Course Outline
Module I: Fundamentals of Large Language Models
1. Mechanisms and architecture of generative models
2. Key concepts – tokens, temperature, and other LLM parameters
3. Context windows and their limitations
4. The hallucination phenomenon and strategies for minimizing it
5. Principles of effective prompt engineering
6. Prompt formatting techniques (few-shot learning, chain-of-thought)
7. The ChatGPT interface – capabilities and limitations
8. The OpenAI platform – playground, models, and API key management
9. Differences between APIs and SDKs
10. Overview of available models and their applications
Module II: Communicating with Models in Python
1. Basics of API requests using the requests library
2. The official OpenAI SDK
3. Handling API responses and formatting results
4. Structured outputs – enforcing a specific response structure
5. Real-time streaming of responses
6. The LangChain framework
7. OpenRouter as an aggregator for accessing various models
Module III: Vector Representation of Text
1. The concept of text embeddings
2. How models understand meaning – vector space
3. Embedding APIs in OpenAI
4. Measuring semantic similarity between texts
5. Practical applications of embeddings
Module IV: Practical Applications of LLMs
1. Automatic document summarization
2. Extracting key information from text
3. Machine translation using LLMs
4. Text classification – sentiment analysis and categorization
Requirements
