Decoding and inference behavior · MLD-0288
Decoding
Choosing outputs from a model's predicted scores or distributions according to a search or sampling rule.
Original reference page · detailed expansion pending. Original definitions, examples, notes and citations are preserved below. New examples are illustrative, not benchmark results.
Original example
Turn next-token probabilities into a response.
Original caveat
The same weights can behave differently under different decoding rules.
Context and aliases
Level: Intermediate. Stages: Inference · Evaluation · Deployment.
Tags: SLM · LLM · NLP.
Aliases: Generation decoding.
Related concepts
Concepts linking here
Original references
- Hugging Face: generation strategies ↗ · Reference link
- Fast Inference from Transformers via Speculative Decoding ↗ · Consulted
- vLLM documentation ↗ · Consulted
- Original reference URL ↗