Skip to content

Commit

Permalink
Remove references to Hessian being in next release
Browse files Browse the repository at this point in the history
In preparation for the release of v0.0.20, which includes the
new double backpropagation feature, enabling calculation of
the Hessian, this commit removes references to the Hessian being
in the next release/coming soon in the documentation.
  • Loading branch information
ar4 committed Sep 26, 2023
1 parent 06c0e0c commit 8b4a429
Show file tree
Hide file tree
Showing 3 changed files with 1 addition and 5 deletions.
2 changes: 1 addition & 1 deletion README.md
Original file line number Diff line number Diff line change
Expand Up @@ -20,7 +20,7 @@ To install it, I recommend first installing PyTorch using the instructions on th
- The PML width for each edge can be set independently, allowing a free surface (no PML) on any side
- Finite difference accuracy can be set by the user
- A region of the model around the sources and receivers currently being propagated can be automatically extracted to avoid the unnecessary computation of propagation in distant parts of the model
- Double backpropagation through the regular scalar propagator, including calculating the Hessian ([coming soon](https://ausargeo.com/deepwave/example_hessian))
- Double backpropagation through the regular scalar propagator, including [calculating the Hessian](https://ausargeo.com/deepwave/example_hessian)

## Quick Example
In a few lines you can make a velocity model, propagate a wave from a source in the top left corner to a receiver in the top right, calculate an objective function, and backpropagate to obtain its gradient with respect to the velocity.
Expand Down
2 changes: 0 additions & 2 deletions docs/example_hessian.py
Original file line number Diff line number Diff line change
@@ -1,5 +1,3 @@
# Double backpropagation to calculate the Hessian, used in this example, is not yet available in the released version of Deepwave. If you would like to beta test it, please write to me: [email protected]

import torch
import torchvision
from scipy.sparse.linalg import eigsh
Expand Down
2 changes: 0 additions & 2 deletions docs/example_hessian.rst
Original file line number Diff line number Diff line change
@@ -1,8 +1,6 @@
Hessian
=======

**Double backpropagation to calculate the Hessian, used in this example, is not yet available in the released version of Deepwave.** If you would like to beta test it, please `write to me <mailto:[email protected]>`_.

Deepwave supports backpropagating up to two times through the regular scalar propagator. One advantage of this is that it makes it possible to calculate Hessian matrices, which are used in optimisation methods such as `Newton-Raphson <https://en.wikipedia.org/wiki/Newton%27s_method_in_optimization>`_. I will demonstrate that in this example.

After setting up a simple two layer model with a single shot, we can calculate the gradient and Hessian of a loss function based on the output receiver data, with respect to the velocity model, using::
Expand Down

0 comments on commit 8b4a429

Please sign in to comment.