GanymedeNil's picture
First model version
89ce1a9
{
"best_metric": 1.4598569869995117,
"best_model_checkpoint": "saves/Gemma-2B/lora/train_2024-03-01-04-36-32/checkpoint-2000",
"epoch": 3.5555555555555554,
"eval_steps": 100,
"global_step": 2000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.02,
"grad_norm": 6.6079277992248535,
"learning_rate": 4.999960939662063e-05,
"loss": 3.747,
"step": 10
},
{
"epoch": 0.04,
"grad_norm": 3.2283411026000977,
"learning_rate": 4.999843759868819e-05,
"loss": 3.5789,
"step": 20
},
{
"epoch": 0.05,
"grad_norm": 41.573001861572266,
"learning_rate": 4.999648464281934e-05,
"loss": 3.1683,
"step": 30
},
{
"epoch": 0.07,
"grad_norm": 4.080965518951416,
"learning_rate": 4.9993750590040575e-05,
"loss": 2.8275,
"step": 40
},
{
"epoch": 0.09,
"grad_norm": 4.576275825500488,
"learning_rate": 4.999023552578632e-05,
"loss": 2.6758,
"step": 50
},
{
"epoch": 0.11,
"grad_norm": 18.012842178344727,
"learning_rate": 4.998593955989626e-05,
"loss": 2.6287,
"step": 60
},
{
"epoch": 0.12,
"grad_norm": 5.738934516906738,
"learning_rate": 4.9980862826611875e-05,
"loss": 2.5284,
"step": 70
},
{
"epoch": 0.14,
"grad_norm": 3.353776216506958,
"learning_rate": 4.9975005484572305e-05,
"loss": 2.2608,
"step": 80
},
{
"epoch": 0.16,
"grad_norm": 4.6298699378967285,
"learning_rate": 4.9968367716809374e-05,
"loss": 2.2475,
"step": 90
},
{
"epoch": 0.18,
"grad_norm": 50.594207763671875,
"learning_rate": 4.996094973074183e-05,
"loss": 2.2007,
"step": 100
},
{
"epoch": 0.18,
"eval_loss": 2.126384735107422,
"eval_runtime": 124.9221,
"eval_samples_per_second": 8.005,
"eval_steps_per_second": 2.001,
"step": 100
},
{
"epoch": 0.2,
"grad_norm": 10.225520133972168,
"learning_rate": 4.995275175816891e-05,
"loss": 1.9414,
"step": 110
},
{
"epoch": 0.21,
"grad_norm": 4.777626991271973,
"learning_rate": 4.994377405526308e-05,
"loss": 1.9729,
"step": 120
},
{
"epoch": 0.23,
"grad_norm": 6.133576393127441,
"learning_rate": 4.993401690256203e-05,
"loss": 2.0237,
"step": 130
},
{
"epoch": 0.25,
"grad_norm": 5.396271228790283,
"learning_rate": 4.992348060495989e-05,
"loss": 2.009,
"step": 140
},
{
"epoch": 0.27,
"grad_norm": 3.4974453449249268,
"learning_rate": 4.991216549169776e-05,
"loss": 2.032,
"step": 150
},
{
"epoch": 0.28,
"grad_norm": 12.256199836730957,
"learning_rate": 4.990007191635334e-05,
"loss": 1.9548,
"step": 160
},
{
"epoch": 0.3,
"grad_norm": 7.5634379386901855,
"learning_rate": 4.988720025682995e-05,
"loss": 1.8164,
"step": 170
},
{
"epoch": 0.32,
"grad_norm": 14.023727416992188,
"learning_rate": 4.987355091534468e-05,
"loss": 1.8517,
"step": 180
},
{
"epoch": 0.34,
"grad_norm": 4.622091293334961,
"learning_rate": 4.985912431841584e-05,
"loss": 2.0255,
"step": 190
},
{
"epoch": 0.36,
"grad_norm": 3.9935083389282227,
"learning_rate": 4.9843920916849645e-05,
"loss": 1.8777,
"step": 200
},
{
"epoch": 0.36,
"eval_loss": 1.8619400262832642,
"eval_runtime": 124.8712,
"eval_samples_per_second": 8.008,
"eval_steps_per_second": 2.002,
"step": 200
},
{
"epoch": 0.37,
"grad_norm": 6.256485939025879,
"learning_rate": 4.982794118572609e-05,
"loss": 1.8885,
"step": 210
},
{
"epoch": 0.39,
"grad_norm": 13.212824821472168,
"learning_rate": 4.981118562438414e-05,
"loss": 1.7744,
"step": 220
},
{
"epoch": 0.41,
"grad_norm": 4.2626118659973145,
"learning_rate": 4.9793654756406085e-05,
"loss": 1.7545,
"step": 230
},
{
"epoch": 0.43,
"grad_norm": 4.217405796051025,
"learning_rate": 4.9775349129601243e-05,
"loss": 1.5633,
