prednet_features.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "nrOqVKJeysuV",
    "colab_type": "text"
   },
   "source": [
    "# PredNet feature extraction\n",
    "\n",
    "In this notebook we extract representations from PredNet models trained on videos. We also generate RDM submissions for fMRI and MEG data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "oRh5U1u_ava3",
    "colab_type": "code",
    "outputId": "f03b6cd3-fb79-4cca-aa5c-e6b71e7bb2fc",
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 122.0
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3Aietf%3Awg%3Aoauth%3A2.0%3Aoob&scope=email%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fdocs.test%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fdrive%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fdrive.photos.readonly%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fpeopleapi.readonly&response_type=code\n",
      "\n",
      "Enter your authorization code:\n",
      "··········\n",
      "Mounted at /content/gdrive\n"
     ]
    }
   ],
   "source": [
    "from google.colab import drive\n",
    "drive.mount('/content/gdrive',force_remount=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ul811sAhsS5T",
    "colab_type": "text"
   },
   "source": [
    "### Copying code and supporting files from Google Drive..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "37wYr4GNbmFc",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "!unzip -q gdrive/My\\ Drive/algonauts/algonauts-master.zip -d /content/\n",
    "!unzip -q gdrive/My\\ Drive/algonauts/algonauts-dataset.zip -d /content/\n",
    "!mv algonauts-master/* /content/\n",
    "!cp Feature_Extract/create_RDMs.py /content/\n",
    "!cp prednet/*.py /content/\n",
    "!cp -r gdrive/My\\ Drive/algonauts/prednet-models /content/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "ZWCbc4Mn65N1",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "#!cp -r gdrive/My\\ Drive/algonauts/prednet-feats /content/\n",
    "!mkdir -p prednet-feats/92images\n",
    "!cp -r gdrive/My\\ Drive/algonauts/prednet-feats/92images/alexnet /content/prednet-feats/92images\n",
    "!mkdir -p prednet-feats/78images\n",
    "!cp -r gdrive/My\\ Drive/algonauts/prednet-feats/78images/alexnet /content/prednet-feats/78images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "YFMgK3wMdDBy",
    "colab_type": "code",
    "outputId": "71f1562f-ca69-4687-d026-7e684c4faf64",
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34.0
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import h5py\n",
    "import numpy as np\n",
    "import scipy.io as sio\n",
    "from skimage.measure import block_reduce\n",
    "import os\n",
    "import glob\n",
    "import zipfile\n",
    "from data import DataGenerator\n",
    "np.random.seed(42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "i0aEQdHA0TaA",
    "colab_type": "text"
   },
   "source": [
    "## Extract PredNet features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "v0mhIwC_FDPx",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "from keras import backend as K\n",
    "from keras.layers import Input\n",
    "from keras.models import Model\n",
    "import prednet_model\n",
    "import prednet_rdm\n",
    "import utils\n",
    "\n",
    "\n",
    "def extract_features(images_path, hidden_dims=[256], \n",
    "                     use_multiprocessing=False, workers=1, \n",
    "                     results_dir='./results', input_shape=[10, 256, 256, 3], \n",
    "                     mask_value=None, data_format=K.image_data_format(), \n",
    "                     **config):\n",
    "  \n",
    "    if not os.path.exists(results_dir): \n",
    "      os.makedirs(results_dir)\n",
    "      \n",
    "    input_width = input_shape[1]\n",
    "    input_height = input_shape[2]\n",
    "    prednet = prednet_model.create_model(**config)\n",
    "    #if weights_file:\n",
    "    #    prednet_model.load_prednet_weights(weights_file, prednet)\n",
    "    prednet.summary()\n",
    "    \n",
    "    layer_config = prednet.layers[1].get_config()\n",
    "    data_format = layer_config['data_format'] if 'data_format' in layer_config else layer_config['dim_ordering']\n",
    "    \n",
    "    def resize(img):\n",
    "        return utils.resize_img(img, target_size=(config['input_height'], \n",
    "                                                  config['input_width']))\n",
    "\n",
    "    print('Creating generator...')\n",
    "    data_generator = DataGenerator(classes=None,\n",
    "                                   seq_length=config['n_timesteps'],\n",
    "                                   target_size=None,\n",
    "                                   rescale=config['rescale'],\n",
    "                                   fn_preprocess=resize,\n",
    "                                   batch_size=config['batch_size'], \n",
    "                                   shuffle=False,\n",
    "                                   return_sources=True,\n",
    "                                   data_format=data_format)\n",
    "    \n",
    "    data_generator = data_generator.flow_from_directory(images_path)\n",
    "    \n",
    "    if len(data_generator) == 0:\n",
    "        return\n",
    "    \n",
    "    n_batches = len(data_generator)\n",
    "    print('Number of batches: {}'.format(n_batches))\n",
    "    \n",
    "    data_iterator = iter(data_generator)\n",
    "    for img_i in range(n_batches):\n",
    "        img_data, y_, sources_ = next(data_iterator)\n",
    "    \n",
    "        features = {}\n",
    "        prednet_rep = prednet.predict(img_data, data_generator.batch_size)\n",
    "        # prednet_rep = prednet_rep[:,:, 0:prednet.output_shape[-1]]\n",
    "        print('prednet rep shape:', prednet_rep.shape)\n",
    "        \n",
    "        rep_layers = prednet.layers[1].unflatten_features(img_data.shape, prednet_rep)\n",
    "        rep = []\n",
    "        \n",
    "        width_index = 3\n",
    "        channel_index = -1\n",
    "        block_size = [1, 1, 1, 1, 1]\n",
    "        if data_format == 'channels_first':\n",
    "            width_index = -1\n",
    "            channel_index = 2\n",
    "        \n",
    "        for layer_index, rep_layer in enumerate(rep_layers):\n",
    "            for timestep_index, timestep_rep in enumerate(rep_layer[0]):\n",
    "                # getting only the last timestep\n",
    "                #if timestep_index < config['n_timesteps']-1:\n",
    "                #    continue\n",
    "                \n",
    "                features[f'prednet_{layer_index}_timestep_{timestep_index}'] = timestep_rep\n",
    "                print(f'prednet_{layer_index}_timestep_{timestep_index} shape: {timestep_rep.shape}')\n",
    "        \n",
    "            # Do avg spatial pooling to make all representation layers have\n",
    "            # the same dimension as the higher-level representation layer.\n",
    "            ratio = int(rep_layer.shape[width_index] / rep_layers[-1].shape[width_index])\n",
    "            if ratio > 1:\n",
    "                block_size[width_index-1] = ratio\n",
    "                block_size[width_index] = ratio\n",
    "                rep_layer = block_reduce(rep_layer, block_size=tuple(block_size), func=np.mean)\n",
    "            rep.append(rep_layer)\n",
    "            \n",
    "        rep = np.concatenate(rep, axis=channel_index)\n",
    "        \n",
    "        if data_format == 'channels_first':\n",
    "            rep = np.transpose(rep, (0, 1, 3, 4, 2))\n",
    "\n",
    "        print('prednet all shape', rep[0][-1].shape)\n",
    "        \n",
    "        # features[f'prednet_all'] = rep[0][-1]\n",
    "                \n",
    "        feature_path = os.path.join(results_dir, f'image_{img_i+1:03d}.mat')\n",
    "        print('Saving feature:', feature_path)\n",
    "        sio.savemat(feature_path, features)\n",
    "\n",
    "\n",
    "def extract_features_finetuned(images_path, hidden_dims=[32], \n",
    "                               use_multiprocessing=False, workers=1, \n",
