AlphaHub

Please feel free to ask any questions here AlphaHub currently supports the following features: 🎯 Automatic shape inference for all layers, so you don't have to worry about tensor shape errors. 🎯 Visual debugging, as soon as a connection is invalid, you get custom solutions in real-time. 🎯 Possibility to create any kind of graph, as the platform is independent of sequential protocols. Go crazy! Upcoming features: 🎯 Custom layers and models can be added as a part of a graph in the canvas. 🎯 You get fully documented notebooks with standard data-pipelines, and your model compiled & loaded for training. Direct upload-and-run with Google Colab. 🔮 Quick fact: Time to design ResNet18 from scratch (without writing a single line of code) - 12 mins! I'm planning to beta launch for limited users so please don't hesitate to reach out. The website has a demo video based on long-term vision. An early preview of all the current features will be released later this week. You can help make AlphaHub the ideal visual programming platform 🏆