"step": 240
},
{
"epoch": 0.44,
"grad_norm": 22.393404006958008,
"learning_rate": 4.9756269315988804e-05,
"loss": 1.8871,
"step": 250
},
{
"epoch": 0.46,
"grad_norm": 3.6576473712921143,
"learning_rate": 4.973641591177991e-05,
"loss": 1.7037,
"step": 260
},
{
"epoch": 0.48,
"grad_norm": 4.2433271408081055,
"learning_rate": 4.971578953735912e-05,
"loss": 1.7631,
"step": 270
},
{
"epoch": 0.5,
"grad_norm": 3.7399721145629883,
"learning_rate": 4.969439083726496e-05,
"loss": 1.7714,
"step": 280
},
{
"epoch": 0.52,
"grad_norm": 4.575680255889893,
"learning_rate": 4.967222048016979e-05,
"loss": 1.8699,
"step": 290
},
{
"epoch": 0.53,
"grad_norm": 7.729683876037598,
"learning_rate": 4.964927915885893e-05,
"loss": 1.6566,
"step": 300
},
{
"epoch": 0.53,
"eval_loss": 1.7350378036499023,
"eval_runtime": 124.9278,
"eval_samples_per_second": 8.005,
"eval_steps_per_second": 2.001,
"step": 300
},
{
"epoch": 0.55,
"grad_norm": 2.755899667739868,
"learning_rate": 4.962556759020898e-05,
"loss": 1.7193,
"step": 310
},
{
"epoch": 0.57,
"grad_norm": 3.513024091720581,
"learning_rate": 4.960108651516545e-05,
"loss": 1.852,
"step": 320
},
{
"epoch": 0.59,
"grad_norm": 3.7794790267944336,
"learning_rate": 4.9575836698719605e-05,
"loss": 1.6785,
"step": 330
},
{
"epoch": 0.6,
"grad_norm": 3.2256739139556885,
"learning_rate": 4.954981892988451e-05,
"loss": 1.6648,
"step": 340
},
{
"epoch": 0.62,
"grad_norm": 2.8756954669952393,
"learning_rate": 4.952303402167047e-05,
"loss": 1.6399,
"step": 350
},
{
"epoch": 0.64,
"grad_norm": 7.057961463928223,
"learning_rate": 4.949548281105951e-05,
"loss": 1.5875,
"step": 360
},
{
"epoch": 0.66,
"grad_norm": 4.63081169128418,
"learning_rate": 4.946716615897932e-05,
"loss": 1.6708,
"step": 370
},
{
"epoch": 0.68,
"grad_norm": 8.755204200744629,
"learning_rate": 4.943808495027631e-05,
"loss": 1.636,
"step": 380
},
{
"epoch": 0.69,
"grad_norm": 10.21866226196289,
"learning_rate": 4.940824009368793e-05,
"loss": 1.5714,
"step": 390
},
{
"epoch": 0.71,
"grad_norm": 5.44133186340332,
"learning_rate": 4.937763252181434e-05,
"loss": 1.4084,
"step": 400
},
{
"epoch": 0.71,
"eval_loss": 1.6840696334838867,
"eval_runtime": 124.8851,
"eval_samples_per_second": 8.007,
"eval_steps_per_second": 2.002,
"step": 400
},
{
"epoch": 0.73,
"grad_norm": 3.056345224380493,
"learning_rate": 4.934626319108923e-05,
"loss": 1.7233,
"step": 410
},
{
"epoch": 0.75,
"grad_norm": 4.303133487701416,
"learning_rate": 4.93141330817499e-05,
"loss": 1.5374,
"step": 420
},
{
"epoch": 0.76,
"grad_norm": 5.2246623039245605,
"learning_rate": 4.9281243197806726e-05,
"loss": 1.8547,
"step": 430
},
{
"epoch": 0.78,
"grad_norm": 3.8070685863494873,
"learning_rate": 4.924759456701167e-05,
"loss": 1.5721,
"step": 440
},
{
"epoch": 0.8,
"grad_norm": 3.243337392807007,
"learning_rate": 4.9213188240826245e-05,
"loss": 1.4322,
"step": 450
},
{
"epoch": 0.82,
"grad_norm": 4.166132926940918,
"learning_rate": 4.917802529438864e-05,
"loss": 1.6621,
"step": 460
},
{
"epoch": 0.84,
"grad_norm": 4.54414701461792,
"learning_rate": 4.9142106826480114e-05,
"loss": 1.6088,
"step": 470
},
{
"epoch": 0.85,
"grad_norm": 9.983458518981934,
"learning_rate": 4.910543395949067e-05,
"loss": 1.6152,
"step": 480
},
{
"epoch": 0.87,
"grad_norm": 6.45111608505249,
"learning_rate": 4.9068007839383946e-05,
"loss": 1.6361,
"step": 490
},
{
"epoch": 0.89,
"grad_norm": 108.82310485839844,
"learning_rate": 4.9029829635661475e-05,
"loss": 1.7045,
"step": 500
},
{
"epoch": 0.89,
"eval_loss": 1.6494970321655273,
"eval_runtime": 124.6904,
"eval_samples_per_second": 8.02,
"eval_steps_per_second": 2.005,
"step": 500
},
{
"epoch": 0.91,
"grad_norm": 5.705786228179932,
"learning_rate": 4.899090054132609e-05,
"loss": 1.738,
"step": 510
},
{
"epoch": 0.92,
"grad_norm": 4.800131320953369,
"learning_rate": 4.895122177284465e-05,
"loss": 1.6218,
"step": 520
},
{
"epoch": 0.94,
"grad_norm": 10.11057186126709,
"learning_rate": 4.891079457011005e-05,
"loss": 1.5169,
"step": 530
},
{
"epoch": 0.96,
"grad_norm": 9.329095840454102,
"learning_rate": 4.8869620196402436e-05,
"loss": 1.7979,
"step": 540
},
{
"epoch": 0.98,
"grad_norm": 3.9115641117095947,
"learning_rate": 4.882769993834978e-05,
"loss": 1.7073,
"step": 550
},
{
"epoch": 1.0,
"grad_norm": 4.80266809463501,
"learning_rate": 4.878503510588765e-05,
"loss": 1.6541,
"step": 560
},
{
"epoch": 1.01,
"grad_norm": 9.07653522491455,
"learning_rate": 4.874162703221823e-05,
"loss": 1.6888,
"step": 570
},
{
"epoch": 1.03,
"grad_norm": 4.492751598358154,
"learning_rate": 4.8697477073768766e-05,
"loss": 1.6448,
"step": 580
},
{
"epoch": 1.05,
"grad_norm": 13.852599143981934,
"learning_rate": 4.8652586610149095e-05,
"loss": 1.6236,
"step": 590
},
{
"epoch": 1.07,
"grad_norm": 5.424524307250977,
"learning_rate": 4.8606957044108556e-05,
"loss": 1.4969,
"step": 600
},
{
"epoch": 1.07,
"eval_loss": 1.6121476888656616,
"eval_runtime": 124.7413,
"eval_samples_per_second": 8.017,
"eval_steps_per_second": 2.004,
"step": 600
},
{
"epoch": 1.08,
"grad_norm": 3.611617088317871,
"learning_rate": 4.856058980149216e-05,
"loss": 1.4571,
"step": 610
},
{
"epoch": 1.1,
"grad_norm": 4.210519313812256,
"learning_rate": 4.851348633119606e-05,
"loss": 1.63,
"step": 620
},
{
"epoch": 1.12,
"grad_norm": 95.43629455566406,
"learning_rate": 4.84656481051222e-05,
"loss": 1.6034,
"step": 630
},
{
"epoch": 1.14,
"grad_norm": 4.3693528175354,
"learning_rate": 4.8417076618132426e-05,
"loss": 1.5791,
"step": 640
},
{
"epoch": 1.16,
"grad_norm": 3.691178321838379,
"learning_rate": 4.836777338800168e-05,
"loss": 1.5327,
"step": 650
},
{
"epoch": 1.17,
"grad_norm": 3.547637939453125,
"learning_rate": 4.8317739955370636e-05,
"loss": 1.4278,
"step": 660
},
{
"epoch": 1.19,
"grad_norm": 3.426717519760132,
"learning_rate": 4.8266977883697515e-05,
"loss": 1.5317,
"step": 670
},
{
"epoch": 1.21,
"grad_norm": 3.004473924636841,
"learning_rate": 4.821548875920927e-05,
"loss": 1.6848,
"step": 680
},
{
"epoch": 1.23,
"grad_norm": 3.686044931411743,
"learning_rate": 4.816327419085196e-05,
"loss": 1.6079,
"step": 690
},
{
"epoch": 1.24,
"grad_norm": 4.130298137664795,
"learning_rate": 4.811033581024056e-05,
"loss": 1.5998,
"step": 700
},
{
"epoch": 1.24,
"eval_loss": 1.5970302820205688,
"eval_runtime": 124.9388,
"eval_samples_per_second": 8.004,
"eval_steps_per_second": 2.001,
"step": 700
},
{
"epoch": 1.26,
"grad_norm": 6.1143059730529785,
"learning_rate": 4.805667527160788e-05,
"loss": 1.554,
"step": 710
},
{
"epoch": 1.28,
"grad_norm": 31.27813148498535,
"learning_rate": 4.800229425175294e-05,
"loss": 1.5824,
"step": 720
},
{
"epoch": 1.3,
"grad_norm": 9.035768508911133,
"learning_rate": 4.7947194449988555e-05,
"loss": 1.547,
"step": 730
},
{
"epoch": 1.32,
"grad_norm": 39.38993835449219,
"learning_rate": 4.7891377588088223e-05,
"loss": 1.5795,
"step": 740
},
{
"epoch": 1.33,
"grad_norm": 7.738800048828125,
"learning_rate": 4.7834845410232356e-05,
"loss": 1.5761,
"step": 750
},
{
"epoch": 1.35,
