diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,30599 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998855180309101,
+  "eval_steps": 500,
+  "global_step": 4367,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0002289639381797367,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 2.1349,
+      "step": 1
+    },
+    {
+      "epoch": 0.0004579278763594734,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 2.1263,
+      "step": 2
+    },
+    {
+      "epoch": 0.0006868918145392101,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 2.1207,
+      "step": 3
+    },
+    {
+      "epoch": 0.0009158557527189468,
+      "grad_norm": 8.66817569732666,
+      "learning_rate": 1.5151515151515152e-07,
+      "loss": 2.1783,
+      "step": 4
+    },
+    {
+      "epoch": 0.0011448196908986834,
+      "grad_norm": 8.842265129089355,
+      "learning_rate": 3.0303030303030305e-07,
+      "loss": 2.0996,
+      "step": 5
+    },
+    {
+      "epoch": 0.0013737836290784202,
+      "grad_norm": 10.951105117797852,
+      "learning_rate": 4.5454545454545457e-07,
+      "loss": 2.2171,
+      "step": 6
+    },
+    {
+      "epoch": 0.0016027475672581567,
+      "grad_norm": 9.442892074584961,
+      "learning_rate": 6.060606060606061e-07,
+      "loss": 2.0768,
+      "step": 7
+    },
+    {
+      "epoch": 0.0018317115054378936,
+      "grad_norm": 8.939762115478516,
+      "learning_rate": 7.575757575757576e-07,
+      "loss": 2.1261,
+      "step": 8
+    },
+    {
+      "epoch": 0.00206067544361763,
+      "grad_norm": 8.166585922241211,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 2.0497,
+      "step": 9
+    },
+    {
+      "epoch": 0.0022896393817973667,
+      "grad_norm": 8.294096946716309,
+      "learning_rate": 1.0606060606060608e-06,
+      "loss": 2.0554,
+      "step": 10
+    },
+    {
+      "epoch": 0.0025186033199771037,
+      "grad_norm": 7.692963123321533,
+      "learning_rate": 1.2121212121212122e-06,
+      "loss": 2.0066,
+      "step": 11
+    },
+    {
+      "epoch": 0.0027475672581568403,
+      "grad_norm": 6.519928932189941,
+      "learning_rate": 1.3636363636363636e-06,
+      "loss": 1.804,
+      "step": 12
+    },
+    {
+      "epoch": 0.002976531196336577,
+      "grad_norm": 6.0504631996154785,
+      "learning_rate": 1.5151515151515152e-06,
+      "loss": 1.728,
+      "step": 13
+    },
+    {
+      "epoch": 0.0032054951345163135,
+      "grad_norm": 5.239335060119629,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.703,
+      "step": 14
+    },
+    {
+      "epoch": 0.0034344590726960505,
+      "grad_norm": 3.2573082447052,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 1.6709,
+      "step": 15
+    },
+    {
+      "epoch": 0.003663423010875787,
+      "grad_norm": 3.028881788253784,
+      "learning_rate": 1.96969696969697e-06,
+      "loss": 1.5926,
+      "step": 16
+    },
+    {
+      "epoch": 0.0038923869490555237,
+      "grad_norm": 2.915285110473633,
+      "learning_rate": 2.1212121212121216e-06,
+      "loss": 1.5392,
+      "step": 17
+    },
+    {
+      "epoch": 0.00412135088723526,
+      "grad_norm": 3.2491884231567383,
+      "learning_rate": 2.2727272727272728e-06,
+      "loss": 1.5468,
+      "step": 18
+    },
+    {
+      "epoch": 0.004350314825414997,
+      "grad_norm": 2.825366497039795,
+      "learning_rate": 2.4242424242424244e-06,
+      "loss": 1.4983,
+      "step": 19
+    },
+    {
+      "epoch": 0.0045792787635947334,
+      "grad_norm": 3.073415994644165,
+      "learning_rate": 2.575757575757576e-06,
+      "loss": 1.4596,
+      "step": 20
+    },
+    {
+      "epoch": 0.004808242701774471,
+      "grad_norm": 4.21063756942749,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 1.4261,
+      "step": 21
+    },
+    {
+      "epoch": 0.0050372066399542075,
+      "grad_norm": 2.9680824279785156,
+      "learning_rate": 2.8787878787878793e-06,
+      "loss": 1.4863,
+      "step": 22
+    },
+    {
+      "epoch": 0.005266170578133944,
+      "grad_norm": 3.912755250930786,
+      "learning_rate": 3.0303030303030305e-06,
+      "loss": 1.4069,
+      "step": 23
+    },
+    {
+      "epoch": 0.005495134516313681,
+      "grad_norm": 1.9926663637161255,
+      "learning_rate": 3.181818181818182e-06,
+      "loss": 1.3404,
+      "step": 24
+    },
+    {
+      "epoch": 0.005724098454493417,
+      "grad_norm": 5.043450355529785,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.402,
+      "step": 25
+    },
+    {
+      "epoch": 0.005953062392673154,
+      "grad_norm": 2.1705479621887207,
+      "learning_rate": 3.4848484848484854e-06,
+      "loss": 1.412,
+      "step": 26
+    },
+    {
+      "epoch": 0.00618202633085289,
+      "grad_norm": 1.6663283109664917,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 1.4681,
+      "step": 27
+    },
+    {
+      "epoch": 0.006410990269032627,
+      "grad_norm": 1.9405096769332886,
+      "learning_rate": 3.7878787878787882e-06,
+      "loss": 1.4265,
+      "step": 28
+    },
+    {
+      "epoch": 0.0066399542072123644,
+      "grad_norm": 2.0428900718688965,
+      "learning_rate": 3.93939393939394e-06,
+      "loss": 1.362,
+      "step": 29
+    },
+    {
+      "epoch": 0.006868918145392101,
+      "grad_norm": 2.2596077919006348,
+      "learning_rate": 4.0909090909090915e-06,
+      "loss": 1.373,
+      "step": 30
+    },
+    {
+      "epoch": 0.007097882083571838,
+      "grad_norm": 1.9293066263198853,
+      "learning_rate": 4.242424242424243e-06,
+      "loss": 1.3973,
+      "step": 31
+    },
+    {
+      "epoch": 0.007326846021751574,
+      "grad_norm": 1.669158697128296,
+      "learning_rate": 4.393939393939394e-06,
+      "loss": 1.3654,
+      "step": 32
+    },
+    {
+      "epoch": 0.007555809959931311,
+      "grad_norm": 1.5311822891235352,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 1.3982,
+      "step": 33
+    },
+    {
+      "epoch": 0.007784773898111047,
+      "grad_norm": 1.653910517692566,
+      "learning_rate": 4.696969696969698e-06,
+      "loss": 1.3749,
+      "step": 34
+    },
+    {
+      "epoch": 0.008013737836290785,
+      "grad_norm": 1.5567728281021118,
+      "learning_rate": 4.848484848484849e-06,
+      "loss": 1.3699,
+      "step": 35
+    },
+    {
+      "epoch": 0.00824270177447052,
+      "grad_norm": 1.6896796226501465,
+      "learning_rate": 5e-06,
+      "loss": 1.3356,
+      "step": 36
+    },
+    {
+      "epoch": 0.008471665712650258,
+      "grad_norm": 1.6226534843444824,
+      "learning_rate": 5.151515151515152e-06,
+      "loss": 1.3602,
+      "step": 37
+    },
+    {
+      "epoch": 0.008700629650829994,
+      "grad_norm": 1.81944739818573,
+      "learning_rate": 5.303030303030303e-06,
+      "loss": 1.3763,
+      "step": 38
+    },
+    {
+      "epoch": 0.008929593589009731,
+      "grad_norm": 1.9701725244522095,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 1.4414,
+      "step": 39
+    },
+    {
+      "epoch": 0.009158557527189467,
+      "grad_norm": 1.9407984018325806,
+      "learning_rate": 5.606060606060606e-06,
+      "loss": 1.3173,
+      "step": 40
+    },
+    {
+      "epoch": 0.009387521465369204,
+      "grad_norm": 1.5994501113891602,
+      "learning_rate": 5.7575757575757586e-06,
+      "loss": 1.3494,
+      "step": 41
+    },
+    {
+      "epoch": 0.009616485403548942,
+      "grad_norm": 1.8734986782073975,
+      "learning_rate": 5.90909090909091e-06,
+      "loss": 1.3658,
+      "step": 42
+    },
+    {
+      "epoch": 0.009845449341728678,
+      "grad_norm": 1.8831076622009277,
+      "learning_rate": 6.060606060606061e-06,
+      "loss": 1.3024,
+      "step": 43
+    },
+    {
+      "epoch": 0.010074413279908415,
+      "grad_norm": 1.6908974647521973,
+      "learning_rate": 6.212121212121213e-06,
+      "loss": 1.4275,
+      "step": 44
+    },
+    {
+      "epoch": 0.01030337721808815,
+      "grad_norm": 1.742285132408142,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 1.3254,
+      "step": 45
+    },
+    {
+      "epoch": 0.010532341156267888,
+      "grad_norm": 1.9381937980651855,
+      "learning_rate": 6.515151515151516e-06,
+      "loss": 1.3249,
+      "step": 46
+    },
+    {
+      "epoch": 0.010761305094447624,
+      "grad_norm": 1.445646047592163,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.325,
+      "step": 47
+    },
+    {
+      "epoch": 0.010990269032627361,
+      "grad_norm": 1.6461410522460938,
+      "learning_rate": 6.818181818181818e-06,
+      "loss": 1.3958,
+      "step": 48
+    },
+    {
+      "epoch": 0.011219232970807097,
+      "grad_norm": 2.5436205863952637,
+      "learning_rate": 6.969696969696971e-06,
+      "loss": 1.2518,
+      "step": 49
+    },
+    {
+      "epoch": 0.011448196908986834,
+      "grad_norm": 1.5389620065689087,
+      "learning_rate": 7.121212121212122e-06,
+      "loss": 1.2919,
+      "step": 50
+    },
+    {
+      "epoch": 0.011677160847166572,
+      "grad_norm": 1.7769535779953003,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 1.2673,
+      "step": 51
+    },
+    {
+      "epoch": 0.011906124785346308,
+      "grad_norm": 2.1329517364501953,
+      "learning_rate": 7.424242424242425e-06,
+      "loss": 1.2995,
+      "step": 52
+    },
+    {
+      "epoch": 0.012135088723526045,
+      "grad_norm": 1.5156912803649902,
+      "learning_rate": 7.5757575757575764e-06,
+      "loss": 1.3062,
+      "step": 53
+    },
+    {
+      "epoch": 0.01236405266170578,
+      "grad_norm": 1.4951452016830444,
+      "learning_rate": 7.727272727272727e-06,
+      "loss": 1.3429,
+      "step": 54
+    },
+    {
+      "epoch": 0.012593016599885518,
+      "grad_norm": 1.3737133741378784,
+      "learning_rate": 7.87878787878788e-06,
+      "loss": 1.2991,
+      "step": 55
+    },
+    {
+      "epoch": 0.012821980538065254,
+      "grad_norm": 1.3582038879394531,
+      "learning_rate": 8.03030303030303e-06,
+      "loss": 1.2792,
+      "step": 56
+    },
+    {
+      "epoch": 0.013050944476244991,
+      "grad_norm": 1.4764114618301392,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 1.3561,
+      "step": 57
+    },
+    {
+      "epoch": 0.013279908414424729,
+      "grad_norm": 1.6100513935089111,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.2303,
+      "step": 58
+    },
+    {
+      "epoch": 0.013508872352604465,
+      "grad_norm": 1.9281264543533325,
+      "learning_rate": 8.484848484848486e-06,
+      "loss": 1.3124,
+      "step": 59
+    },
+    {
+      "epoch": 0.013737836290784202,
+      "grad_norm": 1.6992368698120117,
+      "learning_rate": 8.636363636363637e-06,
+      "loss": 1.3066,
+      "step": 60
+    },
+    {
+      "epoch": 0.013966800228963938,
+      "grad_norm": 1.4923304319381714,
+      "learning_rate": 8.787878787878788e-06,
+      "loss": 1.3137,
+      "step": 61
+    },
+    {
+      "epoch": 0.014195764167143675,
+      "grad_norm": 3.8915576934814453,
+      "learning_rate": 8.93939393939394e-06,
+      "loss": 1.2634,
+      "step": 62
+    },
+    {
+      "epoch": 0.014424728105323411,
+      "grad_norm": 1.6618587970733643,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 1.2467,
+      "step": 63
+    },
+    {
+      "epoch": 0.014653692043503148,
+      "grad_norm": 1.4143376350402832,
+      "learning_rate": 9.242424242424244e-06,
+      "loss": 1.2817,
+      "step": 64
+    },
+    {
+      "epoch": 0.014882655981682884,
+      "grad_norm": 1.594414234161377,
+      "learning_rate": 9.393939393939396e-06,
+      "loss": 1.3015,
+      "step": 65
+    },
+    {
+      "epoch": 0.015111619919862622,
+      "grad_norm": 1.4304990768432617,
+      "learning_rate": 9.545454545454547e-06,
+      "loss": 1.3633,
+      "step": 66
+    },
+    {
+      "epoch": 0.015340583858042359,
+      "grad_norm": 2.2776196002960205,
+      "learning_rate": 9.696969696969698e-06,
+      "loss": 1.3109,
+      "step": 67
+    },
+    {
+      "epoch": 0.015569547796222095,
+      "grad_norm": 1.789172649383545,
+      "learning_rate": 9.84848484848485e-06,
+      "loss": 1.3003,
+      "step": 68
+    },
+    {
+      "epoch": 0.015798511734401832,
+      "grad_norm": 1.5805341005325317,
+      "learning_rate": 1e-05,
+      "loss": 1.3144,
+      "step": 69
+    },
+    {
+      "epoch": 0.01602747567258157,
+      "grad_norm": 1.2724578380584717,
+      "learning_rate": 1.0151515151515152e-05,
+      "loss": 1.2855,
+      "step": 70
+    },
+    {
+      "epoch": 0.016256439610761304,
+      "grad_norm": 1.451402187347412,
+      "learning_rate": 1.0303030303030304e-05,
+      "loss": 1.2442,
+      "step": 71
+    },
+    {
+      "epoch": 0.01648540354894104,
+      "grad_norm": 2.8804028034210205,
+      "learning_rate": 1.0454545454545455e-05,
+      "loss": 1.2494,
+      "step": 72
+    },
+    {
+      "epoch": 0.01671436748712078,
+      "grad_norm": 2.359454870223999,
+      "learning_rate": 1.0606060606060606e-05,
+      "loss": 1.1962,
+      "step": 73
+    },
+    {
+      "epoch": 0.016943331425300516,
+      "grad_norm": 3.0040102005004883,
+      "learning_rate": 1.0757575757575758e-05,
+      "loss": 1.26,
+      "step": 74
+    },
+    {
+      "epoch": 0.017172295363480253,
+      "grad_norm": 2.051539659500122,
+      "learning_rate": 1.0909090909090909e-05,
+      "loss": 1.2601,
+      "step": 75
+    },
+    {
+      "epoch": 0.017401259301659987,
+      "grad_norm": 2.0666730403900146,
+      "learning_rate": 1.1060606060606061e-05,
+      "loss": 1.3369,
+      "step": 76
+    },
+    {
+      "epoch": 0.017630223239839725,
+      "grad_norm": 1.3947479724884033,
+      "learning_rate": 1.1212121212121212e-05,
+      "loss": 1.2414,
+      "step": 77
+    },
+    {
+      "epoch": 0.017859187178019462,
+      "grad_norm": 1.6999517679214478,
+      "learning_rate": 1.1363636363636366e-05,
+      "loss": 1.2773,
+      "step": 78
+    },
+    {
+      "epoch": 0.0180881511161992,
+      "grad_norm": 1.3520228862762451,
+      "learning_rate": 1.1515151515151517e-05,
+      "loss": 1.2515,
+      "step": 79
+    },
+    {
+      "epoch": 0.018317115054378934,
+      "grad_norm": 1.7271336317062378,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 1.285,
+      "step": 80
+    },
+    {
+      "epoch": 0.01854607899255867,
+      "grad_norm": 1.5026074647903442,
+      "learning_rate": 1.181818181818182e-05,
+      "loss": 1.2984,
+      "step": 81
+    },
+    {
+      "epoch": 0.01877504293073841,
+      "grad_norm": 1.455031156539917,
+      "learning_rate": 1.1969696969696971e-05,
+      "loss": 1.1785,
+      "step": 82
+    },
+    {
+      "epoch": 0.019004006868918146,
+      "grad_norm": 2.4134457111358643,
+      "learning_rate": 1.2121212121212122e-05,
+      "loss": 1.2861,
+      "step": 83
+    },
+    {
+      "epoch": 0.019232970807097884,
+      "grad_norm": 1.3893452882766724,
+      "learning_rate": 1.2272727272727274e-05,
+      "loss": 1.3012,
+      "step": 84
+    },
+    {
+      "epoch": 0.019461934745277618,
+      "grad_norm": 2.0484437942504883,
+      "learning_rate": 1.2424242424242425e-05,
+      "loss": 1.2981,
+      "step": 85
+    },
+    {
+      "epoch": 0.019690898683457355,
+      "grad_norm": 1.8571381568908691,
+      "learning_rate": 1.2575757575757576e-05,
+      "loss": 1.2488,
+      "step": 86
+    },
+    {
+      "epoch": 0.019919862621637092,
+      "grad_norm": 1.5972386598587036,
+      "learning_rate": 1.2727272727272728e-05,
+      "loss": 1.2762,
+      "step": 87
+    },
+    {
+      "epoch": 0.02014882655981683,
+      "grad_norm": 1.9522424936294556,
+      "learning_rate": 1.287878787878788e-05,
+      "loss": 1.2176,
+      "step": 88
+    },
+    {
+      "epoch": 0.020377790497996564,
+      "grad_norm": 1.641571044921875,
+      "learning_rate": 1.3030303030303032e-05,
+      "loss": 1.21,
+      "step": 89
+    },
+    {
+      "epoch": 0.0206067544361763,
+      "grad_norm": 2.441856861114502,
+      "learning_rate": 1.3181818181818183e-05,
+      "loss": 1.2518,
+      "step": 90
+    },
+    {
+      "epoch": 0.02083571837435604,
+      "grad_norm": 2.1330080032348633,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.2659,
+      "step": 91
+    },
+    {
+      "epoch": 0.021064682312535776,
+      "grad_norm": 1.3508576154708862,
+      "learning_rate": 1.3484848484848486e-05,
+      "loss": 1.2586,
+      "step": 92
+    },
+    {
+      "epoch": 0.021293646250715514,
+      "grad_norm": 1.7117724418640137,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 1.3094,
+      "step": 93
+    },
+    {
+      "epoch": 0.021522610188895248,
+      "grad_norm": 1.690869927406311,
+      "learning_rate": 1.378787878787879e-05,
+      "loss": 1.1898,
+      "step": 94
+    },
+    {
+      "epoch": 0.021751574127074985,
+      "grad_norm": 2.3677923679351807,
+      "learning_rate": 1.3939393939393942e-05,
+      "loss": 1.2628,
+      "step": 95
+    },
+    {
+      "epoch": 0.021980538065254723,
+      "grad_norm": 1.5887540578842163,
+      "learning_rate": 1.4090909090909092e-05,
+      "loss": 1.2717,
+      "step": 96
+    },
+    {
+      "epoch": 0.02220950200343446,
+      "grad_norm": 2.6575193405151367,
+      "learning_rate": 1.4242424242424245e-05,
+      "loss": 1.2429,
+      "step": 97
+    },
+    {
+      "epoch": 0.022438465941614194,
+      "grad_norm": 1.425506830215454,
+      "learning_rate": 1.4393939393939396e-05,
+      "loss": 1.277,
+      "step": 98
+    },
+    {
+      "epoch": 0.02266742987979393,
+      "grad_norm": 1.729892611503601,
+      "learning_rate": 1.4545454545454546e-05,
+      "loss": 1.2601,
+      "step": 99
+    },
+    {
+      "epoch": 0.02289639381797367,
+      "grad_norm": 2.3861329555511475,
+      "learning_rate": 1.4696969696969699e-05,
+      "loss": 1.275,
+      "step": 100
+    },
+    {
+      "epoch": 0.023125357756153406,
+      "grad_norm": 1.65578031539917,
+      "learning_rate": 1.484848484848485e-05,
+      "loss": 1.248,
+      "step": 101
+    },
+    {
+      "epoch": 0.023354321694333144,
+      "grad_norm": 3.5684421062469482,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.2647,
+      "step": 102
+    },
+    {
+      "epoch": 0.023583285632512878,
+      "grad_norm": 19.274433135986328,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 1.2598,
+      "step": 103
+    },
+    {
+      "epoch": 0.023812249570692615,
+      "grad_norm": 1.5521622896194458,
+      "learning_rate": 1.5303030303030304e-05,
+      "loss": 1.2898,
+      "step": 104
+    },
+    {
+      "epoch": 0.024041213508872353,
+      "grad_norm": 1.7420704364776611,
+      "learning_rate": 1.5454545454545454e-05,
+      "loss": 1.289,
+      "step": 105
+    },
+    {
+      "epoch": 0.02427017744705209,
+      "grad_norm": 1.6716350317001343,
+      "learning_rate": 1.5606060606060605e-05,
+      "loss": 1.2162,
+      "step": 106
+    },
+    {
+      "epoch": 0.024499141385231828,
+      "grad_norm": 1.7894171476364136,
+      "learning_rate": 1.575757575757576e-05,
+      "loss": 1.2512,
+      "step": 107
+    },
+    {
+      "epoch": 0.02472810532341156,
+      "grad_norm": 2.7053062915802,
+      "learning_rate": 1.590909090909091e-05,
+      "loss": 1.2625,
+      "step": 108
+    },
+    {
+      "epoch": 0.0249570692615913,
+      "grad_norm": 1.7497886419296265,
+      "learning_rate": 1.606060606060606e-05,
+      "loss": 1.2179,
+      "step": 109
+    },
+    {
+      "epoch": 0.025186033199771037,
+      "grad_norm": 1.6836780309677124,
+      "learning_rate": 1.6212121212121212e-05,
+      "loss": 1.2076,
+      "step": 110
+    },
+    {
+      "epoch": 0.025414997137950774,
+      "grad_norm": 1.893149733543396,
+      "learning_rate": 1.6363636363636366e-05,
+      "loss": 1.2613,
+      "step": 111
+    },
+    {
+      "epoch": 0.025643961076130508,
+      "grad_norm": 1.6643422842025757,
+      "learning_rate": 1.6515151515151517e-05,
+      "loss": 1.2215,
+      "step": 112
+    },
+    {
+      "epoch": 0.025872925014310245,
+      "grad_norm": 1.624299168586731,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 1.2552,
+      "step": 113
+    },
+    {
+      "epoch": 0.026101888952489983,
+      "grad_norm": 2.46860408782959,
+      "learning_rate": 1.681818181818182e-05,
+      "loss": 1.2901,
+      "step": 114
+    },
+    {
+      "epoch": 0.02633085289066972,
+      "grad_norm": 2.3799362182617188,
+      "learning_rate": 1.6969696969696972e-05,
+      "loss": 1.198,
+      "step": 115
+    },
+    {
+      "epoch": 0.026559816828849458,
+      "grad_norm": 1.3481436967849731,
+      "learning_rate": 1.7121212121212123e-05,
+      "loss": 1.2413,
+      "step": 116
+    },
+    {
+      "epoch": 0.026788780767029192,
+      "grad_norm": 1.9589934349060059,
+      "learning_rate": 1.7272727272727274e-05,
+      "loss": 1.2845,
+      "step": 117
+    },
+    {
+      "epoch": 0.02701774470520893,
+      "grad_norm": 2.889359712600708,
+      "learning_rate": 1.7424242424242425e-05,
+      "loss": 1.2236,
+      "step": 118
+    },
+    {
+      "epoch": 0.027246708643388667,
+      "grad_norm": 2.009254217147827,
+      "learning_rate": 1.7575757575757576e-05,
+      "loss": 1.2147,
+      "step": 119
+    },
+    {
+      "epoch": 0.027475672581568404,
+      "grad_norm": 1.872424602508545,
+      "learning_rate": 1.772727272727273e-05,
+      "loss": 1.2385,
+      "step": 120
+    },
+    {
+      "epoch": 0.027704636519748138,
+      "grad_norm": 1.411062479019165,
+      "learning_rate": 1.787878787878788e-05,
+      "loss": 1.2238,
+      "step": 121
+    },
+    {
+      "epoch": 0.027933600457927876,
+      "grad_norm": 1.9089847803115845,
+      "learning_rate": 1.803030303030303e-05,
+      "loss": 1.3142,
+      "step": 122
+    },
+    {
+      "epoch": 0.028162564396107613,
+      "grad_norm": 1.5556221008300781,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 1.3167,
+      "step": 123
+    },
+    {
+      "epoch": 0.02839152833428735,
+      "grad_norm": 1.295181393623352,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 1.2127,
+      "step": 124
+    },
+    {
+      "epoch": 0.028620492272467088,
+      "grad_norm": 1.6264946460723877,
+      "learning_rate": 1.8484848484848487e-05,
+      "loss": 1.2766,
+      "step": 125
+    },
+    {
+      "epoch": 0.028849456210646822,
+      "grad_norm": 1.6156526803970337,
+      "learning_rate": 1.8636363636363638e-05,
+      "loss": 1.2583,
+      "step": 126
+    },
+    {
+      "epoch": 0.02907842014882656,
+      "grad_norm": 2.654879570007324,
+      "learning_rate": 1.8787878787878792e-05,
+      "loss": 1.223,
+      "step": 127
+    },
+    {
+      "epoch": 0.029307384087006297,
+      "grad_norm": 2.6344661712646484,
+      "learning_rate": 1.8939393939393943e-05,
+      "loss": 1.2516,
+      "step": 128
+    },
+    {
+      "epoch": 0.029536348025186034,
+      "grad_norm": 2.6589064598083496,
+      "learning_rate": 1.9090909090909094e-05,
+      "loss": 1.2435,
+      "step": 129
+    },
+    {
+      "epoch": 0.029765311963365768,
+      "grad_norm": 2.257918357849121,
+      "learning_rate": 1.9242424242424244e-05,
+      "loss": 1.1835,
+      "step": 130
+    },
+    {
+      "epoch": 0.029994275901545506,
+      "grad_norm": 1.8436076641082764,
+      "learning_rate": 1.9393939393939395e-05,
+      "loss": 1.1735,
+      "step": 131
+    },
+    {
+      "epoch": 0.030223239839725243,
+      "grad_norm": 1.7338463068008423,
+      "learning_rate": 1.9545454545454546e-05,
+      "loss": 1.1817,
+      "step": 132
+    },
+    {
+      "epoch": 0.03045220377790498,
+      "grad_norm": 2.028693675994873,
+      "learning_rate": 1.96969696969697e-05,
+      "loss": 1.2663,
+      "step": 133
+    },
+    {
+      "epoch": 0.030681167716084718,
+      "grad_norm": 1.6048945188522339,
+      "learning_rate": 1.984848484848485e-05,
+      "loss": 1.237,
+      "step": 134
+    },
+    {
+      "epoch": 0.030910131654264452,
+      "grad_norm": 1.475034236907959,
+      "learning_rate": 2e-05,
+      "loss": 1.2534,
+      "step": 135
+    },
+    {
+      "epoch": 0.03113909559244419,
+      "grad_norm": 1.4430214166641235,
+      "learning_rate": 1.9999997248541923e-05,
+      "loss": 1.2202,
+      "step": 136
+    },
+    {
+      "epoch": 0.031368059530623923,
+      "grad_norm": 1.705660343170166,
+      "learning_rate": 1.9999988994169196e-05,
+      "loss": 1.2087,
+      "step": 137
+    },
+    {
+      "epoch": 0.031597023468803664,
+      "grad_norm": 1.6410300731658936,
+      "learning_rate": 1.9999975236886367e-05,
+      "loss": 1.1895,
+      "step": 138
+    },
+    {
+      "epoch": 0.0318259874069834,
+      "grad_norm": 2.1050198078155518,
+      "learning_rate": 1.9999955976701005e-05,
+      "loss": 1.2435,
+      "step": 139
+    },
+    {
+      "epoch": 0.03205495134516314,
+      "grad_norm": 2.0697057247161865,
+      "learning_rate": 1.9999931213623708e-05,
+      "loss": 1.2032,
+      "step": 140
+    },
+    {
+      "epoch": 0.03228391528334287,
+      "grad_norm": 1.8335312604904175,
+      "learning_rate": 1.9999900947668106e-05,
+      "loss": 1.2381,
+      "step": 141
+    },
+    {
+      "epoch": 0.03251287922152261,
+      "grad_norm": 1.9289928674697876,
+      "learning_rate": 1.9999865178850847e-05,
+      "loss": 1.2319,
+      "step": 142
+    },
+    {
+      "epoch": 0.03274184315970235,
+      "grad_norm": 2.391378164291382,
+      "learning_rate": 1.9999823907191623e-05,
+      "loss": 1.2736,
+      "step": 143
+    },
+    {
+      "epoch": 0.03297080709788208,
+      "grad_norm": 1.6855164766311646,
+      "learning_rate": 1.9999777132713137e-05,
+      "loss": 1.2463,
+      "step": 144
+    },
+    {
+      "epoch": 0.03319977103606182,
+      "grad_norm": 1.5160446166992188,
+      "learning_rate": 1.999972485544114e-05,
+      "loss": 1.2183,
+      "step": 145
+    },
+    {
+      "epoch": 0.03342873497424156,
+      "grad_norm": 1.6463406085968018,
+      "learning_rate": 1.9999667075404385e-05,
+      "loss": 1.2261,
+      "step": 146
+    },
+    {
+      "epoch": 0.03365769891242129,
+      "grad_norm": 1.453431248664856,
+      "learning_rate": 1.999960379263468e-05,
+      "loss": 1.2804,
+      "step": 147
+    },
+    {
+      "epoch": 0.03388666285060103,
+      "grad_norm": 4.494565486907959,
+      "learning_rate": 1.9999535007166847e-05,
+      "loss": 1.19,
+      "step": 148
+    },
+    {
+      "epoch": 0.034115626788780766,
+      "grad_norm": 2.251443386077881,
+      "learning_rate": 1.999946071903873e-05,
+      "loss": 1.2244,
+      "step": 149
+    },
+    {
+      "epoch": 0.03434459072696051,
+      "grad_norm": 1.86344575881958,
+      "learning_rate": 1.999938092829122e-05,
+      "loss": 1.2541,
+      "step": 150
+    },
+    {
+      "epoch": 0.03457355466514024,
+      "grad_norm": 2.023585796356201,
+      "learning_rate": 1.9999295634968216e-05,
+      "loss": 1.2251,
+      "step": 151
+    },
+    {
+      "epoch": 0.034802518603319975,
+      "grad_norm": 1.6223890781402588,
+      "learning_rate": 1.999920483911666e-05,
+      "loss": 1.2049,
+      "step": 152
+    },
+    {
+      "epoch": 0.035031482541499716,
+      "grad_norm": 1.8935831785202026,
+      "learning_rate": 1.9999108540786513e-05,
+      "loss": 1.2355,
+      "step": 153
+    },
+    {
+      "epoch": 0.03526044647967945,
+      "grad_norm": 1.2119618654251099,
+      "learning_rate": 1.9999006740030774e-05,
+      "loss": 1.2546,
+      "step": 154
+    },
+    {
+      "epoch": 0.035489410417859184,
+      "grad_norm": 1.663476586341858,
+      "learning_rate": 1.999889943690545e-05,
+      "loss": 1.2184,
+      "step": 155
+    },
+    {
+      "epoch": 0.035718374356038925,
+      "grad_norm": 1.5204002857208252,
+      "learning_rate": 1.9998786631469602e-05,
+      "loss": 1.2055,
+      "step": 156
+    },
+    {
+      "epoch": 0.03594733829421866,
+      "grad_norm": 1.5706137418746948,
+      "learning_rate": 1.9998668323785298e-05,
+      "loss": 1.1985,
+      "step": 157
+    },
+    {
+      "epoch": 0.0361763022323984,
+      "grad_norm": 2.2016441822052,
+      "learning_rate": 1.9998544513917646e-05,
+      "loss": 1.2262,
+      "step": 158
+    },
+    {
+      "epoch": 0.036405266170578134,
+      "grad_norm": 1.5397722721099854,
+      "learning_rate": 1.9998415201934775e-05,
+      "loss": 1.1625,
+      "step": 159
+    },
+    {
+      "epoch": 0.03663423010875787,
+      "grad_norm": 1.7281806468963623,
+      "learning_rate": 1.9998280387907845e-05,
+      "loss": 1.2658,
+      "step": 160
+    },
+    {
+      "epoch": 0.03686319404693761,
+      "grad_norm": 1.4419702291488647,
+      "learning_rate": 1.9998140071911044e-05,
+      "loss": 1.1674,
+      "step": 161
+    },
+    {
+      "epoch": 0.03709215798511734,
+      "grad_norm": 1.9832121133804321,
+      "learning_rate": 1.9997994254021584e-05,
+      "loss": 1.2543,
+      "step": 162
+    },
+    {
+      "epoch": 0.03732112192329708,
+      "grad_norm": 1.2950409650802612,
+      "learning_rate": 1.999784293431971e-05,
+      "loss": 1.2494,
+      "step": 163
+    },
+    {
+      "epoch": 0.03755008586147682,
+      "grad_norm": 1.6367920637130737,
+      "learning_rate": 1.9997686112888694e-05,
+      "loss": 1.1819,
+      "step": 164
+    },
+    {
+      "epoch": 0.03777904979965655,
+      "grad_norm": 1.458843469619751,
+      "learning_rate": 1.9997523789814827e-05,
+      "loss": 1.2508,
+      "step": 165
+    },
+    {
+      "epoch": 0.03800801373783629,
+      "grad_norm": 1.2992488145828247,
+      "learning_rate": 1.999735596518744e-05,
+      "loss": 1.2549,
+      "step": 166
+    },
+    {
+      "epoch": 0.038236977676016026,
+      "grad_norm": 1.3639360666275024,
+      "learning_rate": 1.999718263909888e-05,
+      "loss": 1.3374,
+      "step": 167
+    },
+    {
+      "epoch": 0.03846594161419577,
+      "grad_norm": 1.6184186935424805,
+      "learning_rate": 1.9997003811644534e-05,
+      "loss": 1.1888,
+      "step": 168
+    },
+    {
+      "epoch": 0.0386949055523755,
+      "grad_norm": 1.2626984119415283,
+      "learning_rate": 1.9996819482922804e-05,
+      "loss": 1.1682,
+      "step": 169
+    },
+    {
+      "epoch": 0.038923869490555235,
+      "grad_norm": 1.3981817960739136,
+      "learning_rate": 1.9996629653035128e-05,
+      "loss": 1.2318,
+      "step": 170
+    },
+    {
+      "epoch": 0.039152833428734976,
+      "grad_norm": 1.3171190023422241,
+      "learning_rate": 1.999643432208596e-05,
+      "loss": 1.1496,
+      "step": 171
+    },
+    {
+      "epoch": 0.03938179736691471,
+      "grad_norm": 1.2882895469665527,
+      "learning_rate": 1.99962334901828e-05,
+      "loss": 1.2616,
+      "step": 172
+    },
+    {
+      "epoch": 0.03961076130509445,
+      "grad_norm": 1.8390611410140991,
+      "learning_rate": 1.9996027157436154e-05,
+      "loss": 1.3038,
+      "step": 173
+    },
+    {
+      "epoch": 0.039839725243274185,
+      "grad_norm": 1.3590266704559326,
+      "learning_rate": 1.9995815323959576e-05,
+      "loss": 1.2718,
+      "step": 174
+    },
+    {
+      "epoch": 0.04006868918145392,
+      "grad_norm": 1.7429498434066772,
+      "learning_rate": 1.9995597989869625e-05,
+      "loss": 1.2578,
+      "step": 175
+    },
+    {
+      "epoch": 0.04029765311963366,
+      "grad_norm": 1.5614861249923706,
+      "learning_rate": 1.9995375155285906e-05,
+      "loss": 1.1698,
+      "step": 176
+    },
+    {
+      "epoch": 0.040526617057813394,
+      "grad_norm": 2.0290753841400146,
+      "learning_rate": 1.999514682033104e-05,
+      "loss": 1.219,
+      "step": 177
+    },
+    {
+      "epoch": 0.04075558099599313,
+      "grad_norm": 1.4158384799957275,
+      "learning_rate": 1.9994912985130682e-05,
+      "loss": 1.1945,
+      "step": 178
+    },
+    {
+      "epoch": 0.04098454493417287,
+      "grad_norm": 1.7731130123138428,
+      "learning_rate": 1.99946736498135e-05,
+      "loss": 1.1985,
+      "step": 179
+    },
+    {
+      "epoch": 0.0412135088723526,
+      "grad_norm": 1.9962372779846191,
+      "learning_rate": 1.9994428814511205e-05,
+      "loss": 1.2413,
+      "step": 180
+    },
+    {
+      "epoch": 0.041442472810532344,
+      "grad_norm": 1.7486003637313843,
+      "learning_rate": 1.9994178479358526e-05,
+      "loss": 1.1465,
+      "step": 181
+    },
+    {
+      "epoch": 0.04167143674871208,
+      "grad_norm": 1.415873646736145,
+      "learning_rate": 1.9993922644493223e-05,
+      "loss": 1.2388,
+      "step": 182
+    },
+    {
+      "epoch": 0.04190040068689181,
+      "grad_norm": 1.6285479068756104,
+      "learning_rate": 1.9993661310056076e-05,
+      "loss": 1.286,
+      "step": 183
+    },
+    {
+      "epoch": 0.04212936462507155,
+      "grad_norm": 1.3321681022644043,
+      "learning_rate": 1.99933944761909e-05,
+      "loss": 1.1976,
+      "step": 184
+    },
+    {
+      "epoch": 0.04235832856325129,
+      "grad_norm": 1.5006210803985596,
+      "learning_rate": 1.999312214304453e-05,
+      "loss": 1.209,
+      "step": 185
+    },
+    {
+      "epoch": 0.04258729250143103,
+      "grad_norm": 3.087294101715088,
+      "learning_rate": 1.999284431076682e-05,
+      "loss": 1.2405,
+      "step": 186
+    },
+    {
+      "epoch": 0.04281625643961076,
+      "grad_norm": 2.618927240371704,
+      "learning_rate": 1.999256097951067e-05,
+      "loss": 1.2988,
+      "step": 187
+    },
+    {
+      "epoch": 0.043045220377790495,
+      "grad_norm": 1.4773956537246704,
+      "learning_rate": 1.999227214943199e-05,
+      "loss": 1.1566,
+      "step": 188
+    },
+    {
+      "epoch": 0.043274184315970236,
+      "grad_norm": 1.824690818786621,
+      "learning_rate": 1.999197782068972e-05,
+      "loss": 1.2149,
+      "step": 189
+    },
+    {
+      "epoch": 0.04350314825414997,
+      "grad_norm": 1.7128965854644775,
+      "learning_rate": 1.9991677993445832e-05,
+      "loss": 1.1904,
+      "step": 190
+    },
+    {
+      "epoch": 0.04373211219232971,
+      "grad_norm": 1.7998428344726562,
+      "learning_rate": 1.999137266786531e-05,
+      "loss": 1.2846,
+      "step": 191
+    },
+    {
+      "epoch": 0.043961076130509445,
+      "grad_norm": 2.0802230834960938,
+      "learning_rate": 1.9991061844116178e-05,
+      "loss": 1.2073,
+      "step": 192
+    },
+    {
+      "epoch": 0.04419004006868918,
+      "grad_norm": 1.7808719873428345,
+      "learning_rate": 1.9990745522369482e-05,
+      "loss": 1.2883,
+      "step": 193
+    },
+    {
+      "epoch": 0.04441900400686892,
+      "grad_norm": 1.7531901597976685,
+      "learning_rate": 1.9990423702799283e-05,
+      "loss": 1.1568,
+      "step": 194
+    },
+    {
+      "epoch": 0.044647967945048654,
+      "grad_norm": 1.5850685834884644,
+      "learning_rate": 1.9990096385582682e-05,
+      "loss": 1.252,
+      "step": 195
+    },
+    {
+      "epoch": 0.04487693188322839,
+      "grad_norm": 1.3775941133499146,
+      "learning_rate": 1.9989763570899795e-05,
+      "loss": 1.2807,
+      "step": 196
+    },
+    {
+      "epoch": 0.04510589582140813,
+      "grad_norm": 1.3636221885681152,
+      "learning_rate": 1.998942525893377e-05,
+      "loss": 1.2385,
+      "step": 197
+    },
+    {
+      "epoch": 0.04533485975958786,
+      "grad_norm": 1.4277960062026978,
+      "learning_rate": 1.998908144987078e-05,
+      "loss": 1.166,
+      "step": 198
+    },
+    {
+      "epoch": 0.045563823697767604,
+      "grad_norm": 1.4968419075012207,
+      "learning_rate": 1.9988732143900013e-05,
+      "loss": 1.203,
+      "step": 199
+    },
+    {
+      "epoch": 0.04579278763594734,
+      "grad_norm": 1.6300857067108154,
+      "learning_rate": 1.998837734121369e-05,
+      "loss": 1.1981,
+      "step": 200
+    },
+    {
+      "epoch": 0.04602175157412707,
+      "grad_norm": 1.3024663925170898,
+      "learning_rate": 1.9988017042007066e-05,
+      "loss": 1.2181,
+      "step": 201
+    },
+    {
+      "epoch": 0.04625071551230681,
+      "grad_norm": 1.2611545324325562,
+      "learning_rate": 1.99876512464784e-05,
+      "loss": 1.2375,
+      "step": 202
+    },
+    {
+      "epoch": 0.04647967945048655,
+      "grad_norm": 1.7449328899383545,
+      "learning_rate": 1.9987279954828986e-05,
+      "loss": 1.2178,
+      "step": 203
+    },
+    {
+      "epoch": 0.04670864338866629,
+      "grad_norm": 1.4038939476013184,
+      "learning_rate": 1.998690316726315e-05,
+      "loss": 1.225,
+      "step": 204
+    },
+    {
+      "epoch": 0.04693760732684602,
+      "grad_norm": 1.573346495628357,
+      "learning_rate": 1.9986520883988233e-05,
+      "loss": 1.2241,
+      "step": 205
+    },
+    {
+      "epoch": 0.047166571265025756,
+      "grad_norm": 1.1829012632369995,
+      "learning_rate": 1.99861331052146e-05,
+      "loss": 1.1722,
+      "step": 206
+    },
+    {
+      "epoch": 0.0473955352032055,
+      "grad_norm": 1.6095627546310425,
+      "learning_rate": 1.9985739831155637e-05,
+      "loss": 1.2481,
+      "step": 207
+    },
+    {
+      "epoch": 0.04762449914138523,
+      "grad_norm": 1.8268073797225952,
+      "learning_rate": 1.998534106202777e-05,
+      "loss": 1.1972,
+      "step": 208
+    },
+    {
+      "epoch": 0.04785346307956497,
+      "grad_norm": 1.7510496377944946,
+      "learning_rate": 1.9984936798050435e-05,
+      "loss": 1.1889,
+      "step": 209
+    },
+    {
+      "epoch": 0.048082427017744706,
+      "grad_norm": 1.9209846258163452,
+      "learning_rate": 1.9984527039446093e-05,
+      "loss": 1.2393,
+      "step": 210
+    },
+    {
+      "epoch": 0.04831139095592444,
+      "grad_norm": 1.3671090602874756,
+      "learning_rate": 1.998411178644023e-05,
+      "loss": 1.1783,
+      "step": 211
+    },
+    {
+      "epoch": 0.04854035489410418,
+      "grad_norm": 1.4478919506072998,
+      "learning_rate": 1.9983691039261358e-05,
+      "loss": 1.269,
+      "step": 212
+    },
+    {
+      "epoch": 0.048769318832283914,
+      "grad_norm": 1.8099792003631592,
+      "learning_rate": 1.998326479814101e-05,
+      "loss": 1.1632,
+      "step": 213
+    },
+    {
+      "epoch": 0.048998282770463655,
+      "grad_norm": 1.502695083618164,
+      "learning_rate": 1.9982833063313746e-05,
+      "loss": 1.2305,
+      "step": 214
+    },
+    {
+      "epoch": 0.04922724670864339,
+      "grad_norm": 1.1444593667984009,
+      "learning_rate": 1.998239583501714e-05,
+      "loss": 1.1981,
+      "step": 215
+    },
+    {
+      "epoch": 0.04945621064682312,
+      "grad_norm": 1.7091630697250366,
+      "learning_rate": 1.99819531134918e-05,
+      "loss": 1.2201,
+      "step": 216
+    },
+    {
+      "epoch": 0.049685174585002864,
+      "grad_norm": 1.4177758693695068,
+      "learning_rate": 1.9981504898981352e-05,
+      "loss": 1.1817,
+      "step": 217
+    },
+    {
+      "epoch": 0.0499141385231826,
+      "grad_norm": 1.2633055448532104,
+      "learning_rate": 1.998105119173244e-05,
+      "loss": 1.191,
+      "step": 218
+    },
+    {
+      "epoch": 0.05014310246136233,
+      "grad_norm": 1.7247159481048584,
+      "learning_rate": 1.998059199199474e-05,
+      "loss": 1.2037,
+      "step": 219
+    },
+    {
+      "epoch": 0.05037206639954207,
+      "grad_norm": 1.7244751453399658,
+      "learning_rate": 1.9980127300020946e-05,
+      "loss": 1.1898,
+      "step": 220
+    },
+    {
+      "epoch": 0.05060103033772181,
+      "grad_norm": 1.452858567237854,
+      "learning_rate": 1.997965711606677e-05,
+      "loss": 1.2408,
+      "step": 221
+    },
+    {
+      "epoch": 0.05082999427590155,
+      "grad_norm": 1.4057929515838623,
+      "learning_rate": 1.9979181440390954e-05,
+      "loss": 1.2024,
+      "step": 222
+    },
+    {
+      "epoch": 0.05105895821408128,
+      "grad_norm": 1.3213428258895874,
+      "learning_rate": 1.9978700273255254e-05,
+      "loss": 1.2326,
+      "step": 223
+    },
+    {
+      "epoch": 0.051287922152261016,
+      "grad_norm": 1.3907994031906128,
+      "learning_rate": 1.9978213614924456e-05,
+      "loss": 1.169,
+      "step": 224
+    },
+    {
+      "epoch": 0.05151688609044076,
+      "grad_norm": 1.4973134994506836,
+      "learning_rate": 1.9977721465666365e-05,
+      "loss": 1.2163,
+      "step": 225
+    },
+    {
+      "epoch": 0.05174585002862049,
+      "grad_norm": 1.6128813028335571,
+      "learning_rate": 1.9977223825751802e-05,
+      "loss": 1.1489,
+      "step": 226
+    },
+    {
+      "epoch": 0.05197481396680023,
+      "grad_norm": 1.228517770767212,
+      "learning_rate": 1.997672069545462e-05,
+      "loss": 1.1924,
+      "step": 227
+    },
+    {
+      "epoch": 0.052203777904979966,
+      "grad_norm": 1.5707398653030396,
+      "learning_rate": 1.9976212075051683e-05,
+      "loss": 1.2292,
+      "step": 228
+    },
+    {
+      "epoch": 0.0524327418431597,
+      "grad_norm": 1.605344533920288,
+      "learning_rate": 1.997569796482288e-05,
+      "loss": 1.3047,
+      "step": 229
+    },
+    {
+      "epoch": 0.05266170578133944,
+      "grad_norm": 1.1628702878952026,
+      "learning_rate": 1.9975178365051123e-05,
+      "loss": 1.205,
+      "step": 230
+    },
+    {
+      "epoch": 0.052890669719519175,
+      "grad_norm": 1.4649019241333008,
+      "learning_rate": 1.9974653276022347e-05,
+      "loss": 1.2032,
+      "step": 231
+    },
+    {
+      "epoch": 0.053119633657698916,
+      "grad_norm": 1.4579243659973145,
+      "learning_rate": 1.99741226980255e-05,
+      "loss": 1.238,
+      "step": 232
+    },
+    {
+      "epoch": 0.05334859759587865,
+      "grad_norm": 1.4899598360061646,
+      "learning_rate": 1.997358663135255e-05,
+      "loss": 1.2124,
+      "step": 233
+    },
+    {
+      "epoch": 0.053577561534058384,
+      "grad_norm": 1.2446345090866089,
+      "learning_rate": 1.99730450762985e-05,
+      "loss": 1.2593,
+      "step": 234
+    },
+    {
+      "epoch": 0.053806525472238124,
+      "grad_norm": 1.5485501289367676,
+      "learning_rate": 1.997249803316136e-05,
+      "loss": 1.1541,
+      "step": 235
+    },
+    {
+      "epoch": 0.05403548941041786,
+      "grad_norm": 1.386314868927002,
+      "learning_rate": 1.997194550224216e-05,
+      "loss": 1.234,
+      "step": 236
+    },
+    {
+      "epoch": 0.05426445334859759,
+      "grad_norm": 1.917922854423523,
+      "learning_rate": 1.9971387483844956e-05,
+      "loss": 1.2255,
+      "step": 237
+    },
+    {
+      "epoch": 0.05449341728677733,
+      "grad_norm": 1.9101078510284424,
+      "learning_rate": 1.9970823978276818e-05,
+      "loss": 1.1737,
+      "step": 238
+    },
+    {
+      "epoch": 0.05472238122495707,
+      "grad_norm": 2.0706257820129395,
+      "learning_rate": 1.9970254985847842e-05,
+      "loss": 1.1984,
+      "step": 239
+    },
+    {
+      "epoch": 0.05495134516313681,
+      "grad_norm": 1.579801321029663,
+      "learning_rate": 1.9969680506871138e-05,
+      "loss": 1.1953,
+      "step": 240
+    },
+    {
+      "epoch": 0.05518030910131654,
+      "grad_norm": 1.2857059240341187,
+      "learning_rate": 1.9969100541662833e-05,
+      "loss": 1.2072,
+      "step": 241
+    },
+    {
+      "epoch": 0.055409273039496276,
+      "grad_norm": 1.162064552307129,
+      "learning_rate": 1.996851509054208e-05,
+      "loss": 1.1743,
+      "step": 242
+    },
+    {
+      "epoch": 0.05563823697767602,
+      "grad_norm": 1.7716028690338135,
+      "learning_rate": 1.9967924153831054e-05,
+      "loss": 1.1876,
+      "step": 243
+    },
+    {
+      "epoch": 0.05586720091585575,
+      "grad_norm": 1.3083211183547974,
+      "learning_rate": 1.9967327731854937e-05,
+      "loss": 1.2037,
+      "step": 244
+    },
+    {
+      "epoch": 0.05609616485403549,
+      "grad_norm": 1.6880347728729248,
+      "learning_rate": 1.9966725824941933e-05,
+      "loss": 1.112,
+      "step": 245
+    },
+    {
+      "epoch": 0.056325128792215226,
+      "grad_norm": 1.951745629310608,
+      "learning_rate": 1.996611843342327e-05,
+      "loss": 1.2209,
+      "step": 246
+    },
+    {
+      "epoch": 0.05655409273039496,
+      "grad_norm": 2.4489657878875732,
+      "learning_rate": 1.9965505557633188e-05,
+      "loss": 1.1525,
+      "step": 247
+    },
+    {
+      "epoch": 0.0567830566685747,
+      "grad_norm": 1.4747159481048584,
+      "learning_rate": 1.996488719790895e-05,
+      "loss": 1.1597,
+      "step": 248
+    },
+    {
+      "epoch": 0.057012020606754435,
+      "grad_norm": 1.5697540044784546,
+      "learning_rate": 1.9964263354590835e-05,
+      "loss": 1.1907,
+      "step": 249
+    },
+    {
+      "epoch": 0.057240984544934176,
+      "grad_norm": 1.4222530126571655,
+      "learning_rate": 1.9963634028022133e-05,
+      "loss": 1.2253,
+      "step": 250
+    },
+    {
+      "epoch": 0.05746994848311391,
+      "grad_norm": 1.2914453744888306,
+      "learning_rate": 1.9962999218549156e-05,
+      "loss": 1.2175,
+      "step": 251
+    },
+    {
+      "epoch": 0.057698912421293644,
+      "grad_norm": 1.646164059638977,
+      "learning_rate": 1.9962358926521245e-05,
+      "loss": 1.1746,
+      "step": 252
+    },
+    {
+      "epoch": 0.057927876359473385,
+      "grad_norm": 2.148345947265625,
+      "learning_rate": 1.996171315229074e-05,
+      "loss": 1.1377,
+      "step": 253
+    },
+    {
+      "epoch": 0.05815684029765312,
+      "grad_norm": 1.5290459394454956,
+      "learning_rate": 1.9961061896213006e-05,
+      "loss": 1.1186,
+      "step": 254
+    },
+    {
+      "epoch": 0.05838580423583286,
+      "grad_norm": 1.3710248470306396,
+      "learning_rate": 1.9960405158646425e-05,
+      "loss": 1.1941,
+      "step": 255
+    },
+    {
+      "epoch": 0.058614768174012594,
+      "grad_norm": 1.6446150541305542,
+      "learning_rate": 1.9959742939952393e-05,
+      "loss": 1.1667,
+      "step": 256
+    },
+    {
+      "epoch": 0.05884373211219233,
+      "grad_norm": 1.3985366821289062,
+      "learning_rate": 1.9959075240495322e-05,
+      "loss": 1.1506,
+      "step": 257
+    },
+    {
+      "epoch": 0.05907269605037207,
+      "grad_norm": 1.5973879098892212,
+      "learning_rate": 1.9958402060642644e-05,
+      "loss": 1.1326,
+      "step": 258
+    },
+    {
+      "epoch": 0.0593016599885518,
+      "grad_norm": 1.499578833580017,
+      "learning_rate": 1.9957723400764803e-05,
+      "loss": 1.1938,
+      "step": 259
+    },
+    {
+      "epoch": 0.059530623926731537,
+      "grad_norm": 1.3453638553619385,
+      "learning_rate": 1.9957039261235263e-05,
+      "loss": 1.2466,
+      "step": 260
+    },
+    {
+      "epoch": 0.05975958786491128,
+      "grad_norm": 1.7341777086257935,
+      "learning_rate": 1.9956349642430494e-05,
+      "loss": 1.1251,
+      "step": 261
+    },
+    {
+      "epoch": 0.05998855180309101,
+      "grad_norm": 1.2891796827316284,
+      "learning_rate": 1.9955654544729992e-05,
+      "loss": 1.1765,
+      "step": 262
+    },
+    {
+      "epoch": 0.06021751574127075,
+      "grad_norm": 1.1669985055923462,
+      "learning_rate": 1.9954953968516262e-05,
+      "loss": 1.2109,
+      "step": 263
+    },
+    {
+      "epoch": 0.060446479679450486,
+      "grad_norm": 1.5981719493865967,
+      "learning_rate": 1.9954247914174828e-05,
+      "loss": 1.1865,
+      "step": 264
+    },
+    {
+      "epoch": 0.06067544361763022,
+      "grad_norm": 1.8978960514068604,
+      "learning_rate": 1.9953536382094218e-05,
+      "loss": 1.2389,
+      "step": 265
+    },
+    {
+      "epoch": 0.06090440755580996,
+      "grad_norm": 1.3212120532989502,
+      "learning_rate": 1.995281937266599e-05,
+      "loss": 1.1844,
+      "step": 266
+    },
+    {
+      "epoch": 0.061133371493989695,
+      "grad_norm": 2.165839672088623,
+      "learning_rate": 1.995209688628471e-05,
+      "loss": 1.196,
+      "step": 267
+    },
+    {
+      "epoch": 0.061362335432169436,
+      "grad_norm": 1.4239095449447632,
+      "learning_rate": 1.9951368923347945e-05,
+      "loss": 1.177,
+      "step": 268
+    },
+    {
+      "epoch": 0.06159129937034917,
+      "grad_norm": 1.4673502445220947,
+      "learning_rate": 1.9950635484256296e-05,
+      "loss": 1.2372,
+      "step": 269
+    },
+    {
+      "epoch": 0.061820263308528904,
+      "grad_norm": 1.2616180181503296,
+      "learning_rate": 1.9949896569413368e-05,
+      "loss": 1.1906,
+      "step": 270
+    },
+    {
+      "epoch": 0.062049227246708645,
+      "grad_norm": 1.8740971088409424,
+      "learning_rate": 1.9949152179225776e-05,
+      "loss": 1.2036,
+      "step": 271
+    },
+    {
+      "epoch": 0.06227819118488838,
+      "grad_norm": 1.2659952640533447,
+      "learning_rate": 1.9948402314103153e-05,
+      "loss": 1.1836,
+      "step": 272
+    },
+    {
+      "epoch": 0.06250715512306812,
+      "grad_norm": 1.446509599685669,
+      "learning_rate": 1.994764697445815e-05,
+      "loss": 1.2553,
+      "step": 273
+    },
+    {
+      "epoch": 0.06273611906124785,
+      "grad_norm": 1.2037794589996338,
+      "learning_rate": 1.994688616070641e-05,
+      "loss": 1.209,
+      "step": 274
+    },
+    {
+      "epoch": 0.06296508299942759,
+      "grad_norm": 1.1829689741134644,
+      "learning_rate": 1.9946119873266615e-05,
+      "loss": 1.192,
+      "step": 275
+    },
+    {
+      "epoch": 0.06319404693760733,
+      "grad_norm": 1.409515619277954,
+      "learning_rate": 1.994534811256044e-05,
+      "loss": 1.2247,
+      "step": 276
+    },
+    {
+      "epoch": 0.06342301087578707,
+      "grad_norm": 1.238846778869629,
+      "learning_rate": 1.994457087901258e-05,
+      "loss": 1.2204,
+      "step": 277
+    },
+    {
+      "epoch": 0.0636519748139668,
+      "grad_norm": 1.5566067695617676,
+      "learning_rate": 1.9943788173050743e-05,
+      "loss": 1.2295,
+      "step": 278
+    },
+    {
+      "epoch": 0.06388093875214654,
+      "grad_norm": 1.221211552619934,
+      "learning_rate": 1.9942999995105646e-05,
+      "loss": 1.1664,
+      "step": 279
+    },
+    {
+      "epoch": 0.06410990269032628,
+      "grad_norm": 1.3222095966339111,
+      "learning_rate": 1.9942206345611008e-05,
+      "loss": 1.1554,
+      "step": 280
+    },
+    {
+      "epoch": 0.064338866628506,
+      "grad_norm": 1.4909206628799438,
+      "learning_rate": 1.9941407225003577e-05,
+      "loss": 1.2614,
+      "step": 281
+    },
+    {
+      "epoch": 0.06456783056668575,
+      "grad_norm": 1.290727972984314,
+      "learning_rate": 1.9940602633723097e-05,
+      "loss": 1.2267,
+      "step": 282
+    },
+    {
+      "epoch": 0.06479679450486549,
+      "grad_norm": 2.3527672290802,
+      "learning_rate": 1.9939792572212328e-05,
+      "loss": 1.1742,
+      "step": 283
+    },
+    {
+      "epoch": 0.06502575844304521,
+      "grad_norm": 1.4631099700927734,
+      "learning_rate": 1.993897704091705e-05,
+      "loss": 1.2321,
+      "step": 284
+    },
+    {
+      "epoch": 0.06525472238122496,
+      "grad_norm": 1.4212102890014648,
+      "learning_rate": 1.9938156040286027e-05,
+      "loss": 1.208,
+      "step": 285
+    },
+    {
+      "epoch": 0.0654836863194047,
+      "grad_norm": 1.9231419563293457,
+      "learning_rate": 1.9937329570771057e-05,
+      "loss": 1.2423,
+      "step": 286
+    },
+    {
+      "epoch": 0.06571265025758444,
+      "grad_norm": 2.0737898349761963,
+      "learning_rate": 1.9936497632826937e-05,
+      "loss": 1.1687,
+      "step": 287
+    },
+    {
+      "epoch": 0.06594161419576416,
+      "grad_norm": 1.1904573440551758,
+      "learning_rate": 1.993566022691148e-05,
+      "loss": 1.2161,
+      "step": 288
+    },
+    {
+      "epoch": 0.0661705781339439,
+      "grad_norm": 1.472386121749878,
+      "learning_rate": 1.99348173534855e-05,
+      "loss": 1.2219,
+      "step": 289
+    },
+    {
+      "epoch": 0.06639954207212365,
+      "grad_norm": 2.0884242057800293,
+      "learning_rate": 1.993396901301283e-05,
+      "loss": 1.1996,
+      "step": 290
+    },
+    {
+      "epoch": 0.06662850601030337,
+      "grad_norm": 1.3736237287521362,
+      "learning_rate": 1.9933115205960295e-05,
+      "loss": 1.1683,
+      "step": 291
+    },
+    {
+      "epoch": 0.06685746994848311,
+      "grad_norm": 1.3148783445358276,
+      "learning_rate": 1.9932255932797736e-05,
+      "loss": 1.2628,
+      "step": 292
+    },
+    {
+      "epoch": 0.06708643388666286,
+      "grad_norm": 1.330783486366272,
+      "learning_rate": 1.9931391193998015e-05,
+      "loss": 1.1524,
+      "step": 293
+    },
+    {
+      "epoch": 0.06731539782484258,
+      "grad_norm": 1.5354028940200806,
+      "learning_rate": 1.9930520990036984e-05,
+      "loss": 1.2563,
+      "step": 294
+    },
+    {
+      "epoch": 0.06754436176302232,
+      "grad_norm": 1.3578846454620361,
+      "learning_rate": 1.9929645321393505e-05,
+      "loss": 1.2097,
+      "step": 295
+    },
+    {
+      "epoch": 0.06777332570120206,
+      "grad_norm": 1.1557707786560059,
+      "learning_rate": 1.9928764188549462e-05,
+      "loss": 1.1626,
+      "step": 296
+    },
+    {
+      "epoch": 0.06800228963938179,
+      "grad_norm": 1.2463425397872925,
+      "learning_rate": 1.9927877591989727e-05,
+      "loss": 1.2083,
+      "step": 297
+    },
+    {
+      "epoch": 0.06823125357756153,
+      "grad_norm": 1.1633163690567017,
+      "learning_rate": 1.992698553220219e-05,
+      "loss": 1.1593,
+      "step": 298
+    },
+    {
+      "epoch": 0.06846021751574127,
+      "grad_norm": 1.3011326789855957,
+      "learning_rate": 1.992608800967774e-05,
+      "loss": 1.2601,
+      "step": 299
+    },
+    {
+      "epoch": 0.06868918145392101,
+      "grad_norm": 1.7210752964019775,
+      "learning_rate": 1.992518502491028e-05,
+      "loss": 1.2187,
+      "step": 300
+    },
+    {
+      "epoch": 0.06891814539210074,
+      "grad_norm": 1.1964160203933716,
+      "learning_rate": 1.992427657839671e-05,
+      "loss": 1.1478,
+      "step": 301
+    },
+    {
+      "epoch": 0.06914710933028048,
+      "grad_norm": 1.2960009574890137,
+      "learning_rate": 1.992336267063695e-05,
+      "loss": 1.207,
+      "step": 302
+    },
+    {
+      "epoch": 0.06937607326846022,
+      "grad_norm": 1.6698633432388306,
+      "learning_rate": 1.9922443302133906e-05,
+      "loss": 1.133,
+      "step": 303
+    },
+    {
+      "epoch": 0.06960503720663995,
+      "grad_norm": 1.4826613664627075,
+      "learning_rate": 1.9921518473393502e-05,
+      "loss": 1.1586,
+      "step": 304
+    },
+    {
+      "epoch": 0.06983400114481969,
+      "grad_norm": 1.3968387842178345,
+      "learning_rate": 1.9920588184924664e-05,
+      "loss": 1.1621,
+      "step": 305
+    },
+    {
+      "epoch": 0.07006296508299943,
+      "grad_norm": 1.6237884759902954,
+      "learning_rate": 1.9919652437239326e-05,
+      "loss": 1.2013,
+      "step": 306
+    },
+    {
+      "epoch": 0.07029192902117916,
+      "grad_norm": 1.5827672481536865,
+      "learning_rate": 1.9918711230852416e-05,
+      "loss": 1.1804,
+      "step": 307
+    },
+    {
+      "epoch": 0.0705208929593589,
+      "grad_norm": 3.0790855884552,
+      "learning_rate": 1.9917764566281874e-05,
+      "loss": 1.1535,
+      "step": 308
+    },
+    {
+      "epoch": 0.07074985689753864,
+      "grad_norm": 1.3645802736282349,
+      "learning_rate": 1.9916812444048642e-05,
+      "loss": 1.1941,
+      "step": 309
+    },
+    {
+      "epoch": 0.07097882083571837,
+      "grad_norm": 1.1226625442504883,
+      "learning_rate": 1.9915854864676665e-05,
+      "loss": 1.1721,
+      "step": 310
+    },
+    {
+      "epoch": 0.07120778477389811,
+      "grad_norm": 1.386254906654358,
+      "learning_rate": 1.991489182869289e-05,
+      "loss": 1.0778,
+      "step": 311
+    },
+    {
+      "epoch": 0.07143674871207785,
+      "grad_norm": 1.1469475030899048,
+      "learning_rate": 1.9913923336627267e-05,
+      "loss": 1.1466,
+      "step": 312
+    },
+    {
+      "epoch": 0.07166571265025759,
+      "grad_norm": 1.7655390501022339,
+      "learning_rate": 1.9912949389012754e-05,
+      "loss": 1.1651,
+      "step": 313
+    },
+    {
+      "epoch": 0.07189467658843732,
+      "grad_norm": 1.2459030151367188,
+      "learning_rate": 1.9911969986385297e-05,
+      "loss": 1.2548,
+      "step": 314
+    },
+    {
+      "epoch": 0.07212364052661706,
+      "grad_norm": 1.4673657417297363,
+      "learning_rate": 1.991098512928386e-05,
+      "loss": 1.1818,
+      "step": 315
+    },
+    {
+      "epoch": 0.0723526044647968,
+      "grad_norm": 1.3704054355621338,
+      "learning_rate": 1.9909994818250403e-05,
+      "loss": 1.1223,
+      "step": 316
+    },
+    {
+      "epoch": 0.07258156840297653,
+      "grad_norm": 1.2631467580795288,
+      "learning_rate": 1.990899905382988e-05,
+      "loss": 1.1135,
+      "step": 317
+    },
+    {
+      "epoch": 0.07281053234115627,
+      "grad_norm": 1.255850076675415,
+      "learning_rate": 1.990799783657026e-05,
+      "loss": 1.1675,
+      "step": 318
+    },
+    {
+      "epoch": 0.07303949627933601,
+      "grad_norm": 1.2794235944747925,
+      "learning_rate": 1.9906991167022496e-05,
+      "loss": 1.145,
+      "step": 319
+    },
+    {
+      "epoch": 0.07326846021751574,
+      "grad_norm": 1.3548914194107056,
+      "learning_rate": 1.990597904574055e-05,
+      "loss": 1.2148,
+      "step": 320
+    },
+    {
+      "epoch": 0.07349742415569548,
+      "grad_norm": 1.7542753219604492,
+      "learning_rate": 1.990496147328139e-05,
+      "loss": 1.164,
+      "step": 321
+    },
+    {
+      "epoch": 0.07372638809387522,
+      "grad_norm": 1.3537834882736206,
+      "learning_rate": 1.9903938450204972e-05,
+      "loss": 1.2505,
+      "step": 322
+    },
+    {
+      "epoch": 0.07395535203205496,
+      "grad_norm": 1.2525663375854492,
+      "learning_rate": 1.9902909977074267e-05,
+      "loss": 1.2464,
+      "step": 323
+    },
+    {
+      "epoch": 0.07418431597023468,
+      "grad_norm": 1.4234391450881958,
+      "learning_rate": 1.9901876054455217e-05,
+      "loss": 1.1412,
+      "step": 324
+    },
+    {
+      "epoch": 0.07441327990841443,
+      "grad_norm": 1.5655121803283691,
+      "learning_rate": 1.9900836682916796e-05,
+      "loss": 1.2074,
+      "step": 325
+    },
+    {
+      "epoch": 0.07464224384659417,
+      "grad_norm": 1.317477822303772,
+      "learning_rate": 1.989979186303096e-05,
+      "loss": 1.1611,
+      "step": 326
+    },
+    {
+      "epoch": 0.0748712077847739,
+      "grad_norm": 1.2287671566009521,
+      "learning_rate": 1.989874159537266e-05,
+      "loss": 1.1862,
+      "step": 327
+    },
+    {
+      "epoch": 0.07510017172295363,
+      "grad_norm": 1.3680412769317627,
+      "learning_rate": 1.989768588051985e-05,
+      "loss": 1.0958,
+      "step": 328
+    },
+    {
+      "epoch": 0.07532913566113338,
+      "grad_norm": 1.3635417222976685,
+      "learning_rate": 1.9896624719053483e-05,
+      "loss": 1.1807,
+      "step": 329
+    },
+    {
+      "epoch": 0.0755580995993131,
+      "grad_norm": 1.7830880880355835,
+      "learning_rate": 1.9895558111557503e-05,
+      "loss": 1.2397,
+      "step": 330
+    },
+    {
+      "epoch": 0.07578706353749284,
+      "grad_norm": 1.3236726522445679,
+      "learning_rate": 1.9894486058618863e-05,
+      "loss": 1.1608,
+      "step": 331
+    },
+    {
+      "epoch": 0.07601602747567258,
+      "grad_norm": 1.733843207359314,
+      "learning_rate": 1.98934085608275e-05,
+      "loss": 1.2268,
+      "step": 332
+    },
+    {
+      "epoch": 0.07624499141385231,
+      "grad_norm": 1.5161333084106445,
+      "learning_rate": 1.9892325618776353e-05,
+      "loss": 1.2225,
+      "step": 333
+    },
+    {
+      "epoch": 0.07647395535203205,
+      "grad_norm": 1.2248483896255493,
+      "learning_rate": 1.9891237233061354e-05,
+      "loss": 1.1838,
+      "step": 334
+    },
+    {
+      "epoch": 0.0767029192902118,
+      "grad_norm": 1.2949497699737549,
+      "learning_rate": 1.989014340428143e-05,
+      "loss": 1.1763,
+      "step": 335
+    },
+    {
+      "epoch": 0.07693188322839153,
+      "grad_norm": 1.6322613954544067,
+      "learning_rate": 1.9889044133038514e-05,
+      "loss": 1.1773,
+      "step": 336
+    },
+    {
+      "epoch": 0.07716084716657126,
+      "grad_norm": 1.5641744136810303,
+      "learning_rate": 1.9887939419937522e-05,
+      "loss": 1.2079,
+      "step": 337
+    },
+    {
+      "epoch": 0.077389811104751,
+      "grad_norm": 1.1563526391983032,
+      "learning_rate": 1.9886829265586368e-05,
+      "loss": 1.1775,
+      "step": 338
+    },
+    {
+      "epoch": 0.07761877504293074,
+      "grad_norm": 1.9111676216125488,
+      "learning_rate": 1.9885713670595958e-05,
+      "loss": 1.1931,
+      "step": 339
+    },
+    {
+      "epoch": 0.07784773898111047,
+      "grad_norm": 1.3635084629058838,
+      "learning_rate": 1.98845926355802e-05,
+      "loss": 1.207,
+      "step": 340
+    },
+    {
+      "epoch": 0.07807670291929021,
+      "grad_norm": 1.4956107139587402,
+      "learning_rate": 1.988346616115598e-05,
+      "loss": 1.1679,
+      "step": 341
+    },
+    {
+      "epoch": 0.07830566685746995,
+      "grad_norm": 1.2649390697479248,
+      "learning_rate": 1.98823342479432e-05,
+      "loss": 1.1901,
+      "step": 342
+    },
+    {
+      "epoch": 0.07853463079564968,
+      "grad_norm": 1.5861189365386963,
+      "learning_rate": 1.9881196896564735e-05,
+      "loss": 1.2169,
+      "step": 343
+    },
+    {
+      "epoch": 0.07876359473382942,
+      "grad_norm": 1.2659194469451904,
+      "learning_rate": 1.9880054107646467e-05,
+      "loss": 1.2246,
+      "step": 344
+    },
+    {
+      "epoch": 0.07899255867200916,
+      "grad_norm": 1.5740238428115845,
+      "learning_rate": 1.9878905881817254e-05,
+      "loss": 1.2331,
+      "step": 345
+    },
+    {
+      "epoch": 0.0792215226101889,
+      "grad_norm": 1.9337862730026245,
+      "learning_rate": 1.9877752219708956e-05,
+      "loss": 1.1526,
+      "step": 346
+    },
+    {
+      "epoch": 0.07945048654836863,
+      "grad_norm": 1.5918827056884766,
+      "learning_rate": 1.987659312195643e-05,
+      "loss": 1.1238,
+      "step": 347
+    },
+    {
+      "epoch": 0.07967945048654837,
+      "grad_norm": 1.6437872648239136,
+      "learning_rate": 1.9875428589197513e-05,
+      "loss": 1.1845,
+      "step": 348
+    },
+    {
+      "epoch": 0.07990841442472811,
+      "grad_norm": 1.2831597328186035,
+      "learning_rate": 1.9874258622073044e-05,
+      "loss": 1.1319,
+      "step": 349
+    },
+    {
+      "epoch": 0.08013737836290784,
+      "grad_norm": 1.4037307500839233,
+      "learning_rate": 1.9873083221226833e-05,
+      "loss": 1.2522,
+      "step": 350
+    },
+    {
+      "epoch": 0.08036634230108758,
+      "grad_norm": 1.400461196899414,
+      "learning_rate": 1.9871902387305707e-05,
+      "loss": 1.1294,
+      "step": 351
+    },
+    {
+      "epoch": 0.08059530623926732,
+      "grad_norm": 1.286853551864624,
+      "learning_rate": 1.9870716120959462e-05,
+      "loss": 1.1933,
+      "step": 352
+    },
+    {
+      "epoch": 0.08082427017744705,
+      "grad_norm": 1.2215043306350708,
+      "learning_rate": 1.9869524422840893e-05,
+      "loss": 1.1226,
+      "step": 353
+    },
+    {
+      "epoch": 0.08105323411562679,
+      "grad_norm": 1.3354159593582153,
+      "learning_rate": 1.9868327293605778e-05,
+      "loss": 1.1322,
+      "step": 354
+    },
+    {
+      "epoch": 0.08128219805380653,
+      "grad_norm": 1.2361743450164795,
+      "learning_rate": 1.986712473391289e-05,
+      "loss": 1.2118,
+      "step": 355
+    },
+    {
+      "epoch": 0.08151116199198626,
+      "grad_norm": 1.4514862298965454,
+      "learning_rate": 1.986591674442399e-05,
+      "loss": 1.1469,
+      "step": 356
+    },
+    {
+      "epoch": 0.081740125930166,
+      "grad_norm": 1.2745472192764282,
+      "learning_rate": 1.9864703325803818e-05,
+      "loss": 1.1403,
+      "step": 357
+    },
+    {
+      "epoch": 0.08196908986834574,
+      "grad_norm": 1.3737826347351074,
+      "learning_rate": 1.986348447872011e-05,
+      "loss": 1.1907,
+      "step": 358
+    },
+    {
+      "epoch": 0.08219805380652548,
+      "grad_norm": 1.354296326637268,
+      "learning_rate": 1.986226020384359e-05,
+      "loss": 1.1332,
+      "step": 359
+    },
+    {
+      "epoch": 0.0824270177447052,
+      "grad_norm": 1.3757176399230957,
+      "learning_rate": 1.986103050184797e-05,
+      "loss": 1.2131,
+      "step": 360
+    },
+    {
+      "epoch": 0.08265598168288495,
+      "grad_norm": 1.770670771598816,
+      "learning_rate": 1.9859795373409934e-05,
+      "loss": 1.1724,
+      "step": 361
+    },
+    {
+      "epoch": 0.08288494562106469,
+      "grad_norm": 1.3382035493850708,
+      "learning_rate": 1.985855481920917e-05,
+      "loss": 1.1411,
+      "step": 362
+    },
+    {
+      "epoch": 0.08311390955924441,
+      "grad_norm": 1.1877765655517578,
+      "learning_rate": 1.9857308839928346e-05,
+      "loss": 1.2451,
+      "step": 363
+    },
+    {
+      "epoch": 0.08334287349742416,
+      "grad_norm": 1.3735177516937256,
+      "learning_rate": 1.9856057436253105e-05,
+      "loss": 1.1821,
+      "step": 364
+    },
+    {
+      "epoch": 0.0835718374356039,
+      "grad_norm": 1.5288448333740234,
+      "learning_rate": 1.9854800608872096e-05,
+      "loss": 1.1408,
+      "step": 365
+    },
+    {
+      "epoch": 0.08380080137378362,
+      "grad_norm": 1.8656703233718872,
+      "learning_rate": 1.9853538358476933e-05,
+      "loss": 1.2307,
+      "step": 366
+    },
+    {
+      "epoch": 0.08402976531196336,
+      "grad_norm": 1.520883560180664,
+      "learning_rate": 1.9852270685762222e-05,
+      "loss": 1.1503,
+      "step": 367
+    },
+    {
+      "epoch": 0.0842587292501431,
+      "grad_norm": 1.1496034860610962,
+      "learning_rate": 1.9850997591425555e-05,
+      "loss": 1.2023,
+      "step": 368
+    },
+    {
+      "epoch": 0.08448769318832285,
+      "grad_norm": 1.2435383796691895,
+      "learning_rate": 1.9849719076167502e-05,
+      "loss": 1.1091,
+      "step": 369
+    },
+    {
+      "epoch": 0.08471665712650257,
+      "grad_norm": 1.5105100870132446,
+      "learning_rate": 1.9848435140691627e-05,
+      "loss": 1.2194,
+      "step": 370
+    },
+    {
+      "epoch": 0.08494562106468231,
+      "grad_norm": 1.2073237895965576,
+      "learning_rate": 1.9847145785704457e-05,
+      "loss": 1.2124,
+      "step": 371
+    },
+    {
+      "epoch": 0.08517458500286205,
+      "grad_norm": 1.3399255275726318,
+      "learning_rate": 1.9845851011915526e-05,
+      "loss": 1.2134,
+      "step": 372
+    },
+    {
+      "epoch": 0.08540354894104178,
+      "grad_norm": 1.1257169246673584,
+      "learning_rate": 1.9844550820037326e-05,
+      "loss": 1.2316,
+      "step": 373
+    },
+    {
+      "epoch": 0.08563251287922152,
+      "grad_norm": 1.2916021347045898,
+      "learning_rate": 1.9843245210785348e-05,
+      "loss": 1.1766,
+      "step": 374
+    },
+    {
+      "epoch": 0.08586147681740126,
+      "grad_norm": 1.1885099411010742,
+      "learning_rate": 1.9841934184878056e-05,
+      "loss": 1.1737,
+      "step": 375
+    },
+    {
+      "epoch": 0.08609044075558099,
+      "grad_norm": 1.5768221616744995,
+      "learning_rate": 1.98406177430369e-05,
+      "loss": 1.1579,
+      "step": 376
+    },
+    {
+      "epoch": 0.08631940469376073,
+      "grad_norm": 1.2692902088165283,
+      "learning_rate": 1.98392958859863e-05,
+      "loss": 1.1498,
+      "step": 377
+    },
+    {
+      "epoch": 0.08654836863194047,
+      "grad_norm": 1.360911250114441,
+      "learning_rate": 1.9837968614453666e-05,
+      "loss": 1.1672,
+      "step": 378
+    },
+    {
+      "epoch": 0.0867773325701202,
+      "grad_norm": 1.3675583600997925,
+      "learning_rate": 1.9836635929169388e-05,
+      "loss": 1.2173,
+      "step": 379
+    },
+    {
+      "epoch": 0.08700629650829994,
+      "grad_norm": 1.1969804763793945,
+      "learning_rate": 1.9835297830866827e-05,
+      "loss": 1.1914,
+      "step": 380
+    },
+    {
+      "epoch": 0.08723526044647968,
+      "grad_norm": 25.4526424407959,
+      "learning_rate": 1.983395432028233e-05,
+      "loss": 1.2027,
+      "step": 381
+    },
+    {
+      "epoch": 0.08746422438465942,
+      "grad_norm": 1.52403724193573,
+      "learning_rate": 1.9832605398155217e-05,
+      "loss": 1.2035,
+      "step": 382
+    },
+    {
+      "epoch": 0.08769318832283915,
+      "grad_norm": 4.724262237548828,
+      "learning_rate": 1.9831251065227792e-05,
+      "loss": 1.2138,
+      "step": 383
+    },
+    {
+      "epoch": 0.08792215226101889,
+      "grad_norm": 2.621134042739868,
+      "learning_rate": 1.9829891322245326e-05,
+      "loss": 1.1642,
+      "step": 384
+    },
+    {
+      "epoch": 0.08815111619919863,
+      "grad_norm": 3.4685866832733154,
+      "learning_rate": 1.9828526169956083e-05,
+      "loss": 1.1791,
+      "step": 385
+    },
+    {
+      "epoch": 0.08838008013737836,
+      "grad_norm": 1.3899956941604614,
+      "learning_rate": 1.982715560911129e-05,
+      "loss": 1.1847,
+      "step": 386
+    },
+    {
+      "epoch": 0.0886090440755581,
+      "grad_norm": 1.1521110534667969,
+      "learning_rate": 1.9825779640465157e-05,
+      "loss": 1.129,
+      "step": 387
+    },
+    {
+      "epoch": 0.08883800801373784,
+      "grad_norm": 1.7473269701004028,
+      "learning_rate": 1.982439826477487e-05,
+      "loss": 1.1534,
+      "step": 388
+    },
+    {
+      "epoch": 0.08906697195191757,
+      "grad_norm": 1.4069769382476807,
+      "learning_rate": 1.9823011482800584e-05,
+      "loss": 1.1092,
+      "step": 389
+    },
+    {
+      "epoch": 0.08929593589009731,
+      "grad_norm": 1.2800990343093872,
+      "learning_rate": 1.9821619295305432e-05,
+      "loss": 1.1823,
+      "step": 390
+    },
+    {
+      "epoch": 0.08952489982827705,
+      "grad_norm": 1.3700530529022217,
+      "learning_rate": 1.9820221703055528e-05,
+      "loss": 1.1921,
+      "step": 391
+    },
+    {
+      "epoch": 0.08975386376645678,
+      "grad_norm": 1.9944264888763428,
+      "learning_rate": 1.9818818706819955e-05,
+      "loss": 1.1447,
+      "step": 392
+    },
+    {
+      "epoch": 0.08998282770463652,
+      "grad_norm": 2.0315797328948975,
+      "learning_rate": 1.981741030737077e-05,
+      "loss": 1.1634,
+      "step": 393
+    },
+    {
+      "epoch": 0.09021179164281626,
+      "grad_norm": 1.4205044507980347,
+      "learning_rate": 1.9815996505483e-05,
+      "loss": 1.1817,
+      "step": 394
+    },
+    {
+      "epoch": 0.090440755580996,
+      "grad_norm": 1.4898089170455933,
+      "learning_rate": 1.9814577301934647e-05,
+      "loss": 1.1931,
+      "step": 395
+    },
+    {
+      "epoch": 0.09066971951917573,
+      "grad_norm": 1.5351992845535278,
+      "learning_rate": 1.9813152697506696e-05,
+      "loss": 1.2307,
+      "step": 396
+    },
+    {
+      "epoch": 0.09089868345735547,
+      "grad_norm": 1.2297886610031128,
+      "learning_rate": 1.9811722692983088e-05,
+      "loss": 1.161,
+      "step": 397
+    },
+    {
+      "epoch": 0.09112764739553521,
+      "grad_norm": 1.5338091850280762,
+      "learning_rate": 1.981028728915074e-05,
+      "loss": 1.2151,
+      "step": 398
+    },
+    {
+      "epoch": 0.09135661133371493,
+      "grad_norm": 1.4196423292160034,
+      "learning_rate": 1.980884648679955e-05,
+      "loss": 1.2312,
+      "step": 399
+    },
+    {
+      "epoch": 0.09158557527189468,
+      "grad_norm": 1.2989675998687744,
+      "learning_rate": 1.980740028672237e-05,
+      "loss": 1.1532,
+      "step": 400
+    },
+    {
+      "epoch": 0.09181453921007442,
+      "grad_norm": 1.5549927949905396,
+      "learning_rate": 1.9805948689715043e-05,
+      "loss": 1.1157,
+      "step": 401
+    },
+    {
+      "epoch": 0.09204350314825414,
+      "grad_norm": 1.2734893560409546,
+      "learning_rate": 1.9804491696576364e-05,
+      "loss": 1.1663,
+      "step": 402
+    },
+    {
+      "epoch": 0.09227246708643388,
+      "grad_norm": 1.376017689704895,
+      "learning_rate": 1.9803029308108105e-05,
+      "loss": 1.1519,
+      "step": 403
+    },
+    {
+      "epoch": 0.09250143102461363,
+      "grad_norm": 1.1765146255493164,
+      "learning_rate": 1.9801561525115006e-05,
+      "loss": 1.1376,
+      "step": 404
+    },
+    {
+      "epoch": 0.09273039496279337,
+      "grad_norm": 1.1429637670516968,
+      "learning_rate": 1.9800088348404778e-05,
+      "loss": 1.1182,
+      "step": 405
+    },
+    {
+      "epoch": 0.0929593589009731,
+      "grad_norm": 1.3974339962005615,
+      "learning_rate": 1.9798609778788094e-05,
+      "loss": 1.141,
+      "step": 406
+    },
+    {
+      "epoch": 0.09318832283915283,
+      "grad_norm": 1.5140388011932373,
+      "learning_rate": 1.97971258170786e-05,
+      "loss": 1.2011,
+      "step": 407
+    },
+    {
+      "epoch": 0.09341728677733258,
+      "grad_norm": 1.161309838294983,
+      "learning_rate": 1.979563646409291e-05,
+      "loss": 1.1788,
+      "step": 408
+    },
+    {
+      "epoch": 0.0936462507155123,
+      "grad_norm": 1.2854355573654175,
+      "learning_rate": 1.97941417206506e-05,
+      "loss": 1.2321,
+      "step": 409
+    },
+    {
+      "epoch": 0.09387521465369204,
+      "grad_norm": 1.1982879638671875,
+      "learning_rate": 1.9792641587574212e-05,
+      "loss": 1.1316,
+      "step": 410
+    },
+    {
+      "epoch": 0.09410417859187178,
+      "grad_norm": 1.1869289875030518,
+      "learning_rate": 1.979113606568926e-05,
+      "loss": 1.1207,
+      "step": 411
+    },
+    {
+      "epoch": 0.09433314253005151,
+      "grad_norm": 1.159303069114685,
+      "learning_rate": 1.9789625155824226e-05,
+      "loss": 1.0941,
+      "step": 412
+    },
+    {
+      "epoch": 0.09456210646823125,
+      "grad_norm": 1.4857593774795532,
+      "learning_rate": 1.978810885881054e-05,
+      "loss": 1.1209,
+      "step": 413
+    },
+    {
+      "epoch": 0.094791070406411,
+      "grad_norm": 1.3553754091262817,
+      "learning_rate": 1.9786587175482613e-05,
+      "loss": 1.154,
+      "step": 414
+    },
+    {
+      "epoch": 0.09502003434459072,
+      "grad_norm": 1.219914197921753,
+      "learning_rate": 1.9785060106677818e-05,
+      "loss": 1.1627,
+      "step": 415
+    },
+    {
+      "epoch": 0.09524899828277046,
+      "grad_norm": 2.246405839920044,
+      "learning_rate": 1.978352765323648e-05,
+      "loss": 1.2182,
+      "step": 416
+    },
+    {
+      "epoch": 0.0954779622209502,
+      "grad_norm": 1.3677849769592285,
+      "learning_rate": 1.97819898160019e-05,
+      "loss": 1.1858,
+      "step": 417
+    },
+    {
+      "epoch": 0.09570692615912994,
+      "grad_norm": 1.52046537399292,
+      "learning_rate": 1.9780446595820336e-05,
+      "loss": 1.2089,
+      "step": 418
+    },
+    {
+      "epoch": 0.09593589009730967,
+      "grad_norm": 1.4131478071212769,
+      "learning_rate": 1.9778897993541014e-05,
+      "loss": 1.1796,
+      "step": 419
+    },
+    {
+      "epoch": 0.09616485403548941,
+      "grad_norm": 1.3519349098205566,
+      "learning_rate": 1.977734401001611e-05,
+      "loss": 1.148,
+      "step": 420
+    },
+    {
+      "epoch": 0.09639381797366915,
+      "grad_norm": 1.985721230506897,
+      "learning_rate": 1.9775784646100768e-05,
+      "loss": 1.1838,
+      "step": 421
+    },
+    {
+      "epoch": 0.09662278191184888,
+      "grad_norm": 1.2649965286254883,
+      "learning_rate": 1.97742199026531e-05,
+      "loss": 1.1589,
+      "step": 422
+    },
+    {
+      "epoch": 0.09685174585002862,
+      "grad_norm": 1.4421800374984741,
+      "learning_rate": 1.977264978053416e-05,
+      "loss": 1.0934,
+      "step": 423
+    },
+    {
+      "epoch": 0.09708070978820836,
+      "grad_norm": 1.287441611289978,
+      "learning_rate": 1.977107428060799e-05,
+      "loss": 1.1881,
+      "step": 424
+    },
+    {
+      "epoch": 0.09730967372638809,
+      "grad_norm": 1.2831951379776,
+      "learning_rate": 1.9769493403741556e-05,
+      "loss": 1.1283,
+      "step": 425
+    },
+    {
+      "epoch": 0.09753863766456783,
+      "grad_norm": 1.2015918493270874,
+      "learning_rate": 1.976790715080481e-05,
+      "loss": 1.1961,
+      "step": 426
+    },
+    {
+      "epoch": 0.09776760160274757,
+      "grad_norm": 1.4825067520141602,
+      "learning_rate": 1.9766315522670654e-05,
+      "loss": 1.1451,
+      "step": 427
+    },
+    {
+      "epoch": 0.09799656554092731,
+      "grad_norm": 1.2135714292526245,
+      "learning_rate": 1.976471852021495e-05,
+      "loss": 1.0926,
+      "step": 428
+    },
+    {
+      "epoch": 0.09822552947910704,
+      "grad_norm": 1.2095484733581543,
+      "learning_rate": 1.9763116144316506e-05,
+      "loss": 1.1423,
+      "step": 429
+    },
+    {
+      "epoch": 0.09845449341728678,
+      "grad_norm": 1.321748971939087,
+      "learning_rate": 1.9761508395857106e-05,
+      "loss": 1.1583,
+      "step": 430
+    },
+    {
+      "epoch": 0.09868345735546652,
+      "grad_norm": 1.3187049627304077,
+      "learning_rate": 1.9759895275721476e-05,
+      "loss": 1.1745,
+      "step": 431
+    },
+    {
+      "epoch": 0.09891242129364625,
+      "grad_norm": 1.3241872787475586,
+      "learning_rate": 1.97582767847973e-05,
+      "loss": 1.1689,
+      "step": 432
+    },
+    {
+      "epoch": 0.09914138523182599,
+      "grad_norm": 1.2542831897735596,
+      "learning_rate": 1.9756652923975227e-05,
+      "loss": 1.2044,
+      "step": 433
+    },
+    {
+      "epoch": 0.09937034917000573,
+      "grad_norm": 1.2562075853347778,
+      "learning_rate": 1.9755023694148846e-05,
+      "loss": 1.1629,
+      "step": 434
+    },
+    {
+      "epoch": 0.09959931310818546,
+      "grad_norm": 1.2366596460342407,
+      "learning_rate": 1.9753389096214716e-05,
+      "loss": 1.0803,
+      "step": 435
+    },
+    {
+      "epoch": 0.0998282770463652,
+      "grad_norm": 1.4391214847564697,
+      "learning_rate": 1.9751749131072335e-05,
+      "loss": 1.1797,
+      "step": 436
+    },
+    {
+      "epoch": 0.10005724098454494,
+      "grad_norm": 4.435413360595703,
+      "learning_rate": 1.9750103799624165e-05,
+      "loss": 1.1591,
+      "step": 437
+    },
+    {
+      "epoch": 0.10028620492272466,
+      "grad_norm": 1.376836895942688,
+      "learning_rate": 1.974845310277562e-05,
+      "loss": 1.2096,
+      "step": 438
+    },
+    {
+      "epoch": 0.1005151688609044,
+      "grad_norm": 1.4392584562301636,
+      "learning_rate": 1.974679704143507e-05,
+      "loss": 1.0871,
+      "step": 439
+    },
+    {
+      "epoch": 0.10074413279908415,
+      "grad_norm": 1.1275047063827515,
+      "learning_rate": 1.974513561651382e-05,
+      "loss": 1.1411,
+      "step": 440
+    },
+    {
+      "epoch": 0.10097309673726389,
+      "grad_norm": 1.4672117233276367,
+      "learning_rate": 1.974346882892614e-05,
+      "loss": 1.138,
+      "step": 441
+    },
+    {
+      "epoch": 0.10120206067544361,
+      "grad_norm": 1.2267554998397827,
+      "learning_rate": 1.974179667958926e-05,
+      "loss": 1.1537,
+      "step": 442
+    },
+    {
+      "epoch": 0.10143102461362336,
+      "grad_norm": 1.3913313150405884,
+      "learning_rate": 1.9740119169423337e-05,
+      "loss": 1.1623,
+      "step": 443
+    },
+    {
+      "epoch": 0.1016599885518031,
+      "grad_norm": 1.6974395513534546,
+      "learning_rate": 1.97384362993515e-05,
+      "loss": 1.208,
+      "step": 444
+    },
+    {
+      "epoch": 0.10188895248998282,
+      "grad_norm": 1.3473076820373535,
+      "learning_rate": 1.9736748070299813e-05,
+      "loss": 1.1479,
+      "step": 445
+    },
+    {
+      "epoch": 0.10211791642816256,
+      "grad_norm": 1.2066441774368286,
+      "learning_rate": 1.9735054483197295e-05,
+      "loss": 1.2317,
+      "step": 446
+    },
+    {
+      "epoch": 0.1023468803663423,
+      "grad_norm": 1.5624324083328247,
+      "learning_rate": 1.973335553897591e-05,
+      "loss": 1.1282,
+      "step": 447
+    },
+    {
+      "epoch": 0.10257584430452203,
+      "grad_norm": 1.9326655864715576,
+      "learning_rate": 1.9731651238570582e-05,
+      "loss": 1.1848,
+      "step": 448
+    },
+    {
+      "epoch": 0.10280480824270177,
+      "grad_norm": 2.2776360511779785,
+      "learning_rate": 1.9729941582919168e-05,
+      "loss": 1.1599,
+      "step": 449
+    },
+    {
+      "epoch": 0.10303377218088151,
+      "grad_norm": 1.2282005548477173,
+      "learning_rate": 1.9728226572962474e-05,
+      "loss": 1.1843,
+      "step": 450
+    },
+    {
+      "epoch": 0.10326273611906125,
+      "grad_norm": 1.4309104681015015,
+      "learning_rate": 1.972650620964426e-05,
+      "loss": 1.1689,
+      "step": 451
+    },
+    {
+      "epoch": 0.10349170005724098,
+      "grad_norm": 1.427143931388855,
+      "learning_rate": 1.9724780493911227e-05,
+      "loss": 1.1471,
+      "step": 452
+    },
+    {
+      "epoch": 0.10372066399542072,
+      "grad_norm": 1.2180664539337158,
+      "learning_rate": 1.9723049426713018e-05,
+      "loss": 1.1932,
+      "step": 453
+    },
+    {
+      "epoch": 0.10394962793360046,
+      "grad_norm": 1.2000045776367188,
+      "learning_rate": 1.9721313009002228e-05,
+      "loss": 1.1751,
+      "step": 454
+    },
+    {
+      "epoch": 0.10417859187178019,
+      "grad_norm": 1.3412353992462158,
+      "learning_rate": 1.9719571241734395e-05,
+      "loss": 1.147,
+      "step": 455
+    },
+    {
+      "epoch": 0.10440755580995993,
+      "grad_norm": 1.2962464094161987,
+      "learning_rate": 1.9717824125867993e-05,
+      "loss": 1.2452,
+      "step": 456
+    },
+    {
+      "epoch": 0.10463651974813967,
+      "grad_norm": 1.10789155960083,
+      "learning_rate": 1.9716071662364454e-05,
+      "loss": 1.2014,
+      "step": 457
+    },
+    {
+      "epoch": 0.1048654836863194,
+      "grad_norm": 1.4773873090744019,
+      "learning_rate": 1.9714313852188133e-05,
+      "loss": 1.2092,
+      "step": 458
+    },
+    {
+      "epoch": 0.10509444762449914,
+      "grad_norm": 1.4391075372695923,
+      "learning_rate": 1.9712550696306346e-05,
+      "loss": 1.1431,
+      "step": 459
+    },
+    {
+      "epoch": 0.10532341156267888,
+      "grad_norm": 1.7549716234207153,
+      "learning_rate": 1.9710782195689343e-05,
+      "loss": 1.1395,
+      "step": 460
+    },
+    {
+      "epoch": 0.10555237550085861,
+      "grad_norm": 1.5657472610473633,
+      "learning_rate": 1.970900835131031e-05,
+      "loss": 1.132,
+      "step": 461
+    },
+    {
+      "epoch": 0.10578133943903835,
+      "grad_norm": 1.4908957481384277,
+      "learning_rate": 1.9707229164145386e-05,
+      "loss": 1.1533,
+      "step": 462
+    },
+    {
+      "epoch": 0.10601030337721809,
+      "grad_norm": 1.2966023683547974,
+      "learning_rate": 1.9705444635173635e-05,
+      "loss": 1.1855,
+      "step": 463
+    },
+    {
+      "epoch": 0.10623926731539783,
+      "grad_norm": 1.271470546722412,
+      "learning_rate": 1.970365476537707e-05,
+      "loss": 1.2443,
+      "step": 464
+    },
+    {
+      "epoch": 0.10646823125357756,
+      "grad_norm": 1.3924498558044434,
+      "learning_rate": 1.9701859555740647e-05,
+      "loss": 1.1831,
+      "step": 465
+    },
+    {
+      "epoch": 0.1066971951917573,
+      "grad_norm": 1.053683876991272,
+      "learning_rate": 1.9700059007252248e-05,
+      "loss": 1.2135,
+      "step": 466
+    },
+    {
+      "epoch": 0.10692615912993704,
+      "grad_norm": 1.3220330476760864,
+      "learning_rate": 1.9698253120902703e-05,
+      "loss": 1.1772,
+      "step": 467
+    },
+    {
+      "epoch": 0.10715512306811677,
+      "grad_norm": 1.528494119644165,
+      "learning_rate": 1.9696441897685777e-05,
+      "loss": 1.2416,
+      "step": 468
+    },
+    {
+      "epoch": 0.10738408700629651,
+      "grad_norm": 1.3539191484451294,
+      "learning_rate": 1.969462533859817e-05,
+      "loss": 1.1354,
+      "step": 469
+    },
+    {
+      "epoch": 0.10761305094447625,
+      "grad_norm": 1.5194259881973267,
+      "learning_rate": 1.9692803444639517e-05,
+      "loss": 1.208,
+      "step": 470
+    },
+    {
+      "epoch": 0.10784201488265598,
+      "grad_norm": 1.2938241958618164,
+      "learning_rate": 1.9690976216812397e-05,
+      "loss": 1.0596,
+      "step": 471
+    },
+    {
+      "epoch": 0.10807097882083572,
+      "grad_norm": 1.2351398468017578,
+      "learning_rate": 1.968914365612231e-05,
+      "loss": 1.2107,
+      "step": 472
+    },
+    {
+      "epoch": 0.10829994275901546,
+      "grad_norm": 1.381752371788025,
+      "learning_rate": 1.9687305763577705e-05,
+      "loss": 1.1274,
+      "step": 473
+    },
+    {
+      "epoch": 0.10852890669719518,
+      "grad_norm": 1.4342790842056274,
+      "learning_rate": 1.9685462540189955e-05,
+      "loss": 1.1483,
+      "step": 474
+    },
+    {
+      "epoch": 0.10875787063537493,
+      "grad_norm": 1.2768388986587524,
+      "learning_rate": 1.9683613986973373e-05,
+      "loss": 1.2039,
+      "step": 475
+    },
+    {
+      "epoch": 0.10898683457355467,
+      "grad_norm": 1.2173471450805664,
+      "learning_rate": 1.9681760104945203e-05,
+      "loss": 1.1969,
+      "step": 476
+    },
+    {
+      "epoch": 0.10921579851173441,
+      "grad_norm": 1.2570933103561401,
+      "learning_rate": 1.967990089512562e-05,
+      "loss": 1.1701,
+      "step": 477
+    },
+    {
+      "epoch": 0.10944476244991413,
+      "grad_norm": 1.45657479763031,
+      "learning_rate": 1.9678036358537726e-05,
+      "loss": 1.1106,
+      "step": 478
+    },
+    {
+      "epoch": 0.10967372638809388,
+      "grad_norm": 1.4095118045806885,
+      "learning_rate": 1.9676166496207567e-05,
+      "loss": 1.1978,
+      "step": 479
+    },
+    {
+      "epoch": 0.10990269032627362,
+      "grad_norm": 1.445336937904358,
+      "learning_rate": 1.967429130916411e-05,
+      "loss": 1.1609,
+      "step": 480
+    },
+    {
+      "epoch": 0.11013165426445334,
+      "grad_norm": 1.2025765180587769,
+      "learning_rate": 1.9672410798439256e-05,
+      "loss": 1.1806,
+      "step": 481
+    },
+    {
+      "epoch": 0.11036061820263308,
+      "grad_norm": 1.2341455221176147,
+      "learning_rate": 1.9670524965067832e-05,
+      "loss": 1.1765,
+      "step": 482
+    },
+    {
+      "epoch": 0.11058958214081283,
+      "grad_norm": 1.1721457242965698,
+      "learning_rate": 1.96686338100876e-05,
+      "loss": 1.1521,
+      "step": 483
+    },
+    {
+      "epoch": 0.11081854607899255,
+      "grad_norm": 1.2636065483093262,
+      "learning_rate": 1.9666737334539237e-05,
+      "loss": 1.1089,
+      "step": 484
+    },
+    {
+      "epoch": 0.1110475100171723,
+      "grad_norm": 1.7780942916870117,
+      "learning_rate": 1.966483553946637e-05,
+      "loss": 1.1039,
+      "step": 485
+    },
+    {
+      "epoch": 0.11127647395535203,
+      "grad_norm": 1.289139747619629,
+      "learning_rate": 1.9662928425915536e-05,
+      "loss": 1.2103,
+      "step": 486
+    },
+    {
+      "epoch": 0.11150543789353178,
+      "grad_norm": 1.5748257637023926,
+      "learning_rate": 1.9661015994936204e-05,
+      "loss": 1.1882,
+      "step": 487
+    },
+    {
+      "epoch": 0.1117344018317115,
+      "grad_norm": 1.3127802610397339,
+      "learning_rate": 1.9659098247580765e-05,
+      "loss": 1.0901,
+      "step": 488
+    },
+    {
+      "epoch": 0.11196336576989124,
+      "grad_norm": 1.2186063528060913,
+      "learning_rate": 1.9657175184904545e-05,
+      "loss": 1.1095,
+      "step": 489
+    },
+    {
+      "epoch": 0.11219232970807098,
+      "grad_norm": 1.644958734512329,
+      "learning_rate": 1.9655246807965786e-05,
+      "loss": 1.1162,
+      "step": 490
+    },
+    {
+      "epoch": 0.11242129364625071,
+      "grad_norm": 1.3639802932739258,
+      "learning_rate": 1.9653313117825657e-05,
+      "loss": 1.1439,
+      "step": 491
+    },
+    {
+      "epoch": 0.11265025758443045,
+      "grad_norm": 1.1354119777679443,
+      "learning_rate": 1.9651374115548255e-05,
+      "loss": 1.1214,
+      "step": 492
+    },
+    {
+      "epoch": 0.11287922152261019,
+      "grad_norm": 1.509704351425171,
+      "learning_rate": 1.964942980220059e-05,
+      "loss": 1.154,
+      "step": 493
+    },
+    {
+      "epoch": 0.11310818546078992,
+      "grad_norm": 1.605154037475586,
+      "learning_rate": 1.9647480178852606e-05,
+      "loss": 1.2066,
+      "step": 494
+    },
+    {
+      "epoch": 0.11333714939896966,
+      "grad_norm": 1.1039700508117676,
+      "learning_rate": 1.9645525246577168e-05,
+      "loss": 1.2157,
+      "step": 495
+    },
+    {
+      "epoch": 0.1135661133371494,
+      "grad_norm": 1.5542914867401123,
+      "learning_rate": 1.9643565006450055e-05,
+      "loss": 1.1465,
+      "step": 496
+    },
+    {
+      "epoch": 0.11379507727532913,
+      "grad_norm": 1.450024127960205,
+      "learning_rate": 1.9641599459549966e-05,
+      "loss": 1.1535,
+      "step": 497
+    },
+    {
+      "epoch": 0.11402404121350887,
+      "grad_norm": 1.6595662832260132,
+      "learning_rate": 1.9639628606958535e-05,
+      "loss": 1.1892,
+      "step": 498
+    },
+    {
+      "epoch": 0.11425300515168861,
+      "grad_norm": 1.6177374124526978,
+      "learning_rate": 1.9637652449760297e-05,
+      "loss": 1.114,
+      "step": 499
+    },
+    {
+      "epoch": 0.11448196908986835,
+      "grad_norm": 1.6932591199874878,
+      "learning_rate": 1.963567098904272e-05,
+      "loss": 1.1176,
+      "step": 500
+    },
+    {
+      "epoch": 0.11471093302804808,
+      "grad_norm": 1.1896140575408936,
+      "learning_rate": 1.963368422589618e-05,
+      "loss": 1.1573,
+      "step": 501
+    },
+    {
+      "epoch": 0.11493989696622782,
+      "grad_norm": 1.246862769126892,
+      "learning_rate": 1.9631692161413985e-05,
+      "loss": 1.2009,
+      "step": 502
+    },
+    {
+      "epoch": 0.11516886090440756,
+      "grad_norm": 1.422579288482666,
+      "learning_rate": 1.962969479669234e-05,
+      "loss": 1.1447,
+      "step": 503
+    },
+    {
+      "epoch": 0.11539782484258729,
+      "grad_norm": 1.2170311212539673,
+      "learning_rate": 1.962769213283039e-05,
+      "loss": 1.1696,
+      "step": 504
+    },
+    {
+      "epoch": 0.11562678878076703,
+      "grad_norm": 1.365478754043579,
+      "learning_rate": 1.9625684170930172e-05,
+      "loss": 1.1165,
+      "step": 505
+    },
+    {
+      "epoch": 0.11585575271894677,
+      "grad_norm": 1.4074455499649048,
+      "learning_rate": 1.9623670912096656e-05,
+      "loss": 1.1835,
+      "step": 506
+    },
+    {
+      "epoch": 0.1160847166571265,
+      "grad_norm": 1.3482747077941895,
+      "learning_rate": 1.9621652357437723e-05,
+      "loss": 1.1344,
+      "step": 507
+    },
+    {
+      "epoch": 0.11631368059530624,
+      "grad_norm": 1.7042218446731567,
+      "learning_rate": 1.961962850806417e-05,
+      "loss": 1.1432,
+      "step": 508
+    },
+    {
+      "epoch": 0.11654264453348598,
+      "grad_norm": 1.3767738342285156,
+      "learning_rate": 1.9617599365089693e-05,
+      "loss": 1.1334,
+      "step": 509
+    },
+    {
+      "epoch": 0.11677160847166572,
+      "grad_norm": 1.317917823791504,
+      "learning_rate": 1.9615564929630925e-05,
+      "loss": 1.1883,
+      "step": 510
+    },
+    {
+      "epoch": 0.11700057240984545,
+      "grad_norm": 1.2787598371505737,
+      "learning_rate": 1.9613525202807392e-05,
+      "loss": 1.1198,
+      "step": 511
+    },
+    {
+      "epoch": 0.11722953634802519,
+      "grad_norm": 1.6827375888824463,
+      "learning_rate": 1.961148018574154e-05,
+      "loss": 1.1579,
+      "step": 512
+    },
+    {
+      "epoch": 0.11745850028620493,
+      "grad_norm": 1.2459272146224976,
+      "learning_rate": 1.9609429879558726e-05,
+      "loss": 1.1789,
+      "step": 513
+    },
+    {
+      "epoch": 0.11768746422438466,
+      "grad_norm": 1.3571677207946777,
+      "learning_rate": 1.960737428538721e-05,
+      "loss": 1.139,
+      "step": 514
+    },
+    {
+      "epoch": 0.1179164281625644,
+      "grad_norm": 1.3986880779266357,
+      "learning_rate": 1.9605313404358176e-05,
+      "loss": 1.1719,
+      "step": 515
+    },
+    {
+      "epoch": 0.11814539210074414,
+      "grad_norm": 1.2968988418579102,
+      "learning_rate": 1.9603247237605706e-05,
+      "loss": 1.1753,
+      "step": 516
+    },
+    {
+      "epoch": 0.11837435603892386,
+      "grad_norm": 1.434787631034851,
+      "learning_rate": 1.9601175786266796e-05,
+      "loss": 1.1497,
+      "step": 517
+    },
+    {
+      "epoch": 0.1186033199771036,
+      "grad_norm": 1.595548152923584,
+      "learning_rate": 1.9599099051481345e-05,
+      "loss": 1.111,
+      "step": 518
+    },
+    {
+      "epoch": 0.11883228391528335,
+      "grad_norm": 1.3125005960464478,
+      "learning_rate": 1.959701703439217e-05,
+      "loss": 1.1453,
+      "step": 519
+    },
+    {
+      "epoch": 0.11906124785346307,
+      "grad_norm": 1.6637828350067139,
+      "learning_rate": 1.9594929736144978e-05,
+      "loss": 1.1657,
+      "step": 520
+    },
+    {
+      "epoch": 0.11929021179164281,
+      "grad_norm": 1.1127067804336548,
+      "learning_rate": 1.9592837157888396e-05,
+      "loss": 1.2167,
+      "step": 521
+    },
+    {
+      "epoch": 0.11951917572982255,
+      "grad_norm": 1.3453141450881958,
+      "learning_rate": 1.959073930077395e-05,
+      "loss": 1.1839,
+      "step": 522
+    },
+    {
+      "epoch": 0.1197481396680023,
+      "grad_norm": 1.2498703002929688,
+      "learning_rate": 1.958863616595608e-05,
+      "loss": 1.1712,
+      "step": 523
+    },
+    {
+      "epoch": 0.11997710360618202,
+      "grad_norm": 1.6796958446502686,
+      "learning_rate": 1.958652775459212e-05,
+      "loss": 1.1745,
+      "step": 524
+    },
+    {
+      "epoch": 0.12020606754436176,
+      "grad_norm": 1.1889232397079468,
+      "learning_rate": 1.9584414067842304e-05,
+      "loss": 1.1619,
+      "step": 525
+    },
+    {
+      "epoch": 0.1204350314825415,
+      "grad_norm": 1.389387607574463,
+      "learning_rate": 1.9582295106869788e-05,
+      "loss": 1.1106,
+      "step": 526
+    },
+    {
+      "epoch": 0.12066399542072123,
+      "grad_norm": 1.8008201122283936,
+      "learning_rate": 1.958017087284061e-05,
+      "loss": 1.1603,
+      "step": 527
+    },
+    {
+      "epoch": 0.12089295935890097,
+      "grad_norm": 1.4966647624969482,
+      "learning_rate": 1.9578041366923722e-05,
+      "loss": 1.1037,
+      "step": 528
+    },
+    {
+      "epoch": 0.12112192329708071,
+      "grad_norm": 1.1903491020202637,
+      "learning_rate": 1.957590659029097e-05,
+      "loss": 1.1838,
+      "step": 529
+    },
+    {
+      "epoch": 0.12135088723526044,
+      "grad_norm": 1.38507080078125,
+      "learning_rate": 1.957376654411711e-05,
+      "loss": 1.1818,
+      "step": 530
+    },
+    {
+      "epoch": 0.12157985117344018,
+      "grad_norm": 1.4420480728149414,
+      "learning_rate": 1.957162122957978e-05,
+      "loss": 1.1738,
+      "step": 531
+    },
+    {
+      "epoch": 0.12180881511161992,
+      "grad_norm": 1.5511441230773926,
+      "learning_rate": 1.9569470647859544e-05,
+      "loss": 1.183,
+      "step": 532
+    },
+    {
+      "epoch": 0.12203777904979966,
+      "grad_norm": 1.3438152074813843,
+      "learning_rate": 1.9567314800139838e-05,
+      "loss": 1.1897,
+      "step": 533
+    },
+    {
+      "epoch": 0.12226674298797939,
+      "grad_norm": 1.281836748123169,
+      "learning_rate": 1.9565153687607006e-05,
+      "loss": 1.1887,
+      "step": 534
+    },
+    {
+      "epoch": 0.12249570692615913,
+      "grad_norm": 1.3364907503128052,
+      "learning_rate": 1.9562987311450298e-05,
+      "loss": 1.1386,
+      "step": 535
+    },
+    {
+      "epoch": 0.12272467086433887,
+      "grad_norm": 1.3706281185150146,
+      "learning_rate": 1.956081567286185e-05,
+      "loss": 1.1379,
+      "step": 536
+    },
+    {
+      "epoch": 0.1229536348025186,
+      "grad_norm": 1.3901708126068115,
+      "learning_rate": 1.9558638773036694e-05,
+      "loss": 1.124,
+      "step": 537
+    },
+    {
+      "epoch": 0.12318259874069834,
+      "grad_norm": 1.1462208032608032,
+      "learning_rate": 1.955645661317276e-05,
+      "loss": 1.1561,
+      "step": 538
+    },
+    {
+      "epoch": 0.12341156267887808,
+      "grad_norm": 1.1496376991271973,
+      "learning_rate": 1.9554269194470872e-05,
+      "loss": 1.1725,
+      "step": 539
+    },
+    {
+      "epoch": 0.12364052661705781,
+      "grad_norm": 1.095733404159546,
+      "learning_rate": 1.955207651813475e-05,
+      "loss": 1.1946,
+      "step": 540
+    },
+    {
+      "epoch": 0.12386949055523755,
+      "grad_norm": 1.4160492420196533,
+      "learning_rate": 1.9549878585371006e-05,
+      "loss": 1.0487,
+      "step": 541
+    },
+    {
+      "epoch": 0.12409845449341729,
+      "grad_norm": 1.1869254112243652,
+      "learning_rate": 1.9547675397389144e-05,
+      "loss": 1.1571,
+      "step": 542
+    },
+    {
+      "epoch": 0.12432741843159702,
+      "grad_norm": 2.6226820945739746,
+      "learning_rate": 1.9545466955401555e-05,
+      "loss": 1.1698,
+      "step": 543
+    },
+    {
+      "epoch": 0.12455638236977676,
+      "grad_norm": 1.3678830862045288,
+      "learning_rate": 1.954325326062353e-05,
+      "loss": 1.1191,
+      "step": 544
+    },
+    {
+      "epoch": 0.1247853463079565,
+      "grad_norm": 1.2578891515731812,
+      "learning_rate": 1.954103431427325e-05,
+      "loss": 1.1159,
+      "step": 545
+    },
+    {
+      "epoch": 0.12501431024613624,
+      "grad_norm": 1.258697509765625,
+      "learning_rate": 1.9538810117571775e-05,
+      "loss": 1.1422,
+      "step": 546
+    },
+    {
+      "epoch": 0.12524327418431597,
+      "grad_norm": 1.3794115781784058,
+      "learning_rate": 1.9536580671743067e-05,
+      "loss": 1.1376,
