[steramae-nvidia]

Disable templates when missing inference provider for build.nvidia.com since it won't be able to make inference requests for the LLM columns.

Signed-off-by: Sean Teramae steramae@nvidia.com

[steramae-nvidia]

This change is part of the following stack:

• feat(studio): Add column palette component #528
  • feat(studio): DAG for DD #529
    • feat(studio): add DataDesignerJobBuildRoute with CreateFilesetStart #530
      • feat(studio): polish DataDesignerJobBuildRoute columns and canvas #537
        • feat(studio): add model palette and config panel to DAG build route #546
          • feat(studio): Add column fields for samplers #630
            • feat(studio): More templates and visual fixes #647
              • fix(studio): DD embedding template fix #661
                • feat(studio): Support seeding DD #685
                  • fix(studio): Fix empty UI in DD details #693
                  • fix(studio): Disable templates if no provider (w/ hint) #711 ◀

_{Change managed by git-spice.}

[nv-odrulea]

Can we use a more generic validation for the models that does not force the user to depend on build.nvidia.com or other NVIDIA endpoints?

[steramae-nvidia]

Closing to support feedback. Will instead, allow the user to go through to the dag view without any providers added. We will instead show an error immediately on mount to alert the user that they need to set the model to a usable one.