"grad_norm": 3.3933961391448975,
"learning_rate": 4.777759968295369e-05,
"loss": 1.6293,
"step": 760
},
{
"epoch": 1.37,
"grad_norm": 4.511744022369385,
"learning_rate": 4.771964219508222e-05,
"loss": 1.4761,
"step": 770
},
{
"epoch": 1.39,
"grad_norm": 3.566397190093994,
"learning_rate": 4.766097475768919e-05,
"loss": 1.5707,
"step": 780
},
{
"epoch": 1.4,
"grad_norm": 9.365654945373535,
"learning_rate": 4.7601599204030544e-05,
"loss": 1.3932,
"step": 790
},
{
"epoch": 1.42,
"grad_norm": 3.3254847526550293,
"learning_rate": 4.754151738948962e-05,
"loss": 1.6041,
"step": 800
},
{
"epoch": 1.42,
"eval_loss": 1.5639870166778564,
"eval_runtime": 124.923,
"eval_samples_per_second": 8.005,
"eval_steps_per_second": 2.001,
"step": 800
},
{
"epoch": 1.44,
"grad_norm": 3.520264148712158,
"learning_rate": 4.7480731191519224e-05,
"loss": 1.4991,
"step": 810
},
{
"epoch": 1.46,
"grad_norm": 5.3987531661987305,
"learning_rate": 4.741924250958289e-05,
"loss": 1.6856,
"step": 820
},
{
"epoch": 1.48,
"grad_norm": 12.352794647216797,
"learning_rate": 4.7357053265095575e-05,
"loss": 1.4509,
"step": 830
},
{
"epoch": 1.49,
"grad_norm": 9.825531005859375,
"learning_rate": 4.729416540136361e-05,
"loss": 1.6168,
"step": 840
},
{
"epoch": 1.51,
"grad_norm": 10.881526947021484,
"learning_rate": 4.723058088352395e-05,
"loss": 1.5783,
"step": 850
},
{
"epoch": 1.53,
"grad_norm": 6.232407093048096,
"learning_rate": 4.7166301698482815e-05,
"loss": 1.4556,
"step": 860
},
{
"epoch": 1.55,
"grad_norm": 3.3216302394866943,
"learning_rate": 4.710132985485355e-05,
"loss": 1.593,
"step": 870
},
{
"epoch": 1.56,
"grad_norm": 5.219264984130859,
"learning_rate": 4.703566738289389e-05,
"loss": 1.5131,
"step": 880
},
{
"epoch": 1.58,
"grad_norm": 7.875769138336182,
"learning_rate": 4.696931633444251e-05,
"loss": 1.5667,
"step": 890
},
{
"epoch": 1.6,
"grad_norm": 5.77959680557251,
"learning_rate": 4.69022787828549e-05,
"loss": 1.5211,
"step": 900
},
{
"epoch": 1.6,
"eval_loss": 1.5731443166732788,
"eval_runtime": 124.8025,
"eval_samples_per_second": 8.013,
"eval_steps_per_second": 2.003,
"step": 900
},
{
"epoch": 1.62,
"grad_norm": 4.806954383850098,
"learning_rate": 4.683455682293863e-05,
"loss": 1.6824,
"step": 910
},
{
"epoch": 1.64,
"grad_norm": 5.980200290679932,
"learning_rate": 4.676615257088776e-05,
"loss": 1.5989,
"step": 920
},
{
"epoch": 1.65,
"grad_norm": 4.3645429611206055,
"learning_rate": 4.6697068164216896e-05,
"loss": 1.6469,
"step": 930
},
{
"epoch": 1.67,
"grad_norm": 3.2400012016296387,
"learning_rate": 4.662730576169423e-05,
"loss": 1.568,
"step": 940
},
{
"epoch": 1.69,
"grad_norm": 4.331827640533447,
"learning_rate": 4.6556867543274184e-05,
"loss": 1.5236,
"step": 950
},
{
"epoch": 1.71,
"grad_norm": 3.3798201084136963,
"learning_rate": 4.6485755710029256e-05,
"loss": 1.5046,
"step": 960
},
{
"epoch": 1.72,
"grad_norm": 5.440864086151123,
"learning_rate": 4.6413972484081216e-05,
"loss": 1.5816,
"step": 970
},
{
"epoch": 1.74,
"grad_norm": 5.852995872497559,
"learning_rate": 4.6341520108531746e-05,
"loss": 1.4193,
"step": 980
},
{
"epoch": 1.76,
"grad_norm": 4.2782206535339355,
"learning_rate": 4.626840084739224e-05,
"loss": 1.5457,
"step": 990
},
{
"epoch": 1.78,
"grad_norm": 8.631403923034668,
"learning_rate": 4.619461698551315e-05,
"loss": 1.652,
"step": 1000
},
{
"epoch": 1.78,
"eval_loss": 1.5386379957199097,
"eval_runtime": 124.8384,
"eval_samples_per_second": 8.01,
"eval_steps_per_second": 2.003,
"step": 1000
},
{
