Skip to main content
Version: v1.6.1 🚧

Generate a Document summary

Overview​

In the Documents section, you can generate a summary of the Document using Document AI.

note

A summary of the Document using Document AI can not be generated unless the Document has been added to a Collection first. To learn how to add a Document to a Collection, see Add a Document to a Collection.

Instructions​

To generate Document AI, follow these steps:

  1. In the Enterprise h2oGPTe navigation menu, click Documents.
  2. In the Documents grid or list, select the name of the Document you want to create a new summary of.
  3. Click Summarize, Extract, Process under the Document AI section. Create a new summary
  4. In the Summarize, Extract, Process panel, customize the Document AI settings according to your requirements.
    note

    To learn more about each setting, see Document AI settings.

  5. Click Summarize, Extract, Process after finalizing the settings.

Document AI settings​

LLM​

  • This setting enables you to select the Large Language Model (LLM) to generate the Document summary.

  • You can choose from a variety of LLMs by selecting the desired model from the menu.

Enable vision​

In addition to sending document context to the normal Large Language Model (LLM), this setting allows you to pass document context as images to a vision-capable LLM. If you enable the automatic mode, it will select the vision-capable LLM based on the document context and the LLM model.

note

Enabling vision mode can lead to higher latency and cost.

Vision LLM​

This setting allows you to select the LLM for processing images. Selecting automatic mode will pick a vision LLM based on availability and configuration. It typically selects the same LLM for vision-capable models and the default LLM for non-vision models.

Max. number of chunks and approximate cost range​

This setting allows you to configure the number of chunks to be extracted from the Document to summarize its context. Adjust the slider to select the desired number of chunks for the summary.

note

Moving the slider to 100 prompts the Large Language Model (LLM) to utilize 100 chunks from the Document to generate the Document summary.

Prompt Template​

This setting allows you to select a prompt template from the drop-down menu to customize the prompts used for the Collection. You can create your own prompt template on the Prompts page and use it for your Collection.

Personality (System Prompt)​

This setting allows you to customize the personality of the LLM according to your requirements for the Document summary. It aids in shaping the behavior of the generated Document summary.

Example: You are h2oGPTe, an expert question-answering AI system created by H2O.ai that performs like GPT-4 by OpenAI.

Prompts before/after document context​

This setting enables you to specify prompts before and after the Document context to guide the language model (LLM) in generating a Document summary.

Before prompt example: "In order to write a concise single-paragraph or bulleted list summary, pay attention to any chat history, any images given, or any following text: DOCUMENT_CONTEXT"

After prompt example: "Using only any chat history, any images given, or any text above, write a condensed and concise summary of key results (preferably as bullet points)."

Prompt per image batch for vision models​

This setting defines the prompt used to obtain answers for the user queries from a batch of images when vision mode is enabled.

Prompt for final image batch reduction for vision models​

This setting defines the prompt used to obtain the final answer for the user query from the per-image-batch answers when vision mode is enabled.

JSON Schema​

This setting enables you to ensure that the output will conform to the JSON schema. Use {} to let the LLM determine the output schema.

PII Detection​

This setting enables you to detect personally identifiable information (PII) in the Document summary. Click on the drop-down menu to select a PII detection option.

note

The PII detection policies apply in addition to system settings.

Document Metadata to include​

This setting allows you to provide additional Document info as part of the DOCUMENT_CONTEXT. You can select the metadata fields to include in the Document summary by using the drop-down menu.

Keep intermediate results​

If this setting is enabled, the final result will be a list of intermediate results. Otherwise, a global result is created. This is useful for debugging and understanding the process of generating the Document summary.

Feedback