v10_balanced_core_lr1e5_ep2 / trainer_state.json
lldois's picture
Upload v10_balanced_core_lr1e5_ep2 model without checkpoints
218e9cf verified
Raw
History Blame Contribute Delete
43.6 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 1226,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.008159934720522236,
"grad_norm": 43.75,
"learning_rate": 1.0810810810810812e-06,
"loss": 2.7381839752197266,
"step": 5
},
{
"epoch": 0.016319869441044473,
"grad_norm": 38.0,
"learning_rate": 2.432432432432433e-06,
"loss": 2.7479536056518556,
"step": 10
},
{
"epoch": 0.02447980416156671,
"grad_norm": 23.375,
"learning_rate": 3.7837837837837844e-06,
"loss": 2.626136589050293,
"step": 15
},
{
"epoch": 0.032639738882088945,
"grad_norm": 15.75,
"learning_rate": 5.135135135135135e-06,
"loss": 2.284735679626465,
"step": 20
},
{
"epoch": 0.04079967360261118,
"grad_norm": 8.875,
"learning_rate": 6.486486486486487e-06,
"loss": 2.0709344863891603,
"step": 25
},
{
"epoch": 0.04895960832313342,
"grad_norm": 6.09375,
"learning_rate": 7.837837837837838e-06,
"loss": 2.1681032180786133,
"step": 30
},
{
"epoch": 0.05711954304365565,
"grad_norm": 6.375,
"learning_rate": 9.189189189189191e-06,
"loss": 2.0407644271850587,
"step": 35
},
{
"epoch": 0.06527947776417789,
"grad_norm": 4.46875,
"learning_rate": 9.999930187205018e-06,
"loss": 2.01834716796875,
"step": 40
},
{
"epoch": 0.07343941248470012,
"grad_norm": 4.8125,
"learning_rate": 9.999144815650432e-06,
"loss": 2.094622802734375,
"step": 45
},
{
"epoch": 0.08159934720522236,
"grad_norm": 5.40625,
"learning_rate": 9.997486944075234e-06,
"loss": 2.1376014709472657,
"step": 50
},
{
"epoch": 0.08975928192574459,
"grad_norm": 3.453125,
"learning_rate": 9.994956861827514e-06,
"loss": 1.8287012100219726,
"step": 55
},
{
"epoch": 0.09791921664626684,
"grad_norm": 2.8125,
"learning_rate": 9.991555010482156e-06,
"loss": 1.9681983947753907,
"step": 60
},
{
"epoch": 0.10607915136678907,
"grad_norm": 2.828125,
"learning_rate": 9.987281983763789e-06,
"loss": 1.7944761276245118,
"step": 65
},
{
"epoch": 0.1142390860873113,
"grad_norm": 4.625,
"learning_rate": 9.98213852744314e-06,
"loss": 1.7515581130981446,
"step": 70
},
{
"epoch": 0.12239902080783353,
"grad_norm": 3.078125,
"learning_rate": 9.976125539206898e-06,
"loss": 1.7744415283203125,
"step": 75
},
{
"epoch": 0.13055895552835578,
"grad_norm": 2.859375,
"learning_rate": 9.969244068501031e-06,
"loss": 1.712571144104004,
"step": 80
},
{
"epoch": 0.138718890248878,
"grad_norm": 3.890625,
"learning_rate": 9.961495316347625e-06,
"loss": 1.6923364639282226,
"step": 85
},
{
"epoch": 0.14687882496940025,
"grad_norm": 2.890625,
"learning_rate": 9.952880635135266e-06,
"loss": 1.6826797485351563,
"step": 90
},
{
"epoch": 0.15503875968992248,
"grad_norm": 2.5,
"learning_rate": 9.943401528383025e-06,
"loss": 1.6912731170654296,
"step": 95
},
{
"epoch": 0.1631986944104447,
"grad_norm": 8.5,
"learning_rate": 9.933059650478026e-06,
"loss": 1.5941668510437013,
"step": 100
},
{
"epoch": 0.17135862913096694,
"grad_norm": 2.46875,
"learning_rate": 9.921856806386723e-06,
"loss": 1.7211767196655274,
"step": 105
},
{
"epoch": 0.17951856385148918,
"grad_norm": 3.234375,
"learning_rate": 9.909794951339872e-06,
"loss": 1.604559326171875,
"step": 110
},
{
"epoch": 0.18767849857201144,
"grad_norm": 2.53125,
"learning_rate": 9.89687619049129e-06,
"loss": 1.6824333190917968,
"step": 115
},
{
"epoch": 0.19583843329253367,
"grad_norm": 3.703125,
"learning_rate": 9.883102778550434e-06,
"loss": 1.6192102432250977,
"step": 120
},
{
"epoch": 0.2039983680130559,
"grad_norm": 2.734375,
"learning_rate": 9.868477119388897e-06,
"loss": 1.6975128173828125,
"step": 125
},
{
"epoch": 0.21215830273357814,
"grad_norm": 4.125,
"learning_rate": 9.853001765620851e-06,
"loss": 1.634078025817871,
"step": 130
},
{
"epoch": 0.22031823745410037,
"grad_norm": 2.609375,
"learning_rate": 9.836679418157551e-06,
"loss": 1.6162107467651368,
"step": 135
},
{
"epoch": 0.2284781721746226,
"grad_norm": 3.078125,
"learning_rate": 9.819512925735934e-06,
"loss": 1.5213428497314454,
"step": 140
},
{
"epoch": 0.23663810689514483,
"grad_norm": 3.1875,
"learning_rate": 9.801505284421437e-06,
"loss": 1.7047014236450195,
"step": 145
},
{
"epoch": 0.24479804161566707,
