{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Similarity Queries using Nmslib Tutorial" ] }, { "attachments": {}, "cell_type": "markdown", "metadata": {}, "source": [ "This tutorial is about using the ([Non-Metric Space Library (NMSLIB)](https://github.com/nmslib/nmslib \"Link to nmslib repo\")) library for similarity queries with a Word2Vec model built with gensim.\n", "\n", "## Why use Nmslib?\n", "The current implementation for finding k nearest neighbors in a vector space in gensim has linear complexity via brute force in the number of indexed documents, although with extremely low constant factors. The retrieved results are exact, which is an overkill in many applications: approximate results retrieved in sub-linear time may be enough. Nmslib can find approximate nearest neighbors much faster.\n", "Compared to annoy, nmslib has more parameteres to control the build and query time and accuracy. Nmslib can achieve faster and more accurate nearest neighbors search than annoy. This figure shows a comparison between annoy and nmslib indexer with differents parameters. This shows nmslib is better than annoy.\n", "![nmslib.png](nmslib.png)\n", "\n", "## Prerequisites\n", "Additional libraries needed for this tutorial:\n", "- nmslib\n", "- annoy\n", "- psutil\n", "- matplotlib\n", "\n", "## Outline\n", "1. Download Text8 Corpus\n", "2. Build Word2Vec Model\n", "3. Construct NmslibIndex with model & make a similarity query\n", "4. Verify & Evaluate performance\n", "5. Evaluate relationship of parameters to initialization/query time and accuracy, compared with annoy\n", "6. Work with Google's word2vec C formats" ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "scrolled": true }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "CPython 3.6.0\n", "IPython 7.5.0\n", "\n", "gensim 3.7.3\n", "numpy 1.16.2\n", "scipy 1.2.1\n", "psutil 5.6.3\n", "matplotlib 3.1.0\n", "\n", "compiler : GCC 4.2.1 Compatible Apple LLVM 10.0.0 (clang-1000.11.45.5)\n", "system : Darwin\n", "release : 18.2.0\n", "machine : x86_64\n", "processor : i386\n", "CPU cores : 4\n", "interpreter: 64bit\n" ] } ], "source": [ "# pip install watermark\n", "%reload_ext watermark\n", "%watermark -v -m -p gensim,numpy,scipy,psutil,matplotlib" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 1. Download Text8 Corpus" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "--2019-06-27 13:48:42-- https://mattmahoney.net/dc/text8.zip\n", "Resolving mattmahoney.net... 67.195.197.75\n", "Connecting to mattmahoney.net|67.195.197.75|:80... connected.\n", "HTTP request sent, awaiting response... 200 OK\n", "Length: 31344016 (30M) [application/zip]\n", "Saving to: 'text8.zip'\n", "\n", "text8.zip 100%[=====================>] 29.89M 327KB/s in 98s \n", "\n", "2019-06-27 13:50:21 (313 KB/s) - 'text8.zip' saved [31344016/31344016]\n", "\n", "Archive: text8.zip\n", " inflating: text8 \n" ] } ], "source": [ "import os.path\n", "if not os.path.isfile('text8'):\n", " !wget -c https://mattmahoney.net/dc/text8.zip\n", " !unzip text8.zip" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Import & Set up Logging\n", "I'm not going to set up logging due to the verbose input displaying in notebooks, but if you want that, uncomment the lines in the cell below." ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [], "source": [ "LOGS = False\n", "\n", "if LOGS:\n", " import logging\n", " logging.basicConfig(format='%(asctime)s : %(levelname)s : %(message)s', level=logging.INFO)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 2. Build Word2Vec Model" ] }, { "cell_type": "code", "execution_count": 57, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Word2Vec(vocab=71290, size=100, alpha=0.05)\n" ] } ], "source": [ "from gensim.models import Word2Vec, KeyedVectors\n", "from gensim.models.word2vec import Text8Corpus\n", "\n", "# Using params from Word2Vec_FastText_Comparison\n", "\n", "params = {\n", " 'alpha': 0.05,\n", " 'size': 100,\n", " 'window': 5,\n", " 'iter': 5,\n", " 'min_count': 5,\n", " 'sample': 1e-4,\n", " 'sg': 1,\n", " 'hs': 0,\n", " 'negative': 5\n", "}\n", "\n", "model = Word2Vec(Text8Corpus('text8'), **params)\n", "print(model)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "See the [Word2Vec tutorial](word2vec.ipynb) for how to initialize and save this model." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Comparing the traditional implementation, Annoy and Nmslib approximation" ] }, { "cell_type": "code", "execution_count": 14, "metadata": {}, "outputs": [], "source": [ "# Set up the model and vector that we are using in the comparison\n", "from gensim.similarities.index import AnnoyIndexer\n", "from gensim.similarities.nmslib import NmslibIndexer\n", "\n", "model.init_sims()\n", "annoy_index = AnnoyIndexer(model, 300)\n", "nmslib_index = NmslibIndexer(model, {'M': 100, 'indexThreadQty': 1, 'efConstruction': 100}, {'efSearch': 10})" ] }, { "cell_type": "code", "execution_count": 15, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[('the', 1.0), ('of', 0.700629323720932), ('in', 0.6932138800621033), ('which', 0.6132444441318512), ('for', 0.6099657416343689)]\n", "[('the', 0.9999999701976776), ('of', 0.91037717461586), ('in', 0.9058823883533478), ('a', 0.8834112882614136), ('and', 0.8790014386177063)]\n", "[('the', 1.0000001192092896), ('of', 0.82075434923172), ('in', 0.811764657497406), ('a', 0.7668224573135376), ('and', 0.7580028772354126)]\n" ] } ], "source": [ "# Dry run to make sure both indices are fully in RAM\n", "vector = model.wv.syn0norm[0]\n", "print(model.most_similar([vector], topn=5, indexer=annoy_index))\n", "print(model.most_similar([vector], topn=5, indexer=nmslib_index))\n", "print(model.most_similar([vector], topn=5))" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [], "source": [ "import time\n", "import numpy as np" ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [], "source": [ "def avg_query_time(annoy_index=None, queries=1000):\n", " \"\"\"\n", " Average query time of a most_similar method over 1000 random queries,\n", " uses annoy if given an indexer\n", " \"\"\"\n", " total_time = 0\n", " for _ in range(queries):\n", " rand_vec = model.wv.syn0norm[np.random.randint(0, len(model.wv.vocab))]\n", " start_time = time.clock()\n", " model.most_similar([rand_vec], topn=5, indexer=annoy_index)\n", " total_time += time.clock() - start_time\n", " return total_time / queries" ] }, { "cell_type": "code", "execution_count": 18, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Gensim (s/query):\t0.00560\n", "Annoy (s/query):\t0.00099\n", "Nmslib (s/query):\t0.00046\n", "\n", "Nmslib is 12.16 times faster on average on this particular run\n", "\n", "Nmslib is 2.15 times faster on average than annoy on this particular run\n" ] } ], "source": [ "queries = 10000\n", "\n", "gensim_time = avg_query_time(queries=queries)\n", "annoy_time = avg_query_time(annoy_index, queries=queries)\n", "nmslib_time = avg_query_time(nmslib_index, queries=queries)\n", "print(\"Gensim (s/query):\\t{0:.5f}\".format(gensim_time))\n", "print(\"Annoy (s/query):\\t{0:.5f}\".format(annoy_time))\n", "print(\"Nmslib (s/query):\\t{0:.5f}\".format(nmslib_time))\n", "speed_improvement_gensim = gensim_time / nmslib_time\n", "speed_improvement_annoy = annoy_time / nmslib_time\n", "print (\"\\nNmslib is {0:.2f} times faster on average on this particular run\".format(speed_improvement_gensim))\n", "print (\"\\nNmslib is {0:.2f} times faster on average than annoy on this particular run\".format(speed_improvement_annoy))\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 3. Construct Nmslib Index with model & make a similarity