"epoch": 1.8,
"grad_norm": 4.581122875213623,
"learning_rate": 4.612017082851253e-05,
"loss": 1.5746,
"step": 1010
},
{
"epoch": 1.81,
"grad_norm": 3.0373165607452393,
"learning_rate": 4.604506470270403e-05,
"loss": 1.6038,
"step": 1020
},
{
"epoch": 1.83,
"grad_norm": 3.5066914558410645,
"learning_rate": 4.5969300955024167e-05,
"loss": 1.5725,
"step": 1030
},
{
"epoch": 1.85,
"grad_norm": 4.402235507965088,
"learning_rate": 4.589288195295901e-05,
"loss": 1.5469,
"step": 1040
},
{
"epoch": 1.87,
"grad_norm": 4.844370365142822,
"learning_rate": 4.58158100844702e-05,
"loss": 1.5424,
"step": 1050
},
{
"epoch": 1.88,
"grad_norm": 4.146657943725586,
"learning_rate": 4.573808775792033e-05,
"loss": 1.4878,
"step": 1060
},
{
"epoch": 1.9,
"grad_norm": 3.210528612136841,
"learning_rate": 4.5659717401997655e-05,
"loss": 1.6077,
"step": 1070
},
{
"epoch": 1.92,
"grad_norm": 5.2232818603515625,
"learning_rate": 4.5580701465640254e-05,
"loss": 1.4824,
"step": 1080
},
{
"epoch": 1.94,
"grad_norm": 2.8741068840026855,
"learning_rate": 4.550104241795946e-05,
"loss": 1.6172,
"step": 1090
},
{
"epoch": 1.96,
"grad_norm": 8.092519760131836,
"learning_rate": 4.5420742748162734e-05,
"loss": 1.3659,
"step": 1100
},
{
"epoch": 1.96,
"eval_loss": 1.5198711156845093,
"eval_runtime": 124.8546,
"eval_samples_per_second": 8.009,
"eval_steps_per_second": 2.002,
"step": 1100
},
{
"epoch": 1.97,
"grad_norm": 5.068336009979248,
"learning_rate": 4.5339804965475875e-05,
"loss": 1.4661,
"step": 1110
},
{
"epoch": 1.99,
"grad_norm": 13.167552947998047,
"learning_rate": 4.525823159906459e-05,
"loss": 1.411,
"step": 1120
},
{
"epoch": 2.01,
"grad_norm": 4.712369918823242,
"learning_rate": 4.5176025197955494e-05,
"loss": 1.3309,
"step": 1130
},
{
"epoch": 2.03,
"grad_norm": 7.261610507965088,
"learning_rate": 4.509318833095642e-05,
"loss": 1.3892,
"step": 1140
},
{
"epoch": 2.04,
"grad_norm": 3.8006956577301025,
"learning_rate": 4.500972358657618e-05,
"loss": 1.3927,
"step": 1150
},
{
"epoch": 2.06,
"grad_norm": 3.6301958560943604,
"learning_rate": 4.492563357294369e-05,
"loss": 1.4629,
"step": 1160
},
{
"epoch": 2.08,
"grad_norm": 4.353027820587158,
"learning_rate": 4.4840920917726426e-05,
"loss": 1.352,
"step": 1170
},
{
"epoch": 2.1,
"grad_norm": 3.375173807144165,
"learning_rate": 4.475558826804833e-05,
"loss": 1.4096,
"step": 1180
},
{
"epoch": 2.12,
"grad_norm": 6.289668560028076,
"learning_rate": 4.466963829040712e-05,
"loss": 1.4834,
"step": 1190
},
{
"epoch": 2.13,
"grad_norm": 4.517002582550049,
"learning_rate": 4.458307367059092e-05,
"loss": 1.4746,
"step": 1200
},
{
"epoch": 2.13,
"eval_loss": 1.5145190954208374,
"eval_runtime": 124.8898,
"eval_samples_per_second": 8.007,
"eval_steps_per_second": 2.002,
"step": 1200
},
{
"epoch": 2.15,
"grad_norm": 3.195769786834717,
"learning_rate": 4.449589711359438e-05,
"loss": 1.4149,
"step": 1210
},
{
"epoch": 2.17,
"grad_norm": 3.751405715942383,
"learning_rate": 4.440811134353412e-05,
"loss": 1.5501,
"step": 1220
},
{
"epoch": 2.19,
"grad_norm": 4.148709774017334,
"learning_rate": 4.431971910356363e-05,
"loss": 1.5253,
"step": 1230
},
{
"epoch": 2.2,
"grad_norm": 20.003253936767578,
"learning_rate": 4.42307231557875e-05,
"loss": 1.6413,
"step": 1240
},
{
"epoch": 2.22,
"grad_norm": 4.721023082733154,
"learning_rate": 4.414112628117517e-05,
"loss": 1.5608,
"step": 1250
},
{
"epoch": 2.24,
"grad_norm": 4.672358989715576,
"learning_rate": 4.4050931279474015e-05,
"loss": 1.3646,
"step": 1260
},
{
