ggerganov/llama.cpp

Pull Requests Commits

cmake : sync/merge PowerPC build commands (#0)

ggerganov committed 1 year ago

0306aad1

llamafile : ppc64le MMA implementation for Q4_0. (#12489)

amritahs-ibm committed 1 year ago

Verified c7b43ab6

ggml : riscv: add 128-bit RVV support (#12530)

xctan committed 1 year ago

Verified 24feaec0

llama : make loras compatible with repacking (#12593)

ggerganov committed 1 year ago

Verified f28bc4c2

SYCL: implement memset ggml backend buffer interface (#12580)

qnixsynapse committed 1 year ago

Verified f17a3bb4

HIP: Add support for RDNA4 targets (#12372)

slojosic-amd committed 1 year ago

Verified bd40678d

metal : refactor mat-vec code (#12569)

ggerganov committed 1 year ago

Verified b3298fa4

upgrade to llguidance 0.7.10 (#12576)

mmoskal committed 1 year ago

Verified 2447ad8a

clip: Fix llama-llava-clip-quantize-cli quantization error under CUDA backend (#12566)

Ivy233 committed 1 year ago

Verified 02082f15

convert : fix squeeze for ssm_conv tensors (#12573)

ggerganov committed 1 year ago

Verified df4d20cd

ggml : fix MUL_MAT_ID repack with Q8_K (#12544)

ggerganov committed 1 year ago

Verified 5ed38b68

doc: [MUSA] minor changes (#12583)

yeahdongcn committed 1 year ago

Verified fd7855f8

convert: fix Mistral3/Gemma3 model hparams init (#12571)

CISC committed 1 year ago

Verified 53af4dba

run: de-duplicate fmt and format functions and optimize (#11596)

ericcurtin committed 1 year ago

Verified ef19c717

ggml-cpu : update KleidiAI to v1.5.0 (#12568)

eddnjjn committed 1 year ago

Verified 053b3f9a

SYCL: disable Q4_0 reorder optimization (#12560)

qnixsynapse committed 1 year ago

Verified e2f56017

docs : add build instructions for KleidiAI (#12563)

eddnjjn committed 1 year ago

Verified 36ee06dd

ci: [MUSA] add CI and update doc (#12562)

yeahdongcn committed 1 year ago

Verified 3cd3a395

context : fix worst-case reserve outputs (#12545)

ggerganov committed 1 year ago

Verified 2d77d88e

ci: [SYCL] ggml-ci Use main GPU and enable sysman (#12547)

qnixsynapse committed 1 year ago

Verified c95fa362

opencl: simplify kernel embedding logic in cmakefile (#12503)

lhez committed 1 year ago

Verified 2b65ae30

CI: fix SYCL build (#12546)

qnixsynapse committed 1 year ago

Verified 48d7021c

docs: update: improve the Fedoa CUDA guide (#12536)

teihome committed 1 year ago

Verified 3361e2de

llama-vocab : add SuperBPE pre-tokenizer (#12532)

compilade committed 1 year ago

Verified 00d53800

CUDA: Fix clang warnings (#12540)

yeahdongcn committed 1 year ago

Verified 7ea75035

mmap : skip resource limit checks on AIX (#12541)

mehendarkarprajwal committed 1 year ago

Verified c54f6b79

vulkan: fix mul_mat_vec failure in backend tests (#12529)

jeffbolznv committed 1 year ago

Verified 9b169a4d

server : Add verbose output to OAI compatible chat endpoint. (#12246)

mglambda committed 1 year ago

Verified 77f9c6bb

install : add macports (#12518)

IOOI-SqAR committed 1 year ago

Verified 18b663d8

llama : gemma3 : use output tensor if it exists in model weight (#12506)

ngxson committed 1 year ago

Verified fbdfefe7

Newer Older