tinyllama-sft-orca_chat-mix / trainer_state.json
andrewbai's picture
Model save
bb0f4c0 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.999360204734485,
"eval_steps": 500,
"global_step": 781,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0012795905310300703,
"grad_norm": 13.69091815160461,
"learning_rate": 2.5316455696202533e-07,
"loss": 1.8445,
"step": 1
},
{
"epoch": 0.006397952655150352,
"grad_norm": 12.840412054889779,
"learning_rate": 1.2658227848101267e-06,
"loss": 1.6563,
"step": 5
},
{
"epoch": 0.012795905310300703,
"grad_norm": 10.488318286291324,
"learning_rate": 2.5316455696202535e-06,
"loss": 1.645,
"step": 10
},
{
"epoch": 0.019193857965451054,
"grad_norm": 4.368240374497205,
"learning_rate": 3.7974683544303802e-06,
"loss": 1.452,
"step": 15
},
{
"epoch": 0.025591810620601407,
"grad_norm": 1.829258307259485,
"learning_rate": 5.063291139240507e-06,
"loss": 1.3006,
"step": 20
},
{
"epoch": 0.03198976327575176,
"grad_norm": 1.302831212477095,
"learning_rate": 6.329113924050634e-06,
"loss": 1.2503,
"step": 25
},
{
"epoch": 0.03838771593090211,
"grad_norm": 1.2267087511737453,
"learning_rate": 7.5949367088607605e-06,
"loss": 1.1828,
"step": 30
},
{
"epoch": 0.044785668586052464,
"grad_norm": 1.2261121448269945,
"learning_rate": 8.860759493670886e-06,
"loss": 1.1808,
"step": 35
},
{
"epoch": 0.05118362124120281,
"grad_norm": 1.2827887632261594,
"learning_rate": 1.0126582278481014e-05,
"loss": 1.1684,
"step": 40
},
{
"epoch": 0.05758157389635317,
"grad_norm": 1.0440890636094624,
"learning_rate": 1.139240506329114e-05,
"loss": 1.1672,
"step": 45
},
{
"epoch": 0.06397952655150352,
"grad_norm": 1.0835741225896962,
"learning_rate": 1.2658227848101268e-05,
"loss": 1.0953,
"step": 50
},
{
"epoch": 0.07037747920665387,
"grad_norm": 1.1093304538316402,
"learning_rate": 1.3924050632911395e-05,
"loss": 1.0959,
"step": 55
},
{
"epoch": 0.07677543186180422,
"grad_norm": 1.123186995507472,
"learning_rate": 1.5189873417721521e-05,
"loss": 1.0936,
"step": 60
},
{
"epoch": 0.08317338451695458,
"grad_norm": 1.214434182932119,
"learning_rate": 1.6455696202531647e-05,
"loss": 1.1163,
"step": 65
},
{
"epoch": 0.08957133717210493,
"grad_norm": 1.10699408541261,
"learning_rate": 1.7721518987341772e-05,
"loss": 1.1342,
"step": 70
},
{
"epoch": 0.09596928982725528,
"grad_norm": 1.115056919875395,
"learning_rate": 1.89873417721519e-05,
"loss": 1.0592,
"step": 75
},
{
"epoch": 0.10236724248240563,
"grad_norm": 1.2262503515962806,
"learning_rate": 1.999989986294826e-05,
"loss": 1.047,
"step": 80
},
{
"epoch": 0.10876519513755598,
"grad_norm": 1.0071422101656622,
"learning_rate": 1.9996395276708856e-05,
"loss": 1.0533,
"step": 85
},
{
"epoch": 0.11516314779270634,
"grad_norm": 1.1865474148359212,
"learning_rate": 1.9987885843187717e-05,
"loss": 1.0572,
"step": 90
},
{
"epoch": 0.12156110044785669,
"grad_norm": 1.0757586228882932,
"learning_rate": 1.9974375822762117e-05,
"loss": 1.0618,
"step": 95
},
{
"epoch": 0.12795905310300704,
"grad_norm": 0.9404048622546317,
"learning_rate": 1.9955871979429188e-05,
"loss": 1.0659,
"step": 100
},
{
"epoch": 0.1343570057581574,
"grad_norm": 1.0223585826765225,
