mlx-bun library API
    Preparing search index...

    Interface LoadContext

    interface LoadContext {
        path: string;
        prototype: Cache;
        arr(slot: TensorSlot): MlxArray;
        grownArr(slot: TensorSlot, offset: number): MlxArray;
        triple(slots: TensorSlot[], at: number): QuantizedTensor;
    }
    Index
    path prototype
    arr grownArr triple
    path: string
    prototype: Cache

    The restoring graph's own empty cache for this entry's layer (from its makeCache()), borrowed for the load: what the graph composed into its caches (the TurboQuant read a KVCache carries) reaches the restored one.

    Settings
    Member Visibility

    On This Page
    Properties
    Methods