Token Classification/Monolingual/StrombergNLP-Twitter_pos_vcb/NER Project Using StrombergNLP Twitter_pos_vcb Dataset.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Monolingual Named Entity Recognition (Stromberg NLP Twitter P.o.S. Dataset)\n",
    "\n",
    "Dataset Source: https://huggingface.co/datasets/strombergnlp/twitter_pos_vcb"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Import Necessary Libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Git LFS initialized.\n"
     ]
    }
   ],
   "source": [
    "import os, sys\n",
    "from collections import Counter, defaultdict\n",
    "os.environ['TOKENIZERS_PARALLELISM']='false'\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import sklearn\n",
    "from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay\n",
    "\n",
    "import datasets\n",
    "from datasets import load_dataset, DatasetDict\n",
    "\n",
    "import torch\n",
    "from torch.nn.functional import cross_entropy\n",
    "\n",
    "import transformers\n",
    "from transformers import AutoTokenizer, AutoModelForTokenClassification\n",
    "from transformers import DataCollatorForTokenClassification\n",
    "from transformers import Trainer, TrainingArguments\n",
    "\n",
    "import evaluate\n",
    "\n",
    "!git lfs install"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Display Library Versions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "          Python : 3.9.12\n",
      "           NumPy : 1.24.3\n",
      "          Pandas : 2.0.1\n",
      "           Torch : 2.0.0\n",
      "    Transformers : 4.28.1\n",
      "        datasets : 2.11.0\n",
      "         Sklearn : 1.2.2\n"
     ]
    }
   ],
   "source": [
    "print(\"Python :\".rjust(18), sys.version[0:6])\n",
    "print(\"NumPy :\".rjust(18), np.__version__)\n",
    "print(\"Pandas :\".rjust(18), pd.__version__)\n",
    "print(\"Torch :\".rjust(18), torch.__version__)\n",
    "print(\"Transformers :\".rjust(18), transformers.__version__)\n",
    "print(\"datasets :\".rjust(18), datasets.__version__)\n",
    "print(\"Sklearn :\".rjust(18), sklearn.__version__)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Ingest English Subset of Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Found cached dataset twitter_pos_vcb (/Users/briandunn/.cache/huggingface/datasets/strombergnlp___twitter_pos_vcb/twitter-pos-vcb/1.0.0/2c671a933c7ffb84cf124a1f987cd73fb883e6b3628ee4a01f5326f5aefc296a)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c3de061201384d11b382764c1b8092ef",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "DatasetDict({\n",
       "    train: Dataset({\n",
       "        features: ['id', 'tokens', 'pos_tags'],\n",
       "        num_rows: 159492\n",
       "    })\n",
       "})"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = load_dataset(\"strombergnlp/twitter_pos_vcb\")\n",
    "\n",
    "data"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Split Dataset into DatasetDict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training data shape: (119619, 3)\n",
      "Testing data shape: (31899, 3)\n",
      "Validation data shape: (7974, 3)\n"
     ]
    }
   ],
   "source": [
    "# Split Dataset into train/test/eval datasets\n",
    "# then pack all of them into DatasetDict\n",
    "train_testvalid = data['train'].train_test_split(test_size=0.25)\n",
    "test_valid = train_testvalid['test'].train_test_split(test_size=0.80)\n",
    "\n",
    "ds = DatasetDict({\n",
    "    'train': train_testvalid['train'], \n",
    "    'test': test_valid['test'], \n",
    "    'eval': test_valid['train']})\n",
    "\n",
    "print('Training data shape:', ds['train'].shape)\n",
    "print('Testing data shape:', ds['test'].shape)\n",
    "print('Validation data shape:', ds['eval'].shape)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Example"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'id': '96776',\n",
       " 'tokens': ['RT',\n",
       "  '@CAmatty13',\n",
       "  ':',\n",
       "  'We',\n",
       "  'are',\n",
       "  'getting',\n",
       "  'married',\n",
       "  'you',\n",
       "  'just',\n",
       "  'do',\n",
       "  'know',\n",
       "  'yet'],\n",
       " 'pos_tags': [47, 49, 8, 28, 41, 39, 38, 28, 30, 41, 37, 30]}"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "example = ds['train'][12]\n",
    "\n",
    "example"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Display Feature Information About Each Feature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "id: \n",
      "Value(dtype='string', id=None)\n",
      "\n",
      "tokens: \n",
      "Sequence(feature=Value(dtype='string', id=None), length=-1, id=None)\n",
      "\n",
      "pos_tags: \n",
      "Sequence(feature=ClassLabel(names=['\"', \"''\", '#', '$', '(', ')', ',', '.', ':', '``', 'CC', 'CD', 'DT', 'EX', 'FW', 'IN', 'JJ', 'JJR', 'JJS', 'LS', 'MD', 'NN', 'NNP', 'NNPS', 'NNS', 'NN|SYM', 'PDT', 'POS', 'PRP', 'PRP$', 'RB', 'RBR', 'RBS', 'RP', 'SYM', 'TO', 'UH', 'VB', 'VBD', 'VBG', 'VBN', 'VBP', 'VBZ', 'WDT', 'WP', 'WP$', 'WRB', 'RT', 'HT', 'USR', 'URL'], id=None), length=-1, id=None)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for k, v in ds[\"train\"].features.items():\n",
    "    print(f\"{k}: \\n{v}\\n\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Define Tag Values & Conversions Between String & Integer Values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "List of tag values: \n",
      "['\"', \"''\", '#', '$', '(', ')', ',', '.', ':', '``', 'CC', 'CD', 'DT', 'EX', 'FW', 'IN', 'JJ', 'JJR', 'JJS', 'LS', 'MD', 'NN', 'NNP', 'NNPS', 'NNS', 'NN|SYM', 'PDT', 'POS', 'PRP', 'PRP$', 'RB', 'RBR', 'RBS', 'RP', 'SYM', 'TO', 'UH', 'VB', 'VBD', 'VBG', 'VBN', 'VBP', 'VBZ', 'WDT', 'WP', 'WP$', 'WRB', 'RT', 'HT', 'USR', 'URL']\n",
      "Number of NER Tags: \n",
      "51\n",
      "id2label: \n",
      "{0: '\"', 1: \"''\", 2: '#', 3: '$', 4: '(', 5: ')', 6: ',', 7: '.', 8: ':', 9: '``', 10: 'CC', 11: 'CD', 12: 'DT', 13: 'EX', 14: 'FW', 15: 'IN', 16: 'JJ', 17: 'JJR', 18: 'JJS', 19: 'LS', 20: 'MD', 21: 'NN', 22: 'NNP', 23: 'NNPS', 24: 'NNS', 25: 'NN|SYM', 26: 'PDT', 27: 'POS', 28: 'PRP', 29: 'PRP$', 30: 'RB', 31: 'RBR', 32: 'RBS', 33: 'RP', 34: 'SYM', 35: 'TO', 36: 'UH', 37: 'VB', 38: 'VBD', 39: 'VBG', 40: 'VBN', 41: 'VBP', 42: 'VBZ', 43: 'WDT', 44: 'WP', 45: 'WP$', 46: 'WRB', 47: 'RT', 48: 'HT', 49: 'USR', 50: 'URL'}\n",
      "label2id: \n",
      "{'\"': 0, \"''\": 1, '#': 2, '$': 3, '(': 4, ')': 5, ',': 6, '.': 7, ':': 8, '``': 9, 'CC': 10, 'CD': 11, 'DT': 12, 'EX': 13, 'FW': 14, 'IN': 15, 'JJ': 16, 'JJR': 17, 'JJS': 18, 'LS': 19, 'MD': 20, 'NN': 21, 'NNP': 22, 'NNPS': 23, 'NNS': 24, 'NN|SYM': 25, 'PDT': 26, 'POS': 27, 'PRP': 28, 'PRP$': 29, 'RB': 30, 'RBR': 31, 'RBS': 32, 'RP': 33, 'SYM': 34, 'TO': 35, 'UH': 36, 'VB': 37, 'VBD': 38, 'VBG': 39, 'VBN': 40, 'VBP': 41, 'VBZ': 42, 'WDT': 43, 'WP': 44, 'WP$': 45, 'WRB': 46, 'RT': 47, 'HT': 48, 'USR': 49, 'URL': 50}\n"
     ]
    }
   ],
   "source": [
    "pos_tag_values = ds['train'].features[f'pos_tags'].feature.names\n",
    "NUM_OF_LABELS = len(pos_tag_values)\n",
    "\n",
    "label2id = {tag: idx for idx, tag in enumerate(pos_tag_values)}\n",
    "id2label = {idx: tag for idx, tag in enumerate(pos_tag_values)}\n",
    "\n",
    "print(f\"List of tag values: \\n{pos_tag_values}\")\n",
    "print(f\"Number of NER Tags: \\n{NUM_OF_LABELS}\")\n",
    "print(f\"id2label: \\n{id2label}\")\n",
    "print(f\"label2id: \\n{label2id}\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Basic Values/Constants"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "MODEL_CKPT = \"bert-base-cased\"\n",
    "\n",
    "MODEL_NAME = f\"{MODEL_CKPT}-finetuned-Stromberg_NLP_Twitter-PoS\"\n",
    "NUM_OF_EPOCHS = 3\n",
    "\n",
    "BATCH_SIZE = 16\n",
    "STRATEGY = \"epoch\"\n",
    "\n",
    "REPORTS_TO = \"tensorboard\"\n",
    "WEIGHT_DECAY = 0.01\n",
    "\n",
    "LR = 2e-5\n",
    "DEVICE = torch.device(\"cpu\")\n",
    "\n",
    "STEPS = 500"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Function to Tokenize & Align Inputs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "tokenizer = AutoTokenizer.from_pretrained(MODEL_CKPT)\n",
    "\n",
    "def tokenize_and_align_labels(samples):\n",
    "    tokenized_inputs = tokenizer(samples[\"tokens\"], \n",
    "                                      truncation=True, \n",
    "                                      is_split_into_words=True)\n",
    "    \n",
    "    labels = []\n",
    "    \n",
    "    for idx, label in enumerate(samples[f\"pos_tags\"]):\n",
    "        word_ids = tokenized_inputs.word_ids(batch_index=idx)\n",
    "        prev_word_idx = None\n",
    "        label_ids = []\n",
    "        for word_idx in word_ids: # set special tokens to -100\n",
    "            if word_idx is None or word_idx == prev_word_idx:\n",
    "                label_ids.append(-100)\n",
    "            else:\n",
    "                label_ids.append(label[word_idx])\n",
    "            prev_word_idx = word_idx\n",
    "        labels.append(label_ids)\n",
    "    \n",
