Input image dimension requirements unclear

[32-git]

Hi. I am using this repo as a part of my project. And I am currently having a problem with the dimensions of the image I pass to the HME recognizer.

If I use the original image size, I get the following error:
Exception: Kernel shape must have the same length as input, but received kernel of shape (3, 3, 1, 32) and input of shape (None, None, 28, 28, 1).

In my attempts to resize the image, I have to go down to sizes such as 120x120 to avoid getting the above error. But when I go down to that image size, then the handwritten math is barely recognizable.

So instead of constantly guessing, I wanted to know what the actual requirements of the model are regarding image size and whether it would be possible to pass larger images (e.g. 600x600) to the model as well.

Thanks in advance!

[carolreis]

Hi! Just to give you a feedback: I was in a conference these last weeks, I just came back. Within a few days I'll take a look and help you!

[carolreis]

Meanwhile, can you give examples of your images? @32-git

[32-git]

Hi! Thanks for the reply and my apologies if i disturbed you. My images may look something like this

or this

The width of the image may vary as shown above, based on the browser window size when drawing.

Now I would genuinely appreciate it if you could let me know whether these images are fine. If not, what should I pay attention to when passing the images to the model?

[carolreis]

Hi! Don't worry, it's ok to ask me some help.

I'll try to use your images. Did you take a look carefully at the code I used as example?

I'll let you know my results and see what the problem is

But apparently these images should be fine

[32-git]

Great! If the images should be fine, that's a good start. I have in fact looked at your example code and that's what I based my approach off of:

from mathreader import api
import base64 as b64
import sys

hme_recognizer = api.HME_Recognizer()

def pass_image(image):
  try:
    hme_recognizer.load_image(image)
    expression, img = hme_recognizer.recognize()
    sys.stdout.write(f"{expression}")
  except Exception as e:
    sys.stdout.write(f'Exception: {e}')

try:
  data_url = sys.argv[1]
  header, encoded = data_url.split("base64,", 1)
  pass_image(encoded)

except Exception as err:
  print(f'Exception: {err}')

As I said, the error that I mentioned in my first message is what I receive. Although if it helps I could also paste the entire output I get, which shows all the steps of when the code executes.

[carolreis]

Hello, I ran the "example.py" with your first image and I had the correct output:

---

I added both images to:
images/validation/teste.png and images/validation/teste2.png

And, then I ran:

python3 example.py validation/teste.png and python3 example.py validation/teste2.png

---

For simplicity, I didn't create another project importing mathreader.
I did the test inside the repo:

virtualenv -p $(which python3) .
pip install -e .
cd mathreader
python3 example.py validation/teste.png

In your case, you should install it directly by pip(3) install mathreader...

---

Now, I will test it with your code.

But now we know that your images are compatible with the application

[carolreis]

@32-git

This code worked with your images:

teste.py

from mathreader.api import *
from mathreader.config import Configuration
from mathreader.helpers.exceptions import *
import base64
import numpy as np
import cv2
import sys

configs = Configuration()

class Example:

    def __init__(self):

        expression = ""
        hme_recognizer = HME_Recognizer()
        print(sys.argv)

        if sys.argv and len(sys.argv) > 1:
            image_path = configs.package_path + '/images/' + sys.argv[1]
        else:
            image_path = configs.package_path + '/images/validation/21.png'

        # Turning into base64 for testing
        with open(image_path, "rb") as img_file:
            encoded_string = base64.b64encode(img_file.read()).decode('utf-8')

        data_url = f"data:image/png;base64,{encoded_string}"

        header, encoded = data_url.split("base64,", 1)

        try:
            hme_recognizer.load_image(encoded, data_type='base64')
            expression, img = hme_recognizer.recognize()

        except (GrammarError, SintaticError, LexicalError) as e:

            if 'latex_string_original' in e.data:
                expression = e.data['latex_string_original']

            print('[example.py] Exception: ', e.data)
            print('[example.py] Exception: ', e.valor)

        print("\nExpression: ", expression)


Example()

Running:
python3 teste.py validation/teste2.png

[32-git]

Hi thanks a lot for your patience. Relieved to hear that it works.

I tried your code out, but I changed one thing about it. In my project, I receive a base64 string directly from my app which i then send to the python script. Therefore, I skipped the part where you convert the image to b64 and just went straight to the "header, encoded" part.

But I still somehow received an error:

"2024-09-27 15:30:59.928093: I tensorflow/core/util/port.cc:153] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
2024-09-27 15:31:01.464427: I tensorflow/core/util/port.cc:153] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.

...

UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.
  super().__init__(activity_regularizer=activity_regularizer, **kwargs)
Traceback (most recent call last):

[bunch of references to files]

    raise ValueError(
ValueError: Kernel shape must have the same length as input, but received kernel of shape (3, 3, 1, 32) and input of shape (None, None, 28, 28, 1).
"

I shortened the error to what I thought were the most important bits. Is there any chance you get what's going on here?

[carolreis]

@32-git I know it's a bit long, put can you give me the output/value of the encoded image?

[carolreis]

About tensorflow... I don't know why you're having this issue. But here they comment about it: https://stackoverflow.com/questions/77921357/warning-while-using-tensorflow-tensorflow-core-util-port-cc113-onednn-custom

[32-git]

@32-git I know it's a bit long, put can you give me the output/value of the encoded image?

@carolreis
Sure. Here's the data URL with the header (before it's split):

