← Questions
Sharded matrix multiply: forward, backward and planted bugs
0
I was asked this
No reports yet. Sign in to say where you were asked it.
Split a matrix multiplication across several devices and explain how the pieces combine. Write the forward pass and the gradient computation for that layout in NumPy or PyTorch. Then a sharded implementation with planted mistakes is handed over: fix it, using a single-device run as the reference for the gradients.
No official answer yet. Write yours on the right and run it.
Nothing published yet.
Sign in to submit your code; every attempt is kept here.
Write and test your solution herePython in your browser, the question's tests level by level, every submission kept for you.Sign in to start coding