"grad_norm": 3.171875,
"learning_rate": 9.782659637085089e-06,
"loss": 1.6014438629150392,
"step": 150
},
{
"epoch": 0.2529579763361893,
"grad_norm": 2.640625,
"learning_rate": 9.762979272854991e-06,
"loss": 1.605925941467285,
"step": 155
},
{
"epoch": 0.26111791105671156,
"grad_norm": 2.953125,
"learning_rate": 9.742467626542262e-06,
"loss": 1.6351459503173829,
"step": 160
},
{
"epoch": 0.2692778457772338,
"grad_norm": 2.359375,
"learning_rate": 9.721128278041562e-06,
"loss": 1.5994914054870606,
"step": 165
},
{
"epoch": 0.277437780497756,
"grad_norm": 4.125,
"learning_rate": 9.698964951706294e-06,
"loss": 1.6617794036865234,
"step": 170
},
{
"epoch": 0.28559771521827826,
"grad_norm": 2.546875,
"learning_rate": 9.675981515698591e-06,
"loss": 1.5865089416503906,
"step": 175
},
{
"epoch": 0.2937576499388005,
"grad_norm": 2.515625,
"learning_rate": 9.652181981314216e-06,
"loss": 1.6984281539916992,
"step": 180
},
{
"epoch": 0.3019175846593227,
"grad_norm": 3.21875,
"learning_rate": 9.62757050228245e-06,
"loss": 1.6521415710449219,
"step": 185
},
{
"epoch": 0.31007751937984496,
"grad_norm": 2.359375,
"learning_rate": 9.602151374041173e-06,
"loss": 1.6062736511230469,
"step": 190
},
{
"epoch": 0.3182374541003672,
"grad_norm": 3.0,
"learning_rate": 9.575929032987148e-06,
"loss": 1.6013092041015624,
"step": 195
},
{
"epoch": 0.3263973888208894,
"grad_norm": 3.03125,
"learning_rate": 9.548908055701762e-06,
"loss": 1.568859577178955,
"step": 200
},
{
"epoch": 0.33455732354141166,
"grad_norm": 4.78125,
"learning_rate": 9.521093158152271e-06,
"loss": 1.6711809158325195,
"step": 205
},
{
"epoch": 0.3427172582619339,
"grad_norm": 4.03125,
"learning_rate": 9.492489194868714e-06,
"loss": 1.553749179840088,
"step": 210
},
{
"epoch": 0.3508771929824561,
"grad_norm": 3.5,
"learning_rate": 9.463101158096657e-06,
"loss": 1.6366708755493165,
"step": 215
},
{
"epoch": 0.35903712770297835,
"grad_norm": 2.546875,
"learning_rate": 9.432934176925905e-06,
"loss": 1.6099456787109374,
"step": 220
},
{
"epoch": 0.3671970624235006,
"grad_norm": 3.03125,
"learning_rate": 9.401993516395312e-06,
"loss": 1.5125322341918945,
"step": 225
},
{
"epoch": 0.3753569971440229,
"grad_norm": 3.015625,
"learning_rate": 9.370284576573876e-06,
"loss": 1.5324077606201172,
"step": 230
},
{
"epoch": 0.3835169318645451,
"grad_norm": 3.09375,
"learning_rate": 9.337812891618275e-06,
"loss": 1.5691267013549806,
"step": 235
},
{
"epoch": 0.39167686658506734,
"grad_norm": 2.6875,
"learning_rate": 9.304584128806979e-06,
"loss": 1.5658173561096191,
"step": 240
},
{
"epoch": 0.3998368013055896,
"grad_norm": 2.625,
"learning_rate": 9.27060408755115e-06,
"loss": 1.60058650970459,
"step": 245
},
{
"epoch": 0.4079967360261118,
"grad_norm": 3.96875,
"learning_rate": 9.235878698382464e-06,
"loss": 1.5558917999267579,
"step": 250
},
{
"epoch": 0.41615667074663404,
"grad_norm": 3.890625,
"learning_rate": 9.200414021918068e-06,
"loss": 1.569556713104248,
"step": 255
},
{
"epoch": 0.42431660546715627,
"grad_norm": 2.859375,
"learning_rate": 9.164216247802802e-06,
"loss": 1.585373306274414,
"step": 260
},
{
"epoch": 0.4324765401876785,
"grad_norm": 4.09375,
"learning_rate": 9.127291693628945e-06,
"loss": 1.4009552001953125,
"step": 265
},
{
"epoch": 0.44063647490820074,
"grad_norm": 3.328125,
"learning_rate": 9.089646803833589e-06,
"loss": 1.6112878799438477,
"step": 270
},
{
"epoch": 0.44879640962872297,
"grad_norm": 3.859375,
"learning_rate": 9.051288148573895e-06,
"loss": 1.555915641784668,
"step": 275
},
{
"epoch": 0.4569563443492452,
"grad_norm": 2.359375,
"learning_rate": 9.01222242258041e-06,
"loss": 1.5131331443786622,
"step": 280
},
{
"epoch": 0.46511627906976744,
"grad_norm": 2.53125,
"learning_rate": 8.972456443988634e-06,
"loss": 1.4798956871032716,
"step": 285
},
{
"epoch": 0.47327621379028967,
"grad_norm": 3.0625,
"learning_rate": 8.93199715314905e-06,
"loss": 1.5639562606811523,
"step": 290
},
{
"epoch": 0.4814361485108119,
"grad_norm": 3.078125,
"learning_rate": 8.890851611415831e-06,
"loss": 1.5197336196899414,
"step": 295
},
{
"epoch": 0.48959608323133413,
"grad_norm": 3.03125,
"learning_rate": 8.849026999914412e-06,
"loss": 1.623777198791504,
"step": 300
},
{
"epoch": 0.49775601795185637,
"grad_norm": 3.453125,
