 Voir sur TensorFlow.org |  Exécuter dans Google Colab |  Voir sur GitHub |  Télécharger le carnet |  Voir le modèle TF Hub |
Ce didacticiel illustre comment générer des intégrations à partir d'un module TensorFlow Hub (TF-Hub) à partir des données d'entrée et créer un index approximatif des voisins les plus proches (ANN) à l'aide des intégrations extraites. L'index peut ensuite être utilisé pour la recherche et la recherche de similarités en temps réel.
Lorsqu'il s'agit d'un vaste corpus de données, il n'est pas efficace d'effectuer une correspondance exacte en analysant l'ensemble du référentiel pour trouver les éléments les plus similaires à une requête donnée en temps réel. Ainsi, nous utilisons un algorithme de correspondance de similarité approximative qui nous permet de sacrifier un peu de précision dans la recherche des correspondances exactes des voisins les plus proches contre une augmentation significative de la vitesse.
Dans ce didacticiel, nous montrons un exemple de recherche de texte en temps réel sur un corpus de titres d'actualité pour trouver les titres les plus similaires à une requête. Contrairement à la recherche par mot-clé, celle-ci capture la similarité sémantique codée dans l’intégration du texte.
Les étapes de ce tutoriel sont :
- Téléchargez des exemples de données.
- Générer des intégrations pour les données à l'aide d'un module TF-Hub
- Créer un index ANN pour les intégrations
- Utiliser l'index pour la correspondance de similarité
Nous utilisons Apache Beam pour générer les intégrations à partir du module TF-Hub. Nous utilisons également la bibliothèque ANNOY de Spotify pour créer l'index approximatif des voisins les plus proches.
Plus de modèles
Pour les modèles qui ont la même architecture mais qui ont été formés sur un langage différent, reportez-vous à cette collection. Vous trouverez ici toutes les intégrations de texte actuellement hébergées sur tfhub.dev .
Installation
Installez les bibliothèques requises.
pip install -q apache_beampip install -q 'scikit_learn~=0.23.0' # For gaussian_random_matrix.pip install -q annoy
Importez les bibliothèques requises
import os
import sys
import pickle
from collections import namedtuple
from datetime import datetime
import numpy as np
import apache_beam as beam
from apache_beam.transforms import util
import tensorflow as tf
import tensorflow_hub as hub
import annoy
from sklearn.random_projection import gaussian_random_matrix
print('TF version: {}'.format(tf.__version__))
print('TF-Hub version: {}'.format(hub.__version__))
print('Apache Beam version: {}'.format(beam.__version__))
TF version: 2.4.0 TF-Hub version: 0.11.0 Apache Beam version: 2.26.0
1. Téléchargez des exemples de données
L'ensemble de données A Million News Headlines contient des titres d'actualité publiés sur une période de 15 ans et provenant de la réputée Australian Broadcasting Corp. (ABC). Cet ensemble de données d'actualité présente un historique résumé des événements remarquables survenus dans le monde du début 2003 à la fin 2017, avec un accent plus granulaire sur l'Australie.
Format : Données sur deux colonnes séparées par des tabulations : 1) date de publication et 2) texte du titre. Nous ne nous intéressons qu'au texte du titre.
wget 'https://dataverse.harvard.edu/api/access/datafile/3450625?format=tab&gbrecs=true' -O raw.tsvwc -l raw.tsvhead raw.tsv
--2021-01-07 12:50:08-- https://dataverse.harvard.edu/api/access/datafile/3450625?format=tab&gbrecs=true Resolving dataverse.harvard.edu (dataverse.harvard.edu)... 206.191.184.198 Connecting to dataverse.harvard.edu (dataverse.harvard.edu)|206.191.184.198|:443... connected. HTTP request sent, awaiting response... 200 OK Length: 57600231 (55M) [text/tab-separated-values] Saving to: ‘raw.tsv’ raw.tsv 100%[===================>] 54.93M 14.7MB/s in 4.4s 2021-01-07 12:50:14 (12.4 MB/s) - ‘raw.tsv’ saved [57600231/57600231] 1103664 raw.tsv publish_date headline_text 20030219 "aba decides against community broadcasting licence" 20030219 "act fire witnesses must be aware of defamation" 20030219 "a g calls for infrastructure protection summit" 20030219 "air nz staff in aust strike for pay rise" 20030219 "air nz strike to affect australian travellers" 20030219 "ambitious olsson wins triple jump" 20030219 "antic delighted with record breaking barca" 20030219 "aussie qualifier stosur wastes four memphis match" 20030219 "aust addresses un security council over iraq"
Par souci de simplicité, nous conservons uniquement le texte du titre et supprimons la date de publication.
