SO101 Hetero EE Action Decoder

Action decoder checkpoint for Mimic-Video SO101 hetero data.

  • Action space: relative end-effector space
  • Robot: SO101
  • Video input: front + wrist hstack, 5fps
  • Action decoder checkpoint: checkpoints/model/iter_000002750.pt
  • Initialized from Bridge action decoder
  • Uses SO101 video LoRA conditioning
Downloads last month
-
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support