    "    tokenized_inputs[\"labels\"] = labels\n",
    "    return tokenized_inputs"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Apply Above Function to Tokenize Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e98ed091ecce4a0a95a52a85ee49987c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/119619 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7fe2fd95b66b406c95f1c552c976638b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/31899 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1b68bce24e09404aa1d86487bddb15de",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/7974 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "encoded_ds = ds.map(tokenize_and_align_labels, \n",
    "                       batched=True, \n",
    "                       remove_columns=\n",
    "                        [\n",
    "                            'pos_tags', \n",
    "                            'tokens'\n",
    "                        ]\n",
    "                    )"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Define Data Collator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_collator = DataCollatorForTokenClassification(tokenizer=tokenizer)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Load Custom Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Some weights of the model checkpoint at bert-base-cased were not used when initializing BertForTokenClassification: ['cls.seq_relationship.weight', 'cls.seq_relationship.bias', 'cls.predictions.transform.LayerNorm.weight', 'cls.predictions.transform.dense.bias', 'cls.predictions.bias', 'cls.predictions.transform.LayerNorm.bias', 'cls.predictions.transform.dense.weight']\n",
      "- This IS expected if you are initializing BertForTokenClassification from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\n",
      "- This IS NOT expected if you are initializing BertForTokenClassification from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n",
      "Some weights of BertForTokenClassification were not initialized from the model checkpoint at bert-base-cased and are newly initialized: ['classifier.bias', 'classifier.weight']\n",
      "You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n"
     ]
    }
   ],
   "source": [
    "model = (AutoModelForTokenClassification.from_pretrained(\n",
    "    MODEL_CKPT,\n",
    "    num_labels=NUM_OF_LABELS,\n",
    "    id2label=id2label,\n",
    "    label2id=label2id\n",
    "    ).to(DEVICE))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Create Compute Metrics Function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "label_list = pos_tag_values\n",
    "\n",
    "seqeval = evaluate.load(\"seqeval\")\n",
    "\n",
    "labels = [label_list[i] for i in example[f'pos_tags']]\n",
    "\n",
    "def compute_metrics(eval_preds):\n",
    "    predictions, labels = eval_preds\n",
    "    predictions = np.argmax(predictions, \n",
    "                            axis=2)\n",
    "    \n",
    "    true_predictions = [\n",
    "        [label_list[p] for (p, l) in zip(prediction, label) if l != -100]\n",
    "        for prediction, label in zip(predictions, labels)\n",
    "    ]\n",
    "    \n",
    "    true_labels = [\n",
    "        [label_list[l] for (p, l) in zip(prediction, label) if l != -100]\n",
    "        for prediction, label in zip(predictions, labels)\n",
    "    ]\n",
    "    \n",
    "    results = seqeval.compute(predictions=true_predictions, \n",
    "                              references=true_labels)\n",
    "    \n",
    "    return results"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Define TrainingArguments"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "args = TrainingArguments(\n",
    "    MODEL_NAME,\n",
    "    log_level=\"error\",\n",
    "    logging_first_step=True,\n",
    "    learning_rate=LR,\n",
    "    num_train_epochs=NUM_OF_EPOCHS,\n",
    "    per_device_train_batch_size=BATCH_SIZE,\n",
    "    per_device_eval_batch_size=BATCH_SIZE,\n",
    "    evaluation_strategy=STRATEGY,\n",
    "    report_to=REPORTS_TO,\n",
    "    disable_tqdm=False,\n",
    "    logging_steps=STEPS,\n",
    "    weight_decay=WEIGHT_DECAY,\n",
    "    save_strategy=STRATEGY,\n",
    "    hub_private_repo=True,\n",
    "    push_to_hub=True\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Define Trainer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Cloning https://huggingface.co/DunnBC22/bert-base-cased-finetuned-Stromberg_NLP_Twitter-PoS into local empty directory.\n"
     ]
    }
   ],
   "source": [
    "trainer = Trainer(model, \n",
    "                  args=args,\n",
    "                  data_collator=data_collator,\n",
    "                  compute_metrics=compute_metrics,\n",
    "                  tokenizer=tokenizer,\n",
    "                  train_dataset=encoded_ds[\"train\"],\n",
    "                  eval_dataset=encoded_ds[\"test\"]\n",
    "                  )"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Train Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/transformers/optimization.py:391: FutureWarning: This implementation of AdamW is deprecated and will be removed in a future version. Use the PyTorch implementation torch.optim.AdamW instead, or set `no_deprecation_warning=True` to disable this warning\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ba103ac12f81481984a6b35752cec45c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/22431 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'loss': 3.9467, 'learning_rate': 1.9999108376799965e-05, 'epoch': 0.0}\n",
      "{'loss': 0.5493, 'learning_rate': 1.9554188399982172e-05, 'epoch': 0.07}\n",
      "{'loss': 0.1294, 'learning_rate': 1.9108376799964335e-05, 'epoch': 0.13}\n",
      "{'loss': 0.1118, 'learning_rate': 1.8662565199946505e-05, 'epoch': 0.2}\n",
      "{'loss': 0.1006, 'learning_rate': 1.8216753599928672e-05, 'epoch': 0.27}\n",
      "{'loss': 0.0932, 'learning_rate': 1.777094199991084e-05, 'epoch': 0.33}\n",
      "{'loss': 0.0856, 'learning_rate': 1.7325130399893006e-05, 'epoch': 0.4}\n",
      "{'loss': 0.0801, 'learning_rate': 1.6879318799875176e-05, 'epoch': 0.47}\n",
      "{'loss': 0.0791, 'learning_rate': 1.6433507199857343e-05, 'epoch': 0.53}\n",
      "{'loss': 0.0769, 'learning_rate': 1.598769559983951e-05, 'epoch': 0.6}\n",
      "{'loss': 0.0725, 'learning_rate': 1.5541883999821676e-05, 'epoch': 0.67}\n",
      "{'loss': 0.0674, 'learning_rate': 1.5096072399803845e-05, 'epoch': 0.74}\n",
      "{'loss': 0.0683, 'learning_rate': 1.4650260799786012e-05, 'epoch': 0.8}\n",
      "{'loss': 0.064, 'learning_rate': 1.420444919976818e-05, 'epoch': 0.87}\n",
      "{'loss': 0.0617, 'learning_rate': 1.3758637599750347e-05, 'epoch': 0.94}\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d8cdcaf8e6d34e358f72313a425f8227",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1994 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: HT seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: VBP seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: JJ seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: NNS seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: IN seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: PRP seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: . seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: DT seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: NN seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: WRB seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: CC seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: PRP$ seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: RT seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: USR seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: : seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: VBZ seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: VBD seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: TO seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: VB seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: URL seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: RB seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: VBG seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: UH seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: MD seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: WP seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: NNP seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: , seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: VBN seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: POS seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: CD seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: JJS seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: JJR seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: RP seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: ) seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: `` seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: '' seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: ( seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: RBS seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: RBR seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: WDT seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: NNPS seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: EX seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: FW seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: SYM seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: PDT seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/v1.py:57: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'eval_loss': 0.059514690190553665, \"eval_'\": {'precision': 0.9331210191082803, 'recall': 0.9391025641025641, 'f1': 0.9361022364217252, 'number': 312}, 'eval_B': {'precision': 0.9562652403051994, 'recall': 0.9536397866331974, 'f1': 0.9549507089273791, 'number': 25496}, 'eval_BD': {'precision': 0.9716325380424573, 'recall': 0.9322278298485941, 'f1': 0.9515223990433264, 'number': 5548}, 'eval_BG': {'precision': 0.9810585944414941, 'recall': 0.9786332332685855, 'f1': 0.9798444130127298, 'number': 5663}, 'eval_BN': {'precision': 0.8725314183123878, 'recall': 0.9230769230769231, 'f1': 0.8970927549607752, 'number': 2106}, 'eval_BP': {'precision': 0.9555667442885015, 'recall': 0.9585832438916598, 'f1': 0.9570726172465961, 'number': 15839}, 'eval_BR': {'precision': 0.8878504672897196, 'recall': 0.8878504672897196, 'f1': 0.8878504672897196, 'number': 107}, 'eval_BS': {'precision': 0.8589743589743589, 'recall': 1.0, 'f1': 0.9241379310344827, 'number': 67}, 'eval_BZ': {'precision': 0.9792873442162542, 'recall': 0.9834302838004583, 'f1': 0.9813544415127529, 'number': 5673}, 'eval_C': {'precision': 0.998475277717273, 'recall': 0.999128160418483, 'f1': 0.9988016123760759, 'number': 4588}, 'eval_D': {'precision': 0.9818369757826344, 'recall': 0.9885518881950639, 'f1': 0.9851829900726033, 'number': 6726}, 'eval_DT': {'precision': 1.0, 'recall': 0.8, 'f1': 0.888888888888889, 'number': 15}, 'eval_H': {'precision': 0.9391025641025641, 'recall': 0.9105438401775805, 'f1': 0.9246027273751831, 'number': 9010}, 'eval_J': {'precision': 0.9706609264131388, 'recall': 0.9765781663591883, 'f1': 0.973610555777689, 'number': 12467}, 'eval_JR': {'precision': 0.9211538461538461, 'recall': 0.9676767676767677, 'f1': 0.9438423645320195, 'number': 495}, 'eval_JS': {'precision': 0.9226973684210527, 'recall': 0.9756521739130435, 'f1': 0.94843617920541, 'number': 575}, 'eval_N': {'precision': 0.9754043126684636, 'recall': 0.9738394659214408, 'f1': 0.9746212611679399, 'number': 38646}, 'eval_NP': {'precision': 0.9158227848101266, 'recall': 0.9200445080273406, 'f1': 0.9179287923241615, 'number': 6291}, 'eval_NPS': {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 26}, 'eval_NS': {'precision': 0.965710643722116, 'recall': 0.9687979539641943, 'f1': 0.9672518353016277, 'number': 7820}, 'eval_O': {'precision': 0.9972154542290289, 'recall': 0.9989539748953975, 'f1': 0.9980839574986937, 'number': 5736}, 'eval_OS': {'precision': 1.0, 'recall': 0.9928400954653938, 'f1': 0.9964071856287425, 'number': 419}, 'eval_P': {'precision': 0.9771428571428571, 'recall': 0.99079754601227, 'f1': 0.9839228295819935, 'number': 2934}, 'eval_RB': {'precision': 0.9947874899759422, 'recall': 0.9967858577742065, 'f1': 0.99578567128236, 'number': 2489}, 'eval_RL': {'precision': 1.0, 'recall': 0.9997228381374723, 'f1': 0.9998613998613999, 'number': 3608}, 'eval_RP': {'precision': 0.9970126960418223, 'recall': 0.9976222826086957, 'f1': 0.9973173961764407, 'number': 29440}, 'eval_RP$': {'precision': 0.9973994452149791, 'recall': 0.9953287197231834, 'f1': 0.9963630065812261, 'number': 5780}, 'eval_SR': {'precision': 0.9998384665087228, 'recall': 0.9997846336079255, 'f1': 0.9998115493336922, 'number': 18573}, 'eval_T': {'precision': 0.9976884595480421, 'recall': 0.9982050396962375, 'f1': 0.9979466827711155, 'number': 28970}, 'eval_W': {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 1}, 'eval_X': {'precision': 0.8860759493670886, 'recall': 0.9459459459459459, 'f1': 0.9150326797385621, 'number': 74}, 'eval_YM': {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 5}, 'eval__': {'precision': 0.9936347529365444, 'recall': 0.9925925925925926, 'f1': 0.9931133993572506, 'number': 15255}, 'eval_`': {'precision': 0.9540229885057471, 'recall': 0.9595375722543352, 'f1': 0.956772334293948, 'number': 173}, 'eval_overall_precision': 0.9778566086939848, 'eval_overall_recall': 0.9772120171542232, 'eval_overall_f1': 0.9775342066622962, 'eval_overall_accuracy': 0.9820657081787647, 'eval_runtime': 1947.4181, 'eval_samples_per_second': 16.38, 'eval_steps_per_second': 1.024, 'epoch': 1.0}\n",
      "{'loss': 0.0627, 'learning_rate': 1.3312825999732516e-05, 'epoch': 1.0}\n",
      "{'loss': 0.0461, 'learning_rate': 1.286701439971468e-05, 'epoch': 1.07}\n",
      "{'loss': 0.0462, 'learning_rate': 1.242120279969685e-05, 'epoch': 1.14}\n",
      "{'loss': 0.0459, 'learning_rate': 1.1975391199679016e-05, 'epoch': 1.2}\n",
      "{'loss': 0.0444, 'learning_rate': 1.1529579599661185e-05, 'epoch': 1.27}\n",
      "{'loss': 0.0451, 'learning_rate': 1.1083767999643351e-05, 'epoch': 1.34}\n",
      "{'loss': 0.0432, 'learning_rate': 1.063795639962552e-05, 'epoch': 1.4}\n",
      "{'loss': 0.0461, 'learning_rate': 1.0192144799607687e-05, 'epoch': 1.47}\n",
      "{'loss': 0.042, 'learning_rate': 9.746333199589853e-06, 'epoch': 1.54}\n",
      "{'loss': 0.0429, 'learning_rate': 9.300521599572022e-06, 'epoch': 1.6}\n",
      "{'loss': 0.0422, 'learning_rate': 8.85470999955419e-06, 'epoch': 1.67}\n",
      "{'loss': 0.0405, 'learning_rate': 8.408898399536357e-06, 'epoch': 1.74}\n",
      "{'loss': 0.0408, 'learning_rate': 7.963086799518524e-06, 'epoch': 1.81}\n",
      "{'loss': 0.039, 'learning_rate': 7.517275199500692e-06, 'epoch': 1.87}\n",
      "{'loss': 0.0407, 'learning_rate': 7.071463599482859e-06, 'epoch': 1.94}\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8208ffaf77fd4b36b505c005585f5f45",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1994 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: HT seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: VBP seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: JJ seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: NNS seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: IN seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: PRP seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: . seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: DT seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: NN seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: WRB seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: CC seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: PRP$ seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: RT seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: USR seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: : seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: VBZ seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: VBD seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: TO seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: VB seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: URL seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: RB seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: VBG seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: UH seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: MD seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: WP seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: NNP seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: , seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: VBN seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: POS seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: CD seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: JJS seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: JJR seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: RP seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: ) seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: `` seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: '' seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: ( seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: RBS seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: RBR seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: WDT seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: NNPS seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: EX seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: FW seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: SYM seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: PDT seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/v1.py:57: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'eval_loss': 0.05311541631817818, \"eval_'\": {'precision': 0.9605263157894737, 'recall': 0.9358974358974359, 'f1': 0.948051948051948, 'number': 312}, 'eval_B': {'precision': 0.9599141295862608, 'recall': 0.9645826796360213, 'f1': 0.9622427419985915, 'number': 25496}, 'eval_BD': {'precision': 0.9673732718894009, 'recall': 0.9459264599855803, 'f1': 0.95652966372004, 'number': 5548}, 'eval_BG': {'precision': 0.9833863556026865, 'recall': 0.9825180999470245, 'f1': 0.9829520360392192, 'number': 5663}, 'eval_BN': {'precision': 0.8920402561756633, 'recall': 0.9259259259259259, 'f1': 0.9086672879776329, 'number': 2106}, 'eval_BP': {'precision': 0.972785622593068, 'recall': 0.9568785908201275, 'f1': 0.9647665425379548, 'number': 15839}, 'eval_BR': {'precision': 0.9591836734693877, 'recall': 0.8785046728971962, 'f1': 0.9170731707317074, 'number': 107}, 'eval_BS': {'precision': 0.9428571428571428, 'recall': 0.9850746268656716, 'f1': 0.9635036496350364, 'number': 67}, 'eval_BZ': {'precision': 0.988999290276792, 'recall': 0.9825489159175039, 'f1': 0.9857635511539481, 'number': 5673}, 'eval_C': {'precision': 0.999128350403138, 'recall': 0.9993461203138623, 'f1': 0.9992372234935164, 'number': 4588}, 'eval_D': {'precision': 0.9854900799526207, 'recall': 0.9895926256318763, 'f1': 0.9875370919881307, 'number': 6726}, 'eval_DT': {'precision': 1.0, 'recall': 0.8, 'f1': 0.888888888888889, 'number': 15}, 'eval_H': {'precision': 0.9498016997167139, 'recall': 0.9302996670366259, 'f1': 0.9399495374264086, 'number': 9010}, 'eval_J': {'precision': 0.9775892428365616, 'recall': 0.9797064249618994, 'f1': 0.978646688834582, 'number': 12467}, 'eval_JR': {'precision': 0.9124767225325885, 'recall': 0.98989898989899, 'f1': 0.9496124031007753, 'number': 495}, 'eval_JS': {'precision': 0.948073701842546, 'recall': 0.9843478260869565, 'f1': 0.9658703071672354, 'number': 575}, 'eval_N': {'precision': 0.9787719343718411, 'recall': 0.9771257051182528, 'f1': 0.9779481269504189, 'number': 38646}, 'eval_NP': {'precision': 0.9252336448598131, 'recall': 0.9284692417739628, 'f1': 0.9268486194858775, 'number': 6291}, 'eval_NPS': {'precision': 0.5, 'recall': 0.23076923076923078, 'f1': 0.3157894736842105, 'number': 26}, 'eval_NS': {'precision': 0.9653734361177808, 'recall': 0.9768542199488491, 'f1': 0.9710798957605034, 'number': 7820}, 'eval_O': {'precision': 0.9975635224504003, 'recall': 0.999302649930265, 'f1': 0.9984323288625675, 'number': 5736}, 'eval_OS': {'precision': 0.9928571428571429, 'recall': 0.9952267303102625, 'f1': 0.9940405244338498, 'number': 419}, 'eval_P': {'precision': 0.9861205145565335, 'recall': 0.9928425357873211, 'f1': 0.9894701086956521, 'number': 2934}, 'eval_RB': {'precision': 0.9971910112359551, 'recall': 0.9983929288871033, 'f1': 0.9977916081108211, 'number': 2489}, 'eval_RL': {'precision': 1.0, 'recall': 0.9997228381374723, 'f1': 0.9998613998613999, 'number': 3608}, 'eval_RP': {'precision': 0.9986068164055864, 'recall': 0.9982336956521739, 'f1': 0.9984202211690363, 'number': 29440}, 'eval_RP$': {'precision': 0.9963718037318591, 'recall': 0.9977508650519031, 'f1': 0.997060857538036, 'number': 5780}, 'eval_SR': {'precision': 0.9999461555029076, 'recall': 0.9998923168039627, 'f1': 0.9999192354287252, 'number': 18573}, 'eval_T': {'precision': 0.9985499240436404, 'recall': 0.9983431135657577, 'f1': 0.9984465080954189, 'number': 28970}, 'eval_W': {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 1}, 'eval_X': {'precision': 0.9113924050632911, 'recall': 0.972972972972973, 'f1': 0.9411764705882353, 'number': 74}, 'eval_YM': {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 5}, 'eval__': {'precision': 0.9948932827026319, 'recall': 0.9961324156014422, 'f1': 0.9955124635592387, 'number': 15255}, 'eval_`': {'precision': 0.9651162790697675, 'recall': 0.9595375722543352, 'f1': 0.9623188405797102, 'number': 173}, 'eval_overall_precision': 0.9816839981434671, 'eval_overall_recall': 0.9808337197760293, 'eval_overall_f1': 0.9812586747643915, 'eval_overall_accuracy': 0.9849730059835112, 'eval_runtime': 1897.7758, 'eval_samples_per_second': 16.809, 'eval_steps_per_second': 1.051, 'epoch': 2.0}\n",
      "{'loss': 0.0373, 'learning_rate': 6.625651999465027e-06, 'epoch': 2.01}\n",
      "{'loss': 0.0263, 'learning_rate': 6.179840399447195e-06, 'epoch': 2.07}\n",
      "{'loss': 0.0295, 'learning_rate': 5.734028799429362e-06, 'epoch': 2.14}\n",
      "{'loss': 0.0261, 'learning_rate': 5.288217199411529e-06, 'epoch': 2.21}\n",
      "{'loss': 0.0275, 'learning_rate': 4.842405599393697e-06, 'epoch': 2.27}\n",
      "{'loss': 0.0258, 'learning_rate': 4.396593999375864e-06, 'epoch': 2.34}\n",
      "{'loss': 0.0274, 'learning_rate': 3.950782399358032e-06, 'epoch': 2.41}\n",
      "{'loss': 0.0277, 'learning_rate': 3.5049707993401993e-06, 'epoch': 2.47}\n",
      "{'loss': 0.0261, 'learning_rate': 3.059159199322367e-06, 'epoch': 2.54}\n",
      "{'loss': 0.0246, 'learning_rate': 2.613347599304534e-06, 'epoch': 2.61}\n",
      "{'loss': 0.025, 'learning_rate': 2.1675359992867014e-06, 'epoch': 2.67}\n",
      "{'loss': 0.0265, 'learning_rate': 1.7217243992688693e-06, 'epoch': 2.74}\n",
      "{'loss': 0.026, 'learning_rate': 1.2759127992510367e-06, 'epoch': 2.81}\n",
      "{'loss': 0.0238, 'learning_rate': 8.301011992332041e-07, 'epoch': 2.88}\n",
      "{'loss': 0.0246, 'learning_rate': 3.842895992153716e-07, 'epoch': 2.94}\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "07a8bdf7ae944c09bbb8850b2f384f3e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1994 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: HT seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: VBP seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: JJ seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: NNS seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: IN seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: PRP seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: . seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: DT seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: NN seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: WRB seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: CC seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: PRP$ seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: RT seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: USR seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: : seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: VBZ seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: VBD seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: TO seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: VB seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: URL seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: RB seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: VBG seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: UH seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: MD seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: WP seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: NNP seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: , seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: VBN seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: POS seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: CD seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: JJS seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: JJR seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: RP seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: ) seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: `` seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: '' seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: ( seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: RBS seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: RBR seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: WDT seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: NNPS seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: EX seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: FW seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: SYM seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/sequence_labeling.py:171: UserWarning: PDT seems not to be NE tag.\n",
      "  warnings.warn('{} seems not to be NE tag.'.format(chunk))\n",
