robometer_board_insertion_full / trainer_state.json
chomeed's picture
Upload folder using huggingface_hub
a8fe2c7 verified
Raw
History Blame Contribute Delete
175 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.018475750577367205,
"eval_steps": 100,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.8475750577367205e-05,
"grad_norm": 122.84947204589844,
"learning_rate": 0.0,
"loss": 2.4459,
"step": 1
},
{
"epoch": 3.695150115473441e-05,
"grad_norm": 89.25559997558594,
"learning_rate": 2.0000000000000002e-07,
"loss": 1.3932,
"step": 2
},
{
"epoch": 5.542725173210162e-05,
"grad_norm": 32.383846282958984,
"learning_rate": 4.0000000000000003e-07,
"loss": 0.5693,
"step": 3
},
{
"epoch": 7.390300230946882e-05,
"grad_norm": 85.57299041748047,
"learning_rate": 6.000000000000001e-07,
"loss": 0.8119,
"step": 4
},
{
"epoch": 9.237875288683603e-05,
"grad_norm": 52.27479934692383,
"learning_rate": 8.000000000000001e-07,
"loss": 0.6569,
"step": 5
},
{
"epoch": 0.00011085450346420324,
"grad_norm": 45.68242263793945,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.7998,
"step": 6
},
{
"epoch": 0.00012933025404157045,
"grad_norm": 60.590087890625,
"learning_rate": 1.2000000000000002e-06,
"loss": 0.5049,
"step": 7
},
{
"epoch": 0.00014780600461893764,
"grad_norm": 145.44956970214844,
"learning_rate": 1.4000000000000001e-06,
"loss": 0.5006,
"step": 8
},
{
"epoch": 0.00016628175519630485,
"grad_norm": 29.79498863220215,
"learning_rate": 1.6000000000000001e-06,
"loss": 0.4231,
"step": 9
},
{
"epoch": 0.00018475750577367207,
"grad_norm": 75.24263763427734,
"learning_rate": 1.8000000000000001e-06,
"loss": 0.4468,
"step": 10
},
{
"epoch": 0.00020323325635103926,
"grad_norm": 92.08013153076172,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.2879,
"step": 11
},
{
"epoch": 0.00022170900692840647,
"grad_norm": 154.86256408691406,
"learning_rate": 2.2e-06,
"loss": 1.3566,
"step": 12
},
{
"epoch": 0.00024018475750577366,
"grad_norm": 83.2785415649414,
"learning_rate": 2.4000000000000003e-06,
"loss": 0.6362,
"step": 13
},
{
"epoch": 0.0002586605080831409,
"grad_norm": 128.80157470703125,
"learning_rate": 2.6e-06,
"loss": 1.5552,
"step": 14
},
{
"epoch": 0.0002771362586605081,
"grad_norm": 61.4452018737793,
"learning_rate": 2.8000000000000003e-06,
"loss": 0.7372,
"step": 15
},
{
"epoch": 0.0002956120092378753,
"grad_norm": 56.57251739501953,
"learning_rate": 3e-06,
"loss": 0.9798,
"step": 16
},
{
"epoch": 0.0003140877598152425,
"grad_norm": 61.806644439697266,
"learning_rate": 3.2000000000000003e-06,
"loss": 0.8501,
"step": 17
},
{
"epoch": 0.0003325635103926097,
"grad_norm": 106.70343780517578,
"learning_rate": 3.4000000000000005e-06,
"loss": 0.838,
"step": 18
},
{
"epoch": 0.0003510392609699769,
"grad_norm": 149.60189819335938,
"learning_rate": 3.6000000000000003e-06,
"loss": 0.9851,
"step": 19
},
{
"epoch": 0.00036951501154734414,
"grad_norm": 39.54475784301758,
"learning_rate": 3.8000000000000005e-06,
"loss": 0.3503,
"step": 20
},
{
"epoch": 0.0003879907621247113,
"grad_norm": 53.08124923706055,
"learning_rate": 4.000000000000001e-06,
"loss": 0.6706,
"step": 21
},
{
"epoch": 0.0004064665127020785,
"grad_norm": 23.596786499023438,
"learning_rate": 4.2000000000000004e-06,
"loss": 0.3469,
"step": 22
},
{
"epoch": 0.0004249422632794457,
"grad_norm": 59.31620788574219,
"learning_rate": 4.4e-06,
"loss": 0.6243,
"step": 23
},
{
"epoch": 0.00044341801385681294,
"grad_norm": 54.25373840332031,
"learning_rate": 4.600000000000001e-06,
"loss": 0.7036,
"step": 24
},
{
"epoch": 0.00046189376443418013,
"grad_norm": 32.10572814941406,
"learning_rate": 4.800000000000001e-06,
"loss": 0.3416,
"step": 25
},
{
"epoch": 0.0004803695150115473,
"grad_norm": 36.74370193481445,
"learning_rate": 5e-06,
"loss": 0.4767,
"step": 26
},
{
"epoch": 0.0004988452655889145,
"grad_norm": 49.238685607910156,
"learning_rate": 5.2e-06,
"loss": 0.991,
"step": 27
},
{
"epoch": 0.0005173210161662818,
"grad_norm": 31.672143936157227,
"learning_rate": 5.400000000000001e-06,
"loss": 0.4041,
"step": 28
},
{
"epoch": 0.000535796766743649,
"grad_norm": 22.80809211730957,
"learning_rate": 5.600000000000001e-06,
"loss": 0.3008,
"step": 29
},
{
"epoch": 0.0005542725173210162,
"grad_norm": 44.23014450073242,
"learning_rate": 5.8e-06,
"loss": 0.646,
"step": 30
},
{
"epoch": 0.0005727482678983834,
"grad_norm": 24.55451774597168,
"learning_rate": 6e-06,
"loss": 0.3296,
"step": 31
},
{
"epoch": 0.0005912240184757506,
"grad_norm": 25.76312255859375,
"learning_rate": 6.200000000000001e-06,
"loss": 0.3253,
"step": 32
},
{
"epoch": 0.0006096997690531177,
"grad_norm": 64.11470794677734,
"learning_rate": 6.4000000000000006e-06,
"loss": 0.6254,
"step": 33
},
{
"epoch": 0.000628175519630485,
"grad_norm": 13.86672306060791,
"learning_rate": 6.600000000000001e-06,
"loss": 0.2201,
"step": 34
},
{
"epoch": 0.0006466512702078522,
"grad_norm": 20.064128875732422,
"learning_rate": 6.800000000000001e-06,
"loss": 0.3149,
"step": 35
},
{
"epoch": 0.0006651270207852194,
"grad_norm": 6.012679576873779,
"learning_rate": 7e-06,
"loss": 0.1766,
"step": 36
},
{
"epoch": 0.0006836027713625866,
"grad_norm": 36.80343246459961,
"learning_rate": 7.2000000000000005e-06,
"loss": 0.2553,
"step": 37
},
{
"epoch": 0.0007020785219399538,
"grad_norm": 44.268924713134766,
"learning_rate": 7.4e-06,
"loss": 0.3949,
"step": 38
},
{
"epoch": 0.000720554272517321,
"grad_norm": 45.873077392578125,
"learning_rate": 7.600000000000001e-06,
"loss": 0.3814,
"step": 39
},
{
"epoch": 0.0007390300230946883,
"grad_norm": 28.359176635742188,
"learning_rate": 7.800000000000002e-06,
"loss": 0.3959,
"step": 40
},
{
"epoch": 0.0007575057736720555,
"grad_norm": 11.526424407958984,
"learning_rate": 8.000000000000001e-06,
"loss": 0.1898,
"step": 41
},
{
"epoch": 0.0007759815242494227,
"grad_norm": 6.4640793800354,
"learning_rate": 8.2e-06,
"loss": 0.1472,
"step": 42
},
{
"epoch": 0.0007944572748267898,
"grad_norm": 6.329624176025391,
"learning_rate": 8.400000000000001e-06,
"loss": 0.1465,
"step": 43
},
{
"epoch": 0.000812933025404157,
"grad_norm": 27.073734283447266,
"learning_rate": 8.6e-06,
"loss": 0.2869,
"step": 44
},
{
"epoch": 0.0008314087759815242,
"grad_norm": 28.271617889404297,
"learning_rate": 8.8e-06,
"loss": 0.2686,
"step": 45
},
{
"epoch": 0.0008498845265588914,
"grad_norm": 43.489078521728516,
"learning_rate": 9e-06,
"loss": 0.305,
"step": 46
},
{
"epoch": 0.0008683602771362587,
"grad_norm": 28.167417526245117,
"learning_rate": 9.200000000000002e-06,
"loss": 0.3143,
"step": 47
},
{
"epoch": 0.0008868360277136259,
"grad_norm": 7.627122402191162,
"learning_rate": 9.4e-06,
"loss": 0.1353,
"step": 48
},
{
"epoch": 0.0009053117782909931,
"grad_norm": 44.85990524291992,
"learning_rate": 9.600000000000001e-06,
"loss": 0.5191,
"step": 49
},
{
"epoch": 0.0009237875288683603,
"grad_norm": 10.548598289489746,
"learning_rate": 9.800000000000001e-06,
"loss": 0.1422,
"step": 50
},
{
"epoch": 0.0009422632794457275,
"grad_norm": 42.323081970214844,
"learning_rate": 1e-05,
"loss": 0.4841,
"step": 51
},
{
"epoch": 0.0009607390300230946,
"grad_norm": 77.28350830078125,
"learning_rate": 1.02e-05,
"loss": 0.9745,
"step": 52
},
{
"epoch": 0.0009792147806004618,
"grad_norm": 33.764469146728516,
"learning_rate": 1.04e-05,
"loss": 0.3277,
"step": 53
},
{
"epoch": 0.000997690531177829,
"grad_norm": 9.092707633972168,
"learning_rate": 1.0600000000000002e-05,
"loss": 0.1933,
"step": 54
},
{
"epoch": 0.0010161662817551962,
"grad_norm": 38.30156707763672,
"learning_rate": 1.0800000000000002e-05,
"loss": 0.177,
"step": 55
},
{
"epoch": 0.0010346420323325636,
"grad_norm": 6.40720796585083,
"learning_rate": 1.1000000000000001e-05,
"loss": 0.1274,
"step": 56
},
{
"epoch": 0.0010531177829099308,
"grad_norm": 3.1529664993286133,
"learning_rate": 1.1200000000000001e-05,
"loss": 0.1084,
"step": 57
},
{
"epoch": 0.001071593533487298,
"grad_norm": 69.79790496826172,
"learning_rate": 1.14e-05,
"loss": 0.6542,
"step": 58
},
{
"epoch": 0.0010900692840646652,
"grad_norm": 3.393667459487915,
"learning_rate": 1.16e-05,
"loss": 0.1218,
"step": 59
},
{
"epoch": 0.0011085450346420324,
"grad_norm": 43.04346466064453,
"learning_rate": 1.18e-05,
"loss": 0.2968,
"step": 60
},
{
"epoch": 0.0011270207852193995,
"grad_norm": 59.35041427612305,
"learning_rate": 1.2e-05,
"loss": 1.0084,
"step": 61
},
{
"epoch": 0.0011454965357967667,
"grad_norm": 34.094913482666016,
"learning_rate": 1.22e-05,
"loss": 0.1563,
"step": 62
},
{
"epoch": 0.001163972286374134,
"grad_norm": 6.941245079040527,
"learning_rate": 1.2400000000000002e-05,
"loss": 0.1325,
"step": 63
},
{
"epoch": 0.0011824480369515011,
"grad_norm": 39.17277526855469,
"learning_rate": 1.2600000000000001e-05,
"loss": 0.3135,
"step": 64
},
{
"epoch": 0.0012009237875288683,
"grad_norm": 40.0549201965332,
"learning_rate": 1.2800000000000001e-05,
"loss": 0.4354,
"step": 65
},
{
"epoch": 0.0012193995381062355,
"grad_norm": 16.4144287109375,
"learning_rate": 1.3000000000000001e-05,
"loss": 0.1815,
"step": 66
},
{
"epoch": 0.0012378752886836027,
"grad_norm": 20.971471786499023,
"learning_rate": 1.3200000000000002e-05,
"loss": 0.1807,
"step": 67
},
{
"epoch": 0.00125635103926097,
"grad_norm": 13.186601638793945,
"learning_rate": 1.3400000000000002e-05,
"loss": 0.157,
"step": 68
},
{
"epoch": 0.0012748267898383373,
"grad_norm": 47.6711540222168,
"learning_rate": 1.3600000000000002e-05,
"loss": 1.0546,
"step": 69
},
{
"epoch": 0.0012933025404157045,
"grad_norm": 11.16225814819336,
"learning_rate": 1.38e-05,
"loss": 0.1364,
"step": 70
},
{
"epoch": 0.0013117782909930716,
"grad_norm": 72.168212890625,
"learning_rate": 1.4e-05,
"loss": 0.5197,
"step": 71
},
{
"epoch": 0.0013302540415704388,
"grad_norm": 38.61338424682617,
"learning_rate": 1.4200000000000001e-05,
"loss": 0.485,
"step": 72
},
{
"epoch": 0.001348729792147806,
"grad_norm": 3.622647762298584,
"learning_rate": 1.4400000000000001e-05,
"loss": 0.11,
"step": 73
},
{
"epoch": 0.0013672055427251732,
"grad_norm": 22.882038116455078,
"learning_rate": 1.46e-05,
"loss": 0.3079,
"step": 74
},
{
"epoch": 0.0013856812933025404,
"grad_norm": 17.423635482788086,
"learning_rate": 1.48e-05,
"loss": 0.2099,
"step": 75
},
{
"epoch": 0.0014041570438799076,
"grad_norm": 7.647078990936279,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.1603,
"step": 76
},
{
"epoch": 0.0014226327944572748,
"grad_norm": 15.423291206359863,
"learning_rate": 1.5200000000000002e-05,
"loss": 0.1898,
"step": 77
},
{
"epoch": 0.001441108545034642,
"grad_norm": 74.05909729003906,
"learning_rate": 1.54e-05,
"loss": 0.639,
"step": 78
},
{
"epoch": 0.0014595842956120091,
"grad_norm": 20.598119735717773,
"learning_rate": 1.5600000000000003e-05,
"loss": 0.1892,
"step": 79
},
{
"epoch": 0.0014780600461893765,
"grad_norm": 44.88022994995117,
"learning_rate": 1.58e-05,
"loss": 0.4972,
"step": 80
},
{
"epoch": 0.0014965357967667437,
"grad_norm": 25.097993850708008,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.2687,
"step": 81
},
{
"epoch": 0.001515011547344111,
"grad_norm": 4.49970817565918,
"learning_rate": 1.62e-05,
"loss": 0.1333,
"step": 82
},
{
"epoch": 0.0015334872979214781,
"grad_norm": 17.777664184570312,
"learning_rate": 1.64e-05,
"loss": 0.1329,
"step": 83
},
{
"epoch": 0.0015519630484988453,
"grad_norm": 3.763643264770508,
"learning_rate": 1.66e-05,
"loss": 0.1159,
"step": 84
},
{
"epoch": 0.0015704387990762125,
"grad_norm": 22.17643165588379,
"learning_rate": 1.6800000000000002e-05,
"loss": 0.7322,
"step": 85
},
{
"epoch": 0.0015889145496535797,
"grad_norm": 7.4696574211120605,
"learning_rate": 1.7e-05,
"loss": 0.1526,
"step": 86
},
{
"epoch": 0.0016073903002309469,
"grad_norm": 3.037062168121338,
"learning_rate": 1.72e-05,
"loss": 0.1129,
"step": 87
},
{
"epoch": 0.001625866050808314,
"grad_norm": 103.2507553100586,
"learning_rate": 1.7400000000000003e-05,
"loss": 0.5697,
"step": 88
},
{
"epoch": 0.0016443418013856812,
"grad_norm": 21.007707595825195,
"learning_rate": 1.76e-05,
"loss": 0.1509,
"step": 89
},
{
"epoch": 0.0016628175519630484,
"grad_norm": 7.841665744781494,
"learning_rate": 1.7800000000000002e-05,
"loss": 0.1318,
"step": 90
},
{
"epoch": 0.0016812933025404156,
"grad_norm": 160.35418701171875,
"learning_rate": 1.8e-05,
"loss": 1.0556,
"step": 91
},
{
"epoch": 0.0016997690531177828,
"grad_norm": 1.5882081985473633,
"learning_rate": 1.8200000000000002e-05,
"loss": 0.0929,
"step": 92
},
{
"epoch": 0.0017182448036951502,
"grad_norm": 32.12841796875,
"learning_rate": 1.8400000000000003e-05,
"loss": 0.8397,
"step": 93
},
{
"epoch": 0.0017367205542725174,
"grad_norm": 29.697355270385742,
"learning_rate": 1.86e-05,
"loss": 0.7146,
"step": 94
},
{
"epoch": 0.0017551963048498846,
"grad_norm": 4.277217864990234,
"learning_rate": 1.88e-05,
"loss": 0.1319,
"step": 95
},
{
"epoch": 0.0017736720554272518,
"grad_norm": 46.174598693847656,
"learning_rate": 1.9e-05,
"loss": 0.2378,
"step": 96
},
{
"epoch": 0.001792147806004619,
"grad_norm": 30.646949768066406,
"learning_rate": 1.9200000000000003e-05,
"loss": 0.4311,
"step": 97
},
{
"epoch": 0.0018106235565819861,
"grad_norm": 57.86735916137695,
"learning_rate": 1.94e-05,
"loss": 0.3192,
"step": 98
},
{
"epoch": 0.0018290993071593533,
"grad_norm": 41.18842315673828,
"learning_rate": 1.9600000000000002e-05,
"loss": 0.3567,
"step": 99
},
{
"epoch": 0.0018475750577367205,
"grad_norm": 50.436092376708984,
"learning_rate": 1.98e-05,
"loss": 0.3053,
"step": 100
},
{
"epoch": 0.0018660508083140877,
"grad_norm": 66.22702026367188,
"learning_rate": 2e-05,
"loss": 0.5128,
"step": 101
},
{
"epoch": 0.001884526558891455,
"grad_norm": 26.307323455810547,
"learning_rate": 1.9999939076577906e-05,
"loss": 0.4841,
"step": 102
},
{
"epoch": 0.001903002309468822,
"grad_norm": 3.2372958660125732,
"learning_rate": 1.9999756307053947e-05,
"loss": 0.1169,
"step": 103
},
{
"epoch": 0.0019214780600461893,
"grad_norm": 25.138364791870117,
"learning_rate": 1.9999451693655125e-05,
"loss": 0.2984,
"step": 104
},
{
"epoch": 0.0019399538106235567,
"grad_norm": 6.459007740020752,
"learning_rate": 1.9999025240093045e-05,
"loss": 0.1273,
"step": 105
},
{
"epoch": 0.0019584295612009237,
"grad_norm": 36.724830627441406,
"learning_rate": 1.9998476951563914e-05,
"loss": 0.4494,
"step": 106
},
{
"epoch": 0.001976905311778291,
"grad_norm": 5.5757880210876465,
"learning_rate": 1.9997806834748455e-05,
"loss": 0.1242,
"step": 107
},
{
"epoch": 0.001995381062355658,
"grad_norm": 23.667171478271484,
"learning_rate": 1.9997014897811834e-05,
"loss": 0.215,
"step": 108
},
{
"epoch": 0.002013856812933025,
"grad_norm": 20.711095809936523,
"learning_rate": 1.9996101150403543e-05,
"loss": 0.2891,
"step": 109
},
{
"epoch": 0.0020323325635103924,
"grad_norm": 4.896965503692627,
"learning_rate": 1.9995065603657317e-05,
"loss": 0.1316,
"step": 110
},
{
"epoch": 0.00205080831408776,
"grad_norm": 35.64805603027344,
"learning_rate": 1.999390827019096e-05,
"loss": 0.3314,
"step": 111
},
{
"epoch": 0.002069284064665127,
"grad_norm": 12.456117630004883,
"learning_rate": 1.999262916410621e-05,
"loss": 0.1925,
"step": 112
},
{
"epoch": 0.0020877598152424944,
"grad_norm": 30.00156593322754,
"learning_rate": 1.9991228300988586e-05,
"loss": 0.216,
"step": 113
},
{
"epoch": 0.0021062355658198616,
"grad_norm": 26.710689544677734,
"learning_rate": 1.998970569790715e-05,
"loss": 0.4867,
"step": 114
},
{
"epoch": 0.0021247113163972288,
"grad_norm": 4.976332187652588,
"learning_rate": 1.9988061373414342e-05,
"loss": 0.112,
"step": 115
},
{
"epoch": 0.002143187066974596,
"grad_norm": 3.1971662044525146,
"learning_rate": 1.9986295347545738e-05,
"loss": 0.1055,
"step": 116
},
{
"epoch": 0.002161662817551963,
"grad_norm": 1.9513596296310425,
"learning_rate": 1.9984407641819812e-05,
"loss": 0.0871,
"step": 117
},
{
"epoch": 0.0021801385681293303,
"grad_norm": 36.184844970703125,
"learning_rate": 1.9982398279237657e-05,
"loss": 0.3325,
"step": 118
},
{
"epoch": 0.0021986143187066975,
"grad_norm": 5.225800037384033,
"learning_rate": 1.9980267284282718e-05,
"loss": 0.1101,
"step": 119
},
{
"epoch": 0.0022170900692840647,
"grad_norm": 37.07544708251953,
"learning_rate": 1.9978014682920503e-05,
"loss": 0.294,
"step": 120
},
{
"epoch": 0.002235565819861432,
"grad_norm": 11.194981575012207,
"learning_rate": 1.9975640502598243e-05,
"loss": 0.1449,
"step": 121
},
{
"epoch": 0.002254041570438799,
"grad_norm": 34.42378616333008,
"learning_rate": 1.997314477224458e-05,
"loss": 0.3138,
"step": 122
},
{
"epoch": 0.0022725173210161663,
"grad_norm": 3.8231968879699707,
"learning_rate": 1.9970527522269204e-05,
"loss": 0.1173,
"step": 123
},
{
"epoch": 0.0022909930715935335,
"grad_norm": 8.117308616638184,
"learning_rate": 1.9967788784562474e-05,
"loss": 0.1342,
"step": 124
},
{
"epoch": 0.0023094688221709007,
"grad_norm": 40.42792510986328,
"learning_rate": 1.9964928592495046e-05,
"loss": 0.4725,
"step": 125
},
{
"epoch": 0.002327944572748268,
"grad_norm": 3.473273754119873,
"learning_rate": 1.9961946980917457e-05,
"loss": 0.1242,
"step": 126
},
{
"epoch": 0.002346420323325635,
"grad_norm": 34.854549407958984,
"learning_rate": 1.9958843986159705e-05,
"loss": 0.2251,
"step": 127
