Whisper-Timit-fineT-16

This model is a fine-tuned version of openai/whisper-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1388
• Wer: 38.9999

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 1e-05
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 5000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0356	1.73	500	0.0982	61.5030
0.0022	3.46	1000	0.1059	80.2039
0.0018	5.19	1500	0.1167	47.5479
0.0002	6.92	2000	0.1204	49.5247
0.0002	8.65	2500	0.1280	51.4465
0.0001	10.38	3000	0.1316	44.9029
0.0001	12.11	3500	0.1345	42.7538
0.0001	13.84	4000	0.1368	40.0744
0.0001	15.57	4500	0.1382	40.0813
0.0001	17.3	5000	0.1388	38.9999

Framework versions

• Transformers 4.28.0.dev0
• Pytorch 1.13.0
• Datasets 2.1.0
• Tokenizers 0.13.2