This repo contains a CNN for training CIFAR 10 Dataset.
model.py file contains the CNN Model. It has BaseNet model
Model Summary is as Follows -
| Name | Type | Params
------------------------------------
0 | prep | Sequential | 1.9 K
1 | c1 | Sequential | 74.0 K
2 | c2 | Sequential | 221 K
3 | c3 | Sequential | 295 K
4 | c4 | Sequential | 1.2 M
5 | c5 | Sequential | 3.5 M
6 | pool | MaxPool2d | 0
7 | fc | Sequential | 5.1 K
------------------------------------
5.3 M Trainable params
0 Non-trainable params
5.3 M Total params
21.279 Total estimated model params size (MB)
Model Accuracy:
- Training Accuracy: 90.4%
- Test Accuracy: 85.7%
Training Accuracy Graph:
Training Loss Graph:
Incorrect Predictions
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent