tinyllama-1.1b-mt-sft-full / trainer_state.json
martimfasantos's picture
Model save
a776c48 verified
raw
history blame contribute delete
No virus
49.6 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 1478,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0013531799729364006,
"grad_norm": 22.625,
"learning_rate": 1.3513513513513515e-07,
"loss": 2.1988,
"step": 1
},
{
"epoch": 0.006765899864682003,
"grad_norm": 21.875,
"learning_rate": 6.756756756756758e-07,
"loss": 2.2059,
"step": 5
},
{
"epoch": 0.013531799729364006,
"grad_norm": 22.5,
"learning_rate": 1.3513513513513515e-06,
"loss": 2.2336,
"step": 10
},
{
"epoch": 0.02029769959404601,
"grad_norm": 16.25,
"learning_rate": 2.0270270270270273e-06,
"loss": 2.244,
"step": 15
},
{
"epoch": 0.02706359945872801,
"grad_norm": 8.6875,
"learning_rate": 2.702702702702703e-06,
"loss": 2.181,
"step": 20
},
{
"epoch": 0.03382949932341001,
"grad_norm": 6.03125,
"learning_rate": 3.3783783783783788e-06,
"loss": 2.1568,
"step": 25
},
{
"epoch": 0.04059539918809202,
"grad_norm": 5.15625,
"learning_rate": 4.0540540540540545e-06,
"loss": 2.1029,
"step": 30
},
{
"epoch": 0.04736129905277402,
"grad_norm": 4.4375,
"learning_rate": 4.72972972972973e-06,
"loss": 2.0545,
"step": 35
},
{
"epoch": 0.05412719891745602,
"grad_norm": 3.734375,
"learning_rate": 5.405405405405406e-06,
"loss": 2.0671,
"step": 40
},
{
"epoch": 0.06089309878213803,
"grad_norm": 3.546875,
"learning_rate": 6.081081081081082e-06,
"loss": 2.0335,
"step": 45
},
{
"epoch": 0.06765899864682003,
"grad_norm": 3.671875,
"learning_rate": 6.7567567567567575e-06,
"loss": 1.9884,
"step": 50
},
{
"epoch": 0.07442489851150202,
"grad_norm": 4.84375,
"learning_rate": 7.4324324324324324e-06,
"loss": 1.9429,
"step": 55
},
{
"epoch": 0.08119079837618404,
"grad_norm": 3.609375,
"learning_rate": 8.108108108108109e-06,
"loss": 1.8937,
"step": 60
},
{
"epoch": 0.08795669824086604,
"grad_norm": 3.21875,
"learning_rate": 8.783783783783785e-06,
"loss": 1.8555,
"step": 65
},
{
"epoch": 0.09472259810554803,
"grad_norm": 3.015625,
"learning_rate": 9.45945945945946e-06,
"loss": 1.8032,
"step": 70
},
{
"epoch": 0.10148849797023005,
"grad_norm": 2.921875,
"learning_rate": 1.0135135135135136e-05,
"loss": 1.8064,
"step": 75
},
{
"epoch": 0.10825439783491204,
"grad_norm": 2.90625,
"learning_rate": 1.0810810810810812e-05,
"loss": 1.8009,
"step": 80
},
{
"epoch": 0.11502029769959404,
"grad_norm": 2.859375,
"learning_rate": 1.1486486486486488e-05,
"loss": 1.7624,
"step": 85
},
{
"epoch": 0.12178619756427606,
"grad_norm": 2.796875,
"learning_rate": 1.2162162162162164e-05,
"loss": 1.7737,
"step": 90
},
{
"epoch": 0.12855209742895804,
"grad_norm": 2.8125,
"learning_rate": 1.283783783783784e-05,
"loss": 1.7472,
"step": 95
},
{
"epoch": 0.13531799729364005,
"grad_norm": 2.609375,
"learning_rate": 1.3513513513513515e-05,
"loss": 1.7465,
"step": 100
},
{
"epoch": 0.14208389715832206,
"grad_norm": 2.75,
"learning_rate": 1.4189189189189189e-05,
"loss": 1.7326,
"step": 105
},
{
"epoch": 0.14884979702300405,
"grad_norm": 2.5625,
"learning_rate": 1.4864864864864865e-05,
"loss": 1.7463,
"step": 110
},
{
"epoch": 0.15561569688768606,
"grad_norm": 2.609375,
"learning_rate": 1.554054054054054e-05,
"loss": 1.7175,
"step": 115
},
{
"epoch": 0.16238159675236807,
"grad_norm": 2.671875,
"learning_rate": 1.6216216216216218e-05,
"loss": 1.6941,
"step": 120
},
{
"epoch": 0.16914749661705006,
"grad_norm": 2.609375,
"learning_rate": 1.6891891891891896e-05,
"loss": 1.6953,
"step": 125
},
{
"epoch": 0.17591339648173207,
"grad_norm": 2.578125,
"learning_rate": 1.756756756756757e-05,
"loss": 1.7159,
"step": 130
},
{
"epoch": 0.18267929634641408,
"grad_norm": 2.671875,
"learning_rate": 1.8243243243243244e-05,
"loss": 1.7148,
"step": 135
},
{
"epoch": 0.18944519621109607,
"grad_norm": 2.6875,
"learning_rate": 1.891891891891892e-05,
"loss": 1.6967,
"step": 140
},
{
"epoch": 0.19621109607577808,
"grad_norm": 2.578125,
"learning_rate": 1.9594594594594595e-05,
"loss": 1.7066,
"step": 145
},
{
"epoch": 0.2029769959404601,
"grad_norm": 2.578125,
"learning_rate": 1.9999888409903948e-05,
"loss": 1.7105,
"step": 150
},
{
"epoch": 0.20974289580514208,
"grad_norm": 2.65625,
"learning_rate": 1.9998633049924693e-05,
"loss": 1.6862,
"step": 155
},
{
"epoch": 0.2165087956698241,
"grad_norm": 2.46875,
"learning_rate": 1.999598301803528e-05,
"loss": 1.6791,
"step": 160
},
{
"epoch": 0.2232746955345061,
"grad_norm": 2.8125,
"learning_rate": 1.9991938683878746e-05,
"loss": 1.6925,
"step": 165
},
{
"epoch": 0.23004059539918809,
"grad_norm": 2.515625,
"learning_rate": 1.9986500611584133e-05,
"loss": 1.6846,
"step": 170
},
{
"epoch": 0.2368064952638701,
"grad_norm": 2.421875,
"learning_rate": 1.997966955968779e-05,
"loss": 1.6782,
