ggerganov/llama.cpp

Pull Requests Commits

server : add /v1/completion endpoint

ggerganov committed 1 year ago

Verified 0085f949

llama-bench : allow using a different printer for stderr with -oe (#7722)

slaren committed 1 year ago

Verified adc9ff38

Improve hipBLAS support in CMake (#7696)

daniandtheweb committed 1 year ago

Verified 987d743d

refine .gitignore (#7688)

jeffzhou2000 committed 1 year ago

Verified b226c122

Per token attributes (#7685)

jaime-m-p committed 1 year ago

Verified 3b38d486

ggml : prevent builds with -ffinite-math-only (#7726)

ggerganov committed 1 year ago

Verified 6d161694

llama : offload to RPC in addition to other backends (#7640)

rgerganov committed 1 year ago

Verified bde7cd3c

ggml : use OpenMP as a thread pool (#7606)

msy-kato committed 1 year ago

Verified a5735e44

make: fix debug options not being applied to NVCC (#7714)

JohannesGaessler committed 1 year ago

Verified 0b832d53

Vulkan Mixture of Experts (MoE) support (#7628)

0cc4m committed 1 year ago

Verified 3d7ebf63

cmake : add pkg-config spec file for llama.cpp (#7702)

andy-tai committed 1 year ago

Verified a10cda58

llama : MiniCPM support tied embeddings (#7664)

zkh2016 committed 1 year ago

Verified 6f28a333

llama : avoid double token-to-piece cache (#7654)

ggerganov committed 1 year ago

Verified 549279d8

kompute : implement op_getrows_f32 (#6403)

woachk committed 1 year ago

Verified 9e405b6e

fix bug introduced in using calloc (#7701)

airlied committed 1 year ago

Verified 3413ae21

flake.lock: Update (#7686)

ggerganov committed 1 year ago

Verified 1669810d

chore : add ignore rule for generated server themes (#7689)

teleprint-me committed 1 year ago

Verified 7c4e5b7e

[SYCL] Update rpc-server.cpp to include SYCL backend (#7682)

nickp27 committed 1 year ago

Verified 9422c5e3

Fix FlashAttention debug test, FP32 assert (#7684)

JohannesGaessler committed 1 year ago

Verified e141ce62

server : new UI (#7633)

mounta11n committed 1 year ago

Verified 2e666832

SimpleChat: Simple histogram/repeatMatching driven garbageTrimming, Settings UI, Streaming mode, OpenAi Compat (Model, Authorization Bearer), Save/Restore session, Auto Settings UI (#7548)

hanishkvc committed 1 year ago

Verified 2ac95c9d

CUDA: fix Pascal FA, deq. KV to FP16 for batch > 8 (#7681)

JohannesGaessler committed 1 year ago

Verified 750f60c0

CUDA: quantized KV support for FA vec (#7527)

JohannesGaessler committed 1 year ago

Verified 9b596417

server : update js (#7670)

ggerganov committed 1 year ago

Verified a323ec60

convert-hf : Handle NotImplementedError in convert-hf-to-gguf (#7660)

Galunid committed 1 year ago

Verified 0515ad93

scripts: update compare_llama_bench.py [no ci] (#7673)

JohannesGaessler committed 1 year ago

Verified c8047d53

Improve HIP compatibility (#7672)

daniandtheweb committed 1 year ago

Verified 30e238b2

readme : link homebrew discussion

ggerganov committed 1 year ago

Verified 16926dff

ggml : fix loongson compile warnings (#7537)

ggerganov committed 1 year ago

Verified 0c27e6f6

Somehow '**' got lost (#7663)

Galunid committed 1 year ago

Verified 2e32f874

Older