Update README

README.md

@@ -10,3 +10,12 @@ pinned: false
 ---
 
 Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+
+# Create a new evaluation module in HuggingFace space
+Use the evaluate-cli to create a new module
+
+```
+evaluate-cli create "Metric name" --module_type "metric"
+```
+This will create a new space on the HuggingFace hub. You can then clone the space locally and add the code for the computation of the metric.<br/>
+More information on: https://huggingface.co/docs/evaluate/main/en/creating_and_sharing#requirements

code_bleu.py

@@ -79,5 +79,6 @@ 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([push
+        (r, p) for r, p in zip(references, predictions)])/len(references)
         return {"codebleu_score": average_codebleu_score}