"step": 175
},
{
"epoch": 0.2435723951285521,
"grad_norm": 2.734375,
"learning_rate": 1.997144648102759e-05,
"loss": 1.6852,
"step": 180
},
{
"epoch": 0.2503382949932341,
"grad_norm": 2.53125,
"learning_rate": 1.9961832522610004e-05,
"loss": 1.7028,
"step": 185
},
{
"epoch": 0.2571041948579161,
"grad_norm": 2.59375,
"learning_rate": 1.9950829025450116e-05,
"loss": 1.6483,
"step": 190
},
{
"epoch": 0.2638700947225981,
"grad_norm": 2.453125,
"learning_rate": 1.9938437524384572e-05,
"loss": 1.6498,
"step": 195
},
{
"epoch": 0.2706359945872801,
"grad_norm": 2.59375,
"learning_rate": 1.9924659747857485e-05,
"loss": 1.6601,
"step": 200
},
{
"epoch": 0.2774018944519621,
"grad_norm": 2.5,
"learning_rate": 1.990949761767935e-05,
"loss": 1.6657,
"step": 205
},
{
"epoch": 0.28416779431664413,
"grad_norm": 2.609375,
"learning_rate": 1.989295324875897e-05,
"loss": 1.676,
"step": 210
},
{
"epoch": 0.29093369418132614,
"grad_norm": 2.515625,
"learning_rate": 1.9875028948808457e-05,
"loss": 1.65,
"step": 215
},
{
"epoch": 0.2976995940460081,
"grad_norm": 2.484375,
"learning_rate": 1.985572721802134e-05,
"loss": 1.648,
"step": 220
},
{
"epoch": 0.3044654939106901,
"grad_norm": 2.46875,
"learning_rate": 1.9835050748723826e-05,
"loss": 1.6537,
"step": 225
},
{
"epoch": 0.3112313937753721,
"grad_norm": 2.453125,
"learning_rate": 1.981300242499924e-05,
"loss": 1.6459,
"step": 230
},
{
"epoch": 0.31799729364005414,
"grad_norm": 2.5625,
"learning_rate": 1.978958532228576e-05,
"loss": 1.6383,
"step": 235
},
{
"epoch": 0.32476319350473615,
"grad_norm": 2.578125,
"learning_rate": 1.9764802706947423e-05,
"loss": 1.6624,
"step": 240
},
{
"epoch": 0.33152909336941816,
"grad_norm": 2.53125,
"learning_rate": 1.9738658035818495e-05,
"loss": 1.6607,
"step": 245
},
{
"epoch": 0.3382949932341001,
"grad_norm": 2.375,
"learning_rate": 1.9711154955721338e-05,
"loss": 1.6527,
"step": 250
},
{
"epoch": 0.34506089309878213,
"grad_norm": 2.59375,
"learning_rate": 1.9682297302957666e-05,
"loss": 1.6423,
"step": 255
},
{
"epoch": 0.35182679296346414,
"grad_norm": 2.5,
"learning_rate": 1.9652089102773487e-05,
"loss": 1.6288,
"step": 260
},
{
"epoch": 0.35859269282814615,
"grad_norm": 2.4375,
"learning_rate": 1.962053456879761e-05,
"loss": 1.6436,
"step": 265
},
{
"epoch": 0.36535859269282817,
"grad_norm": 2.46875,
"learning_rate": 1.95876381024539e-05,
"loss": 1.6378,
"step": 270
},
{
"epoch": 0.3721244925575101,
"grad_norm": 2.59375,
"learning_rate": 1.9553404292347356e-05,
"loss": 1.6508,
"step": 275
},
{
"epoch": 0.37889039242219213,
"grad_norm": 2.578125,
"learning_rate": 1.9517837913624048e-05,
"loss": 1.6475,
"step": 280
},
{
"epoch": 0.38565629228687415,
"grad_norm": 2.546875,
"learning_rate": 1.948094392730506e-05,
"loss": 1.6353,
"step": 285
},
{
"epoch": 0.39242219215155616,
"grad_norm": 2.40625,
"learning_rate": 1.9442727479594486e-05,
"loss": 1.6464,
"step": 290
},
{
"epoch": 0.39918809201623817,
"grad_norm": 2.328125,
"learning_rate": 1.9403193901161614e-05,
"loss": 1.6157,
"step": 295
},
{
"epoch": 0.4059539918809202,
"grad_norm": 2.5625,
"learning_rate": 1.9362348706397374e-05,
"loss": 1.643,
"step": 300
},
{
"epoch": 0.41271989174560214,
"grad_norm": 2.5,
"learning_rate": 1.932019759264514e-05,
"loss": 1.6225,
"step": 305
},
{
"epoch": 0.41948579161028415,
"grad_norm": 2.625,
"learning_rate": 1.9276746439406046e-05,
"loss": 1.6139,
"step": 310
},
{
"epoch": 0.42625169147496617,
"grad_norm": 2.59375,
"learning_rate": 1.923200130751887e-05,
"loss": 1.6127,
"step": 315
},
{
"epoch": 0.4330175913396482,
"grad_norm": 2.53125,
"learning_rate": 1.918596843831462e-05,
"loss": 1.6245,
"step": 320
},
{
"epoch": 0.4397834912043302,
"grad_norm": 2.84375,
"learning_rate": 1.913865425274597e-05,
"loss": 1.6153,
"step": 325
},
{
"epoch": 0.4465493910690122,
"grad_norm": 2.484375,
"learning_rate": 1.909006535049163e-05,
"loss": 1.6101,
"step": 330
},
{
"epoch": 0.45331529093369416,
"grad_norm": 2.453125,
"learning_rate": 1.9040208509035745e-05,
"loss": 1.5968,
"step": 335
},
{
"epoch": 0.46008119079837617,
"grad_norm": 2.421875,
"learning_rate": 1.8989090682722583e-05,
"loss": 1.6055,
"step": 340
},
{
"epoch": 0.4668470906630582,
"grad_norm": 2.484375,
"learning_rate": 1.8936719001786453e-05,
"loss": 1.6253,
"step": 345
},
{
"epoch": 0.4736129905277402,
"grad_norm": 2.625,
"learning_rate": 1.888310077135716e-05,
"loss": 1.6369,
"step": 350
},
{
"epoch": 0.4803788903924222,
"grad_norm": 2.609375,
"learning_rate": 1.8828243470441026e-05,
"loss": 1.6359,
"step": 355
},
{
"epoch": 0.4871447902571042,
"grad_norm": 2.453125,
"learning_rate": 1.8772154750877696e-05,
"loss": 1.613,
"step": 360
},
{
"epoch": 0.4939106901217862,
"grad_norm": 2.421875,
"learning_rate": 1.8714842436272774e-05,
"loss": 1.603,
