PyPI page
Home page
Author:
None
Summary:
Fast LLM inference with 2.8x speedup using speculative decoding
Latest version:
1.0.0
Required dependencies:
accelerate
|
numpy
|
peft
|
tokenizers
|
torch
|
transformers
Optional dependencies:
black
|
flake8
|
isort
|
mypy
|
pytest
|
pytest-cov
|
sphinx
|
sphinx-rtd-theme
Downloads last day:
0
Downloads last week:
4
Downloads last month:
25