prxy5604's picture
Training in progress, step 200, checkpoint
3b5e9f9 verified
{
"best_metric": 1.047837495803833,
"best_model_checkpoint": "miner_id_24/checkpoint-200",
"epoch": 1.322314049586777,
"eval_steps": 50,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.006611570247933884,
"grad_norm": 0.9148093461990356,
"learning_rate": 5e-06,
"loss": 4.3726,
"step": 1
},
{
"epoch": 0.006611570247933884,
"eval_loss": 1.6257771253585815,
"eval_runtime": 14.0331,
"eval_samples_per_second": 18.171,
"eval_steps_per_second": 9.121,
"step": 1
},
{
"epoch": 0.013223140495867768,
"grad_norm": 0.842186689376831,
"learning_rate": 1e-05,
"loss": 4.5082,
"step": 2
},
{
"epoch": 0.019834710743801654,
"grad_norm": 0.7833890318870544,
"learning_rate": 1.5e-05,
"loss": 4.707,
"step": 3
},
{
"epoch": 0.026446280991735537,
"grad_norm": 1.024601697921753,
"learning_rate": 2e-05,
"loss": 4.7761,
"step": 4
},
{
"epoch": 0.03305785123966942,
"grad_norm": 0.9558944702148438,
"learning_rate": 2.5e-05,
"loss": 5.4576,
"step": 5
},
{
"epoch": 0.03966942148760331,
"grad_norm": 0.9868462681770325,
"learning_rate": 3e-05,
"loss": 4.9784,
"step": 6
},
{
"epoch": 0.04628099173553719,
"grad_norm": 1.144509196281433,
"learning_rate": 3.5e-05,
"loss": 4.9578,
"step": 7
},
{
"epoch": 0.05289256198347107,
"grad_norm": 1.2102956771850586,
"learning_rate": 4e-05,
"loss": 5.301,
"step": 8
},
{
"epoch": 0.05950413223140496,
"grad_norm": 1.0776772499084473,
"learning_rate": 4.5e-05,
"loss": 5.2147,
"step": 9
},
{
"epoch": 0.06611570247933884,
"grad_norm": 1.1019762754440308,
"learning_rate": 5e-05,
"loss": 5.2903,
"step": 10
},
{
"epoch": 0.07272727272727272,
"grad_norm": 1.2564432621002197,
"learning_rate": 5.500000000000001e-05,
"loss": 5.5129,
"step": 11
},
{
"epoch": 0.07933884297520662,
"grad_norm": 1.0597630739212036,
"learning_rate": 6e-05,
"loss": 5.3251,
"step": 12
},
{
"epoch": 0.0859504132231405,
"grad_norm": 1.242934226989746,
"learning_rate": 6.500000000000001e-05,
"loss": 4.8806,
"step": 13
},
{
"epoch": 0.09256198347107437,
"grad_norm": 1.4951207637786865,
"learning_rate": 7e-05,
"loss": 5.2546,
"step": 14
},
{
"epoch": 0.09917355371900827,
"grad_norm": 1.3901413679122925,
"learning_rate": 7.500000000000001e-05,
"loss": 5.3295,
"step": 15
},
{
"epoch": 0.10578512396694215,
"grad_norm": 1.412446141242981,
"learning_rate": 8e-05,
"loss": 5.5863,
"step": 16
},
{
"epoch": 0.11239669421487604,
"grad_norm": 1.4029113054275513,
"learning_rate": 8.5e-05,
"loss": 5.4345,
"step": 17
},
{
"epoch": 0.11900826446280992,
"grad_norm": 1.4717167615890503,
"learning_rate": 9e-05,
"loss": 5.2476,
"step": 18
},
{
"epoch": 0.1256198347107438,
"grad_norm": 1.39309823513031,
"learning_rate": 9.5e-05,
"loss": 5.0354,
"step": 19
},
{
"epoch": 0.1322314049586777,
"grad_norm": 1.5791093111038208,
"learning_rate": 0.0001,
"loss": 5.2546,
"step": 20
},
{
"epoch": 0.13884297520661157,
"grad_norm": 1.4860973358154297,
"learning_rate": 9.999238475781957e-05,
"loss": 5.047,
"step": 21
},
{
"epoch": 0.14545454545454545,
"grad_norm": 1.7063889503479004,
"learning_rate": 9.99695413509548e-05,
"loss": 5.1362,
"step": 22
},
{
"epoch": 0.15206611570247933,
"grad_norm": 1.6598552465438843,
"learning_rate": 9.99314767377287e-05,
"loss": 5.7295,
"step": 23
},
{
"epoch": 0.15867768595041323,
"grad_norm": 1.758675456047058,
"learning_rate": 9.987820251299122e-05,
"loss": 5.0831,
"step": 24
},
{
"epoch": 0.1652892561983471,
"grad_norm": 1.7304877042770386,
"learning_rate": 9.980973490458728e-05,
