| license: other | |
| library_name: docker | |
| tags: | |
| - robotics | |
| - reinforcement-learning | |
| - docker | |
| - x2 | |
| # AIGF X2 RL Containers (private backup) | |
| This private repository contains the four `linux/amd64` OCI images for the | |
| standalone X2 official 29DOF/151D pipeline: | |
| - `aigf/x2-rl-trainer:29dof-latest` | |
| - `aigf/x2-rl-exporter:29dof-latest` | |
| - `aigf/x2-rl-preparer:29dof-latest` | |
| - `aigf/x2-rl-console:29dof-latest` | |
| The images share one OCI blob store, so common CUDA/PyTorch layers are stored | |
| only once. After downloading the repository: | |
| ```bash | |
| tar -C oci-layout -cf - . | docker load | |
| docker image ls 'aigf/x2-rl-*' | |
| ``` | |
| The trainer image bundles NVIDIA Isaac Gym Preview 4. Keep this repository | |
| private and use the image subject to NVIDIA's applicable license terms. | |