mlx-bun library API
    Preparing search index...

    Interface SchedulerMetrics

    interface SchedulerMetrics {
        active: number;
        at: number;
        capacity: number;
        model: string;
        queued: number;
        tokensPerSecond: number;
    }
    Index
    active at capacity model queued tokensPerSecond
    active: number

    Rows decoding now.

    at: number
    capacity: number

    The batch capacity.

    model: string
    queued: number

    Requests waiting for admission or prefill.

    tokensPerSecond: number

    Settings
    Member Visibility

    On This Page
    Properties