},
{
"epoch": 0.0023648960739030022,
"grad_norm": 21.677228927612305,
"learning_rate": 1.99556196460308e-05,
"loss": 0.136,
"step": 128
},
{
"epoch": 0.0023833718244803694,
"grad_norm": 62.770263671875,
"learning_rate": 1.9952273999818312e-05,
"loss": 0.2625,
"step": 129
},
{
"epoch": 0.0024018475750577366,
"grad_norm": 9.687708854675293,
"learning_rate": 1.9948807088287884e-05,
"loss": 0.1311,
"step": 130
},
{
"epoch": 0.002420323325635104,
"grad_norm": 102.37527465820312,
"learning_rate": 1.9945218953682736e-05,
"loss": 0.3084,
"step": 131
},
{
"epoch": 0.002438799076212471,
"grad_norm": 1.700257658958435,
"learning_rate": 1.9941509639723155e-05,
"loss": 0.1036,
"step": 132
},
{
"epoch": 0.002457274826789838,
"grad_norm": 1.113442301750183,
"learning_rate": 1.9937679191605964e-05,
"loss": 0.1035,
"step": 133
},
{
"epoch": 0.0024757505773672053,
"grad_norm": 26.053890228271484,
"learning_rate": 1.9933727656003964e-05,
"loss": 0.1605,
"step": 134
},
{
"epoch": 0.0024942263279445725,
"grad_norm": 0.9373177289962769,
"learning_rate": 1.992965508106537e-05,
"loss": 0.0894,
"step": 135
},
{
"epoch": 0.00251270207852194,
"grad_norm": 1.1894716024398804,
"learning_rate": 1.9925461516413224e-05,
"loss": 0.0962,
"step": 136
},
{
"epoch": 0.0025311778290993073,
"grad_norm": 41.736141204833984,
"learning_rate": 1.9921147013144782e-05,
"loss": 1.3175,
"step": 137
},
{
"epoch": 0.0025496535796766745,
"grad_norm": 12.233323097229004,
"learning_rate": 1.9916711623830904e-05,
"loss": 0.8852,
"step": 138
},
{
"epoch": 0.0025681293302540417,
"grad_norm": 90.36572265625,
"learning_rate": 1.991215540251542e-05,
"loss": 0.5655,
"step": 139
},
{
"epoch": 0.002586605080831409,
"grad_norm": 3.386981725692749,
"learning_rate": 1.9907478404714438e-05,
"loss": 0.1149,
"step": 140
},
{
"epoch": 0.002605080831408776,
"grad_norm": 9.82344913482666,
"learning_rate": 1.9902680687415704e-05,
"loss": 0.1429,
"step": 141
},
{
"epoch": 0.0026235565819861433,
"grad_norm": 17.3359432220459,
"learning_rate": 1.989776230907789e-05,
"loss": 0.1623,
"step": 142
},
{
"epoch": 0.0026420323325635105,
"grad_norm": 12.32984733581543,
"learning_rate": 1.9892723329629885e-05,
"loss": 0.1369,
"step": 143
},
{
"epoch": 0.0026605080831408777,
"grad_norm": 1.4335039854049683,
"learning_rate": 1.988756381047006e-05,
"loss": 0.1111,
"step": 144
},
{
"epoch": 0.002678983833718245,
"grad_norm": 73.8090591430664,
"learning_rate": 1.988228381446553e-05,
"loss": 0.5501,
"step": 145
},
{
"epoch": 0.002697459584295612,
"grad_norm": 0.9194732308387756,
"learning_rate": 1.9876883405951378e-05,
"loss": 0.0846,
"step": 146
},
{
"epoch": 0.0027159353348729792,
"grad_norm": 20.720481872558594,
"learning_rate": 1.987136265072988e-05,
"loss": 0.5338,
"step": 147
},
{
"epoch": 0.0027344110854503464,
"grad_norm": 3.7738983631134033,
"learning_rate": 1.9865721616069695e-05,
"loss": 0.2182,
"step": 148
},
{
"epoch": 0.0027528868360277136,
"grad_norm": 1.4685333967208862,
"learning_rate": 1.985996037070505e-05,
"loss": 0.1003,
"step": 149
},
{
"epoch": 0.002771362586605081,
"grad_norm": 82.18400573730469,
"learning_rate": 1.9854078984834904e-05,
"loss": 0.4944,
"step": 150
},
{
"epoch": 0.002789838337182448,
"grad_norm": 72.0482177734375,
"learning_rate": 1.9848077530122083e-05,
"loss": 0.5362,
"step": 151
},
{
"epoch": 0.002808314087759815,
"grad_norm": 11.315860748291016,
"learning_rate": 1.984195607969242e-05,
"loss": 0.115,
"step": 152
},
{
"epoch": 0.0028267898383371824,
"grad_norm": 68.69886016845703,
"learning_rate": 1.983571470813386e-05,
"loss": 0.4085,
"step": 153
},
{
"epoch": 0.0028452655889145495,
"grad_norm": 15.664894104003906,
"learning_rate": 1.9829353491495545e-05,
"loss": 0.1785,
"step": 154
},
{
"epoch": 0.0028637413394919167,
"grad_norm": 27.369140625,
"learning_rate": 1.982287250728689e-05,
"loss": 0.3354,
"step": 155
},
{
"epoch": 0.002882217090069284,
"grad_norm": 7.389285564422607,
"learning_rate": 1.9816271834476642e-05,
"loss": 0.1562,
"step": 156
},
{
"epoch": 0.002900692840646651,
"grad_norm": 1.90992271900177,
"learning_rate": 1.9809551553491918e-05,
"loss": 0.1102,
"step": 157
},
{
"epoch": 0.0029191685912240183,
"grad_norm": 22.52420425415039,
"learning_rate": 1.9802711746217222e-05,
"loss": 0.4744,
"step": 158
},
{
"epoch": 0.0029376443418013855,
"grad_norm": 12.2611665725708,
"learning_rate": 1.979575249599344e-05,
"loss": 0.1608,
"step": 159
},
{
"epoch": 0.002956120092378753,
"grad_norm": 6.307402610778809,
"learning_rate": 1.9788673887616852e-05,
"loss": 0.1575,
"step": 160
},
{
"epoch": 0.0029745958429561203,
"grad_norm": 30.933067321777344,
"learning_rate": 1.9781476007338058e-05,
"loss": 0.3628,
"step": 161
},
{
"epoch": 0.0029930715935334875,
"grad_norm": 74.32799530029297,
"learning_rate": 1.9774158942860962e-05,
"loss": 0.8721,
"step": 162
},
{
"epoch": 0.0030115473441108547,
"grad_norm": 53.038612365722656,
"learning_rate": 1.9766722783341682e-05,
"loss": 0.4737,
"step": 163
},
{
"epoch": 0.003030023094688222,
"grad_norm": 18.48975944519043,
"learning_rate": 1.9759167619387474e-05,
"loss": 0.294,
"step": 164
},
{
"epoch": 0.003048498845265589,
"grad_norm": 32.237030029296875,
"learning_rate": 1.9751493543055634e-05,
"loss": 0.3167,
"step": 165
},
{
"epoch": 0.0030669745958429562,
"grad_norm": 39.83325958251953,
"learning_rate": 1.9743700647852356e-05,
"loss": 0.3928,
"step": 166
},
{
"epoch": 0.0030854503464203234,
"grad_norm": 28.04949378967285,
"learning_rate": 1.9735789028731603e-05,
"loss": 0.2263,
"step": 167
},
{
"epoch": 0.0031039260969976906,
"grad_norm": 52.016517639160156,
"learning_rate": 1.972775878209397e-05,
"loss": 0.3574,
"step": 168
},
{
"epoch": 0.003122401847575058,
"grad_norm": 7.5934553146362305,
"learning_rate": 1.9719610005785466e-05,
"loss": 0.1389,
"step": 169
},
{
"epoch": 0.003140877598152425,
"grad_norm": 9.168587684631348,
"learning_rate": 1.971134279909636e-05,
"loss": 0.1261,
"step": 170
},
{
"epoch": 0.003159353348729792,
"grad_norm": 1.65561044216156,
"learning_rate": 1.9702957262759964e-05,
"loss": 0.0878,
"step": 171
},
{
"epoch": 0.0031778290993071594,
"grad_norm": 25.38971710205078,
"learning_rate": 1.9694453498951392e-05,
"loss": 0.6207,
"step": 172
},
{
"epoch": 0.0031963048498845265,
"grad_norm": 22.86966323852539,
"learning_rate": 1.9685831611286312e-05,
"loss": 0.4574,
"step": 173
},
{
"epoch": 0.0032147806004618937,
"grad_norm": 58.06105041503906,
"learning_rate": 1.9677091704819714e-05,
"loss": 0.154,
"step": 174
},
{
"epoch": 0.003233256351039261,
"grad_norm": 22.46074867248535,
"learning_rate": 1.9668233886044597e-05,
"loss": 0.6643,
"step": 175
},
{
"epoch": 0.003251732101616628,
"grad_norm": 3.8287947177886963,
"learning_rate": 1.9659258262890683e-05,
"loss": 0.1403,
"step": 176
},
{
"epoch": 0.0032702078521939953,
"grad_norm": 38.93925094604492,
"learning_rate": 1.9650164944723116e-05,
"loss": 0.2734,
"step": 177
},
{
"epoch": 0.0032886836027713625,
"grad_norm": 61.83637237548828,
"learning_rate": 1.96409540423411e-05,
"loss": 0.3221,
"step": 178
},
{
"epoch": 0.0033071593533487297,
"grad_norm": 18.262434005737305,
"learning_rate": 1.9631625667976584e-05,
"loss": 0.2328,
"step": 179
},
{
"epoch": 0.003325635103926097,
"grad_norm": 46.0781364440918,
"learning_rate": 1.9622179935292855e-05,
"loss": 0.3,
"step": 180
},
{
"epoch": 0.003344110854503464,
"grad_norm": 19.867271423339844,
"learning_rate": 1.961261695938319e-05,
"loss": 0.4783,
"step": 181
},
{
"epoch": 0.0033625866050808312,
"grad_norm": 84.92687225341797,
"learning_rate": 1.9602936856769432e-05,
"loss": 0.3775,
"step": 182
},
{
"epoch": 0.0033810623556581984,
"grad_norm": 50.07325744628906,
"learning_rate": 1.9593139745400575e-05,
"loss": 0.584,
"step": 183
},
{
"epoch": 0.0033995381062355656,
"grad_norm": 2.1223649978637695,
"learning_rate": 1.9583225744651334e-05,
"loss": 0.1182,
"step": 184
},
{
"epoch": 0.0034180138568129332,
"grad_norm": 12.416468620300293,
"learning_rate": 1.9573194975320672e-05,
"loss": 0.3262,
"step": 185
},
{
"epoch": 0.0034364896073903004,
"grad_norm": 6.454475402832031,
"learning_rate": 1.9563047559630356e-05,
"loss": 0.1252,
"step": 186
},
{
"epoch": 0.0034549653579676676,
"grad_norm": 36.45076370239258,
"learning_rate": 1.9552783621223437e-05,
"loss": 0.3996,
"step": 187
},
{
"epoch": 0.003473441108545035,
"grad_norm": 3.376037120819092,
"learning_rate": 1.954240328516277e-05,
"loss": 0.1179,
"step": 188
},
{
"epoch": 0.003491916859122402,
"grad_norm": 8.251723289489746,
"learning_rate": 1.9531906677929472e-05,
"loss": 0.1935,
"step": 189
},
{
"epoch": 0.003510392609699769,
"grad_norm": 30.333776473999023,
"learning_rate": 1.9521293927421388e-05,
"loss": 0.1986,
"step": 190
},
{
"epoch": 0.0035288683602771364,
"grad_norm": 42.216583251953125,
"learning_rate": 1.9510565162951538e-05,
"loss": 0.1865,
"step": 191
},
{
"epoch": 0.0035473441108545035,
"grad_norm": 92.4293212890625,
"learning_rate": 1.9499720515246524e-05,
"loss": 0.2779,
"step": 192
},
{
"epoch": 0.0035658198614318707,
"grad_norm": 55.178836822509766,
"learning_rate": 1.9488760116444966e-05,
"loss": 0.7107,
"step": 193
},
{
"epoch": 0.003584295612009238,
"grad_norm": 24.224803924560547,
"learning_rate": 1.947768410009586e-05,
"loss": 0.2005,
"step": 194
},
{
"epoch": 0.003602771362586605,
"grad_norm": 15.010578155517578,
"learning_rate": 1.9466492601156964e-05,
"loss": 0.1484,
"step": 195
},
{
"epoch": 0.0036212471131639723,
"grad_norm": 2.2213423252105713,
"learning_rate": 1.945518575599317e-05,
"loss": 0.1064,
"step": 196
},
{
"epoch": 0.0036397228637413395,
"grad_norm": 0.8339602947235107,
"learning_rate": 1.944376370237481e-05,
"loss": 0.0986,
"step": 197
},
{
"epoch": 0.0036581986143187067,
"grad_norm": 1.2265304327011108,
"learning_rate": 1.943222657947601e-05,
"loss": 0.0974,
"step": 198
},
{
"epoch": 0.003676674364896074,
"grad_norm": 15.473770141601562,
"learning_rate": 1.942057452787297e-05,
"loss": 0.7621,
"step": 199
},
{
"epoch": 0.003695150115473441,
"grad_norm": 0.7640876173973083,
"learning_rate": 1.9408807689542257e-05,
"loss": 0.0902,
"step": 200
},
{
"epoch": 0.0037136258660508082,
"grad_norm": 2.0238685607910156,
"learning_rate": 1.9396926207859085e-05,
"loss": 0.1104,
"step": 201
},
{
"epoch": 0.0037321016166281754,
"grad_norm": 1.7173659801483154,
"learning_rate": 1.938493022759556e-05,
"loss": 0.1029,
"step": 202
},
{
"epoch": 0.0037505773672055426,
"grad_norm": 13.572944641113281,
"learning_rate": 1.937281989491892e-05,
"loss": 0.753,
"step": 203
},
{
"epoch": 0.00376905311778291,
"grad_norm": 58.37710952758789,
"learning_rate": 1.9360595357389735e-05,
"loss": 0.2875,
"step": 204
},
{
"epoch": 0.003787528868360277,
"grad_norm": 39.44758987426758,
"learning_rate": 1.9348256763960146e-05,
"loss": 0.317,
"step": 205
},
{
"epoch": 0.003806004618937644,
"grad_norm": 30.52085304260254,
"learning_rate": 1.9335804264972018e-05,
"loss": 0.3975,
"step": 206
},
{
"epoch": 0.0038244803695150114,
"grad_norm": 3.293384313583374,
"learning_rate": 1.9323238012155125e-05,
"loss": 0.1325,
"step": 207
},
{
"epoch": 0.0038429561200923786,
"grad_norm": 45.91811752319336,
"learning_rate": 1.9310558158625286e-05,
"loss": 0.2431,
"step": 208
},
{
"epoch": 0.003861431870669746,
"grad_norm": 4.664218902587891,
"learning_rate": 1.9297764858882516e-05,
"loss": 0.1618,
"step": 209
},
{
"epoch": 0.0038799076212471134,
"grad_norm": 7.681321144104004,
"learning_rate": 1.9284858268809135e-05,
"loss": 0.1672,
"step": 210
},
{
"epoch": 0.0038983833718244805,
"grad_norm": 2.7926387786865234,
"learning_rate": 1.9271838545667876e-05,
"loss": 0.1343,
"step": 211
},
{
"epoch": 0.003916859122401847,
"grad_norm": 83.28201293945312,
"learning_rate": 1.925870584809995e-05,
"loss": 0.6828,
"step": 212
},
{
"epoch": 0.0039353348729792145,
"grad_norm": 2.0914876461029053,
"learning_rate": 1.9245460336123136e-05,
"loss": 0.1181,
"step": 213
},
{
"epoch": 0.003953810623556582,
"grad_norm": 114.79316711425781,
"learning_rate": 1.923210217112981e-05,
"loss": 0.3275,
"step": 214
},
{
"epoch": 0.003972286374133949,
"grad_norm": 3.838773727416992,
"learning_rate": 1.9218631515885007e-05,
"loss": 0.1013,
"step": 215
},
{
"epoch": 0.003990762124711316,
"grad_norm": 4.88138484954834,
"learning_rate": 1.9205048534524405e-05,
"loss": 0.1471,
"step": 216
},
{
"epoch": 0.004009237875288683,
"grad_norm": 10.948898315429688,
"learning_rate": 1.9191353392552346e-05,
"loss": 0.1944,
"step": 217
},
{
"epoch": 0.00402771362586605,
"grad_norm": 4.260771751403809,
"learning_rate": 1.9177546256839814e-05,
"loss": 0.1101,
"step": 218
},
{
"epoch": 0.004046189376443418,
"grad_norm": 3.144196033477783,
"learning_rate": 1.9163627295622397e-05,
"loss": 0.1206,
"step": 219
},
{
"epoch": 0.004064665127020785,
"grad_norm": 1.3661218881607056,
"learning_rate": 1.914959667849825e-05,
"loss": 0.1051,
"step": 220
},
{
"epoch": 0.004083140877598152,
"grad_norm": 9.742240905761719,
"learning_rate": 1.913545457642601e-05,
"loss": 0.6042,
"step": 221
},
{
"epoch": 0.00410161662817552,
"grad_norm": 20.63588523864746,
"learning_rate": 1.9121201161722732e-05,
"loss": 0.1159,
"step": 222
},
{
"epoch": 0.004120092378752887,
"grad_norm": 2.1614677906036377,
"learning_rate": 1.910683660806177e-05,
"loss": 0.1078,
"step": 223
},
{
"epoch": 0.004138568129330254,
"grad_norm": 19.844690322875977,
"learning_rate": 1.9092361090470688e-05,
"loss": 0.8202,
"step": 224
},
{
"epoch": 0.004157043879907622,
"grad_norm": 3.5642316341400146,
"learning_rate": 1.907777478532909e-05,
"loss": 0.1165,
"step": 225
},
{
"epoch": 0.004175519630484989,
"grad_norm": 5.66938591003418,
"learning_rate": 1.9063077870366504e-05,
"loss": 0.1943,
"step": 226
},
{
"epoch": 0.004193995381062356,
"grad_norm": 22.645231246948242,
"learning_rate": 1.9048270524660197e-05,
"loss": 0.3442,
"step": 227
},
{
"epoch": 0.004212471131639723,
"grad_norm": 90.24234008789062,
"learning_rate": 1.903335292863301e-05,
"loss": 0.764,
"step": 228
},
{
"epoch": 0.00423094688221709,
"grad_norm": 25.336103439331055,
"learning_rate": 1.901832526405114e-05,
"loss": 0.7113,
"step": 229
},
{
"epoch": 0.0042494226327944576,
"grad_norm": 1.2521638870239258,
"learning_rate": 1.9003187714021936e-05,
"loss": 0.0965,
"step": 230
},
{
"epoch": 0.004267898383371825,
"grad_norm": 18.136844635009766,
"learning_rate": 1.8987940462991673e-05,
"loss": 0.2196,
"step": 231
},
{
"epoch": 0.004286374133949192,
"grad_norm": 51.36814880371094,
"learning_rate": 1.8972583696743284e-05,
"loss": 0.2887,
"step": 232
},
{
"epoch": 0.004304849884526559,
"grad_norm": 2.3742902278900146,
"learning_rate": 1.895711760239413e-05,
"loss": 0.1088,
"step": 233
},
{
"epoch": 0.004323325635103926,
"grad_norm": 4.106175899505615,
"learning_rate": 1.8941542368393683e-05,
"loss": 0.1264,
"step": 234
},
{
"epoch": 0.0043418013856812935,
"grad_norm": 73.95805358886719,
"learning_rate": 1.892585818452126e-05,
"loss": 0.2295,
"step": 235
},
{
"epoch": 0.004360277136258661,
"grad_norm": 2.3880388736724854,
"learning_rate": 1.891006524188368e-05,
"loss": 0.1118,
"step": 236
},
{
"epoch": 0.004378752886836028,
"grad_norm": 1.9430997371673584,
"learning_rate": 1.889416373291298e-05,
"loss": 0.1006,
"step": 237
},
{
"epoch": 0.004397228637413395,
"grad_norm": 100.13737487792969,
"learning_rate": 1.8878153851364013e-05,
"loss": 0.2527,
"step": 238
},
{
"epoch": 0.004415704387990762,
"grad_norm": 1.8056846857070923,
"learning_rate": 1.8862035792312148e-05,
"loss": 0.1181,
"step": 239
},
{
"epoch": 0.004434180138568129,
"grad_norm": 4.330031871795654,
"learning_rate": 1.884580975215084e-05,
"loss": 0.1205,
"step": 240
},
{
"epoch": 0.004452655889145497,
"grad_norm": 2.0573930740356445,
"learning_rate": 1.8829475928589272e-05,
"loss": 0.104,
"step": 241
},
{
"epoch": 0.004471131639722864,
"grad_norm": 2.3082385063171387,
"learning_rate": 1.8813034520649923e-05,
"loss": 0.116,
"step": 242
},
{
"epoch": 0.004489607390300231,
"grad_norm": 22.696136474609375,
"learning_rate": 1.879648572866617e-05,
"loss": 0.5436,
"step": 243
},
{
"epoch": 0.004508083140877598,
"grad_norm": 0.9982879161834717,
"learning_rate": 1.8779829754279806e-05,
"loss": 0.0859,
"step": 244
},
{
"epoch": 0.004526558891454965,
"grad_norm": 3.4211270809173584,
"learning_rate": 1.8763066800438638e-05,
"loss": 0.1152,
"step": 245
},
{
"epoch": 0.0045450346420323326,
"grad_norm": 35.30122756958008,
"learning_rate": 1.874619707139396e-05,
"loss": 0.483,
"step": 246
},
{
"epoch": 0.0045635103926097,
"grad_norm": 1.8901995420455933,
"learning_rate": 1.8729220772698096e-05,
"loss": 0.0921,
"step": 247
},
{
"epoch": 0.004581986143187067,
"grad_norm": 21.672452926635742,
"learning_rate": 1.8712138111201898e-05,
"loss": 0.7347,
"step": 248
},
{
"epoch": 0.004600461893764434,
"grad_norm": 17.743106842041016,
"learning_rate": 1.869494929505219e-05,
"loss": 0.1591,
"step": 249
},
{
"epoch": 0.004618937644341801,
"grad_norm": 14.006978034973145,
"learning_rate": 1.8677654533689287e-05,
"loss": 0.5446,
"step": 250
},
{
"epoch": 0.0046374133949191685,
"grad_norm": 20.738595962524414,
"learning_rate": 1.866025403784439e-05,
"loss": 0.4985,
"step": 251
},
{
"epoch": 0.004655889145496536,
