autoprogrammer's picture
Add checkpoint-921
68dd5e6 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 921,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0163265306122449,
"grad_norm": 62.72153091430664,
"learning_rate": 1.4285714285714286e-06,
"loss": 14.1797,
"step": 5
},
{
"epoch": 0.0326530612244898,
"grad_norm": 47.921302795410156,
"learning_rate": 3.2142857142857147e-06,
"loss": 13.6594,
"step": 10
},
{
"epoch": 0.04897959183673469,
"grad_norm": 38.88978958129883,
"learning_rate": 5e-06,
"loss": 11.9582,
"step": 15
},
{
"epoch": 0.0653061224489796,
"grad_norm": 32.92738723754883,
"learning_rate": 6.785714285714287e-06,
"loss": 9.9422,
"step": 20
},
{
"epoch": 0.08163265306122448,
"grad_norm": 26.181106567382812,
"learning_rate": 8.571428571428571e-06,
"loss": 9.0908,
"step": 25
},
{
"epoch": 0.09795918367346938,
"grad_norm": 25.063541412353516,
"learning_rate": 9.99996905885448e-06,
"loss": 8.0087,
"step": 30
},
{
"epoch": 0.11428571428571428,
"grad_norm": 21.642377853393555,
"learning_rate": 9.9988861589696e-06,
"loss": 7.9636,
"step": 35
},
{
"epoch": 0.1306122448979592,
"grad_norm": 21.017885208129883,
"learning_rate": 9.996256584729157e-06,
"loss": 7.4511,
"step": 40
},
{
"epoch": 0.1469387755102041,
"grad_norm": 20.81943130493164,
"learning_rate": 9.992081149733404e-06,
"loss": 7.2858,
"step": 45
},
{
"epoch": 0.16326530612244897,
"grad_norm": 21.193912506103516,
"learning_rate": 9.986361145877783e-06,
"loss": 7.3675,
"step": 50
},
{
"epoch": 0.17959183673469387,
"grad_norm": 21.050230026245117,
"learning_rate": 9.979098342953198e-06,
"loss": 6.7465,
"step": 55
},
{
"epoch": 0.19591836734693877,
"grad_norm": 18.40810775756836,
"learning_rate": 9.970294988098452e-06,
"loss": 6.9341,
"step": 60
},
{
"epoch": 0.21224489795918366,
"grad_norm": 18.563608169555664,
"learning_rate": 9.959953805104953e-06,
"loss": 6.6632,
"step": 65
},
{
"epoch": 0.22857142857142856,
"grad_norm": 21.610551834106445,
"learning_rate": 9.948077993573983e-06,
"loss": 6.7455,
"step": 70
},
{
"epoch": 0.24489795918367346,
"grad_norm": 19.773212432861328,
"learning_rate": 9.934671227926714e-06,
"loss": 6.8586,
"step": 75
},
{
"epoch": 0.2612244897959184,
"grad_norm": 19.932537078857422,
"learning_rate": 9.919737656267335e-06,
"loss": 6.5781,
"step": 80
},
{
"epoch": 0.27755102040816326,
"grad_norm": 17.995580673217773,
"learning_rate": 9.90328189909961e-06,
"loss": 6.7715,
"step": 85
},
{
"epoch": 0.2938775510204082,
"grad_norm": 18.650741577148438,
"learning_rate": 9.885309047897285e-06,
"loss": 6.6833,
"step": 90
},
{
"epoch": 0.31020408163265306,
"grad_norm": 19.030323028564453,
"learning_rate": 9.86582466352875e-06,
"loss": 6.0972,
"step": 95
},
{
"epoch": 0.32653061224489793,
"grad_norm": 20.6298885345459,
"learning_rate": 9.844834774536503e-06,
"loss": 6.4723,
"step": 100
},
{
"epoch": 0.34285714285714286,
"grad_norm": 19.41197395324707,
"learning_rate": 9.822345875271884e-06,
"loss": 6.2123,
"step": 105
},
{
"epoch": 0.35918367346938773,
"grad_norm": 19.00763702392578,
"learning_rate": 9.798364923885703e-06,
"loss": 6.0186,
"step": 110
},
{
"epoch": 0.37551020408163266,
