prxy5605's picture
Training in progress, step 200, checkpoint
4ad78ec verified
{
"best_metric": 5.203610897064209,
"best_model_checkpoint": "miner_id_24/checkpoint-200",
"epoch": 2.030456852791878,
"eval_steps": 50,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01015228426395939,
"grad_norm": 0.46062880754470825,
"learning_rate": 5e-06,
"loss": 5.3841,
"step": 1
},
{
"epoch": 0.01015228426395939,
"eval_loss": 5.5305962562561035,
"eval_runtime": 8.7087,
"eval_samples_per_second": 19.061,
"eval_steps_per_second": 9.531,
"step": 1
},
{
"epoch": 0.02030456852791878,
"grad_norm": 0.5978866815567017,
"learning_rate": 1e-05,
"loss": 5.4053,
"step": 2
},
{
"epoch": 0.030456852791878174,
"grad_norm": 0.4448310434818268,
"learning_rate": 1.5e-05,
"loss": 5.3689,
"step": 3
},
{
"epoch": 0.04060913705583756,
"grad_norm": 0.507191002368927,
"learning_rate": 2e-05,
"loss": 5.4424,
"step": 4
},
{
"epoch": 0.050761421319796954,
"grad_norm": 0.4688023626804352,
"learning_rate": 2.5e-05,
"loss": 5.4252,
"step": 5
},
{
"epoch": 0.06091370558375635,
"grad_norm": 0.4471573531627655,
"learning_rate": 3e-05,
"loss": 5.3927,
"step": 6
},
{
"epoch": 0.07106598984771574,
"grad_norm": 0.45171481370925903,
"learning_rate": 3.5e-05,
"loss": 5.3751,
"step": 7
},
{
"epoch": 0.08121827411167512,
"grad_norm": 0.6176170706748962,
"learning_rate": 4e-05,
"loss": 5.4215,
"step": 8
},
{
"epoch": 0.09137055837563451,
"grad_norm": 0.5099933743476868,
"learning_rate": 4.5e-05,
"loss": 5.3422,
"step": 9
},
{
"epoch": 0.10152284263959391,
"grad_norm": 0.48276761174201965,
"learning_rate": 5e-05,
"loss": 5.4263,
"step": 10
},
{
"epoch": 0.1116751269035533,
"grad_norm": 0.5048362612724304,
"learning_rate": 5.500000000000001e-05,
"loss": 5.3427,
"step": 11
},
{
"epoch": 0.1218274111675127,
"grad_norm": 0.5130631923675537,
"learning_rate": 6e-05,
"loss": 5.3817,
"step": 12
},
{
"epoch": 0.1319796954314721,
"grad_norm": 0.5273154377937317,
"learning_rate": 6.500000000000001e-05,
"loss": 5.3019,
"step": 13
},
{
"epoch": 0.14213197969543148,
"grad_norm": 0.523825466632843,
"learning_rate": 7e-05,
"loss": 5.2535,
"step": 14
},
{
"epoch": 0.15228426395939088,
"grad_norm": 0.6113125681877136,
"learning_rate": 7.500000000000001e-05,
"loss": 5.3172,
"step": 15
},
{
"epoch": 0.16243654822335024,
"grad_norm": 0.5640417337417603,
"learning_rate": 8e-05,
"loss": 5.2758,
"step": 16
},
{
"epoch": 0.17258883248730963,
"grad_norm": 0.4186924397945404,
"learning_rate": 8.5e-05,
"loss": 5.2586,
"step": 17
},
{
"epoch": 0.18274111675126903,
"grad_norm": 0.5193726420402527,
"learning_rate": 9e-05,
"loss": 5.2615,
"step": 18
},
{
"epoch": 0.19289340101522842,
"grad_norm": 0.5647014379501343,
"learning_rate": 9.5e-05,
"loss": 5.2652,
"step": 19
},
{
"epoch": 0.20304568527918782,
"grad_norm": 0.5719868540763855,
"learning_rate": 0.0001,
"loss": 5.1955,
"step": 20
},
{
"epoch": 0.2131979695431472,
"grad_norm": 0.6991127133369446,
"learning_rate": 9.999238475781957e-05,
"loss": 5.1646,
"step": 21
},
{
"epoch": 0.2233502538071066,
"grad_norm": 0.7082369327545166,
"learning_rate": 9.99695413509548e-05,
"loss": 5.1052,
"step": 22
},
{
"epoch": 0.233502538071066,
"grad_norm": 0.7457136511802673,
"learning_rate": 9.99314767377287e-05,
"loss": 5.1863,
"step": 23
},
{
"epoch": 0.2436548223350254,
"grad_norm": 1.6564850807189941,
"learning_rate": 9.987820251299122e-05,
"loss": 5.0987,
"step": 24
},
{
"epoch": 0.25380710659898476,
"grad_norm": 5.510101318359375,
"learning_rate": 9.980973490458728e-05,
"loss": 5.5984,
"step": 25
