train_pointnet_pytorch.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "source": [
    "# Setup"
   ],
   "metadata": {
    "id": "XywsicDnKIUk"
   }
  },
  {
   "cell_type": "code",
   "source": [
    "from google.colab import drive\n",
    "drive.mount('/content/drive')"
   ],
   "metadata": {
    "id": "n_D1dZRCs-_Z",
    "outputId": "6314d6b2-6f63-4469-b578-576874ce8d59",
    "colab": {
     "base_uri": "https://localhost:8080/"
    }
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Mounted at /content/drive\n"
     ]
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [
    "!ls"
   ],
   "metadata": {
    "id": "h3AIlJwQs_qW",
    "outputId": "4d03c08a-4a48-45d1-e3b2-c6dc5463ac31",
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "pycharm": {
     "is_executing": true
    }
   },
   "execution_count": null,
   "outputs": []
  },
  {
   "cell_type": "code",
   "source": [
    "# TODO don't forget to change runtime type to GPT"
   ],
   "metadata": {
    "id": "sRg7rsv95V2v"
   },
   "execution_count": null,
   "outputs": []
  },
  {
   "cell_type": "code",
   "source": [
    "# TODO install our repo\n",
    "!git clone https://github.com/artem-tkachuk/cs182-project-pointnet.git drive/MyDrive/cs182-pointnet-homework"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "KwsVWa2f40dX",
    "outputId": "102ce065-6ada-40bc-ef71-b635fba94865"
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Cloning into 'drive/MyDrive/cs182-pointnet-homework'...\n",
      "remote: Enumerating objects: 42, done.\u001B[K\n",
      "remote: Counting objects: 100% (42/42), done.\u001B[K\n",
      "remote: Compressing objects: 100% (34/34), done.\u001B[K\n",
      "remote: Total 42 (delta 1), reused 36 (delta 1), pack-reused 0\u001B[K\n",
      "Unpacking objects: 100% (42/42), 303.70 KiB | 7.99 MiB/s, done.\n"
     ]
    }
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Obtaining file:///Users/xlsior/Library/Mobile%20Documents/com~apple~CloudDocs/Berkeley/Semesters/Spring%202023/CS182/Project/cs182-project-pointnet\r\n",
      "  Preparing metadata (setup.py) ... \u001B[?25ldone\r\n",
      "\u001B[?25hRequirement already satisfied: torch in /Users/xlsior/opt/anaconda3/envs/cs182-project-pointnet/lib/python3.8/site-packages (from pointnet==0.0.1) (1.12.1)\r\n",
      "Requirement already satisfied: tqdm in /Users/xlsior/opt/anaconda3/envs/cs182-project-pointnet/lib/python3.8/site-packages (from pointnet==0.0.1) (4.65.0)\r\n",
      "Requirement already satisfied: plyfile in /Users/xlsior/opt/anaconda3/envs/cs182-project-pointnet/lib/python3.8/site-packages (from pointnet==0.0.1) (0.9)\r\n",
      "Requirement already satisfied: numpy>=1.17 in /Users/xlsior/opt/anaconda3/envs/cs182-project-pointnet/lib/python3.8/site-packages (from plyfile->pointnet==0.0.1) (1.23.5)\r\n",
      "Requirement already satisfied: typing_extensions in /Users/xlsior/opt/anaconda3/envs/cs182-project-pointnet/lib/python3.8/site-packages (from torch->pointnet==0.0.1) (4.5.0)\r\n",
      "Installing collected packages: pointnet\r\n",
      "  Attempting uninstall: pointnet\r\n",
      "    Found existing installation: pointnet 0.0.1\r\n",
      "    Uninstalling pointnet-0.0.1:\r\n",
      "      Successfully uninstalled pointnet-0.0.1\r\n",
      "  Running setup.py develop for pointnet\r\n",
      "Successfully installed pointnet-0.0.1\r\n"
     ]
    }
   ],
   "source": [
    "!pip install -e ."
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "source": [
    "!cd drive/MyDrive/cs182-pointnet-homework && pip install -e ."
   ],
   "metadata": {
    "id": "XGXDsB9Bt7Ip",
    "outputId": "0184ae94-592e-44b3-e3e5-a8858c9bf6eb",
    "colab": {
     "base_uri": "https://localhost:8080/"
    }
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
      "Obtaining file:///content/drive/MyDrive/cs182-pointnet-homework\n",
      "  Preparing metadata (setup.py) ... \u001B[?25l\u001B[?25hdone\n",
      "Requirement already satisfied: torch in /usr/local/lib/python3.9/dist-packages (from pointnet==0.0.1) (2.0.0+cu118)\n",
      "Requirement already satisfied: tqdm in /usr/local/lib/python3.9/dist-packages (from pointnet==0.0.1) (4.65.0)\n",
      "Collecting plyfile\n",
      "  Downloading plyfile-0.9-py3-none-any.whl (30 kB)\n",
      "Requirement already satisfied: numpy>=1.17 in /usr/local/lib/python3.9/dist-packages (from plyfile->pointnet==0.0.1) (1.22.4)\n",
      "Requirement already satisfied: networkx in /usr/local/lib/python3.9/dist-packages (from torch->pointnet==0.0.1) (3.1)\n",
      "Requirement already satisfied: typing-extensions in /usr/local/lib/python3.9/dist-packages (from torch->pointnet==0.0.1) (4.5.0)\n",
      "Requirement already satisfied: filelock in /usr/local/lib/python3.9/dist-packages (from torch->pointnet==0.0.1) (3.11.0)\n",
      "Requirement already satisfied: jinja2 in /usr/local/lib/python3.9/dist-packages (from torch->pointnet==0.0.1) (3.1.2)\n",
      "Requirement already satisfied: triton==2.0.0 in /usr/local/lib/python3.9/dist-packages (from torch->pointnet==0.0.1) (2.0.0)\n",
      "Requirement already satisfied: sympy in /usr/local/lib/python3.9/dist-packages (from torch->pointnet==0.0.1) (1.11.1)\n",
      "Requirement already satisfied: cmake in /usr/local/lib/python3.9/dist-packages (from triton==2.0.0->torch->pointnet==0.0.1) (3.25.2)\n",
      "Requirement already satisfied: lit in /usr/local/lib/python3.9/dist-packages (from triton==2.0.0->torch->pointnet==0.0.1) (16.0.1)\n",
      "Requirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.9/dist-packages (from jinja2->torch->pointnet==0.0.1) (2.1.2)\n",
      "Requirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.9/dist-packages (from sympy->torch->pointnet==0.0.1) (1.3.0)\n",
      "Installing collected packages: plyfile, pointnet\n",
      "  Running setup.py develop for pointnet\n",
      "Successfully installed plyfile-0.9 pointnet-0.0.1\n"
     ]
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [
    "!bash scripts/download.sh"
   ],
   "metadata": {
    "id": "e8L6o85Vuflh"
   },
   "execution_count": 23,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "scripts/download.sh: line 1: realpath: command not found\r\n",
      "usage: dirname string [...]\r\n",
      "--2023-04-19 18:47:24--  https://shapenet.cs.stanford.edu/ericyi/shapenetcore_partanno_segmentation_benchmark_v0.zip\r\n",
      "Resolving shapenet.cs.stanford.edu (shapenet.cs.stanford.edu)... 171.67.77.19\r\n",
      "Connecting to shapenet.cs.stanford.edu (shapenet.cs.stanford.edu)|171.67.77.19|:443... connected.\r\n",
      "WARNING: cannot verify shapenet.cs.stanford.edu's certificate, issued by ‘CN=InCommon RSA Server CA,OU=InCommon,O=Internet2,L=Ann Arbor,ST=MI,C=US’:\r\n",
      "  Issued certificate has expired.\r\n",
      "HTTP request sent, awaiting response... 200 OK\r\n",
      "Length: 666265910 (635M) [application/zip]\r\n",
      "shapenetcore_partanno_segmentation_benchmark_v0.zip: Read-only file system\r\n",
      "\r\n",
      "Cannot write to ‘shapenetcore_partanno_segmentation_benchmark_v0.zip’ (Success).\r\n",
      "unzip:  cannot find or open shapenetcore_partanno_segmentation_benchmark_v0.zip, shapenetcore_partanno_segmentation_benchmark_v0.zip.zip or shapenetcore_partanno_segmentation_benchmark_v0.zip.ZIP.\r\n",
      "rm: shapenetcore_partanno_segmentation_benchmark_v0.zip: No such file or directory\r\n",
      "/Users/xlsior/Library/Mobile Documents/com~apple~CloudDocs/Berkeley/Semesters/Spring 2023/CS182/Project/cs182-project-pointnet\r\n"
     ]
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [
    "!cd drive/MyDrive/cs182-pointnet-homework/scripts && bash build.sh"
   ],
   "metadata": {
    "id": "W6vLd_yFuBWz",
    "outputId": "b8b89dc1-abbd-4ede-f5dd-02f6c2640bbe",
    "colab": {
     "base_uri": "https://localhost:8080/"
    }
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "/content/drive/MyDrive/cs182-pointnet-homework/scripts\n"
     ]
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [],
   "metadata": {
    "id": "gudOotqG-3Qw"
   },
   "execution_count": null,
   "outputs": []
  },
  {
   "cell_type": "code",
   "source": [
    "showpoints(np.random.randn(2500, 3))"
   ],
   "metadata": {
    "id": "WluXWo6c_Dqj"
   },
   "execution_count": null,
   "outputs": []
  },
  {
   "cell_type": "code",
   "source": [
    "from utils.show3d_balls import showpoints\n",
    "\n",
    "showpoints(np.random.randn(2500, 3))"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "q2VLr7xD8ViG",
    "outputId": "2bc66a5e-d5fe-4096-a330-908375c5ca38"
   },
   "execution_count": 21,
   "outputs": [
    {
     "ename": "OSError",
     "evalue": "no file with expected extension",
     "output_type": "error",
     "traceback": [
      "\u001B[0;31m---------------------------------------------------------------------------\u001B[0m",
      "\u001B[0;31mOSError\u001B[0m                                   Traceback (most recent call last)",
      "Cell \u001B[0;32mIn[21], line 1\u001B[0m\n\u001B[0;32m----> 1\u001B[0m \u001B[38;5;28;01mfrom\u001B[39;00m \u001B[38;5;21;01mutils\u001B[39;00m\u001B[38;5;21;01m.\u001B[39;00m\u001B[38;5;21;01mshow3d_balls\u001B[39;00m \u001B[38;5;28;01mimport\u001B[39;00m showpoints\n\u001B[1;32m      3\u001B[0m showpoints(np\u001B[38;5;241m.\u001B[39mrandom\u001B[38;5;241m.\u001B[39mrandn(\u001B[38;5;241m2500\u001B[39m, \u001B[38;5;241m3\u001B[39m))\n",
      "File \u001B[0;32m~/Library/Mobile Documents/com~apple~CloudDocs/Berkeley/Semesters/Spring 2023/CS182/Project/cs182-project-pointnet/utils/show3d_balls.py:25\u001B[0m\n\u001B[1;32m     22\u001B[0m cv2\u001B[38;5;241m.\u001B[39mmoveWindow(\u001B[38;5;124m'\u001B[39m\u001B[38;5;124mshow3d\u001B[39m\u001B[38;5;124m'\u001B[39m, \u001B[38;5;241m0\u001B[39m, \u001B[38;5;241m0\u001B[39m)\n\u001B[1;32m     23\u001B[0m cv2\u001B[38;5;241m.\u001B[39msetMouseCallback(\u001B[38;5;124m'\u001B[39m\u001B[38;5;124mshow3d\u001B[39m\u001B[38;5;124m'\u001B[39m, onmouse)\n\u001B[0;32m---> 25\u001B[0m dll \u001B[38;5;241m=\u001B[39m \u001B[43mnp\u001B[49m\u001B[38;5;241;43m.\u001B[39;49m\u001B[43mctypeslib\u001B[49m\u001B[38;5;241;43m.\u001B[39;49m\u001B[43mload_library\u001B[49m\u001B[43m(\u001B[49m\u001B[38;5;124;43m'\u001B[39;49m\u001B[38;5;124;43mrender_balls_so\u001B[39;49m\u001B[38;5;124;43m'\u001B[39;49m\u001B[43m,\u001B[49m\u001B[43m \u001B[49m\u001B[38;5;124;43m'\u001B[39;49m\u001B[38;5;124;43m.\u001B[39;49m\u001B[38;5;124;43m'\u001B[39;49m\u001B[43m)\u001B[49m\n\u001B[1;32m     28\u001B[0m \u001B[38;5;28;01mdef\u001B[39;00m \u001B[38;5;21mshowpoints\u001B[39m(xyz, c_gt\u001B[38;5;241m=\u001B[39m\u001B[38;5;28;01mNone\u001B[39;00m, c_pred\u001B[38;5;241m=\u001B[39m\u001B[38;5;28;01mNone\u001B[39;00m, waittime\u001B[38;5;241m=\u001B[39m\u001B[38;5;241m0\u001B[39m,\n\u001B[1;32m     29\u001B[0m                showrot\u001B[38;5;241m=\u001B[39m\u001B[38;5;28;01mFalse\u001B[39;00m, magnifyBlue\u001B[38;5;241m=\u001B[39m\u001B[38;5;241m0\u001B[39m, freezerot\u001B[38;5;241m=\u001B[39m\u001B[38;5;28;01mFalse\u001B[39;00m, background\u001B[38;5;241m=\u001B[39m(\u001B[38;5;241m0\u001B[39m, \u001B[38;5;241m0\u001B[39m, \u001B[38;5;241m0\u001B[39m),\n\u001B[1;32m     30\u001B[0m                normalizecolor\u001B[38;5;241m=\u001B[39m\u001B[38;5;28;01mTrue\u001B[39;00m, ballradius\u001B[38;5;241m=\u001B[39m\u001B[38;5;241m10\u001B[39m):\n\u001B[1;32m     31\u001B[0m     \u001B[38;5;28;01mglobal\u001B[39;00m showsz, mousex, mousey, zoom, changed\n",
