full-lstm-0 / checkpoint-152640 /trainer_state.json
abhinavp's picture
Training in progress, epoch 1, checkpoint
f9c5fa2 verified
raw
history blame
37.5 kB
{
"best_metric": 4.48332405090332,
"best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/full2/lstm/0/checkpoints/checkpoint-152640",
"epoch": 1.0250006060157382,
"eval_steps": 10,
"global_step": 152640,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 4.999998362119627e-05,
"loss": 10.8214,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 4.999161405248948e-05,
"loss": 7.5602,
"step": 512
},
{
"epoch": 0.0,
"learning_rate": 4.998322810497896e-05,
"loss": 7.058,
"step": 1024
},
{
"epoch": 0.0,
"learning_rate": 4.997484215746844e-05,
"loss": 6.9951,
"step": 1536
},
{
"epoch": 0.0,
"learning_rate": 4.996645620995792e-05,
"loss": 6.9498,
"step": 2048
},
{
"epoch": 0.0,
"learning_rate": 4.99580702624474e-05,
"loss": 6.8998,
"step": 2560
},
{
"epoch": 0.0,
"learning_rate": 4.994968431493688e-05,
"loss": 6.7169,
"step": 3072
},
{
"epoch": 0.0,
"learning_rate": 4.994131474623009e-05,
"loss": 6.6079,
"step": 3584
},
{
"epoch": 0.0,
"learning_rate": 4.993292879871958e-05,
"loss": 6.5124,
"step": 4096
},
{
"epoch": 0.0,
"learning_rate": 4.992454285120906e-05,
"loss": 6.4465,
"step": 4608
},
{
"epoch": 0.0,
"learning_rate": 4.991615690369854e-05,
"loss": 6.3968,
"step": 5120
},
{
"epoch": 0.0,
"learning_rate": 4.990777095618801e-05,
"loss": 6.3509,
"step": 5632
},
{
"epoch": 0.0,
"learning_rate": 4.989938500867749e-05,
"loss": 6.2991,
"step": 6144
},
{
"epoch": 0.0,
"learning_rate": 4.989099906116697e-05,
"loss": 6.2314,
"step": 6656
},
{
"epoch": 0.0,
"learning_rate": 4.988261311365645e-05,
"loss": 6.1847,
"step": 7168
},
{
"epoch": 0.0,
"learning_rate": 4.987422716614593e-05,
"loss": 6.1285,
"step": 7680
},
{
"epoch": 0.0,
"learning_rate": 4.986584121863541e-05,
"loss": 6.0873,
"step": 8192
},
{
"epoch": 0.0,
"learning_rate": 4.985745527112489e-05,
"loss": 6.0405,
"step": 8704
},
{
"epoch": 0.0,
"learning_rate": 4.984906932361437e-05,
"loss": 5.9979,
"step": 9216
},
{
"epoch": 0.0,
"learning_rate": 4.984068337610385e-05,
"loss": 5.95,
"step": 9728
},
{
"epoch": 0.0,
"learning_rate": 4.983229742859333e-05,
"loss": 5.9171,
"step": 10240
},
{
"epoch": 0.0,
"learning_rate": 4.982391148108281e-05,
"loss": 5.8771,
"step": 10752
},
{
"epoch": 0.0,
"learning_rate": 4.981552553357229e-05,
"loss": 5.8495,
"step": 11264
},
{
"epoch": 0.0,
"learning_rate": 4.980713958606178e-05,
"loss": 5.8067,
"step": 11776
},
{
"epoch": 0.0,
"learning_rate": 4.979875363855125e-05,
"loss": 5.7943,
"step": 12288
},
{
"epoch": 0.0,
"learning_rate": 4.979036769104073e-05,
"loss": 5.762,
"step": 12800
},
{
"epoch": 0.0,
"learning_rate": 4.9782014501137675e-05,
"loss": 5.7333,
"step": 13312
},
{
"epoch": 0.0,
"learning_rate": 4.9773628553627155e-05,
"loss": 5.6998,
"step": 13824
},
{
"epoch": 0.0,
"learning_rate": 4.9765242606116635e-05,
"loss": 5.6794,
"step": 14336
},
{
"epoch": 0.0,
"learning_rate": 4.9756856658606115e-05,
"loss": 5.6642,
"step": 14848
},
{
"epoch": 0.01,
"learning_rate": 4.9748470711095595e-05,
"loss": 5.6328,
"step": 15360
},
{
"epoch": 0.01,
"learning_rate": 4.9740101142388804e-05,
"loss": 5.6229,
"step": 15872
},
{
"epoch": 0.01,
"learning_rate": 4.9731715194878284e-05,
"loss": 5.5966,
"step": 16384
},
{
"epoch": 0.01,
"learning_rate": 4.9723329247367764e-05,
"loss": 5.5783,
"step": 16896
},
{
"epoch": 0.01,
"learning_rate": 4.9714943299857244e-05,
"loss": 5.5755,
