{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# AMMICO Demonstration Notebook\n",
    "With ammico, you can analyze text on images and image content at the same time. This is a demonstration notebook to showcase the capabilities of ammico.\n",
    "You can run this notebook on google colab or locally / on your own HPC resource. The first cell only runs on google colab; on all other machines, you need to create a conda environment first and install ammico from the Python Package Index using  \n",
    "```pip install ammico```  \n",
    "Alternatively you can install the development version from the GitHub repository  \n",
    "```pip install git+https://github.com/ssciwr/AMMICO.git```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:42:44.420408Z",
     "iopub.status.busy": "2024-02-19T08:42:44.420216Z",
     "iopub.status.idle": "2024-02-19T08:42:44.428568Z",
     "shell.execute_reply": "2024-02-19T08:42:44.428037Z"
    }
   },
   "outputs": [],
   "source": [
    "# if running on google colab\n",
    "# flake8-noqa-cell\n",
    "\n",
    "if \"google.colab\" in str(get_ipython()):\n",
    "    # update python version\n",
    "    # install setuptools\n",
    "    # %pip install setuptools==61 -qqq\n",
    "    # uninstall some pre-installed packages due to incompatibility\n",
    "    %pip uninstall tensorflow-probability dopamine-rl lida pandas-gbq torchaudio torchdata torchtext orbax-checkpoint flex-y -qqq\n",
    "    # install ammico\n",
    "    %pip install git+https://github.com/ssciwr/ammico.git -qqq\n",
    "    # mount google drive for data and API key\n",
    "    from google.colab import drive\n",
    "\n",
    "    drive.mount(\"/content/drive\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Use a test dataset\n",
    "You can download a dataset for test purposes. Skip this step if you use your own data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:42:44.430935Z",
     "iopub.status.busy": "2024-02-19T08:42:44.430571Z",
     "iopub.status.idle": "2024-02-19T08:42:51.757352Z",
     "shell.execute_reply": "2024-02-19T08:42:51.756689Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading readme:   0%|          | 0.00/21.0 [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading readme: 100%|██████████| 21.0/21.0 [00:00<00:00, 178kB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading data:   0%|          | 0.00/59.0k [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading data: 100%|██████████| 59.0k/59.0k [00:00<00:00, 158kB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading data: 100%|██████████| 59.0k/59.0k [00:00<00:00, 158kB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading data:   0%|          | 0.00/48.8k [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading data: 100%|██████████| 48.8k/48.8k [00:00<00:00, 167kB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading data: 100%|██████████| 48.8k/48.8k [00:00<00:00, 166kB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading data:   0%|          | 0.00/43.4k [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading data: 100%|██████████| 43.4k/43.4k [00:00<00:00, 153kB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading data: 100%|██████████| 43.4k/43.4k [00:00<00:00, 152kB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading data:   0%|          | 0.00/315k [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading data: 100%|██████████| 315k/315k [00:00<00:00, 748kB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading data: 100%|██████████| 315k/315k [00:00<00:00, 746kB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading data:   0%|          | 0.00/1.33M [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading data: 100%|██████████| 1.33M/1.33M [00:00<00:00, 3.27MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading data: 100%|██████████| 1.33M/1.33M [00:00<00:00, 3.26MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading data:   0%|          | 0.00/687k [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading data: 100%|██████████| 687k/687k [00:00<00:00, 1.89MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Downloading data: 100%|██████████| 687k/687k [00:00<00:00, 1.89MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Generating train split: 0 examples [00:00, ? examples/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "Generating train split: 6 examples [00:00, 423.54 examples/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "from datasets import load_dataset\n",
    "from pathlib import Path\n",
    "\n",
    "# If the dataset is gated/private, make sure you have run huggingface-cli login\n",
    "dataset = load_dataset(\"iulusoy/test-images\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next you need to provide a path for the saved images - a folder where the data is stored locally. This directory is automatically created if it does not exist."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:42:51.760070Z",
     "iopub.status.busy": "2024-02-19T08:42:51.759727Z",
     "iopub.status.idle": "2024-02-19T08:42:52.642174Z",
     "shell.execute_reply": "2024-02-19T08:42:52.641528Z"
    }
   },
   "outputs": [],
   "source": [
    "data_path = \"./data-test\"\n",
    "data_path = Path(data_path)\n",
    "data_path.mkdir(parents=True, exist_ok=True)\n",
    "# now save the files from the Huggingface dataset as images into the data_path folder\n",
    "for i, image in enumerate(dataset[\"train\"][\"image\"]):\n",
    "    filename = \"img\" + str(i) + \".png\"\n",
    "    image.save(data_path / filename)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Import the ammico package."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:42:52.645250Z",
     "iopub.status.busy": "2024-02-19T08:42:52.644756Z",
     "iopub.status.idle": "2024-02-19T08:43:01.940227Z",
     "shell.execute_reply": "2024-02-19T08:43:01.939551Z"
    }
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import ammico\n",
    "# for displaying a progress bar\n",
    "from tqdm import tqdm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Sometimes you may need to restart a session after installing the correct versions of packages, because `Tensorflow` and `EmotitionDetector` may not work and give an error. You can check it by running this code: \n",
    "```\n",
    "import tensorflow as tf\n",
    "tf.ones([2, 2])\n",
    "```\n",
    "If this code generates an error, you need to restart session. For this please click `Runtime` -> `Restart session`. And rerun the notebook again. All required packages will already be installed, so the execution will be very fast. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Step 0: Create and set a Google Cloud Vision Key\n",
    "\n",
    "Please note that for the [Google Cloud Vision API](https://cloud.google.com/vision/docs/setup) (the TextDetector class) you need to set a key in order to process the images. A key is generated following [these instructions](../set_up_credentials.md). This key is ideally set as an environment variable using for example\n",
    "```\n",
    "os.environ[\n",
    "    \"GOOGLE_APPLICATION_CREDENTIALS\"\n",
    "] = \"/content/drive/MyDrive/misinformation-data/misinformation-campaign-981aa55a3b13.json\"\n",
    "```\n",
    "where you place the key on your Google Drive if running on colab, or place it in a local folder on your machine."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:43:01.943248Z",
     "iopub.status.busy": "2024-02-19T08:43:01.942735Z",
     "iopub.status.idle": "2024-02-19T08:43:01.945881Z",
     "shell.execute_reply": "2024-02-19T08:43:01.945338Z"
    }
   },
   "outputs": [],
   "source": [
    "# os.environ[\"GOOGLE_APPLICATION_CREDENTIALS\"] = \"/content/drive/MyDrive/misinformation-data/misinformation-campaign-981aa55a3b13.json\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Step 1: Read your data into AMMICO\n",
    "The ammico package reads in one or several input files given in a folder for processing. The user can select to read in all image files in a folder, to include subfolders via the `recursive` option, and can select the file extension that should be considered (for example, only \"jpg\" files, or both \"jpg\" and \"png\" files). For reading in the files, the ammico function `find_files` is used, with optional keywords:\n",
    "\n",
    "| input key | input type | possible input values |\n",
    "| --------- | ---------- | --------------------- |\n",
    "`path` | `str` | the directory containing the image files (defaults to the location set by environment variable `AMMICO_DATA_HOME`) |\n",
    "| `pattern` | `str\\|list` | the file extensions to consider (defaults to \"png\", \"jpg\", \"jpeg\", \"gif\", \"webp\", \"avif\", \"tiff\") |\n",
    "| `recursive` | `bool` | include subdirectories recursively (defaults to `True`) |\n",
    "| `limit` | `int` | maximum number of files to read (defaults to `20`, for all images set to `None` or `-1`) |\n",
    "| `random_seed` | `str` | the random seed for shuffling the images; applies when only a few images are read and the selection should be preserved (defaults to `None`) |\n",
    "\n",
    "The `find_files` function returns a nested dict that contains the file ids and the paths to the files and is empty otherwise. This dict is filled step by step with more data as each detector class is run on the data (see below).\n",
    "\n",
    "If you downloaded the test dataset above, you can directly provide the path you already set for the test directory, `data_path`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:43:01.948209Z",
     "iopub.status.busy": "2024-02-19T08:43:01.947847Z",
     "iopub.status.idle": "2024-02-19T08:43:01.952312Z",
     "shell.execute_reply": "2024-02-19T08:43:01.951836Z"
    }
   },
   "outputs": [],
   "source": [
    "image_dict = ammico.find_files(\n",
    "    # path=\"/content/drive/MyDrive/misinformation-data/\",\n",
    "    path=data_path.as_posix(),\n",
    "    limit=15,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Step 2: Inspect the input files using the graphical user interface\n",
    "A Dash user interface is to select the most suitable options for the analysis, before running a complete analysis on the whole data set. The options for each detector module are explained below in the corresponding sections; for example, different models can be selected that will provide slightly different results. This way, the user can interactively explore which settings provide the most accurate results. In the interface, the nested `image_dict` is passed through the `AnalysisExplorer` class. The interface is run on a specific port which is passed using the `port` keyword; if a port is already in use, it will return an error message, in which case the user should select a different port number. \n",
    "The interface opens a dash app inside the Jupyter Notebook and allows selection of the input file in the top left dropdown menu, as well as selection of the detector type in the top right, with options for each detector type as explained below. The output of the detector is shown directly on the right next to the image. This way, the user can directly inspect how updating the options for each detector changes the computed results, and find the best settings for a production run."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:43:01.954766Z",
     "iopub.status.busy": "2024-02-19T08:43:01.954396Z",
     "iopub.status.idle": "2024-02-19T08:43:01.978861Z",
     "shell.execute_reply": "2024-02-19T08:43:01.977627Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "        <iframe\n",
       "            width=\"100%\"\n",
       "            height=\"650\"\n",
       "            src=\"http://127.0.0.1:8055/\"\n",
       "            frameborder=\"0\"\n",
       "            allowfullscreen\n",
       "            \n",
       "        ></iframe>\n",
       "        "
      ],
      "text/plain": [
       "<IPython.lib.display.IFrame at 0x7f966243a8e0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "analysis_explorer = ammico.AnalysisExplorer(image_dict)\n",
    "analysis_explorer.run_server(port=8055)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Step 3: Analyze all images\n",
    "The analysis can be run in production on all images in the data set. Depending on the size of the data set and the computing resources available, this can take some time. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It is also possible to set the dump file  creation `dump_file` in order to save the calculated data every `dump_every` images. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:43:01.981825Z",
     "iopub.status.busy": "2024-02-19T08:43:01.981158Z",
     "iopub.status.idle": "2024-02-19T08:43:01.984935Z",
     "shell.execute_reply": "2024-02-19T08:43:01.983983Z"
    }
   },
   "outputs": [],
   "source": [
    "# dump file name\n",
    "dump_file = \"dump_file.csv\"\n",
    "# dump every N images \n",
    "dump_every = 10"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The desired detector modules are called sequentially in any order, for example the `EmotionDetector`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:43:01.988248Z",
     "iopub.status.busy": "2024-02-19T08:43:01.987632Z",
     "iopub.status.idle": "2024-02-19T08:44:04.645259Z",
     "shell.execute_reply": "2024-02-19T08:44:04.644561Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0/6 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Downloading data from 'https://github.com/serengil/deepface_models/releases/download/v1.0/retinaface.h5' to file '/home/runner/.cache/pooch/3be32af6e4183fa0156bc33bda371147-retinaface.h5'.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 17%|█▋        | 1/6 [00:08<00:41,  8.32s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found missing key(s) {'no_faces', 'age', 'wears_mask', 'emotion (category)', 'multiple_faces', 'emotion', 'gender', 'race', 'face'} in subdict img1 - setting to None.\n",
      "Found missing key(s) {'no_faces', 'age', 'wears_mask', 'emotion (category)', 'multiple_faces', 'emotion', 'gender', 'race', 'face'} in subdict img2 - setting to None.\n",
      "Found missing key(s) {'no_faces', 'age', 'wears_mask', 'emotion (category)', 'multiple_faces', 'emotion', 'gender', 'race', 'face'} in subdict img3 - setting to None.\n",
      "Found missing key(s) {'no_faces', 'age', 'wears_mask', 'emotion (category)', 'multiple_faces', 'emotion', 'gender', 'race', 'face'} in subdict img0 - setting to None.\n",
      "Found missing key(s) {'no_faces', 'age', 'wears_mask', 'emotion (category)', 'multiple_faces', 'emotion', 'gender', 'race', 'face'} in subdict img5 - setting to None.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 33%|███▎      | 2/6 [00:10<00:19,  4.89s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 50%|█████     | 3/6 [00:13<00:11,  3.70s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 67%|██████▋   | 4/6 [00:15<00:06,  3.14s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 83%|████████▎ | 5/6 [00:17<00:02,  2.82s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Downloading data from 'https://github.com/chandrikadeb7/Face-Mask-Detection/raw/v1.0.0/mask_detector.model' to file '/home/runner/.cache/pooch/865b4b1e20f798935b70082440d5fb21-mask_detector.model'.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "1/1 [==============================] - ETA: 0s"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r",
      "1/1 [==============================] - 1s 523ms/step\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Downloading data from 'https://github.com/serengil/deepface_models/releases/download/v1.0/age_model_weights.h5' to file '/home/runner/.cache/pooch/39859d3331cd91ac06154cc306e1acc8-age_model_weights.h5'.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Downloading data from 'https://github.com/serengil/deepface_models/releases/download/v1.0/facial_expression_model_weights.h5' to file '/home/runner/.cache/pooch/dd5d5d6d8f5cecdc0fa6cb34d4d82d16-facial_expression_model_weights.h5'.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Downloading data from 'https://github.com/serengil/deepface_models/releases/download/v1.0/gender_model_weights.h5' to file '/home/runner/.cache/pooch/2e0d8fb96c5ee966ade0f3f2360f6478-gender_model_weights.h5'.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Downloading data from 'https://github.com/serengil/deepface_models/releases/download/v1.0/race_model_single_batch.h5' to file '/home/runner/.cache/pooch/382cb5446128012fa5305ddb9d608751-race_model_single_batch.h5'.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "1/1 [==============================] - ETA: 0s"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r",
      "1/1 [==============================] - 0s 319ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "1/1 [==============================] - ETA: 0s"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r",
      "1/1 [==============================] - 0s 324ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "1/1 [==============================] - ETA: 0s"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r",
      "1/1 [==============================] - 0s 320ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "1/1 [==============================] - ETA: 0s"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r",
      "1/1 [==============================] - 0s 62ms/step\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 6/6 [01:02<00:00, 17.17s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 6/6 [01:02<00:00, 10.44s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "for num, key in tqdm(enumerate(image_dict.keys()), total=len(image_dict)):    # loop through all images\n",
    "    image_dict[key] = ammico.EmotionDetector(image_dict[key]).analyse_image() # analyse image with EmotionDetector and update dict\n",
    "    \n",
    "    if num % dump_every == 0 or num == len(image_dict) - 1:      # save results every dump_every to dump_file\n",
    "        image_df = ammico.get_dataframe(image_dict)\n",
    "        image_df.to_csv(dump_file)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`TextDetector`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:44:04.647938Z",
     "iopub.status.busy": "2024-02-19T08:44:04.647467Z",
     "iopub.status.idle": "2024-02-19T08:45:58.641407Z",
     "shell.execute_reply": "2024-02-19T08:45:58.640478Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0/6 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting en-core-web-md==3.7.1\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Downloading https://github.com/explosion/spacy-models/releases/download/en_core_web_md-3.7.1/en_core_web_md-3.7.1-py3-none-any.whl (42.8 MB)\n",
      "\u001b[?25l     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m0.0/42.8 MB\u001b[0m \u001b[31m?\u001b[0m eta \u001b[36m-:--:--\u001b[0m\r",
      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m0.4/42.8 MB\u001b[0m \u001b[31m11.6 MB/s\u001b[0m eta \u001b[36m0:00:04\u001b[0m\r",
      "\u001b[2K     \u001b[91m╸\u001b[0m\u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m1.0/42.8 MB\u001b[0m \u001b[31m14.9 MB/s\u001b[0m eta \u001b[36m0:00:03\u001b[0m\r",
      "\u001b[2K     \u001b[91m━\u001b[0m\u001b[91m╸\u001b[0m\u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m1.8/42.8 MB\u001b[0m \u001b[31m17.9 MB/s\u001b[0m eta \u001b[36m0:00:03\u001b[0m\r",
      "\u001b[2K     \u001b[91m━━\u001b[0m\u001b[91m╸\u001b[0m\u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m2.8/42.8 MB\u001b[0m \u001b[31m20.5 MB/s\u001b[0m eta \u001b[36m0:00:02\u001b[0m\r",
      "\u001b[2K     \u001b[91m━━━\u001b[0m\u001b[91m╸\u001b[0m\u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m4.1/42.8 MB\u001b[0m \u001b[31m23.5 MB/s\u001b[0m eta \u001b[36m0:00:02\u001b[0m"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "\u001b[2K     \u001b[91m━━━━━\u001b[0m\u001b[90m╺\u001b[0m\u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m5.6/42.8 MB\u001b[0m \u001b[31m26.6 MB/s\u001b[0m eta \u001b[36m0:00:02\u001b[0m\r",
      "\u001b[2K     \u001b[91m━━━━━━\u001b[0m\u001b[91m╸\u001b[0m\u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m7.4/42.8 MB\u001b[0m \u001b[31m30.2 MB/s\u001b[0m eta \u001b[36m0:00:02\u001b[0m\r",
      "\u001b[2K     \u001b[91m━━━━━━━━\u001b[0m\u001b[91m╸\u001b[0m\u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m9.5/42.8 MB\u001b[0m \u001b[31m34.0 MB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m\r",
      "\u001b[2K     \u001b[91m━━━━━━━━━━━\u001b[0m\u001b[90m╺\u001b[0m\u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m12.2/42.8 MB\u001b[0m \u001b[31m49.2 MB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m\r",
      "\u001b[2K     \u001b[91m━━━━━━━━━━━━━━\u001b[0m\u001b[90m╺\u001b[0m\u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m15.3/42.8 MB\u001b[0m \u001b[31m68.1 MB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m\r",
      "\u001b[2K     \u001b[91m━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[90m╺\u001b[0m\u001b[90m━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m19.3/42.8 MB\u001b[0m \u001b[31m92.4 MB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "\u001b[2K     \u001b[91m━━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[91m╸\u001b[0m\u001b[90m━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m23.8/42.8 MB\u001b[0m \u001b[31m115.6 MB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m\r",
      "\u001b[2K     \u001b[91m━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[91m╸\u001b[0m\u001b[90m━━━━━━━━━━━━━\u001b[0m \u001b[32m28.2/42.8 MB\u001b[0m \u001b[31m125.8 MB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m\r",
      "\u001b[2K     \u001b[91m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[91m╸\u001b[0m\u001b[90m━━━━━━━━━\u001b[0m \u001b[32m32.9/42.8 MB\u001b[0m \u001b[31m128.8 MB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m\r",
      "\u001b[2K     \u001b[91m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[90m╺\u001b[0m\u001b[90m━━━━\u001b[0m \u001b[32m37.8/42.8 MB\u001b[0m \u001b[31m134.9 MB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m\r",
      "\u001b[2K     \u001b[91m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[91m╸\u001b[0m \u001b[32m42.8/42.8 MB\u001b[0m \u001b[31m145.2 MB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m\r",
      "\u001b[2K     \u001b[91m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[91m╸\u001b[0m \u001b[32m42.8/42.8 MB\u001b[0m \u001b[31m145.2 MB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "\u001b[2K     \u001b[91m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[91m╸\u001b[0m \u001b[32m42.8/42.8 MB\u001b[0m \u001b[31m145.2 MB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m\r",
      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m42.8/42.8 MB\u001b[0m \u001b[31m60.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[?25hRequirement already satisfied: spacy<3.8.0,>=3.7.2 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from en-core-web-md==3.7.1) (3.7.4)\n",
      "Requirement already satisfied: spacy-legacy<3.1.0,>=3.0.11 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (3.0.12)\n",
      "Requirement already satisfied: spacy-loggers<2.0.0,>=1.0.0 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (1.0.5)\n",
      "Requirement already satisfied: murmurhash<1.1.0,>=0.28.0 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (1.0.10)\n",
      "Requirement already satisfied: cymem<2.1.0,>=2.0.2 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (2.0.8)\n",
      "Requirement already satisfied: preshed<3.1.0,>=3.0.2 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (3.0.9)\n",
      "Requirement already satisfied: thinc<8.3.0,>=8.2.2 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (8.2.3)\n",
      "Requirement already satisfied: wasabi<1.2.0,>=0.9.1 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (1.1.2)\n",
      "Requirement already satisfied: srsly<3.0.0,>=2.4.3 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (2.4.8)\n",
      "Requirement already satisfied: catalogue<2.1.0,>=2.0.6 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (2.0.10)\n",
      "Requirement already satisfied: weasel<0.4.0,>=0.1.0 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (0.3.4)\n",
      "Requirement already satisfied: typer<0.10.0,>=0.3.0 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (0.9.0)\n",
      "Requirement already satisfied: smart-open<7.0.0,>=5.2.1 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (6.4.0)\n",
      "Requirement already satisfied: tqdm<5.0.0,>=4.38.0 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (4.66.2)\n",
      "Requirement already satisfied: requests<3.0.0,>=2.13.0 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (2.31.0)\n",
      "Requirement already satisfied: pydantic!=1.8,!=1.8.1,<3.0.0,>=1.7.4 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (1.10.14)\n",
      "Requirement already satisfied: jinja2 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (3.1.3)\n",
      "Requirement already satisfied: setuptools in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (58.1.0)\n",
      "Requirement already satisfied: packaging>=20.0 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (23.2)\n",
      "Requirement already satisfied: langcodes<4.0.0,>=3.2.0 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (3.3.0)\n",
      "Requirement already satisfied: numpy>=1.19.0 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (1.23.4)\n",
      "Requirement already satisfied: typing-extensions>=4.2.0 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from pydantic!=1.8,!=1.8.1,<3.0.0,>=1.7.4->spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (4.5.0)\n",
      "Requirement already satisfied: charset-normalizer<4,>=2 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from requests<3.0.0,>=2.13.0->spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (3.3.2)\n",
      "Requirement already satisfied: idna<4,>=2.5 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from requests<3.0.0,>=2.13.0->spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (2.10)\n",
      "Requirement already satisfied: urllib3<3,>=1.21.1 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from requests<3.0.0,>=2.13.0->spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (2.2.1)\n",
      "Requirement already satisfied: certifi>=2017.4.17 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from requests<3.0.0,>=2.13.0->spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (2024.2.2)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: blis<0.8.0,>=0.7.8 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from thinc<8.3.0,>=8.2.2->spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (0.7.11)\n",
      "Requirement already satisfied: confection<1.0.0,>=0.0.1 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from thinc<8.3.0,>=8.2.2->spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (0.1.4)\n",
      "Requirement already satisfied: click<9.0.0,>=7.1.1 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from typer<0.10.0,>=0.3.0->spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (8.1.7)\n",
      "Requirement already satisfied: cloudpathlib<0.17.0,>=0.7.0 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from weasel<0.4.0,>=0.1.0->spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (0.16.0)\n",
      "Requirement already satisfied: MarkupSafe>=2.0 in /opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages (from jinja2->spacy<3.8.0,>=3.7.2->en-core-web-md==3.7.1) (2.1.5)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Installing collected packages: en-core-web-md\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Successfully installed en-core-web-md-3.7.1\n",
      "\u001b[38;5;2m✔ Download and installation successful\u001b[0m\n",
      "You can now load the package via spacy.load('en_core_web_md')\n",
      "\u001b[38;5;3m⚠ Restart to reload dependencies\u001b[0m\n",
      "If you are in a Jupyter or Colab notebook, you may need to restart Python in\n",
      "order to load all the package's dependencies. You can do this by selecting the\n",
      "'Restart kernel' or 'Restart runtime' option.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m A new release of pip is available: \u001b[0m\u001b[31;49m23.0.1\u001b[0m\u001b[39;49m -> \u001b[0m\u001b[32;49m24.0\u001b[0m\n",
      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m To update, run: \u001b[0m\u001b[32;49mpip install --upgrade pip\u001b[0m\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "config.json:   0%|          | 0.00/1.80k [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "config.json: 100%|██████████| 1.80k/1.80k [00:00<00:00, 731kB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   0%|          | 0.00/1.22G [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   2%|▏         | 21.0M/1.22G [00:00<00:08, 142MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   3%|▎         | 41.9M/1.22G [00:00<00:06, 175MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   6%|▌         | 73.4M/1.22G [00:00<00:05, 193MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   8%|▊         | 94.4M/1.22G [00:00<00:05, 198MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   9%|▉         | 115M/1.22G [00:00<00:05, 198MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  11%|█         | 136M/1.22G [00:00<00:05, 201MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  13%|█▎        | 157M/1.22G [00:00<00:05, 201MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  15%|█▍        | 178M/1.22G [00:00<00:05, 202MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  16%|█▋        | 199M/1.22G [00:01<00:05, 204MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  18%|█▊        | 220M/1.22G [00:01<00:04, 202MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  20%|█▉        | 241M/1.22G [00:01<00:04, 202MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  21%|██▏       | 262M/1.22G [00:01<00:04, 204MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  23%|██▎       | 283M/1.22G [00:01<00:04, 204MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  25%|██▍       | 304M/1.22G [00:01<00:04, 202MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  27%|██▋       | 325M/1.22G [00:01<00:04, 201MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  28%|██▊       | 346M/1.22G [00:01<00:04, 201MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  30%|███       | 367M/1.22G [00:01<00:04, 202MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  32%|███▏      | 388M/1.22G [00:01<00:04, 202MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  34%|███▍      | 419M/1.22G [00:02<00:03, 206MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  37%|███▋      | 451M/1.22G [00:02<00:03, 205MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  39%|███▊      | 472M/1.22G [00:02<00:03, 205MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  40%|████      | 493M/1.22G [00:02<00:03, 203MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  42%|████▏     | 514M/1.22G [00:02<00:03, 204MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  44%|████▍     | 535M/1.22G [00:03<00:15, 43.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  45%|████▌     | 556M/1.22G [00:04<00:12, 52.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  47%|████▋     | 577M/1.22G [00:04<00:10, 64.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  49%|████▉     | 598M/1.22G [00:04<00:08, 77.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  51%|█████     | 619M/1.22G [00:04<00:06, 90.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  52%|█████▏    | 640M/1.22G [00:06<00:16, 36.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  54%|█████▍    | 661M/1.22G [00:07<00:21, 25.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  56%|█████▌    | 682M/1.22G [00:07<00:16, 33.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  57%|█████▋    | 703M/1.22G [00:07<00:11, 44.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  59%|█████▉    | 724M/1.22G [00:07<00:08, 58.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  61%|██████    | 744M/1.22G [00:07<00:06, 74.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  63%|██████▎   | 765M/1.22G [00:07<00:04, 91.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  64%|██████▍   | 786M/1.22G [00:08<00:03, 110MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  66%|██████▌   | 807M/1.22G [00:08<00:03, 128MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  68%|██████▊   | 828M/1.22G [00:08<00:02, 143MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  69%|██████▉   | 849M/1.22G [00:08<00:02, 156MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  71%|███████   | 870M/1.22G [00:08<00:02, 160MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  73%|███████▎  | 891M/1.22G [00:08<00:01, 170MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  75%|███████▍  | 912M/1.22G [00:08<00:01, 174MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  76%|███████▋  | 933M/1.22G [00:08<00:01, 179MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  78%|███████▊  | 954M/1.22G [00:08<00:01, 184MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  80%|███████▉  | 975M/1.22G [00:09<00:01, 188MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  81%|████████▏ | 996M/1.22G [00:10<00:05, 39.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  83%|████████▎ | 1.02G/1.22G [00:10<00:04, 50.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  85%|████████▍ | 1.04G/1.22G [00:10<00:02, 62.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  87%|████████▋ | 1.06G/1.22G [00:11<00:02, 76.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  88%|████████▊ | 1.08G/1.22G [00:11<00:01, 91.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  90%|█████████ | 1.10G/1.22G [00:11<00:01, 109MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  92%|█████████▏| 1.12G/1.22G [00:11<00:00, 126MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  94%|█████████▎| 1.14G/1.22G [00:11<00:00, 142MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  96%|█████████▌| 1.17G/1.22G [00:11<00:00, 163MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  98%|█████████▊| 1.20G/1.22G [00:11<00:00, 171MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin: 100%|██████████| 1.22G/1.22G [00:11<00:00, 182MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin: 100%|██████████| 1.22G/1.22G [00:11<00:00, 103MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "tokenizer_config.json:   0%|          | 0.00/26.0 [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "tokenizer_config.json: 100%|██████████| 26.0/26.0 [00:00<00:00, 31.8kB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "vocab.json:   0%|          | 0.00/899k [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "vocab.json: 100%|██████████| 899k/899k [00:00<00:00, 4.69MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "vocab.json: 100%|██████████| 899k/899k [00:00<00:00, 4.62MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "merges.txt:   0%|          | 0.00/456k [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "merges.txt: 100%|██████████| 456k/456k [00:00<00:00, 2.35MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "merges.txt: 100%|██████████| 456k/456k [00:00<00:00, 2.32MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "config.json:   0%|          | 0.00/629 [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "config.json: 100%|██████████| 629/629 [00:00<00:00, 128kB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   0%|          | 0.00/268M [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   4%|▍         | 10.5M/268M [00:00<00:06, 42.