teleut-7b-rpink-ckpt232-adpt / trainer_state.json
Fizzarolli's picture
Upload folder using huggingface_hub
bf35252 verified
raw
history blame
40.8 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.8140350877192982,
"eval_steps": 500,
"global_step": 232,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0035087719298245615,
"grad_norm": 0.15189611911773682,
"learning_rate": 2.4000000000000003e-06,
"loss": 2.1379,
"step": 1
},
{
"epoch": 0.007017543859649123,
"grad_norm": 0.1405300348997116,
"learning_rate": 4.800000000000001e-06,
"loss": 2.4083,
"step": 2
},
{
"epoch": 0.010526315789473684,
"grad_norm": 0.12744449079036713,
"learning_rate": 7.2e-06,
"loss": 2.3487,
"step": 3
},
{
"epoch": 0.014035087719298246,
"grad_norm": 0.11590863764286041,
"learning_rate": 9.600000000000001e-06,
"loss": 2.2393,
"step": 4
},
{
"epoch": 0.017543859649122806,
"grad_norm": 0.10541164129972458,
"learning_rate": 1.2e-05,
"loss": 2.0981,
"step": 5
},
{
"epoch": 0.021052631578947368,
"grad_norm": 0.09022955596446991,
"learning_rate": 1.44e-05,
"loss": 2.1626,
"step": 6
},
{
"epoch": 0.02456140350877193,
"grad_norm": 0.0826038271188736,
"learning_rate": 1.6800000000000002e-05,
"loss": 2.0838,
"step": 7
},
{
"epoch": 0.028070175438596492,
"grad_norm": 0.09345665574073792,
"learning_rate": 1.9200000000000003e-05,
"loss": 2.1717,
"step": 8
},
{
"epoch": 0.031578947368421054,
"grad_norm": 0.12961702048778534,
"learning_rate": 2.16e-05,
"loss": 2.0865,
"step": 9
},
{
"epoch": 0.03508771929824561,
"grad_norm": 0.11363445222377777,
"learning_rate": 2.4e-05,
"loss": 2.3286,
"step": 10
},
{
"epoch": 0.03859649122807018,
"grad_norm": 0.10748554766178131,
"learning_rate": 2.64e-05,
"loss": 2.1253,
"step": 11
},
{
"epoch": 0.042105263157894736,
"grad_norm": 0.16965074837207794,
"learning_rate": 2.88e-05,
"loss": 1.953,
"step": 12
},
{
"epoch": 0.0456140350877193,
"grad_norm": 0.10734588652849197,
"learning_rate": 3.12e-05,
"loss": 2.3053,
"step": 13
},
{
"epoch": 0.04912280701754386,
"grad_norm": 0.09110312163829803,
"learning_rate": 3.3600000000000004e-05,
"loss": 2.3143,
"step": 14
},
{
"epoch": 0.05263157894736842,
"grad_norm": 0.09969355165958405,
"learning_rate": 3.6e-05,
"loss": 2.1258,
"step": 15
},
{
"epoch": 0.056140350877192984,
"grad_norm": 0.11016582697629929,
"learning_rate": 3.8400000000000005e-05,
"loss": 2.0992,
"step": 16
},
{
"epoch": 0.05964912280701754,
"grad_norm": 0.10026279091835022,
"learning_rate": 4.08e-05,
"loss": 2.0588,
"step": 17
},
{
"epoch": 0.06315789473684211,
"grad_norm": 0.09350590407848358,
"learning_rate": 4.32e-05,
"loss": 2.1193,
"step": 18
},
{
"epoch": 0.06666666666666667,
"grad_norm": 0.08562514185905457,
"learning_rate": 4.5600000000000004e-05,
"loss": 2.0707,
"step": 19
},
{
"epoch": 0.07017543859649122,
"grad_norm": 0.0910220816731453,
"learning_rate": 4.8e-05,
"loss": 1.9285,
"step": 20
},
{
"epoch": 0.07368421052631578,
"grad_norm": 0.12792207300662994,
"learning_rate": 5.04e-05,
"loss": 2.1708,
"step": 21
},
{
"epoch": 0.07719298245614035,
"grad_norm": 0.09217620640993118,
"learning_rate": 5.28e-05,
"loss": 2.1628,
"step": 22
},
{
"epoch": 0.08070175438596491,
"grad_norm": 0.09265049546957016,
"learning_rate": 5.520000000000001e-05,
"loss": 2.0818,
"step": 23
},
{
"epoch": 0.08421052631578947,
"grad_norm": 0.08625225722789764,
"learning_rate": 5.76e-05,
"loss": 2.2644,
"step": 24
},
{
"epoch": 0.08771929824561403,
"grad_norm": 0.09073711931705475,
"learning_rate": 6e-05,
"loss": 2.1186,
"step": 25
},
{
"epoch": 0.0912280701754386,
"grad_norm": 0.08659897744655609,
"learning_rate": 5.999781002566832e-05,
"loss": 2.0872,
"step": 26
},
{
"epoch": 0.09473684210526316,
"grad_norm": 0.07886387407779694,
"learning_rate": 5.9991240422405795e-05,
"loss": 2.0594,
"step": 27
},
{
"epoch": 0.09824561403508772,
"grad_norm": 0.0848180428147316,
"learning_rate": 5.998029214936324e-05,
"loss": 2.0301,
"step": 28
},
{
"epoch": 0.10175438596491228,
"grad_norm": 0.08976238965988159,
"learning_rate": 5.99649668049698e-05,