"learning_rate": 8.806530618288185e-06,
"loss": 1.5564375877380372,
"step": 305
},
{
"epoch": 0.5059159526723787,
"grad_norm": 2.90625,
"learning_rate": 8.763369883424478e-06,
"loss": 1.4996400833129884,
"step": 310
},
{
"epoch": 0.5140758873929009,
"grad_norm": 2.78125,
"learning_rate": 8.719552328160095e-06,
"loss": 1.622853660583496,
"step": 315
},
{
"epoch": 0.5222358221134231,
"grad_norm": 4.09375,
"learning_rate": 8.675085599966611e-06,
"loss": 1.442308235168457,
"step": 320
},
{
"epoch": 0.5303957568339454,
"grad_norm": 2.78125,
"learning_rate": 8.629977459615655e-06,
"loss": 1.4956160545349122,
"step": 325
},
{
"epoch": 0.5385556915544676,
"grad_norm": 2.765625,
"learning_rate": 8.584235779824433e-06,
"loss": 1.6151552200317383,
"step": 330
},
{
"epoch": 0.5467156262749898,
"grad_norm": 4.0,
"learning_rate": 8.53786854388169e-06,
"loss": 1.4596601486206056,
"step": 335
},
{
"epoch": 0.554875560995512,
"grad_norm": 4.03125,
"learning_rate": 8.4908838442544e-06,
"loss": 1.45312442779541,
"step": 340
},
{
"epoch": 0.5630354957160343,
"grad_norm": 4.0,
"learning_rate": 8.443289881175384e-06,
"loss": 1.479482650756836,
"step": 345
},
{
"epoch": 0.5711954304365565,
"grad_norm": 4.0625,
"learning_rate": 8.395094961212127e-06,
"loss": 1.4872551918029786,
"step": 350
},
{
"epoch": 0.5793553651570787,
"grad_norm": 2.859375,
"learning_rate": 8.346307495817028e-06,
"loss": 1.4293831825256347,
"step": 355
},
{
"epoch": 0.587515299877601,
"grad_norm": 2.71875,
"learning_rate": 8.296935999859357e-06,
"loss": 1.5046165466308594,
"step": 360
},
{
"epoch": 0.5956752345981232,
"grad_norm": 3.171875,
"learning_rate": 8.24698909013915e-06,
"loss": 1.4456731796264648,
"step": 365
},
{
"epoch": 0.6038351693186454,
"grad_norm": 2.78125,
"learning_rate": 8.19647548388333e-06,
"loss": 1.5321898460388184,
"step": 370
},
{
"epoch": 0.6119951040391677,
"grad_norm": 2.640625,
"learning_rate": 8.145403997224277e-06,
"loss": 1.4732604026794434,
"step": 375
},
{
"epoch": 0.6201550387596899,
"grad_norm": 2.75,
"learning_rate": 8.093783543661162e-06,
"loss": 1.5284388542175293,
"step": 380
},
{
"epoch": 0.6283149734802121,
"grad_norm": 2.84375,
"learning_rate": 8.04162313250427e-06,
"loss": 1.4671659469604492,
"step": 385
},
{
"epoch": 0.6364749082007344,
"grad_norm": 3.171875,
"learning_rate": 7.988931867302605e-06,
"loss": 1.4927037239074707,
"step": 390
},
{
"epoch": 0.6446348429212566,
"grad_norm": 2.5625,
"learning_rate": 7.935718944255049e-06,
"loss": 1.551847267150879,
"step": 395
},
{
"epoch": 0.6527947776417788,
"grad_norm": 3.046875,
"learning_rate": 7.881993650605356e-06,
"loss": 1.5500238418579102,
"step": 400
},
{
"epoch": 0.6609547123623011,
"grad_norm": 2.78125,
"learning_rate": 7.827765363021244e-06,
"loss": 1.5144458770751954,
"step": 405
},
{
"epoch": 0.6691146470828233,
"grad_norm": 3.4375,
"learning_rate": 7.773043545957892e-06,
"loss": 1.372291660308838,
"step": 410
},
{
"epoch": 0.6772745818033455,
"grad_norm": 3.65625,
"learning_rate": 7.717837750006106e-06,
"loss": 1.4556828498840333,
"step": 415
},
{
"epoch": 0.6854345165238678,
"grad_norm": 2.125,
"learning_rate": 7.662157610225458e-06,
"loss": 1.5917230606079102,
"step": 420
},
{
"epoch": 0.69359445124439,
"grad_norm": 3.375,
"learning_rate": 7.606012844462678e-06,
"loss": 1.5001279830932617,
"step": 425
},
{
"epoch": 0.7017543859649122,
"grad_norm": 3.234375,
"learning_rate": 7.549413251655604e-06,
"loss": 1.4177436828613281,
"step": 430
},
{
"epoch": 0.7099143206854345,
"grad_norm": 3.203125,
"learning_rate": 7.4923687101229695e-06,
"loss": 1.4507540702819823,
"step": 435
},
{
"epoch": 0.7180742554059567,
"grad_norm": 3.015625,
"learning_rate": 7.4348891758403464e-06,
"loss": 1.6129571914672851,
"step": 440
},
{
"epoch": 0.7262341901264789,
"grad_norm": 3.875,
"learning_rate": 7.376984680702529e-06,
"loss": 1.485245418548584,
"step": 445
},
{
"epoch": 0.7343941248470012,
"grad_norm": 4.28125,
"learning_rate": 7.318665330772665e-06,
"loss": 1.3869742393493651,
"step": 450
},
{
"epoch": 0.7425540595675234,
"grad_norm": 2.78125,
"learning_rate": 7.259941304518453e-06,
"loss": 1.5055102348327636,
"step": 455
},
{
"epoch": 0.7507139942880457,
"grad_norm": 3.40625,
"learning_rate": 7.200822851035687e-06,
"loss": 1.6238336563110352,
"step": 460
