9b-46 / trainer_state.json
furproxy's picture
Upload folder using huggingface_hub
3aed0b2 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 348,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.017241379310344827,
"grad_norm": 0.8816090822219849,
"learning_rate": 1.6666666666666667e-06,
"loss": 2.133296012878418,
"step": 2
},
{
"epoch": 0.034482758620689655,
"grad_norm": 0.29265427589416504,
"learning_rate": 4.9999999999999996e-06,
"loss": 1.8569927215576172,
"step": 4
},
{
"epoch": 0.05172413793103448,
"grad_norm": 0.4113790988922119,
"learning_rate": 8.333333333333334e-06,
"loss": 1.9297298192977905,
"step": 6
},
{
"epoch": 0.06896551724137931,
"grad_norm": 0.501505434513092,
"learning_rate": 1.1666666666666668e-05,
"loss": 4.868577480316162,
"step": 8
},
{
"epoch": 0.08620689655172414,
"grad_norm": 0.43636268377304077,
"learning_rate": 1.5e-05,
"loss": 2.009166955947876,
"step": 10
},
{
"epoch": 0.10344827586206896,
"grad_norm": 3.4641079902648926,
"learning_rate": 1.8333333333333336e-05,
"loss": 2.0880112648010254,
"step": 12
},
{
"epoch": 0.1206896551724138,
"grad_norm": 2.272722005844116,
"learning_rate": 2.1666666666666667e-05,
"loss": 2.3606438636779785,
"step": 14
},
{
"epoch": 0.13793103448275862,
"grad_norm": 0.5703615546226501,
"learning_rate": 2.5e-05,
"loss": 1.9572478532791138,
"step": 16
},
{
"epoch": 0.15517241379310345,
"grad_norm": 0.23337635397911072,
"learning_rate": 2.8333333333333332e-05,
"loss": 1.6358482837677002,
"step": 18
},
{
"epoch": 0.1724137931034483,
"grad_norm": 0.2846831679344177,
"learning_rate": 2.999373614050377e-05,
"loss": 1.5971399545669556,
"step": 20
},
{
"epoch": 0.1896551724137931,
"grad_norm": 0.22005710005760193,
"learning_rate": 2.9943660135527193e-05,
"loss": 1.9199273586273193,
"step": 22
},
{
"epoch": 0.20689655172413793,
"grad_norm": 0.30223548412323,
"learning_rate": 2.984369396041239e-05,
"loss": 1.5466594696044922,
"step": 24
},
{
"epoch": 0.22413793103448276,
"grad_norm": 4.415768146514893,
"learning_rate": 2.9694208595192654e-05,
"loss": 1.039365530014038,
"step": 26
},
{
"epoch": 0.2413793103448276,
"grad_norm": 0.2239375114440918,
"learning_rate": 2.949575878836866e-05,
"loss": 1.457193374633789,
"step": 28
},
{
"epoch": 0.25862068965517243,
"grad_norm": 0.5077499151229858,
"learning_rate": 2.9249080998205997e-05,
"loss": 1.247235894203186,
"step": 30
},
{
"epoch": 0.27586206896551724,
"grad_norm": 0.1902107149362564,
"learning_rate": 2.895509065969754e-05,
"loss": 1.132712721824646,
"step": 32
},
{
"epoch": 0.29310344827586204,
"grad_norm": 0.7180187106132507,
"learning_rate": 2.861487878733334e-05,
"loss": 1.1004750728607178,
"step": 34
},
{
"epoch": 0.3103448275862069,
"grad_norm": 0.19518446922302246,
"learning_rate": 2.8229707926285193e-05,
"loss": 1.2603535652160645,
"step": 36
},
{
"epoch": 0.3275862068965517,
"grad_norm": 0.2706547975540161,
"learning_rate": 2.7801007467031303e-05,
"loss": 1.564131259918213,
"step": 38
},
{
"epoch": 0.3448275862068966,
"grad_norm": 0.11646223813295364,
"learning_rate": 2.7330368340808766e-05,
"loss": 1.3439611196517944,
"step": 40
},
{
"epoch": 0.3620689655172414,
"grad_norm": 0.18652920424938202,
"learning_rate": 2.68195371155793e-05,
"loss": 1.653893232345581,
"step": 42
},
{
"epoch": 0.3793103448275862,