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/seqeval/metrics/v1.py:57: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'eval_loss': 0.05331805720925331, \"eval_'\": {'precision': 0.9580645161290322, 'recall': 0.9519230769230769, 'f1': 0.954983922829582, 'number': 312}, 'eval_B': {'precision': 0.9658270558694287, 'recall': 0.9655240037652966, 'f1': 0.9656755060411109, 'number': 25496}, 'eval_BD': {'precision': 0.9630099728014506, 'recall': 0.9572819033886085, 'f1': 0.9601373949200036, 'number': 5548}, 'eval_BG': {'precision': 0.9836065573770492, 'recall': 0.9853434575313438, 'f1': 0.9844742413549753, 'number': 5663}, 'eval_BN': {'precision': 0.9182209469153515, 'recall': 0.9116809116809117, 'f1': 0.9149392423159399, 'number': 2106}, 'eval_BP': {'precision': 0.9672037914691943, 'recall': 0.9663488856619736, 'f1': 0.9667761495704902, 'number': 15839}, 'eval_BR': {'precision': 0.94, 'recall': 0.8785046728971962, 'f1': 0.9082125603864735, 'number': 107}, 'eval_BS': {'precision': 0.9848484848484849, 'recall': 0.9701492537313433, 'f1': 0.9774436090225564, 'number': 67}, 'eval_BZ': {'precision': 0.9865819209039548, 'recall': 0.9850167459897762, 'f1': 0.9857987121813531, 'number': 5673}, 'eval_C': {'precision': 0.9993461203138623, 'recall': 0.9993461203138623, 'f1': 0.9993461203138623, 'number': 4588}, 'eval_D': {'precision': 0.9876836325864372, 'recall': 0.9895926256318763, 'f1': 0.988637207575195, 'number': 6726}, 'eval_DT': {'precision': 1.0, 'recall': 0.8, 'f1': 0.888888888888889, 'number': 15}, 'eval_H': {'precision': 0.9487382595903587, 'recall': 0.9305216426193119, 'f1': 0.9395416596626883, 'number': 9010}, 'eval_J': {'precision': 0.9803528468323978, 'recall': 0.980588754311382, 'f1': 0.9804707863816818, 'number': 12467}, 'eval_JR': {'precision': 0.9400386847195358, 'recall': 0.9818181818181818, 'f1': 0.9604743083003953, 'number': 495}, 'eval_JS': {'precision': 0.9612141652613828, 'recall': 0.991304347826087, 'f1': 0.9760273972602741, 'number': 575}, 'eval_N': {'precision': 0.9795543362923471, 'recall': 0.9793769083475651, 'f1': 0.9794656142847902, 'number': 38646}, 'eval_NP': {'precision': 0.9330242966751918, 'recall': 0.9278334128119536, 'f1': 0.9304216147286205, 'number': 6291}, 'eval_NPS': {'precision': 0.75, 'recall': 0.23076923076923078, 'f1': 0.3529411764705882, 'number': 26}, 'eval_NS': {'precision': 0.9691858990616282, 'recall': 0.9773657289002557, 'f1': 0.9732586272762003, 'number': 7820}, 'eval_O': {'precision': 0.9984323288625675, 'recall': 0.999302649930265, 'f1': 0.9988672998170254, 'number': 5736}, 'eval_OS': {'precision': 1.0, 'recall': 0.9952267303102625, 'f1': 0.9976076555023923, 'number': 419}, 'eval_P': {'precision': 0.9887869520897044, 'recall': 0.9918200408997955, 'f1': 0.9903011740684022, 'number': 2934}, 'eval_RB': {'precision': 0.9971910112359551, 'recall': 0.9983929288871033, 'f1': 0.9977916081108211, 'number': 2489}, 'eval_RL': {'precision': 1.0, 'recall': 0.9997228381374723, 'f1': 0.9998613998613999, 'number': 3608}, 'eval_RP': {'precision': 0.9979960600502683, 'recall': 0.9980638586956522, 'f1': 0.9980299582215278, 'number': 29440}, 'eval_RP$': {'precision': 0.9975770162686051, 'recall': 0.9972318339100346, 'f1': 0.9974043952240872, 'number': 5780}, 'eval_SR': {'precision': 0.9998923110058152, 'recall': 0.9998384752059442, 'f1': 0.9998653923812088, 'number': 18573}, 'eval_T': {'precision': 0.9987569919204475, 'recall': 0.9984811874352779, 'f1': 0.9986190706345371, 'number': 28970}, 'eval_W': {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 1}, 'eval_X': {'precision': 0.9466666666666667, 'recall': 0.9594594594594594, 'f1': 0.9530201342281879, 'number': 74}, 'eval_YM': {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 5}, 'eval__': {'precision': 0.9951481772882245, 'recall': 0.9949524745984923, 'f1': 0.9950503163208444, 'number': 15255}, 'eval_`': {'precision': 0.9540229885057471, 'recall': 0.9595375722543352, 'f1': 0.956772334293948, 'number': 173}, 'eval_overall_precision': 0.982771581547336, 'eval_overall_recall': 0.9820371214937511, 'eval_overall_f1': 0.9824042142472381, 'eval_overall_accuracy': 0.9859897481208136, 'eval_runtime': 2090.2466, 'eval_samples_per_second': 15.261, 'eval_steps_per_second': 0.954, 'epoch': 3.0}\n",
      "{'train_runtime': 71133.9787, 'train_samples_per_second': 5.045, 'train_steps_per_second': 0.315, 'train_loss': 0.06107173741493567, 'epoch': 3.0}\n"
     ]
    }
   ],
   "source": [
    "train_results = trainer.train()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Push Model To Hub"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Several commits (2) will be pushed upstream.\n",
      "The progress bars may be unreliable.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "dd7407f7190c4b60a18ff90b1c662e54",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Upload file pytorch_model.bin:   0%|          | 1.00/411M [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ac68929a60ee446e803f52c9a18aaefe",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Upload file runs/Jul02_23-03-13_Brians-Mac-mini/events.out.tfevents.1688356998.Brians-Mac-mini.3901.0:   0%|  …"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "To https://huggingface.co/DunnBC22/bert-base-cased-finetuned-Stromberg_NLP_Twitter-PoS\n",
      "   78c886e..6a8949a  main -> main\n",
      "\n",
      "To https://huggingface.co/DunnBC22/bert-base-cased-finetuned-Stromberg_NLP_Twitter-PoS\n",
      "   6a8949a..db40565  main -> main\n",
      "\n"
     ]
    }
   ],
   "source": [
    "kwargs = {\n",
    "    \"tasks\": \"token-classification\",\n",
    "    \"tags\": ['token-classification'],\n",
    "}\n",
    "\n",
    "if args.push_to_hub:\n",
    "    trainer.push_to_hub('All DUNN!!!', **kwargs)\n",
    "else:\n",
    "    trainer.create_model_card(**kwargs)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Save & Log Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "To https://huggingface.co/DunnBC22/bert-base-cased-finetuned-Stromberg_NLP_Twitter-PoS\n",
      "   db40565..8ee8d42  main -> main\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "***** train metrics *****\n",
      "  epoch                    =         3.0\n",
      "  train_loss               =      0.0611\n",
      "  train_runtime            = 19:45:33.97\n",
      "  train_samples_per_second =       5.045\n",
      "  train_steps_per_second   =       0.315\n"
     ]
    }
   ],
   "source": [
    "trainer.save_model()\n",
    "trainer.log_metrics(\"train\", train_results.metrics)\n",
    "trainer.save_metrics(\"train\", train_results.metrics)\n",
    "trainer.save_state()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Error Analysis"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Define Method to Apply to Validation Dataset (& Then Apply it)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "def forward_pass_with_label(batch):\n",
    "    # Convert dict of lists to list of dicts suitable for data collator\n",
    "    features = [dict(zip(batch, t)) for t in zip(*batch.values())]\n",
    "    # Pad inputs and labels and put all tensors on device\n",
    "    batch = data_collator(features)\n",
    "    input_ids = batch[\"input_ids\"].to(DEVICE)\n",
    "    attention_mask = batch[\"attention_mask\"].to(DEVICE)\n",
    "    labels = batch[\"labels\"].to(DEVICE)\n",
    "    with torch.no_grad():\n",
    "        # Pass data through model  \n",
    "        output = trainer.model(input_ids, \n",
    "                               attention_mask\n",
    "                               )\n",
    "        # Logit.size: [batch_size, sequence_length, classes]\n",
    "        predicted_label = torch.argmax(output.logits, \n",
    "                                       axis=-1\n",
    "                                       ).cpu().numpy()\n",
    "        \n",
    "    # Calculate loss per token after flattening batch dimension with view\n",
    "    loss = cross_entropy(output.logits.view(-1, 51), \n",
    "                         labels.view(-1), \n",
    "                         reduction=\"none\")\n",
    "    # Unflatten batch dimension and convert to numpy array\n",
    "    loss = loss.view(len(input_ids), -1).cpu().numpy()\n",
    "\n",
    "    return {\"loss\":loss, \"predicted_label\": predicted_label}"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Apply Above Function to Entire Evaluation Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Parameter 'function'=<function forward_pass_with_label at 0x7fca841d1790> of the transform datasets.arrow_dataset.Dataset._map_single couldn't be hashed properly, a random hash was used instead. Make sure your transforms and parameters are serializable with pickle or dill for the dataset fingerprinting and caching to work. If you reuse this transform, the caching mechanism will consider it to be different from the previous calls and recompute everything. This warning is only showed once. Subsequent hashing failures won't be showed.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e7ed84e43e3c4e8295aa3b49ee368310",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/7974 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "eval_set = encoded_ds['eval']\n",
    "\n",
    "eval_set = eval_set.remove_columns('id')\n",
    "\n",
    "eval_set = eval_set.map(forward_pass_with_label,\n",
    "                        batched=True,\n",
    "                        batch_size=32)\n",
    "\n",
    "eval_df = eval_set.to_pandas()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Clean Up Padding Tokens"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>input_ids</th>\n",
       "      <th>token_type_ids</th>\n",
       "      <th>attention_mask</th>\n",
       "      <th>labels</th>\n",
       "      <th>loss</th>\n",
       "      <th>predicted_label</th>\n",
       "      <th>input_tokens</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>[101, 137, 1337, 2428, 2386, 1181, 2924, 7956,...</td>\n",
       "      <td>[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]</td>\n",