"step": 17408
},
{
"epoch": 0.01,
"learning_rate": 4.9706557352346724e-05,
"loss": 5.558,
"step": 17920
},
{
"epoch": 0.01,
"learning_rate": 4.9698171404836204e-05,
"loss": 5.533,
"step": 18432
},
{
"epoch": 0.01,
"learning_rate": 4.968981821493315e-05,
"loss": 5.5055,
"step": 18944
},
{
"epoch": 0.01,
"learning_rate": 4.968143226742263e-05,
"loss": 5.4963,
"step": 19456
},
{
"epoch": 0.01,
"learning_rate": 4.967304631991211e-05,
"loss": 5.472,
"step": 19968
},
{
"epoch": 0.01,
"learning_rate": 4.966466037240159e-05,
"loss": 5.4702,
"step": 20480
},
{
"epoch": 0.01,
"learning_rate": 4.965627442489107e-05,
"loss": 5.4463,
"step": 20992
},
{
"epoch": 0.01,
"learning_rate": 4.964788847738054e-05,
"loss": 5.4402,
"step": 21504
},
{
"epoch": 0.01,
"learning_rate": 4.963950252987002e-05,
"loss": 5.4154,
"step": 22016
},
{
"epoch": 0.01,
"learning_rate": 4.96311165823595e-05,
"loss": 5.4164,
"step": 22528
},
{
"epoch": 0.01,
"learning_rate": 4.962273063484898e-05,
"loss": 5.4033,
"step": 23040
},
{
"epoch": 0.01,
"learning_rate": 4.961434468733847e-05,
"loss": 5.4023,
"step": 23552
},
{
"epoch": 0.01,
"learning_rate": 4.960597511863168e-05,
"loss": 5.3899,
"step": 24064
},
{
"epoch": 0.01,
"learning_rate": 4.959758917112116e-05,
"loss": 5.3606,
"step": 24576
},
{
"epoch": 0.01,
"learning_rate": 4.958920322361064e-05,
"loss": 5.3523,
"step": 25088
},
{
"epoch": 0.01,
"learning_rate": 4.958083365490385e-05,
"loss": 5.3614,
"step": 25600
},
{
"epoch": 0.01,
"learning_rate": 4.957244770739333e-05,
"loss": 5.342,
"step": 26112
},
{
"epoch": 0.01,
"learning_rate": 4.956406175988281e-05,
"loss": 5.3285,
"step": 26624
},
{
"epoch": 0.01,
"learning_rate": 4.955567581237229e-05,
"loss": 5.3105,
"step": 27136
},
{
"epoch": 0.01,
"learning_rate": 4.954728986486177e-05,
"loss": 5.3129,
"step": 27648
},
{
"epoch": 0.01,
"learning_rate": 4.953890391735125e-05,
"loss": 5.2933,
"step": 28160
},
{
"epoch": 0.01,
"learning_rate": 4.9530517969840727e-05,
"loss": 5.3049,
"step": 28672
},
{
"epoch": 0.01,
"learning_rate": 4.9522132022330207e-05,
"loss": 5.2688,
"step": 29184
},
{
"epoch": 0.01,
"learning_rate": 4.951376245362342e-05,
"loss": 5.2716,
"step": 29696
},
{
"epoch": 0.01,
"learning_rate": 4.95053765061129e-05,
"loss": 5.2677,
"step": 30208
},
{
"epoch": 0.01,
"learning_rate": 4.949699055860238e-05,
"loss": 5.2519,
"step": 30720
},
{
"epoch": 0.01,
"learning_rate": 4.948860461109186e-05,
"loss": 5.2589,
"step": 31232
},
{
"epoch": 0.01,
"learning_rate": 4.948021866358134e-05,
"loss": 5.2281,
"step": 31744
},
{
"epoch": 0.01,
"learning_rate": 4.947184909487455e-05,
"loss": 5.2234,
"step": 32256
},
{
"epoch": 0.01,
"learning_rate": 4.946346314736403e-05,
"loss": 5.2122,
"step": 32768
},
{
"epoch": 0.01,
"learning_rate": 4.945507719985351e-05,
"loss": 5.2322,
"step": 33280
},
{
"epoch": 0.01,
"learning_rate": 4.944669125234299e-05,
"loss": 5.1984,
"step": 33792
},
{
"epoch": 0.01,
"learning_rate": 4.943830530483247e-05,
"loss": 5.2,
"step": 34304
},
{
"epoch": 0.01,
"learning_rate": 4.942991935732195e-05,
"loss": 5.1829,
"step": 34816
},
{
"epoch": 0.01,
"learning_rate": 4.942153340981143e-05,
"loss": 5.176,
"step": 35328
},
{
"epoch": 0.01,
"learning_rate": 4.941316384110464e-05,
"loss": 5.1845,
"step": 35840
},
{
"epoch": 0.01,
"learning_rate": 4.940477789359412e-05,
"loss": 5.1766,
"step": 36352
},
{
"epoch": 0.01,
"learning_rate": 4.93963919460836e-05,
"loss": 5.1683,
"step": 36864
},
{
"epoch": 0.01,
"learning_rate": 4.938800599857309e-05,
