File size: 1,387 Bytes
f77a1a5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
---
library_name: transformers
pipeline_tag: automatic-speech-recognition
license: apache-2.0
language:
- vi
---
Converted to Flax from [vinai/PhoWhisper-tiny](https://huggingface.co/vinai/PhoWhisper-tiny) for light-speed inference on TPU
```py
from whisper_jax import FlaxWhisperPipline
import jax.numpy as jnp
pipeline = FlaxWhisperPipline("hr16/PhoWhisper-tiny-flax", dtype=jnp.bfloat16, batch_size=16)
```

# PhoWhisper:  Automatic  Speech  Recognition for  Vietnamese


We introduce **PhoWhisper** in five versions for Vietnamese automatic speech recognition. PhoWhisper's robustness is achieved through fine-tuning the multilingual [Whisper](https://github.com/openai/whisper)  on an 844-hour dataset that encompasses diverse Vietnamese accents. Our experimental study demonstrates state-of-the-art performances of PhoWhisper on benchmark Vietnamese ASR datasets. Please **cite** our PhoWhisper paper when it is used to help produce published results or is incorporated into other software:

```
@inproceedings{PhoWhisper,
  title     = {{PhoWhisper: Automatic Speech Recognition for Vietnamese}},
  author    = {Thanh-Thien Le and Linh The Nguyen and Dat Quoc Nguyen},
  booktitle = {Proceedings of the ICLR 2024 Tiny Papers track},
  year      = {2024}
}
```

For further information or requests, please go to [PhoWhisper's homepage](https://github.com/VinAIResearch/PhoWhisper)!