devforfu/test.ipynb

## test.ipynb
{
 "cells": [
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "# Problem Statement\n\nWhile training models on large (1-100M samples) datasets, the amount of used RAM gradually increases during a single training epoch. After a few thousands of training batches, the following exception is usually raised:\n```\nRuntimeError: DataLoader worker (pid 31352) is killed by signal: Killed.\n```\nSometimes the format of exception is a little bit different. However, a cause is always the same: the kernel kills training process because it exhausts all available system's memory, doesn't matter how many GB of RAM are available. The plot below was generated when a ResNet18 model was trained on a subset of [Quick Draw Doodle](https://www.kaggle.com/c/quickdraw-doodle-recognition/) competition dataset, with 50000 records per category, `17M = 50K * 340` samples in total. The curve clearly shows a memory leak that happens during the single training epoch.\n\n![](./mem.png)\n\n\nThe goal of this notebook is to create a snippet that would help to reproduce the leakage to understand if it happens within the `fastai` library's codebase, or on `pytorch` side.\n\n## How to reproduce\n\nIt seems that the leakage shows itself during long training periods. It is clearly seen even during a single training epoch on large dataset. Also, the memory is not deallocated even when an epoch is completed. Therefore, to reproduce, one can try to reduce the batch size, increase a number of training epochs, or do both. In other words, the number of training iterations during epoch should be relatively high (> 10000). \n\nThe author tried to variate a large set of training conditions:\n\n\n1. Using the `ImageDataBunch.create` method vs. data block API \n2. Using a custom dataset generating images on the fly agains reading them from persistent storage\n3. Using different data transformations \n4. Using different versions of fastai library (1.0.22 and 1.0.28)\n\nIn every case, the `RuntimeError` exception was eventually raised.\n\nThe dataset selected in this notebook didn't allow author to _exactly_ reproduce the picture shown above. However, it still shows gradual increasing of consumed memory during epoch. The author expects that this value shouldn't increase at all as soon as the data loader's workers are spawned, and the training process is launched.\n\n## Imports"
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "%matplotlib inline\n%reload_ext autoreload\n%autoreload 2",
   "execution_count": 1,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "from multiprocessing import cpu_count",
   "execution_count": 2,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "from fastai import *\nfrom fastai.vision import *\nimport psutil",
   "execution_count": 3,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "device = 1",
   "execution_count": null,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "defaults.device = torch.device(f'cuda:{device}')",
   "execution_count": 4,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "np.random.seed(1)",
   "execution_count": 5,
   "outputs": []
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "## Memory Tracking Callback\n\nTo track memory usage during training process, the `MemoryUsage` callback was created. It calls the `psutil.virtual_memory()` function on each batch end and saves returned metrics into CSV file."
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "class MemoryUsage(Callback):\n\n    def __init__(self, memory_log: str='memory.csv'):\n        self.memory_log = memory_log\n        self.iter = None\n        self._stream = None\n\n    def on_train_begin(self, **kwargs):\n        self.iter = 0\n        if self._stream is not None:\n            self.close()\n        self._stream = open(self.memory_log, 'w')\n        self._stream.write('index,mem_percent,mem_free,mem_available,mem_used\\n')\n        super().on_train_begin(**kwargs)\n\n    def on_train_end(self, **kwargs):\n        self.close()\n\n    def on_batch_end(self, **kwargs):\n        self.iter += 1\n        mem = psutil.virtual_memory()\n        record = [self.iter, mem.percent, mem.free, mem.available, mem.used]\n        string = ','.join([str(x) for x in record])\n        self._stream.write(string + '\\n')\n        return False\n\n    def close(self):\n        self._stream.flush()\n        self._stream.close()",
   "execution_count": 7,
   "outputs": []
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "## Dataset\n\nTo reveal memory leak, the [The Oxford-IIIT Pet Dataset](http://www.robots.ox.ac.uk/~vgg/data/pets/). However, it could be any other dataset as soon as, presumably, the leakage happens on many machines and doesn't depend on hardware and/or dataset size, and only depends on number of training iterations."
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "path = untar_data(URLs.PETS)\npath_anno = path/'annotations'\npath_img = path/'images'\nfnames = get_image_files(path_img)\npat = re.compile(r'/([^/]+)_\\d+.jpg$')",
   "execution_count": 8,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "# tune depending on your hardware\nimage_size = 1500\nbatch_size = 4\nepochs = 10\nnum_workers = cpu_count()",
   "execution_count": 11,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "data = ImageDataBunch.from_name_re(\n    path_img, fnames, pat, \n    ds_tfms=get_transforms(),\n    size=image_size,\n    bs=batch_size, \n    num_workers=num_workers\n)\n_ = data.normalize(imagenet_stats)",
   "execution_count": 14,
   "outputs": []
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "## The Model's Training \n\nOriginally, the issue was recorded when training `ResNet18` model from `torchvision` package. Therefore, the same setting is used here. However, presumably, it doesn't depend on the selected architecture as soon as similar `RuntimeError` exceptions were raised while training with other architectures."
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "arch = models.resnet18",
   "execution_count": 15,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "learn = create_cnn(data, arch)",
   "execution_count": 16,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "learn.fit_one_cycle(epochs, callbacks=[MemoryUsage()])",
   "execution_count": null,
   "outputs": []
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "## Memory Consumption Plots\n"
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "mem = pd.read_csv('memory.csv')\nmem['mem_available'] = mem.mem_available / 1024 ** 3\nmem['mem_used'] = mem.mem_used / 1024 ** 3\nmem.head(4)",
   "execution_count": 18,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 18,
     "data": {
      "text/plain": "   index  mem_percent     mem_free  mem_available  mem_used\n0      1         19.1  21205766144      25.344410  4.428032\n1      2         19.0  21251350528      25.386864  4.429394\n2      3         18.9  21284655104      25.417866  4.442207\n3      4         18.8  21301227520      25.433239  4.470650",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>index</th>\n      <th>mem_percent</th>\n      <th>mem_free</th>\n      <th>mem_available</th>\n      <th>mem_used</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>1</td>\n      <td>19.1</td>\n      <td>21205766144</td>\n      <td>25.344410</td>\n      <td>4.428032</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>2</td>\n      <td>19.0</td>\n      <td>21251350528</td>\n      <td>25.386864</td>\n      <td>4.429394</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>3</td>\n      <td>18.9</td>\n      <td>21284655104</td>\n      <td>25.417866</td>\n      <td>4.442207</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>4</td>\n      <td>18.8</td>\n      <td>21301227520</td>\n      <td>25.433239</td>\n      <td>4.470650</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "f, ax = plt.subplots(2, 1, figsize=(10, 12))\n\nax1, ax2 = ax.flat\n\nmem.plot(x='index', y='mem_percent', ax=ax1)\nmem.plot(x='index', y='mem_available', ax=ax2)\nmem.plot(x='index', y='mem_used', ax=ax2)",
   "execution_count": 19,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 19,
     "data": {
      "text/plain": "<matplotlib.axes._subplots.AxesSubplot at 0x7ff3d0791f98>"
     },
     "metadata": {}
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 720x864 with 2 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "",
   "execution_count": null,
   "outputs": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "name": "python3",
   "display_name": "Python 3",
   "language": "python"
  },
  "language_info": {
   "name": "python",
   "version": "3.7.0",
   "mimetype": "text/x-python",
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "pygments_lexer": "ipython3",
   "nbconvert_exporter": "python",
   "file_extension": ".py"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}