{ "cells": [ { "cell_type": "markdown", "id": "68dda442", "metadata": {}, "source": [ "# Visualize DiffusionDB Images with WizMap\n", "\n", "In this notebook, we demonstrate how to use WizMap to visualize images and text\n", "from the [DiffusionDB](https://huggingface.co/datasets/poloclub/diffusiondb)\n", "dataset.\n" ] }, { "cell_type": "code", "execution_count": 1, "id": "aff7f8bf", "metadata": {}, "outputs": [], "source": [ "# Install wizmap\n", "# !pip install --upgrade wizmap umap-learn" ] }, { "cell_type": "code", "execution_count": 1, "id": "b4c848aa", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "/Users/jayw/.virtualenvs/openai/lib/python3.12/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n", " from .autonotebook import tqdm as notebook_tqdm\n" ] }, { "data": { "text/plain": [ "'0.1.7'" ] }, "execution_count": 1, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from glob import glob\n", "from os.path import exists, join, basename\n", "from tqdm import tqdm\n", "from json import load, dump\n", "from matplotlib import pyplot as plt\n", "from collections import Counter\n", "\n", "from sklearn.feature_extraction.text import CountVectorizer, TfidfTransformer\n", "from scipy.sparse import csr_matrix\n", "from sklearn.neighbors import KernelDensity\n", "from scipy.stats import norm\n", "from typing import Tuple\n", "from io import BytesIO\n", "from umap import UMAP\n", "from openai import AsyncOpenAI\n", "\n", "import pandas as pd\n", "import numpy as np\n", "import json\n", "import requests\n", "import urllib\n", "import wizmap\n", "\n", "SEED = 20230501\n", "\n", "plt.rcParams[\"figure.dpi\"] = 300\n", "wizmap.__version__" ] }, { "cell_type": "markdown", "id": "4bdb12e6", "metadata": {}, "source": [ "## 1. Extract Embeddings\n", "\n", "We use OpenAI's embedding model to get embeddings for 1k stable diffusion image\n", "prompts.\n", "\n", "You can get the `part-000001.json` by downloading the `part-000001.zip` file\n", "from\n", "[diffusiondb](https://huggingface.co/datasets/poloclub/diffusiondb/blob/main/diffusiondb-large-part-1/part-000001.zip).\n" ] }, { "cell_type": "code", "execution_count": 2, "id": "7ce9ada5", "metadata": {}, "outputs": [], "source": [ "# image_data = json.load(open(\"part-000001.json\", \"r\"))" ] }, { "cell_type": "code", "execution_count": null, "id": "31eaf46f", "metadata": {}, "outputs": [], "source": [ "# image_prompts = [(k, v[\"p\"]) for k, v in image_data.items()]\n", "# prompts = [item[1] for item in image_prompts]\n", "# image_names = [item[0] for item in image_prompts]\n", "\n", "# openai_client = AsyncOpenAI(api_key=input(\"OpenAI API Key: \"))\n", "\n", "# embedding_model = \"text-embedding-3-large\"\n", "# results = await openai_client.embeddings.create(input=prompts, model=embedding_model)\n", "\n", "# embeddings = [e.embedding for e in results.data]\n", "\n", "# np.savez(c\n", "# \"embedding_data.npz\",\n", "# embeddings=embeddings,\n", "# image_names=image_names,\n", "# prompts=prompts,\n", "# )" ] }, { "cell_type": "code", "execution_count": 2, "id": "0f118e02", "metadata": {}, "outputs": [], "source": [ "# To save time, we will just load pre-extracted embeddings\n", "EMBEDDING_URL = \"https://huggingface.co/datasets/xiaohk/embeddings/resolve/main/diffusiondb/openai-text-embeddings.npz\"\n", "stream = requests.get(EMBEDDING_URL, stream=True)\n", "embedding_data = np.load(BytesIO(stream.content), allow_pickle=True)" ] }, { "cell_type": "code", "execution_count": 3, "id": "c8046175", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Loaded 1000 embeddings\n" ] } ], "source": [ "image_names = embedding_data[\"image_names\"]\n", "prompts = embedding_data[\"prompts\"]\n", "embeddings = embedding_data[\"embeddings\"]\n", "\n", "print(f\"Loaded {len(embeddings)} embeddings\")" ] }, { "cell_type": "markdown", "id": "2c4270e4", "metadata": {}, "source": [ "## 2. Dimensionality Reduction\n", "\n", "Then, we apply dimensionality reduction techniques (e.g., UMAP, t-SNE, PCA) to\n", "project the embeddings from a 384-dimension space into a 2D space. Here we use\n", "UMAP, but you can use any dimensionality reduction technique you like.\n", "\n", "To save the time to run this notebook, we will use the UMAP's default\n", "parameters. However, it's a good practice to tune the parameters when you are\n", "using WizMap on your own dataset.\n" ] }, { "cell_type": "code", "execution_count": 4, "id": "d3d5f3c4", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "/Users/jayw/.virtualenvs/openai/lib/python3.12/site-packages/sklearn/utils/deprecation.py:151: FutureWarning: 'force_all_finite' was renamed to 'ensure_all_finite' in 1.6 and will be removed in 1.8.\n", " warnings.warn(\n" ] } ], "source": [ "reducer = UMAP(metric=\"cosine\")\n", "embeddings_2d = reducer.fit_transform(embeddings)" ] }, { "cell_type": "code", "execution_count": 5, "id": "a3693d3b", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plt.title(f\"UMAP Projected Embeddings of {len(embeddings)} images\")\n", "plt.scatter(embeddings_2d[:, 0], embeddings_2d[:, 1], s=0.6, alpha=0.2)\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "2a53d874", "metadata": {}, "source": [ "## 3. Generate Two JSON Files for WizMap\n", "\n", "To use WizMap on your embeddings, you need to generate two JSON files.\n", "\n", "- One JSON file encodes the contour plot and multi-level summaries.\n", "- The other JSON file encodes the raw data (e.g., IMDB reviews in this example).\n", "\n", "Fortunately, the `WizMap` Python library makes it extremely easy to generate\n", "these two files.\n" ] }, { "cell_type": "code", "execution_count": 6, "id": "617d967b", "metadata": {}, "outputs": [], "source": [ "# Extract and round the 2D embedding coordinates to 5 decimal places\n", "xs = [round(float(x), 5) for x in embeddings_2d[:, 0]]\n", "ys = [round(float(y), 5) for y in embeddings_2d[:, 1]]" ] }, { "cell_type": "code", "execution_count": 7, "id": "9f23b5ce", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Start generating contours...\n", "Start generating multi-level summaries...\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "1000it [00:00, 89022.69it/s]\n", "100%|██████████| 6/6 [00:00<00:00, 33.79it/s]\n" ] } ], "source": [ "# Here we are writing a json string as the text content\n", "# In the JSON object, we include the\n", "json_strings = []\n", "for i, name in enumerate(image_names):\n", " cur_url_part = f\"{name[:2]}/{name}\"\n", " json_data = {\n", " # thumbnail image name\n", " \"i\": cur_url_part,\n", " # tooltip text\n", " \"t\": prompts[i],\n", " # large thumbnail image name (optional, can be different from thumbnail)\n", " \"li\": cur_url_part,\n", " # link content (optional, these will be shown in the info window when clicked)\n", " # It can be useful to allow users to open external sources to view more information\n", " \"github\": f\"https://github.com/poloclub/diffusiondb-thumbnails/blob/master/{cur_url_part}\",\n", " }\n", " json_strings.append(json.dumps(json_data))\n", "\n", "json_point_content_config = wizmap.JsonPointContentConfig(\n", " groupLabels=None,\n", " textKey=\"t\",\n", " imageKey=\"i\",\n", " imageURLPrefix=\"https://raw.githubusercontent.com/poloclub/diffusiondb-thumbnails/master/\",\n", " largeImageKey=\"li\",\n", " largeImageURLPrefix=\"https://raw.githubusercontent.com/poloclub/diffusiondb-thumbnails/master/\",\n", " linkFieldKeys=[\"github\"],\n", ")\n", "\n", "# Generate the grid data\n", "grid_data = wizmap.generate_grid_dict(\n", " xs=xs,\n", " ys=ys,\n", " texts=json_strings,\n", ")" ] }, { "cell_type": "code", "execution_count": 8, "id": "c96a9dce", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Start generating data list...\n", "Start generating contours...\n", "Start generating multi-level summaries...\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "1000it [00:00, 460204.52it/s]\n", "100%|██████████| 6/6 [00:00<00:00, 37.69it/s]\n" ] } ], "source": [ "data_list = wizmap.generate_data_list(xs, ys, json_strings)\n", "grid_dict = wizmap.generate_grid_dict(\n", " xs,\n", " ys,\n", " json_strings,\n", " \"DiffusionDB Images\",\n", " json_point_content_config=json_point_content_config,\n", ")" ] }, { "cell_type": "code", "execution_count": 9, "id": "b80da89f", "metadata": {}, "outputs": [], "source": [ "# Save the JSON files\n", "wizmap.save_json_files(data_list, grid_dict, output_dir=\"./\")" ] }, { "cell_type": "markdown", "id": "6d6031e9", "metadata": {}, "source": [ "## 4. Host JSON Files and Display WizMap\n", "\n", "After generating these two JSON files (one with `.json` and one with `.ndjson`),\n", "you want to store them somewhere in the network so that you can provide two URLs\n", "to WizMap.\n", "\n", "Depending on your needs, there are many options to store the files.\n", "\n", "1. **Local host**. If you are running WizMap on your local machine, you can\n", " simply start a local server and use ‘local host’ URLs to send your JSON files\n", " to WizMap.\n", "2. **Static website hosting service** (e.g., GitHub page, Vercel, Hugging Face).\n", " You can use many free website hosting services to host your JSON files. A\n", " limitation is that these service usually have file size limits. For example,\n", " you can only include files that are less than 100MB in GitHub.\n", "3. **Cloud storage** (e.g., AWS S3, Cloudflare R2). The most general option is\n", " to put the JSON files on a cloud storage site. There is no size limit, but\n", " you might need to pay for the service.\n", "\n", "Here, we store `data.ndjson` and `grid.json` in\n", "[Hugging Face](https://huggingface.co/datasets/xiaohk/embeddings/tree/main/diffusiondb).\n" ] }, { "cell_type": "code", "execution_count": 10, "id": "77d88f5e", "metadata": {}, "outputs": [], "source": [ "data_url = \"https://huggingface.co/datasets/xiaohk/embeddings/resolve/main/diffusiondb/data.ndjson\"\n", "grid_url = \"https://huggingface.co/datasets/xiaohk/embeddings/resolve/main/diffusiondb/grid.json\"" ] }, { "cell_type": "code", "execution_count": 11, "id": "0bff3261", "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", " \n", " \n", " " ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Display wizmap\n", "wizmap.visualize(data_url, grid_url, height=700)" ] } ], "metadata": { "kernelspec": { "display_name": "openai", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.12.9" } }, "nbformat": 4, "nbformat_minor": 5 }