"grad_norm": 1.4686133861541748,
"learning_rate": 1.864274801953705e-05,
"loss": 0.0936,
"step": 252
},
{
"epoch": 0.004674364896073903,
"grad_norm": 1.796115756034851,
"learning_rate": 1.8625136692072577e-05,
"loss": 0.0941,
"step": 253
},
{
"epoch": 0.00469284064665127,
"grad_norm": 7.412848949432373,
"learning_rate": 1.860742027003944e-05,
"loss": 0.1587,
"step": 254
},
{
"epoch": 0.004711316397228637,
"grad_norm": 31.30906105041504,
"learning_rate": 1.8589598969306646e-05,
"loss": 0.3564,
"step": 255
},
{
"epoch": 0.0047297921478060044,
"grad_norm": 2.334212303161621,
"learning_rate": 1.8571673007021124e-05,
"loss": 0.1191,
"step": 256
},
{
"epoch": 0.004748267898383372,
"grad_norm": 51.09286117553711,
"learning_rate": 1.855364260160507e-05,
"loss": 0.27,
"step": 257
},
{
"epoch": 0.004766743648960739,
"grad_norm": 1.5464048385620117,
"learning_rate": 1.8535507972753275e-05,
"loss": 0.1111,
"step": 258
},
{
"epoch": 0.004785219399538106,
"grad_norm": 1.0696243047714233,
"learning_rate": 1.851726934143048e-05,
"loss": 0.0876,
"step": 259
},
{
"epoch": 0.004803695150115473,
"grad_norm": 27.565221786499023,
"learning_rate": 1.849892692986864e-05,
"loss": 0.4421,
"step": 260
},
{
"epoch": 0.00482217090069284,
"grad_norm": 6.274564266204834,
"learning_rate": 1.848048096156426e-05,
"loss": 0.1106,
"step": 261
},
{
"epoch": 0.004840646651270208,
"grad_norm": 4.153088092803955,
"learning_rate": 1.8461931661275642e-05,
"loss": 0.1076,
"step": 262
},
{
"epoch": 0.004859122401847575,
"grad_norm": 80.05952453613281,
"learning_rate": 1.8443279255020153e-05,
"loss": 0.2082,
"step": 263
},
{
"epoch": 0.004877598152424942,
"grad_norm": 23.937946319580078,
"learning_rate": 1.842452397007148e-05,
"loss": 0.1877,
"step": 264
},
{
"epoch": 0.004896073903002309,
"grad_norm": 5.535915374755859,
"learning_rate": 1.8405666034956842e-05,
"loss": 0.1195,
"step": 265
},
{
"epoch": 0.004914549653579676,
"grad_norm": 15.148236274719238,
"learning_rate": 1.8386705679454243e-05,
"loss": 0.6512,
"step": 266
},
{
"epoch": 0.0049330254041570435,
"grad_norm": 1.7562042474746704,
"learning_rate": 1.836764313458962e-05,
"loss": 0.092,
"step": 267
},
{
"epoch": 0.004951501154734411,
"grad_norm": 35.28544998168945,
"learning_rate": 1.8348478632634067e-05,
"loss": 0.688,
"step": 268
},
{
"epoch": 0.004969976905311778,
"grad_norm": 38.12431716918945,
"learning_rate": 1.8329212407100996e-05,
"loss": 0.8571,
"step": 269
},
{
"epoch": 0.004988452655889145,
"grad_norm": 2.9317550659179688,
"learning_rate": 1.8309844692743283e-05,
"loss": 0.1207,
"step": 270
},
{
"epoch": 0.005006928406466513,
"grad_norm": 35.599605560302734,
"learning_rate": 1.8290375725550417e-05,
"loss": 0.3517,
"step": 271
},
{
"epoch": 0.00502540415704388,
"grad_norm": 13.001343727111816,
"learning_rate": 1.827080574274562e-05,
"loss": 0.3471,
"step": 272
},
{
"epoch": 0.0050438799076212475,
"grad_norm": 19.658357620239258,
"learning_rate": 1.8251134982782952e-05,
"loss": 0.3096,
"step": 273
},
{
"epoch": 0.005062355658198615,
"grad_norm": 8.629194259643555,
"learning_rate": 1.8231363685344422e-05,
"loss": 0.1744,
"step": 274
},
{
"epoch": 0.005080831408775982,
"grad_norm": 4.561062812805176,
"learning_rate": 1.821149209133704e-05,
"loss": 0.1327,
"step": 275
},
{
"epoch": 0.005099307159353349,
"grad_norm": 23.622230529785156,
"learning_rate": 1.819152044288992e-05,
"loss": 0.5279,
"step": 276
},
{
"epoch": 0.005117782909930716,
"grad_norm": 19.12730598449707,
"learning_rate": 1.8171448983351284e-05,
"loss": 0.2452,
"step": 277
},
{
"epoch": 0.0051362586605080834,
"grad_norm": 20.153221130371094,
"learning_rate": 1.815127795728554e-05,
"loss": 0.2621,
"step": 278
},
{
"epoch": 0.005154734411085451,
"grad_norm": 2.64178204536438,
"learning_rate": 1.8131007610470278e-05,
"loss": 0.1153,
"step": 279
},
{
"epoch": 0.005173210161662818,
"grad_norm": 10.701132774353027,
"learning_rate": 1.8110638189893267e-05,
"loss": 0.2623,
"step": 280
},
{
"epoch": 0.005191685912240185,
"grad_norm": 23.051742553710938,
"learning_rate": 1.8090169943749477e-05,
"loss": 0.311,
"step": 281
},
{
"epoch": 0.005210161662817552,
"grad_norm": 19.863746643066406,
"learning_rate": 1.806960312143802e-05,
"loss": 0.3145,
"step": 282
},
{
"epoch": 0.005228637413394919,
"grad_norm": 13.43600082397461,
"learning_rate": 1.804893797355914e-05,
"loss": 0.1315,
"step": 283
},
{
"epoch": 0.005247113163972287,
"grad_norm": 9.73802375793457,
"learning_rate": 1.8028174751911147e-05,
"loss": 0.1668,
"step": 284
},
{
"epoch": 0.005265588914549654,
"grad_norm": 2.9510879516601562,
"learning_rate": 1.8007313709487334e-05,
"loss": 0.1039,
"step": 285
},
{
"epoch": 0.005284064665127021,
"grad_norm": 1.8180186748504639,
"learning_rate": 1.798635510047293e-05,
"loss": 0.1033,
"step": 286
},
{
"epoch": 0.005302540415704388,
"grad_norm": 1.426766037940979,
"learning_rate": 1.7965299180241963e-05,
"loss": 0.0692,
"step": 287
},
{
"epoch": 0.005321016166281755,
"grad_norm": 18.539081573486328,
"learning_rate": 1.7944146205354182e-05,
"loss": 0.3479,
"step": 288
},
{
"epoch": 0.0053394919168591225,
"grad_norm": 34.73767852783203,
"learning_rate": 1.792289643355191e-05,
"loss": 0.2214,
"step": 289
},
{
"epoch": 0.00535796766743649,
"grad_norm": 1.9383809566497803,
"learning_rate": 1.7901550123756906e-05,
"loss": 0.1023,
"step": 290
},
{
"epoch": 0.005376443418013857,
"grad_norm": 1.8322877883911133,
"learning_rate": 1.788010753606722e-05,
"loss": 0.0946,
"step": 291
},
{
"epoch": 0.005394919168591224,
"grad_norm": 17.55063819885254,
"learning_rate": 1.785856893175402e-05,
"loss": 0.1412,
"step": 292
},
{
"epoch": 0.005413394919168591,
"grad_norm": 61.62004852294922,
"learning_rate": 1.78369345732584e-05,
"loss": 1.2193,
"step": 293
},
{
"epoch": 0.0054318706697459584,
"grad_norm": 2.1789324283599854,
"learning_rate": 1.781520472418819e-05,
"loss": 0.0996,
"step": 294
},
{
"epoch": 0.005450346420323326,
"grad_norm": 17.134449005126953,
"learning_rate": 1.7793379649314743e-05,
"loss": 0.7183,
"step": 295
},
{
"epoch": 0.005468822170900693,
"grad_norm": 12.738836288452148,
"learning_rate": 1.777145961456971e-05,
"loss": 0.1244,
"step": 296
},
{
"epoch": 0.00548729792147806,
"grad_norm": 49.77069091796875,
"learning_rate": 1.7749444887041797e-05,
"loss": 0.3325,
"step": 297
},
{
"epoch": 0.005505773672055427,
"grad_norm": 38.23817443847656,
"learning_rate": 1.7727335734973512e-05,
"loss": 0.3567,
"step": 298
},
{
"epoch": 0.005524249422632794,
"grad_norm": 1.9517183303833008,
"learning_rate": 1.7705132427757895e-05,
"loss": 0.0815,
"step": 299
},
{
"epoch": 0.005542725173210162,
"grad_norm": 9.183816909790039,
"learning_rate": 1.7682835235935236e-05,
"loss": 0.7927,
"step": 300
},
{
"epoch": 0.005561200923787529,
"grad_norm": 5.687363624572754,
"learning_rate": 1.766044443118978e-05,
"loss": 0.1235,
"step": 301
},
{
"epoch": 0.005579676674364896,
"grad_norm": 9.396485328674316,
"learning_rate": 1.7637960286346423e-05,
"loss": 0.1612,
"step": 302
},
{
"epoch": 0.005598152424942263,
"grad_norm": 18.12451171875,
"learning_rate": 1.761538307536737e-05,
"loss": 0.189,
"step": 303
},
{
"epoch": 0.00561662817551963,
"grad_norm": 14.55724048614502,
"learning_rate": 1.759271307334881e-05,
"loss": 0.237,
"step": 304
},
{
"epoch": 0.0056351039260969975,
"grad_norm": 6.085824489593506,
"learning_rate": 1.7569950556517566e-05,
"loss": 0.1244,
"step": 305
},
{
"epoch": 0.005653579676674365,
"grad_norm": 1.9505577087402344,
"learning_rate": 1.7547095802227723e-05,
"loss": 0.1485,
"step": 306
},
{
"epoch": 0.005672055427251732,
"grad_norm": 4.278805732727051,
"learning_rate": 1.7524149088957244e-05,
"loss": 0.1321,
"step": 307
},
{
"epoch": 0.005690531177829099,
"grad_norm": 41.136512756347656,
"learning_rate": 1.7501110696304598e-05,
"loss": 0.4183,
"step": 308
},
{
"epoch": 0.005709006928406466,
"grad_norm": 44.14323043823242,
"learning_rate": 1.747798090498532e-05,
"loss": 0.2908,
"step": 309
},
{
"epoch": 0.0057274826789838335,
"grad_norm": 17.839080810546875,
"learning_rate": 1.7454759996828622e-05,
"loss": 0.306,
"step": 310
},
{
"epoch": 0.005745958429561201,
"grad_norm": 10.629524230957031,
"learning_rate": 1.7431448254773943e-05,
"loss": 0.1472,
"step": 311
},
{
"epoch": 0.005764434180138568,
"grad_norm": 2.029240846633911,
"learning_rate": 1.74080459628675e-05,
"loss": 0.0962,
"step": 312
},
{
"epoch": 0.005782909930715935,
"grad_norm": 18.86973762512207,
"learning_rate": 1.7384553406258842e-05,
"loss": 0.1771,
"step": 313
},
{
"epoch": 0.005801385681293302,
"grad_norm": 14.951451301574707,
"learning_rate": 1.7360970871197347e-05,
"loss": 0.181,
"step": 314
},
{
"epoch": 0.005819861431870669,
"grad_norm": 4.3663482666015625,
"learning_rate": 1.7337298645028764e-05,
"loss": 0.1355,
"step": 315
},
{
"epoch": 0.005838337182448037,
"grad_norm": 7.262619495391846,
"learning_rate": 1.7313537016191706e-05,
"loss": 0.1347,
"step": 316
},
{
"epoch": 0.005856812933025404,
"grad_norm": 3.332489252090454,
"learning_rate": 1.7289686274214116e-05,
"loss": 0.1134,
"step": 317
},
{
"epoch": 0.005875288683602771,
"grad_norm": 35.11198425292969,
"learning_rate": 1.7265746709709762e-05,
"loss": 0.6574,
"step": 318
},
{
"epoch": 0.005893764434180138,
"grad_norm": 1.0543757677078247,
"learning_rate": 1.7241718614374678e-05,
"loss": 0.0805,
"step": 319
},
{
"epoch": 0.005912240184757506,
"grad_norm": 18.589736938476562,
"learning_rate": 1.7217602280983622e-05,
"loss": 0.6666,
"step": 320
},
{
"epoch": 0.005930715935334873,
"grad_norm": 5.046590805053711,
"learning_rate": 1.7193398003386514e-05,
"loss": 0.1279,
"step": 321
},
{
"epoch": 0.005949191685912241,
"grad_norm": 0.9419677257537842,
"learning_rate": 1.716910607650483e-05,
"loss": 0.0918,
"step": 322
},
{
"epoch": 0.005967667436489608,
"grad_norm": 2.171743869781494,
"learning_rate": 1.7144726796328034e-05,
"loss": 0.1322,
"step": 323
},
{
"epoch": 0.005986143187066975,
"grad_norm": 58.99452209472656,
"learning_rate": 1.712026045990997e-05,
"loss": 0.6883,
"step": 324
},
{
"epoch": 0.006004618937644342,
"grad_norm": 1.6074374914169312,
"learning_rate": 1.709570736536521e-05,
"loss": 0.1094,
"step": 325
},
{
"epoch": 0.006023094688221709,
"grad_norm": 3.2632040977478027,
"learning_rate": 1.7071067811865477e-05,
"loss": 0.1181,
"step": 326
},
{
"epoch": 0.0060415704387990765,
"grad_norm": 11.132040023803711,
"learning_rate": 1.7046342099635948e-05,
"loss": 0.6684,
"step": 327
},
{
"epoch": 0.006060046189376444,
"grad_norm": 15.503205299377441,
"learning_rate": 1.7021530529951627e-05,
"loss": 0.1506,
"step": 328
},
{
"epoch": 0.006078521939953811,
"grad_norm": 1.1311938762664795,
"learning_rate": 1.6996633405133656e-05,
"loss": 0.0943,
"step": 329
},
{
"epoch": 0.006096997690531178,
"grad_norm": 1.1199091672897339,
"learning_rate": 1.697165102854565e-05,
"loss": 0.0922,
"step": 330
},
{
"epoch": 0.006115473441108545,
"grad_norm": 31.916030883789062,
"learning_rate": 1.6946583704589973e-05,
"loss": 0.4221,
"step": 331
},
{
"epoch": 0.0061339491916859125,
"grad_norm": 33.02965545654297,
"learning_rate": 1.692143173870407e-05,
"loss": 0.3379,
"step": 332
},
{
"epoch": 0.00615242494226328,
"grad_norm": 10.411473274230957,
"learning_rate": 1.68961954373567e-05,
"loss": 0.7348,
"step": 333
},
{
"epoch": 0.006170900692840647,
"grad_norm": 190.42056274414062,
"learning_rate": 1.6870875108044233e-05,
"loss": 0.4543,
"step": 334
},
{
"epoch": 0.006189376443418014,
"grad_norm": 15.432124137878418,
"learning_rate": 1.684547105928689e-05,
"loss": 0.1419,
"step": 335
},
{
"epoch": 0.006207852193995381,
"grad_norm": 1.7345023155212402,
"learning_rate": 1.6819983600624986e-05,
"loss": 0.1004,
"step": 336
},
{
"epoch": 0.006226327944572748,
"grad_norm": 1.2295470237731934,
"learning_rate": 1.6794413042615168e-05,
"loss": 0.0925,
"step": 337
},
{
"epoch": 0.006244803695150116,
"grad_norm": 2.497136116027832,
"learning_rate": 1.6768759696826608e-05,
"loss": 0.1041,
"step": 338
},
{
"epoch": 0.006263279445727483,
"grad_norm": 24.804088592529297,
"learning_rate": 1.6743023875837233e-05,
"loss": 0.4379,
"step": 339
},
{
"epoch": 0.00628175519630485,
"grad_norm": 40.6839599609375,
"learning_rate": 1.6717205893229904e-05,
"loss": 0.3335,
"step": 340
},
{
"epoch": 0.006300230946882217,
"grad_norm": 10.375980377197266,
"learning_rate": 1.6691306063588583e-05,
"loss": 0.1572,
"step": 341
},
{
"epoch": 0.006318706697459584,
"grad_norm": 18.76382064819336,
"learning_rate": 1.6665324702494524e-05,
"loss": 0.2269,
"step": 342
},
{
"epoch": 0.0063371824480369515,
"grad_norm": 37.63737869262695,
"learning_rate": 1.6639262126522417e-05,
"loss": 0.2887,
"step": 343
},
{
"epoch": 0.006355658198614319,
"grad_norm": 6.424595832824707,
"learning_rate": 1.661311865323652e-05,
"loss": 0.1724,
"step": 344
},
{
"epoch": 0.006374133949191686,
"grad_norm": 4.385128498077393,
"learning_rate": 1.6586894601186804e-05,
"loss": 0.1392,
"step": 345
},
{
"epoch": 0.006392609699769053,
"grad_norm": 7.946055889129639,
"learning_rate": 1.6560590289905074e-05,
"loss": 0.1342,
"step": 346
},
{
"epoch": 0.00641108545034642,
"grad_norm": 3.774859666824341,
"learning_rate": 1.6534206039901057e-05,
"loss": 0.106,
"step": 347
},
{
"epoch": 0.0064295612009237875,
"grad_norm": 28.573497772216797,
"learning_rate": 1.650774217265851e-05,
"loss": 0.3729,
"step": 348
},
{
"epoch": 0.006448036951501155,
"grad_norm": 52.259456634521484,
"learning_rate": 1.6481199010631312e-05,
"loss": 0.5048,
"step": 349
},
{
"epoch": 0.006466512702078522,
"grad_norm": 24.54702377319336,
"learning_rate": 1.645457687723951e-05,
"loss": 0.2182,
"step": 350
},
{
"epoch": 0.006484988452655889,
"grad_norm": 35.874534606933594,
"learning_rate": 1.6427876096865394e-05,
"loss": 0.2301,
"step": 351
},
{
"epoch": 0.006503464203233256,
"grad_norm": 2.483865261077881,
"learning_rate": 1.6401096994849558e-05,
"loss": 0.1005,
"step": 352
},
{
"epoch": 0.006521939953810623,
"grad_norm": 8.948705673217773,
"learning_rate": 1.63742398974869e-05,
"loss": 0.1228,
"step": 353
},
{
"epoch": 0.006540415704387991,
"grad_norm": 1.6457048654556274,
"learning_rate": 1.6347305132022677e-05,
"loss": 0.0925,
"step": 354
},
{
"epoch": 0.006558891454965358,
"grad_norm": 1.9249595403671265,
"learning_rate": 1.632029302664851e-05,
"loss": 0.0911,
"step": 355
},
{
"epoch": 0.006577367205542725,
"grad_norm": 3.952897071838379,
"learning_rate": 1.6293203910498375e-05,
"loss": 0.1019,
"step": 356
},
{
"epoch": 0.006595842956120092,
"grad_norm": 2.4536964893341064,
"learning_rate": 1.6266038113644605e-05,
"loss": 0.0973,
"step": 357
},
{
"epoch": 0.006614318706697459,
"grad_norm": 2.164851427078247,
"learning_rate": 1.6238795967093865e-05,
"loss": 0.085,
"step": 358
},
{
"epoch": 0.0066327944572748265,
"grad_norm": 3.3521976470947266,
"learning_rate": 1.6211477802783105e-05,
"loss": 0.0952,
"step": 359
},
{
"epoch": 0.006651270207852194,
"grad_norm": 1.366515040397644,
"learning_rate": 1.6184083953575543e-05,
"loss": 0.083,
"step": 360
},
{
"epoch": 0.006669745958429561,
"grad_norm": 31.498674392700195,
"learning_rate": 1.6156614753256583e-05,
"loss": 0.1175,
"step": 361
},
{
"epoch": 0.006688221709006928,
"grad_norm": 1.7299107313156128,
"learning_rate": 1.6129070536529767e-05,
"loss": 0.0965,
"step": 362
},
{
"epoch": 0.006706697459584295,
"grad_norm": 3.9225969314575195,
"learning_rate": 1.610145163901268e-05,
"loss": 0.1197,
"step": 363
},
{
"epoch": 0.0067251732101616625,
"grad_norm": 1.240811824798584,
"learning_rate": 1.607375839723287e-05,
"loss": 0.093,
"step": 364
},
{
"epoch": 0.00674364896073903,
"grad_norm": 0.7142246961593628,
"learning_rate": 1.6045991148623752e-05,
"loss": 0.0745,
"step": 365
},
{
"epoch": 0.006762124711316397,
"grad_norm": 3.2028310298919678,
"learning_rate": 1.6018150231520486e-05,
"loss": 0.1077,
"step": 366
},
{
"epoch": 0.006780600461893764,
"grad_norm": 1.01662278175354,
"learning_rate": 1.599023598515586e-05,
"loss": 0.0742,
"step": 367
},
{
"epoch": 0.006799076212471131,
"grad_norm": 2.3901126384735107,
"learning_rate": 1.5962248749656158e-05,
"loss": 0.0914,
"step": 368
},
{
"epoch": 0.006817551963048499,
"grad_norm": 1.9054994583129883,
"learning_rate": 1.5934188866037017e-05,
"loss": 0.0723,
"step": 369
},
{
"epoch": 0.0068360277136258665,
"grad_norm": 199.93890380859375,
"learning_rate": 1.5906056676199256e-05,
"loss": 0.9652,
"step": 370
},
{
"epoch": 0.006854503464203234,
"grad_norm": 7.036402225494385,
"learning_rate": 1.5877852522924733e-05,
"loss": 0.0866,
"step": 371
},
{
"epoch": 0.006872979214780601,
"grad_norm": 1.4115077257156372,
"learning_rate": 1.584957674987216e-05,
"loss": 0.0943,
"step": 372
},
{
"epoch": 0.006891454965357968,
"grad_norm": 6.522600173950195,
"learning_rate": 1.5821229701572897e-05,
"loss": 0.1165,
"step": 373
},
{
"epoch": 0.006909930715935335,
"grad_norm": 56.573020935058594,
"learning_rate": 1.5792811723426787e-05,
"loss": 0.4339,
"step": 374
},
{
"epoch": 0.006928406466512702,
"grad_norm": 70.64860534667969,
"learning_rate": 1.5764323161697933e-05,
"loss": 0.6481,
"step": 375
},
{
"epoch": 0.00694688221709007,
"grad_norm": 2.3378095626831055,
"learning_rate": 1.573576436351046e-05,
