whisper-base-mixed_v1

This model is a fine-tuned version of carlot/whisper-base-withoutnoise on the audiofolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.2602
• Cer: 12.0846

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: 4000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Cer
0.0003	125.0	1000	0.2491	11.2790
0.0001	250.0	2000	0.2513	11.2790
0.0001	375.0	3000	0.2576	11.7825
0.0001	500.0	4000	0.2602	12.0846

Framework versions

• Transformers 4.36.0.dev0
• Pytorch 2.1.1
• Datasets 2.12.0
• Tokenizers 0.15.0