After doing these courses, I feel confident creating professional visualizations and dashboards
Shoebur RahmanProcurement Specialist, Giz Bangladesh
Course Description
Deep-Dive into the Transformer Architecture
Transformer models have revolutionized text modeling, kickstarting the generative AI boom by enabling today's large language models (LLMs). In this course, you'll look at the key components in this architecture, including positional encoding, attention mechanisms, and feed-forward sublayers. You'll code these components in a modular way to build your own transformer step-by-step.Implement Attention Mechanisms with PyTorch
The attention mechanism is a key development that helped formalize the transformer architecture. Self-attention allows transformers to better identify relationships between tokens, which improves the quality of generated text. Learn how to create a multi-head attention mechanism class that will form a key building block in your transformer models.Build Your Own Transformer Models
Learn to build encoder-only, decoder-only, and encoder-decoder transformer models. Learn how to choose and code these different transformer architectures for different language tasks, including text classification and sentiment analysis, text generation and completion, and sequence-to-sequence translation.Prerequisites
Curriculum
Course outline
1
The Building Blocks of Transformer Models
Discover what makes the hottest deep learning architecture in AI tick! Learn about the components that make up Transformer models, including the famous self-attention mechanisms described in the renowned paper "Attention is All You Need."
- Transformers with PyTorch50 XP
- Breaking down the Transformer50 XP
- PyTorch Transformers100 XP
- Embedding and positional encoding50 XP
- Creating input embeddings100 XP
- Creating positional encodings100 XP
- Multi-head self-attention50 XP
- Implementing multi-head attention100 XP
- Starting the MultiHeadAttentionClass100 XP
- Adding methods to the MultiHeadAttention class100 XP
2
Building Transformer Architectures
Design transformer encoder and decoder blocks, and combine them with positional encoding, multi-headed attention, and position-wise feed-forward networks to build your very own Transformer architectures. Along the way, you'll develop a deep understanding and appreciation for how transformers work under the hood.
Transformer Models with PyTorch
Course
Complete
