 Jeff Bolz
|
a0f3897d53
vulkan: fix top_k bug when there are ties in the input (#17659)
|
1 month ago |
 Acly
|
e15cd06a94
vulkan : support conv-2d with large output size (#17685)
|
1 month ago |
 Reese Levine
|
fd57b24c0f
ggml webgpu: unary op suppport, code refactoring, ops support (#17764)
|
1 month ago |
|
Jeff Bolz
|
6ab0d64960
vulkan: enable mmvq for q2_k on NVIDIA (#17675)
|
1 month ago |
|
Jeff Bolz
|
93bb92664e
vulkan: set all memory allocations to high priority (#17624)
|
1 month ago |
 Georgi Gerganov
|
8160b38a5f
rpc : fix alloc size logic (#17116)
|
1 month ago |
|
Georgi Gerganov
|
c41bde6fbd
metal : add residency sets keep-alive heartbeat (#17766)
|
1 month ago |
 Johannes Gäßler
|
6016d0bd41
HIP : fix RDNA4 build (#17792)
|
1 month ago |
 Pascal
|
1be97831e4
fix: prevent segfault in tokenizer on highly repetitive input (#17786)
|
1 month ago |
 Adrien Gallouët
|
a6cfc212ed
ci : fix winget workflow (#17790)
|
1 month ago |
 shalinib-ibm
|
3a0d10533a
Q4/Q8 Tiled Gemm Optimization. (#16999)
|
1 month ago |
 Piotr Wilkin (ilintar)
|
6648989673
Add pwilkin to CODEOWNERS for chat files (#17789)
|
1 month ago |
|
Johannes Gäßler
|
e95d0bc8fd
CUDA: fix FA VKQ accumulator overflow (#17746)
|
1 month ago |
 Jiacheng (Jason) Chen
|
668ed76574
HIP: enable WMMA-MMQ INT kernels for RDNA 3 (#17576)
|
1 month ago |
 Sigbjørn Skjæret
|
03d9a77b85
ci : transform release binary root dir in tar to llama-bXXXX (#17773)
|
1 month ago |
 Gabe Goodhart
|
3143a755c8
docs : update ops.md (Metal, BLAS) (#17768)
|
1 month ago |
|
Piotr Wilkin (ilintar)
|
96fe9badfc
Add support for CUMSUM and TRI for CUDA. (#17584)
|
1 month ago |
|
Gabe Goodhart
|
bde188d60f
metal: TRI, FILL, EXPM1, SOFTPLUS (#16623)
|
1 month ago |
 Xuan-Son Nguyen
|
9d0229967a
server: strip content-length header on proxy (#17734)
|
1 month ago |
|
Xuan-Son Nguyen
|
c4c10bfb86
server: move msg diffs tracking to HTTP thread (#17740)
|
1 month ago |
 Daniel Bevenius
|
817d743cc1
examples : add missing code block end marker [no ci] (#17756)
|
1 month ago |
|
Daniel Bevenius
|
bd4ef13476
common : skip model validation when --help is requested (#17755)
|
1 month ago |
 Alberto Cabrera Pérez
|
87a2084c45
ggml-cpu : remove asserts always evaluating to false (#17728)
|
1 month ago |
 SmartestWashingMachine
|
3659aa28e9
convert: use existing local chat_template if mistral-format model has one. (#17749)
|
1 month ago |
|
Adrien Gallouët
|
2a73f81f8a
cmake : simplify build info detection using standard variables (#17423)
|
1 month ago |
|
Sigbjørn Skjæret
|
7dba049b07
ci : disable ggml-ci-x64-amd-* (#17753)
|
1 month ago |
|
Adrien Gallouët
|
83c1171529
common: use native MultiByteToWideChar (#17738)
|
1 month ago |
|
Georgi Gerganov
|
0d1324856f
metal : use params per pipeline instance (#17739)
|
1 month ago |
|
Georgi Gerganov
|
a67ef0f47f
llama : fix sanity checks during quantization (#17721)
|
1 month ago |
|
Adrien Gallouët
|
ef75a89fdb
build : move _WIN32_WINNT definition to headers (#17736)
|
1 month ago |
