diff --git a/experiments.ipynb b/experiments.ipynb new file mode 100644 index 0000000..8c6ce93 --- /dev/null +++ b/experiments.ipynb @@ -0,0 +1,12692 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 1, + "id": "6ea2bcab", + "metadata": {}, + "outputs": [], + "source": [ + "%load_ext autoreload\n", + "%autoreload 2\n", + "%config Completer.use_jedi = False" + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "id": "4e50b369", + "metadata": {}, + "outputs": [], + "source": [ + "import pandas as pd\n", + "import torchvision" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "id": "a87d6b62", + "metadata": { + "scrolled": true + }, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/usr/local/lib/python3.8/site-packages/torch/utils/tensorboard/__init__.py:4: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.\n", + " if not hasattr(tensorboard, '__version__') or LooseVersion(tensorboard.__version__) < LooseVersion('1.15'):\n", + "/usr/local/lib/python3.8/site-packages/torch/utils/tensorboard/__init__.py:4: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.\n", + " if not hasattr(tensorboard, '__version__') or LooseVersion(tensorboard.__version__) < LooseVersion('1.15'):\n", + "/usr/local/lib/python3.8/site-packages/matplotlib/__init__.py:169: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.\n", + " if LooseVersion(module.__version__) < minver:\n", + "/usr/local/lib/python3.8/site-packages/setuptools/_distutils/version.py:351: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.\n", + " other = LooseVersion(other)\n", + "/usr/local/lib/python3.8/site-packages/matplotlib/__init__.py:169: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.\n", + " if LooseVersion(module.__version__) < minver:\n", + "/usr/local/lib/python3.8/site-packages/setuptools/_distutils/version.py:351: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.\n", + " other = LooseVersion(other)\n", + "/usr/local/lib/python3.8/site-packages/matplotlib/__init__.py:169: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.\n", + " if LooseVersion(module.__version__) < minver:\n", + "/usr/local/lib/python3.8/site-packages/setuptools/_distutils/version.py:351: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.\n", + " other = LooseVersion(other)\n", + "/usr/local/lib/python3.8/site-packages/matplotlib/__init__.py:169: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.\n", + " if LooseVersion(module.__version__) < minver:\n", + "/usr/local/lib/python3.8/site-packages/setuptools/_distutils/version.py:351: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.\n", + " other = LooseVersion(other)\n", + "/usr/local/lib/python3.8/site-packages/matplotlib/__init__.py:169: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.\n", + " if LooseVersion(module.__version__) < minver:\n", + "/usr/local/lib/python3.8/site-packages/setuptools/_distutils/version.py:351: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.\n", + " other = LooseVersion(other)\n" + ] + } + ], + "source": [ + "from heuristics.model.dataset import TorchDataset, get_pd_dataset, load_images\n", + "from heuristics.model.settings import IMAGES_DIR\n", + "from heuristics.model.classifier import RoomModel\n", + "from heuristics.model.trainer import TrainerUtils\n", + "from heuristics.model.utils import get_preprocessor\n", + "\n", + "from torch.utils.data import DataLoader\n", + "from torch.optim import AdamW\n", + "from torch import nn\n", + "import torch\n", + "from tqdm import tqdm_notebook\n", + "\n", + "import numpy as np\n", + "import os\n", + "from sklearn.metrics import classification_report\n", + "from transformers import get_linear_schedule_with_warmup\n", + "\n", + "from sklearn.metrics import (\n", + " ConfusionMatrixDisplay,\n", + " accuracy_score,\n", + " classification_report,\n", + " confusion_matrix,\n", + " f1_score,\n", + ")\n", + "import matplotlib.pyplot as plt\n", + "\n", + "from heuristics.model.trainer import predict_img_batch\n", + "\n", + "from PIL import Image\n", + "\n", + "from heuristics.model.metrics import Metrics\n", + "\n", + "from heuristics.model.settings import ROOM_TYPES, VALID_ROOM_TYPES, CLASS_NAME_MAPPING\n", + "\n", + "from heuristics.model.utils import plot_imgs_with_labels, plot_sample, load_images\n", + "\n", + "import logging\n", + "logger = logging.getLogger()\n", + "logger.setLevel(logging.DEBUG)" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "id": "8a24b2df", + "metadata": {}, + "outputs": [], + "source": [ + "# set your own paths\n", + "BASE_DIR_PATH = '/app/'\n", + "BASE_DATA_DIR = '/data/'\n", + "\n", + "CSV_DATA_DIR = os.path.join(BASE_DIR_PATH, '/heuristics/data/')\n", + "TRAIN_IMGS_DATA_DIR = os.path.join(BASE_DATA_DIR, 'train_images')\n", + "TEST_IMGS_DATA_DIR = os.path.join(BASE_DATA_DIR, 'test_images')\n", + "\n", + "TRAIN_DF_DIR = os.path.join(BASE_DIR_PATH, '/app/heuristics/data/AAA_dataset_course_ha_TOLOKA_dataset_new.csv')\n", + "TEST_DF_DIR = os.path.join(BASE_DIR_PATH, '/app/heuristics/data/AAA_dataset_course_ha_TRUE_TEST.csv')" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "id": "43578878", + "metadata": {}, + "outputs": [], + "source": [ + "train_df = pd.read_csv(TRAIN_DF_DIR)\n", + "test_df = pd.read_csv(TEST_DF_DIR)" + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "id": "a1d78137", + "metadata": { + "scrolled": true + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "CPU times: user 8.69 s, sys: 9.99 s, total: 18.7 s\n", + "Wall time: 10.4 s\n" + ] + } + ], + "source": [ + "%%time\n", + "# скачиваем картинки\n", + "load_images(train_df['image'], TRAIN_IMGS_DATA_DIR)\n", + "load_images(test_df['image'], TEST_IMGS_DATA_DIR)" + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "id": "e153bd43", + "metadata": { + "scrolled": true + }, + "outputs": [], + "source": [ + "# определим локальный путь до картинок\n", + "train_df['img_path'] = train_df['image'].map(\n", + " lambda x: os.path.join(TRAIN_IMGS_DATA_DIR, os.path.split(x)[-1])\n", + ")\n", + "test_df['img_path'] = test_df['image'].map(\n", + " lambda x: os.path.join(TEST_IMGS_DATA_DIR, os.path.split(x)[-1])\n", + ")" + ] + }, + { + "cell_type": "code", + "execution_count": 8, + "id": "fee91b5c", + "metadata": { + "scrolled": true + }, + "outputs": [ + { + "data": { + "text/plain": [ + "" + ] + }, + "execution_count": 8, + "metadata": {}, + "output_type": "execute_result" + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "train_df['room_type'] = train_df['result'].map(CLASS_NAME_MAPPING)\n", + "train_df.groupby('room_type').count().sort_values('result').plot.pie(y='result', figsize=(20, 20))" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "f67f56d0", + "metadata": {}, + "outputs": [], + "source": [ + "# картинки из одного айтема не могут лежать в разных сетах" + ] + }, + { + "cell_type": "code", + "execution_count": 9, + "id": "866e1cb9", + "metadata": { + "scrolled": true + }, + "outputs": [ + { + "data": { + "text/plain": [ + "гардеробная / кладовая / постирочная 202\n", + "кабинет 189\n", + "детская 65\n", + "другое 50\n", + "предметы интерьера / быт.