Class VlmModelLocal

java.lang.Object
ai.docling.serve.api.convert.request.options.VlmModelLocal

public class VlmModelLocal extends Object
Options for running a local vision-language model for the VLM pipeline. The parameters refer to a model hosted on Hugging Face.
  • Method Details

    • builder

      public static @NonNull VlmModelLocal.Builder builder()
    • toBuilder

      public @NonNull VlmModelLocal.Builder toBuilder()
    • getRepoId

      public @Nullable String getRepoId()
      Repository id from the Hugging Face Hub.
      Returns:
      the Hugging Face repository ID, or null if not specified
    • getPrompt

      public @Nullable String getPrompt()
      Prompt used when calling the vision-language model.
      Returns:
      the prompt text, or null if not specified
    • getScale

      public @Nullable Integer getScale()
      Scale factor of the images used.
      Returns:
      the image scale factor, or null if not specified
    • getResponseFormat

      public @Nullable ResponseFormat getResponseFormat()
      Type of response generated by the model.
      Returns:
      the response format, or null if not specified
    • getInferenceFramework

      public @Nullable VlmModelLocal.InferenceFramework getInferenceFramework()
      Inference framework to use.
      Returns:
      the inference framework, or null if not specified
    • getTransformersModelType

      public @Nullable VlmModelLocal.TransformersModelType getTransformersModelType()
      Type of transformers auto-model to use.
      Returns:
      the transformers model type, or null if not specified
    • getExtraGenerationConfig

      public Map<String,Object> getExtraGenerationConfig()
      Config from https://huggingface.co/docs/transformers/en/main_classes/text_generation#transformers.GenerationConfig
      Returns:
      the extra generation configuration map
    • toString

      public @NonNull String toString()
      Overrides:
      toString in class Object