      "File \u001B[0;32m~/opt/anaconda3/envs/cs182-project-pointnet/lib/python3.8/site-packages/numpy/ctypeslib.py:163\u001B[0m, in \u001B[0;36mload_library\u001B[0;34m(libname, loader_path)\u001B[0m\n\u001B[1;32m    161\u001B[0m             \u001B[38;5;28;01mraise\u001B[39;00m\n\u001B[1;32m    162\u001B[0m \u001B[38;5;66;03m## if no successful return in the libname_ext loop:\u001B[39;00m\n\u001B[0;32m--> 163\u001B[0m \u001B[38;5;28;01mraise\u001B[39;00m \u001B[38;5;167;01mOSError\u001B[39;00m(\u001B[38;5;124m\"\u001B[39m\u001B[38;5;124mno file with expected extension\u001B[39m\u001B[38;5;124m\"\u001B[39m)\n",
      "\u001B[0;31mOSError\u001B[0m: no file with expected extension"
     ]
    }
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "# Classification training"
   ],
   "metadata": {
    "id": "a-zVW0CWJe0Q"
   }
  },
  {
   "cell_type": "code",
   "source": [
    "!python /content/drive/MyDrive/cs182-pointnet-homework/utils/train_classification.py --dataset /content/drive/MyDrive/cs182-pointnet-homework/shapenetcore_partanno_segmentation_benchmark_v0/ --nepoch=2 --dataset_type shapenet --feature_transform"
   ],
   "metadata": {
    "id": "H5swgN9IukCd",
    "outputId": "b24cc5f0-e1fe-4980-a624-c6d49e39646f",
    "colab": {
     "base_uri": "https://localhost:8080/"
    }
   },
   "execution_count": 2,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Namespace(batchSize=32, num_points=2500, workers=4, nepoch=2, outf='cls', model='', dataset='/content/drive/MyDrive/cs182-pointnet-homework/shapenetcore_partanno_segmentation_benchmark_v0/', dataset_type='shapenet', feature_transform=True)\n",
      "Random Seed:  3419\n",
      "{'Airplane': 0, 'Bag': 1, 'Cap': 2, 'Car': 3, 'Chair': 4, 'Earphone': 5, 'Guitar': 6, 'Knife': 7, 'Lamp': 8, 'Laptop': 9, 'Motorbike': 10, 'Mug': 11, 'Pistol': 12, 'Rocket': 13, 'Skateboard': 14, 'Table': 15}\n",
      "{'Airplane': 4, 'Bag': 2, 'Cap': 2, 'Car': 4, 'Chair': 4, 'Earphone': 3, 'Guitar': 3, 'Knife': 2, 'Lamp': 4, 'Laptop': 2, 'Motorbike': 6, 'Mug': 2, 'Pistol': 3, 'Rocket': 3, 'Skateboard': 3, 'Table': 3} 4\n",
      "{'Airplane': 0, 'Bag': 1, 'Cap': 2, 'Car': 3, 'Chair': 4, 'Earphone': 5, 'Guitar': 6, 'Knife': 7, 'Lamp': 8, 'Laptop': 9, 'Motorbike': 10, 'Mug': 11, 'Pistol': 12, 'Rocket': 13, 'Skateboard': 14, 'Table': 15}\n",
      "{'Airplane': 4, 'Bag': 2, 'Cap': 2, 'Car': 4, 'Chair': 4, 'Earphone': 3, 'Guitar': 3, 'Knife': 2, 'Lamp': 4, 'Laptop': 2, 'Motorbike': 6, 'Mug': 2, 'Pistol': 3, 'Rocket': 3, 'Skateboard': 3, 'Table': 3} 4\n",
      "/usr/local/lib/python3.9/dist-packages/torch/utils/data/dataloader.py:561: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
      "  warnings.warn(_create_warning_msg(\n",
      "12137 2874\n",
      "classes 16\n",
      "/usr/local/lib/python3.9/dist-packages/torch/optim/lr_scheduler.py:139: UserWarning: Detected call of `lr_scheduler.step()` before `optimizer.step()`. In PyTorch 1.1.0 and later, you should call them in the opposite order: `optimizer.step()` before `lr_scheduler.step()`.  Failure to do this will result in PyTorch skipping the first value of the learning rate schedule. See more details at https://pytorch.org/docs/stable/optim.html#how-to-adjust-learning-rate\n",
      "  warnings.warn(\"Detected call of `lr_scheduler.step()` before `optimizer.step()`. \"\n",
      "[0: 0/379] train loss: 2.984598 accuracy: 0.031250\n",
      "[0: 0/379] \u001B[94mtest\u001B[0m loss: 2.759773 accuracy: 0.187500\n",
      "[0: 1/379] train loss: 2.818573 accuracy: 0.125000\n",
      "[0: 2/379] train loss: 2.672199 accuracy: 0.312500\n",
      "[0: 3/379] train loss: 2.345363 accuracy: 0.343750\n",
      "[0: 4/379] train loss: 2.357948 accuracy: 0.343750\n",
      "[0: 5/379] train loss: 2.292125 accuracy: 0.375000\n",
      "[0: 6/379] train loss: 2.171897 accuracy: 0.500000\n",
      "[0: 7/379] train loss: 1.735300 accuracy: 0.656250\n",
      "[0: 8/379] train loss: 1.877696 accuracy: 0.531250\n",
      "[0: 9/379] train loss: 1.682061 accuracy: 0.562500\n",
      "[0: 10/379] train loss: 1.475163 accuracy: 0.718750\n",
      "[0: 10/379] \u001B[94mtest\u001B[0m loss: 2.617384 accuracy: 0.156250\n",
      "[0: 11/379] train loss: 1.859232 accuracy: 0.531250\n",
      "[0: 12/379] train loss: 1.794622 accuracy: 0.500000\n",
      "[0: 13/379] train loss: 1.363952 accuracy: 0.687500\n",
      "[0: 14/379] train loss: 1.819497 accuracy: 0.500000\n",
      "[0: 15/379] train loss: 1.595186 accuracy: 0.625000\n",
      "[0: 16/379] train loss: 0.928573 accuracy: 0.875000\n",
      "[0: 17/379] train loss: 1.235934 accuracy: 0.718750\n",
      "[0: 18/379] train loss: 1.582687 accuracy: 0.468750\n",
      "[0: 19/379] train loss: 1.196744 accuracy: 0.718750\n",
      "[0: 20/379] train loss: 1.561756 accuracy: 0.593750\n",
      "[0: 20/379] \u001B[94mtest\u001B[0m loss: 3.609374 accuracy: 0.187500\n",
      "[0: 21/379] train loss: 1.198104 accuracy: 0.750000\n",
      "[0: 22/379] train loss: 1.271722 accuracy: 0.750000\n",
      "[0: 23/379] train loss: 1.222545 accuracy: 0.625000\n",
      "[0: 24/379] train loss: 1.250868 accuracy: 0.687500\n",
      "[0: 25/379] train loss: 1.376745 accuracy: 0.687500\n",
      "[0: 26/379] train loss: 1.267722 accuracy: 0.687500\n",
      "[0: 27/379] train loss: 1.555996 accuracy: 0.531250\n",
      "[0: 28/379] train loss: 0.882599 accuracy: 0.781250\n",
      "[0: 29/379] train loss: 0.639545 accuracy: 0.875000\n",
      "[0: 30/379] train loss: 1.229346 accuracy: 0.656250\n",
      "[0: 30/379] \u001B[94mtest\u001B[0m loss: 2.909153 accuracy: 0.187500\n",
      "[0: 31/379] train loss: 0.905606 accuracy: 0.718750\n",
      "[0: 32/379] train loss: 1.010792 accuracy: 0.718750\n",
      "[0: 33/379] train loss: 1.138740 accuracy: 0.812500\n",
      "[0: 34/379] train loss: 0.823144 accuracy: 0.750000\n",
      "[0: 35/379] train loss: 0.582907 accuracy: 0.843750\n",
      "[0: 36/379] train loss: 0.981272 accuracy: 0.781250\n",
      "[0: 37/379] train loss: 1.110787 accuracy: 0.718750\n",
      "[0: 38/379] train loss: 1.189903 accuracy: 0.656250\n",
      "[0: 39/379] train loss: 1.482787 accuracy: 0.625000\n",
      "[0: 40/379] train loss: 0.842636 accuracy: 0.750000\n",
      "[0: 40/379] \u001B[94mtest\u001B[0m loss: 2.038788 accuracy: 0.250000\n",
      "[0: 41/379] train loss: 1.463929 accuracy: 0.500000\n",
      "[0: 42/379] train loss: 1.162874 accuracy: 0.656250\n",
      "[0: 43/379] train loss: 0.968038 accuracy: 0.718750\n",
      "[0: 44/379] train loss: 0.836108 accuracy: 0.718750\n",
      "[0: 45/379] train loss: 0.717949 accuracy: 0.875000\n",
      "[0: 46/379] train loss: 1.434099 accuracy: 0.656250\n",
      "[0: 47/379] train loss: 0.954461 accuracy: 0.781250\n",
      "[0: 48/379] train loss: 0.716426 accuracy: 0.812500\n",
      "[0: 49/379] train loss: 0.794396 accuracy: 0.812500\n",
      "[0: 50/379] train loss: 0.708686 accuracy: 0.843750\n",
      "[0: 50/379] \u001B[94mtest\u001B[0m loss: 1.480208 accuracy: 0.531250\n",
      "[0: 51/379] train loss: 1.128860 accuracy: 0.812500\n",
      "[0: 52/379] train loss: 0.707343 accuracy: 0.875000\n",
      "[0: 53/379] train loss: 0.603160 accuracy: 0.812500\n",
      "[0: 54/379] train loss: 1.322605 accuracy: 0.687500\n",
      "[0: 55/379] train loss: 0.979583 accuracy: 0.718750\n",
      "[0: 56/379] train loss: 0.612307 accuracy: 0.812500\n",
      "[0: 57/379] train loss: 0.970639 accuracy: 0.718750\n",
      "[0: 58/379] train loss: 0.722880 accuracy: 0.843750\n",
      "[0: 59/379] train loss: 0.964392 accuracy: 0.750000\n",
      "[0: 60/379] train loss: 1.259061 accuracy: 0.625000\n",
      "[0: 60/379] \u001B[94mtest\u001B[0m loss: 0.849416 accuracy: 0.843750\n",
      "[0: 61/379] train loss: 0.806752 accuracy: 0.718750\n",
      "[0: 62/379] train loss: 1.025444 accuracy: 0.687500\n",
      "[0: 63/379] train loss: 0.723789 accuracy: 0.843750\n",
      "[0: 64/379] train loss: 0.782840 accuracy: 0.843750\n",
      "[0: 65/379] train loss: 0.980262 accuracy: 0.812500\n",
      "[0: 66/379] train loss: 0.872087 accuracy: 0.781250\n",
      "[0: 67/379] train loss: 1.025059 accuracy: 0.781250\n",
      "[0: 68/379] train loss: 1.592350 accuracy: 0.500000\n",
      "[0: 69/379] train loss: 1.257849 accuracy: 0.625000\n",
      "[0: 70/379] train loss: 0.648613 accuracy: 0.843750\n",
      "[0: 70/379] \u001B[94mtest\u001B[0m loss: 0.977785 accuracy: 0.656250\n",
      "[0: 71/379] train loss: 0.720526 accuracy: 0.781250\n",
      "[0: 72/379] train loss: 0.374097 accuracy: 0.968750\n",
      "[0: 73/379] train loss: 0.783310 accuracy: 0.812500\n",
      "[0: 74/379] train loss: 0.722448 accuracy: 0.843750\n",
      "[0: 75/379] train loss: 0.478534 accuracy: 0.906250\n",
      "[0: 76/379] train loss: 0.825253 accuracy: 0.781250\n",
      "[0: 77/379] train loss: 0.677665 accuracy: 0.781250\n",
      "[0: 78/379] train loss: 0.774006 accuracy: 0.750000\n",
      "[0: 79/379] train loss: 0.938934 accuracy: 0.750000\n",
      "[0: 80/379] train loss: 1.062243 accuracy: 0.718750\n",
      "[0: 80/379] \u001B[94mtest\u001B[0m loss: 1.079353 accuracy: 0.625000\n",
      "[0: 81/379] train loss: 0.622089 accuracy: 0.875000\n",
      "[0: 82/379] train loss: 0.325030 accuracy: 0.968750\n",
      "[0: 83/379] train loss: 0.876628 accuracy: 0.750000\n",
      "[0: 84/379] train loss: 0.413872 accuracy: 0.875000\n",
      "[0: 85/379] train loss: 0.914977 accuracy: 0.843750\n",
      "[0: 86/379] train loss: 0.802256 accuracy: 0.812500\n",
      "[0: 87/379] train loss: 0.391248 accuracy: 0.937500\n",
      "[0: 88/379] train loss: 0.695519 accuracy: 0.812500\n",
      "[0: 89/379] train loss: 0.447575 accuracy: 0.875000\n",
      "[0: 90/379] train loss: 0.580214 accuracy: 0.843750\n",
      "[0: 90/379] \u001B[94mtest\u001B[0m loss: 1.136503 accuracy: 0.687500\n",
      "[0: 91/379] train loss: 0.715074 accuracy: 0.875000\n",
      "[0: 92/379] train loss: 0.741522 accuracy: 0.781250\n",
      "[0: 93/379] train loss: 0.851328 accuracy: 0.812500\n",
      "[0: 94/379] train loss: 0.551717 accuracy: 0.875000\n",
      "[0: 95/379] train loss: 0.914390 accuracy: 0.781250\n",
      "[0: 96/379] train loss: 0.655252 accuracy: 0.812500\n",
      "[0: 97/379] train loss: 0.881721 accuracy: 0.812500\n",
      "[0: 98/379] train loss: 0.658027 accuracy: 0.812500\n",
      "[0: 99/379] train loss: 0.596646 accuracy: 0.875000\n",
      "[0: 100/379] train loss: 0.620567 accuracy: 0.812500\n",
      "[0: 100/379] \u001B[94mtest\u001B[0m loss: 0.811483 accuracy: 0.781250\n",
      "[0: 101/379] train loss: 0.655632 accuracy: 0.843750\n",
      "[0: 102/379] train loss: 0.659882 accuracy: 0.843750\n",