},
{
"epoch": 0.2639593908629442,
"grad_norm": 4.215838432312012,
"learning_rate": 9.972609476841367e-05,
"loss": 5.5124,
"step": 26
},
{
"epoch": 0.27411167512690354,
"grad_norm": 1.8195075988769531,
"learning_rate": 9.962730758206611e-05,
"loss": 5.4004,
"step": 27
},
{
"epoch": 0.28426395939086296,
"grad_norm": 1.5931973457336426,
"learning_rate": 9.951340343707852e-05,
"loss": 5.3794,
"step": 28
},
{
"epoch": 0.29441624365482233,
"grad_norm": 0.6483675241470337,
"learning_rate": 9.938441702975689e-05,
"loss": 5.3566,
"step": 29
},
{
"epoch": 0.30456852791878175,
"grad_norm": 0.3973766267299652,
"learning_rate": 9.924038765061042e-05,
"loss": 5.2972,
"step": 30
},
{
"epoch": 0.3147208121827411,
"grad_norm": 0.3534790277481079,
"learning_rate": 9.908135917238321e-05,
"loss": 5.2902,
"step": 31
},
{
"epoch": 0.3248730964467005,
"grad_norm": 0.3970736861228943,
"learning_rate": 9.890738003669029e-05,
"loss": 5.2949,
"step": 32
},
{
"epoch": 0.3350253807106599,
"grad_norm": 0.37264251708984375,
"learning_rate": 9.871850323926177e-05,
"loss": 5.2944,
"step": 33
},
{
"epoch": 0.34517766497461927,
"grad_norm": 0.3872509002685547,
"learning_rate": 9.851478631379982e-05,
"loss": 5.326,
"step": 34
},
{
"epoch": 0.3553299492385787,
"grad_norm": 0.3853321969509125,
"learning_rate": 9.829629131445342e-05,
"loss": 5.2708,
"step": 35
},
{
"epoch": 0.36548223350253806,
"grad_norm": 0.39588385820388794,
"learning_rate": 9.806308479691595e-05,
"loss": 5.3228,
"step": 36
},
{
"epoch": 0.3756345177664975,
"grad_norm": 0.4093726575374603,
"learning_rate": 9.781523779815179e-05,
"loss": 5.2831,
"step": 37
},
{
"epoch": 0.38578680203045684,
"grad_norm": 0.4420938491821289,
"learning_rate": 9.755282581475769e-05,
"loss": 5.2103,
"step": 38
},
{
"epoch": 0.39593908629441626,
"grad_norm": 0.5512685775756836,
"learning_rate": 9.727592877996585e-05,
"loss": 5.1775,
"step": 39
},
{
"epoch": 0.40609137055837563,
"grad_norm": 0.3994300067424774,
"learning_rate": 9.698463103929542e-05,
"loss": 5.2478,
"step": 40
},
{
"epoch": 0.41624365482233505,
"grad_norm": 0.5147640705108643,
"learning_rate": 9.667902132486009e-05,
"loss": 5.1743,
"step": 41
},
{
"epoch": 0.4263959390862944,
"grad_norm": 0.6237267851829529,
"learning_rate": 9.635919272833938e-05,
"loss": 5.2249,
"step": 42
},
{
"epoch": 0.4365482233502538,
"grad_norm": 0.5305359363555908,
"learning_rate": 9.602524267262203e-05,
"loss": 5.2915,
"step": 43
},
{
"epoch": 0.4467005076142132,
"grad_norm": 0.4987878203392029,
"learning_rate": 9.567727288213005e-05,
"loss": 5.253,
"step": 44
},
{
"epoch": 0.45685279187817257,
"grad_norm": 0.6333546042442322,
"learning_rate": 9.53153893518325e-05,
"loss": 5.1612,
"step": 45
},
{
"epoch": 0.467005076142132,
"grad_norm": 0.7043673396110535,
"learning_rate": 9.493970231495835e-05,
"loss": 5.1424,
"step": 46
},
{
"epoch": 0.47715736040609136,
"grad_norm": 0.6457162499427795,
"learning_rate": 9.45503262094184e-05,
"loss": 5.1876,
"step": 47
},
{
"epoch": 0.4873096446700508,
"grad_norm": 0.9820389151573181,
"learning_rate": 9.414737964294636e-05,
"loss": 5.0792,
"step": 48
},
{
"epoch": 0.49746192893401014,
"grad_norm": 2.78731369972229,
"learning_rate": 9.373098535696979e-05,
"loss": 5.5062,
"step": 49
},
{
"epoch": 0.5076142131979695,
"grad_norm": 2.0711777210235596,
"learning_rate": 9.330127018922194e-05,
"loss": 5.4469,
"step": 50
},
{
"epoch": 0.5076142131979695,
"eval_loss": 5.311061859130859,
"eval_runtime": 8.7447,
"eval_samples_per_second": 18.983,
"eval_steps_per_second": 9.491,
