Markus Enzweiler

icon indicating a website link https://markus-enzweiler.de/

icon company Esslingen University of Applied Sciences icon location Esslingen, Germany icon location Professor | Computer Vision | Deep Learning | Robotics

Results 18 comments of Markus Enzweiler

Example of a Convolutional Variational Autoencoder (CVAE) on MNIST

- Tested with mlx 0.0.9 - Requirements and documentation updated - Style checked - Rebased and forced-pushed I think it is ready for review.

Example of a Convolutional Variational Autoencoder (CVAE) on MNIST

Tested with mlx-0.0.11, rebased and force-pushed

Example of a Convolutional Variational Autoencoder (CVAE) on MNIST

@awni Yes. I have made the minor changes requested above and it's ready for review.

Example of a Convolutional Variational Autoencoder (CVAE) on MNIST

Thanks @awni. I have also started a training run.

Example of a Convolutional Variational Autoencoder (CVAE) on MNIST

@awni Great! Thanks for merging. I am getting a throughput of approx. 485 im/s on my M1 which is quite a bit faster than 410 im/s with previous mlx versions....

Using Yolov8 and Anomalib with MLX

I guess you cannot use both at the moment with MLX. You would need to re-implement / port the models you need over to MLX.

Using Yolov8 and Anomalib with MLX

For MaxPooling, also have a look at https://github.com/robertmccraith/mimm/blob/main/mimm/layers/max_pool.py

Using Yolov8 and Anomalib with MLX

I have to pass on that, sorry. But I will follow your progress and am happy to try out and test your implementation.