},
{
"epoch": 0.758873929008568,
"grad_norm": 3.203125,
"learning_rate": 7.141320288259495e-06,
"loss": 1.5736274719238281,
"step": 465
},
{
"epoch": 0.7670338637290902,
"grad_norm": 2.484375,
"learning_rate": 7.081444001163535e-06,
"loss": 1.5016008377075196,
"step": 470
},
{
"epoch": 0.7751937984496124,
"grad_norm": 3.140625,
"learning_rate": 7.021204439947524e-06,
"loss": 1.549778366088867,
"step": 475
},
{
"epoch": 0.7833537331701347,
"grad_norm": 2.703125,
"learning_rate": 6.960612118213357e-06,
"loss": 1.5195813179016113,
"step": 480
},
{
"epoch": 0.7915136678906569,
"grad_norm": 3.625,
"learning_rate": 6.899677611130169e-06,
"loss": 1.460670280456543,
"step": 485
},
{
"epoch": 0.7996736026111791,
"grad_norm": 2.09375,
"learning_rate": 6.838411553588656e-06,
"loss": 1.4845394134521483,
"step": 490
},
{
"epoch": 0.8078335373317014,
"grad_norm": 3.15625,
"learning_rate": 6.776824638344967e-06,
"loss": 1.4958683013916017,
"step": 495
},
{
"epoch": 0.8159934720522236,
"grad_norm": 3.8125,
"learning_rate": 6.714927614154498e-06,
"loss": 1.4073028564453125,
"step": 500
},
{
"epoch": 0.8241534067727458,
"grad_norm": 2.375,
"learning_rate": 6.652731283895917e-06,
"loss": 1.5039354324340821,
"step": 505
},
{
"epoch": 0.8323133414932681,
"grad_norm": 3.390625,
"learning_rate": 6.590246502685738e-06,
"loss": 1.4758033752441406,
"step": 510
},
{
"epoch": 0.8404732762137903,
"grad_norm": 4.5625,
"learning_rate": 6.5274841759837725e-06,
"loss": 1.4331993103027343,
"step": 515
},
{
"epoch": 0.8486332109343125,
"grad_norm": 3.03125,
"learning_rate": 6.464455257689809e-06,
"loss": 1.399909210205078,
"step": 520
},
{
"epoch": 0.8567931456548348,
"grad_norm": 2.296875,
"learning_rate": 6.4011707482318295e-06,
"loss": 1.546139621734619,
"step": 525
},
{
"epoch": 0.864953080375357,
"grad_norm": 2.90625,
"learning_rate": 6.337641692646106e-06,
"loss": 1.4532946586608886,
"step": 530
},
{
"epoch": 0.8731130150958792,
"grad_norm": 3.125,
"learning_rate": 6.273879178649505e-06,
"loss": 1.5350307464599608,
"step": 535
},
{
"epoch": 0.8812729498164015,
"grad_norm": 3.15625,
"learning_rate": 6.209894334704358e-06,
"loss": 1.414911937713623,
"step": 540
},
{
"epoch": 0.8894328845369237,
"grad_norm": 2.796875,
"learning_rate": 6.145698328076215e-06,
"loss": 1.4744022369384766,
"step": 545
},
{
"epoch": 0.8975928192574459,
"grad_norm": 3.46875,
"learning_rate": 6.0813023628848144e-06,
"loss": 1.5095841407775878,
"step": 550
},
{
"epoch": 0.9057527539779682,
"grad_norm": 2.71875,
"learning_rate": 6.0167176781486326e-06,
"loss": 1.4840353012084961,
"step": 555
},
{
"epoch": 0.9139126886984904,
"grad_norm": 2.625,
"learning_rate": 5.951955545823342e-06,
"loss": 1.4900919914245605,
"step": 560
},
{
"epoch": 0.9220726234190126,
"grad_norm": 2.796875,
"learning_rate": 5.88702726883452e-06,
"loss": 1.580379104614258,
"step": 565
},
{
"epoch": 0.9302325581395349,
"grad_norm": 2.828125,
"learning_rate": 5.821944179104941e-06,
"loss": 1.3942214012145997,
"step": 570
},
{
"epoch": 0.9383924928600571,
"grad_norm": 4.15625,
"learning_rate": 5.756717635576834e-06,
"loss": 1.4603987693786622,
"step": 575
},
{
"epoch": 0.9465524275805793,
"grad_norm": 3.109375,
"learning_rate": 5.69135902222939e-06,
"loss": 1.5463579177856446,
"step": 580
},
{
"epoch": 0.9547123623011016,
"grad_norm": 2.921875,
"learning_rate": 5.62587974609193e-06,
"loss": 1.4487386703491212,
"step": 585
},
{
"epoch": 0.9628722970216238,
"grad_norm": 3.15625,
"learning_rate": 5.560291235253037e-06,
"loss": 1.5094460487365722,
"step": 590
},
{
"epoch": 0.971032231742146,
"grad_norm": 2.703125,
"learning_rate": 5.494604936866007e-06,
"loss": 1.4954254150390625,
"step": 595
},
{
"epoch": 0.9791921664626683,
"grad_norm": 3.78125,
"learning_rate": 5.428832315150984e-06,
"loss": 1.4877737045288086,
"step": 600
},
{
"epoch": 0.9873521011831905,
"grad_norm": 3.78125,
"learning_rate": 5.362984849394106e-06,
"loss": 1.3748371124267578,
"step": 605
},
{
"epoch": 0.9955120359037127,
"grad_norm": 2.703125,
"learning_rate": 5.297074031944039e-06,
"loss": 1.45643310546875,
"step": 610
},
{
"epoch": 1.003263973888209,
"grad_norm": 2.484375,
"learning_rate": 5.2311113662062065e-06,
"loss": 1.395163631439209,
"step": 615
},
{
"epoch": 1.011423908608731,
