import os import numpy as np import unicodedata from datasets import load_dataset, Audio from transformers import pipeline import gradio as gr import torch ############### HF ########################### HF_TOKEN = os.getenv("HF_TOKEN") hf_writer = gr.HuggingFaceDatasetSaver(HF_TOKEN, "Urdu-ASR-flags") ############## Inference ############################## transcriber = pipeline("automatic-speech-recognition", model="kingabzpro/wav2vec2-large-xls-r-300m-Urdu") def transcribe(audio): sr, y = audio y = y.astype(np.float32) y /= np.max(np.abs(y)) return transcriber({"sampling_rate": sr, "raw": y})["text"] demo = gr.Interface( transcribe, gr.Audio(sources=["microphone"]), "text", ) ################### Gradio Web APP ################################ title = "Urdu Automatic Speech Recognition" description = """
Fine-tuning XLS-R for Multi-Lingual ASR with 🤗 Transformers