"loss": 2.0395,
"step": 29
},
{
"epoch": 0.10526315789473684,
"grad_norm": 0.08499772846698761,
"learning_rate": 5.994526662669953e-05,
"loss": 2.0563,
"step": 30
},
{
"epoch": 0.10877192982456141,
"grad_norm": 0.08985580503940582,
"learning_rate": 5.992119449074474e-05,
"loss": 2.1382,
"step": 31
},
{
"epoch": 0.11228070175438597,
"grad_norm": 0.08376096189022064,
"learning_rate": 5.9892753911596084e-05,
"loss": 2.2707,
"step": 32
},
{
"epoch": 0.11578947368421053,
"grad_norm": 0.08586399257183075,
"learning_rate": 5.985994904152946e-05,
"loss": 2.0834,
"step": 33
},
{
"epoch": 0.11929824561403508,
"grad_norm": 0.08400382101535797,
"learning_rate": 5.982278466999976e-05,
"loss": 2.0405,
"step": 34
},
{
"epoch": 0.12280701754385964,
"grad_norm": 0.08553475141525269,
"learning_rate": 5.9781266222941616e-05,
"loss": 2.2521,
"step": 35
},
{
"epoch": 0.12631578947368421,
"grad_norm": 0.08234255760908127,
"learning_rate": 5.973539976197728e-05,
"loss": 2.0814,
"step": 36
},
{
"epoch": 0.12982456140350876,
"grad_norm": 0.09046807885169983,
"learning_rate": 5.968519198353153e-05,
"loss": 2.1075,
"step": 37
},
{
"epoch": 0.13333333333333333,
"grad_norm": 0.09342332929372787,
"learning_rate": 5.963065021785414e-05,
"loss": 2.094,
"step": 38
},
{
"epoch": 0.1368421052631579,
"grad_norm": 0.09752500057220459,
"learning_rate": 5.957178242794953e-05,
"loss": 2.008,
"step": 39
},
{
"epoch": 0.14035087719298245,
"grad_norm": 0.08942960947751999,
"learning_rate": 5.950859720841431e-05,
"loss": 2.203,
"step": 40
},
{
"epoch": 0.14385964912280702,
"grad_norm": 0.0905979797244072,
"learning_rate": 5.94411037841824e-05,
"loss": 2.3466,
"step": 41
},
{
"epoch": 0.14736842105263157,
"grad_norm": 0.09027687460184097,
"learning_rate": 5.936931200917826e-05,
"loss": 1.9638,
"step": 42
},
{
"epoch": 0.15087719298245614,
"grad_norm": 0.09753058850765228,
"learning_rate": 5.929323236487816e-05,
"loss": 2.1403,
"step": 43
},
{
"epoch": 0.1543859649122807,
"grad_norm": 0.09758615493774414,
"learning_rate": 5.921287595878001e-05,
"loss": 2.1398,
"step": 44
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.09411653131246567,
"learning_rate": 5.912825452278156e-05,
"loss": 2.151,
"step": 45
},
{
"epoch": 0.16140350877192983,
"grad_norm": 0.10731625556945801,
"learning_rate": 5.903938041146769e-05,
"loss": 2.0521,
"step": 46
},
{
"epoch": 0.1649122807017544,
"grad_norm": 0.09779181331396103,
"learning_rate": 5.894626660030655e-05,
"loss": 2.1397,
"step": 47
},
{
"epoch": 0.16842105263157894,
"grad_norm": 0.09776318818330765,
"learning_rate": 5.8848926683755234e-05,
"loss": 2.1026,
"step": 48
},
{
"epoch": 0.17192982456140352,
"grad_norm": 0.09971348941326141,
"learning_rate": 5.8747374873274985e-05,
"loss": 2.2858,
"step": 49
},
{
"epoch": 0.17543859649122806,
"grad_norm": 0.10052883625030518,
"learning_rate": 5.864162599525637e-05,
"loss": 2.2403,
"step": 50
},
{
"epoch": 0.17894736842105263,
"grad_norm": 0.09604485332965851,
"learning_rate": 5.853169548885461e-05,
"loss": 2.015,
"step": 51
},
{
"epoch": 0.1824561403508772,
"grad_norm": 0.0971733033657074,
"learning_rate": 5.841759940373552e-05,
"loss": 2.0381,
"step": 52
},
{
"epoch": 0.18596491228070175,
"grad_norm": 0.11067637801170349,
"learning_rate": 5.829935439773229e-05,
"loss": 2.1373,
"step": 53
},
{
"epoch": 0.18947368421052632,
"grad_norm": 0.09739149361848831,
"learning_rate": 5.817697773441346e-05,
"loss": 2.2659,
"step": 54
},
{
"epoch": 0.19298245614035087,
"grad_norm": 0.09938070923089981,
"learning_rate": 5.805048728056245e-05,
"loss": 2.0896,
"step": 55
},
{
"epoch": 0.19649122807017544,
"grad_norm": 0.11174876987934113,
"learning_rate": 5.7919901503569066e-05,
"loss": 2.1314,
"step": 56
},
{
"epoch": 0.2,
"grad_norm": 0.10702885687351227,
"learning_rate": 5.77852394687333e-05,
"loss": 1.9618,
"step": 57
},
{
"epoch": 0.20350877192982456,
"grad_norm": 0.10180342197418213,
"learning_rate": 5.764652083648179e-05,
"loss": 2.0258,
"step": 58
},
{
