tensorflow71 《深度学习原理与TensorFlow实战》05 RNN能说会道 02语言模型
2024-05-06 12:06:30
01 基本信息
#《深度学习原理与TensorFlow实战》05 RNN能说会道
# 书源码地址:https://github.com/DeepVisionTeam/TensorFlowBook.git
# 视频讲座地址：http://edu.csdn.net/course/detail/5222
# win10 Tensorflow1.2.0 python3.6.1
# CUDA v8.0 cudnn-8.0-windows10-x64-v5.1
# 本地代码位置：D:\git\DeepLearning\TensorFlowBook\rnn\language_model
# https://github.com/DeepVisionTeam/TensorFlowBook/blob/master/rnn/language_model/tclm.py
# https://github.com/DeepVisionTeam/TensorFlowBook/blob/master/rnn/language_model/tclm_inference.py
# https://github.com/DeepVisionTeam/TensorFlowBook/blob/master/rnn/language_model/tclm_reader.py
# https://github.com/DeepVisionTeam/TensorFlowBook/blob/master/rnn/language_model/tensorflow_code_stat.py
# https://github.com/DeepVisionTeam/TensorFlowBook/blob/master/rnn/language_model/test_tclm_reader.py
# 在tensorflow1.2.0版本上，需要稍作修改02 tclm.py
# 注意需要制定正确的参数
# C:\Python36\python.exe D:/git/DeepLearning/TensorFlowBook/rnn/language_model/tclm.py --data_path=..\..\..\tensorflow --model=small --save_path=.\savepath --use_fp16=True
#
# Copyright 2015 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# =============================================================================="""Example / benchmark for building a PTB LSTM model.Trains the model described in:
(Zaremba, et. al.) Recurrent Neural Network Regularization
http://arxiv.org/abs/1409.2329There are 3 supported model configurations:
===========================================
| config | epochs | train | valid | test
===========================================
| small | 13 | 37.99 | 121.39 | 115.91
| medium | 39 | 48.45 | 86.16 | 82.07
| large | 55 | 37.87 | 82.62 | 78.29
The exact results may vary depending on the random initialization.The hyperparameters used in the model:
- init_scale - the initial scale of the weights
- learning_rate - the initial value of the learning rate
- max_grad_norm - the maximum permissible norm of the gradient
- num_layers - the number of LSTM layers
- num_steps - the number of unrolled steps of LSTM
- hidden_size - the number of LSTM units
- max_epoch - the number of epochs trained with the initial learning rate
- max_max_epoch - the total number of epochs for training
- keep_prob - the probability of keeping weights in the dropout layer
- lr_decay - the decay of the learning rate for each epoch after "max_epoch"
- batch_size - the batch sizeThe data required for this example is in the data/ dir of the
PTB dataset from Tomas Mikolov's webpage:$ wget http://www.fit.vutbr.cz/~imikolov/rnnlm/simple-examples.tgz
$ tar xvf simple-examples.tgzTo run:$ python ptb_word_lm.py --data_path=simple-examples/data/"""
from __future__ import absolute_import
from __future__ import division
from __future__ import print_functionimport timeimport numpy as np
import tclm_reader
import tensorflow as tf
#from tensorflow.contrib.rnn.python.ops import core_rnn_cell
#from tensorflow.contrib.rnn.python.ops import lstm_ops
from tensorflow.contrib import rnnflags = tf.flags
logging = tf.loggingflags.DEFINE_string("model", "small","A type of model. Possible options are: small, medium, large.")
flags.DEFINE_string("data_path", None,"Where the training/test data is stored.")
flags.DEFINE_string("save_path", None, "Model output directory.")
