MCD_Speech

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

• Loss: 0.4702
• Wer: 11.0236

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	Wer
0.0	1000.0	1000	0.3492	12.5984
0.0	2000.0	2000	0.4206	10.2362
0.0	3000.0	3000	0.4483	14.1732
0.0	4000.0	4000	0.4702	11.0236

Framework versions

• Transformers 4.39.3
• Pytorch 2.3.0+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2