marinone94's picture
Training in progress, step 3000
9eafbd0
raw
history blame
4.6 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 31.24935064935065,
"global_step": 3000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.04,
"learning_rate": 3.828125e-06,
"loss": 15.104,
"step": 100
},
{
"epoch": 2.08,
"learning_rate": 7.6953125e-06,
"loss": 12.2149,
"step": 200
},
{
"epoch": 3.12,
"learning_rate": 1.16015625e-05,
"loss": 4.9885,
"step": 300
},
{
"epoch": 4.17,
"learning_rate": 1.55078125e-05,
"loss": 3.3788,
"step": 400
},
{
"epoch": 5.21,
"learning_rate": 1.94140625e-05,
"loss": 3.1705,
"step": 500
},
{
"epoch": 6.25,
"learning_rate": 2.3320312499999995e-05,
"loss": 3.11,
"step": 600
},
{
"epoch": 7.29,
"learning_rate": 2.72265625e-05,
"loss": 3.0526,
"step": 700
},
{
"epoch": 8.33,
"learning_rate": 3.11328125e-05,
"loss": 3.0113,
"step": 800
},
{
"epoch": 9.37,
"learning_rate": 3.5039062499999995e-05,
"loss": 2.9601,
"step": 900
},
{
"epoch": 10.42,
"learning_rate": 3.89453125e-05,
"loss": 2.9099,
"step": 1000
},
{
"epoch": 10.42,
"eval_loss": 2.8369038105010986,
"eval_runtime": 198.5866,
"eval_samples_per_second": 25.405,
"eval_steps_per_second": 0.796,
"eval_wer": 1.0,
"step": 1000
},
{
"epoch": 11.46,
"learning_rate": 4.28515625e-05,
"loss": 2.6653,
"step": 1100
},
{
"epoch": 12.5,
"learning_rate": 4.675781249999999e-05,
"loss": 1.8958,
"step": 1200
},
{
"epoch": 13.54,
"learning_rate": 5.0664062499999996e-05,
"loss": 1.4812,
"step": 1300
},
{
"epoch": 14.58,
"learning_rate": 5.45703125e-05,
"loss": 1.3358,
"step": 1400
},
{
"epoch": 15.62,
"learning_rate": 5.8476562499999996e-05,
"loss": 1.2522,
"step": 1500
},
{
"epoch": 16.66,
"learning_rate": 6.238281249999999e-05,
"loss": 1.1919,
"step": 1600
},
{
"epoch": 17.71,
"learning_rate": 6.62890625e-05,
"loss": 1.1663,
"step": 1700
},
{
"epoch": 18.75,
"learning_rate": 7.019531249999999e-05,
"loss": 1.1289,
"step": 1800
},
{
"epoch": 19.79,
"learning_rate": 7.410156249999999e-05,
"loss": 1.0902,
"step": 1900
},
{
"epoch": 20.83,
"learning_rate": 7.4248046875e-05,
"loss": 1.0745,
"step": 2000
},
{
"epoch": 20.83,
"eval_loss": 0.19572903215885162,
"eval_runtime": 202.06,
"eval_samples_per_second": 24.968,
"eval_steps_per_second": 0.782,
"eval_wer": 0.16725371193237237,
"step": 2000
},
{
"epoch": 21.87,
"learning_rate": 7.327148437499999e-05,
"loss": 1.0485,
"step": 2100
},
{
"epoch": 22.91,
"learning_rate": 7.2294921875e-05,
"loss": 1.0291,
"step": 2200
},
{
"epoch": 23.96,
"learning_rate": 7.1318359375e-05,
"loss": 1.007,
"step": 2300
},
{
"epoch": 25.0,
"learning_rate": 7.0341796875e-05,
"loss": 1.0008,
"step": 2400
},
{
"epoch": 26.04,
"learning_rate": 6.9365234375e-05,
"loss": 0.988,
"step": 2500
},
{
"epoch": 27.08,
"learning_rate": 6.8388671875e-05,
"loss": 0.9766,
"step": 2600
},
{
"epoch": 28.12,
"learning_rate": 6.7412109375e-05,
"loss": 0.9663,
"step": 2700
},
{
"epoch": 29.17,
"learning_rate": 6.6435546875e-05,
"loss": 0.9539,
"step": 2800
},
{
"epoch": 30.21,
"learning_rate": 6.545898437499999e-05,
"loss": 0.9479,
"step": 2900
},
{
"epoch": 31.25,
"learning_rate": 6.448242187499999e-05,
"loss": 0.934,
"step": 3000
},
{
"epoch": 31.25,
"eval_loss": 0.1579357087612152,
"eval_runtime": 198.7733,
"eval_samples_per_second": 25.381,
"eval_steps_per_second": 0.795,
"eval_wer": 0.1388858784003468,
"step": 3000
}
],
"max_steps": 9600,
"num_train_epochs": 100,
"total_flos": 4.660374091191037e+19,
"trial_name": null,
"trial_params": null
}