"grad_norm": 19.79923439025879,
"learning_rate": 9.772899340175362e-06,
"loss": 5.9847,
"step": 115
},
{
"epoch": 0.39183673469387753,
"grad_norm": 17.51770782470703,
"learning_rate": 9.745957003289138e-06,
"loss": 6.0502,
"step": 120
},
{
"epoch": 0.40816326530612246,
"grad_norm": 21.13387107849121,
"learning_rate": 9.71754624928835e-06,
"loss": 6.1906,
"step": 125
},
{
"epoch": 0.42448979591836733,
"grad_norm": 18.06110191345215,
"learning_rate": 9.687675868568145e-06,
"loss": 6.1304,
"step": 130
},
{
"epoch": 0.44081632653061226,
"grad_norm": 18.16826820373535,
"learning_rate": 9.65635510313772e-06,
"loss": 6.0233,
"step": 135
},
{
"epoch": 0.45714285714285713,
"grad_norm": 18.825029373168945,
"learning_rate": 9.623593643760805e-06,
"loss": 5.9259,
"step": 140
},
{
"epoch": 0.47346938775510206,
"grad_norm": 18.966094970703125,
"learning_rate": 9.589401626957309e-06,
"loss": 6.0893,
"step": 145
},
{
"epoch": 0.4897959183673469,
"grad_norm": 19.07234001159668,
"learning_rate": 9.553789631867039e-06,
"loss": 6.2442,
"step": 150
},
{
"epoch": 0.5061224489795918,
"grad_norm": 17.93751335144043,
"learning_rate": 9.516768676976476e-06,
"loss": 5.9128,
"step": 155
},
{
"epoch": 0.5224489795918368,
"grad_norm": 19.36538314819336,
"learning_rate": 9.478350216709609e-06,
"loss": 5.8048,
"step": 160
},
{
"epoch": 0.5387755102040817,
"grad_norm": 19.301006317138672,
"learning_rate": 9.438546137883898e-06,
"loss": 6.276,
"step": 165
},
{
"epoch": 0.5551020408163265,
"grad_norm": 18.383508682250977,
"learning_rate": 9.397368756032445e-06,
"loss": 5.9848,
"step": 170
},
{
"epoch": 0.5714285714285714,
"grad_norm": 20.009305953979492,
"learning_rate": 9.354830811593527e-06,
"loss": 5.9574,
"step": 175
},
{
"epoch": 0.5877551020408164,
"grad_norm": 15.936427116394043,
"learning_rate": 9.31094546596864e-06,
"loss": 6.01,
"step": 180
},
{
"epoch": 0.6040816326530613,
"grad_norm": 19.233423233032227,
"learning_rate": 9.265726297450332e-06,
"loss": 5.9451,
"step": 185
},
{
"epoch": 0.6204081632653061,
"grad_norm": 16.717844009399414,
"learning_rate": 9.219187297021015e-06,
"loss": 5.9017,
"step": 190
},
{
"epoch": 0.636734693877551,
"grad_norm": 17.835121154785156,
"learning_rate": 9.171342864024103e-06,
"loss": 5.5096,
"step": 195
},
{
"epoch": 0.6530612244897959,
"grad_norm": 18.398456573486328,
"learning_rate": 9.122207801708802e-06,
"loss": 5.8136,
"step": 200
},
{
"epoch": 0.6693877551020408,
"grad_norm": 18.077146530151367,
"learning_rate": 9.071797312649934e-06,
"loss": 5.9265,
"step": 205
},
{
"epoch": 0.6857142857142857,
"grad_norm": 15.732425689697266,
"learning_rate": 9.020126994044194e-06,
"loss": 5.8039,
"step": 210
},
{
"epoch": 0.7020408163265306,
"grad_norm": 18.708425521850586,
"learning_rate": 8.96721283288432e-06,
"loss": 5.4268,
"step": 215
},
{
"epoch": 0.7183673469387755,
"grad_norm": 17.379608154296875,
"learning_rate": 8.91307120101266e-06,
"loss": 5.8496,
"step": 220
},
{
"epoch": 0.7346938775510204,
"grad_norm": 17.756458282470703,
"learning_rate": 8.857718850055663e-06,
"loss": 5.8487,
"step": 225
},
{
"epoch": 0.7510204081632653,
"grad_norm": 19.63192367553711,