"loss": 5.3635,
"step": 25
},
{
"epoch": 0.171900826446281,
"grad_norm": 1.9273478984832764,
"learning_rate": 9.972609476841367e-05,
"loss": 5.7037,
"step": 26
},
{
"epoch": 0.17851239669421487,
"grad_norm": 2.0695853233337402,
"learning_rate": 9.962730758206611e-05,
"loss": 4.8804,
"step": 27
},
{
"epoch": 0.18512396694214875,
"grad_norm": 1.8753983974456787,
"learning_rate": 9.951340343707852e-05,
"loss": 5.7339,
"step": 28
},
{
"epoch": 0.19173553719008266,
"grad_norm": 1.8280308246612549,
"learning_rate": 9.938441702975689e-05,
"loss": 5.7124,
"step": 29
},
{
"epoch": 0.19834710743801653,
"grad_norm": 2.103041172027588,
"learning_rate": 9.924038765061042e-05,
"loss": 6.043,
"step": 30
},
{
"epoch": 0.2049586776859504,
"grad_norm": 2.314613103866577,
"learning_rate": 9.908135917238321e-05,
"loss": 6.096,
"step": 31
},
{
"epoch": 0.2115702479338843,
"grad_norm": 2.1027143001556396,
"learning_rate": 9.890738003669029e-05,
"loss": 5.7764,
"step": 32
},
{
"epoch": 0.21818181818181817,
"grad_norm": 2.3914480209350586,
"learning_rate": 9.871850323926177e-05,
"loss": 5.6048,
"step": 33
},
{
"epoch": 0.22479338842975208,
"grad_norm": 3.5736582279205322,
"learning_rate": 9.851478631379982e-05,
"loss": 4.9323,
"step": 34
},
{
"epoch": 0.23140495867768596,
"grad_norm": 2.4157803058624268,
"learning_rate": 9.829629131445342e-05,
"loss": 5.3481,
"step": 35
},
{
"epoch": 0.23801652892561984,
"grad_norm": 3.495314121246338,
"learning_rate": 9.806308479691595e-05,
"loss": 6.1046,
"step": 36
},
{
"epoch": 0.24462809917355371,
"grad_norm": 3.9649362564086914,
"learning_rate": 9.781523779815179e-05,
"loss": 5.6288,
"step": 37
},
{
"epoch": 0.2512396694214876,
"grad_norm": 1.5807100534439087,
"learning_rate": 9.755282581475769e-05,
"loss": 4.219,
"step": 38
},
{
"epoch": 0.2578512396694215,
"grad_norm": 1.2698018550872803,
"learning_rate": 9.727592877996585e-05,
"loss": 4.1076,
"step": 39
},
{
"epoch": 0.2644628099173554,
"grad_norm": 1.089763879776001,
"learning_rate": 9.698463103929542e-05,
"loss": 3.9567,
"step": 40
},
{
"epoch": 0.27107438016528923,
"grad_norm": 0.9421190023422241,
"learning_rate": 9.667902132486009e-05,
"loss": 4.3975,
"step": 41
},
{
"epoch": 0.27768595041322314,
"grad_norm": 0.7669892907142639,
"learning_rate": 9.635919272833938e-05,
"loss": 4.5829,
"step": 42
},
{
"epoch": 0.28429752066115704,
"grad_norm": 0.7751649022102356,
"learning_rate": 9.602524267262203e-05,
"loss": 4.4129,
"step": 43
},
{
"epoch": 0.2909090909090909,
"grad_norm": 0.9909483194351196,
"learning_rate": 9.567727288213005e-05,
"loss": 4.7576,
"step": 44
},
{
"epoch": 0.2975206611570248,
"grad_norm": 0.7943483591079712,
"learning_rate": 9.53153893518325e-05,
"loss": 4.4143,
"step": 45
},
{
"epoch": 0.30413223140495865,
"grad_norm": 0.8910554051399231,
"learning_rate": 9.493970231495835e-05,
"loss": 4.8309,
"step": 46
},
{
"epoch": 0.31074380165289256,
"grad_norm": 0.8945802450180054,
"learning_rate": 9.45503262094184e-05,
"loss": 4.3729,
"step": 47
},
{
"epoch": 0.31735537190082647,
"grad_norm": 0.9367103576660156,
"learning_rate": 9.414737964294636e-05,
"loss": 4.7947,
"step": 48
},
{
"epoch": 0.3239669421487603,
"grad_norm": 0.9118549227714539,
"learning_rate": 9.373098535696979e-05,
"loss": 4.9757,
"step": 49
},
{
"epoch": 0.3305785123966942,
"grad_norm": 0.9432209134101868,
"learning_rate": 9.330127018922194e-05,
"loss": 4.6552,
"step": 50
},
{
"epoch": 0.3305785123966942,
"eval_loss": 1.1729106903076172,
"eval_runtime": 14.1768,