"loss": 0.0894,
"step": 376
},
{
"epoch": 0.006965357967667437,
"grad_norm": 2.4079654216766357,
"learning_rate": 1.570713567684432e-05,
"loss": 0.0869,
"step": 377
},
{
"epoch": 0.006983833718244804,
"grad_norm": 4.931615352630615,
"learning_rate": 1.5678437450531014e-05,
"loss": 0.1477,
"step": 378
},
{
"epoch": 0.007002309468822171,
"grad_norm": 13.358853340148926,
"learning_rate": 1.564967003424938e-05,
"loss": 0.1318,
"step": 379
},
{
"epoch": 0.007020785219399538,
"grad_norm": 11.24197769165039,
"learning_rate": 1.5620833778521306e-05,
"loss": 0.7465,
"step": 380
},
{
"epoch": 0.0070392609699769055,
"grad_norm": 1.6019229888916016,
"learning_rate": 1.5591929034707468e-05,
"loss": 0.098,
"step": 381
},
{
"epoch": 0.007057736720554273,
"grad_norm": 2.7691991329193115,
"learning_rate": 1.556295615500305e-05,
"loss": 0.0926,
"step": 382
},
{
"epoch": 0.00707621247113164,
"grad_norm": 1.3631212711334229,
"learning_rate": 1.553391549243344e-05,
"loss": 0.1031,
"step": 383
},
{
"epoch": 0.007094688221709007,
"grad_norm": 4.861557960510254,
"learning_rate": 1.5504807400849957e-05,
"loss": 0.101,
"step": 384
},
{
"epoch": 0.007113163972286374,
"grad_norm": 13.967028617858887,
"learning_rate": 1.5475632234925505e-05,
"loss": 0.6718,
"step": 385
},
{
"epoch": 0.0071316397228637415,
"grad_norm": 1.688780426979065,
"learning_rate": 1.5446390350150272e-05,
"loss": 0.0799,
"step": 386
},
{
"epoch": 0.007150115473441109,
"grad_norm": 11.237764358520508,
"learning_rate": 1.54170821028274e-05,
"loss": 0.1361,
"step": 387
},
{
"epoch": 0.007168591224018476,
"grad_norm": 4.256268501281738,
"learning_rate": 1.5387707850068633e-05,
"loss": 0.0916,
"step": 388
},
{
"epoch": 0.007187066974595843,
"grad_norm": 1.1430052518844604,
"learning_rate": 1.5358267949789968e-05,
"loss": 0.0818,
"step": 389
},
{
"epoch": 0.00720554272517321,
"grad_norm": 47.733943939208984,
"learning_rate": 1.53287627607073e-05,
"loss": 0.134,
"step": 390
},
{
"epoch": 0.007224018475750577,
"grad_norm": 38.1656608581543,
"learning_rate": 1.529919264233205e-05,
"loss": 0.4995,
"step": 391
},
{
"epoch": 0.007242494226327945,
"grad_norm": 1.2250961065292358,
"learning_rate": 1.5269557954966777e-05,
"loss": 0.0852,
"step": 392
},
{
"epoch": 0.007260969976905312,
"grad_norm": 1.548323631286621,
"learning_rate": 1.5239859059700794e-05,
"loss": 0.0818,
"step": 393
},
{
"epoch": 0.007279445727482679,
"grad_norm": 63.81478500366211,
"learning_rate": 1.5210096318405768e-05,
"loss": 0.2927,
"step": 394
},
{
"epoch": 0.007297921478060046,
"grad_norm": 1.1672543287277222,
"learning_rate": 1.5180270093731305e-05,
"loss": 0.0729,
"step": 395
},
{
"epoch": 0.007316397228637413,
"grad_norm": 3.8235509395599365,
"learning_rate": 1.5150380749100545e-05,
"loss": 0.1308,
"step": 396
},
{
"epoch": 0.0073348729792147805,
"grad_norm": 1.0563135147094727,
"learning_rate": 1.5120428648705716e-05,
"loss": 0.0857,
"step": 397
},
{
"epoch": 0.007353348729792148,
"grad_norm": 1.7499654293060303,
"learning_rate": 1.5090414157503715e-05,
"loss": 0.0991,
"step": 398
},
{
"epoch": 0.007371824480369515,
"grad_norm": 58.56557846069336,
"learning_rate": 1.5060337641211637e-05,
"loss": 0.6564,
"step": 399
},
{
"epoch": 0.007390300230946882,
"grad_norm": 2.0371148586273193,
"learning_rate": 1.5030199466302354e-05,
"loss": 0.0966,
"step": 400
},
{
"epoch": 0.007408775981524249,
"grad_norm": 38.98274230957031,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.6289,
"step": 401
},
{
"epoch": 0.0074272517321016165,
"grad_norm": 2.115440845489502,
"learning_rate": 1.4969739610275556e-05,
"loss": 0.078,
"step": 402
},
{
"epoch": 0.007445727482678984,
"grad_norm": 1.2540347576141357,
"learning_rate": 1.493941866584231e-05,
"loss": 0.0728,
"step": 403
},
{
"epoch": 0.007464203233256351,
"grad_norm": 1.2395659685134888,
"learning_rate": 1.490903753615141e-05,
"loss": 0.074,
"step": 404
},
{
"epoch": 0.007482678983833718,
"grad_norm": 156.56854248046875,
"learning_rate": 1.4878596591387329e-05,
"loss": 0.296,
"step": 405
},
{
"epoch": 0.007501154734411085,
"grad_norm": 45.55757141113281,
"learning_rate": 1.4848096202463373e-05,
"loss": 0.1934,
"step": 406
},
{
"epoch": 0.007519630484988452,
"grad_norm": 1.9094820022583008,
"learning_rate": 1.4817536741017153e-05,
"loss": 0.0774,
"step": 407
},
{
"epoch": 0.00753810623556582,
"grad_norm": 97.25504302978516,
"learning_rate": 1.478691857940607e-05,
"loss": 0.4552,
"step": 408
},
{
"epoch": 0.007556581986143187,
"grad_norm": 1.298113226890564,
"learning_rate": 1.4756242090702756e-05,
"loss": 0.0768,
"step": 409
},
{
"epoch": 0.007575057736720554,
"grad_norm": 3.1004037857055664,
"learning_rate": 1.4725507648690542e-05,
"loss": 0.0953,
"step": 410
},
{
"epoch": 0.007593533487297921,
"grad_norm": 2.0071773529052734,
"learning_rate": 1.469471562785891e-05,
"loss": 0.0872,
"step": 411
},
{
"epoch": 0.007612009237875288,
"grad_norm": 1.5450119972229004,
"learning_rate": 1.4663866403398915e-05,
"loss": 0.0981,
"step": 412
},
{
"epoch": 0.0076304849884526555,
"grad_norm": 55.13638687133789,
"learning_rate": 1.463296035119862e-05,
"loss": 0.2762,
"step": 413
},
{
"epoch": 0.007648960739030023,
"grad_norm": 85.87316131591797,
"learning_rate": 1.4601997847838518e-05,
"loss": 0.3327,
"step": 414
},
{
"epoch": 0.00766743648960739,
"grad_norm": 2.9911603927612305,
"learning_rate": 1.4570979270586944e-05,
"loss": 0.0677,
"step": 415
},
{
"epoch": 0.007685912240184757,
"grad_norm": 1.5465950965881348,
"learning_rate": 1.4539904997395468e-05,
"loss": 0.0818,
"step": 416
},
{
"epoch": 0.007704387990762124,
"grad_norm": 11.56525707244873,
"learning_rate": 1.4508775406894308e-05,
"loss": 0.108,
"step": 417
},
{
"epoch": 0.007722863741339492,
"grad_norm": 1.2624751329421997,
"learning_rate": 1.4477590878387697e-05,
"loss": 0.0737,
"step": 418
},
{
"epoch": 0.0077413394919168595,
"grad_norm": 7.4912800788879395,
"learning_rate": 1.4446351791849276e-05,
"loss": 0.129,
"step": 419
},
{
"epoch": 0.007759815242494227,
"grad_norm": 34.10211944580078,
"learning_rate": 1.4415058527917454e-05,
"loss": 0.1246,
"step": 420
},
{
"epoch": 0.007778290993071594,
"grad_norm": 1.24342942237854,
"learning_rate": 1.4383711467890776e-05,
"loss": 0.0842,
"step": 421
},
{
"epoch": 0.007796766743648961,
"grad_norm": 0.9661096930503845,
"learning_rate": 1.4352310993723277e-05,
"loss": 0.0762,
"step": 422
},
{
"epoch": 0.007815242494226328,
"grad_norm": 2.1040561199188232,
"learning_rate": 1.4320857488019826e-05,
"loss": 0.0825,
"step": 423
},
{
"epoch": 0.007833718244803695,
"grad_norm": 4.445217132568359,
"learning_rate": 1.4289351334031461e-05,
"loss": 0.0998,
"step": 424
},
{
"epoch": 0.007852193995381063,
"grad_norm": 1.1083904504776,
"learning_rate": 1.4257792915650728e-05,
"loss": 0.0716,
"step": 425
},
{
"epoch": 0.007870669745958429,
"grad_norm": 82.57806396484375,
"learning_rate": 1.4226182617406996e-05,
"loss": 0.1667,
"step": 426
},
{
"epoch": 0.007889145496535797,
"grad_norm": 1.0734587907791138,
"learning_rate": 1.4194520824461773e-05,
"loss": 0.0686,
"step": 427
},
{
"epoch": 0.007907621247113163,
"grad_norm": 1.960877776145935,
"learning_rate": 1.4162807922604014e-05,
"loss": 0.0891,
"step": 428
},
{
"epoch": 0.007926096997690531,
"grad_norm": 0.877657413482666,
"learning_rate": 1.413104429824542e-05,
"loss": 0.0744,
"step": 429
},
{
"epoch": 0.007944572748267898,
"grad_norm": 6.067645072937012,
"learning_rate": 1.4099230338415728e-05,
"loss": 0.0956,
"step": 430
},
{
"epoch": 0.007963048498845266,
"grad_norm": 1.915191411972046,
"learning_rate": 1.4067366430758004e-05,
"loss": 0.089,
"step": 431
},
{
"epoch": 0.007981524249422632,
"grad_norm": 43.639793395996094,
"learning_rate": 1.4035452963523903e-05,
"loss": 0.165,
"step": 432
},
{
"epoch": 0.008,
"grad_norm": 1.643744945526123,
"learning_rate": 1.4003490325568953e-05,
"loss": 0.0866,
"step": 433
},
{
"epoch": 0.008018475750577366,
"grad_norm": 1.4863110780715942,
"learning_rate": 1.3971478906347806e-05,
"loss": 0.0949,
"step": 434
},
{
"epoch": 0.008036951501154735,
"grad_norm": 2.204235792160034,
"learning_rate": 1.3939419095909513e-05,
"loss": 0.0906,
"step": 435
},
{
"epoch": 0.0080554272517321,
"grad_norm": 72.22257995605469,
"learning_rate": 1.3907311284892737e-05,
"loss": 0.3142,
"step": 436
},
{
"epoch": 0.008073903002309469,
"grad_norm": 1.8713430166244507,
"learning_rate": 1.3875155864521031e-05,
"loss": 0.0938,
"step": 437
},
{
"epoch": 0.008092378752886835,
"grad_norm": 45.01213073730469,
"learning_rate": 1.3842953226598036e-05,
"loss": 0.6181,
"step": 438
},
{
"epoch": 0.008110854503464203,
"grad_norm": 1.1124855279922485,
"learning_rate": 1.3810703763502744e-05,
"loss": 0.0839,
"step": 439
},
{
"epoch": 0.00812933025404157,
"grad_norm": 34.46360397338867,
"learning_rate": 1.3778407868184674e-05,
"loss": 0.1487,
"step": 440
},
{
"epoch": 0.008147806004618938,
"grad_norm": 59.2484245300293,
"learning_rate": 1.3746065934159123e-05,
"loss": 0.914,
"step": 441
},
{
"epoch": 0.008166281755196304,
"grad_norm": 0.7930397987365723,
"learning_rate": 1.371367835550235e-05,
"loss": 0.0679,
"step": 442
},
{
"epoch": 0.008184757505773672,
"grad_norm": 1.7843949794769287,
"learning_rate": 1.3681245526846782e-05,
"loss": 0.1016,
"step": 443
},
{
"epoch": 0.00820323325635104,
"grad_norm": 1.286493182182312,
"learning_rate": 1.3648767843376196e-05,
"loss": 0.0753,
"step": 444
},
{
"epoch": 0.008221709006928406,
"grad_norm": 12.253741264343262,
"learning_rate": 1.3616245700820922e-05,
"loss": 0.1041,
"step": 445
},
{
"epoch": 0.008240184757505774,
"grad_norm": 1.8959864377975464,
"learning_rate": 1.3583679495453e-05,
"loss": 0.0884,
"step": 446
},
{
"epoch": 0.00825866050808314,
"grad_norm": 2.067258596420288,
"learning_rate": 1.3551069624081372e-05,
"loss": 0.0786,
"step": 447
},
{
"epoch": 0.008277136258660509,
"grad_norm": 37.373207092285156,
"learning_rate": 1.3518416484047018e-05,
"loss": 0.654,
"step": 448
},
{
"epoch": 0.008295612009237875,
"grad_norm": 4.849969387054443,
"learning_rate": 1.3485720473218153e-05,
"loss": 0.1025,
"step": 449
},
{
"epoch": 0.008314087759815243,
"grad_norm": 37.40327072143555,
"learning_rate": 1.3452981989985347e-05,
"loss": 0.4715,
"step": 450
},
{
"epoch": 0.00833256351039261,
"grad_norm": 1.4369239807128906,
"learning_rate": 1.342020143325669e-05,
"loss": 0.0761,
"step": 451
},
{
"epoch": 0.008351039260969978,
"grad_norm": 92.4406509399414,
"learning_rate": 1.3387379202452917e-05,
"loss": 0.4892,
"step": 452
},
{
"epoch": 0.008369515011547344,
"grad_norm": 1.3850326538085938,
"learning_rate": 1.3354515697502552e-05,
"loss": 0.0897,
"step": 453
},
{
"epoch": 0.008387990762124712,
"grad_norm": 66.92438507080078,
"learning_rate": 1.3321611318837033e-05,
"loss": 0.1755,
"step": 454
},
{
"epoch": 0.008406466512702078,
"grad_norm": 2.475630283355713,
"learning_rate": 1.3288666467385834e-05,
"loss": 0.0911,
"step": 455
},
{
"epoch": 0.008424942263279446,
"grad_norm": 65.7319564819336,
"learning_rate": 1.3255681544571568e-05,
"loss": 0.2846,
"step": 456
},
{
"epoch": 0.008443418013856813,
"grad_norm": 1.4315398931503296,
"learning_rate": 1.3222656952305113e-05,
"loss": 0.0927,
"step": 457
},
{
"epoch": 0.00846189376443418,
"grad_norm": 34.59487533569336,
"learning_rate": 1.3189593092980701e-05,
"loss": 0.6073,
"step": 458
},
{
"epoch": 0.008480369515011547,
"grad_norm": 1.1338688135147095,
"learning_rate": 1.3156490369471026e-05,
"loss": 0.0844,
"step": 459
},
{
"epoch": 0.008498845265588915,
"grad_norm": 1.266749620437622,
"learning_rate": 1.3123349185122328e-05,
"loss": 0.0817,
"step": 460
},
{
"epoch": 0.008517321016166281,
"grad_norm": 11.099577903747559,
"learning_rate": 1.3090169943749475e-05,
"loss": 0.0979,
"step": 461
},
{
"epoch": 0.00853579676674365,
"grad_norm": 1.9373276233673096,
"learning_rate": 1.3056953049631059e-05,
"loss": 0.084,
"step": 462
},
{
"epoch": 0.008554272517321016,
"grad_norm": 2.967406988143921,
"learning_rate": 1.3023698907504447e-05,
"loss": 0.0841,
"step": 463
},
{
"epoch": 0.008572748267898384,
"grad_norm": 19.44832992553711,
"learning_rate": 1.2990407922560869e-05,
"loss": 0.1181,
"step": 464
},
{
"epoch": 0.00859122401847575,
"grad_norm": 4.346829891204834,
"learning_rate": 1.2957080500440469e-05,
"loss": 0.0919,
"step": 465
},
{
"epoch": 0.008609699769053118,
"grad_norm": 12.151569366455078,
"learning_rate": 1.2923717047227368e-05,
"loss": 0.1074,
"step": 466
},
{
"epoch": 0.008628175519630485,
"grad_norm": 30.505720138549805,
"learning_rate": 1.2890317969444716e-05,
"loss": 0.79,
"step": 467
},
{
"epoch": 0.008646651270207853,
"grad_norm": 33.179832458496094,
"learning_rate": 1.2856883674049736e-05,
"loss": 0.4086,
"step": 468
},
{
"epoch": 0.008665127020785219,
"grad_norm": 2.661017656326294,
"learning_rate": 1.2823414568428767e-05,
"loss": 0.0874,
"step": 469
},
{
"epoch": 0.008683602771362587,
"grad_norm": 2.305401563644409,
"learning_rate": 1.2789911060392295e-05,
"loss": 0.0884,
"step": 470
},
{
"epoch": 0.008702078521939953,
"grad_norm": 22.346948623657227,
"learning_rate": 1.2756373558169992e-05,
"loss": 0.5532,
"step": 471
},
{
"epoch": 0.008720554272517321,
"grad_norm": 3.430056571960449,
"learning_rate": 1.2722802470405744e-05,
"loss": 0.089,
"step": 472
},
{
"epoch": 0.008739030023094688,
"grad_norm": 42.60260009765625,
"learning_rate": 1.2689198206152657e-05,
"loss": 0.4945,
"step": 473
},
{
"epoch": 0.008757505773672056,
"grad_norm": 30.675615310668945,
"learning_rate": 1.265556117486809e-05,
"loss": 0.4231,
"step": 474
},
{
"epoch": 0.008775981524249422,
"grad_norm": 0.9028281569480896,
"learning_rate": 1.2621891786408648e-05,
"loss": 0.088,
"step": 475
},
{
"epoch": 0.00879445727482679,
"grad_norm": 34.931419372558594,
"learning_rate": 1.2588190451025209e-05,
"loss": 0.1855,
"step": 476
},
{
"epoch": 0.008812933025404156,
"grad_norm": 41.36872482299805,
"learning_rate": 1.2554457579357906e-05,
"loss": 0.2093,
"step": 477
},
{
"epoch": 0.008831408775981524,
"grad_norm": 29.953210830688477,
"learning_rate": 1.252069358243114e-05,
"loss": 0.2949,
"step": 478
},
{
"epoch": 0.00884988452655889,
"grad_norm": 1.8226795196533203,
"learning_rate": 1.2486898871648552e-05,
"loss": 0.0895,
"step": 479
},
{
"epoch": 0.008868360277136259,
"grad_norm": 9.829808235168457,
"learning_rate": 1.2453073858788027e-05,
"loss": 0.6862,
"step": 480
},
{
"epoch": 0.008886836027713625,
"grad_norm": 113.56571960449219,
"learning_rate": 1.2419218955996677e-05,
"loss": 0.3178,
"step": 481
},
{
"epoch": 0.008905311778290993,
"grad_norm": 11.500356674194336,
"learning_rate": 1.238533457578581e-05,
"loss": 0.1303,
"step": 482
},
{
"epoch": 0.00892378752886836,
"grad_norm": 2.5878233909606934,
"learning_rate": 1.23514211310259e-05,
"loss": 0.1055,
"step": 483
},
{
"epoch": 0.008942263279445728,
"grad_norm": 46.66258239746094,
"learning_rate": 1.2317479034941572e-05,
"loss": 0.2863,
"step": 484
},
{
"epoch": 0.008960739030023094,
"grad_norm": 12.498741149902344,
"learning_rate": 1.2283508701106559e-05,
"loss": 0.1641,
"step": 485
},
{
"epoch": 0.008979214780600462,
"grad_norm": 9.274971961975098,
"learning_rate": 1.2249510543438652e-05,
"loss": 0.1329,
"step": 486
},
{
"epoch": 0.008997690531177828,
"grad_norm": 53.34405517578125,
"learning_rate": 1.2215484976194675e-05,
"loss": 0.3297,
"step": 487
},
{
"epoch": 0.009016166281755196,
"grad_norm": 5.238002777099609,
"learning_rate": 1.2181432413965428e-05,
"loss": 0.1091,
"step": 488
},
{
"epoch": 0.009034642032332563,
"grad_norm": 1.4772626161575317,
"learning_rate": 1.2147353271670634e-05,
"loss": 0.09,
"step": 489
},
{
"epoch": 0.00905311778290993,
"grad_norm": 6.821254253387451,
"learning_rate": 1.211324796455389e-05,
"loss": 0.1033,
"step": 490
},
{
"epoch": 0.009071593533487297,
"grad_norm": 1.361069679260254,
"learning_rate": 1.2079116908177592e-05,
"loss": 0.0917,
"step": 491
},
{
"epoch": 0.009090069284064665,
"grad_norm": 7.930213451385498,
"learning_rate": 1.2044960518417902e-05,
"loss": 0.1509,
"step": 492
},
{
"epoch": 0.009108545034642033,
"grad_norm": 2.4302992820739746,
"learning_rate": 1.2010779211459649e-05,
"loss": 0.0943,
"step": 493
},
{
"epoch": 0.0091270207852194,
"grad_norm": 0.8986804485321045,
"learning_rate": 1.1976573403791263e-05,
"loss": 0.086,
"step": 494
},
{
"epoch": 0.009145496535796768,
"grad_norm": 30.513465881347656,
"learning_rate": 1.194234351219972e-05,
"loss": 0.5157,
"step": 495
},
{
"epoch": 0.009163972286374134,
"grad_norm": 1.4492926597595215,
"learning_rate": 1.190808995376545e-05,
"loss": 0.0824,
"step": 496
},
{
"epoch": 0.009182448036951502,
"grad_norm": 102.8843994140625,
"learning_rate": 1.187381314585725e-05,
"loss": 0.273,
"step": 497
},
{
"epoch": 0.009200923787528868,
"grad_norm": 1.4776283502578735,
"learning_rate": 1.1839513506127202e-05,
"loss": 0.0852,
"step": 498
},
{
"epoch": 0.009219399538106236,
"grad_norm": 3.0171475410461426,
"learning_rate": 1.1805191452505602e-05,
"loss": 0.0983,
"step": 499
},
{
"epoch": 0.009237875288683603,
"grad_norm": 19.791242599487305,
"learning_rate": 1.1770847403195836e-05,
"loss": 0.0942,
"step": 500
},
{
"epoch": 0.00925635103926097,
"grad_norm": 1.6341432332992554,
