search

ITMO-NSS-team / torch_DE_solver

Combines power of torch, numerical methods to conquer and solve ALL {O,P}DEs
https://torch-de-solver.readthedocs.io
BSD 3-Clause "New" or "Revised" License
59 stars 9 forks source link

Computation simplification with precision reduction #27

Open SuperSashka opened 1 year ago

SuperSashka commented 1 year ago

In some computer vision applications it is viable to reduce precision from, for example, float64 to float32/float16 or even to int. Maybe, it would be more faster to compute first guess in, for example, float16 and move to float64 to find finer solution.

Cf. Lazy computation mode #25

SuperSashka commented 1 year ago

We got memory issue with tensors of size (52,125,125)~800k points. Gradients are of size 16GB. It seems that we have to somehow mitigate it. =)

SuperSashka commented 1 year ago

We need to up this issue and take this issue as the next target.

We have to make sure that we can set floating point precision as an argument.