gpt2-60 / trainer_state.json
shrinath-suresh's picture
End of training
1d8e01b
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 30.0,
"global_step": 47460,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.32,
"learning_rate": 4.94732406236831e-05,
"loss": 1.2748,
"step": 500
},
{
"epoch": 0.63,
"learning_rate": 4.8946481247366205e-05,
"loss": 1.2878,
"step": 1000
},
{
"epoch": 0.95,
"learning_rate": 4.8419721871049303e-05,
"loss": 1.2988,
"step": 1500
},
{
"epoch": 1.26,
"learning_rate": 4.789296249473241e-05,
"loss": 1.2645,
"step": 2000
},
{
"epoch": 1.58,
"learning_rate": 4.7366203118415506e-05,
"loss": 1.2449,
"step": 2500
},
{
"epoch": 1.9,
"learning_rate": 4.683944374209861e-05,
"loss": 1.2536,
"step": 3000
},
{
"epoch": 2.21,
"learning_rate": 4.631268436578171e-05,
"loss": 1.2103,
"step": 3500
},
{
"epoch": 2.53,
"learning_rate": 4.5785924989464814e-05,
"loss": 1.2159,
"step": 4000
},
{
"epoch": 2.84,
"learning_rate": 4.525916561314791e-05,
"loss": 1.2245,
"step": 4500
},
{
"epoch": 3.16,
"learning_rate": 4.473240623683102e-05,
"loss": 1.1803,
"step": 5000
},
{
"epoch": 3.48,
"learning_rate": 4.420564686051412e-05,
"loss": 1.1739,
"step": 5500
},
{
"epoch": 3.79,
"learning_rate": 4.367888748419722e-05,
"loss": 1.1882,
"step": 6000
},
{
"epoch": 4.11,
"learning_rate": 4.3152128107880325e-05,
"loss": 1.169,
"step": 6500
},
{
"epoch": 4.42,
"learning_rate": 4.262536873156342e-05,
"loss": 1.1294,
"step": 7000
},
{
"epoch": 4.74,
"learning_rate": 4.209860935524653e-05,
"loss": 1.1658,
"step": 7500
},
{
"epoch": 5.06,
"learning_rate": 4.1571849978929626e-05,
"loss": 1.1395,
"step": 8000
},
{
"epoch": 5.37,
"learning_rate": 4.104509060261273e-05,
"loss": 1.0973,
"step": 8500
},
{
"epoch": 5.69,
"learning_rate": 4.051833122629583e-05,
"loss": 1.1112,
"step": 9000
},
{
"epoch": 6.01,
"learning_rate": 3.9991571849978934e-05,
"loss": 1.1368,
"step": 9500
},
{
"epoch": 6.32,
"learning_rate": 3.946481247366203e-05,
"loss": 1.0661,
"step": 10000
},
{
"epoch": 6.64,
"learning_rate": 3.893805309734514e-05,
"loss": 1.0824,
"step": 10500
},
{
"epoch": 6.95,
"learning_rate": 3.8411293721028235e-05,
"loss": 1.1049,
"step": 11000
},
{
"epoch": 7.27,
"learning_rate": 3.788453434471134e-05,
"loss": 1.0503,
"step": 11500
},
{
"epoch": 7.59,
"learning_rate": 3.735777496839444e-05,
"loss": 1.0634,
"step": 12000
},
{
"epoch": 7.9,
"learning_rate": 3.6831015592077536e-05,
"loss": 1.0592,
"step": 12500
},
{
"epoch": 8.22,
"learning_rate": 3.630425621576064e-05,
"loss": 1.0401,
"step": 13000
},
{
"epoch": 8.53,
"learning_rate": 3.577749683944374e-05,
"loss": 1.0138,
"step": 13500
},
{
"epoch": 8.85,
"learning_rate": 3.5250737463126844e-05,
"loss": 1.0445,
"step": 14000
},
{
"epoch": 9.17,
"learning_rate": 3.472397808680995e-05,
"loss": 1.0185,
"step": 14500
},
{
"epoch": 9.48,
"learning_rate": 3.419721871049305e-05,
"loss": 0.9941,
"step": 15000
},
{
"epoch": 9.8,
"learning_rate": 3.367045933417615e-05,
"loss": 1.0206,
"step": 15500
},
{
"epoch": 10.11,
"learning_rate": 3.3143699957859256e-05,
