zephyr-7b-gemma-sft-20p / trainer_state.json
Qin Liu
Model save
2503b9d verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.9955555555555557,
"eval_steps": 500,
"global_step": 1011,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"grad_norm": 72.2578351791039,
"learning_rate": 1.96078431372549e-06,
"loss": 16.2543,
"step": 1
},
{
"epoch": 0.01,
"grad_norm": 69.14793459091129,
"learning_rate": 9.803921568627451e-06,
"loss": 16.1546,
"step": 5
},
{
"epoch": 0.03,
"grad_norm": 61.19418424850778,
"learning_rate": 1.9607843137254903e-05,
"loss": 15.7851,
"step": 10
},
{
"epoch": 0.04,
"grad_norm": 50.95192336423662,
"learning_rate": 2.9411764705882354e-05,
"loss": 13.304,
"step": 15
},
{
"epoch": 0.06,
"grad_norm": 44.16161556810121,
"learning_rate": 3.9215686274509805e-05,
"loss": 10.1286,
"step": 20
},
{
"epoch": 0.07,
"grad_norm": 22.013360927764538,
"learning_rate": 4.901960784313725e-05,
"loss": 6.5136,
"step": 25
},
{
"epoch": 0.09,
"grad_norm": 13.13769527146034,
"learning_rate": 5.882352941176471e-05,
"loss": 4.6002,
"step": 30
},
{
"epoch": 0.1,
"grad_norm": 9.380927030189634,
"learning_rate": 6.862745098039216e-05,
"loss": 3.4775,
"step": 35
},
{
"epoch": 0.12,
"grad_norm": 8.047209493969948,
"learning_rate": 7.843137254901961e-05,
"loss": 2.5528,
"step": 40
},
{
"epoch": 0.13,
"grad_norm": 4.835270040905147,
"learning_rate": 8.823529411764706e-05,
"loss": 1.9181,
"step": 45
},
{
"epoch": 0.15,
"grad_norm": 4.461126160581573,
"learning_rate": 9.80392156862745e-05,
"loss": 1.7189,
"step": 50
},
{
"epoch": 0.16,
"grad_norm": 2.4713562833452647,
"learning_rate": 0.00010784313725490196,
"loss": 1.4816,
"step": 55
},
{
"epoch": 0.18,
"grad_norm": 3.006847832744875,
"learning_rate": 0.00011764705882352942,
"loss": 1.3855,
"step": 60
},
{
"epoch": 0.19,
"grad_norm": 1.6003996046085147,
"learning_rate": 0.00012745098039215687,
"loss": 1.2728,
"step": 65
},
{
"epoch": 0.21,
"grad_norm": 1.6221503977277323,
"learning_rate": 0.0001372549019607843,
"loss": 1.2497,
"step": 70
},
{
"epoch": 0.22,
"grad_norm": 1.584552829167114,
"learning_rate": 0.00014705882352941178,
"loss": 1.2432,
"step": 75
},
{
"epoch": 0.24,
"grad_norm": 1.5050158166503569,
"learning_rate": 0.00015686274509803922,
"loss": 1.1723,
"step": 80
},
{
"epoch": 0.25,
"grad_norm": 1.7107243802135035,
"learning_rate": 0.0001666666666666667,
"loss": 1.083,
"step": 85
},
{
"epoch": 0.27,
"grad_norm": 1.7381735421551534,
"learning_rate": 0.00017647058823529413,
"loss": 1.0726,
"step": 90
},
{
"epoch": 0.28,
"grad_norm": 1.4981023359065837,
"learning_rate": 0.00018627450980392157,
"loss": 1.1057,
"step": 95
},
{
"epoch": 0.3,
"grad_norm": 1.7877995266746756,
"learning_rate": 0.000196078431372549,
"loss": 1.0783,
"step": 100
},
{
"epoch": 0.31,
"grad_norm": 1.5140220512101432,
"learning_rate": 0.00019999462497359466,
"loss": 1.1228,
"step": 105
},
{
"epoch": 0.33,
"grad_norm": 1.329498665050575,
"learning_rate": 0.00019996177968249334,
"loss": 1.0441,
"step": 110
},
{
"epoch": 0.34,
"grad_norm": 1.7238891644878542,
"learning_rate": 0.00019989908502204292,
"loss": 1.1001,
"step": 115
},
{
"epoch": 0.36,
"grad_norm": 1.8266342354047433,
"learning_rate": 0.00019980655971335945,
"loss": 1.0601,