"step": 365
},
{
"epoch": 0.5006765899864682,
"grad_norm": 2.484375,
"learning_rate": 1.865631452090657e-05,
"loss": 1.5804,
"step": 370
},
{
"epoch": 0.5074424898511503,
"grad_norm": 2.953125,
"learning_rate": 1.859657916861899e-05,
"loss": 1.62,
"step": 375
},
{
"epoch": 0.5142083897158322,
"grad_norm": 2.390625,
"learning_rate": 1.8535644711670804e-05,
"loss": 1.6201,
"step": 380
},
{
"epoch": 0.5209742895805142,
"grad_norm": 2.421875,
"learning_rate": 1.8473519649581396e-05,
"loss": 1.6105,
"step": 385
},
{
"epoch": 0.5277401894451962,
"grad_norm": 2.640625,
"learning_rate": 1.8410212647943215e-05,
"loss": 1.6231,
"step": 390
},
{
"epoch": 0.5345060893098782,
"grad_norm": 2.46875,
"learning_rate": 1.834573253721303e-05,
"loss": 1.5854,
"step": 395
},
{
"epoch": 0.5412719891745602,
"grad_norm": 2.46875,
"learning_rate": 1.8280088311480203e-05,
"loss": 1.5792,
"step": 400
},
{
"epoch": 0.5480378890392422,
"grad_norm": 2.546875,
"learning_rate": 1.8213289127212152e-05,
"loss": 1.6033,
"step": 405
},
{
"epoch": 0.5548037889039242,
"grad_norm": 2.53125,
"learning_rate": 1.8145344301977126e-05,
"loss": 1.5986,
"step": 410
},
{
"epoch": 0.5615696887686062,
"grad_norm": 2.484375,
"learning_rate": 1.8076263313144568e-05,
"loss": 1.6238,
"step": 415
},
{
"epoch": 0.5683355886332883,
"grad_norm": 2.53125,
"learning_rate": 1.8006055796563103e-05,
"loss": 1.5919,
"step": 420
},
{
"epoch": 0.5751014884979703,
"grad_norm": 2.390625,
"learning_rate": 1.7934731545216515e-05,
"loss": 1.6125,
"step": 425
},
{
"epoch": 0.5818673883626523,
"grad_norm": 2.46875,
"learning_rate": 1.7862300507857733e-05,
"loss": 1.6184,
"step": 430
},
{
"epoch": 0.5886332882273342,
"grad_norm": 2.4375,
"learning_rate": 1.7788772787621126e-05,
"loss": 1.6034,
"step": 435
},
{
"epoch": 0.5953991880920162,
"grad_norm": 2.53125,
"learning_rate": 1.771415864061326e-05,
"loss": 1.5829,
"step": 440
},
{
"epoch": 0.6021650879566982,
"grad_norm": 2.421875,
"learning_rate": 1.7638468474482297e-05,
"loss": 1.6085,
"step": 445
},
{
"epoch": 0.6089309878213802,
"grad_norm": 2.515625,
"learning_rate": 1.756171284696629e-05,
"loss": 1.6123,
"step": 450
},
{
"epoch": 0.6156968876860622,
"grad_norm": 2.46875,
"learning_rate": 1.7483902464420507e-05,
"loss": 1.6034,
"step": 455
},
{
"epoch": 0.6224627875507442,
"grad_norm": 2.46875,
"learning_rate": 1.7405048180324046e-05,
"loss": 1.5858,
"step": 460
},
{
"epoch": 0.6292286874154263,
"grad_norm": 2.5,
"learning_rate": 1.7325160993765934e-05,
"loss": 1.6065,
"step": 465
},
{
"epoch": 0.6359945872801083,
"grad_norm": 2.515625,
"learning_rate": 1.7244252047910893e-05,
"loss": 1.5963,
"step": 470
},
{
"epoch": 0.6427604871447903,
"grad_norm": 2.484375,
"learning_rate": 1.7162332628445024e-05,
"loss": 1.594,
"step": 475
},
{
"epoch": 0.6495263870094723,
"grad_norm": 2.484375,
"learning_rate": 1.7079414162001617e-05,
"loss": 1.5899,
"step": 480
},
{
"epoch": 0.6562922868741543,
"grad_norm": 2.484375,
"learning_rate": 1.6995508214567275e-05,
"loss": 1.5786,
"step": 485
},
{
"epoch": 0.6630581867388363,
"grad_norm": 2.484375,
"learning_rate": 1.691062648986865e-05,
"loss": 1.5962,
"step": 490
},
{
"epoch": 0.6698240866035182,
"grad_norm": 2.625,
"learning_rate": 1.682478082773989e-05,
"loss": 1.5961,
"step": 495
},
{
"epoch": 0.6765899864682002,
"grad_norm": 2.46875,
"learning_rate": 1.673798320247118e-05,
"loss": 1.6023,
"step": 500
},
{
"epoch": 0.6833558863328822,
"grad_norm": 2.40625,
"learning_rate": 1.6650245721138483e-05,
"loss": 1.6074,
"step": 505
},
{
"epoch": 0.6901217861975643,
"grad_norm": 2.421875,
"learning_rate": 1.6561580621914764e-05,
"loss": 1.5767,
"step": 510
},
{
"epoch": 0.6968876860622463,
"grad_norm": 2.46875,
"learning_rate": 1.6472000272362937e-05,
"loss": 1.6047,
"step": 515
},
{
"epoch": 0.7036535859269283,
"grad_norm": 2.53125,
"learning_rate": 1.6381517167710757e-05,
"loss": 1.5846,
"step": 520
},
{
"epoch": 0.7104194857916103,
"grad_norm": 2.5625,
"learning_rate": 1.629014392910791e-05,
"loss": 1.5756,
"step": 525
},
{
"epoch": 0.7171853856562923,
"grad_norm": 2.453125,
"learning_rate": 1.619789330186555e-05,
"loss": 1.5966,
"step": 530
},
{
"epoch": 0.7239512855209743,
"grad_norm": 2.53125,
"learning_rate": 1.6104778153678467e-05,
"loss": 1.5901,
"step": 535
},
{
"epoch": 0.7307171853856563,
"grad_norm": 2.515625,
"learning_rate": 1.6010811472830253e-05,
"loss": 1.5712,
"step": 540
},
{
"epoch": 0.7374830852503383,
"grad_norm": 2.5,
"learning_rate": 1.591600636638161e-05,
"loss": 1.5877,
"step": 545
},
{
"epoch": 0.7442489851150202,
"grad_norm": 2.453125,
"learning_rate": 1.5820376058342077e-05,
"loss": 1.591,
"step": 550
},
{
"epoch": 0.7510148849797023,
"grad_norm": 2.453125,
"learning_rate": 1.5723933887825492e-05,
