Member Junction
    Preparing search index...

    Interface FixedWindowSegmentationOptions

    Options specific to FixedWindowSegmenter.

    interface FixedWindowSegmentationOptions {
        MaxSegmentTokens?: number;
        MinSegmentTokens?: number;
        OverlapTokens?: number;
        TextStrategy?: "sentence" | "paragraph" | "fixed";
        WindowMs?: number;
        WindowOverlapMs?: number;
    }

    Hierarchy (View Summary)

    Index

    Properties

    MaxSegmentTokens? MinSegmentTokens? OverlapTokens? TextStrategy? WindowMs? WindowOverlapMs?

    Properties

    MaxSegmentTokens?: number

    Hard ceiling on tokens per text segment. The base class splits any oversized segment via TextChunker so no segmenter can exceed it. Default: 512.

    MinSegmentTokens?: number

    Segments whose text estimates below this many tokens are merged forward into the next segment, preventing a spray of near-empty vectors. Default: 0 (off).

    OverlapTokens?: number

    Overlap tokens applied when an oversized segment must be split. Default: 10% of max.

    TextStrategy?: "sentence" | "paragraph" | "fixed"

    Text splitting strategy handed to TextChunker. Default: 'sentence'.

    WindowMs?: number

    Window length in ms for audio/video with no transcript.

    Defaults to 30000 (30s). Multimodal embedders sample a bounded number of frames per call, giving an effective window of roughly half a minute regardless of clip length — windows much longer than this are silently under-sampled, so the default deliberately stays under that ceiling.

    WindowOverlapMs?: number

    Overlap between consecutive AV windows, in ms. Default: 0.

    Settings

    Member Visibility

    On This Page

    Properties