chore: add .gitignore

.gitignore

@@ -1,34 +1,76 @@
-# Data files (trop volumineux pour git)
-data/raw/*.pdf
-data/raw/*.json
-data/processed/*.jsonl
-data/processed/*.json
-data/datasets/*.jsonl
-data/datasets/*.json
-
-# Models
-models/
-
-# Python
+# === Python ===
 __pycache__/
-*.pyc
-.venv/
+*.py[cod]
+*.pyo
+*.egg-info/
+*.egg
+dist/
+build/
+*.whl
 
-# IDE
+# === Virtual environments ===
+.venv/
+venv/
+venv_*/
+env/
+
+# === ML Models & Data ===
+*.pt
+*.pth
+*.onnx
+*.bin
+*.safetensors
+*.h5
+*.hdf5
+*.pkl
+*.pickle
+*.npy
+*.npz
+*.faiss
+models/
+*.tar.gz
+*.zip
+
+# === Documents & Media ===
+*.pdf
+*.docx
+*.xlsx
+*.csv
+*.png
+*.jpg
+*.jpeg
+*.gif
+*.mp3
+*.wav
+*.mp4
+
+# === IDE ===
 .idea/
 .vscode/
+*.swp
+*.swo
+*~
 
-# Aim tracking
-.aim/
-
-# Unsloth compiled cache
-unsloth_compiled_cache/
-
-# llama.cpp build
-llama.cpp/
-
-# Runpod data copies
-runpod/data/
-
-# OS
+# === OS ===
 .DS_Store
+Thumbs.db
+.~lock.*
+
+# === Secrets ===
+.env
+*.env
+credentials.json
+token.pickle
+
+# === Logs & Cache ===
+*.log
+logs/
+.pytest_cache/
+.mypy_cache/
+.ruff_cache/
+htmlcov/
+.coverage
+
+# === Backups ===
+*_backup_*
+backups/