You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Machine Learning Practical - Coursework 2: Analysing problems with the VGG deep neural network architectures (with 8 and 38 hidden layers) on the CIFAR100 dataset by monitoring gradient flow during training. And exploring solutions using batch normalization and residual connections.
A model based on CNN (Convolutional Neural Networks) in order to classify the CIFAR100 dataset, composed by 60,000 32x32 color images labeled over 100 categories.
An ablation study conducted on the CIFAR-100 dataset. Three deep learning architectures are utilized: Convolutional Neural Networks (CNN), Gated Multilayer Perceptrons (gMLP), and Vision Transformers (ViT). The study is built using the PyTorch framework as well as PyTorch Lightning for clean and concise code and Optuna for hyperparameter tuning.