Sandeep Bazar 365 Days of AI
Labs

Courses › Course05

Course05 · Deep Learning

Neural networks from first principles to transformers: build them, train them, debug them, and understand the hardware they run on.

42 days 14 written 0 complete

Days 197-203 · Neural Network Foundations 0/7 complete
  1. Day 197 The Perceptron planned
  2. Day 198 Activation Functions planned
  3. Day 199 Forward Propagation planned
  4. Day 200 Backpropagation planned
  5. Day 201 A Neural Network in Pure NumPy planned
  6. Day 202 PyTorch Tensors planned
  7. Day 203 Training MNIST from Scratch planned

Project: MNIST from Scratch — Implement a two-layer neural network in pure NumPy that reaches at least 95% accuracy on MNIST digits.

Days 204-210 · Training Deep Networks 0/7 complete
  1. Day 204 PyTorch: autograd and nn.Module planned
  2. Day 205 Datasets and DataLoaders planned
  3. Day 206 Optimizers: SGD to Adam planned
  4. Day 207 Learning Rate Schedules planned
  5. Day 208 Dropout, Batch Norm, and Regularization planned
  6. Day 209 Debugging Training Runs planned
  7. Day 210 A Disciplined Training Project planned

Project: Fashion-MNIST Classifier — Train a PyTorch classifier with a proper training loop, LR schedule, and regularization, beating a stated baseline.

Days 211-217 · Convolutional Networks and Vision 0/7 complete
  1. Day 211 Convolutions planned
  2. Day 212 CNN Architectures planned
  3. Day 213 Transfer Learning planned
  4. Day 214 Data Augmentation planned
  5. Day 215 Training a Vision Model End to End planned
  6. Day 216 Beyond Classification: Detection and Segmentation planned
  7. Day 217 Your Own Image Classifier planned

Project: Custom Image Classifier — Fine-tune a pretrained CNN on your own small image dataset with augmentation and an error-case gallery.

Days 218-224 · Sequences and Text 0/7 complete
  1. Day 218 Text Preprocessing and Tokenization planned
  2. Day 219 Word Embeddings planned
  3. Day 220 Recurrent Neural Networks planned
  4. Day 221 LSTMs and GRUs planned
  5. Day 222 Sequence-to-Sequence and Early Attention planned
  6. Day 223 Text Classification with Embeddings planned
  7. Day 224 A Sentiment Analysis Project planned

Project: Sentiment Analyzer — Build a review-sentiment model comparing bag-of-words, embeddings, and a small recurrent network.

Days 225-231 · Transformers 0/7 complete
  1. Day 225 “Attention Is All You Need” planned
  2. Day 226 Self-Attention, Step by Step planned
  3. Day 227 The Transformer Architecture planned
  4. Day 228 Encoder Models: BERT and Friends planned
  5. Day 229 Decoder Models: The GPT Family planned
  6. Day 230 Hugging Face Transformers in Practice planned
  7. Day 231 Fine-Tuning a Small Transformer planned

Project: Fine-Tuned Transformer — Fine-tune a small pretrained transformer on a text classification task with Hugging Face and report results honestly.

Days 232-238 · Training at Scale 0/7 complete
  1. Day 232 GPUs and AI Hardware planned
  2. Day 233 Mixed Precision and Performance planned
  3. Day 234 Distributed Training Concepts planned
  4. Day 235 Experiment Tracking planned
  5. Day 236 Quantization and Distillation planned
  6. Day 237 Scaling Laws and What They Bought Us planned
  7. Day 238 Section Project: Reproducing a Paper planned

Project: Section Project: Reproduce a Result — Reproduce a small published deep-learning result end to end, with experiment tracking and a reproduction report.