]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
kv-cache : better estimate of n_kv for multi-sequence batches (#15610)
authorGeorgi Gerganov <redacted>
Wed, 27 Aug 2025 10:55:12 +0000 (13:55 +0300)
committerGitHub <redacted>
Wed, 27 Aug 2025 10:55:12 +0000 (13:55 +0300)
commit1bded5a3b3376b2aa3ba2f11ade5910c550f354b
tree9f0c1bb4ca1a6d180ec21b8cfe5424cdd4a3ce32
parent1e7489745a74996fc36e8fd05b73aa16bc184e0c
kv-cache : better estimate of n_kv for multi-sequence batches (#15610)

ggml-ci
src/llama-kv-cache.cpp
src/llama-kv-cache.h