      "[0: 103/379] train loss: 0.269969 accuracy: 0.968750\n",
      "[0: 104/379] train loss: 0.655402 accuracy: 0.875000\n",
      "[0: 105/379] train loss: 0.431203 accuracy: 0.937500\n",
      "[0: 106/379] train loss: 1.089968 accuracy: 0.718750\n",
      "[0: 107/379] train loss: 0.620970 accuracy: 0.843750\n",
      "[0: 108/379] train loss: 0.565226 accuracy: 0.875000\n",
      "[0: 109/379] train loss: 1.877333 accuracy: 0.687500\n",
      "[0: 110/379] train loss: 0.386533 accuracy: 0.906250\n",
      "[0: 110/379] \u001B[94mtest\u001B[0m loss: 1.201711 accuracy: 0.656250\n",
      "[0: 111/379] train loss: 0.532029 accuracy: 0.843750\n",
      "[0: 112/379] train loss: 0.868044 accuracy: 0.750000\n",
      "[0: 113/379] train loss: 0.681381 accuracy: 0.812500\n",
      "[0: 114/379] train loss: 0.888795 accuracy: 0.781250\n",
      "[0: 115/379] train loss: 0.468270 accuracy: 0.843750\n",
      "[0: 116/379] train loss: 0.773622 accuracy: 0.812500\n",
      "[0: 117/379] train loss: 0.487127 accuracy: 0.937500\n",
      "[0: 118/379] train loss: 1.102801 accuracy: 0.750000\n",
      "[0: 119/379] train loss: 0.871684 accuracy: 0.781250\n",
      "[0: 120/379] train loss: 0.830808 accuracy: 0.812500\n",
      "[0: 120/379] \u001B[94mtest\u001B[0m loss: 0.601645 accuracy: 0.812500\n",
      "[0: 121/379] train loss: 0.543045 accuracy: 0.875000\n",
      "[0: 122/379] train loss: 1.235712 accuracy: 0.656250\n",
      "[0: 123/379] train loss: 0.763627 accuracy: 0.781250\n",
      "[0: 124/379] train loss: 0.283786 accuracy: 1.000000\n",
      "[0: 125/379] train loss: 1.062139 accuracy: 0.812500\n",
      "[0: 126/379] train loss: 0.477716 accuracy: 0.906250\n",
      "[0: 127/379] train loss: 0.742347 accuracy: 0.812500\n",
      "[0: 128/379] train loss: 0.513967 accuracy: 0.875000\n",
      "[0: 129/379] train loss: 0.662144 accuracy: 0.843750\n",
      "[0: 130/379] train loss: 0.666583 accuracy: 0.812500\n",
      "[0: 130/379] \u001B[94mtest\u001B[0m loss: 0.805385 accuracy: 0.781250\n",
      "[0: 131/379] train loss: 0.558924 accuracy: 0.843750\n",
      "[0: 132/379] train loss: 0.746685 accuracy: 0.812500\n",
      "[0: 133/379] train loss: 0.594745 accuracy: 0.843750\n",
      "[0: 134/379] train loss: 0.456958 accuracy: 0.906250\n",
      "[0: 135/379] train loss: 0.695408 accuracy: 0.750000\n",
      "[0: 136/379] train loss: 0.686423 accuracy: 0.781250\n",
      "[0: 137/379] train loss: 0.936727 accuracy: 0.687500\n",
      "[0: 138/379] train loss: 0.589227 accuracy: 0.781250\n",
      "[0: 139/379] train loss: 0.627099 accuracy: 0.843750\n",
      "[0: 140/379] train loss: 0.671130 accuracy: 0.781250\n",
      "[0: 140/379] \u001B[94mtest\u001B[0m loss: 0.894507 accuracy: 0.718750\n",
      "[0: 141/379] train loss: 0.705796 accuracy: 0.843750\n",
      "[0: 142/379] train loss: 0.717631 accuracy: 0.781250\n",
      "[0: 143/379] train loss: 0.830628 accuracy: 0.781250\n",
      "[0: 144/379] train loss: 0.689712 accuracy: 0.812500\n",
      "[0: 145/379] train loss: 0.778585 accuracy: 0.843750\n",
      "[0: 146/379] train loss: 0.488705 accuracy: 0.875000\n",
      "[0: 147/379] train loss: 0.619571 accuracy: 0.812500\n",
      "[0: 148/379] train loss: 0.609388 accuracy: 0.843750\n",
      "[0: 149/379] train loss: 0.504425 accuracy: 0.843750\n",
      "[0: 150/379] train loss: 0.606021 accuracy: 0.781250\n",
      "[0: 150/379] \u001B[94mtest\u001B[0m loss: 1.006647 accuracy: 0.718750\n",
      "[0: 151/379] train loss: 0.425299 accuracy: 0.875000\n",
      "[0: 152/379] train loss: 0.729130 accuracy: 0.843750\n",
      "[0: 153/379] train loss: 0.726121 accuracy: 0.812500\n",
      "[0: 154/379] train loss: 0.371615 accuracy: 0.906250\n",
      "[0: 155/379] train loss: 0.644586 accuracy: 0.812500\n",
      "[0: 156/379] train loss: 0.461282 accuracy: 0.906250\n",
      "[0: 157/379] train loss: 0.682554 accuracy: 0.812500\n",
      "[0: 158/379] train loss: 0.881612 accuracy: 0.781250\n",
      "[0: 159/379] train loss: 0.313529 accuracy: 0.906250\n",
      "[0: 160/379] train loss: 0.589216 accuracy: 0.781250\n",
      "[0: 160/379] \u001B[94mtest\u001B[0m loss: 0.699998 accuracy: 0.781250\n",
      "[0: 161/379] train loss: 0.586270 accuracy: 0.812500\n",
      "[0: 162/379] train loss: 0.709985 accuracy: 0.718750\n",
      "[0: 163/379] train loss: 0.571608 accuracy: 0.812500\n",
      "[0: 164/379] train loss: 0.444692 accuracy: 0.906250\n",
      "[0: 165/379] train loss: 0.805500 accuracy: 0.843750\n",
      "[0: 166/379] train loss: 1.040623 accuracy: 0.687500\n",
      "[0: 167/379] train loss: 0.629661 accuracy: 0.875000\n",
      "[0: 168/379] train loss: 0.516310 accuracy: 0.843750\n",
      "[0: 169/379] train loss: 0.840019 accuracy: 0.750000\n",
      "[0: 170/379] train loss: 0.725083 accuracy: 0.812500\n",
      "[0: 170/379] \u001B[94mtest\u001B[0m loss: 0.568128 accuracy: 0.812500\n",
      "[0: 171/379] train loss: 0.651807 accuracy: 0.781250\n",
      "[0: 172/379] train loss: 0.504102 accuracy: 0.875000\n",
      "[0: 173/379] train loss: 0.785638 accuracy: 0.781250\n",
      "[0: 174/379] train loss: 0.627932 accuracy: 0.843750\n",
      "[0: 175/379] train loss: 0.965225 accuracy: 0.656250\n",
      "[0: 176/379] train loss: 0.724842 accuracy: 0.843750\n",
      "[0: 177/379] train loss: 0.501105 accuracy: 0.843750\n",
      "[0: 178/379] train loss: 0.863284 accuracy: 0.781250\n",
      "[0: 179/379] train loss: 0.356130 accuracy: 0.906250\n",
      "[0: 180/379] train loss: 1.115882 accuracy: 0.687500\n",
      "[0: 180/379] \u001B[94mtest\u001B[0m loss: 0.639207 accuracy: 0.875000\n",
      "[0: 181/379] train loss: 0.729519 accuracy: 0.812500\n",
      "[0: 182/379] train loss: 0.479283 accuracy: 0.937500\n",
      "[0: 183/379] train loss: 1.153591 accuracy: 0.750000\n",
      "[0: 184/379] train loss: 0.609464 accuracy: 0.750000\n",
      "[0: 185/379] train loss: 0.448372 accuracy: 0.906250\n",
      "[0: 186/379] train loss: 0.472573 accuracy: 0.906250\n",
      "[0: 187/379] train loss: 0.687190 accuracy: 0.781250\n",
      "[0: 188/379] train loss: 0.480456 accuracy: 0.906250\n",
      "[0: 189/379] train loss: 0.686963 accuracy: 0.812500\n",
      "[0: 190/379] train loss: 0.471726 accuracy: 0.843750\n",
      "[0: 190/379] \u001B[94mtest\u001B[0m loss: 1.038011 accuracy: 0.812500\n",
      "[0: 191/379] train loss: 0.701316 accuracy: 0.875000\n",
      "[0: 192/379] train loss: 0.555187 accuracy: 0.843750\n",
      "[0: 193/379] train loss: 0.478193 accuracy: 0.906250\n",
      "[0: 194/379] train loss: 0.627739 accuracy: 0.875000\n",
      "[0: 195/379] train loss: 0.589913 accuracy: 0.843750\n",
      "[0: 196/379] train loss: 0.775230 accuracy: 0.781250\n",
      "[0: 197/379] train loss: 0.523389 accuracy: 0.843750\n",
      "[0: 198/379] train loss: 0.671950 accuracy: 0.843750\n",
      "[0: 199/379] train loss: 0.520278 accuracy: 0.843750\n",
      "[0: 200/379] train loss: 0.427975 accuracy: 0.875000\n",
      "[0: 200/379] \u001B[94mtest\u001B[0m loss: 0.873938 accuracy: 0.781250\n",
      "[0: 201/379] train loss: 0.485167 accuracy: 0.843750\n",
      "[0: 202/379] train loss: 0.475854 accuracy: 0.937500\n",
      "[0: 203/379] train loss: 0.616802 accuracy: 0.812500\n",
      "[0: 204/379] train loss: 0.494185 accuracy: 0.906250\n",
      "[0: 205/379] train loss: 0.567328 accuracy: 0.875000\n",
      "[0: 206/379] train loss: 0.548853 accuracy: 0.812500\n",
      "[0: 207/379] train loss: 0.609744 accuracy: 0.843750\n",
      "[0: 208/379] train loss: 0.629748 accuracy: 0.781250\n",
      "[0: 209/379] train loss: 0.594554 accuracy: 0.843750\n",
      "[0: 210/379] train loss: 0.692084 accuracy: 0.812500\n",
      "[0: 210/379] \u001B[94mtest\u001B[0m loss: 0.674484 accuracy: 0.812500\n",
      "[0: 211/379] train loss: 0.469323 accuracy: 0.875000\n",
      "[0: 212/379] train loss: 0.957646 accuracy: 0.625000\n",
      "[0: 213/379] train loss: 0.323785 accuracy: 0.968750\n",
      "[0: 214/379] train loss: 0.402486 accuracy: 0.937500\n",
      "[0: 215/379] train loss: 0.560519 accuracy: 0.812500\n",
      "[0: 216/379] train loss: 0.499034 accuracy: 0.843750\n",
      "[0: 217/379] train loss: 0.271102 accuracy: 0.968750\n",
      "[0: 218/379] train loss: 0.529804 accuracy: 0.812500\n",
      "[0: 219/379] train loss: 1.451259 accuracy: 0.656250\n",
      "[0: 220/379] train loss: 0.550548 accuracy: 0.843750\n",
      "[0: 220/379] \u001B[94mtest\u001B[0m loss: 0.436405 accuracy: 0.843750\n",
      "[0: 221/379] train loss: 0.940462 accuracy: 0.750000\n",
      "[0: 222/379] train loss: 0.680289 accuracy: 0.812500\n",
      "[0: 223/379] train loss: 0.516395 accuracy: 0.875000\n",
      "[0: 224/379] train loss: 0.967094 accuracy: 0.781250\n",
      "[0: 225/379] train loss: 0.329247 accuracy: 0.968750\n",
      "[0: 226/379] train loss: 0.654403 accuracy: 0.843750\n",
      "[0: 227/379] train loss: 0.773553 accuracy: 0.781250\n",
      "[0: 228/379] train loss: 0.936711 accuracy: 0.687500\n",
      "[0: 229/379] train loss: 0.547843 accuracy: 0.906250\n",
      "[0: 230/379] train loss: 0.867772 accuracy: 0.843750\n",
      "[0: 230/379] \u001B[94mtest\u001B[0m loss: 1.387605 accuracy: 0.656250\n",
      "[0: 231/379] train loss: 0.581306 accuracy: 0.812500\n",
      "[0: 232/379] train loss: 0.571678 accuracy: 0.812500\n",
      "[0: 233/379] train loss: 0.609251 accuracy: 0.812500\n",
      "[0: 234/379] train loss: 0.778610 accuracy: 0.812500\n",
      "[0: 235/379] train loss: 0.481772 accuracy: 0.937500\n",
      "[0: 236/379] train loss: 0.329888 accuracy: 0.906250\n",
      "[0: 237/379] train loss: 0.456451 accuracy: 0.843750\n",
      "[0: 238/379] train loss: 0.541775 accuracy: 0.843750\n",
      "[0: 239/379] train loss: 0.633542 accuracy: 0.812500\n",
      "[0: 240/379] train loss: 0.860100 accuracy: 0.812500\n",
      "[0: 240/379] \u001B[94mtest\u001B[0m loss: 0.426923 accuracy: 0.906250\n",
      "[0: 241/379] train loss: 0.608222 accuracy: 0.812500\n",
      "[0: 242/379] train loss: 0.930059 accuracy: 0.750000\n",
      "[0: 243/379] train loss: 0.854650 accuracy: 0.781250\n",
      "[0: 244/379] train loss: 0.626640 accuracy: 0.843750\n",
      "[0: 245/379] train loss: 0.436926 accuracy: 0.875000\n",
      "[0: 246/379] train loss: 0.813893 accuracy: 0.781250\n",
      "[0: 247/379] train loss: 0.646165 accuracy: 0.781250\n",
      "[0: 248/379] train loss: 0.887788 accuracy: 0.750000\n",
      "[0: 249/379] train loss: 0.970328 accuracy: 0.750000\n",
      "[0: 250/379] train loss: 0.503540 accuracy: 0.843750\n",
      "[0: 250/379] \u001B[94mtest\u001B[0m loss: 0.755730 accuracy: 0.718750\n",
      "[0: 251/379] train loss: 0.580696 accuracy: 0.875000\n",
      "[0: 252/379] train loss: 0.646345 accuracy: 0.843750\n",
      "[0: 253/379] train loss: 0.681613 accuracy: 0.843750\n",
      "[0: 254/379] train loss: 0.563163 accuracy: 0.812500\n",
      "[0: 255/379] train loss: 0.640687 accuracy: 0.875000\n",
      "[0: 256/379] train loss: 0.478087 accuracy: 0.812500\n",
