Downloads · 30 days
5
18% of all-time downloads
phowhisper-vi-telephony-lora-encoder
banhchungtuongot/phowhisper-vi-telephony-lora-encoder
phowhisper-vi-telephony-lora-encoder is a automatic speech recognition model from banhchungtuongot. Use it when you need speech turned into text. It is set up for peft. The card lists the license as apache-2.0.
A PEFT LoRA adapter on vinai/PhoWhisper-small that adds robustness to telephony degradation (8 kHz + µ-law + babble + reverb) while keeping PhoWhisper's strong clean Vietnamese accuracy.
All-time downloads
28
Public
Repo size
21.3 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors21.3 MB · 100%
At a glance
- Task
- Automatic Speech Recognition
- Library
- peft
- License
- apache-2.0
- Access
- Public
- Created
- Jun 19, 2026
- Updated
- Jun 19, 2026
- SHA
- c545c377
Base models
Model card
From the Hugging Face model README
PhoWhisper-small VI — Telephony-robust LoRA (encoder-only)
A PEFT LoRA adapter on vinai/PhoWhisper-small that adds robustness to
telephony degradation (8 kHz + µ-law + babble + reverb) while keeping
PhoWhisper's strong clean Vietnamese accuracy.
Lesson-driven: clean WER on whisper-small was at the FLEURS-vi floor (~0.197), so the lever to reduce both clean and telephony was a stronger Vietnamese base. Swapping to PhoWhisper (clean 0.135) and re-applying the proven recipe — encoder-only LoRA + encoder-consistency loss (anchor adapted encoder ≈ frozen base on clean) + held-out-val selection — reduced both.
Results (FLEURS vi test, beam-5)
| Model | Clean WER | Telephony WER | Trainable | Size |
|---|---|---|---|---|
| Whisper Exp C base | ~0.197 | 0.7419 | — | — |
| LoRA on Exp C (prev best) | 0.1972 | 0.5393 | 2.15% | 21 MB |
| PhoWhisper-small base | 0.1351 | 0.6109 | — | — |
| This adapter (PhoWhisper + LoRA) | 0.1324 | 0.4664 | 2.15% | 21 MB |
vs previous best: clean −33%, telephony −14% (telephony −37% vs original Exp C base). Telephony test = FLEURS vi + telephony_sim (ESC-50 held-out cats 40–49, SNR 0–15).
Usage
from transformers import WhisperForConditionalGeneration, WhisperProcessor
from peft import PeftModel
base = WhisperForConditionalGeneration.from_pretrained("vinai/PhoWhisper-small")
model = PeftModel.from_pretrained(base, "banhchungtuongot/phowhisper-vi-telephony-lora-encoder")
processor = WhisperProcessor.from_pretrained("vinai/PhoWhisper-small")
processor.tokenizer.set_prefix_tokens(language="vi", task="transcribe")
# decode with num_beams=5 for the reported numbers
Training code & analysis: https://github.com/NguyenDucAnforwork/asr-h100-noise-robust
(training/finetune_whisper_tele_lora_v2.py, reports/telephony_lora_finetune.md).
Framework versions
- PEFT 0.13.2
- Task
- Automatic Speech Recognition
- Library
- peft
- License
- apache-2.0
- Languages
- vi
- Created
- Jun 19, 2026
- Updated
- Jun 19, 2026