Training.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "cc6eafae-3f3f-4899-9566-1efc3b338b1f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/home/jupyter\n"
     ]
    }
   ],
   "source": [
    "%cd /home/jupyter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "a0752627-ffef-4218-9d46-48586a632b54",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--2023-05-03 22:43:08--  https://shapenet.cs.stanford.edu/media/modelnet40_normal_resampled.zip\n",
      "Resolving shapenet.cs.stanford.edu (shapenet.cs.stanford.edu)... 171.67.77.19\n",
      "Connecting to shapenet.cs.stanford.edu (shapenet.cs.stanford.edu)|171.67.77.19|:443... connected.\n",
      "WARNING: The certificate of ‘shapenet.cs.stanford.edu’ is not trusted.\n",
      "WARNING: The certificate of ‘shapenet.cs.stanford.edu’ has expired.\n",
      "The certificate has expired\n",
      "HTTP request sent, awaiting response... 200 OK\n",
      "Length: 1705117335 (1.6G) [application/zip]\n",
      "Saving to: ‘modelnet40_normal_resampled.zip’\n",
      "\n",
      "modelnet40_normal_r 100%[===================>]   1.59G  4.83MB/s    in 4m 24s  \n",
      "\n",
      "2023-05-03 22:47:32 (6.15 MB/s) - ‘modelnet40_normal_resampled.zip’ saved [1705117335/1705117335]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "!wget --no-check-certificate https://shapenet.cs.stanford.edu/media/modelnet40_normal_resampled.zip"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "17771671-7055-4b05-9044-c68ae66efbfa",
   "metadata": {},
   "outputs": [],
   "source": [
    "!unzip -q modelnet40_normal_resampled.zip"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "f131059a-d22e-45dc-8cc5-fe761abd7455",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cloning into '3DCTN'...\n",
      "remote: Enumerating objects: 234, done.\u001b[K\n",
      "remote: Counting objects: 100% (136/136), done.\u001b[K\n",
      "remote: Compressing objects: 100% (113/113), done.\u001b[K\n",
      "remote: Total 234 (delta 47), reused 14 (delta 2), pack-reused 98\u001b[K\n",
      "Receiving objects: 100% (234/234), 987.23 KiB | 18.98 MiB/s, done.\n",
      "Resolving deltas: 100% (71/71), done.\n"
     ]
    }
   ],
   "source": [
    "# Don't do this a second time\n",
    "# !git clone https://github.com/SanjayJosh/3DCTN.git"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "ebb905e9-64c0-4c00-acce-d6ab6776e35d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/home/jupyter/3DCTN\n"
     ]
    }
   ],
   "source": [
    "%cd 3DCTN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "6f0e9b41-17cc-4f76-afeb-3e12ffb6f85a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Branch 'feat' set up to track remote branch 'feat' from 'origin'.\n",
      "Switched to a new branch 'feat'\n"
     ]
    }
   ],
   "source": [
    "!git fetch origin\n",
    "!git checkout --track origin/feat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "e67213ba-879c-4055-b709-df7c826e9fde",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: numpy in /opt/conda/lib/python3.7/site-packages (from -r requirements.txt (line 1)) (1.21.6)\n",
      "Collecting msgpack-numpy\n",
      "  Downloading msgpack_numpy-0.4.8-py2.py3-none-any.whl (6.9 kB)\n",
      "Collecting lmdb\n",
      "  Downloading lmdb-1.4.1-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (294 kB)\n",
      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m294.4/294.4 kB\u001b[0m \u001b[31m9.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[?25hCollecting h5py\n",
      "  Downloading h5py-3.8.0-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (4.3 MB)\n",
      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m4.3/4.3 MB\u001b[0m \u001b[31m74.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m:00:01\u001b[0m\n",
      "\u001b[?25hCollecting einops\n",
      "  Downloading einops-0.6.1-py3-none-any.whl (42 kB)\n",
      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m42.2/42.2 kB\u001b[0m \u001b[31m7.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[?25hRequirement already satisfied: msgpack>=0.5.2 in /opt/conda/lib/python3.7/site-packages (from msgpack-numpy->-r requirements.txt (line 2)) (1.0.5)\n",
      "Installing collected packages: lmdb, msgpack-numpy, h5py, einops\n",
      "Successfully installed einops-0.6.1 h5py-3.8.0 lmdb-1.4.1 msgpack-numpy-0.4.8\n"
     ]
    }
   ],
   "source": [
    "!pip install -r requirements.txt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "7d116a8d-e915-4144-ad87-d6aff5a72012",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PARAMETER ...\n",
      "Namespace(batch_size=16, decay_rate=0.0001, epoch=300, gpu='0', learning_rate=0.01, log_dir='pointnet2_cls_msg_github', model='pointnet2_cls_msg', num_category=40, num_point=1024, optimizer='SGD', process_data=True, use_cpu=False, use_normals=True, use_uniform_sample=False)\n",
      "Load dataset ...\n",
      "The size of train data is 9843\n",
      "Load processed data from /home/jupyter/modelnet40_normal_resampled/modelnet40_train_1024pts.dat...\n",
      "The size of test data is 2468\n",
      "Load processed data from /home/jupyter/modelnet40_normal_resampled/modelnet40_test_1024pts.dat...\n",
      "Use pretrain model\n",
      "Epoch 1 (4/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.713415\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.74it/s]\n",
      "Test Instance Accuracy: 0.786694, Class Accuracy: 0.699254\n",
      "Best Instance Accuracy: 0.786694, Class Accuracy: 0.699254\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 2 (5/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.750203\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.56it/s]\n",
      "Test Instance Accuracy: 0.831048, Class Accuracy: 0.767748\n",
      "Best Instance Accuracy: 0.831048, Class Accuracy: 0.767748\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 3 (6/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.778557\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.64it/s]\n",
      "Test Instance Accuracy: 0.835887, Class Accuracy: 0.758669\n",
      "Best Instance Accuracy: 0.835887, Class Accuracy: 0.767748\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 4 (7/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.800813\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.77it/s]\n",
      "Test Instance Accuracy: 0.858468, Class Accuracy: 0.789360\n",
      "Best Instance Accuracy: 0.858468, Class Accuracy: 0.789360\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 5 (8/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.808943\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.77it/s]\n",
      "Test Instance Accuracy: 0.856855, Class Accuracy: 0.791490\n",
      "Best Instance Accuracy: 0.858468, Class Accuracy: 0.791490\n",
      "Epoch 6 (9/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.819411\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.53it/s]\n",
      "Test Instance Accuracy: 0.865726, Class Accuracy: 0.816532\n",
      "Best Instance Accuracy: 0.865726, Class Accuracy: 0.816532\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 7 (10/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.834248\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.48it/s]\n",
      "Test Instance Accuracy: 0.870968, Class Accuracy: 0.825551\n",
      "Best Instance Accuracy: 0.870968, Class Accuracy: 0.825551\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 8 (11/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:18<00:00,  3.10it/s]\n",
      "Train Instance Accuracy: 0.840447\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.73it/s]\n",
      "Test Instance Accuracy: 0.880242, Class Accuracy: 0.828971\n",
      "Best Instance Accuracy: 0.880242, Class Accuracy: 0.828971\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 9 (12/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.847256\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.74it/s]\n",
      "Test Instance Accuracy: 0.897177, Class Accuracy: 0.852496\n",
      "Best Instance Accuracy: 0.897177, Class Accuracy: 0.852496\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 10 (13/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.860874\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.55it/s]\n",
      "Test Instance Accuracy: 0.885887, Class Accuracy: 0.840993\n",
      "Best Instance Accuracy: 0.897177, Class Accuracy: 0.852496\n",
      "Epoch 11 (14/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.860366\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.57it/s]\n",
      "Test Instance Accuracy: 0.887903, Class Accuracy: 0.840418\n",
      "Best Instance Accuracy: 0.897177, Class Accuracy: 0.852496\n",
      "Epoch 12 (15/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:18<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.871850\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.77it/s]\n",
      "Test Instance Accuracy: 0.894355, Class Accuracy: 0.839804\n",
      "Best Instance Accuracy: 0.897177, Class Accuracy: 0.852496\n",
      "Epoch 13 (16/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.875000\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.57it/s]\n",
      "Test Instance Accuracy: 0.899597, Class Accuracy: 0.850729\n",
      "Best Instance Accuracy: 0.899597, Class Accuracy: 0.852496\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 14 (17/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:20<00:00,  3.06it/s]\n",
