npi-only-lstm-1 / checkpoint-381600 /trainer_state.json
abhinavp's picture
Training in progress, epoch 0, checkpoint
e0c8abd verified
raw
history blame
93.1 kB
{
"best_metric": 4.192408084869385,
"best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/npi-only/lstm/1/checkpoints/checkpoint-381600",
"epoch": 0.025000606015738065,
"eval_steps": 10,
"global_step": 381600,
"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": 10.8191,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 4.999161405248948e-05,
"loss": 7.5474,
"step": 512
},
{
"epoch": 0.0,
"learning_rate": 4.998322810497896e-05,
"loss": 7.0562,
"step": 1024
},
{
"epoch": 0.0,
"learning_rate": 4.997484215746844e-05,
"loss": 6.9904,
"step": 1536
},
{
"epoch": 0.0,
"learning_rate": 4.996645620995792e-05,
"loss": 6.9436,
"step": 2048
},
{
"epoch": 0.0,
"learning_rate": 4.99580702624474e-05,
"loss": 6.8394,
"step": 2560
},
{
"epoch": 0.0,
"learning_rate": 4.994968431493688e-05,
"loss": 6.682,
"step": 3072
},
{
"epoch": 0.0,
"learning_rate": 4.994129836742636e-05,
"loss": 6.5858,
"step": 3584
},
{
"epoch": 0.0,
"learning_rate": 4.993292879871958e-05,
"loss": 6.4898,
"step": 4096
},
{
"epoch": 0.0,
"learning_rate": 4.992454285120906e-05,
"loss": 6.4265,
"step": 4608
},
{
"epoch": 0.0,
"learning_rate": 4.991615690369854e-05,
"loss": 6.3737,
"step": 5120
},
{
"epoch": 0.0,
"learning_rate": 4.990777095618801e-05,
"loss": 6.3089,
"step": 5632
},
{
"epoch": 0.0,
"learning_rate": 4.989938500867749e-05,
"loss": 6.2352,
"step": 6144
},
{
"epoch": 0.0,
"learning_rate": 4.989099906116697e-05,
"loss": 6.1763,
"step": 6656
},
{
"epoch": 0.0,
"learning_rate": 4.988261311365645e-05,
"loss": 6.127,
"step": 7168
},
{
"epoch": 0.0,
"learning_rate": 4.987422716614593e-05,
"loss": 6.0632,
"step": 7680
},
{
"epoch": 0.0,
"learning_rate": 4.986584121863541e-05,
"loss": 6.0303,
"step": 8192
},
{
"epoch": 0.0,
"learning_rate": 4.985745527112489e-05,
"loss": 5.9826,
"step": 8704
},
{
"epoch": 0.0,
"learning_rate": 4.984906932361437e-05,
"loss": 5.9348,
"step": 9216
},
{
"epoch": 0.0,
"learning_rate": 4.984068337610385e-05,
"loss": 5.9037,
"step": 9728
},
{
"epoch": 0.0,
"learning_rate": 4.983229742859333e-05,
"loss": 5.8696,
"step": 10240
},
{
"epoch": 0.0,
"learning_rate": 4.982391148108281e-05,
"loss": 5.8321,
"step": 10752
},
{
"epoch": 0.0,
"learning_rate": 4.981552553357229e-05,
"loss": 5.8074,
"step": 11264
},
{
"epoch": 0.0,
"learning_rate": 4.980713958606178e-05,
"loss": 5.7667,
"step": 11776
},
{
"epoch": 0.0,
"learning_rate": 4.979875363855125e-05,
"loss": 5.749,
"step": 12288
},
{
"epoch": 0.0,
"learning_rate": 4.979036769104073e-05,
"loss": 5.7241,
"step": 12800
},
{
"epoch": 0.0,
"learning_rate": 4.978198174353021e-05,
"loss": 5.699,
"step": 13312
},
{
"epoch": 0.0,
"learning_rate": 4.9773612174823426e-05,
"loss": 5.6708,
"step": 13824
},
{
"epoch": 0.0,
"learning_rate": 4.97652262273129e-05,
"loss": 5.638,
"step": 14336
},
{
"epoch": 0.0,
"learning_rate": 4.975684027980238e-05,
"loss": 5.6277,
"step": 14848
},
{
"epoch": 0.01,
"learning_rate": 4.974845433229186e-05,
"loss": 5.6054,
"step": 15360
},
{
"epoch": 0.01,
"learning_rate": 4.974008476358507e-05,
"loss": 5.5934,
"step": 15872
},
{
"epoch": 0.01,
"learning_rate": 4.973169881607455e-05,
"loss": 5.5672,
"step": 16384
},
{
"epoch": 0.01,
"learning_rate": 4.972331286856403e-05,
"loss": 5.5493,
"step": 16896
},
{
"epoch": 0.01,
"learning_rate": 4.9714926921053515e-05,
"loss": 5.5388,
"step": 17408
},
{
"epoch": 0.01,
"learning_rate": 4.9706540973542995e-05,
"loss": 5.5212,
"step": 17920
},
{
"epoch": 0.01,
"learning_rate": 4.9698171404836204e-05,
"loss": 5.5041,
"step": 18432
},
{
"epoch": 0.01,
"learning_rate": 4.968980183612942e-05,
"loss": 5.4724,
"step": 18944
},
{
"epoch": 0.01,
"learning_rate": 4.96814158886189e-05,
"loss": 5.4676,
"step": 19456
},
{
"epoch": 0.01,
"learning_rate": 4.967302994110837e-05,
"loss": 5.4366,
"step": 19968
},
{
"epoch": 0.01,
"learning_rate": 4.966464399359785e-05,
"loss": 5.4384,
"step": 20480
},
{
"epoch": 0.01,
"learning_rate": 4.965625804608733e-05,
"loss": 5.4126,
"step": 20992
},
{
"epoch": 0.01,
"learning_rate": 4.964787209857681e-05,
"loss": 5.4148,
"step": 21504
},
{
"epoch": 0.01,
"learning_rate": 4.963948615106629e-05,
"loss": 5.3857,
"step": 22016
},
{
"epoch": 0.01,
"learning_rate": 4.963110020355577e-05,
"loss": 5.3782,
"step": 22528
},
{
"epoch": 0.01,
"learning_rate": 4.962273063484898e-05,
"loss": 5.3726,
"step": 23040
},
{
"epoch": 0.01,
"learning_rate": 4.961434468733847e-05,
"loss": 5.3643,
"step": 23552
},
{
"epoch": 0.01,
"learning_rate": 4.960595873982795e-05,
"loss": 5.3487,
"step": 24064
},
{
"epoch": 0.01,
"learning_rate": 4.959757279231743e-05,
"loss": 5.3248,
"step": 24576
},
{
"epoch": 0.01,
"learning_rate": 4.958918684480691e-05,
"loss": 5.3222,
"step": 25088
},
{
"epoch": 0.01,
"learning_rate": 4.958081727610012e-05,
"loss": 5.3262,
"step": 25600
},
{
"epoch": 0.01,
"learning_rate": 4.95724313285896e-05,
"loss": 5.3132,
"step": 26112
},
{
"epoch": 0.01,
"learning_rate": 4.956404538107908e-05,
"loss": 5.2916,
"step": 26624
},
{
"epoch": 0.01,
"learning_rate": 4.955565943356856e-05,
"loss": 5.2797,
"step": 27136
},
{
"epoch": 0.01,
"learning_rate": 4.954728986486177e-05,
"loss": 5.2774,
"step": 27648
},
{
"epoch": 0.01,
"learning_rate": 4.953890391735125e-05,
"loss": 5.265,
"step": 28160
},
{
"epoch": 0.01,
"learning_rate": 4.9530517969840727e-05,
"loss": 5.2739,
"step": 28672
},
{
"epoch": 0.01,
"learning_rate": 4.9522132022330207e-05,
"loss": 5.2307,
"step": 29184
},
{
"epoch": 0.01,
"learning_rate": 4.9513746074819686e-05,
"loss": 5.2445,
"step": 29696
},
{
"epoch": 0.01,
"learning_rate": 4.95053765061129e-05,
"loss": 5.2294,
"step": 30208
},
{
"epoch": 0.01,
"learning_rate": 4.949699055860238e-05,
"loss": 5.2149,
"step": 30720
},
{
"epoch": 0.01,
"learning_rate": 4.948860461109186e-05,
"loss": 5.2218,
"step": 31232
},
{
"epoch": 0.01,
"learning_rate": 4.948021866358134e-05,
"loss": 5.2001,
"step": 31744
},
{
"epoch": 0.01,
"learning_rate": 4.947183271607082e-05,
"loss": 5.19,
"step": 32256
},
{
"epoch": 0.01,
"learning_rate": 4.94634467685603e-05,
"loss": 5.1862,
"step": 32768
},
{
"epoch": 0.01,
"learning_rate": 4.945507719985351e-05,
"loss": 5.1966,
"step": 33280
},
{
"epoch": 0.01,
"learning_rate": 4.944669125234299e-05,
"loss": 5.173,
"step": 33792
},
{
"epoch": 0.01,
"learning_rate": 4.943830530483247e-05,
"loss": 5.1667,
"step": 34304
},
{
"epoch": 0.01,
"learning_rate": 4.942991935732195e-05,
"loss": 5.152,
"step": 34816
},
{
"epoch": 0.01,
"learning_rate": 4.942154978861516e-05,
"loss": 5.1446,
"step": 35328
},
{
"epoch": 0.01,
"learning_rate": 4.941316384110464e-05,
"loss": 5.1489,
"step": 35840
},
{
"epoch": 0.01,
"learning_rate": 4.940477789359412e-05,
"loss": 5.1518,
"step": 36352
},
{
"epoch": 0.01,
"learning_rate": 4.93963919460836e-05,
"loss": 5.1258,
"step": 36864
},
{
"epoch": 0.01,
"learning_rate": 4.938800599857309e-05,
"loss": 5.1368,
"step": 37376
},
{
"epoch": 0.01,
"learning_rate": 4.9379636429866296e-05,
"loss": 5.1374,
"step": 37888
},
{
"epoch": 0.01,
"learning_rate": 4.9371250482355776e-05,
"loss": 5.1157,
"step": 38400
},
{
"epoch": 0.01,
"learning_rate": 4.9362880913648985e-05,
"loss": 5.0994,
"step": 38912
},
{
"epoch": 0.01,
"learning_rate": 4.9354494966138465e-05,
"loss": 5.0998,
"step": 39424
},
{
"epoch": 0.01,
"learning_rate": 4.9346109018627945e-05,
"loss": 5.0942,
"step": 39936
},
{
"epoch": 0.01,
"learning_rate": 4.9337723071117425e-05,
"loss": 5.0812,
"step": 40448
},
{
"epoch": 0.01,
"learning_rate": 4.9329337123606905e-05,
"loss": 5.0807,
"step": 40960
},
{
"epoch": 0.01,
"learning_rate": 4.932095117609638e-05,
"loss": 5.0829,
"step": 41472
},
{
"epoch": 0.01,
"learning_rate": 4.931256522858586e-05,
"loss": 5.0821,
"step": 41984
},
{
"epoch": 0.01,
"learning_rate": 4.930417928107534e-05,
"loss": 5.0646,
"step": 42496
},
{
"epoch": 0.01,
"learning_rate": 4.9295809712368554e-05,
"loss": 5.0424,
"step": 43008
},
{
"epoch": 0.01,
"learning_rate": 4.9287423764858034e-05,
"loss": 5.049,
"step": 43520
},
{
"epoch": 0.01,
"learning_rate": 4.9279037817347514e-05,
"loss": 5.0537,
"step": 44032
},
{
"epoch": 0.01,
"learning_rate": 4.9270651869836994e-05,
"loss": 5.042,
"step": 44544
},
{
"epoch": 0.01,
"learning_rate": 4.9262265922326474e-05,
"loss": 5.0399,
"step": 45056
},
{
"epoch": 0.01,
"learning_rate": 4.925389635361968e-05,
"loss": 5.0213,
"step": 45568
},
{
"epoch": 0.02,
"learning_rate": 4.92455267849129e-05,
"loss": 5.0208,
"step": 46080
},
{
"epoch": 0.02,
"learning_rate": 4.923714083740238e-05,
"loss": 5.0139,
"step": 46592
},
{
"epoch": 0.02,
