10250 lines
272 KiB
JSON
10250 lines
272 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 6.18641975308642,
|
|
"eval_steps": 500,
|
|
"global_step": 5670,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.006172839506172839,
|
|
"grad_norm": 10.891036033630371,
|
|
"learning_rate": 2.821869488536155e-07,
|
|
"loss": 0.6832,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.646623432636261,
|
|
"step": 5
|
|
},
|
|
{
|
|
"epoch": 0.012345679012345678,
|
|
"grad_norm": 11.948987007141113,
|
|
"learning_rate": 6.34920634920635e-07,
|
|
"loss": 0.6726,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.608684241771698,
|
|
"step": 10
|
|
},
|
|
{
|
|
"epoch": 0.018518518518518517,
|
|
"grad_norm": 11.665045738220215,
|
|
"learning_rate": 9.876543209876544e-07,
|
|
"loss": 0.6602,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.6334839463233948,
|
|
"step": 15
|
|
},
|
|
{
|
|
"epoch": 0.024691358024691357,
|
|
"grad_norm": 8.376968383789062,
|
|
"learning_rate": 1.3403880070546738e-06,
|
|
"loss": 0.6314,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.6464118361473083,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 0.030864197530864196,
|
|
"grad_norm": 6.368124008178711,
|
|
"learning_rate": 1.6931216931216932e-06,
|
|
"loss": 0.6162,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.6369835734367371,
|
|
"step": 25
|
|
},
|
|
{
|
|
"epoch": 0.037037037037037035,
|
|
"grad_norm": 4.317187309265137,
|
|
"learning_rate": 2.0458553791887124e-06,
|
|
"loss": 0.5712,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.5905289649963379,
|
|
"step": 30
|
|
},
|
|
{
|
|
"epoch": 0.043209876543209874,
|
|
"grad_norm": 3.4061028957366943,
|
|
"learning_rate": 2.3985890652557318e-06,
|
|
"loss": 0.569,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.6018121838569641,
|
|
"step": 35
|
|
},
|
|
{
|
|
"epoch": 0.04938271604938271,
|
|
"grad_norm": 2.1150121688842773,
|
|
"learning_rate": 2.7513227513227516e-06,
|
|
"loss": 0.5323,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.4789934456348419,
|
|
"step": 40
|
|
},
|
|
{
|
|
"epoch": 0.05555555555555555,
|
|
"grad_norm": 1.7288572788238525,
|
|
"learning_rate": 3.104056437389771e-06,
|
|
"loss": 0.5061,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.5093281865119934,
|
|
"step": 45
|
|
},
|
|
{
|
|
"epoch": 0.06172839506172839,
|
|
"grad_norm": 1.1428061723709106,
|
|
"learning_rate": 3.4567901234567904e-06,
|
|
"loss": 0.457,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.5047563910484314,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.06790123456790123,
|
|
"grad_norm": 1.158074975013733,
|
|
"learning_rate": 3.80952380952381e-06,
|
|
"loss": 0.5038,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.5073680281639099,
|
|
"step": 55
|
|
},
|
|
{
|
|
"epoch": 0.07407407407407407,
|
|
"grad_norm": 0.9453456997871399,
|
|
"learning_rate": 4.162257495590829e-06,
|
|
"loss": 0.4499,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.46853289008140564,
|
|
"step": 60
|
|
},
|
|
{
|
|
"epoch": 0.08024691358024691,
|
|
"grad_norm": 0.8852455019950867,
|
|
"learning_rate": 4.514991181657848e-06,
|
|
"loss": 0.4742,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.5242413878440857,
|
|
"step": 65
|
|
},
|
|
{
|
|
"epoch": 0.08641975308641975,
|
|
"grad_norm": 0.6865708231925964,
|
|
"learning_rate": 4.867724867724868e-06,
|
|
"loss": 0.4648,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.5000542998313904,
|
|
"step": 70
|
|
},
|
|
{
|
|
"epoch": 0.09259259259259259,
|
|
"grad_norm": 0.6170792579650879,
|
|
"learning_rate": 5.220458553791887e-06,
|
|
"loss": 0.4393,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.4101174771785736,
|
|
"step": 75
|
|
},
|
|
{
|
|
"epoch": 0.09876543209876543,
|
|
"grad_norm": 0.6837581396102905,
|
|
"learning_rate": 5.573192239858906e-06,
|
|
"loss": 0.4589,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.406427264213562,
|
|
"step": 80
|
|
},
|
|
{
|
|
"epoch": 0.10493827160493827,
|
|
"grad_norm": 0.5824769735336304,
|
|
"learning_rate": 5.925925925925926e-06,
|
|
"loss": 0.4011,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.4628138244152069,
|
|
"step": 85
|
|
},
|
|
{
|
|
"epoch": 0.1111111111111111,
|
|
"grad_norm": 0.5061732530593872,
|
|
"learning_rate": 6.278659611992945e-06,
|
|
"loss": 0.3945,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3808530271053314,
|
|
"step": 90
|
|
},
|
|
{
|
|
"epoch": 0.11728395061728394,
|
|
"grad_norm": 0.5147681832313538,
|
|
"learning_rate": 6.631393298059965e-06,
|
|
"loss": 0.3921,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.38033249974250793,
|
|
"step": 95
|
|
},
|
|
{
|
|
"epoch": 0.12345679012345678,
|
|
"grad_norm": 0.4470911920070648,
|
|
"learning_rate": 6.984126984126984e-06,
|
|
"loss": 0.3677,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.37873291969299316,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.12962962962962962,
|
|
"grad_norm": 0.5028461813926697,
|
|
"learning_rate": 7.3368606701940034e-06,
|
|
"loss": 0.3694,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.34291210770606995,
|
|
"step": 105
|
|
},
|
|
{
|
|
"epoch": 0.13580246913580246,
|
|
"grad_norm": 0.5090969800949097,
|
|
"learning_rate": 7.689594356261023e-06,
|
|
"loss": 0.3763,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.33877086639404297,
|
|
"step": 110
|
|
},
|
|
{
|
|
"epoch": 0.1419753086419753,
|
|
"grad_norm": 0.4665391743183136,
|
|
"learning_rate": 8.042328042328043e-06,
|
|
"loss": 0.3798,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3622983992099762,
|
|
"step": 115
|
|
},
|
|
{
|
|
"epoch": 0.14814814814814814,
|
|
"grad_norm": 0.44902950525283813,
|
|
"learning_rate": 8.395061728395062e-06,
|
|
"loss": 0.3516,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3451531231403351,
|
|
"step": 120
|
|
},
|
|
{
|
|
"epoch": 0.15432098765432098,
|
|
"grad_norm": 0.44352996349334717,
|
|
"learning_rate": 8.747795414462082e-06,
|
|
"loss": 0.3396,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.35839423537254333,
|
|
"step": 125
|
|
},
|
|
{
|
|
"epoch": 0.16049382716049382,
|
|
"grad_norm": 0.4222885072231293,
|
|
"learning_rate": 9.1005291005291e-06,
|
|
"loss": 0.365,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.33761486411094666,
|
|
"step": 130
|
|
},
|
|
{
|
|
"epoch": 0.16666666666666666,
|
|
"grad_norm": 0.5398893356323242,
|
|
"learning_rate": 9.45326278659612e-06,
|
|
"loss": 0.3447,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3309403359889984,
|
|
"step": 135
|
|
},
|
|
{
|
|
"epoch": 0.1728395061728395,
|
|
"grad_norm": 0.39981362223625183,
|
|
"learning_rate": 9.80599647266314e-06,
|
|
"loss": 0.3434,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3704066276550293,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 0.17901234567901234,
|
|
"grad_norm": 0.4901084005832672,
|
|
"learning_rate": 1.015873015873016e-05,
|
|
"loss": 0.3306,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.35292086005210876,
|
|
"step": 145
|
|
},
|
|
{
|
|
"epoch": 0.18518518518518517,
|
|
"grad_norm": 0.41522544622421265,
|
|
"learning_rate": 1.0511463844797178e-05,
|
|
"loss": 0.3112,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.34682130813598633,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 0.19135802469135801,
|
|
"grad_norm": 0.4412677586078644,
|
|
"learning_rate": 1.0864197530864198e-05,
|
|
"loss": 0.3272,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3141136169433594,
|
|
"step": 155
|
|
},
|
|
{
|
|
"epoch": 0.19753086419753085,
|
|
"grad_norm": 0.389218807220459,
|
|
"learning_rate": 1.1216931216931217e-05,
|
|
"loss": 0.3249,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.268544465303421,
|
|
"step": 160
|
|
},
|
|
{
|
|
"epoch": 0.2037037037037037,
|
|
"grad_norm": 0.48689374327659607,
|
|
"learning_rate": 1.1569664902998237e-05,
|
|
"loss": 0.2951,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2812751233577728,
|
|
"step": 165
|
|
},
|
|
{
|
|
"epoch": 0.20987654320987653,
|
|
"grad_norm": 0.3516155183315277,
|
|
"learning_rate": 1.1922398589065256e-05,
|
|
"loss": 0.2944,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2984471619129181,
|
|
"step": 170
|
|
},
|
|
{
|
|
"epoch": 0.21604938271604937,
|
|
"grad_norm": 0.3933318555355072,
|
|
"learning_rate": 1.2275132275132276e-05,
|
|
"loss": 0.312,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3248521387577057,
|
|
"step": 175
|
|
},
|
|
{
|
|
"epoch": 0.2222222222222222,
|
|
"grad_norm": 0.482473760843277,
|
|
"learning_rate": 1.2627865961199294e-05,
|
|
"loss": 0.2954,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.332978218793869,
|
|
"step": 180
|
|
},
|
|
{
|
|
"epoch": 0.22839506172839505,
|
|
"grad_norm": 0.5011610388755798,
|
|
"learning_rate": 1.2980599647266315e-05,
|
|
"loss": 0.3023,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3212778568267822,
|
|
"step": 185
|
|
},
|
|
{
|
|
"epoch": 0.2345679012345679,
|
|
"grad_norm": 0.46970030665397644,
|
|
"learning_rate": 1.3333333333333333e-05,
|
|
"loss": 0.3145,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3365441858768463,
|
|
"step": 190
|
|
},
|
|
{
|
|
"epoch": 0.24074074074074073,
|
|
"grad_norm": 0.4947907030582428,
|
|
"learning_rate": 1.3686067019400354e-05,
|
|
"loss": 0.3003,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3034442961215973,
|
|
"step": 195
|
|
},
|
|
{
|
|
"epoch": 0.24691358024691357,
|
|
"grad_norm": 0.4818788766860962,
|
|
"learning_rate": 1.4038800705467372e-05,
|
|
"loss": 0.2992,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3292461931705475,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 0.25308641975308643,
|
|
"grad_norm": 0.7307603359222412,
|
|
"learning_rate": 1.4391534391534392e-05,
|
|
"loss": 0.3461,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.32606855034828186,
|
|
"step": 205
|
|
},
|
|
{
|
|
"epoch": 0.25925925925925924,
|
|
"grad_norm": 0.6940523982048035,
|
|
"learning_rate": 1.4744268077601411e-05,
|
|
"loss": 0.344,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.33493685722351074,
|
|
"step": 210
|
|
},
|
|
{
|
|
"epoch": 0.2654320987654321,
|
|
"grad_norm": 0.5480557084083557,
|
|
"learning_rate": 1.5097001763668431e-05,
|
|
"loss": 0.3202,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3456254005432129,
|
|
"step": 215
|
|
},
|
|
{
|
|
"epoch": 0.2716049382716049,
|
|
"grad_norm": 0.5187277793884277,
|
|
"learning_rate": 1.544973544973545e-05,
|
|
"loss": 0.3337,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.34756898880004883,
|
|
"step": 220
|
|
},
|
|
{
|
|
"epoch": 0.2777777777777778,
|
|
"grad_norm": 0.5956120491027832,
|
|
"learning_rate": 1.580246913580247e-05,
|
|
"loss": 0.3145,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2839909493923187,
|
|
"step": 225
|
|
},
|
|
{
|
|
"epoch": 0.2839506172839506,
|
|
"grad_norm": 0.48677903413772583,
|
|
"learning_rate": 1.615520282186949e-05,
|
|
"loss": 0.299,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.29133015871047974,
|
|
"step": 230
|
|
},
|
|
{
|
|
"epoch": 0.29012345679012347,
|
|
"grad_norm": 0.6096779108047485,
|
|
"learning_rate": 1.6507936507936507e-05,
|
|
"loss": 0.3176,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3263722062110901,
|
|
"step": 235
|
|
},
|
|
{
|
|
"epoch": 0.2962962962962963,
|
|
"grad_norm": 0.5650277137756348,
|
|
"learning_rate": 1.686067019400353e-05,
|
|
"loss": 0.3071,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.30650031566619873,
|
|
"step": 240
|
|
},
|
|
{
|
|
"epoch": 0.30246913580246915,
|
|
"grad_norm": 0.5755073428153992,
|
|
"learning_rate": 1.7213403880070548e-05,
|
|
"loss": 0.3013,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2889738082885742,
|
|
"step": 245
|
|
},
|
|
{
|
|
"epoch": 0.30864197530864196,
|
|
"grad_norm": 0.47710350155830383,
|
|
"learning_rate": 1.7566137566137566e-05,
|
|
"loss": 0.2636,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2973271310329437,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 0.3148148148148148,
|
|
"grad_norm": 0.5268696546554565,
|
|
"learning_rate": 1.7918871252204585e-05,
|
|
"loss": 0.2869,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2767695486545563,
|
|
"step": 255
|
|
},
|
|
{
|
|
"epoch": 0.32098765432098764,
|
|
"grad_norm": 7.942512512207031,
|
|
"learning_rate": 1.8271604938271607e-05,
|
|
"loss": 0.3257,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.4871584475040436,
|
|
"step": 260
|
|
},
|
|
{
|
|
"epoch": 0.3271604938271605,
|
|
"grad_norm": 0.8912859559059143,
|
|
"learning_rate": 1.8624338624338625e-05,
|
|
"loss": 0.2701,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23925618827342987,
|
|
"step": 265
|
|
},
|
|
{
|
|
"epoch": 0.3333333333333333,
|
|
"grad_norm": 0.5555944442749023,
|
|
"learning_rate": 1.8977072310405644e-05,
|
|
"loss": 0.2738,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27243828773498535,
|
|
"step": 270
|
|
},
|
|
{
|
|
"epoch": 0.3395061728395062,
|
|
"grad_norm": 0.461831271648407,
|
|
"learning_rate": 1.9329805996472662e-05,
|
|
"loss": 0.2906,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3025025427341461,
|
|
"step": 275
|
|
},
|
|
{
|
|
"epoch": 0.345679012345679,
|
|
"grad_norm": 0.5229907631874084,
|
|
"learning_rate": 1.9682539682539684e-05,
|
|
"loss": 0.3063,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27696844935417175,
|
|
"step": 280
|
|
},
|
|
{
|
|
"epoch": 0.35185185185185186,
|
|
"grad_norm": 0.5061423778533936,
|
|
"learning_rate": 2.0035273368606703e-05,
|
|
"loss": 0.2721,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.274807333946228,
|
|
"step": 285
|
|
},
|
|
{
|
|
"epoch": 0.35802469135802467,
|
|
"grad_norm": 0.5906267762184143,
|
|
"learning_rate": 2.0388007054673725e-05,
|
|
"loss": 0.2841,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.29160311818122864,
|
|
"step": 290
|
|
},
|
|
{
|
|
"epoch": 0.36419753086419754,
|
|
"grad_norm": 0.49827638268470764,
|
|
"learning_rate": 2.074074074074074e-05,
|
|
"loss": 0.2698,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2548845708370209,
|
|
"step": 295
|
|
},
|
|
{
|
|
"epoch": 0.37037037037037035,
|
|
"grad_norm": 0.4860912561416626,
|
|
"learning_rate": 2.1093474426807762e-05,
|
|
"loss": 0.292,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.29940351843833923,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 0.006172839506172839,
|
|
"grad_norm": 0.4834241271018982,
|
|
"learning_rate": 2.144620811287478e-05,
|
|
"loss": 0.3019,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.28686702251434326,
|
|
"step": 305
|
|
},
|
|
{
|
|
"epoch": 0.012345679012345678,
|
|
"grad_norm": 0.4784407913684845,
|
|
"learning_rate": 2.1798941798941802e-05,
|
|
"loss": 0.2853,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25928911566734314,
|
|
"step": 310
|
|
},
|
|
{
|
|
"epoch": 0.018518518518518517,
|
|
"grad_norm": 0.49496716260910034,
|
|
"learning_rate": 2.2151675485008818e-05,
|
|
"loss": 0.2775,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2703591287136078,
|
|
"step": 315
|
|
},
|
|
{
|
|
"epoch": 0.024691358024691357,
|
|
"grad_norm": 0.40451133251190186,
|
|
"learning_rate": 2.250440917107584e-05,
|
|
"loss": 0.2723,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.28752318024635315,
|
|
"step": 320
|
|
},
|
|
{
|
|
"epoch": 0.030864197530864196,
|
|
"grad_norm": 0.3546577990055084,
|
|
"learning_rate": 2.2857142857142858e-05,
|
|
"loss": 0.2755,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.30474719405174255,
|
|
"step": 325
|
|
},
|
|
{
|
|
"epoch": 0.037037037037037035,
|
|
"grad_norm": 0.3632960915565491,
|
|
"learning_rate": 2.320987654320988e-05,
|
|
"loss": 0.2656,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.28616902232170105,
|
|
"step": 330
|
|
},
|
|
{
|
|
"epoch": 0.043209876543209874,
|
|
"grad_norm": 0.42180970311164856,
|
|
"learning_rate": 2.3562610229276895e-05,
|
|
"loss": 0.2772,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3029364049434662,
|
|
"step": 335
|
|
},
|
|
{
|
|
"epoch": 0.04938271604938271,
|
|
"grad_norm": 0.36831268668174744,
|
|
"learning_rate": 2.3915343915343917e-05,
|
|
"loss": 0.27,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2365947812795639,
|
|
"step": 340
|
|
},
|
|
{
|
|
"epoch": 0.05555555555555555,
|
|
"grad_norm": 0.47895246744155884,
|
|
"learning_rate": 2.4268077601410936e-05,
|
|
"loss": 0.2645,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26503515243530273,
|
|
"step": 345
|
|
},
|
|
{
|
|
"epoch": 0.06172839506172839,
|
|
"grad_norm": 0.3769896924495697,
|
|
"learning_rate": 2.4620811287477958e-05,
|
|
"loss": 0.2502,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2810387909412384,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 0.06790123456790123,
|
|
"grad_norm": 0.46085411310195923,
|
|
"learning_rate": 2.4973544973544973e-05,
|
|
"loss": 0.2818,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2839004695415497,
|
|
"step": 355
|
|
},
|
|
{
|
|
"epoch": 0.07407407407407407,
|
|
"grad_norm": 0.4568277895450592,
|
|
"learning_rate": 2.5326278659611995e-05,
|
|
"loss": 0.2554,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2613278925418854,
|
|
"step": 360
|
|
},
|
|
{
|
|
"epoch": 0.08024691358024691,
|
|
"grad_norm": 0.5308265089988708,
|
|
"learning_rate": 2.5679012345679013e-05,
|
|
"loss": 0.2748,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3018066883087158,
|
|
"step": 365
|
|
},
|
|
{
|
|
"epoch": 0.08641975308641975,
|
|
"grad_norm": 0.4405893087387085,
|
|
"learning_rate": 2.6031746031746035e-05,
|
|
"loss": 0.2787,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3065824806690216,
|
|
"step": 370
|
|
},
|
|
{
|
|
"epoch": 0.09259259259259259,
|
|
"grad_norm": 0.4083263576030731,
|
|
"learning_rate": 2.638447971781305e-05,
|
|
"loss": 0.2661,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2464752197265625,
|
|
"step": 375
|
|
},
|
|
{
|
|
"epoch": 0.09876543209876543,
|
|
"grad_norm": 0.5209446549415588,
|
|
"learning_rate": 2.6737213403880072e-05,
|
|
"loss": 0.2866,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2574746906757355,
|
|
"step": 380
|
|
},
|
|
{
|
|
"epoch": 0.10493827160493827,
|
|
"grad_norm": 0.48428890109062195,
|
|
"learning_rate": 2.708994708994709e-05,
|
|
"loss": 0.2564,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2907469868659973,
|
|
"step": 385
|
|
},
|
|
{
|
|
"epoch": 0.1111111111111111,
|
|
"grad_norm": 0.4454934298992157,
|
|
"learning_rate": 2.7442680776014113e-05,
|
|
"loss": 0.257,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24461741745471954,
|
|
"step": 390
|
|
},
|
|
{
|
|
"epoch": 0.11728395061728394,
|
|
"grad_norm": 0.4358861744403839,
|
|
"learning_rate": 2.7795414462081128e-05,
|
|
"loss": 0.2622,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2549789845943451,
|
|
"step": 395
|
|
},
|
|
{
|
|
"epoch": 0.12345679012345678,
|
|
"grad_norm": 0.4008908271789551,
|
|
"learning_rate": 2.814814814814815e-05,
|
|
"loss": 0.2495,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25932642817497253,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 0.12962962962962962,
|
|
"grad_norm": 0.47018739581108093,
|
|
"learning_rate": 2.850088183421517e-05,
|
|
"loss": 0.2555,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2385169267654419,
|
|
"step": 405
|
|
},
|
|
{
|
|
"epoch": 0.13580246913580246,
|
|
"grad_norm": 0.5076704025268555,
|
|
"learning_rate": 2.885361552028219e-05,
|
|
"loss": 0.2652,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23661553859710693,
|
|
"step": 410
|
|
},
|
|
{
|
|
"epoch": 0.1419753086419753,
|
|
"grad_norm": 0.4354270398616791,
|
|
"learning_rate": 2.9206349206349206e-05,
|
|
"loss": 0.2726,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26338306069374084,
|
|
"step": 415
|
|
},
|
|
{
|
|
"epoch": 0.14814814814814814,
|
|
"grad_norm": 0.43332722783088684,
|
|
"learning_rate": 2.9559082892416228e-05,
|
|
"loss": 0.2535,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25269973278045654,
|
|
"step": 420
|
|
},
|
|
{
|
|
"epoch": 0.15432098765432098,
|
|
"grad_norm": 0.4248559772968292,
|
|
"learning_rate": 2.9911816578483246e-05,
|
|
"loss": 0.2484,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2629343271255493,
|
|
"step": 425
|
|
},
|
|
{
|
|
"epoch": 0.16049382716049382,
|
|
"grad_norm": 0.4432432949542999,
|
|
"learning_rate": 3.0264550264550268e-05,
|
|
"loss": 0.2706,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24757696688175201,
|
|
"step": 430
|
|
},
|
|
{
|
|
"epoch": 0.16666666666666666,
|
|
"grad_norm": 0.48535019159317017,
|
|
"learning_rate": 3.061728395061729e-05,
|
|
"loss": 0.2585,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24424861371517181,
|
|
"step": 435
|
|
},
|
|
{
|
|
"epoch": 0.1728395061728395,
|
|
"grad_norm": 0.4025910794734955,
|
|
"learning_rate": 3.0970017636684305e-05,
|
|
"loss": 0.2611,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.28406617045402527,
|
|
"step": 440
|
|
},
|
|
{
|
|
"epoch": 0.17901234567901234,
|
|
"grad_norm": 0.5373814105987549,
|
|
"learning_rate": 3.1322751322751324e-05,
|
|
"loss": 0.253,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27086856961250305,
|
|
"step": 445
|
|
},
|
|
{
|
|
"epoch": 0.18518518518518517,
|
|
"grad_norm": 0.4191938638687134,
|
|
"learning_rate": 3.167548500881834e-05,
|
|
"loss": 0.2411,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26918238401412964,
|
|
"step": 450
|
|
},
|
|
{
|
|
"epoch": 0.19135802469135801,
|
|
"grad_norm": 0.42429038882255554,
|
|
"learning_rate": 3.202821869488536e-05,
|
|
"loss": 0.2562,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2502617835998535,
|
|
"step": 455
|
|
},
|
|
{
|
|
"epoch": 0.19753086419753085,
|
|
"grad_norm": 0.35458773374557495,
|
|
"learning_rate": 3.2380952380952386e-05,
|
|
"loss": 0.2572,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21254992485046387,
|
|
"step": 460
|
|
},
|
|
{
|
|
"epoch": 0.2037037037037037,
|
|
"grad_norm": 0.4847748279571533,
|
|
"learning_rate": 3.27336860670194e-05,
|
|
"loss": 0.234,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2244824916124344,
|
|
"step": 465
|
|
},
|
|
{
|
|
"epoch": 0.20987654320987653,
|
|
"grad_norm": 0.34742796421051025,
|
|
"learning_rate": 3.308641975308642e-05,
|
|
"loss": 0.2359,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2404363751411438,
|
|
"step": 470
|
|
},
|
|
{
|
|
"epoch": 0.21604938271604937,
|
|
"grad_norm": 0.36754411458969116,
|
|
"learning_rate": 3.343915343915344e-05,
|
|
"loss": 0.251,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2642151117324829,
|
|
"step": 475
|
|
},
|
|
{
|
|
"epoch": 0.2222222222222222,
|
|
"grad_norm": 0.4316296875476837,
|
|
"learning_rate": 3.379188712522046e-05,
|
|
"loss": 0.2398,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27224355936050415,
|
|
"step": 480
|
|
},
|
|
{
|
|
"epoch": 0.22839506172839505,
|
|
"grad_norm": 0.45226213335990906,
|
|
"learning_rate": 3.414462081128748e-05,
|
|
"loss": 0.2471,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26445767283439636,
|
|
"step": 485
|
|
},
|
|
{
|
|
"epoch": 0.2345679012345679,
|
|
"grad_norm": 0.44799497723579407,
|
|
"learning_rate": 3.44973544973545e-05,
|
|
"loss": 0.2577,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27635422348976135,
|
|
"step": 490
|
|
},
|
|
{
|
|
"epoch": 0.24074074074074073,
|
|
"grad_norm": 0.48871344327926636,
|
|
"learning_rate": 3.4850088183421516e-05,
|
|
"loss": 0.2482,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24841894209384918,
|
|
"step": 495
|
|
},
|
|
{
|
|
"epoch": 0.24691358024691357,
|
|
"grad_norm": 0.4787638783454895,
|
|
"learning_rate": 3.520282186948854e-05,
|
|
"loss": 0.2484,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2760474979877472,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.25308641975308643,
|
|
"grad_norm": 0.4769534170627594,
|
|
"learning_rate": 3.555555555555555e-05,
|
|
"loss": 0.2721,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24780535697937012,
|
|
"step": 505
|
|
},
|
|
{
|
|
"epoch": 0.25925925925925924,
|
|
"grad_norm": 0.5165116786956787,
|
|
"learning_rate": 3.590828924162258e-05,
|
|
"loss": 0.2651,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2576892375946045,
|
|
"step": 510
|
|
},
|
|
{
|
|
"epoch": 0.2654320987654321,
|
|
"grad_norm": 0.4602437913417816,
|
|
"learning_rate": 3.62610229276896e-05,
|
|
"loss": 0.2538,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27468374371528625,
|
|
"step": 515
|
|
},
|
|
{
|
|
"epoch": 0.2716049382716049,
|
|
"grad_norm": 0.4690820574760437,
|
|
"learning_rate": 3.6613756613756616e-05,
|
|
"loss": 0.2662,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.28319332003593445,
|
|
"step": 520
|
|
},
|
|
{
|
|
"epoch": 0.2777777777777778,
|
|
"grad_norm": 0.5172345638275146,
|
|
"learning_rate": 3.6966490299823634e-05,
|
|
"loss": 0.2586,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.231450155377388,
|
|
"step": 525
|
|
},
|
|
{
|
|
"epoch": 0.2839506172839506,
|
|
"grad_norm": 0.43790754675865173,
|
|
"learning_rate": 3.731922398589065e-05,
|
|
"loss": 0.2497,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24846939742565155,
|
|
"step": 530
|
|
},
|
|
{
|
|
"epoch": 0.29012345679012347,
|
|
"grad_norm": 0.5391412377357483,
|
|
"learning_rate": 3.767195767195767e-05,
|
|
"loss": 0.2682,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27478349208831787,
|
|
"step": 535
|
|
},
|
|
{
|
|
"epoch": 0.2962962962962963,
|
|
"grad_norm": 0.5666778087615967,
|
|
"learning_rate": 3.80246913580247e-05,
|
|
"loss": 0.2604,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26373395323753357,
|
|
"step": 540
|
|
},
|
|
{
|
|
"epoch": 0.30246913580246915,
|
|
"grad_norm": 0.6512420177459717,
|
|
"learning_rate": 3.837742504409171e-05,
|
|
"loss": 0.2593,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24866722524166107,
|
|
"step": 545
|
|
},
|
|
{
|
|
"epoch": 0.30864197530864196,
|
|
"grad_norm": 0.5449404120445251,
|
|
"learning_rate": 3.8730158730158734e-05,
|
|
"loss": 0.2276,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25449231266975403,
|
|
"step": 550
|
|
},
|
|
{
|
|
"epoch": 0.3148148148148148,
|
|
"grad_norm": 0.5298479795455933,
|
|
"learning_rate": 3.908289241622575e-05,
|
|
"loss": 0.2475,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24061168730258942,
|
|
"step": 555
|
|
},
|
|
{
|
|
"epoch": 0.32098765432098764,
|
|
"grad_norm": 8.54423999786377,
|
|
"learning_rate": 3.943562610229277e-05,
|
|
"loss": 0.2767,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3316479027271271,
|
|
"step": 560
|
|
},
|
|
{
|
|
"epoch": 0.3271604938271605,
|
|
"grad_norm": 0.6743181943893433,
|
|
"learning_rate": 3.978835978835979e-05,
|
|
"loss": 0.237,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20530176162719727,
|
|
"step": 565
|
|
},
|
|
{
|
|
"epoch": 0.3333333333333333,
|
|
"grad_norm": 0.45386803150177,
|
|
"learning_rate": 3.999998483967383e-05,
|
|
"loss": 0.2399,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2415226250886917,
|
|
"step": 570
|
|
},
|
|
{
|
|
"epoch": 0.3395061728395062,
|
|
"grad_norm": 0.4268856644630432,
|
|
"learning_rate": 3.9999814286268345e-05,
|
|
"loss": 0.2558,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27003559470176697,
|
|
"step": 575
|
|
},
|
|
{
|
|
"epoch": 0.345679012345679,
|
|
"grad_norm": 0.47914114594459534,
|
|
"learning_rate": 3.999945423067106e-05,
|
|
"loss": 0.2744,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2455822080373764,
|
|
"step": 580
|
|
},
|
|
{
|
|
"epoch": 0.35185185185185186,
|
|
"grad_norm": 0.4515032172203064,
|
|
"learning_rate": 3.9998904676293594e-05,
|
|
"loss": 0.2425,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24407601356506348,
|
|
"step": 585
|
|
},
|
|
{
|
|
"epoch": 0.35802469135802467,
|
|
"grad_norm": 0.5433447360992432,
|
|
"learning_rate": 3.999816562834308e-05,
|
|
"loss": 0.2517,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25850483775138855,
|
|
"step": 590
|
|
},
|
|
{
|
|
"epoch": 0.36419753086419754,
|
|
"grad_norm": 0.42169854044914246,
|
|
"learning_rate": 3.999723709382212e-05,
|
|
"loss": 0.2385,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22521519660949707,
|
|
"step": 595
|
|
},
|
|
{
|
|
"epoch": 0.37037037037037035,
|
|
"grad_norm": 0.4280223250389099,
|
|
"learning_rate": 3.9996119081528785e-05,
|
|
"loss": 0.2619,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26947125792503357,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 0.3765432098765432,
|
|
"grad_norm": 0.40860530734062195,
|
|
"learning_rate": 3.999481160205645e-05,
|
|
"loss": 0.2654,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23929916322231293,
|
|
"step": 605
|
|
},
|
|
{
|
|
"epoch": 0.38271604938271603,
|
|
"grad_norm": 0.4502677321434021,
|
|
"learning_rate": 3.999331466779375e-05,
|
|
"loss": 0.2764,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2855079770088196,
|
|
"step": 610
|
|
},
|
|
{
|
|
"epoch": 0.3888888888888889,
|
|
"grad_norm": 0.4336073398590088,
|
|
"learning_rate": 3.999162829292443e-05,
|
|
"loss": 0.2684,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2661067247390747,
|
|
"step": 615
|
|
},
|
|
{
|
|
"epoch": 0.3950617283950617,
|
|
"grad_norm": 0.4738907516002655,
|
|
"learning_rate": 3.998975249342722e-05,
|
|
"loss": 0.2395,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.212646484375,
|
|
"step": 620
|
|
},
|
|
{
|
|
"epoch": 0.4012345679012346,
|
|
"grad_norm": 0.41804760694503784,
|
|
"learning_rate": 3.998768728707571e-05,
|
|
"loss": 0.2383,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23833966255187988,
|
|
"step": 625
|
|
},
|
|
{
|
|
"epoch": 0.4074074074074074,
|
|
"grad_norm": 0.4101686477661133,
|
|
"learning_rate": 3.998543269343811e-05,
|
|
"loss": 0.2514,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23446016013622284,
|
|
"step": 630
|
|
},
|
|
{
|
|
"epoch": 0.41358024691358025,
|
|
"grad_norm": 0.38764819502830505,
|
|
"learning_rate": 3.9982988733877166e-05,
|
|
"loss": 0.2465,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27846357226371765,
|
|
"step": 635
|
|
},
|
|
{
|
|
"epoch": 0.41975308641975306,
|
|
"grad_norm": 0.4104335904121399,
|
|
"learning_rate": 3.9980355431549854e-05,
|
|
"loss": 0.253,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2743143141269684,
|
|
"step": 640
|
|
},
|
|
{
|
|
"epoch": 0.42592592592592593,
|
|
"grad_norm": 0.45527124404907227,
|
|
"learning_rate": 3.997753281140725e-05,
|
|
"loss": 0.244,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24034304916858673,
|
|
"step": 645
|
|
},
|
|
{
|
|
"epoch": 0.43209876543209874,
|
|
"grad_norm": 0.37041789293289185,
|
|
"learning_rate": 3.9974520900194236e-05,
|
|
"loss": 0.2547,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24775992333889008,
|
|
"step": 650
|
|
},
|
|
{
|
|
"epoch": 0.4382716049382716,
|
|
"grad_norm": 0.41022607684135437,
|
|
"learning_rate": 3.9971319726449256e-05,
|
|
"loss": 0.2414,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24325692653656006,
|
|
"step": 655
|
|
},
|
|
{
|
|
"epoch": 0.4444444444444444,
|
|
"grad_norm": 0.5095195174217224,
|
|
"learning_rate": 3.996792932050408e-05,
|
|
"loss": 0.2541,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25377097725868225,
|
|
"step": 660
|
|
},
|
|
{
|
|
"epoch": 0.4506172839506173,
|
|
"grad_norm": 0.4300116300582886,
|
|
"learning_rate": 3.996434971448346e-05,
|
|
"loss": 0.2328,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24578583240509033,
|
|
"step": 665
|
|
},
|
|
{
|
|
"epoch": 0.4567901234567901,
|
|
"grad_norm": 0.4212883412837982,
|
|
"learning_rate": 3.996058094230487e-05,
|
|
"loss": 0.2436,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2376929521560669,
|
|
"step": 670
|
|
},
|
|
{
|
|
"epoch": 0.46296296296296297,
|
|
"grad_norm": 0.4166100025177002,
|
|
"learning_rate": 3.995662303967818e-05,
|
|
"loss": 0.2401,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2225937396287918,
|
|
"step": 675
|
|
},
|
|
{
|
|
"epoch": 0.4691358024691358,
|
|
"grad_norm": 0.425618052482605,
|
|
"learning_rate": 3.99524760441053e-05,
|
|
"loss": 0.2267,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23938322067260742,
|
|
"step": 680
|
|
},
|
|
{
|
|
"epoch": 0.47530864197530864,
|
|
"grad_norm": 0.46085667610168457,
|
|
"learning_rate": 3.994813999487982e-05,
|
|
"loss": 0.24,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22277164459228516,
|
|
"step": 685
|
|
},
|
|
{
|
|
"epoch": 0.48148148148148145,
|
|
"grad_norm": 0.41466468572616577,
|
|
"learning_rate": 3.994361493308668e-05,
|
|
"loss": 0.2624,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25506719946861267,
|
|
"step": 690
|
|
},
|
|
{
|
|
"epoch": 0.4876543209876543,
|
|
"grad_norm": 0.45974433422088623,
|
|
"learning_rate": 3.9938900901601715e-05,
|
|
"loss": 0.2361,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22540944814682007,
|
|
"step": 695
|
|
},
|
|
{
|
|
"epoch": 0.49382716049382713,
|
|
"grad_norm": 0.6547717452049255,
|
|
"learning_rate": 3.993399794509132e-05,
|
|
"loss": 0.2344,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2241024225950241,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 0.5,
|
|
"grad_norm": 0.4765329957008362,
|
|
"learning_rate": 3.992890611001197e-05,
|
|
"loss": 0.2263,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20139180123806,
