benchmark.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "c7beb2e9-48ee-490e-a830-54c06b47c0bb",
   "metadata": {},
   "source": [
    "# 1 class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "b4324361-a3ad-4004-b161-0bfa0fc107ff",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": []
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |█████████████████| 233/233 [762.3ms elapsed, 0s remaining, 305.6 samples/s]      \n",
      " 100% |███████████████████| 51/51 [193.2ms elapsed, 0s remaining, 267.3 samples/s] \n",
      " 100% |███████████████████| 50/50 [193.6ms elapsed, 0s remaining, 260.3 samples/s] \n",
      "Found train dataset labels:\n",
      "['sitting', 'walking', 'waving']\n",
      "Found test dataset labels:\n",
      "['sitting', 'walking', 'waving']\n",
      "Found valid dataset labels:\n",
      "['sitting', 'walking', 'waving']\n",
      " 100% |███████████████████| 74/74 [210.4ms elapsed, 0s remaining, 351.7 samples/s]     \n",
      " 100% |███████████████████| 17/17 [124.7ms elapsed, 0s remaining, 136.4 samples/s]    \n",
      " 100% |███████████████████| 17/17 [129.0ms elapsed, 0s remaining, 131.8 samples/s]    \n",
      "Found difficult_train dataset labels:\n",
      "['sitting', 'walking', 'waving']\n",
      "Found difficult_test dataset labels:\n",
      "['sitting', 'walking', 'waving']\n",
      "Found difficult_valid dataset labels:\n",
      "['sitting', 'walking', 'waving']\n",
      "Label dictionary for complete dataset\n",
      "{'person': 0}\n"
     ]
    }
   ],
   "source": [
    "from helpers.init import initialization\n",
    "cfg, classes, labels_dict, train_dataset_combined, test_dataset_combined, valid_dataset_combined = initialization(merge_to_one_class = True, encoding = True, swin = False)\n",
    "\n",
    "\n",
    "cfg.SOLVER.IMS_PER_BATCH = 2 # batch size\n",
    "cfg.DATALOADER.NUM_WORKERS = 2\n",
    "cfg.DATALOADER.PREFETCH_FACTOR = 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "abfab8ef-2a93-4396-bb50-c587b4eb9a64",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Current model is:    output/0/model_0003999.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/user/seliuninas0/anaconda3/envs/maskdino/lib/python3.8/site-packages/torch/nn/functional.py:718: UserWarning: Named tensors and all their associated APIs are an experimental feature and subject to change. Please do not use them for anything important until they are released as stable. (Triggered internally at  /opt/conda/conda-bld/pytorch_1623448278899/work/c10/core/TensorImpl.h:1156.)\n",
      "  return torch.max_pool2d(input, kernel_size, stride, padding, dilation, ceil_mode)\n",
      "/home/user/seliuninas0/anaconda3/envs/maskdino/lib/python3.8/site-packages/torch/_tensor.py:575: UserWarning: floor_divide is deprecated, and will be removed in a future version of pytorch. It currently rounds toward 0 (like the 'trunc' function NOT 'floor'). This results in incorrect rounding for negative values.\n",
      "To keep the current behavior, use torch.div(a, b, rounding_mode='trunc'), or for actual floor division, use torch.div(a, b, rounding_mode='floor'). (Triggered internally at  /opt/conda/conda-bld/pytorch_1623448278899/work/aten/src/ATen/native/BinaryOps.cpp:467.)\n",
      "  return torch.floor_divide(self, other)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Evaluating detections...\n",
      "  75% |██████████████|----| 51/68 [424.4ms elapsed, 141.5ms remaining, 120.2 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/user/seliuninas0/anaconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [505.3ms elapsed, 0s remaining, 134.6 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [457.3ms elapsed, 0s remaining, 148.7 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      person       0.71      0.43      0.54        95\n",
      "\n",
      "   micro avg       0.71      0.43      0.54        95\n",
      "   macro avg       0.71      0.43      0.54        95\n",
      "weighted avg       0.71      0.43      0.54        95\n",
      "\n",
      "Current model is:    output/0/model_0002499.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n",
      "Evaluating detections...\n",
      "  62% |███████████|-------| 42/68 [433.6ms elapsed, 268.4ms remaining, 96.9 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/user/seliuninas0/anaconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [563.4ms elapsed, 0s remaining, 120.7 samples/s]     \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [480.6ms elapsed, 0s remaining, 141.5 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      person       0.76      0.47      0.58        96\n",
      "\n",
      "   micro avg       0.76      0.47      0.58        96\n",
      "   macro avg       0.76      0.47      0.58        96\n",
      "weighted avg       0.76      0.47      0.58        96\n",
      "\n",
      "Current model is:    output/0/model_0002999.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n",
      "Evaluating detections...\n",
