ggerganov/llama.cpp

Pull Requests Commits

llama : better replace_all

ggerganov committed 289 days ago

Verified bddcc5f9

baby-llama : remove duplicate vector include

danbev committed 289 days ago

01aae2b4

flake.lock: Update (#8847)

ggerganov committed 290 days ago

Verified 4b77ea95

ggml : reading the runtime sve config of the cpu (#8709)

jdomke committed 290 days ago

Verified 76614f35

Fix conversion of unnormalized BF16->BF16 weights (#7843)

CISC committed 291 days ago

Verified b72c20b8

cann: Fix ggml_cann_im2col for 1D im2col (#8819)

MengqingCao committed 291 days ago

Verified e09a800f

[SYCL] Fixing wrong VDR iq4nl value (#8812)

OuadiElfarouki committed 292 days ago

Verified 0fbbd884

ggml-cuda: Adding support for unified memory (#8035)

matteoserva committed 292 days ago

Verified afbb4c13

Build: Only include execinfo.h on linux systems that support it (#8783)

acon96 committed 292 days ago

Verified b7a08fd5

cuda : fix dmmv cols requirement to 2*GGML_CUDA_DMMV_X (#8800)

slaren committed 292 days ago

Verified 7a11eb3a

cann: support q8_0 for Ascend backend (#8805)

wangshuai09 committed 293 days ago

Verified c8a00909

server : update llama-server embedding flag documentation (#8779)

okigan committed 293 days ago

Verified afbbcf3c

Build: Fix potential race condition (#8781)

HanClinto committed 293 days ago

Verified ed9d2854

Adding Gemma 2 2B configs (#8784)

pculliton committed 293 days ago

Verified 398ede5e

cmake : fix use of external ggml (#8787)

iboB committed 293 days ago

Verified 44d28ddd

nix: cuda: rely on propagatedBuildInputs (#8772)

SomeoneSerge committed 294 days ago

Verified 268c5660

py: add_array() will not add to kv store if value is an empty array (#8774)

mofosyne committed 294 days ago

Verified 7e72aa74

added android implementation of ggml_print_backtrace_symbols (#8751)

l3utterfly committed 294 days ago

Verified 7c27a19b

flake.lock: Update (#8729)

ggerganov committed 294 days ago

Verified 140074bb

cann: update cmake (#8765)

wangshuai09 committed 294 days ago

Verified 6e2b6000

[SYCL] Add `TIMESTEP_EMBEDDING` OP (#8707)

zhentaoyu committed 294 days ago

Verified c887d8b0

ggml: bugfix: fix the inactive elements is agnostic for risc-v vector (#8748)

CarterLi999 committed 295 days ago

Verified 75af08c4

cuda : organize vendor-specific headers into vendors directory (#8746)

yeahdongcn committed 295 days ago

Verified 439b3fc7

[SYCL] add conv support (#8688)

airMeng committed 296 days ago

Verified 0832de72

cmake: use 1 more thread for non-ggml in CI (#8740)

JohannesGaessler committed 296 days ago

Verified 6eeaeba1

chore : Fix vulkan related compiler warnings, add help text, improve CLI options (#8477)

teleprint-me committed 296 days ago

Verified 4730faca

llama : refactor session file management (#8699)

compilade committed 297 days ago

Verified 4c676c85

feat: Support Moore Threads GPU (#8383)

yeahdongcn committed 297 days ago

Verified e54c35e4

scripts : sync vulkan-shaders (#0)

ggerganov committed 297 days ago

Verified 5e2727fe

scripts : sync ggml-aarch64 sources

ggerganov committed 297 days ago

Verified 56f20aa2

Older