LazyLLM: Dynamic Token Pruning for Efficient Long Context LLM Inference (arxiv.org)

2 points by usernamesdf 1y ago ↗ HN

0 comments

[ 2.9 ms ] story [ 12.6 ms ] thread

No comments yet.

A showcase of shard-db. Data from the Hacker News API. Refreshed every 15 minutes.

Source · Live DB stats