COMMITS
May 14, 2024
D
chore(deps): bump pypa/cibuildwheel from 2.17.0 to 2.18.0 (#1453)
dependabot[bot] committed
A
Update llama.cpp
Andrei Betlen committed
May 13, 2024
A
feat: Update llama.cpp
Andrei Betlen committed
May 12, 2024
A
fix(ci): Use version without extra platform tag in pep503 index
Andrei Betlen committed
A
chore: Bump version
Andrei Betlen committed
A
Merge branch 'main' of https://github.com/abetlen/llama-cpp-python into main
Andrei Betlen committed
A
feat: Update llama.cpp
Andrei Betlen committed
May 10, 2024
P
docs: Fix typo in README.md (#1444)
Peng Yu committed
A
fix: Enable CUDA backend for llava. Closes #1324
Andrei Betlen committed
A
chore: Bump version
Andrei Betlen committed
A
feat: Update llama.cpp
Andrei Betlen committed
A
fix: Clear kv cache to avoid kv bug when image is evaluated first
Andrei Betlen committed
A
chore: Bump version
Andrei Betlen committed
S
fix(security): Render all jinja templates in immutable sandbox (#1441)
Sigbjørn Skjæret committed
P
Merge pull request from GHSA-56xg-wfcc-g829
Patrick Peng committed
May 9, 2024
S
feat: Support multiple chat templates - step 1 (#1396)
Sigbjørn Skjæret committed
A
chore: Bump version
Andrei Betlen committed
A
fix: free last image embed in llava chat handler
Andrei Betlen committed
May 8, 2024
A
fix: Make leading bos_token optional for image chat formats, fix nanollava system message
Andrei Betlen committed
A
feat: Update llama.cpp
Andrei Betlen committed
A
chore: Bump version
Andrei Betlen committed
S
feat: fill-in-middle support (#1386)
Sigbjørn Skjæret committed
A
feat: Update llama.cpp
Andrei Betlen committed
S
fix: adding missing args in create_completion for functionary chat handler (#1430)
Sarunas Kalade committed
I
docs: update README.md (#1432)
Ikko Eltociear Ashimine committed
B
fix: chat_format log where auto-detected format prints `None` (#1434)
Bruno Alvisio committed
May 5, 2024
A
feat(server): Add support for setting root_path. Closes #1420
Andrei Betlen committed
O
feat(ci): Add docker checks and check deps more frequently (#1426)
Olivier DEBAUCHE committed
A
feat: Update llama.cpp
Andrei Betlen committed
May 4, 2024