      "  59% |███████████\\-------| 40/68 [329.4ms elapsed, 230.6ms remaining, 121.4 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/user/seliuninas0/anaconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [469.9ms elapsed, 0s remaining, 144.7 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [450.3ms elapsed, 0s remaining, 151.0 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      person       0.75      0.43      0.55        95\n",
      "\n",
      "   micro avg       0.75      0.43      0.55        95\n",
      "   macro avg       0.75      0.43      0.55        95\n",
      "weighted avg       0.75      0.43      0.55        95\n",
      "\n",
      "Current model is:    output/0/model_0000999.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n",
      "Evaluating detections...\n",
      "  49% |█████████\\---------| 33/68 [323.6ms elapsed, 343.2ms remaining, 102.0 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/user/seliuninas0/anaconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [592.7ms elapsed, 0s remaining, 114.7 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [572.2ms elapsed, 0s remaining, 118.8 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      person       0.52      0.46      0.49       102\n",
      "\n",
      "   micro avg       0.52      0.46      0.49       102\n",
      "   macro avg       0.52      0.46      0.49       102\n",
      "weighted avg       0.52      0.46      0.49       102\n",
      "\n",
      "Current model is:    output/0/model_0000499.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n",
      "Evaluating detections...\n",
      "  54% |██████████\\--------| 37/68 [320.8ms elapsed, 268.8ms remaining, 115.3 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/user/seliuninas0/anaconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [487.4ms elapsed, 0s remaining, 139.5 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [466.6ms elapsed, 0s remaining, 145.7 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      person       0.59      0.37      0.45        93\n",
      "\n",
      "   micro avg       0.59      0.37      0.45        93\n",
      "   macro avg       0.59      0.37      0.45        93\n",
      "weighted avg       0.59      0.37      0.45        93\n",
      "\n",
      "Current model is:    output/0/model_0003499.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n",
      "Evaluating detections...\n",
      "  59% |███████████\\-------| 40/68 [332.1ms elapsed, 232.4ms remaining, 120.5 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/user/seliuninas0/anaconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [493.3ms elapsed, 0s remaining, 137.8 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [462.3ms elapsed, 0s remaining, 147.1 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      person       0.73      0.45      0.56        95\n",
      "\n",
      "   micro avg       0.73      0.45      0.56        95\n",
      "   macro avg       0.73      0.45      0.56        95\n",
      "weighted avg       0.73      0.45      0.56        95\n",
      "\n",
      "Current model is:    output/0/model_final.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n",
      "Evaluating detections...\n",
      "  78% |██████████████|----| 53/68 [429.2ms elapsed, 121.5ms remaining, 123.5 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/user/seliuninas0/anaconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [496.2ms elapsed, 0s remaining, 137.0 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [470.6ms elapsed, 0s remaining, 144.5 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      person       0.71      0.43      0.54        95\n",
      "\n",
      "   micro avg       0.71      0.43      0.54        95\n",
      "   macro avg       0.71      0.43      0.54        95\n",
      "weighted avg       0.71      0.43      0.54        95\n",
      "\n",
      "Current model is:    output/0/model_0001999.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n",
      "Evaluating detections...\n",
      "  69% |█████████████|-----| 47/68 [431.8ms elapsed, 192.9ms remaining, 108.9 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/user/seliuninas0/anaconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [563.7ms elapsed, 0s remaining, 120.6 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [537.4ms elapsed, 0s remaining, 126.5 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      person       0.68      0.47      0.56       100\n",
      "\n",
      "   micro avg       0.68      0.47      0.56       100\n",
      "   macro avg       0.68      0.47      0.56       100\n",
      "weighted avg       0.68      0.47      0.56       100\n",
      "\n",
      "Current model is:    output/0/model_0001499.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n",
      "Evaluating detections...\n",