техника 44\n", + "Name: label, dtype: int64" + ] + }, + "execution_count": 9, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "test_df[test_df['type'] == 'heuristics']['label'].value_counts()" + ] + }, + { + "cell_type": "code", + "execution_count": 10, + "id": "0b06987d", + "metadata": { + "scrolled": true + }, + "outputs": [], + "source": [ + "columns = [TorchDataset.img_path_column, TorchDataset.label_column, TorchDataset.image_id_column, 'label', 'ratio']\n", + "#print(columns)\n", + "\n", + "# train_df = pd.read_csv('/app/data/TOLOKA_dataset_HA_1.csv')#[columns]\n", + "\n", + "# test_df = pd.read_csv('/app/data/TEST_dataset_HA.csv')#[columns]\n", + "\n", + "# train_df['img_path'] = train_df['image_id_ext'].map(lambda x: f'/data/images_labeled/{int(x)}.jpg')\n", + "# test_df['img_path'] = test_df['image_id_ext'].map(lambda x: f'/data/images_labeled/{int(x)}.jpg')" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "0a8de958", + "metadata": {}, + "outputs": [], + "source": [ + "# train_df_new = pd.concat([train_df[train_df['label'] != 'детская'], \n", + "# train_df[train_df['label'] == 'детская'].sample(100)])" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "e69d57e6", + "metadata": { + "scrolled": true + }, + "outputs": [], + "source": [ + "dataset_train = TorchDataset(image_dir=IMAGES_DIR, df=train_df, transformer=get_preprocessor())\n", + "dataset_test = TorchDataset(image_dir=IMAGES_DIR, df=test_df, transformer=get_preprocessor())\n", + "train_dataloader = DataLoader(dataset_train, batch_size=32, shuffle=True)\n", + "test_dataloader = DataLoader(dataset_test, batch_size=32, shuffle=False)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "31b48f6b", + "metadata": { + "scrolled": true + }, + "outputs": [], + "source": [ + "device = 'cuda:3'\n", + "room_clf = RoomModel(num_classes=len(ROOM_TYPES))\n", + "\n", + "optimizer = AdamW(room_clf.parameters(), lr=0.002)\n", + "room_clf = room_clf.to(device)\n", + "\n", + "trainer = TrainerUtils(device=device, tensorboard_dir='/data/tensorboard',\n", + " experiment_tag='resnet18_baseline_detsk_less')\n", + "trainer.training_loop(\n", + " room_clf,\n", + " train_dataloader,\n", + " test_dataloader,\n", + " optimizer,\n", + " epoch_num=15,\n", + " validate_every=10,\n", + " verbose=True,\n", + ")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "0af53275", + "metadata": {}, + "outputs": [], + "source": [ + "room_clf, meta_info = RoomModel.from_pretrained('/app/data/models/model_resnet18_baseline_detsk_less')\n", + "test_predictions, test_probas, test_targets, _ = trainer.predict(\n", + " room_clf, test_dataloader, with_all_probas=False)\n", + "\n", + "metrics_scorer = Metrics(class_mapping=CLASS_NAME_MAPPING)\n", + "scores_df_base = metrics_scorer.get_accuracies_df(test_targets, test_predictions)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "f6861b7e", + "metadata": {}, + "outputs": [], + "source": [ + "scores_df_base" + ] + }, + { + "attachments": {}, + "cell_type": "markdown", + "id": "b0e7d316", + "metadata": {}, + "source": [ + "### Наша базовая модель с обучением" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "2dbc5ea0", + "metadata": { + "scrolled": true + }, + "outputs": [], + "source": [ + "device = 'cuda:3'\n", + "room_clf, meta_info = RoomModel.from_pretrained('/app/data/models/model_resnet18_baseline_detsk_less')\n", + "\n", + "optimizer = AdamW(room_clf.parameters(), lr=0.002)\n", + "room_clf = room_clf.to(device)\n", + "\n", + "trainer = TrainerUtils(device=device, tensorboard_dir='/data/tensorboard', experiment_tag='model_resnet18_baseline_detsk_less')\n", + "trainer.training_loop(\n", + " room_clf,\n", + " train_dataloader,\n", + " test_dataloader,\n", + " optimizer,\n", + " epoch_num=15,\n", + " validate_every=10,\n", + " verbose=True,\n", + ")\n", + "\n", + "test_predictions, test_probas, test_targets, _ = trainer.predict(room_clf, test_dataloader, with_all_probas=False)\n", + "\n", + "test_df['result_pred'] = test_predictions\n", + "test_df['label_pred'] = test_df['result_pred'].map(CLASS_NAME_MAPPING)\n", + "test_df['proba'] = test_probas\n", + "\n", + "scores_df_base= metrics_scorer.get_accuracies_df(test_targets, test_predictions)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "e8a2e192", + "metadata": {}, + "outputs": [], + "source": [ + "scores_df_base" + ] + }, + { + "cell_type": "markdown", + "id": "0e0419ff", + "metadata": {}, + "source": [ + "### Базовая модель без обучения" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "f5bd5d4d", + "metadata": {}, + "outputs": [], + "source": [ + "device = 'cuda:3'\n", + "room_clf, meta_info = RoomModel.from_pretrained('/app/data/models/model_resnet18_baseline_detsk_less')\n", + "\n", + "optimizer = AdamW(room_clf.parameters(), lr=0.002)\n", + "room_clf = room_clf.to(device)\n", + "\n", + "trainer = TrainerUtils(device=device, tensorboard_dir='/data/tensorboard', experiment_tag='model_resnet18_baseline_detsk_less')\n", + "\n", + "test_predictions, test_probas, test_targets, _ = trainer.predict(room_clf, test_dataloader, with_all_probas=False)\n", + "\n", + "test_df['result_pred'] = test_predictions\n", + "test_df['label_pred'] = test_df['result_pred'].map(CLASS_NAME_MAPPING)\n", + "test_df['proba'] = test_probas\n", + "\n", + "scores_df_base= metrics_scorer.get_accuracies_df(test_targets, test_predictions)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "26438180", + "metadata": {}, + "outputs": [], + "source": [ + "scores_df_base" + ] + }, + { + "attachments": {}, + "cell_type": "markdown", + "id": "cb096690", + "metadata": {}, + "source": [ + "## Эксперименты" + ] + }, + { + "attachments": {}, + "cell_type": "markdown", + "id": "16c0a6f4", + "metadata": {}, + "source": [ + "#### Аугментации" + ] + }, + { + "cell_type": "code", + "execution_count": 17, + "id": "203bc44a", + "metadata": {}, + "outputs": [], + "source": [ + "import math\n", + "from enum import Enum\n", + "from typing import Dict, List, Optional, Tuple, Any\n", + "\n", + "import torch\n", + "from torch import Tensor\n", + "import torchvision.transforms.functional as F\n", + "from torchvision.transforms import functional_tensor as F_t\n", + "from torchvision.transforms import functional_pil as F_pil\n", + "from torchvision.utils import _log_api_usage_once\n", + "from torchvision.transforms.functional import InterpolationMode\n", + "\n", + "\n", + "@torch.jit.unused\n", + "def _is_pil_image(img: Any) -> bool:\n", + " return isinstance(img, Image.Image)\n", + "\n", + "@torch.jit.unused\n", + "def get_dimensions_p(img: Any) -> List[int]:\n", + " if _is_pil_image(img):\n", + " if hasattr(img, \"getbands\"):\n", + " channels = len(img.getbands())\n", + " else:\n", + " channels = img.channels\n", + " width, height = img.size\n", + " return [channels, height, width]\n", + " raise TypeError(f\"Unexpected type {type(img)}\")\n", + "\n", + "def _is_tensor_a_torch_image(x: Tensor) -> bool:\n", + " return x.ndim >= 2\n", + "\n", + "\n", + "def _assert_image_tensor(img: Tensor) -> None:\n", + " if not _is_tensor_a_torch_image(img):\n", + " raise TypeError(\"Tensor is not a torch image.