"learning_rate": 8.801172906240857e-06,
"loss": 5.866,
"step": 230
},
{
"epoch": 0.7673469387755102,
"grad_norm": 16.424760818481445,
"learning_rate": 8.743450865097929e-06,
"loss": 5.746,
"step": 235
},
{
"epoch": 0.7836734693877551,
"grad_norm": 16.6131591796875,
"learning_rate": 8.68457058604555e-06,
"loss": 5.7945,
"step": 240
},
{
"epoch": 0.8,
"grad_norm": 17.163040161132812,
"learning_rate": 8.624550286865592e-06,
"loss": 5.915,
"step": 245
},
{
"epoch": 0.8163265306122449,
"grad_norm": 17.091655731201172,
"learning_rate": 8.563408538066486e-06,
"loss": 5.8607,
"step": 250
},
{
"epoch": 0.8326530612244898,
"grad_norm": 16.500892639160156,
"learning_rate": 8.501164257137431e-06,
"loss": 5.8509,
"step": 255
},
{
"epoch": 0.8489795918367347,
"grad_norm": 16.344823837280273,
"learning_rate": 8.437836702695253e-06,
"loss": 5.5136,
"step": 260
},
{
"epoch": 0.8653061224489796,
"grad_norm": 20.226680755615234,
"learning_rate": 8.373445468525719e-06,
"loss": 5.7412,
"step": 265
},
{
"epoch": 0.8816326530612245,
"grad_norm": 16.34162712097168,
"learning_rate": 8.30801047752113e-06,
"loss": 5.8245,
"step": 270
},
{
"epoch": 0.8979591836734694,
"grad_norm": 17.416240692138672,
"learning_rate": 8.241551975516133e-06,
"loss": 5.9839,
"step": 275
},
{
"epoch": 0.9142857142857143,
"grad_norm": 18.15375328063965,
"learning_rate": 8.17409052502355e-06,
"loss": 5.8697,
"step": 280
},
{
"epoch": 0.9306122448979591,
"grad_norm": 19.698328018188477,
"learning_rate": 8.105646998872275e-06,
"loss": 5.7602,
"step": 285
},
{
"epoch": 0.9469387755102041,
"grad_norm": 16.44139862060547,
"learning_rate": 8.036242573749142e-06,
"loss": 5.4957,
"step": 290
},
{
"epoch": 0.963265306122449,
"grad_norm": 17.497987747192383,
"learning_rate": 7.965898723646777e-06,
"loss": 5.9794,
"step": 295
},
{
"epoch": 0.9795918367346939,
"grad_norm": 16.901426315307617,
"learning_rate": 7.894637213219454e-06,
"loss": 5.4036,
"step": 300
},
{
"epoch": 0.9959183673469387,
"grad_norm": 17.77571678161621,
"learning_rate": 7.82248009104904e-06,
"loss": 5.7303,
"step": 305
},
{
"epoch": 1.0097959183673468,
"grad_norm": 15.3334321975708,
"learning_rate": 7.749449682823077e-06,
"loss": 3.9496,
"step": 310
},
{
"epoch": 1.026122448979592,
"grad_norm": 16.92150115966797,
"learning_rate": 7.675568584427125e-06,
"loss": 4.365,
"step": 315
},
{
"epoch": 1.0424489795918368,
"grad_norm": 18.648591995239258,
"learning_rate": 7.600859654953513e-06,
"loss": 4.1553,
"step": 320
},
{
"epoch": 1.0587755102040817,
"grad_norm": 16.929473876953125,
"learning_rate": 7.525346009628647e-06,
"loss": 4.4238,
"step": 325
},
{
"epoch": 1.0751020408163265,
"grad_norm": 17.300607681274414,
"learning_rate": 7.449051012661073e-06,
"loss": 4.24,
"step": 330
},
{
"epoch": 1.0914285714285714,
"grad_norm": 20.122995376586914,
"learning_rate": 7.371998270012504e-06,
"loss": 4.4822,
"step": 335
},
{
"epoch": 1.1077551020408163,
"grad_norm": 18.745119094848633,
"learning_rate": 7.2942116220940406e-06,
"loss": 4.0978,
"step": 340
},
{
"epoch": 1.1240816326530612,
"grad_norm": 18.034221649169922,
"learning_rate": 7.215715136389862e-06,
"loss": 4.4084,