    "                               results_dir='./results', input_shape=[10, 256, 256, 3], \n",
    "                               mask_value=None, data_format=K.image_data_format(), \n",
    "                               model_json_file=None, model_weights_file=None, model_file=None,\n",
    "                               **config):\n",
    "  \n",
    "    if not os.path.exists(results_dir): \n",
    "      os.makedirs(results_dir)\n",
    "      \n",
    "    #print('Loading model: {}'.format(model_json_file))\n",
    "    model = prednet_rdm.create_model(input_shape, hidden_dims, \n",
    "                                     mask_value=mask_value, **config)\n",
    "    # model.load_weights(model_weights_file)\n",
    "    \n",
    "    config['input_channels'] = input_shape[3]\n",
    "    config['input_width'] = input_shape[1]\n",
    "    config['input_height'] = input_shape[2]\n",
    "    config['model_json_file'] = None\n",
    "    config['output_mode'] = 'representation'\n",
    "    \n",
    "    if 'prednet_json_file' in config:\n",
    "        config['model_weights_file'] = config['prednet_weights_file']\n",
    "        config['model_json_file'] = config['prednet_json_file']\n",
    "        prednet = prednet_model.create_model(**config)\n",
    "    else:\n",
    "        weights_file = config['prednet_weights_file']\n",
    "        config['model_weights_file'] = None\n",
    "        prednet = prednet_model.create_model(**config)\n",
    "    \n",
    "        # Unfortunately, keras 'model.load_weights' will not always work\n",
    "        # with prednet layers, so we have to load the weights 'manually' in this case.\n",
    "        prednet_model.load_prednet_weights(weights_file, prednet)\n",
    "    \n",
    "    prednet.summary()\n",
    "    prednet_layer = prednet.layers[1]\n",
    "    \n",
    "    prednet_out = Input(shape=prednet.output_shape[1:])\n",
    "    last_rep = model.get_layer('crop_last_rep_a')(prednet_out)\n",
    "    lstm_layer = model.get_layer('lstm')\n",
    "    lstm_out = lstm_layer(last_rep)\n",
    "    lstm_model = Model(prednet_out, lstm_out, name='lstm_model')    \n",
    "    lstm_model.load_weights(model_weights_file, by_name=True)\n",
    "    #load_model_weights(model_weights_file, lstm_layer)\n",
    "    lstm_model.summary()\n",
    "    \n",
    "    layer_config = prednet_layer.get_config()\n",
    "    data_format = layer_config['data_format'] if 'data_format' in layer_config else layer_config['dim_ordering']\n",
    "    \n",
    "    input_width = input_shape[1]\n",
    "    input_height = input_shape[2]\n",
    "    \n",
    "    def resize(img):\n",
    "        return utils.resize_img(img, target_size=(input_height, input_width))\n",
    "\n",
    "    print('Creating generator...')\n",
    "    data_generator = DataGenerator(classes=None,\n",
    "                                   seq_length=config['n_timesteps'],\n",
    "                                   target_size=None,\n",
    "                                   rescale=config['rescale'],\n",
    "                                   fn_preprocess=resize,\n",
    "                                   batch_size=config['batch_size'], \n",
    "                                   shuffle=False,\n",
    "                                   return_sources=True,\n",
    "                                   data_format=data_format)\n",
    "    \n",
    "    data_generator = data_generator.flow_from_directory(images_path)\n",
    "    \n",
    "    if len(data_generator) == 0:\n",
    "        return\n",
    "    \n",
    "    n_batches = len(data_generator)\n",
    "    print('Number of batches: {}'.format(n_batches))\n",
    "    \n",
    "    data_iterator = iter(data_generator)\n",
    "    for img_i in range(n_batches):\n",
    "        img_data, y_, sources = next(data_iterator)\n",
    "        print(img_i+1, sources[0][0])\n",
    "    \n",
    "        assert img_data.shape == (1, config['n_timesteps'], 256, 256, 3), img_data.shape\n",
    "        \n",
    "        features = {}\n",
    "        \n",
    "        prednet_rep = prednet.predict(img_data, data_generator.batch_size)\n",
    "        prednet_rep = prednet_rep[:,:, 0:prednet.output_shape[-1]]\n",
    "        print('prednet rep shape:', prednet_rep.shape)\n",
    "        \n",
    "        lstm_rep = lstm_model.predict(prednet_rep, data_generator.batch_size)\n",
    "        print('lstm rep shape:', lstm_rep.shape)\n",
    "        features['lstm'] = lstm_rep\n",
    "        \n",
    "        rep_layers = prednet_layer.unflatten_features(img_data.shape, prednet_rep)\n",
    "        rep = []\n",
    "        \n",
    "        width_index = 3\n",
    "        channel_index = -1\n",
    "        block_size = [1, 1, 1, 1, 1]\n",
    "        if data_format == 'channels_first':\n",
    "            width_index = -1\n",
    "            channel_index = 2\n",
    "\n",
    "        for layer_index, rep_layer in enumerate(rep_layers):\n",
    "            for timestep_index, timestep_rep in enumerate(rep_layer[0]):\n",
    "                # getting only the last timestep\n",
    "                #if timestep_index < config['n_timesteps']-1:\n",
    "                #    continue\n",
    "                \n",
    "                features[f'prednet_{layer_index}_timestep_{timestep_index}'] = timestep_rep\n",
    "                print(f'prednet_{layer_index}_timestep_{timestep_index} shape: {timestep_rep.shape}')\n",
    "                \n",
    "                # features[f'prednet_{layer_index}_timestep_{timestep_index}_lstm'] = np.concatenate([timestep_rep.ravel(), lstm_rep.ravel()])\n",
    "                \n",
    "            # Do avg spatial pooling to make all representation layers have\n",
    "            # the same dimension as the higher-level representation layer.\n",
    "            ratio = int(rep_layer.shape[width_index] / rep_layers[-1].shape[width_index])\n",
    "            if ratio > 1:\n",
    "                block_size[width_index-1] = ratio\n",
    "                block_size[width_index] = ratio\n",
    "                rep_layer = block_reduce(rep_layer, block_size=tuple(block_size), func=np.mean)\n",
    "            rep.append(rep_layer)\n",
    "            \n",
    "        rep = np.concatenate(rep, axis=channel_index)\n",
    "        \n",
    "        if data_format == 'channels_first':\n",
    "            rep = np.transpose(rep, (0, 1, 3, 4, 2))\n",
    "\n",
    "        print('prednet all shape', rep[0][-1].shape)\n",
    "        # features[f'prednet_all'] = rep[0][-1]\n",
    "        # features[f'prednet_all_lstm'] = np.concatenate([rep[0][-1].ravel(), lstm_rep.ravel()])\n",
    "                \n",
    "        feature_path = os.path.join(results_dir, f'image_{img_i+1:03d}.mat')\n",
    "        print('Saving feature:', feature_path)\n",
    "        sio.savemat(feature_path, features)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "xZHyCC9Qm2HO",
    "colab_type": "text"
   },
   "source": [
    "## Configuration for several model variants"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "8W-Z0Aj1oI5f",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "config = {\n",
    "    'n_timesteps': 10,\n",
    "    'batch_size': 1,\n",
    "    'stateful': False,\n",
    "    'train': False,\n",
    "    'rescale': 1./255,\n",
    "    'input_height': 128,\n",
    "    'input_width': 160,\n",
    "    'input_channels': 3\n",
    "}\n",
    "config['output_mode'] = 'representation'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "p6pMzJeDYgjN",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "config['results_dir'] = '/content/prednet-feats/92images/prednet_random'\n",
    "extract_features('/content/Training_Data/92_Image_Set/', **config)\n",
    "config['results_dir'] = '/content/prednet-feats/78images/prednet_random'\n",
    "extract_features('/content/Test_Data/', **config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "8pyr-NU6oZ60",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "config['model_json_file'] = '/content/prednet-models/kitti_keras/prednet_kitti_model.json'\n",