      "[0: 257/379] train loss: 0.504976 accuracy: 0.843750\n",
      "[0: 258/379] train loss: 0.682699 accuracy: 0.781250\n",
      "[0: 259/379] train loss: 0.553783 accuracy: 0.875000\n",
      "[0: 260/379] train loss: 0.481970 accuracy: 0.906250\n",
      "[0: 260/379] \u001B[94mtest\u001B[0m loss: 0.726226 accuracy: 0.812500\n",
      "[0: 261/379] train loss: 0.989575 accuracy: 0.781250\n",
      "[0: 262/379] train loss: 0.425023 accuracy: 0.906250\n",
      "[0: 263/379] train loss: 0.415026 accuracy: 0.906250\n",
      "[0: 264/379] train loss: 0.534948 accuracy: 0.875000\n",
      "[0: 265/379] train loss: 1.594567 accuracy: 0.718750\n",
      "[0: 266/379] train loss: 0.672515 accuracy: 0.812500\n",
      "[0: 267/379] train loss: 0.615484 accuracy: 0.843750\n",
      "[0: 268/379] train loss: 0.602336 accuracy: 0.812500\n",
      "[0: 269/379] train loss: 0.694267 accuracy: 0.812500\n",
      "[0: 270/379] train loss: 0.595660 accuracy: 0.843750\n",
      "[0: 270/379] \u001B[94mtest\u001B[0m loss: 0.710242 accuracy: 0.781250\n",
      "[0: 271/379] train loss: 0.507877 accuracy: 0.843750\n",
      "[0: 272/379] train loss: 0.792665 accuracy: 0.812500\n",
      "[0: 273/379] train loss: 0.868717 accuracy: 0.750000\n",
      "[0: 274/379] train loss: 0.635893 accuracy: 0.843750\n",
      "[0: 275/379] train loss: 0.741768 accuracy: 0.812500\n",
      "[0: 276/379] train loss: 0.380618 accuracy: 0.906250\n",
      "[0: 277/379] train loss: 0.643746 accuracy: 0.843750\n",
      "[0: 278/379] train loss: 0.681990 accuracy: 0.812500\n",
      "[0: 279/379] train loss: 0.514229 accuracy: 0.875000\n",
      "[0: 280/379] train loss: 0.563864 accuracy: 0.812500\n",
      "[0: 280/379] \u001B[94mtest\u001B[0m loss: 1.116700 accuracy: 0.656250\n",
      "[0: 281/379] train loss: 0.332533 accuracy: 0.937500\n",
      "[0: 282/379] train loss: 0.348623 accuracy: 0.906250\n",
      "[0: 283/379] train loss: 0.861792 accuracy: 0.812500\n",
      "[0: 284/379] train loss: 0.703630 accuracy: 0.812500\n",
      "[0: 285/379] train loss: 0.868039 accuracy: 0.750000\n",
      "[0: 286/379] train loss: 0.523157 accuracy: 0.875000\n",
      "[0: 287/379] train loss: 0.579891 accuracy: 0.843750\n",
      "[0: 288/379] train loss: 0.531274 accuracy: 0.875000\n",
      "[0: 289/379] train loss: 0.869728 accuracy: 0.812500\n",
      "[0: 290/379] train loss: 0.616113 accuracy: 0.843750\n",
      "[0: 290/379] \u001B[94mtest\u001B[0m loss: 1.160279 accuracy: 0.718750\n",
      "[0: 291/379] train loss: 0.835343 accuracy: 0.812500\n",
      "[0: 292/379] train loss: 0.514304 accuracy: 0.937500\n",
      "[0: 293/379] train loss: 0.363136 accuracy: 0.906250\n",
      "[0: 294/379] train loss: 0.977522 accuracy: 0.812500\n",
      "[0: 295/379] train loss: 0.999115 accuracy: 0.718750\n",
      "[0: 296/379] train loss: 0.529398 accuracy: 0.875000\n",
      "[0: 297/379] train loss: 0.758087 accuracy: 0.781250\n",
      "[0: 298/379] train loss: 0.496365 accuracy: 0.906250\n",
      "[0: 299/379] train loss: 0.844043 accuracy: 0.781250\n",
      "[0: 300/379] train loss: 1.055995 accuracy: 0.687500\n",
      "[0: 300/379] \u001B[94mtest\u001B[0m loss: 0.744885 accuracy: 0.843750\n",
      "[0: 301/379] train loss: 0.699699 accuracy: 0.750000\n",
      "[0: 302/379] train loss: 0.600069 accuracy: 0.843750\n",
      "[0: 303/379] train loss: 0.705609 accuracy: 0.781250\n",
      "[0: 304/379] train loss: 0.479395 accuracy: 0.875000\n",
      "[0: 305/379] train loss: 0.375306 accuracy: 0.875000\n",
      "[0: 306/379] train loss: 1.077989 accuracy: 0.687500\n",
      "[0: 307/379] train loss: 0.444453 accuracy: 0.906250\n",
      "[0: 308/379] train loss: 0.565487 accuracy: 0.812500\n",
      "[0: 309/379] train loss: 0.569256 accuracy: 0.812500\n",
      "[0: 310/379] train loss: 0.711214 accuracy: 0.843750\n",
      "[0: 310/379] \u001B[94mtest\u001B[0m loss: 0.516668 accuracy: 0.843750\n",
      "[0: 311/379] train loss: 0.493479 accuracy: 0.906250\n",
      "[0: 312/379] train loss: 0.492499 accuracy: 0.937500\n",
      "[0: 313/379] train loss: 0.285076 accuracy: 0.937500\n",
      "[0: 314/379] train loss: 0.866654 accuracy: 0.843750\n",
      "[0: 315/379] train loss: 0.655545 accuracy: 0.906250\n",
      "[0: 316/379] train loss: 0.546469 accuracy: 0.812500\n",
      "[0: 317/379] train loss: 0.489810 accuracy: 0.906250\n",
      "[0: 318/379] train loss: 0.816564 accuracy: 0.781250\n",
      "[0: 319/379] train loss: 0.661146 accuracy: 0.781250\n",
      "[0: 320/379] train loss: 0.541359 accuracy: 0.843750\n",
      "[0: 320/379] \u001B[94mtest\u001B[0m loss: 0.631752 accuracy: 0.812500\n",
      "[0: 321/379] train loss: 0.590129 accuracy: 0.906250\n",
      "[0: 322/379] train loss: 0.599145 accuracy: 0.812500\n",
      "[0: 323/379] train loss: 0.754640 accuracy: 0.812500\n",
      "[0: 324/379] train loss: 0.970167 accuracy: 0.781250\n",
      "[0: 325/379] train loss: 0.542454 accuracy: 0.906250\n",
      "[0: 326/379] train loss: 0.417012 accuracy: 0.875000\n",
      "[0: 327/379] train loss: 0.516170 accuracy: 0.812500\n",
      "[0: 328/379] train loss: 0.469608 accuracy: 0.843750\n",
      "[0: 329/379] train loss: 0.552474 accuracy: 0.843750\n",
      "[0: 330/379] train loss: 0.806218 accuracy: 0.843750\n",
      "[0: 330/379] \u001B[94mtest\u001B[0m loss: 0.819007 accuracy: 0.781250\n",
      "[0: 331/379] train loss: 0.519890 accuracy: 0.843750\n",
      "[0: 332/379] train loss: 0.522483 accuracy: 0.843750\n",
      "[0: 333/379] train loss: 0.567521 accuracy: 0.843750\n",
      "[0: 334/379] train loss: 0.804858 accuracy: 0.750000\n",
      "[0: 335/379] train loss: 0.717981 accuracy: 0.750000\n",
      "[0: 336/379] train loss: 0.706141 accuracy: 0.718750\n",
      "[0: 337/379] train loss: 0.355353 accuracy: 0.937500\n",
      "[0: 338/379] train loss: 0.565842 accuracy: 0.875000\n",
      "[0: 339/379] train loss: 0.193091 accuracy: 0.968750\n",
      "[0: 340/379] train loss: 0.300619 accuracy: 0.968750\n",
      "[0: 340/379] \u001B[94mtest\u001B[0m loss: 0.647304 accuracy: 0.750000\n",
      "[0: 341/379] train loss: 0.483474 accuracy: 0.875000\n",
      "[0: 342/379] train loss: 0.627846 accuracy: 0.875000\n",
      "[0: 343/379] train loss: 0.373951 accuracy: 0.906250\n",
      "[0: 344/379] train loss: 0.710083 accuracy: 0.781250\n",
      "[0: 345/379] train loss: 0.331457 accuracy: 0.937500\n",
      "[0: 346/379] train loss: 1.486977 accuracy: 0.750000\n",
      "[0: 347/379] train loss: 1.135073 accuracy: 0.718750\n",
      "[0: 348/379] train loss: 0.362765 accuracy: 0.906250\n",
      "[0: 349/379] train loss: 0.942370 accuracy: 0.781250\n",
      "[0: 350/379] train loss: 0.636666 accuracy: 0.781250\n",
      "[0: 350/379] \u001B[94mtest\u001B[0m loss: 0.769629 accuracy: 0.781250\n",
      "[0: 351/379] train loss: 0.421125 accuracy: 0.906250\n",
      "[0: 352/379] train loss: 0.632333 accuracy: 0.843750\n",
      "[0: 353/379] train loss: 0.871979 accuracy: 0.750000\n",
      "[0: 354/379] train loss: 0.497097 accuracy: 0.937500\n",
      "[0: 355/379] train loss: 0.403040 accuracy: 0.937500\n",
      "[0: 356/379] train loss: 0.388367 accuracy: 0.968750\n",
      "[0: 357/379] train loss: 0.625436 accuracy: 0.843750\n",
      "[0: 358/379] train loss: 1.191887 accuracy: 0.687500\n",
      "[0: 359/379] train loss: 0.514109 accuracy: 0.875000\n",
      "[0: 360/379] train loss: 0.700062 accuracy: 0.812500\n",
      "[0: 360/379] \u001B[94mtest\u001B[0m loss: 0.849151 accuracy: 0.718750\n",
      "[0: 361/379] train loss: 0.951695 accuracy: 0.750000\n",
      "[0: 362/379] train loss: 0.714305 accuracy: 0.781250\n",
      "[0: 363/379] train loss: 0.550835 accuracy: 0.812500\n",
      "[0: 364/379] train loss: 0.562659 accuracy: 0.875000\n",
      "[0: 365/379] train loss: 0.612509 accuracy: 0.781250\n",
      "[0: 366/379] train loss: 0.628329 accuracy: 0.843750\n",
      "[0: 367/379] train loss: 0.571225 accuracy: 0.875000\n",
      "[0: 368/379] train loss: 0.288964 accuracy: 0.968750\n",
      "[0: 369/379] train loss: 0.675488 accuracy: 0.812500\n",
      "[0: 370/379] train loss: 0.403635 accuracy: 0.937500\n",
      "[0: 370/379] \u001B[94mtest\u001B[0m loss: 0.662075 accuracy: 0.718750\n",
      "[0: 371/379] train loss: 0.761461 accuracy: 0.781250\n",
      "[0: 372/379] train loss: 0.378100 accuracy: 0.875000\n",
      "[0: 373/379] train loss: 0.740165 accuracy: 0.781250\n",
      "[0: 374/379] train loss: 0.706126 accuracy: 0.843750\n",
      "[0: 375/379] train loss: 0.525857 accuracy: 0.875000\n",
      "[0: 376/379] train loss: 0.393528 accuracy: 0.875000\n",
      "[0: 377/379] train loss: 0.332899 accuracy: 0.937500\n",
      "[0: 378/379] train loss: 0.574543 accuracy: 0.906250\n",
      "[0: 379/379] train loss: 2.497179 accuracy: 0.156250\n",
      "[1: 0/379] train loss: 0.903086 accuracy: 0.718750\n",
      "[1: 0/379] \u001B[94mtest\u001B[0m loss: 0.531949 accuracy: 0.812500\n",
      "[1: 1/379] train loss: 0.376883 accuracy: 0.906250\n",
      "[1: 2/379] train loss: 0.767987 accuracy: 0.781250\n",
      "[1: 3/379] train loss: 0.637785 accuracy: 0.875000\n",
      "[1: 4/379] train loss: 0.633189 accuracy: 0.875000\n",
      "[1: 5/379] train loss: 0.482749 accuracy: 0.875000\n",
      "[1: 6/379] train loss: 0.435648 accuracy: 0.937500\n",
      "[1: 7/379] train loss: 0.493830 accuracy: 0.875000\n",
      "[1: 8/379] train loss: 0.700526 accuracy: 0.781250\n",
      "[1: 9/379] train loss: 0.496881 accuracy: 0.812500\n",
      "[1: 10/379] train loss: 0.725925 accuracy: 0.812500\n",
      "[1: 10/379] \u001B[94mtest\u001B[0m loss: 1.049669 accuracy: 0.718750\n",
      "[1: 11/379] train loss: 0.630233 accuracy: 0.843750\n",
      "[1: 12/379] train loss: 0.627347 accuracy: 0.875000\n",
      "[1: 13/379] train loss: 0.458489 accuracy: 0.875000\n",
      "[1: 14/379] train loss: 0.385991 accuracy: 0.906250\n",
      "[1: 15/379] train loss: 0.746426 accuracy: 0.812500\n",
      "[1: 16/379] train loss: 0.354244 accuracy: 0.937500\n",
      "[1: 17/379] train loss: 0.496722 accuracy: 0.906250\n",
      "[1: 18/379] train loss: 0.721410 accuracy: 0.781250\n",
      "[1: 19/379] train loss: 0.487607 accuracy: 0.843750\n",
      "[1: 20/379] train loss: 0.578063 accuracy: 0.843750\n",
      "[1: 20/379] \u001B[94mtest\u001B[0m loss: 0.279755 accuracy: 0.937500\n",
      "[1: 21/379] train loss: 0.847441 accuracy: 0.812500\n",
      "[1: 22/379] train loss: 0.672806 accuracy: 0.781250\n",
      "[1: 23/379] train loss: 0.340600 accuracy: 0.906250\n",
      "[1: 24/379] train loss: 0.502087 accuracy: 0.843750\n",
      "[1: 25/379] train loss: 0.235843 accuracy: 0.968750\n",
      "[1: 26/379] train loss: 0.592444 accuracy: 0.875000\n",
      "[1: 27/379] train loss: 0.245910 accuracy: 0.906250\n",
      "[1: 28/379] train loss: 0.894025 accuracy: 0.781250\n",
      "[1: 29/379] train loss: 0.736787 accuracy: 0.781250\n",
      "[1: 30/379] train loss: 0.565066 accuracy: 0.843750\n",
