Recreate workshop for CHTC (HTCondor)

[qualiaMachine]

Summary

• Adapts the entire GCP/Vertex AI workshop into a CHTC/HTCondor workshop for UW-Madison's Center for High Throughput Computing
• Replaces all 9 episodes: GCP concepts → HTCondor submit files, GPU Lab, DAGMan workflows, CHTC filesystems
• Adds new submit_files/ directory with HTCondor submit files, wrapper scripts, and parameter files
• Adds new helper scripts: preprocess_data.py, evaluate_model.py, aggregate_results.py
• Updates all learner materials (setup, reference, compute guide) and instructor notes for CHTC
• Preserves the same pedagogical structure (Carpentries Sandpaper format) and Titanic dataset examples

Episode mapping (GCP → CHTC)

#	Was (GCP)	Now (CHTC)
01	Overview of GCP for ML	Overview of CHTC for ML
02	Vertex AI Workbench notebooks	Connecting to CHTC (SSH, submit node)
03	GCS data storage	Data management (/home, /staging, SQUID)
04	Vertex AI custom jobs (XGBoost)	HTCondor training jobs (XGBoost)
05	Vertex AI GPU jobs (PyTorch)	CHTC GPU Lab jobs (PyTorch)
06	Vertex AI HP tuning	HTCondor queue + DAGMan HP tuning
07	RAG with Gemini on Vertex AI	RAG on CHTC (API-based + open-source options)
08	gcloud CLI workflows	Advanced HTCondor workflows (DAGMan)
09	GCP billing & cleanup	Resource management & best practices

Test plan

• Verify all 9 episodes render correctly in Sandpaper
• Test HTCondor submit files on a CHTC submit node
• Run training scripts (train_xgboost.py, train_nn.py) locally and via HTCondor
• Test HP tuning pipeline with params.txt
• Verify DAGMan workflow example
• Test RAG pipeline with API key setup
• Review all episodes for accuracy of CHTC-specific details (file paths, GPU Lab settings, runtime limits)

https://claude.ai/code/session_013kTNZ8Y4fXeTMqgJSUhEnn

[github-actions]

🆗 Pre-flight checks passed 😃

This pull request has been checked and contains no modified workflow files or spoofing.

Results of any additional workflows will appear here when they are done.