"grad_norm": 4.15625,
"learning_rate": 5.1651083646350995e-06,
"loss": 1.5137823104858399,
"step": 620
},
{
"epoch": 1.0195838433292534,
"grad_norm": 2.6875,
"learning_rate": 5.099076546725029e-06,
"loss": 1.5231605529785157,
"step": 625
},
{
"epoch": 1.0277437780497756,
"grad_norm": 2.78125,
"learning_rate": 5.033027436999609e-06,
"loss": 1.4491164207458496,
"step": 630
},
{
"epoch": 1.035903712770298,
"grad_norm": 2.5625,
"learning_rate": 4.966972563000392e-06,
"loss": 1.5070039749145507,
"step": 635
},
{
"epoch": 1.04406364749082,
"grad_norm": 2.96875,
"learning_rate": 4.9009234532749725e-06,
"loss": 1.4665255546569824,
"step": 640
},
{
"epoch": 1.0522235822113424,
"grad_norm": 3.9375,
"learning_rate": 4.8348916353649004e-06,
"loss": 1.6022050857543946,
"step": 645
},
{
"epoch": 1.0603835169318645,
"grad_norm": 3.140625,
"learning_rate": 4.768888633793797e-06,
"loss": 1.4543895721435547,
"step": 650
},
{
"epoch": 1.0685434516523868,
"grad_norm": 2.671875,
"learning_rate": 4.702925968055962e-06,
"loss": 1.4940269470214844,
"step": 655
},
{
"epoch": 1.076703386372909,
"grad_norm": 3.65625,
"learning_rate": 4.6370151506058945e-06,
"loss": 1.4373529434204102,
"step": 660
},
{
"epoch": 1.0848633210934313,
"grad_norm": 3.71875,
"learning_rate": 4.571167684849019e-06,
"loss": 1.4962950706481934,
"step": 665
},
{
"epoch": 1.0930232558139534,
"grad_norm": 3.453125,
"learning_rate": 4.5053950631339935e-06,
"loss": 1.4017728805541991,
"step": 670
},
{
"epoch": 1.1011831905344758,
"grad_norm": 3.453125,
"learning_rate": 4.439708764746964e-06,
"loss": 1.5669015884399413,
"step": 675
},
{
"epoch": 1.109343125254998,
"grad_norm": 4.03125,
"learning_rate": 4.374120253908072e-06,
"loss": 1.4162152290344239,
"step": 680
},
{
"epoch": 1.1175030599755202,
"grad_norm": 2.453125,
"learning_rate": 4.3086409777706115e-06,
"loss": 1.5029624938964843,
"step": 685
},
{
"epoch": 1.1256629946960424,
"grad_norm": 4.4375,
"learning_rate": 4.243282364423167e-06,
"loss": 1.4658592224121094,
"step": 690
},
{
"epoch": 1.1338229294165647,
"grad_norm": 4.25,
"learning_rate": 4.17805582089506e-06,
"loss": 1.4464892387390136,
"step": 695
},
{
"epoch": 1.1419828641370868,
"grad_norm": 2.953125,
"learning_rate": 4.112972731165482e-06,
"loss": 1.3933393478393554,
"step": 700
},
{
"epoch": 1.1501427988576092,
"grad_norm": 2.875,
"learning_rate": 4.048044454176658e-06,
"loss": 1.507524585723877,
"step": 705
},
{
"epoch": 1.1583027335781313,
"grad_norm": 3.828125,
"learning_rate": 3.983282321851369e-06,
"loss": 1.4756403923034669,
"step": 710
},
{
"epoch": 1.1664626682986536,
"grad_norm": 3.109375,
"learning_rate": 3.918697637115187e-06,
"loss": 1.4435945510864259,
"step": 715
},
{
"epoch": 1.1746226030191758,
"grad_norm": 3.125,
"learning_rate": 3.8543016719237855e-06,
"loss": 1.5925870895385743,
"step": 720
},
{
"epoch": 1.182782537739698,
"grad_norm": 2.578125,
"learning_rate": 3.7901056652956438e-06,
"loss": 1.391953182220459,
"step": 725
},
{
"epoch": 1.1909424724602202,
"grad_norm": 3.265625,
"learning_rate": 3.7261208213504974e-06,
"loss": 1.4485077857971191,
"step": 730
},
{
"epoch": 1.1991024071807426,
"grad_norm": 2.234375,
"learning_rate": 3.662358307353897e-06,
"loss": 1.4522296905517578,
"step": 735
},
{
"epoch": 1.2072623419012647,
"grad_norm": 2.890625,
"learning_rate": 3.598829251768172e-06,
"loss": 1.4614290237426757,
"step": 740
},
{
"epoch": 1.215422276621787,
"grad_norm": 3.765625,
"learning_rate": 3.5355447423101917e-06,
"loss": 1.3943318367004394,
"step": 745
},
{
"epoch": 1.2235822113423094,
"grad_norm": 3.5,
"learning_rate": 3.4725158240162304e-06,
"loss": 1.4987910270690918,
"step": 750
},
{
"epoch": 1.2317421460628315,
"grad_norm": 3.8125,
"learning_rate": 3.4097534973142633e-06,
"loss": 1.4035530090332031,
"step": 755
},
{
"epoch": 1.2399020807833536,
"grad_norm": 3.1875,
"learning_rate": 3.3472687161040827e-06,
"loss": 1.5458099365234375,
"step": 760
},
{
"epoch": 1.248062015503876,
"grad_norm": 2.390625,
"learning_rate": 3.2850723858455034e-06,
"loss": 1.4949901580810547,
"step": 765
},
{
"epoch": 1.2562219502243983,
"grad_norm": 3.328125,
"learning_rate": 3.2231753616550345e-06,
"loss": 1.381500244140625,
"step": 770
},
{
"epoch": 1.2643818849449204,
"grad_norm": 2.90625,
