whisper-small-atco2-asr

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

• Loss: 0.8297
• Wer: 33.4520

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
0.6955	12.5	50	0.8441	57.4288
0.0631	25.0	100	0.7271	27.8470
0.0034	37.5	150	0.7708	36.8772
0.0008	50.0	200	0.7991	31.8060
0.0006	62.5	250	0.8141	32.2509
0.0004	75.0	300	0.8228	32.8292
0.0004	87.5	350	0.8281	33.2295
0.0004	100.0	400	0.8297	33.4520

Framework versions

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