GraphCanon updated 3w · GitHub synced 3w
Decision brief
openpi is a specialized tool for model training, inference & serving that leverages advanced GPU capabilities and has specific requirements for memory and hardware configurations.
Good fit when
- When you have an NVIDIA GPU with at least 8 GB of VRAM for inference or at least 22.5 GB to fine-tune models using LoRA (Low-Rank Adaptation) on a single GPU.
- If you are working within Ubuntu 22.04 and require the ability to perform model parallelism by configuring `fsdp_devices`.
Avoid when
- When your project is not compatible with Ubuntu 22.04 or if you do not have access to a supported GPU configuration.
- If you need to support multi-node training, as this capability has yet to be implemented in the current version of openpi.
Observed Jul 12, 2026 · Source: enrich:decision_facts
Verify the decision
Maintenance and security
Full trust report- Maintenance
- Steady (46d since push)
- As of 3w
- Provenance
- Not a fork · Organization account
- As of 3w
- Security (OSV)
- No lockfile
- As of 1mo
Public GitHub metadata and optional OSV scans. Signals, not a guarantee. Trust methodology.
Install
pip install openpiSimilar tools
Same-category neighbours. No typed graph edges are catalogued for this tool yet.
Evidence and technical details
Sourced facts, taxonomy, compatibility claims, README excerpt, and machine-readable endpoints.
Overview
A collection of AI models that necessitate specific NVIDIA GPU configurations for inference and fine-tuning, including options for model parallelism. The setup is primarily tested under Ubuntu 22.04.
Capability facts
- Languages
- python
Source: github.language+pyproject.toml · Aug 2, 2026
Categories
Tags
README
Requirements
To run the models in this repository, you will need an NVIDIA GPU with at least the following specifications. These estimations assume a single GPU, but you can also use multiple GPUs with model parallelism to reduce per-GPU memory requirements by configuring fsdp_devices in the training config. Please also note that the current training script does not yet support multi-node training.
| Mode | Memory Required | Example GPU |
|---|---|---|
| Inference | > 8 GB | RTX 4090 |
| Fine-Tuning (LoRA) | > 22.5 GB | RTX 4090 |
| Fine-Tuning (Full) | > 70 GB | A100 (80GB) / H100 |
The repo has been tested with Ubuntu 22.04, we do not currently support other operating systems.
Installation
When cloning this repo, make sure to update submodules:
git clone --recurse-submodules git@github.com:Physical-Intelligence/openpi.git
For agents
This page has a .md twin and JSON over the API.