query\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Creating an indexer\n", "An instance of `NmslibIndexer` needs to be created in order to use Nmslib in gensim. The `NmslibIndexer` class is located in `gensim.similarities.nmslib`\n", "\n", "`NmslibIndexer()` takes three parameters:\n", "\n", "**`model`**: A `Word2Vec` or `Doc2Vec` model\n", "\n", "**`index_params`**: Parameters for building nmslib indexer. `index_params` effects the build time and the index size. The example is `{'M': 100, 'indexThreadQty': 1, 'efConstruction': 100}`. Increasing the value of `M` and `efConstruction` improves the accuracy of search. However this also leads to longer indexing times. `indexThreadQty` is the number of thread. \n", "\n", "**`query_time_params`**: Parameters for querying on nmslib indexer. `query_time_params` effects the query time and the search accuracy. The example is `{'efSearch': 100}`. A larger `efSearch` will give more accurate results, but larger query time. \n", "\n", "More information can be found [here](https://github.com/nmslib/nmslib/blob/master/manual/methods.md). The relationship between parameters, build/query time, and accuracy will be investigated later in the tutorial. \n", "\n", "Now that we are ready to make a query, lets find the top 5 most similar words to \"science\" in the Text8 corpus. To make a similarity query we call `Word2Vec.most_similar` like we would traditionally, but with an added parameter, `indexer`. The only supported indexerers in gensim as of now are Annoy and Nmslib. " ] }, { "cell_type": "code", "execution_count": 19, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Approximate Neighbors\n", "('science', 1.0000000596046448)\n", "('fiction', 0.8769421577453613)\n", "('protoscience', 0.8432635962963104)\n", "('multidisciplinary', 0.835610032081604)\n", "('sciences', 0.8348604440689087)\n", "('astrobiology', 0.8340338170528412)\n", "('actuarial', 0.8339103162288666)\n", "('interdisciplinary', 0.8327268362045288)\n", "('xenobiology', 0.8318319618701935)\n", "('criminology', 0.8261869251728058)\n", "('futurists', 0.82555091381073)\n", "\n", "Normal (not nmslib-indexed) Neighbors\n", "('science', 0.9999998807907104)\n", "('fiction', 0.7538841962814331)\n", "('protoscience', 0.6865270733833313)\n", "('multidisciplinary', 0.6712199449539185)\n", "('sciences', 0.6697208881378174)\n", "('astrobiology', 0.6680675148963928)\n", "('actuarial', 0.6678205132484436)\n", "('interdisciplinary', 0.6654534339904785)\n", "('xenobiology', 0.663663923740387)\n", "('vernor', 0.6569585800170898)\n", "('criminology', 0.652373731136322)\n" ] } ], "source": [ "# Building nmslib indexer\n", "nmslib_index = NmslibIndexer(model, {'M': 100, 'indexThreadQty': 1, 'efConstruction': 100}, {'efSearch': 10})\n", "# Derive the vector for the word \"science\" in our model\n", "vector = model[\"science\"]\n", "# The instance of AnnoyIndexer we just created is passed \n", "approximate_neighbors = model.most_similar([vector], topn=11, indexer=nmslib_index)\n", "\n", "# Neatly print the approximate_neighbors and their corresponding cosine similarity values\n", "print(\"Approximate Neighbors\")\n", "for neighbor in approximate_neighbors:\n", " print(neighbor)\n", "\n", "normal_neighbors = model.most_similar([vector], topn=11)\n", "print(\"\\nNormal (not nmslib-indexed) Neighbors\")\n", "for neighbor in normal_neighbors:\n", " print(neighbor)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Analyzing the results" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The closer the cosine similarity of a vector is to 1, the more similar that word is to our query, which was the vector for \"science\". In this case the results are almostly same." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 4. Verify & Evaluate performance" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Persisting Indexes\n", "You can save and load your indexes from/to disk to prevent having to construct them each time. This will create two files on disk, _fname_ and _fname.d_. Both files are needed to correctly restore all attributes. " ] }, { "cell_type": "code", "execution_count": 23, "metadata": {}, "outputs": [], "source": [ "import os\n", "\n", "fname = '/tmp/mymodel.index'\n", "\n", "# Persist index to disk\n", "nmslib_index.save(fname)\n", "\n", "# Load index back\n", "if os.path.exists(fname):\n", " nmslib_index2 = NmslibIndexer.load(fname)\n", " nmslib_index2.model = model" ] }, { "cell_type": "code", "execution_count": 24, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "('science', 1.0000000596046448)\n", "('fiction', 0.8769421577453613)\n", "('protoscience', 0.8432635962963104)\n", "('multidisciplinary', 0.835610032081604)\n", "('sciences', 0.8348604440689087)\n", "('astrobiology', 0.8340338170528412)\n", "('actuarial', 0.8339103162288666)\n", "('interdisciplinary', 0.8327268362045288)\n", "('xenobiology', 0.8318319618701935)\n", "('criminology', 0.8261869251728058)\n", "('futurists', 0.82555091381073)\n" ] } ], "source": [ "# Results should be identical to above\n", "vector = model[\"science\"]\n", "approximate_neighbors2 = model.most_similar([vector], topn=11, indexer=nmslib_index2)\n", "for neighbor in approximate_neighbors2:\n", " print(neighbor)\n", " \n", "assert approximate_neighbors == approximate_neighbors2" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Be sure to use the same model at load that was used originally, otherwise you will get unexpected behaviors." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Save memory by memory-mapping indices saved to disk\n", "\n", "Nmslib library has a useful feature that indices can be memory-mapped from disk. It saves memory when the same index is used by several processes.\n", "\n", "Below are two snippets of code. First one has a separate index for each process. The second snipped shares the index between two processes via memory-mapping. The second example uses less total RAM as it is shared." ] }, { "cell_type": "code", "execution_count": 26, "metadata": {}, "outputs": [], "source": [ "# Remove verbosity from code below (if logging active)\n", "\n", "if LOGS:\n", " logging.disable(logging.CRITICAL)" ] }, { "cell_type": "code", "execution_count": 27, "metadata": {}, "outputs": [], "source": [ "from multiprocessing import Process\n", "import psutil" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Bad Example: Two processes load the Word2vec model from disk and create there own Nmslib indices from that model. " ] }, { "cell_type": "code", "execution_count": 28, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Process Id: 36262\n", "\n", "Memory used by process 36262: pmem(rss=600539136, vms=6052294656, pfaults=170185, pageins=2)\n", "---\n", "Process Id: 36263\n", "\n", "Memory used by process 36263: pmem(rss=600539136, vms=6052294656, pfaults=170155, pageins=0)\n", "---\n", "CPU times: user 547 ms, sys: 308 ms, total: 856 ms\n", "Wall time: 59.2 s\n" ] } ], "source": [ "%%time\n", "\n", "model.save('/tmp/mymodel.pkl')\n", "\n", "def f(process_id):\n", " print('Process Id: {}'.format(os.getpid()))\n", " process = psutil.Process(os.getpid())\n", " new_model = Word2Vec.load('/tmp/mymodel.pkl')\n", " vector = new_model[\"science\"]\n", " nmslib_index = NmslibIndexer(new_model, {'M': 100, 'indexThreadQty': 1, 'efConstruction': 100}, {'efSearch': 10})\n", " approximate_neighbors = new_model.most_similar([vector], topn=5, indexer=nmslib_index)\n", " print('\\nMemory used by process {}: {}\\n---'.format(os.getpid(), process.memory_info()))\n", "\n", "# Creating and running two parallel process to share the same index file.