|
|
"step": 705
|
|
},
|
|
{
|
|
"epoch": 0.5061728395061729,
|
|
"grad_norm": 0.4123690128326416,
|
|
"learning_rate": 3.992362544460981e-05,
|
|
"loss": 0.2186,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22565175592899323,
|
|
"step": 710
|
|
},
|
|
{
|
|
"epoch": 0.5123456790123457,
|
|
"grad_norm": 0.4858299493789673,
|
|
"learning_rate": 3.991815599892019e-05,
|
|
"loss": 0.2243,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27657151222229004,
|
|
"step": 715
|
|
},
|
|
{
|
|
"epoch": 0.5185185185185185,
|
|
"grad_norm": 0.5484740734100342,
|
|
"learning_rate": 3.991249782476719e-05,
|
|
"loss": 0.24,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25482916831970215,
|
|
"step": 720
|
|
},
|
|
{
|
|
"epoch": 0.5246913580246914,
|
|
"grad_norm": 0.4507368206977844,
|
|
"learning_rate": 3.9906650975763116e-05,
|
|
"loss": 0.2242,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23104362189769745,
|
|
"step": 725
|
|
},
|
|
{
|
|
"epoch": 0.5308641975308642,
|
|
"grad_norm": 0.4715766906738281,
|
|
"learning_rate": 3.990061550730803e-05,
|
|
"loss": 0.241,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2536795437335968,
|
|
"step": 730
|
|
},
|
|
{
|
|
"epoch": 0.5370370370370371,
|
|
"grad_norm": 0.4264748692512512,
|
|
"learning_rate": 3.9894391476589184e-05,
|
|
"loss": 0.2567,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25595971941947937,
|
|
"step": 735
|
|
},
|
|
{
|
|
"epoch": 0.5432098765432098,
|
|
"grad_norm": 0.40620046854019165,
|
|
"learning_rate": 3.98879789425805e-05,
|
|
"loss": 0.2523,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26936662197113037,
|
|
"step": 740
|
|
},
|
|
{
|
|
"epoch": 0.5493827160493827,
|
|
"grad_norm": 0.5521304607391357,
|
|
"learning_rate": 3.988137796604201e-05,
|
|
"loss": 0.247,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22389264404773712,
|
|
"step": 745
|
|
},
|
|
{
|
|
"epoch": 0.5555555555555556,
|
|
"grad_norm": 0.4298575222492218,
|
|
"learning_rate": 3.987458860951925e-05,
|
|
"loss": 0.2365,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25268805027008057,
|
|
"step": 750
|
|
},
|
|
{
|
|
"epoch": 0.5617283950617284,
|
|
"grad_norm": 0.42672857642173767,
|
|
"learning_rate": 3.986761093734274e-05,
|
|
"loss": 0.2641,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23522251844406128,
|
|
"step": 755
|
|
},
|
|
{
|
|
"epoch": 0.5679012345679012,
|
|
"grad_norm": 0.439872145652771,
|
|
"learning_rate": 3.986044501562729e-05,
|
|
"loss": 0.2433,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.208984375,
|
|
"step": 760
|
|
},
|
|
{
|
|
"epoch": 0.5740740740740741,
|
|
"grad_norm": 0.35056906938552856,
|
|
"learning_rate": 3.9853090912271424e-05,
|
|
"loss": 0.2372,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27140045166015625,
|
|
"step": 765
|
|
},
|
|
{
|
|
"epoch": 0.5802469135802469,
|
|
"grad_norm": 0.39765727519989014,
|
|
"learning_rate": 3.984554869695672e-05,
|
|
"loss": 0.251,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26792848110198975,
|
|
"step": 770
|
|
},
|
|
{
|
|
"epoch": 0.5864197530864198,
|
|
"grad_norm": 0.4515121281147003,
|
|
"learning_rate": 3.9837818441147176e-05,
|
|
"loss": 0.2433,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23011119663715363,
|
|
"step": 775
|
|
},
|
|
{
|
|
"epoch": 0.5925925925925926,
|
|
"grad_norm": 0.44911113381385803,
|
|
"learning_rate": 3.982990021808847e-05,
|
|
"loss": 0.2265,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22638726234436035,
|
|
"step": 780
|
|
},
|
|
{
|
|
"epoch": 0.5987654320987654,
|
|
"grad_norm": 0.48707595467567444,
|
|
"learning_rate": 3.9821794102807346e-05,
|
|
"loss": 0.2289,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2399512529373169,
|
|
"step": 785
|
|
},
|
|
{
|
|
"epoch": 0.6049382716049383,
|
|
"grad_norm": 0.40446770191192627,
|
|
"learning_rate": 3.981350017211084e-05,
|
|
"loss": 0.2287,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24059748649597168,
|
|
"step": 790
|
|
},
|
|
{
|
|
"epoch": 0.6111111111111112,
|
|
"grad_norm": 0.40531155467033386,
|
|
"learning_rate": 3.9805018504585596e-05,
|
|
"loss": 0.2401,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2632632553577423,
|
|
"step": 795
|
|
},
|
|
{
|
|
"epoch": 0.6172839506172839,
|
|
"grad_norm": 0.4280650317668915,
|
|
"learning_rate": 3.9796349180597065e-05,
|
|
"loss": 0.227,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2339543104171753,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 0.6234567901234568,
|
|
"grad_norm": 0.3809901177883148,
|
|
"learning_rate": 3.978749228228883e-05,
|
|
"loss": 0.2129,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2053700089454651,
|
|
"step": 805
|
|
},
|
|
{
|
|
"epoch": 0.6296296296296297,
|
|
"grad_norm": 0.3569866418838501,
|
|
"learning_rate": 3.9778447893581745e-05,
|
|
"loss": 0.2226,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20748861134052277,
|
|
"step": 810
|
|
},
|
|
{
|
|
"epoch": 0.6358024691358025,
|
|
"grad_norm": 0.38599514961242676,
|
|
"learning_rate": 3.976921610017317e-05,
|
|
"loss": 0.2412,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27881893515586853,
|
|
"step": 815
|
|
},
|
|
{
|
|
"epoch": 0.6419753086419753,
|
|
"grad_norm": 0.9968234896659851,
|
|
"learning_rate": 3.9759796989536185e-05,
|
|
"loss": 0.324,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.43549323081970215,
|
|
"step": 820
|
|
},
|
|
{
|
|
"epoch": 0.6481481481481481,
|
|
"grad_norm": 0.5863314270973206,
|
|
"learning_rate": 3.9750190650918715e-05,
|
|
"loss": 0.4896,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.4987928867340088,
|
|
"step": 825
|
|
},
|
|
{
|
|
"epoch": 0.654320987654321,
|
|
"grad_norm": 0.5282703638076782,
|
|
"learning_rate": 3.974039717534272e-05,
|
|
"loss": 0.4671,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.4198218882083893,
|
|
"step": 830
|
|
},
|
|
{
|
|
"epoch": 0.6604938271604939,
|
|
"grad_norm": 0.49933239817619324,
|
|
"learning_rate": 3.973041665560331e-05,
|
|
"loss": 0.431,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.44005826115608215,
|
|
"step": 835
|
|
},
|
|
{
|
|
"epoch": 0.6666666666666666,
|
|
"grad_norm": 0.4007917046546936,
|
|
"learning_rate": 3.972024918626789e-05,
|
|
"loss": 0.4485,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.4326408803462982,
|
|
"step": 840
|
|
},
|
|
{
|
|
"epoch": 0.6728395061728395,
|
|
"grad_norm": 0.37785807251930237,
|
|
"learning_rate": 3.9709894863675235e-05,
|
|
"loss": 0.4407,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.40405717492103577,
|
|
"step": 845
|
|
},
|
|
{
|
|
"epoch": 0.6790123456790124,
|
|
"grad_norm": 0.32629433274269104,
|
|
"learning_rate": 3.969935378593461e-05,
|
|
"loss": 0.3872,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.38561010360717773,
|
|
"step": 850
|
|
},
|
|
{
|
|
"epoch": 0.6851851851851852,
|
|
"grad_norm": 0.35349494218826294,
|
|
"learning_rate": 3.968862605292479e-05,
|
|
"loss": 0.3937,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3514169752597809,
|
|
"step": 855
|
|
},
|
|
{
|
|
"epoch": 0.691358024691358,
|
|
"grad_norm": 0.31500837206840515,
|
|
"learning_rate": 3.9677711766293175e-05,
|
|
"loss": 0.3908,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3900444507598877,
|
|
"step": 860
|
|
},
|
|
{
|
|
"epoch": 0.6975308641975309,
|
|
"grad_norm": 0.33248379826545715,
|
|
"learning_rate": 3.966661102945479e-05,
|
|
"loss": 0.3922,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3730594217777252,
|
|
"step": 865
|
|
},
|
|
{
|
|
"epoch": 0.7037037037037037,
|
|
"grad_norm": 0.3747028410434723,
|
|
"learning_rate": 3.9655323947591305e-05,
|
|
"loss": 0.3568,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.34837284684181213,
|
|
"step": 870
|
|
},
|
|
{
|
|
"epoch": 0.7098765432098766,
|
|
"grad_norm": 0.3031326234340668,
|
|
"learning_rate": 3.964385062765005e-05,
|
|
"loss": 0.3703,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3968775272369385,
|
|
"step": 875
|
|
},
|
|
{
|
|
"epoch": 0.7160493827160493,
|
|
"grad_norm": 0.3189447820186615,
|
|
"learning_rate": 3.9632191178343e-05,
|
|
"loss": 0.388,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3416036069393158,
|
|
"step": 880
|
|
},
|
|
{
|
|
"epoch": 0.7222222222222222,
|
|
"grad_norm": 0.3240283727645874,
|
|
"learning_rate": 3.9620345710145735e-05,
|
|
"loss": 0.3848,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.32970184087753296,
|
|
"step": 885
|
|
},
|
|
{
|
|
"epoch": 0.7283950617283951,
|
|
"grad_norm": 0.3894062936306,
|
|
"learning_rate": 3.960831433529642e-05,
|
|
"loss": 0.3571,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.32407477498054504,
|
|
"step": 890
|
|
},
|
|
{
|
|
"epoch": 0.7345679012345679,
|
|
"grad_norm": 0.3772331178188324,
|
|
"learning_rate": 3.95960971677947e-05,
|
|
"loss": 0.3661,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3844054639339447,
|
|
"step": 895
|
|
},
|
|
{
|
|
"epoch": 0.7407407407407407,
|
|
"grad_norm": 0.3585491478443146,
|
|
"learning_rate": 3.958369432340065e-05,
|
|
"loss": 0.3662,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.36225321888923645,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 0.7469135802469136,
|
|
"grad_norm": 0.36180633306503296,
|
|
"learning_rate": 3.957110591963368e-05,
|
|
"loss": 0.3898,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.34740471839904785,
|
|
"step": 905
|
|
},
|
|
{
|
|
"epoch": 0.7530864197530864,
|
|
"grad_norm": 0.3298802375793457,
|
|
"learning_rate": 3.9558332075771384e-05,
|
|
"loss": 0.3383,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.31783321499824524,
|
|
"step": 910
|
|
},
|
|
{
|
|
"epoch": 0.7592592592592593,
|
|
"grad_norm": 0.3728399872779846,
|
|
"learning_rate": 3.954537291284847e-05,
|
|
"loss": 0.3523,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.38530659675598145,
|
|
"step": 915
|
|
},
|
|
{
|
|
"epoch": 0.7654320987654321,
|
|
"grad_norm": 0.331874817609787,
|
|
"learning_rate": 3.953222855365557e-05,
|
|
"loss": 0.3552,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.37151381373405457,
|
|
"step": 920
|
|
},
|
|
{
|
|
"epoch": 0.7716049382716049,
|
|
"grad_norm": 0.36094093322753906,
|
|
"learning_rate": 3.951889912273807e-05,
|
|
"loss": 0.3867,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.4306624233722687,
|
|
"step": 925
|
|
},
|
|
{
|
|
"epoch": 0.7777777777777778,
|
|
"grad_norm": 0.4501263201236725,
|
|
"learning_rate": 3.950538474639498e-05,
|
|
"loss": 0.37,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3891269266605377,
|
|
"step": 930
|
|
},
|
|
{
|
|
"epoch": 0.7839506172839507,
|
|
"grad_norm": 0.3315528333187103,
|
|
"learning_rate": 3.9491685552677674e-05,
|
|
"loss": 0.3742,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.34857630729675293,
|
|
"step": 935
|
|
},
|
|
{
|
|
"epoch": 0.7901234567901234,
|
|
"grad_norm": 0.3106090724468231,
|
|
"learning_rate": 3.9477801671388725e-05,
|
|
"loss": 0.3654,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3707601726055145,
|
|
"step": 940
|
|
},
|
|
{
|
|
"epoch": 0.7962962962962963,
|
|
"grad_norm": 0.3356432020664215,
|
|
"learning_rate": 3.946373323408065e-05,
|
|
"loss": 0.3679,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.4207797944545746,
|
|
"step": 945
|
|
},
|
|
{
|
|
"epoch": 0.8024691358024691,
|
|
"grad_norm": 0.3268519639968872,
|
|
"learning_rate": 3.944948037405466e-05,
|
|
"loss": 0.3736,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.4086616039276123,
|
|
"step": 950
|
|
},
|
|
{
|
|
"epoch": 0.808641975308642,
|
|
"grad_norm": 0.32324424386024475,
|
|
"learning_rate": 3.943504322635944e-05,
|
|
"loss": 0.3631,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.32783088088035583,
|
|
"step": 955
|
|
},
|
|
{
|
|
"epoch": 0.8148148148148148,
|
|
"grad_norm": 0.3585626482963562,
|
|
"learning_rate": 3.94204219277898e-05,
|
|
"loss": 0.3588,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.35287073254585266,
|
|
"step": 960
|
|
},
|
|
{
|
|
"epoch": 0.8209876543209876,
|
|
"grad_norm": 0.31093865633010864,
|
|
"learning_rate": 3.940561661688544e-05,
|
|
"loss": 0.3456,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.31583473086357117,
|
|
"step": 965
|
|
},
|
|
{
|
|
"epoch": 0.8271604938271605,
|
|
"grad_norm": 1.2907553911209106,
|
|
"learning_rate": 3.9390627433929615e-05,
|
|
"loss": 0.3553,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3133183419704437,
|
|
"step": 970
|
|
},
|
|
{
|
|
"epoch": 0.8333333333333334,
|
|
"grad_norm": 0.2955177426338196,
|
|
"learning_rate": 3.937545452094777e-05,
|
|
"loss": 0.3505,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.34256863594055176,
|
|
"step": 975
|
|
},
|
|
{
|
|
"epoch": 0.8395061728395061,
|
|
"grad_norm": 0.30115678906440735,
|
|
"learning_rate": 3.936009802170628e-05,
|
|
"loss": 0.3528,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3723667860031128,
|
|
"step": 980
|
|
},
|
|
{
|
|
"epoch": 0.845679012345679,
|
|
"grad_norm": 0.3079809248447418,
|
|
"learning_rate": 3.934455808171099e-05,
|
|
"loss": 0.3625,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.42024853825569153,
|
|
"step": 985
|
|
},
|
|
{
|
|
"epoch": 0.8518518518518519,
|
|
"grad_norm": 0.30463579297065735,
|
|
"learning_rate": 3.932883484820592e-05,
|
|
"loss": 0.3402,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.29944154620170593,
|
|
"step": 990
|
|
},
|
|
{
|
|
"epoch": 0.8580246913580247,
|
|
"grad_norm": 0.31586068868637085,
|
|
"learning_rate": 3.931292847017179e-05,
|
|
"loss": 0.359,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3427981436252594,
|
|
"step": 995
|
|
},
|
|
{
|
|
"epoch": 0.8641975308641975,
|
|
"grad_norm": 0.31005361676216125,
|
|
"learning_rate": 3.92968390983247e-05,
|
|
"loss": 0.3499,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3636094033718109,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 0.8703703703703703,
|
|
"grad_norm": 0.2949121594429016,
|
|
"learning_rate": 3.928056688511461e-05,
|
|
"loss": 0.356,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3200506865978241,
|
|
"step": 1005
|
|
},
|
|
{
|
|
"epoch": 0.8765432098765432,
|
|
"grad_norm": 1.265420913696289,
|
|
"learning_rate": 3.9264111984723966e-05,
|
|
"loss": 0.343,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3551841080188751,
|
|
"step": 1010
|
|
},
|
|
{
|
|
"epoch": 0.8827160493827161,
|
|
"grad_norm": 0.5925858020782471,
|
|
"learning_rate": 3.9247474553066194e-05,
|
|
"loss": 0.2913,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25881585478782654,
|
|
"step": 1015
|
|
},
|
|
{
|
|
"epoch": 0.8888888888888888,
|
|
"grad_norm": 0.4837743937969208,
|
|
"learning_rate": 3.9230654747784245e-05,
|
|
"loss": 0.277,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27934566140174866,
|
|
"step": 1020
|
|
},
|
|
{
|
|
"epoch": 0.8950617283950617,
|
|
"grad_norm": 0.47393015027046204,
|
|
"learning_rate": 3.9213652728249095e-05,
|
|
"loss": 0.3065,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2953484356403351,
|
|
"step": 1025
|
|
},
|
|
{
|
|
"epoch": 0.9012345679012346,
|
|
"grad_norm": 0.40508612990379333,
|
|
"learning_rate": 3.919646865555824e-05,
|
|
"loss": 0.3107,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2884358763694763,
|
|
"step": 1030
|
|
},
|
|
{
|
|
"epoch": 0.9074074074074074,
|
|
"grad_norm": 0.36504748463630676,
|
|
"learning_rate": 3.917910269253417e-05,
|
|
"loss": 0.2898,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25382792949676514,
|
|
"step": 1035
|
|
},
|
|
{
|
|
"epoch": 0.9135802469135802,
|
|
"grad_norm": 0.36263060569763184,
|
|
"learning_rate": 3.9161555003722816e-05,
|
|
"loss": 0.2873,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3035801649093628,
|
|
"step": 1040
|
|
},
|
|
{
|
|
"epoch": 0.9197530864197531,
|
|
"grad_norm": 0.391148179769516,
|
|
"learning_rate": 3.914382575539199e-05,
|
|
"loss": 0.2853,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.279092937707901,
|
|
"step": 1045
|
|
},
|
|
{
|
|
"epoch": 0.9259259259259259,
|
|
"grad_norm": 0.34417814016342163,
|
|
"learning_rate": 3.912591511552982e-05,
|
|
"loss": 0.2792,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2636912763118744,
|
|
"step": 1050
|
|
},
|
|
{
|
|
"epoch": 0.9320987654320988,
|
|
"grad_norm": 0.4070587754249573,
|
|
"learning_rate": 3.9107823253843165e-05,
|
|
"loss": 0.2964,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3054562509059906,
|
|
"step": 1055
|
|
},
|
|
{
|
|
"epoch": 0.9382716049382716,
|
|
"grad_norm": 0.41854435205459595,
|
|
"learning_rate": 3.9089550341755984e-05,
|
|
"loss": 0.2943,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3032470941543579,
|
|
"step": 1060
|
|
},
|
|
{
|
|
"epoch": 0.9444444444444444,
|
|
"grad_norm": 0.4005330801010132,
|
|
"learning_rate": 3.907109655240774e-05,
|
|
"loss": 0.318,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.34182479977607727,
|
|
"step": 1065
|
|
},
|
|
{
|
|
"epoch": 0.9506172839506173,
|
|
"grad_norm": 0.3375258147716522,
|
|
"learning_rate": 3.9052462060651715e-05,
|
|
"loss": 0.3156,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2863449156284332,
|
|
"step": 1070
|
|
},
|
|
{
|
|
"epoch": 0.9567901234567902,
|
|
"grad_norm": 0.32006749510765076,
|
|
"learning_rate": 3.903364704305341e-05,
|
|
"loss": 0.3119,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.33581599593162537,
|
|
"step": 1075
|
|
},
|
|
{
|
|
"epoch": 0.9629629629629629,
|
|
"grad_norm": 0.44205722212791443,
|
|
"learning_rate": 3.901465167788885e-05,
|
|
"loss": 0.2993,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2816407084465027,
|
|
"step": 1080
|
|
},
|
|
{
|
|
"epoch": 0.9691358024691358,
|
|
"grad_norm": 0.4949347972869873,
|
|
"learning_rate": 3.899547614514283e-05,
|
|
"loss": 0.2981,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.30001768469810486,
|
|
"step": 1085
|
|
},
|
|
{
|
|
"epoch": 0.9753086419753086,
|
|
"grad_norm": 0.34502044320106506,
|
|
"learning_rate": 3.897612062650736e-05,
|
|
"loss": 0.2757,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25090089440345764,
|
|
"step": 1090
|
|
},
|
|
{
|
|
"epoch": 0.9814814814814815,
|
|
"grad_norm": 0.424858957529068,
|
|
"learning_rate": 3.895658530537977e-05,
|
|
"loss": 0.2639,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24759306013584137,
|
|
"step": 1095
|
|
},
|
|
{
|
|
"epoch": 0.9876543209876543,
|
|
"grad_norm": 0.36861452460289,
|
|
"learning_rate": 3.8936870366861104e-05,
|
|
"loss": 0.2641,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2502908706665039,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"epoch": 0.9938271604938271,
|
|
"grad_norm": 0.41577473282814026,
|
|
"learning_rate": 3.8916975997754287e-05,
|
|
"loss": 0.2666,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.28242433071136475,
|
|
"step": 1105
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"grad_norm": 0.740837037563324,
|
|
"learning_rate": 3.889690238656241e-05,
|
|
"loss": 0.2906,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3180903494358063,
|
|
"step": 1110
|
|
},
|
|
{
|
|
"epoch": 1.0061728395061729,
|
|
"grad_norm": 0.4693143963813782,
|
|
"learning_rate": 3.88766497234869e-05,
|
|
"loss": 0.2576,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24789981544017792,
|
|
"step": 1115
|
|
},
|
|
{
|
|
"epoch": 1.0123456790123457,
|
|
"grad_norm": 0.41690993309020996,
|
|
"learning_rate": 3.8856218200425734e-05,
|
|
"loss": 0.2457,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22385604679584503,
|
|
"step": 1120
|
|
},
|
|
{
|
|
"epoch": 1.0185185185185186,
|
|
"grad_norm": 0.4567547142505646,
|
|
"learning_rate": 3.8835608010971645e-05,
|
|
"loss": 0.2386,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23406028747558594,
|
|
"step": 1125
|
|
},
|
|
{
|
|
"epoch": 1.0246913580246915,
|
|
"grad_norm": 0.37873566150665283,
|
|
"learning_rate": 3.8814819350410224e-05,
|
|
"loss": 0.2355,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24889661371707916,
|
|
"step": 1130
|
|
},
|
|
{
|
|
"epoch": 1.0308641975308641,
|
|
"grad_norm": 0.3074162006378174,
|
|
"learning_rate": 3.879385241571817e-05,
|
|
"loss": 0.2388,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2647235691547394,
|
|
"step": 1135
|
|
},
|
|
{
|
|
"epoch": 1.037037037037037,
|
|
"grad_norm": 0.33328789472579956,
|
|
"learning_rate": 3.8772707405561315e-05,
|
|
"loss": 0.2298,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25016680359840393,
|
|
"step": 1140
|
|
},
|
|
{
|
|
"epoch": 1.0432098765432098,
|
|
"grad_norm": 0.38928186893463135,
|
|
"learning_rate": 3.87513845202928e-05,
|
|
"loss": 0.2394,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2616216838359833,
|
|
"step": 1145
|
|
},
|
|
{
|
|
"epoch": 1.0493827160493827,
|
|
"grad_norm": 0.3370906412601471,
|
|
"learning_rate": 3.872988396195119e-05,
|
|
"loss": 0.2353,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20379845798015594,
|
|
"step": 1150
|
|
},
|
|
{
|
|
"epoch": 1.0555555555555556,
|
|
"grad_norm": 0.4154044985771179,
|
|
"learning_rate": 3.870820593425851e-05,
|
|
"loss": 0.2309,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2301466464996338,
|
|
"step": 1155
|
|
},
|
|
{
|
|
"epoch": 1.0617283950617284,
|
|
"grad_norm": 0.3356035351753235,
|
|
"learning_rate": 3.868635064261837e-05,
|
|
"loss": 0.2193,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2465057224035263,
|
|
"step": 1160
|
|
},
|
|
{
|
|
"epoch": 1.0679012345679013,
|
|
"grad_norm": 0.5215581655502319,
|
|
"learning_rate": 3.8664318294113974e-05,
|
|
"loss": 0.2474,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2485499382019043,
|
|
"step": 1165
|
|
},
|
|
{
|
|
"epoch": 1.074074074074074,
|
|
"grad_norm": 0.4347217381000519,
|
|
"learning_rate": 3.864210909750616e-05,
|
|
"loss": 0.2246,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22813738882541656,
|
|
"step": 1170
|
|
},
|
|
{
|
|
"epoch": 1.0802469135802468,
|
|
"grad_norm": 0.4194651246070862,
|
|
"learning_rate": 3.861972326323148e-05,
|
|
"loss": 0.2429,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2675498425960541,
|
|
"step": 1175
|
|
},
|
|
{
|
|
"epoch": 1.0864197530864197,
|
|
"grad_norm": 0.44267013669013977,
|
|
"learning_rate": 3.8597161003400116e-05,
|
|
"loss": 0.2461,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2711530029773712,
|
|
"step": 1180
|
|
},
|
|
{
|
|
"epoch": 1.0925925925925926,
|
|
"grad_norm": 0.3669820725917816,
|
|
"learning_rate": 3.8574422531793943e-05,
|
|
"loss": 0.2359,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2184760421514511,
|
|
"step": 1185
|
|
},
|
|
{
|
|
"epoch": 1.0987654320987654,
|
|
"grad_norm": 0.44124099612236023,
|
|
"learning_rate": 3.8551508063864475e-05,
|
|
"loss": 0.2547,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23093067109584808,
|
|
"step": 1190
|
|
},
|
|
{
|
|
"epoch": 1.1049382716049383,
|
|
"grad_norm": 0.4369843006134033,
|
|
"learning_rate": 3.852841781673082e-05,
|
|
"loss": 0.2283,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2573466897010803,
|
|
"step": 1195
|
|
},
|
|
{
|
|
"epoch": 1.1111111111111112,
|
|
"grad_norm": 0.38986849784851074,
|
|
"learning_rate": 3.8505152009177636e-05,
|
|
"loss": 0.2286,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2178529053926468,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"epoch": 1.117283950617284,
|
|
"grad_norm": 0.40066277980804443,
|
|
"learning_rate": 3.8481710861653036e-05,
|
|
"loss": 0.2345,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22776807844638824,
|
|
"step": 1205
|
|
},
|
|
{
|
|
"epoch": 1.123456790123457,
|
|
"grad_norm": 0.3570098280906677,
|
|
"learning_rate": 3.845809459626654e-05,
|
|
"loss": 0.2231,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23185652494430542,
|
|
"step": 1210
|
|
},
|
|
{
|
|
"epoch": 1.1296296296296295,
|
|
"grad_norm": 0.4078414738178253,
|
|
"learning_rate": 3.843430343678692e-05,
|
|
"loss": 0.2288,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2155301421880722,
|
|
"step": 1215
|
|
},
|
|
{
|
|
"epoch": 1.1358024691358024,
|
|
"grad_norm": 0.44162988662719727,
|
|
"learning_rate": 3.841033760864011e-05,
|
|
"loss": 0.2383,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21442461013793945,
|
|
"step": 1220
|
|
},
|
|
{
|
|
"epoch": 1.1419753086419753,
|
|
"grad_norm": 0.38371214270591736,
|
|
"learning_rate": 3.8386197338907064e-05,
|
|
"loss": 0.2451,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2360832244157791,
|
|
"step": 1225
|
|
},
|
|
{
|
|
"epoch": 1.1481481481481481,
|
|
"grad_norm": 0.39999914169311523,
|
|
"learning_rate": 3.836188285632161e-05,
|
|
"loss": 0.2283,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22600699961185455,
|
|
"step": 1230
|
|
},
|
|
{
|
|
"epoch": 1.154320987654321,
|
|
"grad_norm": 0.3859078884124756,
|
|
"learning_rate": 3.833739439126828e-05,
|
|
"loss": 0.224,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23709864914417267,
|
|
"step": 1235
|
|
},
|
|
{
|
|
"epoch": 1.1604938271604939,
|
|
"grad_norm": 0.38821154832839966,
|
|
"learning_rate": 3.8312732175780124e-05,
|
|
"loss": 0.2445,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22265632450580597,
|
|
"step": 1240
|
|
},
|
|
{
|
|
"epoch": 1.1666666666666667,
|
|
"grad_norm": 0.43809249997138977,
|
|
"learning_rate": 3.828789644353648e-05,
|
|
"loss": 0.234,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.220628559589386,
|
|
"step": 1245
|
|
},
|
|
{
|
|
"epoch": 1.1728395061728394,
|
|
"grad_norm": 0.4124487340450287,
|
|
"learning_rate": 3.826288742986085e-05,
|
|
"loss": 0.237,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2575366199016571,
|
|
"step": 1250
|
|
},
|
|
{
|
|
"epoch": 1.1790123456790123,
|
|
"grad_norm": 0.45238590240478516,
|
|
"learning_rate": 3.823770537171856e-05,
|
|
"loss": 0.2295,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2458871752023697,
|
|
"step": 1255
|
|
},
|
|
{
|
|
"epoch": 1.1851851851851851,
|
|
"grad_norm": 0.3698289394378662,
|
|
"learning_rate": 3.82123505077146e-05,
|
|
"loss": 0.2189,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2447344809770584,
|
|
"step": 1260
|
|
},
|
|
{
|
|
"epoch": 1.191358024691358,
|
|
"grad_norm": 0.38862428069114685,
|
|
"learning_rate": 3.8186823078091296e-05,
|
|
"loss": 0.233,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2283145934343338,
|
|
"step": 1265
|
|
},
|
|
{
|
|
"epoch": 1.1975308641975309,
|
|
"grad_norm": 0.3315979242324829,
|
|
"learning_rate": 3.81611233247261e-05,
|
|
"loss": 0.234,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19199562072753906,
|
|
"step": 1270
|
|
},
|
|
{
|
|
"epoch": 1.2037037037037037,
|
|
"grad_norm": 0.43696150183677673,
|
|
"learning_rate": 3.813525149112926e-05,
|
|
"loss": 0.2129,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2054680585861206,
|
|
"step": 1275
|
|
},
|
|
{
|
|
"epoch": 1.2098765432098766,
|
|
"grad_norm": 0.31870439648628235,
|
|
"learning_rate": 3.810920782244149e-05,
|
|
"loss": 0.2158,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22038424015045166,
|
|
"step": 1280
|
|
},
|
|
{
|
|
"epoch": 1.2160493827160495,
|
|
"grad_norm": 0.3521271049976349,
|
|
"learning_rate": 3.808299256543172e-05,
|
|
"loss": 0.2295,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24346429109573364,
|
|
"step": 1285
|
|
},
|
|
{
|
|
"epoch": 1.2222222222222223,
|
|
"grad_norm": 0.39431071281433105,
|
|
"learning_rate": 3.8056605968494664e-05,
|
|
"loss": 0.2188,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24782423675060272,
|
|
"step": 1290
|
|
},
|
|
{
|
|
"epoch": 1.228395061728395,
|
|
"grad_norm": 0.5386034846305847,
|
|
"learning_rate": 3.803004828164856e-05,
|
|
"loss": 0.2263,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2433486133813858,
|
|
"step": 1295
|
|
},
|
|
{
|
|
"epoch": 1.2345679012345678,
|
|
"grad_norm": 0.4058195948600769,
|
|
"learning_rate": 3.8003319756532736e-05,
|
|
"loss": 0.2359,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25290000438690186,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"epoch": 1.2407407407407407,
|
|
"grad_norm": 0.435697078704834,
|
|
"learning_rate": 3.797642064640525e-05,
|
|
"loss": 0.228,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22705914080142975,
|
|
"step": 1305
|
|
},
|
|
{
|
|
"epoch": 1.2469135802469136,
|
|
"grad_norm": 0.4043845534324646,
|
|
"learning_rate": 3.7949351206140476e-05,
|
|
"loss": 0.2287,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25260791182518005,
|
|
"step": 1310
|
|
},
|
|
{
|
|
"epoch": 1.2530864197530864,
|
|
"grad_norm": 0.40700969099998474,
|
|
"learning_rate": 3.7922111692226716e-05,
|
|
"loss": 0.2385,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21020638942718506,
|
|
"step": 1315
|
|
},
|
|
{
|
|
"epoch": 1.2592592592592593,
|
|
"grad_norm": 0.46284958720207214,
|
|
"learning_rate": 3.789470236276374e-05,
|
|
"loss": 0.2259,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21826355159282684,
|
|
"step": 1320
|
|
},
|
|
{
|
|
"epoch": 1.2654320987654322,
|
|
"grad_norm": 0.4180838167667389,
|
|
"learning_rate": 3.786712347746036e-05,
|
|
"loss": 0.2157,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23369455337524414,
|
|
"step": 1325
|
|
},
|
|
{
|
|
"epoch": 1.2716049382716048,
|
|
"grad_norm": 0.43176138401031494,
|
|
"learning_rate": 3.7839375297631976e-05,
|
|
"loss": 0.2269,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24230773746967316,
|
|
"step": 1330
|
|
},
|
|
{
|
|
"epoch": 1.2777777777777777,
|
|
"grad_norm": 0.45842283964157104,
|
|
"learning_rate": 3.781145808619806e-05,
|
|
"loss": 0.2219,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19564318656921387,
|
|
"step": 1335
|
|
},
|
|
{
|
|
"epoch": 1.2839506172839505,
|
|
"grad_norm": 0.3865016996860504,
|
|
"learning_rate": 3.778337210767971e-05,
|
|
"loss": 0.2169,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.216106116771698,
|
|
"step": 1340
|
|
},
|
|
{
|
|
"epoch": 1.2901234567901234,
|
|
"grad_norm": 0.4970242977142334,
|
|
"learning_rate": 3.775511762819713e-05,
|
|
"loss": 0.2339,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2391725331544876,
|
|
"step": 1345
|
|
},
|
|
{
|
|
"epoch": 1.2962962962962963,
|
|
"grad_norm": 0.4439508020877838,
|
|
"learning_rate": 3.772669491546707e-05,
|
|
"loss": 0.2259,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2305746078491211,
|
|
"step": 1350
|
|
},
|
|
{
|
|
"epoch": 1.3024691358024691,
|
|
"grad_norm": 0.43414968252182007,
|
|
"learning_rate": 3.7698104238800375e-05,
|
|
"loss": 0.2254,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21781843900680542,
|
|
"step": 1355
|
|
},
|
|
{
|
|
"epoch": 1.308641975308642,
|
|
"grad_norm": 0.5810720324516296,
|
|
"learning_rate": 3.766934586909935e-05,
|
|
"loss": 0.1976,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22130520641803741,
|
|
"step": 1360
|
|
},
|
|
{
|
|
"epoch": 1.3148148148148149,
|
|
"grad_norm": 0.43914851546287537,
|
|
"learning_rate": 3.7640420078855224e-05,
|
|
"loss": 0.216,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21131666004657745,
|
|
"step": 1365
|
|
},
|
|
{
|
|
"epoch": 1.3209876543209877,
|
|
"grad_norm": 1.0866725444793701,
|
|
"learning_rate": 3.761132714214558e-05,
|
|
"loss": 0.2374,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2346685379743576,
|
|
"step": 1370
|
|
},
|
|
{
|
|
"epoch": 1.3271604938271606,
|
|
"grad_norm": 0.5554916858673096,
|
|
"learning_rate": 3.758206733463173e-05,
|
|
"loss": 0.2069,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1766139715909958,
|
|
"step": 1375
|
|
},
|
|
{
|
|
"epoch": 1.3333333333333333,
|
|
"grad_norm": 0.41587650775909424,
|
|
"learning_rate": 3.755264093355614e-05,
|
|
"loss": 0.2103,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21381127834320068,
|
|
"step": 1380
|
|
},
|
|
{
|
|
"epoch": 1.3395061728395061,
|
|
"grad_norm": 0.3941839635372162,
|
|
"learning_rate": 3.7523048217739785e-05,
|
|
"loss": 0.2237,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23770034313201904,
|
|
"step": 1385
|
|
},
|
|
{
|
|
"epoch": 1.345679012345679,
|
|
"grad_norm": 0.43083909153938293,
|
|
"learning_rate": 3.749328946757947e-05,
|
|
"loss": 0.2432,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21662195026874542,
|
|
"step": 1390
|
|
},
|
|
{
|
|
"epoch": 1.3518518518518519,
|
|
"grad_norm": 0.4212723672389984,
|
|
"learning_rate": 3.746336496504525e-05,
|
|
"loss": 0.2152,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.216409370303154,
|
|
"step": 1395
|
|
},
|
|
{
|
|
"epoch": 1.3580246913580247,
|
|
"grad_norm": 0.4318290650844574,
|
|
"learning_rate": 3.7433274993677696e-05,
|
|
"loss": 0.2214,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22673408687114716,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"epoch": 1.3641975308641976,
|
|
"grad_norm": 0.3736567795276642,
|
|
"learning_rate": 3.740301983858522e-05,
|
|
"loss": 0.2095,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1957586407661438,
|
|
"step": 1405
|
|
},
|
|
{
|
|
"epoch": 1.3703703703703702,
|
|
"grad_norm": 0.4058263599872589,
|
|
"learning_rate": 3.7372599786441414e-05,
|
|
"loss": 0.2271,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23144106566905975,
|
|
"step": 1410
|
|
},
|
|
{
|
|
"epoch": 1.376543209876543,
|
|
"grad_norm": 0.368058979511261,
|
|
"learning_rate": 3.734201512548226e-05,
|
|
