fix prod bugs

app.py

@@ -16,7 +16,7 @@ AFRIXNLI_TRANSLATE = "afrixnli_translate"
 
 BENCHMARKS = [AFRIMMLU_DIRECT, AFRIMMLU_TRANSLATE, AFRIXNLI_DIRECT, AFRIXNLI_TRANSLATE]
 
-METRICS = ["acc_norm", "acc_norm", "acc_norm", "mc2"]
+METRICS = ["acc_norm", "acc_norm", "acc_norm"]
 
 LANGS = ['amh', 'eng', 'ewe', 'fra', 'hau', 'ibo', 'kin', 'lin', 'lug', 'orm', 'sna', 'sot', 'swa', 'twi', 'wol', 'xho', 'yor', 'zul']
 
@@ -66,11 +66,17 @@ def collect_results():
         pretrained_models.add(pretrained)
 
         for lang_task, perfs in results.items():
-            task, lang = lang_task.split('_')
+            print(lang_task)
+            print(perfs)
+            lang_task = lang_task.split('_')
+            lang = lang_task[-1]
+            task = '_'.join(lang_task[:-1])
+
             assert task in BENCHMARKS
 
             if lang and task:
-                metric = METRICS[BENCHMARKS.index(task)]
+                print(BENCHMARKS.index(task))
+                metric = METRICS[BENCHMARKS.index(task)-1]
                 p = round(perfs[metric] * 100, 1)
                 performance_dict[(pretrained, lang)][task] = p
     return performance_dict, pretrained_models