"step": 50
},
{
"epoch": 0.5177664974619289,
"grad_norm": 0.99236661195755,
"learning_rate": 9.285836503510562e-05,
"loss": 5.3363,
"step": 51
},
{
"epoch": 0.5279187817258884,
"grad_norm": 0.48146185278892517,
"learning_rate": 9.24024048078213e-05,
"loss": 5.3438,
"step": 52
},
{
"epoch": 0.5380710659898477,
"grad_norm": 0.33013349771499634,
"learning_rate": 9.193352839727121e-05,
"loss": 5.3068,
"step": 53
},
{
"epoch": 0.5482233502538071,
"grad_norm": 0.37118765711784363,
"learning_rate": 9.145187862775209e-05,
"loss": 5.306,
"step": 54
},
{
"epoch": 0.5583756345177665,
"grad_norm": 0.4362952411174774,
"learning_rate": 9.09576022144496e-05,
"loss": 5.2849,
"step": 55
},
{
"epoch": 0.5685279187817259,
"grad_norm": 0.3819132149219513,
"learning_rate": 9.045084971874738e-05,
"loss": 5.3258,
"step": 56
},
{
"epoch": 0.5786802030456852,
"grad_norm": 0.3916621804237366,
"learning_rate": 8.993177550236464e-05,
"loss": 5.3383,
"step": 57
},
{
"epoch": 0.5888324873096447,
"grad_norm": 0.36510735750198364,
"learning_rate": 8.940053768033609e-05,
"loss": 5.2607,
"step": 58
},
{
"epoch": 0.5989847715736041,
"grad_norm": 0.40830719470977783,
"learning_rate": 8.885729807284856e-05,
"loss": 5.3212,
"step": 59
},
{
"epoch": 0.6091370558375635,
"grad_norm": 0.4307074248790741,
"learning_rate": 8.83022221559489e-05,
"loss": 5.2596,
"step": 60
},
{
"epoch": 0.6192893401015228,
"grad_norm": 0.3730439245700836,
"learning_rate": 8.773547901113862e-05,
"loss": 5.2363,
"step": 61
},
{
"epoch": 0.6294416243654822,
"grad_norm": 0.4428587555885315,
"learning_rate": 8.715724127386972e-05,
"loss": 5.267,
"step": 62
},
{
"epoch": 0.6395939086294417,
"grad_norm": 0.41558074951171875,
"learning_rate": 8.656768508095853e-05,
"loss": 5.1866,
"step": 63
},
{
"epoch": 0.649746192893401,
"grad_norm": 0.4241006672382355,
"learning_rate": 8.596699001693255e-05,
"loss": 5.1834,
"step": 64
},
{
"epoch": 0.6598984771573604,
"grad_norm": 0.3360666036605835,
"learning_rate": 8.535533905932738e-05,
"loss": 5.2194,
"step": 65
},
{
"epoch": 0.6700507614213198,
"grad_norm": 0.38096800446510315,
"learning_rate": 8.473291852294987e-05,
"loss": 5.2324,
"step": 66
},
{
"epoch": 0.6802030456852792,
"grad_norm": 0.437884122133255,
"learning_rate": 8.409991800312493e-05,
"loss": 5.2587,
"step": 67
},
{
"epoch": 0.6903553299492385,
"grad_norm": 0.48057064414024353,
"learning_rate": 8.345653031794292e-05,
"loss": 5.2374,
"step": 68
},
{
"epoch": 0.700507614213198,
"grad_norm": 0.5798227787017822,
"learning_rate": 8.280295144952536e-05,
"loss": 5.1645,
"step": 69
},
{
"epoch": 0.7106598984771574,
"grad_norm": 0.5949220061302185,
"learning_rate": 8.213938048432697e-05,
"loss": 5.1196,
"step": 70
},
{
"epoch": 0.7208121827411168,
"grad_norm": 0.6557019948959351,
"learning_rate": 8.146601955249188e-05,
"loss": 5.1316,
"step": 71
},
{
"epoch": 0.7309644670050761,
"grad_norm": 0.7752095460891724,
"learning_rate": 8.07830737662829e-05,
"loss": 5.0944,
"step": 72
},
{
"epoch": 0.7411167512690355,
"grad_norm": 0.8926941156387329,
"learning_rate": 8.009075115760243e-05,
"loss": 5.3464,
"step": 73
},
{
"epoch": 0.751269035532995,
"grad_norm": 0.7516569495201111,
"learning_rate": 7.938926261462366e-05,
"loss": 5.3015,
"step": 74
},
{
"epoch": 0.7614213197969543,
"grad_norm": 0.6225847005844116,
"learning_rate": 7.86788218175523e-05,
"loss": 5.3173,
"step": 75
},
{
"epoch": 0.7715736040609137,
"grad_norm": 0.47938278317451477,
"learning_rate": 7.795964517353735e-05,
"loss": 5.3143,
"step": 76
},
{
