roces_success_rate.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "transparent-pierce",
   "metadata": {},
   "source": [
    "## How often does ROCES outperform the state of the art without requiring all examples to be used?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "numerical-intranet",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os, random\n",
    "from utils.simple_solution import SimpleSolution\n",
    "from utils.evaluator import Evaluator\n",
    "from utils.data import Data\n",
    "from ontolearn.knowledge_base import KnowledgeBase\n",
    "from owlapy.render import DLSyntaxObjectRenderer\n",
    "from roces import BaseConceptSynthesis\n",
    "from roces.synthesizer import ConceptSynthesizer\n",
    "from owlapy.parser import DLSyntaxParser\n",
    "from utils.dataset import DatasetInference\n",
    "from torch.utils.data import DataLoader\n",
    "from torch.nn.utils.rnn import pad_sequence\n",
    "import torch.nn.functional as F\n",
    "from tqdm import tqdm\n",
    "import json\n",
    "import torch\n",
    "import numpy as np, time\n",
    "from collections import defaultdict\n",
    "import re\n",
    "from argparse import Namespace\n",
    "\n",
    "\n",
    "def build_roces_vocabulary(data_train, data_test, kb, args):\n",
    "    def add_data_values(path):\n",
    "        print(\"\\n*** Finding relevant data values ***\")\n",
    "        values = set()\n",
    "        for ce, lp in data_train+data_test:\n",
    "            if '[' in ce:\n",
    "                for val in re.findall(\"\\[(.*?)\\]\", ce):\n",
    "                    values.add(val.split(' ')[-1])\n",
    "        print(\"*** Done! ***\\n\")\n",
    "        print(\"Added values: \", values)\n",
    "        print()\n",
    "        return list(values)\n",
    "    renderer = DLSyntaxObjectRenderer()\n",
    "    individuals = [ind.get_iri().as_str().split(\"/\")[-1] for ind in kb.individuals()]\n",
    "    atomic_concepts = list(kb.ontology().classes_in_signature())\n",
    "    atomic_concept_names = [renderer.render(a) for a in atomic_concepts]\n",
    "    role_names = [rel.get_iri().get_remainder() for rel in kb.ontology().object_properties_in_signature()] + \\\n",
    "                 [rel.get_iri().get_remainder() for rel in kb.ontology().data_properties_in_signature()]\n",
    "    vocab = atomic_concept_names + role_names + ['⊔', '⊓', '∃', '∀', '¬', '⊤', '⊥', '.', ' ', '(', ')',\\\n",
    "                                                '⁻', '≤', '≥', 'True', 'False', '{', '}', ':', '[', ']',\n",
    "                                                'double', 'integer', 'date', 'xsd']\n",
    "    quantified_restriction_values = [str(i) for i in range(1,12)]\n",
    "    data_values = add_data_values(args.knowledge_base_path)\n",
    "    vocab = vocab + data_values + quantified_restriction_values\n",
    "    vocab = sorted(set(vocab)) + ['PAD']\n",
    "    print(\"Vocabulary size: \", len(vocab))\n",
    "    num_examples = min(args.num_examples, kb.individuals_count()//2)\n",
    "    return vocab, num_examples\n",
    "\n",
    "\n",
    "def before_pad(arg):\n",
    "    arg_temp = []\n",
    "    for atm in arg:\n",
    "        if atm == 'PAD':\n",
    "            break\n",
    "        arg_temp.append(atm)\n",
    "    return arg_temp\n",
    "\n",
    "def compute_accuracy(prediction, target):\n",
    "    def soft(arg1, arg2):\n",
    "        arg1_ = arg1\n",
    "        arg2_ = arg2\n",
    "        if isinstance(arg1_, str):\n",
    "            arg1_ = set(before_pad(BaseConceptSynthesis.decompose(arg1_)))\n",
    "        else:\n",
    "            arg1_ = set(before_pad(arg1_))\n",
    "        if isinstance(arg2_, str):\n",
    "            arg2_ = set(before_pad(BaseConceptSynthesis.decompose(arg2_)))\n",
    "        else:\n",
    "            arg2_ = set(before_pad(arg2_))\n",
    "        return 100*float(len(arg1_.intersection(arg2_)))/len(arg1_.union(arg2_))\n",
    "\n",
    "    def hard(arg1, arg2):\n",
    "        arg1_ = arg1\n",
    "        arg2_ = arg2\n",
    "        if isinstance(arg1_, str):\n",
    "            arg1_ = before_pad(BaseConceptSynthesis.decompose(arg1_))\n",
    "        else:\n",
    "            arg1_ = before_pad(arg1_)\n",
    "        if isinstance(arg2_, str):\n",
    "            arg2_ = before_pad(BaseConceptSynthesis.decompose(arg2_))\n",
    "        else:\n",
    "            arg2_ = before_pad(arg2_)\n",
    "        return 100*float(sum(map(lambda x,y: x==y, arg1_, arg2_)))/max(len(arg1_), len(arg2_))\n",
    "    soft_acc = sum(map(soft, prediction, target))/len(target)\n",
    "    hard_acc = sum(map(hard, prediction, target))/len(target)\n",
    "    return soft_acc, hard_acc\n",
    "\n",
    "num_examples = 1000\n",
    "def collate_batch(batch):\n",
    "    pos_emb_list = []\n",
    "    neg_emb_list = []\n",
    "    target_labels = []\n",
    "    for pos_emb, neg_emb, label in batch:\n",
    "        if pos_emb.ndim != 2:\n",
    "            pos_emb = pos_emb.reshape(1, -1)\n",
    "        if neg_emb.ndim != 2:\n",
    "            neg_emb = neg_emb.reshape(1, -1)\n",
    "        pos_emb_list.append(pos_emb)\n",
    "        neg_emb_list.append(neg_emb)\n",
    "        target_labels.append(label)\n",
    "    pos_emb_list[0] = F.pad(pos_emb_list[0], (0, 0, 0, num_examples - pos_emb_list[0].shape[0]), \"constant\", 0)\n",
    "    pos_emb_list = pad_sequence(pos_emb_list, batch_first=True, padding_value=0)\n",
    "    neg_emb_list[0] = F.pad(neg_emb_list[0], (0, 0, 0, num_examples - neg_emb_list[0].shape[0]), \"constant\", 0)\n",
    "    neg_emb_list = pad_sequence(neg_emb_list, batch_first=True, padding_value=0)\n",
    "    target_labels = pad_sequence(target_labels, batch_first=True, padding_value=-100)\n",
    "    return pos_emb_list, neg_emb_list, target_labels\n",
    "\n",
    "def predict(kb, test_data, models, embedding_models, repeat_pred, args):\n",
    "    args.path_to_triples = f\"datasets/{kb}/Triples/\"\n",
    "    global num_examples\n",
    "    num_examples = models[0].num_examples\n",
    "    vocab = models[0].vocab\n",
    "    inv_vocab = models[0].inv_vocab\n",
    "    kb_embedding_data = Data(args)\n",
    "    soft_acc, hard_acc = 0.0, 0.0\n",
    "    preds = []\n",
    "    targets = []\n",
    "    k_values = np.linspace(1+0.1*num_examples, 0.95*num_examples, num=50).astype(int)\n",
    "    Scores = []\n",
    "    for j,k in tqdm(enumerate(k_values), total=len(k_values), desc='sampling examples...'):\n",
    "        test_dataset = DatasetInference(test_data, kb_embedding_data, k, vocab, inv_vocab, args, random_sample=True)\n",
    "        for i, (model, embedding_model) in enumerate(zip(models, embedding_models)):\n",
    "            model = model.eval()\n",
    "            scores = []\n",
    "            test_dataset.load_embeddings(embedding_model.eval())\n",
    "            test_dataloader = DataLoader(test_dataset, batch_size=args.batch_size, num_workers=args.num_workers, collate_fn=collate_batch, shuffle=False)\n",
    "            for x1, x2, labels in tqdm(test_dataloader):\n",
    "                if i == 0 and j == 0:\n",
    "                    target_sequence = model.inv_vocab[labels]\n",
    "                    targets.append(target_sequence)  # The target sequence does not depend on the current model\n",
    "                _, sc = model(x1, x2)\n",
    "                scores.append(sc.detach()) \n",
    "            scores = torch.cat(scores, 0)\n",
    "            if i == 0:\n",
    "                all_scores = scores\n",
    "            else:\n",
    "                all_scores = all_scores + scores\n",
    "        all_scores = all_scores / len(models)\n",
    "        Scores.append(all_scores)\n",
    "    pred_sequences = [model.inv_vocab[score.argmax(1)] for score in Scores]\n",
    "    targets = np.concatenate(targets, 0)\n",
    "    return pred_sequences, targets\n",
    "\n",
    "\n",
    "def initialize_synthesizer(vocab, num_examples, num_inds, args):\n",
    "    args.num_inds = num_inds\n",
    "    roces = ConceptSynthesizer(vocab, num_examples, args)\n",
    "    roces.refresh()\n",
    "    return roces.model, roces.embedding_model\n",
    "\n",
    "def synthesize_class_expressions(kb, test_data, vocab, num_examples, num_inds, repeat_pred, args):\n",
    "    args.knowledge_base_path = \"datasets/\" + f\"{kb}/{kb}.owl\"\n",
    "    embs = torch.load(f\"datasets/{kb}/Model_weights/SetTransformer_{args.kb_emb_model}_Emb_inducing_points32.pt\", map_location=torch.device(\"cpu\"))\n",
    "    setattr(args, 'num_entities', embs['emb_ent_real.weight'].shape[0])\n",
    "    setattr(args, 'num_relations', embs['emb_rel_real.weight'].shape[0])\n",
    "    models, embedding_models = [], []\n",
    "    for inds in num_inds:\n",
    "        model, embedding_model = initialize_synthesizer(vocab, num_examples, inds, args)\n",
    "        if args.sampling_strategy != 'uniform':\n",
    "            model.load_state_dict(torch.load(f\"datasets/{kb}/Model_weights/{args.kb_emb_model}_SetTransformer_inducing_points{inds}.pt\", map_location=torch.device(\"cpu\")))\n",
    "            embedding_model.load_state_dict(torch.load(f\"datasets/{kb}/Model_weights/SetTransformer_{args.kb_emb_model}_Emb_inducing_points{inds}.pt\", map_location=torch.device(\"cpu\")))\n",
    "        else:\n",
    "            model.load_state_dict(torch.load(f\"datasets/{kb}/Model_weights/{args.kb_emb_model}_SetTransformer_uniform_inducing_points{inds}.pt\", map_location=torch.device(\"cpu\")))\n",
    "            embedding_model.load_state_dict(torch.load(f\"datasets/{kb}/Model_weights/SetTransformer_{args.kb_emb_model}_Emb_uniform_inducing_points{inds}.pt\", map_location=torch.device(\"cpu\")))\n",
    "        models.append(model)\n",
    "        embedding_models.append(embedding_model)\n",
    "    return predict(kb, test_data, models, embedding_models, repeat_pred, args)\n",
    "\n",
    "    \n",
    "    \n",
    "def run_experiment(kb_name, args, repeat_pred=False, save_results=False, verbose=False):\n",
    "    print('#'*50)\n",
    "    print('ROCES evaluation on {} KB:'.format(kb_name))\n",
    "    print()\n",
    "    kb = KnowledgeBase(path=f\"datasets/{kb_name}/{kb_name}.owl\")\n",