"learning_rate": 1.9932383577419432e-05,
"loss": 1.0638,
"step": 105
},
{
"epoch": 0.14075495841330773,
"grad_norm": 1.2195434324540952,
"learning_rate": 1.9903922376558432e-05,
"loss": 1.0546,
"step": 110
},
{
"epoch": 0.1471529110684581,
"grad_norm": 1.1102939400796445,
"learning_rate": 1.9870502626379127e-05,
"loss": 1.0705,
"step": 115
},
{
"epoch": 0.15355086372360843,
"grad_norm": 1.103687988612398,
"learning_rate": 1.983214105898757e-05,
"loss": 1.0354,
"step": 120
},
{
"epoch": 0.1599488163787588,
"grad_norm": 1.1124001535273442,
"learning_rate": 1.978885688068572e-05,
"loss": 1.0512,
"step": 125
},
{
"epoch": 0.16634676903390916,
"grad_norm": 1.1717239631401808,
"learning_rate": 1.9740671762355548e-05,
"loss": 1.0587,
"step": 130
},
{
"epoch": 0.1727447216890595,
"grad_norm": 1.1785241361491963,
"learning_rate": 1.9687609828609156e-05,
"loss": 1.039,
"step": 135
},
{
"epoch": 0.17914267434420986,
"grad_norm": 0.9884141225353881,
"learning_rate": 1.9629697645710432e-05,
"loss": 1.0216,
"step": 140
},
{
"epoch": 0.1855406269993602,
"grad_norm": 1.0520397277123574,
"learning_rate": 1.9566964208274254e-05,
"loss": 1.0684,
"step": 145
},
{
"epoch": 0.19193857965451055,
"grad_norm": 1.1891427908594712,
"learning_rate": 1.9499440924749878e-05,
"loss": 1.0206,
"step": 150
},
{
"epoch": 0.19833653230966092,
"grad_norm": 1.1154988222532596,
"learning_rate": 1.9427161601695833e-05,
"loss": 1.0037,
"step": 155
},
{
"epoch": 0.20473448496481125,
"grad_norm": 1.105546709789306,
"learning_rate": 1.9350162426854152e-05,
"loss": 0.9942,
"step": 160
},
{
"epoch": 0.21113243761996162,
"grad_norm": 1.1219266808856192,
"learning_rate": 1.926848195103242e-05,
"loss": 1.0426,
"step": 165
},
{
"epoch": 0.21753039027511195,
"grad_norm": 1.2007658640781749,
"learning_rate": 1.9182161068802742e-05,
"loss": 1.0192,
"step": 170
},
{
"epoch": 0.22392834293026231,
"grad_norm": 0.9333222341959819,
"learning_rate": 1.909124299802724e-05,
"loss": 0.9883,
"step": 175
},
{
"epoch": 0.23032629558541268,
"grad_norm": 1.192214367879126,
"learning_rate": 1.8995773258220374e-05,
"loss": 1.023,
"step": 180
},
{
"epoch": 0.236724248240563,
"grad_norm": 0.9148950044155676,
"learning_rate": 1.8895799647758912e-05,
"loss": 1.023,
"step": 185
},
{
"epoch": 0.24312220089571338,
"grad_norm": 0.9829242516952399,
"learning_rate": 1.879137221995095e-05,
"loss": 1.0031,
"step": 190
},
{
"epoch": 0.2495201535508637,
"grad_norm": 1.024194371856946,
"learning_rate": 1.868254325797594e-05,
"loss": 0.9938,
"step": 195
},
{
"epoch": 0.2559181062060141,
"grad_norm": 1.0534251974736057,
"learning_rate": 1.8569367248708343e-05,
"loss": 1.0204,
"step": 200
},
{
"epoch": 0.26231605886116444,
"grad_norm": 1.0216448456279743,
"learning_rate": 1.845190085543795e-05,
"loss": 1.013,
"step": 205
},
{
"epoch": 0.2687140115163148,
"grad_norm": 1.058446110223537,
"learning_rate": 1.8330202889500518e-05,
"loss": 1.038,
"step": 210
},
{
"epoch": 0.2751119641714651,
"grad_norm": 1.1612736755473039,
"learning_rate": 1.8204334280833005e-05,
"loss": 1.0158,
"step": 215
},
{
"epoch": 0.28150991682661547,
"grad_norm": 1.1373744375960468,
"learning_rate": 1.807435804746807e-05,
