Different strokes for different folks, man. Help contribute to the research/dev community by helping affect the research/dev community where it matters.
"But they are focused on deep neural networks" - No. They're really not. If anything, they're focused on making matrix operations faster and faster; where the natural result is deeper and deeper networks. But, why not…
Yawn. Instead of making brand new neural network libraries, why not help contribute to already mature libraries?
Different strokes for different folks, man. Help contribute to the research/dev community by helping affect the research/dev community where it matters.
"But they are focused on deep neural networks" - No. They're really not. If anything, they're focused on making matrix operations faster and faster; where the natural result is deeper and deeper networks. But, why not…
Yawn. Instead of making brand new neural network libraries, why not help contribute to already mature libraries?