MolmoBot-Pi0 โ€” reason-init action policies on real-DROID toys

Three MolmoBot-Pi0 (ฯ€0 / PaliGemma-backbone) action policies fine-tuned on a real-DROID tabletop pick-and-place set. Each is initialized from a different reasoning-only LoRA baked into the backbone before action training.

subfolder reasoning init
full/ all annotation modules
vqa_only/ free-text rationale only
trace_only/ end-effector trace only

Training

Three sequential stages; reasoning is not present at inference:

  1. Reasoning-only LoRA on the backbone, no action loss.
  2. Merge the LoRA into the MolmoBot-Pi0-DROID checkpoint.
  3. Action fine-tune with a flow-matching loss, vision tower frozen.

Action targets are absolute future joint positions (7 arm joints + gripper), horizon 16. Observations are two cameras, letterboxed to 224ร—224.

Contents

Each subfolder contains model.safetensors (flat ฯ€0 checkpoint), metadata.pt, and assets/ with the normalization statistics. Optimizer state is not included. The normalizer comes from the training set's statistics.

Limitations

Research checkpoints from an ablation study, released for reproducibility. They are trained on a single scene with one robot and are not tuned or benchmarked as a general-purpose policy; the three arms are not claimed to differ in downstream performance.

Downloads last month

-

Downloads are not tracked for this model. How to track
Video Preview
loading