wav2vec2-xlsr-53-ft-btb-cy / trainer_state.json
DewiBrynJones's picture
End of training
a71f6ea verified
raw
history blame
8.2 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.6775106082036775,
"eval_steps": 100,
"global_step": 2600,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.14144271570014144,
"eval_loss": 3.605041265487671,
"eval_runtime": 151.6851,
"eval_samples_per_second": 37.288,
"eval_steps_per_second": 4.661,
"eval_wer": 1.0,
"step": 100
},
{
"epoch": 0.2828854314002829,
"eval_loss": 2.980067491531372,
"eval_runtime": 149.8181,
"eval_samples_per_second": 37.752,
"eval_steps_per_second": 4.719,
"eval_wer": 1.0,
"step": 200
},
{
"epoch": 0.4243281471004243,
"eval_loss": 2.634444236755371,
"eval_runtime": 149.2391,
"eval_samples_per_second": 37.899,
"eval_steps_per_second": 4.737,
"eval_wer": 0.9829243632745422,
"step": 300
},
{
"epoch": 0.5657708628005658,
"eval_loss": 1.1875977516174316,
"eval_runtime": 149.4882,
"eval_samples_per_second": 37.836,
"eval_steps_per_second": 4.729,
"eval_wer": 0.8078348927155719,
"step": 400
},
{
"epoch": 0.7072135785007072,
"grad_norm": 1.7146128416061401,
"learning_rate": 0.00029699999999999996,
"loss": 3.377,
"step": 500
},
{
"epoch": 0.7072135785007072,
"eval_loss": 1.006009578704834,
"eval_runtime": 150.0094,
"eval_samples_per_second": 37.704,
"eval_steps_per_second": 4.713,
"eval_wer": 0.7693826130217779,
"step": 500
},
{
"epoch": 0.8486562942008486,
"eval_loss": 0.8393504619598389,
"eval_runtime": 150.017,
"eval_samples_per_second": 37.702,
"eval_steps_per_second": 4.713,
"eval_wer": 0.6380895828986857,
"step": 600
},
{
"epoch": 0.9900990099009901,
"eval_loss": 0.7753380537033081,
"eval_runtime": 150.3627,
"eval_samples_per_second": 37.616,
"eval_steps_per_second": 4.702,
"eval_wer": 0.5870552550913963,
"step": 700
},
{
"epoch": 1.1315417256011315,
"eval_loss": 0.6410965919494629,
"eval_runtime": 150.6587,
"eval_samples_per_second": 37.542,
"eval_steps_per_second": 4.693,
"eval_wer": 0.4923207780327711,
"step": 800
},
{
"epoch": 1.272984441301273,
"eval_loss": 0.632188618183136,
"eval_runtime": 150.208,
"eval_samples_per_second": 37.654,
"eval_steps_per_second": 4.707,
"eval_wer": 0.5022869156328739,
"step": 900
},
{
"epoch": 1.4144271570014144,
"grad_norm": 0.9841262102127075,
"learning_rate": 0.00022928571428571426,
"loss": 0.8318,
"step": 1000
},
{
"epoch": 1.4144271570014144,
"eval_loss": 0.5857909321784973,
"eval_runtime": 151.3029,
"eval_samples_per_second": 37.382,
"eval_steps_per_second": 4.673,
"eval_wer": 0.4563720691370705,
"step": 1000
},
{
"epoch": 1.5558698727015559,
"eval_loss": 0.5580226182937622,
"eval_runtime": 150.4331,
"eval_samples_per_second": 37.598,
"eval_steps_per_second": 4.7,
"eval_wer": 0.43342267015454733,
"step": 1100
},
{
"epoch": 1.6973125884016973,
"eval_loss": 0.5395969748497009,
"eval_runtime": 152.4347,
"eval_samples_per_second": 37.104,
"eval_steps_per_second": 4.638,
"eval_wer": 0.4204073117106129,
"step": 1200
},
{
"epoch": 1.8387553041018387,
"eval_loss": 0.5091754794120789,
"eval_runtime": 150.7597,
"eval_samples_per_second": 37.517,
"eval_steps_per_second": 4.69,
"eval_wer": 0.4033316749851551,
"step": 1300
},
{
"epoch": 1.9801980198019802,
"eval_loss": 0.4941596984863281,
"eval_runtime": 150.8584,
"eval_samples_per_second": 37.492,
"eval_steps_per_second": 4.687,
