Create model container

[trvrggrep]

As a model developer I want to be able to create a docker image that is capable of running my model, so that my model can be executed independently by third parties.

Acceptance criteria

• the container is compatible/can execute the code language that the model was written in
• model developer can import necessary libraries for model execution
• the container is [infernet-compatible](https://docs.ritual.net/infernet/node/containers)
• the container is created on a machine with an http server so that third parties can interface with it as a result of a post request
• model customers are able to make one time requests for inference, or create a stream of inferences at predefined intervals with start and end dates for the stream.

[trvrggrep]

Sprint planning September 25, 2024

attendees: @Davide Silva , @Trevor Jacka

Story update

a.c. done:

• Deployed infernet node with infernet-compatible container
• Deployed Ritual router
• can deploy model to the container
• libraries specified/imported

a.c. outstanding:

• successful http request
• make an onchcain request from Base

planning outcome:

• moving this story to next sprint to address outstanding a.c.

[trvrggrep]

Sprint planning October 9, 2024

attendees: Davide & Trevor

a.c. done:

• successful http request
• make an onchain request from Base

[trvrggrep]

Testing plan:

• Trevor will lead testing

• Testing for the purpose of getting this story passed will consist of:

  • set up a cloud environment capable of running a hello world script
  • setup the model execution container (can I use the terraform?)
  • setup the ritual router (needed for every model?)
  • configure http server
  • deploy and configure model specific contract

• We will use aspects of Paper Aviation's use case for further testing and product discovery, specifically:

  • set up a cloud environment capable of running Llama 8B model
  • download Llama 3.0 8B
  • setup the model execution container (can I use the terraform?)
  • setup the ritual router (needed for every model?)
  • configure http server
  • deploy and configure model specific contract

[trvrggrep]

Sprint planning Oct 16 2024

• Will complete the above QA plan over sprint starting today. Our expectation is the QA plan moves #14, #16, #12, #17, and #4 tickets to complete status