"epoch": 2.26,
"grad_norm": 4.073034286499023,
"learning_rate": 4.396014096912182e-05,
"loss": 1.3499,
"step": 1270
},
{
"epoch": 2.28,
"grad_norm": 3.2312991619110107,
"learning_rate": 4.386875818715874e-05,
"loss": 1.4648,
"step": 1280
},
{
"epoch": 2.29,
"grad_norm": 18.92267417907715,
"learning_rate": 4.3776785789138675e-05,
"loss": 1.4919,
"step": 1290
},
{
"epoch": 2.31,
"grad_norm": 5.677367687225342,
"learning_rate": 4.368422664903997e-05,
"loss": 1.2891,
"step": 1300
},
{
"epoch": 2.31,
"eval_loss": 1.504623532295227,
"eval_runtime": 124.8541,
"eval_samples_per_second": 8.009,
"eval_steps_per_second": 2.002,
"step": 1300
},
{
"epoch": 2.33,
"grad_norm": 5.031940460205078,
"learning_rate": 4.359108365917565e-05,
"loss": 1.4939,
"step": 1310
},
{
"epoch": 2.35,
"grad_norm": 7.701929092407227,
"learning_rate": 4.349735973010305e-05,
"loss": 1.28,
"step": 1320
},
{
"epoch": 2.36,
"grad_norm": 5.7498040199279785,
"learning_rate": 4.3403057790532855e-05,
"loss": 1.4584,
"step": 1330
},
{
"epoch": 2.38,
"grad_norm": 8.7277193069458,
"learning_rate": 4.330818078723755e-05,
"loss": 1.5871,
"step": 1340
},
{
"epoch": 2.4,
"grad_norm": 13.915125846862793,
"learning_rate": 4.32127316849594e-05,
"loss": 1.3794,
"step": 1350
},
{
"epoch": 2.42,
"grad_norm": 2.949733018875122,
"learning_rate": 4.311671346631774e-05,
"loss": 1.3543,
"step": 1360
},
{
"epoch": 2.44,
"grad_norm": 5.377658843994141,
"learning_rate": 4.302012913171584e-05,
"loss": 1.3695,
"step": 1370
},
{
"epoch": 2.45,
"grad_norm": 16.94107437133789,
"learning_rate": 4.292298169924709e-05,
"loss": 1.5168,
"step": 1380
},
{
"epoch": 2.47,
"grad_norm": 4.190367221832275,
"learning_rate": 4.282527420460072e-05,
"loss": 1.4058,
"step": 1390
},
{
"epoch": 2.49,
"grad_norm": 9.269573211669922,
"learning_rate": 4.272700970096696e-05,
"loss": 1.5794,
"step": 1400
},
{
"epoch": 2.49,
"eval_loss": 1.498180627822876,
"eval_runtime": 124.7222,
"eval_samples_per_second": 8.018,
"eval_steps_per_second": 2.004,
"step": 1400
},
{
"epoch": 2.51,
"grad_norm": 3.951293468475342,
"learning_rate": 4.262819125894156e-05,
"loss": 1.56,
"step": 1410
},
{
"epoch": 2.52,
"grad_norm": 3.8725697994232178,
"learning_rate": 4.252882196642992e-05,
"loss": 1.5159,
"step": 1420
},
{
"epoch": 2.54,
"grad_norm": 3.898501396179199,
"learning_rate": 4.242890492855056e-05,
"loss": 1.4659,
"step": 1430
},
{
"epoch": 2.56,
"grad_norm": 5.807662487030029,
"learning_rate": 4.23284432675381e-05,
"loss": 1.5736,
"step": 1440
},
{
"epoch": 2.58,
"grad_norm": 3.529371500015259,
"learning_rate": 4.222744012264566e-05,
"loss": 1.5011,
"step": 1450
},
{
"epoch": 2.6,
"grad_norm": 6.336548805236816,
"learning_rate": 4.212589865004684e-05,
"loss": 1.6629,
"step": 1460
},
{
"epoch": 2.61,
"grad_norm": 6.222330093383789,
"learning_rate": 4.2023822022737016e-05,
"loss": 1.5573,
"step": 1470
},
{
"epoch": 2.63,
"grad_norm": 4.25172233581543,
"learning_rate": 4.192121343043424e-05,
"loss": 1.3817,
"step": 1480
},
{
"epoch": 2.65,
"grad_norm": 4.487111568450928,
"learning_rate": 4.181807607947954e-05,
"loss": 1.5323,
"step": 1490
},
{
"epoch": 2.67,
"grad_norm": 4.656155109405518,
"learning_rate": 4.1714413192736754e-05,
"loss": 1.3678,
"step": 1500
},
{
"epoch": 2.67,
"eval_loss": 1.5049968957901,
"eval_runtime": 124.7803,
"eval_samples_per_second": 8.014,
"eval_steps_per_second": 2.004,
"step": 1500
},
{
"epoch": 2.68,
"grad_norm": 4.431355953216553,
"learning_rate": 4.161022800949177e-05,
"loss": 1.486,
