Configuration Parsing
Warning:
In adapter_config.json: "peft.task_type" must be a string
whisper-model-small-ro-finetune-5k-25-05
This model is a fine-tuned version of openai/whisper-small on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.0945
- Wer: 0.6525
- Cer: 0.4544
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: 2e-05
- train_batch_size: 32
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 64
- optimizer: Use OptimizerNames.ADAMW_TORCH 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: 30
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|---|---|---|---|---|---|
| 3.6234 | 1.0 | 82 | 1.7284 | 0.9467 | 0.6348 |
| 1.4871 | 2.0 | 164 | 1.2036 | 1.0478 | 1.0803 |
| 1.3473 | 3.0 | 246 | 1.1344 | 0.7948 | 0.7811 |
| 1.2769 | 4.0 | 328 | 1.1035 | 0.7285 | 0.5542 |
| 1.2484 | 5.0 | 410 | 1.0945 | 0.6525 | 0.4544 |
Framework versions
- PEFT 0.15.2
- Transformers 4.52.3
- Pytorch 2.6.0+cu124
- Datasets 3.6.0
- Tokenizers 0.21.1
- Downloads last month
- 1
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support
Model tree for victors3136/whisper-model-small-ro-finetune-5k-25-05
Base model
openai/whisper-small