Download README.md from anschrur/dit_block_2: direct link, hf CLI and curl.
- Browser
- Download file 5.11 kB
-
https://huggingface.co/anschrur/dit_block_2/resolve/main/README.md
- Command line
-
hf download hf://anschrur/dit_block_2/README.md
-
curl -L -o README.md https://huggingface.co/anschrur/dit_block_2/resolve/main/README.md
pipeline_tag: robotics
tags:
- imitation-learning
- flow-matching
- diffusion-policy
- so-101
- robot-imitation-glue
dit_block_2 β DiT flow-matching policy, cup-into-ring
Single-camera Diffusion-Transformer flow policy for the SO-10x task pick up the cup, place it in the ring. Trained on the 2026-11-15 clean wrist retake of the cup dataset.
This is not a from_pretrained checkpoint. It is one raw torch.save dict written by
robot_imitation_glue.baseline.train_dit. lerobot's policy registry cannot open it β there is no
config.json / model.safetensors here. Load it through DiTAgent (below).
File
| file | bytes | md5 |
|---|---|---|
policy.ckpt |
66,705,733 | 669479fdbcfa196c7c251c4bf9e7c549 |
Contents: config, image_size, crop_ratio, model_state_dict (16,660,576 params),
proprio_normalizer_state_dict, action_normalizer_state_dict. The two normalisers ship with the weights β
the model is unusable without them, and nothing else (no dataset paths, no credentials) is inside the file.
Task interface
- State / action: 6-D absolute Feetech positions, dataset order
shoulder_pan, shoulder_lift, elbow_flex, wrist_flex, wrist_roll, gripper. Dims 0-4 are degrees; dim 5 is % open (the gripper servo is hard-wired toRANGE_0_100). Not one unit β do not scale the vector. - Vision: one camera,
observation.images.front, recorded 480Γ640 AV1 β resized to 128Γ128 with a 0.9 centre crop. - Chunk:
prediction_horizon = 48counts frames, and the dataset is 30 fps, so one chunk is 1.6 s of motion. A control loop must therefore run at 30 fps; at 10 fps the same chunk stretches to 4.8 s.
Architecture
| hidden dim | 256 |
| layers / heads | 4 / 4 |
| camera encoder | resnet18, ImageNet-pretrained, frozen statistics shipped as encoders.0.mean/std |
| prediction | flow matching, Euler sampler (10 steps is what the deployment script uses) |
| params | 16,660,576 |
Training
| dataset | anschrur/record-test_2_20261115_115045 β 30 episodes, 13,218 frames, 30 fps, 1 task |
| optimiser steps | 41,200 = 200 epochs Γ floor(13,218 / 64), drop_last=True |
| batch / optimiser | 64 / AdamW, lr 1e-4, weight decay 1e-4 |
| schedule | 500-step linear warmup β cosine to 1 % of base (final lr 1e-6), grad-clip 10, seed 42 |
| final train loss | 0.018505 (epoch 200/200) |
| hardware / wall clock | 1Γ RTX 5090 (imec iLab.t GPULab cluster 9), 1.7 h |
| written | 2026-11-15 13:28, /project_ghent/dit/outputs/dit_cup_2/policy.ckpt |
--save-every 20 rewrites the same policy.ckpt path, so these are the epoch-200 weights only β no intermediate
snapshots and no epoch sweep to plot. The dataset declares no held-out split, so train_loss is a training-set
number, not an evaluation result.
What changed against anschrur/dit-cup
Identical architecture and hyperparameters, with the epoch count raised 30 β 200 so the step totals match
(40,710 then vs 41,200 now); the dataset is the only variable. dit-cup trained on record-test_2, whose
wrist_roll labels carry a ~170Β° offset in 32 of 200 episodes. The retake removed it, and the difference shows up
in the action normaliser stored inside the file (degrees for dims 0-4, % open for dim 5):
| dim | dit-cup mean / std |
dit_block_2 mean / std |
|---|---|---|
| shoulder_pan | β2.94 / 18.42 | 0.79 / 16.15 |
| shoulder_lift | β20.40 / 46.84 | 79.98 / 51.82 |
| elbow_flex | 28.94 / 36.92 | β71.29 / 43.93 |
| wrist_flex | 60.48 / 14.90 | 73.85 / 13.39 |
| wrist_roll | β144.60 / 69.53 | 4.37 / 20.51 |
| gripper | 8.01 / 13.20 | 6.85 / 11.46 |
wrist_roll std falls 69.5 β 20.5 once the offset is gone. The other dims move too because the retake is a
different 30-episode recording, not a relabelled subset of the old one β those four joints are not directly
comparable.
Loading
from robot_imitation_glue.baseline.dit_agent import DiTAgent # plain clone on sys.path is enough
agent = DiTAgent(
checkpoint_path="policy.ckpt",
camera_key_mapping={"front": "observation.images.front"}, # live camera key -> policy key
state_key="observation.state",
device="cpu",
n_sampling_steps=10,
)
Closed-loop on an arm β the repo is public, so this needs no token:
python dit_on_robot.py --repo-id anschrur/dit_block_2 --fps 30 --dry-run # arm must NOT move on this one
dit_on_robot.py (thesis glue, not published here) wires DiTAgent to lerobot's SOFollower with a wrap-aware
per-step delta guard, because lerobot's own max_relative_target clamp mishandles the Β±180Β° wrist_roll boundary.
Not measured
No open-loop action MSE, no evaluation episodes, no success rate, and no head-to-head against dit-cup beyond the
normaliser stats above. The wrist_roll label defect in the previous dataset is the reason this run exists; whether
it improves closed-loop behaviour is what the next robot rollout has to answer.