    "config['model_weights_file'] = '/content/prednet-models/kitti_keras/prednet_kitti_weights.hdf5'\n",
    "config['results_dir'] = '/content/prednet-feats/92images/kitti_keras'\n",
    "extract_features('/content/Training_Data/92_Image_Set/', **config)\n",
    "config['results_dir'] = '/content/prednet-feats/78images/kitti_keras'\n",
    "extract_features('/content/Test_Data/', **config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "fUieZhc9oPIK",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "config['model_json_file'] = '/content/prednet-models/prednet_kitti__moments__model__3c/model.json'\n",
    "config['model_weights_file'] = '/content/prednet-models/prednet_kitti__moments__model__3c/weights.hdf5'\n",
    "config['results_dir'] = '/content/prednet-feats/92images/prednet_kitti__moments__model__3c'\n",
    "extract_features('/content/Training_Data/92_Image_Set/', **config)\n",
    "config['results_dir'] = '/content/prednet-feats/78images/prednet_kitti__moments__model__3c'\n",
    "extract_features('/content/Test_Data/', **config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "Z-Dyz4mToRZW",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "config['model_json_file'] = '/content/prednet-models/prednet_kitti__moments__model__10c/model.json'\n",
    "config['model_weights_file'] = '/content/prednet-models/prednet_kitti__moments__model__10c/weights.hdf5'\n",
    "config['results_dir'] = '/content/prednet-feats/92images/prednet_kitti__moments__model__10c'\n",
    "extract_features('/content/Training_Data/92_Image_Set/', **config)\n",
    "config['results_dir'] = '/content/prednet-feats/78images/prednet_kitti__moments__model__10c'\n",
    "extract_features('/content/Test_Data/', **config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "VLLCbuX9pSRL",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "config['model_json_file'] = '/content/prednet-models/prednet_kitti__moments__model__full/model.json'\n",
    "config['model_weights_file'] = '/content/prednet-models/prednet_kitti__moments__model__full/weights.hdf5'\n",
    "config['results_dir'] = '/content/prednet-feats/92images/prednet_kitti__moments__model__full'\n",
    "extract_features('/content/Training_Data/92_Image_Set/', **config)\n",
    "config['results_dir'] = '/content/prednet-feats/78images/prednet_kitti__moments__model__full'\n",
    "extract_features('/content/Test_Data/', **config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "WgjV448C4UjC",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "config['model_json_file'] = '/content/prednet-models/prednet_finetuned_moments__algonauts/model.json'\n",
    "config['model_weights_file'] = '/content/prednet-models/prednet_finetuned_moments__algonauts/weights.hdf5'\n",
    "config['results_dir'] = '/content/results/feats/92images/prednet_finetuned_moments__algonauts'\n",
    "extract_features('/content/Training_Data/92_Image_Set/', **config)\n",
    "config['results_dir'] = '/content/results/feats/78images/prednet_finetuned_moments__algonauts'\n",
    "extract_features('/content/Test_Data/', **config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "jy_vCN_lbaK9",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "config['model_json_file'] = '/content/prednet-models/prednet_random__algonauts/model.json'\n",
    "#config['model_weights_file'] = '/content/prednet-models/prednet_random__algonauts/weights.hdf5'\n",
    "config['results_dir'] = '/content/results/feats/92images/prednet_random__algonauts'\n",
    "extract_features('/content/Training_Data/92_Image_Set/', **config)\n",
    "config['results_dir'] = '/content/results/feats/78images/prednet_random__algonauts'\n",
    "extract_features('/content/Test_Data/', **config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "qWPJNOHcoTJ2",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "config['input_height'] = 256\n",
    "config['input_width'] = 256\n",
    "config['stack_sizes'] = (48, 96, 192, 192)\n",
    "config['results_dir'] = '/content/prednet-feats/92images/prednet_moments__model__4_layers_epoch_1'\n",
    "config['model_json_file'] = '/content/prednet-models/prednet_moments__model__4_layers_epoch_1/model.json'\n",
    "config['model_weights_file'] = '/content/prednet-models/prednet_moments__model__4_layers_epoch_1/weights.hdf5'\n",
    "extract_features('/content/Training_Data/92_Image_Set/', **config)\n",
    "config['results_dir'] = '/content/prednet-feats/78images/prednet_moments__model__4_layers_epoch_1'\n",
    "extract_features('/content/Test_Data/', **config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "cIOMIn0_8-tO",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "config['input_height'] = 256\n",
    "config['input_width'] = 256\n",
    "config['stack_sizes'] = (48, 96, 192)\n",
    "config['output_mode'] = 'representation_and_error'\n",
    "config['results_dir'] = '/content/prednet-feats/92images/prednet_random_rdm_10000'\n",
    "config['model_json_file'] = '/content/prednet-models/prednet_random_rdm_10000/model.json'\n",
    "config['model_weights_file'] = '/content/prednet-models/prednet_random_rdm_10000/weights.hdf5'\n",
    "config['prednet_json_file'] = '/content/prednet-models/prednet_random_rdm_10000/prednet_model.json'\n",
    "config['prednet_weights_file'] = '/content/prednet-models/prednet_random_rdm_10000/prednet_weights.h5'\n",
    "extract_features_finetuned('/content/Training_Data/92_Image_Set/', hidden_dims=[32], **config)\n",
    "config['results_dir'] = '/content/prednet-feats/78images/prednet_random_rdm_10000'\n",
    "extract_features_finetuned('/content/Test_Data/', hidden_dims=[32], **config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "Ann5KBUszs4I",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "config['input_height'] = 256\n",
    "config['input_width'] = 256\n",
    "config['stack_sizes'] = (48, 96, 192, 192)\n",
    "config['output_mode'] = 'representation_and_error'\n",
    "config['results_dir'] = '/content/prednet-feats/92images/prednet_moments_fmri_it_rdm'\n",
    "config['model_json_file'] = '/content/prednet-models/prednet_moments_fmri_it_rdm/model.json'\n",
    "config['model_weights_file'] = '/content/prednet-models/prednet_moments_fmri_it_rdm/weights.hdf5'\n",
    "#config['prednet_json_file'] = '/content/prednet-models/prednet_moments_fmri_it_rdm/prednet_model.json'\n",
    "config['prednet_weights_file'] = '/content/prednet-models/prednet_moments_fmri_it_rdm/prednet_weights.h5'\n",
    "extract_features_finetuned('/content/Training_Data/92_Image_Set/', hidden_dims=[128], **config)\n",
    "config['results_dir'] = '/content/prednet-feats/78images/prednet_moments_fmri_it_rdm'\n",
    "extract_features_finetuned('/content/Test_Data/', hidden_dims=[128], **config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "PKqL7zRHo4SN",
    "colab_type": "code",
    "outputId": "83bc4597-2643-4213-dfda-a3958b5aac61",
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000.0
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PredNet trainable layers: None\n",
      "PredNet trainable units: None\n",
      "Layer a_0 trainable params: 2640\n",
      "Layer a_1 trainable params: 83040\n",
      "Layer a_2 trainable params: 331968\n",
      "Layer a_3 trainable params: 663744\n",
      "Layer ahat_0 trainable params: 84\n",
      "Layer ahat_1 trainable params: 20784\n",
      "Layer ahat_2 trainable params: 83040\n",
      "Layer ahat_3 trainable params: 331968\n",
      "Layer ahat_4 trainable params: 331968\n",
      "Layer c_0 trainable params: 1542\n",
      "Layer c_1 trainable params: 103728\n",
      "Layer c_2 trainable params: 414816\n",
      "Layer c_3 trainable params: 1327296\n",
      "Layer c_4 trainable params: 995520\n",
      "Layer f_0 trainable params: 1542\n",
      "Layer f_1 trainable params: 103728\n",
      "Layer f_2 trainable params: 414816\n",
      "Layer f_3 trainable params: 1327296\n",
      "Layer f_4 trainable params: 995520\n",
      "Layer i_0 trainable params: 1542\n",
      "Layer i_1 trainable params: 103728\n",
      "Layer i_2 trainable params: 414816\n",