"epoch": 0.7817258883248731,
"grad_norm": 0.3764425814151764,
"learning_rate": 7.723195175075136e-05,
"loss": 5.3017,
"step": 77
},
{
"epoch": 0.7918781725888325,
"grad_norm": 0.3587036430835724,
"learning_rate": 7.649596321166024e-05,
"loss": 5.2366,
"step": 78
},
{
"epoch": 0.8020304568527918,
"grad_norm": 0.3250347673892975,
"learning_rate": 7.575190374550272e-05,
"loss": 5.2564,
"step": 79
},
{
"epoch": 0.8121827411167513,
"grad_norm": 0.345753937959671,
"learning_rate": 7.500000000000001e-05,
"loss": 5.2704,
"step": 80
},
{
"epoch": 0.8223350253807107,
"grad_norm": 0.3820163905620575,
"learning_rate": 7.424048101231686e-05,
"loss": 5.3212,
"step": 81
},
{
"epoch": 0.8324873096446701,
"grad_norm": 0.40871894359588623,
"learning_rate": 7.347357813929454e-05,
"loss": 5.2279,
"step": 82
},
{
"epoch": 0.8426395939086294,
"grad_norm": 0.36811622977256775,
"learning_rate": 7.269952498697734e-05,
"loss": 5.263,
"step": 83
},
{
"epoch": 0.8527918781725888,
"grad_norm": 0.3945890963077545,
"learning_rate": 7.191855733945387e-05,
"loss": 5.309,
"step": 84
},
{
"epoch": 0.8629441624365483,
"grad_norm": 0.37406954169273376,
"learning_rate": 7.113091308703498e-05,
"loss": 5.2704,
"step": 85
},
{
"epoch": 0.8730964467005076,
"grad_norm": 0.39915502071380615,
"learning_rate": 7.033683215379002e-05,
"loss": 5.2693,
"step": 86
},
{
"epoch": 0.883248730964467,
"grad_norm": 0.4481967091560364,
"learning_rate": 6.953655642446368e-05,
"loss": 5.2332,
"step": 87
},
{
"epoch": 0.8934010152284264,
"grad_norm": 0.4243590235710144,
"learning_rate": 6.873032967079561e-05,
"loss": 5.2426,
"step": 88
},
{
"epoch": 0.9035532994923858,
"grad_norm": 0.46772047877311707,
"learning_rate": 6.7918397477265e-05,
"loss": 5.1668,
"step": 89
},
{
"epoch": 0.9137055837563451,
"grad_norm": 0.4974723756313324,
"learning_rate": 6.710100716628344e-05,
"loss": 5.17,
"step": 90
},
{
"epoch": 0.9238578680203046,
"grad_norm": 0.4955981373786926,
"learning_rate": 6.627840772285784e-05,
"loss": 5.2168,
"step": 91
},
{
"epoch": 0.934010152284264,
"grad_norm": 0.4544804096221924,
"learning_rate": 6.545084971874738e-05,
"loss": 5.218,
"step": 92
},
{
"epoch": 0.9441624365482234,
"grad_norm": 0.5566396117210388,
"learning_rate": 6.461858523613684e-05,
"loss": 5.1191,
"step": 93
},
{
"epoch": 0.9543147208121827,
"grad_norm": 0.5591332316398621,
"learning_rate": 6.378186779084995e-05,
"loss": 5.0513,
"step": 94
},
{
"epoch": 0.9644670050761421,
"grad_norm": 0.6510249972343445,
"learning_rate": 6.294095225512603e-05,
"loss": 5.0533,
"step": 95
},
{
"epoch": 0.9746192893401016,
"grad_norm": 0.8038920760154724,
"learning_rate": 6.209609477998338e-05,
"loss": 5.0682,
"step": 96
},
{
"epoch": 0.9847715736040609,
"grad_norm": 0.5465517640113831,
"learning_rate": 6.124755271719325e-05,
"loss": 5.3038,
"step": 97
},
{
"epoch": 0.9949238578680203,
"grad_norm": 0.4467344582080841,
"learning_rate": 6.0395584540887963e-05,
"loss": 5.2282,
"step": 98
},
{
"epoch": 1.0050761421319796,
"grad_norm": 0.6803540587425232,
"learning_rate": 5.9540449768827246e-05,
"loss": 7.978,
"step": 99
},
{
"epoch": 1.015228426395939,
"grad_norm": 0.3697003424167633,
"learning_rate": 5.868240888334653e-05,
"loss": 5.1768,
"step": 100
},
{
"epoch": 1.015228426395939,
"eval_loss": 5.2339911460876465,
"eval_runtime": 8.769,
"eval_samples_per_second": 18.93,
"eval_steps_per_second": 9.465,
"step": 100
},
{
"epoch": 1.0253807106598984,
"grad_norm": 0.36876264214515686,
"learning_rate": 5.782172325201155e-05,
"loss": 5.3701,
"step": 101
},
{
