What would be valuable tools for deep learning that nobody has built yet?

As a programmer, I like to think of building deep learning models as programming and experimentation. All the tools that apply in those fields are very relevant to this space as well. TensorFlow for deep learning is like a language for programming. There’s still plenty of work to be done to build the rest of the ecosystem.

For example better debugging tools will help researchers understand why their models aren’t learning, better experimentation management will make it easier for them to run and analyze more experiments.

Previous Post
Next Post