WeightWatcher-ResNet.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# WeightWatcher ResNet\n",
    "\n",
    "https://calculationconsulting.com\n",
    "\n",
    "\n",
    "### Analysis of pyTorch ResNet Models\n",
    "\n",
    "ResNet, with the same number of layers,  is a much smaller model than VGG, with 20X less parameters.\n",
    "\n",
    "We will see the power law exponents $\\alpha$ approach 2 (or less) and do not vary much between layers until the very end, where the power law model breaks down and we have bulk+spikes.  Compare this to VGG, where $\\alpha$ immediately increases with layer depth\n",
    "\n",
    "Moreover, the average exponent $\\langle\\alpha\\langle$ decreaes with test error, whereas for the VGG series, only the weighted alpha $\\hat{\\alpha}$  decreases with test error.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 1,
     "metadata": {
      "image/png": {
       "height": 400,
       "width": 800
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import Image\n",
    "Image(filename='CV-models.png',width=800, height=400)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 2,
     "metadata": {
      "image/png": {
       "height": 400,
       "width": 800
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Image(filename='resnet18.png',width=800, height=400)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Comparison of VGG and ResNet\n",
    "\n",
    "https://towardsdatascience.com/an-overview-of-resnet-and-its-variants-5281e2f56035\n",
    "\n",
    "### Key Idea:  Residual Connections\n",
    "\n",
    "Improves gradient / information flow through the network\n",
    "\n",
    "Because of this, there are much deeper ResNet variants.  In fact, some researchers have trained ResNet variants with 1001 layers!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Results\n",
    "\n",
    "In contrast to the VGG series, the ResNet Average Alpha $\\langle\\alpha\\rangle$ is positively correlated with the Test Error:  smaller average alpha corresponds to better generalization.  Moreover, the layer $\\alpha$ are fairly constant, and near $2.0$, for at least $3/4$ of the network"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Calculations of Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-03-05T23:51:16.544327Z",
     "start_time": "2019-03-05T23:51:16.531457Z"
    },
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# Suppress the powerlaw package warnings\n",
    "# \"powerlaw.py:700: RuntimeWarning: divide by zero encountered in true_divide\"\n",
    "# \"powerlaw.py:700: RuntimeWarning: invalid value encountered in true_divide\"\n",
    "import warnings\n",
    "warnings.simplefilter(action='ignore', category=RuntimeWarning)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-10-01T22:51:12.554859Z",
     "start_time": "2018-10-01T22:51:12.518859Z"
    }
   },
   "source": [
    "### Import WeightWatcher\n",
    "\n",
    "set custom Logging at WARN Level"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-03-06T00:16:41.769719Z",
     "start_time": "2019-03-06T00:16:38.803181Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.4.0'"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import logging\n",
    "logging.basicConfig(level=logging.INFO)\n",
    "logger = logging.getLogger(__name__)\n",
    "logger.setLevel(logging.WARN)\n",
    "\n",
    "import weightwatcher as ww\n",
    "import torchvision.models as models\n",
    "\n",
    "ww.__version__\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "importing Jupyter notebook from ModelPlots.ipynb\n"
     ]
    }
   ],
   "source": [
    "import import_ipynb\n",
    "import ModelPlots as MP"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Create all models now"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "series_name = 'ResNet'\n",
    "all_names = [ 'resnet18', 'resnet34', 'resnet50', 'resnet101','resnet152']\n",
    "\n",
    "colors = ['blue', 'green', 'teal', 'orange', 'red']\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-03-06T00:16:43.249725Z",
     "start_time": "2019-03-06T00:16:43.240611Z"
    }
   },
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "all_models = []\n",
    "all_models.append(models.resnet18(pretrained=True))\n",
    "all_models.append(models.resnet34(pretrained=True))\n",
    "all_models.append(models.resnet50(pretrained=True))\n",
    "all_models.append(models.resnet101(pretrained=True))\n",
    "all_models.append(models.resnet152(pretrained=True))\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Get reported accuracies from pytorch website\n",
    "\n",
    "https://pytorch.org/docs/stable/torchvision/models.html\n",
    "\n",
    "<pre>\n",
    "<table class=\"docutils align-default\">\n",
    "<colgroup>\n",
    "<col style=\"width: 55%\" />\n",
    "<col style=\"width: 22%\" />\n",
    "<col style=\"width: 22%\" />\n",
    "</colgroup>\n",
    "<thead>\n",
    "<thead>\n",
    "<tr class=\"row-odd\"><th class=\"head\"><p>Network</p></th>\n",
    "<th class=\"head\"><p>Top-1 error</p></th>\n",
    "<th class=\"head\"><p>Top-5 error</p></th>\n",
    "</tr>\n",
    "</thead>\n",
    "<tbody>\n",
    "\n",
    "<tr class=\"row-odd\"><td><p>ResNet-18</p></td>\n",
    "<td><p>30.24</p></td>\n",
    "<td><p>10.92</p></td>\n",
    "</tr>\n",
    "<tr class=\"row-even\"><td><p>ResNet-34</p></td>\n",
    "<td><p>26.70</p></td>\n",
    "<td><p>8.58</p></td>\n",
    "</tr>\n",
    "<tr class=\"row-odd\"><td><p>ResNet-50</p></td>\n",
    "<td><p>23.85</p></td>\n",
    "<td><p>7.13</p></td>\n",
    "</tr>\n",
    "<tr class=\"row-even\"><td><p>ResNet-101</p></td>\n",
    "<td><p>22.63</p></td>\n",
    "<td><p>6.44</p></td>\n",
    "</tr>\n",
    "<tr class=\"row-odd\"><td><p>ResNet-152</p></td>\n",
    "<td><p>21.69</p></td>\n",
    "<td><p>5.94</p></td>\n",
    "</tr>\n",
    "\n",
    "</tbody>\n",
    "</table>\n",
    "</pre>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "top1_errors= {\n",
    "    \n",
    "    \"resnet18\": 30.24,\n",
    "    \"resnet34\": 26.70,\n",
    "    \"resnet50\": 23.85,\n",
    "    \"resnet101\": 22.63,\n",
    "    \"resnet152\": 21.69 \n",
    "\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "top5_errors= {\n",
    "    \n",
    "    \"resnet18\": 10.92,\n",
    "    \"resnet34\": 8.58,\n",
    "    \"resnet50\": 7.13,\n",
    "    \"resnet101\": 6.44,\n",
    "    \"resnet152\": 5.94    \n",
    "\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Run WeightWatcher, collect summary and details (as dataframes) for all models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:weightwatcher:\n",
      "\n",
      "python      version 3.8.5 (default, Sep  4 2020, 02:22:02) \n",
      "[Clang 10.0.0 ]\n",
      "numpy       version 1.18.5\n",
      "tensforflow version 2.3.0\n",
      "keras       version 2.4.0\n",
      "INFO:weightwatcher:params {'glorot_fix': False, 'normalize': False, 'conv2d_norm': True, 'randomize': False, 'min_evals': 0, 'max_evals': None, 'plot': False, 'mp_fit': True, 'ww2x': False}\n",
      "INFO:weightwatcher:LAYER: 1 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 1 None \n",
      "INFO:weightwatcher:Randomizing Layer: 1 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 1 None \n",
      "INFO:weightwatcher:Layer id 2  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 7 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 7 None \n",
      "INFO:weightwatcher:Randomizing Layer: 7 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 7 None \n",
      "INFO:weightwatcher:Layer id 8  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 10 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 10 None \n",
      "INFO:weightwatcher:Randomizing Layer: 10 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 10 None \n",
      "INFO:weightwatcher:Layer id 11  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 13 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 13 None \n",
      "INFO:weightwatcher:Randomizing Layer: 13 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 13 None \n",
      "INFO:weightwatcher:Layer id 14  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 16 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 16 None \n",
      "INFO:weightwatcher:Randomizing Layer: 16 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 16 None \n",
      "INFO:weightwatcher:Layer id 17  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 20 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 20 None \n",
      "INFO:weightwatcher:Randomizing Layer: 20 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 20 None \n",
      "INFO:weightwatcher:Layer id 21  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 23 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 23 None \n",
      "INFO:weightwatcher:Randomizing Layer: 23 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 23 None \n",
      "INFO:weightwatcher:Layer id 24  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 26 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 26 None \n",
      "INFO:weightwatcher:Randomizing Layer: 26 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 26 None \n",
      "INFO:weightwatcher:Layer id 27  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 29 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 29 None \n",
      "INFO:weightwatcher:Randomizing Layer: 29 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 29 None \n",
      "INFO:weightwatcher:Layer id 30  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 32 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 32 None \n",
