tensorflow 1 Session.run使用通用语句编码器嵌入语句花费了太多时间

Question

与烧瓶REST 一起使用张量流

如何减少session.run 的时间？

我在REST中使用tf 1/2，而不是在服务器上使用它，而不是服务它。

我试过tensorflow 1和2。

tensorflow 1花了太多时间。

tensorflow 2甚至没有返回文本的向量。

in tensorflow 1初始化需要2-4秒，session.run需要5-8秒.随着我不断地满足要求，时间也在增加。

tensorflow 1

import tensorflow.compat.v1 as tfo
import tensorflow_hub as hub
tfo.disable_eager_execution()

module_url = "https://tfhub.dev/google/universal-sentence-encoder-qa/3"
# Import the Universal Sentence Encoder's TF Hub module
embed = hub.Module(module_url)

def convert_text_to_vector(text):
    # Compute a representation for each message, showing various lengths supported.
    try:
        #text = "qwerty" or ["qwerty"]
        if isinstance(text, str):
            text = [text]
        with tfo.Session() as session:
            t_time = time.time()
            session.run([tfo.global_variables_initializer(), tfo.tables_initializer()])
            m_time = time.time()
            message_embeddings = session.run(embed(text))
            vector_array = message_embeddings.tolist()[0]
        return vector_array
    except Exception as err:
        raise Exception(str(err))

tensorflow 2

它被困在vector_array = embedding_fn(text)

import tensorflow as tf
import tensorflow_hub as hub
module_url = "https://tfhub.dev/google/universal-sentence-encoder-qa/3"
embedding_fn = hub.load(module_url)

@tf.function
def convert_text_to_vector(text):
    try:
        #text = ["qwerty"]
        vector_array = embedding_fn(text)
        return vector_array
    except Exception as err:
        raise Exception(str(err))

Answer 1

对于tensorflow 2版本，我做了很少的修正。基本上，我遵循了您提供的通用句子编码器中的示例。

import tensorflow as tf
import tensorflow_hub as hub
import numpy as np
module_url = "https://tfhub.dev/google/universal-sentence-encoder-qa/3"
embedding_fn = hub.load(module_url)

@tf.function
def convert_text_to_vector(text):
  try:
      vector_array = embedding_fn.signatures['question_encoder'](
          tf.constant(text))
      return vector_array['outputs']
  except Exception as err:
      raise Exception(str(err))

### run the function
vector = convert_text_to_vector(['is this helpful ?'])
print(vector.shape())

Answer 2

from flask import Flask
from flask_restplus import Api, Resource
from werkzeug.utils import cached_property

import tensorflow as tf
import tensorflow_hub as hub
module_url = "https://tfhub.dev/google/universal-sentence-encoder-qa/3"
embedding_fn = hub.load(module_url)


app = Flask(__name__)

@app.route('/embedding', methods=['POST'])
def entry_point(args):
    if args.get("text"):
        text_term = args.get("text")
        if isinstance(text_term, str):
            text_term = [text_term]
        vectors = convert_text_to_vector(text_term)
    return vectors



@tf.function
def convert_text_to_vector(text):
    try:
        vector_array = embedding_fn.signatures['question_encoder'](tf.constant(text))
        return vector_array['outputs']
    except Exception as err:
        raise Exception(str(err))


if __name__ == '__main__':
    app.run(host='0.0.0.0', port=5000, debug=True)

"""
 ----- Requirements.txt ----
flask-restplus==0.13.0
Flask==1.1.1
Werkzeug==0.15.5
tensorboard==2.2.2
tensorboard-plugin-wit==1.6.0.post3
tensorflow==2.2.0
tensorflow-estimator==2.2.0
tensorflow-hub==0.8.0
tensorflow-text==2.2.1
"""