"loss": 1.0188,
"step": 16000
},
{
"epoch": 10.43,
"learning_rate": 3.2616940581542354e-05,
"loss": 0.972,
"step": 16500
},
{
"epoch": 10.75,
"learning_rate": 3.209018120522546e-05,
"loss": 0.9909,
"step": 17000
},
{
"epoch": 11.06,
"learning_rate": 3.156342182890856e-05,
"loss": 0.9792,
"step": 17500
},
{
"epoch": 11.38,
"learning_rate": 3.1036662452591655e-05,
"loss": 0.9532,
"step": 18000
},
{
"epoch": 11.69,
"learning_rate": 3.050990307627476e-05,
"loss": 0.9616,
"step": 18500
},
{
"epoch": 12.01,
"learning_rate": 2.998314369995786e-05,
"loss": 0.9773,
"step": 19000
},
{
"epoch": 12.33,
"learning_rate": 2.9456384323640963e-05,
"loss": 0.9291,
"step": 19500
},
{
"epoch": 12.64,
"learning_rate": 2.8929624947324065e-05,
"loss": 0.9526,
"step": 20000
},
{
"epoch": 12.96,
"learning_rate": 2.8402865571007166e-05,
"loss": 0.9516,
"step": 20500
},
{
"epoch": 13.27,
"learning_rate": 2.7876106194690264e-05,
"loss": 0.9112,
"step": 21000
},
{
"epoch": 13.59,
"learning_rate": 2.7349346818373366e-05,
"loss": 0.9198,
"step": 21500
},
{
"epoch": 13.91,
"learning_rate": 2.6822587442056467e-05,
"loss": 0.9436,
"step": 22000
},
{
"epoch": 14.22,
"learning_rate": 2.629582806573957e-05,
"loss": 0.9024,
"step": 22500
},
{
"epoch": 14.54,
"learning_rate": 2.576906868942267e-05,
"loss": 0.9076,
"step": 23000
},
{
"epoch": 14.85,
"learning_rate": 2.524230931310577e-05,
"loss": 0.9142,
"step": 23500
},
{
"epoch": 15.17,
"learning_rate": 2.4715549936788876e-05,
"loss": 0.9001,
"step": 24000
},
{
"epoch": 15.49,
"learning_rate": 2.4188790560471978e-05,
"loss": 0.8884,
"step": 24500
},
{
"epoch": 15.8,
"learning_rate": 2.366203118415508e-05,
"loss": 0.8909,
"step": 25000
},
{
"epoch": 16.12,
"learning_rate": 2.313527180783818e-05,
"loss": 0.8959,
"step": 25500
},
{
"epoch": 16.43,
"learning_rate": 2.2608512431521282e-05,
"loss": 0.8557,
"step": 26000
},
{
"epoch": 16.75,
"learning_rate": 2.2081753055204384e-05,
"loss": 0.8859,
"step": 26500
},
{
"epoch": 17.07,
"learning_rate": 2.1554993678887485e-05,
"loss": 0.8828,
"step": 27000
},
{
"epoch": 17.38,
"learning_rate": 2.1028234302570587e-05,
"loss": 0.8477,
"step": 27500
},
{
"epoch": 17.7,
"learning_rate": 2.0501474926253688e-05,
"loss": 0.8648,
"step": 28000
},
{
"epoch": 18.02,
"learning_rate": 1.997471554993679e-05,
"loss": 0.8785,
"step": 28500
},
{
"epoch": 18.33,
"learning_rate": 1.944795617361989e-05,
"loss": 0.8449,
"step": 29000
},
{
"epoch": 18.65,
"learning_rate": 1.8921196797302992e-05,
"loss": 0.8559,
"step": 29500
},
{
"epoch": 18.96,
"learning_rate": 1.8394437420986094e-05,
"loss": 0.8546,
"step": 30000
},
{
"epoch": 19.28,
"learning_rate": 1.7867678044669195e-05,
"loss": 0.8199,
"step": 30500
},
{
"epoch": 19.6,
"learning_rate": 1.7340918668352297e-05,
"loss": 0.8308,
"step": 31000
},
{
"epoch": 19.91,
"learning_rate": 1.6814159292035402e-05,
"loss": 0.8645,
"step": 31500
},
{
"epoch": 20.23,
"learning_rate": 1.62873999157185e-05,
"loss": 0.8307,
"step": 32000
},
{
"epoch": 20.54,
"learning_rate": 1.57606405394016e-05,
