Update app.py

app.py

@@ -1,60 +1,25 @@
 import gradio as gr
-import torch
-from transformers import pipeline
+from tone import StreamingCTCPipeline, read_audio
 
-# Load the ASR model
-# Using the t-tech/T-one model for automatic speech recognition
-asr_pipeline = pipeline(
-    "automatic-speech-recognition",
-    model="t-tech/T-one",
-    torch_dtype=torch.float16 if torch.cuda.is_available() else torch.float32,
-    device=0 if torch.cuda.is_available() else -1
-)
+# Load model once at startup
+pipe = StreamingCTCPipeline.from_hugging_face()
 
-def transcribe_audio(audio_file):
-    """
-    Transcribe audio file using the T-one ASR model
-    Args:
-        audio_file: Audio file uploaded by user
-    Returns:
-        str: Transcribed text
-    """
-    if audio_file is None:
+def transcribe(audio_path):
+    if audio_path is None:
         return "Please upload an audio file."
-    
     try:
-        # Transcribe the audio
-        result = asr_pipeline(audio_file)
-        return result["text"]
+        audio = read_audio(audio_path)
+        result = pipe.forward_offline(audio)
+        # result: list of TextPhrase with text, start_time, end_time
+        return "\n".join([x.text for x in result])
     except Exception as e:
-        return f"Error during transcription: {str(e)}"
+        return f"Error: {str(e)}"
 
-# Create Gradio interface
 with gr.Blocks(title="T-one ASR Demo") as demo:
     gr.Markdown("# T-one Automatic Speech Recognition Demo")
     gr.Markdown("Upload an audio file to get real-time transcription using the t-tech/T-one model.")
-    
-    with gr.Row():
-        # Audio input component
-        audio_input = gr.Audio(
-            label="Upload Audio File",
-            type="filepath"
-        )
-        
-        # Text output component
-        text_output = gr.Textbox(
-            label="Transcription",
-            placeholder="Transcribed text will appear here...",
-            lines=5
-        )
-    
-    # Set up the transcription function to run when audio is uploaded
-    audio_input.change(
-        fn=transcribe_audio,
-        inputs=audio_input,
-        outputs=text_output
-    )
+    audio_input = gr.Audio(label="Upload Audio File", type="filepath")
+    text_output = gr.Textbox(label="Transcription", placeholder="Transcribed text will appear here...", lines=5)
+    audio_input.change(transcribe, inputs=audio_input, outputs=text_output)
 
-# Launch the app
-if __name__ == "__main__":
-    demo.launch()
+demo.launch()