"learning_rate": 4.922875488989185e-05,
"loss": 5.0135,
"step": 47104
},
{
"epoch": 0.02,
"learning_rate": 4.922038532118507e-05,
"loss": 5.0129,
"step": 47616
},
{
"epoch": 0.02,
"learning_rate": 4.921199937367455e-05,
"loss": 4.9952,
"step": 48128
},
{
"epoch": 0.02,
"learning_rate": 4.920361342616403e-05,
"loss": 5.012,
"step": 48640
},
{
"epoch": 0.02,
"learning_rate": 4.919522747865351e-05,
"loss": 4.9972,
"step": 49152
},
{
"epoch": 0.02,
"learning_rate": 4.918684153114299e-05,
"loss": 4.9764,
"step": 49664
},
{
"epoch": 0.02,
"learning_rate": 4.917845558363247e-05,
"loss": 4.9733,
"step": 50176
},
{
"epoch": 0.02,
"learning_rate": 4.917006963612195e-05,
"loss": 4.9738,
"step": 50688
},
{
"epoch": 0.02,
"learning_rate": 4.916168368861143e-05,
"loss": 4.971,
"step": 51200
},
{
"epoch": 0.02,
"learning_rate": 4.915329774110091e-05,
"loss": 4.9701,
"step": 51712
},
{
"epoch": 0.02,
"learning_rate": 4.914491179359039e-05,
"loss": 4.9629,
"step": 52224
},
{
"epoch": 0.02,
"learning_rate": 4.913652584607987e-05,
"loss": 4.9552,
"step": 52736
},
{
"epoch": 0.02,
"learning_rate": 4.9128156277373076e-05,
"loss": 4.9505,
"step": 53248
},
{
"epoch": 0.02,
"learning_rate": 4.9119770329862556e-05,
"loss": 4.9288,
"step": 53760
},
{
"epoch": 0.02,
"learning_rate": 4.9111384382352036e-05,
"loss": 4.9351,
"step": 54272
},
{
"epoch": 0.02,
"learning_rate": 4.9102998434841516e-05,
"loss": 4.9326,
"step": 54784
},
{
"epoch": 0.02,
"learning_rate": 4.9094612487330996e-05,
"loss": 4.9279,
"step": 55296
},
{
"epoch": 0.02,
"learning_rate": 4.9086226539820476e-05,
"loss": 4.9228,
"step": 55808
},
{
"epoch": 0.02,
"learning_rate": 4.907784059230996e-05,
"loss": 4.9293,
"step": 56320
},
{
"epoch": 0.02,
"learning_rate": 4.906947102360317e-05,
"loss": 4.9039,
"step": 56832
},
{
"epoch": 0.02,
"learning_rate": 4.906108507609265e-05,
"loss": 4.9066,
"step": 57344
},
{
"epoch": 0.02,
"learning_rate": 4.905269912858213e-05,
"loss": 4.9007,
"step": 57856
},
{
"epoch": 0.02,
"learning_rate": 4.904431318107161e-05,
"loss": 4.9075,
"step": 58368
},
{
"epoch": 0.02,
"learning_rate": 4.903592723356109e-05,
"loss": 4.9122,
"step": 58880
},
{
"epoch": 0.02,
"learning_rate": 4.9027541286050565e-05,
"loss": 4.8949,
"step": 59392
},
{
"epoch": 0.02,
"learning_rate": 4.901917171734378e-05,
"loss": 4.8927,
"step": 59904
},
{
"epoch": 0.02,
"learning_rate": 4.901078576983326e-05,
"loss": 4.8831,
"step": 60416
},
{
"epoch": 0.02,
"learning_rate": 4.900239982232274e-05,
"loss": 4.8897,
"step": 60928
},
{
"epoch": 0.02,
"learning_rate": 4.8994013874812214e-05,
"loss": 4.8835,
"step": 61440
},
{
"epoch": 0.02,
"learning_rate": 4.89856279273017e-05,
"loss": 4.8797,
"step": 61952
},
{
"epoch": 0.02,
"learning_rate": 4.897724197979118e-05,
"loss": 4.8706,
"step": 62464
},
{
"epoch": 0.02,
"learning_rate": 4.896885603228066e-05,
"loss": 4.8661,
"step": 62976
},
{
"epoch": 0.02,
"learning_rate": 4.896047008477014e-05,
"loss": 4.8592,
"step": 63488
},
{
"epoch": 0.02,
"learning_rate": 4.895208413725962e-05,
"loss": 4.8493,
"step": 64000
},
{
"epoch": 0.02,
"learning_rate": 4.894371456855283e-05,
"loss": 4.8547,
"step": 64512
},
{
"epoch": 0.02,
"learning_rate": 4.893532862104231e-05,
"loss": 4.8485,
"step": 65024
},
{
"epoch": 0.02,
"learning_rate": 4.892694267353179e-05,
"loss": 4.8435,
"step": 65536
},
{
"epoch": 0.02,
"learning_rate": 4.891855672602127e-05,
"loss": 4.8358,
"step": 66048
},
{
"epoch": 0.02,
"learning_rate": 4.891017077851075e-05,
"loss": 4.8512,
"step": 66560
},
{
"epoch": 0.02,
"learning_rate": 4.890178483100023e-05,
"loss": 4.8501,
"step": 67072
},
{
"epoch": 0.02,
"learning_rate": 4.889341526229344e-05,
"loss": 4.8396,
"step": 67584
},
{
"epoch": 0.02,
"learning_rate": 4.888504569358665e-05,
"loss": 4.8285,
"step": 68096
},
{
"epoch": 0.02,
"learning_rate": 4.8876659746076134e-05,
"loss": 4.8275,
"step": 68608
},
{
"epoch": 0.02,
"learning_rate": 4.8868273798565614e-05,
"loss": 4.8306,
"step": 69120
},
{
"epoch": 0.02,
"learning_rate": 4.8859887851055094e-05,
"loss": 4.8264,
"step": 69632
},
{
"epoch": 0.02,
"learning_rate": 4.8851501903544574e-05,
"loss": 4.8222,
"step": 70144
},
{
"epoch": 0.02,
"learning_rate": 4.8843115956034054e-05,
"loss": 4.8184,
"step": 70656
},
{
"epoch": 0.02,
"learning_rate": 4.8834730008523534e-05,
"loss": 4.815,
"step": 71168
},
{
"epoch": 0.02,
"learning_rate": 4.8826344061013014e-05,
"loss": 4.8085,
"step": 71680
},
{
"epoch": 0.02,
"learning_rate": 4.8817958113502494e-05,
"loss": 4.8032,
"step": 72192
},
{
"epoch": 0.02,
"learning_rate": 4.8809572165991974e-05,
"loss": 4.8079,
"step": 72704
},
{
"epoch": 0.02,
"learning_rate": 4.8801186218481454e-05,
"loss": 4.7961,
"step": 73216
},
{
"epoch": 0.02,
"learning_rate": 4.879281664977466e-05,
"loss": 4.7909,
"step": 73728
},
{
"epoch": 0.02,
"learning_rate": 4.878443070226414e-05,
"loss": 4.7983,
"step": 74240
},
{
"epoch": 0.02,
"learning_rate": 4.877604475475362e-05,
"loss": 4.7846,
"step": 74752
},
{
"epoch": 0.02,
"learning_rate": 4.87676588072431e-05,
"loss": 4.793,
"step": 75264
},
{
"epoch": 0.02,
"learning_rate": 4.875927285973258e-05,
"loss": 4.782,
"step": 75776
},
{
"epoch": 0.02,
"learning_rate": 4.87509032910258e-05,
"loss": 4.7741,
"step": 76288
},
{
"epoch": 0.03,
"eval_loss": 4.741952896118164,
"eval_runtime": 347.3647,
"eval_samples_per_second": 1098.531,
"eval_steps_per_second": 34.33,
"step": 76320
},
{
"epoch": 1.0,
"learning_rate": 4.874251734351528e-05,
"loss": 4.7684,
"step": 76800
},
{
"epoch": 1.0,
"learning_rate": 4.873413139600475e-05,
"loss": 4.7679,
"step": 77312
},
{
"epoch": 1.0,
"learning_rate": 4.872576182729797e-05,
"loss": 4.7801,
"step": 77824
},
{
"epoch": 1.0,
"learning_rate": 4.871737587978745e-05,
"loss": 4.7668,
"step": 78336
},
{
"epoch": 1.0,
"learning_rate": 4.870898993227693e-05,
"loss": 4.7692,
"step": 78848
},
{
"epoch": 1.0,
"learning_rate": 4.87006039847664e-05,
"loss": 4.756,
"step": 79360
},
{
"epoch": 1.0,
"learning_rate": 4.869221803725588e-05,
"loss": 4.7591,
"step": 79872
},
{
"epoch": 1.0,
"learning_rate": 4.868383208974536e-05,
"loss": 4.743,
"step": 80384
},
{
"epoch": 1.0,
"learning_rate": 4.867544614223484e-05,
"loss": 4.7566,
"step": 80896
},
{
"epoch": 1.0,
"learning_rate": 4.866706019472432e-05,
"loss": 4.7545,
"step": 81408
},
{
"epoch": 1.0,
"learning_rate": 4.86586742472138e-05,
"loss": 4.7603,
"step": 81920
},
{
"epoch": 1.0,
"learning_rate": 4.8650304678507016e-05,
"loss": 4.7405,
"step": 82432
},
{
"epoch": 1.0,
"learning_rate": 4.8641918730996496e-05,
"loss": 4.7387,
"step": 82944
},
{
"epoch": 1.0,
"learning_rate": 4.8633532783485976e-05,
"loss": 4.7343,
"step": 83456
},
{
"epoch": 1.0,
"learning_rate": 4.8625146835975456e-05,
"loss": 4.726,
"step": 83968
},
{
"epoch": 1.0,
"learning_rate": 4.8616760888464936e-05,
"loss": 4.7295,
"step": 84480
},
{
"epoch": 1.0,
"learning_rate": 4.8608374940954416e-05,
"loss": 4.7363,
"step": 84992
},
{
"epoch": 1.0,
"learning_rate": 4.8600005372247625e-05,
"loss": 4.7194,
"step": 85504
},
{
"epoch": 1.0,
"learning_rate": 4.8591619424737105e-05,
"loss": 4.7201,
"step": 86016
},
{
"epoch": 1.0,
"learning_rate": 4.8583233477226585e-05,
"loss": 4.735,
"step": 86528
},
{
"epoch": 1.0,
"learning_rate": 4.8574847529716065e-05,
"loss": 4.7232,
"step": 87040
},
{
"epoch": 1.0,
"learning_rate": 4.8566477961009274e-05,
"loss": 4.7218,
"step": 87552
},
{
"epoch": 1.0,
"learning_rate": 4.8558092013498754e-05,
"loss": 4.7101,
"step": 88064
},
{
"epoch": 1.0,
"learning_rate": 4.854970606598824e-05,
"loss": 4.7232,
"step": 88576
},
{
"epoch": 1.0,
"learning_rate": 4.854132011847772e-05,
"loss": 4.7093,
"step": 89088
},
{
"epoch": 1.0,
"learning_rate": 4.85329341709672e-05,
"loss": 4.7072,
"step": 89600
},
{
"epoch": 1.0,
"learning_rate": 4.852456460226041e-05,
"loss": 4.703,
"step": 90112
},
{
"epoch": 1.0,
"learning_rate": 4.851617865474989e-05,
"loss": 4.6972,
"step": 90624
},
{
"epoch": 1.0,
"learning_rate": 4.850779270723937e-05,
"loss": 4.6902,
"step": 91136
},
{
"epoch": 1.01,
"learning_rate": 4.849940675972885e-05,
"loss": 4.692,
"step": 91648
},
{
"epoch": 1.01,
"learning_rate": 4.849102081221833e-05,
"loss": 4.7025,
"step": 92160
},
{
"epoch": 1.01,
"learning_rate": 4.848263486470781e-05,
"loss": 4.6971,
"step": 92672
},
{
"epoch": 1.01,
"learning_rate": 4.847426529600102e-05,
"loss": 4.6882,
"step": 93184
},
{
"epoch": 1.01,
"learning_rate": 4.84658793484905e-05,
"loss": 4.6979,
"step": 93696
},
{
"epoch": 1.01,
"learning_rate": 4.845749340097998e-05,
"loss": 4.6894,
"step": 94208
},
{
"epoch": 1.01,
"learning_rate": 4.844910745346946e-05,
"loss": 4.6875,
"step": 94720
},
{
"epoch": 1.01,
"learning_rate": 4.8440737884762674e-05,
