Downloads · 30 days
0
lair-nyu/yor_icl_pi05_easy_pnp_v2_absolute_joint_sanity15k
yor_icl_pi05_easy_pnp_v2_absolute_joint_sanity15k is a machine learning model from lair-nyu. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
pi0.5 sanity check (4-task pnp subset, 15k steps, batch 128, 2xH200) -- absolute joint-state action space: both observation and action are icl-dataset's 15-dim observation.state (left/right arm joints + lift), instead…
Downloads · 30 days
0
Access
Public
Updated Sep 8, 2026
Repo size
12.4 GB
Likes
0
Public
Click a slice to open those files.
Other12.4 GB · 100%
From the Hugging Face model README
pi0.5 sanity check (4-task pnp subset, 15k steps, batch 128, 2xH200) -- absolute joint-state action space: both observation and action are icl-dataset's 15-dim observation.state (left/right arm joints + lift), instead of the usual yor_icl_* arms' 20-dim EE-pose action. The action is the future window of observation.state itself (icl-dataset never recorded a joint-space action).
yor_icl_pi05_easy_pnp_v2_absolute_joint_sanity15kparams/ (deployable policy weights) + assets/ (norm stats etc.) only -- train_state/ (optimizer state, needed only to resume this exact training run) is NOT included.