old-full-lstm-4 / checkpoint-152638 /trainer_state.json
abhinavp's picture
Training in progress, epoch 0, checkpoint
49d6808 verified
{
"best_metric": 4.480572700500488,
"best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/full/lstm/4/checkpoints/checkpoint-152638",
"epoch": 0.025000278439663435,
"eval_steps": 10,
"global_step": 152638,
"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": 10.8202,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 4.999161405248948e-05,
"loss": 7.5642,
"step": 512
},
{
"epoch": 0.0,
"learning_rate": 4.998322810497896e-05,
"loss": 7.0561,
"step": 1024
},
{
"epoch": 0.0,
"learning_rate": 4.997484215746844e-05,
"loss": 6.9848,
"step": 1536
},
{
"epoch": 0.0,
"learning_rate": 4.996645620995792e-05,
"loss": 6.9516,
"step": 2048
},
{
"epoch": 0.0,
"learning_rate": 4.99580702624474e-05,
"loss": 6.8317,
"step": 2560
},
{
"epoch": 0.0,
"learning_rate": 4.994968431493688e-05,
"loss": 6.7218,
"step": 3072
},
{
"epoch": 0.0,
"learning_rate": 4.994129836742636e-05,
"loss": 6.6162,
"step": 3584
},
{
"epoch": 0.0,
"learning_rate": 4.993292879871958e-05,
"loss": 6.5336,
"step": 4096
},
{
"epoch": 0.0,
"learning_rate": 4.992454285120906e-05,
"loss": 6.4403,
"step": 4608
},
{
"epoch": 0.0,
"learning_rate": 4.991615690369854e-05,
"loss": 6.3866,
"step": 5120
},
{
"epoch": 0.0,
"learning_rate": 4.990777095618801e-05,
"loss": 6.3152,
"step": 5632
},
{
"epoch": 0.0,
"learning_rate": 4.989938500867749e-05,
"loss": 6.2499,
"step": 6144
},
{
"epoch": 0.0,
"learning_rate": 4.989099906116697e-05,
"loss": 6.185,
"step": 6656
},
{
"epoch": 0.0,
"learning_rate": 4.988261311365645e-05,
"loss": 6.135,
"step": 7168
},
{
"epoch": 0.0,
"learning_rate": 4.987422716614593e-05,
"loss": 6.0919,
"step": 7680
},
{
"epoch": 0.0,
"learning_rate": 4.986585759743914e-05,
"loss": 6.0468,
"step": 8192
},
{
"epoch": 0.0,
"learning_rate": 4.985747164992862e-05,
"loss": 5.9949,
"step": 8704
},
{
"epoch": 0.0,
"learning_rate": 4.98490857024181e-05,
"loss": 5.96,
"step": 9216
},
{
"epoch": 0.0,
"learning_rate": 4.984069975490758e-05,
"loss": 5.926,
"step": 9728
},
{
"epoch": 0.0,
"learning_rate": 4.983231380739706e-05,
"loss": 5.881,
"step": 10240
},
{
"epoch": 0.0,
"learning_rate": 4.9823927859886547e-05,
"loss": 5.8466,
"step": 10752
},
{
"epoch": 0.0,
"learning_rate": 4.9815541912376026e-05,
"loss": 5.8268,
"step": 11264
},
{
"epoch": 0.0,
"learning_rate": 4.9807155964865506e-05,
"loss": 5.7903,
"step": 11776
},
{
"epoch": 0.0,
"learning_rate": 4.9798770017354986e-05,
"loss": 5.7694,
"step": 12288
},
{
"epoch": 0.0,
"learning_rate": 4.9790384069844466e-05,
"loss": 5.7434,
"step": 12800
},
{
"epoch": 0.0,
"learning_rate": 4.9781998122333946e-05,
"loss": 5.7206,
"step": 13312
},
{
"epoch": 0.0,
"learning_rate": 4.9773612174823426e-05,
"loss": 5.6927,
"step": 13824
},
{
"epoch": 0.0,
"learning_rate": 4.9765242606116635e-05,
"loss": 5.6587,
"step": 14336
},
{
"epoch": 0.0,
"learning_rate": 4.9756856658606115e-05,
"loss": 5.6528,
"step": 14848
},
{
"epoch": 0.01,
"learning_rate": 4.9748470711095595e-05,
"loss": 5.6204,
"step": 15360
},
{
"epoch": 0.01,
"learning_rate": 4.974008476358507e-05,
"loss": 5.6044,
"step": 15872
},
{
"epoch": 0.01,
"learning_rate": 4.9731715194878284e-05,
"loss": 5.5801,
"step": 16384
},
{
"epoch": 0.01,
"learning_rate": 4.9723329247367764e-05,
"loss": 5.5763,
"step": 16896
},
{
"epoch": 0.01,
"learning_rate": 4.9714943299857244e-05,
"loss": 5.5452,
"step": 17408
},
{
"epoch": 0.01,