"loss": 4.6746,
"step": 95232
},
{
"epoch": 1.01,
"learning_rate": 4.8432351937252154e-05,
"loss": 4.6753,
"step": 95744
},
{
"epoch": 1.01,
"learning_rate": 4.8423965989741634e-05,
"loss": 4.6641,
"step": 96256
},
{
"epoch": 1.01,
"learning_rate": 4.8415580042231114e-05,
"loss": 4.6747,
"step": 96768
},
{
"epoch": 1.01,
"learning_rate": 4.840721047352432e-05,
"loss": 4.6601,
"step": 97280
},
{
"epoch": 1.01,
"learning_rate": 4.83988245260138e-05,
"loss": 4.6734,
"step": 97792
},
{
"epoch": 1.01,
"learning_rate": 4.839043857850328e-05,
"loss": 4.6556,
"step": 98304
},
{
"epoch": 1.01,
"learning_rate": 4.838205263099276e-05,
"loss": 4.6638,
"step": 98816
},
{
"epoch": 1.01,
"learning_rate": 4.8373666683482236e-05,
"loss": 4.6621,
"step": 99328
},
{
"epoch": 1.01,
"learning_rate": 4.8365280735971716e-05,
"loss": 4.6611,
"step": 99840
},
{
"epoch": 1.01,
"learning_rate": 4.8356894788461196e-05,
"loss": 4.6551,
"step": 100352
},
{
"epoch": 1.01,
"learning_rate": 4.834852521975441e-05,
"loss": 4.6433,
"step": 100864
},
{
"epoch": 1.01,
"learning_rate": 4.834013927224389e-05,
"loss": 4.6412,
"step": 101376
},
{
"epoch": 1.01,
"learning_rate": 4.833175332473337e-05,
"loss": 4.659,
"step": 101888
},
{
"epoch": 1.01,
"learning_rate": 4.832336737722285e-05,
"loss": 4.6489,
"step": 102400
},
{
"epoch": 1.01,
"learning_rate": 4.831498142971233e-05,
"loss": 4.6443,
"step": 102912
},
{
"epoch": 1.01,
"learning_rate": 4.830659548220181e-05,
"loss": 4.6313,
"step": 103424
},
{
"epoch": 1.01,
"learning_rate": 4.829820953469129e-05,
"loss": 4.6415,
"step": 103936
},
{
"epoch": 1.01,
"learning_rate": 4.82898399659845e-05,
"loss": 4.6282,
"step": 104448
},
{
"epoch": 1.01,
"learning_rate": 4.828145401847398e-05,
"loss": 4.6545,
"step": 104960
},
{
"epoch": 1.01,
"learning_rate": 4.827306807096346e-05,
"loss": 4.6175,
"step": 105472
},
{
"epoch": 1.01,
"learning_rate": 4.826468212345294e-05,
"loss": 4.6399,
"step": 105984
},
{
"epoch": 1.01,
"learning_rate": 4.825629617594242e-05,
"loss": 4.6293,
"step": 106496
},
{
"epoch": 1.01,
"learning_rate": 4.82479102284319e-05,
"loss": 4.6187,
"step": 107008
},
{
"epoch": 1.01,
"learning_rate": 4.823952428092138e-05,
"loss": 4.6296,
"step": 107520
},
{
"epoch": 1.01,
"learning_rate": 4.823113833341086e-05,
"loss": 4.6248,
"step": 108032
},
{
"epoch": 1.01,
"learning_rate": 4.8222768764704077e-05,
"loss": 4.6091,
"step": 108544
},
{
"epoch": 1.01,
"learning_rate": 4.8214382817193557e-05,
"loss": 4.6171,
"step": 109056
},
{
"epoch": 1.01,
"learning_rate": 4.8205996869683037e-05,
"loss": 4.6293,
"step": 109568
},
{
"epoch": 1.01,
"learning_rate": 4.8197610922172516e-05,
"loss": 4.6132,
"step": 110080
},
{
"epoch": 1.01,
"learning_rate": 4.8189241353465726e-05,
"loss": 4.6107,
"step": 110592
},
{
"epoch": 1.01,
"learning_rate": 4.8180855405955206e-05,
"loss": 4.6082,
"step": 111104
},
{
"epoch": 1.01,
"learning_rate": 4.8172502216052144e-05,
"loss": 4.6029,
"step": 111616
},
{
"epoch": 1.01,
"learning_rate": 4.8164116268541624e-05,
"loss": 4.6119,
"step": 112128
},
{
"epoch": 1.01,
"learning_rate": 4.8155730321031104e-05,
"loss": 4.6251,
"step": 112640
},
{
"epoch": 1.01,
"learning_rate": 4.8147344373520584e-05,
"loss": 4.5999,
"step": 113152
},
{
"epoch": 1.01,
"learning_rate": 4.8138958426010064e-05,
"loss": 4.6179,
"step": 113664
},
{
"epoch": 1.01,
"learning_rate": 4.813057247849955e-05,
"loss": 4.6179,
"step": 114176
},
{
"epoch": 1.01,
"learning_rate": 4.812218653098903e-05,
"loss": 4.6057,
"step": 114688
},
{
"epoch": 1.01,
"learning_rate": 4.811380058347851e-05,
"loss": 4.5928,
"step": 115200
},
{
"epoch": 1.01,
"learning_rate": 4.810541463596799e-05,
"loss": 4.6043,
"step": 115712
},
{
"epoch": 1.01,
"learning_rate": 4.809702868845747e-05,
"loss": 4.5988,
"step": 116224
},
{
"epoch": 1.01,
"learning_rate": 4.808864274094695e-05,
"loss": 4.5911,
"step": 116736
},
{
"epoch": 1.01,
"learning_rate": 4.808025679343642e-05,
"loss": 4.5924,
"step": 117248
},
{
"epoch": 1.01,
"learning_rate": 4.807188722472964e-05,
"loss": 4.5963,
"step": 117760
},
{
"epoch": 1.01,
"learning_rate": 4.806350127721912e-05,
"loss": 4.6038,
"step": 118272
},
{
"epoch": 1.01,
"learning_rate": 4.80551153297086e-05,
"loss": 4.5919,
"step": 118784
},
{
"epoch": 1.01,
"learning_rate": 4.804672938219807e-05,
"loss": 4.5764,
"step": 119296
},
{
"epoch": 1.01,
"learning_rate": 4.803835981349129e-05,
"loss": 4.5863,
"step": 119808
},
{
"epoch": 1.01,
"learning_rate": 4.802997386598077e-05,
"loss": 4.5889,
"step": 120320
},
{
"epoch": 1.01,
"learning_rate": 4.802158791847025e-05,
"loss": 4.5849,
"step": 120832
},
{
"epoch": 1.01,
"learning_rate": 4.801320197095973e-05,
"loss": 4.5908,
"step": 121344
},
{
"epoch": 1.01,
"learning_rate": 4.8004832402252944e-05,
"loss": 4.5767,
"step": 121856
},
{
"epoch": 1.02,
"learning_rate": 4.7996446454742424e-05,
"loss": 4.5708,
"step": 122368
},
{
"epoch": 1.02,
"learning_rate": 4.79880605072319e-05,
"loss": 4.5729,
"step": 122880
},
{
"epoch": 1.02,
"learning_rate": 4.797967455972138e-05,
"loss": 4.5769,
"step": 123392
},
{
"epoch": 1.02,
"learning_rate": 4.797130499101459e-05,
"loss": 4.5755,
"step": 123904
},
{
"epoch": 1.02,
"learning_rate": 4.796291904350407e-05,
"loss": 4.5669,
"step": 124416
},
{
"epoch": 1.02,
"learning_rate": 4.7954533095993546e-05,
"loss": 4.5889,
"step": 124928
},
{
"epoch": 1.02,
"learning_rate": 4.7946147148483026e-05,
"loss": 4.5758,
"step": 125440
},
{
"epoch": 1.02,
"learning_rate": 4.7937761200972506e-05,
"loss": 4.5584,
"step": 125952
},
{
"epoch": 1.02,
"learning_rate": 4.7929375253461986e-05,
"loss": 4.5614,
"step": 126464
},
{
"epoch": 1.02,
"learning_rate": 4.792098930595147e-05,
"loss": 4.5651,
"step": 126976
},
{
"epoch": 1.02,
"learning_rate": 4.791261973724468e-05,
"loss": 4.5648,
"step": 127488
},
{
"epoch": 1.02,
"learning_rate": 4.790423378973416e-05,
"loss": 4.5645,
"step": 128000
},
{
"epoch": 1.02,
"learning_rate": 4.789584784222364e-05,
"loss": 4.5614,
"step": 128512
},
{
"epoch": 1.02,
"learning_rate": 4.788746189471312e-05,
"loss": 4.556,
"step": 129024
},
{
"epoch": 1.02,
"learning_rate": 4.78790759472026e-05,
"loss": 4.56,
"step": 129536
},
{
"epoch": 1.02,
"learning_rate": 4.787068999969208e-05,
"loss": 4.5356,
"step": 130048
},
{
"epoch": 1.02,
"learning_rate": 4.786230405218156e-05,
"loss": 4.5516,
"step": 130560
},
{
"epoch": 1.02,
"learning_rate": 4.785391810467104e-05,
"loss": 4.5463,
"step": 131072
},
{
"epoch": 1.02,
"learning_rate": 4.784554853596425e-05,
"loss": 4.5496,
"step": 131584
},
{
"epoch": 1.02,
"learning_rate": 4.783717896725746e-05,
"loss": 4.5434,
"step": 132096
},
{
"epoch": 1.02,
"learning_rate": 4.782879301974694e-05,
"loss": 4.5531,
"step": 132608
},
{
"epoch": 1.02,
"learning_rate": 4.782040707223642e-05,
"loss": 4.5343,
"step": 133120
},
{
"epoch": 1.02,
"learning_rate": 4.7812021124725906e-05,
"loss": 4.5372,
"step": 133632
},
{
"epoch": 1.02,
"learning_rate": 4.7803635177215386e-05,
"loss": 4.5391,
"step": 134144
},
{
"epoch": 1.02,
"learning_rate": 4.7795249229704866e-05,
"loss": 4.5392,
"step": 134656
},
{
"epoch": 1.02,
"learning_rate": 4.7786863282194346e-05,
"loss": 4.553,
"step": 135168
},
{
"epoch": 1.02,
"learning_rate": 4.7778477334683826e-05,
"loss": 4.542,
"step": 135680
},
{
"epoch": 1.02,
"learning_rate": 4.7770091387173306e-05,
"loss": 4.5389,
"step": 136192
},
{
"epoch": 1.02,
"learning_rate": 4.7761705439662786e-05,
"loss": 4.5328,
"step": 136704
},
{
"epoch": 1.02,
"learning_rate": 4.7753352249759724e-05,
"loss": 4.5381,
"step": 137216
},
{
"epoch": 1.02,
"learning_rate": 4.7744966302249204e-05,
"loss": 4.5379,
"step": 137728
},
{
"epoch": 1.02,
"learning_rate": 4.7736580354738684e-05,
"loss": 4.5299,
"step": 138240
},
{
"epoch": 1.02,
"learning_rate": 4.7728194407228164e-05,
"loss": 4.5315,
"step": 138752
},
{
"epoch": 1.02,
"learning_rate": 4.7719808459717644e-05,
"loss": 4.526,
"step": 139264
},
{
"epoch": 1.02,
"learning_rate": 4.7711422512207124e-05,
"loss": 4.5235,
"step": 139776
},
{
"epoch": 1.02,
"learning_rate": 4.7703036564696604e-05,
"loss": 4.5155,
"step": 140288
},
{
"epoch": 1.02,
"learning_rate": 4.7694650617186084e-05,
"loss": 4.5208,
"step": 140800
},
{
"epoch": 1.02,
"learning_rate": 4.76862810484793e-05,
"loss": 4.5158,
"step": 141312
},
{
"epoch": 1.02,
"learning_rate": 4.767791147977251e-05,
"loss": 4.5227,
"step": 141824
},
{
"epoch": 1.02,
"learning_rate": 4.766952553226199e-05,
"loss": 4.5068,
"step": 142336
},
{
"epoch": 1.02,
"learning_rate": 4.766113958475147e-05,
"loss": 4.5297,
"step": 142848
},
{
"epoch": 1.02,
"learning_rate": 4.765275363724095e-05,
"loss": 4.5347,
"step": 143360
},
{
