Swapnajeet-21/whisper-small-bn-cv11

Whisper Tiny Bengali

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

• Loss: 0.2100
• Wer: 66.8016

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: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant_with_warmup
• lr_scheduler_warmup_steps: 500
• training_steps: 5000

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.1084	0.5959	500	1.0912	205.3312
0.3323	1.1919	1000	0.3413	93.5779
0.2757	1.7878	1500	0.2803	84.8822
0.2238	2.3838	2000	0.2543	78.7731
0.202	2.9797	2500	0.2323	77.2269
0.172	3.5757	3000	0.2224	71.1526
0.1384	4.1716	3500	0.2171	73.7138
0.14	4.7676	4000	0.2103	74.8553
0.1273	5.3635	4500	0.2117	65.3154
0.1148	5.9595	5000	0.2100	66.8016

Framework versions

• Transformers 4.51.3
• Pytorch 2.5.1+cu118
• Datasets 3.5.0
• Tokenizers 0.21.1