"learning_rate": 4.9706557352346724e-05,
"loss": 5.5375,
"step": 17920
},
{
"epoch": 0.01,
"learning_rate": 4.9698171404836204e-05,
"loss": 5.5231,
"step": 18432
},
{
"epoch": 0.01,
"learning_rate": 4.968980183612942e-05,
"loss": 5.5038,
"step": 18944
},
{
"epoch": 0.01,
"learning_rate": 4.968143226742263e-05,
"loss": 5.4937,
"step": 19456
},
{
"epoch": 0.01,
"learning_rate": 4.967304631991211e-05,
"loss": 5.4722,
"step": 19968
},
{
"epoch": 0.01,
"learning_rate": 4.966466037240159e-05,
"loss": 5.4475,
"step": 20480
},
{
"epoch": 0.01,
"learning_rate": 4.965627442489107e-05,
"loss": 5.4423,
"step": 20992
},
{
"epoch": 0.01,
"learning_rate": 4.964788847738054e-05,
"loss": 5.4142,
"step": 21504
},
{
"epoch": 0.01,
"learning_rate": 4.963950252987002e-05,
"loss": 5.4262,
"step": 22016
},
{
"epoch": 0.01,
"learning_rate": 4.96311165823595e-05,
"loss": 5.3996,
"step": 22528
},
{
"epoch": 0.01,
"learning_rate": 4.962273063484898e-05,
"loss": 5.3943,
"step": 23040
},
{
"epoch": 0.01,
"learning_rate": 4.96143610661422e-05,
"loss": 5.3795,
"step": 23552
},
{
"epoch": 0.01,
"learning_rate": 4.960597511863168e-05,
"loss": 5.3756,
"step": 24064
},
{
"epoch": 0.01,
"learning_rate": 4.959758917112116e-05,
"loss": 5.3695,
"step": 24576
},
{
"epoch": 0.01,
"learning_rate": 4.958920322361064e-05,
"loss": 5.3395,
"step": 25088
},
{
"epoch": 0.01,
"learning_rate": 4.958081727610012e-05,
"loss": 5.3284,
"step": 25600
},
{
"epoch": 0.01,
"learning_rate": 4.957244770739333e-05,
"loss": 5.3211,
"step": 26112
},
{
"epoch": 0.01,
"learning_rate": 4.956406175988281e-05,
"loss": 5.3282,
"step": 26624
},
{
"epoch": 0.01,
"learning_rate": 4.955567581237229e-05,
"loss": 5.3042,
"step": 27136
},
{
"epoch": 0.01,
"learning_rate": 4.954728986486177e-05,
"loss": 5.3108,
"step": 27648
},
{
"epoch": 0.01,
"learning_rate": 4.9538920296154976e-05,
"loss": 5.2807,
"step": 28160
},
{
"epoch": 0.01,
"learning_rate": 4.9530534348644456e-05,
"loss": 5.2767,
"step": 28672
},
{
"epoch": 0.01,
"learning_rate": 4.9522148401133936e-05,
"loss": 5.264,
"step": 29184
},
{
"epoch": 0.01,
"learning_rate": 4.951376245362342e-05,
"loss": 5.2578,
"step": 29696
},
{
"epoch": 0.01,
"learning_rate": 4.95053765061129e-05,
"loss": 5.255,
"step": 30208
},
{
"epoch": 0.01,
"learning_rate": 4.949700693740611e-05,
"loss": 5.2339,
"step": 30720
},
{
"epoch": 0.01,
"learning_rate": 4.948862098989559e-05,
"loss": 5.2284,
"step": 31232
},
{
"epoch": 0.01,
"learning_rate": 4.948023504238507e-05,
"loss": 5.2327,
"step": 31744
},
{
"epoch": 0.01,
"learning_rate": 4.947184909487455e-05,
"loss": 5.2108,
"step": 32256
},
{
"epoch": 0.01,
"learning_rate": 4.946346314736403e-05,
"loss": 5.2101,
"step": 32768
},
{
"epoch": 0.01,
"learning_rate": 4.945507719985351e-05,
"loss": 5.198,
"step": 33280
},
{
"epoch": 0.01,
"learning_rate": 4.944669125234299e-05,
"loss": 5.198,
"step": 33792
},
{
"epoch": 0.01,
"learning_rate": 4.943830530483247e-05,
"loss": 5.182,
"step": 34304
},
{
"epoch": 0.01,
"learning_rate": 4.942993573612568e-05,
"loss": 5.1854,
"step": 34816
},
{
"epoch": 0.01,
"learning_rate": 4.942156616741889e-05,
"loss": 5.1715,
"step": 35328
},
{
"epoch": 0.01,
"learning_rate": 4.941318021990837e-05,
"loss": 5.162,
"step": 35840
},
{
"epoch": 0.01,
"learning_rate": 4.9404794272397856e-05,
"loss": 5.1647,
"step": 36352
},
{
"epoch": 0.01,
"learning_rate": 4.9396408324887336e-05,
"loss": 5.1511,
"step": 36864
},
{
"epoch": 0.01,
"learning_rate": 4.9388022377376816e-05,
"loss": 5.1585,
"step": 37376
