,
+      "learning_rate": 3.580149424794976e-05,
+      "loss": 0.4041,
+      "step": 7498
+    },
+    {
+      "epoch": 2.407770107561406,
+      "grad_norm": 0.6466946005821228,
+      "learning_rate": 3.577473468234156e-05,
+      "loss": 0.2834,
+      "step": 7499
+    },
+    {
+      "epoch": 2.4080911863862577,
+      "grad_norm": 0.7386956214904785,
+      "learning_rate": 3.574798294238713e-05,
+      "loss": 0.3142,
+      "step": 7500
+    },
+    {
+      "epoch": 2.4084122652111093,
+      "grad_norm": 0.9483809471130371,
+      "learning_rate": 3.5721239031346066e-05,
+      "loss": 0.3326,
+      "step": 7501
+    },
+    {
+      "epoch": 2.408733344035961,
+      "grad_norm": 0.9816690683364868,
+      "learning_rate": 3.569450295247706e-05,
+      "loss": 0.3653,
+      "step": 7502
+    },
+    {
+      "epoch": 2.4090544228608124,
+      "grad_norm": 0.8358184099197388,
+      "learning_rate": 3.56677747090378e-05,
+      "loss": 0.3264,
+      "step": 7503
+    },
+    {
+      "epoch": 2.409375501685664,
+      "grad_norm": 0.8334686160087585,
+      "learning_rate": 3.564105430428506e-05,
+      "loss": 0.3277,
+      "step": 7504
+    },
+    {
+      "epoch": 2.4096965805105155,
+      "grad_norm": 0.8004488945007324,
+      "learning_rate": 3.561434174147463e-05,
+      "loss": 0.2976,
+      "step": 7505
+    },
+    {
+      "epoch": 2.4100176593353666,
+      "grad_norm": 0.6906147599220276,
+      "learning_rate": 3.558763702386135e-05,
+      "loss": 0.2825,
+      "step": 7506
+    },
+    {
+      "epoch": 2.410338738160218,
+      "grad_norm": 0.8044391870498657,
+      "learning_rate": 3.556094015469913e-05,
+      "loss": 0.3592,
+      "step": 7507
+    },
+    {
+      "epoch": 2.4106598169850697,
+      "grad_norm": 0.8112503290176392,
+      "learning_rate": 3.553425113724088e-05,
+      "loss": 0.36,
+      "step": 7508
+    },
+    {
+      "epoch": 2.4109808958099213,
+      "grad_norm": 0.9423220753669739,
+      "learning_rate": 3.5507569974738574e-05,
+      "loss": 0.3293,
+      "step": 7509
+    },
+    {
+      "epoch": 2.411301974634773,
+      "grad_norm": 0.973623514175415,
+      "learning_rate": 3.548089667044325e-05,
+      "loss": 0.4033,
+      "step": 7510
+    },
+    {
+      "epoch": 2.4116230534596244,
+      "grad_norm": 0.937961995601654,
+      "learning_rate": 3.545423122760493e-05,
+      "loss": 0.3519,
+      "step": 7511
+    },
+    {
+      "epoch": 2.411944132284476,
+      "grad_norm": 0.8615179061889648,
+      "learning_rate": 3.542757364947281e-05,
+      "loss": 0.4388,
+      "step": 7512
+    },
+    {
+      "epoch": 2.4122652111093275,
+      "grad_norm": 0.753441572189331,
+      "learning_rate": 3.540092393929494e-05,
+      "loss": 0.3198,
+      "step": 7513
+    },
+    {
+      "epoch": 2.412586289934179,
+      "grad_norm": 0.9324014186859131,
+      "learning_rate": 3.537428210031849e-05,
+      "loss": 0.3705,
+      "step": 7514
+    },
+    {
+      "epoch": 2.41290736875903,
+      "grad_norm": 0.7881529927253723,
+      "learning_rate": 3.534764813578982e-05,
+      "loss": 0.2409,
+      "step": 7515
+    },
+    {
+      "epoch": 2.4132284475838817,
+      "grad_norm": 0.6942733526229858,
+      "learning_rate": 3.5321022048954035e-05,
+      "loss": 0.2954,
+      "step": 7516
+    },
+    {
+      "epoch": 2.413549526408733,
+      "grad_norm": 0.7293675541877747,
+      "learning_rate": 3.52944038430556e-05,
+      "loss": 0.3183,
+      "step": 7517
+    },
+    {
+      "epoch": 2.4138706052335848,
+      "grad_norm": 1.0365976095199585,
+      "learning_rate": 3.52677935213377e-05,
+      "loss": 0.3541,
+      "step": 7518
+    },
+    {
+      "epoch": 2.4141916840584363,
+      "grad_norm": 0.6401132345199585,
+      "learning_rate": 3.524119108704286e-05,
+      "loss": 0.2692,
+      "step": 7519
+    },
+    {
+      "epoch": 2.414512762883288,
+      "grad_norm": 0.7052464485168457,
+      "learning_rate": 3.521459654341244e-05,
+      "loss": 0.3227,
+      "step": 7520
+    },
+    {
+      "epoch": 2.4148338417081394,
+      "grad_norm": 0.4697404205799103,
+      "learning_rate": 3.518800989368691e-05,
+      "loss": 0.2497,
+      "step": 7521
+    },
+    {
+      "epoch": 2.415154920532991,
+      "grad_norm": 0.8850452303886414,
+      "learning_rate": 3.516143114110582e-05,
+      "loss": 0.3252,
+      "step": 7522
+    },
+    {
+      "epoch": 2.4154759993578425,
+      "grad_norm": 0.8042798638343811,
+      "learning_rate": 3.51348602889076e-05,
+      "loss": 0.3285,
+      "step": 7523
+    },
+    {
+      "epoch": 2.4157970781826936,
+      "grad_norm": 0.9007611274719238,
+      "learning_rate": 3.510829734032993e-05,
+      "loss": 0.3187,
+      "step": 7524
+    },
+    {
+      "epoch": 2.416118157007545,
+      "grad_norm": 1.6435385942459106,
+      "learning_rate": 3.50817422986094e-05,
+      "loss": 0.2828,
+      "step": 7525
+    },
+    {
+      "epoch": 2.4164392358323967,
+      "grad_norm": 0.31267601251602173,
+      "learning_rate": 3.5055195166981645e-05,
+      "loss": 0.2312,
+      "step": 7526
+    },
+    {
+      "epoch": 2.4167603146572483,
+      "grad_norm": 0.6988046169281006,
+      "learning_rate": 3.5028655948681355e-05,
+      "loss": 0.314,
+      "step": 7527
+    },
+    {
+      "epoch": 2.4170813934821,
+      "grad_norm": 0.6505711078643799,
+      "learning_rate": 3.500212464694227e-05,
+      "loss": 0.289,
+      "step": 7528
+    },
+    {
+      "epoch": 2.4174024723069514,
+      "grad_norm": 0.6511385440826416,
+      "learning_rate": 3.497560126499709e-05,
+      "loss": 0.298,
+      "step": 7529
+    },
+    {
+      "epoch": 2.417723551131803,
+      "grad_norm": 0.5870752930641174,
+      "learning_rate": 3.494908580607774e-05,
+      "loss": 0.7781,
+      "step": 7530
+    },
+    {
+      "epoch": 2.4180446299566545,
+      "grad_norm": 0.5827964544296265,
+      "learning_rate": 3.492257827341492e-05,
+      "loss": 0.5468,
+      "step": 7531
+    },
+    {
+      "epoch": 2.418365708781506,
+      "grad_norm": 0.3877847194671631,
+      "learning_rate": 3.489607867023854e-05,
+      "loss": 0.2175,
+      "step": 7532
+    },
+    {
+      "epoch": 2.418686787606357,
+      "grad_norm": 0.42042702436447144,
+      "learning_rate": 3.4869586999777495e-05,
+      "loss": 0.1292,
+      "step": 7533
+    },
+    {
+      "epoch": 2.4190078664312087,
+      "grad_norm": 0.42274191975593567,
+      "learning_rate": 3.484310326525967e-05,
+      "loss": 0.0746,
+      "step": 7534
+    },
+    {
+      "epoch": 2.4193289452560602,
+      "grad_norm": 0.24631276726722717,
+      "learning_rate": 3.481662746991214e-05,
+      "loss": 0.0749,
+      "step": 7535
+    },
+    {
+      "epoch": 2.419650024080912,
+      "grad_norm": 0.48242461681365967,
+      "learning_rate": 3.479015961696077e-05,
+      "loss": 0.1554,
+      "step": 7536
+    },
+    {
+      "epoch": 2.4199711029057633,
+      "grad_norm": 0.8864598870277405,
+      "learning_rate": 3.4763699709630716e-05,
+      "loss": 0.4802,
+      "step": 7537
+    },
+    {
+      "epoch": 2.420292181730615,
+      "grad_norm": 1.016950249671936,
+      "learning_rate": 3.4737247751145896e-05,
+      "loss": 0.5052,
+      "step": 7538
+    },
+    {
+      "epoch": 2.4206132605554664,
+      "grad_norm": 0.9051849842071533,
+      "learning_rate": 3.4710803744729515e-05,
+      "loss": 0.381,
+      "step": 7539
+    },
+    {
+      "epoch": 2.420934339380318,
+      "grad_norm": 0.9121940732002258,
+      "learning_rate": 3.468436769360368e-05,
+      "loss": 0.3998,
+      "step": 7540
+    },
+    {
+      "epoch": 2.4212554182051695,
+      "grad_norm": 0.698441743850708,
+      "learning_rate": 3.465793960098945e-05,
+      "loss": 0.2916,
+      "step": 7541
+    },
+    {
+      "epoch": 2.4215764970300206,
+      "grad_norm": 0.7434770464897156,
+      "learning_rate": 3.463151947010712e-05,
+      "loss": 0.3105,
+      "step": 7542
+    },
+    {
+      "epoch": 2.421897575854872,
+      "grad_norm": 0.6803128123283386,
+      "learning_rate": 3.460510730417585e-05,
+      "loss": 0.332,
+      "step": 7543
+    },
+    {
+      "epoch": 2.4222186546797237,
+      "grad_norm": 0.7351831197738647,
+      "learning_rate": 3.4578703106413904e-05,
+      "loss": 0.2964,
+      "step": 7544
+    },
+    {
+      "epoch": 2.4225397335045753,
+      "grad_norm": 0.8372403383255005,
+      "learning_rate": 3.455230688003852e-05,
+      "loss": 0.3135,
+      "step": 7545
+    },
+    {
+      "epoch": 2.422860812329427,
+      "grad_norm": 0.7963913679122925,
+      "learning_rate": 3.452591862826603e-05,
+      "loss": 0.3311,
+      "step": 7546
+    },
+    {
+      "epoch": 2.4231818911542784,
+      "grad_norm": 0.9390442967414856,
+      "learning_rate": 3.4499538354311755e-05,
+      "loss": 0.4368,
+      "step": 7547
+    },
+    {
+      "epoch": 2.42350296997913,
+      "grad_norm": 0.8431958556175232,
+      "learning_rate": 3.447316606139004e-05,
+      "loss": 0.3649,
+      "step": 7548
+    },
+    {
+      "epoch": 2.4238240488039815,
+      "grad_norm": 1.0568405389785767,
+      "learning_rate": 3.444680175271428e-05,
+      "loss": 0.369,
+      "step": 7549
+    },
+    {
+      "epoch": 2.424145127628833,
+      "grad_norm": 0.8188615441322327,
+      "learning_rate": 3.442044543149688e-05,
+      "loss": 0.348,
+      "step": 7550
+    },
+    {
+      "epoch": 2.424466206453684,
+      "grad_norm": 1.063718557357788,
+      "learning_rate": 3.439409710094929e-05,
+      "loss": 0.4778,
+      "step": 7551
+    },
+    {
+      "epoch": 2.4247872852785357,
+      "grad_norm": 0.9398375153541565,
+      "learning_rate": 3.4367756764281955e-05,
+      "loss": 0.3605,
+      "step": 7552
+    },
+    {
+      "epoch": 2.4251083641033873,
+      "grad_norm": 0.9994428157806396,
+      "learning_rate": 3.4341424424704375e-05,
+      "loss": 0.3958,
+      "step": 7553
+    },
+    {
+      "epoch": 2.425429442928239,
+      "grad_norm": 0.8988476395606995,
+      "learning_rate": 3.4315100085425034e-05,
+      "loss": 0.3561,
+      "step": 7554
+    },
+    {
+      "epoch": 2.4257505217530904,
+      "grad_norm": 1.0344244241714478,
+      "learning_rate": 3.4288783749651564e-05,
+      "loss": 0.402,
+      "step": 7555
+    },
+    {
+      "epoch": 2.426071600577942,
+      "grad_norm": 1.1983534097671509,
+      "learning_rate": 3.426247542059041e-05,
+      "loss": 0.403,
+      "step": 7556
+    },
+    {
+      "epoch": 2.4263926794027935,
+      "grad_norm": 0.7842648029327393,
+      "learning_rate": 3.423617510144727e-05,
+      "loss": 0.3219,
+      "step": 7557
+    },
+    {
+      "epoch": 2.426713758227645,
+      "grad_norm": 0.8381021022796631,
+      "learning_rate": 3.4209882795426716e-05,
+      "loss": 0.3593,
+      "step": 7558
+    },
+    {
+      "epoch": 2.4270348370524966,
+      "grad_norm": 1.0330874919891357,
+      "learning_rate": 3.418359850573234e-05,
+      "loss": 0.3792,
+      "step": 7559
+    },
+    {
+      "epoch": 2.4273559158773477,
+      "grad_norm": 0.9594546556472778,
+      "learning_rate": 3.415732223556689e-05,
+      "loss": 0.3862,
+      "step": 7560
+    },
+    {
+      "epoch": 2.427676994702199,
+      "grad_norm": 0.9516655206680298,
+      "learning_rate": 3.413105398813195e-05,
+      "loss": 0.297,
+      "step": 7561
+    },
+    {
+      "epoch": 2.4279980735270508,
+      "grad_norm": 0.8917472958564758,
+      "learning_rate": 3.4104793766628304e-05,
+      "loss": 0.2836,
+      "step": 7562
+    },
+    {
+      "epoch": 2.4283191523519023,
+      "grad_norm": 0.5682129263877869,
+      "learning_rate": 3.4078541574255664e-05,
+      "loss": 0.2329,
+      "step": 7563
+    },
+    {
+      "epoch": 2.428640231176754,
+      "grad_norm": 0.8117547631263733,
+      "learning_rate": 3.4052297414212777e-05,
+      "loss": 0.3157,
+      "step": 7564
+    },
+    {
+      "epoch": 2.4289613100016054,
+      "grad_norm": 0.8948675394058228,
+      "learning_rate": 3.40260612896974e-05,
+      "loss": 0.2925,
+      "step": 7565
+    },
+    {
+      "epoch": 2.429282388826457,
+      "grad_norm": 0.8941418528556824,
+      "learning_rate": 3.3999833203906326e-05,
+      "loss": 0.2747,
+      "step": 7566
+    },
+    {
+      "epoch": 2.4296034676513085,
+      "grad_norm": 0.6833550333976746,
+      "learning_rate": 3.397361316003539e-05,
+      "loss": 0.2783,
+      "step": 7567
+    },
+    {
+      "epoch": 2.42992454647616,
+      "grad_norm": 0.571975588798523,
+      "learning_rate": 3.394740116127941e-05,
+      "loss": 0.2373,
+      "step": 7568
+    },
+    {
+      "epoch": 2.430245625301011,
+      "grad_norm": 0.6630203723907471,
+      "learning_rate": 3.3921197210832235e-05,
+      "loss": 0.2824,
+      "step": 7569
+    },
+    {
+      "epoch": 2.4305667041258627,
+      "grad_norm": 0.7997381091117859,
+      "learning_rate": 3.389500131188674e-05,
+      "loss": 0.3572,
+      "step": 7570
+    },
+    {
+      "epoch": 2.4308877829507143,
+      "grad_norm": 0.9815787076950073,
+      "learning_rate": 3.386881346763483e-05,
+      "loss": 0.4214,
+      "step": 7571
+    },
+    {
+      "epoch": 2.431208861775566,
+      "grad_norm": 0.8200189471244812,
+      "learning_rate": 3.3842633681267356e-05,
+      "loss": 0.3118,
+      "step": 7572
+    },
+    {
+      "epoch": 2.4315299406004174,
+      "grad_norm": 0.8650208115577698,
+      "learning_rate": 3.3816461955974365e-05,
+      "loss": 0.3753,
+      "step": 7573
+    },
+    {
+      "epoch": 2.431851019425269,
+      "grad_norm": 0.7072727084159851,
+      "learning_rate": 3.379029829494469e-05,
+      "loss": 0.2879,
+      "step": 7574
+    },
+    {
+      "epoch": 2.4321720982501205,
+      "grad_norm": 0.8465923070907593,
+      "learning_rate": 3.376414270136633e-05,
+      "loss": 0.3194,
+      "step": 7575
+    },
+    {
+      "epoch": 2.432493177074972,
+      "grad_norm": 0.7302094101905823,
+      "learning_rate": 3.373799517842627e-05,
+      "loss": 0.307,
+      "step": 7576
+    },
+    {
+      "epoch": 2.4328142558998236,
+      "grad_norm": 0.8451035022735596,
+      "learning_rate": 3.371185572931048e-05,
+      "loss": 0.3004,
+      "step": 7577
+    },
+    {
+      "epoch": 2.4331353347246747,
+      "grad_norm": 0.870238184928894,
+      "learning_rate": 3.3685724357204054e-05,
+      "loss": 0.2926,
+      "step": 7578
+    },
+    {
+      "epoch": 2.4334564135495262,
+      "grad_norm": 0.525023877620697,
+      "learning_rate": 3.3659601065290893e-05,
+      "loss": 0.282,
+      "step": 7579
+    },
+    {
+      "epoch": 2.433777492374378,
+      "grad_norm": 0.5372437834739685,
+      "learning_rate": 3.363348585675414e-05,
+      "loss": 0.5397,
+      "step": 7580
+    },
+    {
+      "epoch": 2.4340985711992293,
+      "grad_norm": 0.5941777229309082,
+      "learning_rate": 3.360737873477584e-05,
+      "loss": 0.6689,
+      "step": 7581
+    },
+    {
+      "epoch": 2.434419650024081,
+      "grad_norm": 0.35295045375823975,
+      "learning_rate": 3.358127970253704e-05,
+      "loss": 0.2129,
+      "step": 7582
+    },
+    {
+      "epoch": 2.4347407288489324,
+      "grad_norm": 0.43210405111312866,
+      "learning_rate": 3.355518876321787e-05,
+      "loss": 0.1644,
+      "step": 7583
+    },
+    {
+      "epoch": 2.435061807673784,
+      "grad_norm": 0.26534727215766907,
+      "learning_rate": 3.352910591999734e-05,
+      "loss": 0.076,
+      "step": 7584
+    },
+    {
+      "epoch": 2.4353828864986355,
+      "grad_norm": 0.1900699883699417,
+      "learning_rate": 3.3503031176053656e-05,
+      "loss": 0.072,
+      "step": 7585
+    },
+    {
+      "epoch": 2.435703965323487,
+      "grad_norm": 0.45870304107666016,
+      "learning_rate": 3.347696453456393e-05,
+      "loss": 0.1865,
+      "step": 7586
+    },
+    {
+      "epoch": 2.436025044148338,
+      "grad_norm": 0.7371842861175537,
+      "learning_rate": 3.3450905998704275e-05,
+      "loss": 0.341,
+      "step": 7587
+    },
+    {
+      "epoch": 2.4363461229731898,
+      "grad_norm": 0.7692381143569946,
+      "learning_rate": 3.342485557164986e-05,
+      "loss": 0.4095,
+      "step": 7588
+    },
+    {
+      "epoch": 2.4366672017980413,
+      "grad_norm": 0.7908117175102234,
+      "learning_rate": 3.339881325657484e-05,
+      "loss": 0.3606,
+      "step": 7589
+    },
+    {
+      "epoch": 2.436988280622893,
+      "grad_norm": 0.7690086364746094,
+      "learning_rate": 3.3372779056652426e-05,
+      "loss": 0.3458,
+      "step": 7590
+    },
+    {
+      "epoch": 2.4373093594477444,
+      "grad_norm": 0.8492259383201599,
+      "learning_rate": 3.334675297505476e-05,
+      "loss": 0.3107,
+      "step": 7591
+    },
+    {
+      "epoch": 2.437630438272596,
+      "grad_norm": 0.6811227202415466,
+      "learning_rate": 3.3320735014953076e-05,
+      "loss": 0.2832,
+      "step": 7592
+    },
+    {
+      "epoch": 2.4379515170974475,
+      "grad_norm": 0.8364322781562805,
+      "learning_rate": 3.3294725179517574e-05,
+      "loss": 0.322,
+      "step": 7593
+    },
+    {
+      "epoch": 2.438272595922299,
+      "grad_norm": 0.9228085279464722,
+      "learning_rate": 3.326872347191746e-05,
+      "loss": 0.4433,
+      "step": 7594
+    },
+    {
+      "epoch": 2.4385936747471506,
+      "grad_norm": 0.5413333773612976,
+      "learning_rate": 3.3242729895320946e-05,
+      "loss": 0.2079,
+      "step": 7595
+    },
+    {
+      "epoch": 2.4389147535720017,
+      "grad_norm": 0.7873682975769043,
+      "learning_rate": 3.3216744452895354e-05,
+      "loss": 0.3379,
+      "step": 7596
+    },
+    {
+      "epoch": 2.4392358323968533,
+      "grad_norm": 0.885543167591095,
+      "learning_rate": 3.319076714780682e-05,
+      "loss": 0.3296,
+      "step": 7597
+    },
+    {
+      "epoch": 2.439556911221705,
+      "grad_norm": 0.9024078845977783,
+      "learning_rate": 3.316479798322072e-05,
+      "loss": 0.3597,
+      "step": 7598
+    },
+    {
+      "epoch": 2.4398779900465564,
+      "grad_norm": 0.9780003428459167,
+      "learning_rate": 3.313883696230119e-05,
+      "loss": 0.4125,
+      "step": 7599
+    },
+    {
+      "epoch": 2.440199068871408,
+      "grad_norm": 2.2751457691192627,
+      "learning_rate": 3.311288408821159e-05,
+      "loss": 0.402,
+      "step": 7600
+    },
+    {
+      "epoch": 2.4405201476962595,
+      "grad_norm": 1.0487151145935059,
+      "learning_rate": 3.308693936411421e-05,
+      "loss": 0.4024,
+      "step": 7601
+    },
+    {
+      "epoch": 2.440841226521111,
+      "grad_norm": 1.0639290809631348,
+      "learning_rate": 3.306100279317024e-05,
+      "loss": 0.4246,
+      "step": 7602
+    },
+    {
+      "epoch": 2.4411623053459626,
+      "grad_norm": 0.9267032146453857,
+      "learning_rate": 3.303507437854009e-05,
+      "loss": 0.3029,
+      "step": 7603
+    },
+    {
+      "epoch": 2.441483384170814,
+      "grad_norm": 0.8686235547065735,
+      "learning_rate": 3.3009154123382936e-05,
+      "loss": 0.3351,
+      "step": 7604
+    },
+    {
+      "epoch": 2.4418044629956652,
+      "grad_norm": 0.9154815077781677,
+      "learning_rate": 3.2983242030857174e-05,
+      "loss": 0.2886,
+      "step": 7605
+    },
+    {
+      "epoch": 2.4421255418205168,
+      "grad_norm": 0.6667101383209229,
+      "learning_rate": 3.2957338104120096e-05,
+      "loss": 0.312,
+      "step": 7606
+    },
+    {
+      "epoch": 2.4424466206453683,
+      "grad_norm": 0.8355054259300232,
+      "learning_rate": 3.2931442346328004e-05,
+      "loss": 0.384,
+      "step": 7607
+    },
+    {
+      "epoch": 2.44276769947022,
+      "grad_norm": 0.8191713094711304,
+      "learning_rate": 3.290555476063622e-05,
+      "loss": 0.3421,
+      "step": 7608
+    },
+    {
+      "epoch": 2.4430887782950714,
+      "grad_norm": 0.9118049740791321,
+      "learning_rate": 3.287967535019908e-05,
+      "loss": 0.3099,
+      "step": 7609
+    },
+    {
+      "epoch": 2.443409857119923,
+      "grad_norm": 1.0478988885879517,
+      "learning_rate": 3.285380411816988e-05,
+      "loss": 0.4101,
+      "step": 7610
+    },
+    {
+      "epoch": 2.4437309359447745,
+      "grad_norm": 1.4113534688949585,
+      "learning_rate": 3.2827941067700996e-05,
+      "loss": 0.3687,
+      "step": 7611
+    },
+    {
+      "epoch": 2.444052014769626,
+      "grad_norm": 0.7927390336990356,
+      "learning_rate": 3.2802086201943724e-05,
+      "loss": 0.2911,
+      "step": 7612
+    },
+    {
+      "epoch": 2.4443730935944776,
+      "grad_norm": 0.8057661056518555,
+      "learning_rate": 3.277623952404842e-05,
+      "loss": 0.3085,
+      "step": 7613
+    },
+    {
+      "epoch": 2.4446941724193287,
+      "grad_norm": 0.8335930705070496,
+      "learning_rate": 3.275040103716441e-05,
+      "loss": 0.3725,
+      "step": 7614
+    },
+    {
+      "epoch": 2.4450152512441803,
+      "grad_norm": 0.7692309617996216,
+      "learning_rate": 3.272457074444003e-05,
+      "loss": 0.3191,
+      "step": 7615
+    },
+    {
+      "epoch": 2.445336330069032,
+      "grad_norm": 0.9450210928916931,
+      "learning_rate": 3.269874864902269e-05,
+      "loss": 0.3662,
+      "step": 7616
+    },
+    {
+      "epoch": 2.4456574088938834,
+      "grad_norm": 0.6600220203399658,
+      "learning_rate": 3.2672934754058616e-05,
+      "loss": 0.3059,
+      "step": 7617
+    },
+    {
+      "epoch": 2.445978487718735,
+      "grad_norm": 0.8459724187850952,
+      "learning_rate": 3.264712906269328e-05,
+      "loss": 0.3184,
+      "step": 7618
+    },
+    {
+      "epoch": 2.4462995665435865,
+      "grad_norm": 1.1846095323562622,
+      "learning_rate": 3.2621331578070934e-05,
+      "loss": 0.2196,
+      "step": 7619
+    },
+    {
+      "epoch": 2.446620645368438,
+      "grad_norm": 0.5867680311203003,
+      "learning_rate": 3.2595542303334924e-05,
+      "loss": 0.2611,
+      "step": 7620
+    },
+    {
+      "epoch": 2.4469417241932896,
+      "grad_norm": 0.8343913555145264,
+      "learning_rate": 3.2569761241627696e-05,
+      "loss": 0.3502,
+      "step": 7621
+    },
+    {
+      "epoch": 2.447262803018141,
+      "grad_norm": 0.803591251373291,
+      "learning_rate": 3.254398839609044e-05,
+      "loss": 0.2839,
+      "step": 7622
+    },
+    {
+      "epoch": 2.4475838818429922,
+      "grad_norm": 0.7186933755874634,
+      "learning_rate": 3.251822376986363e-05,
+      "loss": 0.3149,
+      "step": 7623
+    },
+    {
+      "epoch": 2.447904960667844,
+      "grad_norm": 0.8497990369796753,
+      "learning_rate": 3.249246736608655e-05,
+      "loss": 0.3579,
+      "step": 7624
+    },
+    {
+      "epoch": 2.4482260394926953,
+      "grad_norm": 0.5227159857749939,
+      "learning_rate": 3.246671918789755e-05,
+      "loss": 0.2571,
+      "step": 7625
+    },
+    {
+      "epoch": 2.448547118317547,
+      "grad_norm": 0.8302345275878906,
+      "learning_rate": 3.244097923843398e-05,
+      "loss": 0.3503,
+      "step": 7626
+    },
+    {
+      "epoch": 2.4488681971423985,
+      "grad_norm": 0.5355742573738098,
+      "learning_rate": 3.2415247520832146e-05,
+      "loss": 0.2695,
+      "step": 7627
+    },
+    {
+      "epoch": 2.44918927596725,
+      "grad_norm": 0.7447862029075623,
+      "learning_rate": 3.23895240382274e-05,
+      "loss": 0.2746,
+      "step": 7628
+    },
+    {
+      "epoch": 2.4495103547921016,
+      "grad_norm": 0.46801793575286865,
+      "learning_rate": 3.236380879375408e-05,
+      "loss": 0.3045,
+      "step": 7629
+    },
+    {
+      "epoch": 2.449831433616953,
+      "grad_norm": 0.5388785004615784,
+      "learning_rate": 3.233810179054548e-05,
+      "loss": 0.9426,
+      "step": 7630
+    },
+    {
+      "epoch": 2.4501525124418047,
+      "grad_norm": 0.4183870553970337,
+      "learning_rate": 3.231240303173394e-05,
+      "loss": 0.399,
+      "step": 7631
+    },
+    {
+      "epoch": 2.4504735912666558,
+      "grad_norm": 0.9477866888046265,
+      "learning_rate": 3.2286712520450765e-05,
+      "loss": 0.2928,
+      "step": 7632
+    },
+    {
+      "epoch": 2.4507946700915073,
+      "grad_norm": 0.34258604049682617,
+      "learning_rate": 3.226103025982628e-05,
+      "loss": 0.0871,
+      "step": 7633
+    },
+    {
+      "epoch": 2.451115748916359,
+      "grad_norm": 0.6076464056968689,
+      "learning_rate": 3.223535625298979e-05,
+      "loss": 0.2203,
+      "step": 7634
+    },
+    {
+      "epoch": 2.4514368277412104,
+      "grad_norm": 0.3136209547519684,
+      "learning_rate": 3.220969050306955e-05,
+      "loss": 0.1264,
+      "step": 7635
+    },
+    {
+      "epoch": 2.451757906566062,
+      "grad_norm": 0.8506674766540527,
+      "learning_rate": 3.218403301319296e-05,
+      "loss": 0.4329,
+      "step": 7636
+    },
+    {
+      "epoch": 2.4520789853909135,
+      "grad_norm": 0.8831260800361633,
+      "learning_rate": 3.21583837864862e-05,
+      "loss": 0.4926,
+      "step": 7637
+    },
+    {
+      "epoch": 2.452400064215765,
+      "grad_norm": 0.8367322683334351,
+      "learning_rate": 3.213274282607457e-05,
+      "loss": 0.3901,
+      "step": 7638
+    },
+    {
+      "epoch": 2.4527211430406166,
+      "grad_norm": 0.824145495891571,
+      "learning_rate": 3.210711013508242e-05,
+      "loss": 0.3007,
+      "step": 7639
+    },
+    {
+      "epoch": 2.453042221865468,
+      "grad_norm": 0.8876867890357971,
+      "learning_rate": 3.208148571663289e-05,
+      "loss": 0.3398,
+      "step": 7640
+    },
+    {
+      "epoch": 2.4533633006903193,
+      "grad_norm": 0.6797581315040588,
+      "learning_rate": 3.205586957384838e-05,
+      "loss": 0.2937,
+      "step": 7641
+    },
+    {
+      "epoch": 2.453684379515171,
+      "grad_norm": 0.7675800919532776,
+      "learning_rate": 3.2030261709849996e-05,
+      "loss": 0.3564,
+      "step": 7642
+    },
+    {
+      "epoch": 2.4540054583400224,
+      "grad_norm": 0.7745330929756165,
+      "learning_rate": 3.200466212775808e-05,
+      "loss": 0.3317,
+      "step": 7643
+    },
+    {
+      "epoch": 2.454326537164874,
+      "grad_norm": 0.7452265024185181,
+      "learning_rate": 3.197907083069184e-05,
+      "loss": 0.3462,
+      "step": 7644
+    },
+    {
+      "epoch": 2.4546476159897255,
+      "grad_norm": 0.8958961963653564,
+      "learning_rate": 3.195348782176948e-05,
+      "loss": 0.3256,
+      "step": 7645
+    },
+    {
+      "epoch": 2.454968694814577,
+      "grad_norm": 0.9540597200393677,
+      "learning_rate": 3.192791310410822e-05,
+      "loss": 0.377,
+      "step": 7646
+    },
+    {
+      "epoch": 2.4552897736394286,
+      "grad_norm": 1.0042481422424316,
+      "learning_rate": 3.190234668082427e-05,
+      "loss": 0.3533,
+      "step": 7647
+    },
+    {
+      "epoch": 2.45561085246428,
+      "grad_norm": 1.0041635036468506,
+      "learning_rate": 3.187678855503282e-05,
+      "loss": 0.3521,
+      "step": 7648
+    },
+    {
+      "epoch": 2.4559319312891317,
+      "grad_norm": 0.8743497133255005,
+      "learning_rate": 3.1851238729848034e-05,
+      "loss": 0.333,
+      "step": 7649
+    },
+    {
+      "epoch": 2.456253010113983,
+      "grad_norm": 0.8867911696434021,
+      "learning_rate": 3.1825697208383096e-05,
+      "loss": 0.3217,
+      "step": 7650
+    },
+    {
+      "epoch": 2.4565740889388343,
+      "grad_norm": 0.6222565770149231,
+      "learning_rate": 3.1800163993750166e-05,
+      "loss": 0.2614,
+      "step": 7651
+    },
+    {
+      "epoch": 2.456895167763686,
+      "grad_norm": 1.0342620611190796,
+      "learning_rate": 3.1774639089060363e-05,
+      "loss": 0.3954,
+      "step": 7652
+    },
+    {
+      "epoch": 2.4572162465885374,
+      "grad_norm": 0.7965120077133179,
+      "learning_rate": 3.174912249742382e-05,
+      "loss": 0.2804,
+      "step": 7653
+    },
+    {
+      "epoch": 2.457537325413389,
+      "grad_norm": 0.7983464002609253,
+      "learning_rate": 3.172361422194974e-05,
+      "loss": 0.2956,
+      "step": 7654
+    },
+    {
+      "epoch": 2.4578584042382405,
+      "grad_norm": 0.8190217018127441,
+      "learning_rate": 3.1698114265746124e-05,
+      "loss": 0.3554,
+      "step": 7655
+    },
+    {
+      "epoch": 2.458179483063092,
+      "grad_norm": 0.7282566428184509,
+      "learning_rate": 3.16726226319201e-05,
+      "loss": 0.2899,
+      "step": 7656
+    },
+    {
+      "epoch": 2.4585005618879436,
+      "grad_norm": 0.8753344416618347,
+      "learning_rate": 3.164713932357776e-05,
+      "loss": 0.347,
+      "step": 7657
+    },
+    {
+      "epoch": 2.458821640712795,
+      "grad_norm": 0.8597754836082458,
+      "learning_rate": 3.162166434382412e-05,
+      "loss": 0.2803,
+      "step": 7658
+    },
+    {
+      "epoch": 2.4591427195376463,
+      "grad_norm": 0.7598649263381958,
+      "learning_rate": 3.159619769576333e-05,
+      "loss": 0.2793,
+      "step": 7659
+    },
+    {
+      "epoch": 2.459463798362498,
+      "grad_norm": 1.1359360218048096,
+      "learning_rate": 3.157073938249829e-05,
+      "loss": 0.3773,
+      "step": 7660
+    },
+    {
+      "epoch": 2.4597848771873494,
+      "grad_norm": 1.1523734331130981,
+      "learning_rate": 3.154528940713113e-05,
+      "loss": 0.3819,
+      "step": 7661
+    },
+    {
+      "epoch": 2.460105956012201,
+      "grad_norm": 0.9038000106811523,
+      "learning_rate": 3.15198477727628e-05,
+      "loss": 0.3285,
+      "step": 7662
+    },
+    {
+      "epoch": 2.4604270348370525,
+      "grad_norm": 0.7829211354255676,
+      "learning_rate": 3.1494414482493304e-05,
+      "loss": 0.3033,
+      "step": 7663
+    },
+    {
+      "epoch": 2.460748113661904,
+      "grad_norm": 0.8830636739730835,
+      "learning_rate": 3.146898953942163e-05,
+      "loss": 0.3084,
+      "step": 7664
+    },
+    {
+      "epoch": 2.4610691924867556,
+      "grad_norm": 0.527309775352478,
+      "learning_rate": 3.144357294664565e-05,
+      "loss": 0.2242,
+      "step": 7665
+    },
+    {
+      "epoch": 2.461390271311607,
+      "grad_norm": 1.0569745302200317,
+      "learning_rate": 3.141816470726238e-05,
+      "loss": 0.3431,
+      "step": 7666
+    },
+    {
+      "epoch": 2.4617113501364587,
+      "grad_norm": 0.7063748240470886,
+      "learning_rate": 3.1392764824367704e-05,
+      "loss": 0.2735,
+      "step": 7667
+    },
+    {
+      "epoch": 2.46203242896131,
+      "grad_norm": 0.5551251769065857,
+      "learning_rate": 3.1367373301056536e-05,
+      "loss": 0.2571,
+      "step": 7668
+    },
+    {
+      "epoch": 2.4623535077861614,
+      "grad_norm": 0.8918856978416443,
+      "learning_rate": 3.134199014042274e-05,
+      "loss": 0.305,
+      "step": 7669
+    },
+    {
+      "epoch": 2.462674586611013,
+      "grad_norm": 0.7437170743942261,
+      "learning_rate": 3.1316615345559185e-05,
+      "loss": 0.2788,
+      "step": 7670
+    },
+    {
+      "epoch": 2.4629956654358645,
+      "grad_norm": 1.070360779762268,
+      "learning_rate": 3.129124891955771e-05,
+      "loss": 0.4669,
+      "step": 7671
+    },
+    {
+      "epoch": 2.463316744260716,
+      "grad_norm": 0.8715602159500122,
+      "learning_rate": 3.126589086550914e-05,
+      "loss": 0.3444,
+      "step": 7672
+    },
+    {
+      "epoch": 2.4636378230855676,
+      "grad_norm": 0.7617392539978027,
+      "learning_rate": 3.124054118650327e-05,
+      "loss": 0.2923,
+      "step": 7673
+    },
+    {
+      "epoch": 2.463958901910419,
+      "grad_norm": 0.5233041644096375,
+      "learning_rate": 3.12151998856289e-05,
+      "loss": 0.2442,
+      "step": 7674
+    },
+    {
+      "epoch": 2.4642799807352707,
+      "grad_norm": 0.776607871055603,
+      "learning_rate": 3.1189866965973766e-05,
+      "loss": 0.3125,
+      "step": 7675
+    },
+    {
+      "epoch": 2.464601059560122,
+      "grad_norm": 0.5157756209373474,
+      "learning_rate": 3.116454243062459e-05,
+      "loss": 0.2563,
+      "step": 7676
+    },
+    {
+      "epoch": 2.4649221383849733,
+      "grad_norm": 0.7321842312812805,
+      "learning_rate": 3.113922628266718e-05,
+      "loss": 0.3024,
+      "step": 7677
+    },
+    {
+      "epoch": 2.465243217209825,
+      "grad_norm": 0.6504801511764526,
+      "learning_rate": 3.111391852518611e-05,
+      "loss": 0.3101,
+      "step": 7678
+    },
+    {
+      "epoch": 2.4655642960346764,
+      "grad_norm": 0.39098212122917175,
+      "learning_rate": 3.108861916126518e-05,
+      "loss": 0.2716,
+      "step": 7679
+    },
+    {
+      "epoch": 2.465885374859528,
+      "grad_norm": 0.5527970194816589,
+      "learning_rate": 3.1063328193986904e-05,
+      "loss": 0.8426,
+      "step": 7680
+    },
+    {
+      "epoch": 2.4662064536843795,
+      "grad_norm": 0.3802039623260498,
+      "learning_rate": 3.103804562643302e-05,
+      "loss": 0.4868,
+      "step": 7681
+    },
+    {
+      "epoch": 2.466527532509231,
+      "grad_norm": 0.40410691499710083,
+      "learning_rate": 3.101277146168412e-05,
+      "loss": 0.2456,
+      "step": 7682
+    },
+    {
+      "epoch": 2.4668486113340826,
+      "grad_norm": 0.2699540853500366,
+      "learning_rate": 3.098750570281969e-05,
+      "loss": 0.0755,
+      "step": 7683
+    },
+    {
+      "epoch": 2.467169690158934,
+      "grad_norm": 0.47554266452789307,
+      "learning_rate": 3.096224835291839e-05,
+      "loss": 0.1443,
+      "step": 7684
+    },
+    {
+      "epoch": 2.4674907689837857,
+      "grad_norm": 0.6371931433677673,
+      "learning_rate": 3.093699941505771e-05,
+      "loss": 0.2333,
+      "step": 7685
+    },
+    {
+      "epoch": 2.467811847808637,
+      "grad_norm": 0.8317315578460693,
+      "learning_rate": 3.0911758892314166e-05,
+      "loss": 0.4051,
+      "step": 7686
+    },
+    {
+      "epoch": 2.4681329266334884,
+      "grad_norm": 0.8877337574958801,
+      "learning_rate": 3.0886526787763234e-05,
+      "loss": 0.4092,
+      "step": 7687
+    },
+    {
+      "epoch": 2.46845400545834,
+      "grad_norm": 0.8633880019187927,
+      "learning_rate": 3.086130310447937e-05,
+      "loss": 0.4231,
+      "step": 7688
+    },
+    {
+      "epoch": 2.4687750842831915,
+      "grad_norm": 0.7211251854896545,
+      "learning_rate": 3.0836087845536e-05,
+      "loss": 0.2923,
+      "step": 7689
+    },
+    {
+      "epoch": 2.469096163108043,
+      "grad_norm": 0.9071685671806335,
+      "learning_rate": 3.081088101400552e-05,
+      "loss": 0.3378,
+      "step": 7690
+    },
+    {
+      "epoch": 2.4694172419328946,
+      "grad_norm": 0.9691846370697021,
+      "learning_rate": 3.078568261295933e-05,
+      "loss": 0.3977,
+      "step": 7691
+    },
+    {
+      "epoch": 2.469738320757746,
+      "grad_norm": 0.9273397326469421,
+      "learning_rate": 3.0760492645467765e-05,
+      "loss": 0.3731,
+      "step": 7692
+    },
+    {
+      "epoch": 2.4700593995825977,
+      "grad_norm": 0.8951248526573181,
+      "learning_rate": 3.073531111460013e-05,
+      "loss": 0.4538,
+      "step": 7693
+    },
+    {
+      "epoch": 2.4703804784074492,
+      "grad_norm": 0.868762731552124,
+      "learning_rate": 3.071013802342475e-05,
+      "loss": 0.374,
+      "step": 7694
+    },
+    {
+      "epoch": 2.4707015572323003,
+      "grad_norm": 0.614288330078125,
+      "learning_rate": 3.068497337500886e-05,
+      "loss": 0.2634,
+      "step": 7695
+    },
+    {
+      "epoch": 2.471022636057152,
+      "grad_norm": 1.016533374786377,
+      "learning_rate": 3.0659817172418693e-05,
+      "loss": 0.3648,
+      "step": 7696
+    },
+    {
+      "epoch": 2.4713437148820034,
+      "grad_norm": 1.118348240852356,
+      "learning_rate": 3.063466941871952e-05,
+      "loss": 0.4014,
+      "step": 7697
+    },
+    {
+      "epoch": 2.471664793706855,
+      "grad_norm": 0.9089675545692444,
+      "learning_rate": 3.060953011697545e-05,
+      "loss": 0.3507,
+      "step": 7698
+    },
+    {
+      "epoch": 2.4719858725317065,
+      "grad_norm": 0.9801891446113586,
+      "learning_rate": 3.058439927024962e-05,
+      "loss": 0.388,
+      "step": 7699
+    },
+    {
+      "epoch": 2.472306951356558,
+      "grad_norm": 1.0153359174728394,
+      "learning_rate": 3.0559276881604236e-05,
+      "loss": 0.4152,
+      "step": 7700
+    },
+    {
+      "epoch": 2.4726280301814096,
+      "grad_norm": 0.6754663586616516,
+      "learning_rate": 3.053416295410026e-05,
+      "loss": 0.279,
+      "step": 7701
+    },
+    {
+      "epoch": 2.472949109006261,
+      "grad_norm": 0.8051523566246033,
+      "learning_rate": 3.0509057490797888e-05,
+      "loss": 0.3163,
+      "step": 7702
+    },
+    {
+      "epoch": 2.4732701878311127,
+      "grad_norm": 0.8289381861686707,
+      "learning_rate": 3.0483960494756016e-05,
+      "loss": 0.281,
+      "step": 7703
+    },
+    {
+      "epoch": 2.473591266655964,
+      "grad_norm": 0.7152822613716125,
+      "learning_rate": 3.045887196903271e-05,
+      "loss": 0.3059,
+      "step": 7704
+    },
+    {
+      "epoch": 2.4739123454808154,
+      "grad_norm": 1.1248416900634766,
+      "learning_rate": 3.0433791916684916e-05,
+      "loss": 0.4753,
+      "step": 7705
+    },
+    {
+      "epoch": 2.474233424305667,
+      "grad_norm": 1.4356248378753662,
+      "learning_rate": 3.0408720340768572e-05,
+      "loss": 0.5532,
+      "step": 7706
+    },
+    {
+      "epoch": 2.4745545031305185,
+      "grad_norm": 0.6347241401672363,
+      "learning_rate": 3.038365724433858e-05,
+      "loss": 0.2673,
+      "step": 7707
+    },
+    {
+      "epoch": 2.47487558195537,
+      "grad_norm": 0.8377917408943176,
+      "learning_rate": 3.035860263044873e-05,
+      "loss": 0.3614,
+      "step": 7708
+    },
+    {
+      "epoch": 2.4751966607802216,
+      "grad_norm": 0.866231381893158,
+      "learning_rate": 3.0333556502151926e-05,
+      "loss": 0.318,
+      "step": 7709
+    },
+    {
+      "epoch": 2.475517739605073,
+      "grad_norm": 0.899551272392273,
+      "learning_rate": 3.0308518862499957e-05,
+      "loss": 0.4259,
+      "step": 7710
+    },
+    {
+      "epoch": 2.4758388184299247,
+      "grad_norm": 0.8847672343254089,
+      "learning_rate": 3.0283489714543556e-05,
+      "loss": 0.3208,
+      "step": 7711
+    },
+    {
+      "epoch": 2.4761598972547763,
+      "grad_norm": 1.1146432161331177,
+      "learning_rate": 3.0258469061332463e-05,
+      "loss": 0.4248,
+      "step": 7712
+    },
+    {
+      "epoch": 2.4764809760796274,
+      "grad_norm": 0.6833528876304626,
+      "learning_rate": 3.023345690591537e-05,
+      "loss": 0.2858,
+      "step": 7713
+    },
+    {
+      "epoch": 2.476802054904479,
+      "grad_norm": 0.6902426481246948,
+      "learning_rate": 3.0208453251339885e-05,
+      "loss": 0.3027,
+      "step": 7714
+    },
+    {
+      "epoch": 2.4771231337293305,
+      "grad_norm": 0.7399451732635498,
+      "learning_rate": 3.018345810065275e-05,
+      "loss": 0.2832,
+      "step": 7715
+    },
+    {
+      "epoch": 2.477444212554182,
+      "grad_norm": 0.6635550856590271,
+      "learning_rate": 3.0158471456899428e-05,
+      "loss": 0.2934,
+      "step": 7716
+    },
+    {
+      "epoch": 2.4777652913790336,
+      "grad_norm": 0.7805300354957581,
+      "learning_rate": 3.0133493323124505e-05,
+      "loss": 0.3349,
+      "step": 7717
+    },
+    {
+      "epoch": 2.478086370203885,
+      "grad_norm": 1.5365228652954102,
+      "learning_rate": 3.0108523702371505e-05,
+      "loss": 0.3909,
+      "step": 7718
+    },
+    {
+      "epoch": 2.4784074490287367,
+      "grad_norm": 0.8297216892242432,
+      "learning_rate": 3.0083562597682847e-05,
+      "loss": 0.2861,
+      "step": 7719
+    },
+    {
+      "epoch": 2.478728527853588,
+      "grad_norm": 0.5565605163574219,
+      "learning_rate": 3.0058610012100074e-05,
+      "loss": 0.2976,
+      "step": 7720
+    },
+    {
+      "epoch": 2.4790496066784398,
+      "grad_norm": 1.5360130071640015,
+      "learning_rate": 3.0033665948663448e-05,
+      "loss": 0.3868,
+      "step": 7721
+    },
+    {
+      "epoch": 2.479370685503291,
+      "grad_norm": 0.8052739500999451,
+      "learning_rate": 3.0008730410412466e-05,
+      "loss": 0.3711,
+      "step": 7722
+    },
+    {
+      "epoch": 2.4796917643281424,
+      "grad_norm": 0.7085245847702026,
+      "learning_rate": 2.9983803400385312e-05,
+      "loss": 0.3142,
+      "step": 7723
+    },
+    {
+      "epoch": 2.480012843152994,
+      "grad_norm": 0.939500629901886,
+      "learning_rate": 2.9958884921619367e-05,
+      "loss": 0.3342,
+      "step": 7724
+    },
+    {
+      "epoch": 2.4803339219778455,
+      "grad_norm": 0.8791356682777405,
+      "learning_rate": 2.993397497715086e-05,
+      "loss": 0.2739,
+      "step": 7725
+    },
+    {
+      "epoch": 2.480655000802697,
+      "grad_norm": 0.8108116984367371,
+      "learning_rate": 2.9909073570014912e-05,
+      "loss": 0.301,
+      "step": 7726
+    },
+    {
+      "epoch": 2.4809760796275486,
+      "grad_norm": 0.6944962739944458,
+      "learning_rate": 2.9884180703245767e-05,
+      "loss": 0.3222,
+      "step": 7727
+    },
+    {
+      "epoch": 2.4812971584524,
+      "grad_norm": 0.3254513144493103,
+      "learning_rate": 2.9859296379876523e-05,
+      "loss": 0.2407,
+      "step": 7728
+    },
+    {
+      "epoch": 2.4816182372772517,
+      "grad_norm": 0.7786895632743835,
+      "learning_rate": 2.9834420602939263e-05,
+      "loss": 0.3052,
+      "step": 7729
+    },
+    {
+      "epoch": 2.4819393161021033,
+      "grad_norm": 0.7794302701950073,
+      "learning_rate": 2.9809553375465004e-05,
+      "loss": 0.7187,
+      "step": 7730
+    },
+    {
+      "epoch": 2.4822603949269544,
+      "grad_norm": 0.33989983797073364,
+      "learning_rate": 2.9784694700483762e-05,
+      "loss": 0.2142,
+      "step": 7731
+    },
+    {
+      "epoch": 2.482581473751806,
+      "grad_norm": 0.5084500908851624,
+      "learning_rate": 2.9759844581024486e-05,
+      "loss": 0.2271,
+      "step": 7732
+    },
+    {
+      "epoch": 2.4829025525766575,
+      "grad_norm": 0.4874570667743683,
+      "learning_rate": 2.9735003020115092e-05,
+      "loss": 0.0856,
+      "step": 7733
+    },
+    {
+      "epoch": 2.483223631401509,
+      "grad_norm": 0.3979865610599518,
+      "learning_rate": 2.9710170020782435e-05,
+      "loss": 0.188,
+      "step": 7734
+    },
+    {
+      "epoch": 2.4835447102263606,
+      "grad_norm": 0.5099960565567017,
+      "learning_rate": 2.968534558605236e-05,
+      "loss": 0.3362,
+      "step": 7735
+    },
+    {
+      "epoch": 2.483865789051212,
+      "grad_norm": 0.5169116854667664,
+      "learning_rate": 2.9660529718949627e-05,
+      "loss": 0.3097,
+      "step": 7736
+    },
+    {
+      "epoch": 2.4841868678760637,
+      "grad_norm": 0.9039494395256042,
+      "learning_rate": 2.9635722422497993e-05,
+      "loss": 0.4115,
+      "step": 7737
+    },
+    {
+      "epoch": 2.4845079467009152,
+      "grad_norm": 0.8627937436103821,
+      "learning_rate": 2.961092369972014e-05,
+      "loss": 0.4387,
+      "step": 7738
+    },
+    {
+      "epoch": 2.484829025525767,
+      "grad_norm": 0.8400534391403198,
+      "learning_rate": 2.9586133553637683e-05,
+      "loss": 0.4016,
+      "step": 7739
+    },
+    {
+      "epoch": 2.485150104350618,
+      "grad_norm": 0.7134156823158264,
+      "learning_rate": 2.9561351987271334e-05,
+      "loss": 0.3185,
+      "step": 7740
+    },
+    {
+      "epoch": 2.4854711831754694,
+      "grad_norm": 0.8354820609092712,
+      "learning_rate": 2.953657900364053e-05,
+      "loss": 0.4111,
+      "step": 7741
+    },
+    {
+      "epoch": 2.485792262000321,
+      "grad_norm": 0.66590815782547,
+      "learning_rate": 2.9511814605763855e-05,
+      "loss": 0.2975,
+      "step": 7742
+    },
+    {
+      "epoch": 2.4861133408251725,
+      "grad_norm": 0.7249523997306824,
+      "learning_rate": 2.9487058796658783e-05,
+      "loss": 0.3311,
+      "step": 7743
+    },
+    {
+      "epoch": 2.486434419650024,
+      "grad_norm": 0.6020844578742981,
+      "learning_rate": 2.9462311579341663e-05,
+      "loss": 0.252,
+      "step": 7744
+    },
+    {
+      "epoch": 2.4867554984748756,
+      "grad_norm": 0.8182882070541382,
+      "learning_rate": 2.9437572956827964e-05,
+      "loss": 0.3021,
+      "step": 7745
+    },
+    {
+      "epoch": 2.487076577299727,
+      "grad_norm": 0.6254271268844604,
+      "learning_rate": 2.94128429321319e-05,
+      "loss": 0.2792,
+      "step": 7746
+    },
+    {
+      "epoch": 2.4873976561245787,
+      "grad_norm": 0.83849036693573,
+      "learning_rate": 2.938812150826684e-05,
+      "loss": 0.3111,
+      "step": 7747
+    },
+    {
+      "epoch": 2.4877187349494303,
+      "grad_norm": 1.0279197692871094,
+      "learning_rate": 2.9363408688245e-05,
+      "loss": 0.3787,
+      "step": 7748
+    },
+    {
+      "epoch": 2.4880398137742814,
+      "grad_norm": 0.6685528755187988,
+      "learning_rate": 2.933870447507753e-05,
+      "loss": 0.3001,
+      "step": 7749
+    },
+    {
+      "epoch": 2.488360892599133,
+      "grad_norm": 0.9342240691184998,
+      "learning_rate": 2.931400887177459e-05,
+      "loss": 0.3955,
+      "step": 7750
+    },
+    {
+      "epoch": 2.4886819714239845,
+      "grad_norm": 0.6949548125267029,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 0.3251,
+      "step": 7751
+    },
+    {
+      "epoch": 2.489003050248836,
+      "grad_norm": 1.0305014848709106,
+      "learning_rate": 2.926464350679756e-05,
+      "loss": 0.3924,
+      "step": 7752
+    },
+    {
+      "epoch": 2.4893241290736876,
+      "grad_norm": 0.8009200692176819,
+      "learning_rate": 2.9239973751138495e-05,
+      "loss": 0.3076,
+      "step": 7753
+    },
+    {
+      "epoch": 2.489645207898539,
+      "grad_norm": 0.944476306438446,
+      "learning_rate": 2.921531261737398e-05,
+      "loss": 0.3573,
+      "step": 7754
+    },
+    {
+      "epoch": 2.4899662867233907,
+      "grad_norm": 1.1194026470184326,
+      "learning_rate": 2.9190660108508917e-05,
+      "loss": 0.419,
+      "step": 7755
+    },
+    {
+      "epoch": 2.4902873655482423,
+      "grad_norm": 0.9375110864639282,
+      "learning_rate": 2.9166016227547133e-05,
+      "loss": 0.3467,
+      "step": 7756
+    },
+    {
+      "epoch": 2.490608444373094,
+      "grad_norm": 0.7370654344558716,
+      "learning_rate": 2.9141380977491373e-05,
+      "loss": 0.3359,
+      "step": 7757
+    },
+    {
+      "epoch": 2.490929523197945,
+      "grad_norm": 0.9551532864570618,
+      "learning_rate": 2.911675436134347e-05,
+      "loss": 0.3477,
+      "step": 7758
+    },
+    {
+      "epoch": 2.4912506020227965,
+      "grad_norm": 0.8018618226051331,
+      "learning_rate": 2.9092136382103973e-05,
+      "loss": 0.2881,
+      "step": 7759
+    },
+    {
+      "epoch": 2.491571680847648,
+      "grad_norm": 0.8496661186218262,
+      "learning_rate": 2.9067527042772636e-05,
+      "loss": 0.3263,
+      "step": 7760
+    },
+    {
+      "epoch": 2.4918927596724996,
+      "grad_norm": 0.5800387859344482,
+      "learning_rate": 2.904292634634793e-05,
+      "loss": 0.2534,
+      "step": 7761
+    },
+    {
+      "epoch": 2.492213838497351,
+      "grad_norm": 0.9090917706489563,
+      "learning_rate": 2.9018334295827388e-05,
+      "loss": 0.3376,
+      "step": 7762
+    },
+    {
+      "epoch": 2.4925349173222027,
+      "grad_norm": 0.84961998462677,
+      "learning_rate": 2.899375089420756e-05,
+      "loss": 0.306,
+      "step": 7763
+    },
+    {
+      "epoch": 2.492855996147054,
+      "grad_norm": 0.7612555623054504,
+      "learning_rate": 2.8969176144483744e-05,
+      "loss": 0.2584,
+      "step": 7764
+    },
+    {
+      "epoch": 2.4931770749719058,
+      "grad_norm": 0.7672126293182373,
+      "learning_rate": 2.894461004965038e-05,
+      "loss": 0.2945,
+      "step": 7765
+    },
+    {
+      "epoch": 2.4934981537967573,
+      "grad_norm": 1.0515177249908447,
+      "learning_rate": 2.8920052612700754e-05,
+      "loss": 0.3602,
+      "step": 7766
+    },
+    {
+      "epoch": 2.4938192326216084,
+      "grad_norm": 0.710355818271637,
+      "learning_rate": 2.8895503836627103e-05,
+      "loss": 0.2777,
+      "step": 7767
+    },
+    {
+      "epoch": 2.49414031144646,
+      "grad_norm": 0.9773350954055786,
+      "learning_rate": 2.887096372442063e-05,
+      "loss": 0.3712,
+      "step": 7768
+    },
+    {
+      "epoch": 2.4944613902713115,
+      "grad_norm": 0.9016016125679016,
+      "learning_rate": 2.8846432279071467e-05,
+      "loss": 0.3312,
+      "step": 7769
+    },
+    {
+      "epoch": 2.494782469096163,
+      "grad_norm": 0.8454534411430359,
+      "learning_rate": 2.88219095035687e-05,
+      "loss": 0.3164,
+      "step": 7770
+    },
+    {
+      "epoch": 2.4951035479210146,
+      "grad_norm": 0.6082969307899475,
+      "learning_rate": 2.879739540090036e-05,
+      "loss": 0.2435,
+      "step": 7771
+    },
+    {
+      "epoch": 2.495424626745866,
+      "grad_norm": 1.3337501287460327,
+      "learning_rate": 2.877288997405341e-05,
+      "loss": 0.315,
+      "step": 7772
+    },
+    {
+      "epoch": 2.4957457055707177,
+      "grad_norm": 0.6798154711723328,
+      "learning_rate": 2.874839322601375e-05,
+      "loss": 0.2806,
+      "step": 7773
+    },
+    {
+      "epoch": 2.4960667843955693,
+      "grad_norm": 1.1664059162139893,
+      "learning_rate": 2.872390515976625e-05,
+      "loss": 0.3415,
+      "step": 7774
+    },
+    {
+      "epoch": 2.496387863220421,
+      "grad_norm": 0.40442001819610596,
+      "learning_rate": 2.869942577829471e-05,
+      "loss": 0.2215,
+      "step": 7775
+    },
+    {
+      "epoch": 2.496708942045272,
+      "grad_norm": 0.6790067553520203,
+      "learning_rate": 2.8674955084581857e-05,
+      "loss": 0.2721,
+      "step": 7776
+    },
+    {
+      "epoch": 2.4970300208701235,
+      "grad_norm": 0.6047253608703613,
+      "learning_rate": 2.865049308160934e-05,
+      "loss": 0.2712,
+      "step": 7777
+    },
+    {
+      "epoch": 2.497351099694975,
+      "grad_norm": 0.2929036617279053,
+      "learning_rate": 2.8626039772357882e-05,
+      "loss": 0.2281,
+      "step": 7778
+    },
+    {
+      "epoch": 2.4976721785198266,
+      "grad_norm": 0.517440140247345,
+      "learning_rate": 2.860159515980695e-05,
+      "loss": 0.2949,
+      "step": 7779
+    },
+    {
+      "epoch": 2.497993257344678,
+      "grad_norm": 0.5559537410736084,
+      "learning_rate": 2.8577159246935037e-05,
+      "loss": 0.6561,
+      "step": 7780
+    },
+    {
+      "epoch": 2.4983143361695297,
+      "grad_norm": 0.4277403652667999,
+      "learning_rate": 2.8552732036719687e-05,
+      "loss": 0.5281,
+      "step": 7781
+    },
+    {
+      "epoch": 2.4986354149943812,
+      "grad_norm": 0.5072173476219177,
+      "learning_rate": 2.852831353213715e-05,
+      "loss": 0.5995,
+      "step": 7782
+    },
+    {
+      "epoch": 2.498956493819233,
+      "grad_norm": 0.4317438006401062,
+      "learning_rate": 2.8503903736162875e-05,
+      "loss": 0.1693,
+      "step": 7783
+    },
+    {
+      "epoch": 2.4992775726440843,
+      "grad_norm": 0.39404723048210144,
+      "learning_rate": 2.8479502651770995e-05,
+      "loss": 0.1665,
+      "step": 7784
+    },
+    {
+      "epoch": 2.4995986514689355,
+      "grad_norm": 0.3772273659706116,
+      "learning_rate": 2.8455110281934803e-05,
+      "loss": 0.1733,
+      "step": 7785
+    },
+    {
+      "epoch": 2.499919730293787,
+      "grad_norm": 0.2818797528743744,
+      "learning_rate": 2.8430726629626413e-05,
+      "loss": 0.0803,
+      "step": 7786
+    },
+    {
+      "epoch": 2.5002408091186386,
+      "grad_norm": 0.6350283026695251,
+      "learning_rate": 2.840635169781688e-05,
+      "loss": 0.3251,
+      "step": 7787
+    },
+    {
+      "epoch": 2.50056188794349,
+      "grad_norm": 0.8361577987670898,
+      "learning_rate": 2.838198548947627e-05,
+      "loss": 0.4189,
+      "step": 7788
+    },
+    {
+      "epoch": 2.5008829667683417,
+      "grad_norm": 0.7367003560066223,
+      "learning_rate": 2.835762800757341e-05,
+      "loss": 0.2988,
+      "step": 7789
+    },
+    {
+      "epoch": 2.501204045593193,
+      "grad_norm": 0.8890562057495117,
+      "learning_rate": 2.8333279255076306e-05,
+      "loss": 0.4109,
+      "step": 7790
+    },
+    {
+      "epoch": 2.5015251244180448,
+      "grad_norm": 0.8579180836677551,
+      "learning_rate": 2.8308939234951726e-05,
+      "loss": 0.3457,
+      "step": 7791
+    },
+    {
+      "epoch": 2.501846203242896,
+      "grad_norm": 0.8420111536979675,
+      "learning_rate": 2.8284607950165442e-05,
+      "loss": 0.3082,
+      "step": 7792
+    },
+    {
+      "epoch": 2.502167282067748,
+      "grad_norm": 0.78440922498703,
+      "learning_rate": 2.826028540368215e-05,
+      "loss": 0.2953,
+      "step": 7793
+    },
+    {
+      "epoch": 2.502488360892599,
+      "grad_norm": 0.5694653391838074,
+      "learning_rate": 2.823597159846547e-05,
+      "loss": 0.2637,
+      "step": 7794
+    },
+    {
+      "epoch": 2.5028094397174505,
+      "grad_norm": 0.77173912525177,
+      "learning_rate": 2.8211666537477933e-05,
+      "loss": 0.3009,
+      "step": 7795
+    },
+    {
+      "epoch": 2.503130518542302,
+      "grad_norm": 0.9021281599998474,
+      "learning_rate": 2.8187370223681132e-05,
+      "loss": 0.3543,
+      "step": 7796
+    },
+    {
+      "epoch": 2.5034515973671536,
+      "grad_norm": 0.8287353515625,
+      "learning_rate": 2.816308266003541e-05,
+      "loss": 0.307,
+      "step": 7797
+    },
+    {
+      "epoch": 2.503772676192005,
+      "grad_norm": 1.0291224718093872,
+      "learning_rate": 2.813880384950016e-05,
+      "loss": 0.419,
+      "step": 7798
+    },
+    {
+      "epoch": 2.5040937550168567,
+      "grad_norm": 0.7518448233604431,
+      "learning_rate": 2.8114533795033683e-05,
+      "loss": 0.3259,
+      "step": 7799
+    },
+    {
+      "epoch": 2.5044148338417083,
+      "grad_norm": 0.9459495544433594,
+      "learning_rate": 2.8090272499593173e-05,
+      "loss": 0.3573,
+      "step": 7800
+    },
+    {
+      "epoch": 2.5047359126665594,
+      "grad_norm": 0.5885202884674072,
+      "learning_rate": 2.8066019966134904e-05,
+      "loss": 0.2741,
+      "step": 7801
+    },
+    {
+      "epoch": 2.5050569914914114,
+      "grad_norm": 0.9306980967521667,
+      "learning_rate": 2.8041776197613844e-05,
+      "loss": 0.3992,
+      "step": 7802
+    },
+    {
+      "epoch": 2.5053780703162625,
+      "grad_norm": 0.9558607339859009,
+      "learning_rate": 2.8017541196984142e-05,
+      "loss": 0.3941,
+      "step": 7803
+    },
+    {
+      "epoch": 2.505699149141114,
+      "grad_norm": 0.9242830276489258,
+      "learning_rate": 2.7993314967198635e-05,
+      "loss": 0.3423,
+      "step": 7804
+    },
+    {
+      "epoch": 2.5060202279659656,
+      "grad_norm": 0.6770075559616089,
+      "learning_rate": 2.7969097511209308e-05,
+      "loss": 0.273,
+      "step": 7805
+    },
+    {
+      "epoch": 2.506341306790817,
+      "grad_norm": 0.7561593055725098,
+      "learning_rate": 2.7944888831966987e-05,
+      "loss": 0.2808,
+      "step": 7806
+    },
+    {
+      "epoch": 2.5066623856156687,
+      "grad_norm": 0.8181188702583313,
+      "learning_rate": 2.7920688932421335e-05,
+      "loss": 0.3705,
+      "step": 7807
+    },
+    {
+      "epoch": 2.5069834644405202,
+      "grad_norm": 0.762374758720398,
+      "learning_rate": 2.7896497815521128e-05,
+      "loss": 0.3188,
+      "step": 7808
+    },
+    {
+      "epoch": 2.5073045432653718,
+      "grad_norm": 1.0830848217010498,
+      "learning_rate": 2.7872315484213952e-05,
+      "loss": 0.4268,
+      "step": 7809
+    },
+    {
+      "epoch": 2.507625622090223,
+      "grad_norm": 1.1398093700408936,
+      "learning_rate": 2.7848141941446347e-05,
+      "loss": 0.4128,
+      "step": 7810
+    },
+    {
+      "epoch": 2.507946700915075,
+      "grad_norm": 0.8498218059539795,
+      "learning_rate": 2.7823977190163786e-05,
