Update README.md

This is a simple CNN model to classify images created using pytorch to classify images from the 'aiornot' dataset. It tries to classify images as being generated by AI or not until a reasonable accuracy is reached. The model is comprised of:

1) Four CNN Conv2D layers with a pooling layer max_pool2D added between each layer
2) Two fully connected layers fc, with a dropout layer between the two layers
3) Softmax is used as the activation function for the final layer of the binary classifier
4) PyTorch allows us to create a custom class with nn.Module

The model was trained for 50 iterations using the Adam optimizer & the BCE loss function. The learning rate will be adjusted on the fly using ReduceLROnPlateau, with a factor of 1/2. The images were normalized using the Min-Max normalizer before being fed to the model.

README.md

@@ -1,3 +1,7 @@
 ---
 license: apache-2.0
----
+datasets:
+- competitions/aiornot
+metrics:
+- accuracy
+---