Views
No views yet
reazon-research/japanese-hubert-base-k2reazon-research/japanese-hubert-base-k2-rs35kh and reazon-research/japanese-hubert-base-k2-rs35kh-bpe, derived from this model.1import librosa
2import torch
3from transformers import AutoFeatureExtractor, AutoModel
4
5feature_extractor = AutoFeatureExtractor.from_pretrained("reazon-research/japanese-hubert-base-k2")
6model = AutoModel.from_pretrained("reazon-research/japanese-hubert-base-k2")
7
8audio, sr = librosa.load(audio_file, sr=16_000)
9inputs = feature_extractor(
10 audio,
11 return_tensors="pt",
12 sampling_rate=sr,
13)
14with torch.inference_mode():
15 outputs = model(**inputs)1@misc{reazon-research-japanese-hubert-base-k2,
2 title={japanese-hubert-base-k2},
3 author={Sasaki, Yuta},
4 url = {https://huggingface.co/reazon-research/japanese-hubert-base-k2},
5 year = {2025}
6}
7
8@article{yang2024k2ssl,
9 title={k2SSL: A faster and better framework for self-supervised speech representation learning},
10 author={Yang, Yifan and Zhuo, Jianheng and Jin, Zengrui and Ma, Ziyang and Yang, Xiaoyu and Yao, Zengwei and Guo, Liyong and Kang, Wei and Kuang, Fangjun and Lin, Long and others},
11 journal={arXiv preprint arXiv:2411.17100},
12 year={2024}
13}