"learning_rate": 1.1736481776669307e-05,
"loss": 0.1053,
"step": 501
},
{
"epoch": 0.009274826789838337,
"grad_norm": 2.8659210205078125,
"learning_rate": 1.1702094991660326e-05,
"loss": 0.1047,
"step": 502
},
{
"epoch": 0.009293302540415705,
"grad_norm": 26.666948318481445,
"learning_rate": 1.1667687467161025e-05,
"loss": 0.0907,
"step": 503
},
{
"epoch": 0.009311778290993071,
"grad_norm": 1.1258450746536255,
"learning_rate": 1.1633259622416224e-05,
"loss": 0.0839,
"step": 504
},
{
"epoch": 0.00933025404157044,
"grad_norm": 4.354061603546143,
"learning_rate": 1.159881187691835e-05,
"loss": 0.1006,
"step": 505
},
{
"epoch": 0.009348729792147806,
"grad_norm": 2.5762939453125,
"learning_rate": 1.156434465040231e-05,
"loss": 0.0846,
"step": 506
},
{
"epoch": 0.009367205542725174,
"grad_norm": 3.2760705947875977,
"learning_rate": 1.1529858362840383e-05,
"loss": 0.0903,
"step": 507
},
{
"epoch": 0.00938568129330254,
"grad_norm": 43.53782272338867,
"learning_rate": 1.1495353434437098e-05,
"loss": 0.2984,
"step": 508
},
{
"epoch": 0.009404157043879908,
"grad_norm": 2.7441537380218506,
"learning_rate": 1.1460830285624119e-05,
"loss": 0.0849,
"step": 509
},
{
"epoch": 0.009422632794457275,
"grad_norm": 2.2179601192474365,
"learning_rate": 1.1426289337055119e-05,
"loss": 0.0743,
"step": 510
},
{
"epoch": 0.009441108545034643,
"grad_norm": 61.454898834228516,
"learning_rate": 1.1391731009600655e-05,
"loss": 0.1248,
"step": 511
},
{
"epoch": 0.009459584295612009,
"grad_norm": 0.9894300103187561,
"learning_rate": 1.1357155724343046e-05,
"loss": 0.0749,
"step": 512
},
{
"epoch": 0.009478060046189377,
"grad_norm": 1.5292553901672363,
"learning_rate": 1.1322563902571227e-05,
"loss": 0.0964,
"step": 513
},
{
"epoch": 0.009496535796766743,
"grad_norm": 142.1166229248047,
"learning_rate": 1.128795596577563e-05,
"loss": 0.2441,
"step": 514
},
{
"epoch": 0.009515011547344111,
"grad_norm": 3.0266716480255127,
"learning_rate": 1.1253332335643043e-05,
"loss": 0.0919,
"step": 515
},
{
"epoch": 0.009533487297921478,
"grad_norm": 142.0424346923828,
"learning_rate": 1.1218693434051475e-05,
"loss": 0.6646,
"step": 516
},
{
"epoch": 0.009551963048498846,
"grad_norm": 3.3894221782684326,
"learning_rate": 1.1184039683065014e-05,
"loss": 0.0985,
"step": 517
},
{
"epoch": 0.009570438799076212,
"grad_norm": 2.3916194438934326,
"learning_rate": 1.1149371504928667e-05,
"loss": 0.0889,
"step": 518
},
{
"epoch": 0.00958891454965358,
"grad_norm": 0.9956536293029785,
"learning_rate": 1.1114689322063255e-05,
"loss": 0.0741,
"step": 519
},
{
"epoch": 0.009607390300230946,
"grad_norm": 48.62922668457031,
"learning_rate": 1.1079993557060228e-05,
"loss": 0.5053,
"step": 520
},
{
"epoch": 0.009625866050808314,
"grad_norm": 28.549036026000977,
"learning_rate": 1.1045284632676535e-05,
"loss": 0.5209,
"step": 521
},
{
"epoch": 0.00964434180138568,
"grad_norm": 3.9586904048919678,
"learning_rate": 1.1010562971829464e-05,
"loss": 0.0886,
"step": 522
},
{
"epoch": 0.009662817551963049,
"grad_norm": 0.8221359848976135,
"learning_rate": 1.0975828997591496e-05,
"loss": 0.0746,
"step": 523
},
{
"epoch": 0.009681293302540415,
"grad_norm": 0.7924201488494873,
"learning_rate": 1.0941083133185146e-05,
"loss": 0.0762,
"step": 524
},
{
"epoch": 0.009699769053117783,
"grad_norm": 18.46653175354004,
"learning_rate": 1.0906325801977804e-05,
"loss": 0.136,
"step": 525
},
{
"epoch": 0.00971824480369515,
"grad_norm": 1.3309564590454102,
"learning_rate": 1.0871557427476585e-05,
"loss": 0.0807,
"step": 526
},
{
"epoch": 0.009736720554272518,
"grad_norm": 8.08139705657959,
"learning_rate": 1.083677843332316e-05,
"loss": 0.0946,
"step": 527
},
{
"epoch": 0.009755196304849884,
"grad_norm": 1.0622408390045166,
"learning_rate": 1.0801989243288588e-05,
"loss": 0.0769,
"step": 528
},
{
"epoch": 0.009773672055427252,
"grad_norm": 1.056046724319458,
"learning_rate": 1.0767190281268187e-05,
"loss": 0.077,
"step": 529
},
{
"epoch": 0.009792147806004618,
"grad_norm": 0.9565911889076233,
"learning_rate": 1.0732381971276318e-05,
"loss": 0.0825,
"step": 530
},
{
"epoch": 0.009810623556581986,
"grad_norm": 1.4339712858200073,
"learning_rate": 1.0697564737441254e-05,
"loss": 0.0761,
"step": 531
},
{
"epoch": 0.009829099307159353,
"grad_norm": 1.2778311967849731,
"learning_rate": 1.0662739004000005e-05,
"loss": 0.0865,
"step": 532
},
{
"epoch": 0.00984757505773672,
"grad_norm": 20.91969108581543,
"learning_rate": 1.0627905195293135e-05,
"loss": 0.1428,
"step": 533
},
{
"epoch": 0.009866050808314087,
"grad_norm": 3.2646594047546387,
"learning_rate": 1.0593063735759619e-05,
"loss": 0.0855,
"step": 534
},
{
"epoch": 0.009884526558891455,
"grad_norm": 1.0020149946212769,
"learning_rate": 1.055821504993164e-05,
"loss": 0.0813,
"step": 535
},
{
"epoch": 0.009903002309468821,
"grad_norm": 1.7093909978866577,
"learning_rate": 1.0523359562429441e-05,
"loss": 0.0818,
"step": 536
},
{
"epoch": 0.00992147806004619,
"grad_norm": 1.646415114402771,
"learning_rate": 1.0488497697956134e-05,
"loss": 0.0943,
"step": 537
},
{
"epoch": 0.009939953810623556,
"grad_norm": 14.313244819641113,
"learning_rate": 1.0453629881292537e-05,
"loss": 0.6456,
"step": 538
},
{
"epoch": 0.009958429561200924,
"grad_norm": 1.3055131435394287,
"learning_rate": 1.0418756537291996e-05,
"loss": 0.0883,
"step": 539
},
{
"epoch": 0.00997690531177829,
"grad_norm": 12.903365135192871,
"learning_rate": 1.03838780908752e-05,
"loss": 0.7288,
"step": 540
},
{
"epoch": 0.009995381062355658,
"grad_norm": 27.646595001220703,
"learning_rate": 1.0348994967025012e-05,
"loss": 0.4649,
"step": 541
},
{
"epoch": 0.010013856812933026,
"grad_norm": 2.27180814743042,
"learning_rate": 1.0314107590781284e-05,
"loss": 0.0737,
"step": 542
},
{
"epoch": 0.010032332563510393,
"grad_norm": 2.8047213554382324,
"learning_rate": 1.0279216387235691e-05,
"loss": 0.0901,
"step": 543
},
{
"epoch": 0.01005080831408776,
"grad_norm": 3.15432071685791,
"learning_rate": 1.0244321781526533e-05,
"loss": 0.0771,
"step": 544
},
{
"epoch": 0.010069284064665127,
"grad_norm": 1.1280345916748047,
"learning_rate": 1.0209424198833571e-05,
"loss": 0.0743,
"step": 545
},
{
"epoch": 0.010087759815242495,
"grad_norm": 8.223532676696777,
"learning_rate": 1.0174524064372837e-05,
"loss": 0.1027,
"step": 546
},
{
"epoch": 0.010106235565819861,
"grad_norm": 1.691619634628296,
"learning_rate": 1.0139621803391454e-05,
"loss": 0.077,
"step": 547
},
{
"epoch": 0.01012471131639723,
"grad_norm": 10.80276107788086,
"learning_rate": 1.010471784116246e-05,
"loss": 0.1217,
"step": 548
},
{
"epoch": 0.010143187066974596,
"grad_norm": 1.4029717445373535,
"learning_rate": 1.0069812602979617e-05,
"loss": 0.1009,
"step": 549
},
{
"epoch": 0.010161662817551964,
"grad_norm": 6.235980987548828,
"learning_rate": 1.0034906514152239e-05,
"loss": 0.0976,
"step": 550
},
{
"epoch": 0.01018013856812933,
"grad_norm": 47.17540740966797,
"learning_rate": 1e-05,
"loss": 0.2387,
"step": 551
},
{
"epoch": 0.010198614318706698,
"grad_norm": 3.9795498847961426,
"learning_rate": 9.965093485847766e-06,
"loss": 0.0852,
"step": 552
},
{
"epoch": 0.010217090069284064,
"grad_norm": 1.7837053537368774,
"learning_rate": 9.930187397020385e-06,
"loss": 0.0828,
"step": 553
},
{
"epoch": 0.010235565819861433,
"grad_norm": 42.76936340332031,
"learning_rate": 9.895282158837545e-06,
"loss": 0.4145,
"step": 554
},
{
"epoch": 0.010254041570438799,
"grad_norm": 25.35696029663086,
"learning_rate": 9.860378196608549e-06,
"loss": 0.5242,
"step": 555
},
{
"epoch": 0.010272517321016167,
"grad_norm": 2.0886199474334717,
"learning_rate": 9.825475935627165e-06,
"loss": 0.0808,
"step": 556
},
{
"epoch": 0.010290993071593533,
"grad_norm": 1.6074731349945068,
"learning_rate": 9.790575801166432e-06,
"loss": 0.0918,
"step": 557
},
{
"epoch": 0.010309468822170901,
"grad_norm": 0.9082817435264587,
"learning_rate": 9.75567821847347e-06,
"loss": 0.0854,
"step": 558
},
{
"epoch": 0.010327944572748268,
"grad_norm": 2.1920785903930664,
"learning_rate": 9.720783612764314e-06,
"loss": 0.0876,
"step": 559
},
{
"epoch": 0.010346420323325636,
"grad_norm": 1.7223331928253174,
"learning_rate": 9.685892409218718e-06,
"loss": 0.0765,
"step": 560
},
{
"epoch": 0.010364896073903002,
"grad_norm": 0.9665731191635132,
"learning_rate": 9.651005032974994e-06,
"loss": 0.0758,
"step": 561
},
{
"epoch": 0.01038337182448037,
"grad_norm": 42.741546630859375,
"learning_rate": 9.616121909124801e-06,
"loss": 0.1765,
"step": 562
},
{
"epoch": 0.010401847575057736,
"grad_norm": 1.223375678062439,
"learning_rate": 9.581243462708007e-06,
"loss": 0.073,
"step": 563
},
{
"epoch": 0.010420323325635104,
"grad_norm": 45.063392639160156,
"learning_rate": 9.546370118707463e-06,
"loss": 0.6437,
"step": 564
},
{
"epoch": 0.01043879907621247,
"grad_norm": 28.241010665893555,
"learning_rate": 9.511502302043867e-06,
"loss": 0.1212,
"step": 565
},
{
"epoch": 0.010457274826789839,
"grad_norm": 0.9051803946495056,
"learning_rate": 9.476640437570562e-06,
"loss": 0.0676,
"step": 566
},
{
"epoch": 0.010475750577367205,
"grad_norm": 20.545042037963867,
"learning_rate": 9.441784950068362e-06,
"loss": 0.0824,
"step": 567
},
{
"epoch": 0.010494226327944573,
"grad_norm": 2.053248167037964,
"learning_rate": 9.406936264240386e-06,
"loss": 0.0839,
"step": 568
},
{
"epoch": 0.01051270207852194,
"grad_norm": 1.8984928131103516,
"learning_rate": 9.372094804706867e-06,
"loss": 0.0771,
"step": 569
},
{
"epoch": 0.010531177829099308,
"grad_norm": 1.2789233922958374,
"learning_rate": 9.337260996000002e-06,
"loss": 0.0856,
"step": 570
},
{
"epoch": 0.010549653579676674,
"grad_norm": 26.48085594177246,
"learning_rate": 9.302435262558748e-06,
"loss": 0.5068,
"step": 571
},
{
"epoch": 0.010568129330254042,
"grad_norm": 68.56343841552734,
"learning_rate": 9.267618028723687e-06,
"loss": 0.3568,
"step": 572
},
{
"epoch": 0.010586605080831408,
"grad_norm": 1.220786213874817,
"learning_rate": 9.232809718731815e-06,
"loss": 0.0683,
"step": 573
},
{
"epoch": 0.010605080831408776,
"grad_norm": 3.250833511352539,
"learning_rate": 9.198010756711413e-06,
"loss": 0.1004,
"step": 574
},
{
"epoch": 0.010623556581986143,
"grad_norm": 1.1265226602554321,
"learning_rate": 9.163221566676847e-06,
"loss": 0.0728,
"step": 575
},
{
"epoch": 0.01064203233256351,
"grad_norm": 2.18690824508667,
"learning_rate": 9.128442572523418e-06,
"loss": 0.0914,
"step": 576
},
{
"epoch": 0.010660508083140877,
"grad_norm": 1.7627544403076172,
"learning_rate": 9.093674198022201e-06,
"loss": 0.0874,
"step": 577
},
{
"epoch": 0.010678983833718245,
"grad_norm": 1.2840354442596436,
"learning_rate": 9.058916866814857e-06,
"loss": 0.0823,
"step": 578
},
{
"epoch": 0.010697459584295611,
"grad_norm": 20.308652877807617,
"learning_rate": 9.024171002408507e-06,
"loss": 0.1156,
"step": 579
},
{
"epoch": 0.01071593533487298,
"grad_norm": 12.058526992797852,
"learning_rate": 8.989437028170537e-06,
"loss": 0.0976,
"step": 580
},
{
"epoch": 0.010734411085450346,
"grad_norm": 77.19605255126953,
"learning_rate": 8.954715367323468e-06,
"loss": 0.2098,
"step": 581
},
{
"epoch": 0.010752886836027714,
"grad_norm": 1.1154943704605103,
"learning_rate": 8.920006442939772e-06,
"loss": 0.0723,
"step": 582
},
{
"epoch": 0.01077136258660508,
"grad_norm": 1.0353710651397705,
"learning_rate": 8.885310677936746e-06,
"loss": 0.0714,
"step": 583
},
{
"epoch": 0.010789838337182448,
"grad_norm": 1.08982253074646,
"learning_rate": 8.850628495071336e-06,
"loss": 0.0813,
"step": 584
},
{
"epoch": 0.010808314087759814,
"grad_norm": 1.1713364124298096,
"learning_rate": 8.815960316934991e-06,
"loss": 0.0753,
"step": 585
},
{
"epoch": 0.010826789838337183,
"grad_norm": 2.217531442642212,
"learning_rate": 8.781306565948528e-06,
"loss": 0.0901,
"step": 586
},
{
"epoch": 0.010845265588914549,
"grad_norm": 122.17057037353516,
"learning_rate": 8.746667664356957e-06,
"loss": 0.2874,
"step": 587
},
{
"epoch": 0.010863741339491917,
"grad_norm": 0.7070766091346741,
"learning_rate": 8.712044034224374e-06,
"loss": 0.0665,
"step": 588
},
{
"epoch": 0.010882217090069283,
"grad_norm": 0.636335015296936,
"learning_rate": 8.677436097428775e-06,
"loss": 0.071,
"step": 589
},
{
"epoch": 0.010900692840646651,
"grad_norm": 2.0750539302825928,
"learning_rate": 8.642844275656957e-06,
"loss": 0.082,
"step": 590
},
{
"epoch": 0.01091916859122402,
"grad_norm": 1.267890214920044,
"learning_rate": 8.60826899039935e-06,
"loss": 0.0637,
"step": 591
},
{
"epoch": 0.010937644341801386,
"grad_norm": 1.7106521129608154,
"learning_rate": 8.573710662944884e-06,
"loss": 0.0722,
"step": 592
},
{
"epoch": 0.010956120092378754,
"grad_norm": 27.218259811401367,
"learning_rate": 8.539169714375885e-06,
"loss": 0.1127,
"step": 593
},
{
"epoch": 0.01097459584295612,
"grad_norm": 0.7766141295433044,
"learning_rate": 8.504646565562907e-06,
"loss": 0.065,
"step": 594
},
{
"epoch": 0.010993071593533488,
"grad_norm": 8.754508972167969,
"learning_rate": 8.47014163715962e-06,
"loss": 0.0902,
"step": 595
},
{
"epoch": 0.011011547344110854,
"grad_norm": 1.9367616176605225,
"learning_rate": 8.43565534959769e-06,
"loss": 0.0897,
"step": 596
},
{
"epoch": 0.011030023094688222,
"grad_norm": 2.6147475242614746,
"learning_rate": 8.401188123081653e-06,
"loss": 0.1039,
"step": 597
},
{
"epoch": 0.011048498845265589,
"grad_norm": 1.3627548217773438,
"learning_rate": 8.366740377583781e-06,
"loss": 0.0675,
"step": 598
},
{
"epoch": 0.011066974595842957,
"grad_norm": 1.5028516054153442,
"learning_rate": 8.332312532838978e-06,
"loss": 0.0879,
"step": 599
},
{
"epoch": 0.011085450346420323,
"grad_norm": 1.2498700618743896,
"learning_rate": 8.297905008339677e-06,
"loss": 0.0817,
"step": 600
},
{
"epoch": 0.011103926096997691,
"grad_norm": 1.7854423522949219,
"learning_rate": 8.263518223330698e-06,
"loss": 0.077,
"step": 601
},
{
"epoch": 0.011122401847575058,
"grad_norm": 1.48270845413208,
"learning_rate": 8.22915259680417e-06,
"loss": 0.0876,
"step": 602
},
{
"epoch": 0.011140877598152426,
"grad_norm": 1.0970808267593384,
"learning_rate": 8.194808547494401e-06,
"loss": 0.0698,
"step": 603
},
{
"epoch": 0.011159353348729792,
"grad_norm": 0.7956103086471558,
"learning_rate": 8.1604864938728e-06,
"loss": 0.0701,
"step": 604
},
{
"epoch": 0.01117782909930716,
"grad_norm": 3.7058029174804688,
"learning_rate": 8.126186854142752e-06,
"loss": 0.1182,
"step": 605
},
{
"epoch": 0.011196304849884526,
"grad_norm": 65.4359130859375,
"learning_rate": 8.091910046234552e-06,
"loss": 0.2825,
"step": 606
},
{
"epoch": 0.011214780600461894,
"grad_norm": 1.5197153091430664,
"learning_rate": 8.057656487800283e-06,
"loss": 0.0818,
"step": 607
},
{
"epoch": 0.01123325635103926,
"grad_norm": 53.51100540161133,
"learning_rate": 8.023426596208739e-06,
"loss": 0.2173,
"step": 608
},
{
"epoch": 0.011251732101616629,
"grad_norm": 16.92856216430664,
"learning_rate": 7.989220788540356e-06,
"loss": 0.094,
"step": 609
},
{
"epoch": 0.011270207852193995,
"grad_norm": 1.8352267742156982,
"learning_rate": 7.955039481582098e-06,
"loss": 0.085,
"step": 610
},
{
"epoch": 0.011288683602771363,
"grad_norm": 2.650439977645874,
"learning_rate": 7.92088309182241e-06,
"loss": 0.0832,
"step": 611
},
{
"epoch": 0.01130715935334873,
"grad_norm": 2.7688019275665283,
"learning_rate": 7.886752035446116e-06,
"loss": 0.0797,
"step": 612
},
{
"epoch": 0.011325635103926097,
"grad_norm": 1.308262825012207,
"learning_rate": 7.852646728329368e-06,
"loss": 0.0687,
"step": 613
},
{
"epoch": 0.011344110854503464,
"grad_norm": 1.2380762100219727,
"learning_rate": 7.818567586034578e-06,
"loss": 0.074,
"step": 614
},
{
"epoch": 0.011362586605080832,
"grad_norm": 1.6561496257781982,
"learning_rate": 7.784515023805328e-06,
"loss": 0.0809,
"step": 615
},
{
"epoch": 0.011381062355658198,
"grad_norm": 4.231128692626953,
"learning_rate": 7.750489456561351e-06,
"loss": 0.1328,
"step": 616
},
{
"epoch": 0.011399538106235566,
"grad_norm": 0.8487908840179443,
"learning_rate": 7.716491298893443e-06,
"loss": 0.0665,
"step": 617
},
{
"epoch": 0.011418013856812933,
"grad_norm": 6.0521769523620605,
"learning_rate": 7.68252096505843e-06,
"loss": 0.0934,
"step": 618
},
{
"epoch": 0.0114364896073903,
"grad_norm": 2.021629571914673,
"learning_rate": 7.6485788689741e-06,
"loss": 0.0844,
"step": 619
},
{
"epoch": 0.011454965357967667,
"grad_norm": 33.949710845947266,
"learning_rate": 7.6146654242141935e-06,
"loss": 0.1208,
"step": 620
},
{
"epoch": 0.011473441108545035,
"grad_norm": 7.253255844116211,
"learning_rate": 7.580781044003324e-06,
"loss": 0.1019,
"step": 621
},
{
"epoch": 0.011491916859122401,
"grad_norm": 2.313133478164673,
"learning_rate": 7.546926141211975e-06,
"loss": 0.0964,
"step": 622
},
{
"epoch": 0.01151039260969977,
"grad_norm": 99.91919708251953,
"learning_rate": 7.513101128351454e-06,
"loss": 0.363,
"step": 623
},
{
"epoch": 0.011528868360277136,
"grad_norm": 1.0669208765029907,
"learning_rate": 7.4793064175688635e-06,
"loss": 0.0774,
"step": 624
},
{
"epoch": 0.011547344110854504,
"grad_norm": 6.171936988830566,
"learning_rate": 7.445542420642097e-06,
"loss": 0.1295,
"step": 625
},
{
"epoch": 0.01156581986143187,
"grad_norm": 35.60003662109375,
