Skip to content

Releases: pvti/NORTON

Release v0.2.0 - Checkpoints for The Ablation Study

01 Feb 08:50
@pvti pvti
v0.2.0
a815f25
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.
Compare
Choose a tag to compare
Release v0.2.0 - Checkpoints for The Ablation Study Latest
Latest

📦 Supplemental Checkpoints
We've added supplemental checkpoints for the ablation study on the role of the rank. These checkpoints provide additional insights into the model's behavior under different rank configurations.

:octocat: Usage

python evaluate.py --ckpt vgg_16_bn_[0.]*100_6.pt -r 6
02/01 09:40:54 AM | args = Namespace(data_dir='~/data', arch='vgg_16_bn', ckpt='app_err/vgg_16_bn/vgg_16_bn_[0.]*100_6.pt', batch_size=256, gpu='0', rank=6, compress_rate='[0.]*100')
Loading data:
Files already downloaded and verified
Files already downloaded and verified
02/01 09:40:55 AM | Loading checkpoint
02/01 09:40:55 AM | Evaluating model:
02/01 09:40:58 AM |  * Acc@1 93.450 Acc@5 99.680

🚀 Future Updates:
We are committed to transparency and reproducibility. Continuous improvements and expansions to our model collection are underway. Stay tuned for more models and fine-tuned variations based on community feedback and demands.

Once again, thank you for your interest in our project. We hope these supplemental checkpoints empower you to accelerate your projects and research in the domain of network compression.

Happy coding and exploring the world of efficient deep learning!

The NORTON Team

Full Changelog: v0.1.0...v0.2.0

Assets 9
Loading

Release v0.1.0 - Checkpoints

28 Jul 12:52
@pvti pvti
v0.1.0
bd25bea
Compare
Choose a tag to compare
Release v0.1.0 - Checkpoints

We are thrilled to announce the first official release of our project, featuring a collection of baseline and compressed checkpoints to support efficient network compression.

Available Checkpoints:

  1. VGG-16-BN/CIFAR-10
  2. ResNet-56/110/CIFAR-10
  3. DenseNet-40/CIFAR-10
  4. ResNet-50/Imagenet
  5. Faster/Mask/KeypointRCNNResNet50FPN/COCO-2017

Usage
To get started with these checkpoints, simply refer to the documentation. Each checkpoint must be loaded with its corresponding decomposition rank and pruning ratio.

Feedback and Contributions:
We value your feedback and contributions to this project. If you encounter any issues, have suggestions for improvements, or would like to contribute models, please don't hesitate to open an issue or submit a pull request on our GitHub repository.

Future Updates:
We are dedicated to the continuous improvement and expansion of our model collection. Keep an eye on our repository for future updates, as we plan to add more models and fine-tuned variations based on community feedback and demands.

Once again, thank you for your interest in our project. We hope these checkpoints empower you to accelerate your projects and research in the domain of network compression.

Happy coding and exploring the world of efficient deep learning!

The NORTON Team

What's Changed

  • Add Faster/Mask/Keypoint-RCNN by @pvtien96 in #1

New Contributors

  • @pvtien96 made their first contribution in #1

Full Changelog: https://github.com/pvtien96/NORTON/commits/v0.1.0

Contributors

pvti
Loading