.gitignore

# Python
venv/
__pycache__/
*.py[cod]
*.pkl
*.egg-info/
dist/
build/

# Environment
.env

# Cache
.cache/
.pytest_cache/

# IDE
.vscode/
.idea/
*.swp

# OS
.DS_Store

# ML models (too large for git — use DVC or download separately)
ml/models/*.pkl
ml/models/*.bin
ml/models/*.pt
ml/models/*.safetensors
ml/models/xlmr_model/
serviceAccountKey.json
*.json.key
docs/*.json

# Dataset pipeline — raw downloads & processed parquet (regenerate via dataset_builder.py)
ml/data/raw/
ml/data/processed/
ml/_smoke_test.py

# Local history persistence (user data — do not commit)
data/history.json
ml/models/