"loss": 5.1671,
"step": 37376
},
{
"epoch": 0.01,
"learning_rate": 4.937962005106256e-05,
"loss": 5.17,
"step": 37888
},
{
"epoch": 0.01,
"learning_rate": 4.937123410355204e-05,
"loss": 5.1369,
"step": 38400
},
{
"epoch": 0.01,
"learning_rate": 4.9362864534845256e-05,
"loss": 5.1444,
"step": 38912
},
{
"epoch": 0.01,
"learning_rate": 4.9354494966138465e-05,
"loss": 5.1229,
"step": 39424
},
{
"epoch": 0.01,
"learning_rate": 4.9346109018627945e-05,
"loss": 5.1172,
"step": 39936
},
{
"epoch": 0.01,
"learning_rate": 4.9337723071117425e-05,
"loss": 5.1179,
"step": 40448
},
{
"epoch": 0.01,
"learning_rate": 4.9329337123606905e-05,
"loss": 5.1183,
"step": 40960
},
{
"epoch": 0.01,
"learning_rate": 4.932095117609638e-05,
"loss": 5.1074,
"step": 41472
},
{
"epoch": 0.01,
"learning_rate": 4.931256522858586e-05,
"loss": 5.1141,
"step": 41984
},
{
"epoch": 0.01,
"learning_rate": 4.930417928107534e-05,
"loss": 5.0963,
"step": 42496
},
{
"epoch": 0.01,
"learning_rate": 4.9295793333564825e-05,
"loss": 5.0776,
"step": 43008
},
{
"epoch": 0.01,
"learning_rate": 4.9287423764858034e-05,
"loss": 5.073,
"step": 43520
},
{
"epoch": 0.01,
"learning_rate": 4.9279037817347514e-05,
"loss": 5.0864,
"step": 44032
},
{
"epoch": 0.01,
"learning_rate": 4.9270651869836994e-05,
"loss": 5.0663,
"step": 44544
},
{
"epoch": 0.01,
"learning_rate": 4.9262265922326474e-05,
"loss": 5.0641,
"step": 45056
},
{
"epoch": 0.01,
"learning_rate": 4.925389635361968e-05,
"loss": 5.0606,
"step": 45568
},
{
"epoch": 0.02,
"learning_rate": 4.924551040610916e-05,
"loss": 5.0481,
"step": 46080
},
{
"epoch": 0.02,
"learning_rate": 4.923712445859864e-05,
"loss": 5.0503,
"step": 46592
},
{
"epoch": 0.02,
"learning_rate": 4.922873851108812e-05,
"loss": 5.044,
"step": 47104
},
{
"epoch": 0.02,
"learning_rate": 4.922036894238133e-05,
"loss": 5.0421,
"step": 47616
},
{
"epoch": 0.02,
"learning_rate": 4.921198299487081e-05,
"loss": 5.0395,
"step": 48128
},
{
"epoch": 0.02,
"learning_rate": 4.920361342616403e-05,
"loss": 5.0307,
"step": 48640
},
{
"epoch": 0.02,
"learning_rate": 4.919522747865351e-05,
"loss": 5.0134,
"step": 49152
},
{
"epoch": 0.02,
"learning_rate": 4.918684153114299e-05,
"loss": 5.0157,
"step": 49664
},
{
"epoch": 0.02,
"learning_rate": 4.917845558363247e-05,
"loss": 5.0042,
"step": 50176
},
{
"epoch": 0.02,
"learning_rate": 4.9170086014925676e-05,
"loss": 5.0036,
"step": 50688
},
{
"epoch": 0.02,
"learning_rate": 4.9161700067415156e-05,
"loss": 5.0109,
"step": 51200
},
{
"epoch": 0.02,
"learning_rate": 4.9153314119904636e-05,
"loss": 4.9903,
"step": 51712
},
{
"epoch": 0.02,
"learning_rate": 4.9144928172394116e-05,
"loss": 4.994,
"step": 52224
},
{
"epoch": 0.02,
"learning_rate": 4.9136558603687325e-05,
"loss": 4.9877,
"step": 52736
},
{
"epoch": 0.02,
"learning_rate": 4.9128172656176805e-05,
"loss": 4.977,
"step": 53248
},
{
"epoch": 0.02,
"learning_rate": 4.9119786708666285e-05,
"loss": 4.9688,
"step": 53760
},
{
"epoch": 0.02,
"learning_rate": 4.9111400761155765e-05,
"loss": 4.9678,
"step": 54272
},
{
"epoch": 0.02,
"learning_rate": 4.9103014813645245e-05,
"loss": 4.9522,
"step": 54784
},
{
"epoch": 0.02,
"learning_rate": 4.909462886613473e-05,
"loss": 4.9542,
"step": 55296
},
{
"epoch": 0.02,
"learning_rate": 4.908625929742794e-05,
"loss": 4.9508,
"step": 55808
},
{
"epoch": 0.02,
"learning_rate": 4.907787334991742e-05,
"loss": 4.9503,
"step": 56320
},
{
"epoch": 0.02,
"learning_rate": 4.90694874024069e-05,
"loss": 4.942,
"step": 56832
},
{
"epoch": 0.02,