    "    with open(f\"datasets/{kb_name}/Test_data/Data.json\", \"r\") as file:\n",
    "        test_data = json.load(file)\n",
    "    with open(f\"datasets/{kb_name}/Train_data/Data.json\", \"r\") as file:\n",
    "        train_data = json.load(file)\n",
    "    vocab, num_examples = build_roces_vocabulary(train_data, test_data, kb, args)\n",
    "    namespace = list(kb.individuals())[0].get_iri().get_namespace()\n",
    "    print(\"KB namespace: \", namespace)\n",
    "    print()\n",
    "    simpleSolution = SimpleSolution(kb)\n",
    "    evaluator = Evaluator(kb)\n",
    "    dl_parser = DLSyntaxParser(namespace = namespace)\n",
    "    all_individuals = set(kb.individuals())\n",
    "    ensemble_models = \"+\".join([\"SetTransformer_I32\", \"SetTransformer_I64\", \"SetTransformer_I128\"])\n",
    "    num_inds = [int(model_name.split(\"I\")[-1]) for model_name in ensemble_models.split(\"+\")]\n",
    "    predictions, targets = synthesize_class_expressions(kb_name, test_data, vocab, num_examples, num_inds, repeat_pred, args)\n",
    "    F1 = [list() for _ in range(len(predictions))]\n",
    "    Acc = [list() for _ in range(len(predictions))]\n",
    "    Predictions = [list() for _ in range(len(predictions))]\n",
    "    for i, k_pred in tqdm(enumerate(predictions), total=len(predictions)):\n",
    "        for j, pb_str in enumerate(targets):\n",
    "            pb_str = \"\".join(before_pad(pb_str))\n",
    "            try:\n",
    "                end_idx = np.where(k_pred[j] == 'PAD')[0][0] # remove padding token\n",
    "            except IndexError:\n",
    "                end_idx = -1\n",
    "            pred = k_pred[j][:end_idx]\n",
    "            try:\n",
    "                prediction = dl_parser.parse(\"\".join(pred.tolist()))\n",
    "            except Exception:\n",
    "                try:\n",
    "                    pred = simpleSolution.predict(k_pred[j].sum())\n",
    "                    prediction = dl_parser.parse(pred)\n",
    "                except Exception:\n",
    "                    print(f\"Could not understand expression {pred}\")\n",
    "            if prediction is None:\n",
    "                prediction = dl_parser.parse('⊤')\n",
    "            target_expression = dl_parser.parse(pb_str) # The target class expression\n",
    "            positive_examples = set(kb.individuals(target_expression))\n",
    "            negative_examples = all_individuals-positive_examples\n",
    "            try:\n",
    "                acc, f1 = evaluator.evaluate(prediction, positive_examples, negative_examples)\n",
    "            except Exception as err:\n",
    "                print(f\"{j}: Parsing error on \", prediction)\n",
    "                print(err)\n",
    "                prediction = dl_parser.parse('⊤')\n",
    "                acc, f1 = evaluator.evaluate(prediction, positive_examples, negative_examples)\n",
    "            if verbose:\n",
    "                print(f'Problem {j}, Target: {pb_str}, Prediction: {simpleSolution.renderer.render(prediction)}, Acc: {acc}, F1: {f1}')\n",
    "                print()\n",
    "            try:\n",
    "                prediction_str = simpleSolution.renderer.render(prediction)\n",
    "            except:\n",
    "                prediction_str = \"Unknown\"\n",
    "            Acc[i].append(acc)\n",
    "            F1[i].append(f1)\n",
    "            Predictions[i].append(prediction_str)\n",
    "    with open(f\"datasets/{kb_name}/RQ3_F1.json\", \"w\") as file:\n",
    "        json.dump({key: val for key,val in enumerate(F1)}, file, ensure_ascii=False)\n",
    "    with open(f\"datasets/{kb_name}/RQ3_Acc.json\", \"w\") as file:\n",
    "        json.dump({key: val for key,val in enumerate(Acc)}, file, ensure_ascii=False)\n",
    "    with open(f\"datasets/{kb_name}/RQ3_Pred.json\", \"w\") as file:\n",
    "        json.dump({key: val for key,val in enumerate(Predictions)}, file, ensure_ascii=False)\n",
    "    return Acc, F1, Predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "labeled-resolution",
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"config.json\") as config:\n",
    "    nces_args = json.load(config)\n",
    "    nces_args = Namespace(**nces_args)\n",
    "nces_args.kb_emb_model = \"ConEx\"\n",
    "nces_args.sampling_strategy = \"original\" # This corresponds to the probability density functions p+ and p- defined in Algorithm 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "extreme-economics",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "##################################################\n",
      "ROCES evaluation on semantic_bible KB:\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "* Owlready2 * Warning: optimized Cython parser module 'owlready2_optimized' is not available, defaulting to slower Python implementation\n",
      "\n",
      "Warning: SQLite3 version 3.40.0 and 3.41.2 have huge performance regressions; please install version 3.41.1 or 3.42!\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "*** Finding relevant data values ***\n",
      "*** Done! ***\n",
      "\n",
      "Added values:  {'35.583', '44.42', '41.013'}\n",
      "\n",
      "Vocabulary size:  126\n",
      "KB namespace:  http://semanticbible.org/ns/2006/NTNames#\n",
      "\n",
      "\n",
      "\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "sampling examples...:   0%|          | 0/50 [00:00<?, ?it/s]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:00<00:00,  2.43it/s]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:00<00:00,  1.49it/s]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.20s/it]\u001b[A\n",
      "sampling examples...:   2%|▏         | 1/50 [00:02<01:53,  2.31s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.22s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.47s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.68s/it]\u001b[A\n",
      "sampling examples...:   4%|▍         | 2/50 [00:06<02:49,  3.53s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.63s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.83s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.96s/it]\u001b[A\n",
      "sampling examples...:   6%|▌         | 3/50 [00:12<03:27,  4.42s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.82s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.83s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.85s/it]\u001b[A\n",
      "sampling examples...:   8%|▊         | 4/50 [00:17<03:44,  4.88s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.07s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.66s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.97s/it]\u001b[A\n",
      "sampling examples...:  10%|█         | 5/50 [00:23<03:53,  5.19s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.68s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.59s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.79s/it]\u001b[A\n",
      "sampling examples...:  12%|█▏        | 6/50 [00:28<03:47,  5.16s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.66s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.43s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.86s/it]\u001b[A\n",
      "sampling examples...:  14%|█▍        | 7/50 [00:33<03:39,  5.11s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.57s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.80s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.71s/it]\u001b[A\n",
      "sampling examples...:  16%|█▌        | 8/50 [00:38<03:34,  5.11s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.66s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.58s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.82s/it]\u001b[A\n",
      "sampling examples...:  18%|█▊        | 9/50 [00:43<03:29,  5.11s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.77s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.59s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.80s/it]\u001b[A\n",
      "sampling examples...:  20%|██        | 10/50 [00:49<03:25,  5.13s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.69s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.62s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.69s/it]\u001b[A\n",
      "sampling examples...:  22%|██▏       | 11/50 [00:54<03:18,  5.10s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.73s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.70s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.71s/it]\u001b[A\n",
      "sampling examples...:  24%|██▍       | 12/50 [00:59<03:14,  5.13s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.69s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.60s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.74s/it]\u001b[A\n",
      "sampling examples...:  26%|██▌       | 13/50 [01:04<03:09,  5.11s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.54s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.67s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.76s/it]\u001b[A\n",
      "sampling examples...:  28%|██▊       | 14/50 [01:09<03:02,  5.08s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.65s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.51s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.82s/it]\u001b[A\n",
      "sampling examples...:  30%|███       | 15/50 [01:14<02:57,  5.06s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.59s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.53s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.70s/it]\u001b[A\n",
      "sampling examples...:  32%|███▏      | 16/50 [01:19<02:49,  4.99s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.53s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.52s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.77s/it]\u001b[A\n",
      "sampling examples...:  34%|███▍      | 17/50 [01:24<02:43,  4.96s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.57s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.69s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.82s/it]\u001b[A\n",
      "sampling examples...:  36%|███▌      | 18/50 [01:29<02:40,  5.01s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.75s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.61s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.80s/it]\u001b[A\n",