"epoch": 0.20701754385964913,
"grad_norm": 0.10502015054225922,
"learning_rate": 5.750376585949748e-05,
"loss": 2.2146,
"step": 59
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.11613918840885162,
"learning_rate": 5.735699537976271e-05,
"loss": 1.9482,
"step": 60
},
{
"epoch": 0.21403508771929824,
"grad_norm": 0.10688231140375137,
"learning_rate": 5.720623082551636e-05,
"loss": 1.9497,
"step": 61
},
{
"epoch": 0.21754385964912282,
"grad_norm": 0.1024007648229599,
"learning_rate": 5.7051494208125385e-05,
"loss": 2.132,
"step": 62
},
{
"epoch": 0.22105263157894736,
"grad_norm": 0.10295075923204422,
"learning_rate": 5.6892808118871113e-05,
"loss": 2.1412,
"step": 63
},
{
"epoch": 0.22456140350877193,
"grad_norm": 0.11542188376188278,
"learning_rate": 5.673019572565103e-05,
"loss": 2.2178,
"step": 64
},
{
"epoch": 0.22807017543859648,
"grad_norm": 0.11483608931303024,
"learning_rate": 5.6563680769596294e-05,
"loss": 2.0975,
"step": 65
},
{
"epoch": 0.23157894736842105,
"grad_norm": 0.1054820865392685,
"learning_rate": 5.6393287561605536e-05,
"loss": 2.1284,
"step": 66
},
{
"epoch": 0.23508771929824562,
"grad_norm": 0.10530298203229904,
"learning_rate": 5.6219040978795544e-05,
"loss": 2.0712,
"step": 67
},
{
"epoch": 0.23859649122807017,
"grad_norm": 0.1356276124715805,
"learning_rate": 5.6040966460869225e-05,
"loss": 2.1189,
"step": 68
},
{
"epoch": 0.24210526315789474,
"grad_norm": 0.10891308635473251,
"learning_rate": 5.5859090006401476e-05,
"loss": 2.0887,
"step": 69
},
{
"epoch": 0.24561403508771928,
"grad_norm": 0.1175449788570404,
"learning_rate": 5.567343816904343e-05,
"loss": 2.0421,
"step": 70
},
{
"epoch": 0.24912280701754386,
"grad_norm": 0.10129985213279724,
"learning_rate": 5.5484038053645636e-05,
"loss": 2.1027,
"step": 71
},
{
"epoch": 0.25263157894736843,
"grad_norm": 0.10736104100942612,
"learning_rate": 5.529091731230084e-05,
"loss": 2.1332,
"step": 72
},
{
"epoch": 0.256140350877193,
"grad_norm": 0.11475091427564621,
"learning_rate": 5.50941041403068e-05,
"loss": 2.1401,
"step": 73
},
{
"epoch": 0.2596491228070175,
"grad_norm": 0.10455591231584549,
"learning_rate": 5.489362727204985e-05,
"loss": 2.1409,
"step": 74
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.09795428812503815,
"learning_rate": 5.468951597680969e-05,
"loss": 2.2061,
"step": 75
},
{
"epoch": 0.26666666666666666,
"grad_norm": 0.11418502777814865,
"learning_rate": 5.4481800054486144e-05,
"loss": 2.2608,
"step": 76
},
{
"epoch": 0.27017543859649124,
"grad_norm": 0.11368659883737564,
"learning_rate": 5.427050983124843e-05,
"loss": 2.1459,
"step": 77
},
{
"epoch": 0.2736842105263158,
"grad_norm": 0.13392335176467896,
"learning_rate": 5.4055676155107554e-05,
"loss": 2.2393,
"step": 78
},
{
"epoch": 0.2771929824561403,
"grad_norm": 0.10770577192306519,
"learning_rate": 5.3837330391412626e-05,
"loss": 1.9827,
"step": 79
},
{
"epoch": 0.2807017543859649,
"grad_norm": 0.10940691083669662,
"learning_rate": 5.36155044182715e-05,
"loss": 1.9657,
"step": 80
},
{
"epoch": 0.28421052631578947,
"grad_norm": 0.1266562044620514,
"learning_rate": 5.339023062189667e-05,
"loss": 2.0797,
"step": 81
},
{
"epoch": 0.28771929824561404,
"grad_norm": 0.11216572672128677,
"learning_rate": 5.3161541891876906e-05,
"loss": 2.0198,
"step": 82
},
{
"epoch": 0.2912280701754386,
"grad_norm": 0.11424341052770615,
"learning_rate": 5.292947161637545e-05,
"loss": 2.0178,
"step": 83
},
{
"epoch": 0.29473684210526313,
"grad_norm": 0.10189955681562424,
"learning_rate": 5.2694053677255404e-05,
"loss": 2.1625,
"step": 84
},
{
"epoch": 0.2982456140350877,
"grad_norm": 0.12291618436574936,
"learning_rate": 5.245532244513303e-05,
"loss": 2.0019,
"step": 85
},
{
"epoch": 0.3017543859649123,
"grad_norm": 0.11346428096294403,
"learning_rate": 5.2213312774359706e-05,
"loss": 2.1314,
"step": 86
},
{
"epoch": 0.30526315789473685,
"grad_norm": 0.11410778015851974,
"learning_rate": 5.196805999793321e-05,
"loss": 2.1738,
"step": 87
},
{
"epoch": 0.3087719298245614,
