ngxson/llama.cpp

Pull Requests Commits

correct output_id for llama-cpp header

ngxson committed 347 days ago

c1f4a78f

fix embeddings and retrieval

ngxson committed 347 days ago

8ec0ff9b

server: fix batch_spec

ngxson committed 349 days ago

50fb3963

ngxson committed 349 days ago

56e82d02

ngxson committed 349 days ago

42062cc2

fix qwzn2vl mrope position input

ngxson committed 349 days ago

c4fea7fe

llama_batch_ext_init with ctx

ngxson committed 349 days ago

d18a79ed

Merge branch 'master' into xsn/private_batch_api

ngxson committed 349 days ago

1434c2c9

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

eddnjjn committed 349 days ago

Verified 053b3f9a

SYCL: disable Q4_0 reorder optimization (#12560)

qnixsynapse committed 349 days ago

Verified e2f56017

docs : add build instructions for KleidiAI (#12563)

eddnjjn committed 349 days ago

Verified 36ee06dd

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

yeahdongcn committed 349 days ago

Verified 3cd3a395

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

ggerganov committed 349 days ago

Verified 2d77d88e

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

qnixsynapse committed 350 days ago

Verified c95fa362

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

lhez committed 350 days ago

Verified 2b65ae30

CI: fix SYCL build (#12546)

qnixsynapse committed 350 days ago

Verified 48d7021c

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

teihome committed 350 days ago

Verified 3361e2de

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

compilade committed 350 days ago

Verified 00d53800

CUDA: Fix clang warnings (#12540)

yeahdongcn committed 350 days ago

Verified 7ea75035

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

mehendarkarprajwal committed 350 days ago

Verified c54f6b79

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

jeffbolznv committed 350 days ago

Verified 9b169a4d

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

mglambda committed 351 days ago

Verified 77f9c6bb

qwen2vl: fix mrope position

ngxson committed 351 days ago

a9efdbbc

Merge branch 'master' into xsn/private_batch_api

ngxson committed 351 days ago

e8827a6f

install : add macports (#12518)

IOOI-SqAR committed 351 days ago

Verified 18b663d8

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

ngxson committed 352 days ago

Verified fbdfefe7

ggml : fix quantized cpy op (#12310)

ggerganov committed 352 days ago

Verified ba932dfb

musa: refine compute capability (#12493)

yeahdongcn committed 352 days ago

Verified fac63a3d

vulkan: Optimize mul_mat_vec p021 and nc shaders (#12505)

jeffbolznv committed 352 days ago

Verified eddfb438

Vulkan: RTE rounding for cpy to quant (#12480)

stduhpf committed 353 days ago

Verified 4375415b

Older