Update app.py

app.py

@@ -4,6 +4,9 @@ import os
 HF_TOKEN = os.getenv('HW_Token')
 hf_writer = gr.HuggingFaceDatasetSaver(HF_TOKEN, "save_audio")
 
+import gradio as gr
+import os
+
 class TextFileReader:
     def __init__(self):
         self.lines = []
@@ -41,18 +44,27 @@ def save_text_lines(file):
 def save_audio_text(audio, text):
     if not os.path.exists("/tmp/recordings"):
         os.makedirs("/tmp/recordings")
-    audio_path = f"/tmp/recordings/line_{reader.current_index}.wav"
-    text_path = f"/tmp/recordings/line_{reader.current_index}.txt"
-    
-    with open(audio_path, "wb") as f:
-        f.write(audio['data'])
+
+    # Debugging to print out the structure of the audio variable
+    print(audio)
+
+    # Check if audio is a dictionary and contains 'data'
+    if isinstance(audio, dict) and 'data' in audio:
+        audio_data = audio['data']
+        audio_path = f"/tmp/recordings/line_{reader.current_index}.wav"
+        text_path = f"/tmp/recordings/line_{reader.current_index}.txt"
         
-    with open(text_path, "w") as f:
-        f.write(text)
-    
-    # Move to the next line after saving
-    next_line = reader.forward_line()
-    return next_line
+        with open(audio_path, "wb") as f:
+            f.write(audio_data)
+            
+        with open(text_path, "w") as f:
+            f.write(text)
+        
+        # Move to the next line after saving
+        next_line = reader.forward_line()
+        return next_line
+    else:
+        return "Audio data is not in the expected format."
 
 # Define the Gradio interface
 with gr.Blocks() as demo:
@@ -70,7 +82,7 @@ with gr.Blocks() as demo:
     generate_button.click(fn=update_output, inputs=file_upload, outputs=current_line)
     
     with gr.Row():
-        audio_record = gr.Audio(sources=["microphone","upload"], type="filepath")
+        audio_record =  gr.Audio(sources=["microphone","upload"], type="filepath")
         save_button = gr.Button("Save Audio and Next Line")
 
     save_button.click(fn=save_audio_text, inputs=[audio_record, current_line], outputs=current_line)