"grad_norm": 0.35720935463905334,
"learning_rate": 2.6270409514418447e-05,
"loss": 1.345064401626587,
"step": 44
},
{
"epoch": 0.39655172413793105,
"grad_norm": 0.18193630874156952,
"learning_rate": 2.568502338038184e-05,
"loss": 1.3374089002609253,
"step": 46
},
{
"epoch": 0.41379310344827586,
"grad_norm": 0.15458394587039948,
"learning_rate": 2.5065551113956345e-05,
"loss": 1.3249835968017578,
"step": 48
},
{
"epoch": 0.43103448275862066,
"grad_norm": 0.3153666853904724,
"learning_rate": 2.441429161116107e-05,
"loss": 1.143474817276001,
"step": 50
},
{
"epoch": 0.4482758620689655,
"grad_norm": 0.1796191781759262,
"learning_rate": 2.3733661732216455e-05,
"loss": 0.91184401512146,
"step": 52
},
{
"epoch": 0.46551724137931033,
"grad_norm": 0.2718876004219055,
"learning_rate": 2.3026187332441612e-05,
"loss": 1.1794605255126953,
"step": 54
},
{
"epoch": 0.4827586206896552,
"grad_norm": 0.154719278216362,
"learning_rate": 2.2294493888664832e-05,
"loss": 1.2965553998947144,
"step": 56
},
{
"epoch": 0.5,
"grad_norm": 0.14152568578720093,
"learning_rate": 2.1541296755933106e-05,
"loss": 1.4213244915008545,
"step": 58
},
{
"epoch": 0.5172413793103449,
"grad_norm": 0.15456536412239075,
"learning_rate": 2.0769391090678594e-05,
"loss": 1.25790274143219,
"step": 60
},
{
"epoch": 0.5344827586206896,
"grad_norm": 0.25589054822921753,
"learning_rate": 1.998164147773771e-05,
"loss": 1.2669916152954102,
"step": 62
},
{
"epoch": 0.5517241379310345,
"grad_norm": 0.19327247142791748,
"learning_rate": 1.9180971299717562e-05,
"loss": 1.0608755350112915,
"step": 64
},
{
"epoch": 0.5689655172413793,
"grad_norm": 0.1933823823928833,
"learning_rate": 1.8370351888160553e-05,
"loss": 1.2603998184204102,
"step": 66
},
{
"epoch": 0.5862068965517241,
"grad_norm": 0.37840721011161804,
"learning_rate": 1.755279149676784e-05,
"loss": 0.9561605453491211,
"step": 68
},
{
"epoch": 0.603448275862069,
"grad_norm": 0.21245446801185608,
"learning_rate": 1.6731324137602643e-05,
"loss": 1.0841741561889648,
"step": 70
},
{
"epoch": 0.6206896551724138,
"grad_norm": 0.2268361747264862,
"learning_rate": 1.590899832170276e-05,
"loss": 0.8117428421974182,
"step": 72
},
{
"epoch": 0.6379310344827587,
"grad_norm": 0.36945435404777527,
"learning_rate": 1.5088865745886681e-05,
"loss": 0.8893211483955383,
"step": 74
},
{
"epoch": 0.6551724137931034,
"grad_norm": 0.43866750597953796,
"learning_rate": 1.4273969967736997e-05,
"loss": 1.256456732749939,
"step": 76
},
{
"epoch": 0.6724137931034483,
"grad_norm": 0.6824928522109985,
"learning_rate": 1.3467335110789045e-05,
"loss": 1.0219424962997437,
"step": 78
},
{
"epoch": 0.6896551724137931,
"grad_norm": 0.1600809097290039,
"learning_rate": 1.267195464184032e-05,
"loss": 1.3244811296463013,
"step": 80
},
{
"epoch": 0.7068965517241379,
"grad_norm": 0.16961514949798584,
"learning_rate": 1.1890780262028785e-05,
"loss": 1.2907791137695312,
"step": 82
},
{
"epoch": 0.7241379310344828,
"grad_norm": 0.12361691892147064,
"learning_rate": 1.1126710952905864e-05,
"loss": 1.0599552392959595,
"step": 84
},
{
"epoch": 0.7413793103448276,
"grad_norm": 0.2644699811935425,
"learning_rate": 1.0382582218154716e-05,
"loss": 0.931882381439209,
"step": 86
},
{
"epoch": 0.7586206896551724,
"grad_norm": 0.3073562681674957,
