metadata
library_name: transformers
language:
- en
license: apache-2.0
base_model: openai/whisper-small
tags:
- generated_from_trainer
datasets:
- darynka-xo/whisper_finetune_dataset
metrics:
- wer
model-index:
- name: Whisper Small En - AIFC
results:
- task:
type: automatic-speech-recognition
name: Automatic Speech Recognition
dataset:
name: whisper_finetune_dataset
type: darynka-xo/whisper_finetune_dataset
config: default
split: None
args: 'config: en, split: test'
metrics:
- type: wer
value: 1.30760286407866
name: Wer
Whisper Small En - AIFC
This model is a fine-tuned version of openai/whisper-small on the whisper_finetune_dataset dataset. It achieves the following results on the evaluation set:
- Loss: 0.0397
- Wer: 1.3076
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: 8
- seed: 42
- 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: 500
- training_steps: 4000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.0043 | 1.1186 | 1000 | 0.0453 | 1.6911 |
| 0.0028 | 2.2371 | 2000 | 0.0396 | 1.4198 |
| 0.0004 | 3.3557 | 3000 | 0.0394 | 1.3464 |
| 0.0003 | 4.4743 | 4000 | 0.0397 | 1.3076 |
Framework versions
- Transformers 4.49.0
- Pytorch 2.6.0+cu124
- Datasets 3.4.1
- Tokenizers 0.21.1