"loss": 0.9954,
"step": 220
},
{
"epoch": 0.28790786948176583,
"grad_norm": 0.8955352969068422,
"learning_rate": 1.7940339263983112e-05,
"loss": 1.0002,
"step": 225
},
{
"epoch": 0.2943058221369162,
"grad_norm": 1.0658172221253226,
"learning_rate": 1.7802345028919728e-05,
"loss": 1.0332,
"step": 230
},
{
"epoch": 0.30070377479206656,
"grad_norm": 0.9629957396185126,
"learning_rate": 1.766044443118978e-05,
"loss": 0.9921,
"step": 235
},
{
"epoch": 0.30710172744721687,
"grad_norm": 1.1794222718295426,
"learning_rate": 1.7514708515485002e-05,
"loss": 1.0208,
"step": 240
},
{
"epoch": 0.31349968010236723,
"grad_norm": 1.099778282186251,
"learning_rate": 1.736521024670737e-05,
"loss": 1.0137,
"step": 245
},
{
"epoch": 0.3198976327575176,
"grad_norm": 1.0490704822754882,
"learning_rate": 1.7212024473438145e-05,
"loss": 0.9886,
"step": 250
},
{
"epoch": 0.32629558541266795,
"grad_norm": 1.0370883879559893,
"learning_rate": 1.705522789046377e-05,
"loss": 1.0097,
"step": 255
},
{
"epoch": 0.3326935380678183,
"grad_norm": 1.0652270730291205,
"learning_rate": 1.6894899000377462e-05,
"loss": 1.0114,
"step": 260
},
{
"epoch": 0.3390914907229686,
"grad_norm": 1.0151841577549323,
"learning_rate": 1.67311180742757e-05,
"loss": 1.0016,
"step": 265
},
{
"epoch": 0.345489443378119,
"grad_norm": 0.9763116996401232,
"learning_rate": 1.65639671115693e-05,
"loss": 1.0346,
"step": 270
},
{
"epoch": 0.35188739603326935,
"grad_norm": 1.1003555211267269,
"learning_rate": 1.6393529798929103e-05,
"loss": 1.0202,
"step": 275
},
{
"epoch": 0.3582853486884197,
"grad_norm": 1.0489351299489513,
"learning_rate": 1.621989146838704e-05,
"loss": 1.0327,
"step": 280
},
{
"epoch": 0.3646833013435701,
"grad_norm": 0.9656565438172054,
"learning_rate": 1.6043139054613326e-05,
"loss": 1.0115,
"step": 285
},
{
"epoch": 0.3710812539987204,
"grad_norm": 0.9370055246045134,
"learning_rate": 1.586336105139127e-05,
"loss": 1.0351,
"step": 290
},
{
"epoch": 0.37747920665387075,
"grad_norm": 0.9340099772063268,
"learning_rate": 1.568064746731156e-05,
"loss": 1.0095,
"step": 295
},
{
"epoch": 0.3838771593090211,
"grad_norm": 0.9928028345305957,
"learning_rate": 1.5495089780708062e-05,
"loss": 0.9924,
"step": 300
},
{
"epoch": 0.3902751119641715,
"grad_norm": 1.0077676917038287,
"learning_rate": 1.530678089385782e-05,
"loss": 1.0119,
"step": 305
},
{
"epoch": 0.39667306461932184,
"grad_norm": 1.0545711896836611,
"learning_rate": 1.5115815086468103e-05,
"loss": 0.9963,
"step": 310
},
{
"epoch": 0.40307101727447214,
"grad_norm": 1.0556109648813998,
"learning_rate": 1.492228796847385e-05,
"loss": 0.9973,
"step": 315
},
{
"epoch": 0.4094689699296225,
"grad_norm": 0.9538130757153889,
"learning_rate": 1.4726296432169095e-05,
"loss": 1.0184,
"step": 320
},
{
"epoch": 0.41586692258477287,
"grad_norm": 1.2110036083338986,
"learning_rate": 1.4527938603696376e-05,
"loss": 0.9805,
"step": 325
},
{
"epoch": 0.42226487523992323,
"grad_norm": 0.994863618345283,
"learning_rate": 1.4327313793918362e-05,
"loss": 0.9871,
"step": 330
},
{
"epoch": 0.4286628278950736,
"grad_norm": 0.9724066831142806,
"learning_rate": 1.4124522448696407e-05,
