COMMITS
April 3, 2026
T
resync core and upstream
Timothy Carambat committed
T
Merge branch 'master' of https://github.com/ggml-org/llama.cpp into prism
Timothy Carambat committed
V
ci : add AMD ZenDNN label to PR labeler (#21345)
Vishal Singh committed
April 2, 2026
S
[HIP] Bump ROCm version to 7.2.1 (#21066)
Slobodan Josic committed
P
fix: gemma 4 template (#21326)
Piotr Wilkin (ilintar) committed
B
tests : add unit test coverage for llama_tensor_get_type (#20112)
Bartowski committed
P
Merge pull request #8 from PrismML-Eng/cpu-fixes
Pasha Khosravi committed
P
some cpu fixes; getting ready for upstream PR; e.g. id 40 is taken by nvfp4 now
Pasha Khosravi committed
Z
ggml-webgpu: add vectorized flash attention (#20709)
Zheyuan Chen committed
R
tests: allow exporting graph ops from HF file without downloading weights (#21182)
Ruben Ortlam committed
X
model, mtmd: fix gguf conversion for audio/vision mmproj (#21309)
Xuan-Son Nguyen committed
A
common : add commentary rules for gpt-oss-20b (#21286)
Aldehir Rojas committed
P
Relax prefill parser to allow space. (#21240)
Piotr Wilkin (ilintar) committed
J
chat : add Granite 4.0 chat template with correct tool_call role mapping (#20804)
Jesus Talavera committed
G
kv-cache : do not quantize SWA KV cache (#21277)
Georgi Gerganov committed
R
Ignore Transfer-Encoding header. (#20269)
Roger Chen committed
G
sync : ggml
Georgi Gerganov committed
G
ggml : bump version to 0.9.11 (ggml/1456)
Georgi Gerganov committed
N
sycl : fix llama_kv_cache hang when kv_cache is huge: 5GB (#21283)
Neo Zhang committed
T
hexagon : add cumsum op support (#21246)
Todor Boinovski committed
April 1, 2026
X
contrib : rewrite AGENTS.md, make it more clear about project values (#21270)
Xuan-Son Nguyen committed
L
opencl: fix leak in Adreno q8_0 path (#21212)
lhez committed
A
server: Bypass API Key validation for WebUI static bundle assets (#21269)
Aleksander Grygier committed
J
CUDA: fix FA kernel selection logic (#21271)
Johannes Gäßler committed
M
kleidiai: add CPU feature detection to CI run script (#20394)
Martin Klacer committed
N
Update Dawn version in WebGPU CI (#20784)
Nikhil Jain committed
A
hexagon: improve RMS_NORM and DIV accuracy (#21251)
Aparna M P committed
T
bump readme
Timothy Carambat committed
T
merge KV rotation
Timothy Carambat committed
J
fix: tool call parsing for LFM2 and LFM2.5 models (#21242)
Jonathan committed