"epoch": 1.0355329949238579,
"grad_norm": 0.37881383299827576,
"learning_rate": 5.695865504800327e-05,
"loss": 5.1475,
"step": 102
},
{
"epoch": 1.0456852791878173,
"grad_norm": 0.4124298691749573,
"learning_rate": 5.6093467170257374e-05,
"loss": 5.2639,
"step": 103
},
{
"epoch": 1.0558375634517767,
"grad_norm": 0.4313642680644989,
"learning_rate": 5.522642316338268e-05,
"loss": 5.2658,
"step": 104
},
{
"epoch": 1.0659898477157361,
"grad_norm": 0.3469761312007904,
"learning_rate": 5.435778713738292e-05,
"loss": 5.1731,
"step": 105
},
{
"epoch": 1.0761421319796955,
"grad_norm": 0.34174656867980957,
"learning_rate": 5.348782368720626e-05,
"loss": 5.2532,
"step": 106
},
{
"epoch": 1.0862944162436547,
"grad_norm": 0.37279707193374634,
"learning_rate": 5.26167978121472e-05,
"loss": 5.3232,
"step": 107
},
{
"epoch": 1.0964467005076142,
"grad_norm": 0.34107351303100586,
"learning_rate": 5.174497483512506e-05,
"loss": 5.2014,
"step": 108
},
{
"epoch": 1.1065989847715736,
"grad_norm": 0.44603461027145386,
"learning_rate": 5.0872620321864185e-05,
"loss": 5.274,
"step": 109
},
{
"epoch": 1.116751269035533,
"grad_norm": 0.43816447257995605,
"learning_rate": 5e-05,
"loss": 5.2644,
"step": 110
},
{
"epoch": 1.1269035532994924,
"grad_norm": 0.42787858843803406,
"learning_rate": 4.912737967813583e-05,
"loss": 5.182,
"step": 111
},
{
"epoch": 1.1370558375634519,
"grad_norm": 0.4469396471977234,
"learning_rate": 4.825502516487497e-05,
"loss": 5.1906,
"step": 112
},
{
"epoch": 1.1472081218274113,
"grad_norm": 0.47741401195526123,
"learning_rate": 4.738320218785281e-05,
"loss": 5.208,
"step": 113
},
{
"epoch": 1.1573604060913705,
"grad_norm": 0.5933973789215088,
"learning_rate": 4.6512176312793736e-05,
"loss": 5.056,
"step": 114
},
{
"epoch": 1.16751269035533,
"grad_norm": 0.5861052870750427,
"learning_rate": 4.564221286261709e-05,
"loss": 5.1204,
"step": 115
},
{
"epoch": 1.1776649746192893,
"grad_norm": 0.6024320125579834,
"learning_rate": 4.477357683661734e-05,
"loss": 5.1356,
"step": 116
},
{
"epoch": 1.1878172588832487,
"grad_norm": 0.5369526147842407,
"learning_rate": 4.390653282974264e-05,
"loss": 5.2165,
"step": 117
},
{
"epoch": 1.1979695431472082,
"grad_norm": 0.4957830309867859,
"learning_rate": 4.3041344951996746e-05,
"loss": 5.1409,
"step": 118
},
{
"epoch": 1.2081218274111676,
"grad_norm": 0.5669883489608765,
"learning_rate": 4.2178276747988446e-05,
"loss": 5.1294,
"step": 119
},
{
"epoch": 1.218274111675127,
"grad_norm": 0.6567976474761963,
"learning_rate": 4.131759111665349e-05,
"loss": 5.127,
"step": 120
},
{
"epoch": 1.2284263959390862,
"grad_norm": 0.7601398825645447,
"learning_rate": 4.045955023117276e-05,
"loss": 4.9877,
"step": 121
},
{
"epoch": 1.2385786802030456,
"grad_norm": 0.8807834982872009,
"learning_rate": 3.960441545911204e-05,
"loss": 5.2077,
"step": 122
},
{
"epoch": 1.248730964467005,
"grad_norm": 0.9801356196403503,
"learning_rate": 3.875244728280676e-05,
"loss": 4.9937,
"step": 123
},
{
"epoch": 1.2588832487309645,
"grad_norm": 0.7310280799865723,
"learning_rate": 3.790390522001662e-05,
"loss": 5.2336,
"step": 124
},
{
"epoch": 1.2690355329949239,
"grad_norm": 0.560065507888794,
"learning_rate": 3.705904774487396e-05,
"loss": 5.4395,
"step": 125
},
{
"epoch": 1.2791878172588833,
"grad_norm": 0.39355024695396423,
"learning_rate": 3.6218132209150045e-05,
"loss": 5.0604,
"step": 126
},
{
"epoch": 1.2893401015228427,
"grad_norm": 0.3939688503742218,
"learning_rate": 3.5381414763863166e-05,
"loss": 5.2652,
"step": 127
},
{
"epoch": 1.299492385786802,
