ACT · dvla-can 250 Hz (delta + trim) — checkpoint 50k

Results (clean in-domain scene set, n=20)

• Repeat variance on the same checkpoint is ±2–3/20 (n=20 eval) — treat 60% as a point estimate.
• Scene set = the corrected in-domain 20 scenes (post contamination fix; membership-checked against the training meta/camera.jsonl).

Contents

050000/pretrained_model/   # lerobot ACT weights + config (inference)
050000/training_state/     # optimizer state (resume)

Eval protocol (matters — measured pitfalls)

• Scene flags OFF: this model predates the mo5 textured scene. Evaluate on the ORIGINAL plain scene (BASKET_SCENE/TEXTURED_OBJS/CAN_TEXTURE all unset) or renders will not match training pixels (policies are render-sensitive: 5% vs 33% across GL backends).
• Eval server on OSMesa with 240 s episode timeout (EGL aborts when GPUs are busy).
• control_freq 250, delta actions, trim-hold preprocessing — as in the dataset name.
• Loader: the DynamicVLA_mujoco patched lerobot (temporal-fusion + n_obs guard patches); see github.com/mickeykang16/DynamicVLA branch mujoco.

Provenance