| license: mit | |
| tags: | |
| - diffusion-policy | |
| - lerobot | |
| - safetensors | |
| - so101 | |
| # Diffusion Policy: so101-block-horizontal-layComb12 | |
| Pretrained Diffusion Policy trained using the [LeRobot](https://github.com/huggingface/lerobot) framework. | |
| ## 📦 Checkpoint Info | |
| - Format: `safetensors` | |
| - Model: Diffusion policy | |
| - Trained on: `so101-block-horizontal-layComb12` dataset | |
| - Steps: 25,000 | |
| - Final loss: ~0.013 | |
| - Config: `train_config.json`, `config.json` | |
| ## 🔧 How to Load | |
| ```python | |
| from huggingface_hub import hf_hub_download | |
| import safetensors.torch as storch | |
| import torch | |
| # Load model weights | |
| model_path = hf_hub_download("tshiamor/diffussion_so101-block-horizontal-layComb12", "model.safetensors") | |
| state_dict = storch.load_file(model_path) | |
| # You can now load state_dict into your model architecture | |