PyPI page
Home page
Author:
Sewon Kim
Summary:
Up to 3.4x faster OmniVoice inference through Triton kernel fusion and CUDA Graph optimization
Latest version:
0.1.0
Required dependencies:
accelerate
|
huggingface-hub
|
numpy
|
omnivoice
|
sageattention
|
soundfile
|
transformers
|
triton
Optional dependencies:
jiwer
|
librosa
|
openai-whisper
|
plotly
|
pre-commit
|
pynvml
|
pytest
|
pytest-cov
|
resemblyzer
|
ruff
|
scipy
|
soundfile
|
streamlit
|
ty
Downloads last day:
41
Downloads last week:
239
Downloads last month:
758