!rm -r corpus
!mkdir corpus
with open('corpus/text.txt', 'w') as out_file:
with open('raw.tsv', 'r') as in_file:
for line in in_file:
headline = line.split('\t')[1].strip().strip('"')
out_file.write(headline+"\n")
rm: cannot remove 'corpus': No such file or directory
tail corpus/text.txt
severe storms forecast for nye in south east queensland snake catcher pleads for people not to kill reptiles south australia prepares for party to welcome new year strikers cool off the heat with big win in adelaide stunning images from the sydney to hobart yacht the ashes smiths warners near miss liven up boxing day test timelapse: brisbanes new year fireworks what 2017 meant to the kids of australia what the papodopoulos meeting may mean for ausus who is george papadopoulos the former trump campaign aide
2. Générez des intégrations pour les données.
Dans ce didacticiel, nous utilisons le modèle de langage de réseau neuronal (NNLM) pour générer des intégrations pour les données principales. Les incorporations de phrases peuvent ensuite être facilement utilisées pour calculer la similarité de signification au niveau de la phrase. Nous exécutons le processus de génération d'intégration à l'aide d'Apache Beam.
Méthode d'extraction d'intégration
embed_fn = None
def generate_embeddings(text, module_url, random_projection_matrix=None):
# Beam will run this function in different processes that need to
# import hub and load embed_fn (if not previously loaded)
global embed_fn
if embed_fn is None:
embed_fn = hub.load(module_url)
embedding = embed_fn(text).numpy()
if random_projection_matrix is not None:
embedding = embedding.dot(random_projection_matrix)
return text, embedding
Convertir en tf.Exemple de méthode
def to_tf_example(entries):
examples = []
text_list, embedding_list = entries
for i in range(len(text_list)):
text = text_list[i]
embedding = embedding_list[i]
features = {
'text': tf.train.Feature(
bytes_list=tf.train.BytesList(value=[text.encode('utf-8')])),
'embedding': tf.train.Feature(
float_list=tf.train.FloatList(value=embedding.tolist()))
}
example = tf.train.Example(
features=tf.train.Features(
feature=features)).SerializeToString(deterministic=True)
examples.append(example)
return examples
Pipeline de poutre
def run_hub2emb(args):
'''Runs the embedding generation pipeline'''
options = beam.options.pipeline_options.PipelineOptions(**args)
args = namedtuple("options", args.keys())(*args.values())
with beam.Pipeline(args.runner, options=options) as pipeline:
(
pipeline
| 'Read sentences from files' >> beam.io.ReadFromText(
file_pattern=args.data_dir)
| 'Batch elements' >> util.BatchElements(
min_batch_size=args.batch_size, max_batch_size=args.batch_size)
| 'Generate embeddings' >> beam.Map(
generate_embeddings, args.module_url, args.random_projection_matrix)
| 'Encode to tf example' >> beam.FlatMap(to_tf_example)
| 'Write to TFRecords files' >> beam.io.WriteToTFRecord(
file_path_prefix='{}/emb'.format(args.output_dir),
file_name_suffix='.tfrecords')
)
Génération d'une matrice de poids de projection aléatoire
La projection aléatoire est une technique simple mais puissante utilisée pour réduire la dimensionnalité d'un ensemble de points situés dans l'espace euclidien. Pour un contexte théorique, voir le lemme de Johnson-Lindenstrauss .
Réduire la dimensionnalité des intégrations avec la projection aléatoire signifie moins de temps nécessaire pour créer et interroger l'index ANN.
