“这个 TensorFlow 二进制文件使用 Intel(R) MKL-DNN 进行了优化......在非 MKL-DNN 操作中启用它们”我能做什么？

Question

我已经使用 conda 安装了 tensorflow。当我测试它时，我得到了这个性能：

import tensorflow as tf
print(tf.__version__)

a = tf.constant([1, 2])
b = tf.constant([3, 4])
print(a + b)

输出

C:\Users\NHoracio\Miniconda3\envs\tesis\python.exe "D:/Dropbox/Codigos/Python3/TensorFlow test/test.py"
2.0.0
2019-12-25 12:13:37.613118: I tensorflow/core/platform/cpu_feature_guard.cc:145] This TensorFlow binary is optimized
with Intel(R) MKL-DNN to use the following CPU instructions in performance critical operations:
AVX AVX2

To enable them in non-MKL-DNN operations, rebuild TensorFlow with the appropriate compiler flags.
2019-12-25 12:13:37.621213: I tensorflow/core/common_runtime/process_util.cc:115] Creating new thread pool with
default inter op setting: 8. Tune using inter_op_parallelism_threads for best performance.

tf.Tensor([4 6], shape=(2,), dtype=int32)

我知道它可以继续以这种方式工作，但我不确定我是否理解该信息。 我想知道我是否可以提高 CPU 的性能，或者我必须为该消息做什么？

配置： 视窗 10 64b。 英特尔 i5 8250。 广发MX130 Python 3.7.4 PyCharm 2019.3.1 康达 4.8.0 TensorFlow 2.0.0

Answer 1

我今天早上更新了 tensorflow，并且那个特定的错误消息不再存在。这是我正在运行的代码：

# smorgasbord of import statements scraped from random Internet sites
from tensorflow.keras import Sequential
from tensorflow.keras.layers import Dense
from tensorflow.keras.constraints import unit_norm
from tensorflow.keras.utils import to_categorical
import numpy as np
from sklearn.preprocessing import LabelEncoder
from sklearn.model_selection import StratifiedKFold
import pandas as pd

# import os
# print(os.getcwd())

# establishing a single random seed for reproducibility of results
seed = 7
np.random.seed(seed)

# load Bezdek Iris dataset
dataframe = pd.read_csv("BezdekIris.csv", header=None)

# split into input (X) and output (Y) variables
dataset = dataframe.values
X = dataset[:, 0:3].astype(float)
Y = dataset[:, 4]

# create dummy indicator output variables to use in Keras
encoder = LabelEncoder()
encoder.fit(Y)
encoded_Y = encoder.transform(Y)

# convert integers to dummy variables (i.e. one hot encoded)
dummy_y = to_categorical(encoded_Y)

# cross-validation train/test methodology scraped from the Internet
kfold = StratifiedKFold(n_splits=10, shuffle=True, random_state=seed)
cvscores = []
for train, test in kfold.split(X, Y):
    # create model (no idea what I am doing here at all...)
    model = Sequential()
    model.add(Dense(12, input_dim=3, activation='relu',
                    kernel_constraint=unit_norm()))
    model.add(Dense(3, activation='relu', kernel_constraint=unit_norm()))
    model.add(Dense(3, activation='softmax', kernel_constraint=unit_norm()))

    # compile and evaluate model (no idea why these options work/if they do)
    model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])
    model.fit(X[train], dummy_y[train], epochs=150, batch_size=10, verbose=0)
    scores = model.evaluate(X[test], dummy_y[test], verbose=0)

    # display each validation fold's results
    print("%s:  %.2f%%" % (model.metrics_names[1], scores[1] * 100))
    cvscores.append(scores[1] * 100)

# display overall results
print("%.2f%% (+/-%.2f%%)" % (np.mean(cvscores), np.std(cvscores)))