      "Layer i_3 trainable params: 1327296\n",
      "Layer i_4 trainable params: 995520\n",
      "Layer o_0 trainable params: 1542\n",
      "Layer o_1 trainable params: 103728\n",
      "Layer o_2 trainable params: 414816\n",
      "Layer o_3 trainable params: 1327296\n",
      "Layer o_4 trainable params: 995520\n",
      "PredNet trainable params: 13220844\n",
      "PredNet non-trainable params: 0\n",
      "Layer 0 shape: (256, 256, 3) (196608)\n",
      "Layer 1 shape: (128, 128, 48) (786432)\n",
      "Layer 2 shape: (64, 64, 96) (393216)\n",
      "Layer 3 shape: (32, 32, 192) (196608)\n",
      "Layer 4 shape: (16, 16, 192) (49152)\n",
      "Layer 0 shape: (256, 256, 3) (196608)\n",
      "Layer 1 shape: (128, 128, 48) (786432)\n",
      "Layer 2 shape: (64, 64, 96) (393216)\n",
      "Layer 3 shape: (32, 32, 192) (196608)\n",
      "Layer 4 shape: (16, 16, 192) (49152)\n",
      "Layer 0 shape: (256, 256, 3) (196608)\n",
      "Layer 1 shape: (128, 128, 48) (786432)\n",
      "Layer 2 shape: (64, 64, 96) (393216)\n",
      "Layer 3 shape: (32, 32, 192) (196608)\n",
      "Layer 4 shape: (16, 16, 192) (49152)\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "image_a (InputLayer)            (None, 10, 256, 256, 0                                            \n",
      "__________________________________________________________________________________________________\n",
      "image_b (InputLayer)            (None, 10, 256, 256, 0                                            \n",
      "__________________________________________________________________________________________________\n",
      "PredNet (Model)                 multiple             13220844    image_a[0][0]                    \n",
      "                                                                 image_b[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "crop_error_a (Lambda)           (None, 10, 5)        0           PredNet[1][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "crop_error_b (Lambda)           (None, 10, 5)        0           PredNet[2][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "crop_rep_a (Lambda)             (None, 10, 1622016)  0           PredNet[1][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "crop_rep_b (Lambda)             (None, 10, 1622016)  0           PredNet[2][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "time_distributed_5 (TimeDistrib (None, 10, 1)        6           crop_error_a[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "time_distributed_6 (TimeDistrib (None, 10, 1)        6           crop_error_b[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "crop_last_rep_a (Lambda)        (None, 10, 49152)    0           crop_rep_a[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "crop_last_rep_b (Lambda)        (None, 10, 49152)    0           crop_rep_b[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "flatten_7 (Flatten)             (None, 10)           0           time_distributed_5[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "flatten_8 (Flatten)             (None, 10)           0           time_distributed_6[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "lstm (Model)                    (None, 32)           6295680     crop_last_rep_a[0][0]            \n",
      "                                                                 crop_last_rep_b[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "dense_10 (Dense)                (None, 1)            11          flatten_7[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "dense_12 (Dense)                (None, 1)            11          flatten_8[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "image_pair_representation (Conc (None, 64)           0           lstm[1][0]                       \n",
      "                                                                 lstm[2][0]                       \n",
      "__________________________________________________________________________________________________\n",
      "prednet_error (Average)         (None, 1)            0           dense_10[0][0]                   \n",
      "                                                                 dense_12[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "rdm_prediction (Dense)          (None, 1)            65          image_pair_representation[0][0]  \n",
      "==================================================================================================\n",
      "Total params: 19,516,623\n",
      "Trainable params: 6,295,745\n",
      "Non-trainable params: 13,220,878\n",
      "__________________________________________________________________________________________________\n",
      "PredNet trainable layers: None\n",
      "PredNet trainable units: None\n",
      "Layer a_0 trainable params: 2640\n",
      "Layer a_1 trainable params: 83040\n",
      "Layer a_2 trainable params: 331968\n",
      "Layer a_3 trainable params: 663744\n",
      "Layer ahat_0 trainable params: 84\n",
      "Layer ahat_1 trainable params: 20784\n",
      "Layer ahat_2 trainable params: 83040\n",
      "Layer ahat_3 trainable params: 331968\n",
      "Layer ahat_4 trainable params: 331968\n",
      "Layer c_0 trainable params: 1542\n",
      "Layer c_1 trainable params: 103728\n",
      "Layer c_2 trainable params: 414816\n",
      "Layer c_3 trainable params: 1327296\n",
      "Layer c_4 trainable params: 995520\n",
      "Layer f_0 trainable params: 1542\n",
      "Layer f_1 trainable params: 103728\n",
      "Layer f_2 trainable params: 414816\n",
      "Layer f_3 trainable params: 1327296\n",
      "Layer f_4 trainable params: 995520\n",
      "Layer i_0 trainable params: 1542\n",
      "Layer i_1 trainable params: 103728\n",
      "Layer i_2 trainable params: 414816\n",
      "Layer i_3 trainable params: 1327296\n",
      "Layer i_4 trainable params: 995520\n",
      "Layer o_0 trainable params: 1542\n",
      "Layer o_1 trainable params: 103728\n",
      "Layer o_2 trainable params: 414816\n",
      "Layer o_3 trainable params: 1327296\n",
      "Layer o_4 trainable params: 995520\n",
      "PredNet trainable params: 13220844\n",
      "PredNet non-trainable params: 0\n",
      "Layer 0 shape: (256, 256, 3) (196608)\n",
      "Layer 1 shape: (128, 128, 48) (786432)\n",
      "Layer 2 shape: (64, 64, 96) (393216)\n",
      "Layer 3 shape: (32, 32, 192) (196608)\n",
      "Layer 4 shape: (16, 16, 192) (49152)\n",
      "58/58 weights loaded from /content/prednet-models/prednet_moments_meg_late_rdm_10000/prednet_weights.h5\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_11 (InputLayer)        (1, 10, 256, 256, 3)      0         \n",
      "_________________________________________________________________\n",
      "prednet_layer (PredNet)      (1, 10, 1622016)          13220844  \n",
      "=================================================================\n",
      "Total params: 13,220,844\n",
      "Trainable params: 13,220,844\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_12 (InputLayer)        (None, 10, 1622016)       0         \n",
      "_________________________________________________________________\n",
      "crop_last_rep_a (Lambda)     (None, 10, 49152)         0         \n",
      "_________________________________________________________________\n",
      "lstm (Model)                 (None, 32)                6295680   \n",
      "=================================================================\n",
      "Total params: 6,295,680\n",
      "Trainable params: 6,295,680\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Creating generator...\n",
      "Found 78 samples belonging to 1 classes in /content/Test_Data/\n",
      "Number of batches: 78\n",
      "1 /content/Test_Data/78images/image_01.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_001.mat\n",
      "2 /content/Test_Data/78images/image_02.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_002.mat\n",
      "3 /content/Test_Data/78images/image_03.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_003.mat\n",
      "4 /content/Test_Data/78images/image_04.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_004.mat\n",
