ggerganov/llama.cpp

Pull Requests Commits

Now fix test-quantize-fns

Iwan Kawrakow committed 1 year ago

68e4fed4

Make quantize_row_iq4_nl do the same thing is quantization on CUDA

Iwan Kawrakow committed 1 year ago

30eef31b

Make quantize_row_iq4_nl do the same thing is quantization on CUDA

Iwan Kawrakow committed 1 year ago

cd4a7c4c

Add ability to use Q5_0, Q5_1, and IQ4_NL for quantized K cache (#6183)

ikawrakow committed 1 year ago

Verified 76aa30a2

Add nvidia and amd backends (#6157)

AidanBeltonS committed 1 year ago

Verified c5b8595e

cuda : fix conflict with std::swap (#6186)

slaren committed 1 year ago

Verified 42e21c68

cuda : print the returned error when CUDA initialization fails (#6185)

slaren committed 1 year ago

Verified 1c51f98a

llava : update MobileVLM-README.md (#6180)

ZiangWu-77 committed 1 year ago

Verified f9c7ba34

llava : add MobileVLM_V2 backup (#6175)

ZiangWu-77 committed 1 year ago

Verified 272935b2

cuda : refactor to remove global resources (#6170)

slaren committed 1 year ago

Verified ccf58aa3

Server: version bump for httplib and json (#6169)

ngxson committed 1 year ago

Verified 91f8ad16

gitignore : ignore curl-related files

ggerganov committed 1 year ago

Verified 6b7e76d2

server : allow to override -ngl in tests (#6170)

ggerganov committed 1 year ago

Verified bc0baab2

Revert "llava : add a MobileVLM_V2-1.7B backup (#6152)"

ggerganov committed 1 year ago

Verified d795988d

llava : add a MobileVLM_V2-1.7B backup (#6152)

ZiangWu-77 committed 1 year ago

Verified f8c4e745

Server: Handle n_keep parameter in the request (#6174)

jkarthic committed 1 year ago

Verified 47cc7a7b

server tests : more pythonic process management; fix bare `except:` (#6146)

cebtenzzre committed 1 year ago

Verified bd60d82d

update readme sycl for new update (#6151)

NeoZhangJianyu committed 1 year ago

Verified 6c0b2877

increase igpu cluster limit (#6159)

abhilash1910 committed 1 year ago

Verified d26e8b66

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

Older