       "      <td>[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]</td>\n",
       "      <td>[IGN, USR, IGN, IGN, IGN, IGN, IGN, IGN, MD, V...</td>\n",
       "      <td>[0.0, 1.5497195e-06, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "      <td>[., USR, NNP, NNP, UH, UH, PRP, PRP, MD, VB, V...</td>\n",
       "      <td>[[CLS], @, That, ##K, ##id, ##d, ##W, ##ill, s...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                           input_ids   \n",
       "0  [101, 137, 1337, 2428, 2386, 1181, 2924, 7956,...  \\\n",
       "\n",
       "                                     token_type_ids   \n",
       "0  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]  \\\n",
       "\n",
       "                                     attention_mask   \n",
       "0  [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]  \\\n",
       "\n",
       "                                              labels   \n",
       "0  [IGN, USR, IGN, IGN, IGN, IGN, IGN, IGN, MD, V...  \\\n",
       "\n",
       "                                                loss   \n",
       "0  [0.0, 1.5497195e-06, 0.0, 0.0, 0.0, 0.0, 0.0, ...  \\\n",
       "\n",
       "                                     predicted_label   \n",
       "0  [., USR, NNP, NNP, UH, UH, PRP, PRP, MD, VB, V...  \\\n",
       "\n",
       "                                        input_tokens  \n",
       "0  [[CLS], @, That, ##K, ##id, ##d, ##W, ##ill, s...  "
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "id2label[-100] = \"IGN\"\n",
    "eval_df[\"input_tokens\"] = eval_df[\"input_ids\"].apply(\n",
    "    lambda x: tokenizer.convert_ids_to_tokens(x))\n",
    "eval_df[\"predicted_label\"] = eval_df[\"predicted_label\"].apply(\n",
    "    lambda x: [id2label[i] for i in x])\n",
    "eval_df[\"labels\"] = eval_df[\"labels\"].apply(\n",
    "    lambda x: [id2label[i] for i in x])\n",
    "eval_df['loss'] = eval_df.apply(\n",
    "    lambda x: x['loss'][:len(x['input_ids'])], axis=1)\n",
    "eval_df['predicted_label'] = eval_df.apply(\n",
    "    lambda x: x['predicted_label'][:len(x['input_ids'])], axis=1)\n",
    "eval_df.head(1)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Unwrap Each Token Within Sample Separately"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>input_ids</th>\n",
       "      <th>token_type_ids</th>\n",
       "      <th>attention_mask</th>\n",
       "      <th>labels</th>\n",
       "      <th>loss</th>\n",
       "      <th>predicted_label</th>\n",
       "      <th>input_tokens</th>\n",
       "      <th>total_loss</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>137</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>USR</td>\n",
       "      <td>0.0</td>\n",
       "      <td>USR</td>\n",
       "      <td>@</td>\n",
       "      <td>0.003804</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1431</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>MD</td>\n",
       "      <td>0.0</td>\n",
       "      <td>MD</td>\n",
       "      <td>should</td>\n",
       "      <td>0.003804</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>189</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>VB</td>\n",
       "      <td>0.0</td>\n",
       "      <td>VB</td>\n",
       "      <td>t</td>\n",
       "      <td>0.003804</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1143</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>PRP</td>\n",
       "      <td>0.0</td>\n",
       "      <td>PRP</td>\n",
       "      <td>me</td>\n",
       "      <td>0.003804</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1171</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>RB</td>\n",
       "      <td>0.0</td>\n",
       "      <td>RB</td>\n",
       "      <td>back</td>\n",
       "      <td>0.003804</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>131</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>.</td>\n",
       "      <td>0.0</td>\n",
       "      <td>.</td>\n",
       "      <td>:</td>\n",
       "      <td>0.003804</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>114</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>)</td>\n",
       "      <td>0.0</td>\n",
       "      <td>)</td>\n",
       "      <td>)</td>\n",
       "      <td>0.003804</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  input_ids token_type_ids attention_mask labels  loss predicted_label   \n",
       "0       137              0              1    USR   0.0             USR  \\\n",
       "0      1431              0              1     MD   0.0              MD   \n",
       "0       189              0              1     VB   0.0              VB   \n",
       "0      1143              0              1    PRP   0.0             PRP   \n",
       "0      1171              0              1     RB   0.0              RB   \n",
       "0       131              0              1      .   0.0               .   \n",
       "0       114              0              1      )   0.0               )   \n",
       "\n",
       "  input_tokens  total_loss  \n",
       "0            @    0.003804  \n",
       "0       should    0.003804  \n",
       "0            t    0.003804  \n",
       "0           me    0.003804  \n",
       "0         back    0.003804  \n",
       "0            :    0.003804  \n",
       "0            )    0.003804  "
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "eval_df_tokens = eval_df.apply(pd.Series.explode)\n",
    "eval_df_tokens = eval_df_tokens.query(\"labels != 'IGN'\")\n",
    "eval_df_tokens[\"loss\"] = eval_df_tokens[\"loss\"].astype(float).round(2)\n",
    "eval_df_tokens.head(7)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### See Which Tokens Have Accumulated Most Loss in Evaluation Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>input_tokens</th>\n",
       "      <td>!</td>\n",
       "      <td>.</td>\n",
       "      <td>[UNK]</td>\n",
       "      <td>B</td>\n",
       "      <td>want</td>\n",
       "      <td>?</td>\n",
       "      <td>have</td>\n",
       "      <td>:</td>\n",
       "      <td>that</td>\n",
       "      <td>G</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1979</td>\n",
       "      <td>3500</td>\n",
       "      <td>107</td>\n",
       "      <td>96</td>\n",
       "      <td>181</td>\n",
       "      <td>822</td>\n",
       "      <td>358</td>\n",
       "      <td>2416</td>\n",
       "      <td>217</td>\n",
       "      <td>43</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.083</td>\n",
       "      <td>0.034</td>\n",
       "      <td>0.573</td>\n",
       "      <td>0.46</td>\n",
       "      <td>0.221</td>\n",
       "      <td>0.044</td>\n",
       "      <td>0.097</td>\n",
       "      <td>0.014</td>\n",
       "      <td>0.149</td>\n",
       "      <td>0.746</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sum</th>\n",
       "      <td>164.3</td>\n",
       "      <td>119.09</td>\n",
       "      <td>61.36</td>\n",
       "      <td>44.12</td>\n",
       "      <td>40.07</td>\n",
       "      <td>36.19</td>\n",
       "      <td>34.76</td>\n",
       "      <td>32.63</td>\n",
       "      <td>32.28</td>\n",
       "      <td>32.08</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  0       1      2      3      4      5      6      7      8   \n",
       "input_tokens      !       .  [UNK]      B   want      ?   have      :   that  \\\n",
       "count          1979    3500    107     96    181    822    358   2416    217   \n",
       "mean          0.083   0.034  0.573   0.46  0.221  0.044  0.097  0.014  0.149   \n",
       "sum           164.3  119.09  61.36  44.12  40.07  36.19  34.76  32.63  32.28   \n",
       "\n",
       "                  9  \n",
       "input_tokens      G  \n",
       "count            43  \n",
       "mean          0.746  \n",
       "sum           32.08  "
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    eval_df_tokens.groupby(\"input_tokens\")[[\"loss\"]]\n",
    "    .agg([\"count\", \"mean\", \"sum\"])\n",
    "    .droplevel(level=0, axis=1)  # Get rid of multi-level columns\n",
    "    .sort_values(by=\"sum\", ascending=False)\n",
    "    .reset_index()\n",
    "    .round(3)\n",
    "    .head(10)\n",
    "    .T\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### See Which Label IDs Have Most Loss in Evaluation Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>33</th>\n",
       "      <th>34</th>\n",
       "      <th>35</th>\n",
       "      <th>36</th>\n",
       "      <th>37</th>\n",
       "      <th>38</th>\n",
       "      <th>39</th>\n",
       "      <th>40</th>\n",
       "      <th>41</th>\n",
       "      <th>42</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>labels</th>\n",
       "      <td>SYM</td>\n",
       "      <td>NNPS</td>\n",
       "      <td>WDT</td>\n",
       "      <td>VBN</td>\n",
       "      <td>''</td>\n",
       "      <td>RBR</td>\n",
       "      <td>JJR</td>\n",
       "      <td>``</td>\n",
       "      <td>NNP</td>\n",
       "      <td>)</td>\n",
