Repository Structure

Repository Map

Pre-Training-and-Fine-Tuning/
├── README.md
├── qwen_alien_romulus_finetune.py        # exported runnable code from notebook
├── qwen_alien_romulus_finetune_fixed.py  # original source, unchanged
├── Qwen_Alien_Romulus_Fine_Tuning.ipynb # original notebook
├── LLM_Training_Fixed_v2.py             # legacy SFT/DPO/ORPO
├── LLM_Training_Fixed_v2.ipynb
├── requirements.txt                     # copied from uploaded 'requirements .txt'
├── requirements_legacy_llm_v2.txt
├── data/alien_romulus_qa.jsonl
├── original_sources/
│   ├── data.zip
│   ├── requirements .txt
│   └── README_uploaded.md
├── wiki/                                 # Wiki Markdown pages
├── diagrams/                             # Mermaid .mmd sources
└── scripts/                              # Wiki publishing tools

Provenance: Original Python and notebook files were preserved. A clean qwen_alien_romulus_finetune.py was exported from the existing Qwen notebook's Python definitions; the original script remains under its uploaded name. The original README was inaccurate about filenames and was moved to original_sources/ for reference. The original dataset archive is preserved without silently publishing its unrelated corpora as training inputs.