ngxson/llama.cpp

Pull Requests Commits

server: fix a race condition cause by "request_completion"

ngxson committed 1 year ago

8f36df8f

server: only add back deferred tasks when one slot is available

ngxson committed 1 year ago

d083c817

server: correct multitask response

ngxson committed 1 year ago

1bd86789

server: move all mutexes away from server.cpp

ngxson committed 1 year ago

d87b48fd

Merge branch 'master' into xsn/clean_up_server

ngxson committed 1 year ago

58fe9cf5

KL-divergence (#5076)

ikawrakow committed 1 year ago

Verified 6f9939d1

ggml : parallelize FP32 conversion when using BLAS (#5045)

ReinForce-II committed 1 year ago

Verified 780e24a2

llava : MobileVLM support (#4954)

XiaotaoChen committed 1 year ago

Verified 3ce7e8f8

flake.nix: add a comment about flakes vs nix

SomeoneSerge committed 1 year ago

b2d80e10

nix: add a comment on the many nixpkgs-with-cuda instances

SomeoneSerge committed 1 year ago

28603cd2

nix: add a comment about makeScope

SomeoneSerge committed 1 year ago

5e97ec91

nix: refactor the cleanSource rules

SomeoneSerge committed 1 year ago

72518707

workflows: nix-ci: drop the redundant "paths" filter

SomeoneSerge committed 1 year ago

fe8b3c0d

workflows: nix-build-aarch64: rate limit

SomeoneSerge committed 1 year ago

f4dd0592

workflows: nix-ci: rebuild on flake.lock updates

SomeoneSerge committed 1 year ago

f7276f75

imatrix : keep intermediate imatrix results (#5077)

ikawrakow committed 1 year ago

Verified 15bceec2

llama : support StableLM 2 1.6B (#5052)

compilade committed 1 year ago

Verified d6bd4d46

finetune : print sample-start/include-sample-start (#5072)

danbev committed 1 year ago

Verified 152d9d05

llama : add Q3_K_XS (#5060)

ikawrakow committed 1 year ago

Verified 66d575c4

ci : fix Windows CI by updating Intel SDE version (#5053)

bobqianic committed 1 year ago

Verified 57744932

llama : add more qwen2 models (#5071)

simonJJJ committed 1 year ago

Verified 3466c6eb

Revert LLAMA_NATIVE to OFF in flake.nix (#5066)

iSma committed 1 year ago

Verified 504dc37b

add safetensors support to convert-lora-to-ggml.py (#5062)

kuronekosaiko committed 1 year ago

Verified 05490fad

add `#include <string>` to unicode.h (#5051)

bobqianic committed 1 year ago

Verified 6c5629d4

server: add comments

ngxson committed 1 year ago

906afe78

server: add llama_server_response_event

ngxson committed 1 year ago

12829b2e

Add ability to evauate multiple choice tasks (#5047)

ikawrakow committed 1 year ago

Verified 7dcbe39d

Slightly faster imatrix (#5050)

ikawrakow committed 1 year ago

Verified 726c0fa9

flake.lock: Update (#5054)

ggerganov committed 1 year ago

Verified 942c0107

convert : partially revert PR #4818 (#5041)

cebtenzzre committed 1 year ago

Verified b43ebde3

Older