      "[1: 30/379] \u001B[94mtest\u001B[0m loss: 0.833337 accuracy: 0.781250\n",
      "[1: 31/379] train loss: 0.738278 accuracy: 0.843750\n",
      "[1: 32/379] train loss: 0.613511 accuracy: 0.812500\n",
      "[1: 33/379] train loss: 0.715220 accuracy: 0.812500\n",
      "[1: 34/379] train loss: 0.591521 accuracy: 0.843750\n",
      "[1: 35/379] train loss: 0.385599 accuracy: 0.906250\n",
      "[1: 36/379] train loss: 0.581067 accuracy: 0.781250\n",
      "[1: 37/379] train loss: 0.355052 accuracy: 0.906250\n",
      "[1: 38/379] train loss: 0.338153 accuracy: 0.968750\n",
      "[1: 39/379] train loss: 0.822665 accuracy: 0.781250\n",
      "[1: 40/379] train loss: 0.308449 accuracy: 0.937500\n",
      "[1: 40/379] \u001B[94mtest\u001B[0m loss: 0.548188 accuracy: 0.843750\n",
      "[1: 41/379] train loss: 0.215880 accuracy: 0.968750\n",
      "[1: 42/379] train loss: 0.524803 accuracy: 0.875000\n",
      "[1: 43/379] train loss: 0.527920 accuracy: 0.843750\n",
      "[1: 44/379] train loss: 1.029285 accuracy: 0.875000\n",
      "[1: 45/379] train loss: 0.361418 accuracy: 0.875000\n",
      "[1: 46/379] train loss: 0.412651 accuracy: 0.875000\n",
      "[1: 47/379] train loss: 0.948829 accuracy: 0.781250\n",
      "[1: 48/379] train loss: 0.442691 accuracy: 0.843750\n",
      "[1: 49/379] train loss: 0.335670 accuracy: 0.937500\n",
      "[1: 50/379] train loss: 0.513594 accuracy: 0.875000\n",
      "[1: 50/379] \u001B[94mtest\u001B[0m loss: 0.967665 accuracy: 0.781250\n",
      "[1: 51/379] train loss: 0.345996 accuracy: 0.843750\n",
      "[1: 52/379] train loss: 0.757281 accuracy: 0.781250\n",
      "[1: 53/379] train loss: 0.550898 accuracy: 0.875000\n",
      "[1: 54/379] train loss: 0.339465 accuracy: 0.937500\n",
      "[1: 55/379] train loss: 0.490338 accuracy: 0.906250\n",
      "[1: 56/379] train loss: 0.446776 accuracy: 0.843750\n",
      "[1: 57/379] train loss: 0.932405 accuracy: 0.750000\n",
      "[1: 58/379] train loss: 0.521816 accuracy: 0.812500\n",
      "[1: 59/379] train loss: 0.505785 accuracy: 0.843750\n",
      "[1: 60/379] train loss: 0.598997 accuracy: 0.812500\n",
      "[1: 60/379] \u001B[94mtest\u001B[0m loss: 0.465314 accuracy: 0.875000\n",
      "[1: 61/379] train loss: 0.581251 accuracy: 0.812500\n",
      "[1: 62/379] train loss: 0.660619 accuracy: 0.812500\n",
      "[1: 63/379] train loss: 0.568227 accuracy: 0.843750\n",
      "[1: 64/379] train loss: 0.349702 accuracy: 0.937500\n",
      "[1: 65/379] train loss: 0.504285 accuracy: 0.906250\n",
      "[1: 66/379] train loss: 0.262966 accuracy: 0.968750\n",
      "[1: 67/379] train loss: 0.518696 accuracy: 0.875000\n",
      "[1: 68/379] train loss: 0.701640 accuracy: 0.750000\n",
      "[1: 69/379] train loss: 0.451254 accuracy: 0.875000\n",
      "[1: 70/379] train loss: 0.384797 accuracy: 0.875000\n",
      "[1: 70/379] \u001B[94mtest\u001B[0m loss: 0.837636 accuracy: 0.718750\n",
      "[1: 71/379] train loss: 0.599110 accuracy: 0.875000\n",
      "[1: 72/379] train loss: 0.399087 accuracy: 0.875000\n",
      "[1: 73/379] train loss: 0.744851 accuracy: 0.843750\n",
      "[1: 74/379] train loss: 0.591542 accuracy: 0.781250\n",
      "[1: 75/379] train loss: 0.653848 accuracy: 0.843750\n",
      "[1: 76/379] train loss: 0.775670 accuracy: 0.812500\n",
      "[1: 77/379] train loss: 0.499010 accuracy: 0.843750\n",
      "[1: 78/379] train loss: 0.453261 accuracy: 0.937500\n",
      "[1: 79/379] train loss: 0.468055 accuracy: 0.906250\n",
      "[1: 80/379] train loss: 0.664465 accuracy: 0.843750\n",
      "[1: 80/379] \u001B[94mtest\u001B[0m loss: 0.517196 accuracy: 0.843750\n",
      "[1: 81/379] train loss: 0.653248 accuracy: 0.781250\n",
      "[1: 82/379] train loss: 0.598195 accuracy: 0.843750\n",
      "[1: 83/379] train loss: 1.136870 accuracy: 0.812500\n",
      "[1: 84/379] train loss: 0.439107 accuracy: 0.875000\n",
      "[1: 85/379] train loss: 0.396411 accuracy: 0.906250\n",
      "[1: 86/379] train loss: 0.349203 accuracy: 0.937500\n",
      "[1: 87/379] train loss: 0.199339 accuracy: 0.968750\n",
      "[1: 88/379] train loss: 0.357529 accuracy: 0.968750\n",
      "[1: 89/379] train loss: 0.385326 accuracy: 0.875000\n",
      "[1: 90/379] train loss: 0.582134 accuracy: 0.875000\n",
      "[1: 90/379] \u001B[94mtest\u001B[0m loss: 0.480679 accuracy: 0.812500\n",
      "[1: 91/379] train loss: 0.225712 accuracy: 0.906250\n",
      "[1: 92/379] train loss: 0.313003 accuracy: 0.937500\n",
      "[1: 93/379] train loss: 0.318724 accuracy: 0.937500\n",
      "[1: 94/379] train loss: 0.590914 accuracy: 0.812500\n",
      "[1: 95/379] train loss: 0.753404 accuracy: 0.812500\n",
      "[1: 96/379] train loss: 0.515469 accuracy: 0.906250\n",
      "[1: 97/379] train loss: 0.462664 accuracy: 0.843750\n",
      "[1: 98/379] train loss: 0.975888 accuracy: 0.687500\n",
      "[1: 99/379] train loss: 0.822982 accuracy: 0.781250\n",
      "[1: 100/379] train loss: 0.625356 accuracy: 0.812500\n",
      "[1: 100/379] \u001B[94mtest\u001B[0m loss: 0.677429 accuracy: 0.812500\n",
      "[1: 101/379] train loss: 0.350273 accuracy: 0.937500\n",
      "[1: 102/379] train loss: 0.555598 accuracy: 0.843750\n",
      "[1: 103/379] train loss: 0.484897 accuracy: 0.875000\n",
      "[1: 104/379] train loss: 0.268618 accuracy: 0.968750\n",
      "[1: 105/379] train loss: 0.461914 accuracy: 0.906250\n",
      "[1: 106/379] train loss: 0.411278 accuracy: 0.937500\n",
      "[1: 107/379] train loss: 0.481364 accuracy: 0.937500\n",
      "[1: 108/379] train loss: 0.851610 accuracy: 0.750000\n",
      "[1: 109/379] train loss: 0.580918 accuracy: 0.875000\n",
      "[1: 110/379] train loss: 0.649919 accuracy: 0.781250\n",
      "[1: 110/379] \u001B[94mtest\u001B[0m loss: 0.360699 accuracy: 0.968750\n",
      "[1: 111/379] train loss: 0.425480 accuracy: 0.906250\n",
      "[1: 112/379] train loss: 0.602405 accuracy: 0.875000\n",
      "[1: 113/379] train loss: 0.672970 accuracy: 0.750000\n",
      "[1: 114/379] train loss: 0.386932 accuracy: 0.906250\n",
      "[1: 115/379] train loss: 0.523155 accuracy: 0.875000\n",
      "[1: 116/379] train loss: 0.396619 accuracy: 0.906250\n",
      "[1: 117/379] train loss: 0.441443 accuracy: 0.843750\n",
      "[1: 118/379] train loss: 0.268826 accuracy: 0.937500\n",
      "[1: 119/379] train loss: 0.319506 accuracy: 0.968750\n",
      "[1: 120/379] train loss: 0.246157 accuracy: 0.937500\n",
      "[1: 120/379] \u001B[94mtest\u001B[0m loss: 0.380146 accuracy: 0.906250\n",
      "[1: 121/379] train loss: 0.652738 accuracy: 0.843750\n",
      "[1: 122/379] train loss: 0.243602 accuracy: 0.937500\n",
      "[1: 123/379] train loss: 0.327872 accuracy: 0.906250\n",
      "[1: 124/379] train loss: 0.265737 accuracy: 0.937500\n",
      "[1: 125/379] train loss: 0.218672 accuracy: 0.937500\n",
      "[1: 126/379] train loss: 0.364222 accuracy: 0.906250\n",
      "[1: 127/379] train loss: 0.633428 accuracy: 0.843750\n",
      "[1: 128/379] train loss: 0.431014 accuracy: 0.875000\n",
      "[1: 129/379] train loss: 0.681378 accuracy: 0.875000\n",
      "[1: 130/379] train loss: 0.449128 accuracy: 0.812500\n",
      "[1: 130/379] \u001B[94mtest\u001B[0m loss: 0.446809 accuracy: 0.843750\n",
      "[1: 131/379] train loss: 0.300216 accuracy: 0.937500\n",
      "[1: 132/379] train loss: 0.272323 accuracy: 0.937500\n",
      "[1: 133/379] train loss: 0.306893 accuracy: 0.968750\n",
      "[1: 134/379] train loss: 0.681042 accuracy: 0.843750\n",
      "[1: 135/379] train loss: 0.583620 accuracy: 0.781250\n",
      "[1: 136/379] train loss: 0.772897 accuracy: 0.781250\n",
      "[1: 137/379] train loss: 0.891414 accuracy: 0.718750\n",
      "[1: 138/379] train loss: 0.770307 accuracy: 0.812500\n",
      "[1: 139/379] train loss: 0.801100 accuracy: 0.812500\n",
      "[1: 140/379] train loss: 0.336363 accuracy: 0.968750\n",
      "[1: 140/379] \u001B[94mtest\u001B[0m loss: 0.857728 accuracy: 0.656250\n",
      "[1: 141/379] train loss: 0.753306 accuracy: 0.781250\n",
      "[1: 142/379] train loss: 0.504704 accuracy: 0.843750\n",
      "[1: 143/379] train loss: 1.051119 accuracy: 0.718750\n",
      "[1: 144/379] train loss: 0.319694 accuracy: 0.937500\n",
      "[1: 145/379] train loss: 0.535656 accuracy: 0.875000\n",
      "[1: 146/379] train loss: 0.694863 accuracy: 0.750000\n",
      "[1: 147/379] train loss: 0.383718 accuracy: 0.875000\n",
      "[1: 148/379] train loss: 0.523292 accuracy: 0.906250\n",
      "[1: 149/379] train loss: 0.364162 accuracy: 0.906250\n",
      "[1: 150/379] train loss: 0.463259 accuracy: 0.843750\n",
      "[1: 150/379] \u001B[94mtest\u001B[0m loss: 1.104759 accuracy: 0.750000\n",
      "[1: 151/379] train loss: 0.384938 accuracy: 0.906250\n",
      "[1: 152/379] train loss: 0.319875 accuracy: 0.968750\n",
      "[1: 153/379] train loss: 0.319210 accuracy: 0.937500\n",
      "[1: 154/379] train loss: 0.605097 accuracy: 0.875000\n",
      "[1: 155/379] train loss: 0.478599 accuracy: 0.843750\n",
      "[1: 156/379] train loss: 0.625734 accuracy: 0.812500\n",
      "[1: 157/379] train loss: 0.735219 accuracy: 0.718750\n",
      "[1: 158/379] train loss: 0.431370 accuracy: 0.906250\n",
      "[1: 159/379] train loss: 0.589846 accuracy: 0.843750\n",
      "[1: 160/379] train loss: 0.279638 accuracy: 0.906250\n",
      "[1: 160/379] \u001B[94mtest\u001B[0m loss: 0.611968 accuracy: 0.812500\n",
      "[1: 161/379] train loss: 0.408834 accuracy: 0.906250\n",
      "[1: 162/379] train loss: 0.592936 accuracy: 0.843750\n",
      "[1: 163/379] train loss: 0.512710 accuracy: 0.843750\n",
      "[1: 164/379] train loss: 0.784729 accuracy: 0.843750\n",
      "[1: 165/379] train loss: 0.408211 accuracy: 0.906250\n",
      "[1: 166/379] train loss: 0.462495 accuracy: 0.843750\n",
      "[1: 167/379] train loss: 0.532661 accuracy: 0.875000\n",
      "[1: 168/379] train loss: 0.727261 accuracy: 0.875000\n",
      "[1: 169/379] train loss: 0.540953 accuracy: 0.875000\n",
      "[1: 170/379] train loss: 0.289228 accuracy: 1.000000\n",
      "[1: 170/379] \u001B[94mtest\u001B[0m loss: 0.542782 accuracy: 0.750000\n",
      "[1: 171/379] train loss: 0.458238 accuracy: 0.875000\n",
      "[1: 172/379] train loss: 0.596314 accuracy: 0.843750\n",
      "[1: 173/379] train loss: 0.792707 accuracy: 0.843750\n",
      "[1: 174/379] train loss: 1.517425 accuracy: 0.656250\n",
      "[1: 175/379] train loss: 0.475452 accuracy: 0.812500\n",
      "[1: 176/379] train loss: 0.427136 accuracy: 0.875000\n",
      "[1: 177/379] train loss: 0.379214 accuracy: 0.906250\n",
      "[1: 178/379] train loss: 0.253252 accuracy: 0.968750\n",
      "[1: 179/379] train loss: 0.261116 accuracy: 0.968750\n",
      "[1: 180/379] train loss: 0.585214 accuracy: 0.812500\n",
      "[1: 180/379] \u001B[94mtest\u001B[0m loss: 0.866407 accuracy: 0.718750\n",
      "[1: 181/379] train loss: 0.390610 accuracy: 0.937500\n",
      "[1: 182/379] train loss: 0.380962 accuracy: 0.968750\n",
      "[1: 183/379] train loss: 0.305981 accuracy: 0.937500\n",
      "[1: 184/379] train loss: 0.657131 accuracy: 0.812500\n",
