whisper-medium-reverse-ml-mft-1

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

• Loss: 0.0811
• Wer: 40.3159
• Cer: 12.1262

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: 0.0002
• train_batch_size: 64
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 128
• optimizer: Use adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 593
• training_steps: 5928

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
0.0582	1.0	1482	0.0768	43.4159	13.4134
0.0399	2.0	2964	0.0669	41.7880	12.4496
0.026	3.0	4446	0.0715	39.8821	11.8874
0.0169	4.0	5928	0.0811	40.3159	12.1262

Framework versions

• Transformers 4.51.1
• Pytorch 2.8.0+cu128
• Datasets 4.4.1
• Tokenizers 0.21.4