Coding Master Class (PyTorch)

CLASS INFO

• Semester: Summer, 2024
• Class time/ Class room: 14:00-17:00/ Mirae Building 607
• Text book:

CONTENTS

• Building AI Models 1 (Fundamentals)
  1. Tensor: storing data and passing it to an AI model
  2. Layer: transforming and extracting information from data
  3. Custom Model: building a custom AI model
  4. RNN: building a nonlinear time-series model using an RNN
  5. FinBert: understanding text embeddings
  6. FinBert: understanding fine-tuning (LoRA)


• Building AI Models 2 (Sequence Inference)
  1. Layer: combining layers
  2. Layer: RNN cell
  3. Layer: Encoder
  4. Layer: Decoder (under construction)
  5. Seq2Seq
  6. Transformer (under construction)


• Building AI Models 3 (Large Language Model)
  1. Loading a GPT2 model (under construction)
  2. Modifying the GPT2 generate function (under construction)


• Fitting and Running AI Models
  1. Models: running in the background
  2. Models: training and monitoring
  3. Models: saving and loading models


• Building an AI Model Pipeline
  1. Development: development environment
  2. Development: remote development environment
  3. Development: automatic execution


• Organizing research materials
  1. PostgreSQL
  2. Flask + Apache + Conda