"loss": 0.2272,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20349471271038055,
|
|
"step": 1415
|
|
},
|
|
{
|
|
"epoch": 1.382716049382716,
|
|
"grad_norm": 0.3930149972438812,
|
|
"learning_rate": 3.7311266145503486e-05,
|
|
"loss": 0.2383,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24719738960266113,
|
|
"step": 1420
|
|
},
|
|
{
|
|
"epoch": 1.3888888888888888,
|
|
"grad_norm": 0.38044217228889465,
|
|
"learning_rate": 3.728035313785774e-05,
|
|
"loss": 0.2325,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2327665239572525,
|
|
"step": 1425
|
|
},
|
|
{
|
|
"epoch": 1.3950617283950617,
|
|
"grad_norm": 0.4118587076663971,
|
|
"learning_rate": 3.724927639545187e-05,
|
|
"loss": 0.2094,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1817220002412796,
|
|
"step": 1430
|
|
},
|
|
{
|
|
"epoch": 1.4012345679012346,
|
|
"grad_norm": 0.3992430567741394,
|
|
"learning_rate": 3.721803621274416e-05,
|
|
"loss": 0.2091,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2075822800397873,
|
|
"step": 1435
|
|
},
|
|
{
|
|
"epoch": 1.4074074074074074,
|
|
"grad_norm": 0.38057300448417664,
|
|
"learning_rate": 3.718663288574152e-05,
|
|
"loss": 0.2222,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20726530253887177,
|
|
"step": 1440
|
|
},
|
|
{
|
|
"epoch": 1.4135802469135803,
|
|
"grad_norm": 0.3660273849964142,
|
|
"learning_rate": 3.7155066711996664e-05,
|
|
"loss": 0.2165,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2451813668012619,
|
|
"step": 1445
|
|
},
|
|
{
|
|
"epoch": 1.4197530864197532,
|
|
"grad_norm": 0.37525391578674316,
|
|
"learning_rate": 3.7123337990605335e-05,
|
|
"loss": 0.2239,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2433450073003769,
|
|
"step": 1450
|
|
},
|
|
{
|
|
"epoch": 1.425925925925926,
|
|
"grad_norm": 0.4378054141998291,
|
|
"learning_rate": 3.709144702220343e-05,
|
|
"loss": 0.215,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21114008128643036,
|
|
"step": 1455
|
|
},
|
|
{
|
|
"epoch": 1.4320987654320987,
|
|
"grad_norm": 0.37331220507621765,
|
|
"learning_rate": 3.705939410896418e-05,
|
|
"loss": 0.2262,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2206510752439499,
|
|
"step": 1460
|
|
},
|
|
{
|
|
"epoch": 1.4382716049382716,
|
|
"grad_norm": 0.36982277035713196,
|
|
"learning_rate": 3.7027179554595255e-05,
|
|
"loss": 0.2152,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21671412885189056,
|
|
"step": 1465
|
|
},
|
|
{
|
|
"epoch": 1.4444444444444444,
|
|
"grad_norm": 0.5762490034103394,
|
|
"learning_rate": 3.699480366433595e-05,
|
|
"loss": 0.2268,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2262163609266281,
|
|
"step": 1470
|
|
},
|
|
{
|
|
"epoch": 1.4506172839506173,
|
|
"grad_norm": 0.37613335251808167,
|
|
"learning_rate": 3.696226674495421e-05,
|
|
"loss": 0.2064,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22146214544773102,
|
|
"step": 1475
|
|
},
|
|
{
|
|
"epoch": 1.4567901234567902,
|
|
"grad_norm": 0.3797090947628021,
|
|
"learning_rate": 3.6929569104743764e-05,
|
|
"loss": 0.2165,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.209309920668602,
|
|
"step": 1480
|
|
},
|
|
{
|
|
"epoch": 1.462962962962963,
|
|
"grad_norm": 0.3595726191997528,
|
|
"learning_rate": 3.6896711053521226e-05,
|
|
"loss": 0.2141,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19577650725841522,
|
|
"step": 1485
|
|
},
|
|
{
|
|
"epoch": 1.4691358024691357,
|
|
"grad_norm": 0.5154550671577454,
|
|
"learning_rate": 3.686369290262311e-05,
|
|
"loss": 0.2005,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21102793514728546,
|
|
"step": 1490
|
|
},
|
|
{
|
|
"epoch": 1.4753086419753085,
|
|
"grad_norm": 0.41036584973335266,
|
|
"learning_rate": 3.683051496490294e-05,
|
|
"loss": 0.2121,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19589179754257202,
|
|
"step": 1495
|
|
},
|
|
{
|
|
"epoch": 1.4814814814814814,
|
|
"grad_norm": 0.4689701497554779,
|
|
"learning_rate": 3.679717755472822e-05,
|
|
"loss": 0.2344,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2274472862482071,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"epoch": 1.4876543209876543,
|
|
"grad_norm": 0.4193136990070343,
|
|
"learning_rate": 3.676368098797749e-05,
|
|
"loss": 0.211,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2021634727716446,
|
|
"step": 1505
|
|
},
|
|
{
|
|
"epoch": 1.4938271604938271,
|
|
"grad_norm": 0.4441111981868744,
|
|
"learning_rate": 3.6730025582037345e-05,
|
|
"loss": 0.2093,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20014601945877075,
|
|
"step": 1510
|
|
},
|
|
{
|
|
"epoch": 1.5,
|
|
"grad_norm": 0.7137706875801086,
|
|
"learning_rate": 3.66962116557994e-05,
|
|
"loss": 0.2027,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1776660680770874,
|
|
"step": 1515
|
|
},
|
|
{
|
|
"epoch": 1.5061728395061729,
|
|
"grad_norm": 0.3689514398574829,
|
|
"learning_rate": 3.66622395296573e-05,
|
|
"loss": 0.1945,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.199792742729187,
|
|
"step": 1520
|
|
},
|
|
{
|
|
"epoch": 1.5123456790123457,
|
|
"grad_norm": 0.5324134826660156,
|
|
"learning_rate": 3.6628109525503646e-05,
|
|
"loss": 0.2009,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24959486722946167,
|
|
"step": 1525
|
|
},
|
|
{
|
|
"epoch": 1.5185185185185186,
|
|
"grad_norm": 0.4831373393535614,
|
|
"learning_rate": 3.659382196672697e-05,
|
|
"loss": 0.2159,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22984512150287628,
|
|
"step": 1530
|
|
},
|
|
{
|
|
"epoch": 1.5246913580246915,
|
|
"grad_norm": 0.4558382034301758,
|
|
"learning_rate": 3.655937717820866e-05,
|
|
"loss": 0.201,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2057984322309494,
|
|
"step": 1535
|
|
},
|
|
{
|
|
"epoch": 1.5308641975308643,
|
|
"grad_norm": 0.42106854915618896,
|
|
"learning_rate": 3.65247754863199e-05,
|
|
"loss": 0.2169,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2275214046239853,
|
|
"step": 1540
|
|
},
|
|
{
|
|
"epoch": 1.5370370370370372,
|
|
"grad_norm": 0.4007282853126526,
|
|
"learning_rate": 3.649001721891855e-05,
|
|
"loss": 0.2323,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2297786921262741,
|
|
"step": 1545
|
|
},
|
|
{
|
|
"epoch": 1.5432098765432098,
|
|
"grad_norm": 0.40163710713386536,
|
|
"learning_rate": 3.6455102705346075e-05,
|
|
"loss": 0.2271,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24219255149364471,
|
|
"step": 1550
|
|
},
|
|
{
|
|
"epoch": 1.5493827160493827,
|
|
"grad_norm": 0.6153103113174438,
|
|
"learning_rate": 3.642003227642437e-05,
|
|
"loss": 0.223,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2019280344247818,
|
|
"step": 1555
|
|
},
|
|
{
|
|
"epoch": 1.5555555555555556,
|
|
"grad_norm": 0.38740089535713196,
|
|
"learning_rate": 3.63848062644527e-05,
|
|
"loss": 0.2128,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22657372057437897,
|
|
"step": 1560
|
|
},
|
|
{
|
|
"epoch": 1.5617283950617284,
|
|
"grad_norm": 0.43764251470565796,
|
|
"learning_rate": 3.6349425003204464e-05,
|
|
"loss": 0.2397,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21528728306293488,
|
|
"step": 1565
|
|
},
|
|
{
|
|
"epoch": 1.567901234567901,
|
|
"grad_norm": 0.4086608588695526,
|
|
"learning_rate": 3.631388882792412e-05,
|
|
"loss": 0.2197,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1848958283662796,
|
|
"step": 1570
|
|
},
|
|
{
|
|
"epoch": 1.574074074074074,
|
|
"grad_norm": 0.3706486225128174,
|
|
"learning_rate": 3.6278198075323945e-05,
|
|
"loss": 0.2146,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24810834228992462,
|
|
"step": 1575
|
|
},
|
|
{
|
|
"epoch": 1.5802469135802468,
|
|
"grad_norm": 0.3584599196910858,
|
|
"learning_rate": 3.624235308358086e-05,
|
|
"loss": 0.2289,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24489863216876984,
|
|
"step": 1580
|
|
},
|
|
{
|
|
"epoch": 1.5864197530864197,
|
|
"grad_norm": 0.4363275468349457,
|
|
"learning_rate": 3.6206354192333264e-05,
|
|
"loss": 0.2198,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20726199448108673,
|
|
"step": 1585
|
|
},
|
|
{
|
|
"epoch": 1.5925925925925926,
|
|
"grad_norm": 0.4050152003765106,
|
|
"learning_rate": 3.617020174267776e-05,
|
|
"loss": 0.2037,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20467142760753632,
|
|
"step": 1590
|
|
},
|
|
{
|
|
"epoch": 1.5987654320987654,
|
|
"grad_norm": 0.40914762020111084,
|
|
"learning_rate": 3.6133896077165934e-05,
|
|
"loss": 0.207,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2183167189359665,
|
|
"step": 1595
|
|
},
|
|
{
|
|
"epoch": 1.6049382716049383,
|
|
"grad_norm": 0.3838684558868408,
|
|
"learning_rate": 3.609743753980116e-05,
|
|
"loss": 0.2072,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21854811906814575,
|
|
"step": 1600
|
|
},
|
|
{
|
|
"epoch": 1.6111111111111112,
|
|
"grad_norm": 0.37509503960609436,
|
|
"learning_rate": 3.606082647603527e-05,
|
|
"loss": 0.2185,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2384641319513321,
|
|
"step": 1605
|
|
},
|
|
{
|
|
"epoch": 1.617283950617284,
|
|
"grad_norm": 0.4176979660987854,
|
|
"learning_rate": 3.602406323276533e-05,
|
|
"loss": 0.2052,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21335192024707794,
|
|
"step": 1610
|
|
},
|
|
{
|
|
"epoch": 1.623456790123457,
|
|
"grad_norm": 0.3778628408908844,
|
|
"learning_rate": 3.5987148158330344e-05,
|
|
"loss": 0.1919,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1846359521150589,
|
|
"step": 1615
|
|
},
|
|
{
|
|
"epoch": 1.6296296296296298,
|
|
"grad_norm": 0.35883358120918274,
|
|
"learning_rate": 3.595008160250791e-05,
|
|
"loss": 0.2014,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18714821338653564,
|
|
"step": 1620
|
|
},
|
|
{
|
|
"epoch": 1.6358024691358026,
|
|
"grad_norm": 0.36836856603622437,
|
|
"learning_rate": 3.5912863916511e-05,
|
|
"loss": 0.2197,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25512298941612244,
|
|
"step": 1625
|
|
},
|
|
{
|
|
"epoch": 1.6419753086419753,
|
|
"grad_norm": 0.7618579864501953,
|
|
"learning_rate": 3.587549545298451e-05,
|
|
"loss": 0.2868,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.371713787317276,
|
|
"step": 1630
|
|
},
|
|
{
|
|
"epoch": 1.6481481481481481,
|
|
"grad_norm": 0.4923429787158966,
|
|
"learning_rate": 3.5837976566002044e-05,
|
|
"loss": 0.4164,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.4260328710079193,
|
|
"step": 1635
|
|
},
|
|
{
|
|
"epoch": 1.654320987654321,
|
|
"grad_norm": 0.49165743589401245,
|
|
"learning_rate": 3.5800307611062446e-05,
|
|
"loss": 0.397,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.35720714926719666,
|
|
"step": 1640
|
|
},
|
|
{
|
|
"epoch": 1.6604938271604939,
|
|
"grad_norm": 0.4694286584854126,
|
|
"learning_rate": 3.576248894508653e-05,
|
|
"loss": 0.3717,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3810012638568878,
|
|
"step": 1645
|
|
},
|
|
{
|
|
"epoch": 1.6666666666666665,
|
|
"grad_norm": 0.39720389246940613,
|
|
"learning_rate": 3.57245209264136e-05,
|
|
"loss": 0.3899,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3790915012359619,
|
|
"step": 1650
|
|
},
|
|
{
|
|
"epoch": 1.6728395061728394,
|
|
"grad_norm": 0.3794459402561188,
|
|
"learning_rate": 3.568640391479815e-05,
|
|
"loss": 0.3859,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3527582585811615,
|
|
"step": 1655
|
|
},
|
|
{
|
|
"epoch": 1.6790123456790123,
|
|
"grad_norm": 0.3121911883354187,
|
|
"learning_rate": 3.56481382714064e-05,
|
|
"loss": 0.34,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.33960476517677307,
|
|
"step": 1660
|
|
},
|
|
{
|
|
"epoch": 1.6851851851851851,
|
|
"grad_norm": 0.3293343484401703,
|
|
"learning_rate": 3.560972435881287e-05,
|
|
"loss": 0.3486,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3094177842140198,
|
|
"step": 1665
|
|
},
|
|
{
|
|
"epoch": 1.691358024691358,
|
|
"grad_norm": 0.304034560918808,
|
|
"learning_rate": 3.557116254099698e-05,
|
|
"loss": 0.3482,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.34977054595947266,
|
|
"step": 1670
|
|
},
|
|
{
|
|
"epoch": 1.6975308641975309,
|
|
"grad_norm": 0.32402095198631287,
|
|
"learning_rate": 3.5532453183339574e-05,
|
|
"loss": 0.3509,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3336471617221832,
|
|
"step": 1675
|
|
},
|
|
{
|
|
"epoch": 1.7037037037037037,
|
|
"grad_norm": 0.29803916811943054,
|
|
"learning_rate": 3.5493596652619456e-05,
|
|
"loss": 0.321,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3121752440929413,
|
|
"step": 1680
|
|
},
|
|
{
|
|
"epoch": 1.7098765432098766,
|
|
"grad_norm": 0.30628249049186707,
|
|
"learning_rate": 3.545459331700993e-05,
|
|
"loss": 0.3349,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.35699906945228577,
|
|
"step": 1685
|
|
},
|
|
{
|
|
"epoch": 1.7160493827160495,
|
|
"grad_norm": 0.3236028254032135,
|
|
"learning_rate": 3.541544354607532e-05,
|
|
"loss": 0.3517,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.30692362785339355,
|
|
"step": 1690
|
|
},
|
|
{
|
|
"epoch": 1.7222222222222223,
|
|
"grad_norm": 0.3192197382450104,
|
|
"learning_rate": 3.537614771076743e-05,
|
|
"loss": 0.3495,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3001939356327057,
|
|
"step": 1695
|
|
},
|
|
{
|
|
"epoch": 1.7283950617283952,
|
|
"grad_norm": 0.3548920154571533,
|
|
"learning_rate": 3.533670618342206e-05,
|
|
"loss": 0.3249,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2963644862174988,
|
|
"step": 1700
|
|
},
|
|
{
|
|
"epoch": 1.734567901234568,
|
|
"grad_norm": 0.3354020118713379,
|
|
"learning_rate": 3.529711933775549e-05,
|
|
"loss": 0.3324,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.35161665081977844,
|
|
"step": 1705
|
|
},
|
|
{
|
|
"epoch": 1.7407407407407407,
|
|
"grad_norm": 0.32942622900009155,
|
|
"learning_rate": 3.525738754886089e-05,
|
|
"loss": 0.3354,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3323695659637451,
|
|
"step": 1710
|
|
},
|
|
{
|
|
"epoch": 1.7469135802469136,
|
|
"grad_norm": 0.36671966314315796,
|
|
"learning_rate": 3.521751119320481e-05,
|
|
"loss": 0.3587,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.31985172629356384,
|
|
"step": 1715
|
|
},
|
|
{
|
|
"epoch": 1.7530864197530864,
|
|
"grad_norm": 0.3378705680370331,
|
|
"learning_rate": 3.51774906486236e-05,
|
|
"loss": 0.3111,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.29154518246650696,
|
|
"step": 1720
|
|
},
|
|
{
|
|
"epoch": 1.7592592592592593,
|
|
"grad_norm": 0.3443194329738617,
|
|
"learning_rate": 3.513732629431984e-05,
|
|
"loss": 0.3242,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3552367389202118,
|
|
"step": 1725
|
|
},
|
|
{
|
|
"epoch": 1.765432098765432,
|
|
"grad_norm": 0.3370530903339386,
|
|
"learning_rate": 3.5097018510858705e-05,
|
|
"loss": 0.3278,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3442824184894562,
|
|
"step": 1730
|
|
},
|
|
{
|
|
"epoch": 1.7716049382716048,
|
|
"grad_norm": 0.3659610450267792,
|
|
"learning_rate": 3.505656768016441e-05,
|
|
"loss": 0.3576,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3976585865020752,
|
|
"step": 1735
|
|
},
|
|
{
|
|
"epoch": 1.7777777777777777,
|
|
"grad_norm": 0.3486379086971283,
|
|
"learning_rate": 3.501597418551658e-05,
|
|
"loss": 0.3415,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3599509000778198,
|
|
"step": 1740
|
|
},
|
|
{
|
|
"epoch": 1.7839506172839505,
|
|
"grad_norm": 0.32802435755729675,
|
|
"learning_rate": 3.4975238411546584e-05,
|
|
"loss": 0.3466,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3233141601085663,
|
|
"step": 1745
|
|
},
|
|
{
|
|
"epoch": 1.7901234567901234,
|
|
"grad_norm": 0.3063686788082123,
|
|
"learning_rate": 3.493436074423393e-05,
|
|
"loss": 0.3389,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.34477224946022034,
|
|
"step": 1750
|
|
},
|
|
{
|
|
"epoch": 1.7962962962962963,
|
|
"grad_norm": 0.3337664306163788,
|
|
"learning_rate": 3.4893341570902604e-05,
|
|
"loss": 0.3418,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.39182186126708984,
|
|
"step": 1755
|
|
},
|
|
{
|
|
"epoch": 1.8024691358024691,
|
|
"grad_norm": 0.3196902871131897,
|
|
"learning_rate": 3.4852181280217364e-05,
|
|
"loss": 0.3476,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3803694546222687,
|
|
"step": 1760
|
|
},
|
|
{
|
|
"epoch": 1.808641975308642,
|
|
"grad_norm": 0.30811813473701477,
|
|
"learning_rate": 3.48108802621801e-05,
|
|
"loss": 0.3369,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3036186397075653,
|
|
"step": 1765
|
|
},
|
|
{
|
|
"epoch": 1.8148148148148149,
|
|
"grad_norm": 0.29382264614105225,
|
|
"learning_rate": 3.476943890812611e-05,
|
|
"loss": 0.3336,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.32850053906440735,
|
|
"step": 1770
|
|
},
|
|
{
|
|
"epoch": 1.8209876543209877,
|
|
"grad_norm": 0.2990680932998657,
|
|
"learning_rate": 3.472785761072043e-05,
|
|
"loss": 0.3215,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.29337310791015625,
|
|
"step": 1775
|
|
},
|
|
{
|
|
"epoch": 1.8271604938271606,
|
|
"grad_norm": 0.3405926823616028,
|
|
"learning_rate": 3.4686136763954063e-05,
|
|
"loss": 0.3326,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.29344442486763,
|
|
"step": 1780
|
|
},
|
|
{
|
|
"epoch": 1.8333333333333335,
|
|
"grad_norm": 0.29642006754875183,
|
|
"learning_rate": 3.464427676314026e-05,
|
|
"loss": 0.3277,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3206281065940857,
|
|
"step": 1785
|
|
},
|
|
{
|
|
"epoch": 1.8395061728395061,
|
|
"grad_norm": 0.2901068925857544,
|
|
"learning_rate": 3.460227800491082e-05,
|
|
"loss": 0.3305,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.34993812441825867,
|
|
"step": 1790
|
|
},
|
|
{
|
|
"epoch": 1.845679012345679,
|
|
"grad_norm": 0.3740207254886627,
|
|
"learning_rate": 3.456014088721226e-05,
|
|
"loss": 0.3403,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.39436253905296326,
|
|
"step": 1795
|
|
},
|
|
{
|
|
"epoch": 1.8518518518518519,
|
|
"grad_norm": 0.2903520166873932,
|
|
"learning_rate": 3.4517865809302096e-05,
|
|
"loss": 0.3191,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.280431866645813,
|
|
"step": 1800
|
|
},
|
|
{
|
|
"epoch": 1.8580246913580247,
|
|
"grad_norm": 0.30218684673309326,
|
|
"learning_rate": 3.4475453171745055e-05,
|
|
"loss": 0.337,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3222458064556122,
|
|
"step": 1805
|
|
},
|
|
{
|
|
"epoch": 1.8641975308641974,
|
|
"grad_norm": 0.3055573105812073,
|
|
"learning_rate": 3.4432903376409246e-05,
|
|
"loss": 0.329,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.34247317910194397,
|
|
"step": 1810
|
|
},
|
|
{
|
|
"epoch": 1.8703703703703702,
|
|
"grad_norm": 0.27665233612060547,
|
|
"learning_rate": 3.439021682646239e-05,
|
|
"loss": 0.3354,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.30159541964530945,
|
|
"step": 1815
|
|
},
|
|
{
|
|
"epoch": 1.876543209876543,
|
|
"grad_norm": 0.7904077172279358,
|
|
"learning_rate": 3.434739392636799e-05,
|
|
"loss": 0.308,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3136962354183197,
|
|
"step": 1820
|
|
},
|
|
{
|
|
"epoch": 1.882716049382716,
|
|
"grad_norm": 0.45841825008392334,
|
|
"learning_rate": 3.430443508188146e-05,
|
|
"loss": 0.2599,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23298458755016327,
|
|
"step": 1825
|
|
},
|
|
{
|
|
"epoch": 1.8888888888888888,
|
|
"grad_norm": 0.42286810278892517,
|
|
"learning_rate": 3.426134070004637e-05,
|
|
"loss": 0.2501,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2525423467159271,
|
|
"step": 1830
|
|
},
|
|
{
|
|
"epoch": 1.8950617283950617,
|
|
"grad_norm": 0.39577388763427734,
|
|
"learning_rate": 3.421811118919049e-05,
|
|
"loss": 0.2771,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.267189621925354,
|
|
"step": 1835
|
|
},
|
|
{
|
|
"epoch": 1.9012345679012346,
|
|
"grad_norm": 0.3696611821651459,
|
|
"learning_rate": 3.417474695892197e-05,
|
|
"loss": 0.2826,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26189497113227844,
|
|
"step": 1840
|
|
},
|
|
{
|
|
"epoch": 1.9074074074074074,
|
|
"grad_norm": 0.34460800886154175,
|
|
"learning_rate": 3.413124842012547e-05,
|
|
"loss": 0.264,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23150493204593658,
|
|
"step": 1845
|
|
},
|
|
{
|
|
"epoch": 1.9135802469135803,
|
|
"grad_norm": 0.3486117720603943,
|
|
"learning_rate": 3.4087615984958225e-05,
|
|
"loss": 0.2634,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.28175386786460876,
|
|
"step": 1850
|
|
},
|
|
{
|
|
"epoch": 1.9197530864197532,
|
|
"grad_norm": 0.36453524231910706,
|
|
"learning_rate": 3.404385006684619e-05,
|
|
"loss": 0.2622,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25572773814201355,
|
|
"step": 1855
|
|
},
|
|
{
|
|
"epoch": 1.925925925925926,
|
|
"grad_norm": 0.36858677864074707,
|
|
"learning_rate": 3.399995108048009e-05,
|
|
"loss": 0.2576,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24438567459583282,
|
|
"step": 1860
|
|
},
|
|
{
|
|
"epoch": 1.932098765432099,
|
|
"grad_norm": 0.3670538067817688,
|
|
"learning_rate": 3.395591944181148e-05,
|
|
"loss": 0.2743,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2811642289161682,
|
|
"step": 1865
|
|
},
|
|
{
|
|
"epoch": 1.9382716049382716,
|
|
"grad_norm": 0.3397423326969147,
|
|
"learning_rate": 3.3911755568048836e-05,
|
|
"loss": 0.2697,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2788163721561432,
|
|
"step": 1870
|
|
},
|
|
{
|
|
"epoch": 1.9444444444444444,
|
|
"grad_norm": 0.39673367142677307,
|
|
"learning_rate": 3.386745987765361e-05,
|
|
"loss": 0.293,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.31592801213264465,
|
|
"step": 1875
|
|
},
|
|
{
|
|
"epoch": 1.9506172839506173,
|
|
"grad_norm": 0.32189682126045227,
|
|
"learning_rate": 3.38230327903362e-05,
|
|
"loss": 0.2921,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2646641433238983,
|
|
"step": 1880
|
|
},
|
|
{
|
|
"epoch": 1.9567901234567902,
|
|
"grad_norm": 0.3050699532032013,
|
|
"learning_rate": 3.3778474727052036e-05,
|
|
"loss": 0.2892,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3114836513996124,
|
|
"step": 1885
|
|
},
|
|
{
|
|
"epoch": 1.9629629629629628,
|
|
"grad_norm": 0.3075443506240845,
|
|
"learning_rate": 3.373378610999758e-05,
|
|
"loss": 0.2786,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26305726170539856,
|
|
"step": 1890
|
|
},
|
|
{
|
|
"epoch": 1.9691358024691357,
|
|
"grad_norm": 0.4711006283760071,
|
|
"learning_rate": 3.3688967362606306e-05,
|
|
"loss": 0.2781,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2802620828151703,
|
|
"step": 1895
|
|
},
|
|
{
|
|
"epoch": 1.9753086419753085,
|
|
"grad_norm": 0.34563252329826355,
|
|
"learning_rate": 3.36440189095447e-05,
|
|
"loss": 0.2564,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2311277836561203,
|
|
"step": 1900
|
|
},
|
|
{
|
|
"epoch": 1.9814814814814814,
|
|
"grad_norm": 0.32549095153808594,
|
|
"learning_rate": 3.3598941176708223e-05,
|
|
"loss": 0.2448,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23143844306468964,
|
|
"step": 1905
|
|
},
|
|
{
|
|
"epoch": 1.9876543209876543,
|
|
"grad_norm": 0.34185269474983215,
|
|
"learning_rate": 3.3553734591217325e-05,
|
|
"loss": 0.2458,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2334810495376587,
|
|
"step": 1910
|
|
},
|
|
{
|
|
"epoch": 1.9938271604938271,
|
|
"grad_norm": 0.7859939932823181,
|
|
"learning_rate": 3.3508399581413305e-05,
|
|
"loss": 0.2481,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2628563344478607,
|
|
"step": 1915
|
|
},
|
|
{
|
|
"epoch": 2.0012345679012347,
|
|
"grad_norm": 0.6735068559646606,
|
|
"learning_rate": 3.3462936576854367e-05,
|
|
"loss": 0.2666,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2805740535259247,
|
|
"step": 1920
|
|
},
|
|
{
|
|
"epoch": 2.0074074074074075,
|
|
"grad_norm": 0.44390788674354553,
|
|
"learning_rate": 3.341734600831144e-05,
|
|
"loss": 0.2322,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22681696712970734,
|
|
"step": 1925
|
|
},
|
|
{
|
|
"epoch": 2.0135802469135804,
|
|
"grad_norm": 0.39590102434158325,
|
|
"learning_rate": 3.337162830776418e-05,
|
|
"loss": 0.2241,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20500679314136505,
|
|
"step": 1930
|
|
},
|
|
{
|
|
"epoch": 2.0197530864197533,
|
|
"grad_norm": 0.4411519169807434,
|
|
"learning_rate": 3.3325783908396824e-05,
|
|
"loss": 0.2184,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2164904624223709,
|
|
"step": 1935
|
|
},
|
|
{
|
|
"epoch": 2.025925925925926,
|
|
"grad_norm": 0.33601024746894836,
|
|
"learning_rate": 3.327981324459412e-05,
|
|
"loss": 0.217,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22954557836055756,
|
|
"step": 1940
|
|
},
|
|
{
|
|
"epoch": 2.0320987654320986,
|
|
"grad_norm": 0.30726346373558044,
|
|
"learning_rate": 3.323371675193719e-05,
|
|
"loss": 0.2203,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24350833892822266,
|
|
"step": 1945
|
|
},
|
|
{
|
|
"epoch": 2.0382716049382714,
|
|
"grad_norm": 0.33996787667274475,
|
|
"learning_rate": 3.318749486719941e-05,
|
|
"loss": 0.2125,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23216162621974945,
|
|
"step": 1950
|
|
},
|
|
{
|
|
"epoch": 2.0444444444444443,
|
|
"grad_norm": 0.3762907087802887,
|
|
"learning_rate": 3.314114802834226e-05,
|
|
"loss": 0.2223,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24356456100940704,
|
|
"step": 1955
|
|
},
|
|
{
|
|
"epoch": 2.050617283950617,
|
|
"grad_norm": 0.34768059849739075,
|
|
"learning_rate": 3.309467667451121e-05,
|
|
"loss": 0.2187,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18856267631053925,
|
|
"step": 1960
|
|
},
|
|
{
|
|
"epoch": 2.05679012345679,
|
|
"grad_norm": 0.4080201983451843,
|
|
"learning_rate": 3.3048081246031515e-05,
|
|
"loss": 0.215,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21309955418109894,
|
|
"step": 1965
|
|
},
|
|
{
|
|
"epoch": 2.062962962962963,
|
|
"grad_norm": 0.3447275757789612,
|
|
"learning_rate": 3.3001362184404055e-05,
|
|
"loss": 0.2047,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2296406477689743,
|
|
"step": 1970
|
|
},
|
|
{
|
|
"epoch": 2.0691358024691358,
|
|
"grad_norm": 0.44411346316337585,
|
|
"learning_rate": 3.295451993230118e-05,
|
|
"loss": 0.2315,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2330566793680191,
|
|
"step": 1975
|
|
},
|
|
{
|
|
"epoch": 2.0753086419753086,
|
|
"grad_norm": 0.396358847618103,
|
|
"learning_rate": 3.290755493356248e-05,
|
|
"loss": 0.21,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21171362698078156,
|
|
"step": 1980
|
|
},
|
|
{
|
|
"epoch": 2.0814814814814815,
|
|
"grad_norm": 0.42596232891082764,
|
|
"learning_rate": 3.286046763319061e-05,
|
|
"loss": 0.2274,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2503734529018402,
|
|
"step": 1985
|
|
},
|
|
{
|
|
"epoch": 2.0876543209876544,
|
|
"grad_norm": 0.38748595118522644,
|
|
"learning_rate": 3.2813258477347015e-05,
|
|
"loss": 0.2304,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25528407096862793,
|
|
"step": 1990
|
|
},
|
|
{
|
|
"epoch": 2.093827160493827,
|
|
"grad_norm": 0.3560836911201477,
|
|
"learning_rate": 3.276592791334779e-05,
|
|
"loss": 0.2211,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20532624423503876,
|
|
"step": 1995
|
|
},
|
|
{
|
|
"epoch": 2.1,
|
|
"grad_norm": 0.4228593409061432,
|
|
"learning_rate": 3.271847638965939e-05,
|
|
"loss": 0.2391,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2177974432706833,
|
|
"step": 2000
|
|
},
|
|
{
|
|
"epoch": 2.106172839506173,
|
|
"grad_norm": 0.41833698749542236,
|
|
"learning_rate": 3.267090435589436e-05,
|
|
"loss": 0.2147,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24089331924915314,
|
|
"step": 2005
|
|
},
|
|
{
|
|
"epoch": 2.112345679012346,
|
|
"grad_norm": 0.38651004433631897,
|
|
"learning_rate": 3.262321226280713e-05,
|
|
"loss": 0.2145,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20465557277202606,
|
|
"step": 2010
|
|
},
|
|
{
|
|
"epoch": 2.1185185185185187,
|
|
"grad_norm": 0.38954806327819824,
|
|
"learning_rate": 3.2575400562289717e-05,
|
|
"loss": 0.2205,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21425573527812958,
|
|
"step": 2015
|
|
},
|
|
{
|
|
"epoch": 2.1246913580246916,
|
|
"grad_norm": 0.3758370876312256,
|
|
"learning_rate": 3.25274697073674e-05,
|
|
"loss": 0.2098,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21829859912395477,
|
|
"step": 2020
|
|
},
|
|
{
|
|
"epoch": 2.1308641975308644,
|
|
"grad_norm": 0.4166395664215088,
|
|
"learning_rate": 3.247942015219451e-05,
|
|
"loss": 0.2154,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20405042171478271,
|
|
"step": 2025
|
|
},
|
|
{
|
|
"epoch": 2.137037037037037,
|
|
"grad_norm": 0.3934132158756256,
|
|
"learning_rate": 3.2431252352050086e-05,
|
|
"loss": 0.2246,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20243589580059052,
|
|
"step": 2030
|
|
},
|
|
{
|
|
"epoch": 2.1432098765432097,
|
|
"grad_norm": 0.38902637362480164,
|
|
"learning_rate": 3.238296676333354e-05,
|
|
"loss": 0.231,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22295449674129486,
|
|
"step": 2035
|
|
},
|
|
{
|
|
"epoch": 2.1493827160493826,
|
|
"grad_norm": 0.3810558617115021,
|
|
"learning_rate": 3.2334563843560376e-05,
|
|
"loss": 0.2156,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2133527249097824,
|
|
"step": 2040
|
|
},
|
|
{
|
|
"epoch": 2.1555555555555554,
|
|
"grad_norm": 0.38075312972068787,
|
|
"learning_rate": 3.228604405135781e-05,
|
|
"loss": 0.2114,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22307901084423065,
|
|
"step": 2045
|
|
},
|
|
{
|
|
"epoch": 2.1617283950617283,
|
|
"grad_norm": 0.3676074743270874,
|
|
"learning_rate": 3.223740784646046e-05,
|
|
"loss": 0.2312,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20981061458587646,
|
|
"step": 2050
|
|
},
|
|
{
|
|
"epoch": 2.167901234567901,
|
|
"grad_norm": 0.43101391196250916,
|
|
"learning_rate": 3.218865568970598e-05,
|
|
"loss": 0.2211,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2079259306192398,
|
|
"step": 2055
|
|
},
|
|
{
|
|
"epoch": 2.174074074074074,
|
|
"grad_norm": 0.3559808135032654,
|
|
"learning_rate": 3.21397880430307e-05,
|
|
"loss": 0.2242,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2437329888343811,
|
|
"step": 2060
|
|
},
|
|
{
|
|
"epoch": 2.180246913580247,
|
|
"grad_norm": 0.4371197521686554,
|
|
"learning_rate": 3.209080536946521e-05,
|
|
"loss": 0.217,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23282241821289062,
|
|
"step": 2065
|
|
},
|
|
{
|
|
"epoch": 2.18641975308642,
|
|
"grad_norm": 0.4523509442806244,
|
|
"learning_rate": 3.204170813313002e-05,
|
|
"loss": 0.2069,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23093490302562714,
|
|
"step": 2070
|
|
},
|
|
{
|
|
"epoch": 2.1925925925925926,
|
|
"grad_norm": 0.39561063051223755,
|
|
"learning_rate": 3.1992496799231156e-05,
|
|
"loss": 0.2202,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2163538932800293,
|
|
"step": 2075
|
|
},
|
|
{
|
|
"epoch": 2.1987654320987655,
|
|
"grad_norm": 0.31616783142089844,
|
|
"learning_rate": 3.194317183405573e-05,
|
|
"loss": 0.2216,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1809944361448288,
|
|
"step": 2080
|
|
},
|
|
{
|
|
"epoch": 2.2049382716049384,
|
|
"grad_norm": 0.42556798458099365,
|
|
"learning_rate": 3.189373370496752e-05,
|
|
"loss": 0.2015,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1941738873720169,
|
|
"step": 2085
|
|
},
|
|
{
|
|
"epoch": 2.2111111111111112,
|
|
"grad_norm": 0.30711138248443604,
|
|
"learning_rate": 3.184418288040257e-05,
|
|
"loss": 0.2044,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2089487910270691,
|
|
"step": 2090
|
|
},
|
|
{
|
|
"epoch": 2.217283950617284,
|
|
"grad_norm": 0.3570137023925781,
|
|
"learning_rate": 3.179451982986473e-05,
|
|
"loss": 0.2177,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23155921697616577,
|
|
"step": 2095
|
|
},
|
|
{
|
|
"epoch": 2.223456790123457,
|
|
"grad_norm": 0.4005148410797119,
|
|
"learning_rate": 3.174474502392123e-05,
|
|
"loss": 0.2077,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23430675268173218,
|
|
"step": 2100
|
|
},
|
|
{
|
|
"epoch": 2.006172839506173,
|
|
"grad_norm": 0.3607727885246277,
|
|
"learning_rate": 3.169485893419817e-05,
|
|
"loss": 0.2122,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20992840826511383,
|
|
"step": 2105
|
|
},
|
|
{
|
|
"epoch": 2.0123456790123457,
|
|
"grad_norm": 0.38729315996170044,
|
|
"learning_rate": 3.164486203337613e-05,
|
|
"loss": 0.2083,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19224704802036285,
|
|
"step": 2110
|
|
},
|
|
{
|
|
"epoch": 2.0185185185185186,
|
|
"grad_norm": 0.44547566771507263,
|
|
"learning_rate": 3.159475479518561e-05,
|
|
"loss": 0.204,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20420944690704346,
|
|
"step": 2115
|
|
},
|
|
{
|
|
"epoch": 2.0246913580246915,
|
|
"grad_norm": 0.35126975178718567,
|
|
"learning_rate": 3.1544537694402595e-05,
|
|
"loss": 0.2035,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21621358394622803,
|
|
"step": 2120
|
|
},
|
|
{
|
|
"epoch": 2.0308641975308643,
|
|
"grad_norm": 0.2993403673171997,
|
|
"learning_rate": 3.149421120684403e-05,
|
|
"loss": 0.2075,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22933651506900787,
