COMMITS
May 27, 2026
L
vulkan: add REPEAT op support for f16 to f16. (llama/23298)
l8bloom committed
O
CUDA: restrict PDL to CTK >= 12.3 due to MSVC issues (llama/23742)
Oliver Simons committed
W
vulkan: avoid preferring transfer queue on AMD UMA devices (llama/22455)
Winston Ma committed
D
ci : only run docker jobs when pushed to master [no ci] (#3828)
Daniel Bevenius committed
D
docs : add AGENTS.md and CONTRIBUTING.md [no ci] (#3826)
Daniel Bevenius committed
May 26, 2026
V
ggml-zendnn : fixed naming of matmul function (llama/20964)
Vladislav committed
J
vulkan: optimize conv2d and implement coopmat1 support (llama/22620)
Jeff Bolz committed
M
hexagon: add support for CONCAT op (llama/23648)
Max Krasnyansky committed
A
SYCL: implement ggml_sycl_pool_vmm (llama/22862)
Alexey Kopytko committed
M
ggml-webgpu: Add MMVQ path for Q4/Q8/Q2_K/Q4_K and clean up legacy MUL_MAT pipeline (llama/23594)
Masashi Yoshimura committed
N
J
CUDA: missing PDL sync for FWHT, better fallback (llama/23690)
Johannes Gäßler committed
T
cli : merge tokens split across UTF-8 boundaries in JSON output (#3751)
texasich committed
May 25, 2026
F
metal : add apple device id (llama/23566)
forforever73 committed
A
CUDA: add fast walsh-hadamard transform (llama/23615)
Aman Gupta committed
G
release : v1.8.5
Georgi Gerganov committed
G
benches : update
Georgi Gerganov committed
G
sync : ggml
Georgi Gerganov committed
G
ggml : bump version to 0.13.0 (ggml/1510)
Georgi Gerganov committed
J
TP: fix ggml context size calculation (llama/22616)
Johannes Gäßler committed
G
K
readme : add AMD ROCm/HIP GPU build instructions (#3823)
Kaihui-AMD committed
G
talk-llama : sync llama.cpp
Georgi Gerganov committed
G
sync : ggml
Georgi Gerganov committed
G
ggml : bump version to 0.12.1 (ggml/1508)
Georgi Gerganov committed
May 28, 2026
D
ci : add ignore for bindings/{ruby, go} in build.yml [no ci] (#3837)
Daniel Bevenius committed
D
ci : fix include paths for bindings-go job [no ci] (#3835)
Daniel Bevenius committed
D
ci : add on push/pull_request paths ruby job (#3833)
Daniel Bevenius committed
D
ci : renable arm64 docker builds (#3832)
Daniel Bevenius committed
D
ci : set GGML_NATIVE=OFF for bindings-java (#3830)
Daniel Bevenius committed