+      "step": 547
+    },
+    {
+      "epoch": 0.1254722381224957,
+      "grad_norm": 1.2176522016525269,
+      "learning_rate": 1.9534345978013972e-05,
+      "loss": 1.1487,
+      "step": 548
+    },
+    {
+      "epoch": 0.12570120206067545,
+      "grad_norm": 1.7151718139648438,
+      "learning_rate": 1.9532106037614218e-05,
+      "loss": 1.1334,
+      "step": 549
+    },
+    {
+      "epoch": 0.12593016599885518,
+      "grad_norm": 1.7247742414474487,
+      "learning_rate": 1.952986085177643e-05,
+      "loss": 1.1493,
+      "step": 550
+    },
+    {
+      "epoch": 0.12615912993703493,
+      "grad_norm": 1.252971887588501,
+      "learning_rate": 1.9527610421736115e-05,
+      "loss": 1.1741,
+      "step": 551
+    },
+    {
+      "epoch": 0.12638809387521466,
+      "grad_norm": 1.716004729270935,
+      "learning_rate": 1.9525354748731665e-05,
+      "loss": 1.1638,
+      "step": 552
+    },
+    {
+      "epoch": 0.12661705781339438,
+      "grad_norm": 1.1862410306930542,
+      "learning_rate": 1.9523093834004358e-05,
+      "loss": 1.1476,
+      "step": 553
+    },
+    {
+      "epoch": 0.12684602175157414,
+      "grad_norm": 1.4751574993133545,
+      "learning_rate": 1.952082767879835e-05,
+      "loss": 1.1203,
+      "step": 554
+    },
+    {
+      "epoch": 0.12707498568975387,
+      "grad_norm": 1.6074382066726685,
+      "learning_rate": 1.9518556284360696e-05,
+      "loss": 1.1421,
+      "step": 555
+    },
+    {
+      "epoch": 0.1273039496279336,
+      "grad_norm": 1.5327030420303345,
+      "learning_rate": 1.9516279651941324e-05,
+      "loss": 1.1892,
+      "step": 556
+    },
+    {
+      "epoch": 0.12753291356611335,
+      "grad_norm": 1.531124472618103,
+      "learning_rate": 1.9513997782793045e-05,
+      "loss": 1.1348,
+      "step": 557
+    },
+    {
+      "epoch": 0.12776187750429308,
+      "grad_norm": 1.5858008861541748,
+      "learning_rate": 1.9511710678171548e-05,
+      "loss": 1.19,
+      "step": 558
+    },
+    {
+      "epoch": 0.1279908414424728,
+      "grad_norm": 1.3005167245864868,
+      "learning_rate": 1.9509418339335415e-05,
+      "loss": 1.1556,
+      "step": 559
+    },
+    {
+      "epoch": 0.12821980538065256,
+      "grad_norm": 1.6737267971038818,
+      "learning_rate": 1.950712076754609e-05,
+      "loss": 1.2304,
+      "step": 560
+    },
+    {
+      "epoch": 0.12844876931883228,
+      "grad_norm": 1.2671421766281128,
+      "learning_rate": 1.950481796406792e-05,
+      "loss": 1.1869,
+      "step": 561
+    },
+    {
+      "epoch": 0.128677733257012,
+      "grad_norm": 1.6968939304351807,
+      "learning_rate": 1.9502509930168113e-05,
+      "loss": 1.1252,
+      "step": 562
+    },
+    {
+      "epoch": 0.12890669719519177,
+      "grad_norm": 1.406265377998352,
+      "learning_rate": 1.950019666711676e-05,
+      "loss": 1.1424,
+      "step": 563
+    },
+    {
+      "epoch": 0.1291356611333715,
+      "grad_norm": 1.4663190841674805,
+      "learning_rate": 1.949787817618683e-05,
+      "loss": 1.1995,
+      "step": 564
+    },
+    {
+      "epoch": 0.12936462507155122,
+      "grad_norm": 1.5765471458435059,
+      "learning_rate": 1.949555445865417e-05,
+      "loss": 1.1302,
+      "step": 565
+    },
+    {
+      "epoch": 0.12959358900973098,
+      "grad_norm": 1.7354494333267212,
+      "learning_rate": 1.94932255157975e-05,
+      "loss": 1.1729,
+      "step": 566
+    },
+    {
+      "epoch": 0.1298225529479107,
+      "grad_norm": 1.379272699356079,
+      "learning_rate": 1.9490891348898423e-05,
+      "loss": 1.1507,
+      "step": 567
+    },
+    {
+      "epoch": 0.13005151688609043,
+      "grad_norm": 1.2708396911621094,
+      "learning_rate": 1.9488551959241405e-05,
+      "loss": 1.1516,
+      "step": 568
+    },
+    {
+      "epoch": 0.13028048082427018,
+      "grad_norm": 1.5942022800445557,
+      "learning_rate": 1.9486207348113803e-05,
+      "loss": 1.1358,
+      "step": 569
+    },
+    {
+      "epoch": 0.1305094447624499,
+      "grad_norm": 1.2558735609054565,
+      "learning_rate": 1.9483857516805823e-05,
+      "loss": 1.154,
+      "step": 570
+    },
+    {
+      "epoch": 0.13073840870062964,
+      "grad_norm": 1.1783373355865479,
+      "learning_rate": 1.9481502466610568e-05,
+      "loss": 1.1798,
+      "step": 571
+    },
+    {
+      "epoch": 0.1309673726388094,
+      "grad_norm": 1.2193893194198608,
+      "learning_rate": 1.9479142198823996e-05,
+      "loss": 1.1767,
+      "step": 572
+    },
+    {
+      "epoch": 0.13119633657698912,
+      "grad_norm": 1.1724305152893066,
+      "learning_rate": 1.9476776714744945e-05,
+      "loss": 1.113,
+      "step": 573
+    },
+    {
+      "epoch": 0.13142530051516887,
+      "grad_norm": 1.3040038347244263,
+      "learning_rate": 1.947440601567512e-05,
+      "loss": 1.1216,
+      "step": 574
+    },
+    {
+      "epoch": 0.1316542644533486,
+      "grad_norm": 1.3851133584976196,
+      "learning_rate": 1.9472030102919102e-05,
+      "loss": 1.1914,
+      "step": 575
+    },
+    {
+      "epoch": 0.13188322839152833,
+      "grad_norm": 1.331370234489441,
+      "learning_rate": 1.946964897778433e-05,
+      "loss": 1.135,
+      "step": 576
+    },
+    {
+      "epoch": 0.13211219232970808,
+      "grad_norm": 1.2417669296264648,
+      "learning_rate": 1.946726264158112e-05,
+      "loss": 1.0824,
+      "step": 577
+    },
+    {
+      "epoch": 0.1323411562678878,
+      "grad_norm": 1.1429842710494995,
+      "learning_rate": 1.9464871095622652e-05,
+      "loss": 1.1327,
+      "step": 578
+    },
+    {
+      "epoch": 0.13257012020606754,
+      "grad_norm": 1.2447996139526367,
+      "learning_rate": 1.9462474341224974e-05,
+      "loss": 1.1308,
+      "step": 579
+    },
+    {
+      "epoch": 0.1327990841442473,
+      "grad_norm": 1.3931182622909546,
+      "learning_rate": 1.9460072379706995e-05,
+      "loss": 1.1119,
+      "step": 580
+    },
+    {
+      "epoch": 0.13302804808242702,
+      "grad_norm": 1.3399399518966675,
+      "learning_rate": 1.9457665212390502e-05,
+      "loss": 1.1726,
+      "step": 581
+    },
+    {
+      "epoch": 0.13325701202060675,
+      "grad_norm": 1.1147905588150024,
+      "learning_rate": 1.9455252840600137e-05,
+      "loss": 1.0929,
+      "step": 582
+    },
+    {
+      "epoch": 0.1334859759587865,
+      "grad_norm": 1.4041277170181274,
+      "learning_rate": 1.9452835265663404e-05,
+      "loss": 1.1418,
+      "step": 583
+    },
+    {
+      "epoch": 0.13371493989696623,
+      "grad_norm": 1.5064622163772583,
+      "learning_rate": 1.945041248891068e-05,
+      "loss": 1.1835,
+      "step": 584
+    },
+    {
+      "epoch": 0.13394390383514596,
+      "grad_norm": 1.390039324760437,
+      "learning_rate": 1.9447984511675192e-05,
+      "loss": 1.089,
+      "step": 585
+    },
+    {
+      "epoch": 0.1341728677733257,
+      "grad_norm": 2.3285298347473145,
+      "learning_rate": 1.944555133529304e-05,
+      "loss": 1.1771,
+      "step": 586
+    },
+    {
+      "epoch": 0.13440183171150544,
+      "grad_norm": 1.5625582933425903,
+      "learning_rate": 1.9443112961103182e-05,
+      "loss": 1.1622,
+      "step": 587
+    },
+    {
+      "epoch": 0.13463079564968516,
+      "grad_norm": 1.6119787693023682,
+      "learning_rate": 1.9440669390447433e-05,
+      "loss": 1.1831,
+      "step": 588
+    },
+    {
+      "epoch": 0.13485975958786492,
+      "grad_norm": 1.1473219394683838,
+      "learning_rate": 1.943822062467047e-05,
+      "loss": 1.0934,
+      "step": 589
+    },
+    {
+      "epoch": 0.13508872352604465,
+      "grad_norm": 1.3341083526611328,
+      "learning_rate": 1.9435766665119823e-05,
+      "loss": 1.149,
+      "step": 590
+    },
+    {
+      "epoch": 0.13531768746422437,
+      "grad_norm": 1.3681514263153076,
+      "learning_rate": 1.943330751314589e-05,
+      "loss": 1.2103,
+      "step": 591
+    },
+    {
+      "epoch": 0.13554665140240413,
+      "grad_norm": 2.2560174465179443,
+      "learning_rate": 1.9430843170101924e-05,
+      "loss": 1.0948,
+      "step": 592
+    },
+    {
+      "epoch": 0.13577561534058386,
+      "grad_norm": 1.165104866027832,
+      "learning_rate": 1.9428373637344027e-05,
+      "loss": 1.1378,
+      "step": 593
+    },
+    {
+      "epoch": 0.13600457927876358,
+      "grad_norm": 1.2633001804351807,
+      "learning_rate": 1.9425898916231166e-05,
+      "loss": 1.0987,
+      "step": 594
+    },
+    {
+      "epoch": 0.13623354321694334,
+      "grad_norm": 1.2817325592041016,
+      "learning_rate": 1.942341900812516e-05,
+      "loss": 1.1535,
+      "step": 595
+    },
+    {
+      "epoch": 0.13646250715512306,
+      "grad_norm": 1.2196224927902222,
+      "learning_rate": 1.9420933914390677e-05,
+      "loss": 1.1257,
+      "step": 596
+    },
+    {
+      "epoch": 0.1366914710933028,
+      "grad_norm": 1.3916923999786377,
+      "learning_rate": 1.941844363639525e-05,
+      "loss": 1.1056,
+      "step": 597
+    },
+    {
+      "epoch": 0.13692043503148255,
+      "grad_norm": 1.571192741394043,
+      "learning_rate": 1.941594817550925e-05,
+      "loss": 1.1894,
+      "step": 598
+    },
+    {
+      "epoch": 0.13714939896966227,
+      "grad_norm": 1.3344855308532715,
+      "learning_rate": 1.9413447533105913e-05,
+      "loss": 1.1814,
+      "step": 599
+    },
+    {
+      "epoch": 0.13737836290784203,
+      "grad_norm": 2.034505605697632,
+      "learning_rate": 1.9410941710561323e-05,
+      "loss": 1.2126,
+      "step": 600
+    },
+    {
+      "epoch": 0.13760732684602175,
+      "grad_norm": 1.2162837982177734,
+      "learning_rate": 1.940843070925441e-05,
+      "loss": 1.1775,
+      "step": 601
+    },
+    {
+      "epoch": 0.13783629078420148,
+      "grad_norm": 1.2769267559051514,
+      "learning_rate": 1.940591453056696e-05,
+      "loss": 1.1882,
+      "step": 602
+    },
+    {
+      "epoch": 0.13806525472238124,
+      "grad_norm": 1.438686490058899,
+      "learning_rate": 1.9403393175883602e-05,
+      "loss": 1.1577,
+      "step": 603
+    },
+    {
+      "epoch": 0.13829421866056096,
+      "grad_norm": 1.3430330753326416,
+      "learning_rate": 1.9400866646591816e-05,
+      "loss": 1.1549,
+      "step": 604
+    },
+    {
+      "epoch": 0.1385231825987407,
+      "grad_norm": 1.135581374168396,
+      "learning_rate": 1.939833494408193e-05,
+      "loss": 1.1509,
+      "step": 605
+    },
+    {
+      "epoch": 0.13875214653692045,
+      "grad_norm": 1.433982014656067,
+      "learning_rate": 1.939579806974712e-05,
+      "loss": 1.1768,
+      "step": 606
+    },
+    {
+      "epoch": 0.13898111047510017,
+      "grad_norm": 1.200698971748352,
+      "learning_rate": 1.939325602498341e-05,
+      "loss": 1.1196,
+      "step": 607
+    },
+    {
+      "epoch": 0.1392100744132799,
+      "grad_norm": 1.2908475399017334,
+      "learning_rate": 1.939070881118966e-05,
+      "loss": 1.137,
+      "step": 608
+    },
+    {
+      "epoch": 0.13943903835145965,
+      "grad_norm": 1.3260984420776367,
+      "learning_rate": 1.9388156429767585e-05,
+      "loss": 1.1106,
+      "step": 609
+    },
+    {
+      "epoch": 0.13966800228963938,
+      "grad_norm": 1.310341238975525,
+      "learning_rate": 1.9385598882121735e-05,
+      "loss": 1.1857,
+      "step": 610
+    },
+    {
+      "epoch": 0.1398969662278191,
+      "grad_norm": 1.109005331993103,
+      "learning_rate": 1.9383036169659513e-05,
+      "loss": 1.1895,
+      "step": 611
+    },
+    {
+      "epoch": 0.14012593016599886,
+      "grad_norm": 1.230418086051941,
+      "learning_rate": 1.938046829379115e-05,
+      "loss": 1.1439,
+      "step": 612
+    },
+    {
+      "epoch": 0.1403548941041786,
+      "grad_norm": 1.380399227142334,
+      "learning_rate": 1.9377895255929734e-05,
+      "loss": 1.1821,
+      "step": 613
+    },
+    {
+      "epoch": 0.14058385804235832,
+      "grad_norm": 1.2389920949935913,
+      "learning_rate": 1.937531705749118e-05,
+      "loss": 1.1463,
+      "step": 614
+    },
+    {
+      "epoch": 0.14081282198053807,
+      "grad_norm": 1.1333352327346802,
+      "learning_rate": 1.9372733699894253e-05,
+      "loss": 1.1391,
+      "step": 615
+    },
+    {
+      "epoch": 0.1410417859187178,
+      "grad_norm": 1.2721819877624512,
+      "learning_rate": 1.937014518456055e-05,
+      "loss": 1.1266,
+      "step": 616
+    },
+    {
+      "epoch": 0.14127074985689753,
+      "grad_norm": 1.2748703956604004,
+      "learning_rate": 1.9367551512914513e-05,
+      "loss": 1.1351,
+      "step": 617
+    },
+    {
+      "epoch": 0.14149971379507728,
+      "grad_norm": 1.4715847969055176,
+      "learning_rate": 1.936495268638342e-05,
+      "loss": 1.1432,
+      "step": 618
+    },
+    {
+      "epoch": 0.141728677733257,
+      "grad_norm": 1.2434061765670776,
+      "learning_rate": 1.9362348706397374e-05,
+      "loss": 1.1471,
+      "step": 619
+    },
+    {
+      "epoch": 0.14195764167143674,
+      "grad_norm": 1.2716186046600342,
+      "learning_rate": 1.935973957438933e-05,
+      "loss": 1.1463,
+      "step": 620
+    },
+    {
+      "epoch": 0.1421866056096165,
+      "grad_norm": 1.4149022102355957,
+      "learning_rate": 1.935712529179507e-05,
+      "loss": 1.1119,
+      "step": 621
+    },
+    {
+      "epoch": 0.14241556954779622,
+      "grad_norm": 1.7188087701797485,
+      "learning_rate": 1.9354505860053215e-05,
+      "loss": 1.0757,
+      "step": 622
+    },
+    {
+      "epoch": 0.14264453348597597,
+      "grad_norm": 1.2718242406845093,
+      "learning_rate": 1.9351881280605212e-05,
+      "loss": 1.1599,
+      "step": 623
+    },
+    {
+      "epoch": 0.1428734974241557,
+      "grad_norm": 1.4921197891235352,
+      "learning_rate": 1.9349251554895347e-05,
+      "loss": 1.1168,
+      "step": 624
+    },
+    {
+      "epoch": 0.14310246136233543,
+      "grad_norm": 1.3648992776870728,
+      "learning_rate": 1.934661668437073e-05,
+      "loss": 1.0909,
+      "step": 625
+    },
+    {
+      "epoch": 0.14333142530051518,
+      "grad_norm": 1.2685736417770386,
+      "learning_rate": 1.934397667048132e-05,
+      "loss": 1.1921,
+      "step": 626
+    },
+    {
+      "epoch": 0.1435603892386949,
+      "grad_norm": 1.4009802341461182,
+      "learning_rate": 1.9341331514679887e-05,
+      "loss": 1.13,
+      "step": 627
+    },
+    {
+      "epoch": 0.14378935317687463,
+      "grad_norm": 2.8500163555145264,
+      "learning_rate": 1.933868121842204e-05,
+      "loss": 1.1513,
+      "step": 628
+    },
+    {
+      "epoch": 0.1440183171150544,
+      "grad_norm": 1.7703518867492676,
+      "learning_rate": 1.933602578316621e-05,
+      "loss": 1.1431,
+      "step": 629
+    },
+    {
+      "epoch": 0.14424728105323412,
+      "grad_norm": 1.2395981550216675,
+      "learning_rate": 1.9333365210373668e-05,
+      "loss": 1.1633,
+      "step": 630
+    },
+    {
+      "epoch": 0.14447624499141384,
+      "grad_norm": 1.3300714492797852,
+      "learning_rate": 1.9330699501508504e-05,
+      "loss": 1.1318,
+      "step": 631
+    },
+    {
+      "epoch": 0.1447052089295936,
+      "grad_norm": 1.4751075506210327,
+      "learning_rate": 1.932802865803763e-05,
+      "loss": 1.0868,
+      "step": 632
+    },
+    {
+      "epoch": 0.14493417286777333,
+      "grad_norm": 1.5677204132080078,
+      "learning_rate": 1.932535268143079e-05,
+      "loss": 1.2013,
+      "step": 633
+    },
+    {
+      "epoch": 0.14516313680595305,
+      "grad_norm": 1.3743131160736084,
+      "learning_rate": 1.9322671573160556e-05,
+      "loss": 1.1175,
+      "step": 634
+    },
+    {
+      "epoch": 0.1453921007441328,
+      "grad_norm": 1.6558480262756348,
+      "learning_rate": 1.9319985334702315e-05,
+      "loss": 1.1751,
+      "step": 635
+    },
+    {
+      "epoch": 0.14562106468231253,
+      "grad_norm": 1.3124221563339233,
+      "learning_rate": 1.9317293967534283e-05,
+      "loss": 1.1122,
+      "step": 636
+    },
+    {
+      "epoch": 0.14585002862049226,
+      "grad_norm": 1.2808423042297363,
+      "learning_rate": 1.9314597473137498e-05,
+      "loss": 1.1417,
+      "step": 637
+    },
+    {
+      "epoch": 0.14607899255867202,
+      "grad_norm": 1.250510334968567,
+      "learning_rate": 1.9311895852995815e-05,
+      "loss": 1.2279,
+      "step": 638
+    },
+    {
+      "epoch": 0.14630795649685174,
+      "grad_norm": 1.366873860359192,
+      "learning_rate": 1.930918910859592e-05,
+      "loss": 1.1324,
+      "step": 639
+    },
+    {
+      "epoch": 0.14653692043503147,
+      "grad_norm": 1.168998122215271,
+      "learning_rate": 1.9306477241427303e-05,
+      "loss": 1.0871,
+      "step": 640
+    },
+    {
+      "epoch": 0.14676588437321123,
+      "grad_norm": 1.3328088521957397,
+      "learning_rate": 1.9303760252982287e-05,
+      "loss": 1.1846,
+      "step": 641
+    },
+    {
+      "epoch": 0.14699484831139095,
+      "grad_norm": 1.1964495182037354,
+      "learning_rate": 1.9301038144756007e-05,
+      "loss": 1.1037,
+      "step": 642
+    },
+    {
+      "epoch": 0.14722381224957068,
+      "grad_norm": 1.2240676879882812,
+      "learning_rate": 1.9298310918246414e-05,
+      "loss": 1.1069,
+      "step": 643
+    },
+    {
+      "epoch": 0.14745277618775043,
+      "grad_norm": 1.3975204229354858,
+      "learning_rate": 1.929557857495428e-05,
+      "loss": 1.1145,
+      "step": 644
+    },
+    {
+      "epoch": 0.14768174012593016,
+      "grad_norm": 1.3508495092391968,
+      "learning_rate": 1.9292841116383192e-05,
+      "loss": 1.1929,
+      "step": 645
+    },
+    {
+      "epoch": 0.14791070406410992,
+      "grad_norm": 1.5650644302368164,
+      "learning_rate": 1.9290098544039546e-05,
+      "loss": 1.1214,
+      "step": 646
+    },
+    {
+      "epoch": 0.14813966800228964,
+      "grad_norm": 1.208836555480957,
+      "learning_rate": 1.9287350859432562e-05,
+      "loss": 1.11,
+      "step": 647
+    },
+    {
+      "epoch": 0.14836863194046937,
+      "grad_norm": 1.2838915586471558,
+      "learning_rate": 1.9284598064074264e-05,
+      "loss": 1.172,
+      "step": 648
+    },
+    {
+      "epoch": 0.14859759587864912,
+      "grad_norm": 1.2721766233444214,
+      "learning_rate": 1.9281840159479494e-05,
+      "loss": 1.1024,
+      "step": 649
+    },
+    {
+      "epoch": 0.14882655981682885,
+      "grad_norm": 1.6511443853378296,
+      "learning_rate": 1.9279077147165903e-05,
+      "loss": 1.14,
+      "step": 650
+    },
+    {
+      "epoch": 0.14905552375500858,
+      "grad_norm": 1.3828959465026855,
+      "learning_rate": 1.9276309028653954e-05,
+      "loss": 1.1819,
+      "step": 651
+    },
+    {
+      "epoch": 0.14928448769318833,
+      "grad_norm": 1.3660774230957031,
+      "learning_rate": 1.927353580546692e-05,
+      "loss": 1.1487,
+      "step": 652
+    },
+    {
+      "epoch": 0.14951345163136806,
+      "grad_norm": 1.296252965927124,
+      "learning_rate": 1.927075747913088e-05,
+      "loss": 1.1442,
+      "step": 653
+    },
+    {
+      "epoch": 0.1497424155695478,
+      "grad_norm": 1.1136078834533691,
+      "learning_rate": 1.9267974051174727e-05,
+      "loss": 1.0958,
+      "step": 654
+    },
+    {
+      "epoch": 0.14997137950772754,
+      "grad_norm": 1.5701806545257568,
+      "learning_rate": 1.9265185523130156e-05,
+      "loss": 1.1934,
+      "step": 655
+    },
+    {
+      "epoch": 0.15020034344590727,
+      "grad_norm": 1.439436912536621,
+      "learning_rate": 1.9262391896531668e-05,
+      "loss": 1.1336,
+      "step": 656
+    },
+    {
+      "epoch": 0.150429307384087,
+      "grad_norm": 1.2246439456939697,
+      "learning_rate": 1.9259593172916577e-05,
+      "loss": 1.1292,
+      "step": 657
+    },
+    {
+      "epoch": 0.15065827132226675,
+      "grad_norm": 1.5522711277008057,
+      "learning_rate": 1.9256789353824996e-05,
+      "loss": 1.1826,
+      "step": 658
+    },
+    {
+      "epoch": 0.15088723526044648,
+      "grad_norm": 1.6734440326690674,
+      "learning_rate": 1.9253980440799843e-05,
+      "loss": 1.0926,
+      "step": 659
+    },
+    {
+      "epoch": 0.1511161991986262,
+      "grad_norm": 1.276139736175537,
+      "learning_rate": 1.9251166435386837e-05,
+      "loss": 1.1941,
+      "step": 660
+    },
+    {
+      "epoch": 0.15134516313680596,
+      "grad_norm": 1.4880852699279785,
+      "learning_rate": 1.9248347339134508e-05,
+      "loss": 1.1254,
+      "step": 661
+    },
+    {
+      "epoch": 0.1515741270749857,
+      "grad_norm": 1.467468500137329,
+      "learning_rate": 1.9245523153594172e-05,
+      "loss": 1.1781,
+      "step": 662
+    },
+    {
+      "epoch": 0.15180309101316541,
+      "grad_norm": 1.3000531196594238,
+      "learning_rate": 1.9242693880319962e-05,
+      "loss": 1.0907,
+      "step": 663
+    },
+    {
+      "epoch": 0.15203205495134517,
+      "grad_norm": 1.569050908088684,
+      "learning_rate": 1.92398595208688e-05,
+      "loss": 1.2279,
+      "step": 664
+    },
+    {
+      "epoch": 0.1522610188895249,
+      "grad_norm": 1.3132572174072266,
+      "learning_rate": 1.9237020076800407e-05,
+      "loss": 1.0786,
+      "step": 665
+    },
+    {
+      "epoch": 0.15248998282770462,
+      "grad_norm": 1.4111486673355103,
+      "learning_rate": 1.9234175549677313e-05,
+      "loss": 1.1556,
+      "step": 666
+    },
+    {
+      "epoch": 0.15271894676588438,
+      "grad_norm": 1.2233771085739136,
+      "learning_rate": 1.923132594106483e-05,
+      "loss": 1.138,
+      "step": 667
+    },
+    {
+      "epoch": 0.1529479107040641,
+      "grad_norm": 1.5644503831863403,
+      "learning_rate": 1.922847125253108e-05,
+      "loss": 1.1911,
+      "step": 668
+    },
+    {
+      "epoch": 0.15317687464224386,
+      "grad_norm": 1.8687934875488281,
+      "learning_rate": 1.922561148564697e-05,
+      "loss": 1.1343,
+      "step": 669
+    },
+    {
+      "epoch": 0.1534058385804236,
+      "grad_norm": 1.1981557607650757,
+      "learning_rate": 1.9222746641986207e-05,
+      "loss": 1.1846,
+      "step": 670
+    },
+    {
+      "epoch": 0.15363480251860331,
+      "grad_norm": 1.6271294355392456,
+      "learning_rate": 1.921987672312529e-05,
+      "loss": 1.167,
+      "step": 671
+    },
+    {
+      "epoch": 0.15386376645678307,
+      "grad_norm": 1.3530622720718384,
+      "learning_rate": 1.9217001730643514e-05,
+      "loss": 1.1491,
+      "step": 672
+    },
+    {
+      "epoch": 0.1540927303949628,
+      "grad_norm": 1.456268072128296,
+      "learning_rate": 1.921412166612296e-05,
+      "loss": 1.1595,
+      "step": 673
+    },
+    {
+      "epoch": 0.15432169433314252,
+      "grad_norm": 1.2416306734085083,
+      "learning_rate": 1.92112365311485e-05,
+      "loss": 1.1396,
+      "step": 674
+    },
+    {
+      "epoch": 0.15455065827132228,
+      "grad_norm": 1.4196218252182007,
+      "learning_rate": 1.920834632730781e-05,
+      "loss": 1.1493,
+      "step": 675
+    },
+    {
+      "epoch": 0.154779622209502,
+      "grad_norm": 1.69566810131073,
+      "learning_rate": 1.920545105619134e-05,
+      "loss": 1.0956,
+      "step": 676
+    },
+    {
+      "epoch": 0.15500858614768173,
+      "grad_norm": 1.6241673231124878,
+      "learning_rate": 1.9202550719392324e-05,
+      "loss": 1.1241,
+      "step": 677
+    },
+    {
+      "epoch": 0.1552375500858615,
+      "grad_norm": 1.2214878797531128,
+      "learning_rate": 1.919964531850681e-05,
+      "loss": 1.1751,
+      "step": 678
+    },
+    {
+      "epoch": 0.1554665140240412,
+      "grad_norm": 2.9216532707214355,
+      "learning_rate": 1.9196734855133603e-05,
+      "loss": 1.1578,
+      "step": 679
+    },
+    {
+      "epoch": 0.15569547796222094,
+      "grad_norm": 1.8954651355743408,
+      "learning_rate": 1.919381933087431e-05,
+      "loss": 1.1785,
+      "step": 680
+    },
+    {
+      "epoch": 0.1559244419004007,
+      "grad_norm": 1.424324870109558,
+      "learning_rate": 1.919089874733332e-05,
+      "loss": 1.0811,
+      "step": 681
+    },
+    {
+      "epoch": 0.15615340583858042,
+      "grad_norm": 1.3659218549728394,
+      "learning_rate": 1.9187973106117808e-05,
+      "loss": 1.1433,
+      "step": 682
+    },
+    {
+      "epoch": 0.15638236977676015,
+      "grad_norm": 1.2462931871414185,
+      "learning_rate": 1.918504240883773e-05,
+      "loss": 1.1522,
+      "step": 683
+    },
+    {
+      "epoch": 0.1566113337149399,
+      "grad_norm": 1.390007495880127,
+      "learning_rate": 1.9182106657105816e-05,
+      "loss": 1.1282,
+      "step": 684
+    },
+    {
+      "epoch": 0.15684029765311963,
+      "grad_norm": 1.1688913106918335,
+      "learning_rate": 1.9179165852537596e-05,
+      "loss": 1.1549,
+      "step": 685
+    },
+    {
+      "epoch": 0.15706926159129936,
+      "grad_norm": 1.6128971576690674,
+      "learning_rate": 1.917621999675136e-05,
+      "loss": 1.1311,
+      "step": 686
+    },
+    {
+      "epoch": 0.1572982255294791,
+      "grad_norm": 1.283479928970337,
+      "learning_rate": 1.9173269091368197e-05,
+      "loss": 1.109,
+      "step": 687
+    },
+    {
+      "epoch": 0.15752718946765884,
+      "grad_norm": 1.3000928163528442,
+      "learning_rate": 1.9170313138011964e-05,
+      "loss": 1.1634,
+      "step": 688
+    },
+    {
+      "epoch": 0.15775615340583857,
+      "grad_norm": 1.204128384590149,
+      "learning_rate": 1.9167352138309294e-05,
+      "loss": 1.0988,
+      "step": 689
+    },
+    {
+      "epoch": 0.15798511734401832,
+      "grad_norm": 1.3165777921676636,
+      "learning_rate": 1.9164386093889598e-05,
+      "loss": 1.1393,
+      "step": 690
+    },
+    {
+      "epoch": 0.15821408128219805,
+      "grad_norm": 1.433110237121582,
+      "learning_rate": 1.9161415006385074e-05,
+      "loss": 1.1621,
+      "step": 691
+    },
+    {
+      "epoch": 0.1584430452203778,
+      "grad_norm": 1.2257025241851807,
+      "learning_rate": 1.915843887743068e-05,
+      "loss": 1.1692,
+      "step": 692
+    },
+    {
+      "epoch": 0.15867200915855753,
+      "grad_norm": 1.773209571838379,
+      "learning_rate": 1.915545770866415e-05,
+      "loss": 1.1176,
+      "step": 693
+    },
+    {
+      "epoch": 0.15890097309673726,
+      "grad_norm": 1.3696967363357544,
+      "learning_rate": 1.9152471501726008e-05,
+      "loss": 1.1944,
+      "step": 694
+    },
+    {
+      "epoch": 0.159129937034917,
+      "grad_norm": 1.1072070598602295,
+      "learning_rate": 1.9149480258259535e-05,
+      "loss": 1.1889,
+      "step": 695
+    },
+    {
+      "epoch": 0.15935890097309674,
+      "grad_norm": 1.5009992122650146,
+      "learning_rate": 1.914648397991078e-05,
+      "loss": 1.125,
+      "step": 696
+    },
+    {
+      "epoch": 0.15958786491127647,
+      "grad_norm": 1.4128434658050537,
+      "learning_rate": 1.9143482668328577e-05,
+      "loss": 1.1271,
+      "step": 697
+    },
+    {
+      "epoch": 0.15981682884945622,
+      "grad_norm": 1.0920000076293945,
+      "learning_rate": 1.9140476325164522e-05,
+      "loss": 1.1275,
+      "step": 698
+    },
+    {
+      "epoch": 0.16004579278763595,
+      "grad_norm": 1.100367546081543,
+      "learning_rate": 1.9137464952072977e-05,
+      "loss": 1.1063,
+      "step": 699
+    },
+    {
+      "epoch": 0.16027475672581568,
+      "grad_norm": 1.1449311971664429,
+      "learning_rate": 1.9134448550711077e-05,
+      "loss": 1.1795,
+      "step": 700
+    },
+    {
+      "epoch": 0.16050372066399543,
+      "grad_norm": 1.3512089252471924,
+      "learning_rate": 1.9131427122738725e-05,
+      "loss": 1.0984,
+      "step": 701
+    },
+    {
+      "epoch": 0.16073268460217516,
+      "grad_norm": 1.5637792348861694,
+      "learning_rate": 1.9128400669818586e-05,
+      "loss": 1.106,
+      "step": 702
+    },
+    {
+      "epoch": 0.16096164854035488,
+      "grad_norm": 1.6217029094696045,
+      "learning_rate": 1.9125369193616085e-05,
+      "loss": 1.1792,
+      "step": 703
+    },
+    {
+      "epoch": 0.16119061247853464,
+      "grad_norm": 1.6088597774505615,
+      "learning_rate": 1.9122332695799432e-05,
+      "loss": 1.1579,
+      "step": 704
+    },
+    {
+      "epoch": 0.16141957641671437,
+      "grad_norm": 1.4255467653274536,
+      "learning_rate": 1.9119291178039573e-05,
+      "loss": 1.1689,
+      "step": 705
+    },
+    {
+      "epoch": 0.1616485403548941,
+      "grad_norm": 2.3765337467193604,
+      "learning_rate": 1.9116244642010234e-05,
+      "loss": 1.2181,
+      "step": 706
+    },
+    {
+      "epoch": 0.16187750429307385,
+      "grad_norm": 1.2266517877578735,
+      "learning_rate": 1.91131930893879e-05,
+      "loss": 1.1672,
+      "step": 707
+    },
+    {
+      "epoch": 0.16210646823125358,
+      "grad_norm": 1.379189372062683,
+      "learning_rate": 1.9110136521851815e-05,
+      "loss": 1.1442,
+      "step": 708
+    },
+    {
+      "epoch": 0.1623354321694333,
+      "grad_norm": 1.462998628616333,
+      "learning_rate": 1.9107074941083987e-05,
+      "loss": 1.1641,
+      "step": 709
+    },
+    {
+      "epoch": 0.16256439610761306,
+      "grad_norm": 1.2941198348999023,
+      "learning_rate": 1.9104008348769164e-05,
+      "loss": 1.1731,
+      "step": 710
+    },
+    {
+      "epoch": 0.16279336004579278,
+      "grad_norm": 1.5538750886917114,
+      "learning_rate": 1.9100936746594878e-05,
+      "loss": 1.0923,
+      "step": 711
+    },
+    {
+      "epoch": 0.1630223239839725,
+      "grad_norm": 1.0926445722579956,
+      "learning_rate": 1.9097860136251402e-05,
+      "loss": 1.2141,
+      "step": 712
+    },
+    {
+      "epoch": 0.16325128792215227,
+      "grad_norm": 1.3583141565322876,
+      "learning_rate": 1.9094778519431773e-05,
+      "loss": 1.1745,
+      "step": 713
+    },
+    {
+      "epoch": 0.163480251860332,
+      "grad_norm": 1.200218915939331,
+      "learning_rate": 1.9091691897831774e-05,
+      "loss": 1.1723,
+      "step": 714
+    },
+    {
+      "epoch": 0.16370921579851175,
+      "grad_norm": 1.428288221359253,
+      "learning_rate": 1.9088600273149947e-05,
+      "loss": 1.1924,
+      "step": 715
+    },
+    {
+      "epoch": 0.16393817973669148,
+      "grad_norm": 1.488163709640503,
+      "learning_rate": 1.9085503647087588e-05,
+      "loss": 1.1357,
+      "step": 716
+    },
+    {
+      "epoch": 0.1641671436748712,
+      "grad_norm": 1.282083511352539,
+      "learning_rate": 1.9082402021348745e-05,
+      "loss": 1.1102,
+      "step": 717
+    },
+    {
+      "epoch": 0.16439610761305096,
+      "grad_norm": 1.33743155002594,
+      "learning_rate": 1.9079295397640215e-05,
+      "loss": 1.2115,
+      "step": 718
+    },
+    {
+      "epoch": 0.16462507155123068,
+      "grad_norm": 1.618739128112793,
+      "learning_rate": 1.9076183777671553e-05,
+      "loss": 1.1347,
+      "step": 719
+    },
+    {
+      "epoch": 0.1648540354894104,
+      "grad_norm": 2.2287635803222656,
+      "learning_rate": 1.907306716315505e-05,
+      "loss": 1.1554,
+      "step": 720
+    },
+    {
+      "epoch": 0.16508299942759017,
+      "grad_norm": 1.8814674615859985,
+      "learning_rate": 1.9069945555805756e-05,
+      "loss": 1.1277,
+      "step": 721
+    },
+    {
+      "epoch": 0.1653119633657699,
+      "grad_norm": 1.54645836353302,
+      "learning_rate": 1.9066818957341463e-05,
+      "loss": 1.1175,
+      "step": 722
+    },
+    {
+      "epoch": 0.16554092730394962,
+      "grad_norm": 1.5343029499053955,
+      "learning_rate": 1.906368736948272e-05,
+      "loss": 1.1077,
+      "step": 723
+    },
+    {
+      "epoch": 0.16576989124212937,
+      "grad_norm": 1.270850658416748,
+      "learning_rate": 1.9060550793952803e-05,
+      "loss": 1.1389,
+      "step": 724
+    },
+    {
+      "epoch": 0.1659988551803091,
+      "grad_norm": 1.866256594657898,
+      "learning_rate": 1.905740923247775e-05,
+      "loss": 1.122,
+      "step": 725
+    },
+    {
+      "epoch": 0.16622781911848883,
+      "grad_norm": 1.560256004333496,
+      "learning_rate": 1.9054262686786332e-05,
+      "loss": 1.1524,
+      "step": 726
+    },
+    {
+      "epoch": 0.16645678305666858,
+      "grad_norm": 1.3043514490127563,
+      "learning_rate": 1.905111115861007e-05,
+      "loss": 1.1863,
+      "step": 727
+    },
+    {
+      "epoch": 0.1666857469948483,
+      "grad_norm": 1.1351722478866577,
+      "learning_rate": 1.9047954649683217e-05,
+      "loss": 1.1149,
+      "step": 728
+    },
+    {
+      "epoch": 0.16691471093302804,
+      "grad_norm": 1.6116639375686646,
+      "learning_rate": 1.9044793161742783e-05,
+      "loss": 1.1352,
+      "step": 729
+    },
+    {
+      "epoch": 0.1671436748712078,
+      "grad_norm": 1.3612065315246582,
+      "learning_rate": 1.9041626696528503e-05,
+      "loss": 1.1206,
+      "step": 730
+    },
+    {
+      "epoch": 0.16737263880938752,
+      "grad_norm": 1.5191378593444824,
+      "learning_rate": 1.903845525578286e-05,
+      "loss": 1.1476,
+      "step": 731
+    },
+    {
+      "epoch": 0.16760160274756725,
+      "grad_norm": 1.3559706211090088,
+      "learning_rate": 1.903527884125106e-05,
+      "loss": 1.0739,
+      "step": 732
+    },
+    {
+      "epoch": 0.167830566685747,
+      "grad_norm": 2.067094087600708,
+      "learning_rate": 1.9032097454681074e-05,
+      "loss": 1.1463,
+      "step": 733
+    },
+    {
+      "epoch": 0.16805953062392673,
+      "grad_norm": 1.6561821699142456,
+      "learning_rate": 1.9028911097823578e-05,
+      "loss": 1.1236,
+      "step": 734
+    },
+    {
+      "epoch": 0.16828849456210646,
+      "grad_norm": 1.81977117061615,
+      "learning_rate": 1.9025719772432006e-05,
+      "loss": 1.1047,
+      "step": 735
+    },
+    {
+      "epoch": 0.1685174585002862,
+      "grad_norm": 1.2570257186889648,
+      "learning_rate": 1.9022523480262517e-05,
+      "loss": 1.1246,
+      "step": 736
+    },
+    {
+      "epoch": 0.16874642243846594,
+      "grad_norm": 1.2872344255447388,
+      "learning_rate": 1.9019322223073997e-05,
+      "loss": 1.1549,
+      "step": 737
+    },
+    {
+      "epoch": 0.1689753863766457,
+      "grad_norm": 1.111267328262329,
+      "learning_rate": 1.901611600262808e-05,
+      "loss": 1.0933,
+      "step": 738
+    },
+    {
+      "epoch": 0.16920435031482542,
+      "grad_norm": 1.3469321727752686,
+      "learning_rate": 1.9012904820689114e-05,
+      "loss": 1.1454,
+      "step": 739
+    },
+    {
+      "epoch": 0.16943331425300515,
+      "grad_norm": 1.3398330211639404,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 1.0773,
+      "step": 740
+    },
+    {
+      "epoch": 0.1696622781911849,
+      "grad_norm": 1.4936128854751587,
+      "learning_rate": 1.9006467579403128e-05,
+      "loss": 1.1834,
+      "step": 741
+    },
+    {
+      "epoch": 0.16989124212936463,
+      "grad_norm": 1.2364333868026733,
+      "learning_rate": 1.9003241523598465e-05,
+      "loss": 1.1359,
+      "step": 742
+    },
+    {
+      "epoch": 0.17012020606754436,
+      "grad_norm": 2.011514663696289,
+      "learning_rate": 1.900001051338547e-05,
+      "loss": 1.1716,
+      "step": 743
+    },
+    {
+      "epoch": 0.1703491700057241,
+      "grad_norm": 1.3461471796035767,
+      "learning_rate": 1.899677455054215e-05,
+      "loss": 1.1082,
+      "step": 744
+    },
+    {
+      "epoch": 0.17057813394390384,
+      "grad_norm": 1.3014931678771973,
+      "learning_rate": 1.8993533636849223e-05,
+      "loss": 1.1362,
+      "step": 745
+    },
+    {
+      "epoch": 0.17080709788208356,
+      "grad_norm": 1.2128814458847046,
+      "learning_rate": 1.8990287774090137e-05,
+      "loss": 1.0883,
+      "step": 746
+    },
+    {
+      "epoch": 0.17103606182026332,
+      "grad_norm": 1.4403022527694702,
+      "learning_rate": 1.8987036964051065e-05,
+      "loss": 1.1065,
+      "step": 747
+    },
+    {
+      "epoch": 0.17126502575844305,
+      "grad_norm": 1.252302885055542,
+      "learning_rate": 1.8983781208520898e-05,
+      "loss": 1.1181,
+      "step": 748
+    },
+    {
+      "epoch": 0.17149398969662277,
+      "grad_norm": 1.318610429763794,
+      "learning_rate": 1.8980520509291255e-05,
+      "loss": 1.1492,
+      "step": 749
+    },
+    {
+      "epoch": 0.17172295363480253,
+      "grad_norm": 1.5634231567382812,
+      "learning_rate": 1.8977254868156465e-05,
+      "loss": 1.0785,
+      "step": 750
+    },
+    {
+      "epoch": 0.17195191757298225,
+      "grad_norm": 2.146759033203125,
+      "learning_rate": 1.8973984286913584e-05,
+      "loss": 1.1331,
+      "step": 751
+    },
+    {
+      "epoch": 0.17218088151116198,
+      "grad_norm": 1.2476396560668945,
+      "learning_rate": 1.897070876736239e-05,
+      "loss": 1.1564,
+      "step": 752
+    },
+    {
+      "epoch": 0.17240984544934174,
+      "grad_norm": 1.4595783948898315,
+      "learning_rate": 1.8967428311305375e-05,
+      "loss": 1.1155,
+      "step": 753
+    },
+    {
+      "epoch": 0.17263880938752146,
+      "grad_norm": 1.1093003749847412,
+      "learning_rate": 1.896414292054774e-05,
+      "loss": 1.1345,
+      "step": 754
+    },
+    {
+      "epoch": 0.1728677733257012,
+      "grad_norm": 1.1664748191833496,
+      "learning_rate": 1.896085259689741e-05,
+      "loss": 1.0583,
+      "step": 755
+    },
+    {
+      "epoch": 0.17309673726388095,
+      "grad_norm": 1.5106245279312134,
+      "learning_rate": 1.8957557342165026e-05,
+      "loss": 1.1689,
+      "step": 756
+    },
+    {
+      "epoch": 0.17332570120206067,
+      "grad_norm": 1.3270502090454102,
+      "learning_rate": 1.8954257158163936e-05,
+      "loss": 1.1115,
+      "step": 757
+    },
+    {
+      "epoch": 0.1735546651402404,
+      "grad_norm": 1.5598188638687134,
+      "learning_rate": 1.895095204671021e-05,
+      "loss": 1.1486,
+      "step": 758
+    },
+    {
+      "epoch": 0.17378362907842015,
+      "grad_norm": 1.5022259950637817,
+      "learning_rate": 1.8947642009622612e-05,
+      "loss": 1.1217,
+      "step": 759
+    },
+    {
+      "epoch": 0.17401259301659988,
+      "grad_norm": 1.3121103048324585,
+      "learning_rate": 1.8944327048722634e-05,
+      "loss": 1.1604,
+      "step": 760
+    },
+    {
+      "epoch": 0.1742415569547796,
+      "grad_norm": 1.4838922023773193,
+      "learning_rate": 1.8941007165834468e-05,
+      "loss": 1.2222,
+      "step": 761
+    },
+    {
+      "epoch": 0.17447052089295936,
+      "grad_norm": 1.4497283697128296,
+      "learning_rate": 1.8937682362785025e-05,
+      "loss": 1.1726,
+      "step": 762
+    },
+    {
+      "epoch": 0.1746994848311391,
+      "grad_norm": 1.3479863405227661,
+      "learning_rate": 1.8934352641403907e-05,
+      "loss": 1.1307,
+      "step": 763
+    },
+    {
+      "epoch": 0.17492844876931885,
+      "grad_norm": 1.622117042541504,
+      "learning_rate": 1.8931018003523438e-05,
+      "loss": 1.0881,
+      "step": 764
+    },
+    {
+      "epoch": 0.17515741270749857,
+      "grad_norm": 1.377864956855774,
+      "learning_rate": 1.892767845097864e-05,
+      "loss": 1.0872,
+      "step": 765
+    },
+    {
+      "epoch": 0.1753863766456783,
+      "grad_norm": 1.4404441118240356,
+      "learning_rate": 1.8924333985607237e-05,
+      "loss": 1.1333,
+      "step": 766
+    },
+    {
+      "epoch": 0.17561534058385805,
+      "grad_norm": 1.5697600841522217,
+      "learning_rate": 1.8920984609249667e-05,
+      "loss": 1.1069,
+      "step": 767
+    },
+    {
+      "epoch": 0.17584430452203778,
+      "grad_norm": 1.4469575881958008,
+      "learning_rate": 1.8917630323749056e-05,
+      "loss": 1.1294,
+      "step": 768
+    },
+    {
+      "epoch": 0.1760732684602175,
+      "grad_norm": 1.2435827255249023,
+      "learning_rate": 1.8914271130951246e-05,
+      "loss": 1.1437,
+      "step": 769
+    },
+    {
+      "epoch": 0.17630223239839726,
+      "grad_norm": 1.4122284650802612,
+      "learning_rate": 1.891090703270477e-05,
+      "loss": 1.1627,
+      "step": 770
+    },
+    {
+      "epoch": 0.176531196336577,
+      "grad_norm": 1.3929946422576904,
+      "learning_rate": 1.8907538030860865e-05,
+      "loss": 1.1148,
+      "step": 771
+    },
+    {
+      "epoch": 0.17676016027475672,
+      "grad_norm": 1.9755219221115112,
+      "learning_rate": 1.890416412727346e-05,
+      "loss": 1.1699,
+      "step": 772
+    },
+    {
+      "epoch": 0.17698912421293647,
+      "grad_norm": 1.6826586723327637,
+      "learning_rate": 1.890078532379919e-05,
+      "loss": 1.0846,
+      "step": 773
+    },
+    {
+      "epoch": 0.1772180881511162,
+      "grad_norm": 1.2816320657730103,
+      "learning_rate": 1.8897401622297375e-05,
+      "loss": 1.2024,
+      "step": 774
+    },
+    {
+      "epoch": 0.17744705208929593,
+      "grad_norm": 1.4006059169769287,
+      "learning_rate": 1.889401302463005e-05,
+      "loss": 1.116,
+      "step": 775
+    },
+    {
+      "epoch": 0.17767601602747568,
+      "grad_norm": 1.476759910583496,
+      "learning_rate": 1.889061953266192e-05,
+      "loss": 1.0944,
+      "step": 776
+    },
+    {
+      "epoch": 0.1779049799656554,
+      "grad_norm": 1.5726763010025024,
+      "learning_rate": 1.8887221148260404e-05,
+      "loss": 1.1472,
+      "step": 777
+    },
+    {
+      "epoch": 0.17813394390383513,
+      "grad_norm": 1.337249517440796,
+      "learning_rate": 1.8883817873295597e-05,
+      "loss": 1.1403,
+      "step": 778
+    },
+    {
+      "epoch": 0.1783629078420149,
+      "grad_norm": 1.338131308555603,
+      "learning_rate": 1.88804097096403e-05,
+      "loss": 1.0678,
+      "step": 779
+    },
+    {
+      "epoch": 0.17859187178019462,
+      "grad_norm": 1.291359543800354,
+      "learning_rate": 1.887699665916999e-05,
+      "loss": 1.1212,
+      "step": 780
+    },
+    {
+      "epoch": 0.17882083571837434,
+      "grad_norm": 1.7371107339859009,
+      "learning_rate": 1.8873578723762845e-05,
+      "loss": 1.0984,
+      "step": 781
+    },
+    {
+      "epoch": 0.1790497996565541,
+      "grad_norm": 1.5285062789916992,
+      "learning_rate": 1.8870155905299725e-05,
+      "loss": 1.1511,
+      "step": 782
+    },
+    {
+      "epoch": 0.17927876359473383,
+      "grad_norm": 1.693925380706787,
+      "learning_rate": 1.8866728205664177e-05,
+      "loss": 1.117,
+      "step": 783
+    },
+    {
+      "epoch": 0.17950772753291355,
+      "grad_norm": 1.2213523387908936,
+      "learning_rate": 1.8863295626742438e-05,
+      "loss": 1.2201,
+      "step": 784
+    },
+    {
+      "epoch": 0.1797366914710933,
+      "grad_norm": 1.410949945449829,
+      "learning_rate": 1.8859858170423423e-05,
+      "loss": 1.1321,
+      "step": 785
+    },
+    {
+      "epoch": 0.17996565540927303,
+      "grad_norm": 1.3400486707687378,
+      "learning_rate": 1.8856415838598738e-05,
+      "loss": 1.2231,
+      "step": 786
+    },
+    {
+      "epoch": 0.1801946193474528,
+      "grad_norm": 1.147056221961975,
+      "learning_rate": 1.885296863316267e-05,
+      "loss": 1.1623,
+      "step": 787
+    },
+    {
+      "epoch": 0.18042358328563252,
+      "grad_norm": 1.1560708284378052,
+      "learning_rate": 1.8849516556012183e-05,
+      "loss": 1.1123,
+      "step": 788
+    },
+    {
+      "epoch": 0.18065254722381224,
+      "grad_norm": 1.350136399269104,
+      "learning_rate": 1.8846059609046935e-05,
+      "loss": 1.1598,
+      "step": 789
+    },
+    {
+      "epoch": 0.180881511161992,
+      "grad_norm": 1.242692232131958,
+      "learning_rate": 1.8842597794169245e-05,
+      "loss": 1.1087,
+      "step": 790
+    },
+    {
+      "epoch": 0.18111047510017173,
+      "grad_norm": 1.2484183311462402,
+      "learning_rate": 1.883913111328413e-05,
+      "loss": 1.2007,
+      "step": 791
+    },
+    {
+      "epoch": 0.18133943903835145,
+      "grad_norm": 1.3073456287384033,
+      "learning_rate": 1.8835659568299268e-05,
+      "loss": 1.1419,
+      "step": 792
+    },
+    {
+      "epoch": 0.1815684029765312,
+      "grad_norm": 1.2637773752212524,
+      "learning_rate": 1.8832183161125026e-05,
+      "loss": 1.1126,
+      "step": 793
+    },
+    {
+      "epoch": 0.18179736691471093,
+      "grad_norm": 2.1271276473999023,
+      "learning_rate": 1.8828701893674435e-05,
+      "loss": 1.0735,
+      "step": 794
+    },
+    {
+      "epoch": 0.18202633085289066,
+      "grad_norm": 1.8395510911941528,
+      "learning_rate": 1.8825215767863215e-05,
+      "loss": 1.1168,
+      "step": 795
+    },
+    {
+      "epoch": 0.18225529479107042,
+      "grad_norm": 1.3857455253601074,
+      "learning_rate": 1.882172478560975e-05,
+      "loss": 1.1164,
+      "step": 796
+    },
+    {
+      "epoch": 0.18248425872925014,
+      "grad_norm": 1.313710331916809,
+      "learning_rate": 1.8818228948835095e-05,
+      "loss": 1.1883,
+      "step": 797
+    },
+    {
+      "epoch": 0.18271322266742987,
+      "grad_norm": 1.5113545656204224,
+      "learning_rate": 1.8814728259462978e-05,
+      "loss": 1.1948,
+      "step": 798
+    },
+    {
+      "epoch": 0.18294218660560962,
+      "grad_norm": 1.3891445398330688,
+      "learning_rate": 1.8811222719419808e-05,
+      "loss": 1.1182,
+      "step": 799
+    },
+    {
+      "epoch": 0.18317115054378935,
+      "grad_norm": 1.672825574874878,
+      "learning_rate": 1.8807712330634645e-05,
+      "loss": 1.1114,
+      "step": 800
+    },
+    {
+      "epoch": 0.18340011448196908,
+      "grad_norm": 1.299664855003357,
+      "learning_rate": 1.880419709503923e-05,
+      "loss": 1.1228,
+      "step": 801
+    },
+    {
+      "epoch": 0.18362907842014883,
+      "grad_norm": 1.6080543994903564,
+      "learning_rate": 1.8800677014567972e-05,
+      "loss": 1.1325,
+      "step": 802
+    },
+    {
+      "epoch": 0.18385804235832856,
+      "grad_norm": 1.4526350498199463,
+      "learning_rate": 1.8797152091157935e-05,
+      "loss": 1.109,
+      "step": 803
+    },
+    {
+      "epoch": 0.1840870062965083,
+      "grad_norm": 1.2981843948364258,
+      "learning_rate": 1.8793622326748857e-05,
+      "loss": 1.1266,
+      "step": 804
+    },
+    {
+      "epoch": 0.18431597023468804,
+      "grad_norm": 1.715675711631775,
+      "learning_rate": 1.879008772328314e-05,
+      "loss": 1.1661,
+      "step": 805
+    },
+    {
+      "epoch": 0.18454493417286777,
+      "grad_norm": 1.5412567853927612,
+      "learning_rate": 1.8786548282705847e-05,
+      "loss": 1.0697,
+      "step": 806
+    },
+    {
+      "epoch": 0.1847738981110475,
+      "grad_norm": 1.3529974222183228,
+      "learning_rate": 1.87830040069647e-05,
+      "loss": 1.1753,
+      "step": 807
+    },
+    {
+      "epoch": 0.18500286204922725,
+      "grad_norm": 1.229610800743103,
+      "learning_rate": 1.877945489801008e-05,
+      "loss": 1.1039,
+      "step": 808
+    },
+    {
+      "epoch": 0.18523182598740698,
+      "grad_norm": 1.704888939857483,
+      "learning_rate": 1.8775900957795042e-05,
+      "loss": 1.1366,
+      "step": 809
+    },
+    {
+      "epoch": 0.18546078992558673,
+      "grad_norm": 1.142831802368164,
+      "learning_rate": 1.877234218827528e-05,
+      "loss": 1.0946,
+      "step": 810
+    },
+    {
+      "epoch": 0.18568975386376646,
+      "grad_norm": 1.304379940032959,
+      "learning_rate": 1.876877859140916e-05,
+      "loss": 1.1063,
+      "step": 811
+    },
+    {
+      "epoch": 0.1859187178019462,
+      "grad_norm": 1.1813594102859497,
+      "learning_rate": 1.8765210169157703e-05,
+      "loss": 1.1067,
+      "step": 812
+    },
+    {
+      "epoch": 0.18614768174012594,
+      "grad_norm": 1.6390926837921143,
+      "learning_rate": 1.876163692348457e-05,
+      "loss": 1.1525,
+      "step": 813
+    },
+    {
+      "epoch": 0.18637664567830567,
+      "grad_norm": 1.6113990545272827,
+      "learning_rate": 1.87580588563561e-05,
+      "loss": 1.1551,
+      "step": 814
+    },
+    {
+      "epoch": 0.1866056096164854,
+      "grad_norm": 1.4802757501602173,
+      "learning_rate": 1.8754475969741272e-05,
+      "loss": 1.1503,
+      "step": 815
+    },
+    {
+      "epoch": 0.18683457355466515,
+      "grad_norm": 1.2753384113311768,
+      "learning_rate": 1.8750888265611708e-05,
+      "loss": 1.1212,
+      "step": 816
+    },
+    {
+      "epoch": 0.18706353749284488,
+      "grad_norm": 1.1978129148483276,
+      "learning_rate": 1.8747295745941705e-05,
+      "loss": 1.1429,
+      "step": 817
+    },
+    {
+      "epoch": 0.1872925014310246,
+      "grad_norm": 1.1764678955078125,
+      "learning_rate": 1.8743698412708187e-05,
+      "loss": 1.1981,
+      "step": 818
+    },
+    {
+      "epoch": 0.18752146536920436,
+      "grad_norm": 1.5134238004684448,
+      "learning_rate": 1.874009626789074e-05,
+      "loss": 1.1221,
+      "step": 819
+    },
+    {
+      "epoch": 0.1877504293073841,
+      "grad_norm": 1.2333041429519653,
+      "learning_rate": 1.87364893134716e-05,
+      "loss": 1.1512,
+      "step": 820
+    },
+    {
+      "epoch": 0.18797939324556381,
+      "grad_norm": 1.1870229244232178,
+      "learning_rate": 1.873287755143563e-05,
+      "loss": 1.1544,
+      "step": 821
+    },
+    {
+      "epoch": 0.18820835718374357,
+      "grad_norm": 1.222976565361023,
+      "learning_rate": 1.872926098377036e-05,
+      "loss": 1.1167,
+      "step": 822
+    },
+    {
+      "epoch": 0.1884373211219233,
+      "grad_norm": 1.2639892101287842,
+      "learning_rate": 1.872563961246596e-05,
+      "loss": 1.1357,
+      "step": 823
+    },
+    {
+      "epoch": 0.18866628506010302,
+      "grad_norm": 1.1272286176681519,
+      "learning_rate": 1.8722013439515236e-05,
+      "loss": 1.1222,
+      "step": 824
+    },
+    {
+      "epoch": 0.18889524899828278,
+      "grad_norm": 1.2517902851104736,
+      "learning_rate": 1.8718382466913642e-05,
+      "loss": 1.0577,
+      "step": 825
+    },
+    {
+      "epoch": 0.1891242129364625,
+      "grad_norm": 1.3976292610168457,
+      "learning_rate": 1.871474669665927e-05,
+      "loss": 1.125,
+      "step": 826
+    },
+    {
+      "epoch": 0.18935317687464223,
+      "grad_norm": 1.212823510169983,
+      "learning_rate": 1.8711106130752855e-05,
+      "loss": 1.1143,
+      "step": 827
+    },
+    {
+      "epoch": 0.189582140812822,
+      "grad_norm": 1.4794899225234985,
+      "learning_rate": 1.8707460771197773e-05,
+      "loss": 1.1415,
+      "step": 828
+    },
+    {
+      "epoch": 0.1898111047510017,
+      "grad_norm": 1.4691826105117798,
+      "learning_rate": 1.8703810620000033e-05,
+      "loss": 1.1055,
+      "step": 829
+    },
+    {
+      "epoch": 0.19004006868918144,
+      "grad_norm": 1.352226972579956,
+      "learning_rate": 1.8700155679168277e-05,
+      "loss": 1.1638,
+      "step": 830
+    },
+    {
+      "epoch": 0.1902690326273612,
+      "grad_norm": 1.3408719301223755,
+      "learning_rate": 1.8696495950713794e-05,
+      "loss": 1.1159,
+      "step": 831
+    },
+    {
+      "epoch": 0.19049799656554092,
+      "grad_norm": 1.0246559381484985,
+      "learning_rate": 1.8692831436650505e-05,
+      "loss": 1.1375,
+      "step": 832
+    },
+    {
+      "epoch": 0.19072696050372068,
+      "grad_norm": 1.1187310218811035,
+      "learning_rate": 1.8689162138994952e-05,
+      "loss": 1.1453,
+      "step": 833
+    },
+    {
+      "epoch": 0.1909559244419004,
+      "grad_norm": 1.2305899858474731,
+      "learning_rate": 1.868548805976633e-05,
+      "loss": 1.1069,
+      "step": 834
+    },
+    {
+      "epoch": 0.19118488838008013,
+      "grad_norm": 1.4860109090805054,
+      "learning_rate": 1.868180920098644e-05,
+      "loss": 1.0497,
+      "step": 835
+    },
+    {
+      "epoch": 0.19141385231825989,
+      "grad_norm": 1.2232182025909424,
+      "learning_rate": 1.867812556467974e-05,
+      "loss": 1.131,
+      "step": 836
+    },
+    {
+      "epoch": 0.1916428162564396,
+      "grad_norm": 1.3128330707550049,
+      "learning_rate": 1.8674437152873296e-05,
+      "loss": 1.2106,
+      "step": 837
+    },
+    {
+      "epoch": 0.19187178019461934,
+      "grad_norm": 1.2296591997146606,
+      "learning_rate": 1.8670743967596817e-05,
+      "loss": 1.1277,
+      "step": 838
+    },
+    {
+      "epoch": 0.1921007441327991,
+      "grad_norm": 1.332777976989746,
+      "learning_rate": 1.8667046010882627e-05,
+      "loss": 1.1003,
+      "step": 839
+    },
+    {
+      "epoch": 0.19232970807097882,
+      "grad_norm": 1.8379496335983276,
+      "learning_rate": 1.866334328476568e-05,
+      "loss": 1.1812,
+      "step": 840
+    },
+    {
+      "epoch": 0.19255867200915855,
+      "grad_norm": 1.1897050142288208,
+      "learning_rate": 1.865963579128356e-05,
+      "loss": 1.0811,
+      "step": 841
+    },
+    {
+      "epoch": 0.1927876359473383,
+      "grad_norm": 1.7345019578933716,
+      "learning_rate": 1.8655923532476463e-05,
+      "loss": 1.1679,
+      "step": 842
+    },
+    {
+      "epoch": 0.19301659988551803,
+      "grad_norm": 1.20344877243042,
+      "learning_rate": 1.865220651038722e-05,
+      "loss": 1.1832,
+      "step": 843
+    },
+    {
+      "epoch": 0.19324556382369776,
+      "grad_norm": 1.2605164051055908,
+      "learning_rate": 1.864848472706127e-05,
+      "loss": 1.1355,
+      "step": 844
+    },
+    {
+      "epoch": 0.1934745277618775,
+      "grad_norm": 1.1412873268127441,
+      "learning_rate": 1.864475818454669e-05,
+      "loss": 1.1202,
+      "step": 845
+    },
+    {
+      "epoch": 0.19370349170005724,
+      "grad_norm": 1.1985411643981934,
+      "learning_rate": 1.8641026884894156e-05,
+      "loss": 1.1509,
+      "step": 846
+    },
+    {
+      "epoch": 0.19393245563823697,
+      "grad_norm": 1.0788958072662354,
+      "learning_rate": 1.8637290830156972e-05,
+      "loss": 1.1141,
+      "step": 847
+    },
+    {
+      "epoch": 0.19416141957641672,
+      "grad_norm": 1.2294611930847168,
+      "learning_rate": 1.8633550022391062e-05,
+      "loss": 1.1473,
+      "step": 848
+    },
+    {
+      "epoch": 0.19439038351459645,
+      "grad_norm": 1.1396217346191406,
+      "learning_rate": 1.8629804463654956e-05,
+      "loss": 1.141,
+      "step": 849
+    },
+    {
+      "epoch": 0.19461934745277618,
+      "grad_norm": 1.0818895101547241,
+      "learning_rate": 1.8626054156009807e-05,
+      "loss": 1.1195,
+      "step": 850
+    },
+    {
+      "epoch": 0.19484831139095593,
+      "grad_norm": 1.287191390991211,
+      "learning_rate": 1.862229910151938e-05,
+      "loss": 1.1387,
+      "step": 851
+    },
+    {
+      "epoch": 0.19507727532913566,
+      "grad_norm": 1.1082041263580322,
+      "learning_rate": 1.8618539302250044e-05,
+      "loss": 1.1418,
+      "step": 852
+    },
+    {
+      "epoch": 0.19530623926731538,
+      "grad_norm": 1.2280734777450562,
+      "learning_rate": 1.8614774760270785e-05,
+      "loss": 1.157,
+      "step": 853
+    },
+    {
+      "epoch": 0.19553520320549514,
+      "grad_norm": 1.1522899866104126,
+      "learning_rate": 1.8611005477653204e-05,
+      "loss": 1.0825,
+      "step": 854
+    },
+    {
+      "epoch": 0.19576416714367487,
+      "grad_norm": 1.2309021949768066,
+      "learning_rate": 1.8607231456471505e-05,
+      "loss": 1.151,
+      "step": 855
+    },
+    {
+      "epoch": 0.19599313108185462,
+      "grad_norm": 1.141215205192566,
+      "learning_rate": 1.8603452698802498e-05,
+      "loss": 1.0634,
+      "step": 856
+    },
+    {
+      "epoch": 0.19622209502003435,
+      "grad_norm": 1.6352757215499878,
+      "learning_rate": 1.85996692067256e-05,
+      "loss": 1.14,
+      "step": 857
+    },
+    {
+      "epoch": 0.19645105895821408,
+      "grad_norm": 1.2608399391174316,
+      "learning_rate": 1.859588098232284e-05,
+      "loss": 1.1651,
+      "step": 858
+    },
+    {
+      "epoch": 0.19668002289639383,
+      "grad_norm": 1.3403522968292236,
+      "learning_rate": 1.859208802767884e-05,
+      "loss": 1.0751,
+      "step": 859
+    },
+    {
+      "epoch": 0.19690898683457356,
+      "grad_norm": 1.5046757459640503,
+      "learning_rate": 1.858829034488084e-05,
+      "loss": 1.1455,
+      "step": 860
+    },
+    {
+      "epoch": 0.19713795077275328,
+      "grad_norm": 1.1532163619995117,
+      "learning_rate": 1.8584487936018663e-05,
+      "loss": 1.0847,
+      "step": 861
+    },
+    {
+      "epoch": 0.19736691471093304,
+      "grad_norm": 1.1010034084320068,
+      "learning_rate": 1.858068080318475e-05,
+      "loss": 1.1729,
+      "step": 862
+    },
+    {
+      "epoch": 0.19759587864911277,
+      "grad_norm": 1.2024052143096924,
+      "learning_rate": 1.857686894847413e-05,
+      "loss": 1.1543,
+      "step": 863
+    },
+    {
+      "epoch": 0.1978248425872925,
+      "grad_norm": 1.4353835582733154,
+      "learning_rate": 1.8573052373984435e-05,
+      "loss": 1.1061,
+      "step": 864
+    },
+    {
+      "epoch": 0.19805380652547225,
+      "grad_norm": 1.281450867652893,
+      "learning_rate": 1.8569231081815895e-05,
+      "loss": 1.0886,
+      "step": 865
+    },
+    {
+      "epoch": 0.19828277046365198,
+      "grad_norm": 1.2481147050857544,
+      "learning_rate": 1.8565405074071338e-05,
+      "loss": 1.0985,
+      "step": 866
+    },
+    {
+      "epoch": 0.1985117344018317,
+      "grad_norm": 1.2200899124145508,
+      "learning_rate": 1.8561574352856176e-05,
+      "loss": 1.1767,
+      "step": 867
+    },
+    {
+      "epoch": 0.19874069834001146,
+      "grad_norm": 1.187233328819275,
+      "learning_rate": 1.855773892027843e-05,
+      "loss": 1.0757,
+      "step": 868
+    },
+    {
+      "epoch": 0.19896966227819118,
+      "grad_norm": 1.319388508796692,
+      "learning_rate": 1.855389877844871e-05,
+      "loss": 1.1752,
+      "step": 869
+    },
+    {
+      "epoch": 0.1991986262163709,
+      "grad_norm": 1.2449707984924316,
+      "learning_rate": 1.8550053929480202e-05,
+      "loss": 1.1535,
+      "step": 870
+    },
+    {
+      "epoch": 0.19942759015455067,
+      "grad_norm": 1.315616250038147,
+      "learning_rate": 1.8546204375488702e-05,
+      "loss": 1.0604,
+      "step": 871
+    },
+    {
+      "epoch": 0.1996565540927304,
+      "grad_norm": 1.368754267692566,
+      "learning_rate": 1.8542350118592585e-05,
+      "loss": 1.1437,
+      "step": 872
+    },
+    {
+      "epoch": 0.19988551803091012,
+      "grad_norm": 1.2946205139160156,
+      "learning_rate": 1.853849116091282e-05,
+      "loss": 1.1325,
+      "step": 873
+    },
+    {
+      "epoch": 0.20011448196908987,
+      "grad_norm": 1.3224684000015259,
+      "learning_rate": 1.853462750457295e-05,
+      "loss": 1.0641,
+      "step": 874
+    },
+    {
+      "epoch": 0.2003434459072696,
+      "grad_norm": 1.159767985343933,
+      "learning_rate": 1.8530759151699122e-05,
+      "loss": 1.0747,
+      "step": 875
+    },
+    {
+      "epoch": 0.20057240984544933,
+      "grad_norm": 1.1444604396820068,
+      "learning_rate": 1.8526886104420056e-05,
+      "loss": 1.1565,
+      "step": 876
+    },
+    {
+      "epoch": 0.20080137378362908,
+      "grad_norm": 1.265729308128357,
+      "learning_rate": 1.8523008364867056e-05,
+      "loss": 1.0907,
+      "step": 877
+    },
+    {
+      "epoch": 0.2010303377218088,
+      "grad_norm": 1.0625101327896118,
+      "learning_rate": 1.851912593517401e-05,
+      "loss": 1.1093,
+      "step": 878
+    },
+    {
+      "epoch": 0.20125930165998857,
+      "grad_norm": 1.4997196197509766,
+      "learning_rate": 1.8515238817477382e-05,
+      "loss": 1.1674,
+      "step": 879
+    },
+    {
+      "epoch": 0.2014882655981683,
+      "grad_norm": 1.2591129541397095,
+      "learning_rate": 1.8511347013916228e-05,
+      "loss": 1.1546,
+      "step": 880
+    },
+    {
+      "epoch": 0.20171722953634802,
+      "grad_norm": 1.1916770935058594,
+      "learning_rate": 1.850745052663217e-05,
+      "loss": 1.094,
+      "step": 881
+    },
+    {
+      "epoch": 0.20194619347452777,
+      "grad_norm": 1.0707792043685913,
+      "learning_rate": 1.850354935776941e-05,
+      "loss": 1.0586,
+      "step": 882
+    },
+    {
+      "epoch": 0.2021751574127075,
+      "grad_norm": 1.159035086631775,
+      "learning_rate": 1.8499643509474738e-05,
+      "loss": 1.1245,
+      "step": 883
+    },
+    {
+      "epoch": 0.20240412135088723,
+      "grad_norm": 1.2937264442443848,
+      "learning_rate": 1.8495732983897504e-05,
+      "loss": 1.1293,
+      "step": 884
+    },
+    {
+      "epoch": 0.20263308528906698,
+      "grad_norm": 1.2372575998306274,
+      "learning_rate": 1.8491817783189636e-05,
+      "loss": 1.1198,
+      "step": 885
+    },
+    {
+      "epoch": 0.2028620492272467,
+      "grad_norm": 1.1078606843948364,
+      "learning_rate": 1.8487897909505637e-05,
+      "loss": 1.1512,
+      "step": 886
+    },
+    {
+      "epoch": 0.20309101316542644,
+      "grad_norm": 1.2595610618591309,
+      "learning_rate": 1.848397336500258e-05,
+      "loss": 1.1433,
+      "step": 887
+    },
+    {
+      "epoch": 0.2033199771036062,
+      "grad_norm": 1.1265288591384888,
+      "learning_rate": 1.848004415184011e-05,
+      "loss": 1.1761,
+      "step": 888
+    },
+    {
+      "epoch": 0.20354894104178592,
+      "grad_norm": 1.1440848112106323,
+      "learning_rate": 1.8476110272180443e-05,
+      "loss": 1.173,
+      "step": 889
+    },
+    {
+      "epoch": 0.20377790497996565,
+      "grad_norm": 1.281594157218933,
+      "learning_rate": 1.8472171728188356e-05,
+      "loss": 1.1097,
+      "step": 890
+    },
+    {
+      "epoch": 0.2040068689181454,
+      "grad_norm": 1.2258485555648804,
+      "learning_rate": 1.8468228522031197e-05,
+      "loss": 1.0735,
+      "step": 891
+    },
+    {
+      "epoch": 0.20423583285632513,
+      "grad_norm": 1.1039539575576782,
+      "learning_rate": 1.8464280655878876e-05,
+      "loss": 1.1628,
+      "step": 892
+    },
+    {
+      "epoch": 0.20446479679450486,
+      "grad_norm": 1.0621813535690308,
+      "learning_rate": 1.846032813190388e-05,
+      "loss": 1.1138,
+      "step": 893
+    },
+    {
+      "epoch": 0.2046937607326846,
+      "grad_norm": 1.2525299787521362,
+      "learning_rate": 1.8456370952281243e-05,
+      "loss": 1.1612,
+      "step": 894
+    },
+    {
+      "epoch": 0.20492272467086434,
+      "grad_norm": 1.553027629852295,
+      "learning_rate": 1.845240911918857e-05,
+      "loss": 1.1157,
+      "step": 895
+    },
+    {
+      "epoch": 0.20515168860904406,
+      "grad_norm": 1.309720754623413,
+      "learning_rate": 1.844844263480602e-05,
+      "loss": 1.1009,
+      "step": 896
+    },
+    {
+      "epoch": 0.20538065254722382,
+      "grad_norm": 1.2567998170852661,
+      "learning_rate": 1.8444471501316324e-05,
+      "loss": 1.136,
+      "step": 897
+    },
+    {
+      "epoch": 0.20560961648540355,
+      "grad_norm": 1.242303490638733,
+      "learning_rate": 1.8440495720904758e-05,
+      "loss": 1.1018,
+      "step": 898
+    },
+    {
+      "epoch": 0.20583858042358327,
+      "grad_norm": 1.5786601305007935,
+      "learning_rate": 1.843651529575916e-05,
+      "loss": 1.126,
+      "step": 899
+    },
+    {
+      "epoch": 0.20606754436176303,
+      "grad_norm": 1.179705262184143,
+      "learning_rate": 1.843253022806993e-05,
+      "loss": 1.1069,
+      "step": 900
+    },
+    {
+      "epoch": 0.20629650829994275,
+      "grad_norm": 1.4622169733047485,
+      "learning_rate": 1.842854052003001e-05,
+      "loss": 1.0582,
+      "step": 901
+    },
+    {
+      "epoch": 0.2065254722381225,
+      "grad_norm": 1.43120276927948,
+      "learning_rate": 1.842454617383491e-05,
+      "loss": 1.0877,
+      "step": 902
+    },
+    {
+      "epoch": 0.20675443617630224,
+      "grad_norm": 1.0542099475860596,
+      "learning_rate": 1.842054719168268e-05,
+      "loss": 1.0746,
+      "step": 903
+    },
+    {
+      "epoch": 0.20698340011448196,
+      "grad_norm": 1.2692487239837646,
+      "learning_rate": 1.841654357577393e-05,
+      "loss": 1.1247,
+      "step": 904
+    },
+    {
+      "epoch": 0.20721236405266172,
+      "grad_norm": 1.1795223951339722,
+      "learning_rate": 1.8412535328311813e-05,
+      "loss": 1.186,
+      "step": 905
+    },
+    {
+      "epoch": 0.20744132799084145,
+      "grad_norm": 1.6526825428009033,
+      "learning_rate": 1.8408522451502038e-05,
+      "loss": 1.2059,
+      "step": 906
+    },
+    {
+      "epoch": 0.20767029192902117,
+      "grad_norm": 2.6972272396087646,
+      "learning_rate": 1.840450494755285e-05,
+      "loss": 1.186,
+      "step": 907
+    },
+    {
+      "epoch": 0.20789925586720093,
+      "grad_norm": 1.294978141784668,
+      "learning_rate": 1.840048281867506e-05,
+      "loss": 1.1069,
+      "step": 908
+    },
+    {
+      "epoch": 0.20812821980538065,
+      "grad_norm": 1.686403512954712,
+      "learning_rate": 1.8396456067082e-05,
+      "loss": 1.1007,
+      "step": 909
+    },
+    {
+      "epoch": 0.20835718374356038,
+      "grad_norm": 1.4886486530303955,
+      "learning_rate": 1.839242469498956e-05,
+      "loss": 1.1646,
+      "step": 910
+    },
+    {
+      "epoch": 0.20858614768174014,
+      "grad_norm": 1.4273146390914917,
+      "learning_rate": 1.8388388704616177e-05,
+      "loss": 1.0601,
+      "step": 911
+    },
+    {
+      "epoch": 0.20881511161991986,
+      "grad_norm": 1.9175933599472046,
+      "learning_rate": 1.8384348098182815e-05,
+      "loss": 1.1559,
+      "step": 912
+    },
+    {
+      "epoch": 0.2090440755580996,
+      "grad_norm": 1.3737765550613403,
+      "learning_rate": 1.8380302877912993e-05,
+      "loss": 1.185,
+      "step": 913
+    },
+    {
+      "epoch": 0.20927303949627934,
+      "grad_norm": 1.3466029167175293,
+      "learning_rate": 1.837625304603275e-05,
+      "loss": 1.14,
+      "step": 914
+    },
+    {
+      "epoch": 0.20950200343445907,
+      "grad_norm": 1.427209734916687,
+      "learning_rate": 1.837219860477069e-05,
+      "loss": 1.1444,
+      "step": 915
+    },
+    {
+      "epoch": 0.2097309673726388,
+      "grad_norm": 1.7554882764816284,
+      "learning_rate": 1.836813955635793e-05,
+      "loss": 1.1354,
+      "step": 916
+    },
+    {
+      "epoch": 0.20995993131081855,
+      "grad_norm": 1.5630890130996704,
+      "learning_rate": 1.8364075903028128e-05,
+      "loss": 1.0917,
+      "step": 917
+    },
+    {
+      "epoch": 0.21018889524899828,
+      "grad_norm": 1.6305307149887085,
+      "learning_rate": 1.836000764701748e-05,
+      "loss": 1.0982,
+      "step": 918
+    },
+    {
+      "epoch": 0.210417859187178,
+      "grad_norm": 1.154215931892395,
+      "learning_rate": 1.8355934790564718e-05,
+      "loss": 1.1199,
+      "step": 919
+    },
+    {
+      "epoch": 0.21064682312535776,
+      "grad_norm": 1.440573811531067,
+      "learning_rate": 1.8351857335911094e-05,
+      "loss": 1.1621,
+      "step": 920
+    },
+    {
+      "epoch": 0.2108757870635375,
+      "grad_norm": 1.2942014932632446,
+      "learning_rate": 1.83477752853004e-05,
+      "loss": 1.1378,
+      "step": 921
+    },
+    {
+      "epoch": 0.21110475100171722,
+      "grad_norm": 3.2026820182800293,
+      "learning_rate": 1.8343688640978955e-05,
+      "loss": 1.2049,
+      "step": 922
+    },
+    {
+      "epoch": 0.21133371493989697,
+      "grad_norm": 1.3892024755477905,
+      "learning_rate": 1.8339597405195607e-05,
+      "loss": 1.1328,
+      "step": 923
+    },
+    {
+      "epoch": 0.2115626788780767,
+      "grad_norm": 1.5560742616653442,
+      "learning_rate": 1.833550158020172e-05,
+      "loss": 1.1212,
+      "step": 924
+    },
+    {
+      "epoch": 0.21179164281625643,
+      "grad_norm": 1.1782360076904297,
+      "learning_rate": 1.83314011682512e-05,
+      "loss": 1.1153,
+      "step": 925
+    },
+    {
+      "epoch": 0.21202060675443618,
+      "grad_norm": 1.5242149829864502,
+      "learning_rate": 1.832729617160047e-05,
+      "loss": 1.1059,
+      "step": 926
+    },
+    {
+      "epoch": 0.2122495706926159,
+      "grad_norm": 1.157873272895813,
+      "learning_rate": 1.8323186592508474e-05,
+      "loss": 1.1489,
+      "step": 927
+    },
+    {
+      "epoch": 0.21247853463079566,
+      "grad_norm": 1.1628938913345337,
+      "learning_rate": 1.8319072433236677e-05,
+      "loss": 1.0962,
+      "step": 928
+    },
+    {
+      "epoch": 0.2127074985689754,
+      "grad_norm": 1.0474616289138794,
+      "learning_rate": 1.831495369604907e-05,
+      "loss": 1.1517,
+      "step": 929
+    },
+    {
+      "epoch": 0.21293646250715512,
+      "grad_norm": 1.7184207439422607,
+      "learning_rate": 1.831083038321215e-05,
+      "loss": 1.1172,
+      "step": 930
+    },
+    {
+      "epoch": 0.21316542644533487,
+      "grad_norm": 1.8269110918045044,
+      "learning_rate": 1.830670249699495e-05,
+      "loss": 1.1521,
+      "step": 931
+    },
+    {
+      "epoch": 0.2133943903835146,
+      "grad_norm": 1.1638559103012085,
+      "learning_rate": 1.830257003966901e-05,
+      "loss": 1.1759,
+      "step": 932
+    },
+    {
+      "epoch": 0.21362335432169433,
+      "grad_norm": 1.1705409288406372,
+      "learning_rate": 1.8298433013508384e-05,
+      "loss": 1.0605,
+      "step": 933
+    },
+    {
+      "epoch": 0.21385231825987408,
+      "grad_norm": 1.2644610404968262,
+      "learning_rate": 1.8294291420789648e-05,
+      "loss": 1.1405,
+      "step": 934
+    },
+    {
+      "epoch": 0.2140812821980538,
+      "grad_norm": 1.0992416143417358,
+      "learning_rate": 1.8290145263791883e-05,
+      "loss": 1.108,
+      "step": 935
+    },
+    {
+      "epoch": 0.21431024613623353,
+      "grad_norm": 1.291539192199707,
+      "learning_rate": 1.8285994544796685e-05,
+      "loss": 1.084,
+      "step": 936
+    },
+    {
+      "epoch": 0.2145392100744133,
+      "grad_norm": 1.244376540184021,
+      "learning_rate": 1.8281839266088155e-05,
+      "loss": 1.1151,
+      "step": 937
+    },
+    {
+      "epoch": 0.21476817401259302,
+      "grad_norm": 1.0517001152038574,
+      "learning_rate": 1.8277679429952913e-05,
+      "loss": 1.1116,
+      "step": 938
+    },
+    {
+      "epoch": 0.21499713795077274,
+      "grad_norm": 1.0727293491363525,
+      "learning_rate": 1.827351503868008e-05,
+      "loss": 1.1381,
+      "step": 939
+    },
+    {
+      "epoch": 0.2152261018889525,
+      "grad_norm": 1.1680247783660889,
+      "learning_rate": 1.826934609456129e-05,
+      "loss": 1.1139,
+      "step": 940
+    },
+    {
+      "epoch": 0.21545506582713222,
+      "grad_norm": 1.4314526319503784,
+      "learning_rate": 1.826517259989067e-05,
+      "loss": 1.1778,
+      "step": 941
+    },
+    {
+      "epoch": 0.21568402976531195,
+      "grad_norm": 1.3765356540679932,
+      "learning_rate": 1.8260994556964865e-05,
+      "loss": 1.1226,
+      "step": 942
+    },
+    {
+      "epoch": 0.2159129937034917,
+      "grad_norm": 1.1225863695144653,
+      "learning_rate": 1.8256811968083016e-05,
+      "loss": 1.0951,
+      "step": 943
+    },
+    {
+      "epoch": 0.21614195764167143,
+      "grad_norm": 1.3155685663223267,
+      "learning_rate": 1.8252624835546768e-05,
+      "loss": 1.1253,
+      "step": 944
+    },
+    {
+      "epoch": 0.21637092157985116,
+      "grad_norm": 1.3809353113174438,
+      "learning_rate": 1.824843316166026e-05,
+      "loss": 1.0972,
+      "step": 945
+    },
+    {
+      "epoch": 0.21659988551803092,
+      "grad_norm": 1.5605497360229492,
+      "learning_rate": 1.8244236948730138e-05,
+      "loss": 1.1479,
+      "step": 946
+    },
+    {
+      "epoch": 0.21682884945621064,
+      "grad_norm": 1.3077151775360107,
+      "learning_rate": 1.8240036199065546e-05,
+      "loss": 1.0822,
+      "step": 947
+    },
+    {
+      "epoch": 0.21705781339439037,
+      "grad_norm": 1.3724560737609863,
+      "learning_rate": 1.8235830914978113e-05,
+      "loss": 1.1366,
+      "step": 948
+    },
+    {
+      "epoch": 0.21728677733257012,
+      "grad_norm": 1.2070374488830566,
+      "learning_rate": 1.8231621098781983e-05,
+      "loss": 1.1371,
+      "step": 949
+    },
+    {
+      "epoch": 0.21751574127074985,
+      "grad_norm": 1.3094055652618408,
+      "learning_rate": 1.822740675279377e-05,
+      "loss": 1.1442,
+      "step": 950
+    },
+    {
+      "epoch": 0.2177447052089296,
+      "grad_norm": 1.3075108528137207,
+      "learning_rate": 1.8223187879332604e-05,
+      "loss": 1.1095,
+      "step": 951
+    },
+    {
+      "epoch": 0.21797366914710933,
+      "grad_norm": 1.2486156225204468,
+      "learning_rate": 1.8218964480720093e-05,
+      "loss": 1.0987,
+      "step": 952
+    },
+    {
+      "epoch": 0.21820263308528906,
+      "grad_norm": 1.1056116819381714,
+      "learning_rate": 1.821473655928033e-05,
+      "loss": 1.116,
+      "step": 953
+    },
+    {
+      "epoch": 0.21843159702346882,
+      "grad_norm": 1.235161542892456,
+      "learning_rate": 1.8210504117339917e-05,
+      "loss": 1.062,
+      "step": 954
+    },
+    {
+      "epoch": 0.21866056096164854,
+      "grad_norm": 1.3450078964233398,
+      "learning_rate": 1.8206267157227918e-05,
+      "loss": 1.0997,
+      "step": 955
+    },
+    {
+      "epoch": 0.21888952489982827,
+      "grad_norm": 1.1602591276168823,
+      "learning_rate": 1.820202568127591e-05,
+      "loss": 1.124,
+      "step": 956
+    },
+    {
+      "epoch": 0.21911848883800802,
+      "grad_norm": 1.4237440824508667,
+      "learning_rate": 1.819777969181793e-05,
+      "loss": 1.148,
+      "step": 957
+    },
+    {
+      "epoch": 0.21934745277618775,
+      "grad_norm": 1.4019017219543457,
+      "learning_rate": 1.819352919119052e-05,
+      "loss": 1.1316,
+      "step": 958
+    },
+    {
+      "epoch": 0.21957641671436748,
+      "grad_norm": 1.2780940532684326,
+      "learning_rate": 1.818927418173269e-05,
+      "loss": 1.1654,
+      "step": 959
+    },
+    {
+      "epoch": 0.21980538065254723,
+      "grad_norm": 1.7911431789398193,
+      "learning_rate": 1.8185014665785936e-05,
+      "loss": 1.0586,
+      "step": 960
+    },
+    {
+      "epoch": 0.22003434459072696,
+      "grad_norm": 2.0923030376434326,
+      "learning_rate": 1.8180750645694236e-05,
+      "loss": 1.146,
+      "step": 961
+    },
+    {
+      "epoch": 0.2202633085289067,
+      "grad_norm": 1.1630254983901978,
+      "learning_rate": 1.8176482123804042e-05,
+      "loss": 1.1488,
+      "step": 962
+    },
+    {
+      "epoch": 0.22049227246708644,
+      "grad_norm": 1.14417564868927,
+      "learning_rate": 1.8172209102464288e-05,
+      "loss": 1.089,
+      "step": 963
+    },
+    {
+      "epoch": 0.22072123640526617,
+      "grad_norm": 1.4543980360031128,
+      "learning_rate": 1.816793158402638e-05,
+      "loss": 1.1678,
+      "step": 964
+    },
+    {
+      "epoch": 0.2209502003434459,
+      "grad_norm": 1.025010108947754,
+      "learning_rate": 1.8163649570844198e-05,
+      "loss": 1.062,
+      "step": 965
+    },
+    {
+      "epoch": 0.22117916428162565,
+      "grad_norm": 1.0538175106048584,
+      "learning_rate": 1.8159363065274104e-05,
+      "loss": 1.1371,
+      "step": 966
+    },
+    {
+      "epoch": 0.22140812821980538,
+      "grad_norm": 1.9237316846847534,
+      "learning_rate": 1.8155072069674923e-05,
+      "loss": 1.1064,
+      "step": 967
+    },
+    {
+      "epoch": 0.2216370921579851,
+      "grad_norm": 1.3368388414382935,
+      "learning_rate": 1.8150776586407957e-05,
+      "loss": 1.0746,
+      "step": 968
+    },
+    {
+      "epoch": 0.22186605609616486,
+      "grad_norm": 1.297314167022705,
+      "learning_rate": 1.814647661783697e-05,
+      "loss": 1.1678,
+      "step": 969
+    },
+    {
+      "epoch": 0.2220950200343446,
+      "grad_norm": 1.2015776634216309,
+      "learning_rate": 1.8142172166328198e-05,
+      "loss": 1.1321,
+      "step": 970
+    },
+    {
+      "epoch": 0.22232398397252431,
+      "grad_norm": 1.1826279163360596,
+      "learning_rate": 1.8137863234250346e-05,
+      "loss": 1.0897,
+      "step": 971
+    },
+    {
+      "epoch": 0.22255294791070407,
+      "grad_norm": 1.9954915046691895,
+      "learning_rate": 1.813354982397459e-05,
+      "loss": 1.0966,
+      "step": 972
+    },
+    {
+      "epoch": 0.2227819118488838,
+      "grad_norm": 1.4948241710662842,
+      "learning_rate": 1.8129231937874555e-05,
+      "loss": 1.0648,
+      "step": 973
+    },
+    {
+      "epoch": 0.22301087578706355,
+      "grad_norm": 1.5026756525039673,
+      "learning_rate": 1.812490957832634e-05,
+      "loss": 1.1085,
+      "step": 974
+    },
+    {
+      "epoch": 0.22323983972524328,
+      "grad_norm": 1.6371681690216064,
+      "learning_rate": 1.8120582747708503e-05,
+      "loss": 1.085,
+      "step": 975
+    },
+    {
+      "epoch": 0.223468803663423,
+      "grad_norm": 1.236262559890747,
+      "learning_rate": 1.8116251448402062e-05,
+      "loss": 1.1093,
+      "step": 976
+    },
+    {
+      "epoch": 0.22369776760160276,
+      "grad_norm": 1.208084225654602,
+      "learning_rate": 1.8111915682790494e-05,
+      "loss": 1.192,
+      "step": 977
+    },
+    {
+      "epoch": 0.2239267315397825,
+      "grad_norm": 1.4316471815109253,
+      "learning_rate": 1.810757545325974e-05,
+      "loss": 1.113,
+      "step": 978
+    },
+    {
+      "epoch": 0.2241556954779622,
+      "grad_norm": 1.1135729551315308,
+      "learning_rate": 1.8103230762198185e-05,
+      "loss": 1.1644,
+      "step": 979
+    },
+    {
+      "epoch": 0.22438465941614197,
+      "grad_norm": 1.1902830600738525,
+      "learning_rate": 1.809888161199668e-05,
+      "loss": 1.1535,
+      "step": 980
+    },
+    {
+      "epoch": 0.2246136233543217,
+      "grad_norm": 1.5543137788772583,
+      "learning_rate": 1.8094528005048527e-05,
+      "loss": 1.077,
+      "step": 981
+    },
+    {
+      "epoch": 0.22484258729250142,
+      "grad_norm": 1.174572229385376,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 1.0657,
+      "step": 982
+    },
+    {
+      "epoch": 0.22507155123068118,
+      "grad_norm": 1.3976424932479858,
+      "learning_rate": 1.8085807430497734e-05,
+      "loss": 1.1355,
+      "step": 983
+    },
+    {
+      "epoch": 0.2253005151688609,
+      "grad_norm": 1.2085835933685303,
+      "learning_rate": 1.808144046769395e-05,
+      "loss": 1.2107,
+      "step": 984
+    },
+    {
+      "epoch": 0.22552947910704063,
+      "grad_norm": 1.3082224130630493,
+      "learning_rate": 1.8077069057741235e-05,
+      "loss": 1.2256,
+      "step": 985
+    },
+    {
+      "epoch": 0.22575844304522039,
+      "grad_norm": 1.0391132831573486,
+      "learning_rate": 1.807269320304514e-05,
+      "loss": 1.1506,
+      "step": 986
+    },
+    {
+      "epoch": 0.2259874069834001,
+      "grad_norm": 1.0561646223068237,
+      "learning_rate": 1.806831290601365e-05,
+      "loss": 1.1065,
+      "step": 987
+    },
+    {
+      "epoch": 0.22621637092157984,
+      "grad_norm": 1.2133493423461914,
+      "learning_rate": 1.8063928169057214e-05,
+      "loss": 1.1006,
+      "step": 988
+    },
+    {
+      "epoch": 0.2264453348597596,
+      "grad_norm": 1.4187254905700684,
+      "learning_rate": 1.8059538994588715e-05,
+      "loss": 1.0748,
+      "step": 989
+    },
+    {
+      "epoch": 0.22667429879793932,
+      "grad_norm": 1.3195863962173462,
+      "learning_rate": 1.8055145385023477e-05,
+      "loss": 1.1143,
+      "step": 990
+    },
+    {
+      "epoch": 0.22690326273611905,
+      "grad_norm": 1.304660677909851,
+      "learning_rate": 1.8050747342779274e-05,
+      "loss": 1.2163,
+      "step": 991
+    },
+    {
+      "epoch": 0.2271322266742988,
+      "grad_norm": 1.0397827625274658,
+      "learning_rate": 1.80463448702763e-05,
+      "loss": 1.0848,
+      "step": 992
+    },
+    {
+      "epoch": 0.22736119061247853,
+      "grad_norm": 1.2411673069000244,
+      "learning_rate": 1.8041937969937206e-05,
+      "loss": 1.0982,
+      "step": 993
+    },
+    {
+      "epoch": 0.22759015455065826,
+      "grad_norm": 1.261560320854187,
+      "learning_rate": 1.803752664418707e-05,
+      "loss": 1.096,
+      "step": 994
+    },
+    {
+      "epoch": 0.227819118488838,
+      "grad_norm": 1.3280490636825562,
+      "learning_rate": 1.803311089545341e-05,
+      "loss": 1.1291,
+      "step": 995
+    },
+    {
+      "epoch": 0.22804808242701774,
+      "grad_norm": 1.3528070449829102,
+      "learning_rate": 1.8028690726166172e-05,
+      "loss": 1.1006,
+      "step": 996
+    },
+    {
+      "epoch": 0.2282770463651975,
+      "grad_norm": 1.1263134479522705,
+      "learning_rate": 1.8024266138757746e-05,
+      "loss": 1.1311,
+      "step": 997
+    },
+    {
+      "epoch": 0.22850601030337722,
+      "grad_norm": 1.3571945428848267,
+      "learning_rate": 1.8019837135662932e-05,
+      "loss": 1.1336,
+      "step": 998
+    },
+    {
+      "epoch": 0.22873497424155695,
+      "grad_norm": 1.3748575448989868,
+      "learning_rate": 1.801540371931898e-05,
+      "loss": 1.1514,
+      "step": 999
+    },
+    {
+      "epoch": 0.2289639381797367,
+      "grad_norm": 1.1356844902038574,
+      "learning_rate": 1.8010965892165568e-05,
+      "loss": 1.1703,
+      "step": 1000
+    },
+    {
+      "epoch": 0.22919290211791643,
+      "grad_norm": 1.219716191291809,
+      "learning_rate": 1.8006523656644787e-05,
+      "loss": 1.1123,
+      "step": 1001
+    },
+    {
+      "epoch": 0.22942186605609616,
+      "grad_norm": 1.0204182863235474,
+      "learning_rate": 1.8002077015201164e-05,
+      "loss": 1.1016,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2296508299942759,
+      "grad_norm": 1.1001499891281128,
+      "learning_rate": 1.7997625970281652e-05,
+      "loss": 1.1063,
+      "step": 1003
+    },
+    {
+      "epoch": 0.22987979393245564,
+      "grad_norm": 1.3529033660888672,
+      "learning_rate": 1.7993170524335614e-05,
+      "loss": 1.1498,
+      "step": 1004
+    },
+    {
+      "epoch": 0.23010875787063537,
+      "grad_norm": 1.0929923057556152,
+      "learning_rate": 1.7988710679814857e-05,
+      "loss": 1.1105,
+      "step": 1005
+    },
+    {
+      "epoch": 0.23033772180881512,
+      "grad_norm": 1.1180248260498047,
+      "learning_rate": 1.798424643917359e-05,
+      "loss": 1.0907,
+      "step": 1006
+    },
+    {
+      "epoch": 0.23056668574699485,
+      "grad_norm": 1.199679970741272,
+      "learning_rate": 1.7979777804868447e-05,
+      "loss": 1.184,
+      "step": 1007
+    },
+    {
+      "epoch": 0.23079564968517458,
+      "grad_norm": 1.0665282011032104,
+      "learning_rate": 1.797530477935848e-05,
+      "loss": 1.1398,
+      "step": 1008
+    },
+    {
+      "epoch": 0.23102461362335433,
+      "grad_norm": 1.1431396007537842,
+      "learning_rate": 1.7970827365105157e-05,
+      "loss": 1.1085,
+      "step": 1009
+    },
+    {
+      "epoch": 0.23125357756153406,
+      "grad_norm": 1.1866995096206665,
+      "learning_rate": 1.796634556457236e-05,
+      "loss": 1.1075,
+      "step": 1010
+    },
+    {
+      "epoch": 0.23148254149971378,
+      "grad_norm": 1.0268393754959106,
+      "learning_rate": 1.7961859380226395e-05,
+      "loss": 1.1128,
+      "step": 1011
+    },
+    {
+      "epoch": 0.23171150543789354,
+      "grad_norm": 1.1771196126937866,
+      "learning_rate": 1.7957368814535963e-05,
+      "loss": 1.0616,
+      "step": 1012
+    },
+    {
+      "epoch": 0.23194046937607327,
+      "grad_norm": 1.1519027948379517,
+      "learning_rate": 1.7952873869972183e-05,
+      "loss": 1.1305,
+      "step": 1013
+    },
+    {
+      "epoch": 0.232169433314253,
+      "grad_norm": 1.0065557956695557,
+      "learning_rate": 1.7948374549008594e-05,
+      "loss": 1.088,
+      "step": 1014
+    },
+    {
+      "epoch": 0.23239839725243275,
+      "grad_norm": 1.4946367740631104,
+      "learning_rate": 1.7943870854121126e-05,
+      "loss": 1.1938,
+      "step": 1015
+    },
+    {
+      "epoch": 0.23262736119061247,
+      "grad_norm": 1.5711524486541748,
+      "learning_rate": 1.793936278778813e-05,
+      "loss": 1.0989,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2328563251287922,
+      "grad_norm": 1.0987305641174316,
+      "learning_rate": 1.793485035249036e-05,
+      "loss": 1.1015,
+      "step": 1017
+    },
+    {
+      "epoch": 0.23308528906697196,
+      "grad_norm": 1.1849825382232666,
+      "learning_rate": 1.793033355071096e-05,
+      "loss": 1.1036,
+      "step": 1018
+    },
+    {
+      "epoch": 0.23331425300515168,
+      "grad_norm": 1.1868687868118286,
+      "learning_rate": 1.79258123849355e-05,
+      "loss": 1.0985,
+      "step": 1019
+    },
+    {
+      "epoch": 0.23354321694333144,
+      "grad_norm": 1.2682559490203857,
+      "learning_rate": 1.792128685765193e-05,
+      "loss": 1.1574,
+      "step": 1020
+    },
+    {
+      "epoch": 0.23377218088151117,
+      "grad_norm": 1.544455647468567,
+      "learning_rate": 1.7916756971350618e-05,
+      "loss": 1.1294,
+      "step": 1021
+    },
+    {
+      "epoch": 0.2340011448196909,
+      "grad_norm": 1.103712558746338,
+      "learning_rate": 1.7912222728524318e-05,
+      "loss": 1.0987,
+      "step": 1022
+    },
+    {
+      "epoch": 0.23423010875787065,
+      "grad_norm": 1.2829042673110962,
+      "learning_rate": 1.7907684131668186e-05,
+      "loss": 1.0943,
+      "step": 1023
+    },
+    {
+      "epoch": 0.23445907269605037,
+      "grad_norm": 1.112669587135315,
+      "learning_rate": 1.7903141183279776e-05,
+      "loss": 1.1518,
+      "step": 1024
+    },
+    {
+      "epoch": 0.2346880366342301,
+      "grad_norm": 1.0655471086502075,
+      "learning_rate": 1.789859388585903e-05,
+      "loss": 1.1347,
+      "step": 1025
+    },
+    {
+      "epoch": 0.23491700057240986,
+      "grad_norm": 1.2130173444747925,
+      "learning_rate": 1.7894042241908293e-05,
+      "loss": 1.1555,
+      "step": 1026
+    },
+    {
+      "epoch": 0.23514596451058958,
+      "grad_norm": 1.1426712274551392,
+      "learning_rate": 1.7889486253932292e-05,
+      "loss": 1.1248,
+      "step": 1027
+    },
+    {
+      "epoch": 0.2353749284487693,
+      "grad_norm": 1.0837373733520508,
+      "learning_rate": 1.7884925924438152e-05,
+      "loss": 1.1218,
+      "step": 1028
+    },
+    {
+      "epoch": 0.23560389238694907,
+      "grad_norm": 1.2106484174728394,
+      "learning_rate": 1.7880361255935385e-05,
+      "loss": 1.1164,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2358328563251288,
+      "grad_norm": 1.155698537826538,
+      "learning_rate": 1.7875792250935883e-05,
+      "loss": 1.1404,
+      "step": 1030
+    },
+    {
+      "epoch": 0.23606182026330852,
+      "grad_norm": 1.0495966672897339,
+      "learning_rate": 1.7871218911953942e-05,
+      "loss": 1.1275,
+      "step": 1031
+    },
+    {
+      "epoch": 0.23629078420148827,
+      "grad_norm": 1.108055591583252,
+      "learning_rate": 1.7866641241506222e-05,
+      "loss": 1.0906,
+      "step": 1032
+    },
+    {
+      "epoch": 0.236519748139668,
+      "grad_norm": 1.1889424324035645,
+      "learning_rate": 1.7862059242111782e-05,
+      "loss": 1.1072,
+      "step": 1033
+    },
+    {
+      "epoch": 0.23674871207784773,
+      "grad_norm": 2.039444923400879,
+      "learning_rate": 1.7857472916292056e-05,
+      "loss": 1.1747,
+      "step": 1034
+    },
+    {
+      "epoch": 0.23697767601602748,
+      "grad_norm": 1.0959917306900024,
+      "learning_rate": 1.785288226657086e-05,
+      "loss": 1.0886,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2372066399542072,
+      "grad_norm": 1.1916193962097168,
+      "learning_rate": 1.7848287295474397e-05,
+      "loss": 1.1372,
+      "step": 1036
+    },
+    {
+      "epoch": 0.23743560389238694,
+      "grad_norm": 1.3053346872329712,
+      "learning_rate": 1.7843688005531227e-05,
+      "loss": 1.1423,
+      "step": 1037
+    },
+    {
+      "epoch": 0.2376645678305667,
+      "grad_norm": 1.1222914457321167,
+      "learning_rate": 1.7839084399272317e-05,
+      "loss": 1.079,
+      "step": 1038
+    },
+    {
+      "epoch": 0.23789353176874642,
+      "grad_norm": 1.1740766763687134,
+      "learning_rate": 1.7834476479230978e-05,
+      "loss": 1.163,
+      "step": 1039
+    },
+    {
+      "epoch": 0.23812249570692615,
+      "grad_norm": 2.025049924850464,
+      "learning_rate": 1.782986424794292e-05,
+      "loss": 1.1855,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2383514596451059,
+      "grad_norm": 1.2098287343978882,
+      "learning_rate": 1.7825247707946212e-05,
+      "loss": 1.1989,
+      "step": 1041
+    },
+    {
+      "epoch": 0.23858042358328563,
+      "grad_norm": 1.0993343591690063,
+      "learning_rate": 1.78206268617813e-05,
+      "loss": 1.1303,
+      "step": 1042
+    },
+    {
+      "epoch": 0.23880938752146538,
+      "grad_norm": 1.1957361698150635,
+      "learning_rate": 1.781600171199099e-05,
+      "loss": 1.1391,
+      "step": 1043
+    },
+    {
+      "epoch": 0.2390383514596451,
+      "grad_norm": 1.0565906763076782,
+      "learning_rate": 1.7811372261120468e-05,
+      "loss": 1.0929,
+      "step": 1044
+    },
+    {
+      "epoch": 0.23926731539782484,
+      "grad_norm": 1.4460198879241943,
+      "learning_rate": 1.780673851171728e-05,
+      "loss": 1.1438,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2394962793360046,
+      "grad_norm": 1.2659794092178345,
+      "learning_rate": 1.7802100466331343e-05,
+      "loss": 1.1188,
+      "step": 1046
+    },
+    {
+      "epoch": 0.23972524327418432,
+      "grad_norm": 1.1114025115966797,
+      "learning_rate": 1.7797458127514934e-05,
+      "loss": 1.1157,
+      "step": 1047
+    },
+    {
+      "epoch": 0.23995420721236405,
+      "grad_norm": 2.1384947299957275,
+      "learning_rate": 1.779281149782269e-05,
+      "loss": 1.0959,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2401831711505438,
+      "grad_norm": 1.1005151271820068,
+      "learning_rate": 1.7788160579811614e-05,
+      "loss": 1.0842,
+      "step": 1049
+    },
+    {
+      "epoch": 0.24041213508872353,
+      "grad_norm": 1.1792573928833008,
+      "learning_rate": 1.7783505376041063e-05,
+      "loss": 1.0822,
+      "step": 1050
+    },
+    {
+      "epoch": 0.24064109902690325,
+      "grad_norm": 1.1286247968673706,
+      "learning_rate": 1.7778845889072764e-05,
+      "loss": 1.1103,
+      "step": 1051
+    },
+    {
+      "epoch": 0.240870062965083,
+      "grad_norm": 1.0358811616897583,
+      "learning_rate": 1.777418212147079e-05,
+      "loss": 1.0796,
+      "step": 1052
+    },
+    {
+      "epoch": 0.24109902690326274,
+      "grad_norm": 1.1496690511703491,
+      "learning_rate": 1.7769514075801573e-05,
+      "loss": 1.1426,
+      "step": 1053
+    },
+    {
+      "epoch": 0.24132799084144246,
+      "grad_norm": 0.995430588722229,
+      "learning_rate": 1.77648417546339e-05,
+      "loss": 1.0874,
+      "step": 1054
+    },
+    {
+      "epoch": 0.24155695477962222,
+      "grad_norm": 1.176997423171997,
+      "learning_rate": 1.7760165160538907e-05,
+      "loss": 1.152,
+      "step": 1055
+    },
+    {
+      "epoch": 0.24178591871780195,
+      "grad_norm": 1.2487221956253052,
+      "learning_rate": 1.775548429609009e-05,
+      "loss": 1.1166,
+      "step": 1056
+    },
+    {
+      "epoch": 0.24201488265598167,
+      "grad_norm": 1.1057595014572144,
+      "learning_rate": 1.7750799163863287e-05,
+      "loss": 1.1729,
+      "step": 1057
+    },
+    {
+      "epoch": 0.24224384659416143,
+      "grad_norm": 1.2305538654327393,
+      "learning_rate": 1.774610976643669e-05,
+      "loss": 1.1503,
+      "step": 1058
+    },
+    {
+      "epoch": 0.24247281053234115,
+      "grad_norm": 1.1498816013336182,
+      "learning_rate": 1.7741416106390828e-05,
+      "loss": 1.0871,
+      "step": 1059
+    },
+    {
+      "epoch": 0.24270177447052088,
+      "grad_norm": 1.7206319570541382,
+      "learning_rate": 1.7736718186308585e-05,
+      "loss": 1.1571,
+      "step": 1060
+    },
+    {
+      "epoch": 0.24293073840870064,
+      "grad_norm": 1.1156868934631348,
+      "learning_rate": 1.7732016008775193e-05,
+      "loss": 1.087,
+      "step": 1061
+    },
+    {
+      "epoch": 0.24315970234688036,
+      "grad_norm": 1.4100158214569092,
+      "learning_rate": 1.7727309576378213e-05,
+      "loss": 1.1383,
+      "step": 1062
+    },
+    {
+      "epoch": 0.2433886662850601,
+      "grad_norm": 1.3357410430908203,
+      "learning_rate": 1.7722598891707562e-05,
+      "loss": 1.1269,
+      "step": 1063
+    },
+    {
+      "epoch": 0.24361763022323984,
+      "grad_norm": 2.102534294128418,
+      "learning_rate": 1.7717883957355484e-05,
+      "loss": 1.0601,
+      "step": 1064
+    },
+    {
+      "epoch": 0.24384659416141957,
+      "grad_norm": 1.1543803215026855,
+      "learning_rate": 1.7713164775916573e-05,
+      "loss": 1.1185,
+      "step": 1065
+    },
+    {
+      "epoch": 0.24407555809959933,
+      "grad_norm": 1.2533857822418213,
+      "learning_rate": 1.7708441349987753e-05,
+      "loss": 1.0979,
+      "step": 1066
+    },
+    {
+      "epoch": 0.24430452203777905,
+      "grad_norm": 1.1019551753997803,
+      "learning_rate": 1.7703713682168288e-05,
+      "loss": 1.0778,
+      "step": 1067
+    },
+    {
+      "epoch": 0.24453348597595878,
+      "grad_norm": 1.085174798965454,
+      "learning_rate": 1.7698981775059767e-05,
+      "loss": 1.1028,
+      "step": 1068
+    },
+    {
+      "epoch": 0.24476244991413854,
+      "grad_norm": 1.204268217086792,
+      "learning_rate": 1.7694245631266124e-05,
+      "loss": 1.0967,
+      "step": 1069
+    },
+    {
+      "epoch": 0.24499141385231826,
+      "grad_norm": 1.129050850868225,
+      "learning_rate": 1.768950525339362e-05,
+      "loss": 1.0997,
+      "step": 1070
+    },
+    {
+      "epoch": 0.245220377790498,
+      "grad_norm": 1.1998200416564941,
+      "learning_rate": 1.7684760644050847e-05,
+      "loss": 1.0759,
+      "step": 1071
+    },
+    {
+      "epoch": 0.24544934172867774,
+      "grad_norm": 2.0567846298217773,
+      "learning_rate": 1.7680011805848715e-05,
+      "loss": 1.0642,
+      "step": 1072
+    },
+    {
+      "epoch": 0.24567830566685747,
+      "grad_norm": 1.1932883262634277,
+      "learning_rate": 1.767525874140048e-05,
+      "loss": 1.0918,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2459072696050372,
+      "grad_norm": 1.18972909450531,
+      "learning_rate": 1.7670501453321705e-05,
+      "loss": 1.0716,
+      "step": 1074
+    },
+    {
+      "epoch": 0.24613623354321695,
+      "grad_norm": 4.0288825035095215,
+      "learning_rate": 1.7665739944230296e-05,
+      "loss": 1.1541,
+      "step": 1075
+    },
+    {
+      "epoch": 0.24636519748139668,
+      "grad_norm": 1.1446605920791626,
+      "learning_rate": 1.766097421674646e-05,
+      "loss": 1.0506,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2465941614195764,
+      "grad_norm": 0.995333731174469,
+      "learning_rate": 1.7656204273492746e-05,
+      "loss": 1.115,
+      "step": 1077
+    },
+    {
+      "epoch": 0.24682312535775616,
+      "grad_norm": 1.2963595390319824,
+      "learning_rate": 1.7651430117094005e-05,
+      "loss": 1.1066,
+      "step": 1078
+    },
+    {
+      "epoch": 0.2470520892959359,
+      "grad_norm": 1.3206766843795776,
+      "learning_rate": 1.7646651750177424e-05,
+      "loss": 1.1593,
+      "step": 1079
+    },
+    {
+      "epoch": 0.24728105323411562,
+      "grad_norm": 1.1858936548233032,
+      "learning_rate": 1.7641869175372493e-05,
+      "loss": 1.1485,
+      "step": 1080
+    },
+    {
+      "epoch": 0.24751001717229537,
+      "grad_norm": 1.3568990230560303,
+      "learning_rate": 1.7637082395311024e-05,
+      "loss": 1.1467,
+      "step": 1081
+    },
+    {
+      "epoch": 0.2477389811104751,
+      "grad_norm": 1.4126697778701782,
+      "learning_rate": 1.7632291412627146e-05,
+      "loss": 1.1229,
+      "step": 1082
+    },
+    {
+      "epoch": 0.24796794504865483,
+      "grad_norm": 1.0935779809951782,
+      "learning_rate": 1.7627496229957288e-05,
+      "loss": 1.1371,
+      "step": 1083
+    },
+    {
+      "epoch": 0.24819690898683458,
+      "grad_norm": 1.2460694313049316,
+      "learning_rate": 1.7622696849940204e-05,
+      "loss": 1.0996,
+      "step": 1084
+    },
+    {
+      "epoch": 0.2484258729250143,
+      "grad_norm": 1.2605793476104736,
+      "learning_rate": 1.7617893275216953e-05,
+      "loss": 1.0923,
+      "step": 1085
+    },
+    {
+      "epoch": 0.24865483686319403,
+      "grad_norm": 1.437913417816162,
+      "learning_rate": 1.76130855084309e-05,
+      "loss": 1.1114,
+      "step": 1086
+    },
+    {
+      "epoch": 0.2488838008013738,
+      "grad_norm": 1.0823016166687012,
+      "learning_rate": 1.7608273552227723e-05,
+      "loss": 1.1255,
+      "step": 1087
+    },
+    {
+      "epoch": 0.24911276473955352,
+      "grad_norm": 1.158138632774353,
+      "learning_rate": 1.7603457409255397e-05,
+      "loss": 1.12,
+      "step": 1088
+    },
+    {
+      "epoch": 0.24934172867773327,
+      "grad_norm": 1.2879137992858887,
+      "learning_rate": 1.7598637082164204e-05,
+      "loss": 1.0848,
+      "step": 1089
+    },
+    {
+      "epoch": 0.249570692615913,
+      "grad_norm": 1.2041521072387695,
+      "learning_rate": 1.759381257360673e-05,
+      "loss": 1.1246,
+      "step": 1090
+    },
+    {
+      "epoch": 0.24979965655409272,
+      "grad_norm": 1.3644989728927612,
+      "learning_rate": 1.7588983886237868e-05,
+      "loss": 1.0443,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2500286204922725,
+      "grad_norm": 1.232747197151184,
+      "learning_rate": 1.75841510227148e-05,
+      "loss": 1.131,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2502575844304522,
+      "grad_norm": 1.1022595167160034,
+      "learning_rate": 1.7579313985697006e-05,
+      "loss": 1.0583,
+      "step": 1093