"grad_norm": 0.11488968133926392,
"learning_rate": 5.1719599922339234e-05,
"loss": 2.0488,
"step": 88
},
{
"epoch": 0.312280701754386,
"grad_norm": 0.11716273427009583,
"learning_rate": 5.1467968822323636e-05,
"loss": 2.2643,
"step": 89
},
{
"epoch": 0.3157894736842105,
"grad_norm": 0.12693484127521515,
"learning_rate": 5.1213203435596425e-05,
"loss": 1.9796,
"step": 90
},
{
"epoch": 0.3192982456140351,
"grad_norm": 0.12479843944311142,
"learning_rate": 5.095534095746812e-05,
"loss": 2.1207,
"step": 91
},
{
"epoch": 0.32280701754385965,
"grad_norm": 0.1232728436589241,
"learning_rate": 5.069441903541923e-05,
"loss": 2.0333,
"step": 92
},
{
"epoch": 0.3263157894736842,
"grad_norm": 0.12169407308101654,
"learning_rate": 5.043047576360392e-05,
"loss": 2.063,
"step": 93
},
{
"epoch": 0.3298245614035088,
"grad_norm": 0.12006253749132156,
"learning_rate": 5.016354967728818e-05,
"loss": 2.0197,
"step": 94
},
{
"epoch": 0.3333333333333333,
"grad_norm": 0.12108758091926575,
"learning_rate": 4.989367974722386e-05,
"loss": 1.9304,
"step": 95
},
{
"epoch": 0.3368421052631579,
"grad_norm": 0.12677016854286194,
"learning_rate": 4.962090537395894e-05,
"loss": 2.1133,
"step": 96
},
{
"epoch": 0.34035087719298246,
"grad_norm": 0.13631866872310638,
"learning_rate": 4.934526638208512e-05,
"loss": 2.0936,
"step": 97
},
{
"epoch": 0.34385964912280703,
"grad_norm": 0.13509684801101685,
"learning_rate": 4.9066803014423566e-05,
"loss": 2.0378,
"step": 98
},
{
"epoch": 0.3473684210526316,
"grad_norm": 0.10955115407705307,
"learning_rate": 4.8785555926149415e-05,
"loss": 1.9979,
"step": 99
},
{
"epoch": 0.3508771929824561,
"grad_norm": 0.10989337414503098,
"learning_rate": 4.85015661788563e-05,
"loss": 2.0235,
"step": 100
},
{
"epoch": 0.3543859649122807,
"grad_norm": 0.10870235413312912,
"learning_rate": 4.821487523456134e-05,
"loss": 2.2916,
"step": 101
},
{
"epoch": 0.35789473684210527,
"grad_norm": 0.11895206570625305,
"learning_rate": 4.792552494965182e-05,
"loss": 2.0428,
"step": 102
},
{
"epoch": 0.36140350877192984,
"grad_norm": 0.1315065175294876,
"learning_rate": 4.7633557568774194e-05,
"loss": 2.2134,
"step": 103
},
{
"epoch": 0.3649122807017544,
"grad_norm": 0.12591607868671417,
"learning_rate": 4.733901571866645e-05,
"loss": 2.2044,
"step": 104
},
{
"epoch": 0.3684210526315789,
"grad_norm": 0.11799575388431549,
"learning_rate": 4.704194240193467e-05,
"loss": 2.0121,
"step": 105
},
{
"epoch": 0.3719298245614035,
"grad_norm": 0.12055860459804535,
"learning_rate": 4.674238099077475e-05,
"loss": 2.2142,
"step": 106
},
{
"epoch": 0.37543859649122807,
"grad_norm": 0.13334396481513977,
"learning_rate": 4.644037522064011e-05,
"loss": 1.9705,
"step": 107
},
{
"epoch": 0.37894736842105264,
"grad_norm": 0.1212170422077179,
"learning_rate": 4.613596918385636e-05,
"loss": 1.9693,
"step": 108
},
{
"epoch": 0.3824561403508772,
"grad_norm": 0.13068757951259613,
"learning_rate": 4.582920732318398e-05,
"loss": 2.0752,
"step": 109
},
{
"epoch": 0.38596491228070173,
"grad_norm": 0.12562724947929382,
"learning_rate": 4.5520134425329705e-05,
"loss": 1.9943,
"step": 110
},
{
"epoch": 0.3894736842105263,
"grad_norm": 0.12634319067001343,
"learning_rate": 4.5208795614407715e-05,
"loss": 2.1028,
"step": 111
},
{
"epoch": 0.3929824561403509,
"grad_norm": 0.11350910365581512,
"learning_rate": 4.4895236345351645e-05,
"loss": 2.2079,
"step": 112
},
{
"epoch": 0.39649122807017545,
"grad_norm": 0.11814025789499283,
"learning_rate": 4.457950239727818e-05,
"loss": 2.1619,
"step": 113
},
{
"epoch": 0.4,
"grad_norm": 0.1307692676782608,
"learning_rate": 4.4261639866803486e-05,
"loss": 2.0255,
"step": 114
},
{
"epoch": 0.40350877192982454,
"grad_norm": 0.12219037115573883,
"learning_rate": 4.394169516131306e-05,
"loss": 2.0005,
"step": 115
},
{
"epoch": 0.4070175438596491,
"grad_norm": 0.11336469650268555,
"learning_rate": 4.3619714992186405e-05,
"loss": 2.1032,
"step": 116
},
{
"epoch": 0.4105263157894737,
"grad_norm": 0.12509691715240479,
"learning_rate": 4.329574636797725e-05,