"epoch": 1.02,
"learning_rate": 4.764438406853416e-05,
"loss": 4.5192,
"step": 143872
},
{
"epoch": 1.02,
"learning_rate": 4.763599812102364e-05,
"loss": 4.5134,
"step": 144384
},
{
"epoch": 1.02,
"learning_rate": 4.762761217351312e-05,
"loss": 4.5143,
"step": 144896
},
{
"epoch": 1.02,
"learning_rate": 4.76192262260026e-05,
"loss": 4.5187,
"step": 145408
},
{
"epoch": 1.02,
"learning_rate": 4.761084027849208e-05,
"loss": 4.5203,
"step": 145920
},
{
"epoch": 1.02,
"learning_rate": 4.760245433098156e-05,
"loss": 4.5122,
"step": 146432
},
{
"epoch": 1.02,
"learning_rate": 4.759406838347104e-05,
"loss": 4.5129,
"step": 146944
},
{
"epoch": 1.02,
"learning_rate": 4.758568243596052e-05,
"loss": 4.5091,
"step": 147456
},
{
"epoch": 1.02,
"learning_rate": 4.757729648845e-05,
"loss": 4.5123,
"step": 147968
},
{
"epoch": 1.02,
"learning_rate": 4.756891054093948e-05,
"loss": 4.5042,
"step": 148480
},
{
"epoch": 1.02,
"learning_rate": 4.756052459342896e-05,
"loss": 4.5134,
"step": 148992
},
{
"epoch": 1.02,
"learning_rate": 4.755213864591844e-05,
"loss": 4.5023,
"step": 149504
},
{
"epoch": 1.02,
"learning_rate": 4.7543785456015375e-05,
"loss": 4.4979,
"step": 150016
},
{
"epoch": 1.02,
"learning_rate": 4.7535399508504855e-05,
"loss": 4.5086,
"step": 150528
},
{
"epoch": 1.02,
"learning_rate": 4.7527013560994335e-05,
"loss": 4.4967,
"step": 151040
},
{
"epoch": 1.02,
"learning_rate": 4.7518627613483815e-05,
"loss": 4.5055,
"step": 151552
},
{
"epoch": 1.02,
"learning_rate": 4.7510241665973295e-05,
"loss": 4.4936,
"step": 152064
},
{
"epoch": 1.02,
"learning_rate": 4.750185571846278e-05,
"loss": 4.4915,
"step": 152576
},
{
"epoch": 1.03,
"eval_loss": 4.463145732879639,
"eval_runtime": 354.3654,
"eval_samples_per_second": 1076.829,
"eval_steps_per_second": 33.652,
"step": 152640
},
{
"epoch": 0.0,
"learning_rate": 4.749346977095226e-05,
"loss": 4.49,
"step": 153088
},
{
"epoch": 0.0,
"learning_rate": 4.748508382344174e-05,
"loss": 4.4806,
"step": 153600
},
{
"epoch": 0.0,
"learning_rate": 4.747669787593122e-05,
"loss": 4.5043,
"step": 154112
},
{
"epoch": 0.0,
"learning_rate": 4.74683119284207e-05,
"loss": 4.4889,
"step": 154624
},
{
"epoch": 0.0,
"learning_rate": 4.745992598091018e-05,
"loss": 4.4979,
"step": 155136
},
{
"epoch": 0.0,
"learning_rate": 4.745154003339966e-05,
"loss": 4.4814,
"step": 155648
},
{
"epoch": 0.0,
"learning_rate": 4.744315408588914e-05,
"loss": 4.4901,
"step": 156160
},
{
"epoch": 0.0,
"learning_rate": 4.743476813837862e-05,
"loss": 4.4715,
"step": 156672
},
{
"epoch": 0.0,
"learning_rate": 4.7426382190868095e-05,
"loss": 4.4877,
"step": 157184
},
{
"epoch": 0.0,
"learning_rate": 4.7417996243357575e-05,
"loss": 4.4914,
"step": 157696
},
{
"epoch": 0.0,
"learning_rate": 4.7409610295847055e-05,
"loss": 4.4922,
"step": 158208
},
{
"epoch": 0.0,
"learning_rate": 4.7401224348336535e-05,
"loss": 4.4831,
"step": 158720
},
{
"epoch": 0.0,
"learning_rate": 4.7392838400826015e-05,
"loss": 4.4717,
"step": 159232
},
{
"epoch": 0.0,
"learning_rate": 4.7384452453315495e-05,
"loss": 4.4745,
"step": 159744
},
{
"epoch": 0.0,
"learning_rate": 4.7376066505804975e-05,
"loss": 4.4694,
"step": 160256
},
{
"epoch": 0.0,
"learning_rate": 4.736768055829446e-05,
"loss": 4.4726,
"step": 160768
},
{
"epoch": 0.0,
"learning_rate": 4.735929461078394e-05,
"loss": 4.4791,
"step": 161280
},
{
"epoch": 0.0,
"learning_rate": 4.735092504207715e-05,
"loss": 4.4653,
"step": 161792
},
{
"epoch": 0.0,
"learning_rate": 4.734253909456663e-05,
"loss": 4.4677,
"step": 162304
},
{
"epoch": 0.0,
"learning_rate": 4.733415314705611e-05,
"loss": 4.4889,
"step": 162816
},
{
"epoch": 0.0,
"learning_rate": 4.732578357834932e-05,
"loss": 4.4729,
"step": 163328
},
{
"epoch": 0.0,
"learning_rate": 4.73173976308388e-05,
"loss": 4.4754,
"step": 163840
},
{
"epoch": 0.0,
"learning_rate": 4.730901168332828e-05,
"loss": 4.4647,
"step": 164352
},
{
"epoch": 0.0,
"learning_rate": 4.730062573581776e-05,
"loss": 4.4819,
"step": 164864
},
{
"epoch": 0.0,
"learning_rate": 4.729223978830724e-05,
"loss": 4.4592,
"step": 165376
},
{
"epoch": 0.0,
"learning_rate": 4.728385384079672e-05,
"loss": 4.4666,
"step": 165888
},
{
"epoch": 0.0,
"learning_rate": 4.72754678932862e-05,
"loss": 4.4612,
"step": 166400
},
{
"epoch": 0.0,
"learning_rate": 4.726708194577568e-05,
"loss": 4.462,
"step": 166912
},
{
"epoch": 0.0,
"learning_rate": 4.725869599826516e-05,
"loss": 4.4481,
"step": 167424
},
{
"epoch": 0.01,
"learning_rate": 4.725031005075464e-05,
"loss": 4.4558,
"step": 167936
},
{
"epoch": 0.01,
"learning_rate": 4.724192410324412e-05,
"loss": 4.4639,
"step": 168448
},
{
"epoch": 0.01,
"learning_rate": 4.7233554534537335e-05,
"loss": 4.4641,
"step": 168960
},
{
"epoch": 0.01,
"learning_rate": 4.722516858702681e-05,
"loss": 4.4589,
"step": 169472
},
{
"epoch": 0.01,
"learning_rate": 4.721678263951629e-05,
"loss": 4.4603,
"step": 169984
},
{
"epoch": 0.01,
"learning_rate": 4.720839669200577e-05,
"loss": 4.4573,
"step": 170496
},
{
"epoch": 0.01,
"learning_rate": 4.7200027123298984e-05,
"loss": 4.4596,
"step": 171008
},
{
"epoch": 0.01,
"learning_rate": 4.719167393339592e-05,
"loss": 4.4456,
"step": 171520
},
{
"epoch": 0.01,
"learning_rate": 4.71832879858854e-05,
"loss": 4.4484,
"step": 172032
},
{
"epoch": 0.01,
"learning_rate": 4.717490203837488e-05,
"loss": 4.4392,
"step": 172544
},
{
"epoch": 0.01,
"learning_rate": 4.716651609086437e-05,
"loss": 4.4543,
"step": 173056
},
{
"epoch": 0.01,
"learning_rate": 4.715813014335385e-05,
"loss": 4.4382,
"step": 173568
},
{
"epoch": 0.01,
"learning_rate": 4.714974419584333e-05,
"loss": 4.4487,
"step": 174080
},
{
"epoch": 0.01,
"learning_rate": 4.714135824833281e-05,
"loss": 4.4401,
"step": 174592
},
{
"epoch": 0.01,
"learning_rate": 4.713297230082228e-05,
"loss": 4.4409,
"step": 175104
},
{
"epoch": 0.01,
"learning_rate": 4.712458635331176e-05,
"loss": 4.449,
"step": 175616
},
{
"epoch": 0.01,
"learning_rate": 4.711621678460498e-05,
"loss": 4.44,
"step": 176128
},
{
"epoch": 0.01,
"learning_rate": 4.710783083709446e-05,
"loss": 4.4411,
"step": 176640
},
{
"epoch": 0.01,
"learning_rate": 4.709944488958393e-05,
"loss": 4.4249,
"step": 177152
},
{
"epoch": 0.01,
"learning_rate": 4.709105894207341e-05,
"loss": 4.4301,
"step": 177664
},
{
"epoch": 0.01,
"learning_rate": 4.7082689373366627e-05,
"loss": 4.4441,
"step": 178176
},
{
"epoch": 0.01,
"learning_rate": 4.7074303425856106e-05,
"loss": 4.4357,
"step": 178688
},
{
"epoch": 0.01,
"learning_rate": 4.7065917478345586e-05,
"loss": 4.436,
"step": 179200
},
{
"epoch": 0.01,
"learning_rate": 4.7057531530835066e-05,
"loss": 4.4212,
"step": 179712
},
{
"epoch": 0.01,
"learning_rate": 4.7049145583324546e-05,
"loss": 4.4338,
"step": 180224
},
{
"epoch": 0.01,
"learning_rate": 4.7040776014617755e-05,
"loss": 4.4206,
"step": 180736
},
{
"epoch": 0.01,
"learning_rate": 4.7032390067107235e-05,
"loss": 4.4457,
"step": 181248
},
{
"epoch": 0.01,
"learning_rate": 4.7024004119596715e-05,
"loss": 4.4139,
"step": 181760
},
{
"epoch": 0.01,
"learning_rate": 4.7015618172086195e-05,
"loss": 4.4341,
"step": 182272
},
{
"epoch": 0.01,
"learning_rate": 4.7007232224575675e-05,
"loss": 4.4279,
"step": 182784
},
{
"epoch": 0.01,
"learning_rate": 4.6998846277065155e-05,
"loss": 4.4125,
"step": 183296
},
{
"epoch": 0.01,
"learning_rate": 4.6990460329554635e-05,
"loss": 4.4232,
"step": 183808
},
{
"epoch": 0.01,
"learning_rate": 4.6982074382044115e-05,
"loss": 4.4271,
"step": 184320
},
{
"epoch": 0.01,
"learning_rate": 4.6973704813337324e-05,
"loss": 4.4103,
"step": 184832
},
{
"epoch": 0.01,
"learning_rate": 4.6965318865826804e-05,
"loss": 4.4185,
"step": 185344
},
{
"epoch": 0.01,
"learning_rate": 4.695693291831629e-05,
"loss": 4.4273,
"step": 185856
},
{
"epoch": 0.01,
"learning_rate": 4.69485633496095e-05,
"loss": 4.4188,
"step": 186368
},
{
"epoch": 0.01,
"learning_rate": 4.694017740209898e-05,
"loss": 4.4127,
"step": 186880
},
{
"epoch": 0.01,
"learning_rate": 4.693179145458846e-05,
"loss": 4.4145,
"step": 187392
},
{
"epoch": 0.01,
"learning_rate": 4.692340550707794e-05,
"loss": 4.4069,
"step": 187904
},
{
"epoch": 0.01,
"learning_rate": 4.691503593837115e-05,
"loss": 4.4192,
"step": 188416
},
{
"epoch": 0.01,
"learning_rate": 4.690664999086063e-05,
"loss": 4.4339,
"step": 188928
},
{
"epoch": 0.01,
"learning_rate": 4.689826404335011e-05,
"loss": 4.4092,
"step": 189440
},
{
"epoch": 0.01,
"learning_rate": 4.688987809583959e-05,
"loss": 4.4267,
"step": 189952
},
{
"epoch": 0.01,
"learning_rate": 4.688149214832907e-05,
"loss": 4.4256,
"step": 190464
},
{
"epoch": 0.01,
"learning_rate": 4.687310620081855e-05,
"loss": 4.4177,
"step": 190976
},
{
"epoch": 0.01,
