I have zero experience in web development so I decided to start a self project to learn some of the basics of front-end and back-end development and server management. So I created a home server to host machine learning capabilities.
I set up a Linux web server with an Angular front end that users can create an account and log in. From there they can use the web UI to design and create deep neural networks. Currently it supports autoencoder, fully connected, convolution, and a host of other network layers. The user can also upload data and train the network. The server hosts a nVidia GTX1080 GPU to train and run the network. The website was served using Flask and Tensorflow was used to create the networks. Successful models could also be deployed as an API endpoint.
My final solution wasn't very scalable but I learned a lot from it!
I set up a Linux web server with an Angular front end that users can create an account and log in. From there they can use the web UI to design and create deep neural networks. Currently it supports autoencoder, fully connected, convolution, and a host of other network layers. The user can also upload data and train the network. The server hosts a nVidia GTX1080 GPU to train and run the network. The website was served using Flask and Tensorflow was used to create the networks. Successful models could also be deployed as an API endpoint.
My final solution wasn't very scalable but I learned a lot from it!