login sign up

Lossless LLM compression for efficient GPU inference via dynamic-length float

arxiv.org · first added by @hn_wayback
saved by
1 person
discussions · 1
Lossless LLM compression for efficient GPU inference via dynamic-length float 411 pts · 117 comments · Apr 2025 411 pts · 117 comments · Apr 2025
0 people saved or upvoted this