      "[1: 185/379] train loss: 0.456593 accuracy: 0.843750\n",
      "[1: 186/379] train loss: 0.703452 accuracy: 0.750000\n",
      "[1: 187/379] train loss: 0.291499 accuracy: 0.906250\n",
      "[1: 188/379] train loss: 0.136026 accuracy: 0.968750\n",
      "[1: 189/379] train loss: 0.365112 accuracy: 0.906250\n",
      "[1: 190/379] train loss: 0.343768 accuracy: 0.906250\n",
      "[1: 190/379] \u001B[94mtest\u001B[0m loss: 0.254261 accuracy: 0.937500\n",
      "[1: 191/379] train loss: 0.377887 accuracy: 0.875000\n",
      "[1: 192/379] train loss: 0.364994 accuracy: 0.875000\n",
      "[1: 193/379] train loss: 0.665280 accuracy: 0.812500\n",
      "[1: 194/379] train loss: 0.560947 accuracy: 0.875000\n",
      "[1: 195/379] train loss: 0.439547 accuracy: 0.906250\n",
      "[1: 196/379] train loss: 0.396190 accuracy: 0.875000\n",
      "[1: 197/379] train loss: 0.316302 accuracy: 0.906250\n",
      "[1: 198/379] train loss: 0.486300 accuracy: 0.843750\n",
      "[1: 199/379] train loss: 0.358969 accuracy: 0.906250\n",
      "[1: 200/379] train loss: 0.528044 accuracy: 0.843750\n",
      "[1: 200/379] \u001B[94mtest\u001B[0m loss: 0.287068 accuracy: 0.937500\n",
      "[1: 201/379] train loss: 0.632860 accuracy: 0.781250\n",
      "[1: 202/379] train loss: 0.412565 accuracy: 0.937500\n",
      "[1: 203/379] train loss: 0.826652 accuracy: 0.687500\n",
      "[1: 204/379] train loss: 0.313192 accuracy: 0.875000\n",
      "[1: 205/379] train loss: 0.795709 accuracy: 0.781250\n",
      "[1: 206/379] train loss: 0.597752 accuracy: 0.843750\n",
      "[1: 207/379] train loss: 0.172399 accuracy: 0.968750\n",
      "[1: 208/379] train loss: 0.521132 accuracy: 0.875000\n",
      "[1: 209/379] train loss: 0.419647 accuracy: 0.906250\n",
      "[1: 210/379] train loss: 0.547517 accuracy: 0.812500\n",
      "[1: 210/379] \u001B[94mtest\u001B[0m loss: 0.586185 accuracy: 0.812500\n",
      "[1: 211/379] train loss: 0.213033 accuracy: 0.937500\n",
      "[1: 212/379] train loss: 0.417342 accuracy: 0.875000\n",
      "[1: 213/379] train loss: 0.245186 accuracy: 0.937500\n",
      "[1: 214/379] train loss: 0.355106 accuracy: 0.937500\n",
      "[1: 215/379] train loss: 0.296235 accuracy: 0.937500\n",
      "[1: 216/379] train loss: 0.527143 accuracy: 0.750000\n",
      "[1: 217/379] train loss: 0.529191 accuracy: 0.875000\n",
      "[1: 218/379] train loss: 0.368938 accuracy: 0.875000\n",
      "[1: 219/379] train loss: 0.329245 accuracy: 0.875000\n",
      "[1: 220/379] train loss: 0.706960 accuracy: 0.812500\n",
      "[1: 220/379] \u001B[94mtest\u001B[0m loss: 0.502114 accuracy: 0.875000\n",
      "[1: 221/379] train loss: 0.964366 accuracy: 0.843750\n",
      "[1: 222/379] train loss: 0.430418 accuracy: 0.906250\n",
      "[1: 223/379] train loss: 0.612709 accuracy: 0.812500\n",
      "[1: 224/379] train loss: 0.503267 accuracy: 0.750000\n",
      "[1: 225/379] train loss: 0.309967 accuracy: 0.968750\n",
      "[1: 226/379] train loss: 0.499644 accuracy: 0.843750\n",
      "[1: 227/379] train loss: 0.203636 accuracy: 0.937500\n",
      "[1: 228/379] train loss: 0.252254 accuracy: 0.968750\n",
      "[1: 229/379] train loss: 0.494214 accuracy: 0.875000\n",
      "[1: 230/379] train loss: 0.606934 accuracy: 0.812500\n",
      "[1: 230/379] \u001B[94mtest\u001B[0m loss: 0.320955 accuracy: 0.875000\n",
      "[1: 231/379] train loss: 0.746220 accuracy: 0.843750\n",
      "[1: 232/379] train loss: 0.338891 accuracy: 0.875000\n",
      "[1: 233/379] train loss: 0.641357 accuracy: 0.812500\n",
      "[1: 234/379] train loss: 0.323820 accuracy: 0.906250\n",
      "[1: 235/379] train loss: 0.309347 accuracy: 0.843750\n",
      "[1: 236/379] train loss: 0.335810 accuracy: 0.937500\n",
      "[1: 237/379] train loss: 0.229807 accuracy: 0.937500\n",
      "[1: 238/379] train loss: 0.305815 accuracy: 0.875000\n",
      "[1: 239/379] train loss: 0.237072 accuracy: 0.968750\n",
      "[1: 240/379] train loss: 0.288141 accuracy: 0.937500\n",
      "[1: 240/379] \u001B[94mtest\u001B[0m loss: 0.536616 accuracy: 0.875000\n",
      "[1: 241/379] train loss: 0.481999 accuracy: 0.812500\n",
      "[1: 242/379] train loss: 0.870731 accuracy: 0.750000\n",
      "[1: 243/379] train loss: 0.321399 accuracy: 0.906250\n",
      "[1: 244/379] train loss: 0.446162 accuracy: 0.875000\n",
      "[1: 245/379] train loss: 0.082324 accuracy: 1.000000\n",
      "[1: 246/379] train loss: 0.223082 accuracy: 0.968750\n",
      "[1: 247/379] train loss: 0.266111 accuracy: 0.968750\n",
      "[1: 248/379] train loss: 0.526450 accuracy: 0.875000\n",
      "[1: 249/379] train loss: 0.674678 accuracy: 0.906250\n",
      "[1: 250/379] train loss: 1.163993 accuracy: 0.750000\n",
      "[1: 250/379] \u001B[94mtest\u001B[0m loss: 0.314846 accuracy: 0.968750\n",
      "[1: 251/379] train loss: 0.278270 accuracy: 0.937500\n",
      "[1: 252/379] train loss: 0.401139 accuracy: 0.906250\n",
      "[1: 253/379] train loss: 0.368462 accuracy: 0.843750\n",
      "[1: 254/379] train loss: 0.353842 accuracy: 0.937500\n",
      "[1: 255/379] train loss: 0.565534 accuracy: 0.812500\n",
      "[1: 256/379] train loss: 0.409841 accuracy: 0.906250\n",
      "[1: 257/379] train loss: 0.393452 accuracy: 0.812500\n",
      "[1: 258/379] train loss: 0.313173 accuracy: 0.906250\n",
      "[1: 259/379] train loss: 0.712263 accuracy: 0.781250\n",
      "[1: 260/379] train loss: 0.611226 accuracy: 0.906250\n",
      "[1: 260/379] \u001B[94mtest\u001B[0m loss: 0.099354 accuracy: 0.968750\n",
      "[1: 261/379] train loss: 0.489988 accuracy: 0.843750\n",
      "[1: 262/379] train loss: 0.597868 accuracy: 0.843750\n",
      "[1: 263/379] train loss: 0.265862 accuracy: 0.937500\n",
      "[1: 264/379] train loss: 0.424859 accuracy: 0.906250\n",
      "[1: 265/379] train loss: 0.392646 accuracy: 0.906250\n",
      "[1: 266/379] train loss: 0.357662 accuracy: 0.906250\n",
      "[1: 267/379] train loss: 0.354884 accuracy: 0.875000\n",
      "[1: 268/379] train loss: 0.384789 accuracy: 0.906250\n",
      "[1: 269/379] train loss: 0.459412 accuracy: 0.937500\n",
      "[1: 270/379] train loss: 0.255364 accuracy: 0.937500\n",
      "[1: 270/379] \u001B[94mtest\u001B[0m loss: 0.443081 accuracy: 0.843750\n",
      "[1: 271/379] train loss: 0.594758 accuracy: 0.812500\n",
      "[1: 272/379] train loss: 0.548975 accuracy: 0.812500\n",
      "[1: 273/379] train loss: 0.426972 accuracy: 0.843750\n",
      "[1: 274/379] train loss: 0.160653 accuracy: 0.968750\n",
      "[1: 275/379] train loss: 0.214212 accuracy: 0.937500\n",
      "[1: 276/379] train loss: 0.188851 accuracy: 1.000000\n",
      "[1: 277/379] train loss: 0.542984 accuracy: 0.906250\n",
      "[1: 278/379] train loss: 0.819228 accuracy: 0.781250\n",
      "[1: 279/379] train loss: 0.549437 accuracy: 0.875000\n",
      "[1: 280/379] train loss: 0.301572 accuracy: 0.968750\n",
      "[1: 280/379] \u001B[94mtest\u001B[0m loss: 0.268041 accuracy: 0.906250\n",
      "[1: 281/379] train loss: 0.445304 accuracy: 0.875000\n",
      "[1: 282/379] train loss: 0.344245 accuracy: 0.875000\n",
      "[1: 283/379] train loss: 0.639627 accuracy: 0.812500\n",
      "[1: 284/379] train loss: 0.658490 accuracy: 0.812500\n",
      "[1: 285/379] train loss: 0.334267 accuracy: 0.968750\n",
      "[1: 286/379] train loss: 0.345201 accuracy: 0.875000\n",
      "[1: 287/379] train loss: 0.623202 accuracy: 0.875000\n",
      "[1: 288/379] train loss: 0.635140 accuracy: 0.781250\n",
      "[1: 289/379] train loss: 0.777282 accuracy: 0.781250\n",
      "[1: 290/379] train loss: 0.213134 accuracy: 0.937500\n",
      "[1: 290/379] \u001B[94mtest\u001B[0m loss: 0.418908 accuracy: 0.937500\n",
      "[1: 291/379] train loss: 0.534623 accuracy: 0.843750\n",
      "[1: 292/379] train loss: 0.579507 accuracy: 0.843750\n",
      "[1: 293/379] train loss: 0.534296 accuracy: 0.812500\n",
      "[1: 294/379] train loss: 0.190568 accuracy: 0.968750\n",
      "[1: 295/379] train loss: 0.347805 accuracy: 0.906250\n",
      "[1: 296/379] train loss: 0.461374 accuracy: 0.812500\n",
      "[1: 297/379] train loss: 0.552003 accuracy: 0.875000\n",
      "[1: 298/379] train loss: 0.623145 accuracy: 0.750000\n",
      "[1: 299/379] train loss: 0.406946 accuracy: 0.906250\n",
      "[1: 300/379] train loss: 0.475912 accuracy: 0.843750\n",
      "[1: 300/379] \u001B[94mtest\u001B[0m loss: 0.592107 accuracy: 0.781250\n",
      "[1: 301/379] train loss: 0.262340 accuracy: 0.937500\n",
      "[1: 302/379] train loss: 0.651195 accuracy: 0.843750\n",
      "[1: 303/379] train loss: 0.502724 accuracy: 0.812500\n",
      "[1: 304/379] train loss: 0.293603 accuracy: 0.937500\n",
      "[1: 305/379] train loss: 0.199582 accuracy: 0.937500\n",
      "[1: 306/379] train loss: 0.361470 accuracy: 0.906250\n",
      "[1: 307/379] train loss: 0.381052 accuracy: 0.906250\n",
      "[1: 308/379] train loss: 0.750706 accuracy: 0.750000\n",
      "[1: 309/379] train loss: 0.438606 accuracy: 0.875000\n",
      "[1: 310/379] train loss: 0.490952 accuracy: 0.875000\n",
      "[1: 310/379] \u001B[94mtest\u001B[0m loss: 1.017266 accuracy: 0.625000\n",
      "[1: 311/379] train loss: 0.319904 accuracy: 0.906250\n",
      "[1: 312/379] train loss: 0.446521 accuracy: 0.875000\n",
      "[1: 313/379] train loss: 0.555718 accuracy: 0.875000\n",
      "[1: 314/379] train loss: 0.431842 accuracy: 0.906250\n",
      "[1: 315/379] train loss: 0.403825 accuracy: 0.906250\n",
      "[1: 316/379] train loss: 0.494576 accuracy: 0.875000\n",
      "[1: 317/379] train loss: 0.310358 accuracy: 0.937500\n",
      "[1: 318/379] train loss: 0.683753 accuracy: 0.843750\n",
      "[1: 319/379] train loss: 0.376049 accuracy: 0.906250\n",
      "[1: 320/379] train loss: 0.118615 accuracy: 1.000000\n",
      "[1: 320/379] \u001B[94mtest\u001B[0m loss: 0.762379 accuracy: 0.750000\n",
      "[1: 321/379] train loss: 0.521321 accuracy: 0.937500\n",
      "[1: 322/379] train loss: 0.611710 accuracy: 0.843750\n",
      "[1: 323/379] train loss: 0.387064 accuracy: 0.875000\n",
      "[1: 324/379] train loss: 0.468968 accuracy: 0.843750\n",
      "[1: 325/379] train loss: 0.160254 accuracy: 1.000000\n",
      "[1: 326/379] train loss: 0.499230 accuracy: 0.875000\n",
      "[1: 327/379] train loss: 0.356620 accuracy: 0.906250\n",
      "[1: 328/379] train loss: 0.580760 accuracy: 0.875000\n",
      "[1: 329/379] train loss: 0.263928 accuracy: 0.906250\n",
      "[1: 330/379] train loss: 0.639778 accuracy: 0.812500\n",
      "[1: 330/379] \u001B[94mtest\u001B[0m loss: 0.566397 accuracy: 0.843750\n",
      "[1: 331/379] train loss: 0.529987 accuracy: 0.875000\n",
      "[1: 332/379] train loss: 0.308875 accuracy: 0.906250\n",
      "[1: 333/379] train loss: 0.261629 accuracy: 0.906250\n",
      "[1: 334/379] train loss: 0.621930 accuracy: 0.843750\n",
      "[1: 335/379] train loss: 0.519706 accuracy: 0.875000\n",
      "[1: 336/379] train loss: 0.281021 accuracy: 0.937500\n",
      "[1: 337/379] train loss: 0.406532 accuracy: 0.875000\n",
      "[1: 338/379] train loss: 0.294442 accuracy: 0.937500\n",
      "[1: 339/379] train loss: 0.304094 accuracy: 0.937500\n",
