FileDownloader is used for visual question answering. It downloads the source files and stores them in the path defined in the FILE_ROOT_PATH environment variable. From there, other nodes, like PromptNode, can access them.

Basic Information

• Pipeline type: Query pipelines.
• Nodes that can precede it in a pipeline: Retriever, EntityExtractor, Ranker, RetrievalScoreAdjuster, JoinDocuments, InterleaveDocuments
• Nodes that can follow it in a pipeline: PromptNode
• Node input: Documents
• Node output:There's no output, but the source files of the documents are saved locally.
• Available node classes: FileDownloader

Usage Example

components:
  - name: ImageDownloader
    type: FileDownloader
    params:
      file_extensions:
        - ".pdf"
        ...
 pipelines:
  - name: query
    nodes:
      - name: BM25Retriever
        inputs: [Query]
      - name: ImageDownloader
        inputs: [BM25Retriever]
      - name: PromptNode
        inputs: [ImageDownloader]
...

Parameters

You can specify the following parameters for FileDownloaderin the pipeline YAML: