DeepONet Project Update: Model Accessibility and Checkpoint Restoration Fix #47

parsa-rahbari-82 · 2024-12-02T16:56:24Z

Description:

This pull request provides a series of important updates to the DeepONet project, primarily focusing on improving model accessibility and bug fixes. The following changes have been made:

Command-Line Model Parameters:
- Added command-line arguments to give users more control over the model configuration without needing to modify the source code directly.
- Users can now adjust critical parameters such as learning rate, number of sensors, and training epochs via the command line using Python's argparse library.
Example usage:

python deeponet_pde.py -p ode -m 200 --num-train 10000 --num-test 3000--lr 1e-3 --epochs 50000 --nn deeponet --activation relu --init Glorot --stacked False
Checkpoint Restoration Bug Fix:
- Resolved a bug related to restoring checkpoints, which was caused by an incorrect file path.
- The model can now properly restore from saved states for both training and inference.
Legendre Transform Flexibility:
- Replaced the hardcoded time value (T=2) with a dynamic time variable.
- This update ensures that the model can work with any time interval, providing flexibility and preventing extrapolation errors.
Compatibility with DeepONet’s Dynamic Systems:
- Updated the structure of X_train and ensured proper formatting for X_test to be compatible with TensorFlow's compat.v1 backend.
- Added necessary changes to avoid type mismatch errors, such as casting arrays in X_test to np.float32.
Example code snippet:

X_test = tuple([arr.astype(dtype=np.float32) for arr in X_test])

Testing:

These changes have been thoroughly tested and confirmed to work reliably with TensorFlow's compat.v1 backend.

MIND Lab:

Supervisor: Dr. Mostafa Abbaszadeh
Team:
- AmirAli Fakhari Zavareh
- Melika Taheri Soufi
- Parsa Rahbari

This reverts commit 38260b9.

.gitignore

src/deeponet_pde.py

src/system.py

parsa-rahbari-82 · 2024-12-15T09:43:01Z

Dear Dr. Lu
Thanks for the review! I’ve addressed the points you raised:

Redundant python 2 removed.
.gitignore file removed.
Using original order of import.

src/spaces.py

parsa-rahbari-82 · 2025-01-18T08:32:32Z

I reordered imports according to PEP 8 guidelines.

requirements.txt

.gitignore

src/test/u_xsin2x.dat

- .gitignore - requirements.txt

lululxvi · 2025-05-24T23:56:37Z

requirements.txt

@@ -1,8 +0,0 @@
-deepxde


Keep this file as the original one.

@AmirAliFakhari It seems the file roolls back to the wrong version.

Dear Mr. Lu Lu,

I hope you're doing well. I've been working on the DeepONet project for the past few months and truly appreciate your feedback throughout the process.

To help streamline the next steps and avoid back-and-forth revisions, would it be possible for you to share a complete list of the issues or changes you'd like to see in the current version? This would really help me address everything thoroughly and ensure that the final result meets your expectations.

lululxvi · 2025-06-25T20:51:52Z

src/deeponet_pde.py

-
+import deepxde as dde
 import numpy as np
 import tensorflow as tf


import argparse import deepxde as dde import itertools import numpy as np import tensorflow as tf

lululxvi · 2025-06-25T20:52:34Z

src/deeponet_pde.py

-from spaces import FinitePowerSeries, FiniteChebyshev, GRF
-from system import LTSystem, ODESystem, DRSystem, CVCSystem, ADVDSystem
-from utils import merge_values, trim_to_65535, mean_squared_error_outlier, safe_test
+from spaces import GRF, FiniteChebyshev, FinitePowerSeries


from spaces import FiniteChebyshev, FinitePowerSeries, GRF

lululxvi · 2025-06-25T20:54:13Z

src/deeponet_pde.py

        X_test = merge_values(X_test)
-    y_pred = model.predict(data.transform_inputs(X_test))
+    X_test = tuple([arr.astype(dtype=np.float32) for arr in X_test])
+    y_pred = model.predict(X_test)


Why no data.transform_inputs? This is for data normalization.

lululxvi · 2025-06-25T20:54:54Z

src/deeponet_pde.py

        # Gravity pendulum
-        # k = 1
-        # return [s[1], - k * np.sin(s[0]) + u]
+        k = 1


There is return in line 108. The code here is not used.

lululxvi · 2025-06-25T20:57:50Z

src/system.py

-from __future__ import print_function
-
 import numpy as np
+import config


config is a local file. Move it below as the original code.

lululxvi · 2025-06-25T20:58:17Z

src/utils.py

-
 import sys
 import time
+import numpy as np


Use the original import format for numpy

parsa-rahbari-82 and others added 14 commits October 31, 2024 13:59

compatibility bug fixed.

4b11912

tensorflow_intel removed from dependencies.

e628ef1

dynamic module import for src

38260b9

fix bug for dynamic import.

b4474e4

Revert "dynamic module import for src"

5ccd287

This reverts commit 38260b9.

tendorflow compatibility fixed.

968cd16

setting up river pollutants example.

0cb134b

smaller model for faster training

003c642

basic commands added.

87ee07e

necessary commands added.

de6c846

redundant prints removed.

ebe3542

Minor tensorflow 2 compatibility solved.

d1b599d

major bugs for checkpoints solved

a4a2616

Legendre transform and deeponet compatibility bugs fixed.

68d5769

lululxvi reviewed Dec 15, 2024

View reviewed changes

.gitignore Outdated Show resolved Hide resolved

lululxvi reviewed Dec 15, 2024

View reviewed changes

src/deeponet_pde.py Outdated Show resolved Hide resolved

lululxvi reviewed Dec 15, 2024

View reviewed changes

src/system.py Outdated Show resolved Hide resolved

parsa-rahbari-82 added 2 commits December 15, 2024 12:54

Stop tracking .gitignore

a1ee9c8

import order resolved. redundant python2 imports removed.

f4a89ee

lululxvi reviewed Jan 18, 2025

View reviewed changes

src/spaces.py Outdated Show resolved Hide resolved

Import ordering fixed.

358c03e

parsa-rahbari-82 added 3 commits February 3, 2025 22:28

fixing default advd example.

b223d3a

including test folder

9c951fe

including test folder for demo.

ca6143c

lululxvi reviewed Mar 18, 2025

View reviewed changes

requirements.txt Show resolved Hide resolved

lululxvi reviewed Mar 18, 2025

View reviewed changes

.gitignore Outdated Show resolved Hide resolved

lululxvi reviewed Mar 18, 2025

View reviewed changes

src/test/u_xsin2x.dat Outdated Show resolved Hide resolved

Redundant files removed.

6981a5d

- .gitignore - requirements.txt

Remove test file.

524ec0e

lululxvi reviewed May 24, 2025

View reviewed changes

Rolling back previous requirements.txt

41e0289

lululxvi reviewed Jun 25, 2025

View reviewed changes

src/utils.py

import sys

import time

import numpy as np

Copy link

Owner

lululxvi Jun 25, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Use the original import format for numpy

DeepONet Project Update: Model Accessibility and Checkpoint Restoration Fix #47

Are you sure you want to change the base?

DeepONet Project Update: Model Accessibility and Checkpoint Restoration Fix #47

Uh oh!

Conversation

parsa-rahbari-82 commented Dec 2, 2024

Description:

Testing:

Uh oh!

Uh oh!

Uh oh!

Uh oh!

parsa-rahbari-82 commented Dec 15, 2024

Uh oh!

Uh oh!

parsa-rahbari-82 commented Jan 18, 2025

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

lululxvi Jun 25, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

lululxvi Jun 25, 2025 •

edited

Loading