"loss": 1.5814,
"step": 555
},
{
"epoch": 0.7577807848443843,
"grad_norm": 2.609375,
"learning_rate": 1.5626693307189334e-05,
"loss": 1.5764,
"step": 560
},
{
"epoch": 0.7645466847090663,
"grad_norm": 2.484375,
"learning_rate": 1.5528667880158338e-05,
"loss": 1.5726,
"step": 565
},
{
"epoch": 0.7713125845737483,
"grad_norm": 2.484375,
"learning_rate": 1.5429871279932514e-05,
"loss": 1.5913,
"step": 570
},
{
"epoch": 0.7780784844384303,
"grad_norm": 2.5,
"learning_rate": 1.533031728727994e-05,
"loss": 1.5674,
"step": 575
},
{
"epoch": 0.7848443843031123,
"grad_norm": 2.421875,
"learning_rate": 1.5230019788614527e-05,
"loss": 1.5727,
"step": 580
},
{
"epoch": 0.7916102841677943,
"grad_norm": 2.4375,
"learning_rate": 1.5128992774059063e-05,
"loss": 1.5762,
"step": 585
},
{
"epoch": 0.7983761840324763,
"grad_norm": 2.484375,
"learning_rate": 1.5027250335493771e-05,
"loss": 1.5678,
"step": 590
},
{
"epoch": 0.8051420838971584,
"grad_norm": 2.5,
"learning_rate": 1.4924806664590702e-05,
"loss": 1.5602,
"step": 595
},
{
"epoch": 0.8119079837618404,
"grad_norm": 2.5625,
"learning_rate": 1.4821676050834166e-05,
"loss": 1.5761,
"step": 600
},
{
"epoch": 0.8186738836265224,
"grad_norm": 2.4375,
"learning_rate": 1.4717872879527578e-05,
"loss": 1.5751,
"step": 605
},
{
"epoch": 0.8254397834912043,
"grad_norm": 3.96875,
"learning_rate": 1.461341162978688e-05,
"loss": 1.5783,
"step": 610
},
{
"epoch": 0.8322056833558863,
"grad_norm": 2.46875,
"learning_rate": 1.4508306872520912e-05,
"loss": 1.5859,
"step": 615
},
{
"epoch": 0.8389715832205683,
"grad_norm": 2.515625,
"learning_rate": 1.4402573268398969e-05,
"loss": 1.5898,
"step": 620
},
{
"epoch": 0.8457374830852503,
"grad_norm": 2.421875,
"learning_rate": 1.4296225565805854e-05,
"loss": 1.5665,
"step": 625
},
{
"epoch": 0.8525033829499323,
"grad_norm": 2.421875,
"learning_rate": 1.4189278598784648e-05,
"loss": 1.5729,
"step": 630
},
{
"epoch": 0.8592692828146143,
"grad_norm": 2.421875,
"learning_rate": 1.4081747284967602e-05,
"loss": 1.5769,
"step": 635
},
{
"epoch": 0.8660351826792964,
"grad_norm": 2.34375,
"learning_rate": 1.3973646623495305e-05,
"loss": 1.5618,
"step": 640
},
{
"epoch": 0.8728010825439784,
"grad_norm": 2.4375,
"learning_rate": 1.3864991692924524e-05,
"loss": 1.589,
"step": 645
},
{
"epoch": 0.8795669824086604,
"grad_norm": 2.484375,
"learning_rate": 1.3755797649124944e-05,
"loss": 1.5733,
"step": 650
},
{
"epoch": 0.8863328822733424,
"grad_norm": 2.453125,
"learning_rate": 1.3646079723165148e-05,
"loss": 1.5596,
"step": 655
},
{
"epoch": 0.8930987821380244,
"grad_norm": 2.5,
"learning_rate": 1.3535853219188064e-05,
"loss": 1.5722,
"step": 660
},
{
"epoch": 0.8998646820027063,
"grad_norm": 2.484375,
"learning_rate": 1.3425133512276284e-05,
"loss": 1.5554,
"step": 665
},
{
"epoch": 0.9066305818673883,
"grad_norm": 2.453125,
"learning_rate": 1.3313936046307411e-05,
"loss": 1.576,
"step": 670
},
{
"epoch": 0.9133964817320703,
"grad_norm": 2.453125,
"learning_rate": 1.320227633179989e-05,
"loss": 1.5668,
"step": 675
},
{
"epoch": 0.9201623815967523,
"grad_norm": 2.25,
"learning_rate": 1.3090169943749475e-05,
"loss": 1.5757,
"step": 680
},
{
"epoch": 0.9269282814614344,
"grad_norm": 2.515625,
"learning_rate": 1.2977632519456745e-05,
"loss": 1.5817,
"step": 685
},
{
"epoch": 0.9336941813261164,
"grad_norm": 2.359375,
"learning_rate": 1.2864679756345905e-05,
"loss": 1.5847,
"step": 690
},
{
"epoch": 0.9404600811907984,
"grad_norm": 2.484375,
"learning_rate": 1.2751327409775227e-05,
"loss": 1.5847,
"step": 695
},
{
"epoch": 0.9472259810554804,
"grad_norm": 2.625,
"learning_rate": 1.2637591290839377e-05,
"loss": 1.5695,
"step": 700
},
{
"epoch": 0.9539918809201624,
"grad_norm": 2.4375,
"learning_rate": 1.2523487264163997e-05,
"loss": 1.566,
"step": 705
},
{
"epoch": 0.9607577807848444,
"grad_norm": 2.34375,
"learning_rate": 1.2409031245692798e-05,
"loss": 1.5487,
"step": 710
},
{
"epoch": 0.9675236806495264,
"grad_norm": 2.453125,
"learning_rate": 1.2294239200467516e-05,
"loss": 1.5825,
"step": 715
},
{
"epoch": 0.9742895805142084,
"grad_norm": 2.546875,
"learning_rate": 1.2179127140400997e-05,
"loss": 1.5651,
"step": 720
},
{
"epoch": 0.9810554803788903,
"grad_norm": 2.40625,
"learning_rate": 1.2063711122043759e-05,
"loss": 1.5487,
"step": 725
},
{
"epoch": 0.9878213802435724,
"grad_norm": 2.46875,
"learning_rate": 1.1948007244344334e-05,
"loss": 1.5604,
"step": 730
},
{
"epoch": 0.9945872801082544,
"grad_norm": 2.375,
"learning_rate": 1.1832031646403654e-05,
"loss": 1.5838,
"step": 735
},
{
"epoch": 1.0,
"eval_loss": 1.6891789436340332,
"eval_runtime": 28.8376,
"eval_samples_per_second": 15.084,
"eval_steps_per_second": 1.907,
"step": 739
},
{
"epoch": 1.0013531799729365,
"grad_norm": 2.265625,