"eval_samples_per_second": 17.987,
"eval_steps_per_second": 9.029,
"step": 50
},
{
"epoch": 0.3371900826446281,
"grad_norm": 0.9313036203384399,
"learning_rate": 9.285836503510562e-05,
"loss": 4.8954,
"step": 51
},
{
"epoch": 0.343801652892562,
"grad_norm": 0.8967971205711365,
"learning_rate": 9.24024048078213e-05,
"loss": 4.2944,
"step": 52
},
{
"epoch": 0.3504132231404959,
"grad_norm": 0.9805819392204285,
"learning_rate": 9.193352839727121e-05,
"loss": 5.1855,
"step": 53
},
{
"epoch": 0.35702479338842974,
"grad_norm": 1.110192060470581,
"learning_rate": 9.145187862775209e-05,
"loss": 5.0659,
"step": 54
},
{
"epoch": 0.36363636363636365,
"grad_norm": 1.0864770412445068,
"learning_rate": 9.09576022144496e-05,
"loss": 4.6465,
"step": 55
},
{
"epoch": 0.3702479338842975,
"grad_norm": 1.6741044521331787,
"learning_rate": 9.045084971874738e-05,
"loss": 4.5788,
"step": 56
},
{
"epoch": 0.3768595041322314,
"grad_norm": 1.2498528957366943,
"learning_rate": 8.993177550236464e-05,
"loss": 4.8934,
"step": 57
},
{
"epoch": 0.3834710743801653,
"grad_norm": 1.2537343502044678,
"learning_rate": 8.940053768033609e-05,
"loss": 4.7502,
"step": 58
},
{
"epoch": 0.39008264462809916,
"grad_norm": 1.2770891189575195,
"learning_rate": 8.885729807284856e-05,
"loss": 4.8936,
"step": 59
},
{
"epoch": 0.39669421487603307,
"grad_norm": 1.4430948495864868,
"learning_rate": 8.83022221559489e-05,
"loss": 4.8503,
"step": 60
},
{
"epoch": 0.4033057851239669,
"grad_norm": 1.5071772336959839,
"learning_rate": 8.773547901113862e-05,
"loss": 5.3403,
"step": 61
},
{
"epoch": 0.4099173553719008,
"grad_norm": 1.624732494354248,
"learning_rate": 8.715724127386972e-05,
"loss": 5.3317,
"step": 62
},
{
"epoch": 0.41652892561983473,
"grad_norm": 1.6274948120117188,
"learning_rate": 8.656768508095853e-05,
"loss": 4.4587,
"step": 63
},
{
"epoch": 0.4231404958677686,
"grad_norm": 1.6330101490020752,
"learning_rate": 8.596699001693255e-05,
"loss": 4.1029,
"step": 64
},
{
"epoch": 0.4297520661157025,
"grad_norm": 1.6284996271133423,
"learning_rate": 8.535533905932738e-05,
"loss": 5.1398,
"step": 65
},
{
"epoch": 0.43636363636363634,
"grad_norm": 1.741026520729065,
"learning_rate": 8.473291852294987e-05,
"loss": 5.2479,
"step": 66
},
{
"epoch": 0.44297520661157025,
"grad_norm": 1.8886727094650269,
"learning_rate": 8.409991800312493e-05,
"loss": 5.6305,
"step": 67
},
{
"epoch": 0.44958677685950416,
"grad_norm": 1.9955488443374634,
"learning_rate": 8.345653031794292e-05,
"loss": 5.6926,
"step": 68
},
{
"epoch": 0.456198347107438,
"grad_norm": 2.239786386489868,
"learning_rate": 8.280295144952536e-05,
"loss": 5.9947,
"step": 69
},
{
"epoch": 0.4628099173553719,
"grad_norm": 2.239755630493164,
"learning_rate": 8.213938048432697e-05,
"loss": 4.4617,
"step": 70
},
{
"epoch": 0.46942148760330576,
"grad_norm": 2.663750648498535,
"learning_rate": 8.146601955249188e-05,
"loss": 4.2143,
"step": 71
},
{
"epoch": 0.47603305785123967,
"grad_norm": 2.472973108291626,
"learning_rate": 8.07830737662829e-05,
"loss": 4.8717,
"step": 72
},
{
"epoch": 0.4826446280991736,
"grad_norm": 3.02223801612854,
"learning_rate": 8.009075115760243e-05,
"loss": 5.2497,
"step": 73
},
{
"epoch": 0.48925619834710743,
"grad_norm": 4.158394813537598,
"learning_rate": 7.938926261462366e-05,
"loss": 4.8917,
"step": 74
},
{
"epoch": 0.49586776859504134,
"grad_norm": 1.0667253732681274,
"learning_rate": 7.86788218175523e-05,
"loss": 3.4973,
"step": 75
},
{
"epoch": 0.5024793388429752,
"grad_norm": 1.0640274286270142,