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  16%|█▌        | 41.9M/268M [00:00<00:01, 123MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  27%|██▋       | 73.4M/268M [00:00<00:01, 164MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  39%|███▉      | 105M/268M [00:00<00:00, 186MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  51%|█████     | 136M/268M [00:00<00:00, 196MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  63%|██████▎   | 168M/268M [00:00<00:00, 201MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  70%|███████   | 189M/268M [00:01<00:00, 202MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  78%|███████▊  | 210M/268M [00:01<00:00, 203MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  90%|█████████ | 241M/268M [00:01<00:00, 206MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin: 100%|██████████| 268M/268M [00:01<00:00, 207MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin: 100%|██████████| 268M/268M [00:01<00:00, 185MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "tokenizer_config.json:   0%|          | 0.00/48.0 [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "tokenizer_config.json: 100%|██████████| 48.0/48.0 [00:00<00:00, 52.1kB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "vocab.txt:   0%|          | 0.00/232k [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "vocab.txt: 100%|██████████| 232k/232k [00:00<00:00, 43.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "config.json:   0%|          | 0.00/998 [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "config.json: 100%|██████████| 998/998 [00:00<00:00, 1.14MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   0%|          | 0.00/1.33G [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   1%|          | 10.5M/1.33G [00:00<01:11, 18.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   2%|▏         | 21.0M/1.33G [00:00<00:43, 30.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   2%|▏         | 31.5M/1.33G [00:01<00:37, 34.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   3%|▎         | 41.9M/1.33G [00:01<00:42, 30.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   4%|▍         | 52.4M/1.33G [00:01<00:33, 38.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   5%|▍         | 62.9M/1.33G [00:01<00:29, 43.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   6%|▌         | 73.4M/1.33G [00:01<00:27, 46.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   6%|▋         | 83.9M/1.33G [00:02<00:28, 44.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   7%|▋         | 94.4M/1.33G [00:02<00:27, 45.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   8%|▊         | 105M/1.33G [00:02<00:25, 47.8MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   9%|▊         | 115M/1.33G [00:02<00:31, 39.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:   9%|▉         | 126M/1.33G [00:03<00:29, 40.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  10%|█         | 136M/1.33G [00:03<00:26, 45.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  11%|█         | 147M/1.33G [00:03<00:27, 43.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  12%|█▏        | 157M/1.33G [00:03<00:26, 44.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  13%|█▎        | 168M/1.33G [00:04<00:26, 44.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  13%|█▎        | 178M/1.33G [00:04<00:23, 49.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  14%|█▍        | 189M/1.33G [00:04<00:22, 51.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  15%|█▍        | 199M/1.33G [00:04<00:23, 48.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  16%|█▌        | 210M/1.33G [00:04<00:22, 49.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  17%|█▋        | 220M/1.33G [00:05<00:21, 51.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  17%|█▋        | 231M/1.33G [00:05<00:24, 44.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  18%|█▊        | 241M/1.33G [00:05<00:26, 41.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  19%|█▉        | 252M/1.33G [00:05<00:27, 39.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  20%|█▉        | 262M/1.33G [00:06<00:24, 43.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  20%|██        | 273M/1.33G [00:06<00:22, 47.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  21%|██        | 283M/1.33G [00:06<00:21, 49.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  22%|██▏       | 294M/1.33G [00:06<00:24, 43.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  23%|██▎       | 304M/1.33G [00:07<00:25, 40.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  24%|██▎       | 315M/1.33G [00:07<00:24, 42.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  24%|██▍       | 325M/1.33G [00:07<00:23, 42.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  25%|██▌       | 336M/1.33G [00:07<00:23, 42.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  26%|██▌       | 346M/1.33G [00:08<00:23, 42.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  27%|██▋       | 357M/1.33G [00:08<00:23, 41.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  28%|██▊       | 367M/1.33G [00:08<00:21, 45.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  28%|██▊       | 377M/1.33G [00:08<00:26, 36.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  29%|██▉       | 388M/1.33G [00:09<00:22, 41.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  30%|██▉       | 398M/1.33G [00:09<00:21, 43.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  31%|███       | 409M/1.33G [00:09<00:22, 40.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  31%|███▏      | 419M/1.33G [00:09<00:24, 37.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  32%|███▏      | 430M/1.33G [00:10<00:21, 41.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  33%|███▎      | 440M/1.33G [00:10<00:21, 41.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  34%|███▍      | 451M/1.33G [00:10<00:18, 48.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  35%|███▍      | 461M/1.33G [00:10<00:20, 42.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  35%|███▌      | 472M/1.33G [00:11<00:19, 44.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  36%|███▌      | 482M/1.33G [00:11<00:27, 30.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  37%|███▋      | 493M/1.33G [00:11<00:23, 35.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  38%|███▊      | 503M/1.33G [00:12<00:25, 32.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  39%|███▊      | 514M/1.33G [00:12<00:28, 29.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  39%|███▉      | 524M/1.33G [00:12<00:24, 33.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  40%|████      | 535M/1.33G [00:13<00:20, 38.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  41%|████      | 545M/1.33G [00:13<00:19, 41.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  42%|████▏     | 556M/1.33G [00:13<00:17, 45.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  42%|████▏     | 566M/1.33G [00:13<00:15, 50.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  43%|████▎     | 577M/1.33G [00:13<00:14, 53.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  44%|████▍     | 587M/1.33G [00:14<00:16, 44.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  45%|████▍     | 598M/1.33G [00:14<00:18, 40.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  46%|████▌     | 608M/1.33G [00:14<00:17, 41.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  46%|████▋     | 619M/1.33G [00:14<00:16, 42.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  47%|████▋     | 629M/1.33G [00:15<00:16, 42.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  48%|████▊     | 640M/1.33G [00:15<00:16, 42.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  49%|████▊     | 650M/1.33G [00:15<00:14, 45.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  50%|████▉     | 661M/1.33G [00:15<00:13, 50.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  50%|█████     | 671M/1.33G [00:16<00:15, 42.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  51%|█████     | 682M/1.33G [00:16<00:13, 48.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  52%|█████▏    | 692M/1.33G [00:16<00:15, 41.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  53%|█████▎    | 703M/1.33G [00:16<00:15, 41.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  53%|█████▎    | 713M/1.33G [00:17<00:15, 40.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  54%|█████▍    | 724M/1.33G [00:17<00:13, 44.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  55%|█████▌    | 734M/1.33G [00:17<00:12, 47.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  56%|█████▌    | 744M/1.33G [00:17<00:13, 44.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  57%|█████▋    | 755M/1.33G [00:18<00:14, 39.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  57%|█████▋    | 765M/1.33G [00:18<00:13, 43.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  58%|█████▊    | 776M/1.33G [00:18<00:12, 43.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  59%|█████▉    | 786M/1.33G [00:18<00:11, 49.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  60%|█████▉    | 797M/1.33G [00:18<00:10, 51.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  61%|██████    | 807M/1.33G [00:18<00:09, 55.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  61%|██████▏   | 818M/1.33G [00:19<00:09, 55.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  62%|██████▏   | 828M/1.33G [00:19<00:10, 46.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  63%|██████▎   | 839M/1.33G [00:19<00:12, 41.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  64%|██████▎   | 849M/1.33G [00:20<00:12, 38.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  64%|██████▍   | 860M/1.33G [00:20<00:11, 41.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  65%|██████▌   | 870M/1.33G [00:20<00:10, 42.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  66%|██████▌   | 881M/1.33G [00:20<00:09, 46.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  67%|██████▋   | 891M/1.33G [00:20<00:08, 51.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  68%|██████▊   | 902M/1.33G [00:21<00:08, 53.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  68%|██████▊   | 912M/1.33G [00:21<00:07, 53.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  69%|██████▉   | 923M/1.33G [00:21<00:08, 48.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  70%|██████▉   | 933M/1.33G [00:21<00:10, 39.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  71%|███████   | 944M/1.33G [00:22<00:09, 42.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  72%|███████▏  | 954M/1.33G [00:22<00:08, 45.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  72%|███████▏  | 965M/1.33G [00:22<00:09, 37.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  73%|███████▎  | 975M/1.33G [00:22<00:08, 40.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  74%|███████▍  | 986M/1.33G [00:23<00:07, 43.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  75%|███████▍  | 996M/1.33G [00:23<00:07, 46.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  75%|███████▌  | 1.01G/1.33G [00:23<00:08, 39.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  76%|███████▌  | 1.02G/1.33G [00:23<00:07, 45.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  77%|███████▋  | 1.03G/1.33G [00:24<00:07, 41.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  78%|███████▊  | 1.04G/1.33G [00:24<00:07, 41.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  79%|███████▊  | 1.05G/1.33G [00:24<00:07, 36.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  79%|███████▉  | 1.06G/1.33G [00:24<00:06, 42.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  80%|████████  | 1.07G/1.33G [00:25<00:05, 46.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  81%|████████  | 1.08G/1.33G [00:25<00:05, 47.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  82%|████████▏ | 1.09G/1.33G [00:25<00:06, 36.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  83%|████████▎ | 1.10G/1.33G [00:25<00:05, 39.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  83%|████████▎ | 1.11G/1.33G [00:26<00:05, 41.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  84%|████████▍ | 1.12G/1.33G [00:26<00:04, 48.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  85%|████████▍ | 1.13G/1.33G [00:26<00:05, 39.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  86%|████████▌ | 1.14G/1.33G [00:26<00:04, 38.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  86%|████████▋ | 1.15G/1.33G [00:27<00:04, 40.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  87%|████████▋ | 1.16G/1.33G [00:27<00:04, 39.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  88%|████████▊ | 1.17G/1.33G [00:27<00:03, 42.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  89%|████████▉ | 1.18G/1.33G [00:27<00:03, 46.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  90%|████████▉ | 1.20G/1.33G [00:28<00:03, 46.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  90%|█████████ | 1.21G/1.33G [00:28<00:02, 46.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  91%|█████████ | 1.22G/1.33G [00:28<00:03, 39.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  92%|█████████▏| 1.23G/1.33G [00:28<00:02, 37.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  93%|█████████▎| 1.24G/1.33G [00:29<00:02, 37.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  94%|█████████▎| 1.25G/1.33G [00:29<00:02, 41.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  94%|█████████▍| 1.26G/1.33G [00:29<00:02, 34.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  95%|█████████▌| 1.27G/1.33G [00:30<00:01, 41.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  96%|█████████▌| 1.28G/1.33G [00:30<00:01, 38.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  97%|█████████▋| 1.29G/1.33G [00:30<00:01, 38.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  97%|█████████▋| 1.30G/1.33G [00:30<00:00, 40.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  98%|█████████▊| 1.31G/1.33G [00:30<00:00, 45.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin:  99%|█████████▉| 1.32G/1.33G [00:31<00:00, 40.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin: 100%|█████████▉| 1.33G/1.33G [00:31<00:00, 47.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "pytorch_model.bin: 100%|██████████| 1.33G/1.33G [00:31<00:00, 42.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "tokenizer_config.json:   0%|          | 0.00/60.0 [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "tokenizer_config.json: 100%|██████████| 60.0/60.0 [00:00<00:00, 66.8kB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "vocab.txt:   0%|          | 0.00/213k [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "vocab.txt: 100%|██████████| 213k/213k [00:00<00:00, 44.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 17%|█▋        | 1/6 [01:05<05:26, 65.29s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 33%|███▎      | 2/6 [01:14<02:09, 32.49s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 50%|█████     | 3/6 [01:24<01:06, 22.03s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 67%|██████▋   | 4/6 [01:35<00:35, 17.55s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 83%|████████▎ | 5/6 [01:44<00:14, 14.69s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 6/6 [01:53<00:00, 12.85s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 6/6 [01:53<00:00, 19.00s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "for num, key in tqdm(enumerate(image_dict.keys()), total=len(image_dict)):  # loop through all images\n",
    "    image_dict[key] = ammico.TextDetector(image_dict[key], analyse_text=True).analyse_image() # analyse image with EmotionDetector and update dict\n",
    "    \n",
    "    if num % dump_every == 0 | num == len(image_dict) - 1:     # save results every dump_every to dump_file\n",
    "        image_df = ammico.get_dataframe(image_dict)\n",
    "        image_df.to_csv(dump_file)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For the computationally demanding `SummaryDetector`, it is best to initialize the model first and then analyze each image while passing the model explicitly. This can be done in a separate loop or in the same loop as for text and emotion detection."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:45:58.645100Z",
     "iopub.status.busy": "2024-02-19T08:45:58.644666Z",
     "iopub.status.idle": "2024-02-19T08:47:40.648780Z",
     "shell.execute_reply": "2024-02-19T08:47:40.648018Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "vocab.txt:   0%|          | 0.00/232k [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "vocab.txt: 100%|██████████| 232k/232k [00:00<00:00, 1.82MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "vocab.txt: 100%|██████████| 232k/232k [00:00<00:00, 1.80MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "tokenizer_config.json:   0%|          | 0.00/28.0 [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "tokenizer_config.json: 100%|██████████| 28.0/28.0 [00:00<00:00, 39.1kB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "config.json:   0%|          | 0.00/570 [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "config.json: 100%|██████████| 570/570 [00:00<00:00, 909kB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0.00/2.50G [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 4.01M/2.50G [00:00<01:35, 28.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 11.6M/2.50G [00:00<00:52, 51.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  1%|          | 16.8M/2.50G [00:00<00:56, 47.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  1%|          | 24.0M/2.50G [00:00<00:56, 46.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  1%|          | 32.0M/2.50G [00:00<00:53, 49.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  2%|▏         | 40.0M/2.50G [00:00<00:51, 51.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  2%|▏         | 48.0M/2.50G [00:01<00:50, 52.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  2%|▏         | 56.0M/2.50G [00:01<00:48, 53.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  2%|▏         | 64.0M/2.50G [00:01<00:47, 55.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  3%|▎         | 72.0M/2.50G [00:01<00:58, 44.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  3%|▎         | 80.0M/2.50G [00:01<00:55, 47.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  3%|▎         | 88.0M/2.50G [00:01<00:52, 49.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  4%|▎         | 96.0M/2.50G [00:02<00:49, 51.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  4%|▍         | 104M/2.50G [00:02<00:44, 57.5MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  4%|▍         | 112M/2.50G [00:02<01:12, 35.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  5%|▍         | 128M/2.50G [00:02<00:56, 45.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  6%|▌         | 144M/2.50G [00:02<00:42, 59.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  6%|▌         | 160M/2.50G [00:03<00:36, 69.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  7%|▋         | 176M/2.50G [00:03<00:29, 84.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  7%|▋         | 192M/2.50G [00:03<00:24, 100MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  8%|▊         | 203M/2.50G [00:03<00:33, 74.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  8%|▊         | 216M/2.50G [00:03<00:34, 72.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  9%|▉         | 232M/2.50G [00:04<00:31, 78.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 10%|▉         | 248M/2.50G [00:04<00:28, 84.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 11%|█         | 272M/2.50G [00:04<00:22, 106MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 11%|█         | 288M/2.50G [00:04<00:20, 114MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 12%|█▏        | 304M/2.50G [00:04<00:22, 106MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 13%|█▎        | 328M/2.50G [00:04<00:18, 125MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 13%|█▎        | 344M/2.50G [00:04<00:19, 122MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 14%|█▍        | 365M/2.50G [00:05<00:16, 144MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 15%|█▍        | 384M/2.50G [00:05<00:16, 139MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 16%|█▌        | 408M/2.50G [00:05<00:14, 157MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 17%|█▋        | 424M/2.50G [00:05<00:17, 128MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 17%|█▋        | 437M/2.50G [00:05<00:19, 112MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 18%|█▊        | 457M/2.50G [00:05<00:16, 130MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 19%|█▊        | 475M/2.50G [00:05<00:15, 145MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 19%|█▉        | 491M/2.50G [00:06<00:15, 142MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 20%|█▉        | 505M/2.50G [00:06<00:15, 141MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 20%|██        | 519M/2.50G [00:06<00:15, 140MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 21%|██        | 544M/2.50G [00:06<00:13, 153MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 22%|██▏       | 570M/2.50G [00:06<00:11, 185MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 23%|██▎       | 589M/2.50G [00:06<00:12, 165MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 24%|██▎       | 606M/2.50G [00:06<00:12, 170MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 24%|██▍       | 623M/2.50G [00:06<00:14, 139MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 25%|██▌       | 647M/2.50G [00:07<00:12, 163MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 26%|██▌       | 664M/2.50G [00:07<00:19, 101MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 27%|██▋       | 687M/2.50G [00:07<00:15, 127MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 27%|██▋       | 704M/2.50G [00:07<00:19, 102MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 28%|██▊       | 728M/2.50G [00:07<00:15, 121MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 29%|██▉       | 746M/2.50G [00:07<00:14, 134MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 30%|██▉       | 762M/2.50G [00:08<00:14, 129MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 31%|███       | 784M/2.50G [00:08<00:12, 152MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 31%|███       | 801M/2.50G [00:08<00:13, 139MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 32%|███▏      | 826M/2.50G [00:08<00:10, 167MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 33%|███▎      | 843M/2.50G [00:08<00:14, 122MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 34%|███▎      | 864M/2.50G [00:08<00:15, 118MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 35%|███▍      | 889M/2.50G [00:09<00:11, 147MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 35%|███▌      | 906M/2.50G [00:09<00:12, 136MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 36%|███▌      | 921M/2.50G [00:09<00:12, 134MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 37%|███▋      | 944M/2.50G [00:09<00:10, 157MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 37%|███▋      | 961M/2.50G [00:09<00:10, 159MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 38%|███▊      | 977M/2.50G [00:09<00:12, 136MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 39%|███▉      | 0.98G/2.50G [00:09<00:09, 166MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 40%|███▉      | 1.00G/2.50G [00:09<00:10, 154MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 41%|████      | 1.02G/2.50G [00:10<00:09, 171MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 41%|████▏     | 1.03G/2.50G [00:10<00:11, 142MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 42%|████▏     | 1.05G/2.50G [00:10<00:12, 124MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 43%|████▎     | 1.07G/2.50G [00:10<00:12, 122MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 44%|████▎     | 1.09G/2.50G [00:10<00:11, 127MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 44%|████▍     | 1.11G/2.50G [00:10<00:11, 126MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 45%|████▌     | 1.13G/2.50G [00:10<00:09, 157MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 46%|████▌     | 1.15G/2.50G [00:11<00:10, 146MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 46%|████▋     | 1.16G/2.50G [00:11<00:10, 134MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 47%|████▋     | 1.18G/2.50G [00:11<00:10, 140MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 48%|████▊     | 1.20G/2.50G [00:11<00:09, 151MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 49%|████▊     | 1.22G/2.50G [00:11<00:09, 149MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 50%|████▉     | 1.24G/2.50G [00:11<00:07, 177MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 50%|█████     | 1.26G/2.50G [00:12<00:10, 132MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 51%|█████     | 1.28G/2.50G [00:12<00:08, 148MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 52%|█████▏    | 1.30G/2.50G [00:12<00:11, 109MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 53%|█████▎    | 1.32G/2.50G [00:12<00:09, 137MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 54%|█████▎    | 1.34G/2.50G [00:12<00:07, 163MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 54%|█████▍    | 1.36G/2.50G [00:12<00:10, 112MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 55%|█████▌    | 1.39G/2.50G [00:13<00:08, 143MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 56%|█████▌    | 1.41G/2.50G [00:13<00:08, 136MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 57%|█████▋    | 1.42G/2.50G [00:13<00:07, 147MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 58%|█████▊    | 1.44G/2.50G [00:13<00:08, 128MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 58%|█████▊    | 1.45G/2.50G [00:13<00:08, 131MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 59%|█████▉    | 1.48G/2.50G [00:13<00:07, 150MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 60%|█████▉    | 1.50G/2.50G [00:13<00:06, 173MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 61%|██████    | 1.52G/2.50G [00:13<00:06, 172MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 62%|██████▏   | 1.54G/2.50G [00:13<00:05, 193MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 62%|██████▏   | 1.56G/2.50G [00:14<00:06, 161MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 63%|██████▎   | 1.58G/2.50G [00:14<00:06, 162MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 64%|██████▎   | 1.60G/2.50G [00:14<00:06, 144MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 64%|██████▍   | 1.61G/2.50G [00:14<00:07, 123MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 65%|██████▍   | 1.63G/2.50G [00:14<00:06, 135MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 66%|██████▌   | 1.64G/2.50G [00:14<00:07, 126MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 66%|██████▌   | 1.65G/2.50G [00:15<00:07, 116MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 67%|██████▋   | 1.67G/2.50G [00:15<00:06, 129MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 67%|██████▋   | 1.69G/2.50G [00:15<00:06, 144MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 68%|██████▊   | 1.71G/2.50G [00:15<00:05, 156MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 69%|██████▉   | 1.72G/2.50G [00:15<00:11, 73.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 70%|██████▉   | 1.75G/2.50G [00:16<00:10, 80.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 71%|███████   | 1.78G/2.50G [00:16<00:07, 109MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 72%|███████▏  | 1.79G/2.50G [00:16<00:07, 108MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 72%|███████▏  | 1.81G/2.50G [00:16<00:07, 104MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 73%|███████▎  | 1.82G/2.50G [00:16<00:06, 107MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 74%|███████▎  | 1.84G/2.50G [00:16<00:05, 129MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 74%|███████▍  | 1.86G/2.50G [00:16<00:05, 124MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 75%|███████▌  | 1.88G/2.50G [00:17<00:04, 155MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 76%|███████▌  | 1.90G/2.50G [00:17<00:04, 132MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 76%|███████▋  | 1.91G/2.50G [00:17<00:04, 133MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 77%|███████▋  | 1.93G/2.50G [00:18<00:10, 61.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 78%|███████▊  | 1.95G/2.50G [00:18<00:08, 69.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 78%|███████▊  | 1.96G/2.50G [00:18<00:08, 67.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 79%|███████▉  | 1.98G/2.50G [00:18<00:06, 87.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 80%|███████▉  | 1.99G/2.50G [00:18<00:05, 100MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 80%|████████  | 2.01G/2.50G [00:18<00:04, 124MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 81%|████████  | 2.03G/2.50G [00:18<00:04, 103MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 81%|████████▏ | 2.04G/2.50G [00:19<00:04, 107MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 82%|████████▏ | 2.05G/2.50G [00:19<00:04, 103MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 83%|████████▎ | 2.07G/2.50G [00:19<00:04, 113MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 83%|████████▎ | 2.09G/2.50G [00:19<00:03, 130MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 84%|████████▍ | 2.10G/2.50G [00:19<00:03, 128MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 85%|████████▌ | 2.13G/2.50G [00:19<00:02, 169MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 86%|████████▌ | 2.15G/2.50G [00:19<00:02, 181MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 87%|████████▋ | 2.17G/2.50G [00:20<00:06, 58.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 87%|████████▋ | 2.18G/2.50G [00:20<00:04, 71.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 88%|████████▊ | 2.20G/2.50G [00:21<00:05, 54.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 88%|████████▊ | 2.21G/2.50G [00:21<00:04, 66.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 89%|████████▉ | 2.23G/2.50G [00:21<00:03, 80.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 90%|████████▉ | 2.24G/2.50G [00:21<00:03, 87.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 90%|█████████ | 2.26G/2.50G [00:21<00:03, 75.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 91%|█████████ | 2.28G/2.50G [00:22<00:02, 82.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 92%|█████████▏| 2.30G/2.50G [00:22<00:02, 98.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 93%|█████████▎| 2.33G/2.50G [00:22<00:01, 104MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 93%|█████████▎| 2.34G/2.50G [00:22<00:01, 103MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 94%|█████████▍| 2.35G/2.50G [00:22<00:01, 104MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 95%|█████████▍| 2.38G/2.50G [00:22<00:01, 116MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 96%|█████████▌| 2.40G/2.50G [00:23<00:00, 142MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 96%|█████████▋| 2.41G/2.50G [00:23<00:00, 118MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 97%|█████████▋| 2.43G/2.50G [00:23<00:00, 123MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 98%|█████████▊| 2.44G/2.50G [00:23<00:00, 101MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 98%|█████████▊| 2.46G/2.50G [00:23<00:00, 118MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 99%|█████████▉| 2.47G/2.50G [00:23<00:00, 111MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|█████████▉| 2.49G/2.50G [00:24<00:00, 121MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 2.50G/2.50G [00:24<00:00, 112MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0/6 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 17%|█▋        | 1/6 [00:12<01:02, 12.47s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 33%|███▎      | 2/6 [00:23<00:46, 11.72s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 50%|█████     | 3/6 [00:35<00:34, 11.65s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 67%|██████▋   | 4/6 [00:47<00:23, 11.81s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 83%|████████▎ | 5/6 [00:58<00:11, 11.56s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 6/6 [01:08<00:00, 11.15s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 6/6 [01:08<00:00, 11.46s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "# clear memory on cuda first? Faces seems to always not release GPU\n",
    "# initialize the models\n",
    "image_summary_detector = ammico.SummaryDetector(subdict = image_dict, analysis_type=\"summary\", model_type=\"base\")\n",
    "\n",
    "# run the analysis without having to re-iniatialize the model\n",
    "for num, key in tqdm(enumerate(image_dict.keys()),total=len(image_dict)):   # loop through all images\n",
    "    image_dict[key] = image_summary_detector.analyse_image(subdict = image_dict[key], analysis_type=\"summary\")  # analyse image with SummaryDetector and update dict\n",
    "    \n",
    "    if num % dump_every == 0 | num == len(image_dict) - 1:    # save results every dump_every to dump_file\n",
    "        image_df = ammico.get_dataframe(image_dict)\n",
    "        image_df.to_csv(dump_file)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Or you can run all Detectors in one loop as for example:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:47:40.651533Z",
     "iopub.status.busy": "2024-02-19T08:47:40.651128Z",
     "iopub.status.idle": "2024-02-19T08:50:22.304514Z",
     "shell.execute_reply": "2024-02-19T08:50:22.303793Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0/6 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 17%|█▋        | 1/6 [00:24<02:01, 24.22s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 33%|███▎      | 2/6 [00:48<01:36, 24.10s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 50%|█████     | 3/6 [01:17<01:19, 26.50s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 67%|██████▋   | 4/6 [01:41<00:51, 25.50s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 83%|████████▎ | 5/6 [02:04<00:24, 24.41s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "1/1 [==============================] - ETA: 0s"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r",
      "1/1 [==============================] - 1s 556ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "1/1 [==============================] - ETA: 0s"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r",
      "1/1 [==============================] - 0s 221ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "1/1 [==============================] - ETA: 0s"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r",
      "1/1 [==============================] - 0s 215ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "1/1 [==============================] - ETA: 0s"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r",
      "1/1 [==============================] - 0s 219ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "1/1 [==============================] - ETA: 0s"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r",
      "1/1 [==============================] - 0s 14ms/step\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 6/6 [02:33<00:00, 26.04s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 6/6 [02:33<00:00, 25.54s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "# initialize the models\n",
    "image_summary_detector = ammico.SummaryDetector(subdict = image_dict, analysis_type=\"summary\", model_type=\"base\")\n",
    "\n",
    "for num, key in tqdm(enumerate(image_dict.keys()),total=len(image_dict)):   # loop through all images\n",
    "    image_dict[key] = ammico.EmotionDetector(image_dict[key]).analyse_image()   # analyse image with EmotionDetector and update dict\n",
    "    image_dict[key] = ammico.TextDetector(image_dict[key], analyse_text=True).analyse_image()   # analyse image with TextDetector and update dict\n",
    "    image_dict[key] = image_summary_detector.analyse_image(subdict = image_dict[key], analysis_type=\"summary\")  # analyse image with SummaryDetector and update dict\n",
    "    \n",
    "    if num % dump_every == 0 | num == len(image_dict) - 1:  # save results every dump_every to dump_file \n",
    "        image_df = ammico.get_dataframe(image_dict)\n",
