PyPI page
Home page
Author:
Lakshmikanthan K (letchupkt)
License:
MIT
Summary:
Adaptive Streaming Vector Quantization KV Cache for LLMs — with Replay + Compare + Multi-Backend Integration (V2.1)
Latest version:
0.2.3
Required dependencies:
numpy
Optional dependencies:
accelerate
|
huggingface-hub
|
llama-cpp-python
|
mlx
|
mlx-lm
|
numpy
|
pytest
|
pytest-timeout
|
requests
|
torch
|
transformers
|
vllm
Downloads last day:
2
Downloads last week:
15
Downloads last month:
163
7-day smoothing