data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAABBkAAAJYCAYAAAAnqz4BAAAAAXNSR0IArs4c6QAAIABJREFUeF7t3U9spddZP/BjqEAs2MAEJm3V2EhI6YakqkqJNFHsBeCoi0wXEUKaYFszUpEQkC5Qy8q2YBFWbdVKgDQj202lSrDozKYe/kh2NCOlFAkm3VB140QCMtBU7YpF1eLf77lwzTseX/te+7x/z+eVqklm7j3veT7PUTrznfOed+7w8PAwuQgQIECAAAECBAgQIECAAAECFxSYEzJcUNDXCRAgQIAAAQIECBAgQIAAgZGAkMFCIECAAAECBAgQIECAAAECBLIICBmyMBqEAAECBAgQIECAAAECBAgQEDJYAwQIECBAgAABAgQIECBAgEAWASFDFkaDECBAgAABAgQIECBAgAABAkIGa4AAAQIECBAgQIAAAQIECBDIIiBkyMJoEAIECBAgQIAAAQIECBAgQEDIYA0QIECAAAECBAgQIECAAAECWQSEDFkYDUKAAAECBAgQIECAAAECBAgIGawBAgQIECBAgAABAgQIECBAIIuAkCELo0EIECBAgAABAgQIECBAgAABIYM1QIAAAQIECBAgQIAAAQIECGQREDJkYTQIAQIECBAgQIAAAQIECBAgIGSwBggQIECAAAECBAgQIECAAIEsAkKGLIwGIUCAAAECBAgQIECAAAECBIQM1gABAgQIECBAgAABAgQIECCQRUDIkIXRIAQIECBAgAABAgQIECBAgICQwRogQIAAAQIECBAgQIAAAQIEsggIGbIwGoQAAQIECBAgQIAAAQIECBAQMlgDBAgQIECAAAECBAgQIECAQBYBIUMWRoMQIECAAAECBAgQIECAAAECQgZrgAABAgQIECBAgAABAgQIEMgiIGTIwmgQAgQIECBAgAABAgQIECBAQMhgDRAgQIAAAQIECBAgQIAAAQJZBIQMWRgNQoAAAQIECBAgQIAAAQIECAgZrAECBAgQIECAAAECBAgQIEAgi4CQIQujQQgQIECAAAECBAgQIECAAAEhgzVAgAABAgQIECBAgAABAgQIZBEQMmRhNAgBAgQIECBAgAABAgQIECAgZLAGCBAgQIAAAQIECBAgQIAAgSwCQoYsjAYhQIAAAQIECBAgQIAAAQIEhAzWAAECBAgQIECAAAECBAgQIJBFQMiQhdEgBAgQIECAAAECBAgQIECAgJDBGiBAgAABAgQIECBAgAABAgSyCAgZsjAahAABAgQIECBAgAABAgQIEBAyWAMECBAgQIAAAQIECBAgQIBAFgEhQxZGgxAgQIAAAQIECBAgQIAAAQJCBmuAAAECBAgQIECAAAECBAgQyCIgZMjCaBACBAgQIECAAAECBAgQIEBAyGANECBAgAABAgQIECBAgAABAlkEhAxZGA1CgAABAgQIECBAgAABAgQICBmsAQIECBAgQIAAAQIECBAgQCCLgJAhC6NBCBAgQIAAAQIECBAgQIAAASGDNUCAAAECBAgQIECAAAECBAhkERAyZGE0CAECBAgQIECAAAECBAgQICBksAYIECBAgAABAgQIECBAgACBLAJChiyMBiFAgAABAgQIECBAgAABAgSEDNYAAQIECBAgQIAAAQIECBAgkEVAyJCF0SAECBAgQIAAAQIECBAgQICAkMEaIECAAAECBAgQIECAAAECBLIICBmyMBqEAAECBAgQIECAAAECBAgQEDJYAwQIECBAgAABAgQIECBAgEAWASFDFkaDECBAgAABAgQIECBAgAABAkIGa4AAAQIECBAgQIAAAQIECBDIIiBkyMJoEAIECBAgQIAAAQIECBAgQEDIYA0QIECAAAECBAgQIECAAAECWQSEDFkYDUKAAAECBAgQIECAAAECBAgIGawBAgQIECBAgAABAgQIECBAIIuAkCELo0EIECBAgAABAgQIECBAgAABIYM1QIAAAQIECBAgQIAAAQIECGQREDJkYTQIAQIECBAgQIAAAQIECBAgIGSwBggQIECAAAECBAgQIECAAIEsAkKGLIwGIUCAAAECBAgQIECAAAECBIQM1gABAgQIECBAgAABAgQIECCQRUDIkIXRIAQIECBAgAABAgQIECBAgICQwRogQIAAAQIECBAgQIAAAQIEsggIGbIwGoQAAQIECBAgQIAAAQIECBAQMlgDBAgQIECAAAECBAgQIECAQBYBIUMWRoMQIECAAAECBAgQIECAAAECQgZrgAABAgQIECBAgAABAgQIEMgiIGTIwmgQAgQIECBAgAABAgQIECBAQMhgDRAgQIAAAQIECBAgQIAAAQJZBIQMWRgNQoAAAQIECBAgQIAAAQIECAgZrAECBAgQIECAAAECBAgQIEAgi4CQIQujQQgQIECAAAECBAgQIECAAAEhgzVAgAABAgQIECBAgAABAgQIZBEQMmRhNAgBAgQIECBAgAABAgQIECAgZLAGCBAgQIAAAQIECBAgQIAAgSwCQoYsjAYhQIAAAQIECBAgQIAAAQIEhAzWAAECBAgQIECAAAECBAgQIJBFQMiQhdEgBAgQIECAAAECBAgQIECAgJDBGiBAgAABAgQIECBAgAABAgSyCAgZsjAahAABAgQIECBAgAABAgQIEBAyWAMECBAgQIAAAQIECBAgQIBAFgEhQxZGgxAgQIAAAQIECBAgQIAAAQJCBmuAAAECBAgQIECAAAECBAgQyCIgZMjCaBACBAgQIECAAAECBAgQIEBAyGANECBAgAABAgQIECBAgAABAlkEhAxZGA1CgAABAgQIECBAgAABAgQICBmsAQIECBAgQIAAAQIECBAgQCCLgJAhC6NBCBAgQIAAAQIECBAgQIAAASGDNUCAAAECBAgQIECAAAECBAhkERAyZGE0CAECBAgQIECAAAECBAgQICBksAYIECBAgAABAgQIECBAgACBLAJChiyMBiFAgAABAgQIECBAgAABAgSEDNYAAQIECBAgQIAAAQIECBAgkEVAyJCF0SAECBAgQIAAAQIECBAgQICAkMEaIECAAAECBAhkEHj77bcnjjLp1876+X/8x39MP/VTP5WeffbZx8a+dOlS+r3f+70MMzcEAQIECBDIJyBkyGdpJAIECBAgQOAcAnX84fykabzzzjuP/fSs9z7t8+co/cJf+dKXviRouLCiAQgQIEAgp4CQIaemsQgQIECgNoHxH+6qP/71X//1xL/lrW0ixwb+5je/OfqZX/3VX816y7rGPWuSP/jBD9J3vvOd9LGPfezoo0P8w/lZDn35dSFDXzplngQIEChHQMhQTq9VSoAAgU4JVMOC+Ofxv4//QDv+ua79zXGnEE2mUwLz8/MT5zPp1yb9/FNPPTUay+MSnWqxyRAgQIDAFAJChimQfIQAAQIETheYFBjEtyI0OP7rPAlUBer4w/lJwifdZ9Z7n/Z5XSVAgAABAgRSEjJYBQQIEKhB4KMf/Wh6991308/93M+NRo+/lTy+nf6s7fBn/fos077oWBES/Od//mf6hV/4hdYCg/Ef7qo/xpwmHYo3i89FPntR20n3rmvcs2p98OBB+uEPf/jI4xL+cH6Wml8nQIAAAQIExgJCBmuBAAECmQUiYPinf/qnzKMOb7hqWBD/PN4eHpXGvx//9eEJqIgAAQIECBAgMDwBIcPweqoiAgRaFnj/+98/2sVQ0nVSIDAODQQGJa0EtRIgQIAAAQKlCwgZSl8B6idAILvAJz7xifT1r399NO7P/uzPpg996EO9f1xivIX+5ZdftsMg+4oxIAECBAgQIEBgOAJChuH0UiUECHREYHNzM21sbIxmEz+ur693ZGamQYAAAQIECBAgQKBeASFDvb5GJ0CgQAEhQ4FNVzIBAgQIECBAgMBIQMhgIRAgQCCzwNraWtre3h6NevXq1fS1r30t8x0MR4AAAQIECBAgQKCbAkKGbvbFrAgQ6LFA9UyGX/u1X0tvvvlmj6sxdQIECBAgQIAAAQLTCwgZprfySQIECEwlELsYYjdDXKurq2lra2uq7/kQAQIECBAgQIAAgb4LCBn63kHzJ0CgcwL7+/tpaWlpNK/FxcW0t7fXuTmaEAECBAgQIECAAIE6BIQMdagakwCBogWEDEW3X/EECBAgQIAAgaIFhAxFt1/xBAjUIfD222+nhYWF0dDz8/Pp4OCgjtsYkwABAgQIECBAgEDnBIQMnWuJCREgMASBubm5ozIODw+HUJIaCBAgQIAAAQIECJwpIGQ4k8gHCBAgMLtA7GSIHQ1xxU6G2NHgIkCAAAECBAgQIDB0ASHD0DusPgIEWhEQMrTC7qYECBAgQIAAAQItCwgZWm6A2xMgMEyBasgQb5eIt0y4CBAgQIAAAQIECAxdQMgw9A6rjwCBVgTiFZbxlom4hAyttMBNCRAgQIAAAQIEWhAQMrSA7pYECAxfYG1tLW1vb48K3draSqurq8MvWoUECBAgQIAAAQLFCwgZil8CAAgQqEPg+eefT/fv3x8Nff369XTz5s06bmNMAgQIECBAgAABAp0SEDJ0qh0mQ4DAUARu3LiRbt26NSpneXk57e7uDqU0dRAgQIAAAQIECBCYKCBksDgIECBQg0CcxxDnMoyvw8PDGu5iSAIECBAgQIAAAQLdEhAydKsfZkOAwIAEvMZyQM1UCgECBAgQIECAwFQCQoapmHyIAAECsws4/HF2M98gQIAAAQIECBDot4CQod/9M3sCBDosEG+XiKAhrsXFxdGrLF0ECBAgQIAAAQIEhiwgZBhyd9VGgECrAsfPZTg4OEjz8/OtzsnNCRAgQIAAAQIECNQpIGSoU9fYBAgUL+BchuKXAAACBAgQIECAQFECQoai2q1YAgSaFnAuQ9Pi7keAAAECBAgQINCmgJChTX33JkBg8ALOZRh8ixVIgAABAgQIECBQERAyWA4ECBCoWWBubu7oDs5lqBnb8AQIECBAgAABAq0KCBla5XdzAgRKEKieyxBvmIg3TbgIECBAgAABAgQIDFFAyDDErqqJAIFOCVTPZVhfX08bGxudmp/JECBAgAABAgQIEMglIGTIJWkcAgQITBBwLoOlQYAAAQIECBAgUIqAkKGUTquTAIFWBZzL0Cq/mxMgQIAAAQIECDQkIGRoCNptCBAoW8C5DGX3X/UECBAgQIAAgVIEhAyldFqdBAi0KuBchlb53ZwAAQIECBAgQKAhASFDQ9BuQ4BA2QLOZSi7/6onQIAAAQIECJQiIGQopdPqJECgdYHqIxMHBwdpfn6+9TmZAAECBAgQIECAAIGcAkKGnJrGIkCAwCkC1ZBha2srra6u8iJAgAABAgQIECAwKAEhw6DaqRgCBLosUD2XIQKGCBpcBAgQIECAAAECBIYkIGQYUjfVQoBApwX29/fT0tLSaI6Li4tpb2+v0/M1OQIECBAgQIAAAQKzCggZZhXzeQIECFxAwLkMF8DzVQIECBAgQIAAgc4LCBk63yITJEBgSAKxkyF2NMTlXIYhdVYtBAgQIECAAAECISBksA4IECDQoMDGxkba3Nwc3dG5DA3CuxUBAgQIECBAgEAjAkKGRpjdhAABAv8j4FwGK4EAAQIECBAgQGDIAkKGIXdXbQQIdFLAuQydbItJESBAgAABAgQIZBAQMmRANAQBAgRmEXAuwyxaPkuAAAECBAgQINAnASFDn7plrgQIDELAuQyDaKMiCBAgQIAAAQIEThAQMlgWBAgQaFjAuQwNg7sdAQIECBAgQIBAYwJChsao3YgAAQL/J+BcBquBAAECBAgQIEBgiAJChiF2VU0ECHRewLkMnW+RCRIgQIAAAQIECJxDQMhwDjRfIUCAwEUFtre309ra2miY1dXVtLW1ddEhfZ8AAQIECBAgQIBA6wJChtZbYAIECJQoUD2XIeo/PDwskUHNBAgQIECAAAECAxMQMgysocohQKA/As5l6E+vzJQAAQIECBAgQGA6ASHDdE4+RYAAgewCzz//fLp///5o3OvXr6ebN29mv4cBCRAgQIAAAQIECDQpIGRoUtu9CBAgUBG4ceNGunXr1uhnlpeX0+7uLh8CBAgQIECAAAECvRYQMvS6fSZPgECfBY6fy3BwcJDm5+f7XJK5EyBAgAABAgQIFC4gZCh8ASifAIF2BZzL0K6/uxMgQIAAAQIECOQVEDLk9TQaAQIEZhKI11jG6yzjWl9fTxsbGzN934cJECBAgAABAgQIdElAyNClbpgLAQLFCUTAEEFDXIuLi2lvb684AwUTIECAAAECBAgMR0DIMJxeqoQAgZ4KeGSip40zbQIECBAgQIAAgccEhAwWBQECBFoWqIYMW1tbaXV1teUZuT0BAgQIECBAgACB8wkIGc7n5lsECBDIJhDnMGxubo7Gi4AhggYXAQIECBAgQIAAgT4KCBn62DVzJkBgUALVV1k6l2FQrVUMAQIECBAgQKA4ASFDcS1XMAECXRRwLkMXu2JOBAgQIECAAAECswoIGWYV83kCBAjUIFB9laVzGWoANiQBAgQIECBAgEAjAkKGRpjdhAABAqcLVF9l6VwGq4UAAQIECBAgQKCvAkKGvnbOvAkQGJRA9VyGKOzg4CDNz88PqkbFECBAgAABAgQIDF9AyDD8HquQAIGeCDiXoSeNMk0CBAgQIECAAIGJAkIGi4MAAQIdEaiey7C+vp7i1ZYuAgQIECBAgAABAn0SEDL0qVvmSoDAoAWq5zJ4leWgW604AgQIECBAgMBgBYQMg22twggQ6KOARyb62DVzJkCAAAECBAgQGAsIGawFAgQIdEigGjJ4lWWHGmMqBAgQIECAAAECUwkIGaZi8iECBAg0IxDnMGxubo5u5lWWzZi7CwECBAgQIECAQD4BIUM+SyMRIEDgwgLVV1k6l+HCnAYgQIAAAQIECBBoWEDI0DC42xEgQOAsAecynCXk1wkQIECAAAECBLoqIGToamfMiwCBYgWef/75dP/+/VH9169fTzdv3izWQuEECBAgQIAAAQL9EhAy9KtfZkuAQAECN27cSLdu3RpVury8nHZ3dwuoWokECBAgQIAAAQJDEBAyDKGLaiBAYFAC1XMZorDDw8NB1acYAgQIECBAgACB4QoIGYbbW5URINBjAecy9Lh5pk6AAAECBAgQKFhAyFBw85VOgEB3BdbW1tL29vZogltbW6PXWboIECBAgAABAgQIdF1AyND1DpkfAQJFCkTAEEFDXBEwRNDgIkCAAAECBAgQINB1ASFD1ztkfgQIFCngXIYi265oAgQIECBAgEDvBYQMvW+hAggQGKqAcxmG2ll1ESBAgAABAgSGKyBkGG5vVUaAQM8FnMvQ8waaPgECBAgQIECgQAEhQ4FNVzIBAv0QcC5DP/pklgQIECBAgAABAv8nIGSwGggQINBRAecydLQxpkWAAAECBAgQIDBRQMhgcRAgQKDDAs5l6HBzTI0AAQIECBAgQOAxASGDRUGAAIEOCziXocPNMTUCBAgQIECAAAEhgzVAgACBPgk4l6FP3TJXAgQIECBAgAABOxmsAQIECHRYwLkMHW6OqREgQIAAAQIECNjJYA0QIECgbwLOZehbx8yXAAECBAgQIFCugJ0M5fZe5QQI9ETAuQw9aZRpEiBAgAABAgQIJCGDRUCAAIGOCziXoeMNMj0CBAgQIECAAIEjASGDxUCAAIGOCziXoeMNMj0CBAgQIECAAAEhgzVAgACBPgk4l6FP3TJXAgQIECBAgEC5AnYylNt7lRMg0COBF198Md29e3c04+vXr6ebN2/2aPamSoAAAQIECBAgUIqAkKGUTquTAIFeC9y4cSPdunVrVMOVK1fSvXv3el2PyRMgQIAAAQIECAxTQMgwzL6qigCBgQlUz2VYXFxMe3t7A6tQOQQIECBAgAABAkMQEDIMoYtqIECgCIG5ubmjOg8ODtL8/HwRdU8q8u233x79UvwY/7t9+3a6dOlSunbtWoogxkWAAAECBAgQINC8gJCheXN3JECAwLkEqoc/xk6Gof1BuhoaVMOD+Od33nnnKEwYf+40xAhgwueFF15Iq6ur5/L2JQIECBAgQIAAgdkFhAyzm/kGAQIEWhFYW1tL29vbo3tvbW314g/Px3cbjP99HBocDxPqgB3v+FhZWRkFD0MLZ+owMyYBAgQIECBA4LwCQobzyvkeAQIEGhaIgCGChrjib+cjaOjC9Qd/8AfprbfeSr/0S7802m0QV5whUfc1Dg/ix/jf97///fS9730v3b9//9RbV3c5jP+57rkanwABAgQIECBQioCQoZROq5MAgd4LdO3wxy984QvpT/7kT0Z/sM91HQ8OnnrqqdHQ4yBh/ONp9xuf0RBeb7zxxpmBxzhoGO90yFWLcQgQIECAAAECJQoIGUrsupoJEOitQBcOf4xw4XOf+9zonISzrkmhQXxv/NjCNMHBWfc5K3SoBg6nnekwnkuc5eDRiouo+y4BAgQIECBQqoCQodTOq5sAgV4KtHn446Rw4ed//ufTxz72sfRbv/VbR2+8qDs4uEjzImSI0GFnZ2fqXQ4OkLyIuO8SIECAAAECJQkIGUrqtloJEOi9wMc//vH0zW9+c1THyy+/nP7qr/6q9pomhQvxKMOnP/3p9Id/+Ie1z6GuG3i0oi5Z4xIgQIAAAQKlCggZSu28ugkQ6KXAiy++mO7evTua+6c+9an0F3/xF7XVMeRwYRLaeJdDnOUwfpPHpM96tKK2pWdgAgQIECBAoMcCQoYeN8/UCRAoT+CVV15JX/nKV0aFx3b/3/md38mOUGK4cFbo4NGK7MvMgAQIECBAgMBABYQMA22ssggQGKbAb/zGb6S/+7u/GxUXOxp+8zd/M1uhwoXTKcePVkTgMN7xcNo3YqfD+I0V40MuszXLQAQIECBAgACBjgoIGTraGNMiQIDASQLPPPNM+ta3vjX6pX/+539Ozz777IWhhAvnI5z10Yqnn346feADH0g3b9483w19iwABAgQIECDQAwEhQw+aZIoECBAYC1y+fDn9x3/8x+hf//3f/z09+eST58b5/d///fTVr341fe9733tkjCEc6HhulHN+cfxazDjHIc5ziLdXTLqih3GexsbGxjnv5msECBAgQIAAge4KCBm62xszI0CAwCMC//3f/51+8id/8ujnfvzjH6ef+ImfOJdS7IB46623hAvn0jv7S9VdDpMerYjHKdbX19Pq6urZA/oEAQIECBAgQKAnAkKGnjTKNAkQIPDuu++m97///SOIX/zFX0wPHz48F0r8bfva2trRd2OsP/7jP+71qyjPBdHglyJoePXVV9M//MM/PNa3CBu2traScxsabIhbESBAgAABArUJCBlqozUwAQIE8go8ePAgfeQjHxkN+iu/8iuP7USY5m6xjX9paenoo7/8y7+cvvOd70zzVZ/JIBBhQ4Q8m5ubj40mbMgAbAgCBAgQIECgdQEhQ+stMAECBAhMJ/A3f/M3aXl5efThX//1X09/+7d/O90X//dTxwOG+Jvzvb29mcbw4TwCp4UN8fhEPEYRoYOLAAECBAgQINA3ASFD3zpmvgQIFCvw5S9/efRKxLiuXbuWXn/99akt4g+1CwsLR5+PP8AeHBxM/X0frEdgUtgQ/YkQSNhQj7tRCRAgQIAAgfoEhAz12RqZAAECWQV+93d/N/3lX/7laMzY0bC7uzvV+AKGqZha/dBpYUMES95E0Wp73JwAAQIECBCYQUDIMAOWjxIgQKBNgXhE4u///u9nDhniDIbqKxXjEQmHDLbZycn3jrAhDuU8/grM2NkgbOhmz8yKAAECBAgQeFRAyGBFECBAoCcC1bBg2qBAwNCT5h6bprChn30zawIECBAgQCAlIYNVQIAAgZ4IxJkK8YfPuKYJGQQMPWnsKdMcv4li3PfxR+1s6H9vVUCAAAECBIYqIGQYamfVRYDA4ATm5uaOajo8PDy1vthyH39AHV9bW1sp3lrg6p9ABAzx+ES89vKksCEOh9Tb/vXVjAkQIECAwFAFhAxD7ay6CBAYlED18Maz3gwR4UKEDOMr/gAaIYOr3wLjwyF3dnZODBuix87a6HePzZ4AAQIECAxBQMgwhC6qgQCBwQvE32TH4w9xnRYyVD8Xn40/dMajFa7hCEx6E8V4bQgbhtNrlRAgQIAAgT4KCBn62DVzJkCgOIHq7oRJOxMEDGUti9PChlgj8RhFBFIuAgQIECBAgECTAkKGJrXdiwABAucU2NjYGD2TH1f84TH+vXpVH6eInz/rkYpzTsPXOigwKWyINRA7WYQNHWyaKREgQIAAgQELCBkG3FylESAwHIHqQY7HQwYBw3D6fJFKTgsbVlZWHgumLnIv3yVAgAABAgQITBIQMlgbBAgQ6IFA9XWUx19f6VWVPWhgg1OMsCFCqXh8pnp57WWDTXArAgQIECBQsICQoeDmK50Agf4ITHp9pYChPz1seqbChqbF3Y8AAQIECBAIASGDdUCAAIGOCxx/HOLw8HA044985CPpwYMHR7M/vsOh42WZXkMCcWhonOcR66h62dnQUAPchgABAgQIFCYgZCis4colQKB/AtWQYXyg42c/+9n0Z3/2Z0fFxGsL440CLgInCcQaiscnJoUNcc6H9WPtECBAgAABAjkEhAw5FI1BgACBGgWqr6aMtwXEjoUnn3wyPXz4cHTXK1eupHv37tU4A0MPRWB8OOTOzs6JOxsirIo15iJAgAABAgQInFdAyHBeOd8jQIBAQwLxN8zxh8K4XnrppXT16tXRwX7ja/z4REPTcZsBCEx6E0WUFrtlhA0DaLISCBAgQIBASwJChpbg3ZYAAQLTCkSocOfOndHH41WEb7zxxtHfQntMYlpFnztJ4LSwIcKteIwiQgcXAQIECBAgQGBaASHDtFI+R4AAgZYEqm+Q+MxnPvPIWQx2MbTUlIHddlLYEAFDPD4hbBhYw5VDgAABAgRqFBAy1IhraAIECOQQWFhYONq58MEPfjD967/+62hYuxhy6BqjKnAnXtlKAAAgAElEQVRa2BC7aDY2NoARIECAAAECBE4VEDJYIAQIEOi4wNzc3IkztIuh443r8fQibIhzP+LQ0erltZc9bqqpEyBAgACBhgSEDA1Buw0BAgTOI1B9feX73ve+9KMf/Wg0jF0M59H0nVkFhA2zivk8AQIECBAgIGSwBggQINBhgerrK6vTtIuhw00b4NS2t7fT5ubmia+99BjFABuuJAIECBAgcAEBIcMF8HyVAAECdQtUX185vpddDHWrG/8kgdjVEKHXpLAhDoeM9eoiQIAAAQIEyhYQMpTdf9UTINBxgQ9/+MPp29/+9iOztIuh400b+PTGh0Pu7OycuLMhQrB4I4WLAAECBAgQKFNAyFBm31VNgEBPBI4f+hivsHzttdd6MnvTHLLApDdRRM1xQKSwYcjdVxsBAgQIEJgsIGSwOggQINBRgVdeeSV95StfOZrd5cuX07vvvtvR2ZpWqQKnhQ3x+EQ8RhGhg4sAAQIECBAoQ0DIUEafVUmAQA8FnnvuufSNb3zjaObxh7WNjY0eVmLKJQhMChsiYIjHJ4QNJawCNRIgQIAAgZSEDFYBAQIEOirwzDPPpG9961uj2f3Mz/xM+q//+q+OztS0CPyfwGlhgzdRWCkECBAgQGD4AkKG4fdYhQQI9FSg+rjEtWvX0uuvv97TSky7RIEIG9bW1kZvpKhesbNB2FDiilAzAQIECJQiIGQopdPqJECglwKvvvrqaN6f//znezl/kyYwKWz44Ac/mD760Y+m27dvQyJAgAABAgQGJCBkGFAzlUKAAAECBLoqsL29nTY3Nx977WUcaPrVr37Vay+72jjzIkCAAAECMwoIGWYE83ECBAgQIEDgfAKxqyEen/ijP/qj9N577x0N4hGK83n6FgECBAgQ6KKAkKGLXTEnAgQIECAwYIEIG+JRoDt37jxSpbBhwE1XGgECBAgUIyBkKKbVCiVAgAABAt0SiLAhHqGIRymq1+rqqldedqtVZkOAAAECBKYWEDJMTeWDBAgQIECAQB0CJ53XYFdDHdLGJECAAAEC9QsIGeo3dgcCBAgQIEDgDIHY1TAOG6ofjbBhb28vxY8uAgQIECBAoPsCQobu98gMCRAgQIBAMQIRNiwtLT3yFgq7Goppv0IJECBAYAACQoYBNFEJBAgQIEBgSAKn7WpYX19PcWaDiwABAgQIEOimgJChm30xKwIECBAgULxAhA1ra2uj115WLwdDFr80ABAgQIBAhwWEDB1ujqkRIECAAAECKW1sbKSdnR2PUFgMBAgQIECgBwJChh40yRQJECBAgEDpAg6GLH0FqJ8AAQIE+iIgZOhLp8yTAAECBAgQGO1mcDCkhUCAAAECBLorIGTobm/MjAABAgQIEDhB4LRdDVtbW2lxcZEbAQIECBAg0JKAkKEleLclQIAAAQIELiYw6WDI8Rso4tWXLgIECBAgQKBZASFDs97uRoAAAQIECGQWiIMhNzc3Hxk1AoaVlZXRoZEuAgQIECBAoDkBIUNz1u5EgAABAgQI1CTgYMiaYA1LgAABAgRmFBAyzAjm4wQIECBAgEB3Bfb399Pa2prXXXa3RWZGgAABAgMXEDIMvMHKI0CAAAECpQk4GLK0jquXAAECBLokIGToUjfMhQABAgQIEMgmcNrBkM5qyMZsIAIECBAg8IiAkMGCIECAAAECBAYrcNquBgdDDrbtCiNAgACBFgWEDC3iuzUBAgQIECDQjMCksGFxcTFtbW0lr7tspg/uQoAAAQLDFxAyDL/HKiRAgAABAgT+V2B7e3v0ussIHcbX5cuX08c//vF0+/ZtTgQIECBAgMAFBYQMFwT0dQIECBAgQKBfApN2NVy5ciXdu3evX8WYLQECBAgQ6JiAkKFjDTEdAgQIECBAoBmBCBuee+659PDhw6MbxmMTBwcHzUzAXQgQIECAwAAFhAwDbKqSCBAgQIAAgekEImh49dVX0507dx4JGvb29pzTMB2hTxEgQIAAgUcEhAwWBAECBAgQIFC8QLzSMs5qGF+xo0HQUPyyAECAAAEC5xAQMpwDzVcIECBAgACB4QmcFDTEmyfiDRQuAgQIECBAYDoBIcN0Tj5FgAABAgQIFCAQj08sLCwcVRo7GlZWVlIEEC4CBAgQIEDgbAEhw9lGPkGAAAECBAgUJBBBw9LS0tFrLgUNBTVfqQQIECBwYQEhw4UJDUCAAAECBAgMTeB40BD1ra+v29EwtEarhwABAgSyCwgZspMakAABAgQIEBiCgKBhCF1UAwECBAg0LSBkaFrc/QgQIECAAIHeCETQsL29/cibJ1ZXV1McCOkiQIAAAQIEHhcQMlgVBAgQIECAAIFTBE4KGrzi0pIhQIAAAQInCwgZrAwCBAgQIECAwBQCJ73icm9vL0Xg4CJAgAABAgT+R0DIYCUQIECAAAECBKYUEDRMCeVjBAgQIFCsgJCh2NYrnAABAgQIEDiPQJzRsLa2dvTV2MkQZzQsLi6eZzjfIUCAAAECgxIQMgyqnYohQIAAAQIEmhCIcxoWFhYeCRpWVla84rIJfPcgQIAAgU4LCBk63R6TI0CAAAECBLoqcPwVl7GjQdDQ1W6ZFwECBAg0JSBkaErafQgQIECAAIHBCRwPGqLA9fV1OxoG12kFESBAgMC0AkKGaaV8jgABAgQIECBwgkAEDXFGw/7+/tGvChosFQIECBAoVUDIUGrn1U2AAAECBAhkE4igIQ6E3NzcPBozDoKMV1y6CBAgQIBASQJChpK6rVYCBAgQIECgNoGTgoY4pyGChvjRRYAAAQIEShAQMpTQZTUSIECAAAECjQlsbGw8sqNB0NAYvRsRIECAQAcEhAwdaIIpECBAgAABAsMSEDQMq5+qIUCAAIHpBYQM01v5JAECBAgQIEBgaoE4oyEOhBxfsaMhDoRcXV2degwfJECAAAECfRMQMvStY+ZLgAABAgQI9Ebg+CsuI2hYWVnxisvedNBECRAgQGBWASHDrGI+T4AAAQIECBCYQeB40BBf9YrLGQB9lAABAgR6JSBk6FW7TJYAAQIECBDoo8BJQcNLL72Ubt++3cdyzJkAAQIECEwUEDJYHAQIECBAgACBBgQiaIgzGvb394/uFo9OxNkNLgIECBAgMBQBIcNQOqkOAgQIECBAoPMCETTEmyd2dnaO5urRic63zQQJECBAYAYBIcMMWD5KgAABAgQIELioQAQNr7zySrp///5oqDgMcmtrKy0uLl50aN8nQIAAAQKtCwgZWm+BCRAgQIAAAQKlCZz01om9vb1R4OAiQIAAAQJ9FhAy9Ll75k6AAAECBAj0ViCChoWFhaP5R8BwcHDQ23pMnAABAgQIhICQwTogQIAAAQIECLQkEIdALi0tHd19dXV19OiEiwABAgQI9FVAyNDXzpk3AQIECBAgMAiBOAhyc3PzqBYHQQ6irYogQIBAsQJChmJbr3ACBAgQIECgKwLVoMFBkF3pinkQIECAwHkEhAznUfMdAgQIECBAgEBGgTifYW1tLcXjE3FF0OAgyIzAhiJAgACBxgSEDI1RuxEBAgQIECBAYLLASW+ccBCkFUOAAAECfRMQMvStY+ZLgAABAgQIDFbg+BsnHAQ52FYrjAABAoMVEDIMtrUKI0CAAAECBPoosL29PXp0Ynw5CLKPXTRnAgQIlCsgZCi39yonQIAAAQIEOirgIMiONsa0CBAgQOBMASHDmUQ+QIAAAQIECBBoVsBBkM16uxsBAgQI5BMQMuSzNBIBAgQIECBAIJuAgyCzURqIAAECBBoUEDI0iO1WBAgQIECAAIFZBI4fBLm8vJx2d3dnGcJnCRAgQIBAowJChka53YwAAQIECBAgMJvA8YMgV1ZWUvyciwABAgQIdFFAyNDFrpgTAQIECBAgQKAicPXq1XTnzp3Rz1y+fDm9+eabaX5+nhEBAgQIEOicgJChcy0xIQIECBAgQIDA4wJPPPFEeu+990a/4LWWVggBAgQIdFVAyNDVzpgXAQIECBAgQKAiUH2tZfz0wcGB3QxWCAECBAh0TkDI0LmWmBABAgQIECBA4GSBhYWFFIdBxmU3g1VCgAABAl0UEDJ0sSvmRIAAAQIECBA4QeD4IZB2M1gmBAgQINA1ASFD1zpiPgQIECBAgACBUwTsZrA8CBAgQKDLAkKGLnfH3AgQIECAAAECxwT29/fT0tLS0c/azWCJECBAgECXBIQMXeqGuRAgQIAAAQIEphBYW1tL8ehEXKurq2lra2uKb/kIAQIECBCoX0DIUL+xOxAgQIAAAQIEsgrEboYIGsaHQNrNkJXXYAQIECBwAQEhwwXwfJUAAQIECBAg0JZAdTfD4uJi2tvba2sq7kuAAAECBI4EhAwWAwECBAgQIECghwKxiyHOZhjvZohHJuLRCRcBAgQIEGhTQMjQpr57EyBAgAABAgQuILCxsZE2NzdHI8Ruhgga5ufnLzCirxIgQIAAgYsJCBku5ufbBAgQIECAAIHWBOxmaI3ejQkQIEBggoCQwdIgQIAAAQIECPRYoLqb4fLly+nNN9+0m6HH/TR1AgQI9F1AyND3Dpo/AQIECBAgULzAE088kd57772Rw/Xr19PNmzeLNwFAgAABAu0ICBnacXdXAgQIECBAgEA2gatXr6Y7d+6MxlteXk67u7vZxjYQAQIECBCYRUDIMIuWzxIgQIAAAQIEOiiwv78/etPE+Do4OPDIRAf7ZEoECBAoQUDIUEKX1UiAAAECBAgMXmBhYeHodZZ7e3ujt024CBAgQIBA0wJChqbF3Y8AAQIECBAgUIPA2tpa2t7eHo28vr6e4kBIFwECBAgQaFpAyNC0uPsRIECAAAECBGoQiIAhgoa4YhdD7GZwESBAgACBpgWEDE2Lux8BAgQIECBAoCaB6iMTzmWoCdmwBAgQIHCqgJDBAiFAgAABAgQIDEQgDn+MQyDj2traSqurqwOpTBkECBAg0BcBIUNfOmWeBAgQIECAAIEzBOIchs3NzdGnImCIoMFFgAABAgSaFBAyNKntXgQIECBAgACBGgWOv8ry8PCwxrsZmgABAgQIPC4gZLAqCBAgQIAAAQIDEnAuw4CaqRQCBAj0UEDI0MOmmTIBAgQIECBAYJKAV1laGwQIECDQpoCQoU199yZAgAABAgQIZBbwKsvMoIYjQIAAgZkEhAwzcfkwAQIECBAgQKD7Ah6Z6H6PzJAAAQJDFRAyDLWz6iJAgAABAgSKFaiGDF5lWewyUDgBAgRaERAytMLupgQIECBAgACB+gS8yrI+WyMTIECAwOkCQgYrhAABAgQIECAwMIHqqyzn5+fTwcHBwCpUDgECBAh0VUDI0NXOmBcBAgQIECBA4AICzmW4AJ6vEiBAgMC5BYQM56bzRQIECBAgQIBAdwWqr7J0LkN3+2RmBAgQGJqAkGFoHVUPAQIECBAgQCCl5FWWlgEBAgQItCEgZGhD3T0JECBAgAABAg0IzM3NHd0lzmWI8xlcBAgQIECgTgEhQ526xiZAgAABAgQItCjgVZYt4rs1AQIEChUQMhTaeGUTIECAAAECwxeonsuwurqa4mwGFwECBAgQqFNAyFCnrrEJECBAgAABAi0KVF9lubi4mPb29lqcjVsTIECAQAkCQoYSuqxGAgQIECBAoFgBr7IstvUKJ0CAQCsCQoZW2N2UAAECBAgQINCMwIsvvpju3r07utn169fTzZs3m7mxuxAgQIBAkQJChiLbrmgCBAgQIECgFIEbN26kW7dujcpdXl5Ou7u7pZSuTgIECBBoQUDI0AK6WxIgQIAAAQIEmhLY3t5OcQBkXA5/bErdfQgQIFCugJCh3N6rnAABAgQIEChAwOGPBTRZiQQIEOiQgJChQ80wFQIECBAgQIBAboFqyDA/P58ODg5y38J4BAgQIEDgSEDIYDEQIECAAAECBAYuMDc3d1Th4eHhwKtVHgECBAi0KSBkaFPfvQkQIECAAAECDQh4jWUDyG5BgAABAiMBIYOFQIAAAQIECBAYuICQYeANVh4BAgQ6JCBk6FAzTIUAAQIECBAgUIfA0tJSirMZ4trb20uLi4t13MaYBAgQIEDATgZrgAABAgQIECAwdIF4hWW8yjKura2t0assXQQIECBAoA4BOxnqUDUmAQIECBAgQKBDAkKGDjXDVAgQIDBwASHDwBusPAIECBAgQIBA7GKIoCGu2MUQuxlcBAgQIECgDgEhQx2qxiRAgAABAgQIdEhAyNChZpgKAQIEBi4gZBh4g5VHgAABAgQIEIhDH+Pwx7ji0Mc4/NFFgAABAgTqEBAy1KFqTAIECBAgQIBAhwSqIcP8/Hw6ODjo0OxMhQABAgSGJCBkGFI31UKAAAECBAgQmCAwNzd39CuHh4ecCBAgQIBALQJChlpYDUqAAAECBAgQ6JbAwsJCevvtt0eTip0MsaPBRYAAAQIEcgsIGXKLGo8AAQIECBAg0EEBIUMHm2JKBAgQGKCAkGGATVUSAQIECBAgQOC4QBz8GGczxBUHP8YBkC4CBAgQIJBbQMiQW9R4BDomEK8tiyvei+4iQIAAgXIF1tbW0vj/E7a2tvz/QrlLQeUECBCoVUDIUCuvwQm0K1A9TdxvKNvthbsTIECgbYHnnnsufeMb3xhN49q1a+n1119ve0ruT4AAAQIDFBAyDLCpSiIwFtjY2Eibm5ujfxUyWBcECBAoW+CTn/xkun379gghdrfF/y+4CBAgQIBAbgEhQ25R4xHokIDnbzvUDFMhQIBAywIROkf4HFf8uL6+3vKM3J4AAQIEhiggZBhiV9VE4P//LVW8pixOEh9f3oluWRAgQKBsASFD2f1XPQECBJoSEDI0Je0+BBoWqIYM8S70eCe6iwABAgTKFRAylNt7lRMgQKBJASFDk9ruRaBBgep5DJ69bRDerQgQINBRASFDRxtjWgQIEBiYgJBhYA1VDoGxQDVkiOdux8/hEiJAgACBMgWEDGX2XdUECBBoWkDI