dp-cupbasket-real

LeRobot diffusion policy (ROBOTIS lerobot-cyclo fork, 288 M parameters) for the bimanual cup handover onto a basket with the real ROBOTIS FFW SG2, trained on real demos only.

Files at the repo root = step 100k (DiffusionPolicy.from_pretrained("padiredd/dp-cupbasket-real")); step-10000/, step-20000/, step-30000/, step-40000/, step-50000/, step-60000/, step-70000/, step-80000/, step-90000/ = intermediate checkpoints (every 10k steps).

data	real teleop demos only: omkarpatil/pick-cup-and-place-on-basket, all 19 episodes (bimanual: the left hand pinches the blue cup's rim, hands it to the right hand, which places it on an upside-down woven basket; lab Cyclo converter, then each camera stretched to 240x240, cam_right_head dropped)
sampling	uniform over all frames (LeRobot default EpisodeAwareSampler)
inputs / outputs	head + left wrist + right wrist cameras, each 240x240, + 22-dim joint state; action = 16 joint targets, 15 Hz
image augmentation	heavy, random order, up to 5 of: brightness 0.6-1.4, contrast 0.6-1.4, saturation 0.3-1.7, hue +-0.1, sharpness 0.3-2.0, affine +-10 deg / 10 % / 0.9-1.1, perspective 0.2, blur, random erasing, grayscale
state noise	none
recipe	100k steps, batch 32, ResNet18 per camera, Adam 1e-4, cosine, DDPM 100 steps, horizon 64 / 32 executed, n_obs_steps 1, seed 1000
hardware	one NVIDIA RTX 4090 24 GB (lab225), shared by the two real-only runs trained at the same time

Task string: "pick up the blue cup with the left hand, hand it to the right hand, and place it on the basket". Evaluation results are added once they are run.