|
|
"step": 2125
|
|
},
|
|
{
|
|
"epoch": 2.037037037037037,
|
|
"grad_norm": 0.3308967053890228,
|
|
"learning_rate": 3.144377580936333e-05,
|
|
"loss": 0.2001,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21940743923187256,
|
|
"step": 2130
|
|
},
|
|
{
|
|
"epoch": 2.04320987654321,
|
|
"grad_norm": 0.3724876940250397,
|
|
"learning_rate": 3.139323197984586e-05,
|
|
"loss": 0.2097,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23085443675518036,
|
|
"step": 2135
|
|
},
|
|
{
|
|
"epoch": 2.049382716049383,
|
|
"grad_norm": 0.3397546112537384,
|
|
"learning_rate": 3.1342580197204373e-05,
|
|
"loss": 0.2067,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17720545828342438,
|
|
"step": 2140
|
|
},
|
|
{
|
|
"epoch": 2.0555555555555554,
|
|
"grad_norm": 0.4186132252216339,
|
|
"learning_rate": 3.129182094137453e-05,
|
|
"loss": 0.2038,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2018803358078003,
|
|
"step": 2145
|
|
},
|
|
{
|
|
"epoch": 2.0617283950617282,
|
|
"grad_norm": 0.33399397134780884,
|
|
"learning_rate": 3.124095469331032e-05,
|
|
"loss": 0.1942,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21815155446529388,
|
|
"step": 2150
|
|
},
|
|
{
|
|
"epoch": 2.067901234567901,
|
|
"grad_norm": 0.43456369638442993,
|
|
"learning_rate": 3.118998193497946e-05,
|
|
"loss": 0.22,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22143904864788055,
|
|
"step": 2155
|
|
},
|
|
{
|
|
"epoch": 2.074074074074074,
|
|
"grad_norm": 0.4060041606426239,
|
|
"learning_rate": 3.113890314935894e-05,
|
|
"loss": 0.1992,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1997821182012558,
|
|
"step": 2160
|
|
},
|
|
{
|
|
"epoch": 2.080246913580247,
|
|
"grad_norm": 0.4097694158554077,
|
|
"learning_rate": 3.1087718820430326e-05,
|
|
"loss": 0.2165,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23854310810565948,
|
|
"step": 2165
|
|
},
|
|
{
|
|
"epoch": 2.0864197530864197,
|
|
"grad_norm": 0.38230136036872864,
|
|
"learning_rate": 3.103642943317526e-05,
|
|
"loss": 0.2188,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2420661300420761,
|
|
"step": 2170
|
|
},
|
|
{
|
|
"epoch": 2.0925925925925926,
|
|
"grad_norm": 0.3648567199707031,
|
|
"learning_rate": 3.09850354735708e-05,
|
|
"loss": 0.2104,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19509001076221466,
|
|
"step": 2175
|
|
},
|
|
{
|
|
"epoch": 2.0987654320987654,
|
|
"grad_norm": 0.4327113926410675,
|
|
"learning_rate": 3.093353742858488e-05,
|
|
"loss": 0.2276,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20813505351543427,
|
|
"step": 2180
|
|
},
|
|
{
|
|
"epoch": 2.1049382716049383,
|
|
"grad_norm": 0.4123539626598358,
|
|
"learning_rate": 3.0881935786171654e-05,
|
|
"loss": 0.2045,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2278442531824112,
|
|
"step": 2185
|
|
},
|
|
{
|
|
"epoch": 2.111111111111111,
|
|
"grad_norm": 0.40057891607284546,
|
|
"learning_rate": 3.083023103526688e-05,
|
|
"loss": 0.2041,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19542978703975677,
|
|
"step": 2190
|
|
},
|
|
{
|
|
"epoch": 2.117283950617284,
|
|
"grad_norm": 0.4086635410785675,
|
|
"learning_rate": 3.077842366578328e-05,
|
|
"loss": 0.2103,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20441095530986786,
|
|
"step": 2195
|
|
},
|
|
{
|
|
"epoch": 2.123456790123457,
|
|
"grad_norm": 0.3568022847175598,
|
|
"learning_rate": 3.072651416860591e-05,
|
|
"loss": 0.2002,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2078150063753128,
|
|
"step": 2200
|
|
},
|
|
{
|
|
"epoch": 2.1296296296296298,
|
|
"grad_norm": 0.4187696874141693,
|
|
"learning_rate": 3.0674503035587534e-05,
|
|
"loss": 0.2051,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1958889216184616,
|
|
"step": 2205
|
|
},
|
|
{
|
|
"epoch": 2.1358024691358026,
|
|
"grad_norm": 0.4033662676811218,
|
|
"learning_rate": 3.0622390759543896e-05,
|
|
"loss": 0.2141,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1929050236940384,
|
|
"step": 2210
|
|
},
|
|
{
|
|
"epoch": 2.1419753086419755,
|
|
"grad_norm": 0.3811688721179962,
|
|
"learning_rate": 3.057017783424911e-05,
|
|
"loss": 0.2206,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21376721560955048,
|
|
"step": 2215
|
|
},
|
|
{
|
|
"epoch": 2.148148148148148,
|
|
"grad_norm": 0.38393881916999817,
|
|
"learning_rate": 3.0517864754430965e-05,
|
|
"loss": 0.2057,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20298965275287628,
|
|
"step": 2220
|
|
},
|
|
{
|
|
"epoch": 2.154320987654321,
|
|
"grad_norm": 0.38883209228515625,
|
|
"learning_rate": 3.0465452015766224e-05,
|
|
"loss": 0.2019,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21197910606861115,
|
|
"step": 2225
|
|
},
|
|
{
|
|
"epoch": 2.1604938271604937,
|
|
"grad_norm": 0.39532989263534546,
|
|
"learning_rate": 3.041294011487593e-05,
|
|
"loss": 0.221,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19919781386852264,
|
|
"step": 2230
|
|
},
|
|
{
|
|
"epoch": 2.1666666666666665,
|
|
"grad_norm": 0.4541357159614563,
|
|
"learning_rate": 3.0360329549320742e-05,
|
|
"loss": 0.2116,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19796842336654663,
|
|
"step": 2235
|
|
},
|
|
{
|
|
"epoch": 2.1728395061728394,
|
|
"grad_norm": 0.3657527267932892,
|
|
"learning_rate": 3.0307620817596146e-05,
|
|
"loss": 0.2147,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23231887817382812,
|
|
"step": 2240
|
|
},
|
|
{
|
|
"epoch": 2.1790123456790123,
|
|
"grad_norm": 0.4421303868293762,
|
|
"learning_rate": 3.0254814419127795e-05,
|
|
"loss": 0.2077,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22234685719013214,
|
|
"step": 2245
|
|
},
|
|
{
|
|
"epoch": 2.185185185185185,
|
|
"grad_norm": 0.3726975619792938,
|
|
"learning_rate": 3.0201910854266744e-05,
|
|
"loss": 0.1981,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22131067514419556,
|
|
"step": 2250
|
|
},
|
|
{
|
|
"epoch": 2.191358024691358,
|
|
"grad_norm": 0.394936203956604,
|
|
"learning_rate": 3.0148910624284725e-05,
|
|
"loss": 0.2108,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20754681527614594,
|
|
"step": 2255
|
|
},
|
|
{
|
|
"epoch": 2.197530864197531,
|
|
"grad_norm": 0.326371431350708,
|
|
"learning_rate": 3.009581423136939e-05,
|
|
"loss": 0.2119,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1721523553133011,
|
|
"step": 2260
|
|
},
|
|
{
|
|
"epoch": 2.2037037037037037,
|
|
"grad_norm": 0.4307393729686737,
|
|
"learning_rate": 3.0042622178619553e-05,
|
|
"loss": 0.1928,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18589818477630615,
|
|
"step": 2265
|
|
},
|
|
{
|
|
"epoch": 2.2098765432098766,
|
|
"grad_norm": 0.30964070558547974,
|
|
"learning_rate": 2.9989334970040418e-05,
|
|
"loss": 0.1961,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20087118446826935,
|
|
"step": 2270
|
|
},
|
|
{
|
|
"epoch": 2.2160493827160495,
|
|
"grad_norm": 0.3598882555961609,
|
|
"learning_rate": 2.9935953110538817e-05,
|
|
"loss": 0.2087,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22222568094730377,
|
|
"step": 2275
|
|
},
|
|
{
|
|
"epoch": 2.2222222222222223,
|
|
"grad_norm": 0.40675249695777893,
|
|
"learning_rate": 2.988247710591841e-05,
|
|
"loss": 0.1992,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22418566048145294,
|
|
"step": 2280
|
|
},
|
|
{
|
|
"epoch": 2.228395061728395,
|
|
"grad_norm": 0.4242406189441681,
|
|
"learning_rate": 2.9828907462874917e-05,
|
|
"loss": 0.2117,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2285754680633545,
|
|
"step": 2285
|
|
},
|
|
{
|
|
"epoch": 2.234567901234568,
|
|
"grad_norm": 0.4023061692714691,
|
|
"learning_rate": 2.9775244688991286e-05,
|
|
"loss": 0.2204,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23574912548065186,
|
|
"step": 2290
|
|
},
|
|
{
|
|
"epoch": 2.240740740740741,
|
|
"grad_norm": 0.4325231909751892,
|
|
"learning_rate": 2.9721489292732903e-05,
|
|
"loss": 0.2136,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2125491052865982,
|
|
"step": 2295
|
|
},
|
|
{
|
|
"epoch": 2.246913580246914,
|
|
"grad_norm": 0.4078216552734375,
|
|
"learning_rate": 2.966764178344276e-05,
|
|
"loss": 0.2143,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23523962497711182,
|
|
"step": 2300
|
|
},
|
|
{
|
|
"epoch": 2.253086419753086,
|
|
"grad_norm": 0.42349499464035034,
|
|
"learning_rate": 2.9613702671336657e-05,
|
|
"loss": 0.2223,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19580888748168945,
|
|
"step": 2305
|
|
},
|
|
{
|
|
"epoch": 2.259259259259259,
|
|
"grad_norm": 0.4151558578014374,
|
|
"learning_rate": 2.9559672467498325e-05,
|
|
"loss": 0.2102,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20334136486053467,
|
|
"step": 2310
|
|
},
|
|
{
|
|
"epoch": 2.265432098765432,
|
|
"grad_norm": 0.6956079602241516,
|
|
"learning_rate": 2.950555168387464e-05,
|
|
"loss": 0.201,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21777981519699097,
|
|
"step": 2315
|
|
},
|
|
{
|
|
"epoch": 2.271604938271605,
|
|
"grad_norm": 0.4452267289161682,
|
|
"learning_rate": 2.9451340833270698e-05,
|
|
"loss": 0.2114,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.226753830909729,
|
|
"step": 2320
|
|
},
|
|
{
|
|
"epoch": 2.2777777777777777,
|
|
"grad_norm": 0.4485831558704376,
|
|
"learning_rate": 2.9397040429345032e-05,
|
|
"loss": 0.2073,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18167626857757568,
|
|
"step": 2325
|
|
},
|
|
{
|
|
"epoch": 2.2839506172839505,
|
|
"grad_norm": 0.37058308720588684,
|
|
"learning_rate": 2.9342650986604698e-05,
|
|
"loss": 0.2027,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2020191103219986,
|
|
"step": 2330
|
|
},
|
|
{
|
|
"epoch": 2.2901234567901234,
|
|
"grad_norm": 0.4976404905319214,
|
|
"learning_rate": 2.9288173020400404e-05,
|
|
"loss": 0.2189,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22280748188495636,
|
|
"step": 2335
|
|
},
|
|
{
|
|
"epoch": 2.2962962962962963,
|
|
"grad_norm": 0.4677009582519531,
|
|
"learning_rate": 2.923360704692165e-05,
|
|
"loss": 0.2108,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21487624943256378,
|
|
"step": 2340
|
|
},
|
|
{
|
|
"epoch": 2.302469135802469,
|
|
"grad_norm": 0.4063774645328522,
|
|
"learning_rate": 2.9178953583191814e-05,
|
|
"loss": 0.2111,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20316989719867706,
|
|
"step": 2345
|
|
},
|
|
{
|
|
"epoch": 2.308641975308642,
|
|
"grad_norm": 0.3789917528629303,
|
|
"learning_rate": 2.9124213147063263e-05,
|
|
"loss": 0.1851,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20726220309734344,
|
|
"step": 2350
|
|
},
|
|
{
|
|
"epoch": 2.314814814814815,
|
|
"grad_norm": 0.4241116940975189,
|
|
"learning_rate": 2.9069386257212442e-05,
|
|
"loss": 0.2018,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19845707714557648,
|
|
"step": 2355
|
|
},
|
|
{
|
|
"epoch": 2.3209876543209877,
|
|
"grad_norm": 0.44856444001197815,
|
|
"learning_rate": 2.901447343313497e-05,
|
|
"loss": 0.2232,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22118651866912842,
|
|
"step": 2360
|
|
},
|
|
{
|
|
"epoch": 2.3271604938271606,
|
|
"grad_norm": 0.5042104125022888,
|
|
"learning_rate": 2.8959475195140695e-05,
|
|
"loss": 0.1938,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.16407765448093414,
|
|
"step": 2365
|
|
},
|
|
{
|
|
"epoch": 2.3333333333333335,
|
|
"grad_norm": 0.4161403179168701,
|
|
"learning_rate": 2.890439206434879e-05,
|
|
"loss": 0.1972,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20118962228298187,
|
|
"step": 2370
|
|
},
|
|
{
|
|
"epoch": 2.3395061728395063,
|
|
"grad_norm": 0.3970435857772827,
|
|
"learning_rate": 2.8849224562682815e-05,
|
|
"loss": 0.2098,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2241845577955246,
|
|
"step": 2375
|
|
},
|
|
{
|
|
"epoch": 2.3456790123456788,
|
|
"grad_norm": 0.3987734615802765,
|
|
"learning_rate": 2.8793973212865734e-05,
|
|
"loss": 0.2292,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2033786177635193,
|
|
"step": 2380
|
|
},
|
|
{
|
|
"epoch": 2.351851851851852,
|
|
"grad_norm": 0.49959614872932434,
|
|
"learning_rate": 2.8738638538415013e-05,
|
|
"loss": 0.2026,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2035222202539444,
|
|
"step": 2385
|
|
},
|
|
{
|
|
"epoch": 2.3580246913580245,
|
|
"grad_norm": 0.4059554934501648,
|
|
"learning_rate": 2.8683221063637617e-05,
|
|
"loss": 0.2074,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21213527023792267,
|
|
"step": 2390
|
|
},
|
|
{
|
|
"epoch": 2.3641975308641974,
|
|
"grad_norm": 0.3537856340408325,
|
|
"learning_rate": 2.8627721313625073e-05,
|
|
"loss": 0.1959,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18329493701457977,
|
|
"step": 2395
|
|
},
|
|
{
|
|
"epoch": 2.3703703703703702,
|
|
"grad_norm": 0.3764045238494873,
|
|
"learning_rate": 2.8572139814248473e-05,
|
|
"loss": 0.2127,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21614117920398712,
|
|
"step": 2400
|
|
},
|
|
{
|
|
"epoch": 2.376543209876543,
|
|
"grad_norm": 0.5354501008987427,
|
|
"learning_rate": 2.8516477092153513e-05,
|
|
"loss": 0.2119,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18935976922512054,
|
|
"step": 2405
|
|
},
|
|
{
|
|
"epoch": 2.382716049382716,
|
|
"grad_norm": 0.37397077679634094,
|
|
"learning_rate": 2.846073367475547e-05,
|
|
"loss": 0.2231,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23228831589221954,
|
|
"step": 2410
|
|
},
|
|
{
|
|
"epoch": 2.388888888888889,
|
|
"grad_norm": 0.3511306643486023,
|
|
"learning_rate": 2.840491009023425e-05,
|
|
"loss": 0.218,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21727554500102997,
|
|
"step": 2415
|
|
},
|
|
{
|
|
"epoch": 2.3950617283950617,
|
|
"grad_norm": 0.3809970021247864,
|
|
"learning_rate": 2.8349006867529338e-05,
|
|
"loss": 0.1955,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1671956330537796,
|
|
"step": 2420
|
|
},
|
|
{
|
|
"epoch": 2.4012345679012346,
|
|
"grad_norm": 0.35577309131622314,
|
|
"learning_rate": 2.8293024536334815e-05,
|
|
"loss": 0.1958,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1941571682691574,
|
|
"step": 2425
|
|
},
|
|
{
|
|
"epoch": 2.4074074074074074,
|
|
"grad_norm": 0.36376234889030457,
|
|
"learning_rate": 2.8236963627094334e-05,
|
|
"loss": 0.2091,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19487722218036652,
|
|
"step": 2430
|
|
},
|
|
{
|
|
"epoch": 2.4135802469135803,
|
|
"grad_norm": 0.3604561388492584,
|
|
"learning_rate": 2.818082467099608e-05,
|
|
"loss": 0.203,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23189707100391388,
|
|
"step": 2435
|
|
},
|
|
{
|
|
"epoch": 2.419753086419753,
|
|
"grad_norm": 0.3463495671749115,
|
|
"learning_rate": 2.812460819996776e-05,
|
|
"loss": 0.2108,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22987087070941925,
|
|
"step": 2440
|
|
},
|
|
{
|
|
"epoch": 2.425925925925926,
|
|
"grad_norm": 0.43438920378685,
|
|
"learning_rate": 2.8068314746671536e-05,
|
|
"loss": 0.2019,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19801346957683563,
|
|
"step": 2445
|
|
},
|
|
{
|
|
"epoch": 2.432098765432099,
|
|
"grad_norm": 0.3817192614078522,
|
|
"learning_rate": 2.801194484449901e-05,
|
|
"loss": 0.2129,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2084825187921524,
|
|
"step": 2450
|
|
},
|
|
{
|
|
"epoch": 2.4382716049382718,
|
|
"grad_norm": 0.36205750703811646,
|
|
"learning_rate": 2.7955499027566132e-05,
|
|
"loss": 0.2024,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2033754587173462,
|
|
"step": 2455
|
|
},
|
|
{
|
|
"epoch": 2.4444444444444446,
|
|
"grad_norm": 0.41000211238861084,
|
|
"learning_rate": 2.7898977830708175e-05,
|
|
"loss": 0.2137,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2139686793088913,
|
|
"step": 2460
|
|
},
|
|
{
|
|
"epoch": 2.450617283950617,
|
|
"grad_norm": 0.37635156512260437,
|
|
"learning_rate": 2.7842381789474647e-05,
|
|
"loss": 0.1939,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2092421054840088,
|
|
"step": 2465
|
|
},
|
|
{
|
|
"epoch": 2.45679012345679,
|
|
"grad_norm": 0.34402555227279663,
|
|
"learning_rate": 2.7785711440124222e-05,
|
|
"loss": 0.2035,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19647137820720673,
|
|
"step": 2470
|
|
},
|
|
{
|
|
"epoch": 2.462962962962963,
|
|
"grad_norm": 0.3757551312446594,
|
|
"learning_rate": 2.7728967319619654e-05,
|
|
"loss": 0.201,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18279297649860382,
|
|
"step": 2475
|
|
},
|
|
{
|
|
"epoch": 2.4691358024691357,
|
|
"grad_norm": 0.49272048473358154,
|
|
"learning_rate": 2.767214996562269e-05,
|
|
"loss": 0.1881,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19929750263690948,
|
|
"step": 2480
|
|
},
|
|
{
|
|
"epoch": 2.4753086419753085,
|
|
"grad_norm": 0.39968743920326233,
|
|
"learning_rate": 2.7615259916488996e-05,
|
|
"loss": 0.1988,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18288390338420868,
|
|
"step": 2485
|
|
},
|
|
{
|
|
"epoch": 2.4814814814814814,
|
|
"grad_norm": 0.3557661175727844,
|
|
"learning_rate": 2.755829771126302e-05,
|
|
"loss": 0.2218,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2156534045934677,
|
|
"step": 2490
|
|
},
|
|
{
|
|
"epoch": 2.4876543209876543,
|
|
"grad_norm": 0.3966319262981415,
|
|
"learning_rate": 2.7501263889672914e-05,
|
|
"loss": 0.1983,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1900152713060379,
|
|
"step": 2495
|
|
},
|
|
{
|
|
"epoch": 2.493827160493827,
|
|
"grad_norm": 0.4476749897003174,
|
|
"learning_rate": 2.7444158992125404e-05,
|
|
"loss": 0.1971,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18784458935260773,
|
|
"step": 2500
|
|
},
|
|
{
|
|
"epoch": 2.5,
|
|
"grad_norm": 0.4738750457763672,
|
|
"learning_rate": 2.7386983559700673e-05,
|
|
"loss": 0.1916,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.16755028069019318,
|
|
"step": 2505
|
|
},
|
|
{
|
|
"epoch": 2.506172839506173,
|
|
"grad_norm": 0.3557210862636566,
|
|
"learning_rate": 2.732973813414725e-05,
|
|
"loss": 0.1821,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18717186152935028,
|
|
"step": 2510
|
|
},
|
|
{
|
|
"epoch": 2.5123456790123457,
|
|
"grad_norm": 0.4317970275878906,
|
|
"learning_rate": 2.7272423257876846e-05,
|
|
"loss": 0.189,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23637473583221436,
|
|
"step": 2515
|
|
},
|
|
{
|
|
"epoch": 2.5185185185185186,
|
|
"grad_norm": 0.4418342411518097,
|
|
"learning_rate": 2.7215039473959233e-05,
|
|
"loss": 0.2032,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2169712781906128,
|
|
"step": 2520
|
|
},
|
|
{
|
|
"epoch": 2.5246913580246915,
|
|
"grad_norm": 0.41437920928001404,
|
|
"learning_rate": 2.7157587326117107e-05,
|
|
"loss": 0.189,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1930430680513382,
|
|
"step": 2525
|
|
},
|
|
{
|
|
"epoch": 2.5308641975308643,
|
|
"grad_norm": 0.40373098850250244,
|
|
"learning_rate": 2.7100067358720924e-05,
|
|
"loss": 0.2043,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21307508647441864,
|
|
"step": 2530
|
|
},
|
|
{
|
|
"epoch": 2.537037037037037,
|
|
"grad_norm": 0.3736327886581421,
|
|
"learning_rate": 2.704248011678373e-05,
|
|
"loss": 0.2196,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21629512310028076,
|
|
"step": 2535
|
|
},
|
|
{
|
|
"epoch": 2.5432098765432096,
|
|
"grad_norm": 0.377960741519928,
|
|
"learning_rate": 2.6984826145956026e-05,
|
|
"loss": 0.2148,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22856013476848602,
|
|
"step": 2540
|
|
},
|
|
{
|
|
"epoch": 2.549382716049383,
|
|
"grad_norm": 0.4671732485294342,
|
|
"learning_rate": 2.6927105992520575e-05,
|
|
"loss": 0.2106,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1916073113679886,
|
|
"step": 2545
|
|
},
|
|
{
|
|
"epoch": 2.5555555555555554,
|
|
"grad_norm": 0.3707594871520996,
|
|
"learning_rate": 2.686932020338723e-05,
|
|
"loss": 0.2008,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21381425857543945,
|
|
"step": 2550
|
|
},
|
|
{
|
|
"epoch": 2.5617283950617287,
|
|
"grad_norm": 0.3798695206642151,
|
|
"learning_rate": 2.6811469326087767e-05,
|
|
"loss": 0.227,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20495067536830902,
|
|
"step": 2555
|
|
},
|
|
{
|
|
"epoch": 2.567901234567901,
|
|
"grad_norm": 0.3864648640155792,
|
|
"learning_rate": 2.6753553908770674e-05,
|
|
"loss": 0.2074,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1736653596162796,
|
|
"step": 2560
|
|
},
|
|
{
|
|
"epoch": 2.574074074074074,
|
|
"grad_norm": 0.3260176181793213,
|
|
"learning_rate": 2.669557450019598e-05,
|
|
"loss": 0.2028,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23535744845867157,
|
|
"step": 2565
|
|
},
|
|
{
|
|
"epoch": 2.580246913580247,
|
|
"grad_norm": 0.33569028973579407,
|
|
"learning_rate": 2.6637531649730024e-05,
|
|
"loss": 0.2169,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23311163485050201,
|
|
"step": 2570
|
|
},
|
|
{
|
|
"epoch": 2.5864197530864197,
|
|
"grad_norm": 0.41907575726509094,
|
|
"learning_rate": 2.6579425907340297e-05,
|
|
"loss": 0.2073,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19486576318740845,
|
|
"step": 2575
|
|
},
|
|
{
|
|
"epoch": 2.5925925925925926,
|
|
"grad_norm": 0.3688010573387146,
|
|
"learning_rate": 2.652125782359017e-05,
|
|
"loss": 0.1918,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19295088946819305,
|
|
"step": 2580
|
|
},
|
|
{
|
|
"epoch": 2.5987654320987654,
|
|
"grad_norm": 0.3946104943752289,
|
|
"learning_rate": 2.6463027949633744e-05,
|
|
"loss": 0.1952,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20642997324466705,
|
|
"step": 2585
|
|
},
|
|
{
|
|
"epoch": 2.6049382716049383,
|
|
"grad_norm": 0.3760945796966553,
|
|
"learning_rate": 2.6404736837210574e-05,
|
|
"loss": 0.1957,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20621974766254425,
|
|
"step": 2590
|
|
},
|
|
{
|
|
"epoch": 2.611111111111111,
|
|
"grad_norm": 0.3818272054195404,
|
|
"learning_rate": 2.6346385038640465e-05,
|
|
"loss": 0.2067,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22466368973255157,
|
|
"step": 2595
|
|
},
|
|
{
|
|
"epoch": 2.617283950617284,
|
|
"grad_norm": 0.4054039418697357,
|
|
"learning_rate": 2.628797310681824e-05,
|
|
"loss": 0.1932,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20212160050868988,
|
|
"step": 2600
|
|
},
|
|
{
|
|
"epoch": 2.623456790123457,
|
|
"grad_norm": 0.3872418999671936,
|
|
"learning_rate": 2.6229501595208496e-05,
|
|
"loss": 0.1808,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17415094375610352,
|
|
"step": 2605
|
|
},
|
|
{
|
|
"epoch": 2.6296296296296298,
|
|
"grad_norm": 0.3373595178127289,
|
|
"learning_rate": 2.6170971057840354e-05,
|
|
"loss": 0.19,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1764603704214096,
|
|
"step": 2610
|
|
},
|
|
{
|
|
"epoch": 2.6358024691358026,
|
|
"grad_norm": 0.34880051016807556,
|
|
"learning_rate": 2.6112382049302215e-05,
|
|
"loss": 0.208,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24108076095581055,
|
|
"step": 2615
|
|
},
|
|
{
|
|
"epoch": 2.6419753086419755,
|
|
"grad_norm": 0.7233880758285522,
|
|
"learning_rate": 2.6053735124736525e-05,
|
|
"loss": 0.2724,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3536733090877533,
|
|
"step": 2620
|
|
},
|
|
{
|
|
"epoch": 2.648148148148148,
|
|
"grad_norm": 0.46082618832588196,
|
|
"learning_rate": 2.599503083983446e-05,
|
|
"loss": 0.3951,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.4049645960330963,
|
|
"step": 2625
|
|
},
|
|
{
|
|
"epoch": 2.6543209876543212,
|
|
"grad_norm": 0.4174596667289734,
|
|
"learning_rate": 2.593626975083074e-05,
|
|
"loss": 0.3766,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.33869409561157227,
|
|
"step": 2630
|
|
},
|
|
{
|
|
"epoch": 2.6604938271604937,
|
|
"grad_norm": 0.4302406311035156,
|
|
"learning_rate": 2.5877452414498283e-05,
|
|
"loss": 0.3529,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.36262741684913635,
|
|
"step": 2635
|
|
},
|
|
{
|
|
"epoch": 2.6666666666666665,
|
|
"grad_norm": 0.341085821390152,
|
|
"learning_rate": 2.5818579388142965e-05,
|
|
"loss": 0.3713,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.36110779643058777,
|
|
"step": 2640
|
|
},
|
|
{
|
|
"epoch": 2.6728395061728394,
|
|
"grad_norm": 0.34470659494400024,
|
|
"learning_rate": 2.575965122959836e-05,
|
|
"loss": 0.3675,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.33468493819236755,
|
|
"step": 2645
|
|
},
|
|
{
|
|
"epoch": 2.6790123456790123,
|
|
"grad_norm": 0.29652518033981323,
|
|
"learning_rate": 2.570066849722039e-05,
|
|
"loss": 0.3227,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3224465548992157,
|
|
"step": 2650
|
|
},
|
|
{
|
|
"epoch": 2.685185185185185,
|
|
"grad_norm": 0.32567688822746277,
|
|
"learning_rate": 2.5641631749882114e-05,
|
|
"loss": 0.3315,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.29401543736457825,
|
|
"step": 2655
|
|
},
|
|
{
|
|
"epoch": 2.691358024691358,
|
|
"grad_norm": 0.33455511927604675,
|
|
"learning_rate": 2.5582541546968375e-05,
|
|
"loss": 0.3315,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.33299720287323,
|
|
"step": 2660
|
|
},
|
|
{
|
|
"epoch": 2.697530864197531,
|
|
"grad_norm": 0.3365607261657715,
|
|
"learning_rate": 2.552339844837052e-05,
|
|
"loss": 0.3344,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.31768491864204407,
|
|
"step": 2665
|
|
},
|
|
{
|
|
"epoch": 2.7037037037037037,
|
|
"grad_norm": 0.29815980792045593,
|
|
"learning_rate": 2.5464203014481083e-05,
|
|
"loss": 0.3057,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.29621538519859314,
|
|
"step": 2670
|
|
},
|
|
{
|
|
"epoch": 2.7098765432098766,
|
|
"grad_norm": 0.28737306594848633,
|
|
"learning_rate": 2.540495580618851e-05,
|
|
"loss": 0.3194,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.33922362327575684,
|
|
"step": 2675
|
|
},
|
|
{
|
|
"epoch": 2.7160493827160495,
|
|
"grad_norm": 0.31547418236732483,
|
|
"learning_rate": 2.5345657384871797e-05,
|
|
"loss": 0.3357,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.29283955693244934,
|
|
"step": 2680
|
|
},
|
|
{
|
|
"epoch": 2.7222222222222223,
|
|
"grad_norm": 0.2928524315357208,
|
|
"learning_rate": 2.5286308312395208e-05,
|
|
"loss": 0.334,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2868933379650116,
|
|
"step": 2685
|
|
},
|
|
{
|
|
"epoch": 2.728395061728395,
|
|
"grad_norm": 0.3741317689418793,
|
|
"learning_rate": 2.522690915110292e-05,
|
|
"loss": 0.31,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2824898958206177,
|
|
"step": 2690
|
|
},
|
|
{
|
|
"epoch": 2.734567901234568,
|
|
"grad_norm": 0.33861666917800903,
|
|
"learning_rate": 2.5167460463813727e-05,
|
|
"loss": 0.3161,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.33588042855262756,
|
|
"step": 2695
|
|
},
|
|
{
|
|
"epoch": 2.7407407407407405,
|
|
"grad_norm": 0.35369300842285156,
|
|
"learning_rate": 2.5107962813815693e-05,
|
|
"loss": 0.3204,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.31716111302375793,
|
|
"step": 2700
|
|
},
|
|
{
|
|
"epoch": 2.746913580246914,
|
|
"grad_norm": 0.32523658871650696,
|
|
"learning_rate": 2.50484167648608e-05,
|
|
"loss": 0.3437,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.30516350269317627,
|
|
"step": 2705
|
|
},
|
|
{
|
|
"epoch": 2.753086419753086,
|
|
"grad_norm": 0.3132164776325226,
|
|
"learning_rate": 2.4988822881159627e-05,
|
|
"loss": 0.2972,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2779521048069,
|
|
"step": 2710
|
|
},
|
|
{
|
|
"epoch": 2.7592592592592595,
|
|
"grad_norm": 0.33597251772880554,
|
|
"learning_rate": 2.4929181727376e-05,
|
|
"loss": 0.3103,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3409285843372345,
|
|
"step": 2715
|
|
},
|
|
{
|
|
"epoch": 2.765432098765432,
|
|
"grad_norm": 0.3103015124797821,
|
|
"learning_rate": 2.486949386862163e-05,
|
|
"loss": 0.3139,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.32940781116485596,
|
|
"step": 2720
|
|
},
|
|
{
|
|
"epoch": 2.771604938271605,
|
|
"grad_norm": 0.3409218192100525,
|
|
"learning_rate": 2.4809759870450774e-05,
|
|
"loss": 0.3432,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3823580741882324,
|
|
"step": 2725
|
|
},
|
|
{
|
|
"epoch": 2.7777777777777777,
|
|
"grad_norm": 0.3353785574436188,
|
|
"learning_rate": 2.4749980298854864e-05,
|
|
"loss": 0.3269,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3456442356109619,
|
|
"step": 2730
|
|
},
|
|
{
|
|
"epoch": 2.7839506172839505,
|
|
"grad_norm": 0.3391750752925873,
|
|
"learning_rate": 2.4690155720257156e-05,
|
|
"loss": 0.3324,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3096297085285187,
|
|
"step": 2735
|
|
},
|
|
{
|
|
"epoch": 2.7901234567901234,
|
|
"grad_norm": 0.3666570484638214,
|
|
"learning_rate": 2.4630286701507346e-05,
|
|
"loss": 0.3255,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.33089083433151245,
|
|
"step": 2740
|
|
},
|
|
{
|
|
"epoch": 2.7962962962962963,
|
|
"grad_norm": 0.3118732273578644,
|
|
"learning_rate": 2.4570373809876228e-05,
|
|
"loss": 0.3286,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.37715280055999756,
|
|
"step": 2745
|
|
},
|
|
{
|
|
"epoch": 2.802469135802469,
|
|
"grad_norm": 0.3451964855194092,
|
|
"learning_rate": 2.4510417613050268e-05,
|
|
"loss": 0.3339,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.36599865555763245,
|
|
"step": 2750
|
|
},
|
|
{
|
|
"epoch": 2.808641975308642,
|
|
"grad_norm": 0.33912158012390137,
|
|
"learning_rate": 2.445041867912629e-05,
|
|
"loss": 0.3235,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2912273705005646,
|
|
"step": 2755
|
|
},
|
|
{
|
|
"epoch": 2.814814814814815,
|
|
"grad_norm": 0.3124856650829315,
|
|
"learning_rate": 2.4390377576606047e-05,
|
|
"loss": 0.32,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.31540587544441223,
|
|
"step": 2760
|
|
},
|
|
{
|
|
"epoch": 2.8209876543209877,
|
|
"grad_norm": 0.489320307970047,
|
|
"learning_rate": 2.4330294874390838e-05,
|
|
"loss": 0.3087,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2815772593021393,
|
|
"step": 2765
|
|
},
|
|
{
|
|
"epoch": 2.8271604938271606,
|
|
"grad_norm": 0.34719735383987427,
|
|
"learning_rate": 2.427017114177615e-05,
|
|
"loss": 0.3195,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.28038203716278076,
|
|
"step": 2770
|
|
},
|
|
{
|
|
"epoch": 2.8333333333333335,
|
|
"grad_norm": 0.291423499584198,
|
|
"learning_rate": 2.421000694844622e-05,
|
|
"loss": 0.315,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3090975284576416,
|
|
"step": 2775
|
|
},
|
|
{
|
|
"epoch": 2.8395061728395063,
|
|
"grad_norm": 0.31119734048843384,
|
|
"learning_rate": 2.414980286446867e-05,
|
|
"loss": 0.3179,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.33642640709877014,
|
|
"step": 2780
|
|
},
|
|
{
|
|
"epoch": 2.8456790123456788,
|
|
"grad_norm": 0.3120616376399994,
|
|
"learning_rate": 2.408955946028909e-05,
|
|
"loss": 0.3279,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.38068172335624695,
|
|
"step": 2785
|
|
},
|
|
{
|
|
"epoch": 2.851851851851852,
|
|
"grad_norm": 0.293904185295105,
|
|
"learning_rate": 2.4029277306725637e-05,
|
|
"loss": 0.3071,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27002063393592834,
|
|
"step": 2790
|
|
},
|
|
{
|
|
"epoch": 2.8580246913580245,
|
|
"grad_norm": 0.2858833074569702,
|
|
"learning_rate": 2.396895697496362e-05,
|
|
"loss": 0.3248,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3099575638771057,
|
|
"step": 2795
|
|
},
|
|
{
|
|
"epoch": 2.8641975308641974,
|
|
"grad_norm": 0.285675585269928,
|
|
"learning_rate": 2.3908599036550113e-05,
|
|
"loss": 0.317,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3297642171382904,
|
|
"step": 2800
|
|
},
|
|
{
|
|
"epoch": 2.8703703703703702,
|
|
"grad_norm": 0.2714674770832062,
|
|
"learning_rate": 2.3848204063388494e-05,
|
|
"loss": 0.3235,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.29032301902770996,
|
|
"step": 2805
|
|
},
|
|
{
|
|
"epoch": 2.876543209876543,
|
|
"grad_norm": 0.7626910209655762,
|
|
"learning_rate": 2.3787772627733067e-05,
|
|
"loss": 0.2957,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.29959654808044434,
|
|
"step": 2810
|
|
},
|
|
{
|
|
"epoch": 2.882716049382716,
|
|
"grad_norm": 0.4776705801486969,
|
|
"learning_rate": 2.372730530218365e-05,
|
|
"loss": 0.2471,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22238855063915253,
|
|
"step": 2815
|
|
},
|
|
{
|
|
"epoch": 2.888888888888889,
|
|
"grad_norm": 0.4362982213497162,
|
|
"learning_rate": 2.366680265968006e-05,
|
|
"loss": 0.2376,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.239817813038826,
|
|
"step": 2820
|
|
},
|
|
{
|
|
"epoch": 2.8950617283950617,
|
|
"grad_norm": 0.41408979892730713,
|
|
"learning_rate": 2.3606265273496814e-05,
|
|
"loss": 0.263,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2526939809322357,
|
|
"step": 2825
|
|
},
|
|
{
|
|
"epoch": 2.9012345679012346,
|
|
"grad_norm": 0.4086337387561798,
|
|
"learning_rate": 2.3545693717237604e-05,
|
|
"loss": 0.2684,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24869686365127563,
|
|
"step": 2830
|
|
},
|
|
{
|
|
"epoch": 2.9074074074074074,
|
|
"grad_norm": 0.3266755938529968,
|
|
"learning_rate": 2.3485088564829884e-05,
|
|
"loss": 0.2508,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22022996842861176,
|
|
"step": 2835
