---
library_name: transformers
license: apache-2.0
base_model: wwwtwwwt/whisper-small-transcription
tags:
- generated_from_trainer
datasets:
- covost2
metrics:
- bleu
model-index:
- name: whisper-tiny-translation
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: covost2
      type: covost2
      config: zh-CN_en
      split: None
      args: zh-CN_en
    metrics:
    - name: Bleu
      type: bleu
      value: 8.913626875649962
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# whisper-tiny-translation

This model is a fine-tuned version of [wwwtwwwt/whisper-small-transcription](https://huggingface.co/wwwtwwwt/whisper-small-transcription) on the covost2 dataset.
It achieves the following results on the evaluation set:
- Loss: 2.1289
- Bleu: 8.9136

## 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: 4
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 2
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Bleu   |
|:-------------:|:------:|:----:|:---------------:|:------:|
| 1.4897        | 1.3407 | 1000 | 2.1289          | 8.9136 |


### Framework versions

- Transformers 4.47.0
- Pytorch 2.5.1+cu121
- Datasets 3.3.1
- Tokenizers 0.21.0