docs: migrate join-use-case API examples to snake_case (0.7.0)

[divyasinghds]

Summary

Sync the join-use-case pages with the snake_case API that landed in tracebloc_package 0.7.0. Split out from #13.

• Rename camelCase API methods to snake_case: upload_model, link_model_dataset, experiment_name, get_training_plan, learning_rate, loss_function, layers_freeze, early_stop_callback, reduce_lr_callback, model_checkpoint_callback, terminate_on_nan_callback, training_classes, data_type
• Rename trainingObject → training
• Correct terminate-on-NaN wording (any NaN loss, not just validation)
• Update pip install snippet to optional-extras syntax: tracebloc_package[pytorch|tensorflow|sklearn|all]

Test plan

• mint dev renders hyperparameters and start-training pages without MDX errors
• Code snippets match the current tracebloc-py-package 0.7.0 public API
• No remaining camelCase references on these two pages

🤖 Generated with Claude Code

---

Note

Low Risk
Low risk documentation-only updates, but incorrect method names or install extras would directly confuse users following the training guide.

Overview
Updates the Join Use Case docs to match the tracebloc_package 0.7.0 snake_case Python API by renaming trainingObject to training and converting all example calls (e.g., upload_model, link_model_dataset, get_training_plan, learning_rate, callbacks, and dataset config setters).

Also revises the installation snippet to use framework-specific pip extras (e.g., tracebloc_package[pytorch|tensorflow|sklearn|all]) and clarifies the terminate-on-NaN callback wording.

^{Reviewed by Cursor Bugbot for commit 5083264. Bugbot is set up for automated code reviews on this repo. Configure here.}