      "Train Instance Accuracy: 0.880386\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.57it/s]\n",
      "Test Instance Accuracy: 0.880645, Class Accuracy: 0.854723\n",
      "Best Instance Accuracy: 0.899597, Class Accuracy: 0.854723\n",
      "Epoch 15 (18/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.886585\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.52it/s]\n",
      "Test Instance Accuracy: 0.903629, Class Accuracy: 0.867693\n",
      "Best Instance Accuracy: 0.903629, Class Accuracy: 0.867693\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 16 (19/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.13it/s]\n",
      "Train Instance Accuracy: 0.891972\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.78it/s]\n",
      "Test Instance Accuracy: 0.895968, Class Accuracy: 0.837037\n",
      "Best Instance Accuracy: 0.903629, Class Accuracy: 0.867693\n",
      "Epoch 17 (20/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.888720\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.75it/s]\n",
      "Test Instance Accuracy: 0.910887, Class Accuracy: 0.868753\n",
      "Best Instance Accuracy: 0.910887, Class Accuracy: 0.868753\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 18 (21/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.896850\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.76it/s]\n",
      "Test Instance Accuracy: 0.906048, Class Accuracy: 0.864631\n",
      "Best Instance Accuracy: 0.910887, Class Accuracy: 0.868753\n",
      "Epoch 19 (22/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.13it/s]\n",
      "Train Instance Accuracy: 0.899695\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.55it/s]\n",
      "Test Instance Accuracy: 0.903226, Class Accuracy: 0.871432\n",
      "Best Instance Accuracy: 0.910887, Class Accuracy: 0.871432\n",
      "Epoch 20 (23/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.13it/s]\n",
      "Train Instance Accuracy: 0.902236\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.79it/s]\n",
      "Test Instance Accuracy: 0.900403, Class Accuracy: 0.858859\n",
      "Best Instance Accuracy: 0.910887, Class Accuracy: 0.871432\n",
      "Epoch 21 (24/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.906606\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.78it/s]\n",
      "Test Instance Accuracy: 0.908871, Class Accuracy: 0.882317\n",
      "Best Instance Accuracy: 0.910887, Class Accuracy: 0.882317\n",
      "Epoch 22 (25/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:18<00:00,  3.10it/s]\n",
      "Train Instance Accuracy: 0.906504\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.76it/s]\n",
      "Test Instance Accuracy: 0.908468, Class Accuracy: 0.874101\n",
      "Best Instance Accuracy: 0.910887, Class Accuracy: 0.882317\n",
      "Epoch 23 (26/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.908333\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.53it/s]\n",
      "Test Instance Accuracy: 0.904435, Class Accuracy: 0.880124\n",
      "Best Instance Accuracy: 0.910887, Class Accuracy: 0.882317\n",
      "Epoch 24 (27/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.13it/s]\n",
      "Train Instance Accuracy: 0.913313\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.72it/s]\n",
      "Test Instance Accuracy: 0.899597, Class Accuracy: 0.870362\n",
      "Best Instance Accuracy: 0.910887, Class Accuracy: 0.882317\n",
      "Epoch 25 (28/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.916260\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.77it/s]\n",
      "Test Instance Accuracy: 0.913710, Class Accuracy: 0.868354\n",
      "Best Instance Accuracy: 0.913710, Class Accuracy: 0.882317\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 26 (29/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.916159\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.78it/s]\n",
      "Test Instance Accuracy: 0.905242, Class Accuracy: 0.873212\n",
      "Best Instance Accuracy: 0.913710, Class Accuracy: 0.882317\n",
      "Epoch 27 (30/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.924695\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.54it/s]\n",
      "Test Instance Accuracy: 0.902823, Class Accuracy: 0.876523\n",
      "Best Instance Accuracy: 0.913710, Class Accuracy: 0.882317\n",
      "Epoch 28 (31/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.918598\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.60it/s]\n",
      "Test Instance Accuracy: 0.904839, Class Accuracy: 0.879793\n",
      "Best Instance Accuracy: 0.913710, Class Accuracy: 0.882317\n",
      "Epoch 29 (32/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.922256\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.77it/s]\n",
      "Test Instance Accuracy: 0.908468, Class Accuracy: 0.881293\n",
      "Best Instance Accuracy: 0.913710, Class Accuracy: 0.882317\n",
      "Epoch 30 (33/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:18<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.922561\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.77it/s]\n",
      "Test Instance Accuracy: 0.921774, Class Accuracy: 0.889711\n",
      "Best Instance Accuracy: 0.921774, Class Accuracy: 0.889711\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 31 (34/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.925305\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.54it/s]\n",
      "Test Instance Accuracy: 0.917339, Class Accuracy: 0.889793\n",
      "Best Instance Accuracy: 0.921774, Class Accuracy: 0.889793\n",
      "Epoch 32 (35/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.13it/s]\n",
      "Train Instance Accuracy: 0.927846\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.57it/s]\n",
      "Test Instance Accuracy: 0.911694, Class Accuracy: 0.886458\n",
      "Best Instance Accuracy: 0.921774, Class Accuracy: 0.889793\n",
      "Epoch 33 (36/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.13it/s]\n",
      "Train Instance Accuracy: 0.925915\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.76it/s]\n",
      "Test Instance Accuracy: 0.903629, Class Accuracy: 0.873608\n",
      "Best Instance Accuracy: 0.921774, Class Accuracy: 0.889793\n",
      "Epoch 34 (37/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:18<00:00,  3.09it/s]\n",
      "Train Instance Accuracy: 0.928049\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.76it/s]\n",
      "Test Instance Accuracy: 0.909274, Class Accuracy: 0.879166\n",
      "Best Instance Accuracy: 0.921774, Class Accuracy: 0.889793\n",
      "Epoch 35 (38/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.931402\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.52it/s]\n",
      "Test Instance Accuracy: 0.922984, Class Accuracy: 0.891596\n",
      "Best Instance Accuracy: 0.922984, Class Accuracy: 0.891596\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 36 (39/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.928455\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.56it/s]\n",
      "Test Instance Accuracy: 0.914919, Class Accuracy: 0.881196\n",
      "Best Instance Accuracy: 0.922984, Class Accuracy: 0.891596\n",
      "Epoch 37 (40/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.13it/s]\n",
      "Train Instance Accuracy: 0.935467\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.78it/s]\n",
      "Test Instance Accuracy: 0.910484, Class Accuracy: 0.898372\n",
      "Best Instance Accuracy: 0.922984, Class Accuracy: 0.898372\n",
      "Epoch 38 (41/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.933028\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.73it/s]\n",
      "Test Instance Accuracy: 0.915323, Class Accuracy: 0.891852\n",
      "Best Instance Accuracy: 0.922984, Class Accuracy: 0.898372\n",
      "Epoch 39 (42/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.935671\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.52it/s]\n",
      "Test Instance Accuracy: 0.914919, Class Accuracy: 0.884546\n",
      "Best Instance Accuracy: 0.922984, Class Accuracy: 0.898372\n",
      "Epoch 40 (43/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.935061\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.52it/s]\n",
      "Test Instance Accuracy: 0.918548, Class Accuracy: 0.874246\n",
      "Best Instance Accuracy: 0.922984, Class Accuracy: 0.898372\n",
      "Epoch 41 (44/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:18<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.937500\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.70it/s]\n",
      "Test Instance Accuracy: 0.922581, Class Accuracy: 0.887976\n",
      "Best Instance Accuracy: 0.922984, Class Accuracy: 0.898372\n",
      "Epoch 42 (45/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.09it/s]\n",
      "Train Instance Accuracy: 0.940041\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.72it/s]\n",
      "Test Instance Accuracy: 0.904839, Class Accuracy: 0.877665\n",
      "Best Instance Accuracy: 0.922984, Class Accuracy: 0.898372\n",
      "Epoch 43 (46/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:18<00:00,  3.10it/s]\n",
      "Train Instance Accuracy: 0.939634\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.50it/s]\n",
      "Test Instance Accuracy: 0.913710, Class Accuracy: 0.873067\n",
      "Best Instance Accuracy: 0.922984, Class Accuracy: 0.898372\n",