"eval_wer": 0.39033236507197766,
"step": 1400
},
{
"epoch": 2.1216407355021216,
"grad_norm": 0.8058004379272461,
"learning_rate": 0.0001582857142857143,
"loss": 0.6486,
"step": 1500
},
{
"epoch": 2.1216407355021216,
"eval_loss": 0.4773481488227844,
"eval_runtime": 151.2881,
"eval_samples_per_second": 37.386,
"eval_steps_per_second": 4.673,
"eval_wer": 0.37783055961226747,
"step": 1500
},
{
"epoch": 2.263083451202263,
"eval_loss": 0.472769170999527,
"eval_runtime": 152.6271,
"eval_samples_per_second": 37.058,
"eval_steps_per_second": 4.632,
"eval_wer": 0.3649756864759031,
"step": 1600
},
{
"epoch": 2.4045261669024045,
"eval_loss": 0.46484073996543884,
"eval_runtime": 150.7659,
"eval_samples_per_second": 37.515,
"eval_steps_per_second": 4.689,
"eval_wer": 0.3623276789009966,
"step": 1700
},
{
"epoch": 2.545968882602546,
"eval_loss": 0.4558813273906708,
"eval_runtime": 151.5484,
"eval_samples_per_second": 37.321,
"eval_steps_per_second": 4.665,
"eval_wer": 0.3528429972236042,
"step": 1800
},
{
"epoch": 2.6874115983026874,
"eval_loss": 0.4480243921279907,
"eval_runtime": 151.9127,
"eval_samples_per_second": 37.232,
"eval_steps_per_second": 4.654,
"eval_wer": 0.3526825119160341,
"step": 1900
},
{
"epoch": 2.828854314002829,
"grad_norm": 0.6075835227966309,
"learning_rate": 8.728571428571428e-05,
"loss": 0.5049,
"step": 2000
},
{
"epoch": 2.828854314002829,
"eval_loss": 0.438320130109787,
"eval_runtime": 151.7706,
"eval_samples_per_second": 37.267,
"eval_steps_per_second": 4.658,
"eval_wer": 0.3383511739500249,
"step": 2000
},
{
"epoch": 2.9702970297029703,
"eval_loss": 0.4345419108867645,
"eval_runtime": 151.6805,
"eval_samples_per_second": 37.289,
"eval_steps_per_second": 4.661,
"eval_wer": 0.33549453547527724,
"step": 2100
},
{
"epoch": 3.1117397454031117,
"eval_loss": 0.4344736337661743,
"eval_runtime": 151.8533,
"eval_samples_per_second": 37.246,
"eval_steps_per_second": 4.656,
"eval_wer": 0.3300219864871371,
"step": 2200
},
{
"epoch": 3.253182461103253,
"eval_loss": 0.42980891466140747,
"eval_runtime": 151.7848,
"eval_samples_per_second": 37.263,
"eval_steps_per_second": 4.658,
"eval_wer": 0.3271813965431465,
"step": 2300
},
{
"epoch": 3.3946251768033946,
"eval_loss": 0.4292232096195221,
"eval_runtime": 151.8596,
"eval_samples_per_second": 37.245,
"eval_steps_per_second": 4.656,
"eval_wer": 0.324565486029754,
"step": 2400
},
{
"epoch": 3.536067892503536,
"grad_norm": 0.6722184419631958,
"learning_rate": 1.614285714285714e-05,
"loss": 0.4131,
"step": 2500
},
{
"epoch": 3.536067892503536,
"eval_loss": 0.42547932267189026,
"eval_runtime": 152.3556,
"eval_samples_per_second": 37.124,
"eval_steps_per_second": 4.64,
"eval_wer": 0.32316926385389416,
"step": 2500
},
{
"epoch": 3.6775106082036775,
"eval_loss": 0.4231884181499481,
"eval_runtime": 151.3029,
"eval_samples_per_second": 37.382,
"eval_steps_per_second": 4.673,
"eval_wer": 0.3216286049012213,
"step": 2600
},
{
"epoch": 3.6775106082036775,
"step": 2600,
"total_flos": 9.931787749443291e+18,
"train_loss": 1.1254758776151217,
"train_runtime": 6709.1009,
"train_samples_per_second": 12.401,
"train_steps_per_second": 0.388
}
],
"logging_steps": 500,
"max_steps": 2600,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 400,
"total_flos": 9.931787749443291e+18,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}