niejunnan25/pi05-robotwin

π₀.₅ RoboTwin Checkpoints for PACE

Fifty task-specific π₀.₅ checkpoints used in PACE: Phase-Aware Chunk Execution for Robot Policies with Action Chunking (paper). Each task has its own model and normalization statistics. PACE is a training-free execution method: Fixed-horizon, PACE and the adaptive baselines in the paper use the same task-specific model weights. PACE calibration results and evaluation code are distributed with the companion codebase.

Contents

pi05-robotwin/
  manifest.json
  tokenizer.model
  beat_block_hammer/
    params/
    assets/robotwin/norm_stats.json
  ...49 other tasks...

Select a task using its standard RoboTwin name. A single task is approximately 12.4 GB; the complete collection is approximately 622 GB. manifest.json lists every task, its OpenPI configuration, asset files, sizes and SHA256 checksums. Model parameters retain the JAX/Orbax OCDBT format. Normalization statistics are the original task-specific statistics; do not replace them with pooled values.

Training

For each task, the pretrained π₀.₅ model was fine-tuned on 50 demo_clean demonstrations. All parameters were updated using AdamW, global batch size 32 and training seed 42. The learning rate warmed up to 2.5e-5 over 1,000 steps, then followed a cosine schedule toward 2.5e-6. These are the checkpoints saved at step 30,000, using EMA weights with decay 0.99.

The policy consumes three CHW-format uint8 camera images and a 14-dimensional dual-arm joint/gripper state. It returns a 50 × 14 chunk of absolute joint/gripper actions. The normalization and action transforms are supplied by the corresponding pi05_robotwin_<task> configuration.

Load with OpenPI-PACE

Use the accompanying OpenPI-PACE source and its installation instructions. The configurations were verified with source revision ba9876639cb7a4bf0c7db00ec39eeb9c8efd8b12. No training demonstrations are required to load these checkpoints for inference.

After downloading one task and preparing the tokenizer cache with the companion pace/scripts/download_model.py utility, run the official OpenPI server from the OpenPI-PACE repository root. Replace the example paths with your download and cache locations:

OPENPI_DATA_HOME=/path/to/openpi-cache CUDA_VISIBLE_DEVICES=0 \
python scripts/serve_policy.py --port 8000 policy:checkpoint \
  --policy.config pi05_robotwin_beat_block_hammer \
  --policy.dir /path/to/pi05-robotwin/beat_block_hammer

The tokenizer cache contains big_vision/paligemma_tokenizer.model, copied from this repository's tokenizer.model. The helper checks its hash and prints the exact server command. Fixed horizons and PACE are selected by the RoboTwin client; the model server always returns a complete predicted chunk.

These checkpoints cover the paper's RoboTwin2.0 task-specific models, not a single jointly trained 50-task policy or the separate LIBERO/real-robot models.

Terms and attribution

See TERMS.md and NOTICE for the upstream model and tokenizer terms. The OpenPI source-code license and the model terms are separate.