+    },
+    {
+      "epoch": 0.25048654836863193,
+      "grad_norm": 1.0636167526245117,
+      "learning_rate": 1.7574472777846276e-05,
+      "loss": 1.0952,
+      "step": 1094
+    },
+    {
+      "epoch": 0.25071551230681166,
+      "grad_norm": 1.6139863729476929,
+      "learning_rate": 1.7569627401826673e-05,
+      "loss": 1.073,
+      "step": 1095
+    },
+    {
+      "epoch": 0.2509444762449914,
+      "grad_norm": 1.2776731252670288,
+      "learning_rate": 1.756477786030458e-05,
+      "loss": 1.0649,
+      "step": 1096
+    },
+    {
+      "epoch": 0.25117344018317117,
+      "grad_norm": 1.2996584177017212,
+      "learning_rate": 1.7559924155948652e-05,
+      "loss": 1.0911,
+      "step": 1097
+    },
+    {
+      "epoch": 0.2514024041213509,
+      "grad_norm": 1.0315338373184204,
+      "learning_rate": 1.7555066291429845e-05,
+      "loss": 1.0903,
+      "step": 1098
+    },
+    {
+      "epoch": 0.2516313680595306,
+      "grad_norm": 1.1848983764648438,
+      "learning_rate": 1.7550204269421395e-05,
+      "loss": 1.1727,
+      "step": 1099
+    },
+    {
+      "epoch": 0.25186033199771035,
+      "grad_norm": 1.332460880279541,
+      "learning_rate": 1.754533809259884e-05,
+      "loss": 1.1459,
+      "step": 1100
+    },
+    {
+      "epoch": 0.2520892959358901,
+      "grad_norm": 1.4366538524627686,
+      "learning_rate": 1.7540467763639994e-05,
+      "loss": 1.163,
+      "step": 1101
+    },
+    {
+      "epoch": 0.25231825987406986,
+      "grad_norm": 1.1349256038665771,
+      "learning_rate": 1.7535593285224958e-05,
+      "loss": 1.1246,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2525472238122496,
+      "grad_norm": 1.2521072626113892,
+      "learning_rate": 1.7530714660036112e-05,
+      "loss": 1.147,
+      "step": 1103
+    },
+    {
+      "epoch": 0.2527761877504293,
+      "grad_norm": 1.1671040058135986,
+      "learning_rate": 1.7525831890758125e-05,
+      "loss": 1.1429,
+      "step": 1104
+    },
+    {
+      "epoch": 0.25300515168860904,
+      "grad_norm": 1.0925365686416626,
+      "learning_rate": 1.7520944980077948e-05,
+      "loss": 1.1264,
+      "step": 1105
+    },
+    {
+      "epoch": 0.25323411562678877,
+      "grad_norm": 1.1112512350082397,
+      "learning_rate": 1.7516053930684804e-05,
+      "loss": 1.1713,
+      "step": 1106
+    },
+    {
+      "epoch": 0.2534630795649685,
+      "grad_norm": 1.217628002166748,
+      "learning_rate": 1.7511158745270197e-05,
+      "loss": 1.0978,
+      "step": 1107
+    },
+    {
+      "epoch": 0.2536920435031483,
+      "grad_norm": 1.3462601900100708,
+      "learning_rate": 1.7506259426527903e-05,
+      "loss": 1.0702,
+      "step": 1108
+    },
+    {
+      "epoch": 0.253921007441328,
+      "grad_norm": 1.3914707899093628,
+      "learning_rate": 1.750135597715398e-05,
+      "loss": 1.1011,
+      "step": 1109
+    },
+    {
+      "epoch": 0.25414997137950773,
+      "grad_norm": 1.2635843753814697,
+      "learning_rate": 1.7496448399846757e-05,
+      "loss": 1.047,
+      "step": 1110
+    },
+    {
+      "epoch": 0.25437893531768746,
+      "grad_norm": 1.217361330986023,
+      "learning_rate": 1.7491536697306828e-05,
+      "loss": 1.1323,
+      "step": 1111
+    },
+    {
+      "epoch": 0.2546078992558672,
+      "grad_norm": 1.3858387470245361,
+      "learning_rate": 1.7486620872237064e-05,
+      "loss": 1.1033,
+      "step": 1112
+    },
+    {
+      "epoch": 0.2548368631940469,
+      "grad_norm": 1.247519612312317,
+      "learning_rate": 1.7481700927342603e-05,
+      "loss": 1.1039,
+      "step": 1113
+    },
+    {
+      "epoch": 0.2550658271322267,
+      "grad_norm": 1.3646209239959717,
+      "learning_rate": 1.7476776865330847e-05,
+      "loss": 1.12,
+      "step": 1114
+    },
+    {
+      "epoch": 0.2552947910704064,
+      "grad_norm": 1.1417921781539917,
+      "learning_rate": 1.7471848688911465e-05,
+      "loss": 1.1152,
+      "step": 1115
+    },
+    {
+      "epoch": 0.25552375500858615,
+      "grad_norm": 2.9017693996429443,
+      "learning_rate": 1.7466916400796398e-05,
+      "loss": 1.117,
+      "step": 1116
+    },
+    {
+      "epoch": 0.2557527189467659,
+      "grad_norm": 1.4879010915756226,
+      "learning_rate": 1.7461980003699835e-05,
+      "loss": 1.0987,
+      "step": 1117
+    },
+    {
+      "epoch": 0.2559816828849456,
+      "grad_norm": 1.6178057193756104,
+      "learning_rate": 1.7457039500338238e-05,
+      "loss": 1.1345,
+      "step": 1118
+    },
+    {
+      "epoch": 0.25621064682312533,
+      "grad_norm": 1.2162644863128662,
+      "learning_rate": 1.7452094893430324e-05,
+      "loss": 1.1063,
+      "step": 1119
+    },
+    {
+      "epoch": 0.2564396107613051,
+      "grad_norm": 1.2172150611877441,
+      "learning_rate": 1.744714618569707e-05,
+      "loss": 1.1393,
+      "step": 1120
+    },
+    {
+      "epoch": 0.25666857469948484,
+      "grad_norm": 1.195317029953003,
+      "learning_rate": 1.74421933798617e-05,
+      "loss": 1.1355,
+      "step": 1121
+    },
+    {
+      "epoch": 0.25689753863766457,
+      "grad_norm": 1.0217382907867432,
+      "learning_rate": 1.7437236478649718e-05,
+      "loss": 1.1258,
+      "step": 1122
+    },
+    {
+      "epoch": 0.2571265025758443,
+      "grad_norm": 1.2835495471954346,
+      "learning_rate": 1.7432275484788852e-05,
+      "loss": 1.1104,
+      "step": 1123
+    },
+    {
+      "epoch": 0.257355466514024,
+      "grad_norm": 1.0786269903182983,
+      "learning_rate": 1.74273104010091e-05,
+      "loss": 1.066,
+      "step": 1124
+    },
+    {
+      "epoch": 0.2575844304522038,
+      "grad_norm": 1.1824384927749634,
+      "learning_rate": 1.74223412300427e-05,
+      "loss": 1.0641,
+      "step": 1125
+    },
+    {
+      "epoch": 0.25781339439038353,
+      "grad_norm": 1.0312379598617554,
+      "learning_rate": 1.7417367974624153e-05,
+      "loss": 1.0843,
+      "step": 1126
+    },
+    {
+      "epoch": 0.25804235832856326,
+      "grad_norm": 1.2044585943222046,
+      "learning_rate": 1.74123906374902e-05,
+      "loss": 1.0899,
+      "step": 1127
+    },
+    {
+      "epoch": 0.258271322266743,
+      "grad_norm": 1.1857991218566895,
+      "learning_rate": 1.740740922137982e-05,
+      "loss": 1.1682,
+      "step": 1128
+    },
+    {
+      "epoch": 0.2585002862049227,
+      "grad_norm": 1.4564883708953857,
+      "learning_rate": 1.7402423729034252e-05,
+      "loss": 1.1455,
+      "step": 1129
+    },
+    {
+      "epoch": 0.25872925014310244,
+      "grad_norm": 1.2754067182540894,
+      "learning_rate": 1.7397434163196967e-05,
+      "loss": 1.1081,
+      "step": 1130
+    },
+    {
+      "epoch": 0.2589582140812822,
+      "grad_norm": 1.149965763092041,
+      "learning_rate": 1.7392440526613684e-05,
+      "loss": 1.1089,
+      "step": 1131
+    },
+    {
+      "epoch": 0.25918717801946195,
+      "grad_norm": 1.430191993713379,
+      "learning_rate": 1.7387442822032354e-05,
+      "loss": 1.1204,
+      "step": 1132
+    },
+    {
+      "epoch": 0.2594161419576417,
+      "grad_norm": 1.3282990455627441,
+      "learning_rate": 1.738244105220318e-05,
+      "loss": 1.1386,
+      "step": 1133
+    },
+    {
+      "epoch": 0.2596451058958214,
+      "grad_norm": 1.1032613515853882,
+      "learning_rate": 1.7377435219878586e-05,
+      "loss": 1.0563,
+      "step": 1134
+    },
+    {
+      "epoch": 0.25987406983400113,
+      "grad_norm": 1.0498651266098022,
+      "learning_rate": 1.737242532781324e-05,
+      "loss": 1.104,
+      "step": 1135
+    },
+    {
+      "epoch": 0.26010303377218086,
+      "grad_norm": 1.2936218976974487,
+      "learning_rate": 1.736741137876405e-05,
+      "loss": 1.043,
+      "step": 1136
+    },
+    {
+      "epoch": 0.26033199771036064,
+      "grad_norm": 1.025429129600525,
+      "learning_rate": 1.736239337549015e-05,
+      "loss": 1.1479,
+      "step": 1137
+    },
+    {
+      "epoch": 0.26056096164854037,
+      "grad_norm": 1.1074275970458984,
+      "learning_rate": 1.7357371320752896e-05,
+      "loss": 1.0875,
+      "step": 1138
+    },
+    {
+      "epoch": 0.2607899255867201,
+      "grad_norm": 1.2737194299697876,
+      "learning_rate": 1.735234521731589e-05,
+      "loss": 1.1193,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2610188895248998,
+      "grad_norm": 2.4709835052490234,
+      "learning_rate": 1.7347315067944955e-05,
+      "loss": 1.1418,
+      "step": 1140
+    },
+    {
+      "epoch": 0.26124785346307955,
+      "grad_norm": 1.4513455629348755,
+      "learning_rate": 1.7342280875408138e-05,
+      "loss": 1.1594,
+      "step": 1141
+    },
+    {
+      "epoch": 0.2614768174012593,
+      "grad_norm": 1.1992151737213135,
+      "learning_rate": 1.7337242642475712e-05,
+      "loss": 1.0609,
+      "step": 1142
+    },
+    {
+      "epoch": 0.26170578133943906,
+      "grad_norm": 1.7710577249526978,
+      "learning_rate": 1.7332200371920173e-05,
+      "loss": 1.1702,
+      "step": 1143
+    },
+    {
+      "epoch": 0.2619347452776188,
+      "grad_norm": 1.3110970258712769,
+      "learning_rate": 1.7327154066516244e-05,
+      "loss": 1.0921,
+      "step": 1144
+    },
+    {
+      "epoch": 0.2621637092157985,
+      "grad_norm": 1.0681921243667603,
+      "learning_rate": 1.7322103729040868e-05,
+      "loss": 1.038,
+      "step": 1145
+    },
+    {
+      "epoch": 0.26239267315397824,
+      "grad_norm": 1.1642048358917236,
+      "learning_rate": 1.731704936227319e-05,
+      "loss": 1.071,
+      "step": 1146
+    },
+    {
+      "epoch": 0.26262163709215797,
+      "grad_norm": 1.287327766418457,
+      "learning_rate": 1.7311990968994598e-05,
+      "loss": 1.102,
+      "step": 1147
+    },
+    {
+      "epoch": 0.26285060103033775,
+      "grad_norm": 1.2213733196258545,
+      "learning_rate": 1.7306928551988683e-05,
+      "loss": 1.1354,
+      "step": 1148
+    },
+    {
+      "epoch": 0.2630795649685175,
+      "grad_norm": 1.3054442405700684,
+      "learning_rate": 1.7301862114041244e-05,
+      "loss": 1.1189,
+      "step": 1149
+    },
+    {
+      "epoch": 0.2633085289066972,
+      "grad_norm": 1.8000365495681763,
+      "learning_rate": 1.72967916579403e-05,
+      "loss": 1.105,
+      "step": 1150
+    },
+    {
+      "epoch": 0.26353749284487693,
+      "grad_norm": 1.155625343322754,
+      "learning_rate": 1.7291717186476088e-05,
+      "loss": 1.1213,
+      "step": 1151
+    },
+    {
+      "epoch": 0.26376645678305666,
+      "grad_norm": 1.052958607673645,
+      "learning_rate": 1.7286638702441037e-05,
+      "loss": 1.1383,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2639954207212364,
+      "grad_norm": 1.0741239786148071,
+      "learning_rate": 1.7281556208629802e-05,
+      "loss": 1.1085,
+      "step": 1153
+    },
+    {
+      "epoch": 0.26422438465941617,
+      "grad_norm": 1.1770488023757935,
+      "learning_rate": 1.7276469707839235e-05,
+      "loss": 1.0881,
+      "step": 1154
+    },
+    {
+      "epoch": 0.2644533485975959,
+      "grad_norm": 1.1173039674758911,
+      "learning_rate": 1.7271379202868394e-05,
+      "loss": 1.162,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2646823125357756,
+      "grad_norm": 1.1196292638778687,
+      "learning_rate": 1.7266284696518537e-05,
+      "loss": 1.1199,
+      "step": 1156
+    },
+    {
+      "epoch": 0.26491127647395535,
+      "grad_norm": 1.097383975982666,
+      "learning_rate": 1.7261186191593135e-05,
+      "loss": 1.1347,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2651402404121351,
+      "grad_norm": 2.0568525791168213,
+      "learning_rate": 1.725608369089785e-05,
+      "loss": 1.0951,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2653692043503148,
+      "grad_norm": 1.1759741306304932,
+      "learning_rate": 1.7250977197240545e-05,
+      "loss": 1.1429,
+      "step": 1159
+    },
+    {
+      "epoch": 0.2655981682884946,
+      "grad_norm": 1.1705988645553589,
+      "learning_rate": 1.7245866713431278e-05,
+      "loss": 1.0861,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2658271322266743,
+      "grad_norm": 1.197718620300293,
+      "learning_rate": 1.724075224228231e-05,
+      "loss": 1.1353,
+      "step": 1161
+    },
+    {
+      "epoch": 0.26605609616485404,
+      "grad_norm": 1.0807304382324219,
+      "learning_rate": 1.7235633786608092e-05,
+      "loss": 1.1407,
+      "step": 1162
+    },
+    {
+      "epoch": 0.26628506010303377,
+      "grad_norm": 1.2779772281646729,
+      "learning_rate": 1.723051134922526e-05,
+      "loss": 1.1031,
+      "step": 1163
+    },
+    {
+      "epoch": 0.2665140240412135,
+      "grad_norm": 1.2724177837371826,
+      "learning_rate": 1.7225384932952655e-05,
+      "loss": 1.0757,
+      "step": 1164
+    },
+    {
+      "epoch": 0.2667429879793932,
+      "grad_norm": 1.0801130533218384,
+      "learning_rate": 1.72202545406113e-05,
+      "loss": 1.1328,
+      "step": 1165
+    },
+    {
+      "epoch": 0.266971951917573,
+      "grad_norm": 1.0200031995773315,
+      "learning_rate": 1.7215120175024405e-05,
+      "loss": 1.0876,
+      "step": 1166
+    },
+    {
+      "epoch": 0.26720091585575273,
+      "grad_norm": 1.1842114925384521,
+      "learning_rate": 1.720998183901737e-05,
+      "loss": 1.1401,
+      "step": 1167
+    },
+    {
+      "epoch": 0.26742987979393246,
+      "grad_norm": 1.0764400959014893,
+      "learning_rate": 1.7204839535417778e-05,
+      "loss": 1.046,
+      "step": 1168
+    },
+    {
+      "epoch": 0.2676588437321122,
+      "grad_norm": 1.2766287326812744,
+      "learning_rate": 1.7199693267055392e-05,
+      "loss": 1.0931,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2678878076702919,
+      "grad_norm": 1.282101035118103,
+      "learning_rate": 1.7194543036762165e-05,
+      "loss": 1.0584,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2681167716084717,
+      "grad_norm": 1.2437055110931396,
+      "learning_rate": 1.7189388847372227e-05,
+      "loss": 1.1001,
+      "step": 1171
+    },
+    {
+      "epoch": 0.2683457355466514,
+      "grad_norm": 1.0838711261749268,
+      "learning_rate": 1.7184230701721876e-05,
+      "loss": 1.0871,
+      "step": 1172
+    },
+    {
+      "epoch": 0.26857469948483115,
+      "grad_norm": 1.4530155658721924,
+      "learning_rate": 1.717906860264961e-05,
+      "loss": 1.1225,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2688036634230109,
+      "grad_norm": 1.1978572607040405,
+      "learning_rate": 1.7173902552996075e-05,
+      "loss": 1.1464,
+      "step": 1174
+    },
+    {
+      "epoch": 0.2690326273611906,
+      "grad_norm": 1.1811645030975342,
+      "learning_rate": 1.7168732555604114e-05,
+      "loss": 1.1627,
+      "step": 1175
+    },
+    {
+      "epoch": 0.26926159129937033,
+      "grad_norm": 1.530895709991455,
+      "learning_rate": 1.716355861331873e-05,
+      "loss": 1.0715,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2694905552375501,
+      "grad_norm": 1.2078356742858887,
+      "learning_rate": 1.7158380728987102e-05,
+      "loss": 1.1287,
+      "step": 1177
+    },
+    {
+      "epoch": 0.26971951917572984,
+      "grad_norm": 1.318663477897644,
+      "learning_rate": 1.715319890545857e-05,
+      "loss": 1.1881,
+      "step": 1178
+    },
+    {
+      "epoch": 0.26994848311390957,
+      "grad_norm": 1.174576759338379,
+      "learning_rate": 1.7148013145584657e-05,
+      "loss": 1.1113,
+      "step": 1179
+    },
+    {
+      "epoch": 0.2701774470520893,
+      "grad_norm": 1.2301723957061768,
+      "learning_rate": 1.7142823452219036e-05,
+      "loss": 1.1051,
+      "step": 1180
+    },
+    {
+      "epoch": 0.270406410990269,
+      "grad_norm": 1.097507119178772,
+      "learning_rate": 1.7137629828217556e-05,
+      "loss": 1.0563,
+      "step": 1181
+    },
+    {
+      "epoch": 0.27063537492844875,
+      "grad_norm": 1.100091814994812,
+      "learning_rate": 1.7132432276438228e-05,
+      "loss": 1.0676,
+      "step": 1182
+    },
+    {
+      "epoch": 0.27086433886662853,
+      "grad_norm": 1.1232244968414307,
+      "learning_rate": 1.712723079974121e-05,
+      "loss": 1.1441,
+      "step": 1183
+    },
+    {
+      "epoch": 0.27109330280480826,
+      "grad_norm": 1.3553615808486938,
+      "learning_rate": 1.712202540098884e-05,
+      "loss": 1.1112,
+      "step": 1184
+    },
+    {
+      "epoch": 0.271322266742988,
+      "grad_norm": 1.2029356956481934,
+      "learning_rate": 1.7116816083045603e-05,
+      "loss": 1.1263,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2715512306811677,
+      "grad_norm": 1.465103268623352,
+      "learning_rate": 1.7111602848778143e-05,
+      "loss": 1.1094,
+      "step": 1186
+    },
+    {
+      "epoch": 0.27178019461934744,
+      "grad_norm": 1.3049366474151611,
+      "learning_rate": 1.710638570105526e-05,
+      "loss": 1.1015,
+      "step": 1187
+    },
+    {
+      "epoch": 0.27200915855752716,
+      "grad_norm": 1.2271827459335327,
+      "learning_rate": 1.7101164642747906e-05,
+      "loss": 1.0745,
+      "step": 1188
+    },
+    {
+      "epoch": 0.27223812249570695,
+      "grad_norm": 1.0381929874420166,
+      "learning_rate": 1.7095939676729184e-05,
+      "loss": 1.1369,
+      "step": 1189
+    },
+    {
+      "epoch": 0.2724670864338867,
+      "grad_norm": 1.3117164373397827,
+      "learning_rate": 1.709071080587435e-05,
+      "loss": 1.1253,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2726960503720664,
+      "grad_norm": 1.409812092781067,
+      "learning_rate": 1.7085478033060808e-05,
+      "loss": 1.1859,
+      "step": 1191
+    },
+    {
+      "epoch": 0.27292501431024613,
+      "grad_norm": 1.3818203210830688,
+      "learning_rate": 1.7080241361168108e-05,
+      "loss": 1.0365,
+      "step": 1192
+    },
+    {
+      "epoch": 0.27315397824842585,
+      "grad_norm": 1.1283059120178223,
+      "learning_rate": 1.707500079307795e-05,
+      "loss": 1.0971,
+      "step": 1193
+    },
+    {
+      "epoch": 0.2733829421866056,
+      "grad_norm": 1.1828192472457886,
+      "learning_rate": 1.7069756331674172e-05,
+      "loss": 1.1046,
+      "step": 1194
+    },
+    {
+      "epoch": 0.27361190612478536,
+      "grad_norm": 1.287385106086731,
+      "learning_rate": 1.7064507979842755e-05,
+      "loss": 1.0766,
+      "step": 1195
+    },
+    {
+      "epoch": 0.2738408700629651,
+      "grad_norm": 1.1248568296432495,
+      "learning_rate": 1.705925574047183e-05,
+      "loss": 1.068,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2740698340011448,
+      "grad_norm": 1.151267647743225,
+      "learning_rate": 1.7053999616451653e-05,
+      "loss": 1.0703,
+      "step": 1197
+    },
+    {
+      "epoch": 0.27429879793932455,
+      "grad_norm": 1.1329370737075806,
+      "learning_rate": 1.7048739610674626e-05,
+      "loss": 1.0707,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2745277618775043,
+      "grad_norm": 1.0887163877487183,
+      "learning_rate": 1.704347572603529e-05,
+      "loss": 1.1708,
+      "step": 1199
+    },
+    {
+      "epoch": 0.27475672581568406,
+      "grad_norm": 1.1625685691833496,
+      "learning_rate": 1.7038207965430316e-05,
+      "loss": 1.0947,
+      "step": 1200
+    },
+    {
+      "epoch": 0.2749856897538638,
+      "grad_norm": 1.0462048053741455,
+      "learning_rate": 1.7032936331758505e-05,
+      "loss": 1.0896,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2752146536920435,
+      "grad_norm": 1.286412000656128,
+      "learning_rate": 1.7027660827920798e-05,
+      "loss": 1.0962,
+      "step": 1202
+    },
+    {
+      "epoch": 0.27544361763022324,
+      "grad_norm": 1.4663245677947998,
+      "learning_rate": 1.702238145682025e-05,
+      "loss": 1.1375,
+      "step": 1203
+    },
+    {
+      "epoch": 0.27567258156840296,
+      "grad_norm": 1.2858197689056396,
+      "learning_rate": 1.701709822136207e-05,
+      "loss": 1.1469,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2759015455065827,
+      "grad_norm": 1.4555660486221313,
+      "learning_rate": 1.7011811124453567e-05,
+      "loss": 1.096,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2761305094447625,
+      "grad_norm": 1.211330771446228,
+      "learning_rate": 1.700652016900419e-05,
+      "loss": 1.0777,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2763594733829422,
+      "grad_norm": 1.0660839080810547,
+      "learning_rate": 1.7001225357925506e-05,
+      "loss": 1.0831,
+      "step": 1207
+    },
+    {
+      "epoch": 0.2765884373211219,
+      "grad_norm": 1.1060619354248047,
+      "learning_rate": 1.6995926694131207e-05,
+      "loss": 1.097,
+      "step": 1208
+    },
+    {
+      "epoch": 0.27681740125930165,
+      "grad_norm": 1.301016926765442,
+      "learning_rate": 1.6990624180537102e-05,
+      "loss": 1.1579,
+      "step": 1209
+    },
+    {
+      "epoch": 0.2770463651974814,
+      "grad_norm": 1.3062912225723267,
+      "learning_rate": 1.6985317820061116e-05,
+      "loss": 1.0382,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2772753291356611,
+      "grad_norm": 1.3300676345825195,
+      "learning_rate": 1.69800076156233e-05,
+      "loss": 1.2066,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2775042930738409,
+      "grad_norm": 1.5877501964569092,
+      "learning_rate": 1.6974693570145818e-05,
+      "loss": 1.0696,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2777332570120206,
+      "grad_norm": 1.235116720199585,
+      "learning_rate": 1.696937568655294e-05,
+      "loss": 1.121,
+      "step": 1213
+    },
+    {
+      "epoch": 0.27796222095020034,
+      "grad_norm": 1.2792831659317017,
+      "learning_rate": 1.6964053967771047e-05,
+      "loss": 1.1357,
+      "step": 1214
+    },
+    {
+      "epoch": 0.27819118488838007,
+      "grad_norm": 1.6012238264083862,
+      "learning_rate": 1.6958728416728644e-05,
+      "loss": 1.0489,
+      "step": 1215
+    },
+    {
+      "epoch": 0.2784201488265598,
+      "grad_norm": 1.0891828536987305,
+      "learning_rate": 1.6953399036356338e-05,
+      "loss": 1.0707,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2786491127647395,
+      "grad_norm": 1.0039647817611694,
+      "learning_rate": 1.6948065829586835e-05,
+      "loss": 1.1229,
+      "step": 1217
+    },
+    {
+      "epoch": 0.2788780767029193,
+      "grad_norm": 1.0819059610366821,
+      "learning_rate": 1.6942728799354963e-05,
+      "loss": 1.1374,
+      "step": 1218
+    },
+    {
+      "epoch": 0.27910704064109904,
+      "grad_norm": 1.178362488746643,
+      "learning_rate": 1.693738794859764e-05,
+      "loss": 1.0953,
+      "step": 1219
+    },
+    {
+      "epoch": 0.27933600457927876,
+      "grad_norm": 1.132528305053711,
+      "learning_rate": 1.6932043280253892e-05,
+      "loss": 1.1012,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2795649685174585,
+      "grad_norm": 1.2141227722167969,
+      "learning_rate": 1.6926694797264844e-05,
+      "loss": 1.1225,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2797939324556382,
+      "grad_norm": 1.3470580577850342,
+      "learning_rate": 1.6921342502573723e-05,
+      "loss": 1.1037,
+      "step": 1222
+    },
+    {
+      "epoch": 0.280022896393818,
+      "grad_norm": 1.2042268514633179,
+      "learning_rate": 1.6915986399125852e-05,
+      "loss": 1.0931,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2802518603319977,
+      "grad_norm": 1.567338228225708,
+      "learning_rate": 1.691062648986865e-05,
+      "loss": 1.0394,
+      "step": 1224
+    },
+    {
+      "epoch": 0.28048082427017745,
+      "grad_norm": 0.9561851024627686,
+      "learning_rate": 1.6905262777751627e-05,
+      "loss": 1.01,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2807097882083572,
+      "grad_norm": 1.5181394815444946,
+      "learning_rate": 1.6899895265726392e-05,
+      "loss": 1.1086,
+      "step": 1226
+    },
+    {
+      "epoch": 0.2809387521465369,
+      "grad_norm": 1.1278475522994995,
+      "learning_rate": 1.689452395674664e-05,
+      "loss": 1.1136,
+      "step": 1227
+    },
+    {
+      "epoch": 0.28116771608471663,
+      "grad_norm": 1.375020146369934,
+      "learning_rate": 1.688914885376816e-05,
+      "loss": 1.1157,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2813966800228964,
+      "grad_norm": 1.1723235845565796,
+      "learning_rate": 1.6883769959748822e-05,
+      "loss": 1.1582,
+      "step": 1229
+    },
+    {
+      "epoch": 0.28162564396107614,
+      "grad_norm": 1.5612130165100098,
+      "learning_rate": 1.6878387277648593e-05,
+      "loss": 1.0827,
+      "step": 1230
+    },
+    {
+      "epoch": 0.28185460789925587,
+      "grad_norm": 1.238914966583252,
+      "learning_rate": 1.6873000810429512e-05,
+      "loss": 1.0996,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2820835718374356,
+      "grad_norm": 1.1172726154327393,
+      "learning_rate": 1.6867610561055707e-05,
+      "loss": 1.1171,
+      "step": 1232
+    },
+    {
+      "epoch": 0.2823125357756153,
+      "grad_norm": 1.265048861503601,
+      "learning_rate": 1.686221653249339e-05,
+      "loss": 1.1057,
+      "step": 1233
+    },
+    {
+      "epoch": 0.28254149971379505,
+      "grad_norm": 1.3867113590240479,
+      "learning_rate": 1.6856818727710847e-05,
+      "loss": 1.1072,
+      "step": 1234
+    },
+    {
+      "epoch": 0.28277046365197483,
+      "grad_norm": 1.2063997983932495,
+      "learning_rate": 1.6851417149678442e-05,
+      "loss": 1.1,
+      "step": 1235
+    },
+    {
+      "epoch": 0.28299942759015456,
+      "grad_norm": 1.0993294715881348,
+      "learning_rate": 1.6846011801368626e-05,
+      "loss": 1.1188,
+      "step": 1236
+    },
+    {
+      "epoch": 0.2832283915283343,
+      "grad_norm": 0.9886915683746338,
+      "learning_rate": 1.6840602685755914e-05,
+      "loss": 1.0935,
+      "step": 1237
+    },
+    {
+      "epoch": 0.283457355466514,
+      "grad_norm": 1.2733272314071655,
+      "learning_rate": 1.6835189805816894e-05,
+      "loss": 1.0806,
+      "step": 1238
+    },
+    {
+      "epoch": 0.28368631940469374,
+      "grad_norm": 1.0870503187179565,
+      "learning_rate": 1.6829773164530226e-05,
+      "loss": 1.1355,
+      "step": 1239
+    },
+    {
+      "epoch": 0.28391528334287347,
+      "grad_norm": 1.102055311203003,
+      "learning_rate": 1.6824352764876653e-05,
+      "loss": 1.1119,
+      "step": 1240
+    },
+    {
+      "epoch": 0.28414424728105325,
+      "grad_norm": 1.2194703817367554,
+      "learning_rate": 1.6818928609838967e-05,
+      "loss": 1.0878,
+      "step": 1241
+    },
+    {
+      "epoch": 0.284373211219233,
+      "grad_norm": 1.2381644248962402,
+      "learning_rate": 1.681350070240204e-05,
+      "loss": 1.0619,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2846021751574127,
+      "grad_norm": 1.302049994468689,
+      "learning_rate": 1.6808069045552793e-05,
+      "loss": 1.117,
+      "step": 1243
+    },
+    {
+      "epoch": 0.28483113909559243,
+      "grad_norm": 1.0947741270065308,
+      "learning_rate": 1.6802633642280233e-05,
+      "loss": 1.0865,
+      "step": 1244
+    },
+    {
+      "epoch": 0.28506010303377216,
+      "grad_norm": 1.4091873168945312,
+      "learning_rate": 1.6797194495575412e-05,
+      "loss": 1.1398,
+      "step": 1245
+    },
+    {
+      "epoch": 0.28528906697195194,
+      "grad_norm": 1.2080914974212646,
+      "learning_rate": 1.679175160843145e-05,
+      "loss": 1.0264,
+      "step": 1246
+    },
+    {
+      "epoch": 0.28551803091013167,
+      "grad_norm": 1.3599348068237305,
+      "learning_rate": 1.6786304983843517e-05,
+      "loss": 1.1261,
+      "step": 1247
+    },
+    {
+      "epoch": 0.2857469948483114,
+      "grad_norm": 1.1633814573287964,
+      "learning_rate": 1.678085462480885e-05,
+      "loss": 1.075,
+      "step": 1248
+    },
+    {
+      "epoch": 0.2859759587864911,
+      "grad_norm": 1.1775692701339722,
+      "learning_rate": 1.6775400534326735e-05,
+      "loss": 1.0378,
+      "step": 1249
+    },
+    {
+      "epoch": 0.28620492272467085,
+      "grad_norm": 1.3439725637435913,
+      "learning_rate": 1.676994271539851e-05,
+      "loss": 1.0889,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2864338866628506,
+      "grad_norm": 1.164843201637268,
+      "learning_rate": 1.6764481171027566e-05,
+      "loss": 1.2002,
+      "step": 1251
+    },
+    {
+      "epoch": 0.28666285060103036,
+      "grad_norm": 1.5093419551849365,
+      "learning_rate": 1.6759015904219348e-05,
+      "loss": 1.0292,
+      "step": 1252
+    },
+    {
+      "epoch": 0.2868918145392101,
+      "grad_norm": 1.3763619661331177,
+      "learning_rate": 1.6753546917981346e-05,
+      "loss": 1.0907,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2871207784773898,
+      "grad_norm": 1.3228497505187988,
+      "learning_rate": 1.6748074215323096e-05,
+      "loss": 1.1251,
+      "step": 1254
+    },
+    {
+      "epoch": 0.28734974241556954,
+      "grad_norm": 1.9618299007415771,
+      "learning_rate": 1.6742597799256182e-05,
+      "loss": 1.0923,
+      "step": 1255
+    },
+    {
+      "epoch": 0.28757870635374927,
+      "grad_norm": 1.724094033241272,
+      "learning_rate": 1.673711767279423e-05,
+      "loss": 1.0645,
+      "step": 1256
+    },
+    {
+      "epoch": 0.287807670291929,
+      "grad_norm": 1.1787453889846802,
+      "learning_rate": 1.6731633838952905e-05,
+      "loss": 1.0869,
+      "step": 1257
+    },
+    {
+      "epoch": 0.2880366342301088,
+      "grad_norm": 1.4934698343276978,
+      "learning_rate": 1.672614630074992e-05,
+      "loss": 1.0895,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2882655981682885,
+      "grad_norm": 1.7817822694778442,
+      "learning_rate": 1.6720655061205013e-05,
+      "loss": 1.0477,
+      "step": 1259
+    },
+    {
+      "epoch": 0.28849456210646823,
+      "grad_norm": 1.1007169485092163,
+      "learning_rate": 1.671516012333997e-05,
+      "loss": 1.1114,
+      "step": 1260
+    },
+    {
+      "epoch": 0.28872352604464796,
+      "grad_norm": 1.1942838430404663,
+      "learning_rate": 1.6709661490178617e-05,
+      "loss": 1.1364,
+      "step": 1261
+    },
+    {
+      "epoch": 0.2889524899828277,
+      "grad_norm": 1.078680157661438,
+      "learning_rate": 1.6704159164746797e-05,
+      "loss": 1.1666,
+      "step": 1262
+    },
+    {
+      "epoch": 0.2891814539210074,
+      "grad_norm": 3.27377986907959,
+      "learning_rate": 1.6698653150072396e-05,
+      "loss": 1.1181,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2894104178591872,
+      "grad_norm": 1.0961107015609741,
+      "learning_rate": 1.6693143449185328e-05,
+      "loss": 1.1408,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2896393817973669,
+      "grad_norm": 1.6030396223068237,
+      "learning_rate": 1.6687630065117536e-05,
+      "loss": 1.1411,
+      "step": 1265
+    },
+    {
+      "epoch": 0.28986834573554665,
+      "grad_norm": 1.1883594989776611,
+      "learning_rate": 1.6682113000902988e-05,
+      "loss": 1.1251,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2900973096737264,
+      "grad_norm": 1.2973169088363647,
+      "learning_rate": 1.667659225957768e-05,
+      "loss": 1.0265,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2903262736119061,
+      "grad_norm": 1.2373725175857544,
+      "learning_rate": 1.6671067844179625e-05,
+      "loss": 1.0575,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2905552375500859,
+      "grad_norm": 1.0662227869033813,
+      "learning_rate": 1.6665539757748866e-05,
+      "loss": 1.1003,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2907842014882656,
+      "grad_norm": 1.0769942998886108,
+      "learning_rate": 1.666000800332747e-05,
+      "loss": 1.0941,
+      "step": 1270
+    },
+    {
+      "epoch": 0.29101316542644534,
+      "grad_norm": 1.3059958219528198,
+      "learning_rate": 1.6654472583959497e-05,
+      "loss": 1.0998,
+      "step": 1271
+    },
+    {
+      "epoch": 0.29124212936462507,
+      "grad_norm": 1.2797632217407227,
+      "learning_rate": 1.664893350269106e-05,
+      "loss": 1.1016,
+      "step": 1272
+    },
+    {
+      "epoch": 0.2914710933028048,
+      "grad_norm": 1.2182438373565674,
+      "learning_rate": 1.6643390762570254e-05,
+      "loss": 1.1192,
+      "step": 1273
+    },
+    {
+      "epoch": 0.2917000572409845,
+      "grad_norm": 1.4072870016098022,
+      "learning_rate": 1.6637844366647216e-05,
+      "loss": 1.1242,
+      "step": 1274
+    },
+    {
+      "epoch": 0.2919290211791643,
+      "grad_norm": 1.3435471057891846,
+      "learning_rate": 1.663229431797407e-05,
+      "loss": 1.1807,
+      "step": 1275
+    },
+    {
+      "epoch": 0.29215798511734403,
+      "grad_norm": 1.090842604637146,
+      "learning_rate": 1.6626740619604967e-05,
+      "loss": 1.1318,
+      "step": 1276
+    },
+    {
+      "epoch": 0.29238694905552376,
+      "grad_norm": 1.192291021347046,
+      "learning_rate": 1.6621183274596063e-05,
+      "loss": 1.09,
+      "step": 1277
+    },
+    {
+      "epoch": 0.2926159129937035,
+      "grad_norm": 1.1859331130981445,
+      "learning_rate": 1.661562228600551e-05,
+      "loss": 0.9964,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2928448769318832,
+      "grad_norm": 1.146068811416626,
+      "learning_rate": 1.6610057656893483e-05,
+      "loss": 1.1096,
+      "step": 1279
+    },
+    {
+      "epoch": 0.29307384087006294,
+      "grad_norm": 1.1786830425262451,
+      "learning_rate": 1.660448939032214e-05,
+      "loss": 1.1826,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2933028048082427,
+      "grad_norm": 1.2743366956710815,
+      "learning_rate": 1.6598917489355665e-05,
+      "loss": 1.0824,
+      "step": 1281
+    },
+    {
+      "epoch": 0.29353176874642245,
+      "grad_norm": 1.339068055152893,
+      "learning_rate": 1.6593341957060218e-05,
+      "loss": 1.1928,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2937607326846022,
+      "grad_norm": 1.1370441913604736,
+      "learning_rate": 1.658776279650397e-05,
+      "loss": 1.1,
+      "step": 1283
+    },
+    {
+      "epoch": 0.2939896966227819,
+      "grad_norm": 1.2556294202804565,
+      "learning_rate": 1.6582180010757082e-05,
+      "loss": 1.0666,
+      "step": 1284
+    },
+    {
+      "epoch": 0.29421866056096163,
+      "grad_norm": 1.174279808998108,
+      "learning_rate": 1.6576593602891727e-05,
+      "loss": 1.0921,
+      "step": 1285
+    },
+    {
+      "epoch": 0.29444762449914136,
+      "grad_norm": 1.3296531438827515,
+      "learning_rate": 1.6571003575982048e-05,
+      "loss": 1.1534,
+      "step": 1286
+    },
+    {
+      "epoch": 0.29467658843732114,
+      "grad_norm": 1.1834275722503662,
+      "learning_rate": 1.656540993310419e-05,
+      "loss": 1.1352,
+      "step": 1287
+    },
+    {
+      "epoch": 0.29490555237550087,
+      "grad_norm": 1.2583208084106445,
+      "learning_rate": 1.6559812677336293e-05,
+      "loss": 1.1401,
+      "step": 1288
+    },
+    {
+      "epoch": 0.2951345163136806,
+      "grad_norm": 1.1927133798599243,
+      "learning_rate": 1.655421181175848e-05,
+      "loss": 1.1334,
+      "step": 1289
+    },
+    {
+      "epoch": 0.2953634802518603,
+      "grad_norm": 1.2883397340774536,
+      "learning_rate": 1.6548607339452853e-05,
+      "loss": 1.1677,
+      "step": 1290
+    },
+    {
+      "epoch": 0.29559244419004005,
+      "grad_norm": 1.8883707523345947,
+      "learning_rate": 1.654299926350351e-05,
+      "loss": 1.0453,
+      "step": 1291
+    },
+    {
+      "epoch": 0.29582140812821983,
+      "grad_norm": 1.2389609813690186,
+      "learning_rate": 1.6537387586996532e-05,
+      "loss": 1.1274,
+      "step": 1292
+    },
+    {
+      "epoch": 0.29605037206639956,
+      "grad_norm": 0.9801892638206482,
+      "learning_rate": 1.6531772313019972e-05,
+      "loss": 1.144,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2962793360045793,
+      "grad_norm": 1.00913667678833,
+      "learning_rate": 1.6526153444663873e-05,
+      "loss": 1.087,
+      "step": 1294
+    },
+    {
+      "epoch": 0.296508299942759,
+      "grad_norm": 1.0659477710723877,
+      "learning_rate": 1.6520530985020243e-05,
+      "loss": 1.0223,
+      "step": 1295
+    },
+    {
+      "epoch": 0.29673726388093874,
+      "grad_norm": 1.5433529615402222,
+      "learning_rate": 1.651490493718309e-05,
+      "loss": 1.1148,
+      "step": 1296
+    },
+    {
+      "epoch": 0.29696622781911847,
+      "grad_norm": 1.1053426265716553,
+      "learning_rate": 1.6509275304248366e-05,
+      "loss": 1.128,
+      "step": 1297
+    },
+    {
+      "epoch": 0.29719519175729825,
+      "grad_norm": 1.0809921026229858,
+      "learning_rate": 1.650364208931401e-05,
+      "loss": 1.0149,
+      "step": 1298
+    },
+    {
+      "epoch": 0.297424155695478,
+      "grad_norm": 1.0818578004837036,
+      "learning_rate": 1.6498005295479946e-05,
+      "loss": 1.0806,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2976531196336577,
+      "grad_norm": 2.2473480701446533,
+      "learning_rate": 1.6492364925848045e-05,
+      "loss": 1.1062,
+      "step": 1300
+    },
+    {
+      "epoch": 0.29788208357183743,
+      "grad_norm": 1.107351541519165,
+      "learning_rate": 1.6486720983522156e-05,
+      "loss": 1.1262,
+      "step": 1301
+    },
+    {
+      "epoch": 0.29811104751001716,
+      "grad_norm": 1.3999285697937012,
+      "learning_rate": 1.6481073471608097e-05,
+      "loss": 1.1821,
+      "step": 1302
+    },
+    {
+      "epoch": 0.2983400114481969,
+      "grad_norm": 1.3839341402053833,
+      "learning_rate": 1.647542239321364e-05,
+      "loss": 1.0979,
+      "step": 1303
+    },
+    {
+      "epoch": 0.29856897538637667,
+      "grad_norm": 1.4304099082946777,
+      "learning_rate": 1.6469767751448538e-05,
+      "loss": 1.1045,
+      "step": 1304
+    },
+    {
+      "epoch": 0.2987979393245564,
+      "grad_norm": 1.2668440341949463,
+      "learning_rate": 1.6464109549424477e-05,
+      "loss": 1.1298,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2990269032627361,
+      "grad_norm": 1.2562881708145142,
+      "learning_rate": 1.645844779025513e-05,
+      "loss": 1.1475,
+      "step": 1306
+    },
+    {
+      "epoch": 0.29925586720091585,
+      "grad_norm": 1.1980136632919312,
+      "learning_rate": 1.6452782477056112e-05,
+      "loss": 1.1703,
+      "step": 1307
+    },
+    {
+      "epoch": 0.2994848311390956,
+      "grad_norm": 1.2106819152832031,
+      "learning_rate": 1.6447113612944993e-05,
+      "loss": 1.1087,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2997137950772753,
+      "grad_norm": 1.3096649646759033,
+      "learning_rate": 1.6441441201041312e-05,
+      "loss": 1.0965,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2999427590154551,
+      "grad_norm": 1.1836016178131104,
+      "learning_rate": 1.6435765244466536e-05,
+      "loss": 1.0473,
+      "step": 1310
+    },
+    {
+      "epoch": 0.3001717229536348,
+      "grad_norm": 1.1866120100021362,
+      "learning_rate": 1.6430085746344107e-05,
+      "loss": 1.1249,
+      "step": 1311
+    },
+    {
+      "epoch": 0.30040068689181454,
+      "grad_norm": 1.4427460432052612,
+      "learning_rate": 1.6424402709799404e-05,
+      "loss": 1.1176,
+      "step": 1312
+    },
+    {
+      "epoch": 0.30062965082999427,
+      "grad_norm": 1.251099705696106,
+      "learning_rate": 1.6418716137959746e-05,
+      "loss": 1.1228,
+      "step": 1313
+    },
+    {
+      "epoch": 0.300858614768174,
+      "grad_norm": 2.566626787185669,
+      "learning_rate": 1.6413026033954418e-05,
+      "loss": 1.1398,
+      "step": 1314
+    },
+    {
+      "epoch": 0.3010875787063538,
+      "grad_norm": 1.1216868162155151,
+      "learning_rate": 1.6407332400914625e-05,
+      "loss": 1.0773,
+      "step": 1315
+    },
+    {
+      "epoch": 0.3013165426445335,
+      "grad_norm": 1.1276382207870483,
+      "learning_rate": 1.6401635241973533e-05,
+      "loss": 1.0415,
+      "step": 1316
+    },
+    {
+      "epoch": 0.30154550658271323,
+      "grad_norm": 1.1274714469909668,
+      "learning_rate": 1.639593456026624e-05,
+      "loss": 1.1136,
+      "step": 1317
+    },
+    {
+      "epoch": 0.30177447052089296,
+      "grad_norm": 1.0719152688980103,
+      "learning_rate": 1.639023035892978e-05,
+      "loss": 1.1102,
+      "step": 1318
+    },
+    {
+      "epoch": 0.3020034344590727,
+      "grad_norm": 1.0603184700012207,
+      "learning_rate": 1.6384522641103133e-05,
+      "loss": 1.1015,
+      "step": 1319
+    },
+    {
+      "epoch": 0.3022323983972524,
+      "grad_norm": 1.0919266939163208,
+      "learning_rate": 1.63788114099272e-05,
+      "loss": 1.0915,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3024613623354322,
+      "grad_norm": 1.3398226499557495,
+      "learning_rate": 1.637309666854483e-05,
+      "loss": 1.1068,
+      "step": 1321
+    },
+    {
+      "epoch": 0.3026903262736119,
+      "grad_norm": 1.094092845916748,
+      "learning_rate": 1.6367378420100798e-05,
+      "loss": 1.1188,
+      "step": 1322
+    },
+    {
+      "epoch": 0.30291929021179165,
+      "grad_norm": 1.0754516124725342,
+      "learning_rate": 1.63616566677418e-05,
+      "loss": 1.1741,
+      "step": 1323
+    },
+    {
+      "epoch": 0.3031482541499714,
+      "grad_norm": 1.155779242515564,
+      "learning_rate": 1.6355931414616477e-05,
+      "loss": 1.1085,
+      "step": 1324
+    },
+    {
+      "epoch": 0.3033772180881511,
+      "grad_norm": 1.186424970626831,
+      "learning_rate": 1.6350202663875385e-05,
+      "loss": 1.0362,
+      "step": 1325
+    },
+    {
+      "epoch": 0.30360618202633083,
+      "grad_norm": 1.1375372409820557,
+      "learning_rate": 1.634447041867101e-05,
+      "loss": 1.0349,
+      "step": 1326
+    },
+    {
+      "epoch": 0.3038351459645106,
+      "grad_norm": 1.346127986907959,
+      "learning_rate": 1.633873468215775e-05,
+      "loss": 1.0951,
+      "step": 1327
+    },
+    {
+      "epoch": 0.30406410990269034,
+      "grad_norm": 1.2048730850219727,
+      "learning_rate": 1.633299545749194e-05,
+      "loss": 1.0937,
+      "step": 1328
+    },
+    {
+      "epoch": 0.30429307384087007,
+      "grad_norm": 1.2353206872940063,
+      "learning_rate": 1.6327252747831824e-05,
+      "loss": 1.1092,
+      "step": 1329
+    },
+    {
+      "epoch": 0.3045220377790498,
+      "grad_norm": 1.1722041368484497,
+      "learning_rate": 1.6321506556337575e-05,
+      "loss": 1.0895,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3047510017172295,
+      "grad_norm": 1.1982693672180176,
+      "learning_rate": 1.6315756886171264e-05,
+      "loss": 1.1342,
+      "step": 1331
+    },
+    {
+      "epoch": 0.30497996565540925,
+      "grad_norm": 1.1630750894546509,
+      "learning_rate": 1.6310003740496887e-05,
+      "loss": 1.1008,
+      "step": 1332
+    },
+    {
+      "epoch": 0.30520892959358903,
+      "grad_norm": 1.6868027448654175,
+      "learning_rate": 1.6304247122480355e-05,
+      "loss": 1.1756,
+      "step": 1333
+    },
+    {
+      "epoch": 0.30543789353176876,
+      "grad_norm": 1.2052215337753296,
+      "learning_rate": 1.629848703528949e-05,
+      "loss": 1.0759,
+      "step": 1334
+    },
+    {
+      "epoch": 0.3056668574699485,
+      "grad_norm": 1.4629309177398682,
+      "learning_rate": 1.6292723482094013e-05,
+      "loss": 1.1375,
+      "step": 1335
+    },
+    {
+      "epoch": 0.3058958214081282,
+      "grad_norm": 0.995369553565979,
+      "learning_rate": 1.6286956466065566e-05,
+      "loss": 1.1433,
+      "step": 1336
+    },
+    {
+      "epoch": 0.30612478534630794,
+      "grad_norm": 1.3381826877593994,
+      "learning_rate": 1.6281185990377683e-05,
+      "loss": 1.0782,
+      "step": 1337
+    },
+    {
+      "epoch": 0.3063537492844877,
+      "grad_norm": 1.22222101688385,
+      "learning_rate": 1.627541205820581e-05,
+      "loss": 1.0767,
+      "step": 1338
+    },
+    {
+      "epoch": 0.30658271322266745,
+      "grad_norm": 1.1784082651138306,
+      "learning_rate": 1.6269634672727296e-05,
+      "loss": 1.0653,
+      "step": 1339
+    },
+    {
+      "epoch": 0.3068116771608472,
+      "grad_norm": 1.214966058731079,
+      "learning_rate": 1.6263853837121384e-05,
+      "loss": 1.0808,
+      "step": 1340
+    },
+    {
+      "epoch": 0.3070406410990269,
+      "grad_norm": 1.5941448211669922,
+      "learning_rate": 1.6258069554569226e-05,
+      "loss": 1.1365,
+      "step": 1341
+    },
+    {
+      "epoch": 0.30726960503720663,
+      "grad_norm": 1.1932793855667114,
+      "learning_rate": 1.6252281828253856e-05,
+      "loss": 1.1333,
+      "step": 1342
+    },
+    {
+      "epoch": 0.30749856897538635,
+      "grad_norm": 1.285779356956482,
+      "learning_rate": 1.6246490661360215e-05,
+      "loss": 1.0826,
+      "step": 1343
+    },
+    {
+      "epoch": 0.30772753291356614,
+      "grad_norm": 1.9943314790725708,
+      "learning_rate": 1.6240696057075138e-05,
+      "loss": 1.0179,
+      "step": 1344
+    },
+    {
+      "epoch": 0.30795649685174586,
+      "grad_norm": 1.3297176361083984,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 1.0836,
+      "step": 1345
+    },
+    {
+      "epoch": 0.3081854607899256,
+      "grad_norm": 1.1271708011627197,
+      "learning_rate": 1.6229096549087434e-05,
+      "loss": 1.0867,
+      "step": 1346
+    },
+    {
+      "epoch": 0.3084144247281053,
+      "grad_norm": 1.003398060798645,
+      "learning_rate": 1.6223291651767922e-05,
+      "loss": 1.119,
+      "step": 1347
+    },
+    {
+      "epoch": 0.30864338866628505,
+      "grad_norm": 1.4007704257965088,
+      "learning_rate": 1.621748332982319e-05,
+      "loss": 1.054,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3088723526044648,
+      "grad_norm": 1.2629410028457642,
+      "learning_rate": 1.6211671586449512e-05,
+      "loss": 1.0702,
+      "step": 1349
+    },
+    {
+      "epoch": 0.30910131654264456,
+      "grad_norm": 1.1390570402145386,
+      "learning_rate": 1.6205856424845038e-05,
+      "loss": 1.0674,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3093302804808243,
+      "grad_norm": 1.0701704025268555,
+      "learning_rate": 1.62000378482098e-05,
+      "loss": 1.1055,
+      "step": 1351
+    },
+    {
+      "epoch": 0.309559244419004,
+      "grad_norm": 1.0888673067092896,
+      "learning_rate": 1.619421585974572e-05,
+      "loss": 1.0241,
+      "step": 1352
+    },
+    {
+      "epoch": 0.30978820835718374,
+      "grad_norm": 1.1905760765075684,
+      "learning_rate": 1.618839046265658e-05,
+      "loss": 1.1075,
+      "step": 1353
+    },
+    {
+      "epoch": 0.31001717229536346,
+      "grad_norm": 1.1428126096725464,
+      "learning_rate": 1.6182561660148053e-05,
+      "loss": 1.1007,
+      "step": 1354
+    },
+    {
+      "epoch": 0.3102461362335432,
+      "grad_norm": 0.9909505844116211,
+      "learning_rate": 1.617672945542768e-05,
+      "loss": 1.0782,
+      "step": 1355
+    },
+    {
+      "epoch": 0.310475100171723,
+      "grad_norm": 1.0805736780166626,
+      "learning_rate": 1.6170893851704875e-05,
+      "loss": 1.1272,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3107040641099027,
+      "grad_norm": 1.3021117448806763,
+      "learning_rate": 1.6165054852190917e-05,
+      "loss": 1.0889,
+      "step": 1357
+    },
+    {
+      "epoch": 0.3109330280480824,
+      "grad_norm": 1.0056633949279785,
+      "learning_rate": 1.6159212460098968e-05,
+      "loss": 1.0785,
+      "step": 1358
+    },
+    {
+      "epoch": 0.31116199198626215,
+      "grad_norm": 1.1618127822875977,
+      "learning_rate": 1.6153366678644035e-05,
+      "loss": 1.0712,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3113909559244419,
+      "grad_norm": 1.7210270166397095,
+      "learning_rate": 1.614751751104301e-05,
+      "loss": 1.0849,
+      "step": 1360
+    },
+    {
+      "epoch": 0.31161991986262166,
+      "grad_norm": 1.262199878692627,
+      "learning_rate": 1.614166496051464e-05,
+      "loss": 1.1274,
+      "step": 1361
+    },
+    {
+      "epoch": 0.3118488838008014,
+      "grad_norm": 1.2134400606155396,
+      "learning_rate": 1.6135809030279534e-05,
+      "loss": 1.1211,
+      "step": 1362
+    },
+    {
+      "epoch": 0.3120778477389811,
+      "grad_norm": 1.4023921489715576,
+      "learning_rate": 1.6129949723560162e-05,
+      "loss": 1.1045,
+      "step": 1363
+    },
+    {
+      "epoch": 0.31230681167716084,
+      "grad_norm": 1.6166713237762451,
+      "learning_rate": 1.612408704358085e-05,
+      "loss": 1.0561,
+      "step": 1364
+    },
+    {
+      "epoch": 0.31253577561534057,
+      "grad_norm": 1.3405228853225708,
+      "learning_rate": 1.6118220993567783e-05,
+      "loss": 1.1804,
+      "step": 1365
+    },
+    {
+      "epoch": 0.3127647395535203,
+      "grad_norm": 1.1368193626403809,
+      "learning_rate": 1.6112351576748994e-05,
+      "loss": 1.1047,
+      "step": 1366
+    },
+    {
+      "epoch": 0.3129937034917001,
+      "grad_norm": 1.1940534114837646,
+      "learning_rate": 1.6106478796354382e-05,
+      "loss": 1.0889,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3132226674298798,
+      "grad_norm": 1.7876060009002686,
+      "learning_rate": 1.6100602655615683e-05,
+      "loss": 1.1257,
+      "step": 1368
+    },
+    {
+      "epoch": 0.31345163136805954,
+      "grad_norm": 1.4505313634872437,
+      "learning_rate": 1.6094723157766493e-05,
+      "loss": 1.1566,
+      "step": 1369
+    },
+    {
+      "epoch": 0.31368059530623926,
+      "grad_norm": 1.2009062767028809,
+      "learning_rate": 1.6088840306042247e-05,
+      "loss": 1.1019,
+      "step": 1370
+    },
+    {
+      "epoch": 0.313909559244419,
+      "grad_norm": 1.2727826833724976,
+      "learning_rate": 1.608295410368023e-05,
+      "loss": 1.1033,
+      "step": 1371
+    },
+    {
+      "epoch": 0.3141385231825987,
+      "grad_norm": 1.1413002014160156,
+      "learning_rate": 1.6077064553919565e-05,
+      "loss": 1.0678,
+      "step": 1372
+    },
+    {
+      "epoch": 0.3143674871207785,
+      "grad_norm": 1.2150791883468628,
+      "learning_rate": 1.6071171660001232e-05,
+      "loss": 1.093,
+      "step": 1373
+    },
+    {
+      "epoch": 0.3145964510589582,
+      "grad_norm": 1.4832680225372314,
+      "learning_rate": 1.6065275425168034e-05,
+      "loss": 1.1531,
+      "step": 1374
+    },
+    {
+      "epoch": 0.31482541499713795,
+      "grad_norm": 1.3526698350906372,
+      "learning_rate": 1.605937585266462e-05,
+      "loss": 1.104,
+      "step": 1375
+    },
+    {
+      "epoch": 0.3150543789353177,
+      "grad_norm": 1.1165345907211304,
+      "learning_rate": 1.6053472945737474e-05,
+      "loss": 1.0644,
+      "step": 1376
+    },
+    {
+      "epoch": 0.3152833428734974,
+      "grad_norm": 1.1160399913787842,
+      "learning_rate": 1.6047566707634918e-05,
+      "loss": 1.1056,
+      "step": 1377
+    },
+    {
+      "epoch": 0.31551230681167713,
+      "grad_norm": 1.4708969593048096,
+      "learning_rate": 1.604165714160711e-05,
+      "loss": 1.049,
+      "step": 1378
+    },
+    {
+      "epoch": 0.3157412707498569,
+      "grad_norm": 1.1713590621948242,
+      "learning_rate": 1.6035744250906026e-05,
+      "loss": 1.1238,
+      "step": 1379
+    },
+    {
+      "epoch": 0.31597023468803664,
+      "grad_norm": 1.135960340499878,
+      "learning_rate": 1.6029828038785486e-05,
+      "loss": 1.1036,
+      "step": 1380
+    },
+    {
+      "epoch": 0.31619919862621637,
+      "grad_norm": 1.1496847867965698,
+      "learning_rate": 1.602390850850113e-05,
+      "loss": 1.1226,
+      "step": 1381
+    },
+    {
+      "epoch": 0.3164281625643961,
+      "grad_norm": 1.4546343088150024,
+      "learning_rate": 1.6017985663310427e-05,
+      "loss": 1.0588,
+      "step": 1382
+    },
+    {
+      "epoch": 0.3166571265025758,
+      "grad_norm": 1.1785930395126343,
+      "learning_rate": 1.6012059506472665e-05,
+      "loss": 1.0823,
+      "step": 1383
+    },
+    {
+      "epoch": 0.3168860904407556,
+      "grad_norm": 1.2727309465408325,
+      "learning_rate": 1.6006130041248968e-05,
+      "loss": 1.1149,
+      "step": 1384
+    },
+    {
+      "epoch": 0.31711505437893533,
+      "grad_norm": 1.122978925704956,
+      "learning_rate": 1.600019727090226e-05,
+      "loss": 1.0961,
+      "step": 1385
+    },
+    {
+      "epoch": 0.31734401831711506,
+      "grad_norm": 1.1095397472381592,
+      "learning_rate": 1.59942611986973e-05,
+      "loss": 1.1687,
+      "step": 1386
+    },
+    {
+      "epoch": 0.3175729822552948,
+      "grad_norm": 1.2510476112365723,
+      "learning_rate": 1.598832182790066e-05,
+      "loss": 1.0486,
+      "step": 1387
+    },
+    {
+      "epoch": 0.3178019461934745,
+      "grad_norm": 1.103319764137268,
+      "learning_rate": 1.5982379161780722e-05,
+      "loss": 1.1006,
+      "step": 1388
+    },
+    {
+      "epoch": 0.31803091013165424,
+      "grad_norm": 1.4000933170318604,
+      "learning_rate": 1.597643320360769e-05,
+      "loss": 1.1743,
+      "step": 1389
+    },
+    {
+      "epoch": 0.318259874069834,
+      "grad_norm": 1.0757001638412476,
+      "learning_rate": 1.5970483956653572e-05,
+      "loss": 1.0575,
+      "step": 1390
+    },
+    {
+      "epoch": 0.31848883800801375,
+      "grad_norm": 1.2366151809692383,
+      "learning_rate": 1.5964531424192187e-05,
+      "loss": 1.0801,
+      "step": 1391
+    },
+    {
+      "epoch": 0.3187178019461935,
+      "grad_norm": 1.175316572189331,
+      "learning_rate": 1.595857560949917e-05,
+      "loss": 1.0543,
+      "step": 1392
+    },
+    {
+      "epoch": 0.3189467658843732,
+      "grad_norm": 1.2209275960922241,
+      "learning_rate": 1.595261651585195e-05,
+      "loss": 1.1517,
+      "step": 1393
+    },
+    {
+      "epoch": 0.31917572982255293,
+      "grad_norm": 1.448627233505249,
+      "learning_rate": 1.5946654146529766e-05,
+      "loss": 1.137,
+      "step": 1394
+    },
+    {
+      "epoch": 0.31940469376073266,
+      "grad_norm": 1.2850122451782227,
+      "learning_rate": 1.5940688504813664e-05,
+      "loss": 1.113,
+      "step": 1395
+    },
+    {
+      "epoch": 0.31963365769891244,
+      "grad_norm": 1.3141731023788452,
+      "learning_rate": 1.5934719593986483e-05,
+      "loss": 1.0964,
+      "step": 1396
+    },
+    {
+      "epoch": 0.31986262163709217,
+      "grad_norm": 1.1482245922088623,
+      "learning_rate": 1.592874741733287e-05,
+      "loss": 1.0624,
+      "step": 1397
+    },
+    {
+      "epoch": 0.3200915855752719,
+      "grad_norm": 1.1503582000732422,
+      "learning_rate": 1.5922771978139255e-05,
+      "loss": 1.0634,
+      "step": 1398
+    },
+    {
+      "epoch": 0.3203205495134516,
+      "grad_norm": 1.4972145557403564,
+      "learning_rate": 1.5916793279693878e-05,
+      "loss": 1.1168,
+      "step": 1399
+    },
+    {
+      "epoch": 0.32054951345163135,
+      "grad_norm": 1.0118694305419922,
+      "learning_rate": 1.5910811325286768e-05,
+      "loss": 1.0714,
+      "step": 1400
+    },
+    {
+      "epoch": 0.3207784773898111,
+      "grad_norm": 1.2603758573532104,
+      "learning_rate": 1.590482611820974e-05,
+      "loss": 1.1209,
+      "step": 1401
+    },
+    {
+      "epoch": 0.32100744132799086,
+      "grad_norm": 0.941050112247467,
+      "learning_rate": 1.5898837661756405e-05,
+      "loss": 1.0597,
+      "step": 1402
+    },
+    {
+      "epoch": 0.3212364052661706,
+      "grad_norm": 1.2156561613082886,
+      "learning_rate": 1.5892845959222164e-05,
+      "loss": 1.101,
+      "step": 1403
+    },
+    {
+      "epoch": 0.3214653692043503,
+      "grad_norm": 1.0653733015060425,
+      "learning_rate": 1.5886851013904193e-05,
+      "loss": 1.0916,
+      "step": 1404
+    },
+    {
+      "epoch": 0.32169433314253004,
+      "grad_norm": 1.1176843643188477,
+      "learning_rate": 1.5880852829101464e-05,
+      "loss": 1.1002,
+      "step": 1405
+    },
+    {
+      "epoch": 0.32192329708070977,
+      "grad_norm": 1.3525538444519043,
+      "learning_rate": 1.5874851408114733e-05,
+      "loss": 1.0686,
+      "step": 1406
+    },
+    {
+      "epoch": 0.32215226101888955,
+      "grad_norm": 1.2710604667663574,
+      "learning_rate": 1.5868846754246524e-05,
+      "loss": 1.0965,
+      "step": 1407
+    },
+    {
+      "epoch": 0.3223812249570693,
+      "grad_norm": 1.1493090391159058,
+      "learning_rate": 1.5862838870801153e-05,
+      "loss": 1.095,
+      "step": 1408
+    },
+    {
+      "epoch": 0.322610188895249,
+      "grad_norm": 1.2418208122253418,
+      "learning_rate": 1.5856827761084698e-05,
+      "loss": 1.0945,
+      "step": 1409
+    },
+    {
+      "epoch": 0.32283915283342873,
+      "grad_norm": 1.0906857252120972,
+      "learning_rate": 1.5850813428405036e-05,
+      "loss": 1.1288,
+      "step": 1410
+    },
+    {
+      "epoch": 0.32306811677160846,
+      "grad_norm": 1.1310313940048218,
+      "learning_rate": 1.58447958760718e-05,
+      "loss": 1.0776,
+      "step": 1411
+    },
+    {
+      "epoch": 0.3232970807097882,
+      "grad_norm": 1.407239317893982,
+      "learning_rate": 1.583877510739639e-05,
+      "loss": 1.0595,
+      "step": 1412
+    },
+    {
+      "epoch": 0.32352604464796797,
+      "grad_norm": 1.0133352279663086,
+      "learning_rate": 1.5832751125691993e-05,
+      "loss": 1.0473,
+      "step": 1413
+    },
+    {
+      "epoch": 0.3237550085861477,
+      "grad_norm": 1.329931616783142,
+      "learning_rate": 1.5826723934273555e-05,
+      "loss": 1.0844,
+      "step": 1414
+    },
+    {
+      "epoch": 0.3239839725243274,
+      "grad_norm": 1.4198311567306519,
+      "learning_rate": 1.5820693536457787e-05,
+      "loss": 1.0686,
+      "step": 1415
+    },
+    {
+      "epoch": 0.32421293646250715,
+      "grad_norm": 1.604586124420166,
+      "learning_rate": 1.5814659935563165e-05,
+      "loss": 1.1034,
+      "step": 1416
+    },
+    {
+      "epoch": 0.3244419004006869,
+      "grad_norm": 1.439122200012207,
+      "learning_rate": 1.5808623134909932e-05,
+      "loss": 1.1126,
+      "step": 1417
+    },
+    {
+      "epoch": 0.3246708643388666,
+      "grad_norm": 1.158563256263733,
+      "learning_rate": 1.5802583137820087e-05,
+      "loss": 1.0924,
+      "step": 1418
+    },
+    {
+      "epoch": 0.3248998282770464,
+      "grad_norm": 1.1076048612594604,
+      "learning_rate": 1.579653994761739e-05,
+      "loss": 1.079,
+      "step": 1419
+    },
+    {
+      "epoch": 0.3251287922152261,
+      "grad_norm": 1.1624640226364136,
+      "learning_rate": 1.5790493567627357e-05,
+      "loss": 1.0644,
+      "step": 1420
+    },
+    {
+      "epoch": 0.32535775615340584,
+      "grad_norm": 1.0896527767181396,
+      "learning_rate": 1.5784444001177262e-05,
+      "loss": 1.1104,
+      "step": 1421
+    },
+    {
+      "epoch": 0.32558672009158557,
+      "grad_norm": 1.096172571182251,
+      "learning_rate": 1.577839125159613e-05,
+      "loss": 1.064,
+      "step": 1422
+    },
+    {
+      "epoch": 0.3258156840297653,
+      "grad_norm": 1.2764605283737183,
+      "learning_rate": 1.577233532221474e-05,
+      "loss": 1.0389,
+      "step": 1423
+    },
+    {
+      "epoch": 0.326044647967945,
+      "grad_norm": 1.2144206762313843,
+      "learning_rate": 1.576627621636561e-05,
+      "loss": 1.0894,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3262736119061248,
+      "grad_norm": 1.1059249639511108,
+      "learning_rate": 1.5760213937383032e-05,
+      "loss": 1.0988,
+      "step": 1425
+    },
+    {
+      "epoch": 0.32650257584430453,
+      "grad_norm": 1.1343713998794556,
+      "learning_rate": 1.5754148488603017e-05,
+      "loss": 1.1237,
+      "step": 1426
+    },
+    {
+      "epoch": 0.32673153978248426,
+      "grad_norm": 1.1411490440368652,
+      "learning_rate": 1.5748079873363327e-05,
+      "loss": 1.14,
+      "step": 1427
+    },
+    {
+      "epoch": 0.326960503720664,
+      "grad_norm": 1.1777209043502808,
+      "learning_rate": 1.5742008095003478e-05,
+      "loss": 1.1442,
+      "step": 1428
+    },
+    {
+      "epoch": 0.3271894676588437,
+      "grad_norm": 1.2743630409240723,
+      "learning_rate": 1.573593315686471e-05,
+      "loss": 1.127,
+      "step": 1429
+    },
+    {
+      "epoch": 0.3274184315970235,
+      "grad_norm": 1.4254614114761353,
+      "learning_rate": 1.5729855062290024e-05,
+      "loss": 1.0446,
+      "step": 1430
+    },
+    {
+      "epoch": 0.3276473955352032,
+      "grad_norm": 1.1998964548110962,
+      "learning_rate": 1.572377381462413e-05,
+      "loss": 1.0988,
+      "step": 1431
+    },
+    {
+      "epoch": 0.32787635947338295,
+      "grad_norm": 1.0931754112243652,
+      "learning_rate": 1.5717689417213495e-05,
+      "loss": 1.0696,
+      "step": 1432
+    },
+    {
+      "epoch": 0.3281053234115627,
+      "grad_norm": 1.1342476606369019,
+      "learning_rate": 1.5711601873406314e-05,
+      "loss": 1.0996,
+      "step": 1433
+    },
+    {
+      "epoch": 0.3283342873497424,
+      "grad_norm": 1.8414933681488037,
+      "learning_rate": 1.5705511186552506e-05,
+      "loss": 1.1078,
+      "step": 1434
+    },
+    {
+      "epoch": 0.32856325128792213,
+      "grad_norm": 1.0369212627410889,
+      "learning_rate": 1.5699417360003725e-05,
+      "loss": 1.0947,
+      "step": 1435
+    },
+    {
+      "epoch": 0.3287922152261019,
+      "grad_norm": 1.1955866813659668,
+      "learning_rate": 1.5693320397113358e-05,
+      "loss": 1.1377,
+      "step": 1436
+    },
+    {
+      "epoch": 0.32902117916428164,
+      "grad_norm": 1.1285663843154907,
+      "learning_rate": 1.568722030123651e-05,
+      "loss": 1.1109,
+      "step": 1437
+    },
+    {
+      "epoch": 0.32925014310246137,
+      "grad_norm": 1.313095211982727,
+      "learning_rate": 1.568111707573001e-05,
+      "loss": 1.1211,
+      "step": 1438
+    },
+    {
+      "epoch": 0.3294791070406411,
+      "grad_norm": 1.4031476974487305,
+      "learning_rate": 1.567501072395241e-05,
+      "loss": 1.0377,
+      "step": 1439
+    },
+    {
+      "epoch": 0.3297080709788208,
+      "grad_norm": 1.2493641376495361,
+      "learning_rate": 1.5668901249263996e-05,
+      "loss": 1.0921,
+      "step": 1440
+    },
+    {
+      "epoch": 0.32993703491700055,
+      "grad_norm": 1.0535095930099487,
+      "learning_rate": 1.5662788655026745e-05,
+      "loss": 1.0488,
+      "step": 1441
+    },
+    {
+      "epoch": 0.33016599885518033,
+      "grad_norm": 1.165377140045166,
+      "learning_rate": 1.565667294460438e-05,
+      "loss": 1.0395,
+      "step": 1442
+    },
+    {
+      "epoch": 0.33039496279336006,
+      "grad_norm": 1.4327694177627563,
+      "learning_rate": 1.5650554121362315e-05,
+      "loss": 1.1258,
+      "step": 1443
+    },
+    {
+      "epoch": 0.3306239267315398,
+      "grad_norm": 1.1685681343078613,
+      "learning_rate": 1.5644432188667695e-05,
+      "loss": 1.0601,
+      "step": 1444
+    },
+    {
+      "epoch": 0.3308528906697195,
+      "grad_norm": 1.4015522003173828,
+      "learning_rate": 1.563830714988936e-05,
+      "loss": 1.0852,
+      "step": 1445
+    },
+    {
+      "epoch": 0.33108185460789924,
+      "grad_norm": 1.0831905603408813,
+      "learning_rate": 1.5632179008397876e-05,
+      "loss": 1.125,
+      "step": 1446
+    },
+    {
+      "epoch": 0.33131081854607897,
+      "grad_norm": 1.9634826183319092,
+      "learning_rate": 1.5626047767565503e-05,
+      "loss": 1.0603,
+      "step": 1447
+    },
+    {
+      "epoch": 0.33153978248425875,
+      "grad_norm": 1.1687532663345337,
+      "learning_rate": 1.561991343076621e-05,
+      "loss": 1.1012,
+      "step": 1448
+    },
+    {
+      "epoch": 0.3317687464224385,
+      "grad_norm": 1.5196926593780518,
+      "learning_rate": 1.5613776001375674e-05,
+      "loss": 1.1178,
+      "step": 1449
+    },
+    {
+      "epoch": 0.3319977103606182,
+      "grad_norm": 1.2472096681594849,
+      "learning_rate": 1.5607635482771272e-05,
+      "loss": 1.0965,
+      "step": 1450
+    },
+    {
+      "epoch": 0.33222667429879793,
+      "grad_norm": 1.207923173904419,
+      "learning_rate": 1.5601491878332077e-05,
+      "loss": 1.0978,
+      "step": 1451
+    },
+    {
+      "epoch": 0.33245563823697766,
+      "grad_norm": 1.1956391334533691,
+      "learning_rate": 1.5595345191438864e-05,
+      "loss": 1.1358,
+      "step": 1452
+    },
+    {
+      "epoch": 0.33268460217515744,
+      "grad_norm": 1.1643397808074951,
+      "learning_rate": 1.5589195425474105e-05,
+      "loss": 1.0615,
+      "step": 1453
+    },
+    {
+      "epoch": 0.33291356611333717,
+      "grad_norm": 1.332839846611023,
+      "learning_rate": 1.5583042583821963e-05,
+      "loss": 1.0744,
+      "step": 1454
+    },
+    {
+      "epoch": 0.3331425300515169,
+      "grad_norm": 1.5693964958190918,
+      "learning_rate": 1.5576886669868297e-05,
+      "loss": 1.095,
+      "step": 1455
+    },
+    {
+      "epoch": 0.3333714939896966,
+      "grad_norm": 1.3151133060455322,
+      "learning_rate": 1.557072768700065e-05,
+      "loss": 1.0147,
+      "step": 1456
+    },
+    {
+      "epoch": 0.33360045792787635,
+      "grad_norm": 2.105339765548706,
+      "learning_rate": 1.5564565638608264e-05,
+      "loss": 1.154,
+      "step": 1457
+    },
+    {
+      "epoch": 0.3338294218660561,
+      "grad_norm": 1.1660069227218628,
+      "learning_rate": 1.5558400528082057e-05,
+      "loss": 1.0835,
+      "step": 1458
+    },
+    {
+      "epoch": 0.33405838580423586,
+      "grad_norm": 1.198183298110962,
+      "learning_rate": 1.5552232358814646e-05,
+      "loss": 1.0679,
+      "step": 1459
+    },
+    {
+      "epoch": 0.3342873497424156,
+      "grad_norm": 0.9822539687156677,
+      "learning_rate": 1.5546061134200316e-05,
+      "loss": 1.0952,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3345163136805953,
+      "grad_norm": 1.082090973854065,
+      "learning_rate": 1.5539886857635037e-05,
+      "loss": 1.0979,
+      "step": 1461
+    },
+    {
+      "epoch": 0.33474527761877504,
+      "grad_norm": 1.2123302221298218,
+      "learning_rate": 1.5533709532516473e-05,
+      "loss": 1.0944,
+      "step": 1462
+    },
+    {
+      "epoch": 0.33497424155695477,
+      "grad_norm": 1.1472247838974,
+      "learning_rate": 1.552752916224395e-05,
+      "loss": 1.1239,
+      "step": 1463
+    },
+    {
+      "epoch": 0.3352032054951345,
+      "grad_norm": 1.1450691223144531,
+      "learning_rate": 1.5521345750218463e-05,
+      "loss": 1.069,
+      "step": 1464
+    },
+    {
+      "epoch": 0.3354321694333143,
+      "grad_norm": 1.3406015634536743,
+      "learning_rate": 1.551515929984271e-05,
+      "loss": 1.0496,
+      "step": 1465
+    },
+    {
+      "epoch": 0.335661133371494,
+      "grad_norm": 1.2106640338897705,
+      "learning_rate": 1.5508969814521026e-05,
+      "loss": 1.1102,
+      "step": 1466
+    },
+    {
+      "epoch": 0.33589009730967373,
+      "grad_norm": 1.483971118927002,
+      "learning_rate": 1.5502777297659447e-05,
+      "loss": 1.0899,
+      "step": 1467
+    },
+    {
+      "epoch": 0.33611906124785346,
+      "grad_norm": 1.171859860420227,
+      "learning_rate": 1.549658175266565e-05,
+      "loss": 1.0654,
+      "step": 1468
+    },
+    {
+      "epoch": 0.3363480251860332,
+      "grad_norm": 2.5037529468536377,
+      "learning_rate": 1.5490383182949e-05,
+      "loss": 1.0884,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3365769891242129,
+      "grad_norm": 1.0418339967727661,
+      "learning_rate": 1.5484181591920516e-05,
+      "loss": 1.0827,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3368059530623927,
+      "grad_norm": 1.2835181951522827,
+      "learning_rate": 1.5477976982992883e-05,
+      "loss": 1.0139,
+      "step": 1471
+    },
+    {
+      "epoch": 0.3370349170005724,
+      "grad_norm": 1.2543227672576904,
+      "learning_rate": 1.547176935958044e-05,
+      "loss": 1.1055,
+      "step": 1472
+    },
+    {
+      "epoch": 0.33726388093875215,
+      "grad_norm": 1.0822360515594482,
+      "learning_rate": 1.54655587250992e-05,
+      "loss": 1.0823,
+      "step": 1473
+    },
+    {
+      "epoch": 0.3374928448769319,
+      "grad_norm": 1.016126036643982,
+      "learning_rate": 1.5459345082966812e-05,
+      "loss": 1.0929,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3377218088151116,
+      "grad_norm": 1.147747278213501,
+      "learning_rate": 1.5453128436602597e-05,
+      "loss": 1.1562,
+      "step": 1475
+    },
+    {
+      "epoch": 0.3379507727532914,
+      "grad_norm": 1.3952897787094116,
+      "learning_rate": 1.5446908789427522e-05,
+      "loss": 1.0573,
+      "step": 1476
+    },
+    {
+      "epoch": 0.3381797366914711,
+      "grad_norm": 1.0825663805007935,
+      "learning_rate": 1.5440686144864207e-05,
+      "loss": 1.1492,
+      "step": 1477
+    },
+    {
+      "epoch": 0.33840870062965084,
+      "grad_norm": 1.1135352849960327,
+      "learning_rate": 1.5434460506336922e-05,
+      "loss": 1.0801,
+      "step": 1478
+    },
+    {
+      "epoch": 0.33863766456783057,
+      "grad_norm": 1.1229122877120972,
+      "learning_rate": 1.5428231877271584e-05,
+      "loss": 1.1142,
+      "step": 1479
+    },
+    {
+      "epoch": 0.3388666285060103,
+      "grad_norm": 1.1263865232467651,
+      "learning_rate": 1.542200026109575e-05,
+      "loss": 1.0628,
+      "step": 1480
+    },
+    {
+      "epoch": 0.33909559244419,
+      "grad_norm": 1.3717758655548096,
+      "learning_rate": 1.5415765661238635e-05,
+      "loss": 1.106,
+      "step": 1481
+    },
+    {
+      "epoch": 0.3393245563823698,
+      "grad_norm": 1.1073044538497925,
+      "learning_rate": 1.540952808113108e-05,
+      "loss": 1.0775,
+      "step": 1482
+    },
+    {
+      "epoch": 0.33955352032054953,
+      "grad_norm": 1.1785542964935303,
+      "learning_rate": 1.5403287524205577e-05,
+      "loss": 1.0809,
+      "step": 1483
+    },
+    {
+      "epoch": 0.33978248425872926,
+      "grad_norm": 1.185001254081726,
+      "learning_rate": 1.539704399389625e-05,
+      "loss": 0.996,
+      "step": 1484
+    },
+    {
+      "epoch": 0.340011448196909,
+      "grad_norm": 1.155920386314392,
+      "learning_rate": 1.5390797493638862e-05,
+      "loss": 1.0949,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3402404121350887,
+      "grad_norm": 1.0656671524047852,
+      "learning_rate": 1.538454802687081e-05,
+      "loss": 1.0234,
+      "step": 1486
+    },
+    {
+      "epoch": 0.34046937607326844,
+      "grad_norm": 1.3848119974136353,
+      "learning_rate": 1.537829559703112e-05,
+      "loss": 1.0846,
+      "step": 1487
+    },
+    {
+      "epoch": 0.3406983400114482,
+      "grad_norm": 1.0920829772949219,
+      "learning_rate": 1.5372040207560457e-05,
+      "loss": 1.1337,
+      "step": 1488
+    },
+    {
+      "epoch": 0.34092730394962795,
+      "grad_norm": 1.3589757680892944,
+      "learning_rate": 1.536578186190111e-05,
+      "loss": 1.0797,
+      "step": 1489
+    },
+    {
+      "epoch": 0.3411562678878077,
+      "grad_norm": 1.4441230297088623,
+      "learning_rate": 1.5359520563496985e-05,
+      "loss": 1.131,
+      "step": 1490
+    },
+    {
+      "epoch": 0.3413852318259874,
+      "grad_norm": 1.3296109437942505,
+      "learning_rate": 1.5353256315793633e-05,
+      "loss": 1.086,
+      "step": 1491
+    },
+    {
+      "epoch": 0.34161419576416713,
+      "grad_norm": 1.065158486366272,
+      "learning_rate": 1.534698912223821e-05,
+      "loss": 1.0374,
+      "step": 1492
+    },
+    {
+      "epoch": 0.34184315970234685,
+      "grad_norm": 2.7994022369384766,
+      "learning_rate": 1.5340718986279505e-05,
+      "loss": 1.0859,
+      "step": 1493
+    },
+    {
+      "epoch": 0.34207212364052664,
+      "grad_norm": 1.2376704216003418,
+      "learning_rate": 1.5334445911367915e-05,
+      "loss": 1.0435,
+      "step": 1494
+    },
+    {
+      "epoch": 0.34230108757870636,
+      "grad_norm": 1.0192632675170898,
+      "learning_rate": 1.5328169900955463e-05,
+      "loss": 1.114,
+      "step": 1495
+    },
+    {
+      "epoch": 0.3425300515168861,
+      "grad_norm": 1.3984559774398804,
+      "learning_rate": 1.5321890958495787e-05,
+      "loss": 1.0808,
+      "step": 1496
+    },
+    {
+      "epoch": 0.3427590154550658,
+      "grad_norm": 0.9953001141548157,
+      "learning_rate": 1.5315609087444135e-05,
+      "loss": 1.0632,
+      "step": 1497
+    },
+    {
+      "epoch": 0.34298797939324555,
+      "grad_norm": 1.4686434268951416,
+      "learning_rate": 1.5309324291257373e-05,
+      "loss": 1.0939,
+      "step": 1498
+    },
+    {
+      "epoch": 0.34321694333142533,
+      "grad_norm": 1.450624704360962,
+      "learning_rate": 1.5303036573393964e-05,
+      "loss": 1.0561,
+      "step": 1499
+    },
+    {
+      "epoch": 0.34344590726960506,
+      "grad_norm": 1.1755778789520264,
+      "learning_rate": 1.529674593731399e-05,
+      "loss": 1.1257,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3436748712077848,
+      "grad_norm": 1.1907165050506592,
+      "learning_rate": 1.5290452386479132e-05,
+      "loss": 1.1242,
+      "step": 1501
+    },
+    {
+      "epoch": 0.3439038351459645,
+      "grad_norm": 1.4106594324111938,
+      "learning_rate": 1.5284155924352678e-05,
+      "loss": 1.1147,
+      "step": 1502
+    },
+    {
+      "epoch": 0.34413279908414424,
+      "grad_norm": 1.3477283716201782,
+      "learning_rate": 1.5277856554399528e-05,
+      "loss": 1.1203,
+      "step": 1503
+    },
+    {
+      "epoch": 0.34436176302232396,
+      "grad_norm": 1.1035650968551636,
+      "learning_rate": 1.5271554280086164e-05,
+      "loss": 1.0926,
+      "step": 1504
+    },
+    {
+      "epoch": 0.34459072696050375,
+      "grad_norm": 1.4535436630249023,
+      "learning_rate": 1.5265249104880675e-05,
+      "loss": 1.0157,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3448196908986835,
+      "grad_norm": 1.184928297996521,
+      "learning_rate": 1.5258941032252747e-05,
+      "loss": 1.0421,
+      "step": 1506
+    },
+    {
+      "epoch": 0.3450486548368632,
+      "grad_norm": 1.1650036573410034,
+      "learning_rate": 1.5252630065673662e-05,
+      "loss": 1.0696,
+      "step": 1507
+    },
+    {
+      "epoch": 0.3452776187750429,
+      "grad_norm": 1.272552490234375,
+      "learning_rate": 1.5246316208616289e-05,
+      "loss": 1.111,
+      "step": 1508
+    },
+    {
+      "epoch": 0.34550658271322265,
+      "grad_norm": 1.1120789051055908,
+      "learning_rate": 1.5239999464555092e-05,
+      "loss": 1.1354,
+      "step": 1509
+    },
+    {
+      "epoch": 0.3457355466514024,
+      "grad_norm": 1.6743152141571045,
+      "learning_rate": 1.5233679836966122e-05,
+      "loss": 1.0393,
+      "step": 1510
+    },
+    {
+      "epoch": 0.34596451058958216,
+      "grad_norm": 1.243407130241394,
+      "learning_rate": 1.5227357329327016e-05,
+      "loss": 1.0646,
+      "step": 1511
+    },
+    {
+      "epoch": 0.3461934745277619,
+      "grad_norm": 1.359520673751831,
+      "learning_rate": 1.5221031945116998e-05,
+      "loss": 1.1385,
+      "step": 1512
+    },
+    {
+      "epoch": 0.3464224384659416,
+      "grad_norm": 1.18605637550354,
+      "learning_rate": 1.5214703687816874e-05,
+      "loss": 1.1231,
+      "step": 1513
+    },
+    {
+      "epoch": 0.34665140240412134,
+      "grad_norm": 1.3147512674331665,
+      "learning_rate": 1.5208372560909031e-05,
+      "loss": 1.1246,
+      "step": 1514
+    },
+    {
+      "epoch": 0.34688036634230107,
+      "grad_norm": 1.420447826385498,
+      "learning_rate": 1.5202038567877436e-05,
+      "loss": 1.1292,
+      "step": 1515
+    },
+    {
+      "epoch": 0.3471093302804808,
+      "grad_norm": 1.1527974605560303,
+      "learning_rate": 1.5195701712207627e-05,
+      "loss": 1.0658,
+      "step": 1516
+    },
+    {
+      "epoch": 0.3473382942186606,
+      "grad_norm": 1.1984782218933105,
+      "learning_rate": 1.5189361997386729e-05,
+      "loss": 1.0923,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3475672581568403,
+      "grad_norm": 1.3671753406524658,
+      "learning_rate": 1.5183019426903434e-05,
+      "loss": 1.2041,
+      "step": 1518
+    },
+    {
+      "epoch": 0.34779622209502004,
+      "grad_norm": 1.3007023334503174,
+      "learning_rate": 1.5176674004247998e-05,
+      "loss": 1.0656,
+      "step": 1519
+    },
+    {
+      "epoch": 0.34802518603319976,
+      "grad_norm": 1.2162976264953613,
+      "learning_rate": 1.517032573291226e-05,
+      "loss": 1.1123,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3482541499713795,
+      "grad_norm": 1.119856834411621,
+      "learning_rate": 1.5163974616389621e-05,
+      "loss": 1.0619,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3484831139095592,
+      "grad_norm": 1.3005276918411255,
+      "learning_rate": 1.5157620658175046e-05,
+      "loss": 1.0622,
+      "step": 1522
+    },
+    {
+      "epoch": 0.348712077847739,
+      "grad_norm": 1.0568853616714478,
+      "learning_rate": 1.515126386176506e-05,
+      "loss": 1.0777,
+      "step": 1523
+    },
+    {
+      "epoch": 0.3489410417859187,
+      "grad_norm": 1.326450228691101,
+      "learning_rate": 1.5144904230657765e-05,
+      "loss": 1.0952,
+      "step": 1524
+    },
+    {
+      "epoch": 0.34917000572409845,
+      "grad_norm": 1.2039971351623535,
+      "learning_rate": 1.5138541768352802e-05,
+      "loss": 1.1044,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3493989696622782,
+      "grad_norm": 1.1191279888153076,
+      "learning_rate": 1.5132176478351386e-05,
+      "loss": 1.0931,
+      "step": 1526
+    },
+    {
+      "epoch": 0.3496279336004579,
+      "grad_norm": 1.0349925756454468,
+      "learning_rate": 1.5125808364156283e-05,
+      "loss": 1.1128,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3498568975386377,
+      "grad_norm": 1.1004794836044312,
+      "learning_rate": 1.5119437429271813e-05,
+      "loss": 1.1096,
+      "step": 1528
+    },
+    {
+      "epoch": 0.3500858614768174,
+      "grad_norm": 1.117506504058838,
+      "learning_rate": 1.5113063677203847e-05,
+      "loss": 1.0509,
+      "step": 1529
+    },
+    {
+      "epoch": 0.35031482541499714,
+      "grad_norm": 1.113121509552002,
+      "learning_rate": 1.5106687111459809e-05,
+      "loss": 1.1017,
+      "step": 1530
+    },
+    {
+      "epoch": 0.35054378935317687,
+      "grad_norm": 3.266343116760254,
+      "learning_rate": 1.5100307735548662e-05,
+      "loss": 1.1525,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3507727532913566,
+      "grad_norm": 1.3672198057174683,
+      "learning_rate": 1.5093925552980934e-05,
+      "loss": 1.0521,
+      "step": 1532
+    },
+    {
+      "epoch": 0.3510017172295363,
+      "grad_norm": 1.2172337770462036,
+      "learning_rate": 1.5087540567268682e-05,
+      "loss": 1.0226,
+      "step": 1533
+    },
+    {
+      "epoch": 0.3512306811677161,
+      "grad_norm": 0.9038141369819641,
+      "learning_rate": 1.5081152781925508e-05,
+      "loss": 1.0518,
+      "step": 1534
+    },
+    {
+      "epoch": 0.35145964510589583,
+      "grad_norm": 1.2456227540969849,
+      "learning_rate": 1.5074762200466557e-05,
+      "loss": 1.121,
+      "step": 1535
+    },
+    {
+      "epoch": 0.35168860904407556,
+      "grad_norm": 1.2485681772232056,
+      "learning_rate": 1.5068368826408515e-05,
+      "loss": 1.0517,
+      "step": 1536
+    },
+    {
+      "epoch": 0.3519175729822553,
+      "grad_norm": 1.2374999523162842,
+      "learning_rate": 1.5061972663269604e-05,
+      "loss": 1.0035,
+      "step": 1537
+    },
+    {
+      "epoch": 0.352146536920435,
+      "grad_norm": 1.0219475030899048,
+      "learning_rate": 1.5055573714569574e-05,
+      "loss": 1.0786,
+      "step": 1538
+    },
+    {
+      "epoch": 0.35237550085861474,
+      "grad_norm": 1.2069019079208374,
+      "learning_rate": 1.5049171983829714e-05,
+      "loss": 1.0511,
+      "step": 1539
+    },
+    {
+      "epoch": 0.3526044647967945,
+      "grad_norm": 1.256385087966919,
+      "learning_rate": 1.5042767474572846e-05,
+      "loss": 1.1174,
+      "step": 1540
+    },
+    {
+      "epoch": 0.35283342873497425,
+      "grad_norm": 1.2299443483352661,
+      "learning_rate": 1.5036360190323315e-05,
+      "loss": 1.0537,
+      "step": 1541
+    },
+    {
+      "epoch": 0.353062392673154,
+      "grad_norm": 1.5519248247146606,
+      "learning_rate": 1.5029950134606991e-05,
+      "loss": 1.131,
+      "step": 1542
+    },
+    {
+      "epoch": 0.3532913566113337,
+      "grad_norm": 1.0618422031402588,
+      "learning_rate": 1.5023537310951284e-05,
+      "loss": 1.1499,
+      "step": 1543
+    },
+    {
+      "epoch": 0.35352032054951343,
+      "grad_norm": 1.1327383518218994,
+      "learning_rate": 1.501712172288511e-05,
+      "loss": 1.1216,
+      "step": 1544
+    },
+    {
+      "epoch": 0.35374928448769316,
+      "grad_norm": 1.2645106315612793,
+      "learning_rate": 1.5010703373938915e-05,
+      "loss": 1.108,
+      "step": 1545
+    },
+    {
+      "epoch": 0.35397824842587294,
+      "grad_norm": 1.7846484184265137,
+      "learning_rate": 1.5004282267644668e-05,
+      "loss": 1.0953,
+      "step": 1546
+    },
+    {
+      "epoch": 0.35420721236405267,
+      "grad_norm": 1.1685068607330322,
+      "learning_rate": 1.4997858407535841e-05,
+      "loss": 1.0658,
+      "step": 1547
+    },
+    {
+      "epoch": 0.3544361763022324,
+      "grad_norm": 0.9860405921936035,
+      "learning_rate": 1.4991431797147433e-05,
+      "loss": 1.053,
+      "step": 1548
+    },
+    {
+      "epoch": 0.3546651402404121,
+      "grad_norm": 1.2382588386535645,
+      "learning_rate": 1.4985002440015959e-05,
+      "loss": 1.112,
+      "step": 1549
+    },
+    {
+      "epoch": 0.35489410417859185,
+      "grad_norm": 1.1630792617797852,
+      "learning_rate": 1.4978570339679435e-05,
+      "loss": 1.0488,
+      "step": 1550
+    },
+    {
+      "epoch": 0.35512306811677163,
+      "grad_norm": 1.154436469078064,
+      "learning_rate": 1.4972135499677395e-05,
+      "loss": 1.0509,
+      "step": 1551
+    },
+    {
+      "epoch": 0.35535203205495136,
+      "grad_norm": 1.0315396785736084,
+      "learning_rate": 1.4965697923550873e-05,
+      "loss": 1.0807,
+      "step": 1552
+    },
+    {
+      "epoch": 0.3555809959931311,
+      "grad_norm": 0.9754356741905212,
+      "learning_rate": 1.4959257614842416e-05,
+      "loss": 1.1137,
+      "step": 1553
+    },
+    {
+      "epoch": 0.3558099599313108,
+      "grad_norm": 1.0511184930801392,
+      "learning_rate": 1.4952814577096073e-05,
+      "loss": 1.1068,
+      "step": 1554
+    },
+    {
+      "epoch": 0.35603892386949054,
+      "grad_norm": 1.3357048034667969,
+      "learning_rate": 1.4946368813857393e-05,
+      "loss": 1.053,
+      "step": 1555
+    },
+    {
+      "epoch": 0.35626788780767027,
+      "grad_norm": 0.9737345576286316,
+      "learning_rate": 1.4939920328673422e-05,
+      "loss": 1.0613,
+      "step": 1556
+    },
+    {
+      "epoch": 0.35649685174585005,
+      "grad_norm": 1.1738935708999634,
+      "learning_rate": 1.4933469125092714e-05,
+      "loss": 1.0548,
+      "step": 1557
+    },
+    {
+      "epoch": 0.3567258156840298,
+      "grad_norm": 1.0252792835235596,
+      "learning_rate": 1.4927015206665311e-05,
+      "loss": 1.1409,
+      "step": 1558
+    },
+    {
+      "epoch": 0.3569547796222095,
+      "grad_norm": 1.1485637426376343,
+      "learning_rate": 1.4920558576942746e-05,
+      "loss": 1.1199,
+      "step": 1559
+    },
+    {
+      "epoch": 0.35718374356038923,
+      "grad_norm": 1.2590608596801758,
+      "learning_rate": 1.4914099239478046e-05,
+      "loss": 1.0939,
+      "step": 1560
+    },
+    {
+      "epoch": 0.35741270749856896,
+      "grad_norm": 1.1943340301513672,
+      "learning_rate": 1.490763719782574e-05,
+      "loss": 1.049,
+      "step": 1561
+    },
+    {
+      "epoch": 0.3576416714367487,
+      "grad_norm": 1.8277630805969238,
+      "learning_rate": 1.4901172455541826e-05,
+      "loss": 1.1446,
+      "step": 1562
+    },
+    {
+      "epoch": 0.35787063537492847,
+      "grad_norm": 1.2777670621871948,
+      "learning_rate": 1.4894705016183803e-05,
+      "loss": 1.0863,
+      "step": 1563
+    },
+    {
+      "epoch": 0.3580995993131082,
+      "grad_norm": 1.1132858991622925,
+      "learning_rate": 1.4888234883310644e-05,
+      "loss": 1.1255,
+      "step": 1564
+    },
+    {
+      "epoch": 0.3583285632512879,
+      "grad_norm": 1.2168593406677246,
+      "learning_rate": 1.4881762060482814e-05,
+      "loss": 1.1271,
+      "step": 1565
+    },
+    {
+      "epoch": 0.35855752718946765,
+      "grad_norm": 1.321545124053955,
+      "learning_rate": 1.4875286551262252e-05,
+      "loss": 1.0964,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3587864911276474,
+      "grad_norm": 1.1392179727554321,
+      "learning_rate": 1.4868808359212373e-05,
+      "loss": 1.0022,
+      "step": 1567
+    },
+    {
+      "epoch": 0.3590154550658271,
+      "grad_norm": 1.227779746055603,
+      "learning_rate": 1.4862327487898075e-05,
+      "loss": 1.1225,
+      "step": 1568
+    },
+    {
+      "epoch": 0.3592444190040069,
+      "grad_norm": 1.1016408205032349,
+      "learning_rate": 1.4855843940885726e-05,
+      "loss": 1.0832,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3594733829421866,
+      "grad_norm": 1.0965155363082886,
+      "learning_rate": 1.4849357721743169e-05,
+      "loss": 1.0853,
+      "step": 1570
+    },
+    {
+      "epoch": 0.35970234688036634,
+      "grad_norm": 1.1590499877929688,
+      "learning_rate": 1.484286883403971e-05,
+      "loss": 1.0385,
+      "step": 1571
+    },
+    {
+      "epoch": 0.35993131081854607,
+      "grad_norm": 1.0855969190597534,
+      "learning_rate": 1.483637728134614e-05,
+      "loss": 1.0943,
+      "step": 1572
+    },
+    {
+      "epoch": 0.3601602747567258,
+      "grad_norm": 1.1280837059020996,
+      "learning_rate": 1.4829883067234699e-05,
+      "loss": 1.0209,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3603892386949056,
+      "grad_norm": 1.0733137130737305,
+      "learning_rate": 1.4823386195279098e-05,
+      "loss": 1.1372,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3606182026330853,
+      "grad_norm": 1.1513261795043945,
+      "learning_rate": 1.4816886669054514e-05,
+      "loss": 1.0584,
+      "step": 1575
+    },
+    {
+      "epoch": 0.36084716657126503,
+      "grad_norm": 1.3716681003570557,
+      "learning_rate": 1.4810384492137582e-05,
+      "loss": 1.1219,
+      "step": 1576
+    },
+    {
+      "epoch": 0.36107613050944476,
+      "grad_norm": 0.9845678806304932,
+      "learning_rate": 1.4803879668106393e-05,
+      "loss": 1.0914,
+      "step": 1577
+    },
+    {
+      "epoch": 0.3613050944476245,
+      "grad_norm": 1.257351040840149,
+      "learning_rate": 1.4797372200540497e-05,
+      "loss": 1.1247,
+      "step": 1578
+    },
+    {
+      "epoch": 0.3615340583858042,
+      "grad_norm": 1.4582107067108154,
+      "learning_rate": 1.4790862093020903e-05,
+      "loss": 1.0401,
+      "step": 1579
+    },
+    {
+      "epoch": 0.361763022323984,
+      "grad_norm": 1.277854323387146,
+      "learning_rate": 1.4784349349130063e-05,
+      "loss": 1.0367,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3619919862621637,
+      "grad_norm": 1.1028213500976562,
+      "learning_rate": 1.4777833972451889e-05,
+      "loss": 1.0808,
+      "step": 1581
+    },
+    {
+      "epoch": 0.36222095020034345,
+      "grad_norm": 1.0931235551834106,
+      "learning_rate": 1.477131596657174e-05,
+      "loss": 1.0286,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3624499141385232,
+      "grad_norm": 1.4668598175048828,
+      "learning_rate": 1.4764795335076414e-05,
+      "loss": 1.1162,
+      "step": 1583
+    },
+    {
+      "epoch": 0.3626788780767029,
+      "grad_norm": 2.652813673019409,
+      "learning_rate": 1.4758272081554168e-05,
+      "loss": 1.1068,
+      "step": 1584
+    },
+    {
+      "epoch": 0.36290784201488263,
+      "grad_norm": 1.1767535209655762,
+      "learning_rate": 1.4751746209594683e-05,
+      "loss": 1.0795,
+      "step": 1585
+    },
+    {
+      "epoch": 0.3631368059530624,
+      "grad_norm": 1.0798718929290771,
+      "learning_rate": 1.47452177227891e-05,
+      "loss": 1.0911,
+      "step": 1586
+    },
+    {
+      "epoch": 0.36336576989124214,
+      "grad_norm": 1.119814395904541,
+      "learning_rate": 1.4738686624729987e-05,
+      "loss": 1.0673,
+      "step": 1587
+    },
+    {
+      "epoch": 0.36359473382942187,
+      "grad_norm": 1.2240833044052124,
+      "learning_rate": 1.4732152919011355e-05,
+      "loss": 1.0708,
+      "step": 1588
+    },
+    {
+      "epoch": 0.3638236977676016,
+      "grad_norm": 1.1120808124542236,
+      "learning_rate": 1.4725616609228648e-05,
+      "loss": 1.0919,
+      "step": 1589
+    },
+    {
+      "epoch": 0.3640526617057813,
+      "grad_norm": 1.2676600217819214,
+      "learning_rate": 1.4719077698978737e-05,
+      "loss": 1.0665,
+      "step": 1590
+    },
+    {
+      "epoch": 0.36428162564396105,
+      "grad_norm": 1.1480255126953125,
+      "learning_rate": 1.4712536191859934e-05,
+      "loss": 1.0787,
+      "step": 1591
+    },
+    {
+      "epoch": 0.36451058958214083,
+      "grad_norm": 1.7230538129806519,
+      "learning_rate": 1.4705992091471975e-05,
+      "loss": 1.1455,
+      "step": 1592
+    },
+    {
+      "epoch": 0.36473955352032056,
+      "grad_norm": 1.1542283296585083,
+      "learning_rate": 1.4699445401416024e-05,
+      "loss": 1.1246,
+      "step": 1593
+    },
+    {
+      "epoch": 0.3649685174585003,
+      "grad_norm": 1.230782389640808,
+      "learning_rate": 1.4692896125294667e-05,
+      "loss": 1.1106,
+      "step": 1594
+    },
+    {
+      "epoch": 0.36519748139668,
+      "grad_norm": 1.2832047939300537,
+      "learning_rate": 1.4686344266711916e-05,
+      "loss": 1.1203,
+      "step": 1595
+    },
+    {
+      "epoch": 0.36542644533485974,
+      "grad_norm": 1.213179111480713,
+      "learning_rate": 1.467978982927321e-05,
+      "loss": 1.1387,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3656554092730395,
+      "grad_norm": 1.2163344621658325,
+      "learning_rate": 1.4673232816585392e-05,
+      "loss": 1.1456,
+      "step": 1597
+    },
+    {
+      "epoch": 0.36588437321121925,
+      "grad_norm": 1.1617780923843384,
+      "learning_rate": 1.4666673232256738e-05,
+      "loss": 1.091,
+      "step": 1598
+    },
+    {
+      "epoch": 0.366113337149399,
+      "grad_norm": 1.3086128234863281,
+      "learning_rate": 1.466011107989693e-05,
+      "loss": 1.096,
+      "step": 1599
+    },
+    {
+      "epoch": 0.3663423010875787,
+      "grad_norm": 1.1292928457260132,
+      "learning_rate": 1.4653546363117063e-05,
+      "loss": 1.0508,
+      "step": 1600
+    },
+    {
+      "epoch": 0.36657126502575843,
+      "grad_norm": 1.2157028913497925,
+      "learning_rate": 1.464697908552965e-05,
+      "loss": 1.0545,
+      "step": 1601
+    },
+    {
+      "epoch": 0.36680022896393816,
+      "grad_norm": 1.3973489999771118,
+      "learning_rate": 1.4640409250748604e-05,
+      "loss": 1.0731,
+      "step": 1602
+    },
+    {
+      "epoch": 0.36702919290211794,
+      "grad_norm": 1.1359471082687378,
+      "learning_rate": 1.4633836862389257e-05,
+      "loss": 1.0881,
+      "step": 1603
+    },
+    {
+      "epoch": 0.36725815684029767,
+      "grad_norm": 1.074826955795288,
+      "learning_rate": 1.4627261924068329e-05,
+      "loss": 1.0443,
+      "step": 1604
+    },
+    {
+      "epoch": 0.3674871207784774,
+      "grad_norm": 1.4347003698349,
+      "learning_rate": 1.4620684439403962e-05,
+      "loss": 1.07,
+      "step": 1605
+    },
+    {
+      "epoch": 0.3677160847166571,
+      "grad_norm": 1.0502759218215942,
+      "learning_rate": 1.4614104412015688e-05,
+      "loss": 1.0734,
+      "step": 1606
+    },
+    {
+      "epoch": 0.36794504865483685,
+      "grad_norm": 1.6279913187026978,
+      "learning_rate": 1.4607521845524439e-05,
+      "loss": 1.0855,
+      "step": 1607
+    },
+    {
+      "epoch": 0.3681740125930166,
+      "grad_norm": 1.2757513523101807,
+      "learning_rate": 1.460093674355255e-05,
+      "loss": 1.0385,
+      "step": 1608
+    },
+    {
+      "epoch": 0.36840297653119636,
+      "grad_norm": 1.1416752338409424,
+      "learning_rate": 1.4594349109723744e-05,
+      "loss": 1.0783,
+      "step": 1609
+    },
+    {
+      "epoch": 0.3686319404693761,
+      "grad_norm": 1.1832513809204102,
+      "learning_rate": 1.4587758947663146e-05,
+      "loss": 1.0836,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3688609044075558,
+      "grad_norm": 1.1325169801712036,
+      "learning_rate": 1.4581166260997259e-05,
+      "loss": 1.0372,
+      "step": 1611
+    },
+    {
+      "epoch": 0.36908986834573554,
+      "grad_norm": 1.3580266237258911,
+      "learning_rate": 1.4574571053353987e-05,
+      "loss": 1.0604,
+      "step": 1612
+    },
+    {
+      "epoch": 0.36931883228391527,
+      "grad_norm": 1.2545275688171387,
+      "learning_rate": 1.4567973328362616e-05,
+      "loss": 1.0859,
+      "step": 1613
+    },
+    {
+      "epoch": 0.369547796222095,
+      "grad_norm": 1.1859338283538818,
+      "learning_rate": 1.4561373089653823e-05,
+      "loss": 1.1461,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3697767601602748,
+      "grad_norm": 1.2535377740859985,
+      "learning_rate": 1.4554770340859661e-05,
+      "loss": 1.0554,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3700057240984545,
+      "grad_norm": 1.490103840827942,
+      "learning_rate": 1.4548165085613569e-05,
+      "loss": 1.0364,
+      "step": 1616
+    },
+    {
+      "epoch": 0.37023468803663423,
+      "grad_norm": 1.1767431497573853,
+      "learning_rate": 1.454155732755036e-05,
+      "loss": 1.0654,
+      "step": 1617
+    },
+    {
+      "epoch": 0.37046365197481396,
+      "grad_norm": 1.525582194328308,
+      "learning_rate": 1.453494707030623e-05,
+      "loss": 1.0878,
+      "step": 1618
+    },
+    {
+      "epoch": 0.3706926159129937,
+      "grad_norm": 1.5364714860916138,
+      "learning_rate": 1.452833431751875e-05,
+      "loss": 1.0832,
+      "step": 1619
+    },
+    {
+      "epoch": 0.37092157985117347,
+      "grad_norm": 1.1335092782974243,
+      "learning_rate": 1.4521719072826858e-05,
+      "loss": 1.0521,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3711505437893532,
+      "grad_norm": 1.278962254524231,
+      "learning_rate": 1.4515101339870871e-05,
+      "loss": 1.0582,
+      "step": 1621
+    },
+    {
+      "epoch": 0.3713795077275329,
+      "grad_norm": 1.424355387687683,
+      "learning_rate": 1.4508481122292475e-05,
+      "loss": 1.0583,
+      "step": 1622
+    },
+    {
+      "epoch": 0.37160847166571265,
+      "grad_norm": 1.1989699602127075,
+      "learning_rate": 1.4501858423734711e-05,
+      "loss": 1.0488,
+      "step": 1623
+    },
+    {
+      "epoch": 0.3718374356038924,
+      "grad_norm": 1.4514950513839722,
+      "learning_rate": 1.4495233247842001e-05,
+      "loss": 1.1174,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3720663995420721,
+      "grad_norm": 1.0482782125473022,
+      "learning_rate": 1.4488605598260129e-05,
+      "loss": 1.085,
+      "step": 1625
+    },
+    {
+      "epoch": 0.3722953634802519,
+      "grad_norm": 1.019201636314392,
+      "learning_rate": 1.448197547863622e-05,
+      "loss": 1.0703,
+      "step": 1626
+    },
+    {
+      "epoch": 0.3725243274184316,
+      "grad_norm": 1.3561880588531494,
+      "learning_rate": 1.4475342892618792e-05,
+      "loss": 1.0594,
+      "step": 1627
+    },
+    {
+      "epoch": 0.37275329135661134,
+      "grad_norm": 1.0671643018722534,
+      "learning_rate": 1.4468707843857683e-05,
+      "loss": 1.0137,
+      "step": 1628
+    },
+    {
+      "epoch": 0.37298225529479107,
+      "grad_norm": 1.2708206176757812,
+      "learning_rate": 1.4462070336004117e-05,
+      "loss": 1.0725,
+      "step": 1629
+    },
+    {
+      "epoch": 0.3732112192329708,
+      "grad_norm": 2.9247419834136963,
+      "learning_rate": 1.4455430372710652e-05,
+      "loss": 1.144,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3734401831711505,
+      "grad_norm": 1.4302781820297241,
+      "learning_rate": 1.4448787957631209e-05,
+      "loss": 1.0814,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3736691471093303,
+      "grad_norm": 1.1992768049240112,
+      "learning_rate": 1.4442143094421054e-05,
+      "loss": 1.0457,
+      "step": 1632
+    },
+    {
+      "epoch": 0.37389811104751003,
+      "grad_norm": 1.1916435956954956,
+      "learning_rate": 1.4435495786736796e-05,
+      "loss": 1.1704,
+      "step": 1633
+    },
+    {
+      "epoch": 0.37412707498568976,
+      "grad_norm": 1.3071752786636353,
+      "learning_rate": 1.4428846038236391e-05,
+      "loss": 1.0534,
+      "step": 1634
+    },
+    {
+      "epoch": 0.3743560389238695,
+      "grad_norm": 1.2490984201431274,
+      "learning_rate": 1.4422193852579144e-05,
+      "loss": 1.0872,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3745850028620492,
+      "grad_norm": 1.0837894678115845,
+      "learning_rate": 1.4415539233425697e-05,
+      "loss": 1.1409,
+      "step": 1636
+    },
+    {
+      "epoch": 0.37481396680022894,
+      "grad_norm": 1.1508346796035767,
+      "learning_rate": 1.4408882184438029e-05,
+      "loss": 1.1122,
+      "step": 1637
+    },
+    {
+      "epoch": 0.3750429307384087,
+      "grad_norm": 1.1703577041625977,
