tanmaylaud's picture
updated model weights
0a1bc74
{
"best_metric": 0.24131617493199414,
"best_model_checkpoint": "/workspace/output_models/hi-mr/wav2vec2-large-xlsr-hindi-marathi-3/checkpoint-2800",
"epoch": 4.444444444444445,
"global_step": 2800,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.16,
"learning_rate": 6.666666666666667e-06,
"loss": 1.1154,
"step": 100
},
{
"epoch": 0.32,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.252,
"step": 200
},
{
"epoch": 0.32,
"eval_loss": 0.49659740924835205,
"eval_runtime": 153.7026,
"eval_samples_per_second": 26.851,
"eval_wer": 0.27688847039129527,
"step": 200
},
{
"epoch": 0.48,
"learning_rate": 1.9999999999999998e-05,
"loss": 0.1238,
"step": 300
},
{
"epoch": 0.63,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.0989,
"step": 400
},
{
"epoch": 0.63,
"eval_loss": 0.49464675784111023,
"eval_runtime": 152.2596,
"eval_samples_per_second": 27.105,
"eval_wer": 0.25449884913161747,
"step": 400
},
{
"epoch": 0.79,
"learning_rate": 2.9743589743589744e-05,
"loss": 0.0872,
"step": 500
},
{
"epoch": 0.95,
"learning_rate": 2.923076923076923e-05,
"loss": 0.0882,
"step": 600
},
{
"epoch": 0.95,
"eval_loss": 0.48914414644241333,
"eval_runtime": 153.1646,
"eval_samples_per_second": 26.945,
"eval_wer": 0.24939840970914418,
"step": 600
},
{
"epoch": 1.11,
"learning_rate": 2.871794871794872e-05,
"loss": 0.0801,
"step": 700
},
{
"epoch": 1.27,
"learning_rate": 2.8205128205128207e-05,
"loss": 0.0769,
"step": 800
},
{
"epoch": 1.27,
"eval_loss": 0.4904399812221527,
"eval_runtime": 151.6576,
"eval_samples_per_second": 27.213,
"eval_wer": 0.24746285833856455,
"step": 800
},
{
"epoch": 1.43,
"learning_rate": 2.7692307692307694e-05,
"loss": 0.0755,
"step": 900
},
{
"epoch": 1.59,
"learning_rate": 2.717948717948718e-05,
"loss": 0.0682,
"step": 1000
},
{
"epoch": 1.59,
"eval_loss": 0.49137312173843384,
"eval_runtime": 150.2065,
"eval_samples_per_second": 27.476,
"eval_wer": 0.24620736555764805,
"step": 1000
},
{
"epoch": 1.75,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.0664,
"step": 1100
},
{
"epoch": 1.9,
"learning_rate": 2.6153846153846157e-05,
"loss": 0.0669,
"step": 1200
},
{
"epoch": 1.9,
"eval_loss": 0.4783032238483429,
"eval_runtime": 152.2438,
"eval_samples_per_second": 27.108,
"eval_wer": 0.24448106298388783,
"step": 1200
},
{
"epoch": 2.06,
"learning_rate": 2.564102564102564e-05,
"loss": 0.0681,
"step": 1300
},
{
"epoch": 2.22,
"learning_rate": 2.512820512820513e-05,
"loss": 0.062,
"step": 1400
},
{
"epoch": 2.22,
"eval_loss": 0.492949515581131,
"eval_runtime": 151.7551,
"eval_samples_per_second": 27.195,
"eval_wer": 0.24312094580456162,
"step": 1400
},
{
"epoch": 2.38,
"learning_rate": 2.4615384615384616e-05,
"loss": 0.0603,
"step": 1500
},
{
"epoch": 2.54,
"learning_rate": 2.4102564102564103e-05,
"loss": 0.0627,
"step": 1600
},
{
"epoch": 2.54,
"eval_loss": 0.48576003313064575,
"eval_runtime": 151.5396,
"eval_samples_per_second": 27.234,
"eval_wer": 0.24445490688428542,
"step": 1600
},
{
"epoch": 2.7,
"learning_rate": 2.358974358974359e-05,
"loss": 0.0601,
"step": 1700
},
{
"epoch": 2.86,
"learning_rate": 2.307692307692308e-05,
"loss": 0.0569,
"step": 1800
},
{
"epoch": 2.86,
"eval_loss": 0.4850601851940155,
"eval_runtime": 152.3874,
"eval_samples_per_second": 27.082,
"eval_wer": 0.24267629211132036,
"step": 1800
},
{
"epoch": 3.02,
"learning_rate": 2.2564102564102566e-05,
"loss": 0.0526,
"step": 1900
},
{
"epoch": 3.17,
"learning_rate": 2.2051282051282052e-05,
"loss": 0.0439,
"step": 2000
},
{
"epoch": 3.17,
"eval_loss": 0.47941696643829346,
"eval_runtime": 152.2838,
"eval_samples_per_second": 27.101,
"eval_wer": 0.2431994141033689,
"step": 2000
},
{
"epoch": 3.33,
"learning_rate": 2.153846153846154e-05,
"loss": 0.0486,
"step": 2100
},
{
"epoch": 3.49,
"learning_rate": 2.1025641025641025e-05,
"loss": 0.0437,
"step": 2200
},
{
"epoch": 3.49,
"eval_loss": 0.481067955493927,
"eval_runtime": 151.9904,
"eval_samples_per_second": 27.153,
"eval_wer": 0.2431994141033689,
"step": 2200
},
{
"epoch": 3.65,
"learning_rate": 2.0512820512820515e-05,
"loss": 0.0384,
"step": 2300
},
{
"epoch": 3.81,
"learning_rate": 1.9999999999999998e-05,
"loss": 0.0415,
"step": 2400
},
{
"epoch": 3.81,
"eval_loss": 0.4836219847202301,
"eval_runtime": 152.3729,
"eval_samples_per_second": 27.085,
"eval_wer": 0.24212701401966938,
"step": 2400
},
{
"epoch": 3.97,
"learning_rate": 1.9487179487179488e-05,
"loss": 0.0398,
"step": 2500
},
{
"epoch": 4.13,
"learning_rate": 1.8974358974358975e-05,
"loss": 0.0399,
"step": 2600
},
{
"epoch": 4.13,
"eval_loss": 0.48345065116882324,
"eval_runtime": 151.6188,
"eval_samples_per_second": 27.22,
"eval_wer": 0.24170851642603056,
"step": 2600
},
{
"epoch": 4.29,
"learning_rate": 1.8461538461538465e-05,
"loss": 0.0367,
"step": 2700
},
{
"epoch": 4.44,
"learning_rate": 1.7948717948717948e-05,
"loss": 0.0361,
"step": 2800
},
{
"epoch": 4.44,
"eval_loss": 0.4902171790599823,
"eval_runtime": 152.2588,
"eval_samples_per_second": 27.105,
"eval_wer": 0.24131617493199414,
"step": 2800
}
],
"max_steps": 6300,
"num_train_epochs": 10,
"total_flos": 2.1761689418766148e+19,
"trial_name": null,
"trial_params": null
}