      "Epoch 44 (47/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.938313\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.57it/s]\n",
      "Test Instance Accuracy: 0.920565, Class Accuracy: 0.883666\n",
      "Best Instance Accuracy: 0.922984, Class Accuracy: 0.898372\n",
      "Epoch 45 (48/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.13it/s]\n",
      "Train Instance Accuracy: 0.942276\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.76it/s]\n",
      "Test Instance Accuracy: 0.913710, Class Accuracy: 0.893467\n",
      "Best Instance Accuracy: 0.922984, Class Accuracy: 0.898372\n",
      "Epoch 46 (49/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.944919\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.74it/s]\n",
      "Test Instance Accuracy: 0.912903, Class Accuracy: 0.881069\n",
      "Best Instance Accuracy: 0.922984, Class Accuracy: 0.898372\n",
      "Epoch 47 (50/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.944614\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.53it/s]\n",
      "Test Instance Accuracy: 0.915323, Class Accuracy: 0.893067\n",
      "Best Instance Accuracy: 0.922984, Class Accuracy: 0.898372\n",
      "Epoch 48 (51/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.945630\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.56it/s]\n",
      "Test Instance Accuracy: 0.915323, Class Accuracy: 0.877405\n",
      "Best Instance Accuracy: 0.922984, Class Accuracy: 0.898372\n",
      "Epoch 49 (52/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.13it/s]\n",
      "Train Instance Accuracy: 0.948171\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.77it/s]\n",
      "Test Instance Accuracy: 0.925403, Class Accuracy: 0.894360\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 50 (53/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:18<00:00,  3.10it/s]\n",
      "Train Instance Accuracy: 0.949187\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.71it/s]\n",
      "Test Instance Accuracy: 0.915726, Class Accuracy: 0.892105\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 51 (54/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.948272\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.53it/s]\n",
      "Test Instance Accuracy: 0.918145, Class Accuracy: 0.890563\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 52 (55/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.947358\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.52it/s]\n",
      "Test Instance Accuracy: 0.909274, Class Accuracy: 0.883865\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 53 (56/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.13it/s]\n",
      "Train Instance Accuracy: 0.946037\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.78it/s]\n",
      "Test Instance Accuracy: 0.913710, Class Accuracy: 0.886722\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 54 (57/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.946646\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.78it/s]\n",
      "Test Instance Accuracy: 0.915726, Class Accuracy: 0.886767\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 55 (58/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.13it/s]\n",
      "Train Instance Accuracy: 0.949695\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.54it/s]\n",
      "Test Instance Accuracy: 0.918548, Class Accuracy: 0.880028\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 56 (59/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.13it/s]\n",
      "Train Instance Accuracy: 0.952541\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.55it/s]\n",
      "Test Instance Accuracy: 0.920565, Class Accuracy: 0.897550\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 57 (60/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.13it/s]\n",
      "Train Instance Accuracy: 0.948679\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.75it/s]\n",
      "Test Instance Accuracy: 0.911290, Class Accuracy: 0.888729\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 58 (61/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.951118\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.78it/s]\n",
      "Test Instance Accuracy: 0.911694, Class Accuracy: 0.895070\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 59 (62/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.951220\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.55it/s]\n",
      "Test Instance Accuracy: 0.914919, Class Accuracy: 0.892464\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 60 (63/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.955488\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.54it/s]\n",
      "Test Instance Accuracy: 0.915323, Class Accuracy: 0.888534\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 61 (64/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.953659\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.75it/s]\n",
      "Test Instance Accuracy: 0.921371, Class Accuracy: 0.892278\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 62 (65/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.955386\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.77it/s]\n",
      "Test Instance Accuracy: 0.920161, Class Accuracy: 0.887377\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 63 (66/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.957825\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.59it/s]\n",
      "Test Instance Accuracy: 0.922984, Class Accuracy: 0.888419\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 64 (67/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.13it/s]\n",
      "Train Instance Accuracy: 0.956504\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.57it/s]\n",
      "Test Instance Accuracy: 0.911290, Class Accuracy: 0.881064\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 65 (68/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.13it/s]\n",
      "Train Instance Accuracy: 0.958130\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.78it/s]\n",
      "Test Instance Accuracy: 0.921371, Class Accuracy: 0.885945\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 66 (69/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.958333\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.75it/s]\n",
      "Test Instance Accuracy: 0.905242, Class Accuracy: 0.886361\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 67 (70/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.959045\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.58it/s]\n",
      "Test Instance Accuracy: 0.920968, Class Accuracy: 0.884442\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 68 (71/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.960569\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.54it/s]\n",
      "Test Instance Accuracy: 0.908871, Class Accuracy: 0.880132\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 69 (72/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.959654\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.77it/s]\n",
      "Test Instance Accuracy: 0.923387, Class Accuracy: 0.895102\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 70 (73/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.964126\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.77it/s]\n",
      "Test Instance Accuracy: 0.922984, Class Accuracy: 0.893268\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 71 (74/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.960569\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.66it/s]\n",
      "Test Instance Accuracy: 0.904839, Class Accuracy: 0.890083\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 72 (75/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.960976\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.55it/s]\n",
      "Test Instance Accuracy: 0.909274, Class Accuracy: 0.876642\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 73 (76/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.961280\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.76it/s]\n",
      "Test Instance Accuracy: 0.910484, Class Accuracy: 0.885271\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 74 (77/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.960569\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.77it/s]\n",
      "Test Instance Accuracy: 0.911290, Class Accuracy: 0.882236\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 75 (78/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.963415\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.75it/s]\n",
      "Test Instance Accuracy: 0.924597, Class Accuracy: 0.891549\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 76 (79/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.966057\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.54it/s]\n",
      "Test Instance Accuracy: 0.919758, Class Accuracy: 0.881705\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 77 (80/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.961687\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.75it/s]\n",
      "Test Instance Accuracy: 0.911694, Class Accuracy: 0.885147\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 78 (81/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.962805\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.76it/s]\n",
      "Test Instance Accuracy: 0.915726, Class Accuracy: 0.880814\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 79 (82/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:18<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.963110\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.74it/s]\n",