"learning_rate": 9.661155560878463e-06,
"loss": 0.7623745203018188,
"step": 88
},
{
"epoch": 0.7758620689655172,
"grad_norm": 0.14720334112644196,
"learning_rate": 8.965108235508627e-06,
"loss": 0.8441999554634094,
"step": 90
},
{
"epoch": 0.7931034482758621,
"grad_norm": 0.29734253883361816,
"learning_rate": 8.29702331236512e-06,
"loss": 0.8531256318092346,
"step": 92
},
{
"epoch": 0.8103448275862069,
"grad_norm": 0.5815638899803162,
"learning_rate": 7.659380091738652e-06,
"loss": 0.8273341655731201,
"step": 94
},
{
"epoch": 0.8275862068965517,
"grad_norm": 1.9160736799240112,
"learning_rate": 7.054544903069566e-06,
"loss": 0.4431428015232086,
"step": 96
},
{
"epoch": 0.8448275862068966,
"grad_norm": 0.16483919322490692,
"learning_rate": 6.484762323367704e-06,
"loss": 1.2755130529403687,
"step": 98
},
{
"epoch": 0.8620689655172413,
"grad_norm": 0.42587000131607056,
"learning_rate": 5.9521468474623304e-06,
"loss": 1.144418716430664,
"step": 100
},
{
"epoch": 0.8793103448275862,
"grad_norm": 0.15736447274684906,
"learning_rate": 5.458675040994306e-06,
"loss": 1.2321348190307617,
"step": 102
},
{
"epoch": 0.896551724137931,
"grad_norm": 0.3543795645236969,
"learning_rate": 5.006178205271082e-06,
"loss": 0.9103032946586609,
"step": 104
},
{
"epoch": 0.9137931034482759,
"grad_norm": 0.1326935738325119,
"learning_rate": 4.596335581205745e-06,
"loss": 1.2454521656036377,
"step": 106
},
{
"epoch": 0.9310344827586207,
"grad_norm": 1.112565517425537,
"learning_rate": 4.230668117560634e-06,
"loss": 1.1100295782089233,
"step": 108
},
{
"epoch": 0.9482758620689655,
"grad_norm": 0.18412314355373383,
"learning_rate": 3.910532826621947e-06,
"loss": 1.0047061443328857,
"step": 110
},
{
"epoch": 0.9655172413793104,
"grad_norm": 0.1519099920988083,
"learning_rate": 3.6371177482516813e-06,
"loss": 0.7593088150024414,
"step": 112
},
{
"epoch": 0.9827586206896551,
"grad_norm": 0.2574796974658966,
"learning_rate": 3.411437541005667e-06,
"loss": 0.8851341605186462,
"step": 114
},
{
"epoch": 1.0,
"grad_norm": 0.162760391831398,
"learning_rate": 3.234329716679273e-06,
"loss": 1.245699405670166,
"step": 116
},
{
"epoch": 1.0172413793103448,
"grad_norm": 0.1872638612985611,
"learning_rate": 3.1064515322545505e-06,
"loss": 1.221221923828125,
"step": 118
},
{
"epoch": 1.0344827586206897,
"grad_norm": 0.21109050512313843,
"learning_rate": 3.0282775507830497e-06,
"loss": 0.7024672627449036,
"step": 120
},
{
"epoch": 1.0517241379310345,
"grad_norm": 0.14990702271461487,
"learning_rate": 3.000097880255948e-06,
"loss": 1.2402740716934204,
"step": 122
},
{
"epoch": 1.0689655172413792,
"grad_norm": 0.23416019976139069,
"learning_rate": 3.0220170969971693e-06,
"loss": 1.1297236680984497,
"step": 124
},
{
"epoch": 1.0862068965517242,
"grad_norm": 0.19604845345020294,
"learning_rate": 3.093953857574826e-06,
"loss": 1.0468320846557617,
"step": 126
},
{
"epoch": 1.103448275862069,
"grad_norm": 0.1591433882713318,
"learning_rate": 3.2156412006712305e-06,
"loss": 1.190972089767456,
"step": 128
},
{
"epoch": 1.1206896551724137,
"grad_norm": 0.21266596019268036,
"learning_rate": 3.386627537791162e-06,
"loss": 0.83074551820755,
"step": 130
},
{
"epoch": 1.1379310344827587,
"grad_norm": 0.1659928411245346,
"learning_rate": 3.606278329131906e-06,
"loss": 1.2269715070724487,
