Data Generator

Choose the model that best fits your generative AI task and available resources. The selected model determines the language understanding, accuracy, and cost of the operation.

Guidance

• Use newer or more capable models for documents with complex structure or ambiguous content.
• Use models which are known to produce reliable structured output.
• Consider cost and performance trade-offs when selecting a model.
• The list of available models is determined by your LLM Connector configuration.

Can be one of the following types:

The Chat Parameters class defines a set of options for tuning the behavior of LLM completion requests in Grooper. These parameters allow users to control the randomness, diversity, and topicality of generated responses, enabling tailored outputs for a wide range of scenarios.

Typical usage involves adjusting one or more parameters to guide the model’s output toward the desired style or content. For example, increasing randomness for creative tasks, or reducing repetition for summarization.

Configuration Guidance

• Adjust either 'Temperature' or 'Top P' to control randomness and diversity, but not both simultaneously.
• Use 'Presence Penalty' and 'Frequency Penalty' together to reduce repetition and encourage broader topic coverage.

Usage in Grooper

Chat Parameters are used throughout Grooper to influence LLM completions for document processing, data extraction, and other AI-driven tasks.

Impact

Proper configuration of ChatParameters can significantly affect the quality, creativity, and reliability of LLM outputs. Experimentation is recommended to find the best settings for your specific use case.

When enabled, the LLM is instructed to return data in a strict JSON format matching the extraction schema. This improves reliability and parsing accuracy, especially for complex or nested data.

• Supported by OpenAI models such as GPT-4.1, GPT-4.1-mini, GPT-4o, GPT-4o-mini, GPT-4 Turbo, and recent GPT-3.5 Turbo models.
• Check documentation for your specific model to verify structured output support.
• When disabled, a best-effort parsing approach is used, which may be less reliable.
• Increase 'Max Tries' if you disable structured output, as the LLM may return unparseable or incomplete responses more frequently.

Controls how many times Grooper will attempt to re-issue a request to the LLM service if an error occurs during extraction. Retries are triggered by issues such as network timeouts, incomplete or malformed responses, or other transient failures that may prevent a successful extraction on the first attempt.

Configuration Guidance

• Increase this value if you experience intermittent connectivity issues or occasional LLM service instability.
• The allowed range is 1 to 8. Setting a value outside this range is not permitted.
• Each retry uses the same prompt and parameters as the original request.
• When 'Use Structured Output' is disabled, increasing 'Max Tries' can help ensure that documents are processed successfully, as the LLM may occasionally return unparseable or incomplete responses. Allowing more retries gives the system additional opportunities to obtain a valid result.

Impact

A higher value for 'Max Tries' can improve reliability in unstable environments, ensuring that temporary issues do not cause extraction to fail. However, increasing the number of retries may also result in longer processing times if the LLM service is unavailable or consistently returns errors.

If all attempts fail, the extraction operation will be unsuccessful, and an error will be logged in the diagnostics.