Module embedding_normalizer 

Vector normalization and transformation for embeddings.

Provides various normalization strategies (L1, L2, LInf, MinMax, ZScore, UnitVariance) for embedding vectors, along with clipping, cosine similarity, and statistics tracking.

Structs

EmbeddingNormalizer

Embedding normalizer with configurable strategies and statistics tracking.

NormStats

Statistics about a single normalization operation.

NormalizerConfig

Configuration for the normalizer.

NormalizerStats

Aggregate statistics across multiple normalization operations.

Enums

NormalizationType

The type of normalization to apply.