"step": 345
},
{
"epoch": 1.140408163265306,
"grad_norm": 18.123455047607422,
"learning_rate": 7.136533100010654e-06,
"loss": 4.3913,
"step": 350
},
{
"epoch": 1.1567346938775511,
"grad_norm": 17.553762435913086,
"learning_rate": 7.0566900121790775e-06,
"loss": 4.569,
"step": 355
},
{
"epoch": 1.173061224489796,
"grad_norm": 17.447336196899414,
"learning_rate": 6.9762105766496315e-06,
"loss": 4.1847,
"step": 360
},
{
"epoch": 1.1893877551020409,
"grad_norm": 18.013084411621094,
"learning_rate": 6.8951196940652045e-06,
"loss": 4.3422,
"step": 365
},
{
"epoch": 1.2057142857142857,
"grad_norm": 16.44858741760254,
"learning_rate": 6.8134424542527215e-06,
"loss": 3.8266,
"step": 370
},
{
"epoch": 1.2220408163265306,
"grad_norm": 17.496063232421875,
"learning_rate": 6.731204128460265e-06,
"loss": 4.1437,
"step": 375
},
{
"epoch": 1.2383673469387755,
"grad_norm": 16.731422424316406,
"learning_rate": 6.648430161538047e-06,
"loss": 4.1942,
"step": 380
},
{
"epoch": 1.2546938775510204,
"grad_norm": 16.141498565673828,
"learning_rate": 6.5651461640656825e-06,
"loss": 4.334,
"step": 385
},
{
"epoch": 1.2710204081632652,
"grad_norm": 18.07802391052246,
"learning_rate": 6.481377904428171e-06,
"loss": 4.3282,
"step": 390
},
{
"epoch": 1.28734693877551,
"grad_norm": 21.77152442932129,
"learning_rate": 6.397151300843065e-06,
"loss": 4.3205,
"step": 395
},
{
"epoch": 1.3036734693877552,
"grad_norm": 18.117504119873047,
"learning_rate": 6.312492413341274e-06,
"loss": 4.2105,
"step": 400
},
{
"epoch": 1.32,
"grad_norm": 17.534467697143555,
"learning_rate": 6.227427435703997e-06,
"loss": 4.4447,
"step": 405
},
{
"epoch": 1.336326530612245,
"grad_norm": 18.127460479736328,
"learning_rate": 6.141982687358255e-06,
"loss": 4.2592,
"step": 410
},
{
"epoch": 1.3526530612244898,
"grad_norm": 18.250205993652344,
"learning_rate": 6.056184605233576e-06,
"loss": 4.3037,
"step": 415
},
{
"epoch": 1.3689795918367347,
"grad_norm": 16.719797134399414,
"learning_rate": 5.970059735582295e-06,
"loss": 3.9873,
"step": 420
},
{
"epoch": 1.3853061224489795,
"grad_norm": 16.509572982788086,
"learning_rate": 5.8836347257660485e-06,
"loss": 4.4724,
"step": 425
},
{
"epoch": 1.4016326530612244,
"grad_norm": 19.0765323638916,
"learning_rate": 5.796936316010984e-06,
"loss": 4.2908,
"step": 430
},
{
"epoch": 1.4179591836734695,
"grad_norm": 18.35072135925293,
"learning_rate": 5.709991331134224e-06,
"loss": 4.1736,
"step": 435
},
{
"epoch": 1.4342857142857142,
"grad_norm": 17.8969669342041,
"learning_rate": 5.622826672244169e-06,
"loss": 4.2511,
"step": 440
},
{
"epoch": 1.4506122448979593,
"grad_norm": 18.110383987426758,
"learning_rate": 5.535469308417198e-06,
"loss": 4.1705,
"step": 445
},
{
"epoch": 1.4669387755102041,
"grad_norm": 18.47060203552246,
"learning_rate": 5.447946268353324e-06,
"loss": 4.2629,
"step": 450
},
{
"epoch": 1.483265306122449,
"grad_norm": 17.598196029663086,
"learning_rate": 5.3602846320134216e-06,
"loss": 4.5002,
"step": 455
},
{
"epoch": 1.4995918367346939,
"grad_norm": 18.4842529296875,
"learning_rate": 5.272511522240574e-06,
"loss": 4.0259,
"step": 460
},
{
"epoch": 1.5159183673469387,
