---
title: latin_whisper-large-v3-turbo
canonical_url: "https://www.modelscope.cn/models/AI-ModelScope/latin_whisper-large-v3-turbo"
md_url: "https://www.modelscope.cn/models/AI-ModelScope/latin_whisper-large-v3-turbo.md"
repository: AI-ModelScope/latin_whisper-large-v3-turbo
last_updated: 2026-10-11
license: mit
model_type:
  - whisper
architectures:
  - WhisperForConditionalGeneration
base_model:
  - openai/whisper-large-v3-turbo
base_model_relation: finetune
library_name:
  - transformer
  - safetensors
  - pytorch
frameworks:
  - pytorch
language:
  - la
downloads: 102
stars: 0
tags:
  - whisper
  - latin
  - classical-latin
  - low-resource
---

# latin_whisper-large-v3-turbo

> latin_whisper-large-v3-turbo - AI-ModelScope 在 ModelScope 开源的模型。latinwhisper-large-v3-turbo

AI-ModelScope/latin_whisper-large-v3-turbo 是 ModelScope 魔搭社区上的机器学习模型，采用 mit 许可，基于 openai/whisper-large-v3-turbo 构建。

- **Repository**: AI-ModelScope/latin_whisper-large-v3-turbo
- **License**: mit
- **Base model**: openai/whisper-large-v3-turbo
- **Tags**: whisper, latin, classical-latin, low-resource
- **Downloads**: 102
- **Stars**: 0
- **Last updated**: 2026-10-11

Source: https://www.modelscope.cn/models/AI-ModelScope/latin_whisper-large-v3-turbo

---

# latin_whisper-large-v3-turbo

`openai/whisper-large-v3-turbo` fine-tuned on [Vox Classica](https://huggingface.co/datasets/Ken-Z/Latin-Audio).

## Results

Test set: 2,060 clips from speakers who do not appear in training
(5.6 h). Decoding: beam size 5, `language="latin"`.
Before scoring, both reference and prediction are normalized: diacritics
(macrons) stripped, lowercased, v→u, j→i, digits and punctuation removed.

| Model | WER % | CER % |
|---|---|---|
| **latin_whisper-large-v3-turbo** (this model) | **13.01** | **2.74** |
| `openai/whisper-large-v3-turbo` (no fine-tuning) | 35.55 | 8.03 |

Splits (`splits.json`) are included in this repository.

## Usage

```python
from transformers import pipeline
asr = pipeline("automatic-speech-recognition", model="Ken-Z/latin_whisper-large-v3-turbo")
print(asr("clip.wav", generate_kwargs={"language": "latin", "task": "transcribe"})["text"])
```

For audio longer than 30 seconds, add `chunk_length_s=30` to the `asr(...)` call.

Output is lowercase Latin without macrons or punctuation, written with u and i (no v or j).

## Training

- Base model: `openai/whisper-large-v3-turbo`
- Data: Vox Classica train split (58.2 h), speaker-held-out validation and test
- Hyperparameters: lr 1e-05, 500 warmup steps, batch 16, 3 epochs, optimizer adamw_bnb_8bit, SpecAugment
- Best checkpoint chosen by validation WER

## Limitations

- Trained in restored Classical pronunciation. Expect higher error on Ecclesiastical pronunciation.
- No macrons (vowel length), punctuation or capitalization in the output.

## Data credits

Vox Classica (CC-BY-4.0)