[ML] Functional tests - support tiny trained models

[pheyos]

Summary

This PR adds support for importing tiny trained PyTorch models during functional tests.

Details

• Add six tiny PyTorch models (config, vocabulary, definition) to x-pack/test/functional/services/ml/resources/trained_model_definitions:
  • pt_tiny_fill_mask
  • pt_tiny_ner
  • pt_tiny_pass_through
  • pt_tiny_text_classification
  • pt_tiny_text_embedding
  • pt_tiny_zero_shot
• Add corresponding service methods to deal with the model files and to import them into the deployment
• Add tests to the model_list suite: import all 6 supported tiny models and checks row contents for all of them
• Some minor service method cleanup

Imported models during the test run look like this:

[elasticmachine]

Pinging @elastic/ml-ui (:ml)

[pheyos]

@elasticmachine merge upstream

[peteharverson]

Great job adding this in! LGTM

[jgowdyelastic]

Added a small comment but generally LGTM

[jgowdyelastic]

Latest changes LGTM

[kibana-ci]

💚 Build Succeeded

• Buildkite Build
• Commit: a063134
• Storybooks Preview
• Webpack Bundle Analyzer

Metrics [docs]

✅ unchanged

History

• 💚 Build #56531 succeeded abee011
• 💚 Build #56304 succeeded a06a9d8

To update your PR or re-run it, just comment with:
@elasticmachine merge upstream

cc @pheyos