"learning_rate": 1.1715800505223918e-05,
"loss": 1.5323,
"step": 740
},
{
"epoch": 1.0081190798376185,
"grad_norm": 2.265625,
"learning_rate": 1.1599330033452078e-05,
"loss": 1.4855,
"step": 745
},
{
"epoch": 1.0148849797023005,
"grad_norm": 2.4375,
"learning_rate": 1.148263647711842e-05,
"loss": 1.531,
"step": 750
},
{
"epoch": 1.0216508795669823,
"grad_norm": 2.390625,
"learning_rate": 1.1365736113370463e-05,
"loss": 1.5063,
"step": 755
},
{
"epoch": 1.0284167794316643,
"grad_norm": 2.515625,
"learning_rate": 1.124864524820251e-05,
"loss": 1.5219,
"step": 760
},
{
"epoch": 1.0351826792963463,
"grad_norm": 2.515625,
"learning_rate": 1.1131380214181205e-05,
"loss": 1.4972,
"step": 765
},
{
"epoch": 1.0419485791610283,
"grad_norm": 2.328125,
"learning_rate": 1.1013957368167343e-05,
"loss": 1.5092,
"step": 770
},
{
"epoch": 1.0487144790257104,
"grad_norm": 2.4375,
"learning_rate": 1.0896393089034336e-05,
"loss": 1.4743,
"step": 775
},
{
"epoch": 1.0554803788903924,
"grad_norm": 2.3125,
"learning_rate": 1.0778703775383559e-05,
"loss": 1.4991,
"step": 780
},
{
"epoch": 1.0622462787550744,
"grad_norm": 2.328125,
"learning_rate": 1.0660905843256995e-05,
"loss": 1.4727,
"step": 785
},
{
"epoch": 1.0690121786197564,
"grad_norm": 2.390625,
"learning_rate": 1.0543015723847402e-05,
"loss": 1.4992,
"step": 790
},
{
"epoch": 1.0757780784844384,
"grad_norm": 2.375,
"learning_rate": 1.042504986120641e-05,
"loss": 1.5143,
"step": 795
},
{
"epoch": 1.0825439783491204,
"grad_norm": 2.453125,
"learning_rate": 1.0307024709950775e-05,
"loss": 1.5027,
"step": 800
},
{
"epoch": 1.0893098782138024,
"grad_norm": 2.25,
"learning_rate": 1.0188956732967208e-05,
"loss": 1.4863,
"step": 805
},
{
"epoch": 1.0960757780784844,
"grad_norm": 2.34375,
"learning_rate": 1.0070862399116016e-05,
"loss": 1.5161,
"step": 810
},
{
"epoch": 1.1028416779431665,
"grad_norm": 2.390625,
"learning_rate": 9.952758180933933e-06,
"loss": 1.5031,
"step": 815
},
{
"epoch": 1.1096075778078485,
"grad_norm": 2.40625,
"learning_rate": 9.834660552336415e-06,
"loss": 1.5017,
"step": 820
},
{
"epoch": 1.1163734776725305,
"grad_norm": 2.359375,
"learning_rate": 9.716585986319769e-06,
"loss": 1.4958,
"step": 825
},
{
"epoch": 1.1231393775372125,
"grad_norm": 2.453125,
"learning_rate": 9.598550952663383e-06,
"loss": 1.5104,
"step": 830
},
{
"epoch": 1.1299052774018945,
"grad_norm": 2.46875,
"learning_rate": 9.480571915632422e-06,
"loss": 1.5208,
"step": 835
},
{
"epoch": 1.1366711772665765,
"grad_norm": 2.40625,
"learning_rate": 9.362665331681294e-06,
"loss": 1.5244,
"step": 840
},
{
"epoch": 1.1434370771312585,
"grad_norm": 2.359375,
"learning_rate": 9.244847647158203e-06,
"loss": 1.5114,
"step": 845
},
{
"epoch": 1.1502029769959405,
"grad_norm": 2.3125,
"learning_rate": 9.127135296011102e-06,
"loss": 1.5153,
"step": 850
},
{
"epoch": 1.1569688768606226,
"grad_norm": 2.359375,
"learning_rate": 9.009544697495373e-06,
"loss": 1.4826,
"step": 855
},
{
"epoch": 1.1637347767253043,
"grad_norm": 2.3125,
"learning_rate": 8.892092253883602e-06,
"loss": 1.5164,
"step": 860
},
{
"epoch": 1.1705006765899864,
"grad_norm": 2.59375,
"learning_rate": 8.774794348177641e-06,
"loss": 1.4995,
"step": 865
},
{
"epoch": 1.1772665764546684,
"grad_norm": 2.4375,
"learning_rate": 8.657667341823449e-06,
"loss": 1.5047,
"step": 870
},
{
"epoch": 1.1840324763193504,
"grad_norm": 2.46875,
"learning_rate": 8.540727572428854e-06,
"loss": 1.5063,
"step": 875
},
{
"epoch": 1.1907983761840324,
"grad_norm": 2.296875,
"learning_rate": 8.423991351484715e-06,
"loss": 1.4832,
"step": 880
},
{
"epoch": 1.1975642760487144,
"grad_norm": 2.359375,
"learning_rate": 8.307474962089676e-06,
"loss": 1.4901,
"step": 885
},
{
"epoch": 1.2043301759133964,
"grad_norm": 2.265625,
"learning_rate": 8.191194656678905e-06,
"loss": 1.5061,
"step": 890
},
{
"epoch": 1.2110960757780784,
"grad_norm": 2.46875,
"learning_rate": 8.07516665475708e-06,
"loss": 1.4967,
"step": 895
},
{
"epoch": 1.2178619756427604,
"grad_norm": 2.46875,
"learning_rate": 7.959407140636034e-06,
"loss": 1.5082,
"step": 900
},
{
"epoch": 1.2246278755074425,
"grad_norm": 2.390625,
"learning_rate": 7.843932261177224e-06,
"loss": 1.485,
"step": 905
},
{
"epoch": 1.2313937753721245,
"grad_norm": 2.328125,
"learning_rate": 7.728758123539498e-06,
"loss": 1.5061,
"step": 910
},
{
"epoch": 1.2381596752368065,
"grad_norm": 2.375,
"learning_rate": 7.613900792932331e-06,
"loss": 1.4953,
"step": 915
},
{
"epoch": 1.2449255751014885,
"grad_norm": 2.421875,
"learning_rate": 7.499376290374994e-06,
"loss": 1.493,
"step": 920
},
{
"epoch": 1.2516914749661705,
"grad_norm": 2.53125,
"learning_rate": 7.385200590461803e-06,
"loss": 1.506,
"step": 925
},
{
"epoch": 1.2584573748308525,
"grad_norm": 2.40625,