"learning_rate": 7.795964517353735e-05,
"loss": 3.6,
"step": 76
},
{
"epoch": 0.509090909090909,
"grad_norm": 0.8343002200126648,
"learning_rate": 7.723195175075136e-05,
"loss": 3.9431,
"step": 77
},
{
"epoch": 0.515702479338843,
"grad_norm": 0.8102278709411621,
"learning_rate": 7.649596321166024e-05,
"loss": 4.2592,
"step": 78
},
{
"epoch": 0.5223140495867769,
"grad_norm": 0.7733540534973145,
"learning_rate": 7.575190374550272e-05,
"loss": 4.4856,
"step": 79
},
{
"epoch": 0.5289256198347108,
"grad_norm": 0.6886125802993774,
"learning_rate": 7.500000000000001e-05,
"loss": 4.4077,
"step": 80
},
{
"epoch": 0.5355371900826447,
"grad_norm": 0.7023859620094299,
"learning_rate": 7.424048101231686e-05,
"loss": 4.4471,
"step": 81
},
{
"epoch": 0.5421487603305785,
"grad_norm": 0.733121931552887,
"learning_rate": 7.347357813929454e-05,
"loss": 4.2295,
"step": 82
},
{
"epoch": 0.5487603305785124,
"grad_norm": 1.4499454498291016,
"learning_rate": 7.269952498697734e-05,
"loss": 4.4026,
"step": 83
},
{
"epoch": 0.5553719008264463,
"grad_norm": 0.795872151851654,
"learning_rate": 7.191855733945387e-05,
"loss": 4.5716,
"step": 84
},
{
"epoch": 0.5619834710743802,
"grad_norm": 0.8914481997489929,
"learning_rate": 7.113091308703498e-05,
"loss": 4.6541,
"step": 85
},
{
"epoch": 0.5685950413223141,
"grad_norm": 0.9090842008590698,
"learning_rate": 7.033683215379002e-05,
"loss": 4.3894,
"step": 86
},
{
"epoch": 0.5752066115702479,
"grad_norm": 0.9541610479354858,
"learning_rate": 6.953655642446368e-05,
"loss": 4.1603,
"step": 87
},
{
"epoch": 0.5818181818181818,
"grad_norm": 1.0280675888061523,
"learning_rate": 6.873032967079561e-05,
"loss": 4.4232,
"step": 88
},
{
"epoch": 0.5884297520661157,
"grad_norm": 1.1509857177734375,
"learning_rate": 6.7918397477265e-05,
"loss": 4.442,
"step": 89
},
{
"epoch": 0.5950413223140496,
"grad_norm": 1.0895159244537354,
"learning_rate": 6.710100716628344e-05,
"loss": 4.802,
"step": 90
},
{
"epoch": 0.6016528925619835,
"grad_norm": 1.1312400102615356,
"learning_rate": 6.627840772285784e-05,
"loss": 4.6077,
"step": 91
},
{
"epoch": 0.6082644628099173,
"grad_norm": 1.2083485126495361,
"learning_rate": 6.545084971874738e-05,
"loss": 4.686,
"step": 92
},
{
"epoch": 0.6148760330578512,
"grad_norm": 1.1744060516357422,
"learning_rate": 6.461858523613684e-05,
"loss": 4.7755,
"step": 93
},
{
"epoch": 0.6214876033057851,
"grad_norm": 1.1983288526535034,
"learning_rate": 6.378186779084995e-05,
"loss": 4.5161,
"step": 94
},
{
"epoch": 0.628099173553719,
"grad_norm": 1.2709901332855225,
"learning_rate": 6.294095225512603e-05,
"loss": 4.6903,
"step": 95
},
{
"epoch": 0.6347107438016529,
"grad_norm": 1.2853686809539795,
"learning_rate": 6.209609477998338e-05,
"loss": 4.8628,
"step": 96
},
{
"epoch": 0.6413223140495867,
"grad_norm": 1.4955391883850098,
"learning_rate": 6.124755271719325e-05,
"loss": 5.4237,
"step": 97
},
{
"epoch": 0.6479338842975206,
"grad_norm": 1.4081076383590698,
"learning_rate": 6.0395584540887963e-05,
"loss": 5.2527,
"step": 98
},
{
"epoch": 0.6545454545454545,
"grad_norm": 1.5048182010650635,
"learning_rate": 5.9540449768827246e-05,
"loss": 4.8745,
"step": 99
},
{
"epoch": 0.6611570247933884,
"grad_norm": 1.615033507347107,
"learning_rate": 5.868240888334653e-05,
"loss": 4.8292,
"step": 100
},
{
"epoch": 0.6611570247933884,
"eval_loss": 1.0880825519561768,
"eval_runtime": 14.1578,
"eval_samples_per_second": 18.011,
"eval_steps_per_second": 9.041,
"step": 100
},
{
"epoch": 0.6677685950413224,
