---
license: apache-2.0
tags:
- vision
- checkpoints
- residual-networks
pretty_name: Checkpoints
---

The Checkpoints dataset as trained and used in [A Symmetry-Aware Exploration of Bayesian Neural Network Posteriors](https://arxiv.org/abs/2310.08287) published at ICLR 2024. All models all trained and uploaded in a float16 format to reduce the memory footprint.

## Usage

### Untar the models

Just untar the desired models available in `models`, for instance with:

```bash
tar -xvf models/cifar10-resnet18/cifar10-resnet18-0-1023.tgz
```

Most of them are regrouped in tar files containing 1024 models each. This will create a new folder containing the models saved as safetensors.

### TorchUncertainty

To load or train models, start by downloading [TorchUncertainty](https://github.com/ENSTA-U2IS-AI/torch-uncertainty) - [Documentation](https://torch-uncertainty.github.io/).
Install the desired version of PyTorch and torchvision, for instance with:

```bash
pip install torch torchvision
```

Then, install TorchUncertainty via pip:

```bash
pip install torch-uncertainty
```

### Loading models

The functions to load the models are available in `scripts`. The script corresponding to Tiny-ImageNet also contains a snippet to evaluate the accuracy of a downloaded model.

**Any questions?** Please feel free to ask in the [GitHub Issues](https://github.com/ENSTA-U2IS-AI/torch-uncertainty/issues) or on our [Discord server](https://discord.gg/HMCawt5MJu).