update config from document prompt (#5580)

Author: dgauldie

packages/gollm/gollm_openai/prompts/config_from_document.py

@@ -23,7 +23,9 @@
 2.	For the description, provide a long-form description of the condition. If the description cannot be created from the user-provided text, set it to an empty string.
 3.	`model_id` id a UUID. If the AMR model has an id, you can use it. Otherwise, you can set as the nil UUID "00000000-0000-0000-0000-000000000000".
 4.	For each state specified in the AMR model, extract information from the text and create an initial semantic object. Do not create new initial semantic objects for states that are not included in the original AMR model. If you cannot find a value or expression for the initial state, do not create an initial semantic object.
-5.	For each parameter specified in the AMR model, extract information from the text and create a parameter semantic object. Do not create new parameter semantic objects for parameters that are not included in the original AMR model. If you cannot find a constant value or distribution for the parameter, do not create a parameter semantic object.
+5.	For each parameter specified in the AMR model, extract information from the text and create a parameter semantic object. Do not create new parameter semantic objects for parameters that are not included in the original AMR model. If you cannot find a value for the parameter, do not create a parameter semantic object. Use the following rules to determine if the parameter is a constant or a distribution:
+    a.	Constant parameters will have a single value. To create a constant parameter, set the distribution type to "Constant" and add the value to the `value` field.
+    b.  Distributed parameters will usually appear as a range (For example, 0.01 - 0.2) or a list of values. To create a distribution parameter, set the distribution type to "Uniform" and add the smallest value in the range or list to the `minimum` field and the largest value to the `maximum` field.
 6. `observableSemanticList` should be an empty list.
 7. `inferredParameterList` should be an empty list.
 8. Determine what page the information was extracted from and set the `extractionPage` value to that page number. If the page number cannot be determined, set it to 0. Only pick one page number. Do not provide a range.