"learning_rate": 4.686472025330803e-05,
"loss": 4.4024,
"step": 191488
},
{
"epoch": 0.01,
"learning_rate": 4.6856350684601245e-05,
"loss": 4.421,
"step": 192000
},
{
"epoch": 0.01,
"learning_rate": 4.6847981115894454e-05,
"loss": 4.411,
"step": 192512
},
{
"epoch": 0.01,
"learning_rate": 4.6839595168383934e-05,
"loss": 4.4095,
"step": 193024
},
{
"epoch": 0.01,
"learning_rate": 4.6831209220873414e-05,
"loss": 4.4062,
"step": 193536
},
{
"epoch": 0.01,
"learning_rate": 4.6822823273362894e-05,
"loss": 4.4113,
"step": 194048
},
{
"epoch": 0.01,
"learning_rate": 4.6814437325852373e-05,
"loss": 4.4249,
"step": 194560
},
{
"epoch": 0.01,
"learning_rate": 4.6806051378341853e-05,
"loss": 4.4079,
"step": 195072
},
{
"epoch": 0.01,
"learning_rate": 4.6797665430831333e-05,
"loss": 4.3941,
"step": 195584
},
{
"epoch": 0.01,
"learning_rate": 4.678929586212454e-05,
"loss": 4.4025,
"step": 196096
},
{
"epoch": 0.01,
"learning_rate": 4.678090991461402e-05,
"loss": 4.4084,
"step": 196608
},
{
"epoch": 0.01,
"learning_rate": 4.67725239671035e-05,
"loss": 4.4079,
"step": 197120
},
{
"epoch": 0.01,
"learning_rate": 4.676413801959298e-05,
"loss": 4.4135,
"step": 197632
},
{
"epoch": 0.01,
"learning_rate": 4.675575207208246e-05,
"loss": 4.3973,
"step": 198144
},
{
"epoch": 0.02,
"learning_rate": 4.674736612457194e-05,
"loss": 4.396,
"step": 198656
},
{
"epoch": 0.02,
"learning_rate": 4.673898017706142e-05,
"loss": 4.3902,
"step": 199168
},
{
"epoch": 0.02,
"learning_rate": 4.67305942295509e-05,
"loss": 4.3978,
"step": 199680
},
{
"epoch": 0.02,
"learning_rate": 4.672220828204038e-05,
"loss": 4.4004,
"step": 200192
},
{
"epoch": 0.02,
"learning_rate": 4.671382233452986e-05,
"loss": 4.399,
"step": 200704
},
{
"epoch": 0.02,
"learning_rate": 4.670545276582307e-05,
"loss": 4.4168,
"step": 201216
},
{
"epoch": 0.02,
"learning_rate": 4.669706681831255e-05,
"loss": 4.3953,
"step": 201728
},
{
"epoch": 0.02,
"learning_rate": 4.668868087080203e-05,
"loss": 4.3918,
"step": 202240
},
{
"epoch": 0.02,
"learning_rate": 4.668029492329151e-05,
"loss": 4.3888,
"step": 202752
},
{
"epoch": 0.02,
"learning_rate": 4.667190897578099e-05,
"loss": 4.3919,
"step": 203264
},
{
"epoch": 0.02,
"learning_rate": 4.66635394070742e-05,
"loss": 4.3973,
"step": 203776
},
{
"epoch": 0.02,
"learning_rate": 4.665515345956368e-05,
"loss": 4.3959,
"step": 204288
},
{
"epoch": 0.02,
"learning_rate": 4.664676751205316e-05,
"loss": 4.3908,
"step": 204800
},
{
"epoch": 0.02,
"learning_rate": 4.663838156454265e-05,
"loss": 4.3875,
"step": 205312
},
{
"epoch": 0.02,
"learning_rate": 4.662999561703213e-05,
"loss": 4.3946,
"step": 205824
},
{
"epoch": 0.02,
"learning_rate": 4.662160966952161e-05,
"loss": 4.3699,
"step": 206336
},
{
"epoch": 0.02,
"learning_rate": 4.661322372201109e-05,
"loss": 4.3875,
"step": 206848
},
{
"epoch": 0.02,
"learning_rate": 4.6604837774500567e-05,
"loss": 4.3795,
"step": 207360
},
{
"epoch": 0.02,
"learning_rate": 4.6596484584597505e-05,
"loss": 4.3851,
"step": 207872
},
{
"epoch": 0.02,
"learning_rate": 4.6588098637086985e-05,
"loss": 4.3804,
"step": 208384
},
{
"epoch": 0.02,
"learning_rate": 4.6579712689576465e-05,
"loss": 4.3902,
"step": 208896
},
{
"epoch": 0.02,
"learning_rate": 4.6571326742065945e-05,
"loss": 4.3719,
"step": 209408
},
{
"epoch": 0.02,
"learning_rate": 4.6562940794555425e-05,
"loss": 4.3729,
"step": 209920
},
{
"epoch": 0.02,
"learning_rate": 4.6554571225848634e-05,
"loss": 4.3828,
"step": 210432
},
{
"epoch": 0.02,
"learning_rate": 4.6546185278338114e-05,
"loss": 4.3745,
"step": 210944
},
{
"epoch": 0.02,
"learning_rate": 4.65377993308276e-05,
"loss": 4.3944,
"step": 211456
},
{
"epoch": 0.02,
"learning_rate": 4.652941338331708e-05,
"loss": 4.3859,
"step": 211968
},
{
"epoch": 0.02,
"learning_rate": 4.652104381461029e-05,
"loss": 4.3842,
"step": 212480
},
{
"epoch": 0.02,
"learning_rate": 4.651265786709977e-05,
"loss": 4.3719,
"step": 212992
},
{
"epoch": 0.02,
"learning_rate": 4.650427191958925e-05,
"loss": 4.3777,
"step": 213504
},
{
"epoch": 0.02,
"learning_rate": 4.649588597207873e-05,
"loss": 4.3802,
"step": 214016
},
{
"epoch": 0.02,
"learning_rate": 4.648750002456821e-05,
"loss": 4.3765,
"step": 214528
},
{
"epoch": 0.02,
"learning_rate": 4.647911407705769e-05,
"loss": 4.3774,
"step": 215040
},
{
"epoch": 0.02,
"learning_rate": 4.647072812954717e-05,
"loss": 4.3667,
"step": 215552
},
{
"epoch": 0.02,
"learning_rate": 4.646234218203665e-05,
"loss": 4.3733,
"step": 216064
},
{
"epoch": 0.02,
"learning_rate": 4.645395623452612e-05,
"loss": 4.363,
"step": 216576
},
{
"epoch": 0.02,
"learning_rate": 4.64455702870156e-05,
"loss": 4.3685,
"step": 217088
},
{
"epoch": 0.02,
"learning_rate": 4.643720071830882e-05,
"loss": 4.365,
"step": 217600
},
{
"epoch": 0.02,
"learning_rate": 4.6428831149602034e-05,
"loss": 4.3689,
"step": 218112
},
{
"epoch": 0.02,
"learning_rate": 4.6420445202091514e-05,
"loss": 4.355,
"step": 218624
},
{
"epoch": 0.02,
"learning_rate": 4.6412059254580994e-05,
"loss": 4.3806,
"step": 219136
},
{
"epoch": 0.02,
"learning_rate": 4.6403673307070474e-05,
"loss": 4.3818,
"step": 219648
},
{
"epoch": 0.02,
"learning_rate": 4.639528735955995e-05,
"loss": 4.3732,
"step": 220160
},
{
"epoch": 0.02,
"learning_rate": 4.638691779085316e-05,
"loss": 4.3591,
"step": 220672
},
{
"epoch": 0.02,
"learning_rate": 4.637853184334264e-05,
"loss": 4.3679,
"step": 221184
},
{
"epoch": 0.02,
"learning_rate": 4.637014589583212e-05,
"loss": 4.3683,
"step": 221696
},
{
"epoch": 0.02,
"learning_rate": 4.6361759948321596e-05,
"loss": 4.3747,
"step": 222208
},
{
"epoch": 0.02,
"learning_rate": 4.6353374000811076e-05,
"loss": 4.3658,
"step": 222720
},
{
"epoch": 0.02,
"learning_rate": 4.6344988053300556e-05,
"loss": 4.3644,
"step": 223232
},
{
"epoch": 0.02,
"learning_rate": 4.6336602105790036e-05,
"loss": 4.3636,
"step": 223744
},
{
"epoch": 0.02,
"learning_rate": 4.632821615827952e-05,
"loss": 4.3695,
"step": 224256
},
{
"epoch": 0.02,
"learning_rate": 4.6319830210769e-05,
"loss": 4.3613,
"step": 224768
},
{
"epoch": 0.02,
"learning_rate": 4.631144426325848e-05,
"loss": 4.3671,
"step": 225280
},
{
"epoch": 0.02,
"learning_rate": 4.630305831574796e-05,
"loss": 4.3593,
"step": 225792
},
{
"epoch": 0.02,
"learning_rate": 4.629467236823744e-05,
"loss": 4.3565,
"step": 226304
},
{
"epoch": 0.02,
"learning_rate": 4.628630279953065e-05,
"loss": 4.3658,
"step": 226816
},
{
"epoch": 0.02,
"learning_rate": 4.627793323082386e-05,
"loss": 4.3593,
"step": 227328
},
{
"epoch": 0.02,
"learning_rate": 4.626954728331334e-05,
"loss": 4.3577,
"step": 227840
},
{
"epoch": 0.02,
"learning_rate": 4.626116133580282e-05,
"loss": 4.3518,
"step": 228352
},
{
"epoch": 0.02,
"learning_rate": 4.625279176709603e-05,
"loss": 4.3483,
"step": 228864
},
{
"epoch": 0.03,
"eval_loss": 4.330287933349609,
"eval_runtime": 302.6809,
"eval_samples_per_second": 1260.704,
"eval_steps_per_second": 39.398,
"step": 228960
},
{
"epoch": 1.0,
"learning_rate": 4.624440581958551e-05,
"loss": 4.3535,
"step": 229376
},
{
"epoch": 1.0,
"learning_rate": 4.623601987207499e-05,
"loss": 4.3428,
"step": 229888
},
{
"epoch": 1.0,
"learning_rate": 4.6227633924564476e-05,
"loss": 4.3615,
"step": 230400
},
{
"epoch": 1.0,
"learning_rate": 4.6219247977053956e-05,
"loss": 4.3486,
"step": 230912
},
{
"epoch": 1.0,
"learning_rate": 4.6210862029543436e-05,
"loss": 4.3612,
"step": 231424
},
{
"epoch": 1.0,
"learning_rate": 4.6202476082032916e-05,
"loss": 4.3447,
"step": 231936
},
{
"epoch": 1.0,
"learning_rate": 4.6194090134522396e-05,
"loss": 4.3496,
"step": 232448
},
{
"epoch": 1.0,
"learning_rate": 4.6185704187011876e-05,
"loss": 4.3354,
"step": 232960
},
{
"epoch": 1.0,
"learning_rate": 4.6177318239501356e-05,
"loss": 4.3521,
"step": 233472
},
{
"epoch": 1.0,
"learning_rate": 4.6168932291990836e-05,
"loss": 4.3534,
"step": 233984
},
{
"epoch": 1.0,
"learning_rate": 4.6160562723284045e-05,
"loss": 4.3559,
"step": 234496
},
{
"epoch": 1.0,
"learning_rate": 4.6152176775773525e-05,
"loss": 4.3493,
"step": 235008
},
{
"epoch": 1.0,
"learning_rate": 4.6143790828263005e-05,
"loss": 4.3382,
"step": 235520
},
{
"epoch": 1.0,
"learning_rate": 4.6135404880752485e-05,
"loss": 4.3415,
"step": 236032
},
{
"epoch": 1.0,
"learning_rate": 4.612701893324196e-05,
"loss": 4.3355,
"step": 236544
},
{
"epoch": 1.0,
"learning_rate": 4.6118649364535174e-05,
"loss": 4.3415,
"step": 237056
},
{
"epoch": 1.0,
"learning_rate": 4.6110263417024654e-05,
"loss": 4.3451,
"step": 237568
},
{
"epoch": 1.0,
"learning_rate": 4.610187746951414e-05,
"loss": 4.3354,
"step": 238080
},
{
"epoch": 1.0,
"learning_rate": 4.609350790080735e-05,
"loss": 4.3352,
"step": 238592
},
{
"epoch": 1.0,