"learning_rate": 7.411809548974792e-06,
"loss": 0.1533,
"step": 626
},
{
"epoch": 0.011584295612009238,
"grad_norm": 38.06922149658203,
"learning_rate": 7.378108213591355e-06,
"loss": 0.4529,
"step": 627
},
{
"epoch": 0.011602771362586604,
"grad_norm": 140.7566680908203,
"learning_rate": 7.344438825131912e-06,
"loss": 0.3357,
"step": 628
},
{
"epoch": 0.011621247113163972,
"grad_norm": 1.9510130882263184,
"learning_rate": 7.310801793847344e-06,
"loss": 0.0875,
"step": 629
},
{
"epoch": 0.011639722863741339,
"grad_norm": 124.25992584228516,
"learning_rate": 7.277197529594257e-06,
"loss": 0.2535,
"step": 630
},
{
"epoch": 0.011658198614318707,
"grad_norm": 5.500129222869873,
"learning_rate": 7.243626441830009e-06,
"loss": 0.1094,
"step": 631
},
{
"epoch": 0.011676674364896073,
"grad_norm": 58.37315368652344,
"learning_rate": 7.210088939607709e-06,
"loss": 0.3436,
"step": 632
},
{
"epoch": 0.011695150115473441,
"grad_norm": 25.11495590209961,
"learning_rate": 7.176585431571235e-06,
"loss": 0.0884,
"step": 633
},
{
"epoch": 0.011713625866050808,
"grad_norm": 45.103607177734375,
"learning_rate": 7.143116325950266e-06,
"loss": 0.1764,
"step": 634
},
{
"epoch": 0.011732101616628176,
"grad_norm": 58.846370697021484,
"learning_rate": 7.109682030555283e-06,
"loss": 0.3387,
"step": 635
},
{
"epoch": 0.011750577367205542,
"grad_norm": 1.893760085105896,
"learning_rate": 7.076282952772634e-06,
"loss": 0.0814,
"step": 636
},
{
"epoch": 0.01176905311778291,
"grad_norm": 24.308069229125977,
"learning_rate": 7.042919499559538e-06,
"loss": 0.5686,
"step": 637
},
{
"epoch": 0.011787528868360276,
"grad_norm": 1.4071834087371826,
"learning_rate": 7.009592077439135e-06,
"loss": 0.0769,
"step": 638
},
{
"epoch": 0.011806004618937644,
"grad_norm": 3.9436771869659424,
"learning_rate": 6.976301092495556e-06,
"loss": 0.0949,
"step": 639
},
{
"epoch": 0.011824480369515012,
"grad_norm": 0.9565290212631226,
"learning_rate": 6.943046950368944e-06,
"loss": 0.0701,
"step": 640
},
{
"epoch": 0.011842956120092379,
"grad_norm": 2.9073293209075928,
"learning_rate": 6.909830056250527e-06,
"loss": 0.0962,
"step": 641
},
{
"epoch": 0.011861431870669747,
"grad_norm": 53.04655456542969,
"learning_rate": 6.876650814877675e-06,
"loss": 0.7598,
"step": 642
},
{
"epoch": 0.011879907621247113,
"grad_norm": 43.70806121826172,
"learning_rate": 6.843509630528977e-06,
"loss": 0.1921,
"step": 643
},
{
"epoch": 0.011898383371824481,
"grad_norm": 0.8185755014419556,
"learning_rate": 6.8104069070193e-06,
"loss": 0.0704,
"step": 644
},
{
"epoch": 0.011916859122401847,
"grad_norm": 8.80150032043457,
"learning_rate": 6.777343047694891e-06,
"loss": 0.0724,
"step": 645
},
{
"epoch": 0.011935334872979216,
"grad_norm": 1.0680536031723022,
"learning_rate": 6.744318455428436e-06,
"loss": 0.0626,
"step": 646
},
{
"epoch": 0.011953810623556582,
"grad_norm": 1.2333645820617676,
"learning_rate": 6.711333532614168e-06,
"loss": 0.0749,
"step": 647
},
{
"epoch": 0.01197228637413395,
"grad_norm": 1.088052749633789,
"learning_rate": 6.67838868116297e-06,
"loss": 0.0682,
"step": 648
},
{
"epoch": 0.011990762124711316,
"grad_norm": 38.39015579223633,
"learning_rate": 6.645484302497452e-06,
"loss": 0.1152,
"step": 649
},
{
"epoch": 0.012009237875288684,
"grad_norm": 2.6487839221954346,
"learning_rate": 6.612620797547087e-06,
"loss": 0.0744,
"step": 650
},
{
"epoch": 0.01202771362586605,
"grad_norm": 1.6891621351242065,
"learning_rate": 6.579798566743314e-06,
"loss": 0.0904,
"step": 651
},
{
"epoch": 0.012046189376443419,
"grad_norm": 3.798185110092163,
"learning_rate": 6.547018010014654e-06,
"loss": 0.1056,
"step": 652
},
{
"epoch": 0.012064665127020785,
"grad_norm": 2.3351097106933594,
"learning_rate": 6.5142795267818505e-06,
"loss": 0.079,
"step": 653
},
{
"epoch": 0.012083140877598153,
"grad_norm": 2.22590970993042,
"learning_rate": 6.481583515952983e-06,
"loss": 0.0949,
"step": 654
},
{
"epoch": 0.01210161662817552,
"grad_norm": 0.8257840871810913,
"learning_rate": 6.448930375918632e-06,
"loss": 0.0646,
"step": 655
},
{
"epoch": 0.012120092378752887,
"grad_norm": 1.594437599182129,
"learning_rate": 6.4163205045469975e-06,
"loss": 0.0815,
"step": 656
},
{
"epoch": 0.012138568129330254,
"grad_norm": 1.2161691188812256,
"learning_rate": 6.383754299179079e-06,
"loss": 0.0726,
"step": 657
},
{
"epoch": 0.012157043879907622,
"grad_norm": 1.9872711896896362,
"learning_rate": 6.351232156623803e-06,
"loss": 0.0869,
"step": 658
},
{
"epoch": 0.012175519630484988,
"grad_norm": 261.1438293457031,
"learning_rate": 6.318754473153221e-06,
"loss": 0.3751,
"step": 659
},
{
"epoch": 0.012193995381062356,
"grad_norm": 1.2866449356079102,
"learning_rate": 6.286321644497655e-06,
"loss": 0.0708,
"step": 660
},
{
"epoch": 0.012212471131639722,
"grad_norm": 1.609950304031372,
"learning_rate": 6.25393406584088e-06,
"loss": 0.088,
"step": 661
},
{
"epoch": 0.01223094688221709,
"grad_norm": 1.0337163209915161,
"learning_rate": 6.22159213181533e-06,
"loss": 0.0616,
"step": 662
},
{
"epoch": 0.012249422632794457,
"grad_norm": 1.0992825031280518,
"learning_rate": 6.18929623649726e-06,
"loss": 0.0823,
"step": 663
},
{
"epoch": 0.012267898383371825,
"grad_norm": 1.4857592582702637,
"learning_rate": 6.157046773401964e-06,
"loss": 0.0929,
"step": 664
},
{
"epoch": 0.012286374133949191,
"grad_norm": 0.9879541397094727,
"learning_rate": 6.124844135478971e-06,
"loss": 0.0723,
"step": 665
},
{
"epoch": 0.01230484988452656,
"grad_norm": 32.7642822265625,
"learning_rate": 6.092688715107265e-06,
"loss": 0.5363,
"step": 666
},
{
"epoch": 0.012323325635103926,
"grad_norm": 34.64175796508789,
"learning_rate": 6.06058090409049e-06,
"loss": 0.1466,
"step": 667
},
{
"epoch": 0.012341801385681294,
"grad_norm": 37.98980712890625,
"learning_rate": 6.028521093652195e-06,
"loss": 0.19,
"step": 668
},
{
"epoch": 0.01236027713625866,
"grad_norm": 6.283631801605225,
"learning_rate": 5.996509674431053e-06,
"loss": 0.0835,
"step": 669
},
{
"epoch": 0.012378752886836028,
"grad_norm": 23.262157440185547,
"learning_rate": 5.9645470364761e-06,
"loss": 0.1091,
"step": 670
},
{
"epoch": 0.012397228637413394,
"grad_norm": 59.45625686645508,
"learning_rate": 5.932633569242e-06,
"loss": 0.5,
"step": 671
},
{
"epoch": 0.012415704387990762,
"grad_norm": 1.9454747438430786,
"learning_rate": 5.900769661584273e-06,
"loss": 0.0891,
"step": 672
},
{
"epoch": 0.012434180138568129,
"grad_norm": 14.132351875305176,
"learning_rate": 5.868955701754584e-06,
"loss": 0.0969,
"step": 673
},
{
"epoch": 0.012452655889145497,
"grad_norm": 1.6335331201553345,
"learning_rate": 5.83719207739599e-06,
"loss": 0.0778,
"step": 674
},
{
"epoch": 0.012471131639722863,
"grad_norm": 4.12040376663208,
"learning_rate": 5.8054791755382286e-06,
"loss": 0.0904,
"step": 675
},
{
"epoch": 0.012489607390300231,
"grad_norm": 1.8990004062652588,
"learning_rate": 5.773817382593008e-06,
"loss": 0.0814,
"step": 676
},
{
"epoch": 0.012508083140877597,
"grad_norm": 4.157707691192627,
"learning_rate": 5.742207084349274e-06,
"loss": 0.091,
"step": 677
},
{
"epoch": 0.012526558891454966,
"grad_norm": 1.2629783153533936,
"learning_rate": 5.710648665968543e-06,
"loss": 0.0725,
"step": 678
},
{
"epoch": 0.012545034642032332,
"grad_norm": 1.3766577243804932,
"learning_rate": 5.679142511980176e-06,
"loss": 0.0846,
"step": 679
},
{
"epoch": 0.0125635103926097,
"grad_norm": 1.1314102411270142,
"learning_rate": 5.647689006276727e-06,
"loss": 0.0653,
"step": 680
},
{
"epoch": 0.012581986143187066,
"grad_norm": 2.0009992122650146,
"learning_rate": 5.616288532109225e-06,
"loss": 0.088,
"step": 681
},
{
"epoch": 0.012600461893764434,
"grad_norm": 0.892760694026947,
"learning_rate": 5.584941472082549e-06,
"loss": 0.0693,
"step": 682
},
{
"epoch": 0.0126189376443418,
"grad_norm": 1.5125826597213745,
"learning_rate": 5.553648208150728e-06,
"loss": 0.0699,
"step": 683
},
{
"epoch": 0.012637413394919169,
"grad_norm": 6.661672115325928,
"learning_rate": 5.522409121612304e-06,
"loss": 0.0862,
"step": 684
},
{
"epoch": 0.012655889145496535,
"grad_norm": 1.3750674724578857,
"learning_rate": 5.491224593105695e-06,
"loss": 0.07,
"step": 685
},
{
"epoch": 0.012674364896073903,
"grad_norm": 4.761937618255615,
"learning_rate": 5.460095002604533e-06,
"loss": 0.1687,
"step": 686
},
{
"epoch": 0.01269284064665127,
"grad_norm": 5.275146484375,
"learning_rate": 5.429020729413062e-06,
"loss": 0.0861,
"step": 687
},
{
"epoch": 0.012711316397228637,
"grad_norm": 4.6939873695373535,
"learning_rate": 5.398002152161484e-06,
"loss": 0.0879,
"step": 688
},
{
"epoch": 0.012729792147806005,
"grad_norm": 8.297616004943848,
"learning_rate": 5.367039648801386e-06,
"loss": 0.0764,
"step": 689
},
{
"epoch": 0.012748267898383372,
"grad_norm": 1.4599788188934326,
"learning_rate": 5.336133596601089e-06,
"loss": 0.0849,
"step": 690
},
{
"epoch": 0.01276674364896074,
"grad_norm": 0.975753664970398,
"learning_rate": 5.305284372141095e-06,
"loss": 0.0665,
"step": 691
},
{
"epoch": 0.012785219399538106,
"grad_norm": 2.5845601558685303,
"learning_rate": 5.274492351309462e-06,
"loss": 0.0923,
"step": 692
},
{
"epoch": 0.012803695150115474,
"grad_norm": 25.161684036254883,
"learning_rate": 5.243757909297247e-06,
"loss": 0.169,
"step": 693
},
{
"epoch": 0.01282217090069284,
"grad_norm": 1.5734552145004272,
"learning_rate": 5.213081420593933e-06,
"loss": 0.0879,
"step": 694
},
{
"epoch": 0.012840646651270209,
"grad_norm": 0.9783046245574951,
"learning_rate": 5.1824632589828465e-06,
"loss": 0.0741,
"step": 695
},
{
"epoch": 0.012859122401847575,
"grad_norm": 1.5255866050720215,
"learning_rate": 5.151903797536631e-06,
"loss": 0.0687,
"step": 696
},
{
"epoch": 0.012877598152424943,
"grad_norm": 1.0195003747940063,
"learning_rate": 5.121403408612672e-06,
"loss": 0.0856,
"step": 697
},
{
"epoch": 0.01289607390300231,
"grad_norm": 34.405548095703125,
"learning_rate": 5.090962463848592e-06,
"loss": 0.164,
"step": 698
},
{
"epoch": 0.012914549653579677,
"grad_norm": 7.046591281890869,
"learning_rate": 5.060581334157693e-06,
"loss": 0.0914,
"step": 699
},
{
"epoch": 0.012933025404157044,
"grad_norm": 2.146393299102783,
"learning_rate": 5.030260389724447e-06,
"loss": 0.0797,
"step": 700
},
{
"epoch": 0.012951501154734412,
"grad_norm": 3.532819986343384,
"learning_rate": 5.000000000000003e-06,
"loss": 0.0681,
"step": 701
},
{
"epoch": 0.012969976905311778,
"grad_norm": 50.249080657958984,
"learning_rate": 4.96980053369765e-06,
"loss": 0.1462,
"step": 702
},
{
"epoch": 0.012988452655889146,
"grad_norm": 1.452976942062378,
"learning_rate": 4.939662358788364e-06,
"loss": 0.074,
"step": 703
},
{
"epoch": 0.013006928406466512,
"grad_norm": 8.007773399353027,
"learning_rate": 4.909585842496287e-06,
"loss": 0.1154,
"step": 704
},
{
"epoch": 0.01302540415704388,
"grad_norm": 0.7397992610931396,
"learning_rate": 4.879571351294287e-06,
"loss": 0.075,
"step": 705
},
{
"epoch": 0.013043879907621247,
"grad_norm": 3.6350057125091553,
"learning_rate": 4.849619250899458e-06,
"loss": 0.087,
"step": 706
},
{
"epoch": 0.013062355658198615,
"grad_norm": 8.616110801696777,
"learning_rate": 4.8197299062687e-06,
"loss": 0.0912,
"step": 707
},
{
"epoch": 0.013080831408775981,
"grad_norm": 4.417272090911865,
"learning_rate": 4.78990368159424e-06,
"loss": 0.0924,
"step": 708
},
{
"epoch": 0.01309930715935335,
"grad_norm": 0.9845334887504578,
"learning_rate": 4.76014094029921e-06,
"loss": 0.0698,
"step": 709
},
{
"epoch": 0.013117782909930716,
"grad_norm": 1.6458173990249634,
"learning_rate": 4.7304420450332244e-06,
"loss": 0.0774,
"step": 710
},
{
"epoch": 0.013136258660508084,
"grad_norm": 21.00367546081543,
"learning_rate": 4.700807357667953e-06,
"loss": 0.583,
"step": 711
},
{
"epoch": 0.01315473441108545,
"grad_norm": 11.204732894897461,
"learning_rate": 4.671237239292699e-06,
"loss": 0.1848,
"step": 712
},
{
"epoch": 0.013173210161662818,
"grad_norm": 42.35819625854492,
"learning_rate": 4.641732050210032e-06,
"loss": 0.181,
"step": 713
},
{
"epoch": 0.013191685912240184,
"grad_norm": 4.871307373046875,
"learning_rate": 4.612292149931369e-06,
"loss": 0.1004,
"step": 714
},
{
"epoch": 0.013210161662817552,
"grad_norm": 1.5525139570236206,
"learning_rate": 4.582917897172603e-06,
"loss": 0.0786,
"step": 715
},
{
"epoch": 0.013228637413394919,
"grad_norm": 1.3929939270019531,
"learning_rate": 4.5536096498497295e-06,
"loss": 0.0863,
"step": 716
},
{
"epoch": 0.013247113163972287,
"grad_norm": 1.1072086095809937,
"learning_rate": 4.524367765074499e-06,
"loss": 0.0739,
"step": 717
},
{
"epoch": 0.013265588914549653,
"grad_norm": 1.17701256275177,
"learning_rate": 4.495192599150045e-06,
"loss": 0.0735,
"step": 718
},
{
"epoch": 0.013284064665127021,
"grad_norm": 32.822998046875,
"learning_rate": 4.46608450756656e-06,
"loss": 0.1069,
"step": 719
},
{
"epoch": 0.013302540415704387,
"grad_norm": 1.1323364973068237,
"learning_rate": 4.437043844996952e-06,
"loss": 0.0819,
"step": 720
},
{
"epoch": 0.013321016166281755,
"grad_norm": 0.8521149158477783,
"learning_rate": 4.408070965292534e-06,
"loss": 0.0713,
"step": 721
},
{
"epoch": 0.013339491916859122,
"grad_norm": 1.0431668758392334,
"learning_rate": 4.379166221478697e-06,
"loss": 0.0697,
"step": 722
},
{
"epoch": 0.01335796766743649,
"grad_norm": 1.2528904676437378,
"learning_rate": 4.350329965750622e-06,
"loss": 0.077,
"step": 723
},
{
"epoch": 0.013376443418013856,
"grad_norm": 1.3092106580734253,
"learning_rate": 4.321562549468991e-06,
"loss": 0.0648,
"step": 724
},
{
"epoch": 0.013394919168591224,
"grad_norm": 1.2038995027542114,
"learning_rate": 4.292864323155684e-06,
"loss": 0.0616,
"step": 725
},
{
"epoch": 0.01341339491916859,
"grad_norm": 4.113232612609863,
"learning_rate": 4.264235636489542e-06,
"loss": 0.0877,
"step": 726
},
{
"epoch": 0.013431870669745959,
"grad_norm": 4.549839019775391,
"learning_rate": 4.235676838302069e-06,
"loss": 0.0834,
"step": 727
},
{
"epoch": 0.013450346420323325,
"grad_norm": 60.654945373535156,
"learning_rate": 4.207188276573214e-06,
"loss": 0.4605,
"step": 728
},
{
"epoch": 0.013468822170900693,
"grad_norm": 8.685653686523438,
"learning_rate": 4.178770298427107e-06,
"loss": 0.138,
"step": 729
},
{
"epoch": 0.01348729792147806,
"grad_norm": 2.5743298530578613,
"learning_rate": 4.150423250127846e-06,
"loss": 0.0926,
"step": 730
},
{
"epoch": 0.013505773672055427,
"grad_norm": 1.18416166305542,
"learning_rate": 4.12214747707527e-06,
"loss": 0.0847,
"step": 731
},
{
"epoch": 0.013524249422632794,
"grad_norm": 1.2563117742538452,
"learning_rate": 4.093943323800746e-06,
"loss": 0.0763,
"step": 732
},
{
"epoch": 0.013542725173210162,
"grad_norm": 1.6285808086395264,
"learning_rate": 4.065811133962987e-06,
"loss": 0.1043,
"step": 733
},
{
"epoch": 0.013561200923787528,
"grad_norm": 1.1073758602142334,
"learning_rate": 4.037751250343841e-06,
"loss": 0.0677,
"step": 734
},
{
"epoch": 0.013579676674364896,
"grad_norm": 67.54512023925781,
"learning_rate": 4.009764014844143e-06,
"loss": 0.7137,
"step": 735
},
{
"epoch": 0.013598152424942262,
"grad_norm": 1.2946701049804688,
"learning_rate": 3.981849768479516e-06,
"loss": 0.0704,
"step": 736
},
{
"epoch": 0.01361662817551963,
"grad_norm": 1.021460771560669,
"learning_rate": 3.954008851376252e-06,
"loss": 0.0779,
"step": 737
},
{
"epoch": 0.013635103926096999,
"grad_norm": 1.0310156345367432,
"learning_rate": 3.9262416027671354e-06,
"loss": 0.0829,
"step": 738
},
{
"epoch": 0.013653579676674365,
"grad_norm": 4.551070213317871,
"learning_rate": 3.898548360987325e-06,
"loss": 0.1096,
"step": 739
},
{
"epoch": 0.013672055427251733,
"grad_norm": 24.569658279418945,
"learning_rate": 3.8709294634702374e-06,
"loss": 0.4991,
"step": 740
},
{
"epoch": 0.0136905311778291,
"grad_norm": 53.204505920410156,
"learning_rate": 3.8433852467434175e-06,
"loss": 0.4099,
"step": 741
},
{
"epoch": 0.013709006928406467,
"grad_norm": 0.9719085693359375,
"learning_rate": 3.81591604642446e-06,
"loss": 0.0785,
"step": 742
},
{
"epoch": 0.013727482678983834,
"grad_norm": 25.742328643798828,
"learning_rate": 3.7885221972168974e-06,
"loss": 0.1041,
"step": 743
},
{
"epoch": 0.013745958429561202,
"grad_norm": 4.599763870239258,
"learning_rate": 3.7612040329061405e-06,
"loss": 0.0822,
"step": 744
},
{
"epoch": 0.013764434180138568,
"grad_norm": 1.4508581161499023,
"learning_rate": 3.7339618863553983e-06,
"loss": 0.0753,
"step": 745
},
{
"epoch": 0.013782909930715936,
"grad_norm": 1.3649834394454956,
"learning_rate": 3.7067960895016277e-06,
"loss": 0.072,
"step": 746
},
{
"epoch": 0.013801385681293302,
"grad_norm": 2.4947659969329834,
"learning_rate": 3.679706973351491e-06,
"loss": 0.0905,
"step": 747
},
{
"epoch": 0.01381986143187067,
"grad_norm": 77.44902038574219,
"learning_rate": 3.6526948679773256e-06,
"loss": 0.3307,
"step": 748
},
{
"epoch": 0.013838337182448037,
"grad_norm": 1.7825913429260254,
"learning_rate": 3.625760102513103e-06,
"loss": 0.0751,
"step": 749
},
{
"epoch": 0.013856812933025405,
"grad_norm": 1.4856318235397339,
"learning_rate": 3.598903005150444e-06,
"loss": 0.0822,
"step": 750
},
{
"epoch": 0.013875288683602771,
"grad_norm": 77.60884094238281,
"learning_rate": 3.5721239031346067e-06,
"loss": 0.9736,
"step": 751
},
