Chapter 6 Complete
You have learned
Training & Inference
You now understand the entire lifecycle of an LLM:
From training through RLHF to sampling strategies
and inference optimizations like Quantization and Speculative Decoding.
Training Loss Curves
RLHF
DPO
Sampling Settings
Top-K & Top-P
Quantization
Speculative Decoding
Continue with Chapter 7
Trends & Future
Look at the evolution of the LLM landscape:
Benchmark evolution, emergent capabilities,
and the scaling of attention complexity.