+      "learning_rate": 1.4402222709279458e-05,
+      "loss": 1.0457,
+      "step": 1638
+    },
+    {
+      "epoch": 0.37527189467658845,
+      "grad_norm": 1.1867021322250366,
+      "learning_rate": 1.439556081161464e-05,
+      "loss": 1.0956,
+      "step": 1639
+    },
+    {
+      "epoch": 0.3755008586147682,
+      "grad_norm": 1.2426538467407227,
+      "learning_rate": 1.438889649510956e-05,
+      "loss": 1.0897,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3757298225529479,
+      "grad_norm": 1.1953974962234497,
+      "learning_rate": 1.4382229763431533e-05,
+      "loss": 1.096,
+      "step": 1641
+    },
+    {
+      "epoch": 0.37595878649112763,
+      "grad_norm": 1.0704538822174072,
+      "learning_rate": 1.4375560620249209e-05,
+      "loss": 1.079,
+      "step": 1642
+    },
+    {
+      "epoch": 0.3761877504293074,
+      "grad_norm": 1.342177152633667,
+      "learning_rate": 1.4368889069232559e-05,
+      "loss": 1.1331,
+      "step": 1643
+    },
+    {
+      "epoch": 0.37641671436748714,
+      "grad_norm": 1.2774497270584106,
+      "learning_rate": 1.4362215114052887e-05,
+      "loss": 1.0949,
+      "step": 1644
+    },
+    {
+      "epoch": 0.37664567830566686,
+      "grad_norm": 1.1422805786132812,
+      "learning_rate": 1.4355538758382805e-05,
+      "loss": 1.087,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3768746422438466,
+      "grad_norm": 1.1978528499603271,
+      "learning_rate": 1.4348860005896266e-05,
+      "loss": 1.0468,
+      "step": 1646
+    },
+    {
+      "epoch": 0.3771036061820263,
+      "grad_norm": 1.2687922716140747,
+      "learning_rate": 1.4342178860268523e-05,
+      "loss": 1.0217,
+      "step": 1647
+    },
+    {
+      "epoch": 0.37733257012020605,
+      "grad_norm": 1.1119877099990845,
+      "learning_rate": 1.433549532517616e-05,
+      "loss": 1.0787,
+      "step": 1648
+    },
+    {
+      "epoch": 0.37756153405838583,
+      "grad_norm": 0.9895734190940857,
+      "learning_rate": 1.4328809404297068e-05,
+      "loss": 1.0796,
+      "step": 1649
+    },
+    {
+      "epoch": 0.37779049799656556,
+      "grad_norm": 1.1445213556289673,
+      "learning_rate": 1.4322121101310454e-05,
+      "loss": 1.0889,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3780194619347453,
+      "grad_norm": 1.115267276763916,
+      "learning_rate": 1.4315430419896836e-05,
+      "loss": 1.0857,
+      "step": 1651
+    },
+    {
+      "epoch": 0.378248425872925,
+      "grad_norm": 1.3437042236328125,
+      "learning_rate": 1.4308737363738035e-05,
+      "loss": 1.0725,
+      "step": 1652
+    },
+    {
+      "epoch": 0.37847738981110474,
+      "grad_norm": 1.1512742042541504,
+      "learning_rate": 1.430204193651719e-05,
+      "loss": 1.0748,
+      "step": 1653
+    },
+    {
+      "epoch": 0.37870635374928446,
+      "grad_norm": 1.109114646911621,
+      "learning_rate": 1.4295344141918734e-05,
+      "loss": 1.1062,
+      "step": 1654
+    },
+    {
+      "epoch": 0.37893531768746425,
+      "grad_norm": 1.2906250953674316,
+      "learning_rate": 1.428864398362841e-05,
+      "loss": 1.0753,
+      "step": 1655
+    },
+    {
+      "epoch": 0.379164281625644,
+      "grad_norm": 1.2104160785675049,
+      "learning_rate": 1.4281941465333255e-05,
+      "loss": 1.0533,
+      "step": 1656
+    },
+    {
+      "epoch": 0.3793932455638237,
+      "grad_norm": 1.1130504608154297,
+      "learning_rate": 1.4275236590721615e-05,
+      "loss": 1.0529,
+      "step": 1657
+    },
+    {
+      "epoch": 0.3796222095020034,
+      "grad_norm": 1.3344502449035645,
+      "learning_rate": 1.4268529363483124e-05,
+      "loss": 1.1511,
+      "step": 1658
+    },
+    {
+      "epoch": 0.37985117344018315,
+      "grad_norm": 1.1640992164611816,
+      "learning_rate": 1.4261819787308708e-05,
+      "loss": 1.0609,
+      "step": 1659
+    },
+    {
+      "epoch": 0.3800801373783629,
+      "grad_norm": 1.4107555150985718,
+      "learning_rate": 1.4255107865890597e-05,
+      "loss": 1.0672,
+      "step": 1660
+    },
+    {
+      "epoch": 0.38030910131654266,
+      "grad_norm": 1.214077115058899,
+      "learning_rate": 1.4248393602922299e-05,
+      "loss": 1.097,
+      "step": 1661
+    },
+    {
+      "epoch": 0.3805380652547224,
+      "grad_norm": 1.1887096166610718,
+      "learning_rate": 1.4241677002098622e-05,
+      "loss": 1.0425,
+      "step": 1662
+    },
+    {
+      "epoch": 0.3807670291929021,
+      "grad_norm": 1.1480146646499634,
+      "learning_rate": 1.4234958067115652e-05,
+      "loss": 1.0934,
+      "step": 1663
+    },
+    {
+      "epoch": 0.38099599313108184,
+      "grad_norm": 1.6082152128219604,
+      "learning_rate": 1.4228236801670762e-05,
+      "loss": 1.0686,
+      "step": 1664
+    },
+    {
+      "epoch": 0.38122495706926157,
+      "grad_norm": 1.2487170696258545,
+      "learning_rate": 1.4221513209462615e-05,
+      "loss": 1.041,
+      "step": 1665
+    },
+    {
+      "epoch": 0.38145392100744135,
+      "grad_norm": 1.2354825735092163,
+      "learning_rate": 1.4214787294191137e-05,
+      "loss": 1.1181,
+      "step": 1666
+    },
+    {
+      "epoch": 0.3816828849456211,
+      "grad_norm": 1.2729392051696777,
+      "learning_rate": 1.4208059059557551e-05,
+      "loss": 1.0707,
+      "step": 1667
+    },
+    {
+      "epoch": 0.3819118488838008,
+      "grad_norm": 1.0977466106414795,
+      "learning_rate": 1.420132850926434e-05,
+      "loss": 1.087,
+      "step": 1668
+    },
+    {
+      "epoch": 0.38214081282198054,
+      "grad_norm": 1.6709105968475342,
+      "learning_rate": 1.419459564701528e-05,
+      "loss": 1.1877,
+      "step": 1669
+    },
+    {
+      "epoch": 0.38236977676016026,
+      "grad_norm": 1.3484159708023071,
+      "learning_rate": 1.41878604765154e-05,
+      "loss": 1.1141,
+      "step": 1670
+    },
+    {
+      "epoch": 0.38259874069834,
+      "grad_norm": 1.3736979961395264,
+      "learning_rate": 1.4181123001471012e-05,
+      "loss": 1.0962,
+      "step": 1671
+    },
+    {
+      "epoch": 0.38282770463651977,
+      "grad_norm": 2.105513572692871,
+      "learning_rate": 1.4174383225589691e-05,
+      "loss": 1.1048,
+      "step": 1672
+    },
+    {
+      "epoch": 0.3830566685746995,
+      "grad_norm": 1.2961084842681885,
+      "learning_rate": 1.4167641152580278e-05,
+      "loss": 1.0663,
+      "step": 1673
+    },
+    {
+      "epoch": 0.3832856325128792,
+      "grad_norm": 1.142513394355774,
+      "learning_rate": 1.416089678615288e-05,
+      "loss": 1.0686,
+      "step": 1674
+    },
+    {
+      "epoch": 0.38351459645105895,
+      "grad_norm": 1.188399314880371,
+      "learning_rate": 1.4154150130018867e-05,
+      "loss": 1.1332,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3837435603892387,
+      "grad_norm": 1.2411218881607056,
+      "learning_rate": 1.4147401187890863e-05,
+      "loss": 1.1118,
+      "step": 1676
+    },
+    {
+      "epoch": 0.3839725243274184,
+      "grad_norm": 1.4901659488677979,
+      "learning_rate": 1.4140649963482763e-05,
+      "loss": 1.0477,
+      "step": 1677
+    },
+    {
+      "epoch": 0.3842014882655982,
+      "grad_norm": 1.0591933727264404,
+      "learning_rate": 1.4133896460509695e-05,
+      "loss": 1.0359,
+      "step": 1678
+    },
+    {
+      "epoch": 0.3844304522037779,
+      "grad_norm": 1.8987795114517212,
+      "learning_rate": 1.412714068268807e-05,
+      "loss": 1.0438,
+      "step": 1679
+    },
+    {
+      "epoch": 0.38465941614195764,
+      "grad_norm": 1.1370187997817993,
+      "learning_rate": 1.4120382633735528e-05,
+      "loss": 1.0796,
+      "step": 1680
+    },
+    {
+      "epoch": 0.38488838008013737,
+      "grad_norm": 1.178002953529358,
+      "learning_rate": 1.4113622317370965e-05,
+      "loss": 1.0361,
+      "step": 1681
+    },
+    {
+      "epoch": 0.3851173440183171,
+      "grad_norm": 0.9956271052360535,
+      "learning_rate": 1.4106859737314532e-05,
+      "loss": 1.0649,
+      "step": 1682
+    },
+    {
+      "epoch": 0.3853463079564968,
+      "grad_norm": 1.285373330116272,
+      "learning_rate": 1.4100094897287618e-05,
+      "loss": 1.1628,
+      "step": 1683
+    },
+    {
+      "epoch": 0.3855752718946766,
+      "grad_norm": 1.119768500328064,
+      "learning_rate": 1.4093327801012854e-05,
+      "loss": 1.1083,
+      "step": 1684
+    },
+    {
+      "epoch": 0.38580423583285633,
+      "grad_norm": 1.304608702659607,
+      "learning_rate": 1.4086558452214121e-05,
+      "loss": 1.0883,
+      "step": 1685
+    },
+    {
+      "epoch": 0.38603319977103606,
+      "grad_norm": 1.1423434019088745,
+      "learning_rate": 1.4079786854616537e-05,
+      "loss": 1.048,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3862621637092158,
+      "grad_norm": 1.159722089767456,
+      "learning_rate": 1.4073013011946449e-05,
+      "loss": 1.0642,
+      "step": 1687
+    },
+    {
+      "epoch": 0.3864911276473955,
+      "grad_norm": 1.1652872562408447,
+      "learning_rate": 1.4066236927931447e-05,
+      "loss": 1.1185,
+      "step": 1688
+    },
+    {
+      "epoch": 0.3867200915855753,
+      "grad_norm": 1.3069536685943604,
+      "learning_rate": 1.4059458606300358e-05,
+      "loss": 1.1017,
+      "step": 1689
+    },
+    {
+      "epoch": 0.386949055523755,
+      "grad_norm": 1.2639847993850708,
+      "learning_rate": 1.405267805078323e-05,
+      "loss": 1.1014,
+      "step": 1690
+    },
+    {
+      "epoch": 0.38717801946193475,
+      "grad_norm": 1.3884036540985107,
+      "learning_rate": 1.4045895265111352e-05,
+      "loss": 1.099,
+      "step": 1691
+    },
+    {
+      "epoch": 0.3874069834001145,
+      "grad_norm": 1.3688089847564697,
+      "learning_rate": 1.4039110253017225e-05,
+      "loss": 1.0532,
+      "step": 1692
+    },
+    {
+      "epoch": 0.3876359473382942,
+      "grad_norm": 1.2900769710540771,
+      "learning_rate": 1.4032323018234592e-05,
+      "loss": 1.0949,
+      "step": 1693
+    },
+    {
+      "epoch": 0.38786491127647393,
+      "grad_norm": 1.3169381618499756,
+      "learning_rate": 1.4025533564498411e-05,
+      "loss": 1.1092,
+      "step": 1694
+    },
+    {
+      "epoch": 0.3880938752146537,
+      "grad_norm": 1.2723472118377686,
+      "learning_rate": 1.401874189554486e-05,
+      "loss": 1.0769,
+      "step": 1695
+    },
+    {
+      "epoch": 0.38832283915283344,
+      "grad_norm": 1.1920543909072876,
+      "learning_rate": 1.4011948015111334e-05,
+      "loss": 1.1574,
+      "step": 1696
+    },
+    {
+      "epoch": 0.38855180309101317,
+      "grad_norm": 1.2525862455368042,
+      "learning_rate": 1.400515192693645e-05,
+      "loss": 1.1647,
+      "step": 1697
+    },
+    {
+      "epoch": 0.3887807670291929,
+      "grad_norm": 9.487869262695312,
+      "learning_rate": 1.3998353634760044e-05,
+      "loss": 1.1278,
+      "step": 1698
+    },
+    {
+      "epoch": 0.3890097309673726,
+      "grad_norm": 1.6089351177215576,
+      "learning_rate": 1.3991553142323156e-05,
+      "loss": 1.1101,
+      "step": 1699
+    },
+    {
+      "epoch": 0.38923869490555235,
+      "grad_norm": 1.3901886940002441,
+      "learning_rate": 1.3984750453368033e-05,
+      "loss": 1.1249,
+      "step": 1700
+    },
+    {
+      "epoch": 0.38946765884373213,
+      "grad_norm": 3.7920846939086914,
+      "learning_rate": 1.397794557163815e-05,
+      "loss": 1.0998,
+      "step": 1701
+    },
+    {
+      "epoch": 0.38969662278191186,
+      "grad_norm": 1.1165345907211304,
+      "learning_rate": 1.3971138500878166e-05,
+      "loss": 1.0719,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3899255867200916,
+      "grad_norm": 1.2537471055984497,
+      "learning_rate": 1.396432924483396e-05,
+      "loss": 1.0601,
+      "step": 1703
+    },
+    {
+      "epoch": 0.3901545506582713,
+      "grad_norm": 1.2301760911941528,
+      "learning_rate": 1.3957517807252607e-05,
+      "loss": 1.1573,
+      "step": 1704
+    },
+    {
+      "epoch": 0.39038351459645104,
+      "grad_norm": 1.645586609840393,
+      "learning_rate": 1.3950704191882388e-05,
+      "loss": 1.1273,
+      "step": 1705
+    },
+    {
+      "epoch": 0.39061247853463077,
+      "grad_norm": 1.3407573699951172,
+      "learning_rate": 1.3943888402472771e-05,
+      "loss": 1.0567,
+      "step": 1706
+    },
+    {
+      "epoch": 0.39084144247281055,
+      "grad_norm": 2.350411891937256,
+      "learning_rate": 1.393707044277443e-05,
+      "loss": 1.0961,
+      "step": 1707
+    },
+    {
+      "epoch": 0.3910704064109903,
+      "grad_norm": 1.187579870223999,
+      "learning_rate": 1.3930250316539237e-05,
+      "loss": 1.0545,
+      "step": 1708
+    },
+    {
+      "epoch": 0.39129937034917,
+      "grad_norm": 1.091801404953003,
+      "learning_rate": 1.3923428027520246e-05,
+      "loss": 1.0505,
+      "step": 1709
+    },
+    {
+      "epoch": 0.39152833428734973,
+      "grad_norm": 1.194849967956543,
+      "learning_rate": 1.3916603579471705e-05,
+      "loss": 1.0593,
+      "step": 1710
+    },
+    {
+      "epoch": 0.39175729822552946,
+      "grad_norm": 1.1145169734954834,
+      "learning_rate": 1.3909776976149047e-05,
+      "loss": 1.051,
+      "step": 1711
+    },
+    {
+      "epoch": 0.39198626216370924,
+      "grad_norm": 1.3400248289108276,
+      "learning_rate": 1.3902948221308903e-05,
+      "loss": 1.0907,
+      "step": 1712
+    },
+    {
+      "epoch": 0.39221522610188897,
+      "grad_norm": 1.3348122835159302,
+      "learning_rate": 1.3896117318709074e-05,
+      "loss": 1.0244,
+      "step": 1713
+    },
+    {
+      "epoch": 0.3924441900400687,
+      "grad_norm": 1.1804814338684082,
+      "learning_rate": 1.388928427210855e-05,
+      "loss": 1.1056,
+      "step": 1714
+    },
+    {
+      "epoch": 0.3926731539782484,
+      "grad_norm": 1.4333142042160034,
+      "learning_rate": 1.3882449085267497e-05,
+      "loss": 1.1095,
+      "step": 1715
+    },
+    {
+      "epoch": 0.39290211791642815,
+      "grad_norm": 1.10541570186615,
+      "learning_rate": 1.3875611761947264e-05,
+      "loss": 1.1461,
+      "step": 1716
+    },
+    {
+      "epoch": 0.3931310818546079,
+      "grad_norm": 1.3302122354507446,
+      "learning_rate": 1.3868772305910376e-05,
+      "loss": 1.043,
+      "step": 1717
+    },
+    {
+      "epoch": 0.39336004579278766,
+      "grad_norm": 1.2536112070083618,
+      "learning_rate": 1.3861930720920518e-05,
+      "loss": 1.0733,
+      "step": 1718
+    },
+    {
+      "epoch": 0.3935890097309674,
+      "grad_norm": 1.1884063482284546,
+      "learning_rate": 1.3855087010742563e-05,
+      "loss": 1.125,
+      "step": 1719
+    },
+    {
+      "epoch": 0.3938179736691471,
+      "grad_norm": 1.491746187210083,
+      "learning_rate": 1.384824117914255e-05,
+      "loss": 1.071,
+      "step": 1720
+    },
+    {
+      "epoch": 0.39404693760732684,
+      "grad_norm": 1.331817865371704,
+      "learning_rate": 1.3841393229887676e-05,
+      "loss": 1.084,
+      "step": 1721
+    },
+    {
+      "epoch": 0.39427590154550657,
+      "grad_norm": 1.232057809829712,
+      "learning_rate": 1.3834543166746317e-05,
+      "loss": 1.1158,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3945048654836863,
+      "grad_norm": 1.0441474914550781,
+      "learning_rate": 1.3827690993488e-05,
+      "loss": 1.1093,
+      "step": 1723
+    },
+    {
+      "epoch": 0.3947338294218661,
+      "grad_norm": 1.1634544134140015,
+      "learning_rate": 1.3820836713883424e-05,
+      "loss": 1.0904,
+      "step": 1724
+    },
+    {
+      "epoch": 0.3949627933600458,
+      "grad_norm": 1.3219141960144043,
+      "learning_rate": 1.3813980331704437e-05,
+      "loss": 1.0814,
+      "step": 1725
+    },
+    {
+      "epoch": 0.39519175729822553,
+      "grad_norm": 1.0469324588775635,
+      "learning_rate": 1.3807121850724045e-05,
+      "loss": 1.0653,
+      "step": 1726
+    },
+    {
+      "epoch": 0.39542072123640526,
+      "grad_norm": 1.1637883186340332,
+      "learning_rate": 1.3800261274716424e-05,
+      "loss": 1.0838,
+      "step": 1727
+    },
+    {
+      "epoch": 0.395649685174585,
+      "grad_norm": 1.1523021459579468,
+      "learning_rate": 1.3793398607456883e-05,
+      "loss": 1.0906,
+      "step": 1728
+    },
+    {
+      "epoch": 0.3958786491127647,
+      "grad_norm": 1.1947762966156006,
+      "learning_rate": 1.3786533852721891e-05,
+      "loss": 1.0937,
+      "step": 1729
+    },
+    {
+      "epoch": 0.3961076130509445,
+      "grad_norm": 1.565125584602356,
+      "learning_rate": 1.3779667014289067e-05,
+      "loss": 1.0973,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3963365769891242,
+      "grad_norm": 1.6443101167678833,
+      "learning_rate": 1.3772798095937172e-05,
+      "loss": 1.0978,
+      "step": 1731
+    },
+    {
+      "epoch": 0.39656554092730395,
+      "grad_norm": 1.2846298217773438,
+      "learning_rate": 1.3765927101446121e-05,
+      "loss": 1.057,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3967945048654837,
+      "grad_norm": 1.245452880859375,
+      "learning_rate": 1.3759054034596953e-05,
+      "loss": 1.1326,
+      "step": 1733
+    },
+    {
+      "epoch": 0.3970234688036634,
+      "grad_norm": 1.2566457986831665,
+      "learning_rate": 1.375217889917187e-05,
+      "loss": 1.1039,
+      "step": 1734
+    },
+    {
+      "epoch": 0.3972524327418432,
+      "grad_norm": 1.0944688320159912,
+      "learning_rate": 1.3745301698954197e-05,
+      "loss": 1.076,
+      "step": 1735
+    },
+    {
+      "epoch": 0.3974813966800229,
+      "grad_norm": 1.7811460494995117,
+      "learning_rate": 1.3738422437728398e-05,
+      "loss": 1.0675,
+      "step": 1736
+    },
+    {
+      "epoch": 0.39771036061820264,
+      "grad_norm": 2.1263771057128906,
+      "learning_rate": 1.3731541119280073e-05,
+      "loss": 1.0859,
+      "step": 1737
+    },
+    {
+      "epoch": 0.39793932455638237,
+      "grad_norm": 1.2063853740692139,
+      "learning_rate": 1.3724657747395957e-05,
+      "loss": 1.1231,
+      "step": 1738
+    },
+    {
+      "epoch": 0.3981682884945621,
+      "grad_norm": 1.1201328039169312,
+      "learning_rate": 1.3717772325863913e-05,
+      "loss": 1.1532,
+      "step": 1739
+    },
+    {
+      "epoch": 0.3983972524327418,
+      "grad_norm": 1.3484069108963013,
+      "learning_rate": 1.3710884858472926e-05,
+      "loss": 1.0621,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3986262163709216,
+      "grad_norm": 1.4419175386428833,
+      "learning_rate": 1.3703995349013113e-05,
+      "loss": 1.0816,
+      "step": 1741
+    },
+    {
+      "epoch": 0.39885518030910133,
+      "grad_norm": 1.026955246925354,
+      "learning_rate": 1.3697103801275714e-05,
+      "loss": 1.0184,
+      "step": 1742
+    },
+    {
+      "epoch": 0.39908414424728106,
+      "grad_norm": 1.369357943534851,
+      "learning_rate": 1.3690210219053088e-05,
+      "loss": 1.0637,
+      "step": 1743
+    },
+    {
+      "epoch": 0.3993131081854608,
+      "grad_norm": 1.0927389860153198,
+      "learning_rate": 1.3683314606138718e-05,
+      "loss": 1.0612,
+      "step": 1744
+    },
+    {
+      "epoch": 0.3995420721236405,
+      "grad_norm": 1.3103286027908325,
+      "learning_rate": 1.3676416966327201e-05,
+      "loss": 1.0001,
+      "step": 1745
+    },
+    {
+      "epoch": 0.39977103606182024,
+      "grad_norm": 1.1909055709838867,
+      "learning_rate": 1.3669517303414254e-05,
+      "loss": 1.0591,
+      "step": 1746
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6235116720199585,
+      "learning_rate": 1.3662615621196697e-05,
+      "loss": 1.1164,
+      "step": 1747
+    },
+    {
+      "epoch": 0.40022896393817975,
+      "grad_norm": 0.9482841491699219,
+      "learning_rate": 1.3655711923472472e-05,
+      "loss": 1.0182,
+      "step": 1748
+    },
+    {
+      "epoch": 0.4004579278763595,
+      "grad_norm": 1.02751886844635,
+      "learning_rate": 1.3648806214040625e-05,
+      "loss": 1.0817,
+      "step": 1749
+    },
+    {
+      "epoch": 0.4006868918145392,
+      "grad_norm": 1.257691502571106,
+      "learning_rate": 1.3641898496701306e-05,
+      "loss": 1.1007,
+      "step": 1750
+    },
+    {
+      "epoch": 0.40091585575271893,
+      "grad_norm": 1.1339843273162842,
+      "learning_rate": 1.3634988775255785e-05,
+      "loss": 1.0948,
+      "step": 1751
+    },
+    {
+      "epoch": 0.40114481969089866,
+      "grad_norm": 2.271824836730957,
+      "learning_rate": 1.362807705350641e-05,
+      "loss": 1.0535,
+      "step": 1752
+    },
+    {
+      "epoch": 0.40137378362907844,
+      "grad_norm": 1.175581693649292,
+      "learning_rate": 1.3621163335256655e-05,
+      "loss": 1.0902,
+      "step": 1753
+    },
+    {
+      "epoch": 0.40160274756725817,
+      "grad_norm": 0.9900003671646118,
+      "learning_rate": 1.3614247624311076e-05,
+      "loss": 1.0428,
+      "step": 1754
+    },
+    {
+      "epoch": 0.4018317115054379,
+      "grad_norm": 1.2661924362182617,
+      "learning_rate": 1.360732992447533e-05,
+      "loss": 1.0597,
+      "step": 1755
+    },
+    {
+      "epoch": 0.4020606754436176,
+      "grad_norm": 1.1470941305160522,
+      "learning_rate": 1.360041023955617e-05,
+      "loss": 1.1798,
+      "step": 1756
+    },
+    {
+      "epoch": 0.40228963938179735,
+      "grad_norm": 1.8923383951187134,
+      "learning_rate": 1.3593488573361442e-05,
+      "loss": 1.0507,
+      "step": 1757
+    },
+    {
+      "epoch": 0.40251860331997713,
+      "grad_norm": 1.0174728631973267,
+      "learning_rate": 1.358656492970008e-05,
+      "loss": 1.0249,
+      "step": 1758
+    },
+    {
+      "epoch": 0.40274756725815686,
+      "grad_norm": 1.094590663909912,
+      "learning_rate": 1.3579639312382105e-05,
+      "loss": 1.0823,
+      "step": 1759
+    },
+    {
+      "epoch": 0.4029765311963366,
+      "grad_norm": 1.2006431818008423,
+      "learning_rate": 1.357271172521863e-05,
+      "loss": 1.0586,
+      "step": 1760
+    },
+    {
+      "epoch": 0.4032054951345163,
+      "grad_norm": 1.0652323961257935,
+      "learning_rate": 1.3565782172021847e-05,
+      "loss": 1.0759,
+      "step": 1761
+    },
+    {
+      "epoch": 0.40343445907269604,
+      "grad_norm": 1.5044901371002197,
+      "learning_rate": 1.3558850656605028e-05,
+      "loss": 1.1158,
+      "step": 1762
+    },
+    {
+      "epoch": 0.40366342301087577,
+      "grad_norm": 1.201694130897522,
+      "learning_rate": 1.355191718278253e-05,
+      "loss": 1.1144,
+      "step": 1763
+    },
+    {
+      "epoch": 0.40389238694905555,
+      "grad_norm": 1.799466609954834,
+      "learning_rate": 1.3544981754369789e-05,
+      "loss": 1.0865,
+      "step": 1764
+    },
+    {
+      "epoch": 0.4041213508872353,
+      "grad_norm": 1.0255722999572754,
+      "learning_rate": 1.3538044375183308e-05,
+      "loss": 1.071,
+      "step": 1765
+    },
+    {
+      "epoch": 0.404350314825415,
+      "grad_norm": 1.148310661315918,
+      "learning_rate": 1.3531105049040667e-05,
+      "loss": 1.0982,
+      "step": 1766
+    },
+    {
+      "epoch": 0.40457927876359473,
+      "grad_norm": 1.1922695636749268,
+      "learning_rate": 1.3524163779760526e-05,
+      "loss": 1.1228,
+      "step": 1767
+    },
+    {
+      "epoch": 0.40480824270177446,
+      "grad_norm": 1.0259567499160767,
+      "learning_rate": 1.3517220571162603e-05,
+      "loss": 1.0426,
+      "step": 1768
+    },
+    {
+      "epoch": 0.4050372066399542,
+      "grad_norm": 1.3682341575622559,
+      "learning_rate": 1.3510275427067688e-05,
+      "loss": 1.11,
+      "step": 1769
+    },
+    {
+      "epoch": 0.40526617057813397,
+      "grad_norm": 1.16379976272583,
+      "learning_rate": 1.3503328351297635e-05,
+      "loss": 1.0049,
+      "step": 1770
+    },
+    {
+      "epoch": 0.4054951345163137,
+      "grad_norm": 1.128479242324829,
+      "learning_rate": 1.3496379347675364e-05,
+      "loss": 1.0751,
+      "step": 1771
+    },
+    {
+      "epoch": 0.4057240984544934,
+      "grad_norm": 1.2324275970458984,
+      "learning_rate": 1.3489428420024851e-05,
+      "loss": 1.0561,
+      "step": 1772
+    },
+    {
+      "epoch": 0.40595306239267315,
+      "grad_norm": 1.2429150342941284,
+      "learning_rate": 1.3482475572171132e-05,
+      "loss": 1.0709,
+      "step": 1773
+    },
+    {
+      "epoch": 0.4061820263308529,
+      "grad_norm": 1.3965815305709839,
+      "learning_rate": 1.3475520807940303e-05,
+      "loss": 1.1079,
+      "step": 1774
+    },
+    {
+      "epoch": 0.4064109902690326,
+      "grad_norm": 1.0004675388336182,
+      "learning_rate": 1.3468564131159515e-05,
+      "loss": 1.0772,
+      "step": 1775
+    },
+    {
+      "epoch": 0.4066399542072124,
+      "grad_norm": 1.2566231489181519,
+      "learning_rate": 1.3461605545656961e-05,
+      "loss": 1.0527,
+      "step": 1776
+    },
+    {
+      "epoch": 0.4068689181453921,
+      "grad_norm": 1.2383748292922974,
+      "learning_rate": 1.3454645055261903e-05,
+      "loss": 1.0754,
+      "step": 1777
+    },
+    {
+      "epoch": 0.40709788208357184,
+      "grad_norm": 1.3026020526885986,
+      "learning_rate": 1.344768266380463e-05,
+      "loss": 1.0667,
+      "step": 1778
+    },
+    {
+      "epoch": 0.40732684602175157,
+      "grad_norm": 1.141505241394043,
+      "learning_rate": 1.3440718375116497e-05,
+      "loss": 1.184,
+      "step": 1779
+    },
+    {
+      "epoch": 0.4075558099599313,
+      "grad_norm": 1.1129202842712402,
+      "learning_rate": 1.3433752193029888e-05,
+      "loss": 1.0463,
+      "step": 1780
+    },
+    {
+      "epoch": 0.4077847738981111,
+      "grad_norm": 0.9912295341491699,
+      "learning_rate": 1.3426784121378233e-05,
+      "loss": 1.1277,
+      "step": 1781
+    },
+    {
+      "epoch": 0.4080137378362908,
+      "grad_norm": 1.029266119003296,
+      "learning_rate": 1.3419814163996007e-05,
+      "loss": 1.0719,
+      "step": 1782
+    },
+    {
+      "epoch": 0.40824270177447053,
+      "grad_norm": 1.1671353578567505,
+      "learning_rate": 1.341284232471872e-05,
+      "loss": 1.0912,
+      "step": 1783
+    },
+    {
+      "epoch": 0.40847166571265026,
+      "grad_norm": 1.2449313402175903,
+      "learning_rate": 1.3405868607382914e-05,
+      "loss": 1.0755,
+      "step": 1784
+    },
+    {
+      "epoch": 0.40870062965083,
+      "grad_norm": 1.078808069229126,
+      "learning_rate": 1.3398893015826166e-05,
+      "loss": 1.0852,
+      "step": 1785
+    },
+    {
+      "epoch": 0.4089295935890097,
+      "grad_norm": 1.0642139911651611,
+      "learning_rate": 1.3391915553887093e-05,
+      "loss": 1.0864,
+      "step": 1786
+    },
+    {
+      "epoch": 0.4091585575271895,
+      "grad_norm": 1.050003170967102,
+      "learning_rate": 1.3384936225405326e-05,
+      "loss": 1.095,
+      "step": 1787
+    },
+    {
+      "epoch": 0.4093875214653692,
+      "grad_norm": 1.2181106805801392,
+      "learning_rate": 1.3377955034221532e-05,
+      "loss": 1.1276,
+      "step": 1788
+    },
+    {
+      "epoch": 0.40961648540354895,
+      "grad_norm": 1.3170802593231201,
+      "learning_rate": 1.3370971984177406e-05,
+      "loss": 1.0547,
+      "step": 1789
+    },
+    {
+      "epoch": 0.4098454493417287,
+      "grad_norm": 1.1851341724395752,
+      "learning_rate": 1.3363987079115656e-05,
+      "loss": 1.0538,
+      "step": 1790
+    },
+    {
+      "epoch": 0.4100744132799084,
+      "grad_norm": 1.218276023864746,
+      "learning_rate": 1.3357000322880024e-05,
+      "loss": 1.0555,
+      "step": 1791
+    },
+    {
+      "epoch": 0.41030337721808813,
+      "grad_norm": 1.2883524894714355,
+      "learning_rate": 1.3350011719315257e-05,
+      "loss": 1.0051,
+      "step": 1792
+    },
+    {
+      "epoch": 0.4105323411562679,
+      "grad_norm": 1.1477938890457153,
+      "learning_rate": 1.334302127226713e-05,
+      "loss": 1.0466,
+      "step": 1793
+    },
+    {
+      "epoch": 0.41076130509444764,
+      "grad_norm": 1.0580675601959229,
+      "learning_rate": 1.333602898558242e-05,
+      "loss": 1.0661,
+      "step": 1794
+    },
+    {
+      "epoch": 0.41099026903262736,
+      "grad_norm": 1.4260722398757935,
+      "learning_rate": 1.3329034863108932e-05,
+      "loss": 1.0466,
+      "step": 1795
+    },
+    {
+      "epoch": 0.4112192329708071,
+      "grad_norm": 1.1878693103790283,
+      "learning_rate": 1.3322038908695466e-05,
+      "loss": 1.1458,
+      "step": 1796
+    },
+    {
+      "epoch": 0.4114481969089868,
+      "grad_norm": 1.031032919883728,
+      "learning_rate": 1.3315041126191845e-05,
+      "loss": 1.0114,
+      "step": 1797
+    },
+    {
+      "epoch": 0.41167716084716655,
+      "grad_norm": 1.2380294799804688,
+      "learning_rate": 1.3308041519448885e-05,
+      "loss": 1.0351,
+      "step": 1798
+    },
+    {
+      "epoch": 0.41190612478534633,
+      "grad_norm": 1.2293623685836792,
+      "learning_rate": 1.3301040092318409e-05,
+      "loss": 1.0952,
+      "step": 1799
+    },
+    {
+      "epoch": 0.41213508872352606,
+      "grad_norm": 1.1843045949935913,
+      "learning_rate": 1.3294036848653247e-05,
+      "loss": 1.0966,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4123640526617058,
+      "grad_norm": 1.0566411018371582,
+      "learning_rate": 1.3287031792307226e-05,
+      "loss": 1.0596,
+      "step": 1801
+    },
+    {
+      "epoch": 0.4125930165998855,
+      "grad_norm": 1.1404145956039429,
+      "learning_rate": 1.3280024927135165e-05,
+      "loss": 1.0058,
+      "step": 1802
+    },
+    {
+      "epoch": 0.41282198053806524,
+      "grad_norm": 1.210115909576416,
+      "learning_rate": 1.3273016256992888e-05,
+      "loss": 1.0835,
+      "step": 1803
+    },
+    {
+      "epoch": 0.413050944476245,
+      "grad_norm": 1.214924693107605,
+      "learning_rate": 1.3266005785737206e-05,
+      "loss": 1.0879,
+      "step": 1804
+    },
+    {
+      "epoch": 0.41327990841442475,
+      "grad_norm": 1.529047966003418,
+      "learning_rate": 1.3258993517225923e-05,
+      "loss": 1.1117,
+      "step": 1805
+    },
+    {
+      "epoch": 0.4135088723526045,
+      "grad_norm": 1.181199312210083,
+      "learning_rate": 1.3251979455317831e-05,
+      "loss": 1.1248,
+      "step": 1806
+    },
+    {
+      "epoch": 0.4137378362907842,
+      "grad_norm": 1.1865763664245605,
+      "learning_rate": 1.3244963603872707e-05,
+      "loss": 1.1102,
+      "step": 1807
+    },
+    {
+      "epoch": 0.4139668002289639,
+      "grad_norm": 1.2326041460037231,
+      "learning_rate": 1.323794596675132e-05,
+      "loss": 1.0253,
+      "step": 1808
+    },
+    {
+      "epoch": 0.41419576416714365,
+      "grad_norm": 1.0496649742126465,
+      "learning_rate": 1.3230926547815413e-05,
+      "loss": 1.0957,
+      "step": 1809
+    },
+    {
+      "epoch": 0.41442472810532344,
+      "grad_norm": 1.4039772748947144,
+      "learning_rate": 1.3223905350927718e-05,
+      "loss": 1.055,
+      "step": 1810
+    },
+    {
+      "epoch": 0.41465369204350316,
+      "grad_norm": 1.393269658088684,
+      "learning_rate": 1.3216882379951932e-05,
+      "loss": 1.1233,
+      "step": 1811
+    },
+    {
+      "epoch": 0.4148826559816829,
+      "grad_norm": 1.2282973527908325,
+      "learning_rate": 1.3209857638752746e-05,
+      "loss": 1.0997,
+      "step": 1812
+    },
+    {
+      "epoch": 0.4151116199198626,
+      "grad_norm": 1.080151081085205,
+      "learning_rate": 1.3202831131195812e-05,
+      "loss": 1.0984,
+      "step": 1813
+    },
+    {
+      "epoch": 0.41534058385804234,
+      "grad_norm": 1.2916617393493652,
+      "learning_rate": 1.3195802861147759e-05,
+      "loss": 1.1306,
+      "step": 1814
+    },
+    {
+      "epoch": 0.41556954779622207,
+      "grad_norm": 1.297864317893982,
+      "learning_rate": 1.318877283247619e-05,
+      "loss": 1.1278,
+      "step": 1815
+    },
+    {
+      "epoch": 0.41579851173440185,
+      "grad_norm": 1.1242778301239014,
+      "learning_rate": 1.3181741049049659e-05,
+      "loss": 1.0858,
+      "step": 1816
+    },
+    {
+      "epoch": 0.4160274756725816,
+      "grad_norm": 1.1434422731399536,
+      "learning_rate": 1.3174707514737706e-05,
+      "loss": 1.0437,
+      "step": 1817
+    },
+    {
+      "epoch": 0.4162564396107613,
+      "grad_norm": 1.0877010822296143,
+      "learning_rate": 1.3167672233410826e-05,
+      "loss": 1.1197,
+      "step": 1818
+    },
+    {
+      "epoch": 0.41648540354894104,
+      "grad_norm": 1.1932451725006104,
+      "learning_rate": 1.3160635208940473e-05,
+      "loss": 1.1059,
+      "step": 1819
+    },
+    {
+      "epoch": 0.41671436748712076,
+      "grad_norm": 1.4846535921096802,
+      "learning_rate": 1.3153596445199063e-05,
+      "loss": 1.0637,
+      "step": 1820
+    },
+    {
+      "epoch": 0.4169433314253005,
+      "grad_norm": 1.1383739709854126,
+      "learning_rate": 1.3146555946059971e-05,
+      "loss": 1.1125,
+      "step": 1821
+    },
+    {
+      "epoch": 0.41717229536348027,
+      "grad_norm": 1.1333290338516235,
+      "learning_rate": 1.3139513715397521e-05,
+      "loss": 1.0098,
+      "step": 1822
+    },
+    {
+      "epoch": 0.41740125930166,
+      "grad_norm": 0.9758977293968201,
+      "learning_rate": 1.3132469757086997e-05,
+      "loss": 1.0442,
+      "step": 1823
+    },
+    {
+      "epoch": 0.4176302232398397,
+      "grad_norm": 1.221085786819458,
+      "learning_rate": 1.3125424075004624e-05,
+      "loss": 1.0796,
+      "step": 1824
+    },
+    {
+      "epoch": 0.41785918717801945,
+      "grad_norm": 1.830741286277771,
+      "learning_rate": 1.3118376673027588e-05,
+      "loss": 1.078,
+      "step": 1825
+    },
+    {
+      "epoch": 0.4180881511161992,
+      "grad_norm": 1.274835467338562,
+      "learning_rate": 1.3111327555034015e-05,
+      "loss": 1.0829,
+      "step": 1826
+    },
+    {
+      "epoch": 0.41831711505437896,
+      "grad_norm": 1.2003624439239502,
+      "learning_rate": 1.3104276724902975e-05,
+      "loss": 1.0118,
+      "step": 1827
+    },
+    {
+      "epoch": 0.4185460789925587,
+      "grad_norm": 1.195296049118042,
+      "learning_rate": 1.3097224186514476e-05,
+      "loss": 1.0357,
+      "step": 1828
+    },
+    {
+      "epoch": 0.4187750429307384,
+      "grad_norm": 1.2735960483551025,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 1.0895,
+      "step": 1829
+    },
+    {
+      "epoch": 0.41900400686891814,
+      "grad_norm": 1.1189090013504028,
+      "learning_rate": 1.3083114000489863e-05,
+      "loss": 1.0777,
+      "step": 1830
+    },
+    {
+      "epoch": 0.41923297080709787,
+      "grad_norm": 1.1308891773223877,
+      "learning_rate": 1.3076056360618465e-05,
+      "loss": 1.0817,
+      "step": 1831
+    },
+    {
+      "epoch": 0.4194619347452776,
+      "grad_norm": 1.2105430364608765,
+      "learning_rate": 1.3068997028019043e-05,
+      "loss": 1.0939,
+      "step": 1832
+    },
+    {
+      "epoch": 0.4196908986834574,
+      "grad_norm": 1.4238054752349854,
+      "learning_rate": 1.3061936006576284e-05,
+      "loss": 1.0822,
+      "step": 1833
+    },
+    {
+      "epoch": 0.4199198626216371,
+      "grad_norm": 1.037213683128357,
+      "learning_rate": 1.3054873300175814e-05,
+      "loss": 1.0473,
+      "step": 1834
+    },
+    {
+      "epoch": 0.42014882655981683,
+      "grad_norm": 1.4768306016921997,
+      "learning_rate": 1.3047808912704178e-05,
+      "loss": 1.0661,
+      "step": 1835
+    },
+    {
+      "epoch": 0.42037779049799656,
+      "grad_norm": 1.2641279697418213,
+      "learning_rate": 1.304074284804885e-05,
+      "loss": 1.0562,
+      "step": 1836
+    },
+    {
+      "epoch": 0.4206067544361763,
+      "grad_norm": 1.1807743310928345,
+      "learning_rate": 1.3033675110098227e-05,
+      "loss": 1.1504,
+      "step": 1837
+    },
+    {
+      "epoch": 0.420835718374356,
+      "grad_norm": 1.099843978881836,
+      "learning_rate": 1.3026605702741625e-05,
+      "loss": 1.0742,
+      "step": 1838
+    },
+    {
+      "epoch": 0.4210646823125358,
+      "grad_norm": 1.1367111206054688,
+      "learning_rate": 1.3019534629869281e-05,
+      "loss": 1.1427,
+      "step": 1839
+    },
+    {
+      "epoch": 0.4212936462507155,
+      "grad_norm": 1.4131443500518799,
+      "learning_rate": 1.3012461895372343e-05,
+      "loss": 1.0812,
+      "step": 1840
+    },
+    {
+      "epoch": 0.42152261018889525,
+      "grad_norm": 1.1341830492019653,
+      "learning_rate": 1.3005387503142884e-05,
+      "loss": 1.0982,
+      "step": 1841
+    },
+    {
+      "epoch": 0.421751574127075,
+      "grad_norm": 1.1615837812423706,
+      "learning_rate": 1.2998311457073878e-05,
+      "loss": 1.0308,
+      "step": 1842
+    },
+    {
+      "epoch": 0.4219805380652547,
+      "grad_norm": 1.2500286102294922,
+      "learning_rate": 1.2991233761059214e-05,
+      "loss": 1.0998,
+      "step": 1843
+    },
+    {
+      "epoch": 0.42220950200343443,
+      "grad_norm": 1.2073861360549927,
+      "learning_rate": 1.2984154418993691e-05,
+      "loss": 1.0588,
+      "step": 1844
+    },
+    {
+      "epoch": 0.4224384659416142,
+      "grad_norm": 1.1415460109710693,
+      "learning_rate": 1.2977073434773009e-05,
+      "loss": 1.0693,
+      "step": 1845
+    },
+    {
+      "epoch": 0.42266742987979394,
+      "grad_norm": 1.4329516887664795,
+      "learning_rate": 1.2969990812293779e-05,
+      "loss": 1.0938,
+      "step": 1846
+    },
+    {
+      "epoch": 0.42289639381797367,
+      "grad_norm": 1.2857966423034668,
+      "learning_rate": 1.29629065554535e-05,
+      "loss": 1.0196,
+      "step": 1847
+    },
+    {
+      "epoch": 0.4231253577561534,
+      "grad_norm": 1.0637885332107544,
+      "learning_rate": 1.2955820668150587e-05,
+      "loss": 1.0602,
+      "step": 1848
+    },
+    {
+      "epoch": 0.4233543216943331,
+      "grad_norm": 1.3474278450012207,
+      "learning_rate": 1.2948733154284343e-05,
+      "loss": 1.135,
+      "step": 1849
+    },
+    {
+      "epoch": 0.42358328563251285,
+      "grad_norm": 1.1365108489990234,
+      "learning_rate": 1.2941644017754964e-05,
+      "loss": 1.1279,
+      "step": 1850
+    },
+    {
+      "epoch": 0.42381224957069263,
+      "grad_norm": 1.1364717483520508,
+      "learning_rate": 1.2934553262463548e-05,
+      "loss": 1.0466,
+      "step": 1851
+    },
+    {
+      "epoch": 0.42404121350887236,
+      "grad_norm": 1.122428059577942,
+      "learning_rate": 1.2927460892312071e-05,
+      "loss": 1.0606,
+      "step": 1852
+    },
+    {
+      "epoch": 0.4242701774470521,
+      "grad_norm": 1.1230154037475586,
+      "learning_rate": 1.2920366911203414e-05,
+      "loss": 1.0651,
+      "step": 1853
+    },
+    {
+      "epoch": 0.4244991413852318,
+      "grad_norm": 1.150088906288147,
+      "learning_rate": 1.2913271323041328e-05,
+      "loss": 1.0204,
+      "step": 1854
+    },
+    {
+      "epoch": 0.42472810532341154,
+      "grad_norm": 1.1731511354446411,
+      "learning_rate": 1.2906174131730458e-05,
+      "loss": 1.1506,
+      "step": 1855
+    },
+    {
+      "epoch": 0.4249570692615913,
+      "grad_norm": 1.1663103103637695,
+      "learning_rate": 1.2899075341176326e-05,
+      "loss": 1.0903,
+      "step": 1856
+    },
+    {
+      "epoch": 0.42518603319977105,
+      "grad_norm": 1.0978206396102905,
+      "learning_rate": 1.289197495528534e-05,
+      "loss": 1.0315,
+      "step": 1857
+    },
+    {
+      "epoch": 0.4254149971379508,
+      "grad_norm": 1.842656135559082,
+      "learning_rate": 1.2884872977964786e-05,
+      "loss": 1.0962,
+      "step": 1858
+    },
+    {
+      "epoch": 0.4256439610761305,
+      "grad_norm": 1.1557258367538452,
+      "learning_rate": 1.2877769413122816e-05,
+      "loss": 1.1225,
+      "step": 1859
+    },
+    {
+      "epoch": 0.42587292501431023,
+      "grad_norm": 1.0806372165679932,
+      "learning_rate": 1.2870664264668467e-05,
+      "loss": 1.0553,
+      "step": 1860
+    },
+    {
+      "epoch": 0.42610188895248996,
+      "grad_norm": 1.3559293746948242,
+      "learning_rate": 1.2863557536511642e-05,
+      "loss": 1.0912,
+      "step": 1861
+    },
+    {
+      "epoch": 0.42633085289066974,
+      "grad_norm": 1.2274169921875,
+      "learning_rate": 1.285644923256311e-05,
+      "loss": 1.1026,
+      "step": 1862
+    },
+    {
+      "epoch": 0.42655981682884947,
+      "grad_norm": 1.467261791229248,
+      "learning_rate": 1.2849339356734513e-05,
+      "loss": 1.0609,
+      "step": 1863
+    },
+    {
+      "epoch": 0.4267887807670292,
+      "grad_norm": 1.08341646194458,
+      "learning_rate": 1.284222791293836e-05,
+      "loss": 1.0122,
+      "step": 1864
+    },
+    {
+      "epoch": 0.4270177447052089,
+      "grad_norm": 1.18088960647583,
+      "learning_rate": 1.2835114905088013e-05,
+      "loss": 1.0016,
+      "step": 1865
+    },
+    {
+      "epoch": 0.42724670864338865,
+      "grad_norm": 3.6483867168426514,
+      "learning_rate": 1.2828000337097703e-05,
+      "loss": 1.0826,
+      "step": 1866
+    },
+    {
+      "epoch": 0.4274756725815684,
+      "grad_norm": 1.3066470623016357,
+      "learning_rate": 1.2820884212882521e-05,
+      "loss": 1.0831,
+      "step": 1867
+    },
+    {
+      "epoch": 0.42770463651974816,
+      "grad_norm": 1.1415867805480957,
+      "learning_rate": 1.2813766536358406e-05,
+      "loss": 1.0411,
+      "step": 1868
+    },
+    {
+      "epoch": 0.4279336004579279,
+      "grad_norm": 1.0170698165893555,
+      "learning_rate": 1.2806647311442153e-05,
+      "loss": 1.0385,
+      "step": 1869
+    },
+    {
+      "epoch": 0.4281625643961076,
+      "grad_norm": 1.0501190423965454,
+      "learning_rate": 1.2799526542051419e-05,
+      "loss": 1.0581,
+      "step": 1870
+    },
+    {
+      "epoch": 0.42839152833428734,
+      "grad_norm": 1.1775119304656982,
+      "learning_rate": 1.2792404232104699e-05,
+      "loss": 1.1704,
+      "step": 1871
+    },
+    {
+      "epoch": 0.42862049227246707,
+      "grad_norm": 1.1157633066177368,
+      "learning_rate": 1.2785280385521342e-05,
+      "loss": 1.0301,
+      "step": 1872
+    },
+    {
+      "epoch": 0.4288494562106468,
+      "grad_norm": 1.531984567642212,
+      "learning_rate": 1.277815500622154e-05,
+      "loss": 1.0217,
+      "step": 1873
+    },
+    {
+      "epoch": 0.4290784201488266,
+      "grad_norm": 4.276288986206055,
+      "learning_rate": 1.2771028098126333e-05,
+      "loss": 1.0464,
+      "step": 1874
+    },
+    {
+      "epoch": 0.4293073840870063,
+      "grad_norm": 1.2262746095657349,
+      "learning_rate": 1.2763899665157591e-05,
+      "loss": 1.0452,
+      "step": 1875
+    },
+    {
+      "epoch": 0.42953634802518603,
+      "grad_norm": 1.0891977548599243,
+      "learning_rate": 1.275676971123804e-05,
+      "loss": 1.0315,
+      "step": 1876
+    },
+    {
+      "epoch": 0.42976531196336576,
+      "grad_norm": 3.1425435543060303,
+      "learning_rate": 1.2749638240291227e-05,
+      "loss": 1.1258,
+      "step": 1877
+    },
+    {
+      "epoch": 0.4299942759015455,
+      "grad_norm": 1.209585428237915,
+      "learning_rate": 1.2742505256241543e-05,
+      "loss": 1.074,
+      "step": 1878
+    },
+    {
+      "epoch": 0.43022323983972527,
+      "grad_norm": 1.5271508693695068,
+      "learning_rate": 1.2735370763014212e-05,
+      "loss": 1.1113,
+      "step": 1879
+    },
+    {
+      "epoch": 0.430452203777905,
+      "grad_norm": 1.2672470808029175,
+      "learning_rate": 1.2728234764535283e-05,
+      "loss": 1.1049,
+      "step": 1880
+    },
+    {
+      "epoch": 0.4306811677160847,
+      "grad_norm": 1.2539156675338745,
+      "learning_rate": 1.2721097264731634e-05,
+      "loss": 1.0472,
+      "step": 1881
+    },
+    {
+      "epoch": 0.43091013165426445,
+      "grad_norm": 1.542236328125,
+      "learning_rate": 1.2713958267530976e-05,
+      "loss": 1.0408,
+      "step": 1882
+    },
+    {
+      "epoch": 0.4311390955924442,
+      "grad_norm": 1.2664871215820312,
+      "learning_rate": 1.2706817776861838e-05,
+      "loss": 1.1148,
+      "step": 1883
+    },
+    {
+      "epoch": 0.4313680595306239,
+      "grad_norm": 1.2938625812530518,
+      "learning_rate": 1.269967579665357e-05,
+      "loss": 1.0531,
+      "step": 1884
+    },
+    {
+      "epoch": 0.4315970234688037,
+      "grad_norm": 1.3394691944122314,
+      "learning_rate": 1.2692532330836346e-05,
+      "loss": 1.0794,
+      "step": 1885
+    },
+    {
+      "epoch": 0.4318259874069834,
+      "grad_norm": 1.4612385034561157,
+      "learning_rate": 1.2685387383341157e-05,
+      "loss": 1.0415,
+      "step": 1886
+    },
+    {
+      "epoch": 0.43205495134516314,
+      "grad_norm": 1.6666114330291748,
+      "learning_rate": 1.2678240958099801e-05,
+      "loss": 1.0744,
+      "step": 1887
+    },
+    {
+      "epoch": 0.43228391528334287,
+      "grad_norm": 1.2210183143615723,
+      "learning_rate": 1.2671093059044899e-05,
+      "loss": 1.0129,
+      "step": 1888
+    },
+    {
+      "epoch": 0.4325128792215226,
+      "grad_norm": 1.5127156972885132,
+      "learning_rate": 1.2663943690109885e-05,
+      "loss": 1.1413,
+      "step": 1889
+    },
+    {
+      "epoch": 0.4327418431597023,
+      "grad_norm": 1.2559911012649536,
+      "learning_rate": 1.2656792855228993e-05,
+      "loss": 1.1116,
+      "step": 1890
+    },
+    {
+      "epoch": 0.4329708070978821,
+      "grad_norm": 1.1968263387680054,
+      "learning_rate": 1.2649640558337266e-05,
+      "loss": 1.1159,
+      "step": 1891
+    },
+    {
+      "epoch": 0.43319977103606183,
+      "grad_norm": 1.3571308851242065,
+      "learning_rate": 1.2642486803370553e-05,
+      "loss": 1.1034,
+      "step": 1892
+    },
+    {
+      "epoch": 0.43342873497424156,
+      "grad_norm": 1.0277183055877686,
+      "learning_rate": 1.2635331594265508e-05,
+      "loss": 1.0414,
+      "step": 1893
+    },
+    {
+      "epoch": 0.4336576989124213,
+      "grad_norm": 0.9604895114898682,
+      "learning_rate": 1.2628174934959583e-05,
+      "loss": 1.066,
+      "step": 1894
+    },
+    {
+      "epoch": 0.433886662850601,
+      "grad_norm": 0.9915218353271484,
+      "learning_rate": 1.2621016829391022e-05,
+      "loss": 1.0006,
+      "step": 1895
+    },
+    {
+      "epoch": 0.43411562678878074,
+      "grad_norm": 1.2676836252212524,
+      "learning_rate": 1.2613857281498878e-05,
+      "loss": 1.0798,
+      "step": 1896
+    },
+    {
+      "epoch": 0.4343445907269605,
+      "grad_norm": 1.2389227151870728,
+      "learning_rate": 1.2606696295222985e-05,
+      "loss": 1.0692,
+      "step": 1897
+    },
+    {
+      "epoch": 0.43457355466514025,
+      "grad_norm": 1.2961418628692627,
+      "learning_rate": 1.2599533874503978e-05,
+      "loss": 1.0435,
+      "step": 1898
+    },
+    {
+      "epoch": 0.43480251860332,
+      "grad_norm": 1.2095304727554321,
+      "learning_rate": 1.2592370023283268e-05,
+      "loss": 1.0942,
+      "step": 1899
+    },
+    {
+      "epoch": 0.4350314825414997,
+      "grad_norm": 0.9683312773704529,
+      "learning_rate": 1.2585204745503072e-05,
+      "loss": 1.1888,
+      "step": 1900
+    },
+    {
+      "epoch": 0.43526044647967943,
+      "grad_norm": 1.135977029800415,
+      "learning_rate": 1.2578038045106383e-05,
+      "loss": 1.1095,
+      "step": 1901
+    },
+    {
+      "epoch": 0.4354894104178592,
+      "grad_norm": 1.0904693603515625,
+      "learning_rate": 1.2570869926036968e-05,
+      "loss": 1.1369,
+      "step": 1902
+    },
+    {
+      "epoch": 0.43571837435603894,
+      "grad_norm": 1.4972336292266846,
+      "learning_rate": 1.2563700392239387e-05,
+      "loss": 1.0466,
+      "step": 1903
+    },
+    {
+      "epoch": 0.43594733829421867,
+      "grad_norm": 1.0306872129440308,
+      "learning_rate": 1.2556529447658974e-05,
+      "loss": 1.1408,
+      "step": 1904
+    },
+    {
+      "epoch": 0.4361763022323984,
+      "grad_norm": 1.2561451196670532,
+      "learning_rate": 1.2549357096241841e-05,
+      "loss": 1.0328,
+      "step": 1905
+    },
+    {
+      "epoch": 0.4364052661705781,
+      "grad_norm": 1.3269802331924438,
+      "learning_rate": 1.2542183341934873e-05,
+      "loss": 1.0962,
+      "step": 1906
+    },
+    {
+      "epoch": 0.43663423010875785,
+      "grad_norm": 1.1158679723739624,
+      "learning_rate": 1.253500818868572e-05,
+      "loss": 1.0914,
+      "step": 1907
+    },
+    {
+      "epoch": 0.43686319404693763,
+      "grad_norm": 1.4163261651992798,
+      "learning_rate": 1.252783164044282e-05,
+      "loss": 1.0665,
+      "step": 1908
+    },
+    {
+      "epoch": 0.43709215798511736,
+      "grad_norm": 1.1557682752609253,
+      "learning_rate": 1.2520653701155361e-05,
+      "loss": 1.0816,
+      "step": 1909
+    },
+    {
+      "epoch": 0.4373211219232971,
+      "grad_norm": 1.1793196201324463,
+      "learning_rate": 1.25134743747733e-05,
+      "loss": 1.0406,
+      "step": 1910
+    },
+    {
+      "epoch": 0.4375500858614768,
+      "grad_norm": 1.505786657333374,
+      "learning_rate": 1.2506293665247366e-05,
+      "loss": 1.1074,
+      "step": 1911
+    },
+    {
+      "epoch": 0.43777904979965654,
+      "grad_norm": 1.6487133502960205,
+      "learning_rate": 1.2499111576529042e-05,
+      "loss": 1.1171,
+      "step": 1912
+    },
+    {
+      "epoch": 0.43800801373783627,
+      "grad_norm": 1.204660177230835,
+      "learning_rate": 1.2491928112570568e-05,
+      "loss": 0.9898,
+      "step": 1913
+    },
+    {
+      "epoch": 0.43823697767601605,
+      "grad_norm": 1.1147689819335938,
+      "learning_rate": 1.2484743277324945e-05,
+      "loss": 0.9693,
+      "step": 1914
+    },
+    {
+      "epoch": 0.4384659416141958,
+      "grad_norm": 1.136470079421997,
+      "learning_rate": 1.2477557074745932e-05,
+      "loss": 1.0785,
+      "step": 1915
+    },
+    {
+      "epoch": 0.4386949055523755,
+      "grad_norm": 1.2387357950210571,
+      "learning_rate": 1.247036950878803e-05,
+      "loss": 1.1218,
+      "step": 1916
+    },
+    {
+      "epoch": 0.43892386949055523,
+      "grad_norm": 1.831175446510315,
+      "learning_rate": 1.24631805834065e-05,
+      "loss": 1.0891,
+      "step": 1917
+    },
+    {
+      "epoch": 0.43915283342873496,
+      "grad_norm": 1.2181199789047241,
+      "learning_rate": 1.2455990302557346e-05,
+      "loss": 1.0932,
+      "step": 1918
+    },
+    {
+      "epoch": 0.4393817973669147,
+      "grad_norm": 1.262529730796814,
+      "learning_rate": 1.2448798670197318e-05,
+      "loss": 1.0351,
+      "step": 1919
+    },
+    {
+      "epoch": 0.43961076130509447,
+      "grad_norm": 1.10959792137146,
+      "learning_rate": 1.2441605690283915e-05,
+      "loss": 1.1082,
+      "step": 1920
+    },
+    {
+      "epoch": 0.4398397252432742,
+      "grad_norm": 1.1682922840118408,
+      "learning_rate": 1.2434411366775367e-05,
+      "loss": 1.1353,
+      "step": 1921
+    },
+    {
+      "epoch": 0.4400686891814539,
+      "grad_norm": 1.2102086544036865,
+      "learning_rate": 1.242721570363066e-05,
+      "loss": 1.0569,
+      "step": 1922
+    },
+    {
+      "epoch": 0.44029765311963365,
+      "grad_norm": 1.1970107555389404,
+      "learning_rate": 1.2420018704809497e-05,
+      "loss": 1.0502,
+      "step": 1923
+    },
+    {
+      "epoch": 0.4405266170578134,
+      "grad_norm": 1.2726011276245117,
+      "learning_rate": 1.2412820374272332e-05,
+      "loss": 1.0845,
+      "step": 1924
+    },
+    {
+      "epoch": 0.44075558099599316,
+      "grad_norm": 1.1796960830688477,
+      "learning_rate": 1.2405620715980345e-05,
+      "loss": 1.0449,
+      "step": 1925
+    },
+    {
+      "epoch": 0.4409845449341729,
+      "grad_norm": 1.0175650119781494,
+      "learning_rate": 1.2398419733895444e-05,
+      "loss": 1.0563,
+      "step": 1926
+    },
+    {
+      "epoch": 0.4412135088723526,
+      "grad_norm": 2.8849644660949707,
+      "learning_rate": 1.2391217431980273e-05,
+      "loss": 1.1089,
+      "step": 1927
+    },
+    {
+      "epoch": 0.44144247281053234,
+      "grad_norm": 1.4662426710128784,
+      "learning_rate": 1.2384013814198197e-05,
+      "loss": 1.1497,
+      "step": 1928
+    },
+    {
+      "epoch": 0.44167143674871207,
+      "grad_norm": 1.102353811264038,
+      "learning_rate": 1.2376808884513306e-05,
+      "loss": 1.1013,
+      "step": 1929
+    },
+    {
+      "epoch": 0.4419004006868918,
+      "grad_norm": 2.3300585746765137,
+      "learning_rate": 1.2369602646890415e-05,
+      "loss": 1.1192,
+      "step": 1930
+    },
+    {
+      "epoch": 0.4421293646250716,
+      "grad_norm": 1.2267271280288696,
+      "learning_rate": 1.2362395105295054e-05,
+      "loss": 1.069,
+      "step": 1931
+    },
+    {
+      "epoch": 0.4423583285632513,
+      "grad_norm": 1.4739381074905396,
+      "learning_rate": 1.235518626369347e-05,
+      "loss": 1.0883,
+      "step": 1932
+    },
+    {
+      "epoch": 0.44258729250143103,
+      "grad_norm": 1.5360037088394165,
+      "learning_rate": 1.2347976126052631e-05,
+      "loss": 1.0633,
+      "step": 1933
+    },
+    {
+      "epoch": 0.44281625643961076,
+      "grad_norm": 1.3187732696533203,
+      "learning_rate": 1.234076469634022e-05,
+      "loss": 1.0849,
+      "step": 1934
+    },
+    {
+      "epoch": 0.4430452203777905,
+      "grad_norm": 1.0551137924194336,
+      "learning_rate": 1.2333551978524618e-05,
+      "loss": 1.0977,
+      "step": 1935
+    },
+    {
+      "epoch": 0.4432741843159702,
+      "grad_norm": 1.3932082653045654,
+      "learning_rate": 1.2326337976574928e-05,
+      "loss": 1.0336,
+      "step": 1936
+    },
+    {
+      "epoch": 0.44350314825415,
+      "grad_norm": 1.1481599807739258,
+      "learning_rate": 1.2319122694460952e-05,
+      "loss": 1.0941,
+      "step": 1937
+    },
+    {
+      "epoch": 0.4437321121923297,
+      "grad_norm": 1.5443516969680786,
+      "learning_rate": 1.2311906136153202e-05,
+      "loss": 1.0848,
+      "step": 1938
+    },
+    {
+      "epoch": 0.44396107613050945,
+      "grad_norm": 1.0437678098678589,
+      "learning_rate": 1.2304688305622889e-05,
+      "loss": 1.1458,
+      "step": 1939
+    },
+    {
+      "epoch": 0.4441900400686892,
+      "grad_norm": 1.0990891456604004,
+      "learning_rate": 1.2297469206841921e-05,
+      "loss": 1.066,
+      "step": 1940
+    },
+    {
+      "epoch": 0.4444190040068689,
+      "grad_norm": 1.1834927797317505,
+      "learning_rate": 1.2290248843782915e-05,
+      "loss": 1.0384,
+      "step": 1941
+    },
+    {
+      "epoch": 0.44464796794504863,
+      "grad_norm": 1.1914305686950684,
+      "learning_rate": 1.2283027220419169e-05,
+      "loss": 1.0453,
+      "step": 1942
+    },
+    {
+      "epoch": 0.4448769318832284,
+      "grad_norm": 1.607132911682129,
+      "learning_rate": 1.2275804340724684e-05,
+      "loss": 1.033,
+      "step": 1943
+    },
+    {
+      "epoch": 0.44510589582140814,
+      "grad_norm": 1.1510486602783203,
+      "learning_rate": 1.2268580208674151e-05,
+      "loss": 1.0786,
+      "step": 1944
+    },
+    {
+      "epoch": 0.44533485975958786,
+      "grad_norm": 1.0266963243484497,
+      "learning_rate": 1.2261354828242948e-05,
+      "loss": 1.0426,
+      "step": 1945
+    },
+    {
+      "epoch": 0.4455638236977676,
+      "grad_norm": 1.0529321432113647,
+      "learning_rate": 1.2254128203407146e-05,
+      "loss": 1.0221,
+      "step": 1946
+    },
+    {
+      "epoch": 0.4457927876359473,
+      "grad_norm": 1.0336111783981323,
+      "learning_rate": 1.2246900338143485e-05,
+      "loss": 1.0285,
+      "step": 1947
+    },
+    {
+      "epoch": 0.4460217515741271,
+      "grad_norm": 1.5184695720672607,
+      "learning_rate": 1.2239671236429413e-05,
+      "loss": 1.0077,
+      "step": 1948
+    },
+    {
+      "epoch": 0.44625071551230683,
+      "grad_norm": 1.1687562465667725,
+      "learning_rate": 1.2232440902243037e-05,
+      "loss": 1.0827,
+      "step": 1949
+    },
+    {
+      "epoch": 0.44647967945048656,
+      "grad_norm": 1.2201441526412964,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 1.0207,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4467086433886663,
+      "grad_norm": 1.1205480098724365,
+      "learning_rate": 1.2217976552369213e-05,
+      "loss": 1.0461,
+      "step": 1951
+    },
+    {
+      "epoch": 0.446937607326846,
+      "grad_norm": 2.142512083053589,
+      "learning_rate": 1.2210742544641377e-05,
+      "loss": 1.1122,
+      "step": 1952
+    },
+    {
+      "epoch": 0.44716657126502574,
+      "grad_norm": 1.682520866394043,
+      "learning_rate": 1.2203507320360458e-05,
+      "loss": 1.0742,
+      "step": 1953
+    },
+    {
+      "epoch": 0.4473955352032055,
+      "grad_norm": 1.2594259977340698,
+      "learning_rate": 1.2196270883507927e-05,
+      "loss": 1.0582,
+      "step": 1954
+    },
+    {
+      "epoch": 0.44762449914138525,
+      "grad_norm": 1.0667154788970947,
+      "learning_rate": 1.218903323806595e-05,
+      "loss": 1.0335,
+      "step": 1955
+    },
+    {
+      "epoch": 0.447853463079565,
+      "grad_norm": 1.5725624561309814,
+      "learning_rate": 1.2181794388017332e-05,
+      "loss": 1.0779,
+      "step": 1956
+    },
+    {
+      "epoch": 0.4480824270177447,
+      "grad_norm": 1.0927304029464722,
+      "learning_rate": 1.2174554337345555e-05,
+      "loss": 1.0951,
+      "step": 1957
+    },
+    {
+      "epoch": 0.4483113909559244,
+      "grad_norm": 1.3435699939727783,
+      "learning_rate": 1.2167313090034756e-05,
+      "loss": 1.0554,
+      "step": 1958
+    },
+    {
+      "epoch": 0.44854035489410415,
+      "grad_norm": 1.0855443477630615,
+      "learning_rate": 1.2160070650069735e-05,
+      "loss": 1.1022,
+      "step": 1959
+    },
+    {
+      "epoch": 0.44876931883228394,
+      "grad_norm": 1.2444900274276733,
+      "learning_rate": 1.2152827021435946e-05,
+      "loss": 1.0268,
+      "step": 1960
+    },
+    {
+      "epoch": 0.44899828277046366,
+      "grad_norm": 1.4539769887924194,
+      "learning_rate": 1.2145582208119497e-05,
+      "loss": 1.1199,
+      "step": 1961
+    },
+    {
+      "epoch": 0.4492272467086434,
+      "grad_norm": 1.8856399059295654,
+      "learning_rate": 1.2138336214107148e-05,
+      "loss": 1.0528,
+      "step": 1962
+    },
+    {
+      "epoch": 0.4494562106468231,
+      "grad_norm": 1.1333612203598022,
+      "learning_rate": 1.2131089043386305e-05,
+      "loss": 1.0599,
+      "step": 1963
+    },
+    {
+      "epoch": 0.44968517458500284,
+      "grad_norm": 1.1188855171203613,
+      "learning_rate": 1.212384069994503e-05,
+      "loss": 1.0356,
+      "step": 1964
+    },
+    {
+      "epoch": 0.44991413852318257,
+      "grad_norm": 1.4274768829345703,
+      "learning_rate": 1.2116591187772026e-05,
+      "loss": 1.0944,
+      "step": 1965
+    },
+    {
+      "epoch": 0.45014310246136235,
+      "grad_norm": 1.637333631515503,
+      "learning_rate": 1.2109340510856633e-05,
+      "loss": 1.098,
+      "step": 1966
+    },
+    {
+      "epoch": 0.4503720663995421,
+      "grad_norm": 1.5847759246826172,
+      "learning_rate": 1.2102088673188845e-05,
+      "loss": 1.0764,
+      "step": 1967
+    },
+    {
+      "epoch": 0.4506010303377218,
+      "grad_norm": 1.5479519367218018,
+      "learning_rate": 1.2094835678759282e-05,
+      "loss": 1.118,
+      "step": 1968
+    },
+    {
+      "epoch": 0.45082999427590154,
+      "grad_norm": 1.1684271097183228,
+      "learning_rate": 1.2087581531559208e-05,
+      "loss": 1.0399,
+      "step": 1969
+    },
+    {
+      "epoch": 0.45105895821408126,
+      "grad_norm": 1.4510235786437988,
+      "learning_rate": 1.2080326235580521e-05,
+      "loss": 1.0827,
+      "step": 1970
+    },
+    {
+      "epoch": 0.45128792215226105,
+      "grad_norm": 1.1053940057754517,
+      "learning_rate": 1.2073069794815748e-05,
+      "loss": 1.0701,
+      "step": 1971
+    },
+    {
+      "epoch": 0.45151688609044077,
+      "grad_norm": 1.083375334739685,
+      "learning_rate": 1.2065812213258051e-05,
+      "loss": 1.0956,
+      "step": 1972
+    },
+    {
+      "epoch": 0.4517458500286205,
+      "grad_norm": 1.1666209697723389,
+      "learning_rate": 1.2058553494901203e-05,
+      "loss": 1.0781,
+      "step": 1973
+    },
+    {
+      "epoch": 0.4519748139668002,
+      "grad_norm": 1.2016594409942627,
+      "learning_rate": 1.2051293643739634e-05,
+      "loss": 1.0406,
+      "step": 1974
+    },
+    {
+      "epoch": 0.45220377790497995,
+      "grad_norm": 1.4277703762054443,
+      "learning_rate": 1.204403266376837e-05,
+      "loss": 1.0543,
+      "step": 1975
+    },
+    {
+      "epoch": 0.4524327418431597,
+      "grad_norm": 1.206367015838623,
+      "learning_rate": 1.2036770558983067e-05,
+      "loss": 1.0423,
+      "step": 1976
+    },
+    {
+      "epoch": 0.45266170578133946,
+      "grad_norm": 1.0165696144104004,
+      "learning_rate": 1.202950733338e-05,
+      "loss": 0.9945,
+      "step": 1977
+    },
+    {
+      "epoch": 0.4528906697195192,
+      "grad_norm": 0.9572392702102661,
+      "learning_rate": 1.2022242990956064e-05,
+      "loss": 1.07,
+      "step": 1978
+    },
+    {
+      "epoch": 0.4531196336576989,
+      "grad_norm": 2.232534885406494,
+      "learning_rate": 1.2014977535708767e-05,
+      "loss": 1.1249,
+      "step": 1979
+    },
+    {
+      "epoch": 0.45334859759587864,
+      "grad_norm": 1.0949493646621704,
+      "learning_rate": 1.2007710971636221e-05,
+      "loss": 0.9862,
+      "step": 1980
+    },
+    {
+      "epoch": 0.45357756153405837,
+      "grad_norm": 1.493613600730896,
+      "learning_rate": 1.2000443302737162e-05,
+      "loss": 1.0943,
+      "step": 1981
+    },
+    {
+      "epoch": 0.4538065254722381,
+      "grad_norm": 1.0247379541397095,
+      "learning_rate": 1.1993174533010928e-05,
+      "loss": 1.0921,
+      "step": 1982
+    },
+    {
+      "epoch": 0.4540354894104179,
+      "grad_norm": 1.1227061748504639,
+      "learning_rate": 1.1985904666457455e-05,
+      "loss": 1.0995,
+      "step": 1983
+    },
+    {
+      "epoch": 0.4542644533485976,
+      "grad_norm": 1.3855504989624023,
+      "learning_rate": 1.1978633707077296e-05,
+      "loss": 1.0868,
+      "step": 1984
+    },
+    {
+      "epoch": 0.45449341728677733,
+      "grad_norm": 1.058700680732727,
+      "learning_rate": 1.1971361658871597e-05,
+      "loss": 1.0761,
+      "step": 1985
+    },
+    {
+      "epoch": 0.45472238122495706,
+      "grad_norm": 1.199990153312683,
+      "learning_rate": 1.1964088525842108e-05,
+      "loss": 1.0743,
+      "step": 1986
+    },
+    {
+      "epoch": 0.4549513451631368,
+      "grad_norm": 1.074346661567688,
+      "learning_rate": 1.1956814311991164e-05,
+      "loss": 1.0791,
+      "step": 1987
+    },
+    {
+      "epoch": 0.4551803091013165,
+      "grad_norm": 1.4295293092727661,
+      "learning_rate": 1.1949539021321713e-05,
+      "loss": 1.0344,
+      "step": 1988
+    },
+    {
+      "epoch": 0.4554092730394963,
+      "grad_norm": 1.2947413921356201,
+      "learning_rate": 1.1942262657837285e-05,
+      "loss": 1.0676,
+      "step": 1989
+    },
+    {
+      "epoch": 0.455638236977676,
+      "grad_norm": 1.270447015762329,
+      "learning_rate": 1.1934985225541998e-05,
+      "loss": 1.0626,
+      "step": 1990
+    },
+    {
+      "epoch": 0.45586720091585575,
+      "grad_norm": 1.2195264101028442,
+      "learning_rate": 1.1927706728440565e-05,
+      "loss": 1.0508,
+      "step": 1991
+    },
+    {
+      "epoch": 0.4560961648540355,
+      "grad_norm": 1.2984168529510498,
+      "learning_rate": 1.192042717053828e-05,
+      "loss": 1.0517,
+      "step": 1992
+    },
+    {
+      "epoch": 0.4563251287922152,
+      "grad_norm": 1.2059777975082397,
+      "learning_rate": 1.1913146555841027e-05,
+      "loss": 1.0639,
+      "step": 1993
+    },
+    {
+      "epoch": 0.456554092730395,
+      "grad_norm": 1.2496461868286133,
+      "learning_rate": 1.1905864888355264e-05,
+      "loss": 1.1124,
+      "step": 1994
+    },
+    {
+      "epoch": 0.4567830566685747,
+      "grad_norm": 1.3210448026657104,
+      "learning_rate": 1.1898582172088027e-05,
+      "loss": 1.1523,
+      "step": 1995
+    },
+    {
+      "epoch": 0.45701202060675444,
+      "grad_norm": 1.288273811340332,
+      "learning_rate": 1.1891298411046943e-05,
+      "loss": 1.0392,
+      "step": 1996
+    },
+    {
+      "epoch": 0.45724098454493417,
+      "grad_norm": 1.6366710662841797,
+      "learning_rate": 1.18840136092402e-05,
+      "loss": 1.1138,
+      "step": 1997
+    },
+    {
+      "epoch": 0.4574699484831139,
+      "grad_norm": 0.9694288372993469,
+      "learning_rate": 1.1876727770676562e-05,
+      "loss": 1.0257,
+      "step": 1998
+    },
+    {
+      "epoch": 0.4576989124212936,
+      "grad_norm": 1.0713609457015991,
+      "learning_rate": 1.1869440899365365e-05,
+      "loss": 1.0785,
+      "step": 1999
+    },
+    {
+      "epoch": 0.4579278763594734,
+      "grad_norm": 0.9268698692321777,
+      "learning_rate": 1.1862152999316517e-05,
+      "loss": 1.0387,
+      "step": 2000
+    },
+    {
+      "epoch": 0.45815684029765313,
+      "grad_norm": 1.066428780555725,
+      "learning_rate": 1.1854864074540484e-05,
+      "loss": 1.0959,
+      "step": 2001
+    },
+    {
+      "epoch": 0.45838580423583286,
+      "grad_norm": 1.2952028512954712,
+      "learning_rate": 1.18475741290483e-05,
+      "loss": 1.0896,
+      "step": 2002
+    },
+    {
+      "epoch": 0.4586147681740126,
+      "grad_norm": 1.1627318859100342,
+      "learning_rate": 1.184028316685157e-05,
+      "loss": 1.0602,
+      "step": 2003
+    },
+    {
+      "epoch": 0.4588437321121923,
+      "grad_norm": 1.2692722082138062,
+      "learning_rate": 1.1832991191962435e-05,
+      "loss": 1.1016,
+      "step": 2004
+    },
+    {
+      "epoch": 0.45907269605037204,
+      "grad_norm": 1.1311182975769043,
+      "learning_rate": 1.182569820839362e-05,
+      "loss": 1.0275,
+      "step": 2005
+    },
+    {
+      "epoch": 0.4593016599885518,
+      "grad_norm": 1.1217960119247437,
+      "learning_rate": 1.1818404220158382e-05,
+      "loss": 1.1007,
+      "step": 2006
+    },
+    {
+      "epoch": 0.45953062392673155,
+      "grad_norm": 1.1474188566207886,
+      "learning_rate": 1.181110923127055e-05,
+      "loss": 1.0613,
+      "step": 2007
+    },
+    {
+      "epoch": 0.4597595878649113,
+      "grad_norm": 1.1906862258911133,
+      "learning_rate": 1.1803813245744495e-05,
+      "loss": 1.1043,
+      "step": 2008
+    },
+    {
+      "epoch": 0.459988551803091,
+      "grad_norm": 1.391391634941101,
+      "learning_rate": 1.179651626759513e-05,
+      "loss": 1.1015,
+      "step": 2009
+    },
+    {
+      "epoch": 0.46021751574127073,
+      "grad_norm": 1.3041538000106812,
+      "learning_rate": 1.1789218300837929e-05,
+      "loss": 1.0601,
+      "step": 2010
+    },
+    {
+      "epoch": 0.46044647967945046,
+      "grad_norm": 1.484880805015564,
+      "learning_rate": 1.1781919349488894e-05,
+      "loss": 1.0303,
+      "step": 2011
+    },
+    {
+      "epoch": 0.46067544361763024,
+      "grad_norm": 1.271309733390808,
+      "learning_rate": 1.177461941756458e-05,
+      "loss": 1.0293,
+      "step": 2012
+    },
+    {
+      "epoch": 0.46090440755580997,
+      "grad_norm": 1.0927857160568237,
+      "learning_rate": 1.1767318509082083e-05,
+      "loss": 1.0673,
+      "step": 2013
+    },
+    {
+      "epoch": 0.4611333714939897,
+      "grad_norm": 1.1591389179229736,
+      "learning_rate": 1.1760016628059026e-05,
+      "loss": 1.1406,
+      "step": 2014
+    },
+    {
+      "epoch": 0.4613623354321694,
+      "grad_norm": 1.0383858680725098,
+      "learning_rate": 1.1752713778513576e-05,
+      "loss": 1.0687,
+      "step": 2015
+    },
+    {
+      "epoch": 0.46159129937034915,
+      "grad_norm": 1.1415190696716309,
+      "learning_rate": 1.1745409964464425e-05,
+      "loss": 1.094,
+      "step": 2016
+    },
+    {
+      "epoch": 0.46182026330852893,
+      "grad_norm": 1.0855273008346558,
+      "learning_rate": 1.1738105189930807e-05,
+      "loss": 1.1396,
+      "step": 2017
+    },
+    {
+      "epoch": 0.46204922724670866,
+      "grad_norm": 1.1211069822311401,
+      "learning_rate": 1.1730799458932473e-05,
+      "loss": 1.0809,
+      "step": 2018
+    },
+    {
+      "epoch": 0.4622781911848884,
+      "grad_norm": 1.233792781829834,
+      "learning_rate": 1.1723492775489709e-05,
+      "loss": 1.0569,
+      "step": 2019
+    },
+    {
+      "epoch": 0.4625071551230681,
+      "grad_norm": 1.1317369937896729,
+      "learning_rate": 1.1716185143623322e-05,
+      "loss": 1.0652,
+      "step": 2020
+    },
+    {
+      "epoch": 0.46273611906124784,
+      "grad_norm": 1.083250880241394,
+      "learning_rate": 1.1708876567354635e-05,
+      "loss": 1.0556,
+      "step": 2021
+    },
+    {
+      "epoch": 0.46296508299942757,
+      "grad_norm": 1.1673812866210938,
+      "learning_rate": 1.1701567050705504e-05,
+      "loss": 1.0599,
+      "step": 2022
+    },
+    {
+      "epoch": 0.46319404693760735,
+      "grad_norm": 1.2285277843475342,
+      "learning_rate": 1.1694256597698288e-05,
+      "loss": 1.0872,
+      "step": 2023
+    },
+    {
+      "epoch": 0.4634230108757871,
+      "grad_norm": 1.4135609865188599,
+      "learning_rate": 1.168694521235587e-05,
+      "loss": 1.0408,
+      "step": 2024
+    },
+    {
+      "epoch": 0.4636519748139668,
+      "grad_norm": 1.0683159828186035,
+      "learning_rate": 1.1679632898701649e-05,
+      "loss": 1.0787,
+      "step": 2025
+    },
+    {
+      "epoch": 0.46388093875214653,
+      "grad_norm": 1.5515210628509521,
+      "learning_rate": 1.1672319660759523e-05,
+      "loss": 1.0343,
+      "step": 2026
+    },
+    {
+      "epoch": 0.46410990269032626,
+      "grad_norm": 1.0499638319015503,
+      "learning_rate": 1.1665005502553912e-05,
+      "loss": 1.0641,
+      "step": 2027
+    },
+    {
+      "epoch": 0.464338866628506,
+      "grad_norm": 1.0508882999420166,
+      "learning_rate": 1.165769042810973e-05,
+      "loss": 1.0326,
+      "step": 2028
+    },
+    {
+      "epoch": 0.46456783056668577,
+      "grad_norm": 1.1158897876739502,
+      "learning_rate": 1.1650374441452403e-05,
+      "loss": 1.0767,
+      "step": 2029
+    },
+    {
+      "epoch": 0.4647967945048655,
+      "grad_norm": 1.1368999481201172,
+      "learning_rate": 1.1643057546607858e-05,
+      "loss": 1.0745,
+      "step": 2030
+    },
+    {
+      "epoch": 0.4650257584430452,
+      "grad_norm": 1.5753467082977295,
+      "learning_rate": 1.1635739747602522e-05,
+      "loss": 1.0078,
+      "step": 2031
+    },
+    {
+      "epoch": 0.46525472238122495,
+      "grad_norm": 1.261248230934143,
+      "learning_rate": 1.1628421048463315e-05,
+      "loss": 0.9994,
+      "step": 2032
+    },
+    {
+      "epoch": 0.4654836863194047,
+      "grad_norm": 1.1466697454452515,
+      "learning_rate": 1.1621101453217656e-05,
+      "loss": 1.0402,
+      "step": 2033
+    },
+    {
+      "epoch": 0.4657126502575844,
+      "grad_norm": 1.1229275465011597,
+      "learning_rate": 1.1613780965893465e-05,
+      "loss": 1.0619,
+      "step": 2034
+    },
+    {
+      "epoch": 0.4659416141957642,
+      "grad_norm": 1.118383526802063,
+      "learning_rate": 1.1606459590519132e-05,
+      "loss": 1.0626,
+      "step": 2035
+    },
+    {
+      "epoch": 0.4661705781339439,
+      "grad_norm": 1.139142632484436,
+      "learning_rate": 1.1599137331123558e-05,
+      "loss": 1.0958,
+      "step": 2036
+    },
+    {
+      "epoch": 0.46639954207212364,
+      "grad_norm": 1.29563570022583,
+      "learning_rate": 1.1591814191736117e-05,
+      "loss": 1.042,
+      "step": 2037
+    },
+    {
+      "epoch": 0.46662850601030337,
+      "grad_norm": 1.1359105110168457,
+      "learning_rate": 1.1584490176386671e-05,
+      "loss": 1.081,
+      "step": 2038
+    },
+    {
+      "epoch": 0.4668574699484831,
+      "grad_norm": 1.0732991695404053,
+      "learning_rate": 1.1577165289105565e-05,
+      "loss": 0.9929,
+      "step": 2039
+    },
+    {
+      "epoch": 0.4670864338866629,
+      "grad_norm": 1.1904350519180298,
+      "learning_rate": 1.1569839533923626e-05,
+      "loss": 1.0244,
+      "step": 2040
+    },
+    {
+      "epoch": 0.4673153978248426,
+      "grad_norm": 1.1562821865081787,
+      "learning_rate": 1.1562512914872152e-05,
+      "loss": 1.1229,
+      "step": 2041
+    },
+    {
+      "epoch": 0.46754436176302233,
+      "grad_norm": 1.1359809637069702,
+      "learning_rate": 1.1555185435982923e-05,
+      "loss": 1.0961,
+      "step": 2042
+    },
+    {
+      "epoch": 0.46777332570120206,
+      "grad_norm": 1.1743630170822144,
+      "learning_rate": 1.1547857101288185e-05,
+      "loss": 1.1129,
+      "step": 2043
+    },
+    {
+      "epoch": 0.4680022896393818,
+      "grad_norm": 1.1456128358840942,
+      "learning_rate": 1.154052791482066e-05,
+      "loss": 1.0655,
+      "step": 2044
+    },
+    {
+      "epoch": 0.4682312535775615,
+      "grad_norm": 1.235987663269043,
+      "learning_rate": 1.153319788061354e-05,
+      "loss": 0.9827,
+      "step": 2045
+    },
+    {
+      "epoch": 0.4684602175157413,
+      "grad_norm": 1.3438743352890015,
+      "learning_rate": 1.1525867002700484e-05,
+      "loss": 1.0906,
+      "step": 2046
+    },
+    {
+      "epoch": 0.468689181453921,
+      "grad_norm": 1.1594905853271484,
+      "learning_rate": 1.1518535285115604e-05,
+      "loss": 1.0886,
+      "step": 2047
+    },
+    {
+      "epoch": 0.46891814539210075,
+      "grad_norm": 1.3096404075622559,
+      "learning_rate": 1.1511202731893493e-05,
+      "loss": 1.108,
+      "step": 2048
+    },
+    {
+      "epoch": 0.4691471093302805,
+      "grad_norm": 1.1685811281204224,
+      "learning_rate": 1.1503869347069186e-05,
+      "loss": 1.0289,
+      "step": 2049
+    },
+    {
+      "epoch": 0.4693760732684602,
+      "grad_norm": 1.7220717668533325,
+      "learning_rate": 1.1496535134678186e-05,
+      "loss": 1.0488,
+      "step": 2050
+    },
+    {
+      "epoch": 0.46960503720663993,
+      "grad_norm": 1.417108416557312,
+      "learning_rate": 1.1489200098756447e-05,
+      "loss": 1.0304,
+      "step": 2051
+    },
+    {
+      "epoch": 0.4698340011448197,
+      "grad_norm": 1.0246206521987915,
+      "learning_rate": 1.1481864243340381e-05,
+      "loss": 1.0297,
+      "step": 2052
+    },
+    {
+      "epoch": 0.47006296508299944,
+      "grad_norm": 1.2509313821792603,
+      "learning_rate": 1.1474527572466847e-05,
+      "loss": 1.1336,
+      "step": 2053
+    },
+    {
+      "epoch": 0.47029192902117917,
+      "grad_norm": 1.0300573110580444,
+      "learning_rate": 1.1467190090173147e-05,
+      "loss": 1.0812,
+      "step": 2054
+    },
+    {
+      "epoch": 0.4705208929593589,
+      "grad_norm": 1.067595362663269,
+      "learning_rate": 1.1459851800497048e-05,
+      "loss": 1.1246,
+      "step": 2055
+    },
+    {
+      "epoch": 0.4707498568975386,
+      "grad_norm": 1.1726340055465698,
+      "learning_rate": 1.145251270747674e-05,
+      "loss": 1.0196,
+      "step": 2056
+    },
+    {
+      "epoch": 0.47097882083571835,
+      "grad_norm": 1.0472357273101807,
+      "learning_rate": 1.1445172815150864e-05,
+      "loss": 1.0323,
+      "step": 2057
+    },
+    {
+      "epoch": 0.47120778477389813,
+      "grad_norm": 2.4607527256011963,
+      "learning_rate": 1.1437832127558508e-05,
+      "loss": 1.073,
+      "step": 2058
+    },
+    {
+      "epoch": 0.47143674871207786,
+      "grad_norm": 0.9941400289535522,
+      "learning_rate": 1.1430490648739185e-05,
+      "loss": 1.0418,
+      "step": 2059
+    },
+    {
+      "epoch": 0.4716657126502576,
+      "grad_norm": 1.4227428436279297,
+      "learning_rate": 1.1423148382732854e-05,
+      "loss": 1.1564,
+      "step": 2060
+    },
+    {
+      "epoch": 0.4718946765884373,
+      "grad_norm": 1.135663390159607,
+      "learning_rate": 1.1415805333579895e-05,
+      "loss": 1.0869,
+      "step": 2061
+    },
+    {
+      "epoch": 0.47212364052661704,
+      "grad_norm": 1.1211762428283691,
+      "learning_rate": 1.1408461505321136e-05,
+      "loss": 1.0782,
+      "step": 2062
+    },
+    {
+      "epoch": 0.4723526044647968,
+      "grad_norm": 1.7367572784423828,
+      "learning_rate": 1.1401116901997815e-05,
+      "loss": 1.0039,
+      "step": 2063
+    },
+    {
+      "epoch": 0.47258156840297655,
+      "grad_norm": 1.2560068368911743,
+      "learning_rate": 1.1393771527651614e-05,
+      "loss": 1.0838,
+      "step": 2064
+    },
+    {
+      "epoch": 0.4728105323411563,
+      "grad_norm": 1.2819138765335083,
+      "learning_rate": 1.1386425386324622e-05,
+      "loss": 0.999,
+      "step": 2065
+    },
+    {
+      "epoch": 0.473039496279336,
+      "grad_norm": 1.2946174144744873,
+      "learning_rate": 1.1379078482059367e-05,
+      "loss": 1.0082,
+      "step": 2066
+    },
+    {
+      "epoch": 0.47326846021751573,
+      "grad_norm": 1.1478831768035889,
+      "learning_rate": 1.1371730818898785e-05,
+      "loss": 1.114,
+      "step": 2067
+    },
+    {
+      "epoch": 0.47349742415569546,
+      "grad_norm": 1.1708440780639648,
+      "learning_rate": 1.1364382400886233e-05,
+      "loss": 1.0527,
+      "step": 2068
+    },
+    {
+      "epoch": 0.47372638809387524,
+      "grad_norm": 1.5635682344436646,
+      "learning_rate": 1.1357033232065484e-05,
+      "loss": 1.0343,
+      "step": 2069
+    },
+    {
+      "epoch": 0.47395535203205497,
+      "grad_norm": 1.4606600999832153,
+      "learning_rate": 1.134968331648073e-05,
+      "loss": 1.0216,
+      "step": 2070
+    },
+    {
+      "epoch": 0.4741843159702347,
+      "grad_norm": 1.3325625658035278,
+      "learning_rate": 1.1342332658176556e-05,
+      "loss": 1.0921,
+      "step": 2071
+    },
+    {
+      "epoch": 0.4744132799084144,
+      "grad_norm": 1.0679417848587036,
+      "learning_rate": 1.1334981261197977e-05,
+      "loss": 1.1267,
+      "step": 2072
+    },
+    {
+      "epoch": 0.47464224384659415,
+      "grad_norm": 1.4468427896499634,
+      "learning_rate": 1.1327629129590402e-05,
+      "loss": 1.0597,
+      "step": 2073
+    },
+    {
+      "epoch": 0.4748712077847739,
+      "grad_norm": 1.1622424125671387,
+      "learning_rate": 1.132027626739965e-05,
+      "loss": 1.0962,
+      "step": 2074
+    },
+    {
+      "epoch": 0.47510017172295366,
+      "grad_norm": 1.3740299940109253,
+      "learning_rate": 1.1312922678671935e-05,
+      "loss": 1.0671,
+      "step": 2075
+    },
+    {
+      "epoch": 0.4753291356611334,
+      "grad_norm": 1.3251038789749146,
+      "learning_rate": 1.1305568367453877e-05,
+      "loss": 1.0959,
+      "step": 2076
+    },
+    {
+      "epoch": 0.4755580995993131,
+      "grad_norm": 1.244946002960205,
+      "learning_rate": 1.1298213337792494e-05,
+      "loss": 1.0705,
+      "step": 2077
+    },
+    {
+      "epoch": 0.47578706353749284,
+      "grad_norm": 1.1645523309707642,
+      "learning_rate": 1.1290857593735196e-05,
+      "loss": 1.097,
+      "step": 2078
+    },
+    {
+      "epoch": 0.47601602747567257,
+      "grad_norm": 1.0410329103469849,
+      "learning_rate": 1.1283501139329787e-05,
+      "loss": 1.0381,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4762449914138523,
+      "grad_norm": 1.137269377708435,
+      "learning_rate": 1.1276143978624457e-05,
+      "loss": 1.0412,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4764739553520321,
+      "grad_norm": 1.1206343173980713,
+      "learning_rate": 1.1268786115667798e-05,
+      "loss": 1.0825,
+      "step": 2081
+    },
+    {
+      "epoch": 0.4767029192902118,
+      "grad_norm": 1.176672339439392,
+      "learning_rate": 1.126142755450878e-05,
+      "loss": 1.0497,
+      "step": 2082
+    },
+    {
+      "epoch": 0.47693188322839153,
+      "grad_norm": 1.2728939056396484,
+      "learning_rate": 1.125406829919675e-05,
+      "loss": 1.0416,
+      "step": 2083
+    },
+    {
+      "epoch": 0.47716084716657126,
+      "grad_norm": 1.3541243076324463,
+      "learning_rate": 1.1246708353781453e-05,
+      "loss": 1.0776,
+      "step": 2084
+    },
+    {
+      "epoch": 0.477389811104751,
+      "grad_norm": 1.3378325700759888,
+      "learning_rate": 1.1239347722312997e-05,
+      "loss": 1.0741,
+      "step": 2085
+    },
+    {
+      "epoch": 0.47761877504293077,
+      "grad_norm": 1.2494585514068604,
+      "learning_rate": 1.1231986408841882e-05,
+      "loss": 1.072,
+      "step": 2086
+    },
+    {
+      "epoch": 0.4778477389811105,
+      "grad_norm": 1.4159393310546875,
+      "learning_rate": 1.1224624417418976e-05,
+      "loss": 1.1049,
+      "step": 2087
+    },
+    {
+      "epoch": 0.4780767029192902,
+      "grad_norm": 1.2381250858306885,
+      "learning_rate": 1.1217261752095518e-05,
+      "loss": 1.0467,
+      "step": 2088
+    },
+    {
+      "epoch": 0.47830566685746995,
+      "grad_norm": 1.128997802734375,
+      "learning_rate": 1.1209898416923129e-05,
+      "loss": 1.0505,
+      "step": 2089
+    },
+    {
+      "epoch": 0.4785346307956497,
+      "grad_norm": 1.1534404754638672,
+      "learning_rate": 1.120253441595378e-05,
+      "loss": 0.9887,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4787635947338294,
+      "grad_norm": 1.047783613204956,
+      "learning_rate": 1.1195169753239825e-05,
+      "loss": 1.0748,
+      "step": 2091
+    },
+    {
+      "epoch": 0.4789925586720092,
+      "grad_norm": 1.0474052429199219,
+      "learning_rate": 1.1187804432833976e-05,
+      "loss": 1.0393,
+      "step": 2092
+    },
+    {
+      "epoch": 0.4792215226101889,
+      "grad_norm": 1.389976978302002,
+      "learning_rate": 1.1180438458789305e-05,
+      "loss": 1.0733,
+      "step": 2093
+    },
+    {
+      "epoch": 0.47945048654836864,
+      "grad_norm": 1.16421377658844,
+      "learning_rate": 1.1173071835159248e-05,
+      "loss": 1.0366,
+      "step": 2094
+    },
+    {
+      "epoch": 0.47967945048654836,
+      "grad_norm": 1.5794599056243896,
+      "learning_rate": 1.1165704565997593e-05,
+      "loss": 1.0504,
+      "step": 2095
+    },
+    {
+      "epoch": 0.4799084144247281,
+      "grad_norm": 1.2971510887145996,
+      "learning_rate": 1.115833665535849e-05,
+      "loss": 1.0369,
+      "step": 2096
+    },
+    {
+      "epoch": 0.4801373783629078,
+      "grad_norm": 1.2570523023605347,
+      "learning_rate": 1.1150968107296438e-05,
+      "loss": 1.046,
+      "step": 2097
+    },
+    {
+      "epoch": 0.4803663423010876,
+      "grad_norm": 1.7457414865493774,
+      "learning_rate": 1.1143598925866286e-05,
+      "loss": 1.0883,
+      "step": 2098
+    },
+    {
+      "epoch": 0.48059530623926733,
+      "grad_norm": 1.236221432685852,
+      "learning_rate": 1.1136229115123232e-05,
+      "loss": 1.0506,
+      "step": 2099
+    },
+    {
+      "epoch": 0.48082427017744706,
+      "grad_norm": 1.4669848680496216,
+      "learning_rate": 1.1128858679122822e-05,
+      "loss": 1.0371,
+      "step": 2100
+    },
+    {
+      "epoch": 0.4810532341156268,
+      "grad_norm": 1.0517163276672363,
+      "learning_rate": 1.1121487621920948e-05,
+      "loss": 1.0899,
+      "step": 2101
+    },
+    {
+      "epoch": 0.4812821980538065,
+      "grad_norm": 1.3375537395477295,
+      "learning_rate": 1.1114115947573834e-05,
+      "loss": 1.0915,
+      "step": 2102
+    },
+    {
+      "epoch": 0.48151116199198624,
+      "grad_norm": 1.2560817003250122,
+      "learning_rate": 1.1106743660138057e-05,
+      "loss": 1.0988,
+      "step": 2103
+    },
+    {
+      "epoch": 0.481740125930166,
+      "grad_norm": 1.2373180389404297,
+      "learning_rate": 1.1099370763670523e-05,
+      "loss": 0.9765,
+      "step": 2104
+    },
+    {
+      "epoch": 0.48196908986834575,
+      "grad_norm": 1.2725884914398193,
+      "learning_rate": 1.1091997262228473e-05,
+      "loss": 1.0098,
+      "step": 2105
+    },
+    {
+      "epoch": 0.4821980538065255,
+      "grad_norm": 1.7785913944244385,
+      "learning_rate": 1.1084623159869488e-05,
+      "loss": 1.0742,
+      "step": 2106
+    },
+    {
+      "epoch": 0.4824270177447052,
+      "grad_norm": 1.4811958074569702,
+      "learning_rate": 1.1077248460651468e-05,
+      "loss": 1.0411,
+      "step": 2107
+    },
+    {
+      "epoch": 0.4826559816828849,
+      "grad_norm": 1.2428529262542725,
+      "learning_rate": 1.1069873168632657e-05,
+      "loss": 1.1439,
+      "step": 2108
+    },
+    {
+      "epoch": 0.4828849456210647,
+      "grad_norm": 1.0219333171844482,
+      "learning_rate": 1.1062497287871606e-05,
+      "loss": 1.0631,
+      "step": 2109
+    },
+    {
+      "epoch": 0.48311390955924444,
+      "grad_norm": 1.0447986125946045,
+      "learning_rate": 1.1055120822427208e-05,
+      "loss": 1.0794,
+      "step": 2110
+    },
+    {
+      "epoch": 0.48334287349742416,
+      "grad_norm": 1.1982734203338623,
+      "learning_rate": 1.1047743776358666e-05,
+      "loss": 1.1079,
+      "step": 2111
+    },
+    {
+      "epoch": 0.4835718374356039,
+      "grad_norm": 1.3162384033203125,
+      "learning_rate": 1.104036615372551e-05,
+      "loss": 1.1181,
+      "step": 2112
+    },
+    {
+      "epoch": 0.4838008013737836,
+      "grad_norm": 1.0572696924209595,
+      "learning_rate": 1.103298795858758e-05,
+      "loss": 1.0138,
+      "step": 2113
+    },
+    {
+      "epoch": 0.48402976531196334,
+      "grad_norm": 1.1792926788330078,
+      "learning_rate": 1.1025609195005038e-05,
+      "loss": 1.0455,
+      "step": 2114
+    },
+    {
+      "epoch": 0.4842587292501431,
+      "grad_norm": 2.336987018585205,
+      "learning_rate": 1.1018229867038358e-05,
+      "loss": 1.1053,
+      "step": 2115
+    },
+    {
+      "epoch": 0.48448769318832285,
+      "grad_norm": 1.0690840482711792,
+      "learning_rate": 1.1010849978748314e-05,
+      "loss": 1.0366,
+      "step": 2116
+    },
+    {
+      "epoch": 0.4847166571265026,
+      "grad_norm": 1.3050801753997803,
+      "learning_rate": 1.1003469534196003e-05,
+      "loss": 1.0121,
+      "step": 2117
+    },
+    {
+      "epoch": 0.4849456210646823,
+      "grad_norm": 1.054900884628296,
+      "learning_rate": 1.099608853744282e-05,
+      "loss": 1.0639,
+      "step": 2118
+    },
+    {
+      "epoch": 0.48517458500286204,
+      "grad_norm": 1.2549351453781128,
+      "learning_rate": 1.0988706992550467e-05,
+      "loss": 1.0572,
+      "step": 2119
+    },
+    {
+      "epoch": 0.48540354894104176,
+      "grad_norm": 1.4606761932373047,
+      "learning_rate": 1.0981324903580945e-05,
+      "loss": 1.0981,
+      "step": 2120
+    },
+    {
+      "epoch": 0.48563251287922155,
+      "grad_norm": 1.153498649597168,
+      "learning_rate": 1.0973942274596557e-05,
+      "loss": 1.0869,
+      "step": 2121
+    },
+    {
+      "epoch": 0.48586147681740127,
+      "grad_norm": 1.411656141281128,
+      "learning_rate": 1.09665591096599e-05,
+      "loss": 1.0528,
+      "step": 2122
+    },
+    {
+      "epoch": 0.486090440755581,
+      "grad_norm": 1.040209174156189,
+      "learning_rate": 1.0959175412833869e-05,
+      "loss": 1.0702,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4863194046937607,
+      "grad_norm": 0.9708045721054077,
+      "learning_rate": 1.0951791188181648e-05,
+      "loss": 1.1015,
+      "step": 2124
+    },
+    {
+      "epoch": 0.48654836863194045,
+      "grad_norm": 1.1500178575515747,
+      "learning_rate": 1.0944406439766719e-05,
+      "loss": 1.0767,
+      "step": 2125
+    },
+    {
+      "epoch": 0.4867773325701202,
+      "grad_norm": 3.3348257541656494,
+      "learning_rate": 1.0937021171652842e-05,
+      "loss": 1.146,
+      "step": 2126
+    },
+    {
+      "epoch": 0.48700629650829996,
+      "grad_norm": 1.2559481859207153,
+      "learning_rate": 1.0929635387904075e-05,
+      "loss": 1.0668,
+      "step": 2127
+    },
+    {
+      "epoch": 0.4872352604464797,
+      "grad_norm": 2.0190439224243164,
+      "learning_rate": 1.092224909258474e-05,
+      "loss": 1.0488,
+      "step": 2128
+    },
+    {
+      "epoch": 0.4874642243846594,
+      "grad_norm": 1.2366149425506592,
+      "learning_rate": 1.0914862289759467e-05,
+      "loss": 1.0883,
+      "step": 2129
+    },
+    {
+      "epoch": 0.48769318832283914,
+      "grad_norm": 1.206701397895813,
+      "learning_rate": 1.0907474983493144e-05,
+      "loss": 1.0949,
+      "step": 2130
+    },
+    {
+      "epoch": 0.48792215226101887,
+      "grad_norm": 1.0292967557907104,
+      "learning_rate": 1.0900087177850946e-05,
+      "loss": 1.0862,
+      "step": 2131
+    },
+    {
+      "epoch": 0.48815111619919865,
+      "grad_norm": 1.2705435752868652,
+      "learning_rate": 1.0892698876898322e-05,
+      "loss": 1.0859,
+      "step": 2132
+    },
+    {
+      "epoch": 0.4883800801373784,
+      "grad_norm": 1.5828287601470947,
+      "learning_rate": 1.0885310084700989e-05,
+      "loss": 1.0841,
+      "step": 2133
+    },
+    {
+      "epoch": 0.4886090440755581,
+      "grad_norm": 1.2173969745635986,
+      "learning_rate": 1.087792080532494e-05,
+      "loss": 1.0741,
+      "step": 2134
+    },
+    {
+      "epoch": 0.48883800801373783,
+      "grad_norm": 1.191078543663025,
+      "learning_rate": 1.087053104283643e-05,
+      "loss": 1.0468,
+      "step": 2135
+    },
+    {
+      "epoch": 0.48906697195191756,
+      "grad_norm": 1.1816260814666748,
+      "learning_rate": 1.0863140801301988e-05,
+      "loss": 1.0349,
+      "step": 2136
+    },
+    {
+      "epoch": 0.4892959358900973,
+      "grad_norm": 1.1615114212036133,
+      "learning_rate": 1.08557500847884e-05,
+      "loss": 1.0635,
+      "step": 2137
+    },
+    {
+      "epoch": 0.48952489982827707,
+      "grad_norm": 1.2740709781646729,
+      "learning_rate": 1.0848358897362713e-05,
+      "loss": 0.9996,
+      "step": 2138
+    },
+    {
+      "epoch": 0.4897538637664568,
+      "grad_norm": 1.2105414867401123,
+      "learning_rate": 1.0840967243092237e-05,
+      "loss": 1.1426,
+      "step": 2139
+    },
+    {
+      "epoch": 0.4899828277046365,
+      "grad_norm": 1.668809175491333,
+      "learning_rate": 1.083357512604454e-05,
+      "loss": 1.0329,
+      "step": 2140
+    },
+    {
+      "epoch": 0.49021179164281625,
+      "grad_norm": 1.245213508605957,
+      "learning_rate": 1.082618255028744e-05,
+      "loss": 1.0716,
+      "step": 2141
+    },
+    {
+      "epoch": 0.490440755580996,
+      "grad_norm": 1.3269602060317993,
+      "learning_rate": 1.0818789519889006e-05,
+      "loss": 1.0416,
+      "step": 2142
+    },
+    {
+      "epoch": 0.4906697195191757,
+      "grad_norm": 1.0687472820281982,
+      "learning_rate": 1.0811396038917568e-05,
+      "loss": 1.0823,
+      "step": 2143
+    },
+    {
+      "epoch": 0.4908986834573555,
+      "grad_norm": 1.3689026832580566,
+      "learning_rate": 1.080400211144169e-05,
+      "loss": 1.0581,
+      "step": 2144
+    },
+    {
+      "epoch": 0.4911276473955352,
+      "grad_norm": 1.1712907552719116,
+      "learning_rate": 1.0796607741530191e-05,
+      "loss": 1.0461,
+      "step": 2145
+    },
+    {
+      "epoch": 0.49135661133371494,
+      "grad_norm": 2.368657350540161,
+      "learning_rate": 1.078921293325213e-05,
+      "loss": 1.1125,
+      "step": 2146
+    },
+    {
+      "epoch": 0.49158557527189467,
+      "grad_norm": 1.0939091444015503,
+      "learning_rate": 1.078181769067681e-05,
+      "loss": 1.1154,
+      "step": 2147
+    },
+    {
+      "epoch": 0.4918145392100744,
+      "grad_norm": 1.1558587551116943,
+      "learning_rate": 1.077442201787377e-05,
+      "loss": 1.049,
+      "step": 2148
+    },
+    {
+      "epoch": 0.4920435031482541,
+      "grad_norm": 1.6609954833984375,
+      "learning_rate": 1.0767025918912785e-05,
+      "loss": 1.0641,
+      "step": 2149
+    },
+    {
+      "epoch": 0.4922724670864339,
+      "grad_norm": 1.109490990638733,
+      "learning_rate": 1.075962939786387e-05,
+      "loss": 1.0407,
+      "step": 2150
+    },
+    {
+      "epoch": 0.49250143102461363,
+      "grad_norm": 1.1975339651107788,
+      "learning_rate": 1.0752232458797262e-05,
+      "loss": 1.1138,
+      "step": 2151
+    },
+    {
+      "epoch": 0.49273039496279336,
+      "grad_norm": 2.040076971054077,
+      "learning_rate": 1.0744835105783443e-05,
+      "loss": 1.0248,
+      "step": 2152
+    },
+    {
+      "epoch": 0.4929593589009731,
+      "grad_norm": 1.3864123821258545,
+      "learning_rate": 1.0737437342893107e-05,
+      "loss": 1.089,
+      "step": 2153
+    },
+    {
+      "epoch": 0.4931883228391528,
+      "grad_norm": 1.306090235710144,
+      "learning_rate": 1.0730039174197185e-05,
+      "loss": 1.0915,
+      "step": 2154
+    },
+    {
+      "epoch": 0.4934172867773326,
+      "grad_norm": 1.2275519371032715,
+      "learning_rate": 1.0722640603766825e-05,
+      "loss": 1.0502,
+      "step": 2155
+    },
+    {
+      "epoch": 0.4936462507155123,
+      "grad_norm": 1.4218353033065796,
+      "learning_rate": 1.0715241635673401e-05,
+      "loss": 1.1642,
+      "step": 2156
+    },
+    {
+      "epoch": 0.49387521465369205,
+      "grad_norm": 1.3406203985214233,
+      "learning_rate": 1.0707842273988498e-05,
+      "loss": 1.1157,
+      "step": 2157
+    },
+    {
+      "epoch": 0.4941041785918718,
+      "grad_norm": 1.8006969690322876,
+      "learning_rate": 1.070044252278393e-05,
+      "loss": 1.0964,
+      "step": 2158
+    },
+    {
+      "epoch": 0.4943331425300515,
+      "grad_norm": 1.0731760263442993,
+      "learning_rate": 1.0693042386131713e-05,
+      "loss": 1.0523,
+      "step": 2159
+    },
+    {
+      "epoch": 0.49456210646823123,
+      "grad_norm": 1.6657426357269287,
+      "learning_rate": 1.0685641868104085e-05,
+      "loss": 1.0756,
+      "step": 2160
+    },
+    {
+      "epoch": 0.494791070406411,
+      "grad_norm": 1.2945160865783691,
+      "learning_rate": 1.0678240972773479e-05,
+      "loss": 1.0694,
+      "step": 2161
+    },
+    {
+      "epoch": 0.49502003434459074,
+      "grad_norm": 1.229271650314331,
+      "learning_rate": 1.0670839704212555e-05,
+      "loss": 1.084,
+      "step": 2162
+    },
+    {
+      "epoch": 0.49524899828277047,
+      "grad_norm": 1.339217185974121,
+      "learning_rate": 1.0663438066494168e-05,
+      "loss": 1.1103,
+      "step": 2163
+    },
+    {
+      "epoch": 0.4954779622209502,
+      "grad_norm": 2.237689971923828,
+      "learning_rate": 1.0656036063691373e-05,
+      "loss": 1.0424,
+      "step": 2164
+    },
+    {
+      "epoch": 0.4957069261591299,
+      "grad_norm": 1.0860393047332764,
+      "learning_rate": 1.064863369987743e-05,
+      "loss": 1.0646,
+      "step": 2165
+    },
+    {
+      "epoch": 0.49593589009730965,
+      "grad_norm": 1.190271258354187,
+      "learning_rate": 1.0641230979125804e-05,
+      "loss": 1.0778,
+      "step": 2166
+    },
+    {
+      "epoch": 0.49616485403548943,
+      "grad_norm": 1.1193439960479736,
+      "learning_rate": 1.0633827905510146e-05,
+      "loss": 1.0855,
+      "step": 2167
+    },
+    {
+      "epoch": 0.49639381797366916,
+      "grad_norm": 1.3502209186553955,
+      "learning_rate": 1.0626424483104302e-05,
+      "loss": 1.0405,
+      "step": 2168
+    },
+    {
+      "epoch": 0.4966227819118489,
+      "grad_norm": 1.0666968822479248,
+      "learning_rate": 1.061902071598232e-05,
+      "loss": 1.0146,
+      "step": 2169
+    },
+    {
+      "epoch": 0.4968517458500286,
+      "grad_norm": 1.291035771369934,
+      "learning_rate": 1.0611616608218429e-05,
+      "loss": 1.058,
+      "step": 2170
+    },
+    {
+      "epoch": 0.49708070978820834,
+      "grad_norm": 1.1638990640640259,
+      "learning_rate": 1.0604212163887044e-05,
+      "loss": 1.0792,
+      "step": 2171
+    },
+    {
+      "epoch": 0.49730967372638807,
+      "grad_norm": 1.4782902002334595,
+      "learning_rate": 1.0596807387062772e-05,
+      "loss": 1.0652,
+      "step": 2172
+    },
+    {
+      "epoch": 0.49753863766456785,
+      "grad_norm": 1.5571351051330566,
+      "learning_rate": 1.0589402281820397e-05,
+      "loss": 1.1116,
+      "step": 2173
+    },
+    {
+      "epoch": 0.4977676016027476,
+      "grad_norm": 1.5529664754867554,
+      "learning_rate": 1.058199685223489e-05,
+      "loss": 1.0767,
+      "step": 2174
+    },
+    {
+      "epoch": 0.4979965655409273,
+      "grad_norm": 1.0995012521743774,
+      "learning_rate": 1.0574591102381395e-05,
+      "loss": 1.0161,
+      "step": 2175
+    },
+    {
+      "epoch": 0.49822552947910703,
+      "grad_norm": 1.1900322437286377,
+      "learning_rate": 1.056718503633523e-05,
+      "loss": 0.9899,
+      "step": 2176
+    },
+    {
+      "epoch": 0.49845449341728676,
+      "grad_norm": 1.1697393655776978,
+      "learning_rate": 1.05597786581719e-05,
+      "loss": 1.0988,
+      "step": 2177
+    },
+    {
+      "epoch": 0.49868345735546654,
+      "grad_norm": 1.67704176902771,
+      "learning_rate": 1.0552371971967064e-05,
+      "loss": 1.0819,
+      "step": 2178
+    },
+    {
+      "epoch": 0.49891242129364627,
+      "grad_norm": 1.1780916452407837,
+      "learning_rate": 1.0544964981796563e-05,
+      "loss": 1.0438,
+      "step": 2179
+    },
+    {
+      "epoch": 0.499141385231826,
+      "grad_norm": 1.136317491531372,
+      "learning_rate": 1.0537557691736402e-05,
+      "loss": 1.0522,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4993703491700057,
+      "grad_norm": 1.273093581199646,
+      "learning_rate": 1.0530150105862748e-05,
+      "loss": 1.0864,
+      "step": 2181
+    },
+    {
+      "epoch": 0.49959931310818545,
+      "grad_norm": 1.1058335304260254,