"learning_rate": 7.271389619133908e-06,
"loss": 1.5135,
"step": 930
},
{
"epoch": 1.2652232746955345,
"grad_norm": 2.34375,
"learning_rate": 7.1579592514578234e-06,
"loss": 1.5076,
"step": 935
},
{
"epoch": 1.2719891745602165,
"grad_norm": 2.359375,
"learning_rate": 7.0449253094110925e-06,
"loss": 1.486,
"step": 940
},
{
"epoch": 1.2787550744248986,
"grad_norm": 2.4375,
"learning_rate": 6.93230355967533e-06,
"loss": 1.5115,
"step": 945
},
{
"epoch": 1.2855209742895806,
"grad_norm": 2.484375,
"learning_rate": 6.820109711436989e-06,
"loss": 1.4928,
"step": 950
},
{
"epoch": 1.2922868741542626,
"grad_norm": 2.3125,
"learning_rate": 6.7083594141961326e-06,
"loss": 1.5128,
"step": 955
},
{
"epoch": 1.2990527740189446,
"grad_norm": 2.28125,
"learning_rate": 6.59706825558357e-06,
"loss": 1.5023,
"step": 960
},
{
"epoch": 1.3058186738836266,
"grad_norm": 2.3125,
"learning_rate": 6.486251759186573e-06,
"loss": 1.489,
"step": 965
},
{
"epoch": 1.3125845737483086,
"grad_norm": 2.328125,
"learning_rate": 6.375925382383561e-06,
"loss": 1.4976,
"step": 970
},
{
"epoch": 1.3193504736129906,
"grad_norm": 2.546875,
"learning_rate": 6.266104514187997e-06,
"loss": 1.5049,
"step": 975
},
{
"epoch": 1.3261163734776726,
"grad_norm": 2.3125,
"learning_rate": 6.156804473101852e-06,
"loss": 1.506,
"step": 980
},
{
"epoch": 1.3328822733423547,
"grad_norm": 2.4375,
"learning_rate": 6.048040504978861e-06,
"loss": 1.5012,
"step": 985
},
{
"epoch": 1.3396481732070367,
"grad_norm": 2.375,
"learning_rate": 5.93982778089796e-06,
"loss": 1.5174,
"step": 990
},
{
"epoch": 1.3464140730717187,
"grad_norm": 2.328125,
"learning_rate": 5.832181395047099e-06,
"loss": 1.5131,
"step": 995
},
{
"epoch": 1.3531799729364005,
"grad_norm": 2.515625,
"learning_rate": 5.72511636261784e-06,
"loss": 1.4879,
"step": 1000
},
{
"epoch": 1.3599458728010825,
"grad_norm": 2.390625,
"learning_rate": 5.618647617710935e-06,
"loss": 1.4899,
"step": 1005
},
{
"epoch": 1.3667117726657645,
"grad_norm": 2.453125,
"learning_rate": 5.512790011253211e-06,
"loss": 1.4927,
"step": 1010
},
{
"epoch": 1.3734776725304465,
"grad_norm": 2.375,
"learning_rate": 5.407558308926083e-06,
"loss": 1.5084,
"step": 1015
},
{
"epoch": 1.3802435723951285,
"grad_norm": 2.53125,
"learning_rate": 5.302967189105941e-06,
"loss": 1.4999,
"step": 1020
},
{
"epoch": 1.3870094722598105,
"grad_norm": 2.3125,
"learning_rate": 5.199031240816715e-06,
"loss": 1.4828,
"step": 1025
},
{
"epoch": 1.3937753721244925,
"grad_norm": 2.4375,
"learning_rate": 5.095764961694923e-06,
"loss": 1.4967,
"step": 1030
},
{
"epoch": 1.4005412719891746,
"grad_norm": 2.34375,
"learning_rate": 4.993182755967412e-06,
"loss": 1.4959,
"step": 1035
},
{
"epoch": 1.4073071718538566,
"grad_norm": 2.421875,
"learning_rate": 4.8912989324422164e-06,
"loss": 1.4905,
"step": 1040
},
{
"epoch": 1.4140730717185386,
"grad_norm": 2.359375,
"learning_rate": 4.7901277025126345e-06,
"loss": 1.5105,
"step": 1045
},
{
"epoch": 1.4208389715832206,
"grad_norm": 2.390625,
"learning_rate": 4.689683178174964e-06,
"loss": 1.5236,
"step": 1050
},
{
"epoch": 1.4276048714479026,
"grad_norm": 2.328125,
"learning_rate": 4.589979370060037e-06,
"loss": 1.4801,
"step": 1055
},
{
"epoch": 1.4343707713125846,
"grad_norm": 2.28125,
"learning_rate": 4.491030185478976e-06,
"loss": 1.5219,
"step": 1060
},
{
"epoch": 1.4411366711772666,
"grad_norm": 2.375,
"learning_rate": 4.392849426483275e-06,
"loss": 1.5072,
"step": 1065
},
{
"epoch": 1.4479025710419486,
"grad_norm": 2.3125,
"learning_rate": 4.295450787939622e-06,
"loss": 1.4897,
"step": 1070
},
{
"epoch": 1.4546684709066307,
"grad_norm": 2.46875,
"learning_rate": 4.198847855619652e-06,
"loss": 1.5098,
"step": 1075
},
{
"epoch": 1.4614343707713127,
"grad_norm": 2.515625,
"learning_rate": 4.1030541043049125e-06,
"loss": 1.5226,
"step": 1080
},
{
"epoch": 1.4682002706359945,
"grad_norm": 2.296875,
"learning_rate": 4.008082895907326e-06,
"loss": 1.4999,
"step": 1085
},
{
"epoch": 1.4749661705006765,
"grad_norm": 2.453125,
"learning_rate": 3.913947477605378e-06,
"loss": 1.5023,
"step": 1090
},
{
"epoch": 1.4817320703653585,
"grad_norm": 2.5,
"learning_rate": 3.8206609799963205e-06,
"loss": 1.5144,
"step": 1095
},
{
"epoch": 1.4884979702300405,
"grad_norm": 2.375,
"learning_rate": 3.72823641526463e-06,
"loss": 1.5185,
"step": 1100
},
{
"epoch": 1.4952638700947225,
"grad_norm": 2.390625,
"learning_rate": 3.636686675367006e-06,
"loss": 1.4987,
"step": 1105
},
{
"epoch": 1.5020297699594045,
"grad_norm": 2.484375,
"learning_rate": 3.5460245302340914e-06,
"loss": 1.5168,
"step": 1110
},
{
"epoch": 1.5087956698240865,
"grad_norm": 2.34375,
"learning_rate": 3.4562626259892605e-06,
"loss": 1.4996,
"step": 1115
},
{
"epoch": 1.5155615696887685,
"grad_norm": 2.328125,
