Update app.py

app.py

@@ -12,8 +12,10 @@ token = os.getenv("HUGGINGFACE_TOKEN")
 def create_dubsync_interface():
     client = Client("Tamiloneto8/Test1", hf_token=token, verbose=True)
 
-    target_langs = ["Assamese","Bengali","Gujarati","Hindi","Kannada",
-                    "Malayalam","Marathi","Odia","Punjabi","Tamil","Telugu"]
+    target_langs = [
+        "Assamese", "Bengali", "Gujarati", "Hindi", "Kannada",
+        "Malayalam", "Marathi", "Odia", "Punjabi", "Tamil", "Telugu"
+    ]
 
     css = """
     .gradio-container { max-width: 1200px; margin: auto; }
@@ -26,10 +28,9 @@ def create_dubsync_interface():
 
         # Step 1 upload and process
         with gr.Row():
-            with gr.Column(scale=1):
-                audio_input = gr.Audio(sources=["upload"], type="filepath", label="🎵 Upload Audio File")
-                lang_dropdown = gr.Dropdown(target_langs, label="🌐 Target Language")
-                process_btn = gr.Button("🎯 Start Processing", variant="primary")
+            audio_input = gr.Audio(sources=["upload"], type="filepath", label="🎵 Upload Audio File")
+            lang_dropdown = gr.Dropdown(target_langs, label="🌐 Target Language")
+            process_btn = gr.Button("🎯 Start Processing", variant="primary")
 
         status_box = gr.Textbox(label="📊 Processing Status", interactive=False)
         chunk_info_state = gr.State([])
@@ -37,23 +38,22 @@ def create_dubsync_interface():
 
         # Step 2 (chunks)
         with gr.Row(visible=False) as edit_section:
-            with gr.Column():
-                nav_row = gr.Row()
+            # Navigation row
+            with gr.Row():
                 prev_btn = gr.Button("◀️ Previous")
                 chunk_counter = gr.Textbox(interactive=False)
                 next_btn = gr.Button("Next ▶️")
-                nav_row.update(prev_btn, chunk_counter, next_btn)
 
-                original_audio = gr.Audio(label="Original Chunk Audio", type="filepath")
-                transcription_text = gr.Textbox(label="Transcription (edit if needed)", lines=2)
-                translation_text = gr.Textbox(label="Translation (edit if needed)", lines=2)
+            original_audio = gr.Audio(label="Original Chunk Audio", type="filepath")
+            transcription_text = gr.Textbox(label="Transcription (edit if needed)", lines=2)
+            translation_text = gr.Textbox(label="Translation (edit if needed)", lines=2)
 
-                with gr.Row():
-                    gen_btn = gr.Button("🔊 Generate Dubbed Chunk")
-                    final_btn = gr.Button("✔️ Finalize Chunk", variant="secondary")
+            with gr.Row():
+                gen_btn = gr.Button("🔊 Generate Dubbed Chunk")
+                final_btn = gr.Button("✔️ Finalize Chunk", variant="secondary")
 
-                dubbed_audio = gr.Audio(label="Dubbed Chunk Audio", type="filepath")
-                progress_text = gr.Textbox(label="Progress", interactive=False)
+            dubbed_audio = gr.Audio(label="Dubbed Chunk Audio", type="filepath")
+            progress_text = gr.Textbox(label="Progress", interactive=False)
 
         # Step 3 (merge)
         merge_btn = gr.Button("🚀 Merge All Finalized Chunks", variant="primary", visible=False)
@@ -63,36 +63,61 @@ def create_dubsync_interface():
         def step1_fn(audio_file, target_language):
             hf_audio = handle_file(audio_file)
             logger.info("Calling /process_audio_pipeline_step1")
-            out = client.predict(audio_file=hf_audio, target_lang=target_language, api_name="/process_audio_pipeline_step1")
+            out = client.predict(
+                audio_file=hf_audio,
+                target_lang=target_language,
+                api_name="/process_audio_pipeline_step1"
+            )
             logger.info("Step1 result: %s", out)
 
             status_text = out[0]
-            chunks = out[1]  # assuming this is chunk info list
+            chunks = out[1] if len(out) > 1 else []
             first = chunks[0] if chunks else {}
-            return status_text, chunks, 0, \
-                   f"Chunk 1 of {len(chunks)}", \
-                   first.get('original_audio'), first.get('transcription'), first.get('translation'), first.get('dubbed_audio'), \
-                   "0 of 0 chunks finalized", True, True
+            return (
+                status_text,
+                chunks,
+                0,
+                f"Chunk 1 of {len(chunks)}",
+                first.get('original_audio'),
+                first.get('transcription'),
+                first.get('translation'),
+                first.get('dubbed_audio'),
+                f"0 of {len(chunks)} chunks finalized",
+                True,  # show edit_section
+                True   # show merge_btn
+            )
 