"step": 1510
},
{
"epoch": 2.7,
"grad_norm": 18.211524963378906,
"learning_rate": 4.150552378535137e-05,
"loss": 1.4498,
"step": 1520
},
{
"epoch": 2.72,
"grad_norm": 5.3755292892456055,
"learning_rate": 4.140030379214147e-05,
"loss": 1.4421,
"step": 1530
},
{
"epoch": 2.74,
"grad_norm": 6.626212120056152,
"learning_rate": 4.1294571317804854e-05,
"loss": 1.4322,
"step": 1540
},
{
"epoch": 2.76,
"grad_norm": 4.030793190002441,
"learning_rate": 4.1188329666298464e-05,
"loss": 1.3433,
"step": 1550
},
{
"epoch": 2.77,
"grad_norm": 6.53309440612793,
"learning_rate": 4.108158215749014e-05,
"loss": 1.5604,
"step": 1560
},
{
"epoch": 2.79,
"grad_norm": 3.76047420501709,
"learning_rate": 4.0974332127054914e-05,
"loss": 1.3259,
"step": 1570
},
{
"epoch": 2.81,
"grad_norm": 4.58742094039917,
"learning_rate": 4.0866582926370725e-05,
"loss": 1.4228,
"step": 1580
},
{
"epoch": 2.83,
"grad_norm": 4.566816806793213,
"learning_rate": 4.0758337922413716e-05,
"loss": 1.3013,
"step": 1590
},
{
"epoch": 2.84,
"grad_norm": 6.218478202819824,
"learning_rate": 4.064960049765304e-05,
"loss": 1.5061,
"step": 1600
},
{
"epoch": 2.84,
"eval_loss": 1.4853577613830566,
"eval_runtime": 124.7889,
"eval_samples_per_second": 8.014,
"eval_steps_per_second": 2.003,
"step": 1600
},
{
"epoch": 2.86,
"grad_norm": 13.811309814453125,
"learning_rate": 4.054037404994516e-05,
"loss": 1.4839,
"step": 1610
},
{
"epoch": 2.88,
"grad_norm": 5.560975074768066,
"learning_rate": 4.043066199242762e-05,
"loss": 1.4765,
"step": 1620
},
{
"epoch": 2.9,
"grad_norm": 35.27302551269531,
"learning_rate": 4.032046775341247e-05,
"loss": 1.4105,
"step": 1630
},
{
"epoch": 2.92,
"grad_norm": 4.9896745681762695,
"learning_rate": 4.020979477627907e-05,
"loss": 1.5688,
"step": 1640
},
{
"epoch": 2.93,
"grad_norm": 3.5250892639160156,
"learning_rate": 4.0098646519366534e-05,
"loss": 1.4484,
"step": 1650
},
{
"epoch": 2.95,
"grad_norm": 5.281729698181152,
"learning_rate": 3.998702645586565e-05,
"loss": 1.6017,
"step": 1660
},
{
"epoch": 2.97,
"grad_norm": 4.667525768280029,
"learning_rate": 3.9874938073710336e-05,
"loss": 1.5006,
"step": 1670
},
{
"epoch": 2.99,
"grad_norm": 4.294438362121582,
"learning_rate": 3.976238487546864e-05,
"loss": 1.4218,
"step": 1680
},
{
"epoch": 3.0,
"grad_norm": 4.070734977722168,
"learning_rate": 3.9649370378233365e-05,
"loss": 1.6569,
"step": 1690
},
{
"epoch": 3.02,
"grad_norm": 4.640359878540039,
"learning_rate": 3.953589811351204e-05,
"loss": 1.5635,
"step": 1700
},
{
"epoch": 3.02,
"eval_loss": 1.478974461555481,
"eval_runtime": 124.6852,
"eval_samples_per_second": 8.02,
"eval_steps_per_second": 2.005,
"step": 1700
},
{
"epoch": 3.04,
"grad_norm": 4.43009090423584,
"learning_rate": 3.94219716271167e-05,
"loss": 1.3304,
"step": 1710
},
{
"epoch": 3.06,
"grad_norm": 4.001712799072266,
"learning_rate": 3.930759447905298e-05,
"loss": 1.3534,
"step": 1720
},
{
"epoch": 3.08,
"grad_norm": 4.664085388183594,
"learning_rate": 3.919277024340891e-05,
"loss": 1.368,
"step": 1730
},
{
"epoch": 3.09,
"grad_norm": 4.42681360244751,
"learning_rate": 3.907750250824327e-05,
"loss": 1.4164,
"step": 1740
},
{
"epoch": 3.11,
"grad_norm": 7.331808567047119,
"learning_rate": 3.8961794875473394e-05,
"loss": 1.4333,
"step": 1750
},
{
"epoch": 3.13,
"grad_norm": 5.612239837646484,
"learning_rate": 3.884565096076269e-05,
"loss": 1.5754,
"step": 1760
},
{
"epoch": 3.15,
"grad_norm": 5.236481666564941,
"learning_rate": 3.872907439340758e-05,
"loss": 1.4017,