+      "learning_rate": 1.052274222825194e-05,
+      "loss": 1.0664,
+      "step": 2182
+    },
+    {
+      "epoch": 0.4998282770463652,
+      "grad_norm": 1.14626944065094,
+      "learning_rate": 1.0515334062980463e-05,
+      "loss": 1.0342,
+      "step": 2183
+    },
+    {
+      "epoch": 0.500057240984545,
+      "grad_norm": 1.7187449932098389,
+      "learning_rate": 1.0507925614124977e-05,
+      "loss": 1.0868,
+      "step": 2184
+    },
+    {
+      "epoch": 0.5002862049227247,
+      "grad_norm": 1.194671630859375,
+      "learning_rate": 1.0500516885762278e-05,
+      "loss": 1.1066,
+      "step": 2185
+    },
+    {
+      "epoch": 0.5005151688609044,
+      "grad_norm": 1.1433876752853394,
+      "learning_rate": 1.0493107881969335e-05,
+      "loss": 1.0736,
+      "step": 2186
+    },
+    {
+      "epoch": 0.5007441327990841,
+      "grad_norm": 1.1823675632476807,
+      "learning_rate": 1.0485698606823258e-05,
+      "loss": 1.0713,
+      "step": 2187
+    },
+    {
+      "epoch": 0.5009730967372639,
+      "grad_norm": 1.105976939201355,
+      "learning_rate": 1.047828906440131e-05,
+      "loss": 1.06,
+      "step": 2188
+    },
+    {
+      "epoch": 0.5012020606754436,
+      "grad_norm": 1.8626004457473755,
+      "learning_rate": 1.0470879258780904e-05,
+      "loss": 1.0403,
+      "step": 2189
+    },
+    {
+      "epoch": 0.5014310246136233,
+      "grad_norm": 0.9595720171928406,
+      "learning_rate": 1.0463469194039584e-05,
+      "loss": 1.0738,
+      "step": 2190
+    },
+    {
+      "epoch": 0.501659988551803,
+      "grad_norm": 1.0298045873641968,
+      "learning_rate": 1.0456058874255055e-05,
+      "loss": 1.0297,
+      "step": 2191
+    },
+    {
+      "epoch": 0.5018889524899828,
+      "grad_norm": 1.3229492902755737,
+      "learning_rate": 1.044864830350515e-05,
+      "loss": 1.0169,
+      "step": 2192
+    },
+    {
+      "epoch": 0.5021179164281626,
+      "grad_norm": 1.3693296909332275,
+      "learning_rate": 1.0441237485867845e-05,
+      "loss": 1.0533,
+      "step": 2193
+    },
+    {
+      "epoch": 0.5023468803663423,
+      "grad_norm": 1.587257981300354,
+      "learning_rate": 1.0433826425421252e-05,
+      "loss": 1.0314,
+      "step": 2194
+    },
+    {
+      "epoch": 0.5025758443045221,
+      "grad_norm": 1.2772449254989624,
+      "learning_rate": 1.0426415126243615e-05,
+      "loss": 1.1135,
+      "step": 2195
+    },
+    {
+      "epoch": 0.5028048082427018,
+      "grad_norm": 1.3835172653198242,
+      "learning_rate": 1.0419003592413308e-05,
+      "loss": 1.0587,
+      "step": 2196
+    },
+    {
+      "epoch": 0.5030337721808815,
+      "grad_norm": 1.1872873306274414,
+      "learning_rate": 1.0411591828008839e-05,
+      "loss": 1.092,
+      "step": 2197
+    },
+    {
+      "epoch": 0.5032627361190612,
+      "grad_norm": 1.5710023641586304,
+      "learning_rate": 1.0404179837108833e-05,
+      "loss": 1.0778,
+      "step": 2198
+    },
+    {
+      "epoch": 0.503491700057241,
+      "grad_norm": 1.2492884397506714,
+      "learning_rate": 1.0396767623792054e-05,
+      "loss": 1.0466,
+      "step": 2199
+    },
+    {
+      "epoch": 0.5037206639954207,
+      "grad_norm": 1.0260891914367676,
+      "learning_rate": 1.0389355192137379e-05,
+      "loss": 1.1209,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5039496279336004,
+      "grad_norm": 1.9055163860321045,
+      "learning_rate": 1.0381942546223805e-05,
+      "loss": 1.0782,
+      "step": 2201
+    },
+    {
+      "epoch": 0.5041785918717802,
+      "grad_norm": 1.1072444915771484,
+      "learning_rate": 1.0374529690130448e-05,
+      "loss": 1.1312,
+      "step": 2202
+    },
+    {
+      "epoch": 0.5044075558099599,
+      "grad_norm": 1.2415649890899658,
+      "learning_rate": 1.0367116627936549e-05,
+      "loss": 1.0609,
+      "step": 2203
+    },
+    {
+      "epoch": 0.5046365197481397,
+      "grad_norm": 1.1317917108535767,
+      "learning_rate": 1.0359703363721443e-05,
+      "loss": 1.0392,
+      "step": 2204
+    },
+    {
+      "epoch": 0.5048654836863194,
+      "grad_norm": 1.2405906915664673,
+      "learning_rate": 1.0352289901564592e-05,
+      "loss": 1.0132,
+      "step": 2205
+    },
+    {
+      "epoch": 0.5050944476244992,
+      "grad_norm": 1.1400092840194702,
+      "learning_rate": 1.034487624554556e-05,
+      "loss": 1.026,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5053234115626789,
+      "grad_norm": 1.1699836254119873,
+      "learning_rate": 1.0337462399744025e-05,
+      "loss": 1.077,
+      "step": 2207
+    },
+    {
+      "epoch": 0.5055523755008586,
+      "grad_norm": 1.1276884078979492,
+      "learning_rate": 1.033004836823976e-05,
+      "loss": 1.1002,
+      "step": 2208
+    },
+    {
+      "epoch": 0.5057813394390384,
+      "grad_norm": 1.2098451852798462,
+      "learning_rate": 1.032263415511264e-05,
+      "loss": 1.006,
+      "step": 2209
+    },
+    {
+      "epoch": 0.5060103033772181,
+      "grad_norm": 1.1760305166244507,
+      "learning_rate": 1.0315219764442657e-05,
+      "loss": 1.0193,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5062392673153978,
+      "grad_norm": 1.0920625925064087,
+      "learning_rate": 1.0307805200309877e-05,
+      "loss": 1.0141,
+      "step": 2211
+    },
+    {
+      "epoch": 0.5064682312535775,
+      "grad_norm": 1.2013732194900513,
+      "learning_rate": 1.0300390466794477e-05,
+      "loss": 1.0508,
+      "step": 2212
+    },
+    {
+      "epoch": 0.5066971951917573,
+      "grad_norm": 1.3105764389038086,
+      "learning_rate": 1.0292975567976719e-05,
+      "loss": 1.0373,
+      "step": 2213
+    },
+    {
+      "epoch": 0.506926159129937,
+      "grad_norm": 1.2512943744659424,
+      "learning_rate": 1.0285560507936962e-05,
+      "loss": 1.0673,
+      "step": 2214
+    },
+    {
+      "epoch": 0.5071551230681167,
+      "grad_norm": 1.5378952026367188,
+      "learning_rate": 1.0278145290755657e-05,
+      "loss": 1.0772,
+      "step": 2215
+    },
+    {
+      "epoch": 0.5073840870062966,
+      "grad_norm": 1.0876376628875732,
+      "learning_rate": 1.0270729920513326e-05,
+      "loss": 1.0603,
+      "step": 2216
+    },
+    {
+      "epoch": 0.5076130509444763,
+      "grad_norm": 1.0932906866073608,
+      "learning_rate": 1.0263314401290589e-05,
+      "loss": 1.0401,
+      "step": 2217
+    },
+    {
+      "epoch": 0.507842014882656,
+      "grad_norm": 1.3708820343017578,
+      "learning_rate": 1.0255898737168147e-05,
+      "loss": 1.073,
+      "step": 2218
+    },
+    {
+      "epoch": 0.5080709788208357,
+      "grad_norm": 1.1124688386917114,
+      "learning_rate": 1.0248482932226775e-05,
+      "loss": 1.0008,
+      "step": 2219
+    },
+    {
+      "epoch": 0.5082999427590155,
+      "grad_norm": 2.375359058380127,
+      "learning_rate": 1.0241066990547328e-05,
+      "loss": 1.0608,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5085289066971952,
+      "grad_norm": 1.3283129930496216,
+      "learning_rate": 1.0233650916210736e-05,
+      "loss": 1.0725,
+      "step": 2221
+    },
+    {
+      "epoch": 0.5087578706353749,
+      "grad_norm": 1.1090928316116333,
+      "learning_rate": 1.0226234713298007e-05,
+      "loss": 1.0388,
+      "step": 2222
+    },
+    {
+      "epoch": 0.5089868345735546,
+      "grad_norm": 1.2009447813034058,
+      "learning_rate": 1.021881838589021e-05,
+      "loss": 1.0969,
+      "step": 2223
+    },
+    {
+      "epoch": 0.5092157985117344,
+      "grad_norm": 1.144832968711853,
+      "learning_rate": 1.021140193806849e-05,
+      "loss": 1.0553,
+      "step": 2224
+    },
+    {
+      "epoch": 0.5094447624499141,
+      "grad_norm": 1.072203516960144,
+      "learning_rate": 1.0203985373914056e-05,
+      "loss": 1.0466,
+      "step": 2225
+    },
+    {
+      "epoch": 0.5096737263880938,
+      "grad_norm": 1.0581434965133667,
+      "learning_rate": 1.019656869750818e-05,
+      "loss": 1.0244,
+      "step": 2226
+    },
+    {
+      "epoch": 0.5099026903262737,
+      "grad_norm": 1.2225676774978638,
+      "learning_rate": 1.0189151912932199e-05,
+      "loss": 1.0231,
+      "step": 2227
+    },
+    {
+      "epoch": 0.5101316542644534,
+      "grad_norm": 1.196749210357666,
+      "learning_rate": 1.0181735024267504e-05,
+      "loss": 1.128,
+      "step": 2228
+    },
+    {
+      "epoch": 0.5103606182026331,
+      "grad_norm": 1.0913293361663818,
+      "learning_rate": 1.0174318035595551e-05,
+      "loss": 1.1325,
+      "step": 2229
+    },
+    {
+      "epoch": 0.5105895821408128,
+      "grad_norm": 1.296054720878601,
+      "learning_rate": 1.0166900950997845e-05,
+      "loss": 1.0734,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5108185460789926,
+      "grad_norm": 1.2411744594573975,
+      "learning_rate": 1.0159483774555945e-05,
+      "loss": 1.033,
+      "step": 2231
+    },
+    {
+      "epoch": 0.5110475100171723,
+      "grad_norm": 1.2513041496276855,
+      "learning_rate": 1.015206651035146e-05,
+      "loss": 1.1326,
+      "step": 2232
+    },
+    {
+      "epoch": 0.511276473955352,
+      "grad_norm": 1.182816743850708,
+      "learning_rate": 1.0144649162466047e-05,
+      "loss": 1.0539,
+      "step": 2233
+    },
+    {
+      "epoch": 0.5115054378935318,
+      "grad_norm": 1.555910587310791,
+      "learning_rate": 1.0137231734981417e-05,
+      "loss": 1.0438,
+      "step": 2234
+    },
+    {
+      "epoch": 0.5117344018317115,
+      "grad_norm": 1.1628386974334717,
+      "learning_rate": 1.012981423197931e-05,
+      "loss": 1.1111,
+      "step": 2235
+    },
+    {
+      "epoch": 0.5119633657698912,
+      "grad_norm": 1.134666919708252,
+      "learning_rate": 1.0122396657541522e-05,
+      "loss": 1.0665,
+      "step": 2236
+    },
+    {
+      "epoch": 0.5121923297080709,
+      "grad_norm": 1.291595458984375,
+      "learning_rate": 1.011497901574988e-05,
+      "loss": 0.956,
+      "step": 2237
+    },
+    {
+      "epoch": 0.5124212936462507,
+      "grad_norm": 1.0975548028945923,
+      "learning_rate": 1.0107561310686247e-05,
+      "loss": 1.0621,
+      "step": 2238
+    },
+    {
+      "epoch": 0.5126502575844305,
+      "grad_norm": 1.4685457944869995,
+      "learning_rate": 1.0100143546432527e-05,
+      "loss": 1.11,
+      "step": 2239
+    },
+    {
+      "epoch": 0.5128792215226102,
+      "grad_norm": 1.5499354600906372,
+      "learning_rate": 1.0092725727070653e-05,
+      "loss": 1.0941,
+      "step": 2240
+    },
+    {
+      "epoch": 0.51310818546079,
+      "grad_norm": 1.6682047843933105,
+      "learning_rate": 1.0085307856682593e-05,
+      "loss": 1.0784,
+      "step": 2241
+    },
+    {
+      "epoch": 0.5133371493989697,
+      "grad_norm": 1.0723470449447632,
+      "learning_rate": 1.007788993935033e-05,
+      "loss": 1.0805,
+      "step": 2242
+    },
+    {
+      "epoch": 0.5135661133371494,
+      "grad_norm": 1.175240397453308,
+      "learning_rate": 1.007047197915589e-05,
+      "loss": 1.098,
+      "step": 2243
+    },
+    {
+      "epoch": 0.5137950772753291,
+      "grad_norm": 1.1597743034362793,
+      "learning_rate": 1.0063053980181305e-05,
+      "loss": 1.0723,
+      "step": 2244
+    },
+    {
+      "epoch": 0.5140240412135089,
+      "grad_norm": 1.0197961330413818,
+      "learning_rate": 1.0055635946508649e-05,
+      "loss": 1.0706,
+      "step": 2245
+    },
+    {
+      "epoch": 0.5142530051516886,
+      "grad_norm": 1.1021794080734253,
+      "learning_rate": 1.0048217882219995e-05,
+      "loss": 1.0715,
+      "step": 2246
+    },
+    {
+      "epoch": 0.5144819690898683,
+      "grad_norm": 1.0522361993789673,
+      "learning_rate": 1.0040799791397444e-05,
+      "loss": 1.0604,
+      "step": 2247
+    },
+    {
+      "epoch": 0.514710933028048,
+      "grad_norm": 1.32160484790802,
+      "learning_rate": 1.0033381678123113e-05,
+      "loss": 1.0638,
+      "step": 2248
+    },
+    {
+      "epoch": 0.5149398969662278,
+      "grad_norm": 1.5277987718582153,
+      "learning_rate": 1.002596354647912e-05,
+      "loss": 1.1117,
+      "step": 2249
+    },
+    {
+      "epoch": 0.5151688609044076,
+      "grad_norm": 1.1318628787994385,
+      "learning_rate": 1.0018545400547609e-05,
+      "loss": 1.0124,
+      "step": 2250
+    },
+    {
+      "epoch": 0.5153978248425873,
+      "grad_norm": 1.8370736837387085,
+      "learning_rate": 1.001112724441072e-05,
+      "loss": 1.0463,
+      "step": 2251
+    },
+    {
+      "epoch": 0.5156267887807671,
+      "grad_norm": 2.369612693786621,
+      "learning_rate": 1.0003709082150598e-05,
+      "loss": 1.0653,
+      "step": 2252
+    },
+    {
+      "epoch": 0.5158557527189468,
+      "grad_norm": 1.059601902961731,
+      "learning_rate": 9.996290917849405e-06,
+      "loss": 1.1053,
+      "step": 2253
+    },
+    {
+      "epoch": 0.5160847166571265,
+      "grad_norm": 1.0960521697998047,
+      "learning_rate": 9.988872755589283e-06,
+      "loss": 1.0594,
+      "step": 2254
+    },
+    {
+      "epoch": 0.5163136805953062,
+      "grad_norm": 1.3870129585266113,
+      "learning_rate": 9.981454599452391e-06,
+      "loss": 1.0288,
+      "step": 2255
+    },
+    {
+      "epoch": 0.516542644533486,
+      "grad_norm": 1.0221267938613892,
+      "learning_rate": 9.974036453520881e-06,
+      "loss": 0.9995,
+      "step": 2256
+    },
+    {
+      "epoch": 0.5167716084716657,
+      "grad_norm": 1.0023499727249146,
+      "learning_rate": 9.966618321876889e-06,
+      "loss": 1.0261,
+      "step": 2257
+    },
+    {
+      "epoch": 0.5170005724098454,
+      "grad_norm": 1.196207880973816,
+      "learning_rate": 9.95920020860256e-06,
+      "loss": 1.0774,
+      "step": 2258
+    },
+    {
+      "epoch": 0.5172295363480252,
+      "grad_norm": 1.261536717414856,
+      "learning_rate": 9.951782117780008e-06,
+      "loss": 1.0645,
+      "step": 2259
+    },
+    {
+      "epoch": 0.5174585002862049,
+      "grad_norm": 1.8544546365737915,
+      "learning_rate": 9.944364053491356e-06,
+      "loss": 1.0541,
+      "step": 2260
+    },
+    {
+      "epoch": 0.5176874642243846,
+      "grad_norm": 1.036802053451538,
+      "learning_rate": 9.936946019818698e-06,
+      "loss": 1.0645,
+      "step": 2261
+    },
+    {
+      "epoch": 0.5179164281625644,
+      "grad_norm": 1.515656590461731,
+      "learning_rate": 9.929528020844114e-06,
+      "loss": 1.0333,
+      "step": 2262
+    },
+    {
+      "epoch": 0.5181453921007442,
+      "grad_norm": 1.1658852100372314,
+      "learning_rate": 9.922110060649672e-06,
+      "loss": 1.0802,
+      "step": 2263
+    },
+    {
+      "epoch": 0.5183743560389239,
+      "grad_norm": 1.0310983657836914,
+      "learning_rate": 9.914692143317412e-06,
+      "loss": 1.0777,
+      "step": 2264
+    },
+    {
+      "epoch": 0.5186033199771036,
+      "grad_norm": 1.1228355169296265,
+      "learning_rate": 9.90727427292935e-06,
+      "loss": 1.0623,
+      "step": 2265
+    },
+    {
+      "epoch": 0.5188322839152834,
+      "grad_norm": 1.1571909189224243,
+      "learning_rate": 9.899856453567474e-06,
+      "loss": 1.0168,
+      "step": 2266
+    },
+    {
+      "epoch": 0.5190612478534631,
+      "grad_norm": 1.233893871307373,
+      "learning_rate": 9.892438689313757e-06,
+      "loss": 1.0401,
+      "step": 2267
+    },
+    {
+      "epoch": 0.5192902117916428,
+      "grad_norm": 2.297332525253296,
+      "learning_rate": 9.885020984250124e-06,
+      "loss": 1.0835,
+      "step": 2268
+    },
+    {
+      "epoch": 0.5195191757298225,
+      "grad_norm": 1.0272998809814453,
+      "learning_rate": 9.877603342458483e-06,
+      "loss": 1.0059,
+      "step": 2269
+    },
+    {
+      "epoch": 0.5197481396680023,
+      "grad_norm": 1.3607406616210938,
+      "learning_rate": 9.870185768020694e-06,
+      "loss": 1.072,
+      "step": 2270
+    },
+    {
+      "epoch": 0.519977103606182,
+      "grad_norm": 1.6629681587219238,
+      "learning_rate": 9.862768265018585e-06,
+      "loss": 1.0697,
+      "step": 2271
+    },
+    {
+      "epoch": 0.5202060675443617,
+      "grad_norm": 1.2140170335769653,
+      "learning_rate": 9.855350837533958e-06,
+      "loss": 1.0702,
+      "step": 2272
+    },
+    {
+      "epoch": 0.5204350314825416,
+      "grad_norm": 1.465503454208374,
+      "learning_rate": 9.847933489648545e-06,
+      "loss": 1.1246,
+      "step": 2273
+    },
+    {
+      "epoch": 0.5206639954207213,
+      "grad_norm": 1.1289547681808472,
+      "learning_rate": 9.840516225444059e-06,
+      "loss": 1.0491,
+      "step": 2274
+    },
+    {
+      "epoch": 0.520892959358901,
+      "grad_norm": 1.1703757047653198,
+      "learning_rate": 9.83309904900216e-06,
+      "loss": 1.0422,
+      "step": 2275
+    },
+    {
+      "epoch": 0.5211219232970807,
+      "grad_norm": 1.3183718919754028,
+      "learning_rate": 9.82568196440445e-06,
+      "loss": 1.1088,
+      "step": 2276
+    },
+    {
+      "epoch": 0.5213508872352605,
+      "grad_norm": 1.2612817287445068,
+      "learning_rate": 9.818264975732497e-06,
+      "loss": 1.0711,
+      "step": 2277
+    },
+    {
+      "epoch": 0.5215798511734402,
+      "grad_norm": 1.238250732421875,
+      "learning_rate": 9.810848087067805e-06,
+      "loss": 1.0843,
+      "step": 2278
+    },
+    {
+      "epoch": 0.5218088151116199,
+      "grad_norm": 1.1225922107696533,
+      "learning_rate": 9.803431302491823e-06,
+      "loss": 1.0751,
+      "step": 2279
+    },
+    {
+      "epoch": 0.5220377790497996,
+      "grad_norm": 1.3378745317459106,
+      "learning_rate": 9.79601462608595e-06,
+      "loss": 1.0344,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5222667429879794,
+      "grad_norm": 1.1414393186569214,
+      "learning_rate": 9.788598061931513e-06,
+      "loss": 1.0469,
+      "step": 2281
+    },
+    {
+      "epoch": 0.5224957069261591,
+      "grad_norm": 1.9631890058517456,
+      "learning_rate": 9.781181614109793e-06,
+      "loss": 1.1085,
+      "step": 2282
+    },
+    {
+      "epoch": 0.5227246708643388,
+      "grad_norm": 1.1475551128387451,
+      "learning_rate": 9.773765286701998e-06,
+      "loss": 1.0458,
+      "step": 2283
+    },
+    {
+      "epoch": 0.5229536348025186,
+      "grad_norm": 1.056517481803894,
+      "learning_rate": 9.766349083789266e-06,
+      "loss": 1.0451,
+      "step": 2284
+    },
+    {
+      "epoch": 0.5231825987406984,
+      "grad_norm": 1.2475084066390991,
+      "learning_rate": 9.758933009452674e-06,
+      "loss": 1.0261,
+      "step": 2285
+    },
+    {
+      "epoch": 0.5234115626788781,
+      "grad_norm": 1.041703701019287,
+      "learning_rate": 9.751517067773228e-06,
+      "loss": 1.0591,
+      "step": 2286
+    },
+    {
+      "epoch": 0.5236405266170578,
+      "grad_norm": 1.334044337272644,
+      "learning_rate": 9.744101262831855e-06,
+      "loss": 1.0793,
+      "step": 2287
+    },
+    {
+      "epoch": 0.5238694905552376,
+      "grad_norm": 1.1858912706375122,
+      "learning_rate": 9.736685598709413e-06,
+      "loss": 1.084,
+      "step": 2288
+    },
+    {
+      "epoch": 0.5240984544934173,
+      "grad_norm": 1.2032268047332764,
+      "learning_rate": 9.72927007948668e-06,
+      "loss": 1.0218,
+      "step": 2289
+    },
+    {
+      "epoch": 0.524327418431597,
+      "grad_norm": 1.705446720123291,
+      "learning_rate": 9.721854709244346e-06,
+      "loss": 1.0121,
+      "step": 2290
+    },
+    {
+      "epoch": 0.5245563823697768,
+      "grad_norm": 1.374306559562683,
+      "learning_rate": 9.71443949206304e-06,
+      "loss": 1.0152,
+      "step": 2291
+    },
+    {
+      "epoch": 0.5247853463079565,
+      "grad_norm": 1.2469069957733154,
+      "learning_rate": 9.707024432023284e-06,
+      "loss": 1.0331,
+      "step": 2292
+    },
+    {
+      "epoch": 0.5250143102461362,
+      "grad_norm": 1.5407803058624268,
+      "learning_rate": 9.699609533205527e-06,
+      "loss": 1.0996,
+      "step": 2293
+    },
+    {
+      "epoch": 0.5252432741843159,
+      "grad_norm": 1.1991982460021973,
+      "learning_rate": 9.692194799690128e-06,
+      "loss": 1.1026,
+      "step": 2294
+    },
+    {
+      "epoch": 0.5254722381224957,
+      "grad_norm": 1.3273327350616455,
+      "learning_rate": 9.684780235557346e-06,
+      "loss": 1.0099,
+      "step": 2295
+    },
+    {
+      "epoch": 0.5257012020606755,
+      "grad_norm": 1.2710803747177124,
+      "learning_rate": 9.67736584488736e-06,
+      "loss": 1.0538,
+      "step": 2296
+    },
+    {
+      "epoch": 0.5259301659988552,
+      "grad_norm": 1.1919000148773193,
+      "learning_rate": 9.669951631760245e-06,
+      "loss": 1.06,
+      "step": 2297
+    },
+    {
+      "epoch": 0.526159129937035,
+      "grad_norm": 1.2854589223861694,
+      "learning_rate": 9.662537600255979e-06,
+      "loss": 1.044,
+      "step": 2298
+    },
+    {
+      "epoch": 0.5263880938752147,
+      "grad_norm": 1.4513083696365356,
+      "learning_rate": 9.65512375445444e-06,
+      "loss": 1.0597,
+      "step": 2299
+    },
+    {
+      "epoch": 0.5266170578133944,
+      "grad_norm": 1.1614129543304443,
+      "learning_rate": 9.647710098435413e-06,
+      "loss": 0.9531,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5268460217515741,
+      "grad_norm": 1.165914535522461,
+      "learning_rate": 9.64029663627856e-06,
+      "loss": 1.0857,
+      "step": 2301
+    },
+    {
+      "epoch": 0.5270749856897539,
+      "grad_norm": 1.8047739267349243,
+      "learning_rate": 9.632883372063458e-06,
+      "loss": 1.029,
+      "step": 2302
+    },
+    {
+      "epoch": 0.5273039496279336,
+      "grad_norm": 1.2753591537475586,
+      "learning_rate": 9.625470309869554e-06,
+      "loss": 0.9939,
+      "step": 2303
+    },
+    {
+      "epoch": 0.5275329135661133,
+      "grad_norm": 1.525112509727478,
+      "learning_rate": 9.618057453776196e-06,
+      "loss": 1.0816,
+      "step": 2304
+    },
+    {
+      "epoch": 0.527761877504293,
+      "grad_norm": 1.4662785530090332,
+      "learning_rate": 9.610644807862625e-06,
+      "loss": 1.0653,
+      "step": 2305
+    },
+    {
+      "epoch": 0.5279908414424728,
+      "grad_norm": 3.009976387023926,
+      "learning_rate": 9.603232376207948e-06,
+      "loss": 0.9941,
+      "step": 2306
+    },
+    {
+      "epoch": 0.5282198053806525,
+      "grad_norm": 1.217646837234497,
+      "learning_rate": 9.59582016289117e-06,
+      "loss": 1.0431,
+      "step": 2307
+    },
+    {
+      "epoch": 0.5284487693188323,
+      "grad_norm": 1.1886101961135864,
+      "learning_rate": 9.588408171991168e-06,
+      "loss": 1.0745,
+      "step": 2308
+    },
+    {
+      "epoch": 0.5286777332570121,
+      "grad_norm": 1.1897996664047241,
+      "learning_rate": 9.580996407586695e-06,
+      "loss": 1.0537,
+      "step": 2309
+    },
+    {
+      "epoch": 0.5289066971951918,
+      "grad_norm": 1.416695237159729,
+      "learning_rate": 9.573584873756387e-06,
+      "loss": 1.0699,
+      "step": 2310
+    },
+    {
+      "epoch": 0.5291356611333715,
+      "grad_norm": 2.577252149581909,
+      "learning_rate": 9.566173574578751e-06,
+      "loss": 1.0533,
+      "step": 2311
+    },
+    {
+      "epoch": 0.5293646250715512,
+      "grad_norm": 1.102662444114685,
+      "learning_rate": 9.558762514132157e-06,
+      "loss": 1.1236,
+      "step": 2312
+    },
+    {
+      "epoch": 0.529593589009731,
+      "grad_norm": 1.6311894655227661,
+      "learning_rate": 9.551351696494854e-06,
+      "loss": 1.1093,
+      "step": 2313
+    },
+    {
+      "epoch": 0.5298225529479107,
+      "grad_norm": 1.327171802520752,
+      "learning_rate": 9.543941125744947e-06,
+      "loss": 1.0756,
+      "step": 2314
+    },
+    {
+      "epoch": 0.5300515168860904,
+      "grad_norm": 1.0158864259719849,
+      "learning_rate": 9.536530805960418e-06,
+      "loss": 1.0776,
+      "step": 2315
+    },
+    {
+      "epoch": 0.5302804808242702,
+      "grad_norm": 1.1998199224472046,
+      "learning_rate": 9.529120741219103e-06,
+      "loss": 1.0484,
+      "step": 2316
+    },
+    {
+      "epoch": 0.5305094447624499,
+      "grad_norm": 1.073457956314087,
+      "learning_rate": 9.521710935598693e-06,
+      "loss": 0.9798,
+      "step": 2317
+    },
+    {
+      "epoch": 0.5307384087006296,
+      "grad_norm": 1.1644235849380493,
+      "learning_rate": 9.514301393176742e-06,
+      "loss": 1.0871,
+      "step": 2318
+    },
+    {
+      "epoch": 0.5309673726388094,
+      "grad_norm": 1.074922800064087,
+      "learning_rate": 9.506892118030668e-06,
+      "loss": 1.0178,
+      "step": 2319
+    },
+    {
+      "epoch": 0.5311963365769892,
+      "grad_norm": 1.2031996250152588,
+      "learning_rate": 9.499483114237726e-06,
+      "loss": 1.1128,
+      "step": 2320
+    },
+    {
+      "epoch": 0.5314253005151689,
+      "grad_norm": 1.1941750049591064,
+      "learning_rate": 9.492074385875025e-06,
+      "loss": 1.0887,
+      "step": 2321
+    },
+    {
+      "epoch": 0.5316542644533486,
+      "grad_norm": 1.6538817882537842,
+      "learning_rate": 9.484665937019539e-06,
+      "loss": 1.0918,
+      "step": 2322
+    },
+    {
+      "epoch": 0.5318832283915284,
+      "grad_norm": 1.1676061153411865,
+      "learning_rate": 9.477257771748061e-06,
+      "loss": 1.0102,
+      "step": 2323
+    },
+    {
+      "epoch": 0.5321121923297081,
+      "grad_norm": 1.8035333156585693,
+      "learning_rate": 9.469849894137254e-06,
+      "loss": 1.0648,
+      "step": 2324
+    },
+    {
+      "epoch": 0.5323411562678878,
+      "grad_norm": 1.2286866903305054,
+      "learning_rate": 9.4624423082636e-06,
+      "loss": 1.0543,
+      "step": 2325
+    },
+    {
+      "epoch": 0.5325701202060675,
+      "grad_norm": 1.29671311378479,
+      "learning_rate": 9.455035018203439e-06,
+      "loss": 1.0887,
+      "step": 2326
+    },
+    {
+      "epoch": 0.5327990841442473,
+      "grad_norm": 1.1480847597122192,
+      "learning_rate": 9.44762802803294e-06,
+      "loss": 1.125,
+      "step": 2327
+    },
+    {
+      "epoch": 0.533028048082427,
+      "grad_norm": 1.277753233909607,
+      "learning_rate": 9.440221341828104e-06,
+      "loss": 1.0196,
+      "step": 2328
+    },
+    {
+      "epoch": 0.5332570120206067,
+      "grad_norm": 1.2395761013031006,
+      "learning_rate": 9.43281496366477e-06,
+      "loss": 1.0753,
+      "step": 2329
+    },
+    {
+      "epoch": 0.5334859759587864,
+      "grad_norm": 1.0635128021240234,
+      "learning_rate": 9.42540889761861e-06,
+      "loss": 1.0173,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5337149398969663,
+      "grad_norm": 1.3224315643310547,
+      "learning_rate": 9.418003147765113e-06,
+      "loss": 0.992,
+      "step": 2331
+    },
+    {
+      "epoch": 0.533943903835146,
+      "grad_norm": 0.9989613890647888,
+      "learning_rate": 9.410597718179603e-06,
+      "loss": 1.0676,
+      "step": 2332
+    },
+    {
+      "epoch": 0.5341728677733257,
+      "grad_norm": 1.0474404096603394,
+      "learning_rate": 9.40319261293723e-06,
+      "loss": 1.0012,
+      "step": 2333
+    },
+    {
+      "epoch": 0.5344018317115055,
+      "grad_norm": 1.0710350275039673,
+      "learning_rate": 9.395787836112958e-06,
+      "loss": 1.0822,
+      "step": 2334
+    },
+    {
+      "epoch": 0.5346307956496852,
+      "grad_norm": 1.1103330850601196,
+      "learning_rate": 9.388383391781576e-06,
+      "loss": 1.05,
+      "step": 2335
+    },
+    {
+      "epoch": 0.5348597595878649,
+      "grad_norm": 1.2348805665969849,
+      "learning_rate": 9.380979284017682e-06,
+      "loss": 1.0656,
+      "step": 2336
+    },
+    {
+      "epoch": 0.5350887235260446,
+      "grad_norm": 1.1286513805389404,
+      "learning_rate": 9.373575516895698e-06,
+      "loss": 1.0303,
+      "step": 2337
+    },
+    {
+      "epoch": 0.5353176874642244,
+      "grad_norm": 1.173068881034851,
+      "learning_rate": 9.366172094489858e-06,
+      "loss": 1.0615,
+      "step": 2338
+    },
+    {
+      "epoch": 0.5355466514024041,
+      "grad_norm": 1.185383915901184,
+      "learning_rate": 9.358769020874198e-06,
+      "loss": 1.0928,
+      "step": 2339
+    },
+    {
+      "epoch": 0.5357756153405838,
+      "grad_norm": 1.1078901290893555,
+      "learning_rate": 9.351366300122569e-06,
+      "loss": 1.081,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5360045792787635,
+      "grad_norm": 1.2607088088989258,
+      "learning_rate": 9.343963936308634e-06,
+      "loss": 1.017,
+      "step": 2341
+    },
+    {
+      "epoch": 0.5362335432169434,
+      "grad_norm": 1.102243423461914,
+      "learning_rate": 9.336561933505836e-06,
+      "loss": 1.0839,
+      "step": 2342
+    },
+    {
+      "epoch": 0.5364625071551231,
+      "grad_norm": 1.3678319454193115,
+      "learning_rate": 9.329160295787447e-06,
+      "loss": 1.073,
+      "step": 2343
+    },
+    {
+      "epoch": 0.5366914710933028,
+      "grad_norm": 1.1074059009552002,
+      "learning_rate": 9.321759027226525e-06,
+      "loss": 1.0501,
+      "step": 2344
+    },
+    {
+      "epoch": 0.5369204350314826,
+      "grad_norm": 1.2822704315185547,
+      "learning_rate": 9.314358131895919e-06,
+      "loss": 1.0713,
+      "step": 2345
+    },
+    {
+      "epoch": 0.5371493989696623,
+      "grad_norm": 1.119148850440979,
+      "learning_rate": 9.306957613868292e-06,
+      "loss": 1.0374,
+      "step": 2346
+    },
+    {
+      "epoch": 0.537378362907842,
+      "grad_norm": 1.2142623662948608,
+      "learning_rate": 9.299557477216073e-06,
+      "loss": 1.0243,
+      "step": 2347
+    },
+    {
+      "epoch": 0.5376073268460217,
+      "grad_norm": 1.343944787979126,
+      "learning_rate": 9.292157726011502e-06,
+      "loss": 1.1066,
+      "step": 2348
+    },
+    {
+      "epoch": 0.5378362907842015,
+      "grad_norm": 1.2517242431640625,
+      "learning_rate": 9.284758364326604e-06,
+      "loss": 1.0373,
+      "step": 2349
+    },
+    {
+      "epoch": 0.5380652547223812,
+      "grad_norm": 1.8333529233932495,
+      "learning_rate": 9.277359396233176e-06,
+      "loss": 1.04,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5382942186605609,
+      "grad_norm": 1.1809799671173096,
+      "learning_rate": 9.269960825802817e-06,
+      "loss": 1.0577,
+      "step": 2351
+    },
+    {
+      "epoch": 0.5385231825987407,
+      "grad_norm": 1.1359210014343262,
+      "learning_rate": 9.262562657106898e-06,
+      "loss": 1.0731,
+      "step": 2352
+    },
+    {
+      "epoch": 0.5387521465369204,
+      "grad_norm": 1.6872447729110718,
+      "learning_rate": 9.255164894216562e-06,
+      "loss": 1.0443,
+      "step": 2353
+    },
+    {
+      "epoch": 0.5389811104751002,
+      "grad_norm": 1.3061941862106323,
+      "learning_rate": 9.247767541202738e-06,
+      "loss": 1.0525,
+      "step": 2354
+    },
+    {
+      "epoch": 0.53921007441328,
+      "grad_norm": 1.4695810079574585,
+      "learning_rate": 9.240370602136134e-06,
+      "loss": 1.0589,
+      "step": 2355
+    },
+    {
+      "epoch": 0.5394390383514597,
+      "grad_norm": 1.203343152999878,
+      "learning_rate": 9.232974081087216e-06,
+      "loss": 1.0574,
+      "step": 2356
+    },
+    {
+      "epoch": 0.5396680022896394,
+      "grad_norm": 1.3898227214813232,
+      "learning_rate": 9.225577982126234e-06,
+      "loss": 1.0697,
+      "step": 2357
+    },
+    {
+      "epoch": 0.5398969662278191,
+      "grad_norm": 1.1392302513122559,
+      "learning_rate": 9.218182309323193e-06,
+      "loss": 1.0639,
+      "step": 2358
+    },
+    {
+      "epoch": 0.5401259301659989,
+      "grad_norm": 0.9744246602058411,
+      "learning_rate": 9.21078706674787e-06,
+      "loss": 1.0744,
+      "step": 2359
+    },
+    {
+      "epoch": 0.5403548941041786,
+      "grad_norm": 1.147935152053833,
+      "learning_rate": 9.203392258469814e-06,
+      "loss": 1.0258,
+      "step": 2360
+    },
+    {
+      "epoch": 0.5405838580423583,
+      "grad_norm": 1.1611018180847168,
+      "learning_rate": 9.195997888558312e-06,
+      "loss": 1.0645,
+      "step": 2361
+    },
+    {
+      "epoch": 0.540812821980538,
+      "grad_norm": 1.4205443859100342,
+      "learning_rate": 9.188603961082436e-06,
+      "loss": 1.0319,
+      "step": 2362
+    },
+    {
+      "epoch": 0.5410417859187178,
+      "grad_norm": 1.160415768623352,
+      "learning_rate": 9.181210480110997e-06,
+      "loss": 1.1,
+      "step": 2363
+    },
+    {
+      "epoch": 0.5412707498568975,
+      "grad_norm": 1.692002296447754,
+      "learning_rate": 9.173817449712563e-06,
+      "loss": 1.0234,
+      "step": 2364
+    },
+    {
+      "epoch": 0.5414997137950772,
+      "grad_norm": 1.0437875986099243,
+      "learning_rate": 9.166424873955462e-06,
+      "loss": 1.0846,
+      "step": 2365
+    },
+    {
+      "epoch": 0.5417286777332571,
+      "grad_norm": 2.888340473175049,
+      "learning_rate": 9.159032756907765e-06,
+      "loss": 1.0399,
+      "step": 2366
+    },
+    {
+      "epoch": 0.5419576416714368,
+      "grad_norm": 1.0444552898406982,
+      "learning_rate": 9.151641102637289e-06,
+      "loss": 1.0821,
+      "step": 2367
+    },
+    {
+      "epoch": 0.5421866056096165,
+      "grad_norm": 1.2841830253601074,
+      "learning_rate": 9.144249915211605e-06,
+      "loss": 1.0579,
+      "step": 2368
+    },
+    {
+      "epoch": 0.5424155695477962,
+      "grad_norm": 1.8997342586517334,
+      "learning_rate": 9.136859198698014e-06,
+      "loss": 1.1244,
+      "step": 2369
+    },
+    {
+      "epoch": 0.542644533485976,
+      "grad_norm": 1.4083105325698853,
+      "learning_rate": 9.12946895716357e-06,
+      "loss": 1.0149,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5428734974241557,
+      "grad_norm": 1.2661566734313965,
+      "learning_rate": 9.122079194675064e-06,
+      "loss": 1.0958,
+      "step": 2371
+    },
+    {
+      "epoch": 0.5431024613623354,
+      "grad_norm": 1.254859209060669,
+      "learning_rate": 9.114689915299015e-06,
+      "loss": 1.1392,
+      "step": 2372
+    },
+    {
+      "epoch": 0.5433314253005151,
+      "grad_norm": 1.7026736736297607,
+      "learning_rate": 9.10730112310168e-06,
+      "loss": 1.0332,
+      "step": 2373
+    },
+    {
+      "epoch": 0.5435603892386949,
+      "grad_norm": 1.1682655811309814,
+      "learning_rate": 9.099912822149056e-06,
+      "loss": 1.0609,
+      "step": 2374
+    },
+    {
+      "epoch": 0.5437893531768746,
+      "grad_norm": 1.0199397802352905,
+      "learning_rate": 9.092525016506858e-06,
+      "loss": 1.0164,
+      "step": 2375
+    },
+    {
+      "epoch": 0.5440183171150543,
+      "grad_norm": 1.0826716423034668,
+      "learning_rate": 9.085137710240536e-06,
+      "loss": 1.0245,
+      "step": 2376
+    },
+    {
+      "epoch": 0.5442472810532342,
+      "grad_norm": 1.1761521100997925,
+      "learning_rate": 9.077750907415264e-06,
+      "loss": 1.0278,
+      "step": 2377
+    },
+    {
+      "epoch": 0.5444762449914139,
+      "grad_norm": 1.112975001335144,
+      "learning_rate": 9.07036461209593e-06,
+      "loss": 1.021,
+      "step": 2378
+    },
+    {
+      "epoch": 0.5447052089295936,
+      "grad_norm": 1.1769195795059204,
+      "learning_rate": 9.06297882834716e-06,
+      "loss": 1.0348,
+      "step": 2379
+    },
+    {
+      "epoch": 0.5449341728677733,
+      "grad_norm": 1.3177459239959717,
+      "learning_rate": 9.055593560233284e-06,
+      "loss": 1.0599,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5451631368059531,
+      "grad_norm": 1.2103499174118042,
+      "learning_rate": 9.048208811818353e-06,
+      "loss": 1.0701,
+      "step": 2381
+    },
+    {
+      "epoch": 0.5453921007441328,
+      "grad_norm": 1.5910011529922485,
+      "learning_rate": 9.040824587166136e-06,
+      "loss": 1.1076,
+      "step": 2382
+    },
+    {
+      "epoch": 0.5456210646823125,
+      "grad_norm": 1.020159125328064,
+      "learning_rate": 9.033440890340102e-06,
+      "loss": 1.0,
+      "step": 2383
+    },
+    {
+      "epoch": 0.5458500286204923,
+      "grad_norm": 1.280670166015625,
+      "learning_rate": 9.026057725403445e-06,
+      "loss": 1.015,
+      "step": 2384
+    },
+    {
+      "epoch": 0.546078992558672,
+      "grad_norm": 1.4677191972732544,
+      "learning_rate": 9.018675096419059e-06,
+      "loss": 1.0787,
+      "step": 2385
+    },
+    {
+      "epoch": 0.5463079564968517,
+      "grad_norm": 1.1645160913467407,
+      "learning_rate": 9.011293007449535e-06,
+      "loss": 1.0954,
+      "step": 2386
+    },
+    {
+      "epoch": 0.5465369204350314,
+      "grad_norm": 1.4413548707962036,
+      "learning_rate": 9.003911462557185e-06,
+      "loss": 1.0522,
+      "step": 2387
+    },
+    {
+      "epoch": 0.5467658843732112,
+      "grad_norm": 1.051669955253601,
+      "learning_rate": 8.996530465804e-06,
+      "loss": 1.0343,
+      "step": 2388
+    },
+    {
+      "epoch": 0.546994848311391,
+      "grad_norm": 1.315771222114563,
+      "learning_rate": 8.98915002125169e-06,
+      "loss": 1.0497,
+      "step": 2389
+    },
+    {
+      "epoch": 0.5472238122495707,
+      "grad_norm": 1.1893324851989746,
+      "learning_rate": 8.981770132961649e-06,
+      "loss": 1.0671,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5474527761877505,
+      "grad_norm": 1.8192094564437866,
+      "learning_rate": 8.974390804994966e-06,
+      "loss": 1.0127,
+      "step": 2391
+    },
+    {
+      "epoch": 0.5476817401259302,
+      "grad_norm": 1.2040550708770752,
+      "learning_rate": 8.96701204141242e-06,
+      "loss": 1.1041,
+      "step": 2392
+    },
+    {
+      "epoch": 0.5479107040641099,
+      "grad_norm": 1.1481995582580566,
+      "learning_rate": 8.959633846274494e-06,
+      "loss": 1.0686,
+      "step": 2393
+    },
+    {
+      "epoch": 0.5481396680022896,
+      "grad_norm": 1.2036935091018677,
+      "learning_rate": 8.952256223641337e-06,
+      "loss": 1.0603,
+      "step": 2394
+    },
+    {
+      "epoch": 0.5483686319404694,
+      "grad_norm": 1.1734176874160767,
+      "learning_rate": 8.944879177572796e-06,
+      "loss": 1.0554,
+      "step": 2395
+    },
+    {
+      "epoch": 0.5485975958786491,
+      "grad_norm": 1.3413574695587158,
+      "learning_rate": 8.9375027121284e-06,
+      "loss": 0.9716,
+      "step": 2396
+    },
+    {
+      "epoch": 0.5488265598168288,
+      "grad_norm": 1.8128085136413574,
+      "learning_rate": 8.930126831367346e-06,
+      "loss": 1.0691,
+      "step": 2397
+    },
+    {
+      "epoch": 0.5490555237550085,
+      "grad_norm": 1.5389128923416138,
+      "learning_rate": 8.922751539348534e-06,
+      "loss": 1.0215,
+      "step": 2398
+    },
+    {
+      "epoch": 0.5492844876931883,
+      "grad_norm": 1.3885033130645752,
+      "learning_rate": 8.915376840130517e-06,
+      "loss": 1.1011,
+      "step": 2399
+    },
+    {
+      "epoch": 0.5495134516313681,
+      "grad_norm": 1.6749422550201416,
+      "learning_rate": 8.908002737771529e-06,
+      "loss": 1.0884,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5497424155695478,
+      "grad_norm": 1.39671790599823,
+      "learning_rate": 8.900629236329482e-06,
+      "loss": 1.0296,
+      "step": 2401
+    },
+    {
+      "epoch": 0.5499713795077276,
+      "grad_norm": 1.2996166944503784,
+      "learning_rate": 8.893256339861946e-06,
+      "loss": 1.0335,
+      "step": 2402
+    },
+    {
+      "epoch": 0.5502003434459073,
+      "grad_norm": 1.482348084449768,
+      "learning_rate": 8.885884052426168e-06,
+      "loss": 1.0564,
+      "step": 2403
+    },
+    {
+      "epoch": 0.550429307384087,
+      "grad_norm": 1.385023593902588,
+      "learning_rate": 8.878512378079057e-06,
+      "loss": 1.0633,
+      "step": 2404
+    },
+    {
+      "epoch": 0.5506582713222667,
+      "grad_norm": 1.3737043142318726,
+      "learning_rate": 8.871141320877181e-06,
+      "loss": 1.0962,
+      "step": 2405
+    },
+    {
+      "epoch": 0.5508872352604465,
+      "grad_norm": 1.2021304368972778,
+      "learning_rate": 8.86377088487677e-06,
+      "loss": 1.0578,
+      "step": 2406
+    },
+    {
+      "epoch": 0.5511161991986262,
+      "grad_norm": 1.1903609037399292,
+      "learning_rate": 8.856401074133718e-06,
+      "loss": 1.0453,
+      "step": 2407
+    },
+    {
+      "epoch": 0.5513451631368059,
+      "grad_norm": 1.103837013244629,
+      "learning_rate": 8.849031892703564e-06,
+      "loss": 0.9953,
+      "step": 2408
+    },
+    {
+      "epoch": 0.5515741270749857,
+      "grad_norm": 1.2637211084365845,
+      "learning_rate": 8.841663344641514e-06,
+      "loss": 1.0946,
+      "step": 2409
+    },
+    {
+      "epoch": 0.5518030910131654,
+      "grad_norm": 1.2190263271331787,
+      "learning_rate": 8.83429543400241e-06,
+      "loss": 1.0259,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5520320549513451,
+      "grad_norm": 1.3978723287582397,
+      "learning_rate": 8.826928164840755e-06,
+      "loss": 1.0372,
+      "step": 2411
+    },
+    {
+      "epoch": 0.552261018889525,
+      "grad_norm": 1.1146042346954346,
+      "learning_rate": 8.819561541210698e-06,
+      "loss": 1.0455,
+      "step": 2412
+    },
+    {
+      "epoch": 0.5524899828277047,
+      "grad_norm": 1.5733097791671753,
+      "learning_rate": 8.812195567166028e-06,
+      "loss": 1.0123,
+      "step": 2413
+    },
+    {
+      "epoch": 0.5527189467658844,
+      "grad_norm": 1.4020459651947021,
+      "learning_rate": 8.804830246760175e-06,
+      "loss": 1.0588,
+      "step": 2414
+    },
+    {
+      "epoch": 0.5529479107040641,
+      "grad_norm": 1.1088488101959229,
+      "learning_rate": 8.797465584046225e-06,
+      "loss": 1.0534,
+      "step": 2415
+    },
+    {
+      "epoch": 0.5531768746422439,
+      "grad_norm": 1.1872738599777222,
+      "learning_rate": 8.790101583076874e-06,
+      "loss": 1.0252,
+      "step": 2416
+    },
+    {
+      "epoch": 0.5534058385804236,
+      "grad_norm": 1.0902388095855713,
+      "learning_rate": 8.78273824790448e-06,
+      "loss": 1.0794,
+      "step": 2417
+    },
+    {
+      "epoch": 0.5536348025186033,
+      "grad_norm": 1.246376633644104,
+      "learning_rate": 8.775375582581027e-06,
+      "loss": 1.0609,
+      "step": 2418
+    },
+    {
+      "epoch": 0.553863766456783,
+      "grad_norm": 1.1201062202453613,
+      "learning_rate": 8.76801359115812e-06,
+      "loss": 1.0713,
+      "step": 2419
+    },
+    {
+      "epoch": 0.5540927303949628,
+      "grad_norm": 1.1334254741668701,
+      "learning_rate": 8.760652277687007e-06,
+      "loss": 1.0681,
+      "step": 2420
+    },
+    {
+      "epoch": 0.5543216943331425,
+      "grad_norm": 1.380236029624939,
+      "learning_rate": 8.75329164621855e-06,
+      "loss": 1.0457,
+      "step": 2421
+    },
+    {
+      "epoch": 0.5545506582713222,
+      "grad_norm": 1.3209164142608643,
+      "learning_rate": 8.745931700803251e-06,
+      "loss": 1.0322,
+      "step": 2422
+    },
+    {
+      "epoch": 0.554779622209502,
+      "grad_norm": 1.5157129764556885,
+      "learning_rate": 8.738572445491225e-06,
+      "loss": 1.0049,
+      "step": 2423
+    },
+    {
+      "epoch": 0.5550085861476818,
+      "grad_norm": 1.1138639450073242,
+      "learning_rate": 8.731213884332205e-06,
+      "loss": 1.0625,
+      "step": 2424
+    },
+    {
+      "epoch": 0.5552375500858615,
+      "grad_norm": 1.2549896240234375,
+      "learning_rate": 8.723856021375543e-06,
+      "loss": 1.0304,
+      "step": 2425
+    },
+    {
+      "epoch": 0.5554665140240412,
+      "grad_norm": 1.2935527563095093,
+      "learning_rate": 8.716498860670218e-06,
+      "loss": 1.0602,
+      "step": 2426
+    },
+    {
+      "epoch": 0.555695477962221,
+      "grad_norm": 1.2371066808700562,
+      "learning_rate": 8.709142406264807e-06,
+      "loss": 1.0347,
+      "step": 2427
+    },
+    {
+      "epoch": 0.5559244419004007,
+      "grad_norm": 1.1444132328033447,
+      "learning_rate": 8.701786662207506e-06,
+      "loss": 1.0418,
+      "step": 2428
+    },
+    {
+      "epoch": 0.5561534058385804,
+      "grad_norm": 1.1240417957305908,
+      "learning_rate": 8.694431632546127e-06,
+      "loss": 0.9742,
+      "step": 2429
+    },
+    {
+      "epoch": 0.5563823697767601,
+      "grad_norm": 1.1536635160446167,
+      "learning_rate": 8.687077321328066e-06,
+      "loss": 1.0736,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5566113337149399,
+      "grad_norm": 1.3496026992797852,
+      "learning_rate": 8.679723732600355e-06,
+      "loss": 1.0392,
+      "step": 2431
+    },
+    {
+      "epoch": 0.5568402976531196,
+      "grad_norm": 1.3216404914855957,
+      "learning_rate": 8.672370870409601e-06,
+      "loss": 1.053,
+      "step": 2432
+    },
+    {
+      "epoch": 0.5570692615912993,
+      "grad_norm": 1.1046510934829712,
+      "learning_rate": 8.665018738802023e-06,
+      "loss": 1.0534,
+      "step": 2433
+    },
+    {
+      "epoch": 0.557298225529479,
+      "grad_norm": 1.7113914489746094,
+      "learning_rate": 8.657667341823449e-06,
+      "loss": 1.092,
+      "step": 2434
+    },
+    {
+      "epoch": 0.5575271894676589,
+      "grad_norm": 1.2329431772232056,
+      "learning_rate": 8.650316683519276e-06,
+      "loss": 1.0156,
+      "step": 2435
+    },
+    {
+      "epoch": 0.5577561534058386,
+      "grad_norm": 1.0992010831832886,
+      "learning_rate": 8.642966767934516e-06,
+      "loss": 1.0104,
+      "step": 2436
+    },
+    {
+      "epoch": 0.5579851173440183,
+      "grad_norm": 1.1873196363449097,
+      "learning_rate": 8.63561759911377e-06,
+      "loss": 0.9733,
+      "step": 2437
+    },
+    {
+      "epoch": 0.5582140812821981,
+      "grad_norm": 1.360651969909668,
+      "learning_rate": 8.628269181101216e-06,
+      "loss": 1.0369,
+      "step": 2438
+    },
+    {
+      "epoch": 0.5584430452203778,
+      "grad_norm": 1.1703392267227173,
+      "learning_rate": 8.620921517940635e-06,
+      "loss": 1.0778,
+      "step": 2439
+    },
+    {
+      "epoch": 0.5586720091585575,
+      "grad_norm": 1.1417691707611084,
+      "learning_rate": 8.61357461367538e-06,
+      "loss": 1.0593,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5589009730967373,
+      "grad_norm": 1.130734920501709,
+      "learning_rate": 8.60622847234839e-06,
+      "loss": 1.0531,
+      "step": 2441
+    },
+    {
+      "epoch": 0.559129937034917,
+      "grad_norm": 1.1957651376724243,
+      "learning_rate": 8.598883098002188e-06,
+      "loss": 1.0961,
+      "step": 2442
+    },
+    {
+      "epoch": 0.5593589009730967,
+      "grad_norm": 1.3269649744033813,
+      "learning_rate": 8.591538494678867e-06,
+      "loss": 1.0608,
+      "step": 2443
+    },
+    {
+      "epoch": 0.5595878649112764,
+      "grad_norm": 1.2087531089782715,
+      "learning_rate": 8.584194666420105e-06,
+      "loss": 1.1281,
+      "step": 2444
+    },
+    {
+      "epoch": 0.5598168288494562,
+      "grad_norm": 1.5364724397659302,
+      "learning_rate": 8.576851617267151e-06,
+      "loss": 1.0366,
+      "step": 2445
+    },
+    {
+      "epoch": 0.560045792787636,
+      "grad_norm": 1.299095630645752,
+      "learning_rate": 8.569509351260817e-06,
+      "loss": 1.0942,
+      "step": 2446
+    },
+    {
+      "epoch": 0.5602747567258157,
+      "grad_norm": 1.0818737745285034,
+      "learning_rate": 8.562167872441493e-06,
+      "loss": 1.1259,
+      "step": 2447
+    },
+    {
+      "epoch": 0.5605037206639955,
+      "grad_norm": 1.1360951662063599,
+      "learning_rate": 8.554827184849139e-06,
+      "loss": 1.1018,
+      "step": 2448
+    },
+    {
+      "epoch": 0.5607326846021752,
+      "grad_norm": 1.383863925933838,
+      "learning_rate": 8.547487292523264e-06,
+      "loss": 1.0885,
+      "step": 2449
+    },
+    {
+      "epoch": 0.5609616485403549,
+      "grad_norm": 1.0860772132873535,
+      "learning_rate": 8.540148199502955e-06,
+      "loss": 1.0822,
+      "step": 2450
+    },
+    {
+      "epoch": 0.5611906124785346,
+      "grad_norm": 1.1337288618087769,
+      "learning_rate": 8.532809909826854e-06,
+      "loss": 1.0814,
+      "step": 2451
+    },
+    {
+      "epoch": 0.5614195764167144,
+      "grad_norm": 1.118790626525879,
+      "learning_rate": 8.525472427533156e-06,
+      "loss": 1.0809,
+      "step": 2452
+    },
+    {
+      "epoch": 0.5616485403548941,
+      "grad_norm": 1.1470918655395508,
+      "learning_rate": 8.518135756659624e-06,
+      "loss": 1.0631,
+      "step": 2453
+    },
+    {
+      "epoch": 0.5618775042930738,
+      "grad_norm": 1.0622974634170532,
+      "learning_rate": 8.510799901243554e-06,
+      "loss": 1.0763,
+      "step": 2454
+    },
+    {
+      "epoch": 0.5621064682312535,
+      "grad_norm": 1.0906604528427124,
+      "learning_rate": 8.503464865321817e-06,
+      "loss": 1.069,
+      "step": 2455
+    },
+    {
+      "epoch": 0.5623354321694333,
+      "grad_norm": 1.1250942945480347,
+      "learning_rate": 8.496130652930818e-06,
+      "loss": 1.0974,
+      "step": 2456
+    },
+    {
+      "epoch": 0.562564396107613,
+      "grad_norm": 1.2028555870056152,
+      "learning_rate": 8.48879726810651e-06,
+      "loss": 1.0351,
+      "step": 2457
+    },
+    {
+      "epoch": 0.5627933600457928,
+      "grad_norm": 1.8054412603378296,
+      "learning_rate": 8.481464714884396e-06,
+      "loss": 1.0841,
+      "step": 2458
+    },
+    {
+      "epoch": 0.5630223239839726,
+      "grad_norm": 1.1389802694320679,
+      "learning_rate": 8.474132997299521e-06,
+      "loss": 1.0583,
+      "step": 2459
+    },
+    {
+      "epoch": 0.5632512879221523,
+      "grad_norm": 1.328384280204773,
+      "learning_rate": 8.466802119386462e-06,
+      "loss": 1.0583,
+      "step": 2460
+    },
+    {
+      "epoch": 0.563480251860332,
+      "grad_norm": 1.3099395036697388,
+      "learning_rate": 8.459472085179342e-06,
+      "loss": 1.0417,
+      "step": 2461
+    },
+    {
+      "epoch": 0.5637092157985117,
+      "grad_norm": 1.1930067539215088,
+      "learning_rate": 8.45214289871182e-06,
+      "loss": 1.0465,
+      "step": 2462
+    },
+    {
+      "epoch": 0.5639381797366915,
+      "grad_norm": 1.0933071374893188,
+      "learning_rate": 8.44481456401708e-06,
+      "loss": 1.0736,
+      "step": 2463
+    },
+    {
+      "epoch": 0.5641671436748712,
+      "grad_norm": 1.0359748601913452,
+      "learning_rate": 8.437487085127851e-06,
+      "loss": 1.0145,
+      "step": 2464
+    },
+    {
+      "epoch": 0.5643961076130509,
+      "grad_norm": 1.7698050737380981,
+      "learning_rate": 8.430160466076378e-06,
+      "loss": 1.0021,
+      "step": 2465
+    },
+    {
+      "epoch": 0.5646250715512307,
+      "grad_norm": 1.2962058782577515,
+      "learning_rate": 8.422834710894434e-06,
+      "loss": 1.05,
+      "step": 2466
+    },
+    {
+      "epoch": 0.5648540354894104,
+      "grad_norm": 1.083851933479309,
+      "learning_rate": 8.415509823613332e-06,
+      "loss": 0.9928,
+      "step": 2467
+    },
+    {
+      "epoch": 0.5650829994275901,
+      "grad_norm": 1.2015832662582397,
+      "learning_rate": 8.408185808263886e-06,
+      "loss": 1.0259,
+      "step": 2468
+    },
+    {
+      "epoch": 0.5653119633657699,
+      "grad_norm": 1.0598255395889282,
+      "learning_rate": 8.400862668876445e-06,
+      "loss": 1.0091,
+      "step": 2469
+    },
+    {
+      "epoch": 0.5655409273039497,
+      "grad_norm": 1.0813366174697876,
+      "learning_rate": 8.393540409480873e-06,
+      "loss": 1.0822,
+      "step": 2470
+    },
+    {
+      "epoch": 0.5657698912421294,
+      "grad_norm": 1.1584067344665527,
+      "learning_rate": 8.38621903410654e-06,
+      "loss": 1.0001,
+      "step": 2471
+    },
+    {
+      "epoch": 0.5659988551803091,
+      "grad_norm": 1.030510425567627,
+      "learning_rate": 8.378898546782344e-06,
+      "loss": 0.9853,
+      "step": 2472
+    },
+    {
+      "epoch": 0.5662278191184889,
+      "grad_norm": 1.317218542098999,
+      "learning_rate": 8.371578951536689e-06,
+      "loss": 1.0422,
+      "step": 2473
+    },
+    {
+      "epoch": 0.5664567830566686,
+      "grad_norm": 1.258917212486267,
+      "learning_rate": 8.364260252397483e-06,
+      "loss": 1.0164,
+      "step": 2474
+    },
+    {
+      "epoch": 0.5666857469948483,
+      "grad_norm": 1.1054089069366455,
+      "learning_rate": 8.356942453392147e-06,
+      "loss": 1.01,
+      "step": 2475
+    },
+    {
+      "epoch": 0.566914710933028,
+      "grad_norm": 1.110060691833496,
+      "learning_rate": 8.349625558547599e-06,
+      "loss": 1.0202,
+      "step": 2476
+    },
+    {
+      "epoch": 0.5671436748712078,
+      "grad_norm": 1.5175520181655884,
+      "learning_rate": 8.342309571890272e-06,
+      "loss": 1.0423,
+      "step": 2477
+    },
+    {
+      "epoch": 0.5673726388093875,
+      "grad_norm": 1.2970167398452759,
+      "learning_rate": 8.334994497446091e-06,
+      "loss": 1.1032,
+      "step": 2478
+    },
+    {
+      "epoch": 0.5676016027475672,
+      "grad_norm": 1.1656423807144165,
+      "learning_rate": 8.327680339240478e-06,
+      "loss": 1.0052,
+      "step": 2479
+    },
+    {
+      "epoch": 0.5678305666857469,
+      "grad_norm": 1.0043197870254517,
+      "learning_rate": 8.320367101298351e-06,
+      "loss": 1.0508,
+      "step": 2480
+    },
+    {
+      "epoch": 0.5680595306239268,
+      "grad_norm": 1.8934231996536255,
+      "learning_rate": 8.313054787644131e-06,
+      "loss": 1.0589,
+      "step": 2481
+    },
+    {
+      "epoch": 0.5682884945621065,
+      "grad_norm": 1.1924573183059692,
+      "learning_rate": 8.305743402301714e-06,
+      "loss": 1.0504,
+      "step": 2482
+    },
+    {
+      "epoch": 0.5685174585002862,
+      "grad_norm": 2.6312386989593506,
+      "learning_rate": 8.298432949294499e-06,
+      "loss": 0.9884,
+      "step": 2483
+    },
+    {
+      "epoch": 0.568746422438466,
+      "grad_norm": 1.0763286352157593,
+      "learning_rate": 8.29112343264537e-06,
+      "loss": 1.075,
+      "step": 2484
+    },
+    {
+      "epoch": 0.5689753863766457,
+      "grad_norm": 2.5799460411071777,
+      "learning_rate": 8.283814856376681e-06,
+      "loss": 1.0406,
+      "step": 2485
+    },
+    {
+      "epoch": 0.5692043503148254,
+      "grad_norm": 1.2226494550704956,
+      "learning_rate": 8.276507224510294e-06,
+      "loss": 1.1118,
+      "step": 2486
+    },
+    {
+      "epoch": 0.5694333142530051,
+      "grad_norm": 1.126726746559143,
+      "learning_rate": 8.26920054106753e-06,
+      "loss": 1.0838,
+      "step": 2487
+    },
+    {
+      "epoch": 0.5696622781911849,
+      "grad_norm": 1.5410242080688477,
+      "learning_rate": 8.261894810069197e-06,
+      "loss": 1.0668,
+      "step": 2488
+    },
+    {
+      "epoch": 0.5698912421293646,
+      "grad_norm": 1.1347763538360596,
+      "learning_rate": 8.25459003553558e-06,
+      "loss": 1.1072,
+      "step": 2489
+    },
+    {
+      "epoch": 0.5701202060675443,
+      "grad_norm": 1.095646858215332,
+      "learning_rate": 8.247286221486429e-06,
+      "loss": 1.0138,
+      "step": 2490
+    },
+    {
+      "epoch": 0.570349170005724,
+      "grad_norm": 1.7708925008773804,
+      "learning_rate": 8.239983371940975e-06,
+      "loss": 1.0268,
+      "step": 2491
+    },
+    {
+      "epoch": 0.5705781339439039,
+      "grad_norm": 1.2089495658874512,
+      "learning_rate": 8.232681490917919e-06,
+      "loss": 1.0141,
+      "step": 2492
+    },
+    {
+      "epoch": 0.5708070978820836,
+      "grad_norm": 1.059664249420166,
+      "learning_rate": 8.22538058243542e-06,
+      "loss": 1.0827,
+      "step": 2493
+    },
+    {
+      "epoch": 0.5710360618202633,
+      "grad_norm": 1.3616527318954468,
+      "learning_rate": 8.218080650511107e-06,
+      "loss": 1.0793,
+      "step": 2494
+    },
+    {
+      "epoch": 0.5712650257584431,
+      "grad_norm": 1.2198219299316406,
+      "learning_rate": 8.210781699162075e-06,
+      "loss": 1.0753,
+      "step": 2495
+    },
+    {
+      "epoch": 0.5714939896966228,
+      "grad_norm": 1.3348844051361084,
+      "learning_rate": 8.203483732404872e-06,
+      "loss": 1.0341,
+      "step": 2496
+    },
+    {
+      "epoch": 0.5717229536348025,
+      "grad_norm": 1.160372257232666,
+      "learning_rate": 8.196186754255509e-06,
+      "loss": 1.0313,
+      "step": 2497
+    },
+    {
+      "epoch": 0.5719519175729822,
+      "grad_norm": 1.0435636043548584,
+      "learning_rate": 8.188890768729452e-06,
+      "loss": 1.0094,
+      "step": 2498
+    },
+    {
+      "epoch": 0.572180881511162,
+      "grad_norm": 1.090612530708313,
+      "learning_rate": 8.181595779841618e-06,
+      "loss": 1.0399,
+      "step": 2499
+    },
+    {
+      "epoch": 0.5724098454493417,
+      "grad_norm": 1.496956467628479,
+      "learning_rate": 8.174301791606384e-06,
+      "loss": 0.9976,
+      "step": 2500
+    },
+    {
+      "epoch": 0.5726388093875214,
+      "grad_norm": 1.1636426448822021,
+      "learning_rate": 8.167008808037568e-06,
+      "loss": 1.0563,
+      "step": 2501
+    },
+    {
+      "epoch": 0.5728677733257012,
+      "grad_norm": 1.6248148679733276,
+      "learning_rate": 8.159716833148432e-06,
+      "loss": 1.0554,
+      "step": 2502
+    },
+    {
+      "epoch": 0.5730967372638809,
+      "grad_norm": 1.1249679327011108,
+      "learning_rate": 8.152425870951701e-06,
+      "loss": 1.0476,
+      "step": 2503
+    },
+    {
+      "epoch": 0.5733257012020607,
+      "grad_norm": 1.0898033380508423,
+      "learning_rate": 8.145135925459518e-06,
+      "loss": 1.1075,
+      "step": 2504
+    },
+    {
+      "epoch": 0.5735546651402404,
+      "grad_norm": 1.1172502040863037,
+      "learning_rate": 8.137847000683485e-06,
+      "loss": 1.0698,
+      "step": 2505
+    },
+    {
+      "epoch": 0.5737836290784202,
+      "grad_norm": 1.1264135837554932,
+      "learning_rate": 8.130559100634639e-06,
+      "loss": 1.0334,
+      "step": 2506
+    },
+    {
+      "epoch": 0.5740125930165999,
+      "grad_norm": 1.164330005645752,
+      "learning_rate": 8.123272229323443e-06,
+      "loss": 0.9811,
+      "step": 2507
+    },
+    {
+      "epoch": 0.5742415569547796,
+      "grad_norm": 1.1172728538513184,
+      "learning_rate": 8.115986390759805e-06,
+      "loss": 1.087,
+      "step": 2508
+    },
+    {
+      "epoch": 0.5744705208929594,
+      "grad_norm": 1.0987147092819214,
+      "learning_rate": 8.108701588953059e-06,
+      "loss": 1.1124,
+      "step": 2509
+    },
+    {
+      "epoch": 0.5746994848311391,
+      "grad_norm": 1.3992830514907837,
+      "learning_rate": 8.101417827911975e-06,
+      "loss": 1.0272,
+      "step": 2510
+    },
+    {
+      "epoch": 0.5749284487693188,
+      "grad_norm": 1.1009325981140137,
+      "learning_rate": 8.094135111644741e-06,
+      "loss": 1.0467,
+      "step": 2511
+    },
+    {
+      "epoch": 0.5751574127074985,
+      "grad_norm": 1.2195725440979004,
+      "learning_rate": 8.086853444158977e-06,
+      "loss": 1.0822,
+      "step": 2512
+    },
+    {
+      "epoch": 0.5753863766456783,
+      "grad_norm": 1.24541175365448,
+      "learning_rate": 8.07957282946172e-06,
+      "loss": 1.0975,
+      "step": 2513
+    },
+    {
+      "epoch": 0.575615340583858,
+      "grad_norm": 1.9513866901397705,
+      "learning_rate": 8.072293271559439e-06,
+      "loss": 1.0539,
+      "step": 2514
+    },
+    {
+      "epoch": 0.5758443045220378,
+      "grad_norm": 1.3148459196090698,
+      "learning_rate": 8.065014774458004e-06,
+      "loss": 1.0509,
+      "step": 2515
+    },
+    {
+      "epoch": 0.5760732684602176,
+      "grad_norm": 1.848658800125122,
+      "learning_rate": 8.05773734216272e-06,
+      "loss": 1.071,
+      "step": 2516
+    },
+    {
+      "epoch": 0.5763022323983973,
+      "grad_norm": 1.1349958181381226,
+      "learning_rate": 8.05046097867829e-06,
+      "loss": 1.0761,
+      "step": 2517
+    },
+    {
+      "epoch": 0.576531196336577,
+      "grad_norm": 1.2456731796264648,
+      "learning_rate": 8.043185688008837e-06,
+      "loss": 1.0541,
+      "step": 2518
+    },
+    {
+      "epoch": 0.5767601602747567,
+      "grad_norm": 1.2999017238616943,
+      "learning_rate": 8.035911474157897e-06,
+      "loss": 1.0678,
+      "step": 2519
+    },
+    {
+      "epoch": 0.5769891242129365,
+      "grad_norm": 1.1113868951797485,
+      "learning_rate": 8.028638341128405e-06,
+      "loss": 1.0296,
+      "step": 2520
+    },
+    {
+      "epoch": 0.5772180881511162,
+      "grad_norm": 1.351574182510376,
+      "learning_rate": 8.021366292922704e-06,
+      "loss": 0.9984,
+      "step": 2521
+    },
+    {
+      "epoch": 0.5774470520892959,
+      "grad_norm": 1.0689737796783447,
+      "learning_rate": 8.014095333542548e-06,
+      "loss": 1.0453,
+      "step": 2522
+    },
+    {
+      "epoch": 0.5776760160274756,
+      "grad_norm": 1.2941296100616455,
+      "learning_rate": 8.006825466989075e-06,
+      "loss": 1.0933,
+      "step": 2523
+    },
+    {
+      "epoch": 0.5779049799656554,
+      "grad_norm": 1.1887894868850708,
+      "learning_rate": 7.999556697262838e-06,
+      "loss": 0.9727,
+      "step": 2524
+    },
+    {
+      "epoch": 0.5781339439038351,
+      "grad_norm": 1.1151421070098877,
+      "learning_rate": 7.992289028363782e-06,
+      "loss": 1.037,
+      "step": 2525
+    },
+    {
+      "epoch": 0.5783629078420148,
+      "grad_norm": 1.1871955394744873,
+      "learning_rate": 7.985022464291236e-06,
+      "loss": 1.0871,
+      "step": 2526
+    },
+    {
+      "epoch": 0.5785918717801947,
+      "grad_norm": 1.3512312173843384,
+      "learning_rate": 7.977757009043941e-06,
+      "loss": 1.0276,
+      "step": 2527
+    },
+    {
+      "epoch": 0.5788208357183744,
+      "grad_norm": 1.1984821557998657,
+      "learning_rate": 7.970492666620002e-06,
+      "loss": 1.0461,
+      "step": 2528
+    },
+    {
+      "epoch": 0.5790497996565541,
+      "grad_norm": 1.1962758302688599,
+      "learning_rate": 7.963229441016938e-06,
+      "loss": 1.0625,
+      "step": 2529
+    },
+    {
+      "epoch": 0.5792787635947338,
+      "grad_norm": 1.174294352531433,
+      "learning_rate": 7.955967336231635e-06,
+      "loss": 0.9852,
+      "step": 2530
+    },
+    {
+      "epoch": 0.5795077275329136,
+      "grad_norm": 1.1075918674468994,
+      "learning_rate": 7.948706356260367e-06,
+      "loss": 1.1239,
+      "step": 2531
+    },
+    {
+      "epoch": 0.5797366914710933,
+      "grad_norm": 1.2722827196121216,
+      "learning_rate": 7.941446505098795e-06,
+      "loss": 1.0525,
+      "step": 2532
+    },
+    {
+      "epoch": 0.579965655409273,
+      "grad_norm": 1.085821509361267,
+      "learning_rate": 7.934187786741956e-06,
+      "loss": 1.0824,
+      "step": 2533
+    },
+    {
+      "epoch": 0.5801946193474528,
+      "grad_norm": 1.2172048091888428,
+      "learning_rate": 7.926930205184254e-06,
+      "loss": 1.0138,
+      "step": 2534
+    },
+    {
+      "epoch": 0.5804235832856325,
+      "grad_norm": 1.2068073749542236,
+      "learning_rate": 7.919673764419479e-06,
+      "loss": 1.0394,
+      "step": 2535
+    },
+    {
+      "epoch": 0.5806525472238122,
+      "grad_norm": 1.9181028604507446,
+      "learning_rate": 7.912418468440794e-06,
+      "loss": 1.0098,
+      "step": 2536
+    },
+    {
+      "epoch": 0.5808815111619919,
+      "grad_norm": 2.7652781009674072,
+      "learning_rate": 7.90516432124072e-06,
+      "loss": 1.1164,
+      "step": 2537
+    },
+    {
+      "epoch": 0.5811104751001718,
+      "grad_norm": 1.5410016775131226,
+      "learning_rate": 7.89791132681116e-06,
+      "loss": 1.0504,
+      "step": 2538
+    },
+    {
+      "epoch": 0.5813394390383515,
+      "grad_norm": 1.0692219734191895,
+      "learning_rate": 7.89065948914337e-06,
+      "loss": 1.0947,
+      "step": 2539
+    },
+    {
+      "epoch": 0.5815684029765312,
+      "grad_norm": 1.367213249206543,
+      "learning_rate": 7.883408812227977e-06,
+      "loss": 1.0337,
+      "step": 2540
+    },
+    {
+      "epoch": 0.581797366914711,
+      "grad_norm": 1.8254352807998657,
+      "learning_rate": 7.876159300054974e-06,
+      "loss": 1.0041,
+      "step": 2541
+    },
+    {
+      "epoch": 0.5820263308528907,
+      "grad_norm": 1.1613630056381226,
+      "learning_rate": 7.868910956613697e-06,
+      "loss": 1.0489,
+      "step": 2542
+    },
+    {
+      "epoch": 0.5822552947910704,
+      "grad_norm": 1.2703049182891846,
+      "learning_rate": 7.861663785892857e-06,
+      "loss": 1.0312,
+      "step": 2543
+    },
+    {
+      "epoch": 0.5824842587292501,
+      "grad_norm": 1.263065218925476,
+      "learning_rate": 7.854417791880508e-06,
+      "loss": 1.031,
+      "step": 2544
+    },
+    {
+      "epoch": 0.5827132226674299,
+      "grad_norm": 1.1359111070632935,
+      "learning_rate": 7.847172978564055e-06,
+      "loss": 1.0146,
+      "step": 2545
+    },
+    {
+      "epoch": 0.5829421866056096,
+      "grad_norm": 1.2447282075881958,
+      "learning_rate": 7.839929349930266e-06,
+      "loss": 1.045,
+      "step": 2546
+    },
+    {
+      "epoch": 0.5831711505437893,
+      "grad_norm": 1.083162546157837,
+      "learning_rate": 7.832686909965248e-06,
+      "loss": 1.0664,
+      "step": 2547
+    },
+    {
+      "epoch": 0.583400114481969,
+      "grad_norm": 1.1830934286117554,
+      "learning_rate": 7.82544566265445e-06,
+      "loss": 0.9864,
+      "step": 2548
+    },
+    {
+      "epoch": 0.5836290784201488,
+      "grad_norm": 1.7827184200286865,
+      "learning_rate": 7.818205611982673e-06,
+      "loss": 1.0378,
+      "step": 2549
+    },
+    {
+      "epoch": 0.5838580423583286,
+      "grad_norm": 1.157387614250183,
+      "learning_rate": 7.810966761934053e-06,
+      "loss": 1.033,
+      "step": 2550
+    },
+    {
+      "epoch": 0.5840870062965083,
+      "grad_norm": 1.540801763534546,
+      "learning_rate": 7.803729116492072e-06,
+      "loss": 1.0447,
+      "step": 2551
+    },
+    {
+      "epoch": 0.5843159702346881,
+      "grad_norm": 1.1215074062347412,
+      "learning_rate": 7.796492679639549e-06,
+      "loss": 1.0606,
+      "step": 2552
+    },
+    {
+      "epoch": 0.5845449341728678,
+      "grad_norm": 1.2133134603500366,
+      "learning_rate": 7.789257455358625e-06,
+      "loss": 1.0275,
+      "step": 2553
+    },
+    {
+      "epoch": 0.5847738981110475,
+      "grad_norm": 1.3863023519515991,
+      "learning_rate": 7.782023447630789e-06,
+      "loss": 1.0156,
+      "step": 2554
+    },
+    {
+      "epoch": 0.5850028620492272,
+      "grad_norm": 1.131239891052246,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.996,
+      "step": 2555
+    },
+    {
+      "epoch": 0.585231825987407,
+      "grad_norm": 1.179404377937317,
+      "learning_rate": 7.767559097756966e-06,
+      "loss": 1.0682,
+      "step": 2556
+    },
+    {
+      "epoch": 0.5854607899255867,
+      "grad_norm": 2.166602373123169,
+      "learning_rate": 7.760328763570589e-06,
+      "loss": 1.0358,
+      "step": 2557
+    },
+    {
+      "epoch": 0.5856897538637664,
+      "grad_norm": 1.535658836364746,
+      "learning_rate": 7.753099661856516e-06,
+      "loss": 1.0759,
+      "step": 2558
+    },
+    {
+      "epoch": 0.5859187178019462,
+      "grad_norm": 1.1416999101638794,
+      "learning_rate": 7.745871796592857e-06,
+      "loss": 1.0636,
+      "step": 2559
+    },
+    {
+      "epoch": 0.5861476817401259,
+      "grad_norm": 1.214870572090149,
+      "learning_rate": 7.738645171757054e-06,
+      "loss": 1.0911,
+      "step": 2560
+    },
+    {
+      "epoch": 0.5863766456783057,
+      "grad_norm": 1.7018429040908813,
+      "learning_rate": 7.731419791325852e-06,
+      "loss": 1.0343,
+      "step": 2561
+    },
+    {
+      "epoch": 0.5866056096164854,
+      "grad_norm": 1.0340094566345215,
+      "learning_rate": 7.72419565927532e-06,
+      "loss": 1.0418,
+      "step": 2562
+    },
+    {
+      "epoch": 0.5868345735546652,
+      "grad_norm": 1.4168850183486938,
+      "learning_rate": 7.716972779580836e-06,
+      "loss": 1.0363,
+      "step": 2563
+    },
+    {
+      "epoch": 0.5870635374928449,
+      "grad_norm": 1.3030794858932495,
+      "learning_rate": 7.709751156217088e-06,
+      "loss": 1.0039,
+      "step": 2564
+    },
+    {
+      "epoch": 0.5872925014310246,
+      "grad_norm": 1.357153296470642,
+      "learning_rate": 7.702530793158079e-06,
+      "loss": 1.1281,
+      "step": 2565
+    },
+    {
+      "epoch": 0.5875214653692044,
+      "grad_norm": 1.1770304441452026,
+      "learning_rate": 7.695311694377116e-06,
+      "loss": 1.037,
+      "step": 2566
+    },
+    {
+      "epoch": 0.5877504293073841,
+      "grad_norm": 1.2851818799972534,
+      "learning_rate": 7.688093863846801e-06,
+      "loss": 1.0687,
+      "step": 2567
+    },
+    {
+      "epoch": 0.5879793932455638,
+      "grad_norm": 1.1500670909881592,
+      "learning_rate": 7.680877305539048e-06,
+      "loss": 1.0786,
+      "step": 2568
+    },
+    {
+      "epoch": 0.5882083571837435,
+      "grad_norm": 1.660108208656311,
+      "learning_rate": 7.673662023425074e-06,
+      "loss": 1.0077,
+      "step": 2569
+    },
+    {
+      "epoch": 0.5884373211219233,
+      "grad_norm": 1.0523499250411987,
+      "learning_rate": 7.666448021475385e-06,
+      "loss": 1.0091,
+      "step": 2570
+    },
+    {
+      "epoch": 0.588666285060103,
+      "grad_norm": 1.180342197418213,
+      "learning_rate": 7.659235303659784e-06,
+      "loss": 1.0668,
+      "step": 2571
+    },
+    {
+      "epoch": 0.5888952489982827,
+      "grad_norm": 1.269478440284729,
+      "learning_rate": 7.65202387394737e-06,
+      "loss": 1.0988,
+      "step": 2572
+    },
+    {
+      "epoch": 0.5891242129364626,
+      "grad_norm": 1.2310328483581543,
+      "learning_rate": 7.644813736306531e-06,
+      "loss": 1.0087,
+      "step": 2573
+    },
+    {
+      "epoch": 0.5893531768746423,
+      "grad_norm": 1.2201181650161743,
+      "learning_rate": 7.637604894704951e-06,
+      "loss": 1.097,
+      "step": 2574
+    },
+    {
+      "epoch": 0.589582140812822,
+      "grad_norm": 1.19076406955719,
+      "learning_rate": 7.630397353109588e-06,
+      "loss": 1.0493,
+      "step": 2575
+    },
+    {
+      "epoch": 0.5898111047510017,
+      "grad_norm": 1.3227699995040894,
+      "learning_rate": 7.623191115486695e-06,
+      "loss": 1.0598,
+      "step": 2576
+    },
+    {
+      "epoch": 0.5900400686891815,
+      "grad_norm": 1.1145106554031372,
+      "learning_rate": 7.615986185801807e-06,
+      "loss": 1.0267,
+      "step": 2577
+    },
+    {
+      "epoch": 0.5902690326273612,
+      "grad_norm": 1.8626641035079956,
+      "learning_rate": 7.608782568019729e-06,
+      "loss": 1.0034,
+      "step": 2578
+    },
+    {
+      "epoch": 0.5904979965655409,
+      "grad_norm": 1.3820207118988037,
+      "learning_rate": 7.601580266104558e-06,
+      "loss": 1.0034,
+      "step": 2579
+    },
+    {
+      "epoch": 0.5907269605037206,
+      "grad_norm": 1.131319522857666,
+      "learning_rate": 7.594379284019659e-06,
+      "loss": 1.0725,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5909559244419004,
+      "grad_norm": 1.295155644416809,
+      "learning_rate": 7.587179625727671e-06,
+      "loss": 1.0437,
+      "step": 2581
+    },
+    {
+      "epoch": 0.5911848883800801,
+      "grad_norm": 1.355495572090149,
+      "learning_rate": 7.579981295190506e-06,
+      "loss": 1.0339,
+      "step": 2582
+    },
+    {
+      "epoch": 0.5914138523182598,
+      "grad_norm": 1.4129412174224854,
+      "learning_rate": 7.572784296369342e-06,
+      "loss": 1.0228,
+      "step": 2583
+    },
+    {
+      "epoch": 0.5916428162564397,
+      "grad_norm": 1.3346372842788696,
+      "learning_rate": 7.565588633224632e-06,
+      "loss": 1.0627,
+      "step": 2584
+    },
+    {
+      "epoch": 0.5918717801946194,
+      "grad_norm": 1.4779704809188843,
+      "learning_rate": 7.558394309716088e-06,
+      "loss": 1.1153,
+      "step": 2585
+    },
+    {
+      "epoch": 0.5921007441327991,
+      "grad_norm": 1.397422194480896,
+      "learning_rate": 7.551201329802684e-06,
+      "loss": 1.0255,
+      "step": 2586
+    },
+    {
+      "epoch": 0.5923297080709788,
+      "grad_norm": 1.3591479063034058,
+      "learning_rate": 7.544009697442656e-06,
+      "loss": 1.0635,
+      "step": 2587
+    },
+    {
+      "epoch": 0.5925586720091586,
+      "grad_norm": 1.7575359344482422,
+      "learning_rate": 7.536819416593504e-06,
+      "loss": 1.0804,
+      "step": 2588
+    },
+    {
+      "epoch": 0.5927876359473383,
+      "grad_norm": 1.3760896921157837,
+      "learning_rate": 7.529630491211972e-06,
+      "loss": 1.0748,
+      "step": 2589
+    },
+    {
+      "epoch": 0.593016599885518,
+      "grad_norm": 1.791036605834961,
+      "learning_rate": 7.522442925254068e-06,
+      "loss": 1.0705,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5932455638236978,
+      "grad_norm": 1.271471619606018,
+      "learning_rate": 7.515256722675058e-06,
+      "loss": 1.0509,
+      "step": 2591
+    },
+    {
+      "epoch": 0.5934745277618775,
+      "grad_norm": 1.1143954992294312,
+      "learning_rate": 7.508071887429433e-06,
+      "loss": 1.0352,
+      "step": 2592
+    },
+    {
+      "epoch": 0.5937034917000572,
+      "grad_norm": 1.292467474937439,
+      "learning_rate": 7.500888423470962e-06,
+      "loss": 1.013,
+      "step": 2593
+    },
+    {
+      "epoch": 0.5939324556382369,
+      "grad_norm": 1.4147076606750488,
+      "learning_rate": 7.493706334752637e-06,
+      "loss": 1.1661,
+      "step": 2594
+    },
+    {
+      "epoch": 0.5941614195764167,
+      "grad_norm": 1.6669336557388306,
+      "learning_rate": 7.486525625226699e-06,
+      "loss": 1.0527,
+      "step": 2595
+    },
+    {
+      "epoch": 0.5943903835145965,
+      "grad_norm": 1.3241775035858154,
+      "learning_rate": 7.479346298844645e-06,
+      "loss": 1.0268,
+      "step": 2596
+    },
+    {
+      "epoch": 0.5946193474527762,
+      "grad_norm": 1.1176531314849854,
+      "learning_rate": 7.472168359557183e-06,
+      "loss": 1.0257,
+      "step": 2597
+    },
+    {
+      "epoch": 0.594848311390956,
+      "grad_norm": 1.194197654724121,
+      "learning_rate": 7.46499181131428e-06,
+      "loss": 1.073,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5950772753291357,
+      "grad_norm": 1.1093398332595825,
+      "learning_rate": 7.4578166580651335e-06,
+      "loss": 0.9844,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5953062392673154,
+      "grad_norm": 1.3040740489959717,
+      "learning_rate": 7.450642903758163e-06,
+      "loss": 1.1006,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5955352032054951,
+      "grad_norm": 1.1911894083023071,
+      "learning_rate": 7.443470552341028e-06,
+      "loss": 1.1261,
+      "step": 2601
+    },
+    {
+      "epoch": 0.5957641671436749,
+      "grad_norm": 1.2414566278457642,
+      "learning_rate": 7.436299607760616e-06,
+      "loss": 1.0389,
+      "step": 2602
+    },
+    {
+      "epoch": 0.5959931310818546,
+      "grad_norm": 1.163601040840149,
+      "learning_rate": 7.429130073963036e-06,
+      "loss": 0.9994,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5962220950200343,
+      "grad_norm": 1.274039626121521,
+      "learning_rate": 7.421961954893622e-06,
+      "loss": 1.0788,
+      "step": 2604
+    },
+    {
+      "epoch": 0.596451058958214,
+      "grad_norm": 1.1208449602127075,
+      "learning_rate": 7.414795254496929e-06,
+      "loss": 1.032,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5966800228963938,
+      "grad_norm": 1.4800190925598145,
+      "learning_rate": 7.4076299767167325e-06,
+      "loss": 1.0526,
+      "step": 2606
+    },
+    {
+      "epoch": 0.5969089868345736,
+      "grad_norm": 1.2557833194732666,
+      "learning_rate": 7.400466125496027e-06,
+      "loss": 1.1016,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5971379507727533,
+      "grad_norm": 1.3543636798858643,
+      "learning_rate": 7.393303704777017e-06,
+      "loss": 1.0346,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5973669147109331,
+      "grad_norm": 1.2229251861572266,
+      "learning_rate": 7.386142718501122e-06,
+      "loss": 1.0234,
+      "step": 2609
+    },
+    {
+      "epoch": 0.5975958786491128,
+      "grad_norm": 1.217532753944397,
+      "learning_rate": 7.378983170608982e-06,
+      "loss": 1.039,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5978248425872925,
+      "grad_norm": 1.3155670166015625,
+      "learning_rate": 7.37182506504042e-06,
+      "loss": 1.0595,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5980538065254722,
+      "grad_norm": 1.3033359050750732,
+      "learning_rate": 7.364668405734493e-06,
+      "loss": 1.0695,
+      "step": 2612
+    },
+    {
+      "epoch": 0.598282770463652,
+      "grad_norm": 1.1440081596374512,
+      "learning_rate": 7.35751319662945e-06,
+      "loss": 1.0842,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5985117344018317,
+      "grad_norm": 1.1137559413909912,
+      "learning_rate": 7.350359441662735e-06,
+      "loss": 1.007,
+      "step": 2614
+    },
+    {
+      "epoch": 0.5987406983400114,
+      "grad_norm": 1.325477957725525,
+      "learning_rate": 7.343207144771012e-06,
+      "loss": 1.0835,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5989696622781912,
+      "grad_norm": 1.1493194103240967,
+      "learning_rate": 7.336056309890116e-06,
+      "loss": 1.0678,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5991986262163709,
+      "grad_norm": 1.2086422443389893,
+      "learning_rate": 7.3289069409551e-06,
+      "loss": 1.1063,
+      "step": 2617
+    },
+    {
+      "epoch": 0.5994275901545506,
+      "grad_norm": 1.3805280923843384,
+      "learning_rate": 7.321759041900204e-06,
+      "loss": 1.0846,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5996565540927304,
+      "grad_norm": 1.2356138229370117,
+      "learning_rate": 7.314612616658847e-06,
+      "loss": 1.0757,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5998855180309102,
+      "grad_norm": 1.2783619165420532,
+      "learning_rate": 7.307467669163655e-06,
+      "loss": 1.0257,
+      "step": 2620
+    },
+    {
+      "epoch": 0.6001144819690899,
+      "grad_norm": 1.2165898084640503,
+      "learning_rate": 7.3003242033464314e-06,
+      "loss": 1.0641,
+      "step": 2621
+    },
+    {
+      "epoch": 0.6003434459072696,
+      "grad_norm": 1.589637279510498,
+      "learning_rate": 7.293182223138164e-06,
+      "loss": 1.1101,
+      "step": 2622
+    },
+    {
+      "epoch": 0.6005724098454494,
+      "grad_norm": 1.0955222845077515,
+      "learning_rate": 7.286041732469024e-06,
+      "loss": 1.0198,
+      "step": 2623
+    },
+    {
+      "epoch": 0.6008013737836291,
+      "grad_norm": 1.270451307296753,
+      "learning_rate": 7.278902735268367e-06,
+      "loss": 1.0718,
+      "step": 2624
+    },
+    {
+      "epoch": 0.6010303377218088,
+      "grad_norm": 1.114689826965332,
+      "learning_rate": 7.27176523546472e-06,
+      "loss": 1.0091,
+      "step": 2625
+    },
+    {
+      "epoch": 0.6012593016599885,
+      "grad_norm": 1.1657631397247314,
+      "learning_rate": 7.264629236985792e-06,
+      "loss": 1.1009,
+      "step": 2626
+    },
+    {
+      "epoch": 0.6014882655981683,
+      "grad_norm": 1.2354736328125,
+      "learning_rate": 7.25749474375846e-06,
+      "loss": 1.0566,
+      "step": 2627
+    },
+    {
+      "epoch": 0.601717229536348,
+      "grad_norm": 1.1252378225326538,
+      "learning_rate": 7.250361759708775e-06,
+      "loss": 1.038,
+      "step": 2628
+    },
+    {
+      "epoch": 0.6019461934745277,
+      "grad_norm": 1.6679290533065796,
+      "learning_rate": 7.243230288761966e-06,
+      "loss": 1.0852,
+      "step": 2629
+    },
+    {
+      "epoch": 0.6021751574127076,
+      "grad_norm": 1.029215693473816,
+      "learning_rate": 7.2361003348424105e-06,
+      "loss": 1.0707,
+      "step": 2630
+    },
+    {
+      "epoch": 0.6024041213508873,
+      "grad_norm": 1.261576771736145,
+      "learning_rate": 7.2289719018736715e-06,
+      "loss": 0.9972,
+      "step": 2631
+    },
+    {
+      "epoch": 0.602633085289067,
+      "grad_norm": 1.4089431762695312,
+      "learning_rate": 7.221844993778464e-06,
+      "loss": 1.0427,
+      "step": 2632
+    },
+    {
+      "epoch": 0.6028620492272467,
+      "grad_norm": 1.418616771697998,
+      "learning_rate": 7.21471961447866e-06,
+      "loss": 1.1127,
+      "step": 2633
+    },
+    {
+      "epoch": 0.6030910131654265,
+      "grad_norm": 1.258452296257019,
+      "learning_rate": 7.207595767895303e-06,
+      "loss": 1.0767,
+      "step": 2634
+    },
+    {
+      "epoch": 0.6033199771036062,
+      "grad_norm": 1.0797663927078247,
+      "learning_rate": 7.200473457948584e-06,
+      "loss": 1.0761,
+      "step": 2635
+    },
+    {
+      "epoch": 0.6035489410417859,
+      "grad_norm": 1.2073185443878174,
+      "learning_rate": 7.193352688557849e-06,
+      "loss": 1.0367,
+      "step": 2636
+    },
+    {
+      "epoch": 0.6037779049799656,
+      "grad_norm": 1.2464630603790283,
+      "learning_rate": 7.186233463641601e-06,
+      "loss": 1.0441,
+      "step": 2637
+    },
+    {
+      "epoch": 0.6040068689181454,
+      "grad_norm": 2.0551130771636963,
+      "learning_rate": 7.179115787117482e-06,
+      "loss": 1.0626,
+      "step": 2638
+    },
+    {
+      "epoch": 0.6042358328563251,
+      "grad_norm": 1.3895354270935059,
+      "learning_rate": 7.171999662902298e-06,
+      "loss": 1.0775,
+      "step": 2639
+    },
+    {
+      "epoch": 0.6044647967945048,
+      "grad_norm": 1.0203369855880737,
+      "learning_rate": 7.164885094911991e-06,
+      "loss": 1.0592,
+      "step": 2640
+    },
+    {
+      "epoch": 0.6046937607326845,
+      "grad_norm": 1.9111573696136475,
+      "learning_rate": 7.157772087061645e-06,
+      "loss": 1.0904,
+      "step": 2641
+    },
+    {
+      "epoch": 0.6049227246708644,
+      "grad_norm": 1.0934786796569824,
+      "learning_rate": 7.150660643265488e-06,
+      "loss": 1.0466,
+      "step": 2642
+    },
+    {
+      "epoch": 0.6051516886090441,
+      "grad_norm": 1.2626543045043945,
+      "learning_rate": 7.143550767436894e-06,
+      "loss": 1.0527,
+      "step": 2643
+    },
+    {
+      "epoch": 0.6053806525472238,
+      "grad_norm": 1.443155288696289,
+      "learning_rate": 7.136442463488362e-06,
+      "loss": 1.0411,
+      "step": 2644
+    },
+    {
+      "epoch": 0.6056096164854036,
+      "grad_norm": 1.5223894119262695,
+      "learning_rate": 7.129335735331537e-06,
+      "loss": 1.0087,
+      "step": 2645
+    },
+    {
+      "epoch": 0.6058385804235833,
+      "grad_norm": 1.1826412677764893,
+      "learning_rate": 7.122230586877188e-06,
+      "loss": 1.0613,
+      "step": 2646
+    },
+    {
+      "epoch": 0.606067544361763,
+      "grad_norm": 3.0586001873016357,
+      "learning_rate": 7.115127022035215e-06,
+      "loss": 1.0597,
+      "step": 2647
+    },
+    {
+      "epoch": 0.6062965082999427,
+      "grad_norm": 1.441426157951355,
+      "learning_rate": 7.108025044714661e-06,
+      "loss": 1.1143,
+      "step": 2648
+    },
+    {
+      "epoch": 0.6065254722381225,
+      "grad_norm": 1.2049249410629272,
+      "learning_rate": 7.100924658823677e-06,
+      "loss": 1.082,
+      "step": 2649
+    },
+    {
+      "epoch": 0.6067544361763022,
+      "grad_norm": 1.1801356077194214,
+      "learning_rate": 7.093825868269546e-06,
+      "loss": 1.0233,
+      "step": 2650
+    },
+    {
+      "epoch": 0.6069834001144819,
+      "grad_norm": 1.2054147720336914,
+      "learning_rate": 7.0867286769586775e-06,
+      "loss": 1.015,
+      "step": 2651
+    },
+    {
+      "epoch": 0.6072123640526617,
+      "grad_norm": 3.0669796466827393,
+      "learning_rate": 7.0796330887965884e-06,
+      "loss": 1.0719,
+      "step": 2652
+    },
+    {
+      "epoch": 0.6074413279908415,
+      "grad_norm": 1.0845890045166016,
+      "learning_rate": 7.072539107687928e-06,
+      "loss": 1.0525,
+      "step": 2653
+    },
+    {
+      "epoch": 0.6076702919290212,
+      "grad_norm": 1.1807281970977783,
+      "learning_rate": 7.065446737536455e-06,
+      "loss": 1.0256,
+      "step": 2654
+    },
+    {
+      "epoch": 0.607899255867201,
+      "grad_norm": 1.2322497367858887,
+      "learning_rate": 7.058355982245038e-06,
+      "loss": 1.0678,
+      "step": 2655
+    },
+    {
+      "epoch": 0.6081282198053807,
+      "grad_norm": 1.065679907798767,
+      "learning_rate": 7.051266845715663e-06,
+      "loss": 1.0384,
+      "step": 2656
+    },
+    {
+      "epoch": 0.6083571837435604,
+      "grad_norm": 1.2442513704299927,
+      "learning_rate": 7.044179331849415e-06,
+      "loss": 1.0565,
+      "step": 2657
+    },
+    {
+      "epoch": 0.6085861476817401,
+      "grad_norm": 1.1693778038024902,
+      "learning_rate": 7.0370934445465026e-06,
+      "loss": 0.9849,
+      "step": 2658
+    },
+    {
+      "epoch": 0.6088151116199199,
+      "grad_norm": 1.1018726825714111,
+      "learning_rate": 7.030009187706228e-06,
+      "loss": 1.0543,
+      "step": 2659
+    },
+    {
+      "epoch": 0.6090440755580996,
+      "grad_norm": 1.5267049074172974,
+      "learning_rate": 7.022926565226995e-06,
+      "loss": 1.0282,
+      "step": 2660
+    },
+    {
+      "epoch": 0.6092730394962793,
+      "grad_norm": 1.2858092784881592,
+      "learning_rate": 7.015845581006311e-06,
+      "loss": 0.991,
+      "step": 2661
+    },
+    {
+      "epoch": 0.609502003434459,
+      "grad_norm": 1.529943585395813,
+      "learning_rate": 7.00876623894079e-06,
+      "loss": 1.0248,
+      "step": 2662
+    },
+    {
+      "epoch": 0.6097309673726388,
+      "grad_norm": 1.2061338424682617,
+      "learning_rate": 7.001688542926126e-06,
+      "loss": 1.0418,
+      "step": 2663
+    },
+    {
+      "epoch": 0.6099599313108185,
+      "grad_norm": 1.2191417217254639,
+      "learning_rate": 6.994612496857118e-06,
+      "loss": 1.0346,
+      "step": 2664
+    },
+    {
+      "epoch": 0.6101888952489983,
+      "grad_norm": 1.3246591091156006,
+      "learning_rate": 6.9875381046276605e-06,
+      "loss": 1.051,
+      "step": 2665
+    },
+    {
+      "epoch": 0.6104178591871781,
+      "grad_norm": 1.1738966703414917,
+      "learning_rate": 6.9804653701307225e-06,
+      "loss": 1.073,
+      "step": 2666
+    },
+    {
+      "epoch": 0.6106468231253578,
+      "grad_norm": 1.06193208694458,
+      "learning_rate": 6.973394297258378e-06,
+      "loss": 1.0468,
+      "step": 2667
+    },
+    {
+      "epoch": 0.6108757870635375,
+      "grad_norm": 1.585971713066101,
+      "learning_rate": 6.966324889901776e-06,
+      "loss": 1.0083,
+      "step": 2668
+    },
+    {
+      "epoch": 0.6111047510017172,
+      "grad_norm": 1.149452805519104,
+      "learning_rate": 6.959257151951153e-06,
+      "loss": 1.0624,
+      "step": 2669
+    },
+    {
+      "epoch": 0.611333714939897,
+      "grad_norm": 1.240146279335022,
+      "learning_rate": 6.952191087295827e-06,
+      "loss": 1.0515,
+      "step": 2670
+    },
+    {
+      "epoch": 0.6115626788780767,
+      "grad_norm": 1.2787004709243774,
+      "learning_rate": 6.945126699824189e-06,
+      "loss": 0.9782,
+      "step": 2671
+    },
+    {
+      "epoch": 0.6117916428162564,
+      "grad_norm": 1.338744878768921,
+      "learning_rate": 6.938063993423718e-06,
+      "loss": 1.0323,
+      "step": 2672
+    },
+    {
+      "epoch": 0.6120206067544361,
+      "grad_norm": 2.4173684120178223,
+      "learning_rate": 6.9310029719809625e-06,
+      "loss": 1.022,
+      "step": 2673
+    },
+    {
+      "epoch": 0.6122495706926159,
+      "grad_norm": 1.3075268268585205,
+      "learning_rate": 6.923943639381539e-06,
+      "loss": 1.0476,
+      "step": 2674
+    },
+    {
+      "epoch": 0.6124785346307956,
+      "grad_norm": 1.0469852685928345,
+      "learning_rate": 6.916885999510137e-06,
+      "loss": 0.9744,
+      "step": 2675
+    },
+    {
+      "epoch": 0.6127074985689754,
+      "grad_norm": 1.373605489730835,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 1.0191,
+      "step": 2676
+    },
+    {
+      "epoch": 0.6129364625071552,
+      "grad_norm": 1.5057294368743896,
+      "learning_rate": 6.902775813485527e-06,
+      "loss": 1.061,
+      "step": 2677
+    },
+    {
+      "epoch": 0.6131654264453349,
+      "grad_norm": 1.0761992931365967,
+      "learning_rate": 6.895723275097031e-06,
+      "loss": 1.0406,
+      "step": 2678
+    },
+    {
+      "epoch": 0.6133943903835146,
+      "grad_norm": 1.1914690732955933,
+      "learning_rate": 6.888672444965988e-06,
+      "loss": 1.0442,
+      "step": 2679
+    },
+    {
+      "epoch": 0.6136233543216943,
+      "grad_norm": 1.2478915452957153,
+      "learning_rate": 6.881623326972412e-06,
+      "loss": 0.9698,
+      "step": 2680
+    },
+    {
+      "epoch": 0.6138523182598741,
+      "grad_norm": 1.6330108642578125,
+      "learning_rate": 6.874575924995378e-06,
+      "loss": 1.0203,
+      "step": 2681
+    },
+    {
+      "epoch": 0.6140812821980538,
+      "grad_norm": 1.1003386974334717,
+      "learning_rate": 6.867530242913008e-06,
+      "loss": 1.0006,
+      "step": 2682
+    },
+    {
+      "epoch": 0.6143102461362335,
+      "grad_norm": 1.3207449913024902,
+      "learning_rate": 6.860486284602479e-06,
+      "loss": 1.1075,
+      "step": 2683
+    },
+    {
+      "epoch": 0.6145392100744133,
+      "grad_norm": 1.2823762893676758,
+      "learning_rate": 6.853444053940034e-06,
+      "loss": 0.9907,
+      "step": 2684
+    },
+    {
+      "epoch": 0.614768174012593,
+      "grad_norm": 1.167484164237976,
+      "learning_rate": 6.846403554800938e-06,
+      "loss": 1.0652,
+      "step": 2685
+    },
+    {
+      "epoch": 0.6149971379507727,
+      "grad_norm": 1.0674535036087036,
+      "learning_rate": 6.839364791059529e-06,
+      "loss": 1.0155,
+      "step": 2686
+    },
+    {
+      "epoch": 0.6152261018889524,
+      "grad_norm": 1.2486037015914917,
+      "learning_rate": 6.832327766589177e-06,
+      "loss": 1.1036,
+      "step": 2687
+    },
+    {
+      "epoch": 0.6154550658271323,
+      "grad_norm": 1.2341771125793457,
+      "learning_rate": 6.825292485262296e-06,
+      "loss": 1.0722,
+      "step": 2688
+    },
+    {
+      "epoch": 0.615684029765312,
+      "grad_norm": 1.7849056720733643,
+      "learning_rate": 6.818258950950346e-06,
+      "loss": 1.0679,
+      "step": 2689
+    },
+    {
+      "epoch": 0.6159129937034917,
+      "grad_norm": 1.1976414918899536,
+      "learning_rate": 6.8112271675238154e-06,
+      "loss": 1.0202,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6161419576416715,
+      "grad_norm": 1.1759425401687622,
+      "learning_rate": 6.804197138852242e-06,
+      "loss": 1.1098,
+      "step": 2691
+    },
+    {
+      "epoch": 0.6163709215798512,
+      "grad_norm": 1.9811475276947021,
+      "learning_rate": 6.797168868804192e-06,
+      "loss": 1.0115,
+      "step": 2692
+    },
+    {
+      "epoch": 0.6165998855180309,
+      "grad_norm": 1.4365376234054565,
+      "learning_rate": 6.790142361247258e-06,
+      "loss": 1.0826,
+      "step": 2693
+    },
+    {
+      "epoch": 0.6168288494562106,
+      "grad_norm": 0.9954733848571777,
+      "learning_rate": 6.7831176200480686e-06,
+      "loss": 1.024,
+      "step": 2694
+    },
+    {
+      "epoch": 0.6170578133943904,
+      "grad_norm": 1.2109805345535278,
+      "learning_rate": 6.776094649072286e-06,
+      "loss": 1.0394,
+      "step": 2695
+    },
+    {
+      "epoch": 0.6172867773325701,
+      "grad_norm": 1.5378385782241821,
+      "learning_rate": 6.769073452184589e-06,
+      "loss": 1.0604,
+      "step": 2696
+    },
+    {
+      "epoch": 0.6175157412707498,
+      "grad_norm": 1.6136606931686401,
+      "learning_rate": 6.762054033248681e-06,
+      "loss": 1.0747,
+      "step": 2697
+    },
+    {
+      "epoch": 0.6177447052089295,
+      "grad_norm": 1.2883368730545044,
+      "learning_rate": 6.755036396127297e-06,
+      "loss": 0.9892,
+      "step": 2698
+    },
+    {
+      "epoch": 0.6179736691471094,
+      "grad_norm": 1.2920852899551392,
+      "learning_rate": 6.748020544682172e-06,
+      "loss": 1.0611,
+      "step": 2699
+    },
+    {
+      "epoch": 0.6182026330852891,
+      "grad_norm": 1.2162961959838867,
+      "learning_rate": 6.7410064827740805e-06,
+      "loss": 0.9927,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6184315970234688,
+      "grad_norm": 1.3566070795059204,
+      "learning_rate": 6.733994214262797e-06,
+      "loss": 0.9985,
+      "step": 2701
+    },
+    {
+      "epoch": 0.6186605609616486,
+      "grad_norm": 1.4097816944122314,
+      "learning_rate": 6.726983743007112e-06,
+      "loss": 1.0129,
+      "step": 2702
+    },
+    {
+      "epoch": 0.6188895248998283,
+      "grad_norm": 1.3612055778503418,
+      "learning_rate": 6.7199750728648395e-06,
+      "loss": 1.0485,
+      "step": 2703
+    },
+    {
+      "epoch": 0.619118488838008,
+      "grad_norm": 1.2182425260543823,
+      "learning_rate": 6.712968207692778e-06,
+      "loss": 1.0371,
+      "step": 2704
+    },
+    {
+      "epoch": 0.6193474527761877,
+      "grad_norm": 1.3078467845916748,
+      "learning_rate": 6.705963151346755e-06,
+      "loss": 0.9789,
+      "step": 2705
+    },
+    {
+      "epoch": 0.6195764167143675,
+      "grad_norm": 1.150083065032959,
+      "learning_rate": 6.698959907681595e-06,
+      "loss": 1.0607,
+      "step": 2706
+    },
+    {
+      "epoch": 0.6198053806525472,
+      "grad_norm": 1.1655128002166748,
+      "learning_rate": 6.6919584805511175e-06,
+      "loss": 1.0277,
+      "step": 2707
+    },
+    {
+      "epoch": 0.6200343445907269,
+      "grad_norm": 1.273138165473938,
+      "learning_rate": 6.684958873808156e-06,
+      "loss": 1.0686,
+      "step": 2708
+    },
+    {
+      "epoch": 0.6202633085289067,
+      "grad_norm": 1.1581090688705444,
+      "learning_rate": 6.6779610913045344e-06,
+      "loss": 1.0464,
+      "step": 2709
+    },
+    {
+      "epoch": 0.6204922724670864,
+      "grad_norm": 1.240431547164917,
+      "learning_rate": 6.670965136891072e-06,
+      "loss": 1.051,
+      "step": 2710
+    },
+    {
+      "epoch": 0.6207212364052662,
+      "grad_norm": 1.4974451065063477,
+      "learning_rate": 6.663971014417585e-06,
+      "loss": 1.1388,
+      "step": 2711
+    },
+    {
+      "epoch": 0.620950200343446,
+      "grad_norm": 1.2878180742263794,
+      "learning_rate": 6.6569787277328745e-06,
+      "loss": 1.0783,
+      "step": 2712
+    },
+    {
+      "epoch": 0.6211791642816257,
+      "grad_norm": 1.712733507156372,
+      "learning_rate": 6.6499882806847445e-06,
+      "loss": 1.007,
+      "step": 2713
+    },
+    {
+      "epoch": 0.6214081282198054,
+      "grad_norm": 1.2414849996566772,
+      "learning_rate": 6.64299967711998e-06,
+      "loss": 1.0117,
+      "step": 2714
+    },
+    {
+      "epoch": 0.6216370921579851,
+      "grad_norm": 1.0478038787841797,
+      "learning_rate": 6.636012920884346e-06,
+      "loss": 1.0522,
+      "step": 2715
+    },
+    {
+      "epoch": 0.6218660560961649,
+      "grad_norm": 1.2143323421478271,
+      "learning_rate": 6.629028015822596e-06,
+      "loss": 1.0254,
+      "step": 2716
+    },
+    {
+      "epoch": 0.6220950200343446,
+      "grad_norm": 1.4381064176559448,
+      "learning_rate": 6.622044965778471e-06,
+      "loss": 1.1201,
+      "step": 2717
+    },
+    {
+      "epoch": 0.6223239839725243,
+      "grad_norm": 1.1734505891799927,
+      "learning_rate": 6.615063774594677e-06,
+      "loss": 1.1097,
+      "step": 2718
+    },
+    {
+      "epoch": 0.622552947910704,
+      "grad_norm": 1.087598204612732,
+      "learning_rate": 6.608084446112909e-06,
+      "loss": 1.0212,
+      "step": 2719
+    },
+    {
+      "epoch": 0.6227819118488838,
+      "grad_norm": 1.2489269971847534,
+      "learning_rate": 6.601106984173835e-06,
+      "loss": 1.0581,
+      "step": 2720
+    },
+    {
+      "epoch": 0.6230108757870635,
+      "grad_norm": 1.7127981185913086,
+      "learning_rate": 6.594131392617087e-06,
+      "loss": 1.0913,
+      "step": 2721
+    },
+    {
+      "epoch": 0.6232398397252433,
+      "grad_norm": 2.1938893795013428,
+      "learning_rate": 6.5871576752812845e-06,
+      "loss": 1.0303,
+      "step": 2722
+    },
+    {
+      "epoch": 0.623468803663423,
+      "grad_norm": 1.364212989807129,
+      "learning_rate": 6.580185836003995e-06,
+      "loss": 1.0338,
+      "step": 2723
+    },
+    {
+      "epoch": 0.6236977676016028,
+      "grad_norm": 1.1102733612060547,
+      "learning_rate": 6.573215878621769e-06,
+      "loss": 1.076,
+      "step": 2724
+    },
+    {
+      "epoch": 0.6239267315397825,
+      "grad_norm": 1.3565003871917725,
+      "learning_rate": 6.566247806970119e-06,
+      "loss": 1.0169,
+      "step": 2725
+    },
+    {
+      "epoch": 0.6241556954779622,
+      "grad_norm": 1.652982234954834,
+      "learning_rate": 6.559281624883506e-06,
+      "loss": 0.9878,
+      "step": 2726
+    },
+    {
+      "epoch": 0.624384659416142,
+      "grad_norm": 1.271844506263733,
+      "learning_rate": 6.552317336195371e-06,
+      "loss": 1.0609,
+      "step": 2727
+    },
+    {
+      "epoch": 0.6246136233543217,
+      "grad_norm": 1.2755295038223267,
+      "learning_rate": 6.5453549447381e-06,
+      "loss": 1.0662,
+      "step": 2728
+    },
+    {
+      "epoch": 0.6248425872925014,
+      "grad_norm": 1.265498399734497,
+      "learning_rate": 6.53839445434304e-06,
+      "loss": 1.0132,
+      "step": 2729
+    },
+    {
+      "epoch": 0.6250715512306811,
+      "grad_norm": 1.5743672847747803,