|
|
},
|
|
{
|
|
"epoch": 2.9135802469135803,
|
|
"grad_norm": 0.3728151321411133,
|
|
"learning_rate": 2.3424450390519454e-05,
|
|
"loss": 0.2502,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2692379653453827,
|
|
"step": 2840
|
|
},
|
|
{
|
|
"epoch": 2.919753086419753,
|
|
"grad_norm": 0.35321909189224243,
|
|
"learning_rate": 2.3363779768865e-05,
|
|
"loss": 0.2492,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.242279052734375,
|
|
"step": 2845
|
|
},
|
|
{
|
|
"epoch": 2.925925925925926,
|
|
"grad_norm": 0.3239139914512634,
|
|
"learning_rate": 2.3303077274732634e-05,
|
|
"loss": 0.2456,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23347532749176025,
|
|
"step": 2850
|
|
},
|
|
{
|
|
"epoch": 2.932098765432099,
|
|
"grad_norm": 0.3423680365085602,
|
|
"learning_rate": 2.3242343483290495e-05,
|
|
"loss": 0.2615,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2678012549877167,
|
|
"step": 2855
|
|
},
|
|
{
|
|
"epoch": 2.9382716049382713,
|
|
"grad_norm": 0.3479361832141876,
|
|
"learning_rate": 2.318157897000325e-05,
|
|
"loss": 0.2562,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26475366950035095,
|
|
"step": 2860
|
|
},
|
|
{
|
|
"epoch": 2.9444444444444446,
|
|
"grad_norm": 0.38138219714164734,
|
|
"learning_rate": 2.312078431062667e-05,
|
|
"loss": 0.2794,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.30102643370628357,
|
|
"step": 2865
|
|
},
|
|
{
|
|
"epoch": 2.950617283950617,
|
|
"grad_norm": 0.3280933201313019,
|
|
"learning_rate": 2.3059960081202162e-05,
|
|
"loss": 0.2785,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25201937556266785,
|
|
"step": 2870
|
|
},
|
|
{
|
|
"epoch": 2.9567901234567904,
|
|
"grad_norm": 0.3382212519645691,
|
|
"learning_rate": 2.2999106858051326e-05,
|
|
"loss": 0.2762,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2974417805671692,
|
|
"step": 2875
|
|
},
|
|
{
|
|
"epoch": 2.962962962962963,
|
|
"grad_norm": 0.2888612449169159,
|
|
"learning_rate": 2.293822521777046e-05,
|
|
"loss": 0.2663,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2504029870033264,
|
|
"step": 2880
|
|
},
|
|
{
|
|
"epoch": 2.9691358024691357,
|
|
"grad_norm": 0.34633317589759827,
|
|
"learning_rate": 2.2877315737225144e-05,
|
|
"loss": 0.2659,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.269275963306427,
|
|
"step": 2885
|
|
},
|
|
{
|
|
"epoch": 2.9753086419753085,
|
|
"grad_norm": 0.3322531282901764,
|
|
"learning_rate": 2.2816378993544738e-05,
|
|
"loss": 0.2445,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2191019058227539,
|
|
"step": 2890
|
|
},
|
|
{
|
|
"epoch": 2.9814814814814814,
|
|
"grad_norm": 0.34253373742103577,
|
|
"learning_rate": 2.275541556411692e-05,
|
|
"loss": 0.2333,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22076310217380524,
|
|
"step": 2895
|
|
},
|
|
{
|
|
"epoch": 2.9876543209876543,
|
|
"grad_norm": 0.3431246280670166,
|
|
"learning_rate": 2.269442602658224e-05,
|
|
"loss": 0.2345,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22343742847442627,
|
|
"step": 2900
|
|
},
|
|
{
|
|
"epoch": 2.993827160493827,
|
|
"grad_norm": 0.3929581046104431,
|
|
"learning_rate": 2.2633410958828614e-05,
|
|
"loss": 0.2374,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25347599387168884,
|
|
"step": 2905
|
|
},
|
|
{
|
|
"epoch": 3.0,
|
|
"grad_norm": 0.6391134262084961,
|
|
"learning_rate": 2.257237093898586e-05,
|
|
"loss": 0.2499,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2522914707660675,
|
|
"step": 2910
|
|
},
|
|
{
|
|
"epoch": 3.006172839506173,
|
|
"grad_norm": 0.4407883584499359,
|
|
"learning_rate": 2.2511306545420235e-05,
|
|
"loss": 0.2119,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2090930938720703,
|
|
"step": 2915
|
|
},
|
|
{
|
|
"epoch": 3.0123456790123457,
|
|
"grad_norm": 0.37867650389671326,
|
|
"learning_rate": 2.245021835672891e-05,
|
|
"loss": 0.2057,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1891670972108841,
|
|
"step": 2920
|
|
},
|
|
{
|
|
"epoch": 3.0185185185185186,
|
|
"grad_norm": 0.44586044549942017,
|
|
"learning_rate": 2.2389106951734566e-05,
|
|
"loss": 0.2,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2007114738225937,
|
|
"step": 2925
|
|
},
|
|
{
|
|
"epoch": 3.0246913580246915,
|
|
"grad_norm": 0.3393300473690033,
|
|
"learning_rate": 2.2327972909479835e-05,
|
|
"loss": 0.1986,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21125547587871552,
|
|
"step": 2930
|
|
},
|
|
{
|
|
"epoch": 3.0308641975308643,
|
|
"grad_norm": 0.35225021839141846,
|
|
"learning_rate": 2.2266816809221832e-05,
|
|
"loss": 0.2025,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22303159534931183,
|
|
"step": 2935
|
|
},
|
|
{
|
|
"epoch": 3.037037037037037,
|
|
"grad_norm": 0.33816292881965637,
|
|
"learning_rate": 2.2205639230426704e-05,
|
|
"loss": 0.1948,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21372048556804657,
|
|
"step": 2940
|
|
},
|
|
{
|
|
"epoch": 3.04320987654321,
|
|
"grad_norm": 0.37574857473373413,
|
|
"learning_rate": 2.2144440752764096e-05,
|
|
"loss": 0.2038,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22463858127593994,
|
|
"step": 2945
|
|
},
|
|
{
|
|
"epoch": 3.049382716049383,
|
|
"grad_norm": 0.36280855536460876,
|
|
"learning_rate": 2.208322195610167e-05,
|
|
"loss": 0.2009,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1719566136598587,
|
|
"step": 2950
|
|
},
|
|
{
|
|
"epoch": 3.0555555555555554,
|
|
"grad_norm": 0.42250558733940125,
|
|
"learning_rate": 2.202198342049962e-05,
|
|
"loss": 0.1981,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19554030895233154,
|
|
"step": 2955
|
|
},
|
|
{
|
|
"epoch": 3.0617283950617282,
|
|
"grad_norm": 0.33952564001083374,
|
|
"learning_rate": 2.196072572620517e-05,
|
|
"loss": 0.1885,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21190857887268066,
|
|
"step": 2960
|
|
},
|
|
{
|
|
"epoch": 3.067901234567901,
|
|
"grad_norm": 0.45724132657051086,
|
|
"learning_rate": 2.189944945364708e-05,
|
|
"loss": 0.2138,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21449393033981323,
|
|
"step": 2965
|
|
},
|
|
{
|
|
"epoch": 3.074074074074074,
|
|
"grad_norm": 0.3910859525203705,
|
|
"learning_rate": 2.1838155183430132e-05,
|
|
"loss": 0.1934,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1937236785888672,
|
|
"step": 2970
|
|
},
|
|
{
|
|
"epoch": 3.080246913580247,
|
|
"grad_norm": 0.4079418480396271,
|
|
"learning_rate": 2.177684349632965e-05,
|
|
"loss": 0.2103,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23122982680797577,
|
|
"step": 2975
|
|
},
|
|
{
|
|
"epoch": 3.0864197530864197,
|
|
"grad_norm": 0.4655308127403259,
|
|
"learning_rate": 2.171551497328599e-05,
|
|
"loss": 0.2121,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23530392348766327,
|
|
"step": 2980
|
|
},
|
|
{
|
|
"epoch": 3.0925925925925926,
|
|
"grad_norm": 0.35715094208717346,
|
|
"learning_rate": 2.1654170195399017e-05,
|
|
"loss": 0.2044,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1895744651556015,
|
|
"step": 2985
|
|
},
|
|
{
|
|
"epoch": 3.0987654320987654,
|
|
"grad_norm": 0.422252893447876,
|
|
"learning_rate": 2.1592809743922616e-05,
|
|
"loss": 0.221,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2027362734079361,
|
|
"step": 2990
|
|
},
|
|
{
|
|
"epoch": 3.1049382716049383,
|
|
"grad_norm": 0.40926796197891235,
|
|
"learning_rate": 2.1531434200259204e-05,
|
|
"loss": 0.1986,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21998877823352814,
|
|
"step": 2995
|
|
},
|
|
{
|
|
"epoch": 3.111111111111111,
|
|
"grad_norm": 0.3874903619289398,
|
|
"learning_rate": 2.147004414595417e-05,
|
|
"loss": 0.198,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18975865840911865,
|
|
"step": 3000
|
|
},
|
|
{
|
|
"epoch": 3.117283950617284,
|
|
"grad_norm": 0.3846581280231476,
|
|
"learning_rate": 2.1408640162690414e-05,
|
|
"loss": 0.2042,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1987636685371399,
|
|
"step": 3005
|
|
},
|
|
{
|
|
"epoch": 3.123456790123457,
|
|
"grad_norm": 0.3576688766479492,
|
|
"learning_rate": 2.13472228322828e-05,
|
|
"loss": 0.1944,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20145122706890106,
|
|
"step": 3010
|
|
},
|
|
{
|
|
"epoch": 3.1296296296296298,
|
|
"grad_norm": 0.4109097421169281,
|
|
"learning_rate": 2.128579273667269e-05,
|
|
"loss": 0.1988,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1912132054567337,
|
|
"step": 3015
|
|
},
|
|
{
|
|
"epoch": 3.1358024691358026,
|
|
"grad_norm": 0.4017428457736969,
|
|
"learning_rate": 2.1224350457922345e-05,
|
|
"loss": 0.2074,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18709485232830048,
|
|
"step": 3020
|
|
},
|
|
{
|
|
"epoch": 3.1419753086419755,
|
|
"grad_norm": 0.38083553314208984,
|
|
"learning_rate": 2.1162896578209517e-05,
|
|
"loss": 0.214,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20727352797985077,
|
|
"step": 3025
|
|
},
|
|
{
|
|
"epoch": 3.148148148148148,
|
|
"grad_norm": 0.38105782866477966,
|
|
"learning_rate": 2.110143167982186e-05,
|
|
"loss": 0.1998,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19632582366466522,
|
|
"step": 3030
|
|
},
|
|
{
|
|
"epoch": 3.154320987654321,
|
|
"grad_norm": 0.3849983215332031,
|
|
"learning_rate": 2.1039956345151413e-05,
|
|
"loss": 0.1958,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2048538774251938,
|
|
"step": 3035
|
|
},
|
|
{
|
|
"epoch": 3.1604938271604937,
|
|
"grad_norm": 0.35731858015060425,
|
|
"learning_rate": 2.0978471156689135e-05,
|
|
"loss": 0.2143,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19307614862918854,
|
|
"step": 3040
|
|
},
|
|
{
|
|
"epoch": 3.1666666666666665,
|
|
"grad_norm": 0.4383801221847534,
|
|
"learning_rate": 2.091697669701933e-05,
|
|
"loss": 0.2055,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19220416247844696,
|
|
"step": 3045
|
|
},
|
|
{
|
|
"epoch": 3.1728395061728394,
|
|
"grad_norm": 0.35691037774086,
|
|
"learning_rate": 2.0855473548814148e-05,
|
|
"loss": 0.2082,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22418153285980225,
|
|
"step": 3050
|
|
},
|
|
{
|
|
"epoch": 3.1790123456790123,
|
|
"grad_norm": 0.4316836893558502,
|
|
"learning_rate": 2.0793962294828073e-05,
|
|
"loss": 0.2016,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21595996618270874,
|
|
"step": 3055
|
|
},
|
|
{
|
|
"epoch": 3.185185185185185,
|
|
"grad_norm": 0.3641813397407532,
|
|
"learning_rate": 2.0732443517892386e-05,
|
|
"loss": 0.1921,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21338899433612823,
|
|
"step": 3060
|
|
},
|
|
{
|
|
"epoch": 3.191358024691358,
|
|
"grad_norm": 0.39384904503822327,
|
|
"learning_rate": 2.0670917800909657e-05,
|
|
"loss": 0.204,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20154981315135956,
|
|
"step": 3065
|
|
},
|
|
{
|
|
"epoch": 3.197530864197531,
|
|
"grad_norm": 0.3160921633243561,
|
|
"learning_rate": 2.0609385726848213e-05,
|
|
"loss": 0.2054,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1659029871225357,
|
|
"step": 3070
|
|
},
|
|
{
|
|
"epoch": 3.2037037037037037,
|
|
"grad_norm": 0.42596593499183655,
|
|
"learning_rate": 2.0547847878736603e-05,
|
|
"loss": 0.1873,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1798311471939087,
|
|
"step": 3075
|
|
},
|
|
{
|
|
"epoch": 3.2098765432098766,
|
|
"grad_norm": 0.30479222536087036,
|
|
"learning_rate": 2.04863048396581e-05,
|
|
"loss": 0.1901,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19535990059375763,
|
|
"step": 3080
|
|
},
|
|
{
|
|
"epoch": 3.2160493827160495,
|
|
"grad_norm": 0.34282127022743225,
|
|
"learning_rate": 2.042475719274517e-05,
|
|
"loss": 0.2029,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21605175733566284,
|
|
"step": 3085
|
|
},
|
|
{
|
|
"epoch": 3.2222222222222223,
|
|
"grad_norm": 0.3848612308502197,
|
|
"learning_rate": 2.036320552117391e-05,
|
|
"loss": 0.1934,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21657271683216095,
|
|
"step": 3090
|
|
},
|
|
{
|
|
"epoch": 3.228395061728395,
|
|
"grad_norm": 0.43970414996147156,
|
|
"learning_rate": 2.0301650408158583e-05,
|
|
"loss": 0.2042,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22150655090808868,
|
|
"step": 3095
|
|
},
|
|
{
|
|
"epoch": 3.234567901234568,
|
|
"grad_norm": 0.398004949092865,
|
|
"learning_rate": 2.0240092436946036e-05,
|
|
"loss": 0.2131,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22805722057819366,
|
|
"step": 3100
|
|
},
|
|
{
|
|
"epoch": 3.240740740740741,
|
|
"grad_norm": 0.4378446638584137,
|
|
"learning_rate": 2.0178532190810215e-05,
|
|
"loss": 0.2065,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20567144453525543,
|
|
"step": 3105
|
|
},
|
|
{
|
|
"epoch": 3.246913580246914,
|
|
"grad_norm": 0.3924199640750885,
|
|
"learning_rate": 2.011697025304661e-05,
|
|
"loss": 0.2074,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2267635315656662,
|
|
"step": 3110
|
|
},
|
|
{
|
|
"epoch": 3.253086419753086,
|
|
"grad_norm": 0.38685929775238037,
|
|
"learning_rate": 2.0055407206966758e-05,
|
|
"loss": 0.2116,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18538208305835724,
|
|
"step": 3115
|
|
},
|
|
{
|
|
"epoch": 3.259259259259259,
|
|
"grad_norm": 0.40124863386154175,
|
|
"learning_rate": 1.9993843635892657e-05,
|
|
"loss": 0.1985,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.191811665892601,
|
|
"step": 3120
|
|
},
|
|
{
|
|
"epoch": 3.265432098765432,
|
|
"grad_norm": 0.36089351773262024,
|
|
"learning_rate": 1.9932280123151334e-05,
|
|
"loss": 0.1907,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20921160280704498,
|
|
"step": 3125
|
|
},
|
|
{
|
|
"epoch": 3.271604938271605,
|
|
"grad_norm": 0.399674654006958,
|
|
"learning_rate": 1.9870717252069215e-05,
|
|
"loss": 0.2001,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2155553549528122,
|
|
"step": 3130
|
|
},
|
|
{
|
|
"epoch": 3.2777777777777777,
|
|
"grad_norm": 0.4354398846626282,
|
|
"learning_rate": 1.980915560596668e-05,
|
|
"loss": 0.1961,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1713920384645462,
|
|
"step": 3135
|
|
},
|
|
{
|
|
"epoch": 3.2839506172839505,
|
|
"grad_norm": 0.36640727519989014,
|
|
"learning_rate": 1.9747595768152486e-05,
|
|
"loss": 0.1922,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19114716351032257,
|
|
"step": 3140
|
|
},
|
|
{
|
|
"epoch": 3.2901234567901234,
|
|
"grad_norm": 0.4540056884288788,
|
|
"learning_rate": 1.968603832191826e-05,
|
|
"loss": 0.2074,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.210588738322258,
|
|
"step": 3145
|
|
},
|
|
{
|
|
"epoch": 3.2962962962962963,
|
|
"grad_norm": 0.4247642755508423,
|
|
"learning_rate": 1.9624483850532954e-05,
|
|
"loss": 0.1997,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20382754504680634,
|
|
"step": 3150
|
|
},
|
|
{
|
|
"epoch": 3.302469135802469,
|
|
"grad_norm": 0.4075118899345398,
|
|
"learning_rate": 1.9562932937237368e-05,
|
|
"loss": 0.2003,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19111143052577972,
|
|
"step": 3155
|
|
},
|
|
{
|
|
"epoch": 3.308641975308642,
|
|
"grad_norm": 0.4129784107208252,
|
|
"learning_rate": 1.9501386165238548e-05,
|
|
"loss": 0.175,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19589219987392426,
|
|
"step": 3160
|
|
},
|
|
{
|
|
"epoch": 3.314814814814815,
|
|
"grad_norm": 0.4156089127063751,
|
|
"learning_rate": 1.9439844117704322e-05,
|
|
"loss": 0.1918,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18867067992687225,
|
|
"step": 3165
|
|
},
|
|
{
|
|
"epoch": 3.3209876543209877,
|
|
"grad_norm": 0.4023275673389435,
|
|
"learning_rate": 1.9378307377757758e-05,
|
|
"loss": 0.2127,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21073518693447113,
|
|
"step": 3170
|
|
},
|
|
{
|
|
"epoch": 3.3271604938271606,
|
|
"grad_norm": 0.49605312943458557,
|
|
"learning_rate": 1.9316776528471617e-05,
|
|
"loss": 0.1838,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.15425068140029907,
|
|
"step": 3175
|
|
},
|
|
{
|
|
"epoch": 3.3333333333333335,
|
|
"grad_norm": 0.4949839115142822,
|
|
"learning_rate": 1.9255252152862857e-05,
|
|
"loss": 0.1877,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1923992782831192,
|
|
"step": 3180
|
|
},
|
|
{
|
|
"epoch": 3.3395061728395063,
|
|
"grad_norm": 0.38800159096717834,
|
|
"learning_rate": 1.9193734833887086e-05,
|
|
"loss": 0.1993,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2134716957807541,
|
|
"step": 3185
|
|
},
|
|
{
|
|
"epoch": 3.3456790123456788,
|
|
"grad_norm": 0.40534597635269165,
|
|
"learning_rate": 1.9132225154433068e-05,
|
|
"loss": 0.2187,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19297607243061066,
|
|
"step": 3190
|
|
},
|
|
{
|
|
"epoch": 3.351851851851852,
|
|
"grad_norm": 0.39850813150405884,
|
|
"learning_rate": 1.9070723697317156e-05,
|
|
"loss": 0.1932,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19484877586364746,
|
|
"step": 3195
|
|
},
|
|
{
|
|
"epoch": 3.3580246913580245,
|
|
"grad_norm": 0.3989926278591156,
|
|
"learning_rate": 1.9009231045277817e-05,
|
|
"loss": 0.1968,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20174063742160797,
|
|
"step": 3200
|
|
},
|
|
{
|
|
"epoch": 3.3641975308641974,
|
|
"grad_norm": 0.3692739009857178,
|
|
"learning_rate": 1.894774778097006e-05,
|
|
"loss": 0.1857,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1730557531118393,
|
|
"step": 3205
|
|
},
|
|
{
|
|
"epoch": 3.3703703703703702,
|
|
"grad_norm": 0.36751773953437805,
|
|
"learning_rate": 1.8886274486959985e-05,
|
|
"loss": 0.202,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20490002632141113,
|
|
"step": 3210
|
|
},
|
|
{
|
|
"epoch": 3.376543209876543,
|
|
"grad_norm": 0.3502015173435211,
|
|
"learning_rate": 1.8824811745719194e-05,
|
|
"loss": 0.2013,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18116652965545654,
|
|
"step": 3215
|
|
},
|
|
{
|
|
"epoch": 3.382716049382716,
|
|
"grad_norm": 0.35877156257629395,
|
|
"learning_rate": 1.8763360139619285e-05,
|
|
"loss": 0.2121,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22162385284900665,
|
|
"step": 3220
|
|
},
|
|
{
|
|
"epoch": 3.388888888888889,
|
|
"grad_norm": 0.3493252992630005,
|
|
"learning_rate": 1.870192025092639e-05,
|
|
"loss": 0.2074,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20627856254577637,
|
|
"step": 3225
|
|
},
|
|
{
|
|
"epoch": 3.3950617283950617,
|
|
"grad_norm": 0.3721823990345001,
|
|
"learning_rate": 1.8640492661795583e-05,
|
|
"loss": 0.1853,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.15576171875,
|
|
"step": 3230
|
|
},
|
|
{
|
|
"epoch": 3.4012345679012346,
|
|
"grad_norm": 0.36171385645866394,
|
|
"learning_rate": 1.8579077954265404e-05,
|
|
"loss": 0.1859,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18401867151260376,
|
|
"step": 3235
|
|
},
|
|
{
|
|
"epoch": 3.4074074074074074,
|
|
"grad_norm": 0.4502449035644531,
|
|
"learning_rate": 1.851767671025234e-05,
|
|
"loss": 0.1991,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18548078835010529,
|
|
"step": 3240
|
|
},
|
|
{
|
|
"epoch": 3.4135802469135803,
|
|
"grad_norm": 0.3481305241584778,
|
|
"learning_rate": 1.8456289511545323e-05,
|
|
"loss": 0.1935,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22206616401672363,
|
|
"step": 3245
|
|
},
|
|
{
|
|
"epoch": 3.419753086419753,
|
|
"grad_norm": 0.375042587518692,
|
|
"learning_rate": 1.8394916939800175e-05,
|
|
"loss": 0.2011,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21954946219921112,
|
|
"step": 3250
|
|
},
|
|
{
|
|
"epoch": 3.425925925925926,
|
|
"grad_norm": 0.3949601948261261,
|
|
"learning_rate": 1.8333559576534142e-05,
|
|
"loss": 0.1919,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18742573261260986,
|
|
"step": 3255
|
|
},
|
|
{
|
|
"epoch": 3.432098765432099,
|
|
"grad_norm": 0.3361015021800995,
|
|
"learning_rate": 1.8272218003120368e-05,
|
|
"loss": 0.2034,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20009279251098633,
|
|
"step": 3260
|
|
},
|
|
{
|
|
"epoch": 3.4382716049382718,
|
|
"grad_norm": 0.34910255670547485,
|
|
"learning_rate": 1.8210892800782377e-05,
|
|
"loss": 0.1928,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1942591816186905,
|
|
"step": 3265
|
|
},
|
|
{
|
|
"epoch": 3.4444444444444446,
|
|
"grad_norm": 0.40600723028182983,
|
|
"learning_rate": 1.8149584550588586e-05,
|
|
"loss": 0.2034,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20267236232757568,
|
|
"step": 3270
|
|
},
|
|
{
|
|
"epoch": 3.450617283950617,
|
|
"grad_norm": 0.3564411401748657,
|
|
"learning_rate": 1.8088293833446754e-05,
|
|
"loss": 0.1843,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19972677528858185,
|
|
"step": 3275
|
|
},
|
|
{
|
|
"epoch": 3.45679012345679,
|
|
"grad_norm": 0.34283682703971863,
|
|
"learning_rate": 1.8027021230098563e-05,
|
|
"loss": 0.1944,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18615026772022247,
|
|
"step": 3280
|
|
},
|
|
{
|
|
"epoch": 3.462962962962963,
|
|
"grad_norm": 0.34584397077560425,
|
|
"learning_rate": 1.796576732111403e-05,
|
|
"loss": 0.1913,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1736493855714798,
|
|
"step": 3285
|
|
},
|
|
{
|
|
"epoch": 3.4691358024691357,
|
|
"grad_norm": 0.38462749123573303,
|
|
"learning_rate": 1.7904532686886033e-05,
|
|
"loss": 0.179,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1906375288963318,
|
|
"step": 3290
|
|
},
|
|
{
|
|
"epoch": 3.4753086419753085,
|
|
"grad_norm": 0.3968222141265869,
|
|
"learning_rate": 1.784331790762483e-05,
|
|
"loss": 0.189,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1732250452041626,
|
|
"step": 3295
|
|
},
|
|
{
|
|
"epoch": 3.4814814814814814,
|
|
"grad_norm": 0.3793950378894806,
|
|
"learning_rate": 1.778212356335256e-05,
|
|
"loss": 0.2121,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20558612048625946,
|
|
"step": 3300
|
|
},
|
|
{
|
|
"epoch": 3.4876543209876543,
|
|
"grad_norm": 0.40723833441734314,
|
|
"learning_rate": 1.772095023389772e-05,
|
|
"loss": 0.1887,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1805504560470581,
|
|
"step": 3305
|
|
},
|
|
{
|
|
"epoch": 3.493827160493827,
|
|
"grad_norm": 0.41483190655708313,
|
|
"learning_rate": 1.7659798498889695e-05,
|
|
"loss": 0.1878,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17890626192092896,
|
|
"step": 3310
|
|
},
|
|
{
|
|
"epoch": 3.5,
|
|
"grad_norm": 0.37749356031417847,
|
|
"learning_rate": 1.759866893775325e-05,
|
|
"loss": 0.1828,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.15856531262397766,
|
|
"step": 3315
|
|
},
|
|
{
|
|
"epoch": 3.506172839506173,
|
|
"grad_norm": 0.33556196093559265,
|
|
"learning_rate": 1.753756212970306e-05,
|
|
"loss": 0.1735,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17780573666095734,
|
|
"step": 3320
|
|
},
|
|
{
|
|
"epoch": 3.5123456790123457,
|
|
"grad_norm": 0.42484769225120544,
|
|
"learning_rate": 1.7476478653738203e-05,
|
|
"loss": 0.1796,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2258736938238144,
|
|
"step": 3325
|
|
},
|
|
{
|
|
"epoch": 3.5185185185185186,
|
|
"grad_norm": 0.42003607749938965,
|
|
"learning_rate": 1.7415419088636683e-05,
|
|
"loss": 0.1936,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2073868066072464,
|
|
"step": 3330
|
|
},
|
|
{
|
|
"epoch": 3.5246913580246915,
|
|
"grad_norm": 0.4104703962802887,
|
|
"learning_rate": 1.7354384012949942e-05,
|
|
"loss": 0.1798,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18249566853046417,
|
|
"step": 3335
|
|
},
|
|
{
|
|
"epoch": 3.5308641975308643,
|
|
"grad_norm": 0.4039440453052521,
|
|
"learning_rate": 1.7293374004997384e-05,
|
|
"loss": 0.1948,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2022232562303543,
|
|
"step": 3340
|
|
},
|
|
{
|
|
"epoch": 3.537037037037037,
|
|
"grad_norm": 0.4063248634338379,
|
|
"learning_rate": 1.7232389642860874e-05,
|
|
"loss": 0.2104,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20621515810489655,
|
|
"step": 3345
|
|
},
|
|
{
|
|
"epoch": 3.5432098765432096,
|
|
"grad_norm": 0.37725409865379333,
|
|
"learning_rate": 1.7171431504379312e-05,
|
|
"loss": 0.2052,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21853919327259064,
|
|
"step": 3350
|
|
},
|
|
{
|
|
"epoch": 3.549382716049383,
|
|
"grad_norm": 0.4433465301990509,
|
|
"learning_rate": 1.71105001671431e-05,
|
|
"loss": 0.2013,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18361617624759674,
|
|
"step": 3355
|
|
},
|
|
{
|
|
"epoch": 3.5555555555555554,
|
|
"grad_norm": 0.35582542419433594,
|
|
"learning_rate": 1.704959620848869e-05,
|
|
"loss": 0.1919,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2045426219701767,
|
|
"step": 3360
|
|
},
|
|
{
|
|
"epoch": 3.5617283950617287,
|
|
"grad_norm": 0.38480621576309204,
|
|
"learning_rate": 1.6988720205493117e-05,
|
|
"loss": 0.2173,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19700050354003906,
|
|
"step": 3365
|
|
},
|
|
{
|
|
"epoch": 3.567901234567901,
|
|
"grad_norm": 0.3868313431739807,
|
|
"learning_rate": 1.692787273496855e-05,
|
|
"loss": 0.1977,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1646321564912796,
|
|
"step": 3370
|
|
},
|
|
{
|
|
"epoch": 3.574074074074074,
|
|
"grad_norm": 0.3655400574207306,
|
|
"learning_rate": 1.6867054373456793e-05,
|
|
"loss": 0.1935,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22567962110042572,
|
|
"step": 3375
|
|
},
|
|
{
|
|
"epoch": 3.580246913580247,
|
|
"grad_norm": 0.3316282629966736,
|
|
"learning_rate": 1.680626569722383e-05,
|
|
"loss": 0.2078,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2229902744293213,
|
|
"step": 3380
|
|
},
|
|
{
|
|
"epoch": 3.5864197530864197,
|
|
"grad_norm": 0.40922486782073975,
|
|
"learning_rate": 1.674550728225438e-05,
|
|
"loss": 0.1978,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18556256592273712,
|
|
"step": 3385
|
|
},
|
|
{
|
|
"epoch": 3.5925925925925926,
|
|
"grad_norm": 0.35142144560813904,
|
|
"learning_rate": 1.668477970424643e-05,
|
|
"loss": 0.1827,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1836051344871521,
|
|
"step": 3390
|
|
},
|
|
{
|
|
"epoch": 3.5987654320987654,
|
|
"grad_norm": 0.4020131826400757,
|
|
"learning_rate": 1.6624083538605787e-05,
|
|
"loss": 0.1861,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19710636138916016,
|
|
"step": 3395
|
|
},
|
|
{
|
|
"epoch": 3.6049382716049383,
|
|
"grad_norm": 0.3597795069217682,
|
|
"learning_rate": 1.65634193604406e-05,
|
|
"loss": 0.1873,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19770848751068115,
|
|
"step": 3400
|
|
},
|
|
{
|
|
"epoch": 3.611111111111111,
|
|
"grad_norm": 0.37317052483558655,
|
|
"learning_rate": 1.6502787744555952e-05,
|
|
"loss": 0.1981,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21495354175567627,
|
|
"step": 3405
|
|
},
|
|
{
|
|
"epoch": 3.617283950617284,
|
|
"grad_norm": 0.435629278421402,
|
|
"learning_rate": 1.6442189265448382e-05,
|
|
"loss": 0.1844,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19390566647052765,
|
|
"step": 3410
|
|
},
|
|
{
|
|
"epoch": 3.623456790123457,
|
|
"grad_norm": 0.3396644592285156,
|
|
"learning_rate": 1.638162449730045e-05,
|
|
"loss": 0.1724,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.16634897887706757,
|
|
"step": 3415
|
|
},
|
|
{
|
|
"epoch": 3.6296296296296298,
|
|
"grad_norm": 0.3257928490638733,
|
|
"learning_rate": 1.632109401397529e-05,
|
|
"loss": 0.1814,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1670076996088028,
|
|
"step": 3420
|
|
},
|
|
{
|
|
"epoch": 3.6358024691358026,
|
|
"grad_norm": 0.3719767928123474,
|
|
"learning_rate": 1.6260598389011217e-05,
|
|
"loss": 0.1991,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23108506202697754,
|
|
"step": 3425
|
|
},
|
|
{
|
|
"epoch": 3.6419753086419755,
|
|
"grad_norm": 0.705763041973114,
|
|
"learning_rate": 1.6200138195616206e-05,
|
|
"loss": 0.2625,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.34224843978881836,
|
|
"step": 3430
|
|
},
|
|
{
|
|
"epoch": 3.648148148148148,
|
|
"grad_norm": 0.4862150549888611,
|
|
"learning_rate": 1.613971400666254e-05,
|
|
"loss": 0.3819,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.39097678661346436,
|
|
"step": 3435
|
|
},
|
|
{
|
|
"epoch": 3.6543209876543212,
|
|
"grad_norm": 0.4186333417892456,
|
|
"learning_rate": 1.607932639468135e-05,
|
|
"loss": 0.3639,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3262823820114136,
|
|
"step": 3440
|
|
},
|
|
{
|
|
"epoch": 3.6604938271604937,
|
|
"grad_norm": 0.420012503862381,
|
|
"learning_rate": 1.6018975931857193e-05,
|
|
"loss": 0.3408,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3500564396381378,
|
|
"step": 3445
|
|
},
|
|
{
|
|
"epoch": 3.6666666666666665,
|
|
"grad_norm": 0.3454051911830902,
|
|
"learning_rate": 1.5958663190022625e-05,
|
|
"loss": 0.3587,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.34928131103515625,
|
|
"step": 3450
|
|
},
|
|
{
|
|
"epoch": 3.6728395061728394,
|
|
"grad_norm": 1.068676233291626,
|
|
"learning_rate": 1.589838874065279e-05,
|
|
"loss": 0.3549,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.32218635082244873,
|
|
"step": 3455
|
|
},
|
|
{
|
|
"epoch": 3.6790123456790123,
|
|
"grad_norm": 0.2928324341773987,
|
|
"learning_rate": 1.5838153154860006e-05,
|
|
"loss": 0.3114,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3112633228302002,
|
|
"step": 3460
|
|
},
|
|
{
|
|
"epoch": 3.685185185185185,
|
|
"grad_norm": 0.33669254183769226,
|
|
"learning_rate": 1.5777957003388348e-05,
|
|
"loss": 0.32,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2840699851512909,
|
|
"step": 3465
|
|
},
|
|
{
|
|
"epoch": 3.691358024691358,
|
|
"grad_norm": 0.2960560917854309,
|
|
"learning_rate": 1.5717800856608245e-05,
|
|
"loss": 0.32,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.32209762930870056,
|
|
"step": 3470
|
|
},
|
|
{
|
|
"epoch": 3.697530864197531,
|
|
"grad_norm": 0.3454954922199249,
|
|
"learning_rate": 1.5657685284511076e-05,
|
|
"loss": 0.323,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3071681261062622,
|
|
"step": 3475
|
|
},
|
|
{
|
|
"epoch": 3.7037037037037037,
|
|
"grad_norm": 0.29561561346054077,
|
|
"learning_rate": 1.559761085670376e-05,
|
|
"loss": 0.2952,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.28527534008026123,
|
|
"step": 3480
|
|
},
|
|
{
|
|
"epoch": 3.7098765432098766,
|
|
"grad_norm": 0.3041757047176361,
|
|
"learning_rate": 1.5537578142403377e-05,
|
|
"loss": 0.3088,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3272123634815216,
|
|
"step": 3485
|
|
},
|
|
{
|
|
"epoch": 3.7160493827160495,
|
|
"grad_norm": 0.32198989391326904,
|
|
"learning_rate": 1.5477587710431737e-05,
|
|
"loss": 0.3254,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.28179481625556946,
|
|
"step": 3490
|
|
},
|
|
{
|
|
"epoch": 3.7222222222222223,
|
|
"grad_norm": 0.2945566475391388,
|
|
"learning_rate": 1.5417640129210057e-05,
|
|
"loss": 0.3228,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2769964635372162,
|
|
"step": 3495
|
|
},
|
|
{
|
|
"epoch": 3.728395061728395,
|
|
"grad_norm": 0.3476250171661377,
|
|
"learning_rate": 1.5357735966753505e-05,
|
|
"loss": 0.2995,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27308252453804016,
|
|
"step": 3500
|
|
},
|
|
{
|
|
"epoch": 3.734567901234568,
|
|
"grad_norm": 0.3318113386631012,
|
|
"learning_rate": 1.5297875790665852e-05,
|
|
"loss": 0.3042,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.32483193278312683,
|
|
"step": 3505
|
|
},
|
|
{
|
|
"epoch": 3.7407407407407405,
|
|
"grad_norm": 0.30800914764404297,
|
|
"learning_rate": 1.5238060168134095e-05,
|
|
"loss": 0.3099,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3072289526462555,
|
|
"step": 3510
|
|
},
|
|
{
|
|
"epoch": 3.746913580246914,
|
|
"grad_norm": 0.3431183397769928,
|
|
"learning_rate": 1.5178289665923078e-05,
|
|
"loss": 0.3326,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2946661114692688,
|
|
"step": 3515
|
|
},
|
|
{
|
|
"epoch": 3.753086419753086,
|
|
"grad_norm": 0.49984854459762573,
|
|
"learning_rate": 1.511856485037012e-05,
|
|
"loss": 0.2873,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26836875081062317,
|
|
"step": 3520
|
|
},
|
|
{
|
|
"epoch": 3.7592592592592595,
|
|
"grad_norm": 0.3298574388027191,
|
|
"learning_rate": 1.5058886287379637e-05,
|
|
"loss": 0.3003,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.33067700266838074,
|
|
"step": 3525
|
|
},
|
|
{
|
|
"epoch": 3.765432098765432,
|
|
"grad_norm": 0.340242862701416,
|
|
"learning_rate": 1.4999254542417819e-05,
|
|
"loss": 0.304,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.31879717111587524,
|
|
"step": 3530
|
|
},
|
|
{
|
|
"epoch": 3.771604938271605,
|
|
"grad_norm": 0.34559252858161926,
|
|
"learning_rate": 1.4939670180507224e-05,
|
|
"loss": 0.3327,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3706111013889313,
|
|
"step": 3535
|
|
},
|
|
{
|
|
"epoch": 3.7777777777777777,
|
|
"grad_norm": 0.35321637988090515,
|
|
"learning_rate": 1.4880133766221461e-05,
|
|
"loss": 0.3167,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.33479252457618713,
|
|
"step": 3540
|
|
},
|
|
{
|
|
"epoch": 3.7839506172839505,
|
|
"grad_norm": 0.3217999339103699,
|
|
"learning_rate": 1.4820645863679794e-05,
|
|
"loss": 0.3222,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.30042847990989685,
|
|
"step": 3545
|
|
},
|
|
{
|
|
"epoch": 3.7901234567901234,