"step": 120
},
{
"epoch": 0.37,
"grad_norm": 1.6200107092414249,
"learning_rate": 0.0001996842313852238,
"loss": 1.0045,
"step": 125
},
{
"epoch": 0.39,
"grad_norm": 1.5172675508165814,
"learning_rate": 0.00019953213656583168,
"loss": 1.006,
"step": 130
},
{
"epoch": 0.4,
"grad_norm": 1.478313589538863,
"learning_rate": 0.0001993503206718859,
"loss": 0.9678,
"step": 135
},
{
"epoch": 0.41,
"grad_norm": 1.8797845296046498,
"learning_rate": 0.0001991388379950346,
"loss": 1.0181,
"step": 140
},
{
"epoch": 0.43,
"grad_norm": 1.8037331639091116,
"learning_rate": 0.00019889775168565943,
"loss": 1.0096,
"step": 145
},
{
"epoch": 0.44,
"grad_norm": 1.4294890809484986,
"learning_rate": 0.0001986271337340182,
"loss": 0.9842,
"step": 150
},
{
"epoch": 0.46,
"grad_norm": 1.4637597540944496,
"learning_rate": 0.0001983270649487481,
"loss": 0.9992,
"step": 155
},
{
"epoch": 0.47,
"grad_norm": 1.519557023555869,
"learning_rate": 0.0001979976349327357,
"loss": 1.036,
"step": 160
},
{
"epoch": 0.49,
"grad_norm": 1.6521810842892943,
"learning_rate": 0.00019763894205636072,
"loss": 0.9961,
"step": 165
},
{
"epoch": 0.5,
"grad_norm": 1.7888048328412252,
"learning_rate": 0.0001972510934281218,
"loss": 0.9796,
"step": 170
},
{
"epoch": 0.52,
"grad_norm": 1.4704141321403639,
"learning_rate": 0.00019683420486265327,
"loss": 0.9766,
"step": 175
},
{
"epoch": 0.53,
"grad_norm": 1.1929169794016272,
"learning_rate": 0.00019638840084614182,
"loss": 1.0146,
"step": 180
},
{
"epoch": 0.55,
"grad_norm": 1.1785071553887134,
"learning_rate": 0.00019591381449915397,
"loss": 1.0056,
"step": 185
},
{
"epoch": 0.56,
"grad_norm": 1.2250942438326087,
"learning_rate": 0.00019541058753688538,
"loss": 0.9809,
"step": 190
},
{
"epoch": 0.58,
"grad_norm": 1.4932596567461078,
"learning_rate": 0.00019487887022684336,
"loss": 0.9827,
"step": 195
},
{
"epoch": 0.59,
"grad_norm": 1.2885727190352654,
"learning_rate": 0.00019431882134397598,
"loss": 1.0027,
"step": 200
},
{
"epoch": 0.61,
"grad_norm": 2.57327759834574,
"learning_rate": 0.00019373060812326052,
"loss": 0.9351,
"step": 205
},
{
"epoch": 0.62,
"grad_norm": 1.3232509112686246,
"learning_rate": 0.00019311440620976597,
"loss": 0.9811,
"step": 210
},
{
"epoch": 0.64,
"grad_norm": 1.3160771870305465,
"learning_rate": 0.0001924703996062038,
"loss": 0.9925,
"step": 215
},
{
"epoch": 0.65,
"grad_norm": 1.2477864155169314,
"learning_rate": 0.00019179878061798347,
"loss": 0.9212,
"step": 220
},
{
"epoch": 0.67,
"grad_norm": 1.093292440232411,
"learning_rate": 0.0001910997497957885,
"loss": 0.9851,
"step": 225
},
{
"epoch": 0.68,
"grad_norm": 1.362427572762746,
"learning_rate": 0.0001903735158756905,
"loss": 0.9893,
"step": 230
},
{
"epoch": 0.7,
"grad_norm": 1.346631495980947,
"learning_rate": 0.00018962029571681886,
"loss": 0.945,
"step": 235
},
{
"epoch": 0.71,
"grad_norm": 1.090827112192976,
"learning_rate": 0.0001888403142366049,
"loss": 1.0224,
"step": 240
},
{
"epoch": 0.73,
"grad_norm": 1.6700272825279434,
"learning_rate": 0.00018803380434362,
"loss": 0.9305,
"step": 245
},
{
"epoch": 0.74,
"grad_norm": 1.0096253684038474,
"learning_rate": 0.00018720100686802694,
"loss": 1.0128,
"step": 250
},
{
"epoch": 0.76,
"grad_norm": 1.7873981051539485,
