import gradio as gr
import logging
from transformers import pipeline

# Set up logging
logging.basicConfig(level=logging.INFO)

# Load Whisper model (tiny version for speed)
asr = pipeline(task="automatic-speech-recognition", model="openai/whisper-tiny.en")

# Function to transcribe audio from a file path
def transcribe_speech(audio_path):
    if audio_path is None:
        logging.error("No audio provided.")
        return "No audio found, please retry."

    try:
        logging.info(f"Received audio file path: {audio_path}")
        output = asr(audio_path)
        return output["text"]
    except Exception as e:
        logging.error(f"Error during transcription: {str(e)}")
        return f"Error processing the audio file: {str(e)}"

# Gradio Interface
with gr.Blocks() as demo:
    gr.Markdown("# 🎤 Simple Speech Recognition App")
    gr.Markdown("Record or upload audio, then click **Transcribe Audio**")

    mic = gr.Audio(label="🎙️ Microphone or Upload", type="filepath")  # This is the key change
    transcribe_button = gr.Button("📝 Transcribe Audio")
    transcription = gr.Textbox(label="🗒️ Transcription", lines=3, placeholder="Transcription will appear here...")

    transcribe_button.click(fn=transcribe_speech, inputs=mic, outputs=transcription)

demo.launch(share=True)