flags.DEFINE_bool("use_fp16", False,"Train using 16-bit floats instead of 32bit floats")FLAGS = flags.FLAGSdef data_type():return tf.float16 if FLAGS.use_fp16 else tf.float32class PTBInput(object):"""The input data."""def __init__(self, config, data, name=None):self.batch_size = batch_size = config.batch_sizeself.num_steps = num_steps = config.num_stepsself.epoch_size = ((len(data) // batch_size) - 1) // num_stepsself.input_data, self.targets = tclm_reader.tensorflow_code_producer(data, batch_size, num_steps, name=name)class PTBModel(object):"""The PTB model."""def __init__(self, is_training, config, input_):self._input = input_batch_size = input_.batch_sizenum_steps = input_.num_stepssize = config.hidden_sizevocab_size = config.vocab_size# Slightly better results can be obtained with forget gate biases# initialized to 1 but the hyperparameters of the model would need to be# different than reported in the paper.lstm_cell = rnn.BasicLSTMCell(num_units=size, state_is_tuple=True)#lstm_cell = core_rnn_cell.BasicLSTMCell(num_units=size, state_is_tuple=True)if is_training and config.keep_prob < 1:lstm_cell = tf.contrib.rnn.DropoutWrapper(lstm_cell, output_keep_prob=config.keep_prob)cell = rnn.MultiRNNCell([lstm_cell] * config.num_layers, state_is_tuple=True)self._initial_state = cell.zero_state(batch_size, data_type())with tf.device("/cpu:0"):embedding = tf.get_variable("embedding", [vocab_size, size], dtype=data_type())inputs = tf.nn.embedding_lookup(embedding, input_.input_data)if is_training and config.keep_prob < 1:inputs = tf.nn.dropout(inputs, config.keep_prob)outputs = []state = self._initial_statewith tf.variable_scope("RNN"):for time_step in range(num_steps):if time_step > 0: tf.get_variable_scope().reuse_variables()(cell_output, state) = cell(inputs[:, time_step, :], state)outputs.append(cell_output)output = tf.reshape(tf.concat(outputs, 1), [-1, size])softmax_w = tf.get_variable("softmax_w", [size, vocab_size], dtype=data_type())softmax_b = tf.get_variable("softmax_b", [vocab_size],dtype=data_type())logits = tf.matmul(output, softmax_w) + softmax_bloss = tf.contrib.legacy_seq2seq.sequence_loss_by_example([logits],[tf.reshape(input_.targets, [-1])],[tf.ones([batch_size * num_steps], dtype=data_type())])self._cost = cost = tf.reduce_sum(loss) / batch_sizeself._final_state = stateif not is_training:returnself._lr = tf.Variable(0.0, trainable=False)tvars = tf.trainable_variables()grads, _ = tf.clip_by_global_norm(tf.gradients(cost, tvars),config.max_grad_norm)optimizer = tf.train.GradientDescentOptimizer(self._lr)self._train_op = optimizer.apply_gradients(zip(grads, tvars),global_step=tf.contrib.framework.get_or_create_global_step())self._new_lr = tf.placeholder(tf.float32, shape=[], name="new_learning_rate")self._lr_update = tf.assign(self._lr, self._new_lr)def assign_lr(self, session, lr_value):session.run(self._lr_update, feed_dict={self._new_lr: lr_value})@propertydef input(self):return self._input@propertydef initial_state(self):return self._initial_state@propertydef cost(self):return self._cost@propertydef final_state(self):return self._final_state@propertydef lr(self):return self._lr@propertydef train_op(self):return self._train_opclass SmallConfig(object):"""Small config."""init_scale = 0.1learning_rate = 1.0max_grad_norm = 5num_layers = 2num_steps = 20hidden_size = 200max_epoch = 4max_max_epoch = 13keep_prob = 1.0lr_decay = 0.5batch_size = 20vocab_size = tclm_reader.VOCAB_SIZEclass MediumConfig(object):"""Medium config."""init_scale = 0.05learning_rate = 1.0max_grad_norm = 5num_layers = 2num_steps = 35hidden_size = 650max_epoch = 6max_max_epoch = 39keep_prob = 0.5lr_decay = 0.8batch_size = 20vocab_size = tclm_reader.VOCAB_SIZEclass LargeConfig(object):"""Large