0LS4+xFoSMChjw1Buw0BAgR6IiBk6EmjTJMAAQI9FxAy9LyBpk/gJAGHPloXBAgQIHBcQMhgTRAgQIBAEwJChiaU3YNAwwIOfWwY3O0IECDQAwEhQw+aZIoECBAYgICQYQBNVAKB4wJx0OPOzs7op69cuZLu3bsHiQABAgQKFxAyFL4AlE+AAIGGBIQMDUG7DYEmBW7cuJFu3bo1uuXy8nLa3d1t8vbuRYAAAQIdFBAydLAppkSAAIEBCggZBthUJRHY3t5Oa2trIwivr7QeCBAgQCAEhAzWAQECBAg0ISBkaELZPQg0LLC/v5/i7RJxLS4upr29vYZn4HYECBAg0DUBIUPXOmI+BAgQGKaAkGGYfVVV4QIOfix8ASifAAECJwgIGSwLAgQIEGhCQMjQhLJ7EGhBYG5u7uiuh4eHLczALQkQIECgSwJChi51w1wIECAwXAEhw3B7q7LCBRYWFlLsaIjr4OAgzc/PFy6ifAIECJQtIGQou/+qJ0CAQFMCQoampN2HQMMCcSZDnM0QV5zJEGczuAgQIECgXAEhQ7m9VzkBAgSaFBAyNKntXgQaFIi3S8RbJuLa2toavWXCRYAAAQLlCggZyu29ygkQINCkgJChSW33ItCgwIsvvpju3r07uuNLL72Ubt++3eDd3YoAAQIEuiYgZOhaR8yHAAECwxQQMgyzr6oikG7cuJFu3bp1JOFcBouCAAECZQsIGcruv+oJECDQlICQoSlp9yHQgsCTTz6ZHj58OLpzHPwYZzM4ALKFRrglAQIEOiBQfYzu6tWr6Wtf+1oHZmUKBAgQIDA0ASHD0DqqHgIVgXi7RBwAOX7LxPr6etrY2GBEgAABAgUKfOITn0hf//rXR5Vfu3Ytvf766wUqKJkAAQIE6hYQMtQtbHwCLQtEqBBbZMeXoKHlhrg9AQIEWhLw1qGW4N2WAAEChQkIGQpruHLLFDgeNDifocx1oGoCBMoViB1tCwsLRwCHh4flYqicAAECBGoVEDLUymtwAt0RiN9cjh+bWFxcHL3W0vkM3emPmRAgQKBOgWrIEP/tj7DZRYAAAQIE6hAQMtShakwCHRTY399PceiX8xk62BxTIkCAQM0C1R1tq6uro6DZRYAAAQIE6hAQMtShakwCHRXw2ERHG2NaBAgQqFmg+t9/Z/PUjG14AgQIFC4gZCh8ASi/PIHqYxNea1le/1VMgECZAg59LLPvqiZAgEAbAkKGNtTdk0CLAl5r2SK+WxMgQKAFAYc+toDulgQIEChYQMhQcPOVXq6A11qW23uVEyBQnoBDH8vruYoJECDQpoCQoU199ybQooDzGVrEd2sCBAg0KHD16tV0586d0R2vXLmS7t271+Dd3YoAAQIEShMQMpTWcfUSqAhUz2d49tln0xe/+MXRb0BdBAgQIDAcgSeeeCK99957o4I+85nPpNdee204xamEAAECBDonIGToXEtMiEBzAvFay5dffvnoN59PP/10+vM///O0uLjY3CTciQABAgRqE6juWrt06VL67ne/W9u9DEyAAAECBEJAyGAdEChcoLqNdkwRIUO8Qz3ePuEiQIAAgf4KVHeseXVlf/to5gQIEOiTgJChT90yVwI1Cdy4cWP0vO54O+34Nqurqyl+UypsqAnesAQIEKhRwIGPNeIamgABAgQmCggZLA4CBI4E1tbWUjxCEb8xrV4RNETgIGywWAgQINAfgaWlpdF/0+Oyi6E/fTNTAgQI9F1AyND3Dpo/gcwC8RvSzc3No9+YjoePgCF+kxqPUggbMqMbjgABApkFqrsYYujDw8PMdzAcAQIECBA4WUDIYGUQIHCiQIQNsbPh+K6GCBnGYQM6AgQIEOimQPz3e3t7ezQ5uxi62SOzIkCAwFAFhAxD7ay6CGQSiN+kxs6Gk8IGh0NmQjYMAQIEMgoc38VwcHBgB1pGX0MRIECAwOkCQgYrhACBqQTiNWg7OzuPhQ0Oh5yKz4cIECDQmED1tZWx+2xvb6+xe7sRAQIECBAQMlgDBAhMLRB/OzY+r+H4zobY1eC8hqkpfZAAAQK1CNjFUAurQQkQIEBgBgEhwwxYPkqAwP8InHU4ZOxucBEgQIBA8wLVXQxxSG88KuEiQIAAAQJNCggZmtR2LwIDE3A45MAaqhwCBHotcHwXQ+wwE/r2uqUmT4AAgV4KCBl62TaTJtAtAYdDdqsfZkOAQJkCdjGU2XdVEyBAoGsCQoaudcR8CPRYIF6ZFrsbjp/X4HDIHjfV1AkQ6I3A3Nzc0Vy9trI3bTNRAgQIDE5AyDC4liqIQLsC48Mhx+9nr87m+vXr6dq1a6MDIl0ECBAgkE/gwx/+cPr2t789GvDSpUvpu9/9br7BjUSAAAECBGYQEDLMgOWjBAhMLzDpcMgYIQ4ji6DhhRde8Lzw9KQ+SYAAgccEItgd7yIb/2IEujdv3qRFgAABAgRaERAytMLupgTKEYiw4bd/+7fTw4cPJxY9Dh1WVlbscihnaaiUAIELCpwUMCwvL6fd3d0LjuzrBAgQIEDg/AJChvPb+SYBAjMIfPazn01vvfVWunv37qnfisAh/he7HGK3g0crZkD2UQIEihGIAHdpaemRer1Nopj2K5QAAQKdFhAydLo9JkdgeALjQyHjzIY33nhjdFDkaVf10YrxPw9PRUUECBCYXuCkgGFvb08oOz2hTxIgQIBAjQJChhpxDU2AwNkCETrE/+I3zbOGDt7/fravTxAgMCyBCGjjDIbxFeFr7GCw62tYfVYNAQIE+iwgZOhz98ydwAAFqoHD+J9PK9N5DgNcBEoiQOAxgfjv4Sc/+cn04MEDAYP1QYAAAQKdFhAydLo9JkeAQDV0OOm1mFUh5zlYLwQIDEFg/N+9+HHSDq/YuRCPSLgIECBAgEDXBIQMXeuI+RAgMFHAeQ4WBwECQxOY9ZGxqP/pp59O//Iv/zI0CvUQIECAwEAEhAwDaaQyCJQoMOtvzquHSDrPocQVo2YC7QvEf7fGu7KmOYdmPOP471eEC88880x67bXX2i/EDAgQIECAwAQBIYOlQYDAYASc5zCYViqEwCAEZg1Cq4GCV/kOYgkoggABAkUKCBmKbLuiCZQh4DyHMvqsSgJdEajuUtjc3Jx6WhEorKysjD4fZy14U8TUdD5IgAABAh0UEDJ0sCmmRIBAfgHnOeQ3NSKBkgWmOZzxJB8H1Ja8atROgACBMgSEDGX0WZUECBwTmHUbs/McLCEC5QrM+t+LqtR4l0L86CyYcteQygkQIFCSgJChpG6rlQCBiQLOc7A4CBAYC1zkcEaPPVhHBAgQIFC6gJCh9BWgfgIEThRwnoOFQaAMgfPuUvDYQxnrQ5UECBAgMLuAkGF2M98gQKAwAec5FNZw5Q5aYLxL4Z133jl6leQ0BVcPZ4zHHuLfXQQIECBAgMDjAkIGq4IAAQIzCsz6N5/Oc5gR2McJZBK4yOGM8YaHp556ytseMvXCMAQIECBQjoCQoZxeq5QAgZoEnOdQE6xhCcwgMGv4Nx56vCMhzlLw+sgZwH2UAAECBAhMEBAyWBoECBDILOA8h8yghiNwgkD1cMadnZ00fqzpLKzqYw9ChbO0/DoBAgQIEJhdQMgwu5lvECBAYGqB85zn8PTTT6cPfOAD6cqVK0fPfccfhlwEShW4yGMPESq88MILdimUunjUTYAAAQKNCwgZGid3QwIEShY475busdn4RPv49/jneGZ8vN17/GsOpCt5hfW/9lmDuWrF1V0KGxsb/cdQAQECBAgQ6KGAkKGHTTNlAgSGIzDreQ7TVn48eIjvxd/mjsOJ+NHuiGk1fa5OgepjD2+88Uba39+f6nbjA1UdzjgVlw8RIECAAIHGBIQMjVG7EQECBM4WiD9w/emf/mn6t3/7t3T58uWj58yn/YPX2Xd4/BN2R5xHzXfOI3DenTzjNeqxh/Oo+w4BAgQIENhDboMAAA62SURBVGhWQMjQrLe7ESBA4NwC423k4z+oxUDxz++8885RGFH9tXPfaMIX7Y7ILdrv8Y4ftFhdn+PKxuvxBz/4Qbpz587UBTuccWoqHyRAgAABAp0TEDJ0riUmRIAAgYsLnBRIxFb0cTARP9odcXHnLo8wbQhQrSECq+oaqYYFJ/18jvrtUsihaAwCBAgQINAdASFDd3phJgQIEGhcwO6IxsmPbtiXECC30HiXQvy4urqae3jjESBAgAABAi0LCBlaboDbEyBAoC8Cpe2OKDUEmHY9Hn+LSfXfx/8chzI+ePAgXbp0KV27ds1ho9Pi+hwBAgQIEOixgJChx80zdQIECHRRoGu7Ix4+fJh++qd/Oj377LOj8yvimiZA6KLtLHOaNgSojlk9d2Oan59lPj5LgAABAgQIlCEgZCijz6okQIBAJwXa3h3RJooQoE199yZAgAABAgTqEhAy1CVrXAIECBDIJtDG7gghQLb2GYgAAQIECBAoSEDIUFCzlUqAAIESBI4HEvfv308/+tGP0vgtBmEwKUA4/vMleKmRAAECBAgQIJBTQMiQU9NYBAgQIECAAAECBAgQIECgYAEhQ8HNVzoBAgQIECBAgAABAgQIEMgpIGTIqWksAgQIECBAgAABAgQIECBQsICQoeDmK50AAQIECBAgQIAAAQIECOQUEDLk1DQWAQIECBAgQIAAAQIECBAoWEDIUHDzlU6AAAECBAgQIECAAAECBHIKCBlyahqLAAECBAgQIECAAAECBAgULCBkKLj5SidAgAABAgQIECBAgAABAjkFhAw5NY1FgAABAgQIECBAgAABAgQKFhAyFNx8pRMgQIAAAQIECBAgQIAAgZwCQoacmsYiQIAAAQIECBAgQIAAAQIFCwgZCm6+0gkQIECAAAECBAgQIECAQE4BIUNOTWMRIECAAAECBAgQIECAAIGCBYQMBTdf6QQIECBAgAABAgQIECBAIKeAkCGnprEIECBAgAABAgQIECBAgEDBAkKGgpuvdAIECBAgQIAAAQIECBAgkFNAyJBT01gECBAgQIAAAQIECBAgQKBgASFDwc1XOgECBAgQIECAAAECBAgQyCkgZMipaSwCBAgQIECAAAECBAgQIFCwgJCh4OYrnQABAgQIECBAgAABAgQI5BQQMuTUNBYBAgQIECBAgAABAgQIEChYQMhQcPOVToAAAQIECBAgQIAAAQIEcgoIGXJqGosAAQIECBAgQIAAAQIECBQsIGQouPlKJ0CAAAECBAgQIECAAAECOQWEDDk1jUWAAAECBAgQIECAAAECBAoWEDIU3HylEyBAgAABAgQIECBAgACBnAJChpyaxiJAgAABAgQIECBAgAABAgULCBkKbr7SCRAgQIAAAQIECBAgQIBATgEhQ05NYxEgQIAAAQIECBAgQIAAgYIFhAwFN1/pBAgQIECAAAECBAgQIEAgp4CQIaemsQgQIECAAAECBAgQIECAQMECQoaCm690AgQIECBAgAABAgQIECCQU0DIkFPTWAQIECBAgAABAgQIECBAoGABIUPBzVc6AQIECBAgQIAAAQIECBDIKSBkyKlpLAIECBAgQIAAAQIECBAgULCAkKHg5iudAAECBAgQIECAAAECBAjkFBAy5NQ0FgECBAgQIECAAAECBAgQKFhAyFBw85VOgAABAgQIECBAgAABAgRyCggZcmoaiwABAgQIECBAgAABAgQIFCwgZCi4+UonQIAAAQIECBAgQIAAAQI5BYQMOTWNRYAAAQIECBAgQIAAAQIEChYQMhTcfKUTIECAAAECBAgQIECAAIGcAkKGnJrGIkCAAAECBAgQIECAAAECBQsIGQpuvtIJECBAgAABAgQIECBAgEBOASFDTk1jESBAgAABAgQIECBAgACBggWEDAU3X+kECBAgQIAAAQIECBAgQCCngJAhp6axCBAgQIAAAQIECBAgQIBAwQJChoKbr3QCBAgQIECAAAECBAgQIJBTQMiQU9NYBAgQIECAAAECBAgQIECgYAEhQ8HNVzoBAgQIECBAgAABAgQIEMgpIGTIqWksAgQIECBAgAABAgQIECBQsICQoeDmK50AAQIECBAgQIAAAQIECOQUEDLk1DQWAQIECBAgQIAAAQIECBAoWEDIUHDzlU6AAAECBAgQIECAAAECBHIKCBlyahqLAAECBAgQIECAAAECBAgULCBkKLj5SidAgAABAgQIECBAgAABAjkFhAw5NY1FgAABAgQIECBAgAABAgQKFhAyFNx8pRMgQIAAAQIECBAgQIAAgZwCQoacmsYiQIAAAQIECBAgQIAAAQIFCwgZCm6+0gkQIECAAAECBAgQIECAQE4BIUNOTWMRIECAAAECBAgQIECAAIGCBYQMBTdf6QQIECBAgAABAgQIECBAIKeAkCGnprEIECBAgAABAgQIECBAgEDBAkKGgpuvdAIECBAgQIAAAQIECBAgkFNAyJBT01gECBAgQIAAAQIECBAgQKBgASFDwc1XOgECBAgQIECAAAECBAgQyCkgZMipaSwCBAgQIECAAAECBAgQIFCwgJCh4OYrnQABAgQIECBAgAABAgQI5BQQMuTUNBYBAgQIECBAgAABAgQIEChYQMhQcPOVToAAAQIECBAgQIAAAQIEcgoIGXJqGosAAQIECBAgQIAAAQIECBQsIGQouPlKJ0CAAAECBAgQIECAAAECOQWEDDk1jUWAAAECBAgQIECAAAECBAoWEDIU3HylEyBAgAABAgQIECBAgACBnAJChpyaxiJAgAABAgQIECBAgAABAgULCBkKbr7SCRAgQIAAAQIECBAgQIBATgEhQ05NYxEgQIAAAQIECBAgQIAAgYIFhAwFN1/pBAgQIECAAAECBAgQIEAgp4CQIaemsQgQIECAAAECBAgQIECAQMECQoaCm690AgQIECBAgAABAgQIECCQU0DIkFPTWAQIECBAgAABAgQIECBAoGABIUPBzVc6AQIECBAgQIAAAQIECBDIKSBkyKlpLAIECBAgQIAAAQIECBAgULCAkKHg5iudAAECBAgQIECAAAECBAjkFBAy5NQ0FgECBAgQIECAAAECBAgQKFhAyFBw85VOgAABAgQIECBAgAABAgRyCggZcmoaiwABAgQIECBAgAABAgQIFCwgZCi4+UonQIAAAQIECBAgQIAAAQI5BYQMOTWNRYAAAQIECBAgQIAAAQIEChYQMhTcfKUTIECAAAECBAgQIECAAIGcAkKGnJrGIkCAAAECBAgQIECAAAECBQsIGQpuvtIJECBAgAABAgQIECBAgEBOASFDTk1jESBAgAABAgQIECBAgACBggWEDAU3X+kECBAgQIAAAQIECBAgQCCngJAhp6axCBAgQIAAAQIECBAgQIBAwQJChoKbr3QCBAgQIECAAAECBAgQIJBTQMiQU9NYBAgQIECAAAECBAgQIECgYAEhQ8HNVzoBAgQIECBAgAABAgQIEMgpIGTIqWksAgQIECBAgAABAgQIECBQsICQoeDmK50AAQIECBAgQIAAAQIECOQUEDLk1DQWAQIECBAgQIAAAQIECBAoWEDIUHDzlU6AAAECBAgQIECAAAECBHIKCBlyahqLAAECBAgQIECAAAECBAgULCBkKLj5SidAgAABAgQIECBAgAABAjkFhAw5NY1FgAABAgQIECBAgAABAgQKFhAyFNx8pRMgQIAAAQIECBAgQIAAgZwCQoacmsYiQIAAAQIECBAgQIAAAQIFCwgZCm6+0gkQIECAAAECBAgQIECAQE4BIUNOTWMRIECAAAECBAgQIECAAIGCBYQMBTdf6QQIECBAgAABAgQIECBAIKeAkCGnprEIECBAgAABAgQIECBAgEDBAkKGgpuvdAIECBAgQIAAAQIECBAgkFNAyJBT01gECBAgQIAAAQIECBAgQKBgASFDwc1XOgECBAgQIECAAAECBAgQyCkgZMipaSwCBAgQIECAAAECBAgQIFCwgJCh4OYrnQABAgQIECBAgAABAgQI5BQQMuTUNBYBAgQIECBAgAABAgQIEChYQMhQcPOVToAAAQIECBAgQIAAAQIEcgoIGXJqGosAAQIECBAgQIAAAQIECBQsIGQouPlKJ0CAAAECBAgQIECAAAECOQWEDDk1jUWAAAECBAgQIECAAAECBAoWEDIU3HylEyBAgAABAgQIECBAgACBnAJChpyaxiJAgAABAgQIECBAgAABAgULCBkKbr7SCRAgQIAAAQIECBAgQIBATgEhQ05NYxEgQIAAAQIECBAgQIAAgYIFhAwFN1/pBAgQIECAAAECBAgQIEAgp4CQIaemsQgQIECAAAECBAgQIECAQMECQoaCm690AgQIECBAgAABAgQIECCQU0DIkFPTWAQIECBAgAABAgQIECBAoGABIUPBzVc6AQIECBAgQIAAAQIECBDIKSBkyKlpLAIECBAgQIAAAQIECBAgULCAkKHg5iudAAECBAgQIECAAAECBAjkFBAy5NQ0FgECBAgQIECAAAECBAgQKFhAyFBw85VOgAABAgQIECBAgAABAgRyCggZcmoaiwABAgQIECBAgAABAgQIFCwgZCi4+UonQIAAAQIECBAgQIAAAQI5BYQMOTWNRYAAAQIECBAgQIAAAQIEChYQMhTcfKUTIECAAAECBAgQIECAAIGcAkKGnJrGIkCAAAECBAgQIECAAAECBQsIGQpuvtIJECBAgAABAgQIECBAgEBOASFDTk1jESBAgAABAgQIECBAgACBggWEDAU3X+kECBAgQIAAAQIECBAgQCCngJAhp6axCBAgQIAAAQIECBAgQIBAwQJChoKbr3QCBAgQIECAAAECBAgQIJBTQMiQU9NYBAgQIECAAAECBAgQIECgYAEhQ8HNVzoBAgQIECBAgAABAgQIEMgpIGTIqWksAgQIECBAgAABAgQIECBQsICQoeDmK50AAQIECBAgQIAAAQIECOQUEDLk1DQWAQIECBAgQIAAAQIECBAoWOD/ATkYgHA8bb8+AAAAAElFTkSuQmCC

