lite-whisper-large-v3-turbo-acc

Model Card for Lite-Whisper large-v3-turbo-acc

<!-- Provide a quick summary of what the model is/does. -->

Lite-Whisper is a compressed version of OpenAI Whisper with LiteASR. See our GitHub repository and paper for details.

Benchmark Results

Following is the average word error rate (WER) evaluated on the ESB datasets:

Quick Start

The easiest way to run our model is to use our integration with HuggingFace Transformers library. We provide model weights for the compressed version of OpenAI Whisper series here.

import librosa 
import torch
from transformers import AutoProcessor, AutoModel

device = "cuda:0"
dtype = torch.float16

# load the compressed Whisper model
model = AutoModel.from_pretrained(
    "efficient-speech/lite-whisper-large-v3-turbo", 
    trust_remote_code=True, 
)
model.to(dtype).to(device)

# we use the same processor as the original model
processor = AutoProcessor.from_pretrained("openai/whisper-large-v3")

# set the path to your audio file
path = "path/to/audio.wav"
audio, _ = librosa.load(path, sr=16000)

input_features = processor(audio, sampling_rate=16000, return_tensors="pt").input_features
input_features = input_features.to(dtype).to(device)

predicted_ids = model.generate(input_features)
transcription = processor.batch_decode(
    predicted_ids, 
    skip_special_tokens=True
)[0]

print(transcription)

Citation

If you use LiteASR in your research, please cite the following paper:

@misc{kamahori2025liteasrefficientautomaticspeech,
      title={LiteASR: Efficient Automatic Speech Recognition with Low-Rank Approximation}, 
      author={Keisuke Kamahori and Jungo Kasai and Noriyuki Kojima and Baris Kasikci},
      year={2025},
      eprint={2502.20583},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2502.20583}, 
}