integrate-pytorch-lightning

Fix Ruff formatting and lint errors in an explicit Python file or directory scope.

Improve docstrings and comments in an explicit Python file or directory scope after implementation is complete.

Define one concrete PyTorch Experiment with selected component paths, phases, configuration or custom phase code, and outputs, without executing it.

Acquire user-selected research materials from direct sources and create co-located Reference records with artifact provenance.

Analyze selected Reference records and their artifacts with evidence locators, implementation details, limitations, comparisons, and review-ready synthesis.

Search papers, code, datasets, documentation, and other research materials, then hand selected sources to add-references.

Define a reusable Benchmark component that evaluates model behavior with selected Dataset derivations, quantitative metrics, and qualitative visualizations.

Define a self-contained, reusable Dataset component, including its source revision and the rules that derive usable inputs.

Draft an evidence-linked report or paper from Reference analyses and experiment results without submission workflows.

Define and implement a self-contained, reproducible PyTorch model component from an aligned model design.

Initialize the current directory as a uv-managed Python project with the preferred configuration.

Run the phases declared by a fixed Experiment.