Microsoft Open Source 3df35280a6 LICENSE committed
2023-07-06 23:26:06 -07:00
2023-07-07 06:26:00 +00:00
2023-07-06 23:26:06 -07:00
Description
To speed up LLMs' inference and enhance LLM's perceive of key information, compress the prompt and KV-Cache, which achieves up to 20x compression with minimal performance loss.
MIT 3 MiB
Languages
Python 99.5%
Makefile 0.5%