Downloads · 30 days
50
30% of all-time downloads
PhoWhisper-large-ct2
vudang449/PhoWhisper-large-ct2
PhoWhisper-large-ct2 is a automatic speech recognition model from vudang449. Use it when you need speech turned into text. The card lists the license as bsd-3-clause.
Bản CTranslate2 (int8-ready) của vinai/PhoWhisper-large — ASR tiếng Việt. Mirror từ kiendt/PhoWhisper-large-ct2 để phục vụ voice-bench (harness benchmark STT/TTS tiếng Việt).
All-time downloads
167
Public
Repo size
3.1 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin3.1 GB · 100%
At a glance
- Task
- Automatic Speech Recognition
- License
- bsd-3-clause
- Access
- Public
- Created
- Jul 5, 2026
- Updated
- Jul 5, 2026
- SHA
- ed85d1b9
Base models
Model card
From the Hugging Face model README
PhoWhisper-large-ct2
Bản CTranslate2 (int8-ready) của vinai/PhoWhisper-large — ASR tiếng Việt. Mirror từ kiendt/PhoWhisper-large-ct2 để phục vụ voice-bench (harness benchmark STT/TTS tiếng Việt).
Model gốc: VinAI Research, license BSD-3-Clause. Mirror này giữ nguyên trọng số, không fine-tune thêm.
Dùng với faster-whisper
from faster_whisper import WhisperModel
model = WhisperModel("vudang449/PhoWhisper-large-ct2", device="cpu", compute_type="int8")
segments, info = model.transcribe("audio.wav", language="vi", beam_size=5)
print(" ".join(s.text for s in segments))
Số đo (voice-bench, Mac Mini M4 24GB, CPU int8, eval VIVOS-test 50 clips + long-form 3:16)
| Chế độ | WER | Latency | Params |
|---|---|---|---|
| Per-utterance (clip ngắn) | 5.50% | p50 7.65s / p90 8.12s | mặc định (beam 5, không VAD) |
| Long-form 195.6s (1 request) | 5.50% | RTF 0.693 (135.6s) | beam 5 + VAD 300/400ms + condition_on_previous_text=False |
Accuracy ceiling trên M4 CPU, nhưng chậm ~5× bản small. Gợi ý: dùng profile theo độ dài — clip ngắn giữ params mặc định, file ≥30s bật VAD + tắt condition (xem configs/service.large.yaml trong voice-bench). Cần nhanh cho interactive: vudang449/PhoWhisper-small-ct2 — WER 6.24%, p50 1.51s.
- Task
- Automatic Speech Recognition
- License
- bsd-3-clause
- Languages
- vi
- Created
- Jul 5, 2026
- Updated
- Jul 5, 2026