Skip to content

llama.cpp backend: GPU offload and sequence-copy fan-out for shared mode - #34

Open
maximelefrancois86 wants to merge 7 commits into
TheoLeeCJ:masterfrom
maximelefrancois86:llamacpp-cuda-parallel
Open

maximelefrancois86 wants to merge 7 commits into
TheoLeeCJ:masterfrom
maximelefrancois86:llamacpp-cuda-parallel

Enhance your code review process with GitHub Actions

GitHub Actions make it easy to automate all your software workflows, now with world-class CI/CD.
Build, test, and deploy your code right from GitHub. Learn more about GitHub Actions.

Linux, macOS, Windows, and containers
Linux, macOS, Windows, and containers
Matrix builds
Matrix builds
Any language
Any language
Live logs
Live logs
Built-in secret store
Built-in secret store
Multi-container testing
Multi-container testing