\n", "p1 = Process(target=f, args=('1',))\n", "p1.start()\n", "p1.join()\n", "p2 = Process(target=f, args=('2',))\n", "p2.start()\n", "p2.join()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Good example. Two processes load both the Word2vec model and index from disk and memory-map the index\n" ] }, { "cell_type": "code", "execution_count": 29, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Process Id: 36265\n", "\n", "Memory used by process 36265: pmem(rss=417345536, vms=5981540352, pfaults=129428, pageins=0)\n", "---\n", "Process Id: 36266\n", "\n", "Memory used by process 36266: pmem(rss=417579008, vms=5981540352, pfaults=129528, pageins=0)\n", "---\n", "CPU times: user 581 ms, sys: 266 ms, total: 847 ms\n", "Wall time: 3.74 s\n" ] } ], "source": [ "%%time\n", "\n", "model.save('/tmp/mymodel.pkl')\n", "\n", "def f(process_id):\n", " print('Process Id: {}'.format(os.getpid()))\n", " process = psutil.Process(os.getpid())\n", " new_model = Word2Vec.load('/tmp/mymodel.pkl')\n", " vector = new_model[\"science\"]\n", " nmslib_index = NmslibIndexer.load('/tmp/mymodel.index')\n", " nmslib_index.model = new_model\n", " approximate_neighbors = new_model.most_similar([vector], topn=5, indexer=nmslib_index)\n", " print('\\nMemory used by process {}: {}\\n---'.format(os.getpid(), process.memory_info()))\n", "\n", "# Creating and running two parallel process to share the same index file.\n", "p1 = Process(target=f, args=('1',))\n", "p1.start()\n", "p1.join()\n", "p2 = Process(target=f, args=('2',))\n", "p2.start()\n", "p2.join()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 5. Evaluate relationship of parameters to initialization/query time and accuracy, compared with annoy\n" ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [], "source": [ "import matplotlib.pyplot as plt\n", "%matplotlib inline" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Build dataset of Initialization times and accuracy measures" ] }, { "cell_type": "code", "execution_count": 14, "metadata": {}, "outputs": [], "source": [ "exact_results = [element[0] for element in model.most_similar([model.wv.syn0norm[0]], topn=100)]" ] }, { "cell_type": "code", "execution_count": 48, "metadata": {}, "outputs": [], "source": [ "# For calculating query time\n", "queries = 1000" ] }, { "cell_type": "code", "execution_count": 49, "metadata": {}, "outputs": [], "source": [ "def create_evaluation_graph(x_values, y_values_init, y_values_accuracy, y_values_query, param_name):\n", " plt.figure(1, figsize=(12, 6))\n", " plt.subplot(231)\n", " plt.plot(x_values, y_values_init)\n", " plt.title(\"{} vs initalization time\".format(param_name))\n", " plt.ylabel(\"Initialization time (s)\")\n", " plt.xlabel(param_name)\n", " plt.subplot(232)\n", " plt.plot(x_values, y_values_accuracy)\n", " plt.title(\"{} vs accuracy\".format(param_name))\n", " plt.ylabel(\"% accuracy\")\n", " plt.xlabel(param_name)\n", " plt.tight_layout()\n", " plt.subplot(233)\n", " plt.plot(y_values_init, y_values_accuracy)\n", " plt.title(\"Initialization time vs accuracy\")\n", " plt.ylabel(\"% accuracy\")\n", " plt.xlabel(\"Initialization time (s)\")\n", " plt.tight_layout()\n", " plt.subplot(234)\n", " plt.plot(x_values, y_values_query)\n", " plt.title(\"{} vs query time\".format(param_name))\n", " plt.ylabel(\"query time\")\n", " plt.xlabel(param_name)\n", " plt.tight_layout()\n", " plt.subplot(235)\n", " plt.plot(y_values_query, y_values_accuracy)\n", " plt.title(\"query time vs accuracy\")\n", " plt.ylabel(\"% accuracy\")\n", " plt.xlabel(\"query time (s)\")\n", " plt.tight_layout()\n", " plt.show()" ] }, { "cell_type": "code", "execution_count": 50, "metadata": {}, "outputs": [], "source": [ "def evaluate_nmslib_performance(parameter, is_parameter_query, parameter_start, parameter_end, parameter_step):\n", " nmslib_x_values = []\n", " nmslib_y_values_init = []\n", " nmslib_y_values_accuracy = []\n", " nmslib_y_values_query = []\n", " index_params = {'M': 100, 'indexThreadQty': 10, 'efConstruction': 100, 'post': 0}\n", " query_params = {'efSearch': 100}\n", " \n", " for x in range(parameter_start, parameter_end, parameter_step):\n", " nmslib_x_values.append(x)\n", " start_time = time.time()\n", " if is_parameter_query:\n", " query_params[parameter] = x\n", " else:\n", " index_params[parameter] = x\n", " nmslib_index = NmslibIndexer(model\n", " , index_params\n", " , query_params)\n", " nmslib_y_values_init.append(time.time() - start_time)\n", " approximate_results = model.most_similar([model.wv.syn0norm[0]], topn=100, indexer=nmslib_index)\n", " top_words = [result[0] for result in approximate_results]\n", " nmslib_y_values_accuracy.append(len(set(top_words).intersection(exact_results)))\n", " nmslib_y_values_query.append(avg_query_time(nmslib_index, queries=queries))\n", " create_evaluation_graph(nmslib_x_values,\n", " nmslib_y_values_init, \n", " nmslib_y_values_accuracy, \n", " nmslib_y_values_query, \n", " parameter)" ] }, { "cell_type": "code", "execution_count": 51, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# Evaluate nmslib indexer, changing the parameter M\n", "evaluate_nmslib_performance(\"M\", False, 50, 401, 50)" ] }, { "cell_type": "code", "execution_count": 52, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# Evaluate nmslib indexer, changing the parameter efConstruction\n", "evaluate_nmslib_performance(\"efConstruction\", False, 50, 1001, 100)" ] }, { "cell_type": "code", "execution_count": 53, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# Evaluate nmslib indexer, changing the parameter efSearch\n", "evaluate_nmslib_performance(\"efSearch\", True, 50, 401, 100)" ] }, { "cell_type": "code", "execution_count": 54, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# Evaluate annoy indexer, changing the parameter num_tree\n", "annoy_x_values = []\n", "annoy_y_values_init = []\n", "annoy_y_values_accuracy = []\n", "annoy_y_values_query = []\n", "\n", "for x in range(100, 401, 50):\n", " annoy_x_values.append(x)\n", " start_time = time.time()\n", " annoy_index = AnnoyIndexer(model, x)\n", " annoy_y_values_init.append(time.time() - start_time)\n", " approximate_results = model.most_similar([model.wv.syn0norm[0]], topn=100, indexer=annoy_index)\n", " top_words = [result[0] for result in approximate_results]\n", " annoy_y_values_accuracy.append(len(set(top_words).intersection(exact_results)))\n", " annoy_y_values_query.append(avg_query_time(annoy_index, queries=queries))\n", "create_evaluation_graph(annoy_x_values,\n", " annoy_y_values_init, \n", " annoy_y_values_accuracy, \n", " annoy_y_values_query, \n", " \"num_tree\")" ] }, { "cell_type": "code", "execution_count": 55, "metadata": { "scrolled": true }, "outputs": [], "source": [ "# nmslib indexer changing the parameter M, efConstruction, efSearch\n", "nmslib_y_values_init = []\n", "nmslib_y_values_accuracy = []\n", "nmslib_y_values_query = []\n", "\n", "for M in [100, 200]:\n", " for efConstruction in [100, 200]:\n", " for efSearch in [100, 200]:\n", " start_time = time.time()\n", " nmslib_index = NmslibIndexer(model, \n", " {'M': M, 'indexThreadQty': 10, 'efConstruction': efConstruction, 'post': 0},\n", " {'efSearch': efSearch})\n", " nmslib_y_values_init.append(time.time() - start_time)\n", " approximate_results = model.most_similar([model.wv.syn0norm[0]], topn=100, indexer=nmslib_index)\n", " top_words = [result[0] for result in approximate_results]\n", " nmslib_y_values_accuracy.append(len(set(top_words).intersection(exact_results)))\n", " nmslib_y_values_query.append(avg_query_time(nmslib_index, queries=queries))\n" ] }, { "cell_type": "code", "execution_count": 56, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# Make a comparison between annoy and nmslib indexer\n", "plt.figure(1, figsize=(12, 6))\n", "plt.subplot(121)\n", "plt.scatter(nmslib_y_values_init, nmslib_y_values_accuracy, label=\"nmslib\", color='r', marker='o')\n", "plt.scatter(annoy_y_values_init, annoy_y_values_accuracy, label=\"annoy\", color='b', marker='x')\n", "plt.legend()\n", "plt.title(\"Initialization time vs accuracy. Upper left is better.