"grad_norm": 0.38597550988197327,
"learning_rate": 3.4549150281252636e-05,
"loss": 5.1938,
"step": 128
},
{
"epoch": 1.3096446700507614,
"grad_norm": 0.4252302050590515,
"learning_rate": 3.372159227714218e-05,
"loss": 5.1874,
"step": 129
},
{
"epoch": 1.3197969543147208,
"grad_norm": 0.4855193793773651,
"learning_rate": 3.289899283371657e-05,
"loss": 5.273,
"step": 130
},
{
"epoch": 1.3299492385786802,
"grad_norm": 0.5743886232376099,
"learning_rate": 3.2081602522734986e-05,
"loss": 5.2788,
"step": 131
},
{
"epoch": 1.3401015228426396,
"grad_norm": 0.4098842740058899,
"learning_rate": 3.12696703292044e-05,
"loss": 5.1825,
"step": 132
},
{
"epoch": 1.350253807106599,
"grad_norm": 0.3755365014076233,
"learning_rate": 3.046344357553632e-05,
"loss": 5.1837,
"step": 133
},
{
"epoch": 1.3604060913705585,
"grad_norm": 0.4349859356880188,
"learning_rate": 2.9663167846209998e-05,
"loss": 5.2206,
"step": 134
},
{
"epoch": 1.3705583756345177,
"grad_norm": 0.48483696579933167,
"learning_rate": 2.886908691296504e-05,
"loss": 5.2736,
"step": 135
},
{
"epoch": 1.380710659898477,
"grad_norm": 0.5594541430473328,
"learning_rate": 2.8081442660546125e-05,
"loss": 5.1473,
"step": 136
},
{
"epoch": 1.3908629441624365,
"grad_norm": 0.478778213262558,
"learning_rate": 2.7300475013022663e-05,
"loss": 5.1136,
"step": 137
},
{
"epoch": 1.401015228426396,
"grad_norm": 0.47473740577697754,
"learning_rate": 2.6526421860705473e-05,
"loss": 5.1958,
"step": 138
},
{
"epoch": 1.4111675126903553,
"grad_norm": 0.504795491695404,
"learning_rate": 2.575951898768315e-05,
"loss": 5.0851,
"step": 139
},
{
"epoch": 1.4213197969543148,
"grad_norm": 0.4611382484436035,
"learning_rate": 2.500000000000001e-05,
"loss": 5.1284,
"step": 140
},
{
"epoch": 1.4314720812182742,
"grad_norm": 0.49524998664855957,
"learning_rate": 2.4248096254497288e-05,
"loss": 5.2106,
"step": 141
},
{
"epoch": 1.4416243654822334,
"grad_norm": 0.4873555302619934,
"learning_rate": 2.350403678833976e-05,
"loss": 5.1443,
"step": 142
},
{
"epoch": 1.451776649746193,
"grad_norm": 0.5808872580528259,
"learning_rate": 2.2768048249248648e-05,
"loss": 5.0184,
"step": 143
},
{
"epoch": 1.4619289340101522,
"grad_norm": 0.695599377155304,
"learning_rate": 2.2040354826462668e-05,
"loss": 5.2417,
"step": 144
},
{
"epoch": 1.4720812182741116,
"grad_norm": 0.7416836023330688,
"learning_rate": 2.132117818244771e-05,
"loss": 5.0167,
"step": 145
},
{
"epoch": 1.482233502538071,
"grad_norm": 0.9077966213226318,
"learning_rate": 2.061073738537635e-05,
"loss": 5.3683,
"step": 146
},
{
"epoch": 1.4923857868020305,
"grad_norm": 0.7044989466667175,
"learning_rate": 1.9909248842397584e-05,
"loss": 4.8825,
"step": 147
},
{
"epoch": 1.50253807106599,
"grad_norm": 0.5296655893325806,
"learning_rate": 1.9216926233717085e-05,
"loss": 5.17,
"step": 148
},
{
"epoch": 1.512690355329949,
"grad_norm": 0.47559142112731934,
"learning_rate": 1.8533980447508137e-05,
"loss": 5.4245,
"step": 149
},
{
"epoch": 1.5228426395939088,
"grad_norm": 0.4813889265060425,
"learning_rate": 1.7860619515673033e-05,
"loss": 5.1568,
"step": 150
},
{
"epoch": 1.5228426395939088,
"eval_loss": 5.212484836578369,
"eval_runtime": 8.7715,
"eval_samples_per_second": 18.925,
"eval_steps_per_second": 9.462,
"step": 150
},
{
"epoch": 1.532994923857868,
"grad_norm": 0.4661964178085327,
"learning_rate": 1.7197048550474643e-05,
"loss": 5.3035,
"step": 151
},
{
"epoch": 1.5431472081218274,
"grad_norm": 0.4161906838417053,
"learning_rate": 1.6543469682057106e-05,
"loss": 5.249,
"step": 152
},
{