Dans ce didacticiel, nous utilisons la projection aléatoire gaussienne de la bibliothèque Scikit-learn .
def generate_random_projection_weights(original_dim, projected_dim):
random_projection_matrix = None
random_projection_matrix = gaussian_random_matrix(
n_components=projected_dim, n_features=original_dim).T
print("A Gaussian random weight matrix was creates with shape of {}".format(random_projection_matrix.shape))
print('Storing random projection matrix to disk...')
with open('random_projection_matrix', 'wb') as handle:
pickle.dump(random_projection_matrix,
handle, protocol=pickle.HIGHEST_PROTOCOL)
return random_projection_matrix
Définir les paramètres
Si vous souhaitez créer un index en utilisant l'espace d'incorporation d'origine sans projection aléatoire, définissez le paramètre projected_dim sur None . Notez que cela ralentira l’étape d’indexation pour les intégrations de grande dimension.
module_url = 'https://tfhub.dev/google/nnlm-en-dim128/2'
projected_dim = 64
Exécuter le pipeline
import tempfile
output_dir = tempfile.mkdtemp()
original_dim = hub.load(module_url)(['']).shape[1]
random_projection_matrix = None
if projected_dim:
random_projection_matrix = generate_random_projection_weights(
original_dim, projected_dim)
args = {
'job_name': 'hub2emb-{}'.format(datetime.utcnow().strftime('%y%m%d-%H%M%S')),
'runner': 'DirectRunner',
'batch_size': 1024,
'data_dir': 'corpus/*.txt',
'output_dir': output_dir,
'module_url': module_url,
'random_projection_matrix': random_projection_matrix,
}
print("Pipeline args are set.")
args
A Gaussian random weight matrix was creates with shape of (128, 64)
Storing random projection matrix to disk...
Pipeline args are set.
/tmpfs/src/tf_docs_env/lib/python3.6/site-packages/sklearn/utils/deprecation.py:86: FutureWarning: Function gaussian_random_matrix is deprecated; gaussian_random_matrix is deprecated in 0.22 and will be removed in version 0.24.
warnings.warn(msg, category=FutureWarning)
{'job_name': 'hub2emb-210107-125029',
'runner': 'DirectRunner',
'batch_size': 1024,
'data_dir': 'corpus/*.txt',
'output_dir': '/tmp/tmp0g361gzp',
'module_url': 'https://tfhub.dev/google/nnlm-en-dim128/2',
'random_projection_matrix': array([[-0.1349755 , -0.12082699, 0.07092581, ..., -0.02680793,
-0.0459312 , -0.20462361],
[-0.06197901, 0.01832142, 0.21362496, ..., 0.06641898,
0.14553738, -0.117217 ],
[ 0.03452009, 0.14239163, 0.01371371, ..., 0.10422342,
0.02966668, -0.07094185],
...,
[ 0.03384223, 0.05102025, 0.01941788, ..., -0.07500625,
0.09584965, -0.08593636],
[ 0.11010087, -0.10597793, 0.06668758, ..., -0.0518654 ,
-0.14681441, 0.08449293],
[ 0.26909502, -0.0291555 , 0.04305639, ..., -0.02295843,
0.1164921 , -0.04828371]])}
print("Running pipeline...")
%time run_hub2emb(args)
print("Pipeline is done.")
WARNING:apache_beam.runners.interactive.interactive_environment:Dependencies required for Interactive Beam PCollection visualization are not available, please use: `pip install apache-beam[interactive]` to install necessary dependencies to enable all data visualization features. Running pipeline... Warning:tensorflow:5 out of the last 5 calls to <function recreate_function.<locals>.restored_function_body at 0x7efcac3599d8> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for more details. Warning:tensorflow:5 out of the last 5 calls to <function recreate_function.<locals>.restored_function_body at 0x7efcac3599d8> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for more details. Warning:tensorflow:6 out of the last 6 calls to <function recreate_function.<locals>.restored_function_body at 0x7efcac475598> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for more details. Warning:tensorflow:6 out of the last 6 calls to <function recreate_function.<locals>.restored_function_body at 0x7efcac475598> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for more details. WARNING:apache_beam.io.tfrecordio:Couldn't find python-snappy so the implementation of _TFRecordUtil._masked_crc32c is not as fast as it could be. CPU times: user 9min 4s, sys: 10min 14s, total: 19min 19s Wall time: 2min 30s Pipeline is done.
ls {output_dir}
emb-00000-of-00001.tfrecords
Lisez quelques-unes des intégrations générées...
