ggml-org/llama.cpp

Pull Requests Commits
ci : add ccache to server builds + fix undefined sanitizer build (#23763)
ggerganov committed 5 days ago
Verified 0d227ec3
docs : fix duplicated "the" in granitevision and model-conversion docs (#23767)
quyentonndbs committed 5 days ago
Verified 1d971bba
convert: add MiniCPM5 tokenizer support (#23384)
zhangtao2-1 committed 5 days ago
Verified 9777256c
server : fix the log message when using SSL (#23393)
rgerganov committed 5 days ago
Verified 7085492c
ggml-zendnn : fixed naming of matmul function (#20964)
truecoder34 committed 5 days ago
Verified b4c0549a
ci : do not allocate ccache for 3rd-party hosted runners (#23730)
ggerganov committed 5 days ago
Verified 0d18aaa9
ci : move [no release] check to dedicated check_release job (#23734)
ggerganov committed 5 days ago
Verified 08bc21b4
ci : add `[no release]` keyword + fix sanitizer builds (#23728)
ggerganov committed 6 days ago
Verified 35a74c8f
ci : move macos jobs to the apple workflow + fix names (#23721)
ggerganov committed 6 days ago
Verified 5190c2ea
vulkan: optimize conv2d and implement coopmat1 support (#22620)
jeffbolznv committed 6 days ago
Verified 7799d31e
ci : remove vulkan SDK dep from webgpu job (#23718)
ggerganov committed 6 days ago
Verified 3a3ed153
hexagon: add support for CONCAT op (#23648)
max-krasnyansky committed 6 days ago
Verified ef66bfab
ci : move more CPU jobs to self-hosted runners (#23715)
ggerganov committed 6 days ago
Verified 678d43d7
ci : move sanitizer jobs to self-hosted runners (#23713)
ggerganov committed 6 days ago
Verified ef41a691
ci : reduce (disable SYCL and CANN builds/releases) (#23705)
ggerganov committed 6 days ago
Verified 3dc7684f
convert : support Gemma4ForCausalLM architecture (#23682)
aoleg committed 6 days ago
Verified dbe9c0c8
models : Attach Mistral3 NVFP4 weight scales (#23629)
michaelw9999 committed 6 days ago
Verified 6fe90def
SYCL: implement ggml_sycl_pool_vmm (#22862)
sanmai committed 6 days ago
Verified 581d020b
tests: test-backend-ops -j <N> to run tests in parallel (#23637)
jeffbolznv committed 6 days ago
Verified 7623de11
model : add support for talkie-1930-13b (#22596)
niklassheth committed 6 days ago
Verified c9d98295
ggml-webgpu: Add MMVQ path for Q4/Q8/Q2_K/Q4_K and clean up legacy MUL_MAT pipeline (#23594)
yomaytk committed 6 days ago
Verified 1506d39e
[WebGPU] Check batch_compute_passes before sending passes when not doing GPU profiling (#23457)
nikhilJain17 committed 6 days ago
Verified 54121f73
CUDA: missing PDL sync for FWHT, better fallback (#23690)
JohannesGaessler committed 6 days ago
Verified 192d8ae8
metal : add apple device id (#23566)
forforever73 committed 6 days ago
Verified 35c9b1f3
snapdragon: bump toolchain docker to v0.7 to fix ui build issues (#23680)
max-krasnyansky committed 6 days ago
Verified 4bead4e3
ci : reduce PR jobs by matching backend paths (#23675)
ggerganov committed 6 days ago
Verified 302e2c26
model: tag ffn_latent as MUL_MAT to fix buft probe (#23664)
ServeurpersoCom committed 7 days ago
Verified 328874d0
CUDA: add fast walsh-hadamard transform (#23615)
am17an committed 7 days ago
Verified c1f1e28d
ui: fix stop/continue during an agentic loop (#23356)
ServeurpersoCom committed 7 days ago
Verified 5a4126ad
convert : add compressed-tensors NVFP4 support (#21095)
michaelw9999 committed 7 days ago
Verified a4d2d4ae
Terms Of Use Privacy Policy Refund Policy Impressum