      "  51% |█████████\\---------| 35/68 [405.1ms elapsed, 381.9ms remaining, 86.4 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/user/seliuninas0/anaconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [690.0ms elapsed, 0s remaining, 98.5 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [636.4ms elapsed, 0s remaining, 106.9 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      person       0.51      0.56      0.54       101\n",
      "\n",
      "   micro avg       0.51      0.56      0.54       101\n",
      "   macro avg       0.51      0.56      0.54       101\n",
      "weighted avg       0.51      0.56      0.54       101\n",
      "\n"
     ]
    }
   ],
   "source": [
    "import os, cv2\n",
    "from helpers.model_trainer import Predictor\n",
    "from helpers.fiftyone_detectron2_bridge import clean_instances, detectron_to_fo, get_fiftyone_dicts\n",
    "\n",
    "path = \"output/0\"\n",
    "cfg.OUTPUT_DIR = path\n",
    "\n",
    "for root, dirs, files in os.walk(path):\n",
    "    for name in files:\n",
    "        if '3_classes' not in root and 'inference' not in root and 'example_detector' not in root:\n",
    "            if name[-3:] == 'pth':\n",
    "                #print(os.path.join(root, name))\n",
    "\n",
    "                print(f'Current model is:    {os.path.join(root, name)}')\n",
    "                print('')\n",
    "        \n",
    "                weight_path = os.path.join(os.path.join(root, name))\n",
    "                cfg.MODEL.WEIGHTS =  weight_path\n",
    "                \n",
    "                #cfg.MODEL.ROI_HEADS.SCORE_THRESH_TEST = 0.1\n",
    "                #cfg.MODEL.ROI_HEADS.NMS_THRESH_TEST = 0.8\n",
    "                \n",
    "                predictor = Predictor(cfg)\n",
    "\n",
    "                test_view = test_dataset_combined\n",
    "                dataset_dicts = get_fiftyone_dicts(test_view, labels_dict)\n",
    "                predictions = {}\n",
    "                for d in dataset_dicts:\n",
    "                    img_w = d[\"width\"]\n",
    "                    img_h = d[\"height\"]\n",
    "                    img = cv2.imread(d[\"file_name\"], cv2.IMREAD_UNCHANGED)\n",
    "                    outputs = predictor(img)\n",
    "                    detections, instances = detectron_to_fo(outputs, img_w, img_h, classes, score_tresh=0.3)\n",
    "                    predictions[d[\"image_id\"]] = detections\n",
    "                \n",
    "                test_dataset_combined.set_values(\"predictions\", predictions, key_field=\"id\")\n",
    "\n",
    "                test_results = test_dataset_combined.evaluate_detections(\n",
    "                    \"predictions\",\n",
    "                    gt_field=\"segmentations\",\n",
    "                    eval_key=\"eval\",\n",
    "                    use_masks=True,\n",
    "                    compute_mAP=True,\n",
    "                    classes=test_dataset_combined.distinct(\"segmentations.detections.label\")\n",
    "                )\n",
    "                test_results.print_report()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c1e7e926-a7c5-4bf1-8990-a016785840fd",
   "metadata": {},
   "source": [
    "# 3 classes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "a9efb79f-f5b1-4a51-8354-aea1399cd470",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": []
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |█████████████████| 233/233 [425.8ms elapsed, 0s remaining, 550.2 samples/s]      \n",
      " 100% |███████████████████| 51/51 [129.8ms elapsed, 0s remaining, 399.6 samples/s] \n",
      " 100% |███████████████████| 50/50 [124.2ms elapsed, 0s remaining, 408.9 samples/s] \n",
      "Found train dataset labels:\n",
      "['sitting', 'walking', 'waving']\n",
      "Found test dataset labels:\n",
      "['sitting', 'walking', 'waving']\n",
      "Found valid dataset labels:\n",
      "['sitting', 'walking', 'waving']\n",
      " 100% |███████████████████| 74/74 [156.3ms elapsed, 0s remaining, 473.3 samples/s]     \n",
      " 100% |███████████████████| 17/17 [53.8ms elapsed, 0s remaining, 315.8 samples/s]     \n",
      " 100% |███████████████████| 17/17 [53.1ms elapsed, 0s remaining, 320.4 samples/s]     \n",
      "Found difficult_train dataset labels:\n",
      "['sitting', 'walking', 'waving']\n",
      "Found difficult_test dataset labels:\n",
      "['sitting', 'walking', 'waving']\n",
      "Found difficult_valid dataset labels:\n",
      "['sitting', 'walking', 'waving']\n",
      "Means of the dataset:  [ 58.33375975 115.78446252 234.51154039]\n",
      "STDs of the dataset:  [76.6960031  74.260651   16.82153164]\n",
      "Label dictionary for complete dataset\n",
      "{'sitting': 0, 'walking': 1, 'waving': 2}\n"
     ]
    }
   ],
   "source": [
    "from helpers.dataset import initialization\n",
    "cfg, classes, labels_dict, train_dataset_combined, test_dataset_combined, valid_dataset_combined = initialization(merge_to_one_class = False, channels = [1, 2, 3])\n",
    "\n",
    "\n",
    "cfg.SOLVER.IMS_PER_BATCH = 2 # batch size\n",
    "cfg.DATALOADER.NUM_WORKERS = 2\n",