\")\n", "plt.ylabel(\"% accuracy\")\n", "plt.xlabel(\"Initialization time (s)\")\n", "plt.subplot(122)\n", "plt.scatter(nmslib_y_values_query, nmslib_y_values_accuracy, label=\"nmslib\", color='r', marker='o')\n", "plt.scatter(annoy_y_values_query, annoy_y_values_accuracy, label=\"annoy\", color='b', marker='x')\n", "plt.legend()\n", "plt.title(\"Query time vs accuracy. Upper left is better.\")\n", "plt.ylabel(\"% accuracy\")\n", "plt.xlabel(\"Query time (s)\")\n", "plt.xlim(min(nmslib_y_values_query+annoy_y_values_query), max(nmslib_y_values_query+annoy_y_values_query))\n", "plt.tight_layout()\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 6. Work with Google word2vec files\n", "\n", "Our model can be exported to a word2vec C format. There is a binary and a plain text word2vec format. Both can be read with a variety of other software, or imported back into gensim as a `KeyedVectors` object." ] }, { "cell_type": "code", "execution_count": 74, "metadata": {}, "outputs": [], "source": [ "# To export our model as text\n", "model.wv.save_word2vec_format('/tmp/vectors.txt', binary=False)" ] }, { "cell_type": "code", "execution_count": 37, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "71290 100\n", "the -0.20548718 0.19478682 -0.15149663 -0.31142342 0.014471135 -0.17996445 -0.07373469 0.09573618 -0.06328416 0.15571225 0.021413572 -0.12776679 -0.16940169 0.15807933 -0.21688043 0.074471496 0.08091913 0.07911484 0.31909388 -0.12297766 0.16993207 -0.02962172 0.08481803 0.12566781 0.02949822 -0.009697897 0.10780254 -0.102594994 -0.03935867 0.2679534 -0.061677158 -0.26071545 0.16285498 0.051780242 -0.1697231 0.24037386 0.0726078 -0.090416454 0.0776138 -0.06611322 0.057015926 0.07859522 -0.1910579 0.2974446 -0.033308737 -0.07360004 0.10797568 0.3595622 0.26797494 -0.062491674 0.21733648 -0.08524646 -0.06860078 -0.01714756 0.1305319 -0.09754544 -0.11249808 0.27328265 0.0041686473 -0.09874534 -0.30283058 0.111191 -0.026302295 -0.095534325 -0.0907799 -0.09120328 0.00068672217 0.31802058 -0.03345536 0.103762306 0.068564445 0.07402255 0.013657822 0.020439595 0.14985266 -0.13516407 0.36674532 0.0077319355 0.24709526 0.07666927 -0.11271039 0.02220251 -0.0235162 0.06409378 -0.10098407 0.23384795 0.094924204 0.061178204 0.19992544 0.29211295 0.004201066 0.12897494 -0.07837112 0.06269808 0.18545003 0.10452479 0.093281575 -0.24360427 0.01996208 0.35048977\n", "of -0.19580448 0.209091 -0.099757686 -0.21272708 0.022300478 -0.21946296 -0.19841547 0.09413904 0.08382151 0.20281556 0.12914163 -0.062828615 -0.24957581 0.29098126 -0.12465087 0.09803499 0.08116001 0.06726196 0.2615518 -0.035177294 0.06914535 -0.09427601 0.0011476864 0.23290296 0.0034714406 0.009549841 0.03912403 0.11168332 -0.10166585 0.34404838 -0.08050078 -0.3933344 0.17402974 -0.15764657 -0.21314052 0.20686424 0.034598276 0.16018851 0.14357902 0.046236232 0.075825214 0.029642927 -0.059190348 0.4163471 -0.1367429 -0.017528763 0.19181107 0.2601198 -0.020112848 -0.23402186 0.2841525 -0.10974068 -0.002565893 0.00070757867 0.13032512 -0.002393167 -0.14120881 0.22138755 0.027622899 0.06942904 -0.39498508 0.1133777 0.19053803 -0.062439334 -0.025348661 -0.11142109 0.015062763 0.3285828 -0.09184951 0.2661699 -0.11710489 -0.15770112 -0.12773664 0.15360866 0.08832063 -0.20914252 0.32392043 -0.023845093 0.3131217 0.08974748 -0.11354328 -0.2037927 -0.06780317 0.20184614 -0.13539118 0.2029387 0.07701099 -0.048417546 0.09797926 0.284204 0.036153372 0.17912139 -0.118080124 -0.025121484 0.10947146 0.09291596 0.1244357 0.006804844 0.025120731 0.28958535\n" ] } ], "source": [ "from smart_open import open\n", "# View the first 3 lines of the exported file\n", "\n", "# The first line has the total number of entries and the vector dimension count. \n", "# The next lines have a key (a string) followed by its vector.