ngxson/llama.cpp

Pull Requests Commits

fix auto scroll

ngxson committed 312 days ago

e61bebb7

add python code interpreter

ngxson committed 312 days ago

27cf968f

redo Settings modal UI

ngxson committed 312 days ago

901d4ac6

add some more comments

ngxson committed 313 days ago

a3547742

server : (webui) fix numeric settings being saved as string

ngxson committed 313 days ago

a54c54ae

Make logging more verbose (#11714)

ericcurtin committed 313 days ago

Verified d2fe216f

llama : fix defrag logic (#11707)

ggerganov committed 313 days ago

Verified ed926d88

vocab : ignore invalid UTF-8 input in the BPE tokenizer (#11729)

cfillion committed 313 days ago

Verified 2d219b38

llama : fix progress dots (#11730)

magicse committed 313 days ago

Verified 333820d7

vulkan: print shared memory size (#11719)

jeffbolznv committed 313 days ago

Verified c026ba3c

llama : add llama_sampler_init for safe usage of llama_sampler_free (#11727)

cfillion committed 313 days ago

Verified 7ee953a6

SYCL: remove XMX info from print devices (#11712)

qnixsynapse committed 313 days ago

Verified ec3bc827

common : add default embeddings presets (#11677)

danbev committed 313 days ago

Verified b7552cfc

ggml : optimize and build warning fix for LoongArch (#11709)

MQ-mengqing committed 313 days ago

Verified 225bbbfa

llama : fix old glm4 models (#11670)

tv1wnd committed 313 days ago

Verified 855cd073

ggerganov committed 313 days ago

Verified 8a59053f

rpc: fix known RCE in rpc-server (ggml/1103)

retr0reg committed 313 days ago

Verified 1d20e53c

server : (webui) migrate project to ReactJS with typescript (#11688)

ngxson committed 314 days ago

Verified 2fb3c32a

docs: update fedora cuda guide for 12.8 release (#11393)

teihome committed 314 days ago

Verified 9ab42dc7

SYCL: Adjust support condition for norm operators (#11674)

qnixsynapse committed 314 days ago

Verified 194b2e69

llama : add log about loading model tensors (#11699)

ggerganov committed 314 days ago

Verified 9dd7a039

build : fix llama.pc (#11658)

angt committed 314 days ago

Verified c0d48432

ggml : fix LoongArch compile error with 128-bit SIMD (#11701)

junchao-loongson committed 314 days ago

Verified 8d4d2be1

vulkan: optimize coopmat2 iq2/iq3 callbacks (#11521)

jeffbolznv committed 314 days ago

Verified 2c6c8df5

vulkan: initial support for IQ4_XS quantization (#11501)

remyoudompheng committed 314 days ago

Verified 8a7e3bf1

vulkan: use smaller combined allocations to avoid fragmentation (#11551)

jeffbolznv committed 314 days ago

Verified 1b598b30

metal : avoid breaking build when metal API predates TARGET_OS_VISION (#11690)

charles-dyfis-net committed 314 days ago

Verified 902368a0

readme : add link to Autopen under UIs (#11684)

blackhole89 committed 314 days ago

Verified c3db0480

metal : adjust support conditions for norm operators (#11671)

ggerganov committed 315 days ago

Verified d774ab3a

CUDA: support for mat. mul. with ne03 != ne13 (#11656)

JohannesGaessler committed 315 days ago

Verified fa62da9b

Older