Understand the GPU acceleration in Tensordiffeq

[pescap]

Check how data is set over all the GPUs in Tensordiffeq.

[pescap]

MirrorStrategy in:

https://github.com/tensordiffeq/TensorDiffEq/blob/5f4b3df2da007cbf9a3b4316034aede2072580ac/tensordiffeq/models.py#L214

[pescap]

It would be interesting to test a few examples from tensordiffeq with an without data paralellization. Try to understand what changes when you add more GPUs!