{
"epoch": 0.01389376443418014,
"grad_norm": 2.233977794647217,
"learning_rate": 3.545423122760493e-06,
"loss": 0.0757,
"step": 752
},
{
"epoch": 0.013912240184757505,
"grad_norm": 108.72965240478516,
"learning_rate": 3.5188009893686916e-06,
"loss": 0.283,
"step": 753
},
{
"epoch": 0.013930715935334874,
"grad_norm": 2.1910760402679443,
"learning_rate": 3.492257827341492e-06,
"loss": 0.0816,
"step": 754
},
{
"epoch": 0.01394919168591224,
"grad_norm": 0.9351925849914551,
"learning_rate": 3.4657939600989453e-06,
"loss": 0.0626,
"step": 755
},
{
"epoch": 0.013967667436489608,
"grad_norm": 2.0676472187042236,
"learning_rate": 3.4394097100949286e-06,
"loss": 0.0873,
"step": 756
},
{
"epoch": 0.013986143187066974,
"grad_norm": 37.4884033203125,
"learning_rate": 3.4131053988131947e-06,
"loss": 0.1745,
"step": 757
},
{
"epoch": 0.014004618937644342,
"grad_norm": 2.072794198989868,
"learning_rate": 3.3868813467634833e-06,
"loss": 0.0835,
"step": 758
},
{
"epoch": 0.014023094688221709,
"grad_norm": 36.6066780090332,
"learning_rate": 3.360737873477584e-06,
"loss": 0.3373,
"step": 759
},
{
"epoch": 0.014041570438799077,
"grad_norm": 2.8456146717071533,
"learning_rate": 3.3346752975054763e-06,
"loss": 0.103,
"step": 760
},
{
"epoch": 0.014060046189376443,
"grad_norm": 1.588493824005127,
"learning_rate": 3.308693936411421e-06,
"loss": 0.077,
"step": 761
},
{
"epoch": 0.014078521939953811,
"grad_norm": 15.633136749267578,
"learning_rate": 3.2827941067700996e-06,
"loss": 0.1053,
"step": 762
},
{
"epoch": 0.014096997690531177,
"grad_norm": 1.303676962852478,
"learning_rate": 3.2569761241627694e-06,
"loss": 0.086,
"step": 763
},
{
"epoch": 0.014115473441108545,
"grad_norm": 2.448899269104004,
"learning_rate": 3.2312403031733943e-06,
"loss": 0.0784,
"step": 764
},
{
"epoch": 0.014133949191685912,
"grad_norm": 72.21338653564453,
"learning_rate": 3.2055869573848374e-06,
"loss": 0.3716,
"step": 765
},
{
"epoch": 0.01415242494226328,
"grad_norm": 3.7171618938446045,
"learning_rate": 3.1800163993750166e-06,
"loss": 0.0838,
"step": 766
},
{
"epoch": 0.014170900692840646,
"grad_norm": 2.813693046569824,
"learning_rate": 3.1545289407131128e-06,
"loss": 0.1063,
"step": 767
},
{
"epoch": 0.014189376443418014,
"grad_norm": 1.1639751195907593,
"learning_rate": 3.1291248919557717e-06,
"loss": 0.067,
"step": 768
},
{
"epoch": 0.01420785219399538,
"grad_norm": 1.7791749238967896,
"learning_rate": 3.103804562643302e-06,
"loss": 0.0902,
"step": 769
},
{
"epoch": 0.014226327944572749,
"grad_norm": 2.4257431030273438,
"learning_rate": 3.0785682612959334e-06,
"loss": 0.0791,
"step": 770
},
{
"epoch": 0.014244803695150115,
"grad_norm": 83.28570556640625,
"learning_rate": 3.0534162954100264e-06,
"loss": 0.331,
"step": 771
},
{
"epoch": 0.014263279445727483,
"grad_norm": 1.1570714712142944,
"learning_rate": 3.028348971454356e-06,
"loss": 0.0561,
"step": 772
},
{
"epoch": 0.01428175519630485,
"grad_norm": 2.306122303009033,
"learning_rate": 3.003366594866345e-06,
"loss": 0.0756,
"step": 773
},
{
"epoch": 0.014300230946882217,
"grad_norm": 42.3394889831543,
"learning_rate": 2.978469470048376e-06,
"loss": 0.2078,
"step": 774
},
{
"epoch": 0.014318706697459584,
"grad_norm": 3.0674984455108643,
"learning_rate": 2.953657900364053e-06,
"loss": 0.0849,
"step": 775
},
{
"epoch": 0.014337182448036952,
"grad_norm": 4.146913528442383,
"learning_rate": 2.9289321881345257e-06,
"loss": 0.0724,
"step": 776
},
{
"epoch": 0.014355658198614318,
"grad_norm": 1.3609325885772705,
"learning_rate": 2.9042926346347932e-06,
"loss": 0.0749,
"step": 777
},
{
"epoch": 0.014374133949191686,
"grad_norm": 4.192856311798096,
"learning_rate": 2.8797395400900362e-06,
"loss": 0.0983,
"step": 778
},
{
"epoch": 0.014392609699769052,
"grad_norm": 1.0222923755645752,
"learning_rate": 2.855273203671969e-06,
"loss": 0.0744,
"step": 779
},
{
"epoch": 0.01441108545034642,
"grad_norm": 1.0213944911956787,
"learning_rate": 2.830893923495173e-06,
"loss": 0.0647,
"step": 780
},
{
"epoch": 0.014429561200923787,
"grad_norm": 86.09981536865234,
"learning_rate": 2.8066019966134907e-06,
"loss": 0.4797,
"step": 781
},
{
"epoch": 0.014448036951501155,
"grad_norm": 1.6531593799591064,
"learning_rate": 2.7823977190163788e-06,
"loss": 0.067,
"step": 782
},
{
"epoch": 0.014466512702078521,
"grad_norm": 1.4805938005447388,
"learning_rate": 2.7582813856253276e-06,
"loss": 0.0629,
"step": 783
},
{
"epoch": 0.01448498845265589,
"grad_norm": 11.277420043945312,
"learning_rate": 2.7342532902902418e-06,
"loss": 0.7421,
"step": 784
},
{
"epoch": 0.014503464203233256,
"grad_norm": 0.828893780708313,
"learning_rate": 2.7103137257858867e-06,
"loss": 0.0766,
"step": 785
},
{
"epoch": 0.014521939953810624,
"grad_norm": 35.175960540771484,
"learning_rate": 2.6864629838082957e-06,
"loss": 0.6106,
"step": 786
},
{
"epoch": 0.014540415704387992,
"grad_norm": 3.607785701751709,
"learning_rate": 2.6627013549712355e-06,
"loss": 0.0852,
"step": 787
},
{
"epoch": 0.014558891454965358,
"grad_norm": 3.943739175796509,
"learning_rate": 2.639029128802657e-06,
"loss": 0.0797,
"step": 788
},
{
"epoch": 0.014577367205542726,
"grad_norm": 28.953777313232422,
"learning_rate": 2.615446593741161e-06,
"loss": 0.0955,
"step": 789
},
{
"epoch": 0.014595842956120092,
"grad_norm": 2.7790236473083496,
"learning_rate": 2.5919540371325005e-06,
"loss": 0.0871,
"step": 790
},
{
"epoch": 0.01461431870669746,
"grad_norm": 3.419975519180298,
"learning_rate": 2.5685517452260566e-06,
"loss": 0.0817,
"step": 791
},
{
"epoch": 0.014632794457274827,
"grad_norm": 1.3086283206939697,
"learning_rate": 2.5452400031713786e-06,
"loss": 0.0799,
"step": 792
},
{
"epoch": 0.014651270207852195,
"grad_norm": 56.54220962524414,
"learning_rate": 2.522019095014683e-06,
"loss": 0.1501,
"step": 793
},
{
"epoch": 0.014669745958429561,
"grad_norm": 1.2110296487808228,
"learning_rate": 2.4988893036954045e-06,
"loss": 0.0815,
"step": 794
},
{
"epoch": 0.014688221709006929,
"grad_norm": 4.748648643493652,
"learning_rate": 2.4758509110427576e-06,
"loss": 0.0909,
"step": 795
},
{
"epoch": 0.014706697459584295,
"grad_norm": 41.92718505859375,
"learning_rate": 2.45290419777228e-06,
"loss": 0.514,
"step": 796
},
{
"epoch": 0.014725173210161664,
"grad_norm": 1.4083735942840576,
"learning_rate": 2.4300494434824373e-06,
"loss": 0.0652,
"step": 797
},
{
"epoch": 0.01474364896073903,
"grad_norm": 1.2323954105377197,
"learning_rate": 2.407286926651192e-06,
"loss": 0.0675,
"step": 798
},
{
"epoch": 0.014762124711316398,
"grad_norm": 4.49576473236084,
"learning_rate": 2.3846169246326345e-06,
"loss": 0.0904,
"step": 799
},
{
"epoch": 0.014780600461893764,
"grad_norm": 2.351994037628174,
"learning_rate": 2.362039713653581e-06,
"loss": 0.1042,
"step": 800
},
{
"epoch": 0.014799076212471132,
"grad_norm": 2.2985846996307373,
"learning_rate": 2.339555568810221e-06,
"loss": 0.083,
"step": 801
},
{
"epoch": 0.014817551963048499,
"grad_norm": 1.1061251163482666,
"learning_rate": 2.317164764064769e-06,
"loss": 0.0741,
"step": 802
},
{
"epoch": 0.014836027713625867,
"grad_norm": 1.2357637882232666,
"learning_rate": 2.2948675722421086e-06,
"loss": 0.0677,
"step": 803
},
{
"epoch": 0.014854503464203233,
"grad_norm": 156.58645629882812,
"learning_rate": 2.27266426502649e-06,
"loss": 0.3675,
"step": 804
},
{
"epoch": 0.014872979214780601,
"grad_norm": 1.0823853015899658,
"learning_rate": 2.2505551129582047e-06,
"loss": 0.077,
"step": 805
},
{
"epoch": 0.014891454965357967,
"grad_norm": 2.457871437072754,
"learning_rate": 2.2285403854302912e-06,
"loss": 0.0772,
"step": 806
},
{
"epoch": 0.014909930715935335,
"grad_norm": 2.225505828857422,
"learning_rate": 2.206620350685257e-06,
"loss": 0.0836,
"step": 807
},
{
"epoch": 0.014928406466512702,
"grad_norm": 3.596954822540283,
"learning_rate": 2.1847952758118118e-06,
"loss": 0.0732,
"step": 808
},
{
"epoch": 0.01494688221709007,
"grad_norm": 70.09001922607422,
"learning_rate": 2.163065426741603e-06,
"loss": 0.2459,
"step": 809
},
{
"epoch": 0.014965357967667436,
"grad_norm": 47.55790710449219,
"learning_rate": 2.1414310682459805e-06,
"loss": 0.3881,
"step": 810
},
{
"epoch": 0.014983833718244804,
"grad_norm": 1.208727240562439,
"learning_rate": 2.119892463932781e-06,
"loss": 0.0828,
"step": 811
},
{
"epoch": 0.01500230946882217,
"grad_norm": 1.7279630899429321,
"learning_rate": 2.098449876243096e-06,
"loss": 0.0718,
"step": 812
},
{
"epoch": 0.015020785219399539,
"grad_norm": 1.1554169654846191,
"learning_rate": 2.0771035664480944e-06,
"loss": 0.082,
"step": 813
},
{
"epoch": 0.015039260969976905,
"grad_norm": 1.9765210151672363,
"learning_rate": 2.0558537946458177e-06,
"loss": 0.0771,
"step": 814
},
{
"epoch": 0.015057736720554273,
"grad_norm": 1.6989619731903076,
"learning_rate": 2.0347008197580376e-06,
"loss": 0.0858,
"step": 815
},
{
"epoch": 0.01507621247113164,
"grad_norm": 1.0549306869506836,
"learning_rate": 2.013644899527074e-06,
"loss": 0.0715,
"step": 816
},
{
"epoch": 0.015094688221709007,
"grad_norm": 1.2516361474990845,
"learning_rate": 1.9926862905126663e-06,
"loss": 0.0781,
"step": 817
},
{
"epoch": 0.015113163972286374,
"grad_norm": 1.1970781087875366,
"learning_rate": 1.9718252480888567e-06,
"loss": 0.0682,
"step": 818
},
{
"epoch": 0.015131639722863742,
"grad_norm": 10.90239143371582,
"learning_rate": 1.95106202644086e-06,
"loss": 0.1104,
"step": 819
},
{
"epoch": 0.015150115473441108,
"grad_norm": 2.855441093444824,
"learning_rate": 1.930396878561983e-06,
"loss": 0.0978,
"step": 820
},
{
"epoch": 0.015168591224018476,
"grad_norm": 1.1766327619552612,
"learning_rate": 1.9098300562505266e-06,
"loss": 0.0751,
"step": 821
},
{
"epoch": 0.015187066974595842,
"grad_norm": 1.1918188333511353,
"learning_rate": 1.8893618101067357e-06,
"loss": 0.0732,
"step": 822
},
{
"epoch": 0.01520554272517321,
"grad_norm": 19.338468551635742,
"learning_rate": 1.8689923895297247e-06,
"loss": 0.1406,
"step": 823
},
{
"epoch": 0.015224018475750577,
"grad_norm": 1.377419114112854,
"learning_rate": 1.848722042714457e-06,
"loss": 0.0707,
"step": 824
},
{
"epoch": 0.015242494226327945,
"grad_norm": 111.94596099853516,
"learning_rate": 1.8285510166487154e-06,
"loss": 0.2613,
"step": 825
},
{
"epoch": 0.015260969976905311,
"grad_norm": 1.2402987480163574,
"learning_rate": 1.808479557110081e-06,
"loss": 0.0604,
"step": 826
},
{
"epoch": 0.01527944572748268,
"grad_norm": 4.572124481201172,
"learning_rate": 1.7885079086629598e-06,
"loss": 0.1049,
"step": 827
},
{
"epoch": 0.015297921478060045,
"grad_norm": 2.0937910079956055,
"learning_rate": 1.7686363146555807e-06,
"loss": 0.0812,
"step": 828
},
{
"epoch": 0.015316397228637414,
"grad_norm": 1.99080228805542,
"learning_rate": 1.7488650172170496e-06,
"loss": 0.0838,
"step": 829
},
{
"epoch": 0.01533487297921478,
"grad_norm": 1.4868416786193848,
"learning_rate": 1.7291942572543806e-06,
"loss": 0.0739,
"step": 830
},
{
"epoch": 0.015353348729792148,
"grad_norm": 1.5787571668624878,
"learning_rate": 1.709624274449584e-06,
"loss": 0.0651,
"step": 831
},
{
"epoch": 0.015371824480369514,
"grad_norm": 1.0979748964309692,
"learning_rate": 1.6901553072567189e-06,
"loss": 0.071,
"step": 832
},
{
"epoch": 0.015390300230946882,
"grad_norm": 1.5331674814224243,
"learning_rate": 1.6707875928990059e-06,
"loss": 0.0734,
"step": 833
},
{
"epoch": 0.015408775981524249,
"grad_norm": 2.15553617477417,
"learning_rate": 1.651521367365936e-06,
"loss": 0.09,
"step": 834
},
{
"epoch": 0.015427251732101617,
"grad_norm": 1.6628950834274292,
"learning_rate": 1.6323568654103838e-06,
"loss": 0.0806,
"step": 835
},
{
"epoch": 0.015445727482678985,
"grad_norm": 3.8797972202301025,
"learning_rate": 1.6132943205457607e-06,
"loss": 0.0912,
"step": 836
},
{
"epoch": 0.015464203233256351,
"grad_norm": 0.9604938626289368,
"learning_rate": 1.5943339650431578e-06,
"loss": 0.0625,
"step": 837
},
{
"epoch": 0.015482678983833719,
"grad_norm": 0.8243102431297302,
"learning_rate": 1.5754760299285255e-06,
"loss": 0.0628,
"step": 838
},
{
"epoch": 0.015501154734411085,
"grad_norm": 1.54539954662323,
"learning_rate": 1.5567207449798517e-06,
"loss": 0.0805,
"step": 839
},
{
"epoch": 0.015519630484988453,
"grad_norm": 1.1079885959625244,
"learning_rate": 1.538068338724361e-06,
"loss": 0.0608,
"step": 840
},
{
"epoch": 0.01553810623556582,
"grad_norm": 1.6011600494384766,
"learning_rate": 1.5195190384357405e-06,
"loss": 0.0792,
"step": 841
},
{
"epoch": 0.015556581986143188,
"grad_norm": 1.851090431213379,
"learning_rate": 1.5010730701313626e-06,
"loss": 0.0893,
"step": 842
},
{
"epoch": 0.015575057736720554,
"grad_norm": 13.419096946716309,
"learning_rate": 1.4827306585695234e-06,
"loss": 0.2043,
"step": 843
},
{
"epoch": 0.015593533487297922,
"grad_norm": 1.3567969799041748,
"learning_rate": 1.4644920272467245e-06,
"loss": 0.0716,
"step": 844
},
{
"epoch": 0.015612009237875289,
"grad_norm": 1.9178869724273682,
"learning_rate": 1.446357398394934e-06,
"loss": 0.0706,
"step": 845
},
{
"epoch": 0.015630484988452657,
"grad_norm": 2.070729970932007,
"learning_rate": 1.4283269929788779e-06,
"loss": 0.072,
"step": 846
},
{
"epoch": 0.015648960739030025,
"grad_norm": 1.4816642999649048,
"learning_rate": 1.4104010306933558e-06,
"loss": 0.0643,
"step": 847
},
{
"epoch": 0.01566743648960739,
"grad_norm": 1.1445708274841309,
"learning_rate": 1.3925797299605649e-06,
"loss": 0.0647,
"step": 848
},
{
"epoch": 0.015685912240184757,
"grad_norm": 1.1432726383209229,
"learning_rate": 1.3748633079274254e-06,
"loss": 0.0592,
"step": 849
},
{
"epoch": 0.015704387990762125,
"grad_norm": 0.963321328163147,
"learning_rate": 1.3572519804629537e-06,
"loss": 0.0722,
"step": 850
},
{
"epoch": 0.015722863741339493,
"grad_norm": 32.5879020690918,
"learning_rate": 1.339745962155613e-06,
"loss": 0.3479,
"step": 851
},
{
"epoch": 0.015741339491916858,
"grad_norm": 19.58377456665039,
"learning_rate": 1.322345466310717e-06,
"loss": 0.0883,
"step": 852
},
{
"epoch": 0.015759815242494226,
"grad_norm": 1.5325170755386353,
"learning_rate": 1.30505070494781e-06,
"loss": 0.0775,
"step": 853
},
{
"epoch": 0.015778290993071594,
"grad_norm": 83.8005142211914,
"learning_rate": 1.2878618887981064e-06,
"loss": 0.1811,
"step": 854
},
{
"epoch": 0.015796766743648962,
"grad_norm": 3.3144283294677734,
"learning_rate": 1.2707792273019049e-06,
"loss": 0.0984,
"step": 855
},
{
"epoch": 0.015815242494226327,
"grad_norm": 1.47489595413208,
"learning_rate": 1.2538029286060428e-06,
"loss": 0.0772,
"step": 856
},
{
"epoch": 0.015833718244803695,
"grad_norm": 106.76734161376953,
"learning_rate": 1.2369331995613664e-06,
"loss": 0.1768,
"step": 857
},
{
"epoch": 0.015852193995381063,
"grad_norm": 1.4664264917373657,
"learning_rate": 1.2201702457201948e-06,
"loss": 0.0979,
"step": 858
},
{
"epoch": 0.01587066974595843,
"grad_norm": 1.7409895658493042,
"learning_rate": 1.2035142713338366e-06,
"loss": 0.0831,
"step": 859
},
{
"epoch": 0.015889145496535795,
"grad_norm": 1.0771201848983765,
"learning_rate": 1.1869654793500784e-06,
"loss": 0.0724,
"step": 860
},
{
"epoch": 0.015907621247113164,
"grad_norm": 108.52374267578125,
"learning_rate": 1.1705240714107301e-06,
"loss": 0.3073,
"step": 861
},
{
"epoch": 0.01592609699769053,
"grad_norm": 0.912402331829071,
"learning_rate": 1.1541902478491607e-06,
"loss": 0.0735,
"step": 862
},
{
"epoch": 0.0159445727482679,
"grad_norm": 0.8605701327323914,
"learning_rate": 1.1379642076878528e-06,
"loss": 0.0729,
"step": 863
},
{
"epoch": 0.015963048498845264,
"grad_norm": 35.42957305908203,
"learning_rate": 1.1218461486359878e-06,
"loss": 0.1683,
"step": 864
},
{
"epoch": 0.015981524249422632,
"grad_norm": 2.626542329788208,
"learning_rate": 1.1058362670870248e-06,
"loss": 0.086,
"step": 865
},
{
"epoch": 0.016,
"grad_norm": 2.617743492126465,
"learning_rate": 1.0899347581163222e-06,
"loss": 0.0741,
"step": 866
},
{
"epoch": 0.01601847575057737,
"grad_norm": 3.514965057373047,
"learning_rate": 1.0741418154787443e-06,
"loss": 0.0859,
"step": 867
},
{
"epoch": 0.016036951501154733,
"grad_norm": 11.416507720947266,
"learning_rate": 1.058457631606319e-06,
"loss": 0.6844,
"step": 868
},
{
"epoch": 0.0160554272517321,
"grad_norm": 3.9940099716186523,
"learning_rate": 1.042882397605871e-06,
"loss": 0.1,
"step": 869
},
{
"epoch": 0.01607390300230947,
"grad_norm": 1.287767767906189,
"learning_rate": 1.0274163032567165e-06,
"loss": 0.0674,
"step": 870
},
{
"epoch": 0.016092378752886837,
"grad_norm": 1.2829840183258057,
"learning_rate": 1.012059537008332e-06,
"loss": 0.0719,
"step": 871
},
{
"epoch": 0.0161108545034642,
"grad_norm": 1.3023672103881836,
"learning_rate": 9.968122859780648e-07,
"loss": 0.0788,
"step": 872
},
{
"epoch": 0.01612933025404157,
"grad_norm": 49.69974136352539,
"learning_rate": 9.816747359488632e-07,
"loss": 0.3605,
"step": 873
},
{
"epoch": 0.016147806004618938,
"grad_norm": 1.385046124458313,
"learning_rate": 9.666470713669918e-07,
"loss": 0.0885,
"step": 874
},
{
"epoch": 0.016166281755196306,
"grad_norm": 65.61194610595703,
"learning_rate": 9.517294753398066e-07,
"loss": 0.5781,
"step": 875
},
{
"epoch": 0.01618475750577367,
"grad_norm": 42.96991729736328,
"learning_rate": 9.369221296335007e-07,
"loss": 0.5182,