"loss": 2.1845,
"step": 117
},
{
"epoch": 0.41403508771929826,
"grad_norm": 0.1294390708208084,
"learning_rate": 4.296983658755032e-05,
"loss": 2.16,
"step": 118
},
{
"epoch": 0.41754385964912283,
"grad_norm": 0.11944102495908737,
"learning_rate": 4.264203323317588e-05,
"loss": 2.1965,
"step": 119
},
{
"epoch": 0.42105263157894735,
"grad_norm": 0.11656481772661209,
"learning_rate": 4.231238416358271e-05,
"loss": 2.2442,
"step": 120
},
{
"epoch": 0.4245614035087719,
"grad_norm": 0.11671917885541916,
"learning_rate": 4.198093750697086e-05,
"loss": 1.9754,
"step": 121
},
{
"epoch": 0.4280701754385965,
"grad_norm": 0.1291894018650055,
"learning_rate": 4.164774165398505e-05,
"loss": 1.9795,
"step": 122
},
{
"epoch": 0.43157894736842106,
"grad_norm": 0.11259818822145462,
"learning_rate": 4.131284525064961e-05,
"loss": 2.1273,
"step": 123
},
{
"epoch": 0.43508771929824563,
"grad_norm": 0.11232686042785645,
"learning_rate": 4.0976297191266376e-05,
"loss": 2.3643,
"step": 124
},
{
"epoch": 0.43859649122807015,
"grad_norm": 0.12415839731693268,
"learning_rate": 4.063814661127607e-05,
"loss": 2.2045,
"step": 125
},
{
"epoch": 0.4421052631578947,
"grad_norm": 0.13372597098350525,
"learning_rate": 4.029844288008475e-05,
"loss": 2.0377,
"step": 126
},
{
"epoch": 0.4456140350877193,
"grad_norm": 0.1277344524860382,
"learning_rate": 3.9957235593855856e-05,
"loss": 1.9898,
"step": 127
},
{
"epoch": 0.44912280701754387,
"grad_norm": 0.17750953137874603,
"learning_rate": 3.961457456826929e-05,
"loss": 1.8503,
"step": 128
},
{
"epoch": 0.45263157894736844,
"grad_norm": 0.12306816130876541,
"learning_rate": 3.927050983124842e-05,
"loss": 2.1545,
"step": 129
},
{
"epoch": 0.45614035087719296,
"grad_norm": 0.12556982040405273,
"learning_rate": 3.892509161565609e-05,
"loss": 2.0301,
"step": 130
},
{
"epoch": 0.45964912280701753,
"grad_norm": 0.11492066085338593,
"learning_rate": 3.8578370351960676e-05,
"loss": 2.0265,
"step": 131
},
{
"epoch": 0.4631578947368421,
"grad_norm": 0.10817985981702805,
"learning_rate": 3.8230396660873376e-05,
"loss": 2.3565,
"step": 132
},
{
"epoch": 0.4666666666666667,
"grad_norm": 0.13673719763755798,
"learning_rate": 3.788122134595761e-05,
"loss": 2.1245,
"step": 133
},
{
"epoch": 0.47017543859649125,
"grad_norm": 0.14079900085926056,
"learning_rate": 3.753089538621185e-05,
"loss": 2.0535,
"step": 134
},
{
"epoch": 0.47368421052631576,
"grad_norm": 0.11784680932760239,
"learning_rate": 3.717946992862674e-05,
"loss": 2.034,
"step": 135
},
{
"epoch": 0.47719298245614034,
"grad_norm": 0.12113582342863083,
"learning_rate": 3.68269962807177e-05,
"loss": 2.0731,
"step": 136
},
{
"epoch": 0.4807017543859649,
"grad_norm": 0.13436686992645264,
"learning_rate": 3.647352590303419e-05,
"loss": 2.1128,
"step": 137
},
{
"epoch": 0.4842105263157895,
"grad_norm": 0.13415394723415375,
"learning_rate": 3.611911040164647e-05,
"loss": 1.8647,
"step": 138
},
{
"epoch": 0.48771929824561405,
"grad_norm": 0.12898415327072144,
"learning_rate": 3.576380152061125e-05,
"loss": 1.9798,
"step": 139
},
{
"epoch": 0.49122807017543857,
"grad_norm": 0.11745478957891464,
"learning_rate": 3.540765113441718e-05,
"loss": 2.1884,
"step": 140
},
{
"epoch": 0.49473684210526314,
"grad_norm": 0.11902904510498047,
"learning_rate": 3.505071124041115e-05,
"loss": 2.2156,
"step": 141
},
{
"epoch": 0.4982456140350877,
"grad_norm": 0.13850358128547668,
"learning_rate": 3.469303395120693e-05,
"loss": 1.8751,
"step": 142
},
{
"epoch": 0.5017543859649123,
"grad_norm": 0.12531642615795135,
"learning_rate": 3.4334671487076636e-05,
"loss": 2.1471,
"step": 143
},
{
"epoch": 0.5052631578947369,
"grad_norm": 0.12197932600975037,
"learning_rate": 3.3975676168326834e-05,
"loss": 1.9089,
"step": 144
},
{
"epoch": 0.5087719298245614,
"grad_norm": 0.11156132072210312,
"learning_rate": 3.361610040765969e-05,
"loss": 2.0513,
"step": 145
},
{
"epoch": 0.512280701754386,
"grad_norm": 0.1183839663863182,
"learning_rate": 3.325599670252099e-05,