"learning_rate": 4.906110145489638e-05,
"loss": 4.9391,
"step": 57344
},
{
"epoch": 0.02,
"learning_rate": 4.905271550738586e-05,
"loss": 4.9304,
"step": 57856
},
{
"epoch": 0.02,
"learning_rate": 4.904432955987534e-05,
"loss": 4.9339,
"step": 58368
},
{
"epoch": 0.02,
"learning_rate": 4.903594361236482e-05,
"loss": 4.9347,
"step": 58880
},
{
"epoch": 0.02,
"learning_rate": 4.902757404365803e-05,
"loss": 4.9285,
"step": 59392
},
{
"epoch": 0.02,
"learning_rate": 4.901918809614751e-05,
"loss": 4.9281,
"step": 59904
},
{
"epoch": 0.02,
"learning_rate": 4.901080214863699e-05,
"loss": 4.9148,
"step": 60416
},
{
"epoch": 0.02,
"learning_rate": 4.900241620112647e-05,
"loss": 4.9169,
"step": 60928
},
{
"epoch": 0.02,
"learning_rate": 4.899403025361595e-05,
"loss": 4.9137,
"step": 61440
},
{
"epoch": 0.02,
"learning_rate": 4.898564430610543e-05,
"loss": 4.8992,
"step": 61952
},
{
"epoch": 0.02,
"learning_rate": 4.897725835859491e-05,
"loss": 4.8968,
"step": 62464
},
{
"epoch": 0.02,
"learning_rate": 4.8968872411084396e-05,
"loss": 4.895,
"step": 62976
},
{
"epoch": 0.02,
"learning_rate": 4.896048646357387e-05,
"loss": 4.8944,
"step": 63488
},
{
"epoch": 0.02,
"learning_rate": 4.895210051606335e-05,
"loss": 4.8814,
"step": 64000
},
{
"epoch": 0.02,
"learning_rate": 4.8943730947356565e-05,
"loss": 4.8867,
"step": 64512
},
{
"epoch": 0.02,
"learning_rate": 4.893534499984604e-05,
"loss": 4.8797,
"step": 65024
},
{
"epoch": 0.02,
"learning_rate": 4.892695905233552e-05,
"loss": 4.8735,
"step": 65536
},
{
"epoch": 0.02,
"learning_rate": 4.8918573104825e-05,
"loss": 4.8651,
"step": 66048
},
{
"epoch": 0.02,
"learning_rate": 4.8910203536118214e-05,
"loss": 4.8823,
"step": 66560
},
{
"epoch": 0.02,
"learning_rate": 4.890181758860769e-05,
"loss": 4.873,
"step": 67072
},
{
"epoch": 0.02,
"learning_rate": 4.889343164109717e-05,
"loss": 4.865,
"step": 67584
},
{
"epoch": 0.02,
"learning_rate": 4.888504569358665e-05,
"loss": 4.8633,
"step": 68096
},
{
"epoch": 0.02,
"learning_rate": 4.8876659746076134e-05,
"loss": 4.8613,
"step": 68608
},
{
"epoch": 0.02,
"learning_rate": 4.8868273798565614e-05,
"loss": 4.8595,
"step": 69120
},
{
"epoch": 0.02,
"learning_rate": 4.8859887851055094e-05,
"loss": 4.8591,
"step": 69632
},
{
"epoch": 0.02,
"learning_rate": 4.8851501903544574e-05,
"loss": 4.8498,
"step": 70144
},
{
"epoch": 0.02,
"learning_rate": 4.8843115956034054e-05,
"loss": 4.8446,
"step": 70656
},
{
"epoch": 0.02,
"learning_rate": 4.8834730008523534e-05,
"loss": 4.8421,
"step": 71168
},
{
"epoch": 0.02,
"learning_rate": 4.882636043981674e-05,
"loss": 4.8299,
"step": 71680
},
{
"epoch": 0.02,
"learning_rate": 4.881799087110995e-05,
"loss": 4.826,
"step": 72192
},
{
"epoch": 0.02,
"learning_rate": 4.880960492359943e-05,
"loss": 4.8261,
"step": 72704
},
{
"epoch": 0.02,
"learning_rate": 4.880121897608891e-05,
"loss": 4.8302,
"step": 73216
},
{
"epoch": 0.02,
"learning_rate": 4.879283302857839e-05,
"loss": 4.8226,
"step": 73728
},
{
"epoch": 0.02,
"learning_rate": 4.878444708106787e-05,
"loss": 4.8218,
"step": 74240
},
{
"epoch": 0.02,
"learning_rate": 4.877606113355735e-05,
"loss": 4.8172,
"step": 74752
},
{
"epoch": 0.02,
"learning_rate": 4.876767518604683e-05,
"loss": 4.8179,
"step": 75264
},
{
"epoch": 0.02,
"learning_rate": 4.875928923853632e-05,
"loss": 4.8237,
"step": 75776
},
{
"epoch": 0.02,
"learning_rate": 4.87509032910258e-05,
"loss": 4.7998,
"step": 76288
},
{
"epoch": 0.03,
"eval_loss": 4.768176555633545,
"eval_runtime": 323.1793,