},
{
"epoch": 0.01,
"learning_rate": 4.9379636429866296e-05,
"loss": 5.1559,
"step": 37888
},
{
"epoch": 0.01,
"learning_rate": 4.9371266861159505e-05,
"loss": 5.1459,
"step": 38400
},
{
"epoch": 0.01,
"learning_rate": 4.9362880913648985e-05,
"loss": 5.1271,
"step": 38912
},
{
"epoch": 0.01,
"learning_rate": 4.9354494966138465e-05,
"loss": 5.1115,
"step": 39424
},
{
"epoch": 0.01,
"learning_rate": 4.9346109018627945e-05,
"loss": 5.1044,
"step": 39936
},
{
"epoch": 0.01,
"learning_rate": 4.9337723071117425e-05,
"loss": 5.1102,
"step": 40448
},
{
"epoch": 0.01,
"learning_rate": 4.9329353502410634e-05,
"loss": 5.1134,
"step": 40960
},
{
"epoch": 0.01,
"learning_rate": 4.9320967554900114e-05,
"loss": 5.1064,
"step": 41472
},
{
"epoch": 0.01,
"learning_rate": 4.9312581607389594e-05,
"loss": 5.0816,
"step": 41984
},
{
"epoch": 0.01,
"learning_rate": 4.9304195659879074e-05,
"loss": 5.0762,
"step": 42496
},
{
"epoch": 0.01,
"learning_rate": 4.9295809712368554e-05,
"loss": 5.085,
"step": 43008
},
{
"epoch": 0.01,
"learning_rate": 4.928744014366177e-05,
"loss": 5.0699,
"step": 43520
},
{
"epoch": 0.01,
"learning_rate": 4.927905419615125e-05,
"loss": 5.0698,
"step": 44032
},
{
"epoch": 0.01,
"learning_rate": 4.927066824864073e-05,
"loss": 5.0613,
"step": 44544
},
{
"epoch": 0.01,
"learning_rate": 4.92622823011302e-05,
"loss": 5.063,
"step": 45056
},
{
"epoch": 0.01,
"learning_rate": 4.925389635361968e-05,
"loss": 5.0524,
"step": 45568
},
{
"epoch": 0.02,
"learning_rate": 4.92455267849129e-05,
"loss": 5.0435,
"step": 46080
},
{
"epoch": 0.02,
"learning_rate": 4.923714083740238e-05,
"loss": 5.0299,
"step": 46592
},
{
"epoch": 0.02,
"learning_rate": 4.922875488989185e-05,
"loss": 5.0439,
"step": 47104
},
{
"epoch": 0.02,
"learning_rate": 4.922038532118507e-05,
"loss": 5.0289,
"step": 47616
},
{
"epoch": 0.02,
"learning_rate": 4.921199937367455e-05,
"loss": 5.0127,
"step": 48128
},
{
"epoch": 0.02,
"learning_rate": 4.920361342616403e-05,
"loss": 5.0113,
"step": 48640
},
{
"epoch": 0.02,
"learning_rate": 4.919522747865351e-05,
"loss": 5.0246,
"step": 49152
},
{
"epoch": 0.02,
"learning_rate": 4.918684153114299e-05,
"loss": 5.0149,
"step": 49664
},
{
"epoch": 0.02,
"learning_rate": 4.917845558363247e-05,
"loss": 5.0059,
"step": 50176
},
{
"epoch": 0.02,
"learning_rate": 4.9170086014925676e-05,
"loss": 4.9942,
"step": 50688
},
{
"epoch": 0.02,
"learning_rate": 4.9161700067415156e-05,
"loss": 4.998,
"step": 51200
},
{
"epoch": 0.02,
"learning_rate": 4.915333049870837e-05,
"loss": 4.9769,
"step": 51712
},
{
"epoch": 0.02,
"learning_rate": 4.914494455119785e-05,
"loss": 4.9922,
"step": 52224
},
{
"epoch": 0.02,
"learning_rate": 4.9136558603687325e-05,
"loss": 4.9944,
"step": 52736
},
{
"epoch": 0.02,
"learning_rate": 4.9128172656176805e-05,
"loss": 4.9695,
"step": 53248
},
{
"epoch": 0.02,
"learning_rate": 4.9119786708666285e-05,
"loss": 4.9626,
"step": 53760
},
{
"epoch": 0.02,
"learning_rate": 4.9111400761155765e-05,
"loss": 4.94,
"step": 54272
},
{
"epoch": 0.02,
"learning_rate": 4.9103014813645245e-05,
"loss": 4.9518,
"step": 54784
},
{
"epoch": 0.02,
"learning_rate": 4.909462886613473e-05,
"loss": 4.9525,
"step": 55296
},
{
"epoch": 0.02,
"learning_rate": 4.908624291862421e-05,
"loss": 4.9467,
"step": 55808
},
{
"epoch": 0.02,
"learning_rate": 4.907785697111369e-05,
"loss": 4.9475,
"step": 56320
},
{
"epoch": 0.02,
"learning_rate": 4.906947102360317e-05,
"loss": 4.9316,
"step": 56832
},
{
"epoch": 0.02,
"learning_rate": 4.906110145489638e-05,