      "Test Instance Accuracy: 0.915726, Class Accuracy: 0.878324\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.898372\n",
      "Epoch 80 (83/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.965650\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.56it/s]\n",
      "Test Instance Accuracy: 0.918145, Class Accuracy: 0.899232\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.899232\n",
      "Epoch 81 (84/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.964939\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.69it/s]\n",
      "Test Instance Accuracy: 0.921774, Class Accuracy: 0.892234\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.899232\n",
      "Epoch 82 (85/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.966362\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.76it/s]\n",
      "Test Instance Accuracy: 0.911290, Class Accuracy: 0.876138\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.899232\n",
      "Epoch 83 (86/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.961585\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.76it/s]\n",
      "Test Instance Accuracy: 0.920565, Class Accuracy: 0.888344\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.899232\n",
      "Epoch 84 (87/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.969004\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.55it/s]\n",
      "Test Instance Accuracy: 0.913710, Class Accuracy: 0.889540\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.899232\n",
      "Epoch 85 (88/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.965447\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.61it/s]\n",
      "Test Instance Accuracy: 0.915323, Class Accuracy: 0.891732\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.899232\n",
      "Epoch 86 (89/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.969411\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.76it/s]\n",
      "Test Instance Accuracy: 0.903226, Class Accuracy: 0.885558\n",
      "Best Instance Accuracy: 0.925403, Class Accuracy: 0.899232\n",
      "Epoch 87 (90/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:18<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.965142\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.73it/s]\n",
      "Test Instance Accuracy: 0.928226, Class Accuracy: 0.900286\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.900286\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 88 (91/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.967378\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.56it/s]\n",
      "Test Instance Accuracy: 0.915323, Class Accuracy: 0.885411\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.900286\n",
      "Epoch 89 (92/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.967886\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.55it/s]\n",
      "Test Instance Accuracy: 0.914113, Class Accuracy: 0.887537\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.900286\n",
      "Epoch 90 (93/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.971341\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.74it/s]\n",
      "Test Instance Accuracy: 0.900403, Class Accuracy: 0.882317\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.900286\n",
      "Epoch 91 (94/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.969817\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.74it/s]\n",
      "Test Instance Accuracy: 0.916129, Class Accuracy: 0.894927\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.900286\n",
      "Epoch 92 (95/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.968089\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.53it/s]\n",
      "Test Instance Accuracy: 0.914919, Class Accuracy: 0.896631\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.900286\n",
      "Epoch 93 (96/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.969614\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.53it/s]\n",
      "Test Instance Accuracy: 0.924597, Class Accuracy: 0.902448\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 94 (97/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.969919\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.77it/s]\n",
      "Test Instance Accuracy: 0.914919, Class Accuracy: 0.887280\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 95 (98/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:18<00:00,  3.10it/s]\n",
      "Train Instance Accuracy: 0.970630\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.72it/s]\n",
      "Test Instance Accuracy: 0.915726, Class Accuracy: 0.896330\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 96 (99/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.971037\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.54it/s]\n",
      "Test Instance Accuracy: 0.918952, Class Accuracy: 0.896296\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 97 (100/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.971850\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.53it/s]\n",
      "Test Instance Accuracy: 0.923387, Class Accuracy: 0.892223\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 98 (101/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.971850\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.75it/s]\n",
      "Test Instance Accuracy: 0.916532, Class Accuracy: 0.881345\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 99 (102/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.972866\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.74it/s]\n",
      "Test Instance Accuracy: 0.914113, Class Accuracy: 0.885154\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 100 (103/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.970833\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.54it/s]\n",
      "Test Instance Accuracy: 0.914919, Class Accuracy: 0.889711\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 101 (104/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.974085\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.52it/s]\n",
      "Test Instance Accuracy: 0.922177, Class Accuracy: 0.892572\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 102 (105/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.973069\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.77it/s]\n",
      "Test Instance Accuracy: 0.909677, Class Accuracy: 0.885400\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 103 (106/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.972459\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.76it/s]\n",
      "Test Instance Accuracy: 0.920161, Class Accuracy: 0.899970\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 104 (107/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.972866\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.54it/s]\n",
      "Test Instance Accuracy: 0.915726, Class Accuracy: 0.899771\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 105 (108/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.973476\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.50it/s]\n",
      "Test Instance Accuracy: 0.918145, Class Accuracy: 0.888140\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 106 (109/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.975203\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.81it/s]\n",
      "Test Instance Accuracy: 0.916935, Class Accuracy: 0.899375\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 107 (110/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.978354\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.79it/s]\n",
      "Test Instance Accuracy: 0.918952, Class Accuracy: 0.893030\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 108 (111/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.13it/s]\n",
      "Train Instance Accuracy: 0.976321\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.55it/s]\n",
      "Test Instance Accuracy: 0.917339, Class Accuracy: 0.892851\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 109 (112/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.973069\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.54it/s]\n",
      "Test Instance Accuracy: 0.911694, Class Accuracy: 0.889869\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 110 (113/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.13it/s]\n",
      "Train Instance Accuracy: 0.976220\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.77it/s]\n",
      "Test Instance Accuracy: 0.906452, Class Accuracy: 0.877544\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 111 (114/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:17<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.976626\n",
      "100%|█████████████████████████████████████████| 155/155 [00:26<00:00,  5.78it/s]\n",
      "Test Instance Accuracy: 0.919355, Class Accuracy: 0.887954\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 112 (115/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:16<00:00,  3.12it/s]\n",
      "Train Instance Accuracy: 0.978252\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.40it/s]\n",
      "Test Instance Accuracy: 0.907661, Class Accuracy: 0.884579\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 113 (116/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:23<00:00,  3.02it/s]\n",
