A free, hands-on LLM programming tutorial series derived from Shanghai Jiao Tong University courses, covering topics from prompt engineering and fine-tuning to AI agents and jailbreak attacks. Learners interact with the material through slides and Jupyter Notebooks, requiring existing programming skills.
Project overview
Covers eight distinct LLM topics across prompt engineering, model training, security, and agents with slides, guides, and runnable Jupyter Notebooks developed with university instructors.
Project type
AI Agent · Model Development · Infrastructure · Prompt Engineering · Learning Resources
Use cases
Learning & Education
Deployment
Refer to project documentation
License
License pending
Best for
Educators, researchers, and developers seeking structured, academic-origin tutorials on LLM programming.
Key capabilities
Contains slides, guides, and Jupyter notebooks teaching how to select, fine-tune, and deploy pre-trained models.
Contains slides, guides, and Jupyter notebooks covering LLM API calls and prompt-based reasoning.
Contains slides, guides, and Jupyter notebooks demonstrating how to edit and validate specific knowledge within language models.
Contains slides, guides, and Jupyter notebooks on distilling mathematical reasoning capabilities into models.
Contains slides, guides, and Jupyter notebooks detailing how to embed invisible text watermarks into language model generations.
Contains slides, guides, and Jupyter notebooks explaining how jailbreak attacks work against LLMs.
Contains slides, guides, and Jupyter notebooks explaining multimodal understanding and generation.
Contains slides, guides, and Jupyter notebooks covering AI agents that automate desktop and GUI tasks.
Limitations and risks
All tips provided in the tutorials are for reference only, and correctness is not fully guaranteed.
GPU, minimum hardware, telemetry, cost dependencies, and data boundary configurations are not documented.
Getting started
Users clone the repository, open the desired Jupyter Notebook, and execute the notebook cells. The setup difficulty is medium because it requires running Jupyter Notebooks and following tutorial guides.
Alternatives and comparisons
Provides educational lecture materials and runnable Python code demonstrating how AI agents work, the mechanics of prompt injection attacks, and defense patterns.