openpi
openpi copied to clipboard
Physical-Intelligence
We fine-tuned the Pi0.5 diffusion model on our Lerobot dataset, which consists of videos showing a robotic arm picking up objects from a designated area and placing them into a...
`run_aloha_sim.sh` ``` Wrapper script that works around Docker Compose GPU bug Overrides CMD to use PYTHONPATH instead of uv run to prevent JAX CUDA package reinstallation Uses docker run --gpus...
I notice that the calvin dataset checkpoint is deleted, what is the reason behind it?
Based on the excellent pi0 work, we hope to be able to deploy it on our own robot arm. I am using Agilex Piper robot arm data to perform LoRA...
Hi there. I noticed that the jax version of pi0 does not support vectorized inference (say, batch size > 1). Does the latest pytorch version of these models work?
Hi authors, I am following official fine-tuning script, command `CUDA_VISIBLE_DEVICES=1 uv run python scripts/train.py pi05_robomimic_lift --exp-name=exp_pi05_robomimic_lift --overwrite` with my own config: ```python TrainConfig( name= "pi05_robomimic_lift", # name = "pi05_genesis", #...
The version of Lerobot is 0.1.0, which is too old and incompatible with the new v3.0 Lerobot dataset. Does OpenPI support a newer version of Lerobot?
Hi, I encounter this issue when I test the Pi0 and Pi0_fast training with the dataset I collected in reality. Running code XLA_PYTHON_CLIENT_MEM_FRACTION=0.9 uv run scripts/train.py 2025-07-18 11:07:06.132422: W external/xla/xla/tsl/framework/bfc_allocator.cc:501]...
I want to ask if there is an official practice of how to deploy PI0.5 on edge devices such as Nvidia Jetson Orin & Thor series? Many Thanks.
On Jetson AGX Orin, I set up the environment as stated and when I try to run serve_policy -libero, this error occurs. INFO:absl:[thread=MainThread] Failed to get flag value for EXPERIMENTAL_ORBAX_USE_DISTRIBUTED_PROCESS_ID....
