ggerganov/llama.cpp

Pull Requests Commits

Add quantize-stats command for testing quantization (#728)

unbounded committed 2 years ago

Verified 62cfc54f

make : add libllama.so target for llama-cpp-python (#797)

bhubbb committed 2 years ago

Verified 698f7b5d

zig : don't link examples/common.cpp for non-example (#814)

iacore committed 2 years ago

Verified c1950c34

llama : always sort logits before nucleus sampling (#812)

ivanstepanovftw committed 2 years ago

Verified 4953e900

Do not crash when it has nothing to say. (#796)

l29ah committed 2 years ago

Verified cc9cee8e

Make docker instructions more explicit (#785)

prusnak committed 2 years ago

Verified d2beca95

ggml : multi-thread ggml_rope() (~3-4 times faster on M1) (#781)

ggerganov committed 2 years ago

Verified eeaa7b04

ggml, llama : avoid heavy V transpose + improvements (#775)

ggerganov committed 2 years ago

Verified 986b6ce9

Update README.md

ggerganov committed 2 years ago

Verified 34162989

llama : define non-positive top_k; top_k range check (#779)

ivanstepanovftw committed 2 years ago

Verified 5a8c4f62

miku.sh : add executable bit (#780)

at8u committed 2 years ago

Verified ff05d05c

media : add logos and banners

ggerganov committed 2 years ago

Verified 62b3e81a

readme : change logo + add bindings + add uis + add wiki

ggerganov committed 2 years ago

Verified 8d10406d

zig : add build.zig (#773)

iacore committed 2 years ago

Verified ed1c214e

make : missing host optimizations in CXXFLAGS (#763)

ivanstepanovftw committed 2 years ago

Verified 0c44427d

readme : update with CMake and windows example (#748)

adithyab94 committed 2 years ago

Verified 594cc95f

examples : add Miku.sh (#724)

at8u committed 2 years ago

Verified 88ed5761

Add Accelerate/BLAS when using Swift (#765)

a10y committed 2 years ago

Verified 58c438cf

Windows: reactive sigint handler after each Ctrl-C (#736)

mgroeber9110 committed 2 years ago

Verified 53dbba76

10+% performance improvement of ggml_vec_dot_q4_0 on AVX2 (#654)

SebastianApel committed 2 years ago

Verified 437e7785

Define non-positive temperature behavior (#720)

ivanstepanovftw committed 2 years ago

Verified cd7fa956

Remove torch GPU dependencies from the Docker.full image (#665)

bsilvereagle committed 2 years ago

Verified a0c05164

Add a missing step to the gpt4all instructions (#690)

ThatcherC committed 2 years ago

Verified d8d4e865

Added api for getting/setting the kv_cache (#685)

chrfalch committed 2 years ago

Verified e986f948

ggml : change ne to int64_t (#626)

Seltsamsel committed 2 years ago

Verified c0bb1d3c

examples : add gpt4all script (#658)

leonardohn committed 2 years ago

Verified 6e7801d0

llama : do not allocate KV cache for "vocab_only == true" (#682)

sw committed 2 years ago

Verified 81040f10

make : use -march=native -mtune=native on x86 (#609)

cmdrf committed 2 years ago

Verified c4f89d8d

fix default params for examples/main (#697)

mvrilo committed 2 years ago

Verified 5b70e7de

py: huggingface -> Hugging Face (#686)

eltociear committed 2 years ago

Verified a717cba8

Newer Older