"step": 132
},
{
"epoch": 1.1551724137931034,
"grad_norm": 0.3175746500492096,
"learning_rate": 3.873778438395723e-06,
"loss": 0.5150075554847717,
"step": 134
},
{
"epoch": 1.1724137931034484,
"grad_norm": 0.31136131286621094,
"learning_rate": 4.1881351578059545e-06,
"loss": 0.47654953598976135,
"step": 136
},
{
"epoch": 1.1896551724137931,
"grad_norm": 0.14184990525245667,
"learning_rate": 4.5481818921007985e-06,
"loss": 1.2832177877426147,
"step": 138
},
{
"epoch": 1.206896551724138,
"grad_norm": 0.25604957342147827,
"learning_rate": 4.952582487833158e-06,
"loss": 1.1527303457260132,
"step": 140
},
{
"epoch": 1.2241379310344827,
"grad_norm": 0.13393063843250275,
"learning_rate": 5.3998361919103724e-06,
"loss": 1.4890146255493164,
"step": 142
},
{
"epoch": 1.2413793103448276,
"grad_norm": 0.12155193835496902,
"learning_rate": 5.888283220972366e-06,
"loss": 1.1686463356018066,
"step": 144
},
{
"epoch": 1.2586206896551724,
"grad_norm": 0.38867977261543274,
"learning_rate": 6.416110920939995e-06,
"loss": 1.1420594453811646,
"step": 146
},
{
"epoch": 1.2758620689655173,
"grad_norm": 0.1920805424451828,
"learning_rate": 6.981360493875075e-06,
"loss": 0.8097314834594727,
"step": 148
},
{
"epoch": 1.293103448275862,
"grad_norm": 0.2275863140821457,
"learning_rate": 7.581934267188444e-06,
"loss": 1.1563327312469482,
"step": 150
},
{
"epoch": 1.3103448275862069,
"grad_norm": 0.6193908452987671,
"learning_rate": 8.215603478219558e-06,
"loss": 0.7376953959465027,
"step": 152
},
{
"epoch": 1.3275862068965516,
"grad_norm": 0.13849061727523804,
"learning_rate": 8.880016545298615e-06,
"loss": 0.8840894103050232,
"step": 154
},
{
"epoch": 1.3448275862068966,
"grad_norm": 0.4044109284877777,
"learning_rate": 9.572707794596881e-06,
"loss": 1.3034617900848389,
"step": 156
},
{
"epoch": 1.3620689655172413,
"grad_norm": 0.1762719750404358,
"learning_rate": 1.0291106610379316e-05,
"loss": 1.2110811471939087,
"step": 158
},
{
"epoch": 1.3793103448275863,
"grad_norm": 0.21315965056419373,
"learning_rate": 1.1032546974702523e-05,
"loss": 1.174965262413025,
"step": 160
},
{
"epoch": 1.396551724137931,
"grad_norm": 0.1833634078502655,
"learning_rate": 1.1794277361155493e-05,
"loss": 0.9785462021827698,
"step": 162
},
{
"epoch": 1.4137931034482758,
"grad_norm": 0.2881973683834076,
"learning_rate": 1.2573470945927114e-05,
"loss": 0.7486613988876343,
"step": 164
},
{
"epoch": 1.4310344827586206,
"grad_norm": 0.1457541584968567,
"learning_rate": 1.3367236098306466e-05,
"loss": 1.2081069946289062,
"step": 166
},
{
"epoch": 1.4482758620689655,
"grad_norm": 0.22875168919563293,
"learning_rate": 1.4172627111685208e-05,
"loss": 1.2631800174713135,
"step": 168
},
{
"epoch": 1.4655172413793103,
"grad_norm": 0.14663054049015045,
"learning_rate": 1.4986655135238434e-05,
"loss": 0.8129775524139404,
"step": 170
},
{
"epoch": 1.4827586206896552,
"grad_norm": 0.18815924227237701,
"learning_rate": 1.5806299265716396e-05,
"loss": 1.2445049285888672,
"step": 172
},
{
"epoch": 1.5,
"grad_norm": 0.23885682225227356,
"learning_rate": 1.6628517758184285e-05,
"loss": 0.7052794694900513,
"step": 174
},
{
"epoch": 1.5172413793103448,
"grad_norm": 0.13497234880924225,
"learning_rate": 1.7450259314106843e-05,
"loss": 1.4724665880203247,
"step": 176
},
{
