inference.py

from transformers import AutoFeatureExtractor, AutoModelForImageClassification
import torch
from datasets import load_dataset

model = AutoModelForImageClassification.from_pretrained(".", trust_remote_code=True)

dataset = load_dataset("huggingface/cats-image")
image = dataset["test"]["image"][0]

feature_extractor = AutoFeatureExtractor.from_pretrained(".")

inputs = feature_extractor(image, return_tensors="pt")

with torch.no_grad():
    logits = model(**inputs).logits

# model predicts one of the 1000 ImageNet classes
predicted_label = logits.argmax(-1).item()
print(model.config.id2label[predicted_label])