      "5 /content/Test_Data/78images/image_05.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_005.mat\n",
      "6 /content/Test_Data/78images/image_06.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_006.mat\n",
      "7 /content/Test_Data/78images/image_07.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_007.mat\n",
      "8 /content/Test_Data/78images/image_08.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_008.mat\n",
      "9 /content/Test_Data/78images/image_09.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_009.mat\n",
      "10 /content/Test_Data/78images/image_10.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_010.mat\n",
      "11 /content/Test_Data/78images/image_11.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_011.mat\n",
      "12 /content/Test_Data/78images/image_12.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_012.mat\n",
      "13 /content/Test_Data/78images/image_13.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_013.mat\n",
      "14 /content/Test_Data/78images/image_14.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_014.mat\n",
      "15 /content/Test_Data/78images/image_15.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_015.mat\n",
      "16 /content/Test_Data/78images/image_16.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_016.mat\n",
      "17 /content/Test_Data/78images/image_17.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_017.mat\n",
      "18 /content/Test_Data/78images/image_18.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_018.mat\n",
      "19 /content/Test_Data/78images/image_19.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_019.mat\n",
      "20 /content/Test_Data/78images/image_20.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_020.mat\n",
      "21 /content/Test_Data/78images/image_21.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_021.mat\n",
      "22 /content/Test_Data/78images/image_22.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_022.mat\n",
      "23 /content/Test_Data/78images/image_23.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_023.mat\n",
      "24 /content/Test_Data/78images/image_24.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_024.mat\n",
      "25 /content/Test_Data/78images/image_25.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_025.mat\n",
      "26 /content/Test_Data/78images/image_26.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_026.mat\n",
      "27 /content/Test_Data/78images/image_27.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_027.mat\n",
      "28 /content/Test_Data/78images/image_28.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_028.mat\n",
      "29 /content/Test_Data/78images/image_29.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_029.mat\n",
      "30 /content/Test_Data/78images/image_30.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_030.mat\n",
      "31 /content/Test_Data/78images/image_31.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_031.mat\n",
      "32 /content/Test_Data/78images/image_32.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_032.mat\n",
      "33 /content/Test_Data/78images/image_33.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_033.mat\n",
      "34 /content/Test_Data/78images/image_34.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_034.mat\n",
      "35 /content/Test_Data/78images/image_35.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_035.mat\n",
      "36 /content/Test_Data/78images/image_36.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_036.mat\n",
      "37 /content/Test_Data/78images/image_37.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_037.mat\n",
      "38 /content/Test_Data/78images/image_38.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_038.mat\n",
      "39 /content/Test_Data/78images/image_39.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_039.mat\n",
      "40 /content/Test_Data/78images/image_40.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_040.mat\n",
      "41 /content/Test_Data/78images/image_41.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_041.mat\n",
      "42 /content/Test_Data/78images/image_42.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_042.mat\n",
      "43 /content/Test_Data/78images/image_43.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_043.mat\n",
      "44 /content/Test_Data/78images/image_44.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_044.mat\n",
      "45 /content/Test_Data/78images/image_45.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_045.mat\n",
      "46 /content/Test_Data/78images/image_46.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_046.mat\n",
      "47 /content/Test_Data/78images/image_47.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_047.mat\n",
      "48 /content/Test_Data/78images/image_48.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_048.mat\n",
      "49 /content/Test_Data/78images/image_49.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_049.mat\n",
      "50 /content/Test_Data/78images/image_50.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_050.mat\n",
      "51 /content/Test_Data/78images/image_51.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_051.mat\n",
      "52 /content/Test_Data/78images/image_52.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_052.mat\n",
      "53 /content/Test_Data/78images/image_53.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_053.mat\n",
      "54 /content/Test_Data/78images/image_54.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_054.mat\n",
      "55 /content/Test_Data/78images/image_55.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_055.mat\n",
      "56 /content/Test_Data/78images/image_56.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_056.mat\n",
      "57 /content/Test_Data/78images/image_57.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_057.mat\n",
      "58 /content/Test_Data/78images/image_58.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_058.mat\n",
      "59 /content/Test_Data/78images/image_59.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_059.mat\n",
      "60 /content/Test_Data/78images/image_60.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_060.mat\n",
      "61 /content/Test_Data/78images/image_61.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_061.mat\n",
      "62 /content/Test_Data/78images/image_62.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_062.mat\n",
      "63 /content/Test_Data/78images/image_63.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_063.mat\n",
      "64 /content/Test_Data/78images/image_64.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_064.mat\n",
      "65 /content/Test_Data/78images/image_65.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_065.mat\n",
      "66 /content/Test_Data/78images/image_66.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_066.mat\n",
      "67 /content/Test_Data/78images/image_67.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_067.mat\n",
      "68 /content/Test_Data/78images/image_68.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_068.mat\n",
      "69 /content/Test_Data/78images/image_69.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_069.mat\n",
      "70 /content/Test_Data/78images/image_70.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_070.mat\n",
      "71 /content/Test_Data/78images/image_71.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_071.mat\n",
      "72 /content/Test_Data/78images/image_72.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_072.mat\n",
      "73 /content/Test_Data/78images/image_73.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_073.mat\n",
      "74 /content/Test_Data/78images/image_74.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_074.mat\n",
      "75 /content/Test_Data/78images/image_75.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_075.mat\n",
      "76 /content/Test_Data/78images/image_76.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_076.mat\n",