"step": 876
},
{
"epoch": 0.01620323325635104,
"grad_norm": 157.11624145507812,
"learning_rate": 9.222252146709143e-07,
"loss": 0.3484,
"step": 877
},
{
"epoch": 0.016221709006928407,
"grad_norm": 1.6983505487442017,
"learning_rate": 9.076389095293148e-07,
"loss": 0.0831,
"step": 878
},
{
"epoch": 0.016240184757505775,
"grad_norm": 1.4686673879623413,
"learning_rate": 8.931633919382299e-07,
"loss": 0.0794,
"step": 879
},
{
"epoch": 0.01625866050808314,
"grad_norm": 3.795792579650879,
"learning_rate": 8.787988382772705e-07,
"loss": 0.0868,
"step": 880
},
{
"epoch": 0.016277136258660507,
"grad_norm": 1.2281461954116821,
"learning_rate": 8.645454235739903e-07,
"loss": 0.0804,
"step": 881
},
{
"epoch": 0.016295612009237875,
"grad_norm": 58.373046875,
"learning_rate": 8.504033215017527e-07,
"loss": 0.1552,
"step": 882
},
{
"epoch": 0.016314087759815243,
"grad_norm": 32.65228271484375,
"learning_rate": 8.363727043776037e-07,
"loss": 0.1371,
"step": 883
},
{
"epoch": 0.016332563510392608,
"grad_norm": 1.0650213956832886,
"learning_rate": 8.224537431601886e-07,
"loss": 0.0768,
"step": 884
},
{
"epoch": 0.016351039260969976,
"grad_norm": 1.0767484903335571,
"learning_rate": 8.086466074476562e-07,
"loss": 0.0626,
"step": 885
},
{
"epoch": 0.016369515011547344,
"grad_norm": 2.329118013381958,
"learning_rate": 7.949514654755963e-07,
"loss": 0.0711,
"step": 886
},
{
"epoch": 0.016387990762124712,
"grad_norm": 39.38077163696289,
"learning_rate": 7.81368484114996e-07,
"loss": 0.494,
"step": 887
},
{
"epoch": 0.01640646651270208,
"grad_norm": 2.8096656799316406,
"learning_rate": 7.678978288701911e-07,
"loss": 0.079,
"step": 888
},
{
"epoch": 0.016424942263279445,
"grad_norm": 37.468963623046875,
"learning_rate": 7.545396638768698e-07,
"loss": 0.6064,
"step": 889
},
{
"epoch": 0.016443418013856813,
"grad_norm": 17.918235778808594,
"learning_rate": 7.412941519000527e-07,
"loss": 0.1081,
"step": 890
},
{
"epoch": 0.01646189376443418,
"grad_norm": 1.0051826238632202,
"learning_rate": 7.281614543321269e-07,
"loss": 0.0843,
"step": 891
},
{
"epoch": 0.01648036951501155,
"grad_norm": 1.2496635913848877,
"learning_rate": 7.151417311908648e-07,
"loss": 0.067,
"step": 892
},
{
"epoch": 0.016498845265588914,
"grad_norm": 7.406235218048096,
"learning_rate": 7.022351411174866e-07,
"loss": 0.0913,
"step": 893
},
{
"epoch": 0.01651732101616628,
"grad_norm": 35.35124206542969,
"learning_rate": 6.894418413747183e-07,
"loss": 0.4679,
"step": 894
},
{
"epoch": 0.01653579676674365,
"grad_norm": 1.2246192693710327,
"learning_rate": 6.767619878448783e-07,
"loss": 0.0685,
"step": 895
},
{
"epoch": 0.016554272517321018,
"grad_norm": 57.691558837890625,
"learning_rate": 6.641957350279838e-07,
"loss": 0.2462,
"step": 896
},
{
"epoch": 0.016572748267898382,
"grad_norm": 54.291725158691406,
"learning_rate": 6.517432360398556e-07,
"loss": 0.2164,
"step": 897
},
{
"epoch": 0.01659122401847575,
"grad_norm": 1.227734088897705,
"learning_rate": 6.394046426102673e-07,
"loss": 0.0748,
"step": 898
},
{
"epoch": 0.01660969976905312,
"grad_norm": 10.872947692871094,
"learning_rate": 6.271801050810856e-07,
"loss": 0.1376,
"step": 899
},
{
"epoch": 0.016628175519630486,
"grad_norm": 1.0088187456130981,
"learning_rate": 6.150697724044407e-07,
"loss": 0.0779,
"step": 900
},
{
"epoch": 0.01664665127020785,
"grad_norm": 12.033712387084961,
"learning_rate": 6.030737921409169e-07,
"loss": 0.0955,
"step": 901
},
{
"epoch": 0.01666512702078522,
"grad_norm": 0.9700973629951477,
"learning_rate": 5.911923104577455e-07,
"loss": 0.0678,
"step": 902
},
{
"epoch": 0.016683602771362587,
"grad_norm": 96.25267028808594,
"learning_rate": 5.794254721270331e-07,
"loss": 0.1404,
"step": 903
},
{
"epoch": 0.016702078521939955,
"grad_norm": 1.1411906480789185,
"learning_rate": 5.677734205239904e-07,
"loss": 0.065,
"step": 904
},
{
"epoch": 0.01672055427251732,
"grad_norm": 1.7712411880493164,
"learning_rate": 5.562362976251901e-07,
"loss": 0.0767,
"step": 905
},
{
"epoch": 0.016739030023094688,
"grad_norm": 70.70386505126953,
"learning_rate": 5.448142440068316e-07,
"loss": 0.3598,
"step": 906
},
{
"epoch": 0.016757505773672056,
"grad_norm": 0.9939737319946289,
"learning_rate": 5.335073988430373e-07,
"loss": 0.0762,
"step": 907
},
{
"epoch": 0.016775981524249424,
"grad_norm": 0.6942830681800842,
"learning_rate": 5.223158999041444e-07,
"loss": 0.066,
"step": 908
},
{
"epoch": 0.01679445727482679,
"grad_norm": 1.6551584005355835,
"learning_rate": 5.112398835550348e-07,
"loss": 0.0808,
"step": 909
},
{
"epoch": 0.016812933025404157,
"grad_norm": 1.2700129747390747,
"learning_rate": 5.002794847534765e-07,
"loss": 0.0739,
"step": 910
},
{
"epoch": 0.016831408775981525,
"grad_norm": 2.7222981452941895,
"learning_rate": 4.894348370484648e-07,
"loss": 0.0982,
"step": 911
},
{
"epoch": 0.016849884526558893,
"grad_norm": 27.60137176513672,
"learning_rate": 4.787060725786141e-07,
"loss": 0.2618,
"step": 912
},
{
"epoch": 0.016868360277136257,
"grad_norm": 3.7117230892181396,
"learning_rate": 4.6809332207053083e-07,
"loss": 0.0833,
"step": 913
},
{
"epoch": 0.016886836027713625,
"grad_norm": 0.9572230577468872,
"learning_rate": 4.575967148372318e-07,
"loss": 0.0716,
"step": 914
},
{
"epoch": 0.016905311778290993,
"grad_norm": 1.1890833377838135,
"learning_rate": 4.4721637877656377e-07,
"loss": 0.0844,
"step": 915
},
{
"epoch": 0.01692378752886836,
"grad_norm": 21.334596633911133,
"learning_rate": 4.3695244036964567e-07,
"loss": 0.6644,
"step": 916
},
{
"epoch": 0.016942263279445726,
"grad_norm": 0.6914374828338623,
"learning_rate": 4.268050246793276e-07,
"loss": 0.0625,
"step": 917
},
{
"epoch": 0.016960739030023094,
"grad_norm": 1.2535415887832642,
"learning_rate": 4.167742553486676e-07,
"loss": 0.0738,
"step": 918
},
{
"epoch": 0.016979214780600462,
"grad_norm": 1.2836554050445557,
"learning_rate": 4.068602545994249e-07,
"loss": 0.0597,
"step": 919
},
{
"epoch": 0.01699769053117783,
"grad_norm": 0.9927970767021179,
"learning_rate": 3.9706314323056936e-07,
"loss": 0.0677,
"step": 920
},
{
"epoch": 0.017016166281755195,
"grad_norm": 1.6001172065734863,
"learning_rate": 3.8738304061681107e-07,
"loss": 0.0845,
"step": 921
},
{
"epoch": 0.017034642032332563,
"grad_norm": 1.4743374586105347,
"learning_rate": 3.7782006470714614e-07,
"loss": 0.0799,
"step": 922
},
{
"epoch": 0.01705311778290993,
"grad_norm": 37.9255485534668,
"learning_rate": 3.68374332023419e-07,
"loss": 0.6246,
"step": 923
},
{
"epoch": 0.0170715935334873,
"grad_norm": 0.7737499475479126,
"learning_rate": 3.590459576589e-07,
"loss": 0.0683,
"step": 924
},
{
"epoch": 0.017090069284064664,
"grad_norm": 0.6625645160675049,
"learning_rate": 3.498350552768859e-07,
"loss": 0.0672,
"step": 925
},
{
"epoch": 0.01710854503464203,
"grad_norm": 4.719531059265137,
"learning_rate": 3.4074173710931804e-07,
"loss": 0.0885,
"step": 926
},
{
"epoch": 0.0171270207852194,
"grad_norm": 151.60169982910156,
"learning_rate": 3.3176611395540625e-07,
"loss": 0.3384,
"step": 927
},
{
"epoch": 0.017145496535796768,
"grad_norm": 1.798681616783142,
"learning_rate": 3.2290829518028867e-07,
"loss": 0.0841,
"step": 928
},
{
"epoch": 0.017163972286374132,
"grad_norm": 2.17362904548645,
"learning_rate": 3.1416838871368925e-07,
"loss": 0.0753,
"step": 929
},
{
"epoch": 0.0171824480369515,
"grad_norm": 42.84740447998047,
"learning_rate": 3.0554650104861137e-07,
"loss": 0.4887,
"step": 930
},
{
"epoch": 0.01720092378752887,
"grad_norm": 161.60919189453125,
"learning_rate": 2.970427372400353e-07,
"loss": 0.2951,
"step": 931
},
{
"epoch": 0.017219399538106236,
"grad_norm": 1.3821405172348022,
"learning_rate": 2.8865720090364037e-07,
"loss": 0.0742,
"step": 932
},
{
"epoch": 0.0172378752886836,
"grad_norm": 70.28323364257812,
"learning_rate": 2.8038999421453827e-07,
"loss": 0.2936,
"step": 933
},
{
"epoch": 0.01725635103926097,
"grad_norm": 7.153317451477051,
"learning_rate": 2.7224121790603517e-07,
"loss": 0.0884,
"step": 934
},
{
"epoch": 0.017274826789838337,
"grad_norm": 1.5210833549499512,
"learning_rate": 2.6421097126839714e-07,
"loss": 0.0809,
"step": 935
},
{
"epoch": 0.017293302540415705,
"grad_norm": 1.937935471534729,
"learning_rate": 2.5629935214764866e-07,
"loss": 0.0887,
"step": 936
},
{
"epoch": 0.017311778290993073,
"grad_norm": 76.13717651367188,
"learning_rate": 2.4850645694436736e-07,
"loss": 0.3424,
"step": 937
},
{
"epoch": 0.017330254041570438,
"grad_norm": 14.685988426208496,
"learning_rate": 2.4083238061252565e-07,
"loss": 0.0868,
"step": 938
},
{
"epoch": 0.017348729792147806,
"grad_norm": 1.4477519989013672,
"learning_rate": 2.332772166583208e-07,
"loss": 0.0734,
"step": 939
},
{
"epoch": 0.017367205542725174,
"grad_norm": 2.444828748703003,
"learning_rate": 2.2584105713904126e-07,
"loss": 0.0711,
"step": 940
},
{
"epoch": 0.017385681293302542,
"grad_norm": 0.9954750537872314,
"learning_rate": 2.1852399266194312e-07,
"loss": 0.0724,
"step": 941
},
{
"epoch": 0.017404157043879907,
"grad_norm": 3.597987413406372,
"learning_rate": 2.1132611238315004e-07,
"loss": 0.1066,
"step": 942
},
{
"epoch": 0.017422632794457275,
"grad_norm": 108.85304260253906,
"learning_rate": 2.0424750400655947e-07,
"loss": 0.2482,
"step": 943
},
{
"epoch": 0.017441108545034643,
"grad_norm": 1.5015740394592285,
"learning_rate": 1.9728825378278248e-07,
"loss": 0.0723,
"step": 944
},
{
"epoch": 0.01745958429561201,
"grad_norm": 1.47672438621521,
"learning_rate": 1.9044844650808468e-07,
"loss": 0.0833,
"step": 945
},
{
"epoch": 0.017478060046189375,
"grad_norm": 1.08296799659729,
"learning_rate": 1.8372816552336025e-07,
"loss": 0.0842,
"step": 946
},
{
"epoch": 0.017496535796766743,
"grad_norm": 2.175466537475586,
"learning_rate": 1.7712749271311392e-07,
"loss": 0.0706,
"step": 947
},
{
"epoch": 0.01751501154734411,
"grad_norm": 1.797810673713684,
"learning_rate": 1.706465085044584e-07,
"loss": 0.0943,
"step": 948
},
{
"epoch": 0.01753348729792148,
"grad_norm": 0.898100733757019,
"learning_rate": 1.6428529186614195e-07,
"loss": 0.071,
"step": 949
},
{
"epoch": 0.017551963048498844,
"grad_norm": 2.9947056770324707,
"learning_rate": 1.580439203075812e-07,
"loss": 0.1002,
"step": 950
},
{
"epoch": 0.017570438799076212,
"grad_norm": 2.4971437454223633,
"learning_rate": 1.519224698779198e-07,
"loss": 0.0852,
"step": 951
},
{
"epoch": 0.01758891454965358,
"grad_norm": 1.6412723064422607,
"learning_rate": 1.4592101516509916e-07,
"loss": 0.0885,
"step": 952
},
{
"epoch": 0.01760739030023095,
"grad_norm": 1.4672108888626099,
"learning_rate": 1.400396292949513e-07,
"loss": 0.0771,
"step": 953
},
{
"epoch": 0.017625866050808313,
"grad_norm": 3.760052442550659,
"learning_rate": 1.3427838393030634e-07,
"loss": 0.0882,
"step": 954
},
{
"epoch": 0.01764434180138568,
"grad_norm": 0.9100709557533264,
"learning_rate": 1.2863734927012094e-07,
"loss": 0.0727,
"step": 955
},
{
"epoch": 0.01766281755196305,
"grad_norm": 7.4524078369140625,
"learning_rate": 1.231165940486234e-07,
"loss": 0.095,
"step": 956
},
{
"epoch": 0.017681293302540417,
"grad_norm": 126.81805419921875,
"learning_rate": 1.1771618553447217e-07,
"loss": 0.2877,
"step": 957
},
{
"epoch": 0.01769976905311778,
"grad_norm": 1.409822702407837,
"learning_rate": 1.1243618952994195e-07,
"loss": 0.0799,
"step": 958
},
{
"epoch": 0.01771824480369515,
"grad_norm": 2.6195473670959473,
"learning_rate": 1.0727667037011668e-07,
"loss": 0.0853,
"step": 959
},
{
"epoch": 0.017736720554272518,
"grad_norm": 17.708559036254883,
"learning_rate": 1.0223769092211012e-07,
"loss": 0.5794,
"step": 960
},
{
"epoch": 0.017755196304849886,
"grad_norm": 3.743419885635376,
"learning_rate": 9.731931258429638e-08,
"loss": 0.082,
"step": 961
},
{
"epoch": 0.01777367205542725,
"grad_norm": 1.3318521976470947,
"learning_rate": 9.252159528556404e-08,
"loss": 0.0728,
"step": 962
},
{
"epoch": 0.01779214780600462,
"grad_norm": 2.438230037689209,
"learning_rate": 8.784459748458318e-08,
"loss": 0.0861,
"step": 963
},
{
"epoch": 0.017810623556581986,
"grad_norm": 2.421828031539917,
"learning_rate": 8.328837616909612e-08,
"loss": 0.0806,
"step": 964
},
{
"epoch": 0.017829099307159355,
"grad_norm": 2.563591480255127,
"learning_rate": 7.885298685522235e-08,
"loss": 0.1074,
"step": 965
},
{
"epoch": 0.01784757505773672,
"grad_norm": 39.273399353027344,
"learning_rate": 7.453848358678018e-08,
"loss": 0.1141,
"step": 966
},
{
"epoch": 0.017866050808314087,
"grad_norm": 0.981799304485321,
"learning_rate": 7.034491893463059e-08,
"loss": 0.0676,
"step": 967
},
{
"epoch": 0.017884526558891455,
"grad_norm": 5.325465679168701,
"learning_rate": 6.627234399603554e-08,
"loss": 0.0928,
"step": 968
},
{
"epoch": 0.017903002309468823,
"grad_norm": 4.879908561706543,
"learning_rate": 6.232080839403631e-08,
"loss": 0.1331,
"step": 969
},
{
"epoch": 0.017921478060046188,
"grad_norm": 1.3284275531768799,
"learning_rate": 5.849036027684607e-08,
"loss": 0.0739,
"step": 970
},
{
"epoch": 0.017939953810623556,
"grad_norm": 1.0023523569107056,
"learning_rate": 5.4781046317267103e-08,
"loss": 0.0661,
"step": 971
},
{
"epoch": 0.017958429561200924,
"grad_norm": 3.2550771236419678,
"learning_rate": 5.119291171211793e-08,
"loss": 0.0764,
"step": 972
},
{
"epoch": 0.017976905311778292,
"grad_norm": 5.131398677825928,
"learning_rate": 4.772600018168816e-08,
"loss": 0.0837,
"step": 973
},
{
"epoch": 0.017995381062355657,
"grad_norm": 5.255235195159912,
"learning_rate": 4.438035396920004e-08,
"loss": 0.1115,
"step": 974
},
{
"epoch": 0.018013856812933025,
"grad_norm": 1.48928964138031,
"learning_rate": 4.115601384029666e-08,
"loss": 0.072,
"step": 975
},
{
"epoch": 0.018032332563510393,
"grad_norm": 3.3543782234191895,
"learning_rate": 3.805301908254455e-08,
"loss": 0.0921,
"step": 976
},
{
"epoch": 0.01805080831408776,
"grad_norm": 1.5796211957931519,
"learning_rate": 3.50714075049563e-08,
"loss": 0.075,
"step": 977
},
{
"epoch": 0.018069284064665125,
"grad_norm": 1.979763388633728,
"learning_rate": 3.22112154375287e-08,
"loss": 0.0837,
"step": 978
},
{
"epoch": 0.018087759815242493,
"grad_norm": 1.811806559562683,
"learning_rate": 2.947247773079753e-08,
"loss": 0.0829,
"step": 979
},
{
"epoch": 0.01810623556581986,
"grad_norm": 2.1670680046081543,
"learning_rate": 2.6855227755419046e-08,
"loss": 0.0817,
"step": 980
},
{
"epoch": 0.01812471131639723,
"grad_norm": 10.522448539733887,
"learning_rate": 2.4359497401758026e-08,
"loss": 0.2201,
"step": 981
},
{
"epoch": 0.018143187066974594,
"grad_norm": 33.86839294433594,
"learning_rate": 2.1985317079500358e-08,
"loss": 0.1112,
"step": 982
},
{
"epoch": 0.018161662817551962,
"grad_norm": 3.0782365798950195,
"learning_rate": 1.973271571728441e-08,
"loss": 0.1167,
"step": 983
},
{
"epoch": 0.01818013856812933,
"grad_norm": 4.700839042663574,
"learning_rate": 1.7601720762346895e-08,
"loss": 0.0801,
"step": 984
},
{
"epoch": 0.0181986143187067,
"grad_norm": 49.40107727050781,
"learning_rate": 1.5592358180189782e-08,
"loss": 0.2201,
"step": 985
},
{
"epoch": 0.018217090069284066,
"grad_norm": 2.221862554550171,
"learning_rate": 1.370465245426167e-08,
"loss": 0.0865,
"step": 986
},
{
"epoch": 0.01823556581986143,
"grad_norm": 1.136305332183838,
"learning_rate": 1.1938626585660252e-08,
"loss": 0.0649,
"step": 987
},
{
"epoch": 0.0182540415704388,
"grad_norm": 2.456942558288574,
"learning_rate": 1.0294302092853647e-08,
"loss": 0.0787,
"step": 988
},
{
"epoch": 0.018272517321016167,
"grad_norm": 1.2830783128738403,
"learning_rate": 8.771699011416169e-09,
"loss": 0.0651,
"step": 989
},
{
"epoch": 0.018290993071593535,
"grad_norm": 8.097618103027344,
"learning_rate": 7.370835893788508e-09,
"loss": 0.1273,
"step": 990
},
{
"epoch": 0.0183094688221709,
"grad_norm": 46.91101837158203,
"learning_rate": 6.091729809042379e-09,
"loss": 0.1365,
"step": 991
},
{
"epoch": 0.018327944572748268,
"grad_norm": 31.30545425415039,
"learning_rate": 4.9343963426840006e-09,
"loss": 0.1328,
"step": 992
},
{
"epoch": 0.018346420323325636,
"grad_norm": 11.6485013961792,
"learning_rate": 3.898849596456477e-09,
"loss": 0.0792,
"step": 993
},
{
"epoch": 0.018364896073903004,
"grad_norm": 4.699264049530029,
"learning_rate": 2.9851021881688314e-09,
"loss": 0.0909,
"step": 994
},
{
"epoch": 0.01838337182448037,
"grad_norm": 1.769290566444397,
"learning_rate": 2.193165251545004e-09,
"loss": 0.0768,
"step": 995
},
{
"epoch": 0.018401847575057736,
"grad_norm": 6.556053638458252,
"learning_rate": 1.5230484360873043e-09,
"loss": 0.0771,
"step": 996
},
{
"epoch": 0.018420323325635105,
"grad_norm": 1.5311815738677979,
"learning_rate": 9.74759906957612e-10,
"loss": 0.0844,
"step": 997
},
{
"epoch": 0.018438799076212473,
"grad_norm": 1.5484660863876343,
"learning_rate": 5.483063448785686e-10,
"loss": 0.0661,
"step": 998
},
{
"epoch": 0.018457274826789837,
"grad_norm": 0.8800917863845825,
"learning_rate": 2.436929460525317e-10,
"loss": 0.0718,
"step": 999
},
{
"epoch": 0.018475750577367205,
"grad_norm": 10.245546340942383,
"learning_rate": 6.092342209607083e-11,
"loss": 0.1362,
"step": 1000
}
],
"logging_steps": 1,
"max_steps": 1000,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 200,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}