Views
No views yet
| 항목 | 세부사항 |
|---|---|
| 기반 모델 | meta-llama/Llama-3.2-1B-Instruct |
| 언어 | 한국어 (Korean) |
| 모델 크기 | ~1B 파라미터 |
| 작업 유형 | Speech-to-Text 음성 멀티모달 |
| 라이선스 | Apache 2.0 |
1git clone https://github.com/bigdefence/bigvox
2cd bigvox1pip install --upgrade pip
2conda install pytorch==2.1.2 torchvision==0.16.2 torchaudio==2.1.2 pytorch-cuda=12.1 -c pytorch -c nvidia
3pip install transformers huggingface_hub1pip install accelerate datasets
2pip install flash-attn --no-build-isolation1pip install -U huggingface_hub
2huggingface-cli download bigdefence/bigvox --local-dir ./checkpoints/bigvoxpip install -U huggingface_hub1from huggingface_hub import snapshot_download
2snapshot_download(
3 repo_id="bigdefence/bigvox",
4 local_dir="./checkpoints/bigvox",
5 resume_download=True
6)1git lfs install
2git clone https://huggingface.co/bigdefence/bigvox./models/speech_encoder/ 디렉토리에 배치 🎤python3 omni_speech/infer/bigvox.py --query_audio test_audio.wav