config."""init_scale = 0.04learning_rate = 1.0max_grad_norm = 10num_layers = 2num_steps = 35hidden_size = 1500max_epoch = 14max_max_epoch = 55keep_prob = 0.35lr_decay = 1 / 1.15batch_size = 20vocab_size = tclm_reader.VOCAB_SIZEclass TestConfig(object):"""Tiny config, for testing."""init_scale = 0.1learning_rate = 1.0max_grad_norm = 1num_layers = 1num_steps = 2hidden_size = 2max_epoch = 1max_max_epoch = 1keep_prob = 1.0lr_decay = 0.5batch_size = 20vocab_size = tclm_reader.VOCAB_SIZEdef run_epoch(session, model, eval_op=None, verbose=False):"""Runs the model on the given data."""start_time = time.time()costs = 0.0iters = 0state = session.run(model.initial_state)fetches = {"cost": model.cost,"final_state": model.final_state,}if eval_op is not None:fetches["eval_op"] = eval_opfor step in range(model.input.epoch_size):feed_dict = {}for i, (c, h) in enumerate(model.initial_state):feed_dict[c] = state[i].cfeed_dict[h] = state[i].hvals = session.run(fetches, feed_dict)cost = vals["cost"]state = vals["final_state"]costs += costiters += model.input.num_stepsif verbose and step % (model.input.epoch_size // 10) == 10:print("%.3f perplexity: %.3f speed: %.0f wps" %(step * 1.0 / model.input.epoch_size, np.exp(costs / iters),iters * model.input.batch_size / (time.time() - start_time)))return np.exp(costs / iters)def get_config():if FLAGS.model == "small":return SmallConfig()elif FLAGS.model == "medium":return MediumConfig()elif FLAGS.model == "large":return LargeConfig()elif FLAGS.model == "test":return TestConfig()else:raise ValueError("Invalid model: %s", FLAGS.model)def main(_):if not FLAGS.data_path:raise ValueError("Must set --data_path to tensorflow source code directory")raw_data = tclm_reader.tensorflow_code_data(FLAGS.data_path)train_data, valid_data, test_data = raw_dataconfig = get_config()eval_config = get_config()eval_config.batch_size = 1eval_config.num_steps = 1with tf.Graph().as_default():initializer = tf.random_uniform_initializer(-config.init_scale,config.init_scale)with tf.name_scope("Train"):train_input = PTBInput(config=config, data=train_data,name="TrainInput")with tf.variable_scope("Model", reuse=None,initializer=initializer):m = PTBModel(is_training=True, config=config,input_=train_input)tf.summary.scalar("Training Loss", m.cost)tf.summary.scalar("Learning Rate", m.lr)with tf.name_scope("Valid"):valid_input = PTBInput(config=config, data=valid_data,name="ValidInput")with tf.variable_scope("Model", reuse=True,initializer=initializer):mvalid = PTBModel(is_training=False, config=config,input_=valid_input)tf.summary.scalar("Validation Loss", mvalid.cost)with tf.name_scope("Test"):test_input = PTBInput(config=eval_config, data=test_data,name="TestInput")with tf.variable_scope("Model", reuse=True,initializer=initializer):mtest = PTBModel(is_training=False, config=eval_config,input_=test_input)sv = tf.train.Supervisor(logdir=FLAGS.save_path)with sv.managed_session() as session:for i in range(config.max_max_epoch):lr_decay = config.lr_decay ** max(i + 1 - config.max_epoch, 0.0)m.assign_lr(session, config.learning_rate * lr_decay)print("Epoch: %d Learning rate: %.3f" % (i + 1, session.run(m.lr)))train_perplexity = run_epoch(session, m, eval_op=m.train_op,verbose=True)print("Epoch: %d Train Perplexity: %.3f" % (i + 1, train_perplexity))valid_perplexity = run_epoch(session, mvalid)print("Epoch: %d Valid Perplexity: %.3f" % (i + 1, valid_perplexity))test_perplexity = run_epoch(session, mtest)print("Test Perplexity: %.3f" % test_perplexity)if FLAGS.save_path:print("Saving model to %s." % FLAGS.save_path)sv.saver.save(session, FLAGS.save_path,global_step=sv.global_step)if __name__ == "__main__":tf.app.run()
'''
...
Epoch: 1 Learning rate: 1.000
0.000 perplexity: 2.667 speed: 98 wps
...
