PyPI page
Home page
Author:
None
License:
MIT
Summary:
TurboQuant+ compression for vLLM. 4.3x weight compression + 3.7x KV cache, zero calibration.
Latest version:
0.13.3
Required dependencies:
click
|
numpy
|
safetensors
|
scipy
|
torch
|
transformers
Optional dependencies:
pytest
|
ruff
|
vllm
Downloads last day:
30
Downloads last week:
645
Downloads last month:
2,234