v0.3.22-cu124-Basic-linux-20260118
·
265 commits
to main
since this release
Bump version to 0.3.22
Optimizations for TFFT (Time To First Token) can reduce response latency in RAG and chat applications in long contexts.
Changlog see here: 0.3.22 Changelog
Signed-off-by: JamePeng jame_peng@sina.com