"loss": 1.0285,
"step": 335
},
{
"epoch": 0.4350607805502239,
"grad_norm": 0.9713521005708377,
"learning_rate": 1.3919666098600753e-05,
"loss": 0.9786,
"step": 340
},
{
"epoch": 0.44145873320537427,
"grad_norm": 1.1651665963666036,
"learning_rate": 1.3712847308077737e-05,
"loss": 0.9695,
"step": 345
},
{
"epoch": 0.44785668586052463,
"grad_norm": 1.1083847488632985,
"learning_rate": 1.350416962409934e-05,
"loss": 0.9868,
"step": 350
},
{
"epoch": 0.454254638515675,
"grad_norm": 1.0523508672708513,
"learning_rate": 1.3293737524320798e-05,
"loss": 1.0132,
"step": 355
},
{
"epoch": 0.46065259117082535,
"grad_norm": 1.1521373650326732,
"learning_rate": 1.3081656364772308e-05,
"loss": 0.98,
"step": 360
},
{
"epoch": 0.46705054382597566,
"grad_norm": 0.9089053301856657,
"learning_rate": 1.2868032327110904e-05,
"loss": 0.982,
"step": 365
},
{
"epoch": 0.473448496481126,
"grad_norm": 1.0069257216134682,
"learning_rate": 1.2652972365459008e-05,
"loss": 0.9836,
"step": 370
},
{
"epoch": 0.4798464491362764,
"grad_norm": 1.106463301558632,
"learning_rate": 1.243658415285622e-05,
"loss": 1.0105,
"step": 375
},
{
"epoch": 0.48624440179142675,
"grad_norm": 0.9231713783345915,
"learning_rate": 1.2218976027351177e-05,
"loss": 0.9722,
"step": 380
},
{
"epoch": 0.4926423544465771,
"grad_norm": 0.9227196904627529,
"learning_rate": 1.2000256937760446e-05,
"loss": 0.9802,
"step": 385
},
{
"epoch": 0.4990403071017274,
"grad_norm": 0.9352639317335375,
"learning_rate": 1.1780536389121668e-05,
"loss": 1.0127,
"step": 390
},
{
"epoch": 0.5054382597568778,
"grad_norm": 1.1193201504587136,
"learning_rate": 1.155992438786818e-05,
"loss": 0.9834,
"step": 395
},
{
"epoch": 0.5118362124120281,
"grad_norm": 1.1237358252844734,
"learning_rate": 1.1338531386752618e-05,
"loss": 0.9917,
"step": 400
},
{
"epoch": 0.5182341650671785,
"grad_norm": 0.9880901116754547,
"learning_rate": 1.1116468229547079e-05,
"loss": 0.9726,
"step": 405
},
{
"epoch": 0.5246321177223289,
"grad_norm": 0.9538208713791894,
"learning_rate": 1.0893846095547493e-05,
"loss": 1.0044,
"step": 410
},
{
"epoch": 0.5310300703774792,
"grad_norm": 1.0075299668140827,
"learning_rate": 1.0670776443910024e-05,
"loss": 1.0038,
"step": 415
},
{
"epoch": 0.5374280230326296,
"grad_norm": 1.0525123195189434,
"learning_rate": 1.0447370957847343e-05,
"loss": 0.9611,
"step": 420
},
{
"epoch": 0.5438259756877799,
"grad_norm": 0.9793000508253443,
"learning_rate": 1.0223741488712732e-05,
"loss": 0.9546,
"step": 425
},
{
"epoch": 0.5502239283429302,
"grad_norm": 1.0313068903836107,
"learning_rate": 1e-05,
"loss": 0.9849,
"step": 430
},
{
"epoch": 0.5566218809980806,
"grad_norm": 1.0401386769180931,
"learning_rate": 9.776258511287271e-06,
"loss": 0.9263,
"step": 435
},
{
"epoch": 0.5630198336532309,
"grad_norm": 1.045623699258006,
"learning_rate": 9.55262904215266e-06,
"loss": 0.9428,
"step": 440
},
{
"epoch": 0.5694177863083814,
"grad_norm": 1.0307153076441182,
"learning_rate": 9.329223556089976e-06,
"loss": 0.9682,
"step": 445
},
{
"epoch": 0.5758157389635317,
"grad_norm": 1.063552223235608,
"learning_rate": 9.10615390445251e-06,
"loss": 0.9632,
"step": 450
},
{
"epoch": 0.582213691618682,
