PhysicsAI got an error

Unknown
Unknown
edited October 2023 in Community Q&A

I'm using hyperworks 2022.3 and I got a license.
I try to use physicsAI with demo files(HVAC), but I got an error.
What should I do?

[17:05:40] (INFO): --------------------------------------------------------------------------------------------------
[17:05:40] (INFO):   1. Building features and labels
[17:05:40] (INFO): --------------------------------------------------------------------------------------------------
[17:05:47] (INFO): Node features:
[17:05:47] (INFO):     name: cae.coord
[17:05:47] (INFO):     type: CONTINOUS
[17:05:47] (INFO):     length: 3
[17:05:47] (INFO): 
[17:05:47] (INFO):     name: cae.part_id
[17:05:47] (INFO):     type: CATEGORICAL
[17:05:47] (INFO):     length: 2
[17:05:47] (INFO): 
[17:05:47] (INFO): Edge features:
[17:05:47] (INFO):     name: cae.direction
[17:05:47] (INFO):     type: CONTINOUS
[17:05:47] (INFO):     length: 4
[17:05:47] (INFO): 
[17:05:47] (INFO): Node labels:
[17:05:47] (INFO):     name: cae.results
[17:05:47] (INFO):     subcase: Flow Solution
[17:05:47] (INFO):     field: pressure
[17:05:47] (INFO):     type: CONTINOUS
[17:05:47] (INFO):     length: 1
[17:05:47] (INFO): 
[17:05:47] (INFO): --------------------------------------------------------------------------------------------------
[17:05:47] (INFO):   2. Transforming training data
[17:05:47] (INFO): --------------------------------------------------------------------------------------------------
[17:05:49] (INFO): Transform: radioss_rigid_rename
[17:05:49] (INFO):     Feature scalers:
[17:05:49] (INFO):         Identity
[17:05:49] (INFO):         RadiossRigidRename
[17:05:49] (INFO):         Identity
[17:05:49] (INFO): 
[17:05:49] (INFO):     Label scalers:
[17:05:49] (INFO):         Identity
[17:05:49] (INFO): 
[17:05:49] (INFO): Transform: binarize_categorical
[17:05:49] (INFO):     Feature scalers:
[17:05:49] (INFO):         Identity
[17:05:49] (INFO):         MultiLabelBinarizerStrict
[17:05:49] (INFO):         Identity
[17:05:49] (INFO): 
[17:05:49] (INFO):     Label scalers:
[17:05:49] (INFO):         Identity
[17:05:49] (INFO): 
[17:05:49] (INFO): Transform: filter_const_comps
[17:05:49] (INFO):     Feature scalers:
[17:05:49] (INFO):         ConstantCompFilter
[17:05:49] (INFO):         ConstantCompFilter
[17:05:49] (INFO):         ConstantCompFilter
[17:05:49] (INFO): 
[17:05:49] (INFO):     Label scalers:
[17:05:49] (INFO):         ConstantCompFilter
[17:05:49] (INFO): 
[17:05:49] (INFO): Transform: cntr_vec_uni_edges_stnd10_res
[17:05:49] (INFO):     Feature scalers:
[17:05:49] (INFO):         CenteredUniformVectorScaler
[17:05:49] (INFO):         Identity
[17:05:49] (INFO):         UniformVectorScaler
[17:05:49] (INFO): 
[17:05:49] (INFO):     Label scalers:
[17:05:49] (INFO):         StandardScaler10
[17:05:49] (INFO): 
[17:05:50] (INFO): --------------------------------------------------------------------------------------------------
[17:05:50] (INFO):   3. Training/Validation split
[17:05:50] (INFO): --------------------------------------------------------------------------------------------------
[17:05:50] (INFO): Fraction     : 1.0
[17:05:50] (INFO): # training   : 7
[17:05:50] (INFO): # validation : 0
[17:05:50] (INFO): 
[17:05:50] (INFO): --------------------------------------------------------------------------------------------------
[17:05:50] (INFO):   4. Initializing model
[17:05:50] (INFO): --------------------------------------------------------------------------------------------------
[17:05:50] (INFO): Total params: 64,201
[17:05:50] (INFO): Trainable params: 64,201
[17:05:50] (INFO): Non-trainable params: 0
[17:05:50] (INFO): 
[17:05:50] (INFO): --------------------------------------------------------------------------------------------------
[17:05:50] (INFO):   5. Training
[17:05:50] (INFO): --------------------------------------------------------------------------------------------------
[17:05:56] (ERROR): *** INTERNAL PROGRAMMING ERROR ***

Tagged:

Answers

  • Adriano A. Koga
    Adriano A. Koga
    Altair Employee
    edited October 2023

    hi,

    what are the training parameters that you're using?

    could you share more details?

  • Unknown
    Unknown
    edited October 2023
    Adriano A. Koga_21884 said:

    hi,

    what are the training parameters that you're using?

    could you share more details?

    Umm.. my co-worker did exactly same jobs with same inputs, and it works normally.

    So, I think it is not problem of inputs parameters.

    (Of course, it is altair provided examples)

     

    Training parameter is like:

    Width: 30

    Epochs: 100

    Depth: 1

    Learning Rate: 1e-3

    Patience: 10 (Ealry Stopping On)

    I did train locally.