"grad_norm": 17.356529235839844,
"learning_rate": 5.184654096368172e-06,
"loss": 4.0298,
"step": 465
},
{
"epoch": 1.5322448979591838,
"grad_norm": 18.3819522857666,
"learning_rate": 5.096739537817324e-06,
"loss": 4.3087,
"step": 470
},
{
"epoch": 1.5485714285714285,
"grad_norm": 17.658527374267578,
"learning_rate": 5.00879504768621e-06,
"loss": 4.3383,
"step": 475
},
{
"epoch": 1.5648979591836736,
"grad_norm": 17.533559799194336,
"learning_rate": 4.92084783633395e-06,
"loss": 4.1381,
"step": 480
},
{
"epoch": 1.5812244897959182,
"grad_norm": 19.04111099243164,
"learning_rate": 4.832925114961629e-06,
"loss": 4.215,
"step": 485
},
{
"epoch": 1.5975510204081633,
"grad_norm": 16.8986759185791,
"learning_rate": 4.745054087193025e-06,
"loss": 4.2722,
"step": 490
},
{
"epoch": 1.6138775510204082,
"grad_norm": 18.75165557861328,
"learning_rate": 4.657261940657732e-06,
"loss": 4.5757,
"step": 495
},
{
"epoch": 1.630204081632653,
"grad_norm": 18.012062072753906,
"learning_rate": 4.569575838579184e-06,
"loss": 4.0448,
"step": 500
},
{
"epoch": 1.646530612244898,
"grad_norm": 17.009748458862305,
"learning_rate": 4.48202291137026e-06,
"loss": 4.0593,
"step": 505
},
{
"epoch": 1.6628571428571428,
"grad_norm": 17.086145401000977,
"learning_rate": 4.394630248239029e-06,
"loss": 4.0783,
"step": 510
},
{
"epoch": 1.679183673469388,
"grad_norm": 16.84444236755371,
"learning_rate": 4.307424888807242e-06,
"loss": 4.0838,
"step": 515
},
{
"epoch": 1.6955102040816326,
"grad_norm": 16.869609832763672,
"learning_rate": 4.220433814744179e-06,
"loss": 4.13,
"step": 520
},
{
"epoch": 1.7118367346938776,
"grad_norm": 17.29819107055664,
"learning_rate": 4.133683941418411e-06,
"loss": 4.0829,
"step": 525
},
{
"epoch": 1.7281632653061223,
"grad_norm": 18.20751953125,
"learning_rate": 4.047202109570086e-06,
"loss": 4.3287,
"step": 530
},
{
"epoch": 1.7444897959183674,
"grad_norm": 17.81183624267578,
"learning_rate": 3.961015077006301e-06,
"loss": 3.9557,
"step": 535
},
{
"epoch": 1.7608163265306123,
"grad_norm": 17.32221794128418,
"learning_rate": 3.875149510322137e-06,
"loss": 4.0355,
"step": 540
},
{
"epoch": 1.7771428571428571,
"grad_norm": 17.272388458251953,
"learning_rate": 3.7896319766499073e-06,
"loss": 4.1266,
"step": 545
},
{
"epoch": 1.793469387755102,
"grad_norm": 16.714859008789062,
"learning_rate": 3.7044889354391934e-06,
"loss": 4.0451,
"step": 550
},
{
"epoch": 1.8097959183673469,
"grad_norm": 17.381494522094727,
"learning_rate": 3.619746730270185e-06,
"loss": 3.9472,
"step": 555
},
{
"epoch": 1.826122448979592,
"grad_norm": 18.791841506958008,
"learning_rate": 3.5354315807028826e-06,
"loss": 4.0682,
"step": 560
},
{
"epoch": 1.8424489795918366,
"grad_norm": 20.267202377319336,
"learning_rate": 3.451569574164658e-06,
"loss": 4.1062,
"step": 565
},
{
"epoch": 1.8587755102040817,
"grad_norm": 18.362844467163086,
"learning_rate": 3.3681866578787124e-06,
"loss": 3.9416,
"step": 570
},
{
"epoch": 1.8751020408163264,
"grad_norm": 18.389270782470703,
"learning_rate": 3.285308630835903e-06,
"loss": 3.9945,
"step": 575
},
{
"epoch": 1.8914285714285715,
"grad_norm": 16.849157333374023,
