https://github.com/deeplearnphysics/pytorch-uresnet
PyTorch Implementation of U-ResNet
https://github.com/deeplearnphysics/pytorch-uresnet
Last synced: about 1 month ago
JSON representation
PyTorch Implementation of U-ResNet
- Host: GitHub
- URL: https://github.com/deeplearnphysics/pytorch-uresnet
- Owner: DeepLearnPhysics
- Created: 2018-01-16T21:55:51.000Z (about 8 years ago)
- Default Branch: master
- Last Pushed: 2018-03-05T21:48:43.000Z (about 8 years ago)
- Last Synced: 2025-04-13T00:35:23.945Z (12 months ago)
- Language: Jupyter Notebook
- Size: 3.32 MB
- Stars: 7
- Watchers: 2
- Forks: 4
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# pytorch-uresnet
PyTorch Implementation of U-ResNet used for track/shower pixel-labeling
## Dependencies
* `ROOT`: data analysis framework. Defines file format, provides python bindings for our code
* `LArCV`: either version 1 and 2
* `pytorch`: network implementation
* `tensorboardX`: interface to log data that can be plotted with Tensorboard
* `tensorflow-tensorboard`: (from tensorflow) for plotting loss/accuracy
* `jupyter notebook`: for testing code and inspecting images
### Known working configuration
* ubuntu 16.10, ROOT 6, python 2.7.12, tensorflow-tensorboard (from tensorflow 1.4.1), cuda 8.0
## Files
* `uresnet.py`: module that defines the network
* `plotlarcv1.ipynb`: notebook used to test file IO commands. For LArCV classic files.
* `test.py.ipynb`: notebook used to test file IO commands. For LArCV2 files.
* `train_wlarcv1.py`: script for training.