\")\n", + "\n", + "def get_dimensions_t(img: Tensor) -> List[int]:\n", + " _assert_image_tensor(img)\n", + " channels = 1 if img.ndim == 2 else img.shape[-3]\n", + " height, width = img.shape[-2:]\n", + " return [channels, height, width]\n", + "\n", + "def get_dimensions(img: Tensor) -> List[int]:\n", + " \"\"\"Returns the dimensions of an image as [channels, height, width].\n", + "\n", + " Args:\n", + " img (PIL Image or Tensor): The image to be checked.\n", + "\n", + " Returns:\n", + " List[int]: The image dimensions.\n", + " \"\"\"\n", + " if not torch.jit.is_scripting() and not torch.jit.is_tracing():\n", + " _log_api_usage_once(get_dimensions)\n", + " if isinstance(img, torch.Tensor):\n", + " return get_dimensions_t(img)\n", + "\n", + " return get_dimensions_p(img)\n", + "\n", + "def _apply_op(\n", + " img: Tensor, op_name: str, magnitude: float, interpolation: InterpolationMode, fill: Optional[List[float]]\n", + "):\n", + " if op_name == \"ShearX\":\n", + " # magnitude should be arctan(magnitude)\n", + " # official autoaug: (1, level, 0, 0, 1, 0)\n", + " # https://github.com/tensorflow/models/blob/dd02069717128186b88afa8d857ce57d17957f03/research/autoaugment/augmentation_transforms.py#L290\n", + " # compared to\n", + " # torchvision: (1, tan(level), 0, 0, 1, 0)\n", + " # https://github.com/pytorch/vision/blob/0c2373d0bba3499e95776e7936e207d8a1676e65/torchvision/transforms/functional.py#L976\n", + " img = F.affine(\n", + " img,\n", + " angle=0.0,\n", + " translate=[0, 0],\n", + " scale=1.0,\n", + " shear=[math.degrees(math.atan(magnitude)), 0.0],\n", + " interpolation=interpolation,\n", + " fill=fill,\n", + " center=[0, 0],\n", + " )\n", + " elif op_name == \"ShearY\":\n", + " # magnitude should be arctan(magnitude)\n", + " # See above\n", + " img = F.affine(\n", + " img,\n", + " angle=0.0,\n", + " translate=[0, 0],\n", + " scale=1.0,\n", + " shear=[0.0, math.degrees(math.atan(magnitude))],\n", + " interpolation=interpolation,\n", + " fill=fill,\n", + " center=[0, 0],\n", + " )\n", + " elif op_name == \"TranslateX\":\n", + " img = F.affine(\n", + " img,\n", + " angle=0.0,\n", + " translate=[int(magnitude), 0],\n", + " scale=1.0,\n", + " interpolation=interpolation,\n", + " shear=[0.0, 0.0],\n", + " fill=fill,\n", + " )\n", + " elif op_name == \"TranslateY\":\n", + " img = F.affine(\n", + " img,\n", + " angle=0.0,\n", + " translate=[0, int(magnitude)],\n", + " scale=1.0,\n", + " interpolation=interpolation,\n", + " shear=[0.0, 0.0],\n", + " fill=fill,\n", + " )\n", + " elif op_name == \"Rotate\":\n", + " img = F.rotate(img, magnitude, interpolation=interpolation, fill=fill)\n", + " elif op_name == \"Brightness\":\n", + " img = F.adjust_brightness(img, 1.0 + magnitude)\n", + " elif op_name == \"Color\":\n", + " img = F.adjust_saturation(img, 1.0 + magnitude)\n", + " elif op_name == \"Contrast\":\n", + " img = F.adjust_contrast(img, 1.0 + magnitude)\n", + " elif op_name == \"Sharpness\":\n", + " img = F.adjust_sharpness(img, 1.0 + magnitude)\n", + " elif op_name == \"Posterize\":\n", + " img = F.posterize(img, int(magnitude))\n", + " elif op_name == \"Solarize\":\n", + " img = F.solarize(img, magnitude)\n", + " elif op_name == \"AutoContrast\":\n", + " img = F.autocontrast(img)\n", + " elif op_name == \"Equalize\":\n", + " img = F.equalize(img)\n", + " elif op_name == \"Invert\":\n", + " img = F.invert(img)\n", + " elif op_name == \"Identity\":\n", + " pass\n", + " else:\n", + " raise ValueError(f\"The provided operator {op_name} is not recognized.\")\n", + " return img\n", + "\n", + "class AugMix(torch.nn.Module):\n", + " r\"\"\"AugMix data augmentation method based on\n", + " `\"AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty\" `_.\n", + " If the image is torch Tensor, it should be of type torch.uint8, and it is expected\n", + " to have [..., 1 or 3, H, W] shape, where ... means an arbitrary number of leading dimensions.\n", + " If img is PIL Image, it is expected to be in mode \"L\" or \"RGB\".\n", + "\n", + " Args:\n", + " severity (int): The severity of base augmentation operators. Default is ``3``.\n", + " mixture_width (int): The number of augmentation chains. Default is ``3``.\n", + " chain_depth (int): The depth of augmentation chains. A negative value denotes stochastic depth sampled from the interval [1, 3].\n", + " Default is ``-1``.\n", + " alpha (float): The hyperparameter for the probability distributions. Default is ``1.0``.\n", + " all_ops (bool): Use all operations (including brightness, contrast, color and sharpness). Default is ``True``.\n", + " interpolation (InterpolationMode): Desired interpolation enum defined by\n", + " :class:`torchvision.transforms.InterpolationMode`. Default is ``InterpolationMode.NEAREST``.\n", + " If input is Tensor, only ``InterpolationMode.NEAREST``, ``InterpolationMode.BILINEAR`` are supported.\n", + " fill (sequence or number, optional): Pixel fill value for the area outside the transformed\n", + " image. If given a number, the value is used for all bands respectively.\n", + " \"\"\"\n", + "\n", + " def __init__(\n", + " self,\n", + " severity: int = 3,\n", + " mixture_width: int = 3,\n", + " chain_depth: int = -1,\n", + " alpha: float = 1.0,\n", + " all_ops: bool = True,\n", + " interpolation: InterpolationMode = InterpolationMode.BILINEAR,\n", + " fill: Optional[List[float]] = None,\n", + " ) -> None:\n", + " super().__init__()\n", + " self._PARAMETER_MAX = 10\n", + " if not (1 <= severity <= self._PARAMETER_MAX):\n", + " raise ValueError(f\"The severity must be between [1, {self._PARAMETER_MAX}]. Got {severity} instead.\")\n", + " self.severity = severity\n", + " self.mixture_width = mixture_width\n", + " self.chain_depth = chain_depth\n", + " self.alpha = alpha\n", + " self.all_ops = all_ops\n", + " self.interpolation = interpolation\n", + " self.fill = fill\n", + "\n", + " def _augmentation_space(self, num_bins: int, image_size: Tuple[int, int]) -> Dict[str, Tuple[Tensor, bool]]:\n", + " s = {\n", + " # op_name: (magnitudes, signed)\n", + " \"ShearX\": (torch.linspace(0.0, 0.3, num_bins), True),\n", + " \"ShearY\": (torch.linspace(0.0, 0.3, num_bins), True),\n", + " \"TranslateX\": (torch.linspace(0.0, image_size[1] / 3.0, num_bins), True),\n", + " \"TranslateY\": (torch.linspace(0.0, image_size[0] / 3.0, num_bins), True),\n", + " \"Rotate\": (torch.linspace(0.0, 30.0, num_bins), True),\n", + " \"Posterize\": (4 - (torch.arange(num_bins) / ((num_bins - 1) / 4)).round().int(), False),\n", + " \"Solarize\": (torch.linspace(255.0, 0.0, num_bins), False),\n", + " \"AutoContrast\": (torch.tensor(0.0), False),\n", + " \"Equalize\": (torch.tensor(0.0), False),\n", + " }\n", + " if self.all_ops:\n", + " s.update(\n", + " {\n", + " \"Brightness\": (torch.linspace(0.0, 0.9, num_bins), True),\n", + " \"Color\": (torch.linspace(0.0, 0.9, num_bins), True),\n", + " \"Contrast\": (torch.linspace(0.0, 0.9, num_bins), True),\n", + " \"Sharpness\": (torch.linspace(0.0, 0.9, num_bins), True),\n", + " }\n", + " )\n", + " return s\n", + "\n", + " @torch.jit.unused\n", + " def _pil_to_tensor(self, img) -> Tensor:\n", + " return F.pil_to_tensor(img)\n", + "\n", + " @torch.jit.unused\n", + " def _tensor_to_pil(self, img: Tensor):\n", + " return F.to_pil_image(img)\n", + "\n", + " def _sample_dirichlet(self, params: Tensor) -> Tensor:\n", + " # Must be on a separate method so that we can overwrite it in tests.