|
|
"grad_norm": 0.29845622181892395,
|
|
"learning_rate": 1.4761207036541883e-05,
|
|
"loss": 0.316,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3213569223880768,
|
|
"step": 3550
|
|
},
|
|
{
|
|
"epoch": 3.7962962962962963,
|
|
"grad_norm": 0.33212509751319885,
|
|
"learning_rate": 1.470181784800235e-05,
|
|
"loss": 0.3186,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.36581292748451233,
|
|
"step": 3555
|
|
},
|
|
{
|
|
"epoch": 3.802469135802469,
|
|
"grad_norm": 0.32539695501327515,
|
|
"learning_rate": 1.464247886078547e-05,
|
|
"loss": 0.3239,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3556390106678009,
|
|
"step": 3560
|
|
},
|
|
{
|
|
"epoch": 3.808641975308642,
|
|
"grad_norm": 0.3083668351173401,
|
|
"learning_rate": 1.4583190637139901e-05,
|
|
"loss": 0.3142,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.28246697783470154,
|
|
"step": 3565
|
|
},
|
|
{
|
|
"epoch": 3.814814814814815,
|
|
"grad_norm": 0.2815481722354889,
|
|
"learning_rate": 1.4523953738833257e-05,
|
|
"loss": 0.3104,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.30584895610809326,
|
|
"step": 3570
|
|
},
|
|
{
|
|
"epoch": 3.8209876543209877,
|
|
"grad_norm": 0.28309598565101624,
|
|
"learning_rate": 1.4464768727146857e-05,
|
|
"loss": 0.2995,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27399513125419617,
|
|
"step": 3575
|
|
},
|
|
{
|
|
"epoch": 3.8271604938271606,
|
|
"grad_norm": 0.33006006479263306,
|
|
"learning_rate": 1.4405636162870378e-05,
|
|
"loss": 0.31,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2714330852031708,
|
|
"step": 3580
|
|
},
|
|
{
|
|
"epoch": 3.8333333333333335,
|
|
"grad_norm": 1.4674088954925537,
|
|
"learning_rate": 1.4346556606296553e-05,
|
|
"loss": 0.3056,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3002880811691284,
|
|
"step": 3585
|
|
},
|
|
{
|
|
"epoch": 3.8395061728395063,
|
|
"grad_norm": 0.28623151779174805,
|
|
"learning_rate": 1.4287530617215857e-05,
|
|
"loss": 0.3089,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.327418714761734,
|
|
"step": 3590
|
|
},
|
|
{
|
|
"epoch": 3.8456790123456788,
|
|
"grad_norm": 0.29594650864601135,
|
|
"learning_rate": 1.422855875491119e-05,
|
|
"loss": 0.3187,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3702506721019745,
|
|
"step": 3595
|
|
},
|
|
{
|
|
"epoch": 3.851851851851852,
|
|
"grad_norm": 0.2914753556251526,
|
|
"learning_rate": 1.41696415781526e-05,
|
|
"loss": 0.2983,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26200807094573975,
|
|
"step": 3600
|
|
},
|
|
{
|
|
"epoch": 3.8580246913580245,
|
|
"grad_norm": 0.2902332842350006,
|
|
"learning_rate": 1.4110779645191984e-05,
|
|
"loss": 0.3156,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.30105137825012207,
|
|
"step": 3605
|
|
},
|
|
{
|
|
"epoch": 3.8641975308641974,
|
|
"grad_norm": 0.2838253080844879,
|
|
"learning_rate": 1.4051973513757785e-05,
|
|
"loss": 0.3086,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3202577531337738,
|
|
"step": 3610
|
|
},
|
|
{
|
|
"epoch": 3.8703703703703702,
|
|
"grad_norm": 0.2753179669380188,
|
|
"learning_rate": 1.39932237410497e-05,
|
|
"loss": 0.3146,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2824477255344391,
|
|
"step": 3615
|
|
},
|
|
{
|
|
"epoch": 3.876543209876543,
|
|
"grad_norm": 0.7304513454437256,
|
|
"learning_rate": 1.3934530883733457e-05,
|
|
"loss": 0.287,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.28868788480758667,
|
|
"step": 3620
|
|
},
|
|
{
|
|
"epoch": 3.882716049382716,
|
|
"grad_norm": 0.4831741750240326,
|
|
"learning_rate": 1.3875895497935465e-05,
|
|
"loss": 0.2381,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2148720771074295,
|
|
"step": 3625
|
|
},
|
|
{
|
|
"epoch": 3.888888888888889,
|
|
"grad_norm": 0.4189258813858032,
|
|
"learning_rate": 1.381731813923759e-05,
|
|
"loss": 0.2288,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.230940043926239,
|
|
"step": 3630
|
|
},
|
|
{
|
|
"epoch": 3.8950617283950617,
|
|
"grad_norm": 0.38482531905174255,
|
|
"learning_rate": 1.3758799362671871e-05,
|
|
"loss": 0.2533,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24300383031368256,
|
|
"step": 3635
|
|
},
|
|
{
|
|
"epoch": 3.9012345679012346,
|
|
"grad_norm": 0.3646378219127655,
|
|
"learning_rate": 1.370033972271529e-05,
|
|
"loss": 0.2588,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23884914815425873,
|
|
"step": 3640
|
|
},
|
|
{
|
|
"epoch": 3.9074074074074074,
|
|
"grad_norm": 0.3003780245780945,
|
|
"learning_rate": 1.3641939773284478e-05,
|
|
"loss": 0.2407,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2111782282590866,
|
|
"step": 3645
|
|
},
|
|
{
|
|
"epoch": 3.9135802469135803,
|
|
"grad_norm": 0.3421449661254883,
|
|
"learning_rate": 1.3583600067730496e-05,
|
|
"loss": 0.2403,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2603928744792938,
|
|
"step": 3650
|
|
},
|
|
{
|
|
"epoch": 3.919753086419753,
|
|
"grad_norm": 0.4100312292575836,
|
|
"learning_rate": 1.3525321158833582e-05,
|
|
"loss": 0.2393,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23270289599895477,
|
|
"step": 3655
|
|
},
|
|
{
|
|
"epoch": 3.925925925925926,
|
|
"grad_norm": 0.3720301687717438,
|
|
"learning_rate": 1.3467103598797911e-05,
|
|
"loss": 0.2365,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22509299218654633,
|
|
"step": 3660
|
|
},
|
|
{
|
|
"epoch": 3.932098765432099,
|
|
"grad_norm": 0.3403886556625366,
|
|
"learning_rate": 1.3408947939246368e-05,
|
|
"loss": 0.2519,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2581208646297455,
|
|
"step": 3665
|
|
},
|
|
{
|
|
"epoch": 3.9382716049382713,
|
|
"grad_norm": 0.32090869545936584,
|
|
"learning_rate": 1.335085473121531e-05,
|
|
"loss": 0.2464,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.254538357257843,
|
|
"step": 3670
|
|
},
|
|
{
|
|
"epoch": 3.9444444444444446,
|
|
"grad_norm": 0.33962109684944153,
|
|
"learning_rate": 1.3292824525149373e-05,
|
|
"loss": 0.2689,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.289763867855072,
|
|
"step": 3675
|
|
},
|
|
{
|
|
"epoch": 3.950617283950617,
|
|
"grad_norm": 0.3646473288536072,
|
|
"learning_rate": 1.3234857870896217e-05,
|
|
"loss": 0.2686,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24301175773143768,
|
|
"step": 3680
|
|
},
|
|
{
|
|
"epoch": 3.9567901234567904,
|
|
"grad_norm": 0.29526692628860474,
|
|
"learning_rate": 1.3176955317701348e-05,
|
|
"loss": 0.2664,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2868461310863495,
|
|
"step": 3685
|
|
},
|
|
{
|
|
"epoch": 3.962962962962963,
|
|
"grad_norm": 0.3017801344394684,
|
|
"learning_rate": 1.3119117414202881e-05,
|
|
"loss": 0.2566,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.241042360663414,
|
|
"step": 3690
|
|
},
|
|
{
|
|
"epoch": 3.9691358024691357,
|
|
"grad_norm": 0.34227466583251953,
|
|
"learning_rate": 1.3061344708426409e-05,
|
|
"loss": 0.2563,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25882720947265625,
|
|
"step": 3695
|
|
},
|
|
{
|
|
"epoch": 3.9753086419753085,
|
|
"grad_norm": 0.4984656572341919,
|
|
"learning_rate": 1.3003637747779712e-05,
|
|
"loss": 0.2353,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20988202095031738,
|
|
"step": 3700
|
|
},
|
|
{
|
|
"epoch": 3.9814814814814814,
|
|
"grad_norm": 0.3332652151584625,
|
|
"learning_rate": 1.294599707904764e-05,
|
|
"loss": 0.2245,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2129083126783371,
|
|
"step": 3705
|
|
},
|
|
{
|
|
"epoch": 3.9876543209876543,
|
|
"grad_norm": 0.4685971736907959,
|
|
"learning_rate": 1.2888423248386935e-05,
|
|
"loss": 0.226,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2160855084657669,
|
|
"step": 3710
|
|
},
|
|
{
|
|
"epoch": 3.993827160493827,
|
|
"grad_norm": 0.3762311041355133,
|
|
"learning_rate": 1.2830916801321007e-05,
|
|
"loss": 0.2291,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24430584907531738,
|
|
"step": 3715
|
|
},
|
|
{
|
|
"epoch": 4.001234567901235,
|
|
"grad_norm": 0.5948289036750793,
|
|
"learning_rate": 1.2773478282734804e-05,
|
|
"loss": 0.2387,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23446960747241974,
|
|
"step": 3720
|
|
},
|
|
{
|
|
"epoch": 4.007407407407407,
|
|
"grad_norm": 0.44916248321533203,
|
|
"learning_rate": 1.2716108236869635e-05,
|
|
"loss": 0.2039,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20267975330352783,
|
|
"step": 3725
|
|
},
|
|
{
|
|
"epoch": 4.01358024691358,
|
|
"grad_norm": 0.3778553903102875,
|
|
"learning_rate": 1.265880720731803e-05,
|
|
"loss": 0.199,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18288271129131317,
|
|
"step": 3730
|
|
},
|
|
{
|
|
"epoch": 4.019753086419753,
|
|
"grad_norm": 0.44959595799446106,
|
|
"learning_rate": 1.2601575737018554e-05,
|
|
"loss": 0.1934,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19528596103191376,
|
|
"step": 3735
|
|
},
|
|
{
|
|
"epoch": 4.025925925925926,
|
|
"grad_norm": 0.3482433259487152,
|
|
"learning_rate": 1.2544414368250699e-05,
|
|
"loss": 0.1917,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20399637520313263,
|
|
"step": 3740
|
|
},
|
|
{
|
|
"epoch": 4.0320987654320986,
|
|
"grad_norm": 0.2997663915157318,
|
|
"learning_rate": 1.2487323642629732e-05,
|
|
"loss": 0.196,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21616096794605255,
|
|
"step": 3745
|
|
},
|
|
{
|
|
"epoch": 4.038271604938272,
|
|
"grad_norm": 0.33535563945770264,
|
|
"learning_rate": 1.2430304101101554e-05,
|
|
"loss": 0.1881,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20673246681690216,
|
|
"step": 3750
|
|
},
|
|
{
|
|
"epoch": 4.044444444444444,
|
|
"grad_norm": 0.3675876557826996,
|
|
"learning_rate": 1.2373356283937593e-05,
|
|
"loss": 0.1969,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2179686576128006,
|
|
"step": 3755
|
|
},
|
|
{
|
|
"epoch": 4.050617283950618,
|
|
"grad_norm": 0.34149879217147827,
|
|
"learning_rate": 1.2316480730729647e-05,
|
|
"loss": 0.1946,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.16587744653224945,
|
|
"step": 3760
|
|
},
|
|
{
|
|
"epoch": 4.05679012345679,
|
|
"grad_norm": 0.46574702858924866,
|
|
"learning_rate": 1.225967798038485e-05,
|
|
"loss": 0.1919,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18906481564044952,
|
|
"step": 3765
|
|
},
|
|
{
|
|
"epoch": 4.062962962962963,
|
|
"grad_norm": 0.32798323035240173,
|
|
"learning_rate": 1.2202948571120463e-05,
|
|
"loss": 0.1825,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20531387627124786,
|
|
"step": 3770
|
|
},
|
|
{
|
|
"epoch": 4.069135802469136,
|
|
"grad_norm": 0.45086562633514404,
|
|
"learning_rate": 1.2146293040458854e-05,
|
|
"loss": 0.2071,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20760412514209747,
|
|
"step": 3775
|
|
},
|
|
{
|
|
"epoch": 4.075308641975309,
|
|
"grad_norm": 0.3985859155654907,
|
|
"learning_rate": 1.2089711925222364e-05,
|
|
"loss": 0.1876,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18711812794208527,
|
|
"step": 3780
|
|
},
|
|
{
|
|
"epoch": 4.0814814814814815,
|
|
"grad_norm": 0.4147719740867615,
|
|
"learning_rate": 1.2033205761528248e-05,
|
|
"loss": 0.2039,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22381742298603058,
|
|
"step": 3785
|
|
},
|
|
{
|
|
"epoch": 4.087654320987654,
|
|
"grad_norm": 0.38564690947532654,
|
|
"learning_rate": 1.1976775084783565e-05,
|
|
"loss": 0.2057,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2285938262939453,
|
|
"step": 3790
|
|
},
|
|
{
|
|
"epoch": 4.093827160493827,
|
|
"grad_norm": 0.38278764486312866,
|
|
"learning_rate": 1.1920420429680124e-05,
|
|
"loss": 0.198,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18325306475162506,
|
|
"step": 3795
|
|
},
|
|
{
|
|
"epoch": 4.1,
|
|
"grad_norm": 0.424329936504364,
|
|
"learning_rate": 1.1864142330189425e-05,
|
|
"loss": 0.2147,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19788433611392975,
|
|
"step": 3800
|
|
},
|
|
{
|
|
"epoch": 4.106172839506173,
|
|
"grad_norm": 0.452142596244812,
|
|
"learning_rate": 1.1807941319557574e-05,
|
|
"loss": 0.1926,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2124447077512741,
|
|
"step": 3805
|
|
},
|
|
{
|
|
"epoch": 4.112345679012345,
|
|
"grad_norm": 0.3925051689147949,
|
|
"learning_rate": 1.1751817930300255e-05,
|
|
"loss": 0.1916,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18426424264907837,
|
|
"step": 3810
|
|
},
|
|
{
|
|
"epoch": 4.118518518518519,
|
|
"grad_norm": 0.38360756635665894,
|
|
"learning_rate": 1.1695772694197655e-05,
|
|
"loss": 0.1981,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19311527907848358,
|
|
"step": 3815
|
|
},
|
|
{
|
|
"epoch": 4.124691358024691,
|
|
"grad_norm": 0.3495960235595703,
|
|
"learning_rate": 1.1639806142289483e-05,
|
|
"loss": 0.1886,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19464744627475739,
|
|
"step": 3820
|
|
},
|
|
{
|
|
"epoch": 4.130864197530864,
|
|
"grad_norm": 0.41566383838653564,
|
|
"learning_rate": 1.1583918804869856e-05,
|
|
"loss": 0.1925,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18602530658245087,
|
|
"step": 3825
|
|
},
|
|
{
|
|
"epoch": 4.137037037037037,
|
|
"grad_norm": 0.4040479063987732,
|
|
"learning_rate": 1.1528111211482342e-05,
|
|
"loss": 0.201,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1808733344078064,
|
|
"step": 3830
|
|
},
|
|
{
|
|
"epoch": 4.14320987654321,
|
|
"grad_norm": 0.37992390990257263,
|
|
"learning_rate": 1.1472383890914926e-05,
|
|
"loss": 0.2078,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2010001689195633,
|
|
"step": 3835
|
|
},
|
|
{
|
|
"epoch": 4.149382716049383,
|
|
"grad_norm": 0.3992995321750641,
|
|
"learning_rate": 1.1416737371194965e-05,
|
|
"loss": 0.1939,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18974147737026215,
|
|
"step": 3840
|
|
},
|
|
{
|
|
"epoch": 4.155555555555556,
|
|
"grad_norm": 0.37349453568458557,
|
|
"learning_rate": 1.1361172179584242e-05,
|
|
"loss": 0.1899,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19742612540721893,
|
|
"step": 3845
|
|
},
|
|
{
|
|
"epoch": 4.161728395061728,
|
|
"grad_norm": 0.37395718693733215,
|
|
"learning_rate": 1.130568884257391e-05,
|
|
"loss": 0.208,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1862659454345703,
|
|
"step": 3850
|
|
},
|
|
{
|
|
"epoch": 4.167901234567902,
|
|
"grad_norm": 0.45303723216056824,
|
|
"learning_rate": 1.1250287885879556e-05,
|
|
"loss": 0.1994,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18550777435302734,
|
|
"step": 3855
|
|
},
|
|
{
|
|
"epoch": 4.174074074074074,
|
|
"grad_norm": 0.36268696188926697,
|
|
"learning_rate": 1.1194969834436203e-05,
|
|
"loss": 0.202,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21677757799625397,
|
|
"step": 3860
|
|
},
|
|
{
|
|
"epoch": 4.1802469135802465,
|
|
"grad_norm": 0.42992910742759705,
|
|
"learning_rate": 1.1139735212393303e-05,
|
|
"loss": 0.1955,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2094179391860962,
|
|
"step": 3865
|
|
},
|
|
{
|
|
"epoch": 4.18641975308642,
|
|
"grad_norm": 0.3596706688404083,
|
|
"learning_rate": 1.1084584543109811e-05,
|
|
"loss": 0.1861,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20604555308818817,
|
|
"step": 3870
|
|
},
|
|
{
|
|
"epoch": 4.192592592592592,
|
|
"grad_norm": 0.3959878385066986,
|
|
"learning_rate": 1.1029518349149218e-05,
|
|
"loss": 0.198,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19569294154644012,
|
|
"step": 3875
|
|
},
|
|
{
|
|
"epoch": 4.1987654320987655,
|
|
"grad_norm": 0.32062244415283203,
|
|
"learning_rate": 1.0974537152274602e-05,
|
|
"loss": 0.199,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1599230021238327,
|
|
"step": 3880
|
|
},
|
|
{
|
|
"epoch": 4.204938271604938,
|
|
"grad_norm": 0.4291191101074219,
|
|
"learning_rate": 1.0919641473443652e-05,
|
|
"loss": 0.1817,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17450173199176788,
|
|
"step": 3885
|
|
},
|
|
{
|
|
"epoch": 4.211111111111111,
|
|
"grad_norm": 0.31011873483657837,
|
|
"learning_rate": 1.0864831832803774e-05,
|
|
"loss": 0.1846,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1902369260787964,
|
|
"step": 3890
|
|
},
|
|
{
|
|
"epoch": 4.217283950617284,
|
|
"grad_norm": 0.3439962565898895,
|
|
"learning_rate": 1.0810108749687154e-05,
|
|
"loss": 0.1971,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2097424417734146,
|
|
"step": 3895
|
|
},
|
|
{
|
|
"epoch": 4.223456790123457,
|
|
"grad_norm": 0.39591050148010254,
|
|
"learning_rate": 1.07554727426058e-05,
|
|
"loss": 0.1881,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20967306196689606,
|
|
"step": 3900
|
|
},
|
|
{
|
|
"epoch": 4.006172839506172,
|
|
"grad_norm": 0.3799671530723572,
|
|
"learning_rate": 1.0700924329246679e-05,
|
|
"loss": 0.192,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19342118501663208,
|
|
"step": 3905
|
|
},
|
|
{
|
|
"epoch": 4.012345679012346,
|
|
"grad_norm": 0.38043487071990967,
|
|
"learning_rate": 1.0646464026466793e-05,
|
|
"loss": 0.1899,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1756172627210617,
|
|
"step": 3910
|
|
},
|
|
{
|
|
"epoch": 4.018518518518518,
|
|
"grad_norm": 0.44888836145401,
|
|
"learning_rate": 1.0592092350288266e-05,
|
|
"loss": 0.1853,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18900108337402344,
|
|
"step": 3915
|
|
},
|
|
{
|
|
"epoch": 4.0246913580246915,
|
|
"grad_norm": 0.3534156084060669,
|
|
"learning_rate": 1.0537809815893454e-05,
|
|
"loss": 0.1837,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1965215802192688,
|
|
"step": 3920
|
|
},
|
|
{
|
|
"epoch": 4.030864197530864,
|
|
"grad_norm": 0.29467928409576416,
|
|
"learning_rate": 1.0483616937620127e-05,
|
|
"loss": 0.1879,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20668862760066986,
|
|
"step": 3925
|
|
},
|
|
{
|
|
"epoch": 4.037037037037037,
|
|
"grad_norm": 0.3228740990161896,
|
|
"learning_rate": 1.04295142289565e-05,
|
|
"loss": 0.1803,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19820208847522736,
|
|
"step": 3930
|
|
},
|
|
{
|
|
"epoch": 4.04320987654321,
|
|
"grad_norm": 0.36807671189308167,
|
|
"learning_rate": 1.0375502202536414e-05,
|
|
"loss": 0.1893,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2100045084953308,
|
|
"step": 3935
|
|
},
|
|
{
|
|
"epoch": 4.049382716049383,
|
|
"grad_norm": 0.37622469663619995,
|
|
"learning_rate": 1.0321581370134503e-05,
|
|
"loss": 0.1871,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.15942655503749847,
|
|
"step": 3940
|
|
},
|
|
{
|
|
"epoch": 4.055555555555555,
|
|
"grad_norm": 0.43238988518714905,
|
|
"learning_rate": 1.0267752242661313e-05,
|
|
"loss": 0.1851,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18185381591320038,
|
|
"step": 3945
|
|
},
|
|
{
|
|
"epoch": 4.061728395061729,
|
|
"grad_norm": 0.3481627404689789,
|
|
"learning_rate": 1.0214015330158449e-05,
|
|
"loss": 0.1757,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1983906477689743,
|
|
"step": 3950
|
|
},
|
|
{
|
|
"epoch": 4.067901234567901,
|
|
"grad_norm": 0.4400255084037781,
|
|
"learning_rate": 1.0160371141793787e-05,
|
|
"loss": 0.2,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19973844289779663,
|
|
"step": 3955
|
|
},
|
|
{
|
|
"epoch": 4.074074074074074,
|
|
"grad_norm": 0.407595157623291,
|
|
"learning_rate": 1.0106820185856599e-05,
|
|
"loss": 0.1807,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17894069850444794,
|
|
"step": 3960
|
|
},
|
|
{
|
|
"epoch": 4.080246913580247,
|
|
"grad_norm": 0.4154496490955353,
|
|
"learning_rate": 1.0053362969752772e-05,
|
|
"loss": 0.1967,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21556039154529572,
|
|
"step": 3965
|
|
},
|
|
{
|
|
"epoch": 4.08641975308642,
|
|
"grad_norm": 0.3922232687473297,
|
|
"learning_rate": 1.0000000000000006e-05,
|
|
"loss": 0.1979,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22001487016677856,
|
|
"step": 3970
|
|
},
|
|
{
|
|
"epoch": 4.092592592592593,
|
|
"grad_norm": 0.3691295385360718,
|
|
"learning_rate": 9.946731782222962e-06,
|
|
"loss": 0.1911,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1770067662000656,
|
|
"step": 3975
|
|
},
|
|
{
|
|
"epoch": 4.098765432098766,
|
|
"grad_norm": 0.430947870016098,
|
|
"learning_rate": 9.893558821148538e-06,
|
|
"loss": 0.2069,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19193418323993683,
|
|
"step": 3980
|
|
},
|
|
{
|
|
"epoch": 4.104938271604938,
|
|
"grad_norm": 0.4151501953601837,
|
|
"learning_rate": 9.840481620601055e-06,
|
|
"loss": 0.1859,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2038847953081131,
|
|
"step": 3985
|
|
},
|
|
{
|
|
"epoch": 4.111111111111111,
|
|
"grad_norm": 0.40241536498069763,
|
|
"learning_rate": 9.787500683497464e-06,
|
|
"loss": 0.1846,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1771858185529709,
|
|
"step": 3990
|
|
},
|
|
{
|
|
"epoch": 4.117283950617284,
|
|
"grad_norm": 0.4031103551387787,
|
|
"learning_rate": 9.734616511842598e-06,
|
|
"loss": 0.191,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18606805801391602,
|
|
"step": 3995
|
|
},
|
|
{
|
|
"epoch": 4.1234567901234565,
|
|
"grad_norm": 0.37315109372138977,
|
|
"learning_rate": 9.681829606724462e-06,
|
|
"loss": 0.182,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18720366060733795,
|
|
"step": 4000
|
|
},
|
|
{
|
|
"epoch": 4.12962962962963,
|
|
"grad_norm": 0.43149131536483765,
|
|
"learning_rate": 9.629140468309397e-06,
|
|
"loss": 0.1851,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18030691146850586,
|
|
"step": 4005
|
|
},
|
|
{
|
|
"epoch": 4.135802469135802,
|
|
"grad_norm": 0.40868282318115234,
|
|
"learning_rate": 9.57654959583739e-06,
|
|
"loss": 0.1936,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1741911619901657,
|
|
"step": 4010
|
|
},
|
|
{
|
|
"epoch": 4.1419753086419755,
|
|
"grad_norm": 0.3822796046733856,
|
|
"learning_rate": 9.524057487617357e-06,
|
|
"loss": 0.2005,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19428043067455292,
|
|
"step": 4015
|
|
},
|
|
{
|
|
"epoch": 4.148148148148148,
|
|
"grad_norm": 0.3896673619747162,
|
|
"learning_rate": 9.471664641022403e-06,
|
|
"loss": 0.1872,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1828230768442154,
|
|
"step": 4020
|
|
},
|
|
{
|
|
"epoch": 4.154320987654321,
|
|
"grad_norm": 0.3884253203868866,
|
|
"learning_rate": 9.419371552485082e-06,
|
|
"loss": 0.1829,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18911968171596527,
|
|
"step": 4025
|
|
},
|
|
{
|
|
"epoch": 4.160493827160494,
|
|
"grad_norm": 0.36914923787117004,
|
|
"learning_rate": 9.36717871749275e-06,
|
|
"loss": 0.2008,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17896592617034912,
|
|
"step": 4030
|
|
},
|
|
{
|
|
"epoch": 4.166666666666667,
|
|
"grad_norm": 0.4583016633987427,
|
|
"learning_rate": 9.315086630582839e-06,
|
|
"loss": 0.1924,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17851655185222626,
|
|
"step": 4035
|
|
},
|
|
{
|
|
"epoch": 4.172839506172839,
|
|
"grad_norm": 0.3633919060230255,
|
|
"learning_rate": 9.263095785338148e-06,
|
|
"loss": 0.1952,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2082824558019638,
|
|
"step": 4040
|
|
},
|
|
{
|
|
"epoch": 4.179012345679013,
|
|
"grad_norm": 0.44875845313072205,
|
|
"learning_rate": 9.211206674382215e-06,
|
|
"loss": 0.1886,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20183007419109344,
|
|
"step": 4045
|
|
},
|
|
{
|
|
"epoch": 4.185185185185185,
|
|
"grad_norm": 0.3686891496181488,
|
|
"learning_rate": 9.159419789374626e-06,
|
|
"loss": 0.1795,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19703872501850128,
|
|
"step": 4050
|
|
},
|
|
{
|
|
"epoch": 4.191358024691358,
|
|
"grad_norm": 0.41455554962158203,
|
|
"learning_rate": 9.107735621006335e-06,
|
|
"loss": 0.1909,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1892962008714676,
|
|
"step": 4055
|
|
},
|
|
{
|
|
"epoch": 4.197530864197531,
|
|
"grad_norm": 0.32353290915489197,
|
|
"learning_rate": 9.05615465899506e-06,
|
|
"loss": 0.1918,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1529625803232193,
|
|
"step": 4060
|
|
},
|
|
{
|
|
"epoch": 4.203703703703703,
|
|
"grad_norm": 0.47212788462638855,
|
|
"learning_rate": 9.004677392080595e-06,
|
|
"loss": 0.1752,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1684436798095703,
|
|
"step": 4065
|
|
},
|
|
{
|
|
"epoch": 4.209876543209877,
|
|
"grad_norm": 0.3098747134208679,
|
|
"learning_rate": 8.953304308020228e-06,
|
|
"loss": 0.1783,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18456339836120605,
|
|
"step": 4070
|
|
},
|
|
{
|
|
"epoch": 4.216049382716049,
|
|
"grad_norm": 0.3594570457935333,
|
|
"learning_rate": 8.902035893584065e-06,
|
|
"loss": 0.1904,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20234109461307526,
|
|
"step": 4075
|
|
},
|
|
{
|
|
"epoch": 4.222222222222222,
|
|
"grad_norm": 0.39718034863471985,
|
|
"learning_rate": 8.850872634550477e-06,
|
|
"loss": 0.1817,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2017226666212082,
|
|
"step": 4080
|
|
},
|
|
{
|
|
"epoch": 4.228395061728395,
|
|
"grad_norm": 0.42863839864730835,
|
|
"learning_rate": 8.79981501570144e-06,
|
|
"loss": 0.1977,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21498949825763702,
|
|
"step": 4085
|
|
},
|
|
{
|
|
"epoch": 4.234567901234568,
|
|
"grad_norm": 0.3937453627586365,
|
|
"learning_rate": 8.748863520817985e-06,
|
|
"loss": 0.206,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21898289024829865,
|
|
"step": 4090
|
|
},
|
|
{
|
|
"epoch": 4.2407407407407405,
|
|
"grad_norm": 0.4446229040622711,
|
|
"learning_rate": 8.698018632675603e-06,
|
|
"loss": 0.2,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19871234893798828,
|
|
"step": 4095
|
|
},
|
|
{
|
|
"epoch": 4.246913580246914,
|
|
"grad_norm": 0.40450623631477356,
|
|
"learning_rate": 8.647280833039632e-06,
|
|
"loss": 0.2005,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21865516901016235,
|
|
"step": 4100
|
|
},
|
|
{
|
|
"epoch": 4.253086419753086,
|
|
"grad_norm": 0.40501993894577026,
|
|
"learning_rate": 8.596650602660766e-06,
|
|
"loss": 0.204,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17829759418964386,
|
|
"step": 4105
|
|
},
|
|
{
|
|
"epoch": 4.2592592592592595,
|
|
"grad_norm": 0.449515700340271,
|
|
"learning_rate": 8.546128421270441e-06,
|
|
"loss": 0.1906,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18337683379650116,
|
|
"step": 4110
|
|
},
|
|
{
|
|
"epoch": 4.265432098765432,
|
|
"grad_norm": 0.36625704169273376,
|
|
"learning_rate": 8.495714767576302e-06,
|
|
"loss": 0.183,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2003413885831833,
|
|
"step": 4115
|
|
},
|
|
{
|
|
"epoch": 4.271604938271605,
|
|
"grad_norm": 0.38719263672828674,
|
|
"learning_rate": 8.445410119257666e-06,
|
|
"loss": 0.192,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20701925456523895,
|
|
"step": 4120
|
|
},
|
|
{
|
|
"epoch": 4.277777777777778,
|
|
"grad_norm": 0.45129629969596863,
|
|
"learning_rate": 8.39521495296104e-06,
|
|
"loss": 0.188,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.16362039744853973,
|
|
"step": 4125
|
|
},
|
|
{
|
|
"epoch": 4.283950617283951,
|
|
"grad_norm": 0.37117305397987366,
|
|
"learning_rate": 8.34512974429553e-06,
|
|
"loss": 0.1844,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18280120193958282,
|
|
"step": 4130
|
|
},
|
|
{
|
|
"epoch": 4.290123456790123,
|
|
"grad_norm": 0.4584495425224304,
|
|
"learning_rate": 8.295154967828367e-06,
|
|
"loss": 0.1987,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19991116225719452,
|
|
"step": 4135
|
|
},
|
|
{
|
|
"epoch": 4.296296296296296,
|
|
"grad_norm": 0.42150580883026123,
|
|
"learning_rate": 8.245291097080463e-06,
|
|
"loss": 0.1909,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1953328400850296,
|
|
"step": 4140
|
|
},
|
|
{
|
|
"epoch": 4.302469135802469,
|
|
"grad_norm": 0.40928173065185547,
|
|
"learning_rate": 8.19553860452183e-06,
|
|
"loss": 0.1922,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18242502212524414,
|
|
"step": 4145
|
|
},
|
|
{
|
|
"epoch": 4.308641975308642,
|
|
"grad_norm": 0.3762543201446533,
|
|
"learning_rate": 8.145897961567155e-06,
|
|
"loss": 0.1678,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1879187375307083,
|
|
"step": 4150
|
|
},
|
|
{
|
|
"epoch": 4.314814814814815,
|
|
"grad_norm": 0.424780935049057,
|
|
"learning_rate": 8.096369638571349e-06,
|
|
"loss": 0.1841,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18120123445987701,
|
|
"step": 4155
|
|
},
|
|
{
|
|
"epoch": 4.320987654320987,
|
|
"grad_norm": 0.3895569443702698,
|
|
"learning_rate": 8.046954104825062e-06,
|
|
"loss": 0.2051,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20310981571674347,
|
|
"step": 4160
|
|
},
|
|
{
|
|
"epoch": 4.327160493827161,
|
|
"grad_norm": 0.4943417012691498,
|
|
"learning_rate": 7.997651828550227e-06,
|
|
"loss": 0.1763,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.146974578499794,
|
|
"step": 4165
|
|
},
|
|
{
|
|
"epoch": 4.333333333333333,
|
|
"grad_norm": 0.399400919675827,
|
|
"learning_rate": 7.948463276895664e-06,
|
|
"loss": 0.1808,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18697625398635864,
|
|
"step": 4170
|
|
},
|
|
{
|
|
"epoch": 4.339506172839506,
|
|
"grad_norm": 0.3938871920108795,
|
|
"learning_rate": 7.899388915932606e-06,
|
|
"loss": 0.1914,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20582695305347443,
|
|
"step": 4175
|
|
},
|
|
{
|
|
"epoch": 4.345679012345679,
|
|
"grad_norm": 0.4100869596004486,
|
|
"learning_rate": 7.850429210650327e-06,
|
|
"loss": 0.2113,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1851186603307724,
|
|
"step": 4180
|
|
},
|
|
{
|
|
"epoch": 4.351851851851852,
|
|
"grad_norm": 0.39267656207084656,
|
|
"learning_rate": 7.801584624951713e-06,
|
|
"loss": 0.1859,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1865524798631668,
|
|
"step": 4185
|
|
},
|
|
{
|
|
"epoch": 4.3580246913580245,
|
|
"grad_norm": 0.4011878967285156,
|
|
"learning_rate": 7.75285562164885e-06,
|
|
"loss": 0.1892,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19341492652893066,
|
|
"step": 4190
|
|
},
|
|
{
|
|
"epoch": 4.364197530864198,
|
|
"grad_norm": 0.36299359798431396,
|
|
"learning_rate": 7.704242662458681e-06,
|
|
"loss": 0.1783,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1660061627626419,
|
|
"step": 4195
|
|
},
|
|
{
|
|
"epoch": 4.37037037037037,
|
|
"grad_norm": 0.41211390495300293,
|
|
"learning_rate": 7.655746207998603e-06,
|
|
"loss": 0.1942,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19698460400104523,
|
|
"step": 4200
|
|
},
|
|
{
|
|
"epoch": 4.3765432098765435,
|
|
"grad_norm": 0.3260882794857025,
|
|
"learning_rate": 7.607366717782099e-06,
|
|
"loss": 0.1933,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17394818365573883,
|
|
"step": 4205
|
|
},
|
|
{
|
|
"epoch": 4.382716049382716,
|
|
"grad_norm": 0.36598140001296997,
|
|
"learning_rate": 7.5591046502143865e-06,
|
|
"loss": 0.204,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21359078586101532,
|
|
"step": 4210
|
|
},
|
|
{
|
|
"epoch": 4.388888888888889,
|
|
"grad_norm": 0.3709130585193634,
|
|
"learning_rate": 7.510960462588104e-06,
|
|
"loss": 0.1998,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19726939499378204,
|
|
"step": 4215
|
|
},
|
|
{
|
|
"epoch": 4.395061728395062,
|
|
"grad_norm": 0.42972633242607117,
|
|
"learning_rate": 7.462934611078942e-06,
|
|
"loss": 0.1777,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1474202424287796,
|
|
"step": 4220
|
|
},
|
|
{
|
|
"epoch": 4.401234567901234,
|
|
"grad_norm": 0.36691904067993164,
|
|
"learning_rate": 7.415027550741323e-06,
|
|
"loss": 0.1787,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1761273890733719,
|
|
"step": 4225
|
|
},
|
|
{
|
|
"epoch": 4.407407407407407,
|
|
"grad_norm": 0.38260477781295776,
|
|
"learning_rate": 7.367239735504113e-06,
|
|
"loss": 0.1918,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17920207977294922,
|
|
"step": 4230
|
|
},
|
|
{
|
|
"epoch": 4.41358024691358,
|
|
"grad_norm": 0.33410242199897766,
|
|
"learning_rate": 7.31957161816631e-06,
|
|
"loss": 0.1866,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21473069489002228,
|
|
"step": 4235
|
|
},
|
|
{
|
|
"epoch": 4.419753086419753,
|
|
"grad_norm": 0.35049325227737427,
|
|
"learning_rate": 7.27202365039273e-06,
|
|
"loss": 0.1939,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21222616732120514,
|
|
"step": 4240
|
|
},
|
|
{
|
|
"epoch": 4.425925925925926,
|
|
"grad_norm": 0.39021584391593933,
|
|
"learning_rate": 7.224596282709773e-06,
|
|
"loss": 0.1844,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17898672819137573,
|
|
"step": 4245
|
|
},
|
|
{
|
|
"epoch": 4.432098765432099,
|
|
"grad_norm": 0.3583777844905853,
|
|
"learning_rate": 7.177289964501118e-06,
|
|
"loss": 0.1957,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19263823330402374,
|
|
"step": 4250
|
|
},
|
|
{
|
|
"epoch": 4.438271604938271,
|
|
"grad_norm": 0.39881429076194763,
|
|
"learning_rate": 7.1301051440034675e-06,
|
|
"loss": 0.1853,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18660461902618408,
|
|
"step": 4255
|
|
},
|
|
{
|
|
"epoch": 4.444444444444445,
|
|
"grad_norm": 0.39390286803245544,
|
|
"learning_rate": 7.083042268302332e-06,
|
|
