1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
| from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import sys
import os
import logging
import json
import math
from six.moves import urllib
import tensorflow as tf
# 读取文件
reader = tf.TextLineReader(skip_header_lines = 0)
# 文件列表
train_input_files = ["/root/xpguo/wnd/1.txt", "/root/xpguo/wnd/2.txt"]
input_file_list = []
for input_file in train_input_files:
if len(input_file) > 0:
input_file_list.append(tf.train.match_filenames_once(input_file))
filename_queue = tf.train.string_input_producer(
tf.concat(input_file_list, axis = 0),
num_epochs = 10, # strings are repeated num_epochs
shuffle = True, # strings are randomly shuffled within each epoch
capacity = 512)
batch_size = 3
(_, records) = reader.read_up_to(filename_queue, num_records = batch_size)
samples = tf.decode_csv(records, record_defaults = column_defaults, field_delim = ',')
label = tf.cast(samples[self.column_dict["label"]], dtype = tf.int32)
feature_dict = {}
for (key, value) in self.column_dict.items():
if key == "label" or value < 0 or value >= len(samples):
continue
if key in ["user_features", "ads_features"]:
feature_dict[key] = tf.string_split(samples[value], delimiter = ';')
if key in ["user_weights", "ads_weights"]:
feature_dict[key] = self.string_to_number(
tf.string_split(samples[value], delimiter = ';'),
dtype = tf.float32)
return feature_dict, label
|