      "77 /content/Test_Data/78images/image_77.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_077.mat\n",
      "78 /content/Test_Data/78images/image_78.jpg\n",
      "prednet rep shape: (1, 10, 1622016)\n",
      "lstm rep shape: (1, 32)\n",
      "prednet_0_timestep_0 shape: (256, 256, 3)\n",
      "prednet_0_timestep_1 shape: (256, 256, 3)\n",
      "prednet_0_timestep_2 shape: (256, 256, 3)\n",
      "prednet_0_timestep_3 shape: (256, 256, 3)\n",
      "prednet_0_timestep_4 shape: (256, 256, 3)\n",
      "prednet_0_timestep_5 shape: (256, 256, 3)\n",
      "prednet_0_timestep_6 shape: (256, 256, 3)\n",
      "prednet_0_timestep_7 shape: (256, 256, 3)\n",
      "prednet_0_timestep_8 shape: (256, 256, 3)\n",
      "prednet_0_timestep_9 shape: (256, 256, 3)\n",
      "prednet_1_timestep_0 shape: (128, 128, 48)\n",
      "prednet_1_timestep_1 shape: (128, 128, 48)\n",
      "prednet_1_timestep_2 shape: (128, 128, 48)\n",
      "prednet_1_timestep_3 shape: (128, 128, 48)\n",
      "prednet_1_timestep_4 shape: (128, 128, 48)\n",
      "prednet_1_timestep_5 shape: (128, 128, 48)\n",
      "prednet_1_timestep_6 shape: (128, 128, 48)\n",
      "prednet_1_timestep_7 shape: (128, 128, 48)\n",
      "prednet_1_timestep_8 shape: (128, 128, 48)\n",
      "prednet_1_timestep_9 shape: (128, 128, 48)\n",
      "prednet_2_timestep_0 shape: (64, 64, 96)\n",
      "prednet_2_timestep_1 shape: (64, 64, 96)\n",
      "prednet_2_timestep_2 shape: (64, 64, 96)\n",
      "prednet_2_timestep_3 shape: (64, 64, 96)\n",
      "prednet_2_timestep_4 shape: (64, 64, 96)\n",
      "prednet_2_timestep_5 shape: (64, 64, 96)\n",
      "prednet_2_timestep_6 shape: (64, 64, 96)\n",
      "prednet_2_timestep_7 shape: (64, 64, 96)\n",
      "prednet_2_timestep_8 shape: (64, 64, 96)\n",
      "prednet_2_timestep_9 shape: (64, 64, 96)\n",
      "prednet_3_timestep_0 shape: (32, 32, 192)\n",
      "prednet_3_timestep_1 shape: (32, 32, 192)\n",
      "prednet_3_timestep_2 shape: (32, 32, 192)\n",
      "prednet_3_timestep_3 shape: (32, 32, 192)\n",
      "prednet_3_timestep_4 shape: (32, 32, 192)\n",
      "prednet_3_timestep_5 shape: (32, 32, 192)\n",
      "prednet_3_timestep_6 shape: (32, 32, 192)\n",
      "prednet_3_timestep_7 shape: (32, 32, 192)\n",
      "prednet_3_timestep_8 shape: (32, 32, 192)\n",
      "prednet_3_timestep_9 shape: (32, 32, 192)\n",
      "prednet_4_timestep_0 shape: (16, 16, 192)\n",
      "prednet_4_timestep_1 shape: (16, 16, 192)\n",
      "prednet_4_timestep_2 shape: (16, 16, 192)\n",
      "prednet_4_timestep_3 shape: (16, 16, 192)\n",
      "prednet_4_timestep_4 shape: (16, 16, 192)\n",
      "prednet_4_timestep_5 shape: (16, 16, 192)\n",
      "prednet_4_timestep_6 shape: (16, 16, 192)\n",
      "prednet_4_timestep_7 shape: (16, 16, 192)\n",
      "prednet_4_timestep_8 shape: (16, 16, 192)\n",
      "prednet_4_timestep_9 shape: (16, 16, 192)\n",
      "prednet all shape (16, 16, 531)\n",
      "Saving feature: /content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000/image_078.mat\n"
     ]
    }
   ],
   "source": [
    "config['input_height'] = 256\n",
    "config['input_width'] = 256\n",
    "config['stack_sizes'] = (48, 96, 192, 192)\n",
    "config['output_mode'] = 'representation_and_error'\n",
    "config['results_dir'] = '/content/prednet-feats/92images/prednet_moments_meg_late_rdm_10000'\n",
    "config['model_json_file'] = '/content/prednet-models/prednet_moments_meg_late_rdm_10000/model.json'\n",
    "config['model_weights_file'] = '/content/prednet-models/prednet_moments_meg_late_rdm_10000/weights.hdf5'\n",
    "#config['prednet_json_file'] = '/content/prednet-models/prednet_moments_meg_late_rdm_10000/prednet_model.json'\n",
    "config['prednet_weights_file'] = '/content/prednet-models/prednet_moments_meg_late_rdm_10000/prednet_weights.h5'\n",
    "#extract_features_finetuned('/content/Training_Data/92_Image_Set/', **config)\n",
    "config['results_dir'] = '/content/prednet-feats/78images/prednet_moments_meg_late_rdm_10000'\n",
    "extract_features_finetuned('/content/Test_Data/', **config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "Z_GPPDWWeZZH",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "!cp -r /content/prednet-feats/ /content/gdrive/My\\ Drive/prednet-feats-$(date +%Y-%m-%dT%H:%M:%S)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "prPFO0n-bprn",
    "colab_type": "text"
   },
   "source": [
    "## Generate RDMs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "1v5ChynyeSjp",
    "colab_type": "code",
    "outputId": "f0388b2a-06d9-4c8b-87ed-04a7dd8a8818",
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34.0
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(15, 118, 118)"
      ]
     },
     "execution_count": 12,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def loadmat(matfile):\n",
    "    try:\n",
    "        f = h5py.File(matfile)\n",
    "    except (IOError, OSError):\n",
    "        return sio.loadmat(matfile)\n",
    "    else:\n",
    "        return {name: np.transpose(f.get(name)) for name in f.keys()}\n",
    "\n",
    "def loadnpy(npyfile):\n",
    "    return np.load(npyfile)\n",
    "\n",
    "def load(data_file):\n",
    "    root, ext = os.path.splitext(data_file)\n",
    "    return {'.npy': loadnpy,\n",
    "            '.mat': loadmat\n",
    "            }.get(ext, loadnpy)(data_file)\n",
    "\n",
    "human_fmri = loadmat('./Training_Data/118_Image_Set/target_fmri.mat')\n",
    "human_fmri['EVC_RDMs'].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "JYFUHMUikStP",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "import create_RDMs\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.decomposition import PCA\n",
    "\n",
    "def get_features_pca(feat_dir):\n",
    "    activations = sorted(glob.glob(feat_dir + \"/*\" + \".mat\"))\n",
    "    \n",
    "    layer_ids = sorted(sio.loadmat(activations[0]).keys())\n",
    "    layer_feats_pca = {}\n",
    "    \n",
    "    for layer_id in layer_ids:\n",
    "      \n",
    "        if layer_id in ['__header__', '__version__', '__globals__']:\n",
    "            continue\n",
    "            \n",
    "        print('Processing features for layer', layer_id)\n",
    "        layer_feats = []\n",
    "        \n",
    "        for img_feats_path in activations:\n",
    "            img_feats = sio.loadmat(img_feats_path)\n",
    "            layer_feats.append(img_feats[layer_id].ravel())\n",
    "            \n",
    "        layer_feats = np.array(layer_feats)\n",
    "        \n",
    "        pca = PCA(n_components=0.99, svd_solver='full')\n",
    "        scaler = StandardScaler()\n",
    "        feats_scaled = scaler.fit_transform(layer_feats)\n",
    "        feats_pca = pca.fit_transform(feats_scaled)\n",
    "        layer_feats_pca[layer_id] = feats_pca\n",
    "            \n",
    "    return layer_feats_pca\n",
    "\n",
    "def get_layers_ncondns(feat_dir):\n",
    "    \"\"\"\n",
    "    to get number of representative layers in the DNN,\n",
    "    and number of images(conditions).\n",
    "    Input:\n",
    "    feat_dir: Directory containing activations generated using generate_features.py\n",
    "    Output:\n",
    "    num_layers: number of layers for which activations were generated\n",
    "    num_condns: number of stimulus images\n",
    "    PS: This function is specific for activations generated using generate_features.py\n",
    "    Write your own function in case you use different script to generate features.\n",
    "    \"\"\"\n",
    "    activations = glob.glob(feat_dir + \"/*\" + \".mat\")\n",
    "    num_condns = len(activations)\n",
    "    feat=sio.loadmat(activations[0])\n",
    "    num_layers = 0\n",
    "    layer_list = []\n",
    "    for key in feat:\n",
    "        if key in ['__header__', '__version__', '__globals__']:\n",
    "            continue\n",
    "            \n",
    "        #if 'prednet_2' not in key:\n",
    "        #    continue\n",
    "            \n",
    "        else:\n",
    "            num_layers+=1\n",
    "            layer_list.append(key)\n",