+      "loss": 0.2717,
+      "step": 7811
+    },
+    {
+      "epoch": 2.508267779739926,
+      "grad_norm": 1.4102894067764282,
+      "learning_rate": 2.7799821233310674e-05,
+      "loss": 0.4987,
+      "step": 7812
+    },
+    {
+      "epoch": 2.5085888585647775,
+      "grad_norm": 1.8042616844177246,
+      "learning_rate": 2.7775674073830337e-05,
+      "loss": 0.3982,
+      "step": 7813
+    },
+    {
+      "epoch": 2.508909937389629,
+      "grad_norm": 0.7573620080947876,
+      "learning_rate": 2.775153571466502e-05,
+      "loss": 0.3681,
+      "step": 7814
+    },
+    {
+      "epoch": 2.5092310162144806,
+      "grad_norm": 1.1396507024765015,
+      "learning_rate": 2.772740615875594e-05,
+      "loss": 0.2719,
+      "step": 7815
+    },
+    {
+      "epoch": 2.509552095039332,
+      "grad_norm": 0.5033369064331055,
+      "learning_rate": 2.770328540904319e-05,
+      "loss": 0.2497,
+      "step": 7816
+    },
+    {
+      "epoch": 2.5098731738641837,
+      "grad_norm": 0.7801833152770996,
+      "learning_rate": 2.7679173468465812e-05,
+      "loss": 0.2883,
+      "step": 7817
+    },
+    {
+      "epoch": 2.5101942526890353,
+      "grad_norm": 0.8423357009887695,
+      "learning_rate": 2.7655070339961776e-05,
+      "loss": 0.3138,
+      "step": 7818
+    },
+    {
+      "epoch": 2.5105153315138864,
+      "grad_norm": 0.672028124332428,
+      "learning_rate": 2.7630976026467968e-05,
+      "loss": 0.2912,
+      "step": 7819
+    },
+    {
+      "epoch": 2.5108364103387384,
+      "grad_norm": 0.8141403198242188,
+      "learning_rate": 2.7606890530920195e-05,
+      "loss": 0.3421,
+      "step": 7820
+    },
+    {
+      "epoch": 2.5111574891635895,
+      "grad_norm": 0.9063519239425659,
+      "learning_rate": 2.7582813856253275e-05,
+      "loss": 0.3336,
+      "step": 7821
+    },
+    {
+      "epoch": 2.511478567988441,
+      "grad_norm": 0.767846941947937,
+      "learning_rate": 2.755874600540078e-05,
+      "loss": 0.3033,
+      "step": 7822
+    },
+    {
+      "epoch": 2.5117996468132926,
+      "grad_norm": 2.2708170413970947,
+      "learning_rate": 2.753468698129533e-05,
+      "loss": 0.4001,
+      "step": 7823
+    },
+    {
+      "epoch": 2.512120725638144,
+      "grad_norm": 1.0831410884857178,
+      "learning_rate": 2.7510636786868514e-05,
+      "loss": 0.3136,
+      "step": 7824
+    },
+    {
+      "epoch": 2.5124418044629957,
+      "grad_norm": 0.8670080900192261,
+      "learning_rate": 2.7486595425050665e-05,
+      "loss": 0.3511,
+      "step": 7825
+    },
+    {
+      "epoch": 2.5127628832878472,
+      "grad_norm": 0.7800338268280029,
+      "learning_rate": 2.746256289877126e-05,
+      "loss": 0.3466,
+      "step": 7826
+    },
+    {
+      "epoch": 2.513083962112699,
+      "grad_norm": 0.871461033821106,
+      "learning_rate": 2.743853921095848e-05,
+      "loss": 0.2794,
+      "step": 7827
+    },
+    {
+      "epoch": 2.51340504093755,
+      "grad_norm": 0.7801231741905212,
+      "learning_rate": 2.741452436453963e-05,
+      "loss": 0.3329,
+      "step": 7828
+    },
+    {
+      "epoch": 2.513726119762402,
+      "grad_norm": 0.9049973487854004,
+      "learning_rate": 2.7390518362440808e-05,
+      "loss": 0.3623,
+      "step": 7829
+    },
+    {
+      "epoch": 2.514047198587253,
+      "grad_norm": 0.6395286321640015,
+      "learning_rate": 2.736652120758708e-05,
+      "loss": 0.7761,
+      "step": 7830
+    },
+    {
+      "epoch": 2.5143682774121046,
+      "grad_norm": 0.4424824118614197,
+      "learning_rate": 2.734253290290242e-05,
+      "loss": 0.6206,
+      "step": 7831
+    },
+    {
+      "epoch": 2.514689356236956,
+      "grad_norm": 0.5469640493392944,
+      "learning_rate": 2.7318553451309726e-05,
+      "loss": 0.3887,
+      "step": 7832
+    },
+    {
+      "epoch": 2.5150104350618077,
+      "grad_norm": 0.3367752134799957,
+      "learning_rate": 2.7294582855730832e-05,
+      "loss": 0.0776,
+      "step": 7833
+    },
+    {
+      "epoch": 2.515331513886659,
+      "grad_norm": 0.5025734305381775,
+      "learning_rate": 2.727062111908647e-05,
+      "loss": 0.2786,
+      "step": 7834
+    },
+    {
+      "epoch": 2.5156525927115108,
+      "grad_norm": 0.4243975281715393,
+      "learning_rate": 2.7246668244296323e-05,
+      "loss": 0.2066,
+      "step": 7835
+    },
+    {
+      "epoch": 2.5159736715363623,
+      "grad_norm": 0.6841888427734375,
+      "learning_rate": 2.722272423427896e-05,
+      "loss": 0.3834,
+      "step": 7836
+    },
+    {
+      "epoch": 2.5162947503612134,
+      "grad_norm": 0.6961179375648499,
+      "learning_rate": 2.7198789091951902e-05,
+      "loss": 0.4068,
+      "step": 7837
+    },
+    {
+      "epoch": 2.5166158291860654,
+      "grad_norm": 0.9804593920707703,
+      "learning_rate": 2.717486282023153e-05,
+      "loss": 0.3961,
+      "step": 7838
+    },
+    {
+      "epoch": 2.5169369080109165,
+      "grad_norm": 0.7120915651321411,
+      "learning_rate": 2.715094542203327e-05,
+      "loss": 0.3416,
+      "step": 7839
+    },
+    {
+      "epoch": 2.517257986835768,
+      "grad_norm": 0.7814658284187317,
+      "learning_rate": 2.7127036900271317e-05,
+      "loss": 0.3016,
+      "step": 7840
+    },
+    {
+      "epoch": 2.5175790656606196,
+      "grad_norm": 0.5942254662513733,
+      "learning_rate": 2.7103137257858868e-05,
+      "loss": 0.281,
+      "step": 7841
+    },
+    {
+      "epoch": 2.517900144485471,
+      "grad_norm": 0.8657404780387878,
+      "learning_rate": 2.707924649770802e-05,
+      "loss": 0.3926,
+      "step": 7842
+    },
+    {
+      "epoch": 2.5182212233103227,
+      "grad_norm": 0.7496175765991211,
+      "learning_rate": 2.7055364622729773e-05,
+      "loss": 0.3238,
+      "step": 7843
+    },
+    {
+      "epoch": 2.5185423021351743,
+      "grad_norm": 0.9622612595558167,
+      "learning_rate": 2.7031491635834137e-05,
+      "loss": 0.3779,
+      "step": 7844
+    },
+    {
+      "epoch": 2.518863380960026,
+      "grad_norm": 1.0655999183654785,
+      "learning_rate": 2.700762753992985e-05,
+      "loss": 0.2991,
+      "step": 7845
+    },
+    {
+      "epoch": 2.519184459784877,
+      "grad_norm": 0.988420844078064,
+      "learning_rate": 2.698377233792476e-05,
+      "loss": 0.4061,
+      "step": 7846
+    },
+    {
+      "epoch": 2.519505538609729,
+      "grad_norm": 0.6470503211021423,
+      "learning_rate": 2.6959926032725535e-05,
+      "loss": 0.2876,
+      "step": 7847
+    },
+    {
+      "epoch": 2.51982661743458,
+      "grad_norm": 0.7641369104385376,
+      "learning_rate": 2.6936088627237765e-05,
+      "loss": 0.2941,
+      "step": 7848
+    },
+    {
+      "epoch": 2.5201476962594316,
+      "grad_norm": 0.9086601138114929,
+      "learning_rate": 2.6912260124366006e-05,
+      "loss": 0.3384,
+      "step": 7849
+    },
+    {
+      "epoch": 2.520468775084283,
+      "grad_norm": 0.7186222672462463,
+      "learning_rate": 2.688844052701359e-05,
+      "loss": 0.2905,
+      "step": 7850
+    },
+    {
+      "epoch": 2.5207898539091347,
+      "grad_norm": 0.9355291724205017,
+      "learning_rate": 2.6864629838082956e-05,
+      "loss": 0.4228,
+      "step": 7851
+    },
+    {
+      "epoch": 2.5211109327339862,
+      "grad_norm": 0.6973735690116882,
+      "learning_rate": 2.6840828060475332e-05,
+      "loss": 0.2988,
+      "step": 7852
+    },
+    {
+      "epoch": 2.521432011558838,
+      "grad_norm": 0.8942835330963135,
+      "learning_rate": 2.681703519709089e-05,
+      "loss": 0.3534,
+      "step": 7853
+    },
+    {
+      "epoch": 2.5217530903836893,
+      "grad_norm": 0.9520301222801208,
+      "learning_rate": 2.679325125082872e-05,
+      "loss": 0.3538,
+      "step": 7854
+    },
+    {
+      "epoch": 2.5220741692085404,
+      "grad_norm": 0.8248050212860107,
+      "learning_rate": 2.676947622458683e-05,
+      "loss": 0.332,
+      "step": 7855
+    },
+    {
+      "epoch": 2.5223952480333924,
+      "grad_norm": 1.0398519039154053,
+      "learning_rate": 2.6745710121262136e-05,
+      "loss": 0.4481,
+      "step": 7856
+    },
+    {
+      "epoch": 2.5227163268582435,
+      "grad_norm": 0.9661288857460022,
+      "learning_rate": 2.672195294375045e-05,
+      "loss": 0.3479,
+      "step": 7857
+    },
+    {
+      "epoch": 2.523037405683095,
+      "grad_norm": 0.7361083030700684,
+      "learning_rate": 2.6698204694946527e-05,
+      "loss": 0.2805,
+      "step": 7858
+    },
+    {
+      "epoch": 2.5233584845079466,
+      "grad_norm": 0.7523530125617981,
+      "learning_rate": 2.6674465377744017e-05,
+      "loss": 0.2939,
+      "step": 7859
+    },
+    {
+      "epoch": 2.523679563332798,
+      "grad_norm": 1.0147556066513062,
+      "learning_rate": 2.6650734995035477e-05,
+      "loss": 0.3913,
+      "step": 7860
+    },
+    {
+      "epoch": 2.5240006421576497,
+      "grad_norm": 0.7158978581428528,
+      "learning_rate": 2.6627013549712355e-05,
+      "loss": 0.2834,
+      "step": 7861
+    },
+    {
+      "epoch": 2.5243217209825013,
+      "grad_norm": 0.8579987287521362,
+      "learning_rate": 2.660330104466513e-05,
+      "loss": 0.3325,
+      "step": 7862
+    },
+    {
+      "epoch": 2.524642799807353,
+      "grad_norm": 0.8115719556808472,
+      "learning_rate": 2.657959748278297e-05,
+      "loss": 0.3175,
+      "step": 7863
+    },
+    {
+      "epoch": 2.524963878632204,
+      "grad_norm": 0.7678377032279968,
+      "learning_rate": 2.655590286695422e-05,
+      "loss": 0.3265,
+      "step": 7864
+    },
+    {
+      "epoch": 2.525284957457056,
+      "grad_norm": 0.8553791642189026,
+      "learning_rate": 2.6532217200065858e-05,
+      "loss": 0.3043,
+      "step": 7865
+    },
+    {
+      "epoch": 2.525606036281907,
+      "grad_norm": 0.8253560662269592,
+      "learning_rate": 2.6508540485004006e-05,
+      "loss": 0.2808,
+      "step": 7866
+    },
+    {
+      "epoch": 2.5259271151067586,
+      "grad_norm": 0.8303504586219788,
+      "learning_rate": 2.6484872724653608e-05,
+      "loss": 0.3263,
+      "step": 7867
+    },
+    {
+      "epoch": 2.52624819393161,
+      "grad_norm": 0.8876441717147827,
+      "learning_rate": 2.646121392189841e-05,
+      "loss": 0.3131,
+      "step": 7868
+    },
+    {
+      "epoch": 2.5265692727564617,
+      "grad_norm": 0.9586053490638733,
+      "learning_rate": 2.6437564079621267e-05,
+      "loss": 0.3133,
+      "step": 7869
+    },
+    {
+      "epoch": 2.5268903515813133,
+      "grad_norm": 1.3426408767700195,
+      "learning_rate": 2.6413923200703794e-05,
+      "loss": 0.3845,
+      "step": 7870
+    },
+    {
+      "epoch": 2.527211430406165,
+      "grad_norm": 0.7227518558502197,
+      "learning_rate": 2.639029128802657e-05,
+      "loss": 0.2797,
+      "step": 7871
+    },
+    {
+      "epoch": 2.5275325092310164,
+      "grad_norm": 0.6822044253349304,
+      "learning_rate": 2.636666834446907e-05,
+      "loss": 0.2739,
+      "step": 7872
+    },
+    {
+      "epoch": 2.5278535880558675,
+      "grad_norm": 0.6981026530265808,
+      "learning_rate": 2.634305437290968e-05,
+      "loss": 0.2966,
+      "step": 7873
+    },
+    {
+      "epoch": 2.5281746668807195,
+      "grad_norm": 0.6758238673210144,
+      "learning_rate": 2.631944937622569e-05,
+      "loss": 0.2845,
+      "step": 7874
+    },
+    {
+      "epoch": 2.5284957457055706,
+      "grad_norm": 0.8002867698669434,
+      "learning_rate": 2.6295853357293298e-05,
+      "loss": 0.3453,
+      "step": 7875
+    },
+    {
+      "epoch": 2.528816824530422,
+      "grad_norm": 0.7498476505279541,
+      "learning_rate": 2.6272266318987603e-05,
+      "loss": 0.3061,
+      "step": 7876
+    },
+    {
+      "epoch": 2.5291379033552737,
+      "grad_norm": 0.633752703666687,
+      "learning_rate": 2.624868826418262e-05,
+      "loss": 0.2688,
+      "step": 7877
+    },
+    {
+      "epoch": 2.529458982180125,
+      "grad_norm": 0.6049948334693909,
+      "learning_rate": 2.6225119195751258e-05,
+      "loss": 0.2902,
+      "step": 7878
+    },
+    {
+      "epoch": 2.5297800610049768,
+      "grad_norm": 0.6882827877998352,
+      "learning_rate": 2.6201559116565345e-05,
+      "loss": 0.3093,
+      "step": 7879
+    },
+    {
+      "epoch": 2.5301011398298283,
+      "grad_norm": 0.4292110800743103,
+      "learning_rate": 2.6178008029495592e-05,
+      "loss": 0.7076,
+      "step": 7880
+    },
+    {
+      "epoch": 2.53042221865468,
+      "grad_norm": 0.42428478598594666,
+      "learning_rate": 2.615446593741161e-05,
+      "loss": 0.3113,
+      "step": 7881
+    },
+    {
+      "epoch": 2.530743297479531,
+      "grad_norm": 0.4815612733364105,
+      "learning_rate": 2.613093284318201e-05,
+      "loss": 0.3371,
+      "step": 7882
+    },
+    {
+      "epoch": 2.531064376304383,
+      "grad_norm": 0.5496947765350342,
+      "learning_rate": 2.6107408749674122e-05,
+      "loss": 0.2977,
+      "step": 7883
+    },
+    {
+      "epoch": 2.531385455129234,
+      "grad_norm": 0.33660855889320374,
+      "learning_rate": 2.6083893659754356e-05,
+      "loss": 0.0772,
+      "step": 7884
+    },
+    {
+      "epoch": 2.5317065339540856,
+      "grad_norm": 0.40665629506111145,
+      "learning_rate": 2.606038757628798e-05,
+      "loss": 0.1418,
+      "step": 7885
+    },
+    {
+      "epoch": 2.532027612778937,
+      "grad_norm": 0.2342071682214737,
+      "learning_rate": 2.603689050213902e-05,
+      "loss": 0.0695,
+      "step": 7886
+    },
+    {
+      "epoch": 2.5323486916037887,
+      "grad_norm": 0.29153773188591003,
+      "learning_rate": 2.6013402440170676e-05,
+      "loss": 0.116,
+      "step": 7887
+    },
+    {
+      "epoch": 2.5326697704286403,
+      "grad_norm": 1.029974341392517,
+      "learning_rate": 2.5989923393244742e-05,
+      "loss": 0.4608,
+      "step": 7888
+    },
+    {
+      "epoch": 2.532990849253492,
+      "grad_norm": 0.7505784034729004,
+      "learning_rate": 2.5966453364222186e-05,
+      "loss": 0.3698,
+      "step": 7889
+    },
+    {
+      "epoch": 2.5333119280783434,
+      "grad_norm": 0.8639885783195496,
+      "learning_rate": 2.5942992355962727e-05,
+      "loss": 0.313,
+      "step": 7890
+    },
+    {
+      "epoch": 2.5336330069031945,
+      "grad_norm": 0.9737856388092041,
+      "learning_rate": 2.5919540371325e-05,
+      "loss": 0.431,
+      "step": 7891
+    },
+    {
+      "epoch": 2.5339540857280465,
+      "grad_norm": 0.6815305352210999,
+      "learning_rate": 2.5896097413166564e-05,
+      "loss": 0.2643,
+      "step": 7892
+    },
+    {
+      "epoch": 2.5342751645528976,
+      "grad_norm": 0.9808143377304077,
+      "learning_rate": 2.5872663484343884e-05,
+      "loss": 0.4601,
+      "step": 7893
+    },
+    {
+      "epoch": 2.534596243377749,
+      "grad_norm": 0.9797862768173218,
+      "learning_rate": 2.584923858771231e-05,
+      "loss": 0.3746,
+      "step": 7894
+    },
+    {
+      "epoch": 2.5349173222026007,
+      "grad_norm": 0.7613912224769592,
+      "learning_rate": 2.582582272612609e-05,
+      "loss": 0.3253,
+      "step": 7895
+    },
+    {
+      "epoch": 2.5352384010274522,
+      "grad_norm": 0.8609127402305603,
+      "learning_rate": 2.580241590243837e-05,
+      "loss": 0.3091,
+      "step": 7896
+    },
+    {
+      "epoch": 2.535559479852304,
+      "grad_norm": 0.8038222789764404,
+      "learning_rate": 2.5779018119501208e-05,
+      "loss": 0.3423,
+      "step": 7897
+    },
+    {
+      "epoch": 2.5358805586771553,
+      "grad_norm": 0.8532456159591675,
+      "learning_rate": 2.575562938016556e-05,
+      "loss": 0.3167,
+      "step": 7898
+    },
+    {
+      "epoch": 2.536201637502007,
+      "grad_norm": 0.8715474009513855,
+      "learning_rate": 2.573224968728123e-05,
+      "loss": 0.3521,
+      "step": 7899
+    },
+    {
+      "epoch": 2.536522716326858,
+      "grad_norm": 0.95112144947052,
+      "learning_rate": 2.5708879043697054e-05,
+      "loss": 0.3424,
+      "step": 7900
+    },
+    {
+      "epoch": 2.53684379515171,
+      "grad_norm": 0.8659119009971619,
+      "learning_rate": 2.5685517452260567e-05,
+      "loss": 0.3874,
+      "step": 7901
+    },
+    {
+      "epoch": 2.537164873976561,
+      "grad_norm": 0.9276009798049927,
+      "learning_rate": 2.566216491581841e-05,
+      "loss": 0.3019,
+      "step": 7902
+    },
+    {
+      "epoch": 2.5374859528014126,
+      "grad_norm": 0.8854154944419861,
+      "learning_rate": 2.5638821437215944e-05,
+      "loss": 0.3359,
+      "step": 7903
+    },
+    {
+      "epoch": 2.537807031626264,
+      "grad_norm": 0.9696468114852905,
+      "learning_rate": 2.561548701929749e-05,
+      "loss": 0.4029,
+      "step": 7904
+    },
+    {
+      "epoch": 2.5381281104511157,
+      "grad_norm": 0.8522638082504272,
+      "learning_rate": 2.5592161664906368e-05,
+      "loss": 0.3034,
+      "step": 7905
+    },
+    {
+      "epoch": 2.5384491892759673,
+      "grad_norm": 0.8860786557197571,
+      "learning_rate": 2.5568845376884587e-05,
+      "loss": 0.2844,
+      "step": 7906
+    },
+    {
+      "epoch": 2.538770268100819,
+      "grad_norm": 0.919465184211731,
+      "learning_rate": 2.554553815807328e-05,
+      "loss": 0.3504,
+      "step": 7907
+    },
+    {
+      "epoch": 2.5390913469256704,
+      "grad_norm": 0.8986783623695374,
+      "learning_rate": 2.5522240011312247e-05,
+      "loss": 0.3556,
+      "step": 7908
+    },
+    {
+      "epoch": 2.5394124257505215,
+      "grad_norm": 0.5819917321205139,
+      "learning_rate": 2.549895093944039e-05,
+      "loss": 0.2474,
+      "step": 7909
+    },
+    {
+      "epoch": 2.5397335045753735,
+      "grad_norm": 0.701481282711029,
+      "learning_rate": 2.547567094529537e-05,
+      "loss": 0.2874,
+      "step": 7910
+    },
+    {
+      "epoch": 2.5400545834002246,
+      "grad_norm": 0.9206550121307373,
+      "learning_rate": 2.5452400031713785e-05,
+      "loss": 0.327,
+      "step": 7911
+    },
+    {
+      "epoch": 2.540375662225076,
+      "grad_norm": 0.8680040836334229,
+      "learning_rate": 2.542913820153113e-05,
+      "loss": 0.3246,
+      "step": 7912
+    },
+    {
+      "epoch": 2.5406967410499277,
+      "grad_norm": 0.9277855157852173,
+      "learning_rate": 2.540588545758179e-05,
+      "loss": 0.2907,
+      "step": 7913
+    },
+    {
+      "epoch": 2.5410178198747793,
+      "grad_norm": 0.8211881518363953,
+      "learning_rate": 2.5382641802699035e-05,
+      "loss": 0.2917,
+      "step": 7914
+    },
+    {
+      "epoch": 2.541338898699631,
+      "grad_norm": 1.1107295751571655,
+      "learning_rate": 2.535940723971505e-05,
+      "loss": 0.4215,
+      "step": 7915
+    },
+    {
+      "epoch": 2.5416599775244824,
+      "grad_norm": 0.9217866659164429,
+      "learning_rate": 2.5336181771460876e-05,
+      "loss": 0.3113,
+      "step": 7916
+    },
+    {
+      "epoch": 2.541981056349334,
+      "grad_norm": 0.8937177062034607,
+      "learning_rate": 2.5312965400766474e-05,
+      "loss": 0.2755,
+      "step": 7917
+    },
+    {
+      "epoch": 2.542302135174185,
+      "grad_norm": 0.5329973697662354,
+      "learning_rate": 2.5289758130460683e-05,
+      "loss": 0.2237,
+      "step": 7918
+    },
+    {
+      "epoch": 2.542623213999037,
+      "grad_norm": 0.9037802815437317,
+      "learning_rate": 2.5266559963371216e-05,
+      "loss": 0.3482,
+      "step": 7919
+    },
+    {
+      "epoch": 2.542944292823888,
+      "grad_norm": 1.0273380279541016,
+      "learning_rate": 2.5243370902324792e-05,
+      "loss": 0.3545,
+      "step": 7920
+    },
+    {
+      "epoch": 2.5432653716487397,
+      "grad_norm": 0.8402252197265625,
+      "learning_rate": 2.5220190950146827e-05,
+      "loss": 0.3273,
+      "step": 7921
+    },
+    {
+      "epoch": 2.543586450473591,
+      "grad_norm": 1.0652894973754883,
+      "learning_rate": 2.5197020109661772e-05,
+      "loss": 0.348,
+      "step": 7922
+    },
+    {
+      "epoch": 2.5439075292984428,
+      "grad_norm": 1.2122594118118286,
+      "learning_rate": 2.5173858383692906e-05,
+      "loss": 0.4131,
+      "step": 7923
+    },
+    {
+      "epoch": 2.5442286081232943,
+      "grad_norm": 1.0098888874053955,
+      "learning_rate": 2.51507057750624e-05,
+      "loss": 0.3763,
+      "step": 7924
+    },
+    {
+      "epoch": 2.544549686948146,
+      "grad_norm": 0.7478073835372925,
+      "learning_rate": 2.512756228659141e-05,
+      "loss": 0.2882,
+      "step": 7925
+    },
+    {
+      "epoch": 2.5448707657729974,
+      "grad_norm": 0.486265629529953,
+      "learning_rate": 2.5104427921099782e-05,
+      "loss": 0.2455,
+      "step": 7926
+    },
+    {
+      "epoch": 2.5451918445978485,
+      "grad_norm": 0.5730571746826172,
+      "learning_rate": 2.508130268140646e-05,
+      "loss": 0.2751,
+      "step": 7927
+    },
+    {
+      "epoch": 2.5455129234227005,
+      "grad_norm": 0.6584112048149109,
+      "learning_rate": 2.5058186570329156e-05,
+      "loss": 0.2707,
+      "step": 7928
+    },
+    {
+      "epoch": 2.5458340022475516,
+      "grad_norm": 0.6145785450935364,
+      "learning_rate": 2.5035079590684497e-05,
+      "loss": 0.323,
+      "step": 7929
+    },
+    {
+      "epoch": 2.546155081072403,
+      "grad_norm": 0.4932810366153717,
+      "learning_rate": 2.5011981745288015e-05,
+      "loss": 0.5853,
+      "step": 7930
+    },
+    {
+      "epoch": 2.5464761598972547,
+      "grad_norm": 0.5000682473182678,
+      "learning_rate": 2.4988893036954043e-05,
+      "loss": 0.3042,
+      "step": 7931
+    },
+    {
+      "epoch": 2.5467972387221063,
+      "grad_norm": 0.4661361873149872,
+      "learning_rate": 2.496581346849596e-05,
+      "loss": 0.2213,
+      "step": 7932
+    },
+    {
+      "epoch": 2.547118317546958,
+      "grad_norm": 0.6509041786193848,
+      "learning_rate": 2.4942743042725892e-05,
+      "loss": 0.3622,
+      "step": 7933
+    },
+    {
+      "epoch": 2.5474393963718094,
+      "grad_norm": 0.3958999216556549,
+      "learning_rate": 2.4919681762454918e-05,
+      "loss": 0.2316,
+      "step": 7934
+    },
+    {
+      "epoch": 2.547760475196661,
+      "grad_norm": 0.45946624875068665,
+      "learning_rate": 2.4896629630492973e-05,
+      "loss": 0.2446,
+      "step": 7935
+    },
+    {
+      "epoch": 2.548081554021512,
+      "grad_norm": 0.3704213500022888,
+      "learning_rate": 2.4873586649648894e-05,
+      "loss": 0.0943,
+      "step": 7936
+    },
+    {
+      "epoch": 2.548402632846364,
+      "grad_norm": 0.6541774272918701,
+      "learning_rate": 2.48505528227304e-05,
+      "loss": 0.2996,
+      "step": 7937
+    },
+    {
+      "epoch": 2.548723711671215,
+      "grad_norm": 0.9613190293312073,
+      "learning_rate": 2.48275281525441e-05,
+      "loss": 0.564,
+      "step": 7938
+    },
+    {
+      "epoch": 2.5490447904960667,
+      "grad_norm": 0.8291346430778503,
+      "learning_rate": 2.480451264189546e-05,
+      "loss": 0.4429,
+      "step": 7939
+    },
+    {
+      "epoch": 2.5493658693209182,
+      "grad_norm": 0.6980516910552979,
+      "learning_rate": 2.4781506293588873e-05,
+      "loss": 0.3154,
+      "step": 7940
+    },
+    {
+      "epoch": 2.54968694814577,
+      "grad_norm": 0.7115036249160767,
+      "learning_rate": 2.4758509110427575e-05,
+      "loss": 0.2834,
+      "step": 7941
+    },
+    {
+      "epoch": 2.5500080269706213,
+      "grad_norm": 0.7645947933197021,
+      "learning_rate": 2.4735521095213687e-05,
+      "loss": 0.3213,
+      "step": 7942
+    },
+    {
+      "epoch": 2.550329105795473,
+      "grad_norm": 0.8161126971244812,
+      "learning_rate": 2.4712542250748304e-05,
+      "loss": 0.3376,
+      "step": 7943
+    },
+    {
+      "epoch": 2.5506501846203244,
+      "grad_norm": 0.9567040801048279,
+      "learning_rate": 2.4689572579831222e-05,
+      "loss": 0.379,
+      "step": 7944
+    },
+    {
+      "epoch": 2.5509712634451756,
+      "grad_norm": 0.878835916519165,
+      "learning_rate": 2.4666612085261342e-05,
+      "loss": 0.3387,
+      "step": 7945
+    },
+    {
+      "epoch": 2.5512923422700275,
+      "grad_norm": 0.9040188193321228,
+      "learning_rate": 2.464366076983623e-05,
+      "loss": 0.3488,
+      "step": 7946
+    },
+    {
+      "epoch": 2.5516134210948787,
+      "grad_norm": 0.9474675059318542,
+      "learning_rate": 2.4620718636352457e-05,
+      "loss": 0.3582,
+      "step": 7947
+    },
+    {
+      "epoch": 2.55193449991973,
+      "grad_norm": 0.6305925250053406,
+      "learning_rate": 2.4597785687605513e-05,
+      "loss": 0.256,
+      "step": 7948
+    },
+    {
+      "epoch": 2.5522555787445818,
+      "grad_norm": 0.7935699820518494,
+      "learning_rate": 2.4574861926389615e-05,
+      "loss": 0.3362,
+      "step": 7949
+    },
+    {
+      "epoch": 2.5525766575694333,
+      "grad_norm": 0.9189661741256714,
+      "learning_rate": 2.4551947355498027e-05,
+      "loss": 0.3227,
+      "step": 7950
+    },
+    {
+      "epoch": 2.552897736394285,
+      "grad_norm": 1.0050448179244995,
+      "learning_rate": 2.45290419777228e-05,
+      "loss": 0.4139,
+      "step": 7951
+    },
+    {
+      "epoch": 2.5532188152191364,
+      "grad_norm": 1.0046634674072266,
+      "learning_rate": 2.4506145795854873e-05,
+      "loss": 0.4224,
+      "step": 7952
+    },
+    {
+      "epoch": 2.553539894043988,
+      "grad_norm": 0.7242197394371033,
+      "learning_rate": 2.4483258812684096e-05,
+      "loss": 0.3028,
+      "step": 7953
+    },
+    {
+      "epoch": 2.553860972868839,
+      "grad_norm": 0.6570572257041931,
+      "learning_rate": 2.4460381030999158e-05,
+      "loss": 0.2526,
+      "step": 7954
+    },
+    {
+      "epoch": 2.554182051693691,
+      "grad_norm": 0.7426683902740479,
+      "learning_rate": 2.443751245358765e-05,
+      "loss": 0.2802,
+      "step": 7955
+    },
+    {
+      "epoch": 2.554503130518542,
+      "grad_norm": 0.7006813883781433,
+      "learning_rate": 2.441465308323605e-05,
+      "loss": 0.3058,
+      "step": 7956
+    },
+    {
+      "epoch": 2.5548242093433937,
+      "grad_norm": 0.7258092164993286,
+      "learning_rate": 2.43918029227297e-05,
+      "loss": 0.2643,
+      "step": 7957
+    },
+    {
+      "epoch": 2.5551452881682453,
+      "grad_norm": 1.4597491025924683,
+      "learning_rate": 2.436896197485282e-05,
+      "loss": 0.3756,
+      "step": 7958
+    },
+    {
+      "epoch": 2.555466366993097,
+      "grad_norm": 0.7586961984634399,
+      "learning_rate": 2.43461302423885e-05,
+      "loss": 0.3493,
+      "step": 7959
+    },
+    {
+      "epoch": 2.5557874458179484,
+      "grad_norm": 1.1637358665466309,
+      "learning_rate": 2.4323307728118738e-05,
+      "loss": 0.4117,
+      "step": 7960
+    },
+    {
+      "epoch": 2.5561085246428,
+      "grad_norm": 0.764044463634491,
+      "learning_rate": 2.4300494434824373e-05,
+      "loss": 0.2827,
+      "step": 7961
+    },
+    {
+      "epoch": 2.5564296034676515,
+      "grad_norm": 0.7120959162712097,
+      "learning_rate": 2.4277690365285112e-05,
+      "loss": 0.2924,
+      "step": 7962
+    },
+    {
+      "epoch": 2.5567506822925026,
+      "grad_norm": 0.6114968061447144,
+      "learning_rate": 2.425489552227964e-05,
+      "loss": 0.2665,
+      "step": 7963
+    },
+    {
+      "epoch": 2.5570717611173546,
+      "grad_norm": 0.710640013217926,
+      "learning_rate": 2.4232109908585377e-05,
+      "loss": 0.3022,
+      "step": 7964
+    },
+    {
+      "epoch": 2.5573928399422057,
+      "grad_norm": 0.6761636137962341,
+      "learning_rate": 2.420933352697865e-05,
+      "loss": 0.2818,
+      "step": 7965
+    },
+    {
+      "epoch": 2.5577139187670572,
+      "grad_norm": 0.927751898765564,
+      "learning_rate": 2.4186566380234798e-05,
+      "loss": 0.363,
+      "step": 7966
+    },
+    {
+      "epoch": 2.5580349975919088,
+      "grad_norm": 0.8839023113250732,
+      "learning_rate": 2.4163808471127812e-05,
+      "loss": 0.3743,
+      "step": 7967
+    },
+    {
+      "epoch": 2.5583560764167603,
+      "grad_norm": 0.7871977090835571,
+      "learning_rate": 2.4141059802430777e-05,
+      "loss": 0.3433,
+      "step": 7968
+    },
+    {
+      "epoch": 2.558677155241612,
+      "grad_norm": 0.7037361860275269,
+      "learning_rate": 2.411832037691545e-05,
+      "loss": 0.2797,
+      "step": 7969
+    },
+    {
+      "epoch": 2.5589982340664634,
+      "grad_norm": 0.6723461151123047,
+      "learning_rate": 2.4095590197352635e-05,
+      "loss": 0.2862,
+      "step": 7970
+    },
+    {
+      "epoch": 2.559319312891315,
+      "grad_norm": 1.6565648317337036,
+      "learning_rate": 2.407286926651192e-05,
+      "loss": 0.3784,
+      "step": 7971
+    },
+    {
+      "epoch": 2.559640391716166,
+      "grad_norm": 0.8423799276351929,
+      "learning_rate": 2.405015758716177e-05,
+      "loss": 0.3592,
+      "step": 7972
+    },
+    {
+      "epoch": 2.559961470541018,
+      "grad_norm": 1.355826735496521,
+      "learning_rate": 2.4027455162069567e-05,
+      "loss": 0.3569,
+      "step": 7973
+    },
+    {
+      "epoch": 2.560282549365869,
+      "grad_norm": 0.8471829891204834,
+      "learning_rate": 2.4004761994001435e-05,
+      "loss": 0.2898,
+      "step": 7974
+    },
+    {
+      "epoch": 2.5606036281907207,
+      "grad_norm": 0.7122255563735962,
+      "learning_rate": 2.3982078085722582e-05,
+      "loss": 0.3251,
+      "step": 7975
+    },
+    {
+      "epoch": 2.5609247070155723,
+      "grad_norm": 0.792010486125946,
+      "learning_rate": 2.3959403439996907e-05,
+      "loss": 0.2844,
+      "step": 7976
+    },
+    {
+      "epoch": 2.561245785840424,
+      "grad_norm": 0.43959325551986694,
+      "learning_rate": 2.3936738059587282e-05,
+      "loss": 0.2479,
+      "step": 7977
+    },
+    {
+      "epoch": 2.5615668646652754,
+      "grad_norm": 0.33454373478889465,
+      "learning_rate": 2.3914081947255397e-05,
+      "loss": 0.2251,
+      "step": 7978
+    },
+    {
+      "epoch": 2.561887943490127,
+      "grad_norm": 0.6790142059326172,
+      "learning_rate": 2.389143510576184e-05,
+      "loss": 0.2833,
+      "step": 7979
+    },
+    {
+      "epoch": 2.5622090223149785,
+      "grad_norm": 0.5404451489448547,
+      "learning_rate": 2.3868797537866016e-05,
+      "loss": 0.6905,
+      "step": 7980
+    },
+    {
+      "epoch": 2.5625301011398296,
+      "grad_norm": 0.36907678842544556,
+      "learning_rate": 2.3846169246326343e-05,
+      "loss": 0.3373,
+      "step": 7981
+    },
+    {
+      "epoch": 2.5628511799646816,
+      "grad_norm": 0.5592865347862244,
+      "learning_rate": 2.3823550233899915e-05,
+      "loss": 0.4475,
+      "step": 7982
+    },
+    {
+      "epoch": 2.5631722587895327,
+      "grad_norm": 0.4797590374946594,
+      "learning_rate": 2.3800940503342828e-05,
+      "loss": 0.1677,
+      "step": 7983
+    },
+    {
+      "epoch": 2.5634933376143842,
+      "grad_norm": 0.6905731558799744,
+      "learning_rate": 2.3778340057409998e-05,
+      "loss": 0.2087,
+      "step": 7984
+    },
+    {
+      "epoch": 2.563814416439236,
+      "grad_norm": 0.22037728130817413,
+      "learning_rate": 2.37557488988552e-05,
+      "loss": 0.0671,
+      "step": 7985
+    },
+    {
+      "epoch": 2.5641354952640873,
+      "grad_norm": 0.41251930594444275,
+      "learning_rate": 2.373316703043119e-05,
+      "loss": 0.1551,
+      "step": 7986
+    },
+    {
+      "epoch": 2.564456574088939,
+      "grad_norm": 0.3948359191417694,
+      "learning_rate": 2.3710594454889378e-05,
+      "loss": 0.1627,
+      "step": 7987
+    },
+    {
+      "epoch": 2.5647776529137905,
+      "grad_norm": 0.5292665362358093,
+      "learning_rate": 2.3688031174980275e-05,
+      "loss": 0.2721,
+      "step": 7988
+    },
+    {
+      "epoch": 2.565098731738642,
+      "grad_norm": 0.8082164525985718,
+      "learning_rate": 2.3665477193453034e-05,
+      "loss": 0.3934,
+      "step": 7989
+    },
+    {
+      "epoch": 2.565419810563493,
+      "grad_norm": 0.9454381465911865,
+      "learning_rate": 2.3642932513055884e-05,
+      "loss": 0.4163,
+      "step": 7990
+    },
+    {
+      "epoch": 2.565740889388345,
+      "grad_norm": 0.889689564704895,
+      "learning_rate": 2.362039713653581e-05,
+      "loss": 0.4446,
+      "step": 7991
+    },
+    {
+      "epoch": 2.566061968213196,
+      "grad_norm": 0.7542763948440552,
+      "learning_rate": 2.359787106663861e-05,
+      "loss": 0.3018,
+      "step": 7992
+    },
+    {
+      "epoch": 2.5663830470380478,
+      "grad_norm": 0.7268791198730469,
+      "learning_rate": 2.35753543061091e-05,
+      "loss": 0.3345,
+      "step": 7993
+    },
+    {
+      "epoch": 2.5667041258628993,
+      "grad_norm": 0.6492055654525757,
+      "learning_rate": 2.3552846857690846e-05,
+      "loss": 0.3103,
+      "step": 7994
+    },
+    {
+      "epoch": 2.567025204687751,
+      "grad_norm": 0.7166942358016968,
+      "learning_rate": 2.3530348724126307e-05,
+      "loss": 0.2754,
+      "step": 7995
+    },
+    {
+      "epoch": 2.5673462835126024,
+      "grad_norm": 0.9371656179428101,
+      "learning_rate": 2.3507859908156827e-05,
+      "loss": 0.5017,
+      "step": 7996
+    },
+    {
+      "epoch": 2.567667362337454,
+      "grad_norm": 1.0456211566925049,
+      "learning_rate": 2.3485380412522585e-05,
+      "loss": 0.3592,
+      "step": 7997
+    },
+    {
+      "epoch": 2.5679884411623055,
+      "grad_norm": 0.6862274408340454,
+      "learning_rate": 2.3462910239962654e-05,
+      "loss": 0.2818,
+      "step": 7998
+    },
+    {
+      "epoch": 2.5683095199871566,
+      "grad_norm": 1.1700456142425537,
+      "learning_rate": 2.3440449393214948e-05,
+      "loss": 0.3444,
+      "step": 7999
+    },
+    {
+      "epoch": 2.5686305988120086,
+      "grad_norm": 0.9991983771324158,
+      "learning_rate": 2.341799787501625e-05,
+      "loss": 0.4639,
+      "step": 8000
+    },
+    {
+      "epoch": 2.5689516776368597,
+      "grad_norm": 0.9430015683174133,
+      "learning_rate": 2.339555568810221e-05,
+      "loss": 0.4327,
+      "step": 8001
+    },
+    {
+      "epoch": 2.5692727564617113,
+      "grad_norm": 0.7065159678459167,
+      "learning_rate": 2.337312283520735e-05,
+      "loss": 0.2867,
+      "step": 8002
+    },
+    {
+      "epoch": 2.569593835286563,
+      "grad_norm": 0.9828401803970337,
+      "learning_rate": 2.3350699319065026e-05,
+      "loss": 0.3369,
+      "step": 8003
+    },
+    {
+      "epoch": 2.5699149141114144,
+      "grad_norm": 2.36687970161438,
+      "learning_rate": 2.3328285142407503e-05,
+      "loss": 0.3983,
+      "step": 8004
+    },
+    {
+      "epoch": 2.570235992936266,
+      "grad_norm": 0.9053291082382202,
+      "learning_rate": 2.3305880307965833e-05,
+      "loss": 0.3354,
+      "step": 8005
+    },
+    {
+      "epoch": 2.5705570717611175,
+      "grad_norm": 1.2159992456436157,
+      "learning_rate": 2.328348481847006e-05,
+      "loss": 0.5254,
+      "step": 8006
+    },
+    {
+      "epoch": 2.570878150585969,
+      "grad_norm": 0.8332436680793762,
+      "learning_rate": 2.3261098676648907e-05,
+      "loss": 0.324,
+      "step": 8007
+    },
+    {
+      "epoch": 2.57119922941082,
+      "grad_norm": 0.9708757400512695,
+      "learning_rate": 2.323872188523013e-05,
+      "loss": 0.3851,
+      "step": 8008
+    },
+    {
+      "epoch": 2.571520308235672,
+      "grad_norm": 1.2236950397491455,
+      "learning_rate": 2.321635444694028e-05,
+      "loss": 0.3511,
+      "step": 8009
+    },
+    {
+      "epoch": 2.5718413870605232,
+      "grad_norm": 0.8926087617874146,
+      "learning_rate": 2.319399636450468e-05,
+      "loss": 0.3669,
+      "step": 8010
+    },
+    {
+      "epoch": 2.572162465885375,
+      "grad_norm": 0.9284133315086365,
+      "learning_rate": 2.3171647640647687e-05,
+      "loss": 0.348,
+      "step": 8011
+    },
+    {
+      "epoch": 2.5724835447102263,
+      "grad_norm": 0.856762170791626,
+      "learning_rate": 2.3149308278092342e-05,
+      "loss": 0.3501,
+      "step": 8012
+    },
+    {
+      "epoch": 2.572804623535078,
+      "grad_norm": 0.7136096358299255,
+      "learning_rate": 2.3126978279560684e-05,
+      "loss": 0.3205,
+      "step": 8013
+    },
+    {
+      "epoch": 2.5731257023599294,
+      "grad_norm": 0.570107102394104,
+      "learning_rate": 2.3104657647773554e-05,
+      "loss": 0.2764,
+      "step": 8014
+    },
+    {
+      "epoch": 2.573446781184781,
+      "grad_norm": 0.7570350766181946,
+      "learning_rate": 2.3082346385450637e-05,
+      "loss": 0.3118,
+      "step": 8015
+    },
+    {
+      "epoch": 2.5737678600096325,
+      "grad_norm": 0.9262259006500244,
+      "learning_rate": 2.3060044495310505e-05,
+      "loss": 0.4234,
+      "step": 8016
+    },
+    {
+      "epoch": 2.5740889388344836,
+      "grad_norm": 0.3663389980792999,
+      "learning_rate": 2.3037751980070555e-05,
+      "loss": 0.2028,
+      "step": 8017
+    },
+    {
+      "epoch": 2.5744100176593356,
+      "grad_norm": 0.8820987343788147,
+      "learning_rate": 2.3015468842447086e-05,
+      "loss": 0.3276,
+      "step": 8018
+    },
+    {
+      "epoch": 2.5747310964841867,
+      "grad_norm": 0.7441263198852539,
+      "learning_rate": 2.2993195085155205e-05,
+      "loss": 0.2819,
+      "step": 8019
+    },
+    {
+      "epoch": 2.5750521753090383,
+      "grad_norm": 0.5265788435935974,
+      "learning_rate": 2.2970930710908935e-05,
+      "loss": 0.2229,
+      "step": 8020
+    },
+    {
+      "epoch": 2.57537325413389,
+      "grad_norm": 0.9287590980529785,
+      "learning_rate": 2.2948675722421086e-05,
+      "loss": 0.3445,
+      "step": 8021
+    },
+    {
+      "epoch": 2.5756943329587414,
+      "grad_norm": 0.8301019072532654,
+      "learning_rate": 2.2926430122403386e-05,
+      "loss": 0.3109,
+      "step": 8022
+    },
+    {
+      "epoch": 2.576015411783593,
+      "grad_norm": 0.7834892868995667,
+      "learning_rate": 2.2904193913566363e-05,
+      "loss": 0.35,
+      "step": 8023
+    },
+    {
+      "epoch": 2.5763364906084445,
+      "grad_norm": 0.5827646851539612,
+      "learning_rate": 2.2881967098619506e-05,
+      "loss": 0.256,
+      "step": 8024
+    },
+    {
+      "epoch": 2.576657569433296,
+      "grad_norm": 0.6390456557273865,
+      "learning_rate": 2.2859749680270982e-05,
+      "loss": 0.272,
+      "step": 8025
+    },
+    {
+      "epoch": 2.576978648258147,
+      "grad_norm": 0.9293005466461182,
+      "learning_rate": 2.2837541661228025e-05,
+      "loss": 0.3369,
+      "step": 8026
+    },
+    {
+      "epoch": 2.577299727082999,
+      "grad_norm": 0.8350579738616943,
+      "learning_rate": 2.2815343044196525e-05,
+      "loss": 0.3416,
+      "step": 8027
+    },
+    {
+      "epoch": 2.5776208059078503,
+      "grad_norm": 0.4396490156650543,
+      "learning_rate": 2.279315383188132e-05,
+      "loss": 0.2917,
+      "step": 8028
+    },
+    {
+      "epoch": 2.577941884732702,
+      "grad_norm": 0.3990170359611511,
+      "learning_rate": 2.277097402698619e-05,
+      "loss": 0.2969,
+      "step": 8029
+    },
+    {
+      "epoch": 2.5782629635575534,
+      "grad_norm": 0.5404382348060608,
+      "learning_rate": 2.2748803632213557e-05,
+      "loss": 0.8456,
+      "step": 8030
+    },
+    {
+      "epoch": 2.578584042382405,
+      "grad_norm": 0.5150883793830872,
+      "learning_rate": 2.2726642650264895e-05,
+      "loss": 0.8014,
+      "step": 8031
+    },
+    {
+      "epoch": 2.5789051212072565,
+      "grad_norm": 0.4833203852176666,
+      "learning_rate": 2.270449108384044e-05,
+      "loss": 0.5567,
+      "step": 8032
+    },
+    {
+      "epoch": 2.579226200032108,
+      "grad_norm": 0.5803343057632446,
+      "learning_rate": 2.2682348935639274e-05,
+      "loss": 0.1951,
+      "step": 8033
+    },
+    {
+      "epoch": 2.5795472788569596,
+      "grad_norm": 0.43384525179862976,
+      "learning_rate": 2.2660216208359365e-05,
+      "loss": 0.2521,
+      "step": 8034
+    },
+    {
+      "epoch": 2.5798683576818107,
+      "grad_norm": 0.3641275465488434,
+      "learning_rate": 2.2638092904697518e-05,
+      "loss": 0.1943,
+      "step": 8035
+    },
+    {
+      "epoch": 2.5801894365066627,
+      "grad_norm": 0.3974734842777252,
+      "learning_rate": 2.2615979027349387e-05,
+      "loss": 0.1705,
+      "step": 8036
+    },
+    {
+      "epoch": 2.5805105153315138,
+      "grad_norm": 0.3340223431587219,
+      "learning_rate": 2.259387457900948e-05,
+      "loss": 0.1161,
+      "step": 8037
+    },
+    {
+      "epoch": 2.5808315941563653,
+      "grad_norm": 0.750033974647522,
+      "learning_rate": 2.2571779562371153e-05,
+      "loss": 0.3098,
+      "step": 8038
+    },
+    {
+      "epoch": 2.581152672981217,
+      "grad_norm": 0.7684246301651001,
+      "learning_rate": 2.254969398012663e-05,
+      "loss": 0.3886,
+      "step": 8039
+    },
+    {
+      "epoch": 2.5814737518060684,
+      "grad_norm": 0.8656273484230042,
+      "learning_rate": 2.2527617834966954e-05,
+      "loss": 0.3571,
+      "step": 8040
+    },
+    {
+      "epoch": 2.58179483063092,
+      "grad_norm": 0.7932442426681519,
+      "learning_rate": 2.2505551129582047e-05,
+      "loss": 0.3677,
+      "step": 8041
+    },
+    {
+      "epoch": 2.5821159094557715,
+      "grad_norm": 0.7484654188156128,
+      "learning_rate": 2.2483493866660676e-05,
+      "loss": 0.3368,
+      "step": 8042
+    },
+    {
+      "epoch": 2.582436988280623,
+      "grad_norm": 0.9018601179122925,
+      "learning_rate": 2.246144604889042e-05,
+      "loss": 0.4161,
+      "step": 8043
+    },
+    {
+      "epoch": 2.582758067105474,
+      "grad_norm": 0.7390075325965881,
+      "learning_rate": 2.2439407678957812e-05,
+      "loss": 0.2773,
+      "step": 8044
+    },
+    {
+      "epoch": 2.583079145930326,
+      "grad_norm": 0.7586901783943176,
+      "learning_rate": 2.241737875954808e-05,
+      "loss": 0.2567,
+      "step": 8045
+    },
+    {
+      "epoch": 2.5834002247551773,
+      "grad_norm": 1.0482172966003418,
+      "learning_rate": 2.2395359293345396e-05,
+      "loss": 0.3444,
+      "step": 8046
+    },
+    {
+      "epoch": 2.583721303580029,
+      "grad_norm": 0.7774665951728821,
+      "learning_rate": 2.237334928303283e-05,
+      "loss": 0.2983,
+      "step": 8047
+    },
+    {
+      "epoch": 2.5840423824048804,
+      "grad_norm": 0.8551390767097473,
+      "learning_rate": 2.235134873129213e-05,
+      "loss": 0.325,
+      "step": 8048
+    },
+    {
+      "epoch": 2.584363461229732,
+      "grad_norm": 0.8655200600624084,
+      "learning_rate": 2.2329357640804117e-05,
+      "loss": 0.3297,
+      "step": 8049
+    },
+    {
+      "epoch": 2.5846845400545835,
+      "grad_norm": 0.7637379765510559,
+      "learning_rate": 2.2307376014248216e-05,
+      "loss": 0.2911,
+      "step": 8050
+    },
+    {
+      "epoch": 2.585005618879435,
+      "grad_norm": 0.9776454567909241,
+      "learning_rate": 2.2285403854302912e-05,
+      "loss": 0.3879,
+      "step": 8051
+    },
+    {
+      "epoch": 2.5853266977042866,
+      "grad_norm": 1.0411674976348877,
+      "learning_rate": 2.2263441163645403e-05,
+      "loss": 0.4912,
+      "step": 8052
+    },
+    {
+      "epoch": 2.5856477765291377,
+      "grad_norm": 1.0440319776535034,
+      "learning_rate": 2.22414879449518e-05,
+      "loss": 0.4109,
+      "step": 8053
+    },
+    {
+      "epoch": 2.5859688553539897,
+      "grad_norm": 0.7270945310592651,
+      "learning_rate": 2.2219544200897025e-05,
+      "loss": 0.3516,
+      "step": 8054
+    },
+    {
+      "epoch": 2.586289934178841,
+      "grad_norm": 0.867685079574585,
+      "learning_rate": 2.219760993415485e-05,
+      "loss": 0.3819,
+      "step": 8055
+    },
+    {
+      "epoch": 2.5866110130036923,
+      "grad_norm": 0.6216245293617249,
+      "learning_rate": 2.2175685147397906e-05,
+      "loss": 0.2304,
+      "step": 8056
+    },
+    {
+      "epoch": 2.586932091828544,
+      "grad_norm": 0.9045018553733826,
+      "learning_rate": 2.2153769843297667e-05,
+      "loss": 0.3466,
+      "step": 8057
+    },
+    {
+      "epoch": 2.5872531706533954,
+      "grad_norm": 1.0591224431991577,
+      "learning_rate": 2.213186402452443e-05,
+      "loss": 0.4561,
+      "step": 8058
+    },
+    {
+      "epoch": 2.587574249478247,
+      "grad_norm": 0.8752357959747314,
+      "learning_rate": 2.210996769374737e-05,
+      "loss": 0.3874,
+      "step": 8059
+    },
+    {
+      "epoch": 2.5878953283030985,
+      "grad_norm": 0.8251414895057678,
+      "learning_rate": 2.2088080853634473e-05,
+      "loss": 0.3303,
+      "step": 8060
+    },
+    {
+      "epoch": 2.58821640712795,
+      "grad_norm": 1.0302598476409912,
+      "learning_rate": 2.2066203506852566e-05,
+      "loss": 0.3676,
+      "step": 8061
+    },
+    {
+      "epoch": 2.588537485952801,
+      "grad_norm": 0.9010094404220581,
+      "learning_rate": 2.204433565606743e-05,
+      "loss": 0.3819,
+      "step": 8062
+    },
+    {
+      "epoch": 2.5888585647776527,
+      "grad_norm": 0.7407838702201843,
+      "learning_rate": 2.202247730394349e-05,
+      "loss": 0.2689,
+      "step": 8063
+    },
+    {
+      "epoch": 2.5891796436025043,
+      "grad_norm": 1.1048128604888916,
+      "learning_rate": 2.200062845314417e-05,
+      "loss": 0.3919,
+      "step": 8064
+    },
+    {
+      "epoch": 2.589500722427356,
+      "grad_norm": 0.9192834496498108,
+      "learning_rate": 2.1978789106331665e-05,
+      "loss": 0.2973,
+      "step": 8065
+    },
+    {
+      "epoch": 2.5898218012522074,
+      "grad_norm": 0.882653534412384,
+      "learning_rate": 2.195695926616702e-05,
+      "loss": 0.279,
+      "step": 8066
+    },
+    {
+      "epoch": 2.590142880077059,
+      "grad_norm": 0.9665889143943787,
+      "learning_rate": 2.1935138935310206e-05,
+      "loss": 0.3333,
+      "step": 8067
+    },
+    {
+      "epoch": 2.5904639589019105,
+      "grad_norm": 0.8618683218955994,
+      "learning_rate": 2.1913328116419873e-05,
+      "loss": 0.3014,
+      "step": 8068
+    },
+    {
+      "epoch": 2.590785037726762,
+      "grad_norm": 0.8841969966888428,
+      "learning_rate": 2.1891526812153672e-05,
+      "loss": 0.3502,
+      "step": 8069
+    },
+    {
+      "epoch": 2.5911061165516136,
+      "grad_norm": 1.0003149509429932,
+      "learning_rate": 2.1869735025168026e-05,
+      "loss": 0.333,
+      "step": 8070
+    },
+    {
+      "epoch": 2.5914271953764647,
+      "grad_norm": 0.9237295985221863,
+      "learning_rate": 2.1847952758118117e-05,
+      "loss": 0.3748,
+      "step": 8071
+    },
+    {
+      "epoch": 2.5917482742013163,
+      "grad_norm": 0.9068896770477295,
+      "learning_rate": 2.1826180013658172e-05,
+      "loss": 0.2793,
+      "step": 8072
+    },
+    {
+      "epoch": 2.592069353026168,
+      "grad_norm": 1.0544371604919434,
+      "learning_rate": 2.1804416794440995e-05,
+      "loss": 0.3318,
+      "step": 8073
+    },
+    {
+      "epoch": 2.5923904318510194,
+      "grad_norm": 1.1038298606872559,
+      "learning_rate": 2.178266310311847e-05,
+      "loss": 0.4288,
+      "step": 8074
+    },
+    {
+      "epoch": 2.592711510675871,
+      "grad_norm": 0.530205488204956,
+      "learning_rate": 2.1760918942341192e-05,
+      "loss": 0.2411,
+      "step": 8075
+    },
+    {
+      "epoch": 2.5930325895007225,
+      "grad_norm": 0.5433615446090698,
+      "learning_rate": 2.173918431475861e-05,
+      "loss": 0.2607,
+      "step": 8076
+    },
+    {
+      "epoch": 2.593353668325574,
+      "grad_norm": 0.8718168139457703,
+      "learning_rate": 2.1717459223019033e-05,
+      "loss": 0.3014,
+      "step": 8077
+    },
+    {
+      "epoch": 2.5936747471504256,
+      "grad_norm": 0.506506085395813,
+      "learning_rate": 2.1695743669769596e-05,
+      "loss": 0.268,
+      "step": 8078
+    },
+    {
+      "epoch": 2.593995825975277,
+      "grad_norm": 0.8781798481941223,
+      "learning_rate": 2.1674037657656266e-05,
+      "loss": 0.3089,
+      "step": 8079
+    },
+    {
+      "epoch": 2.594316904800128,
+      "grad_norm": 0.5046657919883728,
+      "learning_rate": 2.1652341189323866e-05,
+      "loss": 0.7222,
+      "step": 8080
+    },
+    {
+      "epoch": 2.5946379836249798,
+      "grad_norm": 0.40744638442993164,
+      "learning_rate": 2.163065426741603e-05,
+      "loss": 0.4146,
+      "step": 8081
+    },
+    {
+      "epoch": 2.5949590624498313,
+      "grad_norm": 0.49407830834388733,
+      "learning_rate": 2.160897689457526e-05,
+      "loss": 0.3484,
+      "step": 8082
+    },
+    {
+      "epoch": 2.595280141274683,
+      "grad_norm": 0.34928953647613525,
+      "learning_rate": 2.1587309073442863e-05,
+      "loss": 0.0671,
+      "step": 8083
+    },
+    {
+      "epoch": 2.5956012200995344,
+      "grad_norm": 0.22757402062416077,
+      "learning_rate": 2.1565650806658975e-05,
+      "loss": 0.0641,
+      "step": 8084
+    },
+    {
+      "epoch": 2.595922298924386,
+      "grad_norm": 0.3532765805721283,
+      "learning_rate": 2.154400209686268e-05,
+      "loss": 0.1308,
+      "step": 8085
+    },
+    {
+      "epoch": 2.5962433777492375,
+      "grad_norm": 0.7194594144821167,
+      "learning_rate": 2.1522362946691698e-05,
+      "loss": 0.347,
+      "step": 8086
+    },
+    {
+      "epoch": 2.596564456574089,
+      "grad_norm": 0.8896145224571228,
+      "learning_rate": 2.1500733358782786e-05,
+      "loss": 0.4766,
+      "step": 8087
+    },
+    {
+      "epoch": 2.5968855353989406,
+      "grad_norm": 1.069779872894287,
+      "learning_rate": 2.1479113335771383e-05,
+      "loss": 0.4661,
+      "step": 8088
+    },
+    {
+      "epoch": 2.5972066142237917,
+      "grad_norm": 0.7583945989608765,
+      "learning_rate": 2.1457502880291812e-05,
+      "loss": 0.3911,
+      "step": 8089
+    },
+    {
+      "epoch": 2.5975276930486433,
+      "grad_norm": 0.6904090046882629,
+      "learning_rate": 2.1435901994977326e-05,
+      "loss": 0.3297,
+      "step": 8090
+    },
+    {
+      "epoch": 2.597848771873495,
+      "grad_norm": 0.7934258580207825,
+      "learning_rate": 2.1414310682459802e-05,
+      "loss": 0.3471,
+      "step": 8091
+    },
+    {
+      "epoch": 2.5981698506983464,
+      "grad_norm": 1.1645605564117432,
+      "learning_rate": 2.1392728945370222e-05,
+      "loss": 0.365,
+      "step": 8092
+    },
+    {
+      "epoch": 2.598490929523198,
+      "grad_norm": 0.57047039270401,
+      "learning_rate": 2.137115678633811e-05,
+      "loss": 0.2229,
+      "step": 8093
+    },
+    {
+      "epoch": 2.5988120083480495,
+      "grad_norm": 0.8101687431335449,
+      "learning_rate": 2.1349594207992064e-05,
+      "loss": 0.3051,
+      "step": 8094
+    },
+    {
+      "epoch": 2.599133087172901,
+      "grad_norm": 0.785053551197052,
+      "learning_rate": 2.1328041212959403e-05,
+      "loss": 0.3114,
+      "step": 8095
+    },
+    {
+      "epoch": 2.5994541659977526,
+      "grad_norm": 0.7656204104423523,
+      "learning_rate": 2.1306497803866277e-05,
+      "loss": 0.3462,
+      "step": 8096
+    },
+    {
+      "epoch": 2.599775244822604,
+      "grad_norm": 0.861936092376709,
+      "learning_rate": 2.128496398333768e-05,
+      "loss": 0.356,
+      "step": 8097
+    },
+    {
+      "epoch": 2.6000963236474552,
+      "grad_norm": 0.8918510675430298,
+      "learning_rate": 2.126343975399747e-05,
+      "loss": 0.3629,
+      "step": 8098
+    },
+    {
+      "epoch": 2.600417402472307,
+      "grad_norm": 0.8870481848716736,
+      "learning_rate": 2.1241925118468287e-05,
+      "loss": 0.3867,
+      "step": 8099
+    },
+    {
+      "epoch": 2.6007384812971583,
+      "grad_norm": 1.0247498750686646,
+      "learning_rate": 2.1220420079371628e-05,
+      "loss": 0.3675,
+      "step": 8100
+    },
+    {
+      "epoch": 2.60105956012201,
+      "grad_norm": 0.9953125715255737,
+      "learning_rate": 2.119892463932781e-05,
+      "loss": 0.3568,
+      "step": 8101
+    },
+    {
+      "epoch": 2.6013806389468614,
+      "grad_norm": 0.9158168435096741,
+      "learning_rate": 2.1177438800956007e-05,
+      "loss": 0.3369,
+      "step": 8102
+    },
+    {
+      "epoch": 2.601701717771713,
+      "grad_norm": 0.8674771785736084,
+      "learning_rate": 2.115596256687419e-05,
+      "loss": 0.3141,
+      "step": 8103
+    },
+    {
+      "epoch": 2.6020227965965645,
+      "grad_norm": 0.6369979381561279,
+      "learning_rate": 2.113449593969915e-05,
+      "loss": 0.2762,
+      "step": 8104
+    },
+    {
+      "epoch": 2.602343875421416,
+      "grad_norm": 0.8605027794837952,
+      "learning_rate": 2.1113038922046602e-05,
+      "loss": 0.3994,
+      "step": 8105
+    },
+    {
+      "epoch": 2.6026649542462676,
+      "grad_norm": 1.133817434310913,
+      "learning_rate": 2.1091591516530952e-05,
+      "loss": 0.3732,
+      "step": 8106
+    },
+    {
+      "epoch": 2.6029860330711188,
+      "grad_norm": 0.7040675282478333,
+      "learning_rate": 2.107015372576552e-05,
+      "loss": 0.3186,
+      "step": 8107
+    },
+    {
+      "epoch": 2.6033071118959703,
+      "grad_norm": 0.7996187210083008,
+      "learning_rate": 2.1048725552362435e-05,
+      "loss": 0.3545,
+      "step": 8108
+    },
+    {
+      "epoch": 2.603628190720822,
+      "grad_norm": 1.0517889261245728,
+      "learning_rate": 2.1027306998932627e-05,
+      "loss": 0.3864,
+      "step": 8109
+    },
+    {
+      "epoch": 2.6039492695456734,
+      "grad_norm": 0.5958194136619568,
+      "learning_rate": 2.100589806808597e-05,
+      "loss": 0.2543,
+      "step": 8110
+    },
+    {
+      "epoch": 2.604270348370525,
+      "grad_norm": 0.7594183683395386,
+      "learning_rate": 2.098449876243096e-05,
+      "loss": 0.328,
+      "step": 8111
+    },
+    {
+      "epoch": 2.6045914271953765,
+      "grad_norm": 1.2153469324111938,
+      "learning_rate": 2.096310908457513e-05,
+      "loss": 0.3756,
+      "step": 8112
+    },
+    {
+      "epoch": 2.604912506020228,
+      "grad_norm": 0.8676589727401733,
+      "learning_rate": 2.09417290371247e-05,
+      "loss": 0.3697,
+      "step": 8113
+    },
+    {
+      "epoch": 2.6052335848450796,
+      "grad_norm": 0.792874813079834,
+      "learning_rate": 2.0920358622684788e-05,
+      "loss": 0.379,
+      "step": 8114
+    },
+    {
+      "epoch": 2.605554663669931,
+      "grad_norm": 0.6348553895950317,
+      "learning_rate": 2.0898997843859334e-05,
+      "loss": 0.2493,
+      "step": 8115
+    },
+    {
+      "epoch": 2.6058757424947823,
+      "grad_norm": 1.1245969533920288,
+      "learning_rate": 2.0877646703250996e-05,
+      "loss": 0.3336,
+      "step": 8116
+    },
+    {
+      "epoch": 2.606196821319634,
+      "grad_norm": 0.794979453086853,
+      "learning_rate": 2.0856305203461436e-05,
+      "loss": 0.3597,
+      "step": 8117
+    },
+    {
+      "epoch": 2.6065179001444854,
+      "grad_norm": 0.9265166521072388,
+      "learning_rate": 2.0834973347091014e-05,
+      "loss": 0.336,
+      "step": 8118
+    },
+    {
+      "epoch": 2.606838978969337,
+      "grad_norm": 0.7240932583808899,
+      "learning_rate": 2.0813651136738954e-05,
+      "loss": 0.284,
+      "step": 8119
+    },
+    {
+      "epoch": 2.6071600577941885,
+      "grad_norm": 0.5318633913993835,
+      "learning_rate": 2.07923385750033e-05,
+      "loss": 0.2458,
+      "step": 8120
+    },
+    {
+      "epoch": 2.60748113661904,
+      "grad_norm": 0.6563091278076172,
+      "learning_rate": 2.0771035664480942e-05,
+      "loss": 0.2628,
+      "step": 8121
+    },
+    {
+      "epoch": 2.6078022154438916,
+      "grad_norm": 0.9432152509689331,
+      "learning_rate": 2.0749742407767546e-05,
+      "loss": 0.3554,
+      "step": 8122
+    },
+    {
+      "epoch": 2.608123294268743,
+      "grad_norm": 1.121485948562622,
+      "learning_rate": 2.0728458807457662e-05,
+      "loss": 0.3306,
+      "step": 8123
+    },
+    {
+      "epoch": 2.6084443730935947,
+      "grad_norm": 0.8368777632713318,
+      "learning_rate": 2.0707184866144603e-05,
+      "loss": 0.2796,
+      "step": 8124
+    },
+    {
+      "epoch": 2.6087654519184458,
+      "grad_norm": 0.9011041522026062,
+      "learning_rate": 2.068592058642056e-05,
+      "loss": 0.3304,
+      "step": 8125
+    },
+    {
+      "epoch": 2.6090865307432973,
+      "grad_norm": 1.134626865386963,
+      "learning_rate": 2.0664665970876496e-05,
+      "loss": 0.2825,
+      "step": 8126
+    },
+    {
+      "epoch": 2.609407609568149,
+      "grad_norm": 0.7329688668251038,
+      "learning_rate": 2.0643421022102217e-05,
+      "loss": 0.3027,
+      "step": 8127
+    },
+    {
+      "epoch": 2.6097286883930004,
+      "grad_norm": 0.5429527759552002,
+      "learning_rate": 2.0622185742686416e-05,
+      "loss": 0.3091,
+      "step": 8128
+    },
+    {
+      "epoch": 2.610049767217852,
+      "grad_norm": 0.6699220538139343,
+      "learning_rate": 2.0600960135216462e-05,
+      "loss": 0.296,
+      "step": 8129
+    },
+    {
+      "epoch": 2.6103708460427035,
+      "grad_norm": 0.5311281681060791,
+      "learning_rate": 2.0579744202278718e-05,
+      "loss": 0.7533,
+      "step": 8130
+    },
+    {
+      "epoch": 2.610691924867555,
+      "grad_norm": 0.5676338076591492,
+      "learning_rate": 2.0558537946458177e-05,
+      "loss": 0.8681,
+      "step": 8131
+    },
+    {
+      "epoch": 2.6110130036924066,
+      "grad_norm": 0.38879385590553284,
+      "learning_rate": 2.0537341370338857e-05,
+      "loss": 0.1332,
+      "step": 8132
+    },
+    {
+      "epoch": 2.611334082517258,
+      "grad_norm": 0.28156155347824097,
+      "learning_rate": 2.051615447650347e-05,
+      "loss": 0.1392,
+      "step": 8133
+    },
+    {
+      "epoch": 2.6116551613421093,
+      "grad_norm": 0.38800543546676636,
+      "learning_rate": 2.049497726753351e-05,
+      "loss": 0.156,
+      "step": 8134
+    },
+    {
+      "epoch": 2.611976240166961,
+      "grad_norm": 0.2520560026168823,
+      "learning_rate": 2.0473809746009444e-05,
+      "loss": 0.0688,
+      "step": 8135
+    },
+    {
+      "epoch": 2.6122973189918124,
+      "grad_norm": 0.7137389183044434,
+      "learning_rate": 2.045265191451041e-05,
+      "loss": 0.3956,
+      "step": 8136
+    },
+    {
+      "epoch": 2.612618397816664,
+      "grad_norm": 0.8318041563034058,
+      "learning_rate": 2.0431503775614457e-05,
+      "loss": 0.2772,
+      "step": 8137
+    },
+    {
+      "epoch": 2.6129394766415155,
+      "grad_norm": 0.7499494552612305,
+      "learning_rate": 2.0410365331898416e-05,
+      "loss": 0.3192,
+      "step": 8138
+    },
+    {
+      "epoch": 2.613260555466367,
+      "grad_norm": 0.8148148059844971,
+      "learning_rate": 2.0389236585937945e-05,
+      "loss": 0.3618,
+      "step": 8139
+    },
+    {
+      "epoch": 2.6135816342912186,
+      "grad_norm": 1.1199220418930054,
+      "learning_rate": 2.0368117540307496e-05,
+      "loss": 0.3034,
+      "step": 8140
+    },
+    {
+      "epoch": 2.61390271311607,
+      "grad_norm": 0.5880412459373474,
+      "learning_rate": 2.0347008197580374e-05,
+      "loss": 0.2263,
+      "step": 8141
+    },
+    {
+      "epoch": 2.6142237919409217,
+      "grad_norm": 0.8377451300621033,
+      "learning_rate": 2.03259085603287e-05,
+      "loss": 0.4124,
+      "step": 8142
+    },
+    {
+      "epoch": 2.614544870765773,
+      "grad_norm": 0.7712952494621277,
+      "learning_rate": 2.030481863112339e-05,
+      "loss": 0.2779,
+      "step": 8143
+    },
+    {
+      "epoch": 2.6148659495906243,
+      "grad_norm": 0.8210815787315369,
+      "learning_rate": 2.028373841253419e-05,
+      "loss": 0.3206,
+      "step": 8144
+    },
+    {
+      "epoch": 2.615187028415476,
+      "grad_norm": 1.114625334739685,
+      "learning_rate": 2.026266790712965e-05,
+      "loss": 0.4125,
+      "step": 8145
+    },
+    {
+      "epoch": 2.6155081072403275,
+      "grad_norm": 1.0668013095855713,
+      "learning_rate": 2.024160711747717e-05,
+      "loss": 0.4166,
+      "step": 8146
+    },
+    {
+      "epoch": 2.615829186065179,
+      "grad_norm": 1.219161868095398,
+      "learning_rate": 2.0220556046142893e-05,
+      "loss": 0.4567,
+      "step": 8147
+    },
+    {
+      "epoch": 2.6161502648900306,
+      "grad_norm": 0.993944525718689,
+      "learning_rate": 2.019951469569191e-05,
+      "loss": 0.3649,
+      "step": 8148
+    },
+    {
+      "epoch": 2.616471343714882,
+      "grad_norm": 0.8392919898033142,
+      "learning_rate": 2.017848306868797e-05,
+      "loss": 0.2713,
+      "step": 8149
+    },
+    {
+      "epoch": 2.6167924225397337,
+      "grad_norm": 0.9957613945007324,
+      "learning_rate": 2.0157461167693758e-05,
+      "loss": 0.3948,
+      "step": 8150
+    },
+    {
+      "epoch": 2.617113501364585,
+      "grad_norm": 0.8635070323944092,
+      "learning_rate": 2.013644899527074e-05,
+      "loss": 0.3869,
+      "step": 8151
+    },
+    {
+      "epoch": 2.6174345801894363,
+      "grad_norm": 1.053145170211792,
+      "learning_rate": 2.01154465539791e-05,
+      "loss": 0.3391,
+      "step": 8152
+    },
+    {
+      "epoch": 2.617755659014288,
+      "grad_norm": 0.6855107545852661,
+      "learning_rate": 2.009445384637805e-05,
+      "loss": 0.2861,
+      "step": 8153
+    },
+    {
+      "epoch": 2.6180767378391394,
+      "grad_norm": 1.1673146486282349,
+      "learning_rate": 2.0073470875025358e-05,
+      "loss": 0.3544,
+      "step": 8154
+    },
+    {
+      "epoch": 2.618397816663991,
+      "grad_norm": 0.7385457158088684,
+      "learning_rate": 2.005249764247783e-05,
+      "loss": 0.2765,
+      "step": 8155
+    },
+    {
+      "epoch": 2.6187188954888425,
+      "grad_norm": 0.721240222454071,
+      "learning_rate": 2.0031534151290943e-05,
+      "loss": 0.2583,
+      "step": 8156
+    },
+    {
+      "epoch": 2.619039974313694,
+      "grad_norm": 0.8315588235855103,
+      "learning_rate": 2.0010580404019062e-05,
+      "loss": 0.3844,
+      "step": 8157
+    },
+    {
+      "epoch": 2.6193610531385456,
+      "grad_norm": 0.9637236595153809,
+      "learning_rate": 1.9989636403215328e-05,
+      "loss": 0.3219,
+      "step": 8158
+    },
+    {
+      "epoch": 2.619682131963397,
+      "grad_norm": 0.7627900242805481,
+      "learning_rate": 1.9968702151431696e-05,
+      "loss": 0.2788,
+      "step": 8159
+    },
+    {
+      "epoch": 2.6200032107882487,
+      "grad_norm": 0.6353088617324829,
+      "learning_rate": 1.9947777651218946e-05,
+      "loss": 0.256,
+      "step": 8160
+    },
+    {
+      "epoch": 2.6203242896131,
+      "grad_norm": 1.0363852977752686,
+      "learning_rate": 1.9926862905126665e-05,
+      "loss": 0.451,
+      "step": 8161
+    },
+    {
+      "epoch": 2.6206453684379514,
+      "grad_norm": 0.863524854183197,
+      "learning_rate": 1.9905957915703244e-05,
+      "loss": 0.3374,
+      "step": 8162
+    },
+    {
+      "epoch": 2.620966447262803,
+      "grad_norm": 0.8532664775848389,
+      "learning_rate": 1.9885062685495904e-05,
+      "loss": 0.3011,
+      "step": 8163
+    },
+    {
+      "epoch": 2.6212875260876545,
+      "grad_norm": 0.8648016452789307,
+      "learning_rate": 1.9864177217050674e-05,
+      "loss": 0.3211,
+      "step": 8164
+    },
+    {
+      "epoch": 2.621608604912506,
+      "grad_norm": 0.7424303889274597,
+      "learning_rate": 1.9843301512912327e-05,
+      "loss": 0.3291,
+      "step": 8165
+    },
+    {
+      "epoch": 2.6219296837373576,
+      "grad_norm": 0.7464106678962708,
+      "learning_rate": 1.9822435575624608e-05,
+      "loss": 0.3156,
+      "step": 8166
+    },
+    {
+      "epoch": 2.622250762562209,
+      "grad_norm": 1.0930415391921997,
+      "learning_rate": 1.9801579407729866e-05,
+      "loss": 0.5009,
+      "step": 8167
+    },
+    {
+      "epoch": 2.6225718413870607,
+      "grad_norm": 0.78467857837677,
+      "learning_rate": 1.9780733011769447e-05,
+      "loss": 0.2866,
+      "step": 8168
+    },
+    {
+      "epoch": 2.6228929202119122,
+      "grad_norm": 0.7698622345924377,
+      "learning_rate": 1.9759896390283362e-05,
+      "loss": 0.3075,
+      "step": 8169
+    },
+    {
+      "epoch": 2.6232139990367633,
+      "grad_norm": 0.5596573352813721,
+      "learning_rate": 1.9739069545810485e-05,
+      "loss": 0.2454,
+      "step": 8170
+    },
+    {
+      "epoch": 2.623535077861615,
+      "grad_norm": 0.7981199622154236,
+      "learning_rate": 1.9718252480888566e-05,
+      "loss": 0.2827,
+      "step": 8171
+    },
+    {
+      "epoch": 2.6238561566864664,
+      "grad_norm": 0.49611830711364746,
+      "learning_rate": 1.969744519805402e-05,
+      "loss": 0.2539,
+      "step": 8172
+    },
+    {
+      "epoch": 2.624177235511318,
+      "grad_norm": 0.9452134966850281,
+      "learning_rate": 1.9676647699842242e-05,
+      "loss": 0.3056,
+      "step": 8173
+    },
+    {
+      "epoch": 2.6244983143361695,
+      "grad_norm": 0.8190011382102966,
+      "learning_rate": 1.965585998878724e-05,
+      "loss": 0.304,
+      "step": 8174
+    },
+    {
+      "epoch": 2.624819393161021,
+      "grad_norm": 0.6110972762107849,
+      "learning_rate": 1.963508206742202e-05,
+      "loss": 0.2405,
+      "step": 8175
+    },
+    {
+      "epoch": 2.6251404719858726,
+      "grad_norm": 0.8863208293914795,
+      "learning_rate": 1.9614313938278272e-05,
+      "loss": 0.3248,
+      "step": 8176
+    },
+    {
+      "epoch": 2.625461550810724,
+      "grad_norm": 0.5325760841369629,
+      "learning_rate": 1.9593555603886538e-05,
+      "loss": 0.2599,
+      "step": 8177
+    },
+    {
+      "epoch": 2.6257826296355757,
+      "grad_norm": 0.9411792755126953,
+      "learning_rate": 1.9572807066776143e-05,
+      "loss": 0.3166,
+      "step": 8178
+    },
+    {
+      "epoch": 2.626103708460427,
+      "grad_norm": 0.5804592370986938,
+      "learning_rate": 1.955206832947526e-05,
+      "loss": 0.2814,
+      "step": 8179
+    },
+    {
+      "epoch": 2.6264247872852784,
+      "grad_norm": 0.5420529842376709,
+      "learning_rate": 1.9531339394510827e-05,
+      "loss": 0.7411,
+      "step": 8180
+    },
+    {
+      "epoch": 2.62674586611013,
+      "grad_norm": 0.46422919631004333,
+      "learning_rate": 1.9510620264408596e-05,
+      "loss": 0.6111,
+      "step": 8181
+    },
+    {
+      "epoch": 2.6270669449349815,
+      "grad_norm": 0.3627856969833374,
+      "learning_rate": 1.9489910941693133e-05,
+      "loss": 0.2278,
+      "step": 8182
+    },
+    {
+      "epoch": 2.627388023759833,
+      "grad_norm": 0.4854031503200531,
+      "learning_rate": 1.946921142888781e-05,
+      "loss": 0.1976,
+      "step": 8183
+    },
+    {
+      "epoch": 2.6277091025846846,
+      "grad_norm": 0.4990745186805725,
+      "learning_rate": 1.9448521728514802e-05,
+      "loss": 0.2803,
+      "step": 8184
+    },
+    {
+      "epoch": 2.628030181409536,
+      "grad_norm": 0.2989073395729065,
+      "learning_rate": 1.9427841843095063e-05,
+      "loss": 0.0711,
+      "step": 8185
+    },
+    {
+      "epoch": 2.6283512602343877,
+      "grad_norm": 0.5125943422317505,
+      "learning_rate": 1.9407171775148436e-05,
+      "loss": 0.2092,
+      "step": 8186
+    },
+    {
+      "epoch": 2.6286723390592392,
+      "grad_norm": 0.8719359040260315,
+      "learning_rate": 1.938651152719344e-05,
+      "loss": 0.4531,
+      "step": 8187
+    },
+    {
+      "epoch": 2.6289934178840904,
+      "grad_norm": 0.7034810781478882,
+      "learning_rate": 1.9365861101747485e-05,
+      "loss": 0.3558,
+      "step": 8188
+    },
+    {
+      "epoch": 2.629314496708942,
+      "grad_norm": 0.9010298848152161,
+      "learning_rate": 1.9345220501326777e-05,
+      "loss": 0.4044,
+      "step": 8189
+    },
+    {
+      "epoch": 2.6296355755337935,
+      "grad_norm": 0.8850868344306946,
+      "learning_rate": 1.9324589728446262e-05,
+      "loss": 0.3323,
+      "step": 8190
+    },
+    {
+      "epoch": 2.629956654358645,
+      "grad_norm": 0.8857529163360596,
+      "learning_rate": 1.930396878561983e-05,
+      "loss": 0.3986,
+      "step": 8191
+    },
+    {
+      "epoch": 2.6302777331834966,
+      "grad_norm": 0.7102099061012268,
+      "learning_rate": 1.928335767535997e-05,
+      "loss": 0.254,
+      "step": 8192
+    },
+    {
+      "epoch": 2.630598812008348,
+      "grad_norm": 0.7003388404846191,
+      "learning_rate": 1.9262756400178162e-05,
+      "loss": 0.314,
+      "step": 8193
+    },
+    {
+      "epoch": 2.6309198908331997,
+      "grad_norm": 0.9088608026504517,
+      "learning_rate": 1.9242164962584618e-05,
+      "loss": 0.3336,
+      "step": 8194
+    },
+    {
+      "epoch": 2.631240969658051,
+      "grad_norm": 0.8129791617393494,
+      "learning_rate": 1.922158336508825e-05,
+      "loss": 0.3411,
+      "step": 8195
+    },
+    {
+      "epoch": 2.6315620484829028,
+      "grad_norm": 0.9310645461082458,
+      "learning_rate": 1.9201011610196973e-05,
+      "loss": 0.4063,
+      "step": 8196
+    },
+    {
+      "epoch": 2.631883127307754,
+      "grad_norm": 1.2737817764282227,
+      "learning_rate": 1.918044970041729e-05,
+      "loss": 0.4643,
+      "step": 8197
+    },
+    {
+      "epoch": 2.6322042061326054,
+      "grad_norm": 0.9532968997955322,
+      "learning_rate": 1.91598976382547e-05,
+      "loss": 0.4322,
+      "step": 8198
+    },
+    {
+      "epoch": 2.632525284957457,
+      "grad_norm": 0.7169177532196045,
+      "learning_rate": 1.9139355426213347e-05,
+      "loss": 0.2881,
+      "step": 8199
+    },
+    {
+      "epoch": 2.6328463637823085,
+      "grad_norm": 1.0440733432769775,
+      "learning_rate": 1.9118823066796276e-05,
+      "loss": 0.4381,
+      "step": 8200
+    },
+    {
+      "epoch": 2.63316744260716,
+      "grad_norm": 0.9224132299423218,
+      "learning_rate": 1.9098300562505266e-05,
+      "loss": 0.3855,
+      "step": 8201
+    },
+    {
+      "epoch": 2.6334885214320116,
+      "grad_norm": 0.7816334962844849,
+      "learning_rate": 1.9077787915840927e-05,
+      "loss": 0.3378,
+      "step": 8202
+    },
+    {
+      "epoch": 2.633809600256863,
+      "grad_norm": 1.13314688205719,
+      "learning_rate": 1.9057285129302683e-05,
+      "loss": 0.4,
+      "step": 8203
+    },
+    {
+      "epoch": 2.6341306790817147,
+      "grad_norm": 0.848976731300354,
+      "learning_rate": 1.903679220538871e-05,
+      "loss": 0.3194,
+      "step": 8204
+    },
+    {
+      "epoch": 2.6344517579065663,
+      "grad_norm": 1.028289556503296,
+      "learning_rate": 1.9016309146596023e-05,
+      "loss": 0.4086,
+      "step": 8205
+    },
+    {
+      "epoch": 2.6347728367314174,
+      "grad_norm": 1.0954947471618652,
+      "learning_rate": 1.8995835955420416e-05,
+      "loss": 0.4058,
+      "step": 8206
+    },
+    {
+      "epoch": 2.635093915556269,
+      "grad_norm": 1.0540190935134888,
+      "learning_rate": 1.897537263435648e-05,
+      "loss": 0.4228,
+      "step": 8207
+    },
+    {
+      "epoch": 2.6354149943811205,
+      "grad_norm": 1.1869220733642578,
+      "learning_rate": 1.895491918589759e-05,
+      "loss": 0.3598,
+      "step": 8208
+    },
+    {
+      "epoch": 2.635736073205972,
+      "grad_norm": 1.1555882692337036,
+      "learning_rate": 1.8934475612536017e-05,
+      "loss": 0.4349,
+      "step": 8209
+    },
+    {
+      "epoch": 2.6360571520308236,
+      "grad_norm": 0.7538331151008606,
+      "learning_rate": 1.891404191676265e-05,
+      "loss": 0.3082,
+      "step": 8210
+    },
+    {
+      "epoch": 2.636378230855675,
+      "grad_norm": 1.011115312576294,
+      "learning_rate": 1.8893618101067355e-05,
+      "loss": 0.2848,
+      "step": 8211
+    },
+    {
+      "epoch": 2.6366993096805267,
+      "grad_norm": 0.9371142983436584,
+      "learning_rate": 1.8873204167938653e-05,
+      "loss": 0.3994,
+      "step": 8212
+    },
+    {
+      "epoch": 2.6370203885053782,
+      "grad_norm": 0.7361756563186646,
+      "learning_rate": 1.885280011986391e-05,
+      "loss": 0.3152,
+      "step": 8213
+    },
+    {
+      "epoch": 2.63734146733023,
+      "grad_norm": 1.0889180898666382,
+      "learning_rate": 1.883240595932938e-05,
+      "loss": 0.3764,
+      "step": 8214
+    },
+    {
+      "epoch": 2.637662546155081,
+      "grad_norm": 0.6203745603561401,
+      "learning_rate": 1.8812021688819915e-05,
+      "loss": 0.2491,
+      "step": 8215
+    },
+    {
+      "epoch": 2.6379836249799324,
+      "grad_norm": 0.7448567748069763,
+      "learning_rate": 1.879164731081937e-05,
+      "loss": 0.282,
+      "step": 8216
+    },
+    {
+      "epoch": 2.638304703804784,
+      "grad_norm": 0.9264616966247559,
+      "learning_rate": 1.877128282781028e-05,
+      "loss": 0.2982,
+      "step": 8217
+    },
+    {
+      "epoch": 2.6386257826296355,
+      "grad_norm": 0.6099106073379517,
+      "learning_rate": 1.8750928242273968e-05,
+      "loss": 0.2363,
+      "step": 8218
+    },
+    {
+      "epoch": 2.638946861454487,
+      "grad_norm": 0.859263002872467,
+      "learning_rate": 1.8730583556690605e-05,
+      "loss": 0.3094,
+      "step": 8219
+    },
+    {
+      "epoch": 2.6392679402793386,
+      "grad_norm": 0.5748609304428101,
+      "learning_rate": 1.8710248773539118e-05,
+      "loss": 0.2404,
+      "step": 8220
+    },
+    {
+      "epoch": 2.63958901910419,
+      "grad_norm": 0.8924229741096497,
+      "learning_rate": 1.8689923895297245e-05,
+      "loss": 0.3191,
+      "step": 8221
+    },
+    {
+      "epoch": 2.6399100979290417,
+      "grad_norm": 0.8334121108055115,
+      "learning_rate": 1.8669608924441496e-05,
+      "loss": 0.3053,
+      "step": 8222
+    },
+    {
+      "epoch": 2.6402311767538933,
+      "grad_norm": 0.9076601266860962,
+      "learning_rate": 1.8649303863447198e-05,
+      "loss": 0.3409,
+      "step": 8223
+    },
+    {
+      "epoch": 2.6405522555787444,
+      "grad_norm": 0.6396355032920837,
+      "learning_rate": 1.8629008714788464e-05,
+      "loss": 0.3092,
+      "step": 8224
+    },
+    {
+      "epoch": 2.640873334403596,
+      "grad_norm": 0.7007104754447937,
+      "learning_rate": 1.8608723480938206e-05,
+      "loss": 0.3039,
+      "step": 8225
+    },
+    {
+      "epoch": 2.6411944132284475,
+      "grad_norm": 0.4519537389278412,
+      "learning_rate": 1.858844816436809e-05,
+      "loss": 0.2598,
+      "step": 8226
+    },
+    {
+      "epoch": 2.641515492053299,
+      "grad_norm": 0.609611988067627,
+      "learning_rate": 1.8568182767548626e-05,
+      "loss": 0.2982,
+      "step": 8227
+    },
+    {
+      "epoch": 2.6418365708781506,
+      "grad_norm": 0.35319435596466064,
+      "learning_rate": 1.854792729294905e-05,
+      "loss": 0.2584,
+      "step": 8228
+    },
+    {
+      "epoch": 2.642157649703002,
+      "grad_norm": 0.5530852675437927,
+      "learning_rate": 1.852768174303752e-05,
+      "loss": 0.3132,
+      "step": 8229
+    },
+    {
+      "epoch": 2.6424787285278537,
+      "grad_norm": 0.6832270622253418,
+      "learning_rate": 1.8507446120280814e-05,
+      "loss": 0.9194,
+      "step": 8230
+    },
+    {
+      "epoch": 2.6427998073527053,
+      "grad_norm": 0.36571961641311646,
+      "learning_rate": 1.848722042714457e-05,
+      "loss": 0.2325,
+      "step": 8231
+    },
+    {
+      "epoch": 2.643120886177557,
+      "grad_norm": 0.3756738305091858,
+      "learning_rate": 1.8467004666093325e-05,
+      "loss": 0.1744,
+      "step": 8232
+    },
+    {
+      "epoch": 2.643441965002408,
+      "grad_norm": 0.3277255594730377,
+      "learning_rate": 1.8446798839590186e-05,
+      "loss": 0.1241,
+      "step": 8233
+    },
+    {
+      "epoch": 2.6437630438272595,
+      "grad_norm": 0.264514297246933,
+      "learning_rate": 1.8426602950097283e-05,
+      "loss": 0.1122,
+      "step": 8234
+    },
+    {
+      "epoch": 2.644084122652111,
+      "grad_norm": 0.20338915288448334,
+      "learning_rate": 1.8406417000075325e-05,
+      "loss": 0.0738,
+      "step": 8235
+    },
+    {
+      "epoch": 2.6444052014769626,
+      "grad_norm": 0.8854188323020935,
+      "learning_rate": 1.838624099198397e-05,
+      "loss": 0.4599,
+      "step": 8236
+    },
+    {
+      "epoch": 2.644726280301814,
+      "grad_norm": 0.7100738883018494,
+      "learning_rate": 1.8366074928281607e-05,
+      "loss": 0.2707,
+      "step": 8237
+    },
+    {
+      "epoch": 2.6450473591266657,
+      "grad_norm": 0.9927451014518738,
+      "learning_rate": 1.834591881142538e-05,
+      "loss": 0.3854,
+      "step": 8238
+    },
+    {
+      "epoch": 2.645368437951517,
+      "grad_norm": 0.8780126571655273,
+      "learning_rate": 1.8325772643871265e-05,
+      "loss": 0.3619,
+      "step": 8239
+    },
+    {
+      "epoch": 2.6456895167763683,
+      "grad_norm": 0.7575050592422485,
+      "learning_rate": 1.8305636428074014e-05,
+      "loss": 0.3066,
+      "step": 8240
+    },
+    {
+      "epoch": 2.6460105956012203,
+      "grad_norm": 0.599746584892273,
+      "learning_rate": 1.8285510166487152e-05,
+      "loss": 0.2652,
+      "step": 8241
+    },
+    {
+      "epoch": 2.6463316744260714,
+      "grad_norm": 0.9272753596305847,
+      "learning_rate": 1.826539386156302e-05,
+      "loss": 0.4376,
+      "step": 8242
+    },
+    {
+      "epoch": 2.646652753250923,
+      "grad_norm": 0.7078785300254822,
+      "learning_rate": 1.8245287515752708e-05,
+      "loss": 0.2626,
+      "step": 8243
+    },
+    {
+      "epoch": 2.6469738320757745,
+      "grad_norm": 0.8110345602035522,
+      "learning_rate": 1.8225191131506126e-05,
+      "loss": 0.2777,
+      "step": 8244
+    },
+    {
+      "epoch": 2.647294910900626,
+      "grad_norm": 0.7809237241744995,
+      "learning_rate": 1.8205104711271957e-05,
+      "loss": 0.3272,
+      "step": 8245
+    },
+    {
+      "epoch": 2.6476159897254776,
+      "grad_norm": 0.7376668453216553,
+      "learning_rate": 1.818502825749764e-05,
+      "loss": 0.3092,
+      "step": 8246
+    },
+    {
+      "epoch": 2.647937068550329,
+      "grad_norm": 0.645916223526001,
+      "learning_rate": 1.816496177262952e-05,
+      "loss": 0.2277,
+      "step": 8247
+    },
+    {
+      "epoch": 2.6482581473751807,
+      "grad_norm": 1.4311972856521606,
+      "learning_rate": 1.8144905259112542e-05,
+      "loss": 0.4121,
+      "step": 8248
+    },
+    {
+      "epoch": 2.648579226200032,
+      "grad_norm": 0.8135798573493958,
+      "learning_rate": 1.812485871939056e-05,
+      "loss": 0.3105,
+      "step": 8249
+    },
+    {
+      "epoch": 2.648900305024884,
+      "grad_norm": 0.9382086992263794,
+      "learning_rate": 1.81048221559062e-05,
+      "loss": 0.3712,
+      "step": 8250
+    },
+    {
+      "epoch": 2.649221383849735,
+      "grad_norm": 0.9404576420783997,
+      "learning_rate": 1.808479557110081e-05,
+      "loss": 0.3391,
+      "step": 8251
+    },
+    {
+      "epoch": 2.6495424626745865,
+      "grad_norm": 1.0706158876419067,
+      "learning_rate": 1.8064778967414662e-05,
+      "loss": 0.433,
+      "step": 8252
+    },
+    {
+      "epoch": 2.649863541499438,
+      "grad_norm": 0.7485371828079224,
+      "learning_rate": 1.804477234728661e-05,
+      "loss": 0.2812,
+      "step": 8253
+    },
+    {
+      "epoch": 2.6501846203242896,
+      "grad_norm": 1.0219621658325195,
+      "learning_rate": 1.8024775713154473e-05,
+      "loss": 0.4084,
+      "step": 8254
+    },
+    {
+      "epoch": 2.650505699149141,
+      "grad_norm": 0.861611545085907,
+      "learning_rate": 1.8004789067454764e-05,
+      "loss": 0.3148,
+      "step": 8255
+    },
+    {
+      "epoch": 2.6508267779739927,
+      "grad_norm": 1.0790586471557617,
+      "learning_rate": 1.7984812412622787e-05,
+      "loss": 0.4146,
+      "step": 8256
+    },
+    {
+      "epoch": 2.6511478567988442,
+      "grad_norm": 0.6884385347366333,
+      "learning_rate": 1.7964845751092664e-05,
+      "loss": 0.2809,
+      "step": 8257
+    },
+    {
+      "epoch": 2.6514689356236953,
+      "grad_norm": 1.2098215818405151,
+      "learning_rate": 1.794488908529719e-05,
+      "loss": 0.4282,
+      "step": 8258
+    },
+    {
+      "epoch": 2.6517900144485473,
+      "grad_norm": 0.5315207242965698,
+      "learning_rate": 1.792494241766811e-05,
+      "loss": 0.2248,
+      "step": 8259
+    },
+    {
+      "epoch": 2.6521110932733984,
+      "grad_norm": 1.038537859916687,
+      "learning_rate": 1.790500575063584e-05,
+      "loss": 0.4178,
+      "step": 8260
+    },
+    {
+      "epoch": 2.65243217209825,
+      "grad_norm": 0.7643697261810303,
+      "learning_rate": 1.78850790866296e-05,
+      "loss": 0.2673,
+      "step": 8261
+    },
+    {
+      "epoch": 2.6527532509231015,
+      "grad_norm": 0.6102786064147949,
+      "learning_rate": 1.7865162428077386e-05,
+      "loss": 0.2725,
+      "step": 8262
+    },
+    {
+      "epoch": 2.653074329747953,
+      "grad_norm": 0.9418346285820007,
+      "learning_rate": 1.7845255777406e-05,
+      "loss": 0.3239,
+      "step": 8263
+    },
+    {
+      "epoch": 2.6533954085728046,
+      "grad_norm": 0.7261042594909668,
+      "learning_rate": 1.7825359137040988e-05,
+      "loss": 0.2828,
+      "step": 8264
+    },
+    {
+      "epoch": 2.653716487397656,
+      "grad_norm": 0.9733819961547852,
+      "learning_rate": 1.7805472509406696e-05,
+      "loss": 0.3408,
+      "step": 8265
+    },
+    {
+      "epoch": 2.6540375662225077,
+      "grad_norm": 0.5591874718666077,
+      "learning_rate": 1.7785595896926265e-05,
+      "loss": 0.2455,
+      "step": 8266
+    },
+    {
+      "epoch": 2.654358645047359,
+      "grad_norm": 0.7229105234146118,
+      "learning_rate": 1.7765729302021596e-05,
+      "loss": 0.2537,
+      "step": 8267
+    },
+    {
+      "epoch": 2.654679723872211,
+      "grad_norm": 0.6605117917060852,
+      "learning_rate": 1.7745872727113356e-05,
+      "loss": 0.2885,
+      "step": 8268
+    },
+    {
+      "epoch": 2.655000802697062,
+      "grad_norm": 0.5864028334617615,
+      "learning_rate": 1.7726026174621003e-05,
+      "loss": 0.2461,
+      "step": 8269
+    },
+    {
+      "epoch": 2.6553218815219135,
+      "grad_norm": 1.8025933504104614,
+      "learning_rate": 1.7706189646962847e-05,
+      "loss": 0.3684,
+      "step": 8270
+    },
+    {
+      "epoch": 2.655642960346765,
+      "grad_norm": 0.835909903049469,
+      "learning_rate": 1.7686363146555805e-05,
+      "loss": 0.3308,
+      "step": 8271
+    },
+    {
+      "epoch": 2.6559640391716166,
+      "grad_norm": 0.9109052419662476,
+      "learning_rate": 1.7666546675815778e-05,
+      "loss": 0.3649,
+      "step": 8272
+    },
+    {
+      "epoch": 2.656285117996468,
+      "grad_norm": 0.5964946150779724,
+      "learning_rate": 1.7646740237157256e-05,
+      "loss": 0.2541,
+      "step": 8273
+    },
+    {
+      "epoch": 2.6566061968213197,
+      "grad_norm": 0.6849761605262756,
+      "learning_rate": 1.7626943832993647e-05,
+      "loss": 0.29,
+      "step": 8274
+    },
+    {
+      "epoch": 2.6569272756461713,
+      "grad_norm": 0.5830851197242737,
+      "learning_rate": 1.760715746573709e-05,
+      "loss": 0.2506,
+      "step": 8275
+    },
+    {
+      "epoch": 2.6572483544710224,
+      "grad_norm": 0.5915994644165039,
+      "learning_rate": 1.7587381137798432e-05,
+      "loss": 0.299,
+      "step": 8276
+    },
+    {
+      "epoch": 2.6575694332958744,
+      "grad_norm": 0.5111377835273743,
+      "learning_rate": 1.7567614851587443e-05,
+      "loss": 0.2853,
+      "step": 8277
+    },
+    {
+      "epoch": 2.6578905121207255,
+      "grad_norm": 0.5485566854476929,
+      "learning_rate": 1.754785860951249e-05,
+      "loss": 0.2842,
+      "step": 8278
+    },
+    {
+      "epoch": 2.658211590945577,
+      "grad_norm": 0.38606464862823486,
+      "learning_rate": 1.752811241398089e-05,
+      "loss": 0.2609,
+      "step": 8279
+    },
+    {
+      "epoch": 2.6585326697704286,
+      "grad_norm": 0.36825671792030334,
+      "learning_rate": 1.750837626739863e-05,
+      "loss": 0.4696,
+      "step": 8280
+    },
+    {
+      "epoch": 2.65885374859528,
+      "grad_norm": 0.49861788749694824,
+      "learning_rate": 1.7488650172170496e-05,
+      "loss": 0.4966,
+      "step": 8281
+    },
+    {
+      "epoch": 2.6591748274201317,
+      "grad_norm": 0.3955051004886627,
+      "learning_rate": 1.7468934130700044e-05,
+      "loss": 0.1671,
+      "step": 8282
+    },
+    {
+      "epoch": 2.659495906244983,
+      "grad_norm": 0.30318066477775574,
+      "learning_rate": 1.744922814538964e-05,
+      "loss": 0.1232,
+      "step": 8283
+    },
+    {
+      "epoch": 2.6598169850698348,
+      "grad_norm": 0.5016526579856873,
+      "learning_rate": 1.7429532218640377e-05,
+      "loss": 0.1385,
+      "step": 8284
+    },
+    {
+      "epoch": 2.660138063894686,
+      "grad_norm": 0.33723151683807373,
+      "learning_rate": 1.7409846352852143e-05,
+      "loss": 0.122,
+      "step": 8285
+    },
+    {
+      "epoch": 2.660459142719538,
+      "grad_norm": 0.7565352916717529,
+      "learning_rate": 1.7390170550423625e-05,
+      "loss": 0.3913,
+      "step": 8286
+    },
+    {
+      "epoch": 2.660780221544389,
+      "grad_norm": 0.857373833656311,
+      "learning_rate": 1.737050481375223e-05,
+      "loss": 0.4855,
+      "step": 8287
+    },
+    {
+      "epoch": 2.6611013003692405,
+      "grad_norm": 0.7241206765174866,
+      "learning_rate": 1.7350849145234184e-05,
+      "loss": 0.281,
+      "step": 8288
+    },
+    {
+      "epoch": 2.661422379194092,
+      "grad_norm": 0.822955310344696,
+      "learning_rate": 1.7331203547264453e-05,
+      "loss": 0.3799,
+      "step": 8289
+    },
+    {
+      "epoch": 2.6617434580189436,
+      "grad_norm": 0.9299210906028748,
+      "learning_rate": 1.7311568022236845e-05,
+      "loss": 0.3482,
+      "step": 8290
+    },
+    {
+      "epoch": 2.662064536843795,
+      "grad_norm": 0.68746417760849,
+      "learning_rate": 1.7291942572543807e-05,
+      "loss": 0.293,
+      "step": 8291
+    },
+    {
+      "epoch": 2.6623856156686467,
+      "grad_norm": 0.8198730945587158,
+      "learning_rate": 1.7272327200576742e-05,
+      "loss": 0.3485,
+      "step": 8292
+    },
+    {
+      "epoch": 2.6627066944934983,
+      "grad_norm": 1.226529598236084,
+      "learning_rate": 1.7252721908725632e-05,
+      "loss": 0.3973,
+      "step": 8293
+    },
+    {
+      "epoch": 2.6630277733183494,
+      "grad_norm": 0.7077026963233948,
+      "learning_rate": 1.7233126699379343e-05,
+      "loss": 0.2794,
+      "step": 8294
+    },
+    {
+      "epoch": 2.6633488521432014,
+      "grad_norm": 0.8905551433563232,
+      "learning_rate": 1.721354157492555e-05,
+      "loss": 0.3946,
+      "step": 8295
+    },
+    {
+      "epoch": 2.6636699309680525,
+      "grad_norm": 0.9082470536231995,
+      "learning_rate": 1.719396653775056e-05,
+      "loss": 0.3648,
+      "step": 8296
+    },
+    {
+      "epoch": 2.663991009792904,
+      "grad_norm": 0.6751482486724854,
+      "learning_rate": 1.7174401590239587e-05,
+      "loss": 0.2582,
+      "step": 8297
+    },
+    {
+      "epoch": 2.6643120886177556,
+      "grad_norm": 0.916202187538147,
+      "learning_rate": 1.715484673477654e-05,
+      "loss": 0.3537,
+      "step": 8298
+    },
+    {
+      "epoch": 2.664633167442607,
+      "grad_norm": 1.0404547452926636,
+      "learning_rate": 1.7135301973744124e-05,
+      "loss": 0.3866,
+      "step": 8299
+    },
+    {
+      "epoch": 2.6649542462674587,
+      "grad_norm": 0.7834314107894897,
+      "learning_rate": 1.7115767309523812e-05,
+      "loss": 0.3115,
+      "step": 8300
+    },
+    {
+      "epoch": 2.6652753250923102,
+      "grad_norm": 0.8973174691200256,
+      "learning_rate": 1.7096242744495837e-05,
+      "loss": 0.2688,
+      "step": 8301
+    },
+    {
+      "epoch": 2.665596403917162,
+      "grad_norm": 1.4204421043395996,
+      "learning_rate": 1.70767282810392e-05,
+      "loss": 0.5071,
+      "step": 8302
+    },
+    {
+      "epoch": 2.665917482742013,
+      "grad_norm": 1.0806547403335571,
+      "learning_rate": 1.7057223921531707e-05,
+      "loss": 0.3664,
+      "step": 8303
+    },
+    {
+      "epoch": 2.666238561566865,
+      "grad_norm": 1.0438770055770874,
+      "learning_rate": 1.7037729668349877e-05,
+      "loss": 0.3832,
+      "step": 8304
+    },
+    {
+      "epoch": 2.666559640391716,
+      "grad_norm": 0.9235129356384277,
+      "learning_rate": 1.7018245523869036e-05,
+      "loss": 0.345,
+      "step": 8305
+    },
+    {
+      "epoch": 2.6668807192165676,
+      "grad_norm": 0.9867783784866333,
+      "learning_rate": 1.6998771490463262e-05,
+      "loss": 0.3577,
+      "step": 8306
+    },
+    {
+      "epoch": 2.667201798041419,
+      "grad_norm": 0.9603634476661682,
+      "learning_rate": 1.697930757050542e-05,
+      "loss": 0.3107,
+      "step": 8307
+    },
+    {
+      "epoch": 2.6675228768662707,
+      "grad_norm": 0.6079133749008179,
+      "learning_rate": 1.6959853766367118e-05,
+      "loss": 0.2685,
+      "step": 8308
+    },
+    {
+      "epoch": 2.667843955691122,
+      "grad_norm": 1.3194692134857178,
+      "learning_rate": 1.6940410080418723e-05,
+      "loss": 0.3876,
+      "step": 8309
+    },
+    {
+      "epoch": 2.6681650345159738,
+      "grad_norm": 0.7050005793571472,
+      "learning_rate": 1.6920976515029463e-05,
+      "loss": 0.301,
+      "step": 8310
+    },
+    {
+      "epoch": 2.6684861133408253,
+      "grad_norm": 1.0123231410980225,
+      "learning_rate": 1.690155307256719e-05,
+      "loss": 0.4113,
+      "step": 8311
+    },
+    {
+      "epoch": 2.6688071921656764,
+      "grad_norm": 0.8244590163230896,
+      "learning_rate": 1.6882139755398574e-05,
+      "loss": 0.3621,
+      "step": 8312
+    },
+    {
+      "epoch": 2.6691282709905284,
+      "grad_norm": 0.5136908292770386,
+      "learning_rate": 1.686273656588917e-05,
+      "loss": 0.2476,
+      "step": 8313
+    },
+    {
+      "epoch": 2.6694493498153795,
+      "grad_norm": 0.7136945128440857,
+      "learning_rate": 1.6843343506403075e-05,
+      "loss": 0.2703,
+      "step": 8314
+    },
+    {
+      "epoch": 2.669770428640231,
+      "grad_norm": 1.0239125490188599,
+      "learning_rate": 1.6823960579303377e-05,
+      "loss": 0.314,
+      "step": 8315
+    },
+    {
+      "epoch": 2.6700915074650826,
+      "grad_norm": 0.6786704063415527,
+      "learning_rate": 1.680458778695174e-05,
+      "loss": 0.2897,
+      "step": 8316
+    },
+    {
+      "epoch": 2.670412586289934,
+      "grad_norm": 0.6435678601264954,
+      "learning_rate": 1.678522513170875e-05,
+      "loss": 0.274,
+      "step": 8317
+    },
+    {
+      "epoch": 2.6707336651147857,
+      "grad_norm": 0.6759331822395325,
+      "learning_rate": 1.6765872615933677e-05,
+      "loss": 0.2739,
+      "step": 8318
+    },
+    {
+      "epoch": 2.6710547439396373,
+      "grad_norm": 0.7092223167419434,
+      "learning_rate": 1.6746530241984504e-05,
+      "loss": 0.2994,
+      "step": 8319
+    },
+    {
+      "epoch": 2.671375822764489,
+      "grad_norm": 0.9770942330360413,
+      "learning_rate": 1.6727198012218114e-05,
+      "loss": 0.3379,
+      "step": 8320
+    },
+    {
+      "epoch": 2.67169690158934,
+      "grad_norm": 1.3338515758514404,
+      "learning_rate": 1.6707875928990058e-05,
+      "loss": 0.3295,
+      "step": 8321
+    },
+    {
+      "epoch": 2.672017980414192,
+      "grad_norm": 0.6919156312942505,
+      "learning_rate": 1.668856399465466e-05,
+      "loss": 0.2663,
+      "step": 8322
+    },
+    {
+      "epoch": 2.672339059239043,
+      "grad_norm": 0.9007259011268616,
+      "learning_rate": 1.6669262211565028e-05,
+      "loss": 0.3146,
+      "step": 8323
+    },
+    {
+      "epoch": 2.6726601380638946,
+      "grad_norm": 0.6799442172050476,
+      "learning_rate": 1.6649970582073028e-05,
+      "loss": 0.2957,
+      "step": 8324
+    },
+    {
+      "epoch": 2.672981216888746,
+      "grad_norm": 0.8429846167564392,
+      "learning_rate": 1.6630689108529284e-05,
+      "loss": 0.3215,
+      "step": 8325
+    },
+    {
+      "epoch": 2.6733022957135977,
+      "grad_norm": 0.9628200531005859,
+      "learning_rate": 1.661141779328319e-05,
+      "loss": 0.3283,
+      "step": 8326
+    },
+    {
+      "epoch": 2.6736233745384492,
+      "grad_norm": 0.7817652225494385,
+      "learning_rate": 1.6592156638682886e-05,
+      "loss": 0.2815,
+      "step": 8327
+    },
+    {
+      "epoch": 2.6739444533633008,
+      "grad_norm": 0.5728936791419983,
+      "learning_rate": 1.6572905647075298e-05,
+      "loss": 0.2687,
+      "step": 8328
+    },
+    {
+      "epoch": 2.6742655321881523,
+      "grad_norm": 0.3916088342666626,
+      "learning_rate": 1.6553664820806102e-05,
+      "loss": 0.2574,
+      "step": 8329
+    },
+    {
+      "epoch": 2.6745866110130034,
+      "grad_norm": 0.5118910670280457,
+      "learning_rate": 1.6534434162219727e-05,
+      "loss": 0.6336,
+      "step": 8330
+    },
+    {
+      "epoch": 2.6749076898378554,
+      "grad_norm": 0.4536105990409851,
+      "learning_rate": 1.6515213673659357e-05,
+      "loss": 0.4168,
+      "step": 8331
+    },
+    {
+      "epoch": 2.6752287686627065,
+      "grad_norm": 0.442342072725296,
+      "learning_rate": 1.649600335746695e-05,
+      "loss": 0.2928,
+      "step": 8332
+    },
+    {
+      "epoch": 2.675549847487558,
+      "grad_norm": 0.4411477744579315,
+      "learning_rate": 1.6476803215983294e-05,
+      "loss": 0.1311,
+      "step": 8333
+    },
+    {
+      "epoch": 2.6758709263124096,
+      "grad_norm": 0.3371138572692871,
+      "learning_rate": 1.6457613251547754e-05,
+      "loss": 0.1534,
+      "step": 8334
+    },
+    {
+      "epoch": 2.676192005137261,
+      "grad_norm": 0.6045904755592346,
+      "learning_rate": 1.643843346649866e-05,
+      "loss": 0.3301,
+      "step": 8335
+    },
+    {
+      "epoch": 2.6765130839621127,
+      "grad_norm": 0.905532956123352,
+      "learning_rate": 1.6419263863172997e-05,
+      "loss": 0.3937,
+      "step": 8336
+    },
+    {
+      "epoch": 2.6768341627869643,
+      "grad_norm": 0.7415409088134766,
+      "learning_rate": 1.640010444390646e-05,
+      "loss": 0.3073,
+      "step": 8337
+    },
+    {
+      "epoch": 2.677155241611816,
+      "grad_norm": 0.8520146608352661,
+      "learning_rate": 1.6380955211033656e-05,
+      "loss": 0.4127,
+      "step": 8338
+    },
+    {
+      "epoch": 2.677476320436667,
+      "grad_norm": 0.9074903130531311,
+      "learning_rate": 1.6361816166887768e-05,
+      "loss": 0.3889,
+      "step": 8339
+    },
+    {
+      "epoch": 2.677797399261519,
+      "grad_norm": 0.7967313528060913,
+      "learning_rate": 1.634268731380091e-05,
+      "loss": 0.2836,
+      "step": 8340
+    },
+    {
+      "epoch": 2.67811847808637,
+      "grad_norm": 0.7491098046302795,
+      "learning_rate": 1.632356865410384e-05,
+      "loss": 0.3364,
+      "step": 8341
+    },
+    {
+      "epoch": 2.6784395569112216,
+      "grad_norm": 0.7044479250907898,
+      "learning_rate": 1.6304460190126103e-05,
+      "loss": 0.3405,
+      "step": 8342
+    },
+    {
+      "epoch": 2.678760635736073,
+      "grad_norm": 0.7631996273994446,
+      "learning_rate": 1.628536192419603e-05,
+      "loss": 0.3188,
+      "step": 8343
+    },
+    {
+      "epoch": 2.6790817145609247,
+      "grad_norm": 0.9044662714004517,
+      "learning_rate": 1.6266273858640656e-05,
+      "loss": 0.3248,
+      "step": 8344
+    },
+    {
+      "epoch": 2.6794027933857762,
+      "grad_norm": 0.9730323553085327,
+      "learning_rate": 1.6247195995785837e-05,
+      "loss": 0.3272,
+      "step": 8345
+    },
+    {
+      "epoch": 2.679723872210628,
+      "grad_norm": 0.9909996390342712,
+      "learning_rate": 1.622812833795613e-05,
+      "loss": 0.4031,
+      "step": 8346
+    },
+    {
+      "epoch": 2.6800449510354794,
+      "grad_norm": 1.0814063549041748,
+      "learning_rate": 1.6209070887474876e-05,
+      "loss": 0.3565,
+      "step": 8347
+    },
+    {
+      "epoch": 2.6803660298603305,
+      "grad_norm": 0.8834957480430603,
+      "learning_rate": 1.6190023646664175e-05,
+      "loss": 0.4426,
+      "step": 8348
+    },
+    {
+      "epoch": 2.6806871086851825,
+      "grad_norm": 1.0340230464935303,
+      "learning_rate": 1.6170986617844863e-05,
+      "loss": 0.3924,
+      "step": 8349
+    },
+    {
+      "epoch": 2.6810081875100336,
+      "grad_norm": 0.743299663066864,
+      "learning_rate": 1.6151959803336535e-05,
+      "loss": 0.3009,
+      "step": 8350
+    },
+    {
+      "epoch": 2.681329266334885,
+      "grad_norm": 0.802480936050415,
+      "learning_rate": 1.6132943205457606e-05,
+      "loss": 0.3529,
+      "step": 8351
+    },
+    {
+      "epoch": 2.6816503451597367,
+      "grad_norm": 1.1612759828567505,
+      "learning_rate": 1.611393682652511e-05,
+      "loss": 0.377,
+      "step": 8352
+    },
+    {
+      "epoch": 2.681971423984588,
+      "grad_norm": 0.9446073174476624,
+      "learning_rate": 1.6094940668855006e-05,
+      "loss": 0.3516,
+      "step": 8353
+    },
+    {
+      "epoch": 2.6822925028094398,
+      "grad_norm": 0.7460670471191406,
+      "learning_rate": 1.6075954734761845e-05,
+      "loss": 0.3126,
+      "step": 8354
+    },
+    {
+      "epoch": 2.6826135816342913,
+      "grad_norm": 0.7956533432006836,
+      "learning_rate": 1.6056979026559004e-05,
+      "loss": 0.3152,
+      "step": 8355
+    },
+    {
+      "epoch": 2.682934660459143,
+      "grad_norm": 0.5445581674575806,
+      "learning_rate": 1.6038013546558696e-05,
+      "loss": 0.2414,
+      "step": 8356
+    },
+    {
+      "epoch": 2.683255739283994,
+      "grad_norm": 0.8565784692764282,
+      "learning_rate": 1.601905829707171e-05,
+      "loss": 0.2997,
+      "step": 8357
+    },
+    {
+      "epoch": 2.683576818108846,
+      "grad_norm": 0.512998104095459,
+      "learning_rate": 1.600011328040777e-05,
+      "loss": 0.2233,
+      "step": 8358
+    },
+    {
+      "epoch": 2.683897896933697,
+      "grad_norm": 0.8238282203674316,
+      "learning_rate": 1.598117849887518e-05,
+      "loss": 0.3354,
+      "step": 8359
+    },
+    {
+      "epoch": 2.6842189757585486,
+      "grad_norm": 0.8783250451087952,
+      "learning_rate": 1.596225395478116e-05,
+      "loss": 0.3224,
+      "step": 8360
+    },
+    {
+      "epoch": 2.6845400545834,
+      "grad_norm": 0.6157239675521851,
+      "learning_rate": 1.5943339650431576e-05,
+      "loss": 0.2397,
+      "step": 8361
+    },
+    {
+      "epoch": 2.6848611334082517,
+      "grad_norm": 0.7254630327224731,
+      "learning_rate": 1.592443558813109e-05,
+      "loss": 0.249,
+      "step": 8362
+    },
+    {
+      "epoch": 2.6851822122331033,
+      "grad_norm": 1.0518107414245605,
+      "learning_rate": 1.5905541770183096e-05,
+      "loss": 0.3714,
+      "step": 8363
+    },
+    {
+      "epoch": 2.685503291057955,
+      "grad_norm": 0.6833174824714661,
+      "learning_rate": 1.588665819888976e-05,
+      "loss": 0.277,
+      "step": 8364
+    },
+    {
+      "epoch": 2.6858243698828064,
+      "grad_norm": 0.9686453938484192,
+      "learning_rate": 1.586778487655197e-05,
+      "loss": 0.4063,
+      "step": 8365
+    },
+    {
+      "epoch": 2.6861454487076575,
+      "grad_norm": 0.634539008140564,
+      "learning_rate": 1.5848921805469397e-05,
+      "loss": 0.2757,
+      "step": 8366
+    },
+    {
+      "epoch": 2.6864665275325095,
+      "grad_norm": 0.6041567325592041,
+      "learning_rate": 1.5830068987940438e-05,
+      "loss": 0.2413,
+      "step": 8367
+    },
+    {
+      "epoch": 2.6867876063573606,
+      "grad_norm": 0.7220975160598755,
+      "learning_rate": 1.581122642626226e-05,
+      "loss": 0.2697,
+      "step": 8368
+    },
+    {
+      "epoch": 2.687108685182212,
+      "grad_norm": 1.3267475366592407,
+      "learning_rate": 1.579239412273078e-05,
+      "loss": 0.3793,
+      "step": 8369
+    },
+    {
+      "epoch": 2.6874297640070637,
+      "grad_norm": 1.0630850791931152,
+      "learning_rate": 1.577357207964062e-05,
+      "loss": 0.3434,
+      "step": 8370
+    },
+    {
+      "epoch": 2.6877508428319152,
+      "grad_norm": 0.6697249412536621,
+      "learning_rate": 1.5754760299285252e-05,
+      "loss": 0.2852,
+      "step": 8371
+    },
+    {
+      "epoch": 2.688071921656767,
+      "grad_norm": 0.8393073081970215,
+      "learning_rate": 1.5735958783956794e-05,
+      "loss": 0.3452,
+      "step": 8372
+    },
+    {
+      "epoch": 2.6883930004816183,
+      "grad_norm": 0.8739592432975769,
+      "learning_rate": 1.5717167535946144e-05,
+      "loss": 0.3311,
+      "step": 8373
+    },
+    {
+      "epoch": 2.68871407930647,
+      "grad_norm": 0.4763057231903076,
+      "learning_rate": 1.569838655754298e-05,
+      "loss": 0.2477,
+      "step": 8374
+    },
+    {
+      "epoch": 2.689035158131321,
+      "grad_norm": 0.5161868333816528,
+      "learning_rate": 1.567961585103567e-05,
+      "loss": 0.2638,
+      "step": 8375
+    },
+    {
+      "epoch": 2.689356236956173,
+      "grad_norm": 0.5207780003547668,
+      "learning_rate": 1.566085541871145e-05,
+      "loss": 0.2612,
+      "step": 8376
+    },
+    {
+      "epoch": 2.689677315781024,
+      "grad_norm": 0.388313353061676,
+      "learning_rate": 1.564210526285612e-05,
+      "loss": 0.25,
+      "step": 8377
+    },
+    {
+      "epoch": 2.6899983946058756,
+      "grad_norm": 0.6559248566627502,
+      "learning_rate": 1.5623365385754408e-05,
+      "loss": 0.2934,
+      "step": 8378
+    },
+    {
+      "epoch": 2.690319473430727,
+      "grad_norm": 0.6351720094680786,
+      "learning_rate": 1.560463578968967e-05,
+      "loss": 0.2985,
+      "step": 8379
+    },
+    {
+      "epoch": 2.6906405522555787,
+      "grad_norm": 0.6100612282752991,
+      "learning_rate": 1.5585916476944073e-05,
+      "loss": 0.7795,
+      "step": 8380
+    },
+    {
+      "epoch": 2.6909616310804303,
+      "grad_norm": 0.3323669135570526,
+      "learning_rate": 1.5567207449798515e-05,
+      "loss": 0.246,
+      "step": 8381
+    },
+    {
+      "epoch": 2.691282709905282,
+      "grad_norm": 0.4569035768508911,
+      "learning_rate": 1.5548508710532572e-05,
+      "loss": 0.1564,
+      "step": 8382
+    },
+    {
+      "epoch": 2.6916037887301334,
+      "grad_norm": 0.3317892551422119,
+      "learning_rate": 1.5529820261424698e-05,
+      "loss": 0.1498,
+      "step": 8383
+    },
+    {
+      "epoch": 2.6919248675549845,
+      "grad_norm": 0.3487982451915741,
+      "learning_rate": 1.551114210475201e-05,
+      "loss": 0.1257,
+      "step": 8384
+    },
+    {
+      "epoch": 2.6922459463798365,
+      "grad_norm": 0.42391496896743774,
+      "learning_rate": 1.5492474242790366e-05,
+      "loss": 0.1278,
+      "step": 8385
+    },
+    {
+      "epoch": 2.6925670252046876,
+      "grad_norm": 0.3310485780239105,
+      "learning_rate": 1.547381667781439e-05,
+      "loss": 0.1795,
+      "step": 8386
+    },
+    {
+      "epoch": 2.692888104029539,
+      "grad_norm": 0.8500526547431946,
+      "learning_rate": 1.545516941209747e-05,
+      "loss": 0.3601,
+      "step": 8387
+    },
+    {
+      "epoch": 2.6932091828543907,
+      "grad_norm": 0.8416752815246582,
+      "learning_rate": 1.54365324479117e-05,
+      "loss": 0.3523,
+      "step": 8388
+    },
+    {
+      "epoch": 2.6935302616792423,
+      "grad_norm": 0.7443310022354126,
+      "learning_rate": 1.541790578752794e-05,
+      "loss": 0.3137,
+      "step": 8389
+    },
+    {
+      "epoch": 2.693851340504094,
+      "grad_norm": 0.7136545777320862,
+      "learning_rate": 1.539928943321579e-05,
+      "loss": 0.2981,
+      "step": 8390
+    },
+    {
+      "epoch": 2.6941724193289454,
+      "grad_norm": 0.6598768830299377,
+      "learning_rate": 1.538068338724361e-05,
+      "loss": 0.2808,
+      "step": 8391
+    },
+    {
+      "epoch": 2.694493498153797,
+      "grad_norm": 0.8997542858123779,
+      "learning_rate": 1.5362087651878475e-05,
+      "loss": 0.3301,
+      "step": 8392
+    },
+    {
+      "epoch": 2.694814576978648,
+      "grad_norm": 0.9594383835792542,
+      "learning_rate": 1.5343502229386207e-05,
+      "loss": 0.382,
+      "step": 8393
+    },
+    {
+      "epoch": 2.6951356558035,
+      "grad_norm": 0.7930410504341125,
+      "learning_rate": 1.532492712203145e-05,
+      "loss": 0.3238,
+      "step": 8394
+    },
+    {
+      "epoch": 2.695456734628351,
+      "grad_norm": 0.918022096157074,
+      "learning_rate": 1.530636233207743e-05,
+      "loss": 0.3953,
+      "step": 8395
+    },
+    {
+      "epoch": 2.6957778134532027,
+      "grad_norm": 1.3082184791564941,
+      "learning_rate": 1.528780786178631e-05,
+      "loss": 0.4692,
+      "step": 8396
+    },
+    {
+      "epoch": 2.696098892278054,
+      "grad_norm": 0.8029715418815613,
+      "learning_rate": 1.526926371341878e-05,
+      "loss": 0.3129,
+      "step": 8397
+    },
+    {
+      "epoch": 2.6964199711029058,
+      "grad_norm": 0.8471099734306335,
+      "learning_rate": 1.5250729889234482e-05,
+      "loss": 0.3716,
+      "step": 8398
+    },
+    {
+      "epoch": 2.6967410499277573,
+      "grad_norm": 0.8596959710121155,
+      "learning_rate": 1.5232206391491699e-05,
+      "loss": 0.3605,
+      "step": 8399
+    },
+    {
+      "epoch": 2.697062128752609,
+      "grad_norm": 0.6960004568099976,
+      "learning_rate": 1.521369322244739e-05,
+      "loss": 0.288,
+      "step": 8400
+    },
+    {
+      "epoch": 2.6973832075774604,
+      "grad_norm": 0.6006725430488586,
+      "learning_rate": 1.5195190384357404e-05,
+      "loss": 0.2479,
+      "step": 8401
+    },
+    {
+      "epoch": 2.6977042864023115,
+      "grad_norm": 0.9754154086112976,
+      "learning_rate": 1.5176697879476232e-05,
+      "loss": 0.2391,
+      "step": 8402
+    },
+    {
+      "epoch": 2.6980253652271635,
+      "grad_norm": 0.7562586665153503,
+      "learning_rate": 1.5158215710057123e-05,
+      "loss": 0.2767,
+      "step": 8403
+    },
+    {
+      "epoch": 2.6983464440520146,
+      "grad_norm": 0.9271560907363892,
+      "learning_rate": 1.5139743878352075e-05,
+      "loss": 0.3777,
+      "step": 8404
+    },
+    {
+      "epoch": 2.698667522876866,
+      "grad_norm": 0.9956105947494507,
+      "learning_rate": 1.5121282386611824e-05,
+      "loss": 0.4351,
+      "step": 8405
+    },
+    {
+      "epoch": 2.6989886017017177,
+      "grad_norm": 1.1439400911331177,
+      "learning_rate": 1.5102831237085857e-05,
+      "loss": 0.3981,
+      "step": 8406
+    },
+    {
+      "epoch": 2.6993096805265693,
+      "grad_norm": 0.8666257262229919,
+      "learning_rate": 1.5084390432022377e-05,
+      "loss": 0.3045,
+      "step": 8407
+    },
+    {
+      "epoch": 2.699630759351421,
+      "grad_norm": 0.7622800469398499,
+      "learning_rate": 1.5065959973668353e-05,
+      "loss": 0.2846,
+      "step": 8408
+    },
+    {
+      "epoch": 2.6999518381762724,
+      "grad_norm": 0.9066924452781677,
+      "learning_rate": 1.5047539864269478e-05,
+      "loss": 0.2969,
+      "step": 8409
+    },
+    {
+      "epoch": 2.700272917001124,
+      "grad_norm": 0.9989892244338989,
+      "learning_rate": 1.5029130106070167e-05,
+      "loss": 0.3537,
+      "step": 8410
+    },
+    {
+      "epoch": 2.700593995825975,
+      "grad_norm": 1.0656063556671143,
+      "learning_rate": 1.5010730701313625e-05,
+      "loss": 0.3365,
+      "step": 8411
+    },
+    {
+      "epoch": 2.700915074650827,
+      "grad_norm": 0.9919269680976868,
+      "learning_rate": 1.4992341652241737e-05,
+      "loss": 0.3995,
+      "step": 8412
+    },
+    {
+      "epoch": 2.701236153475678,
+      "grad_norm": 0.9192927479743958,
+      "learning_rate": 1.4973962961095134e-05,
+      "loss": 0.3682,
+      "step": 8413
+    },
+    {
+      "epoch": 2.7015572323005297,
+      "grad_norm": 0.9061486124992371,
+      "learning_rate": 1.495559463011329e-05,
+      "loss": 0.3291,
+      "step": 8414
+    },
+    {
+      "epoch": 2.7018783111253812,
+      "grad_norm": 1.0932555198669434,
+      "learning_rate": 1.4937236661534226e-05,
+      "loss": 0.274,
+      "step": 8415
+    },
+    {
+      "epoch": 2.702199389950233,
+      "grad_norm": 1.2957507371902466,
+      "learning_rate": 1.4918889057594876e-05,
+      "loss": 0.2687,
+      "step": 8416
+    },
+    {
+      "epoch": 2.7025204687750843,
+      "grad_norm": 0.8411356806755066,
+      "learning_rate": 1.4900551820530828e-05,
+      "loss": 0.2988,
+      "step": 8417
+    },
+    {
+      "epoch": 2.702841547599936,
+      "grad_norm": 1.028803825378418,
+      "learning_rate": 1.4882224952576373e-05,
+      "loss": 0.393,
+      "step": 8418
+    },
+    {
+      "epoch": 2.7031626264247874,
+      "grad_norm": 1.0516469478607178,
+      "learning_rate": 1.486390845596466e-05,
+      "loss": 0.3689,
+      "step": 8419
+    },
+    {
+      "epoch": 2.7034837052496385,
+      "grad_norm": 0.5976935029029846,
+      "learning_rate": 1.484560233292741e-05,
+      "loss": 0.2572,
+      "step": 8420
+    },
+    {
+      "epoch": 2.7038047840744905,
+      "grad_norm": 1.021976113319397,
+      "learning_rate": 1.4827306585695234e-05,
+      "loss": 0.4232,
+      "step": 8421
+    },
+    {
+      "epoch": 2.7041258628993416,
+      "grad_norm": 0.8594872951507568,
+      "learning_rate": 1.4809021216497399e-05,
+      "loss": 0.3019,
+      "step": 8422
+    },
+    {
+      "epoch": 2.704446941724193,
+      "grad_norm": 0.8935600519180298,
+      "learning_rate": 1.4790746227561925e-05,
+      "loss": 0.2665,
+      "step": 8423
+    },
+    {
+      "epoch": 2.7047680205490447,
+      "grad_norm": 0.5370054244995117,
+      "learning_rate": 1.4772481621115541e-05,
+      "loss": 0.2293,
+      "step": 8424
+    },
+    {
+      "epoch": 2.7050890993738963,
+      "grad_norm": 0.42363736033439636,
+      "learning_rate": 1.4754227399383757e-05,
+      "loss": 0.2423,
+      "step": 8425
+    },
+    {
+      "epoch": 2.705410178198748,
+      "grad_norm": 0.6202889680862427,
+      "learning_rate": 1.4735983564590783e-05,
+      "loss": 0.3039,
+      "step": 8426
+    },
+    {
+      "epoch": 2.7057312570235994,
+      "grad_norm": 0.5840016603469849,
+      "learning_rate": 1.4717750118959584e-05,
+      "loss": 0.3053,
+      "step": 8427
+    },
+    {
+      "epoch": 2.706052335848451,
+      "grad_norm": 0.5942694544792175,
+      "learning_rate": 1.4699527064711838e-05,
+      "loss": 0.2565,
+      "step": 8428
+    },
+    {
+      "epoch": 2.706373414673302,
+      "grad_norm": 0.7443995475769043,
+      "learning_rate": 1.468131440406798e-05,
+      "loss": 0.3169,
+      "step": 8429
+    },
+    {
+      "epoch": 2.706694493498154,
+      "grad_norm": 0.47813400626182556,
+      "learning_rate": 1.466311213924716e-05,
+      "loss": 0.6779,
+      "step": 8430
+    },
+    {
+      "epoch": 2.707015572323005,
+      "grad_norm": 0.406895250082016,
+      "learning_rate": 1.4644920272467244e-05,
+      "loss": 0.5476,
+      "step": 8431
+    },
+    {
+      "epoch": 2.7073366511478567,
+      "grad_norm": 0.3295402228832245,
+      "learning_rate": 1.462673880594494e-05,
+      "loss": 0.155,
+      "step": 8432
+    },
+    {
+      "epoch": 2.7076577299727083,
+      "grad_norm": 0.23888343572616577,
+      "learning_rate": 1.4608567741895495e-05,
+      "loss": 0.0741,
+      "step": 8433
+    },
+    {
+      "epoch": 2.70797880879756,
+      "grad_norm": 0.20620322227478027,
+      "learning_rate": 1.4590407082533097e-05,
+      "loss": 0.0649,
+      "step": 8434
+    },
+    {
+      "epoch": 2.7082998876224114,
+      "grad_norm": 0.3761898875236511,
+      "learning_rate": 1.4572256830070497e-05,
+      "loss": 0.2023,
+      "step": 8435
+    },
+    {
+      "epoch": 2.708620966447263,
+      "grad_norm": 0.45816802978515625,
+      "learning_rate": 1.4554116986719257e-05,
+      "loss": 0.2164,
+      "step": 8436
+    },
+    {
+      "epoch": 2.7089420452721145,
+      "grad_norm": 0.8823957443237305,
+      "learning_rate": 1.4535987554689712e-05,
+      "loss": 0.5551,
+      "step": 8437
+    },
+    {
+      "epoch": 2.7092631240969656,
+      "grad_norm": 0.8065453171730042,
+      "learning_rate": 1.4517868536190803e-05,
+      "loss": 0.396,
+      "step": 8438
+    },
+    {
+      "epoch": 2.7095842029218176,
+      "grad_norm": 0.9334639310836792,
+      "learning_rate": 1.4499759933430346e-05,
+      "loss": 0.4288,
+      "step": 8439
+    },
+    {
+      "epoch": 2.7099052817466687,
+      "grad_norm": 0.9277933239936829,
+      "learning_rate": 1.4481661748614784e-05,
+      "loss": 0.3647,
+      "step": 8440
+    },
+    {
+      "epoch": 2.71022636057152,
+      "grad_norm": 0.7920669317245483,
+      "learning_rate": 1.4463573983949341e-05,
+      "loss": 0.3438,
+      "step": 8441
+    },
+    {
+      "epoch": 2.7105474393963718,
+      "grad_norm": 1.0169240236282349,
+      "learning_rate": 1.4445496641637967e-05,
+      "loss": 0.4013,
+      "step": 8442
+    },
+    {
+      "epoch": 2.7108685182212233,
+      "grad_norm": 0.83201664686203,
+      "learning_rate": 1.4427429723883257e-05,
+      "loss": 0.3804,
+      "step": 8443
+    },
+    {
+      "epoch": 2.711189597046075,
+      "grad_norm": 0.9397286772727966,
+      "learning_rate": 1.4409373232886702e-05,
+      "loss": 0.3005,
+      "step": 8444
+    },
+    {
+      "epoch": 2.7115106758709264,
+      "grad_norm": 0.6609529852867126,
+      "learning_rate": 1.439132717084839e-05,
+      "loss": 0.314,
+      "step": 8445
+    },
+    {
+      "epoch": 2.711831754695778,
+      "grad_norm": 0.8575668334960938,
+      "learning_rate": 1.4373291539967182e-05,
+      "loss": 0.3991,
+      "step": 8446
+    },
+    {
+      "epoch": 2.712152833520629,
+      "grad_norm": 1.156197428703308,
+      "learning_rate": 1.4355266342440677e-05,
+      "loss": 0.3177,
+      "step": 8447
+    },
+    {
+      "epoch": 2.712473912345481,
+      "grad_norm": 0.7570515275001526,
+      "learning_rate": 1.4337251580465172e-05,
+      "loss": 0.3598,
+      "step": 8448
+    },
+    {
+      "epoch": 2.712794991170332,
+      "grad_norm": 0.7249849438667297,
+      "learning_rate": 1.4319247256235714e-05,
+      "loss": 0.2816,
+      "step": 8449
+    },
+    {
+      "epoch": 2.7131160699951837,
+      "grad_norm": 0.9076480865478516,
+      "learning_rate": 1.4301253371946089e-05,
+      "loss": 0.3801,
+      "step": 8450
+    },
+    {
+      "epoch": 2.7134371488200353,
+      "grad_norm": 1.0600674152374268,
+      "learning_rate": 1.4283269929788779e-05,
+      "loss": 0.4087,
+      "step": 8451
+    },
+    {
+      "epoch": 2.713758227644887,
+      "grad_norm": 1.2555755376815796,
+      "learning_rate": 1.426529693195503e-05,
+      "loss": 0.5346,
+      "step": 8452
+    },
+    {
+      "epoch": 2.7140793064697384,
+      "grad_norm": 0.9253405928611755,
+      "learning_rate": 1.4247334380634792e-05,
+      "loss": 0.3705,
+      "step": 8453
+    },
+    {
+      "epoch": 2.71440038529459,
+      "grad_norm": 0.9356733560562134,
+      "learning_rate": 1.4229382278016712e-05,
+      "loss": 0.3621,
+      "step": 8454
+    },
+    {
+      "epoch": 2.7147214641194415,
+      "grad_norm": 1.4253745079040527,
+      "learning_rate": 1.4211440626288286e-05,
+      "loss": 0.4511,
+      "step": 8455
+    },
+    {
+      "epoch": 2.7150425429442926,
+      "grad_norm": 1.0171775817871094,
+      "learning_rate": 1.4193509427635543e-05,
+      "loss": 0.3103,
+      "step": 8456
+    },
+    {
+      "epoch": 2.7153636217691446,
+      "grad_norm": 1.1757081747055054,
+      "learning_rate": 1.4175588684243446e-05,
+      "loss": 0.4486,
+      "step": 8457
+    },
+    {
+      "epoch": 2.7156847005939957,
+      "grad_norm": 0.8698387742042542,
+      "learning_rate": 1.4157678398295481e-05,
+      "loss": 0.3556,
+      "step": 8458
+    },
+    {
+      "epoch": 2.7160057794188472,
+      "grad_norm": 0.8687936067581177,
+      "learning_rate": 1.4139778571974049e-05,
+      "loss": 0.3552,
+      "step": 8459
+    },
+    {
+      "epoch": 2.716326858243699,
+      "grad_norm": 0.8789449334144592,
+      "learning_rate": 1.412188920746017e-05,
+      "loss": 0.3352,
+      "step": 8460
+    },
+    {
+      "epoch": 2.7166479370685503,
+      "grad_norm": 0.8837921023368835,
+      "learning_rate": 1.4104010306933557e-05,
+      "loss": 0.3605,
+      "step": 8461
+    },
+    {
+      "epoch": 2.716969015893402,
+      "grad_norm": 0.7887502312660217,
+      "learning_rate": 1.4086141872572789e-05,
+      "loss": 0.2859,
+      "step": 8462
+    },
+    {
+      "epoch": 2.7172900947182534,
+      "grad_norm": 0.8587145805358887,
+      "learning_rate": 1.406828390655497e-05,
+      "loss": 0.3172,
+      "step": 8463
+    },
+    {
+      "epoch": 2.717611173543105,
+      "grad_norm": 0.6038246750831604,
+      "learning_rate": 1.4050436411056123e-05,
+      "loss": 0.2771,
+      "step": 8464
+    },
+    {
+      "epoch": 2.717932252367956,
+      "grad_norm": 0.7046909332275391,
+      "learning_rate": 1.40325993882509e-05,
+      "loss": 0.2853,
+      "step": 8465
+    },
+    {
+      "epoch": 2.718253331192808,
+      "grad_norm": 0.6671448945999146,
+      "learning_rate": 1.4014772840312663e-05,
+      "loss": 0.2452,
+      "step": 8466
+    },
+    {
+      "epoch": 2.718574410017659,
+      "grad_norm": 0.46538805961608887,
+      "learning_rate": 1.3996956769413538e-05,
+      "loss": 0.2233,
+      "step": 8467
+    },
+    {
+      "epoch": 2.7188954888425108,
+      "grad_norm": 0.845503032207489,
+      "learning_rate": 1.3979151177724347e-05,
+      "loss": 0.329,
+      "step": 8468
+    },
+    {
+      "epoch": 2.7192165676673623,
+      "grad_norm": 0.6878446936607361,
+      "learning_rate": 1.3961356067414666e-05,
+      "loss": 0.255,
+      "step": 8469
+    },
+    {
+      "epoch": 2.719537646492214,
+      "grad_norm": 0.9203073978424072,
+      "learning_rate": 1.3943571440652769e-05,
+      "loss": 0.3479,
+      "step": 8470
+    },
+    {
+      "epoch": 2.7198587253170654,
+      "grad_norm": 0.7054129242897034,
+      "learning_rate": 1.3925797299605647e-05,
+      "loss": 0.3034,
+      "step": 8471
+    },
+    {
+      "epoch": 2.720179804141917,
+      "grad_norm": 1.0850917100906372,
+      "learning_rate": 1.3908033646439033e-05,
+      "loss": 0.3771,
+      "step": 8472
+    },
+    {
+      "epoch": 2.7205008829667685,
+      "grad_norm": 0.6727474927902222,
+      "learning_rate": 1.3890280483317374e-05,
+      "loss": 0.2476,
+      "step": 8473
+    },
+    {
+      "epoch": 2.7208219617916196,
+      "grad_norm": 0.7375319600105286,
+      "learning_rate": 1.3872537812403829e-05,
+      "loss": 0.2564,
+      "step": 8474
+    },
+    {
+      "epoch": 2.7211430406164716,
+      "grad_norm": 0.4288458526134491,
+      "learning_rate": 1.3854805635860336e-05,
+      "loss": 0.2455,
+      "step": 8475
+    },
+    {
+      "epoch": 2.7214641194413227,
+      "grad_norm": 0.5464689135551453,
+      "learning_rate": 1.3837083955847418e-05,
+      "loss": 0.2698,
+      "step": 8476
+    },
+    {
+      "epoch": 2.7217851982661743,
+      "grad_norm": 0.6613762378692627,
+      "learning_rate": 1.3819372774524508e-05,
+      "loss": 0.2394,
+      "step": 8477
+    },
+    {
+      "epoch": 2.722106277091026,
+      "grad_norm": 0.6144346594810486,
+      "learning_rate": 1.38016720940496e-05,
+      "loss": 0.2879,
+      "step": 8478
+    },
+    {
+      "epoch": 2.7224273559158774,
+      "grad_norm": 0.4245906174182892,
+      "learning_rate": 1.3783981916579446e-05,
+      "loss": 0.2576,
+      "step": 8479
+    },
+    {
+      "epoch": 2.722748434740729,
+      "grad_norm": 0.5054341554641724,
+      "learning_rate": 1.3766302244269624e-05,
+      "loss": 0.6253,
+      "step": 8480
+    },
+    {
+      "epoch": 2.7230695135655805,
+      "grad_norm": 0.44886812567710876,
+      "learning_rate": 1.3748633079274253e-05,
+      "loss": 0.3354,
+      "step": 8481
+    },
+    {
+      "epoch": 2.723390592390432,
+      "grad_norm": 0.39985498785972595,
+      "learning_rate": 1.3730974423746335e-05,
+      "loss": 0.2347,
+      "step": 8482
+    },
+    {
+      "epoch": 2.723711671215283,
+      "grad_norm": 0.5271901488304138,
+      "learning_rate": 1.3713326279837501e-05,
+      "loss": 0.2654,
+      "step": 8483
+    },
+    {
+      "epoch": 2.724032750040135,
+      "grad_norm": 0.4039822518825531,
+      "learning_rate": 1.3695688649698124e-05,
+      "loss": 0.1506,
+      "step": 8484
+    },
+    {
+      "epoch": 2.7243538288649862,
+      "grad_norm": 0.4156002104282379,
+      "learning_rate": 1.3678061535477304e-05,
+      "loss": 0.1746,
+      "step": 8485
+    },
+    {
+      "epoch": 2.7246749076898378,
+      "grad_norm": 0.43996649980545044,
+      "learning_rate": 1.3660444939322836e-05,
+      "loss": 0.1855,
+      "step": 8486
+    },
+    {
+      "epoch": 2.7249959865146893,
+      "grad_norm": 1.112030267715454,
+      "learning_rate": 1.3642838863381257e-05,
+      "loss": 0.4318,
+      "step": 8487
+    },
+    {
+      "epoch": 2.725317065339541,
+      "grad_norm": 0.7671384215354919,
+      "learning_rate": 1.362524330979782e-05,
+      "loss": 0.3795,
+      "step": 8488
+    },
+    {
+      "epoch": 2.7256381441643924,
+      "grad_norm": 0.8614218831062317,
+      "learning_rate": 1.3607658280716473e-05,
+      "loss": 0.3608,
+      "step": 8489
+    },
+    {
+      "epoch": 2.725959222989244,
+      "grad_norm": 0.7562094330787659,
+      "learning_rate": 1.3590083778279916e-05,
+      "loss": 0.4333,
+      "step": 8490
+    },
+    {
+      "epoch": 2.7262803018140955,
+      "grad_norm": 0.6444230079650879,
+      "learning_rate": 1.3572519804629536e-05,
+      "loss": 0.2859,
+      "step": 8491
+    },
+    {
+      "epoch": 2.7266013806389466,
+      "grad_norm": 0.707561194896698,
+      "learning_rate": 1.3554966361905464e-05,
+      "loss": 0.2902,
+      "step": 8492
+    },
+    {
+      "epoch": 2.7269224594637986,
+      "grad_norm": 0.7419070601463318,
+      "learning_rate": 1.3537423452246523e-05,
+      "loss": 0.3034,
+      "step": 8493
+    },
+    {
+      "epoch": 2.7272435382886497,
+      "grad_norm": 0.9765730500221252,
+      "learning_rate": 1.3519891077790236e-05,
+      "loss": 0.3747,
+      "step": 8494
+    },
+    {
+      "epoch": 2.7275646171135013,
+      "grad_norm": 0.6046754717826843,
+      "learning_rate": 1.3502369240672941e-05,
+      "loss": 0.2038,
+      "step": 8495
+    },
+    {
+      "epoch": 2.727885695938353,
+      "grad_norm": 0.7997030019760132,
+      "learning_rate": 1.3484857943029572e-05,
+      "loss": 0.3041,
+      "step": 8496
+    },
+    {
+      "epoch": 2.7282067747632044,
+      "grad_norm": 0.879155695438385,
+      "learning_rate": 1.34673571869938e-05,
+      "loss": 0.3071,
+      "step": 8497
+    },
+    {
+      "epoch": 2.728527853588056,
+      "grad_norm": 1.1423804759979248,
+      "learning_rate": 1.3449866974698122e-05,
+      "loss": 0.3656,
+      "step": 8498
+    },
+    {
+      "epoch": 2.7288489324129075,
+      "grad_norm": 0.8787369132041931,
+      "learning_rate": 1.3432387308273575e-05,
+      "loss": 0.4357,
+      "step": 8499
+    },
+    {
+      "epoch": 2.729170011237759,
+      "grad_norm": 1.0522805452346802,
+      "learning_rate": 1.3414918189850089e-05,
+      "loss": 0.4416,
+      "step": 8500
+    },
+    {
+      "epoch": 2.72949109006261,
+      "grad_norm": 0.7044956684112549,
+      "learning_rate": 1.339745962155613e-05,
+      "loss": 0.2899,
+      "step": 8501
+    },
+    {
+      "epoch": 2.729812168887462,
+      "grad_norm": 0.7710807919502258,
+      "learning_rate": 1.338001160551906e-05,
+      "loss": 0.2822,
+      "step": 8502
+    },
+    {
+      "epoch": 2.7301332477123132,
+      "grad_norm": 1.0037212371826172,
+      "learning_rate": 1.3362574143864814e-05,
+      "loss": 0.3884,
+      "step": 8503
+    },
+    {
+      "epoch": 2.730454326537165,
+      "grad_norm": 0.7747420072555542,
+      "learning_rate": 1.3345147238718126e-05,
+      "loss": 0.3585,
+      "step": 8504
+    },
+    {
+      "epoch": 2.7307754053620164,
+      "grad_norm": 0.5802621841430664,
+      "learning_rate": 1.3327730892202383e-05,
+      "loss": 0.2472,
+      "step": 8505
+    },
+    {
+      "epoch": 2.731096484186868,
+      "grad_norm": 0.7853018641471863,
+      "learning_rate": 1.3310325106439726e-05,
+      "loss": 0.2751,
+      "step": 8506
+    },
+    {
+      "epoch": 2.7314175630117195,
+      "grad_norm": 0.5948209166526794,
+      "learning_rate": 1.3292929883550998e-05,
+      "loss": 0.2399,
+      "step": 8507
+    },
+    {
+      "epoch": 2.731738641836571,
+      "grad_norm": 1.0438854694366455,
+      "learning_rate": 1.327554522565576e-05,
+      "loss": 0.3403,
+      "step": 8508
+    },
+    {
+      "epoch": 2.7320597206614226,
+      "grad_norm": 0.662353515625,
+      "learning_rate": 1.3258171134872265e-05,
+      "loss": 0.2787,
+      "step": 8509
+    },
+    {
+      "epoch": 2.7323807994862737,
+      "grad_norm": 0.8576146960258484,
+      "learning_rate": 1.3240807613317507e-05,
+      "loss": 0.3339,
+      "step": 8510
+    },
+    {
+      "epoch": 2.7327018783111257,
+      "grad_norm": 0.8656917214393616,
+      "learning_rate": 1.3223454663107172e-05,
+      "loss": 0.3105,
+      "step": 8511
+    },
+    {
+      "epoch": 2.7330229571359768,
+      "grad_norm": 0.6916074752807617,
+      "learning_rate": 1.3206112286355632e-05,
+      "loss": 0.2808,
+      "step": 8512
+    },
+    {
+      "epoch": 2.7333440359608283,
+      "grad_norm": 0.5651912689208984,
+      "learning_rate": 1.3188780485176088e-05,
+      "loss": 0.2671,
+      "step": 8513
+    },
+    {
+      "epoch": 2.73366511478568,
+      "grad_norm": 0.7664126753807068,
+      "learning_rate": 1.3171459261680297e-05,
+      "loss": 0.316,
+      "step": 8514
+    },
+    {
+      "epoch": 2.7339861936105314,
+      "grad_norm": 0.632695198059082,
+      "learning_rate": 1.3154148617978812e-05,
+      "loss": 0.2613,
+      "step": 8515
+    },
+    {
+      "epoch": 2.734307272435383,
+      "grad_norm": 0.7974990606307983,
+      "learning_rate": 1.3136848556180892e-05,
+      "loss": 0.2982,
+      "step": 8516
+    },
+    {
+      "epoch": 2.7346283512602345,
+      "grad_norm": 0.7887236475944519,
+      "learning_rate": 1.3119559078394461e-05,
+      "loss": 0.2945,
+      "step": 8517
+    },
+    {
+      "epoch": 2.734949430085086,
+      "grad_norm": 0.8384912014007568,
+      "learning_rate": 1.3102280186726269e-05,
+      "loss": 0.3098,
+      "step": 8518
+    },
+    {
+      "epoch": 2.735270508909937,
+      "grad_norm": 0.8313494920730591,
+      "learning_rate": 1.3085011883281606e-05,
+      "loss": 0.3429,
+      "step": 8519
+    },
+    {
+      "epoch": 2.735591587734789,
+      "grad_norm": 0.7698463797569275,
+      "learning_rate": 1.3067754170164614e-05,
+      "loss": 0.276,
+      "step": 8520
+    },
+    {
+      "epoch": 2.7359126665596403,
+      "grad_norm": 0.7834967374801636,
+      "learning_rate": 1.30505070494781e-05,
+      "loss": 0.2935,
+      "step": 8521
+    },
+    {
+      "epoch": 2.736233745384492,
+      "grad_norm": 1.1544512510299683,
+      "learning_rate": 1.3033270523323549e-05,
+      "loss": 0.3743,
+      "step": 8522
+    },
+    {
+      "epoch": 2.7365548242093434,
+      "grad_norm": 1.0003769397735596,
+      "learning_rate": 1.3016044593801202e-05,
+      "loss": 0.3563,
+      "step": 8523
+    },
+    {
+      "epoch": 2.736875903034195,
+      "grad_norm": 1.167098879814148,
+      "learning_rate": 1.2998829263009938e-05,
+      "loss": 0.4428,
+      "step": 8524
+    },
+    {
+      "epoch": 2.7371969818590465,
+      "grad_norm": 0.758287250995636,
+      "learning_rate": 1.2981624533047432e-05,
+      "loss": 0.3105,
+      "step": 8525
+    },
+    {
+      "epoch": 2.737518060683898,
+      "grad_norm": 0.7774193286895752,
+      "learning_rate": 1.296443040601003e-05,
+      "loss": 0.3055,
+      "step": 8526
+    },
+    {
+      "epoch": 2.7378391395087496,
+      "grad_norm": 0.4565434455871582,
+      "learning_rate": 1.294724688399278e-05,
+      "loss": 0.2601,
+      "step": 8527
+    },
+    {
+      "epoch": 2.7381602183336007,
+      "grad_norm": 0.8566820025444031,
+      "learning_rate": 1.293007396908944e-05,
+      "loss": 0.2995,
+      "step": 8528
+    },
+    {
+      "epoch": 2.7384812971584527,
+      "grad_norm": 0.6120618581771851,
+      "learning_rate": 1.2912911663392469e-05,
+      "loss": 0.3114,
+      "step": 8529
+    },
+    {
+      "epoch": 2.738802375983304,
+      "grad_norm": 0.46445560455322266,
+      "learning_rate": 1.2895759968993048e-05,
+      "loss": 0.5424,
+      "step": 8530
+    },
+    {
+      "epoch": 2.7391234548081553,
+      "grad_norm": 0.5213553309440613,
+      "learning_rate": 1.2878618887981064e-05,
+      "loss": 0.6759,
+      "step": 8531
+    },
+    {
+      "epoch": 2.739444533633007,
+      "grad_norm": 0.38170596957206726,
+      "learning_rate": 1.28614884224451e-05,
+      "loss": 0.2189,
+      "step": 8532
+    },
+    {
+      "epoch": 2.7397656124578584,
+      "grad_norm": 0.4311400353908539,
+      "learning_rate": 1.2844368574472454e-05,
+      "loss": 0.2501,
+      "step": 8533
+    },
+    {
+      "epoch": 2.74008669128271,
+      "grad_norm": 0.23123976588249207,
+      "learning_rate": 1.2827259346149122e-05,
+      "loss": 0.0694,
+      "step": 8534
+    },
+    {
+      "epoch": 2.7404077701075615,
+      "grad_norm": 0.32238295674324036,
+      "learning_rate": 1.2810160739559796e-05,
+      "loss": 0.1146,
+      "step": 8535
+    },
+    {
+      "epoch": 2.740728848932413,
+      "grad_norm": 0.1789664924144745,
+      "learning_rate": 1.279307275678795e-05,
+      "loss": 0.063,
+      "step": 8536
+    },
+    {
+      "epoch": 2.741049927757264,
+      "grad_norm": 0.7732048630714417,
+      "learning_rate": 1.2775995399915631e-05,
+      "loss": 0.4496,
+      "step": 8537
+    },
+    {
+      "epoch": 2.7413710065821157,
+      "grad_norm": 1.2547789812088013,
+      "learning_rate": 1.2758928671023718e-05,
+      "loss": 0.4711,
+      "step": 8538
+    },
+    {
+      "epoch": 2.7416920854069673,
+      "grad_norm": 0.7271549105644226,
+      "learning_rate": 1.2741872572191682e-05,
+      "loss": 0.3086,
+      "step": 8539
+    },
+    {
+      "epoch": 2.742013164231819,
+      "grad_norm": 0.7055025696754456,
+      "learning_rate": 1.2724827105497816e-05,
+      "loss": 0.3016,
+      "step": 8540
+    },
+    {
+      "epoch": 2.7423342430566704,
+      "grad_norm": 0.7456016540527344,
+      "learning_rate": 1.2707792273019048e-05,
+      "loss": 0.3477,
+      "step": 8541
+    },
+    {
+      "epoch": 2.742655321881522,
+      "grad_norm": 0.9478852152824402,
+      "learning_rate": 1.2690768076830972e-05,
+      "loss": 0.302,
+      "step": 8542
+    },
+    {
+      "epoch": 2.7429764007063735,
+      "grad_norm": 0.7163116335868835,
+      "learning_rate": 1.2673754519008008e-05,
+      "loss": 0.3128,
+      "step": 8543
+    },
+    {
+      "epoch": 2.743297479531225,
+      "grad_norm": 0.6127589344978333,
+      "learning_rate": 1.2656751601623118e-05,
+      "loss": 0.2145,
+      "step": 8544
+    },
+    {
+      "epoch": 2.7436185583560766,
+      "grad_norm": 0.8434281349182129,
+      "learning_rate": 1.2639759326748135e-05,
+      "loss": 0.2788,
+      "step": 8545
+    },
+    {
+      "epoch": 2.7439396371809277,
+      "grad_norm": 0.6575958728790283,
+      "learning_rate": 1.262277769645348e-05,
+      "loss": 0.2678,
+      "step": 8546
+    },
+    {
+      "epoch": 2.7442607160057793,
+      "grad_norm": 0.9952372312545776,
+      "learning_rate": 1.260580671280832e-05,
+      "loss": 0.3722,
+      "step": 8547
+    },
+    {
+      "epoch": 2.744581794830631,
+      "grad_norm": 1.0888564586639404,
+      "learning_rate": 1.2588846377880525e-05,
+      "loss": 0.4193,
+      "step": 8548
+    },
+    {
+      "epoch": 2.7449028736554824,
+      "grad_norm": 0.8872905373573303,
+      "learning_rate": 1.257189669373664e-05,
+      "loss": 0.3686,
+      "step": 8549
+    },
+    {
+      "epoch": 2.745223952480334,
+      "grad_norm": 0.8427690267562866,
+      "learning_rate": 1.2554957662441957e-05,
+      "loss": 0.3544,
+      "step": 8550
+    },
+    {
+      "epoch": 2.7455450313051855,
+      "grad_norm": 0.8962048888206482,
+      "learning_rate": 1.2538029286060426e-05,
+      "loss": 0.3592,
+      "step": 8551
+    },
+    {
+      "epoch": 2.745866110130037,
+      "grad_norm": 0.9221064448356628,
+      "learning_rate": 1.2521111566654731e-05,
+      "loss": 0.3621,
+      "step": 8552
+    },
+    {
+      "epoch": 2.7461871889548886,
+      "grad_norm": 1.0115526914596558,
+      "learning_rate": 1.2504204506286243e-05,
+      "loss": 0.4147,
+      "step": 8553
+    },
+    {
+      "epoch": 2.74650826777974,
+      "grad_norm": 0.8102017045021057,
+      "learning_rate": 1.2487308107015027e-05,
+      "loss": 0.3009,
+      "step": 8554
+    },
+    {
+      "epoch": 2.746829346604591,
+      "grad_norm": 1.0300239324569702,
+      "learning_rate": 1.2470422370899838e-05,
+      "loss": 0.3574,
+      "step": 8555
+    },
+    {
+      "epoch": 2.7471504254294428,
+      "grad_norm": 1.0161161422729492,
+      "learning_rate": 1.2453547299998225e-05,
+      "loss": 0.3415,
+      "step": 8556
+    },
+    {
+      "epoch": 2.7474715042542943,
+      "grad_norm": 1.0009255409240723,
+      "learning_rate": 1.243668289636628e-05,
+      "loss": 0.3357,
+      "step": 8557
+    },
+    {
+      "epoch": 2.747792583079146,
+      "grad_norm": 0.9013484120368958,
+      "learning_rate": 1.241982916205895e-05,
+      "loss": 0.354,
+      "step": 8558
+    },
+    {
+      "epoch": 2.7481136619039974,
+      "grad_norm": 0.9857357740402222,
+      "learning_rate": 1.2402986099129765e-05,
+      "loss": 0.3988,
+      "step": 8559
+    },
+    {
+      "epoch": 2.748434740728849,
+      "grad_norm": 0.8265845775604248,
+      "learning_rate": 1.2386153709630988e-05,
+      "loss": 0.3291,
+      "step": 8560
+    },
+    {
+      "epoch": 2.7487558195537005,
+      "grad_norm": 1.0555880069732666,
+      "learning_rate": 1.2369331995613665e-05,
+      "loss": 0.3589,
+      "step": 8561
+    },
+    {
+      "epoch": 2.749076898378552,
+      "grad_norm": 0.9744288921356201,
+      "learning_rate": 1.2352520959127379e-05,
+      "loss": 0.3236,
+      "step": 8562
+    },
+    {
+      "epoch": 2.7493979772034036,
+      "grad_norm": 0.7760259509086609,
+      "learning_rate": 1.2335720602220569e-05,
+      "loss": 0.2423,
+      "step": 8563
+    },
+    {
+      "epoch": 2.7497190560282547,
+      "grad_norm": 1.1198222637176514,
+      "learning_rate": 1.2318930926940298e-05,
+      "loss": 0.3958,
+      "step": 8564
+    },
+    {
+      "epoch": 2.7500401348531063,
+      "grad_norm": 1.1928644180297852,
+      "learning_rate": 1.2302151935332329e-05,
+      "loss": 0.4002,
+      "step": 8565
+    },
+    {
+      "epoch": 2.750361213677958,
+      "grad_norm": 0.7599969506263733,
+      "learning_rate": 1.228538362944115e-05,
+      "loss": 0.2612,
+      "step": 8566
+    },
+    {
+      "epoch": 2.7506822925028094,
+      "grad_norm": 1.1186960935592651,
+      "learning_rate": 1.2268626011309858e-05,
+      "loss": 0.3827,
+      "step": 8567
+    },
+    {
+      "epoch": 2.751003371327661,
+      "grad_norm": 0.6915915608406067,
+      "learning_rate": 1.22518790829804e-05,
+      "loss": 0.2565,
+      "step": 8568
+    },
+    {
+      "epoch": 2.7513244501525125,
+      "grad_norm": 0.7428755164146423,
+      "learning_rate": 1.2235142846493308e-05,
+      "loss": 0.2987,
+      "step": 8569
+    },
+    {
+      "epoch": 2.751645528977364,
+      "grad_norm": 0.7877930998802185,
+      "learning_rate": 1.2218417303887842e-05,
+      "loss": 0.3379,
+      "step": 8570
+    },
+    {
+      "epoch": 2.7519666078022156,
+      "grad_norm": 0.9466036558151245,
+      "learning_rate": 1.2201702457201947e-05,
+      "loss": 0.3894,
+      "step": 8571
+    },
+    {
+      "epoch": 2.752287686627067,
+      "grad_norm": 0.8674002289772034,
+      "learning_rate": 1.2184998308472295e-05,
+      "loss": 0.3414,
+      "step": 8572
+    },
+    {
+      "epoch": 2.7526087654519182,
+      "grad_norm": 0.8166511654853821,
+      "learning_rate": 1.2168304859734226e-05,
+      "loss": 0.3156,
+      "step": 8573
+    },
+    {
+      "epoch": 2.75292984427677,
+      "grad_norm": 0.7151662707328796,
+      "learning_rate": 1.2151622113021787e-05,
+      "loss": 0.3031,
+      "step": 8574
+    },
+    {
+      "epoch": 2.7532509231016213,
+      "grad_norm": 0.7391020655632019,
+      "learning_rate": 1.2134950070367723e-05,
+      "loss": 0.2881,
+      "step": 8575
+    },
+    {
+      "epoch": 2.753572001926473,
+      "grad_norm": 0.8440263867378235,
+      "learning_rate": 1.2118288733803473e-05,
+      "loss": 0.3104,
+      "step": 8576
+    },
+    {
+      "epoch": 2.7538930807513244,
+      "grad_norm": 0.8688570261001587,
+      "learning_rate": 1.2101638105359169e-05,
+      "loss": 0.3327,
+      "step": 8577
+    },
+    {
+      "epoch": 2.754214159576176,
+      "grad_norm": 0.35422104597091675,
+      "learning_rate": 1.2084998187063613e-05,
+      "loss": 0.2317,
+      "step": 8578
+    },
+    {
+      "epoch": 2.7545352384010275,
+      "grad_norm": 0.4497572183609009,
+      "learning_rate": 1.206836898094439e-05,
+      "loss": 0.2845,
+      "step": 8579
+    },
+    {
+      "epoch": 2.754856317225879,
+      "grad_norm": 0.48799851536750793,
+      "learning_rate": 1.2051750489027647e-05,
+      "loss": 0.6289,
+      "step": 8580
+    },
+    {
+      "epoch": 2.7551773960507306,
+      "grad_norm": 0.438763827085495,
+      "learning_rate": 1.2035142713338366e-05,
+      "loss": 0.5615,
+      "step": 8581
+    },
+    {
+      "epoch": 2.7554984748755817,
+      "grad_norm": 0.4711540639400482,
+      "learning_rate": 1.2018545655900081e-05,
+      "loss": 0.3515,
+      "step": 8582
+    },
+    {
+      "epoch": 2.7558195537004333,
+      "grad_norm": 0.43280723690986633,
+      "learning_rate": 1.2001959318735156e-05,
+      "loss": 0.16,
+      "step": 8583
+    },
+    {
+      "epoch": 2.756140632525285,
+      "grad_norm": 0.3058044910430908,
+      "learning_rate": 1.1985383703864583e-05,
+      "loss": 0.1416,
+      "step": 8584
+    },
+    {
+      "epoch": 2.7564617113501364,
+      "grad_norm": 0.15330089628696442,
+      "learning_rate": 1.196881881330798e-05,
+      "loss": 0.06,
+      "step": 8585
+    },
+    {
+      "epoch": 2.756782790174988,
+      "grad_norm": 0.45798158645629883,
+      "learning_rate": 1.19522646490838e-05,
+      "loss": 0.1952,
+      "step": 8586
+    },
+    {
+      "epoch": 2.7571038689998395,
+      "grad_norm": 0.8995585441589355,
+      "learning_rate": 1.1935721213209105e-05,
+      "loss": 0.6017,
+      "step": 8587
+    },
+    {
+      "epoch": 2.757424947824691,
+      "grad_norm": 0.8658804297447205,
+      "learning_rate": 1.191918850769964e-05,
+      "loss": 0.4147,
+      "step": 8588
+    },
+    {
+      "epoch": 2.7577460266495426,
+      "grad_norm": 0.7916271090507507,
+      "learning_rate": 1.1902666534569883e-05,
+      "loss": 0.3732,
+      "step": 8589
+    },
+    {
+      "epoch": 2.758067105474394,
+      "grad_norm": 1.541236162185669,
+      "learning_rate": 1.188615529583299e-05,
+      "loss": 0.4061,
+      "step": 8590
+    },
+    {
+      "epoch": 2.7583881842992453,
+      "grad_norm": 0.7521135210990906,
+      "learning_rate": 1.1869654793500784e-05,
+      "loss": 0.3043,
+      "step": 8591
+    },
+    {
+      "epoch": 2.758709263124097,
+      "grad_norm": 0.7983594536781311,
+      "learning_rate": 1.1853165029583823e-05,
+      "loss": 0.3089,
+      "step": 8592
+    },
+    {
+      "epoch": 2.7590303419489484,
+      "grad_norm": 0.6637342572212219,
+      "learning_rate": 1.1836686006091313e-05,
+      "loss": 0.2742,
+      "step": 8593
+    },
+    {
+      "epoch": 2.7593514207738,
+      "grad_norm": 0.8131726384162903,
+      "learning_rate": 1.1820217725031191e-05,
+      "loss": 0.3027,
+      "step": 8594
+    },
+    {
+      "epoch": 2.7596724995986515,
+      "grad_norm": 0.6360102891921997,
+      "learning_rate": 1.1803760188410073e-05,
+      "loss": 0.2646,
+      "step": 8595
+    },
+    {
+      "epoch": 2.759993578423503,
+      "grad_norm": 0.7749152183532715,
+      "learning_rate": 1.1787313398233235e-05,
+      "loss": 0.3449,
+      "step": 8596
+    },
+    {
+      "epoch": 2.7603146572483546,
+      "grad_norm": 0.9877920746803284,
+      "learning_rate": 1.1770877356504683e-05,
+      "loss": 0.3666,
+      "step": 8597
+    },
+    {
+      "epoch": 2.760635736073206,
+      "grad_norm": 0.9534462094306946,
+      "learning_rate": 1.1754452065227084e-05,
+      "loss": 0.3528,
+      "step": 8598
+    },
+    {
+      "epoch": 2.7609568148980577,
+      "grad_norm": 1.1205729246139526,
+      "learning_rate": 1.1738037526401857e-05,
+      "loss": 0.4691,
+      "step": 8599
+    },
+    {
+      "epoch": 2.7612778937229088,
+      "grad_norm": 0.7297707200050354,
+      "learning_rate": 1.1721633742028992e-05,
+      "loss": 0.2858,
+      "step": 8600
+    },
+    {
+      "epoch": 2.7615989725477603,
+      "grad_norm": 0.8408207297325134,
+      "learning_rate": 1.1705240714107302e-05,
+      "loss": 0.3679,
+      "step": 8601
+    },
+    {
+      "epoch": 2.761920051372612,
+      "grad_norm": 1.0112425088882446,
+      "learning_rate": 1.168885844463422e-05,
+      "loss": 0.3732,
+      "step": 8602
+    },
+    {
+      "epoch": 2.7622411301974634,
+      "grad_norm": 1.013698697090149,
+      "learning_rate": 1.1672486935605831e-05,
+      "loss": 0.3725,
+      "step": 8603
+    },
+    {
+      "epoch": 2.762562209022315,
+      "grad_norm": 1.6599117517471313,
+      "learning_rate": 1.1656126189017014e-05,
+      "loss": 0.3795,
+      "step": 8604
+    },
+    {
+      "epoch": 2.7628832878471665,
+      "grad_norm": 0.8416186571121216,
+      "learning_rate": 1.1639776206861196e-05,
+      "loss": 0.2958,
+      "step": 8605
+    },
+    {
+      "epoch": 2.763204366672018,
+      "grad_norm": 0.892895519733429,
+      "learning_rate": 1.1623436991130654e-05,
+      "loss": 0.3505,
+      "step": 8606
+    },
+    {
+      "epoch": 2.7635254454968696,
+      "grad_norm": 0.7338150143623352,
+      "learning_rate": 1.1607108543816248e-05,
+      "loss": 0.3204,
+      "step": 8607
+    },
+    {
+      "epoch": 2.763846524321721,
+      "grad_norm": 0.9701830744743347,
+      "learning_rate": 1.159079086690753e-05,
+      "loss": 0.3527,
+      "step": 8608
+    },
+    {
+      "epoch": 2.7641676031465723,
+      "grad_norm": 1.0651260614395142,
+      "learning_rate": 1.1574483962392767e-05,
+      "loss": 0.3822,
+      "step": 8609
+    },
+    {
+      "epoch": 2.764488681971424,
+      "grad_norm": 0.7772603034973145,
+      "learning_rate": 1.1558187832258926e-05,
+      "loss": 0.3063,
+      "step": 8610
+    },
+    {
+      "epoch": 2.7648097607962754,
+      "grad_norm": 0.5828967690467834,
+      "learning_rate": 1.1541902478491606e-05,
+      "loss": 0.2822,
+      "step": 8611
+    },
+    {
+      "epoch": 2.765130839621127,
+      "grad_norm": 0.8578951954841614,
+      "learning_rate": 1.1525627903075165e-05,
+      "loss": 0.2906,
+      "step": 8612
+    },
+    {
+      "epoch": 2.7654519184459785,
+      "grad_norm": 0.7797196507453918,
+      "learning_rate": 1.1509364107992583e-05,
+      "loss": 0.2938,
+      "step": 8613
+    },
+    {
+      "epoch": 2.76577299727083,
+      "grad_norm": 0.892012894153595,
+      "learning_rate": 1.1493111095225562e-05,
+      "loss": 0.3103,
+      "step": 8614
+    },
+    {
+      "epoch": 2.7660940760956816,
+      "grad_norm": 0.747068464756012,
+      "learning_rate": 1.1476868866754486e-05,
+      "loss": 0.2878,
+      "step": 8615
+    },
+    {
+      "epoch": 2.766415154920533,
+      "grad_norm": 0.9010385274887085,
+      "learning_rate": 1.1460637424558407e-05,
+      "loss": 0.3162,
+      "step": 8616
+    },
+    {
+      "epoch": 2.7667362337453847,
+      "grad_norm": 0.9162553548812866,
+      "learning_rate": 1.1444416770615118e-05,
+      "loss": 0.3087,
+      "step": 8617
+    },
+    {
+      "epoch": 2.767057312570236,
+      "grad_norm": 0.7222566604614258,
+      "learning_rate": 1.1428206906900995e-05,
+      "loss": 0.2889,
+      "step": 8618
+    },
+    {
+      "epoch": 2.7673783913950873,
+      "grad_norm": 0.5064111948013306,
+      "learning_rate": 1.1412007835391236e-05,
+      "loss": 0.2399,
+      "step": 8619
+    },
+    {
+      "epoch": 2.767699470219939,
+      "grad_norm": 0.6242873072624207,
+      "learning_rate": 1.1395819558059572e-05,
+      "loss": 0.2545,
+      "step": 8620
+    },
+    {
+      "epoch": 2.7680205490447904,
+      "grad_norm": 1.012289047241211,
+      "learning_rate": 1.1379642076878527e-05,
+      "loss": 0.3428,
+      "step": 8621
+    },
+    {
+      "epoch": 2.768341627869642,
+      "grad_norm": 0.7770035862922668,
+      "learning_rate": 1.1363475393819311e-05,
+      "loss": 0.3126,
+      "step": 8622
+    },
+    {
+      "epoch": 2.7686627066944935,
+      "grad_norm": 0.8441600799560547,
+      "learning_rate": 1.1347319510851717e-05,
+      "loss": 0.2974,
+      "step": 8623
+    },
+    {
+      "epoch": 2.768983785519345,
+      "grad_norm": 1.1390506029129028,
+      "learning_rate": 1.1331174429944347e-05,
+      "loss": 0.3347,
+      "step": 8624
+    },
+    {
+      "epoch": 2.7693048643441966,
+      "grad_norm": 0.9870839715003967,
+      "learning_rate": 1.1315040153064416e-05,
+      "loss": 0.4109,
+      "step": 8625
+    },
+    {
+      "epoch": 2.769625943169048,
+      "grad_norm": 0.9025225043296814,
+      "learning_rate": 1.129891668217783e-05,
+      "loss": 0.3347,
+      "step": 8626
+    },
+    {
+      "epoch": 2.7699470219938993,
+      "grad_norm": 0.6269119381904602,
+      "learning_rate": 1.1282804019249182e-05,
+      "loss": 0.2896,
+      "step": 8627
+    },
+    {
+      "epoch": 2.770268100818751,
+      "grad_norm": 0.30881941318511963,
+      "learning_rate": 1.1266702166241772e-05,
+      "loss": 0.2314,
+      "step": 8628
+    },
+    {
+      "epoch": 2.7705891796436024,
+      "grad_norm": 0.4127409756183624,
+      "learning_rate": 1.1250611125117527e-05,
+      "loss": 0.2654,
+      "step": 8629
+    },
+    {
+      "epoch": 2.770910258468454,
+      "grad_norm": 0.49982887506484985,
+      "learning_rate": 1.1234530897837126e-05,
+      "loss": 0.7763,
+      "step": 8630
+    },
+    {
+      "epoch": 2.7712313372933055,
+      "grad_norm": 0.3858592212200165,
+      "learning_rate": 1.1218461486359877e-05,
+      "loss": 0.3977,
+      "step": 8631
+    },
+    {
+      "epoch": 2.771552416118157,
+      "grad_norm": 0.5631637573242188,
+      "learning_rate": 1.1202402892643781e-05,
+      "loss": 0.3626,
+      "step": 8632
+    },
+    {
+      "epoch": 2.7718734949430086,
+      "grad_norm": 0.4585270881652832,
+      "learning_rate": 1.1186355118645554e-05,
+      "loss": 0.2111,
+      "step": 8633
+    },
+    {
+      "epoch": 2.77219457376786,
+      "grad_norm": 0.4266338646411896,
+      "learning_rate": 1.1170318166320548e-05,
+      "loss": 0.1454,
+      "step": 8634
+    },
+    {
+      "epoch": 2.7725156525927117,
+      "grad_norm": 0.3712063729763031,
+      "learning_rate": 1.1154292037622838e-05,
+      "loss": 0.1819,
+      "step": 8635
+    },
+    {
+      "epoch": 2.772836731417563,
+      "grad_norm": 0.7830849289894104,
+      "learning_rate": 1.1138276734505104e-05,
+      "loss": 0.4185,
+      "step": 8636
+    },
+    {
+      "epoch": 2.7731578102424144,
+      "grad_norm": 0.7869873046875,
+      "learning_rate": 1.1122272258918865e-05,
+      "loss": 0.3898,
+      "step": 8637
+    },
+    {
+      "epoch": 2.773478889067266,
+      "grad_norm": 0.9816368222236633,
+      "learning_rate": 1.1106278612814125e-05,
+      "loss": 0.3876,
+      "step": 8638
+    },
+    {
+      "epoch": 2.7737999678921175,
+      "grad_norm": 0.8416398167610168,
+      "learning_rate": 1.1090295798139672e-05,
+      "loss": 0.3797,
+      "step": 8639
+    },
+    {
+      "epoch": 2.774121046716969,
+      "grad_norm": 0.7030868530273438,
+      "learning_rate": 1.1074323816843024e-05,
+      "loss": 0.3055,
+      "step": 8640
+    },
+    {
+      "epoch": 2.7744421255418206,
+      "grad_norm": 0.6922499537467957,
+      "learning_rate": 1.1058362670870249e-05,
+      "loss": 0.2845,
+      "step": 8641
+    },
+    {
+      "epoch": 2.774763204366672,
+      "grad_norm": 0.6932337284088135,
+      "learning_rate": 1.1042412362166222e-05,
+      "loss": 0.3024,
+      "step": 8642
+    },
+    {
+      "epoch": 2.7750842831915237,
+      "grad_norm": 0.6935373544692993,
+      "learning_rate": 1.1026472892674378e-05,
+      "loss": 0.2872,
+      "step": 8643
+    },
+    {
+      "epoch": 2.775405362016375,
+      "grad_norm": 0.7587444186210632,
+      "learning_rate": 1.1010544264336942e-05,
+      "loss": 0.3455,
+      "step": 8644
+    },
+    {
+      "epoch": 2.7757264408412263,
+      "grad_norm": 0.6532779335975647,
+      "learning_rate": 1.099462647909475e-05,
+      "loss": 0.2364,
+      "step": 8645
+    },
+    {
+      "epoch": 2.776047519666078,
+      "grad_norm": 0.923891007900238,
+      "learning_rate": 1.0978719538887349e-05,
+      "loss": 0.3352,
+      "step": 8646
+    },
+    {
+      "epoch": 2.7763685984909294,
+      "grad_norm": 0.6212542057037354,
+      "learning_rate": 1.0962823445652959e-05,
+      "loss": 0.2283,
+      "step": 8647
+    },
+    {
+      "epoch": 2.776689677315781,
+      "grad_norm": 1.0696156024932861,
+      "learning_rate": 1.0946938201328416e-05,
+      "loss": 0.4078,
+      "step": 8648
+    },
+    {
+      "epoch": 2.7770107561406325,
+      "grad_norm": 0.7603042125701904,
+      "learning_rate": 1.093106380784934e-05,
+      "loss": 0.324,
+      "step": 8649
+    },
+    {
+      "epoch": 2.777331834965484,
+      "grad_norm": 0.9371538162231445,
+      "learning_rate": 1.0915200267149972e-05,
+      "loss": 0.392,
+      "step": 8650
+    },
+    {
+      "epoch": 2.7776529137903356,
+      "grad_norm": 0.7637982368469238,
+      "learning_rate": 1.0899347581163221e-05,
+      "loss": 0.3077,
+      "step": 8651
+    },
+    {
+      "epoch": 2.777973992615187,
+      "grad_norm": 1.0060869455337524,
+      "learning_rate": 1.08835057518207e-05,
+      "loss": 0.3866,
+      "step": 8652
+    },
+    {
+      "epoch": 2.7782950714400387,
+      "grad_norm": 0.8908117413520813,
+      "learning_rate": 1.0867674781052684e-05,
+      "loss": 0.3209,
+      "step": 8653
+    },
+    {
+      "epoch": 2.77861615026489,
+      "grad_norm": 1.2514095306396484,
+      "learning_rate": 1.0851854670788108e-05,
+      "loss": 0.422,
+      "step": 8654
+    },
+    {
+      "epoch": 2.7789372290897414,
+      "grad_norm": 0.9697837233543396,
+      "learning_rate": 1.0836045422954666e-05,
+      "loss": 0.3297,
+      "step": 8655
+    },
+    {
+      "epoch": 2.779258307914593,
+      "grad_norm": 0.8684007525444031,
+      "learning_rate": 1.0820247039478604e-05,
+      "loss": 0.3689,
+      "step": 8656
+    },
+    {
+      "epoch": 2.7795793867394445,
+      "grad_norm": 1.1245943307876587,
+      "learning_rate": 1.0804459522284926e-05,
+      "loss": 0.3227,
+      "step": 8657
+    },
+    {
+      "epoch": 2.779900465564296,
+      "grad_norm": 0.9986475706100464,
+      "learning_rate": 1.0788682873297307e-05,
+      "loss": 0.3325,
+      "step": 8658
+    },
+    {
+      "epoch": 2.7802215443891476,
+      "grad_norm": 0.8092026114463806,
+      "learning_rate": 1.0772917094438051e-05,
+      "loss": 0.3339,
+      "step": 8659
+    },
+    {
+      "epoch": 2.780542623213999,
+      "grad_norm": 0.6285796165466309,
+      "learning_rate": 1.0757162187628222e-05,
+      "loss": 0.2289,
+      "step": 8660
+    },
+    {
+      "epoch": 2.7808637020388507,
+      "grad_norm": 0.81215900182724,
+      "learning_rate": 1.0741418154787442e-05,
+      "loss": 0.3342,
+      "step": 8661
+    },
+    {
+      "epoch": 2.7811847808637022,
+      "grad_norm": 0.7589025497436523,
+      "learning_rate": 1.0725684997834162e-05,
+      "loss": 0.3424,
+      "step": 8662
+    },
+    {
+      "epoch": 2.7815058596885534,
+      "grad_norm": 0.7941707968711853,
+      "learning_rate": 1.0709962718685318e-05,
+      "loss": 0.342,
+      "step": 8663
+    },
+    {
+      "epoch": 2.781826938513405,
+      "grad_norm": 1.3057998418807983,
+      "learning_rate": 1.0694251319256687e-05,
+      "loss": 0.3437,
+      "step": 8664
+    },
+    {
+      "epoch": 2.7821480173382565,
+      "grad_norm": 1.0559587478637695,
+      "learning_rate": 1.067855080146266e-05,
+      "loss": 0.3157,
+      "step": 8665
+    },
+    {
+      "epoch": 2.782469096163108,
+      "grad_norm": 1.1683728694915771,
+      "learning_rate": 1.0662861167216243e-05,
+      "loss": 0.3433,
+      "step": 8666
+    },
+    {
+      "epoch": 2.7827901749879596,
+      "grad_norm": 0.8493248820304871,
+      "learning_rate": 1.0647182418429225e-05,
+      "loss": 0.3193,
+      "step": 8667
+    },
+    {
+      "epoch": 2.783111253812811,
+      "grad_norm": 1.0568920373916626,
+      "learning_rate": 1.063151455701199e-05,
+      "loss": 0.3286,
+      "step": 8668
+    },
+    {
+      "epoch": 2.7834323326376627,
+      "grad_norm": NaN,
+      "learning_rate": 1.063151455701199e-05,
+      "loss": 0.3763,
+      "step": 8669
+    },
+    {
+      "epoch": 2.783753411462514,
+      "grad_norm": 0.562329888343811,
+      "learning_rate": 1.0615857584873623e-05,
+      "loss": 0.2469,
+      "step": 8670
+    },
+    {
+      "epoch": 2.7840744902873658,
+      "grad_norm": 1.2342668771743774,
+      "learning_rate": 1.0600211503921887e-05,
+      "loss": 0.4084,
+      "step": 8671
+    },
+    {
+      "epoch": 2.784395569112217,
+      "grad_norm": 0.6708762645721436,
+      "learning_rate": 1.0584576316063188e-05,
+      "loss": 0.2886,
+      "step": 8672
+    },
+    {
+      "epoch": 2.7847166479370684,
+      "grad_norm": 0.7978615164756775,
+      "learning_rate": 1.0568952023202638e-05,
+      "loss": 0.3324,
+      "step": 8673
+    },
+    {
+      "epoch": 2.78503772676192,
+      "grad_norm": 0.4819890260696411,
+      "learning_rate": 1.0553338627244025e-05,
+      "loss": 0.2395,
+      "step": 8674
+    },
+    {
+      "epoch": 2.7853588055867715,
+      "grad_norm": 0.7550230026245117,
+      "learning_rate": 1.053773613008977e-05,
+      "loss": 0.2846,
+      "step": 8675
+    },
+    {
+      "epoch": 2.785679884411623,
+      "grad_norm": 0.5349231958389282,
+      "learning_rate": 1.0522144533640998e-05,
+      "loss": 0.2853,
+      "step": 8676
+    },
+    {
+      "epoch": 2.7860009632364746,
+      "grad_norm": 1.028280258178711,
+      "learning_rate": 1.0506563839797501e-05,
+      "loss": 0.3143,
+      "step": 8677
+    },
+    {
+      "epoch": 2.786322042061326,
+      "grad_norm": 0.6815648674964905,
+      "learning_rate": 1.0490994050457748e-05,
+      "loss": 0.2822,
+      "step": 8678
+    },
+    {
+      "epoch": 2.7866431208861777,
+      "grad_norm": 0.6263010501861572,
+      "learning_rate": 1.0475435167518843e-05,
+      "loss": 0.3049,
+      "step": 8679
+    },
+    {
+      "epoch": 2.7869641997110293,
+      "grad_norm": 0.5349334478378296,
+      "learning_rate": 1.0459887192876594e-05,
+      "loss": 0.6824,
+      "step": 8680
+    },
+    {
+      "epoch": 2.7872852785358804,
+      "grad_norm": 0.4282764494419098,
+      "learning_rate": 1.0444350128425529e-05,
+      "loss": 0.4408,
+      "step": 8681
+    },
+    {
+      "epoch": 2.787606357360732,
+      "grad_norm": 0.4315713346004486,
+      "learning_rate": 1.042882397605871e-05,
+      "loss": 0.4115,
+      "step": 8682
+    },
+    {
+      "epoch": 2.7879274361855835,
+      "grad_norm": 0.4827350974082947,
+      "learning_rate": 1.0413308737668005e-05,
+      "loss": 0.1461,
+      "step": 8683
+    },
+    {
+      "epoch": 2.788248515010435,
+      "grad_norm": 0.28072378039360046,
+      "learning_rate": 1.0397804415143909e-05,
+      "loss": 0.1271,
+      "step": 8684
+    },
+    {
+      "epoch": 2.7885695938352866,
+      "grad_norm": 0.3307070732116699,
+      "learning_rate": 1.0382311010375512e-05,
+      "loss": 0.1209,
+      "step": 8685
+    },
+    {
+      "epoch": 2.788890672660138,
+      "grad_norm": 0.18453341722488403,
+      "learning_rate": 1.0366828525250726e-05,
+      "loss": 0.0659,
+      "step": 8686
+    },
+    {
+      "epoch": 2.7892117514849897,
+      "grad_norm": 0.5584306716918945,
+      "learning_rate": 1.0351356961655945e-05,
+      "loss": 0.339,
+      "step": 8687
+    },
+    {
+      "epoch": 2.7895328303098412,
+      "grad_norm": 0.9177886247634888,
+      "learning_rate": 1.0335896321476413e-05,
+      "loss": 0.4866,
+      "step": 8688
+    },
+    {
+      "epoch": 2.7898539091346928,
+      "grad_norm": 1.1755162477493286,
+      "learning_rate": 1.0320446606595934e-05,
+      "loss": 0.5634,
+      "step": 8689
+    },
+    {
+      "epoch": 2.790174987959544,
+      "grad_norm": 0.9318304657936096,
+      "learning_rate": 1.0305007818897006e-05,
+      "loss": 0.3891,
+      "step": 8690
+    },
+    {
+      "epoch": 2.7904960667843954,
+      "grad_norm": 0.7948139309883118,
+      "learning_rate": 1.028957996026081e-05,
+      "loss": 0.3518,
+      "step": 8691
+    },
+    {
+      "epoch": 2.790817145609247,
+      "grad_norm": 0.9322606921195984,
+      "learning_rate": 1.0274163032567163e-05,
+      "loss": 0.3725,
+      "step": 8692
+    },
+    {
+      "epoch": 2.7911382244340985,
+      "grad_norm": 0.8909361958503723,
+      "learning_rate": 1.0258757037694589e-05,
+      "loss": 0.2977,
+      "step": 8693
+    },
+    {
+      "epoch": 2.79145930325895,
+      "grad_norm": 0.964710533618927,
+      "learning_rate": 1.0243361977520249e-05,
+      "loss": 0.4172,
+      "step": 8694
+    },
+    {
+      "epoch": 2.7917803820838016,
+      "grad_norm": 0.9714322686195374,
+      "learning_rate": 1.022797785392e-05,
+      "loss": 0.3906,
+      "step": 8695
+    },
+    {
+      "epoch": 2.792101460908653,
+      "grad_norm": 0.7900899648666382,
+      "learning_rate": 1.0212604668768343e-05,
+      "loss": 0.2676,
+      "step": 8696
+    },
+    {
+      "epoch": 2.7924225397335047,
+      "grad_norm": 0.7134426236152649,
+      "learning_rate": 1.0197242423938446e-05,
+      "loss": 0.2615,
+      "step": 8697
+    },
+    {
+      "epoch": 2.7927436185583563,
+      "grad_norm": 0.9049974679946899,
+      "learning_rate": 1.0181891121302145e-05,
+      "loss": 0.2837,
+      "step": 8698
+    },
+    {
+      "epoch": 2.7930646973832074,
+      "grad_norm": 0.9333779215812683,
+      "learning_rate": 1.0166550762729998e-05,
+      "loss": 0.2789,
+      "step": 8699
+    },
+    {
+      "epoch": 2.793385776208059,
+      "grad_norm": 0.8934162855148315,
+      "learning_rate": 1.0151221350091134e-05,
+      "loss": 0.3497,
+      "step": 8700
+    },
+    {
+      "epoch": 2.7937068550329105,
+      "grad_norm": 0.6106538772583008,
+      "learning_rate": 1.0135902885253401e-05,
+      "loss": 0.2565,
+      "step": 8701
+    },
+    {
+      "epoch": 2.794027933857762,
+      "grad_norm": 0.990350067615509,
+      "learning_rate": 1.0120595370083318e-05,
+      "loss": 0.4381,
+      "step": 8702
+    },
+    {
+      "epoch": 2.7943490126826136,
+      "grad_norm": 0.8355160355567932,
+      "learning_rate": 1.0105298806446028e-05,
+      "loss": 0.3453,
+      "step": 8703
+    },
+    {
+      "epoch": 2.794670091507465,
+      "grad_norm": 0.881846010684967,
+      "learning_rate": 1.009001319620545e-05,
+      "loss": 0.2857,
+      "step": 8704
+    },
+    {
+      "epoch": 2.7949911703323167,
+      "grad_norm": 0.8830981254577637,
+      "learning_rate": 1.0074738541223993e-05,
+      "loss": 0.327,
+      "step": 8705
+    },
+    {
+      "epoch": 2.7953122491571682,
+      "grad_norm": 0.9528623819351196,
+      "learning_rate": 1.0059474843362892e-05,
+      "loss": 0.3359,
+      "step": 8706
+    },
+    {
+      "epoch": 2.79563332798202,
+      "grad_norm": 0.9913058280944824,
+      "learning_rate": 1.0044222104481971e-05,
+      "loss": 0.3238,
+      "step": 8707
+    },
+    {
+      "epoch": 2.795954406806871,
+      "grad_norm": 1.4466516971588135,
+      "learning_rate": 1.0028980326439707e-05,
+      "loss": 0.5732,
+      "step": 8708
+    },
+    {
+      "epoch": 2.7962754856317225,
+      "grad_norm": 0.9511708617210388,
+      "learning_rate": 1.0013749511093307e-05,
+      "loss": 0.3061,
+      "step": 8709
+    },
+    {
+      "epoch": 2.796596564456574,
+      "grad_norm": 0.5760300755500793,
+      "learning_rate": 9.998529660298539e-06,
+      "loss": 0.2568,
+      "step": 8710
+    },
+    {
+      "epoch": 2.7969176432814256,
+      "grad_norm": 0.8406627774238586,
+      "learning_rate": 9.983320775909933e-06,
+      "loss": 0.3046,
+      "step": 8711
+    },
+    {
+      "epoch": 2.797238722106277,
+      "grad_norm": 0.9190598726272583,
+      "learning_rate": 9.968122859780648e-06,
+      "loss": 0.3231,
+      "step": 8712
+    },
+    {
+      "epoch": 2.7975598009311287,
+      "grad_norm": 0.6795236468315125,
+      "learning_rate": 9.952935913762506e-06,
+      "loss": 0.247,
+      "step": 8713
+    },
+    {
+      "epoch": 2.79788087975598,
+      "grad_norm": 0.6721789240837097,
+      "learning_rate": 9.937759939705971e-06,
+      "loss": 0.2513,
+      "step": 8714
+    },
+    {
+      "epoch": 2.7982019585808313,
+      "grad_norm": 1.041229009628296,
+      "learning_rate": 9.922594939460194e-06,
+      "loss": 0.3641,
+      "step": 8715
+    },
+    {
+      "epoch": 2.7985230374056833,
+      "grad_norm": 1.105242371559143,
+      "learning_rate": 9.907440914873e-06,
+      "loss": 0.3677,
+      "step": 8716
+    },
+    {
+      "epoch": 2.7988441162305344,
+      "grad_norm": 0.7640735507011414,
+      "learning_rate": 9.892297867790845e-06,
+      "loss": 0.2991,
+      "step": 8717
+    },
+    {
+      "epoch": 2.799165195055386,
+      "grad_norm": 0.6073140501976013,
+      "learning_rate": 9.877165800058874e-06,
+      "loss": 0.2511,
+      "step": 8718
+    },
+    {
+      "epoch": 2.7994862738802375,
+      "grad_norm": 0.9953632950782776,
+      "learning_rate": 9.86204471352088e-06,
+      "loss": 0.365,
+      "step": 8719
+    },
+    {
+      "epoch": 2.799807352705089,
+      "grad_norm": 0.6479620337486267,
+      "learning_rate": 9.84693461001932e-06,
+      "loss": 0.2704,
+      "step": 8720
+    },
+    {
+      "epoch": 2.8001284315299406,
+      "grad_norm": 0.9347096681594849,
+      "learning_rate": 9.831835491395292e-06,
+      "loss": 0.3145,
+      "step": 8721
+    },
+    {
+      "epoch": 2.800449510354792,
+      "grad_norm": 0.5598117113113403,
+      "learning_rate": 9.816747359488632e-06,
+      "loss": 0.2696,
+      "step": 8722
+    },
+    {
+      "epoch": 2.8007705891796437,
+      "grad_norm": 0.6550912857055664,
+      "learning_rate": 9.801670216137727e-06,
+      "loss": 0.2139,
+      "step": 8723
+    },
+    {
+      "epoch": 2.801091668004495,
+      "grad_norm": 0.6343926787376404,
+      "learning_rate": 9.786604063179728e-06,
+      "loss": 0.2569,
+      "step": 8724
+    },
+    {
+      "epoch": 2.801412746829347,
+      "grad_norm": 1.2895246744155884,
+      "learning_rate": 9.771548902450357e-06,
+      "loss": 0.3091,
+      "step": 8725
+    },
+    {
+      "epoch": 2.801733825654198,
+      "grad_norm": 0.6956881880760193,
+      "learning_rate": 9.756504735784067e-06,
+      "loss": 0.3224,
+      "step": 8726
+    },
+    {
+      "epoch": 2.8020549044790495,
+      "grad_norm": 0.8281194567680359,
+      "learning_rate": 9.74147156501396e-06,
+      "loss": 0.3129,
+      "step": 8727
+    },
+    {
+      "epoch": 2.802375983303901,
+      "grad_norm": 0.6391897201538086,
+      "learning_rate": 9.726449391971714e-06,
+      "loss": 0.3007,
+      "step": 8728
+    },
+    {
+      "epoch": 2.8026970621287526,
+      "grad_norm": 0.5030215978622437,
+      "learning_rate": 9.711438218487834e-06,
+      "loss": 0.2709,
+      "step": 8729
+    },
+    {
+      "epoch": 2.803018140953604,
+      "grad_norm": 0.5472492575645447,
+      "learning_rate": 9.696438046391288e-06,
+      "loss": 0.7123,
+      "step": 8730
+    },
+    {
+      "epoch": 2.8033392197784557,
+      "grad_norm": 0.44840529561042786,
+      "learning_rate": 9.681448877509858e-06,
+      "loss": 0.4614,
+      "step": 8731
+    },
+    {
+      "epoch": 2.8036602986033072,
+      "grad_norm": 0.39549553394317627,
+      "learning_rate": 9.666470713669918e-06,
+      "loss": 0.3304,
+      "step": 8732
+    },
+    {
+      "epoch": 2.8039813774281583,
+      "grad_norm": 0.3779802918434143,
+      "learning_rate": 9.651503556696516e-06,
+      "loss": 0.1788,
+      "step": 8733
+    },
+    {
+      "epoch": 2.8043024562530103,
+      "grad_norm": 0.1729205846786499,
+      "learning_rate": 9.636547408413355e-06,
+      "loss": 0.0618,
+      "step": 8734
+    },
+    {
+      "epoch": 2.8046235350778614,
+      "grad_norm": 0.44540268182754517,
+      "learning_rate": 9.621602270642781e-06,
+      "loss": 0.1766,
+      "step": 8735
+    },
+    {
+      "epoch": 2.804944613902713,
+      "grad_norm": 0.7657910585403442,
+      "learning_rate": 9.606668145205833e-06,
+      "loss": 0.4684,
+      "step": 8736
+    },
+    {
+      "epoch": 2.8052656927275645,
+      "grad_norm": 0.8212659358978271,
+      "learning_rate": 9.591745033922173e-06,
+      "loss": 0.384,
+      "step": 8737
+    },
+    {
+      "epoch": 2.805586771552416,
+      "grad_norm": 0.8744103908538818,
+      "learning_rate": 9.576832938610137e-06,
+      "loss": 0.338,
+      "step": 8738
+    },
+    {
+      "epoch": 2.8059078503772676,
+      "grad_norm": 0.7278825640678406,
+      "learning_rate": 9.561931861086737e-06,
+      "loss": 0.2896,
+      "step": 8739
+    },
+    {
+      "epoch": 2.806228929202119,
+      "grad_norm": 0.5598768591880798,
+      "learning_rate": 9.5470418031676e-06,
+      "loss": 0.2411,
+      "step": 8740
+    },
+    {
+      "epoch": 2.8065500080269707,
+      "grad_norm": 0.7483347058296204,
+      "learning_rate": 9.532162766667042e-06,
+      "loss": 0.2827,
+      "step": 8741
+    },
+    {
+      "epoch": 2.806871086851822,
+      "grad_norm": 0.7367462515830994,
+      "learning_rate": 9.517294753398064e-06,
+      "loss": 0.3007,
+      "step": 8742
+    },
+    {
+      "epoch": 2.807192165676674,
+      "grad_norm": 0.6921778917312622,
+      "learning_rate": 9.502437765172212e-06,
+      "loss": 0.2562,
+      "step": 8743
+    },
+    {
+      "epoch": 2.807513244501525,
+      "grad_norm": 0.8599756956100464,
+      "learning_rate": 9.487591803799856e-06,
+      "loss": 0.3628,
+      "step": 8744
+    },
+    {
+      "epoch": 2.8078343233263765,
+      "grad_norm": 0.8034688234329224,
+      "learning_rate": 9.47275687108986e-06,
+      "loss": 0.3219,
+      "step": 8745
+    },
+    {
+      "epoch": 2.808155402151228,
+      "grad_norm": 1.0342375040054321,
+      "learning_rate": 9.457932968849825e-06,
+      "loss": 0.3947,
+      "step": 8746
+    },
+    {
+      "epoch": 2.8084764809760796,
+      "grad_norm": 1.0900487899780273,
+      "learning_rate": 9.443120098886061e-06,
+      "loss": 0.357,
+      "step": 8747
+    },
+    {
+      "epoch": 2.808797559800931,
+      "grad_norm": 0.7890565991401672,
+      "learning_rate": 9.428318263003378e-06,
+      "loss": 0.2979,
+      "step": 8748
+    },
+    {
+      "epoch": 2.8091186386257827,
+      "grad_norm": 1.0363215208053589,
+      "learning_rate": 9.4135274630054e-06,
+      "loss": 0.4129,
+      "step": 8749
+    },
+    {
+      "epoch": 2.8094397174506343,
+      "grad_norm": 0.8779375553131104,
+      "learning_rate": 9.398747700694322e-06,
+      "loss": 0.2943,
+      "step": 8750
+    },
+    {
+      "epoch": 2.8097607962754854,
+      "grad_norm": 1.0878396034240723,
+      "learning_rate": 9.383978977871021e-06,
+      "loss": 0.3559,
+      "step": 8751
+    },
+    {
+      "epoch": 2.8100818751003374,
+      "grad_norm": 0.8391216397285461,
+      "learning_rate": 9.369221296335006e-06,
+      "loss": 0.3069,
+      "step": 8752
+    },
+    {
+      "epoch": 2.8104029539251885,
+      "grad_norm": 0.6143542528152466,
+      "learning_rate": 9.354474657884472e-06,
+      "loss": 0.225,
+      "step": 8753
+    },
+    {
+      "epoch": 2.81072403275004,
+      "grad_norm": 0.9501350522041321,
+      "learning_rate": 9.339739064316233e-06,
+      "loss": 0.3666,
+      "step": 8754
+    },
+    {
+      "epoch": 2.8110451115748916,
+      "grad_norm": 1.024901032447815,
+      "learning_rate": 9.32501451742579e-06,
+      "loss": 0.3373,
+      "step": 8755
+    },
+    {
+      "epoch": 2.811366190399743,
+      "grad_norm": 1.0849446058273315,
+      "learning_rate": 9.310301019007285e-06,
+      "loss": 0.3409,
+      "step": 8756
+    },
+    {
+      "epoch": 2.8116872692245947,
+      "grad_norm": 0.8314627408981323,
+      "learning_rate": 9.295598570853514e-06,
+      "loss": 0.3095,
+      "step": 8757
+    },
+    {
+      "epoch": 2.812008348049446,
+      "grad_norm": 0.5649183392524719,
+      "learning_rate": 9.280907174755915e-06,
+      "loss": 0.2587,
+      "step": 8758
+    },
+    {
+      "epoch": 2.8123294268742978,
+      "grad_norm": 1.2007514238357544,
+      "learning_rate": 9.266226832504598e-06,
+      "loss": 0.4184,
+      "step": 8759
+    },
+    {
+      "epoch": 2.812650505699149,
+      "grad_norm": 0.562286376953125,
+      "learning_rate": 9.251557545888312e-06,
+      "loss": 0.2122,
+      "step": 8760
+    },
+    {
+      "epoch": 2.812971584524001,
+      "grad_norm": 1.045385479927063,
+      "learning_rate": 9.236899316694459e-06,
+      "loss": 0.3434,
+      "step": 8761
+    },
+    {
+      "epoch": 2.813292663348852,
+      "grad_norm": 1.0073976516723633,
+      "learning_rate": 9.222252146709142e-06,
+      "loss": 0.334,
+      "step": 8762
+    },
+    {
+      "epoch": 2.8136137421737035,
+      "grad_norm": 0.5607120990753174,
+      "learning_rate": 9.207616037717025e-06,
+      "loss": 0.2412,
+      "step": 8763
+    },
+    {
+      "epoch": 2.813934820998555,
+      "grad_norm": 0.7444592714309692,
+      "learning_rate": 9.192990991501482e-06,
+      "loss": 0.2799,
+      "step": 8764
+    },
+    {
+      "epoch": 2.8142558998234066,
+      "grad_norm": 0.9598581194877625,
+      "learning_rate": 9.178377009844563e-06,
+      "loss": 0.3674,
+      "step": 8765
+    },
+    {
+      "epoch": 2.814576978648258,
+      "grad_norm": 0.42026737332344055,
+      "learning_rate": 9.163774094526889e-06,
+      "loss": 0.2069,
+      "step": 8766
+    },
+    {
+      "epoch": 2.8148980574731097,
+      "grad_norm": 0.8369678258895874,
+      "learning_rate": 9.149182247327837e-06,
+      "loss": 0.3303,
+      "step": 8767
+    },
+    {
+      "epoch": 2.8152191362979613,
+      "grad_norm": 0.8033504486083984,
+      "learning_rate": 9.134601470025306e-06,
+      "loss": 0.2976,
+      "step": 8768
+    },
+    {
+      "epoch": 2.8155402151228124,
+      "grad_norm": 0.7782095670700073,
+      "learning_rate": 9.120031764395987e-06,
+      "loss": 0.2766,
+      "step": 8769
+    },
+    {
+      "epoch": 2.8158612939476644,
+      "grad_norm": 0.6862573623657227,
+      "learning_rate": 9.105473132215125e-06,
+      "loss": 0.2771,
+      "step": 8770
+    },
+    {
+      "epoch": 2.8161823727725155,
+      "grad_norm": 1.1740379333496094,
+      "learning_rate": 9.09092557525666e-06,
+      "loss": 0.3947,
+      "step": 8771
+    },
+    {
+      "epoch": 2.816503451597367,
+      "grad_norm": 1.0018607378005981,
+      "learning_rate": 9.076389095293148e-06,
+      "loss": 0.4111,
+      "step": 8772
+    },
+    {
+      "epoch": 2.8168245304222186,
+      "grad_norm": 0.7517522573471069,
+      "learning_rate": 9.061863694095828e-06,
+      "loss": 0.2955,
+      "step": 8773
+    },
+    {
+      "epoch": 2.81714560924707,
+      "grad_norm": 1.8710321187973022,
+      "learning_rate": 9.047349373434566e-06,
+      "loss": 0.2693,
+      "step": 8774
+    },
+    {
+      "epoch": 2.8174666880719217,
+      "grad_norm": 0.8470088839530945,
+      "learning_rate": 9.0328461350779e-06,
+      "loss": 0.3039,
+      "step": 8775
+    },
+    {
+      "epoch": 2.8177877668967732,
+      "grad_norm": 0.5964589715003967,
+      "learning_rate": 9.018353980792993e-06,
+      "loss": 0.2547,
+      "step": 8776
+    },
+    {
+      "epoch": 2.818108845721625,
+      "grad_norm": 0.8444490432739258,
+      "learning_rate": 9.00387291234569e-06,
+      "loss": 0.3071,
+      "step": 8777
+    },
+    {
+      "epoch": 2.818429924546476,
+      "grad_norm": 0.41703271865844727,
+      "learning_rate": 8.989402931500434e-06,
+      "loss": 0.2554,
+      "step": 8778
+    },
+    {
+      "epoch": 2.818751003371328,
+      "grad_norm": 0.3994010090827942,
+      "learning_rate": 8.974944040020362e-06,
+      "loss": 0.2379,
+      "step": 8779
+    },
+    {
+      "epoch": 2.819072082196179,
+      "grad_norm": 0.443518728017807,
+      "learning_rate": 8.960496239667282e-06,
+      "loss": 0.5404,
+      "step": 8780
+    },
+    {
+      "epoch": 2.8193931610210305,
+      "grad_norm": 0.3758715093135834,
+      "learning_rate": 8.946059532201567e-06,
+      "loss": 0.3376,
+      "step": 8781
+    },
+    {
+      "epoch": 2.819714239845882,
+      "grad_norm": 0.3072349727153778,
+      "learning_rate": 8.931633919382298e-06,
+      "loss": 0.1223,
+      "step": 8782
+    },
+    {
+      "epoch": 2.8200353186707336,
+      "grad_norm": 0.3410075306892395,
+      "learning_rate": 8.917219402967202e-06,
+      "loss": 0.11,
+      "step": 8783
+    },
+    {
+      "epoch": 2.820356397495585,
+      "grad_norm": 0.1781407743692398,
+      "learning_rate": 8.90281598471262e-06,
+      "loss": 0.0609,
+      "step": 8784
+    },
+    {
+      "epoch": 2.8206774763204367,
+      "grad_norm": 0.3771248161792755,
+      "learning_rate": 8.888423666373614e-06,
+      "loss": 0.1744,
+      "step": 8785
+    },
+    {
+      "epoch": 2.8209985551452883,
+      "grad_norm": 0.8106180429458618,
+      "learning_rate": 8.87404244970378e-06,
+      "loss": 0.4395,
+      "step": 8786
+    },
+    {
+      "epoch": 2.8213196339701394,
+      "grad_norm": 0.7213160991668701,
+      "learning_rate": 8.85967233645547e-06,
+      "loss": 0.3335,
+      "step": 8787
+    },
+    {
+      "epoch": 2.8216407127949914,
+      "grad_norm": 0.8492469787597656,
+      "learning_rate": 8.845313328379634e-06,
+      "loss": 0.401,
+      "step": 8788
+    },
+    {
+      "epoch": 2.8219617916198425,
+      "grad_norm": 1.2384346723556519,
+      "learning_rate": 8.830965427225868e-06,
+      "loss": 0.3545,
+      "step": 8789
+    },
+    {
+      "epoch": 2.822282870444694,
+      "grad_norm": 0.8457659482955933,
+      "learning_rate": 8.816628634742441e-06,
+      "loss": 0.3346,
+      "step": 8790
+    },
+    {
+      "epoch": 2.8226039492695456,
+      "grad_norm": 0.9293232560157776,
+      "learning_rate": 8.80230295267619e-06,
+      "loss": 0.3992,
+      "step": 8791
+    },
+    {
+      "epoch": 2.822925028094397,
+      "grad_norm": 0.7842317819595337,
+      "learning_rate": 8.787988382772705e-06,
+      "loss": 0.355,
+      "step": 8792
+    },
+    {
+      "epoch": 2.8232461069192487,
+      "grad_norm": 0.9196096062660217,
+      "learning_rate": 8.77368492677616e-06,
+      "loss": 0.3592,
+      "step": 8793
+    },
+    {
+      "epoch": 2.8235671857441003,
+      "grad_norm": 0.7270289659500122,
+      "learning_rate": 8.759392586429393e-06,
+      "loss": 0.3004,
+      "step": 8794
+    },
+    {
+      "epoch": 2.823888264568952,
+      "grad_norm": 0.7461056709289551,
+      "learning_rate": 8.745111363473868e-06,
+      "loss": 0.2673,
+      "step": 8795
+    },
+    {
+      "epoch": 2.824209343393803,
+      "grad_norm": 0.9872536659240723,
+      "learning_rate": 8.730841259649725e-06,
+      "loss": 0.3072,
+      "step": 8796
+    },
+    {
+      "epoch": 2.824530422218655,
+      "grad_norm": 0.9411637783050537,
+      "learning_rate": 8.716582276695728e-06,
+      "loss": 0.3584,
+      "step": 8797
+    },
+    {
+      "epoch": 2.824851501043506,
+      "grad_norm": 1.1243722438812256,
+      "learning_rate": 8.702334416349278e-06,
+      "loss": 0.4163,
+      "step": 8798
+    },
+    {
+      "epoch": 2.8251725798683576,
+      "grad_norm": 0.9346780776977539,
+      "learning_rate": 8.688097680346453e-06,
+      "loss": 0.3681,
+      "step": 8799
+    },
+    {
+      "epoch": 2.825493658693209,
+      "grad_norm": 1.0615766048431396,
+      "learning_rate": 8.67387207042194e-06,
+      "loss": 0.4532,
+      "step": 8800
+    },
+    {
+      "epoch": 2.8258147375180607,
+      "grad_norm": 0.979707658290863,
+      "learning_rate": 8.6596575883091e-06,
+      "loss": 0.3883,
+      "step": 8801
+    },
+    {
+      "epoch": 2.826135816342912,
+      "grad_norm": 0.7535959482192993,
+      "learning_rate": 8.645454235739903e-06,
+      "loss": 0.2761,
+      "step": 8802
+    },
+    {
+      "epoch": 2.8264568951677638,
+      "grad_norm": 0.7734266519546509,
+      "learning_rate": 8.63126201444503e-06,
+      "loss": 0.2963,
+      "step": 8803
+    },
+    {
+      "epoch": 2.8267779739926153,
+      "grad_norm": 0.7798780202865601,
+      "learning_rate": 8.617080926153698e-06,
+      "loss": 0.3009,
+      "step": 8804
+    },
+    {
+      "epoch": 2.8270990528174664,
+      "grad_norm": 0.9741690158843994,
+      "learning_rate": 8.602910972593892e-06,
+      "loss": 0.3527,
+      "step": 8805
+    },
+    {
+      "epoch": 2.8274201316423184,
+      "grad_norm": 0.9816387891769409,
+      "learning_rate": 8.588752155492119e-06,
+      "loss": 0.3465,
+      "step": 8806
+    },
+    {
+      "epoch": 2.8277412104671695,
+      "grad_norm": 0.7801307439804077,
+      "learning_rate": 8.574604476573621e-06,
+      "loss": 0.292,
+      "step": 8807
+    },
+    {
+      "epoch": 2.828062289292021,
+      "grad_norm": 0.5235636234283447,
+      "learning_rate": 8.560467937562277e-06,
+      "loss": 0.2318,
+      "step": 8808
+    },
+    {
+      "epoch": 2.8283833681168726,
+      "grad_norm": 0.6542462110519409,
+      "learning_rate": 8.546342540180508e-06,
+      "loss": 0.2817,
+      "step": 8809
+    },
+    {
+      "epoch": 2.828704446941724,
+      "grad_norm": 0.6339287161827087,
+      "learning_rate": 8.532228286149501e-06,
+      "loss": 0.2691,
+      "step": 8810
+    },
+    {
+      "epoch": 2.8290255257665757,
+      "grad_norm": 0.9842689633369446,
+      "learning_rate": 8.51812517718904e-06,
+      "loss": 0.3822,
+      "step": 8811
+    },
+    {
+      "epoch": 2.8293466045914273,
+      "grad_norm": 0.6883869767189026,
+      "learning_rate": 8.504033215017527e-06,
+      "loss": 0.3071,
+      "step": 8812
+    },
+    {
+      "epoch": 2.829667683416279,
+      "grad_norm": 1.0689724683761597,
+      "learning_rate": 8.489952401352019e-06,
+      "loss": 0.4108,
+      "step": 8813
+    },
+    {
+      "epoch": 2.82998876224113,
+      "grad_norm": 0.9899790287017822,
+      "learning_rate": 8.475882737908248e-06,
+      "loss": 0.3652,
+      "step": 8814
+    },
+    {
+      "epoch": 2.830309841065982,
+      "grad_norm": 0.8097178936004639,
+      "learning_rate": 8.46182422640054e-06,
+      "loss": 0.3144,
+      "step": 8815
+    },
+    {
+      "epoch": 2.830630919890833,
+      "grad_norm": 0.8424767255783081,
+      "learning_rate": 8.447776868541879e-06,
+      "loss": 0.2967,
+      "step": 8816
+    },
+    {
+      "epoch": 2.8309519987156846,
+      "grad_norm": 0.7293074727058411,
+      "learning_rate": 8.433740666043898e-06,
+      "loss": 0.2959,
+      "step": 8817
+    },
+    {
+      "epoch": 2.831273077540536,
+      "grad_norm": 0.6922279000282288,
+      "learning_rate": 8.419715620616874e-06,
+      "loss": 0.2773,
+      "step": 8818
+    },
+    {
+      "epoch": 2.8315941563653877,
+      "grad_norm": 0.5553762316703796,
+      "learning_rate": 8.405701733969706e-06,
+      "loss": 0.2382,
+      "step": 8819
+    },
+    {
+      "epoch": 2.8319152351902392,
+      "grad_norm": 0.7001710534095764,
+      "learning_rate": 8.39169900780995e-06,
+      "loss": 0.2947,
+      "step": 8820
+    },
+    {
+      "epoch": 2.832236314015091,
+      "grad_norm": 0.524242639541626,
+      "learning_rate": 8.377707443843786e-06,
+      "loss": 0.2546,
+      "step": 8821
+    },
+    {
+      "epoch": 2.8325573928399423,
+      "grad_norm": 0.7304323315620422,
+      "learning_rate": 8.363727043776038e-06,
+      "loss": 0.311,
+      "step": 8822
+    },
+    {
+      "epoch": 2.8328784716647935,
+      "grad_norm": 0.5915831923484802,
+      "learning_rate": 8.34975780931021e-06,
+      "loss": 0.2488,
+      "step": 8823
+    },
+    {
+      "epoch": 2.8331995504896454,
+      "grad_norm": 0.8694709539413452,
+      "learning_rate": 8.335799742148387e-06,
+      "loss": 0.368,
+      "step": 8824
+    },
+    {
+      "epoch": 2.8335206293144966,
+      "grad_norm": 0.6131924986839294,
+      "learning_rate": 8.321852843991295e-06,
+      "loss": 0.2581,
+      "step": 8825
+    },
+    {
+      "epoch": 2.833841708139348,
+      "grad_norm": 0.42285412549972534,
+      "learning_rate": 8.307917116538378e-06,
+      "loss": 0.2609,
+      "step": 8826
+    },
+    {
+      "epoch": 2.8341627869641997,
+      "grad_norm": 0.6987262964248657,
+      "learning_rate": 8.293992561487596e-06,
+      "loss": 0.309,
+      "step": 8827
+    },
+    {
+      "epoch": 2.834483865789051,
+      "grad_norm": 0.8132457137107849,
+      "learning_rate": 8.280079180535672e-06,
+      "loss": 0.2866,
+      "step": 8828
+    },
+    {
+      "epoch": 2.8348049446139028,
+      "grad_norm": 0.3915092647075653,
+      "learning_rate": 8.26617697537786e-06,
+      "loss": 0.2629,
+      "step": 8829
+    },
+    {
+      "epoch": 2.8351260234387543,
+      "grad_norm": 0.4772132337093353,
+      "learning_rate": 8.252285947708139e-06,
+      "loss": 0.5484,
+      "step": 8830
+    },
+    {
+      "epoch": 2.835447102263606,
+      "grad_norm": 0.39662283658981323,
+      "learning_rate": 8.238406099219077e-06,
+      "loss": 0.3962,
+      "step": 8831
+    },
+    {
+      "epoch": 2.835768181088457,
+      "grad_norm": 0.4367140233516693,
+      "learning_rate": 8.224537431601886e-06,
+      "loss": 0.319,
+      "step": 8832
+    },
+    {
+      "epoch": 2.836089259913309,
+      "grad_norm": 0.6027937531471252,
+      "learning_rate": 8.21067994654644e-06,
+      "loss": 0.2027,
+      "step": 8833
+    },
+    {
+      "epoch": 2.83641033873816,
+      "grad_norm": 0.3798494040966034,
+      "learning_rate": 8.196833645741186e-06,
+      "loss": 0.155,
+      "step": 8834
+    },
+    {
+      "epoch": 2.8367314175630116,
+      "grad_norm": 0.2580123841762543,
+      "learning_rate": 8.182998530873298e-06,
+      "loss": 0.1179,
+      "step": 8835
+    },
+    {
+      "epoch": 2.837052496387863,
+      "grad_norm": 0.6310969591140747,
+      "learning_rate": 8.169174603628538e-06,
+      "loss": 0.3001,
+      "step": 8836
+    },
+    {
+      "epoch": 2.8373735752127147,
+      "grad_norm": 0.7293077707290649,
+      "learning_rate": 8.15536186569129e-06,
+      "loss": 0.3546,
+      "step": 8837
+    },
+    {
+      "epoch": 2.8376946540375663,
+      "grad_norm": 0.7130358219146729,
+      "learning_rate": 8.1415603187446e-06,
+      "loss": 0.2964,
+      "step": 8838
+    },
+    {
+      "epoch": 2.838015732862418,
+      "grad_norm": 0.7875835299491882,
+      "learning_rate": 8.127769964470156e-06,
+      "loss": 0.339,
+      "step": 8839
+    },
+    {
+      "epoch": 2.8383368116872694,
+      "grad_norm": 0.784997820854187,
+      "learning_rate": 8.113990804548244e-06,
+      "loss": 0.3349,
+      "step": 8840
+    },
+    {
+      "epoch": 2.8386578905121205,
+      "grad_norm": 0.6344702839851379,
+      "learning_rate": 8.100222840657878e-06,
+      "loss": 0.2666,
+      "step": 8841
+    },
+    {
+      "epoch": 2.8389789693369725,
+      "grad_norm": 0.984622061252594,
+      "learning_rate": 8.086466074476563e-06,
+      "loss": 0.432,
+      "step": 8842
+    },
+    {
+      "epoch": 2.8393000481618236,
+      "grad_norm": 0.8080422282218933,
+      "learning_rate": 8.072720507680565e-06,
+      "loss": 0.3034,
+      "step": 8843
+    },
+    {
+      "epoch": 2.839621126986675,
+      "grad_norm": 0.9881112575531006,
+      "learning_rate": 8.058986141944724e-06,
+      "loss": 0.3383,
+      "step": 8844
+    },
+    {
+      "epoch": 2.8399422058115267,
+      "grad_norm": 0.5864750742912292,
+      "learning_rate": 8.045262978942513e-06,
+      "loss": 0.2296,
+      "step": 8845
+    },
+    {
+      "epoch": 2.8402632846363782,
+      "grad_norm": 0.9349332451820374,
+      "learning_rate": 8.031551020346128e-06,
+      "loss": 0.3404,
+      "step": 8846
+    },
+    {
+      "epoch": 2.8405843634612298,
+      "grad_norm": 1.3247158527374268,
+      "learning_rate": 8.017850267826232e-06,
+      "loss": 0.4181,
+      "step": 8847
+    },
+    {
+      "epoch": 2.8409054422860813,
+      "grad_norm": 0.8365796804428101,
+      "learning_rate": 8.004160723052312e-06,
+      "loss": 0.2536,
+      "step": 8848
+    },
+    {
+      "epoch": 2.841226521110933,
+      "grad_norm": 0.7529785633087158,
+      "learning_rate": 7.990482387692311e-06,
+      "loss": 0.3008,
+      "step": 8849
+    },
+    {
+      "epoch": 2.841547599935784,
+      "grad_norm": 1.1152650117874146,
+      "learning_rate": 7.976815263412963e-06,
+      "loss": 0.4836,
+      "step": 8850
+    },
+    {
+      "epoch": 2.841868678760636,
+      "grad_norm": 1.6716127395629883,
+      "learning_rate": 7.963159351879556e-06,
+      "loss": 0.4522,
+      "step": 8851
+    },
+    {
+      "epoch": 2.842189757585487,
+      "grad_norm": 0.8794559836387634,
+      "learning_rate": 7.949514654755962e-06,
+      "loss": 0.3452,
+      "step": 8852
+    },
+    {
+      "epoch": 2.8425108364103386,
+      "grad_norm": 1.2077842950820923,
+      "learning_rate": 7.935881173704819e-06,
+      "loss": 0.3147,
+      "step": 8853
+    },
+    {
+      "epoch": 2.84283191523519,
+      "grad_norm": 0.605907678604126,
+      "learning_rate": 7.922258910387282e-06,
+      "loss": 0.2739,
+      "step": 8854
+    },
+    {
+      "epoch": 2.8431529940600417,
+      "grad_norm": 0.7989472150802612,
+      "learning_rate": 7.908647866463203e-06,
+      "loss": 0.2794,
+      "step": 8855
+    },
+    {
+      "epoch": 2.8434740728848933,
+      "grad_norm": 1.0483945608139038,
+      "learning_rate": 7.895048043591036e-06,
+      "loss": 0.3598,
+      "step": 8856
+    },
+    {
+      "epoch": 2.843795151709745,
+      "grad_norm": 0.4776444137096405,
+      "learning_rate": 7.881459443427886e-06,
+      "loss": 0.2168,
+      "step": 8857
+    },
+    {
+      "epoch": 2.8441162305345964,
+      "grad_norm": 0.8608068227767944,
+      "learning_rate": 7.867882067629472e-06,
+      "loss": 0.3373,
+      "step": 8858
+    },
+    {
+      "epoch": 2.8444373093594475,
+      "grad_norm": 0.8778037428855896,
+      "learning_rate": 7.854315917850163e-06,
+      "loss": 0.2954,
+      "step": 8859
+    },
+    {
+      "epoch": 2.8447583881842995,
+      "grad_norm": 1.0180518627166748,
+      "learning_rate": 7.840760995742946e-06,
+      "loss": 0.3334,
+      "step": 8860
+    },
+    {
+      "epoch": 2.8450794670091506,
+      "grad_norm": 0.9998841285705566,
+      "learning_rate": 7.827217302959467e-06,
+      "loss": 0.3415,
+      "step": 8861
+    },
+    {
+      "epoch": 2.845400545834002,
+      "grad_norm": 0.7613440155982971,
+      "learning_rate": 7.81368484114996e-06,
+      "loss": 0.2876,
+      "step": 8862
+    },
+    {
+      "epoch": 2.8457216246588537,
+      "grad_norm": 0.8815329670906067,
+      "learning_rate": 7.800163611963318e-06,
+      "loss": 0.3486,
+      "step": 8863
+    },
+    {
+      "epoch": 2.8460427034837052,
+      "grad_norm": 0.6765179634094238,
+      "learning_rate": 7.786653617047079e-06,
+      "loss": 0.2835,
+      "step": 8864
+    },
+    {
+      "epoch": 2.846363782308557,
+      "grad_norm": 0.7167448997497559,
+      "learning_rate": 7.77315485804736e-06,
+      "loss": 0.2964,
+      "step": 8865
+    },
+    {
+      "epoch": 2.8466848611334084,
+      "grad_norm": 0.879822850227356,
+      "learning_rate": 7.75966733660901e-06,
+      "loss": 0.3248,
+      "step": 8866
+    },
+    {
+      "epoch": 2.84700593995826,
+      "grad_norm": 0.6780831813812256,
+      "learning_rate": 7.746191054375362e-06,
+      "loss": 0.2852,
+      "step": 8867
+    },
+    {
+      "epoch": 2.847327018783111,
+      "grad_norm": 0.6202150583267212,
+      "learning_rate": 7.73272601298851e-06,
+      "loss": 0.273,
+      "step": 8868
+    },
+    {
+      "epoch": 2.847648097607963,
+      "grad_norm": 0.5386216640472412,
+      "learning_rate": 7.719272214089145e-06,
+      "loss": 0.2531,
+      "step": 8869
+    },
+    {
+      "epoch": 2.847969176432814,
+      "grad_norm": 0.7454226016998291,
+      "learning_rate": 7.7058296593165e-06,
+      "loss": 0.2987,
+      "step": 8870
+    },
+    {
+      "epoch": 2.8482902552576657,
+      "grad_norm": 0.7010564208030701,
+      "learning_rate": 7.692398350308594e-06,
+      "loss": 0.2399,
+      "step": 8871
+    },
+    {
+      "epoch": 2.848611334082517,
+      "grad_norm": 1.0159744024276733,
+      "learning_rate": 7.67897828870191e-06,
+      "loss": 0.3846,
+      "step": 8872
+    },
+    {
+      "epoch": 2.8489324129073688,
+      "grad_norm": 0.6551013588905334,
+      "learning_rate": 7.665569476131706e-06,
+      "loss": 0.2866,
+      "step": 8873
+    },
+    {
+      "epoch": 2.8492534917322203,
+      "grad_norm": 0.5894836783409119,
+      "learning_rate": 7.652171914231776e-06,
+      "loss": 0.2376,
+      "step": 8874
+    },
+    {
+      "epoch": 2.849574570557072,
+      "grad_norm": 1.2114777565002441,
+      "learning_rate": 7.638785604634579e-06,
+      "loss": 0.3111,
+      "step": 8875
+    },
+    {
+      "epoch": 2.8498956493819234,
+      "grad_norm": 0.9398203492164612,
+      "learning_rate": 7.625410548971191e-06,
+      "loss": 0.2726,
+      "step": 8876
+    },
+    {
+      "epoch": 2.8502167282067745,
+      "grad_norm": 0.8695142865180969,
+      "learning_rate": 7.612046748871327e-06,
+      "loss": 0.3361,
+      "step": 8877
+    },
+    {
+      "epoch": 2.8505378070316265,
+      "grad_norm": 0.6532575488090515,
+      "learning_rate": 7.59869420596333e-06,
+      "loss": 0.3271,
+      "step": 8878
+    },
+    {
+      "epoch": 2.8508588858564776,
+      "grad_norm": 0.48674994707107544,
+      "learning_rate": 7.585352921874156e-06,
+      "loss": 0.2893,
+      "step": 8879
+    },
+    {
+      "epoch": 2.851179964681329,
+      "grad_norm": 0.5763227343559265,
+      "learning_rate": 7.572022898229403e-06,
+      "loss": 0.8793,
+      "step": 8880
+    },
+    {
+      "epoch": 2.8515010435061807,
+      "grad_norm": 0.5322999954223633,
+      "learning_rate": 7.558704136653305e-06,
+      "loss": 0.7263,
+      "step": 8881
+    },
+    {
+      "epoch": 2.8518221223310323,
+      "grad_norm": 0.44331344962120056,
+      "learning_rate": 7.545396638768698e-06,
+      "loss": 0.279,
+      "step": 8882
+    },
+    {
+      "epoch": 2.852143201155884,
+      "grad_norm": 0.41767483949661255,
+      "learning_rate": 7.5321004061970405e-06,
+      "loss": 0.2259,
+      "step": 8883
+    },
+    {
+      "epoch": 2.8524642799807354,
+      "grad_norm": 0.3086419403553009,
+      "learning_rate": 7.518815440558513e-06,
+      "loss": 0.1316,
+      "step": 8884
+    },
+    {
+      "epoch": 2.852785358805587,
+      "grad_norm": 0.29793986678123474,
+      "learning_rate": 7.505541743471756e-06,
+      "loss": 0.1329,
+      "step": 8885
+    },
+    {
+      "epoch": 2.853106437630438,
+      "grad_norm": 0.4482042193412781,
+      "learning_rate": 7.492279316554207e-06,
+      "loss": 0.2962,
+      "step": 8886
+    },
+    {
+      "epoch": 2.85342751645529,
+      "grad_norm": 0.780487060546875,
+      "learning_rate": 7.479028161421797e-06,
+      "loss": 0.4438,
+      "step": 8887
+    },
+    {
+      "epoch": 2.853748595280141,
+      "grad_norm": 1.0584594011306763,
+      "learning_rate": 7.465788279689156e-06,
+      "loss": 0.3568,
+      "step": 8888
+    },
+    {
+      "epoch": 2.8540696741049927,
+      "grad_norm": 0.6687178611755371,
+      "learning_rate": 7.45255967296955e-06,
+      "loss": 0.2615,
+      "step": 8889
+    },
+    {
+      "epoch": 2.8543907529298442,
+      "grad_norm": 0.9293543696403503,
+      "learning_rate": 7.439342342874789e-06,
+      "loss": 0.3512,
+      "step": 8890
+    },
+    {
+      "epoch": 2.854711831754696,
+      "grad_norm": 0.8716065287590027,
+      "learning_rate": 7.426136291015417e-06,
+      "loss": 0.3854,
+      "step": 8891
+    },
+    {
+      "epoch": 2.8550329105795473,
+      "grad_norm": 0.711845338344574,
+      "learning_rate": 7.412941519000527e-06,
+      "loss": 0.3156,
+      "step": 8892
+    },
+    {
+      "epoch": 2.855353989404399,
+      "grad_norm": 0.7607097625732422,
+      "learning_rate": 7.399758028437864e-06,
+      "loss": 0.3388,
+      "step": 8893
+    },
+    {
+      "epoch": 2.8556750682292504,
+      "grad_norm": 0.6983038783073425,
+      "learning_rate": 7.386585820933811e-06,
+      "loss": 0.2692,
+      "step": 8894
+    },
+    {
+      "epoch": 2.8559961470541015,
+      "grad_norm": 0.703209638595581,
+      "learning_rate": 7.3734248980933395e-06,
+      "loss": 0.2435,
+      "step": 8895
+    },
+    {
+      "epoch": 2.8563172258789535,
+      "grad_norm": 1.0315760374069214,
+      "learning_rate": 7.360275261520078e-06,
+      "loss": 0.3833,
+      "step": 8896
+    },
+    {
+      "epoch": 2.8566383047038046,
+      "grad_norm": 0.6447276473045349,
+      "learning_rate": 7.347136912816277e-06,
+      "loss": 0.2404,
+      "step": 8897
+    },
+    {
+      "epoch": 2.856959383528656,
+      "grad_norm": 1.043185830116272,
+      "learning_rate": 7.3340098535827905e-06,
+      "loss": 0.4912,
+      "step": 8898
+    },
+    {
+      "epoch": 2.8572804623535077,
+      "grad_norm": 1.1728825569152832,
+      "learning_rate": 7.320894085419116e-06,
+      "loss": 0.4808,
+      "step": 8899
+    },
+    {
+      "epoch": 2.8576015411783593,
+      "grad_norm": 0.8470208644866943,
+      "learning_rate": 7.3077896099233765e-06,
+      "loss": 0.38,
+      "step": 8900
+    },
+    {
+      "epoch": 2.857922620003211,
+      "grad_norm": 0.6861622333526611,
+      "learning_rate": 7.2946964286923046e-06,
+      "loss": 0.2538,
+      "step": 8901
+    },
+    {
+      "epoch": 2.8582436988280624,
+      "grad_norm": 0.7869503498077393,
+      "learning_rate": 7.281614543321269e-06,
+      "loss": 0.2132,
+      "step": 8902
+    },
+    {
+      "epoch": 2.858564777652914,
+      "grad_norm": 0.7390521764755249,
+      "learning_rate": 7.268543955404239e-06,
+      "loss": 0.2631,
+      "step": 8903
+    },
+    {
+      "epoch": 2.858885856477765,
+      "grad_norm": 0.6110241413116455,
+      "learning_rate": 7.255484666533874e-06,
+      "loss": 0.2363,
+      "step": 8904
+    },
+    {
+      "epoch": 2.859206935302617,
+      "grad_norm": 0.7435978651046753,
+      "learning_rate": 7.242436678301367e-06,
+      "loss": 0.3225,
+      "step": 8905
+    },
+    {
+      "epoch": 2.859528014127468,
+      "grad_norm": 0.5957700610160828,
+      "learning_rate": 7.2293999922965705e-06,
+      "loss": 0.2636,
+      "step": 8906
+    },
+    {
+      "epoch": 2.8598490929523197,
+      "grad_norm": 0.6576164364814758,
+      "learning_rate": 7.216374610108012e-06,
+      "loss": 0.3076,
+      "step": 8907
+    },
+    {
+      "epoch": 2.8601701717771713,
+      "grad_norm": 0.7818155288696289,
+      "learning_rate": 7.203360533322734e-06,
+      "loss": 0.3078,
+      "step": 8908
+    },
+    {
+      "epoch": 2.860491250602023,
+      "grad_norm": 0.974690854549408,
+      "learning_rate": 7.190357763526523e-06,
+      "loss": 0.4259,
+      "step": 8909
+    },
+    {
+      "epoch": 2.8608123294268744,
+      "grad_norm": 0.795142412185669,
+      "learning_rate": 7.177366302303667e-06,
+      "loss": 0.3382,
+      "step": 8910
+    },
+    {
+      "epoch": 2.861133408251726,
+      "grad_norm": 1.03029465675354,
+      "learning_rate": 7.164386151237179e-06,
+      "loss": 0.3553,
+      "step": 8911
+    },
+    {
+      "epoch": 2.8614544870765775,
+      "grad_norm": 0.9264429211616516,
+      "learning_rate": 7.151417311908648e-06,
+      "loss": 0.3729,
+      "step": 8912
+    },
+    {
+      "epoch": 2.8617755659014286,
+      "grad_norm": 0.9430894255638123,
+      "learning_rate": 7.138459785898266e-06,
+      "loss": 0.4435,
+      "step": 8913
+    },
+    {
+      "epoch": 2.8620966447262806,
+      "grad_norm": 0.7884198427200317,
+      "learning_rate": 7.125513574784903e-06,
+      "loss": 0.3564,
+      "step": 8914
+    },
+    {
+      "epoch": 2.8624177235511317,
+      "grad_norm": 1.3142826557159424,
+      "learning_rate": 7.112578680145954e-06,
+      "loss": 0.3726,
+      "step": 8915
+    },
+    {
+      "epoch": 2.862738802375983,
+      "grad_norm": 0.9246410727500916,
+      "learning_rate": 7.099655103557556e-06,
+      "loss": 0.3109,
+      "step": 8916
+    },
+    {
+      "epoch": 2.8630598812008348,
+      "grad_norm": 0.8271921277046204,
+      "learning_rate": 7.086742846594385e-06,
+      "loss": 0.2753,
+      "step": 8917
+    },
+    {
+      "epoch": 2.8633809600256863,
+      "grad_norm": 1.1295146942138672,
+      "learning_rate": 7.07384191082977e-06,
+      "loss": 0.3933,
+      "step": 8918
+    },
+    {
+      "epoch": 2.863702038850538,
+      "grad_norm": 2.6274220943450928,
+      "learning_rate": 7.060952297835633e-06,
+      "loss": 0.4857,
+      "step": 8919
+    },
+    {
+      "epoch": 2.8640231176753894,
+      "grad_norm": 0.9029250741004944,
+      "learning_rate": 7.048074009182548e-06,
+      "loss": 0.3297,
+      "step": 8920
+    },
+    {
+      "epoch": 2.864344196500241,
+      "grad_norm": 0.5170316696166992,
+      "learning_rate": 7.035207046439673e-06,
+      "loss": 0.2399,
+      "step": 8921
+    },
+    {
+      "epoch": 2.864665275325092,
+      "grad_norm": 0.6248924732208252,
+      "learning_rate": 7.022351411174866e-06,
+      "loss": 0.2731,
+      "step": 8922
+    },
+    {
+      "epoch": 2.864986354149944,
+      "grad_norm": 0.6657963991165161,
+      "learning_rate": 7.009507104954493e-06,
+      "loss": 0.2576,
+      "step": 8923
+    },
+    {
+      "epoch": 2.865307432974795,
+      "grad_norm": 0.5382866263389587,
+      "learning_rate": 6.996674129343606e-06,
+      "loss": 0.2667,
+      "step": 8924
+    },
+    {
+      "epoch": 2.8656285117996467,
+      "grad_norm": 0.9319119453430176,
+      "learning_rate": 6.9838524859058616e-06,
+      "loss": 0.3681,
+      "step": 8925
+    },
+    {
+      "epoch": 2.8659495906244983,
+      "grad_norm": 0.8236702084541321,
+      "learning_rate": 6.971042176203535e-06,
+      "loss": 0.3227,
+      "step": 8926
+    },
+    {
+      "epoch": 2.86627066944935,
+      "grad_norm": 0.42309141159057617,
+      "learning_rate": 6.958243201797554e-06,
+      "loss": 0.2548,
+      "step": 8927
+    },
+    {
+      "epoch": 2.8665917482742014,
+      "grad_norm": 0.6998513340950012,
+      "learning_rate": 6.945455564247394e-06,
+      "loss": 0.3271,
+      "step": 8928
+    },
+    {
+      "epoch": 2.866912827099053,
+      "grad_norm": 0.48560941219329834,
+      "learning_rate": 6.932679265111231e-06,
+      "loss": 0.2742,
+      "step": 8929
+    },
+    {
+      "epoch": 2.8672339059239045,
+      "grad_norm": 0.45436060428619385,
+      "learning_rate": 6.919914305945774e-06,
+      "loss": 0.6581,
+      "step": 8930
+    },
+    {
+      "epoch": 2.8675549847487556,
+      "grad_norm": 0.4203808903694153,
+      "learning_rate": 6.907160688306425e-06,
+      "loss": 0.3959,
+      "step": 8931
+    },
+    {
+      "epoch": 2.8678760635736076,
+      "grad_norm": 0.45685985684394836,
+      "learning_rate": 6.894418413747183e-06,
+      "loss": 0.3185,
+      "step": 8932
+    },
+    {
+      "epoch": 2.8681971423984587,
+      "grad_norm": 0.4417474865913391,
+      "learning_rate": 6.881687483820609e-06,
+      "loss": 0.2633,
+      "step": 8933
+    },
+    {
+      "epoch": 2.8685182212233102,
+      "grad_norm": 0.34829646348953247,
+      "learning_rate": 6.868967900077972e-06,
+      "loss": 0.1154,
+      "step": 8934
+    },
+    {
+      "epoch": 2.868839300048162,
+      "grad_norm": 0.4470880925655365,
+      "learning_rate": 6.856259664069098e-06,
+      "loss": 0.1363,
+      "step": 8935
+    },
+    {
+      "epoch": 2.8691603788730133,
+      "grad_norm": 0.42514151334762573,
+      "learning_rate": 6.8435627773424495e-06,
+      "loss": 0.1393,
+      "step": 8936
+    },
+    {
+      "epoch": 2.869481457697865,
+      "grad_norm": 0.8983972668647766,
+      "learning_rate": 6.830877241445111e-06,
+      "loss": 0.5135,
+      "step": 8937
+    },
+    {
+      "epoch": 2.8698025365227164,
+      "grad_norm": 1.0805124044418335,
+      "learning_rate": 6.818203057922757e-06,
+      "loss": 0.4579,
+      "step": 8938
+    },
+    {
+      "epoch": 2.870123615347568,
+      "grad_norm": 0.8403233289718628,
+      "learning_rate": 6.805540228319718e-06,
+      "loss": 0.3314,
+      "step": 8939
+    },
+    {
+      "epoch": 2.870444694172419,
+      "grad_norm": 0.9584758281707764,
+      "learning_rate": 6.7928887541789055e-06,
+      "loss": 0.3441,
+      "step": 8940
+    },
+    {
+      "epoch": 2.870765772997271,
+      "grad_norm": 0.9408062100410461,
+      "learning_rate": 6.780248637041875e-06,
+      "loss": 0.3953,
+      "step": 8941
+    },
+    {
+      "epoch": 2.871086851822122,
+      "grad_norm": 0.7710734605789185,
+      "learning_rate": 6.767619878448783e-06,
+      "loss": 0.3212,
+      "step": 8942
+    },
+    {
+      "epoch": 2.8714079306469737,
+      "grad_norm": 0.6842841506004333,
+      "learning_rate": 6.755002479938411e-06,
+      "loss": 0.2995,
+      "step": 8943
+    },
+    {
+      "epoch": 2.8717290094718253,
+      "grad_norm": 0.8255465030670166,
+      "learning_rate": 6.742396443048138e-06,
+      "loss": 0.3496,
+      "step": 8944
+    },
+    {
+      "epoch": 2.872050088296677,
+      "grad_norm": 0.838579535484314,
+      "learning_rate": 6.729801769313981e-06,
+      "loss": 0.3619,
+      "step": 8945
+    },
+    {
+      "epoch": 2.8723711671215284,
+      "grad_norm": 0.7680864930152893,
+      "learning_rate": 6.717218460270536e-06,
+      "loss": 0.3327,
+      "step": 8946
+    },
+    {
+      "epoch": 2.87269224594638,
+      "grad_norm": 0.6783306002616882,
+      "learning_rate": 6.704646517451107e-06,
+      "loss": 0.3324,
+      "step": 8947
+    },
+    {
+      "epoch": 2.8730133247712315,
+      "grad_norm": 0.8792276978492737,
+      "learning_rate": 6.692085942387483e-06,
+      "loss": 0.3289,
+      "step": 8948
+    },
+    {
+      "epoch": 2.8733344035960826,
+      "grad_norm": 0.9213592410087585,
+      "learning_rate": 6.679536736610137e-06,
+      "loss": 0.3607,
+      "step": 8949
+    },
+    {
+      "epoch": 2.8736554824209346,
+      "grad_norm": 0.8916637897491455,
+      "learning_rate": 6.666998901648203e-06,
+      "loss": 0.3322,
+      "step": 8950
+    },
+    {
+      "epoch": 2.8739765612457857,
+      "grad_norm": 0.9735648036003113,
+      "learning_rate": 6.654472439029313e-06,
+      "loss": 0.3434,
+      "step": 8951
+    },
+    {
+      "epoch": 2.8742976400706373,
+      "grad_norm": 0.7319313287734985,
+      "learning_rate": 6.6419573502798374e-06,
+      "loss": 0.3157,
+      "step": 8952
+    },
+    {
+      "epoch": 2.874618718895489,
+      "grad_norm": 0.9477907419204712,
+      "learning_rate": 6.629453636924643e-06,
+      "loss": 0.3392,
+      "step": 8953
+    },
+    {
+      "epoch": 2.8749397977203404,
+      "grad_norm": 0.9957652688026428,
+      "learning_rate": 6.616961300487324e-06,
+      "loss": 0.448,
+      "step": 8954
+    },
+    {
+      "epoch": 2.875260876545192,
+      "grad_norm": 1.1732362508773804,
+      "learning_rate": 6.604480342490004e-06,
+      "loss": 0.438,
+      "step": 8955
+    },
+    {
+      "epoch": 2.8755819553700435,
+      "grad_norm": 0.710220992565155,
+      "learning_rate": 6.592010764453449e-06,
+      "loss": 0.2393,
+      "step": 8956
+    },
+    {
+      "epoch": 2.875903034194895,
+      "grad_norm": 0.8079162836074829,
+      "learning_rate": 6.579552567897051e-06,
+      "loss": 0.3013,
+      "step": 8957
+    },
+    {
+      "epoch": 2.876224113019746,
+      "grad_norm": 0.8686780333518982,
+      "learning_rate": 6.5671057543387985e-06,
+      "loss": 0.3435,
+      "step": 8958
+    },
+    {
+      "epoch": 2.876545191844598,
+      "grad_norm": 1.116487979888916,
+      "learning_rate": 6.554670325295298e-06,
+      "loss": 0.4351,
+      "step": 8959
+    },
+    {
+      "epoch": 2.876866270669449,
+      "grad_norm": 0.6508135199546814,
+      "learning_rate": 6.542246282281772e-06,
+      "loss": 0.311,
+      "step": 8960
+    },
+    {
+      "epoch": 2.8771873494943008,
+      "grad_norm": 1.2992252111434937,
+      "learning_rate": 6.529833626812043e-06,
+      "loss": 0.3634,
+      "step": 8961
+    },
+    {
+      "epoch": 2.8775084283191523,
+      "grad_norm": 0.7349622845649719,
+      "learning_rate": 6.517432360398556e-06,
+      "loss": 0.3065,
+      "step": 8962
+    },
+    {
+      "epoch": 2.877829507144004,
+      "grad_norm": 0.7028359770774841,
+      "learning_rate": 6.5050424845523815e-06,
+      "loss": 0.2529,
+      "step": 8963
+    },
+    {
+      "epoch": 2.8781505859688554,
+      "grad_norm": 0.6974001526832581,
+      "learning_rate": 6.492664000783166e-06,
+      "loss": 0.2822,
+      "step": 8964
+    },
+    {
+      "epoch": 2.878471664793707,
+      "grad_norm": 0.9233234524726868,
+      "learning_rate": 6.480296910599237e-06,
+      "loss": 0.3762,
+      "step": 8965
+    },
+    {
+      "epoch": 2.8787927436185585,
+      "grad_norm": 1.0546296834945679,
+      "learning_rate": 6.467941215507433e-06,
+      "loss": 0.4223,
+      "step": 8966
+    },
+    {
+      "epoch": 2.8791138224434096,
+      "grad_norm": 0.8296759724617004,
+      "learning_rate": 6.455596917013273e-06,
+      "loss": 0.3025,
+      "step": 8967
+    },
+    {
+      "epoch": 2.8794349012682616,
+      "grad_norm": 0.6233950257301331,
+      "learning_rate": 6.443264016620887e-06,
+      "loss": 0.2302,
+      "step": 8968
+    },
+    {
+      "epoch": 2.8797559800931127,
+      "grad_norm": 0.8595796227455139,
+      "learning_rate": 6.430942515832983e-06,
+      "loss": 0.267,
+      "step": 8969
+    },
+    {
+      "epoch": 2.8800770589179643,
+      "grad_norm": 0.9391990900039673,
+      "learning_rate": 6.418632416150927e-06,
+      "loss": 0.3733,
+      "step": 8970
+    },
+    {
+      "epoch": 2.880398137742816,
+      "grad_norm": 0.7740428447723389,
+      "learning_rate": 6.406333719074619e-06,
+      "loss": 0.2623,
+      "step": 8971
+    },
+    {
+      "epoch": 2.8807192165676674,
+      "grad_norm": 0.760464608669281,
+      "learning_rate": 6.394046426102674e-06,
+      "loss": 0.3015,
+      "step": 8972
+    },
+    {
+      "epoch": 2.881040295392519,
+      "grad_norm": 0.8069641590118408,
+      "learning_rate": 6.381770538732224e-06,
+      "loss": 0.324,
+      "step": 8973
+    },
+    {
+      "epoch": 2.8813613742173705,
+      "grad_norm": 0.6506406664848328,
+      "learning_rate": 6.3695060584590625e-06,
+      "loss": 0.2543,
+      "step": 8974
+    },
+    {
+      "epoch": 2.881682453042222,
+      "grad_norm": 0.4887990355491638,
+      "learning_rate": 6.357252986777595e-06,
+      "loss": 0.2442,
+      "step": 8975
+    },
+    {
+      "epoch": 2.882003531867073,
+      "grad_norm": 1.2128475904464722,
+      "learning_rate": 6.345011325180772e-06,
+      "loss": 0.3487,
+      "step": 8976
+    },
+    {
+      "epoch": 2.882324610691925,
+      "grad_norm": 0.8624823689460754,
+      "learning_rate": 6.332781075160243e-06,
+      "loss": 0.301,
+      "step": 8977
+    },
+    {
+      "epoch": 2.8826456895167762,
+      "grad_norm": 0.4960239827632904,
+      "learning_rate": 6.320562238206218e-06,
+      "loss": 0.2894,
+      "step": 8978
+    },
+    {
+      "epoch": 2.882966768341628,
+      "grad_norm": 0.48501670360565186,
+      "learning_rate": 6.308354815807527e-06,
+      "loss": 0.2678,
+      "step": 8979
+    },
+    {
+      "epoch": 2.8832878471664793,
+      "grad_norm": 0.42197084426879883,
+      "learning_rate": 6.296158809451602e-06,
+      "loss": 0.4037,
+      "step": 8980
+    },
+    {
+      "epoch": 2.883608925991331,
+      "grad_norm": 0.36780065298080444,
+      "learning_rate": 6.283974220624489e-06,
+      "loss": 0.3432,
+      "step": 8981
+    },
+    {
+      "epoch": 2.8839300048161824,
+      "grad_norm": 0.3911552131175995,
+      "learning_rate": 6.2718010508108545e-06,
+      "loss": 0.217,
+      "step": 8982
+    },
+    {
+      "epoch": 2.884251083641034,
+      "grad_norm": 0.5778048634529114,
+      "learning_rate": 6.259639301493947e-06,
+      "loss": 0.409,
+      "step": 8983
+    },
+    {
+      "epoch": 2.8845721624658855,
+      "grad_norm": 0.13645879924297333,
+      "learning_rate": 6.2474889741556575e-06,
+      "loss": 0.0592,
+      "step": 8984
+    },
+    {
+      "epoch": 2.8848932412907367,
+      "grad_norm": 0.3099921941757202,
+      "learning_rate": 6.235350070276447e-06,
+      "loss": 0.1358,
+      "step": 8985
+    },
+    {
+      "epoch": 2.8852143201155886,
+      "grad_norm": 0.9487568736076355,
+      "learning_rate": 6.223222591335409e-06,
+      "loss": 0.3908,
+      "step": 8986
+    },
+    {
+      "epoch": 2.8855353989404398,
+      "grad_norm": 0.6402967572212219,
+      "learning_rate": 6.21110653881023e-06,
+      "loss": 0.2223,
+      "step": 8987
+    },
+    {
+      "epoch": 2.8858564777652913,
+      "grad_norm": 1.0103484392166138,
+      "learning_rate": 6.1990019141772605e-06,
+      "loss": 0.4384,
+      "step": 8988
+    },
+    {
+      "epoch": 2.886177556590143,
+      "grad_norm": 0.8888756632804871,
+      "learning_rate": 6.186908718911344e-06,
+      "loss": 0.3182,
+      "step": 8989
+    },
+    {
+      "epoch": 2.8864986354149944,
+      "grad_norm": 0.6990677118301392,
+      "learning_rate": 6.174826954486068e-06,
+      "loss": 0.2554,
+      "step": 8990
+    },
+    {
+      "epoch": 2.886819714239846,
+      "grad_norm": 0.9189555048942566,
+      "learning_rate": 6.1627566223735e-06,
+      "loss": 0.3981,
+      "step": 8991
+    },
+    {
+      "epoch": 2.8871407930646975,
+      "grad_norm": 0.9261887073516846,
+      "learning_rate": 6.1506977240444074e-06,
+      "loss": 0.3965,
+      "step": 8992
+    },
+    {
+      "epoch": 2.887461871889549,
+      "grad_norm": 0.7619607448577881,
+      "learning_rate": 6.138650260968137e-06,
+      "loss": 0.3196,
+      "step": 8993
+    },
+    {
+      "epoch": 2.8877829507144,
+      "grad_norm": 0.7999249696731567,
+      "learning_rate": 6.126614234612593e-06,
+      "loss": 0.2928,
+      "step": 8994
+    },
+    {
+      "epoch": 2.888104029539252,
+      "grad_norm": 0.8483562469482422,
+      "learning_rate": 6.1145896464443685e-06,
+      "loss": 0.3259,
+      "step": 8995
+    },
+    {
+      "epoch": 2.8884251083641033,
+      "grad_norm": 0.9350925087928772,
+      "learning_rate": 6.102576497928614e-06,
+      "loss": 0.3718,
+      "step": 8996
+    },
+    {
+      "epoch": 2.888746187188955,
+      "grad_norm": 1.1752997636795044,
+      "learning_rate": 6.090574790529091e-06,
+      "loss": 0.4583,
+      "step": 8997
+    },
+    {
+      "epoch": 2.8890672660138064,
+      "grad_norm": 1.18094801902771,
+      "learning_rate": 6.078584525708176e-06,
+      "loss": 0.4047,
+      "step": 8998
+    },
+    {
+      "epoch": 2.889388344838658,
+      "grad_norm": 1.1742857694625854,
+      "learning_rate": 6.066605704926831e-06,
+      "loss": 0.4776,
+      "step": 8999
+    },
+    {
+      "epoch": 2.8897094236635095,
+      "grad_norm": 0.905214786529541,
+      "learning_rate": 6.054638329644657e-06,
+      "loss": 0.3768,
+      "step": 9000
+    },
+    {
+      "epoch": 2.890030502488361,
+      "grad_norm": 1.0812019109725952,
+      "learning_rate": 6.042682401319844e-06,
+      "loss": 0.4689,
+      "step": 9001
+    },
+    {
+      "epoch": 2.8903515813132126,
+      "grad_norm": 1.0372034311294556,
+      "learning_rate": 6.030737921409169e-06,
+      "loss": 0.2963,
+      "step": 9002
+    },
+    {
+      "epoch": 2.8906726601380637,
+      "grad_norm": 1.742632508277893,
+      "learning_rate": 6.018804891368035e-06,
+      "loss": 0.3989,
+      "step": 9003
+    },
+    {
+      "epoch": 2.8909937389629157,
+      "grad_norm": 0.8828332424163818,
+      "learning_rate": 6.006883312650457e-06,
+      "loss": 0.3274,
+      "step": 9004
+    },
+    {
+      "epoch": 2.8913148177877668,
+      "grad_norm": 0.6829835772514343,
+      "learning_rate": 5.994973186709041e-06,
+      "loss": 0.2829,
+      "step": 9005
+    },
+    {
+      "epoch": 2.8916358966126183,
+      "grad_norm": 0.8279037475585938,
+      "learning_rate": 5.98307451499498e-06,
+      "loss": 0.3327,
+      "step": 9006
+    },
+    {
+      "epoch": 2.89195697543747,
+      "grad_norm": 0.7003656625747681,
+      "learning_rate": 5.971187298958103e-06,
+      "loss": 0.2629,
+      "step": 9007
+    },
+    {
+      "epoch": 2.8922780542623214,
+      "grad_norm": 0.7397180795669556,
+      "learning_rate": 5.9593115400468636e-06,
+      "loss": 0.2892,
+      "step": 9008
+    },
+    {
+      "epoch": 2.892599133087173,
+      "grad_norm": 0.88568514585495,
+      "learning_rate": 5.947447239708215e-06,
+      "loss": 0.3233,
+      "step": 9009
+    },
+    {
+      "epoch": 2.8929202119120245,
+      "grad_norm": 0.6601244807243347,
+      "learning_rate": 5.935594399387856e-06,
+      "loss": 0.27,
+      "step": 9010
+    },
+    {
+      "epoch": 2.893241290736876,
+      "grad_norm": 1.0693775415420532,
+      "learning_rate": 5.923753020529999e-06,
+      "loss": 0.4009,
+      "step": 9011
+    },
+    {
+      "epoch": 2.893562369561727,
+      "grad_norm": 0.7800514698028564,
+      "learning_rate": 5.911923104577455e-06,
+      "loss": 0.3128,
+      "step": 9012
+    },
+    {
+      "epoch": 2.893883448386579,
+      "grad_norm": 0.9529879689216614,
+      "learning_rate": 5.900104652971694e-06,
+      "loss": 0.3787,
+      "step": 9013
+    },
+    {
+      "epoch": 2.8942045272114303,
+      "grad_norm": 1.1537600755691528,
+      "learning_rate": 5.888297667152731e-06,
+      "loss": 0.531,
+      "step": 9014
+    },
+    {
+      "epoch": 2.894525606036282,
+      "grad_norm": 0.9319456219673157,
+      "learning_rate": 5.8765021485592376e-06,
+      "loss": 0.2995,
+      "step": 9015
+    },
+    {
+      "epoch": 2.8948466848611334,
+      "grad_norm": 0.5598946809768677,
+      "learning_rate": 5.864718098628441e-06,
+      "loss": 0.2649,
+      "step": 9016
+    },
+    {
+      "epoch": 2.895167763685985,
+      "grad_norm": 0.70030277967453,
+      "learning_rate": 5.852945518796205e-06,
+      "loss": 0.2977,
+      "step": 9017
+    },
+    {
+      "epoch": 2.8954888425108365,
+      "grad_norm": 0.9624463319778442,
+      "learning_rate": 5.8411844104969916e-06,
+      "loss": 0.3796,
+      "step": 9018
+    },
+    {
+      "epoch": 2.895809921335688,
+      "grad_norm": 0.5246722102165222,
+      "learning_rate": 5.829434775163833e-06,
+      "loss": 0.2232,
+      "step": 9019
+    },
+    {
+      "epoch": 2.8961310001605396,
+      "grad_norm": 0.5366243720054626,
+      "learning_rate": 5.8176966142283965e-06,
+      "loss": 0.2416,
+      "step": 9020
+    },
+    {
+      "epoch": 2.8964520789853907,
+      "grad_norm": 0.7249617576599121,
+      "learning_rate": 5.805969929120947e-06,
+      "loss": 0.2787,
+      "step": 9021
+    },
+    {
+      "epoch": 2.8967731578102422,
+      "grad_norm": 0.7474877834320068,
+      "learning_rate": 5.7942547212703315e-06,
+      "loss": 0.2923,
+      "step": 9022
+    },
+    {
+      "epoch": 2.897094236635094,
+      "grad_norm": 0.5960254669189453,
+      "learning_rate": 5.78255099210403e-06,
+      "loss": 0.2618,
+      "step": 9023
+    },
+    {
+      "epoch": 2.8974153154599454,
+      "grad_norm": 0.7626693844795227,
+      "learning_rate": 5.770858743048091e-06,
+      "loss": 0.2727,
+      "step": 9024
+    },
+    {
+      "epoch": 2.897736394284797,
+      "grad_norm": 0.5556107759475708,
+      "learning_rate": 5.759177975527186e-06,
+      "loss": 0.258,
+      "step": 9025
+    },
+    {
+      "epoch": 2.8980574731096485,
+      "grad_norm": 0.471110463142395,
+      "learning_rate": 5.747508690964598e-06,
+      "loss": 0.2506,
+      "step": 9026
+    },
+    {
+      "epoch": 2.8983785519345,
+      "grad_norm": 0.47644883394241333,
+      "learning_rate": 5.735850890782157e-06,
+      "loss": 0.2521,
+      "step": 9027
+    },
+    {
+      "epoch": 2.8986996307593516,
+      "grad_norm": 0.49500352144241333,
+      "learning_rate": 5.724204576400371e-06,
+      "loss": 0.275,
+      "step": 9028
+    },
+    {
+      "epoch": 2.899020709584203,
+      "grad_norm": 0.31801772117614746,
+      "learning_rate": 5.7125697492382835e-06,
+      "loss": 0.2572,
+      "step": 9029
+    },
+    {
+      "epoch": 2.899341788409054,
+      "grad_norm": 0.7811435461044312,
+      "learning_rate": 5.700946410713548e-06,
+      "loss": 0.8103,
+      "step": 9030
+    },
+    {
+      "epoch": 2.8996628672339058,
+      "grad_norm": 0.3786062002182007,
+      "learning_rate": 5.6893345622424874e-06,
+      "loss": 0.3159,
+      "step": 9031
+    },
+    {
+      "epoch": 2.8999839460587573,
+      "grad_norm": 0.2506957948207855,
+      "learning_rate": 5.6777342052399045e-06,
+      "loss": 0.0899,
+      "step": 9032
+    },
+    {
+      "epoch": 2.900305024883609,
+      "grad_norm": 0.23247039318084717,
+      "learning_rate": 5.666145341119322e-06,
+      "loss": 0.0615,
+      "step": 9033
+    },
+    {
+      "epoch": 2.9006261037084604,
+      "grad_norm": 0.1694241166114807,
+      "learning_rate": 5.654567971292757e-06,
+      "loss": 0.0589,
+      "step": 9034
+    },
+    {
+      "epoch": 2.900947182533312,
+      "grad_norm": 0.25635042786598206,
+      "learning_rate": 5.643002097170924e-06,
+      "loss": 0.1044,
+      "step": 9035
+    },
+    {
+      "epoch": 2.9012682613581635,
+      "grad_norm": 0.9146557450294495,
+      "learning_rate": 5.6314477201630745e-06,
+      "loss": 0.5639,
+      "step": 9036
+    },
+    {
+      "epoch": 2.901589340183015,
+      "grad_norm": 0.8883252739906311,
+      "learning_rate": 5.619904841677059e-06,
+      "loss": 0.4244,
+      "step": 9037
+    },
+    {
+      "epoch": 2.9019104190078666,
+      "grad_norm": 0.751973032951355,
+      "learning_rate": 5.608373463119354e-06,
+      "loss": 0.3138,
+      "step": 9038
+    },
+    {
+      "epoch": 2.9022314978327177,
+      "grad_norm": 0.7506957054138184,
+      "learning_rate": 5.5968535858950345e-06,
+      "loss": 0.2615,
+      "step": 9039
+    },
+    {
+      "epoch": 2.9025525766575693,
+      "grad_norm": 0.7786146402359009,
+      "learning_rate": 5.585345211407733e-06,
+      "loss": 0.3453,
+      "step": 9040
+    },
+    {
+      "epoch": 2.902873655482421,
+      "grad_norm": 0.9360180497169495,
+      "learning_rate": 5.573848341059739e-06,
+      "loss": 0.4307,
+      "step": 9041
+    },
+    {
+      "epoch": 2.9031947343072724,
+      "grad_norm": 0.9217041730880737,
+      "learning_rate": 5.562362976251901e-06,
+      "loss": 0.3228,
+      "step": 9042
+    },
+    {
+      "epoch": 2.903515813132124,
+      "grad_norm": 0.904606819152832,
+      "learning_rate": 5.550889118383673e-06,
+      "loss": 0.3926,
+      "step": 9043
+    },
+    {
+      "epoch": 2.9038368919569755,
+      "grad_norm": 0.7405612468719482,
+      "learning_rate": 5.5394267688531066e-06,
+      "loss": 0.2862,
+      "step": 9044
+    },
+    {
+      "epoch": 2.904157970781827,
+      "grad_norm": 0.7969362735748291,
+      "learning_rate": 5.52797592905685e-06,
+      "loss": 0.3321,
+      "step": 9045
+    },
+    {
+      "epoch": 2.9044790496066786,
+      "grad_norm": 0.8210546970367432,
+      "learning_rate": 5.516536600390188e-06,
+      "loss": 0.3051,
+      "step": 9046
+    },
+    {
+      "epoch": 2.90480012843153,
+      "grad_norm": 0.796033501625061,
+      "learning_rate": 5.505108784246926e-06,
+      "loss": 0.3624,
+      "step": 9047
+    },
+    {
+      "epoch": 2.9051212072563812,
+      "grad_norm": 0.7903563380241394,
+      "learning_rate": 5.49369248201953e-06,
+      "loss": 0.318,
+      "step": 9048
+    },
+    {
+      "epoch": 2.905442286081233,
+      "grad_norm": 1.0777714252471924,
+      "learning_rate": 5.482287695099031e-06,
+      "loss": 0.4097,
+      "step": 9049
+    },
+    {
+      "epoch": 2.9057633649060843,
+      "grad_norm": 1.2699135541915894,
+      "learning_rate": 5.470894424875062e-06,
+      "loss": 0.3572,
+      "step": 9050
+    },
+    {
+      "epoch": 2.906084443730936,
+      "grad_norm": 0.935076892375946,
+      "learning_rate": 5.4595126727359e-06,
+      "loss": 0.3222,
+      "step": 9051
+    },
+    {
+      "epoch": 2.9064055225557874,
+      "grad_norm": 0.7485237121582031,
+      "learning_rate": 5.448142440068316e-06,
+      "loss": 0.3005,
+      "step": 9052
+    },
+    {
+      "epoch": 2.906726601380639,
+      "grad_norm": 0.9007347822189331,
+      "learning_rate": 5.436783728257788e-06,
+      "loss": 0.3643,
+      "step": 9053
+    },
+    {
+      "epoch": 2.9070476802054905,
+      "grad_norm": 0.8481326699256897,
+      "learning_rate": 5.425436538688322e-06,
+      "loss": 0.3329,
+      "step": 9054
+    },
+    {
+      "epoch": 2.907368759030342,
+      "grad_norm": 0.8909382820129395,
+      "learning_rate": 5.414100872742534e-06,
+      "loss": 0.2989,
+      "step": 9055
+    },
+    {
+      "epoch": 2.9076898378551936,
+      "grad_norm": 0.9575411677360535,
+      "learning_rate": 5.402776731801662e-06,
+      "loss": 0.269,
+      "step": 9056
+    },
+    {
+      "epoch": 2.9080109166800447,
+      "grad_norm": 0.7255709767341614,
+      "learning_rate": 5.39146411724547e-06,
+      "loss": 0.294,
+      "step": 9057
+    },
+    {
+      "epoch": 2.9083319955048963,
+      "grad_norm": 1.5980173349380493,
+      "learning_rate": 5.380163030452412e-06,
+      "loss": 0.2823,
+      "step": 9058
+    },
+    {
+      "epoch": 2.908653074329748,
+      "grad_norm": 0.8485715389251709,
+      "learning_rate": 5.368873472799474e-06,
+      "loss": 0.2583,
+      "step": 9059
+    },
+    {
+      "epoch": 2.9089741531545994,
+      "grad_norm": 0.9129398465156555,
+      "learning_rate": 5.357595445662267e-06,
+      "loss": 0.3554,
+      "step": 9060
+    },
+    {
+      "epoch": 2.909295231979451,
+      "grad_norm": 0.7431888580322266,
+      "learning_rate": 5.346328950414969e-06,
+      "loss": 0.2997,
+      "step": 9061
+    },
+    {
+      "epoch": 2.9096163108043025,
+      "grad_norm": 1.5763301849365234,
+      "learning_rate": 5.335073988430372e-06,
+      "loss": 0.3451,
+      "step": 9062
+    },
+    {
+      "epoch": 2.909937389629154,
+      "grad_norm": 0.8064843416213989,
+      "learning_rate": 5.3238305610798565e-06,
+      "loss": 0.3082,
+      "step": 9063
+    },
+    {
+      "epoch": 2.9102584684540056,
+      "grad_norm": 1.073921799659729,
+      "learning_rate": 5.312598669733404e-06,
+      "loss": 0.3749,
+      "step": 9064
+    },
+    {
+      "epoch": 2.910579547278857,
+      "grad_norm": 0.5479134917259216,
+      "learning_rate": 5.301378315759598e-06,
+      "loss": 0.2452,
+      "step": 9065
+    },
+    {
+      "epoch": 2.9109006261037083,
+      "grad_norm": 0.9387060403823853,
+      "learning_rate": 5.290169500525577e-06,
+      "loss": 0.3453,
+      "step": 9066
+    },
+    {
+      "epoch": 2.91122170492856,
+      "grad_norm": 0.6247801184654236,
+      "learning_rate": 5.278972225397127e-06,
+      "loss": 0.2472,
+      "step": 9067
+    },
+    {
+      "epoch": 2.9115427837534114,
+      "grad_norm": 0.6648100018501282,
+      "learning_rate": 5.267786491738569e-06,
+      "loss": 0.2595,
+      "step": 9068
+    },
+    {
+      "epoch": 2.911863862578263,
+      "grad_norm": 0.5038353204727173,
+      "learning_rate": 5.256612300912911e-06,
+      "loss": 0.2275,
+      "step": 9069
+    },
+    {
+      "epoch": 2.9121849414031145,
+      "grad_norm": 0.5695350170135498,
+      "learning_rate": 5.245449654281631e-06,
+      "loss": 0.2695,
+      "step": 9070
+    },
+    {
+      "epoch": 2.912506020227966,
+      "grad_norm": NaN,
+      "learning_rate": 5.245449654281631e-06,
+      "loss": 0.4011,
+      "step": 9071
+    },
+    {
+      "epoch": 2.9128270990528176,
+      "grad_norm": 1.1442893743515015,
+      "learning_rate": 5.2342985532049084e-06,
+      "loss": 0.3682,
+      "step": 9072
+    },
+    {
+      "epoch": 2.913148177877669,
+      "grad_norm": 0.6914578676223755,
+      "learning_rate": 5.223158999041444e-06,
+      "loss": 0.2844,
+      "step": 9073
+    },
+    {
+      "epoch": 2.9134692567025207,
+      "grad_norm": 0.8676597476005554,
+      "learning_rate": 5.212030993148553e-06,
+      "loss": 0.3133,
+      "step": 9074
+    },
+    {
+      "epoch": 2.9137903355273718,
+      "grad_norm": 0.6819028854370117,
+      "learning_rate": 5.200914536882185e-06,
+      "loss": 0.2754,
+      "step": 9075
+    },
+    {
+      "epoch": 2.9141114143522233,
+      "grad_norm": 0.5778715014457703,
+      "learning_rate": 5.189809631596798e-06,
+      "loss": 0.2804,
+      "step": 9076
+    },
+    {
+      "epoch": 2.914432493177075,
+      "grad_norm": 0.4760773479938507,
+      "learning_rate": 5.178716278645535e-06,
+      "loss": 0.245,
+      "step": 9077
+    },
+    {
+      "epoch": 2.9147535720019264,
+      "grad_norm": 0.41519516706466675,
+      "learning_rate": 5.167634479380068e-06,
+      "loss": 0.2327,
+      "step": 9078
+    },
+    {
+      "epoch": 2.915074650826778,
+      "grad_norm": 0.5341471433639526,
+      "learning_rate": 5.1565642351506845e-06,
+      "loss": 0.2778,
+      "step": 9079
+    },
+    {
+      "epoch": 2.9153957296516295,
+      "grad_norm": 0.6417379379272461,
+      "learning_rate": 5.145505547306251e-06,
+      "loss": 0.9807,
+      "step": 9080
+    },
+    {
+      "epoch": 2.915716808476481,
+      "grad_norm": 0.5737770199775696,
+      "learning_rate": 5.134458417194254e-06,
+      "loss": 0.5596,
+      "step": 9081
+    },
+    {
+      "epoch": 2.9160378873013326,
+      "grad_norm": 0.29216137528419495,
+      "learning_rate": 5.1234228461607304e-06,
+      "loss": 0.1449,
+      "step": 9082
+    },
+    {
+      "epoch": 2.916358966126184,
+      "grad_norm": 0.20528793334960938,
+      "learning_rate": 5.1123988355503475e-06,
+      "loss": 0.06,
+      "step": 9083
+    },
+    {
+      "epoch": 2.9166800449510353,
+      "grad_norm": 0.1566951721906662,
+      "learning_rate": 5.101386386706342e-06,
+      "loss": 0.0578,
+      "step": 9084
+    },
+    {
+      "epoch": 2.917001123775887,
+      "grad_norm": 0.1581335961818695,
+      "learning_rate": 5.0903855009705514e-06,
+      "loss": 0.0592,
+      "step": 9085
+    },
+    {
+      "epoch": 2.9173222026007384,
+      "grad_norm": 0.24517378211021423,
+      "learning_rate": 5.079396179683383e-06,
+      "loss": 0.1089,
+      "step": 9086
+    },
+    {
+      "epoch": 2.91764328142559,
+      "grad_norm": 0.6500630974769592,
+      "learning_rate": 5.068418424183874e-06,
+      "loss": 0.3078,
+      "step": 9087
+    },
+    {
+      "epoch": 2.9179643602504415,
+      "grad_norm": 0.8994437456130981,
+      "learning_rate": 5.057452235809624e-06,
+      "loss": 0.509,
+      "step": 9088
+    },
+    {
+      "epoch": 2.918285439075293,
+      "grad_norm": 0.7024485468864441,
+      "learning_rate": 5.046497615896806e-06,
+      "loss": 0.2671,
+      "step": 9089
+    },
+    {
+      "epoch": 2.9186065179001446,
+      "grad_norm": 0.7663185000419617,
+      "learning_rate": 5.035554565780265e-06,
+      "loss": 0.3624,
+      "step": 9090
+    },
+    {
+      "epoch": 2.918927596724996,
+      "grad_norm": 0.6804545521736145,
+      "learning_rate": 5.024623086793323e-06,
+      "loss": 0.2751,
+      "step": 9091
+    },
+    {
+      "epoch": 2.9192486755498477,
+      "grad_norm": 0.78728187084198,
+      "learning_rate": 5.013703180267959e-06,
+      "loss": 0.3272,
+      "step": 9092
+    },
+    {
+      "epoch": 2.919569754374699,
+      "grad_norm": 0.8548991680145264,
+      "learning_rate": 5.002794847534764e-06,
+      "loss": 0.3579,
+      "step": 9093
+    },
+    {
+      "epoch": 2.9198908331995503,
+      "grad_norm": 1.0403963327407837,
+      "learning_rate": 4.991898089922819e-06,
+      "loss": 0.4054,
+      "step": 9094
+    },
+    {
+      "epoch": 2.920211912024402,
+      "grad_norm": 1.1474127769470215,
+      "learning_rate": 4.981012908759941e-06,
+      "loss": 0.4484,
+      "step": 9095
+    },
+    {
+      "epoch": 2.9205329908492534,
+      "grad_norm": 0.6695740818977356,
+      "learning_rate": 4.97013930537239e-06,
+      "loss": 0.2586,
+      "step": 9096
+    },
+    {
+      "epoch": 2.920854069674105,
+      "grad_norm": 0.9828323125839233,
+      "learning_rate": 4.959277281085129e-06,
+      "loss": 0.3797,
+      "step": 9097
+    },
+    {
+      "epoch": 2.9211751484989565,
+      "grad_norm": 0.9727553725242615,
+      "learning_rate": 4.948426837221631e-06,
+      "loss": 0.3328,
+      "step": 9098
+    },
+    {
+      "epoch": 2.921496227323808,
+      "grad_norm": 0.7968175411224365,
+      "learning_rate": 4.937587975103996e-06,
+      "loss": 0.2932,
+      "step": 9099
+    },
+    {
+      "epoch": 2.9218173061486596,
+      "grad_norm": 0.786050021648407,
+      "learning_rate": 4.926760696052934e-06,
+      "loss": 0.3092,
+      "step": 9100
+    },
+    {
+      "epoch": 2.922138384973511,
+      "grad_norm": 0.9386889338493347,
+      "learning_rate": 4.915945001387667e-06,
+      "loss": 0.3844,
+      "step": 9101
+    },
+    {
+      "epoch": 2.9224594637983623,
+      "grad_norm": 0.8530262112617493,
+      "learning_rate": 4.905140892426097e-06,
+      "loss": 0.3504,
+      "step": 9102
+    },
+    {
+      "epoch": 2.922780542623214,
+      "grad_norm": 0.7439628839492798,
+      "learning_rate": 4.8943483704846475e-06,
+      "loss": 0.3001,
+      "step": 9103
+    },
+    {
+      "epoch": 2.9231016214480654,
+      "grad_norm": 0.8572718501091003,
+      "learning_rate": 4.8835674368783666e-06,
+      "loss": 0.3333,
+      "step": 9104
+    },
+    {
+      "epoch": 2.923422700272917,
+      "grad_norm": 0.9241260886192322,
+      "learning_rate": 4.872798092920872e-06,
+      "loss": 0.3437,
+      "step": 9105
+    },
+    {
+      "epoch": 2.9237437790977685,
+      "grad_norm": 0.8410215973854065,
+      "learning_rate": 4.862040339924378e-06,
+      "loss": 0.3783,
+      "step": 9106
+    },
+    {
+      "epoch": 2.92406485792262,
+      "grad_norm": 1.0775259733200073,
+      "learning_rate": 4.8512941791996726e-06,
+      "loss": 0.3497,
+      "step": 9107
+    },
+    {
+      "epoch": 2.9243859367474716,
+      "grad_norm": 0.4644877016544342,
+      "learning_rate": 4.840559612056183e-06,
+      "loss": 0.2042,
+      "step": 9108
+    },
+    {
+      "epoch": 2.924707015572323,
+      "grad_norm": 0.7072948217391968,
+      "learning_rate": 4.829836639801843e-06,
+      "loss": 0.2708,
+      "step": 9109
+    },
+    {
+      "epoch": 2.9250280943971747,
+      "grad_norm": 0.8432921767234802,
+      "learning_rate": 4.819125263743229e-06,
+      "loss": 0.313,
+      "step": 9110
+    },
+    {
+      "epoch": 2.925349173222026,
+      "grad_norm": 0.8821758031845093,
+      "learning_rate": 4.808425485185486e-06,
+      "loss": 0.4026,
+      "step": 9111
+    },
+    {
+      "epoch": 2.9256702520468774,
+      "grad_norm": 1.0225335359573364,
+      "learning_rate": 4.797737305432337e-06,
+      "loss": 0.463,
+      "step": 9112
+    },
+    {
+      "epoch": 2.925991330871729,
+      "grad_norm": 0.8588323593139648,
+      "learning_rate": 4.7870607257861415e-06,
+      "loss": 0.3394,
+      "step": 9113
+    },
+    {
+      "epoch": 2.9263124096965805,
+      "grad_norm": 0.6533500552177429,
+      "learning_rate": 4.776395747547757e-06,
+      "loss": 0.2785,
+      "step": 9114
+    },
+    {
+      "epoch": 2.926633488521432,
+      "grad_norm": 1.2023701667785645,
+      "learning_rate": 4.765742372016735e-06,
+      "loss": 0.2919,
+      "step": 9115
+    },
+    {
+      "epoch": 2.9269545673462836,
+      "grad_norm": 1.0757431983947754,
+      "learning_rate": 4.755100600491102e-06,
+      "loss": 0.4093,
+      "step": 9116
+    },
+    {
+      "epoch": 2.927275646171135,
+      "grad_norm": 1.0217454433441162,
+      "learning_rate": 4.744470434267567e-06,
+      "loss": 0.3653,
+      "step": 9117
+    },
+    {
+      "epoch": 2.9275967249959867,
+      "grad_norm": 0.9355355501174927,
+      "learning_rate": 4.733851874641382e-06,
+      "loss": 0.3976,
+      "step": 9118
+    },
+    {
+      "epoch": 2.927917803820838,
+      "grad_norm": 1.0048413276672363,
+      "learning_rate": 4.723244922906356e-06,
+      "loss": 0.3092,
+      "step": 9119
+    },
+    {
+      "epoch": 2.9282388826456893,
+      "grad_norm": 0.6854848861694336,
+      "learning_rate": 4.712649580354933e-06,
+      "loss": 0.3157,
+      "step": 9120
+    },
+    {
+      "epoch": 2.928559961470541,
+      "grad_norm": 1.0658174753189087,
+      "learning_rate": 4.702065848278126e-06,
+      "loss": 0.3169,
+      "step": 9121
+    },
+    {
+      "epoch": 2.9288810402953924,
+      "grad_norm": 1.368660807609558,
+      "learning_rate": 4.691493727965512e-06,
+      "loss": 0.4592,
+      "step": 9122
+    },
+    {
+      "epoch": 2.929202119120244,
+      "grad_norm": 0.9535953402519226,
+      "learning_rate": 4.680933220705308e-06,
+      "loss": 0.3418,
+      "step": 9123
+    },
+    {
+      "epoch": 2.9295231979450955,
+      "grad_norm": 0.46225616335868835,
+      "learning_rate": 4.670384327784239e-06,
+      "loss": 0.2457,
+      "step": 9124
+    },
+    {
+      "epoch": 2.929844276769947,
+      "grad_norm": 0.45245954394340515,
+      "learning_rate": 4.659847050487687e-06,
+      "loss": 0.248,
+      "step": 9125
+    },
+    {
+      "epoch": 2.9301653555947986,
+      "grad_norm": 0.48897916078567505,
+      "learning_rate": 4.64932139009957e-06,
+      "loss": 0.261,
+      "step": 9126
+    },
+    {
+      "epoch": 2.93048643441965,
+      "grad_norm": 0.7464796900749207,
+      "learning_rate": 4.638807347902408e-06,
+      "loss": 0.2816,
+      "step": 9127
+    },
+    {
+      "epoch": 2.9308075132445017,
+      "grad_norm": 0.8038036823272705,
+      "learning_rate": 4.628304925177318e-06,
+      "loss": 0.3102,
+      "step": 9128
+    },
+    {
+      "epoch": 2.931128592069353,
+      "grad_norm": 0.49647238850593567,
+      "learning_rate": 4.617814123203967e-06,
+      "loss": 0.2667,
+      "step": 9129
+    },
+    {
+      "epoch": 2.9314496708942044,
+      "grad_norm": 0.3765721917152405,
+      "learning_rate": 4.607334943260655e-06,
+      "loss": 0.3785,
+      "step": 9130
+    },
+    {
+      "epoch": 2.931770749719056,
+      "grad_norm": 0.40889376401901245,
+      "learning_rate": 4.596867386624215e-06,
+      "loss": 0.3646,
+      "step": 9131
+    },
+    {
+      "epoch": 2.9320918285439075,
+      "grad_norm": 0.600988507270813,
+      "learning_rate": 4.586411454570083e-06,
+      "loss": 0.3995,
+      "step": 9132
+    },
+    {
+      "epoch": 2.932412907368759,
+      "grad_norm": 0.4408535659313202,
+      "learning_rate": 4.575967148372317e-06,
+      "loss": 0.1431,
+      "step": 9133
+    },
+    {
+      "epoch": 2.9327339861936106,
+      "grad_norm": 0.45499712228775024,
+      "learning_rate": 4.5655344693034895e-06,
+      "loss": 0.2273,
+      "step": 9134
+    },
+    {
+      "epoch": 2.933055065018462,
+      "grad_norm": 0.5091567039489746,
+      "learning_rate": 4.555113418634805e-06,
+      "loss": 0.123,
+      "step": 9135
+    },
+    {
+      "epoch": 2.9333761438433137,
+      "grad_norm": 0.4760441780090332,
+      "learning_rate": 4.544703997636046e-06,
+      "loss": 0.2153,
+      "step": 9136
+    },
+    {
+      "epoch": 2.9336972226681652,
+      "grad_norm": 0.5263906717300415,
+      "learning_rate": 4.534306207575545e-06,
+      "loss": 0.2508,
+      "step": 9137
+    },
+    {
+      "epoch": 2.9340183014930163,
+      "grad_norm": 0.9112762212753296,
+      "learning_rate": 4.523920049720265e-06,
+      "loss": 0.5172,
+      "step": 9138
+    },
+    {
+      "epoch": 2.934339380317868,
+      "grad_norm": 0.9089072346687317,
+      "learning_rate": 4.513545525335705e-06,
+      "loss": 0.4721,
+      "step": 9139
+    },
+    {
+      "epoch": 2.9346604591427194,
+      "grad_norm": 0.6721816062927246,
+      "learning_rate": 4.5031826356859876e-06,
+      "loss": 0.2626,
+      "step": 9140
+    },
+    {
+      "epoch": 2.934981537967571,
+      "grad_norm": 0.9276089072227478,
+      "learning_rate": 4.492831382033791e-06,
+      "loss": 0.3248,
+      "step": 9141
+    },
+    {
+      "epoch": 2.9353026167924225,
+      "grad_norm": 0.9039839506149292,
+      "learning_rate": 4.482491765640395e-06,
+      "loss": 0.3331,
+      "step": 9142
+    },
+    {
+      "epoch": 2.935623695617274,
+      "grad_norm": 0.983473539352417,
+      "learning_rate": 4.4721637877656375e-06,
+      "loss": 0.3217,
+      "step": 9143
+    },
+    {
+      "epoch": 2.9359447744421256,
+      "grad_norm": 0.7742102146148682,
+      "learning_rate": 4.461847449667955e-06,
+      "loss": 0.3411,
+      "step": 9144
+    },
+    {
+      "epoch": 2.936265853266977,
+      "grad_norm": 0.8041900396347046,
+      "learning_rate": 4.451542752604365e-06,
+      "loss": 0.3223,
+      "step": 9145
+    },
+    {
+      "epoch": 2.9365869320918287,
+      "grad_norm": 0.7607382535934448,
+      "learning_rate": 4.4412496978304515e-06,
+      "loss": 0.2994,
+      "step": 9146
+    },
+    {
+      "epoch": 2.93690801091668,
+      "grad_norm": 0.6458768844604492,
+      "learning_rate": 4.4309682866004125e-06,
+      "loss": 0.2517,
+      "step": 9147
+    },
+    {
+      "epoch": 2.9372290897415314,
+      "grad_norm": 0.8893207311630249,
+      "learning_rate": 4.420698520166988e-06,
+      "loss": 0.3428,
+      "step": 9148
+    },
+    {
+      "epoch": 2.937550168566383,
+      "grad_norm": 1.1111611127853394,
+      "learning_rate": 4.410440399781534e-06,
+      "loss": 0.3199,
+      "step": 9149
+    },
+    {
+      "epoch": 2.9378712473912345,
+      "grad_norm": 0.8783401846885681,
+      "learning_rate": 4.400193926693952e-06,
+      "loss": 0.3904,
+      "step": 9150
+    },
+    {
+      "epoch": 2.938192326216086,
+      "grad_norm": 0.6565034985542297,
+      "learning_rate": 4.389959102152774e-06,
+      "loss": 0.2862,
+      "step": 9151
+    },
+    {
+      "epoch": 2.9385134050409376,
+      "grad_norm": 0.8540425300598145,
+      "learning_rate": 4.379735927405038e-06,
+      "loss": 0.2698,
+      "step": 9152
+    },
+    {
+      "epoch": 2.938834483865789,
+      "grad_norm": 1.0489143133163452,
+      "learning_rate": 4.369524403696457e-06,
+      "loss": 0.3294,
+      "step": 9153
+    },
+    {
+      "epoch": 2.9391555626906407,
+      "grad_norm": 1.0321301221847534,
+      "learning_rate": 4.3593245322712474e-06,
+      "loss": 0.351,
+      "step": 9154
+    },
+    {
+      "epoch": 2.9394766415154923,
+      "grad_norm": 0.7734938263893127,
+      "learning_rate": 4.349136314372204e-06,
+      "loss": 0.3194,
+      "step": 9155
+    },
+    {
+      "epoch": 2.9397977203403434,
+      "grad_norm": 0.8938072919845581,
+      "learning_rate": 4.338959751240801e-06,
+      "loss": 0.3961,
+      "step": 9156
+    },
+    {
+      "epoch": 2.940118799165195,
+      "grad_norm": 0.6044966578483582,
+      "learning_rate": 4.328794844116946e-06,
+      "loss": 0.2416,
+      "step": 9157
+    },
+    {
+      "epoch": 2.9404398779900465,
+      "grad_norm": 0.69122314453125,
+      "learning_rate": 4.318641594239259e-06,
+      "loss": 0.2702,
+      "step": 9158
+    },
+    {
+      "epoch": 2.940760956814898,
+      "grad_norm": 1.0131958723068237,
+      "learning_rate": 4.308500002844862e-06,
+      "loss": 0.362,
+      "step": 9159
+    },
+    {
+      "epoch": 2.9410820356397496,
+      "grad_norm": 0.8207650184631348,
+      "learning_rate": 4.2983700711694665e-06,
+      "loss": 0.3119,
+      "step": 9160
+    },
+    {
+      "epoch": 2.941403114464601,
+      "grad_norm": 0.6501420736312866,
+      "learning_rate": 4.288251800447385e-06,
+      "loss": 0.2507,
+      "step": 9161
+    },
+    {
+      "epoch": 2.9417241932894527,
+      "grad_norm": 0.638103187084198,
+      "learning_rate": 4.278145191911509e-06,
+      "loss": 0.279,
+      "step": 9162
+    },
+    {
+      "epoch": 2.942045272114304,
+      "grad_norm": 0.9142710566520691,
+      "learning_rate": 4.268050246793276e-06,
+      "loss": 0.416,
+      "step": 9163
+    },
+    {
+      "epoch": 2.9423663509391558,
+      "grad_norm": 0.9541748762130737,
+      "learning_rate": 4.257966966322735e-06,
+      "loss": 0.314,
+      "step": 9164
+    },
+    {
+      "epoch": 2.942687429764007,
+      "grad_norm": 0.9409205317497253,
+      "learning_rate": 4.247895351728504e-06,
+      "loss": 0.3532,
+      "step": 9165
+    },
+    {
+      "epoch": 2.9430085085888584,
+      "grad_norm": 0.777114748954773,
+      "learning_rate": 4.237835404237778e-06,
+      "loss": 0.3019,
+      "step": 9166
+    },
+    {
+      "epoch": 2.94332958741371,
+      "grad_norm": 0.796989917755127,
+      "learning_rate": 4.227787125076332e-06,
+      "loss": 0.251,
+      "step": 9167
+    },
+    {
+      "epoch": 2.9436506662385615,
+      "grad_norm": 0.7624577879905701,
+      "learning_rate": 4.217750515468522e-06,
+      "loss": 0.2639,
+      "step": 9168
+    },
+    {
+      "epoch": 2.943971745063413,
+      "grad_norm": 0.6462528109550476,
+      "learning_rate": 4.207725576637256e-06,
+      "loss": 0.2605,
+      "step": 9169
+    },
+    {
+      "epoch": 2.9442928238882646,
+      "grad_norm": 0.8757241368293762,
+      "learning_rate": 4.197712309804058e-06,
+      "loss": 0.2975,
+      "step": 9170
+    },
+    {
+      "epoch": 2.944613902713116,
+      "grad_norm": 0.9458180665969849,
+      "learning_rate": 4.187710716189042e-06,
+      "loss": 0.3727,
+      "step": 9171
+    },
+    {
+      "epoch": 2.9449349815379677,
+      "grad_norm": 0.7695515155792236,
+      "learning_rate": 4.177720797010831e-06,
+      "loss": 0.2814,
+      "step": 9172
+    },
+    {
+      "epoch": 2.9452560603628193,
+      "grad_norm": 1.1303104162216187,
+      "learning_rate": 4.167742553486675e-06,
+      "loss": 0.3553,
+      "step": 9173
+    },
+    {
+      "epoch": 2.9455771391876704,
+      "grad_norm": 0.6950247883796692,
+      "learning_rate": 4.1577759868324125e-06,
+      "loss": 0.2848,
+      "step": 9174
+    },
+    {
+      "epoch": 2.945898218012522,
+      "grad_norm": 0.5095248818397522,
+      "learning_rate": 4.147821098262405e-06,
+      "loss": 0.2629,
+      "step": 9175
+    },
+    {
+      "epoch": 2.9462192968373735,
+      "grad_norm": 0.6144112944602966,
+      "learning_rate": 4.137877888989672e-06,
+      "loss": 0.2893,
+      "step": 9176
+    },
+    {
+      "epoch": 2.946540375662225,
+      "grad_norm": 0.970645546913147,
+      "learning_rate": 4.127946360225721e-06,
+      "loss": 0.3173,
+      "step": 9177
+    },
+    {
+      "epoch": 2.9468614544870766,
+      "grad_norm": 0.5283989310264587,
+      "learning_rate": 4.118026513180695e-06,
+      "loss": 0.2829,
+      "step": 9178
+    },
+    {
+      "epoch": 2.947182533311928,
+      "grad_norm": 0.4269641935825348,
+      "learning_rate": 4.108118349063306e-06,
+      "loss": 0.2613,
+      "step": 9179
+    },
+    {
+      "epoch": 2.9475036121367797,
+      "grad_norm": 0.476938396692276,
+      "learning_rate": 4.09822186908082e-06,
+      "loss": 0.64,
+      "step": 9180
+    },
+    {
+      "epoch": 2.9478246909616312,
+      "grad_norm": 0.4373842477798462,
+      "learning_rate": 4.088337074439097e-06,
+      "loss": 0.4771,
+      "step": 9181
+    },
+    {
+      "epoch": 2.948145769786483,
+      "grad_norm": 0.4393393397331238,
+      "learning_rate": 4.078463966342571e-06,
+      "loss": 0.3074,
+      "step": 9182
+    },
+    {
+      "epoch": 2.948466848611334,
+      "grad_norm": 0.7229725122451782,
+      "learning_rate": 4.068602545994249e-06,
+      "loss": 0.4381,
+      "step": 9183
+    },
+    {
+      "epoch": 2.9487879274361855,
+      "grad_norm": 0.48776933550834656,
+      "learning_rate": 4.0587528145957235e-06,
+      "loss": 0.3591,
+      "step": 9184
+    },
+    {
+      "epoch": 2.949109006261037,
+      "grad_norm": 0.2947094738483429,
+      "learning_rate": 4.048914773347134e-06,
+      "loss": 0.1124,
+      "step": 9185
+    },
+    {
+      "epoch": 2.9494300850858886,
+      "grad_norm": 0.3417683243751526,
+      "learning_rate": 4.039088423447235e-06,
+      "loss": 0.1572,
+      "step": 9186
+    },
+    {
+      "epoch": 2.94975116391074,
+      "grad_norm": 0.2823491394519806,
+      "learning_rate": 4.029273766093333e-06,
+      "loss": 0.131,
+      "step": 9187
+    },
+    {
+      "epoch": 2.9500722427355917,
+      "grad_norm": 0.8070669770240784,
+      "learning_rate": 4.019470802481307e-06,
+      "loss": 0.437,
+      "step": 9188
+    },
+    {
+      "epoch": 2.950393321560443,
+      "grad_norm": 0.7624936699867249,
+      "learning_rate": 4.009679533805633e-06,
+      "loss": 0.3359,
+      "step": 9189
+    },
+    {
+      "epoch": 2.9507144003852948,
+      "grad_norm": 0.7309533357620239,
+      "learning_rate": 3.999899961259335e-06,
+      "loss": 0.3002,
+      "step": 9190
+    },
+    {
+      "epoch": 2.9510354792101463,
+      "grad_norm": 0.8740092515945435,
+      "learning_rate": 3.990132086034026e-06,
+      "loss": 0.4231,
+      "step": 9191
+    },
+    {
+      "epoch": 2.9513565580349974,
+      "grad_norm": 0.9942984580993652,
+      "learning_rate": 3.9803759093199e-06,
+      "loss": 0.4098,
+      "step": 9192
+    },
+    {
+      "epoch": 2.951677636859849,
+      "grad_norm": 0.6029365062713623,
+      "learning_rate": 3.970631432305694e-06,
+      "loss": 0.2482,
+      "step": 9193
+    },
+    {
+      "epoch": 2.9519987156847005,
+      "grad_norm": 0.9003799557685852,
+      "learning_rate": 3.96089865617878e-06,
+      "loss": 0.3181,
+      "step": 9194
+    },
+    {
+      "epoch": 2.952319794509552,
+      "grad_norm": 0.7407945990562439,
+      "learning_rate": 3.951177582125021e-06,
+      "loss": 0.2781,
+      "step": 9195
+    },
+    {
+      "epoch": 2.9526408733344036,
+      "grad_norm": 0.8002344965934753,
+      "learning_rate": 3.941468211328947e-06,
+      "loss": 0.3166,
+      "step": 9196
+    },
+    {
+      "epoch": 2.952961952159255,
+      "grad_norm": 0.7346023321151733,
+      "learning_rate": 3.931770544973601e-06,
+      "loss": 0.2666,
+      "step": 9197
+    },
+    {
+      "epoch": 2.9532830309841067,
+      "grad_norm": 0.8236327767372131,
+      "learning_rate": 3.922084584240582e-06,
+      "loss": 0.3224,
+      "step": 9198
+    },
+    {
+      "epoch": 2.953604109808958,
+      "grad_norm": 0.9807097315788269,
+      "learning_rate": 3.912410330310156e-06,
+      "loss": 0.4151,
+      "step": 9199
+    },
+    {
+      "epoch": 2.95392518863381,
+      "grad_norm": 0.6370247006416321,
+      "learning_rate": 3.902747784361038e-06,
+      "loss": 0.2875,
+      "step": 9200
+    },
+    {
+      "epoch": 2.954246267458661,
+      "grad_norm": 0.8275413513183594,
+      "learning_rate": 3.893096947570618e-06,
+      "loss": 0.3529,
+      "step": 9201
+    },
+    {
+      "epoch": 2.9545673462835125,
+      "grad_norm": 1.0060149431228638,
+      "learning_rate": 3.883457821114811e-06,
+      "loss": 0.3392,
+      "step": 9202
+    },
+    {
+      "epoch": 2.954888425108364,
+      "grad_norm": 0.9771203398704529,
+      "learning_rate": 3.873830406168111e-06,
+      "loss": 0.423,
+      "step": 9203
+    },
+    {
+      "epoch": 2.9552095039332156,
+      "grad_norm": 0.7501790523529053,
+      "learning_rate": 3.864214703903601e-06,
+      "loss": 0.2913,
+      "step": 9204
+    },
+    {
+      "epoch": 2.955530582758067,
+      "grad_norm": 1.0703070163726807,
+      "learning_rate": 3.8546107154929235e-06,
+      "loss": 0.4205,
+      "step": 9205
+    },
+    {
+      "epoch": 2.9558516615829187,
+      "grad_norm": 0.6897678971290588,
+      "learning_rate": 3.845018442106285e-06,
+      "loss": 0.2425,
+      "step": 9206
+    },
+    {
+      "epoch": 2.9561727404077702,
+      "grad_norm": 0.917190670967102,
+      "learning_rate": 3.835437884912474e-06,
+      "loss": 0.392,
+      "step": 9207
+    },
+    {
+      "epoch": 2.9564938192326213,
+      "grad_norm": 1.1160334348678589,
+      "learning_rate": 3.825869045078867e-06,
+      "loss": 0.4136,
+      "step": 9208
+    },
+    {
+      "epoch": 2.9568148980574733,
+      "grad_norm": 1.1579893827438354,
+      "learning_rate": 3.816311923771387e-06,
+      "loss": 0.4582,
+      "step": 9209
+    },
+    {
+      "epoch": 2.9571359768823244,
+      "grad_norm": 0.7560253739356995,
+      "learning_rate": 3.806766522154548e-06,
+      "loss": 0.3024,
+      "step": 9210
+    },
+    {
+      "epoch": 2.957457055707176,
+      "grad_norm": 1.1074497699737549,
+      "learning_rate": 3.797232841391407e-06,
+      "loss": 0.4429,
+      "step": 9211
+    },
+    {
+      "epoch": 2.9577781345320275,
+      "grad_norm": 0.941001832485199,
+      "learning_rate": 3.787710882643658e-06,
+      "loss": 0.3496,
+      "step": 9212
+    },
+    {
+      "epoch": 2.958099213356879,
+      "grad_norm": 0.8937235474586487,
+      "learning_rate": 3.7782006470714616e-06,
+      "loss": 0.325,
+      "step": 9213
+    },
+    {
+      "epoch": 2.9584202921817306,
+      "grad_norm": 1.424865961074829,
+      "learning_rate": 3.7687021358336683e-06,
+      "loss": 0.4345,
+      "step": 9214
+    },
+    {
+      "epoch": 2.958741371006582,
+      "grad_norm": 0.7509331703186035,
+      "learning_rate": 3.759215350087619e-06,
+      "loss": 0.2709,
+      "step": 9215
+    },
+    {
+      "epoch": 2.9590624498314337,
+      "grad_norm": 0.6785710453987122,
+      "learning_rate": 3.7497402909892342e-06,
+      "loss": 0.237,
+      "step": 9216
+    },
+    {
+      "epoch": 2.959383528656285,
+      "grad_norm": 0.7602850794792175,
+      "learning_rate": 3.7402769596930563e-06,
+      "loss": 0.3031,
+      "step": 9217
+    },
+    {
+      "epoch": 2.959704607481137,
+      "grad_norm": 1.575135588645935,
+      "learning_rate": 3.730825357352119e-06,
+      "loss": 0.3567,
+      "step": 9218
+    },
+    {
+      "epoch": 2.960025686305988,
+      "grad_norm": 0.6166884899139404,
+      "learning_rate": 3.721385485118123e-06,
+      "loss": 0.2792,
+      "step": 9219
+    },
+    {
+      "epoch": 2.9603467651308395,
+      "grad_norm": 0.6132175922393799,
+      "learning_rate": 3.711957344141237e-06,
+      "loss": 0.2788,
+      "step": 9220
+    },
+    {
+      "epoch": 2.960667843955691,
+      "grad_norm": 0.7397081255912781,
+      "learning_rate": 3.7025409355702976e-06,
+      "loss": 0.3096,
+      "step": 9221
+    },
+    {
+      "epoch": 2.9609889227805426,
+      "grad_norm": 0.7033382654190063,
+      "learning_rate": 3.693136260552632e-06,
+      "loss": 0.2897,
+      "step": 9222
+    },
+    {
+      "epoch": 2.961310001605394,
+      "grad_norm": 0.62275230884552,
+      "learning_rate": 3.68374332023419e-06,
+      "loss": 0.2907,
+      "step": 9223
+    },
+    {
+      "epoch": 2.9616310804302457,
+      "grad_norm": 0.716752827167511,
+      "learning_rate": 3.6743621157594555e-06,
+      "loss": 0.2795,
+      "step": 9224
+    },
+    {
+      "epoch": 2.9619521592550972,
+      "grad_norm": 1.4299006462097168,
+      "learning_rate": 3.664992648271526e-06,
+      "loss": 0.2909,
+      "step": 9225
+    },
+    {
+      "epoch": 2.9622732380799484,
+      "grad_norm": 0.8173677921295166,
+      "learning_rate": 3.6556349189120097e-06,
+      "loss": 0.3142,
+      "step": 9226
+    },
+    {
+      "epoch": 2.9625943169048004,
+      "grad_norm": 0.9796795845031738,
+      "learning_rate": 3.6462889288211512e-06,
+      "loss": 0.3166,
+      "step": 9227
+    },
+    {
+      "epoch": 2.9629153957296515,
+      "grad_norm": 0.45657387375831604,
+      "learning_rate": 3.6369546791377052e-06,
+      "loss": 0.2369,
+      "step": 9228
+    },
+    {
+      "epoch": 2.963236474554503,
+      "grad_norm": 0.5515535473823547,
+      "learning_rate": 3.627632170999029e-06,
+      "loss": 0.2845,
+      "step": 9229
+    },
+    {
+      "epoch": 2.9635575533793546,
+      "grad_norm": 0.4276106357574463,
+      "learning_rate": 3.6183214055410586e-06,
+      "loss": 0.3697,
+      "step": 9230
+    },
+    {
+      "epoch": 2.963878632204206,
+      "grad_norm": 0.45924896001815796,
+      "learning_rate": 3.6090223838982417e-06,
+      "loss": 0.4946,
+      "step": 9231
+    },
+    {
+      "epoch": 2.9641997110290577,
+      "grad_norm": 0.29669415950775146,
+      "learning_rate": 3.5997351072036945e-06,
+      "loss": 0.142,
+      "step": 9232
+    },
+    {
+      "epoch": 2.964520789853909,
+      "grad_norm": 0.5130549669265747,
+      "learning_rate": 3.5904595765890005e-06,
+      "loss": 0.2495,
+      "step": 9233
+    },
+    {
+      "epoch": 2.9648418686787608,
+      "grad_norm": 0.23706403374671936,
+      "learning_rate": 3.5811957931843554e-06,
+      "loss": 0.0929,
+      "step": 9234
+    },
+    {
+      "epoch": 2.965162947503612,
+      "grad_norm": 0.2668483853340149,
+      "learning_rate": 3.5719437581185454e-06,
+      "loss": 0.1031,
+      "step": 9235
+    },
+    {
+      "epoch": 2.965484026328464,
+      "grad_norm": 0.812019407749176,
+      "learning_rate": 3.562703472518869e-06,
+      "loss": 0.3787,
+      "step": 9236
+    },
+    {
+      "epoch": 2.965805105153315,
+      "grad_norm": 0.8627951741218567,
+      "learning_rate": 3.553474937511281e-06,
+      "loss": 0.372,
+      "step": 9237
+    },
+    {
+      "epoch": 2.9661261839781665,
+      "grad_norm": 0.92494797706604,
+      "learning_rate": 3.5442581542201923e-06,
+      "loss": 0.3974,
+      "step": 9238
+    },
+    {
+      "epoch": 2.966447262803018,
+      "grad_norm": 0.898228108882904,
+      "learning_rate": 3.5350531237686724e-06,
+      "loss": 0.2598,
+      "step": 9239
+    },
+    {
+      "epoch": 2.9667683416278696,
+      "grad_norm": 0.7887357473373413,
+      "learning_rate": 3.5258598472783233e-06,
+      "loss": 0.2791,
+      "step": 9240
+    },
+    {
+      "epoch": 2.967089420452721,
+      "grad_norm": 0.8121463060379028,
+      "learning_rate": 3.516678325869316e-06,
+      "loss": 0.3809,
+      "step": 9241
+    },
+    {
+      "epoch": 2.9674104992775727,
+      "grad_norm": 0.7240415215492249,
+      "learning_rate": 3.5075085606604e-06,
+      "loss": 0.3064,
+      "step": 9242
+    },
+    {
+      "epoch": 2.9677315781024243,
+      "grad_norm": 1.0837162733078003,
+      "learning_rate": 3.4983505527688586e-06,
+      "loss": 0.3971,
+      "step": 9243
+    },
+    {
+      "epoch": 2.9680526569272754,
+      "grad_norm": 0.7732176780700684,
+      "learning_rate": 3.489204303310578e-06,
+      "loss": 0.2647,
+      "step": 9244
+    },
+    {
+      "epoch": 2.9683737357521274,
+      "grad_norm": 0.777204155921936,
+      "learning_rate": 3.480069813400022e-06,
+      "loss": 0.3172,
+      "step": 9245
+    },
+    {
+      "epoch": 2.9686948145769785,
+      "grad_norm": 0.9173356294631958,
+      "learning_rate": 3.470947084150167e-06,
+      "loss": 0.3888,
+      "step": 9246
+    },
+    {
+      "epoch": 2.96901589340183,
+      "grad_norm": 0.6100045442581177,
+      "learning_rate": 3.461836116672612e-06,
+      "loss": 0.2627,
+      "step": 9247
+    },
+    {
+      "epoch": 2.9693369722266816,
+      "grad_norm": 0.8461219668388367,
+      "learning_rate": 3.452736912077503e-06,
+      "loss": 0.342,
+      "step": 9248
+    },
+    {
+      "epoch": 2.969658051051533,
+      "grad_norm": 0.9259991645812988,
+      "learning_rate": 3.443649471473531e-06,
+      "loss": 0.4195,
+      "step": 9249
+    },
+    {
+      "epoch": 2.9699791298763847,
+      "grad_norm": 1.0365054607391357,
+      "learning_rate": 3.434573795967988e-06,
+      "loss": 0.3616,
+      "step": 9250
+    },
+    {
+      "epoch": 2.9703002087012362,
+      "grad_norm": 1.170913815498352,
+      "learning_rate": 3.425509886666711e-06,
+      "loss": 0.4531,
+      "step": 9251
+    },
+    {
+      "epoch": 2.970621287526088,
+      "grad_norm": 0.9523593187332153,
+      "learning_rate": 3.4164577446741174e-06,
+      "loss": 0.3797,
+      "step": 9252
+    },
+    {
+      "epoch": 2.970942366350939,
+      "grad_norm": 0.5297471284866333,
+      "learning_rate": 3.40741737109318e-06,
+      "loss": 0.2028,
+      "step": 9253
+    },
+    {
+      "epoch": 2.971263445175791,
+      "grad_norm": 0.84348064661026,
+      "learning_rate": 3.3983887670254177e-06,
+      "loss": 0.3622,
+      "step": 9254
+    },
+    {
+      "epoch": 2.971584524000642,
+      "grad_norm": 0.9488373398780823,
+      "learning_rate": 3.389371933570995e-06,
+      "loss": 0.3887,
+      "step": 9255
+    },
+    {
+      "epoch": 2.9719056028254935,
+      "grad_norm": 0.7091740369796753,
+      "learning_rate": 3.380366871828522e-06,
+      "loss": 0.2702,
+      "step": 9256
+    },
+    {
+      "epoch": 2.972226681650345,
+      "grad_norm": 0.7415878176689148,
+      "learning_rate": 3.3713735828952987e-06,
+      "loss": 0.2945,
+      "step": 9257
+    },
+    {
+      "epoch": 2.9725477604751966,
+      "grad_norm": 0.7940482497215271,
+      "learning_rate": 3.3623920678670597e-06,
+      "loss": 0.2931,
+      "step": 9258
+    },
+    {
+      "epoch": 2.972868839300048,
+      "grad_norm": 0.745911717414856,
+      "learning_rate": 3.3534223278382405e-06,
+      "loss": 0.278,
+      "step": 9259
+    },
+    {
+      "epoch": 2.9731899181248997,
+      "grad_norm": 0.6489874124526978,
+      "learning_rate": 3.344464363901756e-06,
+      "loss": 0.2768,
+      "step": 9260
+    },
+    {
+      "epoch": 2.9735109969497513,
+      "grad_norm": 0.9306034445762634,
+      "learning_rate": 3.3355181771490772e-06,
+      "loss": 0.3243,
+      "step": 9261
+    },
+    {
+      "epoch": 2.9738320757746024,
+      "grad_norm": 0.9093040823936462,
+      "learning_rate": 3.3265837686703106e-06,
+      "loss": 0.3566,
+      "step": 9262
+    },
+    {
+      "epoch": 2.9741531545994544,
+      "grad_norm": 0.7288286685943604,
+      "learning_rate": 3.3176611395540626e-06,
+      "loss": 0.2532,
+      "step": 9263
+    },
+    {
+      "epoch": 2.9744742334243055,
+      "grad_norm": 1.2597209215164185,
+      "learning_rate": 3.3087502908875413e-06,
+      "loss": 0.4041,
+      "step": 9264
+    },
+    {
+      "epoch": 2.974795312249157,
+      "grad_norm": 0.8495983481407166,
+      "learning_rate": 3.2998512237565005e-06,
+      "loss": 0.3545,
+      "step": 9265
+    },
+    {
+      "epoch": 2.9751163910740086,
+      "grad_norm": 0.9554901719093323,
+      "learning_rate": 3.290963939245262e-06,
+      "loss": 0.3384,
+      "step": 9266
+    },
+    {
+      "epoch": 2.97543746989886,
+      "grad_norm": 0.5555322766304016,
+      "learning_rate": 3.2820884384367146e-06,
+      "loss": 0.2401,
+      "step": 9267
+    },
+    {
+      "epoch": 2.9757585487237117,
+      "grad_norm": 0.5823652744293213,
+      "learning_rate": 3.273224722412327e-06,
+      "loss": 0.2434,
+      "step": 9268
+    },
+    {
+      "epoch": 2.9760796275485633,
+      "grad_norm": 0.7519864439964294,
+      "learning_rate": 3.2643727922520906e-06,
+      "loss": 0.2569,
+      "step": 9269
+    },
+    {
+      "epoch": 2.976400706373415,
+      "grad_norm": 0.6338984966278076,
+      "learning_rate": 3.2555326490346095e-06,
+      "loss": 0.2871,
+      "step": 9270
+    },
+    {
+      "epoch": 2.976721785198266,
+      "grad_norm": 0.7331449389457703,
+      "learning_rate": 3.246704293837011e-06,
+      "loss": 0.2932,
+      "step": 9271
+    },
+    {
+      "epoch": 2.977042864023118,
+      "grad_norm": 0.7755066752433777,
+      "learning_rate": 3.2378877277350116e-06,
+      "loss": 0.3737,
+      "step": 9272
+    },
+    {
+      "epoch": 2.977363942847969,
+      "grad_norm": 1.2024750709533691,
+      "learning_rate": 3.2290829518028862e-06,
+      "loss": 0.2638,
+      "step": 9273
+    },
+    {
+      "epoch": 2.9776850216728206,
+      "grad_norm": 0.7302016019821167,
+      "learning_rate": 3.2202899671134546e-06,
+      "loss": 0.2853,
+      "step": 9274
+    },
+    {
+      "epoch": 2.978006100497672,
+      "grad_norm": 0.6680705547332764,
+      "learning_rate": 3.2115087747381366e-06,
+      "loss": 0.2578,
+      "step": 9275
+    },
+    {
+      "epoch": 2.9783271793225237,
+      "grad_norm": 0.6737269163131714,
+      "learning_rate": 3.2027393757468773e-06,
+      "loss": 0.308,
+      "step": 9276
+    },
+    {
+      "epoch": 2.978648258147375,
+      "grad_norm": 0.5532433986663818,
+      "learning_rate": 3.19398177120821e-06,
+      "loss": 0.2525,
+      "step": 9277
+    },
+    {
+      "epoch": 2.9789693369722268,
+      "grad_norm": 0.49847257137298584,
+      "learning_rate": 3.1852359621892367e-06,
+      "loss": 0.2842,
+      "step": 9278
+    },
+    {
+      "epoch": 2.9792904157970783,
+      "grad_norm": 0.4007408022880554,
+      "learning_rate": 3.1765019497555616e-06,
+      "loss": 0.2705,
+      "step": 9279
+    },
+    {
+      "epoch": 2.9796114946219294,
+      "grad_norm": 0.4380526542663574,
+      "learning_rate": 3.1677797349714544e-06,
+      "loss": 0.4338,
+      "step": 9280
+    },
+    {
+      "epoch": 2.9799325734467814,
+      "grad_norm": 0.45914846658706665,
+      "learning_rate": 3.1590693188996323e-06,
+      "loss": 0.33,
+      "step": 9281
+    },
+    {
+      "epoch": 2.9802536522716325,
+      "grad_norm": 0.43808650970458984,
+      "learning_rate": 3.1503707026014906e-06,
+      "loss": 0.3765,
+      "step": 9282
+    },
+    {
+      "epoch": 2.980574731096484,
+      "grad_norm": 0.36031296849250793,
+      "learning_rate": 3.1416838871368924e-06,
+      "loss": 0.1771,
+      "step": 9283
+    },
+    {
+      "epoch": 2.9808958099213356,
+      "grad_norm": 0.28183841705322266,
+      "learning_rate": 3.1330088735643025e-06,
+      "loss": 0.1466,
+      "step": 9284
+    },
+    {
+      "epoch": 2.981216888746187,
+      "grad_norm": 0.4093777537345886,
+      "learning_rate": 3.124345662940764e-06,
+      "loss": 0.1417,
+      "step": 9285
+    },
+    {
+      "epoch": 2.9815379675710387,
+      "grad_norm": 0.3650287687778473,
+      "learning_rate": 3.115694256321855e-06,
+      "loss": 0.1786,
+      "step": 9286
+    },
+    {
+      "epoch": 2.9818590463958903,
+      "grad_norm": 0.7063071131706238,
+      "learning_rate": 3.1070546547617095e-06,
+      "loss": 0.3095,
+      "step": 9287
+    },
+    {
+      "epoch": 2.982180125220742,
+      "grad_norm": 0.9610477685928345,
+      "learning_rate": 3.0984268593130528e-06,
+      "loss": 0.4365,
+      "step": 9288
+    },
+    {
+      "epoch": 2.982501204045593,
+      "grad_norm": 0.9653912782669067,
+      "learning_rate": 3.0898108710271434e-06,
+      "loss": 0.3772,
+      "step": 9289
+    },
+    {
+      "epoch": 2.982822282870445,
+      "grad_norm": 0.7585698962211609,
+      "learning_rate": 3.081206690953831e-06,
+      "loss": 0.314,
+      "step": 9290
+    },
+    {
+      "epoch": 2.983143361695296,
+      "grad_norm": 0.867253839969635,
+      "learning_rate": 3.072614320141487e-06,
+      "loss": 0.3468,
+      "step": 9291
+    },
+    {
+      "epoch": 2.9834644405201476,
+      "grad_norm": 0.6894168257713318,
+      "learning_rate": 3.064033759637064e-06,
+      "loss": 0.2712,
+      "step": 9292
+    },
+    {
+      "epoch": 2.983785519344999,
+      "grad_norm": 0.5496251583099365,
+      "learning_rate": 3.0554650104861136e-06,
+      "loss": 0.2125,
+      "step": 9293
+    },
+    {
+      "epoch": 2.9841065981698507,
+      "grad_norm": 0.7446759343147278,
+      "learning_rate": 3.046908073732668e-06,
+      "loss": 0.3081,
+      "step": 9294
+    },
+    {
+      "epoch": 2.9844276769947022,
+      "grad_norm": 0.8124382495880127,
+      "learning_rate": 3.0383629504194046e-06,
+      "loss": 0.281,
+      "step": 9295
+    },
+    {
+      "epoch": 2.984748755819554,
+      "grad_norm": 1.0698877573013306,
+      "learning_rate": 3.0298296415874894e-06,
+      "loss": 0.3884,
+      "step": 9296
+    },
+    {
+      "epoch": 2.9850698346444053,
+      "grad_norm": 0.9363170862197876,
+      "learning_rate": 3.0213081482766805e-06,
+      "loss": 0.3386,
+      "step": 9297
+    },
+    {
+      "epoch": 2.9853909134692564,
+      "grad_norm": 0.909624457359314,
+      "learning_rate": 3.012798471525324e-06,
+      "loss": 0.3767,
+      "step": 9298
+    },
+    {
+      "epoch": 2.9857119922941084,
+      "grad_norm": 0.8160519003868103,
+      "learning_rate": 3.0043006123702697e-06,
+      "loss": 0.4078,
+      "step": 9299
+    },
+    {
+      "epoch": 2.9860330711189595,
+      "grad_norm": 0.7235801219940186,
+      "learning_rate": 2.9958145718469777e-06,
+      "loss": 0.3223,
+      "step": 9300
+    },
+    {
+      "epoch": 2.986354149943811,
+      "grad_norm": 0.9839329719543457,
+      "learning_rate": 2.9873403509894203e-06,
+      "loss": 0.4005,
+      "step": 9301
+    },
+    {
+      "epoch": 2.9866752287686626,
+      "grad_norm": 0.7384074926376343,
+      "learning_rate": 2.978877950830172e-06,
+      "loss": 0.2616,
+      "step": 9302
+    },
+    {
+      "epoch": 2.986996307593514,
+      "grad_norm": 0.7919383645057678,
+      "learning_rate": 2.970427372400353e-06,
+      "loss": 0.3024,
+      "step": 9303
+    },
+    {
+      "epoch": 2.9873173864183657,
+      "grad_norm": 0.8400102257728577,
+      "learning_rate": 2.9619886167296384e-06,
+      "loss": 0.3372,
+      "step": 9304
+    },
+    {
+      "epoch": 2.9876384652432173,
+      "grad_norm": 0.610448956489563,
+      "learning_rate": 2.953561684846262e-06,
+      "loss": 0.2516,
+      "step": 9305
+    },
+    {
+      "epoch": 2.987959544068069,
+      "grad_norm": 0.9138132333755493,
+      "learning_rate": 2.9451465777770247e-06,
+      "loss": 0.3427,
+      "step": 9306
+    },
+    {
+      "epoch": 2.98828062289292,
+      "grad_norm": 1.5538538694381714,
+      "learning_rate": 2.936743296547273e-06,
+      "loss": 0.4386,
+      "step": 9307
+    },
+    {
+      "epoch": 2.988601701717772,
+      "grad_norm": 0.7500302195549011,
+      "learning_rate": 2.928351842180921e-06,
+      "loss": 0.2692,
+      "step": 9308
+    },
+    {
+      "epoch": 2.988922780542623,
+      "grad_norm": 0.8792135119438171,
+      "learning_rate": 2.9199722157004616e-06,
+      "loss": 0.3637,
+      "step": 9309
+    },
+    {
+      "epoch": 2.9892438593674746,
+      "grad_norm": 0.4652367830276489,
+      "learning_rate": 2.9116044181269007e-06,
+      "loss": 0.2135,
+      "step": 9310
+    },
+    {
+      "epoch": 2.989564938192326,
+      "grad_norm": 0.9699134826660156,
+      "learning_rate": 2.9032484504798452e-06,
+      "loss": 0.3932,
+      "step": 9311
+    },
+    {
+      "epoch": 2.9898860170171777,
+      "grad_norm": 0.7483917474746704,
+      "learning_rate": 2.8949043137774355e-06,
+      "loss": 0.3142,
+      "step": 9312
+    },
+    {
+      "epoch": 2.9902070958420293,
+      "grad_norm": 0.6890571117401123,
+      "learning_rate": 2.8865720090364034e-06,
+      "loss": 0.3233,
+      "step": 9313
+    },
+    {
+      "epoch": 2.990528174666881,
+      "grad_norm": 0.7219890356063843,
+      "learning_rate": 2.878251537271981e-06,
+      "loss": 0.254,
+      "step": 9314
+    },
+    {
+      "epoch": 2.9908492534917324,
+      "grad_norm": 0.8167535662651062,
+      "learning_rate": 2.8699428994980017e-06,
+      "loss": 0.2907,
+      "step": 9315
+    },
+    {
+      "epoch": 2.9911703323165835,
+      "grad_norm": 0.8420015573501587,
+      "learning_rate": 2.8616460967268667e-06,
+      "loss": 0.3101,
+      "step": 9316
+    },
+    {
+      "epoch": 2.9914914111414355,
+      "grad_norm": 0.5992588996887207,
+      "learning_rate": 2.8533611299694783e-06,
+      "loss": 0.2537,
+      "step": 9317
+    },
+    {
+      "epoch": 2.9918124899662866,
+      "grad_norm": 0.9275438189506531,
+      "learning_rate": 2.845088000235396e-06,
+      "loss": 0.2914,
+      "step": 9318
+    },
+    {
+      "epoch": 2.992133568791138,
+      "grad_norm": 0.6742895245552063,
+      "learning_rate": 2.836826708532603e-06,
+      "loss": 0.2999,
+      "step": 9319
+    },
+    {
+      "epoch": 2.9924546476159897,
+      "grad_norm": 0.9248641729354858,
+      "learning_rate": 2.8285772558677705e-06,
+      "loss": 0.3232,
+      "step": 9320
+    },
+    {
+      "epoch": 2.992775726440841,
+      "grad_norm": 0.8363873362541199,
+      "learning_rate": 2.8203396432460506e-06,
+      "loss": 0.3208,
+      "step": 9321
+    },
+    {
+      "epoch": 2.9930968052656928,
+      "grad_norm": 0.5395601987838745,
+      "learning_rate": 2.8121138716711404e-06,
+      "loss": 0.2489,
+      "step": 9322
+    },
+    {
+      "epoch": 2.9934178840905443,
+      "grad_norm": 0.5238479375839233,
+      "learning_rate": 2.8038999421453826e-06,
+      "loss": 0.2793,
+      "step": 9323
+    },
+    {
+      "epoch": 2.993738962915396,
+      "grad_norm": 0.9257795214653015,
+      "learning_rate": 2.7956978556695767e-06,
+      "loss": 0.3502,
+      "step": 9324
+    },
+    {
+      "epoch": 2.994060041740247,
+      "grad_norm": 0.7510598301887512,
+      "learning_rate": 2.7875076132431344e-06,
+      "loss": 0.3333,
+      "step": 9325
+    },
+    {
+      "epoch": 2.994381120565099,
+      "grad_norm": 0.5098982453346252,
+      "learning_rate": 2.7793292158640126e-06,
+      "loss": 0.2552,
+      "step": 9326
+    },
+    {
+      "epoch": 2.99470219938995,
+      "grad_norm": 0.6713955402374268,
+      "learning_rate": 2.771162664528726e-06,
+      "loss": 0.2947,
+      "step": 9327
+    },
+    {
+      "epoch": 2.9950232782148016,
+      "grad_norm": 0.7259739637374878,
+      "learning_rate": 2.7630079602323442e-06,
+      "loss": 0.3072,
+      "step": 9328
+    },
+    {
+      "epoch": 2.995344357039653,
+      "grad_norm": 0.29132696986198425,
+      "learning_rate": 2.7548651039684846e-06,
+      "loss": 0.247,
+      "step": 9329
+    },
+    {
+      "epoch": 2.9956654358645047,
+      "grad_norm": 0.37000009417533875,
+      "learning_rate": 2.746734096729342e-06,
+      "loss": 0.2994,
+      "step": 9330
+    },
+    {
+      "epoch": 2.9959865146893563,
+      "grad_norm": 0.6013002395629883,
+      "learning_rate": 2.738614939505646e-06,
+      "loss": 0.2101,
+      "step": 9331
+    },
+    {
+      "epoch": 2.996307593514208,
+      "grad_norm": 0.7807649970054626,
+      "learning_rate": 2.7305076332867054e-06,
+      "loss": 0.3249,
+      "step": 9332
+    },
+    {
+      "epoch": 2.9966286723390594,
+      "grad_norm": 0.8677118420600891,
+      "learning_rate": 2.7224121790603517e-06,
+      "loss": 0.3319,
+      "step": 9333
+    },
+    {
+      "epoch": 2.9969497511639105,
+      "grad_norm": 0.7280771136283875,
+      "learning_rate": 2.714328577812997e-06,
+      "loss": 0.3257,
+      "step": 9334
+    },
+    {
+      "epoch": 2.9972708299887625,
+      "grad_norm": 1.6487222909927368,
+      "learning_rate": 2.7062568305295965e-06,
+      "loss": 0.2733,
+      "step": 9335
+    },
+    {
+      "epoch": 2.9975919088136136,
+      "grad_norm": 0.6912175416946411,
+      "learning_rate": 2.6981969381936978e-06,
+      "loss": 0.2817,
+      "step": 9336
+    },
+    {
+      "epoch": 2.997912987638465,
+      "grad_norm": 0.8025344610214233,
+      "learning_rate": 2.690148901787337e-06,
+      "loss": 0.327,
+      "step": 9337
+    },
+    {
+      "epoch": 2.9982340664633167,
+      "grad_norm": 0.6244665384292603,
+      "learning_rate": 2.6821127222911857e-06,
+      "loss": 0.2552,
+      "step": 9338
+    },
+    {
+      "epoch": 2.9985551452881682,
+      "grad_norm": 0.8349512219429016,
+      "learning_rate": 2.6740884006843825e-06,
+      "loss": 0.3398,
+      "step": 9339
+    },
+    {
+      "epoch": 2.99887622411302,
+      "grad_norm": 0.9518395066261292,
+      "learning_rate": 2.66607593794469e-06,
+      "loss": 0.3202,
+      "step": 9340
+    },
+    {
+      "epoch": 2.9991973029378713,
+      "grad_norm": 0.6690359115600586,
+      "learning_rate": 2.6580753350484046e-06,
+      "loss": 0.2847,
+      "step": 9341
+    },
+    {
+      "epoch": 2.999518381762723,
+      "grad_norm": 0.6814465522766113,
+      "learning_rate": 2.650086592970358e-06,
+      "loss": 0.28,
+      "step": 9342
+    },
+    {
+      "epoch": 2.999839460587574,
+      "grad_norm": 1.4310189485549927,
+      "learning_rate": 2.6421097126839712e-06,
+      "loss": 0.4217,
+      "step": 9343
+    },
+    {
+      "epoch": 3.0001605394124256,
+      "grad_norm": 0.43879497051239014,
+      "learning_rate": 2.6341446951612005e-06,
+      "loss": 0.3707,
+      "step": 9344
+    },
+    {
+      "epoch": 3.000481618237277,
+      "grad_norm": 0.40804460644721985,
+      "learning_rate": 2.6261915413725578e-06,
+      "loss": 0.4423,
+      "step": 9345
+    },
+    {
+      "epoch": 3.0008026970621287,
+      "grad_norm": 0.38268548250198364,
+      "learning_rate": 2.618250252287113e-06,
+      "loss": 0.2687,
+      "step": 9346
+    },
+    {
+      "epoch": 3.00112377588698,
+      "grad_norm": 0.2906617224216461,
+      "learning_rate": 2.610320828872481e-06,
+      "loss": 0.1438,
+      "step": 9347
+    },
+    {
+      "epoch": 3.0014448547118318,
+      "grad_norm": 0.2549183964729309,
+      "learning_rate": 2.6024032720948443e-06,
+      "loss": 0.1358,
+      "step": 9348
+    },
+    {
+      "epoch": 3.0017659335366833,
+      "grad_norm": 0.3139497637748718,
+      "learning_rate": 2.59449758291892e-06,
+      "loss": 0.1061,
+      "step": 9349
+    },
+    {
+      "epoch": 3.002087012361535,
+      "grad_norm": 0.11320850253105164,
+      "learning_rate": 2.5866037623080153e-06,
+      "loss": 0.0611,
+      "step": 9350
+    },
+    {
+      "epoch": 3.0024080911863864,
+      "grad_norm": 0.5909608602523804,
+      "learning_rate": 2.5787218112239496e-06,
+      "loss": 0.3883,
+      "step": 9351
+    },
+    {
+      "epoch": 3.002729170011238,
+      "grad_norm": 0.5687462687492371,
+      "learning_rate": 2.570851730627122e-06,
+      "loss": 0.2654,
+      "step": 9352
+    },
+    {
+      "epoch": 3.003050248836089,
+      "grad_norm": 0.6400092840194702,
+      "learning_rate": 2.5629935214764865e-06,
+      "loss": 0.302,
+      "step": 9353
+    },
+    {
+      "epoch": 3.0033713276609406,
+      "grad_norm": 0.46738189458847046,
+      "learning_rate": 2.5551471847295228e-06,
+      "loss": 0.1982,
+      "step": 9354
+    },
+    {
+      "epoch": 3.003692406485792,
+      "grad_norm": 0.5949174165725708,
+      "learning_rate": 2.5473127213422763e-06,
+      "loss": 0.2598,
+      "step": 9355
+    },
+    {
+      "epoch": 3.0040134853106437,
+      "grad_norm": 1.0250693559646606,
+      "learning_rate": 2.5394901322694067e-06,
+      "loss": 0.2021,
+      "step": 9356
+    },
+    {
+      "epoch": 3.0043345641354953,
+      "grad_norm": 0.5591093897819519,
+      "learning_rate": 2.531679418464006e-06,
+      "loss": 0.2286,
+      "step": 9357
+    },
+    {
+      "epoch": 3.004655642960347,
+      "grad_norm": 0.5650901198387146,
+      "learning_rate": 2.5238805808778242e-06,
+      "loss": 0.2416,
+      "step": 9358
+    },
+    {
+      "epoch": 3.0049767217851984,
+      "grad_norm": 0.5326758027076721,
+      "learning_rate": 2.516093620461124e-06,
+      "loss": 0.2221,
+      "step": 9359
+    },
+    {
+      "epoch": 3.00529780061005,
+      "grad_norm": 0.6107519865036011,
+      "learning_rate": 2.508318538162702e-06,
+      "loss": 0.2568,
+      "step": 9360
+    },
+    {
+      "epoch": 3.0056188794349015,
+      "grad_norm": 0.5060980319976807,
+      "learning_rate": 2.5005553349299547e-06,
+      "loss": 0.228,
+      "step": 9361
+    },
+    {
+      "epoch": 3.0059399582597526,
+      "grad_norm": 0.5858919024467468,
+      "learning_rate": 2.4928040117087827e-06,
+      "loss": 0.2552,
+      "step": 9362
+    },
+    {
+      "epoch": 3.006261037084604,
+      "grad_norm": 0.5963205099105835,
+      "learning_rate": 2.4850645694436736e-06,
+      "loss": 0.2336,
+      "step": 9363
+    },
+    {
+      "epoch": 3.0065821159094557,
+      "grad_norm": 0.5636211633682251,
+      "learning_rate": 2.4773370090776626e-06,
+      "loss": 0.1921,
+      "step": 9364
+    },
+    {
+      "epoch": 3.0069031947343072,
+      "grad_norm": 0.5564295649528503,
+      "learning_rate": 2.4696213315523074e-06,
+      "loss": 0.2321,
+      "step": 9365
+    },
+    {
+      "epoch": 3.007224273559159,
+      "grad_norm": 0.5773066282272339,
+      "learning_rate": 2.4619175378077565e-06,
+      "loss": 0.2308,
+      "step": 9366
+    },
+    {
+      "epoch": 3.0075453523840103,
+      "grad_norm": 0.6271586418151855,
+      "learning_rate": 2.4542256287826914e-06,
+      "loss": 0.2407,
+      "step": 9367
+    },
+    {
+      "epoch": 3.007866431208862,
+      "grad_norm": 0.5038358569145203,
+      "learning_rate": 2.446545605414341e-06,
+      "loss": 0.2377,
+      "step": 9368
+    },
+    {
+      "epoch": 3.0081875100337134,
+      "grad_norm": 0.6960586905479431,
+      "learning_rate": 2.4388774686385007e-06,
+      "loss": 0.2911,
+      "step": 9369
+    },
+    {
+      "epoch": 3.008508588858565,
+      "grad_norm": 0.5753246545791626,
+      "learning_rate": 2.4312212193895125e-06,
+      "loss": 0.2706,
+      "step": 9370
+    },
+    {
+      "epoch": 3.008829667683416,
+      "grad_norm": 0.6433031558990479,
+      "learning_rate": 2.4235768586002517e-06,
+      "loss": 0.2616,
+      "step": 9371
+    },
+    {
+      "epoch": 3.0091507465082676,
+      "grad_norm": 0.615394115447998,
+      "learning_rate": 2.415944387202174e-06,
+      "loss": 0.2654,
+      "step": 9372
+    },
+    {
+      "epoch": 3.009471825333119,
+      "grad_norm": 0.7627741694450378,
+      "learning_rate": 2.4083238061252567e-06,
+      "loss": 0.318,
+      "step": 9373
+    },
+    {
+      "epoch": 3.0097929041579707,
+      "grad_norm": 0.6070566773414612,
+      "learning_rate": 2.40071511629808e-06,
+      "loss": 0.2499,
+      "step": 9374
+    },
+    {
+      "epoch": 3.0101139829828223,
+      "grad_norm": 0.8189682364463806,
+      "learning_rate": 2.3931183186477026e-06,
+      "loss": 0.2997,
+      "step": 9375
+    },
+    {
+      "epoch": 3.010435061807674,