    "        image_df.to_csv(dump_file)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The nested dictionary will be updated from containing only the file id's and paths to the image files, to containing all calculated image features."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Step 4: Convert analysis output to pandas dataframe and write csv\n",
    "The content of the nested dictionary can then conveniently be converted into a pandas dataframe for further analysis in Python, or be written as a csv file:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:50:22.307625Z",
     "iopub.status.busy": "2024-02-19T08:50:22.307231Z",
     "iopub.status.idle": "2024-02-19T08:50:22.310969Z",
     "shell.execute_reply": "2024-02-19T08:50:22.310432Z"
    }
   },
   "outputs": [],
   "source": [
    "image_df = ammico.get_dataframe(image_dict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Inspect the dataframe:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:50:22.313326Z",
     "iopub.status.busy": "2024-02-19T08:50:22.312953Z",
     "iopub.status.idle": "2024-02-19T08:50:22.351063Z",
     "shell.execute_reply": "2024-02-19T08:50:22.350534Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>filename</th>\n",
       "      <th>face</th>\n",
       "      <th>multiple_faces</th>\n",
       "      <th>no_faces</th>\n",
       "      <th>wears_mask</th>\n",
       "      <th>age</th>\n",
       "      <th>gender</th>\n",
       "      <th>race</th>\n",
       "      <th>emotion</th>\n",
       "      <th>emotion (category)</th>\n",
       "      <th>...</th>\n",
       "      <th>text_language</th>\n",
       "      <th>text_english</th>\n",
       "      <th>text_clean</th>\n",
       "      <th>text_summary</th>\n",
       "      <th>sentiment</th>\n",
       "      <th>sentiment_score</th>\n",
       "      <th>entity</th>\n",
       "      <th>entity_type</th>\n",
       "      <th>const_image_summary</th>\n",
       "      <th>3_non-deterministic_summary</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>data-test/img4.png</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>[No]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>...</td>\n",
       "      <td>en</td>\n",
       "      <td>MOODOVIN XI</td>\n",
       "      <td>XI</td>\n",
       "      <td>MOODOVIN XI XI: Vladimir Putin, Vladimir Vlad...</td>\n",
       "      <td>POSITIVE</td>\n",
       "      <td>0.66</td>\n",
       "      <td>[MOODOVIN XI]</td>\n",
       "      <td>[ORG]</td>\n",
       "      <td>a river running through a city next to tall bu...</td>\n",
       "      <td>[buildings near a waterway with small boats pa...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>data-test/img1.png</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>[No]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>...</td>\n",
       "      <td>en</td>\n",
       "      <td>SCATTERING THEORY The Quantum Theory of Nonrel...</td>\n",
       "      <td>THEORY The Quantum Theory of Collisions JOHN R...</td>\n",
       "      <td>SCATTERING THEORY The Quantum Theory of Nonre...</td>\n",
       "      <td>POSITIVE</td>\n",
       "      <td>0.91</td>\n",
       "      <td>[Non, ##vist, Col, ##N, R, T, ##AYL, Universit...</td>\n",
       "      <td>[MISC, MISC, MISC, ORG, PER, PER, ORG, ORG]</td>\n",
       "      <td>a close up of a piece of paper with writing on it</td>\n",
       "      <td>[a white paper with some black writing on it, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>data-test/img2.png</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>[No]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>...</td>\n",
       "      <td>en</td>\n",
       "      <td>THE ALGEBRAIC EIGENVALUE PROBLEM DOM NVS TIO M...</td>\n",
       "      <td>THE PROBLEM DOM NVS TIO MINA Monographs on Num...</td>\n",
       "      <td>H. H. W. WILKINSON: The Algebri</td>\n",
       "      <td>NEGATIVE</td>\n",
       "      <td>0.97</td>\n",
       "      <td>[ALGEBRAIC EIGENVAL, NVS TIO MI, J, H, WILKINSON]</td>\n",
       "      <td>[MISC, ORG, ORG, ORG, ORG]</td>\n",
       "      <td>a yellow book with green lettering on it</td>\n",
       "      <td>[an old book with a picture of the slogan of t...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             filename face multiple_faces  no_faces wears_mask     age  \\\n",
       "0  data-test/img4.png   No             No         0       [No]  [None]   \n",
       "1  data-test/img1.png   No             No         0       [No]  [None]   \n",
       "2  data-test/img2.png   No             No         0       [No]  [None]   \n",
       "\n",
       "   gender    race emotion emotion (category)  ... text_language  \\\n",
       "0  [None]  [None]  [None]             [None]  ...            en   \n",
       "1  [None]  [None]  [None]             [None]  ...            en   \n",
       "2  [None]  [None]  [None]             [None]  ...            en   \n",
       "\n",
       "                                        text_english  \\\n",
       "0                                        MOODOVIN XI   \n",
       "1  SCATTERING THEORY The Quantum Theory of Nonrel...   \n",
       "2  THE ALGEBRAIC EIGENVALUE PROBLEM DOM NVS TIO M...   \n",
       "\n",
       "                                          text_clean  \\\n",
       "0                                                 XI   \n",
       "1  THEORY The Quantum Theory of Collisions JOHN R...   \n",
       "2  THE PROBLEM DOM NVS TIO MINA Monographs on Num...   \n",
       "\n",
       "                                        text_summary sentiment  \\\n",
       "0   MOODOVIN XI XI: Vladimir Putin, Vladimir Vlad...  POSITIVE   \n",
       "1   SCATTERING THEORY The Quantum Theory of Nonre...  POSITIVE   \n",
       "2                    H. H. W. WILKINSON: The Algebri  NEGATIVE   \n",
       "\n",
       "  sentiment_score                                             entity  \\\n",
       "0            0.66                                      [MOODOVIN XI]   \n",
       "1            0.91  [Non, ##vist, Col, ##N, R, T, ##AYL, Universit...   \n",
       "2            0.97  [ALGEBRAIC EIGENVAL, NVS TIO MI, J, H, WILKINSON]   \n",
       "\n",
       "                                   entity_type  \\\n",
       "0                                        [ORG]   \n",
       "1  [MISC, MISC, MISC, ORG, PER, PER, ORG, ORG]   \n",
       "2                   [MISC, ORG, ORG, ORG, ORG]   \n",
       "\n",
       "                                 const_image_summary  \\\n",
       "0  a river running through a city next to tall bu...   \n",
       "1  a close up of a piece of paper with writing on it   \n",
       "2           a yellow book with green lettering on it   \n",
       "\n",
       "                         3_non-deterministic_summary  \n",
       "0  [buildings near a waterway with small boats pa...  \n",
       "1  [a white paper with some black writing on it, ...  \n",
       "2  [an old book with a picture of the slogan of t...  \n",
       "\n",
       "[3 rows x 21 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "image_df.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Or write to a csv file:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:50:22.391547Z",
     "iopub.status.busy": "2024-02-19T08:50:22.391161Z",
     "iopub.status.idle": "2024-02-19T08:50:25.022235Z",
     "shell.execute_reply": "2024-02-19T08:50:25.021403Z"
    }
   },
   "outputs": [
    {
     "ename": "OSError",
     "evalue": "Cannot save file into a non-existent directory: '/content/drive/MyDrive/misinformation-data'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mOSError\u001b[0m                                   Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[15], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m \u001b[43mimage_df\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mto_csv\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43m/content/drive/MyDrive/misinformation-data/data_out.csv\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/pandas/util/_decorators.py:333\u001b[0m, in \u001b[0;36mdeprecate_nonkeyword_arguments.<locals>.decorate.<locals>.wrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m    327\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mlen\u001b[39m(args) \u001b[38;5;241m>\u001b[39m num_allow_args:\n\u001b[1;32m    328\u001b[0m     warnings\u001b[38;5;241m.\u001b[39mwarn(\n\u001b[1;32m    329\u001b[0m         msg\u001b[38;5;241m.\u001b[39mformat(arguments\u001b[38;5;241m=\u001b[39m_format_argument_list(allow_args)),\n\u001b[1;32m    330\u001b[0m         \u001b[38;5;167;01mFutureWarning\u001b[39;00m,\n\u001b[1;32m    331\u001b[0m         stacklevel\u001b[38;5;241m=\u001b[39mfind_stack_level(),\n\u001b[1;32m    332\u001b[0m     )\n\u001b[0;32m--> 333\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfunc\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/pandas/core/generic.py:3961\u001b[0m, in \u001b[0;36mNDFrame.to_csv\u001b[0;34m(self, path_or_buf, sep, na_rep, float_format, columns, header, index, index_label, mode, encoding, compression, quoting, quotechar, lineterminator, chunksize, date_format, doublequote, escapechar, decimal, errors, storage_options)\u001b[0m\n\u001b[1;32m   3950\u001b[0m df \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28misinstance\u001b[39m(\u001b[38;5;28mself\u001b[39m, ABCDataFrame) \u001b[38;5;28;01melse\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mto_frame()\n\u001b[1;32m   3952\u001b[0m formatter \u001b[38;5;241m=\u001b[39m DataFrameFormatter(\n\u001b[1;32m   3953\u001b[0m     frame\u001b[38;5;241m=\u001b[39mdf,\n\u001b[1;32m   3954\u001b[0m     header\u001b[38;5;241m=\u001b[39mheader,\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m   3958\u001b[0m     decimal\u001b[38;5;241m=\u001b[39mdecimal,\n\u001b[1;32m   3959\u001b[0m )\n\u001b[0;32m-> 3961\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mDataFrameRenderer\u001b[49m\u001b[43m(\u001b[49m\u001b[43mformatter\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mto_csv\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m   3962\u001b[0m \u001b[43m    \u001b[49m\u001b[43mpath_or_buf\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3963\u001b[0m \u001b[43m    \u001b[49m\u001b[43mlineterminator\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mlineterminator\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3964\u001b[0m \u001b[43m    \u001b[49m\u001b[43msep\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43msep\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3965\u001b[0m \u001b[43m    \u001b[49m\u001b[43mencoding\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mencoding\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3966\u001b[0m \u001b[43m    \u001b[49m\u001b[43merrors\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43merrors\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3967\u001b[0m \u001b[43m    \u001b[49m\u001b[43mcompression\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcompression\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3968\u001b[0m \u001b[43m    \u001b[49m\u001b[43mquoting\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mquoting\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3969\u001b[0m \u001b[43m    \u001b[49m\u001b[43mcolumns\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcolumns\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3970\u001b[0m \u001b[43m    \u001b[49m\u001b[43mindex_label\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mindex_label\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3971\u001b[0m \u001b[43m    \u001b[49m\u001b[43mmode\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mmode\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3972\u001b[0m \u001b[43m    \u001b[49m\u001b[43mchunksize\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mchunksize\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3973\u001b[0m \u001b[43m    \u001b[49m\u001b[43mquotechar\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mquotechar\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3974\u001b[0m \u001b[43m    \u001b[49m\u001b[43mdate_format\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mdate_format\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3975\u001b[0m \u001b[43m    \u001b[49m\u001b[43mdoublequote\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mdoublequote\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3976\u001b[0m \u001b[43m    \u001b[49m\u001b[43mescapechar\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mescapechar\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3977\u001b[0m \u001b[43m    \u001b[49m\u001b[43mstorage_options\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mstorage_options\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3978\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/pandas/io/formats/format.py:1014\u001b[0m, in \u001b[0;36mDataFrameRenderer.to_csv\u001b[0;34m(self, path_or_buf, encoding, sep, columns, index_label, mode, compression, quoting, quotechar, lineterminator, chunksize, date_format, doublequote, escapechar, errors, storage_options)\u001b[0m\n\u001b[1;32m    993\u001b[0m     created_buffer \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mFalse\u001b[39;00m\n\u001b[1;32m    995\u001b[0m csv_formatter \u001b[38;5;241m=\u001b[39m CSVFormatter(\n\u001b[1;32m    996\u001b[0m     path_or_buf\u001b[38;5;241m=\u001b[39mpath_or_buf,\n\u001b[1;32m    997\u001b[0m     lineterminator\u001b[38;5;241m=\u001b[39mlineterminator,\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m   1012\u001b[0m     formatter\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mfmt,\n\u001b[1;32m   1013\u001b[0m )\n\u001b[0;32m-> 1014\u001b[0m \u001b[43mcsv_formatter\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43msave\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1016\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m created_buffer:\n\u001b[1;32m   1017\u001b[0m     \u001b[38;5;28;01massert\u001b[39;00m \u001b[38;5;28misinstance\u001b[39m(path_or_buf, StringIO)\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/pandas/io/formats/csvs.py:251\u001b[0m, in \u001b[0;36mCSVFormatter.save\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    247\u001b[0m \u001b[38;5;250m\u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m    248\u001b[0m \u001b[38;5;124;03mCreate the writer & save.\u001b[39;00m\n\u001b[1;32m    249\u001b[0m \u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m    250\u001b[0m \u001b[38;5;66;03m# apply compression and byte/text conversion\u001b[39;00m\n\u001b[0;32m--> 251\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m \u001b[43mget_handle\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m    252\u001b[0m \u001b[43m    \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfilepath_or_buffer\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    253\u001b[0m \u001b[43m    \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mmode\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    254\u001b[0m \u001b[43m    \u001b[49m\u001b[43mencoding\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mencoding\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    255\u001b[0m \u001b[43m    \u001b[49m\u001b[43merrors\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43merrors\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    256\u001b[0m \u001b[43m    \u001b[49m\u001b[43mcompression\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcompression\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    257\u001b[0m \u001b[43m    \u001b[49m\u001b[43mstorage_options\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstorage_options\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    258\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m \u001b[38;5;28;01mas\u001b[39;00m handles:\n\u001b[1;32m    259\u001b[0m     \u001b[38;5;66;03m# Note: self.encoding is irrelevant here\u001b[39;00m\n\u001b[1;32m    260\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mwriter \u001b[38;5;241m=\u001b[39m csvlib\u001b[38;5;241m.\u001b[39mwriter(\n\u001b[1;32m    261\u001b[0m         handles\u001b[38;5;241m.\u001b[39mhandle,\n\u001b[1;32m    262\u001b[0m         lineterminator\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mlineterminator,\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    267\u001b[0m         quotechar\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mquotechar,\n\u001b[1;32m    268\u001b[0m     )\n\u001b[1;32m    270\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_save()\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/pandas/io/common.py:749\u001b[0m, in \u001b[0;36mget_handle\u001b[0;34m(path_or_buf, mode, encoding, compression, memory_map, is_text, errors, storage_options)\u001b[0m\n\u001b[1;32m    747\u001b[0m \u001b[38;5;66;03m# Only for write methods\u001b[39;00m\n\u001b[1;32m    748\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mr\u001b[39m\u001b[38;5;124m\"\u001b[39m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;129;01min\u001b[39;00m mode \u001b[38;5;129;01mand\u001b[39;00m is_path:\n\u001b[0;32m--> 749\u001b[0m     \u001b[43mcheck_parent_directory\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mstr\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43mhandle\u001b[49m\u001b[43m)\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    751\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m compression:\n\u001b[1;32m    752\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m compression \u001b[38;5;241m!=\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mzstd\u001b[39m\u001b[38;5;124m\"\u001b[39m:\n\u001b[1;32m    753\u001b[0m         \u001b[38;5;66;03m# compression libraries do not like an explicit text-mode\u001b[39;00m\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/pandas/io/common.py:616\u001b[0m, in \u001b[0;36mcheck_parent_directory\u001b[0;34m(path)\u001b[0m\n\u001b[1;32m    614\u001b[0m parent \u001b[38;5;241m=\u001b[39m Path(path)\u001b[38;5;241m.\u001b[39mparent\n\u001b[1;32m    615\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m parent\u001b[38;5;241m.\u001b[39mis_dir():\n\u001b[0;32m--> 616\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mOSError\u001b[39;00m(\u001b[38;5;124mrf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mCannot save file into a non-existent directory: \u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;132;01m{\u001b[39;00mparent\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n",
      "\u001b[0;31mOSError\u001b[0m: Cannot save file into a non-existent directory: '/content/drive/MyDrive/misinformation-data'"
     ]
    }
   ],
   "source": [
    "image_df.to_csv(\"/content/drive/MyDrive/misinformation-data/data_out.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# The detector modules\n",
    "The different detector modules with their options are explained in more detail in this section.\n",
    "## Text detector\n",
    "Text on the images can be extracted using the `TextDetector` class (`text` module). The text is initally extracted using the Google Cloud Vision API and then translated into English with googletrans. The translated text is cleaned of whitespace, linebreaks, and numbers using Python syntax and spaCy. \n",
    "\n",
    "<img src=\"../_static/text_detector.png\" width=\"800\" />\n",
    "\n",
    "The user can set if the text should be further summarized, and analyzed for sentiment and named entity recognition, by setting the keyword `analyse_text` to `True` (the default is `False`). If set, the transformers pipeline is used for each of these tasks, with the default models as of 03/2023. Other models can be selected by setting the optional keyword `model_names` to a list of selected models, on for each task: `model_names=[\"sshleifer/distilbart-cnn-12-6\", \"distilbert-base-uncased-finetuned-sst-2-english\", \"dbmdz/bert-large-cased-finetuned-conll03-english\"]` for summary, sentiment, and ner. To be even more specific, revision numbers can also be selected by specifying the optional keyword `revision_numbers` to a list of revision numbers for each model, for example `revision_numbers=[\"a4f8f3e\", \"af0f99b\", \"f2482bf\"]`. \n",
    "\n",
    "Please note that for the Google Cloud Vision API (the TextDetector class) you need to set a key in order to process the images. This key is ideally set as an environment variable using for example"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:50:25.032719Z",
     "iopub.status.busy": "2024-02-19T08:50:25.032386Z",
     "iopub.status.idle": "2024-02-19T08:50:25.035336Z",
     "shell.execute_reply": "2024-02-19T08:50:25.034770Z"
    }
   },
   "outputs": [],
   "source": [
    "# os.environ[\"GOOGLE_APPLICATION_CREDENTIALS\"] = \"/content/drive/MyDrive/misinformation-data/misinformation-campaign-981aa55a3b13.json\"\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "where you place the key on your Google Drive if running on colab, or place it in a local folder on your machine.\n",
    "\n",
    "Summarizing, the text detection is carried out using the following method call and keywords, where `analyse_text`, `model_names`, and `revision_numbers` are optional:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:50:25.037648Z",
     "iopub.status.busy": "2024-02-19T08:50:25.037344Z",
     "iopub.status.idle": "2024-02-19T08:51:21.184249Z",
     "shell.execute_reply": "2024-02-19T08:51:21.183549Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0/6 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 17%|█▋        | 1/6 [00:09<00:47,  9.51s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 33%|███▎      | 2/6 [00:18<00:36,  9.10s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 50%|█████     | 3/6 [00:27<00:27,  9.07s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 67%|██████▋   | 4/6 [00:37<00:18,  9.49s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 83%|████████▎ | 5/6 [00:46<00:09,  9.38s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 6/6 [00:56<00:00,  9.41s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 6/6 [00:56<00:00,  9.36s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "for num, key in tqdm(enumerate(image_dict.keys()), total=len(image_dict)):  # loop through all images\n",
    "    image_dict[key] = ammico.TextDetector(image_dict[key],                  # analyse image with TextDetector and update dict\n",
    "    analyse_text=True, model_names=[\"sshleifer/distilbart-cnn-12-6\", \n",
    "    \"distilbert-base-uncased-finetuned-sst-2-english\", \n",
    "    \"dbmdz/bert-large-cased-finetuned-conll03-english\"], \n",
    "    revision_numbers=[\"a4f8f3e\", \"af0f99b\", \"f2482bf\"]).analyse_image()\n",
    "    \n",
    "    if num % dump_every == 0 | num == len(image_dict) - 1:   # save results every dump_every to dump_file\n",
    "        image_df = ammico.get_dataframe(image_dict)\n",
    "        image_df.to_csv(dump_file)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The models can be adapted interactively in the notebook interface and the best models can then be used in a subsequent analysis of the whole data set.\n",
    "\n",
    "A detailed description of the output keys and data types is given in the following table.\n",
    "\n",
    "| output key | output type | output value |\n",
    "| ---------- | ----------- | ------------ |\n",
    "| `text` | `str` | the extracted text in the original language |\n",
    "| `text_language` | `str` | the detected dominant language of the extracted text |\n",
    "| `text_english` | `str` | the text translated into English |\n",
    "| `text_clean` | `str` | the text after cleaning from numbers and unrecognizable words |\n",
    "| `text_summary` | `str` | the summary of the text, generated with a transformers model |\n",
    "| `sentiment` | `str` | the detected sentiment, generated with a transformers model |\n",
    "| `sentiment_score` | `float` | the confidence associated with the predicted sentiment |\n",
    "| `entity` | `list[str]` | the detected named entities, generated with a transformers model |\n",
    "| `entity_type` | `list[str]` | the detected entity type |\n",
    "\n",
    "## Image summary and query\n",
    "\n",
    "The `SummaryDetector` can be used to generate image captions (`summary`) as well as visual question answering (`VQA`). \n",
    "\n",
    "<img src=\"../_static/summary_detector.png\" width=\"800\" />\n",
    "\n",
    "This module is based on the [LAVIS](https://github.com/salesforce/LAVIS) library. Since the models can be quite large, an initial object is created which will load the necessary models into RAM/VRAM and then use them in the analysis. The user can specify the type of analysis to be performed using the `analysis_type` keyword. Setting it to `summary` will generate a caption (summary), `questions` will prepare answers (VQA) to a list of questions as set by the user, `summary_and_questions` will do both. Note that the desired analysis type needs to be set here in the initialization of the \n",
    "detector object, and not when running the analysis for each image; the same holds true for the selected model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The implemented models are listed below.\n",
    "\n",
    "| input model name | model |\n",
    "| ---------------- | ----- |\n",
    "| base | BLIP image captioning base, ViT-B/16, pretrained on COCO dataset |\n",
    "| large | BLIP image captioning large, ViT-L/16, pretrained on COCO dataset |\n",
    "| vqa | BLIP base model fine-tuned on VQA v2.0 dataset |\n",
    "| blip2_t5_pretrain_flant5xxl | BLIP2 pretrained on FlanT5<sub>XXL</sub> | \n",
    "| blip2_t5_pretrain_flant5xl | BLIP2 pretrained on FlanT5<sub>XL</sub> | \n",
    "| blip2_t5_caption_coco_flant5xl | BLIP2 pretrained on FlanT5<sub>XL</sub>, fine-tuned on COCO | \n",
    "| blip2_opt_pretrain_opt2.7b | BLIP2 pretrained on OPT-2.7b |\n",
    "| blip2_opt_pretrain_opt6.7b | BLIP2 pretrained on OPT-6.7b | \n",
    "| blip2_opt_caption_coco_opt2.7b | BLIP2 pretrained on OPT-2.7b, fine-tuned on COCO | \n",
    "| blip2_opt_caption_coco_opt6.7b | BLIP2 pretrained on OPT-6.7b, fine-tuned on COCO |\n",
    "\n",
    "Please note that `base`, `large` and `vqa` models can be run on the base TPU video card in Google Colab.\n",
    "To run any advanced `BLIP2` models you need more than 20 gb of video memory, so you need to connect a paid A100 in Google Colab."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First of all, we can run only the summary module `analysis_type`. You can choose a `base` or a `large` model_type. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:51:21.186991Z",
     "iopub.status.busy": "2024-02-19T08:51:21.186599Z",
     "iopub.status.idle": "2024-02-19T08:51:27.937584Z",
     "shell.execute_reply": "2024-02-19T08:51:27.936965Z"
    }
   },
   "outputs": [],
   "source": [
    "image_summary_detector = ammico.SummaryDetector(image_dict, analysis_type=\"summary\", model_type=\"base\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:51:27.945516Z",
     "iopub.status.busy": "2024-02-19T08:51:27.945076Z",
     "iopub.status.idle": "2024-02-19T08:52:37.388418Z",
     "shell.execute_reply": "2024-02-19T08:52:37.387691Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0/6 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 17%|█▋        | 1/6 [00:13<01:06, 13.34s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 33%|███▎      | 2/6 [00:24<00:48, 12.03s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 50%|█████     | 3/6 [00:34<00:33, 11.26s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 67%|██████▋   | 4/6 [00:47<00:23, 11.73s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 83%|████████▎ | 5/6 [00:58<00:11, 11.50s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 6/6 [01:09<00:00, 11.36s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 6/6 [01:09<00:00, 11.57s/it]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "for num, key in tqdm(enumerate(image_dict.keys()),total=len(image_dict)):\n",
    "    image_dict[key] = image_summary_detector.analyse_image(subdict = image_dict[key], analysis_type=\"summary\")\n",
    "    \n",
    "    if num % dump_every == 0 | num == len(image_dict) - 1: \n",
    "        image_df = ammico.get_dataframe(image_dict)\n",
    "        image_df.to_csv(dump_file)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For VQA, a list of questions needs to be passed when carrying out the analysis; these should be given as a list of strings."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:52:37.396555Z",
     "iopub.status.busy": "2024-02-19T08:52:37.396110Z",
     "iopub.status.idle": "2024-02-19T08:52:37.399491Z",
     "shell.execute_reply": "2024-02-19T08:52:37.398848Z"
    }
   },
   "outputs": [],
   "source": [
    "list_of_questions = [\n",
    "    \"How many persons on the picture?\",\n",
    "    \"Are there any politicians in the picture?\",\n",
    "    \"Does the picture show something from medicine?\",\n",
    "]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If you want to execute only the VQA module without captioning, just specify the `analysis_type` as `questions` and `model_type` as `vqa`. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:52:37.405491Z",
     "iopub.status.busy": "2024-02-19T08:52:37.404988Z",
     "iopub.status.idle": "2024-02-19T08:52:56.401148Z",
     "shell.execute_reply": "2024-02-19T08:52:56.400055Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0.00/1.35G [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 4.01M/1.35G [00:00<00:56, 25.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  1%|          | 11.3M/1.35G [00:00<00:31, 45.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  1%|▏         | 17.8M/1.35G [00:00<00:30, 47.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  2%|▏         | 24.0M/1.35G [00:00<00:28, 49.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  2%|▏         | 32.0M/1.35G [00:00<00:26, 53.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  3%|▎         | 40.0M/1.35G [00:00<00:26, 52.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  3%|▎         | 48.0M/1.35G [00:00<00:23, 60.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  4%|▍         | 56.0M/1.35G [00:01<00:22, 61.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  5%|▍         | 64.0M/1.35G [00:01<00:21, 63.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  5%|▌         | 72.0M/1.35G [00:01<00:26, 51.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  6%|▌         | 80.0M/1.35G [00:01<00:30, 44.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  6%|▋         | 88.0M/1.35G [00:01<00:26, 50.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  7%|▋         | 96.0M/1.35G [00:01<00:28, 47.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  8%|▊         | 104M/1.35G [00:02<00:27, 48.3MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  8%|▊         | 112M/1.35G [00:02<00:26, 50.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  9%|▊         | 120M/1.35G [00:02<00:25, 51.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 10%|▉         | 136M/1.35G [00:02<00:17, 74.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 11%|█         | 153M/1.35G [00:02<00:13, 98.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 12%|█▏        | 168M/1.35G [00:02<00:11, 112MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 13%|█▎        | 184M/1.35G [00:02<00:10, 122MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 15%|█▍        | 207M/1.35G [00:02<00:08, 153MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 16%|█▌        | 224M/1.35G [00:03<00:09, 129MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 17%|█▋        | 240M/1.35G [00:03<00:09, 131MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 19%|█▊        | 256M/1.35G [00:03<00:09, 124MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 20%|██        | 280M/1.35G [00:03<00:08, 144MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 21%|██▏       | 294M/1.35G [00:03<00:08, 136MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 23%|██▎       | 312M/1.35G [00:03<00:07, 147MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 24%|██▎       | 326M/1.35G [00:03<00:07, 148MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 26%|██▌       | 352M/1.35G [00:04<00:06, 172MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 27%|██▋       | 369M/1.35G [00:04<00:07, 148MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 28%|██▊       | 388M/1.35G [00:04<00:06, 161MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 30%|██▉       | 413M/1.35G [00:04<00:05, 187MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 32%|███▏      | 437M/1.35G [00:04<00:04, 207MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 33%|███▎      | 461M/1.35G [00:04<00:04, 217MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 35%|███▌      | 486M/1.35G [00:04<00:04, 232MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 37%|███▋      | 511M/1.35G [00:04<00:03, 241MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 39%|███▉      | 537M/1.35G [00:04<00:03, 250MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 41%|████      | 561M/1.35G [00:05<00:03, 244MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 42%|████▏     | 585M/1.35G [00:05<00:03, 236MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 44%|████▍     | 609M/1.35G [00:05<00:03, 241MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 46%|████▌     | 632M/1.35G [00:05<00:03, 200MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 47%|████▋     | 652M/1.35G [00:06<00:16, 47.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 49%|████▊     | 672M/1.35G [00:07<00:14, 52.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 50%|████▉     | 688M/1.35G [00:07<00:11, 62.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 51%|█████     | 704M/1.35G [00:07<00:09, 73.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 52%|█████▏    | 720M/1.35G [00:07<00:08, 85.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 54%|█████▍    | 742M/1.35G [00:07<00:06, 109MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 55%|█████▌    | 763M/1.35G [00:07<00:04, 130MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 57%|█████▋    | 780M/1.35G [00:07<00:04, 133MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 58%|█████▊    | 800M/1.35G [00:07<00:04, 128MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 59%|█████▉    | 816M/1.35G [00:08<00:06, 89.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 61%|██████    | 840M/1.35G [00:08<00:05, 102MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 62%|██████▏   | 858M/1.35G [00:08<00:04, 116MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 63%|██████▎   | 873M/1.35G [00:08<00:04, 125MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 65%|██████▍   | 896M/1.35G [00:08<00:03, 141MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 67%|██████▋   | 918M/1.35G [00:08<00:02, 162MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 68%|██████▊   | 936M/1.35G [00:09<00:03, 144MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 69%|██████▉   | 957M/1.35G [00:09<00:06, 68.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 71%|███████▏  | 985M/1.35G [00:09<00:04, 95.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 73%|███████▎  | 0.98G/1.35G [00:09<00:04, 93.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 74%|███████▎  | 0.99G/1.35G [00:10<00:04, 92.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 75%|███████▍  | 1.00G/1.35G [00:10<00:03, 98.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 76%|███████▌  | 1.02G/1.35G [00:10<00:03, 100MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 78%|███████▊  | 1.05G/1.35G [00:10<00:02, 134MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 80%|███████▉  | 1.07G/1.35G [00:10<00:01, 157MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 81%|████████  | 1.09G/1.35G [00:11<00:04, 68.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 83%|████████▎ | 1.12G/1.35G [00:11<00:02, 93.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 84%|████████▍ | 1.13G/1.35G [00:11<00:02, 104MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 85%|████████▌ | 1.15G/1.35G [00:11<00:01, 113MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 87%|████████▋ | 1.17G/1.35G [00:11<00:01, 112MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 88%|████████▊ | 1.18G/1.35G [00:12<00:03, 52.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 89%|████████▉ | 1.20G/1.35G [00:12<00:02, 66.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 90%|████████▉ | 1.21G/1.35G [00:13<00:03, 45.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 91%|█████████ | 1.23G/1.35G [00:13<00:02, 44.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 91%|█████████▏| 1.23G/1.35G [00:13<00:01, 96.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "ename": "OSError",
     "evalue": "[Errno 28] No space left on device",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mOSError\u001b[0m                                   Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[21], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m image_summary_vqa_detector \u001b[38;5;241m=\u001b[39m \u001b[43mammico\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mSummaryDetector\u001b[49m\u001b[43m(\u001b[49m\u001b[43mimage_dict\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43manalysis_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mquestions\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\n\u001b[1;32m      2\u001b[0m \u001b[43m                                                    \u001b[49m\u001b[43mmodel_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mvqa\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m)\u001b[49m\n\u001b[1;32m      4\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m num, key \u001b[38;5;129;01min\u001b[39;00m tqdm(\u001b[38;5;28menumerate\u001b[39m(image_dict\u001b[38;5;241m.\u001b[39mkeys()),total\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mlen\u001b[39m(image_dict)):\n\u001b[1;32m      5\u001b[0m     image_dict[key] \u001b[38;5;241m=\u001b[39m image_summary_vqa_detector\u001b[38;5;241m.\u001b[39manalyse_image(subdict\u001b[38;5;241m=\u001b[39mimage_dict[key], \n\u001b[1;32m      6\u001b[0m                                                                analysis_type\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mquestions\u001b[39m\u001b[38;5;124m\"\u001b[39m, \n\u001b[1;32m      7\u001b[0m                                                                list_of_questions \u001b[38;5;241m=\u001b[39m list_of_questions)\n",
      "File \u001b[0;32m~/work/AMMICO/AMMICO/ammico/summary.py:141\u001b[0m, in \u001b[0;36mSummaryDetector.__init__\u001b[0;34m(self, subdict, model_type, analysis_type, list_of_questions, summary_model, summary_vis_processors, summary_vqa_model, summary_vqa_vis_processors, summary_vqa_txt_processors, summary_vqa_model_new, summary_vqa_vis_processors_new, summary_vqa_txt_processors_new, device_type)\u001b[0m\n\u001b[1;32m    127\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msummary_vis_processors \u001b[38;5;241m=\u001b[39m summary_vis_processors\n\u001b[1;32m    128\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m (\n\u001b[1;32m    129\u001b[0m     model_type \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mallowed_model_types\n\u001b[1;32m    130\u001b[0m     \u001b[38;5;129;01mand\u001b[39;00m (summary_vqa_model \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m)\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    135\u001b[0m     )\n\u001b[1;32m    136\u001b[0m ):\n\u001b[1;32m    137\u001b[0m     (\n\u001b[1;32m    138\u001b[0m         \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msummary_vqa_model,\n\u001b[1;32m    139\u001b[0m         \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msummary_vqa_vis_processors,\n\u001b[1;32m    140\u001b[0m         \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msummary_vqa_txt_processors,\n\u001b[0;32m--> 141\u001b[0m     ) \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mload_vqa_model\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    142\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m    143\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msummary_vqa_model \u001b[38;5;241m=\u001b[39m summary_vqa_model\n",
      "File \u001b[0;32m~/work/AMMICO/AMMICO/ammico/summary.py:232\u001b[0m, in \u001b[0;36mSummaryDetector.load_vqa_model\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    216\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mload_vqa_model\u001b[39m(\u001b[38;5;28mself\u001b[39m):\n\u001b[1;32m    217\u001b[0m \u001b[38;5;250m    \u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m    218\u001b[0m \u001b[38;5;124;03m    Load blip_vqa model and preprocessors for visual and text inputs from lavis.models.\u001b[39;00m\n\u001b[1;32m    219\u001b[0m \n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    226\u001b[0m \n\u001b[1;32m    227\u001b[0m \u001b[38;5;124;03m    \"\"\"\u001b[39;00m\n\u001b[1;32m    228\u001b[0m     (\n\u001b[1;32m    229\u001b[0m         summary_vqa_model,\n\u001b[1;32m    230\u001b[0m         summary_vqa_vis_processors,\n\u001b[1;32m    231\u001b[0m         summary_vqa_txt_processors,\n\u001b[0;32m--> 232\u001b[0m     ) \u001b[38;5;241m=\u001b[39m \u001b[43mload_model_and_preprocess\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m    233\u001b[0m \u001b[43m        \u001b[49m\u001b[43mname\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mblip_vqa\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[1;32m    234\u001b[0m \u001b[43m        \u001b[49m\u001b[43mmodel_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mvqav2\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[1;32m    235\u001b[0m \u001b[43m        \u001b[49m\u001b[43mis_eval\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\u001b[43m,\u001b[49m\n\u001b[1;32m    236\u001b[0m \u001b[43m        \u001b[49m\u001b[43mdevice\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43msummary_device\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    237\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    238\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m summary_vqa_model, summary_vqa_vis_processors, summary_vqa_txt_processors\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/__init__.py:195\u001b[0m, in \u001b[0;36mload_model_and_preprocess\u001b[0;34m(name, model_type, is_eval, device)\u001b[0m\n\u001b[1;32m    192\u001b[0m model_cls \u001b[38;5;241m=\u001b[39m registry\u001b[38;5;241m.\u001b[39mget_model_class(name)\n\u001b[1;32m    194\u001b[0m \u001b[38;5;66;03m# load model\u001b[39;00m\n\u001b[0;32m--> 195\u001b[0m model \u001b[38;5;241m=\u001b[39m \u001b[43mmodel_cls\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfrom_pretrained\u001b[49m\u001b[43m(\u001b[49m\u001b[43mmodel_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mmodel_type\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    197\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m is_eval:\n\u001b[1;32m    198\u001b[0m     model\u001b[38;5;241m.\u001b[39meval()\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/base_model.py:70\u001b[0m, in \u001b[0;36mBaseModel.from_pretrained\u001b[0;34m(cls, model_type)\u001b[0m\n\u001b[1;32m     60\u001b[0m \u001b[38;5;250m\u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m     61\u001b[0m \u001b[38;5;124;03mBuild a pretrained model from default configuration file, specified by model_type.\u001b[39;00m\n\u001b[1;32m     62\u001b[0m \n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m     67\u001b[0m \u001b[38;5;124;03m    - model (nn.Module): pretrained or finetuned model, depending on the configuration.\u001b[39;00m\n\u001b[1;32m     68\u001b[0m \u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m     69\u001b[0m model_cfg \u001b[38;5;241m=\u001b[39m OmegaConf\u001b[38;5;241m.\u001b[39mload(\u001b[38;5;28mcls\u001b[39m\u001b[38;5;241m.\u001b[39mdefault_config_path(model_type))\u001b[38;5;241m.\u001b[39mmodel\n\u001b[0;32m---> 70\u001b[0m model \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mcls\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfrom_config\u001b[49m\u001b[43m(\u001b[49m\u001b[43mmodel_cfg\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     72\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m model\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/blip_models/blip_vqa.py:373\u001b[0m, in \u001b[0;36mBlipVQA.from_config\u001b[0;34m(cls, cfg)\u001b[0m\n\u001b[1;32m    364\u001b[0m max_txt_len \u001b[38;5;241m=\u001b[39m cfg\u001b[38;5;241m.\u001b[39mget(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mmax_txt_len\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;241m35\u001b[39m)\n\u001b[1;32m    366\u001b[0m model \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mcls\u001b[39m(\n\u001b[1;32m    367\u001b[0m     image_encoder\u001b[38;5;241m=\u001b[39mimage_encoder,\n\u001b[1;32m    368\u001b[0m     text_encoder\u001b[38;5;241m=\u001b[39mtext_encoder,\n\u001b[1;32m    369\u001b[0m     text_decoder\u001b[38;5;241m=\u001b[39mtext_decoder,\n\u001b[1;32m    370\u001b[0m     max_txt_len\u001b[38;5;241m=\u001b[39mmax_txt_len,\n\u001b[1;32m    371\u001b[0m )\n\u001b[0;32m--> 373\u001b[0m \u001b[43mmodel\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mload_checkpoint_from_config\u001b[49m\u001b[43m(\u001b[49m\u001b[43mcfg\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    375\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m model\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/base_model.py:95\u001b[0m, in \u001b[0;36mBaseModel.load_checkpoint_from_config\u001b[0;34m(self, cfg, **kwargs)\u001b[0m\n\u001b[1;32m     91\u001b[0m     finetune_path \u001b[38;5;241m=\u001b[39m cfg\u001b[38;5;241m.\u001b[39mget(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfinetuned\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;28;01mNone\u001b[39;00m)\n\u001b[1;32m     92\u001b[0m     \u001b[38;5;28;01massert\u001b[39;00m (\n\u001b[1;32m     93\u001b[0m         finetune_path \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[1;32m     94\u001b[0m     ), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mFound load_finetuned is True, but finetune_path is None.\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[0;32m---> 95\u001b[0m     \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mload_checkpoint\u001b[49m\u001b[43m(\u001b[49m\u001b[43murl_or_filename\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mfinetune_path\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     96\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m     97\u001b[0m     \u001b[38;5;66;03m# load pre-trained weights\u001b[39;00m\n\u001b[1;32m     98\u001b[0m     pretrain_path \u001b[38;5;241m=\u001b[39m cfg\u001b[38;5;241m.\u001b[39mget(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mpretrained\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;28;01mNone\u001b[39;00m)\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/base_model.py:37\u001b[0m, in \u001b[0;36mBaseModel.load_checkpoint\u001b[0;34m(self, url_or_filename)\u001b[0m\n\u001b[1;32m     30\u001b[0m \u001b[38;5;250m\u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m     31\u001b[0m \u001b[38;5;124;03mLoad from a finetuned checkpoint.\u001b[39;00m\n\u001b[1;32m     32\u001b[0m \n\u001b[1;32m     33\u001b[0m \u001b[38;5;124;03mThis should expect no mismatch in the model keys and the checkpoint keys.\u001b[39;00m\n\u001b[1;32m     34\u001b[0m \u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m     36\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m is_url(url_or_filename):\n\u001b[0;32m---> 37\u001b[0m     cached_file \u001b[38;5;241m=\u001b[39m \u001b[43mdownload_cached_file\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m     38\u001b[0m \u001b[43m        \u001b[49m\u001b[43murl_or_filename\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcheck_hash\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mFalse\u001b[39;49;00m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mprogress\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\n\u001b[1;32m     39\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     40\u001b[0m     checkpoint \u001b[38;5;241m=\u001b[39m torch\u001b[38;5;241m.\u001b[39mload(cached_file, map_location\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mcpu\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[1;32m     41\u001b[0m \u001b[38;5;28;01melif\u001b[39;00m os\u001b[38;5;241m.\u001b[39mpath\u001b[38;5;241m.\u001b[39misfile(url_or_filename):\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/common/dist_utils.py:132\u001b[0m, in \u001b[0;36mdownload_cached_file\u001b[0;34m(url, check_hash, progress)\u001b[0m\n\u001b[1;32m    129\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m cached_file\n\u001b[1;32m    131\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m is_main_process():\n\u001b[0;32m--> 132\u001b[0m     \u001b[43mtimm_hub\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mdownload_cached_file\u001b[49m\u001b[43m(\u001b[49m\u001b[43murl\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcheck_hash\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mprogress\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    134\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m is_dist_avail_and_initialized():\n\u001b[1;32m    135\u001b[0m     dist\u001b[38;5;241m.\u001b[39mbarrier()\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/timm/models/hub.py:51\u001b[0m, in \u001b[0;36mdownload_cached_file\u001b[0;34m(url, check_hash, progress)\u001b[0m\n\u001b[1;32m     49\u001b[0m         r \u001b[38;5;241m=\u001b[39m HASH_REGEX\u001b[38;5;241m.\u001b[39msearch(filename)  \u001b[38;5;66;03m# r is Optional[Match[str]]\u001b[39;00m\n\u001b[1;32m     50\u001b[0m         hash_prefix \u001b[38;5;241m=\u001b[39m r\u001b[38;5;241m.\u001b[39mgroup(\u001b[38;5;241m1\u001b[39m) \u001b[38;5;28;01mif\u001b[39;00m r \u001b[38;5;28;01melse\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[0;32m---> 51\u001b[0m     \u001b[43mdownload_url_to_file\u001b[49m\u001b[43m(\u001b[49m\u001b[43murl\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcached_file\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mhash_prefix\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mprogress\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mprogress\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     52\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m cached_file\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/torch/hub.py:636\u001b[0m, in \u001b[0;36mdownload_url_to_file\u001b[0;34m(url, dst, hash_prefix, progress)\u001b[0m\n\u001b[1;32m    634\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mlen\u001b[39m(buffer) \u001b[38;5;241m==\u001b[39m \u001b[38;5;241m0\u001b[39m:\n\u001b[1;32m    635\u001b[0m     \u001b[38;5;28;01mbreak\u001b[39;00m\n\u001b[0;32m--> 636\u001b[0m \u001b[43mf\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mwrite\u001b[49m\u001b[43m(\u001b[49m\u001b[43mbuffer\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    637\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m hash_prefix \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m    638\u001b[0m     sha256\u001b[38;5;241m.\u001b[39mupdate(buffer)\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/tempfile.py:478\u001b[0m, in \u001b[0;36m_TemporaryFileWrapper.__getattr__.<locals>.func_wrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m    476\u001b[0m \u001b[38;5;129m@_functools\u001b[39m\u001b[38;5;241m.\u001b[39mwraps(func)\n\u001b[1;32m    477\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mfunc_wrapper\u001b[39m(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs):\n\u001b[0;32m--> 478\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfunc\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "\u001b[0;31mOSError\u001b[0m: [Errno 28] No space left on device"
     ]
    }
   ],
   "source": [
    "image_summary_vqa_detector = ammico.SummaryDetector(image_dict, analysis_type=\"questions\", \n",
    "                                                    model_type=\"vqa\")\n",
    "\n",
    "for num, key in tqdm(enumerate(image_dict.keys()),total=len(image_dict)):\n",
    "    image_dict[key] = image_summary_vqa_detector.analyse_image(subdict=image_dict[key], \n",
    "                                                               analysis_type=\"questions\", \n",
    "                                                               list_of_questions = list_of_questions)\n",
    "    if num % dump_every == 0 | num == len(image_dict) - 1: \n",
    "        image_df = ammico.get_dataframe(image_dict)\n",
    "        image_df.to_csv(dump_file)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Or you can specify the analysis type as `summary_and_questions`, then both caption creation and question answers will be generated for each image. In this case, you can choose a `base` or a `large` model_type. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:52:56.408315Z",
     "iopub.status.busy": "2024-02-19T08:52:56.407932Z",
     "iopub.status.idle": "2024-02-19T08:53:18.175891Z",
     "shell.execute_reply": "2024-02-19T08:53:18.175071Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0.00/1.35G [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  1%|          | 8.78M/1.35G [00:00<00:15, 92.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  2%|▏         | 34.1M/1.35G [00:00<00:07, 194MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  4%|▍         | 58.3M/1.35G [00:00<00:06, 221MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  6%|▌         | 82.8M/1.35G [00:00<00:05, 236MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  8%|▊         | 109M/1.35G [00:00<00:05, 250MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 10%|▉         | 137M/1.35G [00:00<00:04, 264MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 12%|█▏        | 162M/1.35G [00:00<00:05, 253MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 14%|█▎        | 186M/1.35G [00:00<00:05, 246MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 15%|█▌        | 212M/1.35G [00:00<00:04, 252MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 17%|█▋        | 236M/1.35G [00:01<00:04, 250MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 19%|█▉        | 263M/1.35G [00:01<00:04, 261MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 21%|██        | 290M/1.35G [00:01<00:04, 267MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 23%|██▎       | 317M/1.35G [00:01<00:04, 272MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 25%|██▍       | 344M/1.35G [00:01<00:03, 275MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 27%|██▋       | 370M/1.35G [00:01<00:03, 267MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 29%|██▉       | 397M/1.35G [00:01<00:03, 272MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 31%|███       | 425M/1.35G [00:01<00:03, 277MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 33%|███▎      | 453M/1.35G [00:01<00:03, 281MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 35%|███▍      | 479M/1.35G [00:01<00:03, 278MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 37%|███▋      | 506M/1.35G [00:02<00:03, 265MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 39%|███▊      | 533M/1.35G [00:02<00:03, 269MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 40%|████      | 558M/1.35G [00:02<00:03, 268MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 42%|████▏     | 584M/1.35G [00:02<00:03, 269MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 44%|████▍     | 610M/1.35G [00:02<00:03, 267MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 46%|████▌     | 635M/1.35G [00:02<00:03, 260MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 48%|████▊     | 660M/1.35G [00:04<00:14, 52.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 49%|████▉     | 683M/1.35G [00:04<00:10, 66.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 51%|█████▏    | 708M/1.35G [00:04<00:08, 86.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 53%|█████▎    | 736M/1.35G [00:04<00:05, 113MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 55%|█████▌    | 763M/1.35G [00:04<00:04, 137MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 57%|█████▋    | 787M/1.35G [00:04<00:03, 158MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 59%|█████▉    | 811M/1.35G [00:04<00:03, 175MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 61%|██████    | 838M/1.35G [00:04<00:02, 200MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 63%|██████▎   | 866M/1.35G [00:04<00:02, 221MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 65%|██████▍   | 893M/1.35G [00:04<00:02, 236MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 67%|██████▋   | 919M/1.35G [00:05<00:01, 248MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 69%|██████▊   | 946M/1.35G [00:05<00:01, 252MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 70%|███████   | 971M/1.35G [00:05<00:04, 102MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 72%|███████▏  | 999M/1.35G [00:05<00:03, 127MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 74%|███████▍  | 1.00G/1.35G [00:05<00:02, 156MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 76%|███████▌  | 1.03G/1.35G [00:06<00:01, 173MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 78%|███████▊  | 1.05G/1.35G [00:06<00:01, 188MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 80%|███████▉  | 1.07G/1.35G [00:06<00:01, 202MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 81%|████████▏ | 1.10G/1.35G [00:06<00:03, 85.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 83%|████████▎ | 1.12G/1.35G [00:07<00:02, 106MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 85%|████████▍ | 1.14G/1.35G [00:07<00:01, 127MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 87%|████████▋ | 1.17G/1.35G [00:07<00:01, 149MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 88%|████████▊ | 1.19G/1.35G [00:08<00:02, 71.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 90%|████████▉ | 1.21G/1.35G [00:08<00:02, 56.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 91%|█████████ | 1.23G/1.35G [00:09<00:02, 51.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 91%|█████████▏| 1.23G/1.35G [00:09<00:00, 144MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "ename": "OSError",
     "evalue": "[Errno 28] No space left on device",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mOSError\u001b[0m                                   Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[22], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m image_summary_vqa_detector \u001b[38;5;241m=\u001b[39m \u001b[43mammico\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mSummaryDetector\u001b[49m\u001b[43m(\u001b[49m\u001b[43mimage_dict\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43manalysis_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43msummary_and_questions\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\n\u001b[1;32m      2\u001b[0m \u001b[43m                                                    \u001b[49m\u001b[43mmodel_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mbase\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m)\u001b[49m\n\u001b[1;32m      3\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m num, key \u001b[38;5;129;01min\u001b[39;00m tqdm(\u001b[38;5;28menumerate\u001b[39m(image_dict\u001b[38;5;241m.\u001b[39mkeys()),total\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mlen\u001b[39m(image_dict)):\n\u001b[1;32m      4\u001b[0m     image_dict[key] \u001b[38;5;241m=\u001b[39m image_summary_vqa_detector\u001b[38;5;241m.\u001b[39manalyse_image(subdict\u001b[38;5;241m=\u001b[39mimage_dict[key], \n\u001b[1;32m      5\u001b[0m                                                                analysis_type\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124msummary_and_questions\u001b[39m\u001b[38;5;124m\"\u001b[39m, \n\u001b[1;32m      6\u001b[0m                                                                list_of_questions \u001b[38;5;241m=\u001b[39m list_of_questions)\n",
      "File \u001b[0;32m~/work/AMMICO/AMMICO/ammico/summary.py:141\u001b[0m, in \u001b[0;36mSummaryDetector.__init__\u001b[0;34m(self, subdict, model_type, analysis_type, list_of_questions, summary_model, summary_vis_processors, summary_vqa_model, summary_vqa_vis_processors, summary_vqa_txt_processors, summary_vqa_model_new, summary_vqa_vis_processors_new, summary_vqa_txt_processors_new, device_type)\u001b[0m\n\u001b[1;32m    127\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msummary_vis_processors \u001b[38;5;241m=\u001b[39m summary_vis_processors\n\u001b[1;32m    128\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m (\n\u001b[1;32m    129\u001b[0m     model_type \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mallowed_model_types\n\u001b[1;32m    130\u001b[0m     \u001b[38;5;129;01mand\u001b[39;00m (summary_vqa_model \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m)\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    135\u001b[0m     )\n\u001b[1;32m    136\u001b[0m ):\n\u001b[1;32m    137\u001b[0m     (\n\u001b[1;32m    138\u001b[0m         \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msummary_vqa_model,\n\u001b[1;32m    139\u001b[0m         \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msummary_vqa_vis_processors,\n\u001b[1;32m    140\u001b[0m         \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msummary_vqa_txt_processors,\n\u001b[0;32m--> 141\u001b[0m     ) \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mload_vqa_model\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    142\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m    143\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msummary_vqa_model \u001b[38;5;241m=\u001b[39m summary_vqa_model\n",
      "File \u001b[0;32m~/work/AMMICO/AMMICO/ammico/summary.py:232\u001b[0m, in \u001b[0;36mSummaryDetector.load_vqa_model\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    216\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mload_vqa_model\u001b[39m(\u001b[38;5;28mself\u001b[39m):\n\u001b[1;32m    217\u001b[0m \u001b[38;5;250m    \u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m    218\u001b[0m \u001b[38;5;124;03m    Load blip_vqa model and preprocessors for visual and text inputs from lavis.models.\u001b[39;00m\n\u001b[1;32m    219\u001b[0m \n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    226\u001b[0m \n\u001b[1;32m    227\u001b[0m \u001b[38;5;124;03m    \"\"\"\u001b[39;00m\n\u001b[1;32m    228\u001b[0m     (\n\u001b[1;32m    229\u001b[0m         summary_vqa_model,\n\u001b[1;32m    230\u001b[0m         summary_vqa_vis_processors,\n\u001b[1;32m    231\u001b[0m         summary_vqa_txt_processors,\n\u001b[0;32m--> 232\u001b[0m     ) \u001b[38;5;241m=\u001b[39m \u001b[43mload_model_and_preprocess\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m    233\u001b[0m \u001b[43m        \u001b[49m\u001b[43mname\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mblip_vqa\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[1;32m    234\u001b[0m \u001b[43m        \u001b[49m\u001b[43mmodel_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mvqav2\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[1;32m    235\u001b[0m \u001b[43m        \u001b[49m\u001b[43mis_eval\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\u001b[43m,\u001b[49m\n\u001b[1;32m    236\u001b[0m \u001b[43m        \u001b[49m\u001b[43mdevice\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43msummary_device\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    237\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    238\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m summary_vqa_model, summary_vqa_vis_processors, summary_vqa_txt_processors\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/__init__.py:195\u001b[0m, in \u001b[0;36mload_model_and_preprocess\u001b[0;34m(name, model_type, is_eval, device)\u001b[0m\n\u001b[1;32m    192\u001b[0m model_cls \u001b[38;5;241m=\u001b[39m registry\u001b[38;5;241m.\u001b[39mget_model_class(name)\n\u001b[1;32m    194\u001b[0m \u001b[38;5;66;03m# load model\u001b[39;00m\n\u001b[0;32m--> 195\u001b[0m model \u001b[38;5;241m=\u001b[39m \u001b[43mmodel_cls\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfrom_pretrained\u001b[49m\u001b[43m(\u001b[49m\u001b[43mmodel_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mmodel_type\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    197\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m is_eval:\n\u001b[1;32m    198\u001b[0m     model\u001b[38;5;241m.\u001b[39meval()\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/base_model.py:70\u001b[0m, in \u001b[0;36mBaseModel.from_pretrained\u001b[0;34m(cls, model_type)\u001b[0m\n\u001b[1;32m     60\u001b[0m \u001b[38;5;250m\u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m     61\u001b[0m \u001b[38;5;124;03mBuild a pretrained model from default configuration file, specified by model_type.\u001b[39;00m\n\u001b[1;32m     62\u001b[0m \n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m     67\u001b[0m \u001b[38;5;124;03m    - model (nn.Module): pretrained or finetuned model, depending on the configuration.\u001b[39;00m\n\u001b[1;32m     68\u001b[0m \u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m     69\u001b[0m model_cfg \u001b[38;5;241m=\u001b[39m OmegaConf\u001b[38;5;241m.\u001b[39mload(\u001b[38;5;28mcls\u001b[39m\u001b[38;5;241m.\u001b[39mdefault_config_path(model_type))\u001b[38;5;241m.\u001b[39mmodel\n\u001b[0;32m---> 70\u001b[0m model \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mcls\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfrom_config\u001b[49m\u001b[43m(\u001b[49m\u001b[43mmodel_cfg\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     72\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m model\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/blip_models/blip_vqa.py:373\u001b[0m, in \u001b[0;36mBlipVQA.from_config\u001b[0;34m(cls, cfg)\u001b[0m\n\u001b[1;32m    364\u001b[0m max_txt_len \u001b[38;5;241m=\u001b[39m cfg\u001b[38;5;241m.\u001b[39mget(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mmax_txt_len\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;241m35\u001b[39m)\n\u001b[1;32m    366\u001b[0m model \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mcls\u001b[39m(\n\u001b[1;32m    367\u001b[0m     image_encoder\u001b[38;5;241m=\u001b[39mimage_encoder,\n\u001b[1;32m    368\u001b[0m     text_encoder\u001b[38;5;241m=\u001b[39mtext_encoder,\n\u001b[1;32m    369\u001b[0m     text_decoder\u001b[38;5;241m=\u001b[39mtext_decoder,\n\u001b[1;32m    370\u001b[0m     max_txt_len\u001b[38;5;241m=\u001b[39mmax_txt_len,\n\u001b[1;32m    371\u001b[0m )\n\u001b[0;32m--> 373\u001b[0m \u001b[43mmodel\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mload_checkpoint_from_config\u001b[49m\u001b[43m(\u001b[49m\u001b[43mcfg\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    375\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m model\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/base_model.py:95\u001b[0m, in \u001b[0;36mBaseModel.load_checkpoint_from_config\u001b[0;34m(self, cfg, **kwargs)\u001b[0m\n\u001b[1;32m     91\u001b[0m     finetune_path \u001b[38;5;241m=\u001b[39m cfg\u001b[38;5;241m.\u001b[39mget(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfinetuned\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;28;01mNone\u001b[39;00m)\n\u001b[1;32m     92\u001b[0m     \u001b[38;5;28;01massert\u001b[39;00m (\n\u001b[1;32m     93\u001b[0m         finetune_path \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[1;32m     94\u001b[0m     ), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mFound load_finetuned is True, but finetune_path is None.\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[0;32m---> 95\u001b[0m     \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mload_checkpoint\u001b[49m\u001b[43m(\u001b[49m\u001b[43murl_or_filename\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mfinetune_path\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     96\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m     97\u001b[0m     \u001b[38;5;66;03m# load pre-trained weights\u001b[39;00m\n\u001b[1;32m     98\u001b[0m     pretrain_path \u001b[38;5;241m=\u001b[39m cfg\u001b[38;5;241m.\u001b[39mget(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mpretrained\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;28;01mNone\u001b[39;00m)\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/base_model.py:37\u001b[0m, in \u001b[0;36mBaseModel.load_checkpoint\u001b[0;34m(self, url_or_filename)\u001b[0m\n\u001b[1;32m     30\u001b[0m \u001b[38;5;250m\u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m     31\u001b[0m \u001b[38;5;124;03mLoad from a finetuned checkpoint.\u001b[39;00m\n\u001b[1;32m     32\u001b[0m \n\u001b[1;32m     33\u001b[0m \u001b[38;5;124;03mThis should expect no mismatch in the model keys and the checkpoint keys.\u001b[39;00m\n\u001b[1;32m     34\u001b[0m \u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m     36\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m is_url(url_or_filename):\n\u001b[0;32m---> 37\u001b[0m     cached_file \u001b[38;5;241m=\u001b[39m \u001b[43mdownload_cached_file\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m     38\u001b[0m \u001b[43m        \u001b[49m\u001b[43murl_or_filename\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcheck_hash\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mFalse\u001b[39;49;00m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mprogress\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\n\u001b[1;32m     39\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     40\u001b[0m     checkpoint \u001b[38;5;241m=\u001b[39m torch\u001b[38;5;241m.\u001b[39mload(cached_file, map_location\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mcpu\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[1;32m     41\u001b[0m \u001b[38;5;28;01melif\u001b[39;00m os\u001b[38;5;241m.\u001b[39mpath\u001b[38;5;241m.\u001b[39misfile(url_or_filename):\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/common/dist_utils.py:132\u001b[0m, in \u001b[0;36mdownload_cached_file\u001b[0;34m(url, check_hash, progress)\u001b[0m\n\u001b[1;32m    129\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m cached_file\n\u001b[1;32m    131\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m is_main_process():\n\u001b[0;32m--> 132\u001b[0m     \u001b[43mtimm_hub\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mdownload_cached_file\u001b[49m\u001b[43m(\u001b[49m\u001b[43murl\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcheck_hash\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mprogress\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    134\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m is_dist_avail_and_initialized():\n\u001b[1;32m    135\u001b[0m     dist\u001b[38;5;241m.\u001b[39mbarrier()\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/timm/models/hub.py:51\u001b[0m, in \u001b[0;36mdownload_cached_file\u001b[0;34m(url, check_hash, progress)\u001b[0m\n\u001b[1;32m     49\u001b[0m         r \u001b[38;5;241m=\u001b[39m HASH_REGEX\u001b[38;5;241m.\u001b[39msearch(filename)  \u001b[38;5;66;03m# r is Optional[Match[str]]\u001b[39;00m\n\u001b[1;32m     50\u001b[0m         hash_prefix \u001b[38;5;241m=\u001b[39m r\u001b[38;5;241m.\u001b[39mgroup(\u001b[38;5;241m1\u001b[39m) \u001b[38;5;28;01mif\u001b[39;00m r \u001b[38;5;28;01melse\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[0;32m---> 51\u001b[0m     \u001b[43mdownload_url_to_file\u001b[49m\u001b[43m(\u001b[49m\u001b[43murl\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcached_file\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mhash_prefix\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mprogress\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mprogress\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     52\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m cached_file\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/torch/hub.py:636\u001b[0m, in \u001b[0;36mdownload_url_to_file\u001b[0;34m(url, dst, hash_prefix, progress)\u001b[0m\n\u001b[1;32m    634\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mlen\u001b[39m(buffer) \u001b[38;5;241m==\u001b[39m \u001b[38;5;241m0\u001b[39m:\n\u001b[1;32m    635\u001b[0m     \u001b[38;5;28;01mbreak\u001b[39;00m\n\u001b[0;32m--> 636\u001b[0m \u001b[43mf\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mwrite\u001b[49m\u001b[43m(\u001b[49m\u001b[43mbuffer\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    637\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m hash_prefix \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m    638\u001b[0m     sha256\u001b[38;5;241m.\u001b[39mupdate(buffer)\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/tempfile.py:478\u001b[0m, in \u001b[0;36m_TemporaryFileWrapper.__getattr__.<locals>.func_wrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m    476\u001b[0m \u001b[38;5;129m@_functools\u001b[39m\u001b[38;5;241m.\u001b[39mwraps(func)\n\u001b[1;32m    477\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mfunc_wrapper\u001b[39m(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs):\n\u001b[0;32m--> 478\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfunc\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "\u001b[0;31mOSError\u001b[0m: [Errno 28] No space left on device"
     ]
    }
   ],
   "source": [
    "image_summary_vqa_detector = ammico.SummaryDetector(image_dict, analysis_type=\"summary_and_questions\", \n",
    "                                                    model_type=\"base\")\n",
    "for num, key in tqdm(enumerate(image_dict.keys()),total=len(image_dict)):\n",
    "    image_dict[key] = image_summary_vqa_detector.analyse_image(subdict=image_dict[key], \n",
    "                                                               analysis_type=\"summary_and_questions\", \n",
    "                                                               list_of_questions = list_of_questions)\n",
    "    if num % dump_every == 0 | num == len(image_dict) - 1: \n",
    "        image_df = ammico.get_dataframe(image_dict)\n",
    "        image_df.to_csv(dump_file)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The output is given as a dictionary with the following keys and data types:\n",
    "\n",
    "| output key | output type | output value |\n",
    "| ---------- | ----------- | ------------ |\n",
    "| `const_image_summary` | `str` | when `analysis_type=\"summary\"` or `\"summary_and_questions\"`, constant image caption (does not change upon re-running the analysis for the same model) |\n",
    "| `3_non-deterministic_summary` | `list[str]` | when `analysis_type=\"summary\"` or s`ummary_and_questions`, three different captions generated with different random seeds |\n",
    "| *a user-defined input question* | `str` | when `analysis_type=\"questions\"` or `summary_and_questions`, the answer to the user-defined input question | \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### BLIP2 models\n",
    "This is very heavy models. They requare approx 60GB of RAM and they can use more than 20GB GPUs memory."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:53:18.184352Z",
     "iopub.status.busy": "2024-02-19T08:53:18.183915Z",
     "iopub.status.idle": "2024-02-19T08:53:49.441778Z",
     "shell.execute_reply": "2024-02-19T08:53:49.440627Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0.00/1.89G [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 4.01M/1.89G [00:00<01:16, 26.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  1%|          | 11.7M/1.89G [00:00<00:42, 47.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  1%|          | 16.6M/1.89G [00:00<00:48, 41.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  1%|          | 24.0M/1.89G [00:00<00:40, 48.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  2%|▏         | 32.0M/1.89G [00:00<00:37, 53.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  2%|▏         | 40.0M/1.89G [00:00<00:32, 61.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  2%|▏         | 48.0M/1.89G [00:00<00:30, 65.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  3%|▎         | 56.0M/1.89G [00:01<00:28, 69.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  3%|▎         | 64.6M/1.89G [00:01<00:26, 75.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  4%|▎         | 72.0M/1.89G [00:01<00:38, 50.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  4%|▍         | 80.0M/1.89G [00:01<00:33, 57.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  5%|▍         | 88.0M/1.89G [00:01<00:30, 63.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  5%|▌         | 101M/1.89G [00:01<00:23, 81.2MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  6%|▌         | 112M/1.89G [00:01<00:22, 85.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  6%|▋         | 122M/1.89G [00:01<00:20, 91.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  7%|▋         | 132M/1.89G [00:02<00:20, 92.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  7%|▋         | 142M/1.89G [00:02<00:19, 98.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  8%|▊         | 156M/1.89G [00:02<00:16, 110MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  9%|▊         | 168M/1.89G [00:02<00:15, 116MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 10%|▉         | 184M/1.89G [00:02<00:14, 125MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 10%|█         | 200M/1.89G [00:02<00:13, 132MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 11%|█         | 216M/1.89G [00:02<00:15, 117MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 12%|█▏        | 240M/1.89G [00:02<00:11, 152MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 13%|█▎        | 256M/1.89G [00:02<00:12, 142MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 14%|█▍        | 274M/1.89G [00:03<00:11, 154MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 15%|█▍        | 289M/1.89G [00:03<00:11, 155MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 16%|█▌        | 310M/1.89G [00:03<00:09, 173MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 17%|█▋        | 327M/1.89G [00:03<00:12, 133MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 18%|█▊        | 348M/1.89G [00:03<00:10, 153MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 19%|█▉        | 372M/1.89G [00:03<00:09, 178MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 21%|██        | 396M/1.89G [00:03<00:08, 198MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 22%|██▏       | 417M/1.89G [00:03<00:07, 203MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 23%|██▎       | 440M/1.89G [00:03<00:07, 203MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 24%|██▍       | 460M/1.89G [00:04<00:07, 204MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 25%|██▍       | 480M/1.89G [00:04<00:07, 201MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 26%|██▌       | 500M/1.89G [00:04<00:10, 143MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 27%|██▋       | 516M/1.89G [00:04<00:10, 140MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 28%|██▊       | 536M/1.89G [00:04<00:09, 153MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 29%|██▊       | 555M/1.89G [00:04<00:08, 165MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 30%|███       | 580M/1.89G [00:04<00:07, 190MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 31%|███       | 602M/1.89G [00:04<00:07, 199MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 32%|███▏      | 622M/1.89G [00:05<00:06, 201MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 33%|███▎      | 646M/1.89G [00:05<00:06, 215MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 35%|███▍      | 669M/1.89G [00:05<00:05, 223MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 36%|███▌      | 690M/1.89G [00:06<00:18, 71.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 37%|███▋      | 709M/1.89G [00:06<00:14, 86.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 38%|███▊      | 728M/1.89G [00:06<00:12, 103MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 39%|███▊      | 746M/1.89G [00:06<00:10, 117MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 40%|███▉      | 764M/1.89G [00:06<00:09, 128MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 41%|████      | 784M/1.89G [00:06<00:08, 145MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 42%|████▏     | 803M/1.89G [00:06<00:07, 158MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 43%|████▎     | 821M/1.89G [00:08<00:38, 30.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 43%|████▎     | 834M/1.89G [00:08<00:33, 34.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 44%|████▍     | 845M/1.89G [00:08<00:29, 39.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 44%|████▍     | 855M/1.89G [00:09<00:38, 29.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 45%|████▍     | 862M/1.89G [00:09<00:34, 32.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 45%|████▌     | 869M/1.89G [00:09<00:31, 35.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 45%|████▌     | 876M/1.89G [00:10<00:28, 39.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 46%|████▌     | 882M/1.89G [00:10<00:25, 43.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 46%|████▌     | 889M/1.89G [00:10<00:27, 39.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 46%|████▋     | 896M/1.89G [00:10<00:23, 46.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 47%|████▋     | 905M/1.89G [00:10<00:19, 55.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 47%|████▋     | 912M/1.89G [00:10<00:23, 45.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 48%|████▊     | 920M/1.89G [00:10<00:20, 51.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 48%|████▊     | 928M/1.89G [00:11<00:18, 55.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 48%|████▊     | 936M/1.89G [00:11<00:17, 59.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 49%|████▉     | 944M/1.89G [00:11<00:17, 60.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 49%|████▉     | 952M/1.89G [00:11<00:16, 63.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 50%|████▉     | 960M/1.89G [00:11<00:15, 66.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 51%|█████     | 976M/1.89G [00:11<00:11, 90.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 51%|█████▏    | 992M/1.89G [00:11<00:11, 83.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 52%|█████▏    | 0.98G/1.89G [00:11<00:09, 102MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 53%|█████▎    | 0.99G/1.89G [00:12<00:10, 90.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 53%|█████▎    | 1.01G/1.89G [00:12<00:09, 97.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 54%|█████▍    | 1.02G/1.89G [00:12<00:08, 113MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 55%|█████▍    | 1.04G/1.89G [00:13<00:19, 47.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 56%|█████▌    | 1.05G/1.89G [00:13<00:14, 63.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 56%|█████▋    | 1.06G/1.89G [00:13<00:15, 56.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 57%|█████▋    | 1.08G/1.89G [00:13<00:15, 57.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 58%|█████▊    | 1.09G/1.89G [00:13<00:11, 72.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 59%|█████▉    | 1.11G/1.89G [00:13<00:09, 89.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 60%|█████▉    | 1.13G/1.89G [00:13<00:07, 103MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 60%|██████    | 1.14G/1.89G [00:14<00:06, 118MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 61%|██████    | 1.15G/1.89G [00:14<00:15, 49.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 62%|██████▏   | 1.17G/1.89G [00:14<00:13, 57.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 62%|██████▏   | 1.18G/1.89G [00:14<00:11, 65.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 63%|██████▎   | 1.19G/1.89G [00:15<00:10, 71.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 63%|██████▎   | 1.20G/1.89G [00:15<00:17, 42.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 64%|██████▍   | 1.21G/1.89G [00:15<00:13, 54.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 65%|██████▍   | 1.22G/1.89G [00:16<00:19, 37.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 65%|██████▍   | 1.23G/1.89G [00:16<00:25, 27.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 65%|██████▌   | 1.23G/1.89G [00:16<00:08, 78.4MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "ename": "OSError",
     "evalue": "[Errno 28] No space left on device",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mOSError\u001b[0m                                   Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[23], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m obj \u001b[38;5;241m=\u001b[39m \u001b[43mammico\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mSummaryDetector\u001b[49m\u001b[43m(\u001b[49m\u001b[43msubdict\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mimage_dict\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43manalysis_type\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43m \u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43msummary_and_questions\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mmodel_type\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43m \u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mblip2_t5_caption_coco_flant5xl\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m)\u001b[49m\n\u001b[1;32m      2\u001b[0m \u001b[38;5;66;03m# list of the new models that can be used:\u001b[39;00m\n\u001b[1;32m      3\u001b[0m \u001b[38;5;66;03m# \"blip2_t5_pretrain_flant5xxl\",\u001b[39;00m\n\u001b[1;32m      4\u001b[0m \u001b[38;5;66;03m# \"blip2_t5_pretrain_flant5xl\",\u001b[39;00m\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m     14\u001b[0m \n\u001b[1;32m     15\u001b[0m \u001b[38;5;66;03m#also you can perform all calculation on cpu if you set device_type= \"cpu\" or gpu if you set device_type= \"cuda\"\u001b[39;00m\n",
      "File \u001b[0;32m~/work/AMMICO/AMMICO/ammico/summary.py:156\u001b[0m, in \u001b[0;36mSummaryDetector.__init__\u001b[0;34m(self, subdict, model_type, analysis_type, list_of_questions, summary_model, summary_vis_processors, summary_vqa_model, summary_vqa_vis_processors, summary_vqa_txt_processors, summary_vqa_model_new, summary_vqa_vis_processors_new, summary_vqa_txt_processors_new, device_type)\u001b[0m\n\u001b[1;32m    145\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msummary_vqa_txt_processors \u001b[38;5;241m=\u001b[39m summary_vqa_txt_processors\n\u001b[1;32m    146\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m (\n\u001b[1;32m    147\u001b[0m     model_type \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mallowed_new_model_types\n\u001b[1;32m    148\u001b[0m     \u001b[38;5;129;01mand\u001b[39;00m (summary_vqa_model_new \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m)\n\u001b[1;32m    149\u001b[0m     \u001b[38;5;129;01mand\u001b[39;00m (summary_vqa_vis_processors_new \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m)\n\u001b[1;32m    150\u001b[0m     \u001b[38;5;129;01mand\u001b[39;00m (summary_vqa_txt_processors_new \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m)\n\u001b[1;32m    151\u001b[0m ):\n\u001b[1;32m    152\u001b[0m     (\n\u001b[1;32m    153\u001b[0m         \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msummary_vqa_model_new,\n\u001b[1;32m    154\u001b[0m         \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msummary_vqa_vis_processors_new,\n\u001b[1;32m    155\u001b[0m         \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msummary_vqa_txt_processors_new,\n\u001b[0;32m--> 156\u001b[0m     ) \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mload_new_model\u001b[49m\u001b[43m(\u001b[49m\u001b[43mmodel_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mmodel_type\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    157\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m    158\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msummary_vqa_model_new \u001b[38;5;241m=\u001b[39m summary_vqa_model_new\n",
      "File \u001b[0;32m~/work/AMMICO/AMMICO/ammico/summary.py:479\u001b[0m, in \u001b[0;36mSummaryDetector.load_new_model\u001b[0;34m(self, model_type)\u001b[0m\n\u001b[1;32m    455\u001b[0m \u001b[38;5;250m\u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m    456\u001b[0m \u001b[38;5;124;03mLoad new BLIP2 models.\u001b[39;00m\n\u001b[1;32m    457\u001b[0m \n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    464\u001b[0m \u001b[38;5;124;03m    txt_processors (dict): preprocessors for text inputs.\u001b[39;00m\n\u001b[1;32m    465\u001b[0m \u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m    466\u001b[0m select_model \u001b[38;5;241m=\u001b[39m {\n\u001b[1;32m    467\u001b[0m     \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mblip2_t5_pretrain_flant5xxl\u001b[39m\u001b[38;5;124m\"\u001b[39m: SummaryDetector\u001b[38;5;241m.\u001b[39mload_model_blip2_t5_pretrain_flant5xxl,\n\u001b[1;32m    468\u001b[0m     \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mblip2_t5_pretrain_flant5xl\u001b[39m\u001b[38;5;124m\"\u001b[39m: SummaryDetector\u001b[38;5;241m.\u001b[39mload_model_blip2_t5_pretrain_flant5xl,\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    473\u001b[0m     \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mblip2_opt_caption_coco_opt6.7b\u001b[39m\u001b[38;5;124m\"\u001b[39m: SummaryDetector\u001b[38;5;241m.\u001b[39mload_model_base_blip2_opt_caption_coco_opt67b,\n\u001b[1;32m    474\u001b[0m }\n\u001b[1;32m    475\u001b[0m (\n\u001b[1;32m    476\u001b[0m     summary_vqa_model,\n\u001b[1;32m    477\u001b[0m     summary_vqa_vis_processors,\n\u001b[1;32m    478\u001b[0m     summary_vqa_txt_processors,\n\u001b[0;32m--> 479\u001b[0m ) \u001b[38;5;241m=\u001b[39m \u001b[43mselect_model\u001b[49m\u001b[43m[\u001b[49m\u001b[43mmodel_type\u001b[49m\u001b[43m]\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[43m)\u001b[49m\n\u001b[1;32m    480\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m summary_vqa_model, summary_vqa_vis_processors, summary_vqa_txt_processors\n",
      "File \u001b[0;32m~/work/AMMICO/AMMICO/ammico/summary.py:543\u001b[0m, in \u001b[0;36mSummaryDetector.load_model_blip2_t5_caption_coco_flant5xl\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    528\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mload_model_blip2_t5_caption_coco_flant5xl\u001b[39m(\u001b[38;5;28mself\u001b[39m):\n\u001b[1;32m    529\u001b[0m \u001b[38;5;250m    \u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m    530\u001b[0m \u001b[38;5;124;03m    Load BLIP2 model with caption_coco_flant5xl architecture.\u001b[39;00m\n\u001b[1;32m    531\u001b[0m \n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    537\u001b[0m \u001b[38;5;124;03m        txt_processors (dict): preprocessors for text inputs.\u001b[39;00m\n\u001b[1;32m    538\u001b[0m \u001b[38;5;124;03m    \"\"\"\u001b[39;00m\n\u001b[1;32m    539\u001b[0m     (\n\u001b[1;32m    540\u001b[0m         summary_vqa_model,\n\u001b[1;32m    541\u001b[0m         summary_vqa_vis_processors,\n\u001b[1;32m    542\u001b[0m         summary_vqa_txt_processors,\n\u001b[0;32m--> 543\u001b[0m     ) \u001b[38;5;241m=\u001b[39m \u001b[43mload_model_and_preprocess\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m    544\u001b[0m \u001b[43m        \u001b[49m\u001b[43mname\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mblip2_t5\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[1;32m    545\u001b[0m \u001b[43m        \u001b[49m\u001b[43mmodel_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mcaption_coco_flant5xl\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[1;32m    546\u001b[0m \u001b[43m        \u001b[49m\u001b[43mis_eval\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\u001b[43m,\u001b[49m\n\u001b[1;32m    547\u001b[0m \u001b[43m        \u001b[49m\u001b[43mdevice\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43msummary_device\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    548\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    549\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m summary_vqa_model, summary_vqa_vis_processors, summary_vqa_txt_processors\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/__init__.py:195\u001b[0m, in \u001b[0;36mload_model_and_preprocess\u001b[0;34m(name, model_type, is_eval, device)\u001b[0m\n\u001b[1;32m    192\u001b[0m model_cls \u001b[38;5;241m=\u001b[39m registry\u001b[38;5;241m.\u001b[39mget_model_class(name)\n\u001b[1;32m    194\u001b[0m \u001b[38;5;66;03m# load model\u001b[39;00m\n\u001b[0;32m--> 195\u001b[0m model \u001b[38;5;241m=\u001b[39m \u001b[43mmodel_cls\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfrom_pretrained\u001b[49m\u001b[43m(\u001b[49m\u001b[43mmodel_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mmodel_type\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    197\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m is_eval:\n\u001b[1;32m    198\u001b[0m     model\u001b[38;5;241m.\u001b[39meval()\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/base_model.py:70\u001b[0m, in \u001b[0;36mBaseModel.from_pretrained\u001b[0;34m(cls, model_type)\u001b[0m\n\u001b[1;32m     60\u001b[0m \u001b[38;5;250m\u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m     61\u001b[0m \u001b[38;5;124;03mBuild a pretrained model from default configuration file, specified by model_type.\u001b[39;00m\n\u001b[1;32m     62\u001b[0m \n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m     67\u001b[0m \u001b[38;5;124;03m    - model (nn.Module): pretrained or finetuned model, depending on the configuration.\u001b[39;00m\n\u001b[1;32m     68\u001b[0m \u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m     69\u001b[0m model_cfg \u001b[38;5;241m=\u001b[39m OmegaConf\u001b[38;5;241m.\u001b[39mload(\u001b[38;5;28mcls\u001b[39m\u001b[38;5;241m.\u001b[39mdefault_config_path(model_type))\u001b[38;5;241m.\u001b[39mmodel\n\u001b[0;32m---> 70\u001b[0m model \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mcls\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfrom_config\u001b[49m\u001b[43m(\u001b[49m\u001b[43mmodel_cfg\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     72\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m model\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/blip2_models/blip2_t5.py:368\u001b[0m, in \u001b[0;36mBlip2T5.from_config\u001b[0;34m(cls, cfg)\u001b[0m\n\u001b[1;32m    364\u001b[0m max_txt_len \u001b[38;5;241m=\u001b[39m cfg\u001b[38;5;241m.\u001b[39mget(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mmax_txt_len\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;241m32\u001b[39m)\n\u001b[1;32m    366\u001b[0m apply_lemmatizer \u001b[38;5;241m=\u001b[39m cfg\u001b[38;5;241m.\u001b[39mget(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mapply_lemmatizer\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;28;01mFalse\u001b[39;00m)\n\u001b[0;32m--> 368\u001b[0m model \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mcls\u001b[39;49m\u001b[43m(\u001b[49m\n\u001b[1;32m    369\u001b[0m \u001b[43m    \u001b[49m\u001b[43mvit_model\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mvit_model\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    370\u001b[0m \u001b[43m    \u001b[49m\u001b[43mimg_size\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mimg_size\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    371\u001b[0m \u001b[43m    \u001b[49m\u001b[43mdrop_path_rate\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mdrop_path_rate\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    372\u001b[0m \u001b[43m    \u001b[49m\u001b[43muse_grad_checkpoint\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43muse_grad_checkpoint\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    373\u001b[0m \u001b[43m    \u001b[49m\u001b[43mvit_precision\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mvit_precision\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    374\u001b[0m \u001b[43m    \u001b[49m\u001b[43mfreeze_vit\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mfreeze_vit\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    375\u001b[0m \u001b[43m    \u001b[49m\u001b[43mnum_query_token\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mnum_query_token\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    376\u001b[0m \u001b[43m    \u001b[49m\u001b[43mt5_model\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mt5_model\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    377\u001b[0m \u001b[43m    \u001b[49m\u001b[43mprompt\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mprompt\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    378\u001b[0m \u001b[43m    \u001b[49m\u001b[43mmax_txt_len\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mmax_txt_len\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    379\u001b[0m \u001b[43m    \u001b[49m\u001b[43mapply_lemmatizer\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mapply_lemmatizer\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    380\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    381\u001b[0m model\u001b[38;5;241m.\u001b[39mload_checkpoint_from_config(cfg)\n\u001b[1;32m    383\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m model\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/blip2_models/blip2_t5.py:61\u001b[0m, in \u001b[0;36mBlip2T5.__init__\u001b[0;34m(self, vit_model, img_size, drop_path_rate, use_grad_checkpoint, vit_precision, freeze_vit, num_query_token, t5_model, prompt, max_txt_len, apply_lemmatizer)\u001b[0m\n\u001b[1;32m     57\u001b[0m \u001b[38;5;28msuper\u001b[39m()\u001b[38;5;241m.\u001b[39m\u001b[38;5;21m__init__\u001b[39m()\n\u001b[1;32m     59\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mtokenizer \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39minit_tokenizer()\n\u001b[0;32m---> 61\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mvisual_encoder, \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mln_vision \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43minit_vision_encoder\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m     62\u001b[0m \u001b[43m    \u001b[49m\u001b[43mvit_model\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mimg_size\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mdrop_path_rate\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43muse_grad_checkpoint\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mvit_precision\u001b[49m\n\u001b[1;32m     63\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     64\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m freeze_vit:\n\u001b[1;32m     65\u001b[0m     \u001b[38;5;28;01mfor\u001b[39;00m name, param \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mvisual_encoder\u001b[38;5;241m.\u001b[39mnamed_parameters():\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/blip2_models/blip2.py:72\u001b[0m, in \u001b[0;36mBlip2Base.init_vision_encoder\u001b[0;34m(cls, model_name, img_size, drop_path_rate, use_grad_checkpoint, precision)\u001b[0m\n\u001b[1;32m     67\u001b[0m \u001b[38;5;28;01massert\u001b[39;00m model_name \u001b[38;5;129;01min\u001b[39;00m [\n\u001b[1;32m     68\u001b[0m     \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124meva_clip_g\u001b[39m\u001b[38;5;124m\"\u001b[39m,\n\u001b[1;32m     69\u001b[0m     \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mclip_L\u001b[39m\u001b[38;5;124m\"\u001b[39m,\n\u001b[1;32m     70\u001b[0m ], \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mvit model must be eva_clip_g or clip_L\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m     71\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m model_name \u001b[38;5;241m==\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124meva_clip_g\u001b[39m\u001b[38;5;124m\"\u001b[39m:\n\u001b[0;32m---> 72\u001b[0m     visual_encoder \u001b[38;5;241m=\u001b[39m \u001b[43mcreate_eva_vit_g\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m     73\u001b[0m \u001b[43m        \u001b[49m\u001b[43mimg_size\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mdrop_path_rate\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43muse_grad_checkpoint\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mprecision\u001b[49m\n\u001b[1;32m     74\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     75\u001b[0m \u001b[38;5;28;01melif\u001b[39;00m model_name \u001b[38;5;241m==\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mclip_L\u001b[39m\u001b[38;5;124m\"\u001b[39m:\n\u001b[1;32m     76\u001b[0m     visual_encoder \u001b[38;5;241m=\u001b[39m create_clip_vit_L(img_size, use_grad_checkpoint, precision)\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/eva_vit.py:430\u001b[0m, in \u001b[0;36mcreate_eva_vit_g\u001b[0;34m(img_size, drop_path_rate, use_checkpoint, precision)\u001b[0m\n\u001b[1;32m    416\u001b[0m model \u001b[38;5;241m=\u001b[39m VisionTransformer(\n\u001b[1;32m    417\u001b[0m     img_size\u001b[38;5;241m=\u001b[39mimg_size,\n\u001b[1;32m    418\u001b[0m     patch_size\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m14\u001b[39m,\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    427\u001b[0m     use_checkpoint\u001b[38;5;241m=\u001b[39muse_checkpoint,\n\u001b[1;32m    428\u001b[0m )  \n\u001b[1;32m    429\u001b[0m url \u001b[38;5;241m=\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mhttps://storage.googleapis.com/sfr-vision-language-research/LAVIS/models/BLIP2/eva_vit_g.pth\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[0;32m--> 430\u001b[0m cached_file \u001b[38;5;241m=\u001b[39m \u001b[43mdownload_cached_file\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m    431\u001b[0m \u001b[43m    \u001b[49m\u001b[43murl\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcheck_hash\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mFalse\u001b[39;49;00m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mprogress\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\n\u001b[1;32m    432\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    433\u001b[0m state_dict \u001b[38;5;241m=\u001b[39m torch\u001b[38;5;241m.\u001b[39mload(cached_file, map_location\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mcpu\u001b[39m\u001b[38;5;124m\"\u001b[39m)    \n\u001b[1;32m    434\u001b[0m interpolate_pos_embed(model,state_dict)\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/common/dist_utils.py:132\u001b[0m, in \u001b[0;36mdownload_cached_file\u001b[0;34m(url, check_hash, progress)\u001b[0m\n\u001b[1;32m    129\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m cached_file\n\u001b[1;32m    131\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m is_main_process():\n\u001b[0;32m--> 132\u001b[0m     \u001b[43mtimm_hub\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mdownload_cached_file\u001b[49m\u001b[43m(\u001b[49m\u001b[43murl\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcheck_hash\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mprogress\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    134\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m is_dist_avail_and_initialized():\n\u001b[1;32m    135\u001b[0m     dist\u001b[38;5;241m.\u001b[39mbarrier()\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/timm/models/hub.py:51\u001b[0m, in \u001b[0;36mdownload_cached_file\u001b[0;34m(url, check_hash, progress)\u001b[0m\n\u001b[1;32m     49\u001b[0m         r \u001b[38;5;241m=\u001b[39m HASH_REGEX\u001b[38;5;241m.\u001b[39msearch(filename)  \u001b[38;5;66;03m# r is Optional[Match[str]]\u001b[39;00m\n\u001b[1;32m     50\u001b[0m         hash_prefix \u001b[38;5;241m=\u001b[39m r\u001b[38;5;241m.\u001b[39mgroup(\u001b[38;5;241m1\u001b[39m) \u001b[38;5;28;01mif\u001b[39;00m r \u001b[38;5;28;01melse\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[0;32m---> 51\u001b[0m     \u001b[43mdownload_url_to_file\u001b[49m\u001b[43m(\u001b[49m\u001b[43murl\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcached_file\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mhash_prefix\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mprogress\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mprogress\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     52\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m cached_file\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/torch/hub.py:636\u001b[0m, in \u001b[0;36mdownload_url_to_file\u001b[0;34m(url, dst, hash_prefix, progress)\u001b[0m\n\u001b[1;32m    634\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mlen\u001b[39m(buffer) \u001b[38;5;241m==\u001b[39m \u001b[38;5;241m0\u001b[39m:\n\u001b[1;32m    635\u001b[0m     \u001b[38;5;28;01mbreak\u001b[39;00m\n\u001b[0;32m--> 636\u001b[0m \u001b[43mf\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mwrite\u001b[49m\u001b[43m(\u001b[49m\u001b[43mbuffer\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    637\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m hash_prefix \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m    638\u001b[0m     sha256\u001b[38;5;241m.\u001b[39mupdate(buffer)\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/tempfile.py:478\u001b[0m, in \u001b[0;36m_TemporaryFileWrapper.__getattr__.<locals>.func_wrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m    476\u001b[0m \u001b[38;5;129m@_functools\u001b[39m\u001b[38;5;241m.\u001b[39mwraps(func)\n\u001b[1;32m    477\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mfunc_wrapper\u001b[39m(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs):\n\u001b[0;32m--> 478\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfunc\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "\u001b[0;31mOSError\u001b[0m: [Errno 28] No space left on device"
     ]
    }
   ],
   "source": [
    "obj = ammico.SummaryDetector(subdict=image_dict, analysis_type = \"summary_and_questions\", model_type = \"blip2_t5_caption_coco_flant5xl\")\n",
    "# list of the new models that can be used:\n",
    "# \"blip2_t5_pretrain_flant5xxl\",\n",
    "# \"blip2_t5_pretrain_flant5xl\",\n",
    "# \"blip2_t5_caption_coco_flant5xl\",\n",
    "# \"blip2_opt_pretrain_opt2.7b\",\n",
    "# \"blip2_opt_pretrain_opt6.7b\",\n",
    "# \"blip2_opt_caption_coco_opt2.7b\",\n",
    "# \"blip2_opt_caption_coco_opt6.7b\",\n",
    "\n",
    "# You can use `pretrain_` model types for zero-shot image-to-text generation with prompts.\n",
    "# Or you can use `caption_coco_`` model types to generate coco-style captions.\n",
    "# `flant5` and `opt` means that the model equipped with FlanT5 and OPT LLMs respectively.\n",
    "\n",
    "#also you can perform all calculation on cpu if you set device_type= \"cpu\" or gpu if you set device_type= \"cuda\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:53:49.445792Z",
     "iopub.status.busy": "2024-02-19T08:53:49.445585Z",
     "iopub.status.idle": "2024-02-19T08:53:49.470611Z",
     "shell.execute_reply": "2024-02-19T08:53:49.470071Z"
    }
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'obj' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[24], line 2\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m key \u001b[38;5;129;01min\u001b[39;00m image_dict:\n\u001b[0;32m----> 2\u001b[0m     image_dict[key] \u001b[38;5;241m=\u001b[39m \u001b[43mobj\u001b[49m\u001b[38;5;241m.\u001b[39manalyse_image(subdict \u001b[38;5;241m=\u001b[39m image_dict[key], analysis_type\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124msummary_and_questions\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[1;32m      4\u001b[0m \u001b[38;5;66;03m# analysis_type can be \u001b[39;00m\n\u001b[1;32m      5\u001b[0m \u001b[38;5;66;03m# \"summary\",\u001b[39;00m\n\u001b[1;32m      6\u001b[0m \u001b[38;5;66;03m# \"questions\",\u001b[39;00m\n\u001b[1;32m      7\u001b[0m \u001b[38;5;66;03m# \"summary_and_questions\".\u001b[39;00m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'obj' is not defined"
     ]
    }
   ],
   "source": [
    "for key in image_dict:\n",