    "    return num_layers,layer_list,num_condns\n",
    "\n",
    "def create_rdm(save_dir,feat_dir,dist):\n",
    "    \"\"\"\n",
    "    Main function to create RDM from activations\n",
    "    Input:\n",
    "    feat_dir: Directory containing activations generated using generate_features.py\n",
    "    save_dir : directory to save the computed RDM\n",
    "    dist : dist used for computing RDM (e.g. 1-Pearson's R)\n",
    "\n",
    "    Output (in submission format):\n",
    "    The model RDMs for each layer are saved in\n",
    "        save_dir/layer_name/submit_fMRI.mat to compare with fMRI RDMs\n",
    "        save_dir/layer_name/submit_MEG.mat to compare with MEG RDMs\n",
    "    \"\"\"\n",
    "\n",
    "    #get number of layers and number of conditions(images) for RDM\n",
    "    num_layers,layer_list, num_condns = get_layers_ncondns(feat_dir)\n",
    "    cwd = os.getcwd()\n",
    "    \n",
    "    #feats = sio.loadmat(os.path.join(feat_dir, 'feats_pca.mat'))\n",
    "    #layer_list = sorted(feats.keys())\n",
    "    \n",
    "    #num_condns = 0\n",
    "    #for layer_id in layer_list:\n",
    "    #    if isinstance(feats[layer_id], np.ndarray) and len(feats[layer_id]) > 0:\n",
    "    #        num_condns = len(feats[layer_id])\n",
    "    #        break\n",
    "            \n",
    "    # loops over layers and create RDM for each layer\n",
    "    for layer in range(num_layers):\n",
    "        layer_id = layer_list[layer]\n",
    "        if layer_id in ['__header__', '__version__', '__globals__']:\n",
    "            continue\n",
    "            \n",
    "        if 'timestep_0' in layer_id:\n",
    "            print('Skipping layer', layer_id)\n",
    "            continue\n",
    "      \n",
    "        os.chdir(cwd)\n",
    "        #RDM is num_condnsxnum_condns matrix, initialized with zeros\n",
    "        RDM = np.zeros((num_condns,num_condns))\n",
    "\n",
    "        #save path for RDMs in  challenge submission format\n",
    "        print(f'Processing layer {layer_id}...')\n",
    "        RDM_dir = os.path.join(save_dir,layer_id)\n",
    "        if not os.path.exists(RDM_dir):\n",
    "            os.makedirs(RDM_dir)\n",
    "        RDM_filename_meg = os.path.join(RDM_dir,'submit_meg.mat')\n",
    "        RDM_filename_fmri = os.path.join(RDM_dir,'submit_fmri.mat')\n",
    "        RDM_filename_meg_zip = os.path.join(RDM_dir,'submit_meg.zip')\n",
    "        RDM_filename_fmri_zip = os.path.join(RDM_dir,'submit_fmri.zip')\n",
    "        #RDM loop\n",
    "        for i in range(num_condns):\n",
    "            for j in range(num_condns):\n",
    "                #get feature for image index i and j\n",
    "                feature_i = create_RDMs.get_features(feat_dir,layer_id,i)\n",
    "                feature_j = create_RDMs.get_features(feat_dir,layer_id,j)\n",
    "                \n",
    "                #feature_i = feats[layer_id][i].ravel()\n",
    "                #feature_j = feats[layer_id][j].ravel()\n",
    "                \n",
    "                # compute distance 1-Pearson's R\n",
    "                if dist == 'pearson':\n",
    "                    RDM[i,j] = 1-np.corrcoef(feature_i,feature_j)[0][1]\n",
    "                else:\n",
    "                    print(\"The\", dist, \"distance measure not implemented, please request through issues\")\n",
    "\n",
    "        #saving RDMs in challenge submission format\n",
    "        rdm_fmri={}\n",
    "        rdm_meg={}\n",
    "        rdm_fmri['EVC_RDMs'] = RDM\n",
    "        rdm_fmri['IT_RDMs'] = RDM\n",
    "        rdm_meg['MEG_RDMs_late'] = RDM\n",
    "        rdm_meg['MEG_RDMs_early'] = RDM\n",
    "        sio.savemat(RDM_filename_fmri,rdm_fmri)\n",
    "        sio.savemat(RDM_filename_meg,rdm_meg)\n",
    "\n",
    "        #creating zipped file for submission\n",
    "        zipfmri = zipfile.ZipFile(RDM_filename_fmri_zip, 'w')\n",
    "        zipmeg = zipfile.ZipFile(RDM_filename_meg_zip, 'w')\n",
    "        os.chdir(RDM_dir)\n",
    "        zipfmri.write('submit_fmri.mat')\n",
    "        zipmeg.write('submit_meg.mat')\n",
    "        zipfmri.close()\n",
    "        zipmeg.close()\n",
    "        \n",
    "    os.chdir(cwd)\n",
    "    \n",
    "def create_rdms(feat_dir, rdm_dir):\n",
    "    \"\"\"\n",
    "    Create RDMs for all features (from all models) in folder feat_dir.\n",
    "    \"\"\"\n",
    "    for feat_path in glob.glob(os.path.join(feat_dir, '*')):\n",
    "        model_name = os.path.basename(feat_path)\n",
    "        print(f'Calculating RDMs for model {model_name}')\n",
    "\n",
    "        rdm_path = os.path.join(rdm_dir, model_name)\n",
    "        if os.path.exists(rdm_path):\n",
    "            print(f'Skipping model {model_name} as folder already exists: {rdm_path}')\n",
    "            continue\n",
    "\n",
    "        create_rdm(os.path.join(rdm_path, 'pearson'), \n",
    "                   feat_path, 'pearson')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "FCo1hPuX2iBw",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "!rm -rf prednet-rdms/92images/prednet_moments_fmri_it_rdm/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "0ES6QEZAgTN6",
    "colab_type": "code",
    "outputId": "cf42078f-976d-4799-d7e4-51bf90210e03",
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 837.0
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Calculating RDMs for model prednet_finetuned_moments__algonauts\n",
      "Skipping model prednet_finetuned_moments__algonauts as folder already exists: /content/results/rdms/92images/prednet_finetuned_moments__algonauts\n",
      "Calculating RDMs for model prednet_random__algonauts\n",
      "Skipping layer prednet_0_timestep_0\n",
      "Processing layer prednet_0_timestep_1...\n",
      "Processing layer prednet_0_timestep_2...\n",
      "Processing layer prednet_0_timestep_3...\n",
      "Processing layer prednet_0_timestep_4...\n",
      "Processing layer prednet_0_timestep_5...\n",
      "Processing layer prednet_0_timestep_6...\n",
      "Processing layer prednet_0_timestep_7...\n",
      "Processing layer prednet_0_timestep_8...\n",
      "Processing layer prednet_0_timestep_9...\n",
      "Skipping layer prednet_1_timestep_0\n",
      "Processing layer prednet_1_timestep_1...\n",
      "Processing layer prednet_1_timestep_2...\n",
      "Processing layer prednet_1_timestep_3...\n",
      "Processing layer prednet_1_timestep_4...\n",
      "Processing layer prednet_1_timestep_5...\n",
      "Processing layer prednet_1_timestep_6...\n",
      "Processing layer prednet_1_timestep_7...\n",
      "Processing layer prednet_1_timestep_8...\n",
      "Processing layer prednet_1_timestep_9...\n",
      "Skipping layer prednet_2_timestep_0\n",
      "Processing layer prednet_2_timestep_1...\n",
      "Processing layer prednet_2_timestep_2...\n",
      "Processing layer prednet_2_timestep_3...\n",
      "Processing layer prednet_2_timestep_4...\n",
      "Processing layer prednet_2_timestep_5...\n",
      "Processing layer prednet_2_timestep_6...\n",
      "Processing layer prednet_2_timestep_7...\n",
      "Processing layer prednet_2_timestep_8...\n",
      "Processing layer prednet_2_timestep_9...\n",
      "Skipping layer prednet_3_timestep_0\n",
      "Processing layer prednet_3_timestep_1...\n",
      "Processing layer prednet_3_timestep_2...\n",
      "Processing layer prednet_3_timestep_3...\n",
      "Processing layer prednet_3_timestep_4...\n",
      "Processing layer prednet_3_timestep_5...\n",
      "Processing layer prednet_3_timestep_6...\n",
      "Processing layer prednet_3_timestep_7...\n",
      "Processing layer prednet_3_timestep_8...\n",
      "Processing layer prednet_3_timestep_9...\n"
     ]
    }
   ],
   "source": [
    "#feats_pca = get_features_pca('prednet-feats/92images/prednet_moments__model__4_layers_epoch_1/')\n",
    "#sio.savemat('prednet-feats/92images/prednet_moments__model__4_layers_epoch_1__pca/feats_pca.mat', feats_pca)\n",
    "\n",
    "feat_dir = '/content/results/feats/92images/'\n",