      "INFO:weightwatcher:Randomizing Layer: 32 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 32 None \n",
      "INFO:weightwatcher:Layer id 33  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 36 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 36 None \n",
      "INFO:weightwatcher:Randomizing Layer: 36 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 36 None \n",
      "INFO:weightwatcher:Layer id 37  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 39 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 39 None \n",
      "INFO:weightwatcher:Randomizing Layer: 39 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 39 None \n",
      "INFO:weightwatcher:Layer id 40  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 42 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 42 None \n",
      "INFO:weightwatcher:Randomizing Layer: 42 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 42 None \n",
      "INFO:weightwatcher:Layer id 43  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 45 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 45 None \n",
      "INFO:weightwatcher:Randomizing Layer: 45 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 45 None \n",
      "INFO:weightwatcher:Layer id 46  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 48 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 48 None \n",
      "INFO:weightwatcher:Randomizing Layer: 48 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 48 None \n",
      "INFO:weightwatcher:Layer id 49  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 52 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 52 None \n",
      "INFO:weightwatcher:Randomizing Layer: 52 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 52 None \n",
      "INFO:weightwatcher:Layer id 53  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 55 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 55 None \n",
      "INFO:weightwatcher:Randomizing Layer: 55 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 55 None \n",
      "INFO:weightwatcher:Layer id 56  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 58 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 58 None \n",
      "INFO:weightwatcher:Randomizing Layer: 58 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 58 None \n",
      "INFO:weightwatcher:Layer id 59  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 61 8  : <class 'torch.nn.modules.conv.Conv2d'>\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:weightwatcher:MP Fitting Layer: 61 None \n",
      "INFO:weightwatcher:Randomizing Layer: 61 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 61 None \n",
      "INFO:weightwatcher:Layer id 62  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 64 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 64 None \n",
      "INFO:weightwatcher:Randomizing Layer: 64 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 64 None \n",
      "INFO:weightwatcher:Layer id 65  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 67 2  : <class 'torch.nn.modules.linear.Linear'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 67 None \n",
      "INFO:weightwatcher:Randomizing Layer: 67 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 67 None \n",
      "INFO:weightwatcher:\n",
      "\n",
      "python      version 3.8.5 (default, Sep  4 2020, 02:22:02) \n",
      "[Clang 10.0.0 ]\n",
      "numpy       version 1.18.5\n",
      "tensforflow version 2.3.0\n",
      "keras       version 2.4.0\n",
      "INFO:weightwatcher:params {'glorot_fix': False, 'normalize': False, 'conv2d_norm': True, 'randomize': False, 'min_evals': 0, 'max_evals': None, 'plot': False, 'mp_fit': True, 'ww2x': False}\n",
      "INFO:weightwatcher:LAYER: 1 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 1 None \n",
      "INFO:weightwatcher:Randomizing Layer: 1 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 1 None \n",
      "INFO:weightwatcher:Layer id 2  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 7 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 7 None \n",
      "INFO:weightwatcher:Randomizing Layer: 7 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 7 None \n",
      "INFO:weightwatcher:Layer id 8  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 10 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 10 None \n",
      "INFO:weightwatcher:Randomizing Layer: 10 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 10 None \n",
      "INFO:weightwatcher:Layer id 11  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 13 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 13 None \n",
      "INFO:weightwatcher:Randomizing Layer: 13 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 13 None \n",
      "INFO:weightwatcher:Layer id 14  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 16 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 16 None \n",
      "INFO:weightwatcher:Randomizing Layer: 16 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 16 None \n",
      "INFO:weightwatcher:Layer id 17  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 19 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 19 None \n",
      "INFO:weightwatcher:Randomizing Layer: 19 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 19 None \n",
      "INFO:weightwatcher:Layer id 20  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 22 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 22 None \n",
      "INFO:weightwatcher:Randomizing Layer: 22 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 22 None \n",
      "INFO:weightwatcher:Layer id 23  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 26 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 26 None \n",
      "INFO:weightwatcher:Randomizing Layer: 26 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 26 None \n",
      "INFO:weightwatcher:Layer id 27  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 29 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 29 None \n",
      "INFO:weightwatcher:Randomizing Layer: 29 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 29 None \n",
      "INFO:weightwatcher:Layer id 30  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 32 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 32 None \n",
      "INFO:weightwatcher:Randomizing Layer: 32 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 32 None \n",
      "INFO:weightwatcher:Layer id 33  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 35 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 35 None \n",
      "INFO:weightwatcher:Randomizing Layer: 35 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 35 None \n",
      "INFO:weightwatcher:Layer id 36  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 38 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 38 None \n",
      "INFO:weightwatcher:Randomizing Layer: 38 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 38 None \n",
      "INFO:weightwatcher:Layer id 39  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 41 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 41 None \n",
      "INFO:weightwatcher:Randomizing Layer: 41 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 41 None \n",
      "INFO:weightwatcher:Layer id 42  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 44 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 44 None \n",
      "INFO:weightwatcher:Randomizing Layer: 44 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 44 None \n",
      "INFO:weightwatcher:Layer id 45  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 47 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 47 None \n",
      "INFO:weightwatcher:Randomizing Layer: 47 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 47 None \n",
      "INFO:weightwatcher:Layer id 48  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 50 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 50 None \n",
      "INFO:weightwatcher:Randomizing Layer: 50 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 50 None \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:weightwatcher:Layer id 51  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 54 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 54 None \n",
      "INFO:weightwatcher:Randomizing Layer: 54 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 54 None \n",
      "INFO:weightwatcher:Layer id 55  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 57 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 57 None \n",
      "INFO:weightwatcher:Randomizing Layer: 57 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 57 None \n",
      "INFO:weightwatcher:Layer id 58  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 60 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 60 None \n",
      "INFO:weightwatcher:Randomizing Layer: 60 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 60 None \n",
      "INFO:weightwatcher:Layer id 61  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 63 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 63 None \n",
      "INFO:weightwatcher:Randomizing Layer: 63 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 63 None \n",
      "INFO:weightwatcher:Layer id 64  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 66 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 66 None \n",
      "INFO:weightwatcher:Randomizing Layer: 66 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 66 None \n",
      "INFO:weightwatcher:Layer id 67  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 69 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 69 None \n",
