app.py

from transformers import AutoModelForCTC, Wav2Vec2Processor
import torch
import gradio as gr

# Load model and processor
model_name = "nada15/wav2vec2-large-xls-r-300m-dm32"
processor = Wav2Vec2Processor.from_pretrained(model_name)
model = AutoModelForCTC.from_pretrained(model_name, ignore_mismatched_sizes=True)

def transcribe(audio):
    inputs = processor(audio, sampling_rate=16000, return_tensors="pt", padding=True)
    logits = model(inputs.input_values).logits
    predicted_ids = torch.argmax(logits, dim=-1)
    transcription = processor.batch_decode(predicted_ids)
    return transcription[0]

# Gradio Interface
interface = gr.Interface(
    fn=transcribe,
    inputs=gr.Audio(source="microphone"),  # Corrected
    outputs="text",
    live=True
)

interface.launch()