ggerganov/llama.cpp

Pull Requests Commits

ggml : IQ3_S improvements (#5829)

ikawrakow committed 2 years ago

Verified bbde6eb2

scripts : add pod-llama.sh

ggerganov committed 2 years ago

Verified ef2cd694

llama : refactor internal quantization functions (#5830)

ngxson committed 2 years ago

Verified 6c32d8c7

llama : fix segfault from unknown model arch name (#5820)

compilade committed 2 years ago

Verified 802da009

Support multiple GPUs (split mode) on SYCL backend (#5806)

NeoZhangJianyu committed 2 years ago

Verified 71564139

workflows : remove nocleanup arg for check-requirements.sh (#5826)

crasm committed 2 years ago

Verified 9bf297a0

build(nix): Introduce flake.formatter for `nix fmt` (#5687)

ditsuke committed 2 years ago

Verified cb5e8f7f

convert-hf-to-gguf : require einops for InternLM2ForCausalLM (#5792)

Nold360 committed 2 years ago

Verified da3b9ba2

llama : add StarCoder2 support (#5795)

pacman100 committed 2 years ago

Verified c29af7e2

server : remove api_like_OAI.py proxy script (#5808)

ggerganov committed 2 years ago

Verified 38d16b14

ggml-vulkan: fix VULKAN_CHECK_RESULTS flag, which was previously broken (#5813)

ddpasa committed 2 years ago

Verified c2224f00

gemma : fix bfloat16 -> float16 conversion issue (#5810)

kunal-vaishnavi committed 2 years ago

Verified e7433867

common : fix flag `--logits-all` to `--all-logits` (#5805)

ensan-hcl committed 2 years ago

Verified f49a5356

llama : cleanup unused mmq flags (#5772)

phymbert committed 2 years ago

Verified 3ab8b3a9

unicode : switch to multimap based nfd_map (#5799)

iamlemec committed 2 years ago

Verified 9600d59e

server: allow to override threads server pool with --threads-http (#5794)

phymbert committed 2 years ago

Verified 5cb02b4a

ci : add Ubuntu 22 Vulkan CI run (#5789)

netrunnereve committed 2 years ago

Verified 6ea0f010

server : fix newlines in help (#5785)

ggerganov committed 2 years ago

Verified f105471e

[SYCL] Use batched mul_mat pathway (#5591)

AidanBeltonS committed 2 years ago

Verified 38d15216

Server: normalize naming (#5779)

ngxson committed 2 years ago

Verified 052051d8

llama : constified `llama_set_state_data`'s `src` (#5774)

MarcusDunn committed 2 years ago

Verified d5ab2975

ci : reduce 3b ppl chunks to 1 to avoid timeout (#5771)

ggerganov committed 2 years ago

Verified 87c91c07

make portability_enumeration_ext apple only (#5757)

netrunnereve committed 2 years ago

Verified 317709b2

llama : remove deprecated API (#5770)

ggerganov committed 2 years ago

Verified 08c5ee87

awq-py : remove (#5768)

ggerganov committed 2 years ago

Verified 78aacf36

ggerganov committed 2 years ago

Verified 8c0e8f4e

add google magika inference example (ggml/748)

slaren committed 2 years ago

Verified 2774b0c9

Introduce backend GUIDs (ggml/743)

uextm committed 2 years ago

Verified 5f706718

server : hit Ctrl+C twice to exit (#5734)

ngxson committed 2 years ago

Verified a693bea1

llama : fix non-quantization of expert gating tensors (#5754)

compilade committed 2 years ago

Verified adcb12a9

Newer Older