"learning_rate": 3.1615884464113443e-06,
"loss": 1.459078025817871,
"step": 775
},
{
"epoch": 1.2725418196654426,
"grad_norm": 2.28125,
"learning_rate": 3.1003223888698318e-06,
"loss": 1.5301389694213867,
"step": 780
},
{
"epoch": 1.280701754385965,
"grad_norm": 2.796875,
"learning_rate": 3.039387881786644e-06,
"loss": 1.3587870597839355,
"step": 785
},
{
"epoch": 1.2888616891064872,
"grad_norm": 3.0625,
"learning_rate": 2.9787955600524756e-06,
"loss": 1.4352481842041016,
"step": 790
},
{
"epoch": 1.2970216238270094,
"grad_norm": 2.40625,
"learning_rate": 2.9185559988364676e-06,
"loss": 1.5293769836425781,
"step": 795
},
{
"epoch": 1.3051815585475315,
"grad_norm": 2.53125,
"learning_rate": 2.8586797117405063e-06,
"loss": 1.4766444206237792,
"step": 800
},
{
"epoch": 1.3133414932680538,
"grad_norm": 2.640625,
"learning_rate": 2.7991771489643126e-06,
"loss": 1.4699063301086426,
"step": 805
},
{
"epoch": 1.3215014279885762,
"grad_norm": 3.140625,
"learning_rate": 2.740058695481548e-06,
"loss": 1.4705145835876465,
"step": 810
},
{
"epoch": 1.3296613627090983,
"grad_norm": 2.921875,
"learning_rate": 2.6813346692273355e-06,
"loss": 1.5600375175476073,
"step": 815
},
{
"epoch": 1.3378212974296206,
"grad_norm": 4.1875,
"learning_rate": 2.6230153192974728e-06,
"loss": 1.4068735122680665,
"step": 820
},
{
"epoch": 1.3459812321501428,
"grad_norm": 4.03125,
"learning_rate": 2.5651108241596544e-06,
"loss": 1.4527854919433594,
"step": 825
},
{
"epoch": 1.354141166870665,
"grad_norm": 2.734375,
"learning_rate": 2.507631289877032e-06,
"loss": 1.4468353271484375,
"step": 830
},
{
"epoch": 1.3623011015911872,
"grad_norm": 2.53125,
"learning_rate": 2.4505867483443978e-06,
"loss": 1.4296448707580567,
"step": 835
},
{
"epoch": 1.3704610363117096,
"grad_norm": 3.046875,
"learning_rate": 2.3939871555373234e-06,
"loss": 1.4662452697753907,
"step": 840
},
{
"epoch": 1.3786209710322317,
"grad_norm": 4.3125,
"learning_rate": 2.3378423897745432e-06,
"loss": 1.4125295639038087,
"step": 845
},
{
"epoch": 1.386780905752754,
"grad_norm": 2.734375,
"learning_rate": 2.282162249993895e-06,
"loss": 1.4153221130371094,
"step": 850
},
{
"epoch": 1.3949408404732762,
"grad_norm": 4.1875,
"learning_rate": 2.2269564540421092e-06,
"loss": 1.501656436920166,
"step": 855
},
{
"epoch": 1.4031007751937985,
"grad_norm": 2.765625,
"learning_rate": 2.172234636978757e-06,
"loss": 1.5204863548278809,
"step": 860
},
{
"epoch": 1.4112607099143206,
"grad_norm": 4.1875,
"learning_rate": 2.118006349394644e-06,
"loss": 1.364470863342285,
"step": 865
},
{
"epoch": 1.419420644634843,
"grad_norm": 4.125,
"learning_rate": 2.0642810557449535e-06,
"loss": 1.4071884155273438,
"step": 870
},
{
"epoch": 1.427580579355365,
"grad_norm": 2.953125,
"learning_rate": 2.0110681326973975e-06,
"loss": 1.4230575561523438,
"step": 875
},
{
"epoch": 1.4357405140758874,
"grad_norm": 3.109375,
"learning_rate": 1.9583768674957313e-06,
"loss": 1.4786928176879883,
"step": 880
},
{
"epoch": 1.4439004487964096,
"grad_norm": 3.078125,
"learning_rate": 1.9062164563388397e-06,
"loss": 1.4713285446166993,
"step": 885
},
{
"epoch": 1.452060383516932,
"grad_norm": 3.0625,
"learning_rate": 1.8545960027757237e-06,
"loss": 1.458634853363037,
"step": 890
},
{
"epoch": 1.460220318237454,
"grad_norm": 3.109375,
"learning_rate": 1.803524516116673e-06,
"loss": 1.3398704528808594,
"step": 895
},
{
"epoch": 1.4683802529579764,
"grad_norm": 3.109375,
"learning_rate": 1.7530109098608521e-06,
"loss": 1.3795308113098144,
"step": 900
},
{
"epoch": 1.4765401876784985,
"grad_norm": 2.78125,
"learning_rate": 1.703064000140644e-06,
"loss": 1.4607046127319336,
"step": 905
},
{
"epoch": 1.4847001223990208,
"grad_norm": 4.0,
"learning_rate": 1.6536925041829738e-06,
"loss": 1.436344528198242,
"step": 910
},
{
"epoch": 1.4928600571195432,
"grad_norm": 3.796875,
"learning_rate": 1.604905038787875e-06,
"loss": 1.4653143882751465,
"step": 915
},
{
"epoch": 1.5010199918400653,
"grad_norm": 2.375,
"learning_rate": 1.556710118824616e-06,
"loss": 1.430887222290039,
"step": 920
},
{
"epoch": 1.5091799265605874,
"grad_norm": 2.609375,
"learning_rate": 1.509116155745603e-06,
"loss": 1.447705364227295,
"step": 925
},
{
"epoch": 1.5173398612811098,
"grad_norm": 3.046875,
"learning_rate": 1.462131456118312e-06,