"learning_rate": 3.202961135812437e-06,
"loss": 3.9355,
"step": 580
},
{
"epoch": 1.9077551020408163,
"grad_norm": 18.02499771118164,
"learning_rate": 3.121169651435903e-06,
"loss": 4.0422,
"step": 585
},
{
"epoch": 1.9240816326530612,
"grad_norm": 17.751028060913086,
"learning_rate": 3.039959484302083e-06,
"loss": 4.1068,
"step": 590
},
{
"epoch": 1.940408163265306,
"grad_norm": 19.579708099365234,
"learning_rate": 2.959355761145001e-06,
"loss": 4.3983,
"step": 595
},
{
"epoch": 1.956734693877551,
"grad_norm": 18.626358032226562,
"learning_rate": 2.8793834210626036e-06,
"loss": 4.2234,
"step": 600
},
{
"epoch": 1.973061224489796,
"grad_norm": 16.147066116333008,
"learning_rate": 2.8000672078005277e-06,
"loss": 4.0854,
"step": 605
},
{
"epoch": 1.9893877551020407,
"grad_norm": 17.97046661376953,
"learning_rate": 2.7214316620962727e-06,
"loss": 4.2465,
"step": 610
},
{
"epoch": 2.003265306122449,
"grad_norm": 17.08619499206543,
"learning_rate": 2.6435011140862167e-06,
"loss": 3.0418,
"step": 615
},
{
"epoch": 2.0195918367346937,
"grad_norm": 15.681780815124512,
"learning_rate": 2.5662996757777716e-06,
"loss": 2.7923,
"step": 620
},
{
"epoch": 2.0359183673469388,
"grad_norm": 18.270936965942383,
"learning_rate": 2.4898512335890425e-06,
"loss": 2.721,
"step": 625
},
{
"epoch": 2.052244897959184,
"grad_norm": 24.456281661987305,
"learning_rate": 2.4141794409582713e-06,
"loss": 2.7695,
"step": 630
},
{
"epoch": 2.0685714285714285,
"grad_norm": 21.39436912536621,
"learning_rate": 2.3393077110253838e-06,
"loss": 2.8396,
"step": 635
},
{
"epoch": 2.0848979591836736,
"grad_norm": 20.58466339111328,
"learning_rate": 2.265259209387867e-06,
"loss": 2.7958,
"step": 640
},
{
"epoch": 2.1012244897959182,
"grad_norm": 19.145729064941406,
"learning_rate": 2.1920568469332458e-06,
"loss": 2.6491,
"step": 645
},
{
"epoch": 2.1175510204081633,
"grad_norm": 20.48813819885254,
"learning_rate": 2.119723272750379e-06,
"loss": 2.6911,
"step": 650
},
{
"epoch": 2.133877551020408,
"grad_norm": 22.411460876464844,
"learning_rate": 2.048280867121722e-06,
"loss": 2.6595,
"step": 655
},
{
"epoch": 2.150204081632653,
"grad_norm": 22.48036003112793,
"learning_rate": 1.9777517345988057e-06,
"loss": 2.9054,
"step": 660
},
{
"epoch": 2.166530612244898,
"grad_norm": 20.426422119140625,
"learning_rate": 1.908157697162993e-06,
"loss": 2.7907,
"step": 665
},
{
"epoch": 2.182857142857143,
"grad_norm": 23.045915603637695,
"learning_rate": 1.8395202874736752e-06,
"loss": 2.6765,
"step": 670
},
{
"epoch": 2.199183673469388,
"grad_norm": 20.717870712280273,
"learning_rate": 1.771860742205988e-06,
"loss": 2.7119,
"step": 675
},
{
"epoch": 2.2155102040816326,
"grad_norm": 18.947322845458984,
"learning_rate": 1.7051999954801058e-06,
"loss": 2.6826,
"step": 680
},
{
"epoch": 2.2318367346938777,
"grad_norm": 19.087261199951172,
"learning_rate": 1.6395586723841328e-06,
"loss": 2.8017,
"step": 685
},
{
"epoch": 2.2481632653061223,
"grad_norm": 22.97345733642578,
"learning_rate": 1.5749570825926437e-06,
"loss": 2.8442,
"step": 690
},
{
"epoch": 2.2644897959183674,
"grad_norm": 19.933149337768555,
"learning_rate": 1.5114152140827744e-06,