"learning_rate": 4.608512195329683e-05,
"loss": 4.3589,
"step": 239104
},
{
"epoch": 1.0,
"learning_rate": 4.607673600578631e-05,
"loss": 4.3413,
"step": 239616
},
{
"epoch": 1.0,
"learning_rate": 4.606835005827578e-05,
"loss": 4.3424,
"step": 240128
},
{
"epoch": 1.0,
"learning_rate": 4.605996411076526e-05,
"loss": 4.3364,
"step": 240640
},
{
"epoch": 1.0,
"learning_rate": 4.605157816325474e-05,
"loss": 4.3488,
"step": 241152
},
{
"epoch": 1.0,
"learning_rate": 4.604319221574422e-05,
"loss": 4.3321,
"step": 241664
},
{
"epoch": 1.0,
"learning_rate": 4.60348062682337e-05,
"loss": 4.342,
"step": 242176
},
{
"epoch": 1.0,
"learning_rate": 4.602642032072318e-05,
"loss": 4.3292,
"step": 242688
},
{
"epoch": 1.0,
"learning_rate": 4.601803437321266e-05,
"loss": 4.3307,
"step": 243200
},
{
"epoch": 1.0,
"learning_rate": 4.600964842570214e-05,
"loss": 4.3206,
"step": 243712
},
{
"epoch": 1.01,
"learning_rate": 4.600127885699536e-05,
"loss": 4.3315,
"step": 244224
},
{
"epoch": 1.01,
"learning_rate": 4.599289290948484e-05,
"loss": 4.3346,
"step": 244736
},
{
"epoch": 1.01,
"learning_rate": 4.598450696197432e-05,
"loss": 4.3382,
"step": 245248
},
{
"epoch": 1.01,
"learning_rate": 4.59761210144638e-05,
"loss": 4.3362,
"step": 245760
},
{
"epoch": 1.01,
"learning_rate": 4.596773506695328e-05,
"loss": 4.3374,
"step": 246272
},
{
"epoch": 1.01,
"learning_rate": 4.595936549824649e-05,
"loss": 4.3297,
"step": 246784
},
{
"epoch": 1.01,
"learning_rate": 4.5950995929539696e-05,
"loss": 4.3383,
"step": 247296
},
{
"epoch": 1.01,
"learning_rate": 4.5942626360832906e-05,
"loss": 4.3191,
"step": 247808
},
{
"epoch": 1.01,
"learning_rate": 4.5934240413322386e-05,
"loss": 4.3283,
"step": 248320
},
{
"epoch": 1.01,
"learning_rate": 4.5925854465811865e-05,
"loss": 4.3142,
"step": 248832
},
{
"epoch": 1.01,
"learning_rate": 4.5917468518301345e-05,
"loss": 4.3263,
"step": 249344
},
{
"epoch": 1.01,
"learning_rate": 4.590908257079083e-05,
"loss": 4.3185,
"step": 249856
},
{
"epoch": 1.01,
"learning_rate": 4.590069662328031e-05,
"loss": 4.3279,
"step": 250368
},
{
"epoch": 1.01,
"learning_rate": 4.589231067576979e-05,
"loss": 4.3237,
"step": 250880
},
{
"epoch": 1.01,
"learning_rate": 4.588392472825927e-05,
"loss": 4.316,
"step": 251392
},
{
"epoch": 1.01,
"learning_rate": 4.587553878074875e-05,
"loss": 4.3293,
"step": 251904
},
{
"epoch": 1.01,
"learning_rate": 4.586715283323823e-05,
"loss": 4.321,
"step": 252416
},
{
"epoch": 1.01,
"learning_rate": 4.585876688572771e-05,
"loss": 4.3188,
"step": 252928
},
{
"epoch": 1.01,
"learning_rate": 4.585038093821719e-05,
"loss": 4.3079,
"step": 253440
},
{
"epoch": 1.01,
"learning_rate": 4.58420113695104e-05,
"loss": 4.3097,
"step": 253952
},
{
"epoch": 1.01,
"learning_rate": 4.583362542199988e-05,
"loss": 4.3225,
"step": 254464
},
{
"epoch": 1.01,
"learning_rate": 4.582523947448936e-05,
"loss": 4.3189,
"step": 254976
},
{
"epoch": 1.01,
"learning_rate": 4.581685352697884e-05,
"loss": 4.3169,
"step": 255488
},
{
"epoch": 1.01,
"learning_rate": 4.580846757946832e-05,
"loss": 4.3046,
"step": 256000
},
{
"epoch": 1.01,
"learning_rate": 4.58000816319578e-05,
"loss": 4.3161,
"step": 256512
},
{
"epoch": 1.01,
"learning_rate": 4.579169568444728e-05,
"loss": 4.3028,
"step": 257024
},
{
"epoch": 1.01,
"learning_rate": 4.578330973693676e-05,
"loss": 4.3234,
"step": 257536
},
{
"epoch": 1.01,
"learning_rate": 4.577494016822997e-05,
"loss": 4.3043,
"step": 258048
},
{
"epoch": 1.01,
"learning_rate": 4.576655422071945e-05,
"loss": 4.3115,
"step": 258560
},
{
"epoch": 1.01,
"learning_rate": 4.575816827320893e-05,
"loss": 4.3148,
"step": 259072
},
{
"epoch": 1.01,
"learning_rate": 4.5749798704502146e-05,
"loss": 4.2994,
"step": 259584
},
{
"epoch": 1.01,
"learning_rate": 4.5741429135795355e-05,
"loss": 4.305,
"step": 260096
},
{
"epoch": 1.01,
"learning_rate": 4.5733043188284835e-05,
"loss": 4.3128,
"step": 260608
},
{
"epoch": 1.01,
"learning_rate": 4.5724657240774315e-05,
"loss": 4.2954,
"step": 261120
},
{
"epoch": 1.01,
"learning_rate": 4.5716271293263794e-05,
"loss": 4.3092,
"step": 261632
},
{
"epoch": 1.01,
"learning_rate": 4.570788534575327e-05,
"loss": 4.3083,
"step": 262144
},
{
"epoch": 1.01,
"learning_rate": 4.5699499398242754e-05,
"loss": 4.3089,
"step": 262656
},
{
"epoch": 1.01,
"learning_rate": 4.5691129829535963e-05,
"loss": 4.2992,
"step": 263168
},
{
"epoch": 1.01,
"learning_rate": 4.5682743882025443e-05,
"loss": 4.3049,
"step": 263680
},
{
"epoch": 1.01,
"learning_rate": 4.5674357934514923e-05,
"loss": 4.2938,
"step": 264192
},
{
"epoch": 1.01,
"learning_rate": 4.56659719870044e-05,
"loss": 4.3051,
"step": 264704
},
{
"epoch": 1.01,
"learning_rate": 4.565758603949388e-05,
"loss": 4.3178,
"step": 265216
},
{
"epoch": 1.01,
"learning_rate": 4.564920009198336e-05,
"loss": 4.3012,
"step": 265728
},
{
"epoch": 1.01,
"learning_rate": 4.564081414447284e-05,
"loss": 4.3143,
"step": 266240
},
{
"epoch": 1.01,
"learning_rate": 4.563244457576605e-05,
"loss": 4.3157,
"step": 266752
},
{
"epoch": 1.01,
"learning_rate": 4.562405862825553e-05,
"loss": 4.3125,
"step": 267264
},
{
"epoch": 1.01,
"learning_rate": 4.561567268074501e-05,
"loss": 4.2891,
"step": 267776
},
{
"epoch": 1.01,
"learning_rate": 4.560728673323449e-05,
"loss": 4.3096,
"step": 268288
},
{
"epoch": 1.01,
"learning_rate": 4.559890078572397e-05,
"loss": 4.3064,
"step": 268800
},
{
"epoch": 1.01,
"learning_rate": 4.559053121701719e-05,
"loss": 4.2971,
"step": 269312
},
{
"epoch": 1.01,
"learning_rate": 4.558214526950667e-05,
"loss": 4.2998,
"step": 269824
},
{
"epoch": 1.01,
"learning_rate": 4.557375932199615e-05,
"loss": 4.3027,
"step": 270336
},
{
"epoch": 1.01,
"learning_rate": 4.556537337448563e-05,
"loss": 4.3156,
"step": 270848
},
{
"epoch": 1.01,
"learning_rate": 4.555698742697511e-05,
"loss": 4.2985,
"step": 271360
},
{
"epoch": 1.01,
"learning_rate": 4.554860147946459e-05,
"loss": 4.2859,
"step": 271872
},
{
"epoch": 1.01,
"learning_rate": 4.554021553195407e-05,
"loss": 4.2978,
"step": 272384
},
{
"epoch": 1.01,
"learning_rate": 4.553182958444355e-05,
"loss": 4.2967,
"step": 272896
},
{
"epoch": 1.01,
"learning_rate": 4.552346001573676e-05,
"loss": 4.3011,
"step": 273408
},
{
"epoch": 1.01,
"learning_rate": 4.551507406822624e-05,
"loss": 4.3043,
"step": 273920
},
{
"epoch": 1.01,
"learning_rate": 4.550668812071572e-05,
"loss": 4.2943,
"step": 274432
},
{
"epoch": 1.02,
"learning_rate": 4.54983021732052e-05,
"loss": 4.2919,
"step": 274944
},
{
"epoch": 1.02,
"learning_rate": 4.548991622569468e-05,
"loss": 4.2842,
"step": 275456
},
{
"epoch": 1.02,
"learning_rate": 4.5481546656987886e-05,
"loss": 4.2899,
"step": 275968
},
{
"epoch": 1.02,
"learning_rate": 4.547316070947737e-05,
"loss": 4.2915,
"step": 276480
},
{
"epoch": 1.02,
"learning_rate": 4.546479114077058e-05,
"loss": 4.3002,
"step": 276992
},
{
"epoch": 1.02,
"learning_rate": 4.545640519326006e-05,
"loss": 4.3132,
"step": 277504
},
{
"epoch": 1.02,
"learning_rate": 4.544801924574954e-05,
"loss": 4.291,
"step": 278016
},
{
"epoch": 1.02,
"learning_rate": 4.543963329823902e-05,
"loss": 4.2851,
"step": 278528
},
{
"epoch": 1.02,
"learning_rate": 4.54312473507285e-05,
"loss": 4.2883,
"step": 279040
},
{
"epoch": 1.02,
"learning_rate": 4.542286140321798e-05,
"loss": 4.2886,
"step": 279552
},
{
"epoch": 1.02,
"learning_rate": 4.5414475455707455e-05,
"loss": 4.2976,
"step": 280064
},
{
"epoch": 1.02,
"learning_rate": 4.5406089508196934e-05,
"loss": 4.288,
"step": 280576
},
{
"epoch": 1.02,
"learning_rate": 4.5397703560686414e-05,
"loss": 4.2905,
"step": 281088
},
{
"epoch": 1.02,
"learning_rate": 4.5389317613175894e-05,
"loss": 4.2851,
"step": 281600
},
{
"epoch": 1.02,
"learning_rate": 4.5380931665665374e-05,
"loss": 4.2931,
"step": 282112
},
{
"epoch": 1.02,
"learning_rate": 4.5372545718154854e-05,
"loss": 4.2693,
"step": 282624
},
{
"epoch": 1.02,
"learning_rate": 4.536417614944807e-05,
"loss": 4.2849,
"step": 283136
},
{
"epoch": 1.02,
"learning_rate": 4.535579020193755e-05,
"loss": 4.2763,
"step": 283648
},
{
"epoch": 1.02,
"learning_rate": 4.534740425442703e-05,
"loss": 4.2886,
"step": 284160
},
{
"epoch": 1.02,
"learning_rate": 4.533901830691651e-05,
"loss": 4.2799,
"step": 284672
},
{
"epoch": 1.02,
"learning_rate": 4.533064873820972e-05,
"loss": 4.2878,
"step": 285184
},
{
"epoch": 1.02,
"learning_rate": 4.53222627906992e-05,
"loss": 4.2733,
"step": 285696
},
{
"epoch": 1.02,
"learning_rate": 4.531389322199241e-05,
"loss": 4.2751,
"step": 286208
},
{
"epoch": 1.02,
"learning_rate": 4.530550727448189e-05,
"loss": 4.2786,
"step": 286720
},
{
"epoch": 1.02,
"learning_rate": 4.529712132697137e-05,
"loss": 4.2808,
"step": 287232