"eval_samples_per_second": 1180.741,
"eval_steps_per_second": 36.899,
"step": 76320
},
{
"epoch": 1.0,
"learning_rate": 4.874253372231901e-05,
"loss": 4.7916,
"step": 76800
},
{
"epoch": 1.0,
"learning_rate": 4.873414777480849e-05,
"loss": 4.7905,
"step": 77312
},
{
"epoch": 1.0,
"learning_rate": 4.872576182729797e-05,
"loss": 4.8073,
"step": 77824
},
{
"epoch": 1.0,
"learning_rate": 4.871737587978745e-05,
"loss": 4.7926,
"step": 78336
},
{
"epoch": 1.0,
"learning_rate": 4.870898993227693e-05,
"loss": 4.7972,
"step": 78848
},
{
"epoch": 1.0,
"learning_rate": 4.87006039847664e-05,
"loss": 4.7881,
"step": 79360
},
{
"epoch": 1.0,
"learning_rate": 4.869221803725588e-05,
"loss": 4.7836,
"step": 79872
},
{
"epoch": 1.0,
"learning_rate": 4.868383208974536e-05,
"loss": 4.7696,
"step": 80384
},
{
"epoch": 1.0,
"learning_rate": 4.8675478899842306e-05,
"loss": 4.7756,
"step": 80896
},
{
"epoch": 1.0,
"learning_rate": 4.8667092952331785e-05,
"loss": 4.7722,
"step": 81408
},
{
"epoch": 1.0,
"learning_rate": 4.865870700482127e-05,
"loss": 4.7934,
"step": 81920
},
{
"epoch": 1.0,
"learning_rate": 4.865032105731075e-05,
"loss": 4.774,
"step": 82432
},
{
"epoch": 1.0,
"learning_rate": 4.864195148860396e-05,
"loss": 4.76,
"step": 82944
},
{
"epoch": 1.0,
"learning_rate": 4.863356554109344e-05,
"loss": 4.7605,
"step": 83456
},
{
"epoch": 1.0,
"learning_rate": 4.862517959358292e-05,
"loss": 4.7603,
"step": 83968
},
{
"epoch": 1.0,
"learning_rate": 4.86167936460724e-05,
"loss": 4.7514,
"step": 84480
},
{
"epoch": 1.0,
"learning_rate": 4.8608407698561874e-05,
"loss": 4.7621,
"step": 84992
},
{
"epoch": 1.0,
"learning_rate": 4.8600021751051354e-05,
"loss": 4.7558,
"step": 85504
},
{
"epoch": 1.0,
"learning_rate": 4.8591635803540834e-05,
"loss": 4.7445,
"step": 86016
},
{
"epoch": 1.0,
"learning_rate": 4.8583249856030314e-05,
"loss": 4.7631,
"step": 86528
},
{
"epoch": 1.0,
"learning_rate": 4.8574863908519794e-05,
"loss": 4.7452,
"step": 87040
},
{
"epoch": 1.0,
"learning_rate": 4.8566477961009274e-05,
"loss": 4.747,
"step": 87552
},
{
"epoch": 1.0,
"learning_rate": 4.8558092013498754e-05,
"loss": 4.7324,
"step": 88064
},
{
"epoch": 1.0,
"learning_rate": 4.854972244479197e-05,
"loss": 4.756,
"step": 88576
},
{
"epoch": 1.0,
"learning_rate": 4.854133649728145e-05,
"loss": 4.7327,
"step": 89088
},
{
"epoch": 1.0,
"learning_rate": 4.853295054977093e-05,
"loss": 4.7311,
"step": 89600
},
{
"epoch": 1.0,
"learning_rate": 4.852458098106414e-05,
"loss": 4.724,
"step": 90112
},
{
"epoch": 1.0,
"learning_rate": 4.851619503355362e-05,
"loss": 4.7258,
"step": 90624
},
{
"epoch": 1.0,
"learning_rate": 4.85078090860431e-05,
"loss": 4.721,
"step": 91136
},
{
"epoch": 1.01,
"learning_rate": 4.849942313853258e-05,
"loss": 4.7177,
"step": 91648
},
{
"epoch": 1.01,
"learning_rate": 4.849105356982579e-05,
"loss": 4.7235,
"step": 92160
},
{
"epoch": 1.01,
"learning_rate": 4.848266762231527e-05,
"loss": 4.7185,
"step": 92672
},
{
"epoch": 1.01,
"learning_rate": 4.847428167480475e-05,
"loss": 4.7143,
"step": 93184
},
{
"epoch": 1.01,
"learning_rate": 4.846589572729423e-05,
"loss": 4.7246,
"step": 93696
},
{
"epoch": 1.01,
"learning_rate": 4.8457526158587444e-05,
"loss": 4.7156,
"step": 94208
},
{
"epoch": 1.01,
"learning_rate": 4.8449140211076924e-05,
"loss": 4.7122,
"step": 94720
},
{
"epoch": 1.01,
"learning_rate": 4.844077064237013e-05,
"loss": 4.6984,
"step": 95232
},
{
"epoch": 1.01,
"learning_rate": 4.843238469485961e-05,
"loss": 4.7004,
"step": 95744