"loss": 2.6767,
"step": 695
},
{
"epoch": 2.280816326530612,
"grad_norm": 19.630264282226562,
"learning_rate": 1.4489527269498749e-06,
"loss": 2.7892,
"step": 700
},
{
"epoch": 2.297142857142857,
"grad_norm": 19.384777069091797,
"learning_rate": 1.3875889473245996e-06,
"loss": 2.6161,
"step": 705
},
{
"epoch": 2.3134693877551022,
"grad_norm": 20.395654678344727,
"learning_rate": 1.3273428613933298e-06,
"loss": 2.7179,
"step": 710
},
{
"epoch": 2.329795918367347,
"grad_norm": 22.20563316345215,
"learning_rate": 1.26823310952377e-06,
"loss": 2.8552,
"step": 715
},
{
"epoch": 2.346122448979592,
"grad_norm": 19.800020217895508,
"learning_rate": 1.2102779804975574e-06,
"loss": 2.5822,
"step": 720
},
{
"epoch": 2.3624489795918366,
"grad_norm": 20.643033981323242,
"learning_rate": 1.1534954058516357e-06,
"loss": 2.677,
"step": 725
},
{
"epoch": 2.3787755102040817,
"grad_norm": 21.90210723876953,
"learning_rate": 1.0979029543301718e-06,
"loss": 2.6476,
"step": 730
},
{
"epoch": 2.3951020408163264,
"grad_norm": 19.81829833984375,
"learning_rate": 1.0435178264487205e-06,
"loss": 2.6254,
"step": 735
},
{
"epoch": 2.4114285714285715,
"grad_norm": 20.124658584594727,
"learning_rate": 9.903568491723176e-07,
"loss": 2.8082,
"step": 740
},
{
"epoch": 2.427755102040816,
"grad_norm": 22.162952423095703,
"learning_rate": 9.384364707091559e-07,
"loss": 2.7515,
"step": 745
},
{
"epoch": 2.444081632653061,
"grad_norm": 20.095272064208984,
"learning_rate": 8.877727554214432e-07,
"loss": 2.5856,
"step": 750
},
{
"epoch": 2.4604081632653063,
"grad_norm": 22.24507713317871,
"learning_rate": 8.383813788550326e-07,
"loss": 2.6821,
"step": 755
},
{
"epoch": 2.476734693877551,
"grad_norm": 20.937358856201172,
"learning_rate": 7.902776228893444e-07,
"loss": 2.5446,
"step": 760
},
{
"epoch": 2.493061224489796,
"grad_norm": 21.72479248046875,
"learning_rate": 7.434763710090991e-07,
"loss": 2.6735,
"step": 765
},
{
"epoch": 2.5093877551020407,
"grad_norm": 21.214231491088867,
"learning_rate": 6.979921036993042e-07,
"loss": 2.7004,
"step": 770
},
{
"epoch": 2.525714285714286,
"grad_norm": 22.42154312133789,
"learning_rate": 6.538388939649442e-07,
"loss": 2.6953,
"step": 775
},
{
"epoch": 2.5420408163265305,
"grad_norm": 19.154542922973633,
"learning_rate": 6.110304029767372e-07,
"loss": 2.6523,
"step": 780
},
{
"epoch": 2.5583673469387755,
"grad_norm": 22.38195037841797,
"learning_rate": 5.695798758443133e-07,
"loss": 2.87,
"step": 785
},
{
"epoch": 2.57469387755102,
"grad_norm": 21.486236572265625,
"learning_rate": 5.295001375181336e-07,
"loss": 2.521,
"step": 790
},
{
"epoch": 2.5910204081632653,
"grad_norm": 20.594417572021484,
"learning_rate": 4.908035888213964e-07,
"loss": 2.4475,
"step": 795
},
{
"epoch": 2.6073469387755104,
"grad_norm": 21.12630844116211,
"learning_rate": 4.5350220261317633e-07,
"loss": 2.7235,
"step": 800
},
{
"epoch": 2.623673469387755,
"grad_norm": 20.065223693847656,
"learning_rate": 4.176075200839791e-07,
"loss": 2.6877,
"step": 805
},
{
"epoch": 2.64,
"grad_norm": 18.989500045776367,
"learning_rate": 3.8313064718485116e-07,
"loss": 2.5682,
"step": 810
},
{
