ggml-org/llama.cpp

Pull Requests Commits

ggerganov committed 1 year ago

Verified f7b0233e

scripts : fix missing key in compare-llama-bench.py (#10332)

ggerganov committed 1 year ago

Verified f245cc28

vulkan: Optimize some mat-vec mul quant shaders (#10296)

jeffbolznv committed 1 year ago

Verified 772703c8

vulkan : add cmake preset debug/release (#10306)

FirstTimeEZ committed 1 year ago

Verified dd3a6ce9

ggml : optimize Q4_0 into Q4_0_X_Y repack (#10324)

eddnjjn committed 1 year ago

Verified 1e58ee13

llama : save number of parameters and the size in llama_model (#10286)

FirstTimeEZ committed 1 year ago

Verified 89e4caaa

Make updates to fix issues with clang-cl builds while using AVX512 flags (#10314)

Srihari-mcw committed 1 year ago

Verified 74d73dc8

scripts: update compare-llama-bench.py (#10319)

JohannesGaessler committed 1 year ago

Verified 4047be74

ggml : fix some build issues

slaren committed 1 year ago

883d206f

cmake : fix ppc64 check (whisper/0)

ggerganov committed 1 year ago

09ecbcb5

ggml : vulkan logs (whisper/2547)

thewh1teagle committed 1 year ago

32250089

ggerganov committed 1 year ago

cbf5541a

AVX BF16 and single scale quant optimizations (#10212)

netrunnereve committed 1 year ago

Verified 18429220

ci: build test musa with cmake (#10298)

yeahdongcn committed 1 year ago

Verified f0204a0e

sycl: Update Intel docker images to use DPC++ 2025.0 (#10305)

Rbiessy committed 1 year ago

Verified 57f8355b

server : (web UI) add copy button for code block, fix api key (#10242)

ngxson committed 1 year ago

Verified 9901068a

cann: dockerfile and doc adjustment (#10302)

noemotiovon committed 1 year ago

Verified 231f9360

scripts : fix regex in sync [no ci]

ggerganov committed 1 year ago

Verified 4802ad35

sycl: Use syclcompat::dp4a (#10267)

Rbiessy committed 1 year ago

Verified 5a54af4d

backend cpu: add online flow for aarch64 Q4_0 GEMV/GEMM kernels (#9921)

chaxu01 committed 1 year ago

Verified 1607a5e5

ggml : build backends as libraries (#10256)

slaren committed 1 year ago

Verified ae8de6d5

CUDA: no -sm row for very small matrices (#10185)

JohannesGaessler committed 1 year ago

Verified 4a8ccb37

speculative : fix out-of-bounds access (#10289)

ggerganov committed 1 year ago

Verified 2a82891a

vulkan: Optimize binary ops (#10270)

jeffbolznv committed 1 year ago

Verified af148c93

vulkan: Use macros to make the mat mul pipeline creation more concise (#10259)

jeffbolznv committed 1 year ago

Verified 66798e42

llama : propagate the results of `graph_compute` (#9525)

Xarbirus committed 1 year ago

Verified fb4a0ec0

ggerganov committed 1 year ago

Verified 5ea926da

docs : update bindings list (#10261)

xuegao-tzx committed 1 year ago

Verified 1ee9eea0

server : add missing docs (#10269)

z80maniac committed 1 year ago

Verified ff7fb670

server : fix incorrect res in validate_model_chat_template (#10272)

jhen0409 committed 1 year ago

Verified 0e712a5a

Older