"epoch": 1.5532994923857868,
"grad_norm": 0.3857448697090149,
"learning_rate": 1.5900081996875083e-05,
"loss": 5.1989,
"step": 153
},
{
"epoch": 1.5634517766497462,
"grad_norm": 0.38317275047302246,
"learning_rate": 1.526708147705013e-05,
"loss": 5.2647,
"step": 154
},
{
"epoch": 1.5736040609137056,
"grad_norm": 0.42127397656440735,
"learning_rate": 1.4644660940672627e-05,
"loss": 5.3172,
"step": 155
},
{
"epoch": 1.5837563451776648,
"grad_norm": 0.40184593200683594,
"learning_rate": 1.4033009983067452e-05,
"loss": 5.1843,
"step": 156
},
{
"epoch": 1.5939086294416245,
"grad_norm": 0.39526352286338806,
"learning_rate": 1.3432314919041478e-05,
"loss": 5.2035,
"step": 157
},
{
"epoch": 1.6040609137055837,
"grad_norm": 0.36701011657714844,
"learning_rate": 1.2842758726130283e-05,
"loss": 5.259,
"step": 158
},
{
"epoch": 1.614213197969543,
"grad_norm": 0.3989609181880951,
"learning_rate": 1.22645209888614e-05,
"loss": 5.1989,
"step": 159
},
{
"epoch": 1.6243654822335025,
"grad_norm": 0.4475120007991791,
"learning_rate": 1.1697777844051105e-05,
"loss": 5.1748,
"step": 160
},
{
"epoch": 1.634517766497462,
"grad_norm": 0.4706570506095886,
"learning_rate": 1.1142701927151456e-05,
"loss": 5.1278,
"step": 161
},
{
"epoch": 1.6446700507614214,
"grad_norm": 0.5085177421569824,
"learning_rate": 1.0599462319663905e-05,
"loss": 5.1414,
"step": 162
},
{
"epoch": 1.6548223350253806,
"grad_norm": 0.5082749128341675,
"learning_rate": 1.006822449763537e-05,
"loss": 5.2022,
"step": 163
},
{
"epoch": 1.6649746192893402,
"grad_norm": 0.5005104541778564,
"learning_rate": 9.549150281252633e-06,
"loss": 5.0471,
"step": 164
},
{
"epoch": 1.6751269035532994,
"grad_norm": 0.5254844427108765,
"learning_rate": 9.042397785550405e-06,
"loss": 5.2574,
"step": 165
},
{
"epoch": 1.6852791878172588,
"grad_norm": 0.55852210521698,
"learning_rate": 8.548121372247918e-06,
"loss": 5.1745,
"step": 166
},
{
"epoch": 1.6954314720812182,
"grad_norm": 0.6532874703407288,
"learning_rate": 8.066471602728803e-06,
"loss": 5.1302,
"step": 167
},
{
"epoch": 1.7055837563451777,
"grad_norm": 0.6552545428276062,
"learning_rate": 7.597595192178702e-06,
"loss": 5.0333,
"step": 168
},
{
"epoch": 1.715736040609137,
"grad_norm": 0.7609627842903137,
"learning_rate": 7.1416349648943894e-06,
"loss": 4.9455,
"step": 169
},
{
"epoch": 1.7258883248730963,
"grad_norm": 0.9393100142478943,
"learning_rate": 6.698729810778065e-06,
"loss": 5.2311,
"step": 170
},
{
"epoch": 1.736040609137056,
"grad_norm": 0.6443814635276794,
"learning_rate": 6.269014643030213e-06,
"loss": 4.9513,
"step": 171
},
{
"epoch": 1.7461928934010151,
"grad_norm": 0.37428995966911316,
"learning_rate": 5.852620357053651e-06,
"loss": 5.1892,
"step": 172
},
{
"epoch": 1.7563451776649748,
"grad_norm": 0.3807058036327362,
"learning_rate": 5.449673790581611e-06,
"loss": 5.4823,
"step": 173
},
{
"epoch": 1.766497461928934,
"grad_norm": 0.3458186984062195,
"learning_rate": 5.060297685041659e-06,
"loss": 5.0147,
"step": 174
},
{
"epoch": 1.7766497461928934,
"grad_norm": 0.3948209285736084,
"learning_rate": 4.684610648167503e-06,
"loss": 5.3152,
"step": 175
},
{
"epoch": 1.7868020304568528,
"grad_norm": 0.3780762553215027,
"learning_rate": 4.322727117869951e-06,
"loss": 5.2797,
"step": 176
},
{
"epoch": 1.796954314720812,
"grad_norm": 0.3970668911933899,
"learning_rate": 3.974757327377981e-06,
"loss": 5.153,
"step": 177
},
{
"epoch": 1.8071065989847717,
"grad_norm": 0.42801252007484436,
"learning_rate": 3.6408072716606346e-06,
"loss": 5.2812,
"step": 178
},
