tr-wikihow-summ fix

app.py

@@ -16,7 +16,6 @@ def cache_results(path):
     json_results = read_results(path)
     results = list()
     for entry in json_results:
-        print(entry)
         row = {
             'model': entry['model']['model'],
             'num_parameters': entry['model']['num_parameters'],
@@ -45,7 +44,6 @@ def cache_datasets(path):
         datasets = json.load(f)
     for key in datasets.keys():
         datasets[key]['dataset'] = key
-    del datasets['tr-wikihow-summ']  # FIXME: There are missing experiments.
     return datasets
 
 

data.py

@@ -65,6 +65,7 @@ DATASET_TASK_DICT = {
     'wmt-tr-en-prompt': Tasks.MACHINE_TRANSLATION,
     'gecturk_generation': Tasks.GRAMMATICAL_ERROR_CORRECTION,
     'turkce_atasozleri': Tasks.MULTIPLE_CHOICE,
+    'turkishmmlu':Tasks.MULTIPLE_CHOICE,
 }
 
 
@@ -103,7 +104,7 @@ DATASET_GROUPS = {
         'description': 'Natural Language Inference (NLI) datasets in Turkish: XNLI, SNLI and MNLI.',
     },
     'SUM': {
-        'datasets': [name for name, task in DATASET_TASK_DICT.items() if task == Tasks.SUMMARIZATION and name != 'tr-wikihow-summ'],  # FIXME: include all the datasets.
+        'datasets': ['wiki_lingua_tr', 'xlsum_tr', 'tr-wikihow-summ', 'mlsum_tr'],
         'description': 'Summarization datasets in Turkish (XLSum, MLSum, WikiLingua and TrWikiHowSumm).',
     },
     'GEC': {