Numpy on the GPU? Speeding up Simple Machine Learning Algorithms with JAX

[@fancyfueko]

anyone interested in trying out JAX (or better yet, a functional approach to parallelization)?

[@glebbash]

Not really, I'm here for Context Free face reveal :)

[@fancyfueko]

😂 thanks for stopping by!

[@xelaxander]

Damn, I could have used this 4 weeks ago for an actual problem. Instead I calculated the derivatives the pedestrian way with pen and paper and wrote them in numba. Could've even used the GPU and saved myself some waiting.

[@xelaxander]

Also: Easy sensitivity analysis!!!

[@fancyfueko]

next time! but its good to calculate the derivatives by hand at least once in life 🙂