"loss": 2.1052,
"step": 146
},
{
"epoch": 0.5157894736842106,
"grad_norm": 0.12653163075447083,
"learning_rate": 3.289541762743546e-05,
"loss": 2.0598,
"step": 147
},
{
"epoch": 0.519298245614035,
"grad_norm": 0.1317450851202011,
"learning_rate": 3.2534415826331014e-05,
"loss": 2.082,
"step": 148
},
{
"epoch": 0.5228070175438596,
"grad_norm": 0.12523813545703888,
"learning_rate": 3.2173044004852874e-05,
"loss": 2.0974,
"step": 149
},
{
"epoch": 0.5263157894736842,
"grad_norm": 0.12494626641273499,
"learning_rate": 3.1811354922668584e-05,
"loss": 2.3148,
"step": 150
},
{
"epoch": 0.5298245614035088,
"grad_norm": 0.11819988489151001,
"learning_rate": 3.144940138576521e-05,
"loss": 2.0978,
"step": 151
},
{
"epoch": 0.5333333333333333,
"grad_norm": 0.12493463605642319,
"learning_rate": 3.108723623873977e-05,
"loss": 2.1228,
"step": 152
},
{
"epoch": 0.5368421052631579,
"grad_norm": 0.11485088616609573,
"learning_rate": 3.072491235708397e-05,
"loss": 2.2135,
"step": 153
},
{
"epoch": 0.5403508771929825,
"grad_norm": 0.12239831686019897,
"learning_rate": 3.0362482639464522e-05,
"loss": 2.076,
"step": 154
},
{
"epoch": 0.543859649122807,
"grad_norm": 0.12548470497131348,
"learning_rate": 3e-05,
"loss": 2.062,
"step": 155
},
{
"epoch": 0.5473684210526316,
"grad_norm": 0.12589600682258606,
"learning_rate": 2.963751736053548e-05,
"loss": 2.0006,
"step": 156
},
{
"epoch": 0.5508771929824562,
"grad_norm": 0.13063469529151917,
"learning_rate": 2.927508764291604e-05,
"loss": 1.8844,
"step": 157
},
{
"epoch": 0.5543859649122806,
"grad_norm": 0.12271241843700409,
"learning_rate": 2.891276376126024e-05,
"loss": 2.0471,
"step": 158
},
{
"epoch": 0.5578947368421052,
"grad_norm": 0.124054454267025,
"learning_rate": 2.8550598614234793e-05,
"loss": 2.0801,
"step": 159
},
{
"epoch": 0.5614035087719298,
"grad_norm": 0.12535397708415985,
"learning_rate": 2.8188645077331414e-05,
"loss": 2.0612,
"step": 160
},
{
"epoch": 0.5649122807017544,
"grad_norm": 0.12145686894655228,
"learning_rate": 2.782695599514713e-05,
"loss": 2.1516,
"step": 161
},
{
"epoch": 0.5684210526315789,
"grad_norm": 0.14351366460323334,
"learning_rate": 2.746558417366899e-05,
"loss": 1.9728,
"step": 162
},
{
"epoch": 0.5719298245614035,
"grad_norm": 0.1383461058139801,
"learning_rate": 2.7104582372564547e-05,
"loss": 2.2342,
"step": 163
},
{
"epoch": 0.5754385964912281,
"grad_norm": 0.12213313579559326,
"learning_rate": 2.6744003297479016e-05,
"loss": 2.0961,
"step": 164
},
{
"epoch": 0.5789473684210527,
"grad_norm": 0.13188298046588898,
"learning_rate": 2.6383899592340312e-05,
"loss": 1.9805,
"step": 165
},
{
"epoch": 0.5824561403508772,
"grad_norm": 0.12241267412900925,
"learning_rate": 2.602432383167318e-05,
"loss": 2.1336,
"step": 166
},
{
"epoch": 0.5859649122807018,
"grad_norm": 0.1236441507935524,
"learning_rate": 2.566532851292337e-05,
"loss": 2.1338,
"step": 167
},
{
"epoch": 0.5894736842105263,
"grad_norm": 0.13339705765247345,
"learning_rate": 2.530696604879307e-05,
"loss": 1.9877,
"step": 168
},
{
"epoch": 0.5929824561403508,
"grad_norm": 0.12486617267131805,
"learning_rate": 2.4949288759588843e-05,
"loss": 2.1196,
"step": 169
},
{
"epoch": 0.5964912280701754,
"grad_norm": 0.12080895155668259,
"learning_rate": 2.459234886558283e-05,
"loss": 2.1327,
"step": 170
},
{
"epoch": 0.6,
"grad_norm": 0.12551304697990417,
"learning_rate": 2.4236198479388747e-05,
"loss": 1.9393,
"step": 171
},
{
"epoch": 0.6035087719298246,
"grad_norm": 0.12770839035511017,
"learning_rate": 2.3880889598353544e-05,
"loss": 2.1024,
"step": 172
},
{
"epoch": 0.6070175438596491,
"grad_norm": 0.1613360196352005,
"learning_rate": 2.352647409696582e-05,
"loss": 1.8259,
"step": 173
},
{
"epoch": 0.6105263157894737,
"grad_norm": 0.12359873205423355,
"learning_rate": 2.31730037192823e-05,
"loss": 2.0849,
"step": 174
},
{
"epoch": 0.6140350877192983,
"grad_norm": 0.124616339802742,
"learning_rate": 2.2820530071373278e-05,
"loss": 2.2074,
"step": 175
},
{
"epoch": 0.6175438596491228,
