Fine_tune_whisper_small

This model is a fine-tuned version of openai/whisper-small on our own recorded dataset (700 audio samples). It achieves the following results on the evaluation set:

• Loss: 0.8225
• Wer: 43.7477

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: 200
• training_steps: 900
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.2991	3.92	200	0.6605	44.1903
0.0185	7.84	400	0.7377	42.8624
0.0026	11.76	600	0.8087	43.0837
0.0011	15.69	800	0.8225	43.7477

Framework versions

• Transformers 4.25.0.dev0
• Pytorch 1.12.1+cu113
• Datasets 2.7.1
• Tokenizers 0.13.2