app.py

import transformers
from transformers import pipeline
import gradio as gr
import os 
import sys 
os.system("pip install evaluate")
os.system("pip install datasets")
os.system("pip install spicy")
os.system("pip install soundfile")
os.system("pip install datasets[audio]")
from evaluate import evaluator
from datasets import load_dataset, Audio


p = pipeline("automatic-speech-recognition")

task_evaluator = evaluator("automatic-speech-recognition")
#url = {"test" : "https://huggingface.co/datasets/mskov/miso_test/blob/main/test_set.parquet"}
data = load_dataset("audiofolder", data_dir="mskov/miso_test/test_set")
results = task_evaluator.compute(
    model_or_pipeline="https://huggingface.co/mskov/whisper_miso",
    data=data,
    input_column="file_name",
    label_column="category",
    metric="wer",
)
print(results)


def transcribe(audio, state=""):
    text = p(audio)["text"]
    state += text + " "
    return state, state

gr.Interface(
    fn=transcribe, 
    inputs=[
        gr.Audio(source="microphone", type="filepath", streaming=True), 
        "state"
    ],
    outputs=[
        "textbox",
        "state"
    ],
    live=True).launch()