    "cfg.DATALOADER.PREFETCH_FACTOR = 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "24dd9341-f770-4ca1-9204-47007ba1a484",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Current model is:    output/20/model_0004499.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/artemoneto/miniconda3/envs/maskdino/lib/python3.8/site-packages/torch/_tensor.py:575: UserWarning: floor_divide is deprecated, and will be removed in a future version of pytorch. It currently rounds toward 0 (like the 'trunc' function NOT 'floor'). This results in incorrect rounding for negative values.\n",
      "To keep the current behavior, use torch.div(a, b, rounding_mode='trunc'), or for actual floor division, use torch.div(a, b, rounding_mode='floor'). (Triggered internally at  /opt/conda/conda-bld/pytorch_1623448278899/work/aten/src/ATen/native/BinaryOps.cpp:467.)\n",
      "  return torch.floor_divide(self, other)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Evaluating detections...\n",
      "  72% |█████████████|-----| 49/68 [420.1ms elapsed, 162.9ms remaining, 116.6 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/artemoneto/miniconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [542.6ms elapsed, 0s remaining, 125.3 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [447.9ms elapsed, 0s remaining, 151.8 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "     sitting       1.00      0.90      0.95        40\n",
      "     walking       0.82      0.82      0.82        28\n",
      "      waving       0.85      0.85      0.85        26\n",
      "\n",
      "   micro avg       0.90      0.86      0.88        94\n",
      "   macro avg       0.89      0.86      0.87        94\n",
      "weighted avg       0.90      0.86      0.88        94\n",
      "\n",
      "Current model is:    output/20/model_0001999.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n",
      "Evaluating detections...\n",
      "  68% |████████████\\------| 46/68 [318.3ms elapsed, 152.2ms remaining, 144.5 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/artemoneto/miniconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [457.1ms elapsed, 0s remaining, 148.8 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [405.2ms elapsed, 0s remaining, 167.8 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "     sitting       0.95      0.88      0.91        41\n",
      "     walking       0.70      0.75      0.72        28\n",
      "      waving       0.69      0.77      0.73        26\n",
      "\n",
      "   micro avg       0.79      0.81      0.80        95\n",
      "   macro avg       0.78      0.80      0.79        95\n",
      "weighted avg       0.80      0.81      0.81        95\n",
      "\n",
      "Current model is:    output/20/model_0003499.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n",
      "Evaluating detections...\n",
      "  66% |████████████\\------| 45/68 [315.2ms elapsed, 161.1ms remaining, 142.8 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/artemoneto/miniconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [448.0ms elapsed, 0s remaining, 151.8 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [384.0ms elapsed, 0s remaining, 177.1 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "     sitting       0.94      0.78      0.85        40\n",
      "     walking       0.90      0.70      0.79        27\n",
      "      waving       0.83      0.73      0.78        26\n",
      "\n",
      "   micro avg       0.90      0.74      0.81        93\n",
      "   macro avg       0.89      0.74      0.81        93\n",
      "weighted avg       0.90      0.74      0.81        93\n",
      "\n",
      "Current model is:    output/20/model_final.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n",
      "Evaluating detections...\n",
      "  72% |█████████████|-----| 49/68 [424.4ms elapsed, 164.6ms remaining, 115.5 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/artemoneto/miniconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [573.8ms elapsed, 0s remaining, 118.5 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [469.8ms elapsed, 0s remaining, 144.8 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "     sitting       1.00      0.90      0.95        40\n",
      "     walking       0.79      0.70      0.75        27\n",
      "      waving       0.85      0.88      0.87        26\n",
      "\n",
      "   micro avg       0.90      0.84      0.87        93\n",
      "   macro avg       0.88      0.83      0.85        93\n",
      "weighted avg       0.90      0.84      0.87        93\n",
      "\n",
      "Current model is:    output/20/model_0000499.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n",
      "Evaluating detections...\n",
      "  63% |████████████\\------| 43/68 [309.5ms elapsed, 179.9ms remaining, 138.9 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/artemoneto/miniconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [477.0ms elapsed, 0s remaining, 142.5 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [504.5ms elapsed, 0s remaining, 134.8 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "     sitting       0.80      0.90      0.85        41\n",