"loss": 0.1956,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19438274204730988,
|
|
"step": 4260
|
|
},
|
|
{
|
|
"epoch": 4.450617283950617,
|
|
"grad_norm": 0.3507928252220154,
|
|
"learning_rate": 7.03610178332774e-06,
|
|
"loss": 0.1767,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19298090040683746,
|
|
"step": 4265
|
|
},
|
|
{
|
|
"epoch": 4.45679012345679,
|
|
"grad_norm": 0.34587815403938293,
|
|
"learning_rate": 6.989284133850076e-06,
|
|
"loss": 0.1869,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17915230989456177,
|
|
"step": 4270
|
|
},
|
|
{
|
|
"epoch": 4.462962962962963,
|
|
"grad_norm": 0.3583889603614807,
|
|
"learning_rate": 6.942589763475807e-06,
|
|
"loss": 0.1842,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.16663770377635956,
|
|
"step": 4275
|
|
},
|
|
{
|
|
"epoch": 4.469135802469136,
|
|
"grad_norm": 0.3552998900413513,
|
|
"learning_rate": 6.896019114643324e-06,
|
|
"loss": 0.1719,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18375253677368164,
|
|
"step": 4280
|
|
},
|
|
{
|
|
"epoch": 4.4753086419753085,
|
|
"grad_norm": 0.4269513785839081,
|
|
"learning_rate": 6.84957262861873e-06,
|
|
"loss": 0.1823,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1659822016954422,
|
|
"step": 4285
|
|
},
|
|
{
|
|
"epoch": 4.481481481481482,
|
|
"grad_norm": 0.35391169786453247,
|
|
"learning_rate": 6.803250745491648e-06,
|
|
"loss": 0.2047,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19803209602832794,
|
|
"step": 4290
|
|
},
|
|
{
|
|
"epoch": 4.487654320987654,
|
|
"grad_norm": 0.37884142994880676,
|
|
"learning_rate": 6.75705390417108e-06,
|
|
"loss": 0.181,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17315763235092163,
|
|
"step": 4295
|
|
},
|
|
{
|
|
"epoch": 4.493827160493828,
|
|
"grad_norm": 0.49155905842781067,
|
|
"learning_rate": 6.710982542381215e-06,
|
|
"loss": 0.1802,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17202763259410858,
|
|
"step": 4300
|
|
},
|
|
{
|
|
"epoch": 4.5,
|
|
"grad_norm": 0.39195021986961365,
|
|
"learning_rate": 6.665037096657314e-06,
|
|
"loss": 0.1758,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1503898948431015,
|
|
"step": 4305
|
|
},
|
|
{
|
|
"epoch": 4.506172839506172,
|
|
"grad_norm": 0.3407750725746155,
|
|
"learning_rate": 6.619218002341561e-06,
|
|
"loss": 0.1664,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17060662806034088,
|
|
"step": 4310
|
|
},
|
|
{
|
|
"epoch": 4.512345679012346,
|
|
"grad_norm": 0.42080408334732056,
|
|
"learning_rate": 6.573525693578913e-06,
|
|
"loss": 0.1723,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21822412312030792,
|
|
"step": 4315
|
|
},
|
|
{
|
|
"epoch": 4.518518518518518,
|
|
"grad_norm": 0.44967561960220337,
|
|
"learning_rate": 6.527960603313031e-06,
|
|
"loss": 0.1857,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1986418515443802,
|
|
"step": 4320
|
|
},
|
|
{
|
|
"epoch": 4.5246913580246915,
|
|
"grad_norm": 0.4166305661201477,
|
|
"learning_rate": 6.4825231632821505e-06,
|
|
"loss": 0.1724,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17389072477817535,
|
|
"step": 4325
|
|
},
|
|
{
|
|
"epoch": 4.530864197530864,
|
|
"grad_norm": 0.3899221122264862,
|
|
"learning_rate": 6.4372138040149904e-06,
|
|
"loss": 0.1872,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19381101429462433,
|
|
"step": 4330
|
|
},
|
|
{
|
|
"epoch": 4.537037037037037,
|
|
"grad_norm": 0.3590966463088989,
|
|
"learning_rate": 6.392032954826666e-06,
|
|
"loss": 0.2029,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19857899844646454,
|
|
"step": 4335
|
|
},
|
|
{
|
|
"epoch": 4.54320987654321,
|
|
"grad_norm": 0.3763622045516968,
|
|
"learning_rate": 6.3469810438146775e-06,
|
|
"loss": 0.1975,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21061484515666962,
|
|
"step": 4340
|
|
},
|
|
{
|
|
"epoch": 4.549382716049383,
|
|
"grad_norm": 0.4324042499065399,
|
|
"learning_rate": 6.3020584978547705e-06,
|
|
"loss": 0.1939,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1769997626543045,
|
|
"step": 4345
|
|
},
|
|
{
|
|
"epoch": 4.555555555555555,
|
|
"grad_norm": 0.35509824752807617,
|
|
"learning_rate": 6.2572657425969344e-06,
|
|
"loss": 0.1841,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19601629674434662,
|
|
"step": 4350
|
|
},
|
|
{
|
|
"epoch": 4.561728395061729,
|
|
"grad_norm": 0.4179898500442505,
|
|
"learning_rate": 6.212603202461385e-06,
|
|
"loss": 0.2096,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19075489044189453,
|
|
"step": 4355
|
|
},
|
|
{
|
|
"epoch": 4.567901234567901,
|
|
"grad_norm": 0.4002719521522522,
|
|
"learning_rate": 6.168071300634515e-06,
|
|
"loss": 0.1907,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1573079377412796,
|
|
"step": 4360
|
|
},
|
|
{
|
|
"epoch": 4.574074074074074,
|
|
"grad_norm": 0.3149775266647339,
|
|
"learning_rate": 6.123670459064888e-06,
|
|
"loss": 0.186,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21867139637470245,
|
|
"step": 4365
|
|
},
|
|
{
|
|
"epoch": 4.580246913580247,
|
|
"grad_norm": 0.3219025731086731,
|
|
"learning_rate": 6.079401098459248e-06,
|
|
"loss": 0.2008,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2158464938402176,
|
|
"step": 4370
|
|
},
|
|
{
|
|
"epoch": 4.58641975308642,
|
|
"grad_norm": 0.39307352900505066,
|
|
"learning_rate": 6.035263638278546e-06,
|
|
"loss": 0.1904,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17891430854797363,
|
|
"step": 4375
|
|
},
|
|
{
|
|
"epoch": 4.592592592592593,
|
|
"grad_norm": 0.3912666141986847,
|
|
"learning_rate": 5.991258496733923e-06,
|
|
"loss": 0.1754,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1765187382698059,
|
|
"step": 4380
|
|
},
|
|
{
|
|
"epoch": 4.598765432098766,
|
|
"grad_norm": 0.3784972131252289,
|
|
"learning_rate": 5.947386090782798e-06,
|
|
"loss": 0.1789,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18979035317897797,
|
|
"step": 4385
|
|
},
|
|
{
|
|
"epoch": 4.604938271604938,
|
|
"grad_norm": 0.3676423132419586,
|
|
"learning_rate": 5.903646836124879e-06,
|
|
"loss": 0.1805,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19081111252307892,
|
|
"step": 4390
|
|
},
|
|
{
|
|
"epoch": 4.611111111111111,
|
|
"grad_norm": 0.36174219846725464,
|
|
"learning_rate": 5.86004114719825e-06,
|
|
"loss": 0.191,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20667438209056854,
|
|
"step": 4395
|
|
},
|
|
{
|
|
"epoch": 4.617283950617284,
|
|
"grad_norm": 0.37002086639404297,
|
|
"learning_rate": 5.816569437175432e-06,
|
|
"loss": 0.1773,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18776558339595795,
|
|
"step": 4400
|
|
},
|
|
{
|
|
"epoch": 4.6234567901234565,
|
|
"grad_norm": 0.3361695110797882,
|
|
"learning_rate": 5.7732321179594575e-06,
|
|
"loss": 0.1656,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1597202867269516,
|
|
"step": 4405
|
|
},
|
|
{
|
|
"epoch": 4.62962962962963,
|
|
"grad_norm": 0.3361755907535553,
|
|
"learning_rate": 5.7300296001800004e-06,
|
|
"loss": 0.1743,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1600361168384552,
|
|
"step": 4410
|
|
},
|
|
{
|
|
"epoch": 4.635802469135802,
|
|
"grad_norm": 0.3459812104701996,
|
|
"learning_rate": 5.6869622931894375e-06,
|
|
"loss": 0.1922,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22340242564678192,
|
|
"step": 4415
|
|
},
|
|
{
|
|
"epoch": 4.6419753086419755,
|
|
"grad_norm": 0.7175085544586182,
|
|
"learning_rate": 5.644030605059025e-06,
|
|
"loss": 0.2554,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.33527031540870667,
|
|
"step": 4420
|
|
},
|
|
{
|
|
"epoch": 4.648148148148148,
|
|
"grad_norm": 0.4298430383205414,
|
|
"learning_rate": 5.60123494257498e-06,
|
|
"loss": 0.3735,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.38197556138038635,
|
|
"step": 4425
|
|
},
|
|
{
|
|
"epoch": 4.654320987654321,
|
|
"grad_norm": 0.416728138923645,
|
|
"learning_rate": 5.558575711234661e-06,
|
|
"loss": 0.3552,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.31877490878105164,
|
|
"step": 4430
|
|
},
|
|
{
|
|
"epoch": 4.660493827160494,
|
|
"grad_norm": 0.4450964331626892,
|
|
"learning_rate": 5.516053315242726e-06,
|
|
"loss": 0.3322,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3422302305698395,
|
|
"step": 4435
|
|
},
|
|
{
|
|
"epoch": 4.666666666666667,
|
|
"grad_norm": 0.33260607719421387,
|
|
"learning_rate": 5.473668157507261e-06,
|
|
"loss": 0.3496,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3397694528102875,
|
|
"step": 4440
|
|
},
|
|
{
|
|
"epoch": 4.672839506172839,
|
|
"grad_norm": 0.34154126048088074,
|
|
"learning_rate": 5.431420639636016e-06,
|
|
"loss": 0.3462,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.31662359833717346,
|
|
"step": 4445
|
|
},
|
|
{
|
|
"epoch": 4.679012345679013,
|
|
"grad_norm": 0.291736364364624,
|
|
"learning_rate": 5.38931116193258e-06,
|
|
"loss": 0.3029,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.30226361751556396,
|
|
"step": 4450
|
|
},
|
|
{
|
|
"epoch": 4.685185185185185,
|
|
"grad_norm": 0.3348061144351959,
|
|
"learning_rate": 5.347340123392561e-06,
|
|
"loss": 0.311,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27620983123779297,
|
|
"step": 4455
|
|
},
|
|
{
|
|
"epoch": 4.6913580246913575,
|
|
"grad_norm": 0.28867459297180176,
|
|
"learning_rate": 5.305507921699837e-06,
|
|
"loss": 0.3114,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.31296804547309875,
|
|
"step": 4460
|
|
},
|
|
{
|
|
"epoch": 4.697530864197531,
|
|
"grad_norm": 0.36192354559898376,
|
|
"learning_rate": 5.263814953222803e-06,
|
|
"loss": 0.3143,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.298537015914917,
|
|
"step": 4465
|
|
},
|
|
{
|
|
"epoch": 4.703703703703704,
|
|
"grad_norm": 0.29650819301605225,
|
|
"learning_rate": 5.222261613010564e-06,
|
|
"loss": 0.2868,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27691400051116943,
|
|
"step": 4470
|
|
},
|
|
{
|
|
"epoch": 4.709876543209877,
|
|
"grad_norm": 0.2901798486709595,
|
|
"learning_rate": 5.18084829478922e-06,
|
|
"loss": 0.3003,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.31805118918418884,
|
|
"step": 4475
|
|
},
|
|
{
|
|
"epoch": 4.716049382716049,
|
|
"grad_norm": 0.32907214760780334,
|
|
"learning_rate": 5.139575390958149e-06,
|
|
"loss": 0.3167,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27357545495033264,
|
|
"step": 4480
|
|
},
|
|
{
|
|
"epoch": 4.722222222222222,
|
|
"grad_norm": 0.2987838089466095,
|
|
"learning_rate": 5.098443292586277e-06,
|
|
"loss": 0.3146,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26974496245384216,
|
|
"step": 4485
|
|
},
|
|
{
|
|
"epoch": 4.728395061728395,
|
|
"grad_norm": 0.34803086519241333,
|
|
"learning_rate": 5.057452389408351e-06,
|
|
"loss": 0.2915,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2650125026702881,
|
|
"step": 4490
|
|
},
|
|
{
|
|
"epoch": 4.734567901234568,
|
|
"grad_norm": 0.32165420055389404,
|
|
"learning_rate": 5.01660306982128e-06,
|
|
"loss": 0.2949,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.316560298204422,
|
|
"step": 4495
|
|
},
|
|
{
|
|
"epoch": 4.7407407407407405,
|
|
"grad_norm": 0.3049926459789276,
|
|
"learning_rate": 4.975895720880446e-06,
|
|
"loss": 0.3017,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2991548180580139,
|
|
"step": 4500
|
|
},
|
|
{
|
|
"epoch": 4.746913580246914,
|
|
"grad_norm": 0.38666319847106934,
|
|
"learning_rate": 4.9353307282960106e-06,
|
|
"loss": 0.3243,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2863841652870178,
|
|
"step": 4505
|
|
},
|
|
{
|
|
"epoch": 4.753086419753086,
|
|
"grad_norm": 0.32655414938926697,
|
|
"learning_rate": 4.894908476429308e-06,
|
|
"loss": 0.2799,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26224279403686523,
|
|
"step": 4510
|
|
},
|
|
{
|
|
"epoch": 4.7592592592592595,
|
|
"grad_norm": 0.32183539867401123,
|
|
"learning_rate": 4.854629348289143e-06,
|
|
"loss": 0.2928,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.32195723056793213,
|
|
"step": 4515
|
|
},
|
|
{
|
|
"epoch": 4.765432098765432,
|
|
"grad_norm": 0.31450143456459045,
|
|
"learning_rate": 4.814493725528224e-06,
|
|
"loss": 0.2961,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.31084248423576355,
|
|
"step": 4520
|
|
},
|
|
{
|
|
"epoch": 4.771604938271605,
|
|
"grad_norm": 0.35228440165519714,
|
|
"learning_rate": 4.774501988439511e-06,
|
|
"loss": 0.3246,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3620688021183014,
|
|
"step": 4525
|
|
},
|
|
{
|
|
"epoch": 4.777777777777778,
|
|
"grad_norm": 0.39064261317253113,
|
|
"learning_rate": 4.734654515952608e-06,
|
|
"loss": 0.3086,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.32669317722320557,
|
|
"step": 4530
|
|
},
|
|
{
|
|
"epoch": 4.783950617283951,
|
|
"grad_norm": 0.30717429518699646,
|
|
"learning_rate": 4.6949516856301755e-06,
|
|
"loss": 0.3142,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.29241785407066345,
|
|
"step": 4535
|
|
},
|
|
{
|
|
"epoch": 4.790123456790123,
|
|
"grad_norm": 0.2958734631538391,
|
|
"learning_rate": 4.655393873664402e-06,
|
|
"loss": 0.3081,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.31273916363716125,
|
|
"step": 4540
|
|
},
|
|
{
|
|
"epoch": 4.796296296296296,
|
|
"grad_norm": 0.34213563799858093,
|
|
"learning_rate": 4.615981454873357e-06,
|
|
"loss": 0.3106,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3580350875854492,
|
|
"step": 4545
|
|
},
|
|
{
|
|
"epoch": 4.802469135802469,
|
|
"grad_norm": 0.3018828332424164,
|
|
"learning_rate": 4.5767148026974925e-06,
|
|
"loss": 0.3161,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3478623330593109,
|
|
"step": 4550
|
|
},
|
|
{
|
|
"epoch": 4.8086419753086425,
|
|
"grad_norm": 0.310220867395401,
|
|
"learning_rate": 4.537594289196101e-06,
|
|
"loss": 0.3066,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2754107415676117,
|
|
"step": 4555
|
|
},
|
|
{
|
|
"epoch": 4.814814814814815,
|
|
"grad_norm": 0.28471076488494873,
|
|
"learning_rate": 4.49862028504378e-06,
|
|
"loss": 0.3027,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.29852402210235596,
|
|
"step": 4560
|
|
},
|
|
{
|
|
"epoch": 4.820987654320987,
|
|
"grad_norm": 0.2731614410877228,
|
|
"learning_rate": 4.459793159526911e-06,
|
|
"loss": 0.2922,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.267426073551178,
|
|
"step": 4565
|
|
},
|
|
{
|
|
"epoch": 4.827160493827161,
|
|
"grad_norm": 0.30923593044281006,
|
|
"learning_rate": 4.4211132805401815e-06,
|
|
"loss": 0.3024,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.263751357793808,
|
|
"step": 4570
|
|
},
|
|
{
|
|
"epoch": 4.833333333333333,
|
|
"grad_norm": 0.28524014353752136,
|
|
"learning_rate": 4.382581014583096e-06,
|
|
"loss": 0.2991,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2971295118331909,
|
|
"step": 4575
|
|
},
|
|
{
|
|
"epoch": 4.839506172839506,
|
|
"grad_norm": 0.290445476770401,
|
|
"learning_rate": 4.344196726756476e-06,
|
|
"loss": 0.3014,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3189544379711151,
|
|
"step": 4580
|
|
},
|
|
{
|
|
"epoch": 4.845679012345679,
|
|
"grad_norm": 0.28536275029182434,
|
|
"learning_rate": 4.305960780759035e-06,
|
|
"loss": 0.3113,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.36156919598579407,
|
|
"step": 4585
|
|
},
|
|
{
|
|
"epoch": 4.851851851851852,
|
|
"grad_norm": 0.2934155762195587,
|
|
"learning_rate": 4.267873538883927e-06,
|
|
"loss": 0.2913,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2559260427951813,
|
|
"step": 4590
|
|
},
|
|
{
|
|
"epoch": 4.8580246913580245,
|
|
"grad_norm": 0.2843273878097534,
|
|
"learning_rate": 4.229935362015276e-06,
|
|
"loss": 0.3082,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.293212890625,
|
|
"step": 4595
|
|
},
|
|
{
|
|
"epoch": 4.864197530864198,
|
|
"grad_norm": 0.2935003936290741,
|
|
"learning_rate": 4.1921466096248164e-06,
|
|
"loss": 0.3013,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3130788505077362,
|
|
"step": 4600
|
|
},
|
|
{
|
|
"epoch": 4.87037037037037,
|
|
"grad_norm": 0.26044461131095886,
|
|
"learning_rate": 4.154507639768435e-06,
|
|
"loss": 0.3075,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2758863568305969,
|
|
"step": 4605
|
|
},
|
|
{
|
|
"epoch": 4.8765432098765435,
|
|
"grad_norm": 0.7349086403846741,
|
|
"learning_rate": 4.117018809082816e-06,
|
|
"loss": 0.2806,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.28194278478622437,
|
|
"step": 4610
|
|
},
|
|
{
|
|
"epoch": 4.882716049382716,
|
|
"grad_norm": 0.4354155659675598,
|
|
"learning_rate": 4.079680472782029e-06,
|
|
"loss": 0.2312,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2082822471857071,
|
|
"step": 4615
|
|
},
|
|
{
|
|
"epoch": 4.888888888888889,
|
|
"grad_norm": 0.4463469088077545,
|
|
"learning_rate": 4.042492984654187e-06,
|
|
"loss": 0.222,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22363901138305664,
|
|
"step": 4620
|
|
},
|
|
{
|
|
"epoch": 4.895061728395062,
|
|
"grad_norm": 0.39360952377319336,
|
|
"learning_rate": 4.005456697058099e-06,
|
|
"loss": 0.2457,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23511290550231934,
|
|
"step": 4625
|
|
},
|
|
{
|
|
"epoch": 4.901234567901234,
|
|
"grad_norm": 0.35958102345466614,
|
|
"learning_rate": 3.968571960919892e-06,
|
|
"loss": 0.2507,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2313954085111618,
|
|
"step": 4630
|
|
},
|
|
{
|
|
"epoch": 4.907407407407407,
|
|
"grad_norm": 0.3145143389701843,
|
|
"learning_rate": 3.9318391257297415e-06,
|
|
"loss": 0.2328,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20464976131916046,
|
|
"step": 4635
|
|
},
|
|
{
|
|
"epoch": 4.91358024691358,
|
|
"grad_norm": 0.34790390729904175,
|
|
"learning_rate": 3.895258539538509e-06,
|
|
"loss": 0.2324,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.252556174993515,
|
|
"step": 4640
|
|
},
|
|
{
|
|
"epoch": 4.919753086419753,
|
|
"grad_norm": 0.34201690554618835,
|
|
"learning_rate": 3.858830548954478e-06,
|
|
"loss": 0.2323,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2253480702638626,
|
|
"step": 4645
|
|
},
|
|
{
|
|
"epoch": 4.925925925925926,
|
|
"grad_norm": 0.3274633586406708,
|
|
"learning_rate": 3.822555499140064e-06,
|
|
"loss": 0.2294,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21859495341777802,
|
|
"step": 4650
|
|
},
|
|
{
|
|
"epoch": 4.932098765432099,
|
|
"grad_norm": 0.33171960711479187,
|
|
"learning_rate": 3.786433733808523e-06,
|
|
"loss": 0.2443,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25110527873039246,
|
|
"step": 4655
|
|
},
|
|
{
|
|
"epoch": 4.938271604938271,
|
|
"grad_norm": 0.32472389936447144,
|
|
"learning_rate": 3.7504655952207137e-06,
|
|
"loss": 0.2385,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24646645784378052,
|
|
"step": 4660
|
|
},
|
|
{
|
|
"epoch": 4.944444444444445,
|
|
"grad_norm": 0.3188411593437195,
|
|
"learning_rate": 3.71465142418187e-06,
|
|
"loss": 0.2609,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2809932827949524,
|
|
"step": 4665
|
|
},
|
|
{
|
|
"epoch": 4.950617283950617,
|
|
"grad_norm": 0.34051513671875,
|
|
"learning_rate": 3.6789915600383297e-06,
|
|
"loss": 0.2604,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2358768731355667,
|
|
"step": 4670
|
|
},
|
|
{
|
|
"epoch": 4.95679012345679,
|
|
"grad_norm": 0.33161598443984985,
|
|
"learning_rate": 3.6434863406743428e-06,
|
|
"loss": 0.2581,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2780393660068512,
|
|
"step": 4675
|
|
},
|
|
{
|
|
"epoch": 4.962962962962963,
|
|
"grad_norm": 0.32056570053100586,
|
|
"learning_rate": 3.6081361025088945e-06,
|
|
"loss": 0.2491,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2343013882637024,
|
|
"step": 4680
|
|
},
|
|
{
|
|
"epoch": 4.969135802469136,
|
|
"grad_norm": 0.3458598554134369,
|
|
"learning_rate": 3.572941180492471e-06,
|
|
"loss": 0.2488,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2514094412326813,
|
|
"step": 4685
|
|
},
|
|
{
|
|
"epoch": 4.9753086419753085,
|
|
"grad_norm": 0.35199108719825745,
|
|
"learning_rate": 3.5379019081039023e-06,
|
|
"loss": 0.2284,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20398132503032684,
|
|
"step": 4690
|
|
},
|
|
{
|
|
"epoch": 4.981481481481482,
|
|
"grad_norm": 0.32564088702201843,
|
|
"learning_rate": 3.5030186173472224e-06,
|
|
"loss": 0.2173,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20684821903705597,
|
|
"step": 4695
|
|
},
|
|
{
|
|
"epoch": 4.987654320987654,
|
|
"grad_norm": 0.34538009762763977,
|
|
"learning_rate": 3.468291638748511e-06,
|
|
"loss": 0.2193,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21137161552906036,
|
|
"step": 4700
|
|
},
|
|
{
|
|
"epoch": 4.993827160493828,
|
|
"grad_norm": 0.4077824056148529,
|
|
"learning_rate": 3.433721301352739e-06,
|
|
"loss": 0.2223,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23706261813640594,
|
|
"step": 4705
|
|
},
|
|
{
|
|
"epoch": 5.0,
|
|
"grad_norm": 0.5916141271591187,
|
|
"learning_rate": 3.39930793272069e-06,
|
|
"loss": 0.2279,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21526634693145752,
|
|
"step": 4710
|
|
},
|
|
{
|
|
"epoch": 5.006172839506172,
|
|
"grad_norm": 0.45316532254219055,
|
|
"learning_rate": 3.365051858925834e-06,
|
|
"loss": 0.1924,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19291605055332184,
|
|
"step": 4715
|
|
},
|
|
{
|
|
"epoch": 5.012345679012346,
|
|
"grad_norm": 0.387811541557312,
|
|
"learning_rate": 3.3309534045512316e-06,
|
|
"loss": 0.1886,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17393136024475098,
|
|
"step": 4720
|
|
},
|
|
{
|
|
"epoch": 5.018518518518518,
|
|
"grad_norm": 0.4478646516799927,
|
|
"learning_rate": 3.2970128926864797e-06,
|
|
"loss": 0.1826,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18674321472644806,
|
|
"step": 4725
|
|
},
|
|
{
|
|
"epoch": 5.0246913580246915,
|
|
"grad_norm": 0.3496834933757782,
|
|
"learning_rate": 3.263230644924631e-06,
|
|
"loss": 0.1804,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1933426409959793,
|
|
"step": 4730
|
|
},
|
|
{
|
|
"epoch": 5.030864197530864,
|
|
"grad_norm": 0.3450537323951721,
|
|
"learning_rate": 3.2296069813591544e-06,
|
|
"loss": 0.1848,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2023792713880539,
|
|
"step": 4735
|
|
},
|
|
{
|
|
"epoch": 5.037037037037037,
|
|
"grad_norm": 0.3337896168231964,
|
|
"learning_rate": 3.1961422205809177e-06,
|
|
"loss": 0.1769,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19460026919841766,
|
|
"step": 4740
|
|
},
|
|
{
|
|
"epoch": 5.04320987654321,
|
|
"grad_norm": 0.3590703308582306,
|
|
"learning_rate": 3.16283667967513e-06,
|
|
"loss": 0.1855,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20627611875534058,
|
|
"step": 4745
|
|
},
|
|
{
|
|
"epoch": 5.049382716049383,
|
|
"grad_norm": 0.33701422810554504,
|
|
"learning_rate": 3.1296906742183707e-06,
|
|
"loss": 0.1834,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1561555713415146,
|
|
"step": 4750
|
|
},
|
|
{
|
|
"epoch": 5.055555555555555,
|
|
"grad_norm": 0.4371582865715027,
|
|
"learning_rate": 3.096704518275595e-06,
|
|
"loss": 0.1811,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17703194916248322,
|
|
"step": 4755
|
|
},
|
|
{
|
|
"epoch": 5.061728395061729,
|
|
"grad_norm": 0.5024053454399109,
|
|
"learning_rate": 3.0638785243971505e-06,
|
|
"loss": 0.1721,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1945503205060959,
|
|
"step": 4760
|
|
},
|
|
{
|
|
"epoch": 5.067901234567901,
|
|
"grad_norm": 0.4448404014110565,
|
|
"learning_rate": 3.031213003615807e-06,
|
|
"loss": 0.1959,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19550740718841553,
|
|
"step": 4765
|
|
},
|
|
{
|
|
"epoch": 5.074074074074074,
|
|
"grad_norm": 0.40806570649147034,
|
|
"learning_rate": 2.9987082654438304e-06,
|
|
"loss": 0.1768,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17459459602832794,
|
|
"step": 4770
|
|
},
|
|
{
|
|
"epoch": 5.080246913580247,
|
|
"grad_norm": 0.4173873960971832,
|
|
"learning_rate": 2.9663646178700434e-06,
|
|
"loss": 0.1927,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21065036952495575,
|
|
"step": 4775
|
|
},
|
|
{
|
|
"epoch": 5.08641975308642,
|
|
"grad_norm": 0.3855539858341217,
|
|
"learning_rate": 2.934182367356888e-06,
|
|
"loss": 0.1934,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2148648053407669,
|
|
"step": 4780
|
|
},
|
|
{
|
|
"epoch": 5.092592592592593,
|
|
"grad_norm": 0.37203601002693176,
|
|
"learning_rate": 2.9021618188375476e-06,
|
|
"loss": 0.1874,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1732834130525589,
|
|
"step": 4785
|
|
},
|
|
{
|
|
"epoch": 5.098765432098766,
|
|
"grad_norm": 0.4391461908817291,
|
|
"learning_rate": 2.8703032757130556e-06,
|
|
"loss": 0.2027,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18861384689807892,
|
|
"step": 4790
|
|
},
|
|
{
|
|
"epoch": 5.104938271604938,
|
|
"grad_norm": 0.41993486881256104,
|
|
"learning_rate": 2.838607039849388e-06,
|
|
"loss": 0.182,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1985810548067093,
|
|
"step": 4795
|
|
},
|
|
{
|
|
"epoch": 5.111111111111111,
|
|
"grad_norm": 0.38813507556915283,
|
|
"learning_rate": 2.80707341157465e-06,
|
|
"loss": 0.1804,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1734999418258667,
|
|
"step": 4800
|
|
},
|
|
{
|
|
"epoch": 5.117283950617284,
|
|
"grad_norm": 0.3901495933532715,
|
|
"learning_rate": 2.7757026896762053e-06,
|
|
"loss": 0.1867,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18226118385791779,
|
|
"step": 4805
|
|
},
|
|
{
|
|
"epoch": 5.1234567901234565,
|
|
"grad_norm": 0.36712223291397095,
|
|
"learning_rate": 2.744495171397832e-06,
|
|
"loss": 0.178,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1825934648513794,
|
|
"step": 4810
|
|
},
|
|
{
|
|
"epoch": 5.12962962962963,
|
|
"grad_norm": 0.42331355810165405,
|
|
"learning_rate": 2.7134511524369324e-06,
|
|
"loss": 0.1809,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17737114429473877,
|
|
"step": 4815
|
|
},
|
|
{
|
|
"epoch": 5.135802469135802,
|
|
"grad_norm": 0.40524235367774963,
|
|
"learning_rate": 2.6825709269417212e-06,
|
|
"loss": 0.1894,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17012222111225128,
|
|
"step": 4820
|
|
},
|
|
{
|
|
"epoch": 5.1419753086419755,
|
|
"grad_norm": 0.3814069926738739,
|
|
"learning_rate": 2.6518547875084412e-06,
|
|
"loss": 0.1961,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19045425951480865,
|
|
"step": 4825
|
|
},
|
|
{
|
|
"epoch": 5.148148148148148,
|
|
"grad_norm": 0.3933735191822052,
|
|
"learning_rate": 2.621303025178574e-06,
|
|
"loss": 0.1833,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17836368083953857,
|
|
"step": 4830
|
|
},
|
|
{
|
|
"epoch": 5.154320987654321,
|
|
"grad_norm": 0.3901108205318451,
|
|
"learning_rate": 2.590915929436113e-06,
|
|
"loss": 0.1787,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18447498977184296,
|
|
"step": 4835
|
|
},
|
|
{
|
|
"epoch": 5.160493827160494,
|
|
"grad_norm": 0.3753919005393982,
|
|
"learning_rate": 2.5606937882047823e-06,
|
|
"loss": 0.1964,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17433185875415802,
|
|
"step": 4840
|
|
},
|
|
{
|
|
"epoch": 5.166666666666667,
|
|
"grad_norm": 0.47514718770980835,
|
|
"learning_rate": 2.5306368878453504e-06,
|
|
"loss": 0.1882,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17399097979068756,
|
|
"step": 4845
|
|
},
|
|
{
|
|
"epoch": 5.172839506172839,
|
|
"grad_norm": 0.3544893264770508,
|
|
"learning_rate": 2.5007455131528868e-06,
|
|
"loss": 0.1908,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20257844030857086,
|
|
"step": 4850
|
|
},
|
|
{
|
|
"epoch": 5.179012345679013,
|
|
"grad_norm": 0.4351579248905182,
|
|
"learning_rate": 2.4710199473540655e-06,
|
|
"loss": 0.1843,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19679565727710724,
|
|
"step": 4855
|
|
},
|
|
{
|
|
"epoch": 5.185185185185185,
|
|
"grad_norm": 0.365847647190094,
|
|
"learning_rate": 2.4414604721044976e-06,
|
|
"loss": 0.1755,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1921578198671341,
|
|
"step": 4860
|
|
},
|
|
{
|
|
"epoch": 5.191358024691358,
|
|
"grad_norm": 0.42758217453956604,
|
|
"learning_rate": 2.4120673674860593e-06,
|
|
"loss": 0.1862,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18506528437137604,
|
|
"step": 4865
|
|
},
|
|
{
|
|
"epoch": 5.197530864197531,
|
|
"grad_norm": 0.3256032168865204,
|
|
"learning_rate": 2.3828409120042183e-06,
|
|
"loss": 0.1872,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.14872629940509796,
|
|
"step": 4870
|
|
},
|
|
{
|
|
"epoch": 5.203703703703703,
|
|
"grad_norm": 0.4426772892475128,
|
|
"learning_rate": 2.353781382585407e-06,
|
|
"loss": 0.1714,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.16431821882724762,
|
|
"step": 4875
|
|
},
|
|
{
|
|
"epoch": 5.209876543209877,
|
|
"grad_norm": 0.3068554103374481,
|
|
"learning_rate": 2.324889054574422e-06,
|
|
"loss": 0.1743,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18106716871261597,
|
|
"step": 4880
|
|
},
|
|
{
|
|
"epoch": 5.216049382716049,
|
|
"grad_norm": 0.35740476846694946,
|
|
"learning_rate": 2.2961642017317744e-06,
|
|
"loss": 0.1861,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19754938781261444,
|
|
"step": 4885
|
|
},
|
|
{
|
|
"epoch": 5.222222222222222,
|
|
"grad_norm": 0.39344117045402527,
|
|
"learning_rate": 2.2676070962311083e-06,
|
|
"loss": 0.1776,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19656796753406525,
|
|
"step": 4890
|
|
},
|
|
{
|
|
"epoch": 5.228395061728395,
|
|
"grad_norm": 0.43272775411605835,
|
|
"learning_rate": 2.239218008656661e-06,
|
|
"loss": 0.192,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2095431685447693,
|
|
"step": 4895
|
|
},
|
|
{
|
|
"epoch": 5.234567901234568,
|
|
"grad_norm": 0.413025826215744,
|
|
"learning_rate": 2.2109972080006337e-06,
|
|
"loss": 0.2003,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21261107921600342,
|
|
"step": 4900
|
|
},
|
|
{
|
|
"epoch": 5.2407407407407405,
|
|
"grad_norm": 0.4527886211872101,
|
|
"learning_rate": 2.182944961660689e-06,
|
|
"loss": 0.1944,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1933744102716446,
|
|
"step": 4905
|
|
},
|
|
{
|
|
"epoch": 5.246913580246914,
|
|
"grad_norm": 0.41296616196632385,
|
|
"learning_rate": 2.155061535437404e-06,
|
|
"loss": 0.195,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21155185997486115,
|
|
"step": 4910
|
|
},
|
|
{
|
|
"epoch": 5.253086419753086,
|
|
"grad_norm": 0.37184858322143555,
|
|
"learning_rate": 2.127347193531757e-06,
|
|
"loss": 0.1965,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1711912900209427,
|
|
"step": 4915
|
|
},
|
|
{
|
|
"epoch": 5.2592592592592595,
|
|
"grad_norm": 0.38625532388687134,
|
|
"learning_rate": 2.099802198542604e-06,
|
|
"loss": 0.1828,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17634202539920807,
|
|
"step": 4920
|
|
},
|
|
{
|
|
"epoch": 5.265432098765432,
|
|
"grad_norm": 0.3992258906364441,
|
|
"learning_rate": 2.072426811464221e-06,
|
|
"loss": 0.1756,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19168704748153687,
|
|
"step": 4925
|
|
},
|
|
{
|
|
"epoch": 5.271604938271605,
|
|
"grad_norm": 0.38087591528892517,
|
|
"learning_rate": 2.0452212916838056e-06,
|
|
"loss": 0.1841,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19795377552509308,
|
|
"step": 4930
|
|
},
|
|
{
|
|
"epoch": 5.277777777777778,
|
|
"grad_norm": 0.4447610080242157,
|
|
"learning_rate": 2.0181858969790347e-06,
|
|
"loss": 0.1798,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.15701837837696075,
|
|
"step": 4935
|
|
},
|
|
{
|
|
"epoch": 5.283950617283951,
|
|
"grad_norm": 0.37163618206977844,
|
|
"learning_rate": 1.9913208835156173e-06,
|
|
"loss": 0.1766,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17546884715557098,
|
|
"step": 4940
|
|
},
|
|
{
|
|
"epoch": 5.290123456790123,
|
|
"grad_norm": 0.4653220474720001,
|
|
"learning_rate": 1.9646265058448553e-06,
|
|
"loss": 0.1905,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19082653522491455,
|
|
"step": 4945
|
|
},
|
|
{
|
|
"epoch": 5.296296296296296,
|
|
"grad_norm": 0.42952677607536316,
|
|
"learning_rate": 1.9381030169012626e-06,
|
|
"loss": 0.1825,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1877564638853073,
|
|
"step": 4950
|
|
},
|
|
{
|
|
"epoch": 5.302469135802469,
|
|
"grad_norm": 0.39721226692199707,
|
|
"learning_rate": 1.9117506680001273e-06,
|
|
"loss": 0.1847,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17412500083446503,
|
|
"step": 4955
|
|
},
|
|
{
|
|
"epoch": 5.308641975308642,
|
|
"grad_norm": 0.39301198720932007,
|
|
"learning_rate": 1.8855697088351688e-06,
|
|
"loss": 0.1609,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17999877035617828,
|
|
"step": 4960
|
|
},
|
|
{
|
|
"epoch": 5.314814814814815,
|
|
"grad_norm": 0.4005042612552643,
|
|
"learning_rate": 1.8595603874761426e-06,
|
|
"loss": 0.1775,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17548508942127228,
|
|
"step": 4965
|
|
},
|
|
{
|
|
"epoch": 5.320987654320987,
|
|
"grad_norm": 0.3552158772945404,
|
|
"learning_rate": 1.8337229503665077e-06,
|
|
"loss": 0.198,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19586153328418732,
|
|
"step": 4970
|
|
},
|
|
{
|
|
"epoch": 5.327160493827161,