[32-git]

Maybe this error output can help too?

<class 'numpy.ndarray'>
[recognize.py] to_recognize |             Showing the image of the expression...

[recognize.py] to_recognize |             Starting image preprocessing...

[preprocessing.py] treatment()
[preprocessing.py] resize_full_image()
[preprocessing.py] normalize()
[preprocessing.py] to_gray_denoise()
[preprocessing.py] invert()
[preprocessing.py] segment()
[preprocessing.py] segment() | contours founded:
1
[preprocessing.py] resize()
[preprocessing.py] segment() |             before line and sqrt processing
[preprocessing.py] segment() |             after line and sqrt processing
[preprocessing.py] segment() |             before resize
[preprocessing.py] segment() | after resize
[preprocessing.py] segment() | before border
[preprocessing.py] segment() | after border
[preprocessing.py] binarization()
[preprocessing.py] invert()
[preprocessing.py] invert()
[preprocessing.py] _255_to_1()
[recognize.py] to_recognize |             Showing preprocessed image

[recognize.py] to_recognize |             Image preprocessing finished.

[recognize.py] to_recognize |             Starting symbol classification...

... segmentation ...
... recognize ...
[recognize.py] to_recognize | Exception:
[api.py] recognize | Exception: Kernel shape must have the same length as input, but received kernel of shape (3, 3, 1, 32) and input of shape (None, None, 28, 28, 1).

[carolreis]

Oh, God. It reminds me I need to improve these logs urgently...

I'll make a test with your base64 and let you know what I found

[32-git]

Hi @carolreis

Were you able to make any progress with the base 64 string?

[mustafa-senyuz]

SAME PROBLEM WITH ME :((

PS I:\ML\MATH\mathoku> & "C:/Program Files/Python311/python.exe" i:/ML/MATH/mathoku/mathreader/test.py
2024-10-15 22:52:10.507499: I tensorflow/core/util/port.cc:153] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable TF_ENABLE_ONEDNN_OPTS=0.
2024-10-15 22:52:11.503528: I tensorflow/core/util/port.cc:153] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable TF_ENABLE_ONEDNN_OPTS=0.

['i:/ML/MATH/mathoku/mathreader/test.py']
[recognize.py] to_recognize | Showing the image of the expression...

[recognize.py] to_recognize | Starting image preprocessing...

[preprocessing.py] treatment()
[preprocessing.py] resize_full_image()
[preprocessing.py] normalize()
[preprocessing.py] to_gray_denoise()
[preprocessing.py] invert()
[preprocessing.py] segment()
[preprocessing.py] segment() | contours founded:
4
[preprocessing.py] resize()
[preprocessing.py] segment() | before line and sqrt processing
hor 4
[preprocessing.py] segment() | after line and sqrt processing
[preprocessing.py] segment() | before resize
[preprocessing.py] segment() | after resize
[preprocessing.py] segment() | before border
[preprocessing.py] segment() | after border
[preprocessing.py] binarization()
[preprocessing.py] invert()
[preprocessing.py] invert()
[preprocessing.py] _255_to_1()
[preprocessing.py] resize()
[preprocessing.py] segment() | before line and sqrt processing
hor 3
[preprocessing.py] segment() | after line and sqrt processing
[preprocessing.py] segment() | before resize
[preprocessing.py] segment() | after resize
[preprocessing.py] segment() | before border
[preprocessing.py] segment() | after border
[preprocessing.py] binarization()
[preprocessing.py] invert()
[preprocessing.py] invert()
[preprocessing.py] _255_to_1()
[preprocessing.py] resize()
[preprocessing.py] segment() | before line and sqrt processing
[preprocessing.py] segment() | after line and sqrt processing
[preprocessing.py] segment() | before resize
[preprocessing.py] segment() | after resize
[preprocessing.py] segment() | before border
[preprocessing.py] segment() | after border
[preprocessing.py] binarization()
[preprocessing.py] invert()
[preprocessing.py] invert()
[preprocessing.py] _255_to_1()
[preprocessing.py] resize()
[preprocessing.py] segment() | before line and sqrt processing
[preprocessing.py] segment() | after line and sqrt processing
[preprocessing.py] segment() | before resize
[preprocessing.py] segment() | after resize
[preprocessing.py] segment() | before border
[preprocessing.py] segment() | after border
[preprocessing.py] binarization()
[preprocessing.py] invert()
[preprocessing.py] invert()
[preprocessing.py] _255_to_1()
[recognize.py] to_recognize | Showing preprocessed image

[recognize.py] to_recognize | Image preprocessing finished.

[recognize.py] to_recognize | Starting symbol classification...

... segmentation ...
... recognize ...
C:\Users\Mustafa\AppData\Roaming\Python\Python311\site-packages\keras\src\layers\convolutional\base_conv.py:107: UserWarning: Do not pass an input_shape/input_dim argument to a layer. When using Sequential models, prefer using an Input(shape) object as the first layer in the model instead.
super().init(activity_regularizer=activity_regularizer, **kwargs)
[recognize.py] to_recognize | Exception:
[api.py] recognize | Exception: Kernel shape must have the same length as input, but received kernel of shape (3, 3, 1, 32) and input of shape (None, None, 28, 28, 1).
Traceback (most recent call last):
File "i:\ML\MATH\mathoku\mathreader\test.py", line 50, in
Example()
File "i:\ML\MATH\mathoku\mathreader\test.py", line 37, in init
expression, img = hme_recognizer.recognize()
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "I:\ML/MATH/mathoku\mathreader\api.py", line 109, in recognize
raise e
File "I:\ML/MATH/mathoku\mathreader\api.py", line 91, in recognize
expression, image = hme.to_recognize()
^^^^^^^^^^^^^^^^^^
File "I:\ML/MATH/mathoku\mathreader\recognize.py", line 85, in to_recognize
raise e
File "I:\ML/MATH/mathoku\mathreader\recognize.py", line 64, in to_recognize
reconhecer = self.__recognize(s['image'])
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "I:\ML/MATH/mathoku\mathreader\recognize.py", line 35, in __recognize
return classification.fit(img)
^^^^^^^^^^^^^^^^^^^^^^^
File "I:\ML/MATH/mathoku\mathreader\classification\classification.py", line 35, in fit
raise e
File "I:\ML/MATH/mathoku\mathreader\classification\classification.py", line 21, in fit
model = load_model(config.package_path + '/ann_models/model/model_11-07-2020_23-54-57.h5')
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Mustafa\AppData\Roaming\Python\Python311\site-packages\keras\src\saving\saving_api.py", line 196, in load_model
return legacy_h5_format.load_model_from_hdf5(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Mustafa\AppData\Roaming\Python\Python311\site-packages\keras\src\legacy\saving\legacy_h5_format.py", line 133, in load_model_from_hdf5
model = saving_utils.model_from_config(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Mustafa\AppData\Roaming\Python\Python311\site-packages\keras\src\legacy\saving\saving_utils.py", line 85, in model_from_config
return serialization.deserialize_keras_object(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Mustafa\AppData\Roaming\Python\Python311\site-packages\keras\src\legacy\saving\serialization.py", line 495, in deserialize_keras_object
deserialized_obj = cls.from_config(
^^^^^^^^^^^^^^^^
File "C:\Users\Mustafa\AppData\Roaming\Python\Python311\site-packages\keras\src\models\sequential.py", line 359, in from_config
model.add(layer)
File "C:\Users\Mustafa\AppData\Roaming\Python\Python311\site-packages\keras\src\models\sequential.py", line 122, in add
self._maybe_rebuild()
File "C:\Users\Mustafa\AppData\Roaming\Python\Python311\site-packages\keras\src\models\sequential.py", line 141, in _maybe_rebuild
self.build(input_shape)
File "C:\Users\Mustafa\AppData\Roaming\Python\Python311\site-packages\keras\src\layers\layer.py", line 226, in build_wrapper
original_build_method(*args, **kwargs)
File "C:\Users\Mustafa\AppData\Roaming\Python\Python311\site-packages\keras\src\models\sequential.py", line 187, in build
x = layer(x)
^^^^^^^^
File "C:\Users\Mustafa\AppData\Roaming\Python\Python311\site-packages\keras\src\utils\traceback_utils.py", line 122, in error_handler
raise e.with_traceback(filtered_tb) from None
File "C:\Users\Mustafa\AppData\Roaming\Python\Python311\site-packages\keras\src\ops\operation_utils.py", line 184, in compute_conv_output_shape
raise ValueError(
ValueError: Kernel shape must have the same length as input, but received kernel of shape (3, 3, 1, 32) and input of shape (None, None, 28, 28, 1).