"learning_rate": 3.367413483184654e-06,
"loss": 1.5008,
"step": 1120
},
{
"epoch": 1.5223274695534506,
"grad_norm": 2.359375,
"learning_rate": 3.279489495054742e-06,
"loss": 1.4882,
"step": 1125
},
{
"epoch": 1.5290933694181326,
"grad_norm": 2.296875,
"learning_rate": 3.1925029257876206e-06,
"loss": 1.4648,
"step": 1130
},
{
"epoch": 1.5358592692828146,
"grad_norm": 2.5,
"learning_rate": 3.1064659088143424e-06,
"loss": 1.5048,
"step": 1135
},
{
"epoch": 1.5426251691474966,
"grad_norm": 2.359375,
"learning_rate": 3.0213904451164623e-06,
"loss": 1.5161,
"step": 1140
},
{
"epoch": 1.5493910690121786,
"grad_norm": 2.421875,
"learning_rate": 2.937288401552063e-06,
"loss": 1.5037,
"step": 1145
},
{
"epoch": 1.5561569688768606,
"grad_norm": 2.296875,
"learning_rate": 2.8541715092005097e-06,
"loss": 1.5058,
"step": 1150
},
{
"epoch": 1.5629228687415426,
"grad_norm": 2.421875,
"learning_rate": 2.7720513617260857e-06,
"loss": 1.5019,
"step": 1155
},
{
"epoch": 1.5696887686062246,
"grad_norm": 2.859375,
"learning_rate": 2.690939413760887e-06,
"loss": 1.5006,
"step": 1160
},
{
"epoch": 1.5764546684709067,
"grad_norm": 2.40625,
"learning_rate": 2.610846979307016e-06,
"loss": 1.5078,
"step": 1165
},
{
"epoch": 1.5832205683355887,
"grad_norm": 2.546875,
"learning_rate": 2.5317852301584642e-06,
"loss": 1.4985,
"step": 1170
},
{
"epoch": 1.5899864682002707,
"grad_norm": 2.34375,
"learning_rate": 2.4537651943427666e-06,
"loss": 1.4675,
"step": 1175
},
{
"epoch": 1.5967523680649527,
"grad_norm": 2.4375,
"learning_rate": 2.376797754582785e-06,
"loss": 1.4989,
"step": 1180
},
{
"epoch": 1.6035182679296347,
"grad_norm": 2.359375,
"learning_rate": 2.3008936467786815e-06,
"loss": 1.4755,
"step": 1185
},
{
"epoch": 1.6102841677943167,
"grad_norm": 2.359375,
"learning_rate": 2.226063458510428e-06,
"loss": 1.4977,
"step": 1190
},
{
"epoch": 1.6170500676589987,
"grad_norm": 2.3125,
"learning_rate": 2.152317627560979e-06,
"loss": 1.4965,
"step": 1195
},
{
"epoch": 1.6238159675236807,
"grad_norm": 2.34375,
"learning_rate": 2.0796664404603416e-06,
"loss": 1.4801,
"step": 1200
},
{
"epoch": 1.6305818673883627,
"grad_norm": 2.421875,
"learning_rate": 2.008120031050753e-06,
"loss": 1.5145,
"step": 1205
},
{
"epoch": 1.6373477672530448,
"grad_norm": 2.46875,
"learning_rate": 1.9376883790731417e-06,
"loss": 1.5032,
"step": 1210
},
{
"epoch": 1.6441136671177268,
"grad_norm": 2.484375,
"learning_rate": 1.8683813087750901e-06,
"loss": 1.4968,
"step": 1215
},
{
"epoch": 1.6508795669824088,
"grad_norm": 2.375,
"learning_rate": 1.8002084875404935e-06,
"loss": 1.512,
"step": 1220
},
{
"epoch": 1.6576454668470908,
"grad_norm": 2.40625,
"learning_rate": 1.7331794245410926e-06,
"loss": 1.5059,
"step": 1225
},
{
"epoch": 1.6644113667117728,
"grad_norm": 2.484375,
"learning_rate": 1.6673034694100655e-06,
"loss": 1.5124,
"step": 1230
},
{
"epoch": 1.6711772665764548,
"grad_norm": 2.296875,
"learning_rate": 1.6025898109378967e-06,
"loss": 1.5248,
"step": 1235
},
{
"epoch": 1.6779431664411368,
"grad_norm": 2.265625,
"learning_rate": 1.5390474757906449e-06,
"loss": 1.5044,
"step": 1240
},
{
"epoch": 1.6847090663058186,
"grad_norm": 2.40625,
"learning_rate": 1.4766853272508785e-06,
"loss": 1.5081,
"step": 1245
},
{
"epoch": 1.6914749661705006,
"grad_norm": 2.375,
"learning_rate": 1.4155120639813392e-06,
"loss": 1.4982,
"step": 1250
},
{
"epoch": 1.6982408660351827,
"grad_norm": 2.375,
"learning_rate": 1.3555362188116173e-06,
"loss": 1.4716,
"step": 1255
},
{
"epoch": 1.7050067658998647,
"grad_norm": 2.484375,
"learning_rate": 1.2967661575479318e-06,
"loss": 1.5055,
"step": 1260
},
{
"epoch": 1.7117726657645467,
"grad_norm": 2.328125,
"learning_rate": 1.2392100778062123e-06,
"loss": 1.5066,
"step": 1265
},
{
"epoch": 1.7185385656292287,
"grad_norm": 2.359375,
"learning_rate": 1.1828760078686563e-06,
"loss": 1.5146,
"step": 1270
},
{
"epoch": 1.7253044654939107,
"grad_norm": 2.34375,
"learning_rate": 1.127771805563882e-06,
"loss": 1.5009,
"step": 1275
},
{
"epoch": 1.7320703653585927,
"grad_norm": 2.4375,
"learning_rate": 1.0739051571708736e-06,
"loss": 1.5149,
"step": 1280
},
{
"epoch": 1.7388362652232747,
"grad_norm": 2.359375,
"learning_rate": 1.0212835763468488e-06,
"loss": 1.5055,
"step": 1285
},
{
"epoch": 1.7456021650879567,
"grad_norm": 2.328125,
"learning_rate": 9.699144030792163e-07,
"loss": 1.5028,
"step": 1290
},
{
"epoch": 1.7523680649526387,
"grad_norm": 2.34375,
"learning_rate": 9.198048026617323e-07,
"loss": 1.4877,
"step": 1295
},
{
"epoch": 1.7591339648173205,
"grad_norm": 2.359375,
"learning_rate": 8.709617646950563e-07,
"loss": 1.5085,
"step": 1300
},
{
"epoch": 1.7658998646820026,
"grad_norm": 2.203125,
"learning_rate": 8.233921021117863e-07,
"loss": 1.4859,
"step": 1305
},
{
"epoch": 1.7726657645466846,
