update prompts and data structure

DARPA-ASKEM · Sep 12, 2024 · b4427bd · b4427bd
1 parent 2663206
commit b4427bd
Show file tree

Hide file tree

Showing 5 changed files with 85 additions and 45 deletions.
diff --git a/api/run.py b/api/run.py
@@ -35,8 +35,10 @@ async def configure_model_from_document(input_model: ConfigureModelDocument):
 @app.post("/model_card")
 async def model_card(input_model: ModelCardModel):
     try:
+        amr = input_model.amr
         research_paper = input_model.research_paper
         response = await model_card_chain(
+            amr=amr,
             research_paper=research_paper
         )  # Use await here
         response = {"response": response}

diff --git a/gollm/openai/prompts/config_from_document.py b/gollm/openai/prompts/config_from_document.py
@@ -1,5 +1,6 @@
 CONFIGURE_FROM_DOCUMENT_PROMPT = """
-You are a helpful agent designed to find multiple model configurations for a given AMR model of various conditions described in a research paper.
+You are a helpful agent designed to find multiple model configurations for a given AMR model of various conditions described in a research paper and the initials and parameters that make up those conditions.
+For context, initials represent the initial state of the system through the initial distribution of tokens across the places, known as the initial marking. Each place corresponds to a variable or state in the system, such as a species concentration in a reaction, and the number of tokens reflects the initial conditions of the ODEs. Parameters define the system's rules and rates of evolution, including transition rates (analogous to reaction rates in ODEs) that determine how quickly tokens move between places. These parameters also include stoichiometric relationships, represented by the weights of arcs connecting places and transitions, dictating how many tokens are consumed or produced when a transition occurs.
 
 Use the following AMR model JSON file as a reference:
 

diff --git a/gollm/openai/prompts/model_card.py b/gollm/openai/prompts/model_card.py
@@ -1,10 +1,10 @@
 INSTRUCTIONS = """
 You are a helpful agent designed to populate metadata of a given AMR model.
 
-You may have access to a document that describes the given AMR model and a JSON representation of the AMR model. Structural information should come from the AMR model.
+You may have access to a document that describes the given AMR model and a JSON representation of the AMR model we want populated. Structural information should come from the AMR model.
 
 You may only have access to the model. Do your best to populate the JSON object specified in the response format with as much information as possible.
-If you cannot answer the entire query, provide as much information as possible. If there is no answer, use the string "null" as a placeholder.
+If you cannot answer the entire query, provide as much information as possible. If there is no answer, populate fields with a null values. Do not leave any fields empty and do not make up information.
 
 Use the following document as a reference:
 

diff --git a/gollm/openai/schemas/model_card.json b/gollm/openai/schemas/model_card.json
@@ -2,11 +2,20 @@
   "$schema": "https://json-schema.org/draft/2020-12/schema",
   "type": "object",
   "properties": {
-    "modelSummary": {
-      "type": "string",
-      "description": "A brief description of the system or process."
+    "summary": {
+      "type": "object",
+      "properties": {
+        "modelSummary": {
+          "type": "string",
+          "description": "A brief description of the system or process."
+        }
+      },
+      "required": [
+        "modelSummary"
+      ],
+      "additionalProperties": false
     },
-    "modelDetails": {
+    "details": {
       "type": "object",
       "properties": {
         "modelDescription": {
@@ -48,30 +57,57 @@
       "additionalProperties": false
     },
     "biasRisksLimitations": {
-      "type": "string",
-      "description": "Describe sources of bias and risk based on the research paper"
+      "type": "object",
+      "properties": {
+        "modelBiasRisksLimitations": {
+          "type": "string",
+          "description": "Describe sources of bias and risk based on the research paper"
+        }
+      },
+      "required": [
+        "modelBiasRisksLimitations"
+      ],
+      "additionalProperties": false
     },
-    "testingDataFactorsMetrics": {
-      "type": "string",
-      "description": "Describe how the model was validated, e.g., through simulation, comparison with real-world data, etc."
+    "testing": {
+      "type": "object",
+      "properties": {
+        "testingDataFactorsMetrics": {
+          "type": "string",
+          "description": "Describe how the model was validated, e.g., through simulation, comparison with real-world data, etc."
+        }
+      },
+      "required": [
+        "testingDataFactorsMetrics"
+      ],
+      "additionalProperties": false
     },
-    "modelSpecs": {
-      "type": "string",
-      "description": "Details about the model's complexity, such as the number of places, transitions, parameter count, and arcs."
+    "specs": {
+      "type": "object",
+      "properties": {
+        "modelSpecs": {
+          "type": "string",
+          "description": "Details about the model's complexity, such as the number of places, transitions, parameter count, and arcs."
+        }
+      },
+      "required": [
+        "modelSpecs"
+      ],
+      "additionalProperties": false
     },
     "glossary": {
       "type": "array",
       "items": {
         "type": "string"
       }
     },
-    "modelCardAuthors": {
+    "authors": {
       "type": "array",
       "items": {
         "type": "string"
       }
     },
-    "howToGetStartedWithTheModel": {
+    "gettingStarted": {
       "type": "object",
       "properties": {
         "examples": {
@@ -86,7 +122,7 @@
       ],
       "additionalProperties": false
     },
-    "citation": {
+    "citations": {
       "type": "object",
       "properties": {
         "references": {
@@ -165,16 +201,16 @@
     }
   },
   "required": [
-    "modelSummary",
-    "modelDetails",
+    "summary",
+    "details",
     "uses",
     "biasRisksLimitations",
-    "testingDataFactorsMetrics",
-    "modelSpecs",
+    "testing",
+    "specs",
     "glossary",
-    "modelCardAuthors",
-    "howToGetStartedWithTheModel",
-    "citation",
+    "authors",
+    "gettingStarted",
+    "citations",
     "moreInformation",
     "structuralInformation"
   ],

