LLMs on Supercomputers¶

Comprehensive course covering prompt engineering, RAG, and fine-tuning. Adapted from TU Wien's AI Factory Austria training materials.

Collection Statistics

Total Notebooks: 15

D0 - Setup¶

Bazzite AI environment setup and configuration

#	Notebook	Description
1	Bazzite-AI Environment Setup

LangChain basics, prompt templates, chaining, evaluation, and optimization

#	Notebook	Description
1	Prompt Engineering Essentials
2	Prompt templates and parsing
3	Chaining
4	LLM Evaluation with evidently.ai
5	LLM as a Judge with evidently.ai
6	Prompt Optimization with Evidently

RAG fundamentals with basic tools and ChromaDB vector databases

#	Notebook	Description
1	RAG Introduction with Ollama OpenAI API
2	RAG with LangChain and ChromaDB using Ollama

Transformer architecture, PyTorch/HuggingFace fine-tuning, quantization, PEFT, and Unsloth

#	Notebook	Description
1	Transformer Anatomy
2	Fine-tuning an LLM with plain PyTorch
3	Fine-tuning an LLM with Hugging Face Trainer
4	Quantization
5	PEFT
6	Once Unsloth updates to support vLLM 0.14.x, enable fast_inference: