Files
llama.cpp-xrip/docs
xripandClaude Opus 5.5 74facb828e Merge upstream/master into master
Resolve llama-context.cpp: keep compute lock on the new llama_batch_ext
encode/decode, and reset both gf_res_prev slots in the shared compute path.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-25 11:33:07 +03:00
..
2026-09-25 10:17:51 +03:00
2026-07-30 16:14:37 +03:00
2026-06-04 08:02:54 +03:00
2026-07-30 16:14:37 +03:00
2026-07-30 16:14:37 +03:00