      "Train Instance Accuracy: 0.975203\n",
      "100%|█████████████████████████████████████████| 155/155 [00:29<00:00,  5.21it/s]\n",
      "Test Instance Accuracy: 0.912097, Class Accuracy: 0.876951\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 114 (117/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:23<00:00,  3.02it/s]\n",
      "Train Instance Accuracy: 0.977744\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.36it/s]\n",
      "Test Instance Accuracy: 0.916532, Class Accuracy: 0.886354\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 115 (118/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:20<00:00,  3.07it/s]\n",
      "Train Instance Accuracy: 0.975203\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.64it/s]\n",
      "Test Instance Accuracy: 0.908065, Class Accuracy: 0.890381\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 116 (119/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.977642\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.67it/s]\n",
      "Test Instance Accuracy: 0.907661, Class Accuracy: 0.892931\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 117 (120/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.977744\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.67it/s]\n",
      "Test Instance Accuracy: 0.906048, Class Accuracy: 0.897492\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 118 (121/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.09it/s]\n",
      "Train Instance Accuracy: 0.978252\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.47it/s]\n",
      "Test Instance Accuracy: 0.916532, Class Accuracy: 0.897933\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 119 (122/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.980386\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.43it/s]\n",
      "Test Instance Accuracy: 0.916532, Class Accuracy: 0.888519\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 120 (123/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:18<00:00,  3.10it/s]\n",
      "Train Instance Accuracy: 0.980488\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.64it/s]\n",
      "Test Instance Accuracy: 0.908065, Class Accuracy: 0.878733\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 121 (124/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.977033\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.66it/s]\n",
      "Test Instance Accuracy: 0.916129, Class Accuracy: 0.887528\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 122 (125/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.976016\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.68it/s]\n",
      "Test Instance Accuracy: 0.912903, Class Accuracy: 0.887529\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 123 (126/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.978354\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.46it/s]\n",
      "Test Instance Accuracy: 0.914113, Class Accuracy: 0.896439\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 124 (127/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.09it/s]\n",
      "Train Instance Accuracy: 0.979472\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.44it/s]\n",
      "Test Instance Accuracy: 0.924597, Class Accuracy: 0.894929\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 125 (128/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.978659\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.64it/s]\n",
      "Test Instance Accuracy: 0.917339, Class Accuracy: 0.896598\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 126 (129/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.979675\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.66it/s]\n",
      "Test Instance Accuracy: 0.918548, Class Accuracy: 0.882745\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 127 (130/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.977744\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.64it/s]\n",
      "Test Instance Accuracy: 0.915726, Class Accuracy: 0.885859\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 128 (131/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.09it/s]\n",
      "Train Instance Accuracy: 0.979675\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.45it/s]\n",
      "Test Instance Accuracy: 0.921774, Class Accuracy: 0.901479\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 129 (132/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.09it/s]\n",
      "Train Instance Accuracy: 0.980894\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.46it/s]\n",
      "Test Instance Accuracy: 0.924194, Class Accuracy: 0.892414\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 130 (133/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:18<00:00,  3.10it/s]\n",
      "Train Instance Accuracy: 0.980183\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.43it/s]\n",
      "Test Instance Accuracy: 0.914919, Class Accuracy: 0.883949\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902448\n",
      "Epoch 131 (134/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:20<00:00,  3.07it/s]\n",
      "Train Instance Accuracy: 0.980691\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.67it/s]\n",
      "Test Instance Accuracy: 0.917339, Class Accuracy: 0.902639\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902639\n",
      "Epoch 132 (135/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:20<00:00,  3.07it/s]\n",
      "Train Instance Accuracy: 0.981707\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.68it/s]\n",
      "Test Instance Accuracy: 0.916129, Class Accuracy: 0.885230\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902639\n",
      "Epoch 133 (136/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.09it/s]\n",
      "Train Instance Accuracy: 0.981911\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.44it/s]\n",
      "Test Instance Accuracy: 0.922177, Class Accuracy: 0.895729\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902639\n",
      "Epoch 134 (137/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.981504\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.46it/s]\n",
      "Test Instance Accuracy: 0.918548, Class Accuracy: 0.890238\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902639\n",
      "Epoch 135 (138/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.09it/s]\n",
      "Train Instance Accuracy: 0.984045\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.66it/s]\n",
      "Test Instance Accuracy: 0.915323, Class Accuracy: 0.888201\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902639\n",
      "Epoch 136 (139/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:20<00:00,  3.07it/s]\n",
      "Train Instance Accuracy: 0.981911\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.65it/s]\n",
      "Test Instance Accuracy: 0.914919, Class Accuracy: 0.886931\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902639\n",
      "Epoch 137 (140/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:20<00:00,  3.07it/s]\n",
      "Train Instance Accuracy: 0.982114\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.60it/s]\n",
      "Test Instance Accuracy: 0.915323, Class Accuracy: 0.893712\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902639\n",
      "Epoch 138 (141/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.983638\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.42it/s]\n",
      "Test Instance Accuracy: 0.912097, Class Accuracy: 0.883402\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902639\n",
      "Epoch 139 (142/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.982724\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.38it/s]\n",
      "Test Instance Accuracy: 0.918145, Class Accuracy: 0.891188\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902639\n",
      "Epoch 140 (143/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.983333\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.56it/s]\n",
      "Test Instance Accuracy: 0.918952, Class Accuracy: 0.897541\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902639\n",
      "Epoch 141 (144/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.985874\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.61it/s]\n",
      "Test Instance Accuracy: 0.910484, Class Accuracy: 0.898534\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902639\n",
      "Epoch 142 (145/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:20<00:00,  3.07it/s]\n",
      "Train Instance Accuracy: 0.981301\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.63it/s]\n",
      "Test Instance Accuracy: 0.916129, Class Accuracy: 0.890266\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902639\n",
      "Epoch 143 (146/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.986280\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.42it/s]\n",
      "Test Instance Accuracy: 0.920161, Class Accuracy: 0.895965\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902639\n",
      "Epoch 144 (147/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.983740\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.43it/s]\n",
      "Test Instance Accuracy: 0.918952, Class Accuracy: 0.898958\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902639\n",
      "Epoch 145 (148/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.984858\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.44it/s]\n",