{
"epoch": 1.8172588832487309,
"grad_norm": 0.3600170314311981,
"learning_rate": 3.3209786751399187e-06,
"loss": 5.226,
"step": 179
},
{
"epoch": 1.8274111675126905,
"grad_norm": 0.40839913487434387,
"learning_rate": 3.0153689607045845e-06,
"loss": 5.2793,
"step": 180
},
{
"epoch": 1.8375634517766497,
"grad_norm": 0.4590075612068176,
"learning_rate": 2.724071220034158e-06,
"loss": 5.2714,
"step": 181
},
{
"epoch": 1.8477157360406091,
"grad_norm": 0.4016427993774414,
"learning_rate": 2.4471741852423237e-06,
"loss": 5.1897,
"step": 182
},
{
"epoch": 1.8578680203045685,
"grad_norm": 0.382449746131897,
"learning_rate": 2.1847622018482283e-06,
"loss": 5.2737,
"step": 183
},
{
"epoch": 1.868020304568528,
"grad_norm": 0.44841504096984863,
"learning_rate": 1.9369152030840556e-06,
"loss": 5.1636,
"step": 184
},
{
"epoch": 1.8781725888324874,
"grad_norm": 0.47819268703460693,
"learning_rate": 1.70370868554659e-06,
"loss": 5.1766,
"step": 185
},
{
"epoch": 1.8883248730964466,
"grad_norm": 0.4346996545791626,
"learning_rate": 1.4852136862001764e-06,
"loss": 5.1172,
"step": 186
},
{
"epoch": 1.8984771573604062,
"grad_norm": 0.4988497495651245,
"learning_rate": 1.2814967607382432e-06,
"loss": 5.1769,
"step": 187
},
{
"epoch": 1.9086294416243654,
"grad_norm": 0.47669705748558044,
"learning_rate": 1.0926199633097157e-06,
"loss": 5.0482,
"step": 188
},
{
"epoch": 1.9187817258883249,
"grad_norm": 0.505896270275116,
"learning_rate": 9.186408276168013e-07,
"loss": 5.2213,
"step": 189
},
{
"epoch": 1.9289340101522843,
"grad_norm": 0.5310911536216736,
"learning_rate": 7.596123493895991e-07,
"loss": 5.1086,
"step": 190
},
{
"epoch": 1.9390862944162437,
"grad_norm": 0.48757103085517883,
"learning_rate": 6.15582970243117e-07,
"loss": 5.039,
"step": 191
},
{
"epoch": 1.9492385786802031,
"grad_norm": 0.6253727078437805,
"learning_rate": 4.865965629214819e-07,
"loss": 5.0935,
"step": 192
},
{
"epoch": 1.9593908629441623,
"grad_norm": 0.7249897122383118,
"learning_rate": 3.7269241793390085e-07,
"loss": 5.0918,
"step": 193
},
{
"epoch": 1.969543147208122,
"grad_norm": 0.9451597929000854,
"learning_rate": 2.7390523158633554e-07,
"loss": 5.0525,
"step": 194
},
{
"epoch": 1.9796954314720812,
"grad_norm": 0.7079738974571228,
"learning_rate": 1.9026509541272275e-07,
"loss": 5.5103,
"step": 195
},
{
"epoch": 1.9898477157360406,
"grad_norm": 0.3672516644001007,
"learning_rate": 1.2179748700879012e-07,
"loss": 5.1374,
"step": 196
},
{
"epoch": 2.0,
"grad_norm": 1.114559292793274,
"learning_rate": 6.852326227130834e-08,
"loss": 6.8999,
"step": 197
},
{
"epoch": 2.010152284263959,
"grad_norm": 0.32475075125694275,
"learning_rate": 3.04586490452119e-08,
"loss": 5.2734,
"step": 198
},
{
"epoch": 2.020304568527919,
"grad_norm": 0.2769867777824402,
"learning_rate": 7.615242180436522e-09,
"loss": 5.2687,
"step": 199
},
{
"epoch": 2.030456852791878,
"grad_norm": 0.2693822979927063,
"learning_rate": 0.0,
"loss": 5.2514,
"step": 200
},
{
"epoch": 2.030456852791878,
"eval_loss": 5.203610897064209,
"eval_runtime": 8.7648,
"eval_samples_per_second": 18.939,
"eval_steps_per_second": 9.47,
"step": 200
}
],
"logging_steps": 1,
"max_steps": 200,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 50,
"stateful_callbacks": {
"EarlyStoppingCallback": {
"args": {
"early_stopping_patience": 5,
"early_stopping_threshold": 0.0
},
"attributes": {
"early_stopping_patience_counter": 0
}
},
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.1363884845760512e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}