example-notebook/Tutorial_06 A very simple example for tensorboard.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 例1 - 在 tensorboard 中记录 tensor 的变化。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "created log_dir path\n"
     ]
    }
   ],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings('ignore')  # 不打印 warning \n",
    "\n",
    "import tensorflow as tf\n",
    "\n",
    "# 设置GPU按需增长\n",
    "config = tf.ConfigProto()\n",
    "config.gpu_options.allow_growth = True\n",
    "sess = tf.Session(config=config)\n",
    "\n",
    "import os\n",
    "import shutil\n",
    "\n",
    "\"\"\"TensorBoard 简单例子。\n",
    "tf.summary.scalar('var_name', var)        # 记录标量的变化\n",
    "tf.summary.histogram('vec_name', vec)     # 记录向量或者矩阵，tensor的数值分布变化。\n",
    "\n",
    "merged = tf.summary.merge_all()           # 把所有的记录并把他们写到 log_dir 中\n",
    "train_writer = tf.summary.FileWriter(log_dir + '/add_example', sess.graph)  # 保存位置\n",
    "\n",
    "运行完后，在命令行中输入 tensorboard --logdir=log_dir_path(你保存到log路径)\n",
    "\"\"\"\n",
    "\n",
    "log_dir = '../summary/graph/'\n",
    "if os.path.exists(log_dir):   # 删掉以前的summary，以免重合\n",
    "    shutil.rmtree(log_dir)\n",
    "os.makedirs(log_dir)\n",
    "print('created log_dir path')\n",
    "\n",
    "with tf.name_scope('add_example'):\n",
    "    a = tf.Variable(tf.truncated_normal([100,1], mean=0.5, stddev=0.5), name='var_a')\n",
    "    tf.summary.histogram('a_hist', a)\n",
    "    b = tf.Variable(tf.truncated_normal([100,1], mean=-0.5, stddev=1.0), name='var_b')\n",
    "    tf.summary.histogram('b_hist', b)\n",
    "    increase_b = tf.assign(b, b + 0.2)\n",
    "    c = tf.add(a, b)\n",
    "    tf.summary.histogram('c_hist', c)\n",
    "    c_mean = tf.reduce_mean(c)\n",
    "    tf.summary.scalar('c_mean', c_mean)\n",
    "merged = tf.summary.merge_all()\n",
    "writer = tf.summary.FileWriter(log_dir+'add_example', sess.graph)\n",
    "\n",
    "\n",
    "sess.run(tf.global_variables_initializer())\n",
    "for step in range(500):\n",
    "    sess.run([merged, increase_b])    # 每步改变一次 b 的值\n",
    "    summary = sess.run(merged)\n",
    "    writer.add_summary(summary, step)\n",
    "writer.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 例2： 分 train 和 test 两部分进行记录。\n",
    "**Restart Kernel 然后继续**。每隔10个step记录一次test结果。一般在网络训练过程中，我们都会隔一段时间跑一次 valid。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "created log_dir path\n",
      "END!\n"
     ]
    }
   ],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings('ignore')  # 不打印 warning \n",
    "\n",
    "import tensorflow as tf\n",
    "\n",
    "# 设置GPU按需增长\n",
    "config = tf.ConfigProto()\n",
    "config.gpu_options.allow_growth = True\n",
    "sess = tf.Session(config=config)\n",
    "\n",
    "import numpy as np\n",
    "import os\n",
    "import shutil\n",
    "\n",
    "\"\"\"TensorBoard 简单例子。\n",
    "tf.summary.scalar('var_name', var)        # 记录标量的变化\n",
    "tf.summary.histogram('vec_name', vec)     # 记录向量或者矩阵，tensor的数值分布变化。\n",
    "\n",
    "merged = tf.summary.merge_all()           # 把所有的记录并把他们写到 log_dir 中\n",
    "train_writer = tf.summary.FileWriter(log_dir + '/train', sess.graph)  # 保存位置\n",
    "test_writer = tf.summary.FileWriter(log_dir + '/test', sess.graph)\n",
    "运行完后，在命令行中输入 tensorboard --logdir=log_dir_path(你保存到log路径)\n",
    "\"\"\"\n",
    "\n",
    "log_dir = '../summary/graph2/'\n",
    "if os.path.exists(log_dir):   # 删掉以前的summary，以免重合\n",
    "    shutil.rmtree(log_dir)\n",
    "os.makedirs(log_dir)\n",
    "print('created log_dir path')\n",
    "\n",
    "a = tf.placeholder(dtype=tf.float32, shape=[100,1], name='a')\n",
    "\n",
    "with tf.name_scope('add_example'):\n",
    "    b = tf.Variable(tf.truncated_normal([100,1], mean=-0.5, stddev=1.0), name='var_b')\n",
    "    tf.summary.histogram('b_hist', b)\n",
    "    increase_b = tf.assign(b, b + 0.2)\n",
    "    c = tf.add(a, b)\n",
    "    tf.summary.histogram('c_hist', c)\n",
    "    c_mean = tf.reduce_mean(c)\n",
    "    tf.summary.scalar('c_mean', c_mean)\n",
    "merged = tf.summary.merge_all()\n",
    "train_writer = tf.summary.FileWriter(log_dir + '/train', sess.graph)  # 保存位置\n",
    "test_writer = tf.summary.FileWriter(log_dir + '/test', sess.graph)\n",
    "\n",
    "\n",
    "sess.run(tf.global_variables_initializer())\n",
    "for step in range(500):\n",
    "    if (step+1) % 10 == 0:\n",
    "        _a = np.random.randn(100,1) \n",
    "        summary, _ = sess.run([merged, increase_b], feed_dict={a: _a})    # 每步改变一次 b 的值\n",
    "        test_writer.add_summary(summary, step)\n",
    "    else:\n",
    "        _a = np.random.randn(100,1) + step*0.2\n",
    "        summary, _ = sess.run([merged, increase_b], feed_dict={a: _a})    # 每步改变一次 b 的值\n",
    "        train_writer.add_summary(summary, step)\n",
    "train_writer.close()\n",
    "test_writer.close()\n",
    "print('END!')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "训练完后，在summary目录下输入 tensorboard --logdir graph2 就能看到结果了。如下图所示："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import Image\n",
    "Image('../figs/graph2.png') "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 关于 tensorboard 的一点心得"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- 1.一定要学会使用 tf.variable_scope() 和 tf.name_scope(),否则稍微复杂一点的网络都会乱七八糟。你可以通过上图中的 graph 来看看自己构建的网络结构。\n",
    "- 2.使用 tensorboard 来看 training 和 validation 的 loss 和 accuracy 变化对于调参非常非常有帮助。\n",
    "- 3.tf.summary.scalar() 来记录某个标量的变化；使用 tf.summary.histogram() 来记录某个矩阵的数值分布。**特别要注意在统计值的分布的时候是要耗费大量的计算的，如果是一些没关紧要的参数就不要记录了，否则影响速度又浪费资源。**\n",
    "\n",
    "想要了解更多，可以参考[详解 TensorBoard－如何调参](http://blog.csdn.net/aliceyangxi1987/article/details/71716596)"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}