         def navigate_fn(idx, chunks, trans, tr, direction):
+            # Save edits
             if chunks and 0 <= idx < len(chunks):
                 chunks[idx]['transcription'] = trans
                 chunks[idx]['translation'] = tr
-            new_idx = max(0, min(len(chunks)-1, idx + direction))
+            new_idx = max(0, min(len(chunks) - 1, idx + direction))
             chunk = chunks[new_idx]
-            return new_idx, f"Chunk {new_idx+1} of {len(chunks)}", chunk.get('original_audio'), chunk.get('transcription'), chunk.get('translation'), chunk.get('dubbed_audio'), f"{sum(1 for c in chunks if c.get('finalized',False))} of {len(chunks)} chunks finalized"
+            return (
+                new_idx,
+                f"Chunk {new_idx + 1} of {len(chunks)}",
+                chunk.get('original_audio'),
+                chunk.get('transcription'),
+                chunk.get('translation'),
+                chunk.get('dubbed_audio'),
+                f"{sum(1 for c in chunks if c.get('finalized', False))} of {len(chunks)} chunks finalized"
+            )
 
         def gen_chunk_fn(idx, chunks, trans, tr):
-            chunk = chunks[idx]
             logger.info("Calling /generate_dubbed_chunk")
-            dubbed = client.predict(transcription=trans, translation=tr, api_name="/generate_dubbed_chunk")
+            dubbed = client.predict(
+                transcription=trans,
+                translation=tr,
+                api_name="/generate_dubbed_chunk"
+            )
             chunks[idx]['dubbed_audio'] = dubbed
             return dubbed, chunks
 
         def finalize_fn(idx, chunks):
             if chunks and chunks[idx].get('dubbed_audio'):
                 chunks[idx]['finalized'] = True
-            return chunks, f"{sum(1 for c in chunks if c.get('finalized',False))} of {len(chunks)} chunks finalized"
+            return chunks, f"{sum(1 for c in chunks if c.get('finalized', False))} of {len(chunks)} chunks finalized"
 
         def merge_fn(chunks):
             logger.info("Calling /merge_audio_files")
@@ -100,20 +125,66 @@ def create_dubsync_interface():
             return out[0], out[1]
 
         # Bind events
-        process_btn.click(step1_fn, [audio_input, lang_dropdown],
-                          [status_box, chunk_info_state, current_index_state,
-                           chunk_counter, original_audio, transcription_text, translation_text,
-                           dubbed_audio, progress_text, edit_section, merge_btn])
-        prev_btn.click(navigate_fn, [current_index_state, chunk_info_state, transcription_text, translation_text, gr.literal(-1)],
-                       [current_index_state, chunk_counter, original_audio, transcription_text, translation_text, dubbed_audio, progress_text])
-        next_btn.click(navigate_fn, [current_index_state, chunk_info_state, transcription_text, translation_text, gr.literal(1)],
-                       [current_index_state, chunk_counter, original_audio, transcription_text, translation_text, dubbed_audio, progress_text])
-
-        gen_btn.click(gen_chunk_fn, [current_index_state, chunk_info_state, transcription_text, translation_text],
-                      [dubbed_audio, chunk_info_state])
-        final_btn.click(finalize_fn, [current_index_state, chunk_info_state],
-                        [chunk_info_state, progress_text])
-        merge_btn.click(merge_fn, [chunk_info_state], [final_output, full_audio])
+        process_btn.click(
+            step1_fn,
+            inputs=[audio_input, lang_dropdown],
+            outputs=[
+                status_box,
+                chunk_info_state,
+                current_index_state,
+                chunk_counter,
+                original_audio,
+                transcription_text,
+                translation_text,
+                dubbed_audio,
+                progress_text,
+                edit_section,
+                merge_btn
+            ]
+        )
+
+        prev_btn.click(
+            navigate_fn,
+            inputs=[current_index_state, chunk_info_state, transcription_text, translation_text, gr.Number(-1)],
+            outputs=[
+                current_index_state,
+                chunk_counter,
+                original_audio,
+                transcription_text,
+                translation_text,
+                dubbed_audio,
+                progress_text
+            ]
+        )
+        next_btn.click(
+            navigate_fn,
+            inputs=[current_index_state, chunk_info_state, transcription_text, translation_text, gr.Number(1)],
+            outputs=[
+                current_index_state,
+                chunk_counter,
+                original_audio,
+                transcription_text,
+                translation_text,
+                dubbed_audio,
+                progress_text
+            ]
+        )
+
+        gen_btn.click(
+            gen_chunk_fn,
+            inputs=[current_index_state, chunk_info_state, transcription_text, translation_text],
+            outputs=[dubbed_audio, chunk_info_state]
+        )
+        final_btn.click(
+            finalize_fn,
+            inputs=[current_index_state, chunk_info_state],
+            outputs=[chunk_info_state, progress_text]
+        )
+        merge_btn.click(
+            merge_fn,
+            inputs=[chunk_info_state],
+            outputs=[final_output, full_audio]
+        )
 
     return demo