This Python project implements a Robotic Transformer model using PyTorch. The Robotic Transformer is a deep learning architecture designed for robot perception and control tasks. It leverages the power of self-attention mechanisms to process sequences of sensor data efficiently.
Before using this project, you should have the following dependencies installed:
- Python 3.11+
- Conda
- Poetry
You can install these packages using pip:
conda create -n rt1-torch python=3.11 -y
conda activate rt1-torch
pip install poetry -U
-
Clone the repository:
git clone git@git.ar.int:dev/ai/ml/agilebrain/lrm/rt1-torch.git
-
Change directory to the project folder:
cd rt1-torch
-
Install the project dependencies:
poetry install
To use the Robotic Transformer model in your own Python code, you can import it as follows:
from robotic_transformer import RT1
# Create an instance of the RoboticTransformer model
model = RT1(num_actions=7, backbone_name='efficientnet-b3').to(device)
# Use the model for your robotic perception or control task
You can adjust the hyperparameters (num_actions
, backbone_name
etc.) to suit your specific application.
This project also includes sample code for training the Robotic Transformer model. You can train the model using your own dataset by modifying the training script and data loaders. To start training, run:
python robotic_transformer/workflow/train.py
Make sure to customize the data loading and training settings in the train.py
script for your specific use case.
Once you have a trained model, you can use it for inference on new data. Modify the inference script to load your trained model weights and perform inference on your robot's sensor data.
To run inference:
python robotic_transformer/workflow/inference.py
If you'd like to contribute to this project, please follow these steps:
- Fork the repository.
- Create a new branch for your feature or bug fix:
git checkout -b feature/my-new-feature
. - Make your changes and commit them:
git commit -m 'Add new feature'
. - Push to the branch:
git push origin feature/my-new-feature
. - Create a pull request on the original repository.
Please ensure your code follows the project's coding style and includes appropriate documentation.
This project is licensed under the MIT License - see the LICENSE file for details.
Happy Robotics with Robotic Transformer! If you have any questions or encounter any issues, feel free to open an issue on the GitHub repository.