"grad_norm": 0.1181364506483078,
"learning_rate": 2.246910461378815e-05,
"loss": 2.1069,
"step": 176
},
{
"epoch": 0.6210526315789474,
"grad_norm": 0.12338215857744217,
"learning_rate": 2.2118778654042395e-05,
"loss": 2.1848,
"step": 177
},
{
"epoch": 0.624561403508772,
"grad_norm": 0.1383160799741745,
"learning_rate": 2.1769603339126626e-05,
"loss": 2.0918,
"step": 178
},
{
"epoch": 0.6280701754385964,
"grad_norm": 0.12015274912118912,
"learning_rate": 2.142162964803933e-05,
"loss": 2.2665,
"step": 179
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.14904998242855072,
"learning_rate": 2.1074908384343916e-05,
"loss": 1.9662,
"step": 180
},
{
"epoch": 0.6350877192982456,
"grad_norm": 0.1324617713689804,
"learning_rate": 2.072949016875158e-05,
"loss": 2.1624,
"step": 181
},
{
"epoch": 0.6385964912280702,
"grad_norm": 0.13385087251663208,
"learning_rate": 2.0385425431730716e-05,
"loss": 2.0126,
"step": 182
},
{
"epoch": 0.6421052631578947,
"grad_norm": 0.13096769154071808,
"learning_rate": 2.0042764406144155e-05,
"loss": 2.0768,
"step": 183
},
{
"epoch": 0.6456140350877193,
"grad_norm": 0.12568296492099762,
"learning_rate": 1.970155711991525e-05,
"loss": 2.216,
"step": 184
},
{
"epoch": 0.6491228070175439,
"grad_norm": 0.14268100261688232,
"learning_rate": 1.936185338872393e-05,
"loss": 2.1907,
"step": 185
},
{
"epoch": 0.6526315789473685,
"grad_norm": 0.14268723130226135,
"learning_rate": 1.9023702808733635e-05,
"loss": 2.0716,
"step": 186
},
{
"epoch": 0.656140350877193,
"grad_norm": 0.15304097533226013,
"learning_rate": 1.8687154749350388e-05,
"loss": 1.8692,
"step": 187
},
{
"epoch": 0.6596491228070176,
"grad_norm": 0.17945235967636108,
"learning_rate": 1.8352258346014953e-05,
"loss": 1.7839,
"step": 188
},
{
"epoch": 0.6631578947368421,
"grad_norm": 0.13728468120098114,
"learning_rate": 1.8019062493029138e-05,
"loss": 2.0884,
"step": 189
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.1320762038230896,
"learning_rate": 1.76876158364173e-05,
"loss": 1.8989,
"step": 190
},
{
"epoch": 0.6701754385964912,
"grad_norm": 0.1560417264699936,
"learning_rate": 1.7357966766824134e-05,
"loss": 1.9468,
"step": 191
},
{
"epoch": 0.6736842105263158,
"grad_norm": 0.14444367587566376,
"learning_rate": 1.7030163412449674e-05,
"loss": 1.9449,
"step": 192
},
{
"epoch": 0.6771929824561403,
"grad_norm": 0.20545677840709686,
"learning_rate": 1.6704253632022765e-05,
"loss": 1.89,
"step": 193
},
{
"epoch": 0.6807017543859649,
"grad_norm": 0.1266920417547226,
"learning_rate": 1.6380285007813596e-05,
"loss": 2.2582,
"step": 194
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.13746094703674316,
"learning_rate": 1.6058304838686945e-05,
"loss": 2.1991,
"step": 195
},
{
"epoch": 0.6877192982456141,
"grad_norm": 0.12731488049030304,
"learning_rate": 1.5738360133196516e-05,
"loss": 2.0734,
"step": 196
},
{
"epoch": 0.6912280701754386,
"grad_norm": 0.12864916026592255,
"learning_rate": 1.542049760272182e-05,
"loss": 2.1831,
"step": 197
},
{
"epoch": 0.6947368421052632,
"grad_norm": 0.13020969927310944,
"learning_rate": 1.5104763654648375e-05,
"loss": 2.0344,
"step": 198
},
{
"epoch": 0.6982456140350877,
"grad_norm": 0.11206473410129547,
"learning_rate": 1.4791204385592297e-05,
"loss": 2.3081,
"step": 199
},
{
"epoch": 0.7017543859649122,
"grad_norm": 0.1320793628692627,
"learning_rate": 1.4479865574670293e-05,
"loss": 2.052,
"step": 200
},
{
"epoch": 0.7052631578947368,
"grad_norm": 0.14232482016086578,
"learning_rate": 1.4170792676816015e-05,
"loss": 1.9167,
"step": 201
},
{
"epoch": 0.7087719298245614,
"grad_norm": 0.1702522486448288,
"learning_rate": 1.3864030816143643e-05,
"loss": 1.9151,
"step": 202
},
{
"epoch": 0.712280701754386,
"grad_norm": 0.13469327986240387,
"learning_rate": 1.3559624779359908e-05,
"loss": 1.9203,
"step": 203
},
{
"epoch": 0.7157894736842105,
"grad_norm": 0.13646726310253143,
"learning_rate": 1.3257619009225253e-05,
"loss": 2.2265,
"step": 204
},
{
"epoch": 0.7192982456140351,
"grad_norm": 0.12395426630973816,
"learning_rate": 1.2958057598065328e-05,
"loss": 2.0885,
"step": 205
},
{
"epoch": 0.7228070175438597,
"grad_norm": 0.12457320839166641,
"learning_rate": 1.2660984281333553e-05,
"loss": 2.1296,
"step": 206
},
{
"epoch": 0.7263157894736842,
"grad_norm": 0.12583313882350922,
"learning_rate": 1.2366442431225809e-05,
"loss": 2.2298,
"step": 207
},
{
"epoch": 0.7298245614035088,
"grad_norm": 0.13265646994113922,
"learning_rate": 1.2074475050348181e-05,
"loss": 2.0604,
"step": 208
},
{
"epoch": 0.7333333333333333,
"grad_norm": 0.13001544773578644,
"learning_rate": 1.1785124765438666e-05,
"loss": 1.66,
"step": 209
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.12221955507993698,
"learning_rate": 1.1498433821143707e-05,
"loss": 2.337,
"step": 210
},
{
"epoch": 0.7403508771929824,
"grad_norm": 0.1360742449760437,
"learning_rate": 1.121444407385059e-05,
"loss": 1.9907,
"step": 211
},
{
"epoch": 0.743859649122807,
"grad_norm": 0.13305602967739105,
"learning_rate": 1.0933196985576443e-05,
"loss": 2.0058,
"step": 212
},
{
"epoch": 0.7473684210526316,
"grad_norm": 0.14633433520793915,
"learning_rate": 1.0654733617914889e-05,
"loss": 2.1349,
"step": 213
},
{
"epoch": 0.7508771929824561,
"grad_norm": 0.1337619572877884,
"learning_rate": 1.0379094626041076e-05,
"loss": 2.1434,
"step": 214
},
{
"epoch": 0.7543859649122807,
"grad_norm": 0.12340360134840012,
"learning_rate": 1.0106320252776151e-05,
"loss": 2.1285,
"step": 215
},
{
"epoch": 0.7578947368421053,
"grad_norm": 0.133439838886261,
"learning_rate": 9.836450322711816e-06,
"loss": 2.2396,
"step": 216
},
{
"epoch": 0.7614035087719299,
"grad_norm": 0.11517748236656189,
"learning_rate": 9.569524236396082e-06,
"loss": 2.2786,
"step": 217
},
{
"epoch": 0.7649122807017544,
"grad_norm": 0.12936294078826904,
"learning_rate": 9.305580964580771e-06,
"loss": 1.9806,
"step": 218
},
{
"epoch": 0.7684210526315789,
"grad_norm": 0.12179510295391083,
"learning_rate": 9.044659042531894e-06,
"loss": 2.3314,
"step": 219
},
{
"epoch": 0.7719298245614035,
"grad_norm": 0.13949505984783173,
"learning_rate": 8.786796564403577e-06,
"loss": 2.0178,
"step": 220
},
{
"epoch": 0.775438596491228,
"grad_norm": 0.14202281832695007,
"learning_rate": 8.53203117767637e-06,
"loss": 2.0716,
"step": 221
},
{
"epoch": 0.7789473684210526,
"grad_norm": 0.1363075077533722,
"learning_rate": 8.280400077660774e-06,
"loss": 1.9615,
"step": 222
},
{
"epoch": 0.7824561403508772,
"grad_norm": 0.14736121892929077,
"learning_rate": 8.031940002066804e-06,
"loss": 1.8541,
"step": 223
},
{
"epoch": 0.7859649122807018,
"grad_norm": 0.1318388134241104,
"learning_rate": 7.78668722564031e-06,
"loss": 2.2338,
"step": 224
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.13814999163150787,
"learning_rate": 7.544677554866964e-06,
"loss": 2.0775,
"step": 225
},
{
"epoch": 0.7929824561403509,
"grad_norm": 0.13002215325832367,
"learning_rate": 7.305946322744591e-06,
"loss": 1.9746,
"step": 226
},
{
"epoch": 0.7964912280701755,
"grad_norm": 0.1340227872133255,
"learning_rate": 7.070528383624548e-06,
"loss": 1.913,
"step": 227
},
{
"epoch": 0.8,
"grad_norm": 0.12861642241477966,
"learning_rate": 6.838458108123103e-06,
"loss": 2.28,
"step": 228
},
{
"epoch": 0.8035087719298246,
"grad_norm": 0.12282972782850266,
"learning_rate": 6.6097693781033316e-06,
"loss": 2.1254,
"step": 229
},
{
"epoch": 0.8070175438596491,
"grad_norm": 0.12134941667318344,
"learning_rate": 6.384495581728498e-06,
"loss": 2.1872,
"step": 230
},
{
"epoch": 0.8105263157894737,
"grad_norm": 0.1427871286869049,
"learning_rate": 6.162669608587376e-06,
"loss": 2.0682,
"step": 231
},
{
"epoch": 0.8140350877192982,
"grad_norm": 0.1252429336309433,
"learning_rate": 5.944323844892447e-06,
"loss": 2.1674,
"step": 232
}
],
"logging_steps": 1,
"max_steps": 285,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 29,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 9.730618611109724e+17,
"train_batch_size": 12,
"trial_name": null,
"trial_params": null
}