You can deploy this model and start earning money today!

l2aggle/
PPTmodel4UnitreeG1

Views

No views yet

PPTmodel4UnitreeG1

This is the PPTmodel4UnitreeG1 model presented in TrajBooster: Boosting Humanoid Whole-Body Manipulation via Trajectory-Centric Learning.

Project page: https://jiachengliu3.github.io/TrajBooster/ Code: https://github.com/jiachengliu3/OpenTrajBooster

This model is a post-pre-trained model specifically designed for Unitree G1 robot applications. The model has been fine-tuned using the Agibot2UnitreeG1Retarget dataset to enhance its performance on robotic whole-body manipulation.

Model Description

This model underwent post-pre-training using specialized robotics data to improve its understanding and generation capabilities for Unitree G1 humanoid robot applications. The training process leveraged the Agibot2UnitreeG1Retarget dataset, which contains motion retargeting data specifically curated for Unitree G1.

Dataset

The model was trained on the Agibot2UnitreeG1Retarget dataset, which provides comprehensive motion retargeting data for converting motion patterns to UnitreeG1 robot format.

Model Files

The model consists of the following files:

config.json - Model configuration
model.safetensors.index.json - SafeTensors index file
model-00001-of-00002.safetensors - Model weights (part 1)
model-00002-of-00002.safetensors - Model weights (part 2)
trainer_state.json - Training state information
training_args.bin - Training arguments
experiment_cfg/ - Experimental configuration files

Download and Usage

Method 1: Using Hugging Face Hub

python

from transformers import AutoModel, AutoTokenizer

# Download and load the model
model = AutoModel.from_pretrained("l2aggle/PPTmodel4UnitreeG1")
tokenizer = AutoTokenizer.from_pretrained("l2aggle/PPTmodel4UnitreeG1")

Method 2: Using Git LFS

bash

# Clone the repository
git clone https://huggingface.co/l2aggle/PPTmodel4UnitreeG1

# Navigate to the model directory
cd PPTmodel4UnitreeG1

Method 3: Direct Download

You can also download individual files directly from the model repository on Hugging Face.

Requirements

Python 3.7+
PyTorch
Transformers library
SafeTensors

Installation

pip install torch transformers safetensors

License

This model is released under the Apache 2.0 license.