      "sampling examples...:  38%|███▊      | 19/50 [01:34<02:37,  5.08s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.70s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.62s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.83s/it]\u001b[A\n",
      "sampling examples...:  40%|████      | 20/50 [01:39<02:33,  5.11s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.75s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.59s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.73s/it]\u001b[A\n",
      "sampling examples...:  42%|████▏     | 21/50 [01:44<02:28,  5.11s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.61s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.57s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.73s/it]\u001b[A\n",
      "sampling examples...:  44%|████▍     | 22/50 [01:49<02:21,  5.07s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.69s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.48s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.82s/it]\u001b[A\n",
      "sampling examples...:  46%|████▌     | 23/50 [01:54<02:16,  5.07s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.62s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.63s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.83s/it]\u001b[A\n",
      "sampling examples...:  48%|████▊     | 24/50 [01:59<02:11,  5.08s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.59s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.58s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.82s/it]\u001b[A\n",
      "sampling examples...:  50%|█████     | 25/50 [02:04<02:06,  5.08s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.64s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.65s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.83s/it]\u001b[A\n",
      "sampling examples...:  52%|█████▏    | 26/50 [02:10<02:02,  5.11s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.62s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.67s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.78s/it]\u001b[A\n",
      "sampling examples...:  54%|█████▍    | 27/50 [02:15<01:57,  5.11s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.57s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.64s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.84s/it]\u001b[A\n",
      "sampling examples...:  56%|█████▌    | 28/50 [02:20<01:52,  5.10s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.77s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.68s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.83s/it]\u001b[A\n",
      "sampling examples...:  58%|█████▊    | 29/50 [02:25<01:48,  5.17s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.65s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.62s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.84s/it]\u001b[A\n",
      "sampling examples...:  60%|██████    | 30/50 [02:30<01:43,  5.16s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.58s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.58s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.88s/it]\u001b[A\n",
      "sampling examples...:  62%|██████▏   | 31/50 [02:35<01:37,  5.14s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.59s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.63s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.88s/it]\u001b[A\n",
      "sampling examples...:  64%|██████▍   | 32/50 [02:40<01:32,  5.14s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.65s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.67s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.81s/it]\u001b[A\n",
      "sampling examples...:  66%|██████▌   | 33/50 [02:46<01:27,  5.14s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.65s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.59s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.77s/it]\u001b[A\n",
      "sampling examples...:  68%|██████▊   | 34/50 [02:51<01:22,  5.13s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.65s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.62s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.79s/it]\u001b[A\n",
      "sampling examples...:  70%|███████   | 35/50 [02:56<01:16,  5.12s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.55s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.60s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.89s/it]\u001b[A\n",
      "sampling examples...:  72%|███████▏  | 36/50 [03:01<01:11,  5.11s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.53s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.66s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.83s/it]\u001b[A\n",
      "sampling examples...:  74%|███████▍  | 37/50 [03:06<01:06,  5.10s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.64s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.67s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.84s/it]\u001b[A\n",
      "sampling examples...:  76%|███████▌  | 38/50 [03:11<01:01,  5.12s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.57s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.63s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.77s/it]\u001b[A\n",
      "sampling examples...:  78%|███████▊  | 39/50 [03:16<00:55,  5.09s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.61s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.54s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.76s/it]\u001b[A\n",
      "sampling examples...:  80%|████████  | 40/50 [03:21<00:50,  5.05s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.58s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.66s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.78s/it]\u001b[A\n",
      "sampling examples...:  82%|████████▏ | 41/50 [03:26<00:45,  5.05s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.71s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.79s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.85s/it]\u001b[A\n",
      "sampling examples...:  84%|████████▍ | 42/50 [03:32<00:41,  5.17s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.78s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.65s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.91s/it]\u001b[A\n",
      "sampling examples...:  86%|████████▌ | 43/50 [03:37<00:36,  5.23s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.78s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.67s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.82s/it]\u001b[A\n",
      "sampling examples...:  88%|████████▊ | 44/50 [03:42<00:31,  5.25s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.66s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.74s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.92s/it]\u001b[A\n",
      "sampling examples...:  90%|█████████ | 45/50 [03:48<00:26,  5.28s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.68s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.63s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.79s/it]\u001b[A\n",
      "sampling examples...:  92%|█████████▏| 46/50 [03:53<00:20,  5.23s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.67s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.62s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.98s/it]\u001b[A\n",
      "sampling examples...:  94%|█████████▍| 47/50 [03:58<00:15,  5.25s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.82s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.74s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.78s/it]\u001b[A\n",
      "sampling examples...:  96%|█████████▌| 48/50 [04:03<00:10,  5.29s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.72s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.64s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.88s/it]\u001b[A\n",
      "sampling examples...:  98%|█████████▊| 49/50 [04:09<00:05,  5.29s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.69s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.59s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.89s/it]\u001b[A\n",
      "sampling examples...: 100%|██████████| 50/50 [04:14<00:00,  5.09s/it]\n",
      "  0%|          | 0/50 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  4%|▍         | 2/50 [00:00<00:13,  3.61it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n",
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  8%|▊         | 4/50 [00:00<00:09,  4.73it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n",
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 50/50 [00:10<00:00,  4.84it/s]\n"
     ]
    }
   ],
   "source": [
    "Acc, F1, Predictions = run_experiment(\"semantic_bible\", nces_args)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "forty-behalf",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "seventh-microphone",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "##################################################\n",
      "ROCES evaluation on carcinogenesis KB:\n",
      "\n",
      "\n",
      "*** Finding relevant data values ***\n",
      "*** Done! ***\n",
      "\n",
      "Added values:  set()\n",
      "\n",
      "Vocabulary size:  198\n",
      "KB namespace:  http://dl-learner.org/carcinogenesis#\n",
      "\n",
      "\n",
      "\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "sampling examples...:   0%|          | 0/50 [00:00<?, ?it/s]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.77s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.07s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.66s/it]\u001b[A\n",
      "sampling examples...:   2%|▏         | 1/50 [00:07<06:08,  7.53s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.81s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.45s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.27s/it]\u001b[A\n",
      "sampling examples...:   4%|▍         | 2/50 [00:14<05:34,  6.96s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.44s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.11s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.65s/it]\u001b[A\n",
      "sampling examples...:   6%|▌         | 3/50 [00:21<05:33,  7.09s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.15s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.39s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.38s/it]\u001b[A\n",