    "    image_dict[key] = obj.analyse_image(subdict = image_dict[key], analysis_type=\"summary_and_questions\")\n",
    "\n",
    "# analysis_type can be \n",
    "# \"summary\",\n",
    "# \"questions\",\n",
    "# \"summary_and_questions\"."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:53:49.473138Z",
     "iopub.status.busy": "2024-02-19T08:53:49.472942Z",
     "iopub.status.idle": "2024-02-19T08:53:49.480775Z",
     "shell.execute_reply": "2024-02-19T08:53:49.480208Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'img4': {'filename': 'data-test/img4.png',\n",
       "  'face': 'No',\n",
       "  'multiple_faces': 'No',\n",
       "  'no_faces': 0,\n",
       "  'wears_mask': ['No'],\n",
       "  'age': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'emotion': [None],\n",
       "  'emotion (category)': [None],\n",
       "  'text': 'MOODOVIN XI',\n",
       "  'text_language': 'en',\n",
       "  'text_english': 'MOODOVIN XI',\n",
       "  'text_clean': 'XI',\n",
       "  'text_summary': ' MOODOVIN XI XI: Vladimir Putin, Vladimir Vladmir Zelizer, Vladimir',\n",
       "  'sentiment': 'POSITIVE',\n",
       "  'sentiment_score': 0.66,\n",
       "  'entity': ['MOODOVIN XI'],\n",
       "  'entity_type': ['ORG'],\n",
       "  'const_image_summary': 'a river running through a city next to tall buildings',\n",
       "  '3_non-deterministic_summary': ['there is a pretty house that sits above the water',\n",
       "   'there is a building with a balcony and lots of plants on the side of it',\n",
       "   'several buildings with a river flowing through it']},\n",
       " 'img1': {'filename': 'data-test/img1.png',\n",
       "  'no_faces': 0,\n",
       "  'age': [None],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': [None],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'face': 'No',\n",
       "  'text': 'SCATTERING THEORY The Quantum Theory of Nonrelativistic Collisions JOHN R. TAYLOR University of Colorado',\n",
       "  'text_language': 'en',\n",
       "  'text_english': 'SCATTERING THEORY The Quantum Theory of Nonrelativistic Collisions JOHN R. TAYLOR University of Colorado',\n",
       "  'text_clean': 'THEORY The Quantum Theory of Collisions JOHN R. TAYLOR University of Colorado',\n",
       "  'text_summary': ' SCATTERING THEORY The Quantum Theory of Nonrelativistic Collisions',\n",
       "  'sentiment': 'POSITIVE',\n",
       "  'sentiment_score': 0.91,\n",
       "  'entity': ['Non',\n",
       "   '##vist',\n",
       "   'Col',\n",
       "   '##N',\n",
       "   'R',\n",
       "   'T',\n",
       "   '##AYL',\n",
       "   'University of Colorado'],\n",
       "  'entity_type': ['MISC', 'MISC', 'MISC', 'ORG', 'PER', 'PER', 'ORG', 'ORG'],\n",
       "  'const_image_summary': 'a close up of a piece of paper with writing on it',\n",
       "  '3_non-deterministic_summary': ['a book opened to the book title for a novel',\n",
       "   'there are many text on this page',\n",
       "   'the text in a book is a handwritten poem']},\n",
       " 'img2': {'filename': 'data-test/img2.png',\n",
       "  'no_faces': 0,\n",
       "  'age': [None],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': [None],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'face': 'No',\n",
       "  'text': 'THE ALGEBRAIC EIGENVALUE PROBLEM DOM NVS TIO MINA Monographs on Numerical Analysis J.. H. WILKINSON',\n",
       "  'text_language': 'en',\n",
       "  'text_english': 'THE ALGEBRAIC EIGENVALUE PROBLEM DOM NVS TIO MINA Monographs on Numerical Analysis J.. H. WILKINSON',\n",
       "  'text_clean': 'THE PROBLEM DOM NVS TIO MINA Monographs on Numerical Analysis J .. H. WILKINSON',\n",
       "  'text_summary': ' H. H. W. WILKINSON: The Algebri',\n",
       "  'sentiment': 'NEGATIVE',\n",
       "  'sentiment_score': 0.97,\n",
       "  'entity': ['ALGEBRAIC EIGENVAL', 'NVS TIO MI', 'J', 'H', 'WILKINSON'],\n",
       "  'entity_type': ['MISC', 'ORG', 'ORG', 'ORG', 'ORG'],\n",
       "  'const_image_summary': 'a yellow book with green lettering on it',\n",
       "  '3_non-deterministic_summary': ['a book cover with green writing on a black background',\n",
       "   'the title page of a book with information from its authors',\n",
       "   'a book about the age - related engineering and engineering']},\n",
       " 'img3': {'filename': 'data-test/img3.png',\n",
       "  'no_faces': 0,\n",
       "  'age': [None],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': [None],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'face': 'No',\n",
       "  'text': 'm OOOO 0000 www.',\n",
       "  'text_language': 'en',\n",
       "  'text_english': 'm OOOO 0000 www.',\n",
       "  'text_clean': 'm www .',\n",
       "  'text_summary': ' www. m OOOO 0000 0000 www.m.m OOOo 0000',\n",
       "  'sentiment': 'NEGATIVE',\n",
       "  'sentiment_score': 0.62,\n",
       "  'entity': [],\n",
       "  'entity_type': [],\n",
       "  'const_image_summary': 'a bus that is sitting on the side of a road',\n",
       "  '3_non-deterministic_summary': ['there are cars and a bus on the side of the road',\n",
       "   'a bus that is sitting in the middle of a street',\n",
       "   'an aerial view of an empty city street with two large buses passing by']},\n",
       " 'img0': {'filename': 'data-test/img0.png',\n",
       "  'no_faces': 0,\n",
       "  'age': [None],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': [None],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'face': 'No',\n",
       "  'text': 'Mathematische Formelsammlung für Ingenieure und Naturwissenschaftler Mit zahlreichen Abbildungen und Rechenbeispielen und einer ausführlichen Integraltafel 3., verbesserte Auflage',\n",
       "  'text_language': 'de',\n",
       "  'text_english': 'Mathematical formula collection for engineers and scientists With numerous illustrations and calculation examples and a detailed integral table 3rd, improved edition',\n",
       "  'text_clean': 'Mathematical formula collection for engineers and scientists With numerous illustrations and calculation examples and a detailed integral table 3rd , improved edition',\n",
       "  'text_summary': ' Mathematical formula collection for engineers and scientists . Includes numerous illustrations and calculation examples . Includes',\n",
       "  'sentiment': 'POSITIVE',\n",
       "  'sentiment_score': 1.0,\n",
       "  'entity': [],\n",
       "  'entity_type': [],\n",
       "  'const_image_summary': 'a close up of an open book with writing on it',\n",
       "  '3_non-deterministic_summary': ['a close up of a book with many languages',\n",
       "   'a book that is opened up in german',\n",
       "   'book about mathemarche formulals and their meaning']},\n",
       " 'img5': {'filename': 'data-test/img5.png',\n",
       "  'no_faces': 1,\n",
       "  'age': [26],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': ['Negative'],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': ['sad'],\n",
       "  'gender': ['Man'],\n",
       "  'race': [None],\n",
       "  'face': 'Yes',\n",
       "  'text': None,\n",
       "  'text_language': 'en',\n",
       "  'text_english': '',\n",
       "  'text_clean': '',\n",
       "  'text_summary': ' CNN.com will feature iReporter photos in a weekly Travel Snapshots gallery .',\n",
       "  'sentiment': 'POSITIVE',\n",
       "  'sentiment_score': 0.75,\n",
       "  'entity': [],\n",
       "  'entity_type': [],\n",
       "  'const_image_summary': 'a person running on a beach near a rock formation',\n",
       "  '3_non-deterministic_summary': ['a woman is running down the beach next to some rocks',\n",
       "   'a woman running along the beach by the ocean',\n",
       "   'there is a person running on the beach next to the ocean']}}"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "image_dict"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can also pass a list of questions to this cell if `analysis_type=\"summary_and_questions\"` or `analysis_type=\"questions\"`. But the format of questions has changed in new models. \n",
    "\n",
    "Here is an example of a list of questions:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:53:49.483664Z",
     "iopub.status.busy": "2024-02-19T08:53:49.483466Z",
     "iopub.status.idle": "2024-02-19T08:53:49.486232Z",
     "shell.execute_reply": "2024-02-19T08:53:49.485670Z"
    }
   },
   "outputs": [],
   "source": [
    "list_of_questions = [\n",
    "    \"Question: Are there people in the image? Answer:\",\n",
    "    \"Question: What is this picture about? Answer:\",\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:53:49.488981Z",
     "iopub.status.busy": "2024-02-19T08:53:49.488787Z",
     "iopub.status.idle": "2024-02-19T08:53:49.509942Z",
     "shell.execute_reply": "2024-02-19T08:53:49.509438Z"
    }
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'obj' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[27], line 2\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m key \u001b[38;5;129;01min\u001b[39;00m image_dict:\n\u001b[0;32m----> 2\u001b[0m     image_dict[key] \u001b[38;5;241m=\u001b[39m \u001b[43mobj\u001b[49m\u001b[38;5;241m.\u001b[39manalyse_image(subdict \u001b[38;5;241m=\u001b[39m image_dict[key], analysis_type\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mquestions\u001b[39m\u001b[38;5;124m\"\u001b[39m, list_of_questions\u001b[38;5;241m=\u001b[39mlist_of_questions)\n",
      "\u001b[0;31mNameError\u001b[0m: name 'obj' is not defined"
     ]
    }
   ],
   "source": [
    "for key in image_dict:\n",
    "    image_dict[key] = obj.analyse_image(subdict = image_dict[key], analysis_type=\"questions\", list_of_questions=list_of_questions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can also pass a question with previous answers as context into this model and pass in questions like this one to get a more accurate answer:\n",
    "\n",
    "You can combine as many questions as you want in a single query as a list."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:53:49.513195Z",
     "iopub.status.busy": "2024-02-19T08:53:49.512999Z",
     "iopub.status.idle": "2024-02-19T08:53:49.515828Z",
     "shell.execute_reply": "2024-02-19T08:53:49.515286Z"
    }
   },
   "outputs": [],
   "source": [
    "list_of_questions = [\n",
    "    \"Question: What country is in the picture? Answer: USA. Question: Why? Answer: Because there is an American flag in the background . Question: Where it comes from? Answer:\",\n",
    "    \"Question: Which city is this? Answer: Frankfurt. Question: Why?\",\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:53:49.520071Z",
     "iopub.status.busy": "2024-02-19T08:53:49.519734Z",
     "iopub.status.idle": "2024-02-19T08:53:49.540462Z",
     "shell.execute_reply": "2024-02-19T08:53:49.539940Z"
    }
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'obj' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[29], line 2\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m key \u001b[38;5;129;01min\u001b[39;00m image_dict:\n\u001b[0;32m----> 2\u001b[0m     image_dict[key] \u001b[38;5;241m=\u001b[39m \u001b[43mobj\u001b[49m\u001b[38;5;241m.\u001b[39manalyse_image(subdict \u001b[38;5;241m=\u001b[39m image_dict[key], analysis_type\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mquestions\u001b[39m\u001b[38;5;124m\"\u001b[39m, list_of_questions\u001b[38;5;241m=\u001b[39mlist_of_questions)\n",
      "\u001b[0;31mNameError\u001b[0m: name 'obj' is not defined"
     ]
    }
   ],
   "source": [
    "for key in image_dict:\n",
    "    image_dict[key] = obj.analyse_image(subdict = image_dict[key], analysis_type=\"questions\", list_of_questions=list_of_questions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:53:49.544909Z",
     "iopub.status.busy": "2024-02-19T08:53:49.544431Z",
     "iopub.status.idle": "2024-02-19T08:53:49.551819Z",
     "shell.execute_reply": "2024-02-19T08:53:49.551270Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'img4': {'filename': 'data-test/img4.png',\n",
       "  'face': 'No',\n",
       "  'multiple_faces': 'No',\n",
       "  'no_faces': 0,\n",
       "  'wears_mask': ['No'],\n",
       "  'age': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'emotion': [None],\n",
       "  'emotion (category)': [None],\n",
       "  'text': 'MOODOVIN XI',\n",
       "  'text_language': 'en',\n",
       "  'text_english': 'MOODOVIN XI',\n",
       "  'text_clean': 'XI',\n",
       "  'text_summary': ' MOODOVIN XI XI: Vladimir Putin, Vladimir Vladmir Zelizer, Vladimir',\n",
       "  'sentiment': 'POSITIVE',\n",
       "  'sentiment_score': 0.66,\n",
       "  'entity': ['MOODOVIN XI'],\n",
       "  'entity_type': ['ORG'],\n",
       "  'const_image_summary': 'a river running through a city next to tall buildings',\n",
       "  '3_non-deterministic_summary': ['there is a pretty house that sits above the water',\n",
       "   'there is a building with a balcony and lots of plants on the side of it',\n",
       "   'several buildings with a river flowing through it']},\n",
       " 'img1': {'filename': 'data-test/img1.png',\n",
       "  'no_faces': 0,\n",
       "  'age': [None],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': [None],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'face': 'No',\n",
       "  'text': 'SCATTERING THEORY The Quantum Theory of Nonrelativistic Collisions JOHN R. TAYLOR University of Colorado',\n",
       "  'text_language': 'en',\n",
       "  'text_english': 'SCATTERING THEORY The Quantum Theory of Nonrelativistic Collisions JOHN R. TAYLOR University of Colorado',\n",
       "  'text_clean': 'THEORY The Quantum Theory of Collisions JOHN R. TAYLOR University of Colorado',\n",
       "  'text_summary': ' SCATTERING THEORY The Quantum Theory of Nonrelativistic Collisions',\n",
       "  'sentiment': 'POSITIVE',\n",
       "  'sentiment_score': 0.91,\n",
       "  'entity': ['Non',\n",
       "   '##vist',\n",
       "   'Col',\n",
       "   '##N',\n",
       "   'R',\n",
       "   'T',\n",
       "   '##AYL',\n",
       "   'University of Colorado'],\n",
       "  'entity_type': ['MISC', 'MISC', 'MISC', 'ORG', 'PER', 'PER', 'ORG', 'ORG'],\n",
       "  'const_image_summary': 'a close up of a piece of paper with writing on it',\n",
       "  '3_non-deterministic_summary': ['a book opened to the book title for a novel',\n",
       "   'there are many text on this page',\n",
       "   'the text in a book is a handwritten poem']},\n",
       " 'img2': {'filename': 'data-test/img2.png',\n",
       "  'no_faces': 0,\n",
       "  'age': [None],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': [None],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'face': 'No',\n",
       "  'text': 'THE ALGEBRAIC EIGENVALUE PROBLEM DOM NVS TIO MINA Monographs on Numerical Analysis J.. H. WILKINSON',\n",
       "  'text_language': 'en',\n",
       "  'text_english': 'THE ALGEBRAIC EIGENVALUE PROBLEM DOM NVS TIO MINA Monographs on Numerical Analysis J.. H. WILKINSON',\n",
       "  'text_clean': 'THE PROBLEM DOM NVS TIO MINA Monographs on Numerical Analysis J .. H. WILKINSON',\n",
       "  'text_summary': ' H. H. W. WILKINSON: The Algebri',\n",
       "  'sentiment': 'NEGATIVE',\n",
       "  'sentiment_score': 0.97,\n",
       "  'entity': ['ALGEBRAIC EIGENVAL', 'NVS TIO MI', 'J', 'H', 'WILKINSON'],\n",
       "  'entity_type': ['MISC', 'ORG', 'ORG', 'ORG', 'ORG'],\n",
       "  'const_image_summary': 'a yellow book with green lettering on it',\n",
       "  '3_non-deterministic_summary': ['a book cover with green writing on a black background',\n",
       "   'the title page of a book with information from its authors',\n",
       "   'a book about the age - related engineering and engineering']},\n",
       " 'img3': {'filename': 'data-test/img3.png',\n",
       "  'no_faces': 0,\n",
       "  'age': [None],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': [None],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'face': 'No',\n",
       "  'text': 'm OOOO 0000 www.',\n",
       "  'text_language': 'en',\n",
       "  'text_english': 'm OOOO 0000 www.',\n",
       "  'text_clean': 'm www .',\n",
       "  'text_summary': ' www. m OOOO 0000 0000 www.m.m OOOo 0000',\n",
       "  'sentiment': 'NEGATIVE',\n",
       "  'sentiment_score': 0.62,\n",
       "  'entity': [],\n",
       "  'entity_type': [],\n",
       "  'const_image_summary': 'a bus that is sitting on the side of a road',\n",
       "  '3_non-deterministic_summary': ['there are cars and a bus on the side of the road',\n",
       "   'a bus that is sitting in the middle of a street',\n",
       "   'an aerial view of an empty city street with two large buses passing by']},\n",
       " 'img0': {'filename': 'data-test/img0.png',\n",
       "  'no_faces': 0,\n",
       "  'age': [None],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': [None],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'face': 'No',\n",
       "  'text': 'Mathematische Formelsammlung für Ingenieure und Naturwissenschaftler Mit zahlreichen Abbildungen und Rechenbeispielen und einer ausführlichen Integraltafel 3., verbesserte Auflage',\n",
       "  'text_language': 'de',\n",
       "  'text_english': 'Mathematical formula collection for engineers and scientists With numerous illustrations and calculation examples and a detailed integral table 3rd, improved edition',\n",
       "  'text_clean': 'Mathematical formula collection for engineers and scientists With numerous illustrations and calculation examples and a detailed integral table 3rd , improved edition',\n",
       "  'text_summary': ' Mathematical formula collection for engineers and scientists . Includes numerous illustrations and calculation examples . Includes',\n",
       "  'sentiment': 'POSITIVE',\n",
       "  'sentiment_score': 1.0,\n",
       "  'entity': [],\n",
       "  'entity_type': [],\n",
       "  'const_image_summary': 'a close up of an open book with writing on it',\n",
       "  '3_non-deterministic_summary': ['a close up of a book with many languages',\n",
       "   'a book that is opened up in german',\n",
       "   'book about mathemarche formulals and their meaning']},\n",
       " 'img5': {'filename': 'data-test/img5.png',\n",
       "  'no_faces': 1,\n",
       "  'age': [26],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': ['Negative'],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': ['sad'],\n",
       "  'gender': ['Man'],\n",
       "  'race': [None],\n",
       "  'face': 'Yes',\n",
       "  'text': None,\n",
       "  'text_language': 'en',\n",
       "  'text_english': '',\n",
       "  'text_clean': '',\n",
       "  'text_summary': ' CNN.com will feature iReporter photos in a weekly Travel Snapshots gallery .',\n",
       "  'sentiment': 'POSITIVE',\n",
       "  'sentiment_score': 0.75,\n",
       "  'entity': [],\n",
       "  'entity_type': [],\n",
       "  'const_image_summary': 'a person running on a beach near a rock formation',\n",
       "  '3_non-deterministic_summary': ['a woman is running down the beach next to some rocks',\n",
       "   'a woman running along the beach by the ocean',\n",
       "   'there is a person running on the beach next to the ocean']}}"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "image_dict"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can also ask sequential questions if you pass the argument `cosequential_questions=True`. This means that the answers to previous questions will be passed as context to the next question. However, this method will work a bit slower, because for each image the answers to the questions will not be calculated simultaneously, but sequentially. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:53:49.554718Z",
     "iopub.status.busy": "2024-02-19T08:53:49.554288Z",
     "iopub.status.idle": "2024-02-19T08:53:49.557147Z",
     "shell.execute_reply": "2024-02-19T08:53:49.556606Z"
    }
   },
   "outputs": [],
   "source": [
    "list_of_questions = [\n",
    "    \"Question: Is this picture taken inside or outside? Answer:\",\n",
    "    \"Question: Why? Answer:\",\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:53:49.559854Z",
     "iopub.status.busy": "2024-02-19T08:53:49.559416Z",
     "iopub.status.idle": "2024-02-19T08:53:49.581420Z",
     "shell.execute_reply": "2024-02-19T08:53:49.580718Z"
    }
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'obj' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[32], line 2\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m key \u001b[38;5;129;01min\u001b[39;00m image_dict:\n\u001b[0;32m----> 2\u001b[0m     image_dict[key] \u001b[38;5;241m=\u001b[39m \u001b[43mobj\u001b[49m\u001b[38;5;241m.\u001b[39manalyse_image(subdict \u001b[38;5;241m=\u001b[39m image_dict[key], analysis_type\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mquestions\u001b[39m\u001b[38;5;124m\"\u001b[39m, list_of_questions\u001b[38;5;241m=\u001b[39mlist_of_questions, consequential_questions\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mTrue\u001b[39;00m)\n",
      "\u001b[0;31mNameError\u001b[0m: name 'obj' is not defined"
     ]
    }
   ],
   "source": [
    "for key in image_dict:\n",
    "    image_dict[key] = obj.analyse_image(subdict = image_dict[key], analysis_type=\"questions\", list_of_questions=list_of_questions, consequential_questions=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:53:49.584997Z",
     "iopub.status.busy": "2024-02-19T08:53:49.584485Z",
     "iopub.status.idle": "2024-02-19T08:53:49.592760Z",
     "shell.execute_reply": "2024-02-19T08:53:49.592039Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'img4': {'filename': 'data-test/img4.png',\n",
       "  'face': 'No',\n",
       "  'multiple_faces': 'No',\n",
       "  'no_faces': 0,\n",
       "  'wears_mask': ['No'],\n",
       "  'age': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'emotion': [None],\n",
       "  'emotion (category)': [None],\n",
       "  'text': 'MOODOVIN XI',\n",
       "  'text_language': 'en',\n",
       "  'text_english': 'MOODOVIN XI',\n",
       "  'text_clean': 'XI',\n",
       "  'text_summary': ' MOODOVIN XI XI: Vladimir Putin, Vladimir Vladmir Zelizer, Vladimir',\n",
       "  'sentiment': 'POSITIVE',\n",
       "  'sentiment_score': 0.66,\n",
       "  'entity': ['MOODOVIN XI'],\n",
       "  'entity_type': ['ORG'],\n",
       "  'const_image_summary': 'a river running through a city next to tall buildings',\n",
       "  '3_non-deterministic_summary': ['there is a pretty house that sits above the water',\n",
       "   'there is a building with a balcony and lots of plants on the side of it',\n",
       "   'several buildings with a river flowing through it']},\n",
       " 'img1': {'filename': 'data-test/img1.png',\n",
       "  'no_faces': 0,\n",
       "  'age': [None],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': [None],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'face': 'No',\n",
       "  'text': 'SCATTERING THEORY The Quantum Theory of Nonrelativistic Collisions JOHN R. TAYLOR University of Colorado',\n",
       "  'text_language': 'en',\n",
       "  'text_english': 'SCATTERING THEORY The Quantum Theory of Nonrelativistic Collisions JOHN R. TAYLOR University of Colorado',\n",
       "  'text_clean': 'THEORY The Quantum Theory of Collisions JOHN R. TAYLOR University of Colorado',\n",
       "  'text_summary': ' SCATTERING THEORY The Quantum Theory of Nonrelativistic Collisions',\n",
       "  'sentiment': 'POSITIVE',\n",
       "  'sentiment_score': 0.91,\n",
       "  'entity': ['Non',\n",
       "   '##vist',\n",
       "   'Col',\n",
       "   '##N',\n",
       "   'R',\n",
       "   'T',\n",
       "   '##AYL',\n",
       "   'University of Colorado'],\n",
       "  'entity_type': ['MISC', 'MISC', 'MISC', 'ORG', 'PER', 'PER', 'ORG', 'ORG'],\n",
       "  'const_image_summary': 'a close up of a piece of paper with writing on it',\n",
       "  '3_non-deterministic_summary': ['a book opened to the book title for a novel',\n",
       "   'there are many text on this page',\n",
       "   'the text in a book is a handwritten poem']},\n",
       " 'img2': {'filename': 'data-test/img2.png',\n",
       "  'no_faces': 0,\n",
       "  'age': [None],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': [None],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'face': 'No',\n",
       "  'text': 'THE ALGEBRAIC EIGENVALUE PROBLEM DOM NVS TIO MINA Monographs on Numerical Analysis J.. H. WILKINSON',\n",
       "  'text_language': 'en',\n",
       "  'text_english': 'THE ALGEBRAIC EIGENVALUE PROBLEM DOM NVS TIO MINA Monographs on Numerical Analysis J.. H. WILKINSON',\n",
       "  'text_clean': 'THE PROBLEM DOM NVS TIO MINA Monographs on Numerical Analysis J .. H. WILKINSON',\n",
       "  'text_summary': ' H. H. W. WILKINSON: The Algebri',\n",
       "  'sentiment': 'NEGATIVE',\n",
       "  'sentiment_score': 0.97,\n",
       "  'entity': ['ALGEBRAIC EIGENVAL', 'NVS TIO MI', 'J', 'H', 'WILKINSON'],\n",
       "  'entity_type': ['MISC', 'ORG', 'ORG', 'ORG', 'ORG'],\n",
       "  'const_image_summary': 'a yellow book with green lettering on it',\n",
       "  '3_non-deterministic_summary': ['a book cover with green writing on a black background',\n",
       "   'the title page of a book with information from its authors',\n",
       "   'a book about the age - related engineering and engineering']},\n",
       " 'img3': {'filename': 'data-test/img3.png',\n",
       "  'no_faces': 0,\n",
       "  'age': [None],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': [None],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'face': 'No',\n",
       "  'text': 'm OOOO 0000 www.',\n",
       "  'text_language': 'en',\n",
       "  'text_english': 'm OOOO 0000 www.',\n",
       "  'text_clean': 'm www .',\n",
       "  'text_summary': ' www. m OOOO 0000 0000 www.m.m OOOo 0000',\n",
       "  'sentiment': 'NEGATIVE',\n",
       "  'sentiment_score': 0.62,\n",
       "  'entity': [],\n",
       "  'entity_type': [],\n",
       "  'const_image_summary': 'a bus that is sitting on the side of a road',\n",
       "  '3_non-deterministic_summary': ['there are cars and a bus on the side of the road',\n",
       "   'a bus that is sitting in the middle of a street',\n",
       "   'an aerial view of an empty city street with two large buses passing by']},\n",
       " 'img0': {'filename': 'data-test/img0.png',\n",
       "  'no_faces': 0,\n",
       "  'age': [None],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': [None],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'face': 'No',\n",
       "  'text': 'Mathematische Formelsammlung für Ingenieure und Naturwissenschaftler Mit zahlreichen Abbildungen und Rechenbeispielen und einer ausführlichen Integraltafel 3., verbesserte Auflage',\n",
       "  'text_language': 'de',\n",
       "  'text_english': 'Mathematical formula collection for engineers and scientists With numerous illustrations and calculation examples and a detailed integral table 3rd, improved edition',\n",
       "  'text_clean': 'Mathematical formula collection for engineers and scientists With numerous illustrations and calculation examples and a detailed integral table 3rd , improved edition',\n",
       "  'text_summary': ' Mathematical formula collection for engineers and scientists . Includes numerous illustrations and calculation examples . Includes',\n",
       "  'sentiment': 'POSITIVE',\n",
       "  'sentiment_score': 1.0,\n",
       "  'entity': [],\n",
       "  'entity_type': [],\n",
       "  'const_image_summary': 'a close up of an open book with writing on it',\n",
       "  '3_non-deterministic_summary': ['a close up of a book with many languages',\n",
       "   'a book that is opened up in german',\n",
       "   'book about mathemarche formulals and their meaning']},\n",
       " 'img5': {'filename': 'data-test/img5.png',\n",
       "  'no_faces': 1,\n",
       "  'age': [26],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': ['Negative'],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': ['sad'],\n",
       "  'gender': ['Man'],\n",
       "  'race': [None],\n",
       "  'face': 'Yes',\n",
       "  'text': None,\n",
       "  'text_language': 'en',\n",
       "  'text_english': '',\n",
       "  'text_clean': '',\n",
       "  'text_summary': ' CNN.com will feature iReporter photos in a weekly Travel Snapshots gallery .',\n",
       "  'sentiment': 'POSITIVE',\n",
       "  'sentiment_score': 0.75,\n",
       "  'entity': [],\n",
       "  'entity_type': [],\n",
       "  'const_image_summary': 'a person running on a beach near a rock formation',\n",
       "  '3_non-deterministic_summary': ['a woman is running down the beach next to some rocks',\n",
       "   'a woman running along the beach by the ocean',\n",
       "   'there is a person running on the beach next to the ocean']}}"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "image_dict"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Detection of faces and facial expression analysis\n",
    "Faces and facial expressions are detected and analyzed using the `EmotionDetector` class from the `faces` module. Initially, it is detected if faces are present on the image using RetinaFace, followed by analysis if face masks are worn (Face-Mask-Detection). The detection of age, gender, race, and emotions is carried out with deepface.\n",
    "\n",
    "<img src=\"../_static/emotion_detector.png\" width=\"800\" />\n",
    "\n",
    "Depending on the features found on the image, the face detection module returns a different analysis content: If no faces are found on the image, all further steps are skipped and the result `\"face\": \"No\", \"multiple_faces\": \"No\", \"no_faces\": 0, \"wears_mask\": [\"No\"], \"age\": [None], \"gender\": [None], \"race\": [None], \"emotion\": [None], \"emotion (category)\": [None]` is returned. If one or several faces are found, up to three faces are analyzed if they are partially concealed by a face mask. If yes, only age and gender are detected; if no, also race, emotion, and dominant emotion are detected. In case of the latter, the output could look like this: `\"face\": \"Yes\", \"multiple_faces\": \"Yes\", \"no_faces\": 2, \"wears_mask\": [\"No\", \"No\"], \"age\": [27, 28], \"gender\": [\"Man\", \"Man\"], \"race\": [\"asian\", None], \"emotion\": [\"angry\", \"neutral\"], \"emotion (category)\": [\"Negative\", \"Neutral\"]`, where for the two faces that are detected (given by `no_faces`), some of the values are returned as a list with the first item for the first (largest) face and the second item for the second (smaller) face (for example, `\"emotion\"` returns a list `[\"angry\", \"neutral\"]` signifying the first face expressing anger, and the second face having a neutral expression).\n",
    "\n",
    "The emotion detection reports the seven facial expressions angry, fear, neutral, sad, disgust, happy and surprise. These emotions are assigned based on the returned confidence of the model (between 0 and 1), with a high confidence signifying a high likelihood of the detected emotion being correct. Emotion recognition is not an easy task, even for a human; therefore, we have added a keyword `emotion_threshold` signifying the % value above which an emotion is counted as being detected. The default is set to 50%, so that a confidence above 0.5 results in an emotion being assigned. If the confidence is lower, no emotion is assigned. \n",
    "\n",
    "From the seven facial expressions, an overall dominating emotion category is identified: negative, positive, or neutral emotion. These are defined with the facial expressions angry, disgust, fear and sad for the negative category, happy for the positive category, and surprise and neutral for the neutral category.\n",
    "\n",
    "A similar threshold as for the emotion recognition is set for the race detection, `race_threshold`, with the default set to 50% so that a confidence for the race above 0.5 only will return a value in the analysis. \n",
    "\n",
    "Summarizing, the face detection is carried out using the following method call and keywords, where `emotion_threshold` and \n",
    "`race_threshold` are optional:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:53:49.596180Z",
     "iopub.status.busy": "2024-02-19T08:53:49.595807Z",
     "iopub.status.idle": "2024-02-19T08:54:11.388417Z",
     "shell.execute_reply": "2024-02-19T08:54:11.387832Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "1/1 [==============================] - ETA: 0s"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r",
      "1/1 [==============================] - 1s 535ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "1/1 [==============================] - ETA: 0s"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r",
      "1/1 [==============================] - 0s 343ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "1/1 [==============================] - ETA: 0s"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r",
      "1/1 [==============================] - 0s 226ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "1/1 [==============================] - ETA: 0s"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r",
      "1/1 [==============================] - 0s 233ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "1/1 [==============================] - ETA: 0s"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r",
      "1/1 [==============================] - 0s 21ms/step\n"
     ]
    }
   ],
   "source": [
    "for key in image_dict.keys():\n",
    "    image_dict[key] = ammico.EmotionDetector(image_dict[key], emotion_threshold=50, race_threshold=50).analyse_image()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The thresholds can be adapted interactively in the notebook interface and the optimal value can then be used in a subsequent analysis of the whole data set.\n",
    "\n",
    "The output keys that are generated are\n",
    "\n",
    "| output key | output type | output value |\n",
    "| ---------- | ----------- | ------------ |\n",
    "| `face` | `str` | if a face is detected |\n",
    "| `multiple_faces` | `str` | if multiple faces are detected |\n",
    "| `no_faces` | `int` | the number of detected faces |\n",
    "| `wears_mask` | `list[str]` | if each of the detected faces wears a face covering, up to three faces |\n",
    "| `age` | `list[int]` | the detected age, up to three faces |\n",
    "| `gender` | `list[str]` | the detected gender, up to three faces |\n",
    "| `race` | `list[str]` | the detected race, up to three faces, if above the confidence threshold |\n",
    "| `emotion` | `list[str]` | the detected emotion, up to three faces, if above the confidence threshold |\n",
    "| `emotion (category)` | `list[str]` | the detected emotion category (positive, negative, or neutral), up to three faces, if above the confidence threshold |"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Image Multimodal Search"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This module shows how to carry out an image multimodal search with the [LAVIS](https://github.com/salesforce/LAVIS) library. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Indexing and extracting features from images in selected folder"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First you need to select a model. You can choose one of the following models: \n",
    "- [blip](https://github.com/salesforce/BLIP)\n",
    "- [blip2](https://huggingface.co/docs/transformers/main/model_doc/blip-2) \n",
    "- [albef](https://github.com/salesforce/ALBEF) \n",
    "- [clip_base](https://github.com/openai/CLIP/blob/main/model-card.md)\n",
    "- [clip_vitl14](https://github.com/mlfoundations/open_clip) \n",
    "- [clip_vitl14_336](https://github.com/mlfoundations/open_clip)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:11.393676Z",
     "iopub.status.busy": "2024-02-19T08:54:11.393305Z",
     "iopub.status.idle": "2024-02-19T08:54:11.396416Z",
     "shell.execute_reply": "2024-02-19T08:54:11.395833Z"
    }
   },
   "outputs": [],
   "source": [
    "model_type = \"blip\"\n",
    "# model_type = \"blip2\"\n",
    "# model_type = \"albef\"\n",
    "# model_type = \"clip_base\"\n",
    "# model_type = \"clip_vitl14\"\n",