      "Test Instance Accuracy: 0.920968, Class Accuracy: 0.901469\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902639\n",
      "Epoch 146 (149/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.09it/s]\n",
      "Train Instance Accuracy: 0.984248\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.65it/s]\n",
      "Test Instance Accuracy: 0.914919, Class Accuracy: 0.892835\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902639\n",
      "Epoch 147 (150/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:20<00:00,  3.07it/s]\n",
      "Train Instance Accuracy: 0.983537\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.52it/s]\n",
      "Test Instance Accuracy: 0.915323, Class Accuracy: 0.886592\n",
      "Best Instance Accuracy: 0.928226, Class Accuracy: 0.902639\n",
      "Epoch 148 (151/300):\n",
      " 59%|████████████████████████▎                | 365/615 [01:58<01:20,  3.12it/s]^C\n",
      " 59%|████████████████████████▎                | 365/615 [01:58<01:21,  3.08it/s]\n",
      "Traceback (most recent call last):\n",
      "  File \"train_classification.py\", line 262, in <module>\n",
      "    main(args)\n",
      "  File \"train_classification.py\", line 207, in main\n",
      "    pred, trans_feat = classifier(points)\n",
      "  File \"/opt/conda/lib/python3.7/site-packages/torch/nn/modules/module.py\", line 1194, in _call_impl\n",
      "    return forward_call(*input, **kwargs)\n",
      "  File \"/home/jupyter/3DCTN/models/pointnet2_cls_msg.py\", line 36, in forward\n",
      "    l2_xyz, l2_points = self.sa2(g1_xyz, g1_points)\n",
      "  File \"/opt/conda/lib/python3.7/site-packages/torch/nn/modules/module.py\", line 1194, in _call_impl\n",
      "    return forward_call(*input, **kwargs)\n",
      "  File \"/home/jupyter/3DCTN/models/pointnet2_utils.py\", line 454, in forward\n",
      "    fps_index=farthest_point_sample(xyz, S)\n",
      "  File \"/home/jupyter/3DCTN/models/pointnet2_utils.py\", line 75, in farthest_point_sample\n",
      "    centroids = torch.zeros(B, npoint, dtype=torch.long).to(device)\n",
      "KeyboardInterrupt\n"
     ]
    }
   ],
   "source": [
    "!python train_classification.py --use_normals --model pointnet2_cls_msg --log_dir pointnet2_cls_msg_github --learning_rate 0.01 --batch_size 16 --optimizer SGD --epoch 300 --process_data\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "a2bf578f-498f-429c-ba7a-c0a887a89527",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tail: cannot open 'nohup.out' for reading: Permission denied\n",
      "tail: no files remaining\n"
     ]
    }
   ],
   "source": [
    "!tail -f nohup.out"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "15190b92-ec30-4d21-b1de-7e96ea407b5f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PARAMETER ...\n",
      "Namespace(batch_size=16, decay_rate=0.0001, epoch=300, gpu='0', learning_rate=0.01, log_dir='pointnet2_cls_msg_github', model='pointnet2_cls_msg', num_category=40, num_point=1024, optimizer='SGD', process_data=True, use_cpu=False, use_normals=True, use_uniform_sample=False)\n",
      "Load dataset ...\n",
      "The size of train data is 9843\n",
      "Load processed data from /home/jupyter/modelnet40_normal_resampled/modelnet40_train_1024pts.dat...\n",
      "The size of test data is 2468\n",
      "Load processed data from /home/jupyter/modelnet40_normal_resampled/modelnet40_test_1024pts.dat...\n",
      "No existing model, starting training from scratch...\n",
      "Epoch 1 (1/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.09it/s]\n",
      "Train Instance Accuracy: 0.393496\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.60it/s]\n",
      "Test Instance Accuracy: 0.537097, Class Accuracy: 0.422809\n",
      "Best Instance Accuracy: 0.537097, Class Accuracy: 0.422809\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 2 (2/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:18<00:00,  3.10it/s]\n",
      "Train Instance Accuracy: 0.583537\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.71it/s]\n",
      "Test Instance Accuracy: 0.643952, Class Accuracy: 0.503122\n",
      "Best Instance Accuracy: 0.643952, Class Accuracy: 0.503122\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 3 (3/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:19<00:00,  3.08it/s]\n",
      "Train Instance Accuracy: 0.666565\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.69it/s]\n",
      "Test Instance Accuracy: 0.736694, Class Accuracy: 0.650207\n",
      "Best Instance Accuracy: 0.736694, Class Accuracy: 0.650207\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 4 (4/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:18<00:00,  3.09it/s]\n",
      "Train Instance Accuracy: 0.717378\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.50it/s]\n",
      "Test Instance Accuracy: 0.779839, Class Accuracy: 0.708641\n",
      "Best Instance Accuracy: 0.779839, Class Accuracy: 0.708641\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 5 (5/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:18<00:00,  3.11it/s]\n",
      "Train Instance Accuracy: 0.745935\n",
      "100%|█████████████████████████████████████████| 155/155 [00:28<00:00,  5.49it/s]\n",
      "Test Instance Accuracy: 0.825403, Class Accuracy: 0.768338\n",
      "Best Instance Accuracy: 0.825403, Class Accuracy: 0.768338\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 6 (6/300):\n",
      "100%|█████████████████████████████████████████| 615/615 [03:18<00:00,  3.10it/s]\n",
      "Train Instance Accuracy: 0.774695\n",
      "100%|█████████████████████████████████████████| 155/155 [00:27<00:00,  5.70it/s]\n",
      "Test Instance Accuracy: 0.843548, Class Accuracy: 0.774300\n",
      "Best Instance Accuracy: 0.843548, Class Accuracy: 0.774300\n",
      "Saving at log/classification/pointnet2_cls_msg_github/checkpoints/best_model.pth\n",
      "Epoch 7 (7/300):\n",
      " 64%|██████████████████████████▎              | 394/615 [02:08<01:10,  3.13it/s]^C\n",
      " 64%|██████████████████████████▎              | 394/615 [02:08<01:11,  3.07it/s]\n",
      "Traceback (most recent call last):\n",
      "  File \"train_classification.py\", line 262, in <module>\n",
      "    main(args)\n",
      "  File \"train_classification.py\", line 214, in main\n",
      "    running_loss += loss.item()\n",
      "KeyboardInterrupt\n"
     ]
    }
   ],
   "source": [
    "!python train_classification.py --use_normals --model pointnet2_cls_msg --log_dir pointnet2_cls_msg_github --learning_rate 0.01 --batch_size 16 --optimizer SGD --epoch 300 --process_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "3fe13bae-7cb9-4870-8480-8228eb3afb95",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Namespace(batch_size=16, decay_rate=0.0001, epoch=300, gpu='0', learning_rate=0.01, log_dir='pointnet2_cls_msg_github', model='pointnet2_cls_msg', num_category=40, num_point=1024, optimizer='SGD', process_data=True, use_cpu=False, use_normals=True, use_uniform_sample=False)\n",
      "PARAMETER ...\n",
      "Namespace(batch_size=16, decay_rate=0.0001, epoch=300, gpu='0', learning_rate=0.01, log_dir='pointnet2_cls_msg_github', model='pointnet2_cls_msg', num_category=40, num_point=1024, optimizer='SGD', process_data=True, use_cpu=False, use_normals=True, use_uniform_sample=False)\n",
      "Load dataset ...\n",
      "The size of train data is 9843\n",
      "Load processed data from /home/jupyter/modelnet40_normal_resampled/modelnet40_train_1024pts.dat...\n",
      "The size of test data is 2468\n",
      "Load processed data from /home/jupyter/modelnet40_normal_resampled/modelnet40_test_1024pts.dat...\n",
      "Use pretrain model\n",
      "Epoch 1 (91/300):\n",
      "  0%|                                                   | 0/615 [00:02<?, ?it/s]\n",
      "Traceback (most recent call last):\n",
      "  File \"train_classification.py\", line 263, in <module>\n",
      "    main(args)\n",
      "  File \"train_classification.py\", line 214, in main\n",
      "    loss.backward()\n",
      "  File \"/opt/conda/lib/python3.7/site-packages/torch/_tensor.py\", line 489, in backward\n",
      "    self, gradient, retain_graph, create_graph, inputs=inputs\n",
      "  File \"/opt/conda/lib/python3.7/site-packages/torch/autograd/__init__.py\", line 199, in backward\n",
      "    allow_unreachable=True, accumulate_grad=True)  # Calls into the C++ engine to run the backward pass\n",
      "torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.25 GiB (GPU 0; 15.78 GiB total capacity; 10.36 GiB already allocated; 988.94 MiB free; 11.66 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation.  See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF\n"
     ]
    }
   ],
   "source": [
    "!python train_classification.py --use_normals --model pointnet2_cls_msg --log_dir pointnet2_cls_msg_github --learning_rate 0.01 --batch_size 16 --optimizer SGD --epoch 300 --process_data\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "74b52783-a68b-4b6e-8e20-4774d33f86e1",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "1d9d1844-2f72-4dcf-9936-729325aee62f",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import sys\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import numpy as np\n",
    "from torch.optim.lr_scheduler import CosineAnnealingLR, CosineAnnealingWarmRestarts\n",
    "\n",
    "import datetime\n",
    "import logging\n",
    "import provider\n",