     

    Log:

    [08:03:56] (INFO): ************************************************************************
    [08:03:56] (INFO): **                                                                    **
    [08:03:56] (INFO): **                                                                    **
    [08:03:56] (INFO): **                     Altair PhysicsAI 2022.3.0.24                   **
    [08:03:56] (INFO): **                                                                    **
    [08:03:56] (INFO): **                 Advanced Machine Learning Software                 **
    [08:03:56] (INFO): **                    from Altair Engineering, Inc.                   **
    [08:03:56] (INFO): **                                                                    **
    [08:03:56] (INFO): ** Build: 2514318                                                     **
    [08:03:56] (INFO): ************************************************************************
    [08:03:56] (INFO): **  COPYRIGHT (C) 1996-2020                 Altair Engineering, Inc.  **
    [08:03:56] (INFO): ** All Rights Reserved.  Copyright notice does not imply publication. **
    [08:03:56] (INFO): **         Contains trade secrets of Altair Engineering, Inc.         **
    [08:03:56] (INFO): ** Decompilation or disassembly of this software strictly prohibited. **
    [08:03:56] (INFO): ************************************************************************
    [08:03:56] (INFO): 
    [08:03:56] (INFO): --------------------------------------------------------------------------------------------------
    [08:03:56] (INFO):   1. Building features and labels
    [08:03:56] (INFO): --------------------------------------------------------------------------------------------------
    [08:04:02] (INFO): Node features:
    [08:04:02] (INFO):     name: cae.coord
    [08:04:02] (INFO):     type: CONTINOUS
    [08:04:02] (INFO):     length: 3
    [08:04:02] (INFO): 
    [08:04:02] (INFO):     name: cae.part_id
    [08:04:02] (INFO):     type: CATEGORICAL
    [08:04:02] (INFO):     length: 2
    [08:04:02] (INFO): 
    [08:04:02] (INFO): Edge features:
    [08:04:02] (INFO):     name: cae.direction
    [08:04:02] (INFO):     type: CONTINOUS
    [08:04:02] (INFO):     length: 4
    [08:04:02] (INFO): 
    [08:04:02] (INFO): Node labels:
    [08:04:02] (INFO):     name: cae.results
    [08:04:02] (INFO):     subcase: Flow Solution
    [08:04:02] (INFO):     field: pressure
    [08:04:02] (INFO):     type: CONTINOUS
    [08:04:02] (INFO):     length: 1
    [08:04:02] (INFO): 
    [08:04:02] (INFO): --------------------------------------------------------------------------------------------------
    [08:04:02] (INFO):   2. Transforming training data
    [08:04:02] (INFO): --------------------------------------------------------------------------------------------------
    [08:04:04] (INFO): Transform: radioss_rigid_rename
    [08:04:04] (INFO):     Feature scalers:
    [08:04:04] (INFO):         Identity
    [08:04:04] (INFO):         RadiossRigidRename
    [08:04:04] (INFO):         Identity
    [08:04:04] (INFO): 
    [08:04:04] (INFO):     Label scalers:
    [08:04:04] (INFO):         Identity
    [08:04:04] (INFO): 
    [08:04:04] (INFO): Transform: binarize_categorical
    [08:04:04] (INFO):     Feature scalers:
    [08:04:04] (INFO):         Identity
    [08:04:04] (INFO):         MultiLabelBinarizerStrict
    [08:04:04] (INFO):         Identity
    [08:04:04] (INFO): 
    [08:04:04] (INFO):     Label scalers:
    [08:04:04] (INFO):         Identity
    [08:04:04] (INFO): 
    [08:04:04] (INFO): Transform: filter_const_comps
    [08:04:04] (INFO):     Feature scalers:
    [08:04:04] (INFO):         ConstantCompFilter
    [08:04:04] (INFO):         ConstantCompFilter
    [08:04:04] (INFO):         ConstantCompFilter
    [08:04:04] (INFO): 
    [08:04:04] (INFO):     Label scalers:
    [08:04:04] (INFO):         ConstantCompFilter
    [08:04:04] (INFO): 
    [08:04:04] (INFO): Transform: cntr_vec_uni_edges_stnd10_res
    [08:04:04] (INFO):     Feature scalers:
    [08:04:04] (INFO):         CenteredUniformVectorScaler
    [08:04:04] (INFO):         Identity
    [08:04:04] (INFO):         UniformVectorScaler
    [08:04:04] (INFO): 
    [08:04:04] (INFO):     Label scalers:
    [08:04:04] (INFO):         StandardScaler10
    [08:04:04] (INFO): 
    [08:04:06] (INFO): --------------------------------------------------------------------------------------------------
    [08:04:06] (INFO):   3. Training/Validation split
    [08:04:06] (INFO): --------------------------------------------------------------------------------------------------
    [08:04:06] (INFO): Fraction     : 1.0
    [08:04:06] (INFO): # training   : 7
    [08:04:06] (INFO): # validation : 0
    [08:04:06] (INFO): 
    [08:04:06] (INFO): --------------------------------------------------------------------------------------------------
    [08:04:06] (INFO):   4. Initializing model
    [08:04:06] (INFO): --------------------------------------------------------------------------------------------------
    [08:04:06] (INFO): Total params: 64,201
    [08:04:06] (INFO): Trainable params: 64,201
    [08:04:06] (INFO): Non-trainable params: 0
    [08:04:06] (INFO): 
    [08:04:06] (INFO): --------------------------------------------------------------------------------------------------
    [08:04:06] (INFO):   5. Training
    [08:04:06] (INFO): --------------------------------------------------------------------------------------------------
    [08:04:12] (ERROR): *** INTERNAL PROGRAMMING ERROR ***

     

Categories