DeepSpeed-FastGen: High-Throughput Text Generation for LLMs (github.com)

3 points by schrodeenger 2y ago ↗ HN

0 comments

[ 9.3 ms ] story [ 334 ms ] thread

No comments yet.

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

Source · Live DB stats