"grad_norm": 1.015053350316954,
"learning_rate": 8.883531770452924e-06,
"loss": 1.0019,
"step": 455
},
{
"epoch": 0.5886116442738324,
"grad_norm": 0.9686964925312078,
"learning_rate": 8.661468613247387e-06,
"loss": 0.9856,
"step": 460
},
{
"epoch": 0.5950095969289827,
"grad_norm": 0.9118252806857438,
"learning_rate": 8.440075612131823e-06,
"loss": 0.9905,
"step": 465
},
{
"epoch": 0.6014075495841331,
"grad_norm": 0.9825555035620782,
"learning_rate": 8.219463610878336e-06,
"loss": 0.9662,
"step": 470
},
{
"epoch": 0.6078055022392834,
"grad_norm": 0.987013717956065,
"learning_rate": 7.999743062239557e-06,
"loss": 0.9803,
"step": 475
},
{
"epoch": 0.6142034548944337,
"grad_norm": 0.9255508396804797,
"learning_rate": 7.781023972648826e-06,
"loss": 0.9847,
"step": 480
},
{
"epoch": 0.6206014075495841,
"grad_norm": 0.9872235531461877,
"learning_rate": 7.563415847143782e-06,
"loss": 0.9794,
"step": 485
},
{
"epoch": 0.6269993602047345,
"grad_norm": 0.9831490559910753,
"learning_rate": 7.347027634540993e-06,
"loss": 0.9923,
"step": 490
},
{
"epoch": 0.6333973128598849,
"grad_norm": 0.9732824895135384,
"learning_rate": 7.131967672889101e-06,
"loss": 0.9801,
"step": 495
},
{
"epoch": 0.6397952655150352,
"grad_norm": 0.8714243008263366,
"learning_rate": 6.918343635227694e-06,
"loss": 0.9939,
"step": 500
},
{
"epoch": 0.6461932181701855,
"grad_norm": 1.0601279930665841,
"learning_rate": 6.706262475679205e-06,
"loss": 0.9775,
"step": 505
},
{
"epoch": 0.6525911708253359,
"grad_norm": 0.95911999094283,
"learning_rate": 6.495830375900665e-06,
"loss": 0.9894,
"step": 510
},
{
"epoch": 0.6589891234804862,
"grad_norm": 0.9761785661849035,
"learning_rate": 6.287152691922264e-06,
"loss": 0.9364,
"step": 515
},
{
"epoch": 0.6653870761356366,
"grad_norm": 1.0110868496188614,
"learning_rate": 6.080333901399252e-06,
"loss": 0.9564,
"step": 520
},
{
"epoch": 0.6717850287907869,
"grad_norm": 0.9389337048437643,
"learning_rate": 5.875477551303596e-06,
"loss": 0.9589,
"step": 525
},
{
"epoch": 0.6781829814459372,
"grad_norm": 1.1057673049959478,
"learning_rate": 5.672686206081638e-06,
"loss": 0.9953,
"step": 530
},
{
"epoch": 0.6845809341010877,
"grad_norm": 0.8785448267158206,
"learning_rate": 5.47206139630363e-06,
"loss": 0.9432,
"step": 535
},
{
"epoch": 0.690978886756238,
"grad_norm": 1.0754908863463288,
"learning_rate": 5.273703567830908e-06,
"loss": 0.9728,
"step": 540
},
{
"epoch": 0.6973768394113884,
"grad_norm": 0.9327335147601428,
"learning_rate": 5.077712031526153e-06,
"loss": 0.9207,
"step": 545
},
{
"epoch": 0.7037747920665387,
"grad_norm": 0.9357200809246364,
"learning_rate": 4.8841849135319015e-06,
"loss": 0.9165,
"step": 550
},
{
"epoch": 0.710172744721689,
"grad_norm": 1.0419371249401577,
"learning_rate": 4.693219106142186e-06,
"loss": 0.9599,
"step": 555
},
{
"epoch": 0.7165706973768394,
"grad_norm": 0.8030703401559492,
"learning_rate": 4.504910219291941e-06,
"loss": 0.9698,
"step": 560
},
{
"epoch": 0.7229686500319897,
"grad_norm": 0.9308184657001143,
"learning_rate": 4.319352532688444e-06,
"loss": 0.9625,
"step": 565
},
{
"epoch": 0.7293666026871402,
"grad_norm": 0.9141662687430406,
