ggerganov/llama.cpp

Pull Requests Commits

server : fix tests expecting old repeat penalty

compilade committed 1 year ago

9a424a38

Remove undeed header file. (#6158)

dranger003 committed 1 year ago

Verified d8b009a9

gguf-split: split and merge gguf per batch of tensors (#6135)

phymbert committed 1 year ago

Verified d0d5de42

common : disable repeat penalties by default (#6127)

ggerganov committed 1 year ago

Verified b80cf3b2

ci : exempt some labels from being tagged as stale (#6140)

slaren committed 1 year ago

Verified 970a4806

common : print usage on '-h' and '--help' (#6145)

dranger003 committed 1 year ago

Verified 4c28b825

flake.lock: Update

github-actions[bot] committed 1 year ago

2d15886b

mpt : implement backwards compatiblity with duped output tensor (#6139)

cebtenzzre committed 1 year ago

Verified d199ca79

clip : fix memory leak (#6138)

felrock committed 1 year ago

Verified 104f5e0f

backend : set max split inputs to GGML_MAX_SRC (#6137)

slaren committed 1 year ago

Verified 5e1b7f94

ci : disable stale issue messages (#6126)

ggerganov committed 1 year ago

Verified ac9ee6a4

ci : temporary disable sanitizer builds (#6128)

ggerganov committed 1 year ago

Verified 4f6d1337

backend : offload large batches to GPU (#6083)

slaren committed 1 year ago

Verified 2bf8d0f7

common : tidy-up argument parsing (#6105)

dranger003 committed 1 year ago

Verified 496bc79b

convert : add support for CamembertModel architecture (#6119)

Royalphax committed 1 year ago

Verified 9b03719a

convert : use f32 outtype for bf16 tensors (#6106)

Artefact2 committed 1 year ago

Verified 3a6efdd0

common: llama_load_model_from_url using --model-url (#6098)

phymbert committed 1 year ago

Verified d01b3c4c

ci : close all stale issues at once (#6115)

ggerganov committed 1 year ago

Verified cd776c37

ggml:fix finding transfer queue family index error (#6094)

GainLee committed 1 year ago

Verified dc0f6125

ggml : add AVX512F SIMD (#6088)

amiralimi committed 1 year ago

Verified c47cf414

gritlm : add initial README.md (#6086)

danbev committed 1 year ago

Verified b5f4ae09

readme : add wllama as a wasm binding (#6100)

ngxson committed 1 year ago

Verified dfbfdd60

common : refactor nested if causing error C1061 on MSVC (#6101)

dranger003 committed 1 year ago

Verified 15961ec0

ci : close inactive issue with workflow (#6053)

phymbert committed 1 year ago

Verified a56d09a4

llama : fix Baichuan2 13B (#6092)

slaren committed 1 year ago

Verified d84c4850

llama : add support for control vectors (#5970)

vgel committed 1 year ago

Verified 877b4d0c

llama : add Command-R support (#6033)

acanis committed 1 year ago

Verified 12247f4c

llava : change API to pure C style for Rust FFI bindgen (#6079)

tinglou committed 1 year ago

Verified 4e9a7f7f

cuda : disable unused cudaLaunchHostFunc code (#6078)

slaren committed 1 year ago

Verified 3020327f

fix set main gpu error (#6073)

NeoZhangJianyu committed 1 year ago

Verified 46acb367

Older