      "     walking       0.70      0.79      0.74        29\n",
      "      waving       0.71      0.81      0.76        27\n",
      "\n",
      "   micro avg       0.75      0.85      0.79        97\n",
      "   macro avg       0.74      0.84      0.78        97\n",
      "weighted avg       0.75      0.85      0.79        97\n",
      "\n",
      "Current model is:    output/20/model_0000999.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n",
      "Evaluating detections...\n",
      "  72% |█████████████\\-----| 49/68 [317.8ms elapsed, 123.2ms remaining, 154.2 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/artemoneto/miniconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [442.5ms elapsed, 0s remaining, 153.7 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [393.7ms elapsed, 0s remaining, 172.7 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "     sitting       1.00      0.65      0.79        40\n",
      "     walking       0.79      0.70      0.75        27\n",
      "      waving       0.71      0.77      0.74        26\n",
      "\n",
      "   micro avg       0.83      0.70      0.76        93\n",
      "   macro avg       0.84      0.71      0.76        93\n",
      "weighted avg       0.86      0.70      0.76        93\n",
      "\n",
      "Current model is:    output/20/model_0001499.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n",
      "Evaluating detections...\n",
      "  63% |████████████\\------| 43/68 [316.1ms elapsed, 183.8ms remaining, 136.0 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/artemoneto/miniconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [487.7ms elapsed, 0s remaining, 139.4 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [499.2ms elapsed, 0s remaining, 136.2 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "     sitting       0.93      0.93      0.93        42\n",
      "     walking       0.88      0.78      0.82        27\n",
      "      waving       0.70      1.00      0.83        26\n",
      "\n",
      "   micro avg       0.83      0.91      0.87        95\n",
      "   macro avg       0.84      0.90      0.86        95\n",
      "weighted avg       0.85      0.91      0.87        95\n",
      "\n",
      "Current model is:    output/20/model_0002999.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n",
      "Evaluating detections...\n",
      "  63% |████████████\\------| 43/68 [311.8ms elapsed, 181.3ms remaining, 137.9 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/artemoneto/miniconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [494.6ms elapsed, 0s remaining, 137.5 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [405.0ms elapsed, 0s remaining, 167.9 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "     sitting       1.00      0.93      0.96        40\n",
      "     walking       0.90      0.67      0.77        27\n",
      "      waving       0.71      0.92      0.80        26\n",
      "\n",
      "   micro avg       0.87      0.85      0.86        93\n",
      "   macro avg       0.87      0.84      0.84        93\n",
      "weighted avg       0.89      0.85      0.86        93\n",
      "\n",
      "Current model is:    output/20/model_0004999.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n",
      "Evaluating detections...\n",
      "  76% |██████████████|----| 52/68 [426.1ms elapsed, 131.1ms remaining, 122.0 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/artemoneto/miniconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [535.8ms elapsed, 0s remaining, 126.9 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [467.4ms elapsed, 0s remaining, 145.5 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "     sitting       1.00      0.90      0.95        40\n",
      "     walking       0.79      0.70      0.75        27\n",
      "      waving       0.85      0.88      0.87        26\n",
      "\n",
      "   micro avg       0.90      0.84      0.87        93\n",
      "   macro avg       0.88      0.83      0.85        93\n",
      "weighted avg       0.90      0.84      0.87        93\n",
      "\n",
      "Current model is:    output/20/model_0003999.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n",
      "Evaluating detections...\n",
      "  75% |██████████████|----| 51/68 [425.1ms elapsed, 141.7ms remaining, 120.0 samples/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/artemoneto/miniconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [537.4ms elapsed, 0s remaining, 126.5 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [442.3ms elapsed, 0s remaining, 153.7 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "     sitting       0.97      0.85      0.91        40\n",
      "     walking       0.69      0.81      0.75        27\n",
      "      waving       0.81      0.81      0.81        26\n",
      "\n",
      "   micro avg       0.83      0.83      0.83        93\n",
      "   macro avg       0.82      0.82      0.82        93\n",
      "weighted avg       0.84      0.83      0.83        93\n",