       "      <td>...</td>\n",
       "      <td>RT</td>\n",
       "      <td>TO</td>\n",
       "      <td>(</td>\n",
       "      <td>WRB</td>\n",
       "      <td>WP</td>\n",
       "      <td>,</td>\n",
       "      <td>USR</td>\n",
       "      <td>HT</td>\n",
       "      <td>URL</td>\n",
       "      <td>POS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>4</td>\n",
       "      <td>549</td>\n",
       "      <td>75</td>\n",
       "      <td>28</td>\n",
       "      <td>140</td>\n",
       "      <td>53</td>\n",
       "      <td>2194</td>\n",
       "      <td>109</td>\n",
       "      <td>...</td>\n",
       "      <td>2047</td>\n",
       "      <td>1470</td>\n",
       "      <td>49</td>\n",
       "      <td>624</td>\n",
       "      <td>383</td>\n",
       "      <td>1076</td>\n",
       "      <td>5175</td>\n",
       "      <td>2011</td>\n",
       "      <td>914</td>\n",
       "      <td>108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>6.53</td>\n",
       "      <td>3.69</td>\n",
       "      <td>3.075</td>\n",
       "      <td>0.288</td>\n",
       "      <td>0.284</td>\n",
       "      <td>0.269</td>\n",
       "      <td>0.209</td>\n",
       "      <td>0.204</td>\n",
       "      <td>0.199</td>\n",
       "      <td>0.139</td>\n",
       "      <td>...</td>\n",
       "      <td>0.005</td>\n",
       "      <td>0.005</td>\n",
       "      <td>0.002</td>\n",
       "      <td>0.002</td>\n",
       "      <td>0.002</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sum</th>\n",
       "      <td>13.06</td>\n",
       "      <td>22.14</td>\n",
       "      <td>12.3</td>\n",
       "      <td>158.29</td>\n",
       "      <td>21.29</td>\n",
       "      <td>7.53</td>\n",
       "      <td>29.19</td>\n",
       "      <td>10.81</td>\n",
       "      <td>437.27</td>\n",
       "      <td>15.14</td>\n",
       "      <td>...</td>\n",
       "      <td>11.19</td>\n",
       "      <td>7.32</td>\n",
       "      <td>0.12</td>\n",
       "      <td>1.48</td>\n",
       "      <td>0.71</td>\n",
       "      <td>0.23</td>\n",
       "      <td>0.34</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4 rows × 43 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           0      1      2       3      4      5      6      7       8    \n",
       "labels    SYM   NNPS    WDT     VBN     ''    RBR    JJR     ``     NNP  \\\n",
       "count       2      6      4     549     75     28    140     53    2194   \n",
       "mean     6.53   3.69  3.075   0.288  0.284  0.269  0.209  0.204   0.199   \n",
       "sum     13.06  22.14   12.3  158.29  21.29   7.53  29.19  10.81  437.27   \n",
       "\n",
       "           9   ...     33     34     35     36     37    38    39    40   41   \n",
       "labels      )  ...     RT     TO      (    WRB     WP     ,   USR    HT  URL  \\\n",
       "count     109  ...   2047   1470     49    624    383  1076  5175  2011  914   \n",
       "mean    0.139  ...  0.005  0.005  0.002  0.002  0.002   0.0   0.0   0.0  0.0   \n",
       "sum     15.14  ...  11.19   7.32   0.12   1.48   0.71  0.23  0.34  0.04  0.0   \n",
       "\n",
       "         42  \n",
       "labels  POS  \n",
       "count   108  \n",
       "mean    0.0  \n",
       "sum     0.0  \n",
       "\n",
       "[4 rows x 43 columns]"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    eval_df_tokens.groupby(\"labels\")[[\"loss\"]] \n",
    "    .agg([\"count\", \"mean\", \"sum\"])\n",
    "    .droplevel(level=0, axis=1)\n",
    "    .sort_values(by=\"mean\", ascending=False)\n",
    "    .reset_index()\n",
    "    .round(3)\n",
    "    .fillna(0)\n",
    "    .T\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Create Function to Display Confusion Matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay\n",
    "\n",
    "def plot_confusion_matrix(y_preds, y_true, labels):\n",
    "    cm = confusion_matrix(y_true, y_preds, normalize=\"true\")\n",
    "    fig, ax = plt.subplots(figsize=(30, 30))\n",
    "    disp = ConfusionMatrixDisplay(confusion_matrix=cm, display_labels=labels)\n",
    "    disp.plot(cmap=\"Blues\", values_format=\".2f\", ax=ax, colorbar=False)\n",
    "    plt.title(\"Normalized confusion matrix\")\n",
    "    plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Display Confusion Matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 3000x3000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "eval_token_list = list(set(eval_df_tokens['labels']))\n",
    "\n",
    "plot_confusion_matrix(eval_df_tokens[\"labels\"], eval_df_tokens[\"predicted_label\"],\n",
    "                      eval_token_list)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Define Function to Display Example Token Sequences Along With Labels & Losses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>11</th>\n",
       "      <th>12</th>\n",
       "      <th>13</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>tokens</th>\n",
       "      <td>Go</td>\n",
       "      <td>cardinal</td>\n",
       "      <td>##s</td>\n",
       "      <td>!</td>\n",
       "      <td>!</td>\n",
       "      <td>!</td>\n",
       "      <td>!</td>\n",
       "      <td>#</td>\n",
       "      <td>game</td>\n",
       "      <td>##2</td>\n",
       "      <td>#</td>\n",
       "      <td>NL</td>\n",
       "      <td>##CS</td>\n",
       "      <td>[SEP]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>labels</th>\n",
       "      <td>VB</td>\n",
       "      <td>NNPS</td>\n",
       "      <td>IGN</td>\n",
       "      <td>UH</td>\n",
       "      <td>.</td>\n",
       "      <td>UH</td>\n",
       "      <td>.</td>\n",
       "      <td>HT</td>\n",
       "      <td>IGN</td>\n",
       "      <td>IGN</td>\n",
       "      <td>HT</td>\n",
       "      <td>IGN</td>\n",
       "      <td>IGN</td>\n",
       "      <td>IGN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>preds</th>\n",
       "      <td>VB</td>\n",
       "      <td>NNS</td>\n",
       "      <td>NNS</td>\n",
       "      <td>.</td>\n",
       "      <td>UH</td>\n",
       "      <td>.</td>\n",
       "      <td>.</td>\n",
       "      <td>HT</td>\n",
       "      <td>HT</td>\n",
       "      <td>HT</td>\n",
       "      <td>HT</td>\n",
       "      <td>NNP</td>\n",
       "      <td>HT</td>\n",
       "      <td>.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>losses</th>\n",
       "      <td>0.03</td>\n",
       "      <td>9.29</td>\n",
       "      <td>0.00</td>\n",
       "      <td>8.61</td>\n",
       "      <td>4.15</td>\n",
       "      <td>5.25</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1     2     3     4     5     6     7     8     9     10   \n",
       "tokens    Go  cardinal   ##s     !     !     !     !     #  game   ##2     #  \\\n",
       "labels    VB      NNPS   IGN    UH     .    UH     .    HT   IGN   IGN    HT   \n",
       "preds     VB       NNS   NNS     .    UH     .     .    HT    HT    HT    HT   \n",
       "losses  0.03      9.29  0.00  8.61  4.15  5.25  0.00  0.00  0.00  0.00  0.00   \n",
       "\n",
       "          11    12     13  \n",
       "tokens    NL  ##CS  [SEP]  \n",
       "labels   IGN   IGN    IGN  \n",
       "preds    NNP    HT      .  \n",
       "losses  0.00  0.00   0.00  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>42</th>\n",
       "      <th>43</th>\n",
       "      <th>44</th>\n",
       "      <th>45</th>\n",
       "      <th>46</th>\n",
       "      <th>47</th>\n",
       "      <th>48</th>\n",
       "      <th>49</th>\n",
       "      <th>50</th>\n",
       "      <th>51</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>tokens</th>\n",
       "      <td>#</td>\n",
       "      <td>W</td>\n",
       "      <td>##EL</td>\n",
       "      <td>##O</td>\n",
       "      <td>##VE</td>\n",
       "      <td>##IF</td>\n",
       "      <td>##Y</td>\n",
       "      <td>##MU</td>\n",
       "      <td>##SI</td>\n",
       "      <td>##C</td>\n",
       "      <td>...</td>\n",
       "      <td>##Y</td>\n",
       "      <td>##MU</td>\n",
       "      <td>##SI</td>\n",
       "      <td>##C</td>\n",
       "      <td>##VI</td>\n",
       "      <td>##DE</td>\n",
       "      <td>##O</td>\n",
       "      <td>!</td>\n",
       "      <td>!</td>\n",
       "      <td>[SEP]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>labels</th>\n",
       "      <td>HT</td>\n",
       "      <td>IGN</td>\n",
       "      <td>IGN</td>\n",
       "      <td>IGN</td>\n",
       "      <td>IGN</td>\n",
       "      <td>IGN</td>\n",
       "      <td>IGN</td>\n",
       "      <td>IGN</td>\n",
       "      <td>IGN</td>\n",
       "      <td>IGN</td>\n",
       "      <td>...</td>\n",
       "      <td>IGN</td>\n",
       "      <td>IGN</td>\n",
       "      <td>IGN</td>\n",
       "      <td>IGN</td>\n",
       "      <td>IGN</td>\n",
       "      <td>IGN</td>\n",
       "      <td>IGN</td>\n",
       "      <td>.</td>\n",
       "      <td>.</td>\n",
       "      <td>IGN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>preds</th>\n",