"grad_norm": 1.748067021369934,
"learning_rate": 5.782172325201155e-05,
"loss": 4.4367,
"step": 101
},
{
"epoch": 0.6743801652892562,
"grad_norm": 1.8786406517028809,
"learning_rate": 5.695865504800327e-05,
"loss": 5.3412,
"step": 102
},
{
"epoch": 0.6809917355371901,
"grad_norm": 1.8346900939941406,
"learning_rate": 5.6093467170257374e-05,
"loss": 4.9618,
"step": 103
},
{
"epoch": 0.687603305785124,
"grad_norm": 2.015852928161621,
"learning_rate": 5.522642316338268e-05,
"loss": 5.5926,
"step": 104
},
{
"epoch": 0.6942148760330579,
"grad_norm": 2.249979019165039,
"learning_rate": 5.435778713738292e-05,
"loss": 4.8259,
"step": 105
},
{
"epoch": 0.7008264462809918,
"grad_norm": 2.263035774230957,
"learning_rate": 5.348782368720626e-05,
"loss": 5.104,
"step": 106
},
{
"epoch": 0.7074380165289256,
"grad_norm": 2.640855550765991,
"learning_rate": 5.26167978121472e-05,
"loss": 5.146,
"step": 107
},
{
"epoch": 0.7140495867768595,
"grad_norm": 3.1351351737976074,
"learning_rate": 5.174497483512506e-05,
"loss": 4.2835,
"step": 108
},
{
"epoch": 0.7206611570247934,
"grad_norm": 2.632106304168701,
"learning_rate": 5.0872620321864185e-05,
"loss": 4.8549,
"step": 109
},
{
"epoch": 0.7272727272727273,
"grad_norm": 2.945842742919922,
"learning_rate": 5e-05,
"loss": 4.8564,
"step": 110
},
{
"epoch": 0.7338842975206612,
"grad_norm": 3.851283550262451,
"learning_rate": 4.912737967813583e-05,
"loss": 5.1433,
"step": 111
},
{
"epoch": 0.740495867768595,
"grad_norm": 0.6680141687393188,
"learning_rate": 4.825502516487497e-05,
"loss": 3.4089,
"step": 112
},
{
"epoch": 0.7471074380165289,
"grad_norm": 0.7174019813537598,
"learning_rate": 4.738320218785281e-05,
"loss": 3.8666,
"step": 113
},
{
"epoch": 0.7537190082644628,
"grad_norm": 0.7039559483528137,
"learning_rate": 4.6512176312793736e-05,
"loss": 3.7843,
"step": 114
},
{
"epoch": 0.7603305785123967,
"grad_norm": 0.7135390043258667,
"learning_rate": 4.564221286261709e-05,
"loss": 3.8314,
"step": 115
},
{
"epoch": 0.7669421487603306,
"grad_norm": 0.7308880686759949,
"learning_rate": 4.477357683661734e-05,
"loss": 4.1674,
"step": 116
},
{
"epoch": 0.7735537190082644,
"grad_norm": 0.7223635315895081,
"learning_rate": 4.390653282974264e-05,
"loss": 3.8675,
"step": 117
},
{
"epoch": 0.7801652892561983,
"grad_norm": 0.7319607138633728,
"learning_rate": 4.3041344951996746e-05,
"loss": 3.9047,
"step": 118
},
{
"epoch": 0.7867768595041322,
"grad_norm": 0.7675497531890869,
"learning_rate": 4.2178276747988446e-05,
"loss": 4.1523,
"step": 119
},
{
"epoch": 0.7933884297520661,
"grad_norm": 0.8165543079376221,
"learning_rate": 4.131759111665349e-05,
"loss": 4.1521,
"step": 120
},
{
"epoch": 0.8,
"grad_norm": 0.8304678797721863,
"learning_rate": 4.045955023117276e-05,
"loss": 4.1744,
"step": 121
},
{
"epoch": 0.8066115702479338,
"grad_norm": 0.8759053945541382,
"learning_rate": 3.960441545911204e-05,
"loss": 4.2122,
"step": 122
},
{
"epoch": 0.8132231404958677,
"grad_norm": 0.8702043890953064,
"learning_rate": 3.875244728280676e-05,
"loss": 4.427,
"step": 123
},
{
"epoch": 0.8198347107438017,
"grad_norm": 0.9653429985046387,
"learning_rate": 3.790390522001662e-05,
"loss": 4.2492,
"step": 124
},
{
"epoch": 0.8264462809917356,
"grad_norm": 0.9605560302734375,
"learning_rate": 3.705904774487396e-05,
"loss": 4.6257,
"step": 125
},
{
"epoch": 0.8330578512396695,
"grad_norm": 0.9582360982894897,
"learning_rate": 3.6218132209150045e-05,
"loss": 3.8951,
"step": 126
},
{
"epoch": 0.8396694214876033,
"grad_norm": 1.1498273611068726,
