ggml-org/llama.cpp

Pull Requests Commits

model : gpt-oss add response_format support (#15494)

aldehir committed 127 days ago

Verified 32732f24

ggml: add `conv3d` op (#15182)

rmatif committed 127 days ago

Verified 92f7f0a5

cuda : add Pad Reflect 1D support (#14659)

YavorGIvanov committed 127 days ago

Verified b1ab9182

llama : remove KV cache defragmentation logic (#15473)

ggerganov committed 127 days ago

Verified 9ebebef6

ggml-cpu: Support Q5_0 and Q5_1 on s390x (#15486)

taronaeo committed 127 days ago

Verified ad5c975c

server : Support multimodal completion and embeddings prompts in JSON format (#15108)

65a committed 127 days ago

Verified 4afb0a74

readme : model : mtdm : lfm2 improvements (#15476)

tdakhran committed 127 days ago

Verified e2886936

CANN: Optimize RMS_NORM using cache (#15419)

noemotiovon committed 127 days ago

Verified a0f98dd6

sched : fix possible use of wrong ids tensor when offloading moe prompt processing (#15488)

slaren committed 128 days ago

Verified 54a241f5

llama : remove deprecated llama_kv_self API (#15472)

ggerganov committed 128 days ago

Verified cd36b5e5

graph : remove build_attn_with_sinks overload (#15469)

ggerganov committed 128 days ago

Verified 3f196be8

vulkan : support conv_2d_dw with f16 weights (#15392)

Acly committed 128 days ago

Verified 97ae5961

vulkan: add exp operation (#15456)

ddwkim committed 128 days ago

Verified 20c2dac8

vulkan: Reuse conversion results in prealloc_y (#15410)

jeffbolznv committed 128 days ago

Verified 96452a3f

examples : fix some typos in examples/model-conversion/README.md (#15477)

DamonFool committed 128 days ago

Verified 9ad5e60d

kv-cache : drop the "unified" prefix (#15467)

ggerganov committed 128 days ago

Verified 715a6db0

examples : install torch-cpu for model conversion tool/example (#15475)

DamonFool committed 128 days ago

Verified ad294df0

ci : enable RVV1.0 native build (#15386)

alitariq4589 committed 128 days ago

Verified 029bb39e

ci : continue file download with wget (#15471)

ggerganov committed 128 days ago

Verified 30649cab

examples : add model conversion tool/example (#15455)

danbev committed 128 days ago

Verified 2758fa10

ci : fix -Werror=return-type in clip.cpp so ci/run.sh can run without issue (#15221)

michaelgiba committed 128 days ago

Verified b108e429

ci : add copilot-instructions.md (#15286)

Copilot committed 128 days ago

Verified 245be739

convert : make Mistral community chat templates optional via parameter (#15420)

juliendenize committed 128 days ago

Verified b2caf67d

common : fix incorrect print of non-ascii characters in the logging (#15466)

DamonFool committed 128 days ago

Verified 2f3dbffb

ggml : fix condition of im2col on Metal backend (#15460)

ngxson committed 128 days ago

Verified 945e1f12

server : fix webui (#15462)

stduhpf committed 128 days ago

Verified 1b0db8f6

examples : remove references to `make` in examples [no ci] (#15457)

danbev committed 128 days ago

Verified 29f538ac

musa: add GGML_UNUSED_VARS (#15446)

yeahdongcn committed 128 days ago

Verified 8ad038c0

sched : copy only the used experts when offloading prompt processing (#15346)

slaren committed 129 days ago

Verified 5682a374

server: fix OpenAI API compatibility for usage statistics in chat streams (#15444)

TeoZosa committed 129 days ago

Verified 1bc664a2

Newer Older