File size: 2,426 Bytes
66925f4 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 |
import gradio as gr
from sentence_transformers import SentenceTransformer
from utils.similarity import get_similar_items
import numpy as np
import pandas as pd
import random
# Create title, description, and article strings
title = "Clothing Similarity Search 👕"
description = "**Transformer-based search engine** to fetch Amazon URLs for similar clothing items given a text description.\n\n**Data Collection**:\nTo scrape quality clothing data containing proper description and URL for the product, Apify's Amazon Product Scraper was used. The scraped data for various clothing categories was downloaded into a CSV file.\n\n**Data Cleaning**:\nPandas was used to clean and preprocess the text data by removing special characters, lowercasing, and applying text normalization techniques.\n\n**Making Embeddings**:\nSentence-transformers library was used to generate embeddings for the cleaned data using the [all-MiniLM-L6-v2](https://example.com/model-card) model. The embeddings were saved into a .npy file for faster similarity search retrieval.\n\n**Cosine Similarity**:\nCosine similarity was used to find the similarity between the query and the product embeddings.\n"
model = SentenceTransformer('model')
embeddings = np.load('data/embeddings.npy')
clothing_data = pd.read_csv('data/clothing_data_preprocessed.csv')
def getURL(text, top_k):
# Call your function to retrieve similar item URLs
similar_urls = get_similar_items(text, embeddings, clothing_data, top_k)
return similar_urls
input_text = gr.components.Textbox(lines=1, label="Input Descriptiont")
input_top_k = gr.components.Slider(label="Number of Recommendations", minimum=1, maximum=10, step=1, default=5)
output_html = gr.outputs.HTML(label="Similar Items")
def process_text(text, top_k):
urls = getURL(text, top_k)
random.shuffle(urls) # Shuffle the URLs for variety
html_links = "<br>".join([f'<a href="{url}" target="_blank">{url}</a>' for url in urls])
return f'<div style="padding: 20px">{html_links}</div>'
iface = gr.Interface(
fn=process_text,
inputs=[input_text, input_top_k],
outputs=output_html,
title=title,
description=description,
examples=[
["casual men's t-shirt", 3],
["stylish summer dress", 5],
["elegant evening gown", 7],
],
theme="default",
layout="vertical",
interpretation="default",
allow_flagging="never",
)
iface.launch()
|