"epoch": 1.5344827586206895,
"grad_norm": 0.5048268437385559,
"learning_rate": 1.826847440488684e-05,
"loss": 0.32006025314331055,
"step": 178
},
{
"epoch": 1.5517241379310345,
"grad_norm": 0.8052246570587158,
"learning_rate": 1.9080126588835274e-05,
"loss": 0.29676616191864014,
"step": 180
},
{
"epoch": 1.5689655172413794,
"grad_norm": 0.43781599402427673,
"learning_rate": 1.9882203779575198e-05,
"loss": 1.1562625169754028,
"step": 182
},
{
"epoch": 1.5862068965517242,
"grad_norm": 0.1728672832250595,
"learning_rate": 2.0671729424061787e-05,
"loss": 0.9202221632003784,
"step": 184
},
{
"epoch": 1.603448275862069,
"grad_norm": 0.5775326490402222,
"learning_rate": 2.144577354873615e-05,
"loss": 1.2155585289001465,
"step": 186
},
{
"epoch": 1.6206896551724137,
"grad_norm": 0.17796771228313446,
"learning_rate": 2.220146363282005e-05,
"loss": 1.167516827583313,
"step": 188
},
{
"epoch": 1.6379310344827587,
"grad_norm": 0.3755885064601898,
"learning_rate": 2.2935995268400318e-05,
"loss": 1.1323333978652954,
"step": 190
},
{
"epoch": 1.6551724137931034,
"grad_norm": 0.21519975364208221,
"learning_rate": 2.3646642567742345e-05,
"loss": 0.9196767807006836,
"step": 192
},
{
"epoch": 1.6724137931034484,
"grad_norm": 0.16918060183525085,
"learning_rate": 2.433076827921117e-05,
"loss": 0.9174919724464417,
"step": 194
},
{
"epoch": 1.6896551724137931,
"grad_norm": 0.20906352996826172,
"learning_rate": 2.4985833574258856e-05,
"loss": 1.3593827486038208,
"step": 196
},
{
"epoch": 1.706896551724138,
"grad_norm": 0.620851457118988,
"learning_rate": 2.560940746915833e-05,
"loss": 0.9581778645515442,
"step": 198
},
{
"epoch": 1.7241379310344827,
"grad_norm": 0.5852736830711365,
"learning_rate": 2.619917584651918e-05,
"loss": 0.6433994174003601,
"step": 200
},
{
"epoch": 1.7413793103448276,
"grad_norm": 0.18479883670806885,
"learning_rate": 2.6752950043105974e-05,
"loss": 1.1657538414001465,
"step": 202
},
{
"epoch": 1.7586206896551724,
"grad_norm": 0.18072469532489777,
"learning_rate": 2.7268674972089096e-05,
"loss": 1.1429758071899414,
"step": 204
},
{
"epoch": 1.7758620689655173,
"grad_norm": 0.2669624388217926,
"learning_rate": 2.7744436749586343e-05,
"loss": 1.2312031984329224,
"step": 206
},
{
"epoch": 1.793103448275862,
"grad_norm": 0.16005530953407288,
"learning_rate": 2.8178469797192326e-05,
"loss": 1.1482793092727661,
"step": 208
},
{
"epoch": 1.8103448275862069,
"grad_norm": 0.24072809517383575,
"learning_rate": 2.856916339413811e-05,
"loss": 1.1373822689056396,
"step": 210
},
{
"epoch": 1.8275862068965516,
"grad_norm": 0.1878095120191574,
"learning_rate": 2.8915067654765562e-05,
"loss": 0.8796662092208862,
"step": 212
},
{
"epoch": 1.8448275862068966,
"grad_norm": 0.21740852296352386,
"learning_rate": 2.9214898909133586e-05,
"loss": 1.0384447574615479,
"step": 214
},
{
"epoch": 1.8620689655172413,
"grad_norm": 0.9018149971961975,
"learning_rate": 2.946754446678865e-05,
"loss": 0.5491471290588379,
"step": 216
},
{
"epoch": 1.8793103448275863,
"grad_norm": 0.13663442432880402,
"learning_rate": 2.967206674602094e-05,
"loss": 0.8381883502006531,
"step": 218
},
{
"epoch": 1.896551724137931,
"grad_norm": 0.17705294489860535,
"learning_rate": 2.98277067532822e-05,
"loss": 0.8339143991470337,
"step": 220
},
{
"epoch": 1.9137931034482758,
