Downloads · 30 days
16
40% of all-time downloads
pi05-openarm-rh56f1-ablation-firstkf-realsynth-20k
RyanL22/pi05-openarm-rh56f1-ablation-firstkf-realsynth-20k
pi05-openarm-rh56f1-ablation-firstkf-realsynth-20k is a robotics model from RyanL22. Use it for the robotics task on the model card, and read the license before you ship it in a product. It is set up for lerobot. The card lists the license as apache-2.0.
LeRobot-native pi05 (v0.6.1) fine-tuned on real teleop v4 (261 ep, 20 Hz) + 324 synthetic RH56F1 episodes generated from the FIRST keyframe only, the rest chained from generated frames (first-keyframe-only ablation);…
All-time downloads
40
Public
Parameters
4.1B
9.4 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors9.4 GB · 100%
Parameter types
How the weights are stored.
BF163.6B · 87%
Base models
Model card
From the Hugging Face model README
pi0.5 ablation (firstkf) — OpenArm + RH56F1, real teleop + ablation synthetic data, step 20k
LeRobot-native pi05 (v0.6.1) fine-tuned on real teleop v4 (261 ep, 20 Hz) + 324 synthetic RH56F1 episodes generated from the FIRST keyframe only, the rest chained from generated frames (first-keyframe-only ablation); synthetic actions = depth-IDM labels with HaWoR wrist IK (same rule set as the 0916 wristIK data).
Synthetic set: 8 tasks (airfryer_2 27, ball 43, bottle 60, bottle_over_shelf 47, bottle_pour_2 44, box 45, coffee_pot 25, doll 33), 82392 frames, 20 fps, 512x288.
Merged: 4 real cells + 8 synthetic cells = 12 cells, shares ∝ sqrt(frames).
The comparison point is the full pipeline (skeleton-flow conditioned, all keyframes) trained the same way: RyanL22/pi05-anyh2r-rh56f1-0916-wristik-30k (and -15k).
Checkpoint: step 20,000 (final) of a 20,000-step run (lab-gpu26, 2x H100, batch 32 x 2 = 64).
Training settings
| value | |
|---|---|
| synthetic action labels | depth IDM + HaWoR wrist IK (0.3 Hz offset filter, excursion reject) |
| vision encoder (SigLIP, 412.4M) | fine-tuned (not frozen) |
| image augmentation | photometric + affine, one draw replayed across the stereo pair |
| cell shares | ∝ sqrt(frames), 12 cells |
| mirror augmentation | off |
| optimizer | AdamW, peak lr 2.5e-5, cosine decay, warmup 1000 |
| precision | bfloat16, gradient checkpointing |
| chunk | 50 actions @ 20 fps (2.5 s), n_obs_steps=1 |
| normalization stats | quantiles; constant dims (head pitch/yaw) widened to mean ± 0.1 rad |
Inputs
observation.images.base_0_rgb<- left ZED view, 288x512;observation.images.left_wrist_0_rgb<- right ZED viewobservation.state/action— 28 dims:neck(2) | left_arm(7) | right_arm(7) | left_hand(6) | right_hand(6)- Rollout: feed the policy state neck as 0.889 / 0.001 (training constant); the synthetic start pose is the same as the 0916 set.
Load
from lerobot.policies.pi05.modeling_pi05 import PI05Policy
policy = PI05Policy.from_pretrained("RyanL22/pi05-openarm-rh56f1-ablation-firstkf-realsynth-20k")
- Task
- Robotics
- Library
- lerobot
- License
- apache-2.0
- Created
- Sep 22, 2026
- Updated
- Sep 22, 2026