Evaluate a trained checkpoint with visualization
Parse the user's request from $ARGUMENTS and run evaluation.
uv run python run_experiment.py eval --path <PATH> [OPTIONS...]
The user may provide:
--path /tmp/.../actor_999.pt)--path /tmp/.../26-03-07-00-26-36_cartpole/)--logdir /tmp/training_runs)If no path is given, check /tmp/training_runs for the most recent run.
Use Glob to verify the path exists and resolve it before running.
| Flag | Default | Description |
|---|---|---|
--ep-len | 10 | Episode length in seconds |
--seed | None | Random seed for reproducible eval |
--out-dir | None | Directory to save videos |