"grad_norm": 0.6685168743133545,
"learning_rate": 2.993388689985295e-05,
"loss": 0.8806865215301514,
"step": 222
},
{
"epoch": 1.9310344827586206,
"grad_norm": 0.27922579646110535,
"learning_rate": 2.9990213145306393e-05,
"loss": 1.1154509782791138,
"step": 224
},
{
"epoch": 1.9482758620689655,
"grad_norm": 0.42925339937210083,
"learning_rate": 2.999647645981429e-05,
"loss": 0.8699172735214233,
"step": 226
},
{
"epoch": 1.9655172413793105,
"grad_norm": 0.30161452293395996,
"learning_rate": 2.995265359986831e-05,
"loss": 1.0112342834472656,
"step": 228
},
{
"epoch": 1.9827586206896552,
"grad_norm": 0.5354682803153992,
"learning_rate": 2.9858907194537958e-05,
"loss": 1.221873164176941,
"step": 230
},
{
"epoch": 2.0,
"grad_norm": 0.1859283596277237,
"learning_rate": 2.9715585141945073e-05,
"loss": 0.8529492616653442,
"step": 232
},
{
"epoch": 2.0172413793103448,
"grad_norm": 0.22067049145698547,
"learning_rate": 2.9523219318194555e-05,
"loss": 1.0910695791244507,
"step": 234
},
{
"epoch": 2.0344827586206895,
"grad_norm": 0.4083843231201172,
"learning_rate": 2.9282523603552557e-05,
"loss": 0.39936938881874084,
"step": 236
},
{
"epoch": 2.0517241379310347,
"grad_norm": 0.22413796186447144,
"learning_rate": 2.899439123319714e-05,
"loss": 0.6455128788948059,
"step": 238
},
{
"epoch": 2.0689655172413794,
"grad_norm": 0.44840389490127563,
"learning_rate": 2.86598914823729e-05,
"loss": 1.0364433526992798,
"step": 240
},
{
"epoch": 2.086206896551724,
"grad_norm": 0.245926633477211,
"learning_rate": 2.8280265698251005e-05,
"loss": 0.6382094025611877,
"step": 242
},
{
"epoch": 2.103448275862069,
"grad_norm": 1.7387171983718872,
"learning_rate": 2.7856922693220947e-05,
"loss": 0.6532158255577087,
"step": 244
},
{
"epoch": 2.1206896551724137,
"grad_norm": 0.3799879848957062,
"learning_rate": 2.739143351670935e-05,
"loss": 0.8184542655944824,
"step": 246
},
{
"epoch": 2.1379310344827585,
"grad_norm": 0.25618430972099304,
"learning_rate": 2.688552562492825e-05,
"loss": 0.675823986530304,
"step": 248
},
{
"epoch": 2.1551724137931036,
"grad_norm": 0.32830527424812317,
"learning_rate": 2.6341076470189065e-05,
"loss": 0.7770854830741882,
"step": 250
},
{
"epoch": 2.1724137931034484,
"grad_norm": 0.15962277352809906,
"learning_rate": 2.5760106533572648e-05,
"loss": 0.713956356048584,
"step": 252
},
{
"epoch": 2.189655172413793,
"grad_norm": 0.42531585693359375,
"learning_rate": 2.5144771826811748e-05,
"loss": 0.4886062741279602,
"step": 254
},
{
"epoch": 2.206896551724138,
"grad_norm": 0.23626545071601868,
"learning_rate": 2.4497355891211547e-05,
"loss": 1.1009202003479004,
"step": 256
},
{
"epoch": 2.2241379310344827,
"grad_norm": 0.29544082283973694,
"learning_rate": 2.3820261323300925e-05,
"loss": 1.0635813474655151,
"step": 258
},
{
"epoch": 2.2413793103448274,
"grad_norm": 0.3267728090286255,
"learning_rate": 2.311600085866322e-05,
"loss": 0.9566066265106201,
"step": 260
},
{
"epoch": 2.2586206896551726,
"grad_norm": 0.09956636279821396,
"learning_rate": 2.238718804703476e-05,
"loss": 0.6884672045707703,
"step": 262
},
{
"epoch": 2.2758620689655173,
"grad_norm": 0.15193510055541992,
"learning_rate": 2.163652755327646e-05,
"loss": 0.9428229928016663,
"step": 264
},
{
"epoch": 2.293103448275862,
"grad_norm": 0.24053514003753662,