    "rdm_dir = '/content/results/rdms/92images/'\n",
    "create_rdms(feat_dir, rdm_dir)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "tTdMPxstyKrd",
    "colab_type": "code",
    "outputId": "361cedd3-7686-4895-df74-31a62fb2b758",
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000.0
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Calculating RDMs for model prednet_finetuned_moments__algonauts\n",
      "Skipping layer prednet_0_timestep_0\n",
      "Processing layer prednet_0_timestep_1...\n",
      "Processing layer prednet_0_timestep_2...\n",
      "Processing layer prednet_0_timestep_3...\n",
      "Processing layer prednet_0_timestep_4...\n",
      "Processing layer prednet_0_timestep_5...\n",
      "Processing layer prednet_0_timestep_6...\n",
      "Processing layer prednet_0_timestep_7...\n",
      "Processing layer prednet_0_timestep_8...\n",
      "Processing layer prednet_0_timestep_9...\n",
      "Skipping layer prednet_1_timestep_0\n",
      "Processing layer prednet_1_timestep_1...\n",
      "Processing layer prednet_1_timestep_2...\n",
      "Processing layer prednet_1_timestep_3...\n",
      "Processing layer prednet_1_timestep_4...\n",
      "Processing layer prednet_1_timestep_5...\n",
      "Processing layer prednet_1_timestep_6...\n",
      "Processing layer prednet_1_timestep_7...\n",
      "Processing layer prednet_1_timestep_8...\n",
      "Processing layer prednet_1_timestep_9...\n",
      "Skipping layer prednet_2_timestep_0\n",
      "Processing layer prednet_2_timestep_1...\n",
      "Processing layer prednet_2_timestep_2...\n",
      "Processing layer prednet_2_timestep_3...\n",
      "Processing layer prednet_2_timestep_4...\n",
      "Processing layer prednet_2_timestep_5...\n",
      "Processing layer prednet_2_timestep_6...\n",
      "Processing layer prednet_2_timestep_7...\n",
      "Processing layer prednet_2_timestep_8...\n",
      "Processing layer prednet_2_timestep_9...\n",
      "Skipping layer prednet_3_timestep_0\n",
      "Processing layer prednet_3_timestep_1...\n",
      "Processing layer prednet_3_timestep_2...\n",
      "Processing layer prednet_3_timestep_3...\n",
      "Processing layer prednet_3_timestep_4...\n",
      "Processing layer prednet_3_timestep_5...\n",
      "Processing layer prednet_3_timestep_6...\n",
      "Processing layer prednet_3_timestep_7...\n",
      "Processing layer prednet_3_timestep_8...\n",
      "Processing layer prednet_3_timestep_9...\n",
      "Calculating RDMs for model prednet_random__algonauts\n",
      "Skipping layer prednet_0_timestep_0\n",
      "Processing layer prednet_0_timestep_1...\n",
      "Processing layer prednet_0_timestep_2...\n",
      "Processing layer prednet_0_timestep_3...\n",
      "Processing layer prednet_0_timestep_4...\n",
      "Processing layer prednet_0_timestep_5...\n",
      "Processing layer prednet_0_timestep_6...\n",
      "Processing layer prednet_0_timestep_7...\n",
      "Processing layer prednet_0_timestep_8...\n",
      "Processing layer prednet_0_timestep_9...\n",
      "Skipping layer prednet_1_timestep_0\n",
      "Processing layer prednet_1_timestep_1...\n",
      "Processing layer prednet_1_timestep_2...\n",
      "Processing layer prednet_1_timestep_3...\n",
      "Processing layer prednet_1_timestep_4...\n",
      "Processing layer prednet_1_timestep_5...\n",
      "Processing layer prednet_1_timestep_6...\n",
      "Processing layer prednet_1_timestep_7...\n",
      "Processing layer prednet_1_timestep_8...\n",
      "Processing layer prednet_1_timestep_9...\n",
      "Skipping layer prednet_2_timestep_0\n",
      "Processing layer prednet_2_timestep_1...\n",
      "Processing layer prednet_2_timestep_2...\n",
      "Processing layer prednet_2_timestep_3...\n",
      "Processing layer prednet_2_timestep_4...\n",
      "Processing layer prednet_2_timestep_5...\n",
      "Processing layer prednet_2_timestep_6...\n",
      "Processing layer prednet_2_timestep_7...\n",
      "Processing layer prednet_2_timestep_8...\n",
      "Processing layer prednet_2_timestep_9...\n",
      "Skipping layer prednet_3_timestep_0\n",
      "Processing layer prednet_3_timestep_1...\n",
      "Processing layer prednet_3_timestep_2...\n",
      "Processing layer prednet_3_timestep_3...\n",
      "Processing layer prednet_3_timestep_4...\n",
      "Processing layer prednet_3_timestep_5...\n",
      "Processing layer prednet_3_timestep_6...\n",
      "Processing layer prednet_3_timestep_7...\n",
      "Processing layer prednet_3_timestep_8...\n",
      "Processing layer prednet_3_timestep_9...\n"
     ]
    }
   ],
   "source": [
    "feat_dir = '/content/results/feats/78images/'\n",
    "rdm_dir = '/content/results/rdms/78images/'\n",
    "create_rdms(feat_dir, rdm_dir)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "fF56bqZuQQC4",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "!cp -r /content/results/rdms /content/gdrive/My\\ Drive/prednet-rdms-$(date +%Y-%m-%dT%H:%M:%S)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "0keVc3Q-1sxy",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "import create_RDMs\n",
    "\n",
    "def merge_feats(save_dir, feat_dir1, feat_dir2, layers1, layers2, dist):\n",
    "    \n",
    "    #get number of layers and number of conditions(images) for RDM\n",
    "    num_layers, layer_list, num_condns = create_RDMs.get_layers_ncondns(feat_dir1)\n",
    "    \n",
    "    if not os.path.exists(save_dir): \n",
    "        os.makedirs(save_dir)\n",
    "    \n",
    "    for img_i in range(num_condns):\n",
    "        features = {}\n",
    "        \n",
    "        for layer1 in layers1:\n",
    "            for layer2 in layers2:\n",
    "                feature1 = create_RDMs.get_features(feat_dir1, layer1, img_i)\n",
    "                feature2 = create_RDMs.get_features(feat_dir2, layer2, img_i)\n",
    "                features[f'{layer1}__{layer2}'] = np.concatenate([feature1, feature2])\n",
    "        \n",
    "        feature_path = os.path.join(save_dir, f'image_{img_i+1:03d}.mat')\n",
    "        # print('Saving feature:', feature_path)\n",
    "        sio.savemat(feature_path, features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "l6lhh5YKacwj",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "feat_dir1 = '/content/prednet-feats/78images/prednet_moments_fmri_it_rdm/'\n",
    "feat_dir2 = '/content/prednet-feats/78images/alexnet/'\n",
    "save_dir = '/content/results/feats/78images/prednet_moments_fmri_it_rdm__alexnet'\n",
    "layers1 = ['prednet_4_timestep_9']\n",
    "layers2 = ['conv2', 'conv4', 'fc8']\n",
    "merge_feats(save_dir, feat_dir1, feat_dir2, layers1, layers2, 'pearson')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "SZJIZdvDvqKE",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "feat_dir1 = '/content/prednet-feats/92images/prednet_kitti__moments__model__10c/'\n",
    "feat_dir2 = '/content/prednet-feats/92images/alexnet/'\n",
    "save_dir = '/content/results/feats/92images/prednet_kitti__moments__model__10c__alexnet'\n",
    "layers1 = ['prednet_3_timestep_9']\n",
    "layers2 = ['conv2', 'conv4', 'fc8']\n",
    "merge_feats(save_dir, feat_dir1, feat_dir2, layers1, layers2, 'pearson')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "id": "D-E0D7mY1iUi",
    "colab_type": "code",
    "colab": {}
   },
   "outputs": [],
   "source": [
    "feat_dir1 = '/content/prednet-feats/78images/prednet_moments__model__4_layers_epoch_1/'\n",
    "feat_dir2 = '/content/prednet-feats/78images/alexnet/'\n",
    "save_dir = '/content/prednet-feats/78images/prednet_moments__model__4_layers_epoch_1__alexnet'\n",
    "layers1 = ['prednet_2_timestep_8', 'prednet_2_timestep_6']\n",
    "layers2 = ['conv4']\n",
    "merge_feats(save_dir, feat_dir1, feat_dir2, layers1, layers2, 'pearson')"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "name": "algonauts-features.ipynb",
   "version": "0.3.2",
   "provenance": [],
   "collapsed_sections": [],
   "toc_visible": true
  },
  "kernelspec": {
   "name": "python3",
   "display_name": "Python 3"
  },
  "accelerator": "GPU"
 },
 "nbformat": 4,
 "nbformat_minor": 0
}