ggml-org/llama.cpp

Pull Requests Commits

cebtenzzre committed 1 year ago

7286b83d

Slight quantization improvement for Q4_K and Q5_K (#5361)

ikawrakow committed 1 year ago

Verified f57fadc0

readme : add phi, orion 14b, internlm2, and yi-VL to readme (#5362)

BarfingLemurs committed 1 year ago

Verified 2e9c0bd6

CUDA: mul_mat_vec_q for batch sizes > 1 (#5351)

JohannesGaessler committed 1 year ago

Verified 2c516611

server : include total "num_slots" in props endpoint (#5349)

jparkerweb committed 1 year ago

Verified 8a79c591

server : add `dynatemp_range` and `dynatemp_exponent` (#5352)

m18coppola committed 1 year ago

Verified 31e79032

server : various fixes for the prompt field in /completion (#5300)

Niall- committed 1 year ago

Verified 4ffc7a17

py : handle byte tokens in `get_token_type` (#5341)

ggerganov committed 1 year ago

Verified 906cff55

make: Use ccache for faster compilation (#5318)

JohannesGaessler committed 2 years ago

Verified 098f6d73

README: updated introduction (#5343)

JohannesGaessler committed 2 years ago

Verified 78b00dda

ggml : make use of ggml-quants.h possible in C++ code (#5338)

ikawrakow committed 2 years ago

Verified c6b39553

ggml : avoid duplicating function calls using MIN/MAX macros (#5325)

tom7 committed 2 years ago

Verified abb61944

iq3_xxs: quards for the no-imatrix situation (#5334)

ikawrakow committed 2 years ago

Verified 89503dcb

py : fix internlm2-hf convert to gguf (#5305)

SolenoidWGT committed 2 years ago

Verified 7e1ae372

iq2_xxs: tune quantization (#5320)

ikawrakow committed 2 years ago

Verified 6fdfa2ec

server : allow to get default generation settings for completion (#5307)

z80maniac committed 2 years ago

Verified a2d60c91

common : add dynamic temperature parameters to main example cli (#5295)

l3utterfly committed 2 years ago

Verified e6f81775

scripts : fix typos, cleanup (#5303)

ggerganov committed 2 years ago

Verified 30679d43

scripts : add non-interactive server-llm.sh (#5303)

garrnizon committed 2 years ago

Verified 4be04c89

readme : add CodeShell models to the supported models list (#5330)

chiranko committed 2 years ago

Verified 5d55b0cd

[SYCL] Fix cpy with dims of 3 (#5289)

AidanBeltonS committed 2 years ago

Verified 4833ac20

flake.lock: Update

github-actions[bot] committed 2 years ago

9392ebd4

Adding some imatrix tools (#5302)

ikawrakow committed 2 years ago

Verified 5ed26e1f

cmake : use set() for LLAMA_WIN_VER (#5298)

crimsonmagick committed 2 years ago

Verified 277fad30

make: add nvcc info print (#5310)

JohannesGaessler committed 2 years ago

Verified 3c0d25c4

make: fix nvcc optimization flags for host code (#5309)

JohannesGaessler committed 2 years ago

Verified 3cc5ed35

add Vulkan support to Nix flake

mschwaig committed 2 years ago

60ecf099

Vulkan Intel Fixes, Optimizations and Debugging Flags (#5301)

0cc4m committed 2 years ago

Verified e920ed39

refactor : switch to emplace_back to avoid extra object (#5291)

MKlimenko committed 2 years ago

Verified 52bb63c7

YaRN : store rope scaling type as int32_t in memory (#5285)

cebtenzzre committed 2 years ago

Verified 1ec3332a

Older