iusztinpaul/hands-on-llms
An educational course teaching how to design, train, and deploy a real-time financial advisor LLM system using QLoRA fine-tuning, streaming pipelines, and vector databases.
Velocity · 7d
+3.2
★ / day
Trend
→steady
star history
This hands-on course covers building a complete LLM system with three pipelines: a training pipeline using QLoRA fine-tuning on an open-source LLM, a streaming real-time pipeline using Beam and Bytewax, and an inference pipeline deployed on AWS. The curriculum includes video lectures, articles, and source code covering LLMOps practices, vector database integration with Qdrant, experiment tracking via Comet ML, and orchestration with LangChain.