Model reference · open weights
japanese-wav2vec2-large is an open-weight embedding model from reazon-research. japanese-wav2vec2-large (BF16) weighs 633 MB; the smallest configuration that runs it is RTX 3060 12 GB.
For self-hosting, compare this build's requirements below with the GPU you plan to rent.
Summary of the reazon-research/japanese-wav2vec2-large model card, 2026-10-01
What it is
| Released by | reazon-research |
|---|---|
| Released | 2024-11-18 |
| Parameters | 316M |
| VRAM | 633 MB for the weights |
What it runs on
| Card | Runs | Memory |
|---|---|---|
| RTX 3060 12 GB | fits | 11.6 GB |
| RTX 4060 Ti 16 GB | fits | 15.4 GB |
| RTX 3090 24 GB | fits | 23.4 GB |
| RTX 4090 24 GB | fits | 23.4 GB |
| RTX 5090 32 GB | fits | 31.0 GB |
| L40S 48 GB | fits | 44.0 GB |
| A100 80 GB | fits | 78.2 GB |
| H100 80 GB | fits | 78.1 GB |
| RTX PRO 6000 Blackwell 96 GB | fits | 93.8 GB |
| DGX Spark (GB10) 128 GB unified | fits | 107 GB |
| H200 141 GB | fits | 138 GB |
| B200 180 GB | fits | 176 GB |
From the model card
This is a Japanese wav2vec 2.0 Large model pre-trained on ReazonSpeech v2.0 corpus.
We also release the CTC model reazon-research/japanese-wav2vec2-large-rs35kh derived from this model.
import librosa
import torch
from transformers import AutoFeatureExtractor, AutoModel
feature_extractor = AutoFeatureExtractor.from_pretrained("reazon-research/japanese-wav2vec2-large")
model = AutoModel.from_pretrained("reazon-research/japanese-wav2vec2-large")
audio, sr = librosa.load(audio_file, sr=16_000)
inputs = feature_extractor(
audio,
return_tensors="pt",
sampling_rate=sr,
)
with torch.inference_mode():
outputs = model(**inputs)
@misc{reazon-research-japanese-wav2vec2-large,
title={japanese-wav2vec2-large},
author={Sasaki, Yuta},
url = {https://huggingface.co/reazon-research/japanese-wav2vec2-large},
year = {2024}
}
Quoted from the model card on Hugging Face. The full card is behind the Hugging Face link above.