nurzhanit/whisper-enhanced-ml

Whisper Small Hi - Sanchit Gandhi

This model is a fine-tuned version of nurzhanit/whisper-enhanced-ml on the Common Voice 11.0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.0003
• Wer: 22.3549

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: 50
• training_steps: 500

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0035	10.0	50	0.0035	22.4403
0.0088	20.0	100	0.0033	22.3976
0.0022	30.0	150	0.0013	22.3549
0.0007	40.0	200	0.0006	22.3549
0.0005	50.0	250	0.0004	22.3549
0.0004	60.0	300	0.0004	22.3549
0.0003	70.0	350	0.0003	22.3549
0.0003	80.0	400	0.0003	22.3549
0.0003	90.0	450	0.0003	22.3549
0.0003	100.0	500	0.0003	22.3549

Framework versions

• Transformers 4.40.0
• Pytorch 2.5.0+cu124
• Datasets 3.0.2
• Tokenizers 0.19.1