fix: py-script execute's return type should be consistent with scorer… (

#125)

.changeset/pink-laws-relax.md

@@ -0,0 +1,5 @@
+---
+"@empiricalrun/core": minor
+---
+
+fix: py-script execute's return type should be consistent with scorer's return type

docs/models/custom.mdx

@@ -30,8 +30,8 @@ signature:
 - **Arguments**
   - inputs: dict of key-value pairs with [sample inputs](../dataset/basics)
   - parameters: dict of key-value pairs with the run parameters
-- **Returns**: a dict with
-  - output (string): The response from the model/application
+- **Returns**: an output dict with
+  - value (string): The response from the model/application
   - metadata (dict): Custom key-value pairs that are passed on to the scorer and
     web reporter
 
@@ -40,7 +40,7 @@ signature:
 def execute(inputs):
     # ...
     return {
-        "output": output,
+        "value": output,
         "metadata": {
             "key": value
         }

examples/rag/rag.py

@@ -24,8 +24,7 @@ def build_query_engine(llm):
     response = query_engine.query(question)
     output = response.response
     contexts = [c.node.get_content() for c in response.source_nodes]
-
     return {
-        "output": output,
+        "value": output,
         "metadata": {"contexts": contexts},
     }

packages/core/src/executors/run/transformers/script.ts

@@ -75,7 +75,7 @@ export const scriptExecutor: Transformer = async (runConfig, sample) => {
 
   return {
     output: {
-      value: result["output"],
+      value: result["value"],
       metadata: result["metadata"],
     },
   };