"loss": 0.8243,
"step": 32500
},
{
"epoch": 20.86,
"learning_rate": 1.5233881163084704e-05,
"loss": 0.8225,
"step": 33000
},
{
"epoch": 21.18,
"learning_rate": 1.4707121786767806e-05,
"loss": 0.8214,
"step": 33500
},
{
"epoch": 21.49,
"learning_rate": 1.4180362410450906e-05,
"loss": 0.8093,
"step": 34000
},
{
"epoch": 21.81,
"learning_rate": 1.3653603034134007e-05,
"loss": 0.817,
"step": 34500
},
{
"epoch": 22.12,
"learning_rate": 1.3126843657817109e-05,
"loss": 0.8373,
"step": 35000
},
{
"epoch": 22.44,
"learning_rate": 1.260008428150021e-05,
"loss": 0.8028,
"step": 35500
},
{
"epoch": 22.76,
"learning_rate": 1.2073324905183313e-05,
"loss": 0.8156,
"step": 36000
},
{
"epoch": 23.07,
"learning_rate": 1.1546565528866415e-05,
"loss": 0.8105,
"step": 36500
},
{
"epoch": 23.39,
"learning_rate": 1.1019806152549515e-05,
"loss": 0.8085,
"step": 37000
},
{
"epoch": 23.7,
"learning_rate": 1.0493046776232618e-05,
"loss": 0.8071,
"step": 37500
},
{
"epoch": 24.02,
"learning_rate": 9.96628739991572e-06,
"loss": 0.7915,
"step": 38000
},
{
"epoch": 24.34,
"learning_rate": 9.43952802359882e-06,
"loss": 0.7973,
"step": 38500
},
{
"epoch": 24.65,
"learning_rate": 8.912768647281922e-06,
"loss": 0.7963,
"step": 39000
},
{
"epoch": 24.97,
"learning_rate": 8.386009270965024e-06,
"loss": 0.7947,
"step": 39500
},
{
"epoch": 25.28,
"learning_rate": 7.859249894648125e-06,
"loss": 0.7929,
"step": 40000
},
{
"epoch": 25.6,
"learning_rate": 7.3324905183312265e-06,
"loss": 0.794,
"step": 40500
},
{
"epoch": 25.92,
"learning_rate": 6.805731142014328e-06,
"loss": 0.7901,
"step": 41000
},
{
"epoch": 26.23,
"learning_rate": 6.2789717656974295e-06,
"loss": 0.7732,
"step": 41500
},
{
"epoch": 26.55,
"learning_rate": 5.752212389380531e-06,
"loss": 0.7862,
"step": 42000
},
{
"epoch": 26.86,
"learning_rate": 5.225453013063633e-06,
"loss": 0.7973,
"step": 42500
},
{
"epoch": 27.18,
"learning_rate": 4.698693636746735e-06,
"loss": 0.7813,
"step": 43000
},
{
"epoch": 27.5,
"learning_rate": 4.171934260429835e-06,
"loss": 0.7834,
"step": 43500
},
{
"epoch": 27.81,
"learning_rate": 3.6451748841129377e-06,
"loss": 0.785,
"step": 44000
},
{
"epoch": 28.13,
"learning_rate": 3.1184155077960387e-06,
"loss": 0.7904,
"step": 44500
},
{
"epoch": 28.45,
"learning_rate": 2.59165613147914e-06,
"loss": 0.7809,
"step": 45000
},
{
"epoch": 28.76,
"learning_rate": 2.064896755162242e-06,
"loss": 0.7804,
"step": 45500
},
{
"epoch": 29.08,
"learning_rate": 1.5381373788453435e-06,
"loss": 0.7755,
"step": 46000
},
{
"epoch": 29.39,
"learning_rate": 1.0113780025284452e-06,
"loss": 0.7789,
"step": 46500
},
{
"epoch": 29.71,
"learning_rate": 4.846186262115466e-07,
"loss": 0.7694,
"step": 47000
},
{
"epoch": 30.0,
"step": 47460,
"total_flos": 4.958800183296e+16,
"train_loss": 0.9465416105698796,
"train_runtime": 17544.8246,
"train_samples_per_second": 5.408,
"train_steps_per_second": 2.705
}
],
"max_steps": 47460,
"num_train_epochs": 30,
"total_flos": 4.958800183296e+16,
"trial_name": null,
"trial_params": null
}