---
license: apache-2.0
language:
- ja
metrics:
- cer
- wer
base_model:
- openai/whisper-medium
tags:
- ctranslate2
- faster-whisper
- whisper
model-index:
- name: whisper-medium-jp-ct2
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: mozilla-foundation/common_voice_17_0 (ja)
      type: mozilla-foundation/common_voice_17_0
      config: ja
      split: test
      args:
        language: ja
    metrics:
    - name: CER
      type: cer
      value: 0.18572446886192148
---

> **This repository contains the CTranslate2 export of the fine-tuned model.**
>
> • Base Transformers model: [drepic/whisper-medium-jp](https://huggingface.co/drepic/whisper-medium-jp)  
> • Use with `faster-whisper`:
>
> ```python
> from faster_whisper import WhisperModel
> model = WhisperModel("drepic/whisper-medium-jp-ct2", device="cuda", compute_type="float16")
> ```

<!-- 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. -->

# OTHER FINETUNES
- Want something more lightweight? Try [drepic/whisper-small-jp-ct2](https://huggingface.co/drepic/whisper-small-jp-ct2)

# whisper-medium-jp

This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on an Japanese youtube based dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4828
- Wer: 0.2254
- Cer: 0.2254

## Model description

Better suited for transcribing japanese youtube content. 

## 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: 4e-06
- train_batch_size: 4
- eval_batch_size: 2
- seed: 42
- distributed_type: multi-GPU
- num_devices: 2
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- total_eval_batch_size: 4
- 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: 400
- num_epochs: 15
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step  | Validation Loss | Wer    | Cer    |
|:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|
| 0.5341        | 1.0   | 7155  | 0.5321          | 0.2416 | 0.2416 |
| 0.5023        | 2.0   | 14310 | 0.5143          | 0.2369 | 0.2369 |
| 0.499         | 3.0   | 21465 | 0.5063          | 0.2337 | 0.2337 |
| 0.4773        | 4.0   | 28620 | 0.5010          | 0.2310 | 0.2310 |
| 0.4775        | 5.0   | 35775 | 0.4944          | 0.2289 | 0.2289 |
| 0.4709        | 6.0   | 42930 | 0.4886          | 0.2288 | 0.2288 |
| 0.4907        | 7.0   | 50085 | 0.4870          | 0.2271 | 0.2271 |
| 0.4855        | 8.0   | 57240 | 0.4868          | 0.2261 | 0.2261 |
| 0.4487        | 9.0   | 64395 | 0.4828          | 0.2254 | 0.2254 |


### Framework versions

- Transformers 4.56.1
- Pytorch 2.8.0+cu128
- Datasets 4.0.0
- Tokenizers 0.22.0