whisper-small-hi

This model is a fine-tuned version of openai/whisper-small on the mozilla-foundation/common_voice_11_0 hi dataset. It achieves the following results on the evaluation set:

• Loss: 0.5355
• Wer: 0.3309

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: 16
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 5000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0534	4.89	1000	0.3375	0.3465
0.0042	9.78	2000	0.4443	0.3402
0.0002	14.67	3000	0.4973	0.3301
0.0001	19.56	4000	0.5254	0.3309
0.0001	24.45	5000	0.5355	0.3309

Framework versions

• Transformers 4.28.0.dev0
• Pytorch 2.0.0+cu117
• Datasets 2.11.0
• Tokenizers 0.13.3