Setup

python -m venv .venv
source .venv/bin/activate
pip install -e ".[dev]"
pre-commit install

Optional extras:

pip install -e ".[nlp]"
pip install -e ".[visualisation]"

Checks

Manual:

ruff check .
ruff format --check .
mypy insideLLMs
pytest

Or helpers:

make check
# or
bash scripts/checks.sh

Golden Path (Offline)

make golden-path

Or run the scripted workflow:

python examples/example_cli_golden_path.py

Run Artifacts Location

To keep outputs local to the repo:

INSIDELLMS_RUN_ROOT=.tmp/insidellms_runs insidellms run examples/experiment.yaml

Note: experiment tracking backends (W&B/MLflow/TensorBoard/local tracking logs) write to separate locations from the deterministic run artifacts. See Experiment Tracking.