fix v_mesuares for clustering

results.py

@@ -33,7 +33,7 @@ TRAIN_SPLIT = ["DanishPoliticalCommentsClassification"]
 # Use "validation" split instead
 VALIDATION_SPLIT = ["AFQMC", "Cmnli", "IFlyTek", "TNews", "MSMARCO", "MSMARCO-PL", "MultilingualSentiment", "Ocnli"]
 # Use "dev" split instead
-DEV_SPLIT = ["CmedqaRetrieval", "CovidRetrieval", "DuRetrieval", "EcomRetrieval", "MedicalRetrieval", "MMarcoReranking", "MMarcoRetrieval", "MSMARCO", "MSMARCO-PL", "T2Reranking", "T2Retrieval", "VideoRetrieval"]
+DEV_SPLIT = ["CmedqaRetrieval", "CovidRetrieval", "DuRetrieval", "EcomRetrieval", "MedicalRetrieval", "MMarcoReranking", "MMarcoRetrieval", "MSMARCO", "MSMARCO-PL", "T2Reranking", "T2Retrieval", "VideoRetrieval", "FloresBitextMining"]
 # Use "test.full" split
 TESTFULL_SPLIT = ["OpusparcusPC"]
 
@@ -149,7 +149,7 @@ class MTEBResults(datasets.GeneratorBasedBuilder):
                             score = {metric: score}
                         for sub_metric, sub_score in score.items():
                             if any(x in sub_metric for x in SKIP_KEYS): continue
-                            if isinstance(sub_score, dict): continue
+                            if isinstance(sub_score, dict) or isinstance(sub_score, list): continue
                             out.append({
                                 "mteb_dataset_name": ds_name,
                                 "eval_language": lang if is_multilingual else "",