      "[1: 340/379] train loss: 0.531061 accuracy: 0.906250\n",
      "[1: 340/379] \u001B[94mtest\u001B[0m loss: 0.520376 accuracy: 0.843750\n",
      "[1: 341/379] train loss: 0.163535 accuracy: 0.937500\n",
      "[1: 342/379] train loss: 0.498127 accuracy: 0.875000\n",
      "[1: 343/379] train loss: 0.619433 accuracy: 0.843750\n",
      "[1: 344/379] train loss: 0.265227 accuracy: 0.968750\n",
      "[1: 345/379] train loss: 0.746791 accuracy: 0.750000\n",
      "[1: 346/379] train loss: 0.232175 accuracy: 0.937500\n",
      "[1: 347/379] train loss: 0.349969 accuracy: 0.906250\n",
      "[1: 348/379] train loss: 0.285853 accuracy: 0.937500\n",
      "[1: 349/379] train loss: 0.481967 accuracy: 0.843750\n",
      "[1: 350/379] train loss: 0.479172 accuracy: 0.875000\n",
      "[1: 350/379] \u001B[94mtest\u001B[0m loss: 0.609026 accuracy: 0.875000\n",
      "[1: 351/379] train loss: 0.377237 accuracy: 0.843750\n",
      "[1: 352/379] train loss: 0.389173 accuracy: 0.843750\n",
      "[1: 353/379] train loss: 0.250013 accuracy: 0.937500\n",
      "[1: 354/379] train loss: 0.204234 accuracy: 0.968750\n",
      "[1: 355/379] train loss: 0.179640 accuracy: 0.968750\n",
      "[1: 356/379] train loss: 0.374350 accuracy: 0.875000\n",
      "[1: 357/379] train loss: 0.559409 accuracy: 0.875000\n",
      "[1: 358/379] train loss: 0.635067 accuracy: 0.812500\n",
      "[1: 359/379] train loss: 0.486377 accuracy: 0.843750\n",
      "[1: 360/379] train loss: 0.605349 accuracy: 0.781250\n",
      "[1: 360/379] \u001B[94mtest\u001B[0m loss: 0.338682 accuracy: 0.906250\n",
      "[1: 361/379] train loss: 0.771049 accuracy: 0.781250\n",
      "[1: 362/379] train loss: 0.333769 accuracy: 0.906250\n",
      "[1: 363/379] train loss: 0.539722 accuracy: 0.812500\n",
      "[1: 364/379] train loss: 0.362698 accuracy: 0.906250\n",
      "[1: 365/379] train loss: 0.323658 accuracy: 0.937500\n",
      "[1: 366/379] train loss: 0.399363 accuracy: 0.843750\n",
      "[1: 367/379] train loss: 0.343046 accuracy: 0.937500\n",
      "[1: 368/379] train loss: 0.351626 accuracy: 0.875000\n",
      "[1: 369/379] train loss: 0.411118 accuracy: 0.812500\n",
      "[1: 370/379] train loss: 0.207290 accuracy: 0.968750\n",
      "[1: 370/379] \u001B[94mtest\u001B[0m loss: 0.471607 accuracy: 0.875000\n",
      "[1: 371/379] train loss: 0.549826 accuracy: 0.875000\n",
      "[1: 372/379] train loss: 0.414166 accuracy: 0.875000\n",
      "[1: 373/379] train loss: 0.123524 accuracy: 1.000000\n",
      "[1: 374/379] train loss: 0.231704 accuracy: 0.937500\n",
      "[1: 375/379] train loss: 0.396067 accuracy: 0.875000\n",
      "[1: 376/379] train loss: 0.420057 accuracy: 0.906250\n",
      "[1: 377/379] train loss: 0.403719 accuracy: 0.906250\n",
      "[1: 378/379] train loss: 0.564113 accuracy: 0.812500\n",
      "[1: 379/379] train loss: 0.110046 accuracy: 0.281250\n",
      "90it [00:28,  3.20it/s]\n",
      "final accuracy 0.860125260960334\n"
     ]
    }
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "# Classification visualization"
   ],
   "metadata": {
    "id": "9hPRYb0iJWP9"
   }
  },
  {
   "cell_type": "code",
   "source": [
    "!python /content/drive/MyDrive/cs182-pointnet-homework/utils/show_cls.py --model=/content/drive/MyDrive/cs182-pointnet-homework/pointnet/model.py"
   ],
   "metadata": {
    "id": "aYACVEK9xF6E",
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "outputId": "e7043fb9-341f-4e5d-e7ee-430efaf16d15"
   },
   "execution_count": 7,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Namespace(model='/content/drive/MyDrive/cs182-pointnet-homework/pointnet/model.py', num_points=2500)\n",
      "{'Airplane': 0, 'Bag': 1, 'Cap': 2, 'Car': 3, 'Chair': 4, 'Earphone': 5, 'Guitar': 6, 'Knife': 7, 'Lamp': 8, 'Laptop': 9, 'Motorbike': 10, 'Mug': 11, 'Pistol': 12, 'Rocket': 13, 'Skateboard': 14, 'Table': 15}\n",
      "{'Airplane': 4, 'Bag': 2, 'Cap': 2, 'Car': 4, 'Chair': 4, 'Earphone': 3, 'Guitar': 3, 'Knife': 2, 'Lamp': 4, 'Laptop': 2, 'Motorbike': 6, 'Mug': 2, 'Pistol': 3, 'Rocket': 3, 'Skateboard': 3, 'Table': 3} 4\n",
      "Traceback (most recent call last):\n",
      "  File \"/content/drive/MyDrive/cs182-pointnet-homework/utils/show_cls.py\", line 35, in <module>\n",
      "    classifier.load_state_dict(torch.load(opt.model))\n",
      "  File \"/usr/local/lib/python3.9/dist-packages/torch/serialization.py\", line 815, in load\n",
      "    return _legacy_load(opened_file, map_location, pickle_module, **pickle_load_args)\n",
      "  File \"/usr/local/lib/python3.9/dist-packages/torch/serialization.py\", line 1033, in _legacy_load\n",
      "    magic_number = pickle_module.load(f, **pickle_load_args)\n",
      "_pickle.UnpicklingError: invalid load key, 'f'.\n"
     ]
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [],
   "metadata": {
    "id": "HlXXpA1fADCh"
   },
   "execution_count": null,
   "outputs": []
  },
  {
   "cell_type": "markdown",
   "source": [
    "# Segmentation training"
   ],
   "metadata": {
    "id": "Hbrc5jUgJaD8"
   }
  },
  {
   "cell_type": "code",
   "source": [
    "!python /content/drive/MyDrive/cs182-pointnet-homework/utils/train_segmentation.py --dataset /content/drive/MyDrive/cs182-pointnet-homework/shapenetcore_partanno_segmentation_benchmark_v0/ --nepoch=2"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "yJe2BGfBG8W5",
    "outputId": "0c05d19b-4345-4248-9de6-37f9f310e872"
   },
   "execution_count": 8,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Namespace(batchSize=32, workers=4, nepoch=2, outf='seg', model='', dataset='/content/drive/MyDrive/cs182-pointnet-homework/shapenetcore_partanno_segmentation_benchmark_v0/', class_choice='Chair', feature_transform=False)\n",
      "Random Seed:  5112\n",
      "{'Chair': 0}\n",
      "{'Airplane': 4, 'Bag': 2, 'Cap': 2, 'Car': 4, 'Chair': 4, 'Earphone': 3, 'Guitar': 3, 'Knife': 2, 'Lamp': 4, 'Laptop': 2, 'Motorbike': 6, 'Mug': 2, 'Pistol': 3, 'Rocket': 3, 'Skateboard': 3, 'Table': 3} 4\n",
      "/usr/local/lib/python3.9/dist-packages/torch/utils/data/dataloader.py:561: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
      "  warnings.warn(_create_warning_msg(\n",
      "{'Chair': 0}\n",
      "{'Airplane': 4, 'Bag': 2, 'Cap': 2, 'Car': 4, 'Chair': 4, 'Earphone': 3, 'Guitar': 3, 'Knife': 2, 'Lamp': 4, 'Laptop': 2, 'Motorbike': 6, 'Mug': 2, 'Pistol': 3, 'Rocket': 3, 'Skateboard': 3, 'Table': 3} 4\n",
      "2658 704\n",
      "classes 4\n",
      "/usr/local/lib/python3.9/dist-packages/torch/optim/lr_scheduler.py:139: UserWarning: Detected call of `lr_scheduler.step()` before `optimizer.step()`. In PyTorch 1.1.0 and later, you should call them in the opposite order: `optimizer.step()` before `lr_scheduler.step()`.  Failure to do this will result in PyTorch skipping the first value of the learning rate schedule. See more details at https://pytorch.org/docs/stable/optim.html#how-to-adjust-learning-rate\n",
      "  warnings.warn(\"Detected call of `lr_scheduler.step()` before `optimizer.step()`. \"\n",
      "[0: 0/83] train loss: 1.687849 accuracy: 0.109488\n",
      "[0: 0/83] \u001B[94mtest\u001B[0m loss: 1.390326 accuracy: 0.050737\n",
      "[0: 1/83] train loss: 1.508113 accuracy: 0.186325\n",
      "[0: 2/83] train loss: 1.463986 accuracy: 0.200862\n",
      "[0: 3/83] train loss: 1.346671 accuracy: 0.301325\n",
      "[0: 4/83] train loss: 1.258082 accuracy: 0.404375\n",
      "[0: 5/83] train loss: 1.166842 accuracy: 0.537000\n",
      "[0: 6/83] train loss: 1.105299 accuracy: 0.541300\n",
      "[0: 7/83] train loss: 0.965996 accuracy: 0.656300\n",
      "[0: 8/83] train loss: 0.939781 accuracy: 0.640513\n",
      "[0: 9/83] train loss: 0.905893 accuracy: 0.673162\n",
      "[0: 10/83] train loss: 0.783329 accuracy: 0.750112\n",
      "[0: 10/83] \u001B[94mtest\u001B[0m loss: 1.157457 accuracy: 0.587200\n",
      "[0: 11/83] train loss: 0.761946 accuracy: 0.774025\n",
      "[0: 12/83] train loss: 0.727742 accuracy: 0.782362\n",
      "[0: 13/83] train loss: 0.653486 accuracy: 0.806900\n",
      "[0: 14/83] train loss: 0.652847 accuracy: 0.777775\n",
      "[0: 15/83] train loss: 0.668327 accuracy: 0.771487\n",
      "[0: 16/83] train loss: 0.605191 accuracy: 0.797863\n",
      "[0: 17/83] train loss: 0.601333 accuracy: 0.804538\n",
      "[0: 18/83] train loss: 0.622272 accuracy: 0.784275\n",
      "[0: 19/83] train loss: 0.586092 accuracy: 0.787050\n",
      "[0: 20/83] train loss: 0.592530 accuracy: 0.790737\n",
      "[0: 20/83] \u001B[94mtest\u001B[0m loss: 0.742089 accuracy: 0.801300\n",
      "[0: 21/83] train loss: 0.543873 accuracy: 0.819025\n",
      "[0: 22/83] train loss: 0.547514 accuracy: 0.815163\n",
      "[0: 23/83] train loss: 0.520850 accuracy: 0.817125\n",
      "[0: 24/83] train loss: 0.544124 accuracy: 0.802900\n",
      "[0: 25/83] train loss: 0.525225 accuracy: 0.818462\n",
      "[0: 26/83] train loss: 0.552314 accuracy: 0.800937\n",
      "[0: 27/83] train loss: 0.511477 accuracy: 0.830712\n",
      "[0: 28/83] train loss: 0.516450 accuracy: 0.814288\n",
      "[0: 29/83] train loss: 0.460195 accuracy: 0.840287\n",
      "[0: 30/83] train loss: 0.539421 accuracy: 0.820037\n",
      "[0: 30/83] \u001B[94mtest\u001B[0m loss: 0.729063 accuracy: 0.651763\n",
      "[0: 31/83] train loss: 0.455319 accuracy: 0.841987\n",
      "[0: 32/83] train loss: 0.424197 accuracy: 0.859525\n",
      "[0: 33/83] train loss: 0.387156 accuracy: 0.871950\n",
      "[0: 34/83] train loss: 0.505723 accuracy: 0.809787\n",
      "[0: 35/83] train loss: 0.592786 accuracy: 0.788800\n",
      "[0: 36/83] train loss: 0.386526 accuracy: 0.873563\n",
      "[0: 37/83] train loss: 0.457591 accuracy: 0.835775\n",
      "[0: 38/83] train loss: 0.419098 accuracy: 0.847413\n",
      "[0: 39/83] train loss: 0.530520 accuracy: 0.812750\n",
      "[0: 40/83] train loss: 0.472950 accuracy: 0.820775\n",
      "[0: 40/83] \u001B[94mtest\u001B[0m loss: 0.534680 accuracy: 0.793863\n",
      "[0: 41/83] train loss: 0.511419 accuracy: 0.817875\n",
      "[0: 42/83] train loss: 0.478776 accuracy: 0.829612\n",
      "[0: 43/83] train loss: 0.445536 accuracy: 0.830812\n",
      "[0: 44/83] train loss: 0.384264 accuracy: 0.865287\n",
      "[0: 45/83] train loss: 0.480030 accuracy: 0.845975\n",
      "[0: 46/83] train loss: 0.407181 accuracy: 0.863762\n",
      "[0: 47/83] train loss: 0.453642 accuracy: 0.837950\n",
      "[0: 48/83] train loss: 0.449523 accuracy: 0.825750\n",
      "[0: 49/83] train loss: 0.424939 accuracy: 0.856313\n",
      "[0: 50/83] train loss: 0.431271 accuracy: 0.836413\n",
      "[0: 50/83] \u001B[94mtest\u001B[0m loss: 1.113706 accuracy: 0.587363\n",
      "[0: 51/83] train loss: 0.441831 accuracy: 0.840150\n",
      "[0: 52/83] train loss: 0.428583 accuracy: 0.842400\n",
      "[0: 53/83] train loss: 0.435754 accuracy: 0.849975\n",
      "[0: 54/83] train loss: 0.469838 accuracy: 0.830013\n",
      "[0: 55/83] train loss: 0.488931 accuracy: 0.828825\n",
      "[0: 56/83] train loss: 0.367116 accuracy: 0.871862\n",
