app.py

import torch

import gradio as gr
import pytube as pt
from transformers import pipeline
from huggingface_hub import model_info
#from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor

MODEL_NAME = "ihanif/wav2vec2-xls-r-300m-pashto"
lang = "ps"

#load pre-trained model and tokenizer
#processor = Wav2Vec2Processor.from_pretrained(MODEL_NAME)
#model = Wav2Vec2ForCTC.from_pretrained(MODEL_NAME)

device = 0 if torch.cuda.is_available() else "cpu"
pipe = pipeline(
    task="automatic-speech-recognition",
    model=MODEL_NAME,
    #chunk_length_s=30,
    device=device,
)

#pipe.model.config.forced_decoder_ids = pipe.tokenizer.get_decoder_prompt_ids(language=lang, task="transcribe")

def transcribe(microphone, file_upload):
    warn_output = ""
    # if (microphone is not None) and (file_upload is not None):
    #     warn_output = (
    #         "WARNING: You've uploaded an audio file and used the microphone. "
    #         "The recorded file from the microphone will be used and the uploaded audio will be discarded.\n"
    #     )

    # elif (microphone is None) and (file_upload is None):
    #     return "ERROR: You have to either use the microphone or upload an audio file"

    if (microphone is None) and (file_upload is None):
        return "ERROR: You have to either use the microphone or upload an audio file"
        
    file = microphone if microphone is not None else file_upload

    text = pipe(file)["text"]
    #transcription = wav2vec_model(audio)["text"]

    return warn_output + text


def _return_yt_html_embed(yt_url):
    video_id = yt_url.split("?v=")[-1]
    HTML_str = (
        f'<center> <iframe width="500" height="320" src="https://www.youtube.com/embed/{video_id}"> </iframe>'
        " </center>"
    )
    return HTML_str


def yt_transcribe(yt_url):
    yt = pt.YouTube(yt_url)
    html_embed_str = _return_yt_html_embed(yt_url)
    stream = yt.streams.filter(only_audio=True)[0]
    stream.download(filename="audio.mp3")

    text = pipe("audio.mp3")["text"]

    return html_embed_str, text


demo = gr.Blocks()

examples=[["example-1.wav","example-2.wav"]]
# examples=["example-1.wav"]

mf_transcribe = gr.Interface(
    fn=transcribe,
    inputs=[
        gr.inputs.Audio(source="microphone", type="filepath", optional=True),
        gr.inputs.Audio(source="upload", type="filepath", optional=True),
    ],
    outputs="text",
    layout="horizontal",
    theme="huggingface",
    title="(Pashto ASR) د پښتو اتوماتیک وینا پیژندنه",
    description=(
        "</p> تاسو کولی شئ یو آډیو فایل اپلوډ کړئ یا په خپل وسیله مایکروفون وکاروئ. مهرباني وکړئ ډاډ ترلاسه کړئ چې تاسو اجازه ورکړې ده<p>"
    ),
    allow_flagging="never",
    examples=examples,
)

yt_transcribe = gr.Interface(
    fn=yt_transcribe,
    inputs=[gr.inputs.Textbox(lines=1, placeholder="Paste the URL to a YouTube video here", label="YouTube URL")],
    outputs=["html", "text"],
    layout="horizontal",
    theme="huggingface",
    title="(Transcribe YouTube) د پښتو اتوماتیک وینا پیژندنه",
    description=(
        "مهرباني وکړئ د خپل غږ په کارولو سره د پښتو لیکلو لپاره لاندې اپلیکیشن وکاروئ. تاسو کولی شئ یو آډیو فایل اپلوډ کړئ یا په خپل وسیله مایکروفون وکاروئ. مهرباني وکړئ ډاډ ترلاسه کړئ چې تاسو اجازه ورکړې ده"
    ),
    allow_flagging="never",
)

with demo:
    gr.TabbedInterface([mf_transcribe, yt_transcribe], ["Transcribe Audio", "Transcribe YouTube"])

demo.launch(enable_queue=False)