DAY 111 / 210
Foundations of LLM Inference Pipelines
This first day of phase-3 establishes core mental models for production inference before optimization layers are added. It directly precedes later days that will extend the existing app/maku routes into inference endpoints.
⏱ 40 min target📝 2 quiz Qs
Resources
- 25 min
Deliverable
Journal entry with 200-word summary of pipeline stages plus one concrete local inference test command that runs successfully
Quiz · 2 questions
1. Which component is responsible for converting raw text into token IDs before model forward pass?
2. Name one common misconception when first running inference on consumer GPUs and state the actual limiting factor.