whisper-tiny.en-atco2-asr

This model is a fine-tuned version of openai/whisper-tiny.en on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.1500
• Wer: 61.5214

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: 128
• eval_batch_size: 128
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 10
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
2.3572	12.5	50	1.9974	100.8897
0.9286	25.0	100	1.5447	246.6637
0.5731	37.5	150	1.3265	233.9413
0.2702	50.0	200	1.2124	65.6584
0.1624	62.5	250	1.1642	62.5
0.0828	75.0	300	1.1518	71.1744
0.0596	87.5	350	1.1501	61.8772
0.0479	100.0	400	1.1500	61.5214

Framework versions

• Transformers 4.36.2
• Pytorch 2.1.2
• Datasets 2.15.0
• Tokenizers 0.15.0