AlinaKozyreva/whisper-small-ru

Whisper Small Ru - Model_ru_3

This model is a fine-tuned version of openai/whisper-small on the Common Voice 13 dataset. It achieves the following results on the evaluation set:

• Loss: 0.2080
• Wer Ortho: 17.4462
• Wer: 13.3014

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
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant_with_warmup
• lr_scheduler_warmup_steps: 50
• training_steps: 4000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer Ortho	Wer
0.2085	0.22	500	0.2366	19.9234	14.9498
0.1875	0.44	1000	0.2176	19.3079	14.5643
0.1688	0.66	1500	0.2095	18.3736	13.9287
0.1678	0.88	2000	0.2038	17.7325	13.4381
0.0853	1.1	2500	0.2036	17.0309	12.7488
0.0822	1.32	3000	0.2046	17.6894	13.2780
0.0775	1.54	3500	0.2051	16.9948	12.7126
0.0727	1.76	4000	0.2080	17.4462	13.3014

Framework versions

• Transformers 4.38.2
• Pytorch 2.3.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2