|
|
"grad_norm": 0.48515811562538147,
|
|
"learning_rate": 1.8080576423210928e-06,
|
|
"loss": 0.1698,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1403854340314865,
|
|
"step": 4975
|
|
},
|
|
{
|
|
"epoch": 5.333333333333333,
|
|
"grad_norm": 0.4558883011341095,
|
|
"learning_rate": 1.7825647065237528e-06,
|
|
"loss": 0.1738,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18035392463207245,
|
|
"step": 4980
|
|
},
|
|
{
|
|
"epoch": 5.339506172839506,
|
|
"grad_norm": 0.38746917247772217,
|
|
"learning_rate": 1.757244384525094e-06,
|
|
"loss": 0.1842,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19885873794555664,
|
|
"step": 4985
|
|
},
|
|
{
|
|
"epoch": 5.345679012345679,
|
|
"grad_norm": 0.3846323788166046,
|
|
"learning_rate": 1.7320969162401757e-06,
|
|
"loss": 0.2048,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17779409885406494,
|
|
"step": 4990
|
|
},
|
|
{
|
|
"epoch": 5.351851851851852,
|
|
"grad_norm": 0.38314008712768555,
|
|
"learning_rate": 1.7071225399462244e-06,
|
|
"loss": 0.1792,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1792631596326828,
|
|
"step": 4995
|
|
},
|
|
{
|
|
"epoch": 5.3580246913580245,
|
|
"grad_norm": 0.41071653366088867,
|
|
"learning_rate": 1.6823214922803832e-06,
|
|
"loss": 0.1816,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18592208623886108,
|
|
"step": 5000
|
|
},
|
|
{
|
|
"epoch": 5.364197530864198,
|
|
"grad_norm": 0.3418760597705841,
|
|
"learning_rate": 1.6576940082374893e-06,
|
|
"loss": 0.1711,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.15865923464298248,
|
|
"step": 5005
|
|
},
|
|
{
|
|
"epoch": 5.37037037037037,
|
|
"grad_norm": 0.36145251989364624,
|
|
"learning_rate": 1.6332403211678173e-06,
|
|
"loss": 0.187,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18976199626922607,
|
|
"step": 5010
|
|
},
|
|
{
|
|
"epoch": 5.3765432098765435,
|
|
"grad_norm": 0.336276113986969,
|
|
"learning_rate": 1.6089606627748765e-06,
|
|
"loss": 0.1857,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1671270728111267,
|
|
"step": 5015
|
|
},
|
|
{
|
|
"epoch": 5.382716049382716,
|
|
"grad_norm": 0.3550458550453186,
|
|
"learning_rate": 1.5848552631132342e-06,
|
|
"loss": 0.1965,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20677323639392853,
|
|
"step": 5020
|
|
},
|
|
{
|
|
"epoch": 5.388888888888889,
|
|
"grad_norm": 0.34708911180496216,
|
|
"learning_rate": 1.5609243505863147e-06,
|
|
"loss": 0.1928,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19033074378967285,
|
|
"step": 5025
|
|
},
|
|
{
|
|
"epoch": 5.395061728395062,
|
|
"grad_norm": 0.35199880599975586,
|
|
"learning_rate": 1.5371681519442372e-06,
|
|
"loss": 0.171,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.140380859375,
|
|
"step": 5030
|
|
},
|
|
{
|
|
"epoch": 5.401234567901234,
|
|
"grad_norm": 0.36190831661224365,
|
|
"learning_rate": 1.5135868922816798e-06,
|
|
"loss": 0.1717,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.16963599622249603,
|
|
"step": 5035
|
|
},
|
|
{
|
|
"epoch": 5.407407407407407,
|
|
"grad_norm": 0.3570379614830017,
|
|
"learning_rate": 1.4901807950357427e-06,
|
|
"loss": 0.1851,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17288438975811005,
|
|
"step": 5040
|
|
},
|
|
{
|
|
"epoch": 5.41358024691358,
|
|
"grad_norm": 0.3677111268043518,
|
|
"learning_rate": 1.4669500819838135e-06,
|
|
"loss": 0.18,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2078167051076889,
|
|
"step": 5045
|
|
},
|
|
{
|
|
"epoch": 5.419753086419753,
|
|
"grad_norm": 0.33986392617225647,
|
|
"learning_rate": 1.443894973241493e-06,
|
|
"loss": 0.1875,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20577071607112885,
|
|
"step": 5050
|
|
},
|
|
{
|
|
"epoch": 5.425925925925926,
|
|
"grad_norm": 0.40088096261024475,
|
|
"learning_rate": 1.4210156872604875e-06,
|
|
"loss": 0.1775,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.171505406498909,
|
|
"step": 5055
|
|
},
|
|
{
|
|
"epoch": 5.432098765432099,
|
|
"grad_norm": 0.35825440287590027,
|
|
"learning_rate": 1.3983124408265547e-06,
|
|
"loss": 0.1887,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1865174025297165,
|
|
"step": 5060
|
|
},
|
|
{
|
|
"epoch": 5.438271604938271,
|
|
"grad_norm": 0.37052100896835327,
|
|
"learning_rate": 1.3757854490574429e-06,
|
|
"loss": 0.1787,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18062949180603027,
|
|
"step": 5065
|
|
},
|
|
{
|
|
"epoch": 5.444444444444445,
|
|
"grad_norm": 0.3871017098426819,
|
|
"learning_rate": 1.3534349254008427e-06,
|
|
"loss": 0.188,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18634474277496338,
|
|
"step": 5070
|
|
},
|
|
{
|
|
"epoch": 5.450617283950617,
|
|
"grad_norm": 0.34788069128990173,
|
|
"learning_rate": 1.3312610816323868e-06,
|
|
"loss": 0.1696,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18630552291870117,
|
|
"step": 5075
|
|
},
|
|
{
|
|
"epoch": 5.45679012345679,
|
|
"grad_norm": 0.44331222772598267,
|
|
"learning_rate": 1.3092641278536266e-06,
|
|
"loss": 0.1802,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1717095822095871,
|
|
"step": 5080
|
|
},
|
|
{
|
|
"epoch": 5.462962962962963,
|
|
"grad_norm": 0.32872918248176575,
|
|
"learning_rate": 1.2874442724900437e-06,
|
|
"loss": 0.1771,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.15947429835796356,
|
|
"step": 5085
|
|
},
|
|
{
|
|
"epoch": 5.469135802469136,
|
|
"grad_norm": 0.34846848249435425,
|
|
"learning_rate": 1.265801722289075e-06,
|
|
"loss": 0.1651,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1762632131576538,
|
|
"step": 5090
|
|
},
|
|
{
|
|
"epoch": 5.4753086419753085,
|
|
"grad_norm": 0.5227166414260864,
|
|
"learning_rate": 1.2443366823181635e-06,
|
|
"loss": 0.1754,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1593904048204422,
|
|
"step": 5095
|
|
},
|
|
{
|
|
"epoch": 5.481481481481482,
|
|
"grad_norm": 0.3328639566898346,
|
|
"learning_rate": 1.2230493559628064e-06,
|
|
"loss": 0.1979,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19116179645061493,
|
|
"step": 5100
|
|
},
|
|
{
|
|
"epoch": 5.487654320987654,
|
|
"grad_norm": 0.3921387493610382,
|
|
"learning_rate": 1.2019399449246217e-06,
|
|
"loss": 0.1737,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.16674624383449554,
|
|
"step": 5105
|
|
},
|
|
{
|
|
"epoch": 5.493827160493828,
|
|
"grad_norm": 0.3921066224575043,
|
|
"learning_rate": 1.1810086492194505e-06,
|
|
"loss": 0.1737,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1660056859254837,
|
|
"step": 5110
|
|
},
|
|
{
|
|
"epoch": 5.5,
|
|
"grad_norm": 0.3634832799434662,
|
|
"learning_rate": 1.1602556671754584e-06,
|
|
"loss": 0.1692,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.14344264566898346,
|
|
"step": 5115
|
|
},
|
|
{
|
|
"epoch": 5.506172839506172,
|
|
"grad_norm": 0.32945385575294495,
|
|
"learning_rate": 1.1396811954312458e-06,
|
|
"loss": 0.1596,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.16342221200466156,
|
|
"step": 5120
|
|
},
|
|
{
|
|
"epoch": 5.512345679012346,
|
|
"grad_norm": 0.41761714220046997,
|
|
"learning_rate": 1.1192854289340005e-06,
|
|
"loss": 0.1652,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2108963578939438,
|
|
"step": 5125
|
|
},
|
|
{
|
|
"epoch": 5.518518518518518,
|
|
"grad_norm": 0.41398733854293823,
|
|
"learning_rate": 1.0990685609376395e-06,
|
|
"loss": 0.1787,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1924719363451004,
|
|
"step": 5130
|
|
},
|
|
{
|
|
"epoch": 5.5246913580246915,
|
|
"grad_norm": 0.42360734939575195,
|
|
"learning_rate": 1.0790307830009806e-06,
|
|
"loss": 0.1656,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.16619162261486053,
|
|
"step": 5135
|
|
},
|
|
{
|
|
"epoch": 5.530864197530864,
|
|
"grad_norm": 0.384074866771698,
|
|
"learning_rate": 1.059172284985932e-06,
|
|
"loss": 0.1799,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18611355125904083,
|
|
"step": 5140
|
|
},
|
|
{
|
|
"epoch": 5.537037037037037,
|
|
"grad_norm": 0.3673054873943329,
|
|
"learning_rate": 1.039493255055679e-06,
|
|
"loss": 0.1958,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1913852095603943,
|
|
"step": 5145
|
|
},
|
|
{
|
|
"epoch": 5.54320987654321,
|
|
"grad_norm": 0.3741569519042969,
|
|
"learning_rate": 1.0199938796729248e-06,
|
|
"loss": 0.1908,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20364321768283844,
|
|
"step": 5150
|
|
},
|
|
{
|
|
"epoch": 5.549382716049383,
|
|
"grad_norm": 0.4210781455039978,
|
|
"learning_rate": 1.000674343598098e-06,
|
|
"loss": 0.1867,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17114388942718506,
|
|
"step": 5155
|
|
},
|
|
{
|
|
"epoch": 5.555555555555555,
|
|
"grad_norm": 0.34998446702957153,
|
|
"learning_rate": 9.815348298876247e-07,
|
|
"loss": 0.177,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18901754915714264,
|
|
"step": 5160
|
|
},
|
|
{
|
|
"epoch": 5.561728395061729,
|
|
"grad_norm": 0.38847973942756653,
|
|
"learning_rate": 9.625755198921793e-07,
|
|
"loss": 0.2024,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18445201218128204,
|
|
"step": 5165
|
|
},
|
|
{
|
|
"epoch": 5.567901234567901,
|
|
"grad_norm": 0.40402284264564514,
|
|
"learning_rate": 9.437965932549642e-07,
|
|
"loss": 0.1834,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.151123046875,
|
|
"step": 5170
|
|
},
|
|
{
|
|
"epoch": 5.574074074074074,
|
|
"grad_norm": 0.32070666551589966,
|
|
"learning_rate": 9.251982279100246e-07,
|
|
"loss": 0.1788,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21087205410003662,
|
|
"step": 5175
|
|
},
|
|
{
|
|
"epoch": 5.580246913580247,
|
|
"grad_norm": 0.3237183690071106,
|
|
"learning_rate": 9.067806000805435e-07,
|
|
"loss": 0.1937,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20807433128356934,
|
|
"step": 5180
|
|
},
|
|
{
|
|
"epoch": 5.58641975308642,
|
|
"grad_norm": 0.399247944355011,
|
|
"learning_rate": 8.885438842771843e-07,
|
|
"loss": 0.1831,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17161689698696136,
|
|
"step": 5185
|
|
},
|
|
{
|
|
"epoch": 5.592592592592593,
|
|
"grad_norm": 0.35707828402519226,
|
|
"learning_rate": 8.704882532964376e-07,
|
|
"loss": 0.1689,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17037604749202728,
|
|
"step": 5190
|
|
},
|
|
{
|
|
"epoch": 5.598765432098766,
|
|
"grad_norm": 0.39152827858924866,
|
|
"learning_rate": 8.526138782189641e-07,
|
|
"loss": 0.1722,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18235178291797638,
|
|
"step": 5195
|
|
},
|
|
{
|
|
"epoch": 5.604938271604938,
|
|
"grad_norm": 0.3638923466205597,
|
|
"learning_rate": 8.349209284080073e-07,
|
|
"loss": 0.1741,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1845027208328247,
|
|
"step": 5200
|
|
},
|
|
{
|
|
"epoch": 5.611111111111111,
|
|
"grad_norm": 0.43305104970932007,
|
|
"learning_rate": 8.174095715077612e-07,
|
|
"loss": 0.1842,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1983155757188797,
|
|
"step": 5205
|
|
},
|
|
{
|
|
"epoch": 5.617283950617284,
|
|
"grad_norm": 0.37263602018356323,
|
|
"learning_rate": 8.000799734417897e-07,
|
|
"loss": 0.1704,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18130941689014435,
|
|
"step": 5210
|
|
},
|
|
{
|
|
"epoch": 5.6234567901234565,
|
|
"grad_norm": 0.3466659188270569,
|
|
"learning_rate": 7.829322984114562e-07,
|
|
"loss": 0.1593,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1533888876438141,
|
|
"step": 5215
|
|
},
|
|
{
|
|
"epoch": 5.62962962962963,
|
|
"grad_norm": 0.346358060836792,
|
|
"learning_rate": 7.659667088943746e-07,
|
|
"loss": 0.168,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1531248837709427,
|
|
"step": 5220
|
|
},
|
|
{
|
|
"epoch": 5.635802469135802,
|
|
"grad_norm": 0.4060136079788208,
|
|
"learning_rate": 7.491833656428582e-07,
|
|
"loss": 0.1855,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21525005996227264,
|
|
"step": 5225
|
|
},
|
|
{
|
|
"epoch": 5.6419753086419755,
|
|
"grad_norm": 0.718070387840271,
|
|
"learning_rate": 7.325824276823934e-07,
|
|
"loss": 0.2485,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3281306028366089,
|
|
"step": 5230
|
|
},
|
|
{
|
|
"epoch": 5.648148148148148,
|
|
"grad_norm": 0.4332717955112457,
|
|
"learning_rate": 7.161640523101531e-07,
|
|
"loss": 0.3659,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.37373587489128113,
|
|
"step": 5235
|
|
},
|
|
{
|
|
"epoch": 5.654320987654321,
|
|
"grad_norm": 0.40375369787216187,
|
|
"learning_rate": 6.999283950934876e-07,
|
|
"loss": 0.3474,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.31121599674224854,
|
|
"step": 5240
|
|
},
|
|
{
|
|
"epoch": 5.660493827160494,
|
|
"grad_norm": 0.40947169065475464,
|
|
"learning_rate": 6.838756098684541e-07,
|
|
"loss": 0.3244,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.33486852049827576,
|
|
"step": 5245
|
|
},
|
|
{
|
|
"epoch": 5.666666666666667,
|
|
"grad_norm": 0.33859407901763916,
|
|
"learning_rate": 6.680058487383645e-07,
|
|
"loss": 0.3418,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.33128443360328674,
|
|
"step": 5250
|
|
},
|
|
{
|
|
"epoch": 5.672839506172839,
|
|
"grad_norm": 0.33771681785583496,
|
|
"learning_rate": 6.523192620723428e-07,
|
|
"loss": 0.3382,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3087645471096039,
|
|
"step": 5255
|
|
},
|
|
{
|
|
"epoch": 5.679012345679013,
|
|
"grad_norm": 0.28529274463653564,
|
|
"learning_rate": 6.368159985038968e-07,
|
|
"loss": 0.2953,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.29404518008232117,
|
|
"step": 5260
|
|
},
|
|
{
|
|
"epoch": 5.685185185185185,
|
|
"grad_norm": 0.313550740480423,
|
|
"learning_rate": 6.214962049295103e-07,
|
|
"loss": 0.3031,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26927483081817627,
|
|
"step": 5265
|
|
},
|
|
{
|
|
"epoch": 5.6913580246913575,
|
|
"grad_norm": 0.28332099318504333,
|
|
"learning_rate": 6.063600265072534e-07,
|
|
"loss": 0.3036,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3055051565170288,
|
|
"step": 5270
|
|
},
|
|
{
|
|
"epoch": 5.697530864197531,
|
|
"grad_norm": 0.32281580567359924,
|
|
"learning_rate": 5.914076066554098e-07,
|
|
"loss": 0.3068,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2920275032520294,
|
|
"step": 5275
|
|
},
|
|
{
|
|
"epoch": 5.703703703703704,
|
|
"grad_norm": 0.29317736625671387,
|
|
"learning_rate": 5.766390870511118e-07,
|
|
"loss": 0.2797,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2694184482097626,
|
|
"step": 5280
|
|
},
|
|
{
|
|
"epoch": 5.709876543209877,
|
|
"grad_norm": 0.28393101692199707,
|
|
"learning_rate": 5.62054607628999e-07,
|
|
"loss": 0.2928,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.31008216738700867,
|
|
"step": 5285
|
|
},
|
|
{
|
|
"epoch": 5.716049382716049,
|
|
"grad_norm": 0.3016740083694458,
|
|
"learning_rate": 5.476543065798967e-07,
|
|
"loss": 0.3092,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.266850084066391,
|
|
"step": 5290
|
|
},
|
|
{
|
|
"epoch": 5.722222222222222,
|
|
"grad_norm": 0.2915099561214447,
|
|
"learning_rate": 5.334383203494975e-07,
|
|
"loss": 0.3069,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26300397515296936,
|
|
"step": 5295
|
|
},
|
|
{
|
|
"epoch": 5.728395061728395,
|
|
"grad_norm": 0.3290761411190033,
|
|
"learning_rate": 5.194067836370886e-07,
|
|
"loss": 0.2841,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.25818413496017456,
|
|
"step": 5300
|
|
},
|
|
{
|
|
"epoch": 5.734567901234568,
|
|
"grad_norm": 0.3215181827545166,
|
|
"learning_rate": 5.055598293942443e-07,
|
|
"loss": 0.2864,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.308118611574173,
|
|
"step": 5305
|
|
},
|
|
{
|
|
"epoch": 5.7407407407407405,
|
|
"grad_norm": 0.31151700019836426,
|
|
"learning_rate": 4.918975888235955e-07,
|
|
"loss": 0.2941,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.291477769613266,
|
|
"step": 5310
|
|
},
|
|
{
|
|
"epoch": 5.746913580246914,
|
|
"grad_norm": 0.3270668387413025,
|
|
"learning_rate": 4.784201913775732e-07,
|
|
"loss": 0.3171,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2795468270778656,
|
|
"step": 5315
|
|
},
|
|
{
|
|
"epoch": 5.753086419753086,
|
|
"grad_norm": 0.45036497712135315,
|
|
"learning_rate": 4.651277647571761e-07,
|
|
"loss": 0.273,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2557394206523895,
|
|
"step": 5320
|
|
},
|
|
{
|
|
"epoch": 5.7592592592592595,
|
|
"grad_norm": 0.31800851225852966,
|
|
"learning_rate": 4.5202043491077374e-07,
|
|
"loss": 0.2854,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.31407395005226135,
|
|
"step": 5325
|
|
},
|
|
{
|
|
"epoch": 5.765432098765432,
|
|
"grad_norm": 0.30804312229156494,
|
|
"learning_rate": 4.390983260329118e-07,
|
|
"loss": 0.2891,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.30289506912231445,
|
|
"step": 5330
|
|
},
|
|
{
|
|
"epoch": 5.771604938271605,
|
|
"grad_norm": 0.3516952395439148,
|
|
"learning_rate": 4.2636156056311773e-07,
|
|
"loss": 0.317,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3543910086154938,
|
|
"step": 5335
|
|
},
|
|
{
|
|
"epoch": 5.777777777777778,
|
|
"grad_norm": 0.3256862461566925,
|
|
"learning_rate": 4.1381025918476145e-07,
|
|
"loss": 0.3012,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3206055462360382,
|
|
"step": 5340
|
|
},
|
|
{
|
|
"epoch": 5.783950617283951,
|
|
"grad_norm": 0.3072642385959625,
|
|
"learning_rate": 4.014445408239076e-07,
|
|
"loss": 0.3068,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.28575754165649414,
|
|
"step": 5345
|
|
},
|
|
{
|
|
"epoch": 5.790123456790123,
|
|
"grad_norm": 0.34158948063850403,
|
|
"learning_rate": 3.892645226481806e-07,
|
|
"loss": 0.301,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.30522072315216064,
|
|
"step": 5350
|
|
},
|
|
{
|
|
"epoch": 5.796296296296296,
|
|
"grad_norm": 0.6767818927764893,
|
|
"learning_rate": 3.772703200656547e-07,
|
|
"loss": 0.3038,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.35004496574401855,
|
|
"step": 5355
|
|
},
|
|
{
|
|
"epoch": 5.802469135802469,
|
|
"grad_norm": 0.30617982149124146,
|
|
"learning_rate": 3.654620467237702e-07,
|
|
"loss": 0.3091,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3403550684452057,
|
|
"step": 5360
|
|
},
|
|
{
|
|
"epoch": 5.8086419753086425,
|
|
"grad_norm": 0.31371399760246277,
|
|
"learning_rate": 3.5383981450825445e-07,
|
|
"loss": 0.2996,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26886507868766785,
|
|
"step": 5365
|
|
},
|
|
{
|
|
"epoch": 5.814814814814815,
|
|
"grad_norm": 0.2771609127521515,
|
|
"learning_rate": 3.424037335420471e-07,
|
|
"loss": 0.2959,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2909230887889862,
|
|
"step": 5370
|
|
},
|
|
{
|
|
"epoch": 5.820987654320987,
|
|
"grad_norm": 0.2863991856575012,
|
|
"learning_rate": 3.3115391218428107e-07,
|
|
"loss": 0.2858,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26095375418663025,
|
|
"step": 5375
|
|
},
|
|
{
|
|
"epoch": 5.827160493827161,
|
|
"grad_norm": 0.29970279335975647,
|
|
"learning_rate": 3.2009045702923404e-07,
|
|
"loss": 0.2957,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.257806658744812,
|
|
"step": 5380
|
|
},
|
|
{
|
|
"epoch": 5.833333333333333,
|
|
"grad_norm": 0.283952921628952,
|
|
"learning_rate": 3.0921347290532755e-07,
|
|
"loss": 0.2923,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2904301881790161,
|
|
"step": 5385
|
|
},
|
|
{
|
|
"epoch": 5.839506172839506,
|
|
"grad_norm": 0.33126378059387207,
|
|
"learning_rate": 2.985230628741387e-07,
|
|
"loss": 0.2956,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3125201463699341,
|
|
"step": 5390
|
|
},
|
|
{
|
|
"epoch": 5.845679012345679,
|
|
"grad_norm": 0.2840425968170166,
|
|
"learning_rate": 2.880193282294097e-07,
|
|
"loss": 0.3048,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3535851538181305,
|
|
"step": 5395
|
|
},
|
|
{
|
|
"epoch": 5.851851851851852,
|
|
"grad_norm": 0.303683340549469,
|
|
"learning_rate": 2.7770236849610663e-07,
|
|
"loss": 0.2848,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24997055530548096,
|
|
"step": 5400
|
|
},
|
|
{
|
|
"epoch": 5.8580246913580245,
|
|
"grad_norm": 0.3636462390422821,
|
|
"learning_rate": 2.675722814294601e-07,
|
|
"loss": 0.3014,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.28643155097961426,
|
|
"step": 5405
|
|
},
|
|
{
|
|
"epoch": 5.864197530864198,
|
|
"grad_norm": 0.2753598988056183,
|
|
"learning_rate": 2.5762916301404597e-07,
|
|
"loss": 0.2947,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.3065018951892853,
|
|
"step": 5410
|
|
},
|
|
{
|
|
"epoch": 5.87037037037037,
|
|
"grad_norm": 0.27454206347465515,
|
|
"learning_rate": 2.4787310746287265e-07,
|
|
"loss": 0.3008,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.26914170384407043,
|
|
"step": 5415
|
|
},
|
|
{
|
|
"epoch": 5.8765432098765435,
|
|
"grad_norm": 0.700809895992279,
|
|
"learning_rate": 2.3830420721649982e-07,
|
|
"loss": 0.2747,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27469417452812195,
|
|
"step": 5420
|
|
},
|
|
{
|
|
"epoch": 5.882716049382716,
|
|
"grad_norm": 0.4362209737300873,
|
|
"learning_rate": 2.289225529421435e-07,
|
|
"loss": 0.2249,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20303870737552643,
|
|
"step": 5425
|
|
},
|
|
{
|
|
"epoch": 5.888888888888889,
|
|
"grad_norm": 0.4066512882709503,
|
|
"learning_rate": 2.1972823353283213e-07,
|
|
"loss": 0.2158,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2180003672838211,
|
|
"step": 5430
|
|
},
|
|
{
|
|
"epoch": 5.895061728395062,
|
|
"grad_norm": 0.37600329518318176,
|
|
"learning_rate": 2.1072133610656077e-07,
|
|
"loss": 0.2389,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2285630702972412,
|
|
"step": 5435
|
|
},
|
|
{
|
|
"epoch": 5.901234567901234,
|
|
"grad_norm": 0.3472577631473541,
|
|
"learning_rate": 2.0190194600546052e-07,
|
|
"loss": 0.2438,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2246277779340744,
|
|
"step": 5440
|
|
},
|
|
{
|
|
"epoch": 5.907407407407407,
|
|
"grad_norm": 0.2951968014240265,
|
|
"learning_rate": 1.932701467949949e-07,
|
|
"loss": 0.2256,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19835145771503448,
|
|
"step": 5445
|
|
},
|
|
{
|
|
"epoch": 5.91358024691358,
|
|
"grad_norm": 0.33108726143836975,
|
|
"learning_rate": 1.8482602026316687e-07,
|
|
"loss": 0.2252,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.24588294327259064,
|
|
"step": 5450
|
|
},
|
|
{
|
|
"epoch": 5.919753086419753,
|
|
"grad_norm": 0.36378446221351624,
|
|
"learning_rate": 1.765696464197464e-07,
|
|
"loss": 0.2248,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21787647902965546,
|
|
"step": 5455
|
|
},
|
|
{
|
|
"epoch": 5.925925925925926,
|
|
"grad_norm": 0.3135976493358612,
|
|
"learning_rate": 1.6850110349550198e-07,
|
|
"loss": 0.2227,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.21238481998443604,
|
|
"step": 5460
|
|
},
|
|
{
|
|
"epoch": 5.932098765432099,
|
|
"grad_norm": 0.3109370768070221,
|
|
"learning_rate": 1.6062046794147468e-07,
|
|
"loss": 0.2375,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2446727305650711,
|
|
"step": 5465
|
|
},
|
|
{
|
|
"epoch": 5.938271604938271,
|
|
"grad_norm": 0.3311823904514313,
|
|
"learning_rate": 1.5292781442824312e-07,
|
|
"loss": 0.2317,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23923824727535248,
|
|
"step": 5470
|
|
},
|
|
{
|
|
"epoch": 5.944444444444445,
|
|
"grad_norm": 0.34650668501853943,
|
|
"learning_rate": 1.4542321584521956e-07,
|
|
"loss": 0.2538,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.27205944061279297,
|
|
"step": 5475
|
|
},
|
|
{
|
|
"epoch": 5.950617283950617,
|
|
"grad_norm": 0.30899497866630554,
|
|
"learning_rate": 1.3810674329995943e-07,
|
|
"loss": 0.2529,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22966128587722778,
|
|
"step": 5480
|
|
},
|
|
{
|
|
"epoch": 5.95679012345679,
|
|
"grad_norm": 0.30232787132263184,
|
|
"learning_rate": 1.3097846611748177e-07,
|
|
"loss": 0.2509,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2702513039112091,
|
|
"step": 5485
|
|
},
|
|
{
|
|
"epoch": 5.962962962962963,
|
|
"grad_norm": 0.28088781237602234,
|
|
"learning_rate": 1.2403845183962536e-07,
|
|
"loss": 0.2423,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.22831811010837555,
|
|
"step": 5490
|
|
},
|
|
{
|
|
"epoch": 5.969135802469136,
|
|
"grad_norm": 0.3470403254032135,
|
|
"learning_rate": 1.1728676622439373e-07,
|
|
"loss": 0.242,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2442747801542282,
|
|
"step": 5495
|
|
},
|
|
{
|
|
"epoch": 5.9753086419753085,
|
|
"grad_norm": 0.32140395045280457,
|
|
"learning_rate": 1.1072347324534438e-07,
|
|
"loss": 0.2216,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19685661792755127,
|
|
"step": 5500
|
|
},
|
|
{
|
|
"epoch": 5.981481481481482,
|
|
"grad_norm": 0.32901397347450256,
|
|
"learning_rate": 1.0434863509097171e-07,
|
|
"loss": 0.2105,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20024986565113068,
|
|
"step": 5505
|
|
},
|
|
{
|
|
"epoch": 5.987654320987654,
|
|
"grad_norm": 0.3855687975883484,
|
|
"learning_rate": 9.816231216412953e-08,
|
|
"loss": 0.2127,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20528362691402435,
|
|
"step": 5510
|
|
},
|
|
{
|
|
"epoch": 5.993827160493828,
|
|
"grad_norm": 0.3800973892211914,
|
|
"learning_rate": 9.216456308145161e-08,
|
|
"loss": 0.2156,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.23067224025726318,
|
|
"step": 5515
|
|
},
|
|
{
|
|
"epoch": 6.001234567901235,
|
|
"grad_norm": 0.7135549187660217,
|
|
"learning_rate": 8.63554446727899e-08,
|
|
"loss": 0.2193,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20058362185955048,
|
|
"step": 5520
|
|
},
|
|
{
|
|
"epoch": 6.007407407407407,
|
|
"grad_norm": 0.465126097202301,
|
|
"learning_rate": 8.07350119806949e-08,
|
|
"loss": 0.1872,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18871080875396729,
|
|
"step": 5525
|
|
},
|
|
{
|
|
"epoch": 6.01358024691358,
|
|
"grad_norm": 0.3813707232475281,
|
|
"learning_rate": 7.530331825987169e-08,
|
|
"loss": 0.1843,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.16995465755462646,
|
|
"step": 5530
|
|
},
|
|
{
|
|
"epoch": 6.019753086419753,
|
|
"grad_norm": 0.44798824191093445,
|
|
"learning_rate": 7.006041497669592e-08,
|
|
"loss": 0.1782,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18314535915851593,
|
|
"step": 5535
|
|
},
|
|
{
|
|
"epoch": 6.025925925925926,
|
|
"grad_norm": 0.3439101278781891,
|
|
"learning_rate": 6.50063518087074e-08,
|
|
"loss": 0.176,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18882863223552704,
|
|
"step": 5540
|
|
},
|
|
{
|
|
"epoch": 6.0320987654320986,
|
|
"grad_norm": 0.29852959513664246,
|
|
"learning_rate": 6.014117664415953e-08,
|
|
"loss": 0.1804,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19732850790023804,
|
|
"step": 5545
|
|
},
|
|
{
|
|
"epoch": 6.038271604938272,
|
|
"grad_norm": 0.33499741554260254,
|
|
"learning_rate": 5.5464935581541756e-08,
|
|
"loss": 0.1722,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1896158903837204,
|
|
"step": 5550
|
|
},
|
|
{
|
|
"epoch": 6.044444444444444,
|
|
"grad_norm": 0.35985612869262695,
|
|
"learning_rate": 5.097767292916445e-08,
|
|
"loss": 0.1806,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20118701457977295,
|
|
"step": 5555
|
|
},
|
|
{
|
|
"epoch": 6.050617283950618,
|
|
"grad_norm": 0.3340224623680115,
|
|
"learning_rate": 4.6679431204730284e-08,
|
|
"loss": 0.1789,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1518581360578537,
|
|
"step": 5560
|
|
},
|
|
{
|
|
"epoch": 6.05679012345679,
|
|
"grad_norm": 0.44487977027893066,
|
|
"learning_rate": 4.2570251134925745e-08,
|
|
"loss": 0.1764,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1720142811536789,
|
|
"step": 5565
|
|
},
|
|
{
|
|
"epoch": 6.062962962962963,
|
|
"grad_norm": 0.34128713607788086,
|
|
"learning_rate": 3.865017165504359e-08,
|
|
"loss": 0.168,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19164510071277618,
|
|
"step": 5570
|
|
},
|
|
{
|
|
"epoch": 6.069135802469136,
|
|
"grad_norm": 0.4419310390949249,
|
|
"learning_rate": 3.4919229908614294e-08,
|
|
"loss": 0.1911,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.19037608802318573,
|
|
"step": 5575
|
|
},
|
|
{
|
|
"epoch": 6.075308641975309,
|
|
"grad_norm": 0.4176490604877472,
|
|
"learning_rate": 3.1377461247046326e-08,
|
|
"loss": 0.1723,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.16997994482517242,
|
|
"step": 5580
|
|
},
|
|
{
|
|
"epoch": 6.0814814814814815,
|
|
"grad_norm": 0.41486406326293945,
|
|
"learning_rate": 2.8024899229295298e-08,
|
|
"loss": 0.1881,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.2046893835067749,
|
|
"step": 5585
|
|
},
|
|
{
|
|
"epoch": 6.087654320987654,
|
|
"grad_norm": 0.3950762450695038,
|
|
"learning_rate": 2.4861575621553112e-08,
|
|
"loss": 0.1884,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.20917779207229614,
|
|
"step": 5590
|
|
},
|
|
{
|
|
"epoch": 6.093827160493827,
|
|
"grad_norm": 0.8507141470909119,
|
|
"learning_rate": 2.188752039693265e-08,
|
|
"loss": 0.1828,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.16898727416992188,
|
|
"step": 5595
|
|
},
|
|
{
|
|
"epoch": 6.1,
|
|
"grad_norm": 0.43243470788002014,
|
|
"learning_rate": 1.9102761735196873e-08,
|
|
"loss": 0.1976,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18492530286312103,
|
|
"step": 5600
|
|
},
|
|
{
|
|
"epoch": 6.106172839506173,
|
|
"grad_norm": 0.4170914590358734,
|
|
"learning_rate": 1.6507326022481285e-08,
|
|
"loss": 0.1775,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.191942036151886,
|
|
"step": 5605
|
|
},
|
|
{
|
|
"epoch": 6.112345679012345,
|
|
"grad_norm": 0.38983917236328125,
|
|
"learning_rate": 1.4101237851056327e-08,
|
|
"loss": 0.1757,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.16892965137958527,
|
|
"step": 5610
|
|
},
|
|
{
|
|
"epoch": 6.118518518518519,
|
|
"grad_norm": 0.3913613259792328,
|
|
"learning_rate": 1.1884520019078694e-08,
|
|
"loss": 0.1819,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17739228904247284,
|
|
"step": 5615
|
|
},
|
|
{
|
|
"epoch": 6.124691358024691,
|
|
"grad_norm": 0.36839035153388977,
|
|
"learning_rate": 9.857193530391495e-09,
|
|
"loss": 0.1735,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17736367881298065,
|
|
"step": 5620
|
|
},
|
|
{
|
|
"epoch": 6.130864197530864,
|
|
"grad_norm": 0.4212496280670166,
|
|
"learning_rate": 8.019277594311091e-09,
|
|
"loss": 0.176,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17365749180316925,
|
|
"step": 5625
|
|
},
|
|
{
|
|
"epoch": 6.137037037037037,
|
|
"grad_norm": 0.40971148014068604,
|
|
"learning_rate": 6.3707896254539e-09,
|
|
"loss": 0.1844,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.16560529172420502,
|
|
"step": 5630
|
|
},
|
|
{
|
|
"epoch": 6.14320987654321,
|
|
"grad_norm": 0.3838355839252472,
|
|
"learning_rate": 4.911745243569854e-09,
|
|
"loss": 0.1912,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1862587183713913,
|
|
"step": 5635
|
|
},
|
|
{
|
|
"epoch": 6.149382716049383,
|
|
"grad_norm": 0.3883192539215088,
|
|
"learning_rate": 3.642158273393648e-09,
|
|
"loss": 0.1788,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17281018197536469,
|
|
"step": 5640
|
|
},
|
|
{
|
|
"epoch": 6.155555555555556,
|
|
"grad_norm": 0.39495953917503357,
|
|
"learning_rate": 2.562040744513716e-09,
|
|
"loss": 0.1743,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.17838597297668457,
|
|
"step": 5645
|
|
},
|
|
{
|
|
"epoch": 6.161728395061728,
|
|
"grad_norm": 0.37174108624458313,
|
|
"learning_rate": 1.6714028912545589e-09,
|
|
"loss": 0.1914,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.16914498805999756,
|
|
"step": 5650
|
|
},
|
|
{
|
|
"epoch": 6.167901234567902,
|
|
"grad_norm": 0.4627566337585449,
|
|
"learning_rate": 9.702531525856984e-10,
|
|
"loss": 0.1834,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.16920219361782074,
|
|
"step": 5655
|
|
},
|
|
{
|
|
"epoch": 6.174074074074074,
|
|
"grad_norm": 0.3644845485687256,
|
|
"learning_rate": 4.5859817203952605e-10,
|
|
"loss": 0.1861,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1958434134721756,
|
|
"step": 5660
|
|
},
|
|
{
|
|
"epoch": 6.1802469135802465,
|
|
"grad_norm": 0.43530961871147156,
|
|
"learning_rate": 1.3644279765134828e-10,
|
|
"loss": 0.1795,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.1919514685869217,
|
|
"step": 5665
|
|
},
|
|
{
|
|
"epoch": 6.18641975308642,
|
|
"grad_norm": 0.3615497350692749,
|
|
"learning_rate": 3.79008190165564e-12,
|
|
"loss": 0.1707,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.18564243614673615,
|
|
"step": 5670
|
|
},
|
|
{
|
|
"epoch": 6.18641975308642,
|
|
"step": 5670,
|
|
"total_flos": 3.2973778191384576e+16,
|
|
"train_loss": 0.0,
|
|
"train_runtime": 0.8183,
|
|
"train_samples_per_second": 13858.168,
|
|
"train_steps_per_second": 6929.084
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 5670,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 7,
|
|
"save_steps": 300,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": true
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 3.2973778191384576e+16,
|
|
"train_batch_size": 1,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|