Open Source Generative AI
This hands-on course teaches participants how to design, build, and deploy practical AI applications using modern Transformer-based architectures. Learners gain experience developing Transformer models from scratch, training and optimizing them with PyTorch, and working with pre-trained open-source large language models. The course also covers prompt engineering, hardware considerations (CPU vs. GPU), fine-tuning, quantization, and real-world AI application development.
Through guided labs and access to a GPU-accelerated environment, participants work with open-source frameworks such as LLaMA to build, optimize, and deploy AI-powered applications securely and efficiently. The course concludes with applied exercises that push model limits and reinforce performance optimization strategies. Successful participants receive an AI certification from Alta3 Research.
By the end of this course, participants will be able to:
- Design, train, and optimize Transformer models using PyTorch
- Understand AI architectures with a focus on Transformer-based models
- Explain tokenization, word embeddings, and positional encoding
- Build real-world AI web applications using open-source LLM frameworks
- Apply advanced prompt engineering techniques
- Install, configure, and use frameworks such as LLaMA
- Fine-tune and quantize models to improve efficiency and performance
- Compare CPU and GPU hardware acceleration and select appropriate infrastructure
- Maximize model performance through testing and optimization
- Distinguish between chat-based and instruction-based AI interaction modes
- Python developers interested in AI and machine learning
- Application developers building AI-enabled systems
- DevSecOps engineers working with AI workloads
- Architects designing AI solutions
- Project managers, managers, or directors overseeing AI initiatives
- Data acquisition specialists supporting AI projects
- Python programming experience or equivalent
- Familiarity with Linux environments
- Basic understanding of software development concepts
- Prior AI or machine learning experience is helpful but not required