      "INFO:weightwatcher:Randomizing Layer: 69 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 69 None \n",
      "INFO:weightwatcher:Layer id 70  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 72 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 72 None \n",
      "INFO:weightwatcher:Randomizing Layer: 72 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 72 None \n",
      "INFO:weightwatcher:Layer id 73  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 75 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 75 None \n",
      "INFO:weightwatcher:Randomizing Layer: 75 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 75 None \n",
      "INFO:weightwatcher:Layer id 76  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 78 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 78 None \n",
      "INFO:weightwatcher:Randomizing Layer: 78 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 78 None \n",
      "INFO:weightwatcher:Layer id 79  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 81 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 81 None \n",
      "INFO:weightwatcher:Randomizing Layer: 81 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 81 None \n",
      "INFO:weightwatcher:Layer id 82  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 84 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 84 None \n",
      "INFO:weightwatcher:Randomizing Layer: 84 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 84 None \n",
      "INFO:weightwatcher:Layer id 85  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 87 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 87 None \n",
      "INFO:weightwatcher:Randomizing Layer: 87 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 87 None \n",
      "INFO:weightwatcher:Layer id 88  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 90 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 90 None \n",
      "INFO:weightwatcher:Randomizing Layer: 90 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 90 None \n",
      "INFO:weightwatcher:Layer id 91  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 94 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 94 None \n",
      "INFO:weightwatcher:Randomizing Layer: 94 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 94 None \n",
      "INFO:weightwatcher:Layer id 95  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 97 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 97 None \n",
      "INFO:weightwatcher:Randomizing Layer: 97 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 97 None \n",
      "INFO:weightwatcher:Layer id 98  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 100 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 100 None \n",
      "INFO:weightwatcher:Randomizing Layer: 100 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 100 None \n",
      "INFO:weightwatcher:Layer id 101  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 103 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 103 None \n",
      "INFO:weightwatcher:Randomizing Layer: 103 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 103 None \n",
      "INFO:weightwatcher:Layer id 104  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 106 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 106 None \n",
      "INFO:weightwatcher:Randomizing Layer: 106 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 106 None \n",
      "INFO:weightwatcher:Layer id 107  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 109 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 109 None \n",
      "INFO:weightwatcher:Randomizing Layer: 109 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 109 None \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:weightwatcher:Layer id 110  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 112 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 112 None \n",
      "INFO:weightwatcher:Randomizing Layer: 112 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 112 None \n",
      "INFO:weightwatcher:Layer id 113  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 115 2  : <class 'torch.nn.modules.linear.Linear'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 115 None \n",
      "INFO:weightwatcher:Randomizing Layer: 115 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 115 None \n",
      "INFO:weightwatcher:\n",
      "\n",
      "python      version 3.8.5 (default, Sep  4 2020, 02:22:02) \n",
      "[Clang 10.0.0 ]\n",
      "numpy       version 1.18.5\n",
      "tensforflow version 2.3.0\n",
      "keras       version 2.4.0\n",
      "INFO:weightwatcher:params {'glorot_fix': False, 'normalize': False, 'conv2d_norm': True, 'randomize': False, 'min_evals': 0, 'max_evals': None, 'plot': False, 'mp_fit': True, 'ww2x': False}\n",
      "INFO:weightwatcher:LAYER: 1 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 1 None \n",
      "INFO:weightwatcher:Randomizing Layer: 1 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 1 None \n",
      "INFO:weightwatcher:Layer id 2  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 7 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 7 None \n",
      "INFO:weightwatcher:Randomizing Layer: 7 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 7 None \n",
      "INFO:weightwatcher:Layer id 8  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 9 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 9 None \n",
      "INFO:weightwatcher:Randomizing Layer: 9 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 9 None \n",
      "INFO:weightwatcher:Layer id 10  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 11 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 11 None \n",
      "INFO:weightwatcher:Randomizing Layer: 11 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 11 None \n",
      "INFO:weightwatcher:Layer id 12  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 15 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 15 None \n",
      "INFO:weightwatcher:Randomizing Layer: 15 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 15 None \n",
      "INFO:weightwatcher:Layer id 16  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 18 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 18 None \n",
      "INFO:weightwatcher:Randomizing Layer: 18 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 18 None \n",
      "INFO:weightwatcher:Layer id 19  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 20 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 20 None \n",
      "INFO:weightwatcher:Randomizing Layer: 20 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 20 None \n",
      "INFO:weightwatcher:Layer id 21  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 22 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 22 None \n",
      "INFO:weightwatcher:Randomizing Layer: 22 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 22 None \n",
      "INFO:weightwatcher:Layer id 23  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 26 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 26 None \n",
      "INFO:weightwatcher:Randomizing Layer: 26 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 26 None \n",
      "INFO:weightwatcher:Layer id 27  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 28 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 28 None \n",
      "INFO:weightwatcher:Randomizing Layer: 28 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 28 None \n",
      "INFO:weightwatcher:Layer id 29  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 30 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 30 None \n",
      "INFO:weightwatcher:Randomizing Layer: 30 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 30 None \n",
      "INFO:weightwatcher:Layer id 31  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 35 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 35 None \n",
      "INFO:weightwatcher:Randomizing Layer: 35 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 35 None \n",
      "INFO:weightwatcher:Layer id 36  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 37 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 37 None \n",
      "INFO:weightwatcher:Randomizing Layer: 37 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 37 None \n",
      "INFO:weightwatcher:Layer id 38  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 39 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 39 None \n",
      "INFO:weightwatcher:Randomizing Layer: 39 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 39 None \n",
      "INFO:weightwatcher:Layer id 40  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 43 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 43 None \n",
      "INFO:weightwatcher:Randomizing Layer: 43 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 43 None \n",
      "INFO:weightwatcher:Layer id 44  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 46 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 46 None \n",
      "INFO:weightwatcher:Randomizing Layer: 46 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 46 None \n",
      "INFO:weightwatcher:Layer id 47  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 48 8  : <class 'torch.nn.modules.conv.Conv2d'>\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:weightwatcher:MP Fitting Layer: 48 None \n",
      "INFO:weightwatcher:Randomizing Layer: 48 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 48 None \n",
      "INFO:weightwatcher:Layer id 49  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 50 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 50 None \n",