"grad_norm": 2.40625,
"learning_rate": 7.771024502261526e-07,
"loss": 1.5006,
"step": 1310
},
{
"epoch": 1.7794316644113666,
"grad_norm": 2.359375,
"learning_rate": 7.320992658084891e-07,
"loss": 1.5159,
"step": 1315
},
{
"epoch": 1.7861975642760486,
"grad_norm": 2.421875,
"learning_rate": 6.883888261845917e-07,
"loss": 1.5117,
"step": 1320
},
{
"epoch": 1.7929634641407306,
"grad_norm": 2.4375,
"learning_rate": 6.459772283601218e-07,
"loss": 1.5026,
"step": 1325
},
{
"epoch": 1.7997293640054126,
"grad_norm": 2.46875,
"learning_rate": 6.048703881701579e-07,
"loss": 1.4971,
"step": 1330
},
{
"epoch": 1.8064952638700946,
"grad_norm": 2.40625,
"learning_rate": 5.650740394540255e-07,
"loss": 1.4969,
"step": 1335
},
{
"epoch": 1.8132611637347766,
"grad_norm": 2.359375,
"learning_rate": 5.265937332554849e-07,
"loss": 1.4996,
"step": 1340
},
{
"epoch": 1.8200270635994586,
"grad_norm": 2.375,
"learning_rate": 4.894348370484648e-07,
"loss": 1.5124,
"step": 1345
},
{
"epoch": 1.8267929634641407,
"grad_norm": 2.5,
"learning_rate": 4.5360253398834765e-07,
"loss": 1.5047,
"step": 1350
},
{
"epoch": 1.8335588633288227,
"grad_norm": 2.359375,
"learning_rate": 4.1910182218900977e-07,
"loss": 1.5211,
"step": 1355
},
{
"epoch": 1.8403247631935047,
"grad_norm": 2.34375,
"learning_rate": 3.8593751402563715e-07,
"loss": 1.5189,
"step": 1360
},
{
"epoch": 1.8470906630581867,
"grad_norm": 2.40625,
"learning_rate": 3.541142354634808e-07,
"loss": 1.4941,
"step": 1365
},
{
"epoch": 1.8538565629228687,
"grad_norm": 2.5,
"learning_rate": 3.236364254125868e-07,
"loss": 1.47,
"step": 1370
},
{
"epoch": 1.8606224627875507,
"grad_norm": 2.421875,
"learning_rate": 2.945083351086331e-07,
"loss": 1.5138,
"step": 1375
},
{
"epoch": 1.8673883626522327,
"grad_norm": 2.328125,
"learning_rate": 2.667340275199426e-07,
"loss": 1.5025,
"step": 1380
},
{
"epoch": 1.8741542625169147,
"grad_norm": 2.5,
"learning_rate": 2.4031737678074985e-07,
"loss": 1.5216,
"step": 1385
},
{
"epoch": 1.8809201623815968,
"grad_norm": 2.296875,
"learning_rate": 2.1526206765082104e-07,
"loss": 1.5059,
"step": 1390
},
{
"epoch": 1.8876860622462788,
"grad_norm": 2.390625,
"learning_rate": 1.915715950014696e-07,
"loss": 1.4874,
"step": 1395
},
{
"epoch": 1.8944519621109608,
"grad_norm": 2.515625,
"learning_rate": 1.6924926332807956e-07,
"loss": 1.5109,
"step": 1400
},
{
"epoch": 1.9012178619756428,
"grad_norm": 2.3125,
"learning_rate": 1.4829818628916525e-07,
"loss": 1.4955,
"step": 1405
},
{
"epoch": 1.9079837618403248,
"grad_norm": 2.25,
"learning_rate": 1.287212862720677e-07,
"loss": 1.5023,
"step": 1410
},
{
"epoch": 1.9147496617050068,
"grad_norm": 2.390625,
"learning_rate": 1.1052129398531508e-07,
"loss": 1.4818,
"step": 1415
},
{
"epoch": 1.9215155615696888,
"grad_norm": 2.25,
"learning_rate": 9.370074807772966e-08,
"loss": 1.4966,
"step": 1420
},
{
"epoch": 1.9282814614343708,
"grad_norm": 2.375,
"learning_rate": 7.826199478431551e-08,
"loss": 1.4801,
"step": 1425
},
{
"epoch": 1.9350473612990529,
"grad_norm": 2.390625,
"learning_rate": 6.420718759900357e-08,
"loss": 1.5037,
"step": 1430
},
{
"epoch": 1.9418132611637349,
"grad_norm": 2.28125,
"learning_rate": 5.1538286974254225e-08,
"loss": 1.4796,
"step": 1435
},
{
"epoch": 1.9485791610284169,
"grad_norm": 2.421875,
"learning_rate": 4.025706004760932e-08,
"loss": 1.5041,
"step": 1440
},
{
"epoch": 1.955345060893099,
"grad_norm": 2.359375,
"learning_rate": 3.036508039520047e-08,
"loss": 1.4933,
"step": 1445
},
{
"epoch": 1.962110960757781,
"grad_norm": 2.375,
"learning_rate": 2.1863727812254653e-08,
"loss": 1.4872,
"step": 1450
},
{
"epoch": 1.968876860622463,
"grad_norm": 2.4375,
"learning_rate": 1.4754188120631453e-08,
"loss": 1.4953,
"step": 1455
},
{
"epoch": 1.975642760487145,
"grad_norm": 2.296875,
"learning_rate": 9.03745300341874e-09,
"loss": 1.5199,
"step": 1460
},
{
"epoch": 1.982408660351827,
"grad_norm": 2.328125,
"learning_rate": 4.714319866607753e-09,
"loss": 1.519,
"step": 1465
},
{
"epoch": 1.989174560216509,
"grad_norm": 2.375,
"learning_rate": 1.7853917278631838e-09,
"loss": 1.5044,
"step": 1470
},
{
"epoch": 1.9959404600811907,
"grad_norm": 2.4375,
"learning_rate": 2.5107713241045994e-10,
"loss": 1.5051,
"step": 1475
},
{
"epoch": 2.0,
"eval_loss": 1.691962480545044,
"eval_runtime": 28.7807,
"eval_samples_per_second": 15.114,
"eval_steps_per_second": 1.911,
"step": 1478
},
{
"epoch": 2.0,
"step": 1478,
"total_flos": 1.502823063832494e+17,
"train_loss": 1.5822677802652403,
"train_runtime": 3974.9532,
"train_samples_per_second": 2.974,
"train_steps_per_second": 0.372
}
],
"logging_steps": 5,
"max_steps": 1478,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.502823063832494e+17,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}