+      "learning_rate": 6.531435868840488e-06,
+      "loss": 1.0412,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6253005151688609,
+      "grad_norm": 1.1162136793136597,
+      "learning_rate": 6.524479192059699e-06,
+      "loss": 1.0259,
+      "step": 2731
+    },
+    {
+      "epoch": 0.6255294791070406,
+      "grad_norm": 1.2399483919143677,
+      "learning_rate": 6.5175244278288705e-06,
+      "loss": 1.0109,
+      "step": 2732
+    },
+    {
+      "epoch": 0.6257584430452203,
+      "grad_norm": 1.169719934463501,
+      "learning_rate": 6.510571579975155e-06,
+      "loss": 0.9764,
+      "step": 2733
+    },
+    {
+      "epoch": 0.6259874069834002,
+      "grad_norm": 1.7997078895568848,
+      "learning_rate": 6.5036206523246404e-06,
+      "loss": 1.0591,
+      "step": 2734
+    },
+    {
+      "epoch": 0.6262163709215799,
+      "grad_norm": 1.2209597826004028,
+      "learning_rate": 6.496671648702366e-06,
+      "loss": 1.0484,
+      "step": 2735
+    },
+    {
+      "epoch": 0.6264453348597596,
+      "grad_norm": 1.2143415212631226,
+      "learning_rate": 6.489724572932314e-06,
+      "loss": 1.0438,
+      "step": 2736
+    },
+    {
+      "epoch": 0.6266742987979393,
+      "grad_norm": 1.148642897605896,
+      "learning_rate": 6.4827794288374e-06,
+      "loss": 1.028,
+      "step": 2737
+    },
+    {
+      "epoch": 0.6269032627361191,
+      "grad_norm": 1.284192681312561,
+      "learning_rate": 6.475836220239475e-06,
+      "loss": 1.0812,
+      "step": 2738
+    },
+    {
+      "epoch": 0.6271322266742988,
+      "grad_norm": 1.0696079730987549,
+      "learning_rate": 6.468894950959336e-06,
+      "loss": 1.1179,
+      "step": 2739
+    },
+    {
+      "epoch": 0.6273611906124785,
+      "grad_norm": 1.2461477518081665,
+      "learning_rate": 6.461955624816696e-06,
+      "loss": 1.0082,
+      "step": 2740
+    },
+    {
+      "epoch": 0.6275901545506583,
+      "grad_norm": 1.2268809080123901,
+      "learning_rate": 6.455018245630214e-06,
+      "loss": 1.0832,
+      "step": 2741
+    },
+    {
+      "epoch": 0.627819118488838,
+      "grad_norm": 1.1206541061401367,
+      "learning_rate": 6.4480828172174714e-06,
+      "loss": 1.0686,
+      "step": 2742
+    },
+    {
+      "epoch": 0.6280480824270177,
+      "grad_norm": 1.9735060930252075,
+      "learning_rate": 6.441149343394975e-06,
+      "loss": 1.0499,
+      "step": 2743
+    },
+    {
+      "epoch": 0.6282770463651974,
+      "grad_norm": 1.1669235229492188,
+      "learning_rate": 6.4342178279781584e-06,
+      "loss": 1.0743,
+      "step": 2744
+    },
+    {
+      "epoch": 0.6285060103033773,
+      "grad_norm": 1.3031091690063477,
+      "learning_rate": 6.427288274781372e-06,
+      "loss": 1.0164,
+      "step": 2745
+    },
+    {
+      "epoch": 0.628734974241557,
+      "grad_norm": 1.2017693519592285,
+      "learning_rate": 6.420360687617897e-06,
+      "loss": 0.9844,
+      "step": 2746
+    },
+    {
+      "epoch": 0.6289639381797367,
+      "grad_norm": 1.155867576599121,
+      "learning_rate": 6.413435070299925e-06,
+      "loss": 1.0669,
+      "step": 2747
+    },
+    {
+      "epoch": 0.6291929021179165,
+      "grad_norm": 1.315988302230835,
+      "learning_rate": 6.406511426638562e-06,
+      "loss": 1.0525,
+      "step": 2748
+    },
+    {
+      "epoch": 0.6294218660560962,
+      "grad_norm": 1.3917714357376099,
+      "learning_rate": 6.3995897604438315e-06,
+      "loss": 1.0961,
+      "step": 2749
+    },
+    {
+      "epoch": 0.6296508299942759,
+      "grad_norm": 1.3738504648208618,
+      "learning_rate": 6.392670075524674e-06,
+      "loss": 1.0626,
+      "step": 2750
+    },
+    {
+      "epoch": 0.6298797939324556,
+      "grad_norm": 1.1680269241333008,
+      "learning_rate": 6.385752375688927e-06,
+      "loss": 1.0226,
+      "step": 2751
+    },
+    {
+      "epoch": 0.6301087578706354,
+      "grad_norm": 1.6877174377441406,
+      "learning_rate": 6.378836664743347e-06,
+      "loss": 1.036,
+      "step": 2752
+    },
+    {
+      "epoch": 0.6303377218088151,
+      "grad_norm": 1.2052552700042725,
+      "learning_rate": 6.3719229464935915e-06,
+      "loss": 1.0776,
+      "step": 2753
+    },
+    {
+      "epoch": 0.6305666857469948,
+      "grad_norm": 1.3558235168457031,
+      "learning_rate": 6.365011224744218e-06,
+      "loss": 1.0233,
+      "step": 2754
+    },
+    {
+      "epoch": 0.6307956496851745,
+      "grad_norm": 1.0907037258148193,
+      "learning_rate": 6.3581015032986945e-06,
+      "loss": 0.9963,
+      "step": 2755
+    },
+    {
+      "epoch": 0.6310246136233543,
+      "grad_norm": 1.4145336151123047,
+      "learning_rate": 6.35119378595938e-06,
+      "loss": 1.0576,
+      "step": 2756
+    },
+    {
+      "epoch": 0.6312535775615341,
+      "grad_norm": 1.3490455150604248,
+      "learning_rate": 6.344288076527532e-06,
+      "loss": 1.0378,
+      "step": 2757
+    },
+    {
+      "epoch": 0.6314825414997138,
+      "grad_norm": 1.4001924991607666,
+      "learning_rate": 6.337384378803309e-06,
+      "loss": 1.108,
+      "step": 2758
+    },
+    {
+      "epoch": 0.6317115054378936,
+      "grad_norm": 1.187118411064148,
+      "learning_rate": 6.330482696585749e-06,
+      "loss": 1.0318,
+      "step": 2759
+    },
+    {
+      "epoch": 0.6319404693760733,
+      "grad_norm": 1.5335361957550049,
+      "learning_rate": 6.323583033672799e-06,
+      "loss": 1.0561,
+      "step": 2760
+    },
+    {
+      "epoch": 0.632169433314253,
+      "grad_norm": 1.461372971534729,
+      "learning_rate": 6.316685393861284e-06,
+      "loss": 0.988,
+      "step": 2761
+    },
+    {
+      "epoch": 0.6323983972524327,
+      "grad_norm": 1.2753512859344482,
+      "learning_rate": 6.309789780946916e-06,
+      "loss": 1.07,
+      "step": 2762
+    },
+    {
+      "epoch": 0.6326273611906125,
+      "grad_norm": 1.0639042854309082,
+      "learning_rate": 6.302896198724288e-06,
+      "loss": 1.122,
+      "step": 2763
+    },
+    {
+      "epoch": 0.6328563251287922,
+      "grad_norm": 1.267246961593628,
+      "learning_rate": 6.29600465098689e-06,
+      "loss": 1.068,
+      "step": 2764
+    },
+    {
+      "epoch": 0.6330852890669719,
+      "grad_norm": 1.2320417165756226,
+      "learning_rate": 6.289115141527077e-06,
+      "loss": 1.0611,
+      "step": 2765
+    },
+    {
+      "epoch": 0.6333142530051517,
+      "grad_norm": 1.517449975013733,
+      "learning_rate": 6.282227674136091e-06,
+      "loss": 0.9906,
+      "step": 2766
+    },
+    {
+      "epoch": 0.6335432169433314,
+      "grad_norm": 1.2116914987564087,
+      "learning_rate": 6.275342252604044e-06,
+      "loss": 1.0568,
+      "step": 2767
+    },
+    {
+      "epoch": 0.6337721808815112,
+      "grad_norm": 1.178650140762329,
+      "learning_rate": 6.2684588807199265e-06,
+      "loss": 1.0065,
+      "step": 2768
+    },
+    {
+      "epoch": 0.6340011448196909,
+      "grad_norm": 1.3236886262893677,
+      "learning_rate": 6.261577562271605e-06,
+      "loss": 0.9728,
+      "step": 2769
+    },
+    {
+      "epoch": 0.6342301087578707,
+      "grad_norm": 1.2734071016311646,
+      "learning_rate": 6.254698301045806e-06,
+      "loss": 1.05,
+      "step": 2770
+    },
+    {
+      "epoch": 0.6344590726960504,
+      "grad_norm": 1.281076431274414,
+      "learning_rate": 6.247821100828131e-06,
+      "loss": 1.0518,
+      "step": 2771
+    },
+    {
+      "epoch": 0.6346880366342301,
+      "grad_norm": 1.1672574281692505,
+      "learning_rate": 6.240945965403049e-06,
+      "loss": 1.0612,
+      "step": 2772
+    },
+    {
+      "epoch": 0.6349170005724099,
+      "grad_norm": 1.2151325941085815,
+      "learning_rate": 6.234072898553882e-06,
+      "loss": 1.0501,
+      "step": 2773
+    },
+    {
+      "epoch": 0.6351459645105896,
+      "grad_norm": 1.215021014213562,
+      "learning_rate": 6.22720190406283e-06,
+      "loss": 1.0501,
+      "step": 2774
+    },
+    {
+      "epoch": 0.6353749284487693,
+      "grad_norm": 1.1667354106903076,
+      "learning_rate": 6.220332985710936e-06,
+      "loss": 1.0068,
+      "step": 2775
+    },
+    {
+      "epoch": 0.635603892386949,
+      "grad_norm": 1.4257439374923706,
+      "learning_rate": 6.213466147278111e-06,
+      "loss": 1.085,
+      "step": 2776
+    },
+    {
+      "epoch": 0.6358328563251288,
+      "grad_norm": 1.5628901720046997,
+      "learning_rate": 6.206601392543121e-06,
+      "loss": 0.9748,
+      "step": 2777
+    },
+    {
+      "epoch": 0.6360618202633085,
+      "grad_norm": 1.2145711183547974,
+      "learning_rate": 6.199738725283578e-06,
+      "loss": 1.0345,
+      "step": 2778
+    },
+    {
+      "epoch": 0.6362907842014882,
+      "grad_norm": 1.2755681276321411,
+      "learning_rate": 6.192878149275954e-06,
+      "loss": 1.068,
+      "step": 2779
+    },
+    {
+      "epoch": 0.636519748139668,
+      "grad_norm": 1.1560249328613281,
+      "learning_rate": 6.186019668295568e-06,
+      "loss": 1.0608,
+      "step": 2780
+    },
+    {
+      "epoch": 0.6367487120778478,
+      "grad_norm": 1.1933772563934326,
+      "learning_rate": 6.179163286116581e-06,
+      "loss": 1.0713,
+      "step": 2781
+    },
+    {
+      "epoch": 0.6369776760160275,
+      "grad_norm": 2.3430261611938477,
+      "learning_rate": 6.172309006511999e-06,
+      "loss": 0.9809,
+      "step": 2782
+    },
+    {
+      "epoch": 0.6372066399542072,
+      "grad_norm": 1.1244544982910156,
+      "learning_rate": 6.165456833253686e-06,
+      "loss": 1.0319,
+      "step": 2783
+    },
+    {
+      "epoch": 0.637435603892387,
+      "grad_norm": 1.41416597366333,
+      "learning_rate": 6.1586067701123255e-06,
+      "loss": 1.0276,
+      "step": 2784
+    },
+    {
+      "epoch": 0.6376645678305667,
+      "grad_norm": 1.5440860986709595,
+      "learning_rate": 6.151758820857455e-06,
+      "loss": 1.0168,
+      "step": 2785
+    },
+    {
+      "epoch": 0.6378935317687464,
+      "grad_norm": 1.2612404823303223,
+      "learning_rate": 6.144912989257441e-06,
+      "loss": 0.9923,
+      "step": 2786
+    },
+    {
+      "epoch": 0.6381224957069261,
+      "grad_norm": 1.4268378019332886,
+      "learning_rate": 6.138069279079484e-06,
+      "loss": 1.0428,
+      "step": 2787
+    },
+    {
+      "epoch": 0.6383514596451059,
+      "grad_norm": 1.4409279823303223,
+      "learning_rate": 6.13122769408963e-06,
+      "loss": 1.0381,
+      "step": 2788
+    },
+    {
+      "epoch": 0.6385804235832856,
+      "grad_norm": 1.529488205909729,
+      "learning_rate": 6.124388238052737e-06,
+      "loss": 1.0536,
+      "step": 2789
+    },
+    {
+      "epoch": 0.6388093875214653,
+      "grad_norm": 1.03910231590271,
+      "learning_rate": 6.1175509147325015e-06,
+      "loss": 0.986,
+      "step": 2790
+    },
+    {
+      "epoch": 0.6390383514596452,
+      "grad_norm": 1.074388027191162,
+      "learning_rate": 6.1107157278914545e-06,
+      "loss": 1.066,
+      "step": 2791
+    },
+    {
+      "epoch": 0.6392673153978249,
+      "grad_norm": 1.1778825521469116,
+      "learning_rate": 6.1038826812909265e-06,
+      "loss": 1.1038,
+      "step": 2792
+    },
+    {
+      "epoch": 0.6394962793360046,
+      "grad_norm": 1.192124843597412,
+      "learning_rate": 6.097051778691099e-06,
+      "loss": 1.0755,
+      "step": 2793
+    },
+    {
+      "epoch": 0.6397252432741843,
+      "grad_norm": 1.2130941152572632,
+      "learning_rate": 6.090223023850954e-06,
+      "loss": 1.0033,
+      "step": 2794
+    },
+    {
+      "epoch": 0.6399542072123641,
+      "grad_norm": 1.1291038990020752,
+      "learning_rate": 6.083396420528298e-06,
+      "loss": 1.04,
+      "step": 2795
+    },
+    {
+      "epoch": 0.6401831711505438,
+      "grad_norm": 1.3231033086776733,
+      "learning_rate": 6.0765719724797586e-06,
+      "loss": 1.0925,
+      "step": 2796
+    },
+    {
+      "epoch": 0.6404121350887235,
+      "grad_norm": 1.6609188318252563,
+      "learning_rate": 6.069749683460765e-06,
+      "loss": 1.0937,
+      "step": 2797
+    },
+    {
+      "epoch": 0.6406410990269032,
+      "grad_norm": 1.21103036403656,
+      "learning_rate": 6.0629295572255695e-06,
+      "loss": 0.951,
+      "step": 2798
+    },
+    {
+      "epoch": 0.640870062965083,
+      "grad_norm": 1.0261344909667969,
+      "learning_rate": 6.056111597527235e-06,
+      "loss": 1.0446,
+      "step": 2799
+    },
+    {
+      "epoch": 0.6410990269032627,
+      "grad_norm": 1.1342116594314575,
+      "learning_rate": 6.0492958081176155e-06,
+      "loss": 1.0956,
+      "step": 2800
+    },
+    {
+      "epoch": 0.6413279908414424,
+      "grad_norm": 1.2922279834747314,
+      "learning_rate": 6.042482192747394e-06,
+      "loss": 1.0102,
+      "step": 2801
+    },
+    {
+      "epoch": 0.6415569547796222,
+      "grad_norm": 1.2681461572647095,
+      "learning_rate": 6.0356707551660434e-06,
+      "loss": 0.9613,
+      "step": 2802
+    },
+    {
+      "epoch": 0.641785918717802,
+      "grad_norm": 1.3258183002471924,
+      "learning_rate": 6.0288614991218366e-06,
+      "loss": 1.0068,
+      "step": 2803
+    },
+    {
+      "epoch": 0.6420148826559817,
+      "grad_norm": 1.2088922262191772,
+      "learning_rate": 6.022054428361852e-06,
+      "loss": 1.0253,
+      "step": 2804
+    },
+    {
+      "epoch": 0.6422438465941614,
+      "grad_norm": 1.0741603374481201,
+      "learning_rate": 6.015249546631969e-06,
+      "loss": 0.9941,
+      "step": 2805
+    },
+    {
+      "epoch": 0.6424728105323412,
+      "grad_norm": 1.408013105392456,
+      "learning_rate": 6.008446857676849e-06,
+      "loss": 1.0913,
+      "step": 2806
+    },
+    {
+      "epoch": 0.6427017744705209,
+      "grad_norm": 1.087931752204895,
+      "learning_rate": 6.001646365239959e-06,
+      "loss": 1.0355,
+      "step": 2807
+    },
+    {
+      "epoch": 0.6429307384087006,
+      "grad_norm": 1.1895891427993774,
+      "learning_rate": 5.994848073063552e-06,
+      "loss": 1.1122,
+      "step": 2808
+    },
+    {
+      "epoch": 0.6431597023468804,
+      "grad_norm": 1.3358407020568848,
+      "learning_rate": 5.988051984888668e-06,
+      "loss": 1.0171,
+      "step": 2809
+    },
+    {
+      "epoch": 0.6433886662850601,
+      "grad_norm": 1.3957514762878418,
+      "learning_rate": 5.9812581044551475e-06,
+      "loss": 1.0719,
+      "step": 2810
+    },
+    {
+      "epoch": 0.6436176302232398,
+      "grad_norm": 2.103109836578369,
+      "learning_rate": 5.974466435501591e-06,
+      "loss": 1.0735,
+      "step": 2811
+    },
+    {
+      "epoch": 0.6438465941614195,
+      "grad_norm": 1.2422171831130981,
+      "learning_rate": 5.967676981765409e-06,
+      "loss": 1.0396,
+      "step": 2812
+    },
+    {
+      "epoch": 0.6440755580995993,
+      "grad_norm": 1.1619694232940674,
+      "learning_rate": 5.960889746982778e-06,
+      "loss": 1.063,
+      "step": 2813
+    },
+    {
+      "epoch": 0.6443045220377791,
+      "grad_norm": 1.4143716096878052,
+      "learning_rate": 5.954104734888653e-06,
+      "loss": 1.0359,
+      "step": 2814
+    },
+    {
+      "epoch": 0.6445334859759588,
+      "grad_norm": 1.296014666557312,
+      "learning_rate": 5.947321949216771e-06,
+      "loss": 1.0085,
+      "step": 2815
+    },
+    {
+      "epoch": 0.6447624499141386,
+      "grad_norm": 1.0912615060806274,
+      "learning_rate": 5.940541393699646e-06,
+      "loss": 1.0417,
+      "step": 2816
+    },
+    {
+      "epoch": 0.6449914138523183,
+      "grad_norm": 1.079500675201416,
+      "learning_rate": 5.933763072068554e-06,
+      "loss": 0.9998,
+      "step": 2817
+    },
+    {
+      "epoch": 0.645220377790498,
+      "grad_norm": 1.2979693412780762,
+      "learning_rate": 5.926986988053557e-06,
+      "loss": 1.0842,
+      "step": 2818
+    },
+    {
+      "epoch": 0.6454493417286777,
+      "grad_norm": 1.6231458187103271,
+      "learning_rate": 5.9202131453834664e-06,
+      "loss": 1.0377,
+      "step": 2819
+    },
+    {
+      "epoch": 0.6456783056668575,
+      "grad_norm": 1.2327483892440796,
+      "learning_rate": 5.913441547785879e-06,
+      "loss": 1.0495,
+      "step": 2820
+    },
+    {
+      "epoch": 0.6459072696050372,
+      "grad_norm": 1.5749176740646362,
+      "learning_rate": 5.906672198987149e-06,
+      "loss": 1.0473,
+      "step": 2821
+    },
+    {
+      "epoch": 0.6461362335432169,
+      "grad_norm": 1.3406803607940674,
+      "learning_rate": 5.899905102712386e-06,
+      "loss": 1.062,
+      "step": 2822
+    },
+    {
+      "epoch": 0.6463651974813966,
+      "grad_norm": 1.0731439590454102,
+      "learning_rate": 5.893140262685469e-06,
+      "loss": 1.0155,
+      "step": 2823
+    },
+    {
+      "epoch": 0.6465941614195764,
+      "grad_norm": 1.2645176649093628,
+      "learning_rate": 5.886377682629037e-06,
+      "loss": 0.9567,
+      "step": 2824
+    },
+    {
+      "epoch": 0.6468231253577561,
+      "grad_norm": 1.1633598804473877,
+      "learning_rate": 5.879617366264476e-06,
+      "loss": 0.9959,
+      "step": 2825
+    },
+    {
+      "epoch": 0.6470520892959359,
+      "grad_norm": 1.266831874847412,
+      "learning_rate": 5.872859317311933e-06,
+      "loss": 1.035,
+      "step": 2826
+    },
+    {
+      "epoch": 0.6472810532341157,
+      "grad_norm": 1.426615595817566,
+      "learning_rate": 5.866103539490307e-06,
+      "loss": 1.0236,
+      "step": 2827
+    },
+    {
+      "epoch": 0.6475100171722954,
+      "grad_norm": 1.0592693090438843,
+      "learning_rate": 5.859350036517242e-06,
+      "loss": 1.0293,
+      "step": 2828
+    },
+    {
+      "epoch": 0.6477389811104751,
+      "grad_norm": 1.2604130506515503,
+      "learning_rate": 5.852598812109139e-06,
+      "loss": 1.0833,
+      "step": 2829
+    },
+    {
+      "epoch": 0.6479679450486548,
+      "grad_norm": 1.2905731201171875,
+      "learning_rate": 5.845849869981137e-06,
+      "loss": 1.109,
+      "step": 2830
+    },
+    {
+      "epoch": 0.6481969089868346,
+      "grad_norm": 1.1893059015274048,
+      "learning_rate": 5.839103213847123e-06,
+      "loss": 1.0075,
+      "step": 2831
+    },
+    {
+      "epoch": 0.6484258729250143,
+      "grad_norm": 1.5178680419921875,
+      "learning_rate": 5.832358847419728e-06,
+      "loss": 1.0779,
+      "step": 2832
+    },
+    {
+      "epoch": 0.648654836863194,
+      "grad_norm": 1.2127212285995483,
+      "learning_rate": 5.8256167744103144e-06,
+      "loss": 1.0206,
+      "step": 2833
+    },
+    {
+      "epoch": 0.6488838008013738,
+      "grad_norm": 1.2147332429885864,
+      "learning_rate": 5.818876998528988e-06,
+      "loss": 1.019,
+      "step": 2834
+    },
+    {
+      "epoch": 0.6491127647395535,
+      "grad_norm": 1.7146003246307373,
+      "learning_rate": 5.812139523484604e-06,
+      "loss": 0.9919,
+      "step": 2835
+    },
+    {
+      "epoch": 0.6493417286777332,
+      "grad_norm": 1.2911676168441772,
+      "learning_rate": 5.805404352984724e-06,
+      "loss": 1.0572,
+      "step": 2836
+    },
+    {
+      "epoch": 0.649570692615913,
+      "grad_norm": 1.3293169736862183,
+      "learning_rate": 5.7986714907356614e-06,
+      "loss": 0.9918,
+      "step": 2837
+    },
+    {
+      "epoch": 0.6497996565540928,
+      "grad_norm": 1.9981685876846313,
+      "learning_rate": 5.791940940442453e-06,
+      "loss": 1.0236,
+      "step": 2838
+    },
+    {
+      "epoch": 0.6500286204922725,
+      "grad_norm": 1.25807785987854,
+      "learning_rate": 5.785212705808865e-06,
+      "loss": 1.0795,
+      "step": 2839
+    },
+    {
+      "epoch": 0.6502575844304522,
+      "grad_norm": 1.261753797531128,
+      "learning_rate": 5.778486790537392e-06,
+      "loss": 1.087,
+      "step": 2840
+    },
+    {
+      "epoch": 0.650486548368632,
+      "grad_norm": 1.0655620098114014,
+      "learning_rate": 5.7717631983292375e-06,
+      "loss": 1.0383,
+      "step": 2841
+    },
+    {
+      "epoch": 0.6507155123068117,
+      "grad_norm": 1.2953264713287354,
+      "learning_rate": 5.76504193288435e-06,
+      "loss": 1.0587,
+      "step": 2842
+    },
+    {
+      "epoch": 0.6509444762449914,
+      "grad_norm": 1.2501654624938965,
+      "learning_rate": 5.758322997901384e-06,
+      "loss": 1.0403,
+      "step": 2843
+    },
+    {
+      "epoch": 0.6511734401831711,
+      "grad_norm": 1.1158286333084106,
+      "learning_rate": 5.751606397077703e-06,
+      "loss": 1.0854,
+      "step": 2844
+    },
+    {
+      "epoch": 0.6514024041213509,
+      "grad_norm": 1.357792615890503,
+      "learning_rate": 5.744892134109406e-06,
+      "loss": 1.1166,
+      "step": 2845
+    },
+    {
+      "epoch": 0.6516313680595306,
+      "grad_norm": 1.070070505142212,
+      "learning_rate": 5.738180212691296e-06,
+      "loss": 1.0345,
+      "step": 2846
+    },
+    {
+      "epoch": 0.6518603319977103,
+      "grad_norm": 1.1419059038162231,
+      "learning_rate": 5.7314706365168806e-06,
+      "loss": 1.0482,
+      "step": 2847
+    },
+    {
+      "epoch": 0.65208929593589,
+      "grad_norm": 1.5696709156036377,
+      "learning_rate": 5.724763409278383e-06,
+      "loss": 1.0254,
+      "step": 2848
+    },
+    {
+      "epoch": 0.6523182598740699,
+      "grad_norm": 1.2937999963760376,
+      "learning_rate": 5.718058534666746e-06,
+      "loss": 1.078,
+      "step": 2849
+    },
+    {
+      "epoch": 0.6525472238122496,
+      "grad_norm": 1.8056284189224243,
+      "learning_rate": 5.711356016371593e-06,
+      "loss": 1.0889,
+      "step": 2850
+    },
+    {
+      "epoch": 0.6527761877504293,
+      "grad_norm": 1.4800915718078613,
+      "learning_rate": 5.704655858081268e-06,
+      "loss": 1.0542,
+      "step": 2851
+    },
+    {
+      "epoch": 0.6530051516886091,
+      "grad_norm": 1.4089901447296143,
+      "learning_rate": 5.6979580634828125e-06,
+      "loss": 1.0553,
+      "step": 2852
+    },
+    {
+      "epoch": 0.6532341156267888,
+      "grad_norm": 1.154541254043579,
+      "learning_rate": 5.691262636261967e-06,
+      "loss": 1.0012,
+      "step": 2853
+    },
+    {
+      "epoch": 0.6534630795649685,
+      "grad_norm": 1.3811694383621216,
+      "learning_rate": 5.684569580103171e-06,
+      "loss": 0.9693,
+      "step": 2854
+    },
+    {
+      "epoch": 0.6536920435031482,
+      "grad_norm": 2.098176956176758,
+      "learning_rate": 5.6778788986895464e-06,
+      "loss": 1.1379,
+      "step": 2855
+    },
+    {
+      "epoch": 0.653921007441328,
+      "grad_norm": 1.450215458869934,
+      "learning_rate": 5.671190595702932e-06,
+      "loss": 1.0369,
+      "step": 2856
+    },
+    {
+      "epoch": 0.6541499713795077,
+      "grad_norm": 1.3346037864685059,
+      "learning_rate": 5.664504674823844e-06,
+      "loss": 1.128,
+      "step": 2857
+    },
+    {
+      "epoch": 0.6543789353176874,
+      "grad_norm": 2.112135171890259,
+      "learning_rate": 5.6578211397314765e-06,
+      "loss": 1.0601,
+      "step": 2858
+    },
+    {
+      "epoch": 0.6546078992558672,
+      "grad_norm": 1.497977614402771,
+      "learning_rate": 5.6511399941037344e-06,
+      "loss": 1.0149,
+      "step": 2859
+    },
+    {
+      "epoch": 0.654836863194047,
+      "grad_norm": 1.1439969539642334,
+      "learning_rate": 5.6444612416171976e-06,
+      "loss": 0.9617,
+      "step": 2860
+    },
+    {
+      "epoch": 0.6550658271322267,
+      "grad_norm": 1.8310574293136597,
+      "learning_rate": 5.637784885947117e-06,
+      "loss": 1.0273,
+      "step": 2861
+    },
+    {
+      "epoch": 0.6552947910704064,
+      "grad_norm": 1.3659378290176392,
+      "learning_rate": 5.631110930767443e-06,
+      "loss": 1.058,
+      "step": 2862
+    },
+    {
+      "epoch": 0.6555237550085862,
+      "grad_norm": 1.355661153793335,
+      "learning_rate": 5.6244393797507944e-06,
+      "loss": 1.0411,
+      "step": 2863
+    },
+    {
+      "epoch": 0.6557527189467659,
+      "grad_norm": 0.9867006540298462,
+      "learning_rate": 5.617770236568469e-06,
+      "loss": 0.9702,
+      "step": 2864
+    },
+    {
+      "epoch": 0.6559816828849456,
+      "grad_norm": 1.066946029663086,
+      "learning_rate": 5.611103504890444e-06,
+      "loss": 1.0521,
+      "step": 2865
+    },
+    {
+      "epoch": 0.6562106468231254,
+      "grad_norm": 1.2276383638381958,
+      "learning_rate": 5.604439188385362e-06,
+      "loss": 0.9953,
+      "step": 2866
+    },
+    {
+      "epoch": 0.6564396107613051,
+      "grad_norm": 1.2933861017227173,
+      "learning_rate": 5.597777290720543e-06,
+      "loss": 1.0319,
+      "step": 2867
+    },
+    {
+      "epoch": 0.6566685746994848,
+      "grad_norm": 1.144028902053833,
+      "learning_rate": 5.591117815561973e-06,
+      "loss": 1.0391,
+      "step": 2868
+    },
+    {
+      "epoch": 0.6568975386376645,
+      "grad_norm": 1.4104443788528442,
+      "learning_rate": 5.584460766574304e-06,
+      "loss": 1.0806,
+      "step": 2869
+    },
+    {
+      "epoch": 0.6571265025758443,
+      "grad_norm": 2.108306884765625,
+      "learning_rate": 5.5778061474208565e-06,
+      "loss": 1.0714,
+      "step": 2870
+    },
+    {
+      "epoch": 0.657355466514024,
+      "grad_norm": 1.3675113916397095,
+      "learning_rate": 5.571153961763613e-06,
+      "loss": 1.0704,
+      "step": 2871
+    },
+    {
+      "epoch": 0.6575844304522038,
+      "grad_norm": 1.1373252868652344,
+      "learning_rate": 5.564504213263205e-06,
+      "loss": 1.0022,
+      "step": 2872
+    },
+    {
+      "epoch": 0.6578133943903836,
+      "grad_norm": 5.982883930206299,
+      "learning_rate": 5.55785690557895e-06,
+      "loss": 1.1152,
+      "step": 2873
+    },
+    {
+      "epoch": 0.6580423583285633,
+      "grad_norm": 1.589943289756775,
+      "learning_rate": 5.551212042368792e-06,
+      "loss": 1.0449,
+      "step": 2874
+    },
+    {
+      "epoch": 0.658271322266743,
+      "grad_norm": 1.2605445384979248,
+      "learning_rate": 5.54456962728935e-06,
+      "loss": 1.0192,
+      "step": 2875
+    },
+    {
+      "epoch": 0.6585002862049227,
+      "grad_norm": 1.27881920337677,
+      "learning_rate": 5.537929663995887e-06,
+      "loss": 1.1066,
+      "step": 2876
+    },
+    {
+      "epoch": 0.6587292501431025,
+      "grad_norm": 2.3701119422912598,
+      "learning_rate": 5.531292156142319e-06,
+      "loss": 1.0039,
+      "step": 2877
+    },
+    {
+      "epoch": 0.6589582140812822,
+      "grad_norm": 2.732001543045044,
+      "learning_rate": 5.5246571073812124e-06,
+      "loss": 1.0069,
+      "step": 2878
+    },
+    {
+      "epoch": 0.6591871780194619,
+      "grad_norm": 1.3857358694076538,
+      "learning_rate": 5.5180245213637785e-06,
+      "loss": 1.0746,
+      "step": 2879
+    },
+    {
+      "epoch": 0.6594161419576416,
+      "grad_norm": 1.0118176937103271,
+      "learning_rate": 5.511394401739874e-06,
+      "loss": 1.0039,
+      "step": 2880
+    },
+    {
+      "epoch": 0.6596451058958214,
+      "grad_norm": 1.1465644836425781,
+      "learning_rate": 5.504766752157997e-06,
+      "loss": 1.0596,
+      "step": 2881
+    },
+    {
+      "epoch": 0.6598740698340011,
+      "grad_norm": 1.1057982444763184,
+      "learning_rate": 5.498141576265289e-06,
+      "loss": 1.0616,
+      "step": 2882
+    },
+    {
+      "epoch": 0.6601030337721809,
+      "grad_norm": 1.2650837898254395,
+      "learning_rate": 5.491518877707527e-06,
+      "loss": 1.045,
+      "step": 2883
+    },
+    {
+      "epoch": 0.6603319977103607,
+      "grad_norm": 1.2822803258895874,
+      "learning_rate": 5.484898660129132e-06,
+      "loss": 1.0242,
+      "step": 2884
+    },
+    {
+      "epoch": 0.6605609616485404,
+      "grad_norm": 1.1227819919586182,
+      "learning_rate": 5.478280927173145e-06,
+      "loss": 0.9838,
+      "step": 2885
+    },
+    {
+      "epoch": 0.6607899255867201,
+      "grad_norm": 2.5325989723205566,
+      "learning_rate": 5.4716656824812505e-06,
+      "loss": 1.0738,
+      "step": 2886
+    },
+    {
+      "epoch": 0.6610188895248998,
+      "grad_norm": 2.1725833415985107,
+      "learning_rate": 5.465052929693774e-06,
+      "loss": 1.0192,
+      "step": 2887
+    },
+    {
+      "epoch": 0.6612478534630796,
+      "grad_norm": 1.7834242582321167,
+      "learning_rate": 5.458442672449644e-06,
+      "loss": 1.0923,
+      "step": 2888
+    },
+    {
+      "epoch": 0.6614768174012593,
+      "grad_norm": 1.2370448112487793,
+      "learning_rate": 5.451834914386435e-06,
+      "loss": 1.0404,
+      "step": 2889
+    },
+    {
+      "epoch": 0.661705781339439,
+      "grad_norm": 1.3691270351409912,
+      "learning_rate": 5.445229659140341e-06,
+      "loss": 1.0226,
+      "step": 2890
+    },
+    {
+      "epoch": 0.6619347452776188,
+      "grad_norm": 1.2318962812423706,
+      "learning_rate": 5.4386269103461785e-06,
+      "loss": 1.039,
+      "step": 2891
+    },
+    {
+      "epoch": 0.6621637092157985,
+      "grad_norm": 1.2632200717926025,
+      "learning_rate": 5.432026671637385e-06,
+      "loss": 1.077,
+      "step": 2892
+    },
+    {
+      "epoch": 0.6623926731539782,
+      "grad_norm": 1.4517078399658203,
+      "learning_rate": 5.425428946646016e-06,
+      "loss": 1.0304,
+      "step": 2893
+    },
+    {
+      "epoch": 0.6626216370921579,
+      "grad_norm": 1.1040912866592407,
+      "learning_rate": 5.418833739002745e-06,
+      "loss": 1.0163,
+      "step": 2894
+    },
+    {
+      "epoch": 0.6628506010303378,
+      "grad_norm": 2.0394599437713623,
+      "learning_rate": 5.4122410523368615e-06,
+      "loss": 1.0289,
+      "step": 2895
+    },
+    {
+      "epoch": 0.6630795649685175,
+      "grad_norm": 1.1909575462341309,
+      "learning_rate": 5.405650890276255e-06,
+      "loss": 1.0632,
+      "step": 2896
+    },
+    {
+      "epoch": 0.6633085289066972,
+      "grad_norm": 1.359932541847229,
+      "learning_rate": 5.39906325644745e-06,
+      "loss": 0.9915,
+      "step": 2897
+    },
+    {
+      "epoch": 0.663537492844877,
+      "grad_norm": 1.3327099084854126,
+      "learning_rate": 5.392478154475565e-06,
+      "loss": 1.0265,
+      "step": 2898
+    },
+    {
+      "epoch": 0.6637664567830567,
+      "grad_norm": 1.2770649194717407,
+      "learning_rate": 5.3858955879843155e-06,
+      "loss": 1.0492,
+      "step": 2899
+    },
+    {
+      "epoch": 0.6639954207212364,
+      "grad_norm": 1.136519193649292,
+      "learning_rate": 5.379315560596038e-06,
+      "loss": 1.027,
+      "step": 2900
+    },
+    {
+      "epoch": 0.6642243846594161,
+      "grad_norm": 1.217628836631775,
+      "learning_rate": 5.372738075931674e-06,
+      "loss": 0.9827,
+      "step": 2901
+    },
+    {
+      "epoch": 0.6644533485975959,
+      "grad_norm": 1.2686570882797241,
+      "learning_rate": 5.366163137610749e-06,
+      "loss": 1.063,
+      "step": 2902
+    },
+    {
+      "epoch": 0.6646823125357756,
+      "grad_norm": 1.29387629032135,
+      "learning_rate": 5.359590749251397e-06,
+      "loss": 1.0409,
+      "step": 2903
+    },
+    {
+      "epoch": 0.6649112764739553,
+      "grad_norm": 1.0357317924499512,
+      "learning_rate": 5.353020914470353e-06,
+      "loss": 1.0741,
+      "step": 2904
+    },
+    {
+      "epoch": 0.665140240412135,
+      "grad_norm": 1.323845624923706,
+      "learning_rate": 5.346453636882939e-06,
+      "loss": 1.1121,
+      "step": 2905
+    },
+    {
+      "epoch": 0.6653692043503149,
+      "grad_norm": 1.1600542068481445,
+      "learning_rate": 5.339888920103074e-06,
+      "loss": 1.0454,
+      "step": 2906
+    },
+    {
+      "epoch": 0.6655981682884946,
+      "grad_norm": 1.1847002506256104,
+      "learning_rate": 5.333326767743263e-06,
+      "loss": 1.0733,
+      "step": 2907
+    },
+    {
+      "epoch": 0.6658271322266743,
+      "grad_norm": 1.0907032489776611,
+      "learning_rate": 5.326767183414609e-06,
+      "loss": 1.0321,
+      "step": 2908
+    },
+    {
+      "epoch": 0.6660560961648541,
+      "grad_norm": 1.4598631858825684,
+      "learning_rate": 5.320210170726796e-06,
+      "loss": 1.0596,
+      "step": 2909
+    },
+    {
+      "epoch": 0.6662850601030338,
+      "grad_norm": 1.2866328954696655,
+      "learning_rate": 5.313655733288083e-06,
+      "loss": 1.0658,
+      "step": 2910
+    },
+    {
+      "epoch": 0.6665140240412135,
+      "grad_norm": 1.1192216873168945,
+      "learning_rate": 5.307103874705335e-06,
+      "loss": 1.0721,
+      "step": 2911
+    },
+    {
+      "epoch": 0.6667429879793932,
+      "grad_norm": 1.190312385559082,
+      "learning_rate": 5.300554598583982e-06,
+      "loss": 1.0158,
+      "step": 2912
+    },
+    {
+      "epoch": 0.666971951917573,
+      "grad_norm": 1.0897172689437866,
+      "learning_rate": 5.294007908528029e-06,
+      "loss": 0.9763,
+      "step": 2913
+    },
+    {
+      "epoch": 0.6672009158557527,
+      "grad_norm": 1.4546432495117188,
+      "learning_rate": 5.287463808140069e-06,
+      "loss": 1.0753,
+      "step": 2914
+    },
+    {
+      "epoch": 0.6674298797939324,
+      "grad_norm": 1.2672525644302368,
+      "learning_rate": 5.280922301021267e-06,
+      "loss": 1.0099,
+      "step": 2915
+    },
+    {
+      "epoch": 0.6676588437321122,
+      "grad_norm": 1.5939594507217407,
+      "learning_rate": 5.274383390771356e-06,
+      "loss": 1.1524,
+      "step": 2916
+    },
+    {
+      "epoch": 0.6678878076702919,
+      "grad_norm": 1.3578407764434814,
+      "learning_rate": 5.267847080988647e-06,
+      "loss": 0.9973,
+      "step": 2917
+    },
+    {
+      "epoch": 0.6681167716084717,
+      "grad_norm": 1.221137285232544,
+      "learning_rate": 5.2613133752700145e-06,
+      "loss": 1.0632,
+      "step": 2918
+    },
+    {
+      "epoch": 0.6683457355466514,
+      "grad_norm": 1.5474742650985718,
+      "learning_rate": 5.254782277210901e-06,
+      "loss": 1.0569,
+      "step": 2919
+    },
+    {
+      "epoch": 0.6685746994848312,
+      "grad_norm": 1.647705078125,
+      "learning_rate": 5.2482537904053185e-06,
+      "loss": 1.0202,
+      "step": 2920
+    },
+    {
+      "epoch": 0.6688036634230109,
+      "grad_norm": 1.6697813272476196,
+      "learning_rate": 5.241727918445836e-06,
+      "loss": 1.0308,
+      "step": 2921
+    },
+    {
+      "epoch": 0.6690326273611906,
+      "grad_norm": 1.1288517713546753,
+      "learning_rate": 5.235204664923586e-06,
+      "loss": 1.0139,
+      "step": 2922
+    },
+    {
+      "epoch": 0.6692615912993704,
+      "grad_norm": 0.9977933168411255,
+      "learning_rate": 5.228684033428265e-06,
+      "loss": 1.0381,
+      "step": 2923
+    },
+    {
+      "epoch": 0.6694905552375501,
+      "grad_norm": 1.2061574459075928,
+      "learning_rate": 5.22216602754811e-06,
+      "loss": 1.0258,
+      "step": 2924
+    },
+    {
+      "epoch": 0.6697195191757298,
+      "grad_norm": 1.1679553985595703,
+      "learning_rate": 5.215650650869941e-06,
+      "loss": 1.0316,
+      "step": 2925
+    },
+    {
+      "epoch": 0.6699484831139095,
+      "grad_norm": 1.209476351737976,
+      "learning_rate": 5.209137906979102e-06,
+      "loss": 1.0115,
+      "step": 2926
+    },
+    {
+      "epoch": 0.6701774470520893,
+      "grad_norm": 1.3832190036773682,
+      "learning_rate": 5.202627799459503e-06,
+      "loss": 1.0837,
+      "step": 2927
+    },
+    {
+      "epoch": 0.670406410990269,
+      "grad_norm": 1.2740558385849,
+      "learning_rate": 5.1961203318936116e-06,
+      "loss": 1.0364,
+      "step": 2928
+    },
+    {
+      "epoch": 0.6706353749284488,
+      "grad_norm": 1.1675924062728882,
+      "learning_rate": 5.1896155078624225e-06,
+      "loss": 1.0948,
+      "step": 2929
+    },
+    {
+      "epoch": 0.6708643388666286,
+      "grad_norm": 1.704378604888916,
+      "learning_rate": 5.183113330945488e-06,
+      "loss": 1.0411,
+      "step": 2930
+    },
+    {
+      "epoch": 0.6710933028048083,
+      "grad_norm": 1.6397212743759155,
+      "learning_rate": 5.176613804720905e-06,
+      "loss": 1.0228,
+      "step": 2931
+    },
+    {
+      "epoch": 0.671322266742988,
+      "grad_norm": 1.287341833114624,
+      "learning_rate": 5.170116932765304e-06,
+      "loss": 1.0504,
+      "step": 2932
+    },
+    {
+      "epoch": 0.6715512306811677,
+      "grad_norm": 1.3021739721298218,
+      "learning_rate": 5.1636227186538625e-06,
+      "loss": 1.0174,
+      "step": 2933
+    },
+    {
+      "epoch": 0.6717801946193475,
+      "grad_norm": 1.6247448921203613,
+      "learning_rate": 5.157131165960289e-06,
+      "loss": 1.0317,
+      "step": 2934
+    },
+    {
+      "epoch": 0.6720091585575272,
+      "grad_norm": 2.791722536087036,
+      "learning_rate": 5.1506422782568345e-06,
+      "loss": 1.0632,
+      "step": 2935
+    },
+    {
+      "epoch": 0.6722381224957069,
+      "grad_norm": 1.033670425415039,
+      "learning_rate": 5.144156059114279e-06,
+      "loss": 1.0681,
+      "step": 2936
+    },
+    {
+      "epoch": 0.6724670864338866,
+      "grad_norm": 1.6212668418884277,
+      "learning_rate": 5.137672512101925e-06,
+      "loss": 1.0611,
+      "step": 2937
+    },
+    {
+      "epoch": 0.6726960503720664,
+      "grad_norm": 1.806549310684204,
+      "learning_rate": 5.131191640787627e-06,
+      "loss": 1.0404,
+      "step": 2938
+    },
+    {
+      "epoch": 0.6729250143102461,
+      "grad_norm": 1.3719223737716675,
+      "learning_rate": 5.124713448737753e-06,
+      "loss": 1.0456,
+      "step": 2939
+    },
+    {
+      "epoch": 0.6731539782484258,
+      "grad_norm": 1.4912806749343872,
+      "learning_rate": 5.11823793951719e-06,
+      "loss": 1.0434,
+      "step": 2940
+    },
+    {
+      "epoch": 0.6733829421866057,
+      "grad_norm": 1.3837188482284546,
+      "learning_rate": 5.111765116689355e-06,
+      "loss": 1.024,
+      "step": 2941
+    },
+    {
+      "epoch": 0.6736119061247854,
+      "grad_norm": 1.8850486278533936,
+      "learning_rate": 5.105294983816203e-06,
+      "loss": 1.0059,
+      "step": 2942
+    },
+    {
+      "epoch": 0.6738408700629651,
+      "grad_norm": 1.2506767511367798,
+      "learning_rate": 5.098827544458178e-06,
+      "loss": 1.0762,
+      "step": 2943
+    },
+    {
+      "epoch": 0.6740698340011448,
+      "grad_norm": 1.9231312274932861,
+      "learning_rate": 5.0923628021742644e-06,
+      "loss": 1.0204,
+      "step": 2944
+    },
+    {
+      "epoch": 0.6742987979393246,
+      "grad_norm": 1.2224328517913818,
+      "learning_rate": 5.085900760521955e-06,
+      "loss": 1.0187,
+      "step": 2945
+    },
+    {
+      "epoch": 0.6745277618775043,
+      "grad_norm": 1.4136698246002197,
+      "learning_rate": 5.079441423057259e-06,
+      "loss": 1.0467,
+      "step": 2946
+    },
+    {
+      "epoch": 0.674756725815684,
+      "grad_norm": 1.244694709777832,
+      "learning_rate": 5.072984793334696e-06,
+      "loss": 1.0298,
+      "step": 2947
+    },
+    {
+      "epoch": 0.6749856897538637,
+      "grad_norm": 1.417830467224121,
+      "learning_rate": 5.066530874907285e-06,
+      "loss": 1.0746,
+      "step": 2948
+    },
+    {
+      "epoch": 0.6752146536920435,
+      "grad_norm": 1.1816048622131348,
+      "learning_rate": 5.060079671326577e-06,
+      "loss": 1.0203,
+      "step": 2949
+    },
+    {
+      "epoch": 0.6754436176302232,
+      "grad_norm": 1.4444321393966675,
+      "learning_rate": 5.053631186142612e-06,
+      "loss": 1.0502,
+      "step": 2950
+    },
+    {
+      "epoch": 0.6756725815684029,
+      "grad_norm": 1.1188122034072876,
+      "learning_rate": 5.0471854229039286e-06,
+      "loss": 1.0968,
+      "step": 2951
+    },
+    {
+      "epoch": 0.6759015455065828,
+      "grad_norm": 1.4322268962860107,
+      "learning_rate": 5.040742385157584e-06,
+      "loss": 1.0962,
+      "step": 2952
+    },
+    {
+      "epoch": 0.6761305094447625,
+      "grad_norm": 1.2078114748001099,
+      "learning_rate": 5.034302076449132e-06,
+      "loss": 1.0694,
+      "step": 2953
+    },
+    {
+      "epoch": 0.6763594733829422,
+      "grad_norm": 4.1897292137146,
+      "learning_rate": 5.027864500322611e-06,
+      "loss": 1.0731,
+      "step": 2954
+    },
+    {
+      "epoch": 0.676588437321122,
+      "grad_norm": 1.3412197828292847,
+      "learning_rate": 5.021429660320565e-06,
+      "loss": 1.074,
+      "step": 2955
+    },
+    {
+      "epoch": 0.6768174012593017,
+      "grad_norm": 2.866015911102295,
+      "learning_rate": 5.014997559984045e-06,
+      "loss": 0.9959,
+      "step": 2956
+    },
+    {
+      "epoch": 0.6770463651974814,
+      "grad_norm": 1.168531894683838,
+      "learning_rate": 5.008568202852569e-06,
+      "loss": 1.0117,
+      "step": 2957
+    },
+    {
+      "epoch": 0.6772753291356611,
+      "grad_norm": 1.4421766996383667,
+      "learning_rate": 5.002141592464162e-06,
+      "loss": 1.0038,
+      "step": 2958
+    },
+    {
+      "epoch": 0.6775042930738409,
+      "grad_norm": 1.2987374067306519,
+      "learning_rate": 4.995717732355335e-06,
+      "loss": 1.03,
+      "step": 2959
+    },
+    {
+      "epoch": 0.6777332570120206,
+      "grad_norm": 1.224198818206787,
+      "learning_rate": 4.989296626061084e-06,
+      "loss": 1.0346,
+      "step": 2960
+    },
+    {
+      "epoch": 0.6779622209502003,
+      "grad_norm": 2.0996627807617188,
+      "learning_rate": 4.982878277114891e-06,
+      "loss": 1.0766,
+      "step": 2961
+    },
+    {
+      "epoch": 0.67819118488838,
+      "grad_norm": 1.3319966793060303,
+      "learning_rate": 4.976462689048718e-06,
+      "loss": 1.002,
+      "step": 2962
+    },
+    {
+      "epoch": 0.6784201488265598,
+      "grad_norm": 1.1799633502960205,
+      "learning_rate": 4.970049865393009e-06,
+      "loss": 1.0168,
+      "step": 2963
+    },
+    {
+      "epoch": 0.6786491127647396,
+      "grad_norm": 1.1029752492904663,
+      "learning_rate": 4.963639809676692e-06,
+      "loss": 1.0343,
+      "step": 2964
+    },
+    {
+      "epoch": 0.6788780767029193,
+      "grad_norm": 1.0645133256912231,
+      "learning_rate": 4.957232525427156e-06,
+      "loss": 1.0062,
+      "step": 2965
+    },
+    {
+      "epoch": 0.6791070406410991,
+      "grad_norm": 1.2702126502990723,
+      "learning_rate": 4.950828016170286e-06,
+      "loss": 1.0674,
+      "step": 2966
+    },
+    {
+      "epoch": 0.6793360045792788,
+      "grad_norm": 1.3423794507980347,
+      "learning_rate": 4.94442628543043e-06,
+      "loss": 1.0065,
+      "step": 2967
+    },
+    {
+      "epoch": 0.6795649685174585,
+      "grad_norm": 1.3491889238357544,
+      "learning_rate": 4.9380273367304e-06,
+      "loss": 1.0039,
+      "step": 2968
+    },
+    {
+      "epoch": 0.6797939324556382,
+      "grad_norm": 1.332449197769165,
+      "learning_rate": 4.931631173591487e-06,
+      "loss": 1.015,
+      "step": 2969
+    },
+    {
+      "epoch": 0.680022896393818,
+      "grad_norm": 1.2349896430969238,
+      "learning_rate": 4.925237799533445e-06,
+      "loss": 1.0143,
+      "step": 2970
+    },
+    {
+      "epoch": 0.6802518603319977,
+      "grad_norm": 1.296811580657959,
+      "learning_rate": 4.918847218074495e-06,
+      "loss": 1.0468,
+      "step": 2971
+    },
+    {
+      "epoch": 0.6804808242701774,
+      "grad_norm": 1.2404383420944214,
+      "learning_rate": 4.912459432731322e-06,
+      "loss": 1.0548,
+      "step": 2972
+    },
+    {
+      "epoch": 0.6807097882083571,
+      "grad_norm": 1.2925716638565063,
+      "learning_rate": 4.906074447019068e-06,
+      "loss": 1.0182,
+      "step": 2973
+    },
+    {
+      "epoch": 0.6809387521465369,
+      "grad_norm": 1.7651426792144775,
+      "learning_rate": 4.899692264451339e-06,
+      "loss": 1.028,
+      "step": 2974
+    },
+    {
+      "epoch": 0.6811677160847167,
+      "grad_norm": 2.1795103549957275,
+      "learning_rate": 4.893312888540195e-06,
+      "loss": 1.0185,
+      "step": 2975
+    },
+    {
+      "epoch": 0.6813966800228964,
+      "grad_norm": 1.4040412902832031,
+      "learning_rate": 4.886936322796154e-06,
+      "loss": 1.015,
+      "step": 2976
+    },
+    {
+      "epoch": 0.6816256439610762,
+      "grad_norm": 2.831342935562134,
+      "learning_rate": 4.880562570728188e-06,
+      "loss": 1.0135,
+      "step": 2977
+    },
+    {
+      "epoch": 0.6818546078992559,
+      "grad_norm": 1.828460454940796,
+      "learning_rate": 4.87419163584372e-06,
+      "loss": 1.0305,
+      "step": 2978
+    },
+    {
+      "epoch": 0.6820835718374356,
+      "grad_norm": 1.3774365186691284,
+      "learning_rate": 4.867823521648613e-06,
+      "loss": 1.0625,
+      "step": 2979
+    },
+    {
+      "epoch": 0.6823125357756153,
+      "grad_norm": 1.7624365091323853,
+      "learning_rate": 4.861458231647202e-06,
+      "loss": 1.0516,
+      "step": 2980
+    },
+    {
+      "epoch": 0.6825414997137951,
+      "grad_norm": 1.3166933059692383,
+      "learning_rate": 4.855095769342241e-06,
+      "loss": 1.0544,
+      "step": 2981
+    },
+    {
+      "epoch": 0.6827704636519748,
+      "grad_norm": 2.098736047744751,
+      "learning_rate": 4.848736138234943e-06,
+      "loss": 1.0258,
+      "step": 2982
+    },
+    {
+      "epoch": 0.6829994275901545,
+      "grad_norm": 1.0946608781814575,
+      "learning_rate": 4.842379341824958e-06,
+      "loss": 1.0407,
+      "step": 2983
+    },
+    {
+      "epoch": 0.6832283915283343,
+      "grad_norm": 1.1885132789611816,
+      "learning_rate": 4.836025383610382e-06,
+      "loss": 1.0462,
+      "step": 2984
+    },
+    {
+      "epoch": 0.683457355466514,
+      "grad_norm": 5.834176540374756,
+      "learning_rate": 4.829674267087742e-06,
+      "loss": 1.0261,
+      "step": 2985
+    },
+    {
+      "epoch": 0.6836863194046937,
+      "grad_norm": 3.0932862758636475,
+      "learning_rate": 4.823325995752005e-06,
+      "loss": 1.0829,
+      "step": 2986
+    },
+    {
+      "epoch": 0.6839152833428735,
+      "grad_norm": 1.27007257938385,
+      "learning_rate": 4.816980573096571e-06,
+      "loss": 1.1001,
+      "step": 2987
+    },
+    {
+      "epoch": 0.6841442472810533,
+      "grad_norm": 1.145909070968628,
+      "learning_rate": 4.810638002613273e-06,
+      "loss": 1.0904,
+      "step": 2988
+    },
+    {
+      "epoch": 0.684373211219233,
+      "grad_norm": 1.4148883819580078,
+      "learning_rate": 4.804298287792374e-06,
+      "loss": 1.0352,
+      "step": 2989
+    },
+    {
+      "epoch": 0.6846021751574127,
+      "grad_norm": 1.51638925075531,
+      "learning_rate": 4.797961432122568e-06,
+      "loss": 1.0851,
+      "step": 2990
+    },
+    {
+      "epoch": 0.6848311390955925,
+      "grad_norm": 2.6201398372650146,
+      "learning_rate": 4.791627439090975e-06,
+      "loss": 1.0082,
+      "step": 2991
+    },
+    {
+      "epoch": 0.6850601030337722,
+      "grad_norm": 1.3478920459747314,
+      "learning_rate": 4.785296312183131e-06,
+      "loss": 1.0192,
+      "step": 2992
+    },
+    {
+      "epoch": 0.6852890669719519,
+      "grad_norm": 2.5813474655151367,
+      "learning_rate": 4.778968054883002e-06,
+      "loss": 1.0566,
+      "step": 2993
+    },
+    {
+      "epoch": 0.6855180309101316,
+      "grad_norm": 1.4003512859344482,
+      "learning_rate": 4.772642670672988e-06,
+      "loss": 1.0256,
+      "step": 2994
+    },
+    {
+      "epoch": 0.6857469948483114,
+      "grad_norm": 1.65157151222229,
+      "learning_rate": 4.766320163033882e-06,
+      "loss": 1.081,
+      "step": 2995
+    },
+    {
+      "epoch": 0.6859759587864911,
+      "grad_norm": 1.354181170463562,
+      "learning_rate": 4.7600005354449075e-06,
+      "loss": 0.9829,
+      "step": 2996
+    },
+    {
+      "epoch": 0.6862049227246708,
+      "grad_norm": 1.2161715030670166,
+      "learning_rate": 4.753683791383713e-06,
+      "loss": 1.0763,
+      "step": 2997
+    },
+    {
+      "epoch": 0.6864338866628507,
+      "grad_norm": 1.2519667148590088,
+      "learning_rate": 4.74736993432634e-06,
+      "loss": 1.0668,
+      "step": 2998
+    },
+    {
+      "epoch": 0.6866628506010304,
+      "grad_norm": 1.2776641845703125,
+      "learning_rate": 4.741058967747254e-06,
+      "loss": 1.0412,
+      "step": 2999
+    },
+    {
+      "epoch": 0.6868918145392101,
+      "grad_norm": 1.4053977727890015,
+      "learning_rate": 4.734750895119327e-06,
+      "loss": 1.079,
+      "step": 3000
+    },
+    {
+      "epoch": 0.6871207784773898,
+      "grad_norm": 1.4398205280303955,
+      "learning_rate": 4.7284457199138374e-06,
+      "loss": 1.0251,
+      "step": 3001
+    },
+    {
+      "epoch": 0.6873497424155696,
+      "grad_norm": 1.4499075412750244,
+      "learning_rate": 4.722143445600477e-06,
+      "loss": 1.0771,
+      "step": 3002
+    },
+    {
+      "epoch": 0.6875787063537493,
+      "grad_norm": 1.4174058437347412,
+      "learning_rate": 4.71584407564732e-06,
+      "loss": 1.0093,
+      "step": 3003
+    },
+    {
+      "epoch": 0.687807670291929,
+      "grad_norm": 1.425048589706421,
+      "learning_rate": 4.70954761352087e-06,
+      "loss": 1.0117,
+      "step": 3004
+    },
+    {
+      "epoch": 0.6880366342301087,
+      "grad_norm": 1.1394058465957642,
+      "learning_rate": 4.703254062686017e-06,
+      "loss": 1.0435,
+      "step": 3005
+    },
+    {
+      "epoch": 0.6882655981682885,
+      "grad_norm": 1.4277303218841553,
+      "learning_rate": 4.696963426606041e-06,
+      "loss": 1.0903,
+      "step": 3006
+    },
+    {
+      "epoch": 0.6884945621064682,
+      "grad_norm": 1.744752287864685,
+      "learning_rate": 4.690675708742628e-06,
+      "loss": 1.0468,
+      "step": 3007
+    },
+    {
+      "epoch": 0.6887235260446479,
+      "grad_norm": 1.4332195520401,
+      "learning_rate": 4.684390912555866e-06,
+      "loss": 1.0169,
+      "step": 3008
+    },
+    {
+      "epoch": 0.6889524899828277,
+      "grad_norm": 1.3215627670288086,
+      "learning_rate": 4.678109041504215e-06,
+      "loss": 1.0733,
+      "step": 3009
+    },
+    {
+      "epoch": 0.6891814539210075,
+      "grad_norm": 1.8956875801086426,
+      "learning_rate": 4.671830099044536e-06,
+      "loss": 1.0317,
+      "step": 3010
+    },
+    {
+      "epoch": 0.6894104178591872,
+      "grad_norm": 1.260425090789795,
+      "learning_rate": 4.665554088632089e-06,
+      "loss": 1.0633,
+      "step": 3011
+    },
+    {
+      "epoch": 0.689639381797367,
+      "grad_norm": 1.27289617061615,
+      "learning_rate": 4.6592810137205e-06,
+      "loss": 1.0074,
+      "step": 3012
+    },
+    {
+      "epoch": 0.6898683457355467,
+      "grad_norm": 1.8777307271957397,
+      "learning_rate": 4.653010877761793e-06,
+      "loss": 1.0374,
+      "step": 3013
+    },
+    {
+      "epoch": 0.6900973096737264,
+      "grad_norm": 1.1621068716049194,
+      "learning_rate": 4.64674368420637e-06,
+      "loss": 1.0394,
+      "step": 3014
+    },
+    {
+      "epoch": 0.6903262736119061,
+      "grad_norm": 1.2484198808670044,
+      "learning_rate": 4.640479436503016e-06,
+      "loss": 1.0031,
+      "step": 3015
+    },
+    {
+      "epoch": 0.6905552375500859,
+      "grad_norm": 1.564257025718689,
+      "learning_rate": 4.634218138098897e-06,
+      "loss": 1.0898,
+      "step": 3016
+    },
+    {
+      "epoch": 0.6907842014882656,
+      "grad_norm": 1.2701921463012695,
+      "learning_rate": 4.6279597924395434e-06,
+      "loss": 1.0803,
+      "step": 3017
+    },
+    {
+      "epoch": 0.6910131654264453,
+      "grad_norm": 1.458882212638855,
+      "learning_rate": 4.621704402968881e-06,
+      "loss": 1.0771,
+      "step": 3018
+    },
+    {
+      "epoch": 0.691242129364625,
+      "grad_norm": 1.2890880107879639,
+      "learning_rate": 4.615451973129196e-06,
+      "loss": 1.0809,
+      "step": 3019
+    },
+    {
+      "epoch": 0.6914710933028048,
+      "grad_norm": 1.1623772382736206,
+      "learning_rate": 4.60920250636114e-06,
+      "loss": 1.0524,
+      "step": 3020
+    },
+    {
+      "epoch": 0.6917000572409846,
+      "grad_norm": 1.1945689916610718,
+      "learning_rate": 4.602956006103752e-06,
+      "loss": 1.0345,
+      "step": 3021
+    },
+    {
+      "epoch": 0.6919290211791643,
+      "grad_norm": 1.1116697788238525,
+      "learning_rate": 4.596712475794427e-06,
+      "loss": 0.9965,
+      "step": 3022
+    },
+    {
+      "epoch": 0.692157985117344,
+      "grad_norm": 1.092020869255066,
+      "learning_rate": 4.590471918868923e-06,
+      "loss": 1.019,
+      "step": 3023
+    },
+    {
+      "epoch": 0.6923869490555238,
+      "grad_norm": 1.2264333963394165,
+      "learning_rate": 4.584234338761368e-06,
+      "loss": 1.0246,
+      "step": 3024
+    },
+    {
+      "epoch": 0.6926159129937035,
+      "grad_norm": 1.1611523628234863,
+      "learning_rate": 4.5779997389042514e-06,
+      "loss": 1.0514,
+      "step": 3025
+    },
+    {
+      "epoch": 0.6928448769318832,
+      "grad_norm": 1.5043096542358398,
+      "learning_rate": 4.571768122728421e-06,
+      "loss": 0.9878,
+      "step": 3026
+    },
+    {
+      "epoch": 0.693073840870063,
+      "grad_norm": 1.4404031038284302,
+      "learning_rate": 4.56553949366308e-06,
+      "loss": 1.0665,
+      "step": 3027
+    },
+    {
+      "epoch": 0.6933028048082427,
+      "grad_norm": 1.6433451175689697,
+      "learning_rate": 4.559313855135795e-06,
+      "loss": 1.0795,
+      "step": 3028
+    },
+    {
+      "epoch": 0.6935317687464224,
+      "grad_norm": 1.2414443492889404,
+      "learning_rate": 4.55309121057248e-06,
+      "loss": 1.0612,
+      "step": 3029
+    },
+    {
+      "epoch": 0.6937607326846021,
+      "grad_norm": 1.7606821060180664,
+      "learning_rate": 4.546871563397409e-06,
+      "loss": 1.0517,
+      "step": 3030
+    },
+    {
+      "epoch": 0.6939896966227819,
+      "grad_norm": 1.563098430633545,
+      "learning_rate": 4.54065491703319e-06,
+      "loss": 1.0475,
+      "step": 3031
+    },
+    {
+      "epoch": 0.6942186605609616,
+      "grad_norm": 1.3073487281799316,
+      "learning_rate": 4.534441274900807e-06,
+      "loss": 1.0028,
+      "step": 3032
+    },
+    {
+      "epoch": 0.6944476244991414,
+      "grad_norm": 1.2197542190551758,
+      "learning_rate": 4.528230640419562e-06,
+      "loss": 1.027,
+      "step": 3033
+    },
+    {
+      "epoch": 0.6946765884373212,
+      "grad_norm": 1.2171616554260254,
+      "learning_rate": 4.522023017007118e-06,
+      "loss": 1.0114,
+      "step": 3034
+    },
+    {
+      "epoch": 0.6949055523755009,
+      "grad_norm": 1.2713412046432495,
+      "learning_rate": 4.515818408079487e-06,
+      "loss": 1.0541,
+      "step": 3035
+    },
+    {
+      "epoch": 0.6951345163136806,
+      "grad_norm": 1.235053539276123,
+      "learning_rate": 4.5096168170510036e-06,
+      "loss": 1.0338,
+      "step": 3036
+    },
+    {
+      "epoch": 0.6953634802518603,
+      "grad_norm": 1.3829389810562134,
+      "learning_rate": 4.503418247334353e-06,
+      "loss": 1.0417,
+      "step": 3037
+    },
+    {
+      "epoch": 0.6955924441900401,
+      "grad_norm": 1.107755422592163,
+      "learning_rate": 4.497222702340559e-06,
+      "loss": 1.0538,
+      "step": 3038
+    },
+    {
+      "epoch": 0.6958214081282198,
+      "grad_norm": 1.2136812210083008,
+      "learning_rate": 4.491030185478976e-06,
+      "loss": 1.0574,
+      "step": 3039
+    },
+    {
+      "epoch": 0.6960503720663995,
+      "grad_norm": 1.9772335290908813,
+      "learning_rate": 4.4848407001572945e-06,
+      "loss": 1.0413,
+      "step": 3040
+    },
+    {
+      "epoch": 0.6962793360045793,
+      "grad_norm": 1.3861563205718994,
+      "learning_rate": 4.4786542497815365e-06,
+      "loss": 1.0276,
+      "step": 3041
+    },
+    {
+      "epoch": 0.696508299942759,
+      "grad_norm": 1.2986854314804077,
+      "learning_rate": 4.472470837756055e-06,
+      "loss": 1.0022,
+      "step": 3042
+    },
+    {
+      "epoch": 0.6967372638809387,
+      "grad_norm": 1.8411191701889038,
+      "learning_rate": 4.466290467483531e-06,
+      "loss": 1.0107,
+      "step": 3043
+    },
+    {
+      "epoch": 0.6969662278191184,
+      "grad_norm": 1.371367335319519,
+      "learning_rate": 4.460113142364961e-06,
+      "loss": 1.017,
+      "step": 3044
+    },
+    {
+      "epoch": 0.6971951917572983,
+      "grad_norm": 1.3172252178192139,
+      "learning_rate": 4.453938865799686e-06,
+      "loss": 1.0403,
+      "step": 3045
+    },
+    {
+      "epoch": 0.697424155695478,
+      "grad_norm": 1.2504808902740479,
+      "learning_rate": 4.447767641185359e-06,
+      "loss": 1.0093,
+      "step": 3046
+    },
+    {
+      "epoch": 0.6976531196336577,
+      "grad_norm": 1.1979867219924927,
+      "learning_rate": 4.441599471917946e-06,
+      "loss": 1.0421,
+      "step": 3047
+    },
+    {
+      "epoch": 0.6978820835718375,
+      "grad_norm": 1.6098335981369019,
+      "learning_rate": 4.435434361391737e-06,
+      "loss": 1.091,
+      "step": 3048
+    },
+    {
+      "epoch": 0.6981110475100172,
+      "grad_norm": 1.1873350143432617,
+      "learning_rate": 4.429272312999353e-06,
+      "loss": 1.0796,
+      "step": 3049
+    },
+    {
+      "epoch": 0.6983400114481969,
+      "grad_norm": 1.3173235654830933,
+      "learning_rate": 4.423113330131708e-06,
+      "loss": 1.0352,
+      "step": 3050
+    },
+    {
+      "epoch": 0.6985689753863766,
+      "grad_norm": 1.3406671285629272,
+      "learning_rate": 4.4169574161780395e-06,
+      "loss": 1.0448,
+      "step": 3051
+    },
+    {
+      "epoch": 0.6987979393245564,
+      "grad_norm": 1.2191475629806519,
+      "learning_rate": 4.4108045745258966e-06,
+      "loss": 1.0402,
+      "step": 3052
+    },
+    {
+      "epoch": 0.6990269032627361,
+      "grad_norm": 1.2168577909469604,
+      "learning_rate": 4.404654808561137e-06,
+      "loss": 1.0253,
+      "step": 3053
+    },
+    {
+      "epoch": 0.6992558672009158,
+      "grad_norm": 9.12889575958252,
+      "learning_rate": 4.398508121667925e-06,
+      "loss": 1.0162,
+      "step": 3054
+    },
+    {
+      "epoch": 0.6994848311390955,
+      "grad_norm": 1.6077773571014404,
+      "learning_rate": 4.39236451722873e-06,
+      "loss": 1.0926,
+      "step": 3055
+    },
+    {
+      "epoch": 0.6997137950772754,
+      "grad_norm": 1.5294177532196045,
+      "learning_rate": 4.386223998624327e-06,
+      "loss": 1.0349,
+      "step": 3056
+    },
+    {
+      "epoch": 0.6999427590154551,
+      "grad_norm": 1.079950213432312,
+      "learning_rate": 4.380086569233796e-06,
+      "loss": 0.9882,
+      "step": 3057
+    },
+    {
+      "epoch": 0.7001717229536348,
+      "grad_norm": 1.1813180446624756,
+      "learning_rate": 4.3739522324344994e-06,
+      "loss": 1.1365,
+      "step": 3058
+    },
+    {
+      "epoch": 0.7004006868918146,
+      "grad_norm": 1.2818435430526733,
+      "learning_rate": 4.3678209916021264e-06,
+      "loss": 1.0429,
+      "step": 3059
+    },
+    {
+      "epoch": 0.7006296508299943,
+      "grad_norm": 1.6467807292938232,
+      "learning_rate": 4.361692850110644e-06,
+      "loss": 1.0304,
+      "step": 3060
+    },
+    {
+      "epoch": 0.700858614768174,
+      "grad_norm": 1.1530770063400269,
+      "learning_rate": 4.355567811332311e-06,
+      "loss": 1.1026,
+      "step": 3061
+    },
+    {
+      "epoch": 0.7010875787063537,
+      "grad_norm": 1.2358185052871704,
+      "learning_rate": 4.3494458786376845e-06,
+      "loss": 1.0106,
+      "step": 3062
+    },
+    {
+      "epoch": 0.7013165426445335,
+      "grad_norm": 2.5719728469848633,
+      "learning_rate": 4.3433270553956245e-06,
+      "loss": 0.9823,
+      "step": 3063
+    },
+    {
+      "epoch": 0.7015455065827132,
+      "grad_norm": 1.4278590679168701,
+      "learning_rate": 4.337211344973256e-06,
+      "loss": 1.0915,
+      "step": 3064
+    },
+    {
+      "epoch": 0.7017744705208929,
+      "grad_norm": 1.1447930335998535,
+      "learning_rate": 4.331098750736008e-06,
+      "loss": 1.0208,
+      "step": 3065
+    },
+    {
+      "epoch": 0.7020034344590727,
+      "grad_norm": 1.5032339096069336,
+      "learning_rate": 4.3249892760475894e-06,
+      "loss": 1.0192,
+      "step": 3066
+    },
+    {
+      "epoch": 0.7022323983972524,
+      "grad_norm": 1.17005455493927,
+      "learning_rate": 4.318882924269994e-06,
+      "loss": 0.9827,
+      "step": 3067
+    },
+    {
+      "epoch": 0.7024613623354322,
+      "grad_norm": 1.24888014793396,
+      "learning_rate": 4.312779698763493e-06,
+      "loss": 1.0159,
+      "step": 3068
+    },
+    {
+      "epoch": 0.7026903262736119,
+      "grad_norm": 1.3852883577346802,
+      "learning_rate": 4.306679602886643e-06,
+      "loss": 1.0457,
+      "step": 3069
+    },
+    {
+      "epoch": 0.7029192902117917,
+      "grad_norm": 1.4950320720672607,
+      "learning_rate": 4.300582639996274e-06,
+      "loss": 1.0019,
+      "step": 3070
+    },
+    {
+      "epoch": 0.7031482541499714,
+      "grad_norm": 1.2468384504318237,
+      "learning_rate": 4.2944888134474995e-06,
+      "loss": 1.0566,
+      "step": 3071
+    },
+    {
+      "epoch": 0.7033772180881511,
+      "grad_norm": 1.0590832233428955,
+      "learning_rate": 4.2883981265936884e-06,
+      "loss": 1.0033,
+      "step": 3072
+    },
+    {
+      "epoch": 0.7036061820263309,
+      "grad_norm": 1.5195868015289307,
+      "learning_rate": 4.282310582786504e-06,
+      "loss": 1.0192,
+      "step": 3073
+    },
+    {
+      "epoch": 0.7038351459645106,
+      "grad_norm": 1.6989542245864868,
+      "learning_rate": 4.276226185375874e-06,
+      "loss": 1.0511,
+      "step": 3074
+    },
+    {
+      "epoch": 0.7040641099026903,
+      "grad_norm": 1.4525377750396729,
+      "learning_rate": 4.270144937709981e-06,
+      "loss": 1.0511,
+      "step": 3075
+    },
+    {
+      "epoch": 0.70429307384087,
+      "grad_norm": 1.3085981607437134,
+      "learning_rate": 4.264066843135292e-06,
+      "loss": 1.0334,
+      "step": 3076
+    },
+    {
+      "epoch": 0.7045220377790498,
+      "grad_norm": 1.2663421630859375,
+      "learning_rate": 4.257991904996527e-06,
+      "loss": 1.0987,
+      "step": 3077
+    },
+    {
+      "epoch": 0.7047510017172295,
+      "grad_norm": 1.1956284046173096,
+      "learning_rate": 4.251920126636676e-06,
+      "loss": 1.0384,
+      "step": 3078
+    },
+    {
+      "epoch": 0.7049799656554093,
+      "grad_norm": 1.3076503276824951,
+      "learning_rate": 4.245851511396988e-06,
+      "loss": 1.1324,
+      "step": 3079
+    },
+    {
+      "epoch": 0.705208929593589,
+      "grad_norm": 1.2820242643356323,
+      "learning_rate": 4.23978606261697e-06,
+      "loss": 1.0065,
+      "step": 3080
+    },
+    {
+      "epoch": 0.7054378935317688,
+      "grad_norm": 1.707970142364502,
+      "learning_rate": 4.233723783634388e-06,
+      "loss": 1.0478,
+      "step": 3081
+    },
+    {
+      "epoch": 0.7056668574699485,
+      "grad_norm": 1.9648287296295166,
+      "learning_rate": 4.227664677785264e-06,
+      "loss": 0.9916,
+      "step": 3082
+    },
+    {
+      "epoch": 0.7058958214081282,
+      "grad_norm": 1.4425021409988403,
+      "learning_rate": 4.221608748403872e-06,
+      "loss": 1.0077,
+      "step": 3083
+    },
+    {
+      "epoch": 0.706124785346308,
+      "grad_norm": 1.1991877555847168,
+      "learning_rate": 4.21555599882274e-06,
+      "loss": 1.085,
+      "step": 3084
+    },
+    {
+      "epoch": 0.7063537492844877,
+      "grad_norm": 1.6011837720870972,
+      "learning_rate": 4.2095064323726485e-06,
+      "loss": 1.0511,
+      "step": 3085
+    },
+    {
+      "epoch": 0.7065827132226674,
+      "grad_norm": 1.2623403072357178,
+      "learning_rate": 4.2034600523826116e-06,
+      "loss": 0.992,
+      "step": 3086
+    },
+    {
+      "epoch": 0.7068116771608471,
+      "grad_norm": 1.3289179801940918,
+      "learning_rate": 4.1974168621799185e-06,
+      "loss": 1.0386,
+      "step": 3087
+    },
+    {
+      "epoch": 0.7070406410990269,
+      "grad_norm": 1.3888287544250488,
+      "learning_rate": 4.191376865090073e-06,
+      "loss": 1.0181,
+      "step": 3088
+    },
+    {
+      "epoch": 0.7072696050372066,
+      "grad_norm": 1.3241780996322632,
+      "learning_rate": 4.1853400644368395e-06,
+      "loss": 1.0285,
+      "step": 3089
+    },
+    {
+      "epoch": 0.7074985689753863,
+      "grad_norm": 1.2323023080825806,
+      "learning_rate": 4.179306463542217e-06,
+      "loss": 1.0286,
+      "step": 3090
+    },
+    {
+      "epoch": 0.7077275329135662,
+      "grad_norm": 1.290425419807434,
+      "learning_rate": 4.173276065726448e-06,
+      "loss": 1.0194,
+      "step": 3091
+    },
+    {
+      "epoch": 0.7079564968517459,
+      "grad_norm": 1.7319308519363403,
+      "learning_rate": 4.167248874308009e-06,
+      "loss": 1.0092,
+      "step": 3092
+    },
+    {
+      "epoch": 0.7081854607899256,
+      "grad_norm": 1.0908766984939575,
+      "learning_rate": 4.161224892603613e-06,
+      "loss": 1.0487,
+      "step": 3093
+    },
+    {
+      "epoch": 0.7084144247281053,
+      "grad_norm": 1.2904776334762573,
+      "learning_rate": 4.155204123928205e-06,
+      "loss": 0.9839,
+      "step": 3094
+    },
+    {
+      "epoch": 0.7086433886662851,
+      "grad_norm": 1.3205946683883667,
+      "learning_rate": 4.149186571594965e-06,
+      "loss": 1.0494,
+      "step": 3095
+    },
+    {
+      "epoch": 0.7088723526044648,
+      "grad_norm": 1.6448436975479126,
+      "learning_rate": 4.143172238915302e-06,
+      "loss": 1.0911,
+      "step": 3096
+    },
+    {
+      "epoch": 0.7091013165426445,
+      "grad_norm": 1.1375172138214111,
+      "learning_rate": 4.137161129198852e-06,
+      "loss": 0.9825,
+      "step": 3097
+    },
+    {
+      "epoch": 0.7093302804808242,
+      "grad_norm": 1.2753404378890991,
+      "learning_rate": 4.131153245753482e-06,
+      "loss": 1.0859,
+      "step": 3098
+    },
+    {
+      "epoch": 0.709559244419004,
+      "grad_norm": 1.3567954301834106,
+      "learning_rate": 4.1251485918852725e-06,
+      "loss": 1.029,
+      "step": 3099
+    },
+    {
+      "epoch": 0.7097882083571837,
+      "grad_norm": 1.057942271232605,
+      "learning_rate": 4.119147170898535e-06,
+      "loss": 1.0965,
+      "step": 3100
+    },
+    {
+      "epoch": 0.7100171722953634,
+      "grad_norm": 1.2383803129196167,
+      "learning_rate": 4.113148986095812e-06,
+      "loss": 1.0577,
+      "step": 3101
+    },
+    {
+      "epoch": 0.7102461362335433,
+      "grad_norm": 1.1256804466247559,
+      "learning_rate": 4.107154040777842e-06,
+      "loss": 0.9798,
+      "step": 3102
+    },
+    {
+      "epoch": 0.710475100171723,
+      "grad_norm": 1.5061252117156982,
+      "learning_rate": 4.101162338243595e-06,
+      "loss": 1.1058,
+      "step": 3103
+    },
+    {
+      "epoch": 0.7107040641099027,
+      "grad_norm": 1.094844102859497,
+      "learning_rate": 4.095173881790265e-06,
+      "loss": 0.9926,
+      "step": 3104
+    },
+    {
+      "epoch": 0.7109330280480824,
+      "grad_norm": 1.3939733505249023,
+      "learning_rate": 4.0891886747132356e-06,
+      "loss": 1.061,
+      "step": 3105
+    },
+    {
+      "epoch": 0.7111619919862622,
+      "grad_norm": 2.069545030593872,
+      "learning_rate": 4.0832067203061244e-06,
+      "loss": 1.0444,
+      "step": 3106
+    },
+    {
+      "epoch": 0.7113909559244419,
+      "grad_norm": 1.2419723272323608,
+      "learning_rate": 4.0772280218607485e-06,
+      "loss": 1.0661,
+      "step": 3107
+    },
+    {
+      "epoch": 0.7116199198626216,
+      "grad_norm": 2.2456400394439697,
+      "learning_rate": 4.071252582667135e-06,
+      "loss": 1.0287,
+      "step": 3108
+    },
+    {
+      "epoch": 0.7118488838008014,
+      "grad_norm": 1.2714403867721558,
+      "learning_rate": 4.065280406013522e-06,
+      "loss": 0.9891,
+      "step": 3109
+    },
+    {
+      "epoch": 0.7120778477389811,
+      "grad_norm": 1.4441180229187012,
+      "learning_rate": 4.059311495186338e-06,
+      "loss": 1.0688,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7123068116771608,
+      "grad_norm": 1.442212700843811,
+      "learning_rate": 4.0533458534702354e-06,
+      "loss": 1.0711,
+      "step": 3111
+    },
+    {
+      "epoch": 0.7125357756153405,
+      "grad_norm": 1.2680224180221558,
+      "learning_rate": 4.0473834841480565e-06,
+      "loss": 1.0097,
+      "step": 3112
+    },
+    {
+      "epoch": 0.7127647395535203,
+      "grad_norm": 1.0516926050186157,
+      "learning_rate": 4.0414243905008325e-06,
+      "loss": 1.0449,
+      "step": 3113
+    },
+    {
+      "epoch": 0.7129937034917001,
+      "grad_norm": 1.1603946685791016,
+      "learning_rate": 4.035468575807812e-06,
+      "loss": 1.0299,
+      "step": 3114
+    },
+    {
+      "epoch": 0.7132226674298798,
+      "grad_norm": 1.0879015922546387,
+      "learning_rate": 4.029516043346432e-06,
+      "loss": 1.0125,
+      "step": 3115
+    },
+    {
+      "epoch": 0.7134516313680596,
+      "grad_norm": 2.325087070465088,
+      "learning_rate": 4.023566796392313e-06,
+      "loss": 1.0965,
+      "step": 3116
+    },
+    {
+      "epoch": 0.7136805953062393,
+      "grad_norm": 1.1928337812423706,
+      "learning_rate": 4.017620838219276e-06,
+      "loss": 1.0108,
+      "step": 3117
+    },
+    {
+      "epoch": 0.713909559244419,
+      "grad_norm": 1.2268211841583252,
+      "learning_rate": 4.011678172099343e-06,
+      "loss": 0.9763,
+      "step": 3118
+    },
+    {
+      "epoch": 0.7141385231825987,
+      "grad_norm": 1.3314909934997559,
+      "learning_rate": 4.005738801302701e-06,
+      "loss": 1.0531,
+      "step": 3119
+    },
+    {
+      "epoch": 0.7143674871207785,
+      "grad_norm": 1.3090684413909912,
+      "learning_rate": 3.999802729097743e-06,
+      "loss": 1.0713,
+      "step": 3120
+    },
+    {
+      "epoch": 0.7145964510589582,
+      "grad_norm": 1.1635836362838745,
+      "learning_rate": 3.993869958751036e-06,
+      "loss": 1.0654,
+      "step": 3121
+    },
+    {
+      "epoch": 0.7148254149971379,
+      "grad_norm": 1.3543496131896973,
+      "learning_rate": 3.9879404935273346e-06,
+      "loss": 1.0163,
+      "step": 3122
+    },
+    {
+      "epoch": 0.7150543789353176,
+      "grad_norm": 1.3396997451782227,
+      "learning_rate": 3.982014336689579e-06,
+      "loss": 1.0441,
+      "step": 3123
+    },
+    {
+      "epoch": 0.7152833428734974,
+      "grad_norm": 1.4751423597335815,
+      "learning_rate": 3.9760914914988716e-06,
+      "loss": 1.0307,
+      "step": 3124
+    },
+    {
+      "epoch": 0.7155123068116772,
+      "grad_norm": 1.3802671432495117,
+      "learning_rate": 3.970171961214515e-06,
+      "loss": 1.0171,
+      "step": 3125
+    },
+    {
+      "epoch": 0.7157412707498569,
+      "grad_norm": 1.8824442625045776,
+      "learning_rate": 3.964255749093979e-06,
+      "loss": 0.969,
+      "step": 3126
+    },
+    {
+      "epoch": 0.7159702346880367,
+      "grad_norm": 1.3549731969833374,
+      "learning_rate": 3.958342858392893e-06,
+      "loss": 1.0393,
+      "step": 3127
+    },
+    {
+      "epoch": 0.7161991986262164,
+      "grad_norm": 1.1898926496505737,
+      "learning_rate": 3.952433292365081e-06,
+      "loss": 1.0948,
+      "step": 3128
+    },
+    {
+      "epoch": 0.7164281625643961,
+      "grad_norm": 1.109851360321045,
+      "learning_rate": 3.94652705426253e-06,
+      "loss": 1.0621,
+      "step": 3129
+    },
+    {
+      "epoch": 0.7166571265025758,
+      "grad_norm": 1.4158560037612915,
+      "learning_rate": 3.940624147335386e-06,
+      "loss": 1.0273,
+      "step": 3130
+    },
+    {
+      "epoch": 0.7168860904407556,
+      "grad_norm": 1.319413661956787,
+      "learning_rate": 3.93472457483197e-06,
+      "loss": 0.9908,
+      "step": 3131
+    },
+    {
+      "epoch": 0.7171150543789353,
+      "grad_norm": 2.840635299682617,
+      "learning_rate": 3.92882833999877e-06,
+      "loss": 1.0255,
+      "step": 3132
+    },
+    {
+      "epoch": 0.717344018317115,
+      "grad_norm": 1.2576954364776611,
+      "learning_rate": 3.9229354460804345e-06,
+      "loss": 1.0252,
+      "step": 3133
+    },
+    {
+      "epoch": 0.7175729822552948,
+      "grad_norm": 1.3615483045578003,
+      "learning_rate": 3.917045896319772e-06,
+      "loss": 1.0324,
+      "step": 3134
+    },
+    {
+      "epoch": 0.7178019461934745,
+      "grad_norm": 1.6669081449508667,
+      "learning_rate": 3.911159693957755e-06,
+      "loss": 1.0628,
+      "step": 3135
+    },
+    {
+      "epoch": 0.7180309101316542,
+      "grad_norm": 1.2302170991897583,
+      "learning_rate": 3.905276842233508e-06,
+      "loss": 1.0182,
+      "step": 3136
+    },
+    {
+      "epoch": 0.718259874069834,
+      "grad_norm": 1.0562607049942017,
+      "learning_rate": 3.899397344384316e-06,
+      "loss": 1.01,
+      "step": 3137
+    },
+    {
+      "epoch": 0.7184888380080138,
+      "grad_norm": 2.51627516746521,
+      "learning_rate": 3.893521203645618e-06,
+      "loss": 1.1183,
+      "step": 3138
+    },
+    {
+      "epoch": 0.7187178019461935,
+      "grad_norm": 1.2331818342208862,
+      "learning_rate": 3.887648423251006e-06,
+      "loss": 1.0323,
+      "step": 3139
+    },
+    {
+      "epoch": 0.7189467658843732,
+      "grad_norm": 1.2647502422332764,
+      "learning_rate": 3.881779006432223e-06,
+      "loss": 1.0546,
+      "step": 3140
+    },
+    {
+      "epoch": 0.719175729822553,
+      "grad_norm": 1.3590573072433472,
+      "learning_rate": 3.875912956419152e-06,
+      "loss": 0.9891,
+      "step": 3141
+    },
+    {
+      "epoch": 0.7194046937607327,
+      "grad_norm": 1.4926403760910034,
+      "learning_rate": 3.870050276439843e-06,
+      "loss": 1.0391,
+      "step": 3142
+    },
+    {
+      "epoch": 0.7196336576989124,
+      "grad_norm": 1.504622459411621,
+      "learning_rate": 3.864190969720469e-06,
+      "loss": 0.9936,
+      "step": 3143
+    },
+    {
+      "epoch": 0.7198626216370921,
+      "grad_norm": 1.0495116710662842,
+      "learning_rate": 3.858335039485362e-06,
+      "loss": 1.0307,
+      "step": 3144
+    },
+    {
+      "epoch": 0.7200915855752719,
+      "grad_norm": 1.2690739631652832,
+      "learning_rate": 3.852482488956992e-06,
+      "loss": 1.0355,
+      "step": 3145
+    },
+    {
+      "epoch": 0.7203205495134516,
+      "grad_norm": 2.033017158508301,
+      "learning_rate": 3.846633321355967e-06,
+      "loss": 1.0948,
+      "step": 3146
+    },
+    {
+      "epoch": 0.7205495134516313,
+      "grad_norm": 1.1862678527832031,
+      "learning_rate": 3.840787539901037e-06,
+      "loss": 1.055,
+      "step": 3147
+    },
+    {
+      "epoch": 0.7207784773898112,
+      "grad_norm": 1.3250430822372437,
+      "learning_rate": 3.834945147809082e-06,
+      "loss": 1.052,
+      "step": 3148
+    },
+    {
+      "epoch": 0.7210074413279909,
+      "grad_norm": 1.126364827156067,
+      "learning_rate": 3.829106148295127e-06,
+      "loss": 1.0445,
+      "step": 3149
+    },
+    {
+      "epoch": 0.7212364052661706,
+      "grad_norm": 1.6518840789794922,
+      "learning_rate": 3.823270544572319e-06,
+      "loss": 1.04,
+      "step": 3150
+    },
+    {
+      "epoch": 0.7214653692043503,
+      "grad_norm": 1.127027988433838,
+      "learning_rate": 3.817438339851947e-06,
+      "loss": 1.0113,
+      "step": 3151
+    },
+    {
+      "epoch": 0.7216943331425301,
+      "grad_norm": 1.3487000465393066,
+      "learning_rate": 3.8116095373434204e-06,
+      "loss": 0.9984,
+      "step": 3152
+    },
+    {
+      "epoch": 0.7219232970807098,
+      "grad_norm": 1.3868497610092163,
+      "learning_rate": 3.805784140254286e-06,
+      "loss": 1.0504,
+      "step": 3153
+    },
+    {
+      "epoch": 0.7221522610188895,
+      "grad_norm": 1.4739601612091064,
+      "learning_rate": 3.799962151790203e-06,
+      "loss": 1.0158,
+      "step": 3154
+    },
+    {
+      "epoch": 0.7223812249570692,
+      "grad_norm": 1.0389552116394043,
+      "learning_rate": 3.794143575154964e-06,
+      "loss": 1.0075,
+      "step": 3155
+    },
+    {
+      "epoch": 0.722610188895249,
+      "grad_norm": 1.0425493717193604,
+      "learning_rate": 3.788328413550493e-06,
+      "loss": 0.9877,
+      "step": 3156
+    },
+    {
+      "epoch": 0.7228391528334287,
+      "grad_norm": 1.1646589040756226,
+      "learning_rate": 3.7825166701768125e-06,
+      "loss": 0.959,
+      "step": 3157
+    },
+    {
+      "epoch": 0.7230681167716084,
+      "grad_norm": 1.1060878038406372,
+      "learning_rate": 3.776708348232081e-06,
+      "loss": 1.0858,
+      "step": 3158
+    },
+    {
+      "epoch": 0.7232970807097882,
+      "grad_norm": 1.2601068019866943,
+      "learning_rate": 3.7709034509125706e-06,
+      "loss": 1.0423,
+      "step": 3159
+    },
+    {
+      "epoch": 0.723526044647968,
+      "grad_norm": 2.284247875213623,
+      "learning_rate": 3.7651019814126656e-06,
+      "loss": 1.0213,
+      "step": 3160
+    },
+    {
+      "epoch": 0.7237550085861477,
+      "grad_norm": 1.5774025917053223,
+      "learning_rate": 3.7593039429248667e-06,
+      "loss": 1.0469,
+      "step": 3161
+    },
+    {
+      "epoch": 0.7239839725243274,
+      "grad_norm": 1.9497007131576538,
+      "learning_rate": 3.7535093386397868e-06,
+      "loss": 1.0541,
+      "step": 3162
+    },
+    {
+      "epoch": 0.7242129364625072,
+      "grad_norm": 1.517136812210083,
+      "learning_rate": 3.7477181717461463e-06,
+      "loss": 0.9846,
+      "step": 3163
+    },
+    {
+      "epoch": 0.7244419004006869,
+      "grad_norm": 1.3717387914657593,
+      "learning_rate": 3.74193044543078e-06,
+      "loss": 1.0027,
+      "step": 3164
+    },
+    {
+      "epoch": 0.7246708643388666,
+      "grad_norm": 1.2264940738677979,
+      "learning_rate": 3.7361461628786167e-06,
+      "loss": 1.1155,
+      "step": 3165
+    },
+    {
+      "epoch": 0.7248998282770464,
+      "grad_norm": 1.5513440370559692,
+      "learning_rate": 3.7303653272727057e-06,
+      "loss": 1.0442,
+      "step": 3166
+    },
+    {
+      "epoch": 0.7251287922152261,
+      "grad_norm": 1.0587714910507202,
+      "learning_rate": 3.7245879417941943e-06,
+      "loss": 1.1098,
+      "step": 3167
+    },
+    {
+      "epoch": 0.7253577561534058,
+      "grad_norm": 1.0646835565567017,
+      "learning_rate": 3.7188140096223225e-06,
+      "loss": 0.9877,
+      "step": 3168
+    },
+    {
+      "epoch": 0.7255867200915855,
+      "grad_norm": 1.0646835565567017,
+      "learning_rate": 3.7188140096223225e-06,
+      "loss": 1.0259,
+      "step": 3169
+    },
+    {
+      "epoch": 0.7258156840297653,
+      "grad_norm": 1.221613883972168,
+      "learning_rate": 3.713043533934435e-06,
+      "loss": 0.96,
+      "step": 3170
+    },
+    {
+      "epoch": 0.7260446479679451,
+      "grad_norm": 1.373716950416565,
+      "learning_rate": 3.707276517905989e-06,
+      "loss": 1.0587,
+      "step": 3171
+    },
+    {
+      "epoch": 0.7262736119061248,
+      "grad_norm": 1.5889724493026733,
+      "learning_rate": 3.701512964710513e-06,
+      "loss": 1.0572,
+      "step": 3172
+    },
+    {
+      "epoch": 0.7265025758443046,
+      "grad_norm": 1.3366655111312866,
+      "learning_rate": 3.695752877519646e-06,
+      "loss": 1.0241,
+      "step": 3173
+    },
+    {
+      "epoch": 0.7267315397824843,
+      "grad_norm": 1.3003686666488647,
+      "learning_rate": 3.689996259503116e-06,
+      "loss": 1.0352,
+      "step": 3174
+    },
+    {
+      "epoch": 0.726960503720664,
+      "grad_norm": 1.2446143627166748,
+      "learning_rate": 3.6842431138287415e-06,
+      "loss": 1.0475,
+      "step": 3175
+    },
+    {
+      "epoch": 0.7271894676588437,
+      "grad_norm": 1.0979483127593994,
+      "learning_rate": 3.67849344366243e-06,
+      "loss": 1.0735,
+      "step": 3176
+    },
+    {
+      "epoch": 0.7274184315970235,
+      "grad_norm": 1.3841794729232788,
+      "learning_rate": 3.672747252168176e-06,
+      "loss": 1.0478,
+      "step": 3177
+    },
+    {
+      "epoch": 0.7276473955352032,
+      "grad_norm": 1.3376761674880981,
+      "learning_rate": 3.6670045425080626e-06,
+      "loss": 1.0732,
+      "step": 3178
+    },
+    {
+      "epoch": 0.7278763594733829,
+      "grad_norm": 2.0364394187927246,
+      "learning_rate": 3.6612653178422564e-06,
+      "loss": 1.05,
+      "step": 3179
+    },
+    {
+      "epoch": 0.7281053234115626,
+      "grad_norm": 1.3383004665374756,
+      "learning_rate": 3.655529581328995e-06,
+      "loss": 1.0408,
+      "step": 3180
+    },
+    {
+      "epoch": 0.7283342873497424,
+      "grad_norm": 1.368977427482605,
+      "learning_rate": 3.6497973361246153e-06,
+      "loss": 1.0406,
+      "step": 3181
+    },
+    {
+      "epoch": 0.7285632512879221,
+      "grad_norm": 1.3485641479492188,
+      "learning_rate": 3.6440685853835266e-06,
+      "loss": 1.0455,
+      "step": 3182
+    },
+    {
+      "epoch": 0.7287922152261019,
+      "grad_norm": 1.3723440170288086,
+      "learning_rate": 3.6383433322582028e-06,
+      "loss": 1.0398,
+      "step": 3183
+    },
+    {
+      "epoch": 0.7290211791642817,
+      "grad_norm": 1.250652551651001,
+      "learning_rate": 3.632621579899207e-06,
+      "loss": 1.0,
+      "step": 3184
+    },
+    {
+      "epoch": 0.7292501431024614,
+      "grad_norm": 1.3991602659225464,
+      "learning_rate": 3.6269033314551725e-06,
+      "loss": 1.0197,
+      "step": 3185
+    },
+    {
+      "epoch": 0.7294791070406411,
+      "grad_norm": 1.7109068632125854,
+      "learning_rate": 3.6211885900728017e-06,
+      "loss": 1.0875,
+      "step": 3186
+    },
+    {
+      "epoch": 0.7297080709788208,
+      "grad_norm": 1.0364817380905151,
+      "learning_rate": 3.6154773588968704e-06,
+      "loss": 1.1141,
+      "step": 3187
+    },
+    {
+      "epoch": 0.7299370349170006,
+      "grad_norm": 1.2963764667510986,
+      "learning_rate": 3.609769641070221e-06,
+      "loss": 1.0609,
+      "step": 3188
+    },
+    {
+      "epoch": 0.7301659988551803,
+      "grad_norm": 1.1506567001342773,
+      "learning_rate": 3.6040654397337614e-06,
+      "loss": 1.0173,
+      "step": 3189
+    },
+    {
+      "epoch": 0.73039496279336,
+      "grad_norm": 1.3218399286270142,
+      "learning_rate": 3.598364758026467e-06,
+      "loss": 1.093,
+      "step": 3190
+    },
+    {
+      "epoch": 0.7306239267315398,
+      "grad_norm": 1.5998913049697876,
+      "learning_rate": 3.5926675990853752e-06,
+      "loss": 1.0072,
+      "step": 3191
+    },
+    {
+      "epoch": 0.7308528906697195,
+      "grad_norm": 1.9323956966400146,
+      "learning_rate": 3.5869739660455847e-06,
+      "loss": 0.9571,
+      "step": 3192
+    },
+    {
+      "epoch": 0.7310818546078992,
+      "grad_norm": 1.1179009675979614,
+      "learning_rate": 3.581283862040257e-06,
+      "loss": 1.0337,
+      "step": 3193
+    },
+    {
+      "epoch": 0.731310818546079,
+      "grad_norm": 1.3210819959640503,
+      "learning_rate": 3.575597290200599e-06,
+      "loss": 1.0332,
+      "step": 3194
+    },
+    {
+      "epoch": 0.7315397824842588,
+      "grad_norm": 1.1699495315551758,
+      "learning_rate": 3.569914253655896e-06,
+      "loss": 1.0276,
+      "step": 3195
+    },
+    {
+      "epoch": 0.7317687464224385,
+      "grad_norm": 1.2185953855514526,
+      "learning_rate": 3.5642347555334665e-06,
+      "loss": 1.0249,
+      "step": 3196
+    },
+    {
+      "epoch": 0.7319977103606182,
+      "grad_norm": 1.1740138530731201,
+      "learning_rate": 3.55855879895869e-06,
+      "loss": 0.9732,
+      "step": 3197
+    },
+    {
+      "epoch": 0.732226674298798,
+      "grad_norm": 1.19003427028656,
+      "learning_rate": 3.552886387055009e-06,
+      "loss": 0.993,
+      "step": 3198
+    },
+    {
+      "epoch": 0.7324556382369777,
+      "grad_norm": 1.3954188823699951,
+      "learning_rate": 3.547217522943892e-06,
+      "loss": 1.0497,
+      "step": 3199
+    },
+    {
+      "epoch": 0.7326846021751574,
+      "grad_norm": 1.2999389171600342,
+      "learning_rate": 3.5415522097448717e-06,
+      "loss": 1.0348,
+      "step": 3200
+    },
+    {
+      "epoch": 0.7329135661133371,
+      "grad_norm": 1.2359155416488647,
+      "learning_rate": 3.5358904505755243e-06,
+      "loss": 1.0308,
+      "step": 3201
+    },
+    {
+      "epoch": 0.7331425300515169,
+      "grad_norm": 1.3111592531204224,
+      "learning_rate": 3.530232248551466e-06,
+      "loss": 1.0899,
+      "step": 3202
+    },
+    {
+      "epoch": 0.7333714939896966,
+      "grad_norm": 1.2703155279159546,
+      "learning_rate": 3.524577606786358e-06,
+      "loss": 1.0127,
+      "step": 3203
+    },
+    {
+      "epoch": 0.7336004579278763,
+      "grad_norm": 1.2910027503967285,
+      "learning_rate": 3.518926528391904e-06,
+      "loss": 1.0473,
+      "step": 3204
+    },
+    {
+      "epoch": 0.733829421866056,
+      "grad_norm": 1.4213367700576782,
+      "learning_rate": 3.513279016477844e-06,
+      "loss": 0.9976,
+      "step": 3205
+    },
+    {
+      "epoch": 0.7340583858042359,
+      "grad_norm": 1.4030628204345703,
+      "learning_rate": 3.5076350741519595e-06,
+      "loss": 1.051,
+      "step": 3206
+    },
+    {
+      "epoch": 0.7342873497424156,
+      "grad_norm": 1.3458659648895264,
+      "learning_rate": 3.5019947045200553e-06,
+      "loss": 1.0091,
+      "step": 3207
+    },
+    {
+      "epoch": 0.7345163136805953,
+      "grad_norm": 1.5768972635269165,
+      "learning_rate": 3.496357910685989e-06,
+      "loss": 1.0405,
+      "step": 3208
+    },
+    {
+      "epoch": 0.7347452776187751,
+      "grad_norm": 1.4291819334030151,
+      "learning_rate": 3.4907246957516416e-06,
+      "loss": 1.0732,
+      "step": 3209
+    },
+    {
+      "epoch": 0.7349742415569548,
+      "grad_norm": 1.2574915885925293,
+      "learning_rate": 3.4850950628169167e-06,
+      "loss": 1.0461,
+      "step": 3210
+    },
+    {
+      "epoch": 0.7352032054951345,
+      "grad_norm": 1.2337363958358765,
+      "learning_rate": 3.479469014979754e-06,
+      "loss": 1.1096,
+      "step": 3211
+    },
+    {
+      "epoch": 0.7354321694333142,
+      "grad_norm": 1.4060138463974,
+      "learning_rate": 3.473846555336131e-06,
+      "loss": 1.0116,
+      "step": 3212
+    },
+    {
+      "epoch": 0.735661133371494,
+      "grad_norm": 1.3836556673049927,
+      "learning_rate": 3.46822768698003e-06,
+      "loss": 1.0388,
+      "step": 3213
+    },
+    {
+      "epoch": 0.7358900973096737,
+      "grad_norm": 1.2583669424057007,
+      "learning_rate": 3.4626124130034713e-06,
+      "loss": 1.0596,
+      "step": 3214
+    },
+    {
+      "epoch": 0.7361190612478534,
+      "grad_norm": 1.593520164489746,
+      "learning_rate": 3.457000736496492e-06,
+      "loss": 1.062,
+      "step": 3215
+    },
+    {
+      "epoch": 0.7363480251860332,
+      "grad_norm": 1.5272831916809082,
+      "learning_rate": 3.4513926605471504e-06,
+      "loss": 1.0255,
+      "step": 3216
+    },
+    {
+      "epoch": 0.736576989124213,
+      "grad_norm": 2.2148725986480713,
+      "learning_rate": 3.445788188241527e-06,
+      "loss": 1.0377,
+      "step": 3217
+    },
+    {
+      "epoch": 0.7368059530623927,
+      "grad_norm": 1.665127158164978,
+      "learning_rate": 3.440187322663707e-06,
+      "loss": 1.0202,
+      "step": 3218
+    },
+    {
+      "epoch": 0.7370349170005724,
+      "grad_norm": 1.2639302015304565,
+      "learning_rate": 3.4345900668958088e-06,
+      "loss": 1.0474,
+      "step": 3219
+    },
+    {
+      "epoch": 0.7372638809387522,
+      "grad_norm": 1.2613798379898071,
+      "learning_rate": 3.428996424017956e-06,
+      "loss": 1.0672,
+      "step": 3220
+    },
+    {
+      "epoch": 0.7374928448769319,
+      "grad_norm": 10.604727745056152,
+      "learning_rate": 3.423406397108273e-06,
+      "loss": 0.9956,
+      "step": 3221
+    },
+    {
+      "epoch": 0.7377218088151116,
+      "grad_norm": 1.3740556240081787,
+      "learning_rate": 3.417819989242915e-06,
+      "loss": 1.0367,
+      "step": 3222
+    },
+    {
+      "epoch": 0.7379507727532914,
+      "grad_norm": 1.2875322103500366,
+      "learning_rate": 3.412237203496036e-06,
+      "loss": 1.0304,
+      "step": 3223
+    },
+    {
+      "epoch": 0.7381797366914711,
+      "grad_norm": 1.4357035160064697,
+      "learning_rate": 3.4066580429397877e-06,
+      "loss": 0.9638,
+      "step": 3224
+    },
+    {
+      "epoch": 0.7384087006296508,
+      "grad_norm": 1.247023344039917,
+      "learning_rate": 3.401082510644337e-06,
+      "loss": 0.9908,
+      "step": 3225
+    },
+    {
+      "epoch": 0.7386376645678305,
+      "grad_norm": 1.2303351163864136,
+      "learning_rate": 3.395510609677861e-06,
+      "loss": 1.0369,
+      "step": 3226
+    },
+    {
+      "epoch": 0.7388666285060103,
+      "grad_norm": 1.5605922937393188,
+      "learning_rate": 3.3899423431065215e-06,
+      "loss": 1.0669,
+      "step": 3227
+    },
+    {
+      "epoch": 0.73909559244419,
+      "grad_norm": 1.551808476448059,
+      "learning_rate": 3.384377713994492e-06,
+      "loss": 0.9855,
+      "step": 3228
+    },
+    {
+      "epoch": 0.7393245563823698,
+      "grad_norm": 1.3470834493637085,
+      "learning_rate": 3.3788167254039417e-06,
+      "loss": 1.0534,
+      "step": 3229
+    },
+    {
+      "epoch": 0.7395535203205496,
+      "grad_norm": 2.2554211616516113,
+      "learning_rate": 3.3732593803950354e-06,
+      "loss": 1.0309,
+      "step": 3230
+    },
+    {
+      "epoch": 0.7397824842587293,
+      "grad_norm": 1.2798614501953125,
+      "learning_rate": 3.3677056820259324e-06,
+      "loss": 0.9978,
+      "step": 3231
+    },
+    {
+      "epoch": 0.740011448196909,
+      "grad_norm": 1.184571385383606,
+      "learning_rate": 3.3621556333527884e-06,
+      "loss": 1.053,
+      "step": 3232
+    },
+    {
+      "epoch": 0.7402404121350887,
+      "grad_norm": 1.5115231275558472,
+      "learning_rate": 3.3566092374297465e-06,
+      "loss": 0.9638,
+      "step": 3233
+    },
+    {
+      "epoch": 0.7404693760732685,
+      "grad_norm": 1.1283025741577148,
+      "learning_rate": 3.3510664973089467e-06,
+      "loss": 1.022,
+      "step": 3234
+    },
+    {
+      "epoch": 0.7406983400114482,
+      "grad_norm": 1.4040166139602661,
+      "learning_rate": 3.3455274160405025e-06,
+      "loss": 1.037,
+      "step": 3235
+    },
+    {
+      "epoch": 0.7409273039496279,
+      "grad_norm": 1.2837587594985962,
+      "learning_rate": 3.3399919966725335e-06,
+      "loss": 1.0439,
+      "step": 3236
+    },
+    {
+      "epoch": 0.7411562678878076,
+      "grad_norm": 1.1922607421875,
+      "learning_rate": 3.3344602422511343e-06,
+      "loss": 0.9965,
+      "step": 3237
+    },
+    {
+      "epoch": 0.7413852318259874,
+      "grad_norm": 1.1945897340774536,
+      "learning_rate": 3.3289321558203767e-06,
+      "loss": 1.0592,
+      "step": 3238
+    },
+    {
+      "epoch": 0.7416141957641671,
+      "grad_norm": 1.1530017852783203,
+      "learning_rate": 3.323407740422323e-06,
+      "loss": 1.0469,
+      "step": 3239
+    },
+    {
+      "epoch": 0.7418431597023469,
+      "grad_norm": 1.1255567073822021,
+      "learning_rate": 3.317886999097014e-06,
+      "loss": 1.0633,
+      "step": 3240
+    },
+    {
+      "epoch": 0.7420721236405267,
+      "grad_norm": 1.1872222423553467,
+      "learning_rate": 3.3123699348824654e-06,
+      "loss": 0.9804,
+      "step": 3241
+    },
+    {
+      "epoch": 0.7423010875787064,
+      "grad_norm": 1.289146900177002,
+      "learning_rate": 3.306856550814673e-06,
+      "loss": 1.0055,
+      "step": 3242
+    },
+    {
+      "epoch": 0.7425300515168861,
+      "grad_norm": 1.5255881547927856,
+      "learning_rate": 3.3013468499276057e-06,
+      "loss": 1.0347,
+      "step": 3243
+    },
+    {
+      "epoch": 0.7427590154550658,
+      "grad_norm": 1.0099772214889526,
+      "learning_rate": 3.2958408352532055e-06,
+      "loss": 1.0634,
+      "step": 3244
+    },
+    {
+      "epoch": 0.7429879793932456,
+      "grad_norm": 1.8835973739624023,
+      "learning_rate": 3.290338509821386e-06,
+      "loss": 1.024,
+      "step": 3245
+    },
+    {
+      "epoch": 0.7432169433314253,
+      "grad_norm": 1.2783852815628052,
+      "learning_rate": 3.2848398766600298e-06,
+      "loss": 1.0316,
+      "step": 3246
+    },
+    {
+      "epoch": 0.743445907269605,
+      "grad_norm": 1.3650398254394531,
+      "learning_rate": 3.2793449387949907e-06,
+      "loss": 1.0741,
+      "step": 3247
+    },
+    {
+      "epoch": 0.7436748712077847,
+      "grad_norm": 1.6553008556365967,
+      "learning_rate": 3.273853699250088e-06,
+      "loss": 0.9974,
+      "step": 3248
+    },
+    {
+      "epoch": 0.7439038351459645,
+      "grad_norm": 1.0704493522644043,
+      "learning_rate": 3.268366161047096e-06,
+      "loss": 1.08,
+      "step": 3249
+    },
+    {
+      "epoch": 0.7441327990841442,
+      "grad_norm": 1.375232219696045,
+      "learning_rate": 3.2628823272057753e-06,
+      "loss": 1.0434,
+      "step": 3250
+    },
+    {
+      "epoch": 0.7443617630223239,
+      "grad_norm": 1.122849464416504,
+      "learning_rate": 3.257402200743821e-06,
+      "loss": 1.0497,
+      "step": 3251
+    },
+    {
+      "epoch": 0.7445907269605038,
+      "grad_norm": 1.2734968662261963,
+      "learning_rate": 3.251925784676907e-06,
+      "loss": 1.0542,
+      "step": 3252
+    },
+    {
+      "epoch": 0.7448196908986835,
+      "grad_norm": 1.3681397438049316,
+      "learning_rate": 3.246453082018658e-06,
+      "loss": 0.9997,
+      "step": 3253
+    },
+    {
+      "epoch": 0.7450486548368632,
+      "grad_norm": 1.1173583269119263,
+      "learning_rate": 3.2409840957806548e-06,
+      "loss": 1.0651,
+      "step": 3254
+    },
+    {
+      "epoch": 0.745277618775043,
+      "grad_norm": 1.235009789466858,
+      "learning_rate": 3.235518828972437e-06,
+      "loss": 0.9709,
+      "step": 3255
+    },
+    {
+      "epoch": 0.7455065827132227,
+      "grad_norm": 1.1784312725067139,
+      "learning_rate": 3.2300572846014945e-06,
+      "loss": 1.0648,
+      "step": 3256
+    },
+    {
+      "epoch": 0.7457355466514024,
+      "grad_norm": 1.190245270729065,
+      "learning_rate": 3.224599465673268e-06,
+      "loss": 1.0308,
+      "step": 3257
+    },
+    {
+      "epoch": 0.7459645105895821,
+      "grad_norm": 1.3046700954437256,
+      "learning_rate": 3.2191453751911505e-06,
+      "loss": 0.9958,
+      "step": 3258
+    },
+    {
+      "epoch": 0.7461934745277619,
+      "grad_norm": 1.363734483718872,
+      "learning_rate": 3.213695016156484e-06,
+      "loss": 1.0139,
+      "step": 3259
+    },
+    {
+      "epoch": 0.7464224384659416,
+      "grad_norm": 1.3983644247055054,
+      "learning_rate": 3.2082483915685526e-06,
+      "loss": 1.0878,
+      "step": 3260
+    },
+    {
+      "epoch": 0.7466514024041213,
+      "grad_norm": 1.4021762609481812,
+      "learning_rate": 3.202805504424592e-06,
+      "loss": 1.0088,
+      "step": 3261
+    },
+    {
+      "epoch": 0.746880366342301,
+      "grad_norm": 1.5322551727294922,
+      "learning_rate": 3.197366357719772e-06,
+      "loss": 1.0558,
+      "step": 3262
+    },
+    {
+      "epoch": 0.7471093302804809,
+      "grad_norm": 1.4262042045593262,
+      "learning_rate": 3.191930954447209e-06,
+      "loss": 1.0341,
+      "step": 3263
+    },
+    {
+      "epoch": 0.7473382942186606,
+      "grad_norm": 1.3308278322219849,
+      "learning_rate": 3.186499297597968e-06,
+      "loss": 1.0309,
+      "step": 3264
+    },
+    {
+      "epoch": 0.7475672581568403,
+      "grad_norm": 1.1862602233886719,
+      "learning_rate": 3.1810713901610367e-06,
+      "loss": 1.0448,
+      "step": 3265
+    },
+    {
+      "epoch": 0.7477962220950201,
+      "grad_norm": 1.3563153743743896,
+      "learning_rate": 3.175647235123347e-06,
+      "loss": 1.0332,
+      "step": 3266
+    },
+    {
+      "epoch": 0.7480251860331998,
+      "grad_norm": 1.3230400085449219,
+      "learning_rate": 3.170226835469774e-06,
+      "loss": 1.0438,
+      "step": 3267
+    },
+    {
+      "epoch": 0.7482541499713795,
+      "grad_norm": 1.3605960607528687,
+      "learning_rate": 3.16481019418311e-06,
+      "loss": 1.001,
+      "step": 3268
+    },
+    {
+      "epoch": 0.7484831139095592,
+      "grad_norm": 1.1855884790420532,
+      "learning_rate": 3.159397314244089e-06,
+      "loss": 0.9966,
+      "step": 3269
+    },
+    {
+      "epoch": 0.748712077847739,
+      "grad_norm": 1.5543913841247559,
+      "learning_rate": 3.1539881986313758e-06,
+      "loss": 0.9855,
+      "step": 3270
+    },
+    {
+      "epoch": 0.7489410417859187,
+      "grad_norm": 1.125836968421936,
+      "learning_rate": 3.1485828503215588e-06,
+      "loss": 1.1038,
+      "step": 3271
+    },
+    {
+      "epoch": 0.7491700057240984,
+      "grad_norm": 2.7458279132843018,
+      "learning_rate": 3.1431812722891598e-06,
+      "loss": 0.9994,
+      "step": 3272
+    },
+    {
+      "epoch": 0.7493989696622781,
+      "grad_norm": 1.4484138488769531,
+      "learning_rate": 3.137783467506613e-06,
+      "loss": 0.9574,
+      "step": 3273
+    },
+    {
+      "epoch": 0.7496279336004579,
+      "grad_norm": 1.4349960088729858,
+      "learning_rate": 3.1323894389442942e-06,
+      "loss": 1.0307,
+      "step": 3274
+    },
+    {
+      "epoch": 0.7498568975386377,
+      "grad_norm": 1.287442922592163,
+      "learning_rate": 3.126999189570493e-06,
+      "loss": 1.0006,
+      "step": 3275
+    },
+    {
+      "epoch": 0.7500858614768174,
+      "grad_norm": 1.1122688055038452,
+      "learning_rate": 3.1216127223514116e-06,
+      "loss": 1.046,
+      "step": 3276
+    },
+    {
+      "epoch": 0.7503148254149972,
+      "grad_norm": 1.196319580078125,
+      "learning_rate": 3.116230040251177e-06,
