Daniel J Walsh comments

Results 1685 comments of


                                            Daniel J Walsh

RAG not saved on OSX

Did it greate a rag_output OCI Image? $ podman images | grep rag_output

``` ./bin/ramalama rag README.md rag_output Building rag_output ... adding vectordb ... 305304ffe66a9a659fe62d7a4e3607a94200501798d80c665781a3b087db8666 ramalama (rag) $ podman images | grep rag_output localhost/rag_output latest 305304ffe66a 17 seconds ago 344 kB $ podman...

RAG not saved on OSX

If you want to just make a directory then add / or ./ to the output path. ``` ./bin/ramalama rag --format markdown README.md ./rag_output tree rag_ramalama (rag) $ tree rag_output...

RAG not saved on OSX

Please give me an example?

RAG not saved on OSX

Did you do: ramalama rag file.md ./output_dir ? The command you showed would create an image called `localhost/output_dir`

rag: ramalama run --rag tries to connect to port 8080 even if a different port is chosen

We should leak the PORT into the container so that scripts will no which to connect to. Interested in opening a PR?

rag: ramalama run --rag tries to connect to port 8080 even if a different port is chosen

I like this alot, although I think in the long run we might move to llama-stack for connecting these services together.

rag: ramalama run --rag tries to connect to port 8080 even if a different port is chosen

@ueno do you know if we have the latest llama.cpp in our containers yet, or do we need to update the release?

rag: ramalama run --rag tries to connect to port 8080 even if a different port is chosen

Then I believe this is fixed, closing, reopen if I am mistaken.

rag: ramalama serve --rag doesn't take into account of RAG

Yes this is something we are working on. Thanks for the issue.