handle large file for HF

src/streamlit_app.py

@@ -383,35 +383,56 @@ def show_home_page():
     )
     st.session_state.annotation_type = annotation_type
     
-    # File upload
+    # File upload with better error handling
     st.subheader("2. Upload Audio File")
+    
+    # Add file size warning for Hugging Face Spaces
+    st.info("💡 **Tip for Hugging Face Spaces:** Large files (>10MB) may fail to upload. Try smaller audio files or compress your audio if you encounter issues.")
+    
     uploaded_file = st.file_uploader(
         "Upload an audio file", 
         type=['wav', 'mp3', 'flac', 'm4a'],
-        help="Supported formats: WAV, MP3, FLAC, M4A"
+        help="Supported formats: WAV, MP3, FLAC, M4A. Keep files under 10MB for best compatibility on Hugging Face Spaces."
     )
     
     if uploaded_file is not None:
-        st.session_state.audio_file = uploaded_file.read()
-        
-        # Save temporary file to get duration
-        with tempfile.NamedTemporaryFile(delete=False, suffix='.wav') as tmp_file:
-            tmp_file.write(st.session_state.audio_file)
-            st.session_state.audio_duration = get_audio_duration(tmp_file.name)
-            os.unlink(tmp_file.name)
-        
-        st.success(f"✅ Audio file uploaded successfully!")
-        st.info(f"Duration: {format_time(st.session_state.audio_duration)}")
-        
-        # Show audio player
-        st.subheader("Audio Preview")
-        audio_html = create_audio_player_html(st.session_state.audio_file)
-        st.components.v1.html(audio_html, height=120)
-        
-        # Continue button
-        if st.button("Continue to Transcription →", type="primary"):
-            st.session_state.current_page = "transcription"
-            st.rerun()
+        try:
+            # Read file data
+            audio_data = uploaded_file.read()
+            
+            # Check file size (warn if >10MB for HF Spaces)
+            file_size_mb = len(audio_data) / (1024 * 1024)
+            if file_size_mb > 10:
+                st.warning(f"⚠️ File size: {file_size_mb:.1f}MB. Large files may cause issues on Hugging Face Spaces.")
+            
+            # Store in session state
+            st.session_state.audio_file = audio_data
+            
+            # Get duration
+            st.session_state.audio_duration = get_audio_duration(audio_data)
+            
+            st.success(f"✅ Audio file uploaded successfully! ({file_size_mb:.1f}MB)")
+            
+            if st.session_state.audio_duration > 0:
+                st.info(f"Duration: {format_time(st.session_state.audio_duration)}")
+            
+            # Show audio player
+            st.subheader("Audio Preview")
+            audio_html = create_audio_player_html(st.session_state.audio_file)
+            st.components.v1.html(audio_html, height=120)
+            
+            # Continue button
+            if st.button("Continue to Transcription →", type="primary"):
+                st.session_state.current_page = "transcription"
+                st.rerun()
+                
+        except Exception as e:
+            st.error(f"❌ Error processing audio file: {str(e)}")
+            st.error("This might be due to:")
+            st.error("- File format not supported")
+            st.error("- File too large for Hugging Face Spaces")
+            st.error("- Corrupted audio file")
+            st.info("Try converting your audio to WAV format and reducing the file size.")
 
 def show_transcription_page():
     """Transcription page - text annotation"""