'''03 tclm_inference.py
# 注意制定参数 --save_path=./tc.lm
from __future__ import absolute_import
from __future__ import division
from __future__ import print_functionimport tclm_reader
import tensorflow as tf
from tclm import *def main(_):# $ python tclm_inference.py --save_path=./tc.lm/if not FLAGS.save_path:raise ValueError("Must set --save_path to language model directory")test_data = [tclm_reader.BOF] * 500config = get_config()eval_config = get_config()eval_config.batch_size = 1eval_config.num_steps = 1with tf.Graph().as_default():initializer = tf.random_uniform_initializer(-config.init_scale,config.init_scale)with tf.name_scope("Train"):test_input = PTBInput(config=config, data=test_data,name="TestInput")with tf.variable_scope("Model", reuse=None,initializer=initializer):mtest = PTBModel(is_training=True, config=eval_config,input_=test_input)tf.summary.scalar("Training Loss", mtest.cost)tf.summary.scalar("Learning Rate", mtest.lr)sv = tf.train.Supervisor(logdir=FLAGS.save_path)with sv.managed_session() as session:ckpt = tf.train.get_checkpoint_state(FLAGS.save_path)if ckpt and ckpt.model_checkpoint_path:sv.saver.restore(session, ckpt.model_checkpoint_path)test_perplexity = run_epoch(session, mtest)print("Test Perplexity: %.3f" % test_perplexity)if __name__ == "__main__":tf.app.run()04 tclm_reader.py
# Copyright 2015 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
from __future__ import absolute_import
from __future__ import division
from __future__ import print_functionimport osimport tensorflow as tfMAX_CHAR = 256
BOF = MAX_CHAR + 1
EOF = MAX_CHAR + 2
VOCAB_SIZE = MAX_CHAR + 3def get_file_suffix(file):index = file.rfind('.')return file[index + 1:] if index >= 0 else ''def read_source_code_data(code_files):data = []for code_file in code_files:file_r = open(code_file, 'r', encoding='UTF-8')curr_data = []curr_data.append(BOF)for dataline in file_r:for c in dataline:curr_data.append(ord(c))curr_data.append(EOF)data.extend(curr_data)file_r.close()return datadef tensorflow_code_data(data_path=None):# find all python source codetensorflow_code_files = []for root, dirs, files in os.walk(data_path):for file in files:code_file = os.path.join(root, file)file_suffix = get_file_suffix(code_file)if file_suffix == 'py':tensorflow_code_files.append(code_file)train_code_file_count = int(len(tensorflow_code_files) * 0.5)valid_code_file_count = int(len(tensorflow_code_files) * 0.3)# test_code_file_count = int(len(tensorflow_code_files) * 0.2)train_data = read_source_code_data(tensorflow_code_files[0: train_code_file_count])valid_data = read_source_code_data(tensorflow_code_files[train_code_file_count: train_code_file_count + valid_code_file_count])test_data = read_source_code_data(tensorflow_code_files[train_code_file_count + valid_code_file_count:])return train_data, valid_data, test_datadef tensorflow_code_producer(raw_data, batch_size, num_steps, name=None):"""Iterate on the raw PTB data.This chunks up raw_data into batches of examples and returns Tensors thatare drawn from these batches.Args:raw_data: one of the raw data outputs from ptb_raw_data.batch_size: int, the batch size.num_steps: int, the number of unrolls.name: the name of this operation (optional).Returns:A pair of Tensors, each shaped [batch_size, num_steps].The second element of the tuple is the same data time-shifted to the right by one.Raises:tf.errors.InvalidArgumentError: if batch_size or num_steps are too high."""with tf.name_scope(name, "TensorflowCodeProducer",[raw_data, batch_size, num_steps]):raw_data = tf.convert_to_tensor(raw_data, name="raw_data",dtype=tf.int32)data_len = tf.size(raw_data)batch_len = data_len // batch_sizedata = tf.reshape(raw_data[0: batch_size * batch_len],[batch_size, batch_len])epoch_size = (batch_len - 1) // num_stepsassertion = tf.assert_positive(epoch_size,message="epoch_size == 0, decrease batch_size or num_steps")with tf.control_dependencies([assertion]):epoch_size = tf.identity(epoch_size, name="epoch_size")i = tf.train.range_input_producer(epoch_size, shuffle=False).dequeue()x = tf.slice(data, [0, i * num_steps], [batch_size, num_steps])x.set_shape([batch_size, num_steps])y = tf.slice(data, [0, i * num_steps + 1], [batch_size, num_steps])y.set_shape([batch_size, num_steps])return x, y05 tensorflow_code_stat.py
“`
import os
source code file count result:
cc 1309
md 1242
py 1235
h 734
html 107
ts 95
sh 81
proto 57
json 49
java 34
def get_file_suffix(file):
index = file.rfind(‘.’)
