https://github.com/ahmedheakl/vgg-fine-grained-pruning

Channel and fine-grained pruning for VGG16 Network and analytics on model latency, computation and accuracy before and after pruning.
https://github.com/ahmedheakl/vgg-fine-grained-pruning

Last synced: 11 days ago
JSON representation

Channel and fine-grained pruning for VGG16 Network and analytics on model latency, computation and accuracy before and after pruning.

• Host: GitHub
• URL: https://github.com/ahmedheakl/vgg-fine-grained-pruning
• Owner: ahmedheakl
• License: mit
• Created: 2024-07-25T18:07:17.000Z (6 months ago)
• Default Branch: main
• Last Pushed: 2024-07-25T18:12:06.000Z (6 months ago)
• Last Synced: 2024-11-14T07:23:35.022Z (2 months ago)
• Language: Jupyter Notebook
• Size: 313 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# VGG Fine-grained Pruning

Channel and fine-grained pruning for VGG16 Network and analytics on model latency, computation, and accuracy before and after pruning.

## Analytics

Here are some analytics about the pruning.

### Weight distribution before fine-grained pruning

![image](https://github.com/user-attachments/assets/9adba591-b35f-4f1a-a6c7-312dd4790b95)

### Weight distribution after fine-grained pruning

![image](https://github.com/user-attachments/assets/b96d4d44-6945-437e-bba0-1ae85cc0bfe6)

### Weights before/after 

![image](https://github.com/user-attachments/assets/0e44866a-ec8d-4520-9549-2416a4f1fac2)

### Scan Analysis of multiple layers

![image](https://github.com/user-attachments/assets/af252904-6835-4440-975b-02efa94af6b5)

### Number of parameters per layer

![image](https://github.com/user-attachments/assets/d27c08a7-fff5-4510-b237-79ec849f68e1)