"loss": 4.9326,
"step": 57344
},
{
"epoch": 0.02,
"learning_rate": 4.905271550738586e-05,
"loss": 4.9338,
"step": 57856
},
{
"epoch": 0.02,
"learning_rate": 4.904432955987534e-05,
"loss": 4.9357,
"step": 58368
},
{
"epoch": 0.02,
"learning_rate": 4.903595999116855e-05,
"loss": 4.9228,
"step": 58880
},
{
"epoch": 0.02,
"learning_rate": 4.902757404365803e-05,
"loss": 4.9119,
"step": 59392
},
{
"epoch": 0.02,
"learning_rate": 4.901918809614751e-05,
"loss": 4.9132,
"step": 59904
},
{
"epoch": 0.02,
"learning_rate": 4.901080214863699e-05,
"loss": 4.9165,
"step": 60416
},
{
"epoch": 0.02,
"learning_rate": 4.900241620112647e-05,
"loss": 4.9104,
"step": 60928
},
{
"epoch": 0.02,
"learning_rate": 4.899403025361595e-05,
"loss": 4.9059,
"step": 61440
},
{
"epoch": 0.02,
"learning_rate": 4.898564430610543e-05,
"loss": 4.8968,
"step": 61952
},
{
"epoch": 0.02,
"learning_rate": 4.897725835859491e-05,
"loss": 4.8881,
"step": 62464
},
{
"epoch": 0.02,
"learning_rate": 4.8968872411084396e-05,
"loss": 4.8811,
"step": 62976
},
{
"epoch": 0.02,
"learning_rate": 4.896048646357387e-05,
"loss": 4.8877,
"step": 63488
},
{
"epoch": 0.02,
"learning_rate": 4.895210051606335e-05,
"loss": 4.8874,
"step": 64000
},
{
"epoch": 0.02,
"learning_rate": 4.894371456855283e-05,
"loss": 4.8881,
"step": 64512
},
{
"epoch": 0.02,
"learning_rate": 4.893532862104231e-05,
"loss": 4.8623,
"step": 65024
},
{
"epoch": 0.02,
"learning_rate": 4.892694267353179e-05,
"loss": 4.876,
"step": 65536
},
{
"epoch": 0.02,
"learning_rate": 4.8918573104825e-05,
"loss": 4.8748,
"step": 66048
},
{
"epoch": 0.02,
"learning_rate": 4.891018715731448e-05,
"loss": 4.8618,
"step": 66560
},
{
"epoch": 0.02,
"learning_rate": 4.890180120980396e-05,
"loss": 4.8774,
"step": 67072
},
{
"epoch": 0.02,
"learning_rate": 4.889341526229344e-05,
"loss": 4.8571,
"step": 67584
},
{
"epoch": 0.02,
"learning_rate": 4.888502931478292e-05,
"loss": 4.8708,
"step": 68096
},
{
"epoch": 0.02,
"learning_rate": 4.8876659746076134e-05,
"loss": 4.858,
"step": 68608
},
{
"epoch": 0.02,
"learning_rate": 4.8868273798565614e-05,
"loss": 4.8387,
"step": 69120
},
{
"epoch": 0.02,
"learning_rate": 4.8859887851055094e-05,
"loss": 4.8374,
"step": 69632
},
{
"epoch": 0.02,
"learning_rate": 4.8851501903544574e-05,
"loss": 4.8375,
"step": 70144
},
{
"epoch": 0.02,
"learning_rate": 4.8843115956034054e-05,
"loss": 4.8411,
"step": 70656
},
{
"epoch": 0.02,
"learning_rate": 4.8834730008523534e-05,
"loss": 4.8453,
"step": 71168
},
{
"epoch": 0.02,
"learning_rate": 4.8826344061013014e-05,
"loss": 4.8415,
"step": 71680
},
{
"epoch": 0.02,
"learning_rate": 4.8817958113502494e-05,
"loss": 4.8348,
"step": 72192
},
{
"epoch": 0.02,
"learning_rate": 4.88095885447957e-05,
"loss": 4.8253,
"step": 72704
},
{
"epoch": 0.02,
"learning_rate": 4.880121897608891e-05,
"loss": 4.8133,
"step": 73216
},
{
"epoch": 0.02,
"learning_rate": 4.879283302857839e-05,
"loss": 4.8126,
"step": 73728
},
{
"epoch": 0.02,
"learning_rate": 4.878444708106787e-05,
"loss": 4.8184,
"step": 74240
},
{
"epoch": 0.02,
"learning_rate": 4.877606113355735e-05,
"loss": 4.8269,
"step": 74752
},
{
"epoch": 0.02,
"learning_rate": 4.876767518604683e-05,
"loss": 4.8172,
"step": 75264
},
{
"epoch": 0.02,
"learning_rate": 4.875928923853632e-05,
"loss": 4.8004,
"step": 75776
},
{
"epoch": 0.02,
"learning_rate": 4.87509032910258e-05,
"loss": 4.8004,
"step": 76288
},
{
"epoch": 0.03,
"eval_loss": 4.7640604972839355,
"eval_runtime": 557.5755,
"eval_samples_per_second": 684.376,