},
{
"epoch": 1.02,
"learning_rate": 4.528873537946085e-05,
"loss": 4.2936,
"step": 287744
},
{
"epoch": 1.02,
"learning_rate": 4.528034943195033e-05,
"loss": 4.2902,
"step": 288256
},
{
"epoch": 1.02,
"learning_rate": 4.527199624204728e-05,
"loss": 4.2856,
"step": 288768
},
{
"epoch": 1.02,
"learning_rate": 4.526361029453675e-05,
"loss": 4.276,
"step": 289280
},
{
"epoch": 1.02,
"learning_rate": 4.525522434702623e-05,
"loss": 4.2792,
"step": 289792
},
{
"epoch": 1.02,
"learning_rate": 4.524683839951571e-05,
"loss": 4.284,
"step": 290304
},
{
"epoch": 1.02,
"learning_rate": 4.523845245200519e-05,
"loss": 4.2738,
"step": 290816
},
{
"epoch": 1.02,
"learning_rate": 4.523006650449467e-05,
"loss": 4.2827,
"step": 291328
},
{
"epoch": 1.02,
"learning_rate": 4.522168055698415e-05,
"loss": 4.2741,
"step": 291840
},
{
"epoch": 1.02,
"learning_rate": 4.521329460947363e-05,
"loss": 4.272,
"step": 292352
},
{
"epoch": 1.02,
"learning_rate": 4.520492504076684e-05,
"loss": 4.2718,
"step": 292864
},
{
"epoch": 1.02,
"learning_rate": 4.519653909325632e-05,
"loss": 4.2723,
"step": 293376
},
{
"epoch": 1.02,
"learning_rate": 4.51881531457458e-05,
"loss": 4.2669,
"step": 293888
},
{
"epoch": 1.02,
"learning_rate": 4.517976719823528e-05,
"loss": 4.2785,
"step": 294400
},
{
"epoch": 1.02,
"learning_rate": 4.517138125072476e-05,
"loss": 4.2603,
"step": 294912
},
{
"epoch": 1.02,
"learning_rate": 4.516299530321425e-05,
"loss": 4.2864,
"step": 295424
},
{
"epoch": 1.02,
"learning_rate": 4.515460935570373e-05,
"loss": 4.2845,
"step": 295936
},
{
"epoch": 1.02,
"learning_rate": 4.514622340819321e-05,
"loss": 4.2832,
"step": 296448
},
{
"epoch": 1.02,
"learning_rate": 4.513785383948642e-05,
"loss": 4.2609,
"step": 296960
},
{
"epoch": 1.02,
"learning_rate": 4.51294678919759e-05,
"loss": 4.2751,
"step": 297472
},
{
"epoch": 1.02,
"learning_rate": 4.512108194446538e-05,
"loss": 4.2797,
"step": 297984
},
{
"epoch": 1.02,
"learning_rate": 4.511269599695486e-05,
"loss": 4.2774,
"step": 298496
},
{
"epoch": 1.02,
"learning_rate": 4.5104326428248066e-05,
"loss": 4.2733,
"step": 299008
},
{
"epoch": 1.02,
"learning_rate": 4.5095940480737546e-05,
"loss": 4.2722,
"step": 299520
},
{
"epoch": 1.02,
"learning_rate": 4.5087554533227026e-05,
"loss": 4.272,
"step": 300032
},
{
"epoch": 1.02,
"learning_rate": 4.5079168585716506e-05,
"loss": 4.2778,
"step": 300544
},
{
"epoch": 1.02,
"learning_rate": 4.5070782638205986e-05,
"loss": 4.2677,
"step": 301056
},
{
"epoch": 1.02,
"learning_rate": 4.5062413069499195e-05,
"loss": 4.2727,
"step": 301568
},
{
"epoch": 1.02,
"learning_rate": 4.505402712198868e-05,
"loss": 4.2694,
"step": 302080
},
{
"epoch": 1.02,
"learning_rate": 4.504565755328189e-05,
"loss": 4.2628,
"step": 302592
},
{
"epoch": 1.02,
"learning_rate": 4.503727160577137e-05,
"loss": 4.2747,
"step": 303104
},
{
"epoch": 1.02,
"learning_rate": 4.502888565826085e-05,
"loss": 4.2687,
"step": 303616
},
{
"epoch": 1.02,
"learning_rate": 4.502049971075033e-05,
"loss": 4.2677,
"step": 304128
},
{
"epoch": 1.02,
"learning_rate": 4.501213014204354e-05,
"loss": 4.2591,
"step": 304640
},
{
"epoch": 1.02,
"learning_rate": 4.500374419453302e-05,
"loss": 4.2585,
"step": 305152
},
{
"epoch": 1.03,
"eval_loss": 4.248311519622803,
"eval_runtime": 297.0154,
"eval_samples_per_second": 1284.752,
"eval_steps_per_second": 40.149,
"step": 305280
},
{
"epoch": 0.0,
"learning_rate": 4.49953582470225e-05,
"loss": 4.2613,
"step": 305664
},
{
"epoch": 0.0,
"learning_rate": 4.498697229951198e-05,
"loss": 4.2525,
"step": 306176
},
{
"epoch": 0.0,
"learning_rate": 4.497858635200146e-05,
"loss": 4.2691,
"step": 306688
},
{
"epoch": 0.0,
"learning_rate": 4.497020040449093e-05,
"loss": 4.2616,
"step": 307200
},
{
"epoch": 0.0,
"learning_rate": 4.496181445698042e-05,
"loss": 4.2749,
"step": 307712
},
{
"epoch": 0.0,
"learning_rate": 4.49534285094699e-05,
"loss": 4.2532,
"step": 308224
},
{
"epoch": 0.0,
"learning_rate": 4.494504256195938e-05,
"loss": 4.2595,
"step": 308736
},
{
"epoch": 0.0,
"learning_rate": 4.493665661444886e-05,
"loss": 4.246,
"step": 309248
},
{
"epoch": 0.0,
"learning_rate": 4.492827066693834e-05,
"loss": 4.2662,
"step": 309760
},
{
"epoch": 0.0,
"learning_rate": 4.491988471942782e-05,
"loss": 4.2653,
"step": 310272
},
{
"epoch": 0.0,
"learning_rate": 4.49114987719173e-05,
"loss": 4.2652,
"step": 310784
},
{
"epoch": 0.0,
"learning_rate": 4.490311282440678e-05,
"loss": 4.2625,
"step": 311296
},
{
"epoch": 0.0,
"learning_rate": 4.489472687689626e-05,
"loss": 4.2492,
"step": 311808
},
{
"epoch": 0.0,
"learning_rate": 4.488634092938574e-05,
"loss": 4.2603,
"step": 312320
},
{
"epoch": 0.0,
"learning_rate": 4.487797136067895e-05,
"loss": 4.2474,
"step": 312832
},
{
"epoch": 0.0,
"learning_rate": 4.486958541316843e-05,
"loss": 4.257,
"step": 313344
},
{
"epoch": 0.0,
"learning_rate": 4.486119946565791e-05,
"loss": 4.2542,
"step": 313856
},
{
"epoch": 0.0,
"learning_rate": 4.485281351814739e-05,
"loss": 4.2492,
"step": 314368
},
{
"epoch": 0.0,
"learning_rate": 4.484442757063687e-05,
"loss": 4.2554,
"step": 314880
},
{
"epoch": 0.0,
"learning_rate": 4.483604162312635e-05,
"loss": 4.2701,
"step": 315392
},
{
"epoch": 0.0,
"learning_rate": 4.482765567561583e-05,
"loss": 4.2526,
"step": 315904
},
{
"epoch": 0.0,
"learning_rate": 4.481926972810531e-05,
"loss": 4.2572,
"step": 316416
},
{
"epoch": 0.0,
"learning_rate": 4.481088378059479e-05,
"loss": 4.2514,
"step": 316928
},
{
"epoch": 0.0,
"learning_rate": 4.480249783308427e-05,
"loss": 4.2623,
"step": 317440
},
{
"epoch": 0.0,
"learning_rate": 4.479411188557375e-05,
"loss": 4.2501,
"step": 317952
},
{
"epoch": 0.0,
"learning_rate": 4.478572593806323e-05,
"loss": 4.2548,
"step": 318464
},
{
"epoch": 0.0,
"learning_rate": 4.477735636935644e-05,
"loss": 4.2465,
"step": 318976
},
{
"epoch": 0.0,
"learning_rate": 4.476898680064965e-05,
"loss": 4.2451,
"step": 319488
},
{
"epoch": 0.0,
"learning_rate": 4.4760600853139126e-05,
"loss": 4.2384,
"step": 320000
},
{
"epoch": 0.0,
"learning_rate": 4.4752214905628606e-05,
"loss": 4.2469,
"step": 320512
},
{
"epoch": 0.01,
"learning_rate": 4.4743828958118086e-05,
"loss": 4.2502,
"step": 321024
},
{
"epoch": 0.01,
"learning_rate": 4.473544301060757e-05,
"loss": 4.2563,
"step": 321536
},
{
"epoch": 0.01,
"learning_rate": 4.472705706309705e-05,
"loss": 4.255,
"step": 322048
},
{
"epoch": 0.01,
"learning_rate": 4.471868749439026e-05,
"loss": 4.2527,
"step": 322560
},
{
"epoch": 0.01,
"learning_rate": 4.471030154687974e-05,
"loss": 4.2489,
"step": 323072
},
{
"epoch": 0.01,
"learning_rate": 4.470191559936922e-05,
"loss": 4.2543,
"step": 323584
},
{
"epoch": 0.01,
"learning_rate": 4.469354603066243e-05,
"loss": 4.2428,
"step": 324096
},
{
"epoch": 0.01,
"learning_rate": 4.468516008315191e-05,
"loss": 4.2443,
"step": 324608
},
{
"epoch": 0.01,
"learning_rate": 4.467677413564139e-05,
"loss": 4.2318,
"step": 325120
},
{
"epoch": 0.01,
"learning_rate": 4.466838818813087e-05,
"loss": 4.243,
"step": 325632
},
{
"epoch": 0.01,
"learning_rate": 4.466000224062035e-05,
"loss": 4.24,
"step": 326144
},
{
"epoch": 0.01,
"learning_rate": 4.465161629310983e-05,
"loss": 4.2485,
"step": 326656
},
{
"epoch": 0.01,
"learning_rate": 4.464323034559931e-05,
"loss": 4.2437,
"step": 327168
},
{
"epoch": 0.01,
"learning_rate": 4.463484439808879e-05,
"loss": 4.2344,
"step": 327680
},
{
"epoch": 0.01,
"learning_rate": 4.462645845057827e-05,
"loss": 4.2501,
"step": 328192
},
{
"epoch": 0.01,
"learning_rate": 4.4618088881871486e-05,
"loss": 4.2415,
"step": 328704
},
{
"epoch": 0.01,
"learning_rate": 4.4609702934360966e-05,
"loss": 4.239,
"step": 329216
},
{
"epoch": 0.01,
"learning_rate": 4.4601316986850446e-05,
"loss": 4.2308,
"step": 329728
},
{
"epoch": 0.01,
"learning_rate": 4.4592931039339926e-05,
"loss": 4.2276,
"step": 330240
},
{
"epoch": 0.01,
"learning_rate": 4.4584545091829406e-05,
"loss": 4.2427,
"step": 330752
},
{
"epoch": 0.01,
"learning_rate": 4.4576175523122615e-05,
"loss": 4.2398,
"step": 331264
},
{
"epoch": 0.01,
"learning_rate": 4.4567789575612095e-05,
"loss": 4.2373,
"step": 331776
},
{
"epoch": 0.01,
"learning_rate": 4.4559403628101575e-05,
"loss": 4.2292,
"step": 332288
},
{
"epoch": 0.01,
"learning_rate": 4.4551017680591055e-05,
"loss": 4.2384,
"step": 332800
},
{
"epoch": 0.01,
"learning_rate": 4.4542631733080535e-05,
"loss": 4.2243,
"step": 333312
},
{
"epoch": 0.01,
"learning_rate": 4.453424578557001e-05,
"loss": 4.2426,
"step": 333824
},
{
"epoch": 0.01,
"learning_rate": 4.4525876216863224e-05,
"loss": 4.2222,
"step": 334336
},
{
"epoch": 0.01,
"learning_rate": 4.4517490269352704e-05,
"loss": 4.2376,
"step": 334848
},
{
"epoch": 0.01,