    "import importlib\n",
    "import shutil\n",
    "import argparse\n",
    "\n",
    "from pathlib import Path\n",
    "from tqdm import tqdm\n",
    "from data_utils.ModelNetDataLoader import ModelNetDataLoader\n",
    "import numpy as np\n",
    "from train_classification import inplace_relu"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "028ed911-f53b-4fbc-9e57-5a5c4a07b474",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "get_model(\n",
       "  (sa1): PointNetSetAbstractionMsg(\n",
       "    (conv_blocks): ModuleList(\n",
       "      (0): ModuleList(\n",
       "        (0): Conv2d(9, 32, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (1): Conv2d(32, 32, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (2): Conv2d(32, 64, kernel_size=(1, 1), stride=(1, 1))\n",
       "      )\n",
       "      (1): ModuleList(\n",
       "        (0): Conv2d(9, 64, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (1): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (2): Conv2d(64, 128, kernel_size=(1, 1), stride=(1, 1))\n",
       "      )\n",
       "      (2): ModuleList(\n",
       "        (0): Conv2d(9, 64, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (1): Conv2d(64, 96, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (2): Conv2d(96, 128, kernel_size=(1, 1), stride=(1, 1))\n",
       "      )\n",
       "    )\n",
       "    (bn_blocks): ModuleList(\n",
       "      (0): ModuleList(\n",
       "        (0): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (1): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "      (1): ModuleList(\n",
       "        (0): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "      (2): ModuleList(\n",
       "        (0): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "    )\n",
       "  )\n",
       "  (global_sa1): Global_Transformer(\n",
       "    (layer): SwinBlock(\n",
       "      (attention_block): WindowAttention(\n",
       "        (linear_input): Sequential(\n",
       "          (0): Linear(in_features=320, out_features=320, bias=True)\n",
       "          (1): BatchNorm2d(320, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (2): ReLU(inplace=True)\n",
       "        )\n",
       "        (linear_out): Sequential(\n",
       "          (0): BatchNorm2d(320, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (1): ReLU(inplace=True)\n",
       "        )\n",
       "        (q_conv): Conv2d(320, 320, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (k_conv): Conv2d(320, 320, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (linear_p): Sequential(\n",
       "          (0): Linear(in_features=3, out_features=3, bias=True)\n",
       "          (1): BatchNorm3d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (2): ReLU(inplace=True)\n",
       "          (3): Linear(in_features=3, out_features=32, bias=True)\n",
       "        )\n",
       "        (linear_dots): Sequential(\n",
       "          (0): Linear(in_features=32, out_features=32, bias=True)\n",
       "          (1): ReLU(inplace=True)\n",
       "          (2): Linear(in_features=32, out_features=32, bias=True)\n",
       "        )\n",
       "        (v_conv): Conv2d(320, 320, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (trans_conv): Conv2d(320, 320, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (after_norm): BatchNorm2d(320, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (act): ReLU(inplace=True)\n",
       "        (softmax): Softmax(dim=-1)\n",
       "      )\n",
       "    )\n",
       "    (avepooling_layer): PatchMerging_avepooling(\n",
       "      (avepooling_layer): AdaptiveAvgPool2d(output_size=(1, 1))\n",
       "    )\n",
       "  )\n",
       "  (sa2): PointNetSetAbstractionMsg(\n",
       "    (conv_blocks): ModuleList(\n",
       "      (0): ModuleList(\n",
       "        (0): Conv2d(643, 64, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (1): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (2): Conv2d(64, 128, kernel_size=(1, 1), stride=(1, 1))\n",
       "      )\n",
       "      (1): ModuleList(\n",
       "        (0): Conv2d(643, 128, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (1): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (2): Conv2d(128, 256, kernel_size=(1, 1), stride=(1, 1))\n",
       "      )\n",
       "      (2): ModuleList(\n",
       "        (0): Conv2d(643, 128, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (1): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (2): Conv2d(128, 256, kernel_size=(1, 1), stride=(1, 1))\n",
       "      )\n",
       "    )\n",
       "    (bn_blocks): ModuleList(\n",
       "      (0): ModuleList(\n",
       "        (0): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "      (1): ModuleList(\n",
       "        (0): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "      (2): ModuleList(\n",
       "        (0): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "    )\n",
       "  )\n",
       "  (global_sa2): Global_Transformer(\n",
       "    (layer): SwinBlock(\n",
       "      (attention_block): WindowAttention(\n",
       "        (linear_input): Sequential(\n",
       "          (0): Linear(in_features=640, out_features=640, bias=True)\n",
       "          (1): BatchNorm2d(640, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (2): ReLU(inplace=True)\n",
       "        )\n",
       "        (linear_out): Sequential(\n",
       "          (0): BatchNorm2d(640, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (1): ReLU(inplace=True)\n",
       "        )\n",
       "        (q_conv): Conv2d(640, 640, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (k_conv): Conv2d(640, 640, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (linear_p): Sequential(\n",
       "          (0): Linear(in_features=3, out_features=3, bias=True)\n",
       "          (1): BatchNorm3d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (2): ReLU(inplace=True)\n",
       "          (3): Linear(in_features=3, out_features=32, bias=True)\n",
       "        )\n",
       "        (linear_dots): Sequential(\n",
       "          (0): Linear(in_features=32, out_features=32, bias=True)\n",
       "          (1): ReLU(inplace=True)\n",
       "          (2): Linear(in_features=32, out_features=32, bias=True)\n",
       "        )\n",
       "        (v_conv): Conv2d(640, 640, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (trans_conv): Conv2d(640, 640, kernel_size=(1, 1), stride=(1, 1))\n",
       "        (after_norm): BatchNorm2d(640, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (act): ReLU(inplace=True)\n",
       "        (softmax): Softmax(dim=-1)\n",
       "      )\n",
       "    )\n",
       "    (avepooling_layer): PatchMerging_avepooling(\n",
       "      (avepooling_layer): AdaptiveAvgPool2d(output_size=(1, 1))\n",
       "    )\n",
       "  )\n",
       "  (sa3): PointNetSetAbstraction(\n",
       "    (mlp_convs): ModuleList(\n",
       "      (0): Conv2d(643, 256, kernel_size=(1, 1), stride=(1, 1))\n",
       "      (1): Conv2d(256, 512, kernel_size=(1, 1), stride=(1, 1))\n",
       "      (2): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1))\n",
       "    )\n",
       "    (mlp_bns): ModuleList(\n",
       "      (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (2): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "    )\n",
       "  )\n",
       "  (fc1): Linear(in_features=1024, out_features=512, bias=True)\n",
       "  (bn1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "  (drop1): Dropout(p=0.4, inplace=False)\n",
       "  (fc2): Linear(in_features=512, out_features=256, bias=True)\n",
       "  (bn2): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "  (drop2): Dropout(p=0.5, inplace=False)\n",
       "  (fc3): Linear(in_features=256, out_features=40, bias=True)\n",
       ")"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "num_class=40\n",
    "model_name=\"pointnet_cls\"\n",
    "BASE_DIR = os.path.dirname(\"/home/jupyter/3DCTN/train_classification.py\")\n",
    "ROOT_DIR = BASE_DIR\n",
    "from models.pointnet2_cls_msg import get_model,get_loss\n",
    "classifier = get_model(40,normal_channel=True)\n",
    "criterion = get_loss()\n",
    "classifier.apply(inplace_relu)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "823ead23-f2c5-4cd8-b00e-abaa711c52b6",
   "metadata": {},
   "outputs": [],
   "source": [
    "classifier = classifier.cuda()\n",
    "criterion = criterion.cuda()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "f186daca-6696-4f2a-ac57-10ff2fcf017c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<All keys matched successfully>"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "checkpoint = torch.load('/home/jupyter/3DCTN/log/classification/pointnet2_cls_msg_github/checkpoints/old_model.pth')\n",
    "classifier.load_state_dict(checkpoint['model_state_dict'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "c7b0d85f-d37f-4d5d-b5f3-9b8f4b05bd9d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/home/jupyter/3DCTN\n"
     ]
    }
   ],
   "source": [
    "!pwd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "b419f6a9-e83f-4496-b6d5-44c07e0fcb34",
   "metadata": {},
   "outputs": [],
   "source": [
    "from train_classification import parse_args\n",
    "from argparse import Namespace\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "d4fc5ab2-a0e4-4e56-9b29-a4cb1f25e740",
   "metadata": {},