       "      <td>HT</td>\n",
       "      <td>NNP</td>\n",
       "      <td>NNP</td>\n",
       "      <td>NNP</td>\n",
       "      <td>NNP</td>\n",
       "      <td>NN</td>\n",
       "      <td>NN</td>\n",
       "      <td>NN</td>\n",
       "      <td>NN</td>\n",
       "      <td>NN</td>\n",
       "      <td>...</td>\n",
       "      <td>NN</td>\n",
       "      <td>NN</td>\n",
       "      <td>NN</td>\n",
       "      <td>NN</td>\n",
       "      <td>NN</td>\n",
       "      <td>NN</td>\n",
       "      <td>NN</td>\n",
       "      <td>.</td>\n",
       "      <td>.</td>\n",
       "      <td>.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>losses</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4 rows × 52 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          0     1     2     3     4     5     6     7     8     9   ...    42   \n",
       "tokens     #     W  ##EL   ##O  ##VE  ##IF   ##Y  ##MU  ##SI   ##C  ...   ##Y  \\\n",
       "labels    HT   IGN   IGN   IGN   IGN   IGN   IGN   IGN   IGN   IGN  ...   IGN   \n",
       "preds     HT   NNP   NNP   NNP   NNP    NN    NN    NN    NN    NN  ...    NN   \n",
       "losses  0.00  0.00  0.00  0.00  0.00  0.00  0.00  0.00  0.00  0.00  ...  0.00   \n",
       "\n",
       "          43    44    45    46    47    48    49    50     51  \n",
       "tokens  ##MU  ##SI   ##C  ##VI  ##DE   ##O     !     !  [SEP]  \n",
       "labels   IGN   IGN   IGN   IGN   IGN   IGN     .     .    IGN  \n",
       "preds     NN    NN    NN    NN    NN    NN     .     .      .  \n",
       "losses  0.00  0.00  0.00  0.00  0.00  0.00  0.00  0.00   0.00  \n",
       "\n",
       "[4 rows x 52 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>25</th>\n",
       "      <th>26</th>\n",
       "      <th>27</th>\n",
       "      <th>28</th>\n",
       "      <th>29</th>\n",
       "      <th>30</th>\n",
       "      <th>31</th>\n",
       "      <th>32</th>\n",
       "      <th>33</th>\n",
       "      <th>34</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>tokens</th>\n",
       "      <td>The</td>\n",
       "      <td>comments</td>\n",
       "      <td>!</td>\n",
       "      <td>!</td>\n",
       "      <td>!</td>\n",
       "      <td>!</td>\n",
       "      <td>!</td>\n",
       "      <td>!</td>\n",
       "      <td>O</td>\n",
       "      <td>##m</td>\n",
       "      <td>...</td>\n",
       "      <td>my</td>\n",
       "      <td>heart</td>\n",
       "      <td>is</td>\n",
       "      <td>going</td>\n",
       "      <td>to</td>\n",
       "      <td>jump</td>\n",
       "      <td>and</td>\n",
       "      <td>run</td>\n",
       "      <td>away</td>\n",
       "      <td>[SEP]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>labels</th>\n",
       "      <td>DT</td>\n",
       "      <td>NNS</td>\n",
       "      <td>.</td>\n",
       "      <td>.</td>\n",
       "      <td>UH</td>\n",
       "      <td>.</td>\n",
       "      <td>UH</td>\n",
       "      <td>.</td>\n",
       "      <td>UH</td>\n",
       "      <td>IGN</td>\n",
       "      <td>...</td>\n",
       "      <td>PRP$</td>\n",
       "      <td>NN</td>\n",
       "      <td>VBZ</td>\n",
       "      <td>VBG</td>\n",
       "      <td>TO</td>\n",
       "      <td>VB</td>\n",
       "      <td>CC</td>\n",
       "      <td>VB</td>\n",
       "      <td>RB</td>\n",
       "      <td>IGN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>preds</th>\n",
       "      <td>DT</td>\n",
       "      <td>NNS</td>\n",
       "      <td>.</td>\n",
       "      <td>UH</td>\n",
       "      <td>.</td>\n",
       "      <td>UH</td>\n",
       "      <td>.</td>\n",
       "      <td>.</td>\n",
       "      <td>UH</td>\n",
       "      <td>UH</td>\n",
       "      <td>...</td>\n",
       "      <td>PRP$</td>\n",
       "      <td>NN</td>\n",
       "      <td>VBZ</td>\n",
       "      <td>VBG</td>\n",
       "      <td>TO</td>\n",
       "      <td>VB</td>\n",
       "      <td>CC</td>\n",
       "      <td>VB</td>\n",
       "      <td>RB</td>\n",
       "      <td>.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>losses</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>5.94</td>\n",
       "      <td>5.39</td>\n",
       "      <td>2.34</td>\n",
       "      <td>3.79</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4 rows × 35 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1     2     3     4     5     6     7     8     9   ...   \n",
       "tokens   The  comments     !     !     !     !     !     !     O   ##m  ...  \\\n",
       "labels    DT       NNS     .     .    UH     .    UH     .    UH   IGN  ...   \n",
       "preds     DT       NNS     .    UH     .    UH     .     .    UH    UH  ...   \n",
       "losses  0.00      0.00  0.00  5.94  5.39  2.34  3.79  0.00  0.00  0.00  ...   \n",
       "\n",
       "          25     26    27     28    29    30    31    32    33     34  \n",
       "tokens    my  heart    is  going    to  jump   and   run  away  [SEP]  \n",
       "labels  PRP$     NN   VBZ    VBG    TO    VB    CC    VB    RB    IGN  \n",
       "preds   PRP$     NN   VBZ    VBG    TO    VB    CC    VB    RB      .  \n",
       "losses  0.00   0.00  0.00   0.00  0.00  0.00  0.00  0.00  0.00   0.00  \n",
       "\n",
       "[4 rows x 35 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def get_samples(df):\n",
    "    for _, row in df.iterrows():\n",
    "        labels, preds, tokens, losses = [], [], [], []\n",
    "        for i, mask in enumerate(row[\"attention_mask\"]):\n",
    "            if i not in {0, len(row[\"attention_mask\"])}:\n",
    "                labels.append(row[\"labels\"][i])\n",
    "                preds.append(row[\"predicted_label\"][i])\n",
    "                tokens.append(row[\"input_tokens\"][i])\n",
    "                losses.append(f\"{row['loss'][i]:.2f}\")\n",
    "        eval_df_tmp = pd.DataFrame({\"tokens\": tokens, \"labels\": labels, \n",
    "                               \"preds\": preds, \"losses\": losses}).T\n",
    "        yield eval_df_tmp\n",
    "\n",
    "eval_df[\"total_loss\"] = eval_df[\"loss\"].apply(sum)\n",
    "eval_df_tmp = eval_df.sort_values(by=\"total_loss\", ascending=False).head(3)\n",
    "\n",
    "for sample in get_samples(eval_df_tmp):\n",
    "    display(sample)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Notes & Other Takeaways From This Project\n",
    "\n",
    "****\n",
    "- While I did not learn that I should use poseval instead of seqeval for evaluating this model until after the first epoch completed, the project still turned out really well!\n",
    "\n",
    "- Once I complete the list of token classification projects that I have scripted/developed, I will come back to this and run it (maybe with only 1 or two epochs) with the poseval evaluator.\n",
    "\n",
    "- Normally, I like to include some nice visualizations, which I (unfortunately) did not do with this project. Because I wanted to focus on the main part of the project (for this project and many others), I forewent the visualizations. I intend on completed some as an acillary project after completing the currently list of projects that I have.\n",
    "\n",
    "****"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Citations\n",
    "\n",
    "- Model Checkpoint\n",
    "\n",
    "    > @article{DBLP:journals/corr/abs-1810-04805, author = {Jacob Devlin and Ming{-}Wei Chang and Kenton Lee and Kristina Toutanova}, title = {{BERT:} Pre-training of Deep Bidirectional Transformers for Language Understanding}, journal = {CoRR}, volume = {abs/1810.04805}, year = {2018}, url = {http://arxiv.org/abs/1810.04805}, archivePrefix = {arXiv}, eprint = {1810.04805}, timestamp = {Tue, 30 Oct 2018 20:39:56 +0100}, biburl = {https://dblp.org/rec/journals/corr/abs-1810-04805.bib}, bibsource = {dblp computer science bibliography, https://dblp.org}}\n",
    "\n",
    "- Dataset\n",
    "\n",
    "    > @inproceedings{derczynski2013twitter, title={Twitter part-of-speech tagging for all: Overcoming sparse and noisy data}, author={Derczynski, Leon and Ritter, Alan and Clark, Sam and Bontcheva, Kalina}, booktitle={Proceedings of the international conference recent advances in natural language ranlp 2013}, pages={198--206}, year={2013}}\n",
    "\n",
    "- Metric (SeqEval)\n",
    "\n",
    "    > @inproceedings{ramshaw-marcus-1995-text, title = \"Text Chunking using Transformation-Based Learning\", author = \"Ramshaw, Lance and Marcus, Mitch\", booktitle = \"Third Workshop on Very Large Corpora\", year = \"1995\", url = \"https://www.aclweb.org/anthology/W95-0107\",}\n",
    "    \n",
    "    > @misc{seqeval, title={{seqeval}: A Python framework for sequence labeling evaluation}, url={https://github.com/chakki-works/seqeval}, note={Software available from https://github.com/chakki-works/seqeval}, author={Hiroki Nakayama}, year={2018},}"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "dl",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.12"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}