      "\n",
      "Current model is:    output/20/model_0002499.pth\n",
      "\n",
      "criterion.weight_dict  {'loss_ce': 4.0, 'loss_mask': 5.0, 'loss_dice': 5.0, 'loss_bbox': 5.0, 'loss_giou': 2.0, 'loss_ce_interm': 4.0, 'loss_mask_interm': 5.0, 'loss_dice_interm': 5.0, 'loss_bbox_interm': 5.0, 'loss_giou_interm': 2.0, 'loss_ce_dn': 4.0, 'loss_mask_dn': 5.0, 'loss_dice_dn': 5.0, 'loss_bbox_dn': 5.0, 'loss_giou_dn': 2.0, 'loss_ce_interm_dn': 4.0, 'loss_mask_interm_dn': 5.0, 'loss_dice_interm_dn': 5.0, 'loss_bbox_interm_dn': 5.0, 'loss_giou_interm_dn': 2.0, 'loss_ce_0': 4.0, 'loss_mask_0': 5.0, 'loss_dice_0': 5.0, 'loss_bbox_0': 5.0, 'loss_giou_0': 2.0, 'loss_ce_interm_0': 4.0, 'loss_mask_interm_0': 5.0, 'loss_dice_interm_0': 5.0, 'loss_bbox_interm_0': 5.0, 'loss_giou_interm_0': 2.0, 'loss_ce_dn_0': 4.0, 'loss_mask_dn_0': 5.0, 'loss_dice_dn_0': 5.0, 'loss_bbox_dn_0': 5.0, 'loss_giou_dn_0': 2.0, 'loss_ce_interm_dn_0': 4.0, 'loss_mask_interm_dn_0': 5.0, 'loss_dice_interm_dn_0': 5.0, 'loss_bbox_interm_dn_0': 5.0, 'loss_giou_interm_dn_0': 2.0, 'loss_ce_1': 4.0, 'loss_mask_1': 5.0, 'loss_dice_1': 5.0, 'loss_bbox_1': 5.0, 'loss_giou_1': 2.0, 'loss_ce_interm_1': 4.0, 'loss_mask_interm_1': 5.0, 'loss_dice_interm_1': 5.0, 'loss_bbox_interm_1': 5.0, 'loss_giou_interm_1': 2.0, 'loss_ce_dn_1': 4.0, 'loss_mask_dn_1': 5.0, 'loss_dice_dn_1': 5.0, 'loss_bbox_dn_1': 5.0, 'loss_giou_dn_1': 2.0, 'loss_ce_interm_dn_1': 4.0, 'loss_mask_interm_dn_1': 5.0, 'loss_dice_interm_dn_1': 5.0, 'loss_bbox_interm_dn_1': 5.0, 'loss_giou_interm_dn_1': 2.0, 'loss_ce_2': 4.0, 'loss_mask_2': 5.0, 'loss_dice_2': 5.0, 'loss_bbox_2': 5.0, 'loss_giou_2': 2.0, 'loss_ce_interm_2': 4.0, 'loss_mask_interm_2': 5.0, 'loss_dice_interm_2': 5.0, 'loss_bbox_interm_2': 5.0, 'loss_giou_interm_2': 2.0, 'loss_ce_dn_2': 4.0, 'loss_mask_dn_2': 5.0, 'loss_dice_dn_2': 5.0, 'loss_bbox_dn_2': 5.0, 'loss_giou_dn_2': 2.0, 'loss_ce_interm_dn_2': 4.0, 'loss_mask_interm_dn_2': 5.0, 'loss_dice_interm_dn_2': 5.0, 'loss_bbox_interm_dn_2': 5.0, 'loss_giou_interm_dn_2': 2.0, 'loss_ce_3': 4.0, 'loss_mask_3': 5.0, 'loss_dice_3': 5.0, 'loss_bbox_3': 5.0, 'loss_giou_3': 2.0, 'loss_ce_interm_3': 4.0, 'loss_mask_interm_3': 5.0, 'loss_dice_interm_3': 5.0, 'loss_bbox_interm_3': 5.0, 'loss_giou_interm_3': 2.0, 'loss_ce_dn_3': 4.0, 'loss_mask_dn_3': 5.0, 'loss_dice_dn_3': 5.0, 'loss_bbox_dn_3': 5.0, 'loss_giou_dn_3': 2.0, 'loss_ce_interm_dn_3': 4.0, 'loss_mask_interm_dn_3': 5.0, 'loss_dice_interm_dn_3': 5.0, 'loss_bbox_interm_dn_3': 5.0, 'loss_giou_interm_dn_3': 2.0, 'loss_ce_4': 4.0, 'loss_mask_4': 5.0, 'loss_dice_4': 5.0, 'loss_bbox_4': 5.0, 'loss_giou_4': 2.0, 'loss_ce_interm_4': 4.0, 'loss_mask_interm_4': 5.0, 'loss_dice_interm_4': 5.0, 'loss_bbox_interm_4': 5.0, 'loss_giou_interm_4': 2.0, 'loss_ce_dn_4': 4.0, 'loss_mask_dn_4': 5.0, 'loss_dice_dn_4': 5.0, 'loss_bbox_dn_4': 5.0, 'loss_giou_dn_4': 2.0, 'loss_ce_interm_dn_4': 4.0, 'loss_mask_interm_dn_4': 5.0, 'loss_dice_interm_dn_4': 5.0, 'loss_bbox_interm_dn_4': 5.0, 'loss_giou_interm_dn_4': 2.0, 'loss_ce_5': 4.0, 'loss_mask_5': 5.0, 'loss_dice_5': 5.0, 'loss_bbox_5': 5.0, 'loss_giou_5': 2.0, 'loss_ce_interm_5': 4.0, 'loss_mask_interm_5': 5.0, 'loss_dice_interm_5': 5.0, 'loss_bbox_interm_5': 5.0, 'loss_giou_interm_5': 2.0, 'loss_ce_dn_5': 4.0, 'loss_mask_dn_5': 5.0, 'loss_dice_dn_5': 5.0, 'loss_bbox_dn_5': 5.0, 'loss_giou_dn_5': 2.0, 'loss_ce_interm_dn_5': 4.0, 'loss_mask_interm_dn_5': 5.0, 'loss_dice_interm_dn_5': 5.0, 'loss_bbox_interm_dn_5': 5.0, 'loss_giou_interm_dn_5': 2.0, 'loss_ce_6': 4.0, 'loss_mask_6': 5.0, 'loss_dice_6': 5.0, 'loss_bbox_6': 5.0, 'loss_giou_6': 2.0, 'loss_ce_interm_6': 4.0, 'loss_mask_interm_6': 5.0, 'loss_dice_interm_6': 5.0, 'loss_bbox_interm_6': 5.0, 'loss_giou_interm_6': 2.0, 'loss_ce_dn_6': 4.0, 'loss_mask_dn_6': 5.0, 'loss_dice_dn_6': 5.0, 'loss_bbox_dn_6': 5.0, 'loss_giou_dn_6': 2.0, 'loss_ce_interm_dn_6': 4.0, 'loss_mask_interm_dn_6': 5.0, 'loss_dice_interm_dn_6': 5.0, 'loss_bbox_interm_dn_6': 5.0, 'loss_giou_interm_dn_6': 2.0, 'loss_ce_7': 4.0, 'loss_mask_7': 5.0, 'loss_dice_7': 5.0, 'loss_bbox_7': 5.0, 'loss_giou_7': 2.0, 'loss_ce_interm_7': 4.0, 'loss_mask_interm_7': 5.0, 'loss_dice_interm_7': 5.0, 'loss_bbox_interm_7': 5.0, 'loss_giou_interm_7': 2.0, 'loss_ce_dn_7': 4.0, 'loss_mask_dn_7': 5.0, 'loss_dice_dn_7': 5.0, 'loss_bbox_dn_7': 5.0, 'loss_giou_dn_7': 2.0, 'loss_ce_interm_dn_7': 4.0, 'loss_mask_interm_dn_7': 5.0, 'loss_dice_interm_dn_7': 5.0, 'loss_bbox_interm_dn_7': 5.0, 'loss_giou_interm_dn_7': 2.0, 'loss_ce_8': 4.0, 'loss_mask_8': 5.0, 'loss_dice_8': 5.0, 'loss_bbox_8': 5.0, 'loss_giou_8': 2.0, 'loss_ce_interm_8': 4.0, 'loss_mask_interm_8': 5.0, 'loss_dice_interm_8': 5.0, 'loss_bbox_interm_8': 5.0, 'loss_giou_interm_8': 2.0, 'loss_ce_dn_8': 4.0, 'loss_mask_dn_8': 5.0, 'loss_dice_dn_8': 5.0, 'loss_bbox_dn_8': 5.0, 'loss_giou_dn_8': 2.0, 'loss_ce_interm_dn_8': 4.0, 'loss_mask_interm_dn_8': 5.0, 'loss_dice_interm_dn_8': 5.0, 'loss_bbox_interm_dn_8': 5.0, 'loss_giou_interm_dn_8': 2.0}\n",
      "Evaluating detections...\n",