},
{
"epoch": 1.01,
"learning_rate": 4.842399874734909e-05,
"loss": 4.691,
"step": 96256
},
{
"epoch": 1.01,
"learning_rate": 4.841561279983857e-05,
"loss": 4.6997,
"step": 96768
},
{
"epoch": 1.01,
"learning_rate": 4.840722685232805e-05,
"loss": 4.6883,
"step": 97280
},
{
"epoch": 1.01,
"learning_rate": 4.839884090481753e-05,
"loss": 4.6954,
"step": 97792
},
{
"epoch": 1.01,
"learning_rate": 4.839045495730701e-05,
"loss": 4.6777,
"step": 98304
},
{
"epoch": 1.01,
"learning_rate": 4.838206900979649e-05,
"loss": 4.6986,
"step": 98816
},
{
"epoch": 1.01,
"learning_rate": 4.837368306228597e-05,
"loss": 4.6873,
"step": 99328
},
{
"epoch": 1.01,
"learning_rate": 4.836529711477545e-05,
"loss": 4.691,
"step": 99840
},
{
"epoch": 1.01,
"learning_rate": 4.835691116726493e-05,
"loss": 4.6881,
"step": 100352
},
{
"epoch": 1.01,
"learning_rate": 4.834854159855814e-05,
"loss": 4.6706,
"step": 100864
},
{
"epoch": 1.01,
"learning_rate": 4.834017202985136e-05,
"loss": 4.6697,
"step": 101376
},
{
"epoch": 1.01,
"learning_rate": 4.833178608234084e-05,
"loss": 4.6849,
"step": 101888
},
{
"epoch": 1.01,
"learning_rate": 4.832340013483032e-05,
"loss": 4.6708,
"step": 102400
},
{
"epoch": 1.01,
"learning_rate": 4.83150141873198e-05,
"loss": 4.6686,
"step": 102912
},
{
"epoch": 1.01,
"learning_rate": 4.830662823980928e-05,
"loss": 4.6595,
"step": 103424
},
{
"epoch": 1.01,
"learning_rate": 4.829824229229876e-05,
"loss": 4.668,
"step": 103936
},
{
"epoch": 1.01,
"learning_rate": 4.828985634478824e-05,
"loss": 4.6501,
"step": 104448
},
{
"epoch": 1.01,
"learning_rate": 4.828147039727771e-05,
"loss": 4.676,
"step": 104960
},
{
"epoch": 1.01,
"learning_rate": 4.827308444976719e-05,
"loss": 4.6407,
"step": 105472
},
{
"epoch": 1.01,
"learning_rate": 4.826469850225667e-05,
"loss": 4.6604,
"step": 105984
},
{
"epoch": 1.01,
"learning_rate": 4.825631255474615e-05,
"loss": 4.6568,
"step": 106496
},
{
"epoch": 1.01,
"learning_rate": 4.824792660723563e-05,
"loss": 4.6465,
"step": 107008
},
{
"epoch": 1.01,
"learning_rate": 4.8239557038528846e-05,
"loss": 4.6573,
"step": 107520
},
{
"epoch": 1.01,
"learning_rate": 4.823118746982206e-05,
"loss": 4.6443,
"step": 108032
},
{
"epoch": 1.01,
"learning_rate": 4.8222801522311535e-05,
"loss": 4.6371,
"step": 108544
},
{
"epoch": 1.01,
"learning_rate": 4.8214415574801015e-05,
"loss": 4.6324,
"step": 109056
},
{
"epoch": 1.01,
"learning_rate": 4.8206029627290495e-05,
"loss": 4.6572,
"step": 109568
},
{
"epoch": 1.01,
"learning_rate": 4.8197643679779975e-05,
"loss": 4.6263,
"step": 110080
},
{
"epoch": 1.01,
"learning_rate": 4.8189274111073184e-05,
"loss": 4.6407,
"step": 110592
},
{
"epoch": 1.01,
"learning_rate": 4.8180888163562664e-05,
"loss": 4.6211,
"step": 111104
},
{
"epoch": 1.01,
"learning_rate": 4.8172502216052144e-05,
"loss": 4.6279,
"step": 111616
},
{
"epoch": 1.01,
"learning_rate": 4.816413264734535e-05,
"loss": 4.6396,
"step": 112128
},
{
"epoch": 1.01,
"learning_rate": 4.815574669983483e-05,
"loss": 4.6376,
"step": 112640
},
{
"epoch": 1.01,
"learning_rate": 4.814736075232432e-05,
"loss": 4.6299,
"step": 113152
},
{
"epoch": 1.01,
"learning_rate": 4.81389748048138e-05,
"loss": 4.6369,
"step": 113664
},
{
"epoch": 1.01,
"learning_rate": 4.813058885730328e-05,
"loss": 4.6391,
"step": 114176
},
{
"epoch": 1.01,
"learning_rate": 4.812220290979276e-05,
"loss": 4.6232,
"step": 114688
},
{
"epoch": 1.01,
"learning_rate": 4.811383334108597e-05,
"loss": 4.629,
"step": 115200
},
{
"epoch": 1.01,