"loss": 1.4594793319702148,
"step": 930
},
{
"epoch": 1.525499796001632,
"grad_norm": 3.8125,
"learning_rate": 1.4157642201755688e-06,
"loss": 1.4479350090026855,
"step": 935
},
{
"epoch": 1.5336597307221542,
"grad_norm": 3.75,
"learning_rate": 1.370022540384347e-06,
"loss": 1.471513271331787,
"step": 940
},
{
"epoch": 1.5418196654426763,
"grad_norm": 3.828125,
"learning_rate": 1.3249144000333908e-06,
"loss": 1.4426036834716798,
"step": 945
},
{
"epoch": 1.5499796001631987,
"grad_norm": 2.453125,
"learning_rate": 1.280447671839906e-06,
"loss": 1.4395041465759277,
"step": 950
},
{
"epoch": 1.558139534883721,
"grad_norm": 3.65625,
"learning_rate": 1.2366301165755228e-06,
"loss": 1.3836175918579101,
"step": 955
},
{
"epoch": 1.5662994696042432,
"grad_norm": 4.65625,
"learning_rate": 1.1934693817118155e-06,
"loss": 1.4375165939331054,
"step": 960
},
{
"epoch": 1.5744594043247653,
"grad_norm": 4.21875,
"learning_rate": 1.1509730000855878e-06,
"loss": 1.4456217765808106,
"step": 965
},
{
"epoch": 1.5826193390452876,
"grad_norm": 2.78125,
"learning_rate": 1.1091483885841703e-06,
"loss": 1.5496036529541015,
"step": 970
},
{
"epoch": 1.59077927376581,
"grad_norm": 2.421875,
"learning_rate": 1.0680028468509506e-06,
"loss": 1.539748477935791,
"step": 975
},
{
"epoch": 1.598939208486332,
"grad_norm": 4.5,
"learning_rate": 1.0275435560113677e-06,
"loss": 1.3937344551086426,
"step": 980
},
{
"epoch": 1.6070991432068542,
"grad_norm": 2.9375,
"learning_rate": 9.877775774195913e-07,
"loss": 1.4053834915161132,
"step": 985
},
{
"epoch": 1.6152590779273766,
"grad_norm": 2.5,
"learning_rate": 9.487118514261068e-07,
"loss": 1.5976146697998046,
"step": 990
},
{
"epoch": 1.623419012647899,
"grad_norm": 4.0625,
"learning_rate": 9.10353196166412e-07,
"loss": 1.5028489112854004,
"step": 995
},
{
"epoch": 1.631578947368421,
"grad_norm": 4.34375,
"learning_rate": 8.72708306371055e-07,
"loss": 1.3941530227661132,
"step": 1000
},
{
"epoch": 1.6397388820889431,
"grad_norm": 3.234375,
"learning_rate": 8.357837521971984e-07,
"loss": 1.5182560920715331,
"step": 1005
},
{
"epoch": 1.6478988168094655,
"grad_norm": 2.90625,
"learning_rate": 7.995859780819337e-07,
"loss": 1.4995549201965332,
"step": 1010
},
{
"epoch": 1.6560587515299878,
"grad_norm": 2.484375,
"learning_rate": 7.64121301617537e-07,
"loss": 1.4941126823425293,
"step": 1015
},
{
"epoch": 1.66421868625051,
"grad_norm": 4.03125,
"learning_rate": 7.293959124488526e-07,
"loss": 1.5346342086791993,
"step": 1020
},
{
"epoch": 1.672378620971032,
"grad_norm": 3.28125,
"learning_rate": 6.954158711930225e-07,
"loss": 1.393368911743164,
"step": 1025
},
{
"epoch": 1.6805385556915544,
"grad_norm": 3.453125,
"learning_rate": 6.621871083817266e-07,
"loss": 1.4414457321166991,
"step": 1030
},
{
"epoch": 1.6886984904120768,
"grad_norm": 2.328125,
"learning_rate": 6.297154234261238e-07,
"loss": 1.5497615814208985,
"step": 1035
},
{
"epoch": 1.696858425132599,
"grad_norm": 2.5,
"learning_rate": 5.980064836046901e-07,
"loss": 1.479388427734375,
"step": 1040
},
{
"epoch": 1.7050183598531212,
"grad_norm": 3.140625,
"learning_rate": 5.670658230740961e-07,
"loss": 1.5294727325439452,
"step": 1045
},
{
"epoch": 1.7131782945736433,
"grad_norm": 3.3125,
"learning_rate": 5.368988419033428e-07,
"loss": 1.4593332290649415,
"step": 1050
},
{
"epoch": 1.7213382292941657,
"grad_norm": 2.921875,
"learning_rate": 5.075108051312882e-07,
"loss": 1.4523781776428222,
"step": 1055
},
{
"epoch": 1.729498164014688,
"grad_norm": 2.359375,
"learning_rate": 4.789068418477288e-07,
"loss": 1.538530445098877,
"step": 1060
},
{
"epoch": 1.7376580987352102,
"grad_norm": 2.796875,
"learning_rate": 4.5109194429823756e-07,
"loss": 1.4923004150390624,
"step": 1065
},
{
"epoch": 1.7458180334557323,
"grad_norm": 3.25,
"learning_rate": 4.2407096701285454e-07,
"loss": 1.4749383926391602,
"step": 1070
},
{
"epoch": 1.7539779681762546,
"grad_norm": 3.171875,
"learning_rate": 3.9784862595882877e-07,
"loss": 1.3908852577209472,
"step": 1075
},
{
"epoch": 1.762137902896777,
"grad_norm": 2.46875,
"learning_rate": 3.7242949771754965e-07,
"loss": 1.504295825958252,
"step": 1080
},
{
"epoch": 1.770297837617299,
"grad_norm": 3.625,
"learning_rate": 3.478180186857871e-07,
"loss": 1.3926816940307618,
"step": 1085
},
