ggerganov/llama.cpp

Pull Requests Commits

speculative : add infill mode

ggerganov committed 282 days ago

Verified b83cae08

vulkan: Fix a vulkan-shaders-gen arugment parsing error (#10484)

sparkleholic committed 283 days ago

Verified 0eb4e12b

Introduce llama-run (#10291)

ericcurtin committed 283 days ago

Verified 0cc63754

ci : build docker images only once daily (#10503)

slaren committed 283 days ago

Verified 50d5cecb

server : add more information about error (#10455)

ggerganov committed 283 days ago

Verified 9fd8c268

server : enable cache_prompt by default (#10501)

ggerganov committed 283 days ago

Verified 47f931c8

metal : enable mat-vec kernels for bs <= 4 (#10491)

ggerganov committed 283 days ago

Verified 106964e3

Rename Olmo1124 to Olmo2 (#10500)

2015aroras committed 283 days ago

Verified 80acb7b4

llama : accept a list of devices to use to offload a model (#10497)

slaren committed 283 days ago

Verified 10bce045

Github: update issue templates [no ci] (#10489)

JohannesGaessler committed 283 days ago

Verified 1f922254

Add download chat feature to server chat (#10481)

brucepro committed 283 days ago

Verified a9a678a6

server : add speculative decoding support (#10455)

ggerganov committed 283 days ago

Verified 9ca2e677

ggml : add support for dynamic loading of backends (#10469)

slaren committed 283 days ago

Verified 5931c1f2

tests : fix compile warning

ggerganov committed 283 days ago

Verified f6d12e7d

metal : minor code formatting

ggerganov committed 283 days ago

Verified b7564411

[SYCL] Fix building Win package for oneAPI 2025.0 update (#10483)

NeoZhangJianyu committed 283 days ago

Verified 5a898779

speculative : refactor and add a simpler example (#10362)

ggerganov committed 283 days ago

Verified d9d54e49

flake.lock: Update (#10470)

ggerganov committed 284 days ago

Verified cce5a900

llama : fix op mul check with command-r-plus (#10476)

slaren committed 284 days ago

Verified dc39012c

convert : XLMRoberta Type Vocab Size (#10458)

gabe-l-hart committed 284 days ago

Verified 9336db46

fix gguf-py: Conversion error when multiple licenses are configured (#9807)

mmngays committed 285 days ago

Verified 96fa2c5e

ggml : do not use ARM features not included in the build (#10457)

slaren committed 285 days ago

Verified 55ed008b

ci: Update oneAPI runtime dll packaging (#10428)

MakeDecisionWorth committed 286 days ago

Verified 6dfcfef0

GitHub: ask for more info in issue templates (#10426)

JohannesGaessler committed 286 days ago

Verified 599b3e0c

CANN: Support Ascend310P to accelerate F32 and F16 Model (#10216)

leo-pony committed 287 days ago

Verified c18610b4

cuda : optimize argmax (#10441)

slaren committed 287 days ago

Verified a5e47592

llama : handle KV shift for recurrent models (#10402)

ggerganov committed 287 days ago

Verified 1bb30bf2

ggerganov committed 287 days ago

Verified 87a533be

ggml/sched : do not skip views in pre-assignments

slaren committed 287 days ago

Verified 59b91728

ggml-opt: fix data corruption (ggml/1022)

JohannesGaessler committed 287 days ago

Verified 02e4eaf2

Older