    "# model_type = \"clip_vitl14_336\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To process the loaded images using the selected model, use the below code:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:11.399454Z",
     "iopub.status.busy": "2024-02-19T08:54:11.399132Z",
     "iopub.status.idle": "2024-02-19T08:54:11.402161Z",
     "shell.execute_reply": "2024-02-19T08:54:11.401648Z"
    }
   },
   "outputs": [],
   "source": [
    "my_obj = ammico.MultimodalSearch(image_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:11.404913Z",
     "iopub.status.busy": "2024-02-19T08:54:11.404591Z",
     "iopub.status.idle": "2024-02-19T08:54:26.394925Z",
     "shell.execute_reply": "2024-02-19T08:54:26.394163Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0.00/1.97G [00:00<?, ?B/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 9.95M/1.97G [00:00<00:20, 104MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  1%|▏         | 25.5M/1.97G [00:00<00:15, 139MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  2%|▏         | 38.7M/1.97G [00:00<00:22, 91.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  2%|▏         | 48.7M/1.97G [00:00<00:33, 62.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  3%|▎         | 56.2M/1.97G [00:00<00:36, 56.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  3%|▎         | 64.0M/1.97G [00:00<00:33, 60.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  4%|▎         | 72.0M/1.97G [00:01<00:31, 65.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  4%|▍         | 80.0M/1.97G [00:01<00:29, 70.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  4%|▍         | 90.6M/1.97G [00:01<00:25, 80.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  5%|▍         | 98.9M/1.97G [00:01<00:25, 79.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  6%|▌         | 112M/1.97G [00:01<00:22, 88.0MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  6%|▋         | 128M/1.97G [00:01<00:20, 98.8MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  7%|▋         | 144M/1.97G [00:01<00:17, 115MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  8%|▊         | 160M/1.97G [00:01<00:16, 115MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  9%|▊         | 176M/1.97G [00:02<00:15, 122MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  9%|▉         | 188M/1.97G [00:02<00:16, 118MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 10%|█         | 205M/1.97G [00:02<00:14, 134MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 11%|█         | 224M/1.97G [00:02<00:15, 118MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 12%|█▏        | 240M/1.97G [00:02<00:14, 128MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 13%|█▎        | 256M/1.97G [00:02<00:14, 129MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 14%|█▍        | 280M/1.97G [00:02<00:11, 153MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 15%|█▍        | 300M/1.97G [00:02<00:10, 167MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 16%|█▌        | 320M/1.97G [00:03<00:10, 168MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 17%|█▋        | 344M/1.97G [00:03<00:09, 184MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 18%|█▊        | 362M/1.97G [00:03<00:10, 170MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 19%|█▉        | 379M/1.97G [00:03<00:10, 159MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 20%|█▉        | 394M/1.97G [00:03<00:11, 154MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 20%|██        | 409M/1.97G [00:03<00:11, 144MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 21%|██▏       | 430M/1.97G [00:03<00:10, 165MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 22%|██▏       | 447M/1.97G [00:03<00:09, 167MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 23%|██▎       | 471M/1.97G [00:03<00:08, 190MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 24%|██▍       | 494M/1.97G [00:04<00:07, 206MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 26%|██▌       | 518M/1.97G [00:04<00:07, 218MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 27%|██▋       | 539M/1.97G [00:04<00:16, 96.6MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 28%|██▊       | 564M/1.97G [00:04<00:12, 123MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 29%|██▉       | 591M/1.97G [00:04<00:09, 152MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 30%|███       | 613M/1.97G [00:04<00:09, 163MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 32%|███▏      | 641M/1.97G [00:05<00:07, 194MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 33%|███▎      | 668M/1.97G [00:05<00:06, 215MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 34%|███▍      | 695M/1.97G [00:05<00:05, 232MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 36%|███▌      | 720M/1.97G [00:05<00:06, 224MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 37%|███▋      | 743M/1.97G [00:05<00:06, 194MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 38%|███▊      | 769M/1.97G [00:05<00:06, 211MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 39%|███▉      | 791M/1.97G [00:05<00:05, 215MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 40%|████      | 813M/1.97G [00:05<00:05, 215MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 41%|████▏     | 834M/1.97G [00:06<00:18, 67.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 43%|████▎     | 858M/1.97G [00:06<00:14, 86.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 43%|████▎     | 876M/1.97G [00:07<00:13, 91.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 44%|████▍     | 896M/1.97G [00:07<00:15, 76.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 46%|████▌     | 920M/1.97G [00:07<00:11, 97.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 46%|████▋     | 935M/1.97G [00:07<00:12, 93.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 47%|████▋     | 952M/1.97G [00:07<00:11, 97.5MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 48%|████▊     | 976M/1.97G [00:08<00:10, 108MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 50%|████▉     | 0.98G/1.97G [00:08<00:07, 141MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 51%|█████     | 1.00G/1.97G [00:08<00:06, 151MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 52%|█████▏    | 1.02G/1.97G [00:08<00:06, 155MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 53%|█████▎    | 1.04G/1.97G [00:09<00:14, 71.1MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 54%|█████▍    | 1.07G/1.97G [00:09<00:10, 95.7MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 55%|█████▌    | 1.09G/1.97G [00:09<00:07, 121MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 56%|█████▋    | 1.11G/1.97G [00:09<00:06, 142MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 57%|█████▋    | 1.13G/1.97G [00:09<00:06, 137MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 58%|█████▊    | 1.15G/1.97G [00:10<00:14, 60.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 59%|█████▉    | 1.16G/1.97G [00:10<00:13, 65.3MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 60%|█████▉    | 1.18G/1.97G [00:10<00:10, 80.0MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 61%|██████    | 1.19G/1.97G [00:11<00:17, 47.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 61%|██████▏   | 1.21G/1.97G [00:11<00:13, 59.9MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 62%|██████▏   | 1.22G/1.97G [00:11<00:16, 50.2MB/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 62%|██████▏   | 1.23G/1.97G [00:11<00:07, 111MB/s] "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "ename": "OSError",
     "evalue": "[Errno 28] No space left on device",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mOSError\u001b[0m                                   Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[37], line 8\u001b[0m\n\u001b[1;32m      1\u001b[0m (\n\u001b[1;32m      2\u001b[0m     model,\n\u001b[1;32m      3\u001b[0m     vis_processors,\n\u001b[1;32m      4\u001b[0m     txt_processors,\n\u001b[1;32m      5\u001b[0m     image_keys,\n\u001b[1;32m      6\u001b[0m     image_names,\n\u001b[1;32m      7\u001b[0m     features_image_stacked,\n\u001b[0;32m----> 8\u001b[0m ) \u001b[38;5;241m=\u001b[39m \u001b[43mmy_obj\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mparsing_images\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m      9\u001b[0m \u001b[43m    \u001b[49m\u001b[43mmodel_type\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\n\u001b[1;32m     10\u001b[0m \u001b[43m    \u001b[49m\u001b[43mpath_to_save_tensors\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43m/content/drive/MyDrive/misinformation-data/\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[1;32m     11\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/work/AMMICO/AMMICO/ammico/multimodal_search.py:363\u001b[0m, in \u001b[0;36mMultimodalSearch.parsing_images\u001b[0;34m(self, model_type, path_to_save_tensors, path_to_load_tensors)\u001b[0m\n\u001b[1;32m    349\u001b[0m select_extract_image_features \u001b[38;5;241m=\u001b[39m {\n\u001b[1;32m    350\u001b[0m     \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mblip2\u001b[39m\u001b[38;5;124m\"\u001b[39m: MultimodalSearch\u001b[38;5;241m.\u001b[39mextract_image_features_blip2,\n\u001b[1;32m    351\u001b[0m     \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mblip\u001b[39m\u001b[38;5;124m\"\u001b[39m: MultimodalSearch\u001b[38;5;241m.\u001b[39mextract_image_features_basic,\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    355\u001b[0m     \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mclip_vitl14_336\u001b[39m\u001b[38;5;124m\"\u001b[39m: MultimodalSearch\u001b[38;5;241m.\u001b[39mextract_image_features_clip,\n\u001b[1;32m    356\u001b[0m }\n\u001b[1;32m    358\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m model_type \u001b[38;5;129;01min\u001b[39;00m select_model\u001b[38;5;241m.\u001b[39mkeys():\n\u001b[1;32m    359\u001b[0m     (\n\u001b[1;32m    360\u001b[0m         model,\n\u001b[1;32m    361\u001b[0m         vis_processors,\n\u001b[1;32m    362\u001b[0m         txt_processors,\n\u001b[0;32m--> 363\u001b[0m     ) \u001b[38;5;241m=\u001b[39m \u001b[43mselect_model\u001b[49m\u001b[43m[\u001b[49m\n\u001b[1;32m    364\u001b[0m \u001b[43m        \u001b[49m\u001b[43mmodel_type\u001b[49m\n\u001b[1;32m    365\u001b[0m \u001b[43m    \u001b[49m\u001b[43m]\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mMultimodalSearch\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mmultimodal_device\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    366\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m    367\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mSyntaxError\u001b[39;00m(\n\u001b[1;32m    368\u001b[0m         \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mPlease, use one of the following models: blip2, blip, albef, clip_base, clip_vitl14, clip_vitl14_336\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m    369\u001b[0m     )\n",
      "File \u001b[0;32m~/work/AMMICO/AMMICO/ammico/multimodal_search.py:55\u001b[0m, in \u001b[0;36mMultimodalSearch.load_feature_extractor_model_blip\u001b[0;34m(self, device)\u001b[0m\n\u001b[1;32m     43\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mload_feature_extractor_model_blip\u001b[39m(\u001b[38;5;28mself\u001b[39m, device: \u001b[38;5;28mstr\u001b[39m \u001b[38;5;241m=\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mcpu\u001b[39m\u001b[38;5;124m\"\u001b[39m):\n\u001b[1;32m     44\u001b[0m \u001b[38;5;250m    \u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m     45\u001b[0m \u001b[38;5;124;03m    Load base blip_feature_extractor model and preprocessors for visual and text inputs from lavis.models.\u001b[39;00m\n\u001b[1;32m     46\u001b[0m \n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m     53\u001b[0m \u001b[38;5;124;03m        txt_processors (dict): preprocessors for text inputs.\u001b[39;00m\n\u001b[1;32m     54\u001b[0m \u001b[38;5;124;03m    \"\"\"\u001b[39;00m\n\u001b[0;32m---> 55\u001b[0m     model, vis_processors, txt_processors \u001b[38;5;241m=\u001b[39m \u001b[43mload_model_and_preprocess\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m     56\u001b[0m \u001b[43m        \u001b[49m\u001b[43mname\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mblip_feature_extractor\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[1;32m     57\u001b[0m \u001b[43m        \u001b[49m\u001b[43mmodel_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mbase\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[1;32m     58\u001b[0m \u001b[43m        \u001b[49m\u001b[43mis_eval\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\u001b[43m,\u001b[49m\n\u001b[1;32m     59\u001b[0m \u001b[43m        \u001b[49m\u001b[43mdevice\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mdevice\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m     60\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     61\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m model, vis_processors, txt_processors\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/__init__.py:195\u001b[0m, in \u001b[0;36mload_model_and_preprocess\u001b[0;34m(name, model_type, is_eval, device)\u001b[0m\n\u001b[1;32m    192\u001b[0m model_cls \u001b[38;5;241m=\u001b[39m registry\u001b[38;5;241m.\u001b[39mget_model_class(name)\n\u001b[1;32m    194\u001b[0m \u001b[38;5;66;03m# load model\u001b[39;00m\n\u001b[0;32m--> 195\u001b[0m model \u001b[38;5;241m=\u001b[39m \u001b[43mmodel_cls\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfrom_pretrained\u001b[49m\u001b[43m(\u001b[49m\u001b[43mmodel_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mmodel_type\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    197\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m is_eval:\n\u001b[1;32m    198\u001b[0m     model\u001b[38;5;241m.\u001b[39meval()\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/base_model.py:70\u001b[0m, in \u001b[0;36mBaseModel.from_pretrained\u001b[0;34m(cls, model_type)\u001b[0m\n\u001b[1;32m     60\u001b[0m \u001b[38;5;250m\u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m     61\u001b[0m \u001b[38;5;124;03mBuild a pretrained model from default configuration file, specified by model_type.\u001b[39;00m\n\u001b[1;32m     62\u001b[0m \n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m     67\u001b[0m \u001b[38;5;124;03m    - model (nn.Module): pretrained or finetuned model, depending on the configuration.\u001b[39;00m\n\u001b[1;32m     68\u001b[0m \u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m     69\u001b[0m model_cfg \u001b[38;5;241m=\u001b[39m OmegaConf\u001b[38;5;241m.\u001b[39mload(\u001b[38;5;28mcls\u001b[39m\u001b[38;5;241m.\u001b[39mdefault_config_path(model_type))\u001b[38;5;241m.\u001b[39mmodel\n\u001b[0;32m---> 70\u001b[0m model \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mcls\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfrom_config\u001b[49m\u001b[43m(\u001b[49m\u001b[43mmodel_cfg\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     72\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m model\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/blip_models/blip_feature_extractor.py:208\u001b[0m, in \u001b[0;36mBlipFeatureExtractor.from_config\u001b[0;34m(cls, cfg)\u001b[0m\n\u001b[1;32m    206\u001b[0m pretrain_path \u001b[38;5;241m=\u001b[39m cfg\u001b[38;5;241m.\u001b[39mget(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mpretrained\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;28;01mNone\u001b[39;00m)\n\u001b[1;32m    207\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m pretrain_path \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[0;32m--> 208\u001b[0m     msg \u001b[38;5;241m=\u001b[39m \u001b[43mmodel\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mload_from_pretrained\u001b[49m\u001b[43m(\u001b[49m\u001b[43murl_or_filename\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mpretrain_path\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    209\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m    210\u001b[0m     warnings\u001b[38;5;241m.\u001b[39mwarn(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mNo pretrained weights are loaded.\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/models/blip_models/blip.py:30\u001b[0m, in \u001b[0;36mBlipBase.load_from_pretrained\u001b[0;34m(self, url_or_filename)\u001b[0m\n\u001b[1;32m     28\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mload_from_pretrained\u001b[39m(\u001b[38;5;28mself\u001b[39m, url_or_filename):\n\u001b[1;32m     29\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m is_url(url_or_filename):\n\u001b[0;32m---> 30\u001b[0m         cached_file \u001b[38;5;241m=\u001b[39m \u001b[43mdownload_cached_file\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m     31\u001b[0m \u001b[43m            \u001b[49m\u001b[43murl_or_filename\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcheck_hash\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mFalse\u001b[39;49;00m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mprogress\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\n\u001b[1;32m     32\u001b[0m \u001b[43m        \u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     33\u001b[0m         checkpoint \u001b[38;5;241m=\u001b[39m torch\u001b[38;5;241m.\u001b[39mload(cached_file, map_location\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mcpu\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[1;32m     34\u001b[0m     \u001b[38;5;28;01melif\u001b[39;00m os\u001b[38;5;241m.\u001b[39mpath\u001b[38;5;241m.\u001b[39misfile(url_or_filename):\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/lavis/common/dist_utils.py:132\u001b[0m, in \u001b[0;36mdownload_cached_file\u001b[0;34m(url, check_hash, progress)\u001b[0m\n\u001b[1;32m    129\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m cached_file\n\u001b[1;32m    131\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m is_main_process():\n\u001b[0;32m--> 132\u001b[0m     \u001b[43mtimm_hub\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mdownload_cached_file\u001b[49m\u001b[43m(\u001b[49m\u001b[43murl\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcheck_hash\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mprogress\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    134\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m is_dist_avail_and_initialized():\n\u001b[1;32m    135\u001b[0m     dist\u001b[38;5;241m.\u001b[39mbarrier()\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/timm/models/hub.py:51\u001b[0m, in \u001b[0;36mdownload_cached_file\u001b[0;34m(url, check_hash, progress)\u001b[0m\n\u001b[1;32m     49\u001b[0m         r \u001b[38;5;241m=\u001b[39m HASH_REGEX\u001b[38;5;241m.\u001b[39msearch(filename)  \u001b[38;5;66;03m# r is Optional[Match[str]]\u001b[39;00m\n\u001b[1;32m     50\u001b[0m         hash_prefix \u001b[38;5;241m=\u001b[39m r\u001b[38;5;241m.\u001b[39mgroup(\u001b[38;5;241m1\u001b[39m) \u001b[38;5;28;01mif\u001b[39;00m r \u001b[38;5;28;01melse\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[0;32m---> 51\u001b[0m     \u001b[43mdownload_url_to_file\u001b[49m\u001b[43m(\u001b[49m\u001b[43murl\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcached_file\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mhash_prefix\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mprogress\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mprogress\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     52\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m cached_file\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/torch/hub.py:636\u001b[0m, in \u001b[0;36mdownload_url_to_file\u001b[0;34m(url, dst, hash_prefix, progress)\u001b[0m\n\u001b[1;32m    634\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mlen\u001b[39m(buffer) \u001b[38;5;241m==\u001b[39m \u001b[38;5;241m0\u001b[39m:\n\u001b[1;32m    635\u001b[0m     \u001b[38;5;28;01mbreak\u001b[39;00m\n\u001b[0;32m--> 636\u001b[0m \u001b[43mf\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mwrite\u001b[49m\u001b[43m(\u001b[49m\u001b[43mbuffer\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    637\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m hash_prefix \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m    638\u001b[0m     sha256\u001b[38;5;241m.\u001b[39mupdate(buffer)\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/tempfile.py:478\u001b[0m, in \u001b[0;36m_TemporaryFileWrapper.__getattr__.<locals>.func_wrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m    476\u001b[0m \u001b[38;5;129m@_functools\u001b[39m\u001b[38;5;241m.\u001b[39mwraps(func)\n\u001b[1;32m    477\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mfunc_wrapper\u001b[39m(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs):\n\u001b[0;32m--> 478\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfunc\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "\u001b[0;31mOSError\u001b[0m: [Errno 28] No space left on device"
     ]
    }
   ],
   "source": [
    "(\n",
    "    model,\n",
    "    vis_processors,\n",
    "    txt_processors,\n",
    "    image_keys,\n",
    "    image_names,\n",
    "    features_image_stacked,\n",
    ") = my_obj.parsing_images(\n",
    "    model_type, \n",
    "    path_to_save_tensors=\"/content/drive/MyDrive/misinformation-data/\",\n",
    "    )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The images are then processed and stored in a numerical representation, a tensor. These tensors do not change for the same image and same model - so if you run this analysis once, and save the tensors giving a path with the keyword `path_to_save_tensors`, a file with filename `.<Number_of_images>_<model_name>_saved_features_image.pt` will be placed there.\n",
    "\n",
    "This can save you time if you want to analyse same images with the same model but different questions. To run using the saved tensors, execute the below code giving the path and name of the tensor file."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:26.399067Z",
     "iopub.status.busy": "2024-02-19T08:54:26.398613Z",
     "iopub.status.idle": "2024-02-19T08:54:26.401763Z",
     "shell.execute_reply": "2024-02-19T08:54:26.401225Z"
    }
   },
   "outputs": [],
   "source": [
    "# (\n",
    "#     model,\n",
    "#     vis_processors,\n",
    "#     txt_processors,\n",
    "#     image_keys,\n",
    "#     image_names,\n",
    "#     features_image_stacked,\n",
    "# ) = my_obj.parsing_images(\n",
    "#     model_type,\n",
    "#     path_to_load_tensors=\"/content/drive/MyDrive/misinformation-data/5_clip_base_saved_features_image.pt\",\n",
    "# )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here we already processed our image folder with 5 images and the `clip_base` model. So you need just to write the name `5_clip_base_saved_features_image.pt` of the saved file that consists of tensors of all images as keyword argument for `path_to_load_tensors`. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Formulate your search queries\n",
    "\n",
    "Next, you need to form search queries. You can search either by image or by text. You can search for a single query, or you can search for several queries at once, the computational time should not be much different. The format of the queries is as follows:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:26.406155Z",
     "iopub.status.busy": "2024-02-19T08:54:26.405642Z",
     "iopub.status.idle": "2024-02-19T08:54:26.409654Z",
     "shell.execute_reply": "2024-02-19T08:54:26.408999Z"
    }
   },
   "outputs": [],
   "source": [
    "import importlib_resources                                                                       # only require for image query example\n",
    "image_example_query = str(importlib_resources.files(\"ammico\") / \"data\" / \"test-crop-image.png\")   # creating the path to the image for the image query example\n",
    "\n",
    "search_query = [\n",
    "    {\"text_input\": \"politician press conference\"},  \n",
    "    {\"text_input\": \"a world map\"},\n",
    "    {\"text_input\": \"a dog\"},            # This is how looks text query\n",
    "    {\"image\": image_example_query},      # This is how looks image query, here `image_example_path` is the path to query image like \"data/test-crop-image.png\"\n",
    "]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can filter your results in 3 different ways:\n",
    "- `filter_number_of_images` limits the number of images found. That is, if the parameter `filter_number_of_images = 10`, then the first 10 images that best match the query will be shown. The other images ranks will be set to `None` and the similarity value to `0`.\n",
    "- `filter_val_limit` limits the output of images with a similarity value not bigger than `filter_val_limit`. That is, if the parameter `filter_val_limit = 0.2`, all images with similarity less than 0.2 will be discarded.\n",
    "- `filter_rel_error` (percentage) limits the output of images with a similarity value not bigger than `100 * abs(current_simularity_value - best_simularity_value_in_current_search)/best_simularity_value_in_current_search < filter_rel_error`. That is, if we set filter_rel_error = 30, it means that if the top1 image have 0.5 similarity value, we discard all image with similarity less than 0.35."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:26.412858Z",
     "iopub.status.busy": "2024-02-19T08:54:26.412664Z",
     "iopub.status.idle": "2024-02-19T08:54:26.436288Z",
     "shell.execute_reply": "2024-02-19T08:54:26.435775Z"
    }
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'model' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[40], line 2\u001b[0m\n\u001b[1;32m      1\u001b[0m similarity, sorted_lists \u001b[38;5;241m=\u001b[39m my_obj\u001b[38;5;241m.\u001b[39mmultimodal_search(\n\u001b[0;32m----> 2\u001b[0m     \u001b[43mmodel\u001b[49m,\n\u001b[1;32m      3\u001b[0m     vis_processors,\n\u001b[1;32m      4\u001b[0m     txt_processors,\n\u001b[1;32m      5\u001b[0m     model_type,\n\u001b[1;32m      6\u001b[0m     image_keys,\n\u001b[1;32m      7\u001b[0m     features_image_stacked,\n\u001b[1;32m      8\u001b[0m     search_query,\n\u001b[1;32m      9\u001b[0m     filter_number_of_images\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m20\u001b[39m,\n\u001b[1;32m     10\u001b[0m )\n",
      "\u001b[0;31mNameError\u001b[0m: name 'model' is not defined"
     ]
    }
   ],
   "source": [
    "similarity, sorted_lists = my_obj.multimodal_search(\n",
    "    model,\n",
    "    vis_processors,\n",
    "    txt_processors,\n",
    "    model_type,\n",
    "    image_keys,\n",
    "    features_image_stacked,\n",
    "    search_query,\n",
    "    filter_number_of_images=20,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:26.443401Z",
     "iopub.status.busy": "2024-02-19T08:54:26.442981Z",
     "iopub.status.idle": "2024-02-19T08:54:26.462491Z",
     "shell.execute_reply": "2024-02-19T08:54:26.461996Z"
    }
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'similarity' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[41], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m \u001b[43msimilarity\u001b[49m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'similarity' is not defined"
     ]
    }
   ],
   "source": [
    "similarity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:26.465916Z",
     "iopub.status.busy": "2024-02-19T08:54:26.465540Z",
     "iopub.status.idle": "2024-02-19T08:54:26.484725Z",
     "shell.execute_reply": "2024-02-19T08:54:26.484175Z"
    }
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'sorted_lists' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[42], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m \u001b[43msorted_lists\u001b[49m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'sorted_lists' is not defined"
     ]
    }
   ],
   "source": [
    "sorted_lists"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After launching `multimodal_search` function, the results of each query will be added to the source dictionary.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:26.487891Z",
     "iopub.status.busy": "2024-02-19T08:54:26.487505Z",
     "iopub.status.idle": "2024-02-19T08:54:26.494730Z",
     "shell.execute_reply": "2024-02-19T08:54:26.494169Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'img4': {'filename': 'data-test/img4.png',\n",
       "  'face': 'No',\n",
       "  'multiple_faces': 'No',\n",
       "  'no_faces': 0,\n",
       "  'wears_mask': ['No'],\n",
       "  'age': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'emotion': [None],\n",
       "  'emotion (category)': [None],\n",
       "  'text': 'MOODOVIN XI',\n",
       "  'text_language': 'en',\n",
       "  'text_english': 'MOODOVIN XI',\n",
       "  'text_clean': 'XI',\n",
       "  'text_summary': ' MOODOVIN XI XI: Vladimir Putin, Vladimir Vladmir Zelizer, Vladimir',\n",
       "  'sentiment': 'POSITIVE',\n",
       "  'sentiment_score': 0.66,\n",
       "  'entity': ['MOODOVIN XI'],\n",
       "  'entity_type': ['ORG'],\n",
       "  'const_image_summary': 'a river running through a city next to tall buildings',\n",
       "  '3_non-deterministic_summary': ['there is a pretty house that sits above the water',\n",
       "   'there is a building with a balcony and lots of plants on the side of it',\n",
       "   'several buildings with a river flowing through it']},\n",
       " 'img1': {'filename': 'data-test/img1.png',\n",
       "  'no_faces': 0,\n",
       "  'age': [None],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': [None],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'face': 'No',\n",
       "  'text': 'SCATTERING THEORY The Quantum Theory of Nonrelativistic Collisions JOHN R. TAYLOR University of Colorado',\n",
       "  'text_language': 'en',\n",
       "  'text_english': 'SCATTERING THEORY The Quantum Theory of Nonrelativistic Collisions JOHN R. TAYLOR University of Colorado',\n",
       "  'text_clean': 'THEORY The Quantum Theory of Collisions JOHN R. TAYLOR University of Colorado',\n",
       "  'text_summary': ' SCATTERING THEORY The Quantum Theory of Nonrelativistic Collisions',\n",
       "  'sentiment': 'POSITIVE',\n",
       "  'sentiment_score': 0.91,\n",
       "  'entity': ['Non',\n",
       "   '##vist',\n",
       "   'Col',\n",
       "   '##N',\n",
       "   'R',\n",
       "   'T',\n",
       "   '##AYL',\n",
       "   'University of Colorado'],\n",
       "  'entity_type': ['MISC', 'MISC', 'MISC', 'ORG', 'PER', 'PER', 'ORG', 'ORG'],\n",
       "  'const_image_summary': 'a close up of a piece of paper with writing on it',\n",
       "  '3_non-deterministic_summary': ['a book opened to the book title for a novel',\n",
       "   'there are many text on this page',\n",
       "   'the text in a book is a handwritten poem']},\n",
       " 'img2': {'filename': 'data-test/img2.png',\n",
       "  'no_faces': 0,\n",
       "  'age': [None],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': [None],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'face': 'No',\n",
       "  'text': 'THE ALGEBRAIC EIGENVALUE PROBLEM DOM NVS TIO MINA Monographs on Numerical Analysis J.. H. WILKINSON',\n",
       "  'text_language': 'en',\n",
       "  'text_english': 'THE ALGEBRAIC EIGENVALUE PROBLEM DOM NVS TIO MINA Monographs on Numerical Analysis J.. H. WILKINSON',\n",
       "  'text_clean': 'THE PROBLEM DOM NVS TIO MINA Monographs on Numerical Analysis J .. H. WILKINSON',\n",
       "  'text_summary': ' H. H. W. WILKINSON: The Algebri',\n",
       "  'sentiment': 'NEGATIVE',\n",
       "  'sentiment_score': 0.97,\n",
       "  'entity': ['ALGEBRAIC EIGENVAL', 'NVS TIO MI', 'J', 'H', 'WILKINSON'],\n",
       "  'entity_type': ['MISC', 'ORG', 'ORG', 'ORG', 'ORG'],\n",
       "  'const_image_summary': 'a yellow book with green lettering on it',\n",
       "  '3_non-deterministic_summary': ['a book cover with green writing on a black background',\n",
       "   'the title page of a book with information from its authors',\n",
       "   'a book about the age - related engineering and engineering']},\n",
       " 'img3': {'filename': 'data-test/img3.png',\n",
       "  'no_faces': 0,\n",
       "  'age': [None],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': [None],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'face': 'No',\n",
       "  'text': 'm OOOO 0000 www.',\n",
       "  'text_language': 'en',\n",
       "  'text_english': 'm OOOO 0000 www.',\n",
       "  'text_clean': 'm www .',\n",
       "  'text_summary': ' www. m OOOO 0000 0000 www.m.m OOOo 0000',\n",
       "  'sentiment': 'NEGATIVE',\n",
       "  'sentiment_score': 0.62,\n",
       "  'entity': [],\n",
       "  'entity_type': [],\n",
       "  'const_image_summary': 'a bus that is sitting on the side of a road',\n",
       "  '3_non-deterministic_summary': ['there are cars and a bus on the side of the road',\n",
       "   'a bus that is sitting in the middle of a street',\n",
       "   'an aerial view of an empty city street with two large buses passing by']},\n",
       " 'img0': {'filename': 'data-test/img0.png',\n",
       "  'no_faces': 0,\n",
       "  'age': [None],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': [None],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': [None],\n",
       "  'gender': [None],\n",
       "  'race': [None],\n",
       "  'face': 'No',\n",
       "  'text': 'Mathematische Formelsammlung für Ingenieure und Naturwissenschaftler Mit zahlreichen Abbildungen und Rechenbeispielen und einer ausführlichen Integraltafel 3., verbesserte Auflage',\n",
       "  'text_language': 'de',\n",
       "  'text_english': 'Mathematical formula collection for engineers and scientists With numerous illustrations and calculation examples and a detailed integral table 3rd, improved edition',\n",
       "  'text_clean': 'Mathematical formula collection for engineers and scientists With numerous illustrations and calculation examples and a detailed integral table 3rd , improved edition',\n",
       "  'text_summary': ' Mathematical formula collection for engineers and scientists . Includes numerous illustrations and calculation examples . Includes',\n",
       "  'sentiment': 'POSITIVE',\n",
       "  'sentiment_score': 1.0,\n",
       "  'entity': [],\n",
       "  'entity_type': [],\n",
       "  'const_image_summary': 'a close up of an open book with writing on it',\n",
       "  '3_non-deterministic_summary': ['a close up of a book with many languages',\n",
       "   'a book that is opened up in german',\n",
       "   'book about mathemarche formulals and their meaning']},\n",
       " 'img5': {'filename': 'data-test/img5.png',\n",
       "  'no_faces': 1,\n",
       "  'age': [26],\n",
       "  'wears_mask': ['No'],\n",
       "  'emotion (category)': ['Negative'],\n",
       "  'multiple_faces': 'No',\n",
       "  'emotion': ['sad'],\n",
       "  'gender': ['Man'],\n",
       "  'race': [None],\n",
       "  'face': 'Yes',\n",
       "  'text': None,\n",
       "  'text_language': 'en',\n",
       "  'text_english': '',\n",
       "  'text_clean': '',\n",
       "  'text_summary': ' CNN.com will feature iReporter photos in a weekly Travel Snapshots gallery .',\n",
       "  'sentiment': 'POSITIVE',\n",
       "  'sentiment_score': 0.75,\n",
       "  'entity': [],\n",
       "  'entity_type': [],\n",
       "  'const_image_summary': 'a person running on a beach near a rock formation',\n",
       "  '3_non-deterministic_summary': ['a woman is running down the beach next to some rocks',\n",
       "   'a woman running along the beach by the ocean',\n",
       "   'there is a person running on the beach next to the ocean']}}"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "image_dict"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A special function was written to present the search results conveniently. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:26.498308Z",
     "iopub.status.busy": "2024-02-19T08:54:26.498116Z",
     "iopub.status.idle": "2024-02-19T08:54:26.561629Z",