      "INFO:weightwatcher:Randomizing Layer: 50 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 50 None \n",
      "INFO:weightwatcher:Layer id 51  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 54 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 54 None \n",
      "INFO:weightwatcher:Randomizing Layer: 54 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 54 None \n",
      "INFO:weightwatcher:Layer id 55  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 56 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 56 None \n",
      "INFO:weightwatcher:Randomizing Layer: 56 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 56 None \n",
      "INFO:weightwatcher:Layer id 57  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 58 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 58 None \n",
      "INFO:weightwatcher:Randomizing Layer: 58 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 58 None \n",
      "INFO:weightwatcher:Layer id 59  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 62 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 62 None \n",
      "INFO:weightwatcher:Randomizing Layer: 62 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 62 None \n",
      "INFO:weightwatcher:Layer id 63  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 64 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 64 None \n",
      "INFO:weightwatcher:Randomizing Layer: 64 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 64 None \n",
      "INFO:weightwatcher:Layer id 65  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 66 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 66 None \n",
      "INFO:weightwatcher:Randomizing Layer: 66 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 66 None \n",
      "INFO:weightwatcher:Layer id 67  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 71 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 71 None \n",
      "INFO:weightwatcher:Randomizing Layer: 71 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 71 None \n",
      "INFO:weightwatcher:Layer id 72  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 73 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 73 None \n",
      "INFO:weightwatcher:Randomizing Layer: 73 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 73 None \n",
      "INFO:weightwatcher:Layer id 74  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 75 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 75 None \n",
      "INFO:weightwatcher:Randomizing Layer: 75 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 75 None \n",
      "INFO:weightwatcher:Layer id 76  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 79 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 79 None \n",
      "INFO:weightwatcher:Randomizing Layer: 79 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 79 None \n",
      "INFO:weightwatcher:Layer id 80  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 82 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 82 None \n",
      "INFO:weightwatcher:Randomizing Layer: 82 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 82 None \n",
      "INFO:weightwatcher:Layer id 83  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 84 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 84 None \n",
      "INFO:weightwatcher:Randomizing Layer: 84 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 84 None \n",
      "INFO:weightwatcher:Layer id 85  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 86 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 86 None \n",
      "INFO:weightwatcher:Randomizing Layer: 86 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 86 None \n",
      "INFO:weightwatcher:Layer id 87  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 90 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 90 None \n",
      "INFO:weightwatcher:Randomizing Layer: 90 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 90 None \n",
      "INFO:weightwatcher:Layer id 91  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 92 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 92 None \n",
      "INFO:weightwatcher:Randomizing Layer: 92 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 92 None \n",
      "INFO:weightwatcher:Layer id 93  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 94 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 94 None \n",
      "INFO:weightwatcher:Randomizing Layer: 94 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 94 None \n",
      "INFO:weightwatcher:Layer id 95  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 98 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 98 None \n",
      "INFO:weightwatcher:Randomizing Layer: 98 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 98 None \n",
      "INFO:weightwatcher:Layer id 99  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 100 8  : <class 'torch.nn.modules.conv.Conv2d'>\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:weightwatcher:MP Fitting Layer: 100 None \n",
      "INFO:weightwatcher:Randomizing Layer: 100 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 100 None \n",
      "INFO:weightwatcher:Layer id 101  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 102 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 102 None \n",
      "INFO:weightwatcher:Randomizing Layer: 102 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 102 None \n",
      "INFO:weightwatcher:Layer id 103  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 106 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 106 None \n",
      "INFO:weightwatcher:Randomizing Layer: 106 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 106 None \n",
      "INFO:weightwatcher:Layer id 107  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 108 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 108 None \n",
      "INFO:weightwatcher:Randomizing Layer: 108 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 108 None \n",
      "INFO:weightwatcher:Layer id 109  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 110 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 110 None \n",
      "INFO:weightwatcher:Randomizing Layer: 110 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 110 None \n",
      "INFO:weightwatcher:Layer id 111  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 114 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 114 None \n",
      "INFO:weightwatcher:Randomizing Layer: 114 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 114 None \n",
      "INFO:weightwatcher:Layer id 115  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 116 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 116 None \n",
      "INFO:weightwatcher:Randomizing Layer: 116 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 116 None \n",
      "INFO:weightwatcher:Layer id 117  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 118 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 118 None \n",
      "INFO:weightwatcher:Randomizing Layer: 118 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 118 None \n",
      "INFO:weightwatcher:Layer id 119  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 123 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 123 None \n",
      "INFO:weightwatcher:Randomizing Layer: 123 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 123 None \n",
      "INFO:weightwatcher:Layer id 124  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 125 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 125 None \n",
      "INFO:weightwatcher:Randomizing Layer: 125 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 125 None \n",
      "INFO:weightwatcher:Layer id 126  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 127 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 127 None \n",
      "INFO:weightwatcher:Randomizing Layer: 127 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 127 None \n",
      "INFO:weightwatcher:Layer id 128  unknown type 1 layer  BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 131 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 131 None \n",
      "INFO:weightwatcher:Randomizing Layer: 131 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 131 None \n",
      "INFO:weightwatcher:Layer id 132  unknown type 1 layer  BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 134 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 134 None \n",
      "INFO:weightwatcher:Randomizing Layer: 134 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 134 None \n",
      "INFO:weightwatcher:Layer id 135  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 136 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 136 None \n",
      "INFO:weightwatcher:Randomizing Layer: 136 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 136 None \n",
      "INFO:weightwatcher:Layer id 137  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 138 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 138 None \n",
      "INFO:weightwatcher:Randomizing Layer: 138 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 138 None \n",
      "INFO:weightwatcher:Layer id 139  unknown type 1 layer  BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 142 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 142 None \n",
      "INFO:weightwatcher:Randomizing Layer: 142 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 142 None \n",
      "INFO:weightwatcher:Layer id 143  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 144 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 144 None \n",