"learning_rate": 0.00018634217048966637,
"loss": 0.953,
"step": 255
},
{
"epoch": 0.77,
"grad_norm": 1.3129933811195174,
"learning_rate": 0.000185457551663799,
"loss": 0.9363,
"step": 260
},
{
"epoch": 0.79,
"grad_norm": 1.2927375510712122,
"learning_rate": 0.00018454741454452603,
"loss": 0.9862,
"step": 265
},
{
"epoch": 0.8,
"grad_norm": 1.1686920395458062,
"learning_rate": 0.00018361203090591071,
"loss": 0.967,
"step": 270
},
{
"epoch": 0.81,
"grad_norm": 1.1863754899949552,
"learning_rate": 0.00018265168006082437,
"loss": 0.9056,
"step": 275
},
{
"epoch": 0.83,
"grad_norm": 1.1941161746530229,
"learning_rate": 0.0001816666487775416,
"loss": 0.9791,
"step": 280
},
{
"epoch": 0.84,
"grad_norm": 1.1930577483488491,
"learning_rate": 0.00018065723119410884,
"loss": 0.9537,
"step": 285
},
{
"epoch": 0.86,
"grad_norm": 1.4589361109870367,
"learning_rate": 0.00017962372873051252,
"loss": 0.9954,
"step": 290
},
{
"epoch": 0.87,
"grad_norm": 1.3290947156035955,
"learning_rate": 0.00017856644999867264,
"loss": 0.922,
"step": 295
},
{
"epoch": 0.89,
"grad_norm": 1.2441967467176946,
"learning_rate": 0.000177485710710289,
"loss": 0.9639,
"step": 300
},
{
"epoch": 0.9,
"grad_norm": 1.1382757452000494,
"learning_rate": 0.00017638183358256696,
"loss": 0.9811,
"step": 305
},
{
"epoch": 0.92,
"grad_norm": 1.147116712031671,
"learning_rate": 0.00017525514824185185,
"loss": 0.9457,
"step": 310
},
{
"epoch": 0.93,
"grad_norm": 1.0220320338471351,
"learning_rate": 0.0001741059911251997,
"loss": 0.9233,
"step": 315
},
{
"epoch": 0.95,
"grad_norm": 1.1472331976712749,
"learning_rate": 0.00017293470537991463,
"loss": 0.9756,
"step": 320
},
{
"epoch": 0.96,
"grad_norm": 1.2410203245777724,
"learning_rate": 0.0001717416407610824,
"loss": 0.943,
"step": 325
},
{
"epoch": 0.98,
"grad_norm": 1.3981773401938695,
"learning_rate": 0.00017052715352713075,
"loss": 0.9477,
"step": 330
},
{
"epoch": 0.99,
"grad_norm": 1.2121344807882255,
"learning_rate": 0.0001692916063334479,
"loss": 0.9328,
"step": 335
},
{
"epoch": 1.0,
"eval_loss": 1.1519585847854614,
"eval_runtime": 157.1947,
"eval_samples_per_second": 14.695,
"eval_steps_per_second": 0.464,
"step": 337
},
{
"epoch": 1.01,
"grad_norm": 1.0908262897233953,
"learning_rate": 0.00016803536812409075,
"loss": 0.9661,
"step": 340
},
{
"epoch": 1.02,
"grad_norm": 1.7756060681085633,
"learning_rate": 0.00016675881402161536,
"loss": 0.889,
"step": 345
},
{
"epoch": 1.04,
"grad_norm": 1.1136805577794664,
"learning_rate": 0.0001654623252150624,
"loss": 0.9091,
"step": 350
},
{
"epoch": 1.05,
"grad_norm": 1.3395157303175598,
"learning_rate": 0.00016414628884613107,
"loss": 0.9285,
"step": 355
},
{
"epoch": 1.07,
"grad_norm": 1.1278036304088261,
"learning_rate": 0.0001628110978935756,
"loss": 0.8679,
"step": 360
},
{
"epoch": 1.08,
"grad_norm": 1.1866907922967995,
"learning_rate": 0.0001614571510558588,
"loss": 0.9031,
"step": 365
},
{
"epoch": 1.1,
"grad_norm": 1.1180348411654952,
"learning_rate": 0.00016008485263209742,
"loss": 0.9585,
"step": 370
},
{
"epoch": 1.11,
"grad_norm": 1.0617311418169892,
"learning_rate": 0.0001586946124013354,
"loss": 0.9212,
"step": 375
},
{
"epoch": 1.13,
"grad_norm": 0.9259620753456954,
