test

code_bleu.py

@@ -69,7 +69,7 @@ class CodeBleu(evaluate.Metric):
             ],
         )
 
-    def compute_codebleu_score(ground_truth, generated_answer, lang="python"):
+    def compute_codebleu_score(self, ground_truth, generated_answer, lang="python"):
         """
         Function to compute CodeBLEU score between ground truth code and generated code
         Has keywords for C, C#, C++, Go, Java, JavaScript, PHP, Python, Ruby, and Rust.
@@ -79,5 +79,5 @@ class CodeBleu(evaluate.Metric):
         return result["codebleu"]
 
     def _compute(self, references, predictions):
-        average_codebleu_score = sum([compute_codebleu_score(r, p) for r, p in zip(references, predictions)])/len(references)
+        average_codebleu_score = sum([self.compute_codebleu_score(r, p) for r, p in zip(references, predictions)])/len(references)
         return {"codebleu_score": average_codebleu_score}