\n", + " return torch._sample_dirichlet(params)\n", + "\n", + " def forward(self, orig_img: Tensor) -> Tensor:\n", + " \"\"\"\n", + " img (PIL Image or Tensor): Image to be transformed.\n", + "\n", + " Returns:\n", + " PIL Image or Tensor: Transformed image.\n", + " \"\"\"\n", + " fill = self.fill\n", + " channels, height, width = get_dimensions(orig_img)\n", + " if isinstance(orig_img, Tensor):\n", + " img = orig_img\n", + " if isinstance(fill, (int, float)):\n", + " fill = [float(fill)] * channels\n", + " elif fill is not None:\n", + " fill = [float(f) for f in fill]\n", + " else:\n", + " img = self._pil_to_tensor(orig_img)\n", + "\n", + " op_meta = self._augmentation_space(self._PARAMETER_MAX, (height, width))\n", + "\n", + " orig_dims = list(img.shape)\n", + " batch = img.view([1] * max(4 - img.ndim, 0) + orig_dims)\n", + " batch_dims = [batch.size(0)] + [1] * (batch.ndim - 1)\n", + "\n", + " # Sample the beta weights for combining the original and augmented image. To get Beta, we use a Dirichlet\n", + " # with 2 parameters. The 1st column stores the weights of the original and the 2nd the ones of augmented image.\n", + " m = self._sample_dirichlet(\n", + " torch.tensor([self.alpha, self.alpha], device=batch.device).expand(batch_dims[0], -1)\n", + " )\n", + "\n", + " # Sample the mixing weights and combine them with the ones sampled from Beta for the augmented images.\n", + " combined_weights = self._sample_dirichlet(\n", + " torch.tensor([self.alpha] * self.mixture_width, device=batch.device).expand(batch_dims[0], -1)\n", + " ) * m[:, 1].view([batch_dims[0], -1])\n", + "\n", + " mix = m[:, 0].view(batch_dims) * batch\n", + " for i in range(self.mixture_width):\n", + " aug = batch\n", + " depth = self.chain_depth if self.chain_depth > 0 else int(torch.randint(low=1, high=4, size=(1,)).item())\n", + " for _ in range(depth):\n", + " op_index = int(torch.randint(len(op_meta), (1,)).item())\n", + " op_name = list(op_meta.keys())[op_index]\n", + " magnitudes, signed = op_meta[op_name]\n", + " magnitude = (\n", + " float(magnitudes[torch.randint(self.severity, (1,), dtype=torch.long)].item())\n", + " if magnitudes.ndim > 0\n", + " else 0.0\n", + " )\n", + " if signed and torch.randint(2, (1,)):\n", + " magnitude *= -1.0\n", + " aug = _apply_op(aug, op_name, magnitude, interpolation=self.interpolation, fill=fill)\n", + " mix.add_(combined_weights[:, i].view(batch_dims) * aug)\n", + " mix = mix.view(orig_dims).to(dtype=img.dtype)\n", + "\n", + " if not isinstance(orig_img, Tensor):\n", + " return self._tensor_to_pil(mix)\n", + " return mix\n", + "\n", + "\n", + " def __repr__(self) -> str:\n", + " s = (\n", + " f\"{self.__class__.__name__}(\"\n", + " f\"severity={self.severity}\"\n", + " f\", mixture_width={self.mixture_width}\"\n", + " f\", chain_depth={self.chain_depth}\"\n", + " f\", alpha={self.alpha}\"\n", + " f\", all_ops={self.all_ops}\"\n", + " f\", interpolation={self.interpolation}\"\n", + " f\", fill={self.fill}\"\n", + " f\")\"\n", + " )\n", + " return s" + ] + }, + { + "cell_type": "code", + "execution_count": 11, + "id": "08a319b0", + "metadata": {}, + "outputs": [ + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "56db773735e249d5a7ca34104523560a", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + " 0%| | 0/143 [00:00\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
indexprecisionrecallf1-scoresupport
21weighted avg0.5339340.5354430.5136614740.0
20macro avg0.5334480.5370320.5155194740.0
19micro avg0.5309620.5354430.5331934740.0
17предметы интерьера / быт.техника0.3795920.3345320.355641278.0
5кабинет0.3272730.0649820.108434277.0
11гардеробная / кладовая / постирочная0.4358970.0639100.111475266.0
16другое0.3162650.7984790.453074263.0
15подъезд / лестничная площадка0.6860990.6000000.640167255.0
10коридор / прихожая0.5480770.6732280.604240254.0
6детская0.3897440.3015870.340045252.0
12балкон / лоджия0.7984190.8015870.800000252.0
2универсальная комната0.3116280.2658730.286938252.0
0кухня / столовая0.5371900.5158730.526316252.0
8туалет0.7617020.7131470.736626251.0
18комната без мебели0.6421400.7649400.698182251.0
3гостиная0.3536230.4880000.410084250.0
14дом снаружи / двор0.7892560.7701610.779592248.0
4спальня0.4943820.5344130.513619247.0
9совмещенный санузел0.6040960.7195120.656772246.0
7ванная комната0.6666670.6122450.638298245.0
13вид из окна / с балкона0.7518520.8285710.788350245.0
1кухня-гостиная0.3416150.3525640.347003156.0
\n", + "" + ], + "text/plain": [ + " index precision recall f1-score \\\n", + "21 weighted avg 0.533934 0.535443 0.513661 \n", + "20 macro avg 0.533448 0.537032 0.515519 \n", + "19 micro avg 0.530962 0.535443 0.533193 \n", + "17 предметы интерьера / быт.техника 0.379592 0.334532 0.355641 \n", + "5 кабинет 0.327273 0.064982 0.108434 \n", + "11 гардеробная / кладовая / постирочная 0.435897 0.063910 0.111475 \n", + "16 другое 0.316265 0.798479 0.453074 \n", + "15 подъезд / лестничная площадка 0.686099 0.600000 0.640167 \n", + "10 коридор / прихожая 0.548077 0.673228 0.604240 \n", + "6 детская 0.389744 0.301587 0.340045 \n", + "12 балкон / лоджия 0.798419 0.801587 0.800000 \n", + "2 универсальная комната 0.311628 0.265873 0.286938 \n", + "0 кухня / столовая 0.537190 0.515873 0.526316 \n", + "8 туалет 0.761702 0.713147 0.736626 \n", + "18 комната без мебели 0.642140 0.764940 0.698182 \n", + "3 гостиная 0.353623 0.488000 0.410084 \n", + "14 дом снаружи / двор 0.789256 0.770161 0.779592 \n", + "4 спальня 0.494382 0.534413 0.513619 \n", + "9 совмещенный санузел 0.604096 0.719512 0.656772 \n", + "7 ванная комната 0.666667 0.612245 0.638298 \n", + "13 вид из окна / с балкона 0.751852 0.828571 0.788350 \n", + "1 кухня-гостиная 0.341615 0.352564 0.347003 \n", + "\n", + " support \n", + "21 4740.0 \n", + "20 4740.0 \n", + "19 4740.0 \n", + "17 278.0 \n", + "5 277.0 \n", + "11 266.0 \n", + "16 263.0 \n", + "15 255.0 \n", + "10 254.0 \n", + "6 252.0 \n", + "12 252.0 \n", + "2 252.0 \n", + "0 252.0 \n", + "8 251.0 \n", + "18 251.0 \n", + "3 250.0 \n", + "14 248.0 \n", + "4 247.0 \n", + "9 246.0 \n", + "7 245.0 \n", + "13 245.0 \n", + "1 156.0 " + ] + }, + "execution_count": 12, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "metrics_scorer = Metrics(class_mapping=CLASS_NAME_MAPPING)\n", + "scores_df_augm= metrics_scorer.get_accuracies_df(test_targets, test_predictions)\n", + "scores_df_augm" + ] + }, + { + "cell_type": "code", + "execution_count": 13, + "id": "5089b424", + "metadata": {}, + "outputs": [ + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "ee6dbb93703f4b20bb829f40ec9d8459", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + " 0%| | 0/143 [00:00\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