"learning_rate": 4.13663894860873e-06,
"loss": 0.9933,
"step": 570
},
{
"epoch": 0.7357645553422905,
"grad_norm": 0.9937683231671001,
"learning_rate": 3.956860945386677e-06,
"loss": 0.9517,
"step": 575
},
{
"epoch": 0.7421625079974408,
"grad_norm": 1.0729326983327145,
"learning_rate": 3.7801085316129615e-06,
"loss": 0.9687,
"step": 580
},
{
"epoch": 0.7485604606525912,
"grad_norm": 0.942218409844499,
"learning_rate": 3.606470201070904e-06,
"loss": 0.9496,
"step": 585
},
{
"epoch": 0.7549584133077415,
"grad_norm": 1.028542441110995,
"learning_rate": 3.4360328884307058e-06,
"loss": 0.9538,
"step": 590
},
{
"epoch": 0.7613563659628919,
"grad_norm": 1.016980822716628,
"learning_rate": 3.2688819257242963e-06,
"loss": 0.9328,
"step": 595
},
{
"epoch": 0.7677543186180422,
"grad_norm": 1.0401770676735838,
"learning_rate": 3.1051009996225434e-06,
"loss": 0.9684,
"step": 600
},
{
"epoch": 0.7741522712731925,
"grad_norm": 1.0575436512016658,
"learning_rate": 2.9447721095362325e-06,
"loss": 0.9287,
"step": 605
},
{
"epoch": 0.780550223928343,
"grad_norm": 1.0250671533063895,
"learning_rate": 2.7879755265618558e-06,
"loss": 0.9549,
"step": 610
},
{
"epoch": 0.7869481765834933,
"grad_norm": 1.0253893757162196,
"learning_rate": 2.6347897532926293e-06,
"loss": 0.9845,
"step": 615
},
{
"epoch": 0.7933461292386437,
"grad_norm": 1.154415856980631,
"learning_rate": 2.485291484515e-06,
"loss": 0.9843,
"step": 620
},
{
"epoch": 0.799744081893794,
"grad_norm": 1.0958121833917867,
"learning_rate": 2.339555568810221e-06,
"loss": 0.9383,
"step": 625
},
{
"epoch": 0.8061420345489443,
"grad_norm": 0.9384925082776849,
"learning_rate": 2.1976549710802754e-06,
"loss": 0.9534,
"step": 630
},
{
"epoch": 0.8125399872040947,
"grad_norm": 1.0343911972749902,
"learning_rate": 2.0596607360168897e-06,
"loss": 0.9734,
"step": 635
},
{
"epoch": 0.818937939859245,
"grad_norm": 0.8737130115934248,
"learning_rate": 1.9256419525319316e-06,
"loss": 0.9679,
"step": 640
},
{
"epoch": 0.8253358925143954,
"grad_norm": 0.9616800124257279,
"learning_rate": 1.7956657191669969e-06,
"loss": 0.9645,
"step": 645
},
{
"epoch": 0.8317338451695457,
"grad_norm": 1.0354353675764634,
"learning_rate": 1.6697971104994847e-06,
"loss": 0.9782,
"step": 650
},
{
"epoch": 0.838131797824696,
"grad_norm": 0.9366269080497807,
"learning_rate": 1.5480991445620541e-06,
"loss": 0.9494,
"step": 655
},
{
"epoch": 0.8445297504798465,
"grad_norm": 1.0820594326475506,
"learning_rate": 1.4306327512916574e-06,
"loss": 0.9456,
"step": 660
},
{
"epoch": 0.8509277031349968,
"grad_norm": 1.0468634102383998,
"learning_rate": 1.3174567420240647e-06,
"loss": 0.9449,
"step": 665
},
{
"epoch": 0.8573256557901472,
"grad_norm": 0.870344280190667,
"learning_rate": 1.2086277800490554e-06,
"loss": 0.9337,
"step": 670
},
{
"epoch": 0.8637236084452975,
"grad_norm": 0.9794345251775067,
"learning_rate": 1.1042003522410882e-06,
"loss": 1.0005,
"step": 675
},
{
"epoch": 0.8701215611004478,
"grad_norm": 1.058395466495503,
"learning_rate": 1.0042267417796292e-06,
"loss": 0.9473,
"step": 680
},
{
"epoch": 0.8765195137555982,
"grad_norm": 0.981770260070343,
"learning_rate": 9.08757001972762e-07,