"eval_steps_per_second": 21.387,
"step": 76319
},
{
"epoch": 0.0,
"learning_rate": 4.874251734351528e-05,
"loss": 4.8081,
"step": 76800
},
{
"epoch": 0.0,
"learning_rate": 4.873413139600475e-05,
"loss": 4.8005,
"step": 77312
},
{
"epoch": 0.0,
"learning_rate": 4.872574544849423e-05,
"loss": 4.7917,
"step": 77824
},
{
"epoch": 0.0,
"learning_rate": 4.871735950098371e-05,
"loss": 4.7901,
"step": 78336
},
{
"epoch": 0.0,
"learning_rate": 4.870897355347319e-05,
"loss": 4.7813,
"step": 78848
},
{
"epoch": 0.0,
"learning_rate": 4.870058760596267e-05,
"loss": 4.7651,
"step": 79360
},
{
"epoch": 0.0,
"learning_rate": 4.869220165845215e-05,
"loss": 4.7739,
"step": 79872
},
{
"epoch": 0.0,
"learning_rate": 4.868381571094163e-05,
"loss": 4.7833,
"step": 80384
},
{
"epoch": 0.0,
"learning_rate": 4.867542976343111e-05,
"loss": 4.7657,
"step": 80896
},
{
"epoch": 0.0,
"learning_rate": 4.866704381592059e-05,
"loss": 4.774,
"step": 81408
},
{
"epoch": 0.0,
"learning_rate": 4.865865786841007e-05,
"loss": 4.7773,
"step": 81920
},
{
"epoch": 0.0,
"learning_rate": 4.865027192089955e-05,
"loss": 4.7754,
"step": 82432
},
{
"epoch": 0.0,
"learning_rate": 4.864188597338903e-05,
"loss": 4.7656,
"step": 82944
},
{
"epoch": 0.0,
"learning_rate": 4.863350002587852e-05,
"loss": 4.7627,
"step": 83456
},
{
"epoch": 0.0,
"learning_rate": 4.8625114078368e-05,
"loss": 4.7588,
"step": 83968
},
{
"epoch": 0.0,
"learning_rate": 4.861672813085747e-05,
"loss": 4.7492,
"step": 84480
},
{
"epoch": 0.0,
"learning_rate": 4.860835856215069e-05,
"loss": 4.7519,
"step": 84992
},
{
"epoch": 0.0,
"learning_rate": 4.859997261464017e-05,
"loss": 4.7529,
"step": 85504
},
{
"epoch": 0.0,
"learning_rate": 4.8591603045933376e-05,
"loss": 4.7516,
"step": 86016
},
{
"epoch": 0.0,
"learning_rate": 4.8583217098422856e-05,
"loss": 4.7499,
"step": 86528
},
{
"epoch": 0.0,
"learning_rate": 4.8574831150912336e-05,
"loss": 4.7374,
"step": 87040
},
{
"epoch": 0.0,
"learning_rate": 4.8566445203401816e-05,
"loss": 4.7492,
"step": 87552
},
{
"epoch": 0.0,
"learning_rate": 4.855805925589129e-05,
"loss": 4.7341,
"step": 88064
},
{
"epoch": 0.0,
"learning_rate": 4.8549689687184505e-05,
"loss": 4.7438,
"step": 88576
},
{
"epoch": 0.0,
"learning_rate": 4.8541303739673985e-05,
"loss": 4.7327,
"step": 89088
},
{
"epoch": 0.0,
"learning_rate": 4.8532917792163465e-05,
"loss": 4.7319,
"step": 89600
},
{
"epoch": 0.0,
"learning_rate": 4.8524531844652945e-05,
"loss": 4.7353,
"step": 90112
},
{
"epoch": 0.0,
"learning_rate": 4.8516145897142425e-05,
"loss": 4.7153,
"step": 90624
},
{
"epoch": 0.0,
"learning_rate": 4.850777632843564e-05,
"loss": 4.7304,
"step": 91136
},
{
"epoch": 0.01,
"learning_rate": 4.8499390380925114e-05,
"loss": 4.7108,
"step": 91648
},
{
"epoch": 0.01,
"learning_rate": 4.8491004433414594e-05,
"loss": 4.7166,
"step": 92160
},
{
"epoch": 0.01,
"learning_rate": 4.8482618485904074e-05,
"loss": 4.712,
"step": 92672
},
{
"epoch": 0.01,
"learning_rate": 4.8474232538393554e-05,
"loss": 4.7209,
"step": 93184
},
{
"epoch": 0.01,
"learning_rate": 4.8465846590883034e-05,
"loss": 4.7075,
"step": 93696
},
{
"epoch": 0.01,
"learning_rate": 4.8457460643372514e-05,
"loss": 4.7017,
"step": 94208
},
{
"epoch": 0.01,
"learning_rate": 4.8449074695861993e-05,
"loss": 4.7185,
"step": 94720
},
{
"epoch": 0.01,
"learning_rate": 4.8440688748351473e-05,
"loss": 4.7052,
"step": 95232
},
{
"epoch": 0.01,
"learning_rate": 4.843231917964469e-05,
"loss": 4.711,
"step": 95744
},
{
"epoch": 0.01,
