ggerganov/llama.cpp

Pull Requests Commits
Vulkan: Fix device info output format specifiers (#10366)
0cc4m committed 351 days ago
Verified 9b75f03c
docker: use GGML_NATIVE=OFF (#10368)
JohannesGaessler committed 352 days ago
Verified 75207b3a
CUDA: fix MMV kernel being used for FP16 src1 (#10357)
JohannesGaessler committed 352 days ago
Verified 76e9e58b
CMake: fix typo in comment [no ci] (#10360)
JohannesGaessler committed 352 days ago
Verified ce2e59ba
llama : only use default buffer types for the KV cache (#10358)
slaren committed 352 days ago
Verified be5cacce
gitignore : ignore local run scripts [no ci]
ggerganov committed 352 days ago
Verified 20a780c7
metal : refactor kernel args into structs (#10238)
ggerganov committed 352 days ago
Verified cf32a9b9
ggml : fix undefined reference to 'getcpu' (#10354)
FirstTimeEZ committed 352 days ago
Verified a4317829
CUDA: remove DMMV, consolidate F16 mult mat vec (#10318)
JohannesGaessler committed 352 days ago
Verified c3ea58ac
CMake: default to -arch=native for CUDA build (#10320)
JohannesGaessler committed 352 days ago
Verified 467576b6
ggml : fix possible buffer use after free in sched reserve (#9930)
slaren committed 352 days ago
Verified eda7e1d4
ggml : inttypes.h -> cinttypes (#0)
ggerganov committed 352 days ago
24203e9d
ggml : adapt AMX to tensor->grad removal (#0)
ggerganov committed 352 days ago
5d9e5997
make : add ggml-opt (#0)
ggerganov committed 352 days ago
a4200caf
tests : remove test-grad0
ggerganov committed 352 days ago
84274a10
ggml : fix compile warnings (#0)
ggerganov committed 352 days ago
68fcb475
ggml: new optimization interface (ggml/988)
JohannesGaessler committed 352 days ago
8a43e940
scripts : update sync
ggerganov committed 352 days ago
5c9a8b22
docs : vulkan build instructions to use git bash mingw64 (#10303)
FirstTimeEZ committed 353 days ago
Verified 0fff7fd7
llama/ex: remove --logdir argument (#10339)
JohannesGaessler committed 353 days ago
Verified 4e54be0e
llamafile : fix include path (#0)
ggerganov committed 353 days ago
db4cfd5d
make : auto-determine dependencies (#0)
ggerganov committed 353 days ago
8ee0d09a
server: (web UI) Add samplers sequence customization (#10255)
MaggotHATE committed 353 days ago
Verified bcdb7a23
scripts : fix missing key in compare-llama-bench.py (#10332)
ggerganov committed 353 days ago
Verified f245cc28
vulkan: Optimize some mat-vec mul quant shaders (#10296)
jeffbolznv committed 353 days ago
Verified 772703c8
vulkan : add cmake preset debug/release (#10306)
FirstTimeEZ committed 353 days ago
Verified dd3a6ce9
ggml : optimize Q4_0 into Q4_0_X_Y repack (#10324)
eddnjjn committed 353 days ago
Verified 1e58ee13
llama : save number of parameters and the size in llama_model (#10286)
FirstTimeEZ committed 353 days ago
Verified 89e4caaa
Make updates to fix issues with clang-cl builds while using AVX512 flags (#10314)
Srihari-mcw committed 354 days ago
Verified 74d73dc8
scripts: update compare-llama-bench.py (#10319)
JohannesGaessler committed 354 days ago
Verified 4047be74
Terms Of Use Privacy Policy Refund Policy Impressum