+      "loss": 1.0686,
+      "step": 3277
+    },
+    {
+      "epoch": 0.7505437893531769,
+      "grad_norm": 1.5629481077194214,
+      "learning_rate": 3.1108511462318437e-06,
+      "loss": 1.0516,
+      "step": 3278
+    },
+    {
+      "epoch": 0.7507727532913566,
+      "grad_norm": 1.118003487586975,
+      "learning_rate": 3.1054760432533626e-06,
+      "loss": 1.0298,
+      "step": 3279
+    },
+    {
+      "epoch": 0.7510017172295363,
+      "grad_norm": 1.118003487586975,
+      "learning_rate": 3.1054760432533626e-06,
+      "loss": 1.0198,
+      "step": 3280
+    },
+    {
+      "epoch": 0.7512306811677161,
+      "grad_norm": 1.319196343421936,
+      "learning_rate": 3.100104734273608e-06,
+      "loss": 1.0704,
+      "step": 3281
+    },
+    {
+      "epoch": 0.7514596451058958,
+      "grad_norm": 1.3036201000213623,
+      "learning_rate": 3.0947372222483762e-06,
+      "loss": 1.0307,
+      "step": 3282
+    },
+    {
+      "epoch": 0.7516886090440755,
+      "grad_norm": 1.2368963956832886,
+      "learning_rate": 3.089373510131354e-06,
+      "loss": 1.0142,
+      "step": 3283
+    },
+    {
+      "epoch": 0.7519175729822553,
+      "grad_norm": 1.583749771118164,
+      "learning_rate": 3.0840136008741505e-06,
+      "loss": 1.0126,
+      "step": 3284
+    },
+    {
+      "epoch": 0.752146536920435,
+      "grad_norm": 1.174692988395691,
+      "learning_rate": 3.0786574974262784e-06,
+      "loss": 1.0675,
+      "step": 3285
+    },
+    {
+      "epoch": 0.7523755008586148,
+      "grad_norm": 1.5759061574935913,
+      "learning_rate": 3.073305202735157e-06,
+      "loss": 1.0071,
+      "step": 3286
+    },
+    {
+      "epoch": 0.7526044647967945,
+      "grad_norm": 1.3044160604476929,
+      "learning_rate": 3.0679567197461135e-06,
+      "loss": 0.9839,
+      "step": 3287
+    },
+    {
+      "epoch": 0.7528334287349743,
+      "grad_norm": 1.2453902959823608,
+      "learning_rate": 3.0626120514023605e-06,
+      "loss": 1.0634,
+      "step": 3288
+    },
+    {
+      "epoch": 0.753062392673154,
+      "grad_norm": 1.130800485610962,
+      "learning_rate": 3.057271200645037e-06,
+      "loss": 1.0402,
+      "step": 3289
+    },
+    {
+      "epoch": 0.7532913566113337,
+      "grad_norm": 1.4683923721313477,
+      "learning_rate": 3.0519341704131666e-06,
+      "loss": 1.0274,
+      "step": 3290
+    },
+    {
+      "epoch": 0.7535203205495135,
+      "grad_norm": 1.3023918867111206,
+      "learning_rate": 3.0466009636436633e-06,
+      "loss": 1.0431,
+      "step": 3291
+    },
+    {
+      "epoch": 0.7537492844876932,
+      "grad_norm": 2.7885706424713135,
+      "learning_rate": 3.0412715832713592e-06,
+      "loss": 1.0291,
+      "step": 3292
+    },
+    {
+      "epoch": 0.7539782484258729,
+      "grad_norm": 1.2981799840927124,
+      "learning_rate": 3.035946032228957e-06,
+      "loss": 0.9623,
+      "step": 3293
+    },
+    {
+      "epoch": 0.7542072123640526,
+      "grad_norm": 1.2832386493682861,
+      "learning_rate": 3.0306243134470668e-06,
+      "loss": 1.0626,
+      "step": 3294
+    },
+    {
+      "epoch": 0.7544361763022324,
+      "grad_norm": 1.5354857444763184,
+      "learning_rate": 3.0253064298541857e-06,
+      "loss": 1.0211,
+      "step": 3295
+    },
+    {
+      "epoch": 0.7546651402404121,
+      "grad_norm": 1.4440540075302124,
+      "learning_rate": 3.0199923843767007e-06,
+      "loss": 1.0351,
+      "step": 3296
+    },
+    {
+      "epoch": 0.7548941041785918,
+      "grad_norm": 1.2906075716018677,
+      "learning_rate": 3.014682179938886e-06,
+      "loss": 1.0806,
+      "step": 3297
+    },
+    {
+      "epoch": 0.7551230681167717,
+      "grad_norm": 1.2191280126571655,
+      "learning_rate": 3.009375819462902e-06,
+      "loss": 1.0078,
+      "step": 3298
+    },
+    {
+      "epoch": 0.7553520320549514,
+      "grad_norm": 1.3844983577728271,
+      "learning_rate": 3.0040733058687956e-06,
+      "loss": 1.056,
+      "step": 3299
+    },
+    {
+      "epoch": 0.7555809959931311,
+      "grad_norm": 1.1661230325698853,
+      "learning_rate": 2.998774642074496e-06,
+      "loss": 1.061,
+      "step": 3300
+    },
+    {
+      "epoch": 0.7558099599313108,
+      "grad_norm": 1.2717007398605347,
+      "learning_rate": 2.993479830995815e-06,
+      "loss": 0.9824,
+      "step": 3301
+    },
+    {
+      "epoch": 0.7560389238694906,
+      "grad_norm": 1.4554436206817627,
+      "learning_rate": 2.9881888755464337e-06,
+      "loss": 1.0202,
+      "step": 3302
+    },
+    {
+      "epoch": 0.7562678878076703,
+      "grad_norm": 1.251201868057251,
+      "learning_rate": 2.9829017786379333e-06,
+      "loss": 1.0595,
+      "step": 3303
+    },
+    {
+      "epoch": 0.75649685174585,
+      "grad_norm": 1.0925774574279785,
+      "learning_rate": 2.97761854317975e-06,
+      "loss": 1.0848,
+      "step": 3304
+    },
+    {
+      "epoch": 0.7567258156840297,
+      "grad_norm": 1.3679059743881226,
+      "learning_rate": 2.972339172079204e-06,
+      "loss": 1.0356,
+      "step": 3305
+    },
+    {
+      "epoch": 0.7569547796222095,
+      "grad_norm": 1.140267014503479,
+      "learning_rate": 2.9670636682414966e-06,
+      "loss": 1.04,
+      "step": 3306
+    },
+    {
+      "epoch": 0.7571837435603892,
+      "grad_norm": 1.2260184288024902,
+      "learning_rate": 2.961792034569686e-06,
+      "loss": 1.0593,
+      "step": 3307
+    },
+    {
+      "epoch": 0.7574127074985689,
+      "grad_norm": 1.3413991928100586,
+      "learning_rate": 2.9565242739647115e-06,
+      "loss": 1.0271,
+      "step": 3308
+    },
+    {
+      "epoch": 0.7576416714367488,
+      "grad_norm": 1.2301392555236816,
+      "learning_rate": 2.9512603893253756e-06,
+      "loss": 1.0802,
+      "step": 3309
+    },
+    {
+      "epoch": 0.7578706353749285,
+      "grad_norm": 1.1220687627792358,
+      "learning_rate": 2.9460003835483497e-06,
+      "loss": 1.0152,
+      "step": 3310
+    },
+    {
+      "epoch": 0.7580995993131082,
+      "grad_norm": 2.1697607040405273,
+      "learning_rate": 2.940744259528173e-06,
+      "loss": 1.0526,
+      "step": 3311
+    },
+    {
+      "epoch": 0.758328563251288,
+      "grad_norm": 1.1174249649047852,
+      "learning_rate": 2.9354920201572457e-06,
+      "loss": 0.9836,
+      "step": 3312
+    },
+    {
+      "epoch": 0.7585575271894677,
+      "grad_norm": 1.1898833513259888,
+      "learning_rate": 2.9302436683258306e-06,
+      "loss": 1.0391,
+      "step": 3313
+    },
+    {
+      "epoch": 0.7587864911276474,
+      "grad_norm": 1.1694586277008057,
+      "learning_rate": 2.9249992069220557e-06,
+      "loss": 1.0203,
+      "step": 3314
+    },
+    {
+      "epoch": 0.7590154550658271,
+      "grad_norm": 1.1864376068115234,
+      "learning_rate": 2.919758638831893e-06,
+      "loss": 1.046,
+      "step": 3315
+    },
+    {
+      "epoch": 0.7592444190040069,
+      "grad_norm": 1.4337023496627808,
+      "learning_rate": 2.9145219669391944e-06,
+      "loss": 1.1258,
+      "step": 3316
+    },
+    {
+      "epoch": 0.7594733829421866,
+      "grad_norm": 1.0952881574630737,
+      "learning_rate": 2.909289194125655e-06,
+      "loss": 1.0528,
+      "step": 3317
+    },
+    {
+      "epoch": 0.7597023468803663,
+      "grad_norm": 1.3980028629302979,
+      "learning_rate": 2.904060323270822e-06,
+      "loss": 1.0476,
+      "step": 3318
+    },
+    {
+      "epoch": 0.759931310818546,
+      "grad_norm": 1.3246995210647583,
+      "learning_rate": 2.898835357252097e-06,
+      "loss": 1.0199,
+      "step": 3319
+    },
+    {
+      "epoch": 0.7601602747567258,
+      "grad_norm": 1.449928879737854,
+      "learning_rate": 2.8936142989447434e-06,
+      "loss": 1.0102,
+      "step": 3320
+    },
+    {
+      "epoch": 0.7603892386949056,
+      "grad_norm": 1.211899757385254,
+      "learning_rate": 2.8883971512218588e-06,
+      "loss": 1.0177,
+      "step": 3321
+    },
+    {
+      "epoch": 0.7606182026330853,
+      "grad_norm": 1.1149224042892456,
+      "learning_rate": 2.8831839169543998e-06,
+      "loss": 1.0456,
+      "step": 3322
+    },
+    {
+      "epoch": 0.760847166571265,
+      "grad_norm": 1.3038837909698486,
+      "learning_rate": 2.877974599011162e-06,
+      "loss": 1.0935,
+      "step": 3323
+    },
+    {
+      "epoch": 0.7610761305094448,
+      "grad_norm": 1.0868488550186157,
+      "learning_rate": 2.8727692002587914e-06,
+      "loss": 1.0169,
+      "step": 3324
+    },
+    {
+      "epoch": 0.7613050944476245,
+      "grad_norm": 1.4088412523269653,
+      "learning_rate": 2.867567723561776e-06,
+      "loss": 1.0274,
+      "step": 3325
+    },
+    {
+      "epoch": 0.7615340583858042,
+      "grad_norm": 1.7433838844299316,
+      "learning_rate": 2.8623701717824435e-06,
+      "loss": 0.9796,
+      "step": 3326
+    },
+    {
+      "epoch": 0.761763022323984,
+      "grad_norm": 1.1715034246444702,
+      "learning_rate": 2.8571765477809645e-06,
+      "loss": 1.0241,
+      "step": 3327
+    },
+    {
+      "epoch": 0.7619919862621637,
+      "grad_norm": 1.7508958578109741,
+      "learning_rate": 2.851986854415347e-06,
+      "loss": 0.9302,
+      "step": 3328
+    },
+    {
+      "epoch": 0.7622209502003434,
+      "grad_norm": 1.3008875846862793,
+      "learning_rate": 2.84680109454143e-06,
+      "loss": 1.0152,
+      "step": 3329
+    },
+    {
+      "epoch": 0.7624499141385231,
+      "grad_norm": 1.7418900728225708,
+      "learning_rate": 2.841619271012901e-06,
+      "loss": 0.97,
+      "step": 3330
+    },
+    {
+      "epoch": 0.7626788780767029,
+      "grad_norm": 1.3994221687316895,
+      "learning_rate": 2.8364413866812733e-06,
+      "loss": 1.1088,
+      "step": 3331
+    },
+    {
+      "epoch": 0.7629078420148827,
+      "grad_norm": 1.5729641914367676,
+      "learning_rate": 2.83126744439589e-06,
+      "loss": 1.0114,
+      "step": 3332
+    },
+    {
+      "epoch": 0.7631368059530624,
+      "grad_norm": 1.295013666152954,
+      "learning_rate": 2.826097447003925e-06,
+      "loss": 1.0756,
+      "step": 3333
+    },
+    {
+      "epoch": 0.7633657698912422,
+      "grad_norm": 1.3600553274154663,
+      "learning_rate": 2.820931397350395e-06,
+      "loss": 1.0254,
+      "step": 3334
+    },
+    {
+      "epoch": 0.7635947338294219,
+      "grad_norm": 1.1421211957931519,
+      "learning_rate": 2.815769298278125e-06,
+      "loss": 1.0173,
+      "step": 3335
+    },
+    {
+      "epoch": 0.7638236977676016,
+      "grad_norm": 2.1977944374084473,
+      "learning_rate": 2.810611152627777e-06,
+      "loss": 1.0828,
+      "step": 3336
+    },
+    {
+      "epoch": 0.7640526617057813,
+      "grad_norm": 1.382452368736267,
+      "learning_rate": 2.8054569632378358e-06,
+      "loss": 1.0648,
+      "step": 3337
+    },
+    {
+      "epoch": 0.7642816256439611,
+      "grad_norm": 1.372998833656311,
+      "learning_rate": 2.80030673294461e-06,
+      "loss": 1.0375,
+      "step": 3338
+    },
+    {
+      "epoch": 0.7645105895821408,
+      "grad_norm": 1.1051013469696045,
+      "learning_rate": 2.795160464582225e-06,
+      "loss": 1.0433,
+      "step": 3339
+    },
+    {
+      "epoch": 0.7647395535203205,
+      "grad_norm": 1.265097737312317,
+      "learning_rate": 2.7900181609826325e-06,
+      "loss": 0.9865,
+      "step": 3340
+    },
+    {
+      "epoch": 0.7649685174585003,
+      "grad_norm": 1.2459075450897217,
+      "learning_rate": 2.784879824975597e-06,
+      "loss": 1.0751,
+      "step": 3341
+    },
+    {
+      "epoch": 0.76519748139668,
+      "grad_norm": 1.2201124429702759,
+      "learning_rate": 2.779745459388705e-06,
+      "loss": 1.0641,
+      "step": 3342
+    },
+    {
+      "epoch": 0.7654264453348597,
+      "grad_norm": 1.1874167919158936,
+      "learning_rate": 2.774615067047346e-06,
+      "loss": 1.0337,
+      "step": 3343
+    },
+    {
+      "epoch": 0.7656554092730395,
+      "grad_norm": 1.3858647346496582,
+      "learning_rate": 2.769488650774741e-06,
+      "loss": 1.0191,
+      "step": 3344
+    },
+    {
+      "epoch": 0.7658843732112193,
+      "grad_norm": 1.2821847200393677,
+      "learning_rate": 2.7643662133919136e-06,
+      "loss": 1.0342,
+      "step": 3345
+    },
+    {
+      "epoch": 0.766113337149399,
+      "grad_norm": 1.4728732109069824,
+      "learning_rate": 2.7592477577176924e-06,
+      "loss": 0.9817,
+      "step": 3346
+    },
+    {
+      "epoch": 0.7663423010875787,
+      "grad_norm": 1.484055519104004,
+      "learning_rate": 2.7541332865687245e-06,
+      "loss": 1.1209,
+      "step": 3347
+    },
+    {
+      "epoch": 0.7665712650257585,
+      "grad_norm": 1.1936273574829102,
+      "learning_rate": 2.749022802759459e-06,
+      "loss": 1.0309,
+      "step": 3348
+    },
+    {
+      "epoch": 0.7668002289639382,
+      "grad_norm": 1.9332703351974487,
+      "learning_rate": 2.7439163091021525e-06,
+      "loss": 1.0065,
+      "step": 3349
+    },
+    {
+      "epoch": 0.7670291929021179,
+      "grad_norm": 1.1286215782165527,
+      "learning_rate": 2.738813808406866e-06,
+      "loss": 0.9393,
+      "step": 3350
+    },
+    {
+      "epoch": 0.7672581568402976,
+      "grad_norm": 1.4261751174926758,
+      "learning_rate": 2.7337153034814636e-06,
+      "loss": 1.051,
+      "step": 3351
+    },
+    {
+      "epoch": 0.7674871207784774,
+      "grad_norm": 1.2641804218292236,
+      "learning_rate": 2.7286207971316094e-06,
+      "loss": 1.0007,
+      "step": 3352
+    },
+    {
+      "epoch": 0.7677160847166571,
+      "grad_norm": 1.276326298713684,
+      "learning_rate": 2.7235302921607665e-06,
+      "loss": 1.0195,
+      "step": 3353
+    },
+    {
+      "epoch": 0.7679450486548368,
+      "grad_norm": 1.2994381189346313,
+      "learning_rate": 2.7184437913701977e-06,
+      "loss": 1.0581,
+      "step": 3354
+    },
+    {
+      "epoch": 0.7681740125930167,
+      "grad_norm": 1.2071455717086792,
+      "learning_rate": 2.713361297558963e-06,
+      "loss": 1.0235,
+      "step": 3355
+    },
+    {
+      "epoch": 0.7684029765311964,
+      "grad_norm": 1.246546745300293,
+      "learning_rate": 2.708282813523917e-06,
+      "loss": 1.0011,
+      "step": 3356
+    },
+    {
+      "epoch": 0.7686319404693761,
+      "grad_norm": 1.2827038764953613,
+      "learning_rate": 2.7032083420597e-06,
+      "loss": 1.0674,
+      "step": 3357
+    },
+    {
+      "epoch": 0.7688609044075558,
+      "grad_norm": 1.9928056001663208,
+      "learning_rate": 2.6981378859587614e-06,
+      "loss": 1.0078,
+      "step": 3358
+    },
+    {
+      "epoch": 0.7690898683457356,
+      "grad_norm": 1.672621250152588,
+      "learning_rate": 2.6930714480113217e-06,
+      "loss": 1.0532,
+      "step": 3359
+    },
+    {
+      "epoch": 0.7693188322839153,
+      "grad_norm": 1.4071228504180908,
+      "learning_rate": 2.688009031005403e-06,
+      "loss": 1.1329,
+      "step": 3360
+    },
+    {
+      "epoch": 0.769547796222095,
+      "grad_norm": 2.0468266010284424,
+      "learning_rate": 2.6829506377268122e-06,
+      "loss": 1.063,
+      "step": 3361
+    },
+    {
+      "epoch": 0.7697767601602747,
+      "grad_norm": 1.948828935623169,
+      "learning_rate": 2.6778962709591382e-06,
+      "loss": 1.0897,
+      "step": 3362
+    },
+    {
+      "epoch": 0.7700057240984545,
+      "grad_norm": 1.4286307096481323,
+      "learning_rate": 2.6728459334837576e-06,
+      "loss": 1.0155,
+      "step": 3363
+    },
+    {
+      "epoch": 0.7702346880366342,
+      "grad_norm": 1.1092466115951538,
+      "learning_rate": 2.667799628079829e-06,
+      "loss": 1.0039,
+      "step": 3364
+    },
+    {
+      "epoch": 0.7704636519748139,
+      "grad_norm": 1.3140974044799805,
+      "learning_rate": 2.6627573575242917e-06,
+      "loss": 1.0641,
+      "step": 3365
+    },
+    {
+      "epoch": 0.7706926159129937,
+      "grad_norm": 1.29413640499115,
+      "learning_rate": 2.6577191245918654e-06,
+      "loss": 1.01,
+      "step": 3366
+    },
+    {
+      "epoch": 0.7709215798511735,
+      "grad_norm": 1.2063698768615723,
+      "learning_rate": 2.6526849320550474e-06,
+      "loss": 1.0791,
+      "step": 3367
+    },
+    {
+      "epoch": 0.7711505437893532,
+      "grad_norm": 1.2607606649398804,
+      "learning_rate": 2.6476547826841106e-06,
+      "loss": 1.0255,
+      "step": 3368
+    },
+    {
+      "epoch": 0.7713795077275329,
+      "grad_norm": 1.187766194343567,
+      "learning_rate": 2.642628679247109e-06,
+      "loss": 1.066,
+      "step": 3369
+    },
+    {
+      "epoch": 0.7716084716657127,
+      "grad_norm": 1.3988637924194336,
+      "learning_rate": 2.6376066245098565e-06,
+      "loss": 1.1261,
+      "step": 3370
+    },
+    {
+      "epoch": 0.7718374356038924,
+      "grad_norm": 1.4937584400177002,
+      "learning_rate": 2.6325886212359496e-06,
+      "loss": 0.9689,
+      "step": 3371
+    },
+    {
+      "epoch": 0.7720663995420721,
+      "grad_norm": 1.2917104959487915,
+      "learning_rate": 2.6275746721867624e-06,
+      "loss": 1.0493,
+      "step": 3372
+    },
+    {
+      "epoch": 0.7722953634802519,
+      "grad_norm": 1.5106794834136963,
+      "learning_rate": 2.6225647801214203e-06,
+      "loss": 0.9849,
+      "step": 3373
+    },
+    {
+      "epoch": 0.7725243274184316,
+      "grad_norm": 1.2963145971298218,
+      "learning_rate": 2.6175589477968234e-06,
+      "loss": 1.0687,
+      "step": 3374
+    },
+    {
+      "epoch": 0.7727532913566113,
+      "grad_norm": 1.9042096138000488,
+      "learning_rate": 2.6125571779676493e-06,
+      "loss": 1.0122,
+      "step": 3375
+    },
+    {
+      "epoch": 0.772982255294791,
+      "grad_norm": 14.797686576843262,
+      "learning_rate": 2.607559473386321e-06,
+      "loss": 1.0177,
+      "step": 3376
+    },
+    {
+      "epoch": 0.7732112192329708,
+      "grad_norm": 1.1996756792068481,
+      "learning_rate": 2.602565836803036e-06,
+      "loss": 1.1003,
+      "step": 3377
+    },
+    {
+      "epoch": 0.7734401831711506,
+      "grad_norm": 1.0620721578598022,
+      "learning_rate": 2.5975762709657506e-06,
+      "loss": 1.0226,
+      "step": 3378
+    },
+    {
+      "epoch": 0.7736691471093303,
+      "grad_norm": 1.3966255187988281,
+      "learning_rate": 2.5925907786201808e-06,
+      "loss": 1.0354,
+      "step": 3379
+    },
+    {
+      "epoch": 0.77389811104751,
+      "grad_norm": 1.4668009281158447,
+      "learning_rate": 2.5876093625098066e-06,
+      "loss": 1.0616,
+      "step": 3380
+    },
+    {
+      "epoch": 0.7741270749856898,
+      "grad_norm": 1.1744608879089355,
+      "learning_rate": 2.5826320253758477e-06,
+      "loss": 1.0438,
+      "step": 3381
+    },
+    {
+      "epoch": 0.7743560389238695,
+      "grad_norm": 1.5106372833251953,
+      "learning_rate": 2.5776587699573007e-06,
+      "loss": 1.066,
+      "step": 3382
+    },
+    {
+      "epoch": 0.7745850028620492,
+      "grad_norm": 1.2056074142456055,
+      "learning_rate": 2.5726895989909063e-06,
+      "loss": 1.0526,
+      "step": 3383
+    },
+    {
+      "epoch": 0.774813966800229,
+      "grad_norm": 1.2208433151245117,
+      "learning_rate": 2.5677245152111497e-06,
+      "loss": 1.0644,
+      "step": 3384
+    },
+    {
+      "epoch": 0.7750429307384087,
+      "grad_norm": 1.3252755403518677,
+      "learning_rate": 2.5627635213502832e-06,
+      "loss": 1.0094,
+      "step": 3385
+    },
+    {
+      "epoch": 0.7752718946765884,
+      "grad_norm": 1.3273080587387085,
+      "learning_rate": 2.5578066201383e-06,
+      "loss": 1.1036,
+      "step": 3386
+    },
+    {
+      "epoch": 0.7755008586147681,
+      "grad_norm": 1.2906368970870972,
+      "learning_rate": 2.552853814302936e-06,
+      "loss": 1.0622,
+      "step": 3387
+    },
+    {
+      "epoch": 0.7757298225529479,
+      "grad_norm": 1.388886570930481,
+      "learning_rate": 2.547905106569677e-06,
+      "loss": 1.0753,
+      "step": 3388
+    },
+    {
+      "epoch": 0.7759587864911276,
+      "grad_norm": 1.221598744392395,
+      "learning_rate": 2.5429604996617653e-06,
+      "loss": 1.0204,
+      "step": 3389
+    },
+    {
+      "epoch": 0.7761877504293074,
+      "grad_norm": 1.2661128044128418,
+      "learning_rate": 2.5380199963001684e-06,
+      "loss": 1.0461,
+      "step": 3390
+    },
+    {
+      "epoch": 0.7764167143674872,
+      "grad_norm": 3.6229851245880127,
+      "learning_rate": 2.5330835992036062e-06,
+      "loss": 1.1383,
+      "step": 3391
+    },
+    {
+      "epoch": 0.7766456783056669,
+      "grad_norm": 1.637839913368225,
+      "learning_rate": 2.528151311088537e-06,
+      "loss": 1.0802,
+      "step": 3392
+    },
+    {
+      "epoch": 0.7768746422438466,
+      "grad_norm": 1.3283582925796509,
+      "learning_rate": 2.523223134669157e-06,
+      "loss": 1.0888,
+      "step": 3393
+    },
+    {
+      "epoch": 0.7771036061820263,
+      "grad_norm": 1.1502894163131714,
+      "learning_rate": 2.518299072657403e-06,
+      "loss": 1.0402,
+      "step": 3394
+    },
+    {
+      "epoch": 0.7773325701202061,
+      "grad_norm": 1.397196888923645,
+      "learning_rate": 2.513379127762937e-06,
+      "loss": 1.048,
+      "step": 3395
+    },
+    {
+      "epoch": 0.7775615340583858,
+      "grad_norm": 1.6322966814041138,
+      "learning_rate": 2.5084633026931727e-06,
+      "loss": 1.0891,
+      "step": 3396
+    },
+    {
+      "epoch": 0.7777904979965655,
+      "grad_norm": 1.2721960544586182,
+      "learning_rate": 2.5035516001532467e-06,
+      "loss": 1.0438,
+      "step": 3397
+    },
+    {
+      "epoch": 0.7780194619347452,
+      "grad_norm": 1.2815996408462524,
+      "learning_rate": 2.4986440228460185e-06,
+      "loss": 1.0351,
+      "step": 3398
+    },
+    {
+      "epoch": 0.778248425872925,
+      "grad_norm": 1.0550155639648438,
+      "learning_rate": 2.4937405734720964e-06,
+      "loss": 1.0648,
+      "step": 3399
+    },
+    {
+      "epoch": 0.7784773898111047,
+      "grad_norm": 1.814061164855957,
+      "learning_rate": 2.488841254729808e-06,
+      "loss": 1.0793,
+      "step": 3400
+    },
+    {
+      "epoch": 0.7787063537492845,
+      "grad_norm": 1.3419748544692993,
+      "learning_rate": 2.4839460693151994e-06,
+      "loss": 1.092,
+      "step": 3401
+    },
+    {
+      "epoch": 0.7789353176874643,
+      "grad_norm": 1.4719575643539429,
+      "learning_rate": 2.4790550199220543e-06,
+      "loss": 1.0483,
+      "step": 3402
+    },
+    {
+      "epoch": 0.779164281625644,
+      "grad_norm": 1.003715991973877,
+      "learning_rate": 2.474168109241877e-06,
+      "loss": 0.984,
+      "step": 3403
+    },
+    {
+      "epoch": 0.7793932455638237,
+      "grad_norm": 1.2123103141784668,
+      "learning_rate": 2.469285339963892e-06,
+      "loss": 1.042,
+      "step": 3404
+    },
+    {
+      "epoch": 0.7796222095020034,
+      "grad_norm": 1.2390992641448975,
+      "learning_rate": 2.4644067147750462e-06,
+      "loss": 0.9107,
+      "step": 3405
+    },
+    {
+      "epoch": 0.7798511734401832,
+      "grad_norm": 1.2249623537063599,
+      "learning_rate": 2.459532236360007e-06,
+      "loss": 1.0057,
+      "step": 3406
+    },
+    {
+      "epoch": 0.7800801373783629,
+      "grad_norm": 1.6221455335617065,
+      "learning_rate": 2.4546619074011603e-06,
+      "loss": 0.9997,
+      "step": 3407
+    },
+    {
+      "epoch": 0.7803091013165426,
+      "grad_norm": 1.1947602033615112,
+      "learning_rate": 2.4497957305786046e-06,
+      "loss": 0.9835,
+      "step": 3408
+    },
+    {
+      "epoch": 0.7805380652547224,
+      "grad_norm": 1.270727515220642,
+      "learning_rate": 2.4449337085701573e-06,
+      "loss": 1.0053,
+      "step": 3409
+    },
+    {
+      "epoch": 0.7807670291929021,
+      "grad_norm": 1.52069890499115,
+      "learning_rate": 2.4400758440513516e-06,
+      "loss": 1.0603,
+      "step": 3410
+    },
+    {
+      "epoch": 0.7809959931310818,
+      "grad_norm": 1.472410798072815,
+      "learning_rate": 2.4352221396954233e-06,
+      "loss": 1.0715,
+      "step": 3411
+    },
+    {
+      "epoch": 0.7812249570692615,
+      "grad_norm": 1.2387964725494385,
+      "learning_rate": 2.430372598173326e-06,
+      "loss": 1.0548,
+      "step": 3412
+    },
+    {
+      "epoch": 0.7814539210074414,
+      "grad_norm": 1.2153174877166748,
+      "learning_rate": 2.4255272221537295e-06,
+      "loss": 1.0096,
+      "step": 3413
+    },
+    {
+      "epoch": 0.7816828849456211,
+      "grad_norm": 1.296165943145752,
+      "learning_rate": 2.4206860143029954e-06,
+      "loss": 1.0653,
+      "step": 3414
+    },
+    {
+      "epoch": 0.7819118488838008,
+      "grad_norm": 1.3834319114685059,
+      "learning_rate": 2.4158489772852035e-06,
+      "loss": 0.9823,
+      "step": 3415
+    },
+    {
+      "epoch": 0.7821408128219806,
+      "grad_norm": 1.151049017906189,
+      "learning_rate": 2.4110161137621325e-06,
+      "loss": 1.051,
+      "step": 3416
+    },
+    {
+      "epoch": 0.7823697767601603,
+      "grad_norm": 2.3822290897369385,
+      "learning_rate": 2.406187426393269e-06,
+      "loss": 1.0159,
+      "step": 3417
+    },
+    {
+      "epoch": 0.78259874069834,
+      "grad_norm": 1.6371638774871826,
+      "learning_rate": 2.401362917835798e-06,
+      "loss": 0.969,
+      "step": 3418
+    },
+    {
+      "epoch": 0.7828277046365197,
+      "grad_norm": 1.29193913936615,
+      "learning_rate": 2.396542590744606e-06,
+      "loss": 1.0759,
+      "step": 3419
+    },
+    {
+      "epoch": 0.7830566685746995,
+      "grad_norm": 1.1832704544067383,
+      "learning_rate": 2.391726447772279e-06,
+      "loss": 1.0444,
+      "step": 3420
+    },
+    {
+      "epoch": 0.7832856325128792,
+      "grad_norm": 1.0548069477081299,
+      "learning_rate": 2.3869144915691033e-06,
+      "loss": 1.0141,
+      "step": 3421
+    },
+    {
+      "epoch": 0.7835145964510589,
+      "grad_norm": 1.3056319952011108,
+      "learning_rate": 2.3821067247830488e-06,
+      "loss": 1.0866,
+      "step": 3422
+    },
+    {
+      "epoch": 0.7837435603892386,
+      "grad_norm": 1.4193148612976074,
+      "learning_rate": 2.3773031500597974e-06,
+      "loss": 1.0747,
+      "step": 3423
+    },
+    {
+      "epoch": 0.7839725243274185,
+      "grad_norm": 1.3721299171447754,
+      "learning_rate": 2.3725037700427168e-06,
+      "loss": 1.0653,
+      "step": 3424
+    },
+    {
+      "epoch": 0.7842014882655982,
+      "grad_norm": 1.2656069993972778,
+      "learning_rate": 2.3677085873728602e-06,
+      "loss": 1.0436,
+      "step": 3425
+    },
+    {
+      "epoch": 0.7844304522037779,
+      "grad_norm": 1.1870973110198975,
+      "learning_rate": 2.3629176046889755e-06,
+      "loss": 1.0881,
+      "step": 3426
+    },
+    {
+      "epoch": 0.7846594161419577,
+      "grad_norm": 2.626173257827759,
+      "learning_rate": 2.3581308246275103e-06,
+      "loss": 1.0576,
+      "step": 3427
+    },
+    {
+      "epoch": 0.7848883800801374,
+      "grad_norm": 1.569612741470337,
+      "learning_rate": 2.353348249822579e-06,
+      "loss": 1.0122,
+      "step": 3428
+    },
+    {
+      "epoch": 0.7851173440183171,
+      "grad_norm": 1.4496647119522095,
+      "learning_rate": 2.3485698829059967e-06,
+      "loss": 1.0401,
+      "step": 3429
+    },
+    {
+      "epoch": 0.7853463079564968,
+      "grad_norm": 1.2720706462860107,
+      "learning_rate": 2.3437957265072587e-06,
+      "loss": 1.0253,
+      "step": 3430
+    },
+    {
+      "epoch": 0.7855752718946766,
+      "grad_norm": 1.1991873979568481,
+      "learning_rate": 2.339025783253541e-06,
+      "loss": 1.075,
+      "step": 3431
+    },
+    {
+      "epoch": 0.7858042358328563,
+      "grad_norm": 1.2989721298217773,
+      "learning_rate": 2.334260055769707e-06,
+      "loss": 1.0634,
+      "step": 3432
+    },
+    {
+      "epoch": 0.786033199771036,
+      "grad_norm": 1.156073808670044,
+      "learning_rate": 2.3294985466782937e-06,
+      "loss": 1.0094,
+      "step": 3433
+    },
+    {
+      "epoch": 0.7862621637092158,
+      "grad_norm": 1.2679800987243652,
+      "learning_rate": 2.324741258599521e-06,
+      "loss": 1.0954,
+      "step": 3434
+    },
+    {
+      "epoch": 0.7864911276473955,
+      "grad_norm": 1.283085823059082,
+      "learning_rate": 2.319988194151287e-06,
+      "loss": 0.9946,
+      "step": 3435
+    },
+    {
+      "epoch": 0.7867200915855753,
+      "grad_norm": 1.6633352041244507,
+      "learning_rate": 2.3152393559491546e-06,
+      "loss": 1.0509,
+      "step": 3436
+    },
+    {
+      "epoch": 0.786949055523755,
+      "grad_norm": 1.3661447763442993,
+      "learning_rate": 2.3104947466063785e-06,
+      "loss": 1.0671,
+      "step": 3437
+    },
+    {
+      "epoch": 0.7871780194619348,
+      "grad_norm": 1.2480249404907227,
+      "learning_rate": 2.305754368733878e-06,
+      "loss": 1.0529,
+      "step": 3438
+    },
+    {
+      "epoch": 0.7874069834001145,
+      "grad_norm": 1.286214828491211,
+      "learning_rate": 2.3010182249402368e-06,
+      "loss": 1.0186,
+      "step": 3439
+    },
+    {
+      "epoch": 0.7876359473382942,
+      "grad_norm": 1.268823504447937,
+      "learning_rate": 2.2962863178317154e-06,
+      "loss": 1.0093,
+      "step": 3440
+    },
+    {
+      "epoch": 0.787864911276474,
+      "grad_norm": 1.215895652770996,
+      "learning_rate": 2.29155865001225e-06,
+      "loss": 1.0696,
+      "step": 3441
+    },
+    {
+      "epoch": 0.7880938752146537,
+      "grad_norm": 1.2569650411605835,
+      "learning_rate": 2.2868352240834304e-06,
+      "loss": 1.0109,
+      "step": 3442
+    },
+    {
+      "epoch": 0.7883228391528334,
+      "grad_norm": 1.3006694316864014,
+      "learning_rate": 2.282116042644519e-06,
+      "loss": 1.0191,
+      "step": 3443
+    },
+    {
+      "epoch": 0.7885518030910131,
+      "grad_norm": 1.2813971042633057,
+      "learning_rate": 2.2774011082924417e-06,
+      "loss": 1.0308,
+      "step": 3444
+    },
+    {
+      "epoch": 0.7887807670291929,
+      "grad_norm": 1.4007090330123901,
+      "learning_rate": 2.2726904236217895e-06,
+      "loss": 1.0365,
+      "step": 3445
+    },
+    {
+      "epoch": 0.7890097309673726,
+      "grad_norm": 1.9624810218811035,
+      "learning_rate": 2.2679839912248104e-06,
+      "loss": 1.0467,
+      "step": 3446
+    },
+    {
+      "epoch": 0.7892386949055524,
+      "grad_norm": 1.6809523105621338,
+      "learning_rate": 2.263281813691417e-06,
+      "loss": 1.025,
+      "step": 3447
+    },
+    {
+      "epoch": 0.7894676588437322,
+      "grad_norm": 1.4995564222335815,
+      "learning_rate": 2.2585838936091753e-06,
+      "loss": 1.0324,
+      "step": 3448
+    },
+    {
+      "epoch": 0.7896966227819119,
+      "grad_norm": 1.4374549388885498,
+      "learning_rate": 2.253890233563316e-06,
+      "loss": 1.0693,
+      "step": 3449
+    },
+    {
+      "epoch": 0.7899255867200916,
+      "grad_norm": 1.204321265220642,
+      "learning_rate": 2.2492008361367133e-06,
+      "loss": 1.0055,
+      "step": 3450
+    },
+    {
+      "epoch": 0.7901545506582713,
+      "grad_norm": 1.4027067422866821,
+      "learning_rate": 2.2445157039099096e-06,
+      "loss": 1.0308,
+      "step": 3451
+    },
+    {
+      "epoch": 0.7903835145964511,
+      "grad_norm": 1.5836561918258667,
+      "learning_rate": 2.2398348394610947e-06,
+      "loss": 1.0295,
+      "step": 3452
+    },
+    {
+      "epoch": 0.7906124785346308,
+      "grad_norm": 1.5846806764602661,
+      "learning_rate": 2.235158245366105e-06,
+      "loss": 1.0415,
+      "step": 3453
+    },
+    {
+      "epoch": 0.7908414424728105,
+      "grad_norm": 1.1941750049591064,
+      "learning_rate": 2.2304859241984313e-06,
+      "loss": 1.0086,
+      "step": 3454
+    },
+    {
+      "epoch": 0.7910704064109902,
+      "grad_norm": 1.2206698656082153,
+      "learning_rate": 2.225817878529214e-06,
+      "loss": 1.0253,
+      "step": 3455
+    },
+    {
+      "epoch": 0.79129937034917,
+      "grad_norm": 1.6405779123306274,
+      "learning_rate": 2.2211541109272383e-06,
+      "loss": 0.983,
+      "step": 3456
+    },
+    {
+      "epoch": 0.7915283342873497,
+      "grad_norm": 1.773629903793335,
+      "learning_rate": 2.216494623958939e-06,
+      "loss": 0.9978,
+      "step": 3457
+    },
+    {
+      "epoch": 0.7917572982255294,
+      "grad_norm": 1.5641894340515137,
+      "learning_rate": 2.2118394201883907e-06,
+      "loss": 1.0389,
+      "step": 3458
+    },
+    {
+      "epoch": 0.7919862621637093,
+      "grad_norm": 1.358713984489441,
+      "learning_rate": 2.207188502177313e-06,
+      "loss": 1.0336,
+      "step": 3459
+    },
+    {
+      "epoch": 0.792215226101889,
+      "grad_norm": 1.1877168416976929,
+      "learning_rate": 2.2025418724850678e-06,
+      "loss": 1.1138,
+      "step": 3460
+    },
+    {
+      "epoch": 0.7924441900400687,
+      "grad_norm": 1.3562843799591064,
+      "learning_rate": 2.197899533668657e-06,
+      "loss": 1.0429,
+      "step": 3461
+    },
+    {
+      "epoch": 0.7926731539782484,
+      "grad_norm": 1.2236860990524292,
+      "learning_rate": 2.1932614882827196e-06,
+      "loss": 1.0147,
+      "step": 3462
+    },
+    {
+      "epoch": 0.7929021179164282,
+      "grad_norm": 1.2904807329177856,
+      "learning_rate": 2.1886277388795363e-06,
+      "loss": 1.0654,
+      "step": 3463
+    },
+    {
+      "epoch": 0.7931310818546079,
+      "grad_norm": 1.3184075355529785,
+      "learning_rate": 2.1839982880090115e-06,
+      "loss": 1.1185,
+      "step": 3464
+    },
+    {
+      "epoch": 0.7933600457927876,
+      "grad_norm": 1.4988008737564087,
+      "learning_rate": 2.1793731382187056e-06,
+      "loss": 1.0399,
+      "step": 3465
+    },
+    {
+      "epoch": 0.7935890097309674,
+      "grad_norm": 1.8388975858688354,
+      "learning_rate": 2.1747522920537913e-06,
+      "loss": 1.0708,
+      "step": 3466
+    },
+    {
+      "epoch": 0.7938179736691471,
+      "grad_norm": 1.239767074584961,
+      "learning_rate": 2.1701357520570797e-06,
+      "loss": 1.0156,
+      "step": 3467
+    },
+    {
+      "epoch": 0.7940469376073268,
+      "grad_norm": 1.1681170463562012,
+      "learning_rate": 2.165523520769024e-06,
+      "loss": 1.0439,
+      "step": 3468
+    },
+    {
+      "epoch": 0.7942759015455065,
+      "grad_norm": 1.1504888534545898,
+      "learning_rate": 2.160915600727688e-06,
+      "loss": 1.0851,
+      "step": 3469
+    },
+    {
+      "epoch": 0.7945048654836864,
+      "grad_norm": 1.3817356824874878,
+      "learning_rate": 2.156311994468774e-06,
+      "loss": 1.0873,
+      "step": 3470
+    },
+    {
+      "epoch": 0.7947338294218661,
+      "grad_norm": 1.3247697353363037,
+      "learning_rate": 2.151712704525608e-06,
+      "loss": 0.9901,
+      "step": 3471
+    },
+    {
+      "epoch": 0.7949627933600458,
+      "grad_norm": 1.5269098281860352,
+      "learning_rate": 2.1471177334291404e-06,
+      "loss": 1.0721,
+      "step": 3472
+    },
+    {
+      "epoch": 0.7951917572982256,
+      "grad_norm": 1.195955753326416,
+      "learning_rate": 2.142527083707946e-06,
+      "loss": 1.0325,
+      "step": 3473
+    },
+    {
+      "epoch": 0.7954207212364053,
+      "grad_norm": 1.3172709941864014,
+      "learning_rate": 2.1379407578882206e-06,
+      "loss": 1.0282,
+      "step": 3474
+    },
+    {
+      "epoch": 0.795649685174585,
+      "grad_norm": 1.4341003894805908,
+      "learning_rate": 2.13335875849378e-06,
+      "loss": 1.098,
+      "step": 3475
+    },
+    {
+      "epoch": 0.7958786491127647,
+      "grad_norm": 1.1495006084442139,
+      "learning_rate": 2.1287810880460636e-06,
+      "loss": 1.0147,
+      "step": 3476
+    },
+    {
+      "epoch": 0.7961076130509445,
+      "grad_norm": 1.1693551540374756,
+      "learning_rate": 2.1242077490641157e-06,
+      "loss": 1.0005,
+      "step": 3477
+    },
+    {
+      "epoch": 0.7963365769891242,
+      "grad_norm": 1.5552064180374146,
+      "learning_rate": 2.119638744064617e-06,
+      "loss": 1.0714,
+      "step": 3478
+    },
+    {
+      "epoch": 0.7965655409273039,
+      "grad_norm": 1.519784688949585,
+      "learning_rate": 2.1150740755618505e-06,
+      "loss": 1.04,
+      "step": 3479
+    },
+    {
+      "epoch": 0.7967945048654836,
+      "grad_norm": 1.1543006896972656,
+      "learning_rate": 2.1105137460677093e-06,
+      "loss": 1.0897,
+      "step": 3480
+    },
+    {
+      "epoch": 0.7970234688036634,
+      "grad_norm": 1.7196766138076782,
+      "learning_rate": 2.1059577580917067e-06,
+      "loss": 1.0252,
+      "step": 3481
+    },
+    {
+      "epoch": 0.7972524327418432,
+      "grad_norm": 1.2323554754257202,
+      "learning_rate": 2.1014061141409715e-06,
+      "loss": 1.0197,
+      "step": 3482
+    },
+    {
+      "epoch": 0.7974813966800229,
+      "grad_norm": 1.1643589735031128,
+      "learning_rate": 2.0968588167202265e-06,
+      "loss": 1.023,
+      "step": 3483
+    },
+    {
+      "epoch": 0.7977103606182027,
+      "grad_norm": 1.2324453592300415,
+      "learning_rate": 2.0923158683318157e-06,
+      "loss": 0.9976,
+      "step": 3484
+    },
+    {
+      "epoch": 0.7979393245563824,
+      "grad_norm": 1.3599796295166016,
+      "learning_rate": 2.087777271475684e-06,
+      "loss": 1.0451,
+      "step": 3485
+    },
+    {
+      "epoch": 0.7981682884945621,
+      "grad_norm": 1.2370396852493286,
+      "learning_rate": 2.0832430286493834e-06,
+      "loss": 0.9734,
+      "step": 3486
+    },
+    {
+      "epoch": 0.7983972524327418,
+      "grad_norm": 1.3840800523757935,
+      "learning_rate": 2.0787131423480722e-06,
+      "loss": 0.9833,
+      "step": 3487
+    },
+    {
+      "epoch": 0.7986262163709216,
+      "grad_norm": 1.1826318502426147,
+      "learning_rate": 2.0741876150645025e-06,
+      "loss": 0.9921,
+      "step": 3488
+    },
+    {
+      "epoch": 0.7988551803091013,
+      "grad_norm": 1.1588865518569946,
+      "learning_rate": 2.0696664492890394e-06,
+      "loss": 1.0858,
+      "step": 3489
+    },
+    {
+      "epoch": 0.799084144247281,
+      "grad_norm": 1.4045047760009766,
+      "learning_rate": 2.0651496475096455e-06,
+      "loss": 1.0409,
+      "step": 3490
+    },
+    {
+      "epoch": 0.7993131081854608,
+      "grad_norm": 1.2877663373947144,
+      "learning_rate": 2.060637212211869e-06,
+      "loss": 0.9848,
+      "step": 3491
+    },
+    {
+      "epoch": 0.7995420721236405,
+      "grad_norm": 1.3673158884048462,
+      "learning_rate": 2.0561291458788736e-06,
+      "loss": 0.9969,
+      "step": 3492
+    },
+    {
+      "epoch": 0.7997710360618203,
+      "grad_norm": 1.3965160846710205,
+      "learning_rate": 2.0516254509914103e-06,
+      "loss": 1.0477,
+      "step": 3493
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1263400316238403,
+      "learning_rate": 2.047126130027819e-06,
+      "loss": 1.0591,
+      "step": 3494
+    },
+    {
+      "epoch": 0.8002289639381798,
+      "grad_norm": 1.2847234010696411,
+      "learning_rate": 2.042631185464039e-06,
+      "loss": 0.9993,
+      "step": 3495
+    },
+    {
+      "epoch": 0.8004579278763595,
+      "grad_norm": 2.184096336364746,
+      "learning_rate": 2.038140619773609e-06,
+      "loss": 1.0074,
+      "step": 3496
+    },
+    {
+      "epoch": 0.8006868918145392,
+      "grad_norm": 1.3483428955078125,
+      "learning_rate": 2.03365443542764e-06,
+      "loss": 1.0989,
+      "step": 3497
+    },
+    {
+      "epoch": 0.800915855752719,
+      "grad_norm": 1.5538264513015747,
+      "learning_rate": 2.029172634894846e-06,
+      "loss": 1.0901,
+      "step": 3498
+    },
+    {
+      "epoch": 0.8011448196908987,
+      "grad_norm": 1.2614027261734009,
+      "learning_rate": 2.024695220641524e-06,
+      "loss": 1.0144,
+      "step": 3499
+    },
+    {
+      "epoch": 0.8013737836290784,
+      "grad_norm": 1.2823106050491333,
+      "learning_rate": 2.020222195131556e-06,
+      "loss": 0.9797,
+      "step": 3500
+    },
+    {
+      "epoch": 0.8016027475672581,
+      "grad_norm": 1.257866382598877,
+      "learning_rate": 2.0157535608264123e-06,
+      "loss": 1.0476,
+      "step": 3501
+    },
+    {
+      "epoch": 0.8018317115054379,
+      "grad_norm": 1.226118803024292,
+      "learning_rate": 2.0112893201851435e-06,
+      "loss": 0.9959,
+      "step": 3502
+    },
+    {
+      "epoch": 0.8020606754436176,
+      "grad_norm": 1.5508838891983032,
+      "learning_rate": 2.0068294756643846e-06,
+      "loss": 0.9996,
+      "step": 3503
+    },
+    {
+      "epoch": 0.8022896393817973,
+      "grad_norm": 1.1606311798095703,
+      "learning_rate": 2.0023740297183536e-06,
+      "loss": 1.0322,
+      "step": 3504
+    },
+    {
+      "epoch": 0.8025186033199772,
+      "grad_norm": 1.516457200050354,
+      "learning_rate": 1.997922984798836e-06,
+      "loss": 1.0247,
+      "step": 3505
+    },
+    {
+      "epoch": 0.8027475672581569,
+      "grad_norm": 2.4328088760375977,
+      "learning_rate": 1.993476343355213e-06,
+      "loss": 1.0883,
+      "step": 3506
+    },
+    {
+      "epoch": 0.8029765311963366,
+      "grad_norm": 1.4516451358795166,
+      "learning_rate": 1.9890341078344343e-06,
+      "loss": 1.0693,
+      "step": 3507
+    },
+    {
+      "epoch": 0.8032054951345163,
+      "grad_norm": 1.3910856246948242,
+      "learning_rate": 1.9845962806810205e-06,
+      "loss": 1.0984,
+      "step": 3508
+    },
+    {
+      "epoch": 0.8034344590726961,
+      "grad_norm": 1.184085488319397,
+      "learning_rate": 1.980162864337071e-06,
+      "loss": 1.0698,
+      "step": 3509
+    },
+    {
+      "epoch": 0.8036634230108758,
+      "grad_norm": 1.3776863813400269,
+      "learning_rate": 1.9757338612422594e-06,
+      "loss": 0.9953,
+      "step": 3510
+    },
+    {
+      "epoch": 0.8038923869490555,
+      "grad_norm": 1.102057933807373,
+      "learning_rate": 1.971309273833828e-06,
+      "loss": 1.0835,
+      "step": 3511
+    },
+    {
+      "epoch": 0.8041213508872352,
+      "grad_norm": 1.2618690729141235,
+      "learning_rate": 1.966889104546591e-06,
+      "loss": 1.0462,
+      "step": 3512
+    },
+    {
+      "epoch": 0.804350314825415,
+      "grad_norm": 1.2333935499191284,
+      "learning_rate": 1.9624733558129304e-06,
+      "loss": 1.0208,
+      "step": 3513
+    },
+    {
+      "epoch": 0.8045792787635947,
+      "grad_norm": 1.2381764650344849,
+      "learning_rate": 1.958062030062795e-06,
+      "loss": 1.0503,
+      "step": 3514
+    },
+    {
+      "epoch": 0.8048082427017744,
+      "grad_norm": 1.2884087562561035,
+      "learning_rate": 1.9536551297237018e-06,
+      "loss": 1.0545,
+      "step": 3515
+    },
+    {
+      "epoch": 0.8050372066399543,
+      "grad_norm": 1.294965147972107,
+      "learning_rate": 1.9492526572207294e-06,
+      "loss": 1.0354,
+      "step": 3516
+    },
+    {
+      "epoch": 0.805266170578134,
+      "grad_norm": 1.5084460973739624,
+      "learning_rate": 1.944854614976521e-06,
+      "loss": 0.9814,
+      "step": 3517
+    },
+    {
+      "epoch": 0.8054951345163137,
+      "grad_norm": 1.1529735326766968,
+      "learning_rate": 1.940461005411288e-06,
+      "loss": 1.0445,
+      "step": 3518
+    },
+    {
+      "epoch": 0.8057240984544934,
+      "grad_norm": 1.3352361917495728,
+      "learning_rate": 1.9360718309427863e-06,
+      "loss": 0.9968,
+      "step": 3519
+    },
+    {
+      "epoch": 0.8059530623926732,
+      "grad_norm": 1.3624452352523804,
+      "learning_rate": 1.931687093986354e-06,
+      "loss": 1.0275,
+      "step": 3520
+    },
+    {
+      "epoch": 0.8061820263308529,
+      "grad_norm": 1.2383012771606445,
+      "learning_rate": 1.9273067969548664e-06,
+      "loss": 1.0416,
+      "step": 3521
+    },
+    {
+      "epoch": 0.8064109902690326,
+      "grad_norm": 1.340103030204773,
+      "learning_rate": 1.922930942258766e-06,
+      "loss": 1.0579,
+      "step": 3522
+    },
+    {
+      "epoch": 0.8066399542072124,
+      "grad_norm": 2.9526262283325195,
+      "learning_rate": 1.918559532306051e-06,
+      "loss": 1.0087,
+      "step": 3523
+    },
+    {
+      "epoch": 0.8068689181453921,
+      "grad_norm": 1.7514522075653076,
+      "learning_rate": 1.91419256950227e-06,
+      "loss": 1.0383,
+      "step": 3524
+    },
+    {
+      "epoch": 0.8070978820835718,
+      "grad_norm": 1.2114307880401611,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.9728,
+      "step": 3525
+    },
+    {
+      "epoch": 0.8073268460217515,
+      "grad_norm": 1.102329969406128,
+      "learning_rate": 1.9054719949514756e-06,
+      "loss": 1.0953,
+      "step": 3526
+    },
+    {
+      "epoch": 0.8075558099599313,
+      "grad_norm": 1.2721331119537354,
+      "learning_rate": 1.9011183880033203e-06,
+      "loss": 1.0068,
+      "step": 3527
+    },
+    {
+      "epoch": 0.8077847738981111,
+      "grad_norm": 1.7620582580566406,
+      "learning_rate": 1.8967692378018155e-06,
+      "loss": 0.9938,
+      "step": 3528
+    },
+    {
+      "epoch": 0.8080137378362908,
+      "grad_norm": 1.253893494606018,
+      "learning_rate": 1.8924245467402612e-06,
+      "loss": 1.0459,
+      "step": 3529
+    },
+    {
+      "epoch": 0.8082427017744706,
+      "grad_norm": 1.549721360206604,
+      "learning_rate": 1.8880843172095066e-06,
+      "loss": 1.0437,
+      "step": 3530
+    },
+    {
+      "epoch": 0.8084716657126503,
+      "grad_norm": 1.4003283977508545,
+      "learning_rate": 1.8837485515979425e-06,
+      "loss": 1.0505,
+      "step": 3531
+    },
+    {
+      "epoch": 0.80870062965083,
+      "grad_norm": 1.227433681488037,
+      "learning_rate": 1.8794172522915022e-06,
+      "loss": 1.0111,
+      "step": 3532
+    },
+    {
+      "epoch": 0.8089295935890097,
+      "grad_norm": 1.263609528541565,
+      "learning_rate": 1.875090421673662e-06,
+      "loss": 1.0436,
+      "step": 3533
+    },
+    {
+      "epoch": 0.8091585575271895,
+      "grad_norm": 1.4125272035598755,
+      "learning_rate": 1.8707680621254487e-06,
+      "loss": 1.061,
+      "step": 3534
+    },
+    {
+      "epoch": 0.8093875214653692,
+      "grad_norm": 1.1441245079040527,
+      "learning_rate": 1.8664501760254128e-06,
+      "loss": 1.0341,
+      "step": 3535
+    },
+    {
+      "epoch": 0.8096164854035489,
+      "grad_norm": 1.3769537210464478,
+      "learning_rate": 1.8621367657496504e-06,
+      "loss": 1.0348,
+      "step": 3536
+    },
+    {
+      "epoch": 0.8098454493417286,
+      "grad_norm": 1.3794772624969482,
+      "learning_rate": 1.8578278336718037e-06,
+      "loss": 1.0156,
+      "step": 3537
+    },
+    {
+      "epoch": 0.8100744132799084,
+      "grad_norm": 1.4838491678237915,
+      "learning_rate": 1.8535233821630338e-06,
+      "loss": 0.9982,
+      "step": 3538
+    },
+    {
+      "epoch": 0.8103033772180882,
+      "grad_norm": 1.272406816482544,
+      "learning_rate": 1.849223413592046e-06,
+      "loss": 1.0323,
+      "step": 3539
+    },
+    {
+      "epoch": 0.8105323411562679,
+      "grad_norm": 1.1917959451675415,
+      "learning_rate": 1.8449279303250777e-06,
+      "loss": 1.0565,
+      "step": 3540
+    },
+    {
+      "epoch": 0.8107613050944477,
+      "grad_norm": 1.2484959363937378,
+      "learning_rate": 1.8406369347258968e-06,
+      "loss": 1.0529,
+      "step": 3541
+    },
+    {
+      "epoch": 0.8109902690326274,
+      "grad_norm": 1.6294903755187988,
+      "learning_rate": 1.8363504291558053e-06,
+      "loss": 1.0013,
+      "step": 3542
+    },
+    {
+      "epoch": 0.8112192329708071,
+      "grad_norm": 1.355380654335022,
+      "learning_rate": 1.8320684159736236e-06,
+      "loss": 0.9698,
+      "step": 3543
+    },
+    {
+      "epoch": 0.8114481969089868,
+      "grad_norm": 1.0993623733520508,
+      "learning_rate": 1.827790897535715e-06,
+      "loss": 1.0257,
+      "step": 3544
+    },
+    {
+      "epoch": 0.8116771608471666,
+      "grad_norm": 1.8703913688659668,
+      "learning_rate": 1.8235178761959626e-06,
+      "loss": 1.0786,
+      "step": 3545
+    },
+    {
+      "epoch": 0.8119061247853463,
+      "grad_norm": 1.225956678390503,
+      "learning_rate": 1.8192493543057676e-06,
+      "loss": 1.0386,
+      "step": 3546
+    },
+    {
+      "epoch": 0.812135088723526,
+      "grad_norm": 1.302952527999878,
+      "learning_rate": 1.8149853342140644e-06,
+      "loss": 1.011,
+      "step": 3547
+    },
+    {
+      "epoch": 0.8123640526617057,
+      "grad_norm": 1.4825295209884644,
+      "learning_rate": 1.8107258182673127e-06,
+      "loss": 1.0508,
+      "step": 3548
+    },
+    {
+      "epoch": 0.8125930165998855,
+      "grad_norm": 1.4160746335983276,
+      "learning_rate": 1.8064708088094829e-06,
+      "loss": 1.0045,
+      "step": 3549
+    },
+    {
+      "epoch": 0.8128219805380652,
+      "grad_norm": 1.3948523998260498,
+      "learning_rate": 1.802220308182071e-06,
+      "loss": 1.0562,
+      "step": 3550
+    },
+    {
+      "epoch": 0.813050944476245,
+      "grad_norm": 1.3531475067138672,
+      "learning_rate": 1.797974318724094e-06,
+      "loss": 0.9929,
+      "step": 3551
+    },
+    {
+      "epoch": 0.8132799084144248,
+      "grad_norm": 1.301820993423462,
+      "learning_rate": 1.7937328427720834e-06,
+      "loss": 1.0954,
+      "step": 3552
+    },
+    {
+      "epoch": 0.8135088723526045,
+      "grad_norm": 1.1766568422317505,
+      "learning_rate": 1.7894958826600884e-06,
+      "loss": 1.0252,
+      "step": 3553
+    },
+    {
+      "epoch": 0.8137378362907842,
+      "grad_norm": 1.2368574142456055,
+      "learning_rate": 1.7852634407196723e-06,
+      "loss": 0.9867,
+      "step": 3554
+    },
+    {
+      "epoch": 0.813966800228964,
+      "grad_norm": 1.2918037176132202,
+      "learning_rate": 1.7810355192799122e-06,
+      "loss": 1.0437,
+      "step": 3555
+    },
+    {
+      "epoch": 0.8141957641671437,
+      "grad_norm": 1.3767991065979004,
+      "learning_rate": 1.7768121206674006e-06,
+      "loss": 1.0281,
+      "step": 3556
+    },
+    {
+      "epoch": 0.8144247281053234,
+      "grad_norm": 1.165218472480774,
+      "learning_rate": 1.7725932472062302e-06,
+      "loss": 0.9958,
+      "step": 3557
+    },
+    {
+      "epoch": 0.8146536920435031,
+      "grad_norm": 1.1615934371948242,
+      "learning_rate": 1.7683789012180196e-06,
+      "loss": 1.0294,
+      "step": 3558
+    },
+    {
+      "epoch": 0.8148826559816829,
+      "grad_norm": 1.3397858142852783,
+      "learning_rate": 1.7641690850218884e-06,
+      "loss": 1.0477,
+      "step": 3559
+    },
+    {
+      "epoch": 0.8151116199198626,
+      "grad_norm": 1.2796862125396729,
+      "learning_rate": 1.7599638009344566e-06,
+      "loss": 0.9805,
+      "step": 3560
+    },
+    {
+      "epoch": 0.8153405838580423,
+      "grad_norm": 1.1992462873458862,
+      "learning_rate": 1.7557630512698642e-06,
+      "loss": 1.0074,
+      "step": 3561
+    },
+    {
+      "epoch": 0.8155695477962221,
+      "grad_norm": 1.330796480178833,
+      "learning_rate": 1.7515668383397433e-06,
+      "loss": 1.0369,
+      "step": 3562
+    },
+    {
+      "epoch": 0.8157985117344019,
+      "grad_norm": 1.7938005924224854,
+      "learning_rate": 1.7473751644532366e-06,
+      "loss": 1.0118,
+      "step": 3563
+    },
+    {
+      "epoch": 0.8160274756725816,
+      "grad_norm": 1.2056411504745483,
+      "learning_rate": 1.7431880319169858e-06,
+      "loss": 1.0215,
+      "step": 3564
+    },
+    {
+      "epoch": 0.8162564396107613,
+      "grad_norm": 1.4476250410079956,
+      "learning_rate": 1.7390054430351366e-06,
+      "loss": 0.9641,
+      "step": 3565
+    },
+    {
+      "epoch": 0.8164854035489411,
+      "grad_norm": 1.1308528184890747,
+      "learning_rate": 1.7348274001093324e-06,
+      "loss": 1.0338,
+      "step": 3566
+    },
+    {
+      "epoch": 0.8167143674871208,
+      "grad_norm": 1.440428614616394,
+      "learning_rate": 1.730653905438714e-06,
+      "loss": 1.0503,
+      "step": 3567
+    },
+    {
+      "epoch": 0.8169433314253005,
+      "grad_norm": 1.2121968269348145,
+      "learning_rate": 1.7264849613199208e-06,
+      "loss": 1.0444,
+      "step": 3568
+    },
+    {
+      "epoch": 0.8171722953634802,
+      "grad_norm": 1.3329687118530273,
+      "learning_rate": 1.722320570047089e-06,
+      "loss": 1.0456,
+      "step": 3569
+    },
+    {
+      "epoch": 0.81740125930166,
+      "grad_norm": 1.237636923789978,
+      "learning_rate": 1.7181607339118488e-06,
+      "loss": 1.0376,
+      "step": 3570
+    },
+    {
+      "epoch": 0.8176302232398397,
+      "grad_norm": 7.456436634063721,
+      "learning_rate": 1.714005455203318e-06,
+      "loss": 1.0065,
+      "step": 3571
+    },
+    {
+      "epoch": 0.8178591871780194,
+      "grad_norm": 1.594241976737976,
+      "learning_rate": 1.7098547362081197e-06,
+      "loss": 0.9686,
+      "step": 3572
+    },
+    {
+      "epoch": 0.8180881511161991,
+      "grad_norm": 1.2208811044692993,
+      "learning_rate": 1.7057085792103534e-06,
+      "loss": 0.965,
+      "step": 3573
+    },
+    {
+      "epoch": 0.818317115054379,
+      "grad_norm": 1.1306706666946411,
+      "learning_rate": 1.701566986491614e-06,
+      "loss": 1.0411,
+      "step": 3574
+    },
+    {
+      "epoch": 0.8185460789925587,
+      "grad_norm": 1.4703606367111206,
+      "learning_rate": 1.697429960330993e-06,
+      "loss": 1.063,
+      "step": 3575
+    },
+    {
+      "epoch": 0.8187750429307384,
+      "grad_norm": 1.7712504863739014,
+      "learning_rate": 1.6932975030050524e-06,
+      "loss": 1.03,
+      "step": 3576
+    },
+    {
+      "epoch": 0.8190040068689182,
+      "grad_norm": 1.5078486204147339,
+      "learning_rate": 1.6891696167878535e-06,
+      "loss": 1.0125,
+      "step": 3577
+    },
+    {
+      "epoch": 0.8192329708070979,
+      "grad_norm": 1.5069183111190796,
+      "learning_rate": 1.6850463039509356e-06,
+      "loss": 1.0039,
+      "step": 3578
+    },
+    {
+      "epoch": 0.8194619347452776,
+      "grad_norm": 1.3289002180099487,
+      "learning_rate": 1.680927566763325e-06,
+      "loss": 1.0696,
+      "step": 3579
+    },
+    {
+      "epoch": 0.8196908986834573,
+      "grad_norm": 1.418169617652893,
+      "learning_rate": 1.6768134074915277e-06,
+      "loss": 1.0456,
+      "step": 3580
+    },
+    {
+      "epoch": 0.8199198626216371,
+      "grad_norm": 1.5103626251220703,
+      "learning_rate": 1.672703828399529e-06,
+      "loss": 1.0815,
+      "step": 3581
+    },
+    {
+      "epoch": 0.8201488265598168,
+      "grad_norm": 1.2844433784484863,
+      "learning_rate": 1.6685988317487988e-06,
+      "loss": 1.0614,
+      "step": 3582
+    },
+    {
+      "epoch": 0.8203777904979965,
+      "grad_norm": 1.187903881072998,
+      "learning_rate": 1.6644984197982828e-06,
+      "loss": 1.0207,
+      "step": 3583
+    },
+    {
+      "epoch": 0.8206067544361763,
+      "grad_norm": 1.2992799282073975,
+      "learning_rate": 1.6604025948043966e-06,
+      "loss": 1.0443,
+      "step": 3584
+    },
+    {
+      "epoch": 0.8208357183743561,
+      "grad_norm": 1.6380910873413086,
+      "learning_rate": 1.6563113590210455e-06,
+      "loss": 1.112,
+      "step": 3585
+    },
+    {
+      "epoch": 0.8210646823125358,
+      "grad_norm": 1.1910545825958252,
+      "learning_rate": 1.652224714699603e-06,
+      "loss": 1.0636,
+      "step": 3586
+    },
+    {
+      "epoch": 0.8212936462507155,
+      "grad_norm": 1.3377147912979126,
+      "learning_rate": 1.6481426640889098e-06,
+      "loss": 1.0045,
+      "step": 3587
+    },
+    {
+      "epoch": 0.8215226101888953,
+      "grad_norm": 1.2766714096069336,
+      "learning_rate": 1.6440652094352838e-06,
+      "loss": 1.017,
+      "step": 3588
+    },
+    {
+      "epoch": 0.821751574127075,
+      "grad_norm": 1.1633610725402832,
+      "learning_rate": 1.6399923529825213e-06,
+      "loss": 1.0152,
+      "step": 3589
+    },
+    {
+      "epoch": 0.8219805380652547,
+      "grad_norm": 1.4114861488342285,
+      "learning_rate": 1.6359240969718748e-06,
+      "loss": 1.0769,
+      "step": 3590
+    },
+    {
+      "epoch": 0.8222095020034345,
+      "grad_norm": 1.3905683755874634,
+      "learning_rate": 1.6318604436420738e-06,
+      "loss": 1.0795,
+      "step": 3591
+    },
+    {
+      "epoch": 0.8224384659416142,
+      "grad_norm": 1.3522437810897827,
+      "learning_rate": 1.6278013952293115e-06,
+      "loss": 1.0389,
+      "step": 3592
+    },
+    {
+      "epoch": 0.8226674298797939,
+      "grad_norm": 1.1281168460845947,
+      "learning_rate": 1.6237469539672479e-06,
+      "loss": 1.0212,
+      "step": 3593
+    },
+    {
+      "epoch": 0.8228963938179736,
+      "grad_norm": 1.1827605962753296,
+      "learning_rate": 1.6196971220870105e-06,
+      "loss": 0.9905,
+      "step": 3594
+    },
+    {
+      "epoch": 0.8231253577561534,
+      "grad_norm": 1.626893401145935,
+      "learning_rate": 1.6156519018171856e-06,
+      "loss": 1.0374,
+      "step": 3595
+    },
+    {
+      "epoch": 0.8233543216943331,
+      "grad_norm": 1.1740373373031616,
+      "learning_rate": 1.6116112953838247e-06,
+      "loss": 1.0092,
+      "step": 3596
+    },
+    {
+      "epoch": 0.8235832856325129,
+      "grad_norm": 1.1388554573059082,
+      "learning_rate": 1.6075753050104426e-06,
+      "loss": 1.0141,
+      "step": 3597
+    },
+    {
+      "epoch": 0.8238122495706927,
+      "grad_norm": 1.3180172443389893,
+      "learning_rate": 1.6035439329180025e-06,
+      "loss": 1.0853,
+      "step": 3598
+    },
+    {
+      "epoch": 0.8240412135088724,
+      "grad_norm": 1.3404852151870728,
+      "learning_rate": 1.5995171813249433e-06,
+      "loss": 1.0734,
+      "step": 3599
+    },
+    {
+      "epoch": 0.8242701774470521,
+      "grad_norm": 1.2810182571411133,
+      "learning_rate": 1.5954950524471513e-06,
+      "loss": 1.0471,
+      "step": 3600
+    },
+    {
+      "epoch": 0.8244991413852318,
+      "grad_norm": 1.1049981117248535,
+      "learning_rate": 1.591477548497966e-06,
+      "loss": 1.0101,
+      "step": 3601
+    },
+    {
+      "epoch": 0.8247281053234116,
+      "grad_norm": 1.3329854011535645,
+      "learning_rate": 1.587464671688187e-06,
+      "loss": 1.058,
+      "step": 3602
+    },
+    {
+      "epoch": 0.8249570692615913,
+      "grad_norm": 1.1104480028152466,
+      "learning_rate": 1.583456424226073e-06,
+      "loss": 1.0087,
+      "step": 3603
+    },
+    {
+      "epoch": 0.825186033199771,
+      "grad_norm": 1.2844616174697876,
+      "learning_rate": 1.5794528083173223e-06,
+      "loss": 1.0039,
+      "step": 3604
+    },
+    {
+      "epoch": 0.8254149971379507,
+      "grad_norm": 1.207993984222412,
+      "learning_rate": 1.575453826165093e-06,
+      "loss": 1.0599,
+      "step": 3605
+    },
+    {
+      "epoch": 0.8256439610761305,
+      "grad_norm": 1.4137972593307495,
+      "learning_rate": 1.5714594799699912e-06,
+      "loss": 1.0705,
+      "step": 3606
+    },
+    {
+      "epoch": 0.8258729250143102,
+      "grad_norm": 1.3582916259765625,
+      "learning_rate": 1.5674697719300735e-06,
+      "loss": 1.0247,
+      "step": 3607
+    },
+    {
+      "epoch": 0.82610188895249,
+      "grad_norm": 1.240177869796753,
+      "learning_rate": 1.5634847042408408e-06,
+      "loss": 1.0207,
+      "step": 3608
+    },
+    {
+      "epoch": 0.8263308528906698,
+      "grad_norm": 1.2895662784576416,
+      "learning_rate": 1.5595042790952442e-06,
+      "loss": 0.9647,
+      "step": 3609
+    },
+    {
+      "epoch": 0.8265598168288495,
+      "grad_norm": 1.8447550535202026,
+      "learning_rate": 1.5555284986836782e-06,
+      "loss": 1.0683,
+      "step": 3610
+    },
+    {
+      "epoch": 0.8267887807670292,
+      "grad_norm": 1.2016741037368774,
+      "learning_rate": 1.551557365193983e-06,
+      "loss": 1.0238,
+      "step": 3611
+    },
+    {
+      "epoch": 0.827017744705209,
+      "grad_norm": 1.8298025131225586,
+      "learning_rate": 1.5475908808114325e-06,
+      "loss": 1.0688,
+      "step": 3612
+    },
+    {
+      "epoch": 0.8272467086433887,
+      "grad_norm": 1.6220577955245972,
+      "learning_rate": 1.5436290477187589e-06,
+      "loss": 1.0334,
+      "step": 3613
+    },
+    {
+      "epoch": 0.8274756725815684,
+      "grad_norm": 1.505828619003296,
+      "learning_rate": 1.539671868096123e-06,
+      "loss": 1.0641,
+      "step": 3614
+    },
+    {
+      "epoch": 0.8277046365197481,
+      "grad_norm": 1.131373643875122,
+      "learning_rate": 1.535719344121125e-06,
+      "loss": 1.0156,
+      "step": 3615
+    },
+    {
+      "epoch": 0.8279336004579279,
+      "grad_norm": 1.5130774974822998,
+      "learning_rate": 1.5317714779688076e-06,
+      "loss": 1.104,
+      "step": 3616
+    },
+    {
+      "epoch": 0.8281625643961076,
+      "grad_norm": 1.2293245792388916,
+      "learning_rate": 1.5278282718116477e-06,
+      "loss": 1.0824,
+      "step": 3617
+    },
+    {
+      "epoch": 0.8283915283342873,
+      "grad_norm": 1.3308606147766113,
+      "learning_rate": 1.5238897278195597e-06,
+      "loss": 1.0208,
+      "step": 3618
+    },
+    {
+      "epoch": 0.828620492272467,
+      "grad_norm": 1.5616285800933838,
+      "learning_rate": 1.5199558481598908e-06,
+      "loss": 1.0327,
+      "step": 3619
+    },
+    {
+      "epoch": 0.8288494562106469,
+      "grad_norm": 1.1585637331008911,
+      "learning_rate": 1.5160266349974207e-06,
+      "loss": 0.9957,
+      "step": 3620
+    },
+    {
+      "epoch": 0.8290784201488266,
+      "grad_norm": 1.2559548616409302,
+      "learning_rate": 1.5121020904943651e-06,
+      "loss": 1.0655,
+      "step": 3621
+    },
+    {
+      "epoch": 0.8293073840870063,
+      "grad_norm": 1.3206498622894287,
+      "learning_rate": 1.5081822168103654e-06,
+      "loss": 1.0465,
+      "step": 3622
+    },
+    {
+      "epoch": 0.829536348025186,
+      "grad_norm": 1.3651444911956787,
+      "learning_rate": 1.5042670161024975e-06,
+      "loss": 1.0738,
+      "step": 3623
+    },
+    {
+      "epoch": 0.8297653119633658,
+      "grad_norm": 1.3507568836212158,
+      "learning_rate": 1.500356490525261e-06,
+      "loss": 1.0927,
+      "step": 3624
+    },
+    {
+      "epoch": 0.8299942759015455,
+      "grad_norm": 1.1345727443695068,
+      "learning_rate": 1.4964506422305902e-06,
+      "loss": 1.0417,
+      "step": 3625
+    },
+    {
+      "epoch": 0.8302232398397252,
+      "grad_norm": 1.2418663501739502,
+      "learning_rate": 1.4925494733678324e-06,
+      "loss": 1.0467,
+      "step": 3626
+    },
+    {
+      "epoch": 0.830452203777905,
+      "grad_norm": 1.6180311441421509,
+      "learning_rate": 1.4886529860837772e-06,
+      "loss": 1.0142,
+      "step": 3627
+    },
+    {
+      "epoch": 0.8306811677160847,
+      "grad_norm": 1.9630390405654907,
+      "learning_rate": 1.4847611825226227e-06,
+      "loss": 1.1022,
+      "step": 3628
+    },
+    {
+      "epoch": 0.8309101316542644,
+      "grad_norm": 1.488043189048767,
+      "learning_rate": 1.4808740648259967e-06,
+      "loss": 1.0521,
+      "step": 3629
+    },
+    {
+      "epoch": 0.8311390955924441,
+      "grad_norm": 1.2796021699905396,
+      "learning_rate": 1.4769916351329495e-06,
+      "loss": 1.0249,
+      "step": 3630
+    },
+    {
+      "epoch": 0.831368059530624,
+      "grad_norm": 1.1457350254058838,
+      "learning_rate": 1.4731138955799474e-06,
+      "loss": 1.1018,
+      "step": 3631
+    },
+    {
+      "epoch": 0.8315970234688037,
+      "grad_norm": 1.1872814893722534,
+      "learning_rate": 1.4692408483008803e-06,
+      "loss": 0.9932,
+      "step": 3632
+    },
+    {
+      "epoch": 0.8318259874069834,
+      "grad_norm": 1.2328518629074097,
+      "learning_rate": 1.465372495427052e-06,
+      "loss": 0.9768,
+      "step": 3633
+    },
+    {
+      "epoch": 0.8320549513451632,
+      "grad_norm": 1.285885214805603,
+      "learning_rate": 1.4615088390871846e-06,
+      "loss": 1.0506,
+      "step": 3634
+    },
+    {
+      "epoch": 0.8322839152833429,
+      "grad_norm": 1.0765000581741333,
+      "learning_rate": 1.457649881407417e-06,
+      "loss": 1.084,
+      "step": 3635
+    },
+    {
+      "epoch": 0.8325128792215226,
+      "grad_norm": 1.3982456922531128,
+      "learning_rate": 1.4537956245113006e-06,
+      "loss": 0.9873,
+      "step": 3636
+    },
+    {
+      "epoch": 0.8327418431597023,
+      "grad_norm": 1.536367654800415,
+      "learning_rate": 1.4499460705198e-06,
+      "loss": 1.0396,
+      "step": 3637
+    },
+    {
+      "epoch": 0.8329708070978821,
+      "grad_norm": 1.5141808986663818,
+      "learning_rate": 1.446101221551296e-06,
+      "loss": 1.0472,
+      "step": 3638
+    },
+    {
+      "epoch": 0.8331997710360618,
+      "grad_norm": 1.16829252243042,
+      "learning_rate": 1.4422610797215707e-06,
+      "loss": 1.0189,
+      "step": 3639
+    },
+    {
+      "epoch": 0.8334287349742415,
+      "grad_norm": 1.6448733806610107,
+      "learning_rate": 1.4384256471438241e-06,
+      "loss": 1.0127,
+      "step": 3640
+    },
+    {
+      "epoch": 0.8336576989124213,
+      "grad_norm": 1.396384358406067,
+      "learning_rate": 1.4345949259286673e-06,
+      "loss": 1.0597,
+      "step": 3641
+    },
+    {
+      "epoch": 0.833886662850601,
+      "grad_norm": 1.3862338066101074,
+      "learning_rate": 1.4307689181841077e-06,
+      "loss": 1.0525,
+      "step": 3642
+    },
+    {
+      "epoch": 0.8341156267887808,
+      "grad_norm": 1.409687876701355,
+      "learning_rate": 1.4269476260155668e-06,
+      "loss": 1.0834,
+      "step": 3643
+    },
+    {
+      "epoch": 0.8343445907269605,
+      "grad_norm": 1.310608983039856,
+      "learning_rate": 1.4231310515258745e-06,
+      "loss": 1.0362,
+      "step": 3644
+    },
+    {
+      "epoch": 0.8345735546651403,
+      "grad_norm": 1.2230523824691772,
+      "learning_rate": 1.4193191968152543e-06,
+      "loss": 0.9915,
+      "step": 3645
+    },
+    {
+      "epoch": 0.83480251860332,
+      "grad_norm": 1.3876858949661255,
+      "learning_rate": 1.4155120639813392e-06,
+      "loss": 0.9453,
+      "step": 3646
+    },
+    {
+      "epoch": 0.8350314825414997,
+      "grad_norm": 1.223950743675232,
+      "learning_rate": 1.4117096551191633e-06,
+      "loss": 1.011,
+      "step": 3647
+    },
+    {
+      "epoch": 0.8352604464796795,
+      "grad_norm": 1.2806096076965332,
+      "learning_rate": 1.4079119723211599e-06,
+      "loss": 1.0136,
+      "step": 3648
+    },
+    {
+      "epoch": 0.8354894104178592,
+      "grad_norm": 1.2026371955871582,
+      "learning_rate": 1.4041190176771635e-06,
+      "loss": 0.9929,
+      "step": 3649
+    },
+    {
+      "epoch": 0.8357183743560389,
+      "grad_norm": 1.4160505533218384,
+      "learning_rate": 1.4003307932744003e-06,
+      "loss": 1.0733,
+      "step": 3650
+    },
+    {
+      "epoch": 0.8359473382942186,
+      "grad_norm": 1.2385631799697876,
+      "learning_rate": 1.396547301197504e-06,
+      "loss": 1.0789,
+      "step": 3651
+    },
+    {
+      "epoch": 0.8361763022323984,
+      "grad_norm": 1.875428557395935,
+      "learning_rate": 1.3927685435284977e-06,
+      "loss": 1.0522,
+      "step": 3652
+    },
+    {
+      "epoch": 0.8364052661705781,
+      "grad_norm": 1.4530951976776123,
+      "learning_rate": 1.388994522346796e-06,
+      "loss": 1.0203,
+      "step": 3653
+    },
+    {
+      "epoch": 0.8366342301087579,
+      "grad_norm": 1.4752883911132812,
+      "learning_rate": 1.3852252397292143e-06,
+      "loss": 0.9734,
+      "step": 3654
+    },
+    {
+      "epoch": 0.8368631940469377,
+      "grad_norm": 1.1157587766647339,
+      "learning_rate": 1.38146069774996e-06,
+      "loss": 0.9831,
+      "step": 3655
+    },
+    {
+      "epoch": 0.8370921579851174,
+      "grad_norm": 1.351131558418274,
+      "learning_rate": 1.377700898480624e-06,
+      "loss": 1.0292,
+      "step": 3656
+    },
+    {
+      "epoch": 0.8373211219232971,
+      "grad_norm": 1.5354865789413452,
+      "learning_rate": 1.373945843990192e-06,
+      "loss": 0.9554,
+      "step": 3657
+    },
+    {
+      "epoch": 0.8375500858614768,
+      "grad_norm": 1.1356061697006226,
+      "learning_rate": 1.3701955363450447e-06,
+      "loss": 1.0667,
+      "step": 3658
+    },
+    {
+      "epoch": 0.8377790497996566,
+      "grad_norm": 1.2410647869110107,
+      "learning_rate": 1.3664499776089401e-06,
+      "loss": 1.0254,
+      "step": 3659
+    },
+    {
+      "epoch": 0.8380080137378363,
+      "grad_norm": 1.2936856746673584,
+      "learning_rate": 1.3627091698430284e-06,
+      "loss": 0.9986,
+      "step": 3660
+    },
+    {
+      "epoch": 0.838236977676016,
+      "grad_norm": 1.256300687789917,
+      "learning_rate": 1.3589731151058461e-06,
+      "loss": 1.0607,
+      "step": 3661
+    },
+    {
+      "epoch": 0.8384659416141957,
+      "grad_norm": 1.4257123470306396,
+      "learning_rate": 1.355241815453312e-06,
+      "loss": 1.0198,
+      "step": 3662
+    },
+    {
+      "epoch": 0.8386949055523755,
+      "grad_norm": 1.3414442539215088,
+      "learning_rate": 1.3515152729387315e-06,
+      "loss": 1.045,
+      "step": 3663
+    },
+    {
+      "epoch": 0.8389238694905552,
+      "grad_norm": 1.2866592407226562,
+      "learning_rate": 1.347793489612782e-06,
+      "loss": 0.9788,
+      "step": 3664
+    },
+    {
+      "epoch": 0.8391528334287349,
+      "grad_norm": 2.8504953384399414,
+      "learning_rate": 1.3440764675235384e-06,
+      "loss": 0.9914,
+      "step": 3665
+    },
+    {
+      "epoch": 0.8393817973669148,
+      "grad_norm": 1.2402368783950806,
+      "learning_rate": 1.3403642087164447e-06,
+      "loss": 1.0267,
+      "step": 3666
+    },
+    {
+      "epoch": 0.8396107613050945,
+      "grad_norm": 1.9054492712020874,
+      "learning_rate": 1.33665671523432e-06,
+      "loss": 1.0689,
+      "step": 3667
+    },
+    {
+      "epoch": 0.8398397252432742,
+      "grad_norm": 1.5320415496826172,
+      "learning_rate": 1.332953989117377e-06,
+      "loss": 1.025,
+      "step": 3668
+    },
+    {
+      "epoch": 0.8400686891814539,
+      "grad_norm": 1.2526971101760864,
+      "learning_rate": 1.3292560324031867e-06,
+      "loss": 1.0578,
+      "step": 3669
+    },
+    {
+      "epoch": 0.8402976531196337,
+      "grad_norm": 1.3369795083999634,
+      "learning_rate": 1.3255628471267056e-06,
+      "loss": 1.0097,
+      "step": 3670
+    },
+    {
+      "epoch": 0.8405266170578134,
+      "grad_norm": 1.3616974353790283,
+      "learning_rate": 1.321874435320264e-06,
+      "loss": 1.0137,
+      "step": 3671
+    },
+    {
+      "epoch": 0.8407555809959931,
+      "grad_norm": 1.156179428100586,
+      "learning_rate": 1.3181907990135624e-06,
+      "loss": 1.0683,
+      "step": 3672
+    },
+    {
+      "epoch": 0.8409845449341729,
+      "grad_norm": 1.5307315587997437,
+      "learning_rate": 1.3145119402336758e-06,
+      "loss": 1.0354,
+      "step": 3673
+    },
+    {
+      "epoch": 0.8412135088723526,
+      "grad_norm": 1.272509217262268,
+      "learning_rate": 1.3108378610050498e-06,
+      "loss": 1.068,
+      "step": 3674
+    },
+    {
+      "epoch": 0.8414424728105323,
+      "grad_norm": 1.1754751205444336,
+      "learning_rate": 1.307168563349499e-06,
+      "loss": 1.0448,
+      "step": 3675
+    },
+    {
+      "epoch": 0.841671436748712,
+      "grad_norm": 1.483634352684021,
+      "learning_rate": 1.3035040492862061e-06,
+      "loss": 1.0705,
+      "step": 3676
+    },
+    {
+      "epoch": 0.8419004006868919,
+      "grad_norm": 1.1576098203659058,
+      "learning_rate": 1.2998443208317246e-06,
+      "loss": 1.1061,
+      "step": 3677
+    },
+    {
+      "epoch": 0.8421293646250716,
+      "grad_norm": 1.2855072021484375,
+      "learning_rate": 1.2961893799999703e-06,
+      "loss": 1.0096,
+      "step": 3678
+    },
+    {
+      "epoch": 0.8423583285632513,
+      "grad_norm": 1.4712800979614258,
+      "learning_rate": 1.2925392288022299e-06,
+      "loss": 0.9488,
+      "step": 3679
+    },
+    {
+      "epoch": 0.842587292501431,
+      "grad_norm": 1.539665937423706,
+      "learning_rate": 1.2888938692471464e-06,
+      "loss": 1.0338,
+      "step": 3680
+    },
+    {
+      "epoch": 0.8428162564396108,
+      "grad_norm": 1.1315950155258179,
+      "learning_rate": 1.28525330334073e-06,
+      "loss": 1.0255,
+      "step": 3681
+    },
+    {
+      "epoch": 0.8430452203777905,
+      "grad_norm": 1.3212192058563232,
+      "learning_rate": 1.2816175330863613e-06,
+      "loss": 0.9955,
+      "step": 3682
+    },
+    {
+      "epoch": 0.8432741843159702,
+      "grad_norm": 1.3658416271209717,
+      "learning_rate": 1.2779865604847674e-06,
+      "loss": 1.0136,
+      "step": 3683
+    },
+    {
+      "epoch": 0.84350314825415,
+      "grad_norm": 1.8101247549057007,
+      "learning_rate": 1.2743603875340426e-06,
+      "loss": 1.0044,
+      "step": 3684
+    },
+    {
+      "epoch": 0.8437321121923297,
+      "grad_norm": 1.4984769821166992,
+      "learning_rate": 1.270739016229642e-06,
+      "loss": 1.0216,
+      "step": 3685
+    },
+    {
+      "epoch": 0.8439610761305094,
+      "grad_norm": 1.2391111850738525,
+      "learning_rate": 1.267122448564374e-06,
+      "loss": 1.0306,
+      "step": 3686
+    },
+    {
+      "epoch": 0.8441900400686891,
+      "grad_norm": 1.167498230934143,
+      "learning_rate": 1.2635106865284063e-06,
+      "loss": 0.9658,
+      "step": 3687
+    },
+    {
+      "epoch": 0.8444190040068689,
+      "grad_norm": 1.7382392883300781,
+      "learning_rate": 1.259903732109261e-06,
+      "loss": 0.9892,
+      "step": 3688
+    },
+    {
+      "epoch": 0.8446479679450487,
+      "grad_norm": 1.2146246433258057,
+      "learning_rate": 1.256301587291815e-06,
+      "loss": 1.0496,
+      "step": 3689
+    },
+    {
+      "epoch": 0.8448769318832284,
+      "grad_norm": 1.7516138553619385,
+      "learning_rate": 1.2527042540583e-06,
+      "loss": 1.0987,
+      "step": 3690
+    },
+    {
+      "epoch": 0.8451058958214082,
+      "grad_norm": 1.5586003065109253,
+      "learning_rate": 1.249111734388292e-06,
+      "loss": 1.0333,
+      "step": 3691
+    },
+    {
+      "epoch": 0.8453348597595879,
+      "grad_norm": 1.4014748334884644,
+      "learning_rate": 1.2455240302587325e-06,
+      "loss": 1.0038,
+      "step": 3692
+    },
+    {
+      "epoch": 0.8455638236977676,
+      "grad_norm": 1.3031091690063477,
+      "learning_rate": 1.2419411436439021e-06,
+      "loss": 1.0135,
+      "step": 3693
+    },
+    {
+      "epoch": 0.8457927876359473,
+      "grad_norm": 1.1367318630218506,
+      "learning_rate": 1.238363076515432e-06,
+      "loss": 1.0154,
+      "step": 3694
+    },
+    {
+      "epoch": 0.8460217515741271,
+      "grad_norm": 1.4096239805221558,
+      "learning_rate": 1.2347898308423012e-06,
+      "loss": 1.0172,
+      "step": 3695
+    },
+    {
+      "epoch": 0.8462507155123068,
+      "grad_norm": 1.2547544240951538,
+      "learning_rate": 1.2312214085908424e-06,
+      "loss": 0.9901,
+      "step": 3696
+    },
+    {
+      "epoch": 0.8464796794504865,
+      "grad_norm": 1.4089115858078003,
+      "learning_rate": 1.227657811724723e-06,
+      "loss": 1.0461,
+      "step": 3697
+    },
+    {
+      "epoch": 0.8467086433886662,
+      "grad_norm": 1.6553548574447632,
+      "learning_rate": 1.2240990422049625e-06,
+      "loss": 1.0419,
+      "step": 3698
+    },
+    {
+      "epoch": 0.846937607326846,
+      "grad_norm": 1.4773091077804565,
+      "learning_rate": 1.2205451019899217e-06,
+      "loss": 1.0423,
+      "step": 3699
+    },
+    {
+      "epoch": 0.8471665712650257,
+      "grad_norm": 1.2673395872116089,
+      "learning_rate": 1.2169959930353049e-06,
+      "loss": 1.0627,
+      "step": 3700
+    },
+    {
+      "epoch": 0.8473955352032055,
+      "grad_norm": 1.941409707069397,
+      "learning_rate": 1.2134517172941563e-06,
+      "loss": 0.9947,
+      "step": 3701
+    },
+    {
+      "epoch": 0.8476244991413853,
+      "grad_norm": 2.137348175048828,
+      "learning_rate": 1.2099122767168602e-06,
+      "loss": 1.009,
+      "step": 3702
+    },
+    {
+      "epoch": 0.847853463079565,
+      "grad_norm": 1.4081718921661377,
+      "learning_rate": 1.2063776732511434e-06,
+      "loss": 0.9911,
+      "step": 3703
+    },
+    {
+      "epoch": 0.8480824270177447,
+      "grad_norm": 1.2987362146377563,
+      "learning_rate": 1.2028479088420686e-06,
+      "loss": 1.0329,
+      "step": 3704
+    },
+    {
+      "epoch": 0.8483113909559244,
+      "grad_norm": 1.554642677307129,
+      "learning_rate": 1.19932298543203e-06,
+      "loss": 1.0258,
+      "step": 3705
+    },
+    {
+      "epoch": 0.8485403548941042,
+      "grad_norm": 1.3535888195037842,
+      "learning_rate": 1.19580290496077e-06,
+      "loss": 1.0652,
+      "step": 3706
+    },
+    {
+      "epoch": 0.8487693188322839,
+      "grad_norm": 1.258093237876892,
+      "learning_rate": 1.1922876693653584e-06,
+      "loss": 1.0323,
+      "step": 3707
+    },
+    {
+      "epoch": 0.8489982827704636,
+      "grad_norm": 1.1866981983184814,
+      "learning_rate": 1.1887772805801967e-06,
+      "loss": 1.0644,
+      "step": 3708
+    },
+    {
+      "epoch": 0.8492272467086434,
+      "grad_norm": 1.5237160921096802,
+      "learning_rate": 1.1852717405370228e-06,
+      "loss": 0.9952,
+      "step": 3709
+    },
+    {
+      "epoch": 0.8494562106468231,
+      "grad_norm": 1.3863621950149536,
+      "learning_rate": 1.1817710511649105e-06,
+      "loss": 0.9737,
+      "step": 3710
+    },
+    {
+      "epoch": 0.8496851745850028,
+      "grad_norm": 1.380892038345337,
+      "learning_rate": 1.1782752143902553e-06,
+      "loss": 0.9958,
+      "step": 3711
+    },
+    {
+      "epoch": 0.8499141385231826,
+      "grad_norm": 1.3540898561477661,
+      "learning_rate": 1.1747842321367886e-06,
+      "loss": 1.0295,
+      "step": 3712
+    },
+    {
+      "epoch": 0.8501431024613624,
+      "grad_norm": 1.2603505849838257,
+      "learning_rate": 1.171298106325568e-06,
+      "loss": 1.0327,
+      "step": 3713
+    },
+    {
+      "epoch": 0.8503720663995421,
+      "grad_norm": 1.7046961784362793,
+      "learning_rate": 1.1678168388749788e-06,
+      "loss": 1.0873,
+      "step": 3714
+    },
+    {
+      "epoch": 0.8506010303377218,
+      "grad_norm": 1.4969935417175293,
+      "learning_rate": 1.1643404317007345e-06,
+      "loss": 1.0087,
+      "step": 3715
+    },
+    {
+      "epoch": 0.8508299942759016,
+      "grad_norm": 1.0810832977294922,
+      "learning_rate": 1.1608688867158724e-06,
+      "loss": 0.9596,
+      "step": 3716
+    },
+    {
+      "epoch": 0.8510589582140813,
+      "grad_norm": 1.7231873273849487,
+      "learning_rate": 1.1574022058307555e-06,
+      "loss": 1.0446,
+      "step": 3717
+    },
+    {
+      "epoch": 0.851287922152261,
+      "grad_norm": 1.4240765571594238,
+      "learning_rate": 1.1539403909530688e-06,
+      "loss": 1.0087,
+      "step": 3718
+    },
+    {
+      "epoch": 0.8515168860904407,
+      "grad_norm": 1.1195824146270752,
+      "learning_rate": 1.1504834439878166e-06,
+      "loss": 1.006,
+      "step": 3719
+    },
+    {
+      "epoch": 0.8517458500286205,
+      "grad_norm": 1.3876497745513916,
+      "learning_rate": 1.1470313668373324e-06,
+      "loss": 1.0934,
+      "step": 3720
+    },
+    {
+      "epoch": 0.8519748139668002,
+      "grad_norm": 1.3672479391098022,
+      "learning_rate": 1.1435841614012666e-06,
+      "loss": 1.0914,
+      "step": 3721
+    },
+    {
+      "epoch": 0.8522037779049799,
+      "grad_norm": 1.1950000524520874,
+      "learning_rate": 1.140141829576582e-06,
+      "loss": 1.0035,
+      "step": 3722
+    },
+    {
+      "epoch": 0.8524327418431596,
+      "grad_norm": 1.6234666109085083,
+      "learning_rate": 1.1367043732575666e-06,
+      "loss": 1.0398,
+      "step": 3723
+    },
+    {
+      "epoch": 0.8526617057813395,
+      "grad_norm": 1.2520289421081543,
+      "learning_rate": 1.1332717943358263e-06,
+      "loss": 0.9795,
+      "step": 3724
+    },
+    {
+      "epoch": 0.8528906697195192,
+      "grad_norm": 1.3402252197265625,
+      "learning_rate": 1.1298440947002775e-06,
+      "loss": 1.0576,
+      "step": 3725
+    },
+    {
+      "epoch": 0.8531196336576989,
+      "grad_norm": 1.3803917169570923,
+      "learning_rate": 1.1264212762371563e-06,
+      "loss": 1.0388,
+      "step": 3726
+    },
+    {
+      "epoch": 0.8533485975958787,
+      "grad_norm": 1.1965564489364624,
+      "learning_rate": 1.1230033408300111e-06,
+      "loss": 1.0416,
+      "step": 3727
+    },
+    {
+      "epoch": 0.8535775615340584,
+      "grad_norm": 1.4472531080245972,
+      "learning_rate": 1.1195902903597023e-06,
+      "loss": 1.0545,
+      "step": 3728
+    },
+    {
+      "epoch": 0.8538065254722381,
+      "grad_norm": 2.9633491039276123,
+      "learning_rate": 1.1161821267044038e-06,
+      "loss": 1.0308,
+      "step": 3729
+    },
+    {
+      "epoch": 0.8540354894104178,
+      "grad_norm": 3.640618324279785,
+      "learning_rate": 1.1127788517395987e-06,
+      "loss": 0.9939,
+      "step": 3730
+    },
+    {
+      "epoch": 0.8542644533485976,
+      "grad_norm": 1.3619310855865479,
+      "learning_rate": 1.1093804673380804e-06,
+      "loss": 1.0465,
+      "step": 3731
+    },
+    {
+      "epoch": 0.8544934172867773,
+      "grad_norm": 1.3596311807632446,
+      "learning_rate": 1.1059869753699547e-06,
+      "loss": 1.0179,
+      "step": 3732
+    },
+    {
+      "epoch": 0.854722381224957,
+      "grad_norm": 1.5578163862228394,
+      "learning_rate": 1.102598377702625e-06,
+      "loss": 0.9975,
+      "step": 3733
+    },
+    {
+      "epoch": 0.8549513451631368,
+      "grad_norm": 1.3997864723205566,
+      "learning_rate": 1.099214676200816e-06,
+      "loss": 1.023,
+      "step": 3734
+    },
+    {
+      "epoch": 0.8551803091013166,
+      "grad_norm": 1.4430853128433228,
+      "learning_rate": 1.0958358727265438e-06,
+      "loss": 1.0851,
+      "step": 3735
+    },
+    {
+      "epoch": 0.8554092730394963,
+      "grad_norm": 1.0692206621170044,
+      "learning_rate": 1.092461969139137e-06,
+      "loss": 1.0646,
+      "step": 3736
+    },
+    {
+      "epoch": 0.855638236977676,
+      "grad_norm": 1.4968456029891968,
+      "learning_rate": 1.089092967295231e-06,
+      "loss": 1.0341,
+      "step": 3737
+    },
+    {
+      "epoch": 0.8558672009158558,
+      "grad_norm": 1.3507084846496582,
+      "learning_rate": 1.0857288690487555e-06,
+      "loss": 1.0051,
+      "step": 3738
+    },
+    {
+      "epoch": 0.8560961648540355,
+      "grad_norm": 1.3483790159225464,
+      "learning_rate": 1.082369676250945e-06,
+      "loss": 1.0424,
+      "step": 3739
+    },
+    {
+      "epoch": 0.8563251287922152,
+      "grad_norm": 1.4069323539733887,
+      "learning_rate": 1.0790153907503364e-06,
+      "loss": 1.0422,
+      "step": 3740
+    },
+    {
+      "epoch": 0.856554092730395,
+      "grad_norm": 1.629030704498291,
+      "learning_rate": 1.0756660143927644e-06,
+      "loss": 1.0405,
+      "step": 3741
+    },
+    {
+      "epoch": 0.8567830566685747,
+      "grad_norm": 1.409432291984558,
+      "learning_rate": 1.0723215490213635e-06,
+      "loss": 1.0972,
+      "step": 3742
+    },
+    {
+      "epoch": 0.8570120206067544,
+      "grad_norm": 1.2074227333068848,
+      "learning_rate": 1.0689819964765646e-06,
+      "loss": 1.0264,
+      "step": 3743
+    },
+    {
+      "epoch": 0.8572409845449341,
+      "grad_norm": 1.379439115524292,
+      "learning_rate": 1.0656473585960946e-06,
+      "loss": 0.9715,
+      "step": 3744
+    },
+    {
+      "epoch": 0.8574699484831139,
+      "grad_norm": 1.0981643199920654,
+      "learning_rate": 1.0623176372149802e-06,
+      "loss": 1.0236,
+      "step": 3745
+    },
+    {
+      "epoch": 0.8576989124212936,
+      "grad_norm": 1.303014874458313,
+      "learning_rate": 1.0589928341655342e-06,
+      "loss": 1.0092,
+      "step": 3746
+    },
+    {
+      "epoch": 0.8579278763594734,
+      "grad_norm": 1.1583458185195923,
+      "learning_rate": 1.0556729512773679e-06,
+      "loss": 1.0262,
+      "step": 3747
+    },
+    {
+      "epoch": 0.8581568402976532,
+      "grad_norm": 1.2592904567718506,
+      "learning_rate": 1.0523579903773917e-06,
+      "loss": 1.0341,
+      "step": 3748
+    },
+    {
+      "epoch": 0.8583858042358329,
+      "grad_norm": 1.2319550514221191,
+      "learning_rate": 1.0490479532897946e-06,
+      "loss": 1.056,
+      "step": 3749
+    },
+    {
+      "epoch": 0.8586147681740126,
+      "grad_norm": 1.1954131126403809,
+      "learning_rate": 1.0457428418360616e-06,
+      "loss": 1.0415,
+      "step": 3750
+    },
+    {
+      "epoch": 0.8588437321121923,
+      "grad_norm": 1.6190708875656128,
+      "learning_rate": 1.0424426578349733e-06,
+      "loss": 1.0591,
+      "step": 3751
+    },
+    {
+      "epoch": 0.8590726960503721,
+      "grad_norm": 1.1923547983169556,
+      "learning_rate": 1.0391474031025895e-06,
+      "loss": 1.0615,
+      "step": 3752
+    },
+    {
+      "epoch": 0.8593016599885518,
+      "grad_norm": 1.4358948469161987,
+      "learning_rate": 1.0358570794522615e-06,
+      "loss": 0.9408,
+      "step": 3753
+    },
+    {
+      "epoch": 0.8595306239267315,
+      "grad_norm": 1.3710299730300903,
+      "learning_rate": 1.0325716886946268e-06,
+      "loss": 1.0515,
+      "step": 3754
+    },
+    {
+      "epoch": 0.8597595878649112,
+      "grad_norm": 1.4297199249267578,
+      "learning_rate": 1.0292912326376091e-06,
+      "loss": 0.9476,
+      "step": 3755
+    },
+    {
+      "epoch": 0.859988551803091,
+      "grad_norm": 1.2614961862564087,
+      "learning_rate": 1.0260157130864178e-06,
+      "loss": 1.1242,
+      "step": 3756
+    },
+    {
+      "epoch": 0.8602175157412707,
+      "grad_norm": 1.7146847248077393,
+      "learning_rate": 1.0227451318435378e-06,
+      "loss": 1.0826,
+      "step": 3757
+    },
+    {
+      "epoch": 0.8604464796794505,
+      "grad_norm": 1.4658128023147583,
+      "learning_rate": 1.019479490708748e-06,
+      "loss": 1.0325,
+      "step": 3758
+    },
+    {
+      "epoch": 0.8606754436176303,
+      "grad_norm": 1.323593258857727,
+      "learning_rate": 1.0162187914791045e-06,
+      "loss": 1.0018,
+      "step": 3759
+    },
+    {
+      "epoch": 0.86090440755581,
+      "grad_norm": 1.5773340463638306,
+      "learning_rate": 1.0129630359489352e-06,
+      "loss": 0.941,
+      "step": 3760
+    },
+    {
+      "epoch": 0.8611333714939897,
+      "grad_norm": 1.2028287649154663,
+      "learning_rate": 1.0097122259098625e-06,
+      "loss": 0.9942,
+      "step": 3761
+    },
+    {
+      "epoch": 0.8613623354321694,
+      "grad_norm": 1.2747315168380737,
+      "learning_rate": 1.0064663631507787e-06,
+      "loss": 1.0371,
+      "step": 3762
+    },
+    {
+      "epoch": 0.8615912993703492,
+      "grad_norm": 1.2724723815917969,
+      "learning_rate": 1.0032254494578519e-06,
+      "loss": 1.0712,
+      "step": 3763
+    },
+    {
+      "epoch": 0.8618202633085289,
+      "grad_norm": 1.3098310232162476,
+      "learning_rate": 9.999894866145287e-07,
+      "loss": 1.0679,
+      "step": 3764
+    },
+    {
+      "epoch": 0.8620492272467086,
+      "grad_norm": 1.2499535083770752,
+      "learning_rate": 9.967584764015392e-07,
+      "loss": 1.0339,
+      "step": 3765
+    },
+    {
+      "epoch": 0.8622781911848884,
+      "grad_norm": 1.4696881771087646,
+      "learning_rate": 9.935324205968744e-07,
+      "loss": 1.0509,
+      "step": 3766
+    },
+    {
+      "epoch": 0.8625071551230681,
+      "grad_norm": 1.3763833045959473,
+      "learning_rate": 9.903113209758098e-07,
+      "loss": 1.0162,
+      "step": 3767
+    },
+    {
+      "epoch": 0.8627361190612478,
+      "grad_norm": 1.3132450580596924,
+      "learning_rate": 9.870951793108863e-07,
+      "loss": 1.031,
+      "step": 3768
+    },
+    {
+      "epoch": 0.8629650829994275,
+      "grad_norm": 1.5283464193344116,
+      "learning_rate": 9.838839973719227e-07,
+      "loss": 1.0361,
+      "step": 3769
+    },
+    {
+      "epoch": 0.8631940469376074,
+      "grad_norm": 1.428132176399231,
+      "learning_rate": 9.806777769260034e-07,
+      "loss": 1.0075,
+      "step": 3770
+    },
+    {
+      "epoch": 0.8634230108757871,
+      "grad_norm": 1.229541540145874,
+      "learning_rate": 9.774765197374847e-07,
+      "loss": 1.0586,
+      "step": 3771
+    },
+    {
+      "epoch": 0.8636519748139668,
+      "grad_norm": 1.8992414474487305,
+      "learning_rate": 9.742802275679941e-07,
+      "loss": 1.0236,
+      "step": 3772
+    },
+    {
+      "epoch": 0.8638809387521466,
+      "grad_norm": 1.2058573961257935,
+      "learning_rate": 9.710889021764235e-07,
+      "loss": 0.9526,
+      "step": 3773
+    },
+    {
+      "epoch": 0.8641099026903263,
+      "grad_norm": 1.2776075601577759,
+      "learning_rate": 9.679025453189273e-07,
+      "loss": 1.0135,
+      "step": 3774
+    },
+    {
+      "epoch": 0.864338866628506,
+      "grad_norm": 1.5076816082000732,
+      "learning_rate": 9.64721158748938e-07,
+      "loss": 1.035,
+      "step": 3775
+    },
+    {
+      "epoch": 0.8645678305666857,
+      "grad_norm": 3.022805690765381,
+      "learning_rate": 9.615447442171444e-07,
+      "loss": 1.0242,
+      "step": 3776
+    },
+    {
+      "epoch": 0.8647967945048655,
+      "grad_norm": 1.037968635559082,
+      "learning_rate": 9.583733034714982e-07,
+      "loss": 1.0302,
+      "step": 3777
+    },
+    {
+      "epoch": 0.8650257584430452,
+      "grad_norm": 1.6912962198257446,
+      "learning_rate": 9.55206838257219e-07,
+      "loss": 0.9981,
+      "step": 3778
+    },
+    {
+      "epoch": 0.8652547223812249,
+      "grad_norm": 1.1713199615478516,
+      "learning_rate": 9.520453503167837e-07,
+      "loss": 0.9741,
+      "step": 3779
+    },
+    {
+      "epoch": 0.8654836863194046,
+      "grad_norm": 1.2378156185150146,
+      "learning_rate": 9.488888413899345e-07,
+      "loss": 1.0817,
+      "step": 3780
+    },
+    {
+      "epoch": 0.8657126502575845,
+      "grad_norm": 1.2005176544189453,
+      "learning_rate": 9.457373132136716e-07,
+      "loss": 1.0458,
+      "step": 3781
+    },
+    {
+      "epoch": 0.8659416141957642,
+      "grad_norm": 2.108745574951172,
+      "learning_rate": 9.425907675222535e-07,
+      "loss": 1.1055,
+      "step": 3782
+    },
+    {
+      "epoch": 0.8661705781339439,
+      "grad_norm": 1.631432294845581,
+      "learning_rate": 9.394492060471994e-07,
+      "loss": 1.0514,
+      "step": 3783
+    },
+    {
+      "epoch": 0.8663995420721237,
+      "grad_norm": 1.4168133735656738,
+      "learning_rate": 9.363126305172831e-07,
+      "loss": 1.0468,
+      "step": 3784
+    },
+    {
+      "epoch": 0.8666285060103034,
+      "grad_norm": 1.2942928075790405,
+      "learning_rate": 9.331810426585364e-07,
+      "loss": 1.0393,
+      "step": 3785
+    },
+    {
+      "epoch": 0.8668574699484831,
+      "grad_norm": 1.3363865613937378,
+      "learning_rate": 9.300544441942461e-07,
+      "loss": 1.0378,
+      "step": 3786
+    },
+    {
+      "epoch": 0.8670864338866628,
+      "grad_norm": 1.569215178489685,
+      "learning_rate": 9.269328368449538e-07,
+      "loss": 1.046,
+      "step": 3787
+    },
+    {
+      "epoch": 0.8673153978248426,
+      "grad_norm": 1.158187985420227,
+      "learning_rate": 9.238162223284497e-07,
+      "loss": 1.0903,
+      "step": 3788
+    },
+    {
+      "epoch": 0.8675443617630223,
+      "grad_norm": 2.184824228286743,
+      "learning_rate": 9.207046023597866e-07,
+      "loss": 1.0087,
+      "step": 3789
+    },
+    {
+      "epoch": 0.867773325701202,
+      "grad_norm": 1.4404144287109375,
+      "learning_rate": 9.175979786512589e-07,
+      "loss": 1.0473,
+      "step": 3790
+    },
+    {
+      "epoch": 0.8680022896393818,
+      "grad_norm": 1.384673833847046,
+      "learning_rate": 9.144963529124163e-07,
+      "loss": 1.0535,
+      "step": 3791
+    },
+    {
+      "epoch": 0.8682312535775615,
+      "grad_norm": 1.3582277297973633,
+      "learning_rate": 9.113997268500574e-07,
+      "loss": 1.0592,
+      "step": 3792
+    },
+    {
+      "epoch": 0.8684602175157413,
+      "grad_norm": 1.1262242794036865,
+      "learning_rate": 9.083081021682305e-07,
+      "loss": 1.0654,
+      "step": 3793
+    },
+    {
+      "epoch": 0.868689181453921,
+      "grad_norm": 1.1574608087539673,
+      "learning_rate": 9.052214805682303e-07,
+      "loss": 1.065,
+      "step": 3794
+    },
+    {
+      "epoch": 0.8689181453921008,
+      "grad_norm": 1.3190420866012573,
+      "learning_rate": 9.021398637485979e-07,
+      "loss": 0.9962,
+      "step": 3795
+    },
+    {
+      "epoch": 0.8691471093302805,
+      "grad_norm": 4.816401958465576,
+      "learning_rate": 8.990632534051224e-07,
+      "loss": 0.9806,
+      "step": 3796
+    },
+    {
+      "epoch": 0.8693760732684602,
+      "grad_norm": 1.8604696989059448,
+      "learning_rate": 8.959916512308387e-07,
+      "loss": 1.0577,
+      "step": 3797
+    },
+    {
+      "epoch": 0.86960503720664,
+      "grad_norm": 1.3026139736175537,
+      "learning_rate": 8.929250589160166e-07,
+      "loss": 1.0274,
+      "step": 3798
+    },
+    {
+      "epoch": 0.8698340011448197,
+      "grad_norm": 1.3048547506332397,
+      "learning_rate": 8.898634781481829e-07,
+      "loss": 1.0608,
+      "step": 3799
+    },
+    {
+      "epoch": 0.8700629650829994,
+      "grad_norm": 1.4611790180206299,
+      "learning_rate": 8.868069106121002e-07,
+      "loss": 1.0704,
+      "step": 3800
+    },
+    {
+      "epoch": 0.8702919290211791,
+      "grad_norm": 1.360741138458252,
+      "learning_rate": 8.837553579897673e-07,
+      "loss": 1.018,
+      "step": 3801
+    },
+    {
+      "epoch": 0.8705208929593589,
+      "grad_norm": 1.1581453084945679,
+      "learning_rate": 8.807088219604288e-07,
+      "loss": 1.0178,
+      "step": 3802
+    },
+    {
+      "epoch": 0.8707498568975386,
+      "grad_norm": 1.437386393547058,
+      "learning_rate": 8.776673042005724e-07,
+      "loss": 1.062,
+      "step": 3803
+    },
+    {
+      "epoch": 0.8709788208357184,
+      "grad_norm": 1.3421489000320435,
+      "learning_rate": 8.74630806383916e-07,
+      "loss": 0.9713,
+      "step": 3804
+    },
+    {
+      "epoch": 0.8712077847738982,
+      "grad_norm": 1.095431923866272,
+      "learning_rate": 8.715993301814174e-07,
+      "loss": 0.9333,
+      "step": 3805
+    },
+    {
+      "epoch": 0.8714367487120779,
+      "grad_norm": 1.4182848930358887,
+      "learning_rate": 8.685728772612778e-07,
+      "loss": 1.0209,
+      "step": 3806
+    },
+    {
+      "epoch": 0.8716657126502576,
+      "grad_norm": 1.2892870903015137,
+      "learning_rate": 8.655514492889249e-07,
+      "loss": 1.048,
+      "step": 3807
+    },
+    {
+      "epoch": 0.8718946765884373,
+      "grad_norm": 1.217413306236267,
+      "learning_rate": 8.625350479270255e-07,
+      "loss": 1.0019,
+      "step": 3808
+    },
+    {
+      "epoch": 0.8721236405266171,
+      "grad_norm": 1.2366745471954346,
+      "learning_rate": 8.595236748354807e-07,
+      "loss": 1.042,
+      "step": 3809
+    },
+    {
+      "epoch": 0.8723526044647968,
+      "grad_norm": 1.248071312904358,
+      "learning_rate": 8.565173316714249e-07,
+      "loss": 1.0241,
+      "step": 3810
+    },
+    {
+      "epoch": 0.8725815684029765,
+      "grad_norm": 1.2598704099655151,
+      "learning_rate": 8.535160200892234e-07,
+      "loss": 1.0653,
+      "step": 3811
+    },
+    {
+      "epoch": 0.8728105323411562,
+      "grad_norm": 1.4132360219955444,
+      "learning_rate": 8.505197417404687e-07,
+      "loss": 1.0395,
+      "step": 3812
+    },
+    {
+      "epoch": 0.873039496279336,
+      "grad_norm": 1.2241437435150146,
+      "learning_rate": 8.475284982739917e-07,
+      "loss": 1.0056,
+      "step": 3813
+    },
+    {
+      "epoch": 0.8732684602175157,
+      "grad_norm": 3.2831687927246094,
+      "learning_rate": 8.445422913358503e-07,
+      "loss": 1.047,
+      "step": 3814
+    },
+    {
+      "epoch": 0.8734974241556954,
+      "grad_norm": 1.535780668258667,
+      "learning_rate": 8.415611225693254e-07,
+      "loss": 1.0187,
+      "step": 3815
+    },
+    {
+      "epoch": 0.8737263880938753,
+      "grad_norm": 1.2877063751220703,
+      "learning_rate": 8.385849936149282e-07,
+      "loss": 0.9162,
+      "step": 3816
+    },
+    {
+      "epoch": 0.873955352032055,
+      "grad_norm": 1.2718114852905273,
+      "learning_rate": 8.356139061104029e-07,
+      "loss": 1.0094,
+      "step": 3817
+    },
+    {
+      "epoch": 0.8741843159702347,
+      "grad_norm": 1.4520769119262695,
+      "learning_rate": 8.326478616907085e-07,
+      "loss": 1.0206,
+      "step": 3818
+    },
+    {
+      "epoch": 0.8744132799084144,
+      "grad_norm": 2.025461435317993,
+      "learning_rate": 8.296868619880372e-07,
+      "loss": 1.0811,
+      "step": 3819
+    },
+    {
+      "epoch": 0.8746422438465942,
+      "grad_norm": 1.302669882774353,
+      "learning_rate": 8.26730908631802e-07,
+      "loss": 1.038,
+      "step": 3820
+    },
+    {
+      "epoch": 0.8748712077847739,
+      "grad_norm": 1.3168150186538696,
+      "learning_rate": 8.237800032486387e-07,
+      "loss": 1.0384,
+      "step": 3821
+    },
+    {
+      "epoch": 0.8751001717229536,
+      "grad_norm": 1.5246906280517578,
+      "learning_rate": 8.208341474624071e-07,
+      "loss": 1.0494,
+      "step": 3822
+    },
+    {
+      "epoch": 0.8753291356611334,
+      "grad_norm": 2.0702121257781982,
+      "learning_rate": 8.178933428941849e-07,
+      "loss": 1.0735,