"learning_rate": 0.00015728684550018064,
"loss": 0.8746,
"step": 380
},
{
"epoch": 1.14,
"grad_norm": 1.1221480241476138,
"learning_rate": 0.00015586197229884184,
"loss": 0.924,
"step": 385
},
{
"epoch": 1.16,
"grad_norm": 0.9589792763467693,
"learning_rate": 0.00015442041827560274,
"loss": 0.8658,
"step": 390
},
{
"epoch": 1.17,
"grad_norm": 1.4212740627448033,
"learning_rate": 0.00015296261388977108,
"loss": 0.8592,
"step": 395
},
{
"epoch": 1.19,
"grad_norm": 1.0480944734011124,
"learning_rate": 0.00015148899445313981,
"loss": 0.9497,
"step": 400
},
{
"epoch": 1.2,
"grad_norm": 1.0225393084718664,
"learning_rate": 0.00015000000000000001,
"loss": 0.9505,
"step": 405
},
{
"epoch": 1.21,
"grad_norm": 0.9776923989427987,
"learning_rate": 0.00014849607515574276,
"loss": 0.9189,
"step": 410
},
{
"epoch": 1.23,
"grad_norm": 0.9981587820990419,
"learning_rate": 0.00014697766900409074,
"loss": 0.9761,
"step": 415
},
{
"epoch": 1.24,
"grad_norm": 1.1240954604826308,
"learning_rate": 0.00014544523495299842,
"loss": 0.9174,
"step": 420
},
{
"epoch": 1.26,
"grad_norm": 1.4387424633384827,
"learning_rate": 0.00014389923059926062,
"loss": 0.905,
"step": 425
},
{
"epoch": 1.27,
"grad_norm": 1.3190393196712298,
"learning_rate": 0.00014234011759187083,
"loss": 0.9378,
"step": 430
},
{
"epoch": 1.29,
"grad_norm": 0.9858027932238713,
"learning_rate": 0.00014076836149416887,
"loss": 0.9164,
"step": 435
},
{
"epoch": 1.3,
"grad_norm": 0.8937095468783902,
"learning_rate": 0.00013918443164482046,
"loss": 0.9296,
"step": 440
},
{
"epoch": 1.32,
"grad_norm": 1.08163207719712,
"learning_rate": 0.0001375888010176686,
"loss": 0.914,
"step": 445
},
{
"epoch": 1.33,
"grad_norm": 1.0364463188268707,
"learning_rate": 0.0001359819460805001,
"loss": 0.8906,
"step": 450
},
{
"epoch": 1.35,
"grad_norm": 0.9266587461121177,
"learning_rate": 0.00013436434665276865,
"loss": 0.9395,
"step": 455
},
{
"epoch": 1.36,
"grad_norm": 1.0746148135081544,
"learning_rate": 0.0001327364857623168,
"loss": 0.9365,
"step": 460
},
{
"epoch": 1.38,
"grad_norm": 1.077761213751867,
"learning_rate": 0.00013109884950114007,
"loss": 0.9157,
"step": 465
},
{
"epoch": 1.39,
"grad_norm": 1.4149324327770756,
"learning_rate": 0.00012945192688023624,
"loss": 0.8864,
"step": 470
},
{
"epoch": 1.41,
"grad_norm": 1.0160083472695705,
"learning_rate": 0.00012779620968358273,
"loss": 0.8757,
"step": 475
},
{
"epoch": 1.42,
"grad_norm": 1.1590619362809342,
"learning_rate": 0.00012613219232128608,
"loss": 0.8985,
"step": 480
},
{
"epoch": 1.44,
"grad_norm": 1.1120230320518538,
"learning_rate": 0.00012446037168194714,
"loss": 0.9227,
"step": 485
},
{
"epoch": 1.45,
"grad_norm": 1.1834009436730764,
"learning_rate": 0.0001227812469842864,
"loss": 0.9204,
"step": 490
},
{
"epoch": 1.47,
"grad_norm": 1.0355132778907057,
"learning_rate": 0.00012109531962807332,
"loss": 0.889,
"step": 495
},
{
"epoch": 1.48,
"grad_norm": 1.0214392183969367,
"learning_rate": 0.00011940309304440433,
"loss": 0.8882,
"step": 500
},
{
"epoch": 1.5,
"grad_norm": 1.0496957066021488,
"learning_rate": 0.00011770507254537453,
"loss": 0.9334,
"step": 505
},
{
"epoch": 1.51,
"grad_norm": 1.0238376160697826,
"learning_rate": 0.00011600176517318741,
