Currently implementing the following tasks:
- Language Model
- Image Captioning
- Pixel CNN
- Conditional Image Generation (GANs)
- Text Summarization
Since many of the listed tasks can be accomplished by different models, we attempt to build an interface for each task, and users can pass in building blocks (such as a certain variants of CNN or RNN). The evaluation and experiments will be done automatically.
We rely on TFLearn for Image preprocessing and data preparation. Our biggest independent contribution is the interface to experiments/tasks.