      "INFO:weightwatcher:Randomizing Layer: 144 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 144 None \n",
      "INFO:weightwatcher:Layer id 145  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 146 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 146 None \n",
      "INFO:weightwatcher:Randomizing Layer: 146 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 146 None \n",
      "INFO:weightwatcher:Layer id 147  unknown type 1 layer  BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 150 2  : <class 'torch.nn.modules.linear.Linear'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 150 None \n",
      "INFO:weightwatcher:Randomizing Layer: 150 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 150 None \n",
      "INFO:weightwatcher:\n",
      "\n",
      "python      version 3.8.5 (default, Sep  4 2020, 02:22:02) \n",
      "[Clang 10.0.0 ]\n",
      "numpy       version 1.18.5\n",
      "tensforflow version 2.3.0\n",
      "keras       version 2.4.0\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:weightwatcher:params {'glorot_fix': False, 'normalize': False, 'conv2d_norm': True, 'randomize': False, 'min_evals': 0, 'max_evals': None, 'plot': False, 'mp_fit': True, 'ww2x': False}\n",
      "INFO:weightwatcher:LAYER: 1 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 1 None \n",
      "INFO:weightwatcher:Randomizing Layer: 1 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 1 None \n",
      "INFO:weightwatcher:Layer id 2  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 7 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 7 None \n",
      "INFO:weightwatcher:Randomizing Layer: 7 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 7 None \n",
      "INFO:weightwatcher:Layer id 8  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 9 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 9 None \n",
      "INFO:weightwatcher:Randomizing Layer: 9 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 9 None \n",
      "INFO:weightwatcher:Layer id 10  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 11 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 11 None \n",
      "INFO:weightwatcher:Randomizing Layer: 11 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 11 None \n",
      "INFO:weightwatcher:Layer id 12  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 15 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 15 None \n",
      "INFO:weightwatcher:Randomizing Layer: 15 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 15 None \n",
      "INFO:weightwatcher:Layer id 16  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 18 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 18 None \n",
      "INFO:weightwatcher:Randomizing Layer: 18 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 18 None \n",
      "INFO:weightwatcher:Layer id 19  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 20 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 20 None \n",
      "INFO:weightwatcher:Randomizing Layer: 20 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 20 None \n",
      "INFO:weightwatcher:Layer id 21  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 22 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 22 None \n",
      "INFO:weightwatcher:Randomizing Layer: 22 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 22 None \n",
      "INFO:weightwatcher:Layer id 23  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 26 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 26 None \n",
      "INFO:weightwatcher:Randomizing Layer: 26 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 26 None \n",
      "INFO:weightwatcher:Layer id 27  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 28 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 28 None \n",
      "INFO:weightwatcher:Randomizing Layer: 28 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 28 None \n",
      "INFO:weightwatcher:Layer id 29  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 30 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 30 None \n",
      "INFO:weightwatcher:Randomizing Layer: 30 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 30 None \n",
      "INFO:weightwatcher:Layer id 31  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 35 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 35 None \n",
      "INFO:weightwatcher:Randomizing Layer: 35 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 35 None \n",
      "INFO:weightwatcher:Layer id 36  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 37 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 37 None \n",
      "INFO:weightwatcher:Randomizing Layer: 37 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 37 None \n",
      "INFO:weightwatcher:Layer id 38  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 39 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 39 None \n",
      "INFO:weightwatcher:Randomizing Layer: 39 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 39 None \n",
      "INFO:weightwatcher:Layer id 40  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 43 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 43 None \n",
      "INFO:weightwatcher:Randomizing Layer: 43 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 43 None \n",
      "INFO:weightwatcher:Layer id 44  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 46 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 46 None \n",
      "INFO:weightwatcher:Randomizing Layer: 46 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 46 None \n",
      "INFO:weightwatcher:Layer id 47  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 48 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 48 None \n",
      "INFO:weightwatcher:Randomizing Layer: 48 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 48 None \n",
      "INFO:weightwatcher:Layer id 49  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 50 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 50 None \n",
      "INFO:weightwatcher:Randomizing Layer: 50 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 50 None \n",
      "INFO:weightwatcher:Layer id 51  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 54 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 54 None \n",
      "INFO:weightwatcher:Randomizing Layer: 54 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 54 None \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:weightwatcher:Layer id 55  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 56 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 56 None \n",
      "INFO:weightwatcher:Randomizing Layer: 56 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 56 None \n",
      "INFO:weightwatcher:Layer id 57  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 58 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 58 None \n",
      "INFO:weightwatcher:Randomizing Layer: 58 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 58 None \n",
      "INFO:weightwatcher:Layer id 59  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 62 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 62 None \n",
      "INFO:weightwatcher:Randomizing Layer: 62 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 62 None \n",
      "INFO:weightwatcher:Layer id 63  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 64 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 64 None \n",
      "INFO:weightwatcher:Randomizing Layer: 64 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 64 None \n",
      "INFO:weightwatcher:Layer id 65  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 66 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 66 None \n",
      "INFO:weightwatcher:Randomizing Layer: 66 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 66 None \n",
      "INFO:weightwatcher:Layer id 67  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 71 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 71 None \n",
      "INFO:weightwatcher:Randomizing Layer: 71 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 71 None \n",
      "INFO:weightwatcher:Layer id 72  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 73 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 73 None \n",
      "INFO:weightwatcher:Randomizing Layer: 73 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 73 None \n",
      "INFO:weightwatcher:Layer id 74  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 75 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 75 None \n",
      "INFO:weightwatcher:Randomizing Layer: 75 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 75 None \n",
      "INFO:weightwatcher:Layer id 76  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 79 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 79 None \n",
      "INFO:weightwatcher:Randomizing Layer: 79 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 79 None \n",
      "INFO:weightwatcher:Layer id 80  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 82 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 82 None \n",
      "INFO:weightwatcher:Randomizing Layer: 82 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 82 None \n",
      "INFO:weightwatcher:Layer id 83  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 84 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 84 None \n",
      "INFO:weightwatcher:Randomizing Layer: 84 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 84 None \n",
      "INFO:weightwatcher:Layer id 85  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 86 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 86 None \n",