indexprecisionrecallf1-scoresupport
21weighted avg0.5414460.5343880.5156874740.0
20macro avg0.5412220.5359310.5177844740.0
19micro avg0.5347270.5343880.5345574740.0
17предметы интерьера / быт.техника0.4025970.3345320.365422278.0
5кабинет0.3773580.0722020.121212277.0
11гардеробная / кладовая / постирочная0.4888890.0827070.141479266.0
16другое0.3156340.8136880.454835263.0
15подъезд / лестничная площадка0.6570250.6235290.639839255.0
10коридор / прихожая0.5769230.6496060.611111254.0
6детская0.4047620.2698410.323810252.0
12балкон / лоджия0.8065840.7777780.791919252.0
2универсальная комната0.2851850.3055560.295019252.0
0кухня / столовая0.5152670.5357140.525292252.0
8туалет0.7428570.7251000.733871251.0
18комната без мебели0.6190480.7768920.689046251.0
3гостиная0.3724140.4320000.400000250.0
14дом снаружи / двор0.7824270.7540320.767967248.0
4спальня0.5289260.5182190.523517247.0
9совмещенный санузел0.6086960.6829270.643678246.0
7ванная комната0.6623380.6244900.642857245.0
13вид из окна / с балкона0.7527270.8448980.796154245.0
1кухня-гостиная0.3835620.3589740.370861156.0
\n", + "" + ], + "text/plain": [ + " index precision recall f1-score \\\n", + "21 weighted avg 0.541446 0.534388 0.515687 \n", + "20 macro avg 0.541222 0.535931 0.517784 \n", + "19 micro avg 0.534727 0.534388 0.534557 \n", + "17 предметы интерьера / быт.техника 0.402597 0.334532 0.365422 \n", + "5 кабинет 0.377358 0.072202 0.121212 \n", + "11 гардеробная / кладовая / постирочная 0.488889 0.082707 0.141479 \n", + "16 другое 0.315634 0.813688 0.454835 \n", + "15 подъезд / лестничная площадка 0.657025 0.623529 0.639839 \n", + "10 коридор / прихожая 0.576923 0.649606 0.611111 \n", + "6 детская 0.404762 0.269841 0.323810 \n", + "12 балкон / лоджия 0.806584 0.777778 0.791919 \n", + "2 универсальная комната 0.285185 0.305556 0.295019 \n", + "0 кухня / столовая 0.515267 0.535714 0.525292 \n", + "8 туалет 0.742857 0.725100 0.733871 \n", + "18 комната без мебели 0.619048 0.776892 0.689046 \n", + "3 гостиная 0.372414 0.432000 0.400000 \n", + "14 дом снаружи / двор 0.782427 0.754032 0.767967 \n", + "4 спальня 0.528926 0.518219 0.523517 \n", + "9 совмещенный санузел 0.608696 0.682927 0.643678 \n", + "7 ванная комната 0.662338 0.624490 0.642857 \n", + "13 вид из окна / с балкона 0.752727 0.844898 0.796154 \n", + "1 кухня-гостиная 0.383562 0.358974 0.370861 \n", + "\n", + " support \n", + "21 4740.0 \n", + "20 4740.0 \n", + "19 4740.0 \n", + "17 278.0 \n", + "5 277.0 \n", + "11 266.0 \n", + "16 263.0 \n", + "15 255.0 \n", + "10 254.0 \n", + "6 252.0 \n", + "12 252.0 \n", + "2 252.0 \n", + "0 252.0 \n", + "8 251.0 \n", + "18 251.0 \n", + "3 250.0 \n", + "14 248.0 \n", + "4 247.0 \n", + "9 246.0 \n", + "7 245.0 \n", + "13 245.0 \n", + "1 156.0 " + ] + }, + "execution_count": 14, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "scores_df_more_augm= metrics_scorer.get_accuracies_df(test_targets, test_predictions)\n", + "scores_df_more_augm" + ] + }, + { + "cell_type": "code", + "execution_count": 15, + "id": "da1600bf", + "metadata": {}, + "outputs": [ + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "566997420686490d9a1d8e0a978659e0", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + " 0%| | 0/143 [00:00 34\u001b[0m trainer\u001b[39m.\u001b[39;49mtraining_loop(\n\u001b[1;32m 35\u001b[0m room_clf,\n\u001b[1;32m 36\u001b[0m train_dataloader,\n\u001b[1;32m 37\u001b[0m test_dataloader,\n\u001b[1;32m 38\u001b[0m optimizer,\n\u001b[1;32m 39\u001b[0m epoch_num\u001b[39m=\u001b[39;49m\u001b[39m100\u001b[39;49m,\n\u001b[1;32m 40\u001b[0m validate_every\u001b[39m=\u001b[39;49m\u001b[39m10\u001b[39;49m,\n\u001b[1;32m 41\u001b[0m verbose\u001b[39m=\u001b[39;49m\u001b[39mTrue\u001b[39;49;00m,\n\u001b[1;32m 42\u001b[0m )\n\u001b[1;32m 44\u001b[0m test_predictions, test_probas, test_targets, _ \u001b[39m=\u001b[39m trainer\u001b[39m.\u001b[39mpredict(room_clf, test_dataloader, with_all_probas\u001b[39m=\u001b[39m\u001b[39mFalse\u001b[39;00m)\n\u001b[1;32m 46\u001b[0m test_df[\u001b[39m'\u001b[39m\u001b[39mresult_pred\u001b[39m\u001b[39m'\u001b[39m] \u001b[39m=\u001b[39m test_predictions\n", + "File \u001b[0;32m/app/heuristics/model/trainer.py:157\u001b[0m, in \u001b[0;36mTrainerUtils.training_loop\u001b[0;34m(self, model, train_dataloader, val_dataloader, optimizer, epoch_num, scheduler, grad_clipping_norm, validate_every, verbose, metrics_scorer)\u001b[0m\n\u001b[1;32m 155\u001b[0m model\u001b[39m.\u001b[39mto(\u001b[39mself\u001b[39m\u001b[39m.\u001b[39mdevice)\n\u001b[1;32m 156\u001b[0m \u001b[39mfor\u001b[39;00m step_number \u001b[39min\u001b[39;00m \u001b[39mrange\u001b[39m(epoch_len):\n\u001b[0;32m--> 157\u001b[0m batch \u001b[39m=\u001b[39m \u001b[39mnext\u001b[39;49m(iter_loader)\n\u001b[1;32m 158\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_global_steps_num \u001b[39m+\u001b[39m\u001b[39m=\u001b[39m \u001b[39m1\u001b[39m\n\u001b[1;32m 159\u001b[0m train_loss, train_accuracy \u001b[39m=\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mtraining_step(\n\u001b[1;32m 160\u001b[0m model, batch, optimizer, scheduler, grad_clipping_norm\n\u001b[1;32m 161\u001b[0m )\n", + "File \u001b[0;32m/usr/local/lib/python3.8/site-packages/torch/utils/data/dataloader.py:530\u001b[0m, in \u001b[0;36m_BaseDataLoaderIter.__next__\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m 528\u001b[0m \u001b[39mif\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_sampler_iter \u001b[39mis\u001b[39;00m \u001b[39mNone\u001b[39;00m:\n\u001b[1;32m 529\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_reset()\n\u001b[0;32m--> 530\u001b[0m data \u001b[39m=\u001b[39m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49m_next_data()\n\u001b[1;32m 531\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_num_yielded \u001b[39m+\u001b[39m\u001b[39m=\u001b[39m \u001b[39m1\u001b[39m\n\u001b[1;32m 532\u001b[0m \u001b[39mif\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_dataset_kind \u001b[39m==\u001b[39m _DatasetKind\u001b[39m.\u001b[39mIterable \u001b[39mand\u001b[39;00m \\\n\u001b[1;32m 533\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_IterableDataset_len_called \u001b[39mis\u001b[39;00m \u001b[39mnot\u001b[39;00m \u001b[39mNone\u001b[39;00m \u001b[39mand\u001b[39;00m \\\n\u001b[1;32m 534\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_num_yielded \u001b[39m>\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_IterableDataset_len_called:\n", + "File \u001b[0;32m/usr/local/lib/python3.8/site-packages/torch/utils/data/dataloader.py:570\u001b[0m, in \u001b[0;36m_SingleProcessDataLoaderIter._next_data\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m 568\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39m_next_data\u001b[39m(\u001b[39mself\u001b[39m):\n\u001b[1;32m 569\u001b[0m index \u001b[39m=\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_next_index() \u001b[39m# may raise StopIteration\u001b[39;00m\n\u001b[0;32m--> 570\u001b[0m data \u001b[39m=\u001b[39m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49m_dataset_fetcher\u001b[39m.