      "sampling examples...:   8%|▊         | 4/50 [00:28<05:23,  7.04s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.41s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.17s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.70s/it]\u001b[A\n",
      "sampling examples...:  10%|█         | 5/50 [00:35<05:21,  7.15s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.15s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.25s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.44s/it]\u001b[A\n",
      "sampling examples...:  12%|█▏        | 6/50 [00:42<05:10,  7.05s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.31s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.12s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.56s/it]\u001b[A\n",
      "sampling examples...:  14%|█▍        | 7/50 [00:49<05:03,  7.05s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.11s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.21s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.46s/it]\u001b[A\n",
      "sampling examples...:  16%|█▌        | 8/50 [00:56<04:53,  6.98s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.31s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.19s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.52s/it]\u001b[A\n",
      "sampling examples...:  18%|█▊        | 9/50 [01:03<04:47,  7.01s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.16s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.17s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.56s/it]\u001b[A\n",
      "sampling examples...:  20%|██        | 10/50 [01:10<04:39,  6.99s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.19s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.17s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.54s/it]\u001b[A\n",
      "sampling examples...:  22%|██▏       | 11/50 [01:17<04:32,  6.98s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.30s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.12s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.41s/it]\u001b[A\n",
      "sampling examples...:  24%|██▍       | 12/50 [01:24<04:23,  6.94s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.26s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.23s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.60s/it]\u001b[A\n",
      "sampling examples...:  26%|██▌       | 13/50 [01:31<04:18,  6.99s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.20s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.27s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.52s/it]\u001b[A\n",
      "sampling examples...:  28%|██▊       | 14/50 [01:38<04:12,  7.00s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.22s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.15s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.51s/it]\u001b[A\n",
      "sampling examples...:  30%|███       | 15/50 [01:45<04:04,  6.98s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.11s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.15s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.60s/it]\u001b[A\n",
      "sampling examples...:  32%|███▏      | 16/50 [01:52<03:56,  6.96s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.29s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.20s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.68s/it]\u001b[A\n",
      "sampling examples...:  34%|███▍      | 17/50 [01:59<03:52,  7.03s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.31s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.20s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.57s/it]\u001b[A\n",
      "sampling examples...:  36%|███▌      | 18/50 [02:06<03:45,  7.06s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.15s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.37s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.55s/it]\u001b[A\n",
      "sampling examples...:  38%|███▊      | 19/50 [02:13<03:39,  7.08s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.24s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.15s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.83s/it]\u001b[A\n",
      "sampling examples...:  40%|████      | 20/50 [02:20<03:34,  7.14s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.18s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.39s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.62s/it]\u001b[A\n",
      "sampling examples...:  42%|████▏     | 21/50 [02:28<03:27,  7.17s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.38s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.14s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.71s/it]\u001b[A\n",
      "sampling examples...:  44%|████▍     | 22/50 [02:35<03:21,  7.19s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.05s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.37s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.50s/it]\u001b[A\n",
      "sampling examples...:  46%|████▌     | 23/50 [02:42<03:12,  7.13s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.38s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.16s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.78s/it]\u001b[A\n",
      "sampling examples...:  48%|████▊     | 24/50 [02:49<03:07,  7.20s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.11s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.36s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.49s/it]\u001b[A\n",
      "sampling examples...:  50%|█████     | 25/50 [02:56<02:58,  7.13s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.26s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.12s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.72s/it]\u001b[A\n",
      "sampling examples...:  52%|█████▏    | 26/50 [03:03<02:51,  7.13s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.23s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.47s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.37s/it]\u001b[A\n",
      "sampling examples...:  54%|█████▍    | 27/50 [03:10<02:43,  7.13s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.53s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.14s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.74s/it]\u001b[A\n",
      "sampling examples...:  56%|█████▌    | 28/50 [03:18<02:38,  7.23s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.12s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.33s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.48s/it]\u001b[A\n",
      "sampling examples...:  58%|█████▊    | 29/50 [03:25<02:30,  7.15s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.38s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.23s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.66s/it]\u001b[A\n",
      "sampling examples...:  60%|██████    | 30/50 [03:32<02:24,  7.20s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.29s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.33s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.56s/it]\u001b[A\n",
      "sampling examples...:  62%|██████▏   | 31/50 [03:39<02:17,  7.21s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.33s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.24s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.71s/it]\u001b[A\n",
      "sampling examples...:  64%|██████▍   | 32/50 [03:47<02:10,  7.25s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.24s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.32s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.47s/it]\u001b[A\n",
      "sampling examples...:  66%|██████▌   | 33/50 [03:54<02:02,  7.20s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.28s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.26s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.70s/it]\u001b[A\n",
      "sampling examples...:  68%|██████▊   | 34/50 [04:01<01:55,  7.23s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.24s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.31s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.54s/it]\u001b[A\n",
      "sampling examples...:  70%|███████   | 35/50 [04:08<01:47,  7.19s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.39s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.18s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.78s/it]\u001b[A\n",
      "sampling examples...:  72%|███████▏  | 36/50 [04:16<01:41,  7.25s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.07s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.39s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.47s/it]\u001b[A\n",
      "sampling examples...:  74%|███████▍  | 37/50 [04:23<01:33,  7.16s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.33s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.25s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.75s/it]\u001b[A\n",
      "sampling examples...:  76%|███████▌  | 38/50 [04:30<01:26,  7.23s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.25s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.48s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.50s/it]\u001b[A\n",
      "sampling examples...:  78%|███████▊  | 39/50 [04:37<01:19,  7.24s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.37s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.15s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.61s/it]\u001b[A\n",
      "sampling examples...:  80%|████████  | 40/50 [04:44<01:12,  7.22s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.19s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.27s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.57s/it]\u001b[A\n",
      "sampling examples...:  82%|████████▏ | 41/50 [04:51<01:04,  7.17s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.36s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.29s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.59s/it]\u001b[A\n",
      "sampling examples...:  84%|████████▍ | 42/50 [04:59<00:57,  7.22s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.23s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.47s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.74s/it]\u001b[A\n",
      "sampling examples...:  86%|████████▌ | 43/50 [05:06<00:51,  7.29s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.24s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.22s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.47s/it]\u001b[A\n",