{
"epoch": 1.7784577723378212,
"grad_norm": 3.59375,
"learning_rate": 3.240184843014088e-07,
"loss": 1.3842317581176757,
"step": 1090
},
{
"epoch": 1.7866177070583436,
"grad_norm": 2.09375,
"learning_rate": 3.0103504829370744e-07,
"loss": 1.5405694007873536,
"step": 1095
},
{
"epoch": 1.794777641778866,
"grad_norm": 3.1875,
"learning_rate": 2.788717219584386e-07,
"loss": 1.4555512428283692,
"step": 1100
},
{
"epoch": 1.802937576499388,
"grad_norm": 2.15625,
"learning_rate": 2.575323734577384e-07,
"loss": 1.4967559814453124,
"step": 1105
},
{
"epoch": 1.8110975112199101,
"grad_norm": 4.1875,
"learning_rate": 2.370207271450098e-07,
"loss": 1.3833650588989257,
"step": 1110
},
{
"epoch": 1.8192574459404325,
"grad_norm": 3.171875,
"learning_rate": 2.173403629149118e-07,
"loss": 1.5348877906799316,
"step": 1115
},
{
"epoch": 1.8274173806609548,
"grad_norm": 3.125,
"learning_rate": 1.9849471557856438e-07,
"loss": 1.4135624885559082,
"step": 1120
},
{
"epoch": 1.835577315381477,
"grad_norm": 2.859375,
"learning_rate": 1.8048707426406742e-07,
"loss": 1.4719861984252929,
"step": 1125
},
{
"epoch": 1.843737250101999,
"grad_norm": 4.4375,
"learning_rate": 1.6332058184245047e-07,
"loss": 1.3041519165039062,
"step": 1130
},
{
"epoch": 1.8518971848225214,
"grad_norm": 4.09375,
"learning_rate": 1.4699823437914985e-07,
"loss": 1.412525749206543,
"step": 1135
},
{
"epoch": 1.8600571195430438,
"grad_norm": 2.171875,
"learning_rate": 1.3152288061110518e-07,
"loss": 1.397139549255371,
"step": 1140
},
{
"epoch": 1.8682170542635659,
"grad_norm": 2.5,
"learning_rate": 1.1689722144956672e-07,
"loss": 1.5287254333496094,
"step": 1145
},
{
"epoch": 1.876376988984088,
"grad_norm": 2.859375,
"learning_rate": 1.031238095087106e-07,
"loss": 1.4039966583251953,
"step": 1150
},
{
"epoch": 1.8845369237046103,
"grad_norm": 3.78125,
"learning_rate": 9.020504866012747e-08,
"loss": 1.400966262817383,
"step": 1155
},
{
"epoch": 1.8926968584251327,
"grad_norm": 3.234375,
"learning_rate": 7.814319361327761e-08,
"loss": 1.4941749572753906,
"step": 1160
},
{
"epoch": 1.9008567931456548,
"grad_norm": 2.234375,
"learning_rate": 6.694034952197448e-08,
"loss": 1.4870512008666992,
"step": 1165
},
{
"epoch": 1.909016727866177,
"grad_norm": 4.25,
"learning_rate": 5.65984716169754e-08,
"loss": 1.3893547058105469,
"step": 1170
},
{
"epoch": 1.9171766625866993,
"grad_norm": 2.984375,
"learning_rate": 4.711936486473334e-08,
"loss": 1.4003395080566405,
"step": 1175
},
{
"epoch": 1.9253365973072216,
"grad_norm": 3.890625,
"learning_rate": 3.85046836523767e-08,
"loss": 1.4021209716796874,
"step": 1180
},
{
"epoch": 1.9334965320277437,
"grad_norm": 2.9375,
"learning_rate": 3.075593149896916e-08,
"loss": 1.5010211944580079,
"step": 1185
},
{
"epoch": 1.9416564667482659,
"grad_norm": 2.078125,
"learning_rate": 2.3874460793102962e-08,
"loss": 1.4050621032714843,
"step": 1190
},
{
"epoch": 1.9498164014687882,
"grad_norm": 3.65625,
"learning_rate": 1.7861472556862103e-08,
"loss": 1.4399273872375489,
"step": 1195
},
{
"epoch": 1.9579763361893106,
"grad_norm": 2.390625,
"learning_rate": 1.2718016236213938e-08,
"loss": 1.5686559677124023,
"step": 1200
},
{
"epoch": 1.9661362709098327,
"grad_norm": 2.75,
"learning_rate": 8.444989517844005e-09,
"loss": 1.4039984703063966,
"step": 1205
},
{
"epoch": 1.9742962056303548,
"grad_norm": 3.375,
"learning_rate": 5.043138172487472e-09,
"loss": 1.416700553894043,
"step": 1210
},
{
"epoch": 1.9824561403508771,
"grad_norm": 2.5625,
"learning_rate": 2.5130559247665832e-09,
"loss": 1.4612144470214843,
"step": 1215
},
{
"epoch": 1.9906160750713995,
"grad_norm": 2.578125,
"learning_rate": 8.551843495696599e-10,
"loss": 1.5142659187316894,
"step": 1220
},
{
"epoch": 1.9987760097919218,
"grad_norm": 3.546875,
"learning_rate": 6.981279498274873e-11,
"loss": 1.3438663482666016,
"step": 1225
},
{
"epoch": 2.0,
"step": 1226,
"total_flos": 5.923294725916508e+17,
"train_loss": 1.5350287483722507,
"train_runtime": 10749.7665,
"train_samples_per_second": 0.456,
"train_steps_per_second": 0.114
}
],
"logging_steps": 5,
"max_steps": 1226,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"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": 5.923294725916508e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}