Update app.py

app.py

@@ -8,6 +8,7 @@ from PIL import Image
 import torch
 from torch.nn.utils.rnn import pad_sequence
 from transformers import BridgeTowerProcessor
+from tqdm import tqdm
 
 from bridgetower_custom import BridgeTowerTextFeatureExtractor, BridgeTowerForITC
 
@@ -134,7 +135,7 @@ def extract_images_and_embeds(video_id, video_path, subtitles, output, expanded=
     batch_list = []
     vtt = webvtt.read(subtitles)
    
-    for idx, caption in enumerate(progress.tqdm(webvtt.read(subtitles), total=vtt.total_length, desc="Generating embeddings")):
+    for idx, caption in enumerate(tqdm(vtt, total=vtt.total_length, desc="Generating embeddings")):
         st_time = str2time(caption.start)
         ed_time = str2time(caption.end)
 
@@ -286,7 +287,7 @@ def get_video_id_from_url(video_url):
     return None
 
 
-def process(video_url, text_query, progress=gr.Progress()):
+def process(video_url, text_query, progress=gr.Progress(track_tqdm=True)):
     tmp_dir = os.environ.get('TMPDIR', '/tmp')
     video_id = get_video_id_from_url(video_url)
     output_dir = os.path.join(tmp_dir, video_id)
@@ -298,7 +299,7 @@ def process(video_url, text_query, progress=gr.Progress()):
         output=output_dir, 
         expanded=False,
         batch_size=8,
-        progress=gr.Progress(),
+        progress=progress,
     )
     frame_paths, transcripts = run_query(video_file, text_query, path=output_dir)
     return video_file, [(image, caption) for image, caption in zip(frame_paths, transcripts)]