      "  90% |█████████████████|-| 61/68 [428.9ms elapsed, 49.2ms remaining, 142.2 samples/s]  "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/artemoneto/miniconda3/envs/maskdino/lib/python3.8/site-packages/shapely/set_operations.py:133: RuntimeWarning: invalid value encountered in intersection\n",
      "  return lib.intersection(a, b, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% |███████████████████| 68/68 [459.5ms elapsed, 0s remaining, 148.0 samples/s]      \n",
      "Performing IoU sweep...\n",
      " 100% |███████████████████| 68/68 [389.8ms elapsed, 0s remaining, 174.4 samples/s]      \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "     sitting       0.97      0.70      0.81        40\n",
      "     walking       0.84      0.59      0.70        27\n",
      "      waving       0.73      0.81      0.77        27\n",
      "\n",
      "   micro avg       0.85      0.70      0.77        94\n",
      "   macro avg       0.85      0.70      0.76        94\n",
      "weighted avg       0.86      0.70      0.77        94\n",
      "\n"
     ]
    }
   ],
   "source": [
    "import os, cv2\n",
    "from helpers.model_trainer import Predictor\n",
    "from helpers.fiftyone_detectron2_bridge import clean_instances, detectron_to_fo, get_fiftyone_dicts\n",
    "\n",
    "path = \"output/20\"\n",
    "cfg.OUTPUT_DIR = path\n",
    "\n",
    "for root, dirs, files in os.walk(path):\n",
    "    for name in files:\n",
    "        if '1_class' not in root and 'inference' not in root and 'example_detector' not in root:\n",
    "            if name[-3:] == 'pth':\n",
    "                #print(os.path.join(root, name))\n",
    "\n",
    "                print(f'Current model is:    {os.path.join(root, name)}')\n",
    "                print('')\n",
    "        \n",
    "                weight_path = os.path.join(os.path.join(root, name))\n",
    "                cfg.MODEL.WEIGHTS =  weight_path\n",
    "                \n",
    "                #cfg.MODEL.ROI_HEADS.SCORE_THRESH_TEST = 0.1\n",
    "                #cfg.MODEL.ROI_HEADS.NMS_THRESH_TEST = 0.8\n",
    "                \n",
    "                predictor = Predictor(cfg)\n",
    "        \n",
    "                test_view = test_dataset_combined\n",
    "                dataset_dicts = get_fiftyone_dicts(test_view, labels_dict)\n",
    "                predictions = {}\n",
    "                for d in dataset_dicts:\n",
    "                    img_w = d[\"width\"]\n",
    "                    img_h = d[\"height\"]\n",
    "                    img = cv2.imread(d[\"file_name\"], cv2.IMREAD_UNCHANGED)\n",
    "                    outputs = predictor(img)\n",
    "                    detections, instances = detectron_to_fo(outputs, img_w, img_h, classes, score_tresh=0.3)\n",
    "                    predictions[d[\"image_id\"]] = detections\n",
    "                \n",
    "                test_dataset_combined.set_values(\"predictions\", predictions, key_field=\"id\")\n",
    "\n",
    "                test_results = test_dataset_combined.evaluate_detections(\n",
    "                    \"predictions\",\n",
    "                    gt_field=\"segmentations\",\n",
    "                    eval_key=\"eval\",\n",
    "                    use_masks=True,\n",
    "                    compute_mAP=True,\n",
    "                    classes=test_dataset_combined.distinct(\"segmentations.detections.label\")\n",
    "                )\n",
    "                test_results.print_report()\n",
    "                "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "a0dc23f7-7a65-47fb-bac7-67ef2b27fc14",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Sample number:  1\n",
      "Detection number:  1\n",
      "\n",
      "Sample number:  2\n",
      "Detection number:  2\n",
      "\n",
      "Sample number:  3\n",
      "Detection number:  3\n",
      "\n",
      "Sample number:  4\n",
      "Detection number:  4\n",
      "\n",
      "Sample number:  5\n",
      "Detection number:  5\n",
      "\n",
      "Sample number:  6\n",
      "Detection number:  6\n",
      "\n",
      "Sample number:  7\n",
      "Detection number:  7\n",
      "\n",
      "Sample number:  8\n",
      "Detection number:  8\n",
      "\n",
      "Sample number:  9\n",
      "Detection number:  9\n",
      "\n",
      "Sample number:  10\n",
      "Detection number:  10\n",
      "\n",
      "Sample number:  11\n",
      "Detection number:  11\n",
      "\n",
      "Sample number:  12\n",
      "Detection number:  12\n",
      "\n",
      "Sample number:  13\n",
      "Detection number:  13\n",
      "\n",
      "Sample number:  14\n",
      "Detection number:  14\n",
      "\n",
      "Sample number:  15\n",
      "Detection number:  15\n",
      "\n",
      "Sample number:  16\n",
      "Detection number:  16\n",
      "\n",
      "Sample number:  17\n",
      "Detection number:  17\n",
      "\n",
      "Sample number:  18\n",
      "Detection number:  18\n",
      "\n",
      "Sample number:  19\n",
      "Detection number:  19\n",
      "\n",
      "Sample number:  20\n",
