whisper-base.en-atco2-asr

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

• Loss: 1.0761
• Wer: 61.4769

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
1.9617	12.5	50	1.6264	47.3310
0.5856	25.0	100	1.2440	43.5498
0.2968	37.5	150	1.1000	55.7829
0.0977	50.0	200	1.0685	44.0391
0.0418	62.5	250	1.0645	51.2456
0.0178	75.0	300	1.0692	41.9929
0.0142	87.5	350	1.0742	61.3434
0.0124	100.0	400	1.0761	61.4769

Framework versions

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