"step": 1770
},
{
"epoch": 3.16,
"grad_norm": 4.995403289794922,
"learning_rate": 3.861206881622419e-05,
"loss": 1.5011,
"step": 1780
},
{
"epoch": 3.18,
"grad_norm": 41.0167236328125,
"learning_rate": 3.8494637885434396e-05,
"loss": 1.4472,
"step": 1790
},
{
"epoch": 3.2,
"grad_norm": 5.136650562286377,
"learning_rate": 3.837678527055168e-05,
"loss": 1.3939,
"step": 1800
},
{
"epoch": 3.2,
"eval_loss": 1.4779504537582397,
"eval_runtime": 124.6728,
"eval_samples_per_second": 8.021,
"eval_steps_per_second": 2.005,
"step": 1800
},
{
"epoch": 3.22,
"grad_norm": 5.178096294403076,
"learning_rate": 3.8258514654266434e-05,
"loss": 1.5265,
"step": 1810
},
{
"epoch": 3.24,
"grad_norm": 6.949739456176758,
"learning_rate": 3.813982973233083e-05,
"loss": 1.3674,
"step": 1820
},
{
"epoch": 3.25,
"grad_norm": 3.84801983833313,
"learning_rate": 3.802073421344339e-05,
"loss": 1.4305,
"step": 1830
},
{
"epoch": 3.27,
"grad_norm": 3.5803613662719727,
"learning_rate": 3.7901231819133105e-05,
"loss": 1.557,
"step": 1840
},
{
"epoch": 3.29,
"grad_norm": 3.509099245071411,
"learning_rate": 3.7781326283643085e-05,
"loss": 1.3611,
"step": 1850
},
{
"epoch": 3.31,
"grad_norm": 41.185279846191406,
"learning_rate": 3.766102135381393e-05,
"loss": 1.3944,
"step": 1860
},
{
"epoch": 3.32,
"grad_norm": 3.797672748565674,
"learning_rate": 3.75403207889666e-05,
"loss": 1.2557,
"step": 1870
},
{
"epoch": 3.34,
"grad_norm": 4.237602233886719,
"learning_rate": 3.741922836078499e-05,
"loss": 1.3583,
"step": 1880
},
{
"epoch": 3.36,
"grad_norm": 4.5037312507629395,
"learning_rate": 3.729774785319801e-05,
"loss": 1.4619,
"step": 1890
},
{
"epoch": 3.38,
"grad_norm": 4.292742729187012,
"learning_rate": 3.717588306226143e-05,
"loss": 1.3986,
"step": 1900
},
{
"epoch": 3.38,
"eval_loss": 1.4685039520263672,
"eval_runtime": 124.7126,
"eval_samples_per_second": 8.018,
"eval_steps_per_second": 2.005,
"step": 1900
},
{
"epoch": 3.4,
"grad_norm": 5.198277950286865,
"learning_rate": 3.705363779603917e-05,
"loss": 1.349,
"step": 1910
},
{
"epoch": 3.41,
"grad_norm": 3.86722469329834,
"learning_rate": 3.693101587448436e-05,
"loss": 1.5053,
"step": 1920
},
{
"epoch": 3.43,
"grad_norm": 8.68099594116211,
"learning_rate": 3.680802112931996e-05,
"loss": 1.3899,
"step": 1930
},
{
"epoch": 3.45,
"grad_norm": 3.8347198963165283,
"learning_rate": 3.6684657403919005e-05,
"loss": 1.4519,
"step": 1940
},
{
"epoch": 3.47,
"grad_norm": 9.875212669372559,
"learning_rate": 3.6560928553184554e-05,
"loss": 1.4788,
"step": 1950
},
{
"epoch": 3.48,
"grad_norm": 8.638535499572754,
"learning_rate": 3.6436838443429175e-05,
"loss": 1.3777,
"step": 1960
},
{
"epoch": 3.5,
"grad_norm": 3.73545503616333,
"learning_rate": 3.631239095225417e-05,
"loss": 1.4962,
"step": 1970
},
{
"epoch": 3.52,
"grad_norm": 8.485962867736816,
"learning_rate": 3.618758996842839e-05,
"loss": 1.4377,
"step": 1980
},
{
"epoch": 3.54,
"grad_norm": 4.3264055252075195,
"learning_rate": 3.60624393917667e-05,
"loss": 1.5695,
"step": 1990
},
{
"epoch": 3.56,
"grad_norm": 3.979128837585449,
"learning_rate": 3.5936943133008183e-05,
"loss": 1.2959,
"step": 2000
},
{
"epoch": 3.56,
"eval_loss": 1.4598569869995117,
"eval_runtime": 124.8036,
"eval_samples_per_second": 8.013,
"eval_steps_per_second": 2.003,
"step": 2000
}
],
"logging_steps": 10,
"max_steps": 5620,
"num_input_tokens_seen": 0,
"num_train_epochs": 10,
"save_steps": 100,
"total_flos": 1.558141521683153e+18,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}