Tupek/gretl refactor #1432
Open
Tupek/gretl refactor #1432
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
chapman39
reviewed
Aug 5, 2025
src/serac/gretl/CMakeLists.txt
Outdated
| @@ -0,0 +1,42 @@ | |||
|
|
|||
| set(smith_gretl_sources | |||
tupek2
force-pushed
the
tupek/gretl_refactor
branch
from
954952c to
9a67661
Compare
btalamini
approved these changes
Aug 27, 2025
src/serac/gretl/checkpoint.hpp
Outdated
| }; | ||
|
|
||
| /// @brief interface to run forward with a linear graph, checkpoint, then automatically backpropagate the sensitivities | ||
| /// given the reverse_callback jvp. |
Member
There was a problem hiding this comment.
Suggested change
| /// given the reverse_callback jvp. | |
| /// given the reverse_callback vjp. |
src/serac/gretl/checkpoint.hpp
Outdated
|
|
||
| namespace gretl { | ||
|
|
||
| struct Unit {}; |
white238
reviewed
Aug 29, 2025
| @@ -0,0 +1,10 @@ | |||
| set(gretl_test_depends serac_gretl gtest) | |||
Member
There was a problem hiding this comment.
Suggested change
| set(gretl_test_depends serac_gretl gtest) | |
| # Copyright (c) Lawrence Livermore National Security, LLC and | |
| # other Serac Project Developers. See the top-level LICENSE file for | |
| # details. | |
| # | |
| # SPDX-License-Identifier: (BSD-3-Clause) | |
| set(gretl_test_depends serac_gretl gtest) |
white238
reviewed
Aug 29, 2025
| @@ -0,0 +1,42 @@ | |||
|
|
|||
Member
There was a problem hiding this comment.
Suggested change
| # Copyright (c) Lawrence Livermore National Security, LLC and | |
| # other Serac Project Developers. See the top-level LICENSE file for | |
| # details. | |
| # | |
| # SPDX-License-Identifier: (BSD-3-Clause) | |
Codecov Report❌ Patch coverage is Additional details and impacted files@@ Coverage Diff @@
## develop #1432 +/- ##
===========================================
+ Coverage 92.03% 92.11% +0.07%
===========================================
Files 199 213 +14
Lines 25200 26142 +942
===========================================
+ Hits 23193 24080 +887
- Misses 2007 2062 +55
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This is an initial implementation of an automatically generated computational graph, based on recording the evaluation, storage, and dependencies between arbitrary types. Compared to other differentiable graph libraries, this one executes during runtime, and can branch dynamically. In addition, it is constructed such that it can be reversed (in the sense of adjoints or back-propagation), to compute the derivative of some final double qoi (quantity of interest) with respect to any input to the graph. Finally, these adjoints require storing (in general) the entire history of the graph. To avoid this, the graph automatically uses 'dynamic-checkpointing' (Minimal Repetition Dynamic Checkpointing Algorithm for Unsteady Adjoint Calculation, Wang, et. al. 2009) to store a near minimal number of states and automatically recomputes states in a near optimal way when back-propagating.