---
language:
- ja
license: apache-2.0
base_model: openai/whisper-small
tags:
- testing
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_16_1
metrics:
- wer
model-index:
- name: Whisper Small JA Test
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: Common Voice 16.1
      type: mozilla-foundation/common_voice_16_1
      config: ja
      split: None
      args: 'config: ja, split: test'
    metrics:
    - name: Wer
      type: wer
      value: 83.60284605433377
---

<!-- 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 Small JA Test

This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the Common Voice 16.1 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.5351
- Wer: 83.6028

## 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: Adam with betas=(0.9,0.999) and epsilon=1e-08
- 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     |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
| 0.2417        | 1.02  | 1000 | 0.5194          | 85.4463 |
| 0.107         | 2.04  | 2000 | 0.5127          | 83.2471 |
| 0.0515        | 3.05  | 3000 | 0.5249          | 83.8616 |
| 0.0306        | 4.07  | 4000 | 0.5351          | 83.6028 |


### Framework versions

- Transformers 4.39.0.dev0
- Pytorch 2.1.0+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2