whisper-finetuned-model

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

• Loss: 1.4620
• Wer Kik: 0.704
• Wer Mean: 0.704
• Cer Kik: 0.371
• Cer Mean: 0.371

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: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 50
• training_steps: 2000

Training results

Training Loss	Epoch	Step	Validation Loss	Wer Kik	Wer Mean	Cer Kik	Cer Mean
2.8621	4.5455	50	2.1788	0.993	0.993	0.517	0.517
1.3099	9.0909	100	1.3917	0.761	0.761	0.388	0.388
0.8879	13.6364	150	1.2341	0.722	0.722	0.364	0.364
0.5424	18.1818	200	1.1984	0.705	0.705	0.359	0.359
0.3523	22.7273	250	1.2556	0.697	0.697	0.355	0.355
0.2287	27.2727	300	1.3024	0.691	0.691	0.355	0.355
0.1315	31.8182	350	1.3603	0.681	0.681	0.354	0.354
0.0808	36.3636	400	1.4313	0.71	0.71	0.375	0.375
0.0492	40.9091	450	1.4620	0.704	0.704	0.371	0.371

Framework versions

• Transformers 4.57.3
• Pytorch 2.9.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.2