domains / families / media / openpi

openpi

separate

Fine-tune and serve Physical Intelligence OpenPI models (pi0, pi0-fast, pi0.5) using JAX or PyTorch backends for robot policy inference across ALOHA, DROID, and LIBERO environments. Use when adapting pi0 models to custom datasets, converting JAX checkpoints to PyTorch, running policy inference servers, or debugging norm stats and GPU memory issues.

When to use

• Fine-tuning pi0, pi0-fast, or pi0.5 on LeRobot or RLDS datasets
• Serving OpenPI policies for ALOHA, DROID, or LIBERO evaluation
• Converting JAX checkpoints to PyTorch format
• Debugging OpenPI training issues (norm stats, memory, config)
• Fine-tuning OpenVLA with continuous action heads and LoRA
• Reproducing OpenVLA-OFT paper results on LIBERO or ALOHA
• Evaluating NVIDIA Cosmos Policy on simulation benchmarks

When not to use

• Fine-tuning OpenVLA with continuous action heads and LoRA
• Reproducing OpenVLA-OFT paper results on LIBERO or ALOHA
• Evaluating NVIDIA Cosmos Policy on simulation benchmarks

Skill page

Source