+      "step": 3823
+    },
+    {
+      "epoch": 0.8755580995993131,
+      "grad_norm": 1.1960113048553467,
+      "learning_rate": 8.149575911622731e-07,
+      "loss": 1.0311,
+      "step": 3824
+    },
+    {
+      "epoch": 0.8757870635374928,
+      "grad_norm": 1.2554943561553955,
+      "learning_rate": 8.12026893882194e-07,
+      "loss": 1.0279,
+      "step": 3825
+    },
+    {
+      "epoch": 0.8760160274756725,
+      "grad_norm": 1.2481484413146973,
+      "learning_rate": 8.091012526666797e-07,
+      "loss": 1.0473,
+      "step": 3826
+    },
+    {
+      "epoch": 0.8762449914138524,
+      "grad_norm": 1.5274428129196167,
+      "learning_rate": 8.061806691256913e-07,
+      "loss": 1.0147,
+      "step": 3827
+    },
+    {
+      "epoch": 0.8764739553520321,
+      "grad_norm": 1.132452130317688,
+      "learning_rate": 8.032651448664009e-07,
+      "loss": 1.0577,
+      "step": 3828
+    },
+    {
+      "epoch": 0.8767029192902118,
+      "grad_norm": 1.3138935565948486,
+      "learning_rate": 8.003546814931917e-07,
+      "loss": 1.0043,
+      "step": 3829
+    },
+    {
+      "epoch": 0.8769318832283916,
+      "grad_norm": 1.1806259155273438,
+      "learning_rate": 7.974492806076762e-07,
+      "loss": 1.0525,
+      "step": 3830
+    },
+    {
+      "epoch": 0.8771608471665713,
+      "grad_norm": 1.3222166299819946,
+      "learning_rate": 7.945489438086651e-07,
+      "loss": 1.092,
+      "step": 3831
+    },
+    {
+      "epoch": 0.877389811104751,
+      "grad_norm": 1.3098158836364746,
+      "learning_rate": 7.91653672692192e-07,
+      "loss": 1.0704,
+      "step": 3832
+    },
+    {
+      "epoch": 0.8776187750429307,
+      "grad_norm": 1.5300372838974,
+      "learning_rate": 7.887634688515e-07,
+      "loss": 1.0192,
+      "step": 3833
+    },
+    {
+      "epoch": 0.8778477389811105,
+      "grad_norm": 1.5282765626907349,
+      "learning_rate": 7.858783338770437e-07,
+      "loss": 1.0689,
+      "step": 3834
+    },
+    {
+      "epoch": 0.8780767029192902,
+      "grad_norm": 1.0936659574508667,
+      "learning_rate": 7.829982693564886e-07,
+      "loss": 1.0448,
+      "step": 3835
+    },
+    {
+      "epoch": 0.8783056668574699,
+      "grad_norm": 1.1930863857269287,
+      "learning_rate": 7.801232768747113e-07,
+      "loss": 0.9976,
+      "step": 3836
+    },
+    {
+      "epoch": 0.8785346307956496,
+      "grad_norm": 1.5860801935195923,
+      "learning_rate": 7.772533580137942e-07,
+      "loss": 0.9733,
+      "step": 3837
+    },
+    {
+      "epoch": 0.8787635947338294,
+      "grad_norm": 1.3473610877990723,
+      "learning_rate": 7.743885143530317e-07,
+      "loss": 1.0056,
+      "step": 3838
+    },
+    {
+      "epoch": 0.8789925586720092,
+      "grad_norm": 1.2198975086212158,
+      "learning_rate": 7.715287474689237e-07,
+      "loss": 1.0589,
+      "step": 3839
+    },
+    {
+      "epoch": 0.8792215226101889,
+      "grad_norm": 2.675382614135742,
+      "learning_rate": 7.686740589351704e-07,
+      "loss": 1.021,
+      "step": 3840
+    },
+    {
+      "epoch": 0.8794504865483687,
+      "grad_norm": 1.468649983406067,
+      "learning_rate": 7.658244503226909e-07,
+      "loss": 1.0378,
+      "step": 3841
+    },
+    {
+      "epoch": 0.8796794504865484,
+      "grad_norm": 1.561955213546753,
+      "learning_rate": 7.629799231995961e-07,
+      "loss": 1.0173,
+      "step": 3842
+    },
+    {
+      "epoch": 0.8799084144247281,
+      "grad_norm": 1.2610421180725098,
+      "learning_rate": 7.601404791312028e-07,
+      "loss": 1.0568,
+      "step": 3843
+    },
+    {
+      "epoch": 0.8801373783629078,
+      "grad_norm": 1.5121142864227295,
+      "learning_rate": 7.573061196800414e-07,
+      "loss": 0.9959,
+      "step": 3844
+    },
+    {
+      "epoch": 0.8803663423010876,
+      "grad_norm": 1.1338526010513306,
+      "learning_rate": 7.544768464058294e-07,
+      "loss": 0.9543,
+      "step": 3845
+    },
+    {
+      "epoch": 0.8805953062392673,
+      "grad_norm": 1.4656891822814941,
+      "learning_rate": 7.51652660865495e-07,
+      "loss": 1.0345,
+      "step": 3846
+    },
+    {
+      "epoch": 0.880824270177447,
+      "grad_norm": 1.067690372467041,
+      "learning_rate": 7.488335646131628e-07,
+      "loss": 1.001,
+      "step": 3847
+    },
+    {
+      "epoch": 0.8810532341156267,
+      "grad_norm": 1.3866279125213623,
+      "learning_rate": 7.460195592001585e-07,
+      "loss": 0.9599,
+      "step": 3848
+    },
+    {
+      "epoch": 0.8812821980538065,
+      "grad_norm": 1.3089559078216553,
+      "learning_rate": 7.432106461750044e-07,
+      "loss": 1.0318,
+      "step": 3849
+    },
+    {
+      "epoch": 0.8815111619919863,
+      "grad_norm": 1.1162192821502686,
+      "learning_rate": 7.40406827083423e-07,
+      "loss": 1.0282,
+      "step": 3850
+    },
+    {
+      "epoch": 0.881740125930166,
+      "grad_norm": 1.7417439222335815,
+      "learning_rate": 7.376081034683325e-07,
+      "loss": 1.0748,
+      "step": 3851
+    },
+    {
+      "epoch": 0.8819690898683458,
+      "grad_norm": 1.5468460321426392,
+      "learning_rate": 7.348144768698485e-07,
+      "loss": 1.0649,
+      "step": 3852
+    },
+    {
+      "epoch": 0.8821980538065255,
+      "grad_norm": 1.8217270374298096,
+      "learning_rate": 7.320259488252757e-07,
+      "loss": 1.0416,
+      "step": 3853
+    },
+    {
+      "epoch": 0.8824270177447052,
+      "grad_norm": 1.378803014755249,
+      "learning_rate": 7.292425208691212e-07,
+      "loss": 1.0032,
+      "step": 3854
+    },
+    {
+      "epoch": 0.882655981682885,
+      "grad_norm": 1.5163785219192505,
+      "learning_rate": 7.26464194533083e-07,
+      "loss": 1.0449,
+      "step": 3855
+    },
+    {
+      "epoch": 0.8828849456210647,
+      "grad_norm": 1.268101692199707,
+      "learning_rate": 7.236909713460482e-07,
+      "loss": 1.0759,
+      "step": 3856
+    },
+    {
+      "epoch": 0.8831139095592444,
+      "grad_norm": 1.5246165990829468,
+      "learning_rate": 7.209228528340972e-07,
+      "loss": 1.0565,
+      "step": 3857
+    },
+    {
+      "epoch": 0.8833428734974241,
+      "grad_norm": 1.3627903461456299,
+      "learning_rate": 7.181598405205082e-07,
+      "loss": 1.0487,
+      "step": 3858
+    },
+    {
+      "epoch": 0.8835718374356039,
+      "grad_norm": 1.2992089986801147,
+      "learning_rate": 7.154019359257381e-07,
+      "loss": 1.0759,
+      "step": 3859
+    },
+    {
+      "epoch": 0.8838008013737836,
+      "grad_norm": 1.353250503540039,
+      "learning_rate": 7.126491405674407e-07,
+      "loss": 1.0197,
+      "step": 3860
+    },
+    {
+      "epoch": 0.8840297653119633,
+      "grad_norm": 1.206063985824585,
+      "learning_rate": 7.099014559604556e-07,
+      "loss": 1.0341,
+      "step": 3861
+    },
+    {
+      "epoch": 0.8842587292501431,
+      "grad_norm": 1.488411784172058,
+      "learning_rate": 7.071588836168109e-07,
+      "loss": 0.992,
+      "step": 3862
+    },
+    {
+      "epoch": 0.8844876931883229,
+      "grad_norm": 1.6317309141159058,
+      "learning_rate": 7.044214250457216e-07,
+      "loss": 1.0674,
+      "step": 3863
+    },
+    {
+      "epoch": 0.8847166571265026,
+      "grad_norm": 4.539884090423584,
+      "learning_rate": 7.016890817535881e-07,
+      "loss": 1.0251,
+      "step": 3864
+    },
+    {
+      "epoch": 0.8849456210646823,
+      "grad_norm": 1.4975309371948242,
+      "learning_rate": 6.989618552439958e-07,
+      "loss": 1.0532,
+      "step": 3865
+    },
+    {
+      "epoch": 0.8851745850028621,
+      "grad_norm": 1.3167531490325928,
+      "learning_rate": 6.962397470177162e-07,
+      "loss": 1.0001,
+      "step": 3866
+    },
+    {
+      "epoch": 0.8854035489410418,
+      "grad_norm": 1.4427505731582642,
+      "learning_rate": 6.935227585726978e-07,
+      "loss": 1.1204,
+      "step": 3867
+    },
+    {
+      "epoch": 0.8856325128792215,
+      "grad_norm": 1.2467824220657349,
+      "learning_rate": 6.908108914040823e-07,
+      "loss": 0.9937,
+      "step": 3868
+    },
+    {
+      "epoch": 0.8858614768174012,
+      "grad_norm": 1.390540361404419,
+      "learning_rate": 6.881041470041849e-07,
+      "loss": 0.9756,
+      "step": 3869
+    },
+    {
+      "epoch": 0.886090440755581,
+      "grad_norm": 1.3013540506362915,
+      "learning_rate": 6.854025268625042e-07,
+      "loss": 1.0061,
+      "step": 3870
+    },
+    {
+      "epoch": 0.8863194046937607,
+      "grad_norm": 1.362518548965454,
+      "learning_rate": 6.827060324657164e-07,
+      "loss": 1.0281,
+      "step": 3871
+    },
+    {
+      "epoch": 0.8865483686319404,
+      "grad_norm": 1.2985994815826416,
+      "learning_rate": 6.800146652976869e-07,
+      "loss": 1.049,
+      "step": 3872
+    },
+    {
+      "epoch": 0.8867773325701203,
+      "grad_norm": 1.0657109022140503,
+      "learning_rate": 6.773284268394464e-07,
+      "loss": 1.0451,
+      "step": 3873
+    },
+    {
+      "epoch": 0.8870062965083,
+      "grad_norm": 1.4140193462371826,
+      "learning_rate": 6.746473185692115e-07,
+      "loss": 0.9702,
+      "step": 3874
+    },
+    {
+      "epoch": 0.8872352604464797,
+      "grad_norm": 1.1937177181243896,
+      "learning_rate": 6.71971341962373e-07,
+      "loss": 1.0247,
+      "step": 3875
+    },
+    {
+      "epoch": 0.8874642243846594,
+      "grad_norm": 1.1080400943756104,
+      "learning_rate": 6.693004984914997e-07,
+      "loss": 1.039,
+      "step": 3876
+    },
+    {
+      "epoch": 0.8876931883228392,
+      "grad_norm": 1.7587668895721436,
+      "learning_rate": 6.666347896263326e-07,
+      "loss": 1.0278,
+      "step": 3877
+    },
+    {
+      "epoch": 0.8879221522610189,
+      "grad_norm": 1.3932582139968872,
+      "learning_rate": 6.639742168337904e-07,
+      "loss": 1.027,
+      "step": 3878
+    },
+    {
+      "epoch": 0.8881511161991986,
+      "grad_norm": 1.1661114692687988,
+      "learning_rate": 6.61318781577962e-07,
+      "loss": 1.0093,
+      "step": 3879
+    },
+    {
+      "epoch": 0.8883800801373783,
+      "grad_norm": 1.9590517282485962,
+      "learning_rate": 6.586684853201153e-07,
+      "loss": 1.0055,
+      "step": 3880
+    },
+    {
+      "epoch": 0.8886090440755581,
+      "grad_norm": 1.1533890962600708,
+      "learning_rate": 6.560233295186802e-07,
+      "loss": 1.0544,
+      "step": 3881
+    },
+    {
+      "epoch": 0.8888380080137378,
+      "grad_norm": 1.3643183708190918,
+      "learning_rate": 6.53383315629268e-07,
+      "loss": 1.0934,
+      "step": 3882
+    },
+    {
+      "epoch": 0.8890669719519175,
+      "grad_norm": 1.3094185590744019,
+      "learning_rate": 6.507484451046575e-07,
+      "loss": 1.0203,
+      "step": 3883
+    },
+    {
+      "epoch": 0.8892959358900973,
+      "grad_norm": 1.2927873134613037,
+      "learning_rate": 6.481187193947913e-07,
+      "loss": 1.0636,
+      "step": 3884
+    },
+    {
+      "epoch": 0.8895248998282771,
+      "grad_norm": 1.2240914106369019,
+      "learning_rate": 6.454941399467873e-07,
+      "loss": 1.0583,
+      "step": 3885
+    },
+    {
+      "epoch": 0.8897538637664568,
+      "grad_norm": 1.765049934387207,
+      "learning_rate": 6.428747082049313e-07,
+      "loss": 1.0131,
+      "step": 3886
+    },
+    {
+      "epoch": 0.8899828277046365,
+      "grad_norm": 1.369654893875122,
+      "learning_rate": 6.402604256106715e-07,
+      "loss": 1.0498,
+      "step": 3887
+    },
+    {
+      "epoch": 0.8902117916428163,
+      "grad_norm": 1.3459523916244507,
+      "learning_rate": 6.37651293602628e-07,
+      "loss": 0.9947,
+      "step": 3888
+    },
+    {
+      "epoch": 0.890440755580996,
+      "grad_norm": 1.3815542459487915,
+      "learning_rate": 6.350473136165836e-07,
+      "loss": 0.9885,
+      "step": 3889
+    },
+    {
+      "epoch": 0.8906697195191757,
+      "grad_norm": 1.2464871406555176,
+      "learning_rate": 6.324484870854874e-07,
+      "loss": 1.0349,
+      "step": 3890
+    },
+    {
+      "epoch": 0.8908986834573555,
+      "grad_norm": 1.1414371728897095,
+      "learning_rate": 6.298548154394501e-07,
+      "loss": 0.9818,
+      "step": 3891
+    },
+    {
+      "epoch": 0.8911276473955352,
+      "grad_norm": 1.2838653326034546,
+      "learning_rate": 6.2726630010575e-07,
+      "loss": 1.0227,
+      "step": 3892
+    },
+    {
+      "epoch": 0.8913566113337149,
+      "grad_norm": 1.3109924793243408,
+      "learning_rate": 6.246829425088229e-07,
+      "loss": 1.0274,
+      "step": 3893
+    },
+    {
+      "epoch": 0.8915855752718946,
+      "grad_norm": 1.4073375463485718,
+      "learning_rate": 6.22104744070271e-07,
+      "loss": 1.0466,
+      "step": 3894
+    },
+    {
+      "epoch": 0.8918145392100744,
+      "grad_norm": 1.345703125,
+      "learning_rate": 6.195317062088513e-07,
+      "loss": 1.0208,
+      "step": 3895
+    },
+    {
+      "epoch": 0.8920435031482542,
+      "grad_norm": 1.3727253675460815,
+      "learning_rate": 6.169638303404912e-07,
+      "loss": 1.0646,
+      "step": 3896
+    },
+    {
+      "epoch": 0.8922724670864339,
+      "grad_norm": 1.132949709892273,
+      "learning_rate": 6.14401117878266e-07,
+      "loss": 0.9631,
+      "step": 3897
+    },
+    {
+      "epoch": 0.8925014310246137,
+      "grad_norm": 1.202333688735962,
+      "learning_rate": 6.118435702324166e-07,
+      "loss": 0.9988,
+      "step": 3898
+    },
+    {
+      "epoch": 0.8927303949627934,
+      "grad_norm": 1.380020022392273,
+      "learning_rate": 6.092911888103404e-07,
+      "loss": 1.0655,
+      "step": 3899
+    },
+    {
+      "epoch": 0.8929593589009731,
+      "grad_norm": 1.3880457878112793,
+      "learning_rate": 6.067439750165916e-07,
+      "loss": 1.004,
+      "step": 3900
+    },
+    {
+      "epoch": 0.8931883228391528,
+      "grad_norm": 1.0868477821350098,
+      "learning_rate": 6.042019302528801e-07,
+      "loss": 0.9799,
+      "step": 3901
+    },
+    {
+      "epoch": 0.8934172867773326,
+      "grad_norm": 1.2509794235229492,
+      "learning_rate": 6.016650559180715e-07,
+      "loss": 1.0058,
+      "step": 3902
+    },
+    {
+      "epoch": 0.8936462507155123,
+      "grad_norm": 1.0911234617233276,
+      "learning_rate": 5.991333534081878e-07,
+      "loss": 1.0277,
+      "step": 3903
+    },
+    {
+      "epoch": 0.893875214653692,
+      "grad_norm": 1.3828800916671753,
+      "learning_rate": 5.966068241164025e-07,
+      "loss": 1.0678,
+      "step": 3904
+    },
+    {
+      "epoch": 0.8941041785918717,
+      "grad_norm": 1.1840590238571167,
+      "learning_rate": 5.940854694330433e-07,
+      "loss": 0.9748,
+      "step": 3905
+    },
+    {
+      "epoch": 0.8943331425300515,
+      "grad_norm": 1.447941541671753,
+      "learning_rate": 5.915692907455905e-07,
+      "loss": 0.9746,
+      "step": 3906
+    },
+    {
+      "epoch": 0.8945621064682312,
+      "grad_norm": 2.082156181335449,
+      "learning_rate": 5.890582894386798e-07,
+      "loss": 1.0355,
+      "step": 3907
+    },
+    {
+      "epoch": 0.894791070406411,
+      "grad_norm": 1.1880125999450684,
+      "learning_rate": 5.86552466894088e-07,
+      "loss": 1.0326,
+      "step": 3908
+    },
+    {
+      "epoch": 0.8950200343445908,
+      "grad_norm": 1.3864256143569946,
+      "learning_rate": 5.840518244907512e-07,
+      "loss": 0.9975,
+      "step": 3909
+    },
+    {
+      "epoch": 0.8952489982827705,
+      "grad_norm": 1.368095874786377,
+      "learning_rate": 5.815563636047539e-07,
+      "loss": 0.9745,
+      "step": 3910
+    },
+    {
+      "epoch": 0.8954779622209502,
+      "grad_norm": 1.2576544284820557,
+      "learning_rate": 5.790660856093245e-07,
+      "loss": 1.0147,
+      "step": 3911
+    },
+    {
+      "epoch": 0.89570692615913,
+      "grad_norm": 1.2153538465499878,
+      "learning_rate": 5.76580991874841e-07,
+      "loss": 1.0201,
+      "step": 3912
+    },
+    {
+      "epoch": 0.8959358900973097,
+      "grad_norm": 1.7341593503952026,
+      "learning_rate": 5.741010837688354e-07,
+      "loss": 0.9843,
+      "step": 3913
+    },
+    {
+      "epoch": 0.8961648540354894,
+      "grad_norm": 1.151334285736084,
+      "learning_rate": 5.716263626559748e-07,
+      "loss": 0.9381,
+      "step": 3914
+    },
+    {
+      "epoch": 0.8963938179736691,
+      "grad_norm": 1.2881344556808472,
+      "learning_rate": 5.691568298980799e-07,
+      "loss": 0.9995,
+      "step": 3915
+    },
+    {
+      "epoch": 0.8966227819118489,
+      "grad_norm": 1.0738816261291504,
+      "learning_rate": 5.666924868541124e-07,
+      "loss": 0.9905,
+      "step": 3916
+    },
+    {
+      "epoch": 0.8968517458500286,
+      "grad_norm": 1.1380635499954224,
+      "learning_rate": 5.64233334880181e-07,
+      "loss": 1.0218,
+      "step": 3917
+    },
+    {
+      "epoch": 0.8970807097882083,
+      "grad_norm": 1.9478411674499512,
+      "learning_rate": 5.617793753295364e-07,
+      "loss": 1.0623,
+      "step": 3918
+    },
+    {
+      "epoch": 0.8973096737263881,
+      "grad_norm": 3.0823843479156494,
+      "learning_rate": 5.593306095525697e-07,
+      "loss": 1.0533,
+      "step": 3919
+    },
+    {
+      "epoch": 0.8975386376645679,
+      "grad_norm": 1.3105992078781128,
+      "learning_rate": 5.568870388968184e-07,
+      "loss": 1.0027,
+      "step": 3920
+    },
+    {
+      "epoch": 0.8977676016027476,
+      "grad_norm": 1.4492088556289673,
+      "learning_rate": 5.544486647069614e-07,
+      "loss": 0.9894,
+      "step": 3921
+    },
+    {
+      "epoch": 0.8979965655409273,
+      "grad_norm": 1.8873605728149414,
+      "learning_rate": 5.520154883248086e-07,
+      "loss": 1.0517,
+      "step": 3922
+    },
+    {
+      "epoch": 0.898225529479107,
+      "grad_norm": 1.1515185832977295,
+      "learning_rate": 5.495875110893223e-07,
+      "loss": 1.0118,
+      "step": 3923
+    },
+    {
+      "epoch": 0.8984544934172868,
+      "grad_norm": 1.1645504236221313,
+      "learning_rate": 5.471647343365982e-07,
+      "loss": 0.9882,
+      "step": 3924
+    },
+    {
+      "epoch": 0.8986834573554665,
+      "grad_norm": 1.2297289371490479,
+      "learning_rate": 5.447471593998666e-07,
+      "loss": 0.9947,
+      "step": 3925
+    },
+    {
+      "epoch": 0.8989124212936462,
+      "grad_norm": 1.699040412902832,
+      "learning_rate": 5.423347876094998e-07,
+      "loss": 0.9676,
+      "step": 3926
+    },
+    {
+      "epoch": 0.899141385231826,
+      "grad_norm": 1.485076665878296,
+      "learning_rate": 5.399276202930071e-07,
+      "loss": 1.0582,
+      "step": 3927
+    },
+    {
+      "epoch": 0.8993703491700057,
+      "grad_norm": 2.078399419784546,
+      "learning_rate": 5.375256587750311e-07,
+      "loss": 1.0306,
+      "step": 3928
+    },
+    {
+      "epoch": 0.8995993131081854,
+      "grad_norm": 1.3613694906234741,
+      "learning_rate": 5.351289043773511e-07,
+      "loss": 1.0463,
+      "step": 3929
+    },
+    {
+      "epoch": 0.8998282770463651,
+      "grad_norm": 1.1392529010772705,
+      "learning_rate": 5.327373584188822e-07,
+      "loss": 1.0335,
+      "step": 3930
+    },
+    {
+      "epoch": 0.900057240984545,
+      "grad_norm": 1.9740110635757446,
+      "learning_rate": 5.303510222156716e-07,
+      "loss": 1.0021,
+      "step": 3931
+    },
+    {
+      "epoch": 0.9002862049227247,
+      "grad_norm": 1.254390835762024,
+      "learning_rate": 5.279698970809011e-07,
+      "loss": 1.0522,
+      "step": 3932
+    },
+    {
+      "epoch": 0.9005151688609044,
+      "grad_norm": 1.3042449951171875,
+      "learning_rate": 5.255939843248792e-07,
+      "loss": 1.0814,
+      "step": 3933
+    },
+    {
+      "epoch": 0.9007441327990842,
+      "grad_norm": 1.3471471071243286,
+      "learning_rate": 5.232232852550568e-07,
+      "loss": 1.0309,
+      "step": 3934
+    },
+    {
+      "epoch": 0.9009730967372639,
+      "grad_norm": 1.9291504621505737,
+      "learning_rate": 5.208578011760079e-07,
+      "loss": 1.0439,
+      "step": 3935
+    },
+    {
+      "epoch": 0.9012020606754436,
+      "grad_norm": 1.2869789600372314,
+      "learning_rate": 5.184975333894349e-07,
+      "loss": 1.1011,
+      "step": 3936
+    },
+    {
+      "epoch": 0.9014310246136233,
+      "grad_norm": 1.4123051166534424,
+      "learning_rate": 5.161424831941797e-07,
+      "loss": 1.0704,
+      "step": 3937
+    },
+    {
+      "epoch": 0.9016599885518031,
+      "grad_norm": 1.2700828313827515,
+      "learning_rate": 5.137926518862013e-07,
+      "loss": 1.053,
+      "step": 3938
+    },
+    {
+      "epoch": 0.9018889524899828,
+      "grad_norm": 1.2948448657989502,
+      "learning_rate": 5.114480407585942e-07,
+      "loss": 1.0921,
+      "step": 3939
+    },
+    {
+      "epoch": 0.9021179164281625,
+      "grad_norm": 1.1770613193511963,
+      "learning_rate": 5.091086511015786e-07,
+      "loss": 1.0008,
+      "step": 3940
+    },
+    {
+      "epoch": 0.9023468803663423,
+      "grad_norm": 1.1652601957321167,
+      "learning_rate": 5.067744842025002e-07,
+      "loss": 1.0086,
+      "step": 3941
+    },
+    {
+      "epoch": 0.9025758443045221,
+      "grad_norm": 1.2411365509033203,
+      "learning_rate": 5.044455413458327e-07,
+      "loss": 1.0527,
+      "step": 3942
+    },
+    {
+      "epoch": 0.9028048082427018,
+      "grad_norm": 1.4037675857543945,
+      "learning_rate": 5.02121823813172e-07,
+      "loss": 1.0761,
+      "step": 3943
+    },
+    {
+      "epoch": 0.9030337721808815,
+      "grad_norm": 1.2515357732772827,
+      "learning_rate": 4.998033328832419e-07,
+      "loss": 0.9981,
+      "step": 3944
+    },
+    {
+      "epoch": 0.9032627361190613,
+      "grad_norm": 1.2006301879882812,
+      "learning_rate": 4.974900698318885e-07,
+      "loss": 0.983,
+      "step": 3945
+    },
+    {
+      "epoch": 0.903491700057241,
+      "grad_norm": 1.3678110837936401,
+      "learning_rate": 4.951820359320802e-07,
+      "loss": 1.0476,
+      "step": 3946
+    },
+    {
+      "epoch": 0.9037206639954207,
+      "grad_norm": 1.4648295640945435,
+      "learning_rate": 4.928792324539089e-07,
+      "loss": 0.9633,
+      "step": 3947
+    },
+    {
+      "epoch": 0.9039496279336005,
+      "grad_norm": 1.2273361682891846,
+      "learning_rate": 4.905816606645896e-07,
+      "loss": 1.038,
+      "step": 3948
+    },
+    {
+      "epoch": 0.9041785918717802,
+      "grad_norm": 1.4598244428634644,
+      "learning_rate": 4.882893218284546e-07,
+      "loss": 1.0626,
+      "step": 3949
+    },
+    {
+      "epoch": 0.9044075558099599,
+      "grad_norm": 1.1587828397750854,
+      "learning_rate": 4.860022172069579e-07,
+      "loss": 1.0123,
+      "step": 3950
+    },
+    {
+      "epoch": 0.9046365197481396,
+      "grad_norm": 1.4061542749404907,
+      "learning_rate": 4.837203480586782e-07,
+      "loss": 0.9873,
+      "step": 3951
+    },
+    {
+      "epoch": 0.9048654836863194,
+      "grad_norm": 1.5064465999603271,
+      "learning_rate": 4.814437156393048e-07,
+      "loss": 1.0661,
+      "step": 3952
+    },
+    {
+      "epoch": 0.9050944476244991,
+      "grad_norm": 1.5875314474105835,
+      "learning_rate": 4.791723212016508e-07,
+      "loss": 0.9952,
+      "step": 3953
+    },
+    {
+      "epoch": 0.9053234115626789,
+      "grad_norm": 1.572794795036316,
+      "learning_rate": 4.769061659956464e-07,
+      "loss": 1.0625,
+      "step": 3954
+    },
+    {
+      "epoch": 0.9055523755008587,
+      "grad_norm": 1.485347032546997,
+      "learning_rate": 4.746452512683375e-07,
+      "loss": 0.9975,
+      "step": 3955
+    },
+    {
+      "epoch": 0.9057813394390384,
+      "grad_norm": 1.1230580806732178,
+      "learning_rate": 4.7238957826388656e-07,
+      "loss": 1.0865,
+      "step": 3956
+    },
+    {
+      "epoch": 0.9060103033772181,
+      "grad_norm": 1.2839785814285278,
+      "learning_rate": 4.7013914822356956e-07,
+      "loss": 1.0588,
+      "step": 3957
+    },
+    {
+      "epoch": 0.9062392673153978,
+      "grad_norm": 1.29798424243927,
+      "learning_rate": 4.6789396238578255e-07,
+      "loss": 1.0536,
+      "step": 3958
+    },
+    {
+      "epoch": 0.9064682312535776,
+      "grad_norm": 1.4684815406799316,
+      "learning_rate": 4.656540219860317e-07,
+      "loss": 1.0095,
+      "step": 3959
+    },
+    {
+      "epoch": 0.9066971951917573,
+      "grad_norm": 3.565049409866333,
+      "learning_rate": 4.6341932825693326e-07,
+      "loss": 1.0161,
+      "step": 3960
+    },
+    {
+      "epoch": 0.906926159129937,
+      "grad_norm": 1.6227142810821533,
+      "learning_rate": 4.611898824282257e-07,
+      "loss": 1.0105,
+      "step": 3961
+    },
+    {
+      "epoch": 0.9071551230681167,
+      "grad_norm": 1.3278828859329224,
+      "learning_rate": 4.5896568572675327e-07,
+      "loss": 0.9983,
+      "step": 3962
+    },
+    {
+      "epoch": 0.9073840870062965,
+      "grad_norm": 1.2562534809112549,
+      "learning_rate": 4.567467393764702e-07,
+      "loss": 1.0176,
+      "step": 3963
+    },
+    {
+      "epoch": 0.9076130509444762,
+      "grad_norm": 1.5097713470458984,
+      "learning_rate": 4.5453304459844525e-07,
+      "loss": 1.0569,
+      "step": 3964
+    },
+    {
+      "epoch": 0.907842014882656,
+      "grad_norm": 1.313685655593872,
+      "learning_rate": 4.5232460261085966e-07,
+      "loss": 1.0136,
+      "step": 3965
+    },
+    {
+      "epoch": 0.9080709788208358,
+      "grad_norm": 1.2489078044891357,
+      "learning_rate": 4.501214146289956e-07,
+      "loss": 1.0279,
+      "step": 3966
+    },
+    {
+      "epoch": 0.9082999427590155,
+      "grad_norm": 1.3103182315826416,
+      "learning_rate": 4.4792348186525e-07,
+      "loss": 1.0426,
+      "step": 3967
+    },
+    {
+      "epoch": 0.9085289066971952,
+      "grad_norm": 1.1325465440750122,
+      "learning_rate": 4.4573080552912874e-07,
+      "loss": 0.9801,
+      "step": 3968
+    },
+    {
+      "epoch": 0.9087578706353749,
+      "grad_norm": 1.951433539390564,
+      "learning_rate": 4.435433868272421e-07,
+      "loss": 0.9769,
+      "step": 3969
+    },
+    {
+      "epoch": 0.9089868345735547,
+      "grad_norm": 1.1956815719604492,
+      "learning_rate": 4.413612269633083e-07,
+      "loss": 1.0412,
+      "step": 3970
+    },
+    {
+      "epoch": 0.9092157985117344,
+      "grad_norm": 1.2107232809066772,
+      "learning_rate": 4.391843271381513e-07,
+      "loss": 0.9577,
+      "step": 3971
+    },
+    {
+      "epoch": 0.9094447624499141,
+      "grad_norm": 1.252339482307434,
+      "learning_rate": 4.370126885497017e-07,
+      "loss": 1.0642,
+      "step": 3972
+    },
+    {
+      "epoch": 0.9096737263880939,
+      "grad_norm": 1.6101003885269165,
+      "learning_rate": 4.3484631239299356e-07,
+      "loss": 0.9944,
+      "step": 3973
+    },
+    {
+      "epoch": 0.9099026903262736,
+      "grad_norm": 1.0703437328338623,
+      "learning_rate": 4.3268519986016444e-07,
+      "loss": 1.0407,
+      "step": 3974
+    },
+    {
+      "epoch": 0.9101316542644533,
+      "grad_norm": 1.2428948879241943,
+      "learning_rate": 4.3052935214045745e-07,
+      "loss": 1.0025,
+      "step": 3975
+    },
+    {
+      "epoch": 0.910360618202633,
+      "grad_norm": 1.3559765815734863,
+      "learning_rate": 4.2837877042021915e-07,
+      "loss": 1.0398,
+      "step": 3976
+    },
+    {
+      "epoch": 0.9105895821408129,
+      "grad_norm": 1.2042170763015747,
+      "learning_rate": 4.26233455882894e-07,
+      "loss": 0.9451,
+      "step": 3977
+    },
+    {
+      "epoch": 0.9108185460789926,
+      "grad_norm": 1.347455620765686,
+      "learning_rate": 4.240934097090299e-07,
+      "loss": 0.9805,
+      "step": 3978
+    },
+    {
+      "epoch": 0.9110475100171723,
+      "grad_norm": 2.1769297122955322,
+      "learning_rate": 4.219586330762815e-07,
+      "loss": 1.0007,
+      "step": 3979
+    },
+    {
+      "epoch": 0.911276473955352,
+      "grad_norm": 1.392369270324707,
+      "learning_rate": 4.198291271593924e-07,
+      "loss": 1.0211,
+      "step": 3980
+    },
+    {
+      "epoch": 0.9115054378935318,
+      "grad_norm": 1.1754943132400513,
+      "learning_rate": 4.177048931302152e-07,
+      "loss": 1.008,
+      "step": 3981
+    },
+    {
+      "epoch": 0.9117344018317115,
+      "grad_norm": 1.4287794828414917,
+      "learning_rate": 4.1558593215769714e-07,
+      "loss": 1.0554,
+      "step": 3982
+    },
+    {
+      "epoch": 0.9119633657698912,
+      "grad_norm": 1.3779202699661255,
+      "learning_rate": 4.134722454078843e-07,
+      "loss": 1.0852,
+      "step": 3983
+    },
+    {
+      "epoch": 0.912192329708071,
+      "grad_norm": 1.3808211088180542,
+      "learning_rate": 4.1136383404392187e-07,
+      "loss": 1.0345,
+      "step": 3984
+    },
+    {
+      "epoch": 0.9124212936462507,
+      "grad_norm": 1.3731261491775513,
+      "learning_rate": 4.092606992260506e-07,
+      "loss": 1.0508,
+      "step": 3985
+    },
+    {
+      "epoch": 0.9126502575844304,
+      "grad_norm": 1.3647085428237915,
+      "learning_rate": 4.0716284211160807e-07,
+      "loss": 1.0739,
+      "step": 3986
+    },
+    {
+      "epoch": 0.9128792215226101,
+      "grad_norm": 1.5696361064910889,
+      "learning_rate": 4.0507026385502747e-07,
+      "loss": 1.0594,
+      "step": 3987
+    },
+    {
+      "epoch": 0.91310818546079,
+      "grad_norm": 1.408075213432312,
+      "learning_rate": 4.029829656078343e-07,
+      "loss": 1.0911,
+      "step": 3988
+    },
+    {
+      "epoch": 0.9133371493989697,
+      "grad_norm": 1.2420895099639893,
+      "learning_rate": 4.0090094851865547e-07,
+      "loss": 0.9449,
+      "step": 3989
+    },
+    {
+      "epoch": 0.9135661133371494,
+      "grad_norm": 1.551822543144226,
+      "learning_rate": 3.9882421373320655e-07,
+      "loss": 0.9516,
+      "step": 3990
+    },
+    {
+      "epoch": 0.9137950772753292,
+      "grad_norm": 2.8030037879943848,
+      "learning_rate": 3.967527623942957e-07,
+      "loss": 1.0588,
+      "step": 3991
+    },
+    {
+      "epoch": 0.9140240412135089,
+      "grad_norm": 1.239384651184082,
+      "learning_rate": 3.9468659564182554e-07,
+      "loss": 1.0079,
+      "step": 3992
+    },
+    {
+      "epoch": 0.9142530051516886,
+      "grad_norm": 1.3504375219345093,
+      "learning_rate": 3.9262571461279097e-07,
+      "loss": 0.9982,
+      "step": 3993
+    },
+    {
+      "epoch": 0.9144819690898683,
+      "grad_norm": 1.2237106561660767,
+      "learning_rate": 3.9057012044127817e-07,
+      "loss": 1.0647,
+      "step": 3994
+    },
+    {
+      "epoch": 0.9147109330280481,
+      "grad_norm": 1.3022246360778809,
+      "learning_rate": 3.885198142584612e-07,
+      "loss": 1.04,
+      "step": 3995
+    },
+    {
+      "epoch": 0.9149398969662278,
+      "grad_norm": 1.3392308950424194,
+      "learning_rate": 3.864747971926086e-07,
+      "loss": 1.0708,
+      "step": 3996
+    },
+    {
+      "epoch": 0.9151688609044075,
+      "grad_norm": 1.2771753072738647,
+      "learning_rate": 3.844350703690758e-07,
+      "loss": 1.0403,
+      "step": 3997
+    },
+    {
+      "epoch": 0.9153978248425872,
+      "grad_norm": 1.0660115480422974,
+      "learning_rate": 3.82400634910306e-07,
+      "loss": 1.0296,
+      "step": 3998
+    },
+    {
+      "epoch": 0.915626788780767,
+      "grad_norm": 1.3524727821350098,
+      "learning_rate": 3.8037149193583257e-07,
+      "loss": 1.0209,
+      "step": 3999
+    },
+    {
+      "epoch": 0.9158557527189468,
+      "grad_norm": 1.2420437335968018,
+      "learning_rate": 3.7834764256227674e-07,
+      "loss": 1.0699,
+      "step": 4000
+    },
+    {
+      "epoch": 0.9160847166571265,
+      "grad_norm": 1.2622954845428467,
+      "learning_rate": 3.7632908790334656e-07,
+      "loss": 1.038,
+      "step": 4001
+    },
+    {
+      "epoch": 0.9163136805953063,
+      "grad_norm": 1.260853886604309,
+      "learning_rate": 3.7431582906983124e-07,
+      "loss": 1.0352,
+      "step": 4002
+    },
+    {
+      "epoch": 0.916542644533486,
+      "grad_norm": 1.2435685396194458,
+      "learning_rate": 3.723078671696168e-07,
+      "loss": 1.0368,
+      "step": 4003
+    },
+    {
+      "epoch": 0.9167716084716657,
+      "grad_norm": 1.2323977947235107,
+      "learning_rate": 3.7030520330766264e-07,
+      "loss": 0.9676,
+      "step": 4004
+    },
+    {
+      "epoch": 0.9170005724098454,
+      "grad_norm": 1.191890001296997,
+      "learning_rate": 3.6830783858601835e-07,
+      "loss": 0.9949,
+      "step": 4005
+    },
+    {
+      "epoch": 0.9172295363480252,
+      "grad_norm": 1.2182968854904175,
+      "learning_rate": 3.6631577410382034e-07,
+      "loss": 1.0146,
+      "step": 4006
+    },
+    {
+      "epoch": 0.9174585002862049,
+      "grad_norm": 1.3352956771850586,
+      "learning_rate": 3.643290109572828e-07,
+      "loss": 1.0326,
+      "step": 4007
+    },
+    {
+      "epoch": 0.9176874642243846,
+      "grad_norm": 1.6930725574493408,
+      "learning_rate": 3.6234755023970447e-07,
+      "loss": 1.0013,
+      "step": 4008
+    },
+    {
+      "epoch": 0.9179164281625644,
+      "grad_norm": 1.251842737197876,
+      "learning_rate": 3.603713930414676e-07,
+      "loss": 1.0314,
+      "step": 4009
+    },
+    {
+      "epoch": 0.9181453921007441,
+      "grad_norm": 1.3096429109573364,
+      "learning_rate": 3.5840054045003346e-07,
+      "loss": 1.024,
+      "step": 4010
+    },
+    {
+      "epoch": 0.9183743560389239,
+      "grad_norm": 1.7176495790481567,
+      "learning_rate": 3.564349935499478e-07,
+      "loss": 0.9741,
+      "step": 4011
+    },
+    {
+      "epoch": 0.9186033199771036,
+      "grad_norm": 1.384275197982788,
+      "learning_rate": 3.5447475342283324e-07,
+      "loss": 1.0004,
+      "step": 4012
+    },
+    {
+      "epoch": 0.9188322839152834,
+      "grad_norm": 1.2257745265960693,
+      "learning_rate": 3.525198211473935e-07,
+      "loss": 0.9716,
+      "step": 4013
+    },
+    {
+      "epoch": 0.9190612478534631,
+      "grad_norm": 1.544185757637024,
+      "learning_rate": 3.5057019779941246e-07,
+      "loss": 1.0022,
+      "step": 4014
+    },
+    {
+      "epoch": 0.9192902117916428,
+      "grad_norm": 2.072894334793091,
+      "learning_rate": 3.4862588445174985e-07,
+      "loss": 0.9794,
+      "step": 4015
+    },
+    {
+      "epoch": 0.9195191757298226,
+      "grad_norm": 1.132996678352356,
+      "learning_rate": 3.466868821743452e-07,
+      "loss": 1.021,
+      "step": 4016
+    },
+    {
+      "epoch": 0.9197481396680023,
+      "grad_norm": 1.3861950635910034,
+      "learning_rate": 3.447531920342173e-07,
+      "loss": 1.0204,
+      "step": 4017
+    },
+    {
+      "epoch": 0.919977103606182,
+      "grad_norm": 1.2314361333847046,
+      "learning_rate": 3.428248150954583e-07,
+      "loss": 1.0447,
+      "step": 4018
+    },
+    {
+      "epoch": 0.9202060675443617,
+      "grad_norm": 1.2820926904678345,
+      "learning_rate": 3.4090175241923593e-07,
+      "loss": 1.0334,
+      "step": 4019
+    },
+    {
+      "epoch": 0.9204350314825415,
+      "grad_norm": 1.4982575178146362,
+      "learning_rate": 3.3898400506379937e-07,
+      "loss": 1.1002,
+      "step": 4020
+    },
+    {
+      "epoch": 0.9206639954207212,
+      "grad_norm": 2.526170492172241,
+      "learning_rate": 3.3707157408446547e-07,
+      "loss": 1.0653,
+      "step": 4021
+    },
+    {
+      "epoch": 0.9208929593589009,
+      "grad_norm": 2.164984703063965,
+      "learning_rate": 3.3516446053363015e-07,
+      "loss": 0.9579,
+      "step": 4022
+    },
+    {
+      "epoch": 0.9211219232970808,
+      "grad_norm": 1.396539330482483,
+      "learning_rate": 3.3326266546076293e-07,
+      "loss": 0.9919,
+      "step": 4023
+    },
+    {
+      "epoch": 0.9213508872352605,
+      "grad_norm": 1.2713897228240967,
+      "learning_rate": 3.3136618991240544e-07,
+      "loss": 1.0049,
+      "step": 4024
+    },
+    {
+      "epoch": 0.9215798511734402,
+      "grad_norm": 1.5097935199737549,
+      "learning_rate": 3.294750349321718e-07,
+      "loss": 1.0192,
+      "step": 4025
+    },
+    {
+      "epoch": 0.9218088151116199,
+      "grad_norm": 1.3932673931121826,
+      "learning_rate": 3.2758920156074624e-07,
+      "loss": 0.9681,
+      "step": 4026
+    },
+    {
+      "epoch": 0.9220377790497997,
+      "grad_norm": 1.4764078855514526,
+      "learning_rate": 3.2570869083589196e-07,
+      "loss": 1.0264,
+      "step": 4027
+    },
+    {
+      "epoch": 0.9222667429879794,
+      "grad_norm": 1.1931257247924805,
+      "learning_rate": 3.238335037924356e-07,
+      "loss": 1.0474,
+      "step": 4028
+    },
+    {
+      "epoch": 0.9224957069261591,
+      "grad_norm": 1.2183254957199097,
+      "learning_rate": 3.219636414622751e-07,
+      "loss": 1.0471,
+      "step": 4029
+    },
+    {
+      "epoch": 0.9227246708643388,
+      "grad_norm": 1.6249024868011475,
+      "learning_rate": 3.20099104874384e-07,
+      "loss": 1.0004,
+      "step": 4030
+    },
+    {
+      "epoch": 0.9229536348025186,
+      "grad_norm": 1.2928696870803833,
+      "learning_rate": 3.1823989505479934e-07,
+      "loss": 0.969,
+      "step": 4031
+    },
+    {
+      "epoch": 0.9231825987406983,
+      "grad_norm": 1.292734980583191,
+      "learning_rate": 3.163860130266283e-07,
+      "loss": 1.0344,
+      "step": 4032
+    },
+    {
+      "epoch": 0.923411562678878,
+      "grad_norm": 1.6928443908691406,
+      "learning_rate": 3.145374598100448e-07,
+      "loss": 1.02,
+      "step": 4033
+    },
+    {
+      "epoch": 0.9236405266170579,
+      "grad_norm": 1.1957359313964844,
+      "learning_rate": 3.1269423642229737e-07,
+      "loss": 1.0593,
+      "step": 4034
+    },
+    {
+      "epoch": 0.9238694905552376,
+      "grad_norm": 1.3396857976913452,
+      "learning_rate": 3.1085634387769124e-07,
+      "loss": 1.0156,
+      "step": 4035
+    },
+    {
+      "epoch": 0.9240984544934173,
+      "grad_norm": 1.7538094520568848,
+      "learning_rate": 3.090237831876053e-07,
+      "loss": 1.0305,
+      "step": 4036
+    },
+    {
+      "epoch": 0.924327418431597,
+      "grad_norm": 1.2242697477340698,
+      "learning_rate": 3.0719655536048276e-07,
+      "loss": 1.027,
+      "step": 4037
+    },
+    {
+      "epoch": 0.9245563823697768,
+      "grad_norm": 1.4524619579315186,
+      "learning_rate": 3.053746614018305e-07,
+      "loss": 1.0006,
+      "step": 4038
+    },
+    {
+      "epoch": 0.9247853463079565,
+      "grad_norm": 1.2049915790557861,
+      "learning_rate": 3.035581023142231e-07,
+      "loss": 1.043,
+      "step": 4039
+    },
+    {
+      "epoch": 0.9250143102461362,
+      "grad_norm": 1.5115872621536255,
+      "learning_rate": 3.0174687909729663e-07,
+      "loss": 1.0647,
+      "step": 4040
+    },
+    {
+      "epoch": 0.925243274184316,
+      "grad_norm": 1.2160587310791016,
+      "learning_rate": 2.9994099274775256e-07,
+      "loss": 1.0233,
+      "step": 4041
+    },
+    {
+      "epoch": 0.9254722381224957,
+      "grad_norm": 1.3652007579803467,
+      "learning_rate": 2.9814044425935605e-07,
+      "loss": 1.0257,
+      "step": 4042
+    },
+    {
+      "epoch": 0.9257012020606754,
+      "grad_norm": 1.5073693990707397,
+      "learning_rate": 2.9634523462293005e-07,
+      "loss": 1.0529,
+      "step": 4043
+    },
+    {
+      "epoch": 0.9259301659988551,
+      "grad_norm": 1.4959176778793335,
+      "learning_rate": 2.945553648263677e-07,
+      "loss": 1.0633,
+      "step": 4044
+    },
+    {
+      "epoch": 0.9261591299370349,
+      "grad_norm": 1.320374846458435,
+      "learning_rate": 2.9277083585461776e-07,
+      "loss": 0.9744,
+      "step": 4045
+    },
+    {
+      "epoch": 0.9263880938752147,
+      "grad_norm": 1.4175662994384766,
+      "learning_rate": 2.909916486896913e-07,
+      "loss": 0.9728,
+      "step": 4046
+    },
+    {
+      "epoch": 0.9266170578133944,
+      "grad_norm": 1.35358464717865,
+      "learning_rate": 2.8921780431065973e-07,
+      "loss": 1.0241,
+      "step": 4047
+    },
+    {
+      "epoch": 0.9268460217515742,
+      "grad_norm": 1.674445390701294,
+      "learning_rate": 2.874493036936554e-07,
+      "loss": 0.9889,
+      "step": 4048
+    },
+    {
+      "epoch": 0.9270749856897539,
+      "grad_norm": 1.1121513843536377,
+      "learning_rate": 2.8568614781186886e-07,
+      "loss": 1.0436,
+      "step": 4049
+    },
+    {
+      "epoch": 0.9273039496279336,
+      "grad_norm": 1.7877289056777954,
+      "learning_rate": 2.839283376355506e-07,
+      "loss": 1.0327,
+      "step": 4050
+    },
+    {
+      "epoch": 0.9275329135661133,
+      "grad_norm": 1.4056133031845093,
+      "learning_rate": 2.82175874132008e-07,
+      "loss": 1.0264,
+      "step": 4051
+    },
+    {
+      "epoch": 0.9277618775042931,
+      "grad_norm": 1.2976934909820557,
+      "learning_rate": 2.804287582656073e-07,
+      "loss": 1.1032,
+      "step": 4052
+    },
+    {
+      "epoch": 0.9279908414424728,
+      "grad_norm": 1.4245258569717407,
+      "learning_rate": 2.7868699099777295e-07,
+      "loss": 1.0262,
+      "step": 4053
+    },
+    {
+      "epoch": 0.9282198053806525,
+      "grad_norm": 1.3631470203399658,
+      "learning_rate": 2.769505732869837e-07,
+      "loss": 1.0073,
+      "step": 4054
+    },
+    {
+      "epoch": 0.9284487693188322,
+      "grad_norm": 2.766822576522827,
+      "learning_rate": 2.752195060887775e-07,
+      "loss": 0.9939,
+      "step": 4055
+    },
+    {
+      "epoch": 0.928677733257012,
+      "grad_norm": 1.360872507095337,
+      "learning_rate": 2.7349379035574217e-07,
+      "loss": 1.0607,
+      "step": 4056
+    },
+    {
+      "epoch": 0.9289066971951918,
+      "grad_norm": 1.0225962400436401,
+      "learning_rate": 2.717734270375272e-07,
+      "loss": 1.0532,
+      "step": 4057
+    },
+    {
+      "epoch": 0.9291356611333715,
+      "grad_norm": 1.3829964399337769,
+      "learning_rate": 2.7005841708083516e-07,
+      "loss": 1.057,
+      "step": 4058
+    },
+    {
+      "epoch": 0.9293646250715513,
+      "grad_norm": 1.5266767740249634,
+      "learning_rate": 2.683487614294189e-07,
+      "loss": 1.0197,
+      "step": 4059
+    },
+    {
+      "epoch": 0.929593589009731,
+      "grad_norm": 1.3430898189544678,
+      "learning_rate": 2.6664446102408924e-07,
+      "loss": 1.046,
+      "step": 4060
+    },
+    {
+      "epoch": 0.9298225529479107,
+      "grad_norm": 1.3258440494537354,
+      "learning_rate": 2.649455168027082e-07,
+      "loss": 1.0198,
+      "step": 4061
+    },
+    {
+      "epoch": 0.9300515168860904,
+      "grad_norm": 1.0601249933242798,
+      "learning_rate": 2.632519297001912e-07,
+      "loss": 1.0252,
+      "step": 4062
+    },
+    {
+      "epoch": 0.9302804808242702,
+      "grad_norm": 1.4658838510513306,
+      "learning_rate": 2.615637006485039e-07,
+      "loss": 1.0577,
+      "step": 4063
+    },
+    {
+      "epoch": 0.9305094447624499,
+      "grad_norm": 1.3812763690948486,
+      "learning_rate": 2.5988083057666534e-07,
+      "loss": 1.0447,
+      "step": 4064
+    },
+    {
+      "epoch": 0.9307384087006296,
+      "grad_norm": 1.4836375713348389,
+      "learning_rate": 2.582033204107437e-07,
+      "loss": 0.9993,
+      "step": 4065
+    },
+    {
+      "epoch": 0.9309673726388094,
+      "grad_norm": 1.3811873197555542,
+      "learning_rate": 2.565311710738616e-07,
+      "loss": 1.0034,
+      "step": 4066
+    },
+    {
+      "epoch": 0.9311963365769891,
+      "grad_norm": 1.320236325263977,
+      "learning_rate": 2.5486438348618416e-07,
+      "loss": 0.9957,
+      "step": 4067
+    },
+    {
+      "epoch": 0.9314253005151688,
+      "grad_norm": 1.1758593320846558,
+      "learning_rate": 2.532029585649343e-07,
+      "loss": 1.0828,
+      "step": 4068
+    },
+    {
+      "epoch": 0.9316542644533486,
+      "grad_norm": 1.2372307777404785,
+      "learning_rate": 2.515468972243795e-07,
+      "loss": 1.0289,
+      "step": 4069
+    },
+    {
+      "epoch": 0.9318832283915284,
+      "grad_norm": 1.5127670764923096,
+      "learning_rate": 2.4989620037583653e-07,
+      "loss": 1.116,
+      "step": 4070
+    },
+    {
+      "epoch": 0.9321121923297081,
+      "grad_norm": 1.2241275310516357,
+      "learning_rate": 2.4825086892766745e-07,
+      "loss": 0.9408,
+      "step": 4071
+    },
+    {
+      "epoch": 0.9323411562678878,
+      "grad_norm": 1.1708391904830933,
+      "learning_rate": 2.466109037852893e-07,
+      "loss": 1.0203,
+      "step": 4072
+    },
+    {
+      "epoch": 0.9325701202060676,
+      "grad_norm": 1.14144766330719,
+      "learning_rate": 2.4497630585115673e-07,
+      "loss": 0.9937,
+      "step": 4073
+    },
+    {
+      "epoch": 0.9327990841442473,
+      "grad_norm": 2.2938714027404785,
+      "learning_rate": 2.4334707602477693e-07,
+      "loss": 1.0365,
+      "step": 4074
+    },
+    {
+      "epoch": 0.933028048082427,
+      "grad_norm": 1.596992015838623,
+      "learning_rate": 2.417232152027016e-07,
+      "loss": 1.0269,
+      "step": 4075
+    },
+    {
+      "epoch": 0.9332570120206067,
+      "grad_norm": 1.3394994735717773,
+      "learning_rate": 2.4010472427852706e-07,
+      "loss": 1.0008,
+      "step": 4076
+    },
+    {
+      "epoch": 0.9334859759587865,
+      "grad_norm": 2.9255053997039795,
+      "learning_rate": 2.384916041428964e-07,
+      "loss": 1.0246,
+      "step": 4077
+    },
+    {
+      "epoch": 0.9337149398969662,
+      "grad_norm": 2.5200510025024414,
+      "learning_rate": 2.3688385568349515e-07,
+      "loss": 1.0253,
+      "step": 4078
+    },
+    {
+      "epoch": 0.9339439038351459,
+      "grad_norm": 1.3866181373596191,
+      "learning_rate": 2.3528147978505334e-07,
+      "loss": 1.0208,
+      "step": 4079
+    },
+    {
+      "epoch": 0.9341728677733258,
+      "grad_norm": 1.7252721786499023,
+      "learning_rate": 2.3368447732934785e-07,
+      "loss": 1.0771,
+      "step": 4080
+    },
+    {
+      "epoch": 0.9344018317115055,
+      "grad_norm": 1.2650654315948486,
+      "learning_rate": 2.3209284919519127e-07,
+      "loss": 1.0407,
+      "step": 4081
+    },
+    {
+      "epoch": 0.9346307956496852,
+      "grad_norm": 1.2546926736831665,
+      "learning_rate": 2.305065962584463e-07,
+      "loss": 1.0684,
+      "step": 4082
+    },
+    {
+      "epoch": 0.9348597595878649,
+      "grad_norm": 1.51604425907135,
+      "learning_rate": 2.289257193920158e-07,
+      "loss": 1.0321,
+      "step": 4083
+    },
+    {
+      "epoch": 0.9350887235260447,
+      "grad_norm": 1.2213822603225708,
+      "learning_rate": 2.2735021946583946e-07,
+      "loss": 1.0326,
+      "step": 4084
+    },
+    {
+      "epoch": 0.9353176874642244,
+      "grad_norm": 1.4243481159210205,
+      "learning_rate": 2.2578009734690264e-07,
+      "loss": 1.0401,
+      "step": 4085
+    },
+    {
+      "epoch": 0.9355466514024041,
+      "grad_norm": 1.6007384061813354,
+      "learning_rate": 2.242153538992331e-07,
+      "loss": 0.9734,
+      "step": 4086
+    },
+    {
+      "epoch": 0.9357756153405838,
+      "grad_norm": 2.3834850788116455,
+      "learning_rate": 2.2265598998389316e-07,
+      "loss": 0.988,
+      "step": 4087
+    },
+    {
+      "epoch": 0.9360045792787636,
+      "grad_norm": 1.524840235710144,
+      "learning_rate": 2.211020064589886e-07,
+      "loss": 1.0204,
+      "step": 4088
+    },
+    {
+      "epoch": 0.9362335432169433,
+      "grad_norm": 1.3676830530166626,
+      "learning_rate": 2.195534041796632e-07,
+      "loss": 1.0464,
+      "step": 4089
+    },
+    {
+      "epoch": 0.936462507155123,
+      "grad_norm": 1.5607987642288208,
+      "learning_rate": 2.180101839981008e-07,
+      "loss": 1.0407,
+      "step": 4090
+    },
+    {
+      "epoch": 0.9366914710933028,
+      "grad_norm": 1.0320096015930176,
+      "learning_rate": 2.1647234676352213e-07,
+      "loss": 1.0463,
+      "step": 4091
+    },
+    {
+      "epoch": 0.9369204350314826,
+      "grad_norm": 1.3179558515548706,
+      "learning_rate": 2.1493989332218468e-07,
+      "loss": 1.0596,
+      "step": 4092
+    },
+    {
+      "epoch": 0.9371493989696623,
+      "grad_norm": 1.3272641897201538,
+      "learning_rate": 2.134128245173872e-07,
+      "loss": 1.0372,
+      "step": 4093
+    },
+    {
+      "epoch": 0.937378362907842,
+      "grad_norm": 1.2094119787216187,
+      "learning_rate": 2.1189114118946196e-07,
+      "loss": 1.032,
+      "step": 4094
+    },
+    {
+      "epoch": 0.9376073268460218,
+      "grad_norm": 1.2302625179290771,
+      "learning_rate": 2.103748441757758e-07,
+      "loss": 0.9675,
+      "step": 4095
+    },
+    {
+      "epoch": 0.9378362907842015,
+      "grad_norm": 1.7436269521713257,
+      "learning_rate": 2.0886393431073794e-07,
+      "loss": 1.0837,
+      "step": 4096
+    },
+    {
+      "epoch": 0.9380652547223812,
+      "grad_norm": 1.3275481462478638,
+      "learning_rate": 2.0735841242578992e-07,
+      "loss": 1.0377,
+      "step": 4097
+    },
+    {
+      "epoch": 0.938294218660561,
+      "grad_norm": 1.3621660470962524,
+      "learning_rate": 2.0585827934940238e-07,
+      "loss": 1.0369,
+      "step": 4098
+    },
+    {
+      "epoch": 0.9385231825987407,
+      "grad_norm": 1.739432692527771,
+      "learning_rate": 2.043635359070928e-07,
+      "loss": 1.0002,
+      "step": 4099
+    },
+    {
+      "epoch": 0.9387521465369204,
+      "grad_norm": 1.273725986480713,
+      "learning_rate": 2.0287418292140204e-07,
+      "loss": 0.9611,
+      "step": 4100
+    },
+    {
+      "epoch": 0.9389811104751001,
+      "grad_norm": 1.3180021047592163,
+      "learning_rate": 2.0139022121190788e-07,
+      "loss": 1.0431,
+      "step": 4101
+    },
+    {
+      "epoch": 0.9392100744132799,
+      "grad_norm": 1.3748250007629395,
+      "learning_rate": 1.9991165159522485e-07,
+      "loss": 0.9487,
+      "step": 4102
+    },
+    {
+      "epoch": 0.9394390383514597,
+      "grad_norm": 1.186481237411499,
+      "learning_rate": 1.9843847488499545e-07,
+      "loss": 1.0343,
+      "step": 4103
+    },
+    {
+      "epoch": 0.9396680022896394,
+      "grad_norm": 1.2737975120544434,
+      "learning_rate": 1.9697069189189678e-07,
+      "loss": 1.0535,
+      "step": 4104
+    },
+    {
+      "epoch": 0.9398969662278192,
+      "grad_norm": 1.1873112916946411,
+      "learning_rate": 1.9550830342363714e-07,
+      "loss": 1.0056,
+      "step": 4105
+    },
+    {
+      "epoch": 0.9401259301659989,
+      "grad_norm": 1.3513420820236206,
+      "learning_rate": 1.9405131028495838e-07,
+      "loss": 1.0278,
+      "step": 4106
+    },
+    {
+      "epoch": 0.9403548941041786,
+      "grad_norm": 1.1599657535552979,
+      "learning_rate": 1.9259971327763028e-07,
+      "loss": 1.0013,
+      "step": 4107
+    },
+    {
+      "epoch": 0.9405838580423583,
+      "grad_norm": 2.088212728500366,
+      "learning_rate": 1.9115351320045495e-07,
+      "loss": 1.0098,
+      "step": 4108
+    },
+    {
+      "epoch": 0.9408128219805381,
+      "grad_norm": 1.281222939491272,
+      "learning_rate": 1.8971271084926245e-07,
+      "loss": 0.9935,
+      "step": 4109
+    },
+    {
+      "epoch": 0.9410417859187178,
+      "grad_norm": 1.147822618484497,
+      "learning_rate": 1.8827730701691749e-07,
+      "loss": 1.0115,
+      "step": 4110
+    },
+    {
+      "epoch": 0.9412707498568975,
+      "grad_norm": 2.1663131713867188,
+      "learning_rate": 1.868473024933082e-07,
+      "loss": 1.1268,
+      "step": 4111
+    },
+    {
+      "epoch": 0.9414997137950772,
+      "grad_norm": 1.2913000583648682,
+      "learning_rate": 1.8542269806535285e-07,
+      "loss": 1.0183,
+      "step": 4112
+    },
+    {
+      "epoch": 0.941728677733257,
+      "grad_norm": 1.3916893005371094,
+      "learning_rate": 1.8400349451700438e-07,
+      "loss": 1.0304,
+      "step": 4113
+    },
+    {
+      "epoch": 0.9419576416714367,
+      "grad_norm": 1.3906985521316528,
+      "learning_rate": 1.8258969262923366e-07,
+      "loss": 1.0088,
+      "step": 4114
+    },
+    {
+      "epoch": 0.9421866056096165,
+      "grad_norm": 1.523729681968689,
+      "learning_rate": 1.8118129318004717e-07,
+      "loss": 1.0386,
+      "step": 4115
+    },
+    {
+      "epoch": 0.9424155695477963,
+      "grad_norm": 1.3909317255020142,
+      "learning_rate": 1.7977829694447278e-07,
+      "loss": 0.9631,
+      "step": 4116
+    },
+    {
+      "epoch": 0.942644533485976,
+      "grad_norm": 1.4848453998565674,
+      "learning_rate": 1.7838070469456958e-07,
+      "loss": 0.9619,
+      "step": 4117
+    },
+    {
+      "epoch": 0.9428734974241557,
+      "grad_norm": 1.3331187963485718,
+      "learning_rate": 1.769885171994201e-07,
+      "loss": 1.0474,
+      "step": 4118
+    },
+    {
+      "epoch": 0.9431024613623354,
+      "grad_norm": 1.6051127910614014,
+      "learning_rate": 1.7560173522513268e-07,
+      "loss": 1.0188,
+      "step": 4119
+    },
+    {
+      "epoch": 0.9433314253005152,
+      "grad_norm": 1.2383366823196411,
+      "learning_rate": 1.742203595348435e-07,
+      "loss": 1.0819,
+      "step": 4120
+    },
+    {
+      "epoch": 0.9435603892386949,
+      "grad_norm": 1.607271671295166,
+      "learning_rate": 1.728443908887112e-07,
+      "loss": 0.9864,
+      "step": 4121
+    },
+    {
+      "epoch": 0.9437893531768746,
+      "grad_norm": 1.5015615224838257,
+      "learning_rate": 1.7147383004391782e-07,
+      "loss": 1.0301,
+      "step": 4122
+    },
+    {
+      "epoch": 0.9440183171150544,
+      "grad_norm": 1.2938905954360962,
+      "learning_rate": 1.7010867775467454e-07,
+      "loss": 0.9931,
+      "step": 4123
+    },
+    {
+      "epoch": 0.9442472810532341,
+      "grad_norm": 1.4441864490509033,
+      "learning_rate": 1.6874893477221376e-07,
+      "loss": 1.0832,
+      "step": 4124
+    },
+    {
+      "epoch": 0.9444762449914138,
+      "grad_norm": 1.309255599975586,
+      "learning_rate": 1.6739460184478694e-07,
+      "loss": 1.0367,
+      "step": 4125
+    },
+    {
+      "epoch": 0.9447052089295936,
+      "grad_norm": 1.3141146898269653,
+      "learning_rate": 1.660456797176735e-07,
+      "loss": 1.0261,
+      "step": 4126
+    },
+    {
+      "epoch": 0.9449341728677734,
+      "grad_norm": 1.307094931602478,
+      "learning_rate": 1.6470216913317628e-07,
+      "loss": 1.0172,
+      "step": 4127
+    },
+    {
+      "epoch": 0.9451631368059531,
+      "grad_norm": 1.344239592552185,
+      "learning_rate": 1.63364070830615e-07,
+      "loss": 1.0416,
+      "step": 4128
+    },
+    {
+      "epoch": 0.9453921007441328,
+      "grad_norm": 1.2404433488845825,
+      "learning_rate": 1.6203138554633625e-07,
+      "loss": 1.0191,
+      "step": 4129
+    },
+    {
+      "epoch": 0.9456210646823126,
+      "grad_norm": 1.1200016736984253,
+      "learning_rate": 1.6070411401370335e-07,
+      "loss": 0.9745,
+      "step": 4130
+    },
+    {
+      "epoch": 0.9458500286204923,
+      "grad_norm": 3.205339193344116,
+      "learning_rate": 1.5938225696310427e-07,
+      "loss": 1.0064,
+      "step": 4131
+    },
+    {
+      "epoch": 0.946078992558672,
+      "grad_norm": 1.468671441078186,
+      "learning_rate": 1.580658151219461e-07,
+      "loss": 1.0288,
+      "step": 4132
+    },
+    {
+      "epoch": 0.9463079564968517,
+      "grad_norm": 1.7323161363601685,
+      "learning_rate": 1.567547892146537e-07,
+      "loss": 0.9814,
+      "step": 4133
+    },
+    {
+      "epoch": 0.9465369204350315,
+      "grad_norm": 1.4511128664016724,
+      "learning_rate": 1.5544917996267562e-07,
+      "loss": 0.9879,
+      "step": 4134
+    },
+    {
+      "epoch": 0.9467658843732112,
+      "grad_norm": 1.2336581945419312,
+      "learning_rate": 1.541489880844782e-07,
+      "loss": 1.0541,
+      "step": 4135
+    },
+    {
+      "epoch": 0.9469948483113909,
+      "grad_norm": 1.2576466798782349,
+      "learning_rate": 1.528542142955436e-07,
+      "loss": 1.0047,
+      "step": 4136
+    },
+    {
+      "epoch": 0.9472238122495706,
+      "grad_norm": 1.4360203742980957,
+      "learning_rate": 1.5156485930837628e-07,
+      "loss": 1.0658,
+      "step": 4137
+    },
+    {
+      "epoch": 0.9474527761877505,
+      "grad_norm": 1.8652851581573486,
+      "learning_rate": 1.5028092383249871e-07,
+      "loss": 1.0568,
+      "step": 4138
+    },
+    {
+      "epoch": 0.9476817401259302,
+      "grad_norm": 1.3658671379089355,
+      "learning_rate": 1.4900240857444792e-07,
+      "loss": 1.095,
+      "step": 4139
+    },
+    {
+      "epoch": 0.9479107040641099,
+      "grad_norm": 1.3814666271209717,
+      "learning_rate": 1.4772931423778003e-07,
+      "loss": 1.089,
+      "step": 4140
+    },
+    {
+      "epoch": 0.9481396680022897,
+      "grad_norm": 1.3728240728378296,
+      "learning_rate": 1.464616415230702e-07,
+      "loss": 1.0431,
+      "step": 4141
+    },
+    {
+      "epoch": 0.9483686319404694,
+      "grad_norm": 1.6586147546768188,
+      "learning_rate": 1.4519939112790705e-07,
+      "loss": 1.0047,
+      "step": 4142
+    },
+    {
+      "epoch": 0.9485975958786491,
+      "grad_norm": 1.2365854978561401,
+      "learning_rate": 1.4394256374689498e-07,
+      "loss": 1.0841,
+      "step": 4143
+    },
+    {
+      "epoch": 0.9488265598168288,
+      "grad_norm": 1.3631367683410645,
+      "learning_rate": 1.426911600716574e-07,
+      "loss": 1.0188,
+      "step": 4144
+    },
+    {
+      "epoch": 0.9490555237550086,
+      "grad_norm": 1.1804825067520142,
+      "learning_rate": 1.4144518079083125e-07,
+      "loss": 1.0877,
+      "step": 4145
+    },
+    {
+      "epoch": 0.9492844876931883,
+      "grad_norm": 1.24893057346344,
+      "learning_rate": 1.4020462659006806e-07,
+      "loss": 1.0689,
+      "step": 4146
+    },
+    {
+      "epoch": 0.949513451631368,
+      "grad_norm": 1.317393183708191,
+      "learning_rate": 1.3896949815203398e-07,
+      "loss": 1.0285,
+      "step": 4147
+    },
+    {
+      "epoch": 0.9497424155695477,
+      "grad_norm": 1.233035683631897,
+      "learning_rate": 1.3773979615640976e-07,
+      "loss": 1.0676,
+      "step": 4148
+    },
+    {
+      "epoch": 0.9499713795077276,
+      "grad_norm": 1.0426533222198486,
+      "learning_rate": 1.3651552127989186e-07,
+      "loss": 1.0019,
+      "step": 4149
+    },
+    {
+      "epoch": 0.9502003434459073,
+      "grad_norm": 1.5621211528778076,
+      "learning_rate": 1.3529667419618475e-07,
+      "loss": 1.0579,
+      "step": 4150
+    },
+    {
+      "epoch": 0.950429307384087,
+      "grad_norm": 1.3285621404647827,
+      "learning_rate": 1.3408325557601297e-07,
+      "loss": 1.0285,
+      "step": 4151
+    },
+    {
+      "epoch": 0.9506582713222668,
+      "grad_norm": 1.1618311405181885,
+      "learning_rate": 1.3287526608711132e-07,
+      "loss": 1.046,
+      "step": 4152
+    },
+    {
+      "epoch": 0.9508872352604465,
+      "grad_norm": 1.317858338356018,
+      "learning_rate": 1.3167270639422357e-07,
+      "loss": 1.0425,
+      "step": 4153
+    },
+    {
+      "epoch": 0.9511161991986262,
+      "grad_norm": 1.2323342561721802,
+      "learning_rate": 1.304755771591093e-07,
+      "loss": 1.0702,
+      "step": 4154
+    },
+    {
+      "epoch": 0.951345163136806,
+      "grad_norm": 1.3113363981246948,
+      "learning_rate": 1.292838790405393e-07,
+      "loss": 1.026,
+      "step": 4155
+    },
+    {
+      "epoch": 0.9515741270749857,
+      "grad_norm": 1.145249366760254,
+      "learning_rate": 1.2809761269429343e-07,
+      "loss": 1.0076,
+      "step": 4156
+    },
+    {
+      "epoch": 0.9518030910131654,
+      "grad_norm": 1.527535319328308,
+      "learning_rate": 1.269167787731662e-07,
+      "loss": 0.9754,
+      "step": 4157
+    },
+    {
+      "epoch": 0.9520320549513451,
+      "grad_norm": 1.321520209312439,
+      "learning_rate": 1.2574137792695895e-07,
+      "loss": 1.0286,
+      "step": 4158
+    },
+    {
+      "epoch": 0.9522610188895249,
+      "grad_norm": 1.5232340097427368,
+      "learning_rate": 1.2457141080248647e-07,
+      "loss": 0.9744,
+      "step": 4159
+    },
+    {
+      "epoch": 0.9524899828277046,
+      "grad_norm": 1.3998576402664185,
+      "learning_rate": 1.2340687804357044e-07,
+      "loss": 1.0796,
+      "step": 4160
+    },
+    {
+      "epoch": 0.9527189467658844,
+      "grad_norm": 1.2090296745300293,
+      "learning_rate": 1.2224778029104377e-07,
+      "loss": 1.0844,
+      "step": 4161
+    },
+    {
+      "epoch": 0.9529479107040641,
+      "grad_norm": 1.319123387336731,
+      "learning_rate": 1.2109411818274851e-07,
+      "loss": 1.0412,
+      "step": 4162
+    },
+    {
+      "epoch": 0.9531768746422439,
+      "grad_norm": 1.2785385847091675,
+      "learning_rate": 1.1994589235353682e-07,
+      "loss": 1.0274,
+      "step": 4163
+    },
+    {
+      "epoch": 0.9534058385804236,
+      "grad_norm": 1.2551246881484985,
+      "learning_rate": 1.1880310343526324e-07,
+      "loss": 1.0395,
+      "step": 4164
+    },
+    {
+      "epoch": 0.9536348025186033,
+      "grad_norm": 1.755634069442749,
+      "learning_rate": 1.1766575205680031e-07,
+      "loss": 1.0167,
+      "step": 4165
+    },
+    {
+      "epoch": 0.9538637664567831,
+      "grad_norm": 1.1135053634643555,
+      "learning_rate": 1.1653383884401959e-07,
+      "loss": 0.9704,
+      "step": 4166
+    },
+    {
+      "epoch": 0.9540927303949628,
+      "grad_norm": 1.2887723445892334,
+      "learning_rate": 1.1540736441980505e-07,
+      "loss": 1.0541,
+      "step": 4167
+    },
+    {
+      "epoch": 0.9543216943331425,
+      "grad_norm": 1.2668262720108032,
+      "learning_rate": 1.1428632940404416e-07,
+      "loss": 1.0312,
+      "step": 4168
+    },
+    {
+      "epoch": 0.9545506582713222,
+      "grad_norm": 1.4527376890182495,
+      "learning_rate": 1.1317073441363458e-07,
+      "loss": 1.0343,
+      "step": 4169
+    },
+    {
+      "epoch": 0.954779622209502,
+      "grad_norm": 1.378683090209961,
+      "learning_rate": 1.1206058006247966e-07,
+      "loss": 0.9851,
+      "step": 4170
+    },
+    {
+      "epoch": 0.9550085861476817,
+      "grad_norm": 1.202966570854187,
+      "learning_rate": 1.109558669614863e-07,
+      "loss": 1.027,
+      "step": 4171
+    },
+    {
+      "epoch": 0.9552375500858615,
+      "grad_norm": 1.4312129020690918,
+      "learning_rate": 1.0985659571856933e-07,
+      "loss": 1.0712,
+      "step": 4172
+    },
+    {
+      "epoch": 0.9554665140240413,
+      "grad_norm": 1.2038168907165527,
+      "learning_rate": 1.0876276693864818e-07,
+      "loss": 1.0783,
+      "step": 4173
+    },
+    {
+      "epoch": 0.955695477962221,
+      "grad_norm": 1.1261094808578491,
+      "learning_rate": 1.0767438122364914e-07,
+      "loss": 1.0094,
+      "step": 4174
+    },
+    {
+      "epoch": 0.9559244419004007,
+      "grad_norm": 2.3489320278167725,
+      "learning_rate": 1.0659143917250092e-07,
+      "loss": 1.0653,
+      "step": 4175
+    },
+    {
+      "epoch": 0.9561534058385804,
+      "grad_norm": 1.5417100191116333,
+      "learning_rate": 1.055139413811379e-07,
+      "loss": 0.9797,
+      "step": 4176
+    },
+    {
+      "epoch": 0.9563823697767602,
+      "grad_norm": 1.4111857414245605,
+      "learning_rate": 1.0444188844249691e-07,
+      "loss": 0.997,
+      "step": 4177
+    },
+    {
+      "epoch": 0.9566113337149399,
+      "grad_norm": 1.4079347848892212,
+      "learning_rate": 1.0337528094651938e-07,
+      "loss": 0.9685,
+      "step": 4178
+    },
+    {
+      "epoch": 0.9568402976531196,
+      "grad_norm": 1.3267221450805664,
+      "learning_rate": 1.0231411948015247e-07,
+      "loss": 1.1016,
+      "step": 4179
+    },
+    {
+      "epoch": 0.9570692615912993,
+      "grad_norm": 1.3851131200790405,
+      "learning_rate": 1.0125840462734238e-07,
+      "loss": 1.0756,
+      "step": 4180
+    },
+    {
+      "epoch": 0.9572982255294791,
+      "grad_norm": 1.3021942377090454,
+      "learning_rate": 1.0020813696904108e-07,
+      "loss": 0.9716,
+      "step": 4181
+    },
+    {
+      "epoch": 0.9575271894676588,
+      "grad_norm": 1.1864266395568848,
+      "learning_rate": 9.916331708320403e-08,
+      "loss": 0.9976,
+      "step": 4182
+    },
+    {
+      "epoch": 0.9577561534058385,
+      "grad_norm": 13.135167121887207,
+      "learning_rate": 9.812394554478355e-08,
+      "loss": 1.0404,
+      "step": 4183
+    },
+    {
+      "epoch": 0.9579851173440184,
+      "grad_norm": 1.3869199752807617,
+      "learning_rate": 9.709002292573876e-08,
+      "loss": 1.0379,
+      "step": 4184
+    },
+    {
+      "epoch": 0.9582140812821981,
+      "grad_norm": 1.8574575185775757,
+      "learning_rate": 9.606154979502791e-08,
+      "loss": 1.0384,
+      "step": 4185
+    },
+    {
+      "epoch": 0.9584430452203778,
+      "grad_norm": 1.1653584241867065,
+      "learning_rate": 9.503852671861158e-08,
+      "loss": 1.0695,
+      "step": 4186
+    },
+    {
+      "epoch": 0.9586720091585575,
+      "grad_norm": 1.499384880065918,
+      "learning_rate": 9.402095425945168e-08,
+      "loss": 1.0062,
+      "step": 4187
+    },
+    {
+      "epoch": 0.9589009730967373,
+      "grad_norm": 1.988095760345459,
+      "learning_rate": 9.300883297750696e-08,
+      "loss": 1.0296,
+      "step": 4188
+    },
+    {
+      "epoch": 0.959129937034917,
+      "grad_norm": 1.2225043773651123,
+      "learning_rate": 9.2002163429743e-08,
+      "loss": 0.9754,
+      "step": 4189
+    },
+    {
+      "epoch": 0.9593589009730967,
+      "grad_norm": 1.299344778060913,
+      "learning_rate": 9.10009461701189e-08,
+      "loss": 1.0025,
+      "step": 4190
+    },
+    {
+      "epoch": 0.9595878649112765,
+      "grad_norm": 1.1545172929763794,
+      "learning_rate": 9.000518174959726e-08,
+      "loss": 1.0316,
+      "step": 4191
+    },
+    {
+      "epoch": 0.9598168288494562,
+      "grad_norm": 1.2487175464630127,
+      "learning_rate": 8.901487071613868e-08,
+      "loss": 1.06,
+      "step": 4192
+    },
+    {
+      "epoch": 0.9600457927876359,
+      "grad_norm": 1.9878804683685303,
+      "learning_rate": 8.803001361470386e-08,
+      "loss": 1.053,
+      "step": 4193
+    },
+    {
+      "epoch": 0.9602747567258156,
+      "grad_norm": 1.282788634300232,
+      "learning_rate": 8.705061098724932e-08,
+      "loss": 0.9953,
+      "step": 4194
+    },
+    {
+      "epoch": 0.9605037206639955,
+      "grad_norm": 1.5588258504867554,
+      "learning_rate": 8.60766633727339e-08,
+      "loss": 1.0122,
+      "step": 4195
+    },
+    {
+      "epoch": 0.9607326846021752,
+      "grad_norm": 2.007049798965454,
+      "learning_rate": 8.510817130711224e-08,
+      "loss": 1.0305,
+      "step": 4196
+    },
+    {
+      "epoch": 0.9609616485403549,
+      "grad_norm": 1.2135330438613892,
+      "learning_rate": 8.41451353233369e-08,
+      "loss": 1.0765,
+      "step": 4197
+    },
+    {
+      "epoch": 0.9611906124785347,
+      "grad_norm": 1.5741900205612183,
+      "learning_rate": 8.318755595135952e-08,
+      "loss": 1.0657,
+      "step": 4198
+    },
+    {
+      "epoch": 0.9614195764167144,
+      "grad_norm": 1.8799227476119995,
+      "learning_rate": 8.223543371812748e-08,
+      "loss": 1.0397,
+      "step": 4199
+    },
+    {
+      "epoch": 0.9616485403548941,
+      "grad_norm": 2.1442928314208984,
+      "learning_rate": 8.128876914758499e-08,
+      "loss": 1.0042,
+      "step": 4200
+    },
+    {
+      "epoch": 0.9618775042930738,
+      "grad_norm": 1.2874436378479004,
+      "learning_rate": 8.034756276067534e-08,
+      "loss": 0.9996,
+      "step": 4201
+    },
+    {
+      "epoch": 0.9621064682312536,
+      "grad_norm": 1.346038579940796,
+      "learning_rate": 7.941181507533424e-08,
+      "loss": 1.0304,
+      "step": 4202
+    },
+    {
+      "epoch": 0.9623354321694333,
+      "grad_norm": 1.7882304191589355,
+      "learning_rate": 7.848152660649866e-08,
+      "loss": 1.0538,
+      "step": 4203
+    },
+    {
+      "epoch": 0.962564396107613,
+      "grad_norm": 1.237015724182129,
+      "learning_rate": 7.755669786609688e-08,
+      "loss": 1.0272,
+      "step": 4204
+    },
+    {
+      "epoch": 0.9627933600457927,
+      "grad_norm": 1.6121935844421387,
+      "learning_rate": 7.663732936305291e-08,
+      "loss": 1.015,
+      "step": 4205
+    },
+    {
+      "epoch": 0.9630223239839725,
+      "grad_norm": 1.267065405845642,
+      "learning_rate": 7.572342160328982e-08,
+      "loss": 1.0515,
+      "step": 4206
+    },
+    {
+      "epoch": 0.9632512879221523,
+      "grad_norm": 1.263258934020996,
+      "learning_rate": 7.481497508972313e-08,
+      "loss": 1.0207,
+      "step": 4207
+    },
+    {
+      "epoch": 0.963480251860332,
+      "grad_norm": 1.343643069267273,
+      "learning_rate": 7.391199032226182e-08,
+      "loss": 1.0493,
+      "step": 4208
+    },
+    {
+      "epoch": 0.9637092157985118,
+      "grad_norm": 1.558370590209961,
+      "learning_rate": 7.301446779781285e-08,
+      "loss": 1.0701,
+      "step": 4209
+    },
+    {
+      "epoch": 0.9639381797366915,
+      "grad_norm": 1.1900969743728638,
+      "learning_rate": 7.212240801027337e-08,
+      "loss": 0.9938,
+      "step": 4210
+    },
+    {
+      "epoch": 0.9641671436748712,
+      "grad_norm": 2.1308228969573975,
+      "learning_rate": 7.123581145053849e-08,
+      "loss": 1.0434,
+      "step": 4211
+    },
+    {
+      "epoch": 0.964396107613051,
+      "grad_norm": 1.5577213764190674,
+      "learning_rate": 7.035467860649348e-08,
+      "loss": 1.0161,
+      "step": 4212
+    },
+    {
+      "epoch": 0.9646250715512307,
+      "grad_norm": 1.1246519088745117,
+      "learning_rate": 6.947900996301826e-08,
+      "loss": 1.0164,
+      "step": 4213
+    },
+    {
+      "epoch": 0.9648540354894104,
+      "grad_norm": 1.6314594745635986,
+      "learning_rate": 6.860880600198627e-08,
+      "loss": 1.0539,
+      "step": 4214
+    },
+    {
+      "epoch": 0.9650829994275901,
+      "grad_norm": 1.6893818378448486,
+      "learning_rate": 6.774406720226335e-08,
+      "loss": 1.0481,
+      "step": 4215
+    },
+    {
+      "epoch": 0.9653119633657699,
+      "grad_norm": 1.3539258241653442,
+      "learning_rate": 6.688479403970883e-08,
+      "loss": 1.0231,
+      "step": 4216
+    },
+    {
+      "epoch": 0.9655409273039496,
+      "grad_norm": 1.230441927909851,
+      "learning_rate": 6.603098698717336e-08,
+      "loss": 1.0163,
+      "step": 4217
+    },
+    {
+      "epoch": 0.9657698912421294,
+      "grad_norm": 1.3217246532440186,
+      "learning_rate": 6.51826465144978e-08,
+      "loss": 0.98,
+      "step": 4218
+    },
+    {
+      "epoch": 0.9659988551803091,
+      "grad_norm": 1.1363701820373535,
+      "learning_rate": 6.433977308851869e-08,
+      "loss": 1.0446,
+      "step": 4219
+    },
+    {
+      "epoch": 0.9662278191184889,
+      "grad_norm": 1.5187417268753052,
+      "learning_rate": 6.35023671730628e-08,
+      "loss": 1.0349,
+      "step": 4220
+    },
+    {
+      "epoch": 0.9664567830566686,
+      "grad_norm": 1.2305715084075928,
+      "learning_rate": 6.267042922894595e-08,
+      "loss": 1.0425,
+      "step": 4221
+    },
+    {
+      "epoch": 0.9666857469948483,
+      "grad_norm": 1.2499189376831055,
+      "learning_rate": 6.18439597139775e-08,
+      "loss": 0.9435,
+      "step": 4222
+    },
+    {
+      "epoch": 0.966914710933028,
+      "grad_norm": 1.2281574010849,
+      "learning_rate": 6.102295908295585e-08,
+      "loss": 1.0268,
+      "step": 4223
+    },
+    {
+      "epoch": 0.9671436748712078,
+      "grad_norm": 1.3302663564682007,
+      "learning_rate": 6.020742778767185e-08,
+      "loss": 1.012,
+      "step": 4224
+    },
+    {
+      "epoch": 0.9673726388093875,
+      "grad_norm": 1.1993911266326904,
+      "learning_rate": 5.93973662769054e-08,
+      "loss": 1.0542,
+      "step": 4225
+    },
+    {
+      "epoch": 0.9676016027475672,
+      "grad_norm": 1.2014667987823486,
+      "learning_rate": 5.859277499642546e-08,
+      "loss": 0.9975,
+      "step": 4226
+    },
+    {
+      "epoch": 0.967830566685747,
+      "grad_norm": 1.501274585723877,
+      "learning_rate": 5.7793654388993426e-08,
+      "loss": 1.0324,
+      "step": 4227
+    },
+    {
+      "epoch": 0.9680595306239267,
+      "grad_norm": 1.7279783487319946,
+      "learning_rate": 5.700000489435753e-08,
+      "loss": 1.0626,
+      "step": 4228
+    },
+    {
+      "epoch": 0.9682884945621064,
+      "grad_norm": 1.46021568775177,
+      "learning_rate": 5.621182694925731e-08,
+      "loss": 1.0041,
+      "step": 4229
+    },
+    {
+      "epoch": 0.9685174585002863,
+      "grad_norm": 1.1859759092330933,
+      "learning_rate": 5.542912098741915e-08,
+      "loss": 1.0243,
+      "step": 4230
+    },
+    {
+      "epoch": 0.968746422438466,
+      "grad_norm": 1.5773413181304932,
+      "learning_rate": 5.465188743956073e-08,
+      "loss": 1.0268,
+      "step": 4231
+    },
+    {
+      "epoch": 0.9689753863766457,
+      "grad_norm": 1.4995766878128052,
+      "learning_rate": 5.388012673338661e-08,
+      "loss": 1.0252,
+      "step": 4232
+    },
+    {
+      "epoch": 0.9692043503148254,
+      "grad_norm": 1.340819239616394,
+      "learning_rate": 5.3113839293590374e-08,
+      "loss": 1.0199,
+      "step": 4233
+    },
+    {
+      "epoch": 0.9694333142530052,
+      "grad_norm": 1.144403338432312,
+      "learning_rate": 5.235302554185362e-08,
+      "loss": 0.9865,
+      "step": 4234
+    },
+    {
+      "epoch": 0.9696622781911849,
+      "grad_norm": 1.210654854774475,
+      "learning_rate": 5.159768589684699e-08,
+      "loss": 0.978,
+      "step": 4235
+    },
+    {
+      "epoch": 0.9698912421293646,
+      "grad_norm": 1.2872427701950073,
+      "learning_rate": 5.084782077422468e-08,
+      "loss": 1.0102,
+      "step": 4236
+    },
+    {
+      "epoch": 0.9701202060675443,
+      "grad_norm": 1.495758295059204,
+      "learning_rate": 5.010343058663325e-08,
+      "loss": 1.0315,
+      "step": 4237
+    },
+    {
+      "epoch": 0.9703491700057241,
+      "grad_norm": 1.3226579427719116,
+      "learning_rate": 4.9364515743705046e-08,
+      "loss": 1.0261,
+      "step": 4238
+    },
+    {
+      "epoch": 0.9705781339439038,
+      "grad_norm": 1.2678083181381226,
+      "learning_rate": 4.863107665205702e-08,
+      "loss": 1.0226,
+      "step": 4239
+    },
+    {
+      "epoch": 0.9708070978820835,
+      "grad_norm": 1.6307822465896606,
+      "learning_rate": 4.79031137152941e-08,
+      "loss": 1.0238,
+      "step": 4240
+    },
+    {
+      "epoch": 0.9710360618202634,
+      "grad_norm": 1.3601518869400024,
+      "learning_rate": 4.71806273340103e-08,
+      "loss": 0.9714,
+      "step": 4241
+    },
+    {
+      "epoch": 0.9712650257584431,
+      "grad_norm": 1.3477270603179932,
+      "learning_rate": 4.646361790578313e-08,
+      "loss": 1.0751,
+      "step": 4242
+    },
+    {
+      "epoch": 0.9714939896966228,
+      "grad_norm": 1.1436761617660522,
+      "learning_rate": 4.575208582517587e-08,
+      "loss": 1.0135,
+      "step": 4243
+    },
+    {
+      "epoch": 0.9717229536348025,
+      "grad_norm": 2.0135879516601562,
+      "learning_rate": 4.504603148373976e-08,
+      "loss": 0.9799,
+      "step": 4244
+    },
+    {
+      "epoch": 0.9719519175729823,
+      "grad_norm": 2.1591763496398926,
+      "learning_rate": 4.4345455270010665e-08,
+      "loss": 1.0147,
+      "step": 4245
+    },
+    {
+      "epoch": 0.972180881511162,
+      "grad_norm": 1.2410547733306885,
+      "learning_rate": 4.365035756950797e-08,
+      "loss": 1.0154,
+      "step": 4246
+    },
+    {
+      "epoch": 0.9724098454493417,
+      "grad_norm": 1.4873212575912476,
+      "learning_rate": 4.296073876474016e-08,
+      "loss": 1.0427,
+      "step": 4247
+    },
+    {
+      "epoch": 0.9726388093875215,
+      "grad_norm": 1.2260034084320068,
+      "learning_rate": 4.227659923519811e-08,
+      "loss": 1.0528,
+      "step": 4248
+    },
+    {
+      "epoch": 0.9728677733257012,
+      "grad_norm": 1.4164505004882812,
+      "learning_rate": 4.159793935735734e-08,
+      "loss": 1.0673,
+      "step": 4249
+    },
+    {
+      "epoch": 0.9730967372638809,
+      "grad_norm": 1.2214871644973755,
+      "learning_rate": 4.092475950468022e-08,
+      "loss": 1.0818,
+      "step": 4250
+    },
+    {
+      "epoch": 0.9733257012020606,
+      "grad_norm": 1.3067419528961182,
+      "learning_rate": 4.025706004760932e-08,
+      "loss": 0.9782,
+      "step": 4251
+    },
+    {
+      "epoch": 0.9735546651402404,
+      "grad_norm": 1.2343987226486206,
+      "learning_rate": 3.9594841353577384e-08,
+      "loss": 1.093,
+      "step": 4252
+    },
+    {
+      "epoch": 0.9737836290784202,
+      "grad_norm": 1.420095443725586,
+      "learning_rate": 3.8938103786995144e-08,
+      "loss": 1.0679,
+      "step": 4253
+    },
+    {
+      "epoch": 0.9740125930165999,
+      "grad_norm": 1.1239619255065918,
+      "learning_rate": 3.8286847709261276e-08,
+      "loss": 1.0286,
+      "step": 4254
+    },
+    {
+      "epoch": 0.9742415569547797,
+      "grad_norm": 1.2686529159545898,
+      "learning_rate": 3.7641073478755786e-08,
+      "loss": 0.9945,
+      "step": 4255
+    },
+    {
+      "epoch": 0.9744705208929594,
+      "grad_norm": 2.8842384815216064,
+      "learning_rate": 3.7000781450844405e-08,
+      "loss": 1.0213,
+      "step": 4256
+    },
+    {
+      "epoch": 0.9746994848311391,
+      "grad_norm": 1.8031467199325562,
+      "learning_rate": 3.636597197787084e-08,
+      "loss": 1.0022,
+      "step": 4257
+    },
+    {
+      "epoch": 0.9749284487693188,
+      "grad_norm": 1.7563955783843994,
+      "learning_rate": 3.573664540916899e-08,
+      "loss": 0.9997,
+      "step": 4258
+    },
+    {
+      "epoch": 0.9751574127074986,
+      "grad_norm": 2.2838706970214844,
+      "learning_rate": 3.5112802091051834e-08,
+      "loss": 1.0434,
+      "step": 4259
+    },
+    {
+      "epoch": 0.9753863766456783,
+      "grad_norm": 1.1979697942733765,
+      "learning_rate": 3.449444236681254e-08,
+      "loss": 1.0448,
+      "step": 4260
+    },
+    {
+      "epoch": 0.975615340583858,
+      "grad_norm": 1.2877243757247925,
+      "learning_rate": 3.388156657673114e-08,
+      "loss": 1.0362,
+      "step": 4261
+    },
+    {
+      "epoch": 0.9758443045220377,
+      "grad_norm": 1.5697972774505615,
+      "learning_rate": 3.327417505806785e-08,
+      "loss": 1.0681,
+      "step": 4262
+    },
+    {
+      "epoch": 0.9760732684602175,
+      "grad_norm": 1.3261923789978027,
+      "learning_rate": 3.267226814506419e-08,
+      "loss": 1.0489,
+      "step": 4263
+    },
+    {
+      "epoch": 0.9763022323983973,
+      "grad_norm": 1.2220418453216553,
+      "learning_rate": 3.2075846168946326e-08,
+      "loss": 1.0713,
+      "step": 4264
+    },
+    {
+      "epoch": 0.976531196336577,
+      "grad_norm": 1.7401745319366455,
+      "learning_rate": 3.148490945791838e-08,
+      "loss": 1.0271,
+      "step": 4265
+    },
+    {
+      "epoch": 0.9767601602747568,
+      "grad_norm": 1.2124825716018677,
+      "learning_rate": 3.089945833716912e-08,
+      "loss": 1.0562,
+      "step": 4266
+    },
+    {
+      "epoch": 0.9769891242129365,
+      "grad_norm": 1.3442974090576172,
+      "learning_rate": 3.03194931288664e-08,
+      "loss": 1.0586,
+      "step": 4267
+    },
+    {
+      "epoch": 0.9772180881511162,
+      "grad_norm": 1.2674909830093384,
+      "learning_rate": 2.9745014152161578e-08,
+      "loss": 0.9957,
+      "step": 4268
+    },
+    {
+      "epoch": 0.9774470520892959,
+      "grad_norm": 1.6897748708724976,
+      "learning_rate": 2.917602172318401e-08,
+      "loss": 0.9973,
+      "step": 4269
+    },
+    {
+      "epoch": 0.9776760160274757,
+      "grad_norm": 1.3805571794509888,
+      "learning_rate": 2.8612516155047674e-08,
+      "loss": 0.9844,
+      "step": 4270
+    },
+    {
+      "epoch": 0.9779049799656554,
+      "grad_norm": 1.2225171327590942,
+      "learning_rate": 2.8054497757842304e-08,
+      "loss": 1.034,
+      "step": 4271
+    },
+    {
+      "epoch": 0.9781339439038351,
+      "grad_norm": 1.321537733078003,
+      "learning_rate": 2.7501966838642258e-08,
+      "loss": 1.0053,
+      "step": 4272
+    },
+    {
+      "epoch": 0.9783629078420149,
+      "grad_norm": 1.2543835639953613,
+      "learning_rate": 2.695492370149988e-08,
+      "loss": 1.0385,
+      "step": 4273
+    },
+    {
+      "epoch": 0.9785918717801946,
+      "grad_norm": 1.189489483833313,
+      "learning_rate": 2.641336864744992e-08,
+      "loss": 0.9956,
+      "step": 4274
+    },
+    {
+      "epoch": 0.9788208357183743,
+      "grad_norm": 1.2529815435409546,
+      "learning_rate": 2.5877301974503998e-08,
+      "loss": 1.0218,
+      "step": 4275
+    },
+    {
+      "epoch": 0.9790497996565541,
+      "grad_norm": 1.217818260192871,
+      "learning_rate": 2.534672397765614e-08,
+      "loss": 0.9919,
+      "step": 4276
+    },
+    {
+      "epoch": 0.9792787635947339,
+      "grad_norm": 1.363022804260254,
+      "learning_rate": 2.482163494887724e-08,
+      "loss": 1.0145,
+      "step": 4277
+    },
+    {
+      "epoch": 0.9795077275329136,
+      "grad_norm": 1.5168522596359253,
+      "learning_rate": 2.430203517712171e-08,
+      "loss": 1.0786,
+      "step": 4278
+    },
+    {
+      "epoch": 0.9797366914710933,
+      "grad_norm": 1.2632395029067993,
+      "learning_rate": 2.3787924948319718e-08,
+      "loss": 1.023,
+      "step": 4279
+    },
+    {
+      "epoch": 0.979965655409273,
+      "grad_norm": 1.2734144926071167,
+      "learning_rate": 2.327930454538274e-08,
+      "loss": 1.0511,
+      "step": 4280
+    },
+    {
+      "epoch": 0.9801946193474528,
+      "grad_norm": 1.4147753715515137,
+      "learning_rate": 2.2776174248199114e-08,
+      "loss": 0.9804,
+      "step": 4281
+    },
+    {
+      "epoch": 0.9804235832856325,
+      "grad_norm": 1.5188604593276978,
+      "learning_rate": 2.227853433363736e-08,
+      "loss": 1.0049,
+      "step": 4282
+    },
+    {
+      "epoch": 0.9806525472238122,
+      "grad_norm": 1.0858154296875,
+      "learning_rate": 2.1786385075545093e-08,
+      "loss": 1.0509,
+      "step": 4283
+    },
+    {
+      "epoch": 0.980881511161992,
+      "grad_norm": 1.3833143711090088,
+      "learning_rate": 2.1299726744747896e-08,
+      "loss": 0.9949,
+      "step": 4284
+    },
+    {
+      "epoch": 0.9811104751001717,
+      "grad_norm": 1.2171403169631958,
+      "learning_rate": 2.0818559609049327e-08,
+      "loss": 1.04,
+      "step": 4285
+    },
+    {
+      "epoch": 0.9813394390383514,
+      "grad_norm": 1.468177080154419,
+      "learning_rate": 2.0342883933232027e-08,
+      "loss": 0.985,
+      "step": 4286
+    },
+    {
+      "epoch": 0.9815684029765313,
+      "grad_norm": 1.3929141759872437,
+      "learning_rate": 1.987269997905661e-08,
+      "loss": 0.9983,
+      "step": 4287
+    },
+    {
+      "epoch": 0.981797366914711,
+      "grad_norm": 1.1607084274291992,
+      "learning_rate": 1.9408008005260548e-08,
+      "loss": 1.0426,
+      "step": 4288
+    },
+    {
+      "epoch": 0.9820263308528907,
+      "grad_norm": 1.3756016492843628,
+      "learning_rate": 1.8948808267560405e-08,
+      "loss": 1.0324,
+      "step": 4289
+    },
+    {
+      "epoch": 0.9822552947910704,
+      "grad_norm": 1.2399051189422607,
+      "learning_rate": 1.8495101018649598e-08,
+      "loss": 0.994,
+      "step": 4290
+    },
+    {
+      "epoch": 0.9824842587292502,
+      "grad_norm": 1.5608595609664917,
+      "learning_rate": 1.8046886508200633e-08,
+      "loss": 1.0074,
+      "step": 4291
+    },
+    {
+      "epoch": 0.9827132226674299,
+      "grad_norm": 1.3537348508834839,
+      "learning_rate": 1.7604164982860662e-08,
+      "loss": 1.0612,
+      "step": 4292
+    },
+    {
+      "epoch": 0.9829421866056096,
+      "grad_norm": 1.5062490701675415,
+      "learning_rate": 1.716693668625591e-08,
+      "loss": 1.0275,
+      "step": 4293
+    },
+    {
+      "epoch": 0.9831711505437893,
+      "grad_norm": 1.5584570169448853,
+      "learning_rate": 1.673520185899058e-08,
+      "loss": 1.0148,
+      "step": 4294
+    },
+    {
+      "epoch": 0.9834001144819691,
+      "grad_norm": 1.1560457944869995,
+      "learning_rate": 1.630896073864352e-08,
+      "loss": 0.986,
+      "step": 4295
+    },
+    {
+      "epoch": 0.9836290784201488,
+      "grad_norm": 1.2290360927581787,
+      "learning_rate": 1.5888213559771548e-08,
+      "loss": 1.0749,
+      "step": 4296
+    },
+    {
+      "epoch": 0.9838580423583285,
+      "grad_norm": 1.4205414056777954,
+      "learning_rate": 1.5472960553909456e-08,
+      "loss": 0.9877,
+      "step": 4297
+    },
+    {
+      "epoch": 0.9840870062965082,
+      "grad_norm": 1.6074973344802856,
+      "learning_rate": 1.5063201949566674e-08,
+      "loss": 1.0362,
+      "step": 4298
+    },
+    {
+      "epoch": 0.9843159702346881,
+      "grad_norm": 1.3286057710647583,
+      "learning_rate": 1.4658937972230613e-08,
+      "loss": 0.9793,
+      "step": 4299
+    },
+    {
+      "epoch": 0.9845449341728678,
+      "grad_norm": 1.3167859315872192,
+      "learning_rate": 1.426016884436332e-08,
+      "loss": 1.039,
+      "step": 4300
+    },
+    {
+      "epoch": 0.9847738981110475,
+      "grad_norm": 1.1874357461929321,
+      "learning_rate": 1.3866894785404816e-08,
+      "loss": 0.9651,
+      "step": 4301
+    },
+    {
+      "epoch": 0.9850028620492273,
+      "grad_norm": 1.4577040672302246,
+      "learning_rate": 1.3479116011769766e-08,
+      "loss": 1.0622,
+      "step": 4302
+    },
+    {
+      "epoch": 0.985231825987407,
+      "grad_norm": 1.3572076559066772,
+      "learning_rate": 1.3096832736850806e-08,
+      "loss": 1.0132,
+      "step": 4303
+    },
+    {
+      "epoch": 0.9854607899255867,
+      "grad_norm": 1.22506844997406,
+      "learning_rate": 1.2720045171014106e-08,
+      "loss": 1.0378,
+      "step": 4304
+    },
+    {
+      "epoch": 0.9856897538637664,
+      "grad_norm": 1.5919448137283325,
+      "learning_rate": 1.2348753521602696e-08,
+      "loss": 1.0365,
+      "step": 4305
+    },
+    {
+      "epoch": 0.9859187178019462,
+      "grad_norm": 1.5855189561843872,
+      "learning_rate": 1.1982957992936472e-08,
+      "loss": 1.0479,
+      "step": 4306
+    },
+    {
+      "epoch": 0.9861476817401259,
+      "grad_norm": 1.5232701301574707,
+      "learning_rate": 1.162265878630886e-08,
+      "loss": 1.0253,
+      "step": 4307
+    },
+    {
+      "epoch": 0.9863766456783056,
+      "grad_norm": 1.51606023311615,
+      "learning_rate": 1.1267856099989039e-08,
+      "loss": 0.9984,
+      "step": 4308
+    },
+    {
+      "epoch": 0.9866056096164854,
+      "grad_norm": 1.5284004211425781,
+      "learning_rate": 1.0918550129223049e-08,
+      "loss": 1.0555,
+      "step": 4309
+    },
+    {
+      "epoch": 0.9868345735546652,
+      "grad_norm": 1.2388665676116943,
+      "learning_rate": 1.0574741066230465e-08,
+      "loss": 0.9951,
+      "step": 4310
+    },
+    {
+      "epoch": 0.9870635374928449,
+      "grad_norm": 1.161252737045288,
+      "learning_rate": 1.0236429100206612e-08,
+      "loss": 1.0634,
+      "step": 4311
+    },
+    {
+      "epoch": 0.9872925014310246,
+      "grad_norm": 1.1357179880142212,
+      "learning_rate": 9.903614417320351e-09,
+      "loss": 1.0417,
+      "step": 4312
+    },
+    {
+      "epoch": 0.9875214653692044,
+      "grad_norm": 1.3556911945343018,
+      "learning_rate": 9.57629720071962e-09,
+      "loss": 1.0326,
+      "step": 4313
+    },
+    {
+      "epoch": 0.9877504293073841,
+      "grad_norm": 1.2056077718734741,
+      "learning_rate": 9.254477630521452e-09,
+      "loss": 0.945,
+      "step": 4314
+    },
+    {
+      "epoch": 0.9879793932455638,
+      "grad_norm": 1.208575963973999,
+      "learning_rate": 8.938155883823074e-09,
+      "loss": 1.0396,
+      "step": 4315
+    },
+    {
+      "epoch": 0.9882083571837436,
+      "grad_norm": 1.3659850358963013,
+      "learning_rate": 8.627332134690802e-09,
+      "loss": 1.0029,
+      "step": 4316
+    },
+    {
+      "epoch": 0.9884373211219233,
+      "grad_norm": 1.363781213760376,
+      "learning_rate": 8.322006554171147e-09,
+      "loss": 1.0407,
+      "step": 4317
+    },
+    {
+      "epoch": 0.988666285060103,
+      "grad_norm": 1.3712806701660156,
+      "learning_rate": 8.02217931028082e-09,
+      "loss": 0.9999,
+      "step": 4318
+    },
+    {
+      "epoch": 0.9888952489982827,
+      "grad_norm": 1.421749234199524,
+      "learning_rate": 7.727850568012286e-09,
+      "loss": 1.0482,
+      "step": 4319
+    },
+    {
+      "epoch": 0.9891242129364625,
+      "grad_norm": 1.1671499013900757,
+      "learning_rate": 7.439020489332649e-09,
+      "loss": 1.044,
+      "step": 4320
+    },
+    {
+      "epoch": 0.9893531768746422,
+      "grad_norm": 1.8170863389968872,
+      "learning_rate": 7.1556892331814394e-09,
+      "loss": 0.9375,
+      "step": 4321
+    },
+    {
+      "epoch": 0.989582140812822,
+      "grad_norm": 1.3475590944290161,
+      "learning_rate": 6.8778569554750484e-09,
+      "loss": 0.9789,
+      "step": 4322
+    },
+    {
+      "epoch": 0.9898111047510018,
+      "grad_norm": 1.19892418384552,
+      "learning_rate": 6.605523809102288e-09,
+      "loss": 1.0129,
+      "step": 4323
+    },
+    {
+      "epoch": 0.9900400686891815,
+      "grad_norm": 1.4870357513427734,
+      "learning_rate": 6.3386899439243925e-09,
+      "loss": 1.0412,
+      "step": 4324
+    },
+    {
+      "epoch": 0.9902690326273612,
+      "grad_norm": 1.529685378074646,
+      "learning_rate": 6.0773555067783485e-09,
+      "loss": 1.0188,
+      "step": 4325
+    },
+    {
+      "epoch": 0.9904979965655409,
+      "grad_norm": 1.2781044244766235,
+      "learning_rate": 5.8215206414746764e-09,
+      "loss": 1.0064,
+      "step": 4326
+    },
+    {
+      "epoch": 0.9907269605037207,
+      "grad_norm": 1.5567418336868286,
+      "learning_rate": 5.571185488797426e-09,
+      "loss": 1.0069,
+      "step": 4327
+    },
+    {
+      "epoch": 0.9909559244419004,
+      "grad_norm": 1.4737282991409302,
+      "learning_rate": 5.326350186503071e-09,
+      "loss": 1.0737,
+      "step": 4328
+    },
+    {
+      "epoch": 0.9911848883800801,
+      "grad_norm": 1.3786917924880981,
+      "learning_rate": 5.087014869322726e-09,
+      "loss": 0.991,
+      "step": 4329
+    },
+    {
+      "epoch": 0.9914138523182598,
+      "grad_norm": 1.3202197551727295,
+      "learning_rate": 4.853179668959928e-09,
+      "loss": 1.0897,
+      "step": 4330
+    },
+    {
+      "epoch": 0.9916428162564396,
+      "grad_norm": 1.275343656539917,
+      "learning_rate": 4.6248447140939675e-09,
+      "loss": 1.0553,
+      "step": 4331
+    },
+    {
+      "epoch": 0.9918717801946193,
+      "grad_norm": 1.365133285522461,
+      "learning_rate": 4.4020101303743345e-09,
+      "loss": 1.024,
+      "step": 4332
+    },
+    {
+      "epoch": 0.9921007441327991,
+      "grad_norm": 1.2677979469299316,
+      "learning_rate": 4.184676040426272e-09,
+      "loss": 1.0297,
+      "step": 4333
+    },
+    {
+      "epoch": 0.9923297080709789,
+      "grad_norm": 1.360640525817871,
+      "learning_rate": 3.972842563845225e-09,
+      "loss": 1.0112,
+      "step": 4334
+    },
+    {
+      "epoch": 0.9925586720091586,
+      "grad_norm": 1.6422892808914185,
+      "learning_rate": 3.7665098172023905e-09,
+      "loss": 1.023,
+      "step": 4335
+    },
+    {
+      "epoch": 0.9927876359473383,
+      "grad_norm": 1.0084567070007324,
+      "learning_rate": 3.5656779140402777e-09,
+      "loss": 0.974,
+      "step": 4336
+    },
+    {
+      "epoch": 0.993016599885518,
+      "grad_norm": 1.3576377630233765,
+      "learning_rate": 3.3703469648760367e-09,
+      "loss": 0.9808,
+      "step": 4337
+    },
+    {
+      "epoch": 0.9932455638236978,
+      "grad_norm": 1.804397463798523,
+      "learning_rate": 3.1805170771970207e-09,
+      "loss": 1.0415,
+      "step": 4338
+    },
+    {
+      "epoch": 0.9934745277618775,
+      "grad_norm": 1.3295351266860962,
+      "learning_rate": 2.9961883554674443e-09,
+      "loss": 1.0812,
+      "step": 4339
+    },
+    {
+      "epoch": 0.9937034917000572,
+      "grad_norm": 2.1188831329345703,
+      "learning_rate": 2.8173609011195035e-09,
+      "loss": 0.964,
+      "step": 4340
+    },
+    {
+      "epoch": 0.993932455638237,
+      "grad_norm": 1.1726047992706299,
+      "learning_rate": 2.6440348125622574e-09,
+      "loss": 0.9524,
+      "step": 4341
+    },
+    {
+      "epoch": 0.9941614195764167,
+      "grad_norm": 1.23749840259552,
+      "learning_rate": 2.476210185173855e-09,
+      "loss": 0.979,
+      "step": 4342
+    },
+    {
+      "epoch": 0.9943903835145964,
+      "grad_norm": 1.6526784896850586,
+      "learning_rate": 2.3138871113081997e-09,
+      "loss": 1.0302,
+      "step": 4343
+    },
+    {
+      "epoch": 0.9946193474527761,
+      "grad_norm": 1.434206485748291,
+      "learning_rate": 2.1570656802905042e-09,
+      "loss": 1.0205,
+      "step": 4344
+    },
+    {
+      "epoch": 0.994848311390956,
+      "grad_norm": 1.5796558856964111,
+      "learning_rate": 2.0057459784161848e-09,
+      "loss": 1.0317,
+      "step": 4345
+    },
+    {
+      "epoch": 0.9950772753291357,
+      "grad_norm": 1.5953503847122192,
+      "learning_rate": 1.859928088957519e-09,
+      "loss": 1.0114,
+      "step": 4346
+    },
+    {
+      "epoch": 0.9953062392673154,
+      "grad_norm": 1.3271796703338623,
+      "learning_rate": 1.7196120921558757e-09,
+      "loss": 1.0554,
+      "step": 4347
+    },
+    {
+      "epoch": 0.9955352032054952,
+      "grad_norm": 3.4375476837158203,
+      "learning_rate": 1.5847980652261563e-09,
+      "loss": 1.0024,
+      "step": 4348
+    },
+    {
+      "epoch": 0.9957641671436749,
+      "grad_norm": 1.3996182680130005,
+      "learning_rate": 1.4554860823556838e-09,
+      "loss": 1.0361,
+      "step": 4349
+    },
+    {
+      "epoch": 0.9959931310818546,
+      "grad_norm": 1.1781070232391357,
+      "learning_rate": 1.3316762147030925e-09,
+      "loss": 1.0539,
+      "step": 4350
+    },
+    {
+      "epoch": 0.9962220950200343,
+      "grad_norm": 1.5741220712661743,
+      "learning_rate": 1.213368530399439e-09,
+      "loss": 0.9569,
+      "step": 4351
+    },
+    {
+      "epoch": 0.9964510589582141,
+      "grad_norm": 1.3308751583099365,
+      "learning_rate": 1.100563094550422e-09,
+      "loss": 0.9695,
+      "step": 4352
+    },
+    {
+      "epoch": 0.9966800228963938,
+      "grad_norm": 1.3270615339279175,
+      "learning_rate": 9.932599692297207e-10,
+      "loss": 1.0147,
+      "step": 4353
+    },
+    {
+      "epoch": 0.9969089868345735,
+      "grad_norm": 1.6848193407058716,
+      "learning_rate": 8.914592134867672e-10,
+      "loss": 0.9697,
+      "step": 4354
+    },
+    {
+      "epoch": 0.9971379507727532,
+      "grad_norm": 1.465806245803833,
+      "learning_rate": 7.95160883341195e-10,
+      "loss": 0.9577,
+      "step": 4355
+    },
+    {
+      "epoch": 0.9973669147109331,
+      "grad_norm": 1.743776798248291,
+      "learning_rate": 7.043650317850592e-10,
+      "loss": 1.0717,
+      "step": 4356
+    },
+    {
+      "epoch": 0.9975958786491128,
+      "grad_norm": 1.4670666456222534,
+      "learning_rate": 6.190717087828368e-10,
+      "loss": 0.9676,
+      "step": 4357
+    },
+    {
+      "epoch": 0.9978248425872925,
+      "grad_norm": 1.3044846057891846,
+      "learning_rate": 5.392809612703165e-10,
+      "loss": 1.0857,
+      "step": 4358
+    },
+    {
+      "epoch": 0.9980538065254723,
+      "grad_norm": 1.9425315856933594,
+      "learning_rate": 4.649928331557085e-10,
+      "loss": 1.061,
+      "step": 4359
+    },
+    {
+      "epoch": 0.998282770463652,
+      "grad_norm": 1.157950520515442,
+      "learning_rate": 3.9620736532075543e-10,
+      "loss": 1.0149,
+      "step": 4360
+    },
+    {
+      "epoch": 0.9985117344018317,
+      "grad_norm": 1.157373309135437,
+      "learning_rate": 3.3292459561518055e-10,
+      "loss": 1.0196,
+      "step": 4361
+    },
+    {
+      "epoch": 0.9987406983400114,
+      "grad_norm": 1.302017331123352,
+      "learning_rate": 2.7514455886334945e-10,
+      "loss": 1.0031,
+      "step": 4362
+    },
+    {
+      "epoch": 0.9989696622781912,
+      "grad_norm": 1.2691115140914917,
+      "learning_rate": 2.2286728686315984e-10,
+      "loss": 1.0935,
+      "step": 4363
+    },
+    {
+      "epoch": 0.9991986262163709,
+      "grad_norm": 1.2917944192886353,
+      "learning_rate": 1.7609280838049026e-10,
+      "loss": 1.0505,
+      "step": 4364
+    },
+    {
+      "epoch": 0.9994275901545506,
+      "grad_norm": 2.0502283573150635,
+      "learning_rate": 1.3482114915475132e-10,
+      "loss": 1.0274,
+      "step": 4365
+    },
+    {
+      "epoch": 0.9996565540927304,
+      "grad_norm": 1.4327635765075684,
+      "learning_rate": 9.905233189888563e-11,
+      "loss": 1.0067,
+      "step": 4366
+    },
+    {
+      "epoch": 0.9998855180309101,
+      "grad_norm": 1.210883617401123,
+      "learning_rate": 6.87863762938168e-11,
+      "loss": 1.04,
+      "step": 4367
+    },
+    {
+      "epoch": 0.9998855180309101,
+      "step": 4367,
+      "total_flos": 2.9591670167426826e+18,
+      "train_loss": 1.0860715442400435,
+      "train_runtime": 124901.2572,
+      "train_samples_per_second": 4.476,
+      "train_steps_per_second": 0.035
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 4367,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50000,
+  "total_flos": 2.9591670167426826e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}