      "sampling examples...:  88%|████████▊ | 44/50 [05:13<00:43,  7.20s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.28s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.23s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.68s/it]\u001b[A\n",
      "sampling examples...:  90%|█████████ | 45/50 [05:20<00:36,  7.20s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.33s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.35s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.61s/it]\u001b[A\n",
      "sampling examples...:  92%|█████████▏| 46/50 [05:28<00:28,  7.24s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.18s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.37s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.54s/it]\u001b[A\n",
      "sampling examples...:  94%|█████████▍| 47/50 [05:35<00:21,  7.20s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.32s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.22s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.62s/it]\u001b[A\n",
      "sampling examples...:  96%|█████████▌| 48/50 [05:42<00:14,  7.21s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.16s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.32s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.56s/it]\u001b[A\n",
      "sampling examples...:  98%|█████████▊| 49/50 [05:49<00:07,  7.18s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.34s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.26s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.56s/it]\u001b[A\n",
      "sampling examples...: 100%|██████████| 50/50 [05:56<00:00,  7.14s/it]\n",
      "100%|██████████| 50/50 [00:28<00:00,  1.76it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "##################################################\n",
      "ROCES evaluation on mutagenesis KB:\n",
      "\n",
      "\n",
      "*** Finding relevant data values ***\n",
      "*** Done! ***\n",
      "\n",
      "Added values:  {'1.002', '-0.529', '-1.031', '7.84', '-0.781', '5.39', '-3.768'}\n",
      "\n",
      "Vocabulary size:  141\n",
      "KB namespace:  http://dl-learner.org/mutagenesis#\n",
      "\n",
      "\n",
      "\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "sampling examples...:   0%|          | 0/50 [00:00<?, ?it/s]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.36s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.76s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.74s/it]\u001b[A\n",
      "sampling examples...:   2%|▏         | 1/50 [00:06<05:37,  6.89s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.79s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.32s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.03s/it]\u001b[A\n",
      "sampling examples...:   4%|▍         | 2/50 [00:13<05:10,  6.47s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.15s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.79s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.55s/it]\u001b[A\n",
      "sampling examples...:   6%|▌         | 3/50 [00:19<05:05,  6.49s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.78s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.17s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.19s/it]\u001b[A\n",
      "sampling examples...:   8%|▊         | 4/50 [00:25<04:52,  6.37s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.20s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.87s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.53s/it]\u001b[A\n",
      "sampling examples...:  10%|█         | 5/50 [00:32<04:50,  6.46s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.78s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.14s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.22s/it]\u001b[A\n",
      "sampling examples...:  12%|█▏        | 6/50 [00:38<04:40,  6.37s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.12s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.97s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.47s/it]\u001b[A\n",
      "sampling examples...:  14%|█▍        | 7/50 [00:45<04:37,  6.44s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.93s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.09s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.32s/it]\u001b[A\n",
      "sampling examples...:  16%|█▌        | 8/50 [00:51<04:30,  6.43s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.09s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.06s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.33s/it]\u001b[A\n",
      "sampling examples...:  18%|█▊        | 9/50 [00:58<04:24,  6.45s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.90s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.01s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.23s/it]\u001b[A\n",
      "sampling examples...:  20%|██        | 10/50 [01:04<04:14,  6.37s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.03s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.12s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.40s/it]\u001b[A\n",
      "sampling examples...:  22%|██▏       | 11/50 [01:10<04:11,  6.44s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.95s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.23s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.34s/it]\u001b[A\n",
      "sampling examples...:  24%|██▍       | 12/50 [01:17<04:06,  6.49s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.00s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.05s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.41s/it]\u001b[A\n",
      "sampling examples...:  26%|██▌       | 13/50 [01:23<04:00,  6.49s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.01s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.06s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.37s/it]\u001b[A\n",
      "sampling examples...:  28%|██▊       | 14/50 [01:30<03:53,  6.48s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.98s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.18s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.42s/it]\u001b[A\n",
      "sampling examples...:  30%|███       | 15/50 [01:37<03:48,  6.53s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.92s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.02s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.33s/it]\u001b[A\n",
      "sampling examples...:  32%|███▏      | 16/50 [01:43<03:39,  6.47s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.01s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.05s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.26s/it]\u001b[A\n",
      "sampling examples...:  34%|███▍      | 17/50 [01:49<03:32,  6.44s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.06s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.23s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.35s/it]\u001b[A\n",
      "sampling examples...:  36%|███▌      | 18/50 [01:56<03:28,  6.52s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.11s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.19s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.42s/it]\u001b[A\n",
      "sampling examples...:  38%|███▊      | 19/50 [02:03<03:24,  6.58s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.98s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.08s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.25s/it]\u001b[A\n",
      "sampling examples...:  40%|████      | 20/50 [02:09<03:15,  6.51s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.06s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.01s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.50s/it]\u001b[A\n",
      "sampling examples...:  42%|████▏     | 21/50 [02:16<03:09,  6.55s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.99s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.32s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.36s/it]\u001b[A\n",
      "sampling examples...:  44%|████▍     | 22/50 [02:22<03:04,  6.59s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.22s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.09s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.49s/it]\u001b[A\n",
      "sampling examples...:  46%|████▌     | 23/50 [02:29<03:00,  6.67s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.98s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.24s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.26s/it]\u001b[A\n",
      "sampling examples...:  48%|████▊     | 24/50 [02:36<02:52,  6.62s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.21s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.84s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.52s/it]\u001b[A\n",
      "sampling examples...:  50%|█████     | 25/50 [02:42<02:45,  6.61s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.84s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.20s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.22s/it]\u001b[A\n",
      "sampling examples...:  52%|█████▏    | 26/50 [02:49<02:36,  6.52s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.06s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.93s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.51s/it]\u001b[A\n",
      "sampling examples...:  54%|█████▍    | 27/50 [02:55<02:30,  6.53s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.92s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.17s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.30s/it]\u001b[A\n",
      "sampling examples...:  56%|█████▌    | 28/50 [03:02<02:23,  6.51s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.14s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.89s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.53s/it]\u001b[A\n",
      "sampling examples...:  58%|█████▊    | 29/50 [03:08<02:17,  6.54s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.93s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.19s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.23s/it]\u001b[A\n",
      "sampling examples...:  60%|██████    | 30/50 [03:15<02:09,  6.49s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.11s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.86s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.55s/it]\u001b[A\n",
      "sampling examples...:  62%|██████▏   | 31/50 [03:21<02:03,  6.51s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.95s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.21s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.31s/it]\u001b[A\n",