"epoch": 2.6563265306122448,
"grad_norm": 20.19880485534668,
"learning_rate": 3.500822511911578e-07,
"loss": 2.6342,
"step": 815
},
{
"epoch": 2.67265306122449,
"grad_norm": 20.03253746032715,
"learning_rate": 3.1847255740208636e-07,
"loss": 2.7869,
"step": 820
},
{
"epoch": 2.688979591836735,
"grad_norm": 21.833900451660156,
"learning_rate": 2.8831134597689604e-07,
"loss": 2.6605,
"step": 825
},
{
"epoch": 2.7053061224489796,
"grad_norm": 21.995954513549805,
"learning_rate": 2.5960794890891093e-07,
"loss": 2.6093,
"step": 830
},
{
"epoch": 2.7216326530612243,
"grad_norm": 20.3529052734375,
"learning_rate": 2.3237124713815285e-07,
"loss": 2.6319,
"step": 835
},
{
"epoch": 2.7379591836734694,
"grad_norm": 20.60895347595215,
"learning_rate": 2.066096678035523e-07,
"loss": 2.4488,
"step": 840
},
{
"epoch": 2.7542857142857144,
"grad_norm": 22.272764205932617,
"learning_rate": 1.823311816355583e-07,
"loss": 2.8126,
"step": 845
},
{
"epoch": 2.770612244897959,
"grad_norm": 21.894479751586914,
"learning_rate": 1.5954330048996326e-07,
"loss": 2.8516,
"step": 850
},
{
"epoch": 2.786938775510204,
"grad_norm": 21.463518142700195,
"learning_rate": 1.3825307502370487e-07,
"loss": 2.6876,
"step": 855
},
{
"epoch": 2.803265306122449,
"grad_norm": 20.55864715576172,
"learning_rate": 1.1846709251337129e-07,
"loss": 2.5887,
"step": 860
},
{
"epoch": 2.819591836734694,
"grad_norm": 21.697608947753906,
"learning_rate": 1.0019147481706626e-07,
"loss": 2.553,
"step": 865
},
{
"epoch": 2.835918367346939,
"grad_norm": 22.482036590576172,
"learning_rate": 8.343187648028772e-08,
"loss": 2.6056,
"step": 870
},
{
"epoch": 2.8522448979591837,
"grad_norm": 24.521743774414062,
"learning_rate": 6.819348298638839e-08,
"loss": 2.9502,
"step": 875
},
{
"epoch": 2.8685714285714283,
"grad_norm": 22.869632720947266,
"learning_rate": 5.448100915216636e-08,
"loss": 2.5782,
"step": 880
},
{
"epoch": 2.8848979591836734,
"grad_norm": 19.174013137817383,
"learning_rate": 4.2298697669084785e-08,
"loss": 2.6187,
"step": 885
},
{
"epoch": 2.9012244897959185,
"grad_norm": 21.829057693481445,
"learning_rate": 3.165031779056582e-08,
"loss": 2.7208,
"step": 890
},
{
"epoch": 2.917551020408163,
"grad_norm": 24.281612396240234,
"learning_rate": 2.2539164165770178e-08,
"loss": 2.6575,
"step": 895
},
{
"epoch": 2.9338775510204083,
"grad_norm": 20.772991180419922,
"learning_rate": 1.4968055820218653e-08,
"loss": 2.7884,
"step": 900
},
{
"epoch": 2.950204081632653,
"grad_norm": 19.215791702270508,
"learning_rate": 8.939335283577599e-09,
"loss": 2.3742,
"step": 905
},
{
"epoch": 2.966530612244898,
"grad_norm": 20.374731063842773,
"learning_rate": 4.4548678648681506e-09,
"loss": 2.5861,
"step": 910
},
{
"epoch": 2.982857142857143,
"grad_norm": 21.597257614135742,
"learning_rate": 1.51604107533454e-09,
"loss": 2.6382,
"step": 915
},
{
"epoch": 2.9991836734693877,
"grad_norm": 19.170747756958008,
"learning_rate": 1.237641991425065e-10,
"loss": 2.5622,
"step": 920
}
],
"logging_steps": 5,
"max_steps": 921,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.906371548002648e+18,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}