      "[0: 57/83] train loss: 0.418025 accuracy: 0.854475\n",
      "[0: 58/83] train loss: 0.363684 accuracy: 0.869537\n",
      "[0: 59/83] train loss: 0.386595 accuracy: 0.872275\n",
      "[0: 60/83] train loss: 0.319498 accuracy: 0.888300\n",
      "[0: 60/83] \u001B[94mtest\u001B[0m loss: 0.482585 accuracy: 0.830812\n",
      "[0: 61/83] train loss: 0.372976 accuracy: 0.863150\n",
      "[0: 62/83] train loss: 0.374579 accuracy: 0.865125\n",
      "[0: 63/83] train loss: 0.304374 accuracy: 0.901012\n",
      "[0: 64/83] train loss: 0.372177 accuracy: 0.867938\n",
      "[0: 65/83] train loss: 0.480331 accuracy: 0.826850\n",
      "[0: 66/83] train loss: 0.330131 accuracy: 0.880637\n",
      "[0: 67/83] train loss: 0.361206 accuracy: 0.872275\n",
      "[0: 68/83] train loss: 0.308694 accuracy: 0.889675\n",
      "[0: 69/83] train loss: 0.349751 accuracy: 0.872025\n",
      "[0: 70/83] train loss: 0.325918 accuracy: 0.881238\n",
      "[0: 70/83] \u001B[94mtest\u001B[0m loss: 0.464231 accuracy: 0.817388\n",
      "[0: 71/83] train loss: 0.394965 accuracy: 0.855387\n",
      "[0: 72/83] train loss: 0.347092 accuracy: 0.880300\n",
      "[0: 73/83] train loss: 0.343333 accuracy: 0.874275\n",
      "[0: 74/83] train loss: 0.409372 accuracy: 0.861825\n",
      "[0: 75/83] train loss: 0.366902 accuracy: 0.871838\n",
      "[0: 76/83] train loss: 0.283748 accuracy: 0.906412\n",
      "[0: 77/83] train loss: 0.358634 accuracy: 0.869988\n",
      "[0: 78/83] train loss: 0.331494 accuracy: 0.886725\n",
      "[0: 79/83] train loss: 0.348888 accuracy: 0.875263\n",
      "[0: 80/83] train loss: 0.369186 accuracy: 0.862088\n",
      "[0: 80/83] \u001B[94mtest\u001B[0m loss: 0.660899 accuracy: 0.731363\n",
      "[0: 81/83] train loss: 0.349931 accuracy: 0.876375\n",
      "[0: 82/83] train loss: 0.378437 accuracy: 0.858950\n",
      "[0: 83/83] train loss: 0.791946 accuracy: 0.048113\n",
      "[1: 0/83] train loss: 0.437498 accuracy: 0.844400\n",
      "[1: 0/83] \u001B[94mtest\u001B[0m loss: 0.570367 accuracy: 0.762725\n",
      "[1: 1/83] train loss: 0.360092 accuracy: 0.877463\n",
      "[1: 2/83] train loss: 0.361009 accuracy: 0.872263\n",
      "[1: 3/83] train loss: 0.317558 accuracy: 0.888725\n",
      "[1: 4/83] train loss: 0.413978 accuracy: 0.848163\n",
      "[1: 5/83] train loss: 0.399597 accuracy: 0.854625\n",
      "[1: 6/83] train loss: 0.346635 accuracy: 0.872900\n",
      "[1: 7/83] train loss: 0.324051 accuracy: 0.890837\n",
      "[1: 8/83] train loss: 0.326866 accuracy: 0.877788\n",
      "[1: 9/83] train loss: 0.306489 accuracy: 0.898000\n",
      "[1: 10/83] train loss: 0.392587 accuracy: 0.856225\n",
      "[1: 10/83] \u001B[94mtest\u001B[0m loss: 0.507932 accuracy: 0.821913\n",
      "[1: 11/83] train loss: 0.312969 accuracy: 0.891088\n",
      "[1: 12/83] train loss: 0.358561 accuracy: 0.876363\n",
      "[1: 13/83] train loss: 0.376745 accuracy: 0.870038\n",
      "[1: 14/83] train loss: 0.305901 accuracy: 0.892050\n",
      "[1: 15/83] train loss: 0.353219 accuracy: 0.883775\n",
      "[1: 16/83] train loss: 0.307283 accuracy: 0.899900\n",
      "[1: 17/83] train loss: 0.314732 accuracy: 0.884162\n",
      "[1: 18/83] train loss: 0.306650 accuracy: 0.896775\n",
      "[1: 19/83] train loss: 0.447072 accuracy: 0.836237\n",
      "[1: 20/83] train loss: 0.326006 accuracy: 0.890387\n",
      "[1: 20/83] \u001B[94mtest\u001B[0m loss: 0.399161 accuracy: 0.833100\n",
      "[1: 21/83] train loss: 0.322937 accuracy: 0.882337\n",
      "[1: 22/83] train loss: 0.318402 accuracy: 0.886975\n",
      "[1: 23/83] train loss: 0.401813 accuracy: 0.866788\n",
      "[1: 24/83] train loss: 0.394835 accuracy: 0.858425\n",
      "[1: 25/83] train loss: 0.298097 accuracy: 0.899913\n",
      "[1: 26/83] train loss: 0.331963 accuracy: 0.884212\n",
      "[1: 27/83] train loss: 0.381280 accuracy: 0.851137\n",
      "[1: 28/83] train loss: 0.357114 accuracy: 0.862000\n",
      "[1: 29/83] train loss: 0.313546 accuracy: 0.884162\n",
      "[1: 30/83] train loss: 0.304844 accuracy: 0.897312\n",
      "[1: 30/83] \u001B[94mtest\u001B[0m loss: 0.428587 accuracy: 0.852912\n",
      "[1: 31/83] train loss: 0.304623 accuracy: 0.891837\n",
      "[1: 32/83] train loss: 0.314532 accuracy: 0.891300\n",
      "[1: 33/83] train loss: 0.271407 accuracy: 0.908687\n",
      "[1: 34/83] train loss: 0.379792 accuracy: 0.873025\n",
      "[1: 35/83] train loss: 0.294883 accuracy: 0.892500\n",
      "[1: 36/83] train loss: 0.359565 accuracy: 0.865738\n",
      "[1: 37/83] train loss: 0.313607 accuracy: 0.888100\n",
      "[1: 38/83] train loss: 0.280035 accuracy: 0.895875\n",
      "[1: 39/83] train loss: 0.327401 accuracy: 0.881950\n",
      "[1: 40/83] train loss: 0.307633 accuracy: 0.885700\n",
      "[1: 40/83] \u001B[94mtest\u001B[0m loss: 0.358457 accuracy: 0.863012\n",
      "[1: 41/83] train loss: 0.328876 accuracy: 0.881550\n",
      "[1: 42/83] train loss: 0.261769 accuracy: 0.905950\n",
      "[1: 43/83] train loss: 0.478477 accuracy: 0.850150\n",
      "[1: 44/83] train loss: 0.366299 accuracy: 0.867350\n",
      "[1: 45/83] train loss: 0.349876 accuracy: 0.867225\n",
      "[1: 46/83] train loss: 0.346231 accuracy: 0.891750\n",
      "[1: 47/83] train loss: 0.293576 accuracy: 0.892362\n",
      "[1: 48/83] train loss: 0.255970 accuracy: 0.907500\n",
      "[1: 49/83] train loss: 0.296744 accuracy: 0.893813\n",
      "[1: 50/83] train loss: 0.311434 accuracy: 0.886088\n",
      "[1: 50/83] \u001B[94mtest\u001B[0m loss: 0.269031 accuracy: 0.899600\n",
      "[1: 51/83] train loss: 0.307558 accuracy: 0.903550\n",
      "[1: 52/83] train loss: 0.331673 accuracy: 0.894038\n",
      "[1: 53/83] train loss: 0.306751 accuracy: 0.892587\n",
      "[1: 54/83] train loss: 0.268834 accuracy: 0.897100\n",
      "[1: 55/83] train loss: 0.321409 accuracy: 0.883975\n",
      "[1: 56/83] train loss: 0.286604 accuracy: 0.899587\n",
      "[1: 57/83] train loss: 0.304845 accuracy: 0.893813\n",
      "[1: 58/83] train loss: 0.324891 accuracy: 0.876062\n",
      "[1: 59/83] train loss: 0.277345 accuracy: 0.906650\n",
      "[1: 60/83] train loss: 0.257939 accuracy: 0.913400\n",
      "[1: 60/83] \u001B[94mtest\u001B[0m loss: 0.327243 accuracy: 0.878038\n",
      "[1: 61/83] train loss: 0.334401 accuracy: 0.876563\n",
      "[1: 62/83] train loss: 0.230512 accuracy: 0.923600\n",
      "[1: 63/83] train loss: 0.298167 accuracy: 0.896388\n",
      "[1: 64/83] train loss: 0.274912 accuracy: 0.904875\n",
      "[1: 65/83] train loss: 0.247860 accuracy: 0.909650\n",
      "[1: 66/83] train loss: 0.242773 accuracy: 0.912013\n",
      "[1: 67/83] train loss: 0.352322 accuracy: 0.879125\n",
      "[1: 68/83] train loss: 0.345362 accuracy: 0.880775\n",
      "[1: 69/83] train loss: 0.268674 accuracy: 0.906400\n",
      "[1: 70/83] train loss: 0.280121 accuracy: 0.901887\n",
      "[1: 70/83] \u001B[94mtest\u001B[0m loss: 0.347013 accuracy: 0.867025\n",
      "[1: 71/83] train loss: 0.248934 accuracy: 0.914362\n",
      "[1: 72/83] train loss: 0.272924 accuracy: 0.905400\n",
      "[1: 73/83] train loss: 0.305282 accuracy: 0.887075\n",
      "[1: 74/83] train loss: 0.206409 accuracy: 0.930013\n",
      "[1: 75/83] train loss: 0.303011 accuracy: 0.883900\n",
      "[1: 76/83] train loss: 0.279223 accuracy: 0.901250\n",
      "[1: 77/83] train loss: 0.329927 accuracy: 0.880487\n",
      "[1: 78/83] train loss: 0.302114 accuracy: 0.889763\n",
      "[1: 79/83] train loss: 0.333905 accuracy: 0.890038\n",
      "[1: 80/83] train loss: 0.252481 accuracy: 0.912275\n",
      "[1: 80/83] \u001B[94mtest\u001B[0m loss: 0.255685 accuracy: 0.915687\n",
      "[1: 81/83] train loss: 0.263363 accuracy: 0.910675\n",
      "[1: 82/83] train loss: 0.292613 accuracy: 0.901487\n",
      "[1: 83/83] train loss: 0.163637 accuracy: 0.059175\n",
      "22it [00:08,  2.70it/s]\n",
      "mIOU for class Chair: 0.6796701601753913\n"
     ]
    }
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "# Segmentation visualization"
   ],
   "metadata": {
    "id": "NLHbgBThJkqj"
   }
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting opencv-contrib-python\r\n",
      "  Downloading opencv_contrib_python-4.7.0.72-cp37-abi3-macosx_10_16_x86_64.whl (63.1 MB)\r\n",
      "\u001B[2K     \u001B[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001B[0m \u001B[32m63.1/63.1 MB\u001B[0m \u001B[31m10.6 MB/s\u001B[0m eta \u001B[36m0:00:00\u001B[0m00:01\u001B[0m00:01\u001B[0m\r\n",
      "\u001B[?25hRequirement already satisfied: numpy>=1.17.0 in /Users/xlsior/opt/anaconda3/envs/cs182-project-pointnet/lib/python3.8/site-packages (from opencv-contrib-python) (1.23.5)\r\n",
      "Installing collected packages: opencv-contrib-python\r\n",
      "Successfully installed opencv-contrib-python-4.7.0.72\r\n"
     ]
    }
   ],
   "source": [
    "!pip install opencv-contrib-python"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001B[31mERROR: Could not find a version that satisfies the requirement ctypes (from versions: none)\u001B[0m\u001B[31m\r\n",
      "\u001B[0m\u001B[31mERROR: No matching distribution found for ctypes\u001B[0m\u001B[31m\r\n",
      "\u001B[0m"
     ]
    }
   ],
   "source": [
    "!pip install ctypes"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "outputs": [],
   "source": [
    "!g++ -std=c++11 ./utils/render_balls_so.cpp -o ./utils/render_balls_so.so -shared -fPIC -O2 -D_GLIBCXX_USE_CXX11_ABI=0"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "source": [
    "!python ./utils/show_seg.py --model=./seg/seg_model_Chair_1.pth --dataset ./shapenetcore_partanno_segmentation_benchmark_v0/ --class_choice 'Chair' --idx 1"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "jep-oD_vHnkP",
    "outputId": "9017d8ce-dbba-404b-df24-62a9bb0e0940"
   },
   "execution_count": 20,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Traceback (most recent call last):\r\n",
      "  File \"./utils/show_seg.py\", line 2, in <module>\r\n",
      "    from show3d_balls import showpoints\r\n",
      "  File \"/Users/xlsior/Library/Mobile Documents/com~apple~CloudDocs/Berkeley/Semesters/Spring 2023/CS182/Project/cs182-project-pointnet/utils/show3d_balls.py\", line 25, in <module>\r\n",
      "    dll = np.ctypeslib.load_library('utils/render_balls_so', '.')\r\n",
      "  File \"/Users/xlsior/opt/anaconda3/envs/cs182-project-pointnet/lib/python3.8/site-packages/numpy/ctypeslib.py\", line 163, in load_library\r\n",
      "    raise OSError(\"no file with expected extension\")\r\n",
      "OSError: no file with expected extension\r\n"
     ]
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [],
   "metadata": {
    "id": "D-Y1-28jK6_P"
   },
   "execution_count": 8,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(2488, 3)\n"
     ]
    }
   ]
  }
 ],
 "metadata": {
  "colab": {
   "provenance": [],
   "collapsed_sections": [
    "XywsicDnKIUk",
    "a-zVW0CWJe0Q",
    "Hbrc5jUgJaD8"
   ]
  },
  "kernelspec": {
   "display_name": "Python 3",
   "name": "python3"
  },
  "accelerator": "GPU",
  "gpuClass": "standard"
 },
 "nbformat": 4,
 "nbformat_minor": 0
}