Tokenization and text representation · MLD-0924
Pre-tokenization
Splitting input into preliminary units before applying a subword algorithm.
Original reference page · detailed expansion pending. Original definitions, examples, notes and citations are preserved below. New examples are illustrative, not benchmark results.
Original example
Separate words or punctuation before BPE.
Original caveat
Pre-tokenization rules affect what later merges are possible.
Context and aliases
Level: Intermediate. Stages: Data · Training · Inference.
Tags: NLP · SLM · LLM.
Aliases: None recorded.
Related concepts
Original references
- Hugging Face: chat templates ↗ · Reference link
- Transformers glossary ↗ · Consulted
- Hugging Face: tokenization algorithms ↗ · Consulted
- Original reference URL ↗