embed_file = os.path.join(output_dir, 'emb-00000-of-00001.tfrecords')
sample = 5
# Create a description of the features.
feature_description = {
'text': tf.io.FixedLenFeature([], tf.string),
'embedding': tf.io.FixedLenFeature([projected_dim], tf.float32)
}
def _parse_example(example):
# Parse the input `tf.Example` proto using the dictionary above.
return tf.io.parse_single_example(example, feature_description)
dataset = tf.data.TFRecordDataset(embed_file)
for record in dataset.take(sample).map(_parse_example):
print("{}: {}".format(record['text'].numpy().decode('utf-8'), record['embedding'].numpy()[:10]))
headline_text: [ 0.07743962 -0.10065071 -0.03604915 0.03902601 0.02538098 -0.01991337 -0.11972483 0.03102058 0.16498186 -0.04299153] aba decides against community broadcasting licence: [ 0.02420221 -0.07736929 0.05655728 -0.18739551 0.11344934 0.12652674 -0.18189304 0.00422473 0.13149698 0.01910412] act fire witnesses must be aware of defamation: [-0.17413895 -0.05418579 0.07769868 0.05096476 0.08622053 0.33112594 0.04067763 0.00448784 0.15882017 0.33829722] a g calls for infrastructure protection summit: [ 0.16939437 -0.18585566 -0.14201084 -0.21779229 -0.1374832 0.14933842 -0.19583155 0.12921487 0.09811856 0.099967 ] air nz staff in aust strike for pay rise: [ 0.0230642 -0.03269081 0.18271443 0.23761444 -0.01575144 0.06109515 -0.01963143 -0.05211507 0.06050447 -0.20023327]
3. Créez l'index ANN pour les intégrations
ANNOY (Approximate Nearest Neighbours Oh Yeah) est une bibliothèque C++ avec des liaisons Python pour rechercher des points dans l'espace proches d'un point de requête donné. Il crée également de grandes structures de données basées sur des fichiers en lecture seule qui sont mappées en mémoire. Il est construit et utilisé par Spotify pour les recommandations musicales. Si vous êtes intéressé, vous pouvez jouer avec d'autres alternatives à ANNOY telles que NGT , FAISS , etc.
def build_index(embedding_files_pattern, index_filename, vector_length,
metric='angular', num_trees=100):
'''Builds an ANNOY index'''
annoy_index = annoy.AnnoyIndex(vector_length, metric=metric)
# Mapping between the item and its identifier in the index
mapping = {}
embed_files = tf.io.gfile.glob(embedding_files_pattern)
num_files = len(embed_files)
print('Found {} embedding file(s).'.format(num_files))
item_counter = 0
for i, embed_file in enumerate(embed_files):
print('Loading embeddings in file {} of {}...'.format(i+1, num_files))
dataset = tf.data.TFRecordDataset(embed_file)
for record in dataset.map(_parse_example):
text = record['text'].numpy().decode("utf-8")
embedding = record['embedding'].numpy()
mapping[item_counter] = text
annoy_index.add_item(item_counter, embedding)
item_counter += 1
if item_counter % 100000 == 0:
print('{} items loaded to the index'.format(item_counter))
print('A total of {} items added to the index'.format(item_counter))
print('Building the index with {} trees...'.format(num_trees))
annoy_index.build(n_trees=num_trees)
print('Index is successfully built.')
print('Saving index to disk...')
annoy_index.save(index_filename)
print('Index is saved to disk.')
print("Index file size: {} GB".format(
round(os.path.getsize(index_filename) / float(1024 ** 3), 2)))
annoy_index.unload()
print('Saving mapping to disk...')
with open(index_filename + '.mapping', 'wb') as handle:
pickle.dump(mapping, handle, protocol=pickle.HIGHEST_PROTOCOL)
print('Mapping is saved to disk.')
