ramalama (llm/ramalama) Updated: 1 month, 1 week ago Add to my watchlist
A tool to simplify the use of local AI modelsRamalama is an open-source developer tool that simplifies the local serving of AI models from any source and facilitates their use for inference in production, all through the familiar language of containers.
Version: 0.23.0 License: MIT
GitHub
| Maintainers | neverpanic |
| Categories | science llm |
| Homepage | https://ramalama.ai/ |
| Platforms | darwin |
| Variants |
|
"ramalama" depends on
lib (1)
run (5)
build (5)
Ports that depend on "ramalama"
No ports
Port notes
ramalama defaults to running AI models in podman containers in a podman machine (i.e., VM) started by libkrun. This is not the podman default, so you will have to change it, either by exporting the CONTAINERS_MACHINE_PROVIDER=libkrun environment variable, or by adding 'provider = "libkrun"' to the '[machine]' section of '$HOME/.config/containers/containers.conf'. See man 7 ramalama-macos for more information. To use MLX as the default inference runtime (Apple Silicon only), set 'runtime = "mlx"' under the '[ramalama]' section of '$HOME/.config/ramalama/ramalama.conf', or export the RAMALAMA_RUNTIME=mlx environment variable, or pass --runtime=mlx on the command line. If podman is installed, ramalama will default to using it. To use MLX, either pass the --nocontainer flag to ramalama or set 'container = false' under the '[machine]' section of '$HOME/.config/ramalama/ramalama.conf'.
Port Health:
Loading Port Health
Installations (30 days)
2
Requested Installations (30 days)
2
Livecheck results
ramalama seems to have been updated (port version: 0.23.0, new version: 0.24.0)
livecheck ran: 1 day, 6 hours ago