Used custom parallel transformer's Encoder architecture to train model on News title & News description separately.
-
Updated
Aug 24, 2024 - Jupyter Notebook
#
model-subclassing
Here are 7 public repositories matching this topic...
BIG PROJECT "The model is trained on images gathered from the COCO dataset, specifically selected to contain only one person in each image."
-
Updated
Sep 25, 2024 - Jupyter Notebook
Trained simple feed forward neural network on MNIST and F-MNIST Dataset. using model subclassing and custom training loop
-
Updated
May 30, 2026 - Jupyter Notebook
Used CNN approach to train model on MNIST and F-MNIST dataset.
-
Updated
Jul 12, 2024 - Jupyter Notebook
Implementation of a custom TensorFlow/Keras model by subclassing tf.keras.Model, including custom layers, a user-defined loss function, and full control over the forward pass. Demonstrates dataset loading, preprocessing, and training on MNIST using the Model Subclassing API for maximum flexibility
python api deep-learning interactive tensorflow numpy keras classification model-subclassing lowlevelapi customsubclassmodel low-level-api-tensorflow custom-layar--model
-
Updated
Aug 17, 2025 - Jupyter Notebook
ResNet-style classifier for Fashion-MNIST built with custom residual layers, model subclassing and a custom training loop (tf.GradientTape) in TensorFlow 2. Reaches 82.76% test accuracy training on only 5,000 images.
python deep-learning tensorflow keras jupyter-notebook image-classification resnet residual-networks fashion-mnist gradient-tape model-subclassing custom-training-loop
-
Updated
Oct 13, 2026 - Jupyter Notebook
Tensorflow 2: for Deep Learning Specialization | Imperial College London
-
Updated
Sep 18, 2023 - Jupyter Notebook
Add this topic to your repo
To associate your repository with the model-subclassing topic, visit your repo's landing page and select "manage topics."