Downloads · 30 days
481
24% of all-time downloads
JackyHoCL/whisper-large-v3-turbo-cantonese
whisper-large-v3-turbo-cantonese is a automatic speech recognition model from JackyHoCL. Use it when you need speech turned into text. It is set up for transformers.
This repository is using dedicated language code for different language (whisper's default behavior), which is a different to JackyHoCL/whisper-large-v3-turbo-cantonese-yue-english
Downloads · 30 days
481
24% of all-time downloads
All-time downloads
2K
Public
Parameters
809M
1.6 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors1.6 GB · 100%
From the Hugging Face model README
This repository is using dedicated language code for different language (whisper's default behavior), which is a different to JackyHoCL/whisper-large-v3-turbo-cantonese-yue-english
2025-10-10: CER:
| Dataset | Evaluation Lang | Dataset Lang | Split | CER(in %) |
|---|---|---|---|---|
| Training | Mixed | Mixed | validation | 5.839 |
| mozilla-foundation/common_voice_22_0 | yue | yue | test | 2.383 |
| mozilla-foundation/common_voice_17_0 | yue | yue | test | 2.134 |
| mozilla-foundation/common_voice_17_0 | en | en | test(2k samples) | 5.68 |
| mozilla-foundation/common_voice_16_1 | zh | zh-CN | test | 6.99 |
| JackyHoCL/cleaned_mixed_cantonese_and_english_speech | yue | yue | test | 5.6 |