"learning_rate": 3.5381414763863166e-05,
"loss": 4.4283,
"step": 127
},
{
"epoch": 0.8462809917355372,
"grad_norm": 1.1258846521377563,
"learning_rate": 3.4549150281252636e-05,
"loss": 4.792,
"step": 128
},
{
"epoch": 0.8528925619834711,
"grad_norm": 1.4514033794403076,
"learning_rate": 3.372159227714218e-05,
"loss": 4.5196,
"step": 129
},
{
"epoch": 0.859504132231405,
"grad_norm": 1.3290873765945435,
"learning_rate": 3.289899283371657e-05,
"loss": 4.5567,
"step": 130
},
{
"epoch": 0.8661157024793389,
"grad_norm": 1.4877480268478394,
"learning_rate": 3.2081602522734986e-05,
"loss": 4.5017,
"step": 131
},
{
"epoch": 0.8727272727272727,
"grad_norm": 1.3383150100708008,
"learning_rate": 3.12696703292044e-05,
"loss": 4.966,
"step": 132
},
{
"epoch": 0.8793388429752066,
"grad_norm": 1.3537639379501343,
"learning_rate": 3.046344357553632e-05,
"loss": 4.9792,
"step": 133
},
{
"epoch": 0.8859504132231405,
"grad_norm": 1.4388115406036377,
"learning_rate": 2.9663167846209998e-05,
"loss": 4.5168,
"step": 134
},
{
"epoch": 0.8925619834710744,
"grad_norm": 1.5659295320510864,
"learning_rate": 2.886908691296504e-05,
"loss": 4.675,
"step": 135
},
{
"epoch": 0.8991735537190083,
"grad_norm": 1.614781141281128,
"learning_rate": 2.8081442660546125e-05,
"loss": 4.4682,
"step": 136
},
{
"epoch": 0.9057851239669421,
"grad_norm": 1.750377893447876,
"learning_rate": 2.7300475013022663e-05,
"loss": 4.9327,
"step": 137
},
{
"epoch": 0.912396694214876,
"grad_norm": 2.194397449493408,
"learning_rate": 2.6526421860705473e-05,
"loss": 5.1882,
"step": 138
},
{
"epoch": 0.9190082644628099,
"grad_norm": 1.799832820892334,
"learning_rate": 2.575951898768315e-05,
"loss": 3.5735,
"step": 139
},
{
"epoch": 0.9256198347107438,
"grad_norm": 1.9374184608459473,
"learning_rate": 2.500000000000001e-05,
"loss": 5.4029,
"step": 140
},
{
"epoch": 0.9322314049586777,
"grad_norm": 2.2784645557403564,
"learning_rate": 2.4248096254497288e-05,
"loss": 5.2612,
"step": 141
},
{
"epoch": 0.9388429752066115,
"grad_norm": 2.151506185531616,
"learning_rate": 2.350403678833976e-05,
"loss": 5.2387,
"step": 142
},
{
"epoch": 0.9454545454545454,
"grad_norm": 2.211778163909912,
"learning_rate": 2.2768048249248648e-05,
"loss": 5.3567,
"step": 143
},
{
"epoch": 0.9520661157024793,
"grad_norm": 2.3867406845092773,
"learning_rate": 2.2040354826462668e-05,
"loss": 4.5479,
"step": 144
},
{
"epoch": 0.9586776859504132,
"grad_norm": 2.5046918392181396,
"learning_rate": 2.132117818244771e-05,
"loss": 4.0853,
"step": 145
},
{
"epoch": 0.9652892561983472,
"grad_norm": 2.8745017051696777,
"learning_rate": 2.061073738537635e-05,
"loss": 4.4699,
"step": 146
},
{
"epoch": 0.971900826446281,
"grad_norm": 3.2906479835510254,
"learning_rate": 1.9909248842397584e-05,
"loss": 5.3134,
"step": 147
},
{
"epoch": 0.9785123966942149,
"grad_norm": 3.933941125869751,
"learning_rate": 1.9216926233717085e-05,
"loss": 5.3385,
"step": 148
},
{
"epoch": 0.9851239669421488,
"grad_norm": 0.7648929953575134,
"learning_rate": 1.8533980447508137e-05,
"loss": 3.8574,
"step": 149
},
{
"epoch": 0.9917355371900827,
"grad_norm": 1.3944673538208008,
"learning_rate": 1.7860619515673033e-05,
"loss": 4.5333,
"step": 150
},
{
"epoch": 0.9917355371900827,
"eval_loss": 1.0583877563476562,
"eval_runtime": 14.1642,
"eval_samples_per_second": 18.003,
"eval_steps_per_second": 9.037,
"step": 150
},
{
"epoch": 0.9983471074380166,
"grad_norm": 2.0663795471191406,
"learning_rate": 1.7197048550474643e-05,
"loss": 4.4136,
"step": 151
},
