whisper-small-norm

This model is a fine-tuned version of openai/whisper-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0000
• Wer: 0.8511

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: 0.0004
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 132
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.2297	3.6364	100	0.5810	27.4468
0.4636	7.2727	200	0.3247	22.7660
0.1455	10.9091	300	0.1490	103.1915
0.0719	14.5455	400	0.0637	15.3191
0.0458	18.1818	500	0.0133	5.1064
0.0084	21.8182	600	0.0003	2.5532
0.0006	25.4545	700	0.0000	0.8511
0.0	29.0909	800	0.0000	0.8511

Framework versions

• Transformers 4.45.0.dev0
• Pytorch 2.4.0
• Datasets 2.21.0
• Tokenizers 0.19.1