      "sampling examples...:  64%|██████▍   | 32/50 [03:28<01:57,  6.51s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.07s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.94s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.44s/it]\u001b[A\n",
      "sampling examples...:  66%|██████▌   | 33/50 [03:34<01:50,  6.51s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.04s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.15s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.38s/it]\u001b[A\n",
      "sampling examples...:  68%|██████▊   | 34/50 [03:41<01:44,  6.53s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.98s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.17s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.30s/it]\u001b[A\n",
      "sampling examples...:  70%|███████   | 35/50 [03:47<01:37,  6.52s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.15s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.11s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.35s/it]\u001b[A\n",
      "sampling examples...:  72%|███████▏  | 36/50 [03:54<01:31,  6.55s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.05s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.07s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.44s/it]\u001b[A\n",
      "sampling examples...:  74%|███████▍  | 37/50 [04:01<01:25,  6.57s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.02s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.11s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.27s/it]\u001b[A\n",
      "sampling examples...:  76%|███████▌  | 38/50 [04:07<01:18,  6.53s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.96s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.09s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.37s/it]\u001b[A\n",
      "sampling examples...:  78%|███████▊  | 39/50 [04:13<01:11,  6.52s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.01s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.98s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.51s/it]\u001b[A\n",
      "sampling examples...:  80%|████████  | 40/50 [04:20<01:05,  6.52s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.94s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.04s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.37s/it]\u001b[A\n",
      "sampling examples...:  82%|████████▏ | 41/50 [04:26<00:58,  6.47s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.07s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.13s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.49s/it]\u001b[A\n",
      "sampling examples...:  84%|████████▍ | 42/50 [04:33<00:52,  6.55s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.98s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.03s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.32s/it]\u001b[A\n",
      "sampling examples...:  86%|████████▌ | 43/50 [04:39<00:45,  6.49s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.02s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.01s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.44s/it]\u001b[A\n",
      "sampling examples...:  88%|████████▊ | 44/50 [04:46<00:38,  6.49s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.01s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.95s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.36s/it]\u001b[A\n",
      "sampling examples...:  90%|█████████ | 45/50 [04:52<00:32,  6.46s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.04s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.03s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.39s/it]\u001b[A\n",
      "sampling examples...:  92%|█████████▏| 46/50 [04:59<00:25,  6.48s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.01s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.13s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.43s/it]\u001b[A\n",
      "sampling examples...:  94%|█████████▍| 47/50 [05:05<00:19,  6.51s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.03s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.13s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.26s/it]\u001b[A\n",
      "sampling examples...:  96%|█████████▌| 48/50 [05:12<00:12,  6.49s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.13s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.03s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.47s/it]\u001b[A\n",
      "sampling examples...:  98%|█████████▊| 49/50 [05:19<00:06,  6.55s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:01<00:00,  1.92s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.15s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.33s/it]\u001b[A\n",
      "sampling examples...: 100%|██████████| 50/50 [05:25<00:00,  6.51s/it]\n",
      " 12%|█▏        | 6/50 [00:05<00:33,  1.30it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 14%|█▍        | 7/50 [00:05<00:30,  1.40it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 16%|█▌        | 8/50 [00:06<00:28,  1.46it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 18%|█▊        | 9/50 [00:06<00:27,  1.51it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 22%|██▏       | 11/50 [00:08<00:24,  1.57it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 30%|███       | 15/50 [00:10<00:21,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 40%|████      | 20/50 [00:13<00:18,  1.63it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 42%|████▏     | 21/50 [00:14<00:17,  1.63it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 44%|████▍     | 22/50 [00:14<00:17,  1.63it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 46%|████▌     | 23/50 [00:15<00:16,  1.63it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 48%|████▊     | 24/50 [00:16<00:16,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 50%|█████     | 25/50 [00:16<00:15,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 52%|█████▏    | 26/50 [00:17<00:14,  1.61it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 54%|█████▍    | 27/50 [00:17<00:14,  1.61it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 56%|█████▌    | 28/50 [00:18<00:13,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 58%|█████▊    | 29/50 [00:19<00:12,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 60%|██████    | 30/50 [00:19<00:12,  1.61it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 62%|██████▏   | 31/50 [00:20<00:11,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 64%|██████▍   | 32/50 [00:21<00:11,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 66%|██████▌   | 33/50 [00:21<00:10,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 68%|██████▊   | 34/50 [00:22<00:09,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 70%|███████   | 35/50 [00:22<00:09,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 72%|███████▏  | 36/50 [00:23<00:08,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 74%|███████▍  | 37/50 [00:24<00:08,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 76%|███████▌  | 38/50 [00:24<00:07,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 78%|███████▊  | 39/50 [00:25<00:06,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 80%|████████  | 40/50 [00:26<00:06,  1.61it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 82%|████████▏ | 41/50 [00:26<00:05,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 84%|████████▍ | 42/50 [00:27<00:04,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 86%|████████▌ | 43/50 [00:27<00:04,  1.61it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 88%|████████▊ | 44/50 [00:28<00:03,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 90%|█████████ | 45/50 [00:29<00:03,  1.61it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 92%|█████████▏| 46/50 [00:29<00:02,  1.61it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 94%|█████████▍| 47/50 [00:30<00:01,  1.61it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 96%|█████████▌| 48/50 [00:30<00:01,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 98%|█████████▊| 49/50 [00:31<00:00,  1.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 50/50 [00:32<00:00,  1.55it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "##################################################\n",
      "ROCES evaluation on vicodi KB:\n",
      "\n",
      "\n",
      "*** Finding relevant data values ***\n",
      "*** Done! ***\n",
      "\n",
      "Added values:  set()\n",
      "\n",
      "Vocabulary size:  243\n",
      "KB namespace:  http://vicodi.org/ontology#\n",
      "\n",
      "\n",
      "\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "sampling examples...:   0%|          | 0/50 [00:00<?, ?it/s]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.88s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.61s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.07s/it]\u001b[A\n",
      "sampling examples...:   2%|▏         | 1/50 [00:08<07:01,  8.60s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.42s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.06s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.77s/it]\u001b[A\n",
      "sampling examples...:   4%|▍         | 2/50 [00:16<06:43,  8.42s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.96s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.64s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.09s/it]\u001b[A\n",
      "sampling examples...:   6%|▌         | 3/50 [00:25<06:41,  8.55s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.55s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.99s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.95s/it]\u001b[A\n",
      "sampling examples...:   8%|▊         | 4/50 [00:34<06:32,  8.54s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.79s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.60s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.30s/it]\u001b[A\n",
      "sampling examples...:  10%|█         | 5/50 [00:42<06:27,  8.61s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.83s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.89s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.06s/it]\u001b[A\n",
      "sampling examples...:  12%|█▏        | 6/50 [00:51<06:22,  8.70s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.76s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.72s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.18s/it]\u001b[A\n",