+      "grad_norm": 0.6750493049621582,
+      "learning_rate": 2.385533414099783e-06,
+      "loss": 0.2817,
+      "step": 9376
+    },
+    {
+      "epoch": 3.0107561406325254,
+      "grad_norm": 0.7114454507827759,
+      "learning_rate": 2.3779604035785273e-06,
+      "loss": 0.3113,
+      "step": 9377
+    },
+    {
+      "epoch": 3.011077219457377,
+      "grad_norm": 0.6532655954360962,
+      "learning_rate": 2.3703992880066638e-06,
+      "loss": 0.2617,
+      "step": 9378
+    },
+    {
+      "epoch": 3.0113982982822285,
+      "grad_norm": 0.44464078545570374,
+      "learning_rate": 2.3628500683055222e-06,
+      "loss": 0.2033,
+      "step": 9379
+    },
+    {
+      "epoch": 3.0117193771070796,
+      "grad_norm": 0.6115095615386963,
+      "learning_rate": 2.355312745394922e-06,
+      "loss": 0.2694,
+      "step": 9380
+    },
+    {
+      "epoch": 3.012040455931931,
+      "grad_norm": 0.3214268982410431,
+      "learning_rate": 2.3477873201932734e-06,
+      "loss": 0.1775,
+      "step": 9381
+    },
+    {
+      "epoch": 3.0123615347567827,
+      "grad_norm": 0.6672443151473999,
+      "learning_rate": 2.3402737936175425e-06,
+      "loss": 0.2725,
+      "step": 9382
+    },
+    {
+      "epoch": 3.0126826135816343,
+      "grad_norm": 0.5313708782196045,
+      "learning_rate": 2.332772166583208e-06,
+      "loss": 0.2661,
+      "step": 9383
+    },
+    {
+      "epoch": 3.013003692406486,
+      "grad_norm": 0.7770663499832153,
+      "learning_rate": 2.325282440004339e-06,
+      "loss": 0.3296,
+      "step": 9384
+    },
+    {
+      "epoch": 3.0133247712313374,
+      "grad_norm": 0.4733831584453583,
+      "learning_rate": 2.3178046147935175e-06,
+      "loss": 0.2151,
+      "step": 9385
+    },
+    {
+      "epoch": 3.013645850056189,
+      "grad_norm": 0.5805613398551941,
+      "learning_rate": 2.3103386918619018e-06,
+      "loss": 0.2395,
+      "step": 9386
+    },
+    {
+      "epoch": 3.0139669288810405,
+      "grad_norm": 0.6747123003005981,
+      "learning_rate": 2.3028846721191876e-06,
+      "loss": 0.3138,
+      "step": 9387
+    },
+    {
+      "epoch": 3.014288007705892,
+      "grad_norm": 0.4454551637172699,
+      "learning_rate": 2.295442556473637e-06,
+      "loss": 0.231,
+      "step": 9388
+    },
+    {
+      "epoch": 3.014609086530743,
+      "grad_norm": 0.5723088979721069,
+      "learning_rate": 2.288012345832047e-06,
+      "loss": 0.2234,
+      "step": 9389
+    },
+    {
+      "epoch": 3.0149301653555947,
+      "grad_norm": 0.5944519639015198,
+      "learning_rate": 2.2805940410997484e-06,
+      "loss": 0.2433,
+      "step": 9390
+    },
+    {
+      "epoch": 3.015251244180446,
+      "grad_norm": 0.32701706886291504,
+      "learning_rate": 2.273187643180652e-06,
+      "loss": 0.2194,
+      "step": 9391
+    },
+    {
+      "epoch": 3.0155723230052978,
+      "grad_norm": 0.37137144804000854,
+      "learning_rate": 2.2657931529772136e-06,
+      "loss": 0.2192,
+      "step": 9392
+    },
+    {
+      "epoch": 3.0158934018301493,
+      "grad_norm": 0.4273146092891693,
+      "learning_rate": 2.2584105713904125e-06,
+      "loss": 0.2554,
+      "step": 9393
+    },
+    {
+      "epoch": 3.016214480655001,
+      "grad_norm": 0.357513964176178,
+      "learning_rate": 2.2510398993198067e-06,
+      "loss": 0.393,
+      "step": 9394
+    },
+    {
+      "epoch": 3.0165355594798524,
+      "grad_norm": 0.5867639183998108,
+      "learning_rate": 2.2436811376634893e-06,
+      "loss": 0.7159,
+      "step": 9395
+    },
+    {
+      "epoch": 3.016856638304704,
+      "grad_norm": 0.3277376890182495,
+      "learning_rate": 2.2363342873180757e-06,
+      "loss": 0.2004,
+      "step": 9396
+    },
+    {
+      "epoch": 3.0171777171295555,
+      "grad_norm": 0.2729596793651581,
+      "learning_rate": 2.2289993491788064e-06,
+      "loss": 0.1336,
+      "step": 9397
+    },
+    {
+      "epoch": 3.0174987959544066,
+      "grad_norm": 0.2100820392370224,
+      "learning_rate": 2.2216763241393767e-06,
+      "loss": 0.092,
+      "step": 9398
+    },
+    {
+      "epoch": 3.017819874779258,
+      "grad_norm": 0.24260498583316803,
+      "learning_rate": 2.2143652130921176e-06,
+      "loss": 0.0919,
+      "step": 9399
+    },
+    {
+      "epoch": 3.0181409536041097,
+      "grad_norm": 0.4891859292984009,
+      "learning_rate": 2.2070660169278166e-06,
+      "loss": 0.2427,
+      "step": 9400
+    },
+    {
+      "epoch": 3.0184620324289613,
+      "grad_norm": 0.6580247282981873,
+      "learning_rate": 2.1997787365358958e-06,
+      "loss": 0.3484,
+      "step": 9401
+    },
+    {
+      "epoch": 3.018783111253813,
+      "grad_norm": 0.6362596154212952,
+      "learning_rate": 2.192503372804278e-06,
+      "loss": 0.2941,
+      "step": 9402
+    },
+    {
+      "epoch": 3.0191041900786644,
+      "grad_norm": 0.7072274088859558,
+      "learning_rate": 2.1852399266194314e-06,
+      "loss": 0.3115,
+      "step": 9403
+    },
+    {
+      "epoch": 3.019425268903516,
+      "grad_norm": 0.5164889097213745,
+      "learning_rate": 2.177988398866415e-06,
+      "loss": 0.2195,
+      "step": 9404
+    },
+    {
+      "epoch": 3.0197463477283675,
+      "grad_norm": 0.5418469905853271,
+      "learning_rate": 2.1707487904287672e-06,
+      "loss": 0.2464,
+      "step": 9405
+    },
+    {
+      "epoch": 3.020067426553219,
+      "grad_norm": 0.6452725529670715,
+      "learning_rate": 2.163521102188648e-06,
+      "loss": 0.2784,
+      "step": 9406
+    },
+    {
+      "epoch": 3.02038850537807,
+      "grad_norm": 0.5608880519866943,
+      "learning_rate": 2.156305335026698e-06,
+      "loss": 0.2553,
+      "step": 9407
+    },
+    {
+      "epoch": 3.0207095842029217,
+      "grad_norm": 0.4648946225643158,
+      "learning_rate": 2.1491014898221582e-06,
+      "loss": 0.1944,
+      "step": 9408
+    },
+    {
+      "epoch": 3.0210306630277732,
+      "grad_norm": 0.5879479050636292,
+      "learning_rate": 2.141909567452793e-06,
+      "loss": 0.2373,
+      "step": 9409
+    },
+    {
+      "epoch": 3.021351741852625,
+      "grad_norm": 0.6517860889434814,
+      "learning_rate": 2.134729568794902e-06,
+      "loss": 0.2796,
+      "step": 9410
+    },
+    {
+      "epoch": 3.0216728206774763,
+      "grad_norm": 1.0598267316818237,
+      "learning_rate": 2.1275614947233624e-06,
+      "loss": 0.249,
+      "step": 9411
+    },
+    {
+      "epoch": 3.021993899502328,
+      "grad_norm": 0.6839745044708252,
+      "learning_rate": 2.120405346111576e-06,
+      "loss": 0.269,
+      "step": 9412
+    },
+    {
+      "epoch": 3.0223149783271794,
+      "grad_norm": 0.6479743123054504,
+      "learning_rate": 2.1132611238315003e-06,
+      "loss": 0.2993,
+      "step": 9413
+    },
+    {
+      "epoch": 3.022636057152031,
+      "grad_norm": 0.5197052955627441,
+      "learning_rate": 2.1061288287536285e-06,
+      "loss": 0.2089,
+      "step": 9414
+    },
+    {
+      "epoch": 3.0229571359768825,
+      "grad_norm": 0.5962384939193726,
+      "learning_rate": 2.0990084617470206e-06,
+      "loss": 0.2327,
+      "step": 9415
+    },
+    {
+      "epoch": 3.0232782148017336,
+      "grad_norm": 1.2533619403839111,
+      "learning_rate": 2.0919000236792607e-06,
+      "loss": 0.3158,
+      "step": 9416
+    },
+    {
+      "epoch": 3.023599293626585,
+      "grad_norm": 0.5709370374679565,
+      "learning_rate": 2.084803515416511e-06,
+      "loss": 0.2074,
+      "step": 9417
+    },
+    {
+      "epoch": 3.0239203724514367,
+      "grad_norm": 0.8780946731567383,
+      "learning_rate": 2.0777189378234143e-06,
+      "loss": 0.356,
+      "step": 9418
+    },
+    {
+      "epoch": 3.0242414512762883,
+      "grad_norm": 0.6024912595748901,
+      "learning_rate": 2.0706462917632673e-06,
+      "loss": 0.2511,
+      "step": 9419
+    },
+    {
+      "epoch": 3.02456253010114,
+      "grad_norm": 0.6366291046142578,
+      "learning_rate": 2.0635855780978044e-06,
+      "loss": 0.2302,
+      "step": 9420
+    },
+    {
+      "epoch": 3.0248836089259914,
+      "grad_norm": 0.6438530683517456,
+      "learning_rate": 2.0565367976873584e-06,
+      "loss": 0.2584,
+      "step": 9421
+    },
+    {
+      "epoch": 3.025204687750843,
+      "grad_norm": 0.5729436874389648,
+      "learning_rate": 2.049499951390832e-06,
+      "loss": 0.237,
+      "step": 9422
+    },
+    {
+      "epoch": 3.0255257665756945,
+      "grad_norm": 0.7282575964927673,
+      "learning_rate": 2.0424750400655947e-06,
+      "loss": 0.2925,
+      "step": 9423
+    },
+    {
+      "epoch": 3.025846845400546,
+      "grad_norm": 0.5626926422119141,
+      "learning_rate": 2.0354620645676504e-06,
+      "loss": 0.2095,
+      "step": 9424
+    },
+    {
+      "epoch": 3.026167924225397,
+      "grad_norm": 0.5786295533180237,
+      "learning_rate": 2.0284610257514937e-06,
+      "loss": 0.2728,
+      "step": 9425
+    },
+    {
+      "epoch": 3.0264890030502487,
+      "grad_norm": 0.48927032947540283,
+      "learning_rate": 2.021471924470175e-06,
+      "loss": 0.201,
+      "step": 9426
+    },
+    {
+      "epoch": 3.0268100818751003,
+      "grad_norm": 0.5696834921836853,
+      "learning_rate": 2.014494761575314e-06,
+      "loss": 0.2512,
+      "step": 9427
+    },
+    {
+      "epoch": 3.027131160699952,
+      "grad_norm": 0.32591769099235535,
+      "learning_rate": 2.0075295379170412e-06,
+      "loss": 0.1816,
+      "step": 9428
+    },
+    {
+      "epoch": 3.0274522395248034,
+      "grad_norm": 0.4733005464076996,
+      "learning_rate": 2.0005762543440445e-06,
+      "loss": 0.2298,
+      "step": 9429
+    },
+    {
+      "epoch": 3.027773318349655,
+      "grad_norm": 0.42047736048698425,
+      "learning_rate": 1.993634911703579e-06,
+      "loss": 0.1959,
+      "step": 9430
+    },
+    {
+      "epoch": 3.0280943971745065,
+      "grad_norm": 0.5498672127723694,
+      "learning_rate": 1.986705510841402e-06,
+      "loss": 0.2253,
+      "step": 9431
+    },
+    {
+      "epoch": 3.028415475999358,
+      "grad_norm": 0.5200008153915405,
+      "learning_rate": 1.979788052601861e-06,
+      "loss": 0.2527,
+      "step": 9432
+    },
+    {
+      "epoch": 3.0287365548242096,
+      "grad_norm": 0.5989855527877808,
+      "learning_rate": 1.9728825378278246e-06,
+      "loss": 0.2299,
+      "step": 9433
+    },
+    {
+      "epoch": 3.0290576336490607,
+      "grad_norm": 0.48008453845977783,
+      "learning_rate": 1.965988967360688e-06,
+      "loss": 0.2221,
+      "step": 9434
+    },
+    {
+      "epoch": 3.029378712473912,
+      "grad_norm": 0.6560676097869873,
+      "learning_rate": 1.9591073420404337e-06,
+      "loss": 0.2343,
+      "step": 9435
+    },
+    {
+      "epoch": 3.0296997912987638,
+      "grad_norm": 0.5400110483169556,
+      "learning_rate": 1.9522376627055583e-06,
+      "loss": 0.2449,
+      "step": 9436
+    },
+    {
+      "epoch": 3.0300208701236153,
+      "grad_norm": 0.4094994366168976,
+      "learning_rate": 1.945379930193125e-06,
+      "loss": 0.2177,
+      "step": 9437
+    },
+    {
+      "epoch": 3.030341948948467,
+      "grad_norm": 0.37665337324142456,
+      "learning_rate": 1.9385341453386995e-06,
+      "loss": 0.2032,
+      "step": 9438
+    },
+    {
+      "epoch": 3.0306630277733184,
+      "grad_norm": 0.6781096458435059,
+      "learning_rate": 1.931700308976436e-06,
+      "loss": 0.2793,
+      "step": 9439
+    },
+    {
+      "epoch": 3.03098410659817,
+      "grad_norm": 0.43711480498313904,
+      "learning_rate": 1.924878421939036e-06,
+      "loss": 0.2372,
+      "step": 9440
+    },
+    {
+      "epoch": 3.0313051854230215,
+      "grad_norm": 0.8434777855873108,
+      "learning_rate": 1.918068485057689e-06,
+      "loss": 0.2299,
+      "step": 9441
+    },
+    {
+      "epoch": 3.031626264247873,
+      "grad_norm": 0.38669654726982117,
+      "learning_rate": 1.911270499162199e-06,
+      "loss": 0.2403,
+      "step": 9442
+    },
+    {
+      "epoch": 3.031947343072724,
+      "grad_norm": 0.29554906487464905,
+      "learning_rate": 1.904484465080847e-06,
+      "loss": 0.2357,
+      "step": 9443
+    },
+    {
+      "epoch": 3.0322684218975757,
+      "grad_norm": 0.5543558597564697,
+      "learning_rate": 1.8977103836405053e-06,
+      "loss": 0.3659,
+      "step": 9444
+    },
+    {
+      "epoch": 3.0325895007224273,
+      "grad_norm": 0.2872603237628937,
+      "learning_rate": 1.8909482556666024e-06,
+      "loss": 0.2276,
+      "step": 9445
+    },
+    {
+      "epoch": 3.032910579547279,
+      "grad_norm": 0.3142801523208618,
+      "learning_rate": 1.8841980819830351e-06,
+      "loss": 0.1512,
+      "step": 9446
+    },
+    {
+      "epoch": 3.0332316583721304,
+      "grad_norm": 0.301505446434021,
+      "learning_rate": 1.8774598634123232e-06,
+      "loss": 0.1045,
+      "step": 9447
+    },
+    {
+      "epoch": 3.033552737196982,
+      "grad_norm": 0.5997371077537537,
+      "learning_rate": 1.8707336007754873e-06,
+      "loss": 0.2805,
+      "step": 9448
+    },
+    {
+      "epoch": 3.0338738160218335,
+      "grad_norm": 0.7243043184280396,
+      "learning_rate": 1.8640192948921054e-06,
+      "loss": 0.3329,
+      "step": 9449
+    },
+    {
+      "epoch": 3.034194894846685,
+      "grad_norm": 0.6885005831718445,
+      "learning_rate": 1.8573169465802898e-06,
+      "loss": 0.3327,
+      "step": 9450
+    },
+    {
+      "epoch": 3.034515973671536,
+      "grad_norm": 0.6366089582443237,
+      "learning_rate": 1.8506265566567094e-06,
+      "loss": 0.2606,
+      "step": 9451
+    },
+    {
+      "epoch": 3.0348370524963877,
+      "grad_norm": 0.5336918234825134,
+      "learning_rate": 1.8439481259365675e-06,
+      "loss": 0.2271,
+      "step": 9452
+    },
+    {
+      "epoch": 3.0351581313212392,
+      "grad_norm": 0.4865361154079437,
+      "learning_rate": 1.8372816552336026e-06,
+      "loss": 0.1939,
+      "step": 9453
+    },
+    {
+      "epoch": 3.035479210146091,
+      "grad_norm": 0.6160674095153809,
+      "learning_rate": 1.8306271453601199e-06,
+      "loss": 0.2288,
+      "step": 9454
+    },
+    {
+      "epoch": 3.0358002889709423,
+      "grad_norm": 0.49999356269836426,
+      "learning_rate": 1.8239845971269266e-06,
+      "loss": 0.1962,
+      "step": 9455
+    },
+    {
+      "epoch": 3.036121367795794,
+      "grad_norm": 0.8584988713264465,
+      "learning_rate": 1.8173540113434194e-06,
+      "loss": 0.3474,
+      "step": 9456
+    },
+    {
+      "epoch": 3.0364424466206454,
+      "grad_norm": 0.5525678396224976,
+      "learning_rate": 1.8107353888175083e-06,
+      "loss": 0.2213,
+      "step": 9457
+    },
+    {
+      "epoch": 3.036763525445497,
+      "grad_norm": 0.5510037541389465,
+      "learning_rate": 1.8041287303556364e-06,
+      "loss": 0.2106,
+      "step": 9458
+    },
+    {
+      "epoch": 3.0370846042703485,
+      "grad_norm": 0.8154999613761902,
+      "learning_rate": 1.7975340367628268e-06,
+      "loss": 0.2556,
+      "step": 9459
+    },
+    {
+      "epoch": 3.0374056830951996,
+      "grad_norm": 0.7554807662963867,
+      "learning_rate": 1.7909513088426255e-06,
+      "loss": 0.3069,
+      "step": 9460
+    },
+    {
+      "epoch": 3.037726761920051,
+      "grad_norm": 0.6828257441520691,
+      "learning_rate": 1.7843805473970798e-06,
+      "loss": 0.277,
+      "step": 9461
+    },
+    {
+      "epoch": 3.0380478407449028,
+      "grad_norm": 1.2714136838912964,
+      "learning_rate": 1.7778217532268714e-06,
+      "loss": 0.2337,
+      "step": 9462
+    },
+    {
+      "epoch": 3.0383689195697543,
+      "grad_norm": 1.116326093673706,
+      "learning_rate": 1.771274927131139e-06,
+      "loss": 0.3795,
+      "step": 9463
+    },
+    {
+      "epoch": 3.038689998394606,
+      "grad_norm": 0.7132155299186707,
+      "learning_rate": 1.7647400699075888e-06,
+      "loss": 0.2681,
+      "step": 9464
+    },
+    {
+      "epoch": 3.0390110772194574,
+      "grad_norm": 0.750069797039032,
+      "learning_rate": 1.7582171823524951e-06,
+      "loss": 0.2974,
+      "step": 9465
+    },
+    {
+      "epoch": 3.039332156044309,
+      "grad_norm": 0.6962805986404419,
+      "learning_rate": 1.751706265260611e-06,
+      "loss": 0.246,
+      "step": 9466
+    },
+    {
+      "epoch": 3.0396532348691605,
+      "grad_norm": 0.6339988708496094,
+      "learning_rate": 1.7452073194253238e-06,
+      "loss": 0.2297,
+      "step": 9467
+    },
+    {
+      "epoch": 3.039974313694012,
+      "grad_norm": 0.7461841702461243,
+      "learning_rate": 1.7387203456384782e-06,
+      "loss": 0.3137,
+      "step": 9468
+    },
+    {
+      "epoch": 3.040295392518863,
+      "grad_norm": 0.6014420986175537,
+      "learning_rate": 1.7322453446905085e-06,
+      "loss": 0.2633,
+      "step": 9469
+    },
+    {
+      "epoch": 3.0406164713437147,
+      "grad_norm": 1.0465008020401,
+      "learning_rate": 1.7257823173703503e-06,
+      "loss": 0.352,
+      "step": 9470
+    },
+    {
+      "epoch": 3.0409375501685663,
+      "grad_norm": 0.4762720465660095,
+      "learning_rate": 1.719331264465529e-06,
+      "loss": 0.224,
+      "step": 9471
+    },
+    {
+      "epoch": 3.041258628993418,
+      "grad_norm": 0.6413834095001221,
+      "learning_rate": 1.712892186762083e-06,
+      "loss": 0.25,
+      "step": 9472
+    },
+    {
+      "epoch": 3.0415797078182694,
+      "grad_norm": 0.46483296155929565,
+      "learning_rate": 1.706465085044584e-06,
+      "loss": 0.2117,
+      "step": 9473
+    },
+    {
+      "epoch": 3.041900786643121,
+      "grad_norm": 0.49583905935287476,
+      "learning_rate": 1.7000499600961505e-06,
+      "loss": 0.198,
+      "step": 9474
+    },
+    {
+      "epoch": 3.0422218654679725,
+      "grad_norm": 0.6456876993179321,
+      "learning_rate": 1.6936468126984572e-06,
+      "loss": 0.2849,
+      "step": 9475
+    },
+    {
+      "epoch": 3.042542944292824,
+      "grad_norm": 0.589216947555542,
+      "learning_rate": 1.6872556436317022e-06,
+      "loss": 0.228,
+      "step": 9476
+    },
+    {
+      "epoch": 3.0428640231176756,
+      "grad_norm": 0.5428295135498047,
+      "learning_rate": 1.680876453674629e-06,
+      "loss": 0.2151,
+      "step": 9477
+    },
+    {
+      "epoch": 3.0431851019425267,
+      "grad_norm": 0.5795981884002686,
+      "learning_rate": 1.6745092436045494e-06,
+      "loss": 0.2326,
+      "step": 9478
+    },
+    {
+      "epoch": 3.043506180767378,
+      "grad_norm": 0.6216808557510376,
+      "learning_rate": 1.6681540141972429e-06,
+      "loss": 0.2701,
+      "step": 9479
+    },
+    {
+      "epoch": 3.0438272595922298,
+      "grad_norm": 0.5757028460502625,
+      "learning_rate": 1.661810766227112e-06,
+      "loss": 0.2414,
+      "step": 9480
+    },
+    {
+      "epoch": 3.0441483384170813,
+      "grad_norm": 0.731472373008728,
+      "learning_rate": 1.6554795004670388e-06,
+      "loss": 0.259,
+      "step": 9481
+    },
+    {
+      "epoch": 3.044469417241933,
+      "grad_norm": 0.6411800384521484,
+      "learning_rate": 1.6491602176884724e-06,
+      "loss": 0.2675,
+      "step": 9482
+    },
+    {
+      "epoch": 3.0447904960667844,
+      "grad_norm": 0.40002959966659546,
+      "learning_rate": 1.6428529186614195e-06,
+      "loss": 0.2122,
+      "step": 9483
+    },
+    {
+      "epoch": 3.045111574891636,
+      "grad_norm": 0.9253550171852112,
+      "learning_rate": 1.636557604154365e-06,
+      "loss": 0.3639,
+      "step": 9484
+    },
+    {
+      "epoch": 3.0454326537164875,
+      "grad_norm": 0.5021322965621948,
+      "learning_rate": 1.6302742749344291e-06,
+      "loss": 0.2332,
+      "step": 9485
+    },
+    {
+      "epoch": 3.045753732541339,
+      "grad_norm": 0.3761097192764282,
+      "learning_rate": 1.6240029317671658e-06,
+      "loss": 0.2103,
+      "step": 9486
+    },
+    {
+      "epoch": 3.04607481136619,
+      "grad_norm": 0.6944969892501831,
+      "learning_rate": 1.6177435754167415e-06,
+      "loss": 0.2706,
+      "step": 9487
+    },
+    {
+      "epoch": 3.0463958901910417,
+      "grad_norm": 0.43862295150756836,
+      "learning_rate": 1.611496206645835e-06,
+      "loss": 0.2505,
+      "step": 9488
+    },
+    {
+      "epoch": 3.0467169690158933,
+      "grad_norm": 0.36449775099754333,
+      "learning_rate": 1.605260826215682e-06,
+      "loss": 0.2377,
+      "step": 9489
+    },
+    {
+      "epoch": 3.047038047840745,
+      "grad_norm": 0.4346637725830078,
+      "learning_rate": 1.5990374348860305e-06,
+      "loss": 0.2281,
+      "step": 9490
+    },
+    {
+      "epoch": 3.0473591266655964,
+      "grad_norm": 0.623367190361023,
+      "learning_rate": 1.5928260334151845e-06,
+      "loss": 0.2755,
+      "step": 9491
+    },
+    {
+      "epoch": 3.047680205490448,
+      "grad_norm": 0.8898089528083801,
+      "learning_rate": 1.5866266225599834e-06,
+      "loss": 0.2915,
+      "step": 9492
+    },
+    {
+      "epoch": 3.0480012843152995,
+      "grad_norm": 0.3457051217556,
+      "learning_rate": 1.580439203075812e-06,
+      "loss": 0.2286,
+      "step": 9493
+    },
+    {
+      "epoch": 3.048322363140151,
+      "grad_norm": 0.48671117424964905,
+      "learning_rate": 1.574263775716578e-06,
+      "loss": 0.3948,
+      "step": 9494
+    },
+    {
+      "epoch": 3.0486434419650026,
+      "grad_norm": 0.5143343210220337,
+      "learning_rate": 1.5681003412347572e-06,
+      "loss": 0.4827,
+      "step": 9495
+    },
+    {
+      "epoch": 3.0489645207898537,
+      "grad_norm": 0.37674856185913086,
+      "learning_rate": 1.561948900381327e-06,
+      "loss": 0.2705,
+      "step": 9496
+    },
+    {
+      "epoch": 3.0492855996147052,
+      "grad_norm": 0.3917763829231262,
+      "learning_rate": 1.555809453905821e-06,
+      "loss": 0.2283,
+      "step": 9497
+    },
+    {
+      "epoch": 3.049606678439557,
+      "grad_norm": 0.4415786564350128,
+      "learning_rate": 1.5496820025563409e-06,
+      "loss": 0.2775,
+      "step": 9498
+    },
+    {
+      "epoch": 3.0499277572644083,
+      "grad_norm": 0.28412747383117676,
+      "learning_rate": 1.543566547079467e-06,
+      "loss": 0.134,
+      "step": 9499
+    },
+    {
+      "epoch": 3.05024883608926,
+      "grad_norm": 0.32522836327552795,
+      "learning_rate": 1.5374630882203588e-06,
+      "loss": 0.1553,
+      "step": 9500
+    },
+    {
+      "epoch": 3.0505699149141114,
+      "grad_norm": 0.37719276547431946,
+      "learning_rate": 1.5313716267226996e-06,
+      "loss": 0.1436,
+      "step": 9501
+    },
+    {
+      "epoch": 3.050890993738963,
+      "grad_norm": 0.7100079655647278,
+      "learning_rate": 1.5252921633287177e-06,
+      "loss": 0.423,
+      "step": 9502
+    },
+    {
+      "epoch": 3.0512120725638145,
+      "grad_norm": 0.6729628443717957,
+      "learning_rate": 1.5192246987791981e-06,
+      "loss": 0.2616,
+      "step": 9503
+    },
+    {
+      "epoch": 3.051533151388666,
+      "grad_norm": 0.7652589082717896,
+      "learning_rate": 1.5131692338134052e-06,
+      "loss": 0.2771,
+      "step": 9504
+    },
+    {
+      "epoch": 3.051854230213517,
+      "grad_norm": 0.6066986918449402,
+      "learning_rate": 1.5071257691692153e-06,
+      "loss": 0.2357,
+      "step": 9505
+    },
+    {
+      "epoch": 3.0521753090383688,
+      "grad_norm": 0.580855131149292,
+      "learning_rate": 1.501094305582984e-06,
+      "loss": 0.2318,
+      "step": 9506
+    },
+    {
+      "epoch": 3.0524963878632203,
+      "grad_norm": 0.5671026110649109,
+      "learning_rate": 1.4950748437896234e-06,
+      "loss": 0.211,
+      "step": 9507
+    },
+    {
+      "epoch": 3.052817466688072,
+      "grad_norm": 0.6296615600585938,
+      "learning_rate": 1.4890673845226133e-06,
+      "loss": 0.2119,
+      "step": 9508
+    },
+    {
+      "epoch": 3.0531385455129234,
+      "grad_norm": 0.6201547980308533,
+      "learning_rate": 1.4830719285139127e-06,
+      "loss": 0.2538,
+      "step": 9509
+    },
+    {
+      "epoch": 3.053459624337775,
+      "grad_norm": 0.9474941492080688,
+      "learning_rate": 1.4770884764940706e-06,
+      "loss": 0.242,
+      "step": 9510
+    },
+    {
+      "epoch": 3.0537807031626265,
+      "grad_norm": 0.6287795901298523,
+      "learning_rate": 1.4711170291921484e-06,
+      "loss": 0.2053,
+      "step": 9511
+    },
+    {
+      "epoch": 3.054101781987478,
+      "grad_norm": 0.6485587358474731,
+      "learning_rate": 1.4651575873357416e-06,
+      "loss": 0.2242,
+      "step": 9512
+    },
+    {
+      "epoch": 3.0544228608123296,
+      "grad_norm": 0.6445651054382324,
+      "learning_rate": 1.4592101516509914e-06,
+      "loss": 0.2214,
+      "step": 9513
+    },
+    {
+      "epoch": 3.0547439396371807,
+      "grad_norm": 0.6522722244262695,
+      "learning_rate": 1.4532747228625854e-06,
+      "loss": 0.2488,
+      "step": 9514
+    },
+    {
+      "epoch": 3.0550650184620323,
+      "grad_norm": 0.6885031461715698,
+      "learning_rate": 1.4473513016937223e-06,
+      "loss": 0.2459,
+      "step": 9515
+    },
+    {
+      "epoch": 3.055386097286884,
+      "grad_norm": 0.6756393909454346,
+      "learning_rate": 1.4414398888661695e-06,
+      "loss": 0.2131,
+      "step": 9516
+    },
+    {
+      "epoch": 3.0557071761117354,
+      "grad_norm": 0.5485737919807434,
+      "learning_rate": 1.4355404851001952e-06,
+      "loss": 0.2285,
+      "step": 9517
+    },
+    {
+      "epoch": 3.056028254936587,
+      "grad_norm": 0.3958723247051239,
+      "learning_rate": 1.4296530911146466e-06,
+      "loss": 0.1763,
+      "step": 9518
+    },
+    {
+      "epoch": 3.0563493337614385,
+      "grad_norm": 0.5892900824546814,
+      "learning_rate": 1.423777707626872e-06,
+      "loss": 0.1842,
+      "step": 9519
+    },
+    {
+      "epoch": 3.05667041258629,
+      "grad_norm": 0.6213316321372986,
+      "learning_rate": 1.4179143353527547e-06,
+      "loss": 0.26,
+      "step": 9520
+    },
+    {
+      "epoch": 3.0569914914111416,
+      "grad_norm": 0.6566430330276489,
+      "learning_rate": 1.412062975006767e-06,
+      "loss": 0.2532,
+      "step": 9521
+    },
+    {
+      "epoch": 3.057312570235993,
+      "grad_norm": 0.5408963561058044,
+      "learning_rate": 1.4062236273018392e-06,
+      "loss": 0.2406,
+      "step": 9522
+    },
+    {
+      "epoch": 3.0576336490608442,
+      "grad_norm": 0.552483856678009,
+      "learning_rate": 1.400396292949513e-06,
+      "loss": 0.2385,
+      "step": 9523
+    },
+    {
+      "epoch": 3.057954727885696,
+      "grad_norm": 0.6494653820991516,
+      "learning_rate": 1.394580972659798e-06,
+      "loss": 0.215,
+      "step": 9524
+    },
+    {
+      "epoch": 3.0582758067105473,
+      "grad_norm": 0.5467716455459595,
+      "learning_rate": 1.3887776671412943e-06,
+      "loss": 0.2344,
+      "step": 9525
+    },
+    {
+      "epoch": 3.058596885535399,
+      "grad_norm": 0.6003792881965637,
+      "learning_rate": 1.3829863771011253e-06,
+      "loss": 0.2197,
+      "step": 9526
+    },
+    {
+      "epoch": 3.0589179643602504,
+      "grad_norm": 0.4577672481536865,
+      "learning_rate": 1.377207103244904e-06,
+      "loss": 0.2218,
+      "step": 9527
+    },
+    {
+      "epoch": 3.059239043185102,
+      "grad_norm": 0.6770538687705994,
+      "learning_rate": 1.3714398462768563e-06,
+      "loss": 0.239,
+      "step": 9528
+    },
+    {
+      "epoch": 3.0595601220099535,
+      "grad_norm": 0.5816624164581299,
+      "learning_rate": 1.3656846068996976e-06,
+      "loss": 0.2556,
+      "step": 9529
+    },
+    {
+      "epoch": 3.059881200834805,
+      "grad_norm": 0.7855013012886047,
+      "learning_rate": 1.359941385814667e-06,
+      "loss": 0.2844,
+      "step": 9530
+    },
+    {
+      "epoch": 3.0602022796596566,
+      "grad_norm": 0.26091650128364563,
+      "learning_rate": 1.3542101837215826e-06,
+      "loss": 0.1777,
+      "step": 9531
+    },
+    {
+      "epoch": 3.0605233584845077,
+      "grad_norm": 0.46976807713508606,
+      "learning_rate": 1.3484910013187524e-06,
+      "loss": 0.2338,
+      "step": 9532
+    },
+    {
+      "epoch": 3.0608444373093593,
+      "grad_norm": 0.46279844641685486,
+      "learning_rate": 1.3427838393030633e-06,
+      "loss": 0.2305,
+      "step": 9533
+    },
+    {
+      "epoch": 3.061165516134211,
+      "grad_norm": 0.6863793730735779,
+      "learning_rate": 1.3370886983698927e-06,
+      "loss": 0.2443,
+      "step": 9534
+    },
+    {
+      "epoch": 3.0614865949590624,
+      "grad_norm": 0.6161332130432129,
+      "learning_rate": 1.3314055792131964e-06,
+      "loss": 0.2431,
+      "step": 9535
+    },
+    {
+      "epoch": 3.061807673783914,
+      "grad_norm": 0.4130920171737671,
+      "learning_rate": 1.3257344825254315e-06,
+      "loss": 0.2002,
+      "step": 9536
+    },
+    {
+      "epoch": 3.0621287526087655,
+      "grad_norm": 0.7032944560050964,
+      "learning_rate": 1.320075408997612e-06,
+      "loss": 0.261,
+      "step": 9537
+    },
+    {
+      "epoch": 3.062449831433617,
+      "grad_norm": 0.563331127166748,
+      "learning_rate": 1.3144283593192752e-06,
+      "loss": 0.2508,
+      "step": 9538
+    },
+    {
+      "epoch": 3.0627709102584686,
+      "grad_norm": 0.2759096026420593,
+      "learning_rate": 1.308793334178493e-06,
+      "loss": 0.2001,
+      "step": 9539
+    },
+    {
+      "epoch": 3.06309198908332,
+      "grad_norm": 0.3822631537914276,
+      "learning_rate": 1.303170334261883e-06,
+      "loss": 0.2032,
+      "step": 9540
+    },
+    {
+      "epoch": 3.0634130679081713,
+      "grad_norm": 0.397834837436676,
+      "learning_rate": 1.2975593602545965e-06,
+      "loss": 0.2111,
+      "step": 9541
+    },
+    {
+      "epoch": 3.063734146733023,
+      "grad_norm": 0.3753139078617096,
+      "learning_rate": 1.2919604128402874e-06,
+      "loss": 0.2241,
+      "step": 9542
+    },
+    {
+      "epoch": 3.0640552255578744,
+      "grad_norm": 0.4659748673439026,
+      "learning_rate": 1.2863734927012095e-06,
+      "loss": 0.2573,
+      "step": 9543
+    },
+    {
+      "epoch": 3.064376304382726,
+      "grad_norm": 0.35863274335861206,
+      "learning_rate": 1.280798600518085e-06,
+      "loss": 0.33,
+      "step": 9544
+    },
+    {
+      "epoch": 3.0646973832075775,
+      "grad_norm": 0.48905041813850403,
+      "learning_rate": 1.275235736970193e-06,
+      "loss": 0.5968,
+      "step": 9545
+    },
+    {
+      "epoch": 3.065018462032429,
+      "grad_norm": 0.3606785237789154,
+      "learning_rate": 1.2696849027353796e-06,
+      "loss": 0.2948,
+      "step": 9546
+    },
+    {
+      "epoch": 3.0653395408572806,
+      "grad_norm": 0.32427453994750977,
+      "learning_rate": 1.26414609848996e-06,
+      "loss": 0.1369,
+      "step": 9547
+    },
+    {
+      "epoch": 3.065660619682132,
+      "grad_norm": 0.3010817766189575,
+      "learning_rate": 1.2586193249088608e-06,
+      "loss": 0.146,
+      "step": 9548
+    },
+    {
+      "epoch": 3.0659816985069837,
+      "grad_norm": 0.3775825500488281,
+      "learning_rate": 1.2531045826654653e-06,
+      "loss": 0.1995,
+      "step": 9549
+    },
+    {
+      "epoch": 3.0663027773318348,
+      "grad_norm": 0.1418655514717102,
+      "learning_rate": 1.2476018724317585e-06,
+      "loss": 0.0591,
+      "step": 9550
+    },
+    {
+      "epoch": 3.0666238561566863,
+      "grad_norm": 0.36233529448509216,
+      "learning_rate": 1.2421111948782149e-06,
+      "loss": 0.1744,
+      "step": 9551
+    },
+    {
+      "epoch": 3.066944934981538,
+      "grad_norm": 0.6624466180801392,
+      "learning_rate": 1.236632550673844e-06,
+      "loss": 0.3614,
+      "step": 9552
+    },
+    {
+      "epoch": 3.0672660138063894,
+      "grad_norm": 0.6916109919548035,
+      "learning_rate": 1.231165940486234e-06,
+      "loss": 0.2959,
+      "step": 9553
+    },
+    {
+      "epoch": 3.067587092631241,
+      "grad_norm": 0.7930703163146973,
+      "learning_rate": 1.225711364981441e-06,
+      "loss": 0.3368,
+      "step": 9554
+    },
+    {
+      "epoch": 3.0679081714560925,
+      "grad_norm": 0.636942982673645,
+      "learning_rate": 1.2202688248241112e-06,
+      "loss": 0.2359,
+      "step": 9555
+    },
+    {
+      "epoch": 3.068229250280944,
+      "grad_norm": 0.5777634978294373,
+      "learning_rate": 1.2148383206773916e-06,
+      "loss": 0.2183,
+      "step": 9556
+    },
+    {
+      "epoch": 3.0685503291057956,
+      "grad_norm": 0.6497059464454651,
+      "learning_rate": 1.2094198532029755e-06,
+      "loss": 0.2373,
+      "step": 9557
+    },
+    {
+      "epoch": 3.068871407930647,
+      "grad_norm": 0.6847310066223145,
+      "learning_rate": 1.20401342306109e-06,
+      "loss": 0.2691,
+      "step": 9558
+    },
+    {
+      "epoch": 3.0691924867554983,
+      "grad_norm": 0.6610705256462097,
+      "learning_rate": 1.1986190309104861e-06,
+      "loss": 0.2512,
+      "step": 9559
+    },
+    {
+      "epoch": 3.06951356558035,
+      "grad_norm": 0.5450610518455505,
+      "learning_rate": 1.193236677408449e-06,
+      "loss": 0.1912,
+      "step": 9560
+    },
+    {
+      "epoch": 3.0698346444052014,
+      "grad_norm": 0.5279379487037659,
+      "learning_rate": 1.1878663632108322e-06,
+      "loss": 0.1994,
+      "step": 9561
+    },
+    {
+      "epoch": 3.070155723230053,
+      "grad_norm": 0.5520129799842834,
+      "learning_rate": 1.1825080889719563e-06,
+      "loss": 0.2005,
+      "step": 9562
+    },
+    {
+      "epoch": 3.0704768020549045,
+      "grad_norm": 0.5607929825782776,
+      "learning_rate": 1.1771618553447216e-06,
+      "loss": 0.1891,
+      "step": 9563
+    },
+    {
+      "epoch": 3.070797880879756,
+      "grad_norm": 0.9774504899978638,
+      "learning_rate": 1.1718276629805625e-06,
+      "loss": 0.3626,
+      "step": 9564
+    },
+    {
+      "epoch": 3.0711189597046076,
+      "grad_norm": 0.7361037135124207,
+      "learning_rate": 1.1665055125294033e-06,
+      "loss": 0.2507,
+      "step": 9565
+    },
+    {
+      "epoch": 3.071440038529459,
+      "grad_norm": 0.8793400526046753,
+      "learning_rate": 1.16119540463977e-06,
+      "loss": 0.2958,
+      "step": 9566
+    },
+    {
+      "epoch": 3.0717611173543107,
+      "grad_norm": 0.8101871609687805,
+      "learning_rate": 1.155897339958667e-06,
+      "loss": 0.2885,
+      "step": 9567
+    },
+    {
+      "epoch": 3.072082196179162,
+      "grad_norm": 1.022641658782959,
+      "learning_rate": 1.1506113191316447e-06,
+      "loss": 0.4274,
+      "step": 9568
+    },
+    {
+      "epoch": 3.0724032750040133,
+      "grad_norm": 0.7252495288848877,
+      "learning_rate": 1.1453373428027992e-06,
+      "loss": 0.2599,
+      "step": 9569
+    },
+    {
+      "epoch": 3.072724353828865,
+      "grad_norm": 0.4876370429992676,
+      "learning_rate": 1.1400754116147271e-06,
+      "loss": 0.2188,
+      "step": 9570
+    },
+    {
+      "epoch": 3.0730454326537164,
+      "grad_norm": 0.6673651933670044,
+      "learning_rate": 1.134825526208605e-06,
+      "loss": 0.3235,
+      "step": 9571
+    },
+    {
+      "epoch": 3.073366511478568,
+      "grad_norm": 0.7081831693649292,
+      "learning_rate": 1.1295876872240873e-06,
+      "loss": 0.2824,
+      "step": 9572
+    },
+    {
+      "epoch": 3.0736875903034195,
+      "grad_norm": 0.8749157786369324,
+      "learning_rate": 1.1243618952994195e-06,
+      "loss": 0.3039,
+      "step": 9573
+    },
+    {
+      "epoch": 3.074008669128271,
+      "grad_norm": 0.5778833031654358,
+      "learning_rate": 1.1191481510713253e-06,
+      "loss": 0.2355,
+      "step": 9574
+    },
+    {
+      "epoch": 3.0743297479531226,
+      "grad_norm": 0.482704758644104,
+      "learning_rate": 1.1139464551750856e-06,
+      "loss": 0.2225,
+      "step": 9575
+    },
+    {
+      "epoch": 3.074650826777974,
+      "grad_norm": 0.5122305750846863,
+      "learning_rate": 1.1087568082445264e-06,
+      "loss": 0.2251,
+      "step": 9576
+    },
+    {
+      "epoch": 3.0749719056028253,
+      "grad_norm": 0.6916018128395081,
+      "learning_rate": 1.103579210911976e-06,
+      "loss": 0.2501,
+      "step": 9577
+    },
+    {
+      "epoch": 3.075292984427677,
+      "grad_norm": 0.5738651156425476,
+      "learning_rate": 1.0984136638083177e-06,
+      "loss": 0.2508,
+      "step": 9578
+    },
+    {
+      "epoch": 3.0756140632525284,
+      "grad_norm": 0.3557813763618469,
+      "learning_rate": 1.0932601675629595e-06,
+      "loss": 0.1737,
+      "step": 9579
+    },
+    {
+      "epoch": 3.07593514207738,
+      "grad_norm": 0.7535400986671448,
+      "learning_rate": 1.0881187228038215e-06,
+      "loss": 0.2763,
+      "step": 9580
+    },
+    {
+      "epoch": 3.0762562209022315,
+      "grad_norm": 0.4553007185459137,
+      "learning_rate": 1.0829893301573913e-06,
+      "loss": 0.226,
+      "step": 9581
+    },
+    {
+      "epoch": 3.076577299727083,
+      "grad_norm": 0.3509460389614105,
+      "learning_rate": 1.0778719902486689e-06,
+      "loss": 0.1899,
+      "step": 9582
+    },
+    {
+      "epoch": 3.0768983785519346,
+      "grad_norm": 0.7953051924705505,
+      "learning_rate": 1.0727667037011668e-06,
+      "loss": 0.3081,
+      "step": 9583
+    },
+    {
+      "epoch": 3.077219457376786,
+      "grad_norm": 0.563300609588623,
+      "learning_rate": 1.0676734711369762e-06,
+      "loss": 0.2307,
+      "step": 9584
+    },
+    {
+      "epoch": 3.0775405362016377,
+      "grad_norm": 0.5179484486579895,
+      "learning_rate": 1.0625922931766785e-06,
+      "loss": 0.2405,
+      "step": 9585
+    },
+    {
+      "epoch": 3.077861615026489,
+      "grad_norm": 0.8609539270401001,
+      "learning_rate": 1.0575231704393895e-06,
+      "loss": 0.2623,
+      "step": 9586
+    },
+    {
+      "epoch": 3.0781826938513404,
+      "grad_norm": 0.571172833442688,
+      "learning_rate": 1.052466103542793e-06,
+      "loss": 0.2477,
+      "step": 9587
+    },
+    {
+      "epoch": 3.078503772676192,
+      "grad_norm": 0.396445095539093,
+      "learning_rate": 1.0474210931030514e-06,
+      "loss": 0.2204,
+      "step": 9588
+    },
+    {
+      "epoch": 3.0788248515010435,
+      "grad_norm": 0.43316546082496643,
+      "learning_rate": 1.0423881397349068e-06,
+      "loss": 0.2267,
+      "step": 9589
+    },
+    {
+      "epoch": 3.079145930325895,
+      "grad_norm": 0.4050913453102112,
+      "learning_rate": 1.0373672440515902e-06,
+      "loss": 0.221,
+      "step": 9590
+    },
+    {
+      "epoch": 3.0794670091507466,
+      "grad_norm": 0.2422318160533905,
+      "learning_rate": 1.0323584066648795e-06,
+      "loss": 0.2087,
+      "step": 9591
+    },
+    {
+      "epoch": 3.079788087975598,
+      "grad_norm": 0.34273797273635864,
+      "learning_rate": 1.0273616281851083e-06,
+      "loss": 0.2479,
+      "step": 9592
+    },
+    {
+      "epoch": 3.0801091668004497,
+      "grad_norm": 0.35393643379211426,
+      "learning_rate": 1.0223769092211012e-06,
+      "loss": 0.2539,
+      "step": 9593
+    },
+    {
+      "epoch": 3.080430245625301,
+      "grad_norm": 0.4257902503013611,
+      "learning_rate": 1.0174042503802493e-06,
+      "loss": 0.4868,
+      "step": 9594
+    },
+    {
+      "epoch": 3.0807513244501523,
+      "grad_norm": 0.4722660779953003,
+      "learning_rate": 1.0124436522684243e-06,
+      "loss": 0.4076,
+      "step": 9595
+    },
+    {
+      "epoch": 3.081072403275004,
+      "grad_norm": 0.2358543872833252,
+      "learning_rate": 1.007495115490087e-06,
+      "loss": 0.0826,
+      "step": 9596
+    },
+    {
+      "epoch": 3.0813934820998554,
+      "grad_norm": 0.2644473612308502,
+      "learning_rate": 1.002558640648199e-06,
+      "loss": 0.102,
+      "step": 9597
+    },
+    {
+      "epoch": 3.081714560924707,
+      "grad_norm": 0.3907872438430786,
+      "learning_rate": 9.976342283442463e-07,
+      "loss": 0.1718,
+      "step": 9598
+    },
+    {
+      "epoch": 3.0820356397495585,
+      "grad_norm": 0.35824838280677795,
+      "learning_rate": 9.927218791782599e-07,
+      "loss": 0.1511,
+      "step": 9599
+    },
+    {
+      "epoch": 3.08235671857441,
+      "grad_norm": 0.6002468466758728,
+      "learning_rate": 9.878215937487834e-07,
+      "loss": 0.3288,
+      "step": 9600
+    },
+    {
+      "epoch": 3.0826777973992616,
+      "grad_norm": 0.827461302280426,
+      "learning_rate": 9.829333726529056e-07,
+      "loss": 0.3523,
+      "step": 9601
+    },
+    {
+      "epoch": 3.082998876224113,
+      "grad_norm": 0.5873133540153503,
+      "learning_rate": 9.78057216486261e-07,
+      "loss": 0.2328,
+      "step": 9602
+    },
+    {
+      "epoch": 3.0833199550489647,
+      "grad_norm": 0.5725942254066467,
+      "learning_rate": 9.731931258429638e-07,
+      "loss": 0.2127,
+      "step": 9603
+    },
+    {
+      "epoch": 3.083641033873816,
+      "grad_norm": 0.584507405757904,
+      "learning_rate": 9.683411013157174e-07,
+      "loss": 0.1985,
+      "step": 9604
+    },
+    {
+      "epoch": 3.0839621126986674,
+      "grad_norm": 0.7739067077636719,
+      "learning_rate": 9.635011434957152e-07,
+      "loss": 0.277,
+      "step": 9605
+    },
+    {
+      "epoch": 3.084283191523519,
+      "grad_norm": 0.5025599598884583,
+      "learning_rate": 9.58673252972675e-07,
+      "loss": 0.1996,
+      "step": 9606
+    },
+    {
+      "epoch": 3.0846042703483705,
+      "grad_norm": 0.6031046509742737,
+      "learning_rate": 9.538574303348813e-07,
+      "loss": 0.2324,
+      "step": 9607
+    },
+    {
+      "epoch": 3.084925349173222,
+      "grad_norm": 0.8453519344329834,
+      "learning_rate": 9.490536761691204e-07,
+      "loss": 0.2988,
+      "step": 9608
+    },
+    {
+      "epoch": 3.0852464279980736,
+      "grad_norm": 0.5942166447639465,
+      "learning_rate": 9.442619910607131e-07,
+      "loss": 0.2195,
+      "step": 9609
+    },
+    {
+      "epoch": 3.085567506822925,
+      "grad_norm": 0.5165850520133972,
+      "learning_rate": 9.394823755935145e-07,
+      "loss": 0.2168,
+      "step": 9610
+    },
+    {
+      "epoch": 3.0858885856477767,
+      "grad_norm": 0.4023417532444,
+      "learning_rate": 9.347148303499142e-07,
+      "loss": 0.1666,
+      "step": 9611
+    },
+    {
+      "epoch": 3.0862096644726282,
+      "grad_norm": 0.44202151894569397,
+      "learning_rate": 9.299593559108033e-07,
+      "loss": 0.1773,
+      "step": 9612
+    },
+    {
+      "epoch": 3.0865307432974793,
+      "grad_norm": 0.6434639096260071,
+      "learning_rate": 9.252159528556403e-07,
+      "loss": 0.2229,
+      "step": 9613
+    },
+    {
+      "epoch": 3.086851822122331,
+      "grad_norm": 0.5662235021591187,
+      "learning_rate": 9.204846217623853e-07,
+      "loss": 0.2579,
+      "step": 9614
+    },
+    {
+      "epoch": 3.0871729009471824,
+      "grad_norm": 0.7555933594703674,
+      "learning_rate": 9.157653632075436e-07,
+      "loss": 0.2733,
+      "step": 9615
+    },
+    {
+      "epoch": 3.087493979772034,
+      "grad_norm": 0.5656288862228394,
+      "learning_rate": 9.110581777661331e-07,
+      "loss": 0.2362,
+      "step": 9616
+    },
+    {
+      "epoch": 3.0878150585968855,
+      "grad_norm": 0.6788967251777649,
+      "learning_rate": 9.063630660117173e-07,
+      "loss": 0.2549,
+      "step": 9617
+    },
+    {
+      "epoch": 3.088136137421737,
+      "grad_norm": 0.45415976643562317,
+      "learning_rate": 9.016800285163718e-07,
+      "loss": 0.1989,
+      "step": 9618
+    },
+    {
+      "epoch": 3.0884572162465886,
+      "grad_norm": 0.4443502128124237,
+      "learning_rate": 8.970090658507291e-07,
+      "loss": 0.1886,
+      "step": 9619
+    },
+    {
+      "epoch": 3.08877829507144,
+      "grad_norm": 0.5720863938331604,
+      "learning_rate": 8.923501785839117e-07,
+      "loss": 0.2065,
+      "step": 9620
+    },
+    {
+      "epoch": 3.0890993738962917,
+      "grad_norm": 0.7446249723434448,
+      "learning_rate": 8.877033672835988e-07,
+      "loss": 0.3329,
+      "step": 9621
+    },
+    {
+      "epoch": 3.089420452721143,
+      "grad_norm": 0.7151133418083191,
+      "learning_rate": 8.830686325160042e-07,
+      "loss": 0.3169,
+      "step": 9622
+    },
+    {
+      "epoch": 3.0897415315459944,
+      "grad_norm": 0.5628482699394226,
+      "learning_rate": 8.784459748458318e-07,
+      "loss": 0.2448,
+      "step": 9623
+    },
+    {
+      "epoch": 3.090062610370846,
+      "grad_norm": 0.7645835280418396,
+      "learning_rate": 8.73835394836342e-07,
+      "loss": 0.2118,
+      "step": 9624
+    },
+    {
+      "epoch": 3.0903836891956975,
+      "grad_norm": 0.9074884653091431,
+      "learning_rate": 8.692368930493521e-07,
+      "loss": 0.2708,
+      "step": 9625
+    },
+    {
+      "epoch": 3.090704768020549,
+      "grad_norm": 0.8181541562080383,
+      "learning_rate": 8.646504700451252e-07,
+      "loss": 0.2557,
+      "step": 9626
+    },
+    {
+      "epoch": 3.0910258468454006,
+      "grad_norm": 0.39952725172042847,
+      "learning_rate": 8.600761263825474e-07,
+      "loss": 0.1937,
+      "step": 9627
+    },
+    {
+      "epoch": 3.091346925670252,
+      "grad_norm": 0.505699634552002,
+      "learning_rate": 8.555138626189618e-07,
+      "loss": 0.2237,
+      "step": 9628
+    },
+    {
+      "epoch": 3.0916680044951037,
+      "grad_norm": 0.8317124843597412,
+      "learning_rate": 8.509636793102682e-07,
+      "loss": 0.316,
+      "step": 9629
+    },
+    {
+      "epoch": 3.0919890833199553,
+      "grad_norm": 0.4344399571418762,
+      "learning_rate": 8.46425577010912e-07,
+      "loss": 0.1989,
+      "step": 9630
+    },
+    {
+      "epoch": 3.0923101621448064,
+      "grad_norm": 0.7013100385665894,
+      "learning_rate": 8.418995562738285e-07,
+      "loss": 0.2483,
+      "step": 9631
+    },
+    {
+      "epoch": 3.092631240969658,
+      "grad_norm": 0.40041857957839966,
+      "learning_rate": 8.373856176505101e-07,
+      "loss": 0.2176,
+      "step": 9632
+    },
+    {
+      "epoch": 3.0929523197945095,
+      "grad_norm": 0.4556305408477783,
+      "learning_rate": 8.328837616909613e-07,
+      "loss": 0.2162,
+      "step": 9633
+    },
+    {
+      "epoch": 3.093273398619361,
+      "grad_norm": 2.067694664001465,
+      "learning_rate": 8.283939889437209e-07,
+      "loss": 0.2219,
+      "step": 9634
+    },
+    {
+      "epoch": 3.0935944774442126,
+      "grad_norm": 0.48048630356788635,
+      "learning_rate": 8.239162999558403e-07,
+      "loss": 0.2146,
+      "step": 9635
+    },
+    {
+      "epoch": 3.093915556269064,
+      "grad_norm": 0.8311915993690491,
+      "learning_rate": 8.194506952729386e-07,
+      "loss": 0.3099,
+      "step": 9636
+    },
+    {
+      "epoch": 3.0942366350939157,
+      "grad_norm": 0.5806686282157898,
+      "learning_rate": 8.14997175439125e-07,
+      "loss": 0.2564,
+      "step": 9637
+    },
+    {
+      "epoch": 3.094557713918767,
+      "grad_norm": 0.674216628074646,
+      "learning_rate": 8.105557409970432e-07,
+      "loss": 0.254,
+      "step": 9638
+    },
+    {
+      "epoch": 3.0948787927436188,
+      "grad_norm": 0.826112687587738,
+      "learning_rate": 8.061263924878604e-07,
+      "loss": 0.2399,
+      "step": 9639
+    },
+    {
+      "epoch": 3.09519987156847,
+      "grad_norm": 0.5507524013519287,
+      "learning_rate": 8.017091304513003e-07,
+      "loss": 0.257,
+      "step": 9640
+    },
+    {
+      "epoch": 3.0955209503933214,
+      "grad_norm": 0.3682684302330017,
+      "learning_rate": 7.973039554255768e-07,
+      "loss": 0.216,
+      "step": 9641
+    },
+    {
+      "epoch": 3.095842029218173,
+      "grad_norm": 0.49590498208999634,
+      "learning_rate": 7.929108679474607e-07,
+      "loss": 0.2547,
+      "step": 9642
+    },
+    {
+      "epoch": 3.0961631080430245,
+      "grad_norm": 0.3100004494190216,
+      "learning_rate": 7.885298685522235e-07,
+      "loss": 0.2364,
+      "step": 9643
+    },
+    {
+      "epoch": 3.096484186867876,
+      "grad_norm": 0.41891974210739136,
+      "learning_rate": 7.841609577736719e-07,
+      "loss": 0.404,
+      "step": 9644
+    },
+    {
+      "epoch": 3.0968052656927276,
+      "grad_norm": 0.5358453989028931,
+      "learning_rate": 7.798041361441688e-07,
+      "loss": 0.6396,
+      "step": 9645
+    },
+    {
+      "epoch": 3.097126344517579,
+      "grad_norm": 0.5173894166946411,
+      "learning_rate": 7.754594041945562e-07,
+      "loss": 0.5229,
+      "step": 9646
+    },
+    {
+      "epoch": 3.0974474233424307,
+      "grad_norm": 0.4027728736400604,
+      "learning_rate": 7.711267624542329e-07,
+      "loss": 0.2377,
+      "step": 9647
+    },
+    {
+      "epoch": 3.0977685021672823,
+      "grad_norm": 0.3841489851474762,
+      "learning_rate": 7.668062114511321e-07,
+      "loss": 0.214,
+      "step": 9648
+    },
+    {
+      "epoch": 3.0980895809921334,
+      "grad_norm": 0.29058584570884705,
+      "learning_rate": 7.624977517116772e-07,
+      "loss": 0.1148,
+      "step": 9649
+    },
+    {
+      "epoch": 3.098410659816985,
+      "grad_norm": 0.2613859474658966,
+      "learning_rate": 7.582013837608592e-07,
+      "loss": 0.0923,
+      "step": 9650
+    },
+    {
+      "epoch": 3.0987317386418365,
+      "grad_norm": 0.5930027365684509,
+      "learning_rate": 7.539171081221596e-07,
+      "loss": 0.2433,
+      "step": 9651
+    },
+    {
+      "epoch": 3.099052817466688,
+      "grad_norm": 0.6712236404418945,
+      "learning_rate": 7.496449253176274e-07,
+      "loss": 0.2822,
+      "step": 9652
+    },
+    {
+      "epoch": 3.0993738962915396,
+      "grad_norm": 0.6282926797866821,
+      "learning_rate": 7.453848358678017e-07,
+      "loss": 0.2422,
+      "step": 9653
+    },
+    {
+      "epoch": 3.099694975116391,
+      "grad_norm": 0.8251811265945435,
+      "learning_rate": 7.411368402917563e-07,
+      "loss": 0.2839,
+      "step": 9654
+    },
+    {
+      "epoch": 3.1000160539412427,
+      "grad_norm": 0.5989024043083191,
+      "learning_rate": 7.369009391070992e-07,
+      "loss": 0.2299,
+      "step": 9655
+    },
+    {
+      "epoch": 3.1003371327660942,
+      "grad_norm": 0.42122694849967957,
+      "learning_rate": 7.326771328299731e-07,
+      "loss": 0.1556,
+      "step": 9656
+    },
+    {
+      "epoch": 3.100658211590946,
+      "grad_norm": 0.7038974761962891,
+      "learning_rate": 7.284654219750331e-07,
+      "loss": 0.2316,
+      "step": 9657
+    },
+    {
+      "epoch": 3.100979290415797,
+      "grad_norm": 0.5096376538276672,
+      "learning_rate": 7.242658070554464e-07,
+      "loss": 0.2069,
+      "step": 9658
+    },
+    {
+      "epoch": 3.1013003692406484,
+      "grad_norm": 0.8389344811439514,
+      "learning_rate": 7.200782885829482e-07,
+      "loss": 0.2828,
+      "step": 9659
+    },
+    {
+      "epoch": 3.1016214480655,
+      "grad_norm": 0.551834762096405,
+      "learning_rate": 7.159028670677526e-07,
+      "loss": 0.1995,
+      "step": 9660
+    },
+    {
+      "epoch": 3.1019425268903515,
+      "grad_norm": 0.7695516347885132,
+      "learning_rate": 7.117395430186414e-07,
+      "loss": 0.2759,
+      "step": 9661
+    },
+    {
+      "epoch": 3.102263605715203,
+      "grad_norm": 0.6411928534507751,
+      "learning_rate": 7.075883169428754e-07,
+      "loss": 0.2384,
+      "step": 9662
+    },
+    {
+      "epoch": 3.1025846845400546,
+      "grad_norm": 0.8904174566268921,
+      "learning_rate": 7.034491893463058e-07,
+      "loss": 0.286,
+      "step": 9663
+    },
+    {
+      "epoch": 3.102905763364906,
+      "grad_norm": 0.7250691056251526,
+      "learning_rate": 6.9932216073324e-07,
+      "loss": 0.2928,
+      "step": 9664
+    },
+    {
+      "epoch": 3.1032268421897578,
+      "grad_norm": 0.5427330136299133,
+      "learning_rate": 6.952072316065761e-07,
+      "loss": 0.2105,
+      "step": 9665
+    },
+    {
+      "epoch": 3.1035479210146093,
+      "grad_norm": 0.7631176710128784,
+      "learning_rate": 6.911044024676683e-07,
+      "loss": 0.2677,
+      "step": 9666
+    },
+    {
+      "epoch": 3.1038689998394604,
+      "grad_norm": 0.6781275272369385,
+      "learning_rate": 6.870136738164612e-07,
+      "loss": 0.2665,
+      "step": 9667
+    },
+    {
+      "epoch": 3.104190078664312,
+      "grad_norm": 0.7830384969711304,
+      "learning_rate": 6.829350461514006e-07,
+      "loss": 0.2469,
+      "step": 9668
+    },
+    {
+      "epoch": 3.1045111574891635,
+      "grad_norm": 0.836956799030304,
+      "learning_rate": 6.788685199694222e-07,
+      "loss": 0.2681,
+      "step": 9669
+    },
+    {
+      "epoch": 3.104832236314015,
+      "grad_norm": 0.6083877086639404,
+      "learning_rate": 6.748140957660631e-07,
+      "loss": 0.2515,
+      "step": 9670
+    },
+    {
+      "epoch": 3.1051533151388666,
+      "grad_norm": 0.5534031987190247,
+      "learning_rate": 6.707717740353059e-07,
+      "loss": 0.2078,
+      "step": 9671
+    },
+    {
+      "epoch": 3.105474393963718,
+      "grad_norm": 0.42929285764694214,
+      "learning_rate": 6.66741555269712e-07,
+      "loss": 0.2037,
+      "step": 9672
+    },
+    {
+      "epoch": 3.1057954727885697,
+      "grad_norm": 0.559162437915802,
+      "learning_rate": 6.627234399603555e-07,
+      "loss": 0.1936,
+      "step": 9673
+    },
+    {
+      "epoch": 3.1061165516134213,
+      "grad_norm": 0.687892735004425,
+      "learning_rate": 6.587174285968223e-07,
+      "loss": 0.2577,
+      "step": 9674
+    },
+    {
+      "epoch": 3.106437630438273,
+      "grad_norm": 0.499457448720932,
+      "learning_rate": 6.547235216672443e-07,
+      "loss": 0.211,
+      "step": 9675
+    },
+    {
+      "epoch": 3.106758709263124,
+      "grad_norm": 0.4725257456302643,
+      "learning_rate": 6.507417196582544e-07,
+      "loss": 0.2347,
+      "step": 9676
+    },
+    {
+      "epoch": 3.1070797880879755,
+      "grad_norm": 0.48947420716285706,
+      "learning_rate": 6.4677202305502e-07,
+      "loss": 0.1987,
+      "step": 9677
+    },
+    {
+      "epoch": 3.107400866912827,
+      "grad_norm": 0.6859459280967712,
+      "learning_rate": 6.428144323412544e-07,
+      "loss": 0.2539,
+      "step": 9678
+    },
+    {
+      "epoch": 3.1077219457376786,
+      "grad_norm": 0.5311869978904724,
+      "learning_rate": 6.388689479991605e-07,
+      "loss": 0.226,
+      "step": 9679
+    },
+    {
+      "epoch": 3.10804302456253,
+      "grad_norm": 0.668393611907959,
+      "learning_rate": 6.349355705094984e-07,
+      "loss": 0.2531,
+      "step": 9680
+    },
+    {
+      "epoch": 3.1083641033873817,
+      "grad_norm": 0.5925018787384033,
+      "learning_rate": 6.310143003515179e-07,
+      "loss": 0.2251,
+      "step": 9681
+    },
+    {
+      "epoch": 3.1086851822122332,
+      "grad_norm": 0.76288902759552,
+      "learning_rate": 6.271051380030368e-07,
+      "loss": 0.2765,
+      "step": 9682
+    },
+    {
+      "epoch": 3.1090062610370848,
+      "grad_norm": 0.8070435523986816,
+      "learning_rate": 6.232080839403631e-07,
+      "loss": 0.3033,
+      "step": 9683
+    },
+    {
+      "epoch": 3.1093273398619363,
+      "grad_norm": 0.6173917055130005,
+      "learning_rate": 6.193231386383391e-07,
+      "loss": 0.2695,
+      "step": 9684
+    },
+    {
+      "epoch": 3.1096484186867874,
+      "grad_norm": 0.8195832967758179,
+      "learning_rate": 6.154503025703417e-07,
+      "loss": 0.2724,
+      "step": 9685
+    },
+    {
+      "epoch": 3.109969497511639,
+      "grad_norm": 0.433586061000824,
+      "learning_rate": 6.115895762082602e-07,
+      "loss": 0.2257,
+      "step": 9686
+    },
+    {
+      "epoch": 3.1102905763364905,
+      "grad_norm": 0.4982379972934723,
+      "learning_rate": 6.07740960022507e-07,
+      "loss": 0.2394,
+      "step": 9687
+    },
+    {
+      "epoch": 3.110611655161342,
+      "grad_norm": 0.7917426824569702,
+      "learning_rate": 6.039044544820404e-07,
+      "loss": 0.2325,
+      "step": 9688
+    },
+    {
