jupyter-naas / naas-models Goto Github PK

We need to auto deploy naas-models version to pypi.

We need to create the Space model with it's validation logic. This also needs to build the proto files into the Python classes using the pydantic BaseModel (there is a protobuf to pydantic module out there).

You also need to setup Semantic release on this project, just for versioning, not pushing on pypi as it's a private repo.

You must create a Makefile and leverage docker to install protoc and needed software to build the proto

We need to have models being created for chats, models and messages

Now that the scheduler-v2-poc is done, we can proceed to the implementation of that in the api.

For that we need to create the Workflow.proto that will contain the resources needed to build a DAG Workflow that will run in Argo Workflow.

We should follow the following documentation from Argo: https://argoproj.github.io/argo-workflows/walk-through/the-structure-of-workflow-specs/

Resources that should be taken in account for creating the protos are:

• io.argoproj.workflow.v1alpha1.Workflow
• io.argoproj.workflow.v1alpha1.WorkflowSpec
• io.argoproj.workflow.v1alpha1.CronWorkflow
• io.argoproj.workflow.v1alpha1.WorkflowSpec
• io.argoproj.workflow.v1alpha1.Template

We must also be able to specify metadata but this will mainly be for us in the backend to pinpoint to which users are the workflow tied to etc.

As part of this you should also create all the CRUD messages (DTO) like WorkflowCreationRequest, WorkflowDeletionRequest etc ...

@viniciusdc please let me know if I am forgetting anything.

When I try to install naas-models, and therefore, naas-python, I get the following error:

I think that in naas_models/python/ we need to add a init.py file.