ngxson/llama.cpp

Pull Requests Commits

no need backticks [no ci]

ngxson committed 349 days ago

90cb25c9

ngxson committed 349 days ago

a160819b

rm cmd in log output [no ci]

ngxson committed 349 days ago

9fe79c06

Apply suggestions from code review [no ci]

ngxson committed 349 days ago

Verified 6bd90cf8

codeowners : (@ngxson) only watch dockerfile

ngxson committed 349 days ago

1714f5ed

github : cmd line to bug report

ngxson committed 349 days ago

29cbff73

Vulkan: Add device-specific blacklist for coopmat for the AMD proprietary driver (#11074)

0cc4m committed 350 days ago

Verified b56f079e

llama : Add support for DeepSeek V3 (#11049)

fairydreaming committed 350 days ago

Verified 9394bbd4

[GGML][RPC] Support for models with non-512-aligned tensors over RPC. (#11047)

matt23654 committed 350 days ago

Verified f922a9c5

llama : add support for the cohere2 model architecture (#10900)

dranger003 committed 350 days ago

Verified 46be9422

ggerganov committed 350 days ago

78c67851

ggml : do not install metal source when embed library (ggml/1054)

ggerganov committed 350 days ago

5e3b08d6

ggml : improve inputs log sched_print_assignments (ggml/1053)

danbev committed 350 days ago

db68c93b

fix: Vulkan shader gen binary path (#11037)

giladgd committed 351 days ago

Verified c31fc8b9

common : disable KV cache shifting automatically for unsupported models (#11053)

MollySophia committed 352 days ago

Verified 4b0c638b

metal : avoid uint (#11019)

ggerganov committed 352 days ago

Verified e7da954e

llama : refactor `src/llama.cpp` (#10902)

ggerganov committed 352 days ago

Verified f66f5829

server: bench: minor fixes (#10765)

phymbert committed 352 days ago

Verified 2f0ee84b

server : allow using LoRA adapters per-request (#10994)

ngxson committed 352 days ago

Verified 0da5d860

readme : add llama-swap to infrastructure section (#11032)

mostlygeek committed 353 days ago

Verified a45433ba

ggml : fixes for AVXVNNI instruction set with MSVC and Clang (#11027)

Srihari-mcw committed 354 days ago

Verified 0827b2c1

server : clean up built-in template detection (#11026)

ngxson committed 354 days ago

Verified 45095a61

server : add OAI compat for /v1/completions (#10974)

ngxson committed 355 days ago

Verified 5896c652

convert : fix Llama-3_1-Nemotron-51B rope settings (#11008)

ymcki committed 355 days ago

Verified bc7b1f86

common, examples, ggml : fix MSYS2 GCC compiler errors and warnings when building with LLAMA_CURL=ON and GGML_OPENCL=ON (#11013)

peter277 committed 355 days ago

Verified 6e1531ac

vulkan: optimize mul_mat for small values of N (#10991)

jeffbolznv committed 355 days ago

Verified 716bd6de

android : fix llama_batch free (#11014)

ag2s20150909 committed 356 days ago

Verified c250ecb3

vulkan: im2col and matmul optimizations for stable diffusion (#10942)

jeffbolznv committed 357 days ago

Verified a813badb

vulkan: Use push constant offset to handle misaligned descriptors (#10987)

jeffbolznv committed 357 days ago

Verified fdd21889

server: added more docs for response_fields field (#10995)

isaac-mcfadyen committed 357 days ago

Verified f865ea14

Older