     "shell.execute_reply": "2024-02-19T08:54:26.560724Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Your search query: politician press conference'"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'--------------------------------------------------'"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'Results:'"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "ename": "KeyError",
     "evalue": "'politician press conference'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyError\u001b[0m                                  Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[44], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m \u001b[43mmy_obj\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mshow_results\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m      2\u001b[0m \u001b[43m    \u001b[49m\u001b[43msearch_query\u001b[49m\u001b[43m[\u001b[49m\u001b[38;5;241;43m0\u001b[39;49m\u001b[43m]\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;66;43;03m# you can change the index to see the results for other queries\u001b[39;49;00m\n\u001b[1;32m      3\u001b[0m \u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/work/AMMICO/AMMICO/ammico/multimodal_search.py:970\u001b[0m, in \u001b[0;36mMultimodalSearch.show_results\u001b[0;34m(self, query, itm, image_gradcam_with_itm)\u001b[0m\n\u001b[1;32m    967\u001b[0m     current_querry_val \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mlist\u001b[39m(query\u001b[38;5;241m.\u001b[39mvalues())[\u001b[38;5;241m0\u001b[39m]\n\u001b[1;32m    968\u001b[0m     current_querry_rank \u001b[38;5;241m=\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mrank \u001b[39m\u001b[38;5;124m\"\u001b[39m \u001b[38;5;241m+\u001b[39m \u001b[38;5;28mlist\u001b[39m(query\u001b[38;5;241m.\u001b[39mvalues())[\u001b[38;5;241m0\u001b[39m]\n\u001b[0;32m--> 970\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m s \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28;43msorted\u001b[39;49m\u001b[43m(\u001b[49m\n\u001b[1;32m    971\u001b[0m \u001b[43m    \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43msubdict\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mitems\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mkey\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mlambda\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mt\u001b[49m\u001b[43m:\u001b[49m\u001b[43m \u001b[49m\u001b[43mt\u001b[49m\u001b[43m[\u001b[49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[43m]\u001b[49m\u001b[43m[\u001b[49m\u001b[43mcurrent_querry_val\u001b[49m\u001b[43m]\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mreverse\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\n\u001b[1;32m    972\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m:\n\u001b[1;32m    973\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m s[\u001b[38;5;241m1\u001b[39m][current_querry_rank] \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m    974\u001b[0m         \u001b[38;5;28;01mbreak\u001b[39;00m\n",
      "File \u001b[0;32m~/work/AMMICO/AMMICO/ammico/multimodal_search.py:971\u001b[0m, in \u001b[0;36mMultimodalSearch.show_results.<locals>.<lambda>\u001b[0;34m(t)\u001b[0m\n\u001b[1;32m    967\u001b[0m     current_querry_val \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mlist\u001b[39m(query\u001b[38;5;241m.\u001b[39mvalues())[\u001b[38;5;241m0\u001b[39m]\n\u001b[1;32m    968\u001b[0m     current_querry_rank \u001b[38;5;241m=\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mrank \u001b[39m\u001b[38;5;124m\"\u001b[39m \u001b[38;5;241m+\u001b[39m \u001b[38;5;28mlist\u001b[39m(query\u001b[38;5;241m.\u001b[39mvalues())[\u001b[38;5;241m0\u001b[39m]\n\u001b[1;32m    970\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m s \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28msorted\u001b[39m(\n\u001b[0;32m--> 971\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msubdict\u001b[38;5;241m.\u001b[39mitems(), key\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mlambda\u001b[39;00m t: \u001b[43mt\u001b[49m\u001b[43m[\u001b[49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[43m]\u001b[49m\u001b[43m[\u001b[49m\u001b[43mcurrent_querry_val\u001b[49m\u001b[43m]\u001b[49m, reverse\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mTrue\u001b[39;00m\n\u001b[1;32m    972\u001b[0m ):\n\u001b[1;32m    973\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m s[\u001b[38;5;241m1\u001b[39m][current_querry_rank] \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m    974\u001b[0m         \u001b[38;5;28;01mbreak\u001b[39;00m\n",
      "\u001b[0;31mKeyError\u001b[0m: 'politician press conference'"
     ]
    }
   ],
   "source": [
    "my_obj.show_results(\n",
    "    search_query[0], # you can change the index to see the results for other queries\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:26.564539Z",
     "iopub.status.busy": "2024-02-19T08:54:26.564124Z",
     "iopub.status.idle": "2024-02-19T08:54:26.742817Z",
     "shell.execute_reply": "2024-02-19T08:54:26.742251Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Your search query: '"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/jpeg": "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",
      "image/png": "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",
      "text/plain": [
       "<PIL.Image.Image image mode=RGB size=400x185>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'--------------------------------------------------'"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'Results:'"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "ename": "KeyError",
     "evalue": "'/home/runner/work/AMMICO/AMMICO/ammico/data/test-crop-image.png'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyError\u001b[0m                                  Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[45], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m \u001b[43mmy_obj\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mshow_results\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m      2\u001b[0m \u001b[43m    \u001b[49m\u001b[43msearch_query\u001b[49m\u001b[43m[\u001b[49m\u001b[38;5;241;43m3\u001b[39;49m\u001b[43m]\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;66;43;03m# you can change the index to see the results for other queries\u001b[39;49;00m\n\u001b[1;32m      3\u001b[0m \u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/work/AMMICO/AMMICO/ammico/multimodal_search.py:970\u001b[0m, in \u001b[0;36mMultimodalSearch.show_results\u001b[0;34m(self, query, itm, image_gradcam_with_itm)\u001b[0m\n\u001b[1;32m    967\u001b[0m     current_querry_val \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mlist\u001b[39m(query\u001b[38;5;241m.\u001b[39mvalues())[\u001b[38;5;241m0\u001b[39m]\n\u001b[1;32m    968\u001b[0m     current_querry_rank \u001b[38;5;241m=\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mrank \u001b[39m\u001b[38;5;124m\"\u001b[39m \u001b[38;5;241m+\u001b[39m \u001b[38;5;28mlist\u001b[39m(query\u001b[38;5;241m.\u001b[39mvalues())[\u001b[38;5;241m0\u001b[39m]\n\u001b[0;32m--> 970\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m s \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28;43msorted\u001b[39;49m\u001b[43m(\u001b[49m\n\u001b[1;32m    971\u001b[0m \u001b[43m    \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43msubdict\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mitems\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mkey\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mlambda\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mt\u001b[49m\u001b[43m:\u001b[49m\u001b[43m \u001b[49m\u001b[43mt\u001b[49m\u001b[43m[\u001b[49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[43m]\u001b[49m\u001b[43m[\u001b[49m\u001b[43mcurrent_querry_val\u001b[49m\u001b[43m]\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mreverse\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\n\u001b[1;32m    972\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m:\n\u001b[1;32m    973\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m s[\u001b[38;5;241m1\u001b[39m][current_querry_rank] \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m    974\u001b[0m         \u001b[38;5;28;01mbreak\u001b[39;00m\n",
      "File \u001b[0;32m~/work/AMMICO/AMMICO/ammico/multimodal_search.py:971\u001b[0m, in \u001b[0;36mMultimodalSearch.show_results.<locals>.<lambda>\u001b[0;34m(t)\u001b[0m\n\u001b[1;32m    967\u001b[0m     current_querry_val \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mlist\u001b[39m(query\u001b[38;5;241m.\u001b[39mvalues())[\u001b[38;5;241m0\u001b[39m]\n\u001b[1;32m    968\u001b[0m     current_querry_rank \u001b[38;5;241m=\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mrank \u001b[39m\u001b[38;5;124m\"\u001b[39m \u001b[38;5;241m+\u001b[39m \u001b[38;5;28mlist\u001b[39m(query\u001b[38;5;241m.\u001b[39mvalues())[\u001b[38;5;241m0\u001b[39m]\n\u001b[1;32m    970\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m s \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28msorted\u001b[39m(\n\u001b[0;32m--> 971\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msubdict\u001b[38;5;241m.\u001b[39mitems(), key\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mlambda\u001b[39;00m t: \u001b[43mt\u001b[49m\u001b[43m[\u001b[49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[43m]\u001b[49m\u001b[43m[\u001b[49m\u001b[43mcurrent_querry_val\u001b[49m\u001b[43m]\u001b[49m, reverse\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mTrue\u001b[39;00m\n\u001b[1;32m    972\u001b[0m ):\n\u001b[1;32m    973\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m s[\u001b[38;5;241m1\u001b[39m][current_querry_rank] \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m    974\u001b[0m         \u001b[38;5;28;01mbreak\u001b[39;00m\n",
      "\u001b[0;31mKeyError\u001b[0m: '/home/runner/work/AMMICO/AMMICO/ammico/data/test-crop-image.png'"
     ]
    }
   ],
   "source": [
    "my_obj.show_results(\n",
    "    search_query[3], # you can change the index to see the results for other queries\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Improve the search results\n",
    "\n",
    "For even better results, a slightly different approach has been prepared that can improve search results. It is quite resource-intensive, so it is applied after the main algorithm has found the most relevant images. This approach works only with text queries and it skips image queries. Among the parameters you can choose 3 models: `\"blip_base\"`, `\"blip_large\"`, `\"blip2_coco\"`. If you get an `Out of Memory` error, try reducing the batch_size value (minimum = 1), which is the number of images being processed simultaneously. With the parameter `need_grad_cam = True/False` you can enable the calculation of the heat map of each image to be processed and save them in `image_gradcam_with_itm`. Thus the `image_text_match_reordering()` function calculates new similarity values and new ranks for each image. The resulting values are added to the general dictionary."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:26.746808Z",
     "iopub.status.busy": "2024-02-19T08:54:26.746601Z",
     "iopub.status.idle": "2024-02-19T08:54:26.749457Z",
     "shell.execute_reply": "2024-02-19T08:54:26.748884Z"
    }
   },
   "outputs": [],
   "source": [
    "itm_model = \"blip_base\"\n",
    "# itm_model = \"blip_large\"\n",
    "# itm_model = \"blip2_coco\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:26.752555Z",
     "iopub.status.busy": "2024-02-19T08:54:26.752084Z",
     "iopub.status.idle": "2024-02-19T08:54:26.773035Z",
     "shell.execute_reply": "2024-02-19T08:54:26.772413Z"
    }
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'image_keys' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[47], line 4\u001b[0m\n\u001b[1;32m      1\u001b[0m itm_scores, image_gradcam_with_itm \u001b[38;5;241m=\u001b[39m my_obj\u001b[38;5;241m.\u001b[39mimage_text_match_reordering(\n\u001b[1;32m      2\u001b[0m     search_query,\n\u001b[1;32m      3\u001b[0m     itm_model,\n\u001b[0;32m----> 4\u001b[0m     \u001b[43mimage_keys\u001b[49m,\n\u001b[1;32m      5\u001b[0m     sorted_lists,\n\u001b[1;32m      6\u001b[0m     batch_size\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m1\u001b[39m,\n\u001b[1;32m      7\u001b[0m     need_grad_cam\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mTrue\u001b[39;00m,\n\u001b[1;32m      8\u001b[0m )\n",
      "\u001b[0;31mNameError\u001b[0m: name 'image_keys' is not defined"
     ]
    }
   ],
   "source": [
    "itm_scores, image_gradcam_with_itm = my_obj.image_text_match_reordering(\n",
    "    search_query,\n",
    "    itm_model,\n",
    "    image_keys,\n",
    "    sorted_lists,\n",
    "    batch_size=1,\n",
    "    need_grad_cam=True,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then using the same output function you can add the `itm=True` argument to output the new image order. Remember that for images querys, an error will be thrown with `itm=True` argument. You can also add the `image_gradcam_with_itm` along with `itm=True` argument to output the heat maps of the calculated images."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:26.775851Z",
     "iopub.status.busy": "2024-02-19T08:54:26.775512Z",
     "iopub.status.idle": "2024-02-19T08:54:26.795328Z",
     "shell.execute_reply": "2024-02-19T08:54:26.794723Z"
    }
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'image_gradcam_with_itm' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[48], line 2\u001b[0m\n\u001b[1;32m      1\u001b[0m my_obj\u001b[38;5;241m.\u001b[39mshow_results(\n\u001b[0;32m----> 2\u001b[0m     search_query[\u001b[38;5;241m0\u001b[39m], itm\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mTrue\u001b[39;00m, image_gradcam_with_itm\u001b[38;5;241m=\u001b[39m\u001b[43mimage_gradcam_with_itm\u001b[49m\n\u001b[1;32m      3\u001b[0m )\n",
      "\u001b[0;31mNameError\u001b[0m: name 'image_gradcam_with_itm' is not defined"
     ]
    }
   ],
   "source": [
    "my_obj.show_results(\n",
    "    search_query[0], itm=True, image_gradcam_with_itm=image_gradcam_with_itm\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Save search results to csv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Convert the dictionary of dictionarys into a dictionary with lists:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:26.799498Z",
     "iopub.status.busy": "2024-02-19T08:54:26.799139Z",
     "iopub.status.idle": "2024-02-19T08:54:26.818722Z",
     "shell.execute_reply": "2024-02-19T08:54:26.818206Z"
    }
   },
   "outputs": [
    {
     "ename": "AttributeError",
     "evalue": "module 'ammico' has no attribute 'append_data_to_dict'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[49], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m outdict \u001b[38;5;241m=\u001b[39m \u001b[43mammico\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mappend_data_to_dict\u001b[49m(image_dict)\n\u001b[1;32m      2\u001b[0m df \u001b[38;5;241m=\u001b[39m ammico\u001b[38;5;241m.\u001b[39mdump_df(outdict)\n",
      "\u001b[0;31mAttributeError\u001b[0m: module 'ammico' has no attribute 'append_data_to_dict'"
     ]
    }
   ],
   "source": [
    "outdict = ammico.append_data_to_dict(image_dict)\n",
    "df = ammico.dump_df(outdict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Check the dataframe:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:26.822275Z",
     "iopub.status.busy": "2024-02-19T08:54:26.821902Z",
     "iopub.status.idle": "2024-02-19T08:54:26.841826Z",
     "shell.execute_reply": "2024-02-19T08:54:26.841214Z"
    }
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'df' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[50], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m \u001b[43mdf\u001b[49m\u001b[38;5;241m.\u001b[39mhead(\u001b[38;5;241m10\u001b[39m)\n",
      "\u001b[0;31mNameError\u001b[0m: name 'df' is not defined"
     ]
    }
   ],
   "source": [
    "df.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Write the csv file:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:26.845662Z",
     "iopub.status.busy": "2024-02-19T08:54:26.845309Z",
     "iopub.status.idle": "2024-02-19T08:54:26.864880Z",
     "shell.execute_reply": "2024-02-19T08:54:26.864321Z"
    }
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'df' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[51], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m \u001b[43mdf\u001b[49m\u001b[38;5;241m.\u001b[39mto_csv(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m/content/drive/MyDrive/misinformation-data/data_out.csv\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n",
      "\u001b[0;31mNameError\u001b[0m: name 'df' is not defined"
     ]
    }
   ],
   "source": [
    "df.to_csv(\"/content/drive/MyDrive/misinformation-data/data_out.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Color analysis of pictures\n",
    "\n",
    "This module shows primary color analysis of color image using K-Means algorithm.\n",
    "The output are N primary colors and their corresponding percentage."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To check the analysis, you can inspect the analyzed elements here. Loading the results takes a moment, so please be patient. If you are sure of what you are doing, you can skip this and directly export a csv file in the step below.\n",
    "Here, we display the color detection results provided by `colorgram` and `colour` libraries. Click on the tabs to see the results in the right sidebar. You may need to increment the `port` number if you are already running several notebook instances on the same server."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:26.868479Z",
     "iopub.status.busy": "2024-02-19T08:54:26.868084Z",
     "iopub.status.idle": "2024-02-19T08:54:26.911851Z",
     "shell.execute_reply": "2024-02-19T08:54:26.911132Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "        <iframe\n",
       "            width=\"100%\"\n",
       "            height=\"650\"\n",
       "            src=\"http://127.0.0.1:8057/\"\n",
       "            frameborder=\"0\"\n",
       "            allowfullscreen\n",
       "            \n",
       "        ></iframe>\n",
       "        "
      ],
      "text/plain": [
       "<IPython.lib.display.IFrame at 0x7f9791095040>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "analysis_explorer = ammico.AnalysisExplorer(image_dict)\n",
    "analysis_explorer.run_server(port = 8057)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Instead of inspecting each of the images, you can also directly carry out the analysis and export the result into a csv. This may take a while depending on how many images you have loaded."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:26.917049Z",
     "iopub.status.busy": "2024-02-19T08:54:26.916702Z",
     "iopub.status.idle": "2024-02-19T08:54:38.080384Z",
     "shell.execute_reply": "2024-02-19T08:54:38.079754Z"
    }
   },
   "outputs": [],
   "source": [
    "for key in image_dict.keys():\n",
    "    image_dict[key] = ammico.colors.ColorDetector(image_dict[key]).analyse_image()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "These steps are required to convert the dictionary of dictionarys into a dictionary with lists, that can be converted into a pandas dataframe and exported to a csv file."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:38.085405Z",
     "iopub.status.busy": "2024-02-19T08:54:38.084973Z",
     "iopub.status.idle": "2024-02-19T08:54:38.089091Z",
     "shell.execute_reply": "2024-02-19T08:54:38.088517Z"
    }
   },
   "outputs": [],
   "source": [
    "df = ammico.get_dataframe(image_dict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Check the dataframe:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:38.097290Z",
     "iopub.status.busy": "2024-02-19T08:54:38.096874Z",
     "iopub.status.idle": "2024-02-19T08:54:38.119692Z",
     "shell.execute_reply": "2024-02-19T08:54:38.119109Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>filename</th>\n",
       "      <th>face</th>\n",
       "      <th>multiple_faces</th>\n",
       "      <th>no_faces</th>\n",
       "      <th>wears_mask</th>\n",
       "      <th>age</th>\n",
       "      <th>gender</th>\n",
       "      <th>race</th>\n",
       "      <th>emotion</th>\n",
       "      <th>emotion (category)</th>\n",
       "      <th>...</th>\n",
       "      <th>blue</th>\n",
       "      <th>yellow</th>\n",
       "      <th>cyan</th>\n",
       "      <th>orange</th>\n",
       "      <th>purple</th>\n",
       "      <th>pink</th>\n",
       "      <th>brown</th>\n",
       "      <th>grey</th>\n",
       "      <th>white</th>\n",
       "      <th>black</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>data-test/img4.png</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>[No]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>...</td>\n",
       "      <td>0.16</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.42</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>data-test/img1.png</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>[No]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.96</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>data-test/img2.png</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>[No]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.15</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>data-test/img3.png</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>[No]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.92</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>data-test/img0.png</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>[No]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.98</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>data-test/img5.png</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "      <td>[No]</td>\n",
       "      <td>[26]</td>\n",
       "      <td>[Man]</td>\n",
       "      <td>[None]</td>\n",
       "      <td>[sad]</td>\n",
       "      <td>[Negative]</td>\n",
       "      <td>...</td>\n",
       "      <td>0.12</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>6 rows × 33 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             filename face multiple_faces  no_faces wears_mask     age  \\\n",
       "0  data-test/img4.png   No             No         0       [No]  [None]   \n",
       "1  data-test/img1.png   No             No         0       [No]  [None]   \n",
       "2  data-test/img2.png   No             No         0       [No]  [None]   \n",
       "3  data-test/img3.png   No             No         0       [No]  [None]   \n",
       "4  data-test/img0.png   No             No         0       [No]  [None]   \n",
       "5  data-test/img5.png  Yes             No         1       [No]    [26]   \n",
       "\n",
       "   gender    race emotion emotion (category)  ...  blue yellow cyan orange  \\\n",
       "0  [None]  [None]  [None]             [None]  ...  0.16   0.00    0      0   \n",
       "1  [None]  [None]  [None]             [None]  ...  0.00   0.00    0      0   \n",
       "2  [None]  [None]  [None]             [None]  ...  0.00   0.75    0      0   \n",
       "3  [None]  [None]  [None]             [None]  ...  0.00   0.00    0      0   \n",
       "4  [None]  [None]  [None]             [None]  ...  0.00   0.00    0      0   \n",
       "5   [Man]  [None]   [sad]         [Negative]  ...  0.12   0.00    0      0   \n",
       "\n",
       "  purple pink  brown  grey white black  \n",
       "0   0.00    0   0.10  0.42  0.05  0.21  \n",
       "1   0.00    0   0.00  0.96  0.00  0.04  \n",
       "2   0.00    0   0.04  0.15  0.00  0.02  \n",
       "3   0.02    0   0.06  0.92  0.01  0.00  \n",
       "4   0.00    0   0.00  0.98  0.00  0.02  \n",
       "5   0.00    0   0.02  0.50  0.00  0.00  \n",
       "\n",
       "[6 rows x 33 columns]"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Write the csv file - here you should provide a file path and file name for the csv file to be written."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-19T08:54:38.124073Z",
     "iopub.status.busy": "2024-02-19T08:54:38.123661Z",
     "iopub.status.idle": "2024-02-19T08:54:38.203962Z",
     "shell.execute_reply": "2024-02-19T08:54:38.203276Z"
    }
   },
   "outputs": [
    {
     "ename": "OSError",
     "evalue": "Cannot save file into a non-existent directory: '/content/drive/MyDrive/misinformation-data'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mOSError\u001b[0m                                   Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[56], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m \u001b[43mdf\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mto_csv\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43m/content/drive/MyDrive/misinformation-data/data_out.csv\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/pandas/util/_decorators.py:333\u001b[0m, in \u001b[0;36mdeprecate_nonkeyword_arguments.<locals>.decorate.<locals>.wrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m    327\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mlen\u001b[39m(args) \u001b[38;5;241m>\u001b[39m num_allow_args:\n\u001b[1;32m    328\u001b[0m     warnings\u001b[38;5;241m.\u001b[39mwarn(\n\u001b[1;32m    329\u001b[0m         msg\u001b[38;5;241m.\u001b[39mformat(arguments\u001b[38;5;241m=\u001b[39m_format_argument_list(allow_args)),\n\u001b[1;32m    330\u001b[0m         \u001b[38;5;167;01mFutureWarning\u001b[39;00m,\n\u001b[1;32m    331\u001b[0m         stacklevel\u001b[38;5;241m=\u001b[39mfind_stack_level(),\n\u001b[1;32m    332\u001b[0m     )\n\u001b[0;32m--> 333\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfunc\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/pandas/core/generic.py:3961\u001b[0m, in \u001b[0;36mNDFrame.to_csv\u001b[0;34m(self, path_or_buf, sep, na_rep, float_format, columns, header, index, index_label, mode, encoding, compression, quoting, quotechar, lineterminator, chunksize, date_format, doublequote, escapechar, decimal, errors, storage_options)\u001b[0m\n\u001b[1;32m   3950\u001b[0m df \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28misinstance\u001b[39m(\u001b[38;5;28mself\u001b[39m, ABCDataFrame) \u001b[38;5;28;01melse\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mto_frame()\n\u001b[1;32m   3952\u001b[0m formatter \u001b[38;5;241m=\u001b[39m DataFrameFormatter(\n\u001b[1;32m   3953\u001b[0m     frame\u001b[38;5;241m=\u001b[39mdf,\n\u001b[1;32m   3954\u001b[0m     header\u001b[38;5;241m=\u001b[39mheader,\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m   3958\u001b[0m     decimal\u001b[38;5;241m=\u001b[39mdecimal,\n\u001b[1;32m   3959\u001b[0m )\n\u001b[0;32m-> 3961\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mDataFrameRenderer\u001b[49m\u001b[43m(\u001b[49m\u001b[43mformatter\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mto_csv\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m   3962\u001b[0m \u001b[43m    \u001b[49m\u001b[43mpath_or_buf\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3963\u001b[0m \u001b[43m    \u001b[49m\u001b[43mlineterminator\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mlineterminator\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3964\u001b[0m \u001b[43m    \u001b[49m\u001b[43msep\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43msep\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3965\u001b[0m \u001b[43m    \u001b[49m\u001b[43mencoding\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mencoding\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3966\u001b[0m \u001b[43m    \u001b[49m\u001b[43merrors\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43merrors\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3967\u001b[0m \u001b[43m    \u001b[49m\u001b[43mcompression\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcompression\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3968\u001b[0m \u001b[43m    \u001b[49m\u001b[43mquoting\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mquoting\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3969\u001b[0m \u001b[43m    \u001b[49m\u001b[43mcolumns\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcolumns\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3970\u001b[0m \u001b[43m    \u001b[49m\u001b[43mindex_label\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mindex_label\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3971\u001b[0m \u001b[43m    \u001b[49m\u001b[43mmode\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mmode\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3972\u001b[0m \u001b[43m    \u001b[49m\u001b[43mchunksize\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mchunksize\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3973\u001b[0m \u001b[43m    \u001b[49m\u001b[43mquotechar\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mquotechar\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3974\u001b[0m \u001b[43m    \u001b[49m\u001b[43mdate_format\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mdate_format\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3975\u001b[0m \u001b[43m    \u001b[49m\u001b[43mdoublequote\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mdoublequote\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3976\u001b[0m \u001b[43m    \u001b[49m\u001b[43mescapechar\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mescapechar\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3977\u001b[0m \u001b[43m    \u001b[49m\u001b[43mstorage_options\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mstorage_options\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   3978\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/pandas/io/formats/format.py:1014\u001b[0m, in \u001b[0;36mDataFrameRenderer.to_csv\u001b[0;34m(self, path_or_buf, encoding, sep, columns, index_label, mode, compression, quoting, quotechar, lineterminator, chunksize, date_format, doublequote, escapechar, errors, storage_options)\u001b[0m\n\u001b[1;32m    993\u001b[0m     created_buffer \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mFalse\u001b[39;00m\n\u001b[1;32m    995\u001b[0m csv_formatter \u001b[38;5;241m=\u001b[39m CSVFormatter(\n\u001b[1;32m    996\u001b[0m     path_or_buf\u001b[38;5;241m=\u001b[39mpath_or_buf,\n\u001b[1;32m    997\u001b[0m     lineterminator\u001b[38;5;241m=\u001b[39mlineterminator,\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m   1012\u001b[0m     formatter\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mfmt,\n\u001b[1;32m   1013\u001b[0m )\n\u001b[0;32m-> 1014\u001b[0m \u001b[43mcsv_formatter\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43msave\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1016\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m created_buffer:\n\u001b[1;32m   1017\u001b[0m     \u001b[38;5;28;01massert\u001b[39;00m \u001b[38;5;28misinstance\u001b[39m(path_or_buf, StringIO)\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/pandas/io/formats/csvs.py:251\u001b[0m, in \u001b[0;36mCSVFormatter.save\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    247\u001b[0m \u001b[38;5;250m\u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m    248\u001b[0m \u001b[38;5;124;03mCreate the writer & save.\u001b[39;00m\n\u001b[1;32m    249\u001b[0m \u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m    250\u001b[0m \u001b[38;5;66;03m# apply compression and byte/text conversion\u001b[39;00m\n\u001b[0;32m--> 251\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m \u001b[43mget_handle\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m    252\u001b[0m \u001b[43m    \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfilepath_or_buffer\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    253\u001b[0m \u001b[43m    \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mmode\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    254\u001b[0m \u001b[43m    \u001b[49m\u001b[43mencoding\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mencoding\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    255\u001b[0m \u001b[43m    \u001b[49m\u001b[43merrors\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43merrors\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    256\u001b[0m \u001b[43m    \u001b[49m\u001b[43mcompression\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcompression\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    257\u001b[0m \u001b[43m    \u001b[49m\u001b[43mstorage_options\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstorage_options\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    258\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m \u001b[38;5;28;01mas\u001b[39;00m handles:\n\u001b[1;32m    259\u001b[0m     \u001b[38;5;66;03m# Note: self.encoding is irrelevant here\u001b[39;00m\n\u001b[1;32m    260\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mwriter \u001b[38;5;241m=\u001b[39m csvlib\u001b[38;5;241m.\u001b[39mwriter(\n\u001b[1;32m    261\u001b[0m         handles\u001b[38;5;241m.\u001b[39mhandle,\n\u001b[1;32m    262\u001b[0m         lineterminator\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mlineterminator,\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    267\u001b[0m         quotechar\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mquotechar,\n\u001b[1;32m    268\u001b[0m     )\n\u001b[1;32m    270\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_save()\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/pandas/io/common.py:749\u001b[0m, in \u001b[0;36mget_handle\u001b[0;34m(path_or_buf, mode, encoding, compression, memory_map, is_text, errors, storage_options)\u001b[0m\n\u001b[1;32m    747\u001b[0m \u001b[38;5;66;03m# Only for write methods\u001b[39;00m\n\u001b[1;32m    748\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mr\u001b[39m\u001b[38;5;124m\"\u001b[39m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;129;01min\u001b[39;00m mode \u001b[38;5;129;01mand\u001b[39;00m is_path:\n\u001b[0;32m--> 749\u001b[0m     \u001b[43mcheck_parent_directory\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mstr\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43mhandle\u001b[49m\u001b[43m)\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    751\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m compression:\n\u001b[1;32m    752\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m compression \u001b[38;5;241m!=\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mzstd\u001b[39m\u001b[38;5;124m\"\u001b[39m:\n\u001b[1;32m    753\u001b[0m         \u001b[38;5;66;03m# compression libraries do not like an explicit text-mode\u001b[39;00m\n",
      "File \u001b[0;32m/opt/hostedtoolcache/Python/3.9.18/x64/lib/python3.9/site-packages/pandas/io/common.py:616\u001b[0m, in \u001b[0;36mcheck_parent_directory\u001b[0;34m(path)\u001b[0m\n\u001b[1;32m    614\u001b[0m parent \u001b[38;5;241m=\u001b[39m Path(path)\u001b[38;5;241m.\u001b[39mparent\n\u001b[1;32m    615\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m parent\u001b[38;5;241m.\u001b[39mis_dir():\n\u001b[0;32m--> 616\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mOSError\u001b[39;00m(\u001b[38;5;124mrf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mCannot save file into a non-existent directory: \u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;132;01m{\u001b[39;00mparent\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n",
      "\u001b[0;31mOSError\u001b[0m: Cannot save file into a non-existent directory: '/content/drive/MyDrive/misinformation-data'"
     ]
    }
   ],
   "source": [
    "df.to_csv(\"/content/drive/MyDrive/misinformation-data/data_out.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Further detector modules\n",
    "Further detector modules exist, also it is possible to carry out a topic analysis on the text data, as well as crop social media posts automatically. These are more experimental features and have their own demonstration notebooks."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "ammico",
   "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": "3.9.18"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}