app.py

from PIL import Image
import torch
import joblib
import numpy as np
from transformers import CLIPProcessor, CLIPModel
import gradio as gr
from config import DEVICE, MODEL_SAVE_PATH

# Load models
clip_model = CLIPModel.from_pretrained("openai/clip-vit-base-patch32").to(DEVICE)
clip_processor = CLIPProcessor.from_pretrained("openai/clip-vit-base-patch32")
classifier = joblib.load(MODEL_SAVE_PATH)
label_encoder = joblib.load("label_encoder.joblib")

def classify_image(image):
    image = image.convert("RGB")
    inputs = clip_processor(images=image, return_tensors="pt").to(DEVICE)

    with torch.no_grad():
        image_features = clip_model.get_image_features(**inputs)
        image_features = image_features / image_features.norm(p=2, dim=-1, keepdim=True)
        features = image_features.cpu().numpy()

    pred = classifier.predict(features)
    label = label_encoder.inverse_transform(pred)[0]
    return f"Predicted Class: {label}"

gr.Interface(
    fn=classify_image,
    inputs=gr.Image(type="pil"),
    outputs="text",
    title="CLIP Model Classifier",
    description="Upload an image. The model will classify it using CLIP + a custom classifier."
).launch()