   "outputs": [],
   "source": [
    "args = Namespace(batch_size=1, decay_rate=0.0001, epoch=300, gpu='0', learning_rate=0.01, log_dir='pointnet2_cls_msg_github', model='pointnet2_cls_msg', num_category=40, num_point=1024, optimizer='SGD', process_data=True, use_cpu=False, use_normals=True, use_uniform_sample=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "913df2e3-6975-4e54-8e4b-d307d18ead30",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The size of test data is 2468\n",
      "Load processed data from /home/jupyter/modelnet40_normal_resampled/modelnet40_test_1024pts.dat...\n"
     ]
    }
   ],
   "source": [
    "data_path = '/home/jupyter/modelnet40_normal_resampled/'\n",
    "from data_utils.ModelNetDataLoader import ModelNetDataLoader\n",
    "test_dataset = ModelNetDataLoader(root=data_path, args=args, split='test', process_data=True)\n",
    "testDataLoader = torch.utils.data.DataLoader(test_dataset, batch_size=args.batch_size, shuffle=False, num_workers=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "9f5e6276-7de2-41f7-9fea-2531de1983eb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2468"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(testDataLoader)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "7d327789-e879-4d9f-ac31-c9069a3d9bb0",
   "metadata": {},
   "outputs": [],
   "source": [
    "def validation_test(model, loader, num_class=40):\n",
    "    mean_correct = []\n",
    "    class_acc = np.zeros((num_class, 3))\n",
    "    classifier = model.eval()\n",
    "\n",
    "    for j, (points, target) in tqdm(enumerate(loader), total=len(loader)):\n",
    "\n",
    "        if not args.use_cpu:\n",
    "            points, target = points.cuda(), target.cuda()\n",
    "\n",
    "        points = points.transpose(2, 1)\n",
    "        pred, _ = classifier(points)\n",
    "        pred_choice = pred.data.max(1)[1]\n",
    "\n",
    "        for cat in np.unique(target.cpu()):\n",
    "            classacc = pred_choice[target == cat].eq(target[target == cat].long().data).cpu().sum()\n",
    "            class_acc[cat, 0] += classacc.item() / float(points[target == cat].size()[0])\n",
    "            class_acc[cat, 1] += 1\n",
    "\n",
    "        correct = pred_choice.eq(target.long().data).cpu().sum()\n",
    "        mean_correct.append(correct.item() / float(points.size()[0]))\n",
    "\n",
    "    class_acc[:, 2] = class_acc[:, 0] / class_acc[:, 1]    \n",
    "\n",
    "    return class_acc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "ffaebcac-4f83-4671-93a5-44cecdcac3cb",
   "metadata": {},
   "outputs": [],
   "source": [
    "def confusion_matrix_create(model, loader, num_class=40):\n",
    "    predictions = []\n",
    "    truths = []\n",
    "    class_acc = np.zeros((num_class, 3))\n",
    "    classifier = model.eval()\n",
    "\n",
    "    for j, (points, target) in tqdm(enumerate(loader), total=len(loader)):\n",
    "\n",
    "        if not args.use_cpu:\n",
    "            points, target = points.cuda(), target.cuda()\n",
    "\n",
    "        points = points.transpose(2, 1)\n",
    "        pred, _ = classifier(points)\n",
    "        pred_choice = pred.data.max(1)[1].cpu().numpy()\n",
    "        \n",
    "        true_choice = target.cpu().numpy()\n",
    "        predictions.append(pred_choice)\n",
    "        truths.append(true_choice)\n",
    "        # if pred_choice != true_choice:\n",
    "        #     print(pred_choice,true_choice)\n",
    "        # for cat in np.unique(target.cpu()):\n",
    "        #     classacc = pred_choice[target == cat].eq(target[target == cat].long().data).cpu().sum()\n",
    "        #     class_acc[cat, 0] += classacc.item() / float(points[target == cat].size()[0])\n",
    "        #     class_acc[cat, 1] += 1\n",
    "\n",
    "        # correct = pred_choice.eq(target.long().data).cpu().sum()\n",
    "        # mean_correct.append(correct.item() / float(points.size()[0]))\n",
    "\n",
    "    # class_acc[:, 2] = class_acc[:, 0] / class_acc[:, 1]    \n",
    "\n",
    "    return [predictions,truths]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "77767a59-5770-4312-8f2b-d7195b49af7b",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2468/2468 [04:18<00:00,  9.55it/s]\n"
     ]
    }
   ],
   "source": [
    "predictions,truths = None, None\n",
    "with torch.no_grad():\n",
    "    predictions,truths = confusion_matrix_create(classifier.eval(), testDataLoader, num_class=num_class)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "85418edc-6114-45b9-af7c-a6e8d7ca5053",
   "metadata": {},
   "outputs": [],
   "source": [
    "class_acc\n",
    "np.save('class_acc.npy',class_acc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "fc8ea93f-626d-40c1-bb28-4d8a1fcb6270",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[100.        , 100.        ,   1.        ],\n",
       "       [ 46.        ,  50.        ,   0.92      ],\n",
       "       [100.        , 100.        ,   1.        ],\n",
       "       [ 16.        ,  20.        ,   0.8       ],\n",
       "       [ 98.        , 100.        ,   0.98      ],\n",
       "       [ 99.        , 100.        ,   0.99      ],\n",
       "       [ 20.        ,  20.        ,   1.        ],\n",
       "       [ 99.        , 100.        ,   0.99      ],\n",
       "       [ 97.        , 100.        ,   0.97      ],\n",
       "       [ 19.        ,  20.        ,   0.95      ],\n",
       "       [ 13.        ,  20.        ,   0.65      ],\n",
       "       [ 18.        ,  20.        ,   0.9       ],\n",
       "       [ 74.        ,  86.        ,   0.86046512],\n",
       "       [ 20.        ,  20.        ,   1.        ],\n",
       "       [ 70.        ,  86.        ,   0.81395349],\n",
       "       [  2.        ,  20.        ,   0.1       ],\n",
       "       [ 97.        , 100.        ,   0.97      ],\n",
       "       [100.        , 100.        ,   1.        ],\n",
       "       [ 20.        ,  20.        ,   1.        ],\n",
       "       [ 17.        ,  20.        ,   0.85      ],\n",
       "       [ 20.        ,  20.        ,   1.        ],\n",
       "       [ 96.        , 100.        ,   0.96      ],\n",
       "       [ 98.        , 100.        ,   0.98      ],\n",
       "       [ 67.        ,  86.        ,   0.77906977],\n",
       "       [ 18.        ,  20.        ,   0.9       ],\n",
       "       [ 90.        , 100.        ,   0.9       ],\n",
       "       [ 88.        , 100.        ,   0.88      ],\n",
       "       [ 20.        ,  20.        ,   1.        ],\n",
       "       [ 95.        , 100.        ,   0.95      ],\n",
       "       [ 16.        ,  20.        ,   0.8       ],\n",
       "       [ 99.        , 100.        ,   0.99      ],\n",
       "       [ 17.        ,  20.        ,   0.85      ],\n",
       "       [ 18.        ,  20.        ,   0.9       ],\n",
       "       [ 94.        , 100.        ,   0.94      ],\n",
       "       [ 19.        ,  20.        ,   0.95      ],\n",
       "       [ 98.        , 100.        ,   0.98      ],\n",
       "       [ 91.        , 100.        ,   0.91      ],\n",
       "       [ 79.        , 100.        ,   0.79      ],\n",
       "       [ 15.        ,  20.        ,   0.75      ],\n",
       "       [ 17.        ,  20.        ,   0.85      ]])"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "class_acc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "id": "842c3fb0-0d4c-47d5-a8bb-fbab2e6283f4",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "from sklearn.metrics import confusion_matrix\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import seaborn as sn\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "id": "ed95d932-0d90-48ac-b458-1bae30690166",
   "metadata": {},
   "outputs": [],
   "source": [
    "def find_classes(dir):\n",
    "    classes = [ name for name in os.listdir(dir) if os.path.isdir(os.path.join(dir, name)) and name != \".ipynb_checkpoints\"]\n",
    "    classes.sort()\n",
    "    class_to_idx = {classes[i]: i for i in range(len(classes))}\n",
    "    return classes, class_to_idx\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "id": "8dac98f4-5105-42fb-a6b6-cda80b3d38c3",
   "metadata": {},
   "outputs": [],
   "source": [
    "all_classes = find_classes(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "id": "dde47e7e-634f-4778-882c-653b1854a5f6",
   "metadata": {},
   "outputs": [],
   "source": [
    "all_classes_list = all_classes[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "id": "72b73650-ed91-4782-8879-8d257a711a22",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x700 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cf_matrix = confusion_matrix(truths,predictions)\n",
    "df_cm = pd.DataFrame(cf_matrix / np.sum(cf_matrix, axis=1)[:, None], index = [i for i in all_classes_list],\n",
    "                     columns = [i for i in all_classes_list])\n",
    "plt.figure(figsize = (12,7))\n",
    "sn.heatmap(df_cm, annot=True)\n",
    "plt.savefig('output.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0d1b9f5e-9146-410b-8632-b6c3f374b69b",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "environment": {
   "kernel": "python3",
   "name": "pytorch-gpu.1-13.m107",
   "type": "gcloud",
   "uri": "gcr.io/deeplearning-platform-release/pytorch-gpu.1-13:m107"
  },
  "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": "3.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}