"learning_rate": 2.0866805120212266e-05,
"loss": 0.6498053669929504,
"step": 266
},
{
"epoch": 2.310344827586207,
"grad_norm": 0.13813968002796173,
"learning_rate": 2.0080877230582565e-05,
"loss": 1.1316256523132324,
"step": 268
},
{
"epoch": 2.3275862068965516,
"grad_norm": 0.3910239636898041,
"learning_rate": 1.9281660506477953e-05,
"loss": 1.0687792301177979,
"step": 270
},
{
"epoch": 2.344827586206897,
"grad_norm": 0.23670506477355957,
"learning_rate": 1.8472120885592557e-05,
"loss": 1.125198245048523,
"step": 272
},
{
"epoch": 2.3620689655172415,
"grad_norm": 0.7727434635162354,
"learning_rate": 1.7655262614464347e-05,
"loss": 0.5221872329711914,
"step": 274
},
{
"epoch": 2.3793103448275863,
"grad_norm": 0.5010542869567871,
"learning_rate": 1.683411709954916e-05,
"loss": 1.0470247268676758,
"step": 276
},
{
"epoch": 2.396551724137931,
"grad_norm": 1.5273538827896118,
"learning_rate": 1.6011731657502948e-05,
"loss": 0.6991406679153442,
"step": 278
},
{
"epoch": 2.413793103448276,
"grad_norm": 0.12472137063741684,
"learning_rate": 1.5191158206420104e-05,
"loss": 1.0698401927947998,
"step": 280
},
{
"epoch": 2.4310344827586206,
"grad_norm": 0.40073373913764954,
"learning_rate": 1.4375441939995862e-05,
"loss": 0.9676493406295776,
"step": 282
},
{
"epoch": 2.4482758620689653,
"grad_norm": 0.29846158623695374,
"learning_rate": 1.3567610026643158e-05,
"loss": 0.8159676194190979,
"step": 284
},
{
"epoch": 2.4655172413793105,
"grad_norm": 0.0978626236319542,
"learning_rate": 1.2770660375502325e-05,
"loss": 0.731606125831604,
"step": 286
},
{
"epoch": 2.4827586206896552,
"grad_norm": 0.1449108123779297,
"learning_rate": 1.1987550511033729e-05,
"loss": 0.9657235145568848,
"step": 288
},
{
"epoch": 2.5,
"grad_norm": 0.7278546690940857,
"learning_rate": 1.1221186597479921e-05,
"loss": 0.6879425644874573,
"step": 290
},
{
"epoch": 2.5172413793103448,
"grad_norm": 0.1562328189611435,
"learning_rate": 1.047441265392866e-05,
"loss": 0.6442292928695679,
"step": 292
},
{
"epoch": 2.5344827586206895,
"grad_norm": 0.412852942943573,
"learning_rate": 9.74999999999999e-06,
"loss": 1.2038459777832031,
"step": 294
},
{
"epoch": 2.5517241379310347,
"grad_norm": 0.8354277014732361,
"learning_rate": 9.05063697132504e-06,
"loss": 0.6161110401153564,
"step": 296
},
{
"epoch": 2.5689655172413794,
"grad_norm": 0.3308437466621399,
"learning_rate": 8.378918942982984e-06,
"loss": 0.4709848463535309,
"step": 298
},
{
"epoch": 2.586206896551724,
"grad_norm": 0.15517888963222504,
"learning_rate": 7.7373386979199e-06,
"loss": 0.9857034683227539,
"step": 300
},
{
"epoch": 2.603448275862069,
"grad_norm": 0.8485265970230103,
"learning_rate": 7.128277176092711e-06,
"loss": 0.93929523229599,
"step": 302
},
{
"epoch": 2.6206896551724137,
"grad_norm": 0.7992085218429565,
"learning_rate": 6.5539946386685424e-06,
"loss": 0.8182238340377808,
"step": 304
},
{
"epoch": 2.637931034482759,
"grad_norm": 0.15321463346481323,
"learning_rate": 6.01662228006993e-06,
"loss": 0.8413806557655334,
"step": 306
},
{
"epoch": 2.655172413793103,
"grad_norm": 0.26136964559555054,
"learning_rate": 5.5181543189938835e-06,
"loss": 0.7553901076316833,
"step": 308
},
{
"epoch": 2.6724137931034484,
"grad_norm": 0.6263846158981323,
"learning_rate": 5.060440597755323e-06,
