policy-training

# policy-training Use this sub-skill when the task is about training or evaluating ACT++, CNNMLP, Diffusion Policy, or the VQ latent model that follows ACT training. ## Typical triggers - "Train ACT on sim_transfer_cube_scripted" - "Evaluate the best checkpoint" - "Why does policy.py fail to import?" - "What does --temporal_agg do?" - "How do I run the latent model training pass?" ## What this sub-skill covers - Step-based training and evaluation through the `imitate_episodes` workflow. - ACT, CNNMLP, and Diffusion policy wrappers. - The DETR-derived backbone / transformer / latent model internals that drive those wrappers. - Dataset loading, normalization, and evaluation statistics. - VQ latent model training from a previously trained ACT checkpoint. ## What it excludes - Simulated data generation, replay, and rendering -> [simulation-data](../simulation-data/SKILL.md). - VINN feature cache and k-selection -> [vinn-offline](../vinn-offline/SKILL.md). - Real robot deployment or ROS/servo control -> root troubleshooting only. - Experimental actuator-network training with hard-coded paths -> reference-only note only. ## Read these first - [Workflow recipes](references/workflows.md) - [Model overview](references/model-overview.md) - [Troubleshooting](references/troubleshooting.md) - [API reference](../../references/api-reference.md) - [Data formats](../../references/data-formats.md) ## Run this helper first Before trying to train, use [check_policy_stack.py](scripts/check_policy_stack.py) to confirm the repo checkout imports, the CUDA backend is visible, and the policy wrapper surface is present.