"loss": 0.9181,
"step": 510
},
{
"epoch": 1.53,
"grad_norm": 1.2008352860962,
"learning_rate": 0.00011429367954874819,
"loss": 0.9335,
"step": 515
},
{
"epoch": 1.54,
"grad_norm": 1.037759449688711,
"learning_rate": 0.00011258132571978555,
"loss": 0.9394,
"step": 520
},
{
"epoch": 1.56,
"grad_norm": 0.98060881335434,
"learning_rate": 0.00011086521500854745,
"loss": 0.9237,
"step": 525
},
{
"epoch": 1.57,
"grad_norm": 1.137978604803647,
"learning_rate": 0.00010914585985911632,
"loss": 0.873,
"step": 530
},
{
"epoch": 1.59,
"grad_norm": 1.0209942184078924,
"learning_rate": 0.00010742377368438914,
"loss": 0.989,
"step": 535
},
{
"epoch": 1.6,
"grad_norm": 1.0005848887814885,
"learning_rate": 0.00010569947071276847,
"loss": 0.9054,
"step": 540
},
{
"epoch": 1.61,
"grad_norm": 1.2509234265685367,
"learning_rate": 0.00010397346583460971,
"loss": 0.9068,
"step": 545
},
{
"epoch": 1.63,
"grad_norm": 1.118860273669544,
"learning_rate": 0.0001022462744484709,
"loss": 0.9185,
"step": 550
},
{
"epoch": 1.64,
"grad_norm": 0.9619304215105235,
"learning_rate": 0.00010051841230721065,
"loss": 0.9152,
"step": 555
},
{
"epoch": 1.66,
"grad_norm": 0.9312722485058544,
"learning_rate": 9.879039536398024e-05,
"loss": 0.9451,
"step": 560
},
{
"epoch": 1.67,
"grad_norm": 0.9758156277152181,
"learning_rate": 9.70627396181561e-05,
"loss": 0.9016,
"step": 565
},
{
"epoch": 1.69,
"grad_norm": 0.8860842880872017,
"learning_rate": 9.533596096125825e-05,
"loss": 0.9069,
"step": 570
},
{
"epoch": 1.7,
"grad_norm": 1.0727044167911228,
"learning_rate": 9.361057502290113e-05,
"loss": 0.8748,
"step": 575
},
{
"epoch": 1.72,
"grad_norm": 0.8634186584908737,
"learning_rate": 9.188709701682247e-05,
"loss": 0.893,
"step": 580
},
{
"epoch": 1.73,
"grad_norm": 1.1477573965693937,
"learning_rate": 9.016604158703654e-05,
"loss": 0.8893,
"step": 585
},
{
"epoch": 1.75,
"grad_norm": 0.9383547560329538,
"learning_rate": 8.844792265415738e-05,
"loss": 0.8651,
"step": 590
},
{
"epoch": 1.76,
"grad_norm": 1.1121008461123547,
"learning_rate": 8.673325326193806e-05,
"loss": 0.9379,
"step": 595
},
{
"epoch": 1.78,
"grad_norm": 1.0975479479362362,
"learning_rate": 8.502254542407186e-05,
"loss": 0.8481,
"step": 600
},
{
"epoch": 1.79,
"grad_norm": 1.113767179456745,
"learning_rate": 8.33163099713009e-05,
"loss": 0.8327,
"step": 605
},
{
"epoch": 1.81,
"grad_norm": 0.89921971591069,
"learning_rate": 8.161505639887817e-05,
"loss": 0.9102,
"step": 610
},
{
"epoch": 1.82,
"grad_norm": 1.0388536628309168,
"learning_rate": 7.991929271442817e-05,
"loss": 0.859,
"step": 615
},
{
"epoch": 1.84,
"grad_norm": 1.0521026013607777,
"learning_rate": 7.822952528625191e-05,
"loss": 0.8731,
"step": 620
},
{
"epoch": 1.85,
"grad_norm": 0.9421918430486114,
"learning_rate": 7.654625869212146e-05,
"loss": 0.9012,
"step": 625
},
{
"epoch": 1.87,
"grad_norm": 0.9484722686972686,
"learning_rate": 7.48699955686089e-05,
"loss": 0.9432,
"step": 630
},
{
"epoch": 1.88,
"grad_norm": 0.9768965209508316,
"learning_rate": 7.320123646099519e-05,
"loss": 0.8999,
"step": 635
},
{
"epoch": 1.9,
"grad_norm": 0.9453535215942106,
"learning_rate": 7.154047967380354e-05,
"loss": 0.9724,
"step": 640
},
{
"epoch": 1.91,
"grad_norm": 1.0424636870577482,