\u001b[39;49mfetch(index) \u001b[39m# may raise StopIteration\u001b[39;00m\n\u001b[1;32m 571\u001b[0m \u001b[39mif\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_pin_memory:\n\u001b[1;32m 572\u001b[0m data \u001b[39m=\u001b[39m _utils\u001b[39m.\u001b[39mpin_memory\u001b[39m.\u001b[39mpin_memory(data)\n", + "File \u001b[0;32m/usr/local/lib/python3.8/site-packages/torch/utils/data/_utils/fetch.py:49\u001b[0m, in \u001b[0;36m_MapDatasetFetcher.fetch\u001b[0;34m(self, possibly_batched_index)\u001b[0m\n\u001b[1;32m 47\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39mfetch\u001b[39m(\u001b[39mself\u001b[39m, possibly_batched_index):\n\u001b[1;32m 48\u001b[0m \u001b[39mif\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mauto_collation:\n\u001b[0;32m---> 49\u001b[0m data \u001b[39m=\u001b[39m [\u001b[39mself\u001b[39m\u001b[39m.\u001b[39mdataset[idx] \u001b[39mfor\u001b[39;00m idx \u001b[39min\u001b[39;00m possibly_batched_index]\n\u001b[1;32m 50\u001b[0m \u001b[39melse\u001b[39;00m:\n\u001b[1;32m 51\u001b[0m data \u001b[39m=\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mdataset[possibly_batched_index]\n", + "File \u001b[0;32m/usr/local/lib/python3.8/site-packages/torch/utils/data/_utils/fetch.py:49\u001b[0m, in \u001b[0;36m\u001b[0;34m(.0)\u001b[0m\n\u001b[1;32m 47\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39mfetch\u001b[39m(\u001b[39mself\u001b[39m, possibly_batched_index):\n\u001b[1;32m 48\u001b[0m \u001b[39mif\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mauto_collation:\n\u001b[0;32m---> 49\u001b[0m data \u001b[39m=\u001b[39m [\u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mdataset[idx] \u001b[39mfor\u001b[39;00m idx \u001b[39min\u001b[39;00m possibly_batched_index]\n\u001b[1;32m 50\u001b[0m \u001b[39melse\u001b[39;00m:\n\u001b[1;32m 51\u001b[0m data \u001b[39m=\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mdataset[possibly_batched_index]\n", + "File \u001b[0;32m/app/heuristics/model/dataset.py:112\u001b[0m, in \u001b[0;36mTorchDataset.__getitem__\u001b[0;34m(self, idx)\u001b[0m\n\u001b[1;32m 109\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mimage_cache[idx] \u001b[39m=\u001b[39m img\n\u001b[1;32m 111\u001b[0m \u001b[39mif\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mcustom_transform:\n\u001b[0;32m--> 112\u001b[0m img \u001b[39m=\u001b[39m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mcustom_transform(img)\n\u001b[1;32m 114\u001b[0m \u001b[39m# item = {\u001b[39;00m\n\u001b[1;32m 115\u001b[0m \u001b[39m# 'img': img,\u001b[39;00m\n\u001b[1;32m 116\u001b[0m \u001b[39m# 'label': item_series['result'],\u001b[39;00m\n\u001b[1;32m 117\u001b[0m \u001b[39m# 'label_name': item_series['label']\u001b[39;00m\n\u001b[1;32m 118\u001b[0m \u001b[39m# }\u001b[39;00m\n\u001b[1;32m 120\u001b[0m \u001b[39mreturn\u001b[39;00m img, label, sample_weight\n", + "File \u001b[0;32m/usr/local/lib/python3.8/site-packages/torchvision/transforms/transforms.py:95\u001b[0m, in \u001b[0;36mCompose.__call__\u001b[0;34m(self, img)\u001b[0m\n\u001b[1;32m 93\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39m__call__\u001b[39m(\u001b[39mself\u001b[39m, img):\n\u001b[1;32m 94\u001b[0m \u001b[39mfor\u001b[39;00m t \u001b[39min\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mtransforms:\n\u001b[0;32m---> 95\u001b[0m img \u001b[39m=\u001b[39m t(img)\n\u001b[1;32m 96\u001b[0m \u001b[39mreturn\u001b[39;00m img\n", + "File \u001b[0;32m/usr/local/lib/python3.8/site-packages/torch/nn/modules/module.py:1110\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m 1106\u001b[0m \u001b[39m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m 1107\u001b[0m \u001b[39m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m 1108\u001b[0m \u001b[39mif\u001b[39;00m \u001b[39mnot\u001b[39;00m (\u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_backward_hooks \u001b[39mor\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_forward_hooks \u001b[39mor\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_forward_pre_hooks \u001b[39mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m 1109\u001b[0m \u001b[39mor\u001b[39;00m _global_forward_hooks \u001b[39mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1110\u001b[0m \u001b[39mreturn\u001b[39;00m forward_call(\u001b[39m*\u001b[39;49m\u001b[39minput\u001b[39;49m, \u001b[39m*\u001b[39;49m\u001b[39m*\u001b[39;49mkwargs)\n\u001b[1;32m 1111\u001b[0m \u001b[39m# Do not call functions when jit is used\u001b[39;00m\n\u001b[1;32m 1112\u001b[0m full_backward_hooks, non_full_backward_hooks \u001b[39m=\u001b[39m [], []\n", + "File \u001b[0;32m/usr/local/lib/python3.8/site-packages/torchvision/transforms/autoaugment.py:361\u001b[0m, in \u001b[0;36mRandAugment.forward\u001b[0;34m(self, img)\u001b[0m\n\u001b[1;32m 359\u001b[0m \u001b[39mif\u001b[39;00m signed \u001b[39mand\u001b[39;00m torch\u001b[39m.\u001b[39mrandint(\u001b[39m2\u001b[39m, (\u001b[39m1\u001b[39m,)):\n\u001b[1;32m 360\u001b[0m magnitude \u001b[39m*\u001b[39m\u001b[39m=\u001b[39m \u001b[39m-\u001b[39m\u001b[39m1.0\u001b[39m\n\u001b[0;32m--> 361\u001b[0m img \u001b[39m=\u001b[39m _apply_op(img, op_name, magnitude, interpolation\u001b[39m=\u001b[39;49m\u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49minterpolation, fill\u001b[39m=\u001b[39;49mfill)\n\u001b[1;32m 363\u001b[0m \u001b[39mreturn\u001b[39;00m img\n", + "File \u001b[0;32m/usr/local/lib/python3.8/site-packages/torchvision/transforms/autoaugment.py:73\u001b[0m, in \u001b[0;36m_apply_op\u001b[0;34m(img, op_name, magnitude, interpolation, fill)\u001b[0m\n\u001b[1;32m 71\u001b[0m img \u001b[39m=\u001b[39m F\u001b[39m.\u001b[39madjust_saturation(img, \u001b[39m1.0\u001b[39m \u001b[39m+\u001b[39m magnitude)\n\u001b[1;32m 72\u001b[0m \u001b[39melif\u001b[39;00m op_name \u001b[39m==\u001b[39m \u001b[39m\"\u001b[39m\u001b[39mContrast\u001b[39m\u001b[39m\"\u001b[39m:\n\u001b[0;32m---> 73\u001b[0m img \u001b[39m=\u001b[39m F\u001b[39m.\u001b[39;49madjust_contrast(img, \u001b[39m1.0\u001b[39;49m \u001b[39m+\u001b[39;49m magnitude)\n\u001b[1;32m 74\u001b[0m \u001b[39melif\u001b[39;00m op_name \u001b[39m==\u001b[39m \u001b[39m\"\u001b[39m\u001b[39mSharpness\u001b[39m\u001b[39m\"\u001b[39m:\n\u001b[1;32m 75\u001b[0m img \u001b[39m=\u001b[39m F\u001b[39m.\u001b[39madjust_sharpness(img, \u001b[39m1.0\u001b[39m \u001b[39m+\u001b[39m magnitude)\n", + "File \u001b[0;32m/usr/local/lib/python3.8/site-packages/torchvision/transforms/functional.py:844\u001b[0m, in \u001b[0;36madjust_contrast\u001b[0;34m(img, contrast_factor)\u001b[0m\n\u001b[1;32m 842\u001b[0m _log_api_usage_once(adjust_contrast)\n\u001b[1;32m 843\u001b[0m \u001b[39mif\u001b[39;00m \u001b[39mnot\u001b[39;00m \u001b[39misinstance\u001b[39m(img, torch\u001b[39m.\u001b[39mTensor):\n\u001b[0;32m--> 844\u001b[0m \u001b[39mreturn\u001b[39;00m F_pil\u001b[39m.\u001b[39;49madjust_contrast(img, contrast_factor)\n\u001b[1;32m 846\u001b[0m \u001b[39mreturn\u001b[39;00m F_t\u001b[39m.