pi05_allslot_st3_s24000

Pi0.5 (pi05, the LeRobot/openpi port) policy checkpoint for EBiM Task 2 — Deformable Material Handling (Thermal Pad Placement).

• Model type: pi05 (paligemma_variant gemma_2b, action_expert gemma_300m, bfloat16, chunk_size 50, n_action_steps 50, num_inference_steps 10).
• Initialization: fine-tuned directly from lerobot/pi05_base (pretrained_path in config.json).
• Training data: 78-episode four-slot scripted-chain demonstrations of the Task 2 thermal-pad lay (local/task2_allslot_st3_s29), recorded with the benchmark's own LeRobot recorder (30 fps, eval_split 0.05, pyav backend).
• Steps: checkpoint at 24 000 optimization steps.
• Action space: 20-dim whole-body action row — base twist (vx, vy, wz), left + right FR3 7-DOF absolute joint targets, left + right gripper open-fraction, spine.height.target (see action_feature_names in config.json).
• Observation: 32-dim state + three 224×224 RGB cameras — observation.images.base_0_rgb, observation.images.left_wrist_0_rgb, observation.images.right_wrist_0_rgb. STATE/ACTION normalization: QUANTILES (pre/post processor pipelines are shipped beside the weights and must be loaded with them).
• Intended use: EBiM Task 2 evaluation. Load with LeRobot ≥0.6 as a pi05 policy — point the deployment config's vla.checkpoint at this directory (see the submission repo's "Fine-tuning on this task's own demonstrations" / LeRobotBackend, which loads the processor pipelines saved beside the weights so inference normalization is identical to training).

Files

file	role
config.json	pi05 policy config (features, variants, sampling)
model.safetensors	weights (9 354 050 752 bytes, md5 a437bf0c8669061923d4cfb4493a420f)
policy_preprocessor.json + policy_preprocessor_step_3_normalizer_processor.safetensors	input pipeline (quantile normalizer)
policy_postprocessor.json + policy_postprocessor_step_0_unnormalizer_processor.safetensors	output pipeline (unnormalizer)
train_config.json	full training configuration for reproducibility