+      "epoch": 3.1109327339861936,
+      "grad_norm": 0.38256534934043884,
+      "learning_rate": 6.000800600542977e-07,
+      "loss": 0.2142,
+      "step": 9689
+    },
+    {
+      "epoch": 3.111253812811045,
+      "grad_norm": 0.7258543372154236,
+      "learning_rate": 5.96267777205295e-07,
+      "loss": 0.246,
+      "step": 9690
+    },
+    {
+      "epoch": 3.1115748916358967,
+      "grad_norm": 0.38089191913604736,
+      "learning_rate": 5.924676063995382e-07,
+      "loss": 0.2342,
+      "step": 9691
+    },
+    {
+      "epoch": 3.1118959704607483,
+      "grad_norm": 0.4571894407272339,
+      "learning_rate": 5.886795481000795e-07,
+      "loss": 0.2346,
+      "step": 9692
+    },
+    {
+      "epoch": 3.1122170492856,
+      "grad_norm": 0.3728485405445099,
+      "learning_rate": 5.849036027684606e-07,
+      "loss": 0.2318,
+      "step": 9693
+    },
+    {
+      "epoch": 3.112538128110451,
+      "grad_norm": 0.40570032596588135,
+      "learning_rate": 5.811397708647803e-07,
+      "loss": 0.3118,
+      "step": 9694
+    },
+    {
+      "epoch": 3.1128592069353025,
+      "grad_norm": 0.4947049617767334,
+      "learning_rate": 5.773880528476494e-07,
+      "loss": 0.5981,
+      "step": 9695
+    },
+    {
+      "epoch": 3.113180285760154,
+      "grad_norm": 0.4747079014778137,
+      "learning_rate": 5.736484491742134e-07,
+      "loss": 0.5451,
+      "step": 9696
+    },
+    {
+      "epoch": 3.1135013645850056,
+      "grad_norm": 0.23416991531848907,
+      "learning_rate": 5.699209603001076e-07,
+      "loss": 0.0842,
+      "step": 9697
+    },
+    {
+      "epoch": 3.113822443409857,
+      "grad_norm": 0.40521523356437683,
+      "learning_rate": 5.662055866795357e-07,
+      "loss": 0.2264,
+      "step": 9698
+    },
+    {
+      "epoch": 3.1141435222347087,
+      "grad_norm": 0.13257575035095215,
+      "learning_rate": 5.62502328765202e-07,
+      "loss": 0.055,
+      "step": 9699
+    },
+    {
+      "epoch": 3.1144646010595602,
+      "grad_norm": 0.25109007954597473,
+      "learning_rate": 5.588111870083346e-07,
+      "loss": 0.1009,
+      "step": 9700
+    },
+    {
+      "epoch": 3.114785679884412,
+      "grad_norm": 0.286777526140213,
+      "learning_rate": 5.551321618586736e-07,
+      "loss": 0.1089,
+      "step": 9701
+    },
+    {
+      "epoch": 3.1151067587092633,
+      "grad_norm": 0.3400633633136749,
+      "learning_rate": 5.514652537645271e-07,
+      "loss": 0.1317,
+      "step": 9702
+    },
+    {
+      "epoch": 3.1154278375341145,
+      "grad_norm": 0.6121630072593689,
+      "learning_rate": 5.478104631726711e-07,
+      "loss": 0.2507,
+      "step": 9703
+    },
+    {
+      "epoch": 3.115748916358966,
+      "grad_norm": 0.6167968511581421,
+      "learning_rate": 5.441677905284381e-07,
+      "loss": 0.2289,
+      "step": 9704
+    },
+    {
+      "epoch": 3.1160699951838176,
+      "grad_norm": 0.735968828201294,
+      "learning_rate": 5.405372362756734e-07,
+      "loss": 0.2501,
+      "step": 9705
+    },
+    {
+      "epoch": 3.116391074008669,
+      "grad_norm": 0.7218773365020752,
+      "learning_rate": 5.369188008567672e-07,
+      "loss": 0.2835,
+      "step": 9706
+    },
+    {
+      "epoch": 3.1167121528335207,
+      "grad_norm": 0.6471768617630005,
+      "learning_rate": 5.333124847125892e-07,
+      "loss": 0.1762,
+      "step": 9707
+    },
+    {
+      "epoch": 3.117033231658372,
+      "grad_norm": 0.7177468538284302,
+      "learning_rate": 5.297182882825879e-07,
+      "loss": 0.2868,
+      "step": 9708
+    },
+    {
+      "epoch": 3.1173543104832238,
+      "grad_norm": 0.540264904499054,
+      "learning_rate": 5.261362120046686e-07,
+      "loss": 0.2281,
+      "step": 9709
+    },
+    {
+      "epoch": 3.1176753893080753,
+      "grad_norm": 0.5251516103744507,
+      "learning_rate": 5.225662563153266e-07,
+      "loss": 0.1886,
+      "step": 9710
+    },
+    {
+      "epoch": 3.117996468132927,
+      "grad_norm": 0.5654820799827576,
+      "learning_rate": 5.190084216495361e-07,
+      "loss": 0.2386,
+      "step": 9711
+    },
+    {
+      "epoch": 3.118317546957778,
+      "grad_norm": 0.5759437680244446,
+      "learning_rate": 5.154627084408059e-07,
+      "loss": 0.1868,
+      "step": 9712
+    },
+    {
+      "epoch": 3.1186386257826295,
+      "grad_norm": 0.6524972319602966,
+      "learning_rate": 5.119291171211793e-07,
+      "loss": 0.2155,
+      "step": 9713
+    },
+    {
+      "epoch": 3.118959704607481,
+      "grad_norm": 0.5860231518745422,
+      "learning_rate": 5.084076481212119e-07,
+      "loss": 0.2216,
+      "step": 9714
+    },
+    {
+      "epoch": 3.1192807834323326,
+      "grad_norm": 0.7092858552932739,
+      "learning_rate": 5.048983018699827e-07,
+      "loss": 0.2829,
+      "step": 9715
+    },
+    {
+      "epoch": 3.119601862257184,
+      "grad_norm": 0.585176944732666,
+      "learning_rate": 5.01401078795094e-07,
+      "loss": 0.196,
+      "step": 9716
+    },
+    {
+      "epoch": 3.1199229410820357,
+      "grad_norm": 0.7186732888221741,
+      "learning_rate": 4.979159793226718e-07,
+      "loss": 0.2327,
+      "step": 9717
+    },
+    {
+      "epoch": 3.1202440199068873,
+      "grad_norm": 0.9595154523849487,
+      "learning_rate": 4.944430038773762e-07,
+      "loss": 0.2656,
+      "step": 9718
+    },
+    {
+      "epoch": 3.120565098731739,
+      "grad_norm": 0.4469728171825409,
+      "learning_rate": 4.909821528823577e-07,
+      "loss": 0.1912,
+      "step": 9719
+    },
+    {
+      "epoch": 3.1208861775565904,
+      "grad_norm": 0.7011276483535767,
+      "learning_rate": 4.875334267593234e-07,
+      "loss": 0.2695,
+      "step": 9720
+    },
+    {
+      "epoch": 3.1212072563814415,
+      "grad_norm": 0.49139729142189026,
+      "learning_rate": 4.840968259284817e-07,
+      "loss": 0.1871,
+      "step": 9721
+    },
+    {
+      "epoch": 3.121528335206293,
+      "grad_norm": 0.7045285105705261,
+      "learning_rate": 4.806723508085864e-07,
+      "loss": 0.241,
+      "step": 9722
+    },
+    {
+      "epoch": 3.1218494140311446,
+      "grad_norm": 0.6836577653884888,
+      "learning_rate": 4.772600018168816e-07,
+      "loss": 0.2571,
+      "step": 9723
+    },
+    {
+      "epoch": 3.122170492855996,
+      "grad_norm": 0.9169760346412659,
+      "learning_rate": 4.738597793691679e-07,
+      "loss": 0.3147,
+      "step": 9724
+    },
+    {
+      "epoch": 3.1224915716808477,
+      "grad_norm": 0.5703161954879761,
+      "learning_rate": 4.704716838797363e-07,
+      "loss": 0.2181,
+      "step": 9725
+    },
+    {
+      "epoch": 3.1228126505056992,
+      "grad_norm": 0.8328591585159302,
+      "learning_rate": 4.670957157614453e-07,
+      "loss": 0.2938,
+      "step": 9726
+    },
+    {
+      "epoch": 3.123133729330551,
+      "grad_norm": 0.545899510383606,
+      "learning_rate": 4.6373187542561035e-07,
+      "loss": 0.2104,
+      "step": 9727
+    },
+    {
+      "epoch": 3.1234548081554023,
+      "grad_norm": 0.6301508545875549,
+      "learning_rate": 4.6038016328211476e-07,
+      "loss": 0.2477,
+      "step": 9728
+    },
+    {
+      "epoch": 3.123775886980254,
+      "grad_norm": 0.7413309216499329,
+      "learning_rate": 4.570405797393762e-07,
+      "loss": 0.2591,
+      "step": 9729
+    },
+    {
+      "epoch": 3.124096965805105,
+      "grad_norm": 0.5572231411933899,
+      "learning_rate": 4.5371312520429144e-07,
+      "loss": 0.2159,
+      "step": 9730
+    },
+    {
+      "epoch": 3.1244180446299565,
+      "grad_norm": 0.4630489647388458,
+      "learning_rate": 4.503978000823028e-07,
+      "loss": 0.1875,
+      "step": 9731
+    },
+    {
+      "epoch": 3.124739123454808,
+      "grad_norm": 0.5202605128288269,
+      "learning_rate": 4.4709460477737607e-07,
+      "loss": 0.2429,
+      "step": 9732
+    },
+    {
+      "epoch": 3.1250602022796596,
+      "grad_norm": 0.5044448971748352,
+      "learning_rate": 4.438035396920004e-07,
+      "loss": 0.2206,
+      "step": 9733
+    },
+    {
+      "epoch": 3.125381281104511,
+      "grad_norm": 0.48041826486587524,
+      "learning_rate": 4.405246052271772e-07,
+      "loss": 0.2322,
+      "step": 9734
+    },
+    {
+      "epoch": 3.1257023599293627,
+      "grad_norm": 0.5572899580001831,
+      "learning_rate": 4.372578017824314e-07,
+      "loss": 0.232,
+      "step": 9735
+    },
+    {
+      "epoch": 3.1260234387542143,
+      "grad_norm": 0.6565282344818115,
+      "learning_rate": 4.3400312975581117e-07,
+      "loss": 0.2185,
+      "step": 9736
+    },
+    {
+      "epoch": 3.126344517579066,
+      "grad_norm": 0.484170526266098,
+      "learning_rate": 4.307605895439104e-07,
+      "loss": 0.2358,
+      "step": 9737
+    },
+    {
+      "epoch": 3.1266655964039174,
+      "grad_norm": 0.6327326893806458,
+      "learning_rate": 4.275301815417909e-07,
+      "loss": 0.2278,
+      "step": 9738
+    },
+    {
+      "epoch": 3.1269866752287685,
+      "grad_norm": 0.8204041719436646,
+      "learning_rate": 4.2431190614309335e-07,
+      "loss": 0.2654,
+      "step": 9739
+    },
+    {
+      "epoch": 3.12730775405362,
+      "grad_norm": 0.3902547359466553,
+      "learning_rate": 4.2110576373993736e-07,
+      "loss": 0.2428,
+      "step": 9740
+    },
+    {
+      "epoch": 3.1276288328784716,
+      "grad_norm": 0.44226837158203125,
+      "learning_rate": 4.179117547229883e-07,
+      "loss": 0.2411,
+      "step": 9741
+    },
+    {
+      "epoch": 3.127949911703323,
+      "grad_norm": 0.4273797273635864,
+      "learning_rate": 4.1472987948143473e-07,
+      "loss": 0.2463,
+      "step": 9742
+    },
+    {
+      "epoch": 3.1282709905281747,
+      "grad_norm": 0.29561299085617065,
+      "learning_rate": 4.115601384029666e-07,
+      "loss": 0.2297,
+      "step": 9743
+    },
+    {
+      "epoch": 3.1285920693530263,
+      "grad_norm": 0.5465667247772217,
+      "learning_rate": 4.084025318738083e-07,
+      "loss": 0.6041,
+      "step": 9744
+    },
+    {
+      "epoch": 3.128913148177878,
+      "grad_norm": 0.5179415941238403,
+      "learning_rate": 4.0525706027870756e-07,
+      "loss": 0.7336,
+      "step": 9745
+    },
+    {
+      "epoch": 3.1292342270027294,
+      "grad_norm": 0.45659175515174866,
+      "learning_rate": 4.021237240009468e-07,
+      "loss": 0.3653,
+      "step": 9746
+    },
+    {
+      "epoch": 3.129555305827581,
+      "grad_norm": 0.45041006803512573,
+      "learning_rate": 3.9900252342228717e-07,
+      "loss": 0.3496,
+      "step": 9747
+    },
+    {
+      "epoch": 3.129876384652432,
+      "grad_norm": 0.29173073172569275,
+      "learning_rate": 3.958934589230467e-07,
+      "loss": 0.1509,
+      "step": 9748
+    },
+    {
+      "epoch": 3.1301974634772836,
+      "grad_norm": 0.362936407327652,
+      "learning_rate": 3.9279653088205584e-07,
+      "loss": 0.1642,
+      "step": 9749
+    },
+    {
+      "epoch": 3.130518542302135,
+      "grad_norm": 0.23348329961299896,
+      "learning_rate": 3.8971173967666807e-07,
+      "loss": 0.0935,
+      "step": 9750
+    },
+    {
+      "epoch": 3.1308396211269867,
+      "grad_norm": 0.5271691083908081,
+      "learning_rate": 3.866390856827495e-07,
+      "loss": 0.2844,
+      "step": 9751
+    },
+    {
+      "epoch": 3.131160699951838,
+      "grad_norm": 0.5729438662528992,
+      "learning_rate": 3.835785692747118e-07,
+      "loss": 0.233,
+      "step": 9752
+    },
+    {
+      "epoch": 3.1314817787766898,
+      "grad_norm": 0.7787706851959229,
+      "learning_rate": 3.805301908254455e-07,
+      "loss": 0.3528,
+      "step": 9753
+    },
+    {
+      "epoch": 3.1318028576015413,
+      "grad_norm": 0.6220112442970276,
+      "learning_rate": 3.774939507063979e-07,
+      "loss": 0.2668,
+      "step": 9754
+    },
+    {
+      "epoch": 3.132123936426393,
+      "grad_norm": 0.5777502059936523,
+      "learning_rate": 3.744698492875398e-07,
+      "loss": 0.2084,
+      "step": 9755
+    },
+    {
+      "epoch": 3.1324450152512444,
+      "grad_norm": 0.6816080808639526,
+      "learning_rate": 3.7145788693732086e-07,
+      "loss": 0.2497,
+      "step": 9756
+    },
+    {
+      "epoch": 3.1327660940760955,
+      "grad_norm": 0.6858428120613098,
+      "learning_rate": 3.6845806402275863e-07,
+      "loss": 0.2512,
+      "step": 9757
+    },
+    {
+      "epoch": 3.133087172900947,
+      "grad_norm": 0.8424202799797058,
+      "learning_rate": 3.654703809093607e-07,
+      "loss": 0.2316,
+      "step": 9758
+    },
+    {
+      "epoch": 3.1334082517257986,
+      "grad_norm": 0.6304720640182495,
+      "learning_rate": 3.6249483796116924e-07,
+      "loss": 0.232,
+      "step": 9759
+    },
+    {
+      "epoch": 3.13372933055065,
+      "grad_norm": 0.9778458476066589,
+      "learning_rate": 3.595314355407609e-07,
+      "loss": 0.3145,
+      "step": 9760
+    },
+    {
+      "epoch": 3.1340504093755017,
+      "grad_norm": 0.593328058719635,
+      "learning_rate": 3.565801740092023e-07,
+      "loss": 0.2213,
+      "step": 9761
+    },
+    {
+      "epoch": 3.1343714882003533,
+      "grad_norm": 0.8587982058525085,
+      "learning_rate": 3.536410537260948e-07,
+      "loss": 0.2974,
+      "step": 9762
+    },
+    {
+      "epoch": 3.134692567025205,
+      "grad_norm": 0.6776645183563232,
+      "learning_rate": 3.50714075049563e-07,
+      "loss": 0.2801,
+      "step": 9763
+    },
+    {
+      "epoch": 3.1350136458500564,
+      "grad_norm": 0.8762674331665039,
+      "learning_rate": 3.4779923833626606e-07,
+      "loss": 0.3066,
+      "step": 9764
+    },
+    {
+      "epoch": 3.135334724674908,
+      "grad_norm": 0.6403717994689941,
+      "learning_rate": 3.4489654394134205e-07,
+      "loss": 0.2094,
+      "step": 9765
+    },
+    {
+      "epoch": 3.135655803499759,
+      "grad_norm": 0.4854280352592468,
+      "learning_rate": 3.4200599221848595e-07,
+      "loss": 0.1921,
+      "step": 9766
+    },
+    {
+      "epoch": 3.1359768823246106,
+      "grad_norm": 0.9062448143959045,
+      "learning_rate": 3.3912758351991593e-07,
+      "loss": 0.3486,
+      "step": 9767
+    },
+    {
+      "epoch": 3.136297961149462,
+      "grad_norm": 0.7571354508399963,
+      "learning_rate": 3.362613181963403e-07,
+      "loss": 0.2229,
+      "step": 9768
+    },
+    {
+      "epoch": 3.1366190399743137,
+      "grad_norm": 0.585010826587677,
+      "learning_rate": 3.3340719659701313e-07,
+      "loss": 0.2403,
+      "step": 9769
+    },
+    {
+      "epoch": 3.1369401187991652,
+      "grad_norm": 0.6055962443351746,
+      "learning_rate": 3.305652190696895e-07,
+      "loss": 0.2285,
+      "step": 9770
+    },
+    {
+      "epoch": 3.137261197624017,
+      "grad_norm": 0.49409618973731995,
+      "learning_rate": 3.277353859606813e-07,
+      "loss": 0.1914,
+      "step": 9771
+    },
+    {
+      "epoch": 3.1375822764488683,
+      "grad_norm": 0.39980348944664,
+      "learning_rate": 3.249176976147683e-07,
+      "loss": 0.1664,
+      "step": 9772
+    },
+    {
+      "epoch": 3.13790335527372,
+      "grad_norm": 0.574813187122345,
+      "learning_rate": 3.2211215437528694e-07,
+      "loss": 0.2273,
+      "step": 9773
+    },
+    {
+      "epoch": 3.138224434098571,
+      "grad_norm": 0.5716437101364136,
+      "learning_rate": 3.1931875658408604e-07,
+      "loss": 0.2453,
+      "step": 9774
+    },
+    {
+      "epoch": 3.1385455129234225,
+      "grad_norm": 0.40409037470817566,
+      "learning_rate": 3.1653750458152666e-07,
+      "loss": 0.1933,
+      "step": 9775
+    },
+    {
+      "epoch": 3.138866591748274,
+      "grad_norm": 0.6817755103111267,
+      "learning_rate": 3.137683987065043e-07,
+      "loss": 0.2179,
+      "step": 9776
+    },
+    {
+      "epoch": 3.1391876705731256,
+      "grad_norm": 0.6137420535087585,
+      "learning_rate": 3.1101143929641585e-07,
+      "loss": 0.2484,
+      "step": 9777
+    },
+    {
+      "epoch": 3.139508749397977,
+      "grad_norm": 0.7133082747459412,
+      "learning_rate": 3.0826662668720364e-07,
+      "loss": 0.2384,
+      "step": 9778
+    },
+    {
+      "epoch": 3.1398298282228287,
+      "grad_norm": 0.6078048348426819,
+      "learning_rate": 3.0553396121330013e-07,
+      "loss": 0.2337,
+      "step": 9779
+    },
+    {
+      "epoch": 3.1401509070476803,
+      "grad_norm": 0.325295627117157,
+      "learning_rate": 3.0281344320768347e-07,
+      "loss": 0.1829,
+      "step": 9780
+    },
+    {
+      "epoch": 3.140471985872532,
+      "grad_norm": 0.5655336976051331,
+      "learning_rate": 3.001050730018218e-07,
+      "loss": 0.2414,
+      "step": 9781
+    },
+    {
+      "epoch": 3.1407930646973834,
+      "grad_norm": 0.6369791626930237,
+      "learning_rate": 2.974088509257511e-07,
+      "loss": 0.2344,
+      "step": 9782
+    },
+    {
+      "epoch": 3.1411141435222345,
+      "grad_norm": 0.42372626066207886,
+      "learning_rate": 2.947247773079753e-07,
+      "loss": 0.2057,
+      "step": 9783
+    },
+    {
+      "epoch": 3.141435222347086,
+      "grad_norm": 0.550973653793335,
+      "learning_rate": 2.9205285247555505e-07,
+      "loss": 0.2484,
+      "step": 9784
+    },
+    {
+      "epoch": 3.1417563011719376,
+      "grad_norm": 0.5076101422309875,
+      "learning_rate": 2.893930767540298e-07,
+      "loss": 0.2296,
+      "step": 9785
+    },
+    {
+      "epoch": 3.142077379996789,
+      "grad_norm": 0.4814938008785248,
+      "learning_rate": 2.867454504675182e-07,
+      "loss": 0.2261,
+      "step": 9786
+    },
+    {
+      "epoch": 3.1423984588216407,
+      "grad_norm": 0.3789692521095276,
+      "learning_rate": 2.841099739386066e-07,
+      "loss": 0.2236,
+      "step": 9787
+    },
+    {
+      "epoch": 3.1427195376464923,
+      "grad_norm": 0.3235059976577759,
+      "learning_rate": 2.81486647488427e-07,
+      "loss": 0.2032,
+      "step": 9788
+    },
+    {
+      "epoch": 3.143040616471344,
+      "grad_norm": 0.4682113826274872,
+      "learning_rate": 2.7887547143662373e-07,
+      "loss": 0.2508,
+      "step": 9789
+    },
+    {
+      "epoch": 3.1433616952961954,
+      "grad_norm": 0.48588764667510986,
+      "learning_rate": 2.762764461013423e-07,
+      "loss": 0.2209,
+      "step": 9790
+    },
+    {
+      "epoch": 3.143682774121047,
+      "grad_norm": 0.4631730914115906,
+      "learning_rate": 2.73689571799296e-07,
+      "loss": 0.2206,
+      "step": 9791
+    },
+    {
+      "epoch": 3.144003852945898,
+      "grad_norm": 0.3348900377750397,
+      "learning_rate": 2.7111484884567717e-07,
+      "loss": 0.2183,
+      "step": 9792
+    },
+    {
+      "epoch": 3.1443249317707496,
+      "grad_norm": 0.2803027927875519,
+      "learning_rate": 2.685522775541904e-07,
+      "loss": 0.2277,
+      "step": 9793
+    },
+    {
+      "epoch": 3.144646010595601,
+      "grad_norm": 0.18889014422893524,
+      "learning_rate": 2.660018582370971e-07,
+      "loss": 0.1372,
+      "step": 9794
+    },
+    {
+      "epoch": 3.1449670894204527,
+      "grad_norm": 0.5592251420021057,
+      "learning_rate": 2.6346359120514863e-07,
+      "loss": 0.627,
+      "step": 9795
+    },
+    {
+      "epoch": 3.145288168245304,
+      "grad_norm": 0.42517322301864624,
+      "learning_rate": 2.609374767676309e-07,
+      "loss": 0.4038,
+      "step": 9796
+    },
+    {
+      "epoch": 3.1456092470701558,
+      "grad_norm": 0.48470279574394226,
+      "learning_rate": 2.584235152323422e-07,
+      "loss": 0.3099,
+      "step": 9797
+    },
+    {
+      "epoch": 3.1459303258950073,
+      "grad_norm": 0.23615846037864685,
+      "learning_rate": 2.5592170690560414e-07,
+      "loss": 0.0814,
+      "step": 9798
+    },
+    {
+      "epoch": 3.146251404719859,
+      "grad_norm": 0.24394987523555756,
+      "learning_rate": 2.534320520922506e-07,
+      "loss": 0.0938,
+      "step": 9799
+    },
+    {
+      "epoch": 3.1465724835447104,
+      "grad_norm": 0.1944095492362976,
+      "learning_rate": 2.5095455109562795e-07,
+      "loss": 0.0707,
+      "step": 9800
+    },
+    {
+      "epoch": 3.1468935623695615,
+      "grad_norm": 0.470803827047348,
+      "learning_rate": 2.484892042176279e-07,
+      "loss": 0.2102,
+      "step": 9801
+    },
+    {
+      "epoch": 3.147214641194413,
+      "grad_norm": 0.5542094111442566,
+      "learning_rate": 2.4603601175864356e-07,
+      "loss": 0.3056,
+      "step": 9802
+    },
+    {
+      "epoch": 3.1475357200192646,
+      "grad_norm": 0.6499555110931396,
+      "learning_rate": 2.4359497401758024e-07,
+      "loss": 0.287,
+      "step": 9803
+    },
+    {
+      "epoch": 3.147856798844116,
+      "grad_norm": 0.690673291683197,
+      "learning_rate": 2.4116609129187786e-07,
+      "loss": 0.2237,
+      "step": 9804
+    },
+    {
+      "epoch": 3.1481778776689677,
+      "grad_norm": 0.6088657975196838,
+      "learning_rate": 2.387493638774774e-07,
+      "loss": 0.2442,
+      "step": 9805
+    },
+    {
+      "epoch": 3.1484989564938193,
+      "grad_norm": 0.6405919194221497,
+      "learning_rate": 2.363447920688655e-07,
+      "loss": 0.248,
+      "step": 9806
+    },
+    {
+      "epoch": 3.148820035318671,
+      "grad_norm": 0.5600770711898804,
+      "learning_rate": 2.339523761590301e-07,
+      "loss": 0.2073,
+      "step": 9807
+    },
+    {
+      "epoch": 3.1491411141435224,
+      "grad_norm": 0.6620354652404785,
+      "learning_rate": 2.315721164394713e-07,
+      "loss": 0.2274,
+      "step": 9808
+    },
+    {
+      "epoch": 3.149462192968374,
+      "grad_norm": 0.8232807517051697,
+      "learning_rate": 2.2920401320022378e-07,
+      "loss": 0.2655,
+      "step": 9809
+    },
+    {
+      "epoch": 3.149783271793225,
+      "grad_norm": 0.48089903593063354,
+      "learning_rate": 2.2684806672982338e-07,
+      "loss": 0.197,
+      "step": 9810
+    },
+    {
+      "epoch": 3.1501043506180766,
+      "grad_norm": 0.6705470085144043,
+      "learning_rate": 2.2450427731534053e-07,
+      "loss": 0.2304,
+      "step": 9811
+    },
+    {
+      "epoch": 3.150425429442928,
+      "grad_norm": 0.7051324248313904,
+      "learning_rate": 2.2217264524236892e-07,
+      "loss": 0.2445,
+      "step": 9812
+    },
+    {
+      "epoch": 3.1507465082677797,
+      "grad_norm": 0.6257637143135071,
+      "learning_rate": 2.1985317079500356e-07,
+      "loss": 0.2012,
+      "step": 9813
+    },
+    {
+      "epoch": 3.1510675870926312,
+      "grad_norm": 0.9062849879264832,
+      "learning_rate": 2.175458542558517e-07,
+      "loss": 0.286,
+      "step": 9814
+    },
+    {
+      "epoch": 3.151388665917483,
+      "grad_norm": 0.5609592795372009,
+      "learning_rate": 2.1525069590607737e-07,
+      "loss": 0.185,
+      "step": 9815
+    },
+    {
+      "epoch": 3.1517097447423343,
+      "grad_norm": 0.5486966371536255,
+      "learning_rate": 2.1296769602532352e-07,
+      "loss": 0.2412,
+      "step": 9816
+    },
+    {
+      "epoch": 3.152030823567186,
+      "grad_norm": 0.7732871770858765,
+      "learning_rate": 2.106968548917676e-07,
+      "loss": 0.2621,
+      "step": 9817
+    },
+    {
+      "epoch": 3.1523519023920374,
+      "grad_norm": 0.5511267781257629,
+      "learning_rate": 2.0843817278209942e-07,
+      "loss": 0.2065,
+      "step": 9818
+    },
+    {
+      "epoch": 3.1526729812168885,
+      "grad_norm": 0.5723970532417297,
+      "learning_rate": 2.0619164997155438e-07,
+      "loss": 0.2028,
+      "step": 9819
+    },
+    {
+      "epoch": 3.15299406004174,
+      "grad_norm": 0.7419432997703552,
+      "learning_rate": 2.0395728673383574e-07,
+      "loss": 0.2796,
+      "step": 9820
+    },
+    {
+      "epoch": 3.1533151388665916,
+      "grad_norm": 0.7188732624053955,
+      "learning_rate": 2.017350833412146e-07,
+      "loss": 0.1864,
+      "step": 9821
+    },
+    {
+      "epoch": 3.153636217691443,
+      "grad_norm": 0.7375403642654419,
+      "learning_rate": 1.9952504006446325e-07,
+      "loss": 0.2474,
+      "step": 9822
+    },
+    {
+      "epoch": 3.1539572965162948,
+      "grad_norm": 0.5861619114875793,
+      "learning_rate": 1.973271571728441e-07,
+      "loss": 0.2157,
+      "step": 9823
+    },
+    {
+      "epoch": 3.1542783753411463,
+      "grad_norm": 0.42802998423576355,
+      "learning_rate": 1.9514143493417625e-07,
+      "loss": 0.2182,
+      "step": 9824
+    },
+    {
+      "epoch": 3.154599454165998,
+      "grad_norm": 0.6591411828994751,
+      "learning_rate": 1.9296787361480216e-07,
+      "loss": 0.246,
+      "step": 9825
+    },
+    {
+      "epoch": 3.1549205329908494,
+      "grad_norm": 0.8120198845863342,
+      "learning_rate": 1.908064734795323e-07,
+      "loss": 0.3286,
+      "step": 9826
+    },
+    {
+      "epoch": 3.155241611815701,
+      "grad_norm": 0.4934924840927124,
+      "learning_rate": 1.8865723479173368e-07,
+      "loss": 0.1993,
+      "step": 9827
+    },
+    {
+      "epoch": 3.155562690640552,
+      "grad_norm": 0.558791995048523,
+      "learning_rate": 1.86520157813308e-07,
+      "loss": 0.2377,
+      "step": 9828
+    },
+    {
+      "epoch": 3.1558837694654036,
+      "grad_norm": 0.8275998830795288,
+      "learning_rate": 1.8439524280462472e-07,
+      "loss": 0.2888,
+      "step": 9829
+    },
+    {
+      "epoch": 3.156204848290255,
+      "grad_norm": 0.5782402157783508,
+      "learning_rate": 1.8228249002461006e-07,
+      "loss": 0.2376,
+      "step": 9830
+    },
+    {
+      "epoch": 3.1565259271151067,
+      "grad_norm": 0.5709177255630493,
+      "learning_rate": 1.8018189973069143e-07,
+      "loss": 0.2141,
+      "step": 9831
+    },
+    {
+      "epoch": 3.1568470059399583,
+      "grad_norm": 0.6545057892799377,
+      "learning_rate": 1.7809347217881966e-07,
+      "loss": 0.2365,
+      "step": 9832
+    },
+    {
+      "epoch": 3.15716808476481,
+      "grad_norm": 0.7144107222557068,
+      "learning_rate": 1.7601720762346897e-07,
+      "loss": 0.2655,
+      "step": 9833
+    },
+    {
+      "epoch": 3.1574891635896614,
+      "grad_norm": 0.4486263394355774,
+      "learning_rate": 1.7395310631762585e-07,
+      "loss": 0.1869,
+      "step": 9834
+    },
+    {
+      "epoch": 3.157810242414513,
+      "grad_norm": 0.5604619979858398,
+      "learning_rate": 1.7190116851280026e-07,
+      "loss": 0.2374,
+      "step": 9835
+    },
+    {
+      "epoch": 3.1581313212393645,
+      "grad_norm": 0.5720604658126831,
+      "learning_rate": 1.698613944589922e-07,
+      "loss": 0.2534,
+      "step": 9836
+    },
+    {
+      "epoch": 3.1584524000642156,
+      "grad_norm": 0.5002235174179077,
+      "learning_rate": 1.678337844047695e-07,
+      "loss": 0.2297,
+      "step": 9837
+    },
+    {
+      "epoch": 3.158773478889067,
+      "grad_norm": 0.519655168056488,
+      "learning_rate": 1.6581833859716788e-07,
+      "loss": 0.249,
+      "step": 9838
+    },
+    {
+      "epoch": 3.1590945577139187,
+      "grad_norm": 0.4107975959777832,
+      "learning_rate": 1.6381505728176872e-07,
+      "loss": 0.2233,
+      "step": 9839
+    },
+    {
+      "epoch": 3.1594156365387702,
+      "grad_norm": 0.30553480982780457,
+      "learning_rate": 1.618239407026767e-07,
+      "loss": 0.2119,
+      "step": 9840
+    },
+    {
+      "epoch": 3.1597367153636218,
+      "grad_norm": 0.3645618259906769,
+      "learning_rate": 1.598449891024978e-07,
+      "loss": 0.2261,
+      "step": 9841
+    },
+    {
+      "epoch": 3.1600577941884733,
+      "grad_norm": 0.3786523938179016,
+      "learning_rate": 1.578782027223502e-07,
+      "loss": 0.2326,
+      "step": 9842
+    },
+    {
+      "epoch": 3.160378873013325,
+      "grad_norm": 0.4503689110279083,
+      "learning_rate": 1.5592358180189782e-07,
+      "loss": 0.2413,
+      "step": 9843
+    },
+    {
+      "epoch": 3.1606999518381764,
+      "grad_norm": 0.5784024000167847,
+      "learning_rate": 1.5398112657929453e-07,
+      "loss": 0.5337,
+      "step": 9844
+    },
+    {
+      "epoch": 3.161021030663028,
+      "grad_norm": 0.42213118076324463,
+      "learning_rate": 1.520508372912288e-07,
+      "loss": 0.4703,
+      "step": 9845
+    },
+    {
+      "epoch": 3.161342109487879,
+      "grad_norm": 0.3703862428665161,
+      "learning_rate": 1.5013271417290143e-07,
+      "loss": 0.2417,
+      "step": 9846
+    },
+    {
+      "epoch": 3.1616631883127306,
+      "grad_norm": 0.321639746427536,
+      "learning_rate": 1.4822675745801429e-07,
+      "loss": 0.1601,
+      "step": 9847
+    },
+    {
+      "epoch": 3.161984267137582,
+      "grad_norm": 0.1442064493894577,
+      "learning_rate": 1.4633296737882607e-07,
+      "loss": 0.0541,
+      "step": 9848
+    },
+    {
+      "epoch": 3.1623053459624337,
+      "grad_norm": 0.10288975387811661,
+      "learning_rate": 1.4445134416607442e-07,
+      "loss": 0.0532,
+      "step": 9849
+    },
+    {
+      "epoch": 3.1626264247872853,
+      "grad_norm": 0.21865668892860413,
+      "learning_rate": 1.425818880490315e-07,
+      "loss": 0.0609,
+      "step": 9850
+    },
+    {
+      "epoch": 3.162947503612137,
+      "grad_norm": 0.6108369827270508,
+      "learning_rate": 1.4072459925548177e-07,
+      "loss": 0.2515,
+      "step": 9851
+    },
+    {
+      "epoch": 3.1632685824369884,
+      "grad_norm": 0.9449090361595154,
+      "learning_rate": 1.3887947801173307e-07,
+      "loss": 0.2917,
+      "step": 9852
+    },
+    {
+      "epoch": 3.16358966126184,
+      "grad_norm": 0.6836480498313904,
+      "learning_rate": 1.3704652454261668e-07,
+      "loss": 0.231,
+      "step": 9853
+    },
+    {
+      "epoch": 3.1639107400866915,
+      "grad_norm": 0.5700660943984985,
+      "learning_rate": 1.3522573907145398e-07,
+      "loss": 0.2352,
+      "step": 9854
+    },
+    {
+      "epoch": 3.1642318189115426,
+      "grad_norm": 0.645478367805481,
+      "learning_rate": 1.33417121820123e-07,
+      "loss": 0.2189,
+      "step": 9855
+    },
+    {
+      "epoch": 3.164552897736394,
+      "grad_norm": 0.7627379894256592,
+      "learning_rate": 1.3162067300898084e-07,
+      "loss": 0.236,
+      "step": 9856
+    },
+    {
+      "epoch": 3.1648739765612457,
+      "grad_norm": 0.6092104911804199,
+      "learning_rate": 1.2983639285693017e-07,
+      "loss": 0.2363,
+      "step": 9857
+    },
+    {
+      "epoch": 3.1651950553860972,
+      "grad_norm": 0.7562824487686157,
+      "learning_rate": 1.2806428158138596e-07,
+      "loss": 0.2824,
+      "step": 9858
+    },
+    {
+      "epoch": 3.165516134210949,
+      "grad_norm": 0.7711445093154907,
+      "learning_rate": 1.2630433939825327e-07,
+      "loss": 0.2944,
+      "step": 9859
+    },
+    {
+      "epoch": 3.1658372130358003,
+      "grad_norm": 0.7503237724304199,
+      "learning_rate": 1.2455656652198277e-07,
+      "loss": 0.2468,
+      "step": 9860
+    },
+    {
+      "epoch": 3.166158291860652,
+      "grad_norm": 0.6290993690490723,
+      "learning_rate": 1.2282096316554858e-07,
+      "loss": 0.2132,
+      "step": 9861
+    },
+    {
+      "epoch": 3.1664793706855034,
+      "grad_norm": 0.7252655625343323,
+      "learning_rate": 1.2109752954042597e-07,
+      "loss": 0.2505,
+      "step": 9862
+    },
+    {
+      "epoch": 3.166800449510355,
+      "grad_norm": 0.5677751898765564,
+      "learning_rate": 1.193862658566025e-07,
+      "loss": 0.2001,
+      "step": 9863
+    },
+    {
+      "epoch": 3.167121528335206,
+      "grad_norm": 0.5677169561386108,
+      "learning_rate": 1.1768717232257809e-07,
+      "loss": 0.205,
+      "step": 9864
+    },
+    {
+      "epoch": 3.1674426071600577,
+      "grad_norm": 0.8806698322296143,
+      "learning_rate": 1.160002491454093e-07,
+      "loss": 0.3299,
+      "step": 9865
+    },
+    {
+      "epoch": 3.167763685984909,
+      "grad_norm": 0.4428505003452301,
+      "learning_rate": 1.1432549653063174e-07,
+      "loss": 0.1925,
+      "step": 9866
+    },
+    {
+      "epoch": 3.1680847648097608,
+      "grad_norm": 0.6543468236923218,
+      "learning_rate": 1.1266291468229328e-07,
+      "loss": 0.2174,
+      "step": 9867
+    },
+    {
+      "epoch": 3.1684058436346123,
+      "grad_norm": 0.6415979862213135,
+      "learning_rate": 1.1101250380300965e-07,
+      "loss": 0.243,
+      "step": 9868
+    },
+    {
+      "epoch": 3.168726922459464,
+      "grad_norm": 0.7214950323104858,
+      "learning_rate": 1.0937426409384221e-07,
+      "loss": 0.2707,
+      "step": 9869
+    },
+    {
+      "epoch": 3.1690480012843154,
+      "grad_norm": 0.7399327754974365,
+      "learning_rate": 1.0774819575442019e-07,
+      "loss": 0.2824,
+      "step": 9870
+    },
+    {
+      "epoch": 3.169369080109167,
+      "grad_norm": 0.6261020302772522,
+      "learning_rate": 1.0613429898287398e-07,
+      "loss": 0.2368,
+      "step": 9871
+    },
+    {
+      "epoch": 3.1696901589340185,
+      "grad_norm": 0.507150411605835,
+      "learning_rate": 1.0453257397585737e-07,
+      "loss": 0.2091,
+      "step": 9872
+    },
+    {
+      "epoch": 3.1700112377588696,
+      "grad_norm": 0.5187735557556152,
+      "learning_rate": 1.0294302092853647e-07,
+      "loss": 0.2175,
+      "step": 9873
+    },
+    {
+      "epoch": 3.170332316583721,
+      "grad_norm": 0.9378407001495361,
+      "learning_rate": 1.013656400345786e-07,
+      "loss": 0.2481,
+      "step": 9874
+    },
+    {
+      "epoch": 3.1706533954085727,
+      "grad_norm": 0.5098487734794617,
+      "learning_rate": 9.980043148619667e-08,
+      "loss": 0.2186,
+      "step": 9875
+    },
+    {
+      "epoch": 3.1709744742334243,
+      "grad_norm": 0.4638996422290802,
+      "learning_rate": 9.824739547410477e-08,
+      "loss": 0.2072,
+      "step": 9876
+    },
+    {
+      "epoch": 3.171295553058276,
+      "grad_norm": 0.7582216858863831,
+      "learning_rate": 9.670653218752934e-08,
+      "loss": 0.3402,
+      "step": 9877
+    },
+    {
+      "epoch": 3.1716166318831274,
+      "grad_norm": 0.42839518189430237,
+      "learning_rate": 9.517784181422019e-08,
+      "loss": 0.2069,
+      "step": 9878
+    },
+    {
+      "epoch": 3.171937710707979,
+      "grad_norm": 0.4199906885623932,
+      "learning_rate": 9.366132454046162e-08,
+      "loss": 0.2111,
+      "step": 9879
+    },
+    {
+      "epoch": 3.1722587895328305,
+      "grad_norm": 0.6500821113586426,
+      "learning_rate": 9.215698055100585e-08,
+      "loss": 0.2415,
+      "step": 9880
+    },
+    {
+      "epoch": 3.1725798683576816,
+      "grad_norm": 1.0988209247589111,
+      "learning_rate": 9.066481002918403e-08,
+      "loss": 0.2793,
+      "step": 9881
+    },
+    {
+      "epoch": 3.172900947182533,
+      "grad_norm": 0.4966912567615509,
+      "learning_rate": 8.918481315678406e-08,
+      "loss": 0.2166,
+      "step": 9882
+    },
+    {
+      "epoch": 3.1732220260073847,
+      "grad_norm": 0.4506468176841736,
+      "learning_rate": 8.771699011416168e-08,
+      "loss": 0.2084,
+      "step": 9883
+    },
+    {
+      "epoch": 3.1735431048322362,
+      "grad_norm": 0.5450963973999023,
+      "learning_rate": 8.626134108016271e-08,
+      "loss": 0.2226,
+      "step": 9884
+    },
+    {
+      "epoch": 3.173864183657088,
+      "grad_norm": 0.6653562784194946,
+      "learning_rate": 8.481786623214528e-08,
+      "loss": 0.2526,
+      "step": 9885
+    },
+    {
+      "epoch": 3.1741852624819393,
+      "grad_norm": 0.3943636417388916,
+      "learning_rate": 8.33865657459909e-08,
+      "loss": 0.2209,
+      "step": 9886
+    },
+    {
+      "epoch": 3.174506341306791,
+      "grad_norm": 0.6992161870002747,
+      "learning_rate": 8.196743979610455e-08,
+      "loss": 0.2377,
+      "step": 9887
+    },
+    {
+      "epoch": 3.1748274201316424,
+      "grad_norm": 0.40512463450431824,
+      "learning_rate": 8.056048855540343e-08,
+      "loss": 0.245,
+      "step": 9888
+    },
+    {
+      "epoch": 3.175148498956494,
+      "grad_norm": 0.5136418342590332,
+      "learning_rate": 7.91657121953171e-08,
+      "loss": 0.2392,
+      "step": 9889
+    },
+    {
+      "epoch": 3.175469577781345,
+      "grad_norm": 0.6193544864654541,
+      "learning_rate": 7.778311088579849e-08,
+      "loss": 0.255,
+      "step": 9890
+    },
+    {
+      "epoch": 3.1757906566061966,
+      "grad_norm": 0.4072697162628174,
+      "learning_rate": 7.641268479531283e-08,
+      "loss": 0.2205,
+      "step": 9891
+    },
+    {
+      "epoch": 3.176111735431048,
+      "grad_norm": 0.3897491693496704,
+      "learning_rate": 7.505443409083768e-08,
+      "loss": 0.2429,
+      "step": 9892
+    },
+    {
+      "epoch": 3.1764328142558997,
+      "grad_norm": 0.46470972895622253,
+      "learning_rate": 7.370835893788508e-08,
+      "loss": 0.2422,
+      "step": 9893
+    },
+    {
+      "epoch": 3.1767538930807513,
+      "grad_norm": 0.39090585708618164,
+      "learning_rate": 7.237445950044608e-08,
+      "loss": 0.3282,
+      "step": 9894
+    },
+    {
+      "epoch": 3.177074971905603,
+      "grad_norm": 0.5572707056999207,
+      "learning_rate": 7.105273594107953e-08,
+      "loss": 0.6177,
+      "step": 9895
+    },
+    {
+      "epoch": 3.1773960507304544,
+      "grad_norm": 0.3608555793762207,
+      "learning_rate": 6.974318842081217e-08,
+      "loss": 0.2933,
+      "step": 9896
+    },
+    {
+      "epoch": 3.177717129555306,
+      "grad_norm": 0.28094372153282166,
+      "learning_rate": 6.844581709921638e-08,
+      "loss": 0.16,
+      "step": 9897
+    },
+    {
+      "epoch": 3.1780382083801575,
+      "grad_norm": 0.4451110363006592,
+      "learning_rate": 6.71606221343768e-08,
+      "loss": 0.261,
+      "step": 9898
+    },
+    {
+      "epoch": 3.1783592872050086,
+      "grad_norm": 0.2656843364238739,
+      "learning_rate": 6.588760368287928e-08,
+      "loss": 0.0902,
+      "step": 9899
+    },
+    {
+      "epoch": 3.17868036602986,
+      "grad_norm": 0.17016752064228058,
+      "learning_rate": 6.462676189985528e-08,
+      "loss": 0.061,
+      "step": 9900
+    },
+    {
+      "epoch": 3.1790014448547117,
+      "grad_norm": 0.3899795413017273,
+      "learning_rate": 6.337809693891527e-08,
+      "loss": 0.2108,
+      "step": 9901
+    },
+    {
+      "epoch": 3.1793225236795633,
+      "grad_norm": 0.6329613924026489,
+      "learning_rate": 6.214160895222643e-08,
+      "loss": 0.2528,
+      "step": 9902
+    },
+    {
+      "epoch": 3.179643602504415,
+      "grad_norm": 0.6943358778953552,
+      "learning_rate": 6.09172980904238e-08,
+      "loss": 0.2953,
+      "step": 9903
+    },
+    {
+      "epoch": 3.1799646813292664,
+      "grad_norm": 0.8991124629974365,
+      "learning_rate": 5.970516450271025e-08,
+      "loss": 0.2841,
+      "step": 9904
+    },
+    {
+      "epoch": 3.180285760154118,
+      "grad_norm": 0.6497676968574524,
+      "learning_rate": 5.850520833676765e-08,
+      "loss": 0.2275,
+      "step": 9905
+    },
+    {
+      "epoch": 3.1806068389789695,
+      "grad_norm": 0.6367692947387695,
+      "learning_rate": 5.7317429738812376e-08,
+      "loss": 0.2095,
+      "step": 9906
+    },
+    {
+      "epoch": 3.180927917803821,
+      "grad_norm": 0.6393100619316101,
+      "learning_rate": 5.6141828853573106e-08,
+      "loss": 0.2308,
+      "step": 9907
+    },
+    {
+      "epoch": 3.181248996628672,
+      "grad_norm": 0.8095628023147583,
+      "learning_rate": 5.497840582429081e-08,
+      "loss": 0.3126,
+      "step": 9908
+    },
+    {
+      "epoch": 3.1815700754535237,
+      "grad_norm": 0.6111844182014465,
+      "learning_rate": 5.382716079271877e-08,
+      "loss": 0.2296,
+      "step": 9909
+    },
+    {
+      "epoch": 3.181891154278375,
+      "grad_norm": 0.4887303113937378,
+      "learning_rate": 5.268809389913365e-08,
+      "loss": 0.1672,
+      "step": 9910
+    },
+    {
+      "epoch": 3.1822122331032268,
+      "grad_norm": 0.6326609253883362,
+      "learning_rate": 5.1561205282335547e-08,
+      "loss": 0.2229,
+      "step": 9911
+    },
+    {
+      "epoch": 3.1825333119280783,
+      "grad_norm": 0.6799615025520325,
+      "learning_rate": 5.0446495079636836e-08,
+      "loss": 0.2754,
+      "step": 9912
+    },
+    {
+      "epoch": 3.18285439075293,
+      "grad_norm": 0.4907290041446686,
+      "learning_rate": 4.934396342684e-08,
+      "loss": 0.1849,
+      "step": 9913
+    },
+    {
+      "epoch": 3.1831754695777814,
+      "grad_norm": 0.7006059288978577,
+      "learning_rate": 4.825361045831534e-08,
+      "loss": 0.2467,
+      "step": 9914
+    },
+    {
+      "epoch": 3.183496548402633,
+      "grad_norm": 0.6643416285514832,
+      "learning_rate": 4.717543630688992e-08,
+      "loss": 0.2323,
+      "step": 9915
+    },
+    {
+      "epoch": 3.1838176272274845,
+      "grad_norm": 0.7749451994895935,
+      "learning_rate": 4.610944110394755e-08,
+      "loss": 0.2579,
+      "step": 9916
+    },
+    {
+      "epoch": 3.1841387060523356,
+      "grad_norm": 0.8002907633781433,
+      "learning_rate": 4.5055624979384316e-08,
+      "loss": 0.2735,
+      "step": 9917
+    },
+    {
+      "epoch": 3.184459784877187,
+      "grad_norm": 0.7704986929893494,
+      "learning_rate": 4.401398806159751e-08,
+      "loss": 0.2775,
+      "step": 9918
+    },
+    {
+      "epoch": 3.1847808637020387,
+      "grad_norm": 0.7502204775810242,
+      "learning_rate": 4.298453047749673e-08,
+      "loss": 0.255,
+      "step": 9919
+    },
+    {
+      "epoch": 3.1851019425268903,
+      "grad_norm": 0.4624794125556946,
+      "learning_rate": 4.196725235253718e-08,
+      "loss": 0.2155,
+      "step": 9920
+    },
+    {
+      "epoch": 3.185423021351742,
+      "grad_norm": 0.8697096705436707,
+      "learning_rate": 4.096215381066415e-08,
+      "loss": 0.3674,
+      "step": 9921
+    },
+    {
+      "epoch": 3.1857441001765934,
+      "grad_norm": 0.8463855385780334,
+      "learning_rate": 3.996923497434635e-08,
+      "loss": 0.3001,
+      "step": 9922
+    },
+    {
+      "epoch": 3.186065179001445,
+      "grad_norm": 0.7046919465065002,
+      "learning_rate": 3.898849596456478e-08,
+      "loss": 0.2382,
+      "step": 9923
+    },
+    {
+      "epoch": 3.1863862578262965,
+      "grad_norm": 0.7259396314620972,
+      "learning_rate": 3.8019936900812735e-08,
+      "loss": 0.2166,
+      "step": 9924
+    },
+    {
+      "epoch": 3.186707336651148,
+      "grad_norm": 0.8264093399047852,
+      "learning_rate": 3.7063557901129144e-08,
+      "loss": 0.2728,
+      "step": 9925
+    },
+    {
+      "epoch": 3.187028415475999,
+      "grad_norm": 0.5185015201568604,
+      "learning_rate": 3.61193590820208e-08,
+      "loss": 0.2227,
+      "step": 9926
+    },
+    {
+      "epoch": 3.1873494943008507,
+      "grad_norm": 0.9947241544723511,
+      "learning_rate": 3.518734055855122e-08,
+      "loss": 0.3701,
+      "step": 9927
+    },
+    {
+      "epoch": 3.1876705731257022,
+      "grad_norm": 0.5121382474899292,
+      "learning_rate": 3.4267502444274015e-08,
+      "loss": 0.1992,
+      "step": 9928
+    },
+    {
+      "epoch": 3.187991651950554,
+      "grad_norm": 0.5972121953964233,
+      "learning_rate": 3.33598448512773e-08,
+      "loss": 0.2321,
+      "step": 9929
+    },
+    {
+      "epoch": 3.1883127307754053,
+      "grad_norm": 0.543809175491333,
+      "learning_rate": 3.246436789015039e-08,
+      "loss": 0.2201,
+      "step": 9930
+    },
+    {
+      "epoch": 3.188633809600257,
+      "grad_norm": 0.8629385232925415,
+      "learning_rate": 3.1581071670006015e-08,
+      "loss": 0.2955,
+      "step": 9931
+    },
+    {
+      "epoch": 3.1889548884251084,
+      "grad_norm": 0.6378942728042603,
+      "learning_rate": 3.070995629846918e-08,
+      "loss": 0.2141,
+      "step": 9932
+    },
+    {
+      "epoch": 3.18927596724996,
+      "grad_norm": 0.5101456642150879,
+      "learning_rate": 2.985102188168831e-08,
+      "loss": 0.2327,
+      "step": 9933
+    },
+    {
+      "epoch": 3.1895970460748115,
+      "grad_norm": 0.6790507435798645,
+      "learning_rate": 2.900426852431304e-08,
+      "loss": 0.2728,
+      "step": 9934
+    },
+    {
+      "epoch": 3.1899181248996626,
+      "grad_norm": 0.4711005985736847,
+      "learning_rate": 2.8169696329527483e-08,
+      "loss": 0.2497,
+      "step": 9935
+    },
+    {
+      "epoch": 3.190239203724514,
+      "grad_norm": 1.078393816947937,
+      "learning_rate": 2.7347305399016977e-08,
+      "loss": 0.2378,
+      "step": 9936
+    },
+    {
+      "epoch": 3.1905602825493657,
+      "grad_norm": 0.6660891175270081,
+      "learning_rate": 2.6537095832990245e-08,
+      "loss": 0.2562,
+      "step": 9937
+    },
+    {
+      "epoch": 3.1908813613742173,
+      "grad_norm": 0.5800371170043945,
+      "learning_rate": 2.573906773016832e-08,
+      "loss": 0.2481,
+      "step": 9938
+    },
+    {
+      "epoch": 3.191202440199069,
+      "grad_norm": 0.47498077154159546,
+      "learning_rate": 2.4953221187784537e-08,
+      "loss": 0.2451,
+      "step": 9939
+    },
+    {
+      "epoch": 3.1915235190239204,
+      "grad_norm": 0.5962786078453064,
+      "learning_rate": 2.417955630159563e-08,
+      "loss": 0.2374,
+      "step": 9940
+    },
+    {
+      "epoch": 3.191844597848772,
+      "grad_norm": 0.3102143108844757,
+      "learning_rate": 2.3418073165870636e-08,
+      "loss": 0.246,
+      "step": 9941
+    },
+    {
+      "epoch": 3.1921656766736235,
+      "grad_norm": 0.6328103542327881,
+      "learning_rate": 2.266877187339089e-08,
+      "loss": 0.2762,
+      "step": 9942
+    },
+    {
+      "epoch": 3.192486755498475,
+      "grad_norm": 0.34649330377578735,
+      "learning_rate": 2.193165251545004e-08,
+      "loss": 0.237,
+      "step": 9943
+    },
+    {
+      "epoch": 3.192807834323326,
+      "grad_norm": 0.34423136711120605,
+      "learning_rate": 2.1206715181876226e-08,
+      "loss": 0.3348,
+      "step": 9944
+    },
+    {
+      "epoch": 3.1931289131481777,
+      "grad_norm": 0.44818225502967834,
+      "learning_rate": 2.0493959960998787e-08,
+      "loss": 0.4411,
+      "step": 9945
+    },
+    {
+      "epoch": 3.1934499919730293,
+      "grad_norm": 0.2298501431941986,
+      "learning_rate": 1.9793386939659376e-08,
+      "loss": 0.1117,
+      "step": 9946
+    },
+    {
+      "epoch": 3.193771070797881,
+      "grad_norm": 0.3223446011543274,
+      "learning_rate": 1.9104996203223037e-08,
+      "loss": 0.167,
+      "step": 9947
+    },
+    {
+      "epoch": 3.1940921496227324,
+      "grad_norm": 0.4785112738609314,
+      "learning_rate": 1.842878783557822e-08,
+      "loss": 0.2142,
+      "step": 9948
+    },
+    {
+      "epoch": 3.194413228447584,
+      "grad_norm": 0.2509891092777252,
+      "learning_rate": 1.7764761919103477e-08,
+      "loss": 0.1088,
+      "step": 9949
+    },
+    {
+      "epoch": 3.1947343072724355,
+      "grad_norm": 0.5531440377235413,
+      "learning_rate": 1.7112918534711865e-08,
+      "loss": 0.2591,
+      "step": 9950
+    },
+    {
+      "epoch": 3.195055386097287,
+      "grad_norm": 0.7335166931152344,
+      "learning_rate": 1.6473257761828732e-08,
+      "loss": 0.3958,
+      "step": 9951
+    },
+    {
+      "epoch": 3.1953764649221386,
+      "grad_norm": 0.8385985493659973,
+      "learning_rate": 1.584577967840284e-08,
+      "loss": 0.305,
+      "step": 9952
+    },
+    {
+      "epoch": 3.1956975437469897,
+      "grad_norm": 0.6894627809524536,
+      "learning_rate": 1.5230484360873044e-08,
+      "loss": 0.2275,
+      "step": 9953
+    },
+    {
+      "epoch": 3.196018622571841,
+      "grad_norm": 0.570214569568634,
+      "learning_rate": 1.4627371884234909e-08,
+      "loss": 0.2196,
+      "step": 9954
+    },
+    {
+      "epoch": 3.1963397013966928,
+      "grad_norm": 0.8872095346450806,
+      "learning_rate": 1.4036442321962995e-08,
+      "loss": 0.3162,
+      "step": 9955
+    },
+    {
+      "epoch": 3.1966607802215443,
+      "grad_norm": 0.6017862558364868,
+      "learning_rate": 1.3457695746055265e-08,
+      "loss": 0.2273,
+      "step": 9956
+    },
+    {
+      "epoch": 3.196981859046396,
+      "grad_norm": 0.6372138857841492,
+      "learning_rate": 1.2891132227033087e-08,
+      "loss": 0.2273,
+      "step": 9957
+    },
+    {
+      "epoch": 3.1973029378712474,
+      "grad_norm": 0.5612054467201233,
+      "learning_rate": 1.2336751833941229e-08,
+      "loss": 0.2014,
+      "step": 9958
+    },
+    {
+      "epoch": 3.197624016696099,
+      "grad_norm": 0.604500949382782,
+      "learning_rate": 1.1794554634314558e-08,
+      "loss": 0.2463,
+      "step": 9959
+    },
+    {
+      "epoch": 3.1979450955209505,
+      "grad_norm": 0.7940091490745544,
+      "learning_rate": 1.126454069423355e-08,
+      "loss": 0.3033,
+      "step": 9960
+    },
+    {
+      "epoch": 3.198266174345802,
+      "grad_norm": 0.7244638204574585,
+      "learning_rate": 1.074671007825767e-08,
+      "loss": 0.2531,
+      "step": 9961
+    },
+    {
+      "epoch": 3.198587253170653,
+      "grad_norm": 0.6201990842819214,
+      "learning_rate": 1.0241062849503103e-08,
+      "loss": 0.2504,
+      "step": 9962
+    },
+    {
+      "epoch": 3.1989083319955047,
+      "grad_norm": 0.5652727484703064,
+      "learning_rate": 9.747599069576119e-09,
+      "loss": 0.2249,
+      "step": 9963
+    },
+    {
+      "epoch": 3.1992294108203563,
+      "grad_norm": 0.810356616973877,
+      "learning_rate": 9.2663187986064e-09,
+      "loss": 0.3258,
+      "step": 9964
+    },
+    {
+      "epoch": 3.199550489645208,
+      "grad_norm": 0.8009127378463745,
+      "learning_rate": 8.797222095224823e-09,
+      "loss": 0.279,
+      "step": 9965
+    },
+    {
+      "epoch": 3.1998715684700594,
+      "grad_norm": 0.5440275073051453,
+      "learning_rate": 8.340309016585668e-09,
+      "loss": 0.2299,
+      "step": 9966
+    },
+    {
+      "epoch": 3.200192647294911,
+      "grad_norm": 0.7199362516403198,
+      "learning_rate": 7.895579618388827e-09,
+      "loss": 0.2572,
+      "step": 9967
+    },
+    {
+      "epoch": 3.2005137261197625,
+      "grad_norm": 0.5794646739959717,
+      "learning_rate": 7.463033954802078e-09,
+      "loss": 0.2327,
+      "step": 9968
+    },
+    {
+      "epoch": 3.200834804944614,
+      "grad_norm": 0.7608715295791626,
+      "learning_rate": 7.042672078527713e-09,
+      "loss": 0.284,
+      "step": 9969
+    },
+    {
+      "epoch": 3.2011558837694656,
+      "grad_norm": 0.6678571701049805,
+      "learning_rate": 6.634494040802519e-09,
+      "loss": 0.2936,
+      "step": 9970
+    },
+    {
+      "epoch": 3.2014769625943167,
+      "grad_norm": 0.7207247614860535,
+      "learning_rate": 6.238499891353389e-09,
+      "loss": 0.2576,
+      "step": 9971
+    },
+    {
+      "epoch": 3.2017980414191682,
+      "grad_norm": 0.7166668176651001,
+      "learning_rate": 5.854689678419512e-09,
+      "loss": 0.3074,
+      "step": 9972
+    },
+    {
+      "epoch": 3.20211912024402,
+      "grad_norm": 0.7045150399208069,
+      "learning_rate": 5.483063448785686e-09,
+      "loss": 0.2643,
+      "step": 9973
+    },
+    {
+      "epoch": 3.2024401990688713,
+      "grad_norm": 0.8350734114646912,
+      "learning_rate": 5.123621247726806e-09,
+      "loss": 0.2232,
+      "step": 9974
+    },
+    {
+      "epoch": 3.202761277893723,
+      "grad_norm": 0.7668434381484985,
+      "learning_rate": 4.776363119030069e-09,
+      "loss": 0.2897,
+      "step": 9975
+    },
+    {
+      "epoch": 3.2030823567185744,
+      "grad_norm": 0.4828073978424072,
+      "learning_rate": 4.4412891050171765e-09,
+      "loss": 0.2087,
+      "step": 9976
+    },
+    {
+      "epoch": 3.203403435543426,
+      "grad_norm": 0.5834342241287231,
+      "learning_rate": 4.1183992465221315e-09,
+      "loss": 0.238,
+      "step": 9977
+    },
+    {
+      "epoch": 3.2037245143682775,
+      "grad_norm": 0.6065945625305176,
+      "learning_rate": 3.807693582869032e-09,
+      "loss": 0.2669,
+      "step": 9978
+    },
+    {
+      "epoch": 3.204045593193129,
+      "grad_norm": 1.0055813789367676,
+      "learning_rate": 3.509172151938689e-09,
+      "loss": 0.3039,
+      "step": 9979
+    },
+    {
+      "epoch": 3.20436667201798,
+      "grad_norm": 0.6315884590148926,
+      "learning_rate": 3.222834990090906e-09,
+      "loss": 0.2177,
+      "step": 9980
+    },
+    {
+      "epoch": 3.2046877508428318,
+      "grad_norm": 0.5141793489456177,
+      "learning_rate": 2.948682132208891e-09,
+      "loss": 0.2139,
+      "step": 9981
+    },
+    {
+      "epoch": 3.2050088296676833,
+      "grad_norm": 0.8965477347373962,
+      "learning_rate": 2.6867136117214587e-09,
+      "loss": 0.2669,
+      "step": 9982
+    },
+    {
+      "epoch": 3.205329908492535,
+      "grad_norm": 0.4767134487628937,
+      "learning_rate": 2.4369294605253166e-09,
+      "loss": 0.2415,
+      "step": 9983
+    },
+    {
+      "epoch": 3.2056509873173864,
+      "grad_norm": 0.5410603880882263,
+      "learning_rate": 2.1993297090627806e-09,
+      "loss": 0.2093,
+      "step": 9984
+    },
+    {
+      "epoch": 3.205972066142238,
+      "grad_norm": 0.5572628378868103,
+      "learning_rate": 1.973914386288467e-09,
+      "loss": 0.2392,
+      "step": 9985
+    },
+    {
+      "epoch": 3.2062931449670895,
+      "grad_norm": 0.5035524964332581,
+      "learning_rate": 1.7606835196692927e-09,
+      "loss": 0.2171,
+      "step": 9986
+    },
+    {
+      "epoch": 3.206614223791941,
+      "grad_norm": 0.4417857229709625,
+      "learning_rate": 1.5596371351733752e-09,
+      "loss": 0.2107,
+      "step": 9987
+    },
+    {
+      "epoch": 3.2069353026167926,
+      "grad_norm": 0.646651029586792,
+      "learning_rate": 1.3707752573255405e-09,
+      "loss": 0.2278,
+      "step": 9988
+    },
+    {
+      "epoch": 3.2072563814416437,
+      "grad_norm": 0.7533266544342041,
+      "learning_rate": 1.1940979091074056e-09,
+      "loss": 0.3198,
+      "step": 9989
+    },
+    {
+      "epoch": 3.2075774602664953,
+      "grad_norm": 0.35283178091049194,
+      "learning_rate": 1.029605112068399e-09,
+      "loss": 0.2025,
+      "step": 9990
+    },
+    {
+      "epoch": 3.207898539091347,
+      "grad_norm": 0.4045920670032501,
+      "learning_rate": 8.772968862369446e-10,
+      "loss": 0.2284,
+      "step": 9991
+    },
+    {
+      "epoch": 3.2082196179161984,
+      "grad_norm": 0.4130226969718933,
+      "learning_rate": 7.371732501759709e-10,
+      "loss": 0.2453,
+      "step": 9992
+    },
+    {
+      "epoch": 3.20854069674105,
+      "grad_norm": 0.35095956921577454,
+      "learning_rate": 6.092342209607083e-10,
+      "loss": 0.2329,
+      "step": 9993
+    },
+    {
+      "epoch": 3.2088617755659015,
+      "grad_norm": 0.28531062602996826,
+      "learning_rate": 4.934798141786879e-10,
+      "loss": 0.1969,
+      "step": 9994
+    },
+    {
+      "epoch": 3.209182854390753,
+      "grad_norm": 0.4474446475505829,
+      "learning_rate": 3.899100439408443e-10,
+      "loss": 0.5176,
+      "step": 9995
+    },
+    {
+      "epoch": 3.2095039332156046,
+      "grad_norm": 0.36056339740753174,
+      "learning_rate": 2.985249228593112e-10,
+      "loss": 0.213,
+      "step": 9996
+    },
+    {
+      "epoch": 3.209825012040456,
+      "grad_norm": 0.27942565083503723,
+      "learning_rate": 2.1932446206962553e-10,
+      "loss": 0.1295,
+      "step": 9997
+    },
+    {
+      "epoch": 3.2101460908653072,
+      "grad_norm": 0.34559550881385803,
+      "learning_rate": 1.5230867123072756e-10,
+      "loss": 0.169,
+      "step": 9998
+    },
+    {
+      "epoch": 3.2104671696901588,
+      "grad_norm": 0.1161576583981514,
+      "learning_rate": 9.74775584916543e-11,
+      "loss": 0.0541,
+      "step": 9999
+    },
+    {
+      "epoch": 3.2107882485150103,
+      "grad_norm": 0.46221232414245605,
+      "learning_rate": 5.483113054705058e-11,
+      "loss": 0.1606,
+      "step": 10000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 10000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.492217387409408e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}