app.py

import pandas as pd, numpy as np
import os
from transformers import CLIPProcessor, CLIPTextModel, CLIPModel

import gradio as gr
import requests




model = CLIPModel.from_pretrained("openai/clip-vit-base-patch32")
processor = CLIPProcessor.from_pretrained("openai/clip-vit-base-patch32")
df = {0: pd.read_csv('data.csv'), 1: pd.read_csv('data2.csv')}
embeddings = {0: np.load('embeddings2.npy'), 1: np.load('embeddings.npy')}
for k in [0, 1]:
  embeddings[k] = np.divide(embeddings[k], np.sqrt(np.sum(embeddings[k]**2, axis=1, keepdims=True)))
  
def compute_text_embeddings(list_of_strings):
    inputs = processor(text=list_of_strings, return_tensors="pt", padding=True)
    return model.get_text_features(**inputs)
    
def download_img(path):
    img_data = requests.get(path).content
    local_path = path.split("/")[-1] + ".jpg"
    with open(local_path, 'wb') as handler:
        handler.write(img_data)
    return local_path
    
def predict(query):
    corpus = 'Movies'
    n_results=3

    text_embeddings = compute_text_embeddings([query]).detach().numpy()
    k = 0 if corpus == 'Unsplash' else 1
    results = np.argsort((embeddings[k]@text_embeddings.T)[:, 0])[-1:-n_results-1:-1]
    paths = [download_img(df[k].iloc[i]['path']) for i in results]
    print(paths)
    return paths

title = "Draw to Search"
iface = gr.Interface(
  fn=predict, 
  inputs=[gr.inputs.Textbox(label="text", lines=3)],
  outputs=[gr.outputs.Image(type="file"), gr.outputs.Image(type="file"), gr.outputs.Image(type="file")],
  title=title,
  examples=[["Sunset"]]
)
iface.launch(debug=True)