      "sampling examples...:  14%|█▍        | 7/50 [01:00<06:13,  8.69s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.75s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.78s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.10s/it]\u001b[A\n",
      "sampling examples...:  16%|█▌        | 8/50 [01:09<06:04,  8.68s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.81s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.87s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.15s/it]\u001b[A\n",
      "sampling examples...:  18%|█▊        | 9/50 [01:17<05:58,  8.74s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.77s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.91s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.11s/it]\u001b[A\n",
      "sampling examples...:  20%|██        | 10/50 [01:26<05:50,  8.76s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.84s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.70s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.18s/it]\u001b[A\n",
      "sampling examples...:  22%|██▏       | 11/50 [01:35<05:41,  8.76s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.64s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.79s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.23s/it]\u001b[A\n",
      "sampling examples...:  24%|██▍       | 12/50 [01:44<05:32,  8.74s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.75s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.87s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.01s/it]\u001b[A\n",
      "sampling examples...:  26%|██▌       | 13/50 [01:52<05:22,  8.73s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.78s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.71s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.12s/it]\u001b[A\n",
      "sampling examples...:  28%|██▊       | 14/50 [02:01<05:13,  8.71s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.72s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.91s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.10s/it]\u001b[A\n",
      "sampling examples...:  30%|███       | 15/50 [02:10<05:05,  8.73s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.82s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.82s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.24s/it]\u001b[A\n",
      "sampling examples...:  32%|███▏      | 16/50 [02:19<04:58,  8.79s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.98s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.87s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.21s/it]\u001b[A\n",
      "sampling examples...:  34%|███▍      | 17/50 [02:28<04:53,  8.88s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.77s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.77s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.01s/it]\u001b[A\n",
      "sampling examples...:  36%|███▌      | 18/50 [02:36<04:41,  8.80s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.75s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.76s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.26s/it]\u001b[A\n",
      "sampling examples...:  38%|███▊      | 19/50 [02:45<04:33,  8.81s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.76s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.80s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.11s/it]\u001b[A\n",
      "sampling examples...:  40%|████      | 20/50 [02:54<04:23,  8.78s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.90s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.84s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.26s/it]\u001b[A\n",
      "sampling examples...:  42%|████▏     | 21/50 [03:03<04:17,  8.86s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.66s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.81s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.04s/it]\u001b[A\n",
      "sampling examples...:  44%|████▍     | 22/50 [03:12<04:05,  8.77s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.86s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.79s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.16s/it]\u001b[A\n",
      "sampling examples...:  46%|████▌     | 23/50 [03:20<03:57,  8.79s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.68s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.86s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.01s/it]\u001b[A\n",
      "sampling examples...:  48%|████▊     | 24/50 [03:29<03:47,  8.73s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.90s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.78s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.20s/it]\u001b[A\n",
      "sampling examples...:  50%|█████     | 25/50 [03:38<03:39,  8.78s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.80s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.06s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.95s/it]\u001b[A\n",
      "sampling examples...:  52%|█████▏    | 26/50 [03:47<03:31,  8.80s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.85s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.82s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.16s/it]\u001b[A\n",
      "sampling examples...:  54%|█████▍    | 27/50 [03:56<03:22,  8.82s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.76s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.84s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:06<00:00,  6.53s/it]\u001b[A\n",
      "sampling examples...:  56%|█████▌    | 28/50 [04:08<03:36,  9.83s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:04<00:00,  4.78s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:04<00:00,  4.95s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:05<00:00,  5.35s/it]\u001b[A\n",
      "sampling examples...:  58%|█████▊    | 29/50 [04:23<03:59, 11.42s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:04<00:00,  4.96s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:05<00:00,  5.14s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.86s/it]\u001b[A\n",
      "sampling examples...:  60%|██████    | 30/50 [04:37<04:04, 12.21s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.84s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.95s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.38s/it]\u001b[A\n",
      "sampling examples...:  62%|██████▏   | 31/50 [04:46<03:34, 11.31s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.78s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.96s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.23s/it]\u001b[A\n",
      "sampling examples...:  64%|██████▍   | 32/50 [04:55<03:11, 10.63s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.81s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.79s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.22s/it]\u001b[A\n",
      "sampling examples...:  66%|██████▌   | 33/50 [05:04<02:51, 10.10s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.85s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.82s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.14s/it]\u001b[A\n",
      "sampling examples...:  68%|██████▊   | 34/50 [05:13<02:35,  9.73s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.80s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.95s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.18s/it]\u001b[A\n",
      "sampling examples...:  70%|███████   | 35/50 [05:22<02:22,  9.50s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.70s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.99s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.11s/it]\u001b[A\n",
      "sampling examples...:  72%|███████▏  | 36/50 [05:31<02:10,  9.31s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.79s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.89s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.14s/it]\u001b[A\n",
      "sampling examples...:  74%|███████▍  | 37/50 [05:40<01:59,  9.18s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.82s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.88s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.16s/it]\u001b[A\n",
      "sampling examples...:  76%|███████▌  | 38/50 [05:49<01:49,  9.10s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.85s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.85s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.20s/it]\u001b[A\n",
      "sampling examples...:  78%|███████▊  | 39/50 [05:58<01:39,  9.05s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.79s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.82s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.25s/it]\u001b[A\n",
      "sampling examples...:  80%|████████  | 40/50 [06:06<01:30,  9.01s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.83s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.75s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.12s/it]\u001b[A\n",
      "sampling examples...:  82%|████████▏ | 41/50 [06:15<01:20,  8.93s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.74s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.81s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.18s/it]\u001b[A\n",
      "sampling examples...:  84%|████████▍ | 42/50 [06:24<01:11,  8.88s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.87s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.90s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.24s/it]\u001b[A\n",
      "sampling examples...:  86%|████████▌ | 43/50 [06:33<01:02,  8.94s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.77s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.84s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.13s/it]\u001b[A\n",
      "sampling examples...:  88%|████████▊ | 44/50 [06:42<00:53,  8.89s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.77s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.82s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.12s/it]\u001b[A\n",
      "sampling examples...:  90%|█████████ | 45/50 [06:51<00:44,  8.85s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.76s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.83s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.28s/it]\u001b[A\n",