"loss": 0.9517,
"step": 685
},
{
"epoch": 0.8829174664107485,
"grad_norm": 0.9326065749239703,
"learning_rate": 8.178389311972612e-07,
"loss": 0.962,
"step": 690
},
{
"epoch": 0.889315419065899,
"grad_norm": 0.9441889253526499,
"learning_rate": 7.315180489675822e-07,
"loss": 0.9256,
"step": 695
},
{
"epoch": 0.8957133717210493,
"grad_norm": 1.022504373128847,
"learning_rate": 6.498375731458529e-07,
"loss": 0.9753,
"step": 700
},
{
"epoch": 0.9021113243761996,
"grad_norm": 1.012774590686695,
"learning_rate": 5.728383983041696e-07,
"loss": 0.9719,
"step": 705
},
{
"epoch": 0.90850927703135,
"grad_norm": 0.9719468902067695,
"learning_rate": 5.005590752501244e-07,
"loss": 0.9751,
"step": 710
},
{
"epoch": 0.9149072296865003,
"grad_norm": 0.922759682421754,
"learning_rate": 4.3303579172574884e-07,
"loss": 0.9438,
"step": 715
},
{
"epoch": 0.9213051823416507,
"grad_norm": 0.9702427160470812,
"learning_rate": 3.7030235428956895e-07,
"loss": 0.9471,
"step": 720
},
{
"epoch": 0.927703134996801,
"grad_norm": 0.9866988086939049,
"learning_rate": 3.1239017139084725e-07,
"loss": 0.9498,
"step": 725
},
{
"epoch": 0.9341010876519513,
"grad_norm": 1.0075946955261035,
"learning_rate": 2.593282376444539e-07,
"loss": 0.9385,
"step": 730
},
{
"epoch": 0.9404990403071017,
"grad_norm": 1.0046835757442745,
"learning_rate": 2.11143119314281e-07,
"loss": 0.9556,
"step": 735
},
{
"epoch": 0.946896992962252,
"grad_norm": 0.9457637995485544,
"learning_rate": 1.6785894101243205e-07,
"loss": 0.9549,
"step": 740
},
{
"epoch": 0.9532949456174025,
"grad_norm": 0.8562754137399341,
"learning_rate": 1.2949737362087156e-07,
"loss": 0.9613,
"step": 745
},
{
"epoch": 0.9596928982725528,
"grad_norm": 1.0842537128116503,
"learning_rate": 9.607762344156946e-08,
"loss": 0.9705,
"step": 750
},
{
"epoch": 0.9660908509277031,
"grad_norm": 0.8944257328647005,
"learning_rate": 6.761642258056977e-08,
"loss": 0.9478,
"step": 755
},
{
"epoch": 0.9724888035828535,
"grad_norm": 0.8604683433937093,
"learning_rate": 4.412802057081278e-08,
"loss": 0.9736,
"step": 760
},
{
"epoch": 0.9788867562380038,
"grad_norm": 1.009449082533386,
"learning_rate": 2.5624177237884017e-08,
"loss": 0.9285,
"step": 765
},
{
"epoch": 0.9852847088931542,
"grad_norm": 0.8974627040364813,
"learning_rate": 1.2114156812284006e-08,
"loss": 0.956,
"step": 770
},
{
"epoch": 0.9916826615483045,
"grad_norm": 1.0253106873389546,
"learning_rate": 3.6047232911462506e-09,
"loss": 0.9377,
"step": 775
},
{
"epoch": 0.9980806142034548,
"grad_norm": 1.0048439402772062,
"learning_rate": 1.0013705174061195e-10,
"loss": 0.9347,
"step": 780
},
{
"epoch": 0.999360204734485,
"eval_loss": 0.949686586856842,
"eval_runtime": 1745.7859,
"eval_samples_per_second": 56.395,
"eval_steps_per_second": 1.763,
"step": 781
},
{
"epoch": 0.999360204734485,
"step": 781,
"total_flos": 113080408473600.0,
"train_loss": 1.0112148089842363,
"train_runtime": 9125.7422,
"train_samples_per_second": 10.958,
"train_steps_per_second": 0.086
}
],
"logging_steps": 5,
"max_steps": 781,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"total_flos": 113080408473600.0,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}