Update app.py

app.py

@@ -2,53 +2,50 @@ from transformers import pipeline
 import gradio as gr
 import os
 
-model_id = "HarshitJoshi/whisper-small-Hindi"  
+model_id = "HarshitJoshi/whisper-small-Hindi"
 pipe = pipeline("automatic-speech-recognition", model=model_id)
 
-def transcribe_speech(filepath):
+def transcribe_speech(source, source_type):
+    if source_type == "Microphone":
+        filepath = source
+    elif source_type == "Upload":
+        filepath = source.name
+    elif source_type == "Example":
+        filepath = os.path.join(example_folder, source)
+    
     output = pipe(
         filepath,
         max_new_tokens=256,
         generate_kwargs={
             "task": "transcribe",
-            "language": "hindi",  
+            "language": "hindi",
         },
-        chunk_length_s=10,  # Reduced chunk length
-        batch_size=4,  # Reduced batch size
+        chunk_length_s=10,
+        batch_size=4,
     )
     return output["text"]
 
-example_folder = "./examples"  
+example_folder = "./examples"
 example_files = [f for f in os.listdir(example_folder) if f.endswith('.wav') or f.endswith('.mp3')]
 
-demo = gr.Blocks()
-
-mic_transcribe = gr.Interface(
-    fn=transcribe_speech,
-    inputs=gr.Audio(type="filepath"),  
-    outputs=gr.Textbox(),
-)
-
-file_transcribe = gr.Interface(
-    fn=transcribe_speech,
-    inputs=gr.Audio(type="filepath"),  
-    outputs=gr.Textbox(),
-)
-
-def play_and_transcribe(filename):
-    filepath = os.path.join(example_folder, filename)
-    transcription = transcribe_speech(filepath)
-    return filepath, transcription
-
-example_transcribe = gr.Interface(
-    fn=play_and_transcribe,
-    inputs=gr.Dropdown(choices=example_files, label="Select an example"),
-    outputs=[gr.Audio(label="Audio Playback"), gr.Textbox(label="Transcription")],
-)
-
-with demo:
-    gr.TabbedInterface(
-        [mic_transcribe, file_transcribe, example_transcribe],
-        ["Transcribe Microphone", "Transcribe Audio File", "Transcribe Example"],
-    )
+def handle_input(mic, upload, example):
+    if mic is not None:
+        return transcribe_speech(mic, "Microphone")
+    elif upload is not None:
+        return transcribe_speech(upload, "Upload")
+    elif example is not None:
+        return transcribe_speech(example, "Example")
+    else:
+        return "Please provide an input."
+
+with gr.Blocks() as demo:
+    mic = gr.Audio(source="microphone", type="filepath", label="Record from Microphone")
+    upload = gr.Audio(source="upload", type="file", label="Upload an Audio File")
+    example = gr.Dropdown(choices=example_files, label="Or Select an Example")
+
+    output = gr.Textbox(label="Transcription")
+
+    submit_btn = gr.Button("Transcribe")
+    submit_btn.click(handle_input, inputs=[mic, upload, example], outputs=output)
+
 demo.launch(debug=True)