      "INFO:weightwatcher:Randomizing Layer: 86 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 86 None \n",
      "INFO:weightwatcher:Layer id 87  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 90 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 90 None \n",
      "INFO:weightwatcher:Randomizing Layer: 90 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 90 None \n",
      "INFO:weightwatcher:Layer id 91  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 92 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 92 None \n",
      "INFO:weightwatcher:Randomizing Layer: 92 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 92 None \n",
      "INFO:weightwatcher:Layer id 93  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 94 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 94 None \n",
      "INFO:weightwatcher:Randomizing Layer: 94 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 94 None \n",
      "INFO:weightwatcher:Layer id 95  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 98 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 98 None \n",
      "INFO:weightwatcher:Randomizing Layer: 98 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 98 None \n",
      "INFO:weightwatcher:Layer id 99  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 100 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 100 None \n",
      "INFO:weightwatcher:Randomizing Layer: 100 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 100 None \n",
      "INFO:weightwatcher:Layer id 101  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 102 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 102 None \n",
      "INFO:weightwatcher:Randomizing Layer: 102 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 102 None \n",
      "INFO:weightwatcher:Layer id 103  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 106 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 106 None \n",
      "INFO:weightwatcher:Randomizing Layer: 106 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 106 None \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:weightwatcher:Layer id 107  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 108 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 108 None \n",
      "INFO:weightwatcher:Randomizing Layer: 108 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 108 None \n",
      "INFO:weightwatcher:Layer id 109  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 110 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 110 None \n",
      "INFO:weightwatcher:Randomizing Layer: 110 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 110 None \n",
      "INFO:weightwatcher:Layer id 111  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 114 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 114 None \n",
      "INFO:weightwatcher:Randomizing Layer: 114 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 114 None \n",
      "INFO:weightwatcher:Layer id 115  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 116 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 116 None \n",
      "INFO:weightwatcher:Randomizing Layer: 116 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 116 None \n",
      "INFO:weightwatcher:Layer id 117  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 118 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 118 None \n",
      "INFO:weightwatcher:Randomizing Layer: 118 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 118 None \n",
      "INFO:weightwatcher:Layer id 119  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 122 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 122 None \n",
      "INFO:weightwatcher:Randomizing Layer: 122 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 122 None \n",
      "INFO:weightwatcher:Layer id 123  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 124 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 124 None \n",
      "INFO:weightwatcher:Randomizing Layer: 124 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 124 None \n",
      "INFO:weightwatcher:Layer id 125  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 126 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 126 None \n",
      "INFO:weightwatcher:Randomizing Layer: 126 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 126 None \n",
      "INFO:weightwatcher:Layer id 127  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 130 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 130 None \n",
      "INFO:weightwatcher:Randomizing Layer: 130 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 130 None \n",
      "INFO:weightwatcher:Layer id 131  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 132 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 132 None \n",
      "INFO:weightwatcher:Randomizing Layer: 132 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 132 None \n",
      "INFO:weightwatcher:Layer id 133  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 134 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 134 None \n",
      "INFO:weightwatcher:Randomizing Layer: 134 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 134 None \n",
      "INFO:weightwatcher:Layer id 135  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 138 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 138 None \n",
      "INFO:weightwatcher:Randomizing Layer: 138 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 138 None \n",
      "INFO:weightwatcher:Layer id 139  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 140 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 140 None \n",
      "INFO:weightwatcher:Randomizing Layer: 140 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 140 None \n",
      "INFO:weightwatcher:Layer id 141  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 142 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 142 None \n",
      "INFO:weightwatcher:Randomizing Layer: 142 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 142 None \n",
      "INFO:weightwatcher:Layer id 143  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 146 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 146 None \n",
      "INFO:weightwatcher:Randomizing Layer: 146 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 146 None \n",
      "INFO:weightwatcher:Layer id 147  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 148 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 148 None \n",
      "INFO:weightwatcher:Randomizing Layer: 148 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 148 None \n",
      "INFO:weightwatcher:Layer id 149  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 150 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 150 None \n",
      "INFO:weightwatcher:Randomizing Layer: 150 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 150 None \n",
      "INFO:weightwatcher:Layer id 151  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 154 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 154 None \n",
      "INFO:weightwatcher:Randomizing Layer: 154 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 154 None \n",
      "INFO:weightwatcher:Layer id 155  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 156 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 156 None \n",
      "INFO:weightwatcher:Randomizing Layer: 156 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:weightwatcher:MP Fitting Random layer: 156 None \n",
      "INFO:weightwatcher:Layer id 157  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 158 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 158 None \n",
      "INFO:weightwatcher:Randomizing Layer: 158 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 158 None \n",
      "INFO:weightwatcher:Layer id 159  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 162 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 162 None \n",
      "INFO:weightwatcher:Randomizing Layer: 162 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 162 None \n",
      "INFO:weightwatcher:Layer id 163  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 164 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 164 None \n",
      "INFO:weightwatcher:Randomizing Layer: 164 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 164 None \n",
      "INFO:weightwatcher:Layer id 165  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 166 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 166 None \n",
      "INFO:weightwatcher:Randomizing Layer: 166 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 166 None \n",
      "INFO:weightwatcher:Layer id 167  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 170 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 170 None \n",
      "INFO:weightwatcher:Randomizing Layer: 170 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 170 None \n",
      "INFO:weightwatcher:Layer id 171  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 172 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 172 None \n",
      "INFO:weightwatcher:Randomizing Layer: 172 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 172 None \n",
      "INFO:weightwatcher:Layer id 173  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 174 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 174 None \n",
      "INFO:weightwatcher:Randomizing Layer: 174 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 174 None \n",
      "INFO:weightwatcher:Layer id 175  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 178 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 178 None \n",
