search

elwinter / nnde

nnde is a package of Python modules which implement a set of neural networks which solve ordinary and partial differential equations. This code is based on the previous package, nnode.
GNU General Public License v3.0
7 stars 5 forks source link

Paper Feedback: Analytic methods for derivative calculations #16

Closed hayesall closed 3 years ago

hayesall commented 3 years ago

The "Statement of need" (lines 19-22) mentioned that existing deep learning frameworks can speed up computation, but cannot perform analytical derivatives needed for the trial function method implemented by nnde.

Screenshot of the paper, where it shows something similar to the text summarized above.

Can this point be clarified or expanded?

For example: PyTorch implements methods for automatic differentiation (autograd) and numerical methods to compute gradients. What is an example where these are insufficient?

taless474 commented 3 years ago

Would you please explain about extra functionalities of nnde with this version of Autograd (not the PyTorch's one) and JAX considering JOSS guidelines?

elwinter commented 3 years ago

We updated the paper to address this issue by describing how the code was started before the widespread adoption of TensorFlow 2. We began to convert the software to use TensorFlow 2 several months ago, and early results are promising, so we wanted to document this particular stage in the project prior to further development.