-
Notifications
You must be signed in to change notification settings - Fork 1.1k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Laromance model in NEML2 #29775
base: next
Are you sure you want to change the base?
Laromance model in NEML2 #29775
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is not set up at all.
verbose = true | ||
device = 'cpu' | ||
|
||
moose_input_types = 'MATERIAL MATERIAL MATERIAL MATERIAL MATERIAL MATERIAL MATERIAL POSTPROCESSOR POSTPROCESSOR' |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is not set up correctly because I am getting the error:
The NEML2 model requires a stateful input variable `old_state/cell_dd`, but its state counterpart on the output axis has not been retrieved by any object. Therefore, there is no way to properly propagate the corresponding stateful data in time. The common solution to this problem is to add a NEML2ToMOOSE retriever such as those called `NEML2To*MOOSEMaterialProperty`.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think you need to save the corresponding neml2 output variable into moose. The output variable should be state/cell_dd
.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks Gary, This fixed the input file problem. I updated the input file. I'm going to work on the problem with the batch sizes now.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Gary gave me a fix for the batch size problem I had. The fix goes in NEML2. It is from linear algebra library changing between libtorch versions. I am using pytorch version 2.1.2.post4 and I need this fix. Gary is using 2.5.1 and does not need the fix.
@dschwen @hugary1995 This is the LAROMANCE implementation in NEML2. I am still converting the radial return mapping input file over to the new syntax. I left a comment about the error I'm getting. |
Job Documentation, step Docs: sync website on a8ec869 wanted to post the following: View the site here This comment will be updated on new commits. |
1e8c32c
to
0a04ca4
Compare
This is based on a grade 91 model from LANL on 2/1/2025. The implementation uses a dummy inteprolation grid, values, and transforms and does not contain any real material data. The actual grade 91 data is in Bison. This implements multilinear interpolation. There are 6 interpolation axes and 3 values at each node. The interpolation axes and interpolated values are transformed. This contains examples where the model only performs interpolation and a full example with radial return done in NEML2. The radial return mapping requries a change to NEML2 in math.cxx that Gary is including in his neml2 udpate closes idaholab#29774
a8ec869
to
e910ca3
Compare
closes #29774