wav2vec2-large-medical-speed

This model is a fine-tuned version of facebook/wav2vec2-large on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.4703
• Wer: 0.1992

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: 0.0001
• train_batch_size: 32
• 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: 1000
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
3.9919	3.97	500	3.0791	0.9895
1.6651	7.94	1000	0.5613	0.4259
0.4694	11.9	1500	0.4658	0.2835
0.2812	15.87	2000	0.5245	0.2460
0.1923	19.84	2500	0.4438	0.2300
0.1406	23.81	3000	0.5112	0.2121
0.1082	27.78	3500	0.4703	0.1992

Framework versions

• Transformers 4.34.1
• Pytorch 2.1.0
• Datasets 2.14.6
• Tokenizers 0.14.1