n4nAI

Topic

Knowledge Distillation

7 posts on knowledge distillation — part of glossary on the n4n AI blog.

GlossaryDefinition

What is knowledge distillation in machine learning?

A practitioner's guide to knowledge distillation — how teacher-student training compresses models, when to use it, and where it falls short.

6 min read
GlossaryComparison

Knowledge distillation vs quantization: the difference

A practitioner's comparison of knowledge distillation vs quantization for LLM compression — covering capabilities, latency, cost, ergonomics, and when to use each.

5 min read
GlossaryDefinition

Knowledge distillation explained: teacher and student

Knowledge distillation explained: train a smaller student model to mimic a larger teacher's output distributions, preserving capability while cutting inference cost.

5 min read
GlossaryAnalysis

How Llama 3.2 1B was distilled from larger Llama models

Technical deep-dive into Meta's Llama 3.2 1B distillation process, covering teacher-student architecture, logit matching, and deployment tradeoffs for edge inference.

5 min read
GlossaryAnalysis

How Gemma 2 uses knowledge distillation

Deep dive into Gemma 2's knowledge distillation pipeline — teacher-student architecture, loss functions, and practical tradeoffs for deploying smaller models.

7 min read
GlossaryAnalysis

How DistilBERT was distilled from BERT

A technical deep-dive into DistilBERT's distillation process from BERT, covering the triple loss function, architectural changes, and practical tradeoffs for engineers.

6 min read
GlossaryComparison

Distillation vs pruning: two ways to shrink a model

A practitioner's comparison of distillation vs pruning for LLM compression — covering quality, latency, tooling, and when to use each.

7 min read