"loss": 0.8772685527801514,
"step": 310
},
{
"epoch": 2.689655172413793,
"grad_norm": 0.1620139181613922,
"learning_rate": 4.6451797174195354e-06,
"loss": 0.9297232031822205,
"step": 312
},
{
"epoch": 2.706896551724138,
"grad_norm": 0.14144213497638702,
"learning_rate": 4.273912734199091e-06,
"loss": 0.8999103307723999,
"step": 314
},
{
"epoch": 2.7241379310344827,
"grad_norm": 0.4446767568588257,
"learning_rate": 3.948017440508606e-06,
"loss": 0.32868385314941406,
"step": 316
},
{
"epoch": 2.7413793103448274,
"grad_norm": 0.2091308832168579,
"learning_rate": 3.6687032519005313e-06,
"loss": 1.1546356678009033,
"step": 318
},
{
"epoch": 2.7586206896551726,
"grad_norm": 0.1578148901462555,
"learning_rate": 3.4370067188567254e-06,
"loss": 0.9378973841667175,
"step": 320
},
{
"epoch": 2.7758620689655173,
"grad_norm": 1.2611632347106934,
"learning_rate": 3.2537876800920677e-06,
"loss": 0.6517216563224792,
"step": 322
},
{
"epoch": 2.793103448275862,
"grad_norm": 0.19654294848442078,
"learning_rate": 3.119726071645127e-06,
"loss": 0.7699644565582275,
"step": 324
},
{
"epoch": 2.810344827586207,
"grad_norm": 0.2935585379600525,
"learning_rate": 3.035319403597751e-06,
"loss": 1.0573527812957764,
"step": 326
},
{
"epoch": 2.8275862068965516,
"grad_norm": 0.18080638349056244,
"learning_rate": 3.000880913787505e-06,
"loss": 1.1476738452911377,
"step": 328
},
{
"epoch": 2.844827586206897,
"grad_norm": 0.1654699146747589,
"learning_rate": 3.016538405364614e-06,
"loss": 1.1468658447265625,
"step": 330
},
{
"epoch": 2.862068965517241,
"grad_norm": 0.17703653872013092,
"learning_rate": 3.082233772507386e-06,
"loss": 1.062970757484436,
"step": 332
},
{
"epoch": 2.8793103448275863,
"grad_norm": 0.37376028299331665,
"learning_rate": 3.1977232160561467e-06,
"loss": 0.603091835975647,
"step": 334
},
{
"epoch": 2.896551724137931,
"grad_norm": 0.19850152730941772,
"learning_rate": 3.3625781482654565e-06,
"loss": 0.7071291208267212,
"step": 336
},
{
"epoch": 2.913793103448276,
"grad_norm": 0.1782890260219574,
"learning_rate": 3.5761867833170895e-06,
"loss": 0.658890962600708,
"step": 338
},
{
"epoch": 2.9310344827586206,
"grad_norm": 0.18476705253124237,
"learning_rate": 3.8377564076912e-06,
"loss": 1.0716102123260498,
"step": 340
},
{
"epoch": 2.9482758620689653,
"grad_norm": 0.30881527066230774,
"learning_rate": 4.146316321970244e-06,
"loss": 1.2615000009536743,
"step": 342
},
{
"epoch": 2.9655172413793105,
"grad_norm": 0.3074623644351959,
"learning_rate": 4.5007214431585365e-06,
"loss": 0.8507893681526184,
"step": 344
},
{
"epoch": 2.9827586206896552,
"grad_norm": 0.21223706007003784,
"learning_rate": 4.8996565541488335e-06,
"loss": 0.6322518587112427,
"step": 346
},
{
"epoch": 3.0,
"grad_norm": 0.4978598952293396,
"learning_rate": 5.341641184566203e-06,
"loss": 0.7810978293418884,
"step": 348
},
{
"epoch": 3.0,
"step": 348,
"total_flos": 2.7756656592022405e+18,
"train_loss": 1.0638441426315526,
"train_runtime": 6969.7468,
"train_samples_per_second": 5.992,
"train_steps_per_second": 0.05
}
],
"logging_steps": 2,
"max_steps": 348,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 99999,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.7756656592022405e+18,
"train_batch_size": 3,
"trial_name": null,
"trial_params": null
}