diff --git a/gollm/openai/tool_utils.py b/gollm/openai/tool_utils.py
@@ -29,6 +29,7 @@
 
 SCRIPT_DIR = os.path.dirname(os.path.abspath(__file__))
 
+
 def escape_curly_braces(text: str):
     """
     Escapes curly braces in a string.
@@ -82,26 +83,26 @@ def model_config_from_document(research_paper: str, amr: str) -> dict:
     return model_config_adapter(output_json)
 
 
-def amr_enrichment_chain(amr: str, research_paper:str) -> dict:
+def amr_enrichment_chain(amr: str, research_paper: str) -> dict:
     amr_param_states = parse_param_initials(amr)
     prompt = ENRICH_PROMPT.format(
-		param_initial_dict=amr_param_states,
-		paper_text=escape_curly_braces(research_paper)
-	)
+        param_initial_dict=amr_param_states,
+        paper_text=escape_curly_braces(research_paper)
+    )
     client = OpenAI()
     output = client.chat.completions.create(
-		model="gpt-4o-2024-05-13",
-		max_tokens=4000,
-		top_p=1,
-		frequency_penalty=0,
-		presence_penalty=0,
-		seed=123,
-		temperature=0,
-		response_format={"type": "json_object"},
-		messages=[
-			{"role": "user", "content": prompt},
-		],
-	)
+        model="gpt-4o-2024-05-13",
+        max_tokens=4000,
+        top_p=1,
+        frequency_penalty=0,
+        presence_penalty=0,
+        seed=123,
+        temperature=0,
+        response_format={"type": "json_object"},
+        messages=[
+            {"role": "user", "content": prompt},
+        ],
+    )
     return postprocess_oai_json(output.choices[0].message.content)
 
 
@@ -125,7 +126,7 @@ def model_card_chain(amr: str = None, research_paper: str = None) -> dict:
     client = OpenAI()
     output = client.chat.completions.create(
         model="gpt-4o-2024-08-06",
-		temperature=0,
+        temperature=0,
         frequency_penalty=0,
         max_tokens=4000,
         presence_penalty=0,
@@ -162,7 +163,7 @@ def condense_chain(query: str, chunks: List[str], max_tokens: int = 16385) -> st
         top_p=1,
         frequency_penalty=0,
         presence_penalty=0,
-		temperature=0,
+        temperature=0,
         seed=123,
         max_tokens=1024,
         messages=[
@@ -180,7 +181,7 @@ def generate_response(instruction: str, response_format: ResponseFormat | None =
         top_p=1,
         frequency_penalty=0,
         presence_penalty=0,
-		temperature=0,
+        temperature=0,
         seed=123,
         max_tokens=2048,
         response_format=response_format,
@@ -261,7 +262,7 @@ def compare_models(amrs: List[str]) -> str:
         frequency_penalty=0,
         presence_penalty=0,
         seed=123,
-		temperature=0,
+        temperature=0,
         max_tokens=2048,
         messages=[
             {"role": "user", "content": prompt},