"learning_rate": 4.450910432184219e-05,
"loss": 4.2347,
"step": 335360
},
{
"epoch": 0.01,
"learning_rate": 4.4500718374331664e-05,
"loss": 4.2276,
"step": 335872
},
{
"epoch": 0.01,
"learning_rate": 4.4492332426821144e-05,
"loss": 4.2258,
"step": 336384
},
{
"epoch": 0.01,
"learning_rate": 4.4483946479310624e-05,
"loss": 4.2342,
"step": 336896
},
{
"epoch": 0.01,
"learning_rate": 4.4475560531800104e-05,
"loss": 4.2197,
"step": 337408
},
{
"epoch": 0.01,
"learning_rate": 4.4467174584289584e-05,
"loss": 4.2319,
"step": 337920
},
{
"epoch": 0.01,
"learning_rate": 4.445880501558279e-05,
"loss": 4.2343,
"step": 338432
},
{
"epoch": 0.01,
"learning_rate": 4.445041906807227e-05,
"loss": 4.2321,
"step": 338944
},
{
"epoch": 0.01,
"learning_rate": 4.444203312056175e-05,
"loss": 4.2243,
"step": 339456
},
{
"epoch": 0.01,
"learning_rate": 4.443364717305123e-05,
"loss": 4.2264,
"step": 339968
},
{
"epoch": 0.01,
"learning_rate": 4.442526122554071e-05,
"loss": 4.2205,
"step": 340480
},
{
"epoch": 0.01,
"learning_rate": 4.441687527803019e-05,
"loss": 4.2244,
"step": 340992
},
{
"epoch": 0.01,
"learning_rate": 4.440848933051967e-05,
"loss": 4.2391,
"step": 341504
},
{
"epoch": 0.01,
"learning_rate": 4.440010338300916e-05,
"loss": 4.2302,
"step": 342016
},
{
"epoch": 0.01,
"learning_rate": 4.439171743549864e-05,
"loss": 4.2382,
"step": 342528
},
{
"epoch": 0.01,
"learning_rate": 4.438334786679185e-05,
"loss": 4.242,
"step": 343040
},
{
"epoch": 0.01,
"learning_rate": 4.437496191928133e-05,
"loss": 4.2396,
"step": 343552
},
{
"epoch": 0.01,
"learning_rate": 4.436657597177081e-05,
"loss": 4.2164,
"step": 344064
},
{
"epoch": 0.01,
"learning_rate": 4.435820640306402e-05,
"loss": 4.2294,
"step": 344576
},
{
"epoch": 0.01,
"learning_rate": 4.43498204555535e-05,
"loss": 4.2363,
"step": 345088
},
{
"epoch": 0.01,
"learning_rate": 4.434143450804298e-05,
"loss": 4.2215,
"step": 345600
},
{
"epoch": 0.01,
"learning_rate": 4.433304856053246e-05,
"loss": 4.2255,
"step": 346112
},
{
"epoch": 0.01,
"learning_rate": 4.432466261302194e-05,
"loss": 4.2276,
"step": 346624
},
{
"epoch": 0.01,
"learning_rate": 4.4316293044315146e-05,
"loss": 4.2445,
"step": 347136
},
{
"epoch": 0.01,
"learning_rate": 4.4307907096804626e-05,
"loss": 4.2279,
"step": 347648
},
{
"epoch": 0.01,
"learning_rate": 4.429953752809784e-05,
"loss": 4.2098,
"step": 348160
},
{
"epoch": 0.01,
"learning_rate": 4.429115158058732e-05,
"loss": 4.2249,
"step": 348672
},
{
"epoch": 0.01,
"learning_rate": 4.42827656330768e-05,
"loss": 4.2226,
"step": 349184
},
{
"epoch": 0.01,
"learning_rate": 4.427437968556628e-05,
"loss": 4.2286,
"step": 349696
},
{
"epoch": 0.01,
"learning_rate": 4.426599373805576e-05,
"loss": 4.2305,
"step": 350208
},
{
"epoch": 0.01,
"learning_rate": 4.425760779054524e-05,
"loss": 4.2194,
"step": 350720
},
{
"epoch": 0.02,
"learning_rate": 4.424922184303472e-05,
"loss": 4.2177,
"step": 351232
},
{
"epoch": 0.02,
"learning_rate": 4.42408358955242e-05,
"loss": 4.2171,
"step": 351744
},
{
"epoch": 0.02,
"learning_rate": 4.423246632681741e-05,
"loss": 4.2146,
"step": 352256
},
{
"epoch": 0.02,
"learning_rate": 4.422408037930689e-05,
"loss": 4.2235,
"step": 352768
},
{
"epoch": 0.02,
"learning_rate": 4.421569443179637e-05,
"loss": 4.2268,
"step": 353280
},
{
"epoch": 0.02,
"learning_rate": 4.420732486308958e-05,
"loss": 4.2394,
"step": 353792
},
{
"epoch": 0.02,
"learning_rate": 4.419893891557907e-05,
"loss": 4.2264,
"step": 354304
},
{
"epoch": 0.02,
"learning_rate": 4.419055296806855e-05,
"loss": 4.2123,
"step": 354816
},
{
"epoch": 0.02,
"learning_rate": 4.418216702055803e-05,
"loss": 4.2169,
"step": 355328
},
{
"epoch": 0.02,
"learning_rate": 4.41737810730475e-05,
"loss": 4.2143,
"step": 355840
},
{
"epoch": 0.02,
"learning_rate": 4.416539512553698e-05,
"loss": 4.2268,
"step": 356352
},
{
"epoch": 0.02,
"learning_rate": 4.415700917802646e-05,
"loss": 4.2219,
"step": 356864
},
{
"epoch": 0.02,
"learning_rate": 4.4148639609319676e-05,
"loss": 4.222,
"step": 357376
},
{
"epoch": 0.02,
"learning_rate": 4.414025366180915e-05,
"loss": 4.2149,
"step": 357888
},
{
"epoch": 0.02,
"learning_rate": 4.413186771429863e-05,
"loss": 4.2235,
"step": 358400
},
{
"epoch": 0.02,
"learning_rate": 4.412348176678811e-05,
"loss": 4.2017,
"step": 358912
},
{
"epoch": 0.02,
"learning_rate": 4.411511219808132e-05,
"loss": 4.2174,
"step": 359424
},
{
"epoch": 0.02,
"learning_rate": 4.4106726250570805e-05,
"loss": 4.2029,
"step": 359936
},
{
"epoch": 0.02,
"learning_rate": 4.4098340303060284e-05,
"loss": 4.2223,
"step": 360448
},
{
"epoch": 0.02,
"learning_rate": 4.4089954355549764e-05,
"loss": 4.2113,
"step": 360960
},
{
"epoch": 0.02,
"learning_rate": 4.4081568408039244e-05,
"loss": 4.2168,
"step": 361472
},
{
"epoch": 0.02,
"learning_rate": 4.4073182460528724e-05,
"loss": 4.2046,
"step": 361984
},
{
"epoch": 0.02,
"learning_rate": 4.4064796513018204e-05,
"loss": 4.2097,
"step": 362496
},
{
"epoch": 0.02,
"learning_rate": 4.4056410565507684e-05,
"loss": 4.2111,
"step": 363008
},
{
"epoch": 0.02,
"learning_rate": 4.404804099680089e-05,
"loss": 4.2109,
"step": 363520
},
{
"epoch": 0.02,
"learning_rate": 4.40396714280941e-05,
"loss": 4.2236,
"step": 364032
},
{
"epoch": 0.02,
"learning_rate": 4.403128548058358e-05,
"loss": 4.2225,
"step": 364544
},
{
"epoch": 0.02,
"learning_rate": 4.402291591187679e-05,
"loss": 4.2176,
"step": 365056
},
{
"epoch": 0.02,
"learning_rate": 4.401452996436627e-05,
"loss": 4.2071,
"step": 365568
},
{
"epoch": 0.02,
"learning_rate": 4.400614401685576e-05,
"loss": 4.2096,
"step": 366080
},
{
"epoch": 0.02,
"learning_rate": 4.399775806934524e-05,
"loss": 4.2192,
"step": 366592
},
{
"epoch": 0.02,
"learning_rate": 4.398937212183472e-05,
"loss": 4.2004,
"step": 367104
},
{
"epoch": 0.02,
"learning_rate": 4.39809861743242e-05,
"loss": 4.2185,
"step": 367616
},
{
"epoch": 0.02,
"learning_rate": 4.397260022681368e-05,
"loss": 4.2013,
"step": 368128
},
{
"epoch": 0.02,
"learning_rate": 4.396421427930316e-05,
"loss": 4.2081,
"step": 368640
},
{
"epoch": 0.02,
"learning_rate": 4.395582833179264e-05,
"loss": 4.2067,
"step": 369152
},
{
"epoch": 0.02,
"learning_rate": 4.394744238428212e-05,
"loss": 4.2057,
"step": 369664
},
{
"epoch": 0.02,
"learning_rate": 4.39390564367716e-05,
"loss": 4.1961,
"step": 370176
},
{
"epoch": 0.02,
"learning_rate": 4.393068686806481e-05,
"loss": 4.214,
"step": 370688
},
{
"epoch": 0.02,
"learning_rate": 4.3922317299358016e-05,
"loss": 4.1948,
"step": 371200
},
{
"epoch": 0.02,
"learning_rate": 4.3913931351847496e-05,
"loss": 4.2185,
"step": 371712
},
{
"epoch": 0.02,
"learning_rate": 4.3905545404336976e-05,
"loss": 4.2125,
"step": 372224
},
{
"epoch": 0.02,
"learning_rate": 4.3897159456826456e-05,
"loss": 4.2228,
"step": 372736
},
{
"epoch": 0.02,
"learning_rate": 4.3888773509315936e-05,
"loss": 4.1906,
"step": 373248
},
{
"epoch": 0.02,
"learning_rate": 4.388038756180542e-05,
"loss": 4.2114,
"step": 373760
},
{
"epoch": 0.02,
"learning_rate": 4.38720016142949e-05,
"loss": 4.2113,
"step": 374272
},
{
"epoch": 0.02,
"learning_rate": 4.386361566678438e-05,
"loss": 4.2114,
"step": 374784
},
{
"epoch": 0.02,
"learning_rate": 4.385522971927386e-05,
"loss": 4.2072,
"step": 375296
},
{
"epoch": 0.02,
"learning_rate": 4.384686015056707e-05,
"loss": 4.2105,
"step": 375808
},
{
"epoch": 0.02,
"learning_rate": 4.383847420305655e-05,
"loss": 4.205,
"step": 376320
},
{
"epoch": 0.02,
"learning_rate": 4.383008825554603e-05,
"loss": 4.2136,
"step": 376832
},
{
"epoch": 0.02,
"learning_rate": 4.3821702308035505e-05,
"loss": 4.1966,
"step": 377344
},
{
"epoch": 0.02,
"learning_rate": 4.3813316360524985e-05,
"loss": 4.2095,
"step": 377856
},
{
"epoch": 0.02,
"learning_rate": 4.3804930413014465e-05,
"loss": 4.2026,
"step": 378368
},
{
"epoch": 0.02,
"learning_rate": 4.3796544465503944e-05,
"loss": 4.2002,
"step": 378880
},
{
"epoch": 0.02,
"learning_rate": 4.3788158517993424e-05,
"loss": 4.2067,
"step": 379392
},
{
"epoch": 0.02,
"learning_rate": 4.3779772570482904e-05,
"loss": 4.2076,
"step": 379904
},
{
"epoch": 0.02,
"learning_rate": 4.377140300177612e-05,
"loss": 4.2003,
"step": 380416
},
{
"epoch": 0.02,
"learning_rate": 4.37630170542656e-05,
"loss": 4.1911,
"step": 380928
},
{
"epoch": 0.02,
"learning_rate": 4.375463110675508e-05,
"loss": 4.1993,
"step": 381440
},
{
"epoch": 0.03,
"eval_loss": 4.192408084869385,
"eval_runtime": 317.2732,
"eval_samples_per_second": 1202.72,
"eval_steps_per_second": 37.586,
"step": 381600
}
],
"logging_steps": 512,
"max_steps": 3052726,
"num_train_epochs": 9223372036854775807,
"save_steps": 10,
"total_flos": 1.579623036124167e+17,
"trial_name": null,
"trial_params": null
}