Markus Enzweiler

Support for pooling layers for CNNs, e.g. MaxPool2d

2

Thanks to the mlx team for creating and sharing mlx. I have managed to get a small CIFAR-10 image classification CNN up and running rather quickly in mlx (inspired by...

[BUG] cifar example fails with ValueError when replacing LayerNorm with BatchNorm with default params

6

When replacing LayerNorm with BatchNorm (merged in https://github.com/ml-explore/mlx/pull/217) with default params, the cifar example fails with ValueError with mlx 0.0.7: ``` File "../mlx-examples/cifar/main.py", line 48, in train_epoch mx.eval(model.parameters(), optimizer.state) ValueError:...

Updated CIFAR-10 ResNet example to use BatchNorm instead of LayerNorm

5

This PR updates the CIFAR-10 ResNet example to use BatchNorm instead of LayerNorm. It is tested with mlx@026ef9a but should probably wait until https://github.com/ml-explore/mlx/pull/385 (remove `retain_graph` flag) appears in an...

Example of a Convolutional Variational Autoencoder (CVAE) on MNIST

5

This PR introduces a (small) Convolutional Variational Autoencoder (CVAE) example on MNIST. Not sure whether this is interesting for mlx-examples or not. It is a port from https://github.com/menzHSE/torch-vae to mlx...