You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more
It would be fascinating to instrument deep learning developer workflows, so as to run the numbers on the productivity increase that comes from using tools that reduce cognitive load vs. tools that lay traps for you to fall into
The first thing that users that recently switched to Keras mention to me is the productivity boost. In the time it would have taken them to debug their way through the implementation of one idea (`zero_grad()` anyone?), they can try out 2, 3 ideas.
It matters because trying more ideas (with fewer mistakes) means you will converge faster towards better ideas (thus winning competitions more often or increasing your paper acceptance rate). I'm thinking Kaggle kernels or Colab would be a way to gather hard data on this...
Hard to remove newbie correlation with poor tooling though ?
I think it’s all about there being enough documentation that highlights what a productive work flow looks like. This allows developers to iterate on what has worked well for others to find what works best for them. Your book Deep Learning with R/Python is a great example imo.
I found using high level framework ultimately you ran into a case where it doesn’t quite support it with ease. This is where a flexible versatile framework can distinguish itself. Easy of extension and good doc r two key things.
In my own experience, productivity goes down in proportional to how novel your idea is. As you are get close to bleeding edge, you have to go under the hood, do lot of hacks, etc. A hard to extend framework will really drag you down. So it depends on where u “sample” this.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
