Parses the result of an LLM call. This method is meant to be implemented by subclasses to define how the output from the LLM should be parsed.
parseResult(generations: Generation[] | ChatGeneration[], callbacks: Callbacks): Promise<Y>| Name | Type | Description |
|---|---|---|
generations* | Generation[] | ChatGeneration[] | The generations from an LLM call. |
callbacks | Callbacks | Optional callbacks. |