{
"epoch": 1.0049586776859505,
"grad_norm": 2.9669547080993652,
"learning_rate": 1.6543469682057106e-05,
"loss": 3.9654,
"step": 152
},
{
"epoch": 1.0115702479338844,
"grad_norm": 0.582588255405426,
"learning_rate": 1.5900081996875083e-05,
"loss": 3.8065,
"step": 153
},
{
"epoch": 1.018181818181818,
"grad_norm": 0.5635128021240234,
"learning_rate": 1.526708147705013e-05,
"loss": 3.506,
"step": 154
},
{
"epoch": 1.024793388429752,
"grad_norm": 0.643647313117981,
"learning_rate": 1.4644660940672627e-05,
"loss": 3.4885,
"step": 155
},
{
"epoch": 1.031404958677686,
"grad_norm": 0.7142451405525208,
"learning_rate": 1.4033009983067452e-05,
"loss": 3.9004,
"step": 156
},
{
"epoch": 1.0380165289256198,
"grad_norm": 0.7921721339225769,
"learning_rate": 1.3432314919041478e-05,
"loss": 3.6416,
"step": 157
},
{
"epoch": 1.0446280991735537,
"grad_norm": 1.8580390214920044,
"learning_rate": 1.2842758726130283e-05,
"loss": 4.0421,
"step": 158
},
{
"epoch": 1.0512396694214876,
"grad_norm": 0.788405179977417,
"learning_rate": 1.22645209888614e-05,
"loss": 4.1674,
"step": 159
},
{
"epoch": 1.0578512396694215,
"grad_norm": 0.8433780670166016,
"learning_rate": 1.1697777844051105e-05,
"loss": 3.8614,
"step": 160
},
{
"epoch": 1.0644628099173554,
"grad_norm": 1.0130237340927124,
"learning_rate": 1.1142701927151456e-05,
"loss": 4.057,
"step": 161
},
{
"epoch": 1.0710743801652893,
"grad_norm": 0.9251139760017395,
"learning_rate": 1.0599462319663905e-05,
"loss": 4.121,
"step": 162
},
{
"epoch": 1.0776859504132232,
"grad_norm": 1.0490461587905884,
"learning_rate": 1.006822449763537e-05,
"loss": 3.7495,
"step": 163
},
{
"epoch": 1.084297520661157,
"grad_norm": 0.962730884552002,
"learning_rate": 9.549150281252633e-06,
"loss": 3.7831,
"step": 164
},
{
"epoch": 1.0909090909090908,
"grad_norm": 1.0215270519256592,
"learning_rate": 9.042397785550405e-06,
"loss": 3.8681,
"step": 165
},
{
"epoch": 1.0975206611570247,
"grad_norm": 1.097220778465271,
"learning_rate": 8.548121372247918e-06,
"loss": 4.1963,
"step": 166
},
{
"epoch": 1.1041322314049586,
"grad_norm": 1.1427369117736816,
"learning_rate": 8.066471602728803e-06,
"loss": 4.4926,
"step": 167
},
{
"epoch": 1.1107438016528925,
"grad_norm": 1.1889334917068481,
"learning_rate": 7.597595192178702e-06,
"loss": 4.1799,
"step": 168
},
{
"epoch": 1.1173553719008265,
"grad_norm": 1.2882307767868042,
"learning_rate": 7.1416349648943894e-06,
"loss": 4.7466,
"step": 169
},
{
"epoch": 1.1239669421487604,
"grad_norm": 1.3270764350891113,
"learning_rate": 6.698729810778065e-06,
"loss": 4.0636,
"step": 170
},
{
"epoch": 1.1305785123966943,
"grad_norm": 1.4401865005493164,
"learning_rate": 6.269014643030213e-06,
"loss": 4.5355,
"step": 171
},
{
"epoch": 1.1371900826446282,
"grad_norm": 1.280537486076355,
"learning_rate": 5.852620357053651e-06,
"loss": 4.0943,
"step": 172
},
{
"epoch": 1.143801652892562,
"grad_norm": 1.436613917350769,
"learning_rate": 5.449673790581611e-06,
"loss": 4.197,
"step": 173
},
{
"epoch": 1.1504132231404958,
"grad_norm": 1.4517698287963867,
"learning_rate": 5.060297685041659e-06,
"loss": 4.1342,
"step": 174
},
{
"epoch": 1.1570247933884297,
"grad_norm": 1.565581202507019,
"learning_rate": 4.684610648167503e-06,
"loss": 4.4184,
"step": 175
},
{
"epoch": 1.1636363636363636,
"grad_norm": 1.619643211364746,
"learning_rate": 4.322727117869951e-06,
"loss": 4.4695,
"step": 176
},
{
"epoch": 1.1702479338842975,
"grad_norm": 1.635844111442566,
"learning_rate": 3.974757327377981e-06,
"loss": 4.2089,
"step": 177
},
{
"epoch": 1.1768595041322314,
"grad_norm": 1.7811866998672485,
"learning_rate": 3.6408072716606346e-06,
"loss": 4.0089,
"step": 178
},
{
"epoch": 1.1834710743801653,
"grad_norm": 1.8893351554870605,
"learning_rate": 3.3209786751399187e-06,
"loss": 3.1828,
"step": 179
},
{
"epoch": 1.1900826446280992,
"grad_norm": 1.8557778596878052,
"learning_rate": 3.0153689607045845e-06,
"loss": 4.2522,
"step": 180
},
{
"epoch": 1.1966942148760331,
"grad_norm": 1.9851038455963135,
"learning_rate": 2.724071220034158e-06,
"loss": 4.7269,
"step": 181
},
{
"epoch": 1.203305785123967,
"grad_norm": 2.1339123249053955,
"learning_rate": 2.4471741852423237e-06,
"loss": 5.0067,
"step": 182
},
{
"epoch": 1.2099173553719007,
"grad_norm": 2.1520884037017822,
"learning_rate": 2.1847622018482283e-06,
"loss": 4.8541,
"step": 183
},
{
"epoch": 1.2165289256198348,
"grad_norm": 2.4593212604522705,
"learning_rate": 1.9369152030840556e-06,
"loss": 4.4116,
"step": 184
},
{
"epoch": 1.2231404958677685,
"grad_norm": 2.4656741619110107,
"learning_rate": 1.70370868554659e-06,
"loss": 3.4118,
"step": 185
},
{
"epoch": 1.2297520661157024,
"grad_norm": 2.7001984119415283,
"learning_rate": 1.4852136862001764e-06,
"loss": 3.9351,
"step": 186
},
{
"epoch": 1.2363636363636363,
"grad_norm": 3.269486904144287,
"learning_rate": 1.2814967607382432e-06,
"loss": 4.4956,
"step": 187
},
{
"epoch": 1.2429752066115702,
"grad_norm": 3.6578261852264404,
"learning_rate": 1.0926199633097157e-06,
"loss": 3.9351,
"step": 188
},
{
"epoch": 1.2495867768595041,
"grad_norm": 2.344083547592163,
"learning_rate": 9.186408276168013e-07,
"loss": 3.9633,
"step": 189
},
{
"epoch": 1.256198347107438,
"grad_norm": 0.6168360710144043,
"learning_rate": 7.596123493895991e-07,
"loss": 3.5821,
"step": 190
},
{
"epoch": 1.262809917355372,
"grad_norm": 0.6475507616996765,
"learning_rate": 6.15582970243117e-07,
"loss": 2.811,
"step": 191
},
{
"epoch": 1.2694214876033059,
"grad_norm": 0.6254196166992188,
"learning_rate": 4.865965629214819e-07,
"loss": 3.8433,
"step": 192
},
{
"epoch": 1.2760330578512398,
"grad_norm": 0.6778838038444519,
"learning_rate": 3.7269241793390085e-07,
"loss": 4.0597,
"step": 193
},
{
"epoch": 1.2826446280991735,
"grad_norm": 0.8316207528114319,
"learning_rate": 2.7390523158633554e-07,
"loss": 3.8344,
"step": 194
},
{
"epoch": 1.2892561983471074,
"grad_norm": 1.414699673652649,
"learning_rate": 1.9026509541272275e-07,
"loss": 4.2579,
"step": 195
},
{
"epoch": 1.2958677685950413,
"grad_norm": 0.8437047004699707,
"learning_rate": 1.2179748700879012e-07,
"loss": 3.9387,
"step": 196
},
{
"epoch": 1.3024793388429752,
"grad_norm": 1.736149787902832,
"learning_rate": 6.852326227130834e-08,
"loss": 4.1031,
"step": 197
},
{
"epoch": 1.309090909090909,
"grad_norm": 0.9786373376846313,
"learning_rate": 3.04586490452119e-08,
"loss": 3.9919,
"step": 198
},
{
"epoch": 1.315702479338843,
"grad_norm": 0.9801681041717529,
"learning_rate": 7.615242180436522e-09,
"loss": 3.9937,
"step": 199
},
{
"epoch": 1.322314049586777,
"grad_norm": 0.9714708924293518,
"learning_rate": 0.0,
"loss": 4.4521,
"step": 200
},
{
"epoch": 1.322314049586777,
"eval_loss": 1.047837495803833,
"eval_runtime": 14.1574,
"eval_samples_per_second": 18.012,
"eval_steps_per_second": 9.041,
"step": 200
}
],
"logging_steps": 1,
"max_steps": 200,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"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": 1.6229100552192e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}