\u001b[39madjust_contrast(img, contrast_factor)\n", + "File \u001b[0;32m/usr/local/lib/python3.8/site-packages/torchvision/transforms/functional_pil.py:68\u001b[0m, in \u001b[0;36madjust_contrast\u001b[0;34m(img, contrast_factor)\u001b[0m\n\u001b[1;32m 65\u001b[0m \u001b[39mif\u001b[39;00m \u001b[39mnot\u001b[39;00m _is_pil_image(img):\n\u001b[1;32m 66\u001b[0m \u001b[39mraise\u001b[39;00m \u001b[39mTypeError\u001b[39;00m(\u001b[39mf\u001b[39m\u001b[39m\"\u001b[39m\u001b[39mimg should be PIL Image. Got \u001b[39m\u001b[39m{\u001b[39;00m\u001b[39mtype\u001b[39m(img)\u001b[39m}\u001b[39;00m\u001b[39m\"\u001b[39m)\n\u001b[0;32m---> 68\u001b[0m enhancer \u001b[39m=\u001b[39m ImageEnhance\u001b[39m.\u001b[39;49mContrast(img)\n\u001b[1;32m 69\u001b[0m img \u001b[39m=\u001b[39m enhancer\u001b[39m.\u001b[39menhance(contrast_factor)\n\u001b[1;32m 70\u001b[0m \u001b[39mreturn\u001b[39;00m img\n", + "File \u001b[0;32m/usr/local/lib/python3.8/site-packages/PIL/ImageEnhance.py:68\u001b[0m, in \u001b[0;36mContrast.__init__\u001b[0;34m(self, image)\u001b[0m\n\u001b[1;32m 66\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mimage \u001b[39m=\u001b[39m image\n\u001b[1;32m 67\u001b[0m mean \u001b[39m=\u001b[39m \u001b[39mint\u001b[39m(ImageStat\u001b[39m.\u001b[39mStat(image\u001b[39m.\u001b[39mconvert(\u001b[39m\"\u001b[39m\u001b[39mL\u001b[39m\u001b[39m\"\u001b[39m))\u001b[39m.\u001b[39mmean[\u001b[39m0\u001b[39m] \u001b[39m+\u001b[39m \u001b[39m0.5\u001b[39m)\n\u001b[0;32m---> 68\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mdegenerate \u001b[39m=\u001b[39m Image\u001b[39m.\u001b[39;49mnew(\u001b[39m\"\u001b[39;49m\u001b[39mL\u001b[39;49m\u001b[39m\"\u001b[39;49m, image\u001b[39m.\u001b[39;49msize, mean)\u001b[39m.\u001b[39;49mconvert(image\u001b[39m.\u001b[39;49mmode)\n\u001b[1;32m 70\u001b[0m \u001b[39mif\u001b[39;00m \u001b[39m\"\u001b[39m\u001b[39mA\u001b[39m\u001b[39m\"\u001b[39m \u001b[39min\u001b[39;00m image\u001b[39m.\u001b[39mgetbands():\n\u001b[1;32m 71\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mdegenerate\u001b[39m.\u001b[39mputalpha(image\u001b[39m.\u001b[39mgetchannel(\u001b[39m\"\u001b[39m\u001b[39mA\u001b[39m\u001b[39m\"\u001b[39m))\n", + "File \u001b[0;32m/usr/local/lib/python3.8/site-packages/PIL/Image.py:1081\u001b[0m, in \u001b[0;36mImage.convert\u001b[0;34m(self, mode, matrix, dither, palette, colors)\u001b[0m\n\u001b[1;32m 1078\u001b[0m dither \u001b[39m=\u001b[39m Dither\u001b[39m.\u001b[39mFLOYDSTEINBERG\n\u001b[1;32m 1080\u001b[0m \u001b[39mtry\u001b[39;00m:\n\u001b[0;32m-> 1081\u001b[0m im \u001b[39m=\u001b[39m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mim\u001b[39m.\u001b[39;49mconvert(mode, dither)\n\u001b[1;32m 1082\u001b[0m \u001b[39mexcept\u001b[39;00m \u001b[39mValueError\u001b[39;00m:\n\u001b[1;32m 1083\u001b[0m \u001b[39mtry\u001b[39;00m:\n\u001b[1;32m 1084\u001b[0m \u001b[39m# normalize source image and try again\u001b[39;00m\n", + "\u001b[0;31mKeyboardInterrupt\u001b[0m: " + ] + } + ], + "source": [ + "from torchvision import transforms\n", + "from heuristics.model.utils import PadCustom, MAX_IMG_SIZE\n", + "\n", + "\n", + "device = 'cuda:3'\n", + "room_clf, meta_info = RoomModel.from_pretrained('/app/data/models/model_resnet18_baseline_detsk_less')\n", + "\n", + "optimizer = AdamW(room_clf.parameters(), lr=0.00002) # 0.490988\n", + "room_clf = room_clf.to(device)\n", + "\n", + "custom_transformer = transforms.Compose(\n", + " [\n", + " # transforms.RandomRotation((0, 10)), # 0.498493\n", + " # transforms.RandomPerspective(distortion_scale=0.1, p=0.1), # 0.492617\n", + " # transforms.Grayscale(num_output_channels=3), # 0.454670\n", + " # transforms.ColorJitter(brightness=0.1, hue=0.1, contrast=0.1), # 0.274195\n", + " # transforms.AutoAugment(transforms.AutoAugmentPolicy.IMAGENET), # 0.473332\n", + " transforms.RandAugment(num_magnitude_bins=50, magnitude=17, num_ops=4), # 0.517\n", + " # transforms.TrivialAugmentWide(), # 0.485550\n", + " PadCustom(MAX_IMG_SIZE),\n", + " transforms.Resize(256),\n", + " transforms.CenterCrop(224),\n", + " transforms.ToTensor(),\n", + " transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),\n", + " ]\n", + " )\n", + "\n", + "dataset_train = TorchDataset(image_dir=IMAGES_DIR, df=train_df, custom_transform=custom_transformer)\n", + "dataset_test = TorchDataset(image_dir=IMAGES_DIR, df=test_df, transformer=get_preprocessor())\n", + "train_dataloader = DataLoader(dataset_train, batch_size=32, shuffle=True)\n", + "test_dataloader = DataLoader(dataset_test, batch_size=32, shuffle=False)\n", + "\n", + "trainer = TrainerUtils(device=device, tensorboard_dir='/data/tensorboard', experiment_tag='esnet18_baseline_detsk_less_augm_many_epochs_rand_more_augm')\n", + "trainer.training_loop(\n", + " room_clf,\n", + " train_dataloader,\n", + " test_dataloader,\n", + " optimizer,\n", + " epoch_num=100,\n", + " validate_every=10,\n", + " verbose=True,\n", + ")\n", + "\n", + "test_predictions, test_probas, test_targets, _ = trainer.predict(room_clf, test_dataloader, with_all_probas=False)\n", + "\n", + "test_df['result_pred'] = test_predictions\n", + "test_df['label_pred'] = test_df['result_pred'].map(CLASS_NAME_MAPPING)\n", + "test_df['proba'] = test_probas\n" + ] + }, + { + "cell_type": "code", + "execution_count": 16, + "id": "2a45604b", + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
indexprecisionrecallf1-scoresupport
21weighted avg0.5414460.5343880.5156874740.0
20macro avg0.5412220.5359310.5177844740.0
19micro avg0.5347270.5343880.5345574740.0
17предметы интерьера / быт.техника0.4025970.3345320.365422278.0
5кабинет0.3773580.0722020.121212277.0
11гардеробная / кладовая / постирочная0.4888890.0827070.141479266.0
16другое0.3156340.8136880.454835263.0
15подъезд / лестничная площадка0.6570250.6235290.639839255.0
10коридор / прихожая0.5769230.6496060.611111254.0
6детская0.4047620.2698410.323810252.0
12балкон / лоджия0.8065840.7777780.791919252.0
2универсальная комната0.2851850.3055560.295019252.0
0кухня / столовая0.5152670.5357140.525292252.0
8туалет0.7428570.7251000.733871251.0
18комната без мебели0.6190480.7768920.689046251.0
3гостиная0.3724140.4320000.400000250.0
14дом снаружи / двор0.7824270.7540320.767967248.0
4спальня0.5289260.5182190.523517247.0
9совмещенный санузел0.6086960.6829270.643678246.0
7ванная комната0.6623380.6244900.642857245.0
13вид из окна / с балкона0.7527270.8448980.796154245.0
1кухня-гостиная0.3835620.3589740.370861156.0
\n", + "
