OntoPrune – Pruning 85% LLM context tokens and 6.7x TTFT on CPU

 (github.com)

1 points | by vigmarcarlo 5 hours ago ago

1 comments