Small bug fix on experiment_to_video_mapping.py

semantic_encoding/_utils/experiment_to_video_mapping.py

@@ -188,18 +188,18 @@ def get_dataset_experiments(dataset_name, cvpr2020_subset=False):
 
                 available_experiments = set([fname.split('/')[-1] for fname in glob(f'{os.path.dirname(os.path.dirname(os.path.dirname(__file__)))}/rl_fast_forward/resources/YouCook2/raw_videos/*/*/*') if len(fname.split('/')[-1].split('.')) == 1])
 
-                experiments = set(experiments).intersection(available_experiments)
+                experiments = list(set(experiments).intersection(available_experiments))
 
         elif dataset_name == 'VaTeX':
             vals = json.load(open(f'{os.path.dirname(os.path.dirname(os.path.abspath(__file__)))}/resources/vatex_validation_v1.0.json'))
-            experiments = set([val['videoID'] for val in vals])  # Validation videos
+            experiments = list(set([val['videoID'] for val in vals]))  # Validation videos
 
         elif dataset_name == 'COIN':
             experiments = set(pd.read_csv(f'{os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))}/rl_fast_forward/resources/COIN/metadata.csv')['vid_id'].tolist())
 
             available_experiments = set([fname.split('/')[-1].split('.mp4')[0] for fname in glob(f'{os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))}/rl_fast_forward/resources/COIN/videos/*/*mp4')])
 
-            experiments = set(experiments).intersection(available_experiments)
+            experiments = list(set(experiments).intersection(available_experiments))
 
         return experiments
 
@@ -255,4 +255,4 @@ def vatex(self, video_name):
         self.start_frame = 1
         self.end_frame = int(video.get(7))
         self.num_frames = self.end_frame - self.start_frame + 1
-        self.document_filename = None
+        self.document_filename = None