app.py

from NatureLM.models import NatureLM
from NatureLM.infer import Pipeline
import gradio as gr

model = NatureLM.from_pretrained("EarthSpeciesProject/NatureLM-audio").eval().to("cuda")
pipeline = Pipeline(model=model)

def transcribe(audio):
    results = pipeline([audio], ["Caption the audio using the species name."])
    return results[0]

gr.Interface(fn=transcribe, inputs=gr.Audio(type="filepath"), outputs="text").launch()