      "INFO:weightwatcher:Randomizing Layer: 178 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 178 None \n",
      "INFO:weightwatcher:Layer id 179  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 180 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 180 None \n",
      "INFO:weightwatcher:Randomizing Layer: 180 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 180 None \n",
      "INFO:weightwatcher:Layer id 181  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 182 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 182 None \n",
      "INFO:weightwatcher:Randomizing Layer: 182 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 182 None \n",
      "INFO:weightwatcher:Layer id 183  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 186 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 186 None \n",
      "INFO:weightwatcher:Randomizing Layer: 186 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 186 None \n",
      "INFO:weightwatcher:Layer id 187  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 188 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 188 None \n",
      "INFO:weightwatcher:Randomizing Layer: 188 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 188 None \n",
      "INFO:weightwatcher:Layer id 189  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 190 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 190 None \n",
      "INFO:weightwatcher:Randomizing Layer: 190 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 190 None \n",
      "INFO:weightwatcher:Layer id 191  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 194 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 194 None \n",
      "INFO:weightwatcher:Randomizing Layer: 194 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 194 None \n",
      "INFO:weightwatcher:Layer id 195  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 196 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 196 None \n",
      "INFO:weightwatcher:Randomizing Layer: 196 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 196 None \n",
      "INFO:weightwatcher:Layer id 197  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 198 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 198 None \n",
      "INFO:weightwatcher:Randomizing Layer: 198 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 198 None \n",
      "INFO:weightwatcher:Layer id 199  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 202 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 202 None \n",
      "INFO:weightwatcher:Randomizing Layer: 202 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 202 None \n",
      "INFO:weightwatcher:Layer id 203  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 204 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 204 None \n",
      "INFO:weightwatcher:Randomizing Layer: 204 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 204 None \n",
      "INFO:weightwatcher:Layer id 205  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 206 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 206 None \n",
      "INFO:weightwatcher:Randomizing Layer: 206 None \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 206 None \n",
      "INFO:weightwatcher:Layer id 207  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 210 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 210 None \n",
      "INFO:weightwatcher:Randomizing Layer: 210 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 210 None \n",
      "INFO:weightwatcher:Layer id 211  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 212 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 212 None \n",
      "INFO:weightwatcher:Randomizing Layer: 212 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 212 None \n",
      "INFO:weightwatcher:Layer id 213  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 214 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 214 None \n",
      "INFO:weightwatcher:Randomizing Layer: 214 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 214 None \n",
      "INFO:weightwatcher:Layer id 215  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 218 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 218 None \n",
      "INFO:weightwatcher:Randomizing Layer: 218 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 218 None \n",
      "INFO:weightwatcher:Layer id 219  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 220 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 220 None \n",
      "INFO:weightwatcher:Randomizing Layer: 220 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 220 None \n",
      "INFO:weightwatcher:Layer id 221  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 222 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 222 None \n",
      "INFO:weightwatcher:Randomizing Layer: 222 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 222 None \n",
      "INFO:weightwatcher:Layer id 223  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 226 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 226 None \n",
      "INFO:weightwatcher:Randomizing Layer: 226 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 226 None \n",
      "INFO:weightwatcher:Layer id 227  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 228 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 228 None \n",
      "INFO:weightwatcher:Randomizing Layer: 228 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 228 None \n",
      "INFO:weightwatcher:Layer id 229  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 230 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 230 None \n",
      "INFO:weightwatcher:Randomizing Layer: 230 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 230 None \n",
      "INFO:weightwatcher:Layer id 231  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 234 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 234 None \n",
      "INFO:weightwatcher:Randomizing Layer: 234 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 234 None \n",
      "INFO:weightwatcher:Layer id 235  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 236 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 236 None \n",
      "INFO:weightwatcher:Randomizing Layer: 236 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 236 None \n",
      "INFO:weightwatcher:Layer id 237  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 238 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 238 None \n",
      "INFO:weightwatcher:Randomizing Layer: 238 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 238 None \n",
      "INFO:weightwatcher:Layer id 239  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 242 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 242 None \n",
      "INFO:weightwatcher:Randomizing Layer: 242 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 242 None \n",
      "INFO:weightwatcher:Layer id 243  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 244 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 244 None \n",
      "INFO:weightwatcher:Randomizing Layer: 244 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 244 None \n",
      "INFO:weightwatcher:Layer id 245  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 246 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 246 None \n",
      "INFO:weightwatcher:Randomizing Layer: 246 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 246 None \n",
      "INFO:weightwatcher:Layer id 247  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 250 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 250 None \n",
      "INFO:weightwatcher:Randomizing Layer: 250 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 250 None \n",
      "INFO:weightwatcher:Layer id 251  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 252 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 252 None \n",
      "INFO:weightwatcher:Randomizing Layer: 252 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 252 None \n",
      "INFO:weightwatcher:Layer id 253  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 254 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 254 None \n",
      "INFO:weightwatcher:Randomizing Layer: 254 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 254 None \n",
      "INFO:weightwatcher:Layer id 255  unknown type 1 layer  BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 259 8  : <class 'torch.nn.modules.conv.Conv2d'>\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:weightwatcher:MP Fitting Layer: 259 None \n",
      "INFO:weightwatcher:Randomizing Layer: 259 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 259 None \n",
      "INFO:weightwatcher:Layer id 260  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 261 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 261 None \n",
      "INFO:weightwatcher:Randomizing Layer: 261 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 261 None \n",
      "INFO:weightwatcher:Layer id 262  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 263 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 263 None \n",
      "INFO:weightwatcher:Randomizing Layer: 263 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 263 None \n",
      "INFO:weightwatcher:Layer id 264  unknown type 1 layer  BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 267 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 267 None \n",
      "INFO:weightwatcher:Randomizing Layer: 267 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 267 None \n",
      "INFO:weightwatcher:Layer id 268  unknown type 1 layer  BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 270 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 270 None \n",
