Issue running mnist_conv.py - Process finished with exit code 132 (interrupted by signal 4:SIGILL)

[schultz77]

Hi,
Amazing tutorial again. Unfortunately i am facing some problems with the code

my system:

Python 3.12.3 [GCC 11.4.0] on linux
I am running it in an old Laptop
linux kernel 5.15.0-102-generic, processor Intel Pentium CPU N3540 @ 2.16GHz × 4, memory 7.6 GiB

in order to get free of errors i had to modify the code:
from:

from keras.utils import np_utils

to:

from keras.utils import to_categorical

After running it i've got following error:

File "../lib/python3.12/site-packages/keras/src/utils/tree.py", line 12, in

from tensorflow.python.trackable.data_structures import ListWrapper

ModuleNotFoundError: No module named 'tensorflow'

After installing tensorflow and running the script:

Process finished with exit code 132 (interrupted by signal 4:SIGILL)

Could you please support me with this?
i am really looking forward to debug the script to get deeper understanding of it.

Thank You in advance!

[schultz77]

I am trying to load the mnist data without tensorflow but i am have save errors with reshape:

def load_mnist():
    base_url = 'http://yann.lecun.com/exdb/mnist/'
    files = ['train-images-idx3-ubyte.gz', 'train-labels-idx1-ubyte.gz',
             't10k-images-idx3-ubyte.gz', 't10k-labels-idx1-ubyte.gz']

    data = {}

    # Download and load training and testing data
    for file in files:
        url = base_url + file
        filename = os.path.join('data', file)
        if not os.path.exists('data'):
            os.makedirs('data')
        if not os.path.exists(filename):
            print('Downloading', file, '...')
            r = requests.get(url)
            with open(filename, 'wb') as f:
                f.write(r.content)

        with gzip.open(filename, 'rb') as f:
            if file.endswith('images-idx3-ubyte.gz'):
                # data[file] = np.frombuffer(f.read(), np.uint8, offset=16).reshape(-1, 28, 28)
                data[file] = np.frombuffer(f.read(), np.uint8, offset=16)
            elif file.endswith('labels-idx1-ubyte.gz'):
                data[file] = np.frombuffer(f.read(), np.uint8, offset=8)

    return data


# Load MNIST dataset 
mnist = load_mnist()

# Extract training data and labels
x_train, y_train = mnist['train-images-idx3-ubyte.gz'], mnist['train-labels-idx1-ubyte.gz']
x_test, y_test = mnist['t10k-images-idx3-ubyte.gz'], mnist['t10k-labels-idx1-ubyte.gz']

Do you have any hint? Thx!

[schultz77]

It runs using this approach. However, I am not 100% sure:

import numpy as np
# from keras.datasets import mnist
# from keras.utils import to_categorical

from sklearn.datasets import fetch_openml

from dense import Dense
from convolutional import Convolutional
from reshape import Reshape
from activations import Sigmoid
from losses import binary_cross_entropy, binary_cross_entropy_prime
from network import train, predict


def to_one_hot(y_value):
    num_classes = np.max(y_value) + 1
    one_hot_matrix = np.eye(num_classes)[y_value.reshape(-1)]
    return one_hot_matrix.reshape(list(y_value.shape) + [num_classes])


def preprocess_data(x_value, y_value, limit):
    zero_indices = np.where(y_value.values == 0)[0][:limit]
    one_indices = np.where(y_value.values == 1)[0][:limit]

    all_indices = np.hstack((zero_indices, one_indices))
    all_indices = np.random.permutation(all_indices)

    x_value = x_value.iloc[all_indices].values.astype("float32") / 255
    x_value = x_value.reshape(len(x_value), 1, 28, 28)

    y_value = y_value.iloc[all_indices].values
    y_value = to_one_hot(y_value)
    y_value = y_value.reshape(len(y_value), 2, 1)

    return x_value, y_value


# Load MNIST data from OpenML
mnist = fetch_openml('mnist_784')

# Separate features and labels
x_data, y_data = mnist['data'], mnist['target']

# Convert string labels to integers
y_data = y_data.astype(int)

# Preprocess the data
x_train, y_train = preprocess_data(x_data[:60000], y_data[:60000], 100)
x_test, y_test = preprocess_data(x_data[60000:], y_data[60000:], 100)

# Print shapes of training and testing data and labels
print("Training data shape:", x_train.shape)
print("Training labels shape:", y_train.shape)
print("Testing data shape:", x_test.shape)
print("Testing labels shape:", y_test.shape)