      "sampling examples...:  92%|█████████▏| 46/50 [07:00<00:35,  8.87s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.88s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.89s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.17s/it]\u001b[A\n",
      "sampling examples...:  94%|█████████▍| 47/50 [07:09<00:26,  8.90s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.80s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.94s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.12s/it]\u001b[A\n",
      "sampling examples...:  96%|█████████▌| 48/50 [07:17<00:17,  8.91s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.94s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.90s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.18s/it]\u001b[A\n",
      "sampling examples...:  98%|█████████▊| 49/50 [07:26<00:08,  8.95s/it]\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  3.00s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:02<00:00,  2.79s/it]\u001b[A\n",
      "\n",
      "  0%|          | 0/1 [00:00<?, ?it/s]\u001b[A\n",
      "100%|██████████| 1/1 [00:03<00:00,  3.17s/it]\u001b[A\n",
      "sampling examples...: 100%|██████████| 50/50 [07:35<00:00,  9.12s/it]\n",
      "  4%|▍         | 2/50 [00:18<06:28,  8.10s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parsing the prediction failed\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 50/50 [01:35<00:00,  1.92s/it]\n"
     ]
    }
   ],
   "source": [
    "for kb_name in [\"carcinogenesis\", \"mutagenesis\", \"vicodi\"]:\n",
    "    run_experiment(kb_name, nces_args);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "resident-narrative",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "overall-corruption",
   "metadata": {},
   "outputs": [],
   "source": [
    "def answer_rq3(baseline = \"EvoLearner\", kb_name=\"semantic_bible\", metric=\"max\"):\n",
    "    with open(f\"datasets/{kb_name}/RQ3_F1.json\") as file:\n",
    "        F1_roces = list(json.load(file).values())\n",
    "    with open(f\"datasets/{kb_name}/Results/{baseline}_results.json\") as file:\n",
    "        F1_baseline = json.load(file)[\"F-measure\"]\n",
    "    if metric == \"max\":\n",
    "        print(\"Exact prediction %\", np.array(F1_roces).max(0).tolist().count(100))\n",
    "        freq = 100*(np.array(F1_roces).max(0) >= np.array(F1_baseline)).mean()\n",
    "        f1 = np.array(F1_roces).max(0).mean()\n",
    "        return freq, f1\n",
    "    elif metric == \"min\":\n",
    "        freq = 100*(np.array(F1_roces).min(0) >= np.array(F1_baseline)).mean()\n",
    "        f1 = np.array(F1_roces).min(0).mean()\n",
    "        return freq, f1\n",
    "    elif metric == \"mean\":\n",
    "        freq = 100*(np.array(F1_roces).mean(0) >= np.array(F1_baseline)).mean()\n",
    "        f1 = np.array(F1_roces).mean(0).mean()\n",
    "        return freq, f1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "french-castle",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Exact prediction % 34\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(63.0, 80.85572)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "answer_rq3()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "derived-electric",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Exact prediction % 80\n",
      "Exact prediction % 47\n",
      "Exact prediction % 34\n",
      "Exact prediction % 68\n",
      "max : [(91.0, 94.83006999999999), (81.0, 92.95210000000002), (63.0, 80.85572), (85.0, 89.70136000000001)]\n",
      "min : [(80.0, 89.52537000000001), (42.0, 72.47313), (44.0, 63.24982), (73.0, 78.16376)]\n",
      "mean : [(82.0, 93.55147239999998), (56.00000000000001, 87.90030000000003), (47.0, 74.7856508), (79.0, 85.0581606)]\n"
     ]
    }
   ],
   "source": [
    "for metric in [\"max\", \"min\", \"mean\"]:\n",
    "    rq3_results = []\n",
    "    for kb_name in [\"carcinogenesis\", \"mutagenesis\", \"semantic_bible\", \"vicodi\"]:\n",
    "        res = answer_rq3(baseline = \"EvoLearner\", kb_name=kb_name, metric=metric)\n",
    "        rq3_results.append(res)\n",
    "    print(metric, \":\", rq3_results)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dutch-bride",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "increasing-museum",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import matplotlib.ticker as ticker\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import json\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "executed-huntington",
   "metadata": {},
   "outputs": [],
   "source": [
    "def violin_plot(names=[\"Carcinogenesis\", \"Mutagenesis\", \"Semantic Bible\", \"Vicodi\"], last=False):\n",
    "    sns.set_theme(style=\"whitegrid\")\n",
    "    #sns.set(style=\"darkgrid\")\n",
    "    min_data = []\n",
    "    mean_data = []\n",
    "    max_data = []\n",
    "    for it, kb_name in enumerate([\"carcinogenesis\", \"mutagenesis\", \"semantic_bible\", \"vicodi\"]):\n",
    "        with open(f\"datasets/{kb_name}/RQ3_F1.json\") as file:\n",
    "            data = np.array(list(json.load(file).values()))\n",
    "        if last and it >= 2:\n",
    "            min_data.append(data.min(0).tolist())\n",
    "            mean_data.append(data.mean(0).tolist())\n",
    "            max_data.append(data.max(0).tolist())\n",
    "        elif (not last) and it <= 1:\n",
    "            min_data.append(data.min(0).tolist())\n",
    "            mean_data.append(data.mean(0).tolist())\n",
    "            max_data.append(data.max(0).tolist())\n",
    "    # Set up the matplotlib figure\n",
    "    fig, axs = plt.subplots(2, 1, figsize=(15, 15))\n",
    "    if last:\n",
    "        names = names[2:]\n",
    "    for i, (ax, name) in enumerate(zip(axs.flat, names)):\n",
    "        sns.violinplot(data=pd.DataFrame(np.array([min_data[i], mean_data[i], max_data[i]]).T, columns=[\"Min\", \"Avg.\", \"Max\"], index=list(range(100))), ax=ax, bw_adjust=.5, cut=1, linewidth=1, palette=['tab:blue', 'tab:orange', 'tab:green'])\n",
    "        ax.set_title(name, fontsize=30, fontweight=\"bold\")\n",
    "        ax.tick_params(axis='both', which='major', labelsize=30)\n",
    "        #ax.legend([\"Min\", \"Mean\", \"Max\"])\n",
    "        #ax.set_xlabel(\"Aggregation\", fontsize=30)\n",
    "        #if i%2 == 0:\n",
    "        ax.set_ylabel(\"F$_1$\", fontsize=30)\n",
    "    sns.despine(left=True, bottom=True)\n",
    "    fig.savefig(f'f1_distribution_last_{last}.pdf', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "familiar-destruction",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x1500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "violin_plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "sunrise-bulletin",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x1500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    " violin_plot(last=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "lovely-reducing",
   "metadata": {},
   "source": [
    "### Standard deviation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dressed-problem",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "checked-gossip",
   "metadata": {},
   "outputs": [],
   "source": [
    "import json, numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "compact-hepatitis",
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(f\"datasets/carcinogenesis/RQ3_F1.json\") as file:\n",
    "    data1 = np.array(list(json.load(file).values()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "according-clinton",
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(f\"datasets/vicodi/RQ3_F1.json\") as file:\n",
    "    data2 = np.array(list(json.load(file).values()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "smoking-testament",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.307669957277256"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data1.std(axis=0).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "finnish-planner",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.980121428655435"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data2.std(axis=0).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "statistical-party",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([9.758500e+01, 5.249042e+01, 1.140000e-01, 1.000000e+02,\n",
       "       9.980596e+01, 1.000000e+02, 1.000000e+02, 1.000000e+02,\n",
       "       9.965400e+01, 1.000000e+02, 7.857100e+01, 9.357100e+01,\n",
       "       1.000000e+02, 1.000000e+02, 1.000000e+02, 1.000000e+02,\n",
       "       1.826900e+01, 9.860432e+01, 9.999144e+01, 1.000000e+02,\n",
       "       1.000000e+02, 1.000000e+02, 1.000000e+02, 1.000000e+02,\n",
       "       2.700060e+01, 1.000000e+02, 4.673362e+01, 9.434000e+01,\n",
       "       1.000000e+02, 1.000000e+02, 1.000000e+02, 9.894700e+01,\n",
       "       9.991426e+01, 1.000000e+02, 9.706090e+01, 9.984180e+01,\n",
       "       1.000000e+02, 9.985000e+01, 1.000000e+02, 9.885100e+01,\n",
       "       1.000000e+02, 9.750400e+01, 9.689994e+01, 1.565944e+01,\n",
       "       1.000000e+02, 1.000000e+02, 2.400000e-02, 6.454140e+00,\n",
       "       9.592000e-02, 9.411800e+01, 1.000000e+02, 1.000000e+02,\n",
       "       1.000000e+02, 1.000000e+02, 1.730000e+00, 1.000000e+02,\n",
       "       1.000000e+02, 1.000000e+02, 1.000000e+02, 1.000000e+02,\n",
       "       9.999800e+01, 1.686000e+00, 1.000000e+02, 1.150000e-01,\n",
       "       9.101016e+01, 1.793206e+01, 7.462662e+01, 1.000000e+02,\n",
       "       9.994690e+01, 1.000000e+02, 9.989100e+01, 9.612306e+01,\n",
       "       9.974422e+01, 1.000000e+02, 1.000000e+02, 1.000000e+02,\n",
       "       1.000000e+02, 1.000000e+02, 1.000000e+02, 1.000000e+02,\n",
       "       1.000000e+02, 9.976900e+01, 1.000000e+02, 1.000000e+02,\n",
       "       3.613610e+01, 1.000000e+02, 9.996784e+01, 9.450500e+01,\n",
       "       9.998290e+01, 9.915500e+01, 4.762000e+00, 1.000000e+02,\n",
       "       1.000000e+02, 6.012660e+00, 7.937440e+01, 9.190562e+01,\n",
       "       9.960976e+01, 9.988200e+01, 1.000000e+02, 1.000000e+02])"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data2.mean(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "limiting-broadway",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "roces",
   "language": "python",
   "name": "roces"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}