File size: 2,600 Bytes
0046a01 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 |
---
license: apache-2.0
pipeline_tag: robotics
library_name: transformers
---
# PPTmodel4UnitreeG1
This is the PPTmodel4UnitreeG1 model presented in [TrajBooster: Boosting Humanoid Whole-Body Manipulation via Trajectory-Centric Learning](https://huggingface.co/papers/2509.11839).
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](https://huggingface.co/datasets/l2aggle/Agibot2UnitreeG1Retarget) 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](https://huggingface.co/datasets/l2aggle/Agibot2UnitreeG1Retarget), 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](https://huggingface.co/l2aggle/PPTmodel4UnitreeG1/tree/main) on Hugging Face.
## Requirements
- Python 3.7+
- PyTorch
- Transformers library
- SafeTensors
## Installation
```bash
pip install torch transformers safetensors
```
## License
This model is released under the Apache 2.0 license. |