"learning_rate": 6.988822112200156e-05,
"loss": 0.8789,
"step": 645
},
{
"epoch": 1.93,
"grad_norm": 0.9224956346414365,
"learning_rate": 6.82449541829174e-05,
"loss": 0.8569,
"step": 650
},
{
"epoch": 1.94,
"grad_norm": 1.1516517831276118,
"learning_rate": 6.661116954891328e-05,
"loss": 0.9306,
"step": 655
},
{
"epoch": 1.96,
"grad_norm": 0.9883581608392334,
"learning_rate": 6.498735508086093e-05,
"loss": 0.871,
"step": 660
},
{
"epoch": 1.97,
"grad_norm": 0.9438307796156034,
"learning_rate": 6.337399566246257e-05,
"loss": 0.9302,
"step": 665
},
{
"epoch": 1.99,
"grad_norm": 0.8930439870325309,
"learning_rate": 6.177157305546078e-05,
"loss": 0.9477,
"step": 670
},
{
"epoch": 2.0,
"grad_norm": 0.9118609372127487,
"learning_rate": 6.018056575578075e-05,
"loss": 0.8771,
"step": 675
},
{
"epoch": 2.0,
"eval_loss": 1.1150912046432495,
"eval_runtime": 156.2319,
"eval_samples_per_second": 14.786,
"eval_steps_per_second": 0.467,
"step": 675
},
{
"epoch": 2.01,
"grad_norm": 0.828432991647057,
"learning_rate": 5.860144885064751e-05,
"loss": 0.8378,
"step": 680
},
{
"epoch": 2.03,
"grad_norm": 1.0448906737327117,
"learning_rate": 5.7034693876721376e-05,
"loss": 0.8421,
"step": 685
},
{
"epoch": 2.04,
"grad_norm": 1.0620773533713173,
"learning_rate": 5.54807686792933e-05,
"loss": 0.8794,
"step": 690
},
{
"epoch": 2.06,
"grad_norm": 1.0869539484563016,
"learning_rate": 5.394013727258254e-05,
"loss": 0.8679,
"step": 695
},
{
"epoch": 2.07,
"grad_norm": 0.9798911637537467,
"learning_rate": 5.2413259701178505e-05,
"loss": 0.8898,
"step": 700
},
{
"epoch": 2.09,
"grad_norm": 0.9047442392594395,
"learning_rate": 5.090059190266779e-05,
"loss": 0.8724,
"step": 705
},
{
"epoch": 2.1,
"grad_norm": 1.0781903357123614,
"learning_rate": 4.940258557148765e-05,
"loss": 0.8573,
"step": 710
},
{
"epoch": 2.12,
"grad_norm": 1.0342896185783423,
"learning_rate": 4.791968802404648e-05,
"loss": 0.8281,
"step": 715
},
{
"epoch": 2.13,
"grad_norm": 0.9191153541590816,
"learning_rate": 4.645234206515171e-05,
"loss": 0.9009,
"step": 720
},
{
"epoch": 2.15,
"grad_norm": 1.1845629044153207,
"learning_rate": 4.5000985855784746e-05,
"loss": 0.857,
"step": 725
},
{
"epoch": 2.16,
"grad_norm": 1.0434089629049308,
"learning_rate": 4.3566052782262735e-05,
"loss": 0.858,
"step": 730
},
{
"epoch": 2.18,
"grad_norm": 0.990812188911313,
"learning_rate": 4.2147971326825966e-05,
"loss": 0.8516,
"step": 735
},
{
"epoch": 2.19,
"grad_norm": 0.905736044179964,
"learning_rate": 4.074716493968975e-05,
"loss": 0.8476,
"step": 740
},
{
"epoch": 2.21,
"grad_norm": 1.016168128211076,
"learning_rate": 3.936405191259891e-05,
"loss": 0.8388,
"step": 745
},
{
"epoch": 2.22,
"grad_norm": 0.9657889846503783,
"learning_rate": 3.79990452539225e-05,
"loss": 0.8796,
"step": 750
},
{
"epoch": 2.24,
"grad_norm": 0.8488273704750955,
"learning_rate": 3.665255256532638e-05,
"loss": 0.8226,
"step": 755
},
{
"epoch": 2.25,
"grad_norm": 0.8515631446002137,
"learning_rate": 3.53249759200601e-05,
"loss": 0.8848,
"step": 760
},
{
"epoch": 2.27,
"grad_norm": 0.9625777666321612,
"learning_rate": 3.401671174289469e-05,
"loss": 0.8581,
"step": 765
},
{
"epoch": 2.28,
"grad_norm": 1.1027956646622936,