print("Mapping file size: {} MB".format(
round(os.path.getsize(index_filename + '.mapping') / float(1024 ** 2), 2)))
embedding_files = "{}/emb-*.tfrecords".format(output_dir)
embedding_dimension = projected_dim
index_filename = "index"
!rm {index_filename}
!rm {index_filename}.mapping
%time build_index(embedding_files, index_filename, embedding_dimension)
rm: cannot remove 'index': No such file or directory rm: cannot remove 'index.mapping': No such file or directory Found 1 embedding file(s). Loading embeddings in file 1 of 1... 100000 items loaded to the index 200000 items loaded to the index 300000 items loaded to the index 400000 items loaded to the index 500000 items loaded to the index 600000 items loaded to the index 700000 items loaded to the index 800000 items loaded to the index 900000 items loaded to the index 1000000 items loaded to the index 1100000 items loaded to the index A total of 1103664 items added to the index Building the index with 100 trees... Index is successfully built. Saving index to disk... Index is saved to disk. Index file size: 1.61 GB Saving mapping to disk... Mapping is saved to disk. Mapping file size: 50.61 MB CPU times: user 9min 54s, sys: 53.9 s, total: 10min 48s Wall time: 5min 5s
ls
corpus random_projection_matrix index raw.tsv index.mapping tf2_semantic_approximate_nearest_neighbors.ipynb
4. Utilisez l'index pour la correspondance de similarité
Nous pouvons désormais utiliser l'index ANN pour rechercher des titres d'actualité sémantiquement proches d'une requête d'entrée.
Charger l'index et les fichiers de mappage
index = annoy.AnnoyIndex(embedding_dimension)
index.load(index_filename, prefault=True)
print('Annoy index is loaded.')
with open(index_filename + '.mapping', 'rb') as handle:
mapping = pickle.load(handle)
print('Mapping file is loaded.')
Annoy index is loaded. /tmpfs/src/tf_docs_env/lib/python3.6/site-packages/ipykernel_launcher.py:1: FutureWarning: The default argument for metric will be removed in future version of Annoy. Please pass metric='angular' explicitly. """Entry point for launching an IPython kernel. Mapping file is loaded.
Méthode de correspondance de similarité
def find_similar_items(embedding, num_matches=5):
'''Finds similar items to a given embedding in the ANN index'''
ids = index.get_nns_by_vector(
embedding, num_matches, search_k=-1, include_distances=False)
items = [mapping[i] for i in ids]
return items
Extraire l'intégration d'une requête donnée
# Load the TF-Hub module
print("Loading the TF-Hub module...")
%time embed_fn = hub.load(module_url)
print("TF-Hub module is loaded.")
random_projection_matrix = None
if os.path.exists('random_projection_matrix'):
print("Loading random projection matrix...")
with open('random_projection_matrix', 'rb') as handle:
random_projection_matrix = pickle.load(handle)
print('random projection matrix is loaded.')
def extract_embeddings(query):
'''Generates the embedding for the query'''
query_embedding = embed_fn([query])[0].numpy()
if random_projection_matrix is not None:
query_embedding = query_embedding.dot(random_projection_matrix)
return query_embedding
Loading the TF-Hub module... CPU times: user 757 ms, sys: 619 ms, total: 1.38 s Wall time: 1.37 s TF-Hub module is loaded. Loading random projection matrix... random projection matrix is loaded.
extract_embeddings("Hello Machine Learning!")[:10]
array([ 0.12164804, 0.0162079 , -0.15466002, -0.14580576, 0.03926325,
-0.10124508, -0.1333948 , 0.0515029 , -0.14688903, -0.09971556])
Entrez une requête pour trouver les articles les plus similaires
query = "confronting global challenges"
print("Generating embedding for the query...")
%time query_embedding = extract_embeddings(query)
print("")
print("Finding relevant items in the index...")
%time items = find_similar_items(query_embedding, 10)
print("")
print("Results:")
print("=========")
for item in items:
print(item)
Generating embedding for the query... CPU times: user 5.18 ms, sys: 596 µs, total: 5.77 ms Wall time: 2.19 ms Finding relevant items in the index... CPU times: user 555 µs, sys: 327 µs, total: 882 µs Wall time: 601 µs Results: ========= confronting global challenges emerging nations to help struggling global economy g7 warns of increasing global economic crisis world struggling to cope with global terrorism companies health to struggle amid global crisis external risks biggest threat to economy asian giants unite to tackle global crisis g7 ministers warn of slowing global growth experts to discuss global warming threat scientists warn of growing natural disasters
Vous voulez en savoir plus ?
Vous pouvez en savoir plus sur TensorFlow sur tensorflow.org et consulter la documentation de l'API TF-Hub sur tensorflow.org/hub . Recherchez les modules TensorFlow Hub disponibles sur tfhub.dev, y compris davantage de modules d'intégration de texte et de modules vectoriels de fonctionnalités d'image.
Consultez également le cours intensif d'apprentissage automatique , qui constitue l'introduction pratique et rapide de Google à l'apprentissage automatique.