return file[index + 1:] if index >= 0 else ”
char frequency in py source code
def tensorflow_code_stat(data_path):
tensorflow_code_files = []
for root, dirs, files in os.walk(data_path):
for file in files:
code_file = os.path.join(root, file)
file_suffix = get_file_suffix(code_file)
if file_suffix == ‘py’:
tensorflow_code_files.append(code_file)
dict_c = {}
for code_file in tensorflow_code_files:file_r = open(code_file, 'r', encoding='UTF-8')for data in file_r:for c in data:if c in dict_c:dict_c[c] += 1else:dict_c[c] = 1file_r.close()for c, c_freq in dict_c.items():print(c + '\t' + str(ord(c)) + '\t' + str(c_freq))
if name == “main“:
# 需要注意这里的目录i相对位置，或者用绝对路径
tensorflow_code_stat(‘../../../tensorflow/tensorflow/’)
”’
35 278507
32 22269644
C 67 195582
o 111 4482778
p 112 2767253
y 121 865716
r 114 4523279
i 105 4313868
g 103 1117762
h 104 1593713
t 116 6038934
2 50 356485
0 48 468573
1 49 448245
5 53 86463
T 84 453378
e 101 8656155
n 110 4455952
s 115 5930738
F 70 171095
l 108 3345283
w 119 594435
A 65 334059
u 117 2178558
. 46 2103344
R 82 262307
v 118 686076
d 100 2250758
- 10 3013538 L 76 184014 c 99 1968624 a 97 5034509 , 44 1604448 V 86 85934 ( 40 1317136 ” 34 811601 ) 41 1314840 ; 59 14841 m 109 1774492 f 102 1967961 x 120 463544 Y 89 39030 b 98 868729
-
58 709226
/ 47 116930
I 73 267985
E 69 342527
N 78 310623
S 83 296461
- 45 162396
U 85 79575
q 113 209850
B 66 73496
W 87 60815
H 72 55985
O 79 195138
D 68 190840
K 75 29106
= 61 1550312
_ 95 2931524
* 42 78463
z 122 196614
’ 39 379320
k 107 322629
7 55 39329
G 71 81447
[ 91 417829
] 93 417002
3 51 203258
+ 43 48685
4 52 114384
% 37 44444
@ 64 30535
P 80 104891
X 88 13313
6 54 91173
9 57 34282
8 56 49653
M 77 100858
{ 123 70433
} 125 7036762 21221
< 60 14353
j 106 61869
| 124 4892
\ 92 69468
J 74 3632
! 33 7608
Z 90 5844
$ 36 1699
` 96 408658
^ 94 4767
~ 126 1074
? 63 2574
Q 81 11146
9 1238
& 38 1267
Т 1058 12
е 1077 17
с 1089 27
т 1090 12
ф 1092 40
ы 1099 40
в 1074 40
你 20320 10
好 22909 10
怎 24590 10
么 20040 10
样 26679 10
а 1072 45
65279 5
Н 1053 5
к 1082 20
и 1080 20
д 1076 15
р 1088 20
ж 1078 5
ј 1112 15
н 1085 5
п 1087 5
о 1086 5
м 1084 5
爱 29233 10
要 35201 5
依 20381 5
法 27861 5
治 27835 5
国 22269 5
是 26159 10
赞 36190 5
美 32654 5
那 37027 5
些 20123 5
谁 35841 5
公 20844 5
义 20041 5
的 30340 5
和 21644 5
惩 24809 5
罚 32602 5
恶 24694 5
人 20154 5
。 12290 5
韩 38889 5
非 38750 5
у 1091 10
г 1075 10
з 1079 10
ñ 241 10
ë 235 10
â 226 10
ô 244 10
à 224 10
æ 230 10
ø 248 10
⚡ 9889 10
tensorflow71 《深度学习原理与TensorFlow实战》05 RNN能说会道 02语言模型相关推荐
- tensorflow70 《深度学习原理与TensorFlow实战》05 RNN能说会道 01 正弦序列预测
01 基本环境 #<深度学习原理与TensorFlow实战>05 RNN能说会道 # 书源码地址:https://github.com/DeepVisionTeam/TensorFlowB ...