"learning_rate": 3.2728150691747115e-05,
"loss": 0.8205,
"step": 770
},
{
"epoch": 2.3,
"grad_norm": 0.959658890342703,
"learning_rate": 3.145967754102691e-05,
"loss": 0.8332,
"step": 775
},
{
"epoch": 2.31,
"grad_norm": 1.014525985812299,
"learning_rate": 3.021167106673928e-05,
"loss": 0.8485,
"step": 780
},
{
"epoch": 2.33,
"grad_norm": 1.0119872246269643,
"learning_rate": 2.898450393337977e-05,
"loss": 0.856,
"step": 785
},
{
"epoch": 2.34,
"grad_norm": 1.321701548482505,
"learning_rate": 2.7778542582653744e-05,
"loss": 0.8304,
"step": 790
},
{
"epoch": 2.36,
"grad_norm": 1.0899857889914895,
"learning_rate": 2.659414712405398e-05,
"loss": 0.8442,
"step": 795
},
{
"epoch": 2.37,
"grad_norm": 1.0994834827590023,
"learning_rate": 2.543167122732918e-05,
"loss": 0.8645,
"step": 800
},
{
"epoch": 2.39,
"grad_norm": 0.8030087173097793,
"learning_rate": 2.429146201687538e-05,
"loss": 0.8567,
"step": 805
},
{
"epoch": 2.4,
"grad_norm": 1.01607316930185,
"learning_rate": 2.3173859968081944e-05,
"loss": 0.9123,
"step": 810
},
{
"epoch": 2.41,
"grad_norm": 0.9294802520942274,
"learning_rate": 2.2079198805662914e-05,
"loss": 0.8374,
"step": 815
},
{
"epoch": 2.43,
"grad_norm": 0.9523609743858644,
"learning_rate": 2.1007805404004242e-05,
"loss": 0.8568,
"step": 820
},
{
"epoch": 2.44,
"grad_norm": 0.8957405654003763,
"learning_rate": 1.995999968955641e-05,
"loss": 0.8873,
"step": 825
},
{
"epoch": 2.46,
"grad_norm": 1.0700819784398359,
"learning_rate": 1.8936094545302095e-05,
"loss": 0.8564,
"step": 830
},
{
"epoch": 2.47,
"grad_norm": 1.0564758507529628,
"learning_rate": 1.7936395717326704e-05,
"loss": 0.8221,
"step": 835
},
{
"epoch": 2.49,
"grad_norm": 1.015095925986807,
"learning_rate": 1.696120172352025e-05,
"loss": 0.8416,
"step": 840
},
{
"epoch": 2.5,
"grad_norm": 1.2346772662125298,
"learning_rate": 1.601080376443763e-05,
"loss": 0.8742,
"step": 845
},
{
"epoch": 2.52,
"grad_norm": 0.9234353972167334,
"learning_rate": 1.5085485636343755e-05,
"loss": 0.9049,
"step": 850
},
{
"epoch": 2.53,
"grad_norm": 1.0570174679975244,
"learning_rate": 1.4185523646469822e-05,
"loss": 0.8367,
"step": 855
},
{
"epoch": 2.55,
"grad_norm": 0.8795669021553979,
"learning_rate": 1.3311186530505838e-05,
"loss": 0.8661,
"step": 860
},
{
"epoch": 2.56,
"grad_norm": 0.9638299921267924,
"learning_rate": 1.2462735372353996e-05,
"loss": 0.8363,
"step": 865
},
{
"epoch": 2.58,
"grad_norm": 1.2030489918077059,
"learning_rate": 1.1640423526166988e-05,
"loss": 0.8583,
"step": 870
},
{
"epoch": 2.59,
"grad_norm": 1.0716487215374508,
"learning_rate": 1.0844496540694515e-05,
"loss": 0.8588,
"step": 875
},
{
"epoch": 2.61,
"grad_norm": 1.0669537623700072,
"learning_rate": 1.007519208596045e-05,
"loss": 0.9143,
"step": 880
},
{
"epoch": 2.62,
"grad_norm": 0.9415770157718919,
"learning_rate": 9.332739882292752e-06,
"loss": 0.906,
"step": 885
},
{
"epoch": 2.64,
"grad_norm": 0.8466881399276756,
"learning_rate": 8.617361631727138e-06,
"loss": 0.8823,
"step": 890
},
{
"epoch": 2.65,
"grad_norm": 1.173003775227567,
"learning_rate": 7.929270951805178e-06,
"loss": 0.8866,
"step": 895
},
{
"epoch": 2.67,
"grad_norm": 0.9695275155903833,
"learning_rate": 7.2686733117863784e-06,
"loss": 0.8717,
"step": 900