"learning_rate": 4.842393323213417e-05,
"loss": 4.6993,
"step": 96256
},
{
"epoch": 0.01,
"learning_rate": 4.841554728462365e-05,
"loss": 4.6847,
"step": 96768
},
{
"epoch": 0.01,
"learning_rate": 4.840716133711313e-05,
"loss": 4.6887,
"step": 97280
},
{
"epoch": 0.01,
"learning_rate": 4.839877538960261e-05,
"loss": 4.6801,
"step": 97792
},
{
"epoch": 0.01,
"learning_rate": 4.839038944209209e-05,
"loss": 4.6946,
"step": 98304
},
{
"epoch": 0.01,
"learning_rate": 4.83820198733853e-05,
"loss": 4.6843,
"step": 98816
},
{
"epoch": 0.01,
"learning_rate": 4.837363392587478e-05,
"loss": 4.6909,
"step": 99328
},
{
"epoch": 0.01,
"learning_rate": 4.836524797836426e-05,
"loss": 4.6826,
"step": 99840
},
{
"epoch": 0.01,
"learning_rate": 4.835686203085374e-05,
"loss": 4.6879,
"step": 100352
},
{
"epoch": 0.01,
"learning_rate": 4.834847608334322e-05,
"loss": 4.6849,
"step": 100864
},
{
"epoch": 0.01,
"learning_rate": 4.834010651463643e-05,
"loss": 4.6609,
"step": 101376
},
{
"epoch": 0.01,
"learning_rate": 4.833173694592964e-05,
"loss": 4.6653,
"step": 101888
},
{
"epoch": 0.01,
"learning_rate": 4.832335099841912e-05,
"loss": 4.6635,
"step": 102400
},
{
"epoch": 0.01,
"learning_rate": 4.83149650509086e-05,
"loss": 4.6761,
"step": 102912
},
{
"epoch": 0.01,
"learning_rate": 4.830657910339808e-05,
"loss": 4.6551,
"step": 103424
},
{
"epoch": 0.01,
"learning_rate": 4.829820953469129e-05,
"loss": 4.6784,
"step": 103936
},
{
"epoch": 0.01,
"learning_rate": 4.828982358718077e-05,
"loss": 4.6518,
"step": 104448
},
{
"epoch": 0.01,
"learning_rate": 4.828143763967025e-05,
"loss": 4.6514,
"step": 104960
},
{
"epoch": 0.01,
"learning_rate": 4.827305169215973e-05,
"loss": 4.6557,
"step": 105472
},
{
"epoch": 0.01,
"learning_rate": 4.826466574464921e-05,
"loss": 4.6498,
"step": 105984
},
{
"epoch": 0.01,
"learning_rate": 4.825627979713869e-05,
"loss": 4.651,
"step": 106496
},
{
"epoch": 0.01,
"learning_rate": 4.824789384962817e-05,
"loss": 4.644,
"step": 107008
},
{
"epoch": 0.01,
"learning_rate": 4.823950790211765e-05,
"loss": 4.6407,
"step": 107520
},
{
"epoch": 0.01,
"learning_rate": 4.8231121954607125e-05,
"loss": 4.6519,
"step": 108032
},
{
"epoch": 0.01,
"learning_rate": 4.822273600709661e-05,
"loss": 4.6366,
"step": 108544
},
{
"epoch": 0.01,
"learning_rate": 4.821435005958609e-05,
"loss": 4.638,
"step": 109056
},
{
"epoch": 0.01,
"learning_rate": 4.82059804908793e-05,
"loss": 4.6354,
"step": 109568
},
{
"epoch": 0.01,
"learning_rate": 4.819759454336878e-05,
"loss": 4.6399,
"step": 110080
},
{
"epoch": 0.01,
"learning_rate": 4.818920859585826e-05,
"loss": 4.6226,
"step": 110592
},
{
"epoch": 0.01,
"learning_rate": 4.818082264834774e-05,
"loss": 4.643,
"step": 111104
},
{
"epoch": 0.01,
"learning_rate": 4.817245307964095e-05,
"loss": 4.6293,
"step": 111616
},
{
"epoch": 0.01,
"learning_rate": 4.816406713213043e-05,
"loss": 4.6273,
"step": 112128
},
{
"epoch": 0.01,
"learning_rate": 4.815568118461991e-05,
"loss": 4.6335,
"step": 112640
},
{
"epoch": 0.01,
"learning_rate": 4.814729523710939e-05,
"loss": 4.6205,
"step": 113152
},
{
"epoch": 0.01,
"learning_rate": 4.813890928959887e-05,
"loss": 4.6381,
"step": 113664
},
{
"epoch": 0.01,
"learning_rate": 4.813053972089208e-05,
"loss": 4.6429,
"step": 114176
},
{
"epoch": 0.01,
"learning_rate": 4.8122153773381565e-05,
"loss": 4.6364,
"step": 114688
},
{
"epoch": 0.01,
"learning_rate": 4.8113784204674774e-05,
"loss": 4.6166,
"step": 115200
},
{
"epoch": 0.01,
"learning_rate": 4.8105398257164254e-05,