" + ], + "text/plain": [ + " index precision recall f1-score \\\n", + "21 weighted avg 0.541446 0.534388 0.515687 \n", + "20 macro avg 0.541222 0.535931 0.517784 \n", + "19 micro avg 0.534727 0.534388 0.534557 \n", + "17 предметы интерьера / быт.техника 0.402597 0.334532 0.365422 \n", + "5 кабинет 0.377358 0.072202 0.121212 \n", + "11 гардеробная / кладовая / постирочная 0.488889 0.082707 0.141479 \n", + "16 другое 0.315634 0.813688 0.454835 \n", + "15 подъезд / лестничная площадка 0.657025 0.623529 0.639839 \n", + "10 коридор / прихожая 0.576923 0.649606 0.611111 \n", + "6 детская 0.404762 0.269841 0.323810 \n", + "12 балкон / лоджия 0.806584 0.777778 0.791919 \n", + "2 универсальная комната 0.285185 0.305556 0.295019 \n", + "0 кухня / столовая 0.515267 0.535714 0.525292 \n", + "8 туалет 0.742857 0.725100 0.733871 \n", + "18 комната без мебели 0.619048 0.776892 0.689046 \n", + "3 гостиная 0.372414 0.432000 0.400000 \n", + "14 дом снаружи / двор 0.782427 0.754032 0.767967 \n", + "4 спальня 0.528926 0.518219 0.523517 \n", + "9 совмещенный санузел 0.608696 0.682927 0.643678 \n", + "7 ванная комната 0.662338 0.624490 0.642857 \n", + "13 вид из окна / с балкона 0.752727 0.844898 0.796154 \n", + "1 кухня-гостиная 0.383562 0.358974 0.370861 \n", + "\n", + " support \n", + "21 4740.0 \n", + "20 4740.0 \n", + "19 4740.0 \n", + "17 278.0 \n", + "5 277.0 \n", + "11 266.0 \n", + "16 263.0 \n", + "15 255.0 \n", + "10 254.0 \n", + "6 252.0 \n", + "12 252.0 \n", + "2 252.0 \n", + "0 252.0 \n", + "8 251.0 \n", + "18 251.0 \n", + "3 250.0 \n", + "14 248.0 \n", + "4 247.0 \n", + "9 246.0 \n", + "7 245.0 \n", + "13 245.0 \n", + "1 156.0 " + ] + }, + "execution_count": 16, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "scores_df_stronger_augm= metrics_scorer.get_accuracies_df(test_targets, test_predictions)\n", + "scores_df_stronger_augm" + ] + }, + { + "cell_type": "code", + "execution_count": 19, + "id": "11d91299", + "metadata": {}, + "outputs": [ + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "3be104bb1457453d84659a9102912f91", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + " 0%| | 0/143 [00:00\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
indexprecisionrecallf1-scoresupport
21weighted avg0.5117850.5107590.4960944740.0
20macro avg0.5123390.5134670.4986684740.0
19micro avg0.5110830.5107590.5109214740.0
17предметы интерьера / быт.техника0.3734440.3237410.346821278.0
5кабинет0.3157890.0649820.107784277.0
11гардеробная / кладовая / постирочная0.2906980.0939850.142045266.0
16другое0.2995660.7870720.433962263.0
15подъезд / лестничная площадка0.5860810.6274510.606061255.0
10коридор / прихожая0.5359480.6456690.585714254.0
6детская0.3370170.2420630.281755252.0
12балкон / лоджия0.7975210.7658730.781377252.0
2универсальная комната0.2812500.2857140.283465252.0
0кухня / столовая0.5132740.4603170.485356252.0
8туалет0.7709250.6972110.732218251.0
18комната без мебели0.6926230.6733070.682828251.0
3гостиная0.3517790.3560000.353877250.0
14дом снаружи / двор0.7333330.7983870.764479248.0
4спальня0.4782610.4898790.484000247.0
9совмещенный санузел0.5818820.6788620.626642246.0
7ванная комната0.6208330.6081630.614433245.0
13вид из окна / с балкона0.8206280.7469390.782051245.0
1кухня-гостиная0.3535910.4102560.379822156.0
\n", + "" + ], + "text/plain": [ + " index precision recall f1-score \\\n", + "21 weighted avg 0.511785 0.510759 0.496094 \n", + "20 macro avg 0.512339 0.513467 0.498668 \n", + "19 micro avg 0.511083 0.510759 0.510921 \n", + "17 предметы интерьера / быт.техника 0.373444 0.323741 0.346821 \n", + "5 кабинет 0.315789 0.064982 0.107784 \n", + "11 гардеробная / кладовая / постирочная 0.290698 0.093985 0.142045 \n", + "16 другое 0.299566 0.787072 0.433962 \n", + "15 подъезд / лестничная площадка 0.586081 0.627451 0.606061 \n", + "10 коридор / прихожая 0.535948 0.645669 0.585714 \n", + "6 детская 0.337017 0.242063 0.281755 \n", + "12 балкон / лоджия 0.797521 0.765873 0.781377 \n", + "2 универсальная комната 0.281250 0.285714 0.283465 \n", + "0 кухня / столовая 0.513274 0.460317 0.485356 \n", + "8 туалет 0.770925 0.697211 0.732218 \n", + "18 комната без мебели 0.692623 0.673307 0.682828 \n", + "3 гостиная 0.351779 0.356000 0.353877 \n", + "14 дом снаружи / двор 0.733333 0.798387 0.764479 \n", + "4 спальня 0.478261 0.489879 0.484000 \n", + "9 совмещенный санузел 0.581882 0.678862 0.626642 \n", + "7 ванная комната 0.620833 0.608163 0.614433 \n", + "13 вид из окна / с балкона 0.820628 0.746939 0.782051 \n", + "1 кухня-гостиная 0.353591 0.410256 0.379822 \n", + "\n", + " support \n", + "21 4740.0 \n", + "20 4740.0 \n", + "19 4740.0 \n", + "17 278.0 \n", + "5 277.0 \n", + "11 266.0 \n", + "16 263.0 \n", + "15 255.0 \n", + "10 254.0 \n", + "6 252.0 \n", + "12 252.0 \n", + "2 252.0 \n", + "0 252.0 \n", + "8 251.0 \n", + "18 251.0 \n", + "3 250.0 \n", + "14 248.0 \n", + "4 247.0 \n", + "9 246.0 \n", + "7 245.0 \n", + "13 245.0 \n", + "1 156.0 " + ] + }, + "execution_count": 20, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "scores_df_augmix= metrics_scorer.get_accuracies_df(test_targets, test_predictions)\n", + "scores_df_augmix" + ] + } + ], + "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": "3.8.6" + }, + "vscode": { + "interpreter": { + "hash": "aee8b7b246df8f9039afb4144a1f6fd8d2ca17a180786b69acc140d282b71a49" + } + } + }, + "nbformat": 4, + "nbformat_minor": 5 +} diff --git a/heuristics/model/dataset.py b/heuristics/model/dataset.py index b4c2b37..3fc1014 100644 --- a/heuristics/model/dataset.py +++ b/heuristics/model/dataset.py @@ -2,10 +2,49 @@ from typing import Dict, Optional import pandas as pd +from avito_ds_swat_utils.storages.dwh import VerticaPandasConnection +from avito_ds_swat_utils.utils.image_manager import ImageManager from PIL import Image from torch.utils.data import Dataset from torchvision import transforms +from .settings import ( + IMAGES_DIR, + LABELS_PATH, + MAX_IMG_SIZE_STR, +) + + +def get_pd_dataset(cache_path=LABELS_PATH): + if cache_path is None or not os.path.exists(cache_path): + with VerticaPandasConnection() as dwh: + df = dwh.sql_to_pd('select * from dsswat.datasets_course_room_type') + + df = df.dropna(subset=['image_id_ext']) + df['image_id_ext'] = df['image_id_ext'].astype(int) + df.reset_index(inplace=True) + if cache_path is not None: + df.to_csv(cache_path, index=False) + else: + df = pd.read_csv(cache_path) + + return df + + +def load_images(image_ids, save_dir=IMAGES_DIR, img_size=MAX_IMG_SIZE_STR, image_save_shards=False): + image_manager = ImageManager() + + image_manager.process_images( + image_id_list=image_ids, + schema='item', + size=img_size, + version=1, + private=True, + image_save_to_disk=True, + image_root_dir=save_dir, + image_save_shards=image_save_shards, + ) + class TorchDataset(Dataset): image_id_column: str = 'image_id_ext' @@ -20,6 +59,7 @@ def __init__( image_dir: str = None, transformer: Optional[transforms.transforms.Compose] = None, normalize_sample_weights: bool = True, + custom_transform : Optional[transforms.transforms.Compose] = None, ): super(TorchDataset, self).__init__() self.image_dir = image_dir @@ -30,6 +70,7 @@ def __init__( # self.df = self.df[self.df['result'] != -1] self.transformer = transformer self.image_cache = {} + self.custom_transform = custom_transform def img_path_from_id(self, image_id): path = os.path.join(self.image_dir, f'{image_id}.{self.img_extension}') @@ -67,6 +108,9 @@ def __getitem__(self, idx): self.image_cache[idx] = img + if self.custom_transform: + img = self.custom_transform(img) + # item = { # 'img': img, # 'label': item_series['result'],