      "INFO:weightwatcher:Randomizing Layer: 270 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 270 None \n",
      "INFO:weightwatcher:Layer id 271  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 272 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 272 None \n",
      "INFO:weightwatcher:Randomizing Layer: 272 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 272 None \n",
      "INFO:weightwatcher:Layer id 273  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 274 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 274 None \n",
      "INFO:weightwatcher:Randomizing Layer: 274 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 274 None \n",
      "INFO:weightwatcher:Layer id 275  unknown type 1 layer  BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 278 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 278 None \n",
      "INFO:weightwatcher:Randomizing Layer: 278 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 278 None \n",
      "INFO:weightwatcher:Layer id 279  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 280 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 280 None \n",
      "INFO:weightwatcher:Randomizing Layer: 280 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 280 None \n",
      "INFO:weightwatcher:Layer id 281  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 282 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 282 None \n",
      "INFO:weightwatcher:Randomizing Layer: 282 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 282 None \n",
      "INFO:weightwatcher:Layer id 283  unknown type 1 layer  BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 286 2  : <class 'torch.nn.modules.linear.Linear'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 286 None \n",
      "INFO:weightwatcher:Randomizing Layer: 286 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 286 None \n",
      "INFO:weightwatcher:\n",
      "\n",
      "python      version 3.8.5 (default, Sep  4 2020, 02:22:02) \n",
      "[Clang 10.0.0 ]\n",
      "numpy       version 1.18.5\n",
      "tensforflow version 2.3.0\n",
      "keras       version 2.4.0\n",
      "INFO:weightwatcher:params {'glorot_fix': False, 'normalize': False, 'conv2d_norm': True, 'randomize': False, 'min_evals': 0, 'max_evals': None, 'plot': False, 'mp_fit': True, 'ww2x': False}\n",
      "INFO:weightwatcher:LAYER: 1 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 1 None \n",
      "INFO:weightwatcher:Randomizing Layer: 1 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 1 None \n",
      "INFO:weightwatcher:Layer id 2  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 7 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 7 None \n",
      "INFO:weightwatcher:Randomizing Layer: 7 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 7 None \n",
      "INFO:weightwatcher:Layer id 8  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 9 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 9 None \n",
      "INFO:weightwatcher:Randomizing Layer: 9 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 9 None \n",
      "INFO:weightwatcher:Layer id 10  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 11 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 11 None \n",
      "INFO:weightwatcher:Randomizing Layer: 11 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 11 None \n",
      "INFO:weightwatcher:Layer id 12  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 15 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 15 None \n",
      "INFO:weightwatcher:Randomizing Layer: 15 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 15 None \n",
      "INFO:weightwatcher:Layer id 16  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 18 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 18 None \n",
      "INFO:weightwatcher:Randomizing Layer: 18 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 18 None \n",
      "INFO:weightwatcher:Layer id 19  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 20 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 20 None \n",
      "INFO:weightwatcher:Randomizing Layer: 20 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 20 None \n",
      "INFO:weightwatcher:Layer id 21  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 22 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 22 None \n",
      "INFO:weightwatcher:Randomizing Layer: 22 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:weightwatcher:MP Fitting Random layer: 22 None \n",
      "INFO:weightwatcher:Layer id 23  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 26 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 26 None \n",
      "INFO:weightwatcher:Randomizing Layer: 26 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 26 None \n",
      "INFO:weightwatcher:Layer id 27  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 28 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 28 None \n",
      "INFO:weightwatcher:Randomizing Layer: 28 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 28 None \n",
      "INFO:weightwatcher:Layer id 29  unknown type 1 layer  BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 30 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 30 None \n",
      "INFO:weightwatcher:Randomizing Layer: 30 None \n",
      "INFO:weightwatcher:generating 5 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 30 None \n",
      "INFO:weightwatcher:Layer id 31  unknown type 1 layer  BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 35 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 35 None \n",
      "INFO:weightwatcher:Randomizing Layer: 35 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 35 None \n",
      "INFO:weightwatcher:Layer id 36  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 37 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 37 None \n",
      "INFO:weightwatcher:Randomizing Layer: 37 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 37 None \n",
      "INFO:weightwatcher:Layer id 38  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 39 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 39 None \n",
      "INFO:weightwatcher:Randomizing Layer: 39 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 39 None \n",
      "INFO:weightwatcher:Layer id 40  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 43 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 43 None \n",
      "INFO:weightwatcher:Randomizing Layer: 43 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 43 None \n",
      "INFO:weightwatcher:Layer id 44  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 46 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 46 None \n",
      "INFO:weightwatcher:Randomizing Layer: 46 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 46 None \n",
      "INFO:weightwatcher:Layer id 47  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 48 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 48 None \n",
      "INFO:weightwatcher:Randomizing Layer: 48 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 48 None \n",
      "INFO:weightwatcher:Layer id 49  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 50 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 50 None \n",
      "INFO:weightwatcher:Randomizing Layer: 50 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 50 None \n",
      "INFO:weightwatcher:Layer id 51  unknown type 1 layer  BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 54 8  : <class 'torch.nn.modules.conv.Conv2d'>\n",
      "INFO:weightwatcher:MP Fitting Layer: 54 None \n",
      "INFO:weightwatcher:Randomizing Layer: 54 None \n",
      "INFO:weightwatcher:generating 1 replicas for each W of the random eigenvalues\n",
      "INFO:weightwatcher:MP Fitting Random layer: 54 None \n",
      "INFO:weightwatcher:Layer id 55  unknown type 1 layer  BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "INFO:weightwatcher:LAYER: 56 8  : <class 'torch.nn.modules.conv.Conv2d'>\n"
     ]
    }
   ],
   "source": [
    "all_details = []\n",
    "all_summaries = []\n",
    "for im, name in enumerate(all_names):\n",
    "    watcher = ww.WeightWatcher(model=all_models[im])\n",
    "    details = watcher.analyze(mp_fit=True)\n",
    "    summary =  watcher.get_summary()\n",
    "    all_summaries.append(summary)\n",
    "    all_details.append(details)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.rcParams.update({'font.size': 16})\n",
    "from pylab import rcParams\n",
    "rcParams['figure.figsize'] = 10,10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "params = {'legend.fontsize': 'x-large',\n",
    "          'figure.figsize': (10,10),\n",
    "         'axes.labelsize': 'x-large',\n",
    "         'axes.titlesize':'xx-large',\n",
    "         'xtick.labelsize':'xx-large',\n",
    "         'ytick.labelsize':'xx-large'}\n",
    "plt.rcParams.update(params)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "MP.plot_all_metrics_vs_test_accuracies( \n",
    "    series_name, all_names, colors, all_summaries, all_details, top1_errors)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "first_n_last_ids = [0, len(all_details)-1]\n",
    "first_n_last_ids"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "MP.plot_all_metric_histograms( \\\n",
    "    series_name, all_names, colors, all_summaries, all_details, first_n_last_ids)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "MP.plot_all_metric_vs_depth( \\\n",
    "    series_name, all_names, colors, all_summaries, all_details, first_n_last_ids)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}