"loss": 4.6117,
"step": 115712
},
{
"epoch": 0.01,
"learning_rate": 4.8097012309653734e-05,
"loss": 4.6079,
"step": 116224
},
{
"epoch": 0.01,
"learning_rate": 4.8088626362143214e-05,
"loss": 4.6183,
"step": 116736
},
{
"epoch": 0.01,
"learning_rate": 4.8080240414632694e-05,
"loss": 4.6282,
"step": 117248
},
{
"epoch": 0.01,
"learning_rate": 4.8071854467122174e-05,
"loss": 4.6213,
"step": 117760
},
{
"epoch": 0.01,
"learning_rate": 4.8063468519611654e-05,
"loss": 4.6067,
"step": 118272
},
{
"epoch": 0.01,
"learning_rate": 4.8055082572101134e-05,
"loss": 4.5945,
"step": 118784
},
{
"epoch": 0.01,
"learning_rate": 4.8046696624590614e-05,
"loss": 4.6138,
"step": 119296
},
{
"epoch": 0.01,
"learning_rate": 4.8038310677080094e-05,
"loss": 4.5986,
"step": 119808
},
{
"epoch": 0.01,
"learning_rate": 4.8029924729569574e-05,
"loss": 4.6017,
"step": 120320
},
{
"epoch": 0.01,
"learning_rate": 4.802155516086278e-05,
"loss": 4.6045,
"step": 120832
},
{
"epoch": 0.01,
"learning_rate": 4.8013185592156e-05,
"loss": 4.6106,
"step": 121344
},
{
"epoch": 0.01,
"learning_rate": 4.800479964464548e-05,
"loss": 4.6002,
"step": 121856
},
{
"epoch": 0.02,
"learning_rate": 4.799641369713496e-05,
"loss": 4.5959,
"step": 122368
},
{
"epoch": 0.02,
"learning_rate": 4.798802774962444e-05,
"loss": 4.5874,
"step": 122880
},
{
"epoch": 0.02,
"learning_rate": 4.797964180211392e-05,
"loss": 4.5988,
"step": 123392
},
{
"epoch": 0.02,
"learning_rate": 4.79712558546034e-05,
"loss": 4.5983,
"step": 123904
},
{
"epoch": 0.02,
"learning_rate": 4.796286990709288e-05,
"loss": 4.5736,
"step": 124416
},
{
"epoch": 0.02,
"learning_rate": 4.795448395958236e-05,
"loss": 4.5826,
"step": 124928
},
{
"epoch": 0.02,
"learning_rate": 4.794609801207184e-05,
"loss": 4.5981,
"step": 125440
},
{
"epoch": 0.02,
"learning_rate": 4.793772844336505e-05,
"loss": 4.5993,
"step": 125952
},
{
"epoch": 0.02,
"learning_rate": 4.792934249585453e-05,
"loss": 4.5856,
"step": 126464
},
{
"epoch": 0.02,
"learning_rate": 4.792095654834401e-05,
"loss": 4.581,
"step": 126976
},
{
"epoch": 0.02,
"learning_rate": 4.791257060083349e-05,
"loss": 4.5896,
"step": 127488
},
{
"epoch": 0.02,
"learning_rate": 4.7904201032126697e-05,
"loss": 4.5702,
"step": 128000
},
{
"epoch": 0.02,
"learning_rate": 4.789581508461618e-05,
"loss": 4.5854,
"step": 128512
},
{
"epoch": 0.02,
"learning_rate": 4.788742913710566e-05,
"loss": 4.5887,
"step": 129024
},
{
"epoch": 0.02,
"learning_rate": 4.7879043189595136e-05,
"loss": 4.5766,
"step": 129536
},
{
"epoch": 0.02,
"learning_rate": 4.7870657242084616e-05,
"loss": 4.5751,
"step": 130048
},
{
"epoch": 0.02,
"learning_rate": 4.7862271294574096e-05,
"loss": 4.5525,
"step": 130560
},
{
"epoch": 0.02,
"learning_rate": 4.7853885347063576e-05,
"loss": 4.559,
"step": 131072
},
{
"epoch": 0.02,
"learning_rate": 4.7845499399553056e-05,
"loss": 4.5642,
"step": 131584
},
{
"epoch": 0.02,
"learning_rate": 4.783714620965e-05,
"loss": 4.5675,
"step": 132096
},
{
"epoch": 0.02,
"learning_rate": 4.782877664094321e-05,
"loss": 4.5715,
"step": 132608
},
{
"epoch": 0.02,
"learning_rate": 4.782039069343269e-05,
"loss": 4.5585,
"step": 133120
},
{
"epoch": 0.02,
"learning_rate": 4.781200474592217e-05,
"loss": 4.5562,
"step": 133632
},
{
"epoch": 0.02,
"learning_rate": 4.780361879841165e-05,
"loss": 4.5631,
"step": 134144
},
{
"epoch": 0.02,
"learning_rate": 4.779523285090114e-05,
"loss": 4.57,
"step": 134656
},
{
"epoch": 0.02,
"learning_rate": 4.778684690339061e-05,
"loss": 4.557,
"step": 135168
},
{
"epoch": 0.02,
"learning_rate": 4.777846095588009e-05,
"loss": 4.5568,
"step": 135680
},
{
"epoch": 0.02,
"learning_rate": 4.777007500836957e-05,
"loss": 4.5543,
"step": 136192
},
{
"epoch": 0.02,
"learning_rate": 4.7761705439662786e-05,
"loss": 4.5576,
"step": 136704
},
{
"epoch": 0.02,
"learning_rate": 4.775331949215226e-05,
"loss": 4.5573,
"step": 137216
},
{
"epoch": 0.02,
"learning_rate": 4.774493354464174e-05,
"loss": 4.5507,
"step": 137728
},
{
"epoch": 0.02,
"learning_rate": 4.773654759713122e-05,
"loss": 4.5492,
"step": 138240
},
{
"epoch": 0.02,
"learning_rate": 4.77281616496207e-05,
"loss": 4.5444,
"step": 138752
},
{
"epoch": 0.02,
"learning_rate": 4.771977570211018e-05,
"loss": 4.5384,
"step": 139264
},
{
"epoch": 0.02,
"learning_rate": 4.7711422512207124e-05,
"loss": 4.5446,
"step": 139776
},
{
"epoch": 0.02,
"learning_rate": 4.7703036564696604e-05,
"loss": 4.5483,
"step": 140288
},
{
"epoch": 0.02,
"learning_rate": 4.7694650617186084e-05,
"loss": 4.5505,
"step": 140800
},
{
"epoch": 0.02,
"learning_rate": 4.7686264669675564e-05,
"loss": 4.5334,
"step": 141312
},
{
"epoch": 0.02,
"learning_rate": 4.767789510096878e-05,
"loss": 4.5408,
"step": 141824
},
{
"epoch": 0.02,
"learning_rate": 4.766950915345826e-05,
"loss": 4.5505,
"step": 142336
},
{
"epoch": 0.02,
"learning_rate": 4.766112320594773e-05,
"loss": 4.5363,
"step": 142848
},
{
"epoch": 0.02,
"learning_rate": 4.765273725843721e-05,
"loss": 4.5483,
"step": 143360
},
{
"epoch": 0.02,
"learning_rate": 4.764435131092669e-05,
"loss": 4.5348,
"step": 143872
},
{
"epoch": 0.02,
"learning_rate": 4.763596536341617e-05,
"loss": 4.5522,
"step": 144384
},
{
"epoch": 0.02,
"learning_rate": 4.762759579470938e-05,
"loss": 4.5377,
"step": 144896
},
{
"epoch": 0.02,
"learning_rate": 4.761920984719886e-05,
"loss": 4.5234,
"step": 145408
},
{
"epoch": 0.02,
"learning_rate": 4.761082389968834e-05,
"loss": 4.5257,
"step": 145920
},
{
"epoch": 0.02,
"learning_rate": 4.760243795217783e-05,
"loss": 4.5236,
"step": 146432
},
{
"epoch": 0.02,
"learning_rate": 4.759405200466731e-05,
"loss": 4.5338,
"step": 146944
},
{
"epoch": 0.02,
"learning_rate": 4.758566605715679e-05,
"loss": 4.5354,
"step": 147456
},
{
"epoch": 0.02,
"learning_rate": 4.757728010964627e-05,
"loss": 4.5352,
"step": 147968
},
{
"epoch": 0.02,
"learning_rate": 4.756889416213575e-05,
"loss": 4.5297,
"step": 148480
},
{
"epoch": 0.02,
"learning_rate": 4.756052459342896e-05,
"loss": 4.5223,
"step": 148992
},
{
"epoch": 0.02,
"learning_rate": 4.755213864591844e-05,
"loss": 4.516,
"step": 149504
},
{
"epoch": 0.02,
"learning_rate": 4.754375269840792e-05,
"loss": 4.5138,
"step": 150016
},
{
"epoch": 0.02,
"learning_rate": 4.75353667508974e-05,
"loss": 4.5242,
"step": 150528
},
{
"epoch": 0.02,
"learning_rate": 4.752698080338688e-05,
"loss": 4.5374,
"step": 151040
},
{
"epoch": 0.02,
"learning_rate": 4.751859485587636e-05,
"loss": 4.5214,
"step": 151552
},
{
"epoch": 0.02,
"learning_rate": 4.7510225287169566e-05,
"loss": 4.5135,
"step": 152064
},
{
"epoch": 0.02,
"learning_rate": 4.7501839339659046e-05,
"loss": 4.5121,
"step": 152576
},
{
"epoch": 0.03,
"eval_loss": 4.480572700500488,
"eval_runtime": 532.969,
"eval_samples_per_second": 715.972,
"eval_steps_per_second": 22.375,
"step": 152638
}
],
"logging_steps": 512,
"max_steps": 3052726,
"num_train_epochs": 9223372036854775807,
"save_steps": 10,
"total_flos": 6.299511641873014e+16,
"trial_name": null,
"trial_params": null
}