Query Workflow

Detailed Query Workflow

  1. Query Reception:

    • Queries are broadcasted via pub/sub channels to all relevant nodes.

    • Each node decides if it has the matching vector slice to respond.

  2. Vector Search and Model Inference:

    • Nodes execute vector similarity searches with Vulkan-accelerated GPUs using ggml libraries.

    • MNN models are loaded dynamically based on the type of query.

  3. Response Generation and Aggregation:

    • Partial responses from each node are aggregated into a final response by a designated aggregator node or function.

PreviousOverviewNextData Storage

Last updated