Large Language Models
GPT, LLaMA, Claude - architecture, training, and applications.
Course Duration: 12 hours
What You'll Learn
- LLM architectures
- Pre-training and scaling
- Instruction tuning
- RLHF (Reinforcement Learning from Human Feedback)
- Open-source vs proprietary models
- LLM applications
Prerequisites
- Transformers & Attention
- NLP fundamentals
Course Modules
- Introduction to LLMs
- GPT Architecture
- Scaling Laws
- Pre-training Process
- Instruction Tuning
- RLHF
- GPT-4 and Capabilities
- Claude and Constitutional AI
- LLaMA and Open Models
- Mistral and Mixtral
- LLM Benchmarks
- Building Applications with LLMs
Models Covered
- GPT-3.5 / GPT-4
- Claude 2 / Claude 3
- LLaMA 2 / LLaMA 3
- Mistral / Mixtral
- Gemini