llama-batch: enable parallel sequences for partial rollback #25278
llama-batch: add n_keep_tail in split_equal for recurrent models
ddaaa143
pwilkin
approved these changes
on 2026-07-04
ggerganov
approved these changes
on 2026-07-08
am17an
merged
230ea9d2
into master 3 days ago
Login to write a write a comment.
Login via GitHub