      "Detection number:  20\n",
      "\n",
      "Sample number:  21\n",
      "Detection number:  21\n",
      "\n",
      "Sample number:  22\n",
      "Detection number:  22\n",
      "\n",
      "Sample number:  23\n",
      "Detection number:  23\n",
      "\n",
      "Sample number:  24\n",
      "Detection number:  24\n",
      "\n",
      "Sample number:  25\n",
      "Detection number:  25\n",
      "\n",
      "Sample number:  26\n",
      "Detection number:  26\n",
      "\n",
      "Sample number:  27\n",
      "Detection number:  27\n",
      "\n",
      "Sample number:  28\n",
      "Detection number:  28\n",
      "Detection number:  29\n",
      "\n",
      "Sample number:  29\n",
      "Detection number:  30\n",
      "Detection number:  31\n",
      "\n",
      "Sample number:  30\n",
      "Detection number:  32\n",
      "Detection number:  33\n",
      "\n",
      "Sample number:  31\n",
      "Detection number:  34\n",
      "Detection number:  35\n",
      "\n",
      "Sample number:  32\n",
      "Detection number:  36\n",
      "Detection number:  37\n",
      "\n",
      "Sample number:  33\n",
      "Detection number:  38\n",
      "Detection number:  39\n",
      "\n",
      "Sample number:  34\n",
      "Detection number:  40\n",
      "Detection number:  41\n",
      "\n",
      "Sample number:  35\n",
      "Detection number:  42\n",
      "Detection number:  43\n",
      "\n",
      "Sample number:  36\n",
      "Detection number:  44\n",
      "Detection number:  45\n",
      "\n",
      "Sample number:  37\n",
      "Detection number:  46\n",
      "Detection number:  47\n",
      "\n",
      "Sample number:  38\n",
      "Detection number:  48\n",
      "Detection number:  49\n",
      "\n",
      "Sample number:  39\n",
      "Detection number:  50\n",
      "Detection number:  51\n",
      "Detection number:  52\n",
      "\n",
      "Sample number:  40\n",
      "Detection number:  53\n",
      "Detection number:  54\n",
      "Detection number:  55\n",
      "\n",
      "Sample number:  41\n",
      "Detection number:  56\n",
      "\n",
      "Sample number:  42\n",
      "Detection number:  57\n",
      "\n",
      "Sample number:  43\n",
      "Detection number:  58\n",
      "\n",
      "Sample number:  44\n",
      "Detection number:  59\n",
      "\n",
      "Sample number:  45\n",
      "Detection number:  60\n",
      "\n",
      "Sample number:  46\n",
      "Detection number:  61\n",
      "Detection number:  62\n",
      "\n",
      "Sample number:  47\n",
      "Detection number:  63\n",
      "Detection number:  64\n",
      "\n",
      "Sample number:  48\n",
      "Detection number:  65\n",
      "Detection number:  66\n",
      "\n",
      "Sample number:  49\n",
      "Detection number:  67\n",
      "Detection number:  68\n",
      "\n",
      "Sample number:  50\n",
      "Detection number:  69\n",
      "Detection number:  70\n",
      "\n",
      "Sample number:  51\n",
      "Detection number:  71\n",
      "Detection number:  72\n",
      "\n",
      "Sample number:  52\n",
      "Detection number:  73\n",
      "Detection number:  74\n",
      "\n",
      "Sample number:  53\n",
      "Detection number:  75\n",
      "Detection number:  76\n",
      "\n",
      "Sample number:  54\n",
      "Detection number:  77\n",
      "Detection number:  78\n",
      "\n",
      "Sample number:  55\n",
      "Detection number:  79\n",
      "Detection number:  80\n",
      "\n",
      "Sample number:  56\n",
      "Detection number:  81\n",
      "\n",
      "Sample number:  57\n",
      "Detection number:  82\n",
      "\n",
      "Sample number:  58\n",
      "Detection number:  83\n",
      "\n",
      "Sample number:  59\n",
      "Detection number:  84\n",
      "\n",
      "Sample number:  60\n",
      "Detection number:  85\n",
      "\n",
      "Sample number:  61\n",
      "Detection number:  86\n",
      "\n",
      "Sample number:  62\n",
      "Detection number:  87\n",
      "\n",
      "Sample number:  63\n",
      "Detection number:  88\n",
      "\n",
      "Sample number:  64\n",
      "Detection number:  89\n",
      "\n",
      "Sample number:  65\n",
      "Detection number:  90\n",
      "\n",
      "Sample number:  66\n",
      "Detection number:  91\n",
      "\n",
      "Sample number:  67\n",
      "Detection number:  92\n",
      "\n",
      "Sample number:  68\n",
      "Detection number:  93\n"
     ]
    }
   ],
   "source": [
    "samples = 0\n",
    "dets = 0\n",
    "\n",
    "for sample in test_view:\n",
    "    \n",
    "    samples += 1\n",
    "    print(\"\\nSample number: \", samples)\n",
    "    detections = sample.detections\n",
    "\n",
    "    for detection in detections.detections:\n",
    "        dets += 1\n",
    "        print(\"Detection number: \", dets)\n",
    "        #print(detection)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "474c79dd-deec-49d5-a0be-ba49b614df5b",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.18"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}