Whisper-base-Arabic / trainer_state.json
YazanSalameh's picture
Upload 9 files
cf0f53d verified
raw
history blame
5.95 kB
{
"best_metric": 34.70818725510414,
"best_model_checkpoint": "/content/drive/MyDrive/whisper-base-ar-checkpoints/cv_mgb2_jordan/checkpoint-7185",
"epoch": 5.0034818941504176,
"eval_steps": 479,
"global_step": 7185,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.33,
"learning_rate": 9.52e-06,
"loss": 0.962,
"step": 479
},
{
"epoch": 0.33,
"eval_loss": 0.627626359462738,
"eval_runtime": 220.8323,
"eval_samples_per_second": 2.78,
"eval_steps_per_second": 0.177,
"eval_wer": 51.04145184574139,
"step": 479
},
{
"epoch": 0.67,
"learning_rate": 9.929457364341086e-06,
"loss": 0.5382,
"step": 958
},
{
"epoch": 0.67,
"eval_loss": 0.5376182794570923,
"eval_runtime": 213.4078,
"eval_samples_per_second": 2.877,
"eval_steps_per_second": 0.183,
"eval_wer": 48.42235512476799,
"step": 958
},
{
"epoch": 1.0,
"learning_rate": 9.855193798449613e-06,
"loss": 0.4603,
"step": 1437
},
{
"epoch": 1.0,
"eval_loss": 0.4931182265281677,
"eval_runtime": 215.5119,
"eval_samples_per_second": 2.849,
"eval_steps_per_second": 0.181,
"eval_wer": 45.88574963910085,
"step": 1437
},
{
"epoch": 1.33,
"learning_rate": 9.78093023255814e-06,
"loss": 0.366,
"step": 1916
},
{
"epoch": 1.33,
"eval_loss": 0.4760296046733856,
"eval_runtime": 220.1923,
"eval_samples_per_second": 2.788,
"eval_steps_per_second": 0.177,
"eval_wer": 40.29696844710249,
"step": 1916
},
{
"epoch": 1.67,
"learning_rate": 9.706666666666668e-06,
"loss": 0.3476,
"step": 2395
},
{
"epoch": 1.67,
"eval_loss": 0.45706605911254883,
"eval_runtime": 221.8166,
"eval_samples_per_second": 2.768,
"eval_steps_per_second": 0.176,
"eval_wer": 42.13239843266653,
"step": 2395
},
{
"epoch": 2.0,
"learning_rate": 9.632713178294575e-06,
"loss": 0.2867,
"step": 2874
},
{
"epoch": 2.0,
"eval_loss": 0.44930991530418396,
"eval_runtime": 216.0554,
"eval_samples_per_second": 2.842,
"eval_steps_per_second": 0.181,
"eval_wer": 36.99731903485255,
"step": 2874
},
{
"epoch": 2.33,
"learning_rate": 9.558449612403102e-06,
"loss": 0.2543,
"step": 3353
},
{
"epoch": 2.33,
"eval_loss": 0.44086378812789917,
"eval_runtime": 216.5946,
"eval_samples_per_second": 2.835,
"eval_steps_per_second": 0.18,
"eval_wer": 37.49226644669004,
"step": 3353
},
{
"epoch": 2.67,
"learning_rate": 9.48418604651163e-06,
"loss": 0.2582,
"step": 3832
},
{
"epoch": 2.67,
"eval_loss": 0.4267967939376831,
"eval_runtime": 222.7442,
"eval_samples_per_second": 2.757,
"eval_steps_per_second": 0.175,
"eval_wer": 37.14167869663848,
"step": 3832
},
{
"epoch": 3.0,
"learning_rate": 9.409922480620157e-06,
"loss": 0.2494,
"step": 4311
},
{
"epoch": 3.0,
"eval_loss": 0.4219926595687866,
"eval_runtime": 220.256,
"eval_samples_per_second": 2.788,
"eval_steps_per_second": 0.177,
"eval_wer": 43.55537224169932,
"step": 4311
},
{
"epoch": 3.34,
"learning_rate": 9.335658914728682e-06,
"loss": 0.1808,
"step": 4790
},
{
"epoch": 3.34,
"eval_loss": 0.44054868817329407,
"eval_runtime": 233.9602,
"eval_samples_per_second": 2.624,
"eval_steps_per_second": 0.167,
"eval_wer": 36.46112600536193,
"step": 4790
},
{
"epoch": 3.67,
"learning_rate": 9.26139534883721e-06,
"loss": 0.1856,
"step": 5269
},
{
"epoch": 3.67,
"eval_loss": 0.43032869696617126,
"eval_runtime": 223.492,
"eval_samples_per_second": 2.747,
"eval_steps_per_second": 0.175,
"eval_wer": 34.74943287275727,
"step": 5269
},
{
"epoch": 4.0,
"learning_rate": 9.187441860465118e-06,
"loss": 0.1435,
"step": 5748
},
{
"epoch": 4.0,
"eval_loss": 0.44085535407066345,
"eval_runtime": 217.814,
"eval_samples_per_second": 2.819,
"eval_steps_per_second": 0.179,
"eval_wer": 40.23510002062281,
"step": 5748
},
{
"epoch": 4.34,
"learning_rate": 9.113178294573645e-06,
"loss": 0.1317,
"step": 6227
},
{
"epoch": 4.34,
"eval_loss": 0.44377487897872925,
"eval_runtime": 209.4289,
"eval_samples_per_second": 2.932,
"eval_steps_per_second": 0.186,
"eval_wer": 39.67828418230563,
"step": 6227
},
{
"epoch": 4.67,
"learning_rate": 9.03891472868217e-06,
"loss": 0.1363,
"step": 6706
},
{
"epoch": 4.67,
"eval_loss": 0.43044084310531616,
"eval_runtime": 212.8255,
"eval_samples_per_second": 2.885,
"eval_steps_per_second": 0.183,
"eval_wer": 36.21365229944318,
"step": 6706
},
{
"epoch": 5.0,
"learning_rate": 8.964806201550388e-06,
"loss": 0.1345,
"step": 7185
},
{
"epoch": 5.0,
"eval_loss": 0.44074973464012146,
"eval_runtime": 204.5774,
"eval_samples_per_second": 3.001,
"eval_steps_per_second": 0.191,
"eval_wer": 34.70818725510414,
"step": 7185
}
],
"logging_steps": 479,
"max_steps": 65000,
"num_input_tokens_seen": 0,
"num_train_epochs": 46,
"save_steps": 479,
"total_flos": 1.4906448617472e+19,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}