Update app.py

app.py

@@ -31,6 +31,7 @@ model.config.eos_token_id = tokenizer.eos_token_id
 model.config.decoder_start_token_id = tokenizer.bos_token_id
 model.config.pad_token_id = tokenizer.pad_token_id
 
+
 model_sum_name = "google-t5/t5-base"
 tokenizer_sum = AutoTokenizer.from_pretrained("google-t5/t5-base")
 model_sum = AutoModelForSeq2SeqLM.from_pretrained("google-t5/t5-base")
@@ -98,19 +99,16 @@ with st.sidebar:
 input_option = st.selectbox("Select input method:", ["Folder Path", "Upload Video", "Upload ZIP"])
 
 video_files = []
-original_names = []
 
 if input_option == "Folder Path":
     folder_path = st.text_input("Enter the folder path containing videos:")
     if folder_path and os.path.isdir(folder_path):
         video_files = [os.path.join(folder_path, f) for f in os.listdir(folder_path) if f.lower().endswith(('mp4', 'avi', 'mov', 'mkv'))]
-        original_names = [f for f in os.listdir(folder_path) if f.lower().endswith(('mp4', 'avi', 'mov', 'mkv'))]
 
 elif input_option == "Upload Video":
     uploaded_files = st.file_uploader("Upload video files", type=["mp4", "avi", "mov", "mkv"], accept_multiple_files=True)
     if uploaded_files:
         for uploaded_file in uploaded_files:
-            original_names.append(uploaded_file.name)
             with tempfile.NamedTemporaryFile(delete=False) as temp_file:
                 temp_file.write(uploaded_file.read())
                 video_files.append(temp_file.name)
@@ -123,23 +121,22 @@ elif input_option == "Upload ZIP":
             with zipfile.ZipFile(temp_file.name, 'r') as zip_ref:
                 zip_ref.extractall("/tmp/videos")
                 video_files = [os.path.join("/tmp/videos", f) for f in zip_ref.namelist() if f.lower().endswith(('mp4', 'avi', 'mov', 'mkv'))]
-                original_names = [f for f in zip_ref.namelist() if f.lower().endswith(('mp4', 'avi', 'mov', 'mkv'))]
 
 if video_files:
     captions = {}
-    for video_file, original_name in zip(video_files, original_names):
+    for video_file in video_files:
         frames, captions_df = process_video(video_file, frame_interval=20)
         
         if frames and not captions_df.empty:
             generated_captions = ' '.join(captions_df['Caption'])
             summary = summarize_pipe(generated_captions)[0]['summary_text']
-            captions[original_name] = summary
+            captions[video_file] = summary
 
     # Display videos in a 4-column grid
     cols = st.columns(4)
     for idx, (video_path, summary) in enumerate(captions.items()):
         with cols[idx % 4]:
-            st.video(video_files[idx])
+            st.video(video_path)
             st.caption(summary)
 
     if query:
@@ -149,15 +146,8 @@ if video_files:
             st.video(video_path)
             st.caption(summary)
 
-    # Save captions to Excel and provide a download button
-    if st.button("Generate Excel"):
+    # Save captions to CSV and provide a download button
+    if st.button("Generate CSV"):
         df = pd.DataFrame(list(captions.items()), columns=['Video', 'Caption'])
-        excel_buffer = io.BytesIO()
-        with pd.ExcelWriter(excel_buffer, engine='xlsxwriter') as writer:
-            df.to_excel(writer, index=False, sheet_name='Captions')
-        st.download_button(
-            label="Download captions as Excel",
-            data=excel_buffer,
-            file_name="captions.xlsx",
-            mime="application/vnd.openxmlformats-officedocument.spreadsheetml.sheet"
-        )
+        csv = df.to_csv(index=False)
+        st.download_button(label="Download captions as CSV", data=csv, file_name="captions.csv", mime="text/csv")