Print feedback on test failure

m-hurlow · m-hurlow · commit b02c85cd9278 · 2024-07-24T13:18:30.000+01:00
diff --git a/evaluation_function/json_tests.py b/evaluation_function/json_tests.py
@@ -18,7 +18,7 @@ def compare(self, eval_result: dict) -> tuple[bool, str]:
         if eval_correct != self.is_correct:
             return (
                 False,
-                f"response \"{self.response}\" with answer \"{self.answer}\" was {'' if eval_correct else 'in'}correct."
+                f"response \"{self.response}\" with answer \"{self.answer}\" was {'' if eval_correct else 'in'}correct: {eval_result['feedback']}."
             )
         
         # Are there any other fields in the eval function result that need to be checked?

Original file line number	Diff line number	Diff line change
`@@ -18,7 +18,7 @@ def compare(self, eval_result: dict) -> tuple[bool, str]:`
`18`	`18`	`if eval_correct != self.is_correct:`
`19`	`19`	`return (`
`20`	`20`	`False,`
`21`		`- f"response \"{self.response}\" with answer \"{self.answer}\" was {'' if eval_correct else 'in'}correct."`
	`21`	`+ f"response \"{self.response}\" with answer \"{self.answer}\" was {'' if eval_correct else 'in'}correct: {eval_result['feedback']}."`
`22`	`22`	`)`
`23`	`23`
`24`	`24`	`# Are there any other fields in the eval function result that need to be checked?`