\n", "with open('/tmp/vectors.txt') as myfile:\n", " for i in range(3):\n", " print(myfile.readline().strip())" ] }, { "cell_type": "code", "execution_count": 38, "metadata": {}, "outputs": [], "source": [ "# To import a word2vec text model\n", "wv = KeyedVectors.load_word2vec_format('/tmp/vectors.txt', binary=False)" ] }, { "cell_type": "code", "execution_count": 39, "metadata": {}, "outputs": [], "source": [ "# To export our model as binary\n", "model.wv.save_word2vec_format('/tmp/vectors.bin', binary=True)" ] }, { "cell_type": "code", "execution_count": 40, "metadata": {}, "outputs": [], "source": [ "# To import a word2vec binary model\n", "wv = KeyedVectors.load_word2vec_format('/tmp/vectors.bin', binary=True)" ] }, { "cell_type": "code", "execution_count": 41, "metadata": {}, "outputs": [], "source": [ "# To create and save Nmslib Index from a loaded `KeyedVectors` object \n", "nmslib_index = NmslibIndexer(wv, \n", " {'M': 100, 'indexThreadQty': 1, 'efConstruction': 100}, {'efSearch': 100})\n", "nmslib_index.save('/tmp/mymodel.index')" ] }, { "cell_type": "code", "execution_count": 44, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Approximate Neighbors\n", "('cat', 1.0)\n", "('cats', 0.8335747122764587)\n", "('meow', 0.8298084139823914)\n", "('leopardus', 0.8215900659561157)\n", "('albino', 0.8207193315029144)\n", "('poodle', 0.8168523907661438)\n", "('saimiri', 0.8167656660079956)\n", "('squirrel', 0.8138661682605743)\n", "('sighthound', 0.8134040832519531)\n", "('proboscis', 0.8130573630332947)\n", "('eared', 0.812840610742569)\n", "\n", "Normal (not Nmslib-indexed) Neighbors\n", "('cat', 1.0)\n", "('cats', 0.6671494245529175)\n", "('meow', 0.6596168279647827)\n", "('leopardus', 0.6431801319122314)\n", "('albino', 0.6414386034011841)\n", "('poodle', 0.633704662322998)\n", "('saimiri', 0.633531391620636)\n", "('squirrel', 0.6277321577072144)\n", "('sighthound', 0.6268081665039062)\n", "('proboscis', 0.6261147260665894)\n", "('eared', 0.6256811618804932)\n" ] } ], "source": [ "# Load and test the saved word vectors and saved nmslib index\n", "wv = KeyedVectors.load_word2vec_format('/tmp/vectors.bin', binary=True)\n", "nmslib_index = NmslibIndexer.load('/tmp/mymodel.index')\n", "nmslib_index.model = wv\n", "\n", "vector = wv[\"cat\"]\n", "approximate_neighbors = wv.most_similar([vector], topn=11, indexer=nmslib_index)\n", "# Neatly print the approximate_neighbors and their corresponding cosine similarity values\n", "print(\"Approximate Neighbors\")\n", "for neighbor in approximate_neighbors:\n", " print(neighbor)\n", "\n", "normal_neighbors = wv.most_similar([vector], topn=11)\n", "print(\"\\nNormal (not Nmslib-indexed) Neighbors\")\n", "for neighbor in normal_neighbors:\n", " print(neighbor)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Recap\n", "In this notebook we used the Nmslib module to build an indexed approximation of our word embeddings. To do so, we did the following steps:\n", "1. Download Text8 Corpus\n", "2. Build Word2Vec Model\n", "3. Construct NmslibIndex with model & make a similarity query\n", "4. Verify & Evaluate performance\n", "5. Evaluate relationship of parameters to initialization/query time and accuracy, compared with annoy\n", "6. Work with Google's word2vec C formats" ] } ], "metadata": { "kernelspec": { "display_name": "Python 3", "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": "3.6.0" } }, "nbformat": 4, "nbformat_minor": 2 }