- tensorflow72 《深度学习原理与TensorFlow实战》05 RNN能说会道 03 对话机器人(chatbot)
01 基本环境 #<深度学习原理与TensorFlow实战>05 RNN能说会道 # 书源码地址:https://github.com/DeepVisionTeam/TensorFlowB ...
- tensorflow63 《深度学习原理与TensorFlow实战》03 Hello TensorFlow
00 基本信息 <深度学习原理与TensorFlow实战>书中涉及到的代码主要来源于: A:Tensorflow/TensorflowModel/TFLean的样例, B:https:// ...
- tensorflow67 《深度学习原理与TensorFlow实战》04 CNN看懂世界 04深度残差网络
00 环境 #<深度学习原理与TensorFlow实战>04 CNN看懂世界 # 书源码地址:https://github.com/DeepVisionTeam/TensorFlowBoo ...
- python神经网络原理pdf_《深度学习原理与 TensorFlow实践》高清完整PDF版 下载
1.封面介绍 2.出版时间 2019年7月 3.推荐理由 本书介绍了深度学习原理与TensorFlow实践.着重讲述了当前学术界和工业界的深度学习核心知识:机器学习概论.神经网络.深度学习.着重讲述了 ...
- 深度学习原理与TensorFlow实践
深度学习原理与TensorFlow实践 王琛,胡振邦,高杰 著 ISBN:9787121312984 包装:平装 开本:16开 用纸:胶版纸 正文语种:中文 出版社:电子工业出版社 出版时间:2017 ...
- 《深度学习原理与TensorFlow实践》喻俨，莫瑜
1. 深度学习简介 2. TensorFlow系统介绍 3. Hello TensorFlow 4. CNN看懂世界 5. RNN能说会道 6. CNN LSTM看图说话 7. 损失函数与优化算法 T ...
- tornado项目搭建_Python深度学习原理及项目实战2019年3月21日上海举办
一.课程背景 众所周知,人工智能是高级计算智能最宽泛的概念,机器学习是研究人工智能的一个工具,深度学习是机器学习的一个子集,是目前研究领域卓有成效的学习方法.深度学习的框架有很多,而TenforFlo ...
- 深度学习利器：TensorFlow与NLP模型
深度学习利器:TensorFlow与NLP模型 享到:微博微信FacebookTwitter有道云笔记邮件分享 稍后阅读 我的阅读清单 前言 自然语言处理(简称NLP),是研究计算机处理人类语言的一门 ...
最新文章
- linux shell 命令执行超时终止
- 使用emu8086学习汇编 int 21h 指令
- windows 包管理器scoop的使用
- 网易搞事情，一首《程序员disco》唱给1024程序员节
- 应用深度学习使用 Tensorflow 对音频进行分类
- PHP web应用的调试
- linux中vi过滤,vim高级编辑（一）
- Taro+react开发（85):taro路由参数
- 微信小程序 body属性的问题
- 【2016年第3期】大数据时代的数据科学家培养
- Openstack 平台部署安装
- 夯实基础，彻底掌握js的核心技术(四)：ES5、ES6对象方法详解
- 从C#程序中调用非受管DLLs
- 关于np.meshgrid
- java swf pdf_基于Java SWFTools实现把pdf转成swf
- Linux下C语言编程（1）：IO编程
- [3维影像]360度杯子环绕拍摄
- Kafka集群搭建(四节点)
- Python编程基础——编程环境准备工作
- 密歇根安娜堡计算机排名,密歇根大学的计算机系统排名，该细心来考查