autokeras

Why we included this project

Most deep learning projects stall at the point where you have data but no clear network architecture. AutoKeras removes that bottleneck by automating architecture search and hyperparameter tuning on top of Keras, so you can go from raw data to a working model with a few lines of code like ak.ImageClassifier().fit(). It handles image, text, and structured data out of the box, which makes it a practical first stop for teams wanting to validate an idea quickly before investing in hand-tuned models. Because it stays close to Keras conventions and runs on TensorFlow, it slots into an existing stack without much friction. The built-in search over model structures and preprocessing is useful, though you give up fine-grained control over every layer if you need it.