LLM inference in C/C++
C++ ggml
Commits on master
[SYCL] Update compute runtime version to 26.x in docker (#24070)
d403f00 Neo Zhang committed 5 months ago
convert : fix conversion for Mistral-Medium-3.5-128B (#24268)
8a963fc David Friehs committed 5 months ago
kv-cache: follow the source cache size when sharing cells (#24267)
f0156d1 Pascal committed 5 months ago
arg: Skip mmproj download when user supplied mmproj (#24239)
465b1f0 konradmb committed 5 months ago
common/chat : fix LFM2/LFM2.5 reasoning round-trip and <think> leak (#24234)
98d5e8b Tarek Dakhran committed 5 months ago
mtmd: support "frame merge" for qwen-vl-based models (#21858)
31e8249 Xuan-Son Nguyen committed 5 months ago
completion : fix format specifier in LOG_INF (#24213)
588f0dc Adrien Gallouët committed 5 months ago
mtmd, server: add "placeholder bitmap" for counting tokens , add */input_tokens API (#23913)
f5c6ae1 Xuan-Son Nguyen committed 5 months ago
vulkan: check coopmat2 features before reporting support (#24186)
5a69c97 Ruben Ortlam committed 5 months ago
context : fix off-by-one comparisons to n_gpu_layers (#24208)
603300b Sigbjørn Skjæret committed 5 months ago
opencl: improve get_rows, cpy, concat and q6_k flat gemv (#24160)
308f61c lhez committed 5 months ago
common/chat : unify and fix LFM2/LFM2.5 tool parser (#24178)
da87e9b Tarek Dakhran committed 5 months ago
vulkan: add fwht support for Intel with shmem reduction (#23964)
e82beaa Ruben Ortlam committed 5 months ago
ui: run npm install when package-lock.json is newer than node_modules (#24171)
2016bf2 Pascal committed 5 months ago