},
{
"epoch": 2.68,
"grad_norm": 0.9278647388922164,
"learning_rate": 6.635765971293484e-06,
"loss": 0.8999,
"step": 905
},
{
"epoch": 2.7,
"grad_norm": 0.9905531986540063,
"learning_rate": 6.030737921409169e-06,
"loss": 0.8772,
"step": 910
},
{
"epoch": 2.71,
"grad_norm": 1.068092774140916,
"learning_rate": 5.453769828241872e-06,
"loss": 0.8974,
"step": 915
},
{
"epoch": 2.73,
"grad_norm": 1.1961165825237712,
"learning_rate": 4.905033978977491e-06,
"loss": 0.86,
"step": 920
},
{
"epoch": 2.74,
"grad_norm": 1.0848450931313236,
"learning_rate": 4.384694230432984e-06,
"loss": 0.8363,
"step": 925
},
{
"epoch": 2.76,
"grad_norm": 1.0929622662706844,
"learning_rate": 3.892905960127546e-06,
"loss": 0.8727,
"step": 930
},
{
"epoch": 2.77,
"grad_norm": 1.0564158056197959,
"learning_rate": 3.4298160198856568e-06,
"loss": 0.8169,
"step": 935
},
{
"epoch": 2.79,
"grad_norm": 1.0136059442310454,
"learning_rate": 2.995562691985898e-06,
"loss": 0.8574,
"step": 940
},
{
"epoch": 2.8,
"grad_norm": 1.092221087850955,
"learning_rate": 2.590275647868867e-06,
"loss": 0.8386,
"step": 945
},
{
"epoch": 2.81,
"grad_norm": 1.0072864796741217,
"learning_rate": 2.2140759094162467e-06,
"loss": 0.8685,
"step": 950
},
{
"epoch": 2.83,
"grad_norm": 0.8433396823974946,
"learning_rate": 1.8670758128126909e-06,
"loss": 0.849,
"step": 955
},
{
"epoch": 2.84,
"grad_norm": 0.9684221245764032,
"learning_rate": 1.5493789750014031e-06,
"loss": 0.8525,
"step": 960
},
{
"epoch": 2.86,
"grad_norm": 0.9141081630302206,
"learning_rate": 1.261080262743297e-06,
"loss": 0.8315,
"step": 965
},
{
"epoch": 2.87,
"grad_norm": 0.9014966731205527,
"learning_rate": 1.0022657642890231e-06,
"loss": 0.8316,
"step": 970
},
{
"epoch": 2.89,
"grad_norm": 0.9464472386186885,
"learning_rate": 7.730127636723539e-07,
"loss": 0.8605,
"step": 975
},
{
"epoch": 2.9,
"grad_norm": 0.896721952301425,
"learning_rate": 5.733897176325665e-07,
"loss": 0.8137,
"step": 980
},
{
"epoch": 2.92,
"grad_norm": 1.0368172773007387,
"learning_rate": 4.034562351727389e-07,
"loss": 0.8664,
"step": 985
},
{
"epoch": 2.93,
"grad_norm": 0.9777429571054969,
"learning_rate": 2.6326305976001055e-07,
"loss": 0.8557,
"step": 990
},
{
"epoch": 2.95,
"grad_norm": 0.96487435678213,
"learning_rate": 1.5285205417319149e-07,
"loss": 0.8691,
"step": 995
},
{
"epoch": 2.96,
"grad_norm": 0.8599085672586297,
"learning_rate": 7.225618800222877e-08,
"loss": 0.8806,
"step": 1000
},
{
"epoch": 2.98,
"grad_norm": 1.0670556057589013,
"learning_rate": 2.1499527803214846e-08,
"loss": 0.8669,
"step": 1005
},
{
"epoch": 2.99,
"grad_norm": 1.0796304331427407,
"learning_rate": 5.972299119250125e-10,
"loss": 0.8486,
"step": 1010
},
{
"epoch": 3.0,
"eval_loss": 1.1171541213989258,
"eval_runtime": 156.2002,
"eval_samples_per_second": 14.789,
"eval_steps_per_second": 0.467,
"step": 1011
},
{
"epoch": 3.0,
"step": 1011,
"total_flos": 2597843400392704.0,
"train_loss": 1.2570129054000658,
"train_runtime": 12020.0319,
"train_samples_per_second": 5.39,
"train_steps_per_second": 0.084
}
],
"logging_steps": 5,
"max_steps": 1011,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"total_flos": 2597843400392704.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}