"learning_rate": 4.810544739357545e-05,
"loss": 4.6164,
"step": 115712
},
{
"epoch": 1.01,
"learning_rate": 4.809706144606493e-05,
"loss": 4.6166,
"step": 116224
},
{
"epoch": 1.01,
"learning_rate": 4.808867549855441e-05,
"loss": 4.6118,
"step": 116736
},
{
"epoch": 1.01,
"learning_rate": 4.808028955104389e-05,
"loss": 4.619,
"step": 117248
},
{
"epoch": 1.01,
"learning_rate": 4.807190360353337e-05,
"loss": 4.6218,
"step": 117760
},
{
"epoch": 1.01,
"learning_rate": 4.806351765602285e-05,
"loss": 4.6265,
"step": 118272
},
{
"epoch": 1.01,
"learning_rate": 4.805513170851233e-05,
"loss": 4.6126,
"step": 118784
},
{
"epoch": 1.01,
"learning_rate": 4.804676213980554e-05,
"loss": 4.5958,
"step": 119296
},
{
"epoch": 1.01,
"learning_rate": 4.803837619229502e-05,
"loss": 4.597,
"step": 119808
},
{
"epoch": 1.01,
"learning_rate": 4.8029990244784504e-05,
"loss": 4.6107,
"step": 120320
},
{
"epoch": 1.01,
"learning_rate": 4.8021604297273984e-05,
"loss": 4.6047,
"step": 120832
},
{
"epoch": 1.01,
"learning_rate": 4.801323472856719e-05,
"loss": 4.6009,
"step": 121344
},
{
"epoch": 1.01,
"learning_rate": 4.800484878105667e-05,
"loss": 4.6055,
"step": 121856
},
{
"epoch": 1.02,
"learning_rate": 4.799646283354615e-05,
"loss": 4.5912,
"step": 122368
},
{
"epoch": 1.02,
"learning_rate": 4.798807688603563e-05,
"loss": 4.5958,
"step": 122880
},
{
"epoch": 1.02,
"learning_rate": 4.797969093852511e-05,
"loss": 4.5975,
"step": 123392
},
{
"epoch": 1.02,
"learning_rate": 4.797132136981832e-05,
"loss": 4.602,
"step": 123904
},
{
"epoch": 1.02,
"learning_rate": 4.79629354223078e-05,
"loss": 4.5987,
"step": 124416
},
{
"epoch": 1.02,
"learning_rate": 4.795456585360101e-05,
"loss": 4.5987,
"step": 124928
},
{
"epoch": 1.02,
"learning_rate": 4.794617990609049e-05,
"loss": 4.5804,
"step": 125440
},
{
"epoch": 1.02,
"learning_rate": 4.793779395857997e-05,
"loss": 4.5855,
"step": 125952
},
{
"epoch": 1.02,
"learning_rate": 4.792940801106946e-05,
"loss": 4.5834,
"step": 126464
},
{
"epoch": 1.02,
"learning_rate": 4.792102206355894e-05,
"loss": 4.5818,
"step": 126976
},
{
"epoch": 1.02,
"learning_rate": 4.791263611604842e-05,
"loss": 4.5961,
"step": 127488
},
{
"epoch": 1.02,
"learning_rate": 4.79042501685379e-05,
"loss": 4.5754,
"step": 128000
},
{
"epoch": 1.02,
"learning_rate": 4.789586422102737e-05,
"loss": 4.5827,
"step": 128512
},
{
"epoch": 1.02,
"learning_rate": 4.788747827351685e-05,
"loss": 4.5811,
"step": 129024
},
{
"epoch": 1.02,
"learning_rate": 4.787909232600633e-05,
"loss": 4.5756,
"step": 129536
},
{
"epoch": 1.02,
"learning_rate": 4.787070637849581e-05,
"loss": 4.5648,
"step": 130048
},
{
"epoch": 1.02,
"learning_rate": 4.7862353188592756e-05,
"loss": 4.5746,
"step": 130560
},
{
"epoch": 1.02,
"learning_rate": 4.7853967241082236e-05,
"loss": 4.5617,
"step": 131072
},
{
"epoch": 1.02,
"learning_rate": 4.7845597672375445e-05,
"loss": 4.5634,
"step": 131584
},
{
"epoch": 1.02,
"learning_rate": 4.7837211724864925e-05,
"loss": 4.5643,
"step": 132096
},
{
"epoch": 1.02,
"learning_rate": 4.782882577735441e-05,
"loss": 4.5623,
"step": 132608
},
{
"epoch": 1.02,
"learning_rate": 4.782043982984389e-05,
"loss": 4.5582,
"step": 133120
},
{
"epoch": 1.02,
"learning_rate": 4.781205388233337e-05,
"loss": 4.5675,
"step": 133632
},
{
"epoch": 1.02,
"learning_rate": 4.7803667934822844e-05,
"loss": 4.5534,
"step": 134144
},
{
"epoch": 1.02,
"learning_rate": 4.7795281987312324e-05,
"loss": 4.563,
"step": 134656
},
{
"epoch": 1.02,
"learning_rate": 4.7786896039801804e-05,
"loss": 4.5706,
"step": 135168
},
{
"epoch": 1.02,
"learning_rate": 4.7778510092291284e-05,
"loss": 4.561,
"step": 135680
},
{
"epoch": 1.02,
"learning_rate": 4.777014052358449e-05,
"loss": 4.5635,
"step": 136192
},
{
"epoch": 1.02,
"learning_rate": 4.776175457607397e-05,
"loss": 4.5564,
"step": 136704
},
{
"epoch": 1.02,
"learning_rate": 4.775336862856345e-05,
"loss": 4.5537,
"step": 137216
},
{
"epoch": 1.02,
"learning_rate": 4.774498268105293e-05,
"loss": 4.5578,
"step": 137728
},
{
"epoch": 1.02,
"learning_rate": 4.773659673354241e-05,
"loss": 4.548,
"step": 138240
},
{
"epoch": 1.02,
"learning_rate": 4.772822716483563e-05,
"loss": 4.546,
"step": 138752
},
{
"epoch": 1.02,
"learning_rate": 4.771984121732511e-05,
"loss": 4.5466,
"step": 139264
},
{
"epoch": 1.02,
"learning_rate": 4.771145526981459e-05,
"loss": 4.5494,
"step": 139776
},
{
"epoch": 1.02,
"learning_rate": 4.770306932230407e-05,
"loss": 4.5447,
"step": 140288
},
{
"epoch": 1.02,
"learning_rate": 4.769468337479355e-05,
"loss": 4.5433,
"step": 140800
},
{
"epoch": 1.02,
"learning_rate": 4.768631380608676e-05,
"loss": 4.5433,
"step": 141312
},
{
"epoch": 1.02,
"learning_rate": 4.767792785857624e-05,
"loss": 4.5372,
"step": 141824
},
{
"epoch": 1.02,
"learning_rate": 4.766955828986945e-05,
"loss": 4.5313,
"step": 142336
},
{
"epoch": 1.02,
"learning_rate": 4.766117234235893e-05,
"loss": 4.5545,
"step": 142848
},
{
"epoch": 1.02,
"learning_rate": 4.765278639484841e-05,
"loss": 4.5452,
"step": 143360
},
{
"epoch": 1.02,
"learning_rate": 4.764440044733789e-05,
"loss": 4.5428,
"step": 143872
},
{
"epoch": 1.02,
"learning_rate": 4.763601449982737e-05,
"loss": 4.5383,
"step": 144384
},
{
"epoch": 1.02,
"learning_rate": 4.762762855231685e-05,
"loss": 4.5383,
"step": 144896
},
{
"epoch": 1.02,
"learning_rate": 4.761924260480633e-05,
"loss": 4.5368,
"step": 145408
},
{
"epoch": 1.02,
"learning_rate": 4.7610856657295813e-05,
"loss": 4.5465,
"step": 145920
},
{
"epoch": 1.02,
"learning_rate": 4.760248708858902e-05,
"loss": 4.537,
"step": 146432
},
{
"epoch": 1.02,
"learning_rate": 4.75941011410785e-05,
"loss": 4.5283,
"step": 146944
},
{
"epoch": 1.02,
"learning_rate": 4.758571519356798e-05,
"loss": 4.5311,
"step": 147456
},
{
"epoch": 1.02,
"learning_rate": 4.757732924605746e-05,
"loss": 4.5245,
"step": 147968
},
{
"epoch": 1.02,
"learning_rate": 4.756895967735067e-05,
"loss": 4.5192,
"step": 148480
},
{
"epoch": 1.02,
"learning_rate": 4.756057372984015e-05,
"loss": 4.5254,
"step": 148992
},
{
"epoch": 1.02,
"learning_rate": 4.755218778232963e-05,
"loss": 4.5308,
"step": 149504
},
{
"epoch": 1.02,
"learning_rate": 4.754380183481911e-05,
"loss": 4.5179,
"step": 150016
},
{
"epoch": 1.02,
"learning_rate": 4.753541588730859e-05,
"loss": 4.5298,
"step": 150528
},
{
"epoch": 1.02,
"learning_rate": 4.752702993979807e-05,
"loss": 4.5252,
"step": 151040
},
{
"epoch": 1.02,
"learning_rate": 4.751864399228755e-05,
"loss": 4.5194,
"step": 151552
},
{
"epoch": 1.02,
"learning_rate": 4.751025804477703e-05,
"loss": 4.5259,
"step": 152064
},
{
"epoch": 1.02,
"learning_rate": 4.750188847607025e-05,
"loss": 4.5113,
"step": 152576
},
{
"epoch": 1.03,
"eval_loss": 4.48332405090332,
"eval_runtime": 318.5739,
"eval_samples_per_second": 1197.81,
"eval_steps_per_second": 37.432,
"step": 152640
}
],
"logging_steps": 512,
"max_steps": 3052726,
"num_train_epochs": 9223372036854775807,
"save_steps": 10,
"total_flos": 6.306206165175117e+16,
"trial_name": null,
"trial_params": null
}