import gradio as gr
from transformers import pipeline
from librosa import load

def transcribe(input_audio):
    speech, _ = load(input_audio, sr=16000, mono=True)
    output = pipe(speech, chunk_length_s=30, stride_length_s=5)['text']
    return output

pipe = pipeline(
    "automatic-speech-recognition",
    model="GetmanY1/wav2vec2-large-sami-cont-pt-22k-finetuned",
    device="cpu"
    )

gradio_app = gr.Interface(
    transcribe,
    gr.Audio(sources=["upload","microphone"]),
    "text",
)

if __name__ == "__main__":
    gradio_app.launch()



# if __name__ == "__main__":
#     gradio_app.launch()