Table 2 MNIST model summary.
From: Deep representation learning using layer-wise VICReg losses
Layer (type) | Output shape | Param # |
|---|---|---|
input_layer (InputLayer) | (None, 784) | 0 |
dense (Dense) | (None, 500) | 392,500 |
dense_1 